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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
e1877bec-add9-4223-acd2-ff15509772d3 | how-to-control-hydrodynamic-force-on-fluidic | 2304.11526 | null | https://arxiv.org/abs/2304.11526v1 | https://arxiv.org/pdf/2304.11526v1.pdf | How to Control Hydrodynamic Force on Fluidic Pinball via Deep Reinforcement Learning | Deep reinforcement learning (DRL) for fluidic pinball, three individually rotating cylinders in the uniform flow arranged in an equilaterally triangular configuration, can learn the efficient flow control strategies due to the validity of self-learning and data-driven state estimation for complex fluid dynamic problems. In this work, we present a DRL-based real-time feedback strategy to control the hydrodynamic force on fluidic pinball, i.e., force extremum and tracking, from cylinders' rotation. By adequately designing reward functions and encoding historical observations, and after automatic learning of thousands of iterations, the DRL-based control was shown to make reasonable and valid control decisions in nonparametric control parameter space, which is comparable to and even better than the optimal policy found through lengthy brute-force searching. Subsequently, one of these results was analyzed by a machine learning model that enabled us to shed light on the basis of decision-making and physical mechanisms of the force tracking process. The finding from this work can control hydrodynamic force on the operation of fluidic pinball system and potentially pave the way for exploring efficient active flow control strategies in other complex fluid dynamic problems. | ['Dixia Fan', 'Zhiyang Jin', 'Hui Xiang', 'Yue Wang', 'Haodong Feng'] | 2023-04-23 | null | null | null | null | ['self-learning'] | ['natural-language-processing'] | [-4.38724667e-01 -1.46921009e-01 -1.46410570e-01 3.43988746e-01
-1.09662481e-01 -5.43163300e-01 4.82021481e-01 1.50639862e-01
-5.05639911e-01 1.17131996e+00 -1.99812949e-01 -2.74496794e-01
-5.60089767e-01 -6.36407256e-01 -7.85351574e-01 -1.11241269e+00
-5.56900740e-01 5.19152582e-01 2.15039268e-01 -4.14616227e-01
6.29230142e-01 7.00086474e-01 -1.43388140e+00 -3.51730973e-01
6.47266150e-01 8.79613400e-01 2.83280730e-01 9.68775153e-01
1.63628533e-01 9.36762393e-01 -3.75181526e-01 9.24780443e-02
5.00634909e-01 -2.56079525e-01 -5.91105938e-01 -9.98512357e-02
-2.02153891e-01 -4.83961135e-01 -3.49277318e-01 3.76463920e-01
4.61425006e-01 7.11160660e-01 8.19481790e-01 -7.79426992e-01
-2.40479752e-01 2.61179268e-01 -2.95734435e-01 3.80528361e-01
2.50298604e-02 6.85386002e-01 7.23334491e-01 -3.78090233e-01
4.36495006e-01 1.15783501e+00 4.10286963e-01 5.61599672e-01
-7.65249789e-01 -4.67482477e-01 -8.90712142e-02 -2.14710757e-02
-8.21691096e-01 -1.22695096e-01 4.23888505e-01 -8.28520179e-01
6.37289762e-01 1.14760101e-01 1.30445504e+00 6.68978930e-01
7.19126403e-01 5.89778781e-01 8.09461176e-01 -4.21823680e-01
4.93069023e-01 5.17073367e-03 -3.58623743e-01 7.79719174e-01
5.01026034e-01 7.87724555e-01 -2.28954986e-01 -1.02288149e-01
1.22109842e+00 -2.95532525e-01 -1.89889684e-01 -5.93555331e-01
-7.71529436e-01 8.93195510e-01 2.27263808e-01 -1.42565100e-02
-2.71446288e-01 2.61518776e-01 5.56641936e-01 2.78615803e-01
-1.76293440e-02 1.44057846e+00 -6.77166224e-01 -4.88370866e-01
-5.63973427e-01 6.96177125e-01 9.49768186e-01 2.96960115e-01
3.66638750e-01 5.47147691e-01 -1.81280270e-01 4.10500616e-01
1.22583494e-01 7.53737152e-01 4.93516028e-01 -1.25751173e+00
2.17214093e-01 2.22301945e-01 6.46751344e-01 -7.96175182e-01
-5.18412948e-01 -2.88404614e-01 -5.51353753e-01 4.71303731e-01
7.90015399e-01 -8.59758556e-01 -4.34596419e-01 1.27214837e+00
3.50753218e-01 3.60622145e-02 5.40525243e-02 1.09964216e+00
4.47840476e-03 7.56716609e-01 -1.57109514e-01 -4.46842879e-01
9.66549337e-01 -6.60130382e-01 -5.34982502e-01 2.46269032e-01
4.83699977e-01 -3.84052873e-01 9.43893611e-01 4.43398088e-01
-1.19111300e+00 -6.10307753e-01 -1.09171641e+00 6.05714560e-01
-1.71269551e-02 -8.47677812e-02 8.00776362e-01 2.93293118e-01
-5.97920835e-01 1.13619411e+00 -1.13335085e+00 7.42017254e-02
2.03040652e-02 2.18367592e-01 1.28809929e-01 6.58955395e-01
-1.35129571e+00 1.03981924e+00 3.47621500e-01 1.14535764e-01
-1.11455309e+00 -8.43848109e-01 -7.27996647e-01 1.63441226e-02
6.01555228e-01 -5.44979334e-01 1.55018079e+00 -3.87631357e-01
-2.08002448e+00 1.43535465e-01 2.31037766e-01 -7.88435638e-01
8.10692310e-01 -3.48178267e-01 1.45654574e-01 3.41668248e-01
-1.09582469e-02 4.59711939e-01 8.47495973e-01 -1.10483217e+00
-4.11230624e-01 1.35748610e-01 -1.38525903e-01 3.41294646e-01
-1.55906484e-01 -5.46378076e-01 3.30496013e-01 -3.20167452e-01
-3.75960439e-01 -1.07484472e+00 -5.06105781e-01 -4.80909869e-02
5.87670617e-02 -3.34136695e-01 6.28842115e-01 -2.63034195e-01
1.13513100e+00 -1.51839054e+00 1.50339752e-01 2.22947553e-01
-9.98289585e-02 6.29609585e-01 3.63817900e-01 7.66095996e-01
1.28474355e-01 -3.07699554e-02 6.97213858e-02 5.05658627e-01
-2.74000853e-01 1.42294437e-01 -6.24112844e-01 4.12929326e-01
2.36790106e-01 8.57808232e-01 -1.04077494e+00 -2.82901734e-01
4.60758984e-01 -7.84188416e-03 -5.38697362e-01 4.76621926e-01
-3.57092112e-01 7.90359795e-01 -7.87263215e-01 1.56040326e-01
3.23948890e-01 -1.24125786e-01 1.66417152e-01 1.41238064e-01
-6.78867340e-01 -2.15175867e-01 -9.30456817e-01 8.72369945e-01
-6.92128181e-01 6.16210699e-01 1.61690325e-01 -1.09291196e+00
1.23851478e+00 2.59123713e-01 9.35625315e-01 -8.47407997e-01
3.37607086e-01 -5.52849378e-03 1.97097287e-01 -9.64425087e-01
4.64676946e-01 -3.06986064e-01 7.35000372e-02 3.68006229e-01
-2.51895815e-01 -7.05202162e-01 5.14713526e-01 -1.63937569e-01
8.59689593e-01 8.39586928e-02 6.43822625e-02 -6.52912140e-01
7.08690405e-01 1.82542130e-02 3.56128842e-01 8.50455165e-01
-2.37032235e-01 -1.60069600e-01 6.52401567e-01 -8.14324021e-01
-1.40831268e+00 -7.35385716e-01 -2.93019533e-01 6.93475068e-01
4.04666662e-01 -7.11525455e-02 -4.20789421e-01 1.74492057e-02
5.80543458e-01 4.31466728e-01 -5.57389736e-01 -2.95145422e-01
-9.95529711e-01 -6.28374457e-01 -1.31759614e-01 5.03184497e-01
4.54767168e-01 -1.15919793e+00 -1.06438351e+00 4.86011207e-01
5.68378270e-01 -8.97989869e-01 -3.83660048e-02 1.36714548e-01
-7.14711785e-01 -1.46885622e+00 -6.09364867e-01 -2.75651902e-01
2.54084975e-01 -1.73735678e-01 7.43980050e-01 1.76939130e-01
-4.58951205e-01 3.96363229e-01 -1.65097877e-01 -2.53023624e-01
-4.29074019e-01 6.62844181e-02 2.31765017e-01 -3.93362582e-01
-3.62028837e-01 -3.59429002e-01 -7.87419736e-01 6.34641886e-01
-4.33099598e-01 -5.73530160e-02 4.17012155e-01 9.48242068e-01
3.08620930e-01 6.82781711e-02 8.62267911e-01 -4.97401536e-01
9.71906185e-01 -3.22872341e-01 -1.22336221e+00 3.72759514e-02
-5.67043006e-01 4.84520286e-01 1.17386580e+00 -5.04073322e-01
-9.87644851e-01 -5.87660745e-02 3.48424166e-02 -5.15772760e-01
3.77064675e-01 -1.15570212e-02 5.29613018e-01 -2.32041478e-01
6.28651440e-01 2.51165926e-01 3.21117193e-01 -1.07784711e-01
1.04510210e-01 3.67277503e-01 1.43160269e-01 -1.17496276e+00
4.75726694e-01 1.61073044e-01 5.27762055e-01 -9.07265842e-01
-5.58528662e-01 -2.58683622e-01 -5.21848857e-01 -4.90436941e-01
7.58879840e-01 -5.54147005e-01 -1.80332935e+00 3.92660201e-01
-6.62887514e-01 -8.41064751e-01 -4.02937025e-01 5.18044829e-01
-9.96212602e-01 1.79615930e-01 -8.48216236e-01 -1.35721195e+00
-3.83531362e-01 -1.00346315e+00 6.05204880e-01 5.59205413e-01
3.26003097e-02 -1.11801398e+00 4.16199327e-01 -7.04373643e-02
4.90418315e-01 4.21799719e-01 6.97945178e-01 -1.81951039e-02
-5.51406145e-01 1.93724781e-02 2.64750987e-01 1.77158862e-01
-1.25942200e-01 2.34691277e-01 -3.33505452e-01 -6.21171713e-01
2.96242367e-02 -4.54895884e-01 5.98311663e-01 5.55320203e-01
1.13825357e+00 -3.24396729e-01 -4.46097314e-01 3.33420932e-01
1.29347003e+00 7.98351943e-01 1.70946509e-01 4.16123807e-01
2.80345887e-01 3.12038839e-01 6.85661972e-01 9.20366645e-01
-7.30904788e-02 4.93847311e-01 2.15158075e-01 1.62010208e-01
2.53718346e-01 -2.73065865e-01 8.18349496e-02 6.24346912e-01
-1.96280390e-01 -1.29302934e-01 -8.78571272e-01 3.13176334e-01
-1.67553222e+00 -9.28188205e-01 1.57932684e-01 2.28811288e+00
6.43553257e-01 2.80516267e-01 1.37559310e-01 -1.73240691e-01
6.03150964e-01 -2.00031906e-01 -9.47489500e-01 -7.56478429e-01
2.07201198e-01 4.40623127e-02 8.23151648e-01 6.15287542e-01
-9.96839106e-01 6.43758237e-01 6.94259167e+00 6.20081484e-01
-1.50457156e+00 -5.19261599e-01 6.25228405e-01 -1.03695579e-01
1.52824625e-01 -1.63256019e-01 -7.98798800e-01 7.45562196e-01
9.09782529e-01 -4.39024895e-01 7.51244247e-01 7.06611574e-01
7.64555037e-01 -4.37203914e-01 -5.90856433e-01 5.96289396e-01
-5.79593658e-01 -1.53495622e+00 -2.36702740e-01 -1.97249532e-01
8.41462016e-01 -4.62946713e-01 -1.67717665e-01 4.59655762e-01
4.66356397e-01 -7.07425654e-01 5.44118464e-01 8.14850032e-01
5.78708529e-01 -6.76143646e-01 4.82392967e-01 4.59917098e-01
-1.03913772e+00 -6.73399210e-01 -4.33243960e-01 -4.86692399e-01
2.54988074e-01 2.54464030e-01 -9.96895075e-01 2.09896490e-01
3.29981923e-01 4.40364629e-01 -8.42013955e-02 1.11650538e+00
-3.11664329e-03 5.96789956e-01 -2.75064260e-01 -9.15718496e-01
3.49063396e-01 -2.89865971e-01 6.31867111e-01 7.00102627e-01
1.40290007e-01 4.21581157e-02 2.94137001e-01 7.01334715e-01
4.52621162e-01 -4.49787453e-02 -4.93034244e-01 6.28894493e-02
3.82291645e-01 1.07498050e+00 -7.57711709e-01 -1.22716531e-01
3.14695388e-01 -1.10390843e-04 3.09492856e-01 1.77968487e-01
-7.62892962e-01 -3.10981393e-01 7.46848643e-01 4.82602566e-01
3.31073791e-01 -6.36071920e-01 -2.72516478e-02 -8.60932827e-01
-2.84392506e-01 -8.31020102e-02 5.93924187e-02 -4.87934053e-01
-1.11177623e+00 2.52791345e-01 2.62379646e-01 -1.10738289e+00
-7.16878712e-01 -8.58754396e-01 -7.85286486e-01 4.90591943e-01
-1.34850705e+00 -2.32859045e-01 2.04142675e-01 1.74585611e-01
4.50149298e-01 -4.36657220e-01 3.44192713e-01 -2.25140527e-01
-5.98930717e-01 2.06847429e-01 7.00295448e-01 1.09270118e-01
2.28721306e-01 -1.28389585e+00 5.62190451e-02 3.94044131e-01
-7.00995982e-01 3.00946265e-01 9.31766629e-01 -7.01492786e-01
-1.76503098e+00 -5.86163998e-01 -2.00653657e-01 2.29576845e-02
8.98048818e-01 -2.17405349e-01 -8.63684833e-01 -1.48067325e-01
6.58861399e-02 3.62967812e-02 4.34477590e-02 -2.87888616e-01
6.49732590e-01 -3.11293751e-01 -9.88023579e-01 4.69536483e-01
5.86053491e-01 7.97337964e-02 -2.88400531e-01 2.68200457e-01
4.67191994e-01 -6.69352174e-01 -7.52501547e-01 4.05599713e-01
6.75825417e-01 -5.97633541e-01 7.76018679e-01 -9.73253548e-01
5.50558984e-01 -2.38452330e-02 4.68898445e-01 -1.43862915e+00
-1.56855494e-01 -1.02228272e+00 -3.27557236e-01 5.45112371e-01
1.59088805e-01 -5.87118089e-01 8.55602741e-01 4.38707411e-01
-2.31534690e-01 -1.39422441e+00 -1.14595163e+00 -7.11753786e-01
5.72341144e-01 -3.06747928e-02 2.83202171e-01 5.82227349e-01
1.49203271e-01 5.10990620e-02 -4.49026912e-01 -5.48711568e-02
3.90208632e-01 2.14655966e-01 6.74659312e-01 -9.91158068e-01
-4.11163837e-01 -4.21054870e-01 -3.49375345e-02 -9.70730245e-01
3.12771797e-01 -4.17811215e-01 1.09591648e-01 -1.10103738e+00
-3.48951042e-01 -6.77182555e-01 1.60464287e-01 1.44061178e-01
1.37234973e-02 -4.70688313e-01 2.79124945e-01 2.37751976e-01
-3.88688534e-01 6.83329463e-01 2.09912896e+00 2.18301907e-01
-4.56556618e-01 3.15159261e-01 -2.73866504e-01 2.83144712e-01
8.82768333e-01 -5.81207909e-02 -4.79534388e-01 7.02848509e-02
2.74423569e-01 8.18520963e-01 7.29466453e-02 -1.05340266e+00
2.10824192e-01 -5.88513315e-01 4.74007100e-01 -1.68563709e-01
2.24497765e-01 -5.33230484e-01 -3.59390050e-01 1.03050053e+00
-4.04163063e-01 6.13561794e-02 3.94177437e-01 6.25414073e-01
1.45376951e-01 -2.93458879e-01 9.25749838e-01 -3.71674299e-01
-4.74243075e-01 2.88875908e-01 -7.99948514e-01 2.23677486e-01
1.23728335e+00 3.14370468e-02 -2.26007596e-01 -3.23307782e-01
-6.44913852e-01 5.48296571e-01 6.09853938e-02 2.43226811e-01
3.32203567e-01 -8.90284479e-01 -5.20473838e-01 5.52604795e-01
-6.28494442e-01 3.26865874e-02 3.15417349e-01 3.95029008e-01
-1.01148963e+00 6.39123023e-01 -3.84982049e-01 -4.91792679e-01
-4.78516132e-01 4.50323045e-01 9.02036548e-01 -5.25275886e-01
-5.49081743e-01 4.05141056e-01 -6.07937090e-02 -1.38698563e-01
-6.89327493e-02 -3.42934459e-01 -1.19427361e-01 -2.66681522e-01
2.50615925e-01 5.92112541e-01 2.94240750e-02 8.02976917e-03
1.42464012e-01 6.69391870e-01 1.75848782e-01 8.08445364e-03
1.25670660e+00 1.69422507e-01 3.49855661e-01 2.43451908e-01
6.69885039e-01 -3.68877321e-01 -2.12967372e+00 4.12028790e-01
-2.03883365e-01 -5.57126641e-01 -1.48771070e-02 -7.09718227e-01
-9.36170995e-01 7.78849363e-01 3.19300860e-01 4.99566257e-01
6.04719341e-01 -3.26315969e-01 6.29160285e-01 6.65852547e-01
2.46662423e-01 -1.12218237e+00 4.63922322e-01 7.79302180e-01
8.65336061e-01 -8.25838327e-01 3.02659348e-03 3.44044380e-02
-6.76751494e-01 1.60434639e+00 1.05111980e+00 -5.30252993e-01
7.52570629e-01 6.15807056e-01 -6.53584301e-02 9.55816582e-02
-1.03203499e+00 2.87713766e-01 -1.01378866e-01 3.64282876e-01
1.32436156e-01 1.18690915e-01 -4.64276433e-01 6.61458150e-02
-2.82162786e-01 -4.12655249e-02 5.52877843e-01 7.15244532e-01
-8.59017253e-01 -8.61138225e-01 -3.33562255e-01 4.24360871e-01
-1.54850796e-01 4.93643910e-01 3.07203889e-01 1.12970030e+00
-1.66873094e-02 5.25994599e-01 3.85745347e-01 1.69439763e-01
2.85245240e-01 -3.54461670e-01 4.44947600e-01 -6.33206442e-02
-5.04179239e-01 -2.04434320e-01 -2.40469933e-01 -2.05894798e-01
8.60994533e-02 -6.51105285e-01 -1.38304210e+00 -2.77286649e-01
-3.19208175e-01 6.39253020e-01 4.57898080e-01 9.44197655e-01
5.99807277e-02 3.07318330e-01 1.01545906e+00 -1.02951431e+00
-8.06375384e-01 -8.42374563e-01 -5.53865731e-01 2.47431606e-01
4.08044845e-01 -1.22267199e+00 -4.69385535e-01 -4.82755363e-01] | [4.470613479614258, 2.1580824851989746] |
bfc1fa12-c70a-4d14-919c-da5c58a96a46 | semi-supervised-neural-architecture-search | 2002.10389 | null | https://arxiv.org/abs/2002.10389v4 | https://arxiv.org/pdf/2002.10389v4.pdf | Semi-Supervised Neural Architecture Search | Neural architecture search (NAS) relies on a good controller to generate better architectures or predict the accuracy of given architectures. However, training the controller requires both abundant and high-quality pairs of architectures and their accuracy, while it is costly to evaluate an architecture and obtain its accuracy. In this paper, we propose SemiNAS, a semi-supervised NAS approach that leverages numerous unlabeled architectures (without evaluation and thus nearly no cost). Specifically, SemiNAS 1) trains an initial accuracy predictor with a small set of architecture-accuracy data pairs; 2) uses the trained accuracy predictor to predict the accuracy of large amount of architectures (without evaluation); and 3) adds the generated data pairs to the original data to further improve the predictor. The trained accuracy predictor can be applied to various NAS algorithms by predicting the accuracy of candidate architectures for them. SemiNAS has two advantages: 1) It reduces the computational cost under the same accuracy guarantee. On NASBench-101 benchmark dataset, it achieves comparable accuracy with gradient-based method while using only 1/7 architecture-accuracy pairs. 2) It achieves higher accuracy under the same computational cost. It achieves 94.02% test accuracy on NASBench-101, outperforming all the baselines when using the same number of architectures. On ImageNet, it achieves 23.5% top-1 error rate (under 600M FLOPS constraint) using 4 GPU-days for search. We further apply it to LJSpeech text to speech task and it achieves 97% intelligibility rate in the low-resource setting and 15% test error rate in the robustness setting, with 9%, 7% improvements over the baseline respectively. | ['Tie-Yan Liu', 'Rui Wang', 'Renqian Luo', 'Enhong Chen', 'Xu Tan', 'Tao Qin'] | 2020-02-24 | null | http://proceedings.neurips.cc/paper/2020/hash/77305c2f862ad1d353f55bf38e5a5183-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/77305c2f862ad1d353f55bf38e5a5183-Paper.pdf | neurips-2020-12 | ['natural-language-transduction'] | ['natural-language-processing'] | [-1.01552382e-01 -2.61107057e-01 -2.12366104e-01 -4.86407220e-01
-1.28091180e+00 -7.45209038e-01 1.35765940e-01 -2.20419809e-01
-5.86361051e-01 5.85584760e-01 -1.27645418e-01 -5.26033282e-01
1.31759912e-01 -5.20719051e-01 -1.04414749e+00 -5.22974491e-01
2.34605744e-01 8.29253614e-01 2.34258816e-01 -1.38362333e-01
2.06715316e-01 1.30116031e-01 -1.55605066e+00 2.96484083e-01
8.38259757e-01 1.71882951e+00 3.70762020e-01 8.64767373e-01
8.92632604e-02 7.85813451e-01 -5.85192621e-01 -3.73991430e-01
7.16474891e-01 -2.34113093e-02 -8.30127239e-01 -3.59560370e-01
7.48839498e-01 -4.05211508e-01 -1.08583318e-02 1.02668178e+00
6.15932584e-01 -9.95440260e-02 4.08313483e-01 -1.17456305e+00
-5.18974185e-01 4.93666589e-01 -4.26932395e-01 3.66138101e-01
-1.49055153e-01 3.77438247e-01 1.23956704e+00 -1.02804661e+00
2.26555139e-01 9.52380836e-01 4.79638249e-01 6.18477404e-01
-9.29017305e-01 -9.85459566e-01 3.62954549e-02 1.03743851e-01
-1.43009651e+00 -7.53610730e-01 2.50807613e-01 -2.33228609e-01
1.42645359e+00 3.67152542e-01 3.69604826e-01 1.06256771e+00
-2.03883305e-01 5.84411323e-01 9.83019888e-01 -2.81914562e-01
4.98312622e-01 1.60040498e-01 4.08229917e-01 8.66275787e-01
5.50810955e-02 3.15848351e-01 -5.10358274e-01 -1.16686866e-01
5.20705402e-01 -4.23409045e-01 -1.39327317e-01 1.29534379e-01
-1.06042421e+00 5.62887311e-01 6.64902270e-01 1.38415694e-01
-3.40507120e-01 1.92413330e-01 3.88822734e-01 6.80477798e-01
9.54089165e-02 7.98081696e-01 -8.23514104e-01 -3.22305888e-01
-9.29590523e-01 4.64957766e-02 7.04719841e-01 1.15696120e+00
6.04762733e-01 5.33983111e-01 -6.53958097e-02 9.23994601e-01
2.00349584e-01 6.31857574e-01 9.58309472e-01 -7.22816348e-01
7.75975823e-01 7.97819674e-01 -1.17115825e-01 -5.93322098e-01
-2.02381343e-01 -9.62104619e-01 -8.99536431e-01 1.40941516e-01
-9.59285069e-03 -1.43670291e-01 -1.11534333e+00 1.75210214e+00
1.48983508e-01 3.98513377e-01 2.19896227e-01 1.01535356e+00
8.55686307e-01 8.06484222e-01 -3.22357118e-01 -1.26840383e-01
1.23180676e+00 -1.50194073e+00 -9.64657813e-02 -2.74047256e-01
5.98623276e-01 -8.24927270e-01 1.51494074e+00 5.35199940e-01
-1.27451837e+00 -6.78736031e-01 -1.34409690e+00 1.37864605e-01
-1.09307401e-01 6.25412703e-01 2.36064464e-01 7.49184191e-01
-1.53576422e+00 3.55994105e-01 -9.15229142e-01 -9.94755849e-02
3.04846495e-01 9.37399209e-01 2.85786670e-02 3.25629741e-01
-7.07537293e-01 6.61253631e-01 3.48979056e-01 -1.57167047e-01
-1.21737349e+00 -8.74462783e-01 -3.76823217e-01 1.05159178e-01
3.35896164e-01 -7.20578969e-01 1.31360626e+00 -1.31928527e+00
-1.84631658e+00 5.78091025e-01 4.45196405e-03 -7.61863828e-01
4.01370645e-01 -3.52720201e-01 -6.29689217e-01 -4.44063455e-01
-1.04652427e-01 9.67741013e-01 7.20631540e-01 -7.36151934e-01
-7.74010420e-01 -3.46053392e-01 -7.34482855e-02 2.04406738e-01
-6.14334404e-01 -2.28922799e-01 -8.11177194e-01 -4.27349448e-01
7.45326281e-02 -1.24764156e+00 -1.98647715e-02 -4.09595639e-01
-5.22097826e-01 -8.50119144e-02 5.52817404e-01 -4.02699322e-01
1.43694389e+00 -2.01167989e+00 -5.36578894e-03 2.52082825e-01
4.43011373e-02 5.69779813e-01 -5.30526817e-01 -2.01148435e-01
9.11032110e-02 1.11034855e-01 -1.43378466e-01 -2.97143102e-01
-1.77476928e-01 -1.07046671e-01 -2.57341057e-01 4.97860909e-02
2.96438217e-01 7.53209889e-01 -4.01196063e-01 -1.96691126e-01
-5.85670359e-02 2.20875815e-01 -9.43735003e-01 3.95865500e-01
-1.86465517e-01 2.07995728e-01 -4.04716402e-01 7.44961679e-01
3.65208000e-01 -4.92880732e-01 7.58839473e-02 -3.31393659e-01
-2.77495533e-02 4.33823496e-01 -1.12369609e+00 1.64578795e+00
-9.29229259e-01 4.75561082e-01 -2.05882370e-01 -7.28926241e-01
1.17874551e+00 1.41170904e-01 -1.23008922e-01 -8.34758759e-01
1.77182809e-01 5.96651018e-01 2.69179583e-01 -1.29761428e-01
9.44603905e-02 6.03131771e-01 1.79913472e-02 5.02010286e-01
-1.48860589e-02 1.20923847e-01 -9.58293527e-02 -1.10720739e-01
1.44728768e+00 -1.38218910e-01 9.35508162e-02 -4.47433650e-01
6.97493434e-01 1.24515273e-01 5.31087339e-01 6.12039447e-01
-9.06930566e-02 4.68803346e-01 8.19259509e-02 -8.78255546e-01
-1.39060795e+00 -9.02163029e-01 -9.42696482e-02 1.25364208e+00
-1.07183091e-01 -1.79486111e-01 -8.03607404e-01 -6.68209434e-01
-3.18086058e-01 7.03851402e-01 -4.83712524e-01 -9.43491459e-02
-6.09127402e-01 -7.12907135e-01 8.18072736e-01 6.79519653e-01
8.93070400e-01 -8.45947623e-01 -6.06414914e-01 1.28136594e-02
1.11777242e-02 -1.08700657e+00 -5.03338277e-01 1.81056514e-01
-9.06776130e-01 -9.48307216e-01 -3.45350623e-01 -9.65297699e-01
5.06479919e-01 -6.51919916e-02 1.47655356e+00 4.46157813e-01
-2.50203647e-02 -2.82293946e-01 -1.62798595e-02 -1.49603948e-01
-5.00268459e-01 4.48030740e-01 2.37077147e-01 -1.13688603e-01
1.36165842e-01 -4.86216635e-01 -7.39797413e-01 5.53345501e-01
-3.42159957e-01 1.32962376e-01 9.07950878e-01 1.03406799e+00
8.68997276e-01 -3.10756326e-01 7.33930290e-01 -6.00000262e-01
4.67882097e-01 -3.42496008e-01 -9.74204838e-01 4.22029227e-01
-1.18013561e+00 3.41529071e-01 8.18271816e-01 -4.64390039e-01
-5.47579825e-01 1.19192012e-01 -1.19126014e-01 -7.68181324e-01
1.89429447e-01 1.76535681e-01 1.68952629e-01 6.58835247e-02
9.58931625e-01 2.24059463e-01 3.11241541e-02 -5.99116981e-01
-3.54358624e-03 8.11234772e-01 6.26364946e-01 -4.31082875e-01
4.92174208e-01 -8.46209973e-02 -2.49991536e-01 -2.79700607e-01
-7.84933388e-01 -3.03205073e-01 -2.24239573e-01 2.58716434e-01
5.50152779e-01 -1.15194321e+00 -7.59765983e-01 1.25584513e-01
-9.51964140e-01 -4.68468308e-01 2.08347321e-01 4.56497788e-01
-2.71462440e-01 -2.12320417e-01 -5.54002225e-01 -5.22293031e-01
-1.19670963e+00 -1.70541763e+00 1.06977201e+00 -1.58320609e-02
-1.89071864e-01 -4.31781322e-01 -1.09755360e-01 4.70003217e-01
5.49164355e-01 -2.06742719e-01 1.08199859e+00 -8.21212769e-01
-5.70849836e-01 -8.95916447e-02 -2.74204701e-01 5.92886746e-01
-2.41312474e-01 3.36073712e-02 -1.01145697e+00 -4.31848884e-01
-8.22217613e-02 -6.37120247e-01 5.85526764e-01 3.48067164e-01
1.31954098e+00 -5.29083848e-01 2.11082883e-02 6.92323446e-01
1.50686085e+00 3.67387503e-01 3.48743021e-01 4.70830917e-01
5.26644230e-01 1.07029572e-01 5.04220307e-01 1.40023902e-01
6.73826039e-02 9.97031450e-01 4.66004699e-01 1.34189546e-01
-2.89439708e-01 3.82589474e-02 4.54888284e-01 1.36934412e+00
1.29461616e-01 -2.11152807e-01 -1.23891830e+00 3.03701550e-01
-1.73840570e+00 -2.32534483e-01 5.30719459e-02 2.43117476e+00
9.61272478e-01 3.14519465e-01 2.51112044e-01 -8.44313279e-02
6.64632797e-01 -3.24621230e-01 -1.08803570e+00 -5.36930025e-01
7.87196308e-03 3.72126043e-01 4.61723864e-01 3.93389344e-01
-8.53794575e-01 9.81160223e-01 6.07912827e+00 8.70207846e-01
-1.26557374e+00 2.68804818e-01 1.28882241e+00 -5.60599208e-01
4.27719243e-02 -3.18340302e-01 -1.04539752e+00 4.99535412e-01
1.43134308e+00 1.46271452e-01 6.63579583e-01 1.30256510e+00
-1.49369404e-01 2.34390900e-01 -1.19264317e+00 1.16602564e+00
1.53161697e-02 -1.44832587e+00 1.04444481e-01 -1.76209062e-01
8.99535835e-01 5.58146179e-01 3.26128006e-01 4.95208472e-01
3.83865625e-01 -1.18028605e+00 8.96013558e-01 -2.23707296e-02
1.09191108e+00 -8.28408718e-01 9.43650305e-01 4.38478202e-01
-1.04459512e+00 -3.00075859e-01 -5.88234186e-01 1.60567418e-01
-4.81467485e-01 1.65004298e-01 -1.20238900e+00 5.72900325e-02
8.81882787e-01 1.10170454e-01 -7.48498976e-01 7.90908337e-01
1.32719025e-01 8.26381862e-01 -5.55810630e-01 -3.55891407e-01
5.78272641e-01 3.56086828e-02 2.45570824e-01 7.74350643e-01
5.60385287e-01 -2.18971252e-01 3.44256945e-02 6.51177406e-01
-3.01551163e-01 2.64859915e-01 -8.35961252e-02 2.40028456e-01
9.30288613e-01 1.12029195e+00 -2.58476168e-01 -4.64054257e-01
-2.06923708e-01 6.78364456e-01 5.87101460e-01 7.43367299e-02
-1.01097429e+00 7.97090754e-02 4.62846756e-01 -8.54968727e-02
5.57786226e-03 1.83147550e-01 -6.20858431e-01 -7.24298775e-01
1.33336410e-01 -1.38499439e+00 3.38563234e-01 -7.94850111e-01
-8.52705121e-01 1.48094451e+00 -6.46020532e-01 -1.30679452e+00
-4.39910829e-01 -5.40102005e-01 -4.65789348e-01 7.29972839e-01
-1.20748484e+00 -8.32851529e-01 -5.51567554e-01 4.64852750e-01
1.06101656e+00 -9.12120998e-01 9.92042005e-01 2.67199963e-01
-9.47823584e-01 1.16494310e+00 1.24033868e-01 6.76931534e-03
5.73718488e-01 -1.07576847e+00 7.01738060e-01 6.68501794e-01
4.88463134e-01 4.04861361e-01 3.39414179e-01 -3.82268727e-01
-1.70926487e+00 -1.33808553e+00 4.69505668e-01 -4.24364328e-01
6.34393692e-01 -2.24757582e-01 -6.94501042e-01 4.64849830e-01
1.41119286e-01 -2.72842795e-02 5.87652802e-01 2.01501444e-01
-4.49891239e-01 -4.97157335e-01 -9.41308856e-01 5.25388002e-01
1.05076110e+00 -1.66590035e-01 -5.89138567e-02 3.63730848e-01
9.27164435e-01 -6.14353955e-01 -7.01854408e-01 6.49289787e-01
5.39983273e-01 -8.46096575e-01 7.26476967e-01 -6.72840595e-01
4.17476118e-01 -1.99972704e-01 -5.21668732e-01 -1.11596572e+00
-1.52149826e-01 -4.30661738e-01 -1.51301175e-01 8.58614266e-01
1.18057013e+00 -6.74360275e-01 9.43399310e-01 5.99891305e-01
-3.02453160e-01 -1.33221912e+00 -1.01807654e+00 -9.23003256e-01
-2.68935170e-02 -4.57551450e-01 1.02848423e+00 6.94016814e-01
-6.32757902e-01 6.07501686e-01 -1.80363178e-01 2.74695635e-01
2.64489889e-01 1.42005593e-01 7.87465394e-01 -8.29068124e-01
-6.22024775e-01 -5.74702919e-01 -2.85689592e-01 -1.02453971e+00
4.73852083e-02 -9.02662575e-01 -7.83076435e-02 -7.15808749e-01
7.78628737e-02 -9.53301907e-01 -5.60584605e-01 6.14222825e-01
5.40331155e-02 3.11504781e-01 -8.57921019e-02 6.43981636e-01
-6.68530703e-01 4.36181515e-01 7.09347963e-01 -1.11762956e-01
-2.90307462e-01 1.32960826e-01 -4.94953692e-01 4.79905277e-01
9.51970160e-01 -2.74980485e-01 -7.29518712e-01 -8.59794438e-01
1.49323493e-01 1.06771803e-02 1.60122082e-01 -1.43766344e+00
2.95573115e-01 2.38259763e-01 2.05485776e-01 -3.50480288e-01
2.45770589e-01 -6.40219033e-01 2.22001851e-01 5.80676019e-01
-4.84359205e-01 7.70512104e-01 2.47331068e-01 4.57616538e-01
-1.83564276e-01 -1.89160153e-01 8.98836672e-01 1.24074891e-01
-7.49174654e-01 3.21791470e-01 2.63667166e-01 3.99730392e-02
9.13614392e-01 -7.29719028e-02 -4.70649809e-01 -2.57195141e-02
-4.86565173e-01 3.33305299e-01 3.58026177e-01 5.09354055e-01
4.87864256e-01 -1.39043391e+00 -6.59468174e-01 2.82089412e-01
-7.21989479e-03 5.47928587e-02 -7.97089860e-02 7.31078625e-01
-7.53166556e-01 6.01591468e-01 -2.68499684e-02 -9.20536757e-01
-1.26107621e+00 3.11869949e-01 4.80007797e-01 -3.70687485e-01
-1.74786493e-01 1.21164322e+00 1.09587990e-01 -4.92360264e-01
4.39030290e-01 -9.19545591e-02 6.45803958e-02 -5.36307573e-01
5.47032058e-01 3.95103991e-01 5.84045529e-01 -3.77783537e-01
-5.12345731e-01 3.94205421e-01 -2.32163697e-01 -1.30779192e-01
1.35867512e+00 2.68691719e-01 2.27436468e-01 2.50544190e-01
1.36434805e+00 -3.76539916e-01 -1.18903840e+00 -2.40823478e-01
-2.04775240e-02 -2.05477789e-01 5.37255824e-01 -1.25864577e+00
-1.44350660e+00 8.56002629e-01 1.09760058e+00 -1.16612330e-01
1.33088923e+00 -3.03148448e-01 7.87852228e-01 5.92151344e-01
4.32430804e-01 -1.12791491e+00 2.86321402e-01 4.49208438e-01
1.11412251e+00 -1.35430121e+00 -1.99352473e-01 -2.22022265e-01
-7.73599446e-01 9.23175991e-01 1.18277001e+00 -1.78812057e-01
4.56466228e-01 5.35965621e-01 8.88957679e-02 -6.85734749e-02
-1.38937986e+00 1.96629092e-01 5.21785080e-01 1.09767675e-01
3.38268131e-01 1.96033508e-01 1.73340321e-01 5.75948358e-01
-6.57180071e-01 -3.37022871e-01 -5.31079993e-02 4.45090979e-01
-3.76258641e-01 -8.88414502e-01 5.54054976e-03 7.30498374e-01
-3.71084303e-01 -5.77498734e-01 -3.66029114e-01 5.10574639e-01
-2.06951454e-01 9.05638993e-01 2.04931736e-01 -9.36860859e-01
3.15882355e-01 3.49346101e-02 -3.45372930e-02 -4.17935371e-01
-9.34942782e-01 5.89333884e-02 2.98509270e-01 -7.53292084e-01
1.50373966e-01 -2.85221517e-01 -9.98368561e-01 -4.51596588e-01
-5.13728261e-01 1.13230504e-01 8.95061672e-01 5.85856020e-01
9.70293224e-01 4.04872626e-01 7.97720075e-01 -3.00181061e-01
-1.07184076e+00 -1.02979934e+00 7.57630020e-02 4.22673821e-02
7.91764073e-03 -5.60266852e-01 -2.19386488e-01 -1.89532280e-01] | [8.675314903259277, 3.288219928741455] |
1a8d9444-bd87-4a96-81d8-57e54eb85e65 | typo-robust-representation-learning-for-dense | 2306.10348 | null | https://arxiv.org/abs/2306.10348v1 | https://arxiv.org/pdf/2306.10348v1.pdf | Typo-Robust Representation Learning for Dense Retrieval | Dense retrieval is a basic building block of information retrieval applications. One of the main challenges of dense retrieval in real-world settings is the handling of queries containing misspelled words. A popular approach for handling misspelled queries is minimizing the representations discrepancy between misspelled queries and their pristine ones. Unlike the existing approaches, which only focus on the alignment between misspelled and pristine queries, our method also improves the contrast between each misspelled query and its surrounding queries. To assess the effectiveness of our proposed method, we compare it against the existing competitors using two benchmark datasets and two base encoders. Our method outperforms the competitors in all cases with misspelled queries. Our code and models are available at https://github. com/panuthept/DST-DenseRetrieval. | ['Sarana Nutanong', 'Ekapol Chuangsuwanich', 'Can Udomcharoenchaikit', 'Peerat Limkonchotiwat', 'Wuttikorn Ponwitayarat', 'Panuthep Tasawong'] | 2023-06-17 | null | null | null | null | ['information-retrieval'] | ['natural-language-processing'] | [-1.73960757e-02 -5.43889344e-01 -3.01312417e-01 -1.95142210e-01
-1.51026869e+00 -7.58368969e-01 7.87958622e-01 1.62025228e-01
-6.69074714e-01 6.06564283e-01 4.56234008e-01 7.53071383e-02
-1.19190790e-01 -6.75059080e-01 -7.07663298e-01 -4.84304518e-01
3.41168553e-01 7.41867721e-01 3.87971699e-01 -5.14549494e-01
2.84798950e-01 2.20519066e-01 -1.54325891e+00 4.97171074e-01
7.43653893e-01 9.06470478e-01 5.15690029e-01 6.79497421e-01
-2.34454051e-01 3.29862624e-01 -6.65484726e-01 -5.98792791e-01
4.42754537e-01 -3.35836969e-02 -5.80086410e-01 -3.26895356e-01
8.38097751e-01 -3.45177203e-01 -9.27485883e-01 1.25248742e+00
6.93509281e-01 2.54837245e-01 4.42633003e-01 -8.10284495e-01
-1.17839706e+00 6.27283871e-01 -4.79659587e-01 4.81555611e-01
3.62979859e-01 -2.29830503e-01 1.34158266e+00 -1.58692801e+00
4.87404764e-01 1.50834632e+00 2.66421944e-01 2.65206188e-01
-7.63446331e-01 -7.65998900e-01 6.27204180e-02 4.02148724e-01
-1.82674074e+00 -7.33779013e-01 2.21044645e-01 8.41714069e-02
8.85183394e-01 5.38547575e-01 2.15645060e-01 1.05515921e+00
-1.06400363e-01 1.20060289e+00 5.76627135e-01 -4.64173049e-01
-1.77426621e-01 2.56658286e-01 4.68762904e-01 2.70326436e-01
3.86377543e-01 3.09600413e-01 -3.22235793e-01 -3.38509738e-01
3.51330131e-01 1.32688403e-01 -5.51700473e-01 -5.19079156e-02
-1.22507763e+00 8.89390230e-01 4.93703842e-01 5.14250875e-01
-2.88887709e-01 2.14506254e-01 1.08842976e-01 3.09404135e-01
3.67132008e-01 5.97217083e-01 -3.38084340e-01 -1.45532385e-01
-1.02463698e+00 3.51316601e-01 6.12399161e-01 1.34208167e+00
8.59918654e-01 -3.87133002e-01 -7.72047937e-01 1.14126873e+00
2.94716328e-01 7.00790524e-01 6.92301571e-01 -5.21670759e-01
5.53355575e-01 4.93519515e-01 3.17300022e-01 -1.08685148e+00
2.25608513e-01 -8.01592588e-01 -5.51073790e-01 -7.80169129e-01
3.15040164e-02 9.88472924e-02 -1.01956785e+00 1.33023071e+00
3.78178656e-02 1.12891026e-01 1.40727028e-01 1.01864600e+00
1.09020412e+00 1.12268376e+00 -3.87423903e-01 6.20282069e-03
1.18475699e+00 -1.36286604e+00 -8.90136361e-01 -5.30662954e-01
4.97655928e-01 -1.24159813e+00 1.28638732e+00 3.90324779e-02
-9.78801131e-01 -3.75727534e-01 -7.03207254e-01 -5.11302829e-01
-5.54650009e-01 3.53824556e-01 3.57368857e-01 1.50371939e-01
-1.08335602e+00 2.11769231e-02 -3.82091671e-01 -3.59622806e-01
1.11153081e-01 1.79438680e-01 -2.43778571e-01 -4.44231659e-01
-1.57470798e+00 9.88329828e-01 4.91860926e-01 1.33902624e-01
-1.02577269e+00 -5.98610640e-01 -6.83939040e-01 3.20711941e-01
4.41727668e-01 -4.92277056e-01 1.60066307e+00 -6.27172828e-01
-6.68469727e-01 6.73958957e-01 -4.61005032e-01 -4.30331975e-01
3.08525592e-01 -6.77612484e-01 -3.67043823e-01 -1.06613271e-01
2.95898169e-02 6.58840001e-01 8.59101772e-01 -1.18067038e+00
-5.68742216e-01 -2.54697531e-01 1.07316315e-01 4.27309304e-01
-4.46909159e-01 -1.46353096e-01 -1.42526793e+00 -8.30962121e-01
-1.55796275e-01 -7.62497783e-01 -9.91083086e-02 -2.03637078e-01
-5.21340013e-01 -3.75140220e-01 9.30436909e-01 -2.65968025e-01
1.51348972e+00 -2.25785017e+00 -3.33979540e-02 9.62986276e-02
1.78599164e-01 6.77539885e-01 -6.34933472e-01 7.15737224e-01
2.91305393e-01 3.54382694e-02 2.03586519e-02 -3.92711759e-01
1.73362911e-01 2.22035035e-01 -6.91121161e-01 1.65985569e-01
-6.59948066e-02 1.09718382e+00 -1.05737984e+00 -4.68648434e-01
1.54501349e-02 7.87298083e-01 -3.95937085e-01 3.14517647e-01
-1.79523259e-01 -4.58686948e-02 -3.98269206e-01 6.25966251e-01
6.33328795e-01 -3.44818354e-01 -2.39241064e-01 -2.47656822e-01
2.29476213e-01 5.18217981e-01 -9.70544219e-01 1.69128311e+00
-5.59750915e-01 6.91047370e-01 -1.04725651e-01 -4.70006585e-01
8.73016179e-01 2.24327564e-01 3.07130873e-01 -9.36063886e-01
-2.30765771e-02 3.97400618e-01 -3.58239740e-01 -2.86209822e-01
1.12322295e+00 1.26564279e-01 -7.13925138e-02 3.95789683e-01
6.96306955e-03 -1.49259686e-01 4.13438797e-01 6.70027375e-01
1.18645787e+00 -5.06470203e-01 5.74881658e-02 1.97342485e-02
3.13897610e-01 -1.56684652e-01 4.21411097e-01 1.22833109e+00
-6.45834431e-02 8.45721722e-01 -2.11135987e-02 -1.27418637e-01
-8.29319417e-01 -1.01509368e+00 -1.51879683e-01 1.29941857e+00
4.58909929e-01 -6.69220090e-01 -3.09394509e-01 -6.30769610e-01
2.63651609e-01 7.39755869e-01 -4.56367135e-01 -4.20243204e-01
-3.72731715e-01 -5.16185045e-01 5.78928769e-01 2.01069325e-01
3.95751625e-01 -9.55690086e-01 -1.19661093e-01 -4.13792022e-02
-4.68887866e-01 -1.00934649e+00 -1.01475024e+00 -5.42299002e-02
-4.67743605e-01 -9.12515581e-01 -9.77133512e-01 -7.83246636e-01
5.15091300e-01 8.87386501e-01 1.50501931e+00 4.54207838e-01
-2.65796900e-01 3.83078039e-01 -6.23468459e-01 -4.25365508e-01
-1.11130282e-01 4.22374308e-01 -1.72506616e-01 -1.35651082e-01
8.39793384e-01 5.43593690e-02 -9.00343418e-01 3.70647281e-01
-1.38560915e+00 -2.76767552e-01 7.35085845e-01 9.96573210e-01
8.31128657e-01 -9.23144892e-02 5.58057368e-01 -9.04740751e-01
9.70284283e-01 -6.93011701e-01 -6.83892310e-01 6.74072921e-01
-7.16555059e-01 2.33935595e-01 8.63146484e-02 -3.37938130e-01
-7.43145823e-01 -1.15266629e-01 -2.11982936e-01 -5.58475018e-01
2.98450619e-01 5.54307818e-01 1.06459036e-01 2.51256585e-01
5.78965545e-01 3.34048778e-01 -3.95051122e-01 -7.36000657e-01
5.02726972e-01 8.72506976e-01 3.55151623e-01 -5.05858123e-01
6.60323560e-01 2.05041021e-01 -7.89024293e-01 -7.15972960e-01
-1.10267401e+00 -1.11841416e+00 -2.45789930e-01 -3.40142660e-02
1.75289676e-01 -9.73908544e-01 -1.10836573e-01 1.26457170e-01
-1.27976823e+00 2.12272063e-01 -3.85253310e-01 3.33280176e-01
-7.04062879e-02 4.61864442e-01 -5.02195418e-01 -5.20831525e-01
-5.47002494e-01 -1.31982744e+00 1.55280912e+00 1.17416032e-01
-1.37120962e-01 -9.02904809e-01 4.61720407e-01 3.33315551e-01
7.73076892e-01 -5.82186222e-01 7.21587896e-01 -1.07623684e+00
-6.23457730e-01 -7.20528305e-01 -4.65865999e-01 2.65160561e-01
2.65636653e-01 -1.59824669e-01 -8.90494525e-01 -6.00179076e-01
-1.26277372e-01 -5.89791656e-01 1.23481798e+00 9.96428449e-03
1.04791582e+00 -6.01099253e-01 -3.48351747e-01 3.89501750e-01
1.49801207e+00 7.80626684e-02 8.27900887e-01 2.15198204e-01
3.40132117e-01 3.15708756e-01 8.96210074e-01 3.67947310e-01
1.11381680e-01 8.15231681e-01 4.33925211e-01 8.25142041e-02
-2.44857699e-01 -4.88512397e-01 2.39134744e-01 1.08349323e+00
6.47062659e-01 -7.47462988e-01 -8.80836189e-01 9.30229127e-01
-1.88866568e+00 -9.11345899e-01 2.51545459e-01 2.24769998e+00
1.04262519e+00 -9.78404358e-02 -3.72517526e-01 -2.72647738e-01
8.31738710e-01 3.89603257e-01 -5.86240888e-01 -5.77603281e-02
-2.72617817e-01 3.89721930e-01 4.01862144e-01 8.47729206e-01
-9.89773929e-01 1.18925297e+00 6.58320427e+00 1.11422658e+00
-8.88203740e-01 6.29698932e-02 3.35599452e-01 -5.55882871e-01
-6.39796674e-01 -9.40348506e-02 -1.25482213e+00 4.55469698e-01
8.47983658e-01 -5.24410427e-01 2.32143015e-01 7.97471881e-01
-2.88173184e-02 1.03360720e-01 -1.21331441e+00 1.03683710e+00
2.41827697e-01 -1.34757662e+00 6.14163518e-01 -3.20818096e-01
6.92655742e-01 3.86679024e-01 3.78677785e-01 7.76589096e-01
3.10518354e-01 -1.02999520e+00 3.71396601e-01 4.77625430e-01
6.80624127e-01 -5.53971946e-01 8.50121677e-01 2.84328431e-01
-9.94100869e-01 1.49725154e-01 -5.83899617e-01 3.73263121e-01
-6.89066872e-02 7.73770571e-01 -7.85309911e-01 1.72847614e-01
6.31600440e-01 7.36809313e-01 -6.85532451e-01 1.31789553e+00
-6.48230240e-02 2.84107298e-01 -4.23365951e-01 -1.20152883e-01
3.89474630e-01 1.14181258e-01 6.21043980e-01 1.51568949e+00
5.10536075e-01 -2.59223074e-01 3.03170774e-02 6.10256910e-01
-5.07814050e-01 3.63172174e-01 -8.14085126e-01 -2.77266115e-01
9.40471947e-01 1.00132322e+00 -5.48115047e-03 -5.17821074e-01
-5.26305377e-01 9.38929021e-01 3.38785648e-01 5.92022121e-01
-6.20274901e-01 -6.29520118e-01 7.69701362e-01 -8.13266262e-03
3.06360722e-01 1.12912029e-01 6.71474263e-02 -1.49798965e+00
2.47229755e-01 -1.20502949e+00 6.21612847e-01 -8.12156975e-01
-1.46607435e+00 7.11444736e-01 -6.04703762e-02 -1.14677453e+00
-1.98018000e-01 -1.79922611e-01 -2.08760187e-01 8.30720723e-01
-1.74539781e+00 -8.57573032e-01 -3.52081597e-01 6.51876211e-01
1.04362059e+00 -1.88953560e-02 7.08910525e-01 7.42690742e-01
-3.15733999e-01 8.18585753e-01 3.94951940e-01 1.04725689e-01
1.03091526e+00 -9.48444366e-01 4.88705963e-01 9.28752542e-01
6.51889265e-01 8.41291368e-01 6.75405085e-01 -5.67417085e-01
-1.45352912e+00 -1.20211172e+00 1.10772634e+00 -4.47358251e-01
6.22838855e-01 -2.61346310e-01 -1.11105692e+00 6.70062482e-01
4.89893526e-01 -5.44955507e-02 5.26623964e-01 7.36605227e-02
-5.79072475e-01 -2.43623048e-01 -7.08261728e-01 6.01660550e-01
6.42372787e-01 -7.53561735e-01 -7.62031138e-01 7.23641217e-01
7.97747195e-01 -5.43364227e-01 -5.47423661e-01 5.87228239e-01
5.52298784e-01 -7.13011980e-01 1.13662255e+00 -4.50899065e-01
1.05621301e-01 -1.57656595e-01 -3.86281312e-01 -1.30761254e+00
-2.95410246e-01 -4.29264218e-01 -3.51755261e-01 9.26667690e-01
5.85082173e-01 -3.63289356e-01 6.93099856e-01 4.71353829e-01
-2.62499638e-02 -8.90016258e-01 -7.13026345e-01 -7.56834626e-01
8.71533155e-02 -3.79952967e-01 5.51879764e-01 4.94024694e-01
-3.99359941e-01 3.93969029e-01 -4.18952346e-01 9.34006646e-02
3.87185484e-01 2.13379309e-01 6.45056486e-01 -6.67020202e-01
-2.29021445e-01 -3.57734710e-01 -2.10362583e-01 -1.64745879e+00
4.78768647e-02 -9.57338214e-01 3.14792216e-01 -1.59428740e+00
3.55042011e-01 -4.54850972e-01 -5.56090593e-01 4.10532534e-01
-2.75630563e-01 4.22171772e-01 1.10403873e-01 4.31659758e-01
-8.79915297e-01 8.66294384e-01 1.12079549e+00 -4.50471401e-01
-2.32031755e-02 2.51557864e-02 -8.10139239e-01 3.09706390e-01
7.43465722e-01 -6.75744891e-01 -5.47948718e-01 -1.04806328e+00
3.35846454e-01 -2.59664536e-01 1.36377484e-01 -7.13742018e-01
4.93416965e-01 5.08661084e-02 -7.96108544e-02 -9.67737138e-01
5.50420225e-01 -8.33723605e-01 -4.27356400e-02 1.05114505e-01
-7.93265820e-01 4.85953122e-01 2.63911217e-01 6.22595906e-01
-8.33677232e-01 -4.91118610e-01 5.74133158e-01 -1.69268608e-01
-6.56928420e-01 5.32731593e-01 -1.22032687e-01 3.93990725e-01
7.43251383e-01 2.90682763e-01 -7.01223314e-01 -7.33475447e-01
-2.63415694e-01 4.68031526e-01 2.44149521e-01 7.59899616e-01
1.14567101e+00 -1.39208925e+00 -8.85395646e-01 1.73775211e-01
4.49214429e-01 1.44147456e-01 -1.07180826e-01 5.17317712e-01
-4.66815144e-01 1.12329221e+00 3.35216373e-01 -3.44050527e-01
-1.42260110e+00 4.96608257e-01 3.37275386e-01 -3.93705010e-01
-3.10410023e-01 1.03822601e+00 4.13994193e-01 -2.34680474e-01
6.94170535e-01 -2.07356453e-01 5.70035279e-02 -3.43006067e-02
8.86239707e-01 3.12395860e-02 3.81400377e-01 -6.01198256e-01
-2.83021003e-01 3.30949605e-01 -8.67570162e-01 -2.04990581e-02
1.09089863e+00 -2.16188982e-01 -1.73472658e-01 3.90426010e-01
1.60380816e+00 1.14909001e-01 -6.80812836e-01 -7.85662651e-01
2.01744601e-01 -9.31223214e-01 2.34373197e-01 -6.57734871e-01
-1.21089113e+00 7.99757540e-01 5.42120039e-01 -8.85802805e-02
1.20932841e+00 1.59859806e-01 1.17626595e+00 8.93169343e-01
3.53050739e-01 -9.48087990e-01 1.87449008e-01 8.18625748e-01
1.23681319e+00 -1.30414236e+00 -9.32827517e-02 -1.46804631e-01
-3.68423462e-01 7.22745657e-01 5.75704336e-01 -6.35739416e-02
6.57682598e-01 4.28107791e-02 1.80800661e-01 -2.93227255e-01
-1.17473340e+00 -3.45558405e-01 6.97938621e-01 1.70436069e-01
4.90673363e-01 -3.23707908e-01 -1.90788969e-01 2.43412688e-01
-1.63112417e-01 -2.32368767e-01 2.02754140e-01 9.11916971e-01
-5.16640723e-01 -1.16501510e+00 -2.00617552e-01 5.81085503e-01
-5.92270195e-01 -7.16156840e-01 -7.17014551e-01 5.95327318e-01
-3.88073981e-01 8.66530836e-01 4.30102050e-02 -3.99296194e-01
3.09090734e-01 -6.73892647e-02 1.22983471e-01 -9.23910737e-01
-6.73104525e-01 9.59648713e-02 -6.77844882e-02 -5.00997424e-01
-1.49169341e-01 -1.88088059e-01 -7.96985149e-01 -1.90640733e-01
-6.36526346e-01 5.42598724e-01 4.65434790e-01 5.33149421e-01
6.05144978e-01 1.04962245e-01 5.04594147e-01 -2.71750003e-01
-7.88753271e-01 -1.05116844e+00 -2.47046992e-01 5.93348801e-01
5.36924660e-01 -4.38383669e-01 -4.44918424e-01 -3.53659511e-01] | [11.44036865234375, 7.665599346160889] |
3b284a61-4f58-4e37-8d2d-6fc0ccd89913 | harnessing-the-power-of-text-image | 2304.10249 | null | https://arxiv.org/abs/2304.10249v1 | https://arxiv.org/pdf/2304.10249v1.pdf | Harnessing the Power of Text-image Contrastive Models for Automatic Detection of Online Misinformation | As growing usage of social media websites in the recent decades, the amount of news articles spreading online rapidly, resulting in an unprecedented scale of potentially fraudulent information. Although a plenty of studies have applied the supervised machine learning approaches to detect such content, the lack of gold standard training data has hindered the development. Analysing the single data format, either fake text description or fake image, is the mainstream direction for the current research. However, the misinformation in real-world scenario is commonly formed as a text-image pair where the news article/news title is described as text content, and usually followed by the related image. Given the strong ability of learning features without labelled data, contrastive learning, as a self-learning approach, has emerged and achieved success on the computer vision. In this paper, our goal is to explore the constrastive learning in the domain of misinformation identification. We developed a self-learning model and carried out the comprehensive experiments on a public data set named COSMOS. Comparing to the baseline classifier, our model shows the superior performance of non-matched image-text pair detection (approximately 10%) when the training data is insufficient. In addition, we observed the stability for contrsative learning and suggested the use of it offers large reductions in the number of training data, whilst maintaining comparable classification results. | ['Siwei Lyu', 'Xi Wu', 'Jinrong Hu', 'Bin Zhu', 'Shu Hu', 'Xin Wang', 'Peng Zheng', 'Hao Chen'] | 2023-04-19 | null | null | null | null | ['misinformation', 'self-learning'] | ['miscellaneous', 'natural-language-processing'] | [ 2.99816132e-01 1.45844713e-01 -3.18520665e-01 -1.83460951e-01
-6.16803944e-01 -2.95100749e-01 1.17095852e+00 2.99927771e-01
-4.21058595e-01 5.75629175e-01 2.16736607e-02 -6.60984442e-02
1.10588208e-01 -6.15456939e-01 -7.10827351e-01 -6.46267235e-01
2.47004583e-01 2.10878104e-01 3.98457617e-01 -1.53529972e-01
6.25456870e-01 4.15779911e-02 -1.71958053e+00 4.90792304e-01
8.75406086e-01 1.22053432e+00 3.75651009e-03 1.03219844e-01
-2.85052299e-01 8.80376339e-01 -6.11489773e-01 -8.58810127e-01
2.78974861e-01 -4.52804059e-01 -6.26756370e-01 3.97660494e-01
6.57030940e-01 -2.77740717e-01 -4.97436643e-01 1.22893202e+00
2.16034129e-01 -4.72499758e-01 6.86038613e-01 -1.46083796e+00
-5.22764444e-01 4.14822102e-01 -7.29510248e-01 3.13177943e-01
2.85032779e-01 -2.99326479e-01 7.89605200e-01 -7.70301700e-01
6.12748384e-01 1.14297760e+00 6.06015921e-01 2.31326997e-01
-8.65982890e-01 -6.26606703e-01 -4.21463251e-01 3.87938201e-01
-1.13744318e+00 -2.02425331e-01 7.87474334e-01 -5.52836239e-01
4.51124072e-01 1.50498748e-01 5.36637843e-01 1.46931779e+00
4.71658677e-01 9.21486795e-01 1.58504498e+00 -6.60916507e-01
-1.69930875e-01 9.15417552e-01 1.10283092e-01 5.80862761e-01
3.95899802e-01 2.05344006e-01 -5.46552598e-01 -2.62814194e-01
2.49042466e-01 1.53067663e-01 -9.41639021e-03 -3.85648817e-01
-9.71264541e-01 9.75125134e-01 2.40305468e-01 5.53727150e-01
3.09029333e-02 -4.34550613e-01 6.84121370e-01 5.51936567e-01
8.35038543e-01 4.35763925e-01 -5.07036299e-02 1.71636760e-01
-1.09620953e+00 2.67409384e-01 8.40691388e-01 5.58480799e-01
4.60339040e-01 -1.51051208e-01 3.05547655e-01 9.08325016e-01
6.75112009e-02 3.77199441e-01 1.10459065e+00 -3.15147370e-01
7.22807229e-01 9.41891849e-01 8.87450501e-02 -1.79806542e+00
-1.16737828e-01 -5.99894643e-01 -8.92011583e-01 7.63953775e-02
4.48553890e-01 1.89956382e-01 -5.06642759e-01 1.05764937e+00
2.13254541e-01 4.79552709e-02 3.58268470e-02 8.58618438e-01
4.76721913e-01 5.57426691e-01 -2.49457523e-01 -3.73943239e-01
1.30181324e+00 -8.89472842e-01 -8.05291295e-01 -2.68714875e-01
7.41441667e-01 -1.08277369e+00 6.77252293e-01 5.70006490e-01
-5.15722275e-01 -3.35183740e-01 -1.10306728e+00 2.42335722e-01
-7.86291182e-01 1.22708231e-01 3.73297155e-01 7.39518166e-01
-5.14954150e-01 6.14329338e-01 -2.65363783e-01 -5.99108636e-01
6.55343354e-01 -5.45242466e-02 -5.73524535e-01 -1.86032578e-01
-1.13808417e+00 9.87444758e-01 6.88446581e-01 -1.67002439e-01
-4.89170730e-01 -3.15653458e-02 -3.13595504e-01 -3.03172588e-01
4.66358215e-01 -1.04678981e-01 8.60914826e-01 -1.66254854e+00
-9.44344878e-01 1.32613862e+00 5.32710731e-01 -7.42509246e-01
1.14505172e+00 -1.05359271e-01 -9.34280276e-01 1.45755827e-01
3.48091930e-01 2.50610113e-01 1.57020140e+00 -1.32701206e+00
-8.94905090e-01 -4.65160996e-01 -3.37863773e-01 1.26396790e-01
-7.03259230e-01 5.27354032e-02 5.64534962e-02 -7.49556303e-01
4.75551456e-01 -8.61015499e-01 3.91504675e-01 -1.71129018e-01
-4.06395435e-01 -1.40953064e-01 1.18265140e+00 -6.77204967e-01
1.20563149e+00 -2.20702505e+00 -3.58170152e-01 -8.36274847e-02
2.61342734e-01 4.73091543e-01 3.77788901e-01 7.20437765e-01
-9.36691239e-02 9.86994803e-02 -2.78755069e-01 -1.49503067e-01
-2.99112946e-01 1.73191819e-02 -5.26226997e-01 7.94730783e-01
2.76670861e-03 4.49504554e-01 -9.24076319e-01 -8.42295587e-01
-1.85896941e-02 -2.98240539e-02 -8.82418305e-02 1.96110010e-01
5.02639711e-02 1.43918589e-01 -4.47867513e-01 6.27179146e-01
8.14574718e-01 -3.14103425e-01 5.75563684e-02 -2.65379399e-01
-2.20780924e-01 5.26346415e-02 -9.59657669e-01 1.00671172e+00
-1.95626430e-02 8.89382720e-01 -1.76156074e-01 -1.41558695e+00
9.01361823e-01 2.33902216e-01 4.04893190e-01 -8.99952769e-01
2.49510661e-01 5.07120967e-01 -1.39911547e-01 -9.59205270e-01
3.88293386e-01 -9.81063098e-02 -1.39083996e-01 4.24057245e-01
-1.70154013e-02 4.10120077e-02 -1.12653613e-01 1.57699928e-01
8.10604870e-01 2.77290288e-02 5.50007463e-01 3.62687139e-03
5.05552471e-01 3.61098111e-01 -1.73892267e-02 6.77048802e-01
-2.87332833e-01 4.08161014e-01 4.20608759e-01 -4.62518334e-01
-1.25114369e+00 -5.43916166e-01 -4.61146325e-01 8.25242341e-01
3.11755538e-01 -1.42652497e-01 -7.30025053e-01 -9.40975428e-01
2.24103883e-01 4.83210742e-01 -3.77528697e-01 -1.41057745e-01
-4.41782624e-01 -9.82654810e-01 7.22827673e-01 -3.27215880e-01
1.06381929e+00 -9.21456099e-01 -3.81082684e-01 1.31402835e-01
-1.74902022e-01 -1.14303923e+00 -2.29373559e-01 -1.22026272e-01
-9.13474739e-01 -1.24811447e+00 -4.55855519e-01 -8.78091216e-01
5.92183471e-01 6.27261341e-01 6.69497967e-01 2.32546136e-01
-3.69163305e-01 1.47678971e-01 -5.53216696e-01 -2.88253516e-01
-9.07090962e-01 -1.28318787e-01 -2.01168898e-02 5.27323186e-01
3.58349591e-01 -1.33465245e-01 -3.55897814e-01 3.15530747e-01
-1.16846168e+00 4.54856753e-02 8.57299507e-01 1.09109843e+00
-8.54594782e-02 2.68708944e-01 6.84468269e-01 -1.08125031e+00
5.15131176e-01 -8.68425310e-01 -3.73939753e-01 5.88595914e-03
-1.10631347e+00 -3.00743226e-02 5.82974613e-01 -3.31078380e-01
-9.75280106e-01 -2.02345461e-01 1.60834059e-01 -8.48612040e-02
-2.89447635e-01 4.91141587e-01 3.07932258e-01 -2.09920168e-01
8.96216273e-01 4.86735702e-01 4.24821675e-01 -4.27064955e-01
-3.39736938e-02 1.37898922e+00 2.94733018e-01 9.42717195e-02
9.46705341e-01 5.24154842e-01 -2.33675241e-01 -9.14743960e-01
-1.00807714e+00 -7.91326284e-01 -4.59271103e-01 -2.73263723e-01
4.46177900e-01 -7.17876375e-01 -2.36337125e-01 7.78131247e-01
-1.08157277e+00 5.53070962e-01 3.83141667e-01 4.11117882e-01
-3.21592063e-01 8.90239120e-01 -5.14264762e-01 -9.36309338e-01
-1.09126829e-01 -8.22789550e-01 6.08176708e-01 -2.94453621e-01
6.88689798e-02 -9.17413235e-01 -1.31159231e-01 8.77861321e-01
2.15194181e-01 2.23990604e-01 6.23358965e-01 -1.26840198e+00
-5.63126326e-01 -7.42507815e-01 -4.90337104e-01 7.34044969e-01
1.35268986e-01 -3.96310598e-01 -1.00556469e+00 -2.48039305e-01
4.26419556e-01 -5.16813219e-01 7.66936779e-01 -3.31306696e-01
8.67989659e-01 -6.40369236e-01 -4.03819412e-01 4.46019657e-02
1.49596286e+00 -1.05672538e-01 4.81217384e-01 7.30843663e-01
5.57975709e-01 7.72212625e-01 7.78975070e-01 3.34576517e-01
1.57015562e-01 6.20118439e-01 6.54901385e-01 5.94134144e-02
1.01597734e-01 -4.22307938e-01 2.49236703e-01 8.19177091e-01
2.69390792e-01 -3.18223625e-01 -8.67478490e-01 3.10612887e-01
-1.91032124e+00 -1.22599435e+00 -3.25067878e-01 2.23535872e+00
5.92394590e-01 3.37318182e-01 7.96628222e-02 4.42252517e-01
9.42462504e-01 2.16749355e-01 -1.99384511e-01 6.28495663e-02
-3.49298328e-01 -5.58749795e-01 8.13268423e-01 5.73282018e-02
-1.44269764e+00 8.28863800e-01 5.84487963e+00 1.09203029e+00
-1.29832304e+00 2.33591706e-01 6.93125427e-01 3.73163521e-01
1.78057939e-01 -1.68284148e-01 -7.31929123e-01 8.68725777e-01
6.84864044e-01 4.64342497e-02 1.68923140e-01 8.64359558e-01
2.98172235e-01 -2.90694594e-01 -5.58703184e-01 1.20577610e+00
7.00484276e-01 -1.21141934e+00 9.66568589e-02 1.09155253e-01
7.07431138e-01 -8.11166018e-02 3.13435569e-02 1.35373041e-01
-3.41995060e-01 -7.15359747e-01 7.33178139e-01 2.22830698e-01
3.90947282e-01 -4.35203075e-01 1.05473936e+00 9.32400286e-01
-2.82220781e-01 -4.31032240e-01 -1.91234380e-01 -9.88056362e-02
-2.86580294e-01 6.11069322e-01 -1.07567585e+00 5.01519561e-01
6.52014315e-01 8.51037204e-01 -8.34159553e-01 1.13833523e+00
1.72004133e-01 6.90378487e-01 5.29179536e-03 -2.18743354e-01
2.40065917e-01 -2.86191106e-01 6.27120495e-01 1.12118769e+00
1.65686056e-01 -3.75082582e-01 1.08891077e-01 4.69940186e-01
-1.95065420e-02 5.36761642e-01 -8.22395325e-01 -2.20157251e-01
1.05891973e-01 1.16350889e+00 -7.28207707e-01 -2.64452517e-01
-5.58684826e-01 1.17105734e+00 3.15929830e-01 -1.16732113e-01
-7.66373575e-01 2.98174862e-02 -1.79241627e-01 4.99985695e-01
2.62396745e-02 1.29505873e-01 -1.05657071e-01 -1.22053790e+00
1.29419327e-01 -1.08094406e+00 3.76159847e-01 -5.00420272e-01
-1.65449250e+00 4.54034537e-01 1.44267567e-02 -1.61720812e+00
-1.82773918e-02 -6.09434903e-01 -2.95235217e-01 3.60280663e-01
-1.46734738e+00 -1.25139380e+00 -3.34350675e-01 2.87429631e-01
5.53513944e-01 -6.17841125e-01 4.93190289e-01 3.50062191e-01
-4.18539256e-01 4.41171706e-01 5.42310655e-01 2.34049648e-01
1.04715931e+00 -8.95506442e-01 -1.93623662e-01 6.23884916e-01
4.97102700e-02 1.67327315e-01 9.86017942e-01 -8.46565425e-01
-1.09923017e+00 -7.85095215e-01 9.53858197e-01 -3.27180743e-01
1.01380324e+00 -2.17358634e-01 -8.96910965e-01 3.38106871e-01
2.07811296e-01 -5.01709804e-02 4.54833478e-01 -5.38114190e-01
-4.79309857e-01 -1.03092641e-01 -1.41354203e+00 3.24707419e-01
7.07838535e-01 -4.88462120e-01 -7.92549849e-01 6.82458699e-01
1.74077466e-01 -1.50815934e-01 -5.00799894e-01 -6.50306325e-03
5.06769300e-01 -1.28369892e+00 6.55146420e-01 -5.08526504e-01
6.76538229e-01 2.87990365e-02 -2.59654596e-03 -1.06054366e+00
1.33748809e-02 -1.93736419e-01 8.17986131e-02 1.23918509e+00
1.98447198e-01 -7.78394878e-01 8.96320164e-01 1.27128502e-02
9.30428728e-02 -4.54291642e-01 -9.82247472e-01 -7.53792405e-01
-1.53310254e-01 -4.20852266e-02 -1.25780270e-01 1.32944536e+00
1.29930213e-01 3.10683876e-01 -6.72242463e-01 -6.16207458e-02
8.53900373e-01 3.45028415e-02 5.91606140e-01 -1.29394913e+00
-1.14684470e-01 -2.94184536e-01 -7.36530840e-01 -7.95178473e-01
7.75210559e-02 -9.04882133e-01 -3.35821390e-01 -1.01100791e+00
2.79273450e-01 -3.54671597e-01 8.71037617e-02 -1.99605618e-02
2.31740228e-03 3.05578887e-01 -9.08508971e-02 8.11388016e-01
-4.39093679e-01 2.04614416e-01 1.21755350e+00 -2.87402660e-01
2.79881030e-01 2.44983241e-01 -2.91542649e-01 8.17862034e-01
7.03462243e-01 -7.27429509e-01 -1.76758707e-01 1.99425951e-01
2.78015703e-01 1.87073406e-02 5.54944038e-01 -9.87342894e-01
2.26971686e-01 7.12860972e-02 2.66692281e-01 -5.61962247e-01
4.92753871e-02 -1.08247614e+00 -1.09873690e-01 6.27269745e-01
-3.24543774e-01 -5.62560223e-02 -3.45830739e-01 8.97641718e-01
-4.65659440e-01 -5.64696074e-01 9.53169823e-01 -4.22667563e-01
-6.86556995e-01 3.93588841e-02 -2.60010093e-01 -4.68475930e-02
1.08999717e+00 -4.63493645e-01 -3.62314761e-01 -2.99081266e-01
-3.25687110e-01 -3.07695955e-01 4.70490277e-01 5.77583075e-01
4.53582436e-01 -1.06914949e+00 -6.80843115e-01 1.19912282e-01
3.97717416e-01 -6.39056861e-01 6.27221353e-03 9.17675734e-01
-5.93286872e-01 3.16289335e-01 -2.35369846e-01 -5.94886065e-01
-1.24950230e+00 7.25666285e-01 2.95120239e-01 -1.20873503e-01
-5.30949473e-01 1.90012127e-01 -2.91198045e-01 -1.44024983e-01
1.29058480e-01 3.38722676e-01 -4.21628684e-01 4.26809192e-01
4.00175840e-01 5.11646211e-01 2.13579133e-01 -9.44768548e-01
4.98264283e-02 1.50539607e-01 -3.47036600e-01 9.21051651e-02
1.10533237e+00 -2.57371187e-01 -3.56143326e-01 5.79501987e-01
1.42755151e+00 -8.88231993e-02 -8.27320993e-01 -2.39872187e-01
4.68938410e-01 -8.14228177e-01 1.21103041e-01 -6.59440398e-01
-6.25900745e-01 4.41644341e-01 8.78948867e-01 9.10878301e-01
5.63151538e-01 -5.65715916e-02 6.80510342e-01 2.66538948e-01
3.65703195e-01 -1.29423594e+00 4.54162270e-01 2.32349262e-01
8.05947244e-01 -1.84386694e+00 1.67844325e-01 -5.31531930e-01
-6.89577162e-01 1.17747533e+00 3.37255061e-01 -3.49148631e-01
6.02440000e-01 -4.36285585e-02 1.10615209e-01 -2.61559069e-01
-2.58705109e-01 1.37066171e-01 1.99552700e-01 3.90829802e-01
4.04861838e-01 -8.40001404e-02 -7.06736386e-01 2.41496608e-01
-2.02174768e-01 -1.30568489e-01 6.43757939e-01 8.61012638e-01
-6.10993087e-01 -1.10302353e+00 -6.20030642e-01 7.92879105e-01
-7.87737906e-01 5.05012758e-02 -4.35479015e-01 9.41564441e-01
2.55481184e-01 9.61096823e-01 -3.71251851e-02 -2.88871557e-01
1.23933226e-01 1.78031638e-01 2.27168337e-01 -4.10598993e-01
-6.37313664e-01 -1.79040566e-01 1.00769483e-01 -2.25281969e-01
-6.21868849e-01 -7.43581712e-01 -7.06291676e-01 -2.54356980e-01
-6.54901564e-01 2.62132697e-02 8.29275191e-01 1.19055045e+00
1.33873150e-01 -1.61872536e-01 1.02001441e+00 -4.71087098e-01
-9.70912158e-01 -1.02563548e+00 -5.78604221e-01 1.00982761e+00
4.06854033e-01 -6.85783327e-01 -8.70556533e-01 2.05011949e-01] | [8.140314102172852, 10.258550643920898] |
a4eff907-37a8-46a5-be2e-047c93952667 | convabuse-data-analysis-and-benchmarks-for | 2109.09483 | null | https://arxiv.org/abs/2109.09483v1 | https://arxiv.org/pdf/2109.09483v1.pdf | ConvAbuse: Data, Analysis, and Benchmarks for Nuanced Abuse Detection in Conversational AI | We present the first English corpus study on abusive language towards three conversational AI systems gathered "in the wild": an open-domain social bot, a rule-based chatbot, and a task-based system. To account for the complexity of the task, we take a more `nuanced' approach where our ConvAI dataset reflects fine-grained notions of abuse, as well as views from multiple expert annotators. We find that the distribution of abuse is vastly different compared to other commonly used datasets, with more sexually tinted aggression towards the virtual persona of these systems. Finally, we report results from bench-marking existing models against this data. Unsurprisingly, we find that there is substantial room for improvement with F1 scores below 90%. | ['Verena Rieser', 'Gavin Abercrombie', 'Amanda Cercas Curry'] | 2021-09-20 | null | null | null | null | ['abuse-detection'] | ['natural-language-processing'] | [-2.26991296e-01 4.55452502e-01 7.11056963e-02 -3.61987174e-01
-6.01202309e-01 -9.60409820e-01 1.11732006e+00 -2.27573082e-01
-5.01056075e-01 7.96859384e-01 6.44604743e-01 -1.93752140e-01
5.73885441e-02 -1.65726408e-01 8.73747617e-02 -5.71255505e-01
-1.52181117e-02 8.90909195e-01 3.71100962e-01 -8.81449699e-01
4.69033301e-01 2.23371357e-01 -6.56884551e-01 3.85846734e-01
5.46466708e-01 4.16443676e-01 -5.07839978e-01 7.49273837e-01
7.34023154e-02 1.61311436e+00 -1.33360720e+00 -1.29698789e+00
1.06437728e-01 -4.32456970e-01 -1.52684748e+00 -3.68249081e-02
5.70695758e-01 -4.10019010e-01 -4.07923877e-01 9.45012689e-01
3.16823184e-01 -2.09045913e-02 6.15970194e-01 -1.50734746e+00
-5.14338315e-01 8.87810349e-01 -5.34787357e-01 6.72465801e-01
5.18800974e-01 8.55753541e-01 1.17701256e+00 -1.28239304e-01
1.12437820e+00 1.46667194e+00 7.27669001e-01 9.15344059e-01
-9.91933644e-01 -5.46069145e-01 -2.25412890e-01 3.78465578e-02
-8.50236535e-01 -7.08829105e-01 7.74794579e-01 -8.76484036e-01
1.00531065e+00 2.81296194e-01 3.97491544e-01 1.88977933e+00
-4.01222631e-02 4.73330081e-01 9.41866219e-01 -8.84242076e-03
-1.54124964e-02 2.44670108e-01 3.73857021e-01 5.48490167e-01
2.14511037e-01 -2.23370939e-01 -6.04638040e-01 -9.53728080e-01
4.01398808e-01 -6.11724734e-01 -7.32916147e-02 -2.13235587e-01
-5.94969988e-01 1.09501278e+00 2.28811115e-01 9.68954563e-01
-2.70282596e-01 1.28977105e-01 9.93219197e-01 4.49753910e-01
6.98047519e-01 8.99983704e-01 -3.19528461e-01 -1.00740111e+00
-3.32334548e-01 5.24976492e-01 1.44906938e+00 6.02792323e-01
1.67658523e-01 3.91291156e-02 1.71648264e-01 1.13055503e+00
1.66834429e-01 7.75950029e-02 3.48046362e-01 -1.44855130e+00
6.48845017e-01 4.92029369e-01 1.17826149e-01 -1.12125707e+00
-2.42094561e-01 7.09819570e-02 -8.59632641e-02 1.29458398e-01
1.08118415e+00 -5.28743982e-01 -2.76970536e-01 1.86570227e+00
1.44636780e-01 -2.35628128e-01 -1.33960336e-01 7.29822636e-01
5.63861966e-01 1.17435679e-01 1.88644275e-01 -2.33371019e-01
1.18929958e+00 -7.91603267e-01 -7.66698658e-01 -5.03289461e-01
9.42694783e-01 -3.45609814e-01 8.48987222e-01 5.07133365e-01
-1.03694725e+00 4.14645970e-01 -8.76761556e-01 -1.66454941e-01
-2.45452002e-01 -6.30497098e-01 5.46049595e-01 9.24830437e-01
-7.66465306e-01 5.27700305e-01 -5.96333802e-01 -6.09620094e-01
5.60493350e-01 -1.01282880e-01 -3.35622609e-01 2.10356161e-01
-9.52082992e-01 1.64667773e+00 -1.09114982e-01 -3.52418214e-01
-9.25289154e-01 -4.52263057e-01 -6.58696294e-01 -2.63366371e-01
6.59366965e-01 -8.87957364e-02 1.66411269e+00 -9.50391650e-01
-1.34738052e+00 1.31537783e+00 3.60200256e-01 -3.97948831e-01
7.74859130e-01 -1.02401175e-01 -2.44340032e-01 -3.34327146e-02
1.72868371e-01 -5.16233221e-02 5.50243497e-01 -1.26590657e+00
-1.54992163e-01 -6.23758495e-01 6.20556712e-01 1.66925997e-01
-2.83597171e-01 8.80154967e-01 3.37389857e-01 -3.30715597e-01
-7.05071330e-01 -8.49406838e-01 -7.87711330e-03 -2.67895311e-01
-4.55304801e-01 -6.06754124e-01 1.00667417e+00 -7.63890147e-01
1.25556922e+00 -2.14963460e+00 4.19155061e-01 -3.98904607e-02
8.20226490e-01 5.14934897e-01 2.07425430e-01 7.48256922e-01
1.90079987e-01 5.05733192e-01 -3.44885647e-01 -4.62968349e-01
1.82915088e-02 4.40304518e-01 -7.45157450e-02 5.15805602e-01
4.54658493e-02 6.86265349e-01 -1.13204205e+00 -6.40410900e-01
-6.39205575e-02 9.84910503e-03 -3.73363107e-01 2.65397787e-01
-1.16536237e-01 3.71592402e-01 -2.86497742e-01 3.36501658e-01
-1.59218498e-02 -8.53933394e-03 3.42665911e-01 5.54680824e-01
6.92069530e-04 7.70525634e-01 -2.28186697e-01 1.52183592e+00
-1.69346318e-01 9.27557588e-01 7.57992923e-01 -4.99710321e-01
6.46475673e-01 4.62890446e-01 1.28490046e-01 -9.81120914e-02
6.29282951e-01 1.77426800e-01 7.98114955e-01 -6.68575704e-01
4.53644454e-01 -2.18758911e-01 -3.70563447e-01 7.46201396e-01
2.83208996e-01 -4.49645936e-01 1.05415799e-01 6.41768754e-01
1.61197162e+00 -2.22155288e-01 4.85023439e-01 -3.60870570e-01
2.45673612e-01 4.56738889e-01 4.11173820e-01 6.24784291e-01
-9.07331765e-01 1.72035232e-01 1.16233528e+00 -4.23194230e-01
-8.94883633e-01 -6.41407847e-01 -5.42897265e-03 1.32384896e+00
-3.55797321e-01 -6.98983550e-01 -1.09184277e+00 -1.04362369e+00
-2.39807963e-01 8.30862820e-01 -6.74038887e-01 -3.42059731e-02
-6.77574813e-01 -4.27533537e-01 1.34317565e+00 2.23848864e-01
3.37200344e-01 -1.27916777e+00 -7.09556878e-01 -5.22654084e-03
-3.67407769e-01 -1.25625110e+00 -1.72968432e-01 -1.65481165e-01
-2.50772238e-01 -1.28855574e+00 -1.04163634e-02 -2.09526390e-01
-1.81996018e-01 -1.29829526e-01 1.36403453e+00 5.59798837e-01
-1.47175997e-01 1.48502558e-01 -5.63553095e-01 -4.96641159e-01
-1.00441647e+00 1.60068601e-01 4.22660038e-02 -3.03362876e-01
6.97467923e-01 -8.28775227e-01 4.46071401e-02 1.92789838e-01
-4.19748485e-01 -4.43499386e-01 4.18806002e-02 6.16860330e-01
-1.06379092e+00 -7.06628263e-01 4.81344044e-01 -1.38038373e+00
1.36618519e+00 -9.46113467e-01 7.18884766e-02 -1.54446647e-01
-1.61037281e-01 -5.73236823e-01 6.17149651e-01 -3.84755403e-01
-1.06329131e+00 -6.76584363e-01 -1.20211676e-01 -2.58521378e-01
-4.51720148e-01 -4.75050434e-02 1.80459395e-01 7.64587000e-02
1.33196962e+00 -3.87870610e-01 2.17868239e-01 -4.82289016e-01
3.33142906e-01 1.19293022e+00 5.51523805e-01 -8.66807640e-01
6.17968380e-01 1.99450418e-01 -6.44964516e-01 -1.11142063e+00
-6.22009337e-01 -2.73181111e-01 -5.21153212e-01 -3.19057763e-01
9.38287437e-01 -3.79927218e-01 -1.03292918e+00 5.92459857e-01
-1.48813820e+00 -6.98990583e-01 -7.07085580e-02 -1.33639887e-01
-6.17004454e-01 4.98594046e-01 -1.24462569e+00 -1.02894568e+00
-1.74079210e-01 -9.89443600e-01 5.34390569e-01 -1.34895161e-01
-1.16358745e+00 -1.02406502e+00 6.75117075e-01 9.28253353e-01
5.25917828e-01 3.15963209e-01 7.66790748e-01 -1.53546441e+00
3.76599193e-01 -1.07521273e-01 -8.08839872e-02 1.34825215e-01
1.40190097e-02 1.52476102e-01 -1.11079705e+00 1.08948007e-01
1.96151957e-01 -1.07606721e+00 7.09270239e-02 -5.11768043e-01
3.39360803e-01 -5.19642770e-01 -1.48710221e-01 -2.22276732e-01
6.41758561e-01 3.85072142e-01 4.91379917e-01 1.46221384e-01
6.57239258e-01 8.51724803e-01 1.76123202e-01 5.31640708e-01
3.24293196e-01 7.01746821e-01 2.44818538e-01 5.68063855e-01
1.43696159e-01 -1.14120007e-01 3.17747533e-01 4.42354381e-01
-3.38825911e-01 -3.27018142e-01 -1.24629557e+00 5.91154873e-01
-1.88228679e+00 -1.35635304e+00 -1.93851918e-01 1.42032707e+00
8.39552701e-01 2.28529856e-01 7.41411090e-01 -9.75173563e-02
5.60732305e-01 6.36873603e-01 -2.31501013e-01 -1.17414033e+00
3.17628294e-01 3.85570452e-02 -6.76866770e-02 7.75342643e-01
-7.07029104e-01 1.08557606e+00 7.21461964e+00 6.96853936e-01
-6.60713196e-01 5.20588696e-01 4.89128679e-01 -2.57222444e-01
8.36678892e-02 -1.19677164e-01 -6.03511222e-02 6.73686326e-01
1.18874574e+00 -2.00935006e-01 6.17119253e-01 9.04829383e-01
-2.06488803e-01 -1.22679614e-01 -1.25374663e+00 6.06413841e-01
4.63900983e-01 -9.13538158e-01 -5.39526343e-01 3.14296186e-01
1.93025395e-01 3.38154882e-01 -4.30873215e-01 4.50154603e-01
1.29219365e+00 -1.12979007e+00 4.97650951e-01 -9.12470091e-03
3.13335836e-01 -3.49156082e-01 5.84130406e-01 8.54440510e-01
-1.03589728e-01 -2.26763010e-01 5.67265972e-02 -6.03241086e-01
2.65231401e-01 -2.04053968e-01 -8.47699463e-01 6.22580573e-02
7.73208618e-01 5.14407992e-01 -4.71575886e-01 2.10900977e-01
-1.70898944e-01 7.44586647e-01 -2.92170197e-01 -4.04006183e-01
4.22252655e-01 -3.36734861e-01 9.23328221e-01 1.27229238e+00
-7.05734730e-01 4.78829622e-01 1.05309799e-01 8.34732175e-01
-1.13099553e-01 -3.06314468e-01 -1.24292600e+00 -3.62769425e-01
4.76424128e-01 1.38166070e+00 -9.25007761e-02 -2.17969894e-01
-2.78526604e-01 8.95754039e-01 8.12622547e-01 -1.38268948e-01
-6.95731938e-01 -7.57061392e-02 9.00833011e-01 -5.59202917e-02
-3.21322739e-01 -2.07149073e-01 -1.90993711e-01 -1.08690166e+00
-2.32301518e-01 -1.25031757e+00 3.51525605e-01 -7.76573896e-01
-1.88676095e+00 8.07865322e-01 1.68113708e-01 -4.88315970e-01
-1.00726867e+00 -5.03435969e-01 -8.09492886e-01 5.31530201e-01
-3.42480600e-01 -1.01142859e+00 6.96504489e-03 4.40059096e-01
4.47633356e-01 -3.56651574e-01 7.98207164e-01 8.91677961e-02
-7.15513766e-01 4.98800844e-01 -5.20875573e-01 7.73298621e-01
6.28924668e-01 -1.04693770e+00 4.29602921e-01 5.42393804e-01
1.10694267e-01 6.81903541e-01 1.13258648e+00 -5.75460613e-01
-9.92951393e-01 -2.12347552e-01 6.32734597e-01 -1.51640069e+00
1.36428297e+00 -4.98664141e-01 -8.69117379e-01 1.23582673e+00
7.75734544e-01 -3.04015964e-01 6.78086400e-01 5.78596711e-01
-8.47114444e-01 6.14694834e-01 -1.56402874e+00 6.53761744e-01
1.46198750e+00 -6.65597081e-01 -1.16992223e+00 4.25531685e-01
4.24809396e-01 -4.23001856e-01 -7.96304345e-01 -1.59745812e-01
4.75885868e-01 -1.22021031e+00 3.33824933e-01 -1.28805852e+00
9.27373707e-01 5.31248212e-01 -1.07548140e-01 -1.61707652e+00
-5.97538650e-02 -8.85722041e-01 1.28919454e-02 1.43189812e+00
3.42349559e-01 -6.00589871e-01 6.16667211e-01 1.13252509e+00
-3.44859362e-02 -5.15343130e-01 -1.08379865e+00 -6.04956508e-01
6.23118401e-01 -3.86776745e-01 7.30567351e-02 1.59196579e+00
1.04193437e+00 9.61328804e-01 -4.84790415e-01 -4.43926245e-01
2.67860949e-01 -4.99788463e-01 1.04143834e+00 -1.27788258e+00
-3.95647049e-01 -5.22132039e-01 -5.06473184e-01 -5.57592154e-01
3.82756770e-01 -5.11908174e-01 -5.33457398e-02 -8.68933320e-01
4.35151428e-01 -2.10564375e-01 6.05674386e-01 4.94563967e-01
1.98249280e-01 5.01310468e-01 3.96162748e-01 1.74989268e-01
-5.93627632e-01 9.78974849e-02 9.60381508e-01 1.53653160e-01
-1.59828346e-02 -3.90175194e-01 -9.80817437e-01 1.18181646e+00
8.17629635e-01 -5.00833750e-01 -2.62993544e-01 -5.18480122e-01
6.47626817e-02 7.76686221e-02 2.39896849e-01 -4.80219185e-01
7.29148835e-02 -5.03144622e-01 -4.41654503e-01 3.23187977e-01
8.05767238e-01 -6.20216370e-01 -2.39386886e-01 3.17029178e-01
-3.83215368e-01 3.41212228e-02 -1.51033074e-01 2.83237636e-01
2.14446068e-01 -4.33674902e-01 7.40905583e-01 -4.29632962e-01
-2.24947378e-01 -1.62828714e-01 -7.91088462e-01 7.44082332e-01
1.09154034e+00 9.40416940e-03 -1.01815188e+00 -8.82216334e-01
-5.23248732e-01 1.98068663e-01 5.70872188e-01 4.06545311e-01
-1.02022029e-01 -6.52780116e-01 -8.33732784e-01 -4.50723231e-01
1.36613593e-01 -4.73955929e-01 -2.03316212e-01 7.43845940e-01
-5.18859267e-01 3.83451656e-02 -3.16953808e-01 -1.84605271e-01
-1.33210671e+00 2.15613082e-01 5.54110229e-01 -2.20744520e-01
-4.07015860e-01 7.26278663e-01 -2.03753740e-01 -6.28984153e-01
-5.88466041e-02 3.79687488e-01 -1.63716719e-01 -5.05520105e-02
5.79619050e-01 6.29959941e-01 -1.80886209e-01 -1.00671864e+00
-4.12972152e-01 -2.62873590e-01 -3.64098012e-01 -4.38874424e-01
1.30405879e+00 1.58860534e-01 -3.74609292e-01 8.32661808e-01
9.08252358e-01 3.49364102e-01 -8.43119740e-01 -1.04824737e-01
2.03527972e-01 -6.97235227e-01 -6.37355506e-01 -9.78809953e-01
-6.87001050e-01 6.86151028e-01 -3.37103337e-01 8.62617314e-01
1.49826795e-01 2.89365470e-01 5.64832449e-01 5.04863739e-01
6.57529473e-01 -1.02107203e+00 1.74024239e-01 7.91839123e-01
1.09363806e+00 -1.06554472e+00 -2.84123626e-02 -5.14141023e-01
-1.19979358e+00 5.37501633e-01 1.26394713e+00 -3.61095041e-01
2.31697604e-01 4.17599201e-01 4.35500801e-01 -8.85416627e-01
-1.07402802e+00 1.33587688e-01 -4.98365015e-01 6.75013542e-01
5.19855380e-01 -5.02997264e-02 -5.12469888e-01 6.99375033e-01
-6.72570288e-01 -4.49162185e-01 8.80079925e-01 8.03737938e-01
-1.90520406e-01 -8.35466564e-01 -2.32894663e-02 3.35801303e-01
-7.31887758e-01 2.80250967e-01 -1.69328189e+00 1.18188357e+00
-1.85172558e-01 1.32250142e+00 9.07597020e-02 -6.39191747e-01
2.46796250e-01 4.57201958e-01 3.90647471e-01 -8.26868176e-01
-1.37362969e+00 -7.02764153e-01 1.14947069e+00 -6.36091232e-01
-2.50398606e-01 -6.46711528e-01 -6.92877173e-01 -1.01038897e+00
-3.81189167e-01 1.45603120e-01 2.13283464e-01 1.41499007e+00
1.37740165e-01 -1.18213795e-01 3.45628828e-01 -6.19126022e-01
-6.74220681e-01 -1.53563344e+00 -3.42725813e-01 1.01319206e+00
2.20545214e-02 -6.09604836e-01 -6.95762455e-01 -4.83538628e-01] | [8.664158821105957, 10.403108596801758] |
b8f60c6a-8d58-4a7f-8e4d-f8d0fa0bc0b4 | d2df2wod-learning-object-proposals-for-weakly | 2212.01376 | null | https://arxiv.org/abs/2212.01376v1 | https://arxiv.org/pdf/2212.01376v1.pdf | D2DF2WOD: Learning Object Proposals for Weakly-Supervised Object Detection via Progressive Domain Adaptation | Weakly-supervised object detection (WSOD) models attempt to leverage image-level annotations in lieu of accurate but costly-to-obtain object localization labels. This oftentimes leads to substandard object detection and localization at inference time. To tackle this issue, we propose D2DF2WOD, a Dual-Domain Fully-to-Weakly Supervised Object Detection framework that leverages synthetic data, annotated with precise object localization, to supplement a natural image target domain, where only image-level labels are available. In its warm-up domain adaptation stage, the model learns a fully-supervised object detector (FSOD) to improve the precision of the object proposals in the target domain, and at the same time learns target-domain-specific and detection-aware proposal features. In its main WSOD stage, a WSOD model is specifically tuned to the target domain. The feature extractor and the object proposal generator of the WSOD model are built upon the fine-tuned FSOD model. We test D2DF2WOD on five dual-domain image benchmarks. The results show that our method results in consistently improved object detection and localization compared with state-of-the-art methods. | ['Vladimir Pavlovic', 'Ricardo Guerrero', 'Yuting Wang'] | 2022-12-02 | null | null | null | null | ['weakly-supervised-object-detection'] | ['computer-vision'] | [ 2.34965533e-01 1.29389480e-01 -3.25541198e-01 -3.61674964e-01
-1.16784084e+00 -5.95619977e-01 7.15066314e-01 1.08640730e-01
-4.33860570e-01 3.41356188e-01 -2.48488605e-01 -3.28551307e-02
4.10429657e-01 -6.07043445e-01 -9.80472505e-01 -6.75378680e-01
1.96895406e-01 6.42897785e-01 9.69215274e-01 2.00249314e-01
7.08829015e-02 3.73892903e-01 -1.51810849e+00 4.23497975e-01
4.21799898e-01 1.28447366e+00 6.25239015e-01 6.49704099e-01
-4.22120467e-03 6.84580207e-01 -5.03918588e-01 -1.17869124e-01
5.31408846e-01 -9.14635137e-02 -6.07631564e-01 2.96498448e-01
7.87164927e-01 -5.94144106e-01 -2.65026212e-01 1.10654438e+00
4.75845963e-01 -3.16906661e-01 7.74124026e-01 -1.31217372e+00
-5.95226467e-01 3.30776036e-01 -9.52320218e-01 3.41115743e-01
1.07648849e-01 6.66965008e-01 8.94757092e-01 -1.47354972e+00
6.67559206e-01 1.36363900e+00 5.97492039e-01 4.72039789e-01
-1.65136552e+00 -7.42865145e-01 2.66607374e-01 -2.37840414e-01
-1.59379148e+00 -5.20443320e-01 6.02741539e-01 -4.56996232e-01
7.18689084e-01 -1.97524697e-01 1.55963048e-01 1.15127134e+00
-1.75955594e-01 1.17799425e+00 1.02431941e+00 -4.06802416e-01
2.12469980e-01 3.86563063e-01 3.89716625e-02 7.83541739e-01
5.23245573e-01 3.74170661e-01 -4.42848831e-01 -1.15728080e-01
7.45919764e-01 2.11967882e-02 2.06034735e-01 -7.64899611e-01
-1.33302629e+00 5.21585107e-01 7.71416485e-01 1.38358753e-02
-3.88206601e-01 4.50020768e-02 2.99241662e-01 9.52795669e-02
5.04142642e-01 3.98341388e-01 -5.52109480e-01 4.49783325e-01
-9.82794702e-01 5.44103801e-01 4.84866679e-01 1.15832710e+00
8.28882039e-01 -2.27194831e-01 -6.84456468e-01 6.01601541e-01
3.38865906e-01 7.33581424e-01 3.38868089e-02 -6.98253632e-01
3.81364197e-01 8.93893242e-01 4.12063658e-01 -5.71700752e-01
-2.32078612e-01 -7.56683111e-01 -3.11947495e-01 4.84155059e-01
6.41288579e-01 1.98480576e-01 -1.19636929e+00 1.70247483e+00
6.88457489e-01 2.20432535e-01 -9.09160171e-03 1.18168175e+00
8.51861537e-01 4.71558064e-01 2.77433991e-01 3.51057321e-01
1.44160366e+00 -1.16739905e+00 -4.04424965e-02 -6.15825653e-01
6.20117903e-01 -7.14129090e-01 1.27399802e+00 1.25161231e-01
-9.18862402e-01 -7.99759030e-01 -8.35319340e-01 -1.16318800e-01
-1.79685518e-01 7.83611834e-01 2.11820543e-01 2.80101746e-01
-6.48439109e-01 9.64434743e-02 -7.30673492e-01 -3.81821245e-01
9.30099010e-01 1.01382561e-01 -4.16525185e-01 -3.01253974e-01
-5.86493313e-01 8.60938132e-01 7.98672080e-01 -2.87574470e-01
-1.69325113e+00 -8.00458848e-01 -7.86086082e-01 -1.65844694e-01
8.11396956e-01 -5.24101317e-01 1.46571052e+00 -8.93338919e-01
-9.35764372e-01 1.36980700e+00 9.80829149e-02 -5.40818155e-01
6.88638866e-01 -8.95025209e-02 -8.93313289e-02 8.31901878e-02
5.62955201e-01 1.08295882e+00 1.24329448e+00 -1.40646362e+00
-1.09615159e+00 -3.92394274e-01 -1.69620162e-03 2.00291313e-02
-1.13685578e-02 -1.43310919e-01 -7.76679218e-01 -6.39130950e-01
-1.34525653e-02 -8.07756901e-01 -1.01799890e-01 5.71285009e-01
-6.71935499e-01 -5.26872218e-01 1.25748694e+00 -2.10647404e-01
6.56893432e-01 -2.27084899e+00 -1.41383305e-01 -1.64289996e-01
3.01371694e-01 4.51607138e-01 -4.96226281e-01 -8.44715163e-02
1.50354013e-01 -4.71261322e-01 -1.10114738e-01 -6.81829929e-01
7.85756186e-02 1.48340106e-01 -3.54033172e-01 6.54754639e-01
8.25521111e-01 1.04106951e+00 -1.11913371e+00 -7.14596093e-01
2.42567554e-01 8.45316201e-02 -4.54372644e-01 4.78628218e-01
-6.19410336e-01 2.87709713e-01 -5.77921927e-01 9.49388027e-01
7.84578800e-01 -4.13438857e-01 -2.15531901e-01 -3.60531658e-01
8.75366502e-04 1.62727430e-01 -1.29902637e+00 1.74733484e+00
-3.13775241e-01 3.46884429e-01 -6.74951226e-02 -8.50736260e-01
8.67138147e-01 -2.44691819e-01 1.14611261e-01 -7.93917537e-01
-5.91074489e-02 3.46651852e-01 -1.46409765e-01 -3.52933586e-01
2.76959032e-01 7.26145059e-02 -1.27046287e-01 4.04002339e-01
5.14906764e-01 -1.52906239e-01 1.75507560e-01 4.74042743e-01
1.23302126e+00 4.67595279e-01 3.90649408e-01 -2.59993792e-01
5.21839082e-01 2.25622416e-01 5.20386517e-01 1.12931478e+00
-2.79596180e-01 7.07196116e-01 4.32115853e-01 -4.55032647e-01
-1.29742324e+00 -1.35306036e+00 -2.83928782e-01 1.41202855e+00
4.07941878e-01 -2.25053523e-02 -6.95629478e-01 -1.28230703e+00
3.39208215e-01 7.08408654e-01 -7.20229626e-01 -2.48313710e-01
-1.60516798e-01 -4.49640900e-01 6.02508008e-01 5.87000489e-01
5.44160962e-01 -8.87692332e-01 -7.02907383e-01 1.08553074e-01
2.34975860e-01 -1.40648663e+00 -5.98747730e-01 4.84334469e-01
-5.18390775e-01 -1.09408295e+00 -8.35519075e-01 -8.69445324e-01
9.12626386e-01 4.26546067e-01 1.17341495e+00 -8.15438628e-02
-6.68341100e-01 7.80994669e-02 -2.23717034e-01 -6.40384734e-01
-5.35908759e-01 5.53242713e-02 1.67016670e-01 1.07110806e-01
6.49490833e-01 1.47960363e-02 -6.90094233e-01 5.62332153e-01
-8.83040488e-01 -5.55003295e-04 9.51841652e-01 9.18855250e-01
7.45869100e-01 -2.95070142e-01 5.40774107e-01 -7.64327765e-01
-1.29226232e-02 -3.44421595e-01 -1.10527718e+00 1.72564566e-01
-5.16605020e-01 1.92782804e-01 3.15749586e-01 -5.96607924e-01
-9.41544771e-01 6.50444269e-01 1.83810696e-01 -6.39715970e-01
-3.44681352e-01 -1.48721024e-01 -2.18432218e-01 -1.64912328e-01
1.20970547e+00 3.36370498e-01 -1.42227069e-01 -5.84116340e-01
3.46141875e-01 6.53749943e-01 7.82549977e-01 -4.89758074e-01
1.12740684e+00 6.01792812e-01 -3.72304827e-01 -4.25892740e-01
-1.48434901e+00 -7.12199628e-01 -5.85929036e-01 -2.33470239e-02
6.77766502e-01 -1.25412762e+00 -2.10494533e-01 5.68965495e-01
-1.17246604e+00 -6.00236773e-01 -6.77800477e-01 2.67331481e-01
-4.13079232e-01 -1.02957867e-01 -2.59755254e-01 -7.70447910e-01
1.14693642e-02 -1.09564960e+00 1.88495326e+00 5.59776165e-02
1.63216472e-01 -4.89671946e-01 -2.27518007e-01 1.67974383e-01
1.42777249e-01 1.51682168e-01 4.07899618e-01 -7.88594306e-01
-7.41895735e-01 -4.80148733e-01 -8.26959550e-01 5.23263454e-01
-1.51966527e-01 -3.60287398e-01 -1.19327056e+00 -3.24786127e-01
-3.87881488e-01 -8.03822815e-01 1.08593571e+00 2.58239448e-01
1.14510131e+00 -1.38596863e-01 -6.49604201e-01 3.52977514e-01
1.35048175e+00 -5.16779006e-01 8.19756016e-02 1.48337185e-01
4.46492761e-01 3.67433637e-01 9.05669034e-01 3.21058661e-01
2.52639800e-01 9.61603642e-01 5.82890928e-01 -2.96407789e-01
-7.12033987e-01 -6.01257026e-01 4.22094405e-01 -4.14405972e-01
4.64046359e-01 -1.58609256e-01 -8.45849395e-01 7.78492510e-01
-1.70466149e+00 -6.96859658e-01 -1.17345676e-02 2.04007816e+00
9.15384412e-01 6.24908566e-01 5.42151392e-01 -1.03606798e-01
7.13259399e-01 -3.26362252e-02 -8.31191540e-01 3.65555108e-01
-8.04436132e-02 -1.84518471e-01 7.87908196e-01 1.85401291e-01
-1.27408779e+00 1.17008650e+00 5.41762066e+00 1.05237389e+00
-1.01558948e+00 4.70160127e-01 4.59684342e-01 -1.17102154e-01
2.21591309e-01 -4.51126173e-02 -1.30494928e+00 4.16900635e-01
3.54967147e-01 1.44057840e-01 -1.53039638e-02 1.36731446e+00
2.39463039e-02 -2.54287183e-01 -1.41027153e+00 8.64629030e-01
-1.28998056e-01 -1.33263874e+00 -4.79455180e-02 -1.27567416e-02
7.40421712e-01 3.12484622e-01 1.10096522e-01 4.68018025e-01
2.96178162e-01 -5.92730403e-01 1.14212537e+00 8.33835155e-02
9.84896421e-01 -3.81713301e-01 6.21907830e-01 7.36379683e-01
-1.28328538e+00 -2.47976080e-01 -4.24402952e-01 1.24995120e-01
5.07318825e-02 6.99388146e-01 -1.37654376e+00 -3.06285499e-03
7.83885539e-01 5.53587258e-01 -7.13316858e-01 1.02971566e+00
-4.24711257e-01 6.19310915e-01 -4.58764434e-01 1.57648608e-01
4.48777378e-01 5.41270137e-01 7.08247960e-01 1.37053847e+00
2.27575339e-02 -1.17722787e-01 6.08586609e-01 1.33788419e+00
-2.37328231e-01 -4.22174186e-01 -5.19161403e-01 1.47221461e-01
6.14704847e-01 1.35720229e+00 -7.90747106e-01 -4.43172991e-01
-2.91566461e-01 9.85410511e-01 3.93867940e-01 2.11158127e-01
-8.75059187e-01 -1.02139644e-01 4.88743842e-01 4.76511687e-01
5.53519428e-01 -1.56078219e-01 -1.91085309e-01 -9.37053263e-01
4.60752770e-02 -6.25506759e-01 4.40028697e-01 -6.09150112e-01
-1.44209743e+00 2.55964577e-01 -7.08118677e-02 -1.18934214e+00
7.54725048e-03 -8.22355390e-01 -3.94931614e-01 7.89551973e-01
-1.71658242e+00 -1.61183524e+00 -4.79494691e-01 5.58723867e-01
7.22220242e-01 -1.88791633e-01 5.41330934e-01 3.35594863e-01
-4.81091768e-01 7.09548652e-01 -1.63440362e-01 3.19286853e-01
7.71419942e-01 -1.32038212e+00 6.05589688e-01 8.20154548e-01
2.56353706e-01 2.53342032e-01 5.93775511e-01 -6.01165235e-01
-1.32982135e+00 -1.63833511e+00 4.86998945e-01 -7.63985157e-01
5.22220314e-01 -8.10006797e-01 -8.26257527e-01 4.44974393e-01
-5.16188264e-01 9.93381858e-01 -3.15562561e-02 -2.85311848e-01
-5.95967591e-01 -2.37942368e-01 -1.30223072e+00 2.68175900e-01
1.21594584e+00 -5.49386382e-01 -6.82590783e-01 5.64916790e-01
9.36685085e-01 -5.92659533e-01 -2.48760134e-01 2.99602598e-01
3.11360478e-01 -7.12653220e-01 1.20274568e+00 -5.76283336e-01
2.19029918e-01 -8.09260905e-01 -1.50638863e-01 -8.35399806e-01
-2.28100449e-01 -1.60864085e-01 -3.44145179e-01 9.78196204e-01
4.25105810e-01 -2.52120793e-01 8.44862461e-01 6.80129454e-02
-3.72612923e-02 -4.24863189e-01 -1.00973225e+00 -1.15001452e+00
-4.39310133e-01 -4.81795192e-01 3.38288069e-01 5.19789875e-01
-6.81176186e-01 3.83288413e-01 -5.54189123e-02 4.97412324e-01
1.05499494e+00 3.06724995e-01 1.12334394e+00 -1.09157729e+00
-3.78682107e-01 -2.28069812e-01 -5.15316784e-01 -1.26013255e+00
-3.66051644e-02 -8.53644729e-01 4.50103223e-01 -1.20068955e+00
4.28322762e-01 -7.09570765e-01 -3.18433851e-01 6.64820910e-01
-2.26242214e-01 7.67392457e-01 8.56974274e-02 2.32833818e-01
-1.04850602e+00 3.06253582e-01 1.12508321e+00 -2.65652239e-01
-1.10018440e-01 7.94764012e-02 -5.70437074e-01 7.64258564e-01
1.73167095e-01 -9.30731475e-01 -1.65298715e-01 -3.79233003e-01
-2.49726713e-01 -3.62961948e-01 1.09700561e+00 -9.56382155e-01
3.23681653e-01 -8.06408655e-03 5.57324290e-01 -6.84086561e-01
1.51970878e-01 -7.01031864e-01 -4.62409586e-01 6.22781813e-01
-4.22259748e-01 -6.24527097e-01 1.00228362e-01 8.64130616e-01
-1.53285889e-02 -5.93018606e-02 1.20293915e+00 -6.37213513e-02
-1.09764278e+00 3.94058287e-01 1.52743146e-01 1.32459849e-01
1.33247316e+00 -8.90623033e-02 -1.31054252e-01 2.08534464e-01
-4.89050448e-01 3.12138617e-01 4.52806443e-01 6.29271686e-01
4.92871642e-01 -1.18944192e+00 -9.41889048e-01 2.30752230e-01
8.67503285e-01 5.63195288e-01 -1.42013654e-01 7.39781737e-01
-6.73093423e-02 1.85085148e-01 1.21558607e-01 -1.24306595e+00
-9.13438976e-01 7.48669922e-01 4.12434071e-01 -1.39614806e-01
-5.91589630e-01 1.19981408e+00 6.02832377e-01 -5.51948667e-01
4.26847190e-01 -4.04237151e-01 3.65542173e-01 -1.09659337e-01
7.03883231e-01 1.76269248e-01 6.03206493e-02 -4.00721073e-01
-5.63938200e-01 2.85431832e-01 -2.63899982e-01 7.04626590e-02
1.27809370e+00 4.46739160e-02 3.76248091e-01 1.64862841e-01
8.89549613e-01 -2.93880761e-01 -1.95913672e+00 -8.28612387e-01
2.82047391e-01 -6.00953043e-01 1.86047599e-01 -1.04928255e+00
-7.98465014e-01 6.74288929e-01 8.31353247e-01 -1.54985338e-01
8.96865070e-01 5.26696920e-01 5.85169673e-01 4.29238737e-01
5.43702960e-01 -1.01698673e+00 5.76727629e-01 1.93514690e-01
7.44539440e-01 -1.70599210e+00 -1.14195235e-01 -1.74712673e-01
-5.47906518e-01 8.78784120e-01 9.92939353e-01 -2.73645669e-01
3.07276964e-01 5.04795671e-01 -1.81433231e-01 -1.86157346e-01
-5.73633075e-01 -4.61599827e-01 4.91854131e-01 6.81212783e-01
-1.38643458e-01 -9.53983217e-02 3.43337208e-01 5.44454157e-01
5.87905467e-01 1.21960811e-01 -4.31729518e-02 7.70423532e-01
-7.38897383e-01 -9.11589324e-01 -5.32826185e-01 4.04040664e-01
-8.64841416e-02 8.64430591e-02 -3.72873783e-01 6.96659207e-01
5.88395774e-01 5.98528802e-01 6.94931997e-03 -1.99347753e-02
4.74593759e-01 -2.33726069e-01 3.34012896e-01 -9.62047935e-01
-2.07721740e-01 -9.47063193e-02 -1.83144748e-01 -7.46740222e-01
-2.21907035e-01 -6.23245239e-01 -1.30746853e+00 3.58358413e-01
-6.20341539e-01 -1.85546786e-01 5.74372590e-01 8.50334167e-01
4.69674349e-01 3.57463926e-01 4.72557634e-01 -1.12367582e+00
-9.39020991e-01 -9.69191909e-01 -5.18805563e-01 3.99222076e-01
7.18963087e-01 -8.77794743e-01 -1.45906331e-02 3.86333689e-02] | [9.390867233276367, 1.3655011653900146] |
99fe3a34-de2b-4a80-901d-1d6048b43747 | towards-remote-fault-detection-by-analyzing | 2209.15498 | null | https://arxiv.org/abs/2209.15498v1 | https://arxiv.org/pdf/2209.15498v1.pdf | Towards remote fault detection by analyzing communication priorities | The ability to detect faults is an important safety feature for event-based multi-agent systems. In most existing algorithms, each agent tries to detect faults by checking its own behavior. But what if one agent becomes unable to recognize misbehavior, for example due to failure in its onboard fault detection? To improve resilience and avoid propagation of individual errors to the multi-agent system, agents should check each other remotely for malfunction or misbehavior. In this paper, we build upon a recently proposed predictive triggering architecture that involves communication priorities shared throughout the network to manage limited bandwidth. We propose a fault detection method that uses these priorities to detect errors in other agents. The resulting algorithms is not only able to detect faults, but can also run on a low-power microcontroller in real-time, as we demonstrate in hardware experiments. | ['Sebastian Trimpe', 'Dominik Baumann', 'Alexander Gräfe'] | 2022-09-30 | null | null | null | null | ['fault-detection'] | ['miscellaneous'] | [ 1.16738945e-01 2.21569031e-01 1.68854877e-01 -3.46486382e-02
-9.92036164e-02 -4.14265841e-01 5.15254617e-01 5.90285301e-01
-1.94544911e-01 9.41863298e-01 -7.59424329e-01 -2.79715776e-01
-1.88101575e-01 -1.15271664e+00 -3.61790329e-01 -3.95949692e-01
-7.14652538e-01 6.05581224e-01 1.10129833e+00 -1.27090052e-01
1.50269251e-02 5.78793585e-01 -1.80870640e+00 -1.31704167e-01
5.95268667e-01 7.71038890e-01 -3.67884263e-02 1.09292507e+00
5.85858524e-01 1.27688384e+00 -1.36290717e+00 7.29236722e-01
-3.02750647e-01 -4.30209786e-01 -6.87053382e-01 2.59448588e-01
-4.26365465e-01 -5.89682519e-01 9.82945263e-02 1.00151861e+00
1.26492679e-01 -1.68225765e-01 1.45254701e-01 -2.28227091e+00
3.25292826e-01 6.97976351e-01 -1.68827653e-01 1.80413872e-01
6.31679177e-01 2.13356450e-01 2.90310562e-01 -1.50848836e-01
2.37736419e-01 9.59908187e-01 4.25015867e-01 3.05439591e-01
-1.06978428e+00 -4.22329962e-01 2.11473122e-01 5.98391771e-01
-1.36406386e+00 -3.99729699e-01 5.58062077e-01 -1.24668822e-01
1.24438977e+00 5.01125753e-01 2.23465830e-01 7.63414800e-01
9.46567535e-01 -7.30271563e-02 1.12299252e+00 -6.79963291e-01
8.82656217e-01 -1.03870727e-01 1.14340179e-01 7.17970490e-01
7.60043621e-01 1.24837860e-01 -2.37227648e-01 -5.41246235e-01
7.05074489e-01 -4.44443990e-03 -2.82892704e-01 2.17202410e-01
-1.31187928e+00 4.46611404e-01 -1.16668366e-01 4.15699333e-01
-7.08600819e-01 4.60118502e-01 4.09998417e-01 5.54229915e-01
-5.49191870e-02 1.53415188e-01 -3.88438851e-01 -6.79414943e-02
-1.27219290e-01 -9.81419235e-02 1.24314749e+00 6.06444776e-01
5.36274374e-01 3.85797620e-01 5.60052693e-01 1.38363361e-01
5.65879047e-01 3.80356729e-01 4.84179974e-01 -7.75773287e-01
-4.44955498e-01 6.87565088e-01 2.15525433e-01 -6.80120230e-01
-8.30698490e-01 1.86978474e-01 -6.66421652e-01 1.30363333e+00
-1.04520202e-01 -3.96750242e-01 -5.21927893e-01 1.29321718e+00
6.25289559e-01 5.80457389e-01 3.72276485e-01 5.96229315e-01
-3.80357727e-02 6.97968960e-01 -3.23901698e-02 -5.91401160e-01
1.36347270e+00 -4.54926163e-01 -8.11549783e-01 4.95219370e-04
2.78312922e-01 -6.38051867e-01 -6.33809716e-03 7.86354899e-01
-9.70950544e-01 -1.51920065e-01 -1.42240584e+00 1.16449034e+00
-9.71782580e-02 -3.87792379e-01 3.04546952e-01 5.31903625e-01
-1.23097622e+00 3.47327650e-01 -1.22876513e+00 -4.87823218e-01
-7.04806864e-01 6.56241834e-01 -1.85963362e-01 4.33402777e-01
-1.19913185e+00 1.11770177e+00 3.89825583e-01 -1.33256570e-01
-1.40340936e+00 1.02268808e-01 -8.17036092e-01 -3.57673764e-02
5.52789271e-01 -4.62276697e-01 1.58491254e+00 -1.15439749e+00
-1.82245505e+00 -1.42915562e-01 2.92213470e-01 -5.47505379e-01
1.87630787e-01 2.72999018e-01 -1.02747810e+00 4.22007918e-01
1.28107537e-02 -1.15640201e-01 9.83776927e-01 -1.10995424e+00
-1.08390439e+00 1.58896279e-02 3.23511690e-01 -3.85153502e-01
9.14110988e-02 2.78499484e-01 6.19894445e-01 -1.84850290e-01
-3.08063664e-02 -8.75870228e-01 -3.05937767e-01 -3.62380385e-01
-1.45483866e-01 -3.35588545e-01 1.10635972e+00 2.40015060e-01
8.33421350e-01 -1.81522167e+00 -4.74419504e-01 5.18555701e-01
3.18963453e-02 1.75361812e-01 6.11111894e-02 8.79622400e-01
-3.14250402e-02 -2.72952795e-01 4.81685735e-02 3.41145664e-01
-1.13090739e-01 4.87037152e-01 -2.21122876e-01 8.28235865e-01
3.79961669e-01 -2.39758059e-01 -7.61748195e-01 -2.36277610e-01
4.43265200e-01 3.29430312e-01 -6.18780740e-02 2.96924293e-01
-2.15105996e-01 1.01900019e-01 -5.57785213e-01 5.38369596e-01
4.70509559e-01 -2.31913552e-01 5.49556911e-01 5.40322721e-01
-5.21476090e-01 3.27061862e-01 -1.70920575e+00 6.70159996e-01
-5.15209556e-01 3.66301686e-01 6.89179540e-01 -8.93153071e-01
6.16603136e-01 1.15241396e+00 4.77977782e-01 -6.05824947e-01
2.46458784e-01 4.19243723e-01 3.46176326e-02 -3.84646982e-01
1.56743422e-01 1.56082124e-01 -1.90940276e-01 9.24186409e-01
-4.19747263e-01 3.38661611e-01 1.36777222e-01 7.06965402e-02
1.99912059e+00 -4.63757217e-01 4.58885700e-01 -1.55481502e-01
7.08968163e-01 3.45217645e-01 7.57221937e-01 7.72883654e-01
-3.11325818e-01 -3.80081266e-01 2.16219857e-01 -4.26211923e-01
-4.70832407e-01 -7.92562723e-01 2.76654482e-01 7.62804568e-01
4.30910140e-01 -4.43389952e-01 -4.98684227e-01 -5.50358176e-01
-1.29396066e-01 6.69167936e-01 -3.26551944e-02 -4.09366846e-01
-6.25464380e-01 -4.85763997e-01 5.47343612e-01 2.35687256e-01
2.60549366e-01 -1.15566397e+00 -1.53455353e+00 1.09547663e+00
3.50247324e-01 -7.49892831e-01 1.47767201e-01 4.74954695e-01
-4.72280473e-01 -1.54504251e+00 2.95638800e-01 -7.07184315e-01
7.18572557e-01 3.44057113e-01 8.55289876e-01 7.90295362e-01
-3.25866669e-01 8.14271808e-01 -5.29017270e-01 -4.48197305e-01
-1.01688695e+00 -6.78147614e-01 4.54728991e-01 -1.20369881e-01
-1.22250013e-01 -6.29215240e-01 -2.53338128e-01 5.91320038e-01
-1.12459326e+00 -6.45336866e-01 2.85808802e-01 4.58183557e-01
1.17381625e-02 9.45595384e-01 1.09695685e+00 -3.87376517e-01
6.92134857e-01 -6.39813423e-01 -1.09142399e+00 6.98396340e-02
-7.29582369e-01 -1.85568243e-01 8.89743447e-01 -4.40707147e-01
-4.65274245e-01 5.50380126e-02 1.21482119e-01 9.78460908e-02
-5.68773091e-01 3.00153673e-01 1.21159106e-01 -3.80880773e-01
3.83885980e-01 1.40856607e-02 2.53500730e-01 1.65847331e-01
-2.86831796e-01 6.07147872e-01 2.90781289e-01 -2.33598500e-01
3.99627417e-01 3.93572032e-01 2.29717091e-01 -8.07661176e-01
4.33556288e-01 1.73855480e-02 2.49464497e-01 -6.76701605e-01
1.67391166e-01 -7.98029244e-01 -1.64951158e+00 5.31823635e-01
-1.43410206e+00 -2.85925925e-01 2.03133553e-01 3.79022717e-01
-2.88600326e-01 1.88598916e-01 -7.31820881e-01 -1.10222077e+00
-1.00804955e-01 -1.16099048e+00 5.15517175e-01 3.71761173e-01
-4.88734871e-01 -9.09471393e-01 3.62138271e-01 -5.78595221e-01
6.22259140e-01 3.44450653e-01 4.68314141e-01 -9.12630141e-01
-5.73935151e-01 -4.74720061e-01 4.29164231e-01 -1.30265560e-02
3.60629976e-01 4.02970314e-01 -4.82936800e-01 -6.44723296e-01
3.18880469e-01 2.91882157e-02 -9.93955806e-02 -7.02079833e-02
2.80334145e-01 -5.29541314e-01 -5.38138151e-01 -5.32162070e-01
1.58787906e+00 7.21026838e-01 3.62960815e-01 6.23356819e-01
-4.07976285e-02 3.13484550e-01 6.21282339e-01 8.63044858e-01
5.40583730e-01 5.72094798e-01 1.05074155e+00 3.20185095e-01
2.01494470e-01 5.82328498e-01 1.05940795e+00 3.83856475e-01
3.00338596e-01 -5.05025089e-01 -8.89419258e-01 4.10849750e-01
-1.98679245e+00 -8.93964887e-01 -4.66633350e-01 2.25479603e+00
5.28494835e-01 3.69403005e-01 1.43477067e-01 5.22231936e-01
1.01035535e+00 -4.06349331e-01 -3.09052318e-01 -5.31988680e-01
2.73399532e-01 -5.87511957e-02 2.84119427e-01 7.13363171e-01
-7.82033861e-01 2.31971025e-01 6.38441753e+00 2.35476866e-02
-1.19889700e+00 2.40057021e-01 -9.08023566e-02 1.10192552e-01
2.59551316e-01 2.46821970e-01 -4.47525114e-01 6.46690130e-01
1.43919635e+00 -3.46356332e-01 1.65237650e-01 7.53645480e-01
5.77074409e-01 -8.12232733e-01 -1.17638564e+00 3.34291518e-01
4.88674678e-02 -8.25340688e-01 -6.18968248e-01 -2.37490192e-01
2.80259222e-01 -1.30489379e-01 -9.09146070e-01 -4.44651097e-02
8.52833033e-01 -6.29369259e-01 9.00399268e-01 3.38346094e-01
-8.07910284e-04 -1.00503600e+00 6.91430330e-01 4.73835111e-01
-1.10238302e+00 -1.34076059e-01 2.32205972e-01 -7.77412355e-01
3.68150890e-01 7.84002721e-01 -1.03224015e+00 5.13111830e-01
5.04125357e-01 -2.33518481e-01 -1.30047411e-01 1.17394102e+00
-4.02283967e-01 5.89266419e-01 -6.93125427e-01 -2.52146065e-01
-1.64096788e-01 4.82263684e-01 8.37444425e-01 6.15749836e-01
3.29857200e-01 2.35776410e-01 7.25890100e-01 2.48651803e-01
5.25223196e-01 -7.38037407e-01 -4.57768917e-01 6.32762730e-01
8.93617988e-01 1.18391907e+00 -9.78755593e-01 -5.48177481e-01
-3.82333487e-01 8.98903906e-01 -2.29622886e-01 8.34106356e-02
-8.86393368e-01 -8.45938444e-01 5.69146812e-01 2.07151771e-02
-5.64659052e-02 -5.23588538e-01 2.07042769e-01 -5.33313334e-01
4.19318974e-02 -9.79516208e-01 4.22557145e-01 -5.89276850e-01
-1.08066666e+00 8.28803241e-01 -3.68440539e-01 -1.11373937e+00
-5.53552568e-01 -4.68859598e-02 -1.08266366e+00 2.64235377e-01
-1.27233267e+00 -4.74643767e-01 -6.20457344e-02 8.88521791e-01
3.42481941e-01 8.21045637e-02 1.04320097e+00 9.36736539e-02
-7.89565921e-01 -3.71563546e-02 -3.10042441e-01 -3.79375935e-01
7.49628305e-01 -9.51979578e-01 -1.61030784e-01 1.12770987e+00
-4.61128473e-01 1.27994016e-01 1.04185891e+00 -8.49961400e-01
-1.76735246e+00 -1.01244855e+00 6.58972025e-01 1.37375876e-01
8.96914721e-01 1.79316521e-01 -8.40902746e-01 9.06981707e-01
5.88980019e-01 4.98901345e-02 2.94095635e-01 -4.57795352e-01
7.41179883e-02 -2.94828326e-01 -1.39484036e+00 3.62292767e-01
-3.63727994e-02 -1.14352681e-01 -4.36789244e-01 3.91628593e-01
3.69247168e-01 -6.04993850e-02 -7.07643330e-01 1.25237584e-01
1.07464138e-02 -1.02720571e+00 2.63975799e-01 7.94934575e-03
-3.86073083e-01 -9.86958861e-01 1.69834256e-01 -1.42702520e+00
-3.19917113e-01 -9.13151681e-01 1.11020848e-01 1.28817058e+00
1.72351182e-01 -1.33142257e+00 2.39385352e-01 5.41252017e-01
-1.31583855e-01 2.00246140e-01 -1.44322503e+00 -9.84307349e-01
-8.45747530e-01 -3.97306234e-01 6.20341480e-01 8.76900017e-01
1.00418663e+00 2.43739516e-01 -4.69428524e-02 1.20756769e+00
7.13645637e-01 -2.07158193e-01 4.18979436e-01 -1.26751328e+00
-3.02314967e-01 -2.31442735e-01 -6.76341891e-01 5.83209880e-02
8.70516673e-02 -2.57802188e-01 5.83250344e-01 -1.25170887e+00
-4.45893973e-01 -4.15619791e-01 -2.06312627e-01 9.67829585e-01
2.20869482e-01 2.03126725e-02 -1.37291551e-01 6.65953010e-02
-1.02312219e+00 -2.85469787e-03 1.23028152e-01 1.42907977e-01
-1.14915811e-01 -3.98084410e-02 2.26740927e-01 8.26004028e-01
1.00032580e+00 -7.21688390e-01 -1.68222919e-01 -1.19337693e-01
1.60194457e-01 8.46193016e-01 8.57440591e-01 -1.56515014e+00
7.86433935e-01 -4.89986926e-01 -6.02211244e-02 -4.52413671e-02
-6.94589689e-02 -1.53998673e+00 6.74196959e-01 1.21344638e+00
1.20770469e-01 6.82681084e-01 1.47029757e-01 6.63151503e-01
-3.10456336e-01 -2.01233044e-01 5.77539742e-01 1.19767092e-01
-7.74328291e-01 -3.33822936e-01 -1.60845137e+00 -8.73562217e-01
1.81304824e+00 3.86869788e-01 -8.27134371e-01 -1.59365326e-01
-5.60619354e-01 4.94370669e-01 5.34260392e-01 1.92138538e-01
7.44944930e-01 -9.71308768e-01 -4.47488815e-01 2.79243231e-01
-1.40555695e-01 -5.21344244e-01 -3.84564325e-02 7.54550040e-01
-5.42872608e-01 1.31777018e-01 -2.73471862e-01 -4.85926837e-01
-1.37161756e+00 5.05710602e-01 4.67037857e-01 1.03712164e-01
-5.47477305e-01 2.29832605e-01 -7.56223977e-01 3.40934813e-01
1.22102164e-01 -1.13214388e-01 -7.70191774e-02 -3.11735630e-01
9.82850373e-01 7.35890269e-01 3.12145829e-01 -1.93165302e-01
-9.69929874e-01 3.86195704e-02 4.98055071e-02 -3.96394074e-01
1.06212735e+00 -2.48919383e-01 -7.32285023e-01 3.79989028e-01
3.04422468e-01 -2.30514213e-01 -1.02576292e+00 2.15428278e-01
1.73008963e-01 -1.85885563e-01 1.36410579e-01 -7.62603939e-01
-7.98559606e-01 -1.50691539e-01 4.71504390e-01 1.22205865e+00
1.18225491e+00 -3.36473107e-01 3.01348954e-01 1.99108511e-01
1.07686853e+00 -9.76807892e-01 1.14080600e-01 4.33236033e-01
5.02629459e-01 -8.57558191e-01 -1.16450273e-01 -4.44375575e-01
-2.61683613e-01 1.43263674e+00 7.46992648e-01 -4.76897389e-01
5.95998645e-01 9.31548536e-01 -8.89182463e-02 -2.92602003e-01
-1.42216229e+00 8.89690593e-02 -1.03967309e+00 9.01576817e-01
-1.92867026e-01 2.65656471e-01 -4.33622092e-01 1.71822175e-01
5.50346255e-01 2.05445260e-01 1.61506820e+00 1.66931891e+00
-1.13250232e+00 -1.32844138e+00 -9.99460638e-01 -2.74666362e-02
-2.36764401e-01 5.18336296e-01 -2.19070971e-01 6.28563762e-01
-6.95248600e-03 1.79126370e+00 3.33136499e-01 -3.46408635e-01
3.71837556e-01 -6.93033636e-02 1.17514320e-01 -5.15058815e-01
-7.67329931e-01 -1.72459874e-02 4.15816277e-01 -7.49133766e-01
-3.52390349e-01 -4.59460646e-01 -1.78539121e+00 -3.96524101e-01
-2.90470004e-01 3.79713058e-01 7.05157101e-01 9.23699021e-01
4.54520136e-01 1.00219369e+00 1.00293171e+00 -6.43022954e-01
-3.25042605e-01 -7.52861917e-01 -7.28503585e-01 -2.99466163e-01
7.58411050e-01 -7.10774004e-01 -4.93208021e-01 -1.88071534e-01] | [5.41487979888916, 2.5037519931793213] |
a0ef5363-865e-4f27-a36b-611a022052d5 | domain-based-punjabi-text-document-clustering | null | null | https://aclanthology.org/C12-3049 | https://aclanthology.org/C12-3049.pdf | Domain Based Punjabi Text Document Clustering | null | ['Saurabh Sharma', 'Vishal Gupta'] | 2012-12-01 | domain-based-punjabi-text-document-clustering-1 | https://aclanthology.org/C12-3049 | https://aclanthology.org/C12-3049.pdf | coling-2012-12 | ['text-clustering'] | ['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] | [-7.339441299438477, 3.840492010116577] |
8ec90d0d-4f35-4fd7-bc1c-7705358529c0 | a-recommender-system-for-equitable-public-art | 2207.14367 | null | https://arxiv.org/abs/2207.14367v1 | https://arxiv.org/pdf/2207.14367v1.pdf | A Recommender System for Equitable Public Art Curation and Installation | The placement of art in public spaces can have a significant impact on who feels a sense of belonging. In cities, public art communicates whose interests and culture are being favored. In this paper, we propose a graph matching approach with local constraints to build a curatorial tool for selecting public art in a way that supports inclusive spaces. We develop a cost matrix by drawing on Schelling's model of segregation. Using the cost matrix as an input, the optimization problem is solved via projected gradient descent to obtain a soft assignment matrix. We discuss regularization terms to set curatorial constraints. Our optimization program allocates artwork to public spaces and walls in a way that de-prioritizes "in-group" preferences, by satisfying minimum representation and exposure criteria. We draw on existing literature to develop a fairness metric for our algorithmic output. Using Tufts University as a testbed, we assess the effectiveness of our approach and discuss its potential pitfalls from both a curatorial and equity standpoint. | ['Dina Deitsch', 'Abiy Tasissa', 'Anna Haensch'] | 2022-07-28 | null | null | null | null | ['graph-matching'] | ['graphs'] | [ 2.81537890e-01 2.93443352e-02 -5.51341414e-01 -4.47395593e-01
-5.65328121e-01 -4.84777540e-01 3.26193273e-01 2.70230234e-01
-3.08679372e-01 6.56935513e-01 7.31312990e-01 -6.14783883e-01
-5.81932306e-01 -1.11048865e+00 -2.90370643e-01 -3.06281149e-01
6.00966871e-01 4.12450880e-01 -2.70559222e-01 -3.59448865e-02
4.69508529e-01 6.24575794e-01 -1.41480529e+00 1.95237890e-01
1.20347035e+00 5.98931670e-01 3.30834687e-02 3.77628684e-01
-1.60916403e-01 6.91777885e-01 -3.67961824e-01 -1.04367793e+00
6.23811245e-01 -3.55399400e-01 -8.33714604e-01 3.68788511e-01
7.46652007e-01 -2.50039361e-02 -3.55517864e-02 7.14379609e-01
6.55949473e-01 3.53406966e-01 7.84448683e-01 -1.36504042e+00
-8.39494169e-01 6.37672007e-01 -4.73117381e-01 -2.65750177e-02
2.83954263e-01 5.24065234e-02 1.67195213e+00 -5.17532825e-01
7.25406349e-01 1.17011142e+00 6.16451144e-01 1.17676191e-01
-1.78488910e+00 -5.90398550e-01 5.40035844e-01 -2.27209345e-01
-1.45362413e+00 -7.06538022e-01 8.17231357e-01 -7.01064825e-01
5.25542796e-01 1.08235466e+00 1.03611457e+00 4.94031936e-01
-3.62404615e-01 4.24489439e-01 1.05738914e+00 -5.31466305e-01
4.35932249e-01 3.49735528e-01 1.06754586e-01 2.91582674e-01
6.06050909e-01 -4.61187631e-01 -5.15510380e-01 -4.67359185e-01
6.02664769e-01 -1.21817380e-01 5.33200614e-02 -4.76344883e-01
-6.10133529e-01 8.82150412e-01 4.49972600e-01 2.75751114e-01
-8.71832818e-02 2.93167353e-01 -3.22056651e-01 1.05976045e-01
6.00903869e-01 5.37305176e-01 1.62489533e-01 1.17001310e-01
-9.33754325e-01 3.40290099e-01 7.21797466e-01 5.35342336e-01
8.94567907e-01 -4.79517788e-01 -4.89795595e-01 1.06487298e+00
6.81706309e-01 4.07752723e-01 -5.01454532e-01 -1.23680961e+00
6.79851055e-01 8.17951143e-01 2.81457603e-01 -1.41341162e+00
-1.01114228e-01 -5.91056705e-01 -2.63130426e-01 3.73997271e-01
5.91515779e-01 -2.45529830e-01 -3.85366023e-01 1.76317859e+00
2.49741942e-01 -1.38686642e-01 -5.64322114e-01 9.22468364e-01
1.64742664e-01 9.71000195e-02 2.01079339e-01 -1.43782556e-01
9.29809272e-01 -6.03707790e-01 -6.71740472e-01 -3.76151711e-01
4.88361746e-01 -8.19145739e-01 1.08293867e+00 -1.77666381e-01
-1.20664394e+00 4.28734235e-02 -6.11833632e-01 4.43679579e-02
-2.99618900e-01 6.91558272e-02 7.01005578e-01 1.34720671e+00
-1.26578164e+00 6.00815475e-01 -3.62958223e-01 -5.95843077e-01
7.73365796e-01 4.16440874e-01 2.20092267e-01 2.56090730e-01
-5.87296247e-01 8.32234323e-01 -4.95538950e-01 -2.11565480e-01
-1.33498600e-02 -7.27317750e-01 -5.74501514e-01 1.07821867e-01
3.27596098e-01 -8.83654952e-01 6.09782755e-01 -1.06727183e+00
-9.29992080e-01 1.00876653e+00 4.61397041e-03 3.98327969e-02
7.38895357e-01 3.15360039e-01 -3.72003675e-01 -2.32518017e-01
5.26405871e-01 4.36567575e-01 4.74731743e-01 -1.48681068e+00
-5.87860882e-01 -3.99764299e-01 4.09069568e-01 3.50111425e-01
-5.97645342e-01 2.38951504e-01 -1.42855614e-01 -6.98737741e-01
9.52327177e-02 -1.00680530e+00 -4.03773218e-01 7.93258399e-02
-6.13309860e-01 3.65579687e-02 1.67667523e-01 -4.20702487e-01
1.91143584e+00 -1.92343664e+00 -9.93841365e-02 1.04239607e+00
2.75117666e-01 -3.94228667e-01 6.70441464e-02 5.46507537e-01
2.22278446e-01 5.78460097e-01 -1.69500619e-01 -4.37324941e-01
5.45436442e-01 -4.11220603e-02 1.53762355e-01 5.16562879e-01
-1.42675102e-01 5.01652896e-01 -6.84203625e-01 -5.89842856e-01
5.22869155e-02 3.21973801e-01 -9.49347377e-01 -2.17058986e-01
2.11773306e-01 8.51979181e-02 -4.96655315e-01 6.42531991e-01
6.63976610e-01 -3.98925930e-01 6.39072061e-01 1.79462239e-01
-4.16476250e-01 3.53492856e-01 -1.49058974e+00 1.17701542e+00
-2.32287884e-01 6.23728931e-01 4.49245930e-01 -5.35313845e-01
8.97614658e-01 -2.43489489e-01 5.56537628e-01 -4.75645155e-01
2.14900926e-01 2.63959765e-01 -1.36314079e-01 -2.73914963e-01
5.92535555e-01 -6.56740963e-02 7.26824477e-02 8.62030268e-01
-8.60589862e-01 6.07874729e-02 1.87493190e-01 3.68154287e-01
1.02489436e+00 -1.78888142e-02 -2.34000057e-01 -6.82108760e-01
2.45024785e-01 2.82287598e-04 5.09094954e-01 6.49932861e-01
-2.30071530e-01 5.67060113e-01 5.80010533e-01 -2.13426828e-01
-9.69754994e-01 -8.93746972e-01 -1.57774150e-01 1.25805426e+00
5.79368509e-02 -2.34731659e-01 -7.03268707e-01 -4.57984954e-01
4.14706230e-01 9.53795433e-01 -5.06807268e-01 1.74306467e-01
-3.51377130e-01 -8.02150190e-01 -1.68583002e-02 1.20133393e-01
5.53047284e-02 -4.46783930e-01 -5.25084257e-01 1.45953912e-02
-2.26858705e-01 -6.52743459e-01 -7.89648652e-01 -2.27670431e-01
-4.76872176e-01 -1.02123916e+00 -6.69202447e-01 -6.45429194e-01
9.63986754e-01 4.31690305e-01 1.25534415e+00 3.16264451e-01
-2.29734197e-01 5.87582946e-01 -1.17122613e-01 -1.57631263e-01
-5.19561097e-02 1.13491155e-01 -1.62026748e-01 4.37974304e-01
3.22402865e-01 -8.30575883e-01 -7.18217731e-01 4.81832534e-01
-5.52284181e-01 3.36658843e-02 2.22646631e-02 2.11867988e-01
2.81803280e-01 4.29337882e-02 5.65162063e-01 -8.17463398e-01
7.73428619e-01 -5.20433068e-01 -5.45767367e-01 3.28095675e-01
-8.67135286e-01 -2.89935321e-01 -1.35983646e-01 -1.06465064e-01
-1.08302438e+00 1.69701912e-02 2.61496007e-01 1.34722441e-01
3.92515898e-01 1.66134298e-01 -4.91521418e-01 -3.26473445e-01
5.04441381e-01 -7.45135844e-01 -2.49003261e-01 -4.74951148e-01
3.14001441e-01 7.40028262e-01 -2.38299817e-01 -9.65967417e-01
7.06887484e-01 5.74584186e-01 -1.68471277e-01 -4.64885026e-01
-6.19646609e-01 -2.38911092e-01 -3.60508889e-01 -5.73243797e-01
7.85900116e-01 -6.41309321e-01 -9.56132114e-01 -1.40515029e-01
-5.83887875e-01 -5.11923254e-01 -3.49281341e-01 3.85154396e-01
-4.98676032e-01 1.11528928e-03 -1.09903835e-01 -1.27000153e+00
1.84851348e-01 -8.96699071e-01 6.04203403e-01 1.68079242e-01
-5.42977095e-01 -9.52849567e-01 1.68875992e-01 9.47502077e-01
4.11212593e-01 4.06456083e-01 7.42249608e-01 -1.81073412e-01
-8.59812140e-01 3.30854893e-01 -3.06365013e-01 -3.22037369e-01
5.28520755e-02 8.41865316e-02 -7.55990565e-01 -6.66979328e-02
-6.01717412e-01 2.10341990e-01 5.28007030e-01 5.34434438e-01
8.72397661e-01 -4.75853235e-01 -4.24731851e-01 5.73402166e-01
1.80537224e+00 2.86829914e-03 6.01384044e-01 7.55933285e-01
6.49320602e-01 9.49762940e-01 2.97969222e-01 7.59400964e-01
6.59734130e-01 7.56851554e-01 3.11131448e-01 -1.39268398e-01
-1.35911396e-02 -2.72660166e-01 -2.05683988e-02 2.38402233e-01
-2.65416831e-01 -2.85964340e-01 -9.81899917e-01 8.07757020e-01
-2.13585758e+00 -1.04024673e+00 -2.21113250e-01 2.21885586e+00
6.88610494e-01 -1.85329863e-03 4.66604680e-01 -1.15691677e-01
1.11981547e+00 1.46659434e-01 -2.44745150e-01 -4.76066202e-01
-1.72134206e-01 -6.08680584e-02 7.78422832e-01 8.16488266e-01
-8.72246683e-01 6.13118112e-01 6.82812738e+00 6.69819772e-01
-4.49479580e-01 4.64235768e-02 1.16065407e+00 -5.07650137e-01
-1.12117040e+00 2.58065820e-01 -5.88964403e-01 4.54391718e-01
4.58184242e-01 -2.35044017e-01 7.77123213e-01 3.65969360e-01
5.85327685e-01 -5.88509403e-02 -9.66568291e-01 9.11088586e-01
-1.52731314e-01 -1.53567004e+00 -2.79802382e-01 4.81443822e-01
1.05135238e+00 -5.20756483e-01 2.36566171e-01 -2.55525559e-01
7.93614447e-01 -1.03692722e+00 8.52700174e-01 4.15134609e-01
6.23781741e-01 -8.91556621e-01 -9.05553694e-04 -1.66011423e-01
-1.29732990e+00 -2.11078286e-01 1.24568967e-02 -4.76684690e-01
2.89203614e-01 6.60393775e-01 -3.74753833e-01 2.14515239e-01
4.33297366e-01 2.50801980e-01 -4.05144125e-01 1.02386880e+00
3.03600207e-02 3.14057887e-01 -2.68446624e-01 -1.89108346e-02
9.23798680e-02 -6.98107064e-01 4.58383113e-01 9.39704537e-01
4.28166419e-01 2.19293416e-01 2.19983935e-01 1.25322139e+00
-1.33377582e-01 5.83071113e-01 -5.77013195e-01 -8.38694274e-02
7.49594748e-01 1.24027359e+00 -8.63234639e-01 1.99287236e-02
-3.46083432e-01 4.84532148e-01 2.99533963e-01 5.10592699e-01
-5.92702806e-01 -1.20341532e-01 9.79058504e-01 7.40959108e-01
8.21010792e-04 6.51007220e-02 -8.67805004e-01 -8.56012940e-01
-4.31206226e-02 -7.12309539e-01 3.68191749e-01 -6.57472372e-01
-1.30169082e+00 -3.44147757e-02 -1.35650322e-01 -7.21830487e-01
5.70680141e-01 -2.20432490e-01 -7.70616651e-01 9.82033968e-01
-1.14518547e+00 -1.18431330e+00 -1.12131357e-01 3.56886268e-01
-1.99863359e-01 6.92574084e-02 2.64684141e-01 6.98951006e-01
-7.56336510e-01 7.05275357e-01 -2.72105057e-02 -1.37173951e-01
5.07578969e-01 -1.05229354e+00 1.03818253e-01 8.03399444e-01
-9.99236256e-02 6.21797681e-01 5.86746275e-01 -8.00119460e-01
-1.13895106e+00 -8.95035625e-01 1.29851532e+00 -4.87695396e-01
7.21838057e-01 -3.13275069e-01 -1.98596805e-01 6.74472392e-01
1.09699525e-01 -5.99283695e-01 1.22983623e+00 5.45735598e-01
1.84278022e-02 -2.33235627e-01 -1.49793243e+00 9.68731642e-01
1.58010042e+00 -4.68340486e-01 -4.96800207e-02 4.78518754e-01
4.85549748e-01 1.19936749e-01 -8.96645308e-01 -3.45212102e-01
6.27705157e-01 -1.07871616e+00 9.97542679e-01 -3.38682532e-01
1.92214549e-01 -1.88430712e-01 -4.92807239e-01 -9.14574146e-01
-8.44039738e-01 -8.82536769e-01 6.38924897e-01 1.58556950e+00
7.25145757e-01 -8.16010892e-01 1.13640845e+00 1.56528842e+00
5.09363897e-02 -7.30729103e-01 -8.74160528e-01 -4.59483117e-01
-1.58525169e-01 -3.57910603e-01 1.09141326e+00 1.17776382e+00
3.96940894e-02 2.43956313e-01 -2.60511190e-01 -3.52860056e-02
7.01931953e-01 2.55397350e-01 6.39091253e-01 -1.37241912e+00
-1.62544236e-01 -8.43976319e-01 -1.03298537e-01 -3.80079001e-01
1.02578029e-01 -1.09777617e+00 -4.76635039e-01 -1.93237209e+00
3.73042792e-01 -1.21049225e+00 -2.51981169e-01 4.73409861e-01
1.47034332e-01 3.52319956e-01 5.31513453e-01 -3.67281120e-03
-3.59463483e-01 1.05963290e-01 1.11555064e+00 -2.61799932e-01
-6.36604130e-01 -1.38600469e-01 -1.43711436e+00 5.24874985e-01
9.13867116e-01 -3.33899349e-01 -4.01343673e-01 -5.65692067e-01
7.57687271e-01 -2.63230056e-01 3.03319812e-01 -6.45692706e-01
1.48380592e-01 -8.35349977e-01 1.67108625e-01 -3.04415703e-01
4.39326257e-01 -1.12562740e+00 6.59758329e-01 2.17582136e-01
-6.08814657e-01 -1.33199126e-01 -2.50167251e-01 3.33601087e-01
4.92181301e-01 -2.02573046e-01 5.14925420e-01 1.50386930e-01
-6.32761121e-02 2.25166008e-01 -2.26458356e-01 1.08659990e-01
1.06813037e+00 -5.23645639e-01 -3.00368100e-01 -5.18053591e-01
-8.21026981e-01 3.62880886e-01 1.12139881e+00 8.90592486e-02
1.81707501e-01 -1.70551980e+00 -7.16162324e-01 -1.11586206e-01
-8.21553022e-02 -7.86843598e-01 -5.28468676e-02 7.48454034e-01
-4.04760927e-01 -6.02701120e-02 -1.17892690e-01 -7.91366175e-02
-1.34237921e+00 1.70871973e-01 1.80284917e-01 5.66268712e-02
-3.20809007e-01 8.20119500e-01 -1.46424115e-01 -3.12211990e-01
1.61606714e-01 1.60184905e-01 -1.73825264e-01 3.88460666e-01
1.32959247e-01 8.62534344e-01 -2.29301766e-01 -7.48252630e-01
-6.41903758e-01 5.82643867e-01 6.27385497e-01 -5.73670328e-01
1.44357717e+00 -4.55882937e-01 -2.94048995e-01 5.02059683e-02
9.36042964e-01 7.32449532e-01 -1.06473887e+00 1.90825671e-01
5.82270361e-02 -1.31303334e+00 7.60734081e-02 -7.02436090e-01
-1.03370583e+00 1.33547008e-01 3.71030450e-01 5.08131325e-01
9.87459242e-01 -1.68476775e-01 4.79258001e-01 9.04416759e-03
8.61390680e-02 -1.72409165e+00 -2.79165298e-01 -7.66690001e-02
5.93394041e-01 -1.02702141e+00 2.61024326e-01 -5.56828916e-01
-5.38006544e-01 5.70834517e-01 4.00792301e-01 -2.64856331e-02
6.94436371e-01 5.90049140e-02 -1.17648348e-01 -1.99418798e-01
-3.11037391e-01 -5.42867541e-01 1.95915818e-01 4.36875075e-01
4.92123842e-01 5.87895274e-01 -7.68771291e-01 4.35468853e-01
-3.50219131e-01 -2.28169367e-01 3.94679129e-01 8.27286363e-01
-6.79786623e-01 -1.44356477e+00 -6.79951549e-01 7.17588842e-01
-3.20182472e-01 -5.65535389e-02 -8.21195245e-01 4.04773802e-01
5.28826177e-01 1.40676117e+00 2.79445559e-01 -2.27784619e-01
2.79117972e-01 -1.45028085e-01 2.49655545e-01 -4.86181855e-01
-8.39248955e-01 2.76707888e-01 9.08224404e-01 -3.80041867e-01
-5.69880545e-01 -8.41716290e-01 -8.28078270e-01 -8.32461476e-01
-2.71669209e-01 1.71578735e-01 6.25091612e-01 5.83160102e-01
5.20896435e-01 4.74712759e-01 6.63470387e-01 -4.59055722e-01
-4.83122393e-02 -1.89230338e-01 -7.13070333e-01 4.36365068e-01
-1.24226667e-01 -6.37792408e-01 -3.54999870e-01 -3.12903851e-01] | [8.563907623291016, 5.4099578857421875] |
89aa1f4c-0d71-4e1f-84d4-ef72ee2f96ff | multi-garment-net-learning-to-dress-3d-people | 1908.06903 | null | https://arxiv.org/abs/1908.06903v2 | https://arxiv.org/pdf/1908.06903v2.pdf | Multi-Garment Net: Learning to Dress 3D People from Images | We present Multi-Garment Network (MGN), a method to predict body shape and clothing, layered on top of the SMPL model from a few frames (1-8) of a video. Several experiments demonstrate that this representation allows higher level of control when compared to single mesh or voxel representations of shape. Our model allows to predict garment geometry, relate it to the body shape, and transfer it to new body shapes and poses. To train MGN, we leverage a digital wardrobe containing 712 digital garments in correspondence, obtained with a novel method to register a set of clothing templates to a dataset of real 3D scans of people in different clothing and poses. Garments from the digital wardrobe, or predicted by MGN, can be used to dress any body shape in arbitrary poses. We will make publicly available the digital wardrobe, the MGN model, and code to dress SMPL with the garments. | ['Gerard Pons-Moll', 'Christian Theobalt', 'Bharat Lal Bhatnagar', 'Garvita Tiwari'] | 2019-08-19 | multi-garment-net-learning-to-dress-3d-people-1 | http://openaccess.thecvf.com/content_ICCV_2019/html/Bhatnagar_Multi-Garment_Net_Learning_to_Dress_3D_People_From_Images_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Bhatnagar_Multi-Garment_Net_Learning_to_Dress_3D_People_From_Images_ICCV_2019_paper.pdf | iccv-2019-10 | ['3d-shape-reconstruction-from-a-single-2d'] | ['computer-vision'] | [ 2.08456382e-01 -3.35053578e-02 1.65922135e-01 -6.39134228e-01
-2.84355938e-01 -6.14072800e-01 2.12854251e-01 -3.40954542e-01
1.41709864e-01 4.76960719e-01 2.51108974e-01 5.18667281e-01
1.09291136e-01 -1.16323102e+00 -1.19369686e+00 -9.43320021e-02
-1.20674498e-01 8.80114257e-01 3.13060224e-01 -3.15719783e-01
-4.83690858e-01 7.62495637e-01 -1.39206517e+00 2.53832608e-01
1.91235673e-02 1.05861211e+00 -1.29815370e-01 7.63099730e-01
5.17610073e-01 -9.21707675e-02 -5.41825704e-02 -8.33157659e-01
8.13582063e-01 -9.91723388e-02 -4.08352852e-01 4.54278320e-01
1.14443731e+00 -7.70823836e-01 -5.52963018e-01 6.62933409e-01
6.12053990e-01 8.20556656e-02 5.13692558e-01 -8.19876194e-01
-4.88530397e-01 4.08976227e-01 -3.61965388e-01 -3.57511669e-01
5.71456552e-01 2.98552271e-02 4.68201041e-01 -8.75502348e-01
9.31966901e-01 1.32941806e+00 1.37885904e+00 7.01935470e-01
-1.40173781e+00 -3.66370112e-01 -5.00758737e-02 -4.26905990e-01
-1.36504614e+00 -3.37555766e-01 1.01696467e+00 -2.93625981e-01
4.31680292e-01 4.27263588e-01 1.52809966e+00 1.23182964e+00
5.94577603e-02 5.77419758e-01 9.54713345e-01 -1.49520576e-01
-2.06802368e-01 -2.61174977e-01 -2.88051724e-01 1.05740380e+00
2.15837583e-02 2.29312509e-01 -5.01709104e-01 -2.26739585e-01
1.61356735e+00 1.31591126e-01 -5.28464988e-02 -6.80183709e-01
-1.29564524e+00 3.50699782e-01 4.84772593e-01 -2.93000668e-01
-3.76721591e-01 5.42237520e-01 7.79781640e-02 5.99681549e-02
6.60629749e-01 -5.34407087e-02 -5.34137428e-01 1.16742238e-01
-8.85568917e-01 7.68147647e-01 8.89535367e-01 1.19586849e+00
6.38054550e-01 -1.12821385e-01 -5.36946617e-02 8.18434834e-01
3.79384041e-01 8.35789323e-01 -4.05881673e-01 -1.06944907e+00
2.96291798e-01 3.85310709e-01 4.84017655e-02 -1.44282997e+00
-5.46012342e-01 -1.16652817e-01 -9.05628443e-01 -1.94025114e-01
3.82363111e-01 -1.42904729e-01 -1.11161518e+00 1.35754776e+00
6.36043072e-01 2.55465627e-01 -6.20130479e-01 1.38880277e+00
1.24239397e+00 3.10555458e-01 -2.07989857e-01 3.33203912e-01
1.36913550e+00 -7.11703122e-01 -9.53444093e-02 -2.45976180e-01
-3.03933024e-01 -8.15977812e-01 6.73176467e-01 4.29311931e-01
-1.59131992e+00 -9.44132507e-01 -8.45981359e-01 -1.94484636e-01
-5.61046833e-03 1.99105203e-01 5.52880526e-01 2.88740784e-01
-1.19633615e+00 1.21115315e+00 -1.09508872e+00 -6.87233508e-01
2.76214957e-01 6.15746737e-01 -3.94800484e-01 -2.00330138e-01
-8.41748655e-01 7.94688702e-01 7.11495355e-02 4.27674323e-01
-1.11227679e+00 -6.13604367e-01 -1.20735765e+00 -5.03805757e-01
2.60544717e-01 -1.18785334e+00 9.59570289e-01 -7.58420885e-01
-1.48140836e+00 1.25187492e+00 3.72468293e-01 -1.14441544e-01
7.78331578e-01 -3.37963015e-01 -4.23131138e-01 1.01335473e-01
-1.96335316e-01 6.76997542e-01 1.00463200e+00 -1.43540633e+00
1.27878413e-01 -4.34926122e-01 1.58686474e-01 6.78277463e-02
2.06163689e-01 -8.18373729e-03 -8.55118334e-01 -1.06457436e+00
4.70642120e-01 -1.08260715e+00 -2.71885067e-01 4.61614937e-01
-5.88733554e-01 4.28718299e-01 4.16334957e-01 -1.13823581e+00
7.28108823e-01 -1.80907893e+00 5.18863559e-01 4.86850202e-01
3.35970744e-02 -8.59727710e-02 -1.49573788e-01 2.46510446e-01
1.05359435e-01 -1.93655178e-01 -2.30840594e-01 -6.81541324e-01
2.01109782e-01 5.31834126e-01 2.43568540e-01 6.19481921e-01
-2.55728071e-03 9.89498198e-01 -8.06141317e-01 -5.32309830e-01
4.00510043e-01 7.59591281e-01 -6.29937589e-01 1.85847089e-01
-3.58636945e-01 4.40640301e-01 -2.04098120e-01 1.05529952e+00
8.27522337e-01 -2.06072386e-02 3.94293129e-01 -9.83133793e-01
2.42053121e-01 -1.07203096e-01 -1.24024856e+00 2.02060986e+00
-2.67600957e-02 3.69390398e-02 2.94939041e-01 -3.54682297e-01
1.10173750e+00 3.16268235e-01 5.29450774e-01 -9.66642126e-02
3.54737103e-01 -1.41460851e-01 -5.19107163e-01 -4.56769019e-01
4.38562244e-01 -1.48922950e-01 -2.40658388e-01 3.07573020e-01
1.28040120e-01 -4.86570984e-01 -7.69562498e-02 -3.89373422e-01
5.88647187e-01 9.54796255e-01 -3.39113809e-02 -5.97679801e-03
-3.62886600e-02 -2.51774788e-01 4.41753834e-01 4.77761775e-01
2.83311605e-01 1.16706312e+00 -2.15675101e-01 -1.25353241e+00
-1.38599479e+00 -1.36863267e+00 9.34960041e-03 1.02711749e+00
3.07752360e-02 -3.86015803e-01 -6.70013726e-01 -3.23853970e-01
3.76754552e-01 -1.77896619e-02 -7.16666639e-01 1.71538323e-01
-1.00260663e+00 -6.52117670e-01 3.80420595e-01 7.65126407e-01
6.37188256e-01 -9.13833380e-01 -3.15885186e-01 2.28255168e-01
-6.50189817e-02 -1.17589319e+00 -8.68945599e-01 -3.57287586e-01
-7.94407666e-01 -9.30977523e-01 -6.63614571e-01 -7.49787331e-01
8.63371372e-01 -2.83736944e-01 1.45834613e+00 3.51035565e-01
-4.13202375e-01 6.31982863e-01 -1.30298227e-01 -1.60644144e-01
-5.87948203e-01 -1.42217472e-01 3.69951695e-01 1.19914107e-01
-2.60632914e-02 -1.07237864e+00 -7.82271922e-01 5.31651914e-01
-5.69302499e-01 6.31986201e-01 6.56797141e-02 2.53010154e-01
9.17352140e-01 -3.87517601e-01 -1.33549765e-01 -4.10948426e-01
1.36272550e-01 -1.43403172e-01 -4.11046743e-01 1.20141149e-01
1.66770935e-01 -4.58427846e-01 2.83687681e-01 -5.44176757e-01
-4.08846110e-01 4.14481640e-01 -4.25898612e-01 -7.12540448e-01
-2.89369971e-01 1.17892027e-01 -6.58064112e-02 -2.49235332e-01
2.06903920e-01 -1.27247591e-02 1.79768056e-01 -9.56908524e-01
5.04533708e-01 6.30771518e-02 1.02793252e+00 -1.04407942e+00
9.78595853e-01 3.72584820e-01 2.40619168e-01 -3.41504723e-01
-7.92872131e-01 7.43068904e-02 -1.40035844e+00 -6.17015064e-01
9.03567612e-01 -8.04559708e-01 -7.11455822e-01 5.13130844e-01
-1.05707121e+00 -4.55537289e-01 -2.32850865e-01 2.50178725e-01
-7.57020473e-01 8.96005109e-02 -9.50529575e-01 -4.77080673e-01
-3.78188431e-01 -6.49213433e-01 1.46152925e+00 3.49558741e-02
-4.43374246e-01 -9.46655512e-01 -8.76911953e-02 2.98990220e-01
5.99151757e-03 1.14662647e+00 6.61874175e-01 -1.92499962e-02
-4.39806342e-01 -2.53631502e-01 1.43158317e-01 3.55760127e-01
3.06975603e-01 1.04889562e-02 -7.23715127e-01 -3.84531140e-01
-4.33046460e-01 -1.50719434e-01 3.05322528e-01 5.96994042e-01
1.18176055e+00 -3.21608871e-01 -2.46296436e-01 1.04277575e+00
1.38977492e+00 -3.75615954e-01 4.01225209e-01 1.01901546e-01
1.04452503e+00 3.99200916e-01 2.62976289e-01 5.08040369e-01
6.28508806e-01 9.75427091e-01 5.20173788e-01 -4.82395589e-01
-4.24626559e-01 -5.96819580e-01 2.39681229e-01 1.02988946e+00
-9.77862716e-01 2.98244745e-01 -7.21105635e-01 2.55938590e-01
-1.58200669e+00 -6.84963107e-01 -3.16339284e-02 2.16013718e+00
8.24163973e-01 2.47328300e-02 6.28183901e-01 -4.59599972e-01
6.65214360e-01 6.52295053e-02 -5.45557559e-01 -2.17410058e-01
1.15605459e-01 3.37975830e-01 4.69093293e-01 2.61451572e-01
-1.15305805e+00 7.47409523e-01 7.12448597e+00 1.03343002e-01
-8.43554556e-01 7.00158533e-03 3.52342784e-01 -3.76572430e-01
4.28328179e-02 -5.21649420e-01 -7.13699162e-01 2.76873082e-01
5.87067366e-01 9.35246572e-02 7.91097641e-01 7.00996757e-01
1.79275200e-01 4.45727766e-01 -1.33892441e+00 9.48968172e-01
3.84171486e-01 -1.30005229e+00 1.47655427e-01 -9.25625861e-02
6.21074498e-01 -9.25300568e-02 -2.25379869e-01 -3.00750416e-02
3.72663379e-01 -8.60982239e-01 1.04000998e+00 1.02058411e+00
1.16762114e+00 -5.97169340e-01 4.72242743e-01 8.47770125e-02
-1.41031837e+00 4.13154483e-01 -3.75629276e-01 -7.29864314e-02
4.27533358e-01 -9.44237807e-04 -8.30691993e-01 7.56258011e-01
8.66775990e-01 8.69033039e-01 -5.25218964e-01 7.92113960e-01
2.30140779e-02 3.28400820e-01 -5.56560636e-01 4.07786280e-01
-4.02673900e-01 -3.15237492e-01 5.84606886e-01 1.00587749e+00
3.97583336e-01 1.38960034e-01 5.38206339e-01 9.25844789e-01
-3.05742323e-01 -2.14428425e-01 -4.67423588e-01 2.48868898e-01
9.79033038e-02 1.38056636e+00 -7.86642194e-01 -4.10573095e-01
-1.59842789e-01 1.14952826e+00 -8.10463279e-02 1.71455652e-01
-8.85802507e-01 4.48033482e-01 6.92424595e-01 6.82181478e-01
2.39622265e-01 -3.82664621e-01 5.40410653e-02 -1.07560134e+00
3.90183628e-02 -8.03806663e-01 4.75690886e-02 -1.20659709e+00
-1.76737714e+00 6.43101513e-01 3.16500574e-01 -1.26348722e+00
1.05297752e-01 -6.29767299e-01 -2.58343935e-01 5.85098088e-01
-7.54126310e-01 -1.88966691e+00 -5.01535594e-01 6.33556843e-01
5.35263777e-01 2.73317397e-01 8.70976865e-01 4.44673032e-01
-2.03345895e-01 2.66418606e-01 -4.52955544e-01 5.60596943e-01
5.17339289e-01 -1.06655204e+00 1.02400148e+00 3.41424793e-01
1.07223846e-01 4.84620303e-01 5.79165936e-01 -1.02170217e+00
-1.57713246e+00 -1.49481797e+00 3.16073328e-01 -9.42366123e-01
1.59418687e-01 -8.15977216e-01 -5.18481135e-01 1.19810545e+00
-2.98157156e-01 3.10875714e-01 5.56141257e-01 -6.66871071e-02
1.86305270e-01 -1.83092475e-01 -1.34230852e+00 5.16200125e-01
1.59936309e+00 -2.67600268e-01 -7.94360876e-01 4.54663008e-01
6.50927663e-01 -1.22483361e+00 -1.40644860e+00 5.56049109e-01
1.22298586e+00 -6.93539500e-01 1.57282913e+00 -4.91104454e-01
5.48880517e-01 -2.24805608e-01 -4.05863166e-01 -1.27627921e+00
-4.29414868e-01 -5.01955330e-01 -3.26479822e-01 8.61810088e-01
-1.81950852e-01 -7.48719648e-02 8.22863638e-01 8.35785389e-01
-1.77797765e-01 -8.31198037e-01 -7.93513894e-01 -5.21683812e-01
-2.37231091e-01 -5.49307704e-01 1.04885197e+00 7.59097695e-01
-8.03950071e-01 -3.37079793e-01 -9.20081019e-01 4.19507027e-01
8.23786795e-01 2.65309542e-01 1.04484928e+00 -1.36250842e+00
-3.76130193e-01 2.52004385e-01 -4.73860115e-01 -1.15465212e+00
-2.20029548e-01 -1.05251765e+00 4.33374345e-02 -1.68874609e+00
1.48836911e-01 -3.72834712e-01 1.33061036e-01 6.80547237e-01
3.32446754e-01 1.14562738e+00 4.50793684e-01 -9.55004096e-02
-3.25819224e-01 1.67530388e-01 1.57658756e+00 -2.14031667e-01
-1.50761217e-01 4.21173200e-02 -3.82565856e-02 1.36370790e+00
3.32869977e-01 -3.45145583e-01 1.16519026e-01 -6.84233844e-01
2.31188267e-01 2.47680396e-01 9.99937534e-01 -1.04915023e+00
-1.43397182e-01 -9.68289599e-02 1.17080152e+00 -8.60027134e-01
7.47027755e-01 -9.53191996e-01 9.56657708e-01 2.19784930e-01
7.62131587e-02 1.70889705e-01 3.24834824e-01 2.55627245e-01
4.90054399e-01 2.74446875e-01 4.64383096e-01 -5.82594037e-01
-4.22092080e-01 7.77635396e-01 1.78232655e-01 -4.95660394e-01
5.90291560e-01 -5.73868513e-01 3.04904848e-01 -8.36370289e-02
-1.63032293e+00 -6.22945167e-02 8.78357887e-01 5.65059066e-01
8.94489050e-01 -1.80412412e+00 -7.18145072e-01 6.01854384e-01
-4.06942248e-01 2.34393343e-01 1.58146352e-01 5.41186094e-01
-8.19641650e-01 -2.63850063e-01 -4.37249273e-01 -7.13722110e-01
-1.48721349e+00 2.99791485e-01 6.10352755e-01 9.69836339e-02
-1.08825123e+00 6.41935945e-01 -2.08313793e-01 -8.74044716e-01
-8.94968063e-02 -5.10152876e-01 2.58719116e-01 -3.78720641e-01
1.62983850e-01 2.13585526e-01 -3.03874537e-02 -9.40831602e-01
-2.75199413e-01 1.21926880e+00 4.90830183e-01 2.40545377e-01
1.71183646e+00 -1.68530345e-01 -2.53241271e-01 4.09619421e-01
8.40012014e-01 1.02184191e-01 -1.59689748e+00 -6.19239919e-02
-6.02230966e-01 -7.09063411e-01 -3.14724445e-01 -8.99018288e-01
-1.19785559e+00 2.91706830e-01 5.48945010e-01 -3.88847068e-02
9.98953104e-01 1.33734360e-01 1.13458526e+00 4.37411517e-02
1.11373866e+00 -9.05607700e-01 -8.93213153e-02 1.95780441e-01
1.27249944e+00 -8.06829810e-01 2.85606623e-01 -5.81854701e-01
-2.34053582e-01 1.40986824e+00 5.50191402e-01 -4.68938202e-01
7.03776062e-01 4.25644070e-01 -1.31230816e-01 -4.50383782e-01
-1.16584130e-01 2.08161011e-01 7.07394242e-01 8.34189296e-01
1.47673041e-01 2.48461306e-01 2.68525958e-01 7.32955515e-01
-6.23455703e-01 3.22686613e-01 1.01939023e-01 8.47006440e-01
1.36029616e-01 -1.09504712e+00 -7.22140908e-01 4.52509671e-01
-2.34851658e-01 1.09695472e-01 -2.31940582e-01 8.58974040e-01
7.77838707e-01 3.04414958e-01 3.59831184e-01 -4.88649875e-01
7.12192655e-01 -1.32738769e-01 1.06331599e+00 -7.17725694e-01
-6.83707535e-01 2.01674521e-01 3.98816675e-01 -7.81329632e-01
-7.72739649e-01 -5.08115470e-01 -8.42347026e-01 -6.14222705e-01
1.46042004e-01 -5.10225892e-01 4.71325517e-01 7.40544379e-01
1.28408343e-01 3.57460529e-01 1.72115430e-01 -1.70065010e+00
3.12966630e-02 -7.16020763e-01 -8.01148236e-01 5.91276169e-01
2.49456555e-01 -6.39542043e-01 2.52456337e-01 6.75760508e-01] | [7.127345561981201, -1.2438609600067139] |
014c15ac-930f-489e-a9de-283a3af616c9 | learning-local-complex-features-using | 2007.05643 | null | https://arxiv.org/abs/2007.05643v2 | https://arxiv.org/pdf/2007.05643v2.pdf | Learning Local Complex Features using Randomized Neural Networks for Texture Analysis | Texture is a visual attribute largely used in many problems of image analysis. Currently, many methods that use learning techniques have been proposed for texture discrimination, achieving improved performance over previous handcrafted methods. In this paper, we present a new approach that combines a learning technique and the Complex Network (CN) theory for texture analysis. This method takes advantage of the representation capacity of CN to model a texture image as a directed network and uses the topological information of vertices to train a randomized neural network. This neural network has a single hidden layer and uses a fast learning algorithm, which is able to learn local CN patterns for texture characterization. Thus, we use the weighs of the trained neural network to compose a feature vector. These feature vectors are evaluated in a classification experiment in four widely used image databases. Experimental results show a high classification performance of the proposed method when compared to other methods, indicating that our approach can be used in many image analysis problems. | ['Jarbas Joaci de Mesquita Sá Junior', 'Leonardo F. S. Scabini', 'Lucas C. Ribas', 'Odemir M. Bruno'] | 2020-07-10 | null | null | null | null | ['texture-classification'] | ['computer-vision'] | [ 3.41195315e-01 -1.82746723e-01 -3.14488798e-01 -3.13538164e-01
5.32920621e-02 -5.52415252e-02 6.02748752e-01 1.80759355e-01
-4.30635124e-01 5.10858119e-01 -3.08092237e-01 -6.33855760e-02
-2.84456104e-01 -1.16291702e+00 -2.89964557e-01 -9.73408222e-01
-1.80291101e-01 4.66695994e-01 4.97098565e-01 -2.67896235e-01
5.04537880e-01 7.65214264e-01 -1.63579404e+00 3.78223151e-01
4.89639193e-01 1.40254486e+00 2.81383485e-01 4.29683417e-01
-3.39614540e-01 1.04847288e+00 -2.37812474e-01 -1.09715015e-02
1.68444440e-01 -3.62303972e-01 -8.03896427e-01 5.11308759e-02
1.04915770e-02 2.73582250e-01 -4.81562801e-02 9.21067536e-01
3.55135709e-01 1.88463643e-01 9.63346958e-01 -7.30480492e-01
-2.39963606e-01 4.39227223e-01 -4.09769267e-01 6.87641427e-02
1.25419483e-01 -3.35852146e-01 7.81209648e-01 -6.82309389e-01
6.43758357e-01 1.12011194e+00 6.66516960e-01 9.66852382e-02
-1.21579874e+00 -3.43155831e-01 -2.40052029e-01 6.29793346e-01
-1.48072898e+00 2.35042777e-02 1.15925181e+00 -5.02149582e-01
5.47898114e-01 2.54470438e-01 8.26459825e-01 7.62937248e-01
4.09185290e-01 6.63137734e-01 1.70624256e+00 -7.87138224e-01
3.27852249e-01 1.68695331e-01 1.88471988e-01 1.02085960e+00
1.66844606e-01 -1.06394194e-01 -1.29106820e-01 5.95760643e-02
8.94417763e-01 -3.68265398e-02 -9.25894752e-02 -5.18065631e-01
-9.56599236e-01 7.75118530e-01 6.61817014e-01 7.49234974e-01
-3.14640641e-01 -1.24833286e-01 3.50510091e-01 2.39814505e-01
3.26714158e-01 2.60911912e-01 -8.97882804e-02 1.84879139e-01
-7.27358818e-01 -3.27621996e-02 8.42378795e-01 1.58146217e-01
8.40193212e-01 -2.08462868e-02 1.83284935e-02 9.95694458e-01
1.19421221e-01 6.91058636e-01 7.31339455e-01 -3.75378758e-01
-6.18571416e-03 8.89982283e-01 -5.05577922e-01 -1.80249023e+00
-5.92569649e-01 -7.32861683e-02 -1.13678014e+00 7.59479523e-01
3.44427198e-01 2.67574638e-01 -7.69868672e-01 1.17542493e+00
1.85807511e-01 -1.50747955e-01 -9.93410051e-02 8.67908657e-01
6.78186893e-01 5.81287205e-01 -1.34335235e-01 4.66557173e-03
1.03879273e+00 -6.76778018e-01 -5.22063017e-01 4.01716083e-01
1.94829509e-01 -8.54961157e-01 8.28930080e-01 8.44146013e-01
-4.98918772e-01 -7.00875938e-01 -1.11981893e+00 3.69034171e-01
-6.63298965e-01 4.45795149e-01 6.71929300e-01 4.80536938e-01
-6.90974712e-01 9.50443566e-01 -7.42259026e-01 -5.13951361e-01
2.30823860e-01 4.76711392e-01 -5.66623390e-01 1.70678735e-01
-7.23305464e-01 8.33419323e-01 8.98690522e-01 3.81384850e-01
-3.39835644e-01 3.50863129e-01 -6.93651438e-01 8.31575990e-02
2.29745612e-01 -2.12431923e-01 6.39759600e-01 -1.32965183e+00
-1.78370249e+00 7.12181807e-01 9.87169221e-02 -9.56509113e-02
2.31098890e-01 3.34746510e-01 -4.41010952e-01 3.10017616e-01
-2.63250679e-01 2.01716110e-01 1.06984532e+00 -1.31223989e+00
-4.35222536e-01 -1.41785279e-01 -2.55781442e-01 -7.09640905e-02
-5.49136519e-01 -3.95878464e-01 -2.29530677e-01 -6.58159196e-01
5.50654650e-01 -7.92570889e-01 -1.03977032e-01 -7.98309743e-02
-4.17907387e-01 -3.56041759e-01 8.40173066e-01 -3.67837191e-01
1.12878704e+00 -1.96721876e+00 3.05960655e-01 1.17597449e+00
2.45244443e-01 4.05710101e-01 -3.97683829e-02 4.54811960e-01
-8.17639828e-02 -8.93061236e-02 -2.10786819e-01 7.91035965e-02
-3.49088073e-01 2.82071590e-01 3.20585579e-01 3.24137926e-01
9.86205712e-02 3.45310926e-01 -4.60175782e-01 -7.16474354e-01
3.78234953e-01 3.74666452e-01 -1.32109150e-01 1.76343516e-01
-1.35101184e-01 2.23550782e-01 -6.83884144e-01 5.56531489e-01
7.91901410e-01 -2.66818460e-02 5.78935921e-01 -5.38138866e-01
-3.75652909e-02 -4.27267402e-01 -1.32116449e+00 9.80601311e-01
-3.34882170e-01 6.50464237e-01 -3.22877258e-01 -1.42216265e+00
1.34847009e+00 4.89851385e-02 2.54422456e-01 -9.23450232e-01
4.74100024e-01 4.09008861e-01 1.52183352e-02 -9.79281902e-01
-2.85326447e-02 -8.50340798e-02 3.15476388e-01 3.25528711e-01
1.08738922e-01 9.49586034e-02 2.27897704e-01 -4.63072300e-01
7.98721492e-01 5.59743345e-02 2.97923595e-01 -4.15168196e-01
1.12089574e+00 -1.60459504e-01 4.11265045e-01 6.18915200e-01
2.76411951e-01 2.82195956e-01 7.22422063e-01 -1.11622226e+00
-1.17278337e+00 -6.23506963e-01 -2.65270263e-01 6.84578240e-01
1.28751263e-01 -5.19369133e-02 -8.33498180e-01 -6.54082954e-01
-1.03246450e-01 -1.41330644e-01 -8.08812082e-01 3.18726152e-02
-6.52605951e-01 -6.83121741e-01 4.08836752e-01 1.49438456e-01
6.90022528e-01 -1.37501335e+00 -4.46546316e-01 1.33355901e-01
9.26804543e-02 -7.78825939e-01 3.81453127e-01 9.88627374e-02
-1.08042443e+00 -1.27917624e+00 -5.99116802e-01 -1.14114606e+00
7.72492886e-01 8.60351101e-02 5.97584307e-01 3.13113064e-01
-3.45952600e-01 8.06378722e-02 -6.07295036e-01 -1.24502063e-01
-3.88919473e-01 2.15820253e-01 -2.46611591e-02 6.93733275e-01
1.68576911e-01 -5.22806227e-01 -3.24923724e-01 2.69670367e-01
-1.04776657e+00 -1.71821173e-02 1.09566581e+00 1.08649004e+00
6.11860514e-01 4.36718822e-01 2.68064111e-01 -9.43790436e-01
5.91407835e-01 -1.26014486e-01 -7.23877132e-01 3.68846923e-01
-4.95628834e-01 4.95981008e-01 7.25173950e-01 -3.80963355e-01
-1.00584269e+00 -4.20222571e-03 -1.76570043e-01 -9.90280136e-02
-2.53357410e-01 8.04520309e-01 3.56321372e-02 -7.70700991e-01
4.99588966e-01 3.29720110e-01 4.20408815e-01 -7.04562485e-01
-1.67642429e-01 6.21329844e-01 1.85397863e-01 -6.20176196e-01
5.17005682e-01 3.72318417e-01 4.91305560e-01 -9.96085465e-01
-3.60312611e-01 -2.41521716e-01 -6.81731641e-01 -4.50652272e-01
8.46254647e-01 -2.40246221e-01 -1.10977542e+00 7.90502727e-01
-7.52170622e-01 -2.93580025e-01 1.92350790e-01 6.48337662e-01
-5.39752603e-01 4.60471749e-01 -4.56154972e-01 -7.10288107e-01
-1.40397176e-01 -9.60624874e-01 4.74788755e-01 2.70004481e-01
2.98507780e-01 -1.28948700e+00 1.72624901e-01 -4.01707888e-02
5.37331641e-01 5.16407311e-01 1.27482402e+00 -4.71283883e-01
-3.62005383e-01 -4.34316754e-01 -2.98039466e-01 5.22624195e-01
5.16586639e-02 1.19659387e-01 -8.37777793e-01 -9.86489356e-02
8.13261494e-02 -3.06812644e-01 1.01288497e+00 8.59248340e-02
1.50319803e+00 -3.45705628e-01 -3.48021865e-01 5.71101844e-01
1.90877044e+00 3.12018126e-01 7.80839443e-01 6.71701908e-01
6.06273234e-01 5.39324105e-01 3.56323242e-01 2.59323001e-01
-1.77769382e-02 5.99346757e-01 3.08837861e-01 -3.62103492e-01
2.73157135e-02 2.15074164e-03 -1.55593872e-01 1.03239214e+00
-7.57037640e-01 -1.13348521e-01 -1.01418459e+00 -1.21744685e-02
-1.90535390e+00 -1.02104163e+00 3.42491753e-02 2.14641547e+00
4.60885674e-01 4.74282116e-01 -3.23066227e-02 6.37277305e-01
6.73821986e-01 3.16080190e-02 1.04819976e-01 -6.22893751e-01
-3.59666735e-01 5.98133981e-01 3.31606269e-01 3.25663805e-01
-1.25892174e+00 8.21656883e-01 6.30856848e+00 1.18594825e+00
-1.71639919e+00 -2.57616609e-01 5.74429572e-01 6.67181969e-01
9.28969905e-02 -2.12842345e-01 -1.48364007e-01 2.43281960e-01
3.85944724e-01 3.16433042e-01 2.97663718e-01 7.13487089e-01
1.53693587e-01 -4.23037142e-01 -6.05970323e-01 1.09851849e+00
2.16857538e-01 -1.11633980e+00 4.25421745e-01 -5.64464144e-02
4.86315340e-01 -4.14130688e-01 2.33949050e-01 -1.24356367e-01
-3.90524477e-01 -1.16015470e+00 2.48519897e-01 8.74512017e-01
6.34371340e-01 -7.67724335e-01 1.28264582e+00 1.39590040e-01
-1.13271189e+00 -1.95496038e-01 -6.18610740e-01 -2.70635307e-01
-4.28997844e-01 6.36618614e-01 -7.77250290e-01 6.86283648e-01
5.57242215e-01 6.78181112e-01 -8.20383728e-01 1.32789660e+00
3.33423279e-02 5.45286477e-01 -2.92890161e-01 -5.29039204e-01
3.02233249e-01 -4.64004964e-01 2.81347185e-01 1.21622193e+00
2.14957371e-01 -4.22942191e-02 8.07337761e-02 5.28889894e-01
3.02370280e-01 7.38235474e-01 -7.90453434e-01 -5.52328937e-02
-1.15938045e-01 1.50700271e+00 -1.27604795e+00 -3.55077654e-01
-1.27150163e-01 7.04157412e-01 4.15146500e-01 2.04002976e-01
-3.96411926e-01 -6.75985634e-01 -1.64732248e-01 2.57226098e-02
4.33721393e-01 -2.21292853e-01 -1.61751822e-01 -9.75429058e-01
-3.30516286e-02 -8.69009018e-01 8.40014517e-02 -6.48656547e-01
-1.15418184e+00 1.00734127e+00 -8.08984935e-02 -1.42358649e+00
7.42428079e-02 -1.27458775e+00 -6.84765279e-01 6.47474051e-01
-1.11266482e+00 -1.27283478e+00 -5.06408930e-01 5.79135895e-01
5.78988194e-02 -5.85961223e-01 9.67392147e-01 2.67722368e-01
-5.45202732e-01 3.97055030e-01 3.30862015e-01 4.10381556e-01
5.18147290e-01 -1.12187636e+00 -4.04131800e-01 4.63162035e-01
9.37612876e-02 3.16834539e-01 3.01084071e-01 -4.91795540e-01
-1.14638710e+00 -5.27133405e-01 9.31998432e-01 1.86705753e-01
3.65216821e-01 -2.80891985e-01 -7.76442111e-01 1.74032852e-01
1.03220068e-01 3.92114855e-02 5.84051073e-01 6.44281656e-02
-2.44528070e-01 -5.57458401e-01 -1.10416746e+00 2.65881181e-01
4.18383509e-01 -3.36846799e-01 -5.11093915e-01 7.75992870e-02
-2.02028155e-01 7.23516801e-03 -8.38416278e-01 5.09201467e-01
9.51587677e-01 -1.18235672e+00 6.13901138e-01 -2.96162099e-01
2.63855845e-01 -3.16091239e-01 3.09998896e-02 -1.12355602e+00
-5.36881387e-01 7.36347586e-02 2.86482394e-01 8.53695393e-01
2.47013733e-01 -6.71004772e-01 8.84874105e-01 -2.08870813e-01
3.62835884e-01 -8.77273500e-01 -6.90690458e-01 -6.13441408e-01
-2.17467472e-01 5.89100234e-02 2.91154802e-01 1.04041576e+00
3.67892277e-03 1.89059556e-01 -3.03164840e-01 -1.96806565e-01
7.00336218e-01 9.26011577e-02 4.11409765e-01 -1.63504589e+00
-1.76119089e-01 -7.21456766e-01 -1.17835474e+00 -2.83509612e-01
7.88653418e-02 -8.27479184e-01 -1.97089717e-01 -1.20897627e+00
1.96226656e-01 -8.12941730e-01 -4.76781487e-01 4.44045067e-01
1.50019661e-01 5.61323881e-01 4.15884443e-02 1.64620951e-01
-3.62008184e-01 3.10843557e-01 1.27600169e+00 -2.15459898e-01
-7.21320435e-02 2.88733114e-02 -9.81199965e-02 8.07474315e-01
9.22673643e-01 -3.80776256e-01 -1.54137462e-01 -1.24982953e-01
1.52412817e-01 -2.38597468e-01 2.92840719e-01 -1.51842880e+00
2.28238240e-01 -1.96095511e-01 8.49216223e-01 -2.00702071e-01
3.79483588e-02 -1.03757679e+00 1.68243095e-01 5.61986983e-01
-1.01304285e-01 -5.18657500e-03 -5.69019234e-03 3.96752685e-01
-6.23731792e-01 -4.60711807e-01 7.95236170e-01 -1.60790309e-01
-9.51996744e-01 1.23980664e-01 -5.41596591e-01 -6.17993116e-01
8.70267451e-01 -3.87048572e-01 -6.27898723e-02 -9.42584649e-02
-9.11557019e-01 -3.60025644e-01 3.11206102e-01 2.41020992e-01
7.46123135e-01 -1.47246170e+00 -2.60500103e-01 4.12593812e-01
2.34052494e-01 -4.59754944e-01 1.34760335e-01 8.18337023e-01
-1.33286059e+00 1.70350626e-01 -9.26142216e-01 -6.92578316e-01
-1.19566751e+00 5.66386223e-01 4.45462227e-01 -2.75955796e-01
-5.57631910e-01 3.10319155e-01 -2.52404064e-01 -3.65832239e-01
4.24633473e-02 -1.40731364e-01 -9.20227349e-01 -6.88481629e-02
2.36092746e-01 4.14050728e-01 1.19382441e-01 -7.37503111e-01
-4.17300276e-02 1.21789300e+00 2.11940408e-01 -6.87298179e-02
1.29737198e+00 2.50961006e-01 -5.49216032e-01 5.50668061e-01
1.45419502e+00 4.87115495e-02 -5.78750432e-01 -2.48604551e-01
2.00609431e-01 -2.78092057e-01 -2.09280085e-02 -6.17104292e-01
-1.00142980e+00 8.93435478e-01 1.08455515e+00 4.07046765e-01
1.26142788e+00 -4.49310988e-01 3.49114388e-01 8.30274582e-01
3.72242123e-01 -1.32320499e+00 2.01352596e-01 5.08351862e-01
6.25868738e-01 -1.28918552e+00 -1.48849711e-01 -4.25421715e-01
-3.69980544e-01 1.84643734e+00 3.85153562e-01 -4.67231065e-01
8.68256390e-01 4.38710637e-02 2.01654792e-01 -2.64562368e-01
-2.51514465e-01 -3.83072376e-01 4.31337595e-01 4.32607770e-01
3.46058041e-01 2.17535660e-01 -7.01258183e-01 1.00888520e-01
1.17588282e-01 3.20284367e-02 3.37274402e-01 9.28880036e-01
-7.05357254e-01 -1.42069256e+00 -4.21583027e-01 3.97405297e-01
-4.28482294e-01 2.85469562e-01 -2.45135903e-01 8.15721691e-01
5.00827670e-01 5.58280885e-01 6.06901497e-02 -6.95103467e-01
2.20004484e-01 -6.10366603e-03 6.31529629e-01 -9.06117037e-02
-3.42362791e-01 1.41667537e-02 -1.11563779e-01 -4.93060023e-01
-7.26029098e-01 -2.27543995e-01 -8.50896358e-01 -2.28827536e-01
-2.86764592e-01 3.43195409e-01 8.98232639e-01 8.61220658e-01
-1.30006447e-01 4.09269810e-01 8.91566336e-01 -1.01501024e+00
-1.07228301e-01 -8.90606701e-01 -7.80375183e-01 5.81528187e-01
2.01734230e-01 -1.03572834e+00 -1.78310692e-01 -1.06094606e-01] | [10.29839038848877, -0.42996010184288025] |
9824b5b2-fced-4ba2-8074-1687027150c5 | an-algorithmic-framework-for-the-optimization | 2303.12797 | null | https://arxiv.org/abs/2303.12797v1 | https://arxiv.org/pdf/2303.12797v1.pdf | An algorithmic framework for the optimization of deep neural networks architectures and hyperparameters | In this paper, we propose an algorithmic framework to automatically generate efficient deep neural networks and optimize their associated hyperparameters. The framework is based on evolving directed acyclic graphs (DAGs), defining a more flexible search space than the existing ones in the literature. It allows mixtures of different classical operations: convolutions, recurrences and dense layers, but also more newfangled operations such as self-attention. Based on this search space we propose neighbourhood and evolution search operators to optimize both the architecture and hyper-parameters of our networks. These search operators can be used with any metaheuristic capable of handling mixed search spaces. We tested our algorithmic framework with an evolutionary algorithm on a time series prediction benchmark. The results demonstrate that our framework was able to find models outperforming the established baseline on numerous datasets. | ['Gilles Cabriel', 'Sandra Claudel', 'El-Ghazali Talbi', 'Julie Keisler'] | 2023-02-27 | null | null | null | null | ['time-series-prediction'] | ['time-series'] | [ 1.93336323e-01 1.17841721e-01 1.36798888e-01 -2.71258593e-01
1.83085173e-01 -4.78585273e-01 7.73387551e-01 -2.15470530e-02
-8.87029409e-01 6.96121335e-01 -2.18464583e-01 -2.45955124e-01
-6.89184725e-01 -9.58573222e-01 -5.93228281e-01 -8.42372894e-01
-4.04203147e-01 6.42934024e-01 2.98639715e-01 -4.06758338e-01
4.16667163e-01 9.10899878e-01 -2.09878683e+00 1.88985720e-01
1.00666928e+00 9.74687576e-01 1.28179088e-01 5.22415280e-01
-2.49193653e-01 3.42675775e-01 -5.63155174e-01 -6.45085156e-01
3.45937222e-01 -1.65969729e-01 -6.21477783e-01 -3.41155350e-01
4.70165312e-02 3.86670470e-01 5.79333445e-03 9.21079636e-01
5.81444621e-01 4.84083742e-01 5.71680963e-01 -1.22951460e+00
-4.52525735e-01 9.47490036e-01 7.23683313e-02 5.39863765e-01
-1.57049149e-01 1.06287651e-01 9.81540918e-01 -5.83277047e-01
6.99622095e-01 1.20939517e+00 7.46573687e-01 7.64825284e-01
-1.35893428e+00 -5.24749160e-01 1.57349836e-02 7.09463298e-01
-1.22460186e+00 -3.26604664e-01 6.66537523e-01 -2.42924809e-01
1.42430186e+00 5.58474481e-01 1.07497859e+00 1.30483425e+00
2.75575399e-01 6.36105835e-01 5.73785663e-01 -6.54902160e-01
6.05382681e-01 1.81887615e-02 3.05027694e-01 7.29192138e-01
2.84138024e-01 2.70179570e-01 -3.14024419e-01 -1.36426702e-01
4.83795017e-01 -2.58451283e-01 -1.78180963e-01 -4.29617733e-01
-7.92698443e-01 9.91425514e-01 4.00334984e-01 8.01356077e-01
-6.11303210e-01 2.22962618e-01 3.72772872e-01 3.55993569e-01
3.58122855e-01 1.05980551e+00 -6.29055262e-01 4.07365374e-02
-6.79077983e-01 3.73312205e-01 1.00332344e+00 3.51958573e-01
5.71807742e-01 2.62469947e-01 -4.03623551e-01 8.88470411e-01
-3.14061940e-02 -1.81295082e-01 9.75152075e-01 -7.35992610e-01
1.82827488e-01 8.11432242e-01 -2.71005511e-01 -1.03594053e+00
-7.39865780e-01 -8.32729101e-01 -9.42274570e-01 2.08307013e-01
3.44579220e-02 -4.73986596e-01 -1.01361310e+00 1.69498003e+00
2.02670455e-01 4.07417923e-01 1.67213138e-02 5.23463607e-01
6.24950886e-01 4.79296386e-01 -7.28907362e-02 -2.73436517e-01
8.49350452e-01 -1.03432667e+00 -6.60806894e-01 -7.62712434e-02
5.12752295e-01 -2.01360494e-01 6.28648102e-01 4.14732695e-01
-1.29156959e+00 -5.25003195e-01 -1.17739892e+00 5.00043809e-01
-6.42814279e-01 -1.18385032e-02 8.23151588e-01 8.01545978e-01
-1.70077360e+00 1.21453238e+00 -9.01387095e-01 -1.93414837e-01
3.36029202e-01 6.30297244e-01 -7.36391032e-03 5.45055926e-01
-1.26824141e+00 1.12304008e+00 1.11960411e+00 4.13807839e-01
-4.11472261e-01 -6.00501597e-01 -5.36624491e-01 4.32839960e-01
2.94250309e-01 -1.01820576e+00 1.01596522e+00 -1.23070896e+00
-1.76632941e+00 7.18039751e-01 2.11959273e-01 -1.15434289e+00
6.41770005e-01 1.19926810e-01 -3.61553848e-01 -2.03208849e-01
-8.32778752e-01 7.71261632e-01 9.17941332e-01 -6.77257419e-01
-4.23541427e-01 -2.49999851e-01 -1.81037232e-01 6.52021095e-02
-7.99419165e-01 -6.78797066e-02 -3.12626898e-01 -7.82621324e-01
-1.29606873e-01 -1.13505554e+00 -5.33722758e-01 -4.53858763e-01
-3.86039078e-01 -3.65771085e-01 4.76978660e-01 -2.09201321e-01
1.54392648e+00 -1.72310722e+00 8.45308840e-01 4.69316572e-01
8.08350146e-02 6.15708351e-01 -2.83993542e-01 2.83208102e-01
-4.26583230e-01 2.90111601e-01 -2.40714133e-01 -2.22015813e-01
1.60531744e-01 3.83066833e-01 -4.03126180e-02 3.36882733e-02
1.24706142e-01 1.03860450e+00 -5.08110106e-01 -2.88480520e-01
-7.40948766e-02 6.73951805e-01 -7.68469334e-01 -6.26590848e-02
-4.63909268e-01 -1.91929955e-02 -2.48390555e-01 1.84974000e-01
3.13070804e-01 -9.74083915e-02 2.67987818e-01 2.16219369e-02
-2.46352628e-01 -1.56664848e-01 -1.16012025e+00 1.57602489e+00
-4.73898321e-01 6.48303866e-01 -3.39658469e-01 -1.14885747e+00
1.21880007e+00 1.02341004e-01 3.96097988e-01 -5.51014841e-01
5.47651410e-01 2.79347688e-01 2.81383365e-01 -4.67118680e-01
3.82556975e-01 3.33108664e-01 5.37186801e-01 3.79193783e-01
5.06525517e-01 1.79774269e-01 6.25388384e-01 -4.65888679e-01
1.10435426e+00 -5.29088788e-02 -1.83190331e-02 -5.21144509e-01
8.50160003e-01 -3.25834036e-01 3.45783085e-01 8.29196632e-01
2.91456908e-01 1.50104359e-01 5.29160976e-01 -9.81910765e-01
-9.60663795e-01 -7.35048115e-01 4.66833590e-03 1.16454649e+00
-5.36985338e-01 -3.94401968e-01 -9.80532527e-01 -5.52388489e-01
-2.83882767e-01 8.40844750e-01 -8.93736660e-01 -4.11163509e-01
-6.62441730e-01 -1.26018655e+00 5.13910592e-01 8.44310820e-02
3.74166489e-01 -1.56164086e+00 -1.07081997e+00 3.44195813e-01
3.23306948e-01 -6.01490021e-01 1.84942950e-02 6.21558726e-01
-1.23635840e+00 -8.21440876e-01 -5.99713504e-01 -7.03048110e-01
4.21573520e-01 -6.32172823e-01 1.36336005e+00 3.84832978e-01
-6.10637069e-01 -7.65418727e-03 -3.47894400e-01 -5.57987571e-01
-5.95854342e-01 6.49990320e-01 -1.10713460e-01 3.20665389e-01
5.97955994e-02 -9.94026542e-01 -4.54627246e-01 2.37622578e-02
-1.07283676e+00 4.01296216e-04 5.57714939e-01 7.90503383e-01
4.44574296e-01 1.79306626e-01 4.23389137e-01 -6.50625706e-01
1.07426620e+00 -4.31730837e-01 -9.29602563e-01 4.38724965e-01
-9.79913950e-01 7.88371563e-01 6.44782603e-01 -6.01569593e-01
-7.75496602e-01 1.37403399e-01 -2.35502064e-01 -5.81964791e-01
-7.47691542e-02 4.11810726e-01 1.10056929e-01 -3.54294568e-01
8.41425061e-01 1.46278828e-01 -2.40952298e-01 -5.60333610e-01
1.62586197e-01 1.15629278e-01 3.07835370e-01 -4.61287886e-01
4.55935359e-01 9.42915902e-02 2.25941464e-01 -5.70796609e-01
-3.18977922e-01 1.48840100e-01 -6.02351904e-01 -2.14629799e-01
7.67069101e-01 -6.24386929e-02 -7.68570781e-01 5.13396442e-01
-1.26013148e+00 -1.98970467e-01 -3.61293793e-01 2.95553654e-01
-3.31118286e-01 -1.96074709e-01 -3.69895935e-01 -7.81687081e-01
-6.08003259e-01 -1.17376828e+00 4.30467039e-01 3.91125947e-01
-8.10715556e-02 -1.14360011e+00 2.67036796e-01 -3.57666522e-01
7.54754782e-01 4.04403776e-01 1.03947139e+00 -1.04283106e+00
-2.36076236e-01 8.35711285e-02 3.09568107e-01 2.76815087e-01
-3.84349376e-01 4.49456722e-01 -7.61304855e-01 -9.01913643e-02
-5.29004224e-02 1.21974446e-01 1.05235577e+00 3.80873322e-01
1.40107262e+00 -4.83369917e-01 -3.55717719e-01 1.02101386e+00
1.40956819e+00 6.40979350e-01 7.38232195e-01 8.02341461e-01
2.51316667e-01 6.66379392e-01 -3.01181190e-02 4.84108597e-01
-1.54780269e-01 6.89292848e-01 7.40859032e-01 2.60111123e-01
3.96833390e-01 2.65157014e-01 7.19794482e-02 4.78997648e-01
-4.46645141e-01 -4.15301114e-01 -1.19378638e+00 3.85693103e-01
-1.90190685e+00 -1.01885295e+00 3.39715630e-02 1.88966775e+00
6.55553341e-01 1.73488766e-01 1.45570949e-01 2.05568805e-01
8.34903121e-01 4.64249263e-03 -6.22866571e-01 -9.21090186e-01
-1.93532318e-01 8.53834152e-01 3.43239963e-01 2.38118291e-01
-1.04475617e+00 7.19650149e-01 6.42944336e+00 6.82957709e-01
-1.16458750e+00 -1.45719618e-01 6.91022098e-01 -5.55489779e-01
-3.31255138e-01 -4.75724459e-01 -7.82362163e-01 3.25922430e-01
1.23188031e+00 -2.48460993e-01 6.60742521e-01 7.16717482e-01
-1.86863244e-01 5.08313775e-01 -8.99898827e-01 9.72771704e-01
-4.97545227e-02 -1.77457285e+00 4.02119607e-02 -6.52770624e-02
8.56669009e-01 2.35528931e-01 8.11818317e-02 2.97889411e-01
4.43854392e-01 -1.19761348e+00 6.81922853e-01 7.19288766e-01
8.58997330e-02 -1.10559177e+00 6.16937578e-01 2.47742251e-01
-7.22169578e-01 -5.80182970e-01 -9.15401801e-02 2.68388689e-01
-1.61825493e-01 3.10902655e-01 -6.73458099e-01 5.86542070e-01
7.96073020e-01 4.39012766e-01 -8.80333006e-01 1.45389605e+00
4.40501384e-02 1.38573259e-01 -3.93223912e-01 -6.48832977e-01
4.92268920e-01 -3.84777546e-01 7.36559093e-01 1.19439566e+00
7.47626305e-01 -3.32757473e-01 -5.56858122e-01 9.42849874e-01
2.93721557e-02 2.38690108e-01 -4.50524926e-01 7.29984120e-02
4.09437418e-01 1.24028873e+00 -1.08546662e+00 -3.61272953e-02
3.35781246e-01 8.56433749e-01 4.50029880e-01 2.84419388e-01
-9.41225410e-01 -4.53070968e-01 5.08022726e-01 -1.88965857e-01
5.93902886e-01 -4.06560767e-03 -1.29891217e-01 -1.00918961e+00
-1.61002681e-01 -8.49476814e-01 5.96486270e-01 -5.66262662e-01
-8.81741703e-01 1.31643283e+00 1.22052468e-01 -7.23800838e-01
-6.47451758e-01 -7.40407825e-01 -6.38350964e-01 6.22868538e-01
-1.23297310e+00 -4.74592179e-01 -1.06162786e-01 5.54701567e-01
3.69347692e-01 -6.43701136e-01 8.03990901e-01 1.74278706e-01
-9.12646174e-01 5.78780472e-01 3.28591466e-02 -4.85320628e-01
-3.41487676e-02 -1.14938152e+00 4.45968121e-01 7.82379746e-01
4.32924837e-01 5.00988603e-01 9.27157462e-01 -1.75123483e-01
-1.20015264e+00 -8.38282824e-01 7.46927381e-01 1.10211164e-01
7.38681376e-01 -2.81641424e-01 -9.23200309e-01 4.17387396e-01
4.82333153e-01 -2.20107585e-01 3.21565449e-01 1.02703691e-01
-1.91410724e-03 -1.39208093e-01 -9.52546895e-01 6.87375784e-01
1.33550274e+00 2.60623991e-01 -3.06964874e-01 2.60065556e-01
6.45307481e-01 -4.89741653e-01 -6.55455709e-01 6.87360764e-01
4.30715233e-01 -1.29822516e+00 1.04739380e+00 -6.94030404e-01
1.60209537e-01 2.65751660e-01 2.57390290e-01 -1.66265202e+00
-3.23229760e-01 -9.75353003e-01 -4.12779897e-01 9.87800956e-01
6.46373570e-01 -8.49187613e-01 7.31763959e-01 4.08983469e-01
-2.77326822e-01 -1.03988492e+00 -1.04040444e+00 -8.53158593e-01
-8.23230445e-02 -3.89438778e-01 9.71514225e-01 8.61771107e-01
-5.33647239e-01 1.61987349e-01 1.00760728e-01 -1.55282632e-01
2.63507932e-01 -2.07950175e-02 8.76236185e-02 -1.72393179e+00
-5.27582586e-01 -1.23642170e+00 -4.40872937e-01 9.99696404e-02
5.56564808e-01 -1.05075347e+00 -3.07972729e-01 -1.10795891e+00
-4.72606778e-01 -4.35145110e-01 -4.98595089e-01 5.47484756e-01
9.95571092e-02 -9.75239351e-02 2.11258888e-01 -2.42620334e-01
-3.59940320e-01 6.51275635e-01 7.77241230e-01 3.84317301e-02
-5.13840616e-01 7.27137551e-02 -3.22608292e-01 6.22758448e-01
1.11666119e+00 -5.20580113e-01 -4.54398066e-01 -5.24147749e-01
5.23960173e-01 -3.04300189e-01 3.47936243e-01 -1.26987278e+00
2.41895363e-01 9.45786685e-02 1.72187731e-01 -1.92339957e-01
6.04640171e-02 -6.82075500e-01 7.33948648e-01 8.58964026e-01
-4.98782188e-01 5.09086728e-01 4.43777442e-01 2.62368262e-01
-1.06285416e-01 -6.50063157e-01 7.30938554e-01 -9.53754187e-02
-7.84422815e-01 2.89487749e-01 -2.69771874e-01 -1.83897763e-01
1.16031241e+00 -2.18935877e-01 -5.95473051e-02 1.02790527e-01
-1.01919627e+00 6.10554628e-02 1.37288630e-01 7.22856104e-01
4.47748125e-01 -1.28704929e+00 -6.76967919e-01 3.35747451e-01
-3.36119533e-01 -3.57047170e-01 1.40325934e-01 5.47090888e-01
-6.71317220e-01 4.91851240e-01 -5.60096085e-01 -4.38911676e-01
-1.38259435e+00 5.55943370e-01 9.00943756e-01 -6.63841665e-01
-3.57964724e-01 9.63866830e-01 -4.59525645e-01 -4.78914112e-01
5.12974441e-01 -3.56341094e-01 -7.85471916e-01 3.99034202e-01
3.10181051e-01 5.68837106e-01 5.02053559e-01 -1.67044550e-01
-3.88793707e-01 2.95586884e-01 2.29333013e-01 -1.05975397e-01
1.91528773e+00 3.91543657e-01 -3.36338729e-01 1.39022082e-01
1.17840374e+00 -6.72188461e-01 -9.93686259e-01 1.16432548e-01
3.63972962e-01 1.30536616e-01 2.87862539e-01 -7.19770849e-01
-1.58052456e+00 5.68498611e-01 7.76308239e-01 7.09415197e-01
1.42624795e+00 -4.44320291e-01 3.73901218e-01 7.60778129e-01
1.92795187e-01 -1.04250801e+00 -8.47397074e-02 7.85504162e-01
1.08875179e+00 -5.68402231e-01 -4.98849422e-01 2.08727673e-01
-2.57554770e-01 1.57269907e+00 6.25963509e-01 -2.44169816e-01
5.57779670e-01 3.84610713e-01 -4.28654492e-01 -1.85131282e-01
-1.27735209e+00 -2.81720877e-01 5.65842986e-01 4.63788390e-01
1.65975943e-01 -3.68814021e-01 -7.22681463e-01 4.48452055e-01
-3.60153824e-01 6.41677231e-02 2.07737327e-01 6.55272663e-01
-1.80794194e-01 -1.26359355e+00 -1.22782290e-01 4.19224799e-01
-4.17061567e-01 -8.58388022e-02 -3.79595935e-01 7.87041187e-01
3.32222313e-01 5.09208620e-01 1.57293648e-01 -5.47785282e-01
4.83849376e-01 2.54497290e-01 6.54154599e-01 -2.95564502e-01
-1.28948426e+00 -3.72506976e-01 1.47855356e-01 -5.67484200e-01
-3.21811616e-01 -6.49413884e-01 -9.68963623e-01 -1.29926533e-01
-2.92690814e-01 2.37628296e-01 9.45548654e-01 8.06082666e-01
5.85915387e-01 8.88498247e-01 4.70527411e-01 -1.13536441e+00
-3.87812823e-01 -7.60586143e-01 -2.66684026e-01 3.87286767e-02
3.40395011e-02 -6.05023205e-01 -3.04738939e-01 -2.90834248e-01] | [8.278223037719727, 3.2424302101135254] |
66dacbf3-71ba-4b27-a70d-bca58e3d49a9 | quantile-filtered-imitation-learning | 2112.00950 | null | https://arxiv.org/abs/2112.00950v1 | https://arxiv.org/pdf/2112.00950v1.pdf | Quantile Filtered Imitation Learning | We introduce quantile filtered imitation learning (QFIL), a novel policy improvement operator designed for offline reinforcement learning. QFIL performs policy improvement by running imitation learning on a filtered version of the offline dataset. The filtering process removes $ s,a $ pairs whose estimated Q values fall below a given quantile of the pushforward distribution over values induced by sampling actions from the behavior policy. The definitions of both the pushforward Q distribution and resulting value function quantile are key contributions of our method. We prove that QFIL gives us a safe policy improvement step with function approximation and that the choice of quantile provides a natural hyperparameter to trade off bias and variance of the improvement step. Empirically, we perform a synthetic experiment illustrating how QFIL effectively makes a bias-variance tradeoff and we see that QFIL performs well on the D4RL benchmark. | ['Joan Bruna', 'Rajesh Ranganath', 'William F. Whitney', 'David Brandfonbrener'] | 2021-12-02 | null | null | null | null | ['d4rl'] | ['robots'] | [-1.76597103e-01 2.86142290e-01 -7.78789043e-01 -1.95513859e-01
-1.06249642e+00 -8.46281230e-01 5.61913490e-01 1.57983556e-01
-8.30074310e-01 1.16543984e+00 1.24293759e-01 -4.67601508e-01
-4.44443375e-01 -6.63446784e-01 -1.01621342e+00 -5.82394660e-01
-5.38437903e-01 3.20840359e-01 1.87934801e-01 -1.12573057e-01
4.76673841e-01 3.15469503e-01 -1.23842847e+00 3.23569635e-03
1.13961506e+00 1.01782691e+00 -1.53817773e-01 8.23927760e-01
1.89981282e-01 6.88786030e-01 -1.03920960e+00 -1.08966030e-01
8.14687729e-01 -4.85004157e-01 -6.20870411e-01 -1.54708207e-01
9.53457728e-02 -8.19412589e-01 -1.61498487e-01 1.14323545e+00
3.41926813e-01 4.40893680e-01 3.52777332e-01 -1.47304404e+00
-2.45939568e-01 9.90711749e-01 -3.00167650e-01 4.38518524e-01
2.02847362e-01 7.06002593e-01 9.46594238e-01 -3.48205894e-01
5.84475994e-01 1.40680039e+00 5.82641959e-01 4.20929223e-01
-1.48089588e+00 -5.89834988e-01 2.75203347e-01 -2.53426969e-01
-8.73474479e-01 -7.41926059e-02 2.06245869e-01 -2.58739412e-01
8.84507298e-01 -1.02570534e-01 7.83389926e-01 8.71689200e-01
5.35419583e-01 7.79382825e-01 1.22262740e+00 -2.87043657e-02
7.75911033e-01 3.22418213e-02 -3.89564097e-01 5.26177943e-01
1.23002745e-01 7.60226846e-01 -3.96416485e-01 -4.51645702e-01
1.12752819e+00 -1.34677321e-01 -2.92410590e-02 -6.81609631e-01
-9.61422503e-01 7.85090506e-01 2.48103872e-01 -2.80925661e-01
-4.71697748e-01 8.49401593e-01 4.67707396e-01 9.69363630e-01
1.53379992e-01 7.52146423e-01 -6.71951413e-01 -7.60590792e-01
-4.17321712e-01 6.74756408e-01 8.49243879e-01 8.68036568e-01
5.98089516e-01 1.10676549e-01 -6.55932724e-01 4.89624381e-01
-1.36692107e-01 7.81700253e-01 3.22927475e-01 -1.73092294e+00
4.22095537e-01 3.14421743e-01 9.48845923e-01 -1.49754614e-01
-7.63768554e-02 -3.47530693e-01 1.43930405e-01 6.20706260e-01
9.09569144e-01 -6.61349177e-01 -6.96653545e-01 1.99913657e+00
3.27749401e-01 -1.56175867e-01 -3.35403495e-02 6.48609877e-01
-4.29247677e-01 3.98555040e-01 1.64864093e-01 -4.41034049e-01
4.94666785e-01 -7.14436114e-01 -5.05273342e-01 1.14360057e-01
4.54833984e-01 -2.19875976e-01 1.51749349e+00 3.22985530e-01
-1.27921581e+00 -3.43085468e-01 -9.04988170e-01 5.23023486e-01
1.58747971e-01 -1.32424071e-01 2.52307951e-01 4.46910053e-01
-8.52154851e-01 1.30156541e+00 -9.02101994e-01 -8.84406120e-02
5.05831420e-01 5.16779363e-01 8.15166607e-02 5.21317363e-01
-9.33955669e-01 9.33829367e-01 4.65400875e-01 -6.40876234e-01
-1.39466274e+00 -1.00212646e+00 -3.65256757e-01 1.01665862e-01
8.67100239e-01 -4.25538242e-01 1.98673689e+00 -9.95685399e-01
-2.01561999e+00 1.93083942e-01 4.47775990e-01 -1.10932410e+00
9.02284324e-01 -4.89867061e-01 3.25740390e-02 1.53454229e-01
-3.95588093e-02 3.76906335e-01 1.28622508e+00 -1.05109072e+00
-1.07670045e+00 -1.28284559e-01 1.60580516e-01 1.49850085e-01
2.67997347e-02 -4.29697990e-01 2.37776354e-01 -4.79376614e-01
-6.10846817e-01 -7.61541843e-01 -4.44898874e-01 -7.67394453e-02
-8.21450576e-02 -3.81758362e-01 4.58012164e-01 -2.76293129e-01
1.22745728e+00 -1.91648829e+00 -2.38974363e-01 5.03869712e-01
-9.36287567e-02 1.24962591e-01 -2.39729151e-01 6.26657307e-01
2.12065920e-01 1.60546958e-01 -1.77700683e-01 2.72831261e-01
4.33605880e-01 2.92219698e-01 -7.57472813e-01 5.96604466e-01
-2.68834177e-02 7.02891231e-01 -1.18007219e+00 9.69786476e-03
-7.51960836e-03 -2.34765187e-01 -9.08722639e-01 4.84496742e-01
-7.84796476e-01 4.87771899e-01 -5.67953229e-01 2.99716622e-01
2.74635136e-01 1.18501514e-01 2.13435352e-01 4.10875231e-01
-2.81478018e-01 2.35199228e-01 -1.18111515e+00 1.29902744e+00
-2.34974891e-01 -8.98263603e-02 1.00926198e-01 -8.86583984e-01
9.95551169e-01 -1.12756580e-01 7.31682122e-01 -6.94623470e-01
2.18536496e-01 2.75044709e-01 7.64548928e-02 -4.75737512e-01
3.70625168e-01 -2.77138650e-01 -5.14017530e-02 7.63747454e-01
6.01877132e-03 -3.40250313e-01 2.67859519e-01 1.21239750e-02
1.26428151e+00 4.83087599e-01 4.43937033e-01 -5.42524993e-01
1.68971121e-01 9.78275947e-03 5.66955447e-01 1.27310801e+00
-7.83057094e-01 -2.02247843e-01 1.17051637e+00 -1.71978906e-01
-1.31061268e+00 -1.41974485e+00 6.46528825e-02 1.40975416e+00
7.10761920e-02 -1.34587988e-01 -7.64981210e-01 -1.13334537e+00
9.19694364e-01 8.45189214e-01 -8.09949398e-01 -4.48376060e-01
-6.01142347e-01 -1.69140071e-01 3.87471378e-01 6.17847741e-01
4.07660753e-01 -9.78574932e-01 -1.01181388e+00 3.66275042e-01
4.99826878e-01 -6.03807747e-01 -8.53285730e-01 2.74027556e-01
-9.87792671e-01 -1.15029633e+00 -3.65215361e-01 -1.87391713e-01
4.99957860e-01 -3.20495397e-01 1.00817001e+00 -3.40082556e-01
2.74970829e-01 6.93521678e-01 -1.33310616e-01 -4.08236414e-01
-6.16318226e-01 -1.73572302e-02 2.50797689e-01 -3.03393871e-01
-8.17932412e-02 -5.44535577e-01 -8.13522160e-01 2.09083647e-01
-6.95538402e-01 -5.96654713e-01 2.09377289e-01 1.04444087e+00
5.45074999e-01 -1.91623002e-01 9.98078644e-01 -6.75914884e-01
1.18733776e+00 -3.66659254e-01 -1.23795927e+00 1.51684448e-01
-9.52814758e-01 5.38488865e-01 1.03449798e+00 -6.03789985e-01
-9.27698016e-01 -1.82214811e-01 5.52219935e-02 -3.97550523e-01
2.83926845e-01 6.13509230e-02 2.08699957e-01 2.01225922e-01
7.27937400e-01 3.91071625e-02 5.85106015e-01 -4.37914759e-01
5.44187248e-01 3.15228999e-01 5.09278595e-01 -1.14109361e+00
5.97892284e-01 3.49848151e-01 -1.80725455e-01 -1.51503265e-01
-6.35013163e-01 -1.57273971e-02 -1.34519979e-01 -1.75671250e-01
1.79275215e-01 -3.95015538e-01 -1.42688513e+00 -5.21260053e-02
-4.27111357e-01 -1.09040582e+00 -1.26283514e+00 2.82233059e-01
-1.27947974e+00 -1.32106602e-01 -4.85239625e-01 -1.18905699e+00
-2.84652174e-01 -9.78807747e-01 6.87742889e-01 5.01117349e-01
5.56271076e-02 -6.90482199e-01 3.46845329e-01 -5.15817523e-01
4.21691895e-01 1.92062497e-01 9.17318463e-01 -6.59758866e-01
-4.31821913e-01 2.43875057e-01 1.23001464e-01 4.28587556e-01
-2.28053192e-03 -2.39687413e-02 -4.77683753e-01 -8.21300387e-01
-2.52779871e-01 -5.20903230e-01 7.09145486e-01 4.74630952e-01
1.27646041e+00 -6.75666153e-01 -3.70368063e-02 4.67324108e-01
1.23763239e+00 6.31333530e-01 1.76093787e-01 5.61914742e-01
8.18360075e-02 -2.23955642e-02 9.99371052e-01 1.06604207e+00
9.96900350e-03 3.03172290e-01 5.49879730e-01 5.20245969e-01
4.48898733e-01 -7.97842860e-01 7.73553729e-01 -7.97530115e-02
1.75650716e-01 1.46033123e-01 -5.87805092e-01 4.79951054e-01
-1.94814587e+00 -1.02378440e+00 9.58082497e-01 2.73554587e+00
1.21152365e+00 5.52725315e-01 9.05395389e-01 -1.45588279e-01
3.76491219e-01 -1.21038653e-01 -1.30960011e+00 -7.83946276e-01
3.99364203e-01 5.98832846e-01 1.06603289e+00 6.93547130e-01
-9.04924273e-01 8.57350647e-01 7.44179869e+00 7.22180486e-01
-9.37829077e-01 -2.33742759e-01 4.93659556e-01 -3.48681033e-01
-4.10639256e-01 -6.44249320e-02 -6.78313494e-01 5.80463767e-01
1.10455799e+00 -7.11790204e-01 8.37942839e-01 1.13435805e+00
5.68469405e-01 -1.89525813e-01 -1.14099979e+00 4.78860587e-01
-9.33348417e-01 -1.27135098e+00 -3.61420870e-01 -1.75009649e-02
8.83744717e-01 6.53565526e-02 2.67531395e-01 8.04545343e-01
1.09259009e+00 -9.81210828e-01 7.53666759e-01 3.89849484e-01
7.17935622e-01 -1.33182049e+00 3.09275955e-01 4.16344315e-01
-7.43245244e-01 -7.01076031e-01 -2.18909532e-01 9.17832926e-03
-3.14401351e-02 2.22436503e-01 -1.05081666e+00 -1.57916412e-01
4.85424727e-01 3.83012414e-01 -8.28227550e-02 1.08462131e+00
-3.53882462e-01 7.89348423e-01 -3.33390057e-01 -2.32790783e-01
3.54971021e-01 -3.92851740e-01 6.43640935e-01 8.65216315e-01
1.15358032e-01 -1.45664901e-01 5.42856336e-01 9.85989988e-01
-2.17183143e-01 -6.65786415e-02 -4.51750159e-01 -3.26588571e-01
7.82769144e-01 6.83306813e-01 -3.71450663e-01 -3.50093663e-01
1.68686956e-01 4.47207391e-01 4.15470898e-01 5.28572917e-01
-8.97057474e-01 -5.56767523e-01 1.06943750e+00 -1.58286527e-01
5.02279043e-01 -3.10011864e-01 -2.09934235e-01 -6.74589038e-01
-1.68780252e-01 -1.09390891e+00 5.64380348e-01 -5.54966293e-02
-1.16229630e+00 -9.18519497e-02 2.36118272e-01 -1.07434440e+00
-6.76348209e-01 -3.89343530e-01 -4.69264984e-01 5.46871960e-01
-1.10927773e+00 -6.64661601e-02 3.77539635e-01 3.28225851e-01
3.82328540e-01 1.63340999e-03 3.20280403e-01 -2.14427829e-01
-4.37036067e-01 7.28529274e-01 6.18577719e-01 -2.46003509e-01
5.64896107e-01 -1.66864491e+00 4.65525627e-01 4.91080433e-01
-3.95587116e-01 5.02399445e-01 9.62518990e-01 -7.09716439e-01
-1.47216773e+00 -8.47292602e-01 -2.23944262e-01 -3.53283942e-01
8.89705360e-01 7.15913530e-03 -6.85518980e-01 7.13662565e-01
7.39811063e-02 -1.98496953e-02 3.67086343e-02 -2.52741873e-01
-1.54804975e-01 -4.42388117e-01 -1.59951568e+00 7.23338246e-01
1.02049494e+00 -4.22435075e-01 -4.15530622e-01 6.84772134e-02
9.07353520e-01 -5.11805773e-01 -1.21604013e+00 3.15907091e-01
6.86197996e-01 -8.32365155e-01 6.82731628e-01 -1.03660440e+00
5.94787262e-02 -8.53419974e-02 4.15284336e-02 -1.86343420e+00
1.16512798e-01 -1.30668592e+00 -4.33475554e-01 6.93001688e-01
2.83415258e-01 -8.69184494e-01 6.55534804e-01 5.01012683e-01
1.20130263e-01 -1.16294575e+00 -9.82010961e-01 -1.15046835e+00
5.94571531e-01 -2.52937943e-01 8.66519570e-01 1.90351099e-01
3.24442238e-01 -3.21599871e-01 -2.42337659e-01 -3.00075322e-01
7.73755133e-01 2.64753610e-01 9.75695193e-01 -5.11538804e-01
-7.90618718e-01 -5.94531298e-01 3.69831860e-01 -1.25384200e+00
-2.71585751e-02 -3.54541242e-01 1.88942507e-01 -1.09231889e+00
-8.99100974e-02 -4.32103485e-01 -4.92139548e-01 3.98649752e-01
-1.32061273e-01 -5.09235859e-01 4.23006326e-01 -4.71592136e-02
-5.60956776e-01 8.26175332e-01 1.60492778e+00 1.33048326e-01
-8.58199537e-01 2.48223260e-01 -5.18094838e-01 5.54388762e-01
1.05482006e+00 -4.00855422e-01 -6.26785755e-01 2.38960106e-02
2.32261457e-02 4.44037765e-01 -5.80562130e-02 -4.67621714e-01
-2.58667171e-01 -8.27343464e-01 1.59245491e-01 -3.66744399e-01
-2.44948477e-01 -3.33826005e-01 -3.90854508e-01 8.34124684e-01
-8.71120811e-01 2.86083490e-01 1.54045567e-01 7.50395000e-01
3.42449248e-01 9.02863666e-02 1.14383459e+00 -2.85529047e-02
-4.45650816e-01 3.31975669e-01 -4.58642066e-01 6.21289372e-01
9.76584494e-01 1.34108335e-01 -3.79889578e-01 -4.75146055e-01
-5.15482008e-01 6.62746072e-01 4.29338545e-01 2.57634014e-01
3.63487929e-01 -1.14694321e+00 -3.65014136e-01 2.77276933e-01
-1.04502425e-01 -4.31542814e-01 -4.95520681e-01 6.88723803e-01
-1.27425224e-01 1.24949850e-01 -2.88842142e-01 -3.11240733e-01
-5.95511317e-01 4.19149667e-01 8.18051934e-01 -3.07984889e-01
-6.96345747e-01 6.89597964e-01 -3.50903839e-01 -3.68886679e-01
6.40062451e-01 -8.34455967e-01 2.61358023e-01 -2.34947279e-01
4.59387720e-01 7.97385812e-01 -3.11583430e-01 2.51152724e-01
-1.06036969e-01 -9.54124518e-03 2.95598116e-02 -6.06366038e-01
1.14872885e+00 7.25280643e-02 4.42316324e-01 2.86014140e-01
7.99149096e-01 -1.27434060e-01 -2.27416849e+00 -2.77749840e-02
3.37230504e-01 -5.87404609e-01 -3.32621157e-01 -9.33961987e-01
-8.00118208e-01 2.75210112e-01 6.91611886e-01 4.62417066e-01
8.45273793e-01 -3.15439731e-01 8.26503098e-01 4.32501882e-01
5.91546297e-01 -1.73659897e+00 2.78650075e-01 6.36893094e-01
7.25644708e-01 -8.92872512e-01 -1.32262874e-02 4.37404990e-01
-7.61045337e-01 8.20175707e-01 7.59264529e-01 -6.31634295e-01
3.45927477e-01 3.95293087e-01 -2.61446014e-02 2.88558334e-01
-9.83021319e-01 -3.16869885e-01 -2.35602200e-01 6.22383118e-01
-8.23029205e-02 2.86939234e-01 -6.09249055e-01 3.10934931e-01
-3.31890196e-01 1.35198161e-01 3.82108688e-01 1.25901306e+00
-8.73707831e-01 -1.26179183e+00 -4.07234609e-01 6.12650514e-01
-3.98679972e-01 5.57760715e-01 -1.12209231e-01 9.84789193e-01
-1.62919149e-01 6.40663803e-01 7.61356801e-02 -2.14999512e-01
5.52311897e-01 9.45479050e-02 7.72970676e-01 -2.02295259e-01
-6.95976973e-01 -2.29151100e-02 -1.67587250e-02 -1.06478190e+00
2.05009818e-01 -5.35738349e-01 -1.61614442e+00 -2.96737731e-01
8.90859067e-02 4.70839351e-01 3.88491958e-01 9.37904656e-01
2.91880101e-01 2.38900587e-01 1.06091404e+00 -4.84922141e-01
-1.84947336e+00 -6.98915839e-01 -7.25113928e-01 4.64725047e-01
7.48071313e-01 -8.81674469e-01 -4.20189351e-01 -4.88392472e-01] | [4.051209449768066, 2.2500553131103516] |
dc8bfdc8-0461-4b1a-ac53-91e880cbe38c | multivariate-regression-with-calibration | null | null | http://papers.nips.cc/paper/5630-multivariate-regression-with-calibration | http://papers.nips.cc/paper/5630-multivariate-regression-with-calibration.pdf | Multivariate Regression with Calibration | We propose a new method named calibrated multivariate regression (CMR) for fitting high dimensional multivariate regression models. Compared to existing methods, CMR calibrates the regularization for each regression task with respect to its noise level so that it is simultaneously tuning insensitive and achieves an improved finite-sample performance. Computationally, we develop an efficient smoothed proximal gradient algorithm which has a worst-case iteration complexity $O(1/\epsilon)$, where $\epsilon$ is a pre-specified numerical accuracy. Theoretically, we prove that CMR achieves the optimal rate of convergence in parameter estimation. We illustrate the usefulness of CMR by thorough numerical simulations and show that CMR consistently outperforms other high dimensional multivariate regression methods. We also apply CMR on a brain activity prediction problem and find that CMR is as competitive as the handcrafted model created by human experts. | ['Tuo Zhao', 'Lie Wang', 'Han Liu'] | 2014-12-01 | null | null | null | neurips-2014-12 | ['activity-prediction', 'activity-prediction'] | ['computer-vision', 'time-series'] | [-8.11656192e-03 -2.14582868e-02 -1.31186366e-01 -2.03029007e-01
-1.39738154e+00 -2.03047946e-01 6.56857863e-02 -2.05946550e-01
-6.74667060e-01 8.64164650e-01 2.41585337e-02 -2.17423692e-01
-3.30371410e-01 -4.67123874e-02 -8.60576034e-01 -7.02276349e-01
-2.45664984e-01 5.01218796e-01 -1.45549074e-01 1.03077926e-01
2.33055606e-01 1.74766332e-01 -8.70413542e-01 -4.42932546e-01
1.08665848e+00 9.24496114e-01 2.32137203e-01 7.47071385e-01
5.48958480e-01 6.48856759e-01 -1.97129324e-01 -3.09937149e-01
3.68730068e-01 -3.40878040e-01 -5.93507767e-01 1.42439693e-01
2.85472661e-01 -7.23705068e-02 -1.96445242e-01 1.06337261e+00
6.94950938e-01 4.16999459e-01 9.01111126e-01 -8.89565527e-01
-3.67343158e-01 4.65923995e-01 -8.53595495e-01 4.78978246e-01
-3.09810750e-02 -5.50727099e-02 7.42878973e-01 -9.69154477e-01
3.75339091e-01 1.07545710e+00 9.57208097e-01 5.37235439e-01
-1.68124557e+00 -7.17178822e-01 4.60752279e-01 -3.50780874e-01
-1.38287449e+00 -4.74938601e-01 7.25649834e-01 -7.35554218e-01
7.73083150e-01 1.73181086e-03 1.76927820e-01 1.02309287e+00
8.20011944e-02 6.58844113e-01 1.12278688e+00 -3.78057688e-01
4.28666472e-01 7.84643292e-02 2.32875481e-01 8.88507605e-01
1.47500992e-01 8.46854448e-02 -3.55395049e-01 -4.48794544e-01
1.06425655e+00 -3.90080452e-01 -3.00784737e-01 -4.41281319e-01
-1.19004893e+00 9.01821733e-01 2.10799724e-01 4.43555675e-02
-4.73251700e-01 5.42507172e-01 2.54664391e-01 2.63013512e-01
6.85336053e-01 4.41375554e-01 -5.83127439e-01 -1.77835837e-01
-8.79401505e-01 1.47128433e-01 5.84528863e-01 9.21876252e-01
3.90688270e-01 3.96051198e-01 1.37401313e-01 1.17325997e+00
1.77233368e-01 5.01405716e-01 4.16614473e-01 -1.39933956e+00
6.33502483e-01 1.57424092e-01 3.85347128e-01 -9.74441528e-01
-7.44220972e-01 -7.73448646e-01 -1.19629085e+00 -1.67715147e-01
4.22052264e-01 -5.61562479e-01 -4.09612328e-01 2.08006954e+00
2.18792081e-01 3.24208289e-01 -1.95503846e-01 9.32754874e-01
1.98041037e-01 3.41215968e-01 2.50111669e-01 -6.64814591e-01
9.25051510e-01 -7.34239817e-01 -5.69013357e-01 -5.55890977e-01
5.38460135e-01 -3.03458124e-01 1.16812265e+00 6.06998444e-01
-1.16633892e+00 -1.30243391e-01 -8.99258137e-01 2.33048722e-01
2.89461523e-01 3.26041609e-01 7.95996249e-01 7.54540324e-01
-7.17260778e-01 6.15381956e-01 -1.05936897e+00 1.06470801e-01
4.00011748e-01 5.36774218e-01 -1.28450662e-01 2.90631264e-01
-8.67385328e-01 7.33536959e-01 -9.97355729e-02 2.98749655e-01
-1.08111870e+00 -1.06986773e+00 -7.07075119e-01 -3.51552010e-01
3.92564118e-01 -7.87721157e-01 1.26901257e+00 -8.28799725e-01
-1.59706891e+00 7.02445865e-01 -5.20822406e-01 -5.66800177e-01
7.92099535e-01 -2.82389283e-01 -1.51907280e-01 6.04727343e-02
1.64034273e-02 2.67018586e-01 1.14562440e+00 -9.73175049e-01
-2.92606354e-01 -4.58780795e-01 -2.60363668e-01 9.47868675e-02
-3.92841071e-01 1.20061807e-01 -6.40748680e-01 -8.39100122e-01
2.49478012e-01 -1.07409823e+00 -8.97063494e-01 -1.41031057e-01
-1.98121771e-01 2.14688554e-01 1.87912777e-01 -7.45324910e-01
1.40848124e+00 -1.94199145e+00 3.94394368e-01 5.16375363e-01
4.40712422e-01 -1.50668800e-01 1.22078761e-01 -3.02024662e-01
-1.72134221e-01 1.03922643e-01 -6.20338559e-01 -3.74937475e-01
-1.21391833e-01 -6.04840331e-02 -1.05358146e-01 9.13450778e-01
-1.13521598e-01 6.50436878e-01 -8.04961264e-01 -3.14956069e-01
-1.31151840e-01 5.90977073e-01 -7.41158724e-01 -5.31585515e-02
7.89195001e-02 6.11231565e-01 -8.48847866e-01 5.28491139e-01
4.57538068e-01 -6.40546024e-01 3.75466019e-01 -2.23308817e-01
3.16663206e-01 -3.49966437e-01 -1.31713414e+00 1.44259953e+00
-6.63960040e-01 5.26345849e-01 3.99933845e-01 -1.24096537e+00
9.19497788e-01 2.22198933e-01 7.34112799e-01 -3.98640573e-01
2.12448165e-01 2.83335090e-01 -5.98644912e-01 -2.62675107e-01
7.51927570e-02 -3.10319602e-01 -7.72031918e-02 2.16536716e-01
-1.55692011e-01 2.87735201e-02 -1.29514277e-01 2.99429167e-02
1.06280982e+00 -2.70317532e-02 4.70845163e-01 -5.20109296e-01
5.86031079e-01 -2.25201547e-01 8.35016847e-01 9.81055379e-01
-3.66588593e-01 3.92956555e-01 8.32652807e-01 -1.12248205e-01
-1.02342820e+00 -1.04321492e+00 -2.36040294e-01 1.24091971e+00
-8.27836543e-02 -1.72921836e-01 -8.98682892e-01 -4.98596996e-01
5.15978038e-02 3.65093052e-01 -5.80411673e-01 4.50627506e-03
-6.93989217e-01 -1.23031616e+00 3.69900107e-01 8.06293786e-01
4.08829361e-01 -6.35774672e-01 -3.32224935e-01 3.39321226e-01
-1.39209805e-02 -1.13647115e+00 -7.79968798e-01 2.97796369e-01
-1.29721594e+00 -9.00177360e-01 -8.03487539e-01 -7.47136295e-01
9.35959041e-01 1.67003516e-02 1.06709778e+00 -2.44282529e-01
-1.87069789e-01 3.80429476e-01 1.76588029e-01 -4.60241027e-02
-7.58735761e-02 2.12766584e-02 4.61271226e-01 -2.02195063e-01
1.32553056e-01 -7.47680366e-01 -6.56982720e-01 4.76004034e-01
-4.80461299e-01 -2.94132471e-01 4.09179360e-01 1.07983375e+00
9.80310857e-01 -1.90000609e-01 7.79916584e-01 -1.14056337e+00
9.94318902e-01 -4.34481770e-01 -1.08554387e+00 3.21318656e-02
-8.55519176e-01 3.55763972e-01 7.71111548e-01 -7.26848125e-01
-1.01308107e+00 4.92807418e-01 2.76468217e-01 -6.51919663e-01
3.57366949e-01 6.08048201e-01 1.85504019e-01 -3.60622942e-01
1.09755874e+00 -4.52837273e-02 -1.26870945e-01 -6.34893000e-01
4.24416989e-01 2.61206806e-01 6.64937735e-01 -9.31978106e-01
6.44172788e-01 2.81856537e-01 1.44727409e-01 -7.92651892e-01
-8.94389808e-01 -2.40711436e-01 -6.31412804e-01 -1.09154083e-01
4.62896854e-01 -1.03116989e+00 -1.02512515e+00 6.38086870e-02
-6.47065580e-01 -5.55096865e-01 2.32870668e-01 8.81451607e-01
-1.04394853e+00 3.51744175e-01 -6.91379011e-01 -9.62115228e-01
-3.49564552e-01 -1.05089986e+00 6.88993156e-01 -2.33712092e-01
-1.96151614e-01 -1.06816542e+00 2.15045922e-02 2.47851968e-01
1.67628974e-01 3.77991527e-01 8.64681482e-01 -3.40442479e-01
-2.09693253e-01 -2.17449680e-01 -8.42916742e-02 3.16268325e-01
-3.45147848e-01 -2.66871184e-01 -6.64885998e-01 -3.78131270e-01
2.05426380e-01 -2.05605030e-01 9.00088489e-01 8.69649470e-01
1.44785666e+00 -4.40823108e-01 -2.86131650e-01 9.53028321e-01
1.39144623e+00 -4.76732813e-02 3.37925643e-01 2.78747499e-01
7.07734704e-01 2.98014909e-01 6.00784361e-01 6.31782413e-01
1.44691035e-01 3.24071467e-01 7.88426697e-02 -6.81708306e-02
4.42900121e-01 -1.19348727e-01 3.37504119e-01 8.01659644e-01
-2.85589069e-01 3.56979728e-01 -7.91814983e-01 3.41683149e-01
-2.12307525e+00 -7.58845091e-01 -4.00682492e-03 2.41521430e+00
8.61160994e-01 2.00818226e-01 2.84915209e-01 -7.29993582e-02
6.76658690e-01 -8.30304772e-02 -9.43446994e-01 -3.20847154e-01
-9.96355861e-02 2.79252410e-01 9.27597225e-01 6.30375147e-01
-1.15462101e+00 9.14897501e-01 8.04881763e+00 8.38992834e-01
-6.38554990e-01 2.77818233e-01 7.55178928e-01 -2.78103560e-01
4.57838997e-02 -2.60578334e-01 -6.59312069e-01 2.72045910e-01
9.18867290e-01 -2.41135538e-01 7.80586541e-01 1.10010755e+00
5.82633734e-01 -2.63479352e-02 -9.73399460e-01 1.25134766e+00
-1.75434515e-01 -1.16029382e+00 -4.47956860e-01 -1.49494618e-01
8.54519010e-01 8.19186401e-03 2.83989906e-01 2.80512899e-01
4.00867909e-01 -1.09340680e+00 3.36551070e-01 5.83446741e-01
5.82672417e-01 -1.00380456e+00 2.83242345e-01 5.98793924e-01
-1.06887484e+00 -3.85784507e-01 -4.96509641e-01 2.81440020e-01
1.00874081e-01 7.00270474e-01 -5.98988175e-01 -3.24069262e-02
6.11684322e-01 6.93148792e-01 -3.38987470e-01 9.26559567e-01
-1.25690058e-01 6.04546309e-01 -4.79558736e-01 1.39436141e-01
1.20283239e-01 -6.07864380e-01 6.02358580e-01 1.38163650e+00
1.28603131e-01 3.10788423e-01 4.23709005e-02 7.92763770e-01
-1.56809345e-01 1.97392046e-01 -2.72093147e-01 9.85031128e-02
4.69968826e-01 9.94033098e-01 -4.47914660e-01 -1.80600554e-01
-2.55244881e-01 7.60112286e-01 6.02176726e-01 5.84531724e-01
-7.74410009e-01 -1.75256073e-01 6.25663817e-01 -1.82539061e-01
4.31309968e-01 -4.30703819e-01 -4.87725884e-01 -1.24805570e+00
1.72118008e-01 -8.89468014e-01 4.61646765e-01 -5.99096417e-01
-1.48249745e+00 4.39110965e-01 -1.26519039e-01 -1.07058442e+00
-3.90892804e-01 -6.07185721e-01 -7.09817838e-03 8.23548794e-01
-1.25004017e+00 -5.09378970e-01 -1.40602728e-02 6.54719591e-01
3.04819435e-01 -1.95833728e-01 7.80590892e-01 2.00493887e-01
-9.25968289e-01 7.98645020e-01 3.62921238e-01 -5.68646267e-02
4.57586586e-01 -1.09525263e+00 1.02383941e-01 7.13609874e-01
-4.16174233e-01 6.37262881e-01 9.29809690e-01 -5.77609360e-01
-1.37275481e+00 -1.04841435e+00 4.34959412e-01 -3.33505690e-01
8.32734406e-01 -2.48285174e-01 -8.70454133e-01 8.74701738e-01
-4.85108405e-01 2.07519129e-01 5.60813189e-01 1.93081707e-01
-2.98873186e-01 -2.15557098e-01 -1.23054338e+00 7.96272397e-01
1.05450749e+00 -6.33543372e-01 -3.90094072e-01 5.67362428e-01
4.74786729e-01 -4.58471358e-01 -1.28206193e+00 4.61527437e-01
4.98469263e-01 -6.61225557e-01 1.17209578e+00 -8.11558247e-01
-6.40173501e-04 -6.14432851e-03 -2.33815983e-01 -1.17928588e+00
-3.51074815e-01 -1.07543969e+00 -5.29742360e-01 5.77602148e-01
7.48982847e-01 -5.61622500e-01 9.58439529e-01 9.20434237e-01
2.81955808e-01 -8.85765970e-01 -9.15446818e-01 -9.06669915e-01
3.38554740e-01 -8.39780390e-01 -1.36453304e-02 9.27072763e-01
1.14044594e-02 2.80977756e-01 -6.88359797e-01 2.43189022e-01
1.08279335e+00 -2.58116096e-01 5.28945744e-01 -1.16280818e+00
-4.88703996e-01 -3.97136182e-01 -1.61519036e-01 -1.19291580e+00
4.67108518e-01 -6.74953341e-01 -1.24612182e-01 -9.41249549e-01
2.41885170e-01 -5.18709958e-01 -4.58435059e-01 1.80885479e-01
-2.64294118e-01 4.35251147e-02 -2.13725865e-01 1.11940034e-01
-5.07088780e-01 5.16164720e-01 1.14865088e+00 1.39259338e-01
-5.83131611e-01 1.99192926e-01 -8.70727897e-01 9.94285882e-01
7.51045048e-01 -6.04617119e-01 -2.91199237e-01 -2.90274680e-01
2.22894236e-01 5.23520947e-01 1.52111620e-01 -6.93474650e-01
2.16481894e-01 -1.29890203e-01 5.68673253e-01 -2.86768109e-01
4.38982517e-01 -5.33648431e-01 -1.09085843e-01 2.14030772e-01
-6.24659359e-01 5.90496585e-02 -3.45187150e-02 8.53094101e-01
2.49836102e-01 -7.36150742e-02 1.07735348e+00 -3.59606668e-02
-3.50778729e-01 3.93312424e-01 -4.04125541e-01 3.69313896e-01
7.40610003e-01 -7.82026276e-02 6.99459165e-02 -3.96206230e-01
-1.01321065e+00 4.33971137e-01 5.92765994e-02 -1.12020597e-01
6.03983164e-01 -1.11700344e+00 -5.37316799e-01 -3.63067351e-02
-2.20848709e-01 -3.68284851e-01 1.45672917e-01 1.27143860e+00
-1.20807476e-01 2.71229923e-01 3.09151918e-01 -5.22427619e-01
-1.01540124e+00 6.23570740e-01 6.40983939e-01 -2.27620870e-01
-6.78860545e-01 7.77824700e-01 5.08898124e-02 -3.69311750e-01
4.86368686e-01 -2.02398032e-01 7.83431157e-02 -2.31340945e-01
3.91165167e-01 6.92107141e-01 -6.54715225e-02 -3.07697445e-01
-3.92103851e-01 7.93072701e-01 9.19260159e-02 -2.81108707e-01
1.48825431e+00 -1.45318866e-01 1.03645861e-01 3.94569755e-01
1.30826545e+00 -9.73589644e-02 -1.77037680e+00 -4.12120044e-01
1.69859767e-01 -3.04751784e-01 3.45982760e-01 -5.63277006e-01
-1.16156960e+00 4.32375789e-01 5.49085438e-01 -4.27793741e-01
1.19852459e+00 -1.80378586e-01 3.31634283e-01 7.77342677e-01
2.37580538e-01 -1.39915490e+00 2.70959940e-02 4.13832098e-01
7.80902982e-01 -1.25865304e+00 2.88704515e-01 -4.04937655e-01
-9.90274847e-01 9.16604996e-01 4.36960816e-01 -6.44233227e-01
9.46882546e-01 2.27264270e-01 -2.93154180e-01 6.88439012e-02
-7.92049468e-01 1.29882902e-01 3.64733607e-01 6.69285953e-01
3.65715832e-01 8.07799399e-02 -4.55614775e-01 9.79434729e-01
-1.52192384e-01 1.49676174e-01 2.31777459e-01 7.35833585e-01
-4.80723202e-01 -6.78478539e-01 -3.15622181e-01 6.19094074e-01
-4.78804201e-01 -2.56553963e-02 -6.99166805e-02 6.98569477e-01
-6.88585937e-01 8.36064756e-01 -2.56698340e-01 -1.00338824e-01
4.34026837e-01 -5.16584851e-02 6.50341630e-01 -9.43129584e-02
-5.12550235e-01 4.71724570e-01 7.03226551e-02 -8.16893995e-01
-2.33025581e-01 -6.15037918e-01 -1.21367025e+00 -2.19673023e-01
-1.05258010e-01 1.09370381e-01 5.73364854e-01 9.81938422e-01
3.07167649e-01 3.22148204e-01 8.60754907e-01 -6.67176127e-01
-8.88683021e-01 -7.30566025e-01 -6.69663608e-01 5.25684357e-02
5.18623471e-01 -7.97839522e-01 -6.17683947e-01 2.37888955e-02] | [7.054462432861328, 4.407219886779785] |
580a4927-929c-41de-b1b7-94714c7d8241 | deep-unsupervised-image-hashing-by-maximizing | 2012.12334 | null | https://arxiv.org/abs/2012.12334v1 | https://arxiv.org/pdf/2012.12334v1.pdf | Deep Unsupervised Image Hashing by Maximizing Bit Entropy | Unsupervised hashing is important for indexing huge image or video collections without having expensive annotations available. Hashing aims to learn short binary codes for compact storage and efficient semantic retrieval. We propose an unsupervised deep hashing layer called Bi-half Net that maximizes entropy of the binary codes. Entropy is maximal when both possible values of the bit are uniformly (half-half) distributed. To maximize bit entropy, we do not add a term to the loss function as this is difficult to optimize and tune. Instead, we design a new parameter-free network layer to explicitly force continuous image features to approximate the optimal half-half bit distribution. This layer is shown to minimize a penalized term of the Wasserstein distance between the learned continuous image features and the optimal half-half bit distribution. Experimental results on the image datasets Flickr25k, Nus-wide, Cifar-10, Mscoco, Mnist and the video datasets Ucf-101 and Hmdb-51 show that our approach leads to compact codes and compares favorably to the current state-of-the-art. | ['Jan van Gemert', 'Yunqiang Li'] | 2020-12-22 | null | null | null | null | ['semantic-retrieval'] | ['natural-language-processing'] | [-6.72674999e-02 1.11247012e-02 -5.15386641e-01 -6.30066395e-01
-1.16862607e+00 -3.57217520e-01 4.54839855e-01 3.72202754e-01
-9.85598743e-01 7.94525862e-01 2.57678300e-01 5.98054416e-02
-4.14053053e-02 -6.27230048e-01 -1.03236079e+00 -8.22381437e-01
-4.46963131e-01 5.90476930e-01 1.34550676e-01 3.42669696e-01
2.94792980e-01 2.23269731e-01 -1.71371520e+00 3.33016813e-01
3.24247390e-01 1.31805122e+00 3.20771217e-01 8.17677975e-01
2.58714497e-01 7.86600351e-01 -2.33154714e-01 -2.75596112e-01
4.12370622e-01 -1.30660519e-01 -9.55471039e-01 -2.82304883e-01
5.87592959e-01 -8.22317421e-01 -7.34496236e-01 1.16233957e+00
5.18213332e-01 2.21756041e-01 8.51934314e-01 -1.17619526e+00
-8.34889591e-01 6.99666202e-01 -2.89339125e-01 7.03299195e-02
1.47548810e-01 -2.06108779e-01 1.31980658e+00 -7.36501932e-01
7.87904322e-01 1.17586279e+00 6.83495760e-01 2.89640456e-01
-1.16984618e+00 -5.61365068e-01 -5.43238461e-01 5.10596216e-01
-1.98981619e+00 -3.60276848e-01 2.44849592e-01 -2.45969743e-01
9.33744848e-01 3.67978737e-02 4.24417347e-01 6.06584847e-01
-2.79119164e-02 7.91687131e-01 7.36246347e-01 -3.30435723e-01
4.46960390e-01 1.12495176e-01 4.55782469e-03 8.33402395e-01
2.02934265e-01 2.65408959e-02 -5.60894310e-01 -2.89496481e-01
4.99098629e-01 2.73335606e-01 -2.93848157e-01 -6.07739151e-01
-1.13296235e+00 1.13264596e+00 7.33474314e-01 1.89576924e-01
-1.76201239e-01 6.55745625e-01 3.94744098e-01 1.30154639e-01
-2.77806763e-02 2.26804838e-01 -1.27469942e-01 -3.61613005e-01
-1.20843875e+00 4.46916431e-01 7.69636273e-01 9.92587686e-01
1.05212641e+00 -5.00930727e-01 -2.71679848e-01 7.84617960e-01
4.45801318e-01 5.19520044e-01 5.08417845e-01 -1.13889742e+00
3.40338379e-01 1.89026475e-01 1.84353799e-01 -9.70605373e-01
-8.86594951e-02 4.24626023e-02 -6.92232490e-01 -3.44370872e-01
1.41057715e-01 4.10950631e-01 -1.06466043e+00 1.66011417e+00
1.01615652e-01 -4.86347936e-02 -7.79998600e-02 1.07215285e+00
6.03150666e-01 9.48839605e-01 -2.21698374e-01 1.32993355e-01
1.22356164e+00 -8.51342261e-01 -5.26478469e-01 1.75470650e-01
7.29982734e-01 -4.97246504e-01 9.76866007e-01 1.81830406e-01
-1.11250198e+00 -3.26522887e-01 -1.20878172e+00 -4.74514157e-01
-4.85952169e-01 -2.33474359e-01 4.24201638e-01 3.91396254e-01
-1.15299392e+00 6.71354771e-01 -8.05825830e-01 -1.19358987e-01
6.04076505e-01 5.74115455e-01 -3.87807131e-01 -3.40738654e-01
-1.28638482e+00 5.86509764e-01 6.82140470e-01 -3.62361401e-01
-1.14707077e+00 -5.37582755e-01 -9.28556681e-01 3.34183812e-01
-2.17298344e-01 -2.08560213e-01 1.14112926e+00 -3.92541677e-01
-1.01656556e+00 8.57483447e-01 2.27073655e-02 -1.13518846e+00
1.63780078e-01 1.15584889e-02 1.12140335e-01 6.47178948e-01
-9.98759083e-03 1.46474528e+00 7.40613639e-01 -9.18801725e-01
-2.76102513e-01 -1.54722854e-01 6.80198297e-02 2.09452733e-01
-8.57662499e-01 -3.19007933e-01 -5.30149281e-01 -5.47439992e-01
-1.35255277e-01 -1.09717155e+00 6.13877252e-02 4.14098114e-01
-3.77739519e-01 -8.01335052e-02 7.91333854e-01 -4.92385030e-01
1.16030216e+00 -2.41100168e+00 3.46351787e-02 2.06289500e-01
-6.58782795e-02 1.25202641e-01 -5.70745021e-02 4.15266871e-01
3.00115228e-01 -4.81560044e-02 -3.82924825e-01 -5.05718112e-01
3.63412499e-01 4.51306611e-01 -2.79197365e-01 7.73519754e-01
-2.76018232e-01 7.24148333e-01 -8.85167718e-01 -6.80994809e-01
2.08069794e-02 8.82188320e-01 -9.77788091e-01 1.75568417e-01
-2.17013285e-01 -3.26556176e-01 -7.68670514e-02 5.22404253e-01
6.92576587e-01 -5.39183080e-01 -9.71923992e-02 -1.58698469e-01
9.16859284e-02 4.66048777e-01 -8.34538341e-01 1.95167911e+00
-9.03945789e-02 7.46842563e-01 -2.14313611e-01 -1.02060497e+00
6.65445387e-01 1.02124318e-01 3.85175854e-01 -7.60136187e-01
6.46267235e-02 2.28413865e-01 -4.25216317e-01 -2.45681792e-01
8.42558563e-01 2.19544154e-02 -1.81877896e-01 4.03396100e-01
1.73885539e-01 -7.55214039e-03 2.41525769e-01 3.67032021e-01
1.02248859e+00 -5.76670706e-01 -7.54557922e-02 -2.93547630e-01
2.16370389e-01 -4.26559210e-01 2.14612871e-01 8.95966709e-01
-1.33496866e-01 8.84103835e-01 3.51084322e-01 -4.61880267e-01
-1.38119495e+00 -9.93610501e-01 -4.59869653e-01 9.80034411e-01
2.99544156e-01 -6.48028195e-01 -8.86679888e-01 -5.09239674e-01
3.38585973e-01 3.27847302e-01 -6.87631369e-01 -2.36694902e-01
-1.89314380e-01 -3.73873770e-01 6.93238497e-01 3.97450417e-01
6.12576306e-01 -9.61383939e-01 -7.57142782e-01 -9.28482637e-02
-4.17194426e-01 -1.13879609e+00 -8.63191128e-01 4.71423477e-01
-5.90724528e-01 -6.91050768e-01 -9.34090316e-01 -8.84120703e-01
6.03474200e-01 1.55680478e-01 9.11597073e-01 1.09173104e-01
-6.96544707e-01 3.11880141e-01 -3.02253664e-01 6.36628196e-02
-4.19385843e-02 3.10492903e-01 -1.11794487e-01 -2.82782584e-01
4.11970347e-01 -2.54486531e-01 -1.05115044e+00 1.86390102e-01
-1.19244206e+00 -4.20715660e-01 4.01340455e-01 1.01254225e+00
9.20599222e-01 -9.84262768e-03 6.49066493e-02 -2.62656659e-01
1.11870356e-01 -4.35360789e-01 -6.70863688e-01 4.64530773e-02
-5.96084654e-01 5.53251803e-01 4.76365447e-01 -3.68286580e-01
-1.46208495e-01 1.48218155e-01 1.48487166e-01 -7.76968360e-01
2.31185734e-01 3.42275023e-01 2.26467356e-01 -1.64447367e-01
3.74086022e-01 4.44612145e-01 -1.16576208e-02 -3.71026158e-01
4.51218575e-01 9.41190064e-01 6.73893392e-01 -4.49323297e-01
4.93782014e-01 6.86076343e-01 -1.39238343e-01 -7.86971509e-01
-6.39127254e-01 -6.58756673e-01 -3.66298795e-01 1.98355958e-01
9.15096521e-01 -1.00041080e+00 -7.63553619e-01 3.08857322e-01
-9.10031796e-01 -3.50133389e-01 -4.57558453e-01 4.67714727e-01
-9.33742404e-01 4.76156443e-01 -8.17472816e-01 -6.68985069e-01
-5.09077847e-01 -1.24975109e+00 1.43874991e+00 1.03484899e-01
2.63562030e-03 -5.66804707e-01 1.75130352e-01 6.11897856e-02
4.52317059e-01 -1.43919498e-01 8.72798979e-01 -4.82410371e-01
-8.19364190e-01 -2.85993248e-01 -5.94488800e-01 4.26433921e-01
-3.14252347e-01 -3.09428453e-01 -7.52349496e-01 -6.67598367e-01
-3.52474481e-01 -1.15117943e+00 1.16832912e+00 3.72052431e-01
1.71032584e+00 -7.60867655e-01 -3.12728658e-02 9.96944487e-01
1.61609137e+00 -9.36674029e-02 8.93225014e-01 4.00168747e-01
3.62149268e-01 5.37143685e-02 5.01871347e-01 7.99313009e-01
6.84306443e-01 7.29127526e-01 5.50385416e-01 4.45402294e-01
-1.55836884e-02 -4.43927705e-01 2.48605177e-01 7.14147627e-01
4.95265633e-01 -2.78709024e-01 -7.22063959e-01 8.05199504e-01
-1.48900616e+00 -1.02241886e+00 6.19715273e-01 2.47189593e+00
1.17857790e+00 8.99976119e-02 2.18158867e-02 1.51845902e-01
5.74900806e-01 3.36603105e-01 -4.52977806e-01 -3.49549323e-01
1.27130728e-02 1.08834438e-01 8.39245379e-01 5.53476453e-01
-1.35795426e+00 6.55738831e-01 6.41053200e+00 9.73497331e-01
-1.00218058e+00 5.57215661e-02 6.27274990e-01 -3.91736269e-01
-3.41906518e-01 -1.26137286e-01 -9.44529355e-01 8.21063995e-01
1.19062495e+00 1.39560446e-01 7.29114950e-01 9.17303562e-01
-4.47476566e-01 4.77270372e-02 -1.23953187e+00 1.45380461e+00
8.80281776e-02 -1.44900572e+00 1.82025090e-01 3.69708657e-01
7.29052782e-01 4.57131833e-01 3.60464543e-01 1.77085564e-01
9.74253565e-02 -1.09162295e+00 8.73308718e-01 3.58994871e-01
1.13186014e+00 -8.79189372e-01 5.90204477e-01 4.39167470e-02
-1.05749869e+00 -2.17996225e-01 -9.21611309e-01 4.36755389e-01
-1.04586675e-03 4.38784242e-01 -8.00959289e-01 -3.25622559e-01
1.06990671e+00 6.34229898e-01 -3.25748861e-01 1.19041991e+00
2.66317099e-01 2.32567593e-01 -6.13981187e-01 -1.34073421e-01
6.50049865e-01 2.47452319e-01 2.38773718e-01 1.39445102e+00
3.96847606e-01 3.44419517e-02 8.79185796e-02 5.91273963e-01
-6.70389473e-01 -1.73227847e-01 -6.32343173e-01 -3.81957293e-01
8.03635776e-01 8.13227415e-01 -5.78931749e-01 -4.07464474e-01
-9.85082984e-02 1.07039356e+00 4.62391734e-01 1.45398632e-01
-7.88234591e-01 -7.36389101e-01 5.60723782e-01 9.50197801e-02
1.00647330e+00 -1.60475031e-01 2.04248995e-01 -9.39619005e-01
-7.68340528e-02 -6.80042386e-01 4.74305540e-01 -6.99457228e-01
-9.88701880e-01 4.59796637e-01 5.06885983e-02 -9.95713413e-01
-3.35475773e-01 -4.04208958e-01 5.24489954e-02 4.20067191e-01
-1.71665788e+00 -5.41318178e-01 -1.65774062e-01 6.70358598e-01
1.10369148e-02 2.39440072e-02 8.69915783e-01 5.91568351e-01
-7.68494457e-02 1.15052116e+00 6.47891343e-01 2.33175308e-01
7.20504999e-01 -1.24931955e+00 1.30192518e-01 1.18679538e-01
1.96861580e-01 6.17492139e-01 5.96146405e-01 -1.98101237e-01
-1.57331002e+00 -1.00659549e+00 9.23359990e-01 3.92350480e-02
5.47083676e-01 -5.07856071e-01 -8.23276460e-01 3.82009238e-01
6.69700578e-02 5.29961050e-01 5.95852613e-01 -4.53252614e-01
-8.58379841e-01 -2.86605388e-01 -1.42577219e+00 2.09039614e-01
6.55177772e-01 -9.38911974e-01 -2.88385719e-01 5.19181430e-01
8.38792741e-01 -1.68415964e-01 -1.08920145e+00 2.77952224e-01
9.15512383e-01 -9.24317658e-01 1.21357870e+00 -2.93322057e-01
5.29246330e-01 -1.12458125e-01 -8.32370102e-01 -8.05543125e-01
-1.94616020e-01 -3.94957930e-01 -4.66988802e-01 7.27786720e-01
2.98792809e-01 -2.39669412e-01 9.29114401e-01 4.58737493e-01
2.05967426e-01 -9.98837948e-01 -1.01627171e+00 -7.57170439e-01
-6.92077056e-02 2.93004643e-02 6.40325785e-01 6.30562842e-01
1.09100252e-01 -3.32702518e-01 -3.82150531e-01 -4.88095358e-02
1.01917899e+00 2.88733393e-02 5.12931585e-01 -7.93125868e-01
-4.09887373e-01 -2.76708484e-01 -8.71834099e-01 -1.32570565e+00
2.08691992e-02 -8.36882830e-01 3.42888683e-01 -1.18923569e+00
6.49859548e-01 -5.85423827e-01 -5.61290860e-01 5.22674978e-01
2.33803257e-01 5.97218692e-01 5.80579713e-02 3.13438535e-01
-1.11155331e+00 8.19078028e-01 8.17477822e-01 -3.55122745e-01
2.50142068e-01 -7.10823059e-01 -3.65785241e-01 3.10299695e-01
4.13075596e-01 -7.82377481e-01 -3.84131670e-01 -5.46200514e-01
2.07381114e-01 -8.04939941e-02 2.84790903e-01 -1.04325187e+00
5.02090454e-01 3.15745682e-01 2.73425996e-01 -8.63901377e-01
5.23948431e-01 -8.31922650e-01 -1.37752995e-01 5.54298937e-01
-8.75643253e-01 -5.49139716e-02 -1.89134970e-01 6.37910068e-01
-2.60395408e-01 -4.36586529e-01 1.03767943e+00 6.65715896e-03
-3.42966467e-01 4.89095122e-01 -8.29409901e-03 1.73835725e-01
8.11567247e-01 -5.98951280e-02 -2.27751836e-01 -6.24918044e-01
-2.68918216e-01 2.86600798e-01 7.27204561e-01 2.85089821e-01
7.86758542e-01 -1.61513770e+00 -4.15258974e-01 3.21536750e-01
2.57173628e-01 5.07573783e-02 1.03292115e-01 3.53303760e-01
-8.62471998e-01 7.45592058e-01 -3.95566076e-02 -7.42117465e-01
-9.64162767e-01 5.90228200e-01 1.42687693e-01 -2.13041574e-01
-4.86250639e-01 1.09322596e+00 -7.53520057e-02 -2.60439247e-01
9.06102359e-01 -3.11493695e-01 2.74793625e-01 -7.92982057e-02
8.57384801e-01 2.09006503e-01 2.54683103e-02 -7.26891577e-01
-3.29973817e-01 3.43799472e-01 -4.00488168e-01 -2.05344096e-01
1.40402758e+00 -1.32297277e-01 -2.11667314e-01 2.35679805e-01
2.17746854e+00 -7.36010075e-01 -1.30634892e+00 -2.65847951e-01
-1.19497977e-01 -6.27786219e-01 1.69766039e-01 -3.16691428e-01
-1.12961495e+00 8.80782247e-01 8.70542169e-01 1.28335968e-01
8.35115969e-01 2.08495148e-02 1.24631834e+00 8.04587662e-01
5.25312364e-01 -1.09934878e+00 2.27786794e-01 5.16241550e-01
5.55726707e-01 -1.06427622e+00 -6.81487396e-02 4.18809354e-01
-2.75858283e-01 8.09322238e-01 1.78083554e-01 -3.58458906e-01
8.50267768e-01 2.58064300e-01 -4.62666959e-01 -2.05592126e-01
-5.86128891e-01 1.19214736e-01 3.52870792e-01 2.05861732e-01
1.93815157e-01 -6.49744347e-02 -1.35507807e-01 -8.98919106e-02
-3.92573535e-01 3.30301118e-03 6.00622594e-02 1.13812065e+00
-6.94610059e-01 -7.58337379e-01 -2.63358921e-01 5.15679538e-01
-7.51497507e-01 -3.28752518e-01 2.62952205e-02 4.33239192e-01
-1.96329400e-01 3.94173980e-01 3.01437736e-01 -2.81421155e-01
-3.47271591e-01 5.27361929e-02 5.07433236e-01 -2.11494461e-01
-6.74146693e-03 -2.13097289e-01 -3.91424805e-01 -6.96092308e-01
-2.40190372e-01 -4.41904008e-01 -1.38364744e+00 -3.80860746e-01
-3.00185263e-01 4.50196505e-01 9.39512253e-01 5.46544194e-01
3.94043744e-01 -2.00959265e-01 6.74862444e-01 -9.92103338e-01
-8.73078108e-01 -6.85369849e-01 -6.92616343e-01 5.52211106e-01
6.88779414e-01 -5.18825054e-01 -5.49588203e-01 1.02158971e-01] | [11.245474815368652, 0.9640119075775146] |
36fd39d6-86eb-418d-938a-8962a657e52b | speechmatrix-a-large-scale-mined-corpus-of-1 | 2211.04508 | null | https://arxiv.org/abs/2211.04508v1 | https://arxiv.org/pdf/2211.04508v1.pdf | SpeechMatrix: A Large-Scale Mined Corpus of Multilingual Speech-to-Speech Translations | We present SpeechMatrix, a large-scale multilingual corpus of speech-to-speech translations mined from real speech of European Parliament recordings. It contains speech alignments in 136 language pairs with a total of 418 thousand hours of speech. To evaluate the quality of this parallel speech, we train bilingual speech-to-speech translation models on mined data only and establish extensive baseline results on EuroParl-ST, VoxPopuli and FLEURS test sets. Enabled by the multilinguality of SpeechMatrix, we also explore multilingual speech-to-speech translation, a topic which was addressed by few other works. We also demonstrate that model pre-training and sparse scaling using Mixture-of-Experts bring large gains to translation performance. The mined data and models are freely available. | ['Holger Schwenk', 'Benoît Sagot', 'Juan Pino', 'Changhan Wang', 'Vedanuj Goswani', 'Ann Lee', 'Jingfei Du', 'Ning Dong', 'Hongyu Gong', 'Paul-Ambroise Duquenne'] | 2022-11-08 | null | null | null | arxiv-2022-10 | ['speech-to-speech-translation'] | ['speech'] | [-4.68200222e-02 3.57836694e-01 -3.19270819e-01 -4.81695235e-01
-1.77245605e+00 -7.86270142e-01 9.98509169e-01 -3.11200202e-01
-4.82862830e-01 9.33586538e-01 7.43527412e-01 -8.63727808e-01
4.52587992e-01 -1.83597818e-01 -8.49793077e-01 -3.71822178e-01
1.25601426e-01 1.04491711e+00 -1.58999547e-01 -5.85847080e-01
-5.38696289e-01 -5.87263182e-02 -6.82123423e-01 7.43838191e-01
8.05503726e-01 1.97304979e-01 2.75877684e-01 7.15901375e-01
-4.91322652e-02 3.72017384e-01 -7.29181468e-01 -9.46191132e-01
4.68419433e-01 -4.98302966e-01 -7.85774171e-01 6.01261221e-02
3.52737069e-01 -1.46070085e-02 -5.17414153e-01 9.06367719e-01
8.57906282e-01 -1.68989554e-01 4.01566207e-01 -7.45122850e-01
-5.10526955e-01 1.20285428e+00 -2.58432567e-01 6.45975411e-01
3.90979171e-01 4.95427251e-02 1.11443138e+00 -1.32919884e+00
8.39784324e-01 1.46518850e+00 4.11066979e-01 3.01031798e-01
-1.33669126e+00 -6.28374755e-01 -2.92684227e-01 -1.49395198e-01
-1.33837318e+00 -1.32665479e+00 1.67431086e-01 -1.78501800e-01
1.74205971e+00 3.65092814e-01 5.16721547e-01 1.64514804e+00
1.72786955e-02 8.50842774e-01 1.10493040e+00 -7.29333580e-01
-2.15027288e-01 1.90840662e-01 -5.54294467e-01 3.36499065e-01
-4.03321922e-01 2.33579054e-01 -8.18482578e-01 -3.02706510e-01
4.22293812e-01 -8.70263636e-01 -1.09098494e-01 2.48209223e-01
-1.62272596e+00 6.59993827e-01 -2.98452497e-01 3.45955074e-01
-1.79852650e-01 -3.23994219e-01 5.89218974e-01 9.93276358e-01
7.93504357e-01 2.09988236e-01 -7.82576859e-01 -5.66273272e-01
-1.08589792e+00 4.32748869e-02 1.03644836e+00 1.29415882e+00
5.25639594e-01 3.78756225e-01 -7.92774092e-03 1.23651409e+00
1.63923487e-01 1.30526161e+00 6.08774602e-01 -8.04830909e-01
1.32047021e+00 -1.84475198e-01 -1.55785561e-01 -3.24771851e-01
-1.50145531e-01 -5.25096416e-01 -5.55974722e-01 -6.50445402e-01
1.97826430e-01 -4.49978143e-01 -7.10208654e-01 1.53275073e+00
2.19521895e-01 -1.92590788e-01 5.22525847e-01 6.28406107e-01
6.24845147e-01 1.00609064e+00 -3.19834501e-01 -5.06548405e-01
1.17666042e+00 -1.31975746e+00 -7.54468560e-01 -5.39912820e-01
7.44035542e-01 -1.48384309e+00 1.03734243e+00 5.93170971e-02
-1.34511626e+00 -4.44790870e-01 -6.89429045e-01 -1.64282862e-02
-2.67853588e-02 3.30473751e-01 2.71947622e-01 5.74377716e-01
-1.18472409e+00 6.11087717e-02 -9.59603906e-01 -5.47758043e-01
-9.87169296e-02 2.37085685e-01 -5.46629250e-01 1.54128179e-01
-1.34991896e+00 1.09116733e+00 1.35744274e-01 -2.13948190e-01
-8.98735106e-01 -4.19696212e-01 -8.38494897e-01 -2.30165094e-01
1.07312553e-01 -5.05346656e-01 1.73088884e+00 -8.27575505e-01
-1.68858588e+00 1.02836132e+00 -3.84342968e-01 -7.23611116e-01
6.81866229e-01 -5.30743636e-02 -9.59606647e-01 -2.58338321e-02
4.19749677e-01 5.46954870e-01 6.24556482e-01 -6.47417605e-01
-7.00951219e-01 -1.25411391e-01 -5.25549829e-01 3.55683267e-01
-2.54489750e-01 6.88694179e-01 -4.71985191e-01 -9.34728026e-01
-1.49690956e-01 -1.09262013e+00 -1.54474288e-01 -8.80476654e-01
-4.19002354e-01 7.37431794e-02 3.71356517e-01 -1.26690388e+00
1.19517028e+00 -2.01580501e+00 2.64715761e-01 5.36664836e-02
-4.72695947e-01 2.88280666e-01 -6.67582452e-01 9.18809772e-01
-1.61043063e-01 1.49914790e-02 -1.86607271e-01 -8.46745193e-01
-1.62810590e-02 4.92202669e-01 -4.84428912e-01 4.02225554e-01
1.77148402e-01 1.08772206e+00 -7.65837610e-01 -4.43563521e-01
3.64510603e-02 4.38084751e-01 -2.89648503e-01 1.00312561e-01
4.28266302e-02 6.46451652e-01 -2.03068871e-02 6.18798792e-01
2.56821722e-01 2.57391900e-01 3.88640314e-01 2.24407345e-01
-2.01931402e-01 1.24512899e+00 -4.39355969e-01 2.04996800e+00
-7.20129848e-01 6.10941470e-01 2.40359098e-01 -7.57533550e-01
7.74196625e-01 9.33955193e-01 2.24930808e-01 -1.11440742e+00
6.72292942e-03 6.98387802e-01 1.86197579e-01 -2.56550759e-01
5.07775366e-01 -2.93478012e-01 -2.10406572e-01 4.93661821e-01
4.93692577e-01 -4.93465453e-01 3.76036793e-01 7.97283724e-02
9.40082252e-01 -6.39928952e-02 2.25992516e-01 -2.96999961e-01
2.16266781e-01 2.15971380e-01 5.24181962e-01 4.48337108e-01
1.13365911e-01 5.86506546e-01 1.48936927e-01 -2.58461416e-01
-1.50561357e+00 -1.12204421e+00 -1.21980362e-01 1.04209721e+00
-7.68582106e-01 -6.77331269e-01 -8.02172184e-01 -6.40035510e-01
-4.40038413e-01 7.22316086e-01 -1.03249885e-01 3.81003976e-01
-9.33188677e-01 -7.85760462e-01 1.05863190e+00 1.44860268e-01
4.28860337e-02 -9.95433152e-01 5.12672722e-01 3.74380678e-01
-8.05176377e-01 -1.72069633e+00 -8.85633469e-01 2.42841810e-01
-7.15600789e-01 -5.93315005e-01 -8.55923772e-01 -9.61553812e-01
2.71259636e-01 9.27370563e-02 1.52006650e+00 -5.62812448e-01
2.76174277e-01 -3.86831202e-02 -3.70701134e-01 -2.55040824e-01
-1.37262642e+00 5.67866981e-01 4.93111849e-01 -2.77404159e-01
3.39094967e-01 -6.92634642e-01 2.09184825e-01 4.84804511e-01
-2.90928036e-01 1.43048957e-01 7.49443889e-01 8.38235140e-01
5.65305412e-01 -5.56850851e-01 5.64971209e-01 -4.96550649e-01
7.58150041e-01 -3.15951258e-01 -6.56717598e-01 2.27562517e-01
-2.16869384e-01 -4.76223826e-02 5.40588021e-01 -3.28340381e-01
-9.21638131e-01 1.72467548e-02 -5.43847322e-01 -3.47077608e-01
2.31512934e-02 6.27485454e-01 -1.76789746e-01 4.30296510e-01
7.84514725e-01 4.85775203e-01 3.85207012e-02 -6.49657547e-01
6.15778863e-01 1.23769057e+00 7.68018663e-01 -6.85395122e-01
8.84839356e-01 -8.69979635e-02 -7.10812986e-01 -1.03905809e+00
-4.00681585e-01 -5.57227373e-01 -5.85192025e-01 4.28621650e-01
6.97893798e-01 -1.54785049e+00 -7.08803460e-02 4.58987011e-03
-1.42344236e+00 -5.83640695e-01 -2.70993352e-01 7.41905928e-01
-7.26859689e-01 1.98638394e-01 -9.11958516e-01 -5.63638449e-01
-5.32729328e-01 -1.33675647e+00 1.30281377e+00 -6.81096435e-01
-5.07290006e-01 -9.10301566e-01 4.88211930e-01 6.32331908e-01
2.76251763e-01 -5.70730150e-01 6.51501656e-01 -9.59112883e-01
-1.84973747e-01 7.52762482e-02 2.02890888e-01 5.58248460e-01
2.64419556e-01 -2.28776425e-01 -7.61843860e-01 -5.48230588e-01
-1.31495431e-01 -3.38370293e-01 5.59888482e-01 1.24275990e-01
-3.81122977e-02 -6.93955719e-01 -8.92181769e-02 4.85941619e-01
6.66472673e-01 5.56095801e-02 3.76534015e-01 2.55011231e-01
3.82240921e-01 5.76122820e-01 4.84895736e-01 1.26290724e-01
5.28609395e-01 9.30522382e-01 -3.16863954e-01 -3.76172692e-01
-3.18011314e-01 -4.97129828e-01 8.43529999e-01 2.25133419e+00
1.98830023e-01 -2.87225962e-01 -1.10820019e+00 9.46301758e-01
-1.61593187e+00 -9.34344769e-01 6.35484755e-02 2.01406908e+00
1.15401399e+00 -6.86390921e-02 4.05971348e-01 -1.33519575e-01
5.12747467e-01 1.83241963e-01 1.62666753e-01 -5.37685215e-01
-6.44886076e-01 4.56494540e-01 4.71163929e-01 9.44008946e-01
-9.20435369e-01 1.45244992e+00 7.30137396e+00 1.10599267e+00
-9.67958629e-01 6.53863966e-01 5.37960529e-01 -3.47611338e-01
-4.55987394e-01 1.36953086e-01 -8.09628129e-01 2.85424650e-01
1.67602241e+00 -3.84501398e-01 7.81277180e-01 4.26521540e-01
5.13542950e-01 5.38062930e-01 -1.05542552e+00 8.32903743e-01
-1.25277355e-01 -1.34212232e+00 -4.21731770e-02 2.18365029e-01
1.03201318e+00 1.07563031e+00 -9.04488266e-02 5.25793970e-01
6.78202927e-01 -9.02436435e-01 9.33658838e-01 -2.27170452e-01
1.11798298e+00 -8.37488234e-01 5.84457994e-01 5.77525556e-01
-1.00848293e+00 2.63348877e-01 -4.47108537e-01 1.63873926e-01
4.42626774e-01 4.08294767e-01 -1.27617216e+00 8.23192239e-01
4.45899338e-01 7.75144517e-01 -3.41597795e-01 4.06567693e-01
-4.11170065e-01 1.17876995e+00 -5.81446886e-01 2.43403792e-01
3.40771794e-01 -3.13034058e-01 7.54834652e-01 1.55483663e+00
5.92018187e-01 -3.91103238e-01 3.06540817e-01 4.83636707e-01
-2.55102634e-01 5.13249397e-01 -6.73567235e-01 -4.02673483e-01
5.79387009e-01 9.31788921e-01 -2.15860218e-01 -3.47978473e-01
-7.25334167e-01 1.09586382e+00 2.23724887e-01 4.44992959e-01
-5.75968683e-01 3.75509560e-02 7.36804962e-01 9.22405273e-02
1.09419897e-01 -4.01686907e-01 9.60330218e-02 -1.40680742e+00
1.53132617e-01 -1.72617638e+00 -3.91015895e-02 -4.82305169e-01
-1.14906514e+00 1.17364502e+00 -2.91654825e-01 -1.18573165e+00
-8.60749900e-01 -2.19287843e-01 -2.09097818e-01 1.19308698e+00
-1.09318662e+00 -1.40907586e+00 6.39834106e-01 4.83190686e-01
9.80411530e-01 -8.62815499e-01 1.11191595e+00 5.91161430e-01
-3.15211147e-01 7.19385207e-01 2.54326910e-01 4.30070966e-01
9.71810400e-01 -1.00096309e+00 1.59875047e+00 8.05206895e-01
8.74835074e-01 5.49928606e-01 7.57583380e-01 -6.30823612e-01
-1.18229139e+00 -1.01496327e+00 1.57801437e+00 -8.10502529e-01
1.13879383e+00 -7.80505002e-01 -5.28282046e-01 9.61664975e-01
7.53407240e-01 -2.59182870e-01 6.63805425e-01 1.83615103e-01
-2.21840262e-01 7.68421739e-02 -4.87832576e-01 6.50768757e-01
1.13081098e+00 -9.52374637e-01 -7.79795587e-01 6.66419864e-01
1.05642307e+00 -5.38010478e-01 -9.77162361e-01 2.90631205e-01
4.63673890e-01 -4.81117100e-01 8.60402465e-01 -6.93142653e-01
2.71215320e-01 -2.13502552e-02 -5.98582387e-01 -1.85583949e+00
6.25550970e-02 -1.15371382e+00 3.32920760e-01 1.18847680e+00
1.20463419e+00 -5.24201393e-01 3.77713323e-01 -2.13487193e-01
-4.89678770e-01 -3.50811690e-01 -1.48517191e+00 -1.11625743e+00
4.62064415e-01 -4.87727761e-01 6.36989355e-01 1.04907572e+00
2.69128084e-01 8.99443448e-01 -5.80107331e-01 2.82368306e-02
2.98006088e-01 -2.70242363e-01 1.03125274e+00 -6.19060099e-01
-7.63160288e-01 -1.92211196e-01 -1.51549987e-02 -1.22298789e+00
4.54484373e-01 -1.30139601e+00 -1.50899859e-02 -1.37656832e+00
-6.84725791e-02 -2.22106844e-01 3.31693530e-01 6.41488791e-01
1.44539714e-01 1.95179790e-01 1.59155279e-01 3.62288475e-01
-2.70224333e-01 6.71684623e-01 1.06160545e+00 -2.44147614e-01
-1.62792295e-01 4.65285704e-02 -2.53477484e-01 2.39665776e-01
6.45821393e-01 -7.43242085e-01 -2.84197122e-01 -8.94093037e-01
8.75073373e-02 6.41781747e-01 -3.77865195e-01 -5.32441914e-01
-1.30996453e-02 5.73031157e-02 -1.03939742e-01 -4.54503119e-01
5.26135683e-01 -4.36388016e-01 1.99524522e-01 1.51551872e-01
-1.43159717e-01 6.34148657e-01 3.78445208e-01 1.91964775e-01
-6.29221559e-01 3.11506033e-01 3.91706437e-01 -1.58219755e-01
-5.14076650e-02 9.57205817e-02 -5.94258368e-01 3.96948427e-01
3.91240925e-01 2.83580840e-01 -2.72526175e-01 -6.74579203e-01
-8.15132678e-01 2.11014282e-02 2.98524261e-01 8.18115413e-01
9.22392830e-02 -1.41797495e+00 -1.42984414e+00 5.45445740e-01
1.59604192e-01 -4.09253150e-01 -3.62017661e-01 1.12599409e+00
-2.32608154e-01 7.69149363e-01 2.01601163e-01 -6.57518446e-01
-1.26740503e+00 1.65808037e-01 2.11406544e-01 -2.66466707e-01
-4.61369067e-01 6.97124362e-01 -2.07441479e-01 -1.23149014e+00
-3.00024580e-02 -1.70084968e-01 5.40534079e-01 -2.21896797e-01
3.08822066e-01 7.22502694e-02 4.42045182e-01 -1.25089824e+00
-3.69109243e-01 1.95813268e-01 -1.33160986e-02 -9.52825129e-01
1.18385148e+00 -2.75103986e-01 -7.76929408e-02 3.78855228e-01
1.17904627e+00 6.18981242e-01 -7.59232104e-01 -3.77476841e-01
1.83246091e-01 -1.95344359e-01 -3.27667177e-01 -8.21263075e-01
-7.67287254e-01 8.48590732e-01 1.16584539e-01 -2.71751463e-01
6.88527107e-01 1.22231878e-02 9.17781830e-01 5.57001710e-01
5.47209382e-01 -1.30005944e+00 -2.66570508e-01 8.93113494e-01
9.72552478e-01 -1.41749930e+00 -4.55358505e-01 -3.56838167e-01
-7.55541146e-01 7.09669948e-01 2.59351972e-02 3.79393578e-01
4.77691650e-01 7.51691520e-01 6.77666664e-01 3.75441283e-01
-9.35705125e-01 -7.48891979e-02 2.48114944e-01 5.32236159e-01
6.52244925e-01 3.15189242e-01 -2.83582181e-01 4.84689355e-01
-9.79269862e-01 -4.68385011e-01 2.11447075e-01 5.86761355e-01
-9.72743034e-02 -1.60264635e+00 -3.43127966e-01 -6.57149106e-02
-6.74753249e-01 -8.46957922e-01 -6.69826329e-01 6.39011443e-01
-4.19848651e-01 1.34252381e+00 -1.18220672e-02 -3.29678833e-01
3.98258507e-01 1.84303135e-01 3.61974210e-01 -7.66085327e-01
-7.22806156e-01 6.96925581e-01 8.41009378e-01 -3.19906443e-01
-8.92607048e-02 -6.24492526e-01 -9.90554690e-01 -4.50719416e-01
-2.12899148e-01 5.40225267e-01 9.09559906e-01 9.58645821e-01
3.17272127e-01 3.22993964e-01 5.78408360e-01 -6.34632289e-01
-8.28491032e-01 -1.41230083e+00 -1.41757563e-01 6.87658712e-02
2.50211656e-01 1.69589788e-01 -2.83742219e-01 1.17518492e-01] | [14.410694122314453, 7.220067977905273] |
86da367d-c636-462e-8745-ababaece95ab | deepclue-enhanced-image-clustering-via-multi | 2206.00359 | null | https://arxiv.org/abs/2206.00359v1 | https://arxiv.org/pdf/2206.00359v1.pdf | DeepCluE: Enhanced Image Clustering via Multi-layer Ensembles in Deep Neural Networks | Deep clustering has recently emerged as a promising technique for complex image clustering. Despite the significant progress, previous deep clustering works mostly tend to construct the final clustering by utilizing a single layer of representation, e.g., by performing $K$-means on the last fully-connected layer or by associating some clustering loss to a specific layer. However, few of them have considered the possibilities and potential benefits of jointly leveraging multi-layer representations for enhancing the deep clustering performance. In light of this, this paper presents a Deep Clustering via Ensembles (DeepCluE) approach, which bridges the gap between deep clustering and ensemble clustering by harnessing the power of multiple layers in deep neural networks. Particularly, we utilize a weight-sharing convolutional neural network as the backbone, which is trained with both the instance-level contrastive learning (via an instance projector) and the cluster-level contrastive learning (via a cluster projector) in an unsupervised manner. Thereafter, multiple layers of feature representations are extracted from the trained network, upon which a set of diversified base clusterings can be generated via a highly efficient clusterer. Then, the reliability of the clusters in multiple base clusterings is automatically estimated by exploiting an entropy-based criterion, based on which the multiple base clusterings are further formulated into a weighted-cluster bipartite graph. By partitioning this bipartite graph via transfer cut, the final image clustering result can therefore be obtained. Experimental results on six image datasets confirm the advantages of our DeepCluE approach over the state-of-the-art deep clustering approaches. | ['Jian-Huang Lai', 'Chang-Dong Wang', 'Xiangji Chen', 'Ding-Hua Chen', 'Dong Huang'] | 2022-06-01 | null | null | null | null | ['image-clustering'] | ['computer-vision'] | [-1.59717843e-01 -4.19368967e-02 2.29691997e-01 -3.56454134e-01
-5.35141587e-01 -4.10258740e-01 5.28831124e-01 2.57809088e-02
-3.38896424e-01 8.36529136e-02 -2.87355930e-01 1.86932907e-01
-3.40925932e-01 -8.39912355e-01 -7.38848805e-01 -1.25864637e+00
-2.13016972e-01 6.04256272e-01 -2.10448913e-02 1.74185932e-01
1.88478917e-01 2.98946887e-01 -1.61649036e+00 3.23300242e-01
1.05603397e+00 9.39844072e-01 2.16945305e-01 1.64777979e-01
-1.44858181e-01 4.90423918e-01 -3.64686787e-01 -2.89271265e-01
2.56441414e-01 -4.60452199e-01 -6.85822368e-01 3.62772226e-01
-1.47534773e-01 1.59638047e-01 -3.32336664e-01 1.03461003e+00
4.36073929e-01 1.72801659e-01 6.00134254e-01 -1.28532481e+00
-5.21681905e-01 9.75233376e-01 -8.04005444e-01 -2.98091382e-01
-9.56180971e-03 2.09506035e-01 1.09425199e+00 -9.67423916e-01
3.71110052e-01 9.84069467e-01 6.82821095e-01 3.35543662e-01
-1.52233279e+00 -8.74702990e-01 2.18825310e-01 1.69881999e-01
-1.74841881e+00 -1.83368757e-01 1.19624197e+00 -5.06882071e-01
7.00545430e-01 -1.97198570e-01 7.89667487e-01 7.66382575e-01
-7.61489719e-02 8.22104871e-01 1.04830575e+00 -1.40457332e-01
3.81168157e-01 1.07777929e-02 4.31086570e-02 7.58829057e-01
6.56281784e-03 -1.79546863e-01 -2.16607437e-01 1.51043132e-01
3.18589866e-01 2.50839084e-01 -1.74482390e-01 -7.40954816e-01
-1.12084639e+00 8.62241566e-01 1.22435844e+00 5.43391645e-01
-4.00615484e-01 2.16468781e-01 3.20494503e-01 -5.92807792e-02
3.78326535e-01 3.51888925e-01 -6.58430532e-02 5.16459763e-01
-1.44309008e+00 -1.50549829e-01 5.86177230e-01 6.31081820e-01
1.29670703e+00 -1.74008951e-01 2.32382752e-02 6.99462593e-01
5.49748480e-01 -5.38403988e-02 3.17583978e-01 -9.23705697e-01
2.08302811e-01 1.18849254e+00 -4.81664777e-01 -1.20893443e+00
-4.92440850e-01 -5.06915390e-01 -1.37880242e+00 2.71950692e-01
-5.56246564e-02 -1.29896224e-01 -7.57717192e-01 1.82922208e+00
2.97609717e-01 4.13777292e-01 -1.04590096e-01 8.32881510e-01
5.59312284e-01 6.31585121e-01 -5.29512987e-02 -1.29854858e-01
9.57593977e-01 -9.92017329e-01 -3.86313587e-01 2.70305812e-01
5.81331074e-01 -4.03282225e-01 7.31484592e-01 4.56080109e-01
-9.67547715e-01 -6.48143828e-01 -1.10632229e+00 1.45131469e-01
-5.43593347e-01 1.04257569e-01 5.99065006e-01 4.87485379e-01
-1.60961449e+00 7.66816080e-01 -1.10330796e+00 -2.34654799e-01
7.79282272e-01 7.86210716e-01 -2.46527418e-01 -1.98177591e-01
-8.54454875e-01 3.96480918e-01 7.12084115e-01 3.66623640e-01
-8.22625697e-01 -4.04235035e-01 -7.29072213e-01 2.95643359e-01
8.85160938e-02 -8.32037389e-01 1.64649323e-01 -1.19256949e+00
-1.52496815e+00 7.43637919e-01 1.80925906e-01 -7.75342360e-02
2.49970108e-01 2.08947688e-01 -2.44383030e-02 4.13482368e-01
1.37103610e-02 9.93398786e-01 8.35024774e-01 -1.69639075e+00
-4.55960393e-01 -5.18787503e-01 -1.58684105e-01 2.51437396e-01
-5.60007453e-01 -2.09768116e-01 -7.78374314e-01 -2.95830458e-01
1.78778902e-01 -1.09142470e+00 -5.23958862e-01 -2.92104393e-01
-6.67771459e-01 -5.12019098e-01 8.58892024e-01 -4.04520661e-01
1.37954152e+00 -2.14563465e+00 8.71799707e-01 7.04483211e-01
5.47425866e-01 1.50474340e-01 -1.41505569e-01 3.24578196e-01
-2.64843106e-01 3.08622539e-01 -6.87890708e-01 -6.92526519e-01
6.53203949e-02 -1.94059759e-02 1.90960363e-01 5.33053815e-01
3.06541502e-01 8.66595268e-01 -8.40694308e-01 -5.48002541e-01
5.21693766e-01 6.95171058e-01 -7.17724204e-01 2.32298389e-01
4.15802523e-02 5.62801719e-01 -2.59433925e-01 3.75538915e-01
7.33045518e-01 -4.22871441e-01 5.95735490e-01 -2.72584558e-01
-9.57579464e-02 -1.99025244e-01 -1.17961276e+00 2.03372335e+00
-3.89189839e-01 2.62151301e-01 1.09086856e-01 -1.62682438e+00
8.49984229e-01 1.37847930e-01 9.57636058e-01 -1.33615360e-01
2.30460301e-01 8.21032673e-02 4.09798063e-02 -9.56543833e-02
5.26680090e-02 1.38693064e-01 -3.12374592e-01 6.20013535e-01
3.76959652e-01 1.72622845e-01 2.26218581e-01 4.74918485e-01
1.13880205e+00 1.39183894e-01 -2.54809588e-01 -4.15931880e-01
6.07804179e-01 -1.68992117e-01 3.16420227e-01 4.42297012e-01
-2.37265415e-02 7.01389551e-01 4.52076167e-01 -3.08369279e-01
-8.79269600e-01 -1.11014986e+00 3.56205143e-02 9.16614294e-01
9.01499614e-02 -3.68768603e-01 -1.15240395e+00 -7.69543171e-01
-2.47020558e-01 1.71114266e-01 -6.79566085e-01 -3.47759277e-01
-4.87975299e-01 -1.09446049e+00 4.45497960e-01 3.73933941e-01
6.56736434e-01 -1.15670311e+00 -3.46282840e-01 4.04867887e-01
-2.96747595e-01 -7.67656028e-01 -9.46688130e-02 4.94780004e-01
-7.43143022e-01 -1.07647300e+00 -5.28128743e-01 -9.37085211e-01
8.56756628e-01 3.02155226e-01 9.21951652e-01 4.78460014e-01
-2.54137814e-01 3.03874969e-01 -3.40968639e-01 3.99252385e-01
-4.69656475e-02 3.87158126e-01 -4.93938429e-03 4.41218525e-01
4.64871854e-01 -9.62411940e-01 -8.91114533e-01 6.92587569e-02
-1.03541934e+00 -3.01560275e-02 7.27715254e-01 7.83062637e-01
5.12555420e-01 4.47866857e-01 6.27896309e-01 -6.38487995e-01
3.68946970e-01 -8.25712383e-01 -5.17178953e-01 1.33524403e-01
-4.69535291e-01 8.80848169e-02 8.56479466e-01 -1.39875203e-01
-7.97329605e-01 4.44824785e-01 -5.37615605e-02 -7.44720757e-01
-1.02613583e-01 6.59763217e-01 -4.27266538e-01 6.91264570e-02
3.31270963e-01 2.86604285e-01 6.66188151e-02 -1.56704351e-01
7.15287268e-01 5.51540554e-01 2.85392135e-01 -5.32032251e-01
9.79999483e-01 5.97935915e-01 -1.79224014e-01 -4.61509645e-01
-5.25410414e-01 -6.47686779e-01 -1.06773305e+00 -4.88862336e-01
1.41390014e+00 -8.74716401e-01 -9.14529562e-01 6.15523100e-01
-7.55236268e-01 -3.67033452e-01 -5.54639287e-03 3.38463724e-01
-5.17666757e-01 3.76284748e-01 -6.97726548e-01 -5.16000390e-01
-1.62974030e-01 -1.42705548e+00 9.82362926e-01 1.71103418e-01
5.65451980e-02 -9.96976554e-01 -9.47803259e-02 2.92968303e-01
9.92899537e-02 5.39671242e-01 9.65602279e-01 -4.45200175e-01
-6.93560600e-01 -1.41992210e-03 -2.80538082e-01 3.88568759e-01
-3.09488606e-02 3.46717954e-01 -9.33007956e-01 -5.68867445e-01
-3.47075790e-01 -4.50493157e-01 1.35729563e+00 4.58295435e-01
1.69532406e+00 -9.26537216e-02 -7.04938471e-01 8.77355814e-01
1.61220086e+00 -9.71939601e-03 6.04468405e-01 5.32640666e-02
1.02410889e+00 6.23367429e-01 -1.43871769e-01 3.01131159e-01
6.58812404e-01 4.08464253e-01 5.87306678e-01 -1.58839911e-01
-3.78160216e-02 1.07238576e-01 2.94046640e-01 1.28505969e+00
-2.41918266e-01 -1.40238091e-01 -9.89054441e-01 5.64197659e-01
-1.97073734e+00 -1.06039560e+00 1.25251457e-01 1.87121391e+00
6.37372375e-01 -7.05585256e-02 2.18486171e-02 3.07156771e-01
1.04763484e+00 1.60844952e-01 -6.17972910e-01 -8.77122059e-02
6.04510643e-02 1.57096222e-01 -5.84594086e-02 1.75238624e-01
-1.29955995e+00 8.82790446e-01 5.03613520e+00 7.86640644e-01
-1.04833221e+00 -1.51075870e-02 7.80329108e-01 -8.21835324e-02
-2.51404226e-01 1.09446933e-02 -4.59480852e-01 6.18442297e-01
7.90526748e-01 2.06484169e-01 7.56809711e-01 6.19046986e-01
-1.36565819e-01 1.82099894e-01 -1.15718007e+00 9.05499756e-01
5.53582376e-03 -1.34000039e+00 1.02090597e-01 3.62012118e-01
9.22561765e-01 2.58721888e-01 2.84489095e-01 5.19611597e-01
6.89938307e-01 -1.05759227e+00 3.67926002e-01 5.39300919e-01
6.65325403e-01 -1.07997680e+00 6.28611267e-01 3.69957447e-01
-1.44294381e+00 -2.26332128e-01 -3.77332002e-01 2.86610842e-01
-2.19856098e-01 9.02055860e-01 -4.10251349e-01 8.83761227e-01
1.08652794e+00 1.00149989e+00 -5.26715696e-01 7.24650264e-01
2.28553079e-02 6.13363504e-01 -3.91142756e-01 2.55789906e-01
4.50799584e-01 -7.59625971e-01 1.12082623e-01 1.32396865e+00
4.12298948e-01 4.67114523e-02 3.61205876e-01 1.09276164e+00
-3.85312974e-01 -1.23465426e-01 -6.22379422e-01 1.68842465e-01
3.58474016e-01 1.82678223e+00 -1.12567973e+00 -2.53427982e-01
-2.49578565e-01 9.89639640e-01 8.60240817e-01 3.58384460e-01
-7.86557853e-01 -3.45072776e-01 4.35394764e-01 -2.37547070e-01
6.71709955e-01 -1.99637771e-01 -3.06516975e-01 -1.15812469e+00
-2.18376428e-01 -5.90219676e-01 4.02322829e-01 -4.69519615e-01
-1.43288720e+00 7.45859683e-01 -2.82556176e-01 -1.15475965e+00
5.95180057e-02 -2.64925867e-01 -8.67489815e-01 5.95347643e-01
-1.38442576e+00 -1.27435589e+00 -3.87966275e-01 9.44198132e-01
1.84880629e-01 -1.11285627e-01 7.10353613e-01 4.10528511e-01
-1.01737773e+00 4.08243179e-01 2.19327122e-01 3.01077753e-01
3.94332945e-01 -1.41836822e+00 -1.68433547e-01 8.00629735e-01
1.18450187e-01 7.93423414e-01 1.57623105e-02 -3.49780470e-01
-1.14675128e+00 -1.45346820e+00 3.97818357e-01 -3.58935028e-01
5.42330086e-01 -6.44478321e-01 -9.82157648e-01 4.42993253e-01
4.68237579e-01 1.26170143e-01 8.98060560e-01 1.46878660e-01
-3.97986203e-01 -2.97215402e-01 -9.43605602e-01 4.82921302e-01
1.05331969e+00 -3.87143880e-01 -2.70166636e-01 1.04880892e-01
6.39552355e-01 1.85163856e-01 -1.12189317e+00 3.33770961e-01
2.49065086e-01 -1.23011875e+00 8.22726369e-01 -2.13354006e-01
6.78384125e-01 -6.43466294e-01 -1.31457346e-02 -1.61155820e+00
-6.69578135e-01 -3.36572826e-01 1.45393834e-01 1.30731666e+00
2.52790153e-01 -1.43291399e-01 6.84600353e-01 1.73872530e-01
-4.52604562e-01 -7.57927477e-01 -9.03154969e-01 -3.57631028e-01
2.49004528e-01 -2.13995814e-01 6.05876923e-01 1.21231306e+00
7.00252429e-02 4.17374611e-01 4.80815908e-03 2.12429404e-01
8.38742733e-01 1.91938385e-01 6.21638000e-01 -1.41482127e+00
-1.19254990e-02 -7.49405742e-01 -2.93510407e-01 -7.07041562e-01
6.58249855e-01 -1.46179795e+00 1.71214521e-01 -1.58751738e+00
3.92736465e-01 -5.69276452e-01 -7.04061627e-01 4.50421214e-01
-2.40977347e-01 4.46707964e-01 3.14327598e-01 2.99047470e-01
-1.03681242e+00 7.48865008e-01 9.50188398e-01 -2.96038270e-01
-1.52609706e-01 -2.93523103e-01 -7.86903024e-01 7.35182762e-01
7.43636668e-01 -4.95900035e-01 -2.85997212e-01 -4.16457891e-01
3.02041650e-01 -1.43605515e-01 3.07463646e-01 -1.31781125e+00
5.48278928e-01 4.20251608e-01 5.58514118e-01 -5.99897325e-01
1.32603005e-01 -9.86094534e-01 8.45963061e-02 4.43701416e-01
-2.50485152e-01 -2.06950545e-01 -1.36925876e-01 7.77695775e-01
-3.04955482e-01 2.60062218e-01 9.30675268e-01 -2.13464960e-01
-4.85535204e-01 5.74364603e-01 -4.02314663e-01 -2.73390204e-01
1.21938193e+00 -2.41787702e-01 3.63361202e-02 7.59768188e-02
-9.70727503e-01 4.45832551e-01 5.18466949e-01 1.53633520e-01
6.01765633e-01 -1.41306019e+00 -6.21438563e-01 2.81778455e-01
-3.51923588e-03 4.24206465e-01 3.20992112e-01 9.62212443e-01
-2.47238800e-01 -4.52104695e-02 -2.58389503e-01 -1.03613532e+00
-5.36110044e-01 8.07810187e-01 3.61646265e-01 -4.27477956e-01
-4.73913670e-01 7.22145677e-01 3.01625699e-01 -7.30550289e-01
1.25139758e-01 8.80454108e-02 -4.08359647e-01 2.59914845e-01
1.05374970e-01 3.07792336e-01 6.83026910e-02 -6.90192461e-01
-5.24831593e-01 7.41389036e-01 6.74672052e-02 8.37771073e-02
1.56311893e+00 -2.72833854e-01 -5.53980470e-01 2.62922525e-01
1.52598739e+00 -5.37630200e-01 -1.46912491e+00 -8.50563422e-02
2.71049030e-02 1.51951864e-01 1.94445010e-02 -4.10862148e-01
-1.73323858e+00 1.09142971e+00 4.26950842e-01 2.36309335e-01
1.41095102e+00 8.58149156e-02 6.03575408e-01 3.02994937e-01
2.43958384e-01 -1.18738306e+00 3.93358350e-01 2.28430122e-01
5.52114129e-01 -1.30865169e+00 -1.81785941e-01 1.38810743e-03
-4.70323443e-01 1.03475857e+00 5.85269153e-01 -5.87221503e-01
1.05304170e+00 1.65912822e-01 -6.16022311e-02 -4.94192362e-01
-6.18528903e-01 -3.10188144e-01 8.88383389e-02 6.22147560e-01
2.90279686e-01 2.09455818e-01 -1.35578718e-02 5.54250479e-01
-1.32755667e-04 -3.28650028e-01 3.28969091e-01 3.04309279e-01
-3.05865109e-01 -8.40313673e-01 -1.29433811e-01 4.19573516e-01
-1.64710611e-01 -8.36520419e-02 -4.15328413e-01 6.68835938e-01
5.20154238e-01 1.10750377e+00 2.90025294e-01 -7.13244140e-01
-1.82568893e-01 -2.63862815e-02 1.80813462e-01 -6.51414335e-01
-8.61267745e-01 6.93233684e-02 -6.05704248e-01 -6.97562635e-01
-9.67563629e-01 -5.28565884e-01 -1.30362618e+00 -2.61908561e-01
-3.40571433e-01 2.05587596e-01 4.90747899e-01 8.77285719e-01
4.66983050e-01 6.61774516e-01 1.09004664e+00 -1.36513531e+00
-7.63056204e-02 -7.09114790e-01 -6.51207685e-01 4.98141557e-01
-4.35360111e-02 -6.16477191e-01 -5.15453994e-01 6.36690110e-02] | [9.121519088745117, 3.2436153888702393] |
7d3dcac8-c249-43cb-ab02-d003d685cfcb | possibility-before-utility-learning-and-using-1 | 2203.12686 | null | https://arxiv.org/abs/2203.12686v1 | https://arxiv.org/pdf/2203.12686v1.pdf | Possibility Before Utility: Learning And Using Hierarchical Affordances | Reinforcement learning algorithms struggle on tasks with complex hierarchical dependency structures. Humans and other intelligent agents do not waste time assessing the utility of every high-level action in existence, but instead only consider ones they deem possible in the first place. By focusing only on what is feasible, or "afforded", at the present moment, an agent can spend more time both evaluating the utility of and acting on what matters. To this end, we present Hierarchical Affordance Learning (HAL), a method that learns a model of hierarchical affordances in order to prune impossible subtasks for more effective learning. Existing works in hierarchical reinforcement learning provide agents with structural representations of subtasks but are not affordance-aware, and by grounding our definition of hierarchical affordances in the present state, our approach is more flexible than the multitude of approaches that ground their subtask dependencies in a symbolic history. While these logic-based methods often require complete knowledge of the subtask hierarchy, our approach is able to utilize incomplete and varying symbolic specifications. Furthermore, we demonstrate that relative to non-affordance-aware methods, HAL agents are better able to efficiently learn complex tasks, navigate environment stochasticity, and acquire diverse skills in the absence of extrinsic supervision -- all of which are hallmarks of human learning. | ['Fei Sha', 'Shariq Iqbal', 'Robby Costales'] | 2022-03-23 | possibility-before-utility-learning-and-using | https://openreview.net/forum?id=7b4zxUnrO2N | https://openreview.net/pdf?id=7b4zxUnrO2N | iclr-2022-4 | ['hierarchical-reinforcement-learning'] | ['methodology'] | [ 2.77592659e-01 3.41580063e-01 -2.70781606e-01 -5.94998859e-02
-4.41021502e-01 -7.71523595e-01 7.21201956e-01 4.04169589e-01
-5.23290217e-01 8.47709239e-01 2.17246100e-01 -4.27358687e-01
-5.15069962e-01 -9.40210342e-01 -6.11836016e-01 -4.64626074e-01
-3.28708231e-01 6.18102312e-01 4.63233918e-01 -6.23426139e-01
3.10554624e-01 4.92961466e-01 -1.71267140e+00 4.02946323e-02
8.29378486e-01 5.75036287e-01 3.01411510e-01 8.11290443e-01
3.92365277e-01 1.09545612e+00 -3.26812088e-01 5.85697442e-02
1.92600816e-01 -3.57525855e-01 -1.06071913e+00 6.44205436e-02
1.90216061e-02 -6.91392124e-01 -2.55075783e-01 8.64031672e-01
-1.08274564e-01 3.69231224e-01 6.94179952e-01 -1.15967834e+00
-3.38789850e-01 1.03319168e+00 2.43957520e-01 1.96464986e-01
7.02387571e-01 6.95387602e-01 1.38318217e+00 -2.22019553e-01
5.86307883e-01 1.38556671e+00 1.74989402e-01 6.19378209e-01
-1.48009551e+00 -1.68073788e-01 6.95206165e-01 8.54833797e-02
-8.43377471e-01 -3.32007080e-01 5.38919508e-01 -6.39266133e-01
1.35687470e+00 2.29250595e-01 9.61646080e-01 8.55090559e-01
1.64288998e-01 9.51217055e-01 1.29463899e+00 -5.45296013e-01
3.72643471e-01 -4.37489212e-01 1.11356981e-01 9.05768037e-01
1.65130571e-01 7.35967994e-01 -6.01185262e-01 -6.71630576e-02
9.55460131e-01 4.30009849e-02 -1.07707947e-01 -8.19787145e-01
-1.47154582e+00 6.19400740e-01 3.87545139e-01 3.64908725e-01
-2.48156503e-01 4.04208243e-01 2.06233248e-01 7.06193209e-01
-4.27727729e-01 1.21945810e+00 -5.71162403e-01 -1.15390204e-01
-4.45611745e-01 6.28017247e-01 8.74247372e-01 8.41290772e-01
7.85471737e-01 4.69374955e-02 -3.57256946e-03 -6.27209023e-02
2.22938642e-01 7.98801109e-02 4.13655728e-01 -1.44112766e+00
2.90684611e-01 7.15486288e-01 4.81341273e-01 -3.81357282e-01
-4.84169453e-01 -1.75958753e-01 -4.27388772e-02 7.16075122e-01
4.22354668e-01 8.53051245e-03 -8.41993153e-01 2.01181269e+00
2.93898582e-01 -1.74431846e-01 2.59645790e-01 8.21714401e-01
4.40496281e-02 3.79450351e-01 2.90518761e-01 -2.21417204e-01
9.95027304e-01 -7.83630550e-01 -3.05743635e-01 -1.86504215e-01
8.07329357e-01 2.40279417e-02 1.36382544e+00 6.41572177e-01
-1.06849504e+00 -3.40824366e-01 -1.40565348e+00 -4.31032851e-02
-1.58410758e-01 -5.98654389e-01 1.07741380e+00 3.70686054e-01
-9.94369030e-01 9.86558199e-01 -1.16027904e+00 -6.29273653e-02
1.95400104e-01 5.54724455e-01 -2.47738123e-01 1.46708488e-01
-1.11350060e+00 1.22892034e+00 6.92279875e-01 -2.27296337e-01
-1.71261525e+00 -3.67845774e-01 -1.03092706e+00 1.56552270e-01
9.12542820e-01 -6.77616954e-01 1.72204971e+00 -8.21965635e-01
-1.49502957e+00 3.31523478e-01 2.39413008e-01 -5.56121349e-01
4.36885297e-01 -1.72003776e-01 5.27056530e-02 2.97258925e-02
-1.33130448e-02 6.66898072e-01 9.00778532e-01 -1.10022748e+00
-8.39087069e-01 -4.35294122e-01 1.06495416e+00 5.96786559e-01
5.59409745e-02 -2.83455431e-01 2.61190504e-01 -2.19454229e-01
2.52621859e-01 -1.01402867e+00 -5.43702900e-01 -2.41472542e-01
-1.97300270e-01 -5.49030721e-01 2.93402702e-01 -2.25620717e-03
1.01754105e+00 -2.00726366e+00 8.30547035e-01 3.46674398e-02
4.46050495e-01 -2.51069129e-01 -9.35266241e-02 5.69925368e-01
1.49547413e-01 1.90526783e-01 -2.51648903e-01 1.10790133e-01
5.13886631e-01 6.29898190e-01 -2.86494434e-01 2.06406936e-01
2.73275048e-01 8.61056209e-01 -1.30563939e+00 -4.56038028e-01
3.05614293e-01 -1.07309699e-01 -9.08021271e-01 2.99509704e-01
-7.92440832e-01 5.46833932e-01 -8.77788126e-01 5.25647461e-01
-3.08855861e-01 3.83344181e-02 4.89657938e-01 4.29905802e-01
-9.00988579e-02 5.94361186e-01 -1.36643887e+00 1.69447041e+00
-5.58401823e-01 1.62180677e-01 -7.04358369e-02 -7.54738748e-01
2.95478344e-01 2.55354375e-01 2.16632485e-01 -5.87414086e-01
-1.81925058e-01 2.75210232e-01 4.32732731e-01 -5.89347780e-01
2.41572559e-01 -3.99583101e-01 -3.39441895e-01 6.18264437e-01
7.58120567e-02 -7.44733810e-01 4.76164073e-01 1.27312481e-01
1.25609994e+00 8.89204979e-01 6.22060776e-01 -1.88419446e-01
2.80658424e-01 1.21611983e-01 6.41681969e-01 9.91559088e-01
-1.88537017e-01 -1.26017824e-01 8.32670152e-01 -6.93600476e-01
-1.07963395e+00 -1.18740213e+00 1.53116390e-01 1.54577482e+00
1.17127314e-01 -3.93361509e-01 -6.71147764e-01 -6.62002563e-01
-5.88544682e-02 9.34183657e-01 -5.89695036e-01 -6.31477311e-02
-8.05431545e-01 -1.83304310e-01 2.82021314e-01 4.91385758e-01
5.69429025e-02 -1.43999612e+00 -1.66962230e+00 2.98533499e-01
1.77316815e-01 -6.90436542e-01 -2.21293584e-01 6.64632201e-01
-9.61631358e-01 -1.21274114e+00 7.92934969e-02 -5.36360800e-01
5.84721148e-01 -2.68491745e-01 1.19034398e+00 5.13682544e-01
-2.05720454e-01 5.63079894e-01 -2.79330403e-01 -3.29025805e-01
-4.57156301e-01 1.01552472e-01 1.77495062e-01 -6.10542893e-01
1.01014212e-01 -7.91835368e-01 -2.86024153e-01 -5.40467277e-02
-8.69615555e-01 1.51755944e-01 5.98807931e-01 9.21327591e-01
4.21110243e-01 3.43564421e-01 2.24835113e-01 -7.58879662e-01
8.30537379e-01 -2.12782070e-01 -7.85610497e-01 3.06712478e-01
-6.49881065e-01 7.14313686e-01 6.02231026e-01 -3.38403404e-01
-9.40177441e-01 1.81036443e-01 3.06878597e-01 1.14363976e-01
-3.90441090e-01 5.07787645e-01 -2.97531068e-01 2.89929330e-01
1.02171838e+00 1.54857665e-01 -1.93028688e-01 -1.16411962e-01
4.64570075e-01 -4.45995405e-02 2.06879705e-01 -1.45374084e+00
6.85137093e-01 -1.85685549e-02 3.42141241e-01 -5.39164543e-01
-8.67330074e-01 -1.03940770e-01 -8.86831224e-01 -1.11318290e-01
7.44153738e-01 -4.78066772e-01 -1.14011741e+00 3.19111645e-02
-6.47838533e-01 -8.80641878e-01 -5.57205439e-01 1.87189981e-01
-1.14008963e+00 1.53830454e-01 -5.10586560e-01 -1.04974163e+00
4.07960206e-01 -1.68507779e+00 8.40180874e-01 -5.84450439e-02
-5.40779889e-01 -6.37908220e-01 -2.25689694e-01 -3.28145213e-02
5.87553270e-02 3.91956002e-01 1.36436427e+00 -5.77475667e-01
-9.69386280e-01 1.41304165e-01 3.82147551e-01 -3.92598622e-02
2.96201296e-02 -1.45927310e-01 -5.27908325e-01 -2.20284089e-01
4.97072563e-02 -8.36851597e-01 5.04289687e-01 1.85984839e-02
8.42716694e-01 -5.51136434e-01 -1.18451573e-01 1.17176905e-01
1.27593338e+00 2.14190230e-01 2.78512746e-01 6.07415318e-01
3.76865327e-01 8.97164643e-01 7.16538846e-01 3.16242278e-01
5.69944978e-01 6.30937576e-01 7.92940080e-01 6.42608285e-01
2.40391448e-01 -5.41468680e-01 2.73505598e-01 1.73069030e-01
-3.40178490e-01 2.78204143e-01 -9.20430541e-01 4.20888573e-01
-2.04312301e+00 -1.03678310e+00 3.11850667e-01 2.17168355e+00
1.24852264e+00 5.70245385e-01 4.12734270e-01 3.07554752e-01
1.22937515e-01 7.00380951e-02 -9.33995187e-01 -4.70590502e-01
4.14932460e-01 1.42325684e-01 1.98257640e-01 8.75467718e-01
-8.92761469e-01 1.08811104e+00 6.64191437e+00 2.63211399e-01
-5.02193928e-01 -1.55933782e-01 2.20308095e-01 -2.36549843e-02
-5.52847266e-01 3.67264897e-01 -4.61958051e-01 -3.24305594e-02
7.71040797e-01 1.09681122e-01 1.06102872e+00 8.49553287e-01
-1.53051421e-01 -3.75019729e-01 -1.87457299e+00 3.21211010e-01
-4.51328933e-01 -9.54038739e-01 -6.66266773e-03 1.21052474e-01
4.35383618e-01 -2.89072305e-01 -2.58788951e-02 5.60164750e-01
9.88954186e-01 -1.29744124e+00 1.18129134e+00 4.95569140e-01
3.66961986e-01 -7.47819483e-01 2.28913710e-01 9.27577734e-01
-9.76940691e-01 -6.61380947e-01 -5.74470647e-02 -6.77046716e-01
-2.25469712e-02 -1.47340909e-01 -6.63024306e-01 2.69778341e-01
2.67584980e-01 2.17807710e-01 -3.17136198e-01 5.16493082e-01
-8.55612159e-01 1.41914248e-01 -3.09078097e-01 -4.63240415e-01
3.88068408e-01 -1.48909679e-02 4.84031320e-01 6.10736430e-01
5.44011686e-03 4.42307085e-01 6.26504421e-01 7.69123256e-01
5.01397610e-01 -4.06660348e-01 -6.72446072e-01 -1.65056705e-01
4.92456287e-01 7.26652741e-01 -4.92319196e-01 -3.02447468e-01
-1.67901382e-01 4.44329143e-01 5.90484500e-01 2.80889094e-01
-4.85061765e-01 1.79850925e-02 5.67389131e-01 3.54707539e-02
2.73315255e-02 -7.87764370e-01 -1.94136530e-01 -1.17302108e+00
-1.93138719e-01 -1.29953051e+00 4.60916609e-01 -7.46340156e-01
-6.46385431e-01 2.63060749e-01 3.29955220e-01 -8.05006802e-01
-6.21677816e-01 -8.10070693e-01 -1.17520988e-01 5.20595253e-01
-1.36420584e+00 -1.01038599e+00 1.91790551e-01 5.50517440e-01
7.58003592e-01 3.92100886e-02 1.08209014e+00 -6.22869849e-01
-6.36695996e-02 -9.29470956e-02 -8.12863290e-01 -2.37560138e-01
2.68584434e-02 -1.61045408e+00 2.14061975e-01 7.68195093e-01
1.96592808e-01 9.58895147e-01 1.01197755e+00 -5.63643456e-01
-1.62780952e+00 -3.69812667e-01 6.20018840e-01 -8.87156188e-01
9.19177711e-01 -1.84868902e-01 -5.96574306e-01 9.76863801e-01
2.41842400e-03 -2.63413340e-01 3.21740776e-01 4.46494699e-01
-4.75486279e-01 1.31000906e-01 -9.89774048e-01 1.03109109e+00
1.45927560e+00 -5.88266551e-01 -1.40977120e+00 1.51077881e-01
1.04381812e+00 -5.29924810e-01 -7.63161659e-01 3.57284427e-01
6.17764294e-01 -1.19477713e+00 8.54321122e-01 -1.04263210e+00
3.78808469e-01 -5.33338666e-01 -3.29330146e-01 -1.22849572e+00
-3.78494799e-01 -8.37060988e-01 -3.23583275e-01 4.46748346e-01
4.95644033e-01 -4.88932252e-01 5.38998902e-01 7.96570003e-01
-2.68778980e-01 -6.40930772e-01 -7.93074369e-01 -6.75612569e-01
1.18700556e-01 -6.13624334e-01 9.20949996e-01 6.35921299e-01
4.53898787e-01 2.54705399e-01 6.10147715e-02 1.90222427e-01
7.24721014e-01 3.52349699e-01 5.21374643e-01 -1.34676659e+00
-6.85321152e-01 -5.65267026e-01 -1.08568467e-01 -9.20372844e-01
2.82434493e-01 -7.92300224e-01 3.04854959e-01 -1.45712578e+00
-1.60755161e-02 -6.21462882e-01 -5.01939297e-01 7.61445761e-01
1.05155230e-01 -5.24911940e-01 3.01028907e-01 2.29976162e-01
-8.21393967e-01 2.26047084e-01 1.58255446e+00 -1.30368859e-01
-3.49911630e-01 -6.00004606e-02 -7.17532575e-01 9.37123597e-01
7.65164554e-01 -2.85669267e-01 -7.95292556e-01 -4.04621631e-01
6.69591010e-01 3.00953776e-01 4.27030236e-01 -9.72108066e-01
1.84965879e-01 -8.20694923e-01 2.22653836e-01 1.51151670e-02
6.93836510e-02 -8.25442970e-01 -3.66907008e-02 8.15754414e-01
-8.47125053e-01 2.41333053e-01 -1.37171298e-01 4.89256442e-01
2.28249609e-01 -4.62835580e-01 3.31325531e-01 -6.80814862e-01
-9.66394722e-01 1.81026980e-01 -6.31916344e-01 -7.30490312e-02
9.96095300e-01 -5.84343970e-02 -4.30249870e-02 -7.43593052e-02
-1.13700116e+00 3.41345578e-01 6.01743281e-01 1.82452276e-01
5.42254627e-01 -9.26954627e-01 -2.08251551e-01 1.44725323e-01
1.82663023e-01 2.31157973e-01 -1.71611294e-01 4.42394942e-01
-3.44217241e-01 3.77934456e-01 -5.08330643e-01 -2.64724046e-01
-6.85767233e-01 9.21413124e-01 4.18338031e-01 -3.88175845e-01
-5.65681815e-01 7.10883260e-01 2.83304453e-01 -4.89328712e-01
2.75035948e-01 -7.50799477e-01 -2.90938407e-01 -1.02910981e-01
2.44945049e-01 5.58075570e-02 -1.38695151e-01 -2.35530719e-01
-1.65301263e-01 2.18703896e-01 6.68950677e-02 -5.46488583e-01
1.27699816e+00 2.89059710e-02 -9.61413160e-02 5.72928727e-01
2.16669679e-01 -2.76306182e-01 -1.77955186e+00 -4.66164537e-02
4.08589423e-01 -3.40879619e-01 -2.04890966e-01 -8.26314628e-01
-1.21199548e-01 8.08026135e-01 -3.33901197e-02 6.30782127e-01
9.50172663e-01 5.35910688e-02 1.53804302e-01 1.08766663e+00
9.63234723e-01 -1.23007584e+00 4.17417496e-01 6.79669201e-01
9.04801130e-01 -9.33987319e-01 9.57665667e-02 4.18147147e-02
-5.99666297e-01 1.16006505e+00 8.31674695e-01 -1.64531425e-01
1.28387839e-01 3.05801749e-01 -5.02260983e-01 -2.10825279e-01
-1.07795978e+00 -5.30178368e-01 -3.26285958e-02 8.32540035e-01
1.62791178e-01 3.08364928e-01 2.44820733e-02 2.21533373e-01
-4.14684147e-01 -2.26559207e-01 5.72732210e-01 1.38373804e+00
-8.36480200e-01 -1.22042537e+00 -2.04340681e-01 3.25621158e-01
-2.59256572e-01 2.66444683e-02 -2.30365172e-01 9.41718936e-01
1.39822707e-01 8.44489217e-01 -2.61242956e-01 -1.21091291e-01
4.64996159e-01 2.23999798e-01 1.10072899e+00 -1.18034852e+00
-3.54158342e-01 -9.43956226e-02 2.47344390e-01 -8.44594955e-01
-3.47585499e-01 -9.49414551e-01 -1.61374545e+00 -2.55470742e-02
1.11157611e-01 7.02419952e-02 2.54828811e-01 1.25345850e+00
-2.57220685e-01 6.47558570e-01 3.87428999e-01 -9.73463655e-01
-1.06177723e+00 -4.49629098e-01 -3.61539751e-01 3.04253668e-01
8.28334570e-01 -8.56285870e-01 -2.99401313e-01 -2.95440331e-02] | [4.1276679039001465, 1.3767915964126587] |
f6e9c26c-d3e4-4c1a-b11c-f556c3fd20da | unsupervised-domain-attention-adaptation | 2007.09344 | null | https://arxiv.org/abs/2007.09344v1 | https://arxiv.org/pdf/2007.09344v1.pdf | Unsupervised Domain Attention Adaptation Network for Caricature Attribute Recognition | Caricature attributes provide distinctive facial features to help research in Psychology and Neuroscience. However, unlike the facial photo attribute datasets that have a quantity of annotated images, the annotations of caricature attributes are rare. To facility the research in attribute learning of caricatures, we propose a caricature attribute dataset, namely WebCariA. Moreover, to utilize models that trained by face attributes, we propose a novel unsupervised domain adaptation framework for cross-modality (i.e., photos to caricatures) attribute recognition, with an integrated inter- and intra-domain consistency learning scheme. Specifically, the inter-domain consistency learning scheme consisting an image-to-image translator to first fill the domain gap between photos and caricatures by generating intermediate image samples, and a label consistency learning module to align their semantic information. The intra-domain consistency learning scheme integrates the common feature consistency learning module with a novel attribute-aware attention-consistency learning module for a more efficient alignment. We did an extensive ablation study to show the effectiveness of the proposed method. And the proposed method also outperforms the state-of-the-art methods by a margin. The implementation of the proposed method is available at https://github.com/KeleiHe/DAAN. | ['Zheng Gu', 'Yang Gao', 'Wen Ji', 'Jing Huo', 'Kelei He'] | 2020-07-18 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/429_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123530018.pdf | eccv-2020-8 | ['caricature'] | ['computer-vision'] | [ 1.17042236e-01 4.96076494e-02 -2.91819870e-01 -9.32776809e-01
-7.96270192e-01 -3.34932595e-01 5.57110965e-01 -2.37597749e-01
-1.47189423e-01 4.58820790e-01 1.37106925e-01 4.49298471e-01
3.40701193e-02 -4.69341516e-01 -8.33510637e-01 -6.85706913e-01
4.97537345e-01 5.71417212e-01 -2.36495018e-01 1.92982003e-01
6.73693717e-02 3.28135043e-01 -1.79048944e+00 4.07331347e-01
1.02179754e+00 1.34493685e+00 2.60709316e-01 -4.89368699e-02
-3.19939226e-01 4.18987930e-01 -1.10617585e-01 -7.93857992e-01
2.69067019e-01 -5.31030416e-01 -9.00008559e-01 4.25013036e-01
9.79705334e-01 -4.45251286e-01 -4.55358252e-02 1.15786064e+00
5.60400128e-01 -7.40457550e-02 8.03943455e-01 -1.88760018e+00
-1.08863854e+00 2.90013522e-01 -7.13212609e-01 -3.36794764e-01
2.56741345e-01 8.89194757e-03 8.38657796e-01 -1.03026152e+00
7.35638201e-01 1.48828661e+00 5.74649453e-01 7.37486541e-01
-1.09247136e+00 -1.22252059e+00 1.79511875e-01 6.02309227e-01
-1.74431431e+00 -6.80292666e-01 1.01546323e+00 -3.57456475e-01
2.62454510e-01 -1.29010186e-01 5.07808864e-01 1.30571365e+00
-2.27361232e-01 8.71454298e-01 1.33488810e+00 -1.28223538e-01
-3.19617987e-02 4.73780811e-01 -1.81074992e-01 7.07863152e-01
-5.38642406e-02 -6.62084296e-02 -2.36664370e-01 -3.54666561e-02
7.14248717e-01 2.14983985e-01 5.08316495e-02 -5.20440340e-01
-9.03473735e-01 6.84319139e-01 4.54665571e-01 3.65268625e-02
-2.21203402e-01 6.53115734e-02 4.47212934e-01 1.35197759e-01
4.22677010e-01 -6.16796948e-02 -4.61926281e-01 2.26492837e-01
-6.39007688e-01 6.85094222e-02 3.48078489e-01 1.48714650e+00
1.04200304e+00 -1.86062947e-01 -1.40251681e-01 1.25389743e+00
4.38342571e-01 5.99299371e-01 3.92768443e-01 -1.00440812e+00
3.47036302e-01 7.30473816e-01 -2.79081732e-01 -8.27758491e-01
-1.71951741e-01 5.44066727e-02 -9.86212730e-01 3.47592384e-02
1.86305970e-01 2.09555939e-01 -7.46737778e-01 1.95420623e+00
6.66327000e-01 4.92731184e-01 3.73902395e-02 9.44240868e-01
9.96149063e-01 3.44625205e-01 5.38656592e-01 -9.32981670e-02
1.62700498e+00 -1.01013744e+00 -8.79109263e-01 -1.63770556e-01
5.53626418e-01 -9.98345256e-01 1.26333249e+00 1.69584066e-01
-8.90535951e-01 -7.80573308e-01 -8.91055286e-01 -3.07931662e-01
-3.05929333e-01 7.03931034e-01 5.51533699e-01 3.83061975e-01
-7.49897659e-01 2.60919482e-01 -2.86731780e-01 -5.80477834e-01
1.07056820e+00 2.55117476e-01 -1.00764382e+00 -2.77284175e-01
-9.24162388e-01 8.82716656e-01 4.97359127e-01 -9.89393964e-02
-1.06846070e+00 -6.89970016e-01 -9.87350583e-01 -1.81255892e-01
2.80355364e-01 -5.60456157e-01 1.05338120e+00 -1.70971870e+00
-1.26693928e+00 1.41821313e+00 -2.33697757e-01 1.12610921e-01
2.69578189e-01 -1.57321274e-01 -6.12090945e-01 1.24867603e-01
2.56143302e-01 9.93831217e-01 9.43007231e-01 -1.31594229e+00
-5.63194931e-01 -6.25037074e-01 -2.30788693e-01 2.65624344e-01
-4.42589790e-01 1.74532384e-02 -5.65200210e-01 -6.05747521e-01
-1.97681095e-02 -1.03187549e+00 4.19212013e-01 5.94945431e-01
-2.07648709e-01 -4.21692312e-01 8.10267806e-01 -7.17374384e-01
8.20398271e-01 -2.33972740e+00 2.21195072e-02 1.71327963e-03
-5.04412390e-02 9.29875076e-02 -5.18053412e-01 1.01075411e-01
-3.05994153e-01 -3.52426469e-02 -2.99284816e-01 -5.86712778e-01
7.63841858e-03 1.86825067e-01 8.02795961e-02 5.43125451e-01
5.25801659e-01 7.15096116e-01 -6.93948507e-01 -9.88207996e-01
7.05253333e-02 4.12319809e-01 -4.72008109e-01 4.40128326e-01
-2.01131422e-02 4.65263784e-01 -5.03839493e-01 1.07483840e+00
1.17871094e+00 2.63340585e-02 1.65119786e-02 -6.59241199e-01
2.53750324e-01 -3.30962330e-01 -1.13712180e+00 2.03000641e+00
-4.83027309e-01 2.05184147e-01 1.68822527e-01 -9.41529095e-01
1.19295251e+00 3.44779342e-01 4.25859481e-01 -8.09420466e-01
3.02648336e-01 2.85339773e-01 -3.11096847e-01 -7.47893810e-01
9.31210667e-02 -1.56645492e-01 1.73862849e-03 2.03931898e-01
4.48173910e-01 -7.25964606e-02 -2.31768414e-01 4.94371839e-02
3.32248062e-01 5.73138475e-01 2.83924937e-01 -3.01976979e-01
7.74712682e-01 -1.45683348e-01 8.16676676e-01 -3.74428146e-02
-4.97186154e-01 7.57283866e-01 2.62393057e-01 -3.80909860e-01
-1.22272587e+00 -9.46412265e-01 -3.72271061e-01 1.08559644e+00
3.72415036e-01 -2.10740656e-01 -1.10506189e+00 -8.52165937e-01
1.33339420e-01 4.62952018e-01 -8.77287328e-01 -3.21489394e-01
-3.33375931e-02 -2.40839258e-01 4.13765788e-01 6.49561048e-01
1.06532240e+00 -1.15615511e+00 3.27779830e-01 -3.06298345e-01
-3.73413444e-01 -1.32233500e+00 -8.44517946e-01 -5.52886009e-01
-5.45059681e-01 -1.05679083e+00 -3.71858567e-01 -1.06320620e+00
7.55257189e-01 3.16722728e-02 9.70395625e-01 4.08390984e-02
-2.70554870e-01 5.35156012e-01 -4.44947571e-01 -4.65037525e-01
-9.23356861e-02 -7.82731399e-02 1.97114646e-01 4.74726588e-01
6.41514003e-01 -5.47461033e-01 -7.16436803e-01 4.96479690e-01
-7.82743335e-01 1.40920520e-01 9.42457557e-01 7.70937920e-01
9.80618894e-01 -5.21927118e-01 6.99819922e-01 -9.72346544e-01
2.88960487e-01 -6.63605571e-01 -2.74889827e-01 3.94190788e-01
-5.28189600e-01 -6.00706451e-02 4.39640254e-01 -6.18567586e-01
-1.15051293e+00 4.62417662e-01 -1.42348751e-01 -7.64484465e-01
-4.92411792e-01 -2.47882176e-02 -9.19911861e-01 1.68706160e-02
3.26079249e-01 2.03444511e-01 3.90591979e-01 -4.28837389e-01
4.72354710e-01 9.15894568e-01 7.12718248e-01 -7.92566478e-01
6.16390586e-01 4.43931431e-01 -1.88781768e-01 -4.24578965e-01
-8.46075296e-01 -3.06061476e-01 -8.05040240e-01 -2.56573021e-01
1.11223483e+00 -1.11863148e+00 -7.38931239e-01 5.91884553e-01
-1.15188670e+00 -1.81864928e-02 -2.40286137e-03 4.91032064e-01
-8.03792775e-01 1.72209740e-01 -2.98589915e-01 -4.08364564e-01
-2.74354607e-01 -1.01106048e+00 1.18605220e+00 4.60218698e-01
-1.13253128e-02 -7.64019728e-01 -1.25663429e-01 5.60675681e-01
1.70589060e-01 9.28293541e-02 7.11767912e-01 -8.45266163e-01
-3.81756663e-01 1.21542096e-01 -4.96788740e-01 5.39040923e-01
2.59794056e-01 2.40364056e-02 -1.25664735e+00 -1.29078031e-01
-3.49485964e-01 -6.45729423e-01 4.95166123e-01 9.31965038e-02
1.60209346e+00 -3.62134129e-01 -3.95413607e-01 8.42610955e-01
1.33152175e+00 1.40116792e-02 9.29046869e-01 2.29254007e-01
8.04570973e-01 8.06865633e-01 1.03121293e+00 3.95909816e-01
6.58785164e-01 7.53684640e-01 4.92243171e-01 -2.46091709e-01
-3.38651568e-01 -4.84657228e-01 1.06897265e-01 5.05555749e-01
1.00253485e-01 2.44390786e-01 -5.74137986e-01 6.38063014e-01
-1.67993820e+00 -9.75892663e-01 1.09717466e-01 2.04592085e+00
1.00289834e+00 -5.11223316e-01 1.05838224e-01 -3.35754812e-01
1.00818801e+00 -2.25855514e-01 -5.74319601e-01 -3.15737784e-01
-4.27506007e-02 -6.57636523e-02 1.41137853e-01 1.83606848e-01
-1.26416624e+00 1.14659607e+00 4.99552774e+00 1.17423987e+00
-8.08722079e-01 2.70330518e-01 6.90921247e-01 2.60109335e-01
-3.51438701e-01 -2.23466866e-02 -8.94444704e-01 6.99450672e-01
3.98642480e-01 1.23859629e-01 3.33727121e-01 1.12295604e+00
-8.71184245e-02 2.24482119e-01 -1.26076591e+00 1.30541241e+00
3.41926277e-01 -9.70114350e-01 2.56848156e-01 -2.11380139e-01
5.30913830e-01 -5.45279741e-01 7.55055919e-02 2.57704943e-01
-3.11837997e-02 -9.82844234e-01 5.02666354e-01 6.94119692e-01
1.18267405e+00 -7.69389272e-01 6.65776551e-01 -3.37295011e-02
-1.28580964e+00 -8.80014338e-03 -6.21324301e-01 3.58794212e-01
-2.66947597e-01 2.48731434e-01 -5.24397314e-01 5.55179060e-01
9.29918885e-01 8.61117005e-01 -6.35423660e-01 8.20650220e-01
-1.36318251e-01 1.49580449e-01 2.56519541e-02 2.91821837e-01
-3.19104828e-03 -5.84744692e-01 -1.50377452e-02 1.01542616e+00
4.46872175e-01 1.55086160e-01 3.24757514e-03 9.41446662e-01
-3.41755211e-01 3.52912456e-01 -6.74820185e-01 1.28666028e-01
1.03025699e+00 1.61859715e+00 -1.00722283e-01 -3.28898400e-01
-6.34355724e-01 1.18583047e+00 4.46050972e-01 1.51723027e-01
-9.16015446e-01 -2.36725822e-01 9.03919935e-01 7.08115101e-02
1.40122101e-01 2.30383083e-01 -2.68973168e-02 -8.98894727e-01
4.53509390e-02 -9.15716171e-01 5.31973362e-01 -1.08709228e+00
-1.79198301e+00 5.76161563e-01 1.11281432e-01 -1.47153127e+00
6.43092990e-02 -4.16550577e-01 -5.43692589e-01 7.98181295e-01
-1.56266773e+00 -1.88939989e+00 -8.39516103e-01 1.07376778e+00
4.15657878e-01 -4.71649170e-01 9.08058703e-01 4.93649989e-01
-5.54235756e-01 1.03776574e+00 -5.63741326e-02 3.29131097e-01
1.34548855e+00 -8.05864990e-01 -1.03700362e-01 3.98599982e-01
-8.25071037e-02 3.70606005e-01 2.95929700e-01 -5.71494102e-01
-1.20136511e+00 -1.34874535e+00 6.00864172e-01 -3.50482136e-01
4.18986619e-01 -3.08996946e-01 -1.09297276e+00 9.47898746e-01
2.76261121e-01 4.86343354e-01 7.51701593e-01 1.73229482e-02
-6.62445068e-01 -6.66639209e-01 -1.39851475e+00 4.45879310e-01
1.25781202e+00 -5.41032195e-01 -4.35610533e-01 1.70089111e-01
4.74358052e-01 -2.12026790e-01 -1.07984293e+00 4.88985330e-01
7.32125401e-01 -6.19546533e-01 8.44742656e-01 -4.47885603e-01
5.61314046e-01 -3.48286659e-01 -2.15196431e-01 -1.15987670e+00
-3.30675036e-01 -1.44368634e-01 2.75138259e-01 2.01503801e+00
1.12983912e-01 -5.11638045e-01 5.95070064e-01 6.84256852e-01
-2.15102628e-01 -5.13634861e-01 -9.47404683e-01 -7.54536092e-01
1.45660520e-01 -9.81627032e-04 9.55718040e-01 1.25429547e+00
-4.10757273e-01 3.40191871e-01 -4.28550720e-01 9.77690741e-02
6.17951870e-01 1.62874430e-01 8.08374822e-01 -1.18896687e+00
1.24766715e-01 -3.13661128e-01 -5.18550575e-01 -6.11960292e-01
6.19945943e-01 -1.08848190e+00 -1.53528079e-01 -1.18028545e+00
6.40098572e-01 -7.53962278e-01 -1.47609979e-01 7.14474797e-01
-2.28567481e-01 3.21575016e-01 5.37282927e-03 2.43645698e-01
-5.64975560e-01 9.46417511e-01 1.46529472e+00 -2.15381771e-01
1.91724911e-01 -2.51760840e-01 -7.89714932e-01 6.32545829e-01
8.54075849e-01 -4.99979466e-01 -5.92935860e-01 -4.36889857e-01
-3.14282000e-01 -2.49152645e-01 5.53522289e-01 -8.58187675e-01
6.19653203e-02 -2.41488919e-01 6.24409378e-01 -3.04041266e-01
4.08281416e-01 -1.17976975e+00 2.35624403e-01 -6.19176626e-02
-3.63983095e-01 5.26812188e-02 3.12946677e-01 5.23725092e-01
-3.27375889e-01 -1.70253754e-01 9.72657919e-01 2.33042650e-02
-1.04771757e+00 8.95439267e-01 2.58197278e-01 1.20386645e-01
1.39541888e+00 -2.39130124e-01 -4.53540683e-01 -1.52421802e-01
-6.23351276e-01 2.47553200e-01 6.90208733e-01 5.88330209e-01
6.76687539e-01 -1.99653721e+00 -8.91565859e-01 2.74418384e-01
6.58212841e-01 -2.38206178e-01 5.62528312e-01 7.65330613e-01
-2.47245342e-01 -1.64135560e-01 -8.51162195e-01 -6.94170177e-01
-1.46834862e+00 7.16282129e-01 3.93747121e-01 4.27022070e-01
-2.89066255e-01 8.39441538e-01 4.34680462e-01 -6.96080029e-01
7.46176317e-02 3.32106769e-01 -2.58426160e-01 1.36872426e-01
4.24700588e-01 2.62709036e-02 -1.15714341e-01 -8.85441840e-01
-5.55008590e-01 9.45553422e-01 -1.27774432e-01 1.77748576e-01
9.83908534e-01 -2.59363681e-01 -2.65439540e-01 1.08802937e-01
1.45074749e+00 -2.98287421e-01 -1.32270145e+00 -4.38537359e-01
-2.75395095e-01 -6.67287171e-01 -2.75121927e-01 -7.49606431e-01
-1.21396291e+00 1.01613629e+00 9.81271148e-01 -5.51903605e-01
1.51078486e+00 2.72735059e-01 6.00589454e-01 1.06850257e-02
1.36725798e-01 -1.21781874e+00 2.20889196e-01 3.09416372e-02
1.09928262e+00 -1.63241303e+00 2.62474120e-02 -6.20957255e-01
-1.16801238e+00 7.83185959e-01 1.17421234e+00 -7.95748234e-02
6.73498273e-01 1.99268740e-02 1.15686402e-01 4.58592847e-02
-6.77452683e-01 -2.88816065e-01 3.26369703e-01 8.83845747e-01
4.09919828e-01 -8.90594628e-03 -3.54927123e-01 9.41616833e-01
3.49429399e-02 3.89397964e-02 1.20895125e-01 4.11584109e-01
-1.65347219e-01 -1.33250189e+00 -1.55269206e-01 2.55956680e-01
-1.79890826e-01 -8.05565193e-02 -4.56334800e-01 1.04083371e+00
4.85480607e-01 7.37175345e-01 3.93729389e-01 -4.87364262e-01
2.75540560e-01 1.42336130e-01 5.62896490e-01 -3.78658324e-01
-2.24643767e-01 -2.18717486e-01 -7.47496709e-02 -6.49403453e-01
-6.50295854e-01 -7.00447202e-01 -1.12504160e+00 -4.29187208e-01
-1.19383663e-01 -1.57716334e-01 5.44070005e-01 8.49886179e-01
6.22994602e-01 8.24229345e-02 8.12736273e-01 -5.13467550e-01
-2.23071992e-01 -1.14463198e+00 -6.55871034e-01 9.57860291e-01
-3.48713212e-02 -8.86944473e-01 -8.84770751e-02 3.24404895e-01] | [13.415716171264648, 0.7379297018051147] |
29ef8ab5-a1fe-4957-81b5-20189b2d31f3 | word-order-sensitive-embedding | null | null | https://aclanthology.org/W16-4910 | https://aclanthology.org/W16-4910.pdf | Word Order Sensitive Embedding Features/Conditional Random Field-based Chinese Grammatical Error Detection | This paper discusses how to adapt two new word embedding features to build a more efficient Chinese Grammatical Error Diagnosis (CGED) systems to assist Chinese foreign learners (CFLs) in improving their written essays. The major idea is to apply word order sensitive Word2Vec approaches including (1) structured skip-gram and (2) continuous window (CWindow) models, because they are more suitable for solving syntax-based problems. The proposed new features were evaluated on the Test of Chinese as a Foreign Language (TOCFL) learner database provided by NLP-TEA-3{\&}CGED shared task. Experimental results showed that the new features did work better than the traditional word order insensitive Word2Vec approaches. Moreover, according to the official evaluation results, our system achieved the lowest (0.1362) false positive (FA) and the highest precision rates in all three measurements. | ['Yih-Ru Wang', 'Yuan-Fu Liao', 'Chin-Kui Lin', 'Wei-Chieh Chou'] | 2016-12-01 | null | null | null | ws-2016-12 | ['grammatical-error-detection'] | ['natural-language-processing'] | [-5.46671510e-01 -4.27791536e-01 9.96021032e-02 -1.47631228e-01
-7.18740642e-01 -3.72565299e-01 4.25184667e-01 5.41787386e-01
-1.10138643e+00 1.01294315e+00 4.77966845e-01 -7.52676547e-01
2.74511762e-02 -6.67307496e-01 -1.59376413e-01 -2.82483429e-01
2.36227810e-01 2.23005727e-01 3.64368796e-01 -6.61065578e-01
7.81981766e-01 3.22949946e-01 -1.25463128e+00 1.27180174e-01
1.69127166e+00 4.68577266e-01 5.81402063e-01 8.26264501e-01
-6.45875037e-01 6.10072136e-01 -7.88494527e-01 -6.65350676e-01
-2.58481920e-01 -3.42244327e-01 -8.42976868e-01 -5.29733837e-01
2.22372547e-01 -5.90063520e-02 -1.51855320e-01 1.19633222e+00
6.34096742e-01 5.61681628e-01 6.52043641e-01 -5.49174190e-01
-1.28845334e+00 6.72530651e-01 -1.13295518e-01 5.05239367e-01
3.61932755e-01 -2.77542621e-01 1.10025430e+00 -1.44021904e+00
6.02187514e-01 1.11792195e+00 5.76793611e-01 6.95746303e-01
-3.86819303e-01 -5.90352356e-01 1.44050196e-01 8.05452704e-01
-1.19083238e+00 6.65596034e-03 3.53638828e-01 -1.79400966e-01
1.43117988e+00 2.61982977e-01 4.02323902e-01 8.65699828e-01
4.82323915e-01 8.69240165e-01 1.16096008e+00 -1.10138953e+00
-1.72630325e-01 6.41456902e-01 1.05416274e+00 7.90506721e-01
5.68147480e-01 -3.69100273e-01 -4.15980190e-01 2.06255466e-01
1.06934071e-01 -3.28160107e-01 -4.86485809e-01 5.77390552e-01
-7.12364972e-01 9.86661315e-01 -4.73509124e-03 9.79239523e-01
3.23239318e-03 -7.98079297e-02 3.72411728e-01 5.56752503e-01
4.37565118e-01 6.22513592e-01 -7.10978746e-01 -4.75147933e-01
-5.10847807e-01 2.65604615e-01 5.34060299e-01 9.42255378e-01
2.72354007e-01 2.51765430e-01 -3.48207027e-01 8.41779709e-01
4.29326445e-01 2.80886173e-01 1.22237504e+00 1.07269967e-02
6.77304029e-01 5.15920758e-01 -4.70008291e-02 -1.06794846e+00
-2.48519585e-01 -4.13400948e-01 -3.87464255e-01 -1.81644067e-01
2.54819661e-01 -3.78311276e-01 -5.40409923e-01 1.16456676e+00
-1.71324268e-01 -2.06275016e-01 1.27924562e-01 6.41873360e-01
1.13845038e+00 8.77811968e-01 2.93982804e-01 -1.86311677e-01
1.33262801e+00 -1.03190649e+00 -1.32214510e+00 1.24577254e-01
1.36868632e+00 -1.15349984e+00 1.29315174e+00 6.17443383e-01
-8.36877882e-01 -7.79085398e-01 -1.13989413e+00 -2.59745747e-01
-8.74787688e-01 2.89269626e-01 4.02330756e-01 1.11507511e+00
-8.60335052e-01 5.23196697e-01 -5.02691865e-01 -2.43438557e-01
-2.02474706e-02 -3.40039209e-02 -4.39360648e-01 -2.13816077e-01
-1.45402420e+00 1.19476318e+00 5.58654368e-01 -3.08160931e-01
-2.82917231e-01 -6.50508821e-01 -8.11306953e-01 1.88545346e-01
-1.93793669e-01 2.37349615e-01 8.29030097e-01 -4.25590068e-01
-1.31056368e+00 4.72154051e-01 -3.23573112e-01 -2.07699016e-01
1.63180143e-01 -6.20657444e-01 -7.21719384e-01 -1.79621041e-01
-8.81561562e-02 -5.51635250e-02 4.63852808e-02 -8.38052392e-01
-6.36176050e-01 -3.55597734e-01 -3.74364287e-01 2.87613720e-01
-9.41068649e-01 3.98123771e-01 -3.87301743e-01 -8.16883206e-01
-1.26700938e-01 -4.90199029e-01 5.58659397e-02 -4.83053982e-01
-1.47553638e-01 -1.04438794e+00 7.51389205e-01 -1.31100476e+00
2.00418043e+00 -1.84012508e+00 -2.16122270e-02 3.00369672e-02
-1.07789949e-01 1.02090359e+00 -1.97854131e-01 5.28227329e-01
1.78616066e-02 5.72162628e-01 1.75490767e-01 -1.86608195e-01
-7.10585341e-02 -7.74276033e-02 1.72790200e-01 1.68043092e-01
9.67867374e-02 7.47134089e-01 -1.01251733e+00 -5.54184079e-01
1.06175885e-01 2.68910825e-01 -5.04654408e-01 3.66449922e-01
1.69499055e-01 -9.00032148e-02 -4.36678141e-01 5.09477556e-01
7.16414988e-01 4.85609829e-01 9.25635174e-02 4.05077279e-01
-3.59274656e-01 4.32726741e-01 -1.24726129e+00 1.25094151e+00
-5.90350688e-01 7.10878372e-01 -5.58111906e-01 -5.87923527e-01
1.36010146e+00 2.96334356e-01 -1.29506439e-01 -7.97734976e-01
2.32569590e-01 3.57546419e-01 6.11380897e-02 -8.44042182e-01
9.16678905e-01 4.02883478e-02 2.82465182e-02 3.51423204e-01
4.67946619e-01 6.04390875e-02 2.59621620e-01 2.18628556e-01
8.99781585e-01 -4.80367281e-02 4.75953579e-01 -6.24940813e-01
1.04397285e+00 -8.24302435e-02 5.25210619e-01 4.32707906e-01
-3.37125003e-01 3.18425804e-01 3.17525387e-01 -2.06891760e-01
-4.76020098e-01 -7.62055755e-01 -2.76461840e-01 1.00992608e+00
-1.91004932e-01 -7.43147492e-01 -5.63697457e-01 -1.06608212e+00
-1.71452254e-01 1.21880138e+00 -3.05347055e-01 9.06987190e-02
-6.41139627e-01 -6.60117567e-01 6.60571933e-01 4.69704807e-01
4.81326073e-01 -9.18605268e-01 -1.14847310e-01 5.05210519e-01
-6.45239875e-02 -7.07409859e-01 -5.49319208e-01 1.59456164e-01
-6.20420635e-01 -1.13785207e+00 -7.35736430e-01 -1.23779142e+00
5.25026679e-01 1.75975442e-01 7.69859910e-01 6.97879255e-01
-3.26252282e-02 1.11132406e-01 -1.21530306e+00 -7.32443571e-01
-1.86149687e-01 6.17194548e-02 2.96843886e-01 -3.96845758e-01
9.97399747e-01 1.62098378e-01 2.62395918e-01 -3.29032868e-01
-6.16740406e-01 -4.79267716e-01 1.13832898e-01 1.22446024e+00
2.59270281e-01 1.06529901e-02 7.22094595e-01 -8.88776779e-01
1.37868655e+00 -2.90718377e-01 -4.25857723e-01 7.95805216e-01
-1.13549471e+00 5.58351986e-02 8.71197283e-01 -1.68659672e-01
-1.32582247e+00 -6.29094124e-01 -5.81525028e-01 1.34109721e-01
4.79038730e-02 6.19760275e-01 -1.49937391e-01 -2.16672048e-01
4.99264747e-01 2.47135222e-01 -4.11952615e-01 -8.79425943e-01
1.03617556e-01 1.06332219e+00 7.00466186e-02 -5.65188408e-01
3.85244608e-01 -7.30688870e-01 -4.27561402e-01 -1.03885603e+00
-4.51069534e-01 -6.05873764e-01 -8.38706017e-01 -1.01905406e-01
1.02016735e+00 -7.77196348e-01 -5.56061208e-01 5.39275944e-01
-1.37687624e+00 1.15091085e-01 2.23735616e-01 9.46074545e-01
3.07429492e-01 7.02476561e-01 -7.68433034e-01 -7.70843029e-01
-5.32158017e-01 -1.12618768e+00 1.63557574e-01 5.43906689e-01
-4.84281220e-02 -1.27392209e+00 2.45426118e-01 1.99581459e-01
4.49978441e-01 -3.07465762e-01 1.31800103e+00 -1.12206745e+00
5.83733097e-02 -7.39483461e-02 -2.01157972e-01 7.38176048e-01
-1.92039967e-01 4.63374667e-02 -7.28749275e-01 -1.34388238e-01
-2.42930830e-01 -3.07750523e-01 8.99826825e-01 -1.95113882e-01
1.26326764e+00 -2.36375615e-01 1.53380156e-01 5.73202908e-01
1.73559165e+00 4.54601943e-01 5.67612529e-01 4.71194565e-01
6.62220657e-01 1.91378549e-01 7.11447477e-01 3.57027203e-01
2.56762594e-01 5.30345917e-01 -5.62394038e-02 4.79190528e-01
-2.70858318e-01 -4.00927514e-01 6.24588668e-01 1.79760337e+00
-1.45577267e-01 -6.29540324e-01 -9.42421317e-01 8.13574374e-01
-1.32587636e+00 -7.33996034e-01 -8.83378148e-01 1.95754743e+00
1.01718819e+00 -1.60807922e-01 -5.29366314e-01 2.12584049e-01
3.70627135e-01 4.20487747e-02 3.83653224e-01 -1.36125386e+00
-2.71025479e-01 1.07226348e+00 3.72367561e-01 7.64491856e-01
-8.53854835e-01 1.37244034e+00 5.20033598e+00 1.09963405e+00
-7.46924579e-01 4.01010513e-01 1.10337473e-01 5.56664884e-01
-6.39722884e-01 -2.55129278e-01 -1.34360623e+00 4.84210730e-01
1.08422971e+00 -3.50215554e-01 -6.45779222e-02 5.71979821e-01
-1.50153697e-01 -7.44716749e-02 -2.95469701e-01 7.04785764e-01
4.03600127e-01 -1.19879186e+00 1.38053164e-01 -3.12195480e-01
7.78454483e-01 -2.26260930e-01 -2.77749598e-01 8.32534730e-01
2.86523342e-01 -1.07286942e+00 2.63720065e-01 4.68706548e-01
7.48113930e-01 -1.03078413e+00 1.34229565e+00 2.75642127e-01
-9.28445339e-01 8.56950879e-02 -7.49226034e-01 -2.02468663e-01
-1.22732259e-01 2.80871361e-01 -6.85132027e-01 6.56029880e-01
6.20981574e-01 5.82320869e-01 -1.01594234e+00 9.06794786e-01
-6.23366535e-01 9.69576001e-01 1.57460213e-01 -9.70283389e-01
4.12089288e-01 -2.25844502e-01 3.63593310e-01 1.56640089e+00
6.44921064e-01 2.09258243e-01 -4.63264346e-01 4.00331676e-01
-1.56094000e-01 8.55132818e-01 -3.49075019e-01 -5.04236296e-03
5.27106762e-01 9.46035445e-01 -1.77940115e-01 -1.83904037e-01
-5.91300368e-01 1.11057305e+00 6.07580543e-01 3.24981809e-02
-5.36730349e-01 -1.26778936e+00 5.21129131e-01 -3.94848019e-01
2.72240937e-01 -3.52331191e-01 -3.76144797e-01 -1.39389384e+00
-6.43086284e-02 -9.03483212e-01 2.29631975e-01 -4.86843407e-01
-1.24965000e+00 6.38931692e-01 -4.25936580e-01 -1.02191532e+00
2.62448221e-01 -1.13643789e+00 -1.00081229e+00 1.26455069e+00
-1.45226336e+00 -7.63238430e-01 -1.91948656e-02 5.07061243e-01
8.45762134e-01 -6.95384264e-01 1.24870229e+00 4.29763883e-01
-8.97153080e-01 1.35001779e+00 5.64367354e-01 3.60117644e-01
8.13325465e-01 -1.52024269e+00 -9.59925726e-02 1.29760540e+00
3.13650101e-01 6.47572875e-01 3.55607361e-01 -7.75129318e-01
-1.16410816e+00 -9.76619124e-01 2.13849568e+00 -2.85548478e-01
4.58544880e-01 1.16435001e-02 -1.14055419e+00 4.78195012e-01
4.22628194e-01 -4.28396463e-01 7.51134992e-01 2.37701327e-01
5.84002845e-02 1.41462028e-01 -1.02617061e+00 5.32219470e-01
3.10051352e-01 -1.88270137e-01 -1.13373351e+00 3.15951675e-01
7.71746695e-01 -1.49550259e-01 -9.79499221e-01 3.52398492e-02
3.41517299e-01 -6.88552499e-01 6.00625277e-01 -9.77779031e-01
5.06263137e-01 -3.91538367e-02 -1.17817007e-01 -1.65824568e+00
-4.80035156e-01 -1.29749939e-01 3.00141033e-02 1.42878795e+00
5.67481518e-01 -4.51626271e-01 2.12012529e-01 4.86157149e-01
-4.93607998e-01 -1.06728470e+00 -9.47339654e-01 -7.47727215e-01
6.20153368e-01 -4.95249480e-01 6.00363791e-01 1.24630260e+00
3.08194518e-01 3.92278619e-02 -1.88898668e-02 -2.13164657e-01
-1.95517568e-04 -7.89350569e-01 2.93786108e-01 -1.03919411e+00
9.98075828e-02 -3.70671451e-01 -4.65390682e-01 -8.15244257e-01
2.99252003e-01 -1.07446027e+00 -1.97150514e-01 -1.27685428e+00
-5.56244031e-02 -4.49679196e-01 -6.30254090e-01 2.91921735e-01
-9.14506197e-01 -1.64517820e-01 3.38847846e-01 -4.69237745e-01
-2.79132664e-01 8.09935093e-01 1.10538924e+00 1.76140144e-01
-5.22674173e-02 -1.28611043e-01 -6.73447430e-01 4.40789133e-01
7.75971293e-01 -5.14096677e-01 3.94831896e-02 -8.07865441e-01
3.09293941e-02 -2.39870369e-01 -4.04118180e-01 -9.55451846e-01
2.39876792e-01 -1.71131060e-01 5.36400557e-01 -5.79437017e-01
-3.28563839e-01 -6.12863421e-01 -8.28153014e-01 5.57922184e-01
-3.13443959e-01 6.20317996e-01 3.43709975e-01 1.92090139e-01
-3.97787541e-01 -1.05219877e+00 5.72206914e-01 -1.48617521e-01
-1.12122953e+00 -7.07713813e-02 -6.08885825e-01 2.74023712e-01
9.89323020e-01 8.83397833e-02 -3.03810000e-01 -1.40259713e-01
-2.20934689e-01 3.48825961e-01 -9.95336324e-02 6.53445721e-01
9.29493546e-01 -1.27975535e+00 -8.40385556e-01 3.81372660e-01
3.00312519e-01 -5.21010756e-01 1.09540552e-01 5.62266171e-01
-1.11134303e+00 8.17862451e-01 -2.85084754e-01 2.82885879e-01
-1.49021733e+00 3.44062597e-02 9.18915123e-02 -6.93201303e-01
-1.52686283e-01 1.37669003e+00 -6.22980237e-01 -7.30301976e-01
3.73111814e-01 -1.21747561e-01 -9.41892445e-01 1.44075781e-01
8.16150188e-01 7.60934353e-01 5.38528025e-01 -5.83634734e-01
-1.75492585e-01 4.36590403e-01 -2.76284069e-01 -5.69847822e-02
1.39136910e+00 1.00361414e-01 -1.00782283e-01 4.14352000e-01
1.18865216e+00 5.24440229e-01 -1.15083925e-01 -2.11930484e-01
2.72253036e-01 -4.34455037e-01 2.63049722e-01 -9.01011229e-01
-6.66377127e-01 1.26167738e+00 6.16579473e-01 -3.29280287e-01
7.70061731e-01 -7.57828653e-01 1.05469573e+00 5.93136132e-01
3.41183573e-01 -1.49854255e+00 -3.71576786e-01 1.16675425e+00
6.40305400e-01 -1.16355157e+00 4.28810418e-02 4.15528081e-02
-7.52308667e-01 1.74146748e+00 9.89437640e-01 -6.90706000e-02
8.30971897e-01 -2.88568825e-01 6.60887510e-02 1.28459707e-01
-5.54855287e-01 -1.88828439e-01 6.04340851e-01 4.78581071e-01
1.12302196e+00 2.86766559e-01 -1.40539598e+00 1.26438892e+00
-9.23415646e-02 -2.74136454e-01 7.90637076e-01 6.91232860e-01
-7.02607632e-01 -1.59902942e+00 -1.39013216e-01 6.31987095e-01
-7.04547167e-01 -4.17983145e-01 -3.12229514e-01 1.00400829e+00
4.02483910e-01 1.13423538e+00 -1.20395869e-01 -6.44346058e-01
3.11539918e-01 5.54789305e-01 2.24458352e-01 -8.08522463e-01
-1.08789384e+00 -4.47543293e-01 1.98235773e-02 -2.57524639e-01
7.90062025e-02 -4.99222130e-01 -1.20031416e+00 -3.05098772e-01
-8.06221366e-01 5.17590880e-01 6.35194957e-01 1.00144362e+00
-9.41787437e-02 6.22048318e-01 5.98248065e-01 1.28777325e-01
-6.62746251e-01 -1.50482082e+00 -8.01154375e-01 3.75289887e-01
-2.11456090e-01 -3.44712675e-01 -3.59056562e-01 -3.11718136e-01] | [11.045891761779785, 10.79604721069336] |
371c111c-adc4-4f85-868b-86a4acd4b9b5 | learnable-graph-convolutional-network-and | 2211.09155 | null | https://arxiv.org/abs/2211.09155v1 | https://arxiv.org/pdf/2211.09155v1.pdf | Learnable Graph Convolutional Network and Feature Fusion for Multi-view Learning | In practical applications, multi-view data depicting objectives from assorted perspectives can facilitate the accuracy increase of learning algorithms. However, given multi-view data, there is limited work for learning discriminative node relationships and graph information simultaneously via graph convolutional network that has drawn the attention from considerable researchers in recent years. Most of existing methods only consider the weighted sum of adjacency matrices, yet a joint neural network of both feature and graph fusion is still under-explored. To cope with these issues, this paper proposes a joint deep learning framework called Learnable Graph Convolutional Network and Feature Fusion (LGCN-FF), consisting of two stages: feature fusion network and learnable graph convolutional network. The former aims to learn an underlying feature representation from heterogeneous views, while the latter explores a more discriminative graph fusion via learnable weights and a parametric activation function dubbed Differentiable Shrinkage Activation (DSA) function. The proposed LGCN-FF is validated to be superior to various state-of-the-art methods in multi-view semi-supervised classification. | ['Shiping Wang', 'Claudia Plant', 'Wenzhong Guo', 'Jie Yao', 'Lele Fu', 'Zhaoliang Chen'] | 2022-11-16 | null | null | null | null | ['multi-view-learning'] | ['computer-vision'] | [ 1.51886910e-01 2.28304446e-01 -2.19334468e-01 -5.39613485e-01
-4.84298021e-01 -2.75651544e-01 5.78723133e-01 1.14765413e-01
8.34823996e-02 5.39166152e-01 3.29823017e-01 2.03630120e-01
-4.19983566e-01 -9.19444799e-01 -6.03443027e-01 -7.12229729e-01
-9.17280316e-02 2.26002365e-01 -1.39648840e-01 -9.97015908e-02
-1.00304514e-01 5.20327151e-01 -1.34745693e+00 1.79337531e-01
8.21749926e-01 1.09468305e+00 -1.35142684e-01 4.44640964e-01
6.10272288e-02 8.98202121e-01 -6.50670901e-02 -5.02948642e-01
3.43564510e-01 -2.55379289e-01 -7.00352728e-01 4.75803286e-01
4.97364283e-01 -1.52022019e-01 -6.16136253e-01 1.17942858e+00
5.00839710e-01 1.67534798e-01 4.78555620e-01 -1.47229135e+00
-1.09386992e+00 7.87870944e-01 -7.10733414e-01 1.42203197e-01
2.78657943e-01 -1.19238615e-01 1.30466151e+00 -9.60978568e-01
5.79066515e-01 1.28380513e+00 5.78706920e-01 4.21191268e-02
-1.00135839e+00 -4.50789243e-01 4.52808321e-01 3.17279905e-01
-1.39392781e+00 -1.34185076e-01 1.33970666e+00 -3.81450117e-01
8.50217104e-01 5.80014810e-02 8.89732897e-01 8.77846599e-01
3.71995836e-01 8.47317219e-01 1.06492901e+00 -1.02654278e-01
-2.06517875e-01 -8.91161636e-02 7.60669783e-02 1.20249796e+00
4.28188413e-01 -9.32569802e-02 -5.05370319e-01 -1.05295949e-01
7.86481082e-01 3.94457489e-01 -3.93661469e-01 -8.41049552e-01
-1.07225621e+00 9.64887559e-01 7.99228132e-01 2.55654544e-01
-5.29333591e-01 1.12737104e-01 4.15578097e-01 3.57340723e-01
8.21455538e-01 1.37168244e-02 -1.52903438e-01 3.98296326e-01
-6.82047904e-01 7.26871286e-03 6.22747600e-01 7.82908380e-01
1.01287043e+00 3.41294736e-01 9.17371958e-02 7.22305357e-01
5.86643159e-01 2.59153694e-01 1.49683684e-01 -5.43970764e-01
7.08690464e-01 1.28182280e+00 -5.03764212e-01 -1.38687551e+00
-6.90106153e-01 -7.72247910e-01 -1.21474946e+00 9.49574783e-02
-4.09410298e-02 -2.53454864e-01 -7.64872611e-01 1.60985339e+00
4.27420378e-01 1.38031870e-01 8.48878771e-02 9.75287557e-01
1.32579005e+00 4.33175117e-01 -2.33025074e-01 -6.19607940e-02
1.09445190e+00 -1.18676817e+00 -7.63622582e-01 -1.56375259e-01
4.94644374e-01 -5.26964605e-01 5.05094588e-01 1.76299989e-01
-1.02789247e+00 -6.92299783e-01 -1.29146802e+00 3.69233941e-03
-4.86056209e-01 2.41123363e-01 9.16335285e-01 2.11383551e-01
-1.15649807e+00 4.55766946e-01 -7.86105752e-01 -3.62948209e-01
6.09110653e-01 4.30440158e-01 -9.32393491e-01 -3.56362104e-01
-1.18822658e+00 6.83613002e-01 4.35642987e-01 3.99578124e-01
-8.15689087e-01 -4.41251695e-01 -1.19651258e+00 2.07123280e-01
5.98764300e-01 -1.00552118e+00 4.78193432e-01 -1.12151265e+00
-1.30304182e+00 7.69285500e-01 3.82596493e-01 -8.94989520e-02
3.02938312e-01 -1.42079353e-01 -5.84892154e-01 1.84080735e-01
-8.26851726e-02 2.89855450e-01 1.04358029e+00 -1.42710614e+00
-2.65712351e-01 -7.90574849e-01 2.81163573e-01 6.06975615e-01
-4.57546532e-01 -2.96358526e-01 -3.71016443e-01 -7.47680724e-01
2.28292957e-01 -7.56924093e-01 -1.39321044e-01 8.41287449e-02
-5.04728258e-01 -3.24274510e-01 1.04990864e+00 -7.81649888e-01
1.16113234e+00 -1.92858970e+00 6.86073184e-01 1.53970107e-01
7.43326485e-01 1.71650425e-01 -1.95824653e-01 6.90424919e-01
-3.38755906e-01 -1.79992706e-01 -1.76149443e-01 -3.18863481e-01
-2.29956791e-01 1.76155761e-01 2.81674922e-01 7.36990154e-01
2.54233748e-01 1.03495967e+00 -8.94966841e-01 -3.85681182e-01
3.16691905e-01 7.22452104e-01 -3.66779625e-01 3.98001283e-01
2.69534916e-01 2.73075819e-01 -7.15876997e-01 8.00737858e-01
8.08632970e-01 -6.30103230e-01 2.35250160e-01 -6.71359181e-01
1.70316070e-01 -4.35622483e-01 -1.31031954e+00 1.91258776e+00
-2.25969970e-01 2.78304070e-01 2.86037803e-01 -1.41269803e+00
1.03990352e+00 1.55940920e-01 7.01043308e-01 -2.38844633e-01
2.78320104e-01 -5.17541775e-03 -3.35191600e-02 -4.78991508e-01
2.96983272e-01 -6.31769896e-02 2.04542041e-01 2.70327568e-01
5.84869802e-01 8.71957093e-02 -7.66850412e-02 2.91951120e-01
9.92742300e-01 2.26725504e-01 6.47828579e-01 -1.76369295e-01
8.57240736e-01 -4.58978564e-01 4.23268348e-01 1.75367966e-01
-5.09222709e-02 5.25511563e-01 4.28910255e-01 -6.47075176e-01
-6.42744184e-01 -9.05436754e-01 3.47653866e-01 8.38070810e-01
3.19119275e-01 -4.56250191e-01 -5.31839252e-01 -1.21538556e+00
1.58512801e-01 8.21970105e-02 -7.59763241e-01 -3.51038575e-01
-4.12180096e-01 -4.82972056e-01 6.54890612e-02 6.64075851e-01
8.75583231e-01 -7.82995999e-01 -1.14721507e-02 -3.57153751e-02
2.28466019e-01 -1.11907530e+00 -5.58011949e-01 3.12356483e-02
-9.04438019e-01 -1.39438403e+00 -6.12403929e-01 -7.38681316e-01
9.17724729e-01 6.23242795e-01 9.96777117e-01 2.23068133e-01
-1.10345870e-01 7.40978777e-01 -4.38717365e-01 -7.16696903e-02
-1.63231224e-01 2.56290138e-01 -1.43135399e-01 5.06125271e-01
4.61249016e-02 -7.68251836e-01 -5.66099703e-01 5.19838817e-02
-9.81832206e-01 2.49792710e-01 6.97544456e-01 8.90758276e-01
5.65071285e-01 -7.40114152e-02 5.98853827e-01 -1.05231297e+00
6.63208604e-01 -5.60588598e-01 -4.03963715e-01 6.11703277e-01
-6.55371308e-01 1.26316771e-01 7.53835320e-01 9.19313356e-02
-8.65966380e-01 1.05079114e-01 -4.26850602e-04 -8.19537461e-01
-1.94413196e-02 9.31694984e-01 -3.58755320e-01 -4.97270763e-01
2.19276592e-01 2.53226012e-01 2.51323044e-01 -2.05303252e-01
6.63507760e-01 4.25838530e-01 1.78942204e-01 -2.17451200e-01
8.12846422e-01 4.03678179e-01 2.14474976e-01 -5.69353342e-01
-9.58579719e-01 -5.18773437e-01 -7.98895299e-01 -5.14123797e-01
9.20849144e-01 -1.15585983e+00 -4.74271029e-01 5.16783416e-01
-8.49943459e-01 2.06909612e-01 -1.20004572e-01 6.51793718e-01
-5.09109139e-01 5.44125676e-01 -4.11580890e-01 -3.97390217e-01
-5.53141296e-01 -1.22030973e+00 1.16569161e+00 1.51059911e-01
3.51194918e-01 -1.37334776e+00 9.28420387e-03 4.91668999e-01
3.67321104e-01 6.41123235e-01 7.55977809e-01 -7.90797353e-01
-5.67391336e-01 -3.75755489e-01 -4.12438124e-01 5.36016881e-01
3.53965253e-01 -1.30409747e-01 -9.15684998e-01 -6.52349114e-01
-7.52211586e-02 -4.20435727e-01 9.32370603e-01 3.60459834e-01
1.06980217e+00 -2.63873607e-01 -3.30191016e-01 8.31571460e-01
1.59213483e+00 -1.38956189e-01 2.87351429e-01 -4.47769463e-02
1.29328525e+00 4.23619270e-01 1.57931030e-01 4.25834090e-01
6.31216168e-01 3.78276885e-01 8.71588588e-01 -2.83291519e-01
-1.88441217e-01 -1.81942314e-01 2.75579020e-02 1.33809924e+00
-4.02949929e-01 -4.46738720e-01 -7.17682898e-01 3.21431667e-01
-2.14736629e+00 -7.90904284e-01 1.56046718e-01 1.66343582e+00
-1.09510287e-03 2.35302355e-02 -9.27815028e-03 1.30163077e-02
8.89357090e-01 7.86183655e-01 -6.71135247e-01 -5.10093458e-02
-3.48222293e-02 -2.26499408e-01 2.99824417e-01 2.99566269e-01
-1.26239800e+00 5.51164627e-01 5.33704138e+00 6.34184778e-01
-1.11283422e+00 6.60866201e-02 5.53698957e-01 4.34928060e-01
-5.78982413e-01 9.86998668e-04 -3.85861665e-01 -4.63846661e-02
4.75437284e-01 4.06488813e-02 5.14669120e-01 9.17211115e-01
-2.78786391e-01 4.65073436e-01 -9.58367109e-01 1.15876377e+00
5.80333412e-01 -1.32253838e+00 3.93044353e-01 1.25355303e-01
7.76461363e-01 7.95040429e-02 -4.10378762e-02 3.29789579e-01
3.19897056e-01 -9.43530977e-01 4.01570976e-01 7.47602999e-01
5.63171864e-01 -7.31497645e-01 8.44782948e-01 1.60425290e-01
-1.67675126e+00 -1.22818686e-01 -1.12687677e-01 2.92487890e-02
1.02032974e-01 5.19575536e-01 -4.18906778e-01 1.46720386e+00
3.42364162e-01 1.50200927e+00 -6.83093190e-01 6.58202112e-01
-4.60255779e-02 3.50067526e-01 1.50568839e-02 2.74856329e-01
4.21729654e-01 -3.87466043e-01 5.08906603e-01 7.15225577e-01
3.43587786e-01 -4.36580740e-02 5.47931850e-01 7.21471488e-01
-3.56409371e-01 1.69116899e-01 -8.33233416e-01 -2.46414766e-01
-1.98713690e-03 1.86237240e+00 -7.38765240e-01 -1.25588819e-01
-9.01674271e-01 7.55408645e-01 8.61985862e-01 3.42315286e-01
-6.44127309e-01 -4.09828126e-02 2.23707348e-01 -7.47077167e-02
2.05996796e-01 -3.21813732e-01 1.27755836e-01 -1.55425572e+00
-8.36149454e-02 -7.49547422e-01 7.20627069e-01 -6.92107141e-01
-1.58682644e+00 7.69907892e-01 1.09390445e-01 -1.37017965e+00
3.21486294e-02 -6.24784052e-01 -5.78080714e-01 7.07119465e-01
-1.59659350e+00 -1.89134526e+00 -7.40046442e-01 6.75977588e-01
5.27511716e-01 -5.26587009e-01 6.50806963e-01 2.60761827e-01
-5.98832011e-01 4.06418979e-01 -8.26513469e-02 6.14097007e-02
4.46448088e-01 -1.26911271e+00 2.67465472e-01 7.27475762e-01
1.48319691e-01 3.33305836e-01 1.64923415e-01 -6.19416416e-01
-1.78120339e+00 -1.22105289e+00 3.47256124e-01 -5.68894409e-02
6.10087454e-01 -1.76149741e-01 -8.47530067e-01 8.39926600e-01
4.69533980e-01 7.68232524e-01 7.07283735e-01 -1.74660049e-02
-3.46548408e-01 -2.93123841e-01 -9.90147769e-01 1.79211006e-01
1.12712276e+00 -6.37190521e-01 -1.77840784e-01 2.41926238e-01
5.68003237e-01 -4.22892869e-01 -1.31216025e+00 5.59204876e-01
4.19783592e-01 -9.38194871e-01 9.13638175e-01 -6.62041306e-01
3.13291430e-01 -2.39206895e-01 -3.45389813e-01 -1.63134897e+00
-5.66936433e-01 -3.01837027e-01 -4.08804089e-01 1.13665235e+00
7.32018724e-02 -7.13093758e-01 6.41675472e-01 1.18558899e-01
-3.02983165e-01 -1.20704949e+00 -7.65470445e-01 -2.92653471e-01
-3.25653434e-01 5.97797781e-02 5.92264533e-01 1.25371957e+00
-1.86563432e-01 5.50552964e-01 -4.59549159e-01 2.00427592e-01
6.86530232e-01 3.75759393e-01 6.44124448e-01 -1.33308756e+00
-1.60224348e-01 -3.45824182e-01 -8.45492780e-01 -5.72085321e-01
3.61553162e-01 -1.28059077e+00 -5.25399268e-01 -1.96301925e+00
4.16678190e-01 -4.85890545e-02 -5.05281627e-01 4.42775071e-01
-3.80163908e-01 -7.03794286e-02 1.80306152e-01 -1.61656388e-03
-8.57934117e-01 9.91319537e-01 1.60343671e+00 -3.07836473e-01
1.92171767e-01 -8.79712850e-02 -7.87026227e-01 7.16403067e-01
4.57729876e-01 -7.50861615e-02 -7.05357432e-01 -3.73625755e-01
2.58590639e-01 4.10890907e-01 3.55632842e-01 -9.16499794e-01
3.30188155e-01 3.74691896e-02 3.94204080e-01 -5.42775095e-01
2.94339716e-01 -1.00694680e+00 3.06549013e-01 2.68095642e-01
-2.06684873e-01 3.99652779e-01 -1.38069630e-01 1.02971065e+00
-5.24768353e-01 2.36150607e-01 4.53737229e-01 -4.01048064e-01
-7.75987387e-01 7.83256710e-01 2.30352387e-01 -5.77634014e-02
1.02290964e+00 -2.51663864e-01 -3.44206959e-01 -3.94260883e-01
-8.48855257e-01 2.94297814e-01 1.14849292e-01 5.54700792e-01
9.17737961e-01 -1.74823523e+00 -7.20036745e-01 3.47862870e-01
2.98814654e-01 -1.88310128e-02 5.99056542e-01 8.49070013e-01
-4.36109632e-01 9.11298916e-02 -2.98774064e-01 -5.51300764e-01
-1.09674573e+00 6.33819222e-01 3.67955685e-01 -5.73144138e-01
-5.72479546e-01 8.29072952e-01 2.10938707e-01 -7.23794103e-01
8.08517542e-03 -1.20968372e-01 -5.58583438e-01 1.52520940e-01
7.21377879e-02 4.31266427e-01 8.88186842e-02 -9.39669788e-01
-2.97154278e-01 7.28020132e-01 -9.21288058e-02 5.50108433e-01
1.58177423e+00 -1.71809644e-01 -2.06660211e-01 3.78204376e-01
1.44771314e+00 -4.13548619e-01 -1.26982081e+00 -4.76880848e-01
-3.68884057e-01 -2.93536037e-01 3.12911481e-01 -3.47508937e-01
-1.59636068e+00 7.86749840e-01 5.21457314e-01 3.74341190e-01
1.09042537e+00 -1.44661814e-01 6.27306223e-01 3.29187393e-01
2.25880340e-01 -7.98849404e-01 2.61640072e-01 2.14785457e-01
1.15092564e+00 -1.56191289e+00 3.77503306e-01 -6.25234187e-01
-6.08143687e-01 1.39437652e+00 7.53318667e-01 -4.31306809e-01
1.06579876e+00 -5.54069467e-02 -1.92801088e-01 -6.89140201e-01
-5.52492738e-01 -1.33870050e-01 7.86540985e-01 3.66938591e-01
4.22496498e-01 5.42804226e-02 -1.60468772e-01 5.62311172e-01
2.44952917e-01 -1.95828006e-01 1.22658208e-01 8.61040294e-01
-1.95444882e-01 -8.25309277e-01 7.55358264e-02 7.71082640e-01
-3.04545432e-01 1.33334041e-01 -3.56070071e-01 9.39164698e-01
7.71861970e-02 6.17361903e-01 -1.89246178e-01 -6.55130625e-01
2.60965198e-01 -1.99855179e-01 4.77219820e-01 -6.08880877e-01
-5.90980589e-01 5.02921902e-02 7.36935362e-02 -5.99622548e-01
-1.01999784e+00 -4.30564404e-01 -8.31298411e-01 -9.89605263e-02
-6.64683759e-01 -4.97405343e-02 2.82385290e-01 9.80644464e-01
3.47148478e-01 8.70641351e-01 7.50854850e-01 -1.03645372e+00
-4.21367913e-01 -8.41288507e-01 -9.48791385e-01 4.84695256e-01
2.53202111e-01 -8.88861358e-01 -2.08179727e-01 -2.16618583e-01] | [7.469359874725342, 6.054861545562744] |
e6ddf494-bb78-4f17-a39e-4d93018009f5 | multiclass-spectral-feature-scaling-method | 1910.07174 | null | https://arxiv.org/abs/1910.07174v1 | https://arxiv.org/pdf/1910.07174v1.pdf | Multiclass spectral feature scaling method for dimensionality reduction | Irregular features disrupt the desired classification. In this paper, we consider aggressively modifying scales of features in the original space according to the label information to form well-separated clusters in low-dimensional space. The proposed method exploits spectral clustering to derive scaling factors that are used to modify the features. Specifically, we reformulate the Laplacian eigenproblem of the spectral clustering as an eigenproblem of a linear matrix pencil whose eigenvector has the scaling factors. Numerical experiments show that the proposed method outperforms well-established supervised dimensionality reduction methods for toy problems with more samples than features and real-world problems with more features than samples. | ['Akira Imakura', 'Xiucai Ye', 'Tetsuya Sakurai', 'Momo Matsuda', 'Keiichi Morikuni'] | 2019-10-16 | null | null | null | null | ['supervised-dimensionality-reduction'] | ['computer-vision'] | [ 5.80605939e-02 2.27976013e-02 -1.42790809e-01 -4.09292549e-01
-7.31715798e-01 -8.40336800e-01 4.06120300e-01 -2.67450988e-01
-3.85663033e-01 5.20660937e-01 8.19700137e-02 9.19504687e-02
-7.06979394e-01 -1.88370645e-01 -1.45497382e-01 -1.12205970e+00
-1.80287659e-01 3.76537710e-01 -1.79961115e-01 1.59893468e-01
5.31435192e-01 7.56564319e-01 -1.41160452e+00 6.30614981e-02
8.68243158e-01 7.75309265e-01 -3.13014269e-01 6.11736715e-01
1.11462295e-01 -1.54537946e-01 -3.87346059e-01 3.24273914e-01
8.58682454e-01 -3.07125986e-01 -6.84740603e-01 8.93274784e-01
4.49901104e-01 3.21823597e-01 -3.50436121e-01 1.34094012e+00
2.19557673e-01 5.02187312e-01 1.21489334e+00 -1.52595365e+00
-7.80708194e-01 2.63575554e-01 -1.08257031e+00 2.01978236e-01
1.06112570e-01 -2.53806144e-01 8.99915099e-01 -1.06443644e+00
3.63875329e-01 1.40933919e+00 5.85206270e-01 4.43890274e-01
-1.80085218e+00 -4.49665219e-01 -2.23314706e-02 2.34844416e-01
-1.66923022e+00 -4.09720957e-01 9.71207857e-01 -7.21393943e-01
7.37912953e-01 3.69834453e-01 4.23734456e-01 4.86888319e-01
6.60771355e-02 3.46420974e-01 9.73276436e-01 -5.35687268e-01
4.61712152e-01 1.03708647e-01 5.51181614e-01 5.47364235e-01
4.79540288e-01 -8.77402648e-02 -1.17243208e-01 -7.30161965e-01
4.20274585e-01 9.85963866e-02 -1.36978090e-01 -1.05216801e+00
-1.28264010e+00 1.14611471e+00 7.42574036e-02 3.86060148e-01
-2.13138312e-01 -2.18601953e-02 1.68402717e-01 3.84076118e-01
3.81211430e-01 7.24344015e-01 -3.73834342e-01 7.60428160e-02
-5.76267779e-01 -8.56877863e-02 5.61587870e-01 8.07651639e-01
1.00288904e+00 -7.99429044e-02 2.82092720e-01 8.50617349e-01
1.31526977e-01 5.19201159e-01 7.67462909e-01 -1.24791920e+00
3.09546620e-01 8.52523685e-01 2.69395679e-01 -1.21533835e+00
-8.19729865e-01 -4.51662987e-02 -1.17197967e+00 -1.47866577e-01
2.54803419e-01 -2.60850340e-01 -6.38107777e-01 1.45729256e+00
7.02859104e-01 4.19494212e-01 1.17791407e-01 8.65902722e-01
-9.29146111e-02 5.99115252e-01 -5.61666727e-01 -5.43479323e-01
9.80464458e-01 -6.97250366e-01 -7.34751821e-01 2.36028116e-02
7.81735718e-01 -7.84766674e-01 1.10748363e+00 3.96763504e-01
-7.54267752e-01 -3.85722548e-01 -1.07989013e+00 3.73457938e-01
-1.88862503e-01 2.01948524e-01 5.92143953e-01 8.10306251e-01
-9.12537813e-01 6.80092335e-01 -7.95961559e-01 -3.76446545e-01
7.51767606e-02 6.84715688e-01 -4.95256871e-01 2.35357910e-01
-7.89758325e-01 2.57964104e-01 3.07090759e-01 8.91539529e-02
-1.32961914e-01 -7.25016594e-01 -7.96811163e-01 -1.72605738e-01
2.40710750e-01 -2.98749864e-01 7.89720476e-01 -7.70734847e-01
-1.56804240e+00 7.20890939e-01 -1.04606077e-01 1.32508412e-01
2.29406774e-01 -4.73783948e-02 -3.52373809e-01 1.60699710e-01
2.30883643e-01 7.47948736e-02 1.45597446e+00 -1.23258007e+00
-6.84585631e-01 -5.24110496e-01 -4.78930980e-01 1.82476386e-01
-7.98351407e-01 -4.33872074e-01 -2.78319627e-01 -7.34750748e-01
4.48627323e-01 -1.32602298e+00 -5.23757875e-01 -4.62943584e-01
-6.16791904e-01 -2.31537014e-01 1.09336054e+00 -7.48992525e-03
1.39794528e+00 -2.58281922e+00 5.69280803e-01 8.37599993e-01
5.91417193e-01 -3.78261320e-02 -3.02307427e-01 3.81330907e-01
-5.20279288e-01 6.01917505e-02 -2.16826141e-01 -1.56901583e-01
4.14285772e-02 1.42267764e-01 6.71714470e-02 1.12088573e+00
-9.53483433e-02 2.99358785e-01 -7.95477271e-01 -3.52198243e-01
2.35175237e-01 8.48132595e-02 -7.70896256e-01 -8.86693597e-02
5.60168982e-01 1.80170625e-01 -5.98549783e-01 1.26082778e-01
8.57365608e-01 -3.58540237e-01 1.99297607e-01 -4.45654690e-01
1.47273079e-01 -3.88025254e-01 -1.84281564e+00 1.34375000e+00
-1.18113421e-01 4.33351934e-01 3.24434876e-01 -1.41852200e+00
6.51773214e-01 -2.17161015e-01 1.10750663e+00 6.02266639e-02
7.60414973e-02 -8.34146440e-02 3.29865888e-02 -3.66484672e-01
1.19902022e-01 -1.69823363e-01 -3.01064640e-01 4.98478651e-01
-1.48672462e-01 -1.22373901e-01 3.45364630e-01 2.97968924e-01
1.21891737e+00 -8.11398864e-01 4.07964170e-01 -7.15289295e-01
7.19511926e-01 -2.03908205e-01 5.96031427e-01 6.25653446e-01
-3.14028621e-01 3.74676704e-01 5.50160646e-01 -1.20816477e-01
-1.15597999e+00 -1.04316306e+00 -4.74243462e-01 9.07068849e-01
1.59181297e-01 -6.04051948e-01 -1.07411146e+00 -8.68142307e-01
2.89287627e-01 3.68083060e-01 -7.65858352e-01 -4.94627714e-01
-3.92617583e-01 -9.92427528e-01 2.34590232e-01 8.23161528e-02
1.48168281e-01 -2.51046032e-01 -1.10791601e-01 -4.10324037e-02
2.81220794e-01 -7.29750097e-01 -8.68362427e-01 4.07210767e-01
-7.39105642e-01 -1.34152973e+00 -4.13409412e-01 -9.33123887e-01
1.19984472e+00 5.23253977e-01 3.84866625e-01 -2.65491039e-01
-5.92661619e-01 6.36511087e-01 -3.85845512e-01 2.65765995e-01
-1.63702428e-01 2.14295655e-01 7.17933357e-01 4.30428594e-01
5.58742404e-01 -4.60696846e-01 -4.42071497e-01 6.45263314e-01
-1.20960379e+00 -3.83439720e-01 3.91020864e-01 1.10465515e+00
6.29546762e-01 5.25222063e-01 7.78371990e-01 -9.44362700e-01
8.86064708e-01 -4.10252273e-01 -6.43314481e-01 1.31129339e-01
-8.91712070e-01 4.33633745e-01 9.42235947e-01 -7.42080569e-01
-7.74168372e-01 7.75189579e-01 6.40726149e-01 -4.36042100e-01
-1.82303339e-01 5.94370179e-02 -2.21357018e-01 -3.47832978e-01
1.01987207e+00 8.96492749e-02 7.37432670e-03 -5.23933351e-01
8.08626354e-01 9.04628396e-01 1.27539501e-01 -4.80321527e-01
1.29537344e+00 6.24792635e-01 3.70011657e-01 -9.16588664e-01
-7.11164892e-01 -8.15572143e-01 -1.21049368e+00 1.73774719e-01
4.88734484e-01 -4.31400657e-01 -6.15176201e-01 1.05023518e-01
-5.06683648e-01 3.54018547e-02 -5.46323955e-01 5.48919916e-01
-8.07273924e-01 7.10112453e-01 -4.97837096e-01 -4.36164111e-01
2.05291599e-01 -8.57742786e-01 9.78003442e-01 -1.21026538e-01
-1.38887614e-01 -1.03085089e+00 4.83641058e-01 -6.09466098e-02
-7.40662068e-02 4.95604463e-02 1.05488122e+00 -5.14783740e-01
1.29384845e-01 -4.47483450e-01 -3.47794108e-02 4.46844846e-01
7.44493067e-01 -3.89936455e-02 -5.51074028e-01 -6.25200450e-01
2.81901062e-01 2.00348087e-02 6.73202395e-01 3.92514855e-01
1.15472925e+00 -4.36800212e-01 -5.62659323e-01 7.68645346e-01
1.11241174e+00 1.91767216e-01 6.78918848e-04 1.60950989e-01
7.62525141e-01 5.83756685e-01 7.73202538e-01 8.55614722e-01
-2.05979407e-01 4.47510511e-01 -2.27492854e-01 3.38051422e-03
3.39731336e-01 2.94022828e-01 1.77837521e-01 1.09957767e+00
4.49572384e-01 4.09883261e-01 -8.26703548e-01 2.97609180e-01
-2.20546031e+00 -7.52685964e-01 -2.17211977e-01 2.00678205e+00
7.24216163e-01 -2.52768308e-01 2.42178559e-01 2.97858149e-01
1.20879853e+00 -8.03440213e-02 -8.01444829e-01 -1.82275414e-01
-3.30417305e-02 -3.68325621e-01 5.27657986e-01 6.72590971e-01
-1.44951189e+00 8.94720554e-01 7.20057631e+00 9.22711015e-01
-8.20975721e-01 -1.52046025e-01 2.96915829e-01 -2.94340700e-01
7.21478760e-02 -1.79621026e-01 -7.31119037e-01 3.40134591e-01
5.34949839e-01 -6.36140287e-01 8.81904006e-01 8.87792230e-01
3.02913696e-01 2.00474843e-01 -1.29548132e+00 1.43516386e+00
1.40381336e-01 -9.65565860e-01 1.03960782e-01 1.60996258e-01
1.09676826e+00 -3.32719982e-01 3.68427604e-01 3.25766541e-02
2.76942402e-01 -5.92004955e-01 -1.78628822e-03 3.83312613e-01
6.53514922e-01 -9.37785447e-01 2.26096332e-01 2.04465479e-01
-1.25191510e+00 -3.24990392e-01 -7.42265642e-01 -2.61841547e-02
-1.65093914e-01 9.45603073e-01 -7.74476290e-01 -1.40340105e-02
4.58018273e-01 1.01013374e+00 -8.02530587e-01 9.31832910e-01
3.93824637e-01 3.58719617e-01 -5.46977639e-01 1.11475110e-01
2.43799642e-01 -7.20850289e-01 5.73220670e-01 1.13985908e+00
2.74137437e-01 3.45537543e-01 3.18471998e-01 3.49118352e-01
-6.25086427e-02 3.59482944e-01 -6.30157590e-01 -7.41944015e-02
4.98922735e-01 1.53607476e+00 -8.81961703e-01 -2.27921665e-01
-3.54039013e-01 1.11636496e+00 4.43446249e-01 6.49630487e-01
-4.67044801e-01 -6.69613659e-01 8.88850689e-01 6.44305721e-02
2.93537915e-01 -2.93436289e-01 -2.69434422e-01 -1.37561452e+00
-1.82517737e-01 -7.66977906e-01 5.74111819e-01 -1.02129556e-01
-1.84645462e+00 3.17954659e-01 -1.39248483e-02 -1.40620506e+00
-2.13632911e-01 -5.52972555e-01 -2.20416084e-01 3.92321765e-01
-8.02661121e-01 -3.87656987e-01 9.81474575e-03 1.09845924e+00
2.61354417e-01 -3.54668498e-01 7.60690689e-01 1.47702321e-01
-8.57232928e-01 5.47212064e-01 1.12847173e+00 -1.17534712e-01
9.12126601e-01 -1.49015248e+00 1.17701245e-02 6.49841785e-01
-1.08488202e-01 6.79491580e-01 4.51959521e-01 -6.07078969e-01
-1.56863856e+00 -1.18383896e+00 2.41763994e-01 -2.51786500e-01
1.02141595e+00 -4.64018732e-01 -6.94299698e-01 6.59129620e-01
-1.87583432e-01 2.35431641e-01 1.17949903e+00 1.08534746e-01
-3.41233283e-01 -2.37489969e-01 -1.48888731e+00 4.64050978e-01
1.05152071e+00 -3.93336177e-01 -5.59004366e-01 6.28459573e-01
3.88731658e-01 1.36121362e-01 -1.03355682e+00 2.38820642e-01
1.68120295e-01 -5.28254926e-01 9.49341536e-01 -8.50530386e-01
-2.71053970e-01 -4.96842444e-01 -2.51233608e-01 -1.67873359e+00
-1.05001462e+00 -8.80139410e-01 -1.27066165e-01 1.03619015e+00
1.52540714e-01 -7.51198947e-01 1.03280151e+00 6.64709091e-01
3.33852708e-01 -4.50021595e-01 -1.15265357e+00 -9.28106785e-01
1.12323001e-01 -1.37488857e-01 2.28307515e-01 1.31667960e+00
5.84487140e-01 5.41928589e-01 -1.16697095e-01 2.57492006e-01
1.02661479e+00 1.16050780e-01 6.77377939e-01 -1.42005467e+00
-2.74341375e-01 -4.44393069e-01 -7.59067833e-01 -7.25391567e-01
6.16742373e-01 -9.66132224e-01 -1.25036806e-01 -8.86246026e-01
3.95451725e-01 -4.68373626e-01 -2.02036873e-01 3.10891181e-01
-2.33093023e-01 1.66592747e-01 1.44048646e-01 3.81407619e-01
-8.63389134e-01 6.00971460e-01 9.11868751e-01 -1.66143402e-01
-5.67860961e-01 -5.67623824e-02 -5.82690239e-01 8.58211398e-01
7.28270471e-01 -6.88926280e-01 -4.07148242e-01 1.32520288e-01
1.30119517e-01 -2.84336418e-01 -2.63574898e-01 -8.89175951e-01
1.60657555e-01 -4.66645271e-01 5.00930071e-01 -4.30150241e-01
1.55830845e-01 -1.15056765e+00 -4.83386293e-02 3.06740075e-01
-4.65629846e-01 -1.14432745e-01 -3.60249169e-02 8.40101063e-01
-8.69763196e-02 -3.04262847e-01 1.07781613e+00 4.60862517e-01
-1.44855976e-01 6.37504086e-02 -7.35005975e-01 2.66645588e-02
1.58233595e+00 -1.56420410e-01 -2.25308035e-02 -2.41553709e-01
-1.10298240e+00 2.65706778e-01 5.66316307e-01 2.16202259e-01
5.10140657e-01 -1.64665401e+00 -3.31337035e-01 3.71148914e-01
1.15529813e-01 -2.46356696e-01 9.54996049e-02 7.17701554e-01
-1.45865604e-01 3.79295707e-01 -1.17462892e-02 -6.90417528e-01
-1.23963964e+00 1.08965206e+00 3.36486340e-01 -7.81374704e-03
-4.03816968e-01 4.77384478e-01 3.26882750e-01 -6.33758545e-01
1.90945163e-01 -2.22063780e-01 -8.68979245e-02 2.76216656e-01
4.47004825e-01 9.02016580e-01 -8.39704126e-02 -6.12776995e-01
-3.95792484e-01 9.43035305e-01 -1.03433013e-01 -2.69750506e-03
1.41192615e+00 -5.11429667e-01 -3.64298761e-01 4.75017726e-01
1.77293861e+00 6.38880506e-02 -1.31559646e+00 -3.32694799e-01
2.48203710e-01 -7.35622704e-01 4.60320106e-03 -1.20958179e-01
-9.55906153e-01 3.15808803e-01 7.12360084e-01 6.09840751e-01
1.10190332e+00 1.06272206e-01 3.81946743e-01 9.73026991e-01
3.44096422e-02 -1.48735476e+00 2.78407454e-01 3.12425077e-01
7.29055166e-01 -9.06636834e-01 5.16352989e-03 -6.84680521e-01
-4.23800111e-01 1.10858500e+00 4.85631138e-01 -7.02827394e-01
1.07829154e+00 2.02479452e-01 -1.30484998e-01 -6.14349917e-02
-4.25807178e-01 7.44056106e-02 5.19580781e-01 7.08929658e-01
-2.53560860e-02 9.69418362e-02 -4.45065856e-01 7.19144166e-01
-2.34006777e-01 -5.61718166e-01 5.30610979e-01 6.07772827e-01
-3.85461926e-01 -1.08891940e+00 -8.92663002e-01 8.61851037e-01
-5.30260243e-02 2.61104345e-01 -7.67072737e-01 4.31120515e-01
-1.83579568e-02 1.03685093e+00 1.52136609e-01 -4.33649659e-01
3.62422287e-01 5.76847829e-02 4.17865157e-01 -7.07211733e-01
-1.56965375e-01 3.76326174e-01 -3.93868715e-01 -6.22467279e-01
-3.60241592e-01 -1.02047598e+00 -1.33338141e+00 -1.86209261e-01
-4.75247800e-01 5.77024698e-01 4.64985937e-01 6.88347280e-01
6.02918565e-01 1.62648648e-01 1.38110077e+00 -1.07687652e+00
-8.44717026e-01 -7.94344187e-01 -1.24998164e+00 6.58285916e-01
4.24485326e-01 -8.33158910e-01 -8.95854473e-01 1.88717440e-01] | [7.785927772521973, 4.3043212890625] |
74a6c08b-5610-446f-b622-a0316657a3d8 | implicit-u-net-for-volumetric-medical-image | 2206.15217 | null | https://arxiv.org/abs/2206.15217v1 | https://arxiv.org/pdf/2206.15217v1.pdf | Implicit U-Net for volumetric medical image segmentation | U-Net has been the go-to architecture for medical image segmentation tasks, however computational challenges arise when extending the U-Net architecture to 3D images. We propose the Implicit U-Net architecture that adapts the efficient Implicit Representation paradigm to supervised image segmentation tasks. By combining a convolutional feature extractor with an implicit localization network, our implicit U-Net has 40% less parameters than the equivalent U-Net. Moreover, we propose training and inference procedures to capitalize sparse predictions. When comparing to an equivalent fully convolutional U-Net, Implicit U-Net reduces by approximately 30% inference and training time as well as training memory footprint while achieving comparable results in our experiments with two different abdominal CT scan datasets. | ['Giacomo Tarroni', 'Sergio Naval Marimont'] | 2022-06-30 | null | null | null | null | ['volumetric-medical-image-segmentation'] | ['medical'] | [ 6.92222953e-01 1.03869939e+00 -5.17984092e-01 -5.00551939e-01
-9.91141498e-01 -2.02772208e-03 1.28541857e-01 3.55923414e-01
-4.44969237e-01 5.62847257e-01 1.21798575e-01 -6.78223372e-01
1.74036235e-01 -9.08398986e-01 -9.71605182e-01 -2.67398119e-01
-2.42879465e-01 4.86523181e-01 3.48338902e-01 1.96231097e-01
-9.39818993e-02 2.51855224e-01 -8.36631417e-01 4.99470145e-01
6.92399561e-01 1.26684356e+00 1.95329800e-01 1.00065196e+00
-2.56930962e-02 1.21999228e+00 -1.26453653e-01 9.08219889e-02
2.66770571e-01 -1.32372886e-01 -1.27657831e+00 -7.97483101e-02
4.70575154e-01 -9.44066226e-01 -3.75978023e-01 8.12764764e-01
6.97426915e-01 2.82096416e-01 5.01512706e-01 -4.04829890e-01
-5.39881945e-01 6.96967781e-01 -1.51747212e-01 3.28096837e-01
1.76356062e-01 -7.76711702e-02 7.32590437e-01 -5.56095541e-01
7.19728589e-01 9.81511593e-01 1.34188867e+00 6.34148955e-01
-1.42390275e+00 -1.09768860e-01 -5.74489869e-02 -2.75924087e-01
-1.12843919e+00 -1.97761849e-01 1.26010969e-01 -2.84211993e-01
1.29461288e+00 1.68116823e-01 3.08181196e-01 7.08864748e-01
2.31975585e-01 1.23245144e+00 7.90379286e-01 -3.33614051e-01
1.27289534e-01 -9.02425423e-02 7.14813471e-01 1.19320369e+00
2.13805452e-01 1.37525871e-01 3.77696976e-02 -1.05936080e-01
1.41964376e+00 1.17250852e-01 -1.90886796e-01 -2.25513726e-01
-9.57452178e-01 9.86500084e-01 1.00777543e+00 2.47891501e-01
-2.29673669e-01 6.15220964e-01 7.40776300e-01 4.60631698e-02
4.65039700e-01 5.82616866e-01 -6.28838062e-01 4.33116183e-02
-1.10341203e+00 -5.50878793e-02 7.62770116e-01 1.26081252e+00
8.75479758e-01 -2.45984253e-02 -3.30707401e-01 5.42919099e-01
-6.59707710e-02 -8.28016847e-02 5.27477324e-01 -9.79267001e-01
3.94635163e-02 5.91602802e-01 -3.64510506e-01 -6.93995059e-01
-8.14138234e-01 -8.39227438e-01 -9.40510511e-01 7.43474215e-02
9.42658558e-02 -3.77147049e-01 -1.58784616e+00 1.25210440e+00
9.99560654e-02 5.77572823e-01 -1.92971259e-01 7.48717308e-01
1.31872499e+00 2.82171100e-01 8.47909451e-02 3.46296310e-01
1.14509928e+00 -1.37304163e+00 -5.58692634e-01 -9.31581631e-02
1.12655509e+00 -4.08650726e-01 7.19050229e-01 6.81734309e-02
-1.19415867e+00 -5.27825594e-01 -9.83819485e-01 -7.34417200e-01
-2.76765913e-01 1.70277283e-01 9.08335626e-01 5.84416687e-01
-1.33004355e+00 8.53897393e-01 -1.20880783e+00 -2.29270756e-01
5.51510572e-01 8.58409345e-01 -3.14043075e-01 1.20553985e-01
-1.07881725e+00 7.57257640e-01 7.35832155e-01 3.91565636e-02
-7.69444346e-01 -7.64487565e-01 -1.34290600e+00 2.50484198e-01
4.34310794e-01 -9.93522167e-01 1.59752214e+00 -8.07993412e-01
-1.58963597e+00 6.59167588e-01 -8.44926387e-02 -1.04276001e+00
3.52930129e-01 -2.47297198e-01 2.82852799e-01 4.83072490e-01
1.96566880e-01 8.74754369e-01 3.36396426e-01 -8.14027131e-01
-2.97476679e-01 5.87396808e-02 1.84975773e-01 5.82501702e-02
2.63773888e-01 -5.63265800e-01 -7.74266660e-01 -8.53135586e-01
4.03100222e-01 -8.26707065e-01 -1.15677798e+00 2.55364209e-01
-1.79012522e-01 1.99944422e-01 6.37033522e-01 -7.18170822e-01
8.55997622e-01 -1.75708091e+00 -2.30045557e-01 3.67590517e-01
3.90789121e-01 4.15440112e-01 -8.68358314e-02 -3.66742581e-01
1.16239255e-03 -5.08873463e-02 -5.66649735e-01 -3.63073260e-01
-3.72352570e-01 8.69107068e-01 3.64411436e-02 2.10578978e-01
2.54180342e-01 1.24887168e+00 -1.16232848e+00 -6.61240458e-01
2.44254068e-01 4.28411871e-01 -1.26541519e+00 8.67269933e-02
-1.32119626e-01 4.10580158e-01 -5.39018631e-01 8.21636021e-01
4.88562852e-01 -7.46276140e-01 4.17160600e-01 -9.85853821e-02
2.25775719e-01 5.14828682e-01 -6.09991550e-01 2.21436381e+00
-5.67630529e-01 2.61443049e-01 2.47960269e-01 -1.32423687e+00
6.42450750e-01 5.58868945e-01 6.99506879e-01 -7.81147718e-01
6.20315254e-01 3.82226318e-01 -2.41747603e-01 -3.05784464e-01
4.83716965e-01 -2.04952136e-02 -9.16889384e-02 3.19838017e-01
5.90195775e-01 -8.22737440e-02 5.31981736e-02 2.08526343e-01
1.21509969e+00 3.80278170e-01 6.25288904e-01 -5.20332158e-01
4.23147947e-01 6.94041178e-02 6.04447544e-01 1.14512944e+00
-1.41358316e-01 7.84753799e-01 6.04126632e-01 -9.03604746e-01
-5.27437925e-01 -9.70585883e-01 -5.72672009e-01 8.07471037e-01
-2.09277123e-02 -3.83969367e-01 -7.51757920e-01 -9.83412147e-01
1.72655936e-02 4.03423220e-01 -1.13588512e+00 7.85178393e-02
-8.10625136e-01 -5.74387848e-01 7.11674988e-01 1.09207380e+00
3.55953127e-01 -8.98583353e-01 -7.30795026e-01 5.02257943e-01
7.82604516e-02 -1.19742155e+00 -3.90292495e-01 7.57835448e-01
-1.57259727e+00 -1.22086310e+00 -7.35641479e-01 -9.54305947e-01
1.17355955e+00 -2.61262894e-01 1.34964144e+00 1.45531535e-01
-6.23404026e-01 2.02024519e-01 -3.94169182e-01 -2.61555731e-01
-1.85414970e-01 4.46316332e-01 -4.03343797e-01 -8.45729351e-01
1.08852595e-01 -2.44395494e-01 -9.35203910e-01 2.44042501e-02
-8.13551068e-01 1.75350443e-01 8.47037077e-01 1.29392004e+00
1.16613829e+00 -6.98366940e-01 3.47292364e-01 -1.60567856e+00
6.05150275e-02 -5.83094656e-01 -2.10808069e-01 -2.32038349e-01
-5.78722119e-01 -9.77130607e-03 2.81569719e-01 -1.18319415e-01
-8.67652416e-01 3.77241045e-01 -7.28320837e-01 -5.25615573e-01
8.07706937e-02 6.44907892e-01 6.22657239e-01 -3.20722878e-01
6.28870308e-01 8.54176506e-02 2.03160629e-01 -6.01228714e-01
5.32973409e-01 2.75308549e-01 7.01284051e-01 -5.44655025e-01
2.66653746e-01 3.35805327e-01 1.05849572e-01 -7.20277548e-01
-1.08237219e+00 -5.11830628e-01 -9.40360427e-01 4.98786986e-01
1.05447519e+00 -8.43817353e-01 -4.16578919e-01 -1.06806882e-01
-1.06967211e+00 -6.54056966e-01 -7.64243364e-01 5.09749889e-01
-7.95208097e-01 3.44668150e-01 -1.26191533e+00 -4.21388038e-02
-8.81832361e-01 -1.59007120e+00 1.11544335e+00 -4.66669239e-02
-3.96617651e-01 -1.18539989e+00 5.89490421e-02 1.08827636e-01
6.26843393e-01 3.95382851e-01 8.96312654e-01 -5.25508523e-01
-6.64887190e-01 -3.17940086e-01 -5.47223628e-01 3.55753422e-01
2.34003499e-01 -8.10257554e-01 -8.81939530e-01 -2.33118013e-01
-2.07186833e-01 -4.31148171e-01 1.00879180e+00 8.19870591e-01
1.85117924e+00 -4.26347882e-01 -4.55095053e-01 1.21016967e+00
1.28831840e+00 -8.17600340e-02 5.32823980e-01 4.86512743e-02
5.00802279e-01 -7.03604072e-02 2.65479416e-01 6.35586530e-02
8.25530514e-02 2.59304374e-01 4.92465317e-01 -7.93979108e-01
-3.99486184e-01 -1.47108272e-01 -1.42711803e-01 9.20343041e-01
-1.54773340e-01 5.27530432e-01 -7.62896895e-01 7.08778620e-01
-1.90108097e+00 -5.41647136e-01 1.77681088e-01 1.82304728e+00
1.01736557e+00 1.97665647e-01 -2.80883223e-01 -2.63373107e-01
1.91607863e-01 -1.57276437e-01 -4.42688525e-01 -6.07730031e-01
4.64379668e-01 9.41491306e-01 7.55554378e-01 5.65616667e-01
-1.32467175e+00 6.82549119e-01 7.84532785e+00 7.07105815e-01
-1.29336166e+00 3.92065167e-01 1.03842306e+00 -9.82186347e-02
-9.84619334e-02 -2.51390725e-01 -3.21418136e-01 -3.24937627e-02
7.68610537e-01 1.25408068e-01 -1.99955642e-01 1.18217814e+00
-2.41489246e-01 5.47358170e-02 -1.27865219e+00 8.90745699e-01
-9.51664597e-02 -1.95024335e+00 1.28690898e-01 -7.11683109e-02
7.43150413e-01 5.53440690e-01 -1.17133118e-01 5.95035195e-01
3.96942198e-01 -1.30876088e+00 2.25051820e-01 1.37556225e-01
1.23591805e+00 -4.67821419e-01 1.08381164e+00 2.01875076e-01
-1.22791564e+00 3.72314662e-01 -4.16283965e-01 -1.24255583e-01
1.20054334e-01 3.75825882e-01 -1.32001603e+00 6.26332045e-01
7.30633736e-01 6.58967614e-01 -2.28329077e-01 9.18492198e-01
-9.91790369e-02 5.28082252e-01 -3.11818719e-01 5.26317239e-01
7.76240230e-01 4.86597382e-02 2.86794662e-01 1.55907106e+00
1.85269222e-01 4.81050044e-01 4.50831503e-01 7.11171329e-01
-3.79320145e-01 -3.92058305e-03 -6.45349741e-01 2.51624942e-01
-2.85004258e-01 1.21318090e+00 -9.31129217e-01 -7.25281894e-01
-5.30020773e-01 1.12018001e+00 4.39691782e-01 1.78942561e-01
-5.69739163e-01 -4.25770134e-01 4.64025468e-01 1.72075972e-01
4.11345422e-01 -6.81452006e-02 -4.31238532e-01 -1.00991237e+00
-5.91759384e-01 -2.83171356e-01 5.30117512e-01 -2.97552496e-01
-9.76931632e-01 8.31170321e-01 -3.65403503e-01 -1.05063689e+00
-2.75713474e-01 -9.99835312e-01 -4.25377101e-01 6.61890864e-01
-1.82054615e+00 -1.00369942e+00 -9.88268852e-02 5.39944172e-01
3.95788193e-01 3.60477239e-01 1.25631118e+00 3.64429235e-01
-2.46952295e-01 5.58063149e-01 -1.33212820e-01 8.61706674e-01
3.22884649e-01 -1.49950027e+00 6.28426969e-01 4.51481998e-01
-2.01703399e-01 1.04963732e+00 1.69091746e-01 -5.34572124e-01
-1.19244969e+00 -1.47067845e+00 3.75420511e-01 -1.75094217e-01
4.18110549e-01 -1.06306002e-02 -6.57096088e-01 1.39529550e+00
2.56068408e-01 8.14557672e-01 1.00712180e+00 3.32624674e-01
-6.95124269e-02 4.27867234e-01 -1.34865320e+00 2.77075499e-01
9.21692431e-01 -4.63744849e-01 -7.95958221e-01 2.53580183e-01
8.83203864e-01 -1.07030022e+00 -1.34185135e+00 7.13829696e-01
4.26437914e-01 -5.61825275e-01 9.95378673e-01 -4.53237385e-01
5.31993032e-01 1.48510367e-01 7.17327669e-02 -8.94636929e-01
-6.58217609e-01 -2.86962599e-01 -4.86007839e-01 1.33002371e-01
4.86474305e-01 -6.55970931e-01 1.13481045e+00 3.98138791e-01
-6.40450716e-01 -1.41482544e+00 -8.11393142e-01 -2.22880974e-01
1.82939202e-01 -3.80157650e-01 2.31168941e-01 9.08543050e-01
-1.35847703e-02 1.32185847e-01 -6.67677596e-02 -5.81965223e-02
7.00764656e-01 2.07759887e-01 3.98193479e-01 -9.17247891e-01
-5.15438437e-01 -2.28059605e-01 -4.28329200e-01 -1.69914067e+00
-1.36744395e-01 -1.32817161e+00 2.17430592e-02 -1.78731883e+00
1.33489490e-01 -8.48093092e-01 -6.62110031e-01 8.87855589e-01
-1.75973043e-01 7.14196146e-01 -2.74411240e-03 -6.76931590e-02
-5.15724659e-01 3.59224886e-01 1.50619948e+00 -3.69310319e-01
-1.53188109e-01 -3.16654220e-02 -5.31969190e-01 9.48207021e-01
5.21175861e-01 -2.24450082e-01 -4.53890890e-01 -8.46956313e-01
-1.86545253e-01 4.66413260e-01 5.98137081e-02 -9.66491044e-01
1.14440523e-01 3.85402739e-01 5.79686761e-01 -8.15211833e-01
1.94490731e-01 -9.25128818e-01 -5.53693175e-01 1.03627419e+00
-3.66837770e-01 -3.20139170e-01 5.11676669e-01 3.71630609e-01
-3.77757639e-01 -2.98081815e-01 6.61541522e-01 -7.65751779e-01
-8.01536798e-01 7.80790865e-01 -5.28172016e-01 -1.97389320e-01
9.68269408e-01 -4.21172082e-01 -8.06051940e-02 1.08894862e-01
-1.31798303e+00 3.98265153e-01 1.25100270e-01 -1.01411097e-01
7.96873093e-01 -9.57516789e-01 -3.73465449e-01 3.72975558e-01
-2.14434236e-01 7.39453435e-01 3.15013468e-01 1.05017459e+00
-1.26938987e+00 8.42630744e-01 -1.73676714e-01 -8.12077284e-01
-9.48940516e-01 3.30668449e-01 5.32821655e-01 -7.70077527e-01
-1.18815565e+00 9.84750032e-01 5.17487764e-01 -1.03212285e+00
6.06482401e-02 -8.35566819e-01 -5.75913191e-02 -3.23186725e-01
4.52871293e-01 2.80627757e-01 2.64653385e-01 -2.16323480e-01
-2.85662830e-01 1.42423630e-01 -3.46884280e-01 4.52365369e-01
1.36236024e+00 2.66740471e-01 -1.28047824e-01 -6.94009587e-02
1.42065024e+00 -5.06906629e-01 -1.05845046e+00 -5.29063106e-01
-1.28184587e-01 -1.42616816e-02 4.66250449e-01 -8.32886040e-01
-1.34125578e+00 7.37086773e-01 4.99130666e-01 -3.65121037e-01
8.80495906e-01 -1.52019873e-01 1.08220959e+00 5.98821819e-01
4.99298096e-01 -7.94300556e-01 -1.74194470e-01 7.90726423e-01
3.79678756e-01 -1.19768465e+00 2.92403162e-01 -7.41859496e-01
-3.22060883e-01 1.21525240e+00 6.43719852e-01 -6.03686988e-01
9.72451866e-01 5.41278303e-01 2.67946482e-01 -3.69058728e-01
-4.50000852e-01 -2.46677428e-01 3.37248474e-01 2.85754919e-01
8.31241608e-01 2.06110716e-01 -2.03162327e-01 6.29633427e-01
-1.27326408e-02 5.22881210e-01 2.92314768e-01 1.25103486e+00
-2.07302228e-01 -1.02222788e+00 -1.10518308e-02 8.36409271e-01
-6.50669098e-01 -1.96914002e-01 2.71233559e-01 7.92121172e-01
2.07906678e-01 2.41227642e-01 2.31397167e-01 5.96508235e-02
1.44580919e-02 -1.93073049e-01 6.98625445e-01 -1.11221862e+00
-8.72366548e-01 -5.12235016e-02 7.57949008e-03 -1.09454906e+00
-3.25061738e-01 -1.17604434e-01 -1.68476963e+00 1.14636466e-01
-3.22333068e-01 9.30808261e-02 3.44396502e-01 8.37704599e-01
3.27060729e-01 1.14727318e+00 -7.07506686e-02 -1.11042261e+00
-4.21534777e-01 -8.68447363e-01 -4.15647715e-01 1.09272823e-01
4.44888562e-01 -4.62266862e-01 2.48677477e-01 -4.48846743e-02] | [14.556987762451172, -2.6047840118408203] |
4a474dd9-c60a-47e8-b6fc-15f95abb2000 | interactive-learning-with-corrective-feedback | 1810.00466 | null | http://arxiv.org/abs/1810.00466v1 | http://arxiv.org/pdf/1810.00466v1.pdf | Interactive Learning with Corrective Feedback for Policies based on Deep Neural Networks | Deep Reinforcement Learning (DRL) has become a powerful strategy to solve
complex decision making problems based on Deep Neural Networks (DNNs). However,
it is highly data demanding, so unfeasible in physical systems for most
applications. In this work, we approach an alternative Interactive Machine
Learning (IML) strategy for training DNN policies based on human corrective
feedback, with a method called Deep COACH (D-COACH). This approach not only
takes advantage of the knowledge and insights of human teachers as well as the
power of DNNs, but also has no need of a reward function (which sometimes
implies the need of external perception for computing rewards). We combine Deep
Learning with the COrrective Advice Communicated by Humans (COACH) framework,
in which non-expert humans shape policies by correcting the agent's actions
during execution. The D-COACH framework has the potential to solve complex
problems without much data or time required. Experimental results validated the
efficiency of the framework in three different problems (two simulated, one
with a real robot), with state spaces of low and high dimensions, showing the
capacity to successfully learn policies for continuous action spaces like in
the Car Racing and Cart-Pole problems faster than with DRL. | ['Javier Ruiz-del-Solar', 'Rodrigo Pérez-Dattari', 'Carlos Celemin', 'Jens Kober'] | 2018-09-30 | null | null | null | null | ['carracing-v0'] | ['playing-games'] | [-7.02819973e-02 5.36976874e-01 2.10857868e-01 1.13216855e-01
2.05677658e-01 -3.83378834e-01 6.76675260e-01 1.35371566e-01
-8.76408696e-01 9.61295545e-01 -5.01819193e-01 -4.88629818e-01
-3.58732432e-01 -7.56586194e-01 -5.86220205e-01 -7.99648821e-01
-1.54524475e-01 6.33872926e-01 6.07839167e-01 -6.14208102e-01
3.66152227e-01 8.02351654e-01 -1.72735715e+00 -1.42750099e-01
9.10780609e-01 8.94105315e-01 5.88903725e-01 6.63305283e-01
4.17209342e-02 1.22248352e+00 -5.60194552e-01 1.98001966e-01
3.74476552e-01 -1.82338849e-01 -6.67235613e-01 -2.76664849e-02
-1.66632950e-01 -5.02857685e-01 -9.27080661e-02 8.89536619e-01
5.15932441e-01 4.74832177e-01 4.07231063e-01 -1.32882440e+00
-3.33911479e-01 3.22187215e-01 -1.74051777e-01 1.03231557e-02
1.55091971e-01 7.65224814e-01 4.13882911e-01 -3.40961158e-01
3.37328523e-01 1.51370037e+00 4.22517300e-01 8.82230461e-01
-1.11824143e+00 -3.56547743e-01 1.47566050e-01 5.30814111e-01
-6.44494593e-01 1.63611758e-03 4.61942762e-01 -6.28575385e-01
1.03780961e+00 -1.38993904e-01 9.08132732e-01 1.17610204e+00
2.62556821e-01 8.37563992e-01 1.34569216e+00 -5.29893219e-01
7.18793750e-01 3.29905719e-01 -2.31356472e-01 5.67468643e-01
-1.23633876e-01 8.28104913e-01 -5.35713248e-02 2.08107665e-01
1.05624092e+00 8.97952244e-02 6.17839769e-02 -7.56867528e-01
-9.38363552e-01 8.53254855e-01 3.84915650e-01 3.92879635e-01
-7.38928258e-01 3.55178148e-01 4.40188885e-01 7.65755534e-01
-5.19708209e-02 7.03629792e-01 -6.42750680e-01 -3.17602396e-01
-1.18646465e-01 4.44587827e-01 7.34703958e-01 4.87862289e-01
6.14737451e-01 4.97662306e-01 -5.78134768e-02 5.53003609e-01
2.05491975e-01 3.26876163e-01 5.76904118e-01 -1.25943041e+00
5.88869266e-02 6.65742159e-01 5.14519989e-01 -8.43960464e-01
-7.22026229e-01 -1.30510196e-01 -7.18150795e-01 1.23394430e+00
5.21802723e-01 -4.49372292e-01 -7.86456347e-01 1.59679282e+00
5.10435581e-01 -2.77237073e-02 1.96832448e-01 9.32925999e-01
1.61958069e-01 5.26835680e-01 1.66472599e-01 -1.13981120e-01
8.39010656e-01 -8.20041180e-01 -6.55496657e-01 -2.72656530e-01
5.92829585e-01 -1.18375130e-01 1.15782869e+00 9.75146770e-01
-9.04646516e-01 -8.11593950e-01 -8.61766160e-01 2.81376272e-01
-5.18094361e-01 -1.19358838e-01 5.90704083e-01 3.14693063e-01
-1.19554687e+00 1.06732082e+00 -9.76800680e-01 -1.73566505e-01
2.55091071e-01 6.21935606e-01 -1.07547693e-01 2.91850537e-01
-1.26708317e+00 1.40056920e+00 6.32034063e-01 1.06650345e-01
-1.33070791e+00 -3.62582445e-01 -6.19516194e-01 1.66233540e-01
8.50970924e-01 -3.04654181e-01 1.66424632e+00 -1.23443544e+00
-2.22283888e+00 1.85194731e-01 8.42154324e-01 -7.90727079e-01
8.31399918e-01 -5.42549014e-01 1.72673576e-02 7.84470588e-02
-3.78390968e-01 8.39097559e-01 8.82110238e-01 -1.21453476e+00
-5.51694930e-01 -1.79877639e-01 4.51361179e-01 -2.12744344e-02
-4.51500118e-02 -4.03918564e-01 3.82557899e-01 -2.47811034e-01
-3.99485081e-01 -1.01604521e+00 -6.18395209e-01 -2.20977943e-02
7.47661069e-02 -7.17692316e-01 9.19731259e-01 -3.48021418e-01
8.64063144e-01 -2.09479499e+00 2.39885777e-01 2.21012190e-01
1.36180431e-01 1.01469910e+00 -2.31725842e-01 6.68865800e-01
-9.82399210e-02 -3.91387492e-01 -8.40592757e-02 1.99313313e-01
1.81646615e-01 7.92405188e-01 -6.12104610e-02 9.51832309e-02
1.84803843e-01 6.17487371e-01 -1.20741642e+00 -1.55617043e-01
4.95732605e-01 2.06557006e-01 -5.54900646e-01 5.19015014e-01
-6.15144193e-01 6.83003306e-01 -3.89126420e-01 1.58274025e-02
1.96544334e-01 2.16627046e-01 2.04383060e-01 4.38654512e-01
-2.48751953e-01 3.57632972e-02 -1.38289976e+00 1.17884171e+00
-6.05952740e-01 4.51924860e-01 2.94624090e-01 -1.37879241e+00
1.11955476e+00 5.06684840e-01 3.92346263e-01 -1.03788280e+00
2.65759945e-01 9.14821774e-02 3.60889196e-01 -9.68022227e-01
2.97037736e-02 -1.72422856e-01 1.51198581e-01 4.55860823e-01
2.70556927e-01 -3.61464411e-01 1.63314059e-01 -1.11393943e-01
1.18548596e+00 3.90867323e-01 3.18711340e-01 -2.62332052e-01
6.75938904e-01 -3.28731313e-02 6.16828084e-01 8.81879210e-01
-3.12628478e-01 -3.60335380e-01 8.00295174e-01 -7.78729498e-01
-1.09048438e+00 -6.86630011e-01 5.77343166e-01 1.19101214e+00
-9.51320454e-02 1.11327745e-01 -8.76474619e-01 -6.73291683e-01
2.01112125e-02 9.69017982e-01 -5.76990962e-01 -3.56761366e-01
-6.52578533e-01 3.16951913e-03 7.09134713e-02 5.15749514e-01
3.55717659e-01 -1.80344713e+00 -1.26041007e+00 5.45107424e-01
5.52198946e-01 -7.93812275e-01 1.19251087e-01 6.14996433e-01
-8.95808458e-01 -1.15878916e+00 -5.62666953e-01 -4.97391641e-01
5.57627916e-01 -1.58954665e-01 9.13652360e-01 1.26620799e-01
-1.99656770e-01 7.81483173e-01 -3.54916394e-01 -6.22337699e-01
-8.34175527e-01 -3.41313750e-01 4.42965955e-01 -4.02660340e-01
1.57138810e-01 -7.45144367e-01 -3.51500750e-01 3.29976857e-01
-8.52706075e-01 1.16273627e-01 7.84168959e-01 1.02352846e+00
1.61735103e-01 1.12246230e-01 7.70275950e-01 -8.39077830e-01
1.00826478e+00 -1.10215612e-01 -1.23080301e+00 4.51637395e-02
-8.43319237e-01 2.10948691e-01 9.46350992e-01 -8.22669625e-01
-9.63986576e-01 9.02394131e-02 -4.28919606e-02 -5.17041266e-01
-5.06917000e-01 1.84086874e-01 1.28473625e-01 -3.31251323e-02
6.82436168e-01 3.32325995e-02 2.46908873e-01 -4.59098190e-01
3.26074600e-01 4.03915226e-01 4.17952210e-01 -6.34387970e-01
4.34609771e-01 -1.53003156e-01 2.31243908e-01 -6.45799637e-01
-5.94214141e-01 1.66364703e-02 -7.75491714e-01 -5.32378674e-01
7.90835321e-01 -3.06516558e-01 -1.45384717e+00 5.72025120e-01
-1.07831836e+00 -9.81244326e-01 -6.46170199e-01 5.72917879e-01
-9.30736959e-01 1.35911256e-01 -6.04348719e-01 -1.13956165e+00
2.37268656e-02 -1.21092653e+00 2.52177805e-01 4.62166697e-01
8.95148814e-02 -9.66736615e-01 8.51305574e-02 -2.89633088e-02
4.76715446e-01 2.48162895e-01 9.85969603e-01 -6.68529809e-01
-4.30448323e-01 1.10919699e-01 1.09394774e-01 6.94222152e-01
-2.27697447e-01 -1.96720466e-01 -8.53688776e-01 -3.28917176e-01
5.39785847e-02 -7.45465040e-01 3.54686588e-01 2.62535699e-02
1.23168981e+00 -3.62708539e-01 1.41498610e-01 -2.59551883e-01
1.15759861e+00 6.88093960e-01 5.33173561e-01 4.67357814e-01
5.03814816e-01 8.54003072e-01 8.06490242e-01 5.88972330e-01
7.60010481e-02 5.74333549e-01 8.60622942e-01 -4.89135925e-03
2.85952508e-01 -1.53104752e-01 4.94277954e-01 4.65312749e-01
-4.36186492e-01 7.27940574e-02 -9.44073439e-01 3.34707439e-01
-2.21654272e+00 -8.41914177e-01 -1.49500100e-02 2.00641441e+00
6.79966152e-01 3.51833254e-01 1.98946610e-01 3.30645502e-01
4.26447421e-01 -3.97232383e-01 -8.89198065e-01 -1.03566635e+00
5.38201094e-01 1.50759727e-01 2.90658116e-01 3.81933391e-01
-8.24610233e-01 8.41348946e-01 5.40227127e+00 7.48934090e-01
-1.00915277e+00 -5.77950589e-02 2.00483382e-01 3.45070511e-01
3.50988865e-01 -2.89549083e-01 -4.46364015e-01 3.31303149e-01
1.08349705e+00 3.12459975e-01 8.63759518e-01 1.14195657e+00
5.24673998e-01 -5.97776890e-01 -1.23239541e+00 7.94465780e-01
-6.53482199e-01 -8.95671487e-01 -5.61855614e-01 -1.19274447e-03
5.25798678e-01 -9.34649184e-02 -2.20277198e-02 8.99158359e-01
7.84611762e-01 -8.01519990e-01 6.17590308e-01 7.11647153e-01
2.67684817e-01 -9.48870301e-01 8.52227390e-01 1.08430898e+00
-4.31094825e-01 -6.42739117e-01 -3.35200876e-01 -3.66891176e-01
-2.57905394e-01 1.30835369e-01 -1.11498392e+00 1.45362362e-01
6.66131675e-01 4.35894877e-01 -2.98511356e-01 8.65901649e-01
-5.81134498e-01 4.11723673e-01 -1.80272982e-01 -4.55656677e-01
6.09125972e-01 -2.34590471e-01 4.19059753e-01 8.10132027e-01
1.42538309e-01 2.30775148e-01 2.42750466e-01 7.45057464e-01
5.90923071e-01 -3.10800046e-01 -7.72193551e-01 -5.92181496e-02
1.08586624e-01 1.07628179e+00 -5.27789116e-01 -2.49779329e-01
-9.94358882e-02 7.58704543e-01 4.72834736e-01 1.87931597e-01
-5.49381256e-01 -2.88819969e-01 4.66948926e-01 4.80705388e-02
3.35660100e-01 -4.24401253e-01 2.05601931e-01 -4.58400637e-01
-4.22947586e-01 -9.22391355e-01 2.40449719e-02 -8.80792081e-01
-1.07275939e+00 3.87427002e-01 2.22831406e-02 -1.21433210e+00
-7.59713233e-01 -9.58013237e-01 -5.33262432e-01 6.34491563e-01
-1.24928617e+00 -4.23895836e-01 -1.00293718e-01 7.12732732e-01
3.99637520e-01 -3.21841210e-01 8.46837342e-01 1.02855191e-01
-4.43441629e-01 5.31814918e-02 5.12417294e-02 -1.67709246e-01
3.22461009e-01 -1.61176896e+00 -2.61067208e-02 2.59919554e-01
-3.16703856e-01 8.12247396e-02 8.41540396e-01 -2.69611925e-01
-1.29733407e+00 -6.82014823e-01 2.31296048e-01 -6.91214725e-02
5.87305009e-01 -2.37735331e-01 -1.06721079e+00 3.31461161e-01
4.09782022e-01 -1.07816689e-01 2.13736534e-01 -1.37867093e-01
2.90625930e-01 -2.06191704e-01 -1.24655604e+00 5.97714603e-01
5.58046997e-01 -1.64433971e-01 -7.28780150e-01 1.95394456e-01
6.79419756e-01 -3.30725431e-01 -5.12456357e-01 3.41378599e-01
2.26291314e-01 -1.20948601e+00 7.35545754e-01 -6.93179607e-01
1.79110572e-01 -2.47953326e-01 4.03826386e-01 -1.77940917e+00
-2.04142809e-01 -5.58561146e-01 -3.23880374e-01 7.68293083e-01
7.63052329e-02 -6.80114150e-01 5.02139091e-01 5.97317874e-01
-2.44186550e-01 -9.48579490e-01 -8.38959992e-01 -8.44811440e-01
-1.07410744e-01 -4.22295511e-01 3.61624718e-01 7.61999965e-01
-5.79509735e-02 2.34178856e-01 -3.68524045e-01 3.04627474e-02
2.33168349e-01 -3.53862613e-01 8.56582880e-01 -1.32272279e+00
-6.38153374e-01 -3.81489575e-01 -2.70406276e-01 -8.60251009e-01
1.76878467e-01 -3.48713428e-01 2.54733294e-01 -1.44423664e+00
-5.54068089e-01 -5.43213844e-01 -3.67989659e-01 7.30303049e-01
1.64727166e-01 -6.54550195e-01 3.98134857e-01 -2.05636919e-01
-7.21241653e-01 6.92350566e-01 1.66660213e+00 1.17526703e-01
-3.68084431e-01 2.69428670e-01 -1.02848582e-01 9.33255196e-01
9.72902596e-01 -4.95061636e-01 -5.48513353e-01 -1.93538368e-01
1.89436167e-01 4.32237893e-01 5.60932994e-01 -1.31542337e+00
1.85663685e-01 -4.86120939e-01 3.05175066e-01 -2.88836598e-01
1.59537509e-01 -1.20231915e+00 -1.74266025e-01 1.07457173e+00
-3.32763433e-01 1.73331663e-01 3.82315964e-01 6.09730899e-01
-2.18621865e-02 -5.20258427e-01 9.36214745e-01 -3.60425532e-01
-1.07482350e+00 -8.95998925e-02 -8.78808975e-01 -2.83362716e-01
1.41010809e+00 -3.38299014e-02 -1.03893496e-01 -3.85773122e-01
-1.00997353e+00 6.14851892e-01 -4.44253273e-02 4.10209239e-01
5.57214856e-01 -9.38170433e-01 -1.47429869e-01 2.54608214e-01
-3.35677236e-01 1.50463253e-01 1.73662826e-01 8.03363860e-01
-5.52979410e-01 3.11268061e-01 -7.08265305e-01 -5.08013070e-01
-8.37086201e-01 1.01974988e+00 6.20603979e-01 -4.01397973e-01
-6.62920952e-01 5.55377483e-01 4.89897057e-02 -7.96494544e-01
5.31649530e-01 -4.37023073e-01 -5.64794481e-01 -1.18532531e-01
4.97963399e-01 5.36834657e-01 1.77056212e-02 2.00569242e-01
1.74318656e-01 1.90606397e-02 5.43168485e-02 -1.34633079e-01
1.63212252e+00 2.25802347e-01 2.26637810e-01 4.58600968e-01
2.69194216e-01 -7.01955914e-01 -1.79704058e+00 -8.85726064e-02
2.46466771e-01 -9.47946310e-02 2.07862128e-02 -1.14629233e+00
-8.19665074e-01 1.24233925e+00 8.13282907e-01 5.35133123e-01
1.01915610e+00 -4.53777969e-01 4.02329057e-01 8.75090718e-01
7.46209383e-01 -1.60490584e+00 5.53240776e-01 7.65443742e-01
9.73413706e-01 -1.21056271e+00 -3.62829983e-01 2.46572852e-01
-8.54313552e-01 1.31859016e+00 1.10435164e+00 -3.49646568e-01
4.15136218e-01 2.02557653e-01 7.68654197e-02 -2.11396646e-02
-1.04271638e+00 -2.17308700e-01 -2.67662585e-01 8.23580325e-01
-1.61121354e-01 1.63860127e-01 -1.44731894e-01 3.53987128e-01
2.16194302e-01 2.16363519e-01 6.24876142e-01 1.11521697e+00
-8.36585641e-01 -1.20033228e+00 -4.42880124e-01 1.41084865e-01
1.00202359e-01 4.91961420e-01 -1.15124315e-01 1.00640023e+00
5.57930648e-01 7.44245768e-01 -1.17157862e-01 -2.30365142e-01
5.59971929e-01 -1.55373588e-01 5.84794462e-01 -5.93997180e-01
-8.05177629e-01 -1.18540607e-01 -1.68649271e-01 -8.01712155e-01
-2.58473366e-01 -4.67314780e-01 -1.39345670e+00 -1.66228607e-01
4.52700295e-02 1.19702376e-01 6.64681554e-01 9.66539800e-01
1.60648942e-01 8.10814023e-01 6.60202086e-01 -1.11632788e+00
-1.06233811e+00 -1.05072653e+00 -5.85870862e-01 1.14459194e-01
3.62021327e-01 -1.07406819e+00 -1.23799622e-01 -2.52563179e-01] | [4.259618282318115, 1.7542214393615723] |
51b8af86-161d-4a13-9ddc-492ec9b2c679 | interact-interaction-network-inference-from | 1801.03011 | null | http://arxiv.org/abs/1801.03011v3 | http://arxiv.org/pdf/1801.03011v3.pdf | INtERAcT: Interaction Network Inference from Vector Representations of Words | In recent years, the number of biomedical publications has steadfastly grown,
resulting in a rich source of untapped new knowledge. Most biomedical facts are
however not readily available, but buried in the form of unstructured text, and
hence their exploitation requires the time-consuming manual curation of
published articles. Here we present INtERAcT, a novel approach to extract
protein-protein interactions from a corpus of biomedical articles related to a
broad range of scientific domains in a completely unsupervised way. INtERAcT
exploits vector representation of words, computed on a corpus of domain
specific knowledge, and implements a new metric that estimates an interaction
score between two molecules in the space where the corresponding words are
embedded. We demonstrate the power of INtERAcT by reconstructing the molecular
pathways associated to 10 different cancer types using a corpus of
disease-specific articles for each cancer type. We evaluate INtERAcT using
STRING database as a benchmark, and show that our metric outperforms currently
adopted approaches for similarity computation at the task of identifying known
molecular interactions in all studied cancer types. Furthermore, our approach
does not require text annotation, manual curation or the definition of semantic
rules based on expert knowledge, and hence it can be easily and efficiently
applied to different scientific domains. Our findings suggest that INtERAcT may
increase our capability to summarize the understanding of a specific disease
using the published literature in an automated and completely unsupervised
fashion. | [] | 2018-04-16 | null | null | null | null | ['text-annotation'] | ['natural-language-processing'] | [ 5.34392715e-01 -2.76989583e-02 -1.08986296e-01 -2.28934243e-01
-5.41554272e-01 -7.62857795e-01 5.34736693e-01 1.17450130e+00
-6.71110928e-01 1.15887320e+00 1.67326242e-01 -3.28124195e-01
-4.55192506e-01 -7.74420500e-01 -5.38921237e-01 -8.77642155e-01
-4.08372562e-03 7.18568563e-01 2.96015263e-01 -1.44038945e-01
3.78258526e-01 5.57689130e-01 -1.44949150e+00 2.57050991e-01
9.37373877e-01 5.79448581e-01 3.67115617e-01 4.71520662e-01
-3.88407886e-01 1.53207630e-01 -6.34545207e-01 -4.87959504e-01
-3.18798333e-01 -6.07584178e-01 -7.63417065e-01 -5.08565366e-01
-2.85287112e-01 6.81456566e-01 1.46510139e-01 9.39606607e-01
5.07129967e-01 -1.95179954e-01 9.39689994e-01 -6.44935906e-01
-1.80140316e-01 3.81225139e-01 -2.38889992e-01 1.05154058e-02
5.96885264e-01 -7.82236457e-02 1.09982181e+00 -8.10440183e-01
1.19621289e+00 7.79693007e-01 5.04808843e-01 1.69915229e-01
-1.33418083e+00 -3.88127208e-01 -4.24090415e-01 2.70276666e-01
-1.40025103e+00 -2.55659848e-01 4.88499403e-01 -6.01297677e-01
1.25948310e+00 4.58599716e-01 5.95737636e-01 9.03797925e-01
3.84676367e-01 -1.23109734e-02 8.53844881e-01 -5.32078624e-01
4.98642147e-01 1.35551602e-01 3.50189775e-01 7.91894197e-01
7.77036071e-01 -3.84811670e-01 -5.75902462e-01 -6.79646075e-01
-3.83753851e-02 -5.41638881e-02 -4.10014540e-01 -3.19852442e-01
-1.42965460e+00 6.23501897e-01 -1.70573115e-01 8.38673055e-01
-3.39705318e-01 -2.29160264e-01 7.78238058e-01 2.32852131e-01
4.54675853e-01 7.47041702e-01 -5.85918546e-01 -1.62220821e-02
-6.67800307e-01 1.89634319e-02 1.15345967e+00 5.01433849e-01
6.45948350e-01 -7.40862131e-01 3.15536350e-01 6.30085528e-01
2.83570141e-01 2.09308073e-01 8.28767657e-01 -1.30820841e-01
8.49530324e-02 9.14888084e-01 1.62880309e-03 -1.32038665e+00
-4.90242094e-01 -1.43966764e-01 -7.11396158e-01 -9.43034738e-02
4.90038007e-01 1.07907221e-01 -4.07726705e-01 1.56811345e+00
5.10570943e-01 1.15470894e-01 4.57656354e-01 4.10255879e-01
9.06593323e-01 3.49788040e-01 3.79097044e-01 -4.35154349e-01
1.66024995e+00 -3.52551877e-01 -6.99303389e-01 3.78916532e-01
8.20356727e-01 -8.10808599e-01 4.13901061e-01 5.71738005e-01
-6.80560470e-01 -1.49563566e-01 -1.13778937e+00 -3.24203596e-02
-1.14026439e+00 9.18356255e-02 7.52933443e-01 4.05967683e-01
-6.78632736e-01 7.93148458e-01 -7.34567940e-01 -7.47527838e-01
3.54226947e-01 4.38819051e-01 -9.71460760e-01 1.49435818e-01
-1.22692060e+00 1.07514870e+00 7.75761366e-01 -8.85461643e-02
-1.79279864e-01 -6.40925884e-01 -6.00015223e-01 7.08433017e-02
3.24598134e-01 -7.86372781e-01 3.37463260e-01 -5.03103614e-01
-1.16985607e+00 1.10885501e+00 -3.04035723e-01 -4.68118608e-01
1.16726644e-01 1.11491755e-01 -5.43604553e-01 3.69480073e-01
3.27519625e-02 9.14046913e-02 2.45737076e-01 -7.96493769e-01
-3.11234951e-01 -4.26930189e-01 -3.06356311e-01 -5.45155304e-03
-5.16144156e-01 -4.73672599e-02 -2.98238814e-01 -5.61354995e-01
-2.20454931e-01 -7.59517550e-01 -4.14857537e-01 6.09610453e-02
-3.26651841e-01 -3.22484195e-01 3.77191544e-01 -4.46037143e-01
1.19223416e+00 -1.98911345e+00 6.14107370e-01 5.03562927e-01
4.00911242e-01 4.69227254e-01 7.70178363e-02 9.46875334e-01
-2.61420310e-01 2.51572400e-01 -6.36967719e-01 2.37740502e-01
-2.97862500e-01 3.21193874e-01 -1.53588096e-03 6.06678605e-01
5.53398393e-02 7.18966842e-01 -1.09783864e+00 -7.24715352e-01
7.70761892e-02 5.93176484e-01 -2.91508049e-01 -1.68900117e-02
-3.62366349e-01 2.73043573e-01 -8.25250566e-01 4.27627623e-01
4.44220155e-01 -3.42316657e-01 7.01020181e-01 -2.84074008e-01
-6.44057319e-02 4.18723226e-02 -9.16279733e-01 1.91819859e+00
-2.02369288e-01 3.79136920e-01 -4.41638172e-01 -1.46487808e+00
9.34566677e-01 4.83399332e-01 7.15238690e-01 -6.95785508e-02
1.52780995e-01 3.86351824e-01 3.01824813e-03 -6.28242612e-01
-1.18829049e-01 3.54688540e-02 3.17856856e-03 2.81686693e-01
2.14537606e-01 7.69724920e-02 4.81347471e-01 1.37368754e-01
1.42361414e+00 -4.78955731e-02 9.43453610e-01 -4.29884166e-01
9.37874973e-01 1.54057279e-01 2.44459420e-01 3.81146222e-01
2.43704051e-01 1.33557200e-01 6.70784056e-01 -4.86435324e-01
-7.08651721e-01 -9.32343781e-01 -4.06819701e-01 5.74100554e-01
-2.72241514e-02 -6.53494895e-01 -6.62153780e-01 -6.74875259e-01
8.47200304e-02 3.27885181e-01 -7.23152578e-01 -6.97911605e-02
-2.11413011e-01 -1.12476134e+00 6.58158302e-01 -2.22371161e-01
2.79776324e-02 -9.07496333e-01 -5.51533163e-01 3.90192747e-01
3.05223893e-02 -9.67072546e-01 -3.69765386e-02 3.83082896e-01
-7.63492942e-01 -1.50314105e+00 -6.80214226e-01 -7.21112311e-01
9.32119370e-01 -1.69131339e-01 9.08986688e-01 5.99346533e-02
-8.86217415e-01 3.19889490e-03 -3.47881675e-01 -5.64293623e-01
-5.58439016e-01 -7.27877319e-02 6.64668605e-02 -6.85671046e-02
5.64893663e-01 -6.27829850e-01 -4.61384207e-01 8.31325203e-02
-1.06291008e+00 -8.42238814e-02 7.05671549e-01 9.46156204e-01
6.79302096e-01 8.69424641e-03 8.60618830e-01 -1.11192954e+00
5.07592082e-01 -7.52354860e-01 -5.58622479e-01 3.52963358e-01
-6.06754065e-01 4.46315438e-01 6.20737433e-01 -2.00555936e-01
-7.09057808e-01 3.47679973e-01 -2.50561148e-01 3.09700131e-01
-2.89868563e-01 1.04365313e+00 -2.02875271e-01 -1.59129322e-01
6.88486338e-01 2.87151664e-01 1.29355565e-01 -4.73813474e-01
2.66261220e-01 6.64175808e-01 3.75767469e-01 -3.46605718e-01
4.38467830e-01 4.72252071e-01 5.78262925e-01 -9.84925926e-01
-3.89359057e-01 -8.30881357e-01 -7.11199760e-01 2.11372524e-01
9.55623686e-01 -5.02191305e-01 -1.01905680e+00 8.36915299e-02
-1.18313003e+00 4.10322785e-01 -1.66630950e-02 6.03416026e-01
-3.64073575e-01 8.17491233e-01 -2.60217220e-01 -3.51378202e-01
-4.70409244e-01 -1.00214803e+00 9.88688052e-01 -2.24833116e-01
-6.37813509e-01 -1.11902606e+00 4.97152120e-01 1.60715640e-01
-1.41447127e-01 5.26022673e-01 1.46273184e+00 -1.20979738e+00
-1.41795725e-01 -3.07910979e-01 -1.10462405e-01 -4.11538221e-02
5.30381322e-01 -1.76600162e-02 -6.24840200e-01 7.98601955e-02
-2.54028141e-01 8.58360529e-02 8.07432413e-01 -1.25026211e-01
1.02043772e+00 -2.15815336e-01 -8.82612348e-01 1.34171709e-01
1.54281175e+00 3.07924151e-01 4.95560765e-01 1.34908348e-01
4.47666228e-01 7.42000163e-01 4.93999720e-01 3.16514105e-01
-1.81926519e-01 6.82893991e-01 6.44319355e-02 -2.35975124e-02
3.18749011e-01 1.68200672e-01 -1.25335440e-01 8.71933877e-01
-3.70617926e-01 -4.01582122e-01 -9.53216195e-01 6.04931056e-01
-1.81285894e+00 -8.91233742e-01 -1.24787182e-01 2.33662534e+00
1.37116539e+00 6.05433211e-02 -1.40632465e-01 1.37487710e-01
4.53040957e-01 -4.13270414e-01 -3.24240506e-01 -3.94720674e-01
-1.71397671e-01 7.13372469e-01 4.97123301e-01 3.85803401e-01
-7.70927906e-01 6.77526593e-01 6.31346083e+00 8.51946175e-01
-9.66427028e-01 -1.76505476e-01 2.51993448e-01 2.22332582e-01
-1.84878811e-01 -1.47360358e-02 -4.65213299e-01 5.04447162e-01
1.18532252e+00 -6.14440322e-01 -1.94941491e-01 6.45815492e-01
2.40173221e-01 -2.61261046e-01 -1.17634511e+00 8.14958155e-01
-2.90045179e-02 -1.70324636e+00 1.44393578e-01 5.23212701e-02
3.61634731e-01 -1.43696219e-01 -2.99598038e-01 -3.80105853e-01
3.51646505e-02 -1.01195276e+00 -1.65232927e-01 8.97017002e-01
6.06730759e-01 -7.27782071e-01 9.04309452e-01 1.43956259e-01
-9.34343636e-01 4.08272773e-01 -3.55897248e-01 2.02043995e-01
-1.87160164e-01 9.66240644e-01 -1.20654988e+00 9.69125390e-01
3.43564063e-01 8.57969999e-01 -3.91473472e-01 9.13441241e-01
1.38537899e-01 4.15326416e-01 -2.51135468e-01 -4.86774027e-01
1.97242759e-02 -1.84711397e-01 5.24455428e-01 1.54409981e+00
3.41180950e-01 1.02069020e-01 -3.55136581e-02 4.31780249e-01
6.54333606e-02 8.10722709e-01 -7.26419270e-01 -4.78828907e-01
2.29206666e-01 1.19571960e+00 -1.03313184e+00 -4.83257711e-01
-3.26989561e-01 1.00361061e+00 1.54493913e-01 1.32686406e-01
-4.38019544e-01 -7.88580477e-01 6.82081640e-01 -1.46972731e-01
1.82776988e-01 1.83326285e-02 1.60458107e-02 -9.62912381e-01
-2.25907475e-01 -7.98112214e-01 3.70297462e-01 -3.46018732e-01
-1.23647857e+00 5.42771339e-01 -3.04539762e-02 -1.09425437e+00
-1.41212448e-01 -8.92628014e-01 -1.18059419e-01 7.81734169e-01
-1.23607302e+00 -6.38960779e-01 9.44097042e-02 3.35987598e-01
1.67332605e-01 -2.77278006e-01 1.50859785e+00 3.38094324e-01
-3.69385928e-01 2.33514637e-01 6.16310418e-01 -2.51727283e-01
9.59185839e-01 -1.10750008e+00 3.42013203e-02 1.01728946e-01
8.17115232e-02 1.00792170e+00 1.02880085e+00 -6.84736013e-01
-1.46751046e+00 -8.44701767e-01 1.27333331e+00 -4.33416128e-01
8.19082379e-01 -4.62928832e-01 -9.51819181e-01 1.01509757e-01
8.32103193e-02 -2.16615751e-01 1.32596004e+00 -5.56446873e-02
-3.68348598e-01 1.88700706e-01 -1.23594916e+00 6.00310028e-01
9.09948707e-01 -4.33088243e-01 -7.82146692e-01 7.20678210e-01
5.33735335e-01 6.07318245e-02 -1.24277270e+00 1.72286361e-01
5.21350145e-01 -4.77996141e-01 1.06842172e+00 -7.34246492e-01
1.95855737e-01 -4.08116221e-01 1.23426020e-01 -1.09532166e+00
2.44081132e-02 -4.76173609e-01 1.71377212e-01 7.70599842e-01
6.24443948e-01 -8.56060207e-01 5.76143503e-01 1.00916579e-01
8.53667855e-02 -8.14668238e-01 -1.15083969e+00 -3.90143454e-01
-1.31761804e-01 -1.31830443e-02 3.16543341e-01 1.16634750e+00
6.90921128e-01 3.57304215e-01 2.72151381e-02 -7.31810555e-02
3.98036212e-01 2.75162142e-02 4.82097298e-01 -1.73786294e+00
-3.77459437e-01 -3.97624701e-01 -1.00688314e+00 -6.63275048e-02
2.08989725e-01 -1.12770283e+00 -1.97244406e-01 -1.43699467e+00
2.68621504e-01 -5.02735563e-02 -5.29583693e-01 5.44295311e-01
3.76590751e-02 2.32110709e-01 -6.00281298e-01 1.08961791e-01
-3.83899152e-01 7.80944005e-02 6.71537340e-01 -2.15585992e-01
2.53856182e-02 -4.08695459e-01 -6.88999653e-01 8.23845625e-01
5.07224381e-01 -5.98085582e-01 -3.29553813e-01 4.84844685e-01
5.23855567e-01 -2.26499140e-01 1.42026961e-01 -7.88733304e-01
2.89840728e-01 -1.84416279e-01 3.29663068e-01 -3.75894189e-01
4.74479347e-02 -1.00547254e+00 5.62151551e-01 7.57087111e-01
-3.12039644e-01 -2.38820836e-01 2.94473052e-01 8.53497803e-01
-1.50094613e-01 -4.18488473e-01 4.86621946e-01 -1.06422119e-01
-5.28690696e-01 -5.74049391e-02 -5.58850765e-01 -3.31300199e-01
1.19654584e+00 -9.57918018e-02 -1.90110616e-02 2.83672333e-01
-9.04838681e-01 -2.53477186e-01 3.88693094e-01 1.23827368e-01
4.61971670e-01 -7.56226182e-01 -5.30314207e-01 -7.93096870e-02
4.91895646e-01 -5.46370506e-01 8.41618702e-02 7.09230661e-01
-6.89784229e-01 6.24058425e-01 -1.30047679e-01 -4.13273364e-01
-1.65894902e+00 8.20928395e-01 -7.93460682e-02 -4.82379913e-01
-5.32352030e-01 4.18809563e-01 1.07630573e-01 -1.78408071e-01
-3.06458808e-02 -3.01235139e-01 -6.15301847e-01 3.50757182e-01
4.97940689e-01 2.51910202e-02 2.50768870e-01 -5.29426575e-01
-6.03049576e-01 7.03475714e-01 -2.31114119e-01 1.34751216e-01
1.33165371e+00 2.52718955e-01 -5.04604518e-01 4.80682611e-01
1.35059798e+00 2.28288770e-01 -6.65067732e-02 -1.43750198e-02
4.33556527e-01 -8.83981660e-02 -3.74619842e-01 -8.75689745e-01
-4.88784671e-01 3.30658555e-01 3.68316829e-01 -1.45830408e-01
1.01445663e+00 1.09888673e-01 3.72935265e-01 9.27122116e-01
3.14287782e-01 -8.39690924e-01 -3.64667416e-01 2.24372745e-01
6.17793679e-01 -9.31365490e-01 2.86552012e-01 -6.48558736e-01
-1.99085161e-01 1.20397472e+00 -2.58417577e-01 1.95877403e-01
7.14429438e-01 3.22764903e-01 -2.05688760e-01 -4.71548349e-01
-6.80964351e-01 -3.49348709e-02 3.09414089e-01 4.58638579e-01
7.96041965e-01 -2.16415599e-01 -1.20773685e+00 5.17506719e-01
2.14894742e-01 1.26208022e-01 2.06115782e-01 9.96318161e-01
-4.34496582e-01 -1.57492507e+00 -5.27179278e-02 4.66835737e-01
-8.07624340e-01 -3.67040813e-01 -6.81277514e-01 6.35788143e-01
1.77303255e-01 7.06858039e-01 -3.54617029e-01 -1.13724833e-02
1.23122349e-01 2.66460508e-01 2.91006207e-01 -6.75529420e-01
-4.11088228e-01 -1.40289426e-01 2.00052842e-01 -2.14048624e-01
-6.71603262e-01 -6.64869547e-01 -1.59193671e+00 7.30326250e-02
-4.16128367e-01 7.86421299e-01 9.23581600e-01 1.05595636e+00
6.64048314e-01 5.03751814e-01 2.24794149e-01 -2.48661935e-01
-2.79591642e-02 -6.58008397e-01 -3.78981411e-01 3.00644308e-01
-1.10463724e-02 -6.82671547e-01 -1.97685301e-01 4.03367817e-01] | [8.330918312072754, 8.558834075927734] |
8c57bed5-7c68-4bf9-bce8-c3b38acef80b | improving-continual-relation-extraction-by | 2305.06620 | null | https://arxiv.org/abs/2305.06620v1 | https://arxiv.org/pdf/2305.06620v1.pdf | Improving Continual Relation Extraction by Distinguishing Analogous Semantics | Continual relation extraction (RE) aims to learn constantly emerging relations while avoiding forgetting the learned relations. Existing works store a small number of typical samples to re-train the model for alleviating forgetting. However, repeatedly replaying these samples may cause the overfitting problem. We conduct an empirical study on existing works and observe that their performance is severely affected by analogous relations. To address this issue, we propose a novel continual extraction model for analogous relations. Specifically, we design memory-insensitive relation prototypes and memory augmentation to overcome the overfitting problem. We also introduce integrated training and focal knowledge distillation to enhance the performance on analogous relations. Experimental results show the superiority of our model and demonstrate its effectiveness in distinguishing analogous relations and overcoming overfitting. | ['Wei Hu', 'Yuanning Cui', 'Wenzheng Zhao'] | 2023-05-11 | null | null | null | null | ['relation-extraction', 'continual-relation-extraction'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.68823709e-02 2.83808947e-01 -6.25719726e-01 -3.10465962e-01
-1.83256105e-01 -1.07488692e-01 4.63415772e-01 2.26114035e-01
-3.80611271e-01 9.89244223e-01 1.03029914e-01 -3.84621650e-01
-4.96515781e-01 -9.16379750e-01 -4.94246453e-01 -3.32727611e-01
-9.79441963e-03 4.49130803e-01 2.99105257e-01 -4.24037009e-01
1.14668626e-02 4.47301149e-01 -1.16297221e+00 8.50339606e-02
1.24397814e+00 6.44276857e-01 -6.61968440e-02 1.51149690e-01
-9.47333947e-02 1.14619493e+00 -7.59353459e-01 -5.62020004e-01
-1.51459510e-02 -2.27419645e-01 -1.19333422e+00 -3.84808220e-02
-1.34288400e-01 -4.05458301e-01 -7.68693089e-01 7.88731575e-01
2.37846911e-01 4.42950100e-01 1.96955264e-01 -1.19937801e+00
-1.28305423e+00 1.00069678e+00 -5.21063268e-01 6.27430618e-01
3.00000578e-01 -1.18117817e-01 8.34296823e-01 -1.18725669e+00
3.61910999e-01 1.11255765e+00 8.46138358e-01 4.41667140e-01
-1.18015873e+00 -1.00714135e+00 5.67779779e-01 4.49669808e-01
-1.74773300e+00 -6.47637546e-01 8.36285591e-01 1.49596497e-01
1.34185934e+00 1.98559597e-01 5.48291743e-01 1.05470645e+00
-5.83376735e-02 7.72459686e-01 7.33161330e-01 -5.63139081e-01
-9.76002291e-02 5.34982122e-02 6.53931558e-01 5.45634091e-01
6.82707548e-01 5.68362251e-02 -7.08438396e-01 -1.19587094e-01
7.23937571e-01 1.91217110e-01 -4.05175269e-01 3.24970717e-03
-8.75272512e-01 3.66347492e-01 4.97665554e-01 4.28370774e-01
-1.19425200e-01 -3.47913325e-01 3.07510018e-01 6.71024621e-01
4.68839586e-01 6.86695933e-01 -7.18864679e-01 1.02126569e-01
-4.49579954e-01 -7.44672567e-02 6.07880235e-01 1.40457129e+00
8.58032644e-01 -1.31830230e-01 -3.10625166e-01 1.13002658e+00
-2.42138520e-01 9.44754481e-02 7.86097109e-01 -5.54936290e-01
6.17193222e-01 1.15535557e+00 -7.97639564e-02 -9.21018660e-01
-5.02333045e-01 -5.91118217e-01 -1.10031343e+00 -5.76758802e-01
6.66720718e-02 1.38431400e-01 -8.36711645e-01 1.72122765e+00
1.53893039e-01 3.73075515e-01 2.08961785e-01 5.29547989e-01
8.16802323e-01 3.01115781e-01 2.08235562e-01 -6.69716537e-01
1.13867116e+00 -1.22159863e+00 -1.05740559e+00 -5.13757348e-01
9.41143394e-01 -4.63794649e-01 1.40864086e+00 6.27394393e-02
-8.35631132e-01 -6.36529267e-01 -1.20983183e+00 -2.27756768e-01
-3.84372085e-01 1.00921758e-01 1.14559352e+00 4.07125235e-01
-4.62386400e-01 6.99853599e-01 -8.57711077e-01 -2.96191573e-01
4.41863120e-01 5.18487275e-01 -2.73819447e-01 2.63868701e-02
-1.64358163e+00 1.17398822e+00 8.31851900e-01 2.82154292e-01
-2.64334202e-01 -8.26887071e-01 -7.86677122e-01 -1.44278221e-02
8.77608180e-01 -6.84939861e-01 1.08906031e+00 -2.21670732e-01
-1.14322734e+00 5.22337973e-01 -2.78587550e-01 -5.70834160e-01
7.96487033e-02 -7.83963382e-01 -9.76537585e-01 -4.06436503e-01
-1.67329758e-01 8.49223137e-02 4.08799857e-01 -1.08319747e+00
-3.85888159e-01 -1.62277579e-01 2.22146109e-01 1.34204611e-01
-9.72438395e-01 -3.49676520e-01 -4.89434570e-01 -9.02709961e-01
3.98602158e-01 -6.08839273e-01 -1.01286292e-01 -6.22258902e-01
-5.48294783e-01 -4.81561035e-01 8.53594780e-01 -4.10739392e-01
1.95688117e+00 -2.11137938e+00 -1.70821488e-01 -5.36321588e-02
4.48733866e-01 6.01512671e-01 -1.79202959e-01 2.68079400e-01
-1.68371394e-01 6.97028637e-02 -1.39571846e-01 -2.46076822e-01
-5.52671134e-01 6.37006223e-01 -5.66130877e-01 -5.59004806e-02
4.56124842e-01 1.17179358e+00 -1.06075108e+00 -5.41656315e-01
-1.98402599e-01 4.29135025e-01 -1.69259682e-01 2.79058605e-01
1.61767248e-02 1.42821580e-01 -2.91248888e-01 6.96891308e-01
6.72407806e-01 -4.76036221e-01 4.04496878e-01 -2.92096674e-01
4.62561429e-01 6.25246882e-01 -9.74798977e-01 1.60096085e+00
-4.83717889e-01 2.97106147e-01 -8.27867091e-01 -8.33494544e-01
1.31873071e+00 1.19405404e-01 1.34033293e-01 -8.41816366e-01
-2.66958363e-02 4.08420302e-02 1.71979085e-01 -4.77322042e-01
7.85007536e-01 -1.35347143e-01 2.96260506e-01 3.76663953e-01
3.67856100e-02 3.01080763e-01 5.83898537e-02 3.36905301e-01
1.15497482e+00 1.13520429e-01 5.71904719e-01 8.89059529e-02
4.61631089e-01 -2.76477039e-01 1.03447282e+00 7.38027513e-01
-1.58913404e-01 2.04059437e-01 3.33696187e-01 -5.31156182e-01
-5.71303070e-01 -8.73284519e-01 6.68673813e-02 9.20682073e-01
3.71226966e-01 -1.00842977e+00 -5.59890270e-02 -9.96710896e-01
-3.77667719e-03 9.51642454e-01 -6.60303831e-01 -1.02989554e+00
-8.39805663e-01 -1.09423923e+00 5.12455285e-01 9.76879954e-01
7.83953786e-01 -9.42170560e-01 3.06208432e-02 1.94604337e-01
-3.56052965e-01 -1.12624609e+00 -2.71702439e-01 1.88132361e-01
-1.10809302e+00 -9.51278210e-01 -8.49825069e-02 -7.91613102e-01
7.14814186e-01 6.24429941e-01 1.31413639e+00 4.13842291e-01
1.94295570e-01 -7.90786371e-02 -4.49519038e-01 -1.95830151e-01
-2.29265749e-01 6.16078079e-01 3.09467286e-01 -4.14770305e-01
8.03997338e-01 -1.14497900e+00 -4.80574667e-01 3.16941381e-01
-7.60328770e-01 -1.17755793e-01 7.73661196e-01 9.78575170e-01
3.21893483e-01 3.29679728e-01 8.85789454e-01 -1.27728271e+00
1.04575551e+00 -5.09197235e-01 -4.77473661e-02 7.32700348e-01
-1.28904164e+00 3.09703797e-01 5.45746803e-01 -1.03963625e+00
-1.20540953e+00 -2.85069764e-01 3.93034369e-01 -3.98004830e-01
2.09573165e-01 5.35074174e-01 -2.70759583e-01 1.14358202e-01
6.89169586e-01 1.09573916e-01 -3.94928604e-01 -5.41081190e-01
4.61693436e-01 4.37194139e-01 7.86230862e-01 -7.68114328e-01
7.16937542e-01 1.18419476e-01 -1.87127382e-01 -3.45804960e-01
-1.38899994e+00 -1.78670064e-01 -5.87326825e-01 2.72264838e-01
-8.28649625e-02 -8.51649940e-01 -6.24328494e-01 5.57725057e-02
-1.17182291e+00 -1.32008702e-01 -6.32771075e-01 3.39392126e-01
-7.14402646e-02 3.67390662e-01 -8.67530465e-01 -8.65431368e-01
-5.80337048e-01 -4.92230237e-01 3.92581522e-01 3.95584226e-01
-5.91037393e-01 -8.70985389e-01 1.10156126e-01 6.94106845e-03
2.46529609e-01 -2.40182713e-01 1.10778272e+00 -8.19068611e-01
-3.69369745e-01 -1.48384228e-01 -3.96161228e-01 1.17371507e-01
5.71124494e-01 -2.77693331e-01 -9.29899752e-01 -7.82376230e-02
-1.49896380e-03 -3.31090808e-01 1.16042578e+00 -4.57869709e-01
8.74240398e-01 -3.95642221e-01 -6.43008530e-01 4.01112109e-01
8.55771363e-01 4.22206908e-01 7.43329406e-01 4.01561558e-01
8.03837538e-01 3.51295024e-01 8.35761189e-01 1.87001884e-01
6.18563354e-01 4.30075765e-01 -2.08831742e-01 1.22345142e-01
-3.15969259e-01 -5.85548520e-01 2.09482908e-02 1.12074494e+00
-2.06223920e-01 1.05991736e-01 -8.86733532e-01 6.76507175e-01
-2.10532904e+00 -5.36048412e-01 1.40992165e-01 2.12717795e+00
1.54883015e+00 5.71788073e-01 -1.40866488e-01 5.95610619e-01
5.53043067e-01 -2.00857192e-01 -5.55703521e-01 -2.15563625e-02
-4.24209207e-01 4.84095424e-01 -6.76333681e-02 6.33648694e-01
-9.02663469e-01 1.39872360e+00 6.21675491e+00 8.30171049e-01
-7.12497771e-01 5.43696620e-02 4.58774209e-01 5.12203090e-02
-2.81692326e-01 3.79613698e-01 -8.99952173e-01 3.39179225e-02
7.10987508e-01 -2.81065822e-01 2.42883548e-01 6.99455380e-01
-4.64498609e-01 -7.98895955e-02 -1.16125846e+00 9.45707262e-01
-1.99114040e-01 -1.15205240e+00 3.33905905e-01 -4.41405356e-01
5.57138860e-01 -7.31342018e-01 -8.40026215e-02 7.96882570e-01
4.13613498e-01 -8.35119843e-01 2.89182970e-03 7.82375515e-01
5.46744525e-01 -8.21110249e-01 6.13851190e-01 1.91435814e-01
-1.12514985e+00 -4.04169373e-02 -4.54104513e-01 -4.45391655e-01
-1.19231664e-01 8.48619044e-01 -7.53656030e-01 7.97807634e-01
3.97819817e-01 7.19740331e-01 -1.03926897e+00 7.20127642e-01
-5.23748279e-01 5.40160358e-01 -2.41429329e-01 1.40134349e-01
-5.05230248e-01 1.73001781e-01 3.04303050e-01 1.01511323e+00
6.91661239e-02 3.15141112e-01 -7.35640526e-02 7.08685696e-01
-1.32720903e-01 -5.76450340e-02 -5.23828804e-01 -5.83442934e-02
1.03884065e+00 9.31823313e-01 -6.00677848e-01 -4.43788290e-01
-3.35715264e-01 9.38076079e-01 1.01241112e+00 5.49003243e-01
-5.42662203e-01 -5.92497528e-01 3.99111390e-01 -3.10624763e-02
2.16272697e-02 -3.33361983e-01 -5.33519745e-01 -1.31871343e+00
2.84762919e-01 -5.09954691e-01 6.48853898e-01 -5.84366441e-01
-1.32759297e+00 6.68828070e-01 9.86383632e-02 -7.93252587e-01
6.74820738e-03 5.56700788e-02 -3.54913682e-01 5.42061687e-01
-1.30214882e+00 -1.07516086e+00 -3.25904697e-01 4.66221273e-01
2.62824208e-01 -9.37544480e-02 9.41101491e-01 5.23609817e-01
-1.12149930e+00 1.11238885e+00 -5.51183999e-01 1.12794578e-01
8.20632458e-01 -9.23470974e-01 4.70601529e-01 8.19011092e-01
2.63524652e-01 1.40620673e+00 4.73495036e-01 -1.00994253e+00
-1.03707218e+00 -1.08379638e+00 1.24819636e+00 -4.94370729e-01
8.09570193e-01 -2.01359063e-01 -1.51415348e+00 1.09061813e+00
-7.99203962e-02 8.02194700e-02 7.06724286e-01 8.33106279e-01
-8.64178479e-01 -3.93524855e-01 -7.84709811e-01 8.52781534e-01
1.66007674e+00 -5.60627043e-01 -1.02645910e+00 1.02942698e-01
1.05959427e+00 -2.20782220e-01 -1.06273925e+00 1.03102136e+00
3.65914375e-01 -6.73736155e-01 9.47000444e-01 -8.77566338e-01
2.17278197e-01 -2.18902931e-01 2.95071363e-01 -1.21543014e+00
-5.19996166e-01 -7.41273224e-01 -1.19695389e+00 1.72163773e+00
4.97758776e-01 -7.43691921e-01 7.20731795e-01 6.26187205e-01
2.82258004e-01 -9.97952342e-01 -5.54186285e-01 -1.08718419e+00
-8.05524737e-02 -3.11194867e-01 9.80124891e-01 1.33179963e+00
3.88207287e-01 1.15539765e+00 -5.81596375e-01 1.29921481e-01
2.22324625e-01 2.51613140e-01 7.14227378e-01 -1.11521840e+00
-3.07246447e-01 -1.72842830e-01 -6.49576709e-02 -1.22533178e+00
1.26583487e-01 -7.41905749e-01 -3.01450908e-01 -1.17013168e+00
3.29348713e-01 -7.99028754e-01 -7.93952942e-01 7.98859239e-01
-8.68077815e-01 -8.05023313e-02 -1.95147589e-01 5.10540485e-01
-7.83240020e-01 8.31032455e-01 1.32592797e+00 -4.47194912e-02
-5.48862219e-01 3.43196280e-02 -1.05512059e+00 4.10436153e-01
9.10053968e-01 -5.33571243e-01 -9.33491766e-01 -2.45045692e-01
4.02420044e-01 -2.95760572e-01 9.21344385e-02 -8.63616347e-01
4.64102775e-01 -4.31438535e-02 3.07670385e-01 -5.58519006e-01
1.89206824e-01 -5.98760068e-01 -8.56436118e-02 2.66997755e-01
-2.84715474e-01 -1.85513087e-02 3.36898446e-01 6.01131320e-01
-9.05122682e-02 6.21975912e-03 2.48065412e-01 3.01844656e-01
-6.06097460e-01 1.52905613e-01 1.35637000e-01 1.28758281e-01
7.48229086e-01 7.24503547e-02 -5.99502742e-01 -1.43032640e-01
-7.99707055e-01 3.15121710e-01 1.33632105e-02 6.52599692e-01
7.95657039e-01 -1.60946357e+00 -3.20622295e-01 3.58908713e-01
3.26928854e-01 2.78265327e-01 5.85104413e-02 7.31060863e-01
1.80751830e-01 2.41279066e-01 1.48284629e-01 -5.09128384e-02
-9.41514134e-01 1.12759280e+00 1.40601933e-01 -5.88842511e-01
-7.56181657e-01 9.25559878e-01 -1.71458662e-01 -2.54235595e-01
2.84540176e-01 -5.29121280e-01 -1.03342682e-01 -1.78401053e-01
6.61244333e-01 4.42813396e-01 2.38227919e-01 6.11675270e-02
-4.34567302e-01 1.84445798e-01 -8.48231137e-01 3.91928047e-01
1.10927856e+00 -1.81543425e-01 -2.07485244e-01 6.60187244e-01
6.27553940e-01 1.98939539e-04 -9.76672590e-01 -8.42031777e-01
5.87555110e-01 -2.43913710e-01 -4.03632730e-01 -6.50539398e-01
-9.24192369e-01 3.80282849e-01 8.51500705e-02 1.94677249e-01
1.31580257e+00 -6.44525737e-02 1.03990102e+00 9.06611502e-01
3.13404560e-01 -8.83968890e-01 1.28810018e-01 7.06221640e-01
7.93078840e-01 -1.11328208e+00 4.41073567e-01 -8.90332401e-01
-3.37723970e-01 9.32055175e-01 1.12699544e+00 9.53031480e-02
8.14124286e-01 5.79343081e-01 -1.00543179e-01 -1.42001927e-01
-9.44037080e-01 -1.89705074e-01 1.80489749e-01 6.29763722e-01
4.64977056e-01 -1.26008034e-01 -5.63274264e-01 1.00372005e+00
-2.62051195e-01 1.90192237e-01 8.34445208e-02 1.26299739e+00
-1.74652189e-01 -1.25719547e+00 4.08178940e-02 3.32822204e-01
6.48514628e-02 -3.74350369e-01 -6.31627381e-01 8.47988427e-01
1.41526937e-01 9.61104274e-01 -2.01137066e-01 -9.63213623e-01
7.04734325e-01 2.69468307e-01 6.15328312e-01 -5.38048208e-01
-5.27489543e-01 -3.36977839e-01 2.38204166e-01 -2.72447288e-01
-3.17877620e-01 -1.05715461e-01 -1.39290142e+00 -4.58168954e-01
-7.26272285e-01 3.21931839e-01 -4.25345451e-01 9.70976412e-01
6.24447942e-01 7.87439823e-01 2.99574822e-01 7.46941864e-02
-4.26930279e-01 -1.31660354e+00 -3.97821784e-01 4.36282933e-01
2.42705882e-01 -1.00645041e+00 -9.64533836e-02 -1.82655856e-01] | [9.17829418182373, 8.531557083129883] |
6d6c2f28-199c-4f0a-bd56-6c7e2b7fea5e | deep-reinforcement-learning-for-on-line | 2009.10321 | null | https://arxiv.org/abs/2009.10321v1 | https://arxiv.org/pdf/2009.10321v1.pdf | Deep Reinforcement Learning for On-line Dialogue State Tracking | Dialogue state tracking (DST) is a crucial module in dialogue management. It is usually cast as a supervised training problem, which is not convenient for on-line optimization. In this paper, a novel companion teaching based deep reinforcement learning (DRL) framework for on-line DST optimization is proposed. To the best of our knowledge, this is the first effort to optimize the DST module within DRL framework for on-line task-oriented spoken dialogue systems. In addition, dialogue policy can be further jointly updated. Experiments show that on-line DST optimization can effectively improve the dialogue manager performance while keeping the flexibility of using predefined policy. Joint training of both DST and policy can further improve the performance. | ['Xiang Zhou', 'Zhi Chen', 'Kai Yu', 'Lu Chen'] | 2020-09-22 | null | null | null | null | ['dialogue-management'] | ['natural-language-processing'] | [-3.68923217e-01 3.21246088e-01 -9.64549035e-02 -3.87115061e-01
-3.77824068e-01 -5.09457946e-01 7.30161369e-01 3.17035288e-01
-5.43165743e-01 8.65097225e-01 3.23471129e-01 -5.37531853e-01
4.64337543e-02 -5.34666359e-01 -5.77648692e-02 -3.68723005e-01
2.99710575e-02 6.82530642e-01 4.57409501e-01 -9.36877429e-01
3.51570517e-01 6.20639026e-01 -1.15142238e+00 1.65605903e-01
1.21035862e+00 5.27728319e-01 4.53907341e-01 1.00298727e+00
-5.19063532e-01 1.00787973e+00 -9.81728673e-01 1.68529063e-01
-1.58626810e-01 -5.31809688e-01 -1.30111301e+00 2.97106773e-01
-9.10516083e-02 -7.70868003e-01 -5.06177843e-01 5.46360433e-01
8.93478513e-01 4.67753083e-01 1.64073035e-01 -1.09106839e+00
1.20773472e-01 7.93162107e-01 -5.39687127e-02 5.14386483e-02
6.82548225e-01 3.22181761e-01 7.59163320e-01 -3.77739161e-01
1.96431458e-01 1.69507921e+00 2.45915055e-01 7.89837003e-01
-7.54094779e-01 -1.90933436e-01 4.59710360e-01 2.03845337e-01
-7.05007792e-01 -5.54232359e-01 7.21913040e-01 -1.07273728e-01
1.30538416e+00 5.35241552e-02 8.10936928e-01 6.82411015e-01
2.54241098e-02 1.31541407e+00 1.15968287e+00 -7.01374650e-01
3.58158559e-01 3.37951303e-01 1.66822001e-01 8.94603372e-01
-7.15351582e-01 -2.14625746e-01 -5.49808860e-01 -1.65624589e-01
7.77360022e-01 -5.59741437e-01 -2.57792175e-01 -3.63094449e-01
-9.20831382e-01 1.05184960e+00 1.37219606e-02 2.63486952e-01
-5.31636961e-02 -3.12467068e-01 6.90208435e-01 7.86982954e-01
2.45901018e-01 6.22958660e-01 -6.58949792e-01 -9.08261657e-01
-4.02978092e-01 2.47415110e-01 1.69614351e+00 6.70037508e-01
5.20884633e-01 3.90384436e-01 -5.13889551e-01 1.30580449e+00
3.24440688e-01 2.23331705e-01 5.87701499e-01 -1.10808802e+00
1.97096184e-01 6.40439630e-01 3.22278291e-01 -3.89016598e-01
-6.50674999e-01 1.59175724e-01 -5.44267535e-01 1.44162580e-01
4.46503162e-01 -7.71157742e-01 -3.28828484e-01 1.42021751e+00
6.08702481e-01 -7.96596110e-02 3.12349409e-01 7.63729811e-01
8.83050203e-01 9.88252580e-01 -1.74851298e-01 -3.79958600e-01
1.26633835e+00 -1.39624774e+00 -1.26548219e+00 -2.55929939e-02
1.00301218e+00 -7.79219151e-01 1.04534709e+00 3.83311898e-01
-1.05162561e+00 -3.57612312e-01 -7.86424994e-01 8.22778046e-02
-1.74299404e-01 4.37698811e-02 4.99998182e-01 6.64662600e-01
-1.05257916e+00 4.94351655e-01 -9.71964180e-01 -5.96041024e-01
-2.41850898e-01 5.22776306e-01 3.42629962e-02 3.04016829e-01
-1.57751584e+00 1.38422954e+00 3.99471581e-01 8.89079198e-02
-7.90525258e-01 -1.59950301e-01 -8.86578500e-01 2.85202079e-02
8.39391589e-01 -1.04030743e-01 2.16308618e+00 -3.91776145e-01
-2.87933755e+00 3.09553742e-01 -2.52287686e-02 -4.15430069e-01
4.92329925e-01 -4.34552550e-01 -1.99811250e-01 -2.09166054e-02
-2.84637511e-01 4.53979790e-01 6.07132137e-01 -9.92145777e-01
-8.11999559e-01 -5.78983575e-02 5.72631955e-01 7.88546681e-01
-4.07800287e-01 -8.47747177e-02 -4.57050145e-01 -1.15654722e-01
-3.57376993e-01 -9.64537621e-01 -3.95731628e-01 -3.26802790e-01
-1.96965724e-01 -7.14850962e-01 1.07897162e+00 -6.44037902e-01
1.41734862e+00 -1.62388456e+00 8.72874781e-02 -2.10178867e-01
-2.20143586e-01 8.64471555e-01 -1.00781426e-01 8.01942527e-01
4.58830744e-01 -2.79463261e-01 3.83957545e-03 -1.82717562e-01
2.52701193e-01 5.87211668e-01 1.01990020e-02 8.70945454e-02
1.27798662e-01 7.32956946e-01 -9.97748137e-01 -5.31246483e-01
5.42849422e-01 -1.67713970e-01 -6.47583008e-01 8.22514057e-01
-6.34920776e-01 7.40539134e-01 -5.15917420e-01 9.68574807e-02
3.35455716e-01 7.25388452e-02 4.27955508e-01 1.74950119e-02
-4.29728568e-01 6.14132345e-01 -1.14085138e+00 1.66496253e+00
-7.00411022e-01 5.27412057e-01 3.82475615e-01 -1.13149750e+00
1.00749528e+00 5.31143248e-01 5.07753134e-01 -7.05018044e-01
4.72561494e-02 -2.74548620e-01 3.55250239e-01 -7.53317416e-01
7.69223571e-01 2.04255894e-01 -3.11665684e-01 8.24316204e-01
3.14326048e-01 -3.89723837e-01 2.29963183e-01 1.61129668e-01
7.08052933e-01 5.00613935e-02 5.78888118e-01 -4.52634022e-02
8.07971418e-01 3.18224914e-02 4.41463798e-01 8.91593158e-01
-4.18278903e-01 -2.62085944e-01 7.53554761e-01 -1.57149717e-01
-8.68802369e-01 -3.72428775e-01 2.02276316e-02 1.48013115e+00
-1.59596279e-01 -4.78424162e-01 -1.04435074e+00 -7.95877218e-01
-6.35201409e-02 8.31792474e-01 -4.07510661e-02 -9.94915441e-02
-7.94870257e-01 -4.46567446e-01 5.43368101e-01 1.54013500e-01
9.97401118e-01 -1.20619714e+00 -4.97733355e-01 7.61927664e-01
-2.63578761e-02 -1.18034160e+00 -6.09266341e-01 2.42465064e-01
-6.82668507e-01 -7.43783474e-01 -5.76068103e-01 -7.63410091e-01
2.90467292e-01 1.51990578e-01 6.96431160e-01 1.55918866e-01
1.04276203e-01 5.48954248e-01 -4.43986297e-01 -3.99188042e-01
-1.05523908e+00 4.04095680e-01 4.82700355e-02 -2.99965709e-01
5.74693549e-03 -3.54993008e-02 -3.20586264e-01 4.74651307e-01
-5.15729845e-01 1.44703194e-01 1.16251886e-01 1.26541638e+00
-1.25888065e-01 1.35055184e-01 7.96529949e-01 -8.84640872e-01
1.36932576e+00 9.03105587e-02 -7.78271854e-01 4.89986807e-01
-6.85984492e-01 3.92937273e-01 6.79435670e-01 -2.75673419e-01
-1.50253856e+00 -5.84594570e-02 -4.29961354e-01 1.52458087e-01
-3.46094817e-01 6.14433527e-01 -1.03053756e-01 -2.13051483e-01
1.81908533e-01 2.89544553e-01 4.67198163e-01 -5.66159010e-01
2.33087182e-01 1.10041821e+00 1.19489945e-01 -6.89069390e-01
1.90501601e-01 -2.16513574e-01 -5.41547000e-01 -1.01663113e+00
-8.14544976e-01 -5.06269395e-01 -7.87401795e-01 -3.65975708e-01
5.77289939e-01 -7.00814903e-01 -1.14246678e+00 8.12168121e-01
-9.11395609e-01 -1.07036602e+00 6.95969090e-02 2.70065844e-01
-6.67442679e-01 7.60223985e-01 -8.65130901e-01 -1.15043950e+00
-3.28964621e-01 -1.29160559e+00 6.90581083e-01 5.78811944e-01
-3.33434790e-01 -1.34714305e+00 1.84536219e-01 3.75605315e-01
3.75424892e-01 -4.89056140e-01 6.45850956e-01 -9.81939197e-01
-1.45115495e-01 3.28699201e-02 2.16487542e-01 3.19342941e-01
4.25102144e-01 -6.42232150e-02 -7.81713963e-01 -5.81660807e-01
1.93094507e-01 -7.36852348e-01 1.89467922e-01 5.88680953e-02
7.94750452e-01 -5.29287517e-01 1.33705288e-01 1.78174023e-02
6.76593840e-01 6.13813460e-01 1.56840786e-01 6.00185812e-01
5.72121918e-01 7.02568293e-01 1.23682487e+00 1.00615060e+00
7.72775471e-01 9.90432322e-01 -1.06950132e-02 -2.04021372e-02
2.53518850e-01 -2.68007256e-02 7.31679142e-01 1.09603989e+00
4.66074318e-01 -2.53371626e-01 -1.12276852e+00 2.63086051e-01
-2.25929570e+00 -7.27847099e-01 3.26216668e-01 1.88953364e+00
1.30442381e+00 4.53320406e-02 2.95952857e-01 -8.49869102e-02
5.74151099e-01 2.89561421e-01 -5.37391484e-01 -1.06118810e+00
3.43315452e-01 4.68043573e-02 -2.91616712e-02 9.42516506e-01
-1.00364673e+00 1.30423892e+00 6.00497437e+00 6.49513602e-01
-1.15857339e+00 -6.83934614e-02 6.84253052e-02 3.10795993e-01
1.50069639e-01 7.81980343e-03 -9.38900769e-01 2.89636463e-01
9.82170701e-01 -3.66022587e-01 5.26687801e-01 7.35262692e-01
6.62393451e-01 -3.90302479e-01 -8.18575501e-01 6.66854560e-01
-2.95032650e-01 -8.29905272e-01 -2.10361451e-01 -9.54609364e-02
4.76416975e-01 -3.96442324e-01 -3.75232011e-01 8.66593659e-01
7.10509777e-01 -6.76648617e-01 3.18177044e-02 2.26228312e-01
2.41887018e-01 -1.05670607e+00 7.53500938e-01 8.77370417e-01
-8.11291456e-01 -1.31924689e-01 -2.25071713e-01 5.66078499e-02
1.82362303e-01 -5.68163916e-02 -1.58798552e+00 3.73644173e-01
3.46263081e-01 5.09992778e-01 -2.62927979e-01 9.79974270e-01
-3.16332877e-01 7.30232358e-01 -2.00052708e-01 -5.00029683e-01
5.16708970e-01 -2.00856611e-01 5.85824966e-01 1.24267828e+00
-1.47450760e-01 2.79696107e-01 8.15601766e-01 1.81071058e-01
2.60697037e-01 1.69131443e-01 -4.29096550e-01 -2.41503656e-01
5.21543980e-01 1.21864665e+00 -2.91379690e-01 -3.06200713e-01
-2.68078476e-01 9.71720397e-01 4.40451741e-01 1.30261183e-01
-5.43985069e-01 -2.69893259e-01 5.63602149e-01 -3.84027064e-01
3.09169948e-01 -3.89363140e-01 1.75393030e-01 -1.14129627e+00
-5.82221329e-01 -1.23518169e+00 2.13962317e-01 -3.90393823e-01
-9.58674610e-01 3.41492712e-01 -5.34688458e-02 -1.03800786e+00
-7.22437978e-01 -2.60577410e-01 -5.85903227e-01 7.68311679e-01
-1.56990659e+00 -6.20529413e-01 -1.31065756e-01 6.56751394e-01
8.96886826e-01 -5.20044208e-01 1.30686533e+00 -1.72482103e-01
-1.01724732e+00 7.06644714e-01 4.36284751e-01 1.87700659e-01
8.60789835e-01 -1.74157691e+00 1.40924126e-01 2.63523698e-01
-4.61309016e-01 3.83926481e-01 6.73152983e-01 -4.51977372e-01
-1.42043626e+00 -7.71225572e-01 4.82860744e-01 3.24643582e-01
7.52179682e-01 -1.77691087e-01 -1.12663758e+00 2.35940292e-01
8.30109775e-01 -7.85477221e-01 6.92140818e-01 1.68661624e-01
4.83342588e-01 1.24333054e-01 -8.96238387e-01 8.15507174e-01
1.18172109e-01 -5.78038335e-01 -6.12831712e-01 4.81779248e-01
6.98633611e-01 -8.96095097e-01 -1.48280728e+00 5.54974042e-02
1.61817461e-01 -7.25103676e-01 5.80088317e-01 -5.56475103e-01
-5.19024804e-02 6.98419139e-02 4.58306760e-01 -1.83622980e+00
3.43654118e-02 -1.14697874e+00 -3.34551871e-01 1.37934637e+00
5.05765621e-03 -6.01085305e-01 7.25044608e-01 7.03102887e-01
-2.81376481e-01 -7.74117112e-01 -8.05618703e-01 -5.22394359e-01
-2.45162006e-02 5.20416312e-02 4.72355962e-01 8.08980107e-01
4.94736195e-01 8.72268140e-01 -5.62064052e-01 4.50875089e-02
1.66432157e-01 6.11337535e-02 1.06078684e+00 -8.91367197e-01
-3.53388608e-01 -3.62653226e-01 2.41125017e-01 -1.54413605e+00
2.82182872e-01 -1.96571440e-01 4.41459507e-01 -1.62235975e+00
-5.31971753e-01 -4.00299668e-01 1.22265983e-02 5.00974774e-01
-2.86727428e-01 -1.10468197e+00 2.10628092e-01 -1.13274390e-02
-7.94352114e-01 1.09718227e+00 1.69812584e+00 -4.06741239e-02
-6.22740984e-01 4.24651265e-01 -3.03221792e-01 4.73815292e-01
1.23360419e+00 -1.71928570e-01 -6.65579677e-01 -4.52770703e-02
-5.44705629e-01 9.23862815e-01 -3.26646566e-01 -6.56508803e-01
5.66159189e-01 -3.39946181e-01 -3.44750226e-01 -6.68369710e-01
4.36237663e-01 -4.26380068e-01 -7.55954325e-01 4.48982388e-01
-6.03564084e-01 -1.82506889e-01 4.72570568e-01 3.29029739e-01
-3.31835330e-01 -5.80062330e-01 9.02295053e-01 -3.57157826e-01
-9.28710222e-01 -4.30495776e-02 -1.10502732e+00 5.81548968e-03
9.83867705e-01 -1.11395985e-01 -1.85418963e-01 -9.32123065e-01
-5.96522510e-01 1.00920141e+00 8.68558511e-02 2.87065536e-01
5.64647794e-01 -8.21722150e-01 -5.18144906e-01 -1.83925241e-01
-2.05727354e-01 1.92832500e-01 2.50014424e-01 6.29539549e-01
-4.59914446e-01 7.18869984e-01 -3.10572416e-01 -5.36640048e-01
-1.47130513e+00 6.44599274e-02 6.80943191e-01 -7.08134234e-01
-4.87907052e-01 5.29954851e-01 -2.33074084e-01 -1.09508574e+00
6.38607681e-01 -3.05198450e-02 -8.30788314e-01 1.64309576e-01
3.62988055e-01 2.54109383e-01 3.71022783e-02 -1.42492980e-01
1.34647414e-01 -9.57508311e-02 -5.20988822e-01 -4.07090992e-01
1.25738943e+00 -5.08199453e-01 -1.62446573e-01 6.73853219e-01
7.43825376e-01 -3.52542013e-01 -1.21575832e+00 -4.73426759e-01
1.85698152e-01 -1.58971101e-01 1.62741050e-01 -1.00670862e+00
-6.36376023e-01 8.27124655e-01 5.24617434e-01 4.08259183e-01
7.40869403e-01 -5.53125203e-01 8.43142033e-01 1.08994925e+00
3.35093468e-01 -1.64003634e+00 4.51140732e-01 1.17308986e+00
9.43180919e-01 -1.31976843e+00 -9.61365849e-02 -1.12228200e-01
-1.08802354e+00 1.31559956e+00 1.33957696e+00 2.92338908e-01
4.89711851e-01 2.66109437e-01 1.70216084e-01 5.15508056e-02
-1.23906851e+00 -1.86835334e-01 -9.64782760e-02 5.14517665e-01
5.53833544e-01 -2.22207308e-01 -2.69797653e-01 5.47619909e-02
-3.84638645e-02 -1.67578712e-01 7.59412885e-01 1.32187736e+00
-8.13226819e-01 -1.64719331e+00 -3.52614045e-01 3.16688955e-01
-1.07086457e-01 1.93063751e-01 -4.32267219e-01 7.17128217e-01
-6.55700922e-01 1.21207333e+00 -1.96329892e-01 -1.53505072e-01
6.35332108e-01 1.84963986e-01 3.87011319e-01 -8.22129846e-01
-1.08769047e+00 3.11008722e-01 7.07695663e-01 -2.58396626e-01
-2.73538291e-01 -4.62697655e-01 -1.43150902e+00 -3.32665831e-01
-5.08187711e-01 5.24162471e-01 3.73875558e-01 1.01944327e+00
1.03825174e-01 6.45999253e-01 1.01287186e+00 -3.69111210e-01
-1.09136891e+00 -1.20530272e+00 -4.88589793e-01 -1.01380914e-01
2.54251242e-01 -6.34820938e-01 7.53909945e-02 -4.60201800e-01] | [13.07951831817627, 8.028826713562012] |
52c1c168-19a5-4a96-a242-dd89a6af5137 | masked-pre-training-of-transformers-for | 2304.07434 | null | https://arxiv.org/abs/2304.07434v1 | https://arxiv.org/pdf/2304.07434v1.pdf | Masked Pre-Training of Transformers for Histology Image Analysis | In digital pathology, whole slide images (WSIs) are widely used for applications such as cancer diagnosis and prognosis prediction. Visual transformer models have recently emerged as a promising method for encoding large regions of WSIs while preserving spatial relationships among patches. However, due to the large number of model parameters and limited labeled data, applying transformer models to WSIs remains challenging. Inspired by masked language models, we propose a pretext task for training the transformer model without labeled data to address this problem. Our model, MaskHIT, uses the transformer output to reconstruct masked patches and learn representative histological features based on their positions and visual features. The experimental results demonstrate that MaskHIT surpasses various multiple instance learning approaches by 3% and 2% on survival prediction and cancer subtype classification tasks, respectively. Furthermore, MaskHIT also outperforms two of the most recent state-of-the-art transformer-based methods. Finally, a comparison between the attention maps generated by the MaskHIT model with pathologist's annotations indicates that the model can accurately identify clinically relevant histological structures in each task. | ['Saeed Hassanpour', 'Arief A. Suriawinata', 'Liesbeth Hondelink', 'Shuai Jiang'] | 2023-04-14 | null | null | null | null | ['whole-slide-images', 'multiple-instance-learning'] | ['computer-vision', 'methodology'] | [ 4.87950116e-01 4.77018237e-01 -4.14768338e-01 -1.77482635e-01
-1.17651200e+00 -2.63106376e-01 3.96714866e-01 4.76811200e-01
-3.17001939e-01 5.05587459e-01 3.01385909e-01 -2.22982943e-01
-9.71003622e-02 -5.67879558e-01 -6.34322822e-01 -1.01845467e+00
1.18499786e-01 3.04661423e-01 6.08573258e-01 9.05938596e-02
1.34475037e-01 3.40232462e-01 -1.26471758e+00 9.13780510e-01
8.22385609e-01 1.14289892e+00 5.56168616e-01 2.83670127e-01
-3.75497609e-01 1.02990890e+00 -5.17811179e-01 -3.53478819e-01
-2.20585674e-01 -4.40214455e-01 -8.72180343e-01 -1.08063325e-01
4.55354542e-01 1.09625712e-01 -6.19572103e-01 1.08071184e+00
4.21323776e-01 -2.84171730e-01 7.49717355e-01 -8.40798795e-01
-1.00153065e+00 6.39386952e-01 -7.37882257e-01 6.27685010e-01
-2.09868386e-01 1.11842565e-02 9.75160599e-01 -1.00088799e+00
9.09218729e-01 8.14016461e-01 6.94261670e-01 5.24236560e-01
-1.32332885e+00 -6.35139823e-01 3.40763211e-01 5.40104270e-01
-1.81221306e+00 -4.30952758e-01 6.46308005e-01 -3.11689526e-01
9.48660970e-01 5.55860698e-01 6.45638764e-01 9.05059874e-01
5.43353260e-01 1.04027557e+00 1.30611432e+00 -4.33700293e-01
5.29429354e-02 3.24712098e-01 9.53702778e-02 9.43375409e-01
-1.56052873e-01 -9.28759277e-02 -6.05734289e-01 -1.01039894e-02
7.71495402e-01 2.71621644e-01 -7.13185906e-01 -2.49145538e-01
-1.51113129e+00 5.83056390e-01 9.43501711e-01 5.75299799e-01
-6.90662414e-02 -5.06603047e-02 3.38966966e-01 9.52820554e-02
7.87858129e-01 2.90740401e-01 -8.08330160e-03 4.82540846e-01
-1.01140428e+00 -3.52971941e-01 5.20493202e-02 7.64262021e-01
4.72170472e-01 -2.47710377e-01 -6.53023422e-01 8.11360717e-01
1.08857222e-01 2.42782980e-01 7.95135379e-01 -1.91899151e-01
1.12227455e-01 1.00513697e+00 -3.67083967e-01 -8.73609245e-01
-4.69934464e-01 -8.63127947e-01 -1.18453574e+00 -3.82893085e-02
1.88505948e-01 8.28588545e-01 -1.26962125e+00 1.41787386e+00
1.00629754e-01 4.49821353e-01 -1.24410413e-01 8.02414596e-01
1.25821328e+00 4.36977118e-01 1.92495316e-01 -1.75549984e-01
1.41973233e+00 -1.03541350e+00 -8.06747496e-01 -2.72776067e-01
7.55819619e-01 -4.29861933e-01 1.17451525e+00 5.38248606e-02
-9.14445519e-01 -2.90651023e-01 -7.94950902e-01 -2.53456265e-01
-2.87102252e-01 4.96640086e-01 2.74020046e-01 1.34062290e-01
-1.32388091e+00 2.85057694e-01 -1.03891098e+00 -5.03741503e-01
1.02533615e+00 3.27781737e-01 -5.80836952e-01 -1.74116418e-01
-6.16016805e-01 8.22032928e-01 -1.06729545e-01 1.49970531e-01
-1.07917309e+00 -9.83155072e-01 -8.65853846e-01 3.62988085e-01
6.00513108e-02 -3.55265886e-01 1.11070693e+00 -8.49548578e-01
-1.06780112e+00 1.39151943e+00 -5.29970646e-01 -3.01794052e-01
4.40024376e-01 4.48023111e-01 -2.82298416e-01 4.23457831e-01
4.70212400e-02 6.91524684e-01 4.31067437e-01 -1.16983509e+00
-4.50360596e-01 -5.13113856e-01 -2.84469843e-01 1.11131199e-01
-7.39413798e-01 -2.64175624e-01 -7.16597855e-01 -7.94542849e-01
8.67697224e-02 -5.98611712e-01 -3.92105728e-01 5.36216676e-01
-6.53318763e-01 -4.76469696e-02 6.59045219e-01 -7.35678017e-01
1.22297895e+00 -2.33808374e+00 3.29273105e-01 2.48295650e-01
5.71255505e-01 1.05850972e-01 -2.83455908e-01 1.30642638e-01
-1.32570133e-01 1.36975959e-01 -2.27474794e-01 -4.92300421e-01
-2.77491689e-01 9.27606747e-02 -2.31932625e-01 6.45336509e-01
1.18791712e-02 1.26257944e+00 -8.54203403e-01 -8.52947891e-01
4.07000780e-02 4.72954601e-01 -2.52548486e-01 2.30565816e-01
1.67881340e-01 4.68289346e-01 -2.07247153e-01 7.37509668e-01
4.95393187e-01 -8.73744071e-01 3.67764443e-01 -5.18566430e-01
1.61665514e-01 1.94104314e-01 -2.28305623e-01 1.75449085e+00
-4.00191337e-01 7.79375017e-01 -3.89280282e-02 -1.17601824e+00
6.25749290e-01 3.74565154e-01 3.03297758e-01 -9.70722139e-01
-6.44180104e-02 6.54219985e-02 -9.40033197e-02 -5.90834260e-01
-1.98330022e-02 -3.83700252e-01 1.06803715e-01 2.34210983e-01
4.68793847e-02 3.98692667e-01 -1.42455891e-01 3.10112447e-01
1.31550801e+00 -3.40464026e-01 4.45303351e-01 -4.65921074e-01
4.68283385e-01 5.06614000e-02 7.27036357e-01 5.06048083e-01
-2.85625309e-01 8.14435780e-01 5.27669370e-01 -4.36510980e-01
-6.64477348e-01 -1.09468913e+00 -2.81351507e-01 8.52996111e-01
2.81021267e-01 -1.51453540e-01 -4.97455090e-01 -1.01942420e+00
-3.07253376e-02 2.76584744e-01 -1.28302109e+00 -3.59628648e-01
-4.38254863e-01 -7.70251274e-01 4.42619979e-01 7.56411493e-01
2.77914137e-01 -8.45581114e-01 -3.50976646e-01 -1.55885294e-02
-3.49009752e-01 -9.21257317e-01 -5.92926860e-01 2.12277159e-01
-6.92289591e-01 -1.11029363e+00 -1.03886271e+00 -1.29463291e+00
1.29453516e+00 5.89313507e-01 1.01102829e+00 3.98272365e-01
-8.10257673e-01 1.32302180e-01 -2.28474841e-01 -2.27280706e-01
-4.21272248e-01 1.47651494e-01 -5.77506304e-01 1.64127260e-01
2.32724831e-01 -3.84832561e-01 -9.13712442e-01 3.35074395e-01
-1.05929649e+00 4.51638430e-01 9.79643524e-01 1.20036852e+00
1.07691789e+00 -1.81322575e-01 4.46597755e-01 -1.23546827e+00
1.35733232e-01 -3.72149110e-01 -2.38523915e-01 7.09502637e-01
-2.79125750e-01 -2.97955889e-02 4.91849005e-01 -4.75493789e-01
-9.22099292e-01 -8.36154744e-02 -4.05416191e-02 -5.05185723e-01
1.95926607e-01 6.27014577e-01 5.31448759e-02 -3.51565808e-01
6.84088051e-01 5.23883164e-01 2.03682631e-01 -3.50582063e-01
-2.19345182e-01 5.05028605e-01 4.65902746e-01 1.81493059e-01
4.31448638e-01 7.91301131e-01 -1.92579642e-01 -4.84917223e-01
-1.10962164e+00 -4.56340522e-01 -4.64450777e-01 -8.45066831e-02
7.37037539e-01 -8.49238575e-01 -4.90070313e-01 3.15706462e-01
-7.72336066e-01 -5.42198718e-01 -2.60049909e-01 8.85736644e-02
-2.69807160e-01 2.35546276e-01 -8.92979145e-01 -2.83850968e-01
-3.73187214e-01 -1.17030954e+00 1.21049702e+00 4.87857424e-02
7.06449524e-02 -1.10771620e+00 -9.60402861e-02 2.11985826e-01
4.57431495e-01 1.29685193e-01 1.41181290e+00 -3.88954133e-01
-6.37709081e-01 -2.10353032e-01 -4.20793444e-01 -1.59439862e-01
3.36826473e-01 -3.70469928e-01 -1.03769624e+00 -5.50559640e-01
-3.70042622e-01 -2.04487696e-01 1.26040304e+00 4.26903874e-01
1.67932618e+00 -3.99062485e-01 -1.18865287e+00 6.76683426e-01
1.50566614e+00 -1.40627213e-02 6.45593345e-01 7.15256780e-02
5.51907599e-01 5.67520559e-01 4.39904690e-01 -3.82032469e-02
4.22862172e-01 6.82224512e-01 5.01933217e-01 -7.46781528e-01
-4.86223429e-01 -2.68689007e-01 -3.92135158e-02 7.20966518e-01
2.61030406e-01 -2.05044314e-01 -1.04963171e+00 9.38148439e-01
-1.77003384e+00 -8.88316751e-01 1.98417678e-01 2.08287787e+00
8.05390537e-01 -7.80959204e-02 -2.53090322e-01 9.31964442e-02
5.78886092e-01 6.53248504e-02 -5.60959518e-01 2.10476145e-01
-8.97405967e-02 1.79409221e-01 4.61633414e-01 2.83367127e-01
-1.01039577e+00 7.28838801e-01 6.43330240e+00 1.07886159e+00
-1.42287123e+00 4.18570429e-01 1.07917345e+00 -2.59933025e-01
-5.02110124e-01 -4.74000514e-01 -5.52946210e-01 3.48445654e-01
5.39814770e-01 -9.59315822e-02 9.35339183e-02 4.27016675e-01
-5.01706153e-02 3.60742025e-02 -1.07788146e+00 8.78465354e-01
2.94619411e-01 -1.90993071e+00 3.34044307e-01 2.35889554e-01
6.22540355e-01 1.17109539e-02 4.35897827e-01 2.00363934e-01
1.66173801e-01 -1.38390815e+00 5.86815298e-01 4.62229639e-01
1.27371085e+00 -4.30851310e-01 1.00596642e+00 1.94024175e-01
-1.16563022e+00 -1.34204298e-01 -4.93861705e-01 5.75022876e-01
-3.18289548e-01 5.00771344e-01 -1.04765296e+00 4.39337343e-01
7.65345991e-01 9.13900971e-01 -8.84476185e-01 1.27640772e+00
-1.60651840e-02 6.27903759e-01 1.58979133e-01 6.38700500e-02
1.38291791e-01 5.04876375e-01 1.56526804e-01 1.27251637e+00
4.83596057e-01 3.83191220e-02 1.42838806e-01 8.24440837e-01
1.23481089e-02 1.74048424e-01 -4.78799909e-01 4.07329313e-02
4.34091181e-01 1.39236188e+00 -1.01052260e+00 -2.14372948e-01
-4.02845740e-01 1.07166922e+00 7.79382467e-01 4.70277607e-01
-6.29981101e-01 -1.96503282e-01 4.35563505e-01 4.29082632e-01
4.14531440e-01 3.80246907e-01 -3.11718285e-01 -9.22885895e-01
-1.15097538e-01 -6.46885276e-01 4.85124260e-01 -6.27188027e-01
-1.21007657e+00 1.06896043e+00 -3.35823655e-01 -1.48195755e+00
2.37228498e-01 -6.79942608e-01 -5.68655431e-01 5.48966765e-01
-1.90953934e+00 -1.54646790e+00 -5.13494492e-01 4.58291829e-01
3.72212976e-01 -1.13605186e-01 1.07802689e+00 1.51896819e-01
-6.87362313e-01 9.19686437e-01 2.43897766e-01 1.60068989e-01
6.72625721e-01 -1.22617114e+00 5.97554632e-03 4.67424780e-01
1.18941672e-01 3.39410394e-01 3.74591857e-01 -3.03805798e-01
-9.89796221e-01 -1.32784200e+00 9.28732574e-01 -1.50762513e-01
4.42751616e-01 -5.33827603e-01 -1.19081807e+00 6.85698748e-01
3.79977286e-01 6.07388914e-01 1.14416575e+00 -2.07141861e-01
-3.91536683e-01 -3.05783570e-01 -1.22132421e+00 6.32126749e-01
1.10909617e+00 -4.70440328e-01 -9.90352705e-02 3.58337611e-01
4.59043205e-01 -5.19951344e-01 -8.49158704e-01 3.65317315e-01
2.27942541e-01 -8.25920403e-01 9.08432066e-01 -4.30839658e-01
4.69025254e-01 -2.57551104e-01 8.57088640e-02 -1.46329749e+00
-7.63106763e-01 -3.45121585e-02 2.08199397e-01 1.06199932e+00
5.15904009e-01 -6.22478008e-01 8.29384208e-01 1.41468763e-01
-3.68332058e-01 -1.09639299e+00 -1.33220732e+00 -4.53673065e-01
5.07172532e-02 1.23072065e-01 4.07452494e-01 8.45584989e-01
2.98531413e-01 7.76064321e-02 3.63563076e-02 2.31385574e-01
5.46971321e-01 3.65602970e-01 1.60196245e-01 -1.06900752e+00
-2.23696277e-01 -6.06922746e-01 -7.11378276e-01 -8.10214400e-01
2.74464101e-01 -1.45746052e+00 7.77396606e-03 -1.77298999e+00
7.25515604e-01 -6.01140440e-01 -8.92489076e-01 9.79743600e-01
-3.74999076e-01 8.43231320e-01 -1.08062021e-01 4.08259660e-01
-6.45576417e-01 6.02584064e-01 1.44118619e+00 -7.19784498e-01
3.18388343e-01 -3.65064889e-01 -9.94597495e-01 4.07252222e-01
4.00070608e-01 -3.90414447e-01 -4.04512674e-01 -5.30387700e-01
-1.17069863e-01 9.77564696e-03 5.65846026e-01 -8.54939938e-01
4.03431058e-01 -7.06628636e-02 5.99194109e-01 -4.71023768e-01
3.24833095e-01 -6.97124720e-01 -4.32977863e-02 5.50836265e-01
-5.92124999e-01 -1.55058160e-01 2.82886952e-01 7.34497190e-01
-5.12453377e-01 1.95122510e-01 8.30947101e-01 -1.54279530e-01
-5.61095953e-01 4.34800029e-01 -4.99717534e-01 -1.10473678e-01
1.17997491e+00 -3.28351259e-01 -6.79216981e-01 -1.52694685e-02
-7.45205224e-01 2.59424299e-01 4.04136032e-01 2.07044542e-01
9.83100057e-01 -1.28847587e+00 -7.36562252e-01 2.83107698e-01
5.90508640e-01 7.27726445e-02 7.13995934e-01 1.18246758e+00
-3.87499928e-01 5.31642437e-01 -1.06564909e-01 -8.08392286e-01
-1.56270635e+00 7.45141983e-01 4.41114396e-01 -7.91833758e-01
-9.02957320e-01 1.21797478e+00 1.07004404e+00 -2.40115896e-01
2.66882241e-01 -3.95327032e-01 -3.19952518e-01 -3.16943258e-01
8.02437663e-01 -2.20404997e-01 3.22517514e-01 -4.63724256e-01
-5.87020755e-01 4.97875571e-01 -6.54347360e-01 3.00932497e-01
1.39383173e+00 -4.93795425e-03 -5.52123450e-02 3.31052214e-01
1.17304623e+00 -1.04965113e-01 -1.22606182e+00 -7.53027022e-01
-9.94771495e-02 -4.07733053e-01 3.41935724e-01 -1.05291831e+00
-1.37866139e+00 1.06795788e+00 7.52943635e-01 -1.58136055e-01
1.34131491e+00 3.72417212e-01 5.28871000e-01 -1.52677596e-01
3.16365838e-01 -5.04769862e-01 2.09283561e-01 -1.31221622e-01
1.01311696e+00 -1.11956632e+00 -2.49843076e-01 -7.77216196e-01
-7.21843123e-01 7.64778733e-01 7.40505397e-01 1.49439424e-01
5.53596139e-01 4.92523551e-01 4.30234641e-01 -2.74605811e-01
-1.01964164e+00 -3.53999622e-02 5.62591136e-01 7.03873336e-01
4.74408895e-01 -2.87432950e-02 -5.48578426e-02 5.85817635e-01
2.03463361e-01 -5.53452261e-02 1.86638162e-01 7.05923080e-01
-2.35885635e-01 -7.75717735e-01 -1.45376012e-01 7.00837672e-01
-5.09888053e-01 -1.76960781e-01 -3.23771685e-01 6.33574247e-01
-2.06227690e-01 4.59438771e-01 3.18410426e-01 -2.66429007e-01
1.38906807e-01 -2.10562155e-01 4.83299077e-01 -6.91206396e-01
-6.69539988e-01 2.22041756e-02 -3.24623764e-01 -4.07704353e-01
-2.68157691e-01 -3.51412624e-01 -1.29003191e+00 1.64494842e-01
-2.39256367e-01 9.38975886e-02 9.75670516e-02 7.79685318e-01
5.14446855e-01 9.29776371e-01 4.22238111e-01 -6.29155755e-01
-3.06335241e-01 -8.56983244e-01 -5.78395426e-01 3.32518399e-01
4.82540637e-01 -6.99695587e-01 -8.28787126e-03 8.86286050e-03] | [15.068456649780273, -2.8323702812194824] |
72610613-7f17-4bfc-b5ec-4267a31ea022 | hiporank-incorporating-hierarchical-and | 2005.00513 | null | https://arxiv.org/abs/2005.00513v2 | https://arxiv.org/pdf/2005.00513v2.pdf | Discourse-Aware Unsupervised Summarization of Long Scientific Documents | We propose an unsupervised graph-based ranking model for extractive summarization of long scientific documents. Our method assumes a two-level hierarchical graph representation of the source document, and exploits asymmetrical positional cues to determine sentence importance. Results on the PubMed and arXiv datasets show that our approach outperforms strong unsupervised baselines by wide margins in automatic metrics and human evaluation. In addition, it achieves performance comparable to many state-of-the-art supervised approaches which are trained on hundreds of thousands of examples. These results suggest that patterns in the discourse structure are a strong signal for determining importance in scientific articles. | ['Andrei Mircea', 'Jackie C. K. Cheung', 'Yue Dong'] | 2020-05-01 | null | null | null | null | ['unsupervised-extractive-summarization'] | ['natural-language-processing'] | [ 3.66770267e-01 9.35814440e-01 -6.56399846e-01 -3.33956808e-01
-1.30815101e+00 -7.55535364e-01 8.56245816e-01 1.28375936e+00
-3.81456554e-01 9.97528195e-01 1.37407827e+00 -2.94180036e-01
-4.18570131e-01 -4.82079327e-01 -6.97789133e-01 -3.16107750e-01
-2.24018112e-01 5.84349632e-01 1.92036465e-01 -2.26186842e-01
1.19926763e+00 4.47716974e-02 -9.28365529e-01 3.76163006e-01
1.21522725e+00 8.82069543e-02 -1.30381003e-01 9.63816404e-01
-4.58415776e-01 1.08459556e+00 -1.17701292e+00 -5.81451952e-01
-4.31537837e-01 -4.36537236e-01 -1.15684485e+00 -3.10597152e-01
8.97784531e-01 1.21424101e-01 -3.30621898e-01 9.84608829e-01
5.76329947e-01 -3.69629450e-02 8.86358619e-01 -4.60084915e-01
-5.80633879e-01 1.22099066e+00 -7.71376371e-01 8.40123475e-01
8.73795569e-01 -4.58156765e-01 1.81330490e+00 -4.69504058e-01
1.31604600e+00 1.27585447e+00 4.11906749e-01 6.23329096e-02
-1.15875542e+00 -4.15767096e-02 2.27098018e-01 9.08436924e-02
-7.25827515e-01 -4.54456478e-01 7.37144887e-01 -5.19694448e-01
9.62282181e-01 5.14611006e-01 5.73279977e-01 9.36543286e-01
6.62011981e-01 6.18450522e-01 8.23385894e-01 -5.11732459e-01
1.15487732e-01 -5.76612234e-01 9.67023075e-01 6.40472353e-01
9.18675125e-01 -7.38633871e-01 -1.08069968e+00 -7.96290815e-01
-8.59310180e-02 -6.72193408e-01 -3.27495515e-01 2.85246432e-01
-1.49962890e+00 7.36902833e-01 1.19939372e-01 4.70906883e-01
-4.07445639e-01 1.07686073e-01 5.15846372e-01 1.01043701e-01
8.20949793e-01 1.21032405e+00 -1.01711713e-01 -2.63562292e-01
-1.23788571e+00 2.53221571e-01 1.10344362e+00 8.70422184e-01
2.32589677e-01 -4.09677923e-01 -7.61541128e-01 4.49426085e-01
-4.63417359e-02 -1.08029261e-01 2.41112143e-01 -1.16691720e+00
5.72877109e-01 7.52478957e-01 -1.30932555e-01 -1.21417665e+00
-4.27403897e-01 -7.14487016e-01 -6.65423989e-01 -4.78402942e-01
2.51271367e-01 7.75764789e-03 -5.29260039e-01 1.25776780e+00
1.91419423e-01 -2.30968371e-01 1.04520410e-01 4.13824588e-01
1.48129761e+00 7.00412035e-01 -3.29728276e-02 -5.96679032e-01
1.28490901e+00 -7.58926690e-01 -1.11172175e+00 1.54052861e-02
6.14959300e-01 -8.75214159e-01 7.71611810e-01 4.40515965e-01
-1.35905766e+00 -9.79053881e-03 -1.10761333e+00 -5.45810163e-01
1.16043594e-02 -6.62098676e-02 6.35892510e-01 1.13313444e-01
-1.26910889e+00 9.40588534e-01 -3.14664781e-01 -2.32125863e-01
5.76731145e-01 4.81229313e-02 -1.33939520e-01 1.16988011e-02
-1.01731694e+00 7.21090496e-01 2.79133916e-01 -2.88675785e-01
-4.07525867e-01 -9.31627810e-01 -6.70599341e-01 1.50391549e-01
1.89906403e-01 -9.35880423e-01 1.10293472e+00 -1.61456019e-01
-1.17980492e+00 8.54286969e-01 -4.30111349e-01 -4.36140388e-01
1.38286069e-01 -2.13831961e-01 7.70660564e-02 6.63917661e-01
4.11721051e-01 8.70521441e-02 3.11913699e-01 -9.93366480e-01
-2.55647063e-01 -1.95031375e-01 1.23952948e-01 3.70078087e-01
-4.85638171e-01 3.07510376e-01 -3.87971014e-01 -6.99428737e-01
-1.25428379e-01 -3.76154900e-01 -3.54310483e-01 -7.85452247e-01
-1.00453603e+00 -7.59300768e-01 1.12520441e-01 -8.09244931e-01
1.64306176e+00 -1.36950886e+00 5.55705369e-01 2.19693676e-01
8.06125283e-01 -5.45855343e-01 -3.12086083e-02 7.93882012e-01
2.58324474e-01 5.98750532e-01 -1.22320838e-01 -4.14773636e-02
-6.60407543e-02 -2.06765071e-01 -2.02022895e-01 4.15286571e-01
-1.80736229e-01 9.21512842e-01 -1.50758624e+00 -9.99896348e-01
-6.77659214e-01 -5.68928309e-02 -3.75770062e-01 2.06387341e-02
-3.30618203e-01 2.44743958e-01 -7.61627555e-01 5.06965458e-01
1.24817453e-01 -7.46325552e-01 6.25496745e-01 -1.37737617e-01
-4.54974398e-02 1.01673150e+00 -2.55661011e-01 1.62465799e+00
5.24184763e-01 9.25465167e-01 -1.38797507e-01 -8.66128922e-01
4.76825088e-01 1.30836353e-01 4.00418609e-01 -1.83424085e-01
-1.47284001e-01 3.29592638e-02 2.38205001e-01 -3.08444560e-01
7.88031876e-01 5.42358279e-01 -4.13098186e-01 6.20637119e-01
1.27598375e-01 -2.69733548e-01 9.40463364e-01 1.52417839e+00
1.65003335e+00 -2.84329861e-01 5.00373006e-01 -9.05169129e-01
3.74759406e-01 1.40067711e-01 3.62041473e-01 1.04596913e+00
2.22486153e-01 4.37850833e-01 1.34162521e+00 -2.60295626e-02
-9.02602613e-01 -6.40750885e-01 2.63211168e-02 1.11753726e+00
-7.12761432e-02 -1.36414444e+00 -8.76360059e-01 -7.41135001e-01
6.25454485e-02 9.58078146e-01 -8.58612776e-01 1.01763375e-01
-6.68673933e-01 -6.91690326e-01 4.51102704e-01 3.20017673e-02
-2.01634750e-01 -6.80146635e-01 -4.03112322e-02 4.78262194e-02
-3.84193987e-01 -1.05861819e+00 -7.10148394e-01 -1.58669308e-01
-1.03687370e+00 -1.28491950e+00 -5.08581460e-01 -6.30263031e-01
9.92604434e-01 2.31989369e-01 1.64738321e+00 2.53040582e-01
-1.41745865e-01 5.46826959e-01 -2.04657748e-01 -5.53977251e-01
-5.30949354e-01 4.45301563e-01 -2.37311274e-01 -7.06453085e-01
8.96973684e-02 -4.82120335e-01 -7.54977643e-01 -4.84610677e-01
-6.72905087e-01 -9.82491150e-02 6.60307288e-01 7.09291458e-01
4.56940830e-01 -5.22697628e-01 8.11837137e-01 -1.57173109e+00
1.43301725e+00 -5.26536882e-01 -1.82867512e-01 4.14195567e-01
-9.51781988e-01 4.07750398e-01 4.02859509e-01 1.59693405e-01
-9.18419421e-01 -6.85635388e-01 1.23452984e-01 5.14043510e-01
3.87513429e-01 1.19330108e+00 3.28426421e-01 1.15591615e-01
8.84238064e-01 -1.63436998e-02 -2.10767359e-01 -3.95530790e-01
5.51899195e-01 3.52250844e-01 4.54161972e-01 -5.58309853e-01
4.41404819e-01 1.40094846e-01 1.80586472e-01 -9.54428792e-01
-1.43118036e+00 -6.49206161e-01 -5.14945626e-01 -3.27124566e-01
7.89445877e-01 -6.74190164e-01 -4.10744160e-01 -2.04184785e-01
-1.51894891e+00 2.25902990e-01 -2.28106156e-01 2.46353939e-01
-1.31037617e-02 9.25494611e-01 -1.02401209e+00 -3.90806168e-01
-7.83167303e-01 -5.86153924e-01 1.14379632e+00 2.81621039e-01
-6.71463609e-01 -1.10339034e+00 3.42788100e-01 6.22608900e-01
3.96446772e-02 5.61579525e-01 1.05046439e+00 -8.06543350e-01
-4.78599250e-01 -2.39269465e-01 -2.23043449e-02 -3.77285004e-01
6.96074292e-02 2.66068488e-01 -4.56097603e-01 -1.69164747e-01
-4.62619096e-01 -3.29833776e-01 1.27694428e+00 6.78967655e-01
1.15715456e+00 -8.93016279e-01 -4.76355225e-01 -7.04727694e-02
8.54852736e-01 -5.18441200e-01 4.24824059e-01 3.06285858e-01
7.16349304e-01 7.50498652e-01 4.04875696e-01 2.30785146e-01
5.80620229e-01 1.41058639e-01 1.02985077e-01 1.05677776e-01
-1.79715887e-01 -1.44716159e-01 2.04228088e-01 1.69901991e+00
-2.90932536e-01 -4.23760831e-01 -9.08385575e-01 5.84299505e-01
-1.84956491e+00 -1.17901886e+00 -5.95493674e-01 1.88428569e+00
1.33664727e+00 4.83750641e-01 1.05969971e-02 -2.86839813e-01
6.20187998e-01 6.31435215e-01 -2.05414489e-01 -4.76508588e-01
-4.14342523e-01 -2.04386488e-02 4.45634097e-01 6.97862923e-01
-7.51278996e-01 7.55740106e-01 8.22668743e+00 7.83319473e-01
-3.41704875e-01 -9.49570760e-02 7.45179772e-01 -1.24293871e-01
-8.22138846e-01 7.05133155e-02 -6.46459043e-01 2.56189048e-01
1.10859406e+00 -1.10126877e+00 -1.88377112e-01 3.90830100e-01
2.63305724e-01 -2.75096953e-01 -9.96826291e-01 4.25754726e-01
5.41021347e-01 -2.02376628e+00 3.74021679e-01 9.95350033e-02
9.70983267e-01 3.37196738e-02 -4.69426990e-01 -1.46987975e-01
6.97968066e-01 -7.80092001e-01 3.02310765e-01 6.95879102e-01
5.75451910e-01 -5.07429779e-01 7.54407108e-01 2.13840231e-01
-4.97880965e-01 4.63818640e-01 -5.30097783e-01 8.29825327e-02
1.20736204e-01 1.26919377e+00 -8.25270414e-01 9.10263300e-01
4.76654381e-01 9.26821649e-01 -9.23651874e-01 9.40156162e-01
-6.42325997e-01 1.12760508e+00 9.94345024e-02 -5.73080242e-01
9.59929377e-02 -1.00039087e-01 1.14925742e+00 1.81320620e+00
-1.90864019e-02 3.47441107e-01 -1.29779465e-02 4.11512703e-01
-8.98135543e-01 3.86235029e-01 -7.17816055e-01 -5.87268949e-01
3.34915817e-01 1.42510629e+00 -7.77664363e-01 -7.55163252e-01
1.11530237e-01 5.37157893e-01 7.81155586e-01 2.66619116e-01
-1.53259665e-01 -7.79254556e-01 -1.10313430e-01 -8.03962275e-02
6.19324064e-03 -2.99235702e-01 -3.17471445e-01 -1.21823728e+00
-9.87847000e-02 -8.71242642e-01 8.02950919e-01 -5.22487462e-01
-1.23791170e+00 3.18369418e-01 -1.85196791e-02 -7.83596873e-01
-3.41299206e-01 -2.69480646e-01 -9.19211090e-01 3.98086548e-01
-1.43026161e+00 -5.28120816e-01 -6.90989289e-03 -1.16744019e-01
4.66489464e-01 -1.00283951e-01 7.95191944e-01 -3.19000870e-01
-4.20190215e-01 2.54893780e-01 3.15848470e-01 3.25076319e-02
8.59291613e-01 -1.73809505e+00 4.54305261e-01 8.74626696e-01
2.86333889e-01 1.04483533e+00 1.09055638e+00 -9.47894871e-01
-1.68611455e+00 -6.22133851e-01 1.43846250e+00 -7.36432076e-01
8.75222325e-01 -1.84506401e-01 -7.80919671e-01 3.64647329e-01
1.06274521e+00 -6.27103508e-01 1.16032064e+00 6.45891309e-01
-3.31394076e-01 2.55140603e-01 -6.07900500e-01 6.01404369e-01
1.18787599e+00 -1.79771036e-01 -1.11559629e+00 9.97183383e-01
8.95564497e-01 -4.01081711e-01 -1.11345351e+00 1.48334712e-01
2.99495250e-01 -5.48337758e-01 7.33510733e-01 -1.00797546e+00
1.17247570e+00 2.60097813e-02 4.89106148e-01 -1.45070982e+00
-7.54137039e-01 -1.06686425e+00 -5.41033506e-01 1.11253297e+00
7.86724210e-01 -3.47583443e-01 5.22190928e-01 3.07945520e-01
-3.55969548e-01 -7.21594989e-01 -7.17944562e-01 -3.30011874e-01
1.26084954e-01 2.92720199e-01 -8.98760483e-02 9.80964303e-01
8.46801877e-01 1.10724175e+00 5.36011197e-02 -4.22355086e-01
1.02493775e+00 3.45599055e-01 5.27734041e-01 -1.44579422e+00
-2.66391207e-02 -8.29028070e-01 -3.19177032e-01 -8.11360180e-01
3.76326382e-01 -1.19980729e+00 -1.42645672e-01 -2.55618024e+00
7.42006660e-01 3.28480989e-01 -3.84374797e-01 4.06378843e-02
-6.99473441e-01 -1.34740934e-01 -2.52088428e-01 5.03996491e-01
-1.39144588e+00 3.85332108e-01 1.14526713e+00 -4.83833283e-01
9.48401690e-02 -4.89655644e-01 -1.38102078e+00 6.72639072e-01
6.28515780e-01 -5.93123674e-01 -1.69976145e-01 -3.98809999e-01
7.04631746e-01 -8.31186846e-02 -2.52959281e-01 -3.60676855e-01
7.38141179e-01 -2.36074656e-01 2.66874641e-01 -7.36858010e-01
-3.42000455e-01 4.39777404e-01 -4.52839315e-01 2.61516154e-01
-1.10740173e+00 1.17790677e-01 4.32812758e-02 8.06567013e-01
-1.91447601e-01 -2.61132121e-01 6.93149865e-02 -1.24885723e-01
5.21465763e-02 -1.09815329e-01 -4.81627464e-01 6.83166802e-01
3.12561810e-01 3.05201173e-01 -1.09000731e+00 -5.81988752e-01
-1.68644667e-01 4.20000970e-01 2.01883376e-01 1.69910461e-01
5.33544660e-01 -9.85342920e-01 -1.29545796e+00 -9.48556364e-01
6.08101711e-02 -1.69376343e-01 -2.11968392e-01 9.05381918e-01
-5.92504144e-01 5.71805954e-01 2.42716998e-01 -3.04550469e-01
-1.38200629e+00 1.76849782e-01 -4.47715759e-01 -8.44031513e-01
-7.25464880e-01 7.42958128e-01 -9.80665907e-02 -3.57089080e-02
1.43116027e-01 -3.78731862e-02 -7.19925761e-01 2.94643283e-01
6.30118191e-01 3.62701863e-01 1.90212838e-02 -3.29085916e-01
-3.55526805e-01 2.88710296e-01 -3.00519258e-01 -7.76834264e-02
1.55266178e+00 6.10104129e-02 -8.69096160e-01 5.95111132e-01
9.57500100e-01 9.23868775e-01 -2.27210104e-01 -1.76821813e-01
7.39765525e-01 -1.08348250e-01 2.12686151e-01 -7.88479388e-01
-3.59107494e-01 3.62121791e-01 -6.42491996e-01 5.20455956e-01
4.60104108e-01 3.38836700e-01 3.04660141e-01 7.59491324e-01
-1.22594401e-01 -1.40827966e+00 1.82937473e-01 5.46582639e-01
1.22147870e+00 -9.21702981e-01 1.01907527e+00 -6.19351804e-01
-4.81862813e-01 1.36091459e+00 1.38573959e-01 -1.01941124e-01
2.61523366e-01 1.44039184e-01 -2.17081204e-01 -8.26534033e-01
-1.07682145e+00 2.16446117e-01 9.52503204e-01 -6.45998344e-02
1.16544211e+00 -2.06119284e-01 -1.19697928e+00 5.47807515e-01
-4.70964104e-01 -4.05543953e-01 1.03918743e+00 9.35304880e-01
-5.98221362e-01 -9.53206003e-01 -2.24809110e-01 9.99587893e-01
-9.99638796e-01 -3.57211292e-01 -1.00016356e+00 3.59147012e-01
-8.33052933e-01 1.03433919e+00 -3.35406870e-01 1.27423614e-01
1.23185523e-01 -7.44417496e-03 5.64515829e-01 -7.77696490e-01
-8.45267713e-01 1.97540894e-01 7.87623346e-01 -4.76919860e-01
-6.37921154e-01 -7.24036992e-01 -1.29289436e+00 -3.34700257e-01
-1.23235382e-01 8.98826361e-01 4.18268889e-01 5.61981022e-01
8.77462029e-01 9.30578351e-01 3.08220655e-01 -7.20500827e-01
-6.22937620e-01 -1.12022400e+00 -2.80173272e-01 3.91233385e-01
2.07356721e-01 -1.55379668e-01 -5.76857746e-01 1.15723655e-01] | [12.455068588256836, 9.470772743225098] |
58fcd758-82ba-4133-81e4-dc56ad4bbf3e | spatio-temporal-transformer-network-for-video | null | null | http://openaccess.thecvf.com/content_ECCV_2018/html/Tae_Hyun_Kim_Spatio-temporal_Transformer_Network_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Tae_Hyun_Kim_Spatio-temporal_Transformer_Network_ECCV_2018_paper.pdf | Spatio-temporal Transformer Network for Video Restoration | State-of-the-art video restoration methods integrate optical flow estimation networks to utilize temporal information. However, these networks typically consider only a pair of consecutive frames and hence are not capable of capturing long-range temporal dependencies and fall short of establishing correspondences across several timesteps. To alleviate these problems, we propose a novel Spatio-temporal Transformer Network (STTN) which handles multiple frames at once and thereby manages to mitigate the common nuisance of occlusions in optical flow estimation. Our proposed STTN comprises a module that estimates optical flow in both space and time and a resampling layer that selectively warps target frames using the estimated flow. In our experiments, we demonstrate the efficiency of the proposed network and show state-of-the-art restoration results in video super-resolution and video deblurring. | ['Tae Hyun Kim', 'Michael Hirsch', 'Mehdi S. M. Sajjadi', 'Bernhard Scholkopf'] | 2018-09-01 | null | null | null | eccv-2018-9 | ['video-restoration'] | ['computer-vision'] | [ 3.72890085e-02 -5.60861290e-01 -2.35890925e-01 -2.50487924e-02
-1.82094261e-01 -3.33632916e-01 4.98474628e-01 -4.05405700e-01
-3.32563639e-01 9.91048932e-01 3.63754690e-01 5.30231511e-04
-1.75464943e-01 -7.05728590e-01 -5.51344573e-01 -5.22182882e-01
-2.13238373e-01 -1.54513121e-01 5.67773521e-01 -3.44314761e-02
3.32945198e-01 7.28612185e-01 -1.47340882e+00 2.17313260e-01
8.94824564e-01 8.12526226e-01 3.14935237e-01 7.99692094e-01
-3.99919115e-02 1.19557750e+00 -3.49708974e-01 5.61046414e-02
3.82985473e-01 -5.18028975e-01 -7.84365654e-01 2.97578335e-01
9.96736109e-01 -8.56454074e-01 -9.96471465e-01 9.70979989e-01
3.24272275e-01 4.48313147e-01 4.95417938e-02 -1.10161769e+00
-5.67135632e-01 1.40414298e-01 -5.95083773e-01 1.00141644e+00
4.78484750e-01 2.26833418e-01 5.81085026e-01 -7.46861696e-01
8.94795477e-01 1.25822127e+00 6.35886550e-01 5.87521970e-01
-1.35940385e+00 -6.43252850e-01 1.56866968e-01 4.56900269e-01
-1.09062421e+00 -8.00049126e-01 8.03182662e-01 -3.83900166e-01
7.97884643e-01 -7.02195689e-02 7.32904613e-01 8.64266932e-01
2.02659488e-01 5.48531711e-01 8.79744232e-01 -4.96853739e-02
-7.80761093e-02 -5.11499405e-01 -1.27140716e-01 5.60439229e-01
-2.49162633e-02 4.13054138e-01 -8.87958825e-01 1.27837136e-01
1.52973020e+00 3.57365026e-03 -7.81345367e-01 -2.44212553e-01
-1.37278283e+00 3.90092373e-01 3.94616812e-01 3.21344584e-01
-3.74577105e-01 3.13184053e-01 4.29634720e-01 2.98328936e-01
5.62308550e-01 9.03184712e-02 -1.29723743e-01 -1.77179933e-01
-1.42713785e+00 8.74569193e-02 4.51557547e-01 7.47358441e-01
8.95934522e-01 3.66367549e-01 -1.02067240e-01 6.84032619e-01
1.26662739e-02 1.30960226e-01 3.72738689e-01 -1.46405828e+00
3.05330336e-01 -6.02910519e-02 3.98608238e-01 -1.13895988e+00
-2.02701822e-01 -2.87577242e-01 -1.07380152e+00 2.88783133e-01
4.06540543e-01 8.26440826e-02 -7.41795123e-01 1.73225152e+00
3.20880562e-01 1.10931730e+00 -5.98905422e-02 1.21320975e+00
6.06095374e-01 6.04776144e-01 -3.99661763e-03 -6.75610304e-01
9.34759557e-01 -1.18174720e+00 -9.52889085e-01 -9.29138139e-02
8.01851079e-02 -9.01433051e-01 5.81845760e-01 6.81458339e-02
-1.42144167e+00 -8.69123995e-01 -8.53372514e-01 -2.00889170e-01
2.14692757e-01 -1.52208179e-01 3.14877152e-01 1.12975366e-01
-1.32599926e+00 1.03414226e+00 -1.12398016e+00 -3.18748295e-01
3.46164703e-01 2.31144518e-01 -5.08516669e-01 -3.05663615e-01
-1.08075321e+00 7.62983680e-01 1.94693118e-01 2.82780617e-01
-9.96463418e-01 -9.20051336e-01 -8.06807101e-01 6.48025125e-02
3.06013614e-01 -9.92621720e-01 8.25914919e-01 -9.52272356e-01
-1.61771107e+00 2.86785454e-01 -6.23617113e-01 -3.19390833e-01
7.16254592e-01 -3.86859328e-01 -5.66224337e-01 5.72641253e-01
1.00906052e-01 5.21340728e-01 1.05483794e+00 -1.10662162e+00
-7.53291845e-01 1.15063749e-02 1.66500345e-01 1.15384042e-01
-2.21982732e-01 4.28787842e-02 -3.69462162e-01 -8.55016470e-01
1.08170837e-01 -6.63618982e-01 -7.86244124e-02 3.62107038e-01
-1.40840068e-01 2.33220831e-01 1.23367095e+00 -7.19858408e-01
1.25207889e+00 -1.97462690e+00 3.13875943e-01 -3.30837905e-01
2.96706617e-01 5.68743110e-01 -2.90847242e-01 1.88938692e-01
-1.55080423e-01 -1.36898294e-01 -7.54115358e-02 -4.28970337e-01
-6.39994681e-01 3.01944971e-01 -4.00665194e-01 6.50764406e-01
5.56501746e-02 5.44540584e-01 -1.21881974e+00 -4.66772288e-01
6.78103983e-01 9.30320978e-01 -5.80958962e-01 3.36466610e-01
2.04894748e-02 1.00982594e+00 -2.93044895e-01 3.01684886e-01
9.31143045e-01 -2.75849432e-01 9.31663662e-02 -6.24221087e-01
-5.53660989e-01 3.51594657e-01 -1.17810369e+00 1.92487156e+00
-3.40791106e-01 9.17047441e-01 2.34964490e-01 -6.49911046e-01
4.36793119e-01 4.79103357e-01 9.41570878e-01 -4.16216105e-01
-9.10005718e-02 1.02753930e-01 -2.55534738e-01 -6.19138896e-01
7.07191885e-01 3.39177288e-02 8.38773608e-01 4.16372895e-01
1.01413466e-02 2.43096679e-01 4.92497474e-01 -1.14546251e-02
1.21592355e+00 4.60750103e-01 6.82973266e-02 -9.87053066e-02
8.77731621e-01 -2.70239413e-01 9.27061737e-01 5.70883274e-01
-5.75036407e-01 7.12792575e-01 1.84623063e-01 -8.20498049e-01
-1.25320804e+00 -1.09306467e+00 -1.98026989e-02 6.80698872e-01
4.46859211e-01 -4.36743408e-01 -4.42859173e-01 -4.20538545e-01
-1.83759466e-01 2.56601989e-01 -4.56899762e-01 1.12105444e-01
-1.02328587e+00 -3.77074778e-01 2.91089684e-01 3.77334267e-01
8.87742579e-01 -9.13529158e-01 -7.63331354e-01 3.91960293e-01
-5.82787454e-01 -1.60876918e+00 -7.92117536e-01 -6.79446161e-01
-1.12314916e+00 -1.13103569e+00 -7.04627991e-01 -6.67684317e-01
5.51128745e-01 7.72839963e-01 1.04768002e+00 2.13078618e-01
-9.12878588e-02 5.35215512e-02 -2.16887876e-01 7.42962837e-01
-3.75905216e-01 -1.41555920e-01 5.99662513e-02 2.20307499e-01
-1.73422232e-01 -1.21840525e+00 -9.71030951e-01 5.17454088e-01
-1.08858371e+00 1.89982906e-01 9.91476476e-02 8.22473645e-01
3.58183354e-01 -1.57820748e-03 2.84360170e-01 -3.56530368e-01
3.02224040e-01 -2.07359105e-01 -7.16964841e-01 1.08956136e-01
-2.76168615e-01 -1.17152400e-01 7.53045261e-01 -5.86533248e-01
-1.13668370e+00 -1.22216709e-01 2.30505750e-01 -1.00728762e+00
-6.65309280e-02 1.09849460e-01 3.12653363e-01 -4.03077126e-01
2.97006994e-01 2.79092759e-01 -5.46631915e-03 -5.05611956e-01
2.40182251e-01 1.79663017e-01 9.60389853e-01 -3.76154274e-01
8.60009193e-01 9.62008893e-01 2.14703158e-01 -7.90845335e-01
-7.21527398e-01 -4.40717131e-01 -8.88429284e-01 -4.49833214e-01
8.23074460e-01 -9.16971743e-01 -8.43672335e-01 6.21548116e-01
-1.45018411e+00 -1.17171645e-01 -1.93578914e-01 6.41942263e-01
-5.96065879e-01 7.23113477e-01 -1.06171370e+00 -4.12693948e-01
-1.59603715e-01 -1.18523598e+00 7.41062403e-01 3.40384603e-01
9.22944993e-02 -1.03653693e+00 2.01533929e-01 1.61496297e-01
6.44878924e-01 1.66824818e-01 4.84834045e-01 3.65097672e-01
-1.08983374e+00 3.45013022e-01 -5.24876177e-01 3.04391176e-01
4.83089834e-01 3.33720446e-02 -8.21439624e-01 -5.55856824e-01
-4.71459962e-02 1.63249448e-01 9.68099833e-01 6.78372800e-01
9.02524590e-01 -1.37265742e-01 -9.98727456e-02 9.62855339e-01
1.54958498e+00 -3.81854363e-02 1.04467154e+00 4.32136357e-01
9.29976344e-01 3.67831439e-01 4.27331448e-01 4.66693819e-01
2.92092085e-01 8.34042192e-01 4.31381673e-01 2.51148697e-02
-5.88991463e-01 -2.20177658e-02 3.97016525e-01 7.69292951e-01
-5.13807297e-01 -1.93281710e-01 -5.00277877e-01 8.01607490e-01
-1.97696376e+00 -1.30182493e+00 1.42340092e-02 2.06407547e+00
7.05550849e-01 -2.23380864e-01 -1.20816097e-01 -3.46816070e-02
9.97026980e-01 7.11137891e-01 -5.08623481e-01 6.68551326e-02
-2.12930188e-01 2.50760447e-02 4.37245637e-01 8.49244058e-01
-1.06310487e+00 1.05619085e+00 6.52020216e+00 3.80160451e-01
-1.19320369e+00 1.44371539e-01 2.23535091e-01 -1.85051441e-01
-5.63483313e-02 1.59094781e-01 -3.37473214e-01 3.66743207e-01
6.58966780e-01 -9.56967846e-02 7.15470135e-01 1.17513567e-01
7.21770883e-01 -1.90251991e-01 -9.11748648e-01 8.92666399e-01
-1.13185994e-01 -1.69379246e+00 4.20879312e-02 -7.75643215e-02
8.39521408e-01 3.07427421e-02 -1.69835806e-01 -4.98767018e-01
1.11572564e-01 -7.68536389e-01 5.76492131e-01 6.17239475e-01
7.51476109e-01 -5.10432839e-01 3.56455415e-01 -2.85835154e-02
-1.43429506e+00 -3.13565731e-02 -2.40077972e-01 -3.31453942e-02
6.78416729e-01 6.32317543e-01 -1.40971705e-01 5.83572149e-01
7.54406095e-01 1.43951833e+00 -1.25160381e-01 1.21571720e+00
-1.87767267e-01 1.90126210e-01 -9.03207287e-02 9.77479398e-01
9.39766690e-02 -3.21484953e-01 9.25907314e-01 9.13473248e-01
4.86030728e-01 1.02733895e-01 1.88625172e-01 7.62912929e-01
2.85476632e-02 -3.84365916e-01 -1.90688774e-01 2.95862198e-01
5.01781881e-01 9.65222180e-01 -5.62686145e-01 -5.14756262e-01
-5.79647243e-01 1.11924696e+00 2.02377096e-01 6.73070431e-01
-7.74572849e-01 4.71365042e-02 1.21162355e+00 2.25375831e-01
5.05007088e-01 -5.47044098e-01 1.92104250e-01 -1.71942616e+00
9.58538204e-02 -5.04975677e-01 1.85871124e-01 -1.02402055e+00
-1.13628423e+00 6.70969725e-01 -2.09629953e-01 -1.51254368e+00
-1.92063227e-01 -1.29779398e-01 -3.49826008e-01 8.20971489e-01
-2.13978004e+00 -8.90441358e-01 -6.22841716e-01 8.94953489e-01
6.24525368e-01 2.00381517e-01 3.86134475e-01 6.32820427e-01
-6.47068799e-01 4.39724512e-02 2.26303209e-02 1.74695224e-01
1.01855922e+00 -7.37302840e-01 4.92786884e-01 1.36415493e+00
-2.67392874e-01 5.51687539e-01 8.03682268e-01 -6.15485668e-01
-1.15254462e+00 -1.07849097e+00 8.03358614e-01 1.40655205e-01
7.06433535e-01 2.84712881e-01 -1.07823813e+00 7.53298759e-01
2.07260400e-01 8.08396399e-01 -1.18641807e-02 -5.30492663e-01
-4.01266009e-01 -1.82815671e-01 -1.07315087e+00 4.37689513e-01
1.25634265e+00 -7.44713545e-01 -2.00401649e-01 3.34859379e-02
5.40693760e-01 -5.50947130e-01 -9.61549759e-01 3.49827170e-01
6.10049367e-01 -1.49717247e+00 1.23728716e+00 -3.90849054e-01
6.65592015e-01 -5.92859030e-01 3.80747207e-02 -1.22521245e+00
-4.15127128e-01 -9.83376205e-01 -3.63677472e-01 1.02672327e+00
-3.49957079e-01 -5.46552002e-01 7.23511219e-01 3.42955440e-01
-1.18179135e-01 -2.06480563e-01 -1.11673748e+00 -6.82572782e-01
-4.15514976e-01 -2.25720853e-01 3.32183599e-01 1.18129396e+00
-3.41306299e-01 -8.21852088e-02 -8.02875459e-01 3.91343981e-01
9.54415858e-01 1.11224718e-01 5.56236029e-01 -9.36737895e-01
-1.42262712e-01 -2.45161414e-01 -5.63955963e-01 -1.34178901e+00
3.83745104e-01 -2.68301934e-01 -7.25586526e-03 -1.37419581e+00
6.14173412e-02 -1.67844594e-01 -3.34428042e-01 1.82963729e-01
-1.31377578e-01 4.37994123e-01 3.49148154e-01 5.76837659e-01
-3.59693140e-01 4.55469608e-01 1.67449498e+00 1.05134510e-01
-2.53995508e-01 -2.86180466e-01 -2.55700555e-02 6.05960488e-01
4.62994158e-01 -2.96735615e-01 -4.83737707e-01 -9.31762695e-01
-2.99736977e-01 6.26224875e-01 6.88327372e-01 -1.04724228e+00
4.03383642e-01 -4.17000234e-01 2.52347678e-01 -3.96591932e-01
3.84740591e-01 -7.49046922e-01 3.02140176e-01 2.76936293e-01
-2.12210268e-01 2.60772765e-01 1.50581881e-01 6.45277679e-01
-4.51660424e-01 2.99023449e-01 1.00522339e+00 -1.09724842e-01
-8.34751546e-01 6.04667544e-01 -2.93316394e-01 -6.96608284e-03
6.73025966e-01 -2.29741678e-01 -5.87253332e-01 -4.54755306e-01
-7.17760086e-01 -5.33817075e-02 7.13189960e-01 4.07702148e-01
6.47253692e-01 -1.25151432e+00 -6.44915104e-01 2.05710754e-01
-4.57001656e-01 -1.32642910e-01 6.06640100e-01 1.00755119e+00
-7.60646284e-01 1.94091663e-01 -6.86213255e-01 -6.35516346e-01
-1.14551568e+00 4.49558407e-01 6.43191397e-01 -2.27186471e-01
-1.00915062e+00 5.88317335e-01 1.65170655e-01 2.14827076e-01
-7.47257769e-02 -1.32105872e-01 -2.39210963e-01 -1.17019013e-01
7.81840324e-01 7.33867347e-01 -2.25050434e-01 -9.56786096e-01
-2.48612225e-01 8.85201812e-01 1.04941227e-01 -5.41816875e-02
1.34560192e+00 -7.09732234e-01 -4.50552583e-01 8.74289870e-02
1.13168383e+00 -2.02308968e-01 -1.94110799e+00 -4.32473838e-01
-4.29373562e-01 -1.09857595e+00 3.50931376e-01 -2.48649210e-01
-1.58128548e+00 6.73260689e-01 4.87143427e-01 5.58892824e-02
1.50915658e+00 -5.18944681e-01 1.29303348e+00 -1.55669376e-01
1.96920365e-01 -6.96321428e-01 -5.02610132e-02 5.46369076e-01
4.95870590e-01 -1.05901349e+00 2.34208271e-01 -7.55865157e-01
6.95456862e-02 1.44551301e+00 4.34246302e-01 -3.17397326e-01
6.00344837e-01 2.57828474e-01 6.23328052e-02 1.54403463e-01
-7.99328566e-01 -5.68577126e-02 9.42679048e-02 5.86217165e-01
3.61074805e-01 -4.37568128e-01 -1.97155908e-01 -5.70077300e-01
2.55110621e-01 4.38262761e-01 8.41738224e-01 9.12103534e-01
-1.86823100e-01 -1.11369264e+00 -3.10593039e-01 -1.18834205e-01
-4.20068622e-01 -1.43657818e-01 3.66785139e-01 6.54408157e-01
-4.92894389e-02 9.54980731e-01 2.44159043e-01 -2.59997606e-01
1.30576953e-01 -4.09312755e-01 6.69766724e-01 4.02209014e-02
-3.64204049e-01 2.33981952e-01 4.96805087e-02 -1.12378275e+00
-1.18347991e+00 -5.94108880e-01 -9.55153227e-01 -7.46557653e-01
7.54067721e-03 -1.31214097e-01 1.79094970e-01 9.53028083e-01
5.05942881e-01 6.71222270e-01 5.66787899e-01 -1.27330315e+00
8.18043798e-02 -7.56167412e-01 -4.87452835e-01 5.10284543e-01
1.06673300e+00 -6.17738068e-01 -4.30493027e-01 4.85362560e-01] | [10.908451080322266, -1.78080415725708] |
186e22c5-ddb4-4deb-a6c0-6d755b22fb04 | deep-video-generation-prediction-and | 1711.08682 | null | http://arxiv.org/abs/1711.08682v3 | http://arxiv.org/pdf/1711.08682v3.pdf | Deep Video Generation, Prediction and Completion of Human Action Sequences | Current deep learning results on video generation are limited while there are
only a few first results on video prediction and no relevant significant
results on video completion. This is due to the severe ill-posedness inherent
in these three problems. In this paper, we focus on human action videos, and
propose a general, two-stage deep framework to generate human action videos
with no constraints or arbitrary number of constraints, which uniformly address
the three problems: video generation given no input frames, video prediction
given the first few frames, and video completion given the first and last
frames. To make the problem tractable, in the first stage we train a deep
generative model that generates a human pose sequence from random noise. In the
second stage, a skeleton-to-image network is trained, which is used to generate
a human action video given the complete human pose sequence generated in the
first stage. By introducing the two-stage strategy, we sidestep the original
ill-posed problems while producing for the first time high-quality video
generation/prediction/completion results of much longer duration. We present
quantitative and qualitative evaluation to show that our two-stage approach
outperforms state-of-the-art methods in video generation, prediction and video
completion. Our video result demonstration can be viewed at
https://iamacewhite.github.io/supp/index.html | ['Chi-Keung Tang', 'Yu-Wing Tai', 'Haoye Cai', 'Chunyan Bai'] | 2017-11-23 | deep-video-generation-prediction-and-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Chunyan_Bai_Deep_Video_Generation_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Chunyan_Bai_Deep_Video_Generation_ECCV_2018_paper.pdf | eccv-2018-9 | ['human-action-generation'] | ['computer-vision'] | [ 5.29989481e-01 1.62876040e-01 2.28826385e-02 3.85892019e-02
-7.72463977e-01 -2.95427144e-01 6.16629779e-01 -6.53455555e-01
-2.62431920e-01 7.71204174e-01 3.09926450e-01 -4.15019654e-02
2.89796501e-01 -5.78713536e-01 -9.77526546e-01 -6.08709753e-01
-3.54532413e-02 2.76393771e-01 2.82614321e-01 -4.07783687e-02
3.51532698e-02 1.64582849e-01 -1.66178703e+00 5.41675091e-01
5.38666964e-01 1.03923345e+00 4.07669663e-01 1.14521933e+00
4.48592275e-01 1.07735491e+00 -4.44326967e-01 -4.40396070e-01
4.26560104e-01 -9.06035542e-01 -6.38369501e-01 4.70906466e-01
4.25610304e-01 -1.01347435e+00 -6.49916828e-01 7.98068285e-01
6.49518073e-01 3.26758623e-01 5.26684344e-01 -1.46698785e+00
-3.62875223e-01 2.50849366e-01 -3.50914568e-01 -1.28513828e-01
9.53669906e-01 5.19057035e-01 8.01325679e-01 -9.48190212e-01
8.58067632e-01 1.13572538e+00 4.88684654e-01 1.01129377e+00
-9.07724917e-01 -5.34411073e-01 4.29433882e-02 2.66471267e-01
-1.22857702e+00 -5.86659968e-01 6.27935350e-01 -6.28801405e-01
7.34076679e-01 1.42322242e-01 9.43603754e-01 1.46783531e+00
2.41901651e-01 1.02693748e+00 5.29578745e-01 -2.62286782e-01
1.77245200e-01 -4.43797261e-01 -6.12758338e-01 7.27889776e-01
-1.73150539e-01 2.96989113e-01 -5.26679754e-01 1.94172621e-01
1.04200816e+00 1.56836852e-01 -4.79954273e-01 -2.68078953e-01
-1.29416847e+00 7.34677196e-01 7.61561766e-02 1.95198860e-02
-6.39509678e-01 3.21476012e-01 3.67804438e-01 1.45076618e-01
2.68003076e-01 1.86212286e-01 -6.48641288e-02 -3.45134556e-01
-1.37458003e+00 7.34335184e-01 6.81169450e-01 1.03155982e+00
4.98332858e-01 2.34435946e-01 -3.39733750e-01 5.03215611e-01
3.20653886e-01 4.68845755e-01 3.50799292e-01 -1.39210093e+00
4.99161333e-01 -2.08896454e-02 3.38404626e-01 -9.50958371e-01
-1.27030596e-01 -2.38853674e-02 -7.85694301e-01 2.82078058e-01
3.82380068e-01 -5.03582239e-01 -1.09020770e+00 1.57072461e+00
2.87740022e-01 4.73049730e-01 -1.24577247e-02 1.22662139e+00
8.69608402e-01 9.09753144e-01 -7.76343746e-03 -3.33990574e-01
1.12130094e+00 -1.28342867e+00 -7.40440309e-01 -6.80909753e-02
3.46475571e-01 -9.10965741e-01 7.31994748e-01 4.39072371e-01
-1.55574429e+00 -8.48992705e-01 -8.27805817e-01 -1.79831758e-01
1.78939655e-01 2.64335990e-01 3.44031960e-01 1.42733231e-01
-1.19827342e+00 6.35607064e-01 -9.10627723e-01 -2.04108059e-01
3.34025115e-01 1.43638745e-01 -3.92323226e-01 -1.86396331e-01
-1.23969281e+00 4.85631317e-01 2.80875593e-01 1.98976904e-01
-1.44008875e+00 -4.64237422e-01 -9.96175230e-01 -2.49260627e-02
5.48544407e-01 -1.01452541e+00 1.45216441e+00 -1.18769073e+00
-1.70893955e+00 5.38718462e-01 -1.33658335e-01 -4.82010722e-01
9.93394256e-01 -5.76266110e-01 -9.26473439e-02 5.13718069e-01
9.18580592e-02 1.11705041e+00 9.54767287e-01 -1.14506996e+00
-6.93456888e-01 2.11775422e-01 2.79525787e-01 3.17345053e-01
2.14571610e-01 1.60960946e-02 -9.29515004e-01 -9.37324882e-01
-3.29216838e-01 -1.08972216e+00 -3.04235429e-01 2.36472534e-03
-3.73207808e-01 3.41858901e-03 6.30642176e-01 -9.22913969e-01
1.29684818e+00 -2.00591826e+00 6.48403764e-01 -3.24973494e-01
5.42212185e-03 3.49696577e-01 -3.69611382e-01 4.92839217e-01
-1.36142641e-01 1.83345731e-02 -1.84401125e-01 -6.64991617e-01
-3.66359986e-02 1.81055628e-02 -4.33320582e-01 1.69193342e-01
2.62955248e-01 8.98080111e-01 -9.91614938e-01 -4.96555775e-01
3.53933066e-01 5.50503373e-01 -8.67570043e-01 7.00020075e-01
-4.53974068e-01 5.57272434e-01 -2.08246142e-01 6.03067160e-01
3.48833799e-01 -1.59860164e-01 1.40733361e-01 -3.33970636e-01
-2.37221364e-02 1.23172551e-02 -1.09512711e+00 1.94116235e+00
-2.80460209e-01 6.10546470e-01 -8.69059786e-02 -8.58291209e-01
3.73693287e-01 7.22988307e-01 7.50199378e-01 -4.28733021e-01
1.85554728e-01 5.81891499e-02 -1.43193245e-01 -8.64461780e-01
5.25367022e-01 -1.98615342e-01 5.49264550e-02 3.17283005e-01
1.39839694e-01 -1.50889769e-01 5.52708685e-01 4.17073190e-01
1.02060401e+00 8.11768293e-01 1.43583775e-01 3.94297004e-01
4.30188447e-01 -1.17023617e-01 6.33158505e-01 5.23588121e-01
-1.87053591e-01 1.29632652e+00 5.55024505e-01 -4.05744344e-01
-1.28689802e+00 -1.00514364e+00 5.86221516e-01 7.38916278e-01
1.15723073e-01 -6.43130362e-01 -9.56991911e-01 -6.09542966e-01
-4.46218729e-01 4.74670231e-01 -5.50387859e-01 -4.72485758e-02
-8.31612110e-01 -2.82142520e-01 3.56772661e-01 5.79888046e-01
4.33201790e-01 -1.38682973e+00 -7.68968582e-01 2.24717364e-01
-5.79833090e-01 -1.29967058e+00 -5.89739084e-01 -4.78850812e-01
-6.23148382e-01 -1.17371404e+00 -1.17579603e+00 -7.27147043e-01
5.81962228e-01 4.44585234e-02 1.01776612e+00 3.01167727e-01
-2.42075071e-01 2.72614628e-01 -5.30672252e-01 2.81602400e-03
-5.83044469e-01 -3.01746309e-01 4.35677730e-02 -3.94500308e-02
-1.58344850e-01 -3.70339841e-01 -8.69542241e-01 2.50270009e-01
-1.13493323e+00 6.46539330e-01 6.36977255e-01 8.82192135e-01
5.27840793e-01 -5.78623526e-02 2.85298675e-01 -4.81473893e-01
2.24453524e-01 -4.35463667e-01 -3.42259675e-01 2.39962079e-02
1.11925518e-02 -1.79816470e-01 7.58699477e-01 -5.34412503e-01
-1.01218951e+00 4.39357400e-01 -3.73968601e-01 -8.96697640e-01
-2.12975085e-01 3.11485469e-01 -2.11050898e-01 5.04694402e-01
4.30671901e-01 2.90700823e-01 6.98046237e-02 -3.07998717e-01
3.01921397e-01 3.14377427e-01 7.11077392e-01 -3.20407361e-01
7.72532165e-01 3.06452274e-01 -8.27435926e-02 -6.91703975e-01
-7.14345932e-01 -8.96975026e-02 -6.59384549e-01 -5.55291235e-01
1.20818245e+00 -1.13557589e+00 -6.01725042e-01 6.43320560e-01
-1.25259387e+00 -7.41181552e-01 -2.10532874e-01 5.08880854e-01
-1.04721403e+00 6.53190613e-01 -8.04741681e-01 -6.27737939e-01
-2.14574501e-01 -1.27477443e+00 1.24191356e+00 8.07881132e-02
-2.52697885e-01 -7.27069318e-01 -2.43587699e-02 5.24347603e-01
8.54159370e-02 5.39747238e-01 2.09789440e-01 -1.24618500e-01
-8.38257194e-01 -1.02736555e-01 4.42265831e-02 4.44563448e-01
-5.25335819e-02 2.05977872e-01 -4.58484918e-01 -3.73773932e-01
-1.78174943e-01 -4.70731199e-01 7.61991978e-01 4.79837835e-01
1.13455677e+00 -4.14137423e-01 -1.02591127e-01 7.36617684e-01
1.18880510e+00 2.68114448e-01 9.34475243e-01 3.63855585e-02
7.81038344e-01 4.20920104e-01 8.98080051e-01 6.26980186e-01
1.70779988e-01 8.46109986e-01 3.65840256e-01 -9.43043754e-02
-4.17203099e-01 -5.52659750e-01 6.39081120e-01 5.42726040e-01
-5.07598519e-01 -5.87061763e-01 -6.06562495e-01 4.10968870e-01
-2.05147338e+00 -1.41493773e+00 1.15547642e-01 2.00807953e+00
6.44868493e-01 8.31818804e-02 3.97396654e-01 3.58461477e-02
7.51906931e-01 1.96087465e-01 -3.78838807e-01 1.22882128e-01
1.72860980e-01 5.44399656e-02 3.79335657e-02 5.87639868e-01
-1.10786045e+00 1.01536798e+00 6.10834980e+00 8.35655987e-01
-1.06225348e+00 -1.46999463e-01 5.90314090e-01 -4.65265095e-01
-1.53332531e-01 -2.88566761e-02 -5.17118692e-01 6.16174459e-01
8.14471722e-01 1.56349409e-02 4.26402688e-01 6.39935493e-01
6.62387371e-01 -1.39053404e-01 -1.17647457e+00 1.03542936e+00
2.10030764e-01 -1.44862783e+00 3.23731780e-01 -1.11909155e-02
8.17536116e-01 -3.70919108e-01 -1.46603122e-01 2.71554917e-01
5.83511852e-02 -9.86577809e-01 1.07214308e+00 7.02890337e-01
1.01577985e+00 -5.97655356e-01 5.97872913e-01 2.92571455e-01
-1.22365975e+00 -6.16671033e-02 -1.49086729e-01 -1.56919032e-01
8.32495272e-01 3.64877015e-01 -4.74245757e-01 6.34309292e-01
5.51400304e-01 8.62079978e-01 -1.88130856e-01 9.43004012e-01
-4.39365059e-01 4.26777124e-01 3.53411352e-03 3.47775817e-01
1.86537400e-01 -1.27675325e-01 5.73522568e-01 1.14558470e+00
6.80622637e-01 3.73605907e-01 3.20363641e-01 7.67707169e-01
-1.45332469e-02 -2.51022220e-01 -6.14579201e-01 -5.42796850e-02
1.19358115e-01 1.15039968e+00 -5.87405860e-01 -6.35131121e-01
-4.43534553e-01 1.36603177e+00 2.64306180e-02 5.30487597e-01
-1.37608600e+00 -2.11643443e-01 4.28752244e-01 3.60169739e-01
3.78077626e-01 -3.44206393e-01 2.07592651e-01 -1.43409932e+00
8.68950114e-02 -1.07752323e+00 2.32832044e-01 -1.00374222e+00
-7.76268542e-01 6.64674342e-01 7.40507618e-02 -1.50582802e+00
-8.89823914e-01 -4.06215340e-01 -5.62015474e-01 5.25074720e-01
-9.69495893e-01 -1.12322724e+00 -4.18305248e-01 6.75830066e-01
1.02168655e+00 -1.10214278e-01 4.58006501e-01 5.16544759e-01
-5.31060040e-01 3.33641440e-01 -3.60593975e-01 2.72883654e-01
5.74921846e-01 -9.53360379e-01 4.77616459e-01 1.15562820e+00
-1.02410637e-01 2.48504028e-01 8.63571823e-01 -8.54358613e-01
-1.22035003e+00 -1.08482790e+00 7.44439483e-01 -2.55915493e-01
4.00983006e-01 -3.37716699e-01 -4.98210073e-01 6.91591322e-01
2.46714354e-01 -9.62629616e-02 3.28328729e-01 -6.73265457e-01
1.76527932e-01 3.26240718e-01 -7.27914989e-01 8.37753177e-01
1.21098149e+00 -2.50194341e-01 -3.13218236e-01 3.36079419e-01
7.74422646e-01 -6.34254158e-01 -7.13690937e-01 5.19724190e-01
7.55010068e-01 -1.16977453e+00 9.48880613e-01 -4.42945987e-01
1.13744593e+00 -3.34169865e-01 -1.36943325e-01 -9.88181949e-01
-2.17810914e-01 -9.99810576e-01 -3.45783442e-01 8.70809197e-01
3.31587702e-01 7.05047771e-02 8.21671546e-01 5.57817042e-01
-2.77846694e-01 -9.63410974e-01 -5.36955833e-01 -5.99460304e-01
-2.71166086e-01 -3.45503896e-01 1.66997522e-01 5.13336658e-01
-2.36374587e-01 2.15141684e-01 -1.04608023e+00 -1.16880707e-01
4.12268639e-01 1.16640806e-01 1.07580590e+00 -4.90140170e-01
-6.20125473e-01 -3.49873789e-02 -2.62404889e-01 -1.46098924e+00
5.12161553e-02 -4.05649751e-01 3.53663146e-01 -1.62595773e+00
2.70356715e-01 1.33159235e-01 1.80214226e-01 2.73666531e-01
-3.87095302e-01 4.73263890e-01 5.83643436e-01 3.06584179e-01
-7.60143757e-01 7.54664242e-01 1.53183305e+00 1.86693102e-01
-1.08026713e-01 -1.24383410e-02 -2.84325480e-01 6.95404887e-01
5.79592168e-01 -1.72497511e-01 -5.71259916e-01 -4.17121738e-01
-9.47020855e-03 6.46658003e-01 6.16636574e-01 -1.31730497e+00
1.34948837e-02 -2.18421832e-01 5.71041286e-01 -4.66835588e-01
6.41526997e-01 -5.36761940e-01 5.15079677e-01 6.58824980e-01
-2.73748696e-01 2.65995085e-01 -5.79132959e-02 4.90443975e-01
-2.76069760e-01 -4.84591015e-02 7.42727876e-01 -2.31639445e-01
-7.47404337e-01 5.89646161e-01 -5.42246759e-01 -4.01831605e-02
1.08760619e+00 -2.90711433e-01 3.32310908e-02 -7.69245327e-01
-1.06503057e+00 1.33095652e-01 5.91895342e-01 6.80837750e-01
8.59920025e-01 -1.40609598e+00 -8.93604517e-01 1.17292084e-01
-3.79132301e-01 7.19466284e-02 5.80019057e-01 7.30226159e-01
-8.60147178e-01 2.32916266e-01 -4.34364498e-01 -5.06881297e-01
-1.22298813e+00 6.99862182e-01 1.32235676e-01 -3.03409129e-01
-6.51800036e-01 6.72195673e-01 4.46504176e-01 4.89116944e-02
2.17426583e-01 -1.67592205e-02 -5.52957915e-02 -1.63377091e-01
4.90233719e-01 3.54147881e-01 -4.38732117e-01 -8.55440319e-01
-4.49557044e-02 4.71570671e-01 1.46731764e-01 -4.36187744e-01
1.11212158e+00 -1.42356530e-01 3.52698207e-01 2.01569609e-02
1.08456850e+00 -2.38564283e-01 -1.83720183e+00 2.53560781e-01
-5.84540486e-01 -5.76165199e-01 -3.76860738e-01 -5.30673623e-01
-1.21676373e+00 8.52940321e-01 3.19029033e-01 -8.38582143e-02
1.33258879e+00 -2.69546717e-01 1.14584708e+00 -6.45276904e-02
1.93301052e-01 -1.08496153e+00 5.38489580e-01 4.50729162e-01
1.15764070e+00 -1.01224351e+00 3.32610905e-02 -4.03078884e-01
-8.12864423e-01 1.01564133e+00 8.54226649e-01 -2.05674991e-01
2.99965292e-01 2.69031823e-01 -6.51361942e-02 1.03518836e-01
-9.29572701e-01 -6.04282878e-02 2.58927107e-01 4.44033921e-01
5.82442343e-01 -1.13908231e-01 -3.82522017e-01 3.96150410e-01
-2.55845964e-01 4.10395920e-01 6.00947380e-01 7.88457990e-01
-3.01297218e-01 -1.04102886e+00 -2.75989562e-01 1.49526879e-01
-5.97551107e-01 -2.64023263e-02 -1.68340296e-01 7.19711721e-01
2.56545693e-01 9.20650661e-01 -3.46706770e-02 -5.39666474e-01
1.97508186e-01 -1.18264541e-01 5.47687173e-01 -4.63970572e-01
-2.53573269e-01 4.09278423e-01 2.24609166e-01 -1.02761483e+00
-5.62952101e-01 -6.92802310e-01 -1.18192697e+00 -2.87270010e-01
6.68527605e-03 -2.26199278e-03 2.26130337e-01 7.42858589e-01
3.68319750e-01 5.50623834e-01 4.21138257e-01 -1.59945905e+00
-2.70436496e-01 -8.50385368e-01 -1.71661556e-01 7.42448986e-01
3.23932528e-01 -5.49668014e-01 -2.56708920e-01 6.76533699e-01] | [10.756230354309082, -0.6050357222557068] |
e958a8f7-a2ac-40ac-bb1e-d47ca2241eca | meta-learning-adversarial-domain-adaptation | 2107.12262 | null | https://arxiv.org/abs/2107.12262v1 | https://arxiv.org/pdf/2107.12262v1.pdf | Meta-Learning Adversarial Domain Adaptation Network for Few-Shot Text Classification | Meta-learning has emerged as a trending technique to tackle few-shot text classification and achieved state-of-the-art performance. However, existing solutions heavily rely on the exploitation of lexical features and their distributional signatures on training data, while neglecting to strengthen the model's ability to adapt to new tasks. In this paper, we propose a novel meta-learning framework integrated with an adversarial domain adaptation network, aiming to improve the adaptive ability of the model and generate high-quality text embedding for new classes. Extensive experiments are conducted on four benchmark datasets and our method demonstrates clear superiority over the state-of-the-art models in all the datasets. In particular, the accuracy of 1-shot and 5-shot classification on the dataset of 20 Newsgroups is boosted from 52.1% to 59.6%, and from 68.3% to 77.8%, respectively. | ['Aoying Zhou', 'Ming Gao', 'Minghui Qiu', 'Dongxiang Zhang', 'Zeqiu Fan', 'Chengcheng Han'] | 2021-07-26 | null | https://aclanthology.org/2021.findings-acl.145 | https://aclanthology.org/2021.findings-acl.145.pdf | findings-acl-2021-8 | ['few-shot-text-classification'] | ['natural-language-processing'] | [ 5.77505790e-02 -2.02293515e-01 -3.01928341e-01 -4.08312023e-01
-5.96661568e-01 -6.94637373e-02 8.38545203e-01 2.47907013e-01
-6.63698196e-01 8.62622440e-01 1.64815813e-01 2.19644129e-01
9.21493769e-03 -8.40220094e-01 -2.84866393e-01 -6.70630276e-01
2.30997294e-01 2.03551605e-01 3.43205959e-01 -4.63513255e-01
3.60707998e-01 -9.25883129e-02 -1.69127607e+00 4.99850303e-01
8.80496562e-01 9.63845968e-01 -5.09132519e-02 4.03434664e-01
-5.19220412e-01 3.98840785e-01 -6.72283947e-01 -6.41176224e-01
-1.77893625e-03 -2.82873720e-01 -4.33737785e-01 -2.97508448e-01
-7.76353106e-02 -3.64543051e-02 -5.18045366e-01 1.05737674e+00
6.51959479e-01 3.86960655e-01 7.65106201e-01 -1.09495533e+00
-1.00123584e+00 7.46889353e-01 -6.36257648e-01 2.67209053e-01
1.21433765e-01 1.16414979e-01 9.12949264e-01 -9.98106241e-01
5.63917637e-01 9.03775990e-01 5.37882328e-01 8.78945529e-01
-1.00506508e+00 -9.28176403e-01 1.33274704e-01 4.51745063e-01
-1.20424557e+00 -4.74106789e-01 9.43397164e-01 -2.91063279e-01
7.67908275e-01 -6.08680286e-02 3.04197818e-01 1.34660757e+00
2.12663770e-01 8.09567571e-01 1.11496019e+00 -6.62222266e-01
4.06818777e-01 5.17666101e-01 3.09663713e-01 2.97534108e-01
1.77041426e-01 -4.64727730e-02 -5.25414467e-01 -6.89630955e-02
-6.37419298e-02 1.70859054e-01 -7.62313381e-02 -3.81257772e-01
-9.35973942e-01 9.76249874e-01 3.52864176e-01 6.73021019e-01
-3.14746827e-01 -3.28145415e-01 8.09586644e-01 1.10893190e-01
1.05875361e+00 3.40875357e-01 -4.70044106e-01 -2.38563597e-01
-6.99329317e-01 1.77863210e-01 4.85731095e-01 7.07742751e-01
5.05838454e-01 1.87040374e-01 -2.96033263e-01 1.09457481e+00
4.25062738e-02 2.63487279e-01 1.20122898e+00 -1.44215867e-01
5.13176143e-01 8.25731516e-01 -1.66831300e-01 -1.04323506e+00
-4.41647440e-01 -8.26128900e-01 -8.98142278e-01 -5.20338863e-02
-3.26917996e-03 -2.13070497e-01 -9.26688671e-01 1.69690073e+00
4.18963611e-01 3.66089046e-01 3.71011168e-01 3.92821103e-01
8.24395955e-01 9.49871719e-01 3.45268428e-01 -2.42944226e-01
1.09489679e+00 -1.11049426e+00 -7.06635952e-01 -3.06893319e-01
7.39285350e-01 -5.19032657e-01 1.16182339e+00 9.14483666e-02
-7.35986233e-01 -8.03257167e-01 -1.20525038e+00 3.43484551e-01
-7.60552764e-01 -2.28310853e-01 2.70910352e-01 7.71325469e-01
-4.23227757e-01 5.65058112e-01 -2.89537340e-01 -3.91931027e-01
6.00697577e-01 2.16773171e-02 -3.90998930e-01 -1.79511011e-01
-1.54159880e+00 9.63686466e-01 8.48160803e-01 -4.38999265e-01
-6.29933834e-01 -9.41324711e-01 -5.34974813e-01 2.06782222e-01
2.84819096e-01 -6.14952981e-01 1.09379101e+00 -8.17884088e-01
-1.53445041e+00 7.85882056e-01 2.89637804e-01 -5.08918405e-01
5.83950639e-01 -1.55211031e-01 -8.28577459e-01 3.20091210e-02
1.91360302e-02 2.86749959e-01 1.03927922e+00 -1.06587613e+00
-9.13505077e-01 -4.95516479e-01 -1.64668873e-01 1.75330222e-01
-1.19613671e+00 -1.91003576e-01 6.41279593e-02 -7.69176126e-01
-3.89660627e-01 -6.48746014e-01 8.75610393e-03 -2.17571139e-01
-5.03097884e-02 -3.60228151e-01 1.02780104e+00 -4.95126218e-01
1.44797850e+00 -2.12754703e+00 1.27219588e-01 -3.63905340e-01
-3.33340988e-02 8.56915176e-01 -1.03231698e-01 5.48424542e-01
3.41699123e-02 -5.94752207e-02 -7.71028548e-02 -3.89642298e-01
1.34895965e-01 -3.44210975e-02 -4.01101500e-01 2.43936867e-01
-3.25539596e-02 8.42833042e-01 -8.27942550e-01 -2.26938874e-01
2.60088384e-01 4.21768695e-01 -3.45151246e-01 1.46544367e-01
-7.69755319e-02 1.31620467e-01 -4.36798513e-01 2.95466542e-01
5.30801415e-01 2.46595070e-02 -5.90751693e-02 1.03915028e-01
8.51621479e-02 -1.32671818e-01 -8.63756359e-01 1.71685016e+00
-6.44928038e-01 2.00716659e-01 -5.12560487e-01 -1.16753042e+00
1.15857732e+00 2.66922444e-01 2.36223593e-01 -7.88306117e-01
3.83213699e-01 6.39971048e-02 -8.67647976e-02 -5.06979167e-01
4.83006567e-01 -4.71319377e-01 -2.06909925e-01 4.70706284e-01
3.42850208e-01 3.64650697e-01 4.24603783e-02 -1.74365565e-02
7.31638730e-01 -1.55512825e-01 4.56850976e-01 -1.91763788e-01
8.18045497e-01 -1.86416343e-01 2.47448757e-01 5.43293357e-01
-3.38945895e-01 1.89393193e-01 1.05903812e-01 -4.60491091e-01
-1.12919724e+00 -6.61500514e-01 -1.71848685e-01 1.44457173e+00
-4.64313030e-02 -1.92865327e-01 -6.39117301e-01 -9.59669054e-01
2.55346894e-02 1.26021791e+00 -9.70959783e-01 -8.29185724e-01
-2.20931947e-01 -9.73094583e-01 3.85880262e-01 4.83993858e-01
6.39967144e-01 -8.88298929e-01 -4.20674950e-01 3.96264821e-01
4.27293638e-03 -9.33042943e-01 -3.31853688e-01 -4.71550934e-02
-8.88468683e-01 -8.11589539e-01 -9.62172151e-01 -8.08174551e-01
4.05060053e-01 3.90493631e-01 5.46112418e-01 -1.32673353e-01
-2.19087437e-01 1.68513712e-02 -8.12368810e-01 -4.26356345e-01
-5.37421703e-01 4.57907647e-01 2.27431923e-01 3.60021919e-01
5.87681293e-01 -4.39561218e-01 -2.91952223e-01 2.03113616e-01
-1.01404667e+00 -8.00973251e-02 3.59705955e-01 1.26525533e+00
6.52001128e-02 2.21424520e-01 1.18852568e+00 -1.12428784e+00
7.70723820e-01 -7.19457209e-01 -1.97369739e-01 2.33809352e-01
-1.02102435e+00 -5.30166924e-02 1.04541576e+00 -6.38490856e-01
-1.29393208e+00 -4.01885748e-01 -1.38251796e-01 -2.29288995e-01
-9.89328846e-02 5.28906643e-01 -8.60168114e-02 1.91112325e-01
8.34269464e-01 5.36574006e-01 -9.92701128e-02 -4.20513123e-01
5.37390888e-01 1.13971317e+00 2.86439776e-01 -2.90328473e-01
8.57706130e-01 4.49582875e-01 -3.25028598e-01 -6.79024339e-01
-1.09788895e+00 -6.24955654e-01 -7.71879137e-01 1.56751294e-02
6.52816534e-01 -7.72943735e-01 -1.46219030e-01 6.74286485e-01
-7.62549043e-01 1.09762549e-02 -3.56312782e-01 3.79549772e-01
-3.98699939e-01 3.58568311e-01 -2.39152789e-01 -7.23755538e-01
-8.11196208e-01 -6.85955048e-01 5.98869741e-01 3.31187546e-01
5.13036400e-02 -1.17849135e+00 2.45325580e-01 2.60223925e-01
7.56857693e-01 5.44285364e-02 1.04145277e+00 -1.21564436e+00
3.24866116e-01 -6.31621480e-01 -1.03247635e-01 4.60152000e-01
7.54000023e-02 -2.34875008e-01 -1.26717746e+00 -3.99445564e-01
1.38590544e-01 -3.19179446e-01 8.24240625e-01 -7.32661085e-03
9.76048648e-01 -2.12799326e-01 -3.67226720e-01 3.10351461e-01
1.37368977e+00 3.65795553e-01 5.49671829e-01 6.54784858e-01
3.34184200e-01 4.89394724e-01 7.01340973e-01 6.54092193e-01
2.46618703e-01 6.97409570e-01 2.97923893e-01 4.98403043e-01
-5.96184507e-02 -4.35480297e-01 1.95274159e-01 9.93036866e-01
2.49449819e-01 -1.83650687e-01 -8.55211318e-01 4.40374047e-01
-1.87031007e+00 -1.10469067e+00 2.01537415e-01 2.08764625e+00
7.48182356e-01 4.06096220e-01 -1.25906207e-02 3.16283137e-01
9.40807581e-01 5.05703568e-01 -5.30637980e-01 -2.52773255e-01
5.69054671e-02 2.30060667e-01 3.44448052e-02 3.25020440e-02
-1.28926802e+00 1.02195895e+00 5.35267687e+00 9.86670971e-01
-1.00525153e+00 4.04498279e-01 5.24520516e-01 -1.19786881e-01
1.66705847e-02 -3.22442025e-01 -9.33600783e-01 7.16194928e-01
1.10067964e+00 -7.05109477e-01 8.08663145e-02 1.14430392e+00
-3.54716688e-01 3.52700055e-01 -7.07280278e-01 8.86685729e-01
5.49308300e-01 -1.27063668e+00 2.10085928e-01 -2.66923308e-01
1.00113142e+00 -1.86568007e-01 2.99974918e-01 9.55696523e-01
-1.34473205e-01 -7.42856324e-01 4.08538789e-01 4.67243195e-01
7.30872691e-01 -1.04718852e+00 1.03091371e+00 7.33319163e-01
-8.27120900e-01 -3.51549238e-01 -6.52884960e-01 -3.87287438e-02
1.24701977e-01 3.70097995e-01 -6.85737669e-01 5.42454839e-01
5.56460321e-01 7.17661798e-01 -7.43655324e-01 8.39357793e-01
-1.44577667e-01 4.46954548e-01 1.44320413e-01 -5.53534567e-01
2.65393704e-01 2.40030453e-01 4.13440526e-01 1.10280526e+00
1.57155603e-01 4.22947183e-02 1.05660222e-02 3.46778363e-01
-3.09492767e-01 6.19495332e-01 -6.18192315e-01 -8.91309902e-02
3.48016649e-01 1.26609039e+00 -1.59275219e-01 -5.91569006e-01
-4.31605875e-01 8.63423049e-01 4.41322088e-01 -1.18513824e-02
-8.57232273e-01 -9.33776379e-01 3.70844007e-01 -8.12794939e-02
4.28259075e-01 2.30404645e-01 -1.95251957e-01 -1.38244200e+00
-1.02326848e-01 -8.15210819e-01 5.35643458e-01 -5.18764019e-01
-1.65297198e+00 7.27707565e-01 -1.39860049e-01 -1.17846537e+00
-1.92311659e-01 -5.43485105e-01 -6.64870799e-01 6.43168926e-01
-1.63467669e+00 -1.27612770e+00 -2.88661182e-01 4.79631931e-01
7.17855036e-01 -8.25312257e-01 9.90885794e-01 3.48619998e-01
-4.44287598e-01 9.31088269e-01 7.10333765e-01 5.17734326e-02
8.44691098e-01 -8.32873285e-01 3.76056761e-01 6.61273181e-01
2.00921800e-02 3.72947276e-01 5.97037494e-01 -3.75679761e-01
-1.16754675e+00 -1.20523953e+00 8.73252809e-01 -3.55557114e-01
8.65605533e-01 -2.26995841e-01 -1.21501637e+00 4.15588707e-01
2.30190158e-01 -3.85736339e-02 9.27188635e-01 1.16910964e-01
-6.03278458e-01 -2.96340108e-01 -1.29097843e+00 6.97006404e-01
7.12174952e-01 -4.12500948e-01 -9.78917181e-01 2.03593567e-01
7.82363653e-01 -9.60153490e-02 -8.06819558e-01 3.07303160e-01
6.37078643e-01 -8.08187902e-01 8.17149103e-01 -1.12095737e+00
5.97602785e-01 1.51807189e-01 -3.45394850e-01 -1.59235227e+00
-5.96411824e-01 -1.08642995e-01 -3.70027363e-01 1.46102560e+00
2.39619583e-01 -9.00574982e-01 5.45123994e-01 3.69204313e-01
-9.59126502e-02 -7.30179489e-01 -1.10542786e+00 -7.36792505e-01
3.82147402e-01 -2.88875133e-01 5.14356971e-01 1.28886580e+00
2.85565853e-01 7.26561904e-01 -4.92904127e-01 -3.23827296e-01
5.50597906e-01 -1.11413803e-02 8.39031279e-01 -1.35672581e+00
-1.95459574e-01 -4.79034603e-01 -5.64192295e-01 -4.17224139e-01
4.44373548e-01 -1.12741077e+00 -3.23471367e-01 -1.29629302e+00
3.43645483e-01 -1.77419081e-01 -7.78343201e-01 2.97879964e-01
-3.90589088e-01 1.63608432e-01 2.53123939e-01 3.92321609e-02
-5.75606227e-01 1.05119240e+00 9.67739761e-01 -3.77495795e-01
-7.54610747e-02 1.94450438e-01 -8.31779838e-01 6.38174057e-01
1.10009098e+00 -6.20653212e-01 -4.24140543e-01 -2.07099795e-01
-1.43945336e-01 -2.82603920e-01 -6.64726924e-03 -1.07068515e+00
2.71198571e-01 -4.45493571e-02 4.54096496e-01 -2.49753281e-01
3.34649295e-01 -6.62458658e-01 -2.16516405e-01 6.32973194e-01
-5.55610895e-01 -1.93684384e-01 2.69429684e-01 9.00440574e-01
-1.43213660e-01 -6.70277178e-01 1.07275176e+00 -3.08878943e-02
-8.93728077e-01 3.50248009e-01 3.43599007e-03 3.40269774e-01
1.42416167e+00 -1.00392643e-02 -5.85813403e-01 -4.79495041e-02
-4.79985029e-01 -2.98074968e-02 2.66007006e-01 9.79345679e-01
5.22108078e-01 -1.43238139e+00 -7.39728272e-01 2.93565571e-01
6.09893560e-01 -5.87116599e-01 6.56737328e-01 4.43390816e-01
-6.12145923e-02 4.14721787e-01 -3.42167348e-01 -2.63419658e-01
-9.81123745e-01 9.30402160e-01 3.19199450e-03 -3.71348590e-01
-6.35107696e-01 6.25552714e-01 -1.75904065e-01 -4.17921841e-01
1.38324633e-01 1.46400064e-01 -4.43300158e-01 2.99498379e-01
8.44007969e-01 4.96004641e-01 4.10521850e-02 -5.77986658e-01
-2.14004457e-01 5.37633419e-01 -4.91012752e-01 3.76830921e-02
1.43017876e+00 -4.82696034e-02 3.33635658e-01 7.60171235e-01
1.32177520e+00 -2.56214112e-01 -9.27625716e-01 -5.46985567e-01
3.21644768e-02 -5.48174381e-01 4.55080383e-02 -8.70734692e-01
-7.43034422e-01 1.15316474e+00 7.29328811e-01 2.19017491e-01
7.93340743e-01 -3.85204613e-01 1.02774906e+00 3.29051882e-01
4.50011134e-01 -1.28752291e+00 3.72733325e-01 5.17245233e-01
5.21910965e-01 -1.33770275e+00 -4.14572842e-02 2.17434213e-01
-7.91643500e-01 1.13202298e+00 6.18504405e-01 -2.90461211e-03
7.74487317e-01 -2.85991609e-01 -1.14105837e-02 2.42838740e-01
-9.41843092e-01 7.66153783e-02 3.32357258e-01 5.75155318e-01
2.39081949e-01 -1.75326541e-02 -3.98815244e-01 8.36715877e-01
1.42698005e-01 8.22126269e-02 3.41213793e-01 8.67703497e-01
-8.55088949e-01 -1.06282043e+00 1.20313749e-01 5.04632473e-01
-5.13386190e-01 -4.21416797e-02 7.69297853e-02 7.30804861e-01
-4.90440801e-02 9.07502890e-01 -1.72218755e-01 -5.45870602e-01
4.36731637e-01 6.74784064e-01 9.85024273e-02 -8.42787206e-01
-5.23987293e-01 -4.27477658e-01 -1.77222416e-01 1.27608344e-01
-3.43679994e-01 -3.42156708e-01 -8.71833146e-01 -2.40122125e-01
-4.46042567e-01 2.39187911e-01 6.80830419e-01 9.25115764e-01
4.74567503e-01 6.55544996e-01 8.59355927e-01 -3.88123065e-01
-1.11052990e+00 -1.20379770e+00 -5.61813176e-01 5.98922372e-01
-5.11579625e-02 -8.36993992e-01 -4.76701230e-01 -1.51908085e-01] | [10.198762893676758, 3.521707057952881] |
34f87b13-c613-402a-93b8-08c3c1c7cc10 | traditional-methods-in-edge-corner-and | 2208.07714 | null | https://arxiv.org/abs/2208.07714v1 | https://arxiv.org/pdf/2208.07714v1.pdf | Traditional methods in Edge, Corner and Boundary detection | This is a review paper of traditional approaches for edge, corner, and boundary detection methods. There are many real-world applications of edge, corner, and boundary detection methods. For instance, in medical image analysis, edge detectors are used to extract the features from the given image. In modern innovations like autonomous vehicles, edge detection and segmentation are the most crucial things. If we want to detect motion or track video, corner detectors help. I tried to compare the results of detectors stage-wise wherever it is possible and also discussed the importance of image prepossessing to minimise the noise. Real-world images are used to validate detector performance and limitations. | ['Sai Pavan Tadem'] | 2022-08-12 | null | null | null | null | ['edge-detection', 'boundary-detection'] | ['computer-vision', 'computer-vision'] | [ 9.03196484e-02 -4.30089712e-01 -2.84719616e-01 -8.38462710e-02
-6.04562052e-02 -3.16874534e-01 1.28903553e-01 1.83915168e-01
-6.02713883e-01 4.42442745e-01 -1.55425310e-01 -3.32590163e-01
1.11691065e-01 -5.34688711e-01 -8.85958672e-02 -5.75980783e-01
-1.01791777e-01 2.09637750e-02 8.68545890e-01 -6.49644202e-03
7.07424819e-01 8.45736265e-01 -1.33082235e+00 1.66942939e-01
5.01936913e-01 8.32054734e-01 1.37962967e-01 1.03813362e+00
-3.05645972e-01 2.87057579e-01 -3.50611985e-01 2.42410358e-02
4.37321514e-01 -5.36197543e-01 -3.92203093e-01 5.61808586e-01
9.31998901e-03 -1.62062854e-01 -1.75933719e-01 1.12236464e+00
4.09273714e-01 -7.49296993e-02 8.43661427e-01 -1.25580609e+00
6.51415214e-02 5.09911776e-02 -9.56452012e-01 7.02744424e-01
4.88092780e-01 -1.05079515e-02 2.73176670e-01 -9.56474185e-01
9.40667033e-01 8.66204381e-01 9.25283909e-01 4.32662576e-01
-7.23177969e-01 -3.54732066e-01 -2.68754303e-01 2.81216234e-01
-1.51168728e+00 -3.37088615e-01 6.36692584e-01 -3.21908534e-01
7.60192752e-01 2.68272966e-01 8.00076902e-01 3.69473815e-01
9.54046786e-01 9.72811341e-01 8.13909888e-01 -6.91066742e-01
1.47947311e-01 3.59540612e-01 2.17129126e-01 9.35464978e-01
6.56466842e-01 1.32507369e-01 -6.05007559e-02 5.25263809e-02
9.51412380e-01 2.01488301e-01 -3.86501431e-01 -3.24462295e-01
-1.22831345e+00 5.75609863e-01 1.70253888e-01 5.80625534e-01
-4.08614755e-01 1.16689354e-02 4.86603081e-01 9.71477330e-02
-2.85920803e-04 1.37718543e-01 -1.69931024e-01 -7.95475245e-02
-1.22286248e+00 -1.28009528e-01 6.65719151e-01 9.40656662e-01
4.54994828e-01 -1.11110903e-01 9.88890007e-02 3.21098924e-01
3.99928987e-01 3.08493406e-01 5.08212984e-01 -8.31562757e-01
-1.25330940e-01 7.44491816e-01 7.27509288e-03 -1.14867198e+00
-6.58395231e-01 7.24563897e-02 -6.91742539e-01 7.52869427e-01
4.94002163e-01 -2.38129258e-01 -1.41450727e+00 3.50704640e-01
4.83445168e-01 2.10502759e-01 -1.83314145e-01 7.48375416e-01
1.05818391e+00 4.84447420e-01 -1.14557981e-01 -2.99572647e-01
1.67477441e+00 -8.32047820e-01 -1.00650871e+00 -4.27373350e-01
8.60103786e-01 -1.11465788e+00 1.31643519e-01 4.89222616e-01
-8.40175390e-01 -4.37418759e-01 -1.01330793e+00 1.99937254e-01
-5.27796507e-01 1.50975093e-01 6.35677278e-01 6.63399875e-01
-7.51129508e-01 3.93793643e-01 -7.45987594e-01 -6.06904089e-01
1.25762716e-01 4.38272744e-01 -5.91207564e-01 5.95253445e-02
-8.45000207e-01 8.67488325e-01 4.99130815e-01 2.91800052e-01
-5.42289317e-02 7.89742842e-02 -8.21605623e-01 -3.06535304e-01
2.98400849e-01 -5.35599530e-01 9.41542089e-01 -8.89036953e-01
-1.08364642e+00 1.01562691e+00 -3.22520822e-01 -1.32760510e-01
6.77927375e-01 6.76919892e-02 -4.19844627e-01 3.90481681e-01
-8.12977627e-02 5.08918405e-01 9.30880189e-01 -9.87447500e-01
-1.39553738e+00 1.82802435e-02 -4.30764109e-01 2.84828227e-02
4.01907057e-01 6.48505017e-02 -6.39018357e-01 -3.85042340e-01
5.47978759e-01 -7.96651006e-01 -4.69539642e-01 1.18064575e-01
-4.10570741e-01 -1.31018162e-01 1.29152787e+00 -7.24896014e-01
1.22725368e+00 -2.39725184e+00 -4.51366723e-01 4.53632981e-01
1.98168576e-01 3.86704236e-01 3.14701259e-01 2.96924591e-01
1.18838161e-01 6.02286588e-03 -9.03575274e-04 1.17493212e-01
-5.88612854e-01 6.40853494e-02 5.21287978e-01 7.32411385e-01
-6.51061758e-02 6.80292010e-01 -6.69427097e-01 -1.18570960e+00
5.74377596e-01 6.14579260e-01 -1.67278722e-01 -3.80554646e-01
5.41516244e-01 3.98249507e-01 -5.23621678e-01 1.00683725e+00
4.73960370e-01 1.83812797e-01 -4.03355449e-01 -2.72969246e-01
-2.35841721e-01 -4.45255905e-01 -1.77811182e+00 1.08260977e+00
-1.78183317e-02 1.17471027e+00 2.56583214e-01 -9.73630965e-01
7.86027849e-01 5.12739003e-01 6.72356904e-01 -5.85229456e-01
5.55282712e-01 4.91319329e-01 1.44159213e-01 -7.96386302e-01
6.87456131e-01 4.61797304e-02 9.29189846e-02 -6.13965187e-03
-4.48303223e-01 -1.29330575e-01 4.50386494e-01 2.98704598e-02
9.23034012e-01 5.04338071e-02 9.49397922e-01 -1.72287807e-01
6.12573385e-01 5.12138188e-01 6.72536612e-01 6.85163498e-01
-6.48868680e-01 8.63442481e-01 1.57867491e-01 -5.22880554e-01
-9.27415013e-01 -5.62318742e-01 -3.10733616e-01 3.60053420e-01
5.78275084e-01 -8.17177296e-02 -8.41775000e-01 -5.26950896e-01
-7.50956312e-02 1.59520641e-01 -5.32590449e-01 1.54178282e-02
-5.42662501e-01 -4.14214969e-01 8.13529491e-02 6.43989801e-01
5.36178946e-01 -1.05647743e+00 -1.37322962e+00 4.81783867e-01
2.02932790e-01 -1.04868174e+00 -5.76254785e-01 3.00125062e-01
-1.21604955e+00 -1.04208922e+00 -8.02507102e-01 -1.08264005e+00
9.94662166e-01 7.54293442e-01 7.09351182e-01 5.36461771e-01
-9.31041420e-01 4.70255435e-01 -4.77316916e-01 -3.24148268e-01
-2.25589126e-01 -2.66471624e-01 -1.33350685e-01 -3.16909313e-01
4.67069626e-01 5.99788986e-02 -8.78860056e-01 4.76642281e-01
-6.45549715e-01 -2.28006035e-01 8.20910096e-01 7.52926290e-01
5.69933772e-01 3.21743816e-01 -9.46772993e-02 -8.47949147e-01
5.59688032e-01 -8.81580263e-02 -5.20240366e-01 2.04427078e-01
-5.16345322e-01 -2.04426274e-01 5.63486293e-02 -3.65560621e-01
-7.18335390e-01 5.67978501e-01 -7.63199329e-02 -9.98085290e-02
-4.15738016e-01 2.41026267e-01 1.54665336e-01 -2.84445226e-01
6.72890842e-01 1.35922683e-02 -8.06894749e-02 -7.96180144e-02
1.98410600e-01 1.14814305e+00 5.58213770e-01 3.45140189e-01
4.20760989e-01 8.23588371e-01 1.10827424e-01 -1.43727589e+00
-1.50305957e-01 -1.27810037e+00 -9.37620819e-01 -5.54072380e-01
1.00711679e+00 -3.51898581e-01 -5.30431509e-01 2.49591991e-01
-1.12224090e+00 1.63989931e-01 -1.88830346e-01 7.45992422e-01
-4.05762404e-01 4.10683066e-01 -3.29918563e-01 -1.08151484e+00
-4.21973884e-01 -1.15085435e+00 9.58448410e-01 9.57477391e-01
-2.71277577e-01 -1.32764876e+00 -2.58607537e-01 -7.55846500e-02
2.93513924e-01 3.70275646e-01 1.90176010e-01 -4.58159298e-01
-4.07633781e-01 -9.65068519e-01 -7.72891268e-02 -1.62659019e-01
3.96335512e-01 5.61063826e-01 -7.41456449e-01 -1.28735807e-02
5.14651351e-02 7.35114396e-01 8.23429406e-01 8.61293018e-01
3.74265760e-01 1.79479435e-01 -1.21021163e+00 4.87981528e-01
1.45484376e+00 9.40357864e-01 8.87661040e-01 4.92586970e-01
4.05432850e-01 3.95323128e-01 8.04814696e-01 1.38357401e-01
-2.37795159e-01 4.33839828e-01 -1.42335230e-02 -4.30687398e-01
-1.41249672e-01 1.33527935e-01 8.84962827e-02 4.97955501e-01
-6.98021352e-02 -1.02957025e-01 -8.96117508e-01 6.66569710e-01
-1.63129294e+00 -1.01768744e+00 -7.30478525e-01 2.09811211e+00
1.45325080e-01 5.12924075e-01 2.21252471e-01 4.76778656e-01
9.53181386e-01 -5.28092325e-01 -2.91268408e-01 -5.71287453e-01
2.05342516e-01 -1.40043885e-01 1.00848007e+00 4.73769128e-01
-1.37472558e+00 7.45574832e-01 7.39903402e+00 6.12090290e-01
-1.30966759e+00 -2.47951731e-01 5.96271455e-01 5.85533619e-01
3.62392247e-01 5.06995134e-02 -1.07436728e+00 3.33478421e-01
3.88168663e-01 -3.19474712e-02 -2.48914212e-01 9.68978226e-01
2.61909127e-01 -1.08180690e+00 -7.94052005e-01 1.21180439e+00
2.87560113e-02 -1.03818429e+00 -6.69040859e-01 -5.95228076e-02
4.48283195e-01 -4.19059098e-01 -4.30260777e-01 -1.53149784e-01
-2.90996999e-01 -8.09107065e-01 3.09739619e-01 5.56658566e-01
2.22008005e-01 -5.76849997e-01 1.05425119e+00 3.09610397e-01
-1.48440111e+00 1.58181787e-03 -2.71760851e-01 -5.57168983e-02
5.76601982e-01 6.33301556e-01 -9.81985986e-01 2.53094584e-01
6.26588523e-01 3.93138707e-01 -4.04926807e-01 1.88432014e+00
-1.85153842e-01 1.27986938e-01 -6.84826553e-01 -1.49585322e-01
3.19047689e-01 -3.18770289e-01 6.64263964e-01 1.51098371e+00
2.42446586e-01 1.37023047e-01 1.29960239e-01 2.52474815e-01
6.49841905e-01 2.48040468e-01 -9.29885149e-01 1.31577283e-01
3.03035170e-01 1.62091792e+00 -1.94828069e+00 -4.24233913e-01
-5.10819256e-01 1.29295337e+00 -6.04249537e-01 1.92042127e-01
-6.16130531e-01 -8.91771674e-01 3.16155702e-01 3.92606109e-01
3.48121434e-01 -3.71432960e-01 -3.97395760e-01 -5.52700102e-01
-3.66963506e-01 -4.09050047e-01 5.07124424e-01 -5.03493369e-01
-6.81490242e-01 3.90736759e-01 9.83961895e-02 -1.46642280e+00
-2.45275185e-01 -7.72207916e-01 -9.61324871e-01 4.31973845e-01
-1.06100833e+00 -6.76358402e-01 -3.16040069e-01 1.53000221e-01
8.65571260e-01 1.70441449e-01 9.58114937e-02 2.55957186e-01
-4.28620875e-01 2.89010316e-01 2.64751852e-01 5.57899296e-01
3.78975093e-01 -9.87330079e-01 4.28789496e-01 1.11311412e+00
2.74362773e-01 4.67671037e-01 9.31498349e-01 -8.79489601e-01
-1.21141183e+00 -4.72553164e-01 6.75217927e-01 -1.42076194e-01
1.43924281e-01 5.89286834e-02 -7.85318494e-01 4.34004843e-01
2.02697814e-01 3.73915553e-01 1.76514462e-01 -6.21179581e-01
6.33353531e-01 1.47462115e-01 -1.37411368e+00 6.65550530e-01
7.61130691e-01 8.73027369e-02 -5.83739579e-01 5.69369830e-02
-1.51236057e-01 -5.96442640e-01 -4.51476485e-01 2.73260981e-01
8.67604911e-01 -1.22533631e+00 8.38604748e-01 -1.06597245e-01
-2.95878500e-01 -5.49576819e-01 4.61154610e-01 -1.10405946e+00
-3.11520904e-01 -5.22251785e-01 3.27683508e-01 9.62392807e-01
4.56689388e-01 -6.28391445e-01 1.06651235e+00 4.23350960e-01
3.41448896e-02 -4.74615186e-01 -9.10026431e-01 -8.26540112e-01
-6.20928645e-01 -1.65541664e-01 -2.20659286e-01 7.03282595e-01
2.13833585e-01 3.65343578e-02 -8.98116007e-02 -5.55110388e-02
5.85830629e-01 -1.87335432e-01 6.40751004e-01 -1.22058737e+00
2.39493206e-01 -5.07784843e-01 -1.12489092e+00 -8.60852778e-01
-6.13495767e-01 -3.41124922e-01 2.94393957e-01 -2.12655354e+00
-8.17737635e-03 -2.00845212e-01 2.31173962e-01 1.76198334e-01
-3.33462387e-01 5.78683436e-01 6.51298761e-02 6.00888692e-02
-3.61067861e-01 -4.38235283e-01 1.23808575e+00 1.42426357e-01
-4.85180587e-01 1.90515801e-01 -2.29870126e-01 1.05218327e+00
8.67242813e-01 -6.26225471e-01 -1.39428943e-01 2.24049985e-01
2.48762425e-02 2.31161285e-02 -5.26515283e-02 -1.17674053e+00
6.61833346e-01 -1.18813924e-01 5.78195512e-01 -8.39374602e-01
1.03534825e-01 -1.24318087e+00 1.85410678e-01 9.96195555e-01
4.91929471e-01 3.58289629e-01 5.39165288e-02 4.37930256e-01
-2.32814893e-01 -7.87007153e-01 9.51974928e-01 -4.55407560e-01
-1.41042137e+00 -1.29197225e-01 -6.96972549e-01 -1.81913733e-01
1.75212646e+00 -1.25718939e+00 1.45652115e-01 -4.56507206e-01
-7.01986492e-01 2.50254869e-01 4.63867188e-01 1.55937985e-01
9.20646548e-01 -1.06386602e+00 -4.83559161e-01 4.76406068e-01
1.49018727e-02 -9.08480659e-02 -4.73479219e-02 1.28820479e+00
-1.22508979e+00 2.53419608e-01 -3.30126226e-01 -8.52363527e-01
-1.71027756e+00 7.66348541e-01 4.85357642e-01 2.77386397e-01
-1.12377834e+00 6.73535705e-01 -1.18470490e-01 4.59316880e-01
1.48868725e-01 -5.51197350e-01 -3.92699510e-01 1.99889198e-01
6.44094050e-01 6.31473541e-01 -1.02148689e-01 -7.02274263e-01
-6.99893892e-01 1.12866366e+00 -1.25888944e-01 -1.17469661e-01
6.50969565e-01 -3.09088260e-01 3.30678314e-01 4.03981328e-01
7.82675803e-01 1.21732756e-01 -9.89240706e-01 2.25163952e-01
4.27725732e-01 -6.25875652e-01 2.17591926e-01 -1.44501701e-01
-1.11963201e+00 7.05411553e-01 9.70593572e-01 4.00376201e-01
1.12286448e+00 -1.11322217e-01 6.54942989e-01 4.34554256e-02
3.77997458e-01 -1.48963630e+00 -3.31240982e-01 2.12193504e-01
2.65633792e-01 -1.35829139e+00 3.97054702e-01 -6.18953943e-01
-5.94953001e-01 1.40669370e+00 3.04321706e-01 -3.52061093e-01
9.17255461e-01 8.40368807e-01 3.61158520e-01 -1.99793771e-01
-1.49118081e-01 -5.04338801e-01 2.59305865e-01 7.83532381e-01
5.55413425e-01 -1.85450345e-01 -8.08973074e-01 -3.20688725e-01
4.65686262e-01 1.85120314e-01 5.93223572e-01 1.35259330e+00
-1.03312135e+00 -8.99279952e-01 -9.37659442e-01 4.97125417e-01
-6.58238173e-01 2.56361663e-01 -3.55224192e-01 1.21177578e+00
5.92264906e-02 9.58774388e-01 5.86759821e-02 -2.38479689e-01
3.75639677e-01 1.74274333e-02 2.76953399e-01 -2.07859948e-01
-3.78157258e-01 2.88721740e-01 -1.22337259e-01 -4.70488310e-01
-3.48867267e-01 -6.31890237e-01 -1.75856483e+00 -3.04732602e-02
-7.43283451e-01 1.68888196e-01 9.46998000e-01 9.45289016e-01
7.11243525e-02 4.03356135e-01 3.87952924e-01 -9.02557611e-01
1.98144227e-01 -6.84991777e-01 -7.57679641e-01 3.74570429e-01
2.88906842e-01 -6.50874555e-01 -5.11133432e-01 3.20235848e-01] | [9.137112617492676, -1.3715037107467651] |
2fcc176d-1199-47e7-ab76-7570752e06fa | detecting-melanoma-fairly-skin-tone-detection | 2202.02832 | null | https://arxiv.org/abs/2202.02832v4 | https://arxiv.org/pdf/2202.02832v4.pdf | Detecting Melanoma Fairly: Skin Tone Detection and Debiasing for Skin Lesion Classification | Convolutional Neural Networks have demonstrated human-level performance in the classification of melanoma and other skin lesions, but evident performance disparities between differing skin tones should be addressed before widespread deployment. In this work, we propose an efficient yet effective algorithm for automatically labelling the skin tone of lesion images, and use this to annotate the benchmark ISIC dataset. We subsequently use these automated labels as the target for two leading bias unlearning techniques towards mitigating skin tone bias. Our experimental results provide evidence that our skin tone detection algorithm outperforms existing solutions and that unlearning skin tone may improve generalisation and can reduce the performance disparity between melanoma detection in lighter and darker skin tones. | ['Amir Atapour-Abarghouei', 'Peter J. Bevan'] | 2022-02-06 | null | null | null | null | ['skin-lesion-classification'] | ['medical'] | [ 1.15856826e+00 1.55760139e-01 -4.90063667e-01 -3.78829300e-01
-1.10166979e+00 -4.59596574e-01 5.38190484e-01 -8.30754191e-02
-5.88308811e-01 4.97280657e-01 3.53063226e-01 -3.82359058e-01
-6.90992698e-02 -4.01896685e-01 8.37359801e-02 -8.61082256e-01
2.38239855e-01 -3.26075912e-01 2.00520083e-01 -1.64300799e-01
3.35283846e-01 3.30972999e-01 -1.24032068e+00 7.51728356e-01
5.43496490e-01 9.73756135e-01 -4.68817204e-01 1.29275835e+00
4.32229154e-02 8.03907812e-01 -1.00326514e+00 -2.75687963e-01
2.17829019e-01 -5.18611372e-01 -8.80433738e-01 -1.26279797e-02
1.16127276e+00 -1.75210446e-01 -3.96147631e-02 1.12426412e+00
1.05086505e+00 -6.22200429e-01 6.30641639e-01 -7.32969105e-01
-3.29145283e-01 4.15824711e-01 -9.14889514e-01 4.80343997e-01
-5.74889369e-02 3.89479220e-01 6.83265448e-01 -1.48547947e-01
5.31083703e-01 8.94966781e-01 1.25886381e+00 1.09106767e+00
-1.22769845e+00 -5.98403931e-01 -3.15432280e-01 2.14925352e-02
-1.37113869e+00 -4.93696272e-01 6.05296314e-01 -3.16053063e-01
7.07716763e-01 9.44832981e-01 6.45175099e-01 1.17111182e+00
2.19828203e-01 4.19325173e-01 1.89221168e+00 -8.33891332e-01
-1.10470831e-01 3.08320403e-01 -4.34919298e-02 6.72726214e-01
9.70431268e-02 4.04464811e-01 -4.25030977e-01 -1.53969064e-01
5.54120719e-01 -9.65595543e-01 -5.95601685e-02 4.49829459e-01
-9.05842721e-01 7.94077694e-01 6.08091950e-01 1.62674874e-01
-7.30782151e-02 5.33316672e-01 7.06291378e-01 1.64137453e-01
5.53201258e-01 7.37905264e-01 -1.20439641e-01 8.34563375e-02
-8.75235617e-01 -1.63363084e-01 4.26071018e-01 -3.28082368e-02
1.69650391e-01 3.92011600e-03 -4.55913037e-01 1.19000173e+00
-1.26401767e-01 2.30230182e-01 3.54121029e-01 -8.98551524e-01
-3.56221110e-01 2.28039131e-01 -3.76732111e-01 -6.99965715e-01
-6.73290789e-01 -3.71221542e-01 -7.65472174e-01 4.83672351e-01
4.90390062e-01 -3.67728800e-01 -1.50528121e+00 1.25143385e+00
1.83000728e-01 6.32565692e-02 -1.46402389e-01 9.81683731e-01
9.75639760e-01 1.15213491e-01 7.51614332e-01 9.63597819e-02
1.44354808e+00 -5.33171535e-01 -6.22453570e-01 -1.70292065e-01
6.91101849e-01 -1.02449226e+00 8.53794456e-01 7.91577578e-01
-7.03768730e-01 -5.53972721e-01 -1.25333261e+00 -2.95083765e-02
-2.32515231e-01 1.54725611e-01 8.64722550e-01 1.43858695e+00
-1.33895206e+00 2.76426375e-01 -4.50096071e-01 -3.72267932e-01
7.72986889e-01 3.04841191e-01 -4.19781869e-03 3.47793773e-02
-1.17933655e+00 1.15188646e+00 2.96498179e-01 2.36316174e-01
-4.15201128e-01 -9.85886395e-01 -7.32281923e-01 -7.57406294e-01
-1.71524480e-01 -1.82415500e-01 1.33566415e+00 -1.62411642e+00
-1.33102965e+00 1.43284321e+00 2.94060349e-01 -5.63452244e-01
5.60496509e-01 3.11259329e-01 -8.55892777e-01 3.74652386e-01
-3.27441514e-01 1.33146822e+00 9.32497621e-01 -1.28907537e+00
-8.97256076e-01 2.88032919e-01 -1.97385862e-01 2.05710009e-01
-3.62439364e-01 2.36529931e-01 -1.75157622e-01 -6.52013898e-01
-5.62338114e-01 -1.04130852e+00 -2.85904586e-01 4.71456945e-01
-4.61058825e-01 2.16045883e-02 4.16959941e-01 -1.00012338e+00
1.19465184e+00 -2.19526148e+00 -7.13610053e-01 6.55731261e-01
1.37576386e-01 4.46106881e-01 -3.75692159e-01 -2.68242747e-01
-2.81397402e-01 4.19717550e-01 -6.90258741e-02 3.80976617e-01
-1.17380351e-01 5.03276149e-03 1.10215925e-01 4.94437307e-01
6.09483898e-01 7.35892236e-01 -7.91702986e-01 -6.65626347e-01
4.93158370e-01 8.03184092e-01 -2.24078670e-01 -4.37741786e-01
8.85337591e-02 8.01858306e-02 3.52542043e-01 8.48786473e-01
7.36021280e-01 8.91243592e-02 1.44060194e-01 -4.88268346e-01
-9.48188361e-03 8.32319334e-02 -6.85408652e-01 1.25248480e+00
-5.51027596e-01 1.21513152e+00 -1.56465456e-01 -2.31850818e-01
7.56261528e-01 3.12336057e-01 3.61044019e-01 -9.65413392e-01
-5.20987716e-03 3.63663912e-01 7.06526697e-01 -6.24211073e-01
3.28314692e-01 -3.67017746e-01 -4.16000560e-02 2.81527024e-02
-8.03629458e-02 -4.40320909e-01 1.05843700e-01 -2.33286634e-01
9.35883939e-01 -2.51383513e-01 5.98436177e-01 -4.42228556e-01
3.67120087e-01 2.47030959e-01 6.84936345e-02 7.60631502e-01
-7.95081317e-01 6.89180851e-01 5.90909421e-01 -5.91686845e-01
-8.40683401e-01 -8.34312201e-01 -8.19382668e-01 1.16056001e+00
-1.16498247e-01 1.74729735e-01 -9.98951137e-01 -8.72933626e-01
-1.16643794e-01 4.45815057e-01 -1.35527074e+00 -3.67390007e-01
-2.63595223e-01 -1.32891989e+00 1.22122896e+00 4.52290863e-01
3.69569182e-01 -1.02018404e+00 -7.79253781e-01 -1.90518528e-01
1.86766386e-01 -6.06472254e-01 -3.90957922e-01 6.55962348e-01
-6.19154982e-02 -1.20455122e+00 -9.82115984e-01 -7.83658206e-01
8.00891519e-01 -7.64123425e-02 1.03071272e+00 4.14077878e-01
-1.45731914e+00 2.66017020e-01 -2.12856054e-01 -8.16584229e-01
-7.72780001e-01 1.03407063e-01 -5.20350754e-01 -1.69515356e-01
6.98780715e-01 4.22627956e-01 -9.47271466e-01 1.14086956e-01
-1.07220626e+00 -1.98576748e-02 7.76537120e-01 9.87783551e-01
4.93144900e-01 2.21948013e-01 3.74757290e-01 -1.28142917e+00
6.67064011e-01 -3.96380126e-02 -4.66114432e-02 1.32213041e-01
-3.38438153e-01 -3.01782697e-01 3.01573649e-02 -5.35094202e-01
-1.13639164e+00 2.56415069e-01 -5.26481807e-01 3.88375312e-01
-6.25406802e-01 1.79451689e-01 5.06304741e-01 -9.25723732e-01
1.57819593e+00 -2.83203959e-01 -2.08843410e-01 1.31769493e-01
1.54020846e-01 7.55785584e-01 6.55511796e-01 -1.33265825e-02
4.72688705e-01 4.45426464e-01 1.03084676e-01 -7.97432542e-01
-1.14615309e+00 -4.65606391e-01 -3.49776655e-01 -5.22308946e-01
9.16007638e-01 -8.53421330e-01 -5.14845215e-02 6.16104484e-01
-6.07529938e-01 -8.53240490e-01 -2.60619789e-01 -1.03075646e-01
-2.82948744e-02 1.80586755e-01 -6.55132174e-01 -7.16849685e-01
-4.98058587e-01 -8.23718786e-01 1.17783284e+00 5.84352493e-01
-8.47222209e-01 -1.46394539e+00 3.21066678e-01 2.79722989e-01
6.06244743e-01 7.12435424e-01 8.24893415e-01 -1.37997344e-01
4.95652497e-01 -2.03375056e-01 -4.88811404e-01 4.00664300e-01
4.79146361e-01 4.78833258e-01 -1.62989640e+00 -2.28656545e-01
-6.57259166e-01 -5.83606124e-01 1.22207224e+00 6.86843038e-01
1.44509459e+00 1.59343094e-01 -2.76808459e-02 8.37505341e-01
1.61272919e+00 -4.85105440e-02 8.74158323e-01 4.81138408e-01
3.40965211e-01 8.31390500e-01 2.90646434e-01 -7.83990696e-03
-2.41925061e-01 2.61666536e-01 2.75678545e-01 -1.11127949e+00
-1.07116258e+00 2.02158064e-01 -4.71548438e-02 2.35089168e-01
6.96387291e-02 3.20993885e-02 -1.01141310e+00 6.46698713e-01
-8.58782113e-01 -6.11291885e-01 -2.69367605e-01 1.73800337e+00
1.43876374e+00 3.52344096e-01 2.68936813e-01 3.27968299e-01
7.70755827e-01 -8.30341205e-02 -5.34858286e-01 -8.56656849e-01
-2.38031194e-01 8.95192027e-01 9.61699665e-01 6.01794720e-01
-1.55006373e+00 8.31607699e-01 8.14814758e+00 1.00107062e+00
-1.60279918e+00 -2.06307694e-01 1.13094974e+00 -3.75291817e-02
-6.55960068e-02 -6.52711332e-01 -4.06663477e-01 2.42214769e-01
8.11223745e-01 4.35993612e-01 1.30628273e-01 3.44155252e-01
-1.00323543e-01 -4.76652086e-01 -6.42855048e-01 7.95663297e-01
1.66793540e-01 -1.24363661e+00 -1.99256718e-01 -1.36787266e-01
1.03363013e+00 -1.94446385e-01 6.53150737e-01 -1.87511340e-01
2.86421120e-01 -1.36152196e+00 1.82212889e-01 5.07893503e-01
1.58507657e+00 -8.26210141e-01 1.10173225e+00 -6.55257165e-01
-6.40390575e-01 2.59343773e-01 -2.02512845e-01 3.11317086e-01
-6.02621317e-01 4.84389067e-01 -1.34532785e+00 4.87836599e-02
5.92452228e-01 1.34240970e-01 -1.26090813e+00 1.31170881e+00
-2.81382143e-01 8.52141559e-01 -1.16295546e-01 -1.16174243e-01
3.17202359e-01 6.65156782e-01 -1.21026732e-01 1.80399001e+00
-1.68385953e-01 -3.62964809e-01 -1.43105313e-01 2.21502036e-01
3.00321102e-01 4.60228512e-05 -1.11655414e-01 4.06476483e-02
1.89753503e-01 1.67248917e+00 -1.04449916e+00 -2.03072488e-01
-1.25092611e-01 7.75930107e-01 -2.80419528e-01 1.43794894e-01
-6.50145292e-01 -6.30227923e-01 4.83966798e-01 -1.33406386e-01
-2.94278949e-01 6.54087365e-01 -9.08807218e-01 -2.51882881e-01
-3.99275064e-01 -1.16134202e+00 7.10865617e-01 -6.65860176e-01
-1.33730423e+00 6.06024623e-01 -4.25233185e-01 -1.02726138e+00
1.11564994e-01 -1.05689502e+00 -6.35011554e-01 9.79307711e-01
-1.96374190e+00 -1.28299844e+00 -7.25041568e-01 4.08316821e-01
2.98800647e-01 -2.89532430e-02 1.08581233e+00 -9.26049426e-02
-3.54417086e-01 1.00265110e+00 -4.73206580e-01 2.21441254e-01
1.27317297e+00 -1.79842436e+00 2.97124714e-01 6.18422627e-01
8.17563459e-02 4.88164350e-02 6.38201177e-01 -5.14472008e-01
-5.89887619e-01 -1.04687190e+00 3.92487377e-01 -2.82293230e-01
7.71380723e-01 -4.36089784e-02 -6.24603450e-01 -1.53221816e-01
5.78324258e-01 -1.63629428e-02 1.35308003e+00 7.40647539e-02
-5.29600322e-01 -7.40680322e-02 -1.32896972e+00 7.27550089e-01
4.46622103e-01 -4.90944356e-01 1.75267532e-02 3.46774399e-01
6.79423660e-02 -5.12564957e-01 -9.04522777e-01 5.33386350e-01
5.62784076e-01 -6.61758363e-01 7.45354116e-01 -3.37239414e-01
4.69616860e-01 -2.13546567e-02 2.14237154e-01 -1.43160462e+00
-3.81959856e-01 -5.95283329e-01 4.24862683e-01 7.66485214e-01
6.25724792e-01 -2.23079965e-01 1.01424873e+00 2.85288543e-01
1.15168095e-02 -6.25805616e-01 -8.53496253e-01 -4.33672071e-02
4.28671628e-01 -3.52782965e-01 -1.27423301e-01 1.01443398e+00
-1.18846707e-01 -2.28483438e-01 -3.48294675e-01 -1.98809728e-02
5.80366790e-01 -6.34930611e-01 2.50482500e-01 -8.13088119e-01
-6.77525774e-02 -8.75664055e-01 -6.32196665e-01 1.08702876e-01
-1.37306169e-01 -8.29184830e-01 3.33424151e-01 -1.12547493e+00
2.57276744e-01 -5.06301224e-01 -5.95801890e-01 9.06449080e-01
-5.75906873e-01 1.39477992e+00 -1.81956291e-01 -3.39294076e-01
-3.20070535e-01 -6.56765461e-01 1.24103856e+00 -4.56936151e-01
8.11416805e-02 -3.66466433e-01 -1.10668397e+00 5.80741644e-01
1.01571572e+00 -1.78625986e-01 -1.27969965e-01 -2.84126580e-01
1.65708885e-01 -7.09641755e-01 5.69999635e-01 -1.06550825e+00
4.76728147e-03 -2.02372357e-01 1.01851749e+00 -3.42015031e-04
-7.68837482e-02 -4.29181546e-01 -2.17542663e-01 9.05875862e-01
-1.03479445e+00 -6.27468467e-01 7.21352458e-01 4.39153239e-02
9.12461504e-02 -2.77818143e-01 1.46486652e+00 1.37112411e-02
-9.57805455e-01 -1.47676229e-01 -4.90708590e-01 -7.86287710e-02
8.53353739e-01 -1.93491727e-01 -5.77552021e-01 -6.79052155e-03
-6.65548205e-01 -1.80759177e-01 2.37392113e-01 3.99889559e-01
3.04835588e-01 -1.22375643e+00 -9.43055391e-01 1.48885772e-01
3.99468243e-01 -6.23203158e-01 4.81220663e-01 8.28030527e-01
-1.00308251e+00 2.60575533e-01 -5.05406976e-01 -7.81130731e-01
-1.56817770e+00 -2.29408611e-02 8.99916232e-01 -3.53388451e-02
-2.06722066e-01 1.34480572e+00 -1.92515180e-01 -2.27961928e-01
3.02184016e-01 -1.09081499e-01 -1.57103598e-01 -2.06813902e-01
7.20263481e-01 5.70616685e-02 3.66447508e-01 -3.45477134e-01
-1.80582568e-01 5.45002043e-01 -3.69400650e-01 -1.00294404e-01
7.20354438e-01 4.20707941e-01 1.81764364e-02 1.87400162e-01
1.07383311e+00 5.02877496e-02 -1.28502691e+00 1.31409019e-01
7.73789361e-02 -5.46703935e-01 5.95906794e-01 -1.77608037e+00
-1.04554796e+00 8.68549764e-01 1.44616783e+00 4.05857533e-01
1.51491439e+00 -2.68965095e-01 4.25372541e-01 -4.38515730e-02
-3.25529873e-01 -1.55482376e+00 -3.79881337e-02 -1.06251597e-01
5.98896861e-01 -1.33453310e+00 1.11289598e-01 -7.17703819e-01
-7.49351978e-01 1.24831331e+00 7.53188670e-01 -2.69312531e-01
2.00338900e-01 7.01944709e-01 1.03039944e+00 -4.80202548e-02
-7.98610032e-01 -4.25194144e-01 8.12682927e-01 1.25426471e+00
7.40182817e-01 1.57352641e-01 -2.61692315e-01 -2.77648628e-01
-9.27350074e-02 -1.95143238e-01 4.55252886e-01 6.61005199e-01
-2.15124875e-01 -9.36112165e-01 -4.11972046e-01 8.06046844e-01
-9.11970556e-01 -3.10868621e-01 -1.14241338e+00 9.59289670e-01
6.49225473e-01 6.49725914e-01 1.53812900e-01 -2.79530883e-01
6.04761317e-02 -7.03346208e-02 6.10694230e-01 -5.53353727e-01
-1.04938269e+00 4.59045798e-01 6.46174073e-01 -3.12408090e-01
-6.23424470e-01 -4.04845804e-01 -6.75383210e-01 2.19778359e-01
-3.44444454e-01 -4.77776825e-01 6.23621225e-01 4.95832413e-01
-3.01944405e-01 9.05544221e-01 5.71624041e-01 -2.14128137e-01
-2.95699716e-01 -8.74090254e-01 -3.72193515e-01 4.69696820e-01
6.23610973e-01 -2.43689433e-01 -3.33246320e-01 3.51079673e-01] | [15.672615051269531, -2.946049213409424] |
1a196919-7dc8-4d14-968e-4ee0efcf856c | memory-aware-curriculum-federated-learning | 2107.02504 | null | https://arxiv.org/abs/2107.02504v2 | https://arxiv.org/pdf/2107.02504v2.pdf | Memory-aware curriculum federated learning for breast cancer classification | For early breast cancer detection, regular screening with mammography imaging is recommended. Routinary examinations result in datasets with a predominant amount of negative samples. A potential solution to such class-imbalance is joining forces across multiple institutions. Developing a collaborative computer-aided diagnosis system is challenging in different ways. Patient privacy and regulations need to be carefully respected. Data across institutions may be acquired from different devices or imaging protocols, leading to heterogeneous non-IID data. Also, for learning-based methods, new optimization strategies working on distributed data are required. Recently, federated learning has emerged as an effective tool for collaborative learning. In this setting, local models perform computation on their private data to update the global model. The order and the frequency of local updates influence the final global model. Hence, the order in which samples are locally presented to the optimizers plays an important role. In this work, we define a memory-aware curriculum learning method for the federated setting. Our curriculum controls the order of the training samples paying special attention to those that are forgotten after the deployment of the global model. Our approach is combined with unsupervised domain adaptation to deal with domain shift while preserving data privacy. We evaluate our method with three clinical datasets from different vendors. Our results verify the effectiveness of federated adversarial learning for the multi-site breast cancer classification. Moreover, we show that our proposed memory-aware curriculum method is beneficial to further improve classification performance. Our code is publicly available at: https://github.com/ameliajimenez/curriculum-federated-learning. | ['Gemma Piella', 'Diana Mateus', 'Miguel A. González Ballester', 'Mickael Tardy', 'Amelia Jiménez-Sánchez'] | 2021-07-06 | null | null | null | null | ['breast-cancer-detection', 'breast-cancer-detection'] | ['knowledge-base', 'medical'] | [ 5.47008030e-02 -2.21168855e-03 -5.51550567e-01 -4.37142134e-01
-9.43328023e-01 -6.25648975e-01 2.71687005e-02 4.24078763e-01
-4.95210886e-01 7.31781483e-01 -4.55574803e-02 -5.94362557e-01
-3.00939322e-01 -1.01433253e+00 -8.72294545e-01 -1.01327264e+00
3.56793739e-02 4.86738354e-01 1.18323527e-02 5.28556518e-02
-1.16330281e-01 5.41648448e-01 -8.69403541e-01 5.69804668e-01
9.39093828e-01 1.00120378e+00 -1.45000890e-02 6.42825842e-01
-9.16977897e-02 7.39974678e-01 -5.10508478e-01 -7.21058369e-01
5.96230567e-01 -2.56785929e-01 -6.58087194e-01 -8.31357837e-02
2.08166495e-01 -5.02163291e-01 -3.73797625e-01 1.19926381e+00
7.14051306e-01 -2.38855213e-01 2.68009245e-01 -1.24184978e+00
-1.06778711e-01 5.90320766e-01 -3.02863836e-01 4.63570841e-02
-2.58206278e-01 6.17594868e-02 5.10206163e-01 -4.79890853e-01
7.28306770e-01 5.62227309e-01 6.15709245e-01 5.35597026e-01
-1.10287356e+00 -1.03811026e+00 3.56045701e-02 2.66301990e-01
-1.16747844e+00 -3.41946900e-01 8.18037331e-01 -2.47428894e-01
1.06086567e-01 5.48542202e-01 5.91725886e-01 1.17154062e+00
4.81843084e-01 6.42729580e-01 1.08379138e+00 -2.94012874e-01
5.41506171e-01 4.92030829e-01 1.10657305e-01 5.54495633e-01
5.85321784e-01 1.22023612e-01 -4.48208481e-01 -7.69942760e-01
3.99990946e-01 3.80791843e-01 -1.74998954e-01 -7.85075128e-01
-7.77777016e-01 8.05820823e-01 3.30415279e-01 1.22979417e-01
-2.68265247e-01 -1.86584190e-01 6.07533336e-01 3.97629529e-01
3.06193799e-01 1.73246250e-01 -5.15838623e-01 3.94681662e-01
-7.30964720e-01 2.44072571e-01 8.21778119e-01 8.20308626e-01
6.26630306e-01 -5.32628715e-01 -7.21460581e-02 6.35484695e-01
3.00096273e-02 4.17434603e-01 5.84097981e-01 -8.45554233e-01
5.18814921e-01 5.55697680e-01 -1.34603996e-02 -1.28864944e+00
-2.57678568e-01 -5.38273871e-01 -1.19083834e+00 -8.84692296e-02
4.86904413e-01 -3.18319529e-01 -6.30321980e-01 1.60033643e+00
7.75488615e-01 2.56551862e-01 1.15739942e-01 7.64540136e-01
3.41116011e-01 1.95119798e-01 1.34232521e-01 -1.95510894e-01
1.30642188e+00 -4.84158874e-01 -8.18326592e-01 1.19258180e-01
8.33075643e-01 -5.86391211e-01 4.50963467e-01 4.75866318e-01
-7.22635508e-01 -1.07201770e-01 -1.01957297e+00 3.21224511e-01
-2.48323128e-01 -4.77926387e-03 6.65267587e-01 1.06097698e+00
-6.37281835e-01 3.96741748e-01 -1.25587773e+00 -7.84058571e-02
8.87554467e-01 6.33334517e-01 -4.72904831e-01 -4.17474777e-01
-1.08958936e+00 3.59540254e-01 2.51020223e-01 -1.94444910e-01
-7.74305224e-01 -1.01604569e+00 -5.80641150e-01 -1.22545041e-01
4.46176231e-01 -5.81264734e-01 1.18426204e+00 -1.07488430e+00
-1.17205608e+00 7.28694260e-01 1.51785105e-01 -7.07181334e-01
9.39791083e-01 1.97312117e-01 -5.06791115e-01 2.14086965e-01
-2.17144471e-02 -4.79912534e-02 7.55634487e-01 -1.15950227e+00
-7.53216684e-01 -6.22941136e-01 -2.03876093e-01 -1.22453414e-01
-8.51575851e-01 -1.92854792e-01 -4.78864670e-01 -7.19511688e-01
7.07841218e-02 -9.42663729e-01 -5.69048047e-01 3.09781998e-01
-2.36697137e-01 4.92719948e-01 7.91769743e-01 -7.11478174e-01
1.19919586e+00 -2.34142494e+00 -2.07791492e-01 5.59422255e-01
3.15728992e-01 1.78100944e-01 7.53841698e-02 7.43047372e-02
3.18219587e-02 1.43643945e-01 -2.50209540e-01 -2.20414162e-01
-3.13081354e-01 2.02584609e-01 -9.41178109e-03 6.91567302e-01
1.13259293e-02 5.09114802e-01 -8.24832499e-01 -7.54868805e-01
-1.19597830e-01 1.76437259e-01 -9.37839270e-01 1.81265369e-01
-3.38778421e-02 7.11808741e-01 -8.44899058e-01 7.50841618e-01
1.09322000e+00 -4.68526185e-01 8.47863793e-01 -2.39672691e-01
3.41591984e-01 -2.07294837e-01 -1.27811646e+00 1.72992647e+00
-2.90728778e-01 4.67049964e-02 4.55133677e-01 -1.05823088e+00
6.44589186e-01 3.85309577e-01 8.37057650e-01 -7.14663923e-01
2.23904952e-01 4.14998114e-01 -3.44973691e-02 -4.32911277e-01
4.07492697e-01 -7.65848383e-02 -1.43034995e-01 3.31626207e-01
-1.20117009e-01 3.40935856e-01 -3.07296067e-01 2.63514787e-01
1.40227532e+00 -4.07025635e-01 3.01659822e-01 -1.88223541e-01
4.20391589e-01 1.96101204e-01 1.04878306e+00 7.04406381e-01
-4.76721019e-01 3.79416674e-01 5.00232339e-01 -3.89494747e-01
-8.18464756e-01 -7.20096350e-01 -4.33761030e-01 6.86542988e-01
2.37555608e-01 -1.14617564e-01 -4.63311344e-01 -1.04073191e+00
2.24420235e-01 2.67334968e-01 -5.42819083e-01 -3.44995856e-01
-6.38753295e-01 -8.82244408e-01 5.82205534e-01 2.92385042e-01
3.30602705e-01 -5.49738467e-01 -5.69936812e-01 2.06059575e-01
-1.54494643e-01 -8.10563445e-01 -4.83714759e-01 3.81223768e-01
-1.06891263e+00 -1.32403493e+00 -6.51352882e-01 -5.70428014e-01
9.66594815e-01 -1.09953888e-01 8.07387352e-01 -4.28372398e-02
-3.02175850e-01 2.60548860e-01 -3.35421175e-01 -4.52768326e-01
-7.26604939e-01 1.62303120e-01 -1.55466184e-01 4.39142972e-01
1.51572496e-01 -2.63972074e-01 -8.56562734e-01 4.37933326e-01
-1.13988745e+00 -2.70311862e-01 3.64282817e-01 1.28604996e+00
8.14265966e-01 1.36551619e-01 4.82178837e-01 -1.56183064e+00
2.60731190e-01 -8.09075236e-01 -7.36274600e-01 3.04181695e-01
-4.83422369e-01 -3.52536254e-02 7.88707078e-01 -5.99111259e-01
-1.10963237e+00 2.77139783e-01 6.94689900e-02 -4.99670893e-01
1.10022418e-01 3.88272762e-01 -4.68974799e-01 -2.57046849e-01
6.76866412e-01 -5.84972575e-02 2.64979094e-01 -2.64726371e-01
-5.77748455e-02 7.77312756e-01 2.25078225e-01 -6.43325329e-01
5.87427676e-01 6.96035087e-01 -1.47656342e-02 -2.78742105e-01
-2.85525680e-01 -3.48956436e-01 -3.19945291e-02 9.21112821e-02
3.58753413e-01 -9.64317381e-01 -5.47878563e-01 5.66347897e-01
-8.27130616e-01 -2.72768438e-01 -2.22894758e-01 4.98990655e-01
-2.62285978e-01 2.76314169e-01 -5.39475977e-01 -4.68440145e-01
-4.54917282e-01 -1.17424965e+00 5.99549353e-01 1.14014693e-01
-1.03573008e-02 -9.36461806e-01 8.93264487e-02 4.89675879e-01
4.98085439e-01 3.59463692e-01 8.80896449e-01 -9.48683441e-01
-7.30484247e-01 -4.66021717e-01 2.21211433e-01 2.55226910e-01
2.58919269e-01 -3.46559644e-01 -7.95959294e-01 -6.54476106e-01
2.15334147e-01 -1.90662608e-01 4.15216774e-01 3.39857996e-01
1.62238121e+00 -4.35380965e-01 -6.60841167e-01 9.65885222e-01
1.52793014e+00 2.21133292e-01 4.05952930e-01 4.08903658e-01
4.46397096e-01 4.53598261e-01 8.88885379e-01 7.88652599e-01
1.19903721e-01 4.12641257e-01 5.26152432e-01 -8.66231546e-02
3.32419962e-01 -1.27980351e-01 2.92779282e-02 4.45878595e-01
4.28652704e-01 -1.85391799e-01 -9.99327362e-01 5.33770800e-01
-1.78193045e+00 -7.79647827e-01 1.72617555e-01 2.45169950e+00
1.04642200e+00 -1.96159035e-01 -3.17930669e-01 -4.42435890e-02
6.80662334e-01 -2.08453923e-01 -8.11035752e-01 -1.14638582e-01
3.13260406e-02 3.77335221e-01 9.90550637e-01 2.76222043e-02
-1.06602335e+00 4.25668716e-01 4.96081781e+00 8.78263175e-01
-1.45540071e+00 5.24002910e-01 9.76155162e-01 -2.41536781e-01
-3.25258672e-01 -2.33234480e-01 -5.85977316e-01 6.01953387e-01
8.77880216e-01 -2.29127854e-01 -4.40986734e-03 9.62922752e-01
-6.81587681e-02 3.51379439e-02 -9.68342960e-01 9.55664098e-01
-3.10621738e-01 -1.54981458e+00 -2.82855690e-01 2.82840222e-01
8.87031674e-01 -1.14991494e-01 1.78586602e-01 8.19565877e-02
2.41963968e-01 -6.79274201e-01 3.16426426e-01 3.07011396e-01
8.66307914e-01 -9.49479580e-01 7.91521370e-01 3.64313483e-01
-7.36322641e-01 -3.03689718e-01 -1.99693382e-01 6.17587984e-01
-4.11261082e-01 7.33002901e-01 -8.80196214e-01 8.56718063e-01
7.99667239e-01 2.75077313e-01 -3.84593010e-01 9.22408760e-01
3.93153220e-01 6.11917973e-01 -2.37824604e-01 1.97263166e-01
-4.47189391e-01 1.32566512e-01 3.95787776e-01 8.30114305e-01
3.53167653e-01 9.62932929e-02 1.22866847e-01 3.24353844e-01
-2.51633734e-01 3.12573671e-01 -6.12042427e-01 9.38215926e-02
7.15400338e-01 1.22577918e+00 -4.28585351e-01 2.77355146e-02
-3.21849197e-01 8.87206435e-01 1.69464946e-01 8.91344398e-02
-8.99892867e-01 -1.19611375e-01 7.72743583e-01 2.25423113e-01
1.66625008e-01 1.81415960e-01 -3.25501710e-01 -1.07950735e+00
8.01066607e-02 -1.35125732e+00 8.10239732e-01 1.87800065e-01
-1.48714590e+00 4.60748762e-01 -2.58305758e-01 -1.39114189e+00
-1.11648336e-01 -4.06014860e-01 -2.08879933e-01 6.73850656e-01
-1.43977821e+00 -8.88742685e-01 -3.08941633e-01 9.34010565e-01
-9.34522748e-02 -3.40085059e-01 1.02084935e+00 7.28490710e-01
-5.69755077e-01 1.25476277e+00 4.13385183e-01 3.38109136e-01
1.00193059e+00 -8.31335247e-01 -3.47051978e-01 7.89572835e-01
-3.43254656e-01 5.74348032e-01 2.78790146e-01 -6.99722290e-01
-1.67007923e+00 -1.40945435e+00 3.58990490e-01 -2.12430581e-01
3.85385990e-01 -3.09040070e-01 -9.00275469e-01 6.00448787e-01
-1.38602600e-01 5.75970769e-01 1.06825662e+00 -9.67051759e-02
-1.25353754e-01 -6.62344158e-01 -1.68331134e+00 3.81320119e-01
6.05596006e-01 -3.23663563e-01 2.44680673e-01 4.27393377e-01
5.56658745e-01 -7.22540498e-01 -1.05199873e+00 4.64735895e-01
3.82533371e-01 -8.04789186e-01 6.88762307e-01 -5.97981155e-01
3.06685090e-01 -6.31297156e-02 -1.91082671e-01 -9.90409791e-01
-7.71179795e-02 -5.75120866e-01 -1.46654308e-01 1.07148457e+00
2.50192165e-01 -9.48226452e-01 1.32498097e+00 8.99865866e-01
2.08897635e-01 -6.68383420e-01 -1.06419909e+00 -5.29592752e-01
2.14133576e-01 -2.99645931e-01 8.42994630e-01 1.16426492e+00
-2.33267322e-01 -5.48942327e-01 -2.78204203e-01 5.24913192e-01
7.81756639e-01 4.85126488e-02 8.56809318e-01 -8.07072699e-01
-6.82699144e-01 6.20807819e-02 -4.32907492e-01 -3.47978592e-01
-4.43916209e-02 -1.03971696e+00 -3.75205815e-01 -7.07821310e-01
3.04456085e-01 -1.03892350e+00 -5.46015441e-01 5.67462146e-01
-1.45207942e-01 -8.96657184e-02 5.65121025e-02 1.69424713e-01
-3.77282858e-01 1.70420840e-01 1.20030439e+00 -3.68781567e-01
-2.19839558e-01 3.20769966e-01 -7.30271637e-01 2.53508389e-01
1.07573497e+00 -9.13753510e-01 -2.82881588e-01 -3.39150131e-01
2.85410844e-02 3.29601586e-01 3.36116105e-01 -8.43079448e-01
5.88192701e-01 -2.69906133e-01 3.79413575e-01 -3.71751517e-01
-4.73458357e-02 -1.36831903e+00 6.46065652e-01 8.34063232e-01
-2.00943321e-01 -3.14503580e-01 2.47303680e-01 7.12565720e-01
-2.67102629e-01 -1.05349384e-01 8.17388237e-01 -4.80244607e-02
-2.57203043e-01 6.19976282e-01 -7.08398372e-02 -8.76243562e-02
1.28018165e+00 3.17805111e-02 -1.77736029e-01 -6.79334998e-03
-6.63378358e-01 4.04052258e-01 5.06668031e-01 6.43948093e-02
3.79043996e-01 -1.09872413e+00 -6.40187144e-01 4.39929992e-01
1.54718503e-01 3.15496683e-01 5.62499583e-01 8.85080755e-01
-4.45290506e-01 1.12085447e-01 -4.84149903e-02 -4.82369661e-01
-1.40925586e+00 4.92627442e-01 4.98499960e-01 -6.88827515e-01
-2.54956484e-01 6.18948877e-01 -6.08470626e-02 -7.27114916e-01
3.44865322e-01 -9.10779908e-02 3.45193177e-01 -5.38185351e-02
4.36312646e-01 2.60121197e-01 5.06996453e-01 -9.25031397e-03
-3.74468535e-01 -2.01377831e-02 -5.23403287e-01 1.89835042e-01
1.29107547e+00 1.83088914e-01 -7.93987960e-02 -1.18063815e-01
1.30572879e+00 2.57603317e-01 -9.48070168e-01 -3.13791901e-01
-2.31451809e-01 -6.33195281e-01 8.69839862e-02 -7.32334614e-01
-1.48013115e+00 3.52048784e-01 1.03862798e+00 -2.23517016e-01
1.52200413e+00 -3.57532799e-01 6.72350705e-01 1.89077199e-01
7.62305617e-01 -9.21764970e-01 -1.25206143e-01 -4.85794544e-02
3.77507567e-01 -1.44310200e+00 9.10094380e-02 -4.27027643e-01
-5.13604879e-01 8.40198040e-01 4.86402810e-01 1.96009457e-01
9.02562141e-01 6.00924194e-01 1.95848852e-01 1.00151040e-01
-5.33391178e-01 5.56184292e-01 -2.07112163e-01 5.18572152e-01
3.82942590e-03 2.91778207e-01 -3.81648213e-01 1.14332414e+00
1.92408860e-01 3.72176975e-01 2.66737223e-01 1.31832957e+00
2.49576017e-01 -1.71106791e+00 -7.39626408e-01 5.18267691e-01
-7.68283606e-01 2.22161576e-01 2.12863064e-03 5.10478020e-01
2.47840509e-01 7.23953843e-01 -6.44229949e-02 -2.08794042e-01
2.23659441e-01 1.86011510e-03 2.64923602e-01 -4.66843516e-01
-9.07261431e-01 -9.04076323e-02 -2.23204359e-01 -5.81712067e-01
-1.14657834e-01 -7.62627184e-01 -1.28026581e+00 -3.94257843e-01
-2.59352088e-01 1.35283604e-01 7.88455665e-01 3.48945886e-01
6.61336660e-01 5.21679640e-01 1.01885450e+00 -1.93810880e-01
-1.09873295e+00 -2.64159769e-01 -5.96137822e-01 3.98607343e-01
3.49046946e-01 -1.12815760e-01 -1.54635057e-01 1.97307076e-02] | [6.088781356811523, 6.451897621154785] |
7ef6e1c7-3a85-47bf-81aa-5a73479b34c5 | self-concordant-analysis-of-frank-wolfe-1 | null | null | https://proceedings.icml.cc/static/paper_files/icml/2020/2292-Paper.pdf | https://proceedings.icml.cc/static/paper_files/icml/2020/2292-Paper.pdf | Self-concordant analysis of Frank-Wolfe algorithm | Projection-free optimization via different variants of the Frank-Wolfe (FW) method has become one of the cornerstones in optimization for machine learning since in many cases the linear minimization oracle is much cheaper to implement than projections and some sparsity needs to be preserved. In a number of applications, e.g. Poisson inverse problems or quantum state tomography, the loss is given by a
self-concordant (SC) function having unbounded curvature, implying absence of theoretical guaranteesfor the existing FW methods. We use the
theory of SC functions to provide a new adaptive step size for FW methods and prove global convergence rate O(1/k), k being the iteration counter. If the problem can be represented by a local linear minimization oracle, we are the first
to propose a FW method with linear convergence rate without assuming neither strong convexity nor a Lipschitz continuous gradient. | ['Kamil Safin', 'Mathias Staudigl', 'Petr Ostroukhov', 'Shimrit Shtern', 'Pavel Dvurechenskii'] | null | null | https://proceedings.icml.cc/static/paper_files/icml/2020/2292-Paper.pdf | https://proceedings.icml.cc/static/paper_files/icml/2020/2292-Paper.pdf | icml-2020-1 | ['quantum-state-tomography'] | ['medical'] | [ 1.89796403e-01 4.02295947e-01 -1.47562504e-01 -2.92214960e-01
-8.71632516e-01 -6.34620130e-01 3.45278740e-01 -4.86428104e-02
-6.90265596e-01 1.01737344e+00 -5.97652346e-02 -4.92020428e-01
-2.47300357e-01 -6.62040353e-01 -9.31407452e-01 -1.05606186e+00
6.27309084e-02 5.70731401e-01 2.43904721e-02 -1.97628900e-01
3.32947284e-01 4.12144154e-01 -6.75123870e-01 -3.06456119e-01
7.44324148e-01 9.77835834e-01 -4.90110889e-02 6.09928727e-01
2.23531798e-02 4.26491916e-01 1.00726321e-01 -4.45194542e-01
7.53777266e-01 -6.76604152e-01 -8.83155465e-01 3.37236896e-02
3.55882496e-01 -6.67490344e-03 -2.52399355e-01 1.47305107e+00
3.06474656e-01 1.69649392e-01 4.93019730e-01 -8.50115716e-01
-2.06739426e-01 4.53287065e-01 -4.99896824e-01 -7.74850398e-02
4.01916765e-02 -2.49412119e-01 1.02023828e+00 -1.03287005e+00
6.87612116e-01 9.07083631e-01 8.47845256e-01 5.08647680e-01
-1.62012017e+00 -3.35029691e-01 -1.72445953e-01 1.93898827e-01
-1.16217136e+00 -5.36512375e-01 6.11121237e-01 -1.51036337e-01
4.18822199e-01 5.96307933e-01 3.94485295e-01 4.62797552e-01
1.18188508e-01 5.22572339e-01 1.48222458e+00 -5.80557287e-01
3.06185067e-01 5.93061030e-01 -1.20971557e-02 9.60393786e-01
3.16415548e-01 -1.33535534e-01 -3.46179038e-01 -3.34836900e-01
7.55497575e-01 -5.79501502e-02 -6.26328409e-01 -6.93286717e-01
-1.15407777e+00 1.13736904e+00 3.09471756e-01 1.09644823e-01
-9.60489437e-02 2.82459140e-01 1.59988225e-01 6.35373831e-01
3.13018411e-01 3.34041208e-01 -1.39761254e-01 -4.24716510e-02
-9.80363190e-01 1.83140010e-01 1.11812460e+00 6.59365594e-01
8.34082067e-01 -9.93759036e-02 2.53815979e-01 3.45662296e-01
2.02908128e-01 6.21007323e-01 8.54083896e-02 -1.26615012e+00
5.83347380e-01 -1.96946412e-02 4.16253090e-01 -8.00947249e-01
-2.30067730e-01 -3.56410265e-01 -9.09903228e-01 2.76625395e-01
7.98134208e-01 -1.09457880e-01 -3.73999178e-01 1.88487124e+00
4.67804044e-01 -3.90968956e-02 -1.41312286e-01 9.37500894e-01
-1.86459109e-01 6.26668692e-01 -8.25623333e-01 -5.74972391e-01
7.62000442e-01 -6.44852877e-01 -5.38049340e-01 -6.98124319e-02
7.84125507e-01 -6.64617777e-01 1.04864049e+00 5.17424464e-01
-1.27095878e+00 1.77158058e-01 -1.06176877e+00 -1.51758105e-01
1.42218381e-01 -1.44646257e-01 6.33135378e-01 7.81952441e-01
-9.46423948e-01 7.67276466e-01 -1.00711906e+00 -1.26118258e-01
3.16185087e-01 4.35143292e-01 -6.78314984e-01 -9.84132141e-02
-5.09088993e-01 7.87411571e-01 2.84214109e-01 2.11753607e-01
-7.66769350e-01 -6.38033688e-01 -6.06181860e-01 -7.78814703e-02
4.37225699e-01 -4.64268267e-01 7.31570423e-01 -5.87970436e-01
-1.70364630e+00 7.84425259e-01 -2.86605120e-01 -6.12231731e-01
1.11625719e+00 -1.41847253e-01 3.26544583e-01 1.01582624e-01
-3.21345925e-02 -1.80624440e-01 7.99374402e-01 -8.29657733e-01
-1.49092615e-01 -5.54864466e-01 1.15139931e-01 2.19111696e-01
-2.29873225e-01 -1.91529393e-01 3.97417061e-02 -1.27647042e-01
6.00825071e-01 -1.05932605e+00 -4.98163462e-01 3.16221178e-01
-3.73610914e-01 1.61869287e-01 5.51631033e-01 -4.68539447e-01
7.97891259e-01 -1.99392319e+00 4.04938608e-01 3.43942553e-01
2.07592189e-01 -1.33095294e-01 1.32856727e-01 4.92367148e-01
9.23012272e-02 -1.44169688e-01 -8.07932615e-01 -4.74611908e-01
1.70119599e-01 9.34995860e-02 -3.59623909e-01 1.34148133e+00
-2.69738615e-01 6.88615918e-01 -8.18249762e-01 -1.58470586e-01
4.75158319e-02 2.00492874e-01 -8.74066234e-01 -1.09693527e-01
9.08608213e-02 5.39135754e-01 -3.63961101e-01 1.34324282e-01
8.22168350e-01 -3.00271124e-01 1.31989881e-01 7.64664635e-02
-2.88204819e-01 2.61281163e-01 -1.53804243e+00 1.59878194e+00
-6.26125753e-01 5.46494782e-01 8.67364466e-01 -1.33214641e+00
4.83454376e-01 2.36954570e-01 4.51277077e-01 -2.03923196e-01
-1.42290637e-01 6.19635046e-01 -3.08074087e-01 -1.02851287e-01
7.44532794e-02 -8.00540090e-01 1.19385295e-01 4.57221985e-01
-1.19558424e-01 -1.92594945e-01 -5.71640730e-02 1.81739375e-01
1.18772554e+00 -2.94268161e-01 2.45861232e-01 -6.78161860e-01
5.97700298e-01 -9.75475013e-02 5.95047235e-01 8.36082518e-01
3.62650007e-02 7.50687361e-01 5.81834614e-01 -2.87308395e-01
-1.07324302e+00 -1.01637745e+00 -4.85783339e-01 6.45425260e-01
1.63707267e-02 -1.95306927e-01 -5.94006002e-01 -4.26362753e-01
-2.01516673e-01 3.78142953e-01 -4.87334728e-01 2.60756761e-02
-6.08986318e-01 -1.00277269e+00 3.78911257e-01 8.19500163e-02
7.39899397e-01 -4.88358855e-01 -3.45215976e-01 1.85642943e-01
8.59590545e-02 -1.02799773e+00 -6.84161723e-01 3.72365743e-01
-1.06670892e+00 -1.04103768e+00 -7.12894857e-01 -5.18832386e-01
8.36388409e-01 1.21176906e-01 6.42872274e-01 -2.83734262e-01
7.17067719e-02 2.65886337e-01 2.01839685e-01 -9.03428495e-02
-1.13783762e-01 -1.07520588e-01 1.66261539e-01 1.68986201e-01
-1.47510037e-01 -6.85820460e-01 -5.89147389e-01 1.32790297e-01
-8.72840106e-01 -7.39934742e-02 4.82285172e-01 1.17063332e+00
7.92962551e-01 -4.73884977e-02 -1.12566343e-02 -1.27854168e+00
4.81317103e-01 -2.62107491e-01 -1.17440438e+00 2.94225961e-02
-7.45743692e-01 2.95041323e-01 9.43021476e-01 -3.45261395e-01
-8.28921616e-01 3.86299521e-01 -1.35766849e-01 -1.66251466e-01
3.08932960e-01 5.34841061e-01 -1.68701366e-01 -6.99692965e-01
5.83450019e-01 3.95905942e-01 4.37536277e-02 -5.29381394e-01
5.10550916e-01 1.93330079e-01 5.63209534e-01 -6.06552362e-01
1.03710449e+00 1.05034971e+00 5.64890921e-01 -1.07234359e+00
-9.52358723e-01 -5.83703995e-01 -4.28211182e-01 1.89198986e-01
6.22819662e-01 -4.04878378e-01 -8.09645474e-01 6.14667982e-02
-9.83510673e-01 -2.19904840e-01 -5.24574280e-01 7.63585210e-01
-8.21808040e-01 6.37431085e-01 -5.88282228e-01 -1.10708499e+00
-1.35934770e-01 -8.17894518e-01 5.28206110e-01 -9.80038717e-02
2.77369171e-01 -1.12400138e+00 1.96428895e-01 2.58688658e-01
5.63516140e-01 2.20546499e-01 6.89928353e-01 -3.06097567e-01
-6.84967279e-01 -4.52633202e-01 -8.97664055e-02 5.35087168e-01
-2.56448627e-01 -6.45601273e-01 -6.57992542e-01 -3.89431208e-01
9.66873705e-01 -2.54631430e-01 9.00570810e-01 3.61843646e-01
1.10661221e+00 -8.03285122e-01 2.98260548e-03 1.18837583e+00
1.65454352e+00 -5.47080398e-01 5.14951289e-01 6.19039983e-02
6.11162007e-01 3.10268104e-01 1.40601471e-01 4.28768784e-01
3.79834548e-02 4.04484630e-01 3.96840423e-01 1.50735155e-01
2.87943006e-01 -1.26811996e-01 5.03532827e-01 1.10657024e+00
-1.04491204e-01 4.43190396e-01 -4.50316668e-01 4.40916359e-01
-1.83748090e+00 -1.03799689e+00 -4.18516308e-01 2.93911982e+00
9.75288630e-01 1.20788477e-01 -4.71735597e-02 1.79090068e-01
3.18075508e-01 6.41326234e-02 -4.43090230e-01 -3.92376959e-01
-1.98060513e-01 4.78031278e-01 1.18244076e+00 1.04144096e+00
-8.30731452e-01 6.46987557e-01 6.60595560e+00 7.16224730e-01
-1.10907137e+00 5.79126179e-01 3.67841959e-01 -1.07165553e-01
-3.52467328e-01 3.67197722e-01 -6.62351251e-01 4.32634771e-01
6.50026321e-01 -6.32367954e-02 7.20528841e-01 8.74785125e-01
1.74965426e-01 -9.37510431e-02 -9.39746737e-01 1.01499796e+00
-2.26600423e-01 -1.34680736e+00 -6.50477171e-01 5.69690287e-01
1.02083898e+00 3.27775240e-01 5.62463738e-02 -1.43942460e-01
2.42512114e-02 -1.02606070e+00 3.68386269e-01 2.67533481e-01
6.38627112e-01 -6.75130665e-01 5.86768389e-01 5.71455359e-01
-7.88540006e-01 7.26212561e-02 -7.74546683e-01 -7.80310109e-02
4.10453260e-01 9.81440067e-01 -5.44114292e-01 2.19054610e-01
1.91353410e-01 2.20879763e-01 2.40398929e-01 1.12689102e+00
-1.56831294e-01 8.38053882e-01 -1.03664613e+00 6.26581684e-02
3.54919583e-01 -1.01771951e+00 9.81468141e-01 1.07499874e+00
3.44996393e-01 1.65847242e-01 3.50013450e-02 8.73541474e-01
-2.48923406e-01 2.85042614e-01 -4.79402661e-01 -1.18486263e-01
7.93954954e-02 1.24504650e+00 -6.69397533e-01 3.27082649e-02
-4.96574938e-01 1.05498338e+00 1.92155987e-01 2.97378421e-01
-5.26031375e-01 -4.27964658e-01 4.69807297e-01 3.34288090e-01
4.30459917e-01 -6.54693902e-01 -5.10115147e-01 -1.42362392e+00
4.50503916e-01 -5.28319716e-01 2.37029344e-01 1.22882523e-01
-1.08915031e+00 1.17461346e-01 -1.35330975e-01 -5.87860465e-01
-1.46231975e-03 -6.90207064e-01 -7.44384050e-01 8.84231031e-01
-1.47387946e+00 -6.51470482e-01 1.85478479e-01 6.38785422e-01
2.80693918e-02 1.64021045e-01 6.71991527e-01 2.24802211e-01
-2.55321264e-01 3.88103694e-01 7.14959800e-01 -2.01285750e-01
3.75681460e-01 -1.58388507e+00 -1.53730720e-01 1.06126416e+00
2.51571059e-01 6.17674828e-01 9.92650151e-01 -3.30849141e-01
-1.83922040e+00 -6.76373422e-01 8.80614340e-01 -3.54391992e-01
9.09867644e-01 -4.93579328e-01 -7.49127030e-01 6.36849940e-01
-2.42511898e-01 4.13674802e-01 3.10722470e-01 4.77834344e-02
-1.08232364e-01 -1.96308449e-01 -1.21193779e+00 2.45684028e-01
7.98789740e-01 -6.99107289e-01 -1.66324839e-01 7.99268186e-01
2.38231793e-01 -3.73865455e-01 -7.27516532e-01 1.58425987e-01
4.11275566e-01 -9.97807741e-01 8.67441475e-01 -4.05688852e-01
1.94157716e-02 -2.78957129e-01 -3.31165522e-01 -9.26148713e-01
6.08105771e-03 -1.07290006e+00 -3.47017229e-01 5.34087718e-01
4.42942053e-01 -1.08240724e+00 9.82198894e-01 5.59069932e-01
-6.01891391e-02 -8.59710693e-01 -1.24285972e+00 -1.03555715e+00
2.52327085e-01 -3.93550754e-01 -1.01932064e-01 7.61653304e-01
5.41845299e-02 1.90015152e-01 -5.98067164e-01 1.36875346e-01
9.69372690e-01 1.40979201e-01 6.41868234e-01 -9.25382435e-01
-8.98862004e-01 -3.82448882e-01 -3.76232624e-01 -1.41446030e+00
-1.19593874e-01 -1.13642204e+00 1.11184023e-01 -1.10796738e+00
3.13185990e-01 -5.84769130e-01 -2.11628646e-01 2.07407866e-02
1.98514879e-01 2.30600864e-01 8.11166465e-02 2.07328364e-01
-4.33858722e-01 6.39445662e-01 1.12073171e+00 8.52598771e-02
-1.54191494e-01 4.53689754e-01 -4.93378431e-01 7.90908098e-01
4.98120010e-01 -7.79594004e-01 -2.43078649e-01 -2.76322067e-01
5.40480494e-01 6.56917915e-02 1.84675172e-01 -8.07265699e-01
4.18172628e-01 -2.18173280e-01 -7.89225399e-02 -1.30879611e-01
5.66646218e-01 -5.59824467e-01 1.63009703e-01 5.80170214e-01
-2.11709559e-01 -1.30605802e-01 -4.36288238e-01 6.63973749e-01
5.05403206e-02 -7.74744868e-01 1.15710568e+00 -2.20689714e-01
1.78115740e-01 3.84057939e-01 5.92011623e-02 2.82344759e-01
6.31895065e-01 1.53102130e-01 1.17467605e-01 -6.41138077e-01
-6.13980711e-01 -1.05384551e-01 6.47050977e-01 -6.23351097e-01
5.69129050e-01 -1.26537931e+00 -8.14830184e-01 3.02775372e-02
-4.24396217e-01 5.16847298e-02 -1.11432046e-01 1.61742914e+00
-7.44774580e-01 5.29791176e-01 2.73605376e-01 -4.06284541e-01
-9.73542511e-01 3.12484980e-01 3.70053440e-01 -3.21426600e-01
-8.31322610e-01 1.16385198e+00 2.28645056e-01 -3.64180297e-01
9.82636288e-02 -1.00578189e-01 6.77338719e-01 -4.19723481e-01
6.60209239e-01 4.83023137e-01 1.17511451e-01 -4.53258604e-01
-1.98128253e-01 3.96653324e-01 -6.61262963e-03 -5.85549772e-01
1.51597011e+00 -6.21749610e-02 -4.13805455e-01 3.87958258e-01
1.52178776e+00 3.66387635e-01 -1.17476380e+00 -2.90664703e-01
8.29819441e-02 -5.49631953e-01 1.65524825e-01 -1.23015314e-01
-9.78724003e-01 9.51106966e-01 4.31100160e-01 1.87615693e-01
8.77645910e-01 -2.17998549e-02 8.51027548e-01 7.95403600e-01
5.27556777e-01 -1.19692183e+00 -2.64464825e-01 5.04381895e-01
6.53980255e-01 -1.25994921e+00 1.76185429e-01 -2.91067153e-01
-2.32716784e-01 1.14185655e+00 -1.44940943e-01 -5.76222479e-01
7.27087498e-01 1.65828720e-01 -3.94732505e-01 5.48019856e-02
-4.07460958e-01 -7.03023598e-02 3.34131926e-01 3.61520350e-02
1.81494683e-01 2.06006300e-02 -8.60451460e-01 5.49775101e-02
-2.03625098e-01 -3.00263315e-01 6.00605786e-01 7.60119915e-01
-5.13384998e-01 -1.34140944e+00 -1.71172500e-01 5.76961398e-01
-6.63608730e-01 -1.57750770e-01 -1.32039174e-01 5.20845711e-01
-3.47170413e-01 5.80025911e-01 -5.25101125e-01 1.61639854e-01
-1.61006987e-01 8.73483568e-02 9.80744362e-01 -5.55439591e-01
-6.38138205e-02 3.95568274e-02 -3.06306720e-01 -7.11074233e-01
-2.73950607e-01 -7.06169665e-01 -1.16206610e+00 -5.57987332e-01
-5.70633352e-01 3.35189044e-01 9.71209824e-01 1.16350794e+00
-1.25399679e-01 -3.73508245e-01 7.81838417e-01 -4.92577463e-01
-1.21867454e+00 -6.23398364e-01 -7.46307731e-01 2.08200693e-01
4.61805582e-01 -2.28053391e-01 -7.06957340e-01 -3.28704029e-01] | [6.592894554138184, 4.54079008102417] |
32aeb8ac-7445-47c6-8282-57e3a4de5b9f | guided-deep-decoder-unsupervised-image-pair | 2007.11766 | null | https://arxiv.org/abs/2007.11766v1 | https://arxiv.org/pdf/2007.11766v1.pdf | Guided Deep Decoder: Unsupervised Image Pair Fusion | The fusion of input and guidance images that have a tradeoff in their information (e.g., hyperspectral and RGB image fusion or pansharpening) can be interpreted as one general problem. However, previous studies applied a task-specific handcrafted prior and did not address the problems with a unified approach. To address this limitation, in this study, we propose a guided deep decoder network as a general prior. The proposed network is composed of an encoder-decoder network that exploits multi-scale features of a guidance image and a deep decoder network that generates an output image. The two networks are connected by feature refinement units to embed the multi-scale features of the guidance image into the deep decoder network. The proposed network allows the network parameters to be optimized in an unsupervised way without training data. Our results show that the proposed network can achieve state-of-the-art performance in various image fusion problems. | ['wei he', 'Naoto Yokoya', 'Danfeng Hong', 'Tatsumi Uezato'] | 2020-07-23 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/4749_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123510086.pdf | eccv-2020-8 | ['pansharpening'] | ['computer-vision'] | [ 7.17157781e-01 1.19945090e-02 1.61851183e-01 -5.16142309e-01
-7.37111390e-01 -1.95627376e-01 5.50234497e-01 -3.06511104e-01
-4.19839263e-01 4.02074218e-01 2.99267769e-02 -9.76867899e-02
-6.35371953e-02 -9.23858166e-01 -6.45297229e-01 -1.07605135e+00
6.07500613e-01 -1.58457950e-01 1.70654729e-01 -3.29775900e-01
1.25654012e-01 3.62412810e-01 -1.58004963e+00 1.69284999e-01
1.25072038e+00 1.31247592e+00 7.04518139e-01 5.52944660e-01
1.21526584e-01 4.62859869e-01 -3.22673649e-01 -1.58910230e-01
4.23377752e-01 -5.32658935e-01 -3.12280178e-01 3.72686148e-01
1.49979740e-01 -4.10673976e-01 -5.02353847e-01 1.50068879e+00
6.32709920e-01 1.18197821e-01 6.11325026e-01 -1.19198263e+00
-8.40577543e-01 2.46898323e-01 -7.64404953e-01 -2.14769468e-01
5.35448119e-02 2.36025110e-01 7.48384058e-01 -6.42291427e-01
8.10308978e-02 9.79321122e-01 3.67055863e-01 1.98293507e-01
-1.08382654e+00 -6.49120152e-01 1.28897473e-01 5.15004843e-02
-1.41029894e+00 -3.46029371e-01 9.97446120e-01 -3.39040160e-01
6.20020211e-01 -9.13102627e-02 5.29470325e-01 6.93181932e-01
3.95004034e-01 5.51899612e-01 1.20036173e+00 -3.63596261e-01
-2.89067701e-02 4.55247574e-02 -8.39823782e-02 8.10467601e-01
3.16336364e-01 1.55564517e-01 -1.77393436e-01 2.59211719e-01
7.56129622e-01 1.98766559e-01 -4.21591878e-01 -8.85194987e-02
-1.00346756e+00 8.30205441e-01 8.49720299e-01 2.36088708e-01
-5.25080621e-01 -2.92805210e-02 -9.57958251e-02 1.31164119e-01
2.16127560e-01 3.36422056e-01 -3.88607383e-01 5.00414193e-01
-1.03051555e+00 -1.20157287e-01 5.30183792e-01 7.52364933e-01
1.18380857e+00 1.62250400e-01 -3.04026417e-02 8.22574914e-01
7.00748563e-01 6.16115570e-01 5.90922296e-01 -8.18980694e-01
3.97223324e-01 6.92173123e-01 7.53848115e-04 -1.07152784e+00
-5.46789110e-01 -4.16997939e-01 -1.06747139e+00 4.07811821e-01
-8.21910650e-02 -5.59973001e-01 -1.25958943e+00 1.75608575e+00
1.76358908e-01 2.13770717e-01 4.06319648e-01 1.13439906e+00
1.03302908e+00 8.26933801e-01 -1.27348974e-01 -1.69099987e-01
1.28195357e+00 -1.31957102e+00 -8.28079045e-01 -3.78732532e-01
2.39907444e-01 -8.32740188e-01 7.88778663e-01 3.82669806e-01
-9.99226332e-01 -7.50804961e-01 -1.35880017e+00 -2.41206244e-01
-5.18904388e-01 4.49247777e-01 4.70776886e-01 4.32689160e-01
-1.09176767e+00 4.07611817e-01 -8.10470819e-01 -4.54066545e-01
1.45978332e-01 5.33921659e-01 -3.28193218e-01 1.17909685e-01
-1.17039537e+00 7.79944122e-01 7.20510483e-01 2.66280174e-01
-8.42139840e-01 -2.61293173e-01 -1.01014352e+00 1.03567541e-01
4.50592160e-01 -8.62119257e-01 1.11692500e+00 -1.12420285e+00
-1.83364093e+00 6.30572081e-01 4.34900150e-02 -2.13892579e-01
4.83836494e-02 -1.86002940e-01 -5.91693163e-01 1.77190006e-01
-5.94389774e-02 8.19640279e-01 9.92342055e-01 -1.32951081e+00
-8.87132704e-01 -2.47820109e-01 1.47680789e-01 6.39643192e-01
-3.28667969e-01 -2.86776274e-01 -6.06957197e-01 -7.51162291e-01
3.63531858e-01 -7.27883935e-01 -3.47460061e-01 -9.51421186e-02
-4.56705570e-01 3.02585870e-01 9.76774454e-01 -6.44215465e-01
8.86315107e-01 -2.20521569e+00 1.61877126e-01 3.23064849e-02
4.44459692e-02 3.86917233e-01 -3.29183191e-01 2.65743285e-01
-6.41764775e-02 -2.99007948e-02 -5.62817454e-01 -2.05633864e-01
1.78255420e-02 1.91777259e-01 -1.35823831e-01 4.89685506e-01
1.97994485e-01 7.72726238e-01 -8.30548465e-01 -2.69065976e-01
4.51075286e-01 7.40143955e-01 -3.94252896e-01 6.46989703e-01
-1.73502952e-01 5.11162519e-01 -4.24315572e-01 6.07052445e-01
9.35028136e-01 -2.62595952e-01 -3.65212709e-02 -6.79931700e-01
-1.91947564e-01 -1.98192075e-02 -1.19493759e+00 1.95532298e+00
-2.91539103e-01 3.93130720e-01 3.55558097e-01 -1.08948600e+00
9.95711625e-01 3.01486313e-01 4.74908322e-01 -5.65492034e-01
4.30604666e-01 1.95241585e-01 9.22960863e-02 -5.35516620e-01
5.71403086e-01 -2.85397917e-01 1.59634009e-01 3.33207786e-01
2.37462074e-01 -5.13040721e-01 -1.64366495e-02 -2.31023088e-01
5.79529405e-01 1.32288232e-01 2.47164205e-01 -6.82289898e-03
7.80537903e-01 -2.68563807e-01 6.53055608e-01 5.11483908e-01
1.66473210e-01 8.61646891e-01 1.76537991e-01 -8.87671262e-02
-8.96286845e-01 -7.50070035e-01 -9.30760503e-02 8.00621450e-01
4.78427380e-01 -2.14172676e-01 -7.00928450e-01 -4.26308751e-01
-1.86124131e-01 4.19394791e-01 -4.28112358e-01 -3.55955958e-01
5.44247665e-02 -1.00623918e+00 2.19958127e-01 3.41640353e-01
1.09554219e+00 -7.98558414e-01 -5.80689609e-01 1.37399197e-01
-1.05311304e-01 -1.23783100e+00 -3.30487251e-01 4.19559896e-01
-7.03653634e-01 -1.12498486e+00 -6.17355108e-01 -7.73190379e-01
7.88953066e-01 5.03395259e-01 6.26140773e-01 -1.04046583e-01
5.05582765e-02 1.44769162e-01 -4.14134622e-01 -4.17606711e-01
-1.65714234e-01 1.61960602e-01 -2.62326151e-01 4.12103623e-01
2.13721618e-01 -6.12589121e-01 -7.47007966e-01 3.02894533e-01
-1.36558151e+00 4.41306829e-01 7.84825087e-01 8.59440446e-01
6.82572782e-01 3.85642111e-01 3.77236575e-01 -6.01987839e-01
6.74351513e-01 -2.91451782e-01 -8.62938166e-01 3.70188057e-01
-5.81170142e-01 2.29580238e-01 4.94304419e-01 -6.77416392e-04
-1.29252970e+00 3.67759615e-01 -2.70431697e-01 -2.70604491e-01
-2.11259350e-01 6.65211916e-01 -5.18327355e-01 -3.97076964e-01
3.13239783e-01 2.68282503e-01 -5.27540334e-02 -4.38536704e-01
4.71117496e-01 8.44069421e-01 7.66398847e-01 -4.09049630e-01
7.96046853e-01 3.41914117e-01 4.30437401e-02 -5.81090927e-01
-9.65842545e-01 -2.52010554e-01 -7.13437796e-01 -6.82665557e-02
1.13705063e+00 -1.16740370e+00 -4.08752620e-01 7.02602565e-01
-1.08402634e+00 -3.16477776e-01 -8.61258060e-02 5.82805634e-01
-4.64769304e-01 4.20481980e-01 -4.30513024e-01 -5.76061189e-01
-3.82651985e-01 -1.51632178e+00 1.08279777e+00 7.55020618e-01
5.43542266e-01 -9.01154160e-01 -1.44661620e-01 1.64690554e-01
4.51705426e-01 4.26804245e-01 4.84879762e-01 -2.55382240e-01
-6.11138940e-01 1.80662330e-02 -5.55169463e-01 5.71780086e-01
5.64976633e-01 8.39276053e-03 -1.16100681e+00 -3.35462749e-01
3.62040192e-01 -3.43109757e-01 1.21814239e+00 3.57026547e-01
1.35405910e+00 -1.70988500e-01 -1.21756002e-01 1.23562717e+00
1.63702857e+00 3.51564318e-01 7.05221355e-01 2.81399727e-01
7.33522177e-01 2.70348996e-01 4.41562206e-01 4.12639081e-01
3.97406787e-01 3.89727771e-01 8.28574657e-01 -5.77272296e-01
2.29381789e-02 -5.57931811e-02 2.72482187e-01 7.30039060e-01
-1.23319566e-01 -3.51768464e-01 -6.40329659e-01 3.12156200e-01
-1.95318937e+00 -5.82964063e-01 7.81432688e-02 1.89206266e+00
7.04798520e-01 -2.20611468e-01 -3.56017381e-01 -2.19543707e-02
7.34174013e-01 4.22949672e-01 -7.27494776e-01 -1.24676056e-01
-2.79193312e-01 5.86647838e-02 6.47727072e-01 4.40739304e-01
-1.30998957e+00 8.79840970e-01 6.29812574e+00 5.35407841e-01
-1.42270160e+00 -9.68310330e-03 4.77162421e-01 2.80767649e-01
-3.20684940e-01 -1.98470727e-02 -5.55657029e-01 3.03976476e-01
6.40027761e-01 1.20680019e-01 5.33199549e-01 3.93599242e-01
6.67295381e-02 -2.24346533e-01 -7.66847908e-01 1.22891080e+00
9.83841196e-02 -9.17042077e-01 -1.53261563e-02 3.62054370e-02
9.50032890e-01 8.86846259e-02 3.04391146e-01 -1.18977785e-01
3.20036054e-01 -9.10975277e-01 4.74667609e-01 7.72559047e-01
7.16535807e-01 -5.94828725e-01 8.64122272e-01 2.51486510e-01
-1.16900074e+00 -9.01341885e-02 -4.20024127e-01 2.06130389e-02
1.83458060e-01 7.01302290e-01 -3.42570245e-01 9.62673724e-01
4.24008399e-01 9.55589235e-01 -5.82498014e-01 9.77046132e-01
-5.61090291e-01 3.01295280e-01 -2.67862111e-01 5.92832088e-01
4.28036928e-01 -5.97048879e-01 2.59011656e-01 1.10499275e+00
7.29016542e-01 2.79322505e-01 2.87859857e-01 8.05690467e-01
-6.26667663e-02 -1.23010628e-01 -5.71097076e-01 -1.04257442e-01
1.33817140e-02 1.63513041e+00 -5.91896653e-01 -2.98481643e-01
-6.87535465e-01 9.76470590e-01 3.00399661e-02 5.82989931e-01
-8.77030969e-01 -5.09651005e-01 5.43750644e-01 -1.41472280e-01
3.95227045e-01 -1.04217961e-01 -3.05409640e-01 -1.43269956e+00
-3.84070933e-01 -8.27277780e-01 2.83939719e-01 -1.22584867e+00
-1.15402210e+00 8.15173388e-01 -1.57994851e-01 -1.17386532e+00
-1.39301680e-02 -7.86388099e-01 -5.73089123e-01 1.08896124e+00
-1.92984235e+00 -1.15182149e+00 -8.26545000e-01 7.75036395e-01
8.54234472e-02 -3.25301617e-01 6.26683474e-01 2.44251341e-01
-6.99146986e-01 2.09606126e-01 1.33992538e-01 -2.29953174e-02
7.38996625e-01 -1.20870268e+00 -1.13654211e-01 1.11304319e+00
-2.34065965e-01 4.37413782e-01 4.46897477e-01 -4.68182951e-01
-1.26884079e+00 -1.25281894e+00 3.64772707e-01 3.39153618e-01
4.47147012e-01 -7.85265397e-03 -8.20637882e-01 5.87182879e-01
7.81357944e-01 9.19367895e-02 6.78373218e-01 -3.75418216e-01
-2.00360920e-02 -4.58976537e-01 -1.13318396e+00 3.97068322e-01
7.64337182e-01 -6.41303837e-01 -4.97521937e-01 2.20629409e-01
7.53145278e-01 -6.23057485e-01 -7.56902397e-01 7.31707394e-01
3.56611967e-01 -9.26874161e-01 7.20906079e-01 -3.97091843e-02
4.72715318e-01 -8.63360584e-01 -4.03841913e-01 -1.44978905e+00
-3.80047709e-01 -4.67160493e-01 2.40387376e-02 1.12362194e+00
2.71657526e-01 -7.05320954e-01 4.27712560e-01 4.74204689e-01
-4.09637541e-01 -5.95985591e-01 -3.18871111e-01 -2.46569186e-01
-3.95735860e-01 -1.20847166e-01 9.42549825e-01 7.82542884e-01
-4.70440149e-01 3.96974981e-01 -3.94897550e-01 6.02833152e-01
4.95564133e-01 2.49010235e-01 5.96234143e-01 -1.07992280e+00
-1.43956825e-01 -4.18988615e-01 -2.09400401e-01 -1.15331531e+00
1.02823503e-01 -8.72462094e-01 1.97140709e-01 -1.87573504e+00
2.11578548e-01 -2.08111748e-01 -5.84670067e-01 6.74396992e-01
-3.03015560e-01 3.04226607e-01 2.27982670e-01 -6.90967366e-02
-4.33634520e-01 8.23269010e-01 1.43632865e+00 -3.44464868e-01
-2.50593752e-01 -2.52578914e-01 -1.11304617e+00 7.45186090e-01
9.21672344e-01 -2.77057141e-01 -5.27806342e-01 -7.60811329e-01
1.17946707e-01 9.56922024e-03 3.07861686e-01 -1.11244059e+00
4.77221906e-01 -1.72160119e-01 4.18062150e-01 -6.67166412e-01
3.76201302e-01 -1.10515797e+00 1.47069067e-01 2.77644426e-01
-1.10806108e-01 -9.28865448e-02 1.11151196e-01 4.32547778e-01
-5.28517127e-01 -2.89363980e-01 8.31220627e-01 -5.91693744e-02
-6.48525119e-01 4.41328675e-01 -1.10256732e-01 -5.41374147e-01
8.94573331e-01 -2.15453193e-01 -2.60599852e-01 -4.16469246e-01
-5.72654843e-01 3.15439820e-01 5.23266435e-01 2.10352600e-01
7.04495490e-01 -1.31066537e+00 -5.35098910e-01 4.44747746e-01
-7.03226179e-02 3.05939287e-01 1.51049241e-01 7.26232529e-01
-4.03174460e-01 2.38576978e-01 -4.32703823e-01 -6.19283378e-01
-8.84943664e-01 4.75495160e-01 6.25994503e-01 -2.19583288e-01
-3.34308267e-01 5.14054835e-01 5.53558111e-01 -5.66889882e-01
7.76407942e-02 -4.20953810e-01 -2.92646497e-01 -4.02070694e-02
3.59787196e-01 -7.44496807e-02 2.90805753e-02 -7.31852233e-01
-5.36131971e-02 7.67488897e-01 1.36122555e-01 -8.19798782e-02
1.43948662e+00 -1.92700505e-01 -2.64455080e-01 7.41472766e-02
1.14354086e+00 -3.97508800e-01 -1.70845437e+00 -3.37226301e-01
-4.96474594e-01 -3.08539122e-01 4.69024152e-01 -7.55512178e-01
-1.67302585e+00 9.72647965e-01 7.15059519e-01 1.68487590e-04
1.82319427e+00 -4.03623521e-01 6.61334515e-01 3.94394130e-01
3.37154679e-02 -1.10008037e+00 7.20212162e-02 5.91957808e-01
7.66533077e-01 -1.43293893e+00 4.14532274e-02 -2.70044923e-01
-5.07568717e-01 1.27539885e+00 7.22825110e-01 1.19797643e-02
7.27742016e-01 1.82743669e-01 7.07110092e-02 -2.05099866e-01
-4.20256495e-01 -6.31506085e-01 5.87405145e-01 5.44933021e-01
3.69305938e-01 -4.89094630e-02 -3.34403872e-01 7.63588846e-01
-5.79965636e-02 1.55679554e-01 3.99341226e-01 8.02134693e-01
-6.32611275e-01 -9.96645868e-01 -3.83413345e-01 3.65433425e-01
-3.44789803e-01 -9.93953422e-02 -7.32111335e-02 4.35895860e-01
4.86170650e-01 1.10726011e+00 -7.12035373e-02 -6.61345422e-01
2.87580669e-01 -1.54694304e-01 3.93262893e-01 -5.12373328e-01
-4.40563500e-01 3.73062044e-01 -3.29418153e-01 -3.52815211e-01
-7.94791162e-01 -2.09394276e-01 -1.12168753e+00 8.84353503e-05
-4.92705971e-01 2.84294803e-02 7.19190538e-01 9.90754545e-01
2.95507878e-01 8.39571297e-01 6.89996481e-01 -9.38130558e-01
-3.34666930e-02 -8.61653984e-01 -8.00305247e-01 2.71936297e-01
4.52239543e-01 -6.00897491e-01 -7.00457245e-02 3.10962647e-01] | [10.450265884399414, -1.8429944515228271] |
edda8296-8cd1-45a4-a2f1-82ed06fc4b79 | airborne-lidar-point-cloud-classification | 2004.09057 | null | https://arxiv.org/abs/2004.09057v1 | https://arxiv.org/pdf/2004.09057v1.pdf | Airborne LiDAR Point Cloud Classification with Graph Attention Convolution Neural Network | Airborne light detection and ranging (LiDAR) plays an increasingly significant role in urban planning, topographic mapping, environmental monitoring, power line detection and other fields thanks to its capability to quickly acquire large-scale and high-precision ground information. To achieve point cloud classification, previous studies proposed point cloud deep learning models that can directly process raw point clouds based on PointNet-like architectures. And some recent works proposed graph convolution neural network based on the inherent topology of point clouds. However, the above point cloud deep learning models only pay attention to exploring local geometric structures, yet ignore global contextual relationships among all points. In this paper, we present a graph attention convolution neural network (GACNN) that can be directly applied to the classification of unstructured 3D point clouds obtained by airborne LiDAR. Specifically, we first introduce a graph attention convolution module that incorporates global contextual information and local structural features. Based on the proposed graph attention convolution module, we further design an end-to-end encoder-decoder network, named GACNN, to capture multiscale features of the point clouds and therefore enable more accurate airborne point cloud classification. Experiments on the ISPRS 3D labeling dataset show that the proposed model achieves a new state-of-the-art performance in terms of average F1 score (71.5\%) and a satisfying overall accuracy (83.2\%). Additionally, experiments further conducted on the 2019 Data Fusion Contest Dataset by comparing with other prevalent point cloud deep learning models demonstrate the favorable generalization capability of the proposed model. | ['Congcong Wen', 'Xiaojing Yao', 'Xiang Li', 'Tianhe Chi', 'Ling Peng'] | 2020-04-20 | null | null | null | null | ['line-detection'] | ['computer-vision'] | [-1.22124888e-01 -4.02524322e-01 2.69117132e-02 -4.90410060e-01
-5.15718818e-01 -2.29378074e-01 4.66436088e-01 2.67478168e-01
-7.38597959e-02 3.86036754e-01 -2.88898051e-01 -6.39820755e-01
-2.72234976e-01 -1.26822102e+00 -7.58049309e-01 -4.89295900e-01
-1.61048889e-01 4.33856070e-01 -3.23387049e-02 -1.13866121e-01
3.55574116e-02 1.05859935e+00 -1.43956733e+00 -6.81523886e-03
1.09182060e+00 1.25860083e+00 4.35905486e-01 3.30734909e-01
-5.45918643e-01 2.99892426e-01 -3.90137792e-01 -2.79882044e-01
3.04893136e-01 3.88258249e-01 -2.83054829e-01 3.26423682e-02
7.44842291e-01 -2.34221146e-01 -4.34428096e-01 1.29296386e+00
4.69967157e-01 1.14359565e-01 4.07892555e-01 -1.43785155e+00
-1.22059619e+00 2.38293767e-01 -7.43243158e-01 1.98808089e-01
-1.54621884e-01 2.32995778e-01 1.06867683e+00 -1.12599277e+00
1.43898964e-01 1.25742614e+00 7.38603711e-01 1.02932028e-01
-8.35876882e-01 -1.15660036e+00 4.24342185e-01 2.76384860e-01
-1.73859310e+00 1.48158699e-01 1.00044429e+00 -3.77414107e-01
1.12755620e+00 3.90894664e-03 7.71687984e-01 4.46229905e-01
2.16931000e-01 4.77074534e-01 8.12614977e-01 -5.80447242e-02
-1.15978736e-02 -2.78510183e-01 2.15100527e-01 8.69637907e-01
4.92148489e-01 1.88947201e-01 -1.19834393e-01 1.15160450e-01
7.48047948e-01 4.95895028e-01 -1.59618914e-01 -1.89730301e-01
-9.18457448e-01 9.21434820e-01 1.53484750e+00 2.29024678e-01
-5.16469955e-01 5.38056433e-01 6.70923293e-02 7.12626651e-02
6.13458455e-01 -4.13901769e-02 -1.86262771e-01 6.01172209e-01
-6.99581027e-01 1.47906289e-01 6.48076013e-02 1.30660844e+00
1.03581476e+00 3.55564654e-01 3.31528373e-02 5.78607082e-01
7.56450355e-01 8.71457696e-01 -2.11493328e-01 -4.27014977e-01
7.02950954e-01 9.76705909e-01 -3.28278631e-01 -1.12782538e+00
-4.51123238e-01 -7.84438550e-01 -1.14167464e+00 3.52835536e-01
-4.02037799e-01 1.19961165e-01 -1.18574095e+00 1.38702309e+00
2.84675330e-01 6.52172685e-01 2.71269921e-02 1.02273273e+00
1.35714412e+00 7.05175221e-01 2.54842579e-01 2.63330251e-01
1.25259471e+00 -5.31327128e-01 -5.28165638e-01 -1.99702740e-01
3.46607178e-01 -5.47733426e-01 8.78972054e-01 -1.79427713e-02
-6.16861582e-01 -8.90198171e-01 -1.21124983e+00 -2.37700820e-01
-6.61240458e-01 1.04181454e-01 8.43521535e-01 4.48227227e-01
-1.04571629e+00 3.48708659e-01 -7.34728396e-01 -3.72301519e-01
1.00094771e+00 6.38364255e-01 -1.24409132e-01 -3.40492487e-01
-8.77358139e-01 5.15922487e-01 2.76330948e-01 5.29292107e-01
-6.75372005e-01 -7.34981239e-01 -8.66270006e-01 3.37070793e-01
1.89973623e-01 -9.89319026e-01 8.39913845e-01 -2.92370647e-01
-9.65673923e-01 5.84743559e-01 5.10027707e-02 -4.57147449e-01
-7.85642769e-03 -1.52960271e-01 -5.45553207e-01 -8.97173956e-02
1.53365359e-01 8.95579219e-01 5.67513108e-01 -1.36029673e+00
-7.28605986e-01 -6.20605111e-01 5.30255698e-02 1.72573447e-01
-5.88569827e-02 -3.32035452e-01 -3.24870735e-01 -4.91721451e-01
4.01207328e-01 -8.81212831e-01 -2.69545734e-01 1.38686344e-01
-4.78691310e-01 -5.03845811e-01 1.23965871e+00 -1.79127708e-01
7.28701890e-01 -2.14234114e+00 -2.95136601e-01 2.61219352e-01
5.72672546e-01 3.18776608e-01 -1.75761446e-01 1.86222360e-01
-1.76838353e-01 2.46811241e-01 -4.13168907e-01 -3.31383407e-01
5.94930425e-02 1.87540561e-01 -4.69281524e-01 4.17806774e-01
5.20256817e-01 1.31146550e+00 -7.34806120e-01 -3.02471310e-01
7.44615197e-01 6.74941421e-01 -4.27335829e-01 -6.25581481e-03
-2.60416597e-01 2.24257097e-01 -5.83604693e-01 9.65638876e-01
1.20447648e+00 -3.89130563e-01 -3.36776227e-01 -3.03206921e-01
-9.50315073e-02 -2.77013257e-02 -8.19048584e-01 1.84925795e+00
-5.13712585e-01 6.65699124e-01 -5.60800508e-02 -6.76068783e-01
1.32655346e+00 -1.41790938e-02 5.51994622e-01 -5.88515162e-01
2.22287074e-01 1.26114458e-01 -4.51419353e-02 -4.06404510e-02
4.89951670e-01 -2.55047858e-01 1.46629359e-03 -9.54766050e-02
-4.57698014e-03 -5.01403391e-01 -4.81837600e-01 1.08294331e-01
8.40295494e-01 -2.73341209e-01 -6.83649704e-02 -9.15898085e-02
6.00834608e-01 9.07276198e-02 2.99788505e-01 4.40778553e-01
-6.81014210e-02 6.13914967e-01 -2.38394797e-01 -6.33841038e-01
-7.59197414e-01 -1.13899171e+00 -1.59256727e-01 3.87977958e-01
3.20515126e-01 -2.41391420e-01 -2.18049258e-01 -6.19630098e-01
3.65273833e-01 7.41787910e-01 -2.54421115e-01 -2.20381320e-02
-4.71909910e-01 -4.55693871e-01 2.71558255e-01 7.11136818e-01
8.92410934e-01 -8.25711429e-01 -9.56472307e-02 1.23291261e-01
2.04115182e-01 -1.49570847e+00 -2.61833407e-02 -8.37506875e-02
-1.00732303e+00 -9.66748297e-01 -1.71603993e-01 -7.52627552e-01
4.28022772e-01 7.92981505e-01 1.09156895e+00 3.14083993e-01
-5.19718975e-02 1.12981163e-01 -1.41099855e-01 -8.24230790e-01
2.25897580e-01 2.05278546e-01 -1.18781626e-01 9.78609640e-03
8.27558398e-01 -6.95657194e-01 -3.84537548e-01 -2.10547019e-02
-6.77917778e-01 -9.81402546e-02 6.85579062e-01 5.94164848e-01
8.32714498e-01 2.74184495e-01 4.63591337e-01 -6.07517660e-01
5.74210525e-01 -5.49223065e-01 -9.55581427e-01 -3.91738042e-02
-5.14622450e-01 -3.34013730e-01 4.10949618e-01 3.11493963e-01
-6.22910857e-01 6.06711432e-02 -2.81549722e-01 -1.27376914e+00
-2.77710736e-01 6.45390511e-01 -3.16331953e-01 -5.00129104e-01
2.14080960e-01 5.86185642e-02 -3.03998917e-01 -3.66302848e-01
4.52486396e-01 5.70908010e-01 5.53902805e-01 -4.00571257e-01
1.25991833e+00 6.09529257e-01 4.50227201e-01 -8.92764390e-01
-5.40682793e-01 -4.14552748e-01 -7.26677656e-01 -1.48998871e-01
1.11784554e+00 -1.30879903e+00 -1.01713347e+00 1.81229100e-01
-1.37412918e+00 7.76084140e-02 -1.76530480e-01 5.60566425e-01
-2.49011457e-01 2.84270763e-01 -1.95370108e-01 -7.45307386e-01
-6.04817688e-01 -1.13902915e+00 1.51634133e+00 1.71104223e-01
4.30305213e-01 -9.45257604e-01 -3.58643621e-01 1.35643467e-01
3.11936110e-01 4.51910526e-01 1.13167214e+00 -4.11398262e-01
-1.41677964e+00 -1.65451512e-01 -8.54346216e-01 3.09460849e-01
9.46597531e-02 -6.63945228e-02 -9.84523714e-01 -3.11596334e-01
-2.13835731e-01 1.90356329e-01 9.74055707e-01 4.28647518e-01
1.45505273e+00 1.31532684e-01 -4.35134053e-01 9.07978833e-01
1.83105862e+00 9.89330634e-02 5.16660988e-01 3.02182175e-02
1.33499920e+00 2.01107427e-01 4.37405139e-01 1.41207054e-01
4.69307125e-01 4.45598125e-01 1.10614455e+00 -2.48434126e-01
-1.75839052e-01 -3.62860173e-01 -1.34403124e-01 8.33654106e-01
-1.07637435e-01 -3.34793240e-01 -1.10484910e+00 5.41650712e-01
-1.70798528e+00 -7.22391784e-01 -4.91578728e-01 1.83103752e+00
-2.76158959e-01 2.39717379e-01 -3.58961016e-01 4.14573252e-02
7.88133740e-01 2.63827592e-01 -5.12081504e-01 -2.43551657e-01
-5.80763444e-02 5.72968841e-01 7.86268055e-01 3.65956664e-01
-1.17555821e+00 1.09680414e+00 5.15380430e+00 7.72970438e-01
-1.33620203e+00 1.11934431e-01 1.23164080e-01 6.36181235e-02
-3.92039686e-01 -2.46137843e-01 -7.47818351e-01 3.66411090e-01
6.66909695e-01 -6.54343292e-02 1.84410453e-01 8.39714468e-01
1.12973481e-01 6.16880178e-01 -8.31359088e-01 1.23990703e+00
-2.70311892e-01 -1.57634997e+00 4.68588680e-01 4.00451601e-01
6.35639429e-01 5.89859962e-01 2.09365800e-01 3.55316252e-01
4.31956530e-01 -1.17925453e+00 5.58106184e-01 4.01320577e-01
9.03651178e-01 -8.64039958e-01 7.95557737e-01 2.50044674e-01
-1.70928657e+00 -1.65550157e-01 -7.01302230e-01 -6.43689409e-02
1.87875003e-01 7.70220220e-01 -1.05843985e+00 1.05684447e+00
7.17096150e-01 1.11669195e+00 -4.74604934e-01 1.18141413e+00
-3.53845209e-01 3.72429252e-01 -4.49613005e-01 4.39418480e-02
6.67713940e-01 -4.15860474e-01 5.99214256e-01 9.54932749e-01
6.70652449e-01 3.64053667e-01 3.85604411e-01 1.25484562e+00
-4.23196733e-01 -1.02041692e-01 -1.03331995e+00 6.88878298e-02
6.32250845e-01 1.24138975e+00 -4.86954719e-01 -1.70036465e-01
-6.69945836e-01 3.00381869e-01 2.29782850e-01 3.64350498e-01
-8.87628555e-01 -3.62627149e-01 1.06330276e+00 4.64923382e-02
5.17149091e-01 -6.60209000e-01 -5.82173347e-01 -8.92497420e-01
-9.73024219e-02 -2.84326106e-01 2.21455395e-01 -1.16617966e+00
-1.49598491e+00 7.67186999e-01 -6.55540302e-02 -1.46517026e+00
2.82980800e-01 -7.87112951e-01 -8.20984483e-01 1.16877723e+00
-2.06149745e+00 -1.71285844e+00 -9.54573274e-01 7.01615095e-01
4.28550452e-01 -1.82085425e-01 3.70595098e-01 4.93009210e-01
-3.57241094e-01 2.52030730e-01 -2.30750173e-01 2.04375997e-01
1.20758601e-01 -1.10214710e+00 9.97048974e-01 7.99291611e-01
3.11360419e-01 4.88758713e-01 4.93071489e-02 -7.50579655e-01
-1.51710117e+00 -1.89423692e+00 7.00952709e-01 -2.89709598e-01
4.21832621e-01 -4.00339961e-01 -1.10634601e+00 7.29199469e-01
6.61103353e-02 2.85999894e-01 4.26890254e-01 6.23940974e-02
-2.14112833e-01 -3.99564624e-01 -1.17425299e+00 2.66332746e-01
1.36238623e+00 -6.38601422e-01 -3.45671088e-01 5.14918149e-01
1.21955407e+00 -4.52580512e-01 -7.30694354e-01 8.61005843e-01
4.69446667e-02 -7.09028006e-01 1.15827346e+00 -5.15865266e-01
2.16390148e-01 -6.54383421e-01 -5.64728200e-01 -1.07530928e+00
-8.12078595e-01 7.35810772e-02 9.56572145e-02 1.00660086e+00
1.85503542e-01 -7.70216763e-01 9.54285860e-01 1.12978570e-01
-7.16877580e-01 -7.28159606e-01 -1.05756915e+00 -7.08836734e-01
3.52472402e-02 -8.35509360e-01 1.34832013e+00 1.02183473e+00
-9.35354412e-01 5.23410022e-01 1.23237327e-01 9.39779818e-01
8.11151206e-01 4.77966368e-01 7.97231495e-01 -1.65200186e+00
3.22499275e-01 -5.47134697e-01 -8.76695633e-01 -1.09655809e+00
3.91346365e-01 -1.31966698e+00 -2.60625362e-01 -1.89176905e+00
-2.59568781e-01 -8.99692774e-01 -4.71574962e-01 3.82152617e-01
-8.34587589e-02 3.47264469e-01 4.59432304e-01 1.55352145e-01
-2.87374169e-01 8.09389710e-01 1.18947935e+00 -6.31124973e-01
-5.36775663e-02 -1.33235119e-02 -5.73924959e-01 4.43213165e-01
6.06467366e-01 -3.47381949e-01 -3.15514088e-01 -1.05253887e+00
2.13597044e-02 -2.63893962e-01 7.69088864e-01 -1.36913455e+00
3.90255988e-01 3.63680571e-02 4.65238303e-01 -1.29484975e+00
4.87334818e-01 -1.25461042e+00 2.26445124e-01 2.98254460e-01
3.17645311e-01 2.17747509e-01 4.64717984e-01 8.60996604e-01
-3.55140030e-01 2.72242129e-01 5.50833821e-01 -1.50819672e-02
-1.04007483e+00 1.02603805e+00 1.96270704e-01 -5.98448336e-01
1.02598560e+00 -2.55407155e-01 -3.32248926e-01 -1.77512005e-01
-4.57922280e-01 4.15839434e-01 2.97226489e-01 5.85857213e-01
9.59827304e-01 -1.67241108e+00 -7.56589830e-01 5.04640758e-01
2.70809263e-01 5.69316804e-01 2.35359073e-01 3.62823844e-01
-6.63184464e-01 5.71631253e-01 -1.05248719e-01 -1.19693851e+00
-1.07946932e+00 6.25536501e-01 2.78760880e-01 1.17164090e-01
-7.31089890e-01 8.94375205e-01 3.34543616e-01 -5.97303391e-01
-1.03132129e-01 -9.02421474e-01 -6.99733347e-02 -3.20130080e-01
-3.35415043e-02 9.11145136e-02 4.77018595e-01 -7.28997231e-01
-5.29396772e-01 9.58421648e-01 2.97837883e-01 4.08947557e-01
1.34895539e+00 9.64263305e-02 -3.81562486e-02 1.18420646e-01
1.17235780e+00 -2.99892575e-01 -8.69806767e-01 -4.51254010e-01
-2.58851171e-01 -7.85615623e-01 5.35382390e-01 -4.45602864e-01
-1.46944392e+00 1.45865083e+00 8.55169773e-01 1.69481531e-01
1.11271775e+00 -1.00213148e-01 8.81808579e-01 4.65191096e-01
4.70812738e-01 -3.43793750e-01 -2.57383406e-01 5.69550633e-01
8.13346863e-01 -1.37212121e+00 1.13884836e-01 -6.99183702e-01
-1.91365331e-01 9.90097702e-01 6.52520061e-01 -2.59198815e-01
8.78428042e-01 -2.88675725e-02 -1.56176910e-01 -6.94325626e-01
-5.96272111e-01 -5.78076422e-01 3.88649732e-01 6.88469470e-01
3.45389135e-02 3.19015712e-01 3.44553649e-01 4.05696392e-01
-2.68354416e-01 3.63947302e-02 -1.71766989e-02 7.79694200e-01
-6.25474632e-01 -6.94209158e-01 -2.82206953e-01 5.63713074e-01
1.10442780e-01 -2.89545298e-01 -2.57429898e-01 8.50373983e-01
3.68526161e-01 8.76255095e-01 5.21474719e-01 -6.66528761e-01
6.22564852e-01 -2.26999491e-01 2.61780292e-01 -6.90658748e-01
-4.61580753e-01 -1.56544104e-01 -3.22927475e-01 -3.48654360e-01
-3.25759947e-01 -3.44321489e-01 -1.31420839e+00 -5.89949608e-01
-4.90042567e-01 4.18749526e-02 9.98007178e-01 7.12642491e-01
6.74746692e-01 7.99079239e-01 5.86207390e-01 -9.05921817e-01
-2.91632526e-02 -7.87115693e-01 -5.56748569e-01 1.22589812e-01
4.72670853e-01 -7.76130557e-01 2.89447960e-02 -4.56278175e-01] | [7.974843978881836, -3.44391131401062] |
2924a0e1-6699-43c0-ada2-be072771d967 | turn-level-dialog-evaluation-with-dialog | 2011.06395 | null | https://arxiv.org/abs/2011.06395v1 | https://arxiv.org/pdf/2011.06395v1.pdf | Turn-level Dialog Evaluation with Dialog-level Weak Signals for Bot-Human Hybrid Customer Service Systems | We developed a machine learning approach that quantifies multiple aspects of the success or values in Customer Service contacts, at anytime during the interaction. Specifically, the value/reward function regarding to the turn-level behaviors across human agents, chatbots and other hybrid dialog systems is characterized by the incremental information and confidence gain between sentences, based on the token-level predictions from a multi-task neural network trained with only weak signals in dialog-level attributes/states. The resulting model, named Value Profiler, serves as a goal-oriented dialog manager that enhances conversations by regulating automated decisions with its reward and state predictions. It supports both real-time monitoring and scalable offline customer experience evaluation, for both bot- and human-handled contacts. We show how it improves Amazon customer service quality in several applications. | ['Ruofeng Wen'] | 2020-10-25 | null | null | null | null | ['goal-oriented-dialog'] | ['natural-language-processing'] | [-1.58723965e-01 4.01795268e-01 -3.45261246e-01 -9.68880594e-01
-9.94580388e-01 -7.09306419e-01 7.11989939e-01 1.48891807e-01
-2.91840404e-01 1.01376450e+00 3.00334126e-01 -2.23680690e-01
-6.04959950e-02 -6.80471897e-01 1.24029279e-01 -5.88692188e-01
1.51266247e-01 1.36436141e+00 1.74707353e-01 -7.13578701e-01
1.85139060e-01 3.35483193e-01 -9.24481332e-01 8.76092076e-01
3.47241580e-01 1.16034818e+00 -5.13296314e-02 1.13829184e+00
-5.69902778e-01 1.53547430e+00 -9.73816276e-01 -6.45839036e-01
-1.70398802e-01 -1.57118335e-01 -1.29599190e+00 5.03783710e-02
-3.65093559e-01 -5.80251694e-01 -7.58674517e-02 6.50738358e-01
3.78715098e-01 1.82997927e-01 4.30789500e-01 -1.76873851e+00
-4.48427439e-01 1.23124409e+00 1.22928165e-01 1.57203928e-01
6.38295829e-01 7.79119134e-01 1.46616185e+00 -2.18911946e-01
4.19173270e-01 1.53959572e+00 4.51141626e-01 7.17662573e-01
-1.35303700e+00 2.93412451e-02 8.61832052e-02 1.54892221e-01
-3.78431290e-01 -3.92257929e-01 4.64539796e-01 -3.97453636e-01
1.51945341e+00 1.61856487e-01 1.74425378e-01 1.49008811e+00
8.40489790e-02 9.01353717e-01 1.11787522e+00 -1.14285849e-01
1.71629101e-01 8.17708313e-01 3.30051988e-01 4.09819812e-01
-8.03651750e-01 -1.97284624e-01 -5.49504638e-01 -5.80396473e-01
2.76180953e-01 -1.02384336e-01 4.44830120e-01 2.35976100e-01
-9.55473542e-01 1.00593460e+00 -1.98283076e-01 4.81491178e-01
-6.63171947e-01 -1.21129595e-01 5.17819524e-01 9.23767626e-01
4.46642816e-01 6.90176368e-01 -1.03428257e+00 -1.08692992e+00
-1.69571154e-02 2.39231929e-01 1.75278819e+00 1.20789266e+00
6.76139176e-01 -5.11467122e-02 -8.56844544e-01 1.08066380e+00
-6.26255423e-02 3.13516915e-01 4.24798965e-01 -1.38598740e+00
3.88979465e-01 7.15889692e-01 6.38493896e-01 -2.25648209e-01
-6.17955208e-01 2.12820321e-01 -2.47862875e-01 6.95578456e-02
6.71254277e-01 -6.94609284e-01 2.87975240e-02 1.76611745e+00
3.28409553e-01 -5.67523122e-01 7.47912973e-02 2.55703270e-01
5.65024793e-01 4.88587648e-01 1.03145398e-01 -3.52918684e-01
1.45658565e+00 -1.09351063e+00 -1.08640862e+00 -1.03117891e-01
7.74399877e-01 -4.23324108e-01 1.37670004e+00 5.05431592e-01
-1.28333628e+00 -3.32199156e-01 -3.71053547e-01 1.95029423e-01
-3.78472924e-01 -3.77180338e-01 6.44461453e-01 5.74260950e-01
-1.14697325e+00 6.90991819e-01 -3.01454395e-01 -1.63103074e-01
-2.17030719e-01 4.45138842e-01 1.78266212e-01 6.66000247e-01
-1.23650432e+00 1.43783367e+00 -2.54931092e-01 -1.91652909e-01
-9.38963890e-01 -3.55952919e-01 -4.39560115e-01 4.37194496e-01
5.55499494e-01 -1.49972737e-01 2.31807542e+00 -6.99899852e-01
-2.52050209e+00 6.46647096e-01 -1.88006535e-02 -4.71184671e-01
5.42739511e-01 -6.49780780e-02 -2.65376002e-01 -1.39504299e-01
-1.10855795e-01 3.49502325e-01 5.47334731e-01 -9.29673672e-01
-1.02563906e+00 -1.77193061e-01 3.93246293e-01 7.53504783e-02
-2.30763137e-01 3.95279646e-01 2.04763383e-01 3.29459250e-01
-8.88633668e-01 -7.53753901e-01 -2.67186791e-01 -6.63986742e-01
-5.22316158e-01 -8.43323886e-01 6.07793868e-01 -3.99512917e-01
7.70812035e-01 -1.74364376e+00 -9.30597633e-02 5.64219477e-03
1.21006034e-01 6.53519928e-02 -3.76851141e-01 7.48138249e-01
7.13828504e-01 -4.81538969e-04 3.46139312e-01 -6.75790966e-01
3.20185393e-01 1.78162694e-01 -1.20965034e-01 -9.91105661e-02
3.70234907e-01 9.72952724e-01 -9.58615184e-01 -3.28259468e-01
4.86173183e-01 3.87973487e-02 -4.79282111e-01 1.15220034e+00
-7.51690149e-01 6.92227721e-01 -5.36040962e-01 5.55284202e-01
1.24308139e-01 -1.81853592e-01 4.67975080e-01 4.16372567e-01
-2.13903889e-01 7.99282491e-01 -4.94622171e-01 1.05454612e+00
-1.01589644e+00 2.50459164e-01 4.60130662e-01 -6.88774347e-01
1.11628222e+00 5.25908649e-01 5.92435241e-01 -4.47132140e-01
5.11696160e-01 -2.11343542e-01 1.68572888e-01 -6.50976300e-01
7.08652496e-01 1.41574005e-02 -5.83294630e-01 7.34534919e-01
3.45143169e-01 -2.64350742e-01 -7.48949721e-02 3.28950584e-01
1.37996805e+00 -3.68187606e-01 4.37861055e-01 1.82153985e-01
4.97052580e-01 4.08711806e-02 3.04804713e-01 1.03549719e+00
-7.63556898e-01 -1.50580093e-01 1.17050958e+00 -1.02691747e-01
-9.74269927e-01 -1.01162875e+00 1.41570061e-01 2.14014602e+00
-2.11275041e-01 -6.92932457e-02 -5.97007215e-01 -6.92455530e-01
6.03525192e-02 1.14585161e+00 -3.05088669e-01 -3.20698857e-01
-5.76695383e-01 -2.42329419e-01 4.44617420e-01 4.43976372e-01
3.20622265e-01 -1.77166820e+00 -1.87373668e-01 6.81927562e-01
-3.56873095e-01 -1.50006604e+00 -5.79311371e-01 4.57570642e-01
-4.17266905e-01 -7.03088939e-01 5.99759929e-02 -4.12431061e-01
-2.93181777e-01 -2.51074702e-01 1.28780591e+00 -1.14169560e-01
2.04551905e-01 6.71761334e-01 -5.31011045e-01 -5.81382215e-02
-1.13118136e+00 2.09380686e-01 -8.20358843e-03 1.36997089e-01
6.27875090e-01 -6.61199749e-01 -1.92507908e-01 7.18038738e-01
-2.38762591e-02 -5.44353008e-01 2.61921883e-01 1.13844621e+00
-2.64515489e-01 -6.23208940e-01 1.20379567e+00 -1.00403702e+00
1.43334496e+00 -5.15442073e-01 -4.21891302e-01 5.07967591e-01
-7.48467267e-01 1.57634377e-01 7.46985674e-01 -4.93013620e-01
-1.48268712e+00 -5.52973486e-02 -4.07641411e-01 1.36760175e-01
-4.43595827e-01 -2.59507541e-02 -1.03719503e-01 5.11961102e-01
4.81017679e-01 7.60867894e-02 1.76779017e-01 -4.65712190e-01
7.45127022e-01 1.38152981e+00 4.78738695e-01 -6.50047779e-01
7.27462992e-02 -1.70770079e-01 -6.26601338e-01 -6.52720392e-01
-8.49100053e-01 -8.50208521e-01 -6.37516677e-01 -5.35188913e-01
8.99181545e-01 -3.38863581e-01 -2.17440343e+00 4.48940903e-01
-1.45269322e+00 -8.68561625e-01 -3.55128706e-01 -3.21560912e-03
-9.71313655e-01 -7.81090483e-02 -1.20027936e+00 -1.52460837e+00
-5.25635839e-01 -1.15916121e+00 9.06612098e-01 2.36143738e-01
-5.62467039e-01 -1.02334452e+00 -1.93455890e-02 5.65538108e-01
7.38904595e-01 -3.23613495e-01 1.05230975e+00 -1.52701414e+00
-3.59506011e-01 -2.77510464e-01 1.76822856e-01 5.33549309e-01
1.99866071e-01 -1.65444165e-02 -1.24844253e+00 7.74404481e-02
2.40083799e-01 -1.08458447e+00 1.65389106e-01 2.66768038e-01
9.54044104e-01 -7.25468516e-01 -1.10721868e-02 -3.31596226e-01
5.43379545e-01 5.60680389e-01 8.47049430e-02 7.04996800e-03
-1.03918742e-02 9.99107063e-01 7.24977195e-01 1.06609094e+00
5.52638888e-01 9.36661422e-01 5.17463446e-01 3.89008343e-01
4.96197194e-01 -1.37960836e-01 6.96954250e-01 4.26247358e-01
2.12307349e-01 -3.32535833e-01 -4.29557562e-01 2.09943429e-01
-2.05194235e+00 -1.15252507e+00 -3.44591849e-02 1.89780915e+00
1.26975572e+00 3.45125884e-01 8.50019217e-01 -4.17178571e-01
7.97669411e-01 1.26081288e-01 -8.07929873e-01 -1.21863508e+00
1.95768043e-01 -1.68742728e-03 1.45629346e-01 9.10169184e-01
-6.80581927e-01 1.19880068e+00 7.12494326e+00 5.04409730e-01
-4.57807541e-01 5.38483024e-01 7.72369504e-01 -2.94065028e-01
-7.33986273e-02 -2.91896194e-01 -1.04609311e+00 2.58256733e-01
1.49624443e+00 -2.36549288e-01 8.91836882e-01 1.40575325e+00
5.10067642e-01 3.88897769e-02 -1.38990796e+00 3.54829252e-01
-3.45158398e-01 -1.15988731e+00 -5.35363793e-01 -7.41601884e-02
1.42045349e-01 9.38254595e-02 1.31269142e-01 1.01600695e+00
1.41763055e+00 -7.49988139e-01 5.20833611e-01 3.26965064e-01
5.39285362e-01 -4.21968043e-01 8.73674095e-01 8.06767166e-01
-7.95514822e-01 -6.52175665e-01 -2.73734275e-02 -4.70708497e-03
7.52872586e-01 2.10934594e-01 -1.56965744e+00 -3.24373066e-01
2.72374809e-01 1.29969284e-01 1.25608817e-01 1.51689187e-01
-1.01624377e-01 4.97844756e-01 -6.59792544e-03 -7.30674803e-01
3.26471329e-01 -3.71155053e-01 4.16494250e-01 1.32709026e+00
-3.22394490e-01 4.21845131e-02 4.40243125e-01 9.52706158e-01
-1.21934056e-01 -1.60303742e-01 -5.50107181e-01 -7.90766180e-02
7.77914345e-01 1.78873301e+00 -1.22195214e-01 -5.05726218e-01
-2.31982812e-01 7.88575053e-01 5.37855506e-01 5.59681617e-02
-7.70419180e-01 -1.25824943e-01 9.05439198e-01 -3.85650337e-01
-5.07164225e-02 3.82447504e-02 -3.71107101e-01 -6.43007636e-01
-3.74706537e-01 -1.06284225e+00 3.70724976e-01 -4.76503372e-01
-1.57447755e+00 9.81717706e-01 -2.42268339e-01 -6.39956117e-01
-9.96684253e-01 -7.08723426e-01 -9.36677516e-01 6.61811054e-01
-1.05752325e+00 -9.53756690e-01 -1.17192864e-02 6.10838175e-01
8.07423413e-01 -3.16332012e-01 1.04359770e+00 -9.76617634e-02
-5.18684208e-01 7.11610377e-01 -1.28548235e-01 1.17719829e-01
6.35440111e-01 -1.67450535e+00 2.49151632e-01 -3.98894131e-01
-2.22892597e-01 2.15982780e-01 8.20251286e-01 -2.60578722e-01
-1.03648567e+00 -6.62429988e-01 1.09688354e+00 -8.74472976e-01
9.86560106e-01 -4.93501455e-01 -6.94749892e-01 8.70486856e-01
5.61848581e-01 -7.75389314e-01 6.36114597e-01 8.65265071e-01
-3.15970570e-01 -2.30749305e-02 -1.55634224e+00 4.20158148e-01
8.09159100e-01 -8.12497735e-01 -5.73509455e-01 7.22736657e-01
1.09895182e+00 -2.23272353e-01 -1.10756552e+00 -2.05219314e-01
4.25615370e-01 -1.18799055e+00 5.69758415e-01 -1.27842402e+00
2.92180061e-01 8.95986855e-01 -1.68000281e-01 -1.45366848e+00
-2.41884783e-01 -1.05257380e+00 -1.33732259e-01 1.34582984e+00
7.91055143e-01 -7.30376422e-01 5.99241197e-01 1.23741436e+00
-1.49887592e-01 -6.90879643e-01 -1.04525411e+00 -5.97529531e-01
-6.17129207e-02 -3.83666039e-01 7.18471408e-01 4.83862191e-01
7.78298497e-01 7.40764022e-01 -7.70480037e-01 -2.86878258e-01
-6.01185858e-02 8.61520022e-02 7.92514384e-01 -1.02147150e+00
-5.46020448e-01 -4.28024113e-01 3.48460466e-01 -1.24182129e+00
4.92239922e-01 -5.28979242e-01 3.75301898e-01 -1.06145418e+00
9.34889764e-02 -3.10013890e-01 -3.29851992e-02 5.90031326e-01
1.86407015e-01 -7.20732033e-01 2.82036155e-01 -5.89296557e-02
-8.83294761e-01 5.91347933e-01 1.05088723e+00 -1.16945185e-01
-6.00404918e-01 8.03788364e-01 -5.38567662e-01 5.76249540e-01
9.42200243e-01 -3.58785093e-01 4.71073203e-02 2.19043985e-01
-7.83713534e-02 7.06300974e-01 5.19551933e-02 -2.44402230e-01
1.26530126e-01 -5.96287012e-01 -5.43049991e-01 -4.72995222e-01
9.48644996e-01 -5.04748464e-01 -5.40737510e-01 2.80192971e-01
-1.30174136e+00 9.23573747e-02 -3.74877512e-01 4.43067074e-01
8.07895511e-03 -4.06669378e-01 8.49506676e-01 -3.71709108e-01
-1.91768810e-01 2.01868489e-01 -9.00190949e-01 6.79776445e-02
7.74930596e-01 2.91893989e-01 -6.66593194e-01 -1.25737834e+00
-8.57222736e-01 6.16283000e-01 -3.80887449e-01 4.14307535e-01
1.75058231e-01 -9.15875018e-01 -5.48401594e-01 -2.16139242e-01
8.80267695e-02 -6.76854789e-01 1.03104852e-01 7.45707154e-01
2.53079265e-01 4.98158634e-01 -1.97666571e-01 -5.37432313e-01
-1.26696479e+00 3.21459532e-01 5.34190655e-01 -8.43490303e-01
-1.40268758e-01 7.75760233e-01 -6.46768272e-01 -7.39811242e-01
6.30861342e-01 -2.63361752e-01 -4.58319157e-01 2.49455318e-01
4.48151201e-01 4.79017258e-01 -1.46777287e-01 -2.40757808e-01
7.62943327e-02 -5.71034491e-01 -1.21890061e-01 -5.84812582e-01
1.18551433e+00 -2.73960948e-01 -6.84273466e-02 9.24123287e-01
9.33652818e-01 -3.42403293e-01 -1.24647665e+00 -4.88402933e-01
4.55947042e-01 -3.25391471e-01 -5.32090008e-01 -1.27038026e+00
-5.05757689e-01 7.23202527e-01 1.58464685e-01 1.04986942e+00
4.03473079e-01 3.86677891e-01 9.56336379e-01 1.05374300e+00
5.35407484e-01 -1.51506627e+00 6.82025552e-01 9.42454875e-01
8.98289382e-01 -1.56148183e+00 -8.98466051e-01 -3.47679481e-02
-1.56804943e+00 8.92243445e-01 9.76872563e-01 4.30656582e-01
5.01540840e-01 5.93709648e-01 3.90258729e-01 -4.61067148e-02
-1.66300642e+00 -1.78441420e-01 -6.01960361e-01 8.38630140e-01
4.44780886e-01 3.00310671e-01 1.36936814e-01 8.42402518e-01
-1.95042685e-01 -2.18117252e-01 4.67258304e-01 4.28340048e-01
-7.85204530e-01 -1.36250973e+00 9.97339636e-02 6.85535014e-01
-2.31354013e-01 -8.07506964e-02 -9.18398917e-01 3.27283144e-01
-5.08336544e-01 1.85178566e+00 -3.40880379e-02 -7.02475607e-01
4.93428051e-01 5.73249936e-01 -4.16367315e-02 -7.24815428e-01
-1.59165978e+00 -3.02992284e-01 9.89914596e-01 -6.86609864e-01
-5.58457617e-03 -6.41667128e-01 -1.27119207e+00 -6.19372249e-01
-3.98961455e-01 3.84986728e-01 4.33036208e-01 9.91095245e-01
2.55426615e-01 2.94151157e-01 1.49098349e+00 -6.39341772e-01
-1.65566695e+00 -1.68594420e+00 -7.16767013e-01 4.70595926e-01
1.66133478e-01 -4.56910223e-01 -6.02907360e-01 -4.59718198e-01] | [12.818747520446777, 7.981148719787598] |
5f09fb0f-1f11-4725-a167-cff12bc9cfd9 | decoupling-features-and-coordinates-for-few | 1911.11534 | null | https://arxiv.org/abs/1911.11534v4 | https://arxiv.org/pdf/1911.11534v4.pdf | Decoupling Features and Coordinates for Few-shot RGB Relocalization | Cross-scene model adaption is crucial for camera relocalization in real scenarios. It is often preferable that a pre-learned model can be fast adapted to a novel scene with as few training samples as possible. The existing state-of-the-art approaches, however, can hardly support such few-shot scene adaption due to the entangling of image feature extraction and scene coordinate regression. To address this issue, we approach camera relocalization with a decoupled solution where feature extraction, coordinate regression, and pose estimation are performed separately. Our key insight is that feature encoder used for coordinate regression should be learned by removing the distracting factor of coordinate systems, such that feature encoder is learned from multiple scenes for general feature representation and more important, view-insensitive capability. With this feature prior, and combined with a coordinate regressor, few-shot observations in a new scene are much easier to connect with the 3D world than the one with existing integrated solution. Experiments have shown the superiority of our approach compared to the state-of-the-art methods, producing higher accuracy on several scenes with diverse visual appearance and viewpoint distribution. | ['Songyin Wu', 'Shanghang Zhang', 'Yixin Zhuang', 'Siyan Dong', 'Kai Xu', 'Baoquan Chen'] | 2019-11-26 | null | null | null | null | ['camera-relocalization'] | ['computer-vision'] | [-1.00697996e-02 -2.95875043e-01 -7.59347007e-02 -4.06347483e-01
-5.48870206e-01 -5.04569352e-01 4.82462853e-01 -3.06943357e-01
-3.43050510e-01 5.21027565e-01 -3.84383649e-02 2.35818312e-01
4.65080738e-02 -3.73904765e-01 -7.45029211e-01 -6.57141924e-01
5.79415381e-01 3.48103106e-01 5.57944357e-01 -2.09403753e-01
3.05647761e-01 8.72407854e-01 -1.45929730e+00 -1.82064205e-01
5.95130861e-01 7.84659386e-01 5.85112989e-01 6.13114774e-01
1.82974875e-01 5.42441666e-01 -1.12473920e-01 -2.99239933e-01
4.55080330e-01 -2.57866383e-01 -3.47076952e-01 6.83367550e-01
6.42157078e-01 -3.26199621e-01 -4.61938679e-01 9.87902224e-01
4.23015982e-01 5.59348345e-01 4.83806670e-01 -8.36266458e-01
-3.46837819e-01 -1.75464883e-01 -8.69106293e-01 -2.71552689e-02
5.19778550e-01 1.45947486e-01 6.36288226e-01 -1.07865083e+00
8.22357297e-01 9.59498167e-01 5.07359445e-01 3.67982745e-01
-1.37138939e+00 -3.37322414e-01 6.44685209e-01 3.31995308e-01
-1.65286708e+00 -5.90164006e-01 1.07463133e+00 -4.14261758e-01
9.55106616e-01 1.65638089e-01 7.98217773e-01 9.42101538e-01
2.38579348e-01 4.38380957e-01 6.57411993e-01 -4.89946157e-01
1.36303425e-01 5.44394791e-01 -2.58133858e-01 6.71664596e-01
2.09928647e-01 7.64001207e-03 -4.24576759e-01 2.78679371e-01
9.89504993e-01 4.43215668e-01 -4.04564291e-01 -1.23561025e+00
-1.34850466e+00 5.85129499e-01 5.55091083e-01 7.37851784e-02
-3.06774676e-01 -1.09537831e-02 2.01723129e-01 1.41761452e-01
3.53771001e-01 6.49242580e-01 -5.57510257e-01 -9.65801999e-03
-8.04585218e-01 -3.89102637e-03 4.61027771e-01 1.37376499e+00
9.92430687e-01 1.15435004e-01 3.17627788e-01 6.39229774e-01
1.96861833e-01 4.51121598e-01 4.43868846e-01 -9.68803048e-01
3.25533241e-01 4.14596021e-01 9.20079574e-02 -1.04164898e+00
-3.42123061e-01 -4.20237720e-01 -7.70734966e-01 3.19949538e-01
1.20321564e-01 -5.54398708e-02 -6.59109414e-01 1.46017098e+00
6.44192159e-01 6.82534203e-02 1.37702627e-02 9.15773213e-01
3.19448650e-01 5.20805836e-01 -4.83550698e-01 -3.51821899e-01
1.12861216e+00 -1.07903409e+00 -4.43718374e-01 -4.77071434e-01
5.16218305e-01 -1.04441988e+00 1.04103160e+00 4.13954496e-01
-8.25908601e-01 -7.72863507e-01 -1.04491496e+00 -1.49571016e-01
-1.79062843e-01 1.48283020e-01 6.19508564e-01 2.85207570e-01
-7.90192008e-01 1.72496289e-01 -7.82542288e-01 -9.84250009e-01
-2.18965951e-02 3.18632871e-01 -8.50040555e-01 -1.89655975e-01
-5.23333669e-01 1.10654795e+00 2.62900025e-01 -4.41197231e-02
-6.45928800e-01 -4.22582120e-01 -9.47595477e-01 -1.16547868e-01
6.55054510e-01 -1.00207472e+00 1.00123405e+00 -1.07840109e+00
-1.86719704e+00 5.67771256e-01 -2.38604590e-01 2.36350801e-02
4.13821816e-01 -3.77663076e-01 -1.68590635e-01 8.60228464e-02
8.63597095e-02 4.37200904e-01 1.09613478e+00 -1.55176842e+00
-6.37885392e-01 -3.75883967e-01 2.01571867e-01 7.01069117e-01
-1.55239061e-01 -1.05839156e-01 -8.30918491e-01 -3.67569387e-01
4.70385134e-01 -9.98091459e-01 -4.06703919e-01 2.85419852e-01
-2.13659093e-01 3.43869567e-01 8.68719697e-01 -2.60080904e-01
8.21750998e-01 -2.32949805e+00 5.20734012e-01 -8.24005157e-02
-3.47383767e-02 -2.28842981e-02 -8.41395333e-02 4.49277282e-01
-2.54938900e-01 -4.30380881e-01 3.20935965e-01 -6.05529368e-01
-3.71222705e-01 5.83597831e-02 -1.95635334e-01 8.60728920e-01
-3.25398408e-02 4.16253984e-01 -8.09227347e-01 -2.90762305e-01
7.31208861e-01 3.72349203e-01 -7.87216008e-01 4.25210536e-01
-1.24618402e-02 7.85331011e-01 -4.41125840e-01 5.29249668e-01
7.34925151e-01 -1.58637807e-01 -4.87673134e-02 -4.31503385e-01
-3.13494384e-01 -1.86558813e-01 -1.33476174e+00 2.21889210e+00
-5.73193014e-01 5.98676026e-01 -7.15137720e-02 -8.38071108e-01
9.08902407e-01 3.20340879e-02 5.44616938e-01 -4.07484472e-01
1.42129242e-01 -1.55788854e-01 -2.49741241e-01 -6.13900125e-01
5.27106166e-01 7.22462013e-02 1.23755867e-02 8.92186537e-02
1.58937678e-01 -6.51411593e-01 -1.98044866e-01 1.30453750e-01
6.02697670e-01 5.15893579e-01 6.53077364e-01 -7.80196264e-02
5.95113575e-01 2.44562123e-02 7.10679293e-01 4.85163510e-01
5.68170426e-03 8.63938332e-01 -4.44398597e-02 -5.50469697e-01
-1.11807263e+00 -8.18077803e-01 1.61693301e-02 7.92100787e-01
6.32407069e-01 -5.03157914e-01 -5.20402551e-01 -7.42578924e-01
-3.50443393e-01 6.69943869e-01 -4.33127970e-01 -2.20725492e-01
-5.63285291e-01 -3.62997204e-01 -2.96827972e-01 4.06478822e-01
4.80384916e-01 -4.26881641e-01 -7.51775384e-01 5.51405735e-02
5.64490482e-02 -1.27920401e+00 -6.29047215e-01 2.19012573e-01
-8.35519671e-01 -1.13154209e+00 -4.93371874e-01 -5.14029622e-01
9.50307131e-01 9.99202847e-01 7.71323681e-01 -3.09008807e-01
-2.07876384e-01 7.13971019e-01 -5.05974233e-01 1.13322260e-02
6.38020635e-02 1.36630848e-01 3.05213958e-01 2.39883929e-01
1.31093964e-01 -6.13094091e-01 -6.08904183e-01 4.42952484e-01
-5.83822548e-01 3.93698514e-01 7.57688701e-01 8.79263818e-01
6.63037300e-01 -8.60993788e-02 -9.39342752e-02 -7.53425479e-01
-1.45699412e-01 -1.60123393e-01 -7.67973363e-01 3.10289264e-01
-4.37584937e-01 -1.27129965e-02 7.47029066e-01 -6.50478482e-01
-1.29741526e+00 7.29091644e-01 1.52599588e-01 -7.00888634e-01
-2.42518067e-01 1.07305765e-01 -5.35952806e-01 -3.88898104e-01
6.96210861e-01 2.91116446e-01 -1.77048057e-01 -4.10535157e-01
6.85093880e-01 2.72859991e-01 4.30754811e-01 -4.14657921e-01
1.12920046e+00 6.99345469e-01 1.64228022e-01 -9.91160631e-01
-8.61030757e-01 -6.21659279e-01 -1.41281295e+00 -2.78282076e-01
9.39579070e-01 -1.14769506e+00 -3.71315390e-01 2.05658525e-01
-1.26604569e+00 -2.80370098e-02 -4.77963477e-01 9.14382756e-01
-6.45626426e-01 4.59118098e-01 7.06776008e-02 -5.35325944e-01
2.05350935e-01 -1.26828277e+00 1.10604274e+00 2.99031645e-01
-3.49248648e-02 -9.57693577e-01 1.69840530e-01 1.29009843e-01
2.69542158e-01 1.49826318e-01 4.55329210e-01 -3.25460315e-01
-8.71532381e-01 -5.41883111e-01 -1.55114710e-01 2.33962491e-01
2.72698313e-01 1.70563638e-01 -1.09440899e+00 -5.04121184e-01
3.80931050e-01 -4.74318862e-02 5.74705303e-01 3.22520524e-01
8.07684839e-01 -2.54835300e-02 -4.32561696e-01 1.12004292e+00
1.62855244e+00 1.84354737e-01 4.21261549e-01 3.58614475e-01
1.01900995e+00 3.48228633e-01 9.22008216e-01 4.18950856e-01
4.53919470e-01 1.11297143e+00 6.00055933e-01 -1.74160957e-01
-2.96135210e-02 -2.04417810e-01 3.36919159e-01 9.04410899e-01
-1.39092162e-01 4.65443581e-02 -7.02201009e-01 2.04996184e-01
-2.04619980e+00 -7.46134222e-01 1.31278068e-01 2.31730413e+00
3.07020456e-01 -9.17141587e-02 -2.45948836e-01 -4.13984239e-01
5.22538662e-01 1.39530674e-01 -5.81634462e-01 2.51600873e-02
-5.81334718e-02 -5.42485893e-01 5.76570332e-01 5.71092725e-01
-1.04986787e+00 1.07231367e+00 6.06220865e+00 4.85168725e-01
-1.26766717e+00 1.65603876e-01 3.95852357e-01 -2.43818387e-01
6.86911568e-02 5.09525239e-01 -8.88784111e-01 9.02901739e-02
2.52608240e-01 -1.32951858e-02 5.34540415e-01 1.12805378e+00
1.30846202e-01 -2.75391906e-01 -1.33065462e+00 1.36008573e+00
6.49296701e-01 -9.73141253e-01 -1.23289175e-01 4.57501672e-02
8.24985862e-01 1.22763567e-01 -1.25922307e-01 1.63607836e-01
-4.24923487e-02 -4.60401326e-01 8.22365642e-01 8.71711552e-01
7.22069263e-01 -5.46286821e-01 6.00626588e-01 4.91733134e-01
-1.26692140e+00 -1.37060866e-01 -7.74344444e-01 -8.71407613e-02
3.53849679e-01 4.15864289e-01 -5.89770496e-01 8.35357487e-01
5.53251863e-01 1.02252901e+00 -7.60761321e-01 1.02196693e+00
-1.36800736e-01 -2.07565278e-01 -3.51767361e-01 2.79416442e-01
-9.83674824e-02 -3.52450341e-01 6.45234346e-01 7.89542019e-01
4.98602420e-01 1.10123530e-01 3.89751613e-01 4.51883852e-01
3.55131119e-01 1.02431759e-01 -1.02126265e+00 7.11907923e-01
2.48096749e-01 1.33215618e+00 -6.76363289e-01 -2.23255634e-01
-8.24259222e-01 1.23692107e+00 4.53936696e-01 5.09868681e-01
-8.97862673e-01 -1.10631421e-01 6.09730780e-01 2.84070402e-01
5.18856704e-01 -5.48218191e-01 4.30484824e-02 -1.67889249e+00
-6.49169683e-02 -6.65112495e-01 3.42887677e-02 -1.02251351e+00
-1.13125503e+00 7.87762642e-01 2.40974113e-01 -1.78981161e+00
-4.36181128e-01 -4.91422981e-01 -5.10071218e-01 6.26816154e-01
-1.43222558e+00 -1.50026345e+00 -5.28164208e-01 8.66420984e-01
9.49624717e-01 -2.97692508e-01 7.24235177e-01 8.04528594e-02
-5.97373188e-01 3.98777783e-01 1.87833846e-01 -3.42258424e-01
1.06199074e+00 -9.15333211e-01 6.95848167e-02 1.16223526e+00
2.47476637e-01 5.50950527e-01 6.84209228e-01 -3.91155362e-01
-1.70134163e+00 -1.09915411e+00 4.71729428e-01 -7.36489773e-01
4.72503275e-01 -6.52007282e-01 -6.81633770e-01 8.18948209e-01
2.91694384e-02 3.59512866e-01 3.92418176e-01 1.51276514e-01
-1.91364601e-01 -4.93917435e-01 -7.51536965e-01 6.77468598e-01
1.03614104e+00 -5.16969621e-01 -4.00938749e-01 2.84453422e-01
7.15467155e-01 -5.33487618e-01 -6.10999823e-01 1.38578892e-01
5.18710434e-01 -1.08897138e+00 1.00857019e+00 -2.85461962e-01
-1.32286832e-01 -6.35066867e-01 -5.66301107e-01 -1.30842328e+00
-6.93872750e-01 -6.46418810e-01 1.35555491e-02 1.16745603e+00
6.46307394e-02 -5.11568308e-01 5.23566425e-01 5.93320727e-01
-1.66209146e-01 -5.56417048e-01 -8.49029005e-01 -7.95557201e-01
-4.32443678e-01 -2.11536527e-01 4.45751578e-01 8.41531336e-01
-1.95836425e-01 7.82215595e-01 -6.77878797e-01 3.60698074e-01
5.13860404e-01 2.66911447e-01 1.53922927e+00 -1.13302255e+00
-5.19004583e-01 2.43021805e-05 -7.65533566e-01 -1.26400006e+00
6.16624281e-02 -3.41079086e-01 4.54894714e-02 -1.24502182e+00
3.78531218e-01 -2.45998129e-01 -3.25710215e-02 1.65597826e-01
-7.70414621e-02 1.00391172e-01 4.03302610e-01 3.35878909e-01
-6.43851936e-01 7.72022367e-01 1.31989264e+00 7.36882612e-02
-3.76284838e-01 4.23488282e-02 -5.30677319e-01 9.23216462e-01
4.45648253e-01 -3.99780661e-01 -6.99662387e-01 -6.53631806e-01
2.16293216e-01 4.40771133e-03 2.31584683e-01 -1.20473433e+00
5.84997296e-01 -5.05144536e-01 6.11521125e-01 -5.42077363e-01
7.51118720e-01 -1.30133760e+00 4.00303245e-01 7.11804256e-02
1.06898502e-01 1.65753648e-01 1.82081405e-02 6.87210739e-01
-1.45825148e-01 -1.33565634e-01 8.65851581e-01 -1.84581608e-01
-8.99150848e-01 4.59686667e-01 -1.32476673e-01 -4.50927764e-01
1.30264008e+00 -5.47997117e-01 8.00706968e-02 -4.47445363e-01
-7.44333327e-01 -1.59757376e-01 1.07671738e+00 4.21274185e-01
6.70946240e-01 -1.24978232e+00 -4.61248219e-01 6.75833583e-01
4.62433487e-01 4.01541770e-01 3.27205002e-01 9.60663915e-01
-5.27506709e-01 2.80319571e-01 -2.27423221e-01 -9.70644176e-01
-1.17740357e+00 1.01033068e+00 2.40834385e-01 2.83356514e-02
-8.87179434e-01 5.87531388e-01 7.32085526e-01 -3.34525645e-01
5.17760366e-02 -2.20748425e-01 7.73140565e-02 -1.92707911e-01
2.97453731e-01 5.40461205e-02 -8.81708413e-02 -8.87512088e-01
-2.31600955e-01 1.24683034e+00 -3.02402109e-01 -2.96659786e-02
1.30218136e+00 -7.75863588e-01 9.70185399e-02 7.40269244e-01
1.08258760e+00 2.01845348e-01 -1.75168490e+00 -3.11917424e-01
-4.48681295e-01 -9.56760526e-01 3.00895944e-02 -3.18114251e-01
-9.51563299e-01 8.78559709e-01 6.66609228e-01 -1.90349743e-01
1.21455920e+00 -5.88246323e-02 1.77029759e-01 5.92856288e-01
5.08719981e-01 -1.06433761e+00 2.72899002e-01 4.89963472e-01
9.16256785e-01 -1.47103226e+00 4.03584927e-01 -4.64794785e-01
-8.15579236e-01 1.18400896e+00 8.06582749e-01 -2.41845235e-01
6.95297301e-01 -1.86540876e-02 -1.05313985e-02 -1.59817383e-01
-6.52694583e-01 2.55278535e-02 4.20058221e-01 6.73360825e-01
7.82318637e-02 -2.87504315e-01 5.94750762e-01 2.21787453e-01
1.16291784e-01 -4.74733621e-01 4.39058185e-01 6.79077268e-01
-3.10059756e-01 -9.69195902e-01 -3.11197221e-01 -4.21115570e-02
1.58539444e-01 8.70197192e-02 -1.76260442e-01 1.07898057e+00
6.23504855e-02 6.12405419e-01 5.77418767e-02 -3.13723862e-01
4.64261860e-01 -2.38085493e-01 7.19524980e-01 -9.20900881e-01
-8.19687396e-02 4.73004758e-01 -2.65384614e-01 -9.77153480e-01
-3.71166855e-01 -8.16473663e-01 -7.76988268e-01 -3.31701450e-02
-7.64698505e-01 -2.40423843e-01 7.15899408e-01 8.86297643e-01
4.05071080e-01 4.80157763e-01 1.00927818e+00 -1.51725376e+00
-2.68981606e-01 -7.84847558e-01 -6.29732728e-01 2.73342043e-01
5.50543189e-01 -7.94161558e-01 -2.43638605e-01 2.66203642e-01] | [8.060348510742188, -2.3362889289855957] |
6a25ebab-f5e3-460e-8aab-64cf1d8785d0 | backdoor-attack-with-sparse-and-invisible | 2306.06209 | null | https://arxiv.org/abs/2306.06209v1 | https://arxiv.org/pdf/2306.06209v1.pdf | Backdoor Attack with Sparse and Invisible Trigger | Deep neural networks (DNNs) are vulnerable to backdoor attacks, where the adversary manipulates a small portion of training data such that the victim model predicts normally on the benign samples but classifies the triggered samples as the target class. The backdoor attack is an emerging yet threatening training-phase threat, leading to serious risks in DNN-based applications. In this paper, we revisit the trigger patterns of existing backdoor attacks. We reveal that they are either visible or not sparse and therefore are not stealthy enough. More importantly, it is not feasible to simply combine existing methods to design an effective sparse and invisible backdoor attack. To address this problem, we formulate the trigger generation as a bi-level optimization problem with sparsity and invisibility constraints and propose an effective method to solve it. The proposed method is dubbed sparse and invisible backdoor attack (SIBA). We conduct extensive experiments on benchmark datasets under different settings, which verify the effectiveness of our attack and its resistance to existing backdoor defenses. The codes for reproducing main experiments are available at \url{https://github.com/YinghuaGao/SIBA}. | ['Qian Wang', 'Shu-Tao Xia', 'Xueluan Gong', 'Yiming Li', 'Yinghua Gao'] | 2023-05-11 | null | null | null | null | ['backdoor-attack'] | ['adversarial'] | [ 2.13233948e-01 -7.71291628e-02 -2.93952286e-01 3.12923267e-02
-3.24609250e-01 -1.10263622e+00 4.96086478e-01 -3.36330771e-01
-2.34071508e-01 6.16286933e-01 -2.64416575e-01 -8.40022445e-01
-4.74652983e-02 -7.97295570e-01 -9.19865251e-01 -9.47084963e-01
-5.24838343e-02 -1.30760670e-01 2.69718111e-01 -1.70596182e-01
1.71171039e-01 5.82405567e-01 -9.11775827e-01 1.53295293e-01
5.75834632e-01 9.53654468e-01 -1.64042354e-01 3.09253544e-01
2.10162029e-01 6.60274386e-01 -1.00383401e+00 -4.87536341e-01
6.57981455e-01 -3.59982312e-01 -3.96791369e-01 -5.53687096e-01
3.64096344e-01 -5.61133981e-01 -8.04972947e-01 1.54390407e+00
3.45984876e-01 -1.51701972e-01 2.26664409e-01 -1.59428728e+00
-5.55390060e-01 7.21260428e-01 -6.84071422e-01 5.85238397e-01
5.96734360e-02 2.09533706e-01 6.36159718e-01 -6.72455609e-01
2.79743135e-01 1.12571037e+00 2.65524983e-01 9.92291749e-01
-8.11114907e-01 -1.54076552e+00 3.89137685e-01 1.22062042e-01
-1.15599537e+00 -5.41228354e-01 9.31778252e-01 -2.40178078e-01
3.42110515e-01 6.54079974e-01 4.96055424e-01 1.79591823e+00
2.24879742e-01 7.50431478e-01 1.00196207e+00 -2.08724905e-02
3.38116497e-01 1.66528523e-01 2.00729907e-01 6.08415663e-01
6.38258874e-01 8.37802649e-01 -3.73314828e-01 -6.77706301e-01
6.00025415e-01 2.69491941e-01 -5.67107499e-01 6.13945201e-02
-8.19910049e-01 9.99098182e-01 4.52095509e-01 1.05607755e-01
1.53226620e-02 1.06114671e-01 2.61165082e-01 3.51270616e-01
2.18960583e-01 1.51688129e-01 -7.73632526e-02 2.00855121e-01
-7.14615047e-01 2.84324378e-01 7.45145202e-01 6.22980714e-01
3.41044575e-01 3.23781252e-01 9.19387788e-02 2.45953590e-01
3.19987983e-01 5.31383097e-01 1.99377552e-01 -4.33663428e-01
6.76987708e-01 9.63504910e-02 -1.60420015e-01 -1.29391563e+00
1.40598314e-02 -7.31778622e-01 -7.89270103e-01 8.51853788e-02
2.60210901e-01 -4.91652846e-01 -7.84706354e-01 1.86021972e+00
5.00619113e-01 7.97854304e-01 -1.04745798e-01 7.91582167e-01
5.85613251e-01 8.18549752e-01 -1.74138516e-01 -2.01991454e-01
1.19242311e+00 -7.11889327e-01 -7.79722035e-01 -2.95217514e-01
5.01701772e-01 -4.80362177e-01 8.04022789e-01 4.54435229e-01
-6.62079155e-01 3.70800495e-02 -1.34536123e+00 5.29254198e-01
-2.92222947e-01 -2.81342506e-01 6.45677507e-01 1.06312549e+00
-6.43276095e-01 2.55133748e-01 -7.78477907e-01 2.79204398e-01
7.96751678e-01 4.83811945e-01 -2.88744092e-01 -1.03614993e-01
-1.53424513e+00 2.37637833e-01 4.69009012e-01 3.04237574e-01
-1.57021177e+00 -6.22358918e-01 -5.00326693e-01 -1.96038038e-01
7.23396361e-01 -1.81549385e-01 8.75774980e-01 -4.96322840e-01
-1.22577882e+00 5.69702625e-01 3.57558876e-01 -6.50468826e-01
5.96497059e-01 -2.85533875e-01 -4.92681175e-01 7.41222575e-02
-1.79516792e-01 6.06465824e-02 1.13494778e+00 -1.28364587e+00
-3.39956820e-01 -4.58448470e-01 2.45958865e-01 -3.75349998e-01
-8.91937733e-01 2.68072516e-01 -3.91778611e-02 -8.38433385e-01
-1.69318557e-01 -9.96856928e-01 -1.57197967e-01 -1.15001827e-01
-1.03517652e+00 1.23760931e-01 1.46093595e+00 -2.26012558e-01
1.28939557e+00 -2.40151763e+00 -2.31255338e-01 3.09246600e-01
5.40882230e-01 8.28717351e-01 -1.13948449e-01 4.19575393e-01
-1.45126700e-01 3.60830277e-01 -1.52705878e-01 -1.36837333e-01
-2.59359628e-01 2.59145230e-01 -1.25145507e+00 7.89188206e-01
-2.75591999e-01 5.61749458e-01 -7.71515667e-01 -1.22841736e-02
-1.04296930e-01 4.88681883e-01 -4.74671632e-01 2.01343745e-01
5.28493188e-02 3.35101604e-01 -8.95773172e-01 8.22368085e-01
9.42592740e-01 3.29317488e-02 -1.22409821e-01 -6.32497221e-02
1.77030131e-01 1.27043128e-01 -9.13342237e-01 7.44061530e-01
4.60307114e-02 4.51038599e-01 1.34304374e-01 -9.22867119e-01
8.92350554e-01 3.26187164e-01 4.99220863e-02 -1.88059643e-01
5.63619375e-01 3.21355402e-01 1.23400658e-01 -1.69626251e-01
3.26386094e-02 -1.44375652e-01 -4.68374267e-02 4.69483167e-01
-2.07732886e-01 6.22271419e-01 -2.52496988e-01 3.19326520e-01
1.29802513e+00 -5.23192823e-01 3.47788706e-02 -2.61999637e-01
4.71658081e-01 -3.35812479e-01 8.28642488e-01 9.85369265e-01
-4.79262978e-01 1.63686112e-01 7.46574998e-01 -3.70760947e-01
-3.96294117e-01 -1.14262486e+00 -4.34502810e-02 6.83998525e-01
3.87658298e-01 -3.01407546e-01 -8.62678707e-01 -1.09444797e+00
7.84920901e-02 5.74038208e-01 -6.27156675e-01 -7.35350728e-01
-4.44804907e-01 -6.87216222e-01 1.08815515e+00 1.83001578e-01
6.56377614e-01 -7.84888148e-01 -3.96120042e-01 -2.13883907e-01
2.49861497e-02 -1.23940504e+00 -5.46683967e-01 1.45629093e-01
-7.18367636e-01 -1.15591526e+00 -5.53019822e-01 -4.58263695e-01
7.27415800e-01 5.57776093e-01 4.21133667e-01 2.08259791e-01
-2.66054153e-01 -2.02267081e-01 -2.17858240e-01 -5.68278730e-01
-3.33696425e-01 -1.27997279e-01 4.90964264e-01 2.54052907e-01
2.29911610e-01 -6.31213963e-01 -5.24407923e-01 3.62639755e-01
-1.10387981e+00 -4.40284133e-01 4.34837699e-01 7.13911057e-01
4.55025822e-01 2.88217187e-01 5.16966403e-01 -1.12235773e+00
7.44238734e-01 -7.74673879e-01 -9.68049943e-01 -7.81035691e-04
-1.86052397e-01 -3.21734637e-01 9.18054104e-01 -8.66494298e-01
-5.05440712e-01 -8.13560188e-02 -2.39513263e-01 -1.09984124e+00
-1.59873739e-02 2.41850674e-01 -4.83668238e-01 -4.46805894e-01
5.69528341e-01 4.12172556e-01 -3.39634717e-01 -4.56863374e-01
4.30862382e-02 4.85325187e-01 4.51991498e-01 -5.77691078e-01
1.52281737e+00 7.51363516e-01 -6.77495673e-02 -8.13590050e-01
-8.61695707e-01 9.62730795e-02 2.41628706e-01 -1.92484602e-01
4.89154696e-01 -6.55253232e-01 -7.65233874e-01 5.84760249e-01
-1.07507086e+00 -1.65388167e-01 1.16942242e-01 3.43347609e-01
9.47971717e-02 4.43099976e-01 -5.94733119e-01 -8.86408150e-01
-4.44474041e-01 -1.25030911e+00 5.18071234e-01 2.59397179e-01
2.06615239e-01 -7.75409162e-01 -4.05196063e-02 4.29027170e-01
2.07285970e-01 5.87083757e-01 7.48076558e-01 -1.07692158e+00
-6.66195273e-01 -5.02626896e-01 3.18615772e-02 3.10829043e-01
6.69509992e-02 8.22981969e-02 -1.02821994e+00 -5.62491000e-01
5.86366892e-01 -3.18215936e-01 9.56944168e-01 9.46902186e-02
1.39996159e+00 -7.89115131e-01 -4.77188975e-01 9.87494171e-01
1.22157347e+00 4.79082972e-01 5.24889112e-01 4.07042086e-01
7.99721301e-01 4.02101487e-01 3.26693386e-01 5.02365828e-01
-2.59647161e-01 4.19925272e-01 1.03518045e+00 5.14758676e-02
5.35831511e-01 -4.98440921e-01 6.28784537e-01 4.21397656e-01
2.76445061e-01 -5.76011717e-01 -7.70891309e-01 2.16171831e-01
-1.53885603e+00 -1.02596509e+00 7.15978518e-02 2.20987940e+00
6.34690642e-01 5.24583161e-01 6.18454590e-02 2.95982271e-01
7.94534683e-01 4.77565259e-01 -5.73444784e-01 -3.80341381e-01
-6.74730837e-02 2.22004596e-02 6.63843334e-01 3.68595630e-01
-1.19746709e+00 8.87698591e-01 5.29651642e+00 1.11830699e+00
-1.24256968e+00 2.44495258e-01 7.84082890e-01 -4.52096999e-01
-4.19836998e-01 6.41818941e-02 -1.05892336e+00 6.08927071e-01
6.98891401e-01 -2.69064695e-01 4.94122177e-01 8.69460166e-01
3.07409111e-02 5.37546754e-01 -9.13635194e-01 8.38845789e-01
-6.00894690e-02 -1.41394472e+00 2.42111921e-01 5.31301200e-01
5.60889363e-01 1.07250549e-01 5.50480783e-01 1.37630597e-01
2.38870010e-01 -1.05298126e+00 7.37997055e-01 -1.28957376e-01
5.94714165e-01 -1.04875112e+00 3.62001240e-01 6.28997266e-01
-9.58130956e-01 -4.21664447e-01 -4.49569702e-01 2.29555681e-01
-1.67266056e-01 5.97201288e-01 -2.39722803e-01 3.48210782e-01
5.27084827e-01 3.88001114e-01 -1.64715052e-01 6.97269142e-01
-4.33251172e-01 9.72118139e-01 -4.49783564e-01 6.05053385e-04
5.03165066e-01 3.92630063e-02 1.09123325e+00 7.15575933e-01
1.50472283e-01 1.20701797e-01 1.78615019e-01 1.09344447e+00
-3.55195969e-01 -2.43674144e-01 -9.86253381e-01 -3.44293594e-01
8.11131954e-01 1.12947404e+00 -6.23929203e-01 7.98626393e-02
-3.25921588e-02 5.97844422e-01 1.23581201e-01 4.37249810e-01
-1.13161790e+00 -4.75029737e-01 8.70245099e-01 -7.50819920e-03
1.19790696e-02 -6.29274771e-02 -9.52162221e-02 -1.33273327e+00
8.55607763e-02 -1.18508291e+00 4.83932287e-01 -1.16835840e-01
-1.20718348e+00 8.21635365e-01 -1.32274508e-01 -1.22796762e+00
3.14829916e-01 -5.25574625e-01 -1.06139565e+00 5.79849482e-01
-1.29940820e+00 -7.67402470e-01 1.27278194e-01 9.19558525e-01
4.32725586e-02 -4.28893089e-01 5.66092610e-01 3.78779978e-01
-1.13139129e+00 1.10897183e+00 1.44896850e-01 3.82808596e-01
2.31583521e-01 -5.31721056e-01 2.87916601e-01 1.22711754e+00
2.93412209e-01 9.88066614e-01 7.74555266e-01 -6.89321697e-01
-1.64332950e+00 -1.21486461e+00 3.44534874e-01 -2.38193408e-01
8.50483716e-01 -7.90339530e-01 -8.69119942e-01 6.83255374e-01
-7.86500722e-02 3.86943161e-01 8.43953013e-01 -4.03132319e-01
-6.84278071e-01 -1.13142163e-01 -1.20686495e+00 7.99107790e-01
9.65918422e-01 -4.52210426e-01 -2.45040208e-01 4.86985087e-01
9.71637249e-01 -3.26996595e-01 -1.72530025e-01 2.89563954e-01
3.35071892e-01 -9.26488578e-01 1.01084721e+00 -6.97189510e-01
3.16411793e-01 -7.24816369e-03 -2.69847363e-01 -8.09635341e-01
5.41940220e-02 -1.08002782e+00 -6.12524688e-01 1.05794716e+00
1.15694508e-01 -1.25168920e+00 1.12052953e+00 5.46830669e-02
1.96518600e-01 -1.05164802e+00 -1.23717356e+00 -1.21948481e+00
1.32789806e-01 -3.21686655e-01 7.23975718e-01 1.18251765e+00
-1.81032598e-01 -1.15023531e-01 -6.71078384e-01 6.14162505e-01
1.07417274e+00 -1.75079852e-01 6.33330941e-01 -8.74217927e-01
-4.57102269e-01 -3.57001692e-01 -2.95778066e-01 -8.50962341e-01
1.89142421e-01 -8.69003713e-01 -3.76203954e-01 -6.24811530e-01
-4.62716119e-03 -3.86974186e-01 -7.28196025e-01 7.02087164e-01
6.88751182e-03 2.24667892e-01 2.81988829e-01 2.79268444e-01
-1.03004865e-01 6.70262575e-01 1.03659201e+00 -1.77024305e-01
-2.43668519e-02 3.36848408e-01 -1.02950203e+00 6.60149038e-01
1.04711831e+00 -9.37278926e-01 -8.38850379e-01 -3.36906403e-01
7.07649887e-02 1.70734197e-01 6.85132384e-01 -8.70311081e-01
1.20683201e-01 -3.25215101e-01 -3.04299220e-02 -5.19229770e-01
3.12074691e-01 -8.71852756e-01 5.06358780e-02 1.00832582e+00
-2.94752717e-01 -1.72715217e-01 3.07016134e-01 8.92682731e-01
-2.64775623e-02 -9.79942232e-02 7.56739557e-01 1.42205209e-01
-1.82744578e-01 8.11425865e-01 -2.13925391e-01 2.11519644e-01
1.21218181e+00 -1.50652081e-01 -8.06282759e-01 -2.24297211e-01
-2.97748953e-01 2.12604448e-01 1.10249043e-01 5.09091973e-01
9.67809141e-01 -1.29094338e+00 -5.52086234e-01 4.25950229e-01
-5.09807281e-02 -4.09937948e-01 2.78476655e-01 5.74096501e-01
-3.50224227e-01 4.62308645e-01 -4.71818261e-03 -1.52973354e-01
-1.45618045e+00 7.58477390e-01 5.08502185e-01 -7.83419907e-02
-6.88245118e-01 1.19426918e+00 6.09184384e-01 -1.51322603e-01
4.55377400e-01 1.44006640e-01 -1.14909343e-01 -3.38196397e-01
7.75285125e-01 1.15119301e-01 -2.73872435e-01 -4.45445240e-01
-5.94768107e-01 -2.23620031e-02 -4.74228114e-01 1.32454082e-01
1.08738172e+00 4.29553151e-01 -4.08289023e-02 -1.22454599e-01
1.38772023e+00 1.33366838e-01 -9.71622050e-01 -2.07091406e-01
-3.84234011e-01 -8.06812525e-01 1.42870367e-01 -4.15689230e-01
-1.56613386e+00 8.61287355e-01 4.48557198e-01 3.74620974e-01
1.03190076e+00 -2.43573293e-01 1.26395798e+00 5.03965616e-01
3.94681364e-01 -5.12010872e-01 8.99272189e-02 3.84882033e-01
7.14012861e-01 -6.25118673e-01 -1.78396881e-01 -4.63597149e-01
-1.90391541e-01 7.98528075e-01 7.73239911e-01 -3.24607432e-01
9.23100650e-01 5.48457921e-01 -1.23559339e-02 -1.84610188e-01
-7.80257106e-01 6.14260554e-01 -1.74436808e-01 3.46175283e-01
-3.62451166e-01 4.37101871e-02 -9.89353731e-02 9.35391665e-01
-2.25995794e-01 -2.23549455e-01 5.33536553e-01 9.62200403e-01
-3.81992906e-01 -1.08136904e+00 -6.10428810e-01 3.04076910e-01
-7.26099789e-01 1.45541374e-02 -6.28618658e-01 4.81273621e-01
1.07295580e-01 9.34343100e-01 -4.40307677e-01 -1.05469656e+00
-1.55666908e-02 -4.51501340e-01 4.26233932e-02 -4.93016571e-01
-4.93880391e-01 -1.09651111e-01 -2.19201192e-01 -6.00652158e-01
3.16189051e-01 -2.59916782e-01 -9.54892755e-01 -5.58206797e-01
-5.48851609e-01 2.59956896e-01 4.10817981e-01 7.41899252e-01
1.72121808e-01 3.76666397e-01 1.14743817e+00 -4.16787177e-01
-1.09736919e+00 -2.58720279e-01 -6.24748051e-01 -6.77272826e-02
4.79758054e-01 -7.04063952e-01 -8.32985401e-01 -5.87072313e-01] | [5.713866710662842, 7.706539154052734] |
acf101cc-da87-43de-aa73-967bfafc674c | synergy-between-3dmm-and-3d-landmarks-for | 2110.09772 | null | https://arxiv.org/abs/2110.09772v2 | https://arxiv.org/pdf/2110.09772v2.pdf | Synergy between 3DMM and 3D Landmarks for Accurate 3D Facial Geometry | This work studies learning from a synergy process of 3D Morphable Models (3DMM) and 3D facial landmarks to predict complete 3D facial geometry, including 3D alignment, face orientation, and 3D face modeling. Our synergy process leverages a representation cycle for 3DMM parameters and 3D landmarks. 3D landmarks can be extracted and refined from face meshes built by 3DMM parameters. We next reverse the representation direction and show that predicting 3DMM parameters from sparse 3D landmarks improves the information flow. Together we create a synergy process that utilizes the relation between 3D landmarks and 3DMM parameters, and they collaboratively contribute to better performance. We extensively validate our contribution on full tasks of facial geometry prediction and show our superior and robust performance on these tasks for various scenarios. Particularly, we adopt only simple and widely-used network operations to attain fast and accurate facial geometry prediction. Codes and data: https://choyingw.github.io/works/SynergyNet/ | ['Ulrich Neumann', 'Qiangeng Xu', 'Cho-Ying Wu'] | 2021-10-19 | null | null | null | null | ['3d-face-reconstruction', '3d-face-modeling', 'head-pose-estimation', 'face-alignment'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [-3.95375192e-01 3.08670402e-01 -4.38079000e-01 -6.11012399e-01
-6.05729342e-01 -3.70610267e-01 4.62564379e-01 -3.23566109e-01
2.32398674e-01 5.52190654e-02 4.69592392e-01 -1.05798729e-01
-2.40883231e-02 -6.57500029e-01 -7.57242560e-01 -1.67097434e-01
-3.19647968e-01 6.08298659e-01 -3.39962512e-01 -1.22648954e-01
2.37097442e-01 1.14884460e+00 -1.46635580e+00 5.07460460e-02
4.34239328e-01 1.08616900e+00 -5.23165345e-01 4.08960253e-01
-1.36452556e-01 9.90053415e-02 2.25762144e-01 -5.03044426e-01
6.84386194e-01 5.32130934e-02 -8.07503164e-01 2.17482433e-01
8.68911386e-01 -3.83884221e-01 -2.24270672e-01 6.10077441e-01
6.18690133e-01 -2.62631029e-02 8.23957384e-01 -1.40323138e+00
-5.12816787e-01 9.25363749e-02 -8.41796637e-01 -3.35377634e-01
5.01545906e-01 -3.50398682e-02 6.66823387e-01 -1.54973483e+00
6.97300911e-01 1.60581589e+00 1.08654523e+00 6.55183375e-01
-1.05290103e+00 -1.10508692e+00 2.06101805e-01 -3.35380137e-02
-1.77938342e+00 -1.06206739e+00 9.69728768e-01 -4.55363542e-01
6.34082437e-01 1.04989335e-01 9.29092407e-01 7.50676870e-01
1.47945076e-01 4.02371019e-01 6.41287386e-01 -2.05277368e-01
-9.35115814e-02 -3.68406296e-01 -4.05933559e-01 1.50793898e+00
-6.68336824e-02 -1.35272020e-03 -8.26620579e-01 -2.71999419e-01
1.28436279e+00 5.74595444e-02 1.04888998e-01 -2.87882566e-01
-4.85206485e-01 6.55801833e-01 3.95085067e-01 -2.11028457e-01
-2.33742326e-01 2.72833824e-01 -8.94103572e-02 1.69997394e-01
8.99940133e-01 8.29835981e-02 -5.17708659e-01 -3.61210480e-02
-9.56882060e-01 1.61200956e-01 7.77190328e-01 1.22419024e+00
1.19332969e+00 -1.79994050e-02 3.07901323e-01 9.68450308e-01
7.70286977e-01 7.13415027e-01 -1.48041502e-01 -1.42043197e+00
2.20065817e-01 9.12153304e-01 -3.27995777e-01 -1.40232313e+00
-5.89091182e-01 3.68366428e-02 -8.24824393e-01 1.42336726e-01
1.92568898e-01 -9.69170853e-02 -9.70781922e-01 1.64486396e+00
8.33608091e-01 5.32687962e-01 -6.33798718e-01 7.50054598e-01
1.14519060e+00 2.56480068e-01 -1.76111087e-01 2.56898612e-01
1.03393734e+00 -7.60218024e-01 -1.59702823e-01 1.05181873e-01
7.58989930e-01 -8.20624888e-01 5.99996030e-01 8.27392265e-02
-1.47231269e+00 -4.11908656e-01 -5.97250164e-01 -3.57073367e-01
-8.60370770e-02 7.66332727e-03 8.33691418e-01 4.17788178e-01
-1.51751149e+00 7.75198996e-01 -9.17999923e-01 -1.98382318e-01
1.11150169e+00 7.86942124e-01 -8.18029165e-01 -2.09204152e-01
-6.10354960e-01 8.36452305e-01 -4.36383128e-01 4.19883937e-01
-9.13464844e-01 -1.16611600e+00 -1.09749615e+00 -2.99171090e-01
1.11987591e-01 -9.03884470e-01 9.60097015e-01 -4.73787904e-01
-1.61342621e+00 1.32344472e+00 -5.90461612e-01 1.95262596e-01
2.80666679e-01 -1.14469260e-01 4.74587083e-02 1.74450025e-01
-1.60223693e-01 9.72906291e-01 9.56437707e-01 -1.06488872e+00
-1.63890928e-01 -7.18615413e-01 -2.60877639e-01 1.61481217e-01
-2.90170014e-01 -1.14932038e-01 -8.59298944e-01 -3.52291912e-01
5.72383821e-01 -8.55012476e-01 -3.53579313e-01 6.29675984e-01
-3.93576115e-01 -2.33141452e-01 5.21762967e-01 -7.81180918e-01
8.90876710e-01 -1.90502942e+00 1.68440148e-01 7.36521423e-01
6.45262659e-01 3.52243483e-02 -4.75512892e-01 1.87160656e-01
-1.69386342e-03 4.83326793e-01 3.00731678e-02 -1.05972552e+00
4.91221920e-02 3.70767042e-02 6.11490980e-02 6.78481162e-01
3.52281660e-01 1.13083625e+00 -5.44523060e-01 -7.70621538e-01
2.19403014e-01 1.02734494e+00 -9.22465146e-01 -1.39329955e-02
-6.70754258e-03 5.68980098e-01 -6.62101090e-01 1.14849043e+00
9.16447461e-01 -3.75847340e-01 8.88094679e-02 -3.71736676e-01
1.74565271e-01 7.29073212e-02 -1.03071165e+00 1.97623491e+00
-5.87275505e-01 2.63081819e-01 3.61711949e-01 -5.32629907e-01
1.28439605e+00 1.75644949e-01 8.50551665e-01 -5.54339290e-01
3.29860091e-01 2.30810374e-01 -5.96917212e-01 -1.75100312e-01
1.36884317e-01 1.28805608e-01 3.83984834e-01 5.01040697e-01
3.42340231e-01 -3.73286813e-01 -3.35899800e-01 -6.30171373e-02
7.19589531e-01 4.94941294e-01 2.33756110e-01 -3.58531266e-01
3.80207300e-01 -5.02670407e-01 5.60570002e-01 2.23688688e-02
2.19250899e-02 6.13527656e-01 5.36195993e-01 -7.14341760e-01
-9.36533809e-01 -9.65046823e-01 -2.62328953e-01 9.88131821e-01
1.29809706e-02 -7.86451161e-01 -6.31442249e-01 -6.47005320e-01
4.57258403e-01 -1.60062000e-01 -9.33196604e-01 2.19617411e-02
-6.58458591e-01 -5.22317410e-01 5.03423035e-01 3.92533690e-01
8.32045376e-02 -5.57195067e-01 1.78162575e-01 -3.86826009e-01
3.12216431e-01 -8.68786752e-01 -6.98904157e-01 -4.82581794e-01
-1.09581876e+00 -1.23239863e+00 -3.65623444e-01 -8.16559792e-01
1.17421675e+00 9.41057131e-02 1.12966526e+00 7.18666732e-01
-3.88991207e-01 5.62489212e-01 -2.02747975e-02 -2.09564656e-01
-1.45815656e-01 8.24752450e-02 1.64190695e-01 1.68368310e-01
1.70449302e-01 -1.13256514e+00 -6.80418134e-01 4.21811193e-01
-2.25144237e-01 2.67118901e-01 2.79344380e-01 1.48774743e-01
9.04644728e-01 -4.86608386e-01 3.26128423e-01 -8.56406510e-01
-3.96443531e-02 -5.60366035e-01 -6.02315366e-01 -3.58568691e-03
-4.95044887e-01 -8.86060894e-02 2.35883191e-01 -8.86585340e-02
-6.54600441e-01 2.41778418e-01 -4.76171017e-01 -7.63259828e-01
-2.25138903e-01 1.36194810e-01 -3.17831725e-01 -7.55268991e-01
3.70584935e-01 -2.04829603e-01 4.94374990e-01 -8.17861795e-01
6.01408899e-01 3.38625103e-01 4.90359068e-02 -7.81841338e-01
9.50528324e-01 7.43187129e-01 6.36499763e-01 -7.08118141e-01
-8.02771866e-01 -5.99636137e-02 -1.13925231e+00 -3.77286017e-01
4.30558294e-01 -9.68271136e-01 -9.24339712e-01 3.97929817e-01
-1.13524902e+00 -3.56909603e-01 -4.82222140e-02 7.07518458e-02
-6.00102067e-01 2.44934678e-01 -5.25870264e-01 -6.15448058e-01
-4.76460218e-01 -1.04674566e+00 1.43409157e+00 1.51247695e-01
-3.18067700e-01 -1.32824063e+00 -8.54471177e-02 3.00485432e-01
3.17600012e-01 4.49567497e-01 9.07401800e-01 -2.17178017e-01
-5.66638768e-01 -9.12772864e-02 -4.61967215e-02 3.77992727e-02
2.96646982e-01 3.22692245e-01 -9.11676168e-01 -1.28158465e-01
-5.23594260e-01 -1.46434575e-01 5.90582728e-01 4.27686363e-01
1.39145625e+00 -3.35213840e-01 -4.01470870e-01 1.46012092e+00
1.11582673e+00 -3.51538032e-01 2.66221434e-01 -2.20431656e-01
1.08693314e+00 6.83365464e-01 1.64360300e-01 6.12848163e-01
8.04637551e-01 6.74675763e-01 5.62311351e-01 -3.35817784e-01
-4.32172954e-01 -5.73052645e-01 1.34015754e-01 9.73863304e-01
-6.40619278e-01 5.40157974e-01 -9.49170291e-01 2.30417684e-01
-1.52373123e+00 -4.60487783e-01 1.25531688e-01 2.03408766e+00
8.88598919e-01 -4.10391033e-01 1.95848510e-01 -3.33109498e-01
4.82077628e-01 2.37926692e-01 -6.73349738e-01 -1.71711177e-01
1.13574132e-01 5.60868442e-01 2.94732481e-01 8.96993458e-01
-9.27051544e-01 1.25652599e+00 6.65443468e+00 7.78569162e-01
-9.37116027e-01 6.79408535e-02 7.98004985e-01 -4.72488821e-01
-5.81337690e-01 -1.00703657e-01 -1.13804054e+00 2.27743551e-01
6.42138124e-01 -1.23303309e-01 4.71702427e-01 7.05489516e-01
3.48508447e-01 4.34997201e-01 -1.18964624e+00 1.23950565e+00
1.80045277e-01 -1.64425886e+00 4.96479213e-01 4.18845296e-01
8.56593013e-01 -7.00481236e-02 2.14408010e-01 -1.52009547e-01
3.56971234e-01 -1.46980464e+00 4.36255008e-01 8.42682898e-01
1.24167001e+00 -7.95278728e-01 2.27051347e-01 -4.89125289e-02
-1.40072322e+00 2.58999050e-01 -3.54144007e-01 1.12839028e-01
5.77697195e-02 4.87518370e-01 -8.02936494e-01 4.90034163e-01
5.60075223e-01 1.26283848e+00 -5.73460758e-01 8.57501686e-01
-1.55201495e-01 2.96795964e-01 -5.44652879e-01 4.76010025e-01
-2.24986017e-01 -4.17026848e-01 2.91604757e-01 9.26170468e-01
4.14503276e-01 1.99263439e-01 -1.85190737e-02 8.74283135e-01
-5.48558414e-01 3.14803123e-01 -8.26509774e-01 3.35708559e-01
8.61267924e-01 1.41648638e+00 -4.67572331e-01 1.00613251e-01
-3.95727247e-01 4.83652681e-01 5.34592927e-01 1.72318995e-01
-6.05863810e-01 2.68646598e-01 1.27229512e+00 3.75255138e-01
5.08554094e-02 -4.79562223e-01 -4.60165858e-01 -9.39058363e-01
-5.35719767e-02 -5.30927777e-01 1.03075385e-01 -5.33260643e-01
-1.30351579e+00 4.14771765e-01 -1.47207245e-01 -9.50254261e-01
-4.64424267e-02 -6.04813755e-01 -4.97616470e-01 9.09998477e-01
-1.72719479e+00 -1.54517436e+00 -3.85744631e-01 7.57103682e-01
1.46742105e-01 -1.08561687e-01 9.44429100e-01 2.74150610e-01
-6.64006174e-01 9.40186560e-01 -3.93807441e-01 2.79641956e-01
5.87355375e-01 -6.81240261e-01 7.45869517e-01 3.13083142e-01
2.50170350e-01 6.20143592e-01 -1.59783185e-01 -7.41061449e-01
-1.79400694e+00 -1.33719742e+00 8.44435930e-01 -9.35363233e-01
2.94209361e-01 -5.97970545e-01 -6.18318141e-01 8.97622526e-01
-5.12014270e-01 2.61929601e-01 8.72871757e-01 3.66531700e-01
-6.11076951e-01 -1.35301530e-01 -1.07210946e+00 6.56601787e-01
1.88839328e+00 -5.64536393e-01 -4.50584963e-02 3.59182179e-01
4.65452015e-01 -6.17788792e-01 -1.36395025e+00 4.74962801e-01
8.98491800e-01 -7.48179138e-01 1.09919906e+00 -8.24528337e-01
4.19955075e-01 -4.21224535e-02 -1.69323131e-01 -1.06482422e+00
-2.35022351e-01 -9.43520010e-01 -3.67305577e-01 1.00792503e+00
5.13874054e-01 -5.27830482e-01 1.15787709e+00 8.09474289e-01
-1.67498276e-01 -1.33397055e+00 -1.06459725e+00 -3.00518602e-01
1.55979618e-01 -5.61518550e-01 1.15815723e+00 8.93557608e-01
-2.61837423e-01 -9.71810073e-02 -3.36707503e-01 1.45015001e-01
5.79505384e-01 2.26712182e-01 9.48672354e-01 -1.50737166e+00
4.57725883e-01 -5.72095513e-01 -5.86467624e-01 -1.28458560e+00
6.10654056e-01 -1.41866922e+00 -5.54219782e-01 -1.31115806e+00
8.41967165e-02 -7.97925115e-01 1.11674167e-01 8.78133774e-01
2.28545070e-01 6.11491501e-01 1.35086074e-01 1.45861655e-01
-5.40473223e-01 6.65209472e-01 1.60601366e+00 3.15896839e-01
-2.16036037e-01 -1.30781680e-01 -8.96181881e-01 9.75728810e-01
5.57167649e-01 -1.38731465e-01 -1.86771333e-01 -7.78371632e-01
1.96073905e-01 -1.16771208e-02 2.88076192e-01 -6.11081302e-01
1.73421144e-01 -2.37582296e-01 7.67610550e-01 -3.66027892e-01
7.36566246e-01 -7.13425279e-01 8.30483884e-02 -1.61431208e-01
1.67994704e-02 -3.93296257e-02 2.13962317e-01 1.46420613e-01
2.02659324e-01 3.92915189e-01 8.04744065e-01 -1.56100675e-01
-5.49629450e-01 1.26629519e+00 2.97348976e-01 -9.08067897e-02
7.98645377e-01 -4.71591443e-01 1.42725352e-02 -4.55444485e-01
-1.02933073e+00 3.36325407e-01 7.62317479e-01 4.76107031e-01
9.56741750e-01 -1.62114441e+00 -7.98491836e-01 7.54179418e-01
-3.51358980e-01 4.23291624e-01 2.03023091e-01 9.85447168e-01
-4.22466397e-01 1.60948619e-01 -2.95739949e-01 -5.35037279e-01
-1.26495540e+00 -7.10156560e-02 6.72576487e-01 4.23358858e-01
-5.29351711e-01 1.14801693e+00 7.65952189e-03 -7.00059652e-01
2.25908756e-01 -8.21169168e-02 -1.44288793e-01 1.06210165e-01
4.32331800e-01 4.92151082e-01 3.32331359e-02 -1.03654897e+00
-4.59151149e-01 1.40888834e+00 2.00950071e-01 2.60540545e-01
1.66954267e+00 -1.82740957e-01 -3.22956830e-01 -3.91424485e-02
1.44816184e+00 1.47290409e-01 -1.47741866e+00 -2.10570604e-01
-2.32851818e-01 -6.56687737e-01 -6.10174388e-02 -3.19497734e-01
-1.61484897e+00 7.62655199e-01 9.69750211e-02 -4.82913524e-01
1.11693978e+00 3.12204629e-01 7.95714319e-01 3.15840133e-02
5.94259858e-01 -6.88512981e-01 -3.35623473e-02 5.51936448e-01
9.58486378e-01 -1.09179270e+00 1.66556671e-01 -8.65545273e-01
-1.62686437e-01 1.14890635e+00 5.88034093e-01 -2.84367263e-01
1.38981283e+00 2.42077962e-01 4.37622517e-02 -5.23738742e-01
-6.87845290e-01 3.69263515e-02 5.93628526e-01 6.39740884e-01
4.44754869e-01 -6.91153333e-02 2.34101325e-01 5.60080290e-01
-5.38396299e-01 -7.51651973e-02 -7.05412328e-02 5.09941161e-01
-8.36082846e-02 -1.21811235e+00 -1.54101804e-01 4.92403567e-01
-1.22979097e-01 -4.52532135e-02 -4.46159333e-01 7.45978892e-01
1.29119411e-01 5.90462089e-01 2.04438120e-01 -6.70229733e-01
2.45713830e-01 1.81910783e-01 7.77048230e-01 -6.68916583e-01
-8.14650729e-02 -1.40202090e-01 -1.67303562e-01 -1.19923317e+00
-3.50077271e-01 -6.03060782e-01 -1.23416412e+00 -8.58957946e-01
7.35105798e-02 -1.39611319e-01 8.38691890e-01 7.24805593e-01
1.01839530e+00 -2.14635506e-01 9.85004604e-01 -1.40500069e+00
-1.04594149e-01 -6.74111068e-01 -6.39351785e-01 2.23329604e-01
3.34489018e-01 -9.36956942e-01 -3.52412492e-01 1.11443341e-01] | [13.26046085357666, 0.12581495940685272] |
a224e8be-b610-4026-9efd-3d19d7cb3890 | embedded-graph-theory | 1709.04710 | null | http://arxiv.org/abs/1709.04710v1 | http://arxiv.org/pdf/1709.04710v1.pdf | Embedded-Graph Theory | In this paper, we propose a new type of graph, denoted as "embedded-graph",
and its theory, which employs a distributed representation to describe the
relations on the graph edges. Embedded-graphs can express linguistic and
complicated relations, which cannot be expressed by the existing edge-graphs or
weighted-graphs. We introduce the mathematical definition of embedded-graph,
translation, edge distance, and graph similarity. We can transform an
embedded-graph into a weighted-graph and a weighted-graph into an edge-graph by
the translation method and by threshold calculation, respectively. The edge
distance of an embedded-graph is a distance based on the components of a target
vector, and it is calculated through cosine similarity with the target vector.
The graph similarity is obtained considering the relations with linguistic
complexity. In addition, we provide some examples and data structures for
embedded-graphs in this paper. | ['Atsushi Yokoyama'] | 2017-09-14 | null | null | null | null | ['graph-similarity'] | ['graphs'] | [-7.82457143e-02 2.44907275e-01 -1.21318094e-01 -4.08031225e-01
1.84376121e-01 -5.33567727e-01 4.59491611e-01 4.34996873e-01
-7.44302422e-02 2.19197959e-01 2.20226273e-01 -2.80790538e-01
-5.18324673e-01 -1.26681983e+00 7.57768983e-03 -3.42015356e-01
-4.54997063e-01 1.79120302e-01 5.20996451e-01 -6.52433872e-01
1.31958932e-01 2.84422010e-01 -1.05352736e+00 6.63885102e-02
8.24125648e-01 7.70316243e-01 1.39777064e-02 3.66033375e-01
-7.79834270e-01 7.70981431e-01 -5.62536657e-01 -6.46359265e-01
-4.97830994e-02 -7.14794338e-01 -5.44257283e-01 2.49796715e-02
-2.70248324e-01 3.50287437e-01 -4.41852987e-01 1.39195180e+00
2.53911465e-01 3.57974097e-02 6.52547777e-01 -1.70835447e+00
-1.04213774e+00 7.62852073e-01 -2.65774786e-01 -1.69177920e-01
9.13948596e-01 -5.82239628e-01 1.02339268e+00 -6.88587010e-01
7.06971645e-01 1.38477111e+00 7.11991906e-01 -3.44463289e-02
-7.79068828e-01 -2.58074850e-01 3.27233523e-01 5.10113597e-01
-1.68542850e+00 2.81461656e-01 9.09013629e-01 -3.95842224e-01
8.34982514e-01 6.62215829e-01 9.75227654e-01 2.12678835e-01
3.67340654e-01 1.04962759e-01 8.23722720e-01 -7.20296323e-01
3.23254708e-03 -1.32209122e-01 4.28653300e-01 1.02587342e+00
6.71573043e-01 -4.32752855e-02 -1.10736946e-02 -1.77067742e-01
2.88854986e-01 -2.33736243e-02 -4.57380891e-01 -2.29631901e-01
-1.03109634e+00 7.04987526e-01 5.02830386e-01 8.80534351e-01
2.06953228e-01 -1.47956833e-01 4.30467099e-01 5.89031816e-01
5.07239163e-01 -1.24317175e-02 1.40480682e-01 4.31426793e-01
-4.38646555e-01 -2.12479576e-01 9.86853600e-01 1.44703531e+00
8.82689416e-01 4.46280688e-02 1.52747296e-02 5.22380233e-01
6.02024972e-01 5.54847002e-01 4.64387476e-01 -3.39511096e-01
3.59800190e-01 1.42929554e+00 -4.54334199e-01 -1.88220620e+00
-5.42909145e-01 -4.31586653e-02 -9.81572151e-01 -1.72468796e-01
-2.38073543e-02 1.54019803e-01 -3.58196229e-01 1.56987882e+00
3.45310450e-01 2.21447721e-01 -8.32098275e-02 6.65962756e-01
1.30107296e+00 5.96851468e-01 -3.10605228e-01 -6.17237628e-01
1.30431640e+00 -8.59943390e-01 -1.25400984e+00 1.67978406e-01
8.77949893e-01 -7.15204179e-01 7.89436877e-01 3.06134056e-02
-8.36432457e-01 -3.92903030e-01 -1.19046140e+00 7.22556710e-02
-9.75787103e-01 -3.81596833e-01 2.38786921e-01 6.24647915e-01
-1.31624067e+00 4.24014807e-01 -1.99087352e-01 -6.10332847e-01
-4.28841859e-01 2.32585803e-01 -6.91929758e-01 1.29573175e-03
-1.39733648e+00 9.28659976e-01 5.53773761e-01 4.33017053e-02
2.66485304e-01 -8.20572674e-02 -1.29010725e+00 6.41819164e-02
4.27683651e-01 -4.46425110e-01 4.21162665e-01 -5.94435632e-01
-9.99777317e-01 7.37018943e-01 -5.59876673e-02 8.31551030e-02
-8.06713179e-02 6.23245895e-01 -1.31768906e+00 4.38391492e-02
-6.53099418e-02 -2.29175761e-01 4.60168540e-01 -1.05393100e+00
-4.48002994e-01 -3.92089248e-01 5.56191914e-02 3.59399840e-02
-4.64616656e-01 1.44178107e-01 -5.79173326e-01 -7.34650612e-01
3.29609782e-01 -5.62131763e-01 -9.58735943e-02 -3.50335598e-01
-4.73894596e-01 -3.81680846e-01 8.76743913e-01 -6.11856997e-01
2.14606380e+00 -2.21924424e+00 1.98851138e-01 8.17051053e-01
6.35563791e-01 3.04551095e-01 -3.59456152e-01 9.76336062e-01
-3.48546982e-01 4.47105199e-01 -3.01997215e-01 6.24764025e-01
1.66421369e-01 3.53007585e-01 1.44477606e-01 4.86090600e-01
-1.15258031e-01 9.41819847e-01 -1.25396216e+00 -8.98468435e-01
7.93868974e-02 3.64832878e-01 -2.72114370e-02 1.73403010e-01
1.86713174e-01 -3.72558177e-01 -4.44884688e-01 3.93542677e-01
6.49343550e-01 -3.08057349e-02 6.31199896e-01 -6.50830507e-01
-1.53102189e-01 -3.50940526e-01 -1.42606652e+00 1.28054702e+00
-2.29687646e-01 2.41725862e-01 2.43772529e-02 -1.06452811e+00
1.61653543e+00 1.88433066e-01 4.39677745e-01 -6.04382455e-01
3.44635040e-01 2.47631982e-01 8.45895186e-02 -7.63979793e-01
5.24879277e-01 8.97666663e-02 -1.71940267e-01 4.09090340e-01
-1.44942388e-01 -2.17644230e-01 6.71243489e-01 5.17067611e-01
1.08272231e+00 -1.82351153e-02 6.51926756e-01 -5.79742432e-01
9.75178123e-01 -2.31856316e-01 3.43571633e-01 -4.91850115e-02
2.11941347e-01 2.04794064e-01 6.76049829e-01 -3.28249425e-01
-5.28437734e-01 -1.08538508e+00 9.36900377e-02 5.57397962e-01
5.27584672e-01 -1.20595729e+00 -5.56385636e-01 -6.94392443e-01
2.99527291e-02 3.76341969e-01 -3.66278797e-01 -5.31061351e-01
-2.21877500e-01 -3.80599797e-01 2.91832656e-01 2.81181008e-01
3.91365439e-01 -5.68721056e-01 1.60832942e-01 2.22487867e-01
-9.75549594e-02 -8.48638594e-01 -1.07653689e+00 -3.70860040e-01
-5.34010410e-01 -1.41243875e+00 -3.50630403e-01 -1.35257006e+00
9.75608170e-01 2.45981947e-01 1.14542115e+00 5.41726232e-01
-7.93543532e-02 3.90260965e-01 -7.93460071e-01 -5.16000763e-02
-2.87497878e-01 -2.92465687e-01 6.14755787e-03 1.94679558e-01
4.12659079e-01 -5.68760395e-01 -2.02701673e-01 4.91853207e-01
-1.04580367e+00 -5.46187721e-02 2.14651182e-01 7.05937207e-01
5.94533980e-01 3.83185357e-01 4.36980277e-01 -7.99310088e-01
1.22871554e+00 -4.04757291e-01 -5.15876949e-01 8.00788403e-01
-9.31469142e-01 2.24725604e-01 6.61669910e-01 -3.58209103e-01
-5.04426897e-01 -4.01739269e-01 8.80409703e-02 -1.11338027e-01
6.15320444e-01 1.08891428e+00 -4.33761299e-01 -4.97734338e-01
3.16845208e-01 5.15181683e-02 -1.20474929e-02 -1.52642831e-01
8.39737177e-01 8.25007439e-01 2.43218541e-01 -4.55244064e-01
8.34386230e-01 1.31239206e-01 4.29440200e-01 -6.55871511e-01
-2.31781855e-01 -4.38539892e-01 -6.09567285e-01 -4.63051856e-01
7.15650499e-01 -3.77393663e-01 -7.55926371e-01 1.10748485e-01
-1.26177788e+00 1.63667142e-01 -1.91535100e-01 5.52600980e-01
-3.74079049e-02 7.49772310e-01 -5.41660249e-01 -6.03915989e-01
-3.13669354e-01 -7.66756177e-01 5.86073101e-01 -4.09656540e-02
-1.58562660e-01 -1.54505002e+00 1.33458018e-01 -4.70359981e-01
3.13610554e-01 5.63912868e-01 1.09183598e+00 -7.52456486e-01
9.85095501e-02 -4.43492532e-01 -3.10954332e-01 1.51083604e-01
4.67070311e-01 3.29815149e-01 1.87982582e-02 -1.29048154e-01
-1.88522384e-01 4.95956600e-01 2.93872297e-01 -1.49566174e-01
7.29597449e-01 -4.63323802e-01 -4.47456032e-01 5.79523504e-01
1.57985532e+00 5.29585004e-01 4.60420698e-01 -1.01402830e-02
6.84503138e-01 6.25070333e-01 5.91419697e-01 1.53690487e-01
5.85212052e-01 6.42807662e-01 1.11340187e-01 2.43628919e-02
2.73398776e-02 -5.08767247e-01 1.08162969e-01 2.07296300e+00
-2.78022349e-01 -3.54226738e-01 -7.96644509e-01 2.44527608e-01
-1.96386075e+00 -9.33505774e-01 -8.14137220e-01 2.03033829e+00
4.43801075e-01 2.61425576e-03 2.47844146e-03 2.78825045e-01
1.26659238e+00 1.75202146e-01 -5.03611416e-02 -8.94526303e-01
-2.83185214e-01 8.21660087e-02 2.89772809e-01 9.96496975e-01
-4.15401876e-01 6.91328406e-01 6.92878437e+00 8.20914924e-01
-7.92708218e-01 -9.83752906e-02 1.41404355e-02 6.29351318e-01
-8.65472555e-01 1.37247995e-01 -1.62300050e-01 4.76561606e-01
4.36923504e-01 -1.02482498e+00 5.49758136e-01 5.40583074e-01
-1.82490006e-01 3.10255140e-01 -8.38506877e-01 1.18608558e+00
3.53678435e-01 -8.60671461e-01 2.90982962e-01 -2.01105952e-01
3.81170899e-01 -6.60152793e-01 -3.69320750e-01 1.39430687e-01
2.23502383e-01 -8.03436935e-01 2.11649314e-01 7.23002851e-01
1.03204799e+00 -5.54693878e-01 9.12027955e-01 2.36331373e-02
-2.26973724e+00 3.31708729e-01 -3.52564692e-01 -1.94530338e-01
2.94554204e-01 6.58006907e-01 -1.21495016e-01 1.52787769e+00
3.74657393e-01 8.69462490e-01 -4.04275209e-01 8.00470173e-01
-3.15114737e-01 2.14472953e-02 -1.22349635e-01 -4.38713998e-01
5.98501600e-02 -1.19986558e+00 3.50938469e-01 1.17966568e+00
7.33203888e-01 3.24418068e-01 2.05091566e-01 7.14249909e-01
1.12880982e-01 6.39438629e-01 -9.84795570e-01 -2.27531582e-01
6.77472532e-01 1.35884917e+00 -6.03494585e-01 -4.77598578e-01
-7.13383079e-01 6.36918247e-01 1.33330464e-01 3.48742366e-01
-8.04636598e-01 -1.11425304e+00 7.49397948e-02 1.15451582e-01
-1.59088179e-01 -4.20865446e-01 2.32998341e-01 -9.90579188e-01
6.17227182e-02 -5.29887080e-01 6.72243059e-01 -1.08464074e+00
-1.49028349e+00 1.07583904e+00 3.66826475e-01 -1.35078609e+00
1.22042997e-02 -6.98059022e-01 -7.33258784e-01 8.60785782e-01
-9.72274959e-01 -8.11758816e-01 -5.64651549e-01 1.03258729e+00
-4.51794982e-01 -2.44942963e-01 9.40554917e-01 2.84757525e-01
-3.03379208e-01 4.20736045e-01 -1.35044113e-01 2.42684543e-01
3.59678030e-01 -1.16478300e+00 2.59591550e-01 6.68211162e-01
1.83545068e-01 5.51897287e-01 4.53739017e-01 -7.96129704e-01
-1.43969333e+00 -1.06132078e+00 1.42865896e+00 2.11667746e-01
1.04740119e+00 -2.49553442e-01 -7.12204933e-01 8.88531923e-01
4.01249498e-01 3.41601700e-01 6.91587031e-01 -7.29009062e-02
-4.49117571e-01 -2.59516627e-01 -1.25618267e+00 8.24932337e-01
1.66110706e+00 -5.90701461e-01 -8.31994414e-01 4.27819461e-01
1.15401065e+00 -1.87262684e-01 -1.32782924e+00 3.53207469e-01
4.05429393e-01 -8.52031708e-01 8.34353030e-01 -4.05508667e-01
-3.78547162e-01 -6.43075168e-01 -1.17689058e-01 -1.54311895e+00
-8.41334999e-01 -5.43919921e-01 1.59149155e-01 1.25048029e+00
1.79013208e-01 -1.10522711e+00 1.77110255e-01 2.06328668e-02
-9.56976488e-02 -7.82259464e-01 -6.60728335e-01 -1.21422899e+00
-4.34868932e-01 -2.17315227e-01 1.22009647e+00 1.17940414e+00
9.04533148e-01 5.25112391e-01 -3.85980383e-02 3.21163274e-02
4.31227267e-01 1.74478069e-01 1.97843760e-01 -1.19887662e+00
8.37243497e-02 -5.99406958e-01 -1.22759485e+00 -6.17405713e-01
2.53153086e-01 -1.33568847e+00 -5.75796187e-01 -1.98952925e+00
-3.04823518e-01 -2.40992248e-01 -1.80700079e-01 2.74542868e-01
1.84001341e-01 3.19099464e-02 1.17501326e-01 -7.48051405e-02
-4.75462884e-01 4.68680263e-01 1.38861656e+00 -3.20575058e-01
-1.60642534e-01 -4.64206517e-01 -4.05877024e-01 5.94695866e-01
7.67575681e-01 -2.26557955e-01 -6.66743636e-01 -2.13703364e-01
4.51186717e-01 1.07193291e-01 -2.50350893e-01 -6.71711922e-01
4.07043159e-01 -2.84990549e-01 -5.38790822e-01 -2.76950687e-01
-7.77064310e-03 -1.19707763e+00 6.25557184e-01 7.13953733e-01
-6.74350262e-02 6.48350894e-01 -3.42389107e-01 3.88073534e-01
-5.70599258e-01 -2.00382814e-01 1.61304802e-01 5.46820760e-02
-7.16620922e-01 4.88135308e-01 1.93040110e-02 1.28904790e-01
1.31259227e+00 -3.56605738e-01 -5.40878952e-01 -4.47062135e-01
-9.58410680e-01 2.44045168e-01 3.39549839e-01 2.56543428e-01
9.75202501e-01 -2.11971211e+00 -5.53883731e-01 1.87394813e-01
3.30031753e-01 -3.32028657e-01 -1.79604605e-01 7.80475616e-01
-8.16911757e-01 4.28394675e-02 -9.59296599e-02 -1.19361192e-01
-1.55129004e+00 9.81716990e-01 2.26594597e-01 -1.19447269e-01
-4.87415344e-01 2.22436577e-01 -9.61350575e-02 -4.49758679e-01
4.68448885e-02 -4.59231764e-01 -4.43362623e-01 1.38299642e-02
3.99193287e-01 4.76639271e-01 -6.53721541e-02 -1.04453242e+00
-5.95774710e-01 1.18467927e+00 6.45764828e-01 1.09960996e-01
8.29527080e-01 -4.93731350e-02 -9.30481136e-01 5.17182827e-01
1.34381652e+00 2.91383922e-01 1.28282741e-01 -2.89065003e-01
1.81030586e-01 -4.79522705e-01 -3.67206901e-01 -2.82534003e-01
-1.10890532e+00 4.48969662e-01 1.94520026e-01 1.00696194e+00
1.19955480e+00 1.33690890e-02 6.51357830e-01 2.33056888e-01
5.51824987e-01 -8.45396817e-01 -1.84496194e-01 4.26716030e-01
1.07254112e+00 -5.90668619e-01 6.45079389e-02 -1.01612830e+00
-3.45220178e-01 1.37955368e+00 4.29545492e-01 -1.73756182e-01
1.32114446e+00 3.87453914e-01 -1.32489130e-01 -3.10792178e-01
-3.85272354e-01 -5.14202416e-01 5.72057962e-01 8.38007390e-01
4.17664438e-01 2.74333745e-01 -1.12749577e+00 7.76008248e-01
-2.76703805e-01 -1.71325386e-01 2.30113104e-01 6.30301952e-01
-6.02886677e-01 -1.33626974e+00 -2.82725215e-01 2.86386877e-01
2.99933076e-01 -1.50817875e-02 -7.24189401e-01 9.53267515e-01
3.41984034e-01 1.35040712e+00 2.16899589e-01 -1.17807686e+00
6.95279241e-01 -2.22419143e-01 3.97012115e-01 -5.85758805e-01
-3.34471881e-01 -4.07684624e-01 2.25384951e-01 -5.52516639e-01
-5.98718941e-01 1.60960838e-01 -1.55629373e+00 -7.80614614e-01
-5.91005921e-01 6.08705401e-01 3.62723082e-01 7.02528715e-01
2.49519303e-01 5.72776556e-01 6.11751735e-01 -8.64959881e-02
-1.02314211e-01 -7.10881472e-01 -1.13621116e+00 7.73724616e-01
-3.55277151e-01 -4.38559771e-01 -4.97426540e-01 -2.05671832e-01] | [7.313633918762207, 5.829979419708252] |
0b979a77-5354-47a7-986f-755637c1f7ce | clips-stylometry-investigation-csi-corpus-a | null | null | https://aclanthology.org/L14-1001 | https://aclanthology.org/L14-1001.pdf | CLiPS Stylometry Investigation (CSI) corpus: A Dutch corpus for the detection of age, gender, personality, sentiment and deception in text | We present the CLiPS Stylometry Investigation (CSI) corpus, a new Dutch corpus containing reviews and essays written by university students. It is designed to serve multiple purposes: detection of age, gender, authorship, personality, sentiment, deception, topic and genre. Another major advantage is its planned yearly expansion with each year{'}s new students. The corpus currently contains about 305,000 tokens spread over 749 documents. The average review length is 128 tokens; the average essay length is 1126 tokens. The corpus will be made available on the CLiPS website (www.clips.uantwerpen.be/datasets) and can freely be used for academic research purposes. An initial deception detection experiment was performed on this data. Deception detection is the task of automatically classifying a text as being either truthful or deceptive, in our case by examining the writing style of the author. This task has never been investigated for Dutch before. We performed a supervised machine learning experiment using the SVM algorithm in a 10-fold cross-validation setup. The only features were the token unigrams present in the training data. Using this simple method, we reached a state-of-the-art F-score of 72.2{\%}. | ['Walter Daelemans', 'Ben Verhoeven'] | 2014-05-01 | null | null | null | lrec-2014-5 | ['deception-detection'] | ['miscellaneous'] | [-1.99571162e-01 2.05754399e-01 -2.19778836e-01 -2.51652330e-01
-6.01858318e-01 -7.15117812e-01 9.19761837e-01 4.32854265e-01
-5.90673268e-01 8.66035104e-01 1.21753216e-01 -3.39196652e-01
1.78552836e-01 -1.92401066e-01 -1.29722238e-01 -4.52649981e-01
3.97814631e-01 3.74154866e-01 -1.53913125e-01 -6.52405396e-02
1.15226972e+00 4.75621730e-01 -1.59666431e+00 5.75417466e-02
1.15894210e+00 8.08263898e-01 -1.58714086e-01 8.88753891e-01
4.28797156e-02 7.96645105e-01 -1.35543692e+00 -1.16767669e+00
-1.95540518e-01 -5.21160364e-01 -7.34642744e-01 4.83148843e-02
6.70431733e-01 -2.35785246e-01 -1.22692652e-01 1.01520133e+00
4.71599489e-01 3.60701531e-02 7.64855802e-01 -1.11382473e+00
-7.29528189e-01 4.97282773e-01 -3.98506731e-01 3.96014780e-01
7.82215178e-01 -2.84118891e-01 1.13741696e+00 -9.22491550e-01
7.17948616e-01 1.05616212e+00 2.26751044e-01 6.74533665e-01
-8.28798890e-01 -8.31118703e-01 -2.70678133e-01 3.63369375e-01
-9.84843850e-01 -7.17267454e-01 7.12924302e-01 -5.30515015e-01
6.09410107e-01 2.96561867e-01 6.55204475e-01 1.51996613e+00
3.10415477e-01 8.94077420e-01 1.35021722e+00 -7.46390283e-01
2.80979007e-01 9.10292208e-01 4.16607618e-01 6.37513936e-01
4.10502493e-01 -4.46832269e-01 -9.69381273e-01 -1.78720936e-01
2.37754554e-01 -6.32617891e-01 -1.99553021e-03 2.28499159e-01
-9.52615857e-01 1.07554817e+00 -4.68644530e-01 5.33639610e-01
1.38571531e-01 -4.15753663e-01 8.61280978e-01 6.59015477e-01
7.61091292e-01 6.68742597e-01 -3.61403704e-01 -7.75131941e-01
-1.10067880e+00 3.14112812e-01 1.30634856e+00 6.09235883e-01
2.73749623e-02 4.92091030e-02 1.72215238e-01 1.08966160e+00
1.17829740e-01 4.37232882e-01 9.57411468e-01 -7.80362487e-01
4.53466088e-01 5.80063343e-01 8.54237229e-02 -1.07715368e+00
-1.69669479e-01 -9.05033946e-02 -3.68775308e-01 9.68942568e-02
7.67961323e-01 -2.70528406e-01 -3.01299930e-01 1.14904022e+00
3.21361087e-02 -6.07313693e-01 -5.99696338e-02 6.06512368e-01
1.05468011e+00 3.57342184e-01 -1.35987371e-01 -3.70680630e-01
1.41129553e+00 -6.58121109e-01 -8.65795732e-01 5.91955334e-02
6.77134454e-01 -1.06950271e+00 8.08088958e-01 9.84156013e-01
-1.23132598e+00 -3.23318690e-01 -1.03545725e+00 -5.21642528e-02
-6.33607864e-01 5.74429393e-01 4.75380868e-01 1.41988134e+00
-7.07826257e-01 6.13196731e-01 -3.96927357e-01 -4.06660169e-01
3.49489808e-01 4.28525284e-02 -4.65377659e-01 2.66465217e-01
-1.14667916e+00 1.09772277e+00 -1.62084326e-01 -1.01498000e-01
-2.42380843e-01 -3.86038363e-01 -9.82166767e-01 -1.90754041e-01
1.20966714e-02 2.49060258e-01 1.07544148e+00 -1.14269280e+00
-1.75188434e+00 1.53851259e+00 -2.63718784e-01 -1.56565577e-01
7.70607352e-01 -1.39789015e-01 -7.76784360e-01 1.48514420e-01
3.09665769e-01 2.34051049e-02 1.02429080e+00 -6.95144951e-01
-4.12053555e-01 -6.23449922e-01 -3.82073015e-01 -6.65108040e-02
-5.33950865e-01 5.92729986e-01 1.02870665e-01 -8.31809878e-01
-2.15989903e-01 -8.21000874e-01 5.92167497e-01 -4.50439453e-01
-4.29676771e-01 -5.62838912e-01 4.74683106e-01 -1.11643469e+00
1.42689645e+00 -2.12198949e+00 -2.89585646e-02 1.04513951e-01
1.96494862e-01 4.31455433e-01 3.17831248e-01 5.97600460e-01
8.76187757e-02 1.60010755e-01 2.44744606e-02 -5.24109483e-01
3.62281591e-01 -2.09571958e-01 4.15999666e-02 7.96950221e-01
-6.69343844e-02 5.24331629e-01 -7.15081692e-01 -5.75830698e-01
-1.84719443e-01 1.26396239e-01 1.00856207e-01 -4.11824435e-02
4.45741594e-01 -9.67791583e-03 -1.55594647e-01 9.08531487e-01
5.56572497e-01 4.38843697e-01 -1.02624655e-01 6.00289404e-01
-3.78689706e-01 4.21287030e-01 -9.66557264e-01 1.14345491e+00
-3.92710388e-01 1.32384753e+00 1.12912573e-01 -7.61495650e-01
1.45791686e+00 2.02074841e-01 -1.01853691e-01 -4.18811828e-01
3.31229866e-01 5.60873926e-01 2.33588725e-01 -5.55916011e-01
9.68358517e-01 -8.19776580e-02 -4.18716997e-01 5.16423821e-01
-1.36908721e-02 -1.88191608e-01 6.22438014e-01 3.90370935e-01
8.46457303e-01 -2.67635167e-01 2.89681345e-01 -4.48392093e-01
8.79900515e-01 -1.59969285e-01 3.18309605e-01 4.35884774e-01
-6.41888976e-01 4.21174556e-01 1.32456172e+00 -1.52084082e-01
-9.12979782e-01 -6.29468024e-01 -5.34273922e-01 1.02769089e+00
-3.76736760e-01 -4.46348667e-01 -8.60732198e-01 -7.42935717e-01
1.09509893e-01 9.37258363e-01 -7.74944067e-01 -8.01189095e-02
-2.35887751e-01 -4.44730401e-01 6.39758766e-01 2.82322932e-02
3.20303142e-01 -1.31040120e+00 -5.47339797e-01 -2.27184787e-01
4.21937592e-02 -8.70597184e-01 -4.16855633e-01 4.36099991e-02
-4.88677561e-01 -1.09323215e+00 -7.69682527e-01 -7.75236964e-01
5.18284321e-01 -3.50103557e-01 9.38593507e-01 9.18347090e-02
-3.55498523e-01 2.50729620e-01 -5.91405094e-01 -8.36141527e-01
-6.67072415e-01 4.00054246e-01 2.37679392e-01 -1.52075455e-01
6.91944003e-01 -1.33519366e-01 -1.94195867e-01 1.07917756e-01
-4.78645563e-01 -5.65840542e-01 2.78467566e-01 9.43221927e-01
-3.23966235e-01 -2.95804560e-01 9.23251867e-01 -1.01984513e+00
9.97794449e-01 -2.12032750e-01 -2.89003134e-01 -5.49424477e-02
-6.70704544e-01 -4.62389171e-01 7.15111375e-01 -3.87905538e-01
-1.04459298e+00 -4.02332634e-01 -9.69055742e-02 2.08621040e-01
-4.48002070e-01 3.44967186e-01 -4.89588715e-02 6.89360425e-02
4.83088493e-01 1.83961987e-01 2.49783218e-01 -3.63155097e-01
-3.24215770e-01 1.30601394e+00 3.79095584e-01 -3.12739521e-01
5.06637275e-01 -1.90509364e-01 -2.24268764e-01 -1.26051915e+00
-8.30741346e-01 -6.38378441e-01 -1.01079726e+00 -3.92658502e-01
3.85870427e-01 -6.22478485e-01 -1.02162993e+00 8.62657726e-01
-1.01803589e+00 1.66794229e-02 1.22113608e-01 3.27598900e-01
-1.25913516e-01 5.50747871e-01 -6.30681694e-01 -1.03553319e+00
-4.36722100e-01 -6.44062936e-01 6.49014950e-01 2.06830278e-01
-8.40456724e-01 -1.20067823e+00 5.56886755e-02 8.13666046e-01
-6.12700805e-02 1.91896170e-01 7.49715745e-01 -1.14880180e+00
6.58033848e-01 -6.97359085e-01 4.54933606e-02 6.76323652e-01
-2.37661943e-01 4.12497789e-01 -8.76111209e-01 -1.18537694e-01
5.29526584e-02 -6.03031278e-01 7.17969596e-01 -1.64040439e-02
9.49704885e-01 -3.25757504e-01 2.08789825e-01 -1.29962757e-01
9.22346890e-01 -9.70457308e-03 5.51372290e-01 4.55984056e-01
4.99018103e-01 8.22044969e-01 5.47604978e-01 8.11320484e-01
2.72626162e-01 3.99915934e-01 1.05640166e-01 6.34669244e-01
3.75423282e-01 2.48969942e-02 7.97147393e-01 8.49937558e-01
-5.17618284e-03 -1.34608313e-01 -9.04913962e-01 7.08679259e-01
-1.39933217e+00 -1.13758194e+00 -6.20110154e-01 2.19041300e+00
8.22031319e-01 3.14522296e-01 4.75083470e-01 6.17018998e-01
6.80063903e-01 1.52266026e-01 -2.37483963e-01 -1.54759133e+00
-4.00357693e-03 2.80782580e-01 2.62481660e-01 3.59644681e-01
-8.03521037e-01 7.85829663e-01 5.61490679e+00 9.20699835e-01
-1.05619514e+00 -1.83853447e-01 6.97683752e-01 -3.14298749e-01
1.26143068e-01 -5.91784298e-01 -9.95360792e-01 7.96106815e-01
1.25799465e+00 -4.08360273e-01 -1.03872446e-02 7.38536894e-01
2.79270500e-01 -8.08420300e-01 -8.80817950e-01 9.05620396e-01
7.35748827e-01 -7.99852788e-01 -3.31275970e-01 1.84526280e-01
6.25977993e-01 -4.79366004e-01 1.61196962e-01 4.16371822e-01
-2.48680055e-01 -9.84649658e-01 9.94700611e-01 2.89341599e-01
8.14668655e-01 -1.02206385e+00 9.87807214e-01 4.93182451e-01
-2.62018859e-01 -4.12634537e-02 -2.96750069e-01 -4.60550636e-01
-2.24417627e-01 7.61579633e-01 -6.73290670e-01 1.68183774e-01
5.77132404e-01 8.48377228e-01 -8.85971785e-01 6.56142116e-01
-3.94209415e-01 8.36309195e-01 1.88948706e-01 -7.88523197e-01
8.43832269e-02 -3.02838862e-01 5.99017322e-01 1.32286382e+00
1.37745410e-01 -2.21040189e-01 -3.88980776e-01 4.60823923e-01
-1.93089366e-01 3.84528369e-01 -2.71416366e-01 -1.56378150e-01
5.48174083e-01 1.52196896e+00 -7.86633790e-01 -2.29426444e-01
-1.43488765e-01 1.24219465e+00 2.38731116e-01 -1.16763830e-01
-5.06525278e-01 -1.15131080e+00 3.70946705e-01 1.32363096e-01
2.28513449e-01 -4.01338702e-03 -7.04757631e-01 -1.03092551e+00
1.63550034e-01 -9.18754995e-01 1.05607435e-01 -3.54933172e-01
-1.41778159e+00 2.74850696e-01 -3.10109079e-01 -7.72510231e-01
-2.16125831e-01 -1.01249421e+00 -8.48561883e-01 9.90785003e-01
-9.43436444e-01 -7.65942276e-01 -1.02520771e-01 1.90699711e-01
5.75336039e-01 -5.97936034e-01 8.68561566e-01 2.15018224e-02
-9.72799420e-01 9.96639609e-01 4.07766849e-01 2.56728172e-01
1.12563598e+00 -1.69452810e+00 6.62895665e-02 4.21691000e-01
-1.76240474e-01 4.33894128e-01 8.12597930e-01 -5.11644900e-01
-9.24110889e-01 -3.88404638e-01 1.77657306e+00 -7.38766789e-01
1.07044339e+00 -6.10589921e-01 -7.17704415e-01 1.74131721e-01
2.09913582e-01 -4.86823827e-01 1.08719945e+00 2.39936799e-01
-3.11956555e-01 1.63260400e-01 -1.32950103e+00 3.18627745e-01
4.56674725e-01 -4.80502129e-01 -6.47710085e-01 2.83027828e-01
-1.60462767e-01 -4.69572514e-01 -1.05873549e+00 -4.51486737e-01
1.00252557e+00 -1.24464762e+00 2.77877510e-01 -3.67962420e-01
1.25302780e+00 5.23189604e-01 4.23620015e-01 -1.36040878e+00
-4.96367663e-02 -6.75548553e-01 -1.02747507e-01 1.57342458e+00
4.98491943e-01 -6.89419985e-01 8.61045659e-01 7.31214106e-01
-1.16448915e-02 -8.82452011e-01 -9.84498620e-01 -8.23240995e-01
5.02495050e-01 -9.09899920e-02 1.22368731e-01 1.04175055e+00
8.08747113e-01 3.30411434e-01 -1.43330380e-01 -6.13348901e-01
3.64484340e-01 -1.34070456e-01 6.13884091e-01 -1.42550576e+00
1.02586828e-01 -8.85830998e-01 -4.22458708e-01 -2.46471420e-01
6.52606308e-01 -6.69688523e-01 -2.91054100e-01 -1.19790351e+00
7.03407452e-02 2.15692483e-02 3.18128645e-01 1.56969294e-01
-1.09784603e-01 3.07783037e-01 1.21849649e-01 8.74021500e-02
-3.99178237e-01 3.18717450e-01 1.18215108e+00 1.16230771e-01
-2.23042350e-02 2.89946198e-01 -7.91828454e-01 5.53487301e-01
1.09834826e+00 -2.26666376e-01 9.66830105e-02 3.04937631e-01
1.72175840e-01 -1.43741816e-01 1.37211815e-01 -6.26240313e-01
1.89835802e-02 5.65391928e-02 5.92128158e-01 -4.65261638e-01
3.31859440e-01 -2.77571321e-01 -5.63463688e-01 4.12484497e-01
-4.12866205e-01 1.60926938e-01 7.67494589e-02 1.26879051e-01
-2.58161128e-01 -9.77086186e-01 5.75128376e-01 -1.64639458e-01
-3.78521010e-02 -3.46050739e-01 -9.76160824e-01 4.92315106e-02
1.00909615e+00 -4.99575019e-01 -4.23355877e-01 -4.71913189e-01
-4.93172735e-01 -8.24123248e-02 5.92327178e-01 3.77286702e-01
3.24926883e-01 -9.33016360e-01 -9.35182691e-01 2.51778923e-02
2.03798458e-01 -7.84479797e-01 -1.77654438e-02 9.06143129e-01
-5.57570159e-01 3.82514983e-01 -3.68606627e-01 9.44676176e-02
-1.77202213e+00 1.20370604e-01 -1.47588462e-01 -1.40455484e-01
-2.10572585e-01 7.84490287e-01 -6.71146393e-01 -1.66853756e-01
1.71406105e-01 3.18643659e-01 -6.73523903e-01 7.82841444e-01
6.65544271e-01 1.05589712e+00 2.56095737e-01 -9.52390194e-01
-2.68441230e-01 2.25438569e-02 -3.63116086e-01 -3.38252485e-01
1.15033960e+00 3.08955275e-03 -4.57077026e-01 8.72474730e-01
1.33174789e+00 5.21752834e-01 -4.00006771e-01 3.28765810e-01
7.69904554e-02 -6.43654525e-01 -6.93865791e-02 -8.97178590e-01
-6.53360307e-01 7.31817842e-01 -3.91505379e-03 7.05068648e-01
4.51420873e-01 -3.24239224e-01 6.01356685e-01 1.65116444e-01
-1.06283866e-01 -1.66460371e+00 1.30796701e-01 7.52199769e-01
1.00069857e+00 -1.45859575e+00 2.59084105e-01 -2.47794479e-01
-8.83701265e-01 1.34015191e+00 5.36248922e-01 -2.68154256e-02
4.87912863e-01 -1.52157202e-01 1.07195519e-01 -1.40524149e-01
-5.89521229e-01 2.83559740e-01 3.76241177e-01 5.47939122e-01
8.14452648e-01 2.03185409e-01 -1.05197942e+00 9.31769848e-01
-8.48394871e-01 -4.50537294e-01 1.29258716e+00 6.86418116e-01
-3.44874650e-01 -1.13123262e+00 -5.68960488e-01 8.16427350e-01
-1.01136065e+00 1.25668228e-01 -1.10423875e+00 7.33681977e-01
-1.18374802e-01 1.05194414e+00 7.69518986e-02 -2.45181814e-01
1.72570169e-01 3.37549895e-01 4.56971169e-01 -5.70493102e-01
-9.80820835e-01 -4.02541757e-01 5.46797335e-01 7.86946043e-02
-1.31985098e-01 -1.47423065e+00 -7.29020715e-01 -9.29154336e-01
-3.12745154e-01 3.88150573e-01 1.01583016e+00 8.61621439e-01
-1.49275541e-01 1.27009943e-01 7.14224517e-01 -5.82495153e-01
-6.00865424e-01 -1.26741707e+00 -1.09944212e+00 2.11307585e-01
1.11470252e-01 -4.51385319e-01 -8.09227943e-01 7.72321969e-02] | [8.342473030090332, 10.437302589416504] |
1ae7e657-fa64-43a4-8661-63dcb9eb67e1 | automatic-3d-registration-of-dental-cbct-and | 2305.10132 | null | https://arxiv.org/abs/2305.10132v2 | https://arxiv.org/pdf/2305.10132v2.pdf | Automatic 3D Registration of Dental CBCT and Face Scan Data using 2D Projection images | This paper presents a fully automatic registration method of dental cone-beam computed tomography (CBCT) and face scan data. It can be used for a digital platform of 3D jaw-teeth-face models in a variety of applications, including 3D digital treatment planning and orthognathic surgery. Difficulties in accurately merging facial scans and CBCT images are due to the different image acquisition methods and limited area of correspondence between the two facial surfaces. In addition, it is difficult to use machine learning techniques because they use face-related 3D medical data with radiation exposure, which are difficult to obtain for training. The proposed method addresses these problems by reusing an existing machine-learning-based 2D landmark detection algorithm in an open-source library and developing a novel mathematical algorithm that identifies paired 3D landmarks from knowledge of the corresponding 2D landmarks. A main contribution of this study is that the proposed method does not require annotated training data of facial landmarks because it uses a pre-trained facial landmark detection algorithm that is known to be robust and generalized to various 2D face image models. Note that this reduces a 3D landmark detection problem to a 2D problem of identifying the corresponding landmarks on two 2D projection images generated from two different projection angles. Here, the 3D landmarks for registration were selected from the sub-surfaces with the least geometric change under the CBCT and face scan environments. For the final fine-tuning of the registration, the Iterative Closest Point method was applied, which utilizes geometrical information around the 3D landmarks. The experimental results show that the proposed method achieved an averaged surface distance error of 0.74 mm for three pairs of CBCT and face scan datasets. | ['Kiwan Jeon', 'Jin Keun Seo', 'Sang-Hwy Lee', 'Chang Min Hyun', 'Hyoung Suk Park'] | 2023-05-17 | null | null | null | null | ['facial-landmark-detection'] | ['computer-vision'] | [ 1.15686111e-01 2.86843896e-01 5.30360034e-04 -4.01517600e-01
-9.21241522e-01 3.66327651e-02 2.68752158e-01 1.33190006e-01
-4.53483939e-01 2.77757049e-02 -1.60985604e-01 -1.83866292e-01
-3.47893685e-01 -8.25236678e-01 -2.92751193e-01 -6.10098183e-01
5.60281500e-02 1.08169806e+00 3.90570283e-01 -1.08763248e-01
4.38043445e-01 1.16696048e+00 -1.79983032e+00 -2.84718871e-01
3.67428899e-01 9.02907431e-01 2.77877539e-01 8.34851712e-02
-3.77884895e-01 -2.71238834e-01 -1.50754556e-01 -2.84801424e-01
4.84339505e-01 -3.31993133e-01 -4.65703040e-01 2.27273121e-01
5.09913921e-01 -4.02151942e-01 4.42842782e-01 7.68530011e-01
7.07338810e-01 -5.90902865e-02 7.93059409e-01 -9.82006192e-01
-5.83851039e-02 -2.13714555e-01 -9.34534729e-01 -1.34599343e-01
4.87096697e-01 -3.48355561e-01 3.22092772e-01 -1.02794003e+00
6.22525811e-01 1.23843634e+00 7.74290740e-01 7.84737825e-01
-9.01634693e-01 -7.64161408e-01 -4.75615144e-01 5.41372923e-03
-1.57504058e+00 -3.98701519e-01 8.07088077e-01 -6.25036657e-01
6.53270602e-01 3.07346731e-01 7.20405400e-01 8.42811167e-02
1.96166083e-01 -1.47242635e-01 1.43896985e+00 -1.05455697e+00
2.80962348e-01 1.88050404e-01 -2.27006264e-02 9.95336235e-01
1.36645585e-01 2.59643972e-01 -2.05994435e-02 -3.39103758e-01
1.01699138e+00 4.49419990e-02 -2.03203693e-01 -4.00034040e-01
-3.34446251e-01 8.03314865e-01 3.44185054e-01 6.02225423e-01
-3.57004702e-01 -3.60468715e-01 3.57613154e-02 -2.54222244e-01
4.58360523e-01 -3.36588264e-01 -1.74265072e-01 2.71334469e-01
-8.35572660e-01 -3.08780730e-01 5.11666000e-01 7.82111168e-01
7.07822740e-01 -2.93907076e-01 4.95160699e-01 1.01182878e+00
9.33019161e-01 5.61385393e-01 7.05058575e-01 -4.87852454e-01
7.80324861e-02 7.06285954e-01 -4.98115271e-01 -8.57277215e-01
-5.34907877e-01 2.62046307e-01 -3.73783261e-01 5.05872011e-01
4.19395775e-01 3.55644375e-01 -1.19147682e+00 1.16135442e+00
9.97743130e-01 7.20141679e-02 -2.15507865e-01 7.47084200e-01
9.47702527e-01 1.53542176e-01 -1.23748824e-01 -4.74334955e-01
1.64852130e+00 -4.09039438e-01 -5.60174465e-01 -6.47795126e-02
4.47062016e-01 -1.16595745e+00 9.30255592e-01 -2.19354946e-02
-1.01608038e+00 -1.93809241e-01 -8.87057960e-01 -1.01147965e-01
-1.73367918e-01 1.78603474e-02 2.16924950e-01 9.90105808e-01
-9.17374313e-01 4.49184805e-01 -9.92016137e-01 -5.49645603e-01
4.32093680e-01 7.54696190e-01 -6.40243471e-01 -1.36763126e-01
-4.04868305e-01 1.21500742e+00 -6.43038526e-02 2.56896704e-01
-1.83756217e-01 -5.82626760e-01 -8.23335528e-01 -2.64523476e-01
3.49907912e-02 -4.39549476e-01 1.12508905e+00 -4.25771892e-01
-1.95404375e+00 1.54441106e+00 -2.57442355e-01 3.58312070e-01
5.36171734e-01 1.42369062e-01 -3.12047839e-01 2.05642074e-01
1.96568727e-01 1.50739223e-01 4.90405649e-01 -1.17090082e+00
-2.55241811e-01 -1.25022471e+00 -5.04513144e-01 2.61444092e-01
3.34333032e-02 2.24671379e-01 -8.32123458e-01 -9.68831107e-02
9.82832134e-01 -8.53994310e-01 -3.13276201e-01 4.14086163e-01
-2.07154691e-01 -1.86994359e-01 7.34533906e-01 -6.62068903e-01
5.91002345e-01 -2.06786704e+00 -4.47202951e-01 8.02328706e-01
-2.42360085e-01 -2.60237401e-04 2.54318744e-01 2.41377369e-01
-2.08101109e-01 -7.81120807e-02 -2.64710695e-01 -4.65717643e-01
-4.13394958e-01 1.96318761e-01 4.52102393e-01 9.09035027e-01
-5.60416937e-01 2.21041486e-01 -4.83004332e-01 -9.34226513e-01
3.99238139e-01 8.69355679e-01 -2.92471737e-01 -1.42682912e-02
2.82086819e-01 5.35644352e-01 -5.51130772e-01 6.72247767e-01
1.09852302e+00 2.96787351e-01 1.68387383e-01 -2.34182760e-01
-1.16917863e-01 3.98283862e-02 -1.31961453e+00 1.64756191e+00
-6.94592059e-01 2.34282501e-02 2.36354202e-01 -5.96118867e-01
1.23962331e+00 6.11019313e-01 9.44087625e-01 -7.62275279e-01
4.18943226e-01 6.67065501e-01 -2.67719090e-01 -4.57619786e-01
-1.41606733e-01 -7.99130142e-01 4.95363951e-01 5.77557266e-01
-1.51056454e-01 -8.47422421e-01 -4.20298934e-01 -5.68416953e-01
2.24743992e-01 1.09210899e-02 5.40950418e-01 -3.25134307e-01
8.55886161e-01 -2.37729684e-01 4.47659999e-01 -1.18416339e-01
7.89302588e-02 7.54341781e-01 -1.62161797e-01 -3.63046020e-01
-6.66789293e-01 -7.74983346e-01 -8.80699217e-01 2.12293833e-01
-2.25358214e-02 -5.05737625e-02 -8.62564266e-01 -4.04569805e-01
-2.57294923e-01 4.49718773e-01 -3.77082884e-01 3.04525286e-01
-8.35409582e-01 -7.42686272e-01 1.16425350e-01 1.16730444e-01
1.43927753e-01 -5.61028540e-01 -5.41593254e-01 -7.61279836e-02
3.27038944e-01 -8.27258050e-01 -4.44091529e-01 -2.05988884e-01
-1.44305694e+00 -1.57271814e+00 -6.32187724e-01 -1.09059763e+00
1.23488688e+00 2.55956920e-03 5.80017567e-01 4.12793010e-01
-4.41904068e-01 5.65960765e-01 3.66992317e-02 -5.75041711e-01
-5.86316824e-01 -4.54280347e-01 2.44693495e-02 -1.51074857e-01
5.73396564e-01 -5.77167332e-01 -4.46329772e-01 7.36185551e-01
-7.40898907e-01 -3.16473812e-01 2.98679322e-01 5.99217772e-01
1.05807710e+00 1.07393950e-01 1.95961744e-01 -7.56852806e-01
3.96294951e-01 -9.10895616e-02 -9.35754001e-01 2.68239528e-01
-7.73067117e-01 -1.59904540e-01 9.90342796e-02 -3.44471559e-02
-1.12479579e+00 4.74739760e-01 -5.89578569e-01 -7.38847926e-02
-3.59685808e-01 1.46736428e-01 -2.67946780e-01 -3.53521585e-01
6.12919748e-01 -3.09563670e-02 5.86578012e-01 -7.19357669e-01
9.46758166e-02 8.46200228e-01 3.31297934e-01 -4.92376685e-01
7.59481966e-01 8.15384269e-01 5.35108447e-01 -9.07324791e-01
-2.24361986e-01 -5.98571658e-01 -1.14849234e+00 -2.12365612e-01
8.32114756e-01 -5.47310114e-01 -7.32895732e-01 3.47798407e-01
-1.03460109e+00 2.19499141e-01 -1.20063759e-01 8.99352014e-01
-5.79824984e-01 6.01652801e-01 -3.18012595e-01 -6.52733862e-01
-6.20940566e-01 -1.51518154e+00 1.17906702e+00 2.76337743e-01
2.52036452e-02 -1.07124722e+00 3.73530090e-01 4.17922825e-01
7.49021694e-02 2.55079925e-01 1.00353801e+00 -4.31113929e-01
-1.81853846e-01 -5.52330554e-01 3.25267673e-01 8.31675231e-02
4.96001124e-01 6.70793727e-02 -1.05436778e+00 -7.60900453e-02
5.55265009e-01 8.65956247e-02 9.91239473e-02 5.22086143e-01
8.50429893e-01 1.26959339e-01 -6.31800652e-01 5.68914533e-01
1.57116139e+00 6.08891189e-01 6.52876616e-01 4.48882908e-01
3.60667855e-01 8.59916389e-01 5.18562078e-01 1.68329269e-01
2.85256147e-01 1.16520274e+00 4.48650569e-01 -1.63622230e-01
-2.08663493e-01 -1.93987787e-01 -9.04531106e-02 7.41961598e-01
-4.37933058e-01 6.98259234e-01 -1.05737484e+00 1.56314939e-01
-1.04557192e+00 -6.52918875e-01 -3.28660369e-01 2.62930202e+00
6.96433842e-01 -3.52278262e-01 -6.56818971e-02 3.55339944e-01
9.00143385e-01 -5.98805130e-01 -2.30273589e-01 -3.62282306e-01
4.56418425e-01 8.94341648e-01 4.01051551e-01 8.40978205e-01
-6.73178971e-01 4.47871566e-01 5.78724241e+00 6.98800266e-01
-1.29927933e+00 2.97257632e-01 3.42180520e-01 1.78059101e-01
-1.23073578e-01 -7.85807222e-02 -9.19975519e-01 1.14785731e-01
6.61124229e-01 7.70817846e-02 -1.92614406e-01 9.34776366e-01
3.10522646e-01 -3.68194222e-01 -1.03804147e+00 1.06407630e+00
1.78814143e-01 -9.44088101e-01 -1.67509615e-01 3.66508573e-01
2.97372639e-01 -2.33152092e-01 -4.88650315e-02 -3.53393376e-01
-3.15019310e-01 -1.03798032e+00 4.17304397e-01 1.27570361e-01
1.05733597e+00 -4.31079656e-01 5.52500844e-01 3.66353989e-01
-1.07962310e+00 3.92413884e-01 -3.28306258e-01 3.32902312e-01
3.18987161e-01 4.53675121e-01 -1.42977285e+00 5.56527019e-01
4.85058218e-01 -3.52913626e-02 -1.94641516e-01 1.26039147e+00
-1.37899891e-01 1.33817066e-02 -7.92151093e-01 3.62806022e-01
-2.10855737e-01 -4.75694001e-01 2.43282974e-01 5.30496895e-01
7.68737614e-01 3.57325912e-01 -1.90717712e-01 4.76214260e-01
2.63328791e-01 7.88891137e-01 -6.24602258e-01 8.58242750e-01
5.70717812e-01 1.15602207e+00 -1.20220137e+00 2.08605811e-01
-5.29937625e-01 5.18438280e-01 -2.21373022e-01 -3.52819532e-01
-4.37093973e-01 1.61099017e-01 1.41021237e-01 6.63406551e-01
-6.68484718e-02 -3.87462415e-02 -2.29465276e-01 -6.01076305e-01
4.81337830e-02 -3.28081012e-01 4.64726001e-01 -5.13332367e-01
-1.02936518e+00 8.79656911e-01 3.75348032e-01 -1.22117162e+00
-4.23452258e-01 -7.41094470e-01 -7.12628782e-01 9.98415768e-01
-1.32127500e+00 -1.21989918e+00 -2.37892777e-01 7.84574091e-01
2.27911055e-01 8.05946663e-02 1.14554703e+00 4.11450952e-01
-3.58629584e-01 5.66028178e-01 1.70730248e-01 3.61303128e-02
4.21805710e-01 -7.88974941e-01 -1.62250802e-01 3.07083845e-01
1.37347763e-03 4.77534503e-01 1.74483418e-01 -7.05566704e-01
-1.15966129e+00 -5.17912567e-01 8.92763853e-01 -2.72254378e-01
1.43450564e-02 -9.69444513e-02 -4.81264502e-01 7.22568750e-01
-3.17611635e-01 2.60526717e-01 8.96571040e-01 -1.90896884e-01
5.36467545e-02 -1.16800204e-01 -1.80453467e+00 2.01472804e-01
6.01901591e-01 -2.94563711e-01 -5.36373973e-01 5.73254108e-01
-1.12821847e-01 -7.73108661e-01 -1.04210556e+00 3.38653952e-01
7.32041240e-01 -9.60676908e-01 1.07818294e+00 6.64727238e-04
-2.14969158e-01 -3.05101603e-01 -9.84876305e-02 -1.00918674e+00
3.69229615e-01 -2.02342287e-01 7.75202811e-01 9.91014481e-01
2.15685874e-01 -7.77025104e-01 1.02665830e+00 8.49715531e-01
-2.59526670e-01 -7.23059714e-01 -1.64599574e+00 -6.34338200e-01
1.31398335e-01 -2.64049470e-01 5.65807104e-01 6.72281027e-01
-8.16992819e-02 -2.52761729e-02 4.33904707e-01 3.59205574e-01
5.70238173e-01 1.59621537e-01 6.21850431e-01 -1.57668865e+00
1.17453761e-01 -4.00426745e-01 -7.05177188e-01 -5.73803902e-01
-2.98618712e-02 -1.03305793e+00 -7.79808387e-02 -1.42284656e+00
-8.68539214e-02 -8.95928919e-01 3.56939822e-01 3.29800397e-01
3.29427719e-01 2.51708657e-01 -9.07943323e-02 4.16530579e-01
7.08829284e-01 3.82194281e-01 1.51244128e+00 2.93035507e-01
-4.28244054e-01 4.73600864e-01 -2.51759410e-01 1.09405470e+00
4.80098158e-01 -6.86030984e-01 -4.60567415e-01 -3.33939463e-01
-4.65121388e-01 1.38308853e-01 2.19698716e-02 -7.39860535e-01
2.68106103e-01 6.10786304e-02 1.72995284e-01 -6.78509891e-01
6.57524228e-01 -1.38832045e+00 4.28223282e-01 6.24910176e-01
4.33195859e-01 6.82262704e-02 2.15869233e-01 1.62401706e-01
-7.25633092e-03 -7.50823259e-01 1.12260497e+00 -3.82045358e-01
-2.51426280e-01 2.82600462e-01 3.46767455e-02 -5.32702088e-01
1.43183541e+00 -6.68808460e-01 2.61995494e-01 -1.02709234e-01
-9.08128023e-01 -5.39394915e-01 6.26235425e-01 -1.41971618e-01
8.33659172e-01 -1.26717305e+00 -5.17009735e-01 5.09347796e-01
-1.11174643e-01 3.61637920e-01 2.49567583e-01 1.00147891e+00
-8.74149740e-01 4.16392893e-01 -2.27107033e-01 -9.59548295e-01
-1.62428939e+00 2.67065346e-01 6.91990674e-01 7.62743352e-04
-6.30672455e-01 6.17090821e-01 -8.49006549e-02 -7.50520885e-01
2.10878011e-02 -4.25537288e-01 -3.34138662e-01 3.27395946e-02
3.98112327e-01 4.02738184e-01 7.77411461e-01 -1.26466572e+00
-5.77387691e-01 1.64946210e+00 9.91585180e-02 -1.01949520e-01
1.45864403e+00 -6.08088188e-02 -2.33179122e-01 -1.70316771e-01
1.34807301e+00 3.55620623e-01 -5.38893402e-01 -2.02157736e-01
-1.16450123e-01 -6.71477795e-01 3.59476715e-01 -4.64068890e-01
-1.41086137e+00 8.53922546e-01 1.08989394e+00 -4.45276380e-01
1.18207347e+00 1.99698493e-01 4.87673521e-01 -2.29573756e-01
6.35734439e-01 -8.39641690e-01 -3.41571003e-01 -1.24283038e-01
9.08500373e-01 -1.03238845e+00 3.26546997e-01 -1.03102195e+00
3.67697924e-02 1.49566126e+00 4.60884959e-01 3.04878324e-01
1.24156785e+00 2.04571739e-01 3.52951854e-01 -7.03942358e-01
2.16528356e-01 -4.68700454e-02 3.46105397e-01 7.80063450e-01
4.77831602e-01 -1.62224263e-01 -7.33935535e-01 1.89057857e-01
-3.70563656e-01 1.32149383e-01 3.49027097e-01 1.14296365e+00
-2.53575295e-01 -1.41186905e+00 -1.05687642e+00 1.18468486e-01
-4.20117408e-01 2.04285115e-01 2.07436144e-01 1.15505683e+00
1.66376561e-01 7.09082544e-01 2.53815442e-01 1.10992163e-01
5.57595432e-01 -1.08680613e-02 7.82908082e-01 -7.78113008e-01
-1.61204979e-01 4.36483264e-01 -3.47171485e-01 -4.33550507e-01
-6.49786890e-01 -7.42430925e-01 -1.47963381e+00 -1.24972217e-01
-7.82208860e-01 9.86286625e-02 1.31835103e+00 8.96411955e-01
8.61707479e-02 -2.60214061e-01 7.11680233e-01 -9.37200248e-01
-4.05831188e-01 -8.10005486e-01 -7.07185805e-01 2.41238043e-01
-1.77756131e-01 -1.09402907e+00 -3.06264222e-01 -5.83321229e-02] | [13.773301124572754, -2.2109780311584473] |
d02cd3b8-fb7a-4950-8e3f-6df05011c7c3 | intelligible-lip-to-speech-synthesis-with | 2305.19603 | null | https://arxiv.org/abs/2305.19603v1 | https://arxiv.org/pdf/2305.19603v1.pdf | Intelligible Lip-to-Speech Synthesis with Speech Units | In this paper, we propose a novel Lip-to-Speech synthesis (L2S) framework, for synthesizing intelligible speech from a silent lip movement video. Specifically, to complement the insufficient supervisory signal of the previous L2S model, we propose to use quantized self-supervised speech representations, named speech units, as an additional prediction target for the L2S model. Therefore, the proposed L2S model is trained to generate multiple targets, mel-spectrogram and speech units. As the speech units are discrete while mel-spectrogram is continuous, the proposed multi-target L2S model can be trained with strong content supervision, without using text-labeled data. Moreover, to accurately convert the synthesized mel-spectrogram into a waveform, we introduce a multi-input vocoder that can generate a clear waveform even from blurry and noisy mel-spectrogram by referring to the speech units. Extensive experimental results confirm the effectiveness of the proposed method in L2S. | ['Yong Man Ro', 'Minsu Kim', 'Jeongsoo Choi'] | 2023-05-31 | null | null | null | null | ['lip-to-speech-synthesis', 'speech-synthesis'] | ['computer-vision', 'speech'] | [ 5.10435164e-01 4.08232749e-01 -4.62500393e-01 -1.37341440e-01
-9.43728149e-01 -2.72817284e-01 4.52839881e-01 -5.18050373e-01
2.51458019e-01 8.87718558e-01 5.19788086e-01 -3.77350986e-01
3.95807475e-01 -2.78729498e-01 -6.08640254e-01 -6.79965317e-01
7.04712212e-01 -4.12593126e-01 1.11064233e-01 4.09378894e-02
4.78363335e-02 1.59747809e-01 -1.88332176e+00 4.59569186e-01
1.09414625e+00 1.11644495e+00 6.44841135e-01 9.32561696e-01
-1.19058937e-01 9.76426303e-01 -8.78231525e-01 -1.65092889e-02
2.54041981e-02 -9.43414509e-01 -3.77668142e-01 5.10993063e-01
1.38020426e-01 -3.85813415e-01 -2.51057655e-01 1.30868578e+00
5.01872003e-01 2.51534998e-01 7.18131363e-01 -1.23965073e+00
-6.34695470e-01 4.88760620e-01 -2.97443699e-02 -2.17354655e-01
4.20293778e-01 3.15850824e-01 7.40284681e-01 -1.11857831e+00
4.67679203e-01 1.47979677e+00 2.29348898e-01 8.60860407e-01
-1.01948214e+00 -6.53127670e-01 1.22436695e-01 1.87653214e-01
-1.22627020e+00 -1.21414936e+00 1.03636813e+00 -2.42680609e-01
4.96256232e-01 3.72293919e-01 3.48123252e-01 1.22212017e+00
-2.71633714e-01 1.17825389e+00 1.08050370e+00 -6.15616798e-01
2.33961865e-01 3.89120430e-01 -3.71154249e-01 3.68198425e-01
-3.30667555e-01 1.78804800e-01 -5.68938911e-01 2.13988736e-01
6.28238916e-01 -3.00301164e-01 -7.14717984e-01 7.03098392e-03
-1.03896737e+00 5.58382690e-01 -4.12715599e-02 2.42710486e-01
-2.55651802e-01 -4.30835076e-02 4.58429843e-01 2.59299904e-01
6.09072626e-01 -1.58897206e-01 -1.06833428e-01 -1.83400035e-01
-1.22443819e+00 -2.27237552e-01 4.56631839e-01 9.80954826e-01
3.90502840e-01 6.69653535e-01 -2.78538883e-01 9.07503188e-01
5.69191515e-01 8.07234526e-01 9.65994656e-01 -8.35889459e-01
5.92041969e-01 1.00794427e-01 2.97001362e-01 -5.67293465e-01
1.45028517e-01 -6.14741966e-02 -8.69521022e-01 4.01667893e-01
3.08991540e-02 -2.13551983e-01 -9.29102242e-01 1.68824840e+00
3.20054561e-01 3.62201214e-01 7.31635571e-01 8.90116394e-01
9.76157486e-01 1.18090606e+00 -1.92664385e-01 -9.02652979e-01
9.28911090e-01 -1.32522655e+00 -1.45206714e+00 -9.01806802e-02
1.04813822e-01 -9.12553370e-01 1.30903614e+00 2.30820626e-01
-1.28705049e+00 -1.09095013e+00 -1.02983332e+00 -8.21997970e-02
-5.82128279e-02 7.71243155e-01 -2.23528475e-01 5.71280479e-01
-9.80104983e-01 2.67303497e-01 -5.31135023e-01 1.08762570e-01
3.42510710e-03 -1.09065011e-01 -1.37521699e-01 2.74961263e-01
-1.30989242e+00 6.72886848e-01 5.17642736e-01 8.69099647e-02
-1.07038105e+00 -3.91440302e-01 -1.27991211e+00 1.49906591e-01
3.16393942e-01 -1.74091116e-01 1.33761883e+00 -1.13866115e+00
-2.31063700e+00 3.53207350e-01 -6.83450282e-01 -3.71139288e-01
4.40856785e-01 1.26312256e-01 -7.35246837e-01 4.72675741e-01
-6.92666620e-02 5.54842174e-01 1.57800055e+00 -1.55448174e+00
-6.58878565e-01 1.84183925e-01 -3.93268466e-01 5.08157015e-01
-2.79000640e-01 1.63762104e-02 -1.99396998e-01 -1.01578379e+00
3.92875299e-02 -5.78282297e-01 2.99636662e-01 -1.37873679e-01
-6.01086974e-01 -3.14158678e-01 1.00503719e+00 -8.89284492e-01
1.26255774e+00 -2.52275610e+00 8.76133516e-02 -3.37386400e-01
-1.70209676e-01 6.97834373e-01 -1.57273754e-01 1.51182979e-01
-6.31356314e-02 1.60472710e-02 -3.62296462e-01 -8.03514481e-01
-4.06304300e-02 -3.42933759e-02 -6.22924328e-01 1.88071683e-01
3.96554828e-01 6.59736276e-01 -8.56679916e-01 -6.03026927e-01
5.31622052e-01 6.13072932e-01 -1.81854755e-01 6.12896085e-01
-2.30157867e-01 4.95605916e-01 -1.07894532e-01 5.39774060e-01
6.18732035e-01 1.59891292e-01 -6.55472800e-02 -2.28850797e-01
-2.57721335e-01 5.39968550e-01 -1.04683626e+00 1.52125502e+00
-7.69788086e-01 7.67865777e-01 4.79448110e-01 -8.27257216e-01
1.25019741e+00 9.05558765e-01 1.82183564e-01 -3.34908187e-01
1.04806945e-02 3.40186626e-01 -2.22426638e-01 -7.84906209e-01
2.73725808e-01 -2.94281691e-01 3.64470124e-01 8.76940191e-02
1.78657040e-01 -5.55470407e-01 -5.69273569e-02 -2.51255155e-01
3.23633999e-01 2.25819767e-01 2.31112108e-01 1.63819537e-01
1.18487465e+00 -5.40166497e-01 4.92460907e-01 1.31893456e-01
-3.69800091e-01 7.14570224e-01 1.43411830e-01 3.64882052e-01
-1.02387786e+00 -1.04152560e+00 1.63886584e-02 8.99000645e-01
2.26628393e-01 -2.61458635e-01 -1.02183986e+00 -5.05741358e-01
-4.85962600e-01 8.16810846e-01 -1.68269500e-01 -2.16180041e-01
-4.10872996e-01 1.09799147e-01 6.57364964e-01 2.92701602e-01
4.38544482e-01 -1.35842049e+00 -2.12125868e-01 3.54690969e-01
-5.17058372e-01 -1.23725879e+00 -9.37142849e-01 -6.46857843e-02
-4.93280768e-01 -6.95808887e-01 -1.25229073e+00 -1.16297913e+00
4.87122327e-01 2.14915588e-01 2.94235826e-01 -2.67278582e-01
1.64269328e-01 2.26373933e-02 -4.36465561e-01 -3.53561997e-01
-1.00466478e+00 -2.94030756e-01 4.13951337e-01 5.83783925e-01
-1.10023327e-01 -5.07255077e-01 -3.75957310e-01 2.23625958e-01
-8.54300976e-01 4.33400333e-01 5.34543991e-01 1.01875114e+00
5.45687854e-01 1.35438204e-01 1.13945198e+00 -9.82395411e-02
8.59136522e-01 -2.12253481e-01 -4.67013150e-01 2.42115319e-01
-2.83852637e-01 -1.48264363e-01 1.26022029e+00 -7.70126045e-01
-1.39485121e+00 -1.17362421e-02 -2.48250335e-01 -9.95399237e-01
-3.93524647e-01 2.76309788e-01 -5.36571860e-01 3.27987283e-01
3.56138587e-01 8.85866344e-01 3.38508308e-01 -6.12758398e-01
5.70352197e-01 1.53108382e+00 9.72501755e-01 -1.51483461e-01
6.55673325e-01 1.81236625e-01 -5.21190405e-01 -1.19838727e+00
-4.45690274e-01 -3.35521311e-01 -2.85706103e-01 -1.79521486e-01
8.83267701e-01 -9.61364388e-01 -7.71960199e-01 6.43359482e-01
-1.42939258e+00 -4.02294040e-01 -3.24980676e-01 7.35901654e-01
-9.51168180e-01 5.57575643e-01 -5.52457094e-01 -1.48850977e+00
-4.03520644e-01 -1.52465045e+00 1.18770397e+00 4.28872436e-01
4.09450307e-02 -7.34440267e-01 -2.30395079e-01 5.80224156e-01
2.01946959e-01 -8.50815251e-02 5.28896689e-01 -4.61271524e-01
-2.47569129e-01 2.02538878e-01 -5.85972890e-03 9.65810955e-01
8.31152439e-01 -1.20490249e-02 -1.19608653e+00 -2.21458212e-01
3.35266203e-01 -6.08205974e-01 6.25881732e-01 3.37158650e-01
1.10183454e+00 -7.63563097e-01 1.02103256e-01 5.68387508e-01
8.55078042e-01 5.88625014e-01 5.49760103e-01 -2.88340807e-01
5.05844653e-01 6.53121054e-01 7.28717327e-01 2.14330122e-01
1.95346907e-01 5.72531164e-01 1.40133679e-01 -2.20027760e-01
-7.25629091e-01 -8.16382229e-01 9.96004641e-01 1.34867418e+00
2.20794603e-01 -2.83917785e-01 -2.60947645e-01 5.97626507e-01
-1.57134259e+00 -9.68419433e-01 2.31252372e-01 2.06121969e+00
1.12180769e+00 1.36997744e-01 2.50791549e-04 6.18321359e-01
1.10890555e+00 4.11319405e-01 -7.41465330e-01 -3.17369014e-01
-6.30283281e-02 6.75036386e-02 -4.35974486e-02 8.62095892e-01
-8.66825223e-01 1.09806538e+00 5.80408430e+00 1.45148063e+00
-1.59586358e+00 -9.81864110e-02 4.23355997e-01 4.31359820e-02
-3.97419661e-01 -3.83008212e-01 -6.45423055e-01 8.50499392e-01
9.46937501e-01 -4.14949089e-01 6.73671067e-01 8.37011814e-01
8.78529787e-01 1.46817923e-01 -9.58775222e-01 1.17781794e+00
2.63201624e-01 -1.20571637e+00 1.68225020e-01 -3.83029461e-01
7.16508925e-01 -6.02810562e-01 3.10284257e-01 2.34554425e-01
-1.68441579e-01 -1.05231214e+00 1.03155923e+00 2.80197620e-01
1.47999763e+00 -5.13076842e-01 2.51025915e-01 5.89923203e-01
-1.39166355e+00 1.16644211e-01 -1.45055622e-01 2.82133549e-01
3.13141733e-01 9.68848467e-02 -1.11373568e+00 5.99413574e-01
2.32935101e-01 6.32853091e-01 1.26013234e-02 6.60823882e-01
-4.28241730e-01 8.18402708e-01 -5.19955903e-02 -3.89870889e-02
1.33015513e-01 -9.89429727e-02 6.62539959e-01 1.08324409e+00
3.30970794e-01 9.32326242e-02 1.02643326e-01 8.40361476e-01
-1.68027934e-02 2.62541920e-01 -5.74603677e-01 -2.45900467e-01
5.87309122e-01 8.11947644e-01 -1.55028820e-01 -5.93324363e-01
-1.88528299e-01 9.99335587e-01 -3.26732218e-01 7.31600642e-01
-8.15366328e-01 -7.91763484e-01 5.79868197e-01 -1.62383258e-01
2.43792728e-01 9.75547475e-04 2.53248848e-02 -1.18268406e+00
1.72750384e-01 -1.13241482e+00 -3.40156406e-01 -1.26619112e+00
-8.85379314e-01 7.33212888e-01 -2.50109792e-01 -1.54730260e+00
-6.00156426e-01 -3.58176172e-01 -6.86452866e-01 1.12121248e+00
-1.94233298e+00 -1.18124640e+00 -9.75378454e-02 8.41956735e-01
1.21940887e+00 -4.38570470e-01 6.38700724e-01 6.88291043e-02
-2.99052596e-01 6.71104789e-01 1.90236002e-01 1.86359789e-02
7.30787396e-01 -9.64062512e-01 2.55163103e-01 8.43083024e-01
9.34229866e-02 3.93932462e-01 6.99872971e-01 -5.66564143e-01
-1.04263043e+00 -1.24387598e+00 7.12839127e-01 2.74290204e-01
4.98710334e-01 -3.69990140e-01 -8.13073516e-01 3.76791388e-01
2.72770613e-01 5.72447723e-04 4.97350872e-01 -8.31456125e-01
-2.83167548e-02 -1.46555036e-01 -1.09666836e+00 5.78946412e-01
7.19119668e-01 -1.06576097e+00 -8.94682169e-01 2.40785792e-01
1.22247958e+00 -4.67293888e-01 -5.43342352e-01 4.22173917e-01
2.33386129e-01 -7.43479431e-01 7.36185670e-01 -2.73223609e-01
4.53831971e-01 -5.94395041e-01 -1.69110239e-01 -1.62357509e+00
2.23120660e-01 -1.09724581e+00 -2.10983694e-01 1.38364792e+00
3.45697314e-01 -4.93026972e-01 5.83966911e-01 -1.07951067e-01
-4.72317129e-01 -3.30783188e-01 -1.05517137e+00 -9.67648625e-01
-2.27313146e-01 -3.35087091e-01 5.23908913e-01 6.47483468e-01
3.03590000e-01 2.94494152e-01 -8.28420281e-01 3.15170884e-01
5.55993915e-01 -5.22667430e-02 7.03799427e-01 -6.06266081e-01
-1.43982723e-01 -3.25028509e-01 -8.27697068e-02 -1.60130668e+00
4.13682401e-01 -7.96094716e-01 5.53517282e-01 -1.41415870e+00
-5.29794872e-01 -7.90008530e-02 -9.87966955e-02 2.38878012e-01
-3.23897421e-01 6.30783811e-02 4.60809171e-01 1.38878495e-01
-6.27764091e-02 1.12930989e+00 1.58323896e+00 -3.75880808e-01
-4.50147182e-01 3.16956311e-01 -2.59876281e-01 6.82007909e-01
5.88414848e-01 -4.09612432e-02 -8.19511950e-01 -6.11702166e-03
-8.20361614e-01 6.60537124e-01 1.44802898e-01 -9.09331679e-01
3.32484603e-01 -3.80311221e-01 -7.45645687e-02 -7.66658127e-01
7.31532454e-01 -7.08499312e-01 -2.53713727e-01 2.72223115e-01
-6.14303052e-01 -6.49370193e-01 5.18721305e-02 4.37820822e-01
-6.86601222e-01 -3.08446199e-01 8.89690459e-01 1.88422516e-01
-4.13867831e-01 1.19213745e-01 -4.57935005e-01 -1.10552125e-01
9.41055298e-01 -2.86532909e-01 -1.68495446e-01 -7.01208830e-01
-8.85738790e-01 -3.67817171e-02 1.49070859e-01 4.87991631e-01
9.43535089e-01 -1.49649608e+00 -6.21849000e-01 4.16070700e-01
-1.29875526e-01 1.03551395e-01 3.19238275e-01 3.33948016e-01
-1.21706754e-01 5.10489106e-01 -3.99076715e-02 -5.10912478e-01
-1.24773812e+00 7.56048679e-01 1.77134633e-01 3.49567115e-01
-4.84759688e-01 5.43077588e-01 2.97566503e-01 -1.37191594e-01
4.90949035e-01 -6.35785103e-01 -3.50961506e-01 -6.46208301e-02
6.69414878e-01 2.38094881e-01 -2.21116304e-01 -7.75209785e-01
-1.26481056e-03 3.48160356e-01 4.71825689e-01 -5.91364205e-01
9.15878057e-01 -4.40100044e-01 1.87318072e-01 7.34485447e-01
1.20102370e+00 4.53859150e-01 -1.41393316e+00 3.69935706e-02
-3.72992635e-01 -2.08187222e-01 -3.72307748e-02 -4.90247697e-01
-7.90295064e-01 1.04423678e+00 3.37910265e-01 3.76417398e-01
1.23386765e+00 -3.07143599e-01 9.16917682e-01 7.61196911e-02
-4.49468121e-02 -1.07557237e+00 2.32098266e-01 1.71332508e-01
1.03837824e+00 -1.08650017e+00 -6.34337306e-01 -5.55805087e-01
-8.80791843e-01 1.21303678e+00 5.70836246e-01 2.77131140e-01
3.20671916e-01 2.02367991e-01 3.75161797e-01 6.48594618e-01
-7.50599086e-01 -2.86938757e-01 1.62442997e-01 8.05830419e-01
1.36166453e-01 -4.14822847e-02 -1.46741658e-01 5.81469119e-01
-4.09616679e-01 2.84487814e-01 6.30359828e-01 3.62280190e-01
-5.91712952e-01 -9.24462020e-01 -6.22970819e-01 -1.78365469e-01
-1.35014445e-01 -2.32843235e-01 -1.88093279e-02 1.97622627e-01
1.05288766e-01 1.42040896e+00 -5.09317219e-03 -3.65669012e-01
1.30442500e-01 4.11342621e-01 1.06461924e-02 -5.51714361e-01
1.41618922e-02 5.17093778e-01 -1.20356068e-01 -1.30992055e-01
-3.26566041e-01 -2.53596991e-01 -1.29021668e+00 3.00095499e-01
-3.52296323e-01 3.57427269e-01 7.22136915e-01 8.89338672e-01
1.49543449e-01 6.56369805e-01 1.25025702e+00 -9.01724339e-01
-8.23868990e-01 -1.25282252e+00 -4.82942641e-01 4.04732764e-01
9.77255940e-01 -4.43827718e-01 -7.07260728e-01 5.86592793e-01] | [14.547938346862793, 5.40910005569458] |
a7626a4e-521f-4c39-99ad-27bc0b9a85a8 | distributional-reinforcement-learning-for-2 | null | null | https://openreview.net/forum?id=19drPzGV691 | https://openreview.net/pdf?id=19drPzGV691 | Distributional Reinforcement Learning for Risk-Sensitive Policies | We address the problem of learning a risk-sensitive policy based on the CVaR risk measure using distributional reinforcement learning. In particular, we show that applying the distributional Bellman optimality operator with respect to a risk-based action-selection strategy overestimates the dynamic, Markovian CVaR. The resulting policies can however still be overly conservative and one often prefers to learn an optimal policy based on the static, non-Markovian CVaR. To this end, we propose a modification to the existing algorithm and show that it can indeed learn a proper CVaR-optimized policy. Our proposed approach is a simple extension of standard distributional RL algorithms and can therefore take advantage of many of the recent advances in deep RL. On both synthetic and real data, we empirically show that our proposed algorithm is able to produce a family of risk-averse policies that achieves a better tradeoff between risk and the expected return.
| ['Ilyas Malik', 'Shiau Hong Lim'] | 2021-01-01 | null | null | null | null | ['distributional-reinforcement-learning'] | ['methodology'] | [-1.87728003e-01 3.80285203e-01 -3.81972551e-01 -2.62064874e-01
-1.15593481e+00 -7.22187817e-01 5.54802716e-01 2.39102557e-01
-1.06655848e+00 1.11468387e+00 6.16233088e-02 -5.03963351e-01
-4.48194355e-01 -8.30079138e-01 -5.54544151e-01 -9.06517625e-01
-2.92113394e-01 6.14615917e-01 -2.22489517e-02 -7.07884580e-02
4.24384236e-01 4.90105897e-01 -1.36217511e+00 -3.29845250e-01
8.91872525e-01 1.17153966e+00 9.99210477e-02 6.76284790e-01
1.71059757e-01 6.63276136e-01 -6.24711454e-01 -3.93118769e-01
3.86139989e-01 -4.32376921e-01 -5.82153380e-01 -5.66851735e-01
-3.17554772e-02 -3.67661178e-01 2.93108702e-01 1.08071160e+00
5.99627495e-01 4.58301157e-01 8.60322535e-01 -1.04088402e+00
-2.65792102e-01 8.49338531e-01 -5.00604570e-01 2.29247525e-01
-1.97106209e-02 1.67197585e-01 1.24423468e+00 -1.32784665e-01
3.66587400e-01 1.49303818e+00 4.78451490e-01 5.73074460e-01
-1.42003322e+00 -4.70874131e-01 2.93609768e-01 -2.64602333e-01
-1.02332401e+00 -1.20555937e-01 4.68011796e-01 -2.07143769e-01
6.72177494e-01 -9.33397282e-03 5.32478511e-01 1.31030464e+00
5.03490925e-01 7.96704233e-01 1.70384312e+00 -3.61739278e-01
7.95538425e-01 3.87947746e-02 -3.33952785e-01 4.30290103e-01
4.52775389e-01 8.04714501e-01 9.97001156e-02 -4.23643857e-01
5.83374381e-01 -2.22161055e-01 4.92599420e-02 -8.59569967e-01
-7.33350158e-01 1.18838096e+00 -3.67381796e-02 1.18926294e-01
-6.35327756e-01 5.03837407e-01 4.18663144e-01 5.33919275e-01
3.26476872e-01 5.79687059e-01 -4.46925610e-01 -4.70513195e-01
-7.48092234e-01 7.58952975e-01 1.00214577e+00 4.59557593e-01
3.31689924e-01 2.63011843e-01 -6.25655115e-01 4.56730336e-01
3.81908178e-01 5.86412132e-01 2.70059228e-01 -1.44424450e+00
2.47908667e-01 -1.75462395e-01 7.84847140e-01 -5.82878470e-01
-3.83441031e-01 -5.60919821e-01 -2.20694542e-01 7.81047404e-01
5.79491079e-01 -5.08962095e-01 -5.11349320e-01 2.05141592e+00
1.34585053e-01 -2.87215132e-02 2.69387752e-01 5.24110734e-01
-4.53864008e-01 2.56584138e-01 3.45367193e-01 -7.33304381e-01
6.86143696e-01 -2.34849334e-01 -7.55221605e-01 1.19040422e-01
4.66546327e-01 -3.52141172e-01 1.41472340e+00 5.38667858e-01
-1.09115565e+00 -1.54368803e-02 -1.05467713e+00 6.57597840e-01
-3.70580517e-03 -3.92815679e-01 4.55927461e-01 9.84516919e-01
-9.00950789e-01 1.08214140e+00 -7.39200592e-01 -1.70530349e-01
5.92589915e-01 8.38676915e-02 3.16533029e-01 2.76317686e-01
-1.02796841e+00 1.07737136e+00 5.10874987e-01 -3.01148683e-01
-1.42286277e+00 -5.41658401e-01 -5.48804641e-01 -4.10267673e-02
1.05580676e+00 -4.82674181e-01 1.75942874e+00 -9.55222785e-01
-2.10489345e+00 3.48535299e-01 5.52883148e-01 -1.00304163e+00
1.01870966e+00 -4.80498135e-01 4.61937003e-02 2.76202653e-02
-1.24946617e-01 1.21358663e-01 1.05832827e+00 -1.20026362e+00
-6.67193711e-01 -8.66878629e-02 2.30299443e-01 2.01930434e-01
5.33811115e-02 -1.25775293e-01 6.37800515e-01 -7.13408768e-01
-8.03813994e-01 -9.62725639e-01 -5.09097397e-01 -4.10326630e-01
-1.65791094e-01 -2.79406518e-01 2.97709852e-01 -1.37897223e-01
1.03804362e+00 -1.89273918e+00 8.17246810e-02 3.52395236e-01
-1.36983335e-01 1.48654893e-01 1.35657694e-02 3.72998267e-01
1.64135903e-01 2.40053192e-01 -4.05122191e-01 -2.38034189e-01
4.35946703e-01 4.11964953e-01 -5.52605152e-01 7.75781155e-01
-1.76521510e-01 6.63155675e-01 -1.12548125e+00 -3.15236092e-01
1.29501760e-01 1.43036693e-01 -5.48605680e-01 3.48015517e-01
-5.98466814e-01 3.72565269e-01 -6.12294495e-01 1.50769338e-01
3.91549528e-01 4.78218526e-01 2.83297688e-01 5.51961660e-01
-2.84724385e-01 3.24940681e-03 -1.09310091e+00 1.32338047e+00
-5.91434240e-01 9.79821458e-02 3.64352087e-03 -1.22293830e+00
7.50870526e-01 6.66779503e-02 5.39067864e-01 -3.98483783e-01
4.50231999e-01 2.57449746e-01 -3.89870927e-02 -8.31588879e-02
1.75790936e-01 -7.01888442e-01 -2.06690192e-01 8.73198271e-01
1.55916475e-02 -1.50828073e-02 -2.02003226e-01 -1.57563657e-01
8.96964431e-01 5.70781529e-01 6.76628649e-01 -6.41684890e-01
2.48788491e-01 -5.02602100e-01 5.51771462e-01 1.29977536e+00
-5.27329147e-01 3.38048935e-02 1.08852530e+00 -5.49524762e-02
-8.84296834e-01 -1.32288253e+00 -5.76346554e-03 1.14968348e+00
-3.10736120e-01 -3.77517641e-02 -7.81382740e-01 -1.13535714e+00
3.00221801e-01 1.43539166e+00 -7.06902862e-01 -3.80825907e-01
-3.34223688e-01 -6.06927574e-01 4.92746651e-01 4.20131624e-01
1.28905892e-01 -1.18808055e+00 -1.25519812e+00 1.54041678e-01
2.57118434e-01 -4.99859303e-01 -3.08309555e-01 3.16207856e-01
-7.20724821e-01 -8.20069194e-01 -7.70041406e-01 6.51803762e-02
9.59864557e-02 -4.03636992e-01 1.10258949e+00 -4.78653461e-01
2.18747228e-01 8.57844353e-01 -1.40519187e-01 -8.48368406e-01
-6.28416359e-01 -8.85243267e-02 3.04487795e-01 7.43982568e-02
1.24242187e-01 -5.34670234e-01 -6.14592016e-01 -6.65870607e-02
-7.34005272e-01 -7.45758474e-01 4.96473163e-01 7.53749311e-01
7.37364888e-01 1.68109909e-01 8.91290188e-01 -1.00567901e+00
1.19130147e+00 -4.89154845e-01 -1.16523612e+00 2.18152031e-01
-1.19510174e+00 7.51491547e-01 7.34788239e-01 -5.08228898e-01
-1.29970515e+00 -8.69929641e-02 9.15568396e-02 -3.99863452e-01
3.78057174e-02 2.39852220e-01 -1.54238880e-01 1.24922141e-01
5.04360974e-01 -4.23375107e-02 3.15162651e-02 -3.98470402e-01
6.82269096e-01 3.55520785e-01 1.87012091e-01 -1.16365647e+00
5.40025115e-01 2.92767256e-01 2.20351800e-01 -3.25396687e-01
-9.58758891e-01 2.36056864e-01 -2.43612751e-01 -2.48563513e-01
7.18815863e-01 -3.93315881e-01 -1.22521734e+00 1.46680232e-02
-5.62114358e-01 -6.93938315e-01 -9.06038880e-01 5.47056198e-01
-1.51644242e+00 2.51736075e-01 -1.95572168e-01 -1.50358212e+00
-6.93706423e-02 -9.90983665e-01 6.91749513e-01 3.90030630e-02
-2.21109372e-02 -1.11306274e+00 4.39514995e-01 -4.63833302e-01
3.81176144e-01 5.33119977e-01 9.59138632e-01 -7.74212599e-01
-1.33589640e-01 1.91302270e-01 2.84186035e-01 5.82434833e-01
-1.83084249e-01 -4.53193001e-02 -8.26130271e-01 -4.88866031e-01
2.07605079e-01 -6.14735067e-01 9.11006391e-01 6.60143435e-01
1.02126789e+00 -6.18844390e-01 1.46938071e-01 3.49839389e-01
1.59256601e+00 4.35286850e-01 3.87752980e-01 5.64088583e-01
1.93819016e-01 6.10578775e-01 1.12604153e+00 8.23661625e-01
1.91092074e-01 5.10588229e-01 6.32991672e-01 5.93622446e-01
6.64466381e-01 -6.29978180e-01 7.27291763e-01 -9.35089663e-02
-1.95342749e-01 6.97740465e-02 -5.56375206e-01 2.79972851e-01
-2.06784725e+00 -1.17452300e+00 7.16400921e-01 2.77858377e+00
9.29809213e-01 2.48718143e-01 6.93432808e-01 -2.11200386e-01
3.57355118e-01 1.61686867e-01 -8.25019717e-01 -9.32658315e-01
7.42154866e-02 3.97170842e-01 1.00323534e+00 7.31575906e-01
-1.01021099e+00 9.62265074e-01 7.27685928e+00 8.93139005e-01
-7.72490799e-01 7.05473870e-02 6.37786627e-01 -2.81968713e-01
-4.81760740e-01 -6.57243282e-02 -6.22632027e-01 3.70785683e-01
1.18683481e+00 -4.98442441e-01 6.30405426e-01 1.14655542e+00
3.06296378e-01 -2.18824163e-01 -1.17668176e+00 6.59890115e-01
-4.46282685e-01 -9.05222714e-01 -3.14161748e-01 2.77175188e-01
6.17079854e-01 -3.55036259e-01 3.48183334e-01 6.25480711e-01
1.04306328e+00 -1.27495384e+00 9.79464173e-01 7.66660690e-01
4.55390722e-01 -1.45757484e+00 7.61325181e-01 4.81945276e-01
-5.94041169e-01 -4.99913186e-01 -4.26553190e-01 3.77978124e-02
-7.94129074e-02 3.89137238e-01 -5.98063707e-01 2.97218770e-01
4.11407441e-01 4.04818773e-01 -4.14921306e-02 7.98997879e-01
-2.69093484e-01 5.61242461e-01 -2.15170667e-01 -1.49283320e-01
6.12879813e-01 -3.46303821e-01 6.06972873e-01 9.54620779e-01
3.72494251e-01 -2.55604804e-01 2.91458368e-01 9.20041919e-01
2.16196865e-01 1.23815224e-01 -9.18253422e-01 -1.46896496e-01
2.83897370e-01 9.80709374e-01 -5.44077814e-01 -1.49895743e-01
1.36592740e-03 5.16197324e-01 6.24561012e-01 3.63774836e-01
-8.22867572e-01 -7.01409504e-02 9.01469588e-01 -3.50070149e-01
5.76974452e-01 -1.27802759e-01 -4.61447611e-02 -8.29904020e-01
-1.85416922e-01 -8.78184438e-01 7.28391528e-01 -8.41642022e-02
-1.30834675e+00 1.95675984e-01 3.68584126e-01 -7.98319161e-01
-7.06362307e-01 -5.03721893e-01 -3.98237050e-01 7.88811743e-01
-1.52366626e+00 -3.58777404e-01 6.62173331e-01 5.07889450e-01
1.47231385e-01 -2.00649992e-01 7.43131816e-01 -4.00981396e-01
-3.82223666e-01 4.99246895e-01 3.62104028e-01 -4.66999322e-01
5.05320072e-01 -1.71948957e+00 -2.17036381e-01 5.83024085e-01
-1.86228380e-02 2.20773548e-01 1.05784845e+00 -4.84148592e-01
-1.09081221e+00 -1.02292740e+00 1.76253691e-01 -4.46924716e-01
8.28272402e-01 -8.86297598e-02 -5.67179084e-01 7.70411491e-01
2.27816507e-01 -1.87641993e-01 4.37995315e-01 5.50933443e-02
-2.46516094e-01 -2.62987494e-01 -1.53068364e+00 7.67900884e-01
8.53279293e-01 -1.45877525e-01 -7.34781325e-01 5.88808060e-02
7.45164216e-01 -1.87644865e-02 -8.33204746e-01 3.38394791e-01
6.42076731e-01 -1.15301049e+00 8.67603421e-01 -9.45243955e-01
-1.40168443e-01 1.37605995e-01 -2.34804034e-01 -1.67568207e+00
-9.17341933e-02 -1.18652868e+00 -1.36634141e-01 9.80681956e-01
-3.91319469e-02 -9.04056549e-01 4.60090011e-01 4.40298140e-01
3.31650496e-01 -8.95156264e-01 -1.24882984e+00 -1.20646715e+00
7.21343815e-01 -6.24935508e-01 4.30351287e-01 3.62782598e-01
-1.84130788e-01 -7.14464858e-02 -5.45847595e-01 -1.09904148e-01
1.10200667e+00 1.46467239e-01 3.53772730e-01 -1.13465095e+00
-5.62491953e-01 -7.73385108e-01 1.44329503e-01 -5.69704473e-01
6.19312167e-01 -5.48073590e-01 2.12183490e-01 -1.00532019e+00
-1.29973739e-01 -5.23226619e-01 -7.38848567e-01 2.90965229e-01
4.13644500e-02 -5.56103170e-01 3.46023828e-01 -2.26863325e-01
-6.11252427e-01 8.06871116e-01 1.12270999e+00 2.28545204e-01
-3.87950212e-01 4.43081588e-01 -9.15704429e-01 8.61814797e-01
1.08876026e+00 -7.93752193e-01 -7.27759182e-01 3.27534825e-01
3.30534101e-01 3.25236678e-01 1.91613242e-01 -5.64904332e-01
-4.36125904e-01 -7.81143546e-01 2.08550179e-03 -1.03418462e-01
-6.14712052e-02 -7.10538864e-01 -2.74498522e-01 6.86992645e-01
-8.92162442e-01 6.34932965e-02 -4.38949913e-02 9.98456001e-01
4.05472875e-01 -3.78429711e-01 1.12785614e+00 -3.35605711e-01
-2.19967529e-01 2.91414291e-01 -5.12469709e-01 4.48673010e-01
1.21352363e+00 3.49305689e-01 7.66199008e-02 -6.69908881e-01
-4.96637106e-01 9.02054757e-02 3.37238520e-01 1.05542295e-01
4.36059058e-01 -1.09284484e+00 -5.77084959e-01 -2.35307917e-01
-4.92348447e-02 -5.09957969e-01 -1.24321736e-01 6.87316775e-01
-6.40943125e-02 4.24781084e-01 -2.40427762e-01 -1.13691859e-01
-6.72725677e-01 8.13944757e-01 5.90135455e-01 -5.84385872e-01
-6.53833210e-01 3.90647709e-01 1.78601909e-02 -1.75585046e-01
3.92848670e-01 -4.24445271e-01 -1.60294369e-01 8.29099566e-02
4.48343009e-01 5.21575511e-01 -4.52883273e-01 -3.17390144e-01
-8.70934129e-02 2.80671805e-01 2.25340962e-01 -7.04813361e-01
1.23166811e+00 -1.84849367e-01 3.91258419e-01 7.81420290e-01
8.16011608e-01 1.27902821e-01 -1.79389548e+00 -2.48839483e-02
4.40558255e-01 -5.79953074e-01 1.15774147e-01 -8.51776123e-01
-9.41834152e-01 7.51821578e-01 7.02187777e-01 2.09586203e-01
9.11071539e-01 -2.50076056e-01 3.39764833e-01 5.44547856e-01
6.63094044e-01 -1.45089459e+00 -9.80469957e-02 2.46735543e-01
7.79164493e-01 -9.86757934e-01 -8.16016365e-03 4.34869409e-01
-1.10641956e+00 9.33731437e-01 2.44172662e-01 -5.38501203e-01
6.63859606e-01 3.00051093e-01 2.26767380e-02 2.90486723e-01
-9.07331288e-01 -5.07468402e-01 -1.53579459e-01 9.00706947e-01
3.45512331e-02 5.01217544e-01 -5.60960472e-01 5.26812792e-01
-1.46316960e-01 -7.24196956e-02 5.36192179e-01 8.58633101e-01
-5.33309042e-01 -1.24137676e+00 -2.18669116e-01 3.06291878e-01
-7.75118470e-01 2.16490597e-01 -1.34051248e-01 8.59715879e-01
-4.01700288e-01 7.07479537e-01 -1.35328054e-01 1.23058092e-02
2.84817547e-01 1.29889905e-01 6.84474885e-01 -3.18705529e-01
-3.11911464e-01 1.21561579e-01 -2.07654852e-02 -8.53951395e-01
-4.37982023e-01 -8.77070367e-01 -1.07312894e+00 -1.80499077e-01
3.23938459e-01 7.61551857e-02 4.54495460e-01 1.00229824e+00
1.17533803e-01 2.54856408e-01 7.89277971e-01 -5.50920427e-01
-1.67178226e+00 -6.10775530e-01 -8.77707541e-01 1.82039112e-01
4.71879601e-01 -1.11813343e+00 -4.87882018e-01 -7.33249843e-01] | [4.222787380218506, 2.5596389770507812] |
e41a3608-c0ca-4272-8a88-96dd1d2d7413 | how-do-human-users-teach-a-continual-learning | 2307.00123 | null | https://arxiv.org/abs/2307.00123v1 | https://arxiv.org/pdf/2307.00123v1.pdf | How Do Human Users Teach a Continual Learning Robot in Repeated Interactions? | Continual learning (CL) has emerged as an important avenue of research in recent years, at the intersection of Machine Learning (ML) and Human-Robot Interaction (HRI), to allow robots to continually learn in their environments over long-term interactions with humans. Most research in continual learning, however, has been robot-centered to develop continual learning algorithms that can quickly learn new information on static datasets. In this paper, we take a human-centered approach to continual learning, to understand how humans teach continual learning robots over the long term and if there are variations in their teaching styles. We conducted an in-person study with 40 participants that interacted with a continual learning robot in 200 sessions. In this between-participant study, we used two different CL models deployed on a Fetch mobile manipulator robot. An extensive qualitative and quantitative analysis of the data collected in the study shows that there is significant variation among the teaching styles of individual users indicating the need for personalized adaptation to their distinct teaching styles. The results also show that although there is a difference in the teaching styles between expert and non-expert users, the style does not have an effect on the performance of the continual learning robot. Finally, our analysis shows that the constrained experimental setups that have been widely used to test most continual learning techniques are not adequate, as real users interact with and teach continual learning robots in a variety of ways. Our code is available at https://github.com/aliayub7/cl_hri. | ['Chrystopher L. Nehaniv', 'Kerstin Dautenhahn', 'Patrick Holthaus', 'Zachary De Francesco', 'Jainish Mehta', 'Ali Ayub'] | 2023-06-30 | null | null | null | null | ['continual-learning'] | ['methodology'] | [-3.95936668e-01 2.62363821e-01 -1.41276166e-01 -4.06444550e-01
-1.98199973e-01 -7.03236163e-01 3.74649048e-01 1.73757210e-01
-5.28355420e-01 6.22645557e-01 -2.67833099e-02 -4.93342906e-01
-7.36448765e-01 -2.42634371e-01 -8.84914994e-01 -1.93461314e-01
-4.25831914e-01 6.45682335e-01 8.86338353e-02 -4.61982459e-01
1.55855164e-01 5.66438317e-01 -1.70585942e+00 2.00787500e-01
7.45836914e-01 2.97864396e-02 5.90353608e-01 5.15989065e-01
1.50236309e-01 7.82846987e-01 -4.62388456e-01 1.95387974e-01
3.63479912e-01 -2.33380616e-01 -1.17083180e+00 1.74276292e-01
4.67435986e-01 -4.04645264e-01 -1.52815759e-01 5.76515436e-01
3.85816604e-01 6.31325901e-01 6.95362687e-01 -1.59802353e+00
-5.60749650e-01 5.77888310e-01 -2.57468581e-01 -7.31444731e-02
7.88388789e-01 5.02670228e-01 6.07388258e-01 -5.79551458e-01
4.85915035e-01 1.59573436e+00 7.61136293e-01 4.97269303e-01
-1.28044224e+00 -7.75796950e-01 8.09298456e-03 2.61991560e-01
-1.15339684e+00 -1.46044165e-01 3.97953808e-01 -7.87937224e-01
7.39542603e-01 -1.49403885e-01 8.67635965e-01 1.03462613e+00
4.63799506e-01 6.65789425e-01 1.06991053e+00 -8.07587087e-01
1.99473277e-01 7.44446814e-01 3.97403985e-01 5.37417889e-01
1.39301077e-01 2.92228580e-01 -5.03862143e-01 6.65258393e-02
9.40137684e-01 1.69927984e-01 -5.13304509e-02 -7.40989149e-01
-1.07561445e+00 6.51244581e-01 3.39775145e-01 5.78606129e-01
-5.70489943e-01 1.44385785e-01 2.57891923e-01 1.00291753e+00
-1.88968882e-01 6.32246494e-01 -6.22902751e-01 -5.33844471e-01
1.66562304e-01 3.11518162e-01 1.07839620e+00 1.13498569e+00
9.35912669e-01 -5.49839973e-01 3.11582536e-01 8.77884328e-01
3.40407103e-01 4.15171646e-02 6.49377763e-01 -1.40219474e+00
-2.00126134e-02 5.28915048e-01 3.99589121e-01 -8.45089197e-01
-5.57216763e-01 7.57780746e-02 2.62276471e-01 7.24007547e-01
4.51629847e-01 -5.04581213e-01 -3.76564711e-01 1.54591680e+00
4.05198127e-01 -4.33758348e-01 -5.51360473e-02 5.32381177e-01
2.62741119e-01 4.44697887e-01 5.96915483e-01 -1.43056288e-01
1.10696924e+00 -1.03375351e+00 -5.38558543e-01 -4.32847470e-01
9.62379098e-01 -7.93403625e-01 1.94188237e+00 6.59122705e-01
-5.73125660e-01 -1.00243139e+00 -8.83161247e-01 2.95178086e-01
-3.27710539e-01 -1.17630489e-01 8.71477008e-01 7.49923944e-01
-1.06629527e+00 6.03976607e-01 -1.02439344e+00 -1.15009475e+00
-1.19870953e-01 5.09255767e-01 -3.68699491e-01 -1.08508460e-01
-6.43189132e-01 1.15110600e+00 4.45697367e-01 -4.37438488e-01
-9.70205665e-01 -5.77335715e-01 -6.36565030e-01 -2.12358668e-01
4.57453609e-01 -2.74198115e-01 1.83223164e+00 -1.28788579e+00
-1.55597126e+00 2.68908411e-01 3.15015823e-01 3.00306082e-02
4.79732007e-01 -6.44418180e-01 9.26697478e-02 -2.02677473e-01
2.49449030e-01 9.06266689e-01 2.08915293e-01 -1.46136439e+00
-8.62895072e-01 -3.38969827e-01 4.44915295e-01 6.47242308e-01
-6.23744786e-01 -2.62173176e-01 5.08540682e-02 -2.36117333e-01
-2.10037917e-01 -1.48873889e+00 2.32988801e-02 -1.68866202e-01
4.10195500e-01 -7.70510018e-01 1.12053239e+00 -2.11200342e-01
6.33148789e-01 -2.29517412e+00 4.87291552e-02 9.03177038e-02
-7.73904100e-02 -6.20231926e-02 -1.74785122e-01 8.93568814e-01
-8.35465416e-02 -2.44565815e-01 3.26410145e-01 -1.66870534e-01
2.11832076e-02 2.35353455e-01 2.21378073e-01 3.34886342e-01
-5.58454633e-01 5.73651493e-01 -1.10152519e+00 -1.65964276e-01
2.07493737e-01 2.70642847e-01 -5.51921427e-01 4.53322500e-01
-7.93306455e-02 6.54661417e-01 -1.58412963e-01 3.39225888e-01
3.36308628e-02 7.30275810e-02 3.36903602e-01 2.32678950e-01
-3.38711172e-01 -1.72086880e-01 -9.63303268e-01 1.70254648e+00
-6.58269107e-01 7.41776586e-01 -9.10337940e-02 -5.53746760e-01
4.29673702e-01 5.31695664e-01 6.31093621e-01 -5.70733309e-01
3.86180319e-02 2.71734387e-01 3.29144746e-01 -9.97972786e-01
4.87907976e-01 -9.62003469e-02 9.71208140e-02 7.24349916e-01
1.67040214e-01 -8.60544816e-02 1.64363369e-01 -1.17126182e-02
1.09697473e+00 2.75802314e-01 1.40043229e-01 -4.35042411e-01
1.07506644e-02 4.30448353e-01 2.46331185e-01 7.53413618e-01
-5.45798540e-01 -1.23700872e-01 4.45051752e-02 -3.12174767e-01
-6.43595636e-01 -8.22680473e-01 2.39180386e-01 1.66934407e+00
2.12644130e-01 -9.52492654e-02 -6.75881267e-01 -3.79776776e-01
-1.28894439e-03 1.24816000e+00 -3.34952533e-01 -2.45839044e-01
-4.18530315e-01 -2.29538351e-01 1.33846700e-01 2.28295878e-01
3.58741164e-01 -1.44508040e+00 -1.07624960e+00 1.86606809e-01
-6.07792325e-02 -5.15206575e-01 -3.72402966e-01 3.35714102e-01
-9.65347409e-01 -1.08633029e+00 -2.72678047e-01 -1.25043559e+00
7.24731266e-01 7.36664712e-01 6.69260383e-01 3.31412293e-02
-1.48165271e-01 1.61850774e+00 -6.57423139e-01 -7.69778073e-01
-6.46902502e-01 3.11250746e-01 5.62752545e-01 -7.08240628e-01
4.07744825e-01 -7.62516856e-01 -4.28320020e-01 5.87651908e-01
-6.33461773e-01 -1.60619199e-01 9.20664787e-01 5.09719968e-01
-1.24636844e-01 3.90488297e-01 9.25045311e-01 -9.36598241e-01
1.10686398e+00 -7.24214554e-01 -2.29186848e-01 3.97337794e-01
-1.00802112e+00 -4.23816979e-01 1.36024639e-01 -9.20028031e-01
-1.25538743e+00 2.01076165e-01 2.29990333e-01 -8.99851546e-02
-6.96204960e-01 7.47838974e-01 1.13276288e-01 -2.78993875e-01
1.24153090e+00 -3.61751884e-01 4.05334473e-01 -2.53565669e-01
2.75175750e-01 8.87823224e-01 2.40458459e-01 -9.21392560e-01
6.95158899e-01 -1.35204256e-01 -7.14269638e-01 -9.99456286e-01
-2.63984650e-01 -5.07899284e-01 -9.49707985e-01 -7.30533183e-01
4.59054947e-01 -8.47093880e-01 -1.14191580e+00 5.30509591e-01
-8.09078276e-01 -1.57369602e+00 -3.68659139e-01 8.37622285e-01
-8.54836285e-01 -1.67958084e-02 -5.61669946e-01 -9.51450109e-01
1.37264490e-01 -1.22580194e+00 3.85440111e-01 4.21802044e-01
-8.58042479e-01 -1.14445007e+00 1.04683168e-01 4.17863637e-01
5.20197690e-01 -1.11996651e-01 9.83136475e-01 -7.30919898e-01
-2.32649937e-01 -1.47491753e-01 4.45867062e-01 8.92418250e-02
7.12822080e-01 -3.92880231e-01 -5.46732843e-01 -6.74540520e-01
2.30955146e-02 -9.06747222e-01 -3.46311480e-01 -2.15133391e-02
4.06761229e-01 -2.09766448e-01 -3.28613997e-01 -4.21273768e-01
1.15200317e+00 7.71103561e-01 4.15517330e-01 7.46363342e-01
5.56439042e-01 9.04002845e-01 1.10320961e+00 -1.22141987e-02
4.26088750e-01 3.95973474e-01 -9.24188793e-02 3.84224176e-01
1.37006432e-01 -1.71576962e-01 7.36186683e-01 3.27625424e-01
-5.49631380e-02 8.75517875e-02 -1.08101809e+00 3.96789908e-01
-1.99969244e+00 -9.25964534e-01 9.20893475e-02 2.15100336e+00
6.66551471e-01 2.90495697e-02 4.55151796e-01 -1.05419047e-01
3.89835387e-01 -6.85610712e-01 -5.33201039e-01 -3.57599795e-01
7.74432182e-01 -3.26209873e-01 3.57674211e-01 6.38438284e-01
-7.06488311e-01 8.72727334e-01 6.37898922e+00 3.67915750e-01
-1.01349306e+00 9.38530732e-03 3.96726102e-01 1.15465887e-01
1.87835157e-01 -1.09001793e-01 -3.75235081e-01 -4.01060656e-02
1.07254779e+00 -2.37864181e-01 6.78752780e-01 1.14676428e+00
5.90413928e-01 -5.50843179e-01 -1.40756917e+00 5.66794276e-01
-1.56780928e-01 -3.30968559e-01 -3.53905678e-01 9.06231329e-02
5.02934277e-01 -1.94339395e-01 9.79751125e-02 1.03582585e+00
3.97442728e-01 -7.33269870e-01 7.61988163e-01 2.80406177e-01
2.56473690e-01 -6.60734475e-01 3.42126936e-01 8.27844620e-01
-7.04677165e-01 -8.22541177e-01 -1.24488465e-01 -3.58360946e-01
-1.85056664e-02 -3.00799519e-01 -1.39856696e+00 6.77514076e-03
1.03165996e+00 2.07742453e-01 -6.15014136e-01 8.27978015e-01
-7.77975991e-02 6.21401131e-01 -3.49493586e-02 -6.99388906e-02
4.61842231e-02 -2.58584172e-01 3.27044994e-01 8.54365468e-01
1.23446368e-01 2.80640930e-01 4.40185845e-01 4.44748163e-01
3.45704108e-01 1.87600419e-01 -8.67966831e-01 2.88173120e-04
6.04030132e-01 1.14299476e+00 -9.03596044e-01 1.03256322e-01
-5.20077765e-01 1.00367320e+00 3.35987091e-01 3.73930484e-01
-5.28727055e-01 -3.32544208e-01 3.74372929e-01 2.50288457e-01
-2.77989149e-01 -7.53403664e-01 -7.69459382e-02 -2.84369528e-01
-2.63357490e-01 -1.32123196e+00 2.33345151e-01 -7.35711455e-01
-1.25210989e+00 1.41796544e-01 5.89901567e-01 -1.17972076e+00
-3.11981410e-01 -4.66773808e-01 -4.76437330e-01 4.26803827e-01
-7.62554765e-01 -1.11214483e+00 -6.61577940e-01 4.58558768e-01
1.01244438e+00 -2.62323439e-01 7.48624861e-01 2.80696988e-01
-3.75230879e-01 5.29735625e-01 -3.83314560e-03 -4.38149303e-01
1.07777405e+00 -9.86461937e-01 -2.43776500e-01 1.71635319e-02
-1.83980480e-01 1.37191093e+00 8.63777459e-01 -1.01492548e+00
-1.63357210e+00 -6.63810313e-01 1.15958549e-01 -5.85415483e-01
4.50641513e-01 -6.84646219e-02 -1.00454998e+00 1.40251660e+00
5.91293514e-01 -8.09154987e-01 8.69568050e-01 4.22703475e-01
2.75665730e-01 2.81953532e-02 -1.20084596e+00 7.20250189e-01
1.22532082e+00 -2.16135710e-01 -6.84534252e-01 4.58616614e-01
8.81546080e-01 -2.90377170e-01 -7.64218748e-01 2.29480147e-01
9.29206371e-01 -7.31283784e-01 6.63891256e-01 -2.78396934e-01
6.53662309e-02 -2.40261964e-02 1.16610169e-01 -1.60402036e+00
-5.23919821e-01 -4.85120118e-01 4.16525275e-01 1.29243708e+00
3.81437182e-01 -7.95035303e-01 5.91061473e-01 1.05639338e+00
-3.78182441e-01 -5.25010347e-01 -2.74559528e-01 -1.01415455e+00
1.81713670e-01 -2.78052211e-01 1.20240726e-01 8.98635149e-01
6.10775113e-01 4.47743148e-01 1.21777102e-01 7.70697221e-02
2.32254520e-01 -5.37203908e-01 1.30875278e+00 -1.37677872e+00
-3.49414438e-01 -1.62278935e-01 -8.88308063e-02 -8.59312236e-01
1.71456635e-01 -6.09922945e-01 3.70818585e-01 -1.45133233e+00
1.02199979e-01 -7.61456609e-01 8.26570913e-02 6.95255458e-01
-2.96397004e-02 -5.21728456e-01 1.12981558e-01 5.31507790e-01
-5.27191758e-01 5.03478885e-01 9.86164689e-01 1.58783272e-01
-9.58982646e-01 1.06133647e-01 -5.93636453e-01 7.44357467e-01
1.00326669e+00 -3.41506094e-01 -1.12553096e+00 -3.42988133e-01
-3.08894040e-03 -1.94880441e-01 2.36348271e-01 -1.42036438e+00
6.39283180e-01 -5.29630065e-01 1.16521269e-01 7.74244294e-02
9.98142064e-02 -1.15365589e+00 4.03727561e-01 6.66686416e-01
-2.09870651e-01 1.47880048e-01 6.27567112e-01 4.82567161e-01
2.82853872e-01 -3.89894813e-01 4.60025758e-01 -3.23185623e-01
-9.31112647e-01 -1.31332114e-01 -1.00886691e+00 -4.25049990e-01
1.41120315e+00 -2.89014786e-01 5.19571565e-02 -7.19420195e-01
-6.36544585e-01 5.56451917e-01 5.76051593e-01 7.19036937e-01
2.89035410e-01 -1.00517762e+00 -1.65800929e-01 -2.83597857e-01
1.24033295e-01 8.64354055e-03 2.14045838e-01 7.07282424e-01
-5.18187881e-01 3.28287214e-01 -4.76953804e-01 -4.86157894e-01
-1.25930464e+00 3.42877328e-01 3.79905909e-01 3.10957611e-01
-5.85733891e-01 4.75025088e-01 3.07077654e-02 -1.04372215e+00
6.46889150e-01 1.49875149e-01 -2.86180824e-01 2.54502803e-01
1.67366520e-01 8.28966022e-01 -2.07565337e-01 -2.84839779e-01
2.60479510e-01 2.16059417e-01 -1.16235204e-01 -5.49325287e-01
1.38558853e+00 -3.71793777e-01 9.26222280e-02 1.27054775e+00
6.96995258e-01 -2.39672363e-01 -1.34295034e+00 -2.49278657e-02
3.71303439e-01 -4.30133253e-01 -6.05327487e-01 -8.91325057e-01
-4.20827478e-01 3.67942095e-01 1.07009840e+00 9.85875428e-02
5.56603491e-01 2.38885917e-02 4.52908576e-01 1.14010501e+00
9.49530303e-01 -1.38554943e+00 7.88719296e-01 6.02654278e-01
1.28656530e+00 -1.33189428e+00 -1.41097950e-02 -2.76106000e-01
-6.48226023e-01 1.04397821e+00 1.14532447e+00 3.04655433e-01
7.27835596e-01 -2.43762419e-01 3.14119905e-01 -4.93940115e-01
-3.81873667e-01 3.36448133e-01 -2.38478258e-01 7.14213729e-01
8.33539665e-01 1.35562733e-01 -3.70987326e-01 3.32740456e-01
-5.03110111e-01 3.37809443e-01 8.05909514e-01 1.50371683e+00
-6.73427522e-01 -1.04250193e+00 -5.83630860e-01 1.64278209e-01
3.37638050e-01 7.54993260e-01 -2.93338448e-01 1.15845764e+00
1.33153915e-01 1.29028344e+00 -1.13515288e-01 -4.90926147e-01
8.43991518e-01 3.90211403e-01 5.59572279e-01 -1.13144624e+00
-5.99415958e-01 -3.17937613e-01 -5.03481030e-02 -3.04317713e-01
-2.33715922e-01 -1.20046389e+00 -1.64619994e+00 -4.13601547e-01
-1.92588687e-01 1.96084157e-01 6.70393407e-01 7.25941420e-01
1.21100023e-01 4.92982745e-01 5.21250129e-01 -1.10718727e+00
-5.84552884e-01 -1.33225596e+00 -5.29905200e-01 2.86774397e-01
2.76518031e-03 -8.78381133e-01 -3.31726938e-01 1.63107395e-01] | [4.498420238494873, 1.0794414281845093] |
b6b8ec19-a370-4719-bf04-c7a938be50a5 | unsupervised-meta-learning-via-few-shot-1 | 2303.00996 | null | https://arxiv.org/abs/2303.00996v1 | https://arxiv.org/pdf/2303.00996v1.pdf | Unsupervised Meta-Learning via Few-shot Pseudo-supervised Contrastive Learning | Unsupervised meta-learning aims to learn generalizable knowledge across a distribution of tasks constructed from unlabeled data. Here, the main challenge is how to construct diverse tasks for meta-learning without label information; recent works have proposed to create, e.g., pseudo-labeling via pretrained representations or creating synthetic samples via generative models. However, such a task construction strategy is fundamentally limited due to heavy reliance on the immutable pseudo-labels during meta-learning and the quality of the representations or the generated samples. To overcome the limitations, we propose a simple yet effective unsupervised meta-learning framework, coined Pseudo-supervised Contrast (PsCo), for few-shot classification. We are inspired by the recent self-supervised learning literature; PsCo utilizes a momentum network and a queue of previous batches to improve pseudo-labeling and construct diverse tasks in a progressive manner. Our extensive experiments demonstrate that PsCo outperforms existing unsupervised meta-learning methods under various in-domain and cross-domain few-shot classification benchmarks. We also validate that PsCo is easily scalable to a large-scale benchmark, while recent prior-art meta-schemes are not. | ['Jinwoo Shin', 'Hankook Lee', 'Huiwon Jang'] | 2023-03-02 | unsupervised-meta-learning-via-few-shot | https://openreview.net/forum?id=i0Fgim2tB0i | https://openreview.net/pdf?id=i0Fgim2tB0i | 6th-workshop-on-meta-learning-at-neurips-2022 | ['cross-domain-few-shot'] | ['computer-vision'] | [ 4.15560097e-01 -6.20204769e-03 -4.06556994e-01 -3.95002425e-01
-1.03261244e+00 -2.71444827e-01 7.72313058e-01 -2.08794072e-01
-2.03469485e-01 1.04688060e+00 2.71045178e-01 2.08671302e-01
2.71601416e-03 -6.45092785e-01 -7.34738469e-01 -7.21669316e-01
6.70817792e-01 7.25997031e-01 1.00827720e-02 -9.28118005e-02
1.67688623e-01 -2.69901007e-01 -1.90262973e+00 4.13674176e-01
1.11421585e+00 7.21037686e-01 2.79085100e-01 2.98631579e-01
-2.95433909e-01 1.07262552e+00 -6.72160923e-01 -4.41269577e-01
2.12279424e-01 -9.23857391e-01 -7.98331678e-01 6.58129275e-01
2.68048465e-01 -9.39320549e-02 -6.09920882e-02 1.01867366e+00
5.97653747e-01 4.39183265e-01 1.09386241e+00 -1.33465159e+00
-8.73641193e-01 7.74699271e-01 -3.91063958e-01 -1.17435962e-01
-1.28902048e-01 2.73535907e-01 9.27964330e-01 -9.64011073e-01
8.18356812e-01 9.90091801e-01 6.62572265e-01 1.02365506e+00
-1.33525288e+00 -5.77542186e-01 8.83092284e-02 9.81940106e-02
-1.01173091e+00 -6.11747384e-01 9.17285562e-01 -5.74138224e-01
5.63575864e-01 -2.94323172e-02 2.76049316e-01 1.80652869e+00
-1.60723671e-01 9.03909028e-01 1.55158222e+00 -5.95334947e-01
7.40353048e-01 2.30625093e-01 2.30253220e-01 5.58172882e-01
1.22864723e-01 -4.99643870e-02 -6.26413167e-01 -3.22116077e-01
4.55779463e-01 2.96057761e-01 4.49154638e-02 -6.01840436e-01
-1.35999274e+00 8.22659194e-01 1.04426339e-01 1.62542507e-01
-2.15319440e-01 -2.90965512e-02 6.16692662e-01 5.01302779e-01
8.70000362e-01 6.56200945e-01 -4.46093112e-01 -1.07218772e-01
-1.06622207e+00 8.05600733e-02 7.83127069e-01 1.19765019e+00
9.48113561e-01 2.71168143e-01 -5.30180573e-01 1.22591186e+00
4.20571938e-02 1.31534502e-01 1.06887579e+00 -9.73856688e-01
5.58668554e-01 3.89519721e-01 2.69374490e-01 -2.25944385e-01
-1.11568272e-01 -6.27666473e-01 -8.31070364e-01 1.55624272e-02
2.28594810e-01 -3.28745097e-01 -1.33191657e+00 1.71087670e+00
3.37363750e-01 4.59963441e-01 3.16672862e-01 6.60155058e-01
9.37568903e-01 4.85868365e-01 3.16995680e-01 -3.44249398e-01
8.80607605e-01 -1.64090109e+00 -5.78629196e-01 -3.41252357e-01
5.79392135e-01 -3.41075867e-01 1.16936970e+00 1.57864481e-01
-9.67383623e-01 -9.13804412e-01 -1.06661439e+00 1.52596325e-01
-3.44173312e-01 9.29881930e-02 4.35859472e-01 5.35890698e-01
-7.74844468e-01 7.98735678e-01 -6.25984669e-01 -3.19329619e-01
5.62086463e-01 -1.09415561e-01 -4.20924537e-02 -2.27544814e-01
-1.06165051e+00 6.89392567e-01 7.76335597e-01 -4.16540831e-01
-1.44794858e+00 -7.74041712e-01 -8.83379400e-01 4.96364832e-02
5.30064166e-01 -9.97974455e-01 1.32605588e+00 -1.14770770e+00
-1.68951905e+00 8.96438599e-01 4.71821316e-02 -3.75212342e-01
6.20165229e-01 -4.96125501e-03 -3.02205652e-01 -1.15642369e-01
3.09805244e-01 8.33681703e-01 1.24401963e+00 -1.53671789e+00
-5.74610353e-01 -1.64216816e-01 -9.89826471e-02 2.80979812e-01
-4.48745549e-01 -4.68511045e-01 -2.02800900e-01 -8.42368901e-01
-9.89928246e-02 -1.03503847e+00 -4.24295157e-01 -4.87095147e-01
-4.09612954e-01 -2.77982444e-01 4.80024785e-01 -1.10692337e-01
8.45657885e-01 -2.03200102e+00 2.22995445e-01 -3.92234296e-01
2.91308701e-01 4.51636851e-01 -4.38209742e-01 5.17463148e-01
5.53152077e-02 -1.19731344e-01 -4.77009356e-01 -8.13151717e-01
1.63597167e-01 2.26405710e-01 -4.49162394e-01 1.57608718e-01
1.48625344e-01 1.13882053e+00 -1.36507678e+00 -5.61050296e-01
1.47002816e-01 1.18367910e-01 -2.73253798e-01 3.88616025e-01
-5.40870011e-01 7.11719871e-01 -4.26790118e-01 7.01971114e-01
4.00665492e-01 -7.06371784e-01 1.64645419e-01 1.45699710e-01
1.89187184e-01 1.28293842e-01 -8.70839238e-01 2.14299560e+00
-6.27468348e-01 2.02874780e-01 -4.47389543e-01 -1.26266193e+00
1.05259407e+00 3.31996143e-01 5.04269004e-01 -5.25223553e-01
2.09876880e-01 3.80127460e-01 -4.19081897e-01 -4.61529225e-01
5.21451771e-01 -4.01264399e-01 -2.48891115e-03 7.57435739e-01
6.54986203e-01 -4.32261936e-02 3.84569854e-01 1.19062088e-01
1.02385890e+00 4.40014869e-01 2.87769318e-01 -9.78700891e-02
5.75618595e-02 2.66968608e-01 6.58883691e-01 1.01900637e+00
-2.45462522e-01 1.00136364e+00 1.47672519e-01 -2.70083338e-01
-1.23249686e+00 -1.06080508e+00 -9.81380865e-02 1.51888406e+00
-5.63680977e-02 -2.44011819e-01 -6.97596312e-01 -1.04079950e+00
-1.36067435e-01 8.06352735e-01 -7.93223023e-01 -2.95849055e-01
-3.71081270e-02 -8.52727413e-01 2.59770930e-01 4.68306720e-01
5.24752915e-01 -1.29970419e+00 -2.99171269e-01 4.57633138e-01
-2.01506421e-01 -9.92289722e-01 -2.77474463e-01 4.31390345e-01
-1.09626245e+00 -8.97571683e-01 -1.11704099e+00 -9.89219487e-01
8.94404173e-01 5.50410628e-01 1.21246827e+00 -4.24624681e-01
-7.97881633e-02 3.35394681e-01 -6.57845855e-01 -2.00352862e-01
-5.56077421e-01 2.96601832e-01 -1.21241435e-02 1.31615952e-01
3.44234675e-01 -7.25279868e-01 -4.04621392e-01 3.42212141e-01
-9.53156173e-01 3.70486081e-01 5.62528431e-01 1.37414837e+00
7.49283254e-01 -5.87890036e-02 1.07315075e+00 -1.71936560e+00
4.71832454e-01 -9.91334260e-01 -1.69921443e-01 4.31002110e-01
-7.66920507e-01 1.66036919e-01 8.38917196e-01 -7.03739166e-01
-1.37639010e+00 7.41830990e-02 1.98755428e-01 -7.53604949e-01
-2.18693927e-01 4.30909336e-01 -2.05771104e-02 2.71941423e-01
1.06868052e+00 4.80611414e-01 -2.26932347e-01 -5.02683878e-01
5.74917912e-01 7.06642389e-01 3.72039407e-01 -8.08669150e-01
7.06514955e-01 6.26067877e-01 -2.58597463e-01 -5.63599348e-01
-1.58504796e+00 -6.48938000e-01 -8.45376551e-01 -7.26584047e-02
5.99408388e-01 -1.26457930e+00 1.91048920e-01 5.05841017e-01
-7.58141935e-01 -7.06112146e-01 -5.71162760e-01 3.19945633e-01
-7.85047412e-01 1.50647298e-01 -5.05838037e-01 -5.92004061e-01
-4.04584944e-01 -9.25979853e-01 1.07525551e+00 1.06451124e-01
-9.70249921e-02 -1.13789153e+00 3.82564336e-01 5.77847242e-01
4.16758627e-01 3.01647574e-01 6.69117749e-01 -7.43692756e-01
-2.00980321e-01 1.04633979e-01 -1.65388305e-02 4.75127310e-01
3.89775097e-01 -3.98264110e-01 -1.22883856e+00 -4.29132670e-01
7.01142475e-02 -1.17069459e+00 1.14472938e+00 2.41569161e-01
1.18325627e+00 -1.37514561e-01 -2.00427398e-01 6.09147012e-01
1.34427762e+00 -1.74800515e-01 3.34030867e-01 4.11071420e-01
7.53178716e-01 6.32126629e-01 7.32924759e-01 6.25916719e-01
3.66254628e-01 2.62119859e-01 -3.79012451e-02 1.80738166e-01
-3.56342077e-01 -4.74171430e-01 2.58697897e-01 1.10405493e+00
1.68109071e-02 -1.76459014e-01 -6.30560815e-01 6.90240920e-01
-2.21131730e+00 -1.01078403e+00 2.30109319e-01 2.06296325e+00
1.06762600e+00 1.45289198e-01 1.44726753e-01 -7.40680993e-02
8.37018907e-01 3.23991746e-01 -8.89359891e-01 5.06645665e-02
-2.45357044e-02 2.41667941e-01 8.10065120e-02 -1.07792750e-01
-1.19857335e+00 1.05464149e+00 6.06890345e+00 9.60777104e-01
-9.39797163e-01 6.77332938e-01 7.33321249e-01 -1.94486722e-01
-3.41889560e-01 -1.22866742e-02 -8.09329808e-01 5.78372478e-01
9.32390928e-01 -1.78449273e-01 2.72471040e-01 1.18097544e+00
-3.21558535e-01 1.82004437e-01 -1.10304165e+00 9.60677743e-01
3.47447127e-01 -1.23998868e+00 1.90226547e-02 -2.44187489e-01
1.75430143e+00 1.67244643e-01 1.62427276e-01 8.86895955e-01
7.69618869e-01 -6.47803783e-01 6.78019822e-01 3.94641578e-01
7.90379465e-01 -5.32817900e-01 4.06002313e-01 5.86820781e-01
-8.95014644e-01 -1.09150104e-01 -8.11378360e-01 2.99581867e-02
3.22860219e-02 6.99823201e-01 -7.51219988e-01 5.27589619e-01
3.51384193e-01 8.89090717e-01 -6.25519812e-01 9.26808834e-01
-3.57864380e-01 7.48954654e-01 3.35942268e-01 6.52325004e-02
3.43381852e-01 -1.70536451e-02 3.21153760e-01 1.04631722e+00
2.38737762e-01 -1.52791813e-01 4.37046707e-01 9.85743284e-01
-3.24505448e-01 -7.00781122e-02 -4.63570565e-01 -2.22921312e-01
5.27216852e-01 1.36662793e+00 -6.94206178e-01 -7.67082989e-01
-3.93026739e-01 9.86273825e-01 7.06286073e-01 4.36168373e-01
-7.76839852e-01 -2.39276528e-01 1.74850240e-01 -1.36510372e-01
3.04902107e-01 -2.81598605e-02 -2.64239162e-01 -1.66656339e+00
-1.43915594e-01 -9.05777872e-01 4.78586793e-01 -6.52160466e-01
-1.78512096e+00 5.59124053e-01 -4.90399040e-02 -1.65210056e+00
-4.07622606e-01 -1.96307614e-01 -5.13613760e-01 5.70505142e-01
-1.73520041e+00 -1.30181396e+00 -4.97038335e-01 5.34622967e-01
1.12019455e+00 -5.10153711e-01 8.17271173e-01 1.09165115e-02
-6.37065649e-01 4.57956702e-01 5.19958913e-01 -2.85628214e-02
1.01721668e+00 -1.33888614e+00 5.46458006e-01 5.32392740e-01
3.46537352e-01 4.78857040e-01 4.76048172e-01 -7.12646544e-01
-1.17654443e+00 -1.55726624e+00 4.72185016e-01 -5.28897583e-01
5.85486174e-01 -3.81955832e-01 -9.14702654e-01 6.71211243e-01
1.63277850e-01 2.85644323e-01 8.95605564e-01 2.09452778e-01
-5.42640209e-01 5.58573790e-02 -9.67066169e-01 4.70107019e-01
1.31717074e+00 -4.40945715e-01 -8.56274664e-01 5.80639482e-01
7.37089515e-01 -2.85265744e-01 -6.81556940e-01 3.78830284e-01
2.50096112e-01 -8.83428514e-01 7.12158620e-01 -8.17163646e-01
8.49116743e-01 -1.38495386e-01 8.12717900e-03 -1.85522091e+00
-4.55594629e-01 -5.39749980e-01 -3.86038750e-01 1.48831391e+00
4.33270872e-01 -4.76272255e-01 9.29508448e-01 4.25279200e-01
-3.91474903e-01 -4.97462392e-01 -5.17902255e-01 -1.10035026e+00
-4.50593121e-02 -1.35732040e-01 5.21885276e-01 1.34847605e+00
-3.21870595e-02 6.26247764e-01 -7.47858763e-01 -2.79529661e-01
1.08940876e+00 4.49011892e-01 9.37629521e-01 -1.27291501e+00
-5.89543819e-01 -1.63322553e-01 1.25754446e-01 -7.84637332e-01
5.47854483e-01 -1.16802335e+00 2.38522753e-01 -1.49628174e+00
6.49745584e-01 -6.37902200e-01 -6.62374377e-01 5.85442603e-01
-4.68352079e-01 3.16011935e-01 1.76997026e-04 6.45960212e-01
-1.17535424e+00 8.79331112e-01 1.53046906e+00 -3.06159645e-01
-3.35214794e-01 1.63364075e-02 -7.38768339e-01 5.50235510e-01
7.94808149e-01 -7.34204412e-01 -6.80490851e-01 -4.64866757e-01
-7.74243101e-02 -4.25660945e-02 1.07632399e-01 -1.23436487e+00
1.61669478e-01 -2.97186017e-01 3.61561358e-01 -2.96121538e-01
3.69202018e-01 -3.60365301e-01 6.58811331e-02 2.54945308e-01
-6.09001994e-01 -4.77763683e-01 -3.03908437e-01 7.85523057e-01
-2.67051965e-01 -5.96627295e-01 9.16806817e-01 -5.79648495e-01
-9.98826325e-01 4.23180014e-01 7.44859949e-02 5.97185731e-01
1.04279506e+00 -2.11134199e-02 -4.79573041e-01 -9.71897244e-02
-7.94737637e-01 1.85675681e-01 5.08630931e-01 4.43478286e-01
3.56539756e-01 -1.53390157e+00 -7.37113178e-01 -1.33335009e-01
5.64080656e-01 7.83920437e-02 4.64489371e-01 4.83053535e-01
-2.84531172e-02 1.40723094e-01 -3.79775405e-01 -5.81676602e-01
-6.04249120e-01 7.88405597e-01 -7.01943710e-02 -3.75389040e-01
-5.29125929e-01 7.13066638e-01 1.10308431e-01 -7.15351939e-01
1.72229871e-01 8.06438252e-02 -1.08422644e-01 2.60067105e-01
5.44740617e-01 5.67389250e-01 -3.40630524e-02 -2.53907621e-01
2.23344237e-01 1.80270299e-01 -2.53141612e-01 -2.12100986e-02
1.44991648e+00 -6.59978986e-02 2.29700580e-01 9.50686753e-01
1.09873128e+00 -4.88826245e-01 -1.74343383e+00 -5.72221160e-01
-1.67535860e-02 -3.72777224e-01 -2.15779886e-01 -6.98178768e-01
-8.33966196e-01 7.71078706e-01 2.06235319e-01 1.75901614e-02
8.16914916e-01 -7.83294067e-02 8.03305030e-01 6.13688111e-01
6.23479664e-01 -1.44796097e+00 6.93752050e-01 5.87657154e-01
4.33498651e-01 -1.71180034e+00 -6.17197901e-02 -8.80941600e-02
-1.12629783e+00 8.56470168e-01 8.77165198e-01 -1.23958170e-01
4.22364473e-01 -1.85492784e-01 1.13256142e-01 4.90681157e-02
-9.20479476e-01 -4.17416304e-01 2.19938755e-01 6.04155242e-01
2.68619567e-01 1.93897225e-02 -1.23473026e-01 6.65294647e-01
8.39573070e-02 1.95874318e-01 3.76536250e-01 1.17307520e+00
-5.55137873e-01 -1.19233859e+00 -9.74574089e-02 6.16519928e-01
-1.38754144e-01 1.90769439e-03 -2.57622600e-01 4.02443379e-01
1.13548234e-01 9.31622744e-01 -1.85022220e-01 -2.98389226e-01
2.89252456e-02 4.11590695e-01 4.49776918e-01 -1.26922059e+00
-3.61586004e-01 -6.24697991e-02 -4.88670953e-02 -5.04157059e-02
-7.54900753e-01 -4.50580865e-01 -7.72033811e-01 1.72552884e-01
-3.14871490e-01 2.89610267e-01 3.18302691e-01 1.07823348e+00
4.57394242e-01 6.30971730e-01 9.50645268e-01 -9.55265164e-01
-8.75128150e-01 -1.29790652e+00 -7.48418152e-01 7.96133637e-01
-4.24381979e-02 -9.30248380e-01 -4.03902829e-01 2.87088305e-01] | [9.95498275756836, 3.0043439865112305] |
bdbeac6d-cf06-4f9f-a751-a1482bb8ccf7 | unsupervised-brain-lesion-segmentation-from | 1811.09655 | null | http://arxiv.org/abs/1811.09655v1 | http://arxiv.org/pdf/1811.09655v1.pdf | Unsupervised brain lesion segmentation from MRI using a convolutional autoencoder | Lesions that appear hyperintense in both Fluid Attenuated Inversion Recovery
(FLAIR) and T2-weighted magnetic resonance images (MRIs) of the human brain are
common in the brains of the elderly population and may be caused by ischemia or
demyelination. Lesions are biomarkers for various neurodegenerative diseases,
making accurate quantification of them important for both disease diagnosis and
progression. Automatic lesion detection using supervised learning requires
manually annotated images, which can often be impractical to acquire.
Unsupervised lesion detection, on the other hand, does not require any manual
delineation; however, these methods can be challenging to construct due to the
variability in lesion load, placement of lesions, and voxel intensities. Here
we present a novel approach to address this problem using a convolutional
autoencoder, which learns to segment brain lesions as well as the white matter,
gray matter, and cerebrospinal fluid by reconstructing FLAIR images as conical
combinations of softmax layer outputs generated from the corresponding T1, T2,
and FLAIR images. Some of the advantages of this model are that it accurately
learns to segment lesions regardless of lesion load, and it can be used to
quickly and robustly segment new images that were not in the training set.
Comparisons with state-of-the-art segmentation methods evaluated on ground
truth manual labels indicate that the proposed method works well for generating
accurate lesion segmentations without the need for manual annotations. | ['Lotta M. Ellingsen', 'Vilmundur Gudnason', 'Hans E. Atlason', 'Sigurdur Sigurdsson', 'Askell Love'] | 2018-11-23 | null | null | null | null | ['brain-lesion-segmentation-from-mri'] | ['medical'] | [ 3.75253171e-01 -9.59881693e-02 3.04102525e-02 -4.31253165e-01
-4.91743356e-01 -2.53856272e-01 3.44804525e-01 1.61278397e-01
-8.20811868e-01 8.64453554e-01 7.72628486e-02 -1.78460523e-01
-1.33398309e-01 -6.04586601e-01 -4.03873503e-01 -7.97623515e-01
-3.41016293e-01 8.53468597e-01 5.40291250e-01 1.67602643e-01
1.89392298e-01 5.65495431e-01 -1.20266724e+00 1.31047517e-01
1.30037749e+00 7.36147046e-01 6.52794719e-01 3.02831829e-01
-1.17281176e-01 7.31104851e-01 -5.15835047e-01 1.38288617e-01
-5.27880304e-02 -3.42584133e-01 -9.39875245e-01 2.43523166e-01
3.73951167e-01 -6.38740957e-01 -1.32722914e-01 1.12745166e+00
5.56846619e-01 1.83091536e-01 7.55948246e-01 -6.22196257e-01
-5.51161647e-01 6.21586442e-01 -5.21914899e-01 7.38017797e-01
-3.38369012e-01 1.74928740e-01 2.43902043e-01 -7.09596753e-01
6.83799684e-01 7.50099719e-01 5.82337856e-01 3.49186748e-01
-1.00007570e+00 -5.37439287e-01 6.52007526e-04 4.75388885e-01
-9.71761107e-01 -1.86597615e-01 3.39303374e-01 -1.06505144e+00
9.22897577e-01 1.75202917e-02 6.44538224e-01 6.12168550e-01
3.91167402e-01 6.77757442e-01 1.33922386e+00 -2.85307646e-01
3.02639723e-01 -4.84727591e-01 3.10728818e-01 6.26742423e-01
2.96285659e-01 -8.27796087e-02 2.70173877e-01 -4.57695536e-02
8.52157593e-01 1.09954178e-01 -5.32577753e-01 -2.94500947e-01
-1.53138781e+00 6.99377954e-01 7.62373984e-01 7.97698975e-01
-8.78620863e-01 -4.46192808e-02 4.20449078e-01 -2.53106058e-01
3.97785664e-01 3.58881325e-01 2.34616101e-02 2.84549505e-01
-1.49901330e+00 -2.25379542e-02 -2.49189045e-03 -1.82239786e-02
4.55284119e-01 1.46264642e-01 6.00297116e-02 1.06991816e+00
1.24974847e-01 3.84698749e-01 1.27965128e+00 -8.73030961e-01
3.38253826e-02 5.38562119e-01 -9.74480361e-02 -5.17638624e-01
-8.64754498e-01 -4.67096388e-01 -8.02720666e-01 5.27060509e-01
4.57215071e-01 -3.30662578e-01 -1.44378948e+00 1.39857161e+00
1.25986829e-01 4.12716251e-03 -3.32004845e-01 1.22836530e+00
4.39345449e-01 2.34290808e-01 2.55177945e-01 -1.36114284e-01
1.01363993e+00 -8.43673170e-01 -5.95011353e-01 -5.12151420e-01
7.25951910e-01 -2.32925475e-01 5.90327799e-01 -2.43905615e-02
-1.10119402e+00 -1.49220794e-01 -1.06615186e+00 1.97559029e-01
-2.52872616e-01 9.15502235e-02 4.83557880e-01 4.83409941e-01
-1.10966110e+00 5.19428372e-01 -1.25534225e+00 -1.02147065e-01
5.97531259e-01 5.20099044e-01 -4.74352360e-01 -2.78857499e-02
-1.10256231e+00 1.46354330e+00 6.51235402e-01 2.43963405e-01
-5.56453049e-01 -7.13545024e-01 -5.90453327e-01 -2.01589301e-01
6.58435673e-02 -3.47066283e-01 1.03841186e+00 -1.02672625e+00
-1.03292894e+00 9.64618683e-01 -2.19808612e-02 -6.84123695e-01
4.33631331e-01 9.90089551e-02 -3.57452601e-01 4.65444237e-01
1.85515672e-01 8.37111235e-01 6.26402855e-01 -1.00897074e+00
-4.44256455e-01 -7.22329736e-01 -2.58343101e-01 8.17510486e-02
1.83303908e-01 2.65154690e-01 1.86180905e-01 -5.15382707e-01
3.11568886e-01 -8.13706398e-01 -4.22764689e-01 -1.03508145e-01
-3.40604365e-01 1.32149890e-01 7.32234716e-01 -1.41836011e+00
6.85585797e-01 -1.77111995e+00 -3.13488320e-02 3.76743168e-01
5.77694893e-01 4.84621495e-01 1.09650590e-01 -4.96097028e-01
-4.39348698e-01 4.71566468e-02 -8.65377247e-01 2.19505772e-01
-2.06496924e-01 3.93061899e-02 2.43617460e-01 7.92942166e-01
-6.42317086e-02 1.04011178e+00 -8.61100972e-01 -3.52755964e-01
3.46284270e-01 5.72834551e-01 -9.28034037e-02 -5.99570461e-02
4.18183357e-02 8.24612856e-01 -1.25589967e-02 3.19720626e-01
3.97331387e-01 -2.49050766e-01 2.68946737e-01 -1.55399427e-01
-1.32751837e-01 -3.08002322e-03 -9.08930659e-01 1.25774646e+00
-8.84044543e-02 6.94759667e-01 6.80787936e-02 -1.44329715e+00
6.15350842e-01 4.66651529e-01 9.11817908e-01 -8.84663165e-01
4.82857347e-01 6.19016349e-01 6.32330656e-01 -6.18242979e-01
-2.22880945e-01 -4.30175900e-01 5.11282146e-01 9.05427277e-01
3.62243354e-02 1.37109607e-01 5.40520608e-01 -3.72022450e-01
8.26714456e-01 -2.16114566e-01 -1.69966388e-02 -2.04347447e-01
5.14876187e-01 -1.75033465e-01 3.93872142e-01 4.90358621e-01
-3.67447853e-01 6.17020547e-01 9.12729651e-02 -4.63927954e-01
-1.01177454e+00 -1.17795885e+00 -3.68886888e-01 7.15945303e-01
-2.03288689e-01 5.50749302e-01 -1.18749917e+00 -4.72890139e-01
-3.05999547e-01 3.99506927e-01 -6.33174002e-01 -1.63035430e-02
-9.40042555e-01 -1.28155577e+00 3.01283270e-01 7.24727273e-01
5.18151641e-01 -1.23881972e+00 -9.78460670e-01 3.47968608e-01
-4.48178053e-01 -1.01188958e+00 -4.15756106e-01 6.85469508e-02
-1.22055113e+00 -1.30057549e+00 -1.62091422e+00 -1.11957729e+00
9.43789780e-01 -1.09957181e-01 7.19638705e-01 2.46178329e-01
-5.24376988e-01 1.66863680e-01 -6.54615164e-02 -1.41343012e-01
-2.99670517e-01 1.59534421e-02 8.46158564e-02 1.17341466e-01
1.72547519e-01 -5.99230170e-01 -7.55372643e-01 2.11744964e-01
-1.12656820e+00 2.53761709e-02 6.15202725e-01 6.27674997e-01
8.77391040e-01 -3.37050147e-02 6.71927631e-01 -6.52849793e-01
8.65435004e-01 -5.16865730e-01 -2.70428300e-01 3.49345535e-01
-5.30476630e-01 -4.66937795e-02 1.11037798e-01 -4.68314648e-01
-8.28644335e-01 -1.55168995e-01 -3.59983258e-02 -6.95651174e-02
-4.24283564e-01 5.88910341e-01 2.97850102e-01 -3.26786280e-01
7.11496890e-01 2.08314374e-01 3.92816126e-01 -3.17600161e-01
1.11098155e-01 6.85979009e-01 9.17527318e-01 -1.47561431e-01
2.13749960e-01 5.04076719e-01 -2.52422810e-01 -6.96854889e-01
-3.66439432e-01 -4.27631378e-01 -1.11353958e+00 -3.37524682e-01
1.05761170e+00 -3.14201117e-01 -6.36168569e-02 6.34439647e-01
-8.08687508e-01 -7.06339955e-01 -3.43956985e-02 8.96035373e-01
-4.94717836e-01 5.38460970e-01 -6.38594747e-01 -2.34898761e-01
-6.75207317e-01 -1.59666753e+00 4.38587785e-01 2.55608439e-01
-2.74205834e-01 -1.20633161e+00 -1.43270984e-01 4.69861060e-01
8.10108304e-01 4.38696057e-01 1.41170049e+00 -5.61099648e-01
-1.38510913e-01 -3.26167345e-01 -2.62019664e-01 4.84205395e-01
3.19003552e-01 -5.55006742e-01 -3.35600227e-01 -1.64359510e-01
8.47034678e-02 -1.95944875e-01 8.63071680e-01 8.96315515e-01
1.00863194e+00 1.58693552e-01 -1.57188386e-01 2.84404755e-01
1.05015445e+00 4.53640103e-01 5.37577689e-01 5.07873774e-01
5.45154691e-01 6.25838459e-01 -1.70573801e-01 -1.31729051e-01
2.27661446e-01 4.49177444e-01 3.01165462e-01 -2.95689225e-01
-5.38730860e-01 5.86826026e-01 6.20056912e-02 6.86964750e-01
-3.06320995e-01 3.43987823e-01 -1.23552763e+00 7.91862249e-01
-1.36257827e+00 -8.80538821e-01 -2.15562701e-01 2.04094982e+00
9.35367286e-01 1.49873493e-03 2.65271753e-01 2.56119877e-01
1.08906794e+00 -9.17843133e-02 -8.98921728e-01 -1.19053312e-01
-1.90204859e-01 3.90174538e-01 4.74430352e-01 6.85810506e-01
-9.96428907e-01 5.29587090e-01 6.55122948e+00 1.28016919e-02
-1.75079405e+00 4.38731670e-01 5.76006114e-01 -9.81375277e-02
5.42417131e-02 -4.21303093e-01 -9.27874222e-02 6.27084374e-01
8.62295330e-01 1.55866295e-01 5.25726080e-01 1.94149077e-01
2.31419921e-01 -3.79804760e-01 -7.03855455e-01 7.22957134e-01
1.75291374e-01 -1.12812483e+00 -3.00056905e-01 -2.37467140e-01
6.58015847e-01 5.03644943e-01 1.06427789e-01 3.24787013e-02
-1.66071489e-01 -1.14680421e+00 6.41091943e-01 7.39619136e-01
8.52208197e-01 -4.34053034e-01 8.45629454e-01 2.15930820e-01
-4.79734361e-01 -2.08943523e-02 2.68856110e-03 2.97389835e-01
4.51100707e-01 6.44461632e-01 -9.69156444e-01 -1.69679180e-01
4.88212198e-01 9.60721299e-02 -5.37721992e-01 1.69234908e+00
-2.38386512e-01 5.93415916e-01 -2.55121738e-01 4.11039203e-01
3.95138651e-01 -7.28859678e-02 4.50668931e-01 1.00553071e+00
3.20421934e-01 -8.33405778e-02 2.77879834e-01 1.07692873e+00
4.16037403e-02 1.97027013e-01 1.69497892e-01 1.23191722e-01
-1.45293050e-03 9.91846383e-01 -1.15258718e+00 -3.58229011e-01
-3.12323987e-01 8.06811273e-01 1.11995876e-01 5.58311462e-01
-3.91376913e-01 -2.57363558e-01 -6.19392209e-02 4.43670511e-01
-5.21231629e-02 -2.76453942e-01 -3.57994437e-01 -9.15113628e-01
5.68320490e-02 -6.75222099e-01 6.39982447e-02 -6.51903033e-01
-9.96729195e-01 7.67970741e-01 5.68514802e-02 -5.56660235e-01
-4.36962903e-01 -6.75101042e-01 -5.36385536e-01 1.07062066e+00
-1.51263857e+00 -5.94144881e-01 -2.66807109e-01 4.54779029e-01
1.75763234e-01 -1.26361232e-02 7.30116785e-01 4.65109795e-01
-6.04840934e-01 1.61243618e-01 1.78190157e-01 3.27764302e-01
5.76197565e-01 -1.32550752e+00 7.20795244e-02 9.00613070e-01
-2.70065457e-01 4.58897293e-01 3.01303625e-01 -9.00546134e-01
-3.83240134e-01 -1.05454254e+00 6.84148133e-01 3.71114135e-01
5.55060446e-01 2.86616504e-01 -1.20859635e+00 7.03696549e-01
-9.37281176e-02 3.02942157e-01 6.50174797e-01 -3.74713957e-01
7.82908797e-02 3.75461102e-01 -1.38084853e+00 3.44962299e-01
3.87537390e-01 -4.97492760e-01 -8.04550707e-01 4.95533377e-01
-3.62078547e-02 -4.93070126e-01 -9.43052888e-01 4.47042286e-01
4.02388424e-01 -7.54835844e-01 9.71335053e-01 -3.43246460e-01
1.46580040e-01 -1.01815052e-01 3.28018516e-01 -1.45045638e+00
-4.65292245e-01 1.25720486e-01 -1.30117580e-01 3.93445581e-01
2.34993085e-01 -7.60138869e-01 7.80534863e-01 8.61997843e-01
-3.47090542e-01 -7.17279255e-01 -7.62756586e-01 -5.41822731e-01
4.59944814e-01 -5.58535755e-02 5.00489235e-01 8.70417058e-01
-1.52416661e-01 -2.94148803e-01 1.79620445e-01 -1.81627739e-03
7.57349730e-01 -1.14531137e-01 -3.88082474e-01 -1.48974860e+00
8.73255581e-02 -9.73868668e-01 -4.18482542e-01 -2.70708859e-01
3.71893257e-01 -1.46684182e+00 1.56636521e-01 -1.97808313e+00
9.35713798e-02 -5.79487205e-01 -5.69279015e-01 7.04439521e-01
-2.32570291e-01 5.09155214e-01 1.67746153e-02 4.52294707e-01
-1.06858000e-01 6.52811751e-02 1.47789431e+00 -3.42256188e-01
-2.28804946e-01 -2.42570952e-01 -3.18805248e-01 9.49893653e-01
9.80941653e-01 -2.42788151e-01 -2.68822163e-01 -7.77721167e-01
-4.13173139e-01 -4.14283536e-02 3.90259326e-01 -1.25754225e+00
1.50911853e-01 5.35002649e-02 7.32231975e-01 -2.81911045e-01
-3.96922007e-02 -4.99505401e-01 -1.11132473e-01 5.76088071e-01
-2.22409487e-01 -1.21065654e-01 2.59859432e-02 -5.57605214e-02
-6.90921545e-02 -7.01427877e-01 1.13761020e+00 -5.04335582e-01
-6.96786642e-01 4.96806234e-01 -9.16836739e-01 1.33850882e-02
9.23887849e-01 -2.64840543e-01 -1.39930367e-01 -1.81914002e-01
-1.01084423e+00 -5.02793938e-02 2.17188433e-01 3.02451581e-01
5.92354715e-01 -1.06356120e+00 -7.60190248e-01 -6.60265377e-03
-3.97109598e-01 -1.68919042e-01 5.02439439e-01 1.50002193e+00
-8.48322332e-01 5.17689645e-01 -7.98846960e-01 -5.44461370e-01
-7.23818779e-01 2.43796051e-01 9.91771400e-01 -2.72488624e-01
-8.91450822e-01 5.88189483e-01 -7.98021555e-02 -3.10498089e-01
2.60974824e-01 -3.41365695e-01 -5.29249907e-01 9.13120434e-02
9.72391605e-01 4.35752332e-01 3.99502426e-01 -1.04586363e+00
-1.47448227e-01 3.21254581e-01 -3.06626558e-01 -7.07500875e-02
1.36804760e+00 -7.96318725e-02 -4.19274420e-01 1.47489578e-01
9.19696689e-01 -5.85018456e-01 -1.02939630e+00 -3.95017564e-01
1.69360340e-01 1.21891826e-01 7.05451787e-01 -1.11069715e+00
-1.53568196e+00 9.90087032e-01 1.30247831e+00 -1.34218246e-01
8.42918634e-01 -2.40392059e-01 1.29785645e+00 5.36818169e-02
2.91996777e-01 -9.81793404e-01 -2.58763164e-01 3.26280862e-01
8.18939447e-01 -1.00276625e+00 -3.16995829e-01 -1.02789234e-03
-3.60159725e-01 1.22147548e+00 3.99679065e-01 -2.62754828e-01
4.78423417e-01 2.52334267e-01 2.37990975e-01 -1.75446928e-01
2.63487995e-01 -2.06417650e-01 3.62125665e-01 8.41077507e-01
5.20528734e-01 1.73964098e-01 -5.31530976e-01 5.19428909e-01
3.16931866e-02 1.89368650e-01 3.85383099e-01 9.02140975e-01
-6.90388322e-01 -1.10535574e+00 -6.36980236e-01 1.07279658e+00
-7.00871944e-01 6.98620826e-02 2.27816124e-02 3.95661652e-01
4.39561456e-01 5.45007885e-01 8.63992795e-02 2.34253392e-01
-1.38934841e-02 4.27878618e-01 7.45145977e-01 -5.04464805e-01
-4.04944330e-01 -1.55650735e-01 -3.10461998e-01 -1.33942530e-01
-4.46640402e-01 -8.00634027e-01 -1.83824444e+00 2.62500048e-01
-1.15465835e-01 -5.99836744e-02 8.00998628e-01 1.43585110e+00
-1.35570401e-02 6.27915978e-01 -5.92862889e-02 -1.07323694e+00
-1.71343341e-01 -1.18318987e+00 -5.84641695e-01 5.40847361e-01
2.49656245e-01 -9.50684905e-01 -3.93189937e-02 2.75092036e-01] | [14.162829399108887, -2.122345209121704] |
11275abe-2e50-4272-8749-820135635b9c | improving-opinion-based-question-answering | 2306.07499 | null | https://arxiv.org/abs/2306.07499v1 | https://arxiv.org/pdf/2306.07499v1.pdf | Improving Opinion-based Question Answering Systems Through Label Error Detection and Overwrite | Label error is a ubiquitous problem in annotated data. Large amounts of label error substantially degrades the quality of deep learning models. Existing methods to tackle the label error problem largely focus on the classification task, and either rely on task specific architecture or require non-trivial additional computations, which is undesirable or even unattainable for industry usage. In this paper, we propose LEDO: a model-agnostic and computationally efficient framework for Label Error Detection and Overwrite. LEDO is based on Monte Carlo Dropout combined with uncertainty metrics, and can be easily generalized to multiple tasks and data sets. Applying LEDO to an industry opinion-based question answering system demonstrates it is effective at improving accuracy in all the core models. Specifically, LEDO brings 1.1% MRR gain for the retrieval model, 1.5% PR AUC improvement for the machine reading comprehension model, and 0.9% rise in the Average Precision for the ranker, on top of the strong baselines with a large-scale social media dataset. Importantly, LEDO is computationally efficient compared to methods that require loss function change, and cost-effective as the resulting data can be used in the same continuous training pipeline for production. Further analysis shows that these gains come from an improved decision boundary after cleaning the label errors existed in the training data. | ['Pranab Mohanty', 'Gagan Aneja', 'Nikita Bhalla', 'Hanwen Zha', 'Debojeet Chatterjee', 'Stanislav Peshterliev', 'Shashank Jain', 'Ahmed K. Mohamed', 'Xiao Yang'] | 2023-06-13 | null | null | null | null | ['reading-comprehension', 'machine-reading-comprehension'] | ['natural-language-processing', 'natural-language-processing'] | [ 3.30658704e-01 2.48142824e-01 -5.25022708e-02 -4.72984463e-01
-1.40722644e+00 -6.19275510e-01 4.08881307e-01 5.43617189e-01
-6.57962620e-01 7.31223881e-01 -1.41080990e-01 -3.23397189e-01
-1.67728946e-01 -3.92973363e-01 -1.01670754e+00 -3.85060698e-01
4.32756692e-01 6.63220584e-01 1.65553302e-01 7.73016810e-02
1.39001176e-01 -1.13809161e-01 -1.47202170e+00 2.59574264e-01
1.13500547e+00 1.19356287e+00 -1.02020733e-01 4.72019404e-01
-3.89142819e-02 1.01536858e+00 -6.59977019e-01 -8.12264860e-01
1.42949435e-03 -2.29181007e-01 -9.16322708e-01 -1.72287107e-01
9.84107971e-01 -4.18604583e-01 6.47534337e-03 8.94286335e-01
6.47108734e-01 7.23305345e-02 8.08784902e-01 -1.07830191e+00
-7.51133621e-01 5.82031429e-01 -3.46739948e-01 -1.21942967e-01
3.56738828e-02 1.21459045e-01 1.49071872e+00 -8.58048141e-01
3.56013954e-01 1.14068878e+00 9.58015680e-01 6.53196454e-01
-1.43429077e+00 -6.70888245e-01 1.44119948e-01 1.10781744e-01
-1.07836699e+00 -3.24820310e-01 8.93038660e-02 -4.95214581e-01
9.06079948e-01 1.53960153e-01 2.14395821e-01 1.31371129e+00
2.26156414e-01 9.69687581e-01 1.21006262e+00 -4.83528733e-01
2.24374622e-01 2.25400150e-01 6.55873477e-01 7.96343446e-01
3.55735391e-01 -1.61086526e-02 -6.90279722e-01 -4.82588857e-02
4.88517769e-02 -6.48070499e-02 -3.09201032e-01 -2.21339583e-01
-8.27477157e-01 8.09958994e-01 4.73655790e-01 -3.67788851e-01
1.05189063e-01 5.44917405e-01 2.87410885e-01 3.63856584e-01
9.21438754e-01 7.41334140e-01 -7.85933197e-01 -1.85682222e-01
-8.68773460e-01 3.41836333e-01 7.64932036e-01 7.75050461e-01
5.69201529e-01 -4.27712530e-01 -6.05852902e-01 9.37611282e-01
2.72711277e-01 5.13426960e-01 3.30753744e-01 -1.14264548e+00
3.39253962e-01 6.55809641e-01 2.77578354e-01 -6.40002012e-01
-5.39890587e-01 -7.28726625e-01 -2.96774894e-01 1.44087806e-01
6.99791968e-01 -1.45653635e-01 -1.17983913e+00 1.98791683e+00
1.59596875e-01 -3.37333083e-02 -1.46075428e-01 6.37651503e-01
7.44570196e-01 3.34137768e-01 2.81988412e-01 2.15275541e-01
1.46901214e+00 -1.29278374e+00 -8.34847093e-01 -4.60388422e-01
1.17124808e+00 -6.59466088e-01 1.44746661e+00 6.05123460e-01
-8.62517715e-01 -3.09131593e-01 -1.15786064e+00 -4.16959375e-01
-3.20866525e-01 1.29655793e-01 7.10053027e-01 6.62912309e-01
-8.01484108e-01 6.79566026e-01 -5.72899759e-01 -5.58030829e-02
6.17602348e-01 2.97622323e-01 -3.08178156e-03 -4.00561452e-01
-1.20391941e+00 9.81910825e-01 1.61025941e-01 -5.54923117e-02
-8.73741806e-01 -9.86481071e-01 -6.52910650e-01 2.21090123e-01
5.76639950e-01 -6.27728641e-01 1.83023369e+00 -8.22699010e-01
-1.26762021e+00 8.26430380e-01 -1.17271602e-01 -4.46910322e-01
6.84526742e-01 -7.40733564e-01 -9.76986960e-02 -2.24110365e-01
3.96127030e-02 7.65138805e-01 6.75164521e-01 -9.73484337e-01
-6.62207127e-01 -2.17505708e-01 3.82419652e-03 1.71634689e-01
-4.00193453e-01 -1.78826466e-01 -3.89375806e-01 -4.28061545e-01
-5.60068041e-02 -9.69787359e-01 5.13276607e-02 -1.01864345e-01
-1.47698492e-01 -6.62058175e-01 1.84537902e-01 -8.30085993e-01
1.21516681e+00 -2.04858971e+00 -1.62704438e-01 -1.75841898e-01
2.69055009e-01 2.26884916e-01 -1.76996499e-01 4.34328876e-02
2.62127757e-01 3.63031864e-01 -2.79189616e-01 -4.64048982e-01
2.46311262e-01 -1.98822588e-01 -3.04546297e-01 2.06407890e-01
5.20019531e-01 9.45895016e-01 -1.08159590e+00 -2.42239103e-01
-3.25695544e-01 2.30949402e-01 -7.73876071e-01 1.38340771e-01
-6.02153540e-01 1.38897493e-01 -5.36747836e-02 4.78792399e-01
3.66652906e-01 -6.00349486e-01 -7.52870291e-02 8.37542266e-02
4.84634966e-01 6.86323285e-01 -7.76027322e-01 1.74900627e+00
-7.05860019e-01 5.90940535e-01 -1.44167647e-01 -3.84873182e-01
6.65657461e-01 1.87592223e-01 -2.19150987e-02 -7.00891733e-01
1.29792571e-01 5.02063930e-01 -4.85910699e-02 -3.69875103e-01
4.49981540e-01 4.93681729e-02 -2.23335214e-02 5.83050787e-01
1.99035391e-01 -2.44373500e-01 8.63491893e-02 2.30366662e-01
1.26342797e+00 1.71883494e-01 -3.13998610e-01 -1.80330977e-01
-3.06134056e-02 2.23155189e-02 6.32817328e-01 9.68413055e-01
-1.59177139e-01 5.98656118e-01 4.62405473e-01 -6.01062700e-02
-9.42030668e-01 -9.17435110e-01 -4.20906425e-01 1.38492858e+00
-6.69621527e-02 -3.87018144e-01 -9.39720213e-01 -1.04097474e+00
1.88401982e-01 9.39714909e-01 -4.71064240e-01 -3.80066723e-01
-1.89758092e-01 -8.90353203e-01 4.91682976e-01 4.39888149e-01
5.42919576e-01 -7.90124774e-01 -1.09780595e-01 1.46597669e-01
-2.70834923e-01 -8.90263498e-01 -4.01253521e-01 3.22853714e-01
-8.39230299e-01 -1.00905573e+00 -4.93957639e-01 -7.05046892e-01
5.78986943e-01 -5.60674556e-02 1.49730468e+00 7.38278776e-02
-1.65167049e-01 2.23835468e-01 -3.71321976e-01 -8.40220451e-01
-3.32496107e-01 4.23271984e-01 -1.33983642e-01 -8.56252164e-02
7.51464367e-01 -2.84221731e-02 -6.01360738e-01 3.07513893e-01
-6.83727920e-01 -1.75927430e-01 4.40632731e-01 1.22869897e+00
3.79115462e-01 -1.97425097e-01 9.70884621e-01 -1.28246891e+00
7.21189320e-01 -4.50538456e-01 -7.03402221e-01 4.42975312e-01
-1.31639302e+00 2.08294645e-01 1.97553307e-01 -3.11684251e-01
-1.03330672e+00 -3.79511639e-02 -1.10701002e-01 2.61281375e-02
2.30519027e-01 4.19537216e-01 -2.97316778e-02 2.73072958e-01
9.19617474e-01 -5.04607916e-01 -5.26236333e-02 -5.76353610e-01
3.87466609e-01 8.88072729e-01 1.50271207e-01 -4.24135476e-01
2.67669469e-01 1.07428044e-01 -3.63702357e-01 -8.94252062e-02
-1.67561603e+00 -3.39949638e-01 -1.17453761e-01 -1.36960968e-02
8.20542812e-01 -1.16964102e+00 -6.87538803e-01 5.74804485e-01
-1.00702226e+00 -5.74285746e-01 -2.19930857e-01 3.90715241e-01
-1.43351614e-01 8.55142698e-02 -7.58887470e-01 -7.16726005e-01
-4.41222399e-01 -1.12063587e+00 1.04387355e+00 1.57481253e-01
-4.25983846e-01 -9.29086745e-01 -1.73817769e-01 9.68698561e-01
4.32472587e-01 -1.57247186e-01 1.07658541e+00 -8.32724512e-01
-5.53459466e-01 -4.34407383e-01 -3.34585011e-01 8.03608656e-01
-2.92774737e-01 -1.79745346e-01 -1.38383102e+00 -4.21350718e-01
-9.40423235e-02 -9.52879608e-01 1.21675766e+00 1.59179911e-01
1.25513935e+00 4.04637307e-02 -1.62180007e-01 2.19700187e-01
1.05924845e+00 -2.05108300e-01 2.85660177e-01 1.86193198e-01
7.30956018e-01 6.68521404e-01 6.14656687e-01 -3.97050418e-02
3.49749684e-01 6.51789427e-01 3.98412704e-01 -7.44910631e-03
-2.63706654e-01 -4.31204915e-01 5.03578305e-01 9.61551726e-01
6.24087691e-01 -3.80396754e-01 -1.07031620e+00 5.99838197e-01
-1.90447223e+00 -3.47952604e-01 -4.04208750e-01 2.23441505e+00
1.18467259e+00 3.23701829e-01 -3.96822721e-01 -9.44669768e-02
5.43054342e-01 -1.50292560e-01 -8.34902525e-01 -4.27364379e-01
4.59744297e-02 3.80792439e-01 6.34485245e-01 6.18165612e-01
-1.07181454e+00 8.88946176e-01 6.47436810e+00 1.09249365e+00
-6.12617075e-01 5.14310777e-01 8.32331181e-01 -3.16205412e-01
-3.63260061e-01 -3.14219855e-02 -9.69037712e-01 3.87413800e-01
1.28609419e+00 2.15315148e-01 3.00523996e-01 8.04131091e-01
-1.16412163e-01 -3.21199626e-01 -1.39245152e+00 8.70632470e-01
8.31480697e-02 -9.28885579e-01 -1.85973570e-01 -2.23385785e-02
1.00109899e+00 1.77242801e-01 1.85947120e-01 7.45740116e-01
6.19866312e-01 -1.18089175e+00 7.81093657e-01 3.95639390e-01
7.35463560e-01 -6.88429952e-01 1.08382893e+00 3.86931628e-01
-3.10486823e-01 -2.46624440e-01 -3.94348472e-01 -2.94250846e-01
5.72347306e-02 1.06929231e+00 -8.32435310e-01 1.40441179e-01
7.12402225e-01 4.31490630e-01 -8.77883017e-01 9.81665194e-01
-6.36764467e-01 1.16633499e+00 -2.34347761e-01 -2.58564323e-01
1.76482439e-01 1.25691965e-01 8.99079964e-02 1.03070772e+00
6.81648850e-02 -3.00058663e-01 -2.25713830e-02 8.02277863e-01
-8.12462151e-01 1.68027118e-01 -2.96008170e-01 8.36588442e-02
5.91503382e-01 1.20006180e+00 -3.17797244e-01 -4.04870808e-01
-3.22783858e-01 1.01257896e+00 6.79589510e-01 1.77938819e-01
-1.01262081e+00 -5.25268018e-01 3.39715153e-01 5.91966845e-02
5.95827810e-02 2.63052821e-01 -5.14453471e-01 -1.05770850e+00
1.14077948e-01 -1.12824714e+00 1.88172117e-01 -6.91702068e-01
-1.53228247e+00 3.25480610e-01 -4.47573304e-01 -7.12984145e-01
-1.98510941e-02 -7.19027281e-01 -5.97033910e-02 9.00716722e-01
-1.53859997e+00 -7.94541121e-01 -3.27926606e-01 -7.18121929e-03
4.36974913e-01 3.01831141e-02 8.43048871e-01 6.21393204e-01
-8.08493972e-01 1.04340804e+00 4.12970603e-01 -4.04936895e-02
1.19940782e+00 -1.49134958e+00 4.53934342e-01 7.40592480e-01
1.06680155e-01 5.72848678e-01 4.33490962e-01 -5.17169714e-01
-8.75412345e-01 -1.28046548e+00 1.23569191e+00 -1.21421707e+00
6.28690779e-01 -5.69124818e-01 -8.30415249e-01 6.57496572e-01
1.00471057e-01 -2.47309849e-01 7.41969526e-01 6.14949763e-01
-7.26429820e-01 -1.21879838e-01 -1.09931850e+00 4.92539883e-01
1.01240706e+00 -6.68334961e-01 -4.99125153e-01 7.98810661e-01
1.00651014e+00 -3.81701171e-01 -8.07797670e-01 5.01360178e-01
3.45205516e-01 -6.93336606e-01 4.86196846e-01 -8.44489634e-01
6.69972301e-01 -7.38911936e-03 -8.40493217e-02 -1.21228480e+00
-3.19155276e-01 -3.72809947e-01 -8.11590105e-02 1.42193878e+00
9.63955879e-01 -4.74871188e-01 5.99892735e-01 9.74706531e-01
-2.03060135e-01 -8.17905068e-01 -7.94046342e-01 -7.75999844e-01
3.94595951e-01 -4.98334438e-01 3.23017865e-01 8.42262506e-01
-3.67109746e-01 8.29506874e-01 -2.62254357e-01 3.36613394e-02
6.34390116e-01 -3.05536330e-01 4.28770542e-01 -1.42091262e+00
-5.03667295e-01 -3.11423421e-01 8.89726430e-02 -1.18717837e+00
2.05486685e-01 -1.18052173e+00 4.41751480e-01 -1.57369030e+00
4.10565108e-01 -6.80961907e-01 -4.04391289e-01 6.65362537e-01
-6.29652023e-01 3.65473837e-01 4.19422314e-02 1.88992590e-01
-8.39917421e-01 4.99652863e-01 8.58177602e-01 -2.85550475e-01
1.27144948e-01 -4.15158160e-02 -7.92376459e-01 6.84770644e-01
8.01003098e-01 -7.86295891e-01 -4.37537670e-01 -8.42701197e-01
7.25328088e-01 -5.54088593e-01 2.98222393e-01 -8.83951962e-01
-7.17530102e-02 4.80166554e-01 1.65598374e-02 -1.15765475e-01
1.69443980e-01 -6.65650368e-01 -1.58597887e-01 4.38230932e-01
-8.15654159e-01 -1.32768452e-01 1.34569153e-01 7.77280688e-01
-1.43886975e-03 -7.01776862e-01 6.41445994e-01 6.30437955e-02
-2.59281278e-01 -1.04837798e-01 -1.21140398e-01 5.21976531e-01
6.37311339e-01 3.44676286e-01 -5.98099828e-01 -2.00046957e-01
-5.09502530e-01 5.15096843e-01 2.52580673e-01 4.62950975e-01
-2.72850785e-02 -1.15101099e+00 -8.59154642e-01 -2.43912622e-01
3.68850768e-01 1.50336415e-01 8.72224644e-02 7.68324673e-01
-4.61645067e-01 3.77137095e-01 4.27060783e-01 -6.53762758e-01
-9.71021891e-01 1.69960901e-01 3.74839664e-01 -6.39199018e-01
-1.49952114e-01 1.29655278e+00 -4.83051762e-02 -6.63851261e-01
6.64027631e-01 -4.67956930e-01 2.24556834e-01 3.47470462e-01
4.77024972e-01 4.51766044e-01 5.63043118e-01 1.70076281e-01
-6.57395944e-02 2.06196561e-01 -4.09112751e-01 -7.05263838e-02
1.13274014e+00 -6.70902506e-02 3.95172369e-03 6.81558192e-01
1.07742131e+00 -2.96961349e-02 -1.49728811e+00 -3.28651458e-01
3.07792187e-01 -6.12238906e-02 4.01862383e-01 -1.49946594e+00
-7.85582364e-01 9.97517467e-01 6.86070383e-01 5.59652075e-02
6.94753647e-01 -9.58574414e-02 8.21211576e-01 5.44937849e-01
2.34075770e-01 -1.27388585e+00 1.42105892e-01 6.34320915e-01
6.57211781e-01 -1.63683391e+00 -7.28283748e-02 -2.34716982e-01
-5.43546021e-01 4.64159548e-01 7.01765776e-01 2.34671578e-01
3.65392596e-01 -8.26245397e-02 2.16752872e-01 -1.58914685e-01
-9.85005915e-01 8.19838699e-03 3.07858407e-01 2.07498521e-01
6.69073105e-01 9.26377345e-03 -5.11114478e-01 6.91510737e-01
-1.08226333e-02 -7.18260631e-02 3.88828427e-01 6.74171448e-01
-4.23385799e-01 -9.92781520e-01 3.02793197e-02 7.89667308e-01
-5.88619173e-01 -4.48039889e-01 -2.98724085e-01 4.61536020e-01
1.22565836e-01 1.23507857e+00 1.52368275e-02 -3.07433069e-01
3.76899809e-01 6.49235427e-01 4.08419847e-01 -8.61354649e-01
-8.28158975e-01 -2.95426458e-01 4.09629732e-01 -5.92387915e-01
-7.48472735e-02 -5.51721811e-01 -1.22448313e+00 -2.81806290e-01
-9.21263278e-01 2.01072656e-02 6.98741376e-01 7.80407727e-01
6.82998598e-01 6.79787159e-01 1.78885028e-01 -2.45149568e-01
-1.03586078e+00 -1.27939415e+00 -3.41947228e-01 6.48725390e-01
1.50897026e-01 -7.78318226e-01 -7.06187427e-01 -1.65865719e-01] | [9.295044898986816, 4.300482273101807] |
405b7341-f1f2-4f1a-b44d-5ecea7b7650a | background-subtraction-with-real-time | 1811.10020 | null | http://arxiv.org/abs/1811.10020v2 | http://arxiv.org/pdf/1811.10020v2.pdf | Background Subtraction with Real-time Semantic Segmentation | Accurate and fast foreground object extraction is very important for object
tracking and recognition in video surveillance. Although many background
subtraction (BGS) methods have been proposed in the recent past, it is still
regarded as a tough problem due to the variety of challenging situations that
occur in real-world scenarios. In this paper, we explore this problem from a
new perspective and propose a novel background subtraction framework with
real-time semantic segmentation (RTSS). Our proposed framework consists of two
components, a traditional BGS segmenter $\mathcal{B}$ and a real-time semantic
segmenter $\mathcal{S}$. The BGS segmenter $\mathcal{B}$ aims to construct
background models and segments foreground objects. The real-time semantic
segmenter $\mathcal{S}$ is used to refine the foreground segmentation outputs
as feedbacks for improving the model updating accuracy. $\mathcal{B}$ and
$\mathcal{S}$ work in parallel on two threads. For each input frame $I_t$, the
BGS segmenter $\mathcal{B}$ computes a preliminary foreground/background
(FG/BG) mask $B_t$. At the same time, the real-time semantic segmenter
$\mathcal{S}$ extracts the object-level semantics ${S}_t$. Then, some specific
rules are applied on ${B}_t$ and ${S}_t$ to generate the final detection
${D}_t$. Finally, the refined FG/BG mask ${D}_t$ is fed back to update the
background model. Comprehensive experiments evaluated on the CDnet 2014 dataset
demonstrate that our proposed method achieves state-of-the-art performance
among all unsupervised background subtraction methods while operating at
real-time, and even performs better than some deep learning based supervised
algorithms. In addition, our proposed framework is very flexible and has the
potential for generalization. | ['Arjan Kuijper', 'Xiang Chen', 'Ming Zhu', 'Michael Goesele', 'Dongdong Zeng'] | 2018-11-25 | null | null | null | null | ['foreground-segmentation'] | ['computer-vision'] | [ 5.24979651e-01 -4.33813930e-01 1.00065991e-01 -4.18489158e-01
-4.86382365e-01 -2.81048656e-01 2.09998146e-01 1.46854162e-01
-4.89650458e-01 4.94307280e-01 -7.19667912e-01 -5.36514223e-01
1.04585022e-01 -1.05676353e+00 -6.67520881e-01 -9.90876436e-01
-4.61529978e-02 2.70972639e-01 1.05546570e+00 2.92866840e-04
1.03418240e-02 4.51952577e-01 -1.57336211e+00 1.11262493e-01
9.41041410e-01 1.43522644e+00 3.65420252e-01 6.88355863e-01
-5.44366539e-01 7.27741718e-01 -9.37849522e-01 -4.14823502e-01
5.24916530e-01 -4.96720701e-01 -2.77883112e-01 2.46435896e-01
4.59886789e-01 -5.19267678e-01 -2.90653855e-01 1.52565598e+00
1.98679686e-01 2.01517329e-01 2.25495055e-01 -1.16270387e+00
1.87887311e-01 1.87143967e-01 -1.07305479e+00 7.16672659e-01
-1.63107216e-01 4.49941725e-01 2.25254327e-01 -5.90572059e-01
3.66037250e-01 1.18426049e+00 3.69475186e-01 4.95461196e-01
-8.82631779e-01 -1.22953045e+00 6.95415139e-01 9.97067466e-02
-1.23192406e+00 -7.03964159e-02 7.15793252e-01 -4.73967880e-01
4.91398335e-01 3.36519212e-01 8.01635563e-01 4.04581696e-01
-1.09121669e-03 1.02295721e+00 1.03671920e+00 -2.27765769e-01
4.02719259e-01 -1.53587058e-01 5.46499968e-01 9.32667851e-01
4.36490297e-01 -1.85211033e-01 -2.30379179e-01 1.07241765e-01
8.56721461e-01 2.84239918e-01 -3.45473081e-01 -1.11938924e-01
-8.64342809e-01 5.08579731e-01 3.43635023e-01 1.19559646e-01
-3.27102989e-01 1.56802595e-01 3.11360836e-01 -2.15936512e-01
2.96188027e-01 -4.07835722e-01 -4.42759275e-01 1.61819801e-01
-1.24598587e+00 1.96259633e-01 5.04988730e-01 9.49019551e-01
8.97176683e-01 2.89864600e-01 -1.35544002e-01 5.78891754e-01
2.36625478e-01 7.44560063e-01 -1.27295300e-01 -9.73413229e-01
4.61557239e-01 8.72355342e-01 2.37819716e-01 -8.70118082e-01
-2.51961321e-01 -4.71932054e-01 -1.03285515e+00 6.10290051e-01
6.28264248e-01 -2.13775858e-01 -1.26203871e+00 1.50799394e+00
6.96463466e-01 5.26597023e-01 -2.30429962e-01 7.47834206e-01
8.41349602e-01 9.47602868e-01 4.02373731e-01 -5.49229622e-01
1.48705292e+00 -8.23080003e-01 -5.39276063e-01 -5.04634261e-01
9.26492438e-02 -8.28984857e-01 6.33828461e-01 4.72812772e-01
-1.00208914e+00 -8.16981077e-01 -9.04238999e-01 5.64734578e-01
-2.10229710e-01 1.76688686e-01 4.96203601e-01 9.78235006e-01
-7.81711042e-01 2.03751117e-01 -1.13473582e+00 -2.57516950e-01
5.96534789e-01 4.98660654e-01 2.22143993e-01 -1.59878746e-01
-8.68348956e-01 5.16283989e-01 4.92618978e-01 3.78005207e-01
-1.08651781e+00 -2.75749385e-01 -7.14452386e-01 -2.16022000e-01
9.22352970e-01 -4.33114707e-01 1.01448786e+00 -1.08713102e+00
-1.12748706e+00 9.60377991e-01 -3.11633974e-01 -4.10974830e-01
4.58000451e-01 4.28665206e-02 -5.71280062e-01 2.52129436e-01
1.41588673e-01 5.17237484e-01 9.53454494e-01 -1.24509144e+00
-1.31046748e+00 -6.16633713e-01 1.74759910e-01 -3.66199273e-03
8.14392418e-03 3.97682607e-01 -1.01765728e+00 -7.66664267e-01
4.46882397e-01 -6.48802578e-01 -2.39411399e-01 -1.73871852e-02
-1.81160316e-01 5.21716326e-02 1.11419749e+00 -9.08137798e-01
1.38949537e+00 -2.06126523e+00 -1.25074148e-01 2.72017837e-01
3.73426557e-01 9.63210762e-01 2.75559843e-01 -3.68647546e-01
1.82335392e-01 -2.79205173e-01 -4.80406374e-01 -1.30010486e-01
-4.78675455e-01 1.14604130e-01 2.57647373e-02 3.43258768e-01
8.36766213e-02 3.79890412e-01 -7.88434148e-01 -5.91187418e-01
5.60714185e-01 2.20273092e-01 -3.12828660e-01 1.37445644e-01
-5.05438924e-01 4.80716944e-01 -6.35933697e-01 8.51759374e-01
1.12791491e+00 -6.78503588e-02 8.69258270e-02 -1.37241855e-01
-4.52257425e-01 -1.55725896e-01 -1.67822802e+00 9.49289382e-01
3.88615876e-01 3.14166874e-01 7.28336990e-01 -1.22797048e+00
1.02363122e+00 -1.22470476e-01 5.09288132e-01 -7.01908410e-01
4.06952471e-01 6.19681366e-02 1.31766587e-01 -2.28697643e-01
3.82381827e-01 -6.87756091e-02 1.56327799e-01 1.30623475e-01
-2.94250101e-01 -3.47378254e-02 4.63210672e-01 1.67049617e-01
1.07345712e+00 3.86833519e-01 1.34176835e-01 -2.06655428e-01
9.52847362e-01 1.90276295e-01 1.09872139e+00 8.71835887e-01
-6.69986367e-01 1.40416667e-01 3.17212939e-01 -4.55285013e-01
-4.20626968e-01 -1.24675262e+00 2.55467504e-01 1.04642498e+00
7.17271030e-01 -2.16046989e-01 -1.22740853e+00 -5.98216593e-01
-3.86160970e-01 6.00958407e-01 -4.00996625e-01 -1.87624991e-02
-8.91917288e-01 -1.07221997e+00 4.04822886e-01 5.92059553e-01
1.07996798e+00 -1.11118472e+00 -9.72719252e-01 3.91336143e-01
-1.75602257e-01 -1.07194495e+00 -3.53481412e-01 7.34911636e-02
-8.44696105e-01 -1.16235960e+00 -6.74912870e-01 -5.75814784e-01
6.95899069e-01 5.75827956e-01 7.90626585e-01 3.07947963e-01
-5.06785989e-01 9.75097269e-02 -2.59177417e-01 -6.74493551e-01
-3.84345412e-01 -3.92592758e-01 -3.69781196e-01 2.20763817e-01
5.58568478e-01 -1.22134805e-01 -8.56945515e-01 4.97965991e-01
-1.08545852e+00 1.68961763e-01 5.13531685e-01 3.75601828e-01
8.29035938e-01 5.32597184e-01 1.25232935e-01 -7.61342645e-01
-1.56006604e-01 -2.24794224e-02 -1.28185165e+00 1.64743856e-01
-6.98709488e-02 -5.00019014e-01 5.13241947e-01 -3.46908420e-01
-1.12879825e+00 -9.15094614e-02 -1.36956256e-02 -4.85502332e-01
-4.01848376e-01 8.66123289e-03 -5.20670712e-01 1.30121469e-01
3.38399917e-01 4.65725631e-01 -1.92438915e-01 -3.50553751e-01
-1.03682891e-01 3.57422501e-01 7.54314363e-01 -4.06339169e-01
7.81589687e-01 7.27608502e-01 -1.84672639e-01 -8.65416408e-01
-5.60320020e-01 -4.95596647e-01 -4.56227422e-01 -5.33974946e-01
1.15219784e+00 -8.32939923e-01 -6.89038515e-01 1.04649484e+00
-8.89487982e-01 -5.94825923e-01 -7.15398714e-02 2.71325976e-01
-5.76447770e-02 4.80912387e-01 -3.93322140e-01 -1.16635668e+00
-2.77725756e-01 -1.53394568e+00 9.22339201e-01 8.42016637e-01
1.97994456e-01 -5.56756079e-01 -6.77420974e-01 6.21774554e-01
2.90375262e-01 3.22115272e-01 5.66730976e-01 -2.63240755e-01
-1.15009391e+00 -1.99607700e-01 -6.91382468e-01 5.65726638e-01
1.94025278e-01 1.47887066e-01 -6.99875057e-01 -2.36620992e-01
1.21334381e-01 4.47305948e-01 1.01515353e+00 6.37666821e-01
1.15587604e+00 3.13736759e-02 -5.30658960e-01 4.37305570e-01
1.22252607e+00 9.71170843e-01 7.08834171e-01 2.90444434e-01
5.59627593e-01 3.01125139e-01 8.18097889e-01 4.06943440e-01
3.52395087e-01 4.42290276e-01 5.15075922e-01 -2.85870880e-01
-1.52836755e-01 4.06915843e-01 4.85186189e-01 2.78815627e-01
-3.06914691e-02 -7.92652369e-02 -9.71244991e-01 3.56070489e-01
-1.67954922e+00 -1.02782905e+00 -3.03746521e-01 2.34966683e+00
7.64824808e-01 5.32814503e-01 2.00057417e-01 9.69613865e-02
1.17274821e+00 3.08042824e-01 -6.88485265e-01 1.50991499e-01
-9.17792991e-02 6.50130570e-01 5.27312100e-01 3.05064738e-01
-1.24798119e+00 1.10704076e+00 3.14695287e+00 9.25848544e-01
-1.17590392e+00 3.97658199e-02 8.21998775e-01 -4.33844849e-02
3.36965889e-01 5.76956235e-02 -1.14390850e+00 8.36339593e-01
2.85455197e-01 4.07113254e-01 1.87249541e-01 8.02825153e-01
2.72037745e-01 -8.14946651e-01 -6.03313208e-01 1.04406416e+00
6.15167022e-02 -1.20149755e+00 -1.54073954e-01 -3.49186659e-01
5.61926842e-01 -2.71151185e-01 -2.37876669e-01 3.54139805e-01
2.79646963e-01 -5.19393861e-01 9.24634874e-01 1.23148851e-01
5.75877011e-01 -6.23800993e-01 5.96088588e-01 5.35324872e-01
-1.75721681e+00 -8.11570510e-02 -9.08340588e-02 1.11153826e-01
2.84520864e-01 6.84632838e-01 -3.11891854e-01 4.65594500e-01
1.15231097e+00 2.98457354e-01 -5.89655995e-01 9.25687313e-01
-1.11584209e-01 6.87120616e-01 -3.05071861e-01 1.78114921e-02
2.59682506e-01 -5.08471668e-01 6.10352099e-01 1.40460324e+00
2.57176743e-03 5.88821113e-01 5.46108305e-01 6.62639976e-01
8.14602971e-02 -1.03161111e-01 1.09112054e-01 1.74696386e-01
2.32054397e-01 1.16895509e+00 -1.56215060e+00 -8.62471104e-01
-3.64304334e-01 8.79418612e-01 -2.88708299e-01 4.00156647e-01
-1.20043683e+00 -3.96917492e-01 7.35896528e-01 2.95256823e-01
4.68202859e-01 -2.18832731e-01 -1.58692673e-01 -1.07581043e+00
3.99351455e-02 -9.24284220e-01 6.45080745e-01 -5.09010255e-01
-7.09719837e-01 4.63805020e-01 1.33577317e-01 -9.74356711e-01
5.35662353e-01 -7.00849771e-01 -5.34332812e-01 8.05406749e-01
-1.36935699e+00 -7.98374116e-01 -8.28429282e-01 8.18421483e-01
6.73781574e-01 -7.05328807e-02 2.85708755e-01 4.49991047e-01
-9.44476128e-01 1.85486495e-01 -3.30342911e-02 5.99663377e-01
3.47145468e-01 -9.61254060e-01 3.39826167e-01 1.38982379e+00
-3.60360324e-01 5.40302336e-01 6.06641710e-01 -9.99721527e-01
-1.04022682e+00 -1.34443092e+00 1.25214592e-01 1.13576032e-01
2.98885882e-01 -2.39553943e-01 -8.96770120e-01 6.22295201e-01
-1.58312365e-01 8.93322825e-02 2.53188968e-01 -6.77136302e-01
-6.33437326e-03 -5.18044889e-01 -1.30333471e+00 6.63142622e-01
9.35127258e-01 6.81394711e-02 -3.26072603e-01 4.21630070e-02
4.44699377e-01 -6.95920289e-01 -2.84611583e-01 6.55775785e-01
2.44019777e-01 -1.22627866e+00 9.72328722e-01 -2.50611007e-02
-2.27901012e-01 -9.25155938e-01 -1.61551461e-01 -3.80231887e-01
1.03541844e-01 -7.02669680e-01 7.63163194e-02 1.24972522e+00
-1.06153846e-01 -6.94897473e-01 9.43743229e-01 6.81872189e-01
-1.28888741e-01 -4.51705188e-01 -9.01021004e-01 -6.25108540e-01
-2.64072239e-01 -6.09007597e-01 3.62372488e-01 6.05905771e-01
-9.64526355e-01 -3.14157814e-01 2.86538713e-02 4.46878105e-01
1.00381708e+00 2.67138571e-01 9.43574667e-01 -1.08518386e+00
-1.93943843e-01 -5.15309155e-01 -3.39066744e-01 -1.13779020e+00
-2.49291137e-01 -4.40243930e-01 9.08539519e-02 -1.54260480e+00
1.95350349e-01 -7.54965842e-01 -3.81530344e-01 3.22293252e-01
-5.48285365e-01 3.00363630e-01 3.07964146e-01 -9.73419473e-02
-8.32268417e-01 1.26788747e-02 1.03309357e+00 -2.64518559e-01
-2.81313419e-01 2.94746786e-01 -4.17996645e-01 1.20539832e+00
6.65130615e-01 -3.39784652e-01 -1.56355441e-01 -3.29593390e-01
-4.91226882e-01 2.31055304e-01 6.17520273e-01 -1.08171356e+00
2.19943032e-01 -4.17146713e-01 6.36112869e-01 -9.49376643e-01
3.68987739e-01 -7.15634644e-01 1.73170373e-01 5.79044163e-01
4.37675446e-01 -8.99529755e-02 4.09062415e-01 4.57303405e-01
-1.24794077e-02 -5.64798154e-02 1.36786127e+00 -3.11768115e-01
-9.58324492e-01 4.53816295e-01 -3.41383994e-01 9.12722796e-02
1.38811088e+00 -6.41383290e-01 -1.74030975e-01 -3.97841372e-02
-6.51065469e-01 2.36133158e-01 3.89092505e-01 2.70386010e-01
4.56282198e-01 -6.64794028e-01 -4.35526669e-01 3.59821737e-01
-2.33527124e-01 5.27678609e-01 5.41125774e-01 9.33787167e-01
-6.88437462e-01 2.12768801e-02 -2.37579402e-02 -8.78112793e-01
-1.39603257e+00 4.15723830e-01 4.62702036e-01 -2.82091230e-01
-4.17891920e-01 1.18336618e+00 4.19844091e-01 8.29414129e-02
4.41086978e-01 -4.59288865e-01 1.65330097e-01 5.06492965e-02
7.47420073e-01 6.62558615e-01 -7.97055662e-02 -7.94787228e-01
-5.39822996e-01 5.18094063e-01 7.13066831e-02 -6.19370043e-02
9.90678847e-01 -1.39631078e-01 -3.75438839e-01 -3.01179346e-02
6.89038694e-01 -1.96999907e-01 -1.45486724e+00 -2.57795960e-01
7.42614968e-03 -5.84096849e-01 -2.96312362e-01 -7.27323771e-01
-1.44374514e+00 8.73805702e-01 8.40371370e-01 -9.87483561e-02
1.46392787e+00 -1.65375248e-01 8.60306323e-01 -2.89531127e-02
5.21734178e-01 -1.16305113e+00 1.14021003e-01 3.97644579e-01
1.62415072e-01 -1.10326076e+00 4.52630855e-02 -5.78346908e-01
-3.64829183e-01 9.03152943e-01 8.60785067e-01 6.87807575e-02
5.01228571e-01 4.86412138e-01 1.75589502e-01 -4.74736355e-02
3.84932794e-02 -3.51624519e-01 -3.88641618e-02 4.61385608e-01
-8.29685200e-03 -7.37930462e-02 -7.88143873e-02 5.84975719e-01
2.27105945e-01 -1.10116077e-03 2.04943269e-01 1.18195629e+00
-8.14329207e-01 -1.09921813e+00 -8.00998926e-01 4.97110456e-01
-4.32805091e-01 1.55365884e-01 -5.51917776e-03 7.91136026e-01
8.01189721e-01 1.13202822e+00 5.40246032e-02 9.61350277e-02
9.65465307e-02 1.31535232e-01 3.90303731e-01 -4.68154073e-01
-6.89285815e-01 5.67752242e-01 -3.21443379e-01 -6.23339653e-01
-5.37101209e-01 -8.02199125e-01 -1.75419688e+00 -1.91995665e-01
-3.60874027e-01 1.50202541e-03 5.17961919e-01 8.09615493e-01
-8.99536759e-02 6.83112442e-01 1.85251310e-01 -7.60325909e-01
2.60285400e-02 -5.39621532e-01 -5.20453811e-01 4.40283209e-01
1.27702862e-01 -7.05833614e-01 -4.24248837e-02 3.37777108e-01] | [8.998948097229004, -0.5708959698677063] |
24009884-caf8-4be4-b38d-1ed5c84ad098 | heterogeneous-domain-adaptation-for-iot | 2301.09801 | null | https://arxiv.org/abs/2301.09801v1 | https://arxiv.org/pdf/2301.09801v1.pdf | Heterogeneous Domain Adaptation for IoT Intrusion Detection: A Geometric Graph Alignment Approach | Data scarcity hinders the usability of data-dependent algorithms when tackling IoT intrusion detection (IID). To address this, we utilise the data rich network intrusion detection (NID) domain to facilitate more accurate intrusion detection for IID domains. In this paper, a Geometric Graph Alignment (GGA) approach is leveraged to mask the geometric heterogeneities between domains for better intrusion knowledge transfer. Specifically, each intrusion domain is formulated as a graph where vertices and edges represent intrusion categories and category-wise interrelationships, respectively. The overall shape is preserved via a confused discriminator incapable to identify adjacency matrices between different intrusion domain graphs. A rotation avoidance mechanism and a centre point matching mechanism is used to avoid graph misalignment due to rotation and symmetry, respectively. Besides, category-wise semantic knowledge is transferred to act as vertex-level alignment. To exploit the target data, a pseudo-label election mechanism that jointly considers network prediction, geometric property and neighbourhood information is used to produce fine-grained pseudo-label assignment. Upon aligning the intrusion graphs geometrically from different granularities, the transferred intrusion knowledge can boost IID performance. Comprehensive experiments on several intrusion datasets demonstrate state-of-the-art performance of the GGA approach and validate the usefulness of GGA constituting components. | ['Chengzhong Xu', 'Kejiang Ye', 'Yang Wang', 'Hao Dai', 'Jiashu Wu'] | 2023-01-24 | null | null | null | null | ['network-intrusion-detection'] | ['miscellaneous'] | [ 3.88989002e-01 -1.03799857e-01 -2.76833922e-01 -2.82644808e-01
1.88552842e-01 -6.88414872e-01 6.94625318e-01 5.03983974e-01
-4.01095971e-02 3.78421694e-01 -2.97077298e-01 -5.16596138e-01
-6.91402495e-01 -1.19899476e+00 -1.24170586e-01 -4.30186927e-01
-1.95671692e-01 6.56417489e-01 2.56375998e-01 -2.91749775e-01
3.98817986e-01 1.13251007e+00 -1.18320715e+00 -3.20620425e-02
8.21894884e-01 1.19086349e+00 -3.73321861e-01 -2.29633562e-02
-6.75165534e-01 1.74185038e-01 -6.93046629e-01 -3.26151282e-01
7.80440807e-01 -6.69682771e-02 -5.14476717e-01 3.51845473e-01
2.33583450e-01 2.03043908e-01 -4.02245760e-01 1.23083055e+00
1.86465040e-01 7.19223320e-02 8.24648440e-01 -1.71093500e+00
-2.07143873e-01 1.41121224e-01 -8.21180344e-01 3.96812081e-01
4.19726372e-01 8.22857320e-02 7.01377273e-01 -4.35458809e-01
5.67317009e-01 1.23477232e+00 3.99673998e-01 2.64104843e-01
-1.19753373e+00 -7.70745814e-01 3.76335293e-01 1.75935552e-01
-1.60571790e+00 -8.37511907e-04 1.30855322e+00 -1.24374591e-01
8.61219168e-01 4.90706414e-01 5.94853044e-01 8.73479486e-01
1.38536647e-01 -6.99417666e-02 8.41316879e-01 -3.77455622e-01
4.32063043e-01 7.99265504e-02 1.95552930e-01 6.57750607e-01
7.95694113e-01 2.21324544e-02 -2.40472659e-01 -5.03818095e-02
7.06943691e-01 3.86545926e-01 1.68610424e-01 -7.31972933e-01
-6.59749448e-01 5.78240454e-01 9.19243157e-01 4.10974890e-01
-4.80672807e-01 -7.70382881e-01 4.46177483e-01 5.46234369e-01
2.22982213e-01 4.72937524e-01 -1.50644660e-01 4.03388262e-01
-5.74582696e-01 -3.95032465e-01 7.77438939e-01 8.91664922e-01
9.95562971e-01 1.30830020e-01 2.30894610e-01 5.84909916e-01
4.93562430e-01 3.78188729e-01 1.46432936e-01 -7.58396164e-02
5.07495821e-01 1.56272709e+00 -5.99331021e-01 -1.77361453e+00
-8.39640975e-01 -5.19180298e-01 -1.29811430e+00 4.44178134e-02
2.96769708e-01 3.62830341e-01 -9.78612185e-01 1.41326141e+00
8.50617707e-01 5.15206575e-01 -2.59457976e-02 6.83612525e-01
5.04726768e-01 -1.80131420e-02 2.59586185e-01 -7.31033236e-02
1.56369424e+00 -3.10296655e-01 -3.70452881e-01 -2.96152025e-01
6.48344874e-01 -6.06796443e-01 5.03273129e-01 4.25430853e-03
-3.09247166e-01 -4.48542029e-01 -1.24148202e+00 5.34020424e-01
-1.01875532e+00 -6.74759805e-01 4.70588058e-01 1.03967583e+00
-5.36604762e-01 2.24670261e-01 -5.11975110e-01 -6.22850955e-01
6.54698074e-01 6.28928423e-01 -4.36036676e-01 -2.34316498e-01
-1.17628002e+00 6.41689777e-01 8.05625796e-01 -2.15497822e-01
-1.72340125e-01 -6.49236679e-01 -6.71308517e-01 -2.84866333e-01
3.92930716e-01 -6.58395290e-01 1.19419627e-01 -5.57614505e-01
-1.15536332e+00 8.92136633e-01 2.81900883e-01 -6.29329324e-01
1.90910310e-01 6.43766403e-01 -1.18141305e+00 3.74426126e-01
4.88768565e-03 2.77068406e-01 1.10206437e+00 -1.31329668e+00
-6.64808452e-01 -8.45357418e-01 1.97644606e-02 2.13036284e-01
-5.27896643e-01 -3.01355928e-01 -2.97872454e-01 -6.78590715e-01
6.24479353e-01 -8.79021764e-01 -2.20055997e-01 -3.05172086e-01
-7.86272883e-01 -4.11879048e-02 1.45151675e+00 -2.32910201e-01
1.46628642e+00 -2.02894449e+00 5.44196321e-03 1.29011285e+00
6.60204172e-01 3.97338629e-01 -7.65577704e-02 3.29071879e-01
-2.37291142e-01 1.41725376e-01 -3.25303793e-01 9.64708105e-02
7.34166428e-03 2.97085911e-01 -1.96420670e-01 5.62182963e-01
2.12165654e-01 6.89739406e-01 -5.72370350e-01 -3.42067271e-01
6.65519059e-01 2.87017465e-01 -3.18129897e-01 -1.29406780e-01
-1.22554332e-01 5.24653971e-01 -7.55911112e-01 9.26227152e-01
1.13808882e+00 -1.80231303e-01 4.68198121e-01 -6.16242588e-01
1.20353796e-01 1.99427739e-01 -1.41062677e+00 1.20708096e+00
-3.13215375e-01 -2.48912767e-01 4.62172292e-02 -1.29846764e+00
1.54397130e+00 -1.38553157e-01 7.97117889e-01 -1.08735240e+00
2.70057917e-01 1.35596737e-01 1.07335225e-01 6.56185895e-02
-3.02489717e-02 1.06700912e-01 -2.68599898e-01 4.76436615e-01
-2.91686088e-01 8.95030946e-02 8.11395571e-02 2.01991230e-01
1.16467321e+00 -7.75403798e-01 5.41433811e-01 -4.30666625e-01
9.28054690e-01 5.70814498e-02 2.76714712e-01 6.74307168e-01
-1.97804213e-01 5.95841929e-02 2.01651633e-01 -6.96692646e-01
-6.48310244e-01 -1.21514404e+00 -1.94078475e-01 6.45656466e-01
5.26952446e-01 -2.92928934e-01 -4.32520598e-01 -1.11294460e+00
1.63231954e-01 4.70599413e-01 -4.79049683e-01 -5.66354752e-01
-4.83354211e-01 -7.80468583e-01 4.87936974e-01 8.42171311e-02
7.44186401e-01 -7.49416471e-01 -2.48243406e-01 2.28242487e-01
3.08095723e-01 -1.14499509e+00 -1.91126227e-01 2.36676842e-01
-7.88400769e-01 -1.48379827e+00 1.84847832e-01 -7.08012462e-01
8.30295682e-01 5.30876398e-01 8.19640040e-01 2.34862953e-01
-3.90744507e-01 2.91189343e-01 -4.01382864e-01 -1.96510315e-01
-2.28378326e-01 1.79400042e-01 3.28168809e-01 2.01741159e-01
7.69061863e-01 -1.00207484e+00 -5.95949352e-01 6.55403256e-01
-1.02173591e+00 -4.05792505e-01 6.25739515e-01 3.81135702e-01
5.34847736e-01 5.34331560e-01 5.83833098e-01 -7.32472777e-01
5.67599177e-01 -7.54431009e-01 -5.98122120e-01 1.95486024e-01
-6.91231728e-01 -1.11097671e-01 8.63871038e-01 -5.19066393e-01
-5.93221486e-01 -1.72477394e-01 2.96122849e-01 -4.45873797e-01
-5.25386930e-01 2.69426286e-01 -7.14241803e-01 -5.01417279e-01
5.75021267e-01 3.32116149e-02 3.59672308e-02 -3.80751610e-01
3.26854140e-01 6.45540535e-01 2.88649946e-01 -4.97254461e-01
1.24411547e+00 7.00258315e-01 6.35849535e-01 -9.45765853e-01
-4.33283061e-01 -5.91852009e-01 -1.03606129e+00 -6.95886537e-02
6.20127439e-01 -4.74626452e-01 -6.18462086e-01 1.76861241e-01
-6.95376158e-01 3.52555126e-01 -2.44093046e-01 1.92277044e-01
1.47951066e-01 5.03608406e-01 -4.90168072e-02 -4.81275260e-01
-3.16674531e-01 -9.09768164e-01 6.51943207e-01 2.68465906e-01
-1.57953277e-01 -1.22834468e+00 -1.11584194e-01 1.29694134e-01
3.16484690e-01 4.66560692e-01 1.30238676e+00 -1.38766718e+00
-4.56429124e-01 -4.30643529e-01 -6.48236275e-01 -7.78255537e-02
6.85006618e-01 -2.78671086e-01 -3.84790897e-01 -3.42898637e-01
-3.56869921e-02 4.49664652e-01 3.49204123e-01 -1.68663397e-01
9.25880432e-01 -2.55658984e-01 -6.40213072e-01 5.38342476e-01
1.16271067e+00 5.24025321e-01 4.37329561e-01 6.95046306e-01
1.13502109e+00 6.36384726e-01 3.68726522e-01 5.80974996e-01
3.24659795e-01 7.49684393e-01 6.34594917e-01 -8.14887136e-02
-2.21030653e-01 -1.96114168e-01 -3.35492760e-01 6.25960886e-01
3.96520466e-01 -2.25017965e-01 -1.06448543e+00 -2.67999303e-02
-1.44306207e+00 -5.19850135e-01 -1.99584812e-01 2.19162822e+00
1.00636482e-01 5.97735286e-01 2.75334865e-01 3.41810763e-01
1.06704366e+00 1.69291392e-01 -7.66819477e-01 -3.57719630e-01
-1.68500587e-01 3.45417917e-01 5.05292118e-01 2.56517857e-01
-1.14065075e+00 8.95791411e-01 4.55370188e+00 9.78574336e-01
-1.06979132e+00 -2.84143656e-01 2.96139956e-01 6.07971251e-01
-1.63845822e-01 -1.38430610e-01 -5.89931846e-01 3.91401231e-01
6.13950253e-01 8.81198719e-02 3.82007539e-01 5.26346266e-01
-1.06155060e-01 3.13638926e-01 -5.09865463e-01 8.90914857e-01
-1.61106020e-01 -1.06752348e+00 4.81211752e-01 4.35431480e-01
4.78108883e-01 -3.59766811e-01 1.08631924e-01 5.10680266e-02
1.10207617e-01 -7.12203503e-01 1.47182820e-02 3.21223110e-01
7.71812677e-01 -1.17529953e+00 4.59272742e-01 -4.26355228e-02
-1.77095389e+00 -1.83506206e-01 -2.12115139e-01 1.51262671e-01
-6.41390830e-02 7.02739954e-01 -1.22972655e+00 1.27296782e+00
3.27516228e-01 5.89159727e-01 -7.26223111e-01 8.32188427e-01
1.66095402e-02 1.06302388e-01 -6.28406525e-01 3.50884944e-01
2.04946011e-01 -6.23661280e-01 9.11998987e-01 8.00091326e-01
7.98679441e-02 2.10914090e-02 6.15110159e-01 3.30206960e-01
-8.85757655e-02 2.96628952e-01 -8.19373965e-01 1.56078935e-01
8.77241850e-01 1.11022830e+00 -1.27060091e+00 -6.76051825e-02
-3.66641551e-01 7.42479861e-01 4.56113145e-02 9.49613005e-02
-6.34184718e-01 -1.92807868e-01 1.01643014e+00 3.19897145e-01
1.59683570e-01 -2.35781252e-01 -7.25255907e-01 -8.67175639e-01
-2.99505800e-01 -9.07291412e-01 8.92396510e-01 5.52613512e-02
-1.67358994e+00 6.82923317e-01 1.39408913e-02 -1.46677625e+00
2.29103297e-01 -4.74713743e-01 -7.53534973e-01 5.51315844e-01
-1.28549671e+00 -1.25922287e+00 -3.51407409e-01 1.02040362e+00
6.70449436e-02 -3.39618146e-01 8.24267626e-01 3.50853682e-01
-9.16138411e-01 7.18247354e-01 -2.40089655e-01 2.06119902e-02
3.34361851e-01 -8.89459968e-01 3.74941558e-01 8.79433870e-01
1.74175918e-01 5.88119924e-01 4.26649719e-01 -1.15175664e+00
-1.22217381e+00 -1.48660672e+00 2.95036346e-01 -6.28843084e-02
8.40725124e-01 -3.46106410e-01 -9.98724043e-01 3.47099274e-01
-3.91508043e-01 8.78592506e-02 6.74358785e-01 -7.69043416e-02
-7.93289840e-01 -4.55134273e-01 -1.59775472e+00 6.84569597e-01
1.34335876e+00 -5.27135134e-01 -4.71902072e-01 2.23416731e-01
5.58012426e-01 1.26655698e-01 -9.69666958e-01 7.64670968e-01
1.61516860e-01 -7.54291773e-01 1.38030195e+00 -5.59271872e-01
-5.09085834e-01 -5.94825745e-01 -2.93474644e-01 -8.47265184e-01
-2.03680933e-01 -3.76922369e-01 -2.85581112e-01 1.32328546e+00
1.23393141e-01 -1.03004968e+00 8.35343480e-01 3.63216341e-01
7.16275573e-02 -3.89276356e-01 -1.10502136e+00 -9.82525826e-01
-3.98745537e-01 -3.22286189e-01 9.78671670e-01 1.20926499e+00
-2.39205495e-01 3.64257693e-01 2.82325614e-02 7.01615930e-01
8.57226014e-01 1.07359998e-01 6.80943727e-01 -1.77311242e+00
3.41642767e-01 -6.89199924e-01 -1.00780857e+00 -5.22660851e-01
1.15544096e-01 -1.16157031e+00 -8.29159498e-01 -1.06859374e+00
-5.15291631e-01 -8.33559930e-01 -6.12065911e-01 5.34244239e-01
3.77629608e-01 5.94973147e-01 1.17998049e-01 3.27759922e-01
-6.56051338e-01 2.28334755e-01 1.02023089e+00 -1.21015981e-01
-5.34594893e-01 1.09684907e-01 -6.63946927e-01 5.87790072e-01
1.22219634e+00 -5.19444823e-01 -5.78560233e-01 2.12781727e-01
-2.39306718e-01 -2.78142095e-01 2.72353292e-01 -1.18703425e+00
4.01099980e-01 -1.06238544e-01 4.50561404e-01 -5.11002898e-01
1.75029233e-01 -1.35570014e+00 3.29646349e-01 6.45729780e-01
3.55567902e-01 1.84527546e-01 3.34519774e-01 8.68587911e-01
-6.06182180e-02 2.57718861e-01 8.26857626e-01 1.68191791e-01
-8.62371624e-01 5.72237015e-01 -5.57756647e-02 3.11614852e-03
1.33888221e+00 -6.57181859e-01 -2.30945513e-01 2.24141330e-01
-5.13578951e-01 -7.61543587e-02 4.79183346e-01 5.93954325e-01
6.76433027e-01 -1.32332695e+00 -2.39639074e-01 6.81855500e-01
2.93763489e-01 -1.39928982e-01 3.46606731e-01 6.89979255e-01
-2.66969442e-01 3.54921758e-01 -5.23824930e-01 -6.08684361e-01
-1.15437925e+00 8.43907416e-01 1.27457827e-01 -5.13631165e-01
-6.68950081e-01 4.18772280e-01 8.75823125e-02 -5.89500070e-01
3.77425700e-02 2.78910637e-01 -3.18523526e-01 3.31727654e-01
4.26730514e-01 4.08344448e-01 4.64335531e-01 -8.34457934e-01
-8.66483748e-01 5.74553609e-01 -3.06991339e-01 4.94266182e-01
9.95553613e-01 -5.17514527e-01 -1.71608314e-01 -2.91947991e-01
1.14656556e+00 -7.40193948e-02 -8.04865420e-01 -3.08579713e-01
3.88285756e-01 -4.62630302e-01 -2.71980375e-01 -7.53902197e-01
-9.97618020e-01 3.87567073e-01 6.47009730e-01 8.00710797e-01
1.33959079e+00 -1.69523746e-01 5.79651833e-01 7.92978853e-02
3.02455723e-01 -6.91150129e-01 2.41862372e-01 4.31662381e-01
3.55605245e-01 -9.62860823e-01 -1.01262428e-01 -7.40645885e-01
-3.55575025e-01 1.17792249e+00 7.00525522e-01 -2.22956315e-01
8.40120733e-01 3.34706120e-02 -7.79389068e-02 -5.35022140e-01
8.90015438e-02 -4.81068760e-01 5.42518318e-01 1.09896743e+00
-5.01191795e-01 1.51763633e-01 3.70167941e-02 2.66762018e-01
-2.52205670e-01 -8.97211194e-01 -5.53817376e-02 9.47283983e-01
-4.73269910e-01 -1.32213891e+00 -5.15446424e-01 4.89902288e-01
1.63894184e-02 9.41904634e-02 -6.21695757e-01 1.00552845e+00
2.02501446e-01 1.20306790e+00 1.90172911e-01 -7.19535232e-01
5.10783613e-01 2.84591001e-02 2.10234731e-01 -4.02308404e-01
-4.04644400e-01 -2.00502172e-01 -1.48588672e-01 -5.28671026e-01
-1.56702980e-01 -1.21915132e-01 -1.25692642e+00 -4.73874867e-01
-1.15323067e-01 3.89442667e-02 6.10830784e-01 9.91380334e-01
6.26423657e-01 6.11709535e-01 9.67647672e-01 -4.11264807e-01
-7.24392608e-02 -3.89774650e-01 -5.60758889e-01 5.73345780e-01
-1.82119422e-02 -9.48288858e-01 -4.09113824e-01 -6.11754119e-01] | [6.576463222503662, 5.881117820739746] |
5014e810-6cba-4b7f-af64-3fec601e6baa | small-sample-hyperspectral-image | null | null | https://doi.org/10.3390/s23052499 | https://www.mdpi.com/1424-8220/23/5/2499 | Small Sample Hyperspectral Image Classification Based on the Random Patches Network and Recursive Filtering | In recent years, different deep learning frameworks were introduced for hyperspectral image (HSI) classification. However, the proposed network models have a higher model complexity, and do not provide high classification accuracy if few-shot learning is used. This paper presents an HSI classification method that combines random patches network (RPNet) and recursive filtering (RF) to obtain informative deep features. The proposed method first convolves image bands with random patches to extract multi-level deep RPNet features. Thereafter, the RPNet feature set is subjected to dimension reduction through principal component analysis (PCA), and the extracted components are filtered using the RF procedure. Finally, the HSI spectral features and the obtained RPNet–RF features are combined to classify the HSI using a support vector machine (SVM) classifier. In order to test the performance of the proposed RPNet–RF method, some experiments were performed on three widely known datasets using a few training samples for each class, and classification results were compared with those obtained by other advanced HSI classification methods adopted for small training samples. The comparison showed that the RPNet–RF classification is characterized by higher values of such evaluation metrics as overall accuracy and Kappa coefficient. | ['Dmitry Uchaev', 'Denis Uchaev'] | 2023-02-23 | null | null | null | sensors-2023-2 | ['few-shot-image-classification', 'dimensionality-reduction'] | ['computer-vision', 'methodology'] | [ 5.98682046e-01 -4.85331982e-01 2.25979201e-02 -2.62877107e-01
-4.29553092e-01 -1.48852959e-01 3.89407486e-01 1.78337712e-02
-3.47262859e-01 8.20840359e-01 -3.62887457e-02 1.16578288e-01
-8.06328118e-01 -1.36134636e+00 -1.16420142e-01 -1.23061550e+00
-1.51703626e-01 -1.41730011e-01 1.04601823e-01 -7.54065886e-02
2.69055992e-01 8.51449370e-01 -1.95354211e+00 5.63698001e-02
1.16999352e+00 1.45315945e+00 1.34765491e-01 4.66348827e-01
2.48412527e-02 7.64495730e-01 -4.10283417e-01 2.73205906e-01
4.54014957e-01 -3.49743098e-01 -5.53057849e-01 1.31274924e-01
1.08761065e-01 -2.19026923e-01 -3.03206980e-01 1.10725701e+00
6.91255510e-01 8.79977703e-01 7.87892401e-01 -9.39391315e-01
-6.31696522e-01 3.35625589e-01 -5.27886152e-01 1.26805171e-01
-1.10522181e-01 1.70793980e-01 8.49620163e-01 -8.90691638e-01
1.78515255e-01 8.92496109e-01 8.78479660e-01 -5.14112562e-02
-1.14542902e+00 -6.75129175e-01 -4.39954609e-01 6.42515540e-01
-1.48040056e+00 -1.07022345e-01 9.11305010e-01 -4.62331682e-01
8.21627140e-01 2.60854542e-01 9.32071567e-01 6.09481514e-01
1.37797073e-01 3.29132527e-01 1.33009052e+00 -3.71377617e-01
5.02566636e-01 7.62284473e-02 7.98782945e-01 5.05131185e-01
3.50575238e-01 2.42359877e-01 -4.09158021e-02 1.33533217e-02
3.69598418e-01 3.21524143e-01 -4.61549759e-01 -8.39117169e-02
-5.82779109e-01 1.02004409e+00 1.07295263e+00 7.08487391e-01
-7.21780539e-01 -5.61977446e-01 3.59874099e-01 1.09063163e-01
3.84816140e-01 3.24379116e-01 -1.79816678e-01 5.44126928e-01
-9.62746859e-01 -3.33782375e-01 6.72557831e-01 9.36029255e-02
1.19776881e+00 1.94304690e-01 -1.68065727e-01 1.22346842e+00
1.63887277e-01 5.52230597e-01 6.03006482e-01 -4.85916793e-01
-6.35587275e-02 6.13358796e-01 -3.01747471e-01 -1.56320715e+00
-5.52659154e-01 -5.67102849e-01 -1.34667635e+00 2.40428492e-01
-2.77027667e-01 -1.49726182e-01 -9.65566635e-01 1.13146579e+00
-1.24192365e-01 1.47435129e-01 5.59320807e-01 8.77559364e-01
1.26212716e+00 1.24812829e+00 3.12113971e-01 -2.35587403e-01
9.29752886e-01 -6.86861455e-01 -4.06712472e-01 -1.39374256e-01
4.86337654e-02 -4.04154718e-01 1.04185236e+00 4.41135079e-01
-2.81678706e-01 -9.29317474e-01 -1.23744380e+00 2.91762352e-01
-6.24182343e-01 5.36359191e-01 6.86886311e-01 5.95613301e-01
-8.20647597e-01 8.57809603e-01 -5.08713484e-01 -6.64855301e-01
7.07147539e-01 3.38279426e-01 -3.21044087e-01 -2.56017327e-01
-1.17398751e+00 5.80203414e-01 7.38705873e-01 4.02328402e-01
-7.45472491e-01 -5.32640755e-01 -8.18391442e-01 3.73875737e-01
-1.51635585e-02 -2.04314306e-01 5.12235224e-01 -1.12308288e+00
-1.60829556e+00 3.73938143e-01 1.77002341e-01 -1.73474789e-01
-1.91951692e-01 7.05840066e-02 -5.69481194e-01 4.29973722e-01
-9.27683190e-02 2.46147677e-01 7.21293747e-01 -1.02638578e+00
-4.64052409e-01 -4.31055397e-01 -5.18942587e-02 2.51980305e-01
-6.02745295e-01 -3.58293325e-01 5.31245947e-01 -1.74686506e-01
3.73746783e-01 -5.00551820e-01 -1.37497619e-01 -2.87547886e-01
-2.78385490e-01 -2.28497669e-01 9.79196310e-01 -7.73510754e-01
6.96874321e-01 -2.44908333e+00 -9.76272598e-02 6.44604504e-01
-2.49494135e-01 6.69121504e-01 -2.73692936e-01 2.22487047e-01
-2.20741928e-01 7.02544302e-02 -5.10286093e-01 4.86587137e-01
-2.59550661e-01 -8.75618681e-02 -1.37607371e-02 4.89288330e-01
1.19095959e-01 4.84875381e-01 -7.25083768e-01 -2.87265450e-01
6.83272779e-01 6.91004932e-01 -1.15152366e-01 2.70124733e-01
6.48292825e-02 7.50048757e-02 -2.44689628e-01 7.57679760e-01
1.04965234e+00 3.66841666e-02 -1.14649594e-01 -7.93288648e-01
-2.72559434e-01 -3.55734020e-01 -1.05248654e+00 1.03487325e+00
-3.20084572e-01 7.60106325e-01 -2.04355136e-01 -1.24280357e+00
1.42870796e+00 1.38782293e-01 5.57286561e-01 -4.14388090e-01
3.03111851e-01 1.33207038e-01 -1.93358809e-02 -6.95685327e-01
2.70430088e-01 -2.26686448e-01 4.36107069e-01 -6.72965422e-02
2.39566222e-01 -1.25995502e-01 1.72750443e-01 -4.65943992e-01
6.73315227e-01 -1.52947724e-01 4.07121688e-01 -2.95938164e-01
9.63458240e-01 3.62940282e-01 4.05205309e-01 5.26593566e-01
-1.12580620e-01 2.25113451e-01 8.29930678e-02 -6.20958805e-01
-7.44827867e-01 -6.41793549e-01 -4.41171378e-01 8.75460804e-01
7.31292740e-02 2.39208132e-01 -3.03116739e-01 -3.66249979e-01
-1.84138611e-01 6.76538289e-01 -3.07278425e-01 -2.27284327e-01
9.44331735e-02 -1.44558132e+00 3.77531618e-01 1.56568214e-01
1.21266985e+00 -1.36734593e+00 -5.57365537e-01 1.84142053e-01
3.59291621e-02 -4.26560402e-01 6.93516135e-01 2.56569207e-01
-8.01237881e-01 -1.19066429e+00 -6.26162231e-01 -9.09766436e-01
4.32913005e-01 5.83144486e-01 4.16044593e-01 -2.31079131e-01
-5.19129097e-01 -2.94621345e-02 -4.72793877e-01 -1.33767039e-01
-3.66983041e-02 7.08865225e-02 -2.65320450e-01 2.65964687e-01
5.79413533e-01 -7.61261940e-01 -6.16791844e-01 -2.39374265e-02
-1.04723799e+00 -2.39079058e-01 9.15916860e-01 9.24931645e-01
4.91647840e-01 1.03734839e+00 4.76265758e-01 -5.17561376e-01
5.57892859e-01 -3.95868838e-01 -5.94411731e-01 5.82879335e-02
-4.56882447e-01 -4.90965366e-01 8.13804030e-01 -1.71512857e-01
-1.34610116e+00 1.07800001e-02 -2.55919516e-01 -1.67163476e-01
-5.23398697e-01 1.02143359e+00 -8.85452628e-02 -5.23018599e-01
9.67533290e-01 4.72860157e-01 4.02317457e-02 -3.10481787e-01
1.03415661e-02 9.12362158e-01 3.18988621e-01 5.04588522e-02
7.20146060e-01 2.27112770e-01 1.64825901e-01 -1.19818211e+00
-9.03388560e-01 -3.71926814e-01 -7.28692174e-01 -2.12938264e-01
8.88768435e-01 -8.82677019e-01 -5.45741916e-01 8.00162017e-01
-7.19072998e-01 -1.19126044e-01 -2.12516591e-01 7.51989067e-01
-2.10999146e-01 3.23973328e-01 -5.26856065e-01 -8.17684412e-01
-6.77276611e-01 -8.12338769e-01 3.96854252e-01 7.47359157e-01
4.58413482e-01 -6.49957538e-01 -2.17984412e-02 -7.23763704e-02
6.69918001e-01 2.96516448e-01 1.20327878e+00 -6.56094432e-01
-1.46404833e-01 -5.21482110e-01 -5.76460540e-01 8.32969904e-01
2.54651397e-01 2.59577155e-01 -1.20813036e+00 -1.86671406e-01
1.70375586e-01 -3.01918328e-01 1.10472655e+00 6.91400826e-01
1.44242334e+00 -1.76479802e-01 -5.29310144e-02 8.63420129e-01
2.26803136e+00 5.12943208e-01 8.65831017e-01 4.17068154e-01
6.17822289e-01 3.95960480e-01 6.25971437e-01 3.47109586e-01
-2.87610799e-01 -9.86541733e-02 4.55092013e-01 -2.43745193e-01
6.26956522e-02 1.73063904e-01 1.15696065e-01 5.13994932e-01
-4.67834175e-01 -7.71788582e-02 -8.42381597e-01 1.12427019e-01
-1.57356727e+00 -1.35813034e+00 -5.84988669e-02 1.88711321e+00
2.23504603e-01 -2.15827420e-01 -1.95408180e-01 7.21229255e-01
8.89798462e-01 4.47578043e-01 -5.06935298e-01 8.88914242e-02
-4.04840946e-01 5.95156252e-01 4.79698420e-01 1.70241266e-01
-1.42653811e+00 7.10110903e-01 5.65126705e+00 7.29180336e-01
-1.54966271e+00 -5.11298515e-02 5.19703805e-01 4.26998705e-01
2.24998057e-01 -1.71337664e-01 -3.53414863e-01 6.49990812e-02
7.97404051e-01 -1.90709159e-01 5.79001307e-01 8.08786571e-01
1.97498485e-01 -2.92664677e-01 -2.99120337e-01 1.09939361e+00
1.09864779e-01 -1.06175065e+00 9.17481259e-02 -1.76775157e-01
6.76875353e-01 8.89638439e-02 -6.22173436e-02 4.55751300e-01
-8.08340758e-02 -8.70106757e-01 -5.76132536e-02 9.64538753e-01
5.09729803e-01 -9.49606717e-01 1.15389717e+00 1.80500865e-01
-1.22106028e+00 -6.98431194e-01 -1.10329568e+00 -4.74579148e-02
-4.85083461e-01 8.43279898e-01 -5.65845668e-01 8.85760486e-01
6.75757945e-01 1.06608224e+00 -6.67689800e-01 1.41601408e+00
-3.01250499e-02 6.30839884e-01 -4.37353104e-02 -2.35657558e-01
2.63725907e-01 -5.77834904e-01 3.45292777e-01 1.04818654e+00
6.81757987e-01 2.84372389e-01 7.21584111e-02 7.62210429e-01
2.15435684e-01 4.53413814e-01 -6.42414987e-01 -2.39682376e-01
2.92981893e-01 1.56556201e+00 -8.49239349e-01 -4.31621373e-01
-2.51865059e-01 5.79346359e-01 -1.27483606e-01 4.28580940e-01
-6.05567098e-01 -9.75721419e-01 2.04220906e-01 -2.79533714e-01
2.53046304e-01 6.51655570e-02 3.41599807e-02 -9.99661267e-01
-4.45209205e-01 -6.38309002e-01 5.19411981e-01 -1.09648681e+00
-1.54074037e+00 7.78226256e-01 -1.25065848e-01 -1.26396215e+00
2.36469626e-01 -7.95841634e-01 -6.18058503e-01 1.06569469e+00
-1.58844614e+00 -8.20577860e-01 -1.12788701e+00 5.95193863e-01
3.70266855e-01 -4.80737478e-01 1.08205652e+00 1.26639202e-01
-7.79500008e-01 -5.07760905e-02 3.93235803e-01 1.63812310e-01
1.78100094e-01 -7.28138804e-01 -5.35180449e-01 8.41742814e-01
-4.18542325e-01 1.31174147e-01 1.48270592e-01 -4.63648856e-01
-1.12866414e+00 -1.42837310e+00 2.60861993e-01 7.13692069e-01
2.66542614e-01 3.01265746e-01 -9.80568528e-01 2.65598089e-01
-2.56188307e-02 3.13253254e-01 9.87114370e-01 -1.95526540e-01
-1.15934864e-01 -6.88477516e-01 -1.58847010e+00 5.38438186e-02
3.43654424e-01 -4.74784106e-01 -4.47636992e-01 3.83439004e-01
3.80392879e-01 3.77567768e-01 -9.77149069e-01 7.62921929e-01
4.83897269e-01 -1.11148369e+00 8.58566523e-01 -1.42814517e-01
4.58388031e-01 -4.05369937e-01 -4.10303205e-01 -1.49536705e+00
-9.10741866e-01 3.37095678e-01 4.23639327e-01 1.04463398e+00
1.58791810e-01 -7.56643534e-01 5.46743035e-01 2.25904696e-02
-9.31150317e-02 -2.92188644e-01 -5.34498215e-01 -5.63091576e-01
-4.27040935e-01 -9.57008302e-02 6.72805130e-01 8.63228917e-01
-2.84369498e-01 3.60598207e-01 -9.20959860e-02 3.83090496e-01
6.88568294e-01 2.12905005e-01 2.96950519e-01 -1.73160112e+00
4.47718464e-02 -4.50325489e-01 -6.92895055e-01 -2.61752214e-02
1.65307090e-01 -9.89139497e-01 1.19513772e-01 -1.83967960e+00
2.61847168e-01 -4.14027005e-01 -6.36197686e-01 5.48004270e-01
5.71212098e-02 4.78994191e-01 8.05507302e-02 3.63603771e-01
1.23953953e-01 6.34493589e-01 8.65182459e-01 -3.91444325e-01
-5.76163292e-01 2.54038006e-01 -5.74592412e-01 5.78246236e-01
1.07951391e+00 -2.70681590e-01 -3.20563942e-01 2.37754942e-03
-2.81102687e-01 -5.72082922e-02 4.16357964e-01 -1.94113314e+00
1.16485484e-01 -1.92498729e-01 1.00254083e+00 -7.45799005e-01
2.46809080e-01 -8.25205207e-01 3.56519938e-01 6.37276947e-01
-1.43271461e-01 -7.98970520e-01 2.25901574e-01 2.69080371e-01
-2.45946825e-01 -4.51023787e-01 1.10806823e+00 -1.53249398e-01
-1.16479850e+00 4.44214135e-01 -4.72342342e-01 -9.06564057e-01
1.05189526e+00 -3.49737912e-01 -3.38706166e-01 -9.10950731e-03
-8.90863478e-01 -4.47726488e-01 4.07574838e-03 -1.42802402e-01
8.18905413e-01 -1.17603409e+00 -5.49073815e-01 4.15880322e-01
1.35008432e-02 -1.52532354e-01 4.70376819e-01 7.22704411e-01
-8.00021529e-01 1.88825577e-01 -7.65639484e-01 -6.10673368e-01
-1.13302398e+00 4.60887849e-01 6.40759766e-01 1.01827681e-01
-5.39500237e-01 5.65092325e-01 -3.73817831e-01 -5.38842201e-01
1.13710091e-01 9.83219221e-02 -8.65376174e-01 3.88842911e-01
4.65730935e-01 6.42643154e-01 1.37291208e-01 -6.75068021e-01
-2.08811954e-01 7.74458587e-01 5.59373438e-01 1.33513927e-01
1.89914513e+00 2.02142507e-01 -5.14243484e-01 3.49531472e-01
1.27280164e+00 -4.38076288e-01 -8.43069673e-01 -2.92507589e-01
-2.96552956e-01 -4.94156450e-01 4.53721374e-01 -9.23819780e-01
-1.14918256e+00 1.00325537e+00 1.21194339e+00 3.53579462e-01
1.65757334e+00 -5.62512338e-01 2.57016867e-01 7.24905074e-01
1.42182812e-01 -8.79887938e-01 -2.73739845e-01 5.05950630e-01
6.68366790e-01 -1.07883155e+00 1.37907997e-01 -2.98915893e-01
-3.48246574e-01 1.60889351e+00 6.18963182e-01 -3.29396516e-01
9.81087804e-01 -3.60623449e-01 -2.66590476e-01 -2.19796598e-01
-4.12442982e-02 -4.64201003e-01 1.98860496e-01 7.38799930e-01
2.62917072e-01 1.34760141e-01 -2.75047600e-01 5.95139623e-01
-1.35938346e-01 3.04459393e-01 3.82544726e-01 7.16882467e-01
-1.07424796e+00 -3.26645851e-01 -4.07833874e-01 8.50836039e-01
-4.86056320e-02 -3.22387889e-02 -6.24502636e-02 5.52037120e-01
3.25965971e-01 1.02365971e+00 3.06408480e-02 -7.01057315e-01
2.45103046e-01 1.20552219e-01 1.06755771e-01 -5.33394814e-01
-2.12692261e-01 -1.56781301e-01 -3.17917734e-01 -1.05873741e-01
-7.29076266e-01 -2.74752021e-01 -8.93501759e-01 -2.51369804e-01
-3.41902643e-01 3.77074391e-01 7.73265064e-01 8.06883752e-01
-1.52040292e-02 5.27761638e-01 1.07070363e+00 -1.23304665e+00
-4.61028785e-01 -1.18529427e+00 -1.04682064e+00 1.16495699e-01
1.91790044e-01 -4.71630335e-01 -8.92406344e-01 -2.82016277e-01] | [9.910750389099121, -1.5504357814788818] |
5a9ebd7c-3656-483d-bf6a-1a76be4e9ee6 | simvtp-simple-video-text-pre-training-with | 2212.03490 | null | https://arxiv.org/abs/2212.03490v1 | https://arxiv.org/pdf/2212.03490v1.pdf | SimVTP: Simple Video Text Pre-training with Masked Autoencoders | This paper presents SimVTP: a Simple Video-Text Pretraining framework via masked autoencoders. We randomly mask out the spatial-temporal tubes of input video and the word tokens of input text and then feed them into a unified autencoder to reconstruct the missing pixels and words. Our SimVTP has several properties: 1) Thanks to the unified autoencoder, SimVTP reconstructs the masked signal of one modality with the help from another modality, which implicitly learns the cross-modal alignment between video tubes and text tokens. 2) SimVTP not only benefits from a high video masking ratio (e.g. 90%) due to the temporal redundancy of video, but also needs a high text masking ratio (e.g. 75%), which is much higher than BERT (e.g. 15%), to achieve optimal performance. This is because the aid of video modality makes text reconstruction less challenging, which thus needs a higher mask ratio to make the pretext harder for useful feature learning. 3) Equipping SimVTP with video-text contrastive learning (VTC) and video-text matching (VTM), which are two commonly used cross-modal training strategies, could further improve the transferable performance significantly. 4) SimVTP is dataefficent, e.g., pre-training only on 10% data of WebVid-2M, SimVTP achieves surprisingly good results (43.8 R@1) on MSRVTT, which is far above recent state-of-the-art methods pre-trained on both CC3M and WebVid-2M. We transfer our pre-trained model to various downstream tasks and achieve superior performance. The codes and models will be released at https://github.com/mayuelala/SimVTP. | ['Xiu Li', 'Yin Shan', 'Tianyu Yang', 'Yue Ma'] | 2022-12-07 | null | null | null | null | ['moment-retrieval'] | ['computer-vision'] | [ 1.57446519e-01 -2.63356678e-02 -3.16909075e-01 2.45312825e-02
-6.73705101e-01 -5.38028181e-01 6.85996056e-01 -5.69960594e-01
-5.71039855e-01 2.35814139e-01 2.34503895e-01 -4.11101341e-01
4.39192593e-01 -5.60607612e-01 -1.16262507e+00 -6.39906168e-01
2.05341443e-01 2.65587389e-01 3.90471965e-01 -1.59492388e-01
-2.03492671e-01 -1.49664506e-01 -1.74693930e+00 8.07211220e-01
8.10397744e-01 9.03527260e-01 6.75465822e-01 7.33207166e-01
-3.01303387e-01 8.41847718e-01 -3.69524777e-01 -4.98816818e-01
5.84891200e-01 -4.71770465e-01 -6.19876385e-01 1.07691094e-01
6.34607434e-01 -7.18068182e-01 -1.00175464e+00 8.29179108e-01
2.67685771e-01 1.33924156e-01 6.96888983e-01 -1.13956583e+00
-5.90869308e-01 8.25628459e-01 -7.76328802e-01 1.05149239e-01
1.73825085e-01 4.33665812e-01 1.16101158e+00 -1.08664703e+00
5.38163126e-01 1.07616639e+00 4.38414603e-01 6.34169042e-01
-1.13011491e+00 -9.90143538e-01 1.53310686e-01 2.82696098e-01
-1.29012585e+00 -5.23151696e-01 4.88693744e-01 -4.81141567e-01
9.92356181e-01 3.09550196e-01 5.83243847e-01 1.37078869e+00
9.11968574e-02 1.08305275e+00 9.62655962e-01 -3.29325199e-01
-2.02255443e-01 1.51763707e-01 -2.80880332e-01 7.63422608e-01
-1.52882129e-01 1.91796988e-01 -7.26687431e-01 4.54545200e-01
9.26494598e-01 1.74728274e-01 -5.36890984e-01 3.54937874e-02
-1.31906927e+00 6.34668946e-01 3.85984480e-01 5.69003403e-01
-9.55160633e-02 1.18244395e-01 2.68385768e-01 6.85742021e-01
2.01341048e-01 9.90502983e-02 -3.95007610e-01 -1.84450284e-01
-1.20179355e+00 -1.86321482e-01 4.94356960e-01 1.09598207e+00
9.14138198e-01 4.21964467e-01 -2.35721454e-01 8.36621404e-01
2.80635655e-01 8.40976000e-01 7.55009830e-01 -8.26007664e-01
1.07317746e+00 3.54538172e-01 -3.56760532e-01 -7.41024315e-01
1.75366251e-04 -1.28365561e-01 -1.01026964e+00 1.48392871e-01
4.64224339e-01 -1.26279652e-01 -1.22027707e+00 1.54840970e+00
-1.34652883e-01 3.54557306e-01 6.90265000e-02 9.19226110e-01
1.05388308e+00 1.02340055e+00 -1.71792224e-01 -6.98549077e-02
1.46960759e+00 -9.93674397e-01 -6.46864355e-01 -4.50416505e-01
4.35870945e-01 -8.75679076e-01 1.33520830e+00 2.21312597e-01
-1.05064321e+00 -7.05827057e-01 -1.11296666e+00 -1.68291762e-01
-1.64783984e-01 6.78732395e-02 3.00206602e-01 3.34255308e-01
-1.03817058e+00 5.84899843e-01 -8.40861499e-01 -1.11207314e-01
3.49704206e-01 3.09622169e-01 -4.42653239e-01 -3.76645178e-01
-1.15127850e+00 5.31599939e-01 4.32540536e-01 -2.23414183e-01
-9.37648356e-01 -5.83758295e-01 -9.20568407e-01 1.06543943e-01
5.25144815e-01 -4.89104569e-01 1.10420156e+00 -1.18969953e+00
-1.43575251e+00 8.09421897e-01 -3.21958512e-01 -3.52182627e-01
7.22223878e-01 -1.18700862e-01 -4.14046645e-01 5.38161695e-01
1.03316568e-01 9.85244215e-01 1.44532359e+00 -9.62167203e-01
-5.55671513e-01 -1.21928109e-02 -9.16476101e-02 4.25740987e-01
-4.90915447e-01 -7.31837451e-02 -1.07427704e+00 -1.01625752e+00
9.56394747e-02 -8.88245106e-01 2.25423455e-01 -8.84372815e-02
-2.23083973e-01 1.04822338e-01 9.83012021e-01 -9.44385052e-01
1.00835991e+00 -2.27684903e+00 3.52040023e-01 -6.11762665e-02
3.94780427e-01 2.15636551e-01 -4.85722661e-01 3.32105368e-01
-2.57475555e-01 4.87345532e-02 -1.50543064e-01 -5.07123590e-01
-2.13067517e-01 2.77988970e-01 -3.13550889e-01 3.30102235e-01
6.34121522e-02 1.10044801e+00 -4.77622360e-01 -4.87014323e-01
4.85265523e-01 3.09370518e-01 -7.18595088e-01 2.46542901e-01
-3.26003313e-01 4.04433817e-01 -1.07532375e-01 4.89055753e-01
6.35870039e-01 -3.28967243e-01 1.48909837e-01 -4.02290434e-01
-5.48457354e-02 1.76502615e-01 -1.08528984e+00 1.62401819e+00
-4.59460169e-01 9.86225307e-01 3.99011150e-02 -1.02882671e+00
3.55703354e-01 4.69697475e-01 5.55483162e-01 -9.89353955e-01
3.13100904e-01 7.28664026e-02 -2.23878980e-01 -5.48645675e-01
4.45402652e-01 -4.84308898e-02 1.77391768e-01 5.22997737e-01
4.50893581e-01 7.29428977e-02 1.38389423e-01 4.07463133e-01
9.90210831e-01 3.11085116e-02 -4.80516739e-02 1.67510882e-02
3.59143525e-01 -2.79554427e-01 3.41475546e-01 7.60379493e-01
1.29637616e-02 7.63430476e-01 2.74060905e-01 -5.22179380e-02
-1.20863175e+00 -8.83154690e-01 8.77932161e-02 1.02851617e+00
1.93514571e-01 -5.12767375e-01 -6.68293297e-01 -6.94526553e-01
-1.32031143e-01 3.88627768e-01 -4.70268041e-01 -7.67423660e-02
-6.53133988e-01 -3.67213130e-01 5.40112197e-01 5.98130167e-01
8.37078214e-01 -8.74993742e-01 -4.94700372e-02 -5.17259538e-02
-5.89825094e-01 -1.49348307e+00 -8.23358655e-01 2.18707681e-01
-8.71430218e-01 -8.56130362e-01 -8.54107738e-01 -8.88477206e-01
5.18354535e-01 7.46082187e-01 9.35043037e-01 3.14916104e-01
1.04131065e-01 2.77927577e-01 -5.58969557e-01 3.62501815e-02
-4.93468642e-01 -1.12858027e-01 -4.17151563e-02 1.28704146e-01
1.85230762e-01 -5.68216920e-01 -4.71336544e-01 4.48826015e-01
-1.19634485e+00 5.00134349e-01 7.82475054e-01 1.01513362e+00
4.90827918e-01 1.10872544e-01 -2.29349211e-01 -5.51097453e-01
-1.35472447e-01 -3.37918222e-01 -5.76995134e-01 6.79571927e-02
-2.23789498e-01 1.06478572e-01 7.48843372e-01 -8.17260504e-01
-8.30772042e-01 2.73633073e-03 -3.14352602e-01 -1.01939881e+00
6.52358169e-03 2.21357688e-01 -1.97005883e-01 7.30539560e-02
5.22584796e-01 5.94104946e-01 8.15754309e-02 -3.82666320e-01
3.76117736e-01 6.80717587e-01 5.13619900e-01 -3.31759661e-01
1.18046057e+00 4.55922157e-01 -4.86527622e-01 -8.81486118e-01
-4.61377472e-01 -4.21284020e-01 -3.50790501e-01 -1.42207295e-01
1.00009894e+00 -1.28752625e+00 -5.36044300e-01 5.51119328e-01
-8.20269465e-01 -7.37915993e-01 -1.07943855e-01 6.78063273e-01
-4.16355997e-01 5.90200722e-01 -8.13234687e-01 -3.52571517e-01
-1.73668951e-01 -1.26286769e+00 1.02633011e+00 -1.31152034e-01
8.05348009e-02 -6.24186814e-01 -4.19572711e-01 6.21589899e-01
1.06712319e-01 -3.54169577e-01 6.73687935e-01 -3.97981912e-01
-8.93876195e-01 4.75249998e-02 -2.60140777e-01 4.48727816e-01
4.07149680e-02 -1.94551185e-01 -1.06462586e+00 -5.18025756e-01
1.08674839e-01 -2.61344463e-01 1.19195271e+00 3.33817989e-01
1.11750329e+00 -5.20645440e-01 -2.16331467e-01 8.87883842e-01
1.27155256e+00 1.35012463e-01 8.47467840e-01 2.19075203e-01
9.35953140e-01 4.14338410e-01 4.09687430e-01 1.82994887e-01
3.11580896e-01 7.29542494e-01 3.92536879e-01 -2.26979971e-01
-4.94817823e-01 -4.56923544e-01 8.08112681e-01 1.15102744e+00
9.67042595e-02 -3.05551469e-01 -7.54296005e-01 2.97877729e-01
-1.86209500e+00 -1.02360559e+00 -8.85844529e-02 1.92953432e+00
7.87584841e-01 1.27709031e-01 6.21780530e-02 2.43177846e-01
6.63987935e-01 3.53843123e-01 -4.09906179e-01 7.98006654e-02
-2.93245852e-01 2.21361369e-01 5.34196973e-01 4.66696829e-01
-1.02316666e+00 1.02122152e+00 5.54994202e+00 1.17498946e+00
-1.21239591e+00 2.84743786e-01 4.74665999e-01 -2.56614417e-01
-3.81080925e-01 -1.71064407e-01 -6.44120932e-01 8.39945793e-01
9.01455760e-01 3.27828288e-01 7.98258305e-01 4.66595262e-01
-5.92172891e-02 1.23093590e-01 -1.08513200e+00 1.24247503e+00
7.64491633e-02 -1.35608149e+00 8.61718059e-02 -5.88424429e-02
6.62531018e-01 3.42951268e-01 1.65884137e-01 4.94446009e-01
2.35167950e-01 -8.93445611e-01 9.67170298e-01 -9.66083705e-02
1.27752090e+00 -3.77379328e-01 5.09928405e-01 3.54781300e-01
-1.32853222e+00 -1.31290734e-01 -3.30712438e-01 2.75487423e-01
-1.73552126e-01 5.18342853e-01 -5.64246655e-01 6.32517278e-01
1.03441739e+00 9.48497117e-01 -3.42726588e-01 5.17675757e-01
-2.69400895e-01 7.66732633e-01 -4.56053734e-01 1.78542525e-01
2.11929023e-01 -1.55401617e-01 5.47431946e-01 1.23628724e+00
3.42043966e-01 6.67877379e-04 5.88817745e-02 6.95741296e-01
-3.98238689e-01 5.69535941e-02 -6.18232787e-01 -6.80702105e-02
3.79062235e-01 9.89671528e-01 -4.51797515e-01 -4.42959875e-01
-9.05530691e-01 1.33010209e+00 9.18890983e-02 6.37817085e-01
-9.59994495e-01 -1.41186059e-01 5.48910975e-01 8.45531449e-02
8.00537586e-01 -1.17117435e-01 -1.74862057e-01 -1.79866564e+00
2.87343770e-01 -1.23015463e+00 3.55429411e-01 -7.63427556e-01
-9.37049985e-01 7.10728943e-01 -2.38848910e-01 -1.58134210e+00
-2.07054332e-01 -6.41402960e-01 -4.10825521e-01 7.66496778e-01
-1.37838733e+00 -1.18075061e+00 -3.16535980e-01 9.68998849e-01
8.76203537e-01 -3.56509387e-01 4.19758201e-01 4.86745834e-01
-5.46631038e-01 9.20043707e-01 1.34788290e-01 3.45046848e-01
7.90615439e-01 -9.34407890e-01 4.67951566e-01 1.11215949e+00
4.25345749e-01 2.20014021e-01 4.93445784e-01 -5.60966074e-01
-1.64149165e+00 -1.16434705e+00 4.95492756e-01 -4.59047347e-01
6.96111739e-01 -6.54348195e-01 -1.02425933e+00 8.57489347e-01
4.18740332e-01 -1.85251191e-01 3.40660989e-01 -2.60056615e-01
-7.07533479e-01 -1.83996245e-01 -7.63797283e-01 9.59313750e-01
1.04174900e+00 -7.35111415e-01 -6.40886784e-01 1.31499991e-01
9.53517735e-01 -4.95561302e-01 -6.05762005e-01 3.05679947e-01
5.41128397e-01 -9.79602695e-01 8.94358695e-01 -1.97004721e-01
6.81101322e-01 -2.80049652e-01 -4.79695946e-01 -1.11398911e+00
-1.31159738e-01 -5.96690714e-01 -3.52497905e-01 1.21248186e+00
3.76972914e-01 -5.90969563e-01 8.00612986e-01 2.24055335e-01
-1.11927100e-01 -5.41061342e-01 -9.24492240e-01 -7.72720993e-01
1.10554956e-01 -8.09237719e-01 3.50392193e-01 1.09912932e+00
-1.78377286e-01 1.95160165e-01 -7.66214669e-01 -7.16302264e-03
3.00778955e-01 -1.38398772e-02 6.96635008e-01 -7.19847202e-01
-6.57823205e-01 -3.89687687e-01 -1.78403839e-01 -1.60859072e+00
1.94688767e-01 -1.01056218e+00 3.31820808e-02 -1.07720721e+00
2.08573446e-01 -7.37650096e-02 -8.48569274e-02 7.45294511e-01
-2.93607920e-01 4.52431023e-01 4.53169495e-01 4.71248984e-01
-2.76896060e-01 6.15096152e-01 1.24040639e+00 -3.12048018e-01
-5.78934364e-02 -3.18121076e-01 -5.65516293e-01 4.56579626e-01
7.12664604e-01 -4.27935064e-01 -3.71681929e-01 -7.27870107e-01
-1.10210128e-01 2.82921076e-01 2.45829746e-01 -8.86045277e-01
1.40460879e-01 2.19616536e-02 5.69397867e-01 -6.25838459e-01
5.09831607e-01 -9.53830481e-01 8.41918886e-02 3.84574264e-01
-1.71654746e-02 -5.60583919e-02 4.02756780e-01 4.92606103e-01
-3.04878294e-01 -9.28385779e-02 6.24607921e-01 -6.74957931e-02
-8.01427543e-01 3.13038647e-01 -6.32560372e-01 -7.50295864e-03
6.22062862e-01 -3.38921130e-01 -3.17189932e-01 -6.25723422e-01
-2.71209240e-01 2.79326349e-01 5.19876599e-01 5.58395207e-01
7.96596229e-01 -1.29948068e+00 -6.85950875e-01 5.12056470e-01
-4.04096097e-02 -1.46968186e-01 4.96257097e-01 8.85180712e-01
-3.10543120e-01 3.07162315e-01 -1.12931341e-01 -8.35879803e-01
-1.33970618e+00 6.66674852e-01 1.88995555e-01 -1.45519361e-01
-9.91573870e-01 6.86593592e-01 6.10095382e-01 -1.95617229e-01
2.78948009e-01 -2.53684908e-01 5.61248958e-02 -8.36478770e-02
6.20368958e-01 5.55552877e-02 8.24669749e-03 -5.52644610e-01
-1.67359814e-01 7.89754927e-01 -1.87265575e-01 -2.87792146e-01
1.16666377e+00 -1.37919515e-01 2.46136799e-01 1.15628839e-01
1.45108747e+00 7.57418647e-02 -1.43283761e+00 -4.89140064e-01
-4.87776458e-01 -5.12447000e-01 1.66353345e-01 -4.45009798e-01
-1.48076785e+00 8.69602799e-01 4.06991363e-01 -1.26705453e-01
1.32554340e+00 1.42126307e-01 9.13759470e-01 3.17735881e-01
1.03282779e-02 -9.01737571e-01 4.48756427e-01 5.65835238e-01
7.98295617e-01 -1.29758692e+00 -1.43519491e-01 -3.77114683e-01
-8.57723713e-01 8.84261429e-01 6.28460288e-01 7.60734007e-02
4.21823174e-01 3.93889248e-01 1.04240596e-01 5.22982217e-02
-8.41470540e-01 -3.62583101e-01 5.10363221e-01 3.40992570e-01
2.31286272e-01 -1.49288461e-01 2.99291253e-01 5.10555863e-01
-1.52884901e-01 -3.25704694e-01 3.70070368e-01 6.94917321e-01
-2.64096230e-01 -1.00108945e+00 -4.01240081e-01 5.54986775e-01
-5.18103063e-01 -4.29736912e-01 -1.56322613e-01 8.54182839e-01
-4.78912331e-03 9.45080400e-01 1.19582959e-01 -8.92680824e-01
2.23131523e-01 -7.84939248e-03 3.76706570e-01 -3.78182173e-01
-4.59884435e-01 5.16534269e-01 -1.61747128e-01 -6.29503608e-01
-3.33094746e-01 -4.15968776e-01 -1.15486813e+00 -5.93233526e-01
-2.85102189e-01 -5.64381257e-02 4.17562097e-01 8.59482110e-01
2.15852290e-01 4.71794039e-01 6.55632854e-01 -8.96639764e-01
-2.05080867e-01 -9.01003361e-01 -4.42119569e-01 4.83136475e-01
4.55491155e-01 -4.90820289e-01 -4.92870092e-01 2.44794428e-01] | [9.842999458312988, 0.8859527111053467] |
aa1fa858-0e9b-495d-96e6-a26213392124 | deep-neural-network-and-data-augmentation | 1903.00389 | null | http://arxiv.org/abs/1903.00389v1 | http://arxiv.org/pdf/1903.00389v1.pdf | Deep Neural Network and Data Augmentation Methodology for off-axis iris segmentation in wearable headsets | A data augmentation methodology is presented and applied to generate a large
dataset of off-axis iris regions and train a low-complexity deep neural
network. Although of low complexity the resulting network achieves a high level
of accuracy in iris region segmentation for challenging off-axis eye-patches.
Interestingly, this network is also shown to achieve high levels of performance
for regular, frontal, segmentation of iris regions, comparing favorably with
state-of-the-art techniques of significantly higher complexity. Due to its
lower complexity, this network is well suited for deployment in embedded
applications such as augmented and mixed reality headsets. | ['Viktor Varkarakis', 'Shabab Bazrafkan', 'Peter Corcoran'] | 2019-03-01 | null | null | null | null | ['iris-segmentation'] | ['medical'] | [ 4.63037133e-01 7.63910711e-01 -2.98643172e-01 -4.31164593e-01
-4.10853833e-01 -3.25565398e-01 2.62108892e-01 -2.43021473e-01
-1.88861310e-01 4.35150653e-01 -1.39794692e-01 -5.21754682e-01
-6.91451356e-02 -2.65110791e-01 -5.14971852e-01 -4.08054888e-01
7.66224414e-02 5.31518519e-01 -1.75891563e-01 -1.71553731e-01
8.95058550e-03 9.38593209e-01 -1.81452489e+00 -2.35089753e-02
7.86357462e-01 9.29498613e-01 -7.69238234e-01 7.99688518e-01
3.86512578e-01 3.74059826e-01 -5.63773632e-01 -2.40320265e-01
7.39404917e-01 -5.16080379e-01 -8.07868958e-01 4.86368179e-01
1.54281116e+00 -4.94906217e-01 -1.17894717e-01 7.95709491e-01
6.87651217e-01 -1.05003178e-01 1.23493597e-01 -5.99305689e-01
-2.21154138e-01 5.16392104e-02 -8.37913752e-01 3.45005095e-01
3.55548561e-01 2.64005125e-01 5.67377508e-01 -3.88759702e-01
7.37954319e-01 6.19227946e-01 8.73384416e-01 6.29932940e-01
-1.25957096e+00 -4.57396090e-01 -1.56190470e-01 -7.27924168e-01
-1.19038749e+00 -5.18153429e-01 2.43909463e-01 -3.61357629e-01
1.10917461e+00 3.88895929e-01 1.15463924e+00 4.74208951e-01
6.19102642e-02 7.45159984e-01 1.56865275e+00 -4.98604417e-01
-1.73956946e-01 3.60878408e-01 1.36181153e-02 8.55514646e-01
4.34024855e-02 6.17488682e-01 -1.20637715e-01 9.17825624e-02
1.20832503e+00 -2.33111292e-01 -1.32534146e-01 -3.42276335e-01
-1.05222905e+00 5.33898354e-01 6.88564003e-01 2.83268541e-01
-3.42889935e-01 -1.56084731e-01 1.94170251e-01 1.08038977e-01
6.04183912e-01 8.00947607e-01 -4.37454730e-01 -1.41854778e-01
-1.23593652e+00 2.15660691e-01 4.06525463e-01 7.55509913e-01
2.85881907e-01 1.39591306e-01 -1.64748147e-01 6.42137051e-01
2.04647720e-01 5.72693683e-02 2.39051521e-01 -6.14344120e-01
-7.91053325e-02 8.81773591e-01 -9.81518775e-02 -5.00865459e-01
-9.08758402e-01 -9.98606920e-01 -7.60134578e-01 7.51514077e-01
5.91936171e-01 -3.11277658e-01 -1.79516292e+00 1.17300010e+00
6.03050113e-01 1.76189870e-01 -1.18750080e-01 1.20502329e+00
1.00590730e+00 3.52800339e-02 -2.84905016e-01 1.04314454e-01
1.34315705e+00 -8.84905934e-01 -4.61074352e-01 -2.49922648e-01
5.03514528e-01 -9.01335776e-01 8.23090613e-01 5.63664079e-01
-1.45990920e+00 -5.12286127e-01 -9.27285969e-01 -3.90822649e-01
-2.27839932e-01 5.73553503e-01 1.04070735e+00 1.08079898e+00
-1.43735850e+00 3.27978104e-01 -4.80008364e-01 -4.59415078e-01
8.26717913e-01 1.17172742e+00 -3.61913413e-01 4.48994130e-01
-3.45574141e-01 5.98610342e-01 1.61937013e-01 2.77266651e-01
-2.92806864e-01 -9.93337810e-01 -9.40015316e-01 -3.43873143e-01
9.07076448e-02 -8.39093804e-01 1.36022425e+00 -1.16685069e+00
-1.86113203e+00 1.54304934e+00 -2.51664780e-02 -7.16185331e-01
4.56011742e-01 -3.39811444e-02 -4.96262431e-01 3.30155104e-01
-4.29856658e-01 1.19707775e+00 9.97834742e-01 -6.77520156e-01
-9.11807597e-01 -6.91060066e-01 2.93849766e-01 2.24861354e-01
1.50240853e-01 3.47936660e-01 -3.90792668e-01 -4.38482672e-01
3.07125181e-01 -1.06480932e+00 -4.81584996e-01 1.39603794e-01
-6.96299136e-01 1.76514357e-01 5.90890408e-01 -6.45504475e-01
8.37226272e-01 -1.99863863e+00 -1.79662168e-01 3.37823957e-01
5.30935943e-01 5.23874521e-01 1.84741542e-01 -5.17603993e-01
-5.28772593e-01 -7.80806877e-03 -8.83027688e-02 -3.94457579e-01
-5.38865387e-01 -1.52876258e-01 1.62624136e-01 7.88490415e-01
1.94122404e-01 9.24046874e-01 -4.06098902e-01 -2.18599916e-01
7.29836285e-01 6.90516353e-01 -5.77772498e-01 -4.14955020e-02
-2.06808150e-01 8.04610729e-01 9.70096141e-02 1.01723778e+00
6.01324141e-01 -3.51524979e-01 -2.31067345e-01 -7.65015632e-02
2.36052014e-02 1.30427510e-01 -1.04398751e+00 1.76155102e+00
-2.82153428e-01 9.66220260e-01 2.29778424e-01 -4.62619483e-01
7.28496015e-01 2.74100810e-01 5.05428076e-01 -6.59050465e-01
6.50337815e-01 1.68102294e-01 3.73310208e-01 -4.93670031e-02
6.70558989e-01 -1.92343727e-01 3.79683077e-01 1.75382748e-01
1.54551581e-01 -4.12247442e-02 -3.35339941e-02 -1.46020219e-01
5.52849948e-01 2.12625384e-01 3.50018620e-01 -1.77949548e-01
5.03335416e-01 4.32887971e-02 1.92946032e-01 3.82207900e-01
-3.17908823e-01 8.93986821e-01 3.28604519e-01 -8.38438928e-01
-1.20747626e+00 -5.07370949e-01 -6.81096077e-01 4.81964588e-01
-2.68489961e-02 -6.53699972e-03 -1.02051449e+00 -5.59248030e-01
-4.03750390e-02 -3.05984300e-02 -8.60335529e-01 4.38158691e-01
-3.49503189e-01 -7.59913683e-01 4.33246374e-01 3.87359738e-01
3.42737705e-01 -8.96902144e-01 -5.48222780e-01 -1.22481667e-01
7.26683676e-01 -1.33540297e+00 -2.40721837e-01 -2.58130431e-01
-1.05559802e+00 -1.33917439e+00 -9.41121221e-01 -7.93956697e-01
1.04611421e+00 -8.96482095e-02 1.26520646e+00 1.33324534e-01
-8.12730610e-01 2.18041003e-01 1.62909374e-01 -6.05096400e-01
-1.31421775e-01 1.28760844e-01 -4.23800258e-04 1.74724564e-01
7.07660615e-01 -2.96193719e-01 -8.47888112e-01 2.85484612e-01
-7.25732982e-01 6.87652156e-02 7.02318013e-01 9.48614955e-01
8.47489476e-01 -2.27642938e-01 -2.68809702e-02 -1.06102848e+00
3.77514184e-01 4.30699401e-02 -1.08379281e+00 -2.50320017e-01
-8.17841411e-01 -6.38885915e-01 2.26462886e-01 -3.33875000e-01
-7.34435022e-01 3.91026527e-01 -1.67284667e-01 -2.01617852e-01
-7.57837653e-01 -1.94457591e-01 5.40555656e-01 -8.42658937e-01
8.79704475e-01 -2.66890258e-01 5.95909178e-01 -3.20701212e-01
3.70604455e-01 6.11949563e-01 6.18606091e-01 1.73227906e-01
5.10832369e-01 6.44907713e-01 2.40236938e-01 -8.46067965e-01
-8.90564322e-01 -5.18853009e-01 -7.79758036e-01 2.49542436e-03
8.59134734e-01 -9.06848371e-01 -5.98320484e-01 5.94277859e-01
-5.43465197e-01 -2.09925681e-01 -6.71367705e-01 4.84890640e-01
-5.86586356e-01 -6.88453577e-03 -5.66028178e-01 -5.79790354e-01
-4.81059968e-01 -1.31666970e+00 1.21930456e+00 1.03010809e+00
-1.38622597e-01 -9.50406909e-01 -3.54215913e-02 6.92324996e-01
1.27948552e-01 6.55111492e-01 5.07072270e-01 -6.47841215e-01
-6.80596948e-01 -5.44829786e-01 -3.40205252e-01 3.92892450e-01
-7.53156915e-02 3.37401778e-01 -1.20243645e+00 -4.67431754e-01
-5.03572166e-01 -2.28843287e-01 2.52828777e-01 1.00480175e+00
1.12433887e+00 -9.49771609e-03 -3.00869107e-01 1.24645126e+00
1.27320635e+00 1.90200895e-01 8.70411456e-01 2.66812295e-01
3.92717272e-01 5.03465772e-01 5.07095098e-01 -4.96199168e-02
-3.51130888e-02 7.30155349e-01 4.33085859e-01 -1.29991937e+00
-2.87450045e-01 1.66231349e-01 -5.83960891e-01 -1.20737784e-01
-1.05024956e-01 1.75839975e-01 -8.96852195e-01 7.78341949e-01
-1.23885918e+00 -4.01239395e-01 -3.19166839e-01 2.20653057e+00
5.41507900e-01 1.42494977e-01 7.68166363e-01 -1.58036221e-02
5.00176251e-01 -1.98444813e-01 -5.97915351e-01 -6.90323830e-01
-1.78941339e-01 7.31325269e-01 5.30161560e-01 3.33532184e-01
-1.24054408e+00 9.82404351e-01 8.06531239e+00 5.25204003e-01
-1.24673712e+00 -1.50425673e-01 1.10005283e+00 -4.73916858e-01
4.04411882e-01 -2.53577530e-01 -1.05592418e+00 1.25695780e-01
1.01427710e+00 4.48790431e-01 3.32699358e-01 6.53581500e-01
-7.43824942e-03 -2.72910267e-01 -7.40558803e-01 1.03818488e+00
1.06311455e-01 -1.53681731e+00 -4.28265929e-01 6.45074904e-01
9.54190314e-01 3.06950033e-01 3.86398107e-01 -7.04545528e-02
-1.31867826e-01 -1.39790893e+00 -1.79088131e-01 3.31517786e-01
1.22785020e+00 -9.51679826e-01 8.65262508e-01 -2.02533036e-01
-4.55881029e-01 -9.63749662e-02 -3.95416878e-02 -3.45631279e-02
-8.85690749e-02 5.32715976e-01 -1.02532232e+00 3.20061445e-01
6.36482656e-01 5.53005099e-01 -9.04232025e-01 1.45604599e+00
2.02462718e-01 3.65872294e-01 -6.82875335e-01 2.95966476e-01
2.77940720e-01 -3.59561980e-01 4.89739984e-01 8.22338521e-01
-8.42931047e-02 4.54574451e-02 -1.30451977e-01 7.08644331e-01
2.21105516e-02 3.25493932e-01 -7.75263309e-01 -8.73001888e-02
-1.94390818e-01 1.46356499e+00 -7.16819406e-01 -3.34966630e-01
-2.72457570e-01 6.45876765e-01 -8.63630548e-02 1.60761744e-01
-5.10405779e-01 -3.66612345e-01 8.21131587e-01 5.37063599e-01
8.27196687e-02 2.43100286e-01 -7.04124093e-01 -7.13994026e-01
1.19955108e-01 -1.18170416e+00 1.48592934e-01 -7.14897513e-01
-6.18761122e-01 1.06889570e+00 -4.10756856e-01 -1.45926118e+00
-4.21068847e-01 -7.92233288e-01 -3.16748500e-01 1.07355762e+00
-1.46438992e+00 -1.56286669e+00 -5.40269971e-01 3.28188628e-01
1.52418852e-01 -7.28607535e-01 8.09601426e-01 1.79648921e-01
-7.69959390e-01 8.65505159e-01 -1.90244809e-01 1.58805728e-01
4.40491229e-01 -1.40988445e+00 6.22770190e-01 9.09684181e-01
1.26922071e-01 7.08868802e-01 4.78509128e-01 -3.45184356e-01
-1.03069699e+00 -7.91578233e-01 2.62771040e-01 -4.86699790e-01
1.92670912e-01 -7.33830333e-02 -4.65269953e-01 7.68321872e-01
3.48299950e-01 4.47926015e-01 8.19522917e-01 4.55620319e-01
1.31175622e-01 1.40964715e-02 -1.59450722e+00 5.03933430e-01
6.33472383e-01 -4.02868629e-01 -3.02008271e-01 6.91102207e-01
3.33511174e-01 -1.43065834e+00 -1.00568640e+00 6.84780836e-01
6.17783368e-01 -1.44470334e+00 8.76004040e-01 -6.54118180e-01
3.25039774e-01 -1.98210075e-01 6.78347409e-01 -8.91512334e-01
-4.34343293e-02 -1.21945250e+00 -2.99162835e-01 7.40990698e-01
5.87446749e-01 -7.40703821e-01 1.53954327e+00 4.37568158e-01
1.61692090e-02 -1.12222373e+00 -9.67714012e-01 -2.52850950e-01
-2.51556277e-01 7.42809698e-02 4.69671965e-01 7.01776862e-01
-4.28407609e-01 1.57169029e-01 -9.82316285e-02 1.88217014e-01
4.54613656e-01 -6.16299408e-03 1.01307726e+00 -1.45056951e+00
4.59363498e-02 -6.01599395e-01 -1.14092815e+00 -1.01251292e+00
-1.99175462e-01 -4.30158138e-01 -4.81754869e-01 -1.14300227e+00
-4.21511412e-01 -5.04295766e-01 -5.23282960e-02 4.28692251e-01
1.28201276e-01 1.26143742e+00 -3.12261730e-01 -1.76415578e-01
6.56387955e-02 -1.41448051e-01 1.40678084e+00 1.05330117e-01
-7.33392060e-01 4.47338969e-01 -8.22824001e-01 6.73579335e-01
7.53101766e-01 6.39187247e-02 -3.75277311e-01 7.73594305e-02
7.74456328e-03 -1.21896446e-01 2.05622867e-01 -1.23242199e+00
4.24554609e-02 3.80498469e-01 6.43724024e-01 -6.79699659e-01
5.42389035e-01 -8.25402319e-01 -9.36721116e-02 1.01932429e-01
-8.18977058e-02 -2.27342710e-01 7.47482300e-01 1.02016903e-01
-3.20001006e-01 1.48025393e-01 1.20902514e+00 -3.74658294e-02
-2.66740888e-01 4.35186923e-01 3.71479988e-03 -1.99239418e-01
1.26399457e+00 -9.72132802e-01 -1.75444648e-01 -1.90472305e-01
-1.01041591e+00 -9.42910090e-02 8.22941661e-01 4.04221028e-01
4.73305881e-01 -8.43350887e-01 -4.83189642e-01 9.66073215e-01
1.03531308e-01 2.86288023e-01 2.77434736e-01 1.39407170e+00
-9.58546698e-01 7.17044294e-01 -3.10660481e-01 -1.13958335e+00
-1.63359594e+00 3.26621443e-01 1.12487769e+00 3.98094058e-02
-1.14427984e+00 9.29032207e-01 -6.27184063e-02 -5.50150931e-01
2.57952362e-01 -5.51539063e-01 -3.55692625e-01 -1.80193171e-01
7.26243138e-01 -2.75591612e-02 4.41560686e-01 -6.18083060e-01
8.87226462e-02 8.50137055e-01 -2.89803982e-01 2.88084388e-01
1.05046034e+00 1.53004497e-01 -2.36564040e-01 -2.93281674e-01
6.84303105e-01 -2.22728285e-03 -1.03879464e+00 -3.38996127e-02
-3.57526064e-01 -6.94785476e-01 6.10139430e-01 -1.24154341e+00
-1.31934786e+00 5.84858477e-01 1.27136111e+00 -6.72577098e-02
1.51431298e+00 -3.45429718e-01 5.36412597e-01 -1.12930067e-01
9.19467583e-02 -8.93895745e-01 -6.14363313e-01 -3.07322573e-02
4.48420107e-01 -1.36277151e+00 2.69224316e-01 -5.11995435e-01
-3.94556880e-01 1.05554569e+00 7.89400458e-01 -2.74884105e-01
6.01127267e-01 3.85246009e-01 5.96737921e-01 -5.54853559e-01
-3.94478589e-02 -4.96012539e-01 1.03984904e+00 8.82535160e-01
5.26829898e-01 -1.42694563e-01 -7.46701956e-02 -3.74087423e-01
-3.88727963e-01 9.06556845e-02 6.90670848e-01 6.35076046e-01
1.44871742e-01 -9.58237827e-01 -2.42522895e-01 8.03417623e-01
-8.20542336e-01 -9.92638916e-02 -6.18049502e-01 1.14081836e+00
2.27696016e-01 4.96973395e-01 3.97547752e-01 -8.52453187e-02
3.14830840e-01 -2.07577318e-01 6.27156854e-01 -6.82726026e-01
-1.28479409e+00 2.57968068e-01 8.03072900e-02 -8.48478198e-01
-5.81959724e-01 -3.53503168e-01 -8.23565245e-01 -4.96675298e-02
-2.85375476e-01 -3.46990049e-01 8.60520184e-01 6.07764423e-01
6.71042264e-01 5.60371161e-01 4.13832992e-01 -7.59515882e-01
6.64637238e-03 -9.80013251e-01 -7.12504387e-01 1.71091989e-01
9.20462668e-01 -3.94053102e-01 -1.83802664e-01 -3.08618784e-01] | [3.743356943130493, -3.6324100494384766] |
fb5a5cb9-6568-4e97-8862-7c2185554511 | modelling-depth-for-nonparametric-foreground | 1609.09240 | null | http://arxiv.org/abs/1609.09240v1 | http://arxiv.org/pdf/1609.09240v1.pdf | Modelling depth for nonparametric foreground segmentation using RGBD devices | The problem of detecting changes in a scene and segmenting the foreground
from background is still challenging, despite previous work. Moreover, new RGBD
capturing devices include depth cues, which could be incorporated to improve
foreground segmentation. In this work, we present a new nonparametric approach
where a unified model mixes the device multiple information cues. In order to
unify all the device channel cues, a new probabilistic depth data model is also
proposed where we show how handle the inaccurate data to improve foreground
segmentation. A new RGBD video dataset is presented in order to introduce a new
standard for comparison purposes of this kind of algorithms. Results show that
the proposed approach can handle several practical situations and obtain good
results in all cases. | ['Antoni Jaume-i-Capó', 'Gabriel Moyà-Alcover', 'Ahmed Elgammal', 'Javier Varona'] | 2016-09-29 | null | null | null | null | ['foreground-segmentation'] | ['computer-vision'] | [ 6.04390025e-01 -9.88291875e-02 -1.91030711e-01 -3.98804933e-01
-6.15553856e-01 -5.08161902e-01 3.58439684e-01 -1.43272594e-01
-3.78043890e-01 7.98116505e-01 -3.73345643e-01 -2.96044517e-02
-6.92257434e-02 -6.25994086e-01 -4.93807077e-01 -1.04989207e+00
3.14769685e-01 3.22840989e-01 1.10151589e+00 4.65614110e-01
1.23471588e-01 7.14956820e-01 -1.80148590e+00 2.61703014e-01
4.90307897e-01 9.72108662e-01 4.81485277e-01 7.85256207e-01
-3.40213627e-01 4.32117075e-01 -6.64884508e-01 -1.43489167e-01
4.86886203e-01 -4.18507457e-01 -4.66548800e-01 6.06310308e-01
3.28025877e-01 -5.78929484e-01 3.52500752e-02 8.85848463e-01
6.10898972e-01 5.23317233e-02 6.76260114e-01 -1.24481654e+00
1.40002638e-01 2.30969325e-01 -7.89326608e-01 4.03372645e-01
5.81671715e-01 -1.60747826e-01 4.02149171e-01 -4.20107692e-01
5.38329959e-01 1.18536103e+00 4.88611400e-01 4.96294349e-01
-1.11287308e+00 -2.96362966e-01 4.47886050e-01 1.07036695e-01
-1.12816727e+00 -2.59751678e-02 1.17010605e+00 -4.66365993e-01
4.28304434e-01 3.82688284e-01 7.01190233e-01 1.10801840e+00
-3.46124955e-02 1.14251602e+00 1.45839584e+00 -4.23852682e-01
3.35563123e-01 1.20189428e-01 2.99371839e-01 3.55937123e-01
5.09322882e-01 -1.43635750e-01 -3.77103537e-01 8.69052298e-03
8.48926306e-01 1.16555512e-01 -2.92888403e-01 -6.68055594e-01
-9.25686836e-01 4.09709811e-01 4.59359437e-02 6.42193854e-01
-1.94054380e-01 1.53008997e-01 -3.70464957e-04 -2.69835263e-01
5.17691314e-01 2.44998503e-02 -3.82914573e-01 -2.50179946e-01
-1.24733543e+00 7.18190446e-02 6.35042012e-01 8.74559939e-01
8.35002363e-01 -1.07535318e-01 -8.50690007e-02 4.33679461e-01
5.82397342e-01 7.58764744e-01 6.66627660e-02 -1.25173187e+00
3.42146188e-01 4.81509447e-01 1.36741444e-01 -9.24969733e-01
-6.14448488e-01 5.50911576e-02 -5.93113840e-01 3.36062223e-01
7.81717062e-01 1.10347643e-01 -1.10524130e+00 9.61342871e-01
5.47807097e-01 5.12058794e-01 -1.43188789e-01 8.84981334e-01
7.18376338e-01 7.24799931e-01 -1.98475882e-01 -4.45239782e-01
9.86283541e-01 -4.03601468e-01 -1.04485321e+00 3.23486067e-02
1.07528135e-01 -7.98937798e-01 7.49223590e-01 8.83884966e-01
-1.08010650e+00 -5.75175405e-01 -9.70683932e-01 -1.04345761e-01
-4.84065086e-01 7.63058513e-02 6.97923243e-01 1.20796752e+00
-9.58061814e-01 5.10442853e-01 -1.17776775e+00 -4.53504115e-01
1.97175682e-01 3.71784270e-01 1.05697855e-01 -1.62477344e-01
-7.74234772e-01 6.29667222e-01 4.85003918e-01 3.56427819e-01
-6.40324533e-01 -2.41872832e-01 -6.95558846e-01 -5.94429553e-01
5.56776702e-01 -4.78210390e-01 1.02427685e+00 -9.30287600e-01
-1.91109097e+00 9.28803563e-01 -2.47876093e-01 -2.13300303e-01
7.42993772e-01 -6.14718720e-02 -7.06381351e-02 3.95451725e-01
-1.59208089e-01 3.34394127e-01 8.67220879e-01 -1.77776992e+00
-7.36364424e-01 -3.66932362e-01 1.82947755e-01 2.18611006e-02
1.13606036e-01 -1.37954399e-01 -9.35889840e-01 -4.53929871e-01
4.80571091e-01 -9.15290475e-01 -1.74062446e-01 -1.74951166e-01
-4.64448929e-01 3.86377349e-02 8.80260646e-01 -4.93581593e-01
8.92976105e-01 -2.03212595e+00 2.01815471e-01 3.36861998e-01
1.40579566e-01 7.12792203e-03 4.31889981e-01 -7.15545565e-02
3.60417128e-01 -4.87621054e-02 -6.20041251e-01 -6.99907422e-01
-1.06732838e-01 6.91198051e-01 2.54349783e-03 7.20118940e-01
-3.11277970e-03 3.33494544e-01 -7.89490044e-01 -5.98742425e-01
8.64696801e-01 6.93665504e-01 -1.62253618e-01 9.65114236e-02
-1.47598565e-01 6.64473712e-01 -5.65361142e-01 1.00765395e+00
1.15852463e+00 1.97739258e-01 -8.19281116e-02 -1.40001997e-01
-1.30271092e-01 -2.91548222e-01 -1.74007261e+00 1.64216280e+00
-6.48462027e-02 6.39747143e-01 1.34844899e-01 -7.45372653e-01
7.77235270e-01 5.33574373e-02 8.02536249e-01 -4.10803348e-01
4.94985998e-01 1.49966121e-01 -3.49908501e-01 -5.26901007e-01
6.28773689e-01 -2.10426494e-01 1.91817209e-02 5.58769666e-02
-5.97607195e-02 -4.43540394e-01 9.11040008e-02 -1.35988683e-01
8.30709934e-01 5.37691534e-01 2.07221359e-01 -1.30232915e-01
5.15521467e-01 -8.80025327e-02 6.90829813e-01 1.00999296e+00
-4.75248605e-01 1.02887416e+00 4.56332326e-01 -2.11680785e-01
-4.80119437e-01 -1.07445180e+00 -1.95423186e-01 5.16134918e-01
7.27381587e-01 1.18094489e-01 -8.18586111e-01 -5.48553348e-01
-1.13616332e-01 4.04395133e-01 -5.52529514e-01 5.04692018e-01
-3.87081593e-01 -1.13945472e+00 3.06690842e-01 5.24535716e-01
5.83929300e-01 -5.72548270e-01 -8.98610353e-01 -1.85962413e-02
-2.96597481e-01 -1.37551141e+00 2.09167704e-01 5.07955313e-01
-1.26149929e+00 -1.04970551e+00 -1.00008941e+00 -1.92754552e-01
4.07662123e-01 5.04665017e-01 8.12298536e-01 -1.25363305e-01
-3.21121961e-01 7.30325460e-01 -6.34817600e-01 -3.19205165e-01
-2.20014751e-01 -8.29214230e-02 -1.81229159e-01 2.37916186e-01
3.61739695e-01 -2.79318035e-01 -3.34696740e-01 3.20526004e-01
-1.16045904e+00 -8.08253139e-02 5.06716549e-01 1.88564807e-01
9.15529251e-01 1.70322984e-01 -2.65353262e-01 -8.38334262e-01
-4.31168592e-03 -3.56602430e-01 -8.94752681e-01 2.09289212e-02
-2.39736736e-01 -1.03613334e-02 5.36460020e-02 -4.53914285e-01
-1.19629788e+00 5.37559807e-01 -7.91675895e-02 -4.98012453e-01
-6.26903832e-01 -8.81686360e-02 -7.74496436e-01 -1.51258588e-01
2.84460664e-01 6.49639145e-02 -3.50955516e-01 -6.71627462e-01
2.46106505e-01 5.69728851e-01 3.99892509e-01 -6.14882350e-01
7.43883014e-01 9.33049142e-01 2.48088390e-01 -1.10660064e+00
-5.04047096e-01 -7.62345612e-01 -1.05107474e+00 -5.74959397e-01
1.16440547e+00 -7.31816113e-01 -4.55683678e-01 6.37614965e-01
-1.31443369e+00 -4.61204588e-01 -2.05897138e-01 6.61908746e-01
-6.83489501e-01 6.08552754e-01 -5.38179398e-01 -1.26424217e+00
6.75823748e-01 -1.39826763e+00 1.33551407e+00 4.84502941e-01
4.09340769e-01 -1.01980555e+00 -6.09570220e-02 2.78490514e-01
8.05191845e-02 6.33811414e-01 1.40517980e-01 -2.80626625e-01
-1.09206235e+00 -8.63693878e-02 -1.68352991e-01 4.70324486e-01
2.48720691e-01 2.72998512e-01 -1.39296031e+00 1.73581392e-01
3.24441075e-01 4.13118094e-01 9.05884981e-01 7.10791588e-01
1.05655468e+00 5.82123041e-01 -4.41060185e-01 6.07200325e-01
1.48512042e+00 5.57624042e-01 8.92610669e-01 5.60844481e-01
8.23502362e-01 5.82576931e-01 7.95651078e-01 4.29067403e-01
2.23339379e-01 7.05653846e-01 5.19663036e-01 -2.32962489e-01
-1.95946574e-01 3.63632590e-01 3.36620718e-01 5.24605036e-01
-5.26138186e-01 -5.00635386e-01 -7.62727797e-01 2.46625990e-01
-1.88969874e+00 -7.89049506e-01 -6.98287606e-01 2.06774426e+00
3.74013126e-01 2.78882474e-01 2.72038221e-01 6.43431783e-01
7.42846310e-01 1.08853385e-01 -3.04598987e-01 -1.20182358e-01
-3.76254588e-01 -4.88805398e-02 9.55710888e-01 7.92910397e-01
-1.30083179e+00 5.29070318e-01 6.81951809e+00 7.21715808e-01
-9.53670919e-01 2.46169224e-01 4.66354638e-01 1.70446992e-01
-9.60641652e-02 -1.39558716e-02 -1.01627231e+00 6.00531518e-01
6.70893550e-01 5.68716705e-01 -8.22517127e-02 6.49279177e-01
2.51987517e-01 -9.71155465e-01 -6.82979584e-01 1.23976910e+00
3.73187304e-01 -7.46097267e-01 -2.60352284e-01 1.76119655e-01
6.60722613e-01 -3.78095180e-01 -2.25907177e-01 -2.54292578e-01
-1.46094695e-01 -3.87378037e-01 5.84057868e-01 9.71156180e-01
7.91000128e-02 -5.90428352e-01 1.04126155e+00 2.65893936e-01
-9.95240092e-01 2.86766618e-01 -2.23796308e-01 -3.13740820e-01
4.75570142e-01 1.06443739e+00 -6.89389586e-01 9.00900126e-01
8.27125609e-01 6.30356312e-01 -7.93655276e-01 1.34011853e+00
-1.82127818e-01 3.56111795e-01 -6.01228178e-01 -1.15547739e-01
4.16141702e-03 -2.58323759e-01 5.38451374e-01 1.23410583e+00
4.94772524e-01 -2.73154639e-02 2.54129708e-01 6.43956006e-01
5.83746314e-01 -4.92990837e-02 -6.08540356e-01 4.03557241e-01
-1.56193271e-01 1.20463407e+00 -1.58607876e+00 -2.14818090e-01
-4.09031212e-01 1.30203748e+00 -5.02464056e-01 4.48284715e-01
-1.02402532e+00 -3.24792080e-02 2.95921534e-01 2.71259964e-01
2.94423789e-01 -6.23652697e-01 -2.76246071e-01 -1.17761922e+00
2.85170134e-02 -2.94414490e-01 2.05222875e-01 -8.03954303e-01
-9.90625024e-01 3.04101616e-01 4.94854003e-01 -1.31907797e+00
-1.64454114e-02 -1.12171733e+00 -6.36217669e-02 5.24098635e-01
-1.73494232e+00 -1.03653061e+00 -5.14425933e-01 7.12609410e-01
4.44807649e-01 2.72063851e-01 2.14751273e-01 4.25977230e-01
-6.16000950e-01 1.30242957e-02 2.39519715e-01 -1.75401211e-01
6.07436717e-01 -1.46382296e+00 -2.98580825e-01 1.25608087e+00
1.83862031e-01 2.38769799e-01 9.13194358e-01 -7.34675884e-01
-1.24431980e+00 -6.53936982e-01 3.31447572e-01 -7.35523760e-01
3.39817971e-01 -5.49572170e-01 -7.15795755e-01 4.94352698e-01
2.81967651e-02 5.95156811e-02 7.44071364e-01 -2.45282233e-01
2.06988335e-01 -1.73196346e-01 -1.20823300e+00 1.17600262e-01
7.28387713e-01 -2.80718774e-01 -5.42774200e-01 -7.54297674e-02
5.23850322e-01 -6.13927245e-01 -6.11928523e-01 6.06783032e-01
2.40361884e-01 -1.44506288e+00 9.21083450e-01 3.85260046e-01
-2.06502229e-01 -8.97410572e-01 -5.39812922e-01 -6.73577011e-01
3.15481275e-01 -5.80446839e-01 -3.70320886e-01 1.44852436e+00
7.59971887e-02 -4.08694118e-01 9.67430592e-01 6.83117509e-01
7.68300192e-03 -1.76324323e-01 -1.15756071e+00 -8.47767889e-01
-3.59384120e-01 -1.01175308e+00 1.43593401e-01 5.27508736e-01
-6.89894438e-01 -1.89064726e-01 -4.12957132e-01 4.71715868e-01
8.12910140e-01 8.71347040e-02 7.84554780e-01 -1.40114391e+00
-2.86878854e-01 -3.24241519e-01 -7.29184031e-01 -1.24813986e+00
-1.11215860e-01 -3.99364293e-01 3.37761700e-01 -1.59146094e+00
1.50435403e-01 -4.78452027e-01 -1.77530110e-01 -3.61896455e-02
-2.09462628e-01 3.99439305e-01 3.04540873e-01 -9.50688347e-02
-7.23997414e-01 2.71303684e-01 9.60417926e-01 -1.78287979e-02
-3.79168689e-01 3.95621359e-01 -1.88587755e-01 1.00027037e+00
5.75273752e-01 -5.79116702e-01 -3.77105296e-01 -3.14063162e-01
-9.08378884e-02 -1.29359022e-01 5.22872090e-01 -1.22349465e+00
8.70465562e-02 -9.23986286e-02 5.93012631e-01 -9.69847679e-01
6.85821533e-01 -1.26790202e+00 1.48066387e-01 2.80975789e-01
3.39921385e-01 -2.49002039e-01 3.19631934e-01 8.39655399e-01
-2.38212183e-01 -4.94395107e-01 6.97634220e-01 -7.07568079e-02
-7.65539110e-01 -7.11887702e-02 -6.34293437e-01 -3.69621396e-01
1.27223110e+00 -7.50085056e-01 1.37216896e-01 -1.93329975e-01
-9.61349607e-01 -1.46077231e-01 5.98863959e-01 1.90066844e-01
4.10251647e-01 -1.08566868e+00 7.73806050e-02 -9.50437784e-03
-1.17910922e-01 1.49980530e-01 3.75827283e-01 1.00256968e+00
-6.57852471e-01 1.28636748e-01 -7.37684518e-02 -1.06179380e+00
-1.20288444e+00 6.01958036e-01 2.73149639e-01 -6.11835569e-02
-2.86235988e-01 6.61537647e-01 1.36614181e-02 -6.17689043e-02
4.10047829e-01 -1.11711490e+00 -5.47156297e-02 4.44257826e-01
5.08405089e-01 5.43365002e-01 -3.66217433e-03 -6.41477883e-01
-4.65848595e-01 9.50315118e-01 6.59063816e-01 -2.50091434e-01
1.00857544e+00 -6.44821942e-01 1.97792854e-02 1.15291166e+00
8.77520144e-01 1.72845721e-01 -1.62065899e+00 2.44761966e-02
1.67609349e-01 -7.07358718e-01 5.27406223e-02 -3.65955621e-01
-1.04724503e+00 1.16336012e+00 1.05930936e+00 4.85309422e-01
1.44450152e+00 -4.13593911e-02 3.20506275e-01 1.24323681e-01
6.72417998e-01 -1.09045112e+00 -6.16589002e-02 2.67685384e-01
2.63342470e-01 -1.48110592e+00 1.21489383e-01 -8.58544052e-01
-4.68664914e-01 1.43484151e+00 3.20905536e-01 -2.61241081e-03
6.58273458e-01 5.61912298e-01 2.40194146e-02 1.73623815e-01
2.53477097e-01 -8.71690869e-01 1.00017451e-01 8.56327236e-01
1.24375775e-01 -2.28296936e-01 -1.39343113e-01 7.03578889e-01
5.00095308e-01 1.19362228e-01 6.37981832e-01 1.12391949e+00
-4.57414746e-01 -1.32381070e+00 -8.74933124e-01 -2.42674109e-02
-5.76533377e-01 4.00952220e-01 -5.03152430e-01 1.11110759e+00
4.93678272e-01 1.09378147e+00 -1.85150176e-01 -3.16770315e-01
3.54862988e-01 -2.63305008e-03 6.95806146e-01 -3.98946851e-01
-1.92784205e-01 5.87061465e-01 -1.88426062e-01 -7.93453813e-01
-1.37534595e+00 -9.69030738e-01 -1.15611351e+00 1.34187683e-01
-5.25675952e-01 -2.37646997e-01 8.20564091e-01 9.37953472e-01
-6.61164001e-02 6.43306553e-01 3.35327774e-01 -1.45607579e+00
5.00334442e-01 -5.80092311e-01 -9.23136473e-01 1.19581878e-01
3.79129648e-01 -1.08565986e+00 -5.97431183e-01 3.22618216e-01] | [8.814130783081055, -1.0674192905426025] |
cd568d8e-96c6-44cd-8a61-6649d81f197d | an-exploration-of-neural-radiance-field-scene | 2210.12268 | null | https://arxiv.org/abs/2210.12268v1 | https://arxiv.org/pdf/2210.12268v1.pdf | An Exploration of Neural Radiance Field Scene Reconstruction: Synthetic, Real-world and Dynamic Scenes | This project presents an exploration into 3D scene reconstruction of synthetic and real-world scenes using Neural Radiance Field (NeRF) approaches. We primarily take advantage of the reduction in training and rendering time of neural graphic primitives multi-resolution hash encoding, to reconstruct static video game scenes and real-world scenes, comparing and observing reconstruction detail and limitations. Additionally, we explore dynamic scene reconstruction using Neural Radiance Fields for Dynamic Scenes(D-NeRF). Finally, we extend the implementation of D-NeRF, originally constrained to handle synthetic scenes to also handle real-world dynamic scenes. | ['Zheng Xin Yong', 'Wasiwasi Mgonzo', 'Tuluhan Akbulut', 'Benedict Quartey'] | 2022-10-21 | null | null | null | null | ['3d-scene-reconstruction'] | ['computer-vision'] | [ 4.36419219e-01 -4.24460739e-01 7.39379764e-01 -4.44361985e-01
-5.56678832e-01 -4.28419709e-01 7.42538095e-01 -2.87207901e-01
-3.40697110e-01 5.58292985e-01 2.43739560e-01 -4.51376677e-01
-1.45184295e-02 -1.37965703e+00 -6.20902181e-01 -2.82818884e-01
-3.19741756e-01 4.20303583e-01 4.45211440e-01 -3.97565365e-01
3.26584041e-01 1.16370010e+00 -1.91618693e+00 5.26476085e-01
-4.25810628e-02 7.53261447e-01 5.34306705e-01 1.29846799e+00
-1.57220334e-01 1.32414126e+00 -7.33761728e-01 -2.49114390e-02
6.21481538e-01 -2.70512789e-01 -6.72885358e-01 -4.13037360e-01
6.70033693e-01 -8.67738724e-01 -7.96948910e-01 7.27033734e-01
7.81741083e-01 7.45474219e-01 2.29744807e-01 -8.60350132e-01
-2.45826945e-01 -4.33937572e-02 -4.56247300e-01 4.34151590e-01
8.90183151e-01 1.89998850e-01 2.41374373e-01 -6.50361896e-01
8.28877866e-01 1.44132602e+00 1.03351164e+00 2.19601378e-01
-1.02416623e+00 -6.18193567e-01 -5.23222983e-01 -1.76105294e-02
-1.55199826e+00 -3.44593257e-01 4.31556523e-01 -1.31045714e-01
1.82933438e+00 7.38413274e-01 8.06722462e-01 7.09018826e-01
5.76907694e-01 1.66536883e-01 1.13776767e+00 -2.50240475e-01
2.17326924e-01 -1.79309770e-01 -1.61935046e-01 7.87325978e-01
-2.45512098e-01 4.59160030e-01 -5.05663991e-01 -2.94717640e-01
1.62232554e+00 -4.07921910e-01 -4.09669489e-01 -1.02974540e-02
-1.20472503e+00 6.96628213e-01 5.19027948e-01 1.60258636e-02
-5.72951376e-01 5.43586731e-01 3.80885035e-01 1.92273512e-01
4.06919122e-01 5.71093976e-01 -1.01982921e-01 -4.96872723e-01
-7.42075920e-01 5.36140859e-01 7.48165190e-01 7.89842188e-01
8.85982037e-01 6.39489949e-01 2.18116179e-01 6.27435088e-01
5.98561252e-03 2.70360023e-01 2.83033133e-01 -1.49548674e+00
4.50597890e-02 -3.02036077e-01 -1.60875078e-02 -1.07079768e+00
-6.66854978e-01 1.68356597e-01 -8.39101136e-01 7.83907413e-01
-4.28581648e-02 2.05847457e-01 -8.08289468e-01 1.30505240e+00
3.64449620e-01 4.86918300e-01 1.79958031e-01 7.90296376e-01
1.07843506e+00 1.45937264e+00 -5.05042747e-02 1.45467699e-01
1.00424969e+00 -6.42085075e-01 -3.79485965e-01 -3.84791419e-02
3.68953347e-01 -7.06875503e-01 1.12259626e+00 5.74648440e-01
-1.30838156e+00 -8.22137356e-01 -1.03535426e+00 -5.55284262e-01
-2.66433030e-01 -5.80368936e-01 1.32941639e+00 6.96857512e-01
-1.61159647e+00 5.70638001e-01 -6.52473867e-01 -1.60984099e-01
9.45334882e-02 4.07418549e-01 -3.82987440e-01 -2.25278184e-01
-1.03838933e+00 7.87626922e-01 6.22657001e-01 -1.74182072e-01
-8.74836326e-01 -7.14454591e-01 -9.57150877e-01 7.54732126e-03
-3.15309823e-01 -8.72217000e-01 1.42150843e+00 -6.23069584e-01
-1.66855645e+00 6.27713919e-01 2.76013196e-01 -2.25167602e-01
1.55848578e-01 3.65981013e-02 -5.12578070e-01 3.02458405e-01
-3.48402530e-01 8.20205033e-01 1.16403028e-01 -1.41404867e+00
-4.05280352e-01 8.97251666e-02 4.68832463e-01 7.81887114e-01
3.72150600e-01 9.62959528e-02 -2.57518113e-01 -4.00651723e-01
1.91121891e-01 -5.01933694e-01 -5.60304642e-01 1.92670926e-01
1.52546048e-01 4.01858836e-01 1.01615000e+00 -6.07087553e-01
5.19647181e-01 -2.12701130e+00 -2.95946419e-01 2.94859082e-01
5.11703677e-02 -2.08429635e-01 -1.47287071e-01 2.17530742e-01
-2.04314828e-01 -3.23319197e-01 -5.85153922e-02 -2.59296000e-01
-1.51033327e-01 4.70505387e-01 -5.39798439e-01 5.41507661e-01
-4.35097069e-01 4.27943408e-01 -6.58571303e-01 -3.18247646e-01
6.52779460e-01 9.14755583e-01 -8.01381648e-01 1.27021268e-01
-2.59915441e-01 5.06970644e-01 -1.55739337e-01 3.51611733e-01
1.05734837e+00 -4.58175596e-03 -4.81430627e-02 1.72876686e-01
-4.33185577e-01 4.83343631e-01 -1.37192631e+00 2.11032939e+00
-6.95022821e-01 1.12295961e+00 6.42460287e-02 -2.42866144e-01
8.35250378e-01 8.56438726e-02 5.46689570e-01 -1.23666251e+00
-2.52753086e-02 -1.99773699e-01 -6.96250081e-01 -3.10041070e-01
1.65480924e+00 -4.37389821e-01 -7.70302117e-02 6.07212782e-01
-1.64601132e-01 -8.71596634e-01 -3.63470227e-01 1.20474257e-01
1.16256869e+00 6.23909175e-01 2.96369195e-01 -1.33664668e-01
-1.80329066e-02 8.79695639e-02 -1.52375037e-02 1.04473257e+00
2.63829350e-01 8.88579845e-01 -4.54136804e-02 -7.87539780e-01
-1.26445496e+00 -1.28985238e+00 -2.28590757e-01 1.11758327e+00
3.76561940e-01 -4.83000040e-01 -3.88481081e-01 2.29193032e-01
-3.70945156e-01 1.11608005e+00 -3.00022840e-01 3.29402089e-01
-1.05934095e+00 -8.24305832e-01 7.69180536e-01 2.78677821e-01
5.16277134e-01 -1.25809932e+00 -1.60441387e+00 1.57200292e-01
-2.93399226e-02 -9.70225573e-01 4.95971501e-01 4.18092400e-01
-6.90716803e-01 -5.93294919e-01 -2.76098132e-01 -4.10768598e-01
-1.36875123e-01 4.99199480e-01 1.52129042e+00 -2.59869490e-02
-6.48094475e-01 5.80577135e-01 -1.91812944e-02 -6.97836950e-02
-4.83796477e-01 -5.77453971e-01 -2.35138699e-01 -8.39982390e-01
-1.67763293e-01 -8.74584615e-01 -2.93750048e-01 1.93173289e-02
-1.26335752e+00 6.00817084e-01 -7.18972981e-02 2.13524371e-01
7.64416456e-01 5.89520514e-01 -6.57273591e-01 -7.59590566e-01
2.95670539e-01 -4.95167136e-01 -8.76163781e-01 -1.40781119e-01
1.49293453e-01 -1.70996040e-01 6.43593311e-01 -2.87388891e-01
-1.60626125e+00 -1.02647975e-01 -6.85729682e-01 -3.34767014e-01
-2.35191405e-01 1.06493898e-01 1.82384372e-01 -2.91809708e-01
1.09252799e+00 1.57534733e-01 -7.93999016e-01 -2.12965816e-01
3.53615373e-01 2.75536805e-01 1.28576791e+00 -6.66668355e-01
5.63503087e-01 5.83912015e-01 9.44738984e-02 -9.84678924e-01
-1.24668486e-01 -2.58597910e-01 -3.27323228e-01 -3.91247660e-01
8.62590492e-01 -1.05636072e+00 -6.65018141e-01 5.14940202e-01
-1.63182151e+00 -9.29116607e-01 -6.97162867e-01 6.33721590e-01
-1.14714217e+00 2.17037853e-02 -9.24012661e-01 -8.62859905e-01
7.20169917e-02 -1.08917820e+00 1.29750359e+00 1.48169547e-01
9.27567855e-02 -8.18890035e-01 7.02005804e-01 -2.81227082e-01
5.14848292e-01 7.11798310e-01 9.54553008e-01 3.70546520e-01
-1.17459464e+00 1.37576789e-01 -5.29656351e-01 -4.28453296e-01
-3.39878738e-01 -1.83456197e-01 -1.35513592e+00 -1.83811605e-01
3.63961905e-01 -4.49621141e-01 4.18772489e-01 4.82095838e-01
1.42172575e+00 -1.20597698e-01 1.21739067e-01 1.37004304e+00
2.03839207e+00 4.89149064e-01 1.46565723e+00 6.07309639e-01
7.05747485e-01 4.16601002e-01 3.42619866e-01 9.03809786e-01
4.18278813e-01 7.25299180e-01 4.55076188e-01 -2.42835790e-01
-3.84258300e-01 -3.64532053e-01 -5.84615162e-03 6.31967962e-01
-2.70124018e-01 -6.13184988e-01 -8.92820239e-01 -1.06763698e-01
-1.24767506e+00 -1.22377253e+00 -3.57049346e-01 1.94985580e+00
4.54627991e-01 3.08959140e-03 -2.18923956e-01 4.44339309e-03
4.80054468e-01 6.49508715e-01 -2.75855005e-01 -8.22908640e-01
-4.12994862e-01 6.40419006e-01 7.50962317e-01 6.39830172e-01
-7.05402613e-01 9.24936116e-01 7.93994045e+00 6.83591068e-01
-1.14746141e+00 1.01975478e-01 4.63789463e-01 -6.92631304e-02
-7.57209957e-01 9.81641188e-03 -3.15330684e-01 -2.22757265e-01
1.48322916e+00 9.36407223e-02 8.72331858e-01 8.49342942e-01
-9.16413739e-02 -5.49713135e-01 -5.87380707e-01 1.27772915e+00
7.81854317e-02 -1.55877614e+00 5.54200783e-02 -5.35656847e-02
4.91167843e-01 3.18313897e-01 3.35414298e-02 3.54336977e-01
7.30869412e-01 -1.20251966e+00 8.53831112e-01 6.78695440e-01
1.26793063e+00 -8.64355922e-01 7.91877508e-02 1.62492573e-01
-1.16729546e+00 2.85303384e-01 -8.42834890e-01 -3.81683975e-01
3.94605011e-01 1.67390242e-01 -8.50283325e-01 5.26610017e-01
1.09873009e+00 3.36658001e-01 -5.36177933e-01 1.10121274e+00
4.46692765e-01 -6.15665726e-02 -6.22878253e-01 1.70925230e-01
2.04392746e-01 -4.50079590e-02 2.61958867e-01 1.39596343e+00
4.60585922e-01 6.01082563e-01 -7.64207393e-02 5.84992468e-01
4.67568845e-01 -2.35461220e-01 -1.20745420e+00 5.11784732e-01
-8.67981091e-02 7.65333831e-01 -8.76286387e-01 -2.78060526e-01
-4.89668250e-02 1.40021276e+00 4.81606796e-02 3.55876356e-01
-1.09035242e+00 -2.88491756e-01 5.52033782e-01 2.14652106e-01
-4.85004559e-02 -7.49390781e-01 -4.00670588e-01 -8.12503159e-01
-5.74052691e-01 -5.54188371e-01 2.07448915e-01 -1.69649196e+00
-5.48955142e-01 9.95012045e-01 4.44732577e-01 -1.02800179e+00
-5.12629330e-01 -4.10344273e-01 -1.89545855e-01 8.80429149e-01
-1.40210247e+00 -7.93489933e-01 -6.63298428e-01 8.32632124e-01
5.47799885e-01 1.04498237e-01 1.26567829e+00 2.62044787e-01
1.41294271e-01 -1.50047727e-02 2.44805858e-01 -4.18641031e-01
4.26712595e-02 -1.06609750e+00 9.86680329e-01 7.95161188e-01
1.33678555e-01 1.75749123e-01 7.94042647e-01 -5.27462602e-01
-1.46667051e+00 -8.27074409e-01 7.39279687e-02 -1.37230009e-01
7.92550519e-02 -5.69730282e-01 -8.57878983e-01 6.67368293e-01
1.93058044e-01 -5.65759279e-02 6.56902969e-01 -4.06466097e-01
-5.42026877e-01 4.37311977e-01 -1.56197166e+00 6.21046305e-01
1.15292203e+00 -8.45919907e-01 9.64056142e-03 3.99374992e-01
8.64932597e-01 -1.08734167e+00 -7.73961723e-01 3.65521044e-01
6.14660501e-01 -1.72959888e+00 1.44623232e+00 1.72938928e-02
3.37473691e-01 -4.26764756e-01 -7.53003776e-01 -8.30407262e-01
-5.42788923e-01 -5.25302052e-01 1.96402177e-01 3.91606808e-01
-3.13261181e-01 -2.97442973e-01 7.12515712e-01 4.64731365e-01
-3.39789450e-01 2.58258153e-02 -1.10673392e+00 -2.95329362e-01
-2.47073904e-01 -1.01917911e+00 8.29595566e-01 9.17866230e-01
-7.63648510e-01 -5.56907021e-02 -5.04257500e-01 1.47983521e-01
2.80639142e-01 9.09361914e-02 9.41646159e-01 -8.43688011e-01
-8.19659710e-01 -1.63216740e-01 -6.11739337e-01 -1.08624268e+00
-1.45901889e-02 -6.55544043e-01 4.22967821e-01 -1.60468638e+00
-1.23511620e-01 -6.58758640e-01 1.71324536e-01 8.07394087e-02
4.16777164e-01 7.36389160e-01 2.36237764e-01 1.37376964e-01
-5.67000210e-01 4.65702266e-01 1.06194973e+00 4.03907388e-01
-1.41587287e-01 -3.36254299e-01 -7.97467977e-02 6.18699074e-01
5.46571016e-01 -3.53969634e-01 -6.01867914e-01 -8.08468640e-01
4.76290137e-01 7.49468803e-01 4.26424593e-01 -1.41011763e+00
1.51456371e-01 -1.84890553e-01 7.27274120e-01 -8.43300581e-01
1.02009630e+00 -8.50485861e-01 9.57564294e-01 2.76246607e-01
-3.55107971e-02 7.56596625e-01 7.29100168e-01 2.31512979e-01
2.96443015e-01 -1.90371737e-01 1.04644179e+00 -6.91343665e-01
-1.19169557e+00 1.40854921e-02 -6.79621816e-01 -4.40399945e-01
6.47037387e-01 -7.01624095e-01 -6.01743460e-01 -6.81211114e-01
-2.50949472e-01 -6.84551418e-01 8.34903538e-01 3.06684077e-02
9.92283165e-01 -1.25627279e+00 -4.97476190e-01 4.31832105e-01
-2.14385360e-01 3.27884704e-01 7.83652842e-01 -9.25560296e-02
-1.78587174e+00 1.74795926e-01 -6.72858894e-01 -5.49115121e-01
-1.12734306e+00 5.53004801e-01 4.73875850e-01 -1.09282389e-01
-1.13627994e+00 8.51073265e-01 5.18572628e-01 -6.62449002e-01
1.04630524e-02 -2.36288458e-01 1.83188006e-01 -6.22854590e-01
6.68158889e-01 3.48165631e-01 -8.69078655e-03 -7.42212474e-01
8.25659037e-02 3.91830623e-01 5.81540465e-01 -5.98613083e-01
1.39162350e+00 6.04257919e-03 -1.55757844e-01 4.31918681e-01
1.31331110e+00 -3.03640645e-02 -1.20468831e+00 1.04558520e-01
-4.73286927e-01 -8.66075158e-01 3.72237533e-01 -6.88011527e-01
-6.72448516e-01 6.27287626e-01 8.94075036e-01 -8.65840986e-02
1.49933791e+00 -3.65915000e-01 7.99481153e-01 6.10735536e-01
8.57859612e-01 -4.90436554e-01 -1.30160600e-01 6.79095268e-01
6.48463070e-01 -3.46863925e-01 1.65818021e-01 -3.52469265e-01
-3.68274987e-01 1.25555396e+00 4.33035165e-01 -6.66163340e-02
4.96834874e-01 1.00453329e+00 -7.10412040e-02 -3.95099759e-01
-4.52338099e-01 2.50430942e-01 -4.02306616e-01 8.59413445e-01
2.02633485e-01 -2.45142039e-02 6.18259549e-01 -4.69374478e-01
-9.14268672e-01 -9.65680182e-02 8.72857690e-01 1.13051581e+00
-3.92949969e-01 -5.58198392e-01 -5.98536193e-01 9.34133306e-02
-2.28132069e-01 -2.91442364e-01 1.45474926e-01 9.28046048e-01
3.24752368e-02 2.61485219e-01 7.39520490e-01 -4.33959931e-01
3.40206146e-01 -4.37006503e-01 8.58837903e-01 -4.25969422e-01
-9.11558628e-01 4.87933457e-02 1.65631503e-01 -8.42451096e-01
-3.94195497e-01 -3.35699886e-01 -1.30802178e+00 -6.90210521e-01
7.08547607e-02 -2.73330986e-01 1.13292217e+00 1.22851215e-01
-1.21800177e-01 6.77664757e-01 5.12754321e-01 -1.33932459e+00
2.54468858e-01 -4.10422534e-01 -7.18420565e-01 -1.01732105e-01
2.15312794e-01 -1.80574089e-01 9.38258842e-02 -3.19294840e-01] | [9.346010208129883, -2.984679698944092] |
8990b0f8-d3a4-4dd1-899c-12a4ab8e661e | proteinbert-a-universal-deep-learning-model | null | null | https://doi.org/10.1093/bioinformatics/btac020 | https://academic.oup.com/bioinformatics/article-pdf/38/8/2102/49009610/btac020.pdf | ProteinBERT: a universal deep-learning model of protein sequence and function | Self-supervised deep language modeling has shown unprecedented success across natural language tasks, and has recently been repurposed to biological sequences. However, existing models and pretraining methods are designed and optimized for text analysis. We introduce ProteinBERT, a deep language model specifically designed for proteins. Our pretraining scheme combines language modeling with a novel task of Gene Ontology (GO) annotation prediction. We introduce novel architectural elements that make the model highly efficient and flexible to long sequences. The architecture of ProteinBERT consists of both local and global representations, allowing end-to-end processing of these types of inputs and outputs. ProteinBERT obtains near state-of-the-art performance, and sometimes exceeds it, on multiple benchmarks covering diverse protein properties (including protein structure, post-translational modifications and biophysical attributes), despite using a far smaller and faster model than competing deep-learning methods. Overall, ProteinBERT provides an efficient framework for rapidly training protein predictors, even with limited labeled data. | ['Michal Linial', 'Nadav Rappoport', 'Yam Peleg', 'Dan Ofer', 'Nadav Brandes'] | 2022-02-10 | null | null | null | bioinformatics-volume-38-issue-8-2022-2 | ['protein-secondary-structure-prediction', 'protein-structure-prediction'] | ['medical', 'miscellaneous'] | [ 1.61591753e-01 -8.53687990e-03 -5.65726817e-01 -7.76847422e-01
-8.71352315e-01 -5.43956876e-01 3.33747685e-01 6.51365459e-01
-5.37955642e-01 9.66792643e-01 1.49836183e-01 -3.36144000e-01
2.76596934e-01 -4.41875488e-01 -9.99115527e-01 -7.56908417e-01
-2.76483834e-01 8.98330629e-01 -8.32514081e-04 -7.88672417e-02
-1.41155452e-01 5.08234918e-01 -1.26598752e+00 7.84750700e-01
4.06104118e-01 8.08834791e-01 3.52483839e-01 7.26918578e-01
-4.72716749e-01 9.74109292e-01 -7.67147169e-02 -2.42083132e-01
-2.21061796e-01 -2.56786883e-01 -1.10135007e+00 -1.95314005e-01
1.65812805e-01 1.69384599e-01 -1.81478232e-01 5.96138239e-01
6.11991823e-01 1.51879350e-02 6.43569291e-01 -7.27398396e-01
-8.88206363e-01 4.66668367e-01 -2.76198149e-01 1.32012337e-01
2.49949738e-01 3.30688328e-01 1.27324200e+00 -1.07774639e+00
8.34620893e-01 1.21363509e+00 9.09910381e-01 5.81739843e-01
-1.66586292e+00 -2.08658874e-01 9.70186368e-02 -2.11398210e-02
-1.32593274e+00 -4.07864749e-01 2.99434997e-02 -5.59332609e-01
1.87161863e+00 -1.77992105e-01 1.48888752e-02 1.08240104e+00
6.26978040e-01 7.91436911e-01 7.30318487e-01 -3.35580647e-01
1.20085433e-01 -1.79376185e-01 3.34485978e-01 1.05396426e+00
-2.20758662e-01 -2.67373860e-01 -6.88460767e-01 -5.48547983e-01
1.15046196e-01 8.71224776e-02 -2.54356745e-03 -4.98302937e-01
-1.23608136e+00 6.61636710e-01 1.67138785e-01 1.65096238e-01
-4.58215714e-01 2.36870900e-01 7.17889309e-01 4.23160851e-01
6.77743375e-01 4.76864994e-01 -1.38267648e+00 -1.07787527e-01
-6.45503581e-01 2.34731555e-01 8.93882334e-01 8.30674171e-01
8.14407587e-01 -2.00171024e-01 -1.39853135e-01 1.08701992e+00
2.59897023e-01 1.69451952e-01 7.64566958e-01 -3.35880578e-01
6.59912229e-02 6.27025008e-01 -3.69903482e-02 -4.96440053e-01
-8.84022772e-01 -3.43682140e-01 -6.43163979e-01 -1.60936825e-02
3.02375466e-01 6.67365715e-02 -1.00251913e+00 1.82984400e+00
5.11993915e-02 -7.91292265e-02 2.39117205e-01 6.40589237e-01
1.00146914e+00 8.10225904e-01 8.84705067e-01 -4.94738780e-02
1.35674071e+00 -1.13604116e+00 -4.42489862e-01 -3.40719938e-01
1.13891220e+00 -5.90592802e-01 1.07515442e+00 3.59611481e-01
-1.00234103e+00 -4.24391240e-01 -8.33348870e-01 -7.67662108e-01
-7.52633989e-01 9.70297977e-02 7.56076276e-01 1.05312608e-01
-9.39555705e-01 7.93501854e-01 -9.53473389e-01 -6.56029701e-01
3.47368300e-01 6.35289669e-01 -6.24408603e-01 -1.80376451e-02
-1.02510393e+00 9.54417765e-01 7.22229779e-01 -4.45703954e-01
-8.38039160e-01 -7.01606631e-01 -7.33163595e-01 2.96014786e-01
1.04931362e-01 -6.55196905e-01 1.35746527e+00 -6.24396145e-01
-1.50066400e+00 1.39505529e+00 -4.53310430e-01 -7.96007097e-01
5.72699830e-02 -3.30706984e-01 -2.75295019e-01 -7.05964118e-02
-1.52633607e-01 9.60478246e-01 1.77323043e-01 -7.41707802e-01
-2.76181072e-01 -1.67300895e-01 -4.08230513e-01 -1.43582653e-02
-2.40069360e-01 3.29869121e-01 -4.01323348e-01 -5.94067156e-01
-1.43010154e-01 -6.82837307e-01 -5.07816553e-01 1.94036111e-01
-3.04261833e-01 -5.27288914e-01 2.74796575e-01 -5.35766244e-01
7.70830572e-01 -1.83883595e+00 1.98721051e-01 -1.28103092e-01
3.62037659e-01 4.13230926e-01 -4.46261793e-01 7.42432594e-01
-4.77132976e-01 6.08521923e-02 -2.73203999e-01 -4.88945782e-01
3.28731760e-02 3.07987601e-01 -1.70855582e-01 4.35862124e-01
4.23727036e-01 1.24222362e+00 -8.21437955e-01 -3.33313555e-01
-1.21410131e-01 5.41821420e-01 -4.89760458e-01 2.97952712e-01
-8.25574100e-01 8.00158232e-02 -3.54298085e-01 7.75444329e-01
3.55642557e-01 -8.91998887e-01 5.90571284e-01 1.42091224e-02
7.37877861e-02 6.23960018e-01 -1.72362745e-01 1.89054465e+00
-1.36363119e-01 3.33975941e-01 -1.14817411e-01 -1.18987250e+00
1.16880548e+00 3.34731907e-01 7.93085992e-01 -3.77144277e-01
1.59898773e-01 1.60107031e-01 -2.28799712e-02 -6.06495678e-01
2.27138158e-02 -5.56336939e-02 2.14472085e-01 4.62636799e-01
5.80230236e-01 3.69790435e-01 2.22832710e-01 2.48577213e-03
1.30437303e+00 6.15089297e-01 6.95817530e-01 -4.74966049e-01
5.99676907e-01 1.89128727e-01 5.80428600e-01 5.30554235e-01
-4.55118455e-02 3.57519776e-01 5.16321659e-01 -9.33043480e-01
-1.22516763e+00 -7.21489906e-01 9.09370463e-03 1.99218750e+00
-3.03790420e-01 -7.20241010e-01 -5.66139698e-01 -7.46185660e-01
1.55896962e-01 4.26578343e-01 -2.92822987e-01 -4.14794385e-01
-3.47068399e-01 -1.14361751e+00 6.64653182e-01 4.77115065e-01
-1.37194945e-02 -1.29695070e+00 -1.19710661e-01 6.79997087e-01
7.49303773e-02 -1.10375571e+00 -2.49965921e-01 9.98701274e-01
-7.88950801e-01 -9.33904827e-01 -4.71494615e-01 -1.23211277e+00
4.07589167e-01 -1.14052564e-01 1.49314117e+00 1.78346392e-02
-3.41935128e-01 -9.18934420e-02 -8.16909075e-02 -5.85347652e-01
-6.33210540e-01 4.13279891e-01 1.67737305e-01 -3.09095562e-01
9.73614275e-01 -4.95374620e-01 -3.74138832e-01 1.21303305e-01
-7.63020158e-01 2.63278902e-01 7.91769683e-01 1.08118141e+00
1.02718496e+00 -5.48113644e-01 7.91328728e-01 -1.07293093e+00
3.69225204e-01 -6.46125376e-01 -3.48158538e-01 4.90897954e-01
-6.00037575e-01 3.92915636e-01 8.99016023e-01 -3.00778449e-01
-7.18721747e-01 4.95039582e-01 -4.27837789e-01 6.21803552e-02
-3.81861180e-01 8.50326598e-01 -1.21468872e-01 -1.12761058e-01
7.97924280e-01 4.20154244e-01 2.04256982e-01 -9.20457184e-01
3.47684830e-01 4.70597982e-01 5.45221686e-01 -7.33395636e-01
8.19740593e-02 3.04726660e-01 1.94515917e-03 -6.29393160e-01
-9.32282507e-01 -5.60086727e-01 -8.75536740e-01 5.68864107e-01
7.26747870e-01 -1.06051099e+00 -9.19412553e-01 4.84227419e-01
-1.02379215e+00 -5.49531937e-01 3.34218852e-02 1.70456201e-01
-7.24164188e-01 3.20782363e-01 -1.05128133e+00 -3.75034690e-01
-6.36111021e-01 -9.97246861e-01 1.21189630e+00 -3.15678954e-01
-3.21579427e-01 -9.95441318e-01 1.28616124e-01 2.41200089e-01
1.80253908e-01 6.68680072e-02 1.33025312e+00 -1.25385559e+00
-2.91192740e-01 -1.43152446e-01 -2.08368406e-01 2.37385750e-01
2.24275328e-02 -2.41013035e-01 -9.41049516e-01 -3.28531384e-01
-5.42631149e-01 -9.46434200e-01 1.24057209e+00 1.72483370e-01
1.42619228e+00 -1.35893270e-01 -5.99387825e-01 7.23113418e-01
1.24487519e+00 -4.54553217e-02 3.33948284e-01 3.15244585e-01
6.32150471e-01 3.95407587e-01 3.48835975e-01 1.57020435e-01
3.31284881e-01 6.23433173e-01 8.94331560e-02 -4.38402355e-01
1.47897378e-01 -2.08386764e-01 3.74197632e-01 6.00596130e-01
3.54426116e-01 -6.22051358e-01 -1.24873602e+00 3.57621431e-01
-2.04341388e+00 -6.39126003e-01 7.38169476e-02 1.79629111e+00
1.21045280e+00 9.45441350e-02 1.32466648e-02 -5.74720740e-01
2.67339885e-01 -1.00386277e-01 -8.98488760e-01 -5.83901405e-01
-4.17831719e-01 4.16102380e-01 5.31730473e-01 2.71074176e-01
-1.32664502e+00 1.25032794e+00 7.72843266e+00 7.72921026e-01
-1.09878039e+00 6.15317486e-02 1.03127778e+00 -7.02073378e-03
6.09917752e-02 -2.45502993e-01 -8.66276920e-01 2.81321019e-01
1.43258417e+00 2.20108014e-02 1.52509347e-01 1.02309453e+00
2.13425264e-01 4.00355697e-01 -1.32756388e+00 7.26050377e-01
-7.57412463e-02 -1.67570615e+00 3.70072350e-02 -1.17734946e-01
4.81925994e-01 8.83893251e-01 -2.78199434e-01 4.16735351e-01
6.36754632e-01 -1.41222560e+00 1.58190668e-01 5.29416144e-01
7.61204898e-01 -6.67100787e-01 8.13072383e-01 4.84710693e-01
-8.82072091e-01 8.59336555e-03 -6.51595831e-01 -2.72979904e-02
-4.89492044e-02 5.93345046e-01 -1.00807953e+00 2.26310551e-01
6.06120646e-01 8.99430811e-01 -4.35566396e-01 6.31040335e-01
1.24372147e-01 7.28613675e-01 -2.25337505e-01 -4.82915267e-02
2.80271202e-01 5.33588044e-03 -6.19415455e-02 1.65086687e+00
-1.98363945e-01 1.61413029e-01 8.27042341e-01 7.63702452e-01
-5.30324817e-01 5.68914294e-01 -4.15082246e-01 -5.36223769e-01
1.52774662e-01 1.09333634e+00 -5.05302072e-01 -5.08128703e-01
-7.15495646e-01 9.19909418e-01 9.66650248e-01 4.98110533e-01
-5.70116520e-01 -2.10915700e-01 1.14930427e+00 -1.21038415e-01
5.05455919e-02 -2.95642763e-01 -2.23167762e-01 -9.87670124e-01
-3.33009392e-01 -9.72144604e-01 4.60991591e-01 -6.59403384e-01
-1.64461184e+00 4.65121150e-01 -4.26754951e-01 -6.01348817e-01
-1.90173849e-01 -1.19840086e+00 -8.82217586e-02 9.09807086e-01
-1.46308494e+00 -1.40333068e+00 2.38796696e-01 1.41348476e-02
5.71327686e-01 -5.35124600e-01 1.47759295e+00 3.09273660e-01
-5.56420326e-01 6.03155851e-01 6.21020079e-01 -5.21907322e-02
9.48130548e-01 -1.20295072e+00 8.15329671e-01 3.02264571e-01
2.83964556e-02 8.71620893e-01 6.29457593e-01 -5.64044833e-01
-1.29096663e+00 -1.38466489e+00 1.34517348e+00 -5.70001543e-01
6.95987940e-01 -6.46477163e-01 -1.28105021e+00 7.83493996e-01
3.89048830e-02 1.50686577e-01 1.19131351e+00 2.81672150e-01
-5.96332431e-01 2.63107806e-01 -9.23570633e-01 3.12979311e-01
1.07354069e+00 -7.21974254e-01 -3.82707089e-01 9.73079562e-01
9.48306859e-01 -3.01913083e-01 -9.53419328e-01 4.22339857e-01
5.14272749e-01 -5.45286059e-01 1.12841940e+00 -1.47414494e+00
4.59082544e-01 -1.02637313e-01 -5.51673323e-02 -9.45944846e-01
-7.94586539e-01 -5.77128172e-01 -2.95293450e-01 8.69352758e-01
7.69369841e-01 -5.16531229e-01 8.78845572e-01 3.43021601e-01
-4.43082660e-01 -1.15084100e+00 -4.74200845e-01 -5.93797147e-01
2.96126932e-01 -1.60516843e-01 4.54502821e-01 9.19058800e-01
3.41517270e-01 6.08582020e-01 -4.16334718e-01 -1.75983801e-01
1.11349985e-01 1.06134035e-01 5.87769270e-01 -1.18659925e+00
-4.35456902e-01 -3.75741959e-01 -4.18265283e-01 -1.35652089e+00
5.48696041e-01 -1.27043796e+00 1.77738190e-01 -1.39459348e+00
5.14406681e-01 1.68569870e-02 -7.48127878e-01 1.12655056e+00
-2.15139359e-01 1.88274950e-01 -4.18716729e-01 2.74920225e-01
-8.04514706e-01 4.14309502e-01 7.46467710e-01 -2.93760747e-01
-6.84137121e-02 -4.11881894e-01 -6.99739158e-01 6.34675622e-01
8.66667926e-01 -3.89151096e-01 -1.15928315e-01 -3.92750710e-01
2.17012361e-01 -4.54970121e-01 7.01410174e-02 -7.64749646e-01
3.49990018e-02 -2.79257476e-01 4.86524254e-01 -6.02059007e-01
3.59809220e-01 -4.07431483e-01 -1.46413654e-01 4.08082694e-01
-9.20265317e-01 4.54321988e-02 3.12735856e-01 6.98808372e-01
-7.51004592e-02 1.23730913e-01 8.89139712e-01 -2.47827694e-01
-7.65578091e-01 5.50590575e-01 -7.12966681e-01 -2.23929852e-01
6.08878970e-01 2.27596477e-01 -2.60620147e-01 -3.66270803e-02
-1.00190902e+00 3.78050178e-01 4.69857931e-01 4.24593776e-01
2.52218246e-01 -8.70773137e-01 -6.52572572e-01 8.54918584e-02
5.74875712e-01 -2.88696676e-01 -9.36779156e-02 5.46989560e-01
-7.86980391e-01 8.31896961e-01 -5.72224632e-02 -6.10394835e-01
-1.35667145e+00 8.34952235e-01 3.59286129e-01 -4.50928271e-01
-6.30310833e-01 1.05648768e+00 4.31780010e-01 -9.24833179e-01
2.41319060e-01 -4.38560128e-01 -9.11938176e-02 -4.68366802e-01
5.04612684e-01 -2.39542827e-01 3.13619405e-01 -4.94287997e-01
-4.06890333e-01 1.98841721e-01 -4.12214875e-01 5.72455108e-01
1.53622937e+00 7.10611269e-02 -3.87224108e-01 5.40641129e-01
1.37319744e+00 -5.42872787e-01 -1.03917527e+00 -5.61588168e-01
5.51676512e-01 2.52108485e-01 -2.64056951e-01 -1.09905815e+00
-4.47251678e-01 8.50779295e-01 5.31057298e-01 -4.52096462e-01
6.61681652e-01 9.16635320e-02 9.79986191e-01 1.10059154e+00
3.10241133e-01 -8.84298742e-01 -1.05833374e-01 9.32528555e-01
3.77095133e-01 -1.44241071e+00 -7.75641715e-03 -3.13440502e-01
-4.45931524e-01 1.08140397e+00 5.39307475e-01 1.04030669e-01
5.64910591e-01 4.12090987e-01 2.58955598e-01 -3.81963670e-01
-1.40483904e+00 -9.38776284e-02 3.36385190e-01 5.09826779e-01
1.32613444e+00 2.57499255e-02 -3.86904031e-01 7.28152335e-01
1.93427190e-01 1.56905696e-01 -7.15673342e-02 9.01207447e-01
-6.89778149e-01 -1.39175928e+00 1.69335261e-01 4.37152922e-01
-7.71227479e-01 -4.30839360e-01 -7.03492761e-01 3.99489433e-01
3.74538521e-03 6.21390641e-01 -1.66146040e-01 -1.41692489e-01
-1.49108479e-02 6.86842144e-01 5.28288111e-02 -9.02960479e-01
-5.91730475e-01 1.28674135e-01 2.66421944e-01 -6.31677210e-01
-2.28743151e-01 -3.26428473e-01 -1.76836658e+00 -1.50440186e-01
1.01773739e-02 2.68617898e-01 5.01451135e-01 1.03801703e+00
9.24025774e-01 5.07863224e-01 -1.21814653e-03 -6.74231827e-01
-4.63524342e-01 -9.73504186e-01 -3.81372660e-01 4.60756809e-01
1.91085353e-01 -4.99423295e-01 3.01161677e-01 4.22737658e-01] | [4.710461139678955, 5.698302745819092] |
e1ba7b01-0300-4cf6-a8c6-49df9eed6e38 | ai-enabled-prediction-of-esports-player | 2012.03491 | null | https://arxiv.org/abs/2012.03491v2 | https://arxiv.org/pdf/2012.03491v2.pdf | AI-enabled Prediction of eSports Player Performance Using the Data from Heterogeneous Sensors | The emerging progress of eSports lacks the tools for ensuring high-quality analytics and training in Pro and amateur eSports teams. We report on an Artificial Intelligence (AI) enabled solution for predicting the eSports player in-game performance using exclusively the data from sensors. For this reason, we collected the physiological, environmental, and the game chair data from Pro and amateur players. The player performance is assessed from the game logs in a multiplayer game for each moment of time using a recurrent neural network. We have investigated that attention mechanism improves the generalization of the network and provides the straightforward feature importance as well. The best model achieves ROC AUC score 0.73. The prediction of the performance of particular player is realized although his data are not utilized in the training set. The proposed solution has a number of promising applications for Pro eSports teams and amateur players, such as a learning tool or a performance monitoring system. | ['Anton Stepanov', 'Andrey Somov', 'Evgeny Burnaev', 'Anton Smerdov'] | 2020-12-07 | null | null | null | null | ['sensor-modeling', 'skills-evaluation', 'skills-assessment', 'fps-games'] | ['computer-vision', 'computer-vision', 'computer-vision', 'playing-games'] | [-2.18913123e-01 3.08542456e-02 2.16937270e-02 2.20356137e-02
-3.89883637e-01 -3.50645453e-01 -2.63950199e-01 3.03354356e-02
-7.31301427e-01 5.01573980e-01 -1.54787973e-01 2.35600516e-01
-5.94696522e-01 -7.07368493e-01 -1.91844940e-01 -6.06121898e-01
-1.48028024e-02 3.00268769e-01 2.37218723e-01 -5.15699446e-01
3.34147215e-01 3.55206490e-01 -1.96466124e+00 3.53371382e-01
3.89160872e-01 1.19826913e+00 2.24395409e-01 8.94566774e-01
5.46614707e-01 1.01265430e+00 -6.67025983e-01 -4.19217855e-01
5.28106987e-01 -3.24751973e-01 -2.65463412e-01 -4.81650412e-01
-1.78243101e-01 1.33810446e-01 -3.63494635e-01 5.84167957e-01
7.88347721e-01 3.64530832e-01 3.86699662e-02 -1.13623798e+00
2.72336379e-02 6.60950243e-01 -2.74978489e-01 5.42465150e-01
2.65555978e-01 3.46650183e-01 8.27534378e-01 -3.60804558e-01
2.26371005e-01 3.87877852e-01 9.21451867e-01 5.02138019e-01
-7.52887547e-01 -8.06693316e-01 -1.85718879e-01 6.06969178e-01
-1.22739708e+00 -2.06400424e-01 1.16712534e+00 -4.74634558e-01
7.70187676e-01 2.57568270e-01 1.34052157e+00 1.04249918e+00
3.14386964e-01 4.56533700e-01 1.20529354e+00 8.90304446e-02
3.80558699e-01 3.77841979e-01 3.22909117e-01 2.71430492e-01
-6.49496680e-03 2.21272156e-01 -1.28232956e+00 1.94445297e-01
7.76535213e-01 2.22460856e-03 2.52920836e-01 2.94287056e-01
-4.75415200e-01 5.54459870e-01 1.85695320e-01 2.80352861e-01
-9.76149380e-01 2.84088671e-01 6.84191227e-01 3.94571364e-01
1.22190714e-01 8.87510598e-01 -4.02470797e-01 -1.09440291e+00
-8.86774957e-01 3.71012956e-01 6.15064740e-01 1.74856484e-01
6.35003224e-02 4.29146707e-01 8.07309747e-02 6.74537241e-01
-6.07006326e-02 2.55336612e-01 9.14442599e-01 -1.11387670e+00
1.91598728e-01 8.95730376e-01 -1.45525977e-01 -1.27855241e+00
-8.07906926e-01 -8.48967731e-01 -1.59988388e-01 2.08283722e-01
1.65753752e-01 -1.69726044e-01 -1.52070031e-01 1.55614781e+00
5.17914519e-02 1.93890557e-01 -1.04712777e-01 1.09736454e+00
8.78962815e-01 2.95190275e-01 1.47926062e-01 -2.84276575e-01
1.34010351e+00 -5.40515602e-01 -7.40868986e-01 -4.33185458e-01
2.28318706e-01 -7.26798847e-02 9.13294733e-01 1.03642833e+00
-1.32480335e+00 -9.50344384e-01 -9.12666142e-01 5.05941987e-01
-1.65352196e-01 1.06710270e-01 8.64052534e-01 9.00345504e-01
-4.97962952e-01 9.75687027e-01 -8.41662943e-01 -1.99412882e-01
-6.91463798e-02 8.09782565e-01 -4.01667893e-01 8.42105508e-01
-1.18888378e+00 8.24744940e-01 4.80336607e-01 2.76960701e-01
-6.81134045e-01 -6.49576426e-01 -5.11303484e-01 -2.89444928e-03
2.58983374e-01 -1.82726175e-01 9.68629956e-01 -1.00409365e+00
-1.90345478e+00 9.59694445e-01 5.37965953e-01 -7.49468982e-01
3.89708757e-01 -3.63743991e-01 -4.07642573e-01 2.38846205e-02
1.84803014e-03 1.91141609e-02 3.09855223e-01 -3.43569189e-01
-5.54578841e-01 -8.24380577e-01 -2.24146485e-01 4.99975115e-01
-3.56545687e-01 3.00742425e-02 -9.79874879e-02 -2.91157216e-01
3.71662229e-01 -9.96517599e-01 5.99462092e-02 -7.06656873e-01
1.84505954e-02 -1.47225216e-01 1.87609047e-01 -7.29090631e-01
1.55788672e+00 -2.15681481e+00 4.02595382e-03 3.78815025e-01
2.28076950e-01 5.13827860e-01 3.66346449e-01 4.52706903e-01
-1.62755087e-01 -2.35788748e-01 5.85538924e-01 1.59753442e-01
-2.09847659e-01 -3.41219306e-02 1.80844754e-01 2.10444838e-01
-3.62937212e-01 9.10637319e-01 -3.61262351e-01 -1.47996828e-01
1.94866255e-01 -3.34937051e-02 -4.88384575e-01 3.91992867e-01
2.77021676e-01 4.56273973e-01 -4.21341509e-01 3.57470810e-01
-4.03798185e-02 4.36721474e-01 2.08990201e-02 -4.11160737e-02
-2.65377760e-01 1.32335901e-01 -1.29188490e+00 1.58352518e+00
-1.97999448e-01 3.95802319e-01 -8.30909014e-02 -1.10728681e+00
1.46504688e+00 3.81936997e-01 9.96168733e-01 -1.16851258e+00
5.86884975e-01 1.65685520e-01 4.17943329e-01 -9.51701462e-01
6.59142435e-01 -2.36644462e-01 -3.05885464e-01 3.21995556e-01
8.66487026e-02 3.58412653e-01 1.45202652e-01 -3.52735013e-01
1.17136788e+00 3.47661316e-01 1.41117424e-01 -3.47922817e-02
3.09284568e-01 9.94776860e-02 7.87174284e-01 6.59498572e-01
-4.83761638e-01 1.68992698e-01 5.08314013e-01 -4.65931743e-01
-7.81609535e-01 -7.83655763e-01 2.88134456e-01 1.47242403e+00
-3.70876119e-02 -4.13189769e-01 -6.60436749e-01 3.47614497e-01
-1.64664879e-01 3.04905087e-01 -7.59099722e-01 -6.97400033e-01
-4.38266039e-01 -5.38813472e-01 8.15840781e-01 8.34750354e-01
5.23787796e-01 -1.35868883e+00 -9.77727175e-01 6.38858140e-01
-1.48741081e-01 -9.58468854e-01 2.21512407e-01 4.66711342e-01
-8.77653241e-01 -8.53100717e-01 9.50930081e-03 -3.94897133e-01
-5.71516275e-01 -2.22647309e-01 7.01745749e-01 -4.54180598e-01
-1.78944454e-01 3.41978610e-01 -6.27778322e-02 -8.49396408e-01
5.31050973e-02 2.50575542e-01 5.47926426e-01 9.11244825e-02
6.16323292e-01 -1.12252390e+00 -5.81443608e-01 3.53731245e-01
-3.74720506e-02 -9.26562771e-03 4.13251251e-01 4.67953324e-01
6.04934394e-01 -9.36276689e-02 1.04557395e+00 -6.19055390e-01
1.06121933e+00 -5.63176155e-01 -1.75561771e-01 -2.86935806e-01
-7.39935279e-01 -4.53643769e-01 5.98857582e-01 -6.09777510e-01
-5.77575207e-01 -1.15800396e-01 -1.61354244e-01 -4.65259075e-01
-7.23609105e-02 5.14521480e-01 9.26506072e-02 2.27491394e-01
9.93448734e-01 1.11220598e-01 8.39958042e-02 -5.20818830e-01
-4.09896761e-01 6.78966582e-01 1.03886843e+00 -5.71180224e-01
1.77100688e-01 -1.25982864e-02 -1.83758866e-02 -5.63812912e-01
-4.95814025e-01 -7.33303905e-01 -3.98150474e-01 -1.10834277e+00
8.03088248e-01 -9.87733066e-01 -1.68715179e+00 3.42651963e-01
-5.57310879e-01 8.08598660e-03 -6.59445941e-01 9.62960005e-01
-7.84053922e-01 -4.67071652e-01 -6.13513887e-01 -1.38849020e+00
-6.85561180e-01 -7.01457322e-01 4.17389601e-01 6.94964111e-01
-5.08433402e-01 -4.68923628e-01 4.55281675e-01 9.67435122e-01
2.12974727e-01 2.64506072e-01 1.94351897e-01 -9.16512728e-01
1.70713156e-01 -8.46578240e-01 5.22989333e-01 2.24154785e-01
-3.30134392e-01 -2.63014168e-01 -1.21021426e+00 1.44694805e-01
2.69974977e-01 -3.75491887e-01 2.98499078e-01 4.20277119e-01
9.96179938e-01 2.61686921e-01 2.78625339e-01 3.88764709e-01
1.16767645e+00 5.05764425e-01 6.17803752e-01 4.86319214e-01
4.64478433e-01 6.64540648e-01 5.62887788e-01 5.97804666e-01
8.54052827e-02 6.28182352e-01 2.95416653e-01 3.10063362e-01
3.29503983e-01 -4.71760869e-01 5.78096747e-01 9.35979605e-01
-9.33498204e-01 1.79883376e-01 -6.90623403e-01 4.38127190e-01
-2.03816628e+00 -1.18908656e+00 -2.98245549e-01 2.24767041e+00
2.95775235e-01 3.60533625e-01 7.08429277e-01 7.37091839e-01
5.05718708e-01 -3.08622271e-01 -8.18034947e-01 -1.00779653e+00
1.77106977e-01 5.16128480e-01 3.96825373e-01 -2.45391786e-01
-5.91503024e-01 7.90094793e-01 6.16905594e+00 7.99226642e-01
-1.09217739e+00 1.32357001e-01 2.29320407e-01 -6.50840044e-01
4.07363206e-01 -5.09332120e-01 -5.40459335e-01 4.64074790e-01
1.51277566e+00 -1.94790229e-01 4.07371432e-01 1.00653005e+00
7.09934533e-01 -2.05468833e-01 -6.64038897e-01 1.27885747e+00
-1.49944704e-02 -8.45611334e-01 -8.79770756e-01 1.98371500e-01
2.01821536e-01 -9.90134999e-02 1.55064687e-01 6.54439390e-01
-1.83347821e-01 -1.06564176e+00 5.75359762e-01 9.43826675e-01
3.59517753e-01 -9.43534791e-01 9.46006536e-01 5.26793957e-01
-9.28489625e-01 -5.68834841e-01 -1.80093851e-02 -9.26069379e-01
-3.06997329e-01 -5.84155880e-03 -4.80925053e-01 2.39206016e-01
7.80467451e-01 4.92938101e-01 -4.82115746e-01 7.90345013e-01
1.47270590e-01 9.51957941e-01 -4.64392096e-01 -2.26639360e-01
1.26134798e-01 -2.45089501e-01 5.54812551e-01 5.12132406e-01
1.08995751e-01 2.89070874e-01 4.91586588e-02 8.33110392e-01
4.52549189e-01 4.68654603e-01 -6.82235062e-01 -1.00144289e-01
1.99453890e-01 1.24807549e+00 -5.62001586e-01 1.02665894e-01
-7.49667361e-02 4.08609182e-01 2.07096517e-01 -1.57548025e-01
-7.74536788e-01 -1.39974520e-01 6.59143865e-01 4.78920937e-01
-2.50080258e-01 9.58924461e-03 -7.45656312e-01 -7.17924237e-01
1.65515319e-01 -7.80339181e-01 4.01417047e-01 -8.41095269e-01
-9.33404326e-01 4.01666701e-01 -2.32729524e-01 -1.31284916e+00
-4.73353744e-01 -6.68421388e-01 -7.95827925e-01 5.80824852e-01
-5.15488982e-01 -6.54167652e-01 -3.91332239e-01 5.64470053e-01
4.05120850e-01 -4.92367923e-01 6.93310678e-01 3.89099687e-01
-1.00682032e+00 4.29701149e-01 -3.26648355e-01 4.24587950e-02
3.52812111e-01 -8.58454466e-01 -3.26490462e-01 4.34748441e-01
-2.18333378e-02 2.24842891e-01 8.17968726e-01 -4.73577589e-01
-1.39967310e+00 -4.00368124e-01 2.90473998e-01 -3.68515670e-01
6.19094789e-01 -1.25335738e-01 -5.32746971e-01 4.24936593e-01
-3.10949326e-01 -4.36213136e-01 1.21880090e+00 4.18278784e-01
3.66122514e-01 -5.79849303e-01 -9.04335916e-01 4.03280407e-01
8.95816207e-01 -5.58237910e-01 -6.08220220e-01 -1.66091561e-01
9.49954391e-02 -4.46103901e-01 -1.15618205e+00 3.12569261e-01
1.06024039e+00 -1.23182356e+00 6.98578298e-01 -8.53046119e-01
4.58273143e-01 -1.16410730e-02 2.79623549e-02 -1.02541411e+00
-1.92743003e-01 -4.86626267e-01 -1.01679891e-01 9.31473851e-01
2.11969793e-01 -3.61991554e-01 1.11063468e+00 1.14861655e+00
-1.15616307e-01 -9.24280524e-01 -7.62422085e-01 -4.20882463e-01
-2.76090711e-01 -1.09797561e+00 1.47066250e-01 4.42530781e-01
5.04664719e-01 2.98291087e-01 -6.98232472e-01 -2.60099381e-01
2.71202832e-01 -8.50150958e-02 8.47052574e-01 -1.52810514e+00
-6.06978595e-01 -2.43658826e-01 -1.25991857e+00 -2.63223410e-01
-2.22995803e-01 -6.51923060e-01 -2.89183289e-01 -1.02146971e+00
4.49918443e-04 -4.93084267e-02 -7.04789340e-01 2.11428016e-01
1.17181942e-01 5.06701350e-01 1.93421289e-01 -1.20087145e-02
-6.07436121e-01 2.34907612e-01 9.90247488e-01 4.24804121e-01
-6.48629129e-01 5.23177922e-01 -8.84377301e-01 7.54001796e-01
1.10314596e+00 -6.02258563e-01 -5.25130570e-01 1.65150017e-01
6.07353389e-01 5.17059624e-01 2.78952599e-01 -1.58940232e+00
4.50609475e-01 -1.21593110e-01 3.70881259e-01 -1.73904985e-01
7.80914009e-01 -6.39586270e-01 4.96860623e-01 4.53670830e-01
-3.60954732e-01 4.88233566e-02 2.12245300e-01 3.47336769e-01
-2.59715945e-01 -1.97449833e-01 4.89067763e-01 -3.66593242e-01
-7.74432957e-01 9.13524032e-02 -5.92313409e-01 -6.80321008e-02
1.17829680e+00 -8.64859998e-01 1.46738589e-01 -3.74405861e-01
-1.29202688e+00 1.65497229e-01 -2.24816978e-01 3.48658830e-01
2.77656645e-01 -1.03877366e+00 -4.95454967e-01 2.03395441e-01
-5.91235422e-03 -6.41201675e-01 7.15267777e-01 1.00612521e+00
-3.78896922e-01 1.94152653e-01 -9.63208020e-01 -3.82581204e-01
-1.25862873e+00 1.18959568e-01 6.90084815e-01 -1.99557900e-01
-5.24391472e-01 5.93623221e-01 -4.18512523e-01 -1.93769962e-01
1.13078184e-01 2.55948365e-01 -7.32326329e-01 2.47865319e-01
5.14444828e-01 7.03414142e-01 3.55112106e-01 -4.94249195e-01
-2.14146256e-01 4.42279935e-01 1.73294395e-01 -2.44140387e-01
1.58755946e+00 1.23051524e-01 3.18348020e-01 1.14197958e+00
4.60564345e-01 -8.79682153e-02 -8.85167003e-01 1.24381609e-01
-1.84799835e-01 -1.83393002e-01 7.66340867e-02 -6.94819331e-01
-1.22760594e+00 8.69084418e-01 1.07078838e+00 2.11720914e-01
1.05902648e+00 -3.81541878e-01 5.60957551e-01 3.73266608e-01
5.87779164e-01 -1.74089420e+00 7.41000921e-02 2.43295446e-01
4.83740181e-01 -9.54579711e-01 -1.21627204e-01 1.84599340e-01
-1.17909420e+00 9.89206970e-01 1.02512801e+00 -6.07301950e-01
5.60253859e-01 3.06296468e-01 1.84488431e-01 -4.77062792e-01
-9.45186257e-01 -3.65572155e-01 1.33037448e-01 5.85755825e-01
4.41791326e-01 3.40019464e-01 -6.10434115e-01 1.66322637e+00
-9.02339756e-01 3.04441899e-01 5.00267684e-01 5.12122571e-01
-4.44178402e-01 -4.76756364e-01 -2.86345363e-01 6.60702586e-01
-7.50211358e-01 3.76648873e-01 -3.50134194e-01 6.81358576e-01
3.70065033e-01 1.03919959e+00 9.46918502e-02 -1.08973336e+00
7.80571163e-01 4.09808755e-01 2.31697246e-01 -4.35511738e-01
-1.40343785e+00 -1.61582232e-01 1.01689428e-01 -8.63617718e-01
-2.19476372e-01 -5.47116160e-01 -1.16765237e+00 -5.07720172e-01
-1.29966378e-01 3.99831027e-01 9.00941730e-01 7.07280278e-01
2.72697002e-01 8.15976441e-01 6.31523550e-01 -6.04504526e-01
4.99244072e-02 -1.20458090e+00 -1.13937902e+00 3.08541268e-01
-4.64944065e-01 -7.15586543e-01 -3.16427387e-02 -3.11519980e-01] | [6.826514720916748, 0.39057713747024536] |
5bd49e9b-a20e-44b1-bf42-5cd3b252d1e4 | keep-the-conversation-going-fixing-162-out-of | 2304.00385 | null | https://arxiv.org/abs/2304.00385v1 | https://arxiv.org/pdf/2304.00385v1.pdf | Keep the Conversation Going: Fixing 162 out of 337 bugs for $0.42 each using ChatGPT | Automated Program Repair (APR) aims to automatically generate patches for buggy programs. Recent APR work has been focused on leveraging modern Large Language Models (LLMs) to directly generate patches for APR. Such LLM-based APR tools work by first constructing an input prompt built using the original buggy code and then queries the LLM to generate patches. While the LLM-based APR tools are able to achieve state-of-the-art results, it still follows the classic Generate and Validate repair paradigm of first generating lots of patches and then validating each one afterwards. This not only leads to many repeated patches that are incorrect but also miss the crucial information in test failures as well as in plausible patches. To address these limitations, we propose ChatRepair, the first fully automated conversation-driven APR approach that interleaves patch generation with instant feedback to perform APR in a conversational style. ChatRepair first feeds the LLM with relevant test failure information to start with, and then learns from both failures and successes of earlier patching attempts of the same bug for more powerful APR. For earlier patches that failed to pass all tests, we combine the incorrect patches with their corresponding relevant test failure information to construct a new prompt for the LLM to generate the next patch. In this way, we can avoid making the same mistakes. For earlier patches that passed all the tests, we further ask the LLM to generate alternative variations of the original plausible patches. In this way, we can further build on and learn from earlier successes to generate more plausible patches to increase the chance of having correct patches. While our approach is general, we implement ChatRepair using state-of-the-art dialogue-based LLM -- ChatGPT. By calculating the cost of accessing ChatGPT, we can fix 162 out of 337 bugs for \$0.42 each! | ['Lingming Zhang', 'Chunqiu Steven Xia'] | 2023-04-01 | null | null | null | null | ['program-repair', 'program-repair'] | ['computer-code', 'reasoning'] | [-6.85829297e-02 6.93114460e-01 9.24221873e-02 -4.48006168e-02
-1.40688407e+00 -6.87457740e-01 4.58383895e-02 4.26908791e-01
4.44169939e-01 8.21732223e-01 -1.30143389e-01 -7.42362440e-01
1.53759256e-01 -9.08698797e-01 -9.91028130e-01 -1.72552288e-01
-1.83445543e-01 4.94158864e-01 6.13629699e-01 -4.09872681e-01
5.02888381e-01 -3.07810605e-01 -1.59545958e+00 7.45906711e-01
1.13504088e+00 -2.52080057e-02 3.26299995e-01 1.00867307e+00
-1.96370736e-01 9.45567191e-01 -1.19911921e+00 -4.36096340e-01
-7.75496736e-02 -8.87642622e-01 -1.09291351e+00 -9.31160972e-02
1.24693312e-01 -2.26407498e-01 4.44336295e-01 9.04796720e-01
2.73205429e-01 -3.51539761e-01 -1.30913243e-01 -1.36901820e+00
-4.20971751e-01 1.07701719e+00 -5.07060587e-01 -1.03854299e-01
1.04189980e+00 5.36254346e-01 9.95056689e-01 -5.27230501e-01
5.72958469e-01 1.06553090e+00 8.65995467e-01 7.61280715e-01
-1.39446020e+00 -1.41648918e-01 -3.49257961e-02 -1.38531089e-01
-1.11683941e+00 -3.55007835e-02 5.34235120e-01 -4.18124616e-01
1.43199384e+00 4.95026320e-01 6.26382232e-01 9.77950394e-01
5.17670631e-01 3.06988657e-01 9.03990388e-01 -6.90827191e-01
2.61003762e-01 1.71169117e-01 3.49286079e-01 9.93994713e-01
1.17434829e-01 -7.59839490e-02 -1.96229145e-01 -9.23682749e-01
2.95924425e-01 -2.79039830e-01 -2.43556216e-01 3.72241408e-01
-9.93229747e-01 7.53307164e-01 8.31275210e-02 4.42783237e-01
-2.22705677e-01 1.53864324e-01 2.01865494e-01 6.36125922e-01
2.34572157e-01 9.82981324e-01 -6.49890006e-01 -4.72070724e-01
-8.08739781e-01 5.23182571e-01 1.34463131e+00 9.40583467e-01
1.22919869e+00 -2.63165534e-01 -2.52027482e-01 6.63055122e-01
2.26287201e-01 2.77136751e-02 3.00535470e-01 -7.82289743e-01
5.80626249e-01 1.17126083e+00 3.42424847e-02 -8.35105002e-01
-1.31494313e-01 -1.45361081e-01 -7.76065141e-02 3.06468308e-01
1.67972386e-01 -2.95168310e-01 -5.33315659e-01 1.50537097e+00
3.65473777e-01 5.53350560e-02 -8.65494385e-02 5.62677979e-01
3.35577995e-01 7.75965989e-01 -4.73867148e-01 -8.79504755e-02
1.11507607e+00 -1.06773710e+00 -6.72606975e-02 -3.64207208e-01
1.16712534e+00 -8.51541996e-01 1.40873933e+00 5.70240557e-01
-9.97240186e-01 -2.81296045e-01 -1.05070341e+00 4.76985693e-01
2.24623978e-01 -4.30229455e-02 4.53810006e-01 5.61799824e-01
-1.43601155e+00 7.16342509e-01 -7.59248257e-01 -2.16194957e-01
-7.10287318e-02 1.43829674e-01 -1.35746717e-01 -3.48338068e-01
-7.09124267e-01 5.62233746e-01 -1.29930049e-01 -1.01187050e-01
-1.05719638e+00 -7.93430686e-01 -8.84341776e-01 -6.46596076e-03
7.07991004e-01 -5.27565777e-01 1.63490260e+00 -8.97154808e-01
-1.29395199e+00 1.70517370e-01 -2.14297131e-01 -2.08165213e-01
1.74137980e-01 -2.79807989e-02 4.95534716e-03 -2.45631486e-01
4.25732553e-01 4.74729747e-01 6.93162799e-01 -1.11011422e+00
-5.14166832e-01 1.92644909e-01 5.18389702e-01 -4.15256530e-01
2.05735013e-01 2.22855866e-01 -1.66325778e-01 -1.69367060e-01
-3.17748010e-01 -9.08390105e-01 -3.43643427e-01 -6.22036874e-01
-5.92537224e-01 -2.94610798e-01 4.77796286e-01 -8.23200345e-01
1.50994372e+00 -2.14147878e+00 3.06325853e-01 1.06351823e-01
3.20271730e-01 5.42593822e-02 -7.37049758e-01 8.60750318e-01
-4.17450350e-03 5.77004373e-01 -3.51590365e-01 -3.11204314e-01
1.13425508e-01 2.10744351e-01 -3.51680636e-01 -2.98459958e-02
7.68407643e-01 7.46989310e-01 -9.73792970e-01 -1.50377333e-01
-2.95887351e-01 -3.81857716e-03 -1.36815345e+00 4.95306402e-01
-1.04568446e+00 2.86613375e-01 -2.22024128e-01 6.71867013e-01
4.23591524e-01 -7.68356249e-02 1.11951299e-01 6.23534679e-01
-1.81292608e-01 6.08788013e-01 -9.76967752e-01 1.31201291e+00
-5.82405984e-01 3.59341562e-01 2.29756199e-02 -6.74121022e-01
9.77726340e-01 4.29388195e-01 -9.03838426e-02 -2.86282241e-01
-4.79264617e-01 4.72390175e-01 1.37378991e-01 -7.97926426e-01
4.18080658e-01 2.54155964e-01 -4.44113165e-01 8.46277356e-01
6.20395243e-02 -4.66397166e-01 2.38143921e-01 2.91334301e-01
1.95188212e+00 2.58281291e-01 8.10803846e-02 1.90456405e-01
4.43998277e-01 3.01343501e-01 7.30981827e-01 1.00790668e+00
3.22561949e-01 6.71542048e-01 1.37162507e+00 -3.06001216e-01
-1.01369226e+00 -6.08053327e-01 4.90484327e-01 7.69481719e-01
-2.71995515e-01 -1.17447782e+00 -1.01580441e+00 -1.22083890e+00
-2.99791664e-01 8.69827509e-01 -6.34512544e-01 -1.50806561e-01
-6.74950600e-01 -3.21314067e-01 6.83685124e-01 -4.60423827e-02
1.53156906e-01 -1.44812751e+00 -5.33489287e-01 4.94219899e-01
-3.65464747e-01 -9.26103294e-02 -4.23245192e-01 2.51923259e-02
-3.91163468e-01 -1.47535467e+00 -9.31421816e-02 -6.96566105e-01
8.94684792e-01 -2.40635592e-02 1.24866700e+00 1.07702684e+00
-4.95615453e-01 2.06273451e-01 -8.23085606e-01 6.97455779e-02
-1.27017403e+00 1.39024794e-01 -6.47202134e-01 -3.09858799e-01
-5.06382883e-02 -6.05003715e-01 6.27425592e-03 4.73952442e-01
-8.69310260e-01 -1.63706169e-01 5.77858508e-01 1.08253372e+00
2.34241307e-01 1.06168725e-01 6.80714488e-01 -9.38311636e-01
8.16022396e-01 -7.59347200e-01 -7.85098135e-01 3.30656916e-01
-3.20974857e-01 1.38240173e-01 7.50543296e-01 -7.07677662e-01
-7.07359314e-01 -3.78220171e-01 -5.44511795e-01 5.52508086e-02
-2.23739058e-01 1.02999401e+00 1.47141665e-01 -1.83391329e-02
1.05689621e+00 1.16035730e-01 -1.59113869e-01 -3.43345165e-01
-3.38174030e-02 4.97150570e-01 1.45326599e-01 -9.26620483e-01
5.89311123e-01 -5.62030733e-01 -6.68108702e-01 -4.98076916e-01
-3.39154661e-01 -7.73333088e-02 1.54046342e-01 -1.72569171e-01
4.16947782e-01 -2.49100298e-01 -5.53466737e-01 2.86429465e-01
-1.54401886e+00 -7.56384134e-01 -2.86950767e-01 -2.03332275e-01
-3.59571964e-01 4.02857959e-01 -6.70014977e-01 -7.70269752e-01
-1.77831993e-01 -1.54392695e+00 1.25195396e+00 6.94630966e-02
-7.19279349e-01 -5.29545903e-01 6.38664186e-01 1.55453086e-01
6.31622672e-01 5.41245304e-02 1.42394662e+00 -6.95377648e-01
-8.17487478e-01 -4.68394220e-01 3.14454287e-01 2.02242836e-01
8.53562029e-04 3.10923249e-01 -4.23634350e-01 -2.31191337e-01
1.79397181e-01 -2.22982183e-01 8.09757635e-02 -2.11983323e-01
4.66300964e-01 -7.89279044e-01 -3.13814491e-01 7.16990558e-03
1.30863488e+00 5.65404855e-02 9.12605107e-01 3.13108593e-01
3.11067939e-01 5.96868157e-01 5.91369569e-01 4.26513553e-01
4.02766287e-01 5.04024804e-01 4.56221700e-01 2.36914337e-01
8.14920813e-02 -3.40534806e-01 8.25355291e-01 6.94994152e-01
3.81496549e-01 -6.73617795e-02 -1.14231730e+00 8.49340796e-01
-1.91964412e+00 -8.64875078e-01 -4.43875492e-01 2.13085818e+00
1.21967840e+00 1.32864192e-01 1.12986624e-01 1.18887261e-01
5.68021655e-01 -3.05668235e-01 -3.53040360e-02 -7.42561102e-01
4.71013635e-01 4.89066839e-01 -4.09728765e-01 8.13454568e-01
-4.53954160e-01 8.57281983e-01 5.32344341e+00 5.34959078e-01
-1.02261174e+00 1.83426082e-01 2.91575462e-01 1.65386260e-01
-7.79097140e-01 8.64162385e-01 -7.94629395e-01 3.89653385e-01
1.07763350e+00 -1.75740167e-01 7.83522487e-01 9.60389078e-01
8.37295037e-03 -6.02017522e-01 -1.30962729e+00 1.03800423e-01
1.99282914e-01 -1.33852279e+00 -2.91302502e-01 -1.25502169e-01
7.78099775e-01 -7.52713978e-02 -4.54578936e-01 7.93652475e-01
3.57370406e-01 -7.76410580e-01 6.51361823e-01 4.42363709e-01
7.70452917e-02 -6.99795306e-01 9.92455602e-01 7.93367863e-01
-8.07306528e-01 -1.47561640e-01 -3.33240271e-01 -3.61572176e-01
-4.50212620e-02 8.11553180e-01 -1.33710444e+00 6.17951155e-01
6.50621295e-01 2.46722117e-01 -9.26568270e-01 1.38941276e+00
-5.96957624e-01 8.96407783e-01 -1.73899323e-01 -2.25837693e-01
-1.24541663e-01 7.33373538e-02 7.59565592e-01 1.01241553e+00
5.80397308e-01 -3.22039485e-01 4.53421831e-01 1.46637571e+00
2.96414495e-01 -7.34595582e-02 -5.85128725e-01 -1.75590010e-03
5.41287005e-01 1.18162036e+00 -3.57816666e-01 -3.53345543e-01
-4.27155107e-01 6.89443350e-01 5.44310868e-01 1.48863196e-01
-8.32283318e-01 -6.39680922e-01 3.62384021e-01 2.24233106e-01
3.39626342e-01 -4.13804129e-02 -4.27676216e-02 -1.01317120e+00
5.99023700e-01 -1.60014713e+00 8.30355063e-02 -7.71369874e-01
-1.14382923e+00 9.40441430e-01 -3.20227183e-02 -9.25265968e-01
-5.02167821e-01 2.40822837e-01 -1.27128553e+00 8.78514886e-01
-9.77164090e-01 -7.94237614e-01 -2.59767398e-02 1.37370870e-01
6.63641512e-01 2.58434236e-01 9.97883856e-01 1.97454393e-02
-5.37654757e-01 6.39743268e-01 -8.48085403e-01 -3.11245918e-01
6.74190640e-01 -1.23406768e+00 6.74114525e-01 1.07645464e+00
-1.39291301e-01 9.24392760e-01 6.84735239e-01 -1.06729043e+00
-1.49104249e+00 -1.06753874e+00 1.23754001e+00 -5.95704079e-01
8.07398856e-01 -4.86664206e-01 -1.23667073e+00 5.39893210e-01
3.73723894e-01 -5.98822653e-01 2.01121286e-01 2.17701569e-02
-3.95627886e-01 1.28288314e-01 -9.73461747e-01 7.46827483e-01
3.95943701e-01 -4.05367404e-01 -7.27135241e-01 5.31769574e-01
9.48302567e-01 -2.58822441e-01 -5.44202805e-01 1.39627963e-01
-1.21730983e-01 -1.14233088e+00 1.47684112e-01 -7.34868422e-02
7.19164968e-01 -6.46728873e-01 4.07915451e-02 -1.48271167e+00
1.17128201e-01 -1.20140433e+00 2.87483424e-01 1.53788340e+00
1.00212252e+00 -9.68450367e-01 4.62811410e-01 4.95041728e-01
-5.78820884e-01 -8.43567669e-01 -8.16477537e-01 -4.92211759e-01
-1.83062822e-01 -5.23040652e-01 7.55638361e-01 6.99686229e-01
7.59469509e-01 1.55321211e-01 -2.10714608e-01 5.11973262e-01
1.34629369e-01 1.51177540e-01 1.20441830e+00 -8.92072022e-01
-1.28051436e+00 -1.37768939e-01 -5.44889942e-02 -5.50139666e-01
7.19948411e-02 -7.05091059e-01 6.79836810e-01 -1.33221662e+00
3.21260422e-01 -6.01399958e-01 5.68452120e-01 1.11867857e+00
-4.52525973e-01 -2.01876294e-02 -6.30345419e-02 1.71798840e-03
-4.71044600e-01 -1.47927120e-01 7.39934027e-01 -1.19787559e-01
-6.53933048e-01 7.25095123e-02 -7.84696102e-01 4.42710966e-01
6.34534240e-01 -9.06906843e-01 -1.59987062e-01 -2.69208223e-01
7.83600509e-01 7.42019951e-01 5.87936103e-01 -9.70959961e-01
1.50466472e-01 -3.92995290e-02 -4.25319225e-01 -2.39761248e-01
-3.38297874e-01 -2.11644903e-01 4.75643039e-01 5.04935801e-01
-2.00572610e-01 3.85599315e-01 3.67890984e-01 2.77593821e-01
-3.10924262e-01 -7.13195324e-01 1.67637199e-01 -1.84750810e-01
-3.31102759e-02 -1.99227482e-01 -8.95440817e-01 -6.96593001e-02
1.01574242e+00 3.12546804e-03 -7.92765915e-01 -2.29369313e-01
-5.85796297e-01 2.22366691e-01 9.33890224e-01 3.90653402e-01
4.74926144e-01 -6.78253353e-01 -6.78420365e-01 5.75504661e-01
1.69536933e-01 -2.10219678e-02 1.37496321e-02 1.05109143e+00
-4.74148244e-01 -8.12050626e-02 2.36610487e-01 -7.15564430e-01
-1.16874731e+00 5.11117339e-01 1.93633914e-01 -4.29195940e-01
-5.24989963e-01 6.98547065e-01 -1.64247528e-01 -7.24465966e-01
-1.13347583e-01 -5.87871253e-01 2.68690497e-01 -4.87436831e-01
5.23499966e-01 -1.23569898e-01 4.49803352e-01 2.55200058e-01
-3.10673088e-01 2.97589272e-01 -2.52207126e-02 -2.21668765e-01
1.49069917e+00 5.36487937e-01 -8.18228304e-01 2.19330683e-01
7.66054630e-01 5.97439229e-01 -9.33612227e-01 3.57426375e-01
1.41445175e-01 -4.50188935e-01 -7.03226984e-01 -9.90916193e-01
-5.44475436e-01 4.43064064e-01 -3.58790547e-01 5.72791517e-01
8.66902292e-01 3.44288647e-01 4.59578544e-01 2.99253434e-01
9.61990118e-01 -4.06141192e-01 4.26396072e-01 6.91100597e-01
9.82010186e-01 -6.68396413e-01 -5.00255227e-01 -4.37818557e-01
-3.62037987e-01 1.06229281e+00 8.77337575e-01 -2.81446874e-01
-1.42062873e-01 6.09827399e-01 -2.16914132e-01 -1.95072934e-01
-1.30089343e+00 3.69959205e-01 -2.37963662e-01 6.52171791e-01
3.59866560e-01 -1.48304641e-01 -4.67040211e-01 8.88790905e-01
-2.92040050e-01 -8.79274085e-02 1.07094669e+00 1.40184009e+00
-4.83694702e-01 -1.80472207e+00 -6.42714739e-01 4.41537708e-01
-1.75828189e-01 -2.84857363e-01 -7.81770825e-01 6.41981721e-01
5.15321344e-02 1.22223485e+00 -3.38269740e-01 -6.66705191e-01
3.58252585e-01 2.33587727e-01 4.71017212e-01 -1.27410674e+00
-1.15753019e+00 1.88021678e-02 6.58952892e-01 -6.69846475e-01
2.58780479e-01 -5.63989639e-01 -1.14064288e+00 -3.10223460e-01
-6.19797170e-01 5.78366518e-01 3.56677741e-01 9.71013665e-01
7.57493854e-01 5.93433440e-01 6.89781249e-01 -8.51763129e-01
-5.82502663e-01 -9.39924300e-01 1.20193191e-01 -3.37334648e-02
3.71232122e-01 -2.48286083e-01 -5.55255592e-01 -1.29094690e-01] | [7.640715599060059, 7.714723110198975] |
d8b4f71a-abc7-4859-99e9-743c5eac87d9 | identifying-training-stop-point-with-noisy | 2012.13435 | null | https://arxiv.org/abs/2012.13435v2 | https://arxiv.org/pdf/2012.13435v2.pdf | Identifying Training Stop Point with Noisy Labeled Data | Training deep neural networks (DNNs) with noisy labels is a challenging problem due to over-parameterization. DNNs tend to essentially fit on clean samples at a higher rate in the initial stages, and later fit on the noisy samples at a relatively lower rate. Thus, with a noisy dataset, the test accuracy increases initially and drops in the later stages. To find an early stopping point at the maximum obtainable test accuracy (MOTA), recent studies assume either that i) a clean validation set is available or ii) the noise ratio is known, or, both. However, often a clean validation set is unavailable, and the noise estimation can be inaccurate. To overcome these issues, we provide a novel training solution, free of these conditions. We analyze the rate of change of the training accuracy for different noise ratios under different conditions to identify a training stop region. We further develop a heuristic algorithm based on a small-learning assumption to find a training stop point (TSP) at or close to MOTA. To the best of our knowledge, our method is the first to rely solely on the \textit{training behavior}, while utilizing the entire training set, to automatically find a TSP. We validated the robustness of our algorithm (AutoTSP) through several experiments on CIFAR-10, CIFAR-100, and a real-world noisy dataset for different noise ratios, noise types, and architectures. | ['Ganesh Sankaranarayanan', 'Babak Namazi', 'Venkat Devarajan', 'Sree Ram Kamabattula'] | 2020-12-24 | null | null | null | null | ['noise-estimation'] | ['medical'] | [ 9.08181295e-02 -2.66445458e-01 3.73362191e-02 -4.84586090e-01
-8.18354905e-01 -5.09427369e-01 1.61881104e-01 2.31094025e-02
-6.40393615e-01 9.14191127e-01 -6.93447709e-01 -3.62098187e-01
-3.11563700e-01 -7.22096741e-01 -7.62828946e-01 -9.73165631e-01
2.42353946e-01 5.23355722e-01 2.95392096e-01 2.14844167e-01
-6.04060218e-02 3.16540122e-01 -1.40800142e+00 7.14734942e-02
1.03314447e+00 1.19580364e+00 2.56227523e-01 2.66500622e-01
-2.04216957e-01 5.50409615e-01 -8.12093973e-01 -1.98425606e-01
3.70521933e-01 -4.28279251e-01 -5.49986124e-01 6.89990744e-02
2.83376187e-01 -1.18128143e-01 -1.52699575e-01 1.50310874e+00
5.05747974e-01 2.58541673e-01 4.47705746e-01 -1.10835469e+00
-2.04820946e-01 7.96412587e-01 -3.96630883e-01 1.88194677e-01
-3.90485883e-01 1.62927434e-01 7.54487097e-01 -8.14560115e-01
3.33839059e-01 6.81616724e-01 6.39618456e-01 6.59931302e-01
-1.27883458e+00 -8.93360853e-01 3.42385739e-01 -1.24503560e-01
-1.61542368e+00 -4.37509954e-01 7.13701189e-01 -2.00057328e-01
3.76756161e-01 1.34248221e-02 4.72774625e-01 1.36238241e+00
-1.81705624e-01 6.25333965e-01 1.13784575e+00 -3.60863090e-01
7.08719492e-01 2.90229440e-01 5.82423091e-01 4.72855985e-01
5.58052778e-01 3.50170247e-02 -7.93179423e-02 -6.45119920e-02
5.53870559e-01 -1.40165776e-01 -3.90338212e-01 -7.23792687e-02
-5.86353302e-01 6.42556369e-01 2.99476266e-01 6.00988269e-01
-1.84156805e-01 -1.05628870e-01 2.28674620e-01 5.27101219e-01
2.81470656e-01 5.12098730e-01 -6.45785332e-01 -2.24632442e-01
-1.03940916e+00 -5.93402535e-02 9.57400560e-01 6.76236868e-01
7.76236296e-01 2.23809376e-01 -2.34384462e-02 1.08333349e+00
-5.39886504e-02 3.13159794e-01 6.14136577e-01 -6.60514057e-01
7.01261759e-01 4.02872831e-01 1.00687310e-01 -7.01038003e-01
-4.56683904e-01 -1.10257614e+00 -9.84180331e-01 2.08769649e-01
9.50990558e-01 -5.15524745e-01 -1.14187896e+00 1.99406815e+00
1.48548469e-01 2.45662555e-01 1.92868635e-01 9.52225804e-01
6.16345525e-01 4.64658946e-01 -1.87448874e-01 -3.39333147e-01
8.97471189e-01 -5.35687208e-01 -6.07813060e-01 -3.98437947e-01
8.46354783e-01 -3.27749372e-01 1.07084846e+00 8.56647432e-01
-5.89513242e-01 -6.05040967e-01 -1.17949593e+00 5.52128553e-01
-2.76336789e-01 4.02329564e-01 2.19949961e-01 9.25228834e-01
-7.07632959e-01 8.62100363e-01 -9.62400973e-01 -9.06628892e-02
4.63757455e-01 4.16000485e-01 -1.75388634e-01 -2.98209339e-01
-1.10331297e+00 4.96012360e-01 4.80271935e-01 4.95521277e-01
-7.26496518e-01 -3.42314780e-01 -4.60211158e-01 1.11711159e-01
5.63409269e-01 -2.44495124e-01 1.07561159e+00 -1.39699745e+00
-1.42489350e+00 3.24379861e-01 1.13496222e-01 -5.24133205e-01
6.89046204e-01 -1.55364454e-01 -3.23041379e-01 -2.06104994e-01
-2.04384983e-01 4.16156828e-01 6.09089136e-01 -1.38106394e+00
-3.57781351e-01 -3.61521840e-01 -6.96367696e-02 -7.26003759e-03
-3.92944694e-01 -2.34684259e-01 -4.68154699e-01 -4.92456228e-01
3.92515898e-01 -9.68999147e-01 -2.27858007e-01 -4.46064413e-01
-5.58178365e-01 -1.12354107e-01 5.87934375e-01 -1.08651847e-01
1.18206358e+00 -2.31451535e+00 -2.40989894e-01 7.00913966e-01
2.47449577e-01 5.88320315e-01 -2.33772308e-01 -2.09810197e-01
-1.31085366e-01 3.60942483e-01 -2.67331600e-01 -2.87142485e-01
-3.19132537e-01 4.20601279e-01 -9.80417058e-02 3.80503237e-01
2.25641564e-01 3.43875349e-01 -9.71992075e-01 -3.36486787e-01
-5.43281436e-02 2.31108963e-01 -3.71734291e-01 -1.37966216e-01
6.59826677e-03 2.30197877e-01 -6.24228895e-01 4.24370885e-01
7.67215312e-01 -2.56327689e-01 2.62671620e-01 -1.33733585e-01
-5.52922077e-02 1.19075244e-02 -1.57687783e+00 9.87296522e-01
-3.43841046e-01 6.27220869e-01 -3.88757773e-02 -1.25702941e+00
1.19465208e+00 3.38972151e-01 4.74820346e-01 -5.63353777e-01
4.92042124e-01 5.79510033e-01 3.55682045e-01 -3.49051535e-01
3.14630233e-02 -8.33344236e-02 2.38109857e-01 1.91815406e-01
1.07160375e-01 3.20107818e-01 2.98689604e-01 -2.94526637e-01
1.29929304e+00 -1.94844306e-01 -1.57338172e-01 -9.78820622e-02
1.76324651e-01 -3.19447368e-01 7.94443727e-01 1.12821817e+00
-3.46521676e-01 8.07714462e-01 7.78042912e-01 -3.61225933e-01
-7.98803627e-01 -8.19577634e-01 -4.91981626e-01 4.72065091e-01
1.04343267e-02 -1.08055741e-01 -9.16929483e-01 -7.94337571e-01
-3.58818859e-01 7.21766770e-01 -5.23119867e-01 -2.29081020e-01
-4.31966662e-01 -1.10027385e+00 5.37924230e-01 4.14572269e-01
7.09691048e-01 -8.85074317e-01 -2.27791041e-01 2.65699595e-01
-5.57340123e-02 -1.23213470e+00 -4.92436439e-02 7.86538780e-01
-8.95195425e-01 -1.05565250e+00 -5.75698435e-01 -6.99522018e-01
9.33557928e-01 4.26816870e-04 8.01841736e-01 3.27884763e-01
2.94531167e-01 -2.73912489e-01 -4.22406346e-01 -3.09625357e-01
-4.05307084e-01 2.04625979e-01 2.49421909e-01 1.06405526e-01
2.40571469e-01 -4.45954442e-01 -2.19200507e-01 6.19582832e-01
-8.78469229e-01 -2.57026732e-01 5.10181010e-01 9.77172434e-01
7.86135912e-01 6.85202241e-01 7.14097440e-01 -8.55987668e-01
4.45864320e-01 -5.39501905e-01 -8.38404119e-01 2.20607981e-01
-4.63612765e-01 1.89895332e-01 1.15168476e+00 -9.18665171e-01
-6.97269738e-01 1.95452839e-01 -2.97078192e-01 -7.83892632e-01
-3.57046306e-01 5.69032669e-01 -4.21215206e-01 -1.74702443e-02
9.29910243e-01 -2.64702737e-03 -2.81693101e-01 -5.19169986e-01
-1.98576316e-01 6.26130044e-01 2.62683272e-01 -6.01625144e-01
8.31675410e-01 1.87512830e-01 -1.75767943e-01 -6.61157370e-01
-1.04310334e+00 -2.70186484e-01 -5.29099226e-01 -2.17505246e-01
2.97021985e-01 -5.57574391e-01 -4.27514881e-01 6.52740479e-01
-8.06176901e-01 -5.73650479e-01 -2.78273582e-01 6.79197371e-01
-2.94853784e-02 2.79667050e-01 -3.89419496e-01 -9.04323995e-01
-1.57147393e-01 -1.23472226e+00 4.52755928e-01 2.64459074e-01
9.76540372e-02 -7.45844483e-01 -2.51865000e-01 -8.31465796e-03
1.36322945e-01 8.05589482e-02 6.03546917e-01 -1.06764972e+00
-3.61435205e-01 -2.43903473e-01 -1.40563086e-01 7.82916367e-01
2.28621796e-01 1.63960084e-01 -1.01188195e+00 -2.77963787e-01
2.26762354e-01 -3.44651639e-01 1.01438272e+00 5.84583402e-01
1.40124953e+00 -2.69943982e-01 -2.28913650e-01 6.36651695e-01
1.48134339e+00 4.12663549e-01 5.26016831e-01 4.01355594e-01
4.59628433e-01 2.31284305e-01 4.03482020e-01 3.17259043e-01
-2.55797029e-01 5.04380941e-01 3.46861631e-01 -4.71504815e-02
4.17034179e-02 4.09797952e-02 1.87051296e-01 6.49657905e-01
1.00425839e-01 -6.63354814e-01 -9.94101524e-01 4.06668335e-01
-1.64774299e+00 -5.98003089e-01 -8.19607228e-02 2.40086150e+00
9.05410051e-01 6.41140282e-01 -1.27945036e-01 5.94187796e-01
6.87681913e-01 -2.91894019e-01 -7.11840332e-01 -1.29759815e-02
-1.54944375e-01 8.59476030e-02 6.15622282e-01 2.81628966e-01
-1.12826860e+00 7.51474500e-01 5.54892921e+00 1.04506826e+00
-1.36500108e+00 -6.49229586e-02 9.80788946e-01 -1.87893629e-01
-3.36414650e-02 -2.08204433e-01 -8.79798710e-01 5.92971027e-01
8.95726860e-01 2.89314777e-01 3.57705057e-01 9.50852513e-01
1.74405918e-01 -9.27273184e-02 -1.05869722e+00 9.50320125e-01
-3.16894084e-01 -8.27255785e-01 -2.61944145e-01 -7.17480332e-02
7.33967662e-01 2.50017017e-01 2.07116846e-02 4.81956899e-01
2.70620406e-01 -7.84318566e-01 7.51967669e-01 3.67556304e-01
5.96329391e-01 -7.86689878e-01 1.12641692e+00 5.81340134e-01
-7.81745195e-01 -2.86375880e-01 -5.61655164e-01 2.77865082e-01
-3.06429327e-01 1.14614928e+00 -7.48772860e-01 2.97284245e-01
6.62595510e-01 2.82256186e-01 -4.35088694e-01 1.37522531e+00
-1.45745963e-01 1.02882123e+00 -7.46999323e-01 -2.39250600e-01
2.38509208e-01 -1.87180638e-01 5.07122338e-01 8.95400167e-01
5.46674013e-01 1.34882852e-01 1.01055689e-01 9.08146858e-01
-2.59735078e-01 5.95411509e-02 -2.93533444e-01 1.20630994e-01
5.16053915e-01 1.06786203e+00 -1.07072127e+00 -1.68196499e-01
-1.75671756e-01 4.80248779e-01 3.71789515e-01 7.09270716e-01
-8.85564685e-01 -3.88654977e-01 3.92368615e-01 2.10669786e-02
2.20995814e-01 -2.39595443e-01 -5.46850443e-01 -1.12925971e+00
1.67884305e-01 -7.68120110e-01 2.94098437e-01 -4.62360233e-01
-1.24854517e+00 9.28329885e-01 -1.48239106e-01 -1.33532596e+00
-5.04066199e-02 -6.17417812e-01 -4.23050463e-01 6.13056719e-01
-1.37811971e+00 -2.65097618e-01 -2.83675373e-01 3.42500895e-01
4.91278023e-01 -1.16383366e-01 4.28082079e-01 4.54724312e-01
-1.09692764e+00 9.03790355e-01 4.90191728e-01 3.73025477e-01
4.99749064e-01 -9.96509314e-01 -4.34336625e-02 9.62729156e-01
2.34938145e-01 4.64823782e-01 7.09065557e-01 -4.61276472e-01
-8.43657374e-01 -1.13325989e+00 4.60713029e-01 -1.09397076e-01
5.16319513e-01 -4.17290628e-01 -1.25836146e+00 3.77133101e-01
-4.49598312e-01 1.12125106e-01 3.23435366e-01 2.32821226e-01
-1.08893193e-01 -5.48049212e-01 -1.11710525e+00 5.11217833e-01
7.81760514e-01 -5.01344129e-02 -1.72436669e-01 2.34694272e-01
4.08324718e-01 -4.93225098e-01 -5.85035563e-01 5.62178910e-01
3.45490783e-01 -8.37887883e-01 5.06805658e-01 -2.38316610e-01
2.44598184e-02 -2.32968166e-01 -3.63894366e-02 -1.48021126e+00
-1.28592908e-01 -4.06197667e-01 7.11929649e-02 1.30374575e+00
6.28744841e-01 -7.15481520e-01 8.57047260e-01 6.00002170e-01
-1.53632984e-01 -8.70393395e-01 -1.18376970e+00 -1.03557670e+00
-4.98529747e-02 -7.71161735e-01 4.65896279e-01 8.84475350e-01
-5.49089253e-01 5.80912232e-02 -7.20054954e-02 3.36508393e-01
4.21826422e-01 -2.42141247e-01 4.29817230e-01 -1.43768322e+00
-1.61668330e-01 -4.86787379e-01 -1.80248335e-01 -9.34391737e-01
1.29063681e-01 -5.56433320e-01 3.46747816e-01 -1.17015028e+00
-1.39117330e-01 -9.84377980e-01 -5.43914974e-01 7.62566984e-01
-6.61348253e-02 8.72325376e-02 -5.00238277e-02 2.41124898e-01
-5.00324249e-01 4.42477018e-01 1.01797509e+00 4.97918911e-02
-4.64263380e-01 4.73818660e-01 -5.75332761e-01 7.54176259e-01
1.01560438e+00 -7.42959678e-01 -5.23140252e-01 -3.97952408e-01
2.45811284e-01 -2.28730232e-01 2.00839803e-01 -1.45148444e+00
9.62295830e-02 -5.80066293e-02 4.09254432e-01 -3.94747823e-01
1.46591529e-01 -9.77401316e-01 1.70838721e-02 2.98497617e-01
-2.49796897e-01 -4.45781827e-01 2.92523921e-01 4.60632861e-01
-6.58328533e-02 -7.96985388e-01 8.94302249e-01 1.49146140e-01
-3.46709937e-01 2.43740827e-01 -3.39676708e-01 5.52473031e-02
7.82448113e-01 -4.39507633e-01 -2.42852628e-01 -1.13631591e-01
-7.51921892e-01 2.42051065e-01 2.14887649e-01 2.04944223e-01
4.35008138e-01 -1.08822966e+00 -5.06569028e-01 3.49186063e-01
-1.83260858e-01 2.19983548e-01 1.73540488e-01 7.35940635e-01
-1.96608603e-01 5.89279197e-02 1.21272534e-01 -6.77742839e-01
-9.32998896e-01 3.35754037e-01 8.04921150e-01 -3.16457063e-01
-4.26509857e-01 7.10945129e-01 6.77438304e-02 -2.42681921e-01
5.07611692e-01 -6.73235178e-01 -2.47242033e-01 -4.02837098e-02
3.60546887e-01 1.56698525e-01 5.59500337e-01 -3.10803413e-01
-1.37130558e-01 4.67944026e-01 -2.01903302e-02 7.41343871e-02
1.25447440e+00 6.71479180e-02 4.07329232e-01 5.38134992e-01
1.09649134e+00 -1.89509705e-01 -1.31304204e+00 -3.49979669e-01
-7.27369413e-02 -3.04954261e-01 2.13914812e-01 -6.60920560e-01
-1.47273552e+00 6.21458292e-01 6.86536014e-01 4.40457940e-01
1.19604409e+00 -2.25180209e-01 6.49441421e-01 6.65815413e-01
2.57550925e-01 -1.30830050e+00 5.39129600e-02 6.35821581e-01
5.22373199e-01 -1.21057487e+00 -3.56136233e-01 -1.06238157e-01
-4.82933104e-01 1.21386290e+00 8.20832551e-01 5.57878092e-02
7.83805549e-01 4.70450401e-01 1.69500172e-01 5.68111204e-02
-6.59769773e-01 -1.59174651e-01 -2.57735662e-02 3.35711479e-01
6.00863844e-02 -1.60657376e-01 -1.31717101e-01 9.46396530e-01
-2.56578088e-01 7.06072152e-02 4.36408013e-01 7.84079313e-01
-5.10367155e-01 -9.49278712e-01 -4.06462818e-01 6.37779236e-01
-5.69132388e-01 -1.24417290e-01 -2.22716361e-01 8.15941513e-01
3.13425034e-01 1.10991240e+00 9.98035371e-02 -4.99258786e-01
5.03680706e-01 1.47394449e-01 2.70764142e-01 -4.24950242e-01
-3.99967670e-01 8.65396783e-02 1.01045705e-02 -1.16557561e-01
-1.93689615e-01 -5.03394425e-01 -1.29401445e+00 -4.08759899e-02
-7.63900936e-01 3.20541382e-01 6.51187241e-01 1.10054708e+00
-7.38307927e-03 5.44872820e-01 7.80163467e-01 -4.87552851e-01
-6.77385032e-01 -9.70784307e-01 -6.79199457e-01 2.15431511e-01
3.42959195e-01 -6.86406612e-01 -7.70611346e-01 -3.26058030e-01] | [9.262702941894531, 3.8676228523254395] |
7754214c-cdaf-4a87-9fda-0bbe9e837175 | education-to-skill-mapping-using-hierarchical | null | null | https://www.mdpi.com/2076-3417/11/13/5868 | https://www.mdpi.com/2076-3417/11/13/5868/pdf | Education-to-Skill Mapping Using Hierarchical Classification and Transformer Neural Network | Skills gained from vocational or higher education form an essential component of country’s economy, determining the structure of the national labor force. Therefore, knowledge on how people’s education converts to jobs enables data-driven choices concerning human resources within an ever-changing job market. Moreover, the relationship between education and occupation is also relevant in times of global crises, such as the COVID-19 pandemic. Healthcare system overload and skill shortage on one hand, and job losses related to lock-downs on the other, have exposed a necessity to identify target groups with relevant education backgrounds in order to facilitate their occupational transitions. However, the relationship between education and employment is complex and difficult to model. This study aims to propose the methodology that would allow us to model education-to-skill mapping. Multiple challenges arising from administrative datasets, namely imbalanced data, complex labeling, hierarchical structure and textual data, were addressed using six neural network-based algorithms of incremental complexity. The final proposed mathematical model incorporates the textual data from descriptions of education programs that are transformed into embeddings, utilizing transformer neural networks. The output of the final model is constructed as the hierarchical classification task. The effectiveness of the proposed model is demonstrated using experiments on national level data, which covers whole population of Lithuania. Finally, we provide the recommendations for the usage of proposed model. This model can be used for practical applications and scenario forecasting. Some possible applications for such model usage are demonstrated and described in this article. The code for this research has been made available on GitHub. | ['Linas Petkevičius', 'Vilija Kuodytė'] | 2021-06-24 | null | null | null | applied-sciences-2021-6 | ['occupation-prediction'] | ['natural-language-processing'] | [ 1.18857883e-01 5.55099808e-02 -3.38045388e-01 -3.17278177e-01
-3.66990753e-02 -3.00827920e-01 5.08431494e-01 5.10342717e-01
-5.99404633e-01 8.05857599e-01 6.01585329e-01 -5.07977307e-01
-7.57053673e-01 -1.15621769e+00 -4.19818103e-01 -5.31926811e-01
3.75989407e-01 6.93096817e-01 -4.31495905e-01 -2.91694701e-01
3.35015118e-01 7.17496097e-01 -1.87575603e+00 -1.25017166e-01
1.07204688e+00 6.41866922e-01 3.98707181e-01 2.64461190e-01
-4.10864830e-01 7.45131969e-01 -4.87971187e-01 -4.48208034e-01
1.64773539e-01 -2.33779609e-01 -8.33039105e-01 6.80599660e-02
1.04570352e-01 -9.16423649e-02 -1.51003554e-01 1.08393645e+00
6.03726983e-01 3.64874065e-01 8.53733599e-01 -1.05025899e+00
-7.75374770e-01 5.03980041e-01 -1.63559034e-01 3.62991095e-01
3.17592949e-01 -4.08656031e-01 4.78043675e-01 -8.77007484e-01
5.17694712e-01 1.18296790e+00 5.92886031e-01 5.51070534e-02
-8.50978732e-01 -7.17361093e-01 -1.89349368e-01 4.48998719e-01
-1.17569804e+00 -2.22587198e-01 7.97797084e-01 -8.71619225e-01
8.48514497e-01 1.85981467e-01 8.38476837e-01 6.36162937e-01
2.06001222e-01 1.36758342e-01 1.38832545e+00 -6.71444476e-01
-2.96545923e-01 7.24571288e-01 2.59703606e-01 7.63938785e-01
4.97872919e-01 4.50255349e-02 -3.80332619e-01 1.78951025e-01
4.84517246e-01 3.91132176e-01 1.41033486e-01 -1.09404579e-01
-7.45086014e-01 8.68927598e-01 2.38898024e-01 7.07893431e-01
-5.36874771e-01 -4.95244324e-01 3.26705605e-01 3.82141650e-01
4.39168632e-01 3.34162891e-01 -3.34808975e-01 -1.64146736e-01
-9.90912557e-01 1.87861934e-01 5.37909210e-01 6.60786569e-01
6.20026052e-01 5.30577935e-02 -1.36611247e-02 7.77271688e-01
6.87742382e-02 3.80332440e-01 6.13594174e-01 -6.57667279e-01
4.35957104e-01 1.00443327e+00 -6.07292317e-02 -1.54066634e+00
-5.79356372e-01 -7.87324846e-01 -1.09797478e+00 -1.35881037e-01
4.94733512e-01 -8.99137929e-02 -7.23303199e-01 1.50655365e+00
3.32300544e-01 -1.15802355e-01 1.83224574e-01 6.40390217e-01
8.60788167e-01 5.70304334e-01 5.78473322e-03 -1.99514896e-01
1.57393503e+00 -5.93647063e-01 -9.40952837e-01 6.50828481e-02
3.78441781e-01 -7.99874365e-01 8.19811046e-01 4.78144027e-02
-1.26110184e+00 -9.11692739e-01 -6.03419542e-01 -9.95383337e-02
-8.77512753e-01 2.25780979e-01 5.05398214e-01 7.57167578e-01
-8.80316257e-01 5.70877671e-01 -5.36789954e-01 -5.47906339e-01
2.06934646e-01 4.72000569e-01 -2.53466755e-01 8.57382342e-02
-1.55586839e+00 1.15386796e+00 4.94164646e-01 2.39923313e-01
-2.25506842e-01 -6.03351414e-01 -9.35763657e-01 2.92403877e-01
1.01113291e-02 -5.13660133e-01 6.26499474e-01 -8.40723813e-01
-8.55261147e-01 6.99454248e-01 -1.47876620e-01 -2.19228283e-01
4.90169793e-01 2.26716340e-01 -6.18350983e-01 -2.87134796e-01
1.20619252e-01 2.67705768e-01 1.78545818e-01 -7.82520711e-01
-1.02501678e+00 -7.21813083e-01 -1.14828810e-01 4.82907683e-01
-9.00429606e-01 1.75885350e-01 -6.48080409e-02 -3.88745606e-01
-6.75866157e-02 -6.27322853e-01 -9.04823095e-02 -9.44186687e-01
3.99250537e-03 -4.28893536e-01 3.73754352e-01 -1.08467627e+00
1.56198823e+00 -1.77301657e+00 1.28649995e-01 3.72302592e-01
-8.49037692e-02 1.27906159e-01 3.46684277e-01 5.77557981e-01
-1.01136684e-01 -2.01881658e-02 6.81961849e-02 -5.19493483e-02
2.22868741e-01 7.88347200e-02 1.00832514e-01 3.13493937e-01
9.11187977e-02 5.09497583e-01 -7.41898000e-01 -5.20903826e-01
3.71331632e-01 4.81757969e-01 -3.43492866e-01 1.48028821e-01
5.92221260e-01 5.05259037e-01 -1.47952616e-01 7.51543224e-01
5.74267626e-01 1.65105566e-01 3.49338055e-01 -7.30801793e-03
-5.17153919e-01 2.27875054e-01 -1.19279623e+00 1.14318311e+00
-6.24150753e-01 5.31922460e-01 -1.08719438e-01 -1.40539491e+00
1.04564404e+00 3.55283290e-01 3.71607542e-01 -9.88595188e-01
2.69154876e-01 1.98146284e-01 1.67352989e-01 -6.45541608e-01
7.42136657e-01 -1.63204715e-01 -1.45923868e-01 1.35918453e-01
-1.04422905e-01 1.22383349e-01 4.78630692e-01 -1.10808119e-01
5.54673493e-01 2.17002202e-02 3.40272665e-01 -5.01322448e-01
8.97328019e-01 4.51009646e-02 6.37494564e-01 3.25098306e-01
-1.88600153e-01 -1.78229675e-01 3.98379534e-01 -7.46314764e-01
-1.20586371e+00 -6.54995561e-01 -2.92299062e-01 1.02705193e+00
-2.58561552e-01 3.87661755e-01 -7.22754121e-01 -2.10746318e-01
-1.00081572e-02 5.41296482e-01 -4.43105698e-01 -3.03744432e-02
-5.11327684e-01 -5.73598385e-01 3.05738837e-01 2.37678602e-01
5.01754344e-01 -1.34040761e+00 -7.57487535e-01 1.37283504e-01
-2.66527325e-01 -7.80515730e-01 -1.03729263e-01 1.30942523e-01
-7.56289005e-01 -8.53030682e-01 -5.96763492e-01 -1.12418258e+00
8.56404781e-01 -1.34906232e-01 1.24363279e+00 1.12081848e-01
-2.58011401e-01 2.42510378e-01 -8.44313502e-02 -5.76773286e-01
-4.02238697e-01 4.60076600e-01 5.34799993e-01 -2.31188700e-01
7.69246161e-01 -6.41396940e-01 -4.75039631e-01 -2.85955127e-02
-9.15333807e-01 7.01423809e-02 4.90367711e-01 8.35141659e-01
3.37528884e-01 8.17304432e-01 8.80487263e-01 -1.01524305e+00
9.88994479e-01 -7.35793471e-01 -5.51571488e-01 1.44677147e-01
-9.88761187e-01 -1.02019601e-01 7.28538752e-01 -1.66075528e-01
-1.36658740e+00 -7.42363110e-02 -3.35518122e-01 -3.04649379e-02
-3.20909411e-01 8.15729022e-01 -2.97022257e-02 4.89801615e-01
2.83792883e-01 2.70289272e-01 -1.02818668e-01 -5.97298443e-01
-6.08321168e-02 1.20059025e+00 6.13170803e-01 -6.50483847e-01
7.21282542e-01 4.49063256e-02 7.74246603e-02 -7.60774553e-01
-5.12680173e-01 -4.61509019e-01 -8.47072959e-01 -3.59020114e-01
7.03470826e-01 -7.77213633e-01 -7.74943769e-01 2.91525215e-01
-9.14832532e-01 1.35969862e-01 -1.30239591e-01 5.32487810e-01
-2.15878978e-01 -7.52528980e-02 -3.38434815e-01 -1.09847701e+00
-2.78894603e-01 -1.06154335e+00 3.05481076e-01 5.78988194e-01
-1.81952402e-01 -1.38447464e+00 -8.76891091e-02 9.23596978e-01
3.89094442e-01 2.90918410e-01 1.34000158e+00 -6.20095253e-01
-2.60004699e-01 9.82949324e-03 -8.12508091e-02 3.82892251e-01
3.95400435e-01 -3.03195775e-01 -5.69168687e-01 -3.10885668e-01
-1.17653951e-01 -2.68824130e-01 5.03259361e-01 3.51529539e-01
9.66330767e-01 -3.21392983e-01 8.45355466e-02 2.59925336e-01
1.44811988e+00 4.35163081e-01 4.56861377e-01 4.41357821e-01
5.59322178e-01 1.16988313e+00 8.35924089e-01 3.89498085e-01
6.47925258e-01 4.59536701e-01 4.90629338e-02 -2.96628118e-01
-6.25427365e-02 -1.74605519e-01 7.89381489e-02 1.35039723e+00
-3.93199146e-01 1.57645315e-01 -1.14854741e+00 9.01711762e-01
-1.47450757e+00 -1.12388968e+00 -5.95558472e-02 2.04306436e+00
9.13825631e-01 1.63862899e-01 1.38025254e-01 4.55486089e-01
5.76806128e-01 -2.28247464e-01 -3.25921327e-01 -9.12661850e-01
3.12976331e-01 3.95090371e-01 6.20238423e-01 5.23567915e-01
-8.42432916e-01 6.96202755e-01 5.48744345e+00 6.63335800e-01
-8.78566206e-01 1.17633194e-01 6.57097697e-01 2.07112536e-01
-2.96490133e-01 -3.60474885e-01 -9.04117703e-01 5.35483420e-01
1.11968529e+00 -3.13592434e-01 6.51224554e-01 4.47249472e-01
6.60577893e-01 8.90606456e-03 -5.86098135e-01 7.81413019e-01
-9.01837498e-02 -1.06020236e+00 -1.96869984e-01 2.20800951e-01
7.70243168e-01 -4.81711000e-01 2.88737446e-01 5.38891613e-01
1.06107622e-01 -1.14561605e+00 5.03070474e-01 6.53276205e-01
8.32256496e-01 -1.25558329e+00 1.06967783e+00 6.25018716e-01
-1.06985557e+00 -4.35536206e-01 -3.45856011e-01 -7.35504687e-01
-1.26265571e-01 5.54482937e-01 -9.58664417e-01 8.02465439e-01
6.57555878e-01 3.87425423e-01 -4.78526145e-01 4.24975485e-01
2.81978667e-01 4.69500333e-01 1.88372761e-01 -1.24069070e-02
3.52461748e-02 -5.05152524e-01 1.91109218e-02 1.18450987e+00
8.54641914e-01 1.49886273e-02 -3.61830927e-02 5.99413097e-01
-1.40201762e-01 3.90954256e-01 -9.54131186e-01 -3.37552220e-01
7.06337333e-01 1.18041480e+00 -7.51367271e-01 -2.38699690e-01
-5.35714447e-01 5.86655140e-01 2.78451562e-01 7.87101761e-02
-6.26638114e-01 -5.18838406e-01 6.03579104e-01 4.08282578e-01
-2.10928708e-01 4.15038466e-02 -2.32112318e-01 -6.84988976e-01
-1.29207835e-01 -9.95615602e-01 3.18829745e-01 -1.75713912e-01
-9.04007435e-01 3.74900758e-01 -5.18865511e-02 -8.50595772e-01
-2.68547744e-01 -5.91459811e-01 -4.21195887e-02 1.12459242e+00
-1.45997167e+00 -9.22807455e-01 -1.68666959e-01 3.98953140e-01
5.63737929e-01 -5.45029879e-01 8.64300907e-01 9.73747611e-01
-7.12911665e-01 5.06166041e-01 1.35112748e-01 1.48498833e-01
3.79356951e-01 -1.23436046e+00 -1.99353263e-01 8.67194414e-01
-1.91165134e-01 6.67557001e-01 6.25270724e-01 -7.81583130e-01
-9.81365502e-01 -1.13239586e+00 1.76719379e+00 -3.38071883e-01
4.66946423e-01 -6.96725696e-02 -6.00682735e-01 4.38610196e-01
2.70424068e-01 -6.39047801e-01 7.65990615e-01 2.01849997e-01
3.98538947e-01 -1.56237990e-01 -1.28381324e+00 3.42482567e-01
9.47469413e-01 -4.35520351e-01 -6.08490884e-01 1.98592886e-01
4.03437227e-01 -2.89642185e-01 -1.25836682e+00 3.24650586e-01
6.86097443e-01 -1.12229347e+00 1.03166628e+00 -6.16623819e-01
6.05320692e-01 4.32949737e-02 5.42062102e-03 -1.30943382e+00
-6.23319387e-01 1.09590098e-01 1.00121655e-01 1.34695005e+00
1.91893160e-01 -6.53820992e-01 7.58600473e-01 4.95004177e-01
-5.52591197e-02 -8.47932458e-01 -8.29159319e-01 -5.51524401e-01
1.88104168e-01 7.01381713e-02 9.33485448e-01 1.43824208e+00
-1.92895561e-01 1.31457821e-01 -3.57456088e-01 5.09117804e-02
3.18633139e-01 -5.30224154e-03 4.93964523e-01 -1.59300745e+00
7.50566944e-02 -4.23691154e-01 -4.27501500e-01 -2.39889324e-01
2.48135075e-01 -9.50057685e-01 -4.79354650e-01 -1.84002650e+00
4.69382614e-01 -4.51009452e-01 -5.04052162e-01 2.32244328e-01
2.84355506e-02 8.52633193e-02 1.81812093e-01 -1.70432165e-01
1.06357768e-01 1.29836500e-01 1.24699950e+00 -1.38552725e-01
-3.45946491e-01 9.76623446e-02 -8.11674416e-01 6.80459797e-01
9.77570355e-01 -4.33139026e-01 -4.86059755e-01 -2.34083071e-01
5.12067974e-01 1.34794742e-01 -1.31949782e-01 -1.01741433e+00
2.70848036e-01 -4.53585386e-01 5.02220273e-01 -6.63572073e-01
1.64495572e-01 -1.41977644e+00 5.50286889e-01 6.93596005e-01
-2.07299113e-01 5.15109479e-01 -6.50985986e-02 -3.75148654e-02
-3.65254402e-01 -4.31202888e-01 5.23221433e-01 -9.86888409e-02
-4.09278810e-01 5.26392609e-02 -4.03562427e-01 -2.68755525e-01
9.04413223e-01 -3.25663954e-01 -2.97617912e-02 -2.53044426e-01
-7.59144902e-01 2.24402547e-01 2.96115518e-01 5.99466264e-01
3.75962079e-01 -1.62108994e+00 -6.59808815e-01 3.21323395e-01
-3.52331340e-01 -2.50329107e-01 2.61098236e-01 8.69233370e-01
-5.75080156e-01 8.49870265e-01 -7.08154559e-01 6.06429912e-02
-1.51788759e+00 5.44041216e-01 1.18087545e-01 -5.92377126e-01
-4.07817774e-02 3.53063583e-01 -2.75208145e-01 -9.36469495e-01
3.43120784e-01 -2.79720366e-01 -9.18238699e-01 5.07618070e-01
2.15403065e-01 8.12493980e-01 1.69237033e-01 -9.03878391e-01
-2.57382631e-01 3.76820862e-01 4.18177068e-01 5.77937327e-02
1.32205415e+00 -2.83516169e-01 -3.38884741e-01 5.15883744e-01
9.06114340e-01 1.10403664e-01 -4.98646349e-01 -2.05500036e-01
1.28733948e-01 -5.28979123e-01 4.72192392e-02 -7.86856055e-01
-9.00136471e-01 9.06474352e-01 8.24007988e-01 3.82549673e-01
1.44636619e+00 -4.40117240e-01 5.83206475e-01 1.80915296e-01
6.04122654e-02 -1.48105848e+00 -3.17977071e-01 5.35135150e-01
3.64278883e-01 -1.15510178e+00 -1.60387918e-01 -3.50567624e-02
-2.73963690e-01 9.55717325e-01 3.79892260e-01 5.51805079e-01
4.74100709e-01 6.22496642e-02 1.09790854e-01 -9.64949206e-02
-3.14893633e-01 -3.07603568e-01 2.41808638e-01 5.35161197e-01
6.78304434e-01 4.12722766e-01 -7.47364461e-01 6.81467235e-01
-5.28457880e-01 1.87951416e-01 4.22581822e-01 6.84718847e-01
-5.10528028e-01 -1.33236337e+00 -7.59881258e-01 6.89772189e-01
-6.80352569e-01 -1.72556881e-02 -9.77963433e-02 7.06959784e-01
9.06934679e-01 9.40437555e-01 3.10779691e-01 -2.86459118e-01
4.91724432e-01 3.98960948e-01 3.09325337e-01 -3.72382879e-01
-5.85711300e-01 -3.61009389e-01 1.21889740e-01 1.86347365e-01
-5.24450898e-01 -8.13327193e-01 -1.14238453e+00 -7.68124759e-01
1.25945792e-01 3.92944515e-01 8.46463919e-01 9.09722090e-01
1.87930375e-01 9.41681325e-01 5.88680446e-01 -5.32618523e-01
-4.33955133e-01 -1.14811325e+00 -5.78831255e-01 4.24085408e-01
5.52954823e-02 -7.79979229e-01 -5.03692813e-02 2.84121884e-03] | [9.737175941467285, 9.45209789276123] |
7479199e-a8f3-4de2-bcd8-e20464d3a738 | greedy-layer-pruning-decreasing-inference | 2105.14839 | null | https://arxiv.org/abs/2105.14839v2 | https://arxiv.org/pdf/2105.14839v2.pdf | Greedy-layer Pruning: Speeding up Transformer Models for Natural Language Processing | Fine-tuning transformer models after unsupervised pre-training reaches a very high performance on many different natural language processing tasks. Unfortunately, transformers suffer from long inference times which greatly increases costs in production. One possible solution is to use knowledge distillation, which solves this problem by transferring information from large teacher models to smaller student models. Knowledge distillation maintains high performance and reaches high compression rates, nevertheless, the size of the student model is fixed after pre-training and can not be changed individually for a given downstream task and use-case to reach a desired performance/speedup ratio. Another solution to reduce the size of models in a much more fine-grained and computationally cheaper fashion is to prune layers after the pre-training. The price to pay is that the performance of layer-wise pruning algorithms is not on par with state-of-the-art knowledge distillation methods. In this paper, Greedy-layer pruning is introduced to (1) outperform current state-of-the-art for layer-wise pruning, (2) close the performance gap when compared to knowledge distillation, while (3) providing a method to adapt the model size dynamically to reach a desired performance/speedup tradeoff without the need of additional pre-training phases. Our source code is available on https://github.com/deepopinion/greedy-layer-pruning. | ['Antonio Rodriguez-Sanchez', 'Stefan Engl', 'Sebastian Stabinger', 'David Peer'] | 2021-05-31 | null | null | null | null | ['unsupervised-pre-training'] | ['methodology'] | [ 2.28901370e-03 3.35938901e-01 -2.27350220e-01 -2.92326719e-01
-6.65349126e-01 -5.07892609e-01 3.57871264e-01 4.30840582e-01
-6.29311502e-01 6.81287885e-01 -2.24178731e-02 -6.51599050e-01
-2.70784795e-01 -9.93082345e-01 -6.43035114e-01 -4.52861667e-01
2.86207646e-01 8.62964094e-01 6.73863649e-01 1.45544317e-02
1.29207760e-01 4.81293440e-01 -1.64231884e+00 3.21646392e-01
1.03240824e+00 8.25104058e-01 4.51015800e-01 6.56999052e-01
-3.82303089e-01 7.23514616e-01 -4.23888654e-01 -6.54891253e-01
3.14303815e-01 -1.78800792e-01 -1.06366253e+00 -2.67570257e-01
4.02611732e-01 -1.86246961e-01 -9.22929719e-02 9.31796670e-01
3.26397419e-01 1.83781087e-01 3.65573645e-01 -8.33103180e-01
-1.72848612e-01 1.16677427e+00 -4.20599192e-01 3.36391717e-01
-2.51677454e-01 -2.75267158e-02 1.04377532e+00 -7.30280697e-01
3.59534860e-01 1.11553240e+00 5.71334243e-01 3.43727142e-01
-1.29813910e+00 -7.91922867e-01 2.18131751e-01 2.81650186e-01
-1.53218210e+00 -5.16949236e-01 4.06455159e-01 -3.31094682e-01
1.43153989e+00 1.96227953e-01 9.14390087e-01 4.69845563e-01
-2.98290789e-01 8.70441318e-01 8.48929048e-01 -5.77072859e-01
1.16655111e-01 2.05277383e-01 1.88588545e-01 1.01489460e+00
4.06567752e-01 -1.64007604e-01 -5.16090751e-01 -8.83813202e-02
7.34145999e-01 -5.64375184e-02 -1.28290921e-01 -3.45833987e-01
-8.65997851e-01 7.64155388e-01 3.59931469e-01 5.20226419e-01
-2.39979833e-01 2.30241641e-01 3.91721338e-01 5.07939696e-01
3.07485282e-01 8.38994205e-01 -1.11365974e+00 -5.02049208e-01
-1.62504458e+00 2.46328831e-01 9.50301409e-01 7.18284070e-01
9.04178262e-01 1.59735493e-02 -1.45298213e-01 9.93878603e-01
5.58178984e-02 1.24587566e-01 6.01192057e-01 -9.50217962e-01
5.81790984e-01 7.25173354e-01 -3.37302715e-01 -5.37725687e-01
-1.53530777e-01 -6.64584756e-01 -6.37028754e-01 7.36456588e-02
5.72465658e-01 -1.58400297e-01 -1.05984259e+00 1.55667615e+00
2.71151751e-01 1.34837732e-01 6.68907687e-02 3.90916705e-01
6.22712731e-01 6.28171682e-01 1.10139072e-01 -1.54837638e-01
1.47452390e+00 -1.21398079e+00 -1.78552181e-01 -4.91570026e-01
8.69228482e-01 -7.21030653e-01 1.14691699e+00 7.20995247e-01
-1.40008330e+00 -2.70190030e-01 -9.40798461e-01 -2.22397134e-01
-4.94018167e-01 8.75407755e-02 7.31511652e-01 6.54158175e-01
-8.82466316e-01 7.24888861e-01 -1.04335582e+00 -1.51214629e-01
5.44124007e-01 6.18191063e-01 -1.81600988e-01 -1.69924200e-02
-9.87109065e-01 1.13781607e+00 8.97578180e-01 -2.54842311e-01
-6.81220233e-01 -1.19944084e+00 -8.14303219e-01 5.98256648e-01
7.25108743e-01 -9.51599419e-01 1.62357950e+00 -6.45739138e-01
-1.72534823e+00 5.90642929e-01 -1.20249271e-01 -6.98712349e-01
4.20741767e-01 -3.94214779e-01 1.86554492e-01 -4.55851257e-02
-2.67477900e-01 7.04425871e-01 6.15470350e-01 -8.34873617e-01
-7.33808994e-01 -8.22152868e-02 2.31314555e-01 3.13560426e-01
-4.91866678e-01 -1.16775028e-01 -8.16758037e-01 -5.68270922e-01
7.62327686e-02 -7.47910202e-01 -3.66695464e-01 -2.82563716e-01
-1.50054693e-01 -2.86507219e-01 6.83022022e-01 -4.83532578e-01
1.45425379e+00 -1.86534476e+00 1.75459042e-01 9.71158817e-02
3.16288561e-01 8.40079248e-01 -3.10388170e-02 2.01171860e-01
6.68972656e-02 2.20077872e-01 -2.64935374e-01 -3.96029830e-01
-9.23220292e-02 4.08857137e-01 -2.03629181e-01 -1.28115937e-01
7.68358484e-02 8.93307209e-01 -9.79092777e-01 -6.11122131e-01
1.71744570e-01 3.53072673e-01 -1.05363333e+00 -7.69883767e-02
-4.80369538e-01 2.28647757e-02 -3.59412909e-01 3.42253745e-01
3.90065074e-01 -1.77692503e-01 2.18774468e-01 -7.51145408e-02
-3.60424891e-02 9.25492227e-01 -1.23384380e+00 1.62349033e+00
-7.30893314e-01 4.82510239e-01 1.16543911e-01 -1.27892637e+00
7.46519804e-01 2.22012356e-01 1.65685371e-01 -5.18963397e-01
7.56938085e-02 2.46722952e-01 1.82892814e-01 -1.94420964e-01
4.63377059e-01 -1.84505776e-01 -2.98415637e-03 3.76454294e-01
2.62946457e-01 -4.32957470e-01 6.20914519e-01 2.35209763e-01
1.22855020e+00 1.10447340e-01 2.74358153e-01 -2.56426632e-01
4.14892495e-01 5.84711283e-02 7.48767674e-01 6.62090957e-01
2.85781264e-01 2.19830602e-01 5.99281311e-01 -3.70073915e-01
-9.63657260e-01 -9.23821449e-01 -5.57915270e-02 1.34688890e+00
-4.16455507e-01 -8.03918481e-01 -7.99909711e-01 -6.18152082e-01
1.27567902e-01 8.01652789e-01 -1.99239239e-01 -1.99651599e-01
-6.49924576e-01 -5.28443217e-01 8.33521962e-01 6.08355522e-01
7.61909544e-01 -9.46447790e-01 -7.69240081e-01 3.13672990e-01
-2.91991830e-02 -1.03603578e+00 2.47094985e-02 7.35124767e-01
-1.29954326e+00 -9.51330483e-01 -3.41987759e-01 -6.66934848e-01
6.93125784e-01 2.22074706e-02 1.35603786e+00 1.25847638e-01
-1.00627542e-01 -1.05284221e-01 -1.27986342e-01 -4.52558160e-01
-3.17648858e-01 6.21961176e-01 -5.76080428e-03 -7.51197159e-01
3.78842533e-01 -7.76732981e-01 -3.67651165e-01 -1.65131371e-02
-7.68716455e-01 3.90078008e-01 8.33931983e-01 7.35785365e-01
6.43159926e-01 5.70719421e-01 2.50050038e-01 -1.12889385e+00
6.19823635e-01 -4.26300652e-02 -7.75671005e-01 3.30293864e-01
-9.34229314e-01 4.73386824e-01 7.14552164e-01 -4.82408822e-01
-9.85345185e-01 -2.37490609e-02 -5.40637448e-02 -5.86307347e-01
3.13217863e-02 6.97317004e-01 5.46417534e-02 2.33909979e-01
6.41547084e-01 6.38220981e-02 -2.47013211e-01 -7.48207211e-01
5.05372167e-01 3.40197831e-01 4.10511315e-01 -8.75723481e-01
7.45575607e-01 -5.60039431e-02 -2.60688186e-01 -6.62740946e-01
-1.04839253e+00 -3.18882138e-01 -7.12828398e-01 4.81935740e-01
3.49763215e-01 -9.01147366e-01 -4.51959133e-01 2.46351525e-01
-9.09523487e-01 -8.29343140e-01 -5.94007432e-01 3.21123749e-01
-3.13073367e-01 1.32733509e-01 -6.62690282e-01 -4.22495574e-01
-6.36246562e-01 -1.05327833e+00 6.60161734e-01 2.39971608e-01
-2.52465606e-01 -9.00453269e-01 -7.25320354e-02 6.00550115e-01
4.90788072e-01 -3.33092719e-01 1.06473362e+00 -7.24227190e-01
-5.37292898e-01 -7.51396567e-02 -2.17802852e-01 4.81667340e-01
-2.19241142e-01 -6.41785040e-02 -7.84130216e-01 -1.91644624e-01
-1.77550435e-01 -3.37618202e-01 1.10084248e+00 1.99997872e-01
1.35696244e+00 -4.61732745e-01 -3.97916913e-01 6.24855578e-01
1.18435526e+00 -1.75836086e-01 4.84956682e-01 4.04205978e-01
6.21022463e-01 1.51584119e-01 4.65299457e-01 2.16649055e-01
3.73461246e-01 5.12031555e-01 7.79832294e-03 7.10053965e-02
-2.89351910e-01 -3.92937154e-01 2.31804669e-01 9.39591467e-01
-5.97319240e-03 5.15900031e-02 -1.20463037e+00 7.84000576e-01
-1.90488303e+00 -7.47574091e-01 2.68612176e-01 2.17760420e+00
1.32535231e+00 5.29538631e-01 7.72435442e-02 3.85678172e-01
2.31909201e-01 -1.82513535e-01 -4.29531038e-01 -7.94243813e-01
3.15899044e-01 6.53599143e-01 5.70951104e-01 6.45701826e-01
-7.79905498e-01 1.43890560e+00 5.72143459e+00 1.09341896e+00
-1.10118997e+00 2.77025163e-01 4.19848382e-01 -5.46620011e-01
-7.57819042e-02 2.75132596e-01 -1.15077662e+00 1.45840660e-01
1.15260780e+00 -3.04108739e-01 5.71722269e-01 9.87321317e-01
-4.12725657e-02 -2.22195715e-01 -1.03848660e+00 8.10384691e-01
-3.32088292e-01 -1.40732765e+00 3.85471024e-02 -4.35379148e-02
6.56176031e-01 3.58037382e-01 -2.85682321e-01 8.41521382e-01
6.19211257e-01 -8.28644574e-01 5.34319341e-01 2.22093508e-01
6.29489958e-01 -8.63067389e-01 5.59265614e-01 6.61099732e-01
-1.18172669e+00 -1.70127660e-01 -5.23983896e-01 -2.79022366e-01
-7.57658333e-02 9.41963732e-01 -9.81460094e-01 1.81859374e-01
8.48557413e-01 2.99100578e-01 -6.50300801e-01 1.13590288e+00
-6.08348846e-01 9.05123949e-01 -8.23358774e-01 1.11719593e-01
3.80207717e-01 5.26385829e-02 2.34524459e-01 1.51323450e+00
1.30646259e-01 1.06240869e-01 1.15540631e-01 7.78207064e-01
-2.61218578e-01 -1.17700677e-02 -3.37388486e-01 -1.67852819e-01
7.87014961e-01 1.20217144e+00 -6.66089714e-01 -6.80337310e-01
-2.09644362e-01 6.23783827e-01 8.24886978e-01 8.14542845e-02
-7.04890430e-01 -5.15664995e-01 5.59526443e-01 3.96553218e-01
7.64699221e-01 -3.10160756e-01 -4.29861188e-01 -1.13309836e+00
-5.29630631e-02 -8.07065845e-01 4.81692910e-01 -4.64632809e-01
-7.33376563e-01 2.89380729e-01 2.18749076e-01 -4.93689924e-01
-2.85146505e-01 -4.13998008e-01 -4.97734040e-01 7.39593744e-01
-1.53593242e+00 -9.65053380e-01 -7.98001140e-02 4.98690873e-01
5.82871318e-01 -2.99366284e-03 8.94428611e-01 2.75070608e-01
-5.19997835e-01 8.90804529e-01 -4.24249582e-02 3.86377200e-02
4.42606896e-01 -1.24858427e+00 2.58555919e-01 8.11712146e-01
2.74951398e-01 8.06991220e-01 4.34806317e-01 -4.19449389e-01
-1.20117414e+00 -9.64025915e-01 1.20931637e+00 -3.04038227e-01
7.16804266e-01 -3.01231623e-01 -1.02344227e+00 7.25437045e-01
-4.13442850e-02 -1.40063107e-01 6.66813552e-01 6.47312582e-01
-5.53327382e-01 -2.05899000e-01 -1.05746400e+00 5.64754188e-01
8.98359001e-01 -3.74505043e-01 -8.77394915e-01 1.15382485e-01
6.63453758e-01 -4.65808392e-01 -9.99796391e-01 4.57742751e-01
4.69181657e-01 -8.85310829e-01 9.37266290e-01 -3.56595695e-01
4.07276869e-01 -2.16118217e-01 1.44255951e-01 -1.20288682e+00
-4.48452473e-01 -6.60778403e-01 -5.64282417e-01 1.24889314e+00
7.08604157e-01 -6.10865533e-01 1.08548474e+00 6.54160500e-01
-5.96049540e-02 -1.19716692e+00 -7.38667548e-01 -7.61772752e-01
2.66010314e-01 -5.49776852e-01 6.15758002e-01 8.65338862e-01
1.73887417e-01 6.97335184e-01 1.05748087e-01 -4.27841172e-02
3.62171024e-01 1.66166887e-01 8.15565944e-01 -1.29738867e+00
-6.29345059e-01 -7.94531107e-01 -1.76111653e-01 -1.23577309e+00
-5.59205748e-02 -1.06958485e+00 -2.43373424e-01 -1.70514083e+00
2.16919467e-01 -8.42807114e-01 -3.29548150e-01 1.23538673e+00
-1.23161696e-01 7.95540512e-02 4.01793182e-01 8.71909708e-02
-5.17581820e-01 3.68904799e-01 1.03110301e+00 -7.50501677e-02
-4.91688699e-01 -1.37972891e-01 -8.42519522e-01 9.10631478e-01
1.04260206e+00 -6.22336686e-01 -6.40581191e-01 -6.80194736e-01
2.82928795e-01 -1.73429385e-01 -2.97510438e-02 -1.13535333e+00
4.27267820e-01 6.79206476e-02 2.24836200e-01 -4.27383602e-01
3.18049490e-01 -6.68907881e-01 -7.19997808e-02 5.95753133e-01
-1.65459812e-01 -8.01503733e-02 5.87959766e-01 1.98834449e-01
-1.26048997e-01 -6.08550191e-01 9.43574190e-01 -2.87686616e-01
-5.58075309e-01 1.28283128e-01 -3.84242624e-01 1.75474212e-01
7.66260147e-01 -3.08989614e-01 -2.22854331e-01 1.24510288e-01
-7.55020916e-01 3.21706444e-01 3.39726835e-01 3.19098473e-01
2.52127320e-01 -7.08771288e-01 -5.09787023e-01 2.43518308e-01
-2.00644985e-01 5.65344393e-01 -1.03045531e-01 6.95839465e-01
-6.31517649e-01 5.70692837e-01 8.58310685e-02 -3.89679492e-01
-1.41306269e+00 3.67550701e-01 2.94159859e-01 -9.77362871e-01
-7.54087150e-01 1.15826631e+00 -9.09769386e-02 -5.93529701e-01
3.38562250e-01 -5.35002828e-01 1.74188659e-01 2.93410737e-02
4.15388227e-01 2.62836397e-01 3.88937205e-01 8.14980119e-02
-2.78528720e-01 2.93274581e-01 -5.11250436e-01 3.50651480e-02
1.51456761e+00 4.12931442e-01 -9.09017026e-02 1.12520792e-01
8.19655597e-01 -1.48682818e-01 -1.01083374e+00 -5.38303792e-01
-2.47595608e-02 -2.34172076e-01 5.13270199e-01 -1.02123392e+00
-1.29995608e+00 9.07404721e-01 1.23309806e-01 2.19132868e-03
1.12523401e+00 1.04457811e-02 8.35005879e-01 8.30251753e-01
2.68491775e-01 -1.16056693e+00 -1.42129138e-01 8.46699834e-01
3.97952974e-01 -8.36242795e-01 4.03068393e-01 -3.32424194e-01
-4.16977853e-01 9.41958785e-01 6.45349443e-01 -9.02153552e-02
7.03318715e-01 6.22077346e-01 -2.31049180e-01 -2.20467538e-01
-1.04042935e+00 -2.83947945e-01 1.47405967e-01 2.66489267e-01
5.29436648e-01 1.27969071e-01 -1.84511945e-01 5.97949266e-01
-7.73660660e-01 1.86776027e-01 2.24463586e-02 1.04223895e+00
-7.29312181e-01 -1.41353405e+00 -1.40690252e-01 7.42305100e-01
-5.47098339e-01 -4.57407326e-01 -1.85177490e-01 6.97794855e-01
3.08607250e-01 8.00619304e-01 -2.26124469e-02 -2.06381619e-01
3.95905882e-01 3.74100953e-01 8.24862659e-01 -9.40570712e-01
-9.56327319e-01 -1.21648215e-01 2.24127650e-01 -3.85392785e-01
9.97706410e-03 -3.76142412e-01 -1.34994221e+00 -5.72699964e-01
-5.45680761e-01 3.55607480e-01 6.15702808e-01 9.60511327e-01
5.30163527e-01 6.15586877e-01 -2.03979492e-01 -7.68403471e-01
-5.10633767e-01 -9.55789208e-01 -1.58368275e-01 -1.95270069e-02
-1.19208641e-01 -7.08940446e-01 -1.37867868e-01 -1.38147660e-02] | [8.775311470031738, 3.5990819931030273] |
73780f12-eece-4f78-a806-c57573af5051 | higher-order-integration-of-hierarchical | null | null | http://openaccess.thecvf.com/content_iccv_2017/html/Cai_Higher-Order_Integration_of_ICCV_2017_paper.html | http://openaccess.thecvf.com/content_ICCV_2017/papers/Cai_Higher-Order_Integration_of_ICCV_2017_paper.pdf | Higher-Order Integration of Hierarchical Convolutional Activations for Fine-Grained Visual Categorization | The success of fine-grained visual categorization (FGVC) extremely relies on the modeling of appearance and interactions of various semantic parts. This makes FGVC very challenging because: (i) part annotation and detection require expert guidance and are very expensive; (ii) parts are of different sizes; and (iii) the part interactions are complex and of higher-order. To address these issues, we propose an end-to-end framework based on higher-order integration of hierarchical convolutional activations for FGVC. By treating the convolutional activations as local descriptors, hierarchical convolutional activations can serve as a representation of local parts from different scales. A polynomial kernel based predictor is proposed to capture higher-order statistics of convolutional activations for modeling part interaction. To model inter-layer part interactions, we extend polynomial predictor to integrate hierarchical activations via kernel fusion. Our work also provides a new perspective for combining convolutional activations from multiple layers. While hypercolumns simply concatenate maps from different layers, and holistically-nested network uses weighted fusion to combine side-outputs, our approach exploits higher-order intra-layer and inter-layer relations for better integration of hierarchical convolutional features. The proposed framework yields more discriminative representation and achieves competitive results on the widely used FGVC datasets.
| ['WangMeng Zuo', 'Sijia Cai', 'Lei Zhang'] | 2017-10-01 | null | null | null | iccv-2017-10 | ['fine-grained-visual-categorization'] | ['computer-vision'] | [-1.87146813e-01 -2.15299904e-01 -1.27381191e-01 -3.87054622e-01
-5.05575716e-01 -5.92873454e-01 4.14076000e-01 4.29417133e-01
-2.80652434e-01 2.68962830e-01 1.81308985e-01 1.39679924e-01
4.09762450e-02 -7.64775574e-01 -7.62185693e-01 -6.57047093e-01
-1.47651255e-01 -2.27273498e-02 7.64620483e-01 -2.33372953e-02
9.61050913e-02 8.08696449e-01 -1.85298514e+00 9.14687932e-01
7.36150324e-01 1.64707756e+00 2.33323902e-01 6.08077705e-01
-5.07071614e-01 8.04992318e-01 -5.06387174e-01 -2.33030140e-01
2.86813319e-01 5.30075803e-02 -8.24008882e-01 2.76460439e-01
8.18006873e-01 -8.75013471e-02 -3.32085751e-02 1.00022840e+00
1.95538089e-01 -1.31485254e-01 8.61181557e-01 -1.36192381e+00
-6.28964543e-01 4.06931132e-01 -8.15647423e-01 -7.68911391e-02
-1.33451670e-01 -1.11025624e-01 1.11745358e+00 -8.75247896e-01
2.81659216e-01 1.50420034e+00 7.36788392e-01 9.10201520e-02
-1.37284327e+00 -5.86191058e-01 8.47468197e-01 3.64386767e-01
-1.54179060e+00 -9.07344520e-02 7.62829602e-01 -7.56042957e-01
8.41116905e-01 2.98193336e-01 6.43622279e-01 5.92754543e-01
2.04473004e-01 8.58051181e-01 1.05518377e+00 -2.41428196e-01
1.58955287e-02 2.43065909e-01 5.14527261e-01 8.94193232e-01
1.88160211e-01 -3.68076026e-01 -1.91603214e-01 -2.03143314e-01
7.83200085e-01 4.98862445e-01 5.21800620e-03 -5.20210803e-01
-1.00744927e+00 7.69535720e-01 9.59811628e-01 3.02561462e-01
-5.69564104e-01 3.93483788e-01 5.06066203e-01 4.13657762e-02
2.03893721e-01 3.92329805e-02 -8.53231668e-01 5.19416153e-01
-8.42055202e-01 1.57753974e-02 5.24496436e-01 8.53123009e-01
1.17756629e+00 -2.01434404e-01 -5.06122291e-01 1.18119621e+00
4.14413095e-01 7.13053644e-02 4.99518692e-01 -6.37886703e-01
2.96663314e-01 1.19576108e+00 -2.59399831e-01 -9.82144833e-01
-4.46598202e-01 -6.25827491e-01 -1.16654420e+00 4.49421018e-01
2.16885373e-01 4.10384089e-01 -1.08105874e+00 1.44830668e+00
3.02203149e-01 -5.08980043e-02 -2.53991157e-01 8.10526907e-01
1.08875573e+00 4.34152573e-01 5.18922865e-01 7.30731860e-02
1.78232563e+00 -1.24334300e+00 -3.63173187e-01 -6.11562952e-02
3.82488728e-01 -7.82363772e-01 8.23974371e-01 2.35881552e-01
-9.43888247e-01 -1.07960820e+00 -9.98526096e-01 -2.86508024e-01
-7.95650780e-01 7.99341679e-01 6.88049436e-01 7.24281818e-02
-9.90031898e-01 4.41746444e-01 -6.72989368e-01 -1.14392206e-01
5.94545662e-01 5.05225599e-01 -6.10278130e-01 1.25547841e-01
-7.46872067e-01 5.64744473e-01 5.51404417e-01 1.02357924e-01
-5.46654582e-01 -7.00646222e-01 -8.34568441e-01 3.19248438e-01
1.50494680e-01 -4.60568249e-01 8.10573876e-01 -1.10603333e+00
-1.06884789e+00 6.43205285e-01 -1.40544653e-01 -2.26504967e-01
3.30570340e-01 -1.34605259e-01 -6.80630133e-02 2.45089456e-01
4.67961617e-02 8.78240347e-01 9.74766910e-01 -1.33625114e+00
-8.18238616e-01 -4.66716826e-01 9.23099592e-02 2.81357300e-02
-4.15633172e-01 -1.28167108e-01 -7.88437903e-01 -7.90438950e-01
3.48113120e-01 -7.62500226e-01 -2.46509790e-01 2.97701329e-01
-3.49622637e-01 -4.21331853e-01 9.77186084e-01 -5.88219166e-01
1.21218109e+00 -2.16188216e+00 1.47796810e-01 -9.81918885e-04
4.85788465e-01 2.39728898e-01 -1.13350563e-01 3.19353938e-01
4.68950197e-02 5.83188348e-02 7.01861754e-02 -3.82473588e-01
3.12682129e-02 -7.73374811e-02 -3.70555297e-02 2.09483340e-01
4.78074640e-01 9.97657895e-01 -2.61835933e-01 -7.73796082e-01
4.15935248e-01 6.93982124e-01 -5.28395653e-01 8.62021148e-02
9.41777751e-02 -1.50364935e-01 -4.91173178e-01 7.46943653e-01
9.53227460e-01 -4.16573077e-01 -1.90895975e-01 -1.04157293e+00
-1.17059574e-01 -3.95581335e-01 -1.33385718e+00 1.38253880e+00
-5.67071557e-01 2.93163508e-01 1.23687319e-01 -8.30483019e-01
8.33972275e-01 -2.18002722e-02 3.01705807e-01 -3.22691590e-01
2.04925537e-01 -3.08086053e-02 -1.39113188e-01 -1.87455714e-01
2.34942794e-01 1.40816376e-01 -1.67085379e-01 -6.16676882e-02
4.30482268e-01 7.90435523e-02 1.14907995e-01 9.53180343e-02
9.05226886e-01 1.14752226e-01 4.71211106e-01 -3.00516099e-01
8.24836969e-01 -2.04666376e-01 5.46969533e-01 5.64200640e-01
-3.55727792e-01 5.87229848e-01 4.26969051e-01 -5.53000569e-01
-7.13481128e-01 -9.20673549e-01 -1.30813539e-01 1.41008210e+00
3.13798010e-01 -4.59207237e-01 -5.89225531e-01 -9.46614683e-01
3.56449485e-01 -4.73566800e-02 -1.01851046e+00 3.01481783e-02
-3.77214521e-01 -5.39028466e-01 4.84044045e-01 1.04690206e+00
7.02230096e-01 -7.93895066e-01 -4.37250197e-01 1.94440395e-01
-1.31956488e-01 -1.37509894e+00 -4.11385685e-01 5.50669789e-01
-7.99950540e-01 -1.23540843e+00 -8.64157796e-01 -9.81900096e-01
7.14743972e-01 4.26941723e-01 9.18783069e-01 1.09444723e-01
-7.50636101e-01 2.64951795e-01 -3.45780283e-01 -2.44557306e-01
1.10343829e-01 -1.35354981e-01 -2.52986878e-01 3.56934488e-01
1.48906589e-01 -3.53613645e-01 -7.95124054e-01 3.49011898e-01
-8.34633052e-01 5.78914136e-02 1.03836417e+00 8.52482617e-01
8.70327890e-01 1.34432390e-01 1.44985810e-01 -5.75295389e-01
1.71438754e-01 -1.62156865e-01 -6.27418160e-01 5.82189381e-01
-1.90771967e-01 1.95105731e-01 8.44451964e-01 -5.97490489e-01
-1.09115517e+00 3.88991326e-01 3.33693102e-02 -7.94793844e-01
-4.11145180e-01 -2.34701522e-02 -1.78642273e-01 -3.63488138e-01
4.23366964e-01 1.48629084e-01 -4.52706933e-01 -7.03432977e-01
5.18244028e-01 5.58096170e-01 4.02486920e-01 -5.28920710e-01
6.79058850e-01 5.08888125e-01 8.86834413e-03 -8.35606396e-01
-5.33530593e-01 -6.86887324e-01 -1.11703753e+00 -6.74333572e-02
1.15767860e+00 -1.11870980e+00 -8.79189491e-01 5.42542756e-01
-1.13491225e+00 -1.35789618e-01 7.00681806e-02 2.26865828e-01
-8.12742114e-02 3.71032774e-01 -8.00158620e-01 -7.20977187e-01
-4.51870829e-01 -1.27396810e+00 1.52901185e+00 3.10843349e-01
2.19938979e-01 -7.26116121e-01 -4.42001671e-01 1.46723986e-01
3.22872579e-01 1.42788813e-01 9.50601995e-01 -4.25991178e-01
-5.07632494e-01 -3.89601558e-01 -7.76989281e-01 5.65265298e-01
2.49282029e-02 2.50542939e-01 -1.00248408e+00 -2.01055571e-01
-4.83066559e-01 -4.67271715e-01 1.16046572e+00 3.72929662e-01
1.47531259e+00 -2.06587926e-01 -4.74926084e-01 5.87812126e-01
1.66831613e+00 -1.57420382e-01 2.83398986e-01 4.96206284e-02
1.08696187e+00 8.59982431e-01 4.29145873e-01 4.86629605e-01
5.37104189e-01 7.82666743e-01 4.88776326e-01 -4.21066940e-01
-2.61367768e-01 -3.76833491e-02 1.73262835e-01 5.21652460e-01
-1.74948141e-01 3.15144002e-01 -5.46384931e-01 3.65704775e-01
-1.92390001e+00 -6.56445384e-01 -4.01733279e-01 1.88399899e+00
4.63103265e-01 -5.34932464e-02 1.53989226e-01 6.56969324e-02
6.70170724e-01 -5.84014468e-02 -3.54994893e-01 -2.71833450e-01
-1.55951604e-01 1.86540976e-01 6.68799877e-01 2.12128565e-01
-1.54421628e+00 1.07191443e+00 5.60938358e+00 1.17498946e+00
-1.13678122e+00 1.24973163e-01 6.90528035e-01 2.40658551e-01
1.97956353e-01 -1.32696539e-01 -9.91893351e-01 2.35916808e-01
8.63081589e-02 5.43290377e-01 -5.79958633e-02 1.03495777e+00
-2.92795628e-01 -1.31289616e-01 -8.07293117e-01 1.02705038e+00
-4.82853167e-02 -1.11713064e+00 2.18864307e-01 9.80222672e-02
7.04364777e-01 -1.52491286e-01 -2.12458208e-01 4.57154900e-01
2.96343237e-01 -8.09469938e-01 9.64100063e-01 4.13516462e-01
5.31558871e-01 -6.65908933e-01 7.97210813e-01 2.98393160e-01
-1.99614882e+00 -3.83991063e-01 -6.94640040e-01 1.96063921e-01
-3.73547047e-01 4.01386976e-01 -4.50094670e-01 5.63383639e-01
9.30959284e-01 4.55564857e-01 -9.26437199e-01 1.04691160e+00
-9.59843248e-02 1.94340929e-01 -6.40300736e-02 1.31938204e-01
2.56971061e-01 1.40738770e-01 -1.91532314e-01 1.51711214e+00
-3.72508019e-02 3.70489024e-02 4.32362586e-01 6.78766429e-01
8.33715349e-02 3.32783729e-01 -1.27340674e-01 7.90502578e-02
1.82977617e-01 1.74580073e+00 -9.46828783e-01 -4.64568049e-01
-6.16500258e-01 1.26609707e+00 6.52837098e-01 4.43360597e-01
-7.60706007e-01 -3.95559102e-01 8.20852637e-01 -5.77975698e-02
8.12559307e-01 -9.95693356e-02 -3.05734485e-01 -1.29350054e+00
-8.53331760e-02 -5.60021341e-01 4.81083661e-01 -4.74851459e-01
-1.46596527e+00 8.06798697e-01 -7.34401122e-02 -1.21032178e+00
1.81911170e-01 -9.11283076e-01 -5.24672270e-01 8.60654950e-01
-1.38079107e+00 -1.74597311e+00 -6.11196756e-01 8.37236524e-01
5.83187163e-01 6.52571768e-02 9.10078943e-01 1.86152264e-01
-5.01610398e-01 8.07383835e-01 -3.27666759e-01 2.61345625e-01
4.89385277e-01 -1.22496855e+00 -3.79695185e-02 6.30624175e-01
1.31414473e-01 4.94165510e-01 2.17233956e-01 -5.20794630e-01
-1.13064826e+00 -1.25679481e+00 5.94276547e-01 -6.37877360e-02
4.57663894e-01 -4.81287330e-01 -9.60060716e-01 3.69524240e-01
-1.28650784e-01 4.64860886e-01 6.01216495e-01 2.07038820e-01
-8.71754706e-01 -4.04466987e-01 -1.04025555e+00 2.77003884e-01
7.20676064e-01 -6.57033086e-01 -3.82200450e-01 -2.56401822e-02
5.77462316e-01 1.08866513e-01 -9.48358238e-01 5.72266996e-01
6.99211121e-01 -9.32516575e-01 1.09828377e+00 -4.56344783e-01
3.58766675e-01 -5.39838552e-01 -4.22836840e-01 -9.77170050e-01
-8.12370181e-01 2.27810934e-01 -1.53831616e-01 1.15158081e+00
8.86661261e-02 -3.13407540e-01 6.75729513e-01 3.96687239e-01
-7.79362917e-02 -9.96264517e-01 -4.81543839e-01 -5.93417406e-01
-1.48161024e-01 -3.73708040e-01 4.11002785e-01 5.88964760e-01
-2.42225558e-01 1.82191968e-01 -1.78291038e-01 3.24986011e-01
6.50367320e-01 3.20970386e-01 6.81577563e-01 -1.41301978e+00
-4.61143732e-01 -7.97961473e-01 -8.93391728e-01 -9.77066100e-01
4.33407142e-04 -7.63251066e-01 8.17338601e-02 -1.71000004e+00
4.95562702e-01 -4.14869606e-01 -6.67035401e-01 8.04168999e-01
-3.33189696e-01 7.80139744e-01 4.39091384e-01 2.92771965e-01
-6.91522241e-01 4.15055156e-01 1.13362706e+00 -2.41403788e-01
1.87373646e-02 -1.18508726e-01 -6.29291236e-01 9.34495687e-01
4.66338634e-01 8.68498255e-03 -1.99484169e-01 -1.18004382e-01
-3.24842721e-01 -2.03321129e-01 6.35311604e-01 -1.15260446e+00
3.02916825e-01 1.17740341e-01 1.01250720e+00 -9.54365611e-01
4.22853976e-01 -1.02193856e+00 -8.67138617e-03 4.09716368e-01
-2.16329992e-01 -2.36621335e-01 3.76513451e-01 4.80002910e-01
-5.69512486e-01 8.80593210e-02 8.88518095e-01 -1.84001252e-01
-8.77519786e-01 4.57418293e-01 -2.29045004e-01 -4.11096781e-01
1.07826447e+00 -1.46135762e-01 -1.19944796e-01 -9.68569610e-03
-8.70500863e-01 1.46188349e-01 2.85020620e-01 4.25899535e-01
5.22128642e-01 -1.44174528e+00 -4.35476750e-01 2.09418640e-01
3.94420147e-01 -1.01936050e-01 6.02845669e-01 8.40894580e-01
-3.52440178e-01 3.82467061e-01 -3.79390121e-01 -7.42344320e-01
-1.65948629e+00 6.11270368e-01 2.57942021e-01 -3.22936028e-01
-2.73059428e-01 1.30989444e+00 1.07564318e+00 -1.83641881e-01
5.06798208e-01 -4.90203857e-01 -4.64558929e-01 4.27265167e-01
5.89603722e-01 2.79807866e-01 5.59468940e-02 -9.45786953e-01
-5.45476735e-01 9.18882906e-01 -2.63526380e-01 3.31886828e-01
1.14790881e+00 -5.17132925e-03 -1.78622738e-01 2.14842364e-01
1.64689124e+00 -1.30575046e-01 -1.37765944e+00 -4.05802250e-01
-2.69373924e-01 -2.41741061e-01 9.09985900e-02 -7.96525717e-01
-1.09643257e+00 1.14850020e+00 7.15537131e-01 3.50807719e-02
1.32781506e+00 2.74450749e-01 3.82549763e-01 1.38167977e-01
3.40942502e-01 -9.20126081e-01 2.49536753e-01 4.04054046e-01
1.04072738e+00 -1.18495452e+00 2.96635479e-02 -6.91769540e-01
-7.46044397e-01 1.29997003e+00 8.59369457e-01 -2.58259505e-01
8.66078317e-01 1.22658990e-01 -4.57828306e-03 -3.88484299e-02
-6.20741725e-01 -6.00604713e-01 9.30403352e-01 2.60299861e-01
4.79714096e-01 2.21190199e-01 -2.64846742e-01 8.79480243e-01
2.57241398e-01 -2.22033724e-01 -2.64070272e-01 7.13488817e-01
-5.13500631e-01 -1.10374963e+00 -4.53895062e-01 5.17621160e-01
-5.42354167e-01 -9.37800780e-02 -5.01900196e-01 7.12545216e-01
7.29808271e-01 7.74605691e-01 -3.08492832e-04 -5.90149760e-01
2.56604105e-01 -1.07871937e-02 4.96608377e-01 -5.34530818e-01
-7.86883771e-01 4.96234298e-01 -2.39458770e-01 -7.27788150e-01
-1.78368375e-01 -3.56648654e-01 -1.24778259e+00 1.04343884e-01
-3.23245466e-01 -1.94207449e-02 6.74189091e-01 8.26861203e-01
3.81730169e-01 7.84087658e-01 5.36165178e-01 -1.22876775e+00
-3.34078372e-01 -1.01996017e+00 -6.84203029e-01 5.19639730e-01
3.47572803e-01 -8.05682063e-01 -1.48012072e-01 -5.79324216e-02] | [9.62134838104248, 1.9685039520263672] |
27f4cb7e-3e06-4b4b-83fa-043d2351d85f | everlight-indoor-outdoor-editable-hdr | 2304.13207 | null | https://arxiv.org/abs/2304.13207v1 | https://arxiv.org/pdf/2304.13207v1.pdf | EverLight: Indoor-Outdoor Editable HDR Lighting Estimation | Because of the diversity in lighting environments, existing illumination estimation techniques have been designed explicitly on indoor or outdoor environments. Methods have focused specifically on capturing accurate energy (e.g., through parametric lighting models), which emphasizes shading and strong cast shadows; or producing plausible texture (e.g., with GANs), which prioritizes plausible reflections. Approaches which provide editable lighting capabilities have been proposed, but these tend to be with simplified lighting models, offering limited realism. In this work, we propose to bridge the gap between these recent trends in the literature, and propose a method which combines a parametric light model with 360{\deg} panoramas, ready to use as HDRI in rendering engines. We leverage recent advances in GAN-based LDR panorama extrapolation from a regular image, which we extend to HDR using parametric spherical gaussians. To achieve this, we introduce a novel lighting co-modulation method that injects lighting-related features throughout the generator, tightly coupling the original or edited scene illumination within the panorama generation process. In our representation, users can easily edit light direction, intensity, number, etc. to impact shading while providing rich, complex reflections while seamlessly blending with the edits. Furthermore, our method encompasses indoor and outdoor environments, demonstrating state-of-the-art results even when compared to domain-specific methods. | ['Jean-François Lalonde', 'Jonathan Eisenmann', 'Yannick Hold-Geoffroy', 'Mohammad Reza Karimi Dastjerdi'] | 2023-04-26 | null | null | null | null | ['lighting-estimation'] | ['computer-vision'] | [ 5.50499558e-01 -2.13372171e-01 3.95654738e-01 -2.46092856e-01
-3.22231829e-01 -6.33178055e-01 8.68138552e-01 -4.43753600e-01
2.46808097e-01 6.70886338e-01 3.05888683e-01 -1.44575849e-01
2.32726589e-01 -1.06199026e+00 -4.32315677e-01 -6.08033597e-01
2.92539626e-01 2.90672183e-02 5.74908443e-02 -4.27264869e-01
-2.25451756e-02 6.74233377e-01 -1.94374979e+00 8.01326260e-02
1.13800979e+00 7.39318609e-01 1.59949988e-01 1.05291700e+00
-3.09419632e-01 8.06228220e-01 -8.28188419e-01 -3.51765364e-01
4.37279135e-01 -7.86922276e-01 -4.95958626e-02 1.08388789e-01
7.29453921e-01 -4.74846214e-01 -5.39192632e-02 5.31338096e-01
5.96336305e-01 5.57900816e-02 3.37263107e-01 -9.86409307e-01
-6.22942567e-01 -9.80126932e-02 -5.73191345e-01 -6.24754965e-01
6.66277170e-01 2.54107594e-01 6.88950598e-01 -7.61912465e-01
5.25631547e-01 1.14290905e+00 7.28182733e-01 5.47062337e-01
-1.52761197e+00 -4.78057593e-01 -5.65438718e-02 -5.38846999e-02
-1.14440572e+00 -4.59016830e-01 1.11648500e+00 -2.23227054e-01
6.66576743e-01 9.88816261e-01 1.07576215e+00 1.09269238e+00
1.86268777e-01 3.17604512e-01 1.82192147e+00 -8.00948799e-01
2.07391962e-01 1.30635351e-01 -4.91519153e-01 4.34552789e-01
-2.16612473e-01 2.97942907e-01 -5.43240726e-01 -9.08169001e-02
1.00239468e+00 -1.71746179e-01 -6.23147428e-01 -3.47537428e-01
-8.86904180e-01 5.68263948e-01 3.21214467e-01 -3.44433710e-02
-1.59978107e-01 2.54320681e-01 -9.66196656e-02 1.01005189e-01
6.98017955e-01 3.55148643e-01 -1.52576372e-01 -1.50597617e-01
-1.23766720e+00 3.03805470e-01 7.96698868e-01 9.16642547e-01
1.09674776e+00 1.62433431e-01 -4.94327962e-01 1.00248599e+00
3.87295514e-01 7.25760043e-01 -1.50214195e-01 -1.12162578e+00
-5.42796105e-02 2.90140420e-01 1.45884007e-01 -6.36297584e-01
-2.37273604e-01 -1.41157404e-01 -6.61195457e-01 8.67860615e-01
1.68202281e-01 -1.34050809e-02 -1.11387992e+00 1.48477435e+00
4.43040758e-01 -1.96391717e-03 -2.38113374e-01 7.51547992e-01
6.52543902e-01 6.95969641e-01 -2.46476173e-01 -1.65189996e-01
1.20018816e+00 -1.10969532e+00 -9.12908018e-01 1.21600188e-01
3.85025283e-03 -1.23983359e+00 1.56768799e+00 7.65094578e-01
-1.20539331e+00 -4.74322110e-01 -9.64696467e-01 -4.66071367e-01
-2.85315812e-01 6.08054698e-02 6.76082432e-01 1.24433565e+00
-1.45996857e+00 5.49979150e-01 -5.45883596e-01 -2.70994872e-01
5.52392229e-02 -7.08192289e-02 4.23517287e-01 -2.97254086e-01
-9.11580324e-01 9.55191553e-01 -2.87091017e-01 -7.91490898e-02
-6.64569318e-01 -1.06228077e+00 -7.76213288e-01 -7.38004223e-02
9.71525460e-02 -8.53716373e-01 1.06360364e+00 -1.13455939e+00
-2.38100815e+00 5.82654595e-01 -1.95784688e-01 -2.35275505e-03
8.08952510e-01 -2.84905434e-01 -4.15791661e-01 -1.57180458e-01
-4.00506943e-01 7.24886954e-01 1.01060212e+00 -1.83453667e+00
-2.34301686e-01 4.32853848e-02 3.93207133e-01 3.24592888e-01
-2.00400740e-01 2.21695472e-02 -4.27057892e-01 -8.02330136e-01
-2.52262533e-01 -9.16500747e-01 -1.86061725e-01 2.77193129e-01
-4.56448704e-01 3.09180230e-01 1.14442301e+00 -6.85819685e-01
1.25582528e+00 -1.96060133e+00 -6.04368374e-03 2.55964518e-01
9.54279304e-03 -4.39614542e-02 1.48963809e-01 6.34948194e-01
5.02380282e-02 -1.75548181e-01 -4.03726727e-01 -1.01547146e+00
2.95442432e-01 3.78347933e-01 -5.25622547e-01 3.34073544e-01
-2.38859028e-01 6.54944897e-01 -6.80812955e-01 -2.52807289e-01
7.97157228e-01 1.28084779e+00 -6.14041567e-01 3.82875919e-01
-2.87597060e-01 7.93416202e-01 1.81515098e-01 5.24688184e-01
9.75085557e-01 2.62067497e-01 5.82556538e-02 -3.41370434e-01
-6.92982912e-01 2.34812796e-01 -1.23215175e+00 1.77013123e+00
-1.19390464e+00 7.35450566e-01 2.42989048e-01 1.01620875e-01
9.50308263e-01 7.39401625e-03 2.33242229e-01 -6.37708962e-01
-1.86637685e-01 1.84930190e-01 -6.21096790e-01 1.06186852e-01
8.69409323e-01 -9.22320783e-02 4.07949507e-01 3.23506296e-01
-3.98035735e-01 -9.19806421e-01 -4.49841358e-02 -8.06034952e-02
7.73719847e-01 8.62358272e-01 1.30905181e-01 -2.25372210e-01
2.35924423e-01 -3.41991723e-01 9.37978700e-02 5.72132170e-01
4.97355580e-01 1.27583694e+00 1.85131244e-02 -1.33082181e-01
-1.13338327e+00 -1.14525390e+00 -3.12338382e-01 9.70737875e-01
-7.80096650e-02 -5.72629809e-01 -1.02330196e+00 -9.90094021e-02
-3.70653689e-01 1.17808592e+00 -5.52711844e-01 3.81020457e-01
-7.01227307e-01 -5.96799970e-01 1.81862444e-01 4.82757799e-02
6.01424396e-01 -8.60871971e-01 -1.04217339e+00 4.50983085e-02
2.32696593e-01 -8.06469560e-01 -3.68942410e-01 9.60643217e-02
-4.63202357e-01 -7.27324843e-01 -1.02357745e+00 -5.16283065e-02
5.95071971e-01 1.95651352e-01 1.52481902e+00 1.79447457e-02
-6.39419258e-01 8.34118187e-01 -3.86833519e-01 -3.91864657e-01
-4.27397132e-01 -3.86593640e-01 -4.00842458e-01 7.07759336e-02
-4.42659259e-01 -7.76398003e-01 -1.00872374e+00 3.40113729e-01
-1.18541467e+00 6.46313608e-01 1.26545653e-01 3.94958347e-01
5.34560919e-01 -2.26852491e-01 -9.09207836e-02 -9.45542574e-01
4.33600038e-01 -4.97804619e-02 -7.30094612e-01 2.02233970e-01
-5.87880492e-01 -2.77923107e-01 9.05216932e-01 -1.76677600e-01
-1.52320099e+00 -2.73638934e-01 -2.36368135e-01 -3.77363622e-01
-2.64515489e-01 -2.31513143e-01 -3.43104452e-01 -4.39205855e-01
5.56720614e-01 2.20567286e-01 -5.37433445e-01 -4.64311302e-01
7.77196646e-01 3.87227088e-01 3.83473396e-01 -8.56236994e-01
9.68165755e-01 7.74778783e-01 1.73927024e-01 -9.48272049e-01
-5.69781780e-01 -3.52640264e-02 -2.66043693e-01 -5.03239095e-01
6.67352557e-01 -6.11060202e-01 -3.40762317e-01 6.04697168e-01
-1.07863748e+00 -9.22084510e-01 -8.64923477e-01 8.96619409e-02
-7.01604962e-01 3.29155982e-01 -5.03036141e-01 -9.77563500e-01
-1.23135358e-01 -1.15144074e+00 1.32103777e+00 3.96447301e-01
-1.29727170e-01 -1.02439463e+00 5.03282547e-01 4.30608429e-02
9.70298052e-01 6.40320003e-01 6.86392248e-01 6.62721217e-01
-8.23704958e-01 3.67473006e-01 -1.34188145e-01 3.87210459e-01
4.07906175e-01 4.77943897e-01 -1.50970674e+00 -1.00849308e-01
8.64248071e-03 2.53115781e-02 5.81967890e-01 4.85762030e-01
1.25787640e+00 -2.33386591e-01 3.56948636e-02 1.02004600e+00
1.63372755e+00 -2.22970434e-02 1.18946087e+00 3.03632975e-01
9.91416335e-01 7.35857189e-01 3.05121005e-01 6.45936608e-01
4.71495599e-01 1.07124448e+00 3.91642243e-01 -7.11509526e-01
-7.62504280e-01 -2.15131342e-01 2.73414940e-01 6.80361331e-01
-3.76148045e-01 -4.30289835e-01 -3.78026038e-01 -1.42336721e-02
-1.42855179e+00 -7.31539547e-01 -3.67691159e-01 2.43687344e+00
1.07312071e+00 -2.70254195e-01 -1.48836508e-01 5.89881167e-02
2.74542749e-01 5.22749901e-01 -2.30285972e-01 -8.37725401e-01
-3.87378186e-01 6.39936507e-01 5.38488984e-01 7.33938754e-01
-5.04777312e-01 6.59420013e-01 6.24711847e+00 8.52167964e-01
-1.19946587e+00 1.69516087e-01 6.22258484e-01 -1.28448993e-01
-1.12519825e+00 2.11460844e-01 -5.01462877e-01 3.93177956e-01
6.07250273e-01 2.37248555e-01 9.96896327e-01 5.22604525e-01
5.79130650e-01 -3.98649901e-01 -8.80509377e-01 9.29582715e-01
2.63830155e-01 -1.01805246e+00 -1.74076736e-01 1.01699457e-01
1.06950092e+00 -3.95859450e-01 2.16785073e-01 -1.86510626e-02
4.82484847e-01 -1.13781178e+00 7.00614214e-01 7.44731724e-01
1.30093241e+00 -5.62349796e-01 -1.79626971e-01 -5.00100330e-02
-1.13755226e+00 4.43078816e-01 -9.47582275e-02 3.85951735e-02
4.37777311e-01 9.93712604e-01 -5.16919434e-01 6.92976654e-01
6.14581048e-01 3.11102033e-01 -5.29950321e-01 8.79586875e-01
-5.58010459e-01 5.04769087e-01 -5.51076889e-01 3.31092209e-01
-1.97138786e-01 -6.89206183e-01 4.33355182e-01 1.38688624e+00
5.72914302e-01 -1.27572939e-01 5.28899804e-02 1.10708320e+00
1.52541682e-01 1.42963573e-01 -4.92909580e-01 7.20436811e-01
1.57272294e-01 1.55449986e+00 -5.92073083e-01 -1.10438295e-01
-3.27428520e-01 1.39674091e+00 -1.21185653e-01 6.83171570e-01
-1.05554998e+00 -3.13425332e-01 6.54438853e-01 4.18688744e-01
7.15532675e-02 -4.49541867e-01 -4.93175000e-01 -9.46956098e-01
-5.46880476e-02 -8.35569441e-01 -2.01526552e-01 -1.21386278e+00
-9.46070015e-01 5.35907149e-01 -3.49034905e-03 -9.81902122e-01
3.17060761e-02 -3.67521793e-01 -7.55666316e-01 1.02340806e+00
-1.96125376e+00 -1.61034763e+00 -7.26377368e-01 6.86234951e-01
6.40105665e-01 5.45899689e-01 9.79047835e-01 4.04079437e-01
-1.74687803e-01 3.35431099e-01 2.02305585e-01 -6.51252806e-01
8.33375990e-01 -1.52166414e+00 4.16246444e-01 8.62869263e-01
6.49602637e-02 5.07164240e-01 9.48721707e-01 -3.71442109e-01
-1.49513924e+00 -9.22562063e-01 4.04866755e-01 -4.19420779e-01
1.23548739e-01 -6.40501380e-01 -6.22565329e-01 3.88937205e-01
6.25007451e-01 -1.01974741e-01 5.17771304e-01 -1.21612497e-01
-2.82965899e-01 -2.51772076e-01 -1.34839821e+00 1.02182305e+00
1.05559266e+00 -3.71992439e-01 5.28317094e-02 2.44732931e-01
6.50077343e-01 -7.39823222e-01 -7.08266497e-01 3.76613855e-01
6.35724127e-01 -1.71571493e+00 1.11060500e+00 3.75881135e-01
3.42747271e-01 -7.41001964e-01 -1.93886042e-01 -1.57845664e+00
-1.12412564e-01 -1.14091790e+00 -1.87206537e-01 1.55271316e+00
-5.48124909e-02 -6.83492661e-01 4.26433474e-01 6.42091334e-01
-4.28964168e-01 -5.76977432e-01 -6.98621631e-01 -4.34947371e-01
-3.18471223e-01 -5.74102342e-01 8.39206219e-01 7.68437922e-01
-6.74980879e-01 -2.42854282e-01 -6.46604359e-01 -1.85704380e-02
7.24313855e-01 2.67960310e-01 1.17249560e+00 -8.36546421e-01
-5.57232320e-01 -4.13306206e-01 2.48768017e-01 -1.11477089e+00
-2.21686825e-01 -2.88276553e-01 1.52399883e-01 -1.69169426e+00
-1.32610455e-01 -8.53591919e-01 3.04320604e-01 1.92449883e-01
-1.54997870e-01 6.96019411e-01 2.36810833e-01 5.93470968e-02
-8.09551477e-02 6.63494349e-01 1.29785275e+00 2.34484300e-01
-3.87793899e-01 -2.14306146e-01 -3.67178142e-01 6.78429842e-01
7.11261272e-01 -5.36913984e-02 -4.26106334e-01 -5.09987772e-01
3.43904972e-01 -4.44494218e-01 5.50137877e-01 -1.21168113e+00
-1.71045527e-01 -1.68957412e-01 4.64537770e-01 -3.21432739e-01
6.65949523e-01 -8.43660712e-01 8.00097406e-01 -1.79598629e-02
-5.19543476e-02 -1.38862848e-01 1.19090654e-01 2.50050128e-01
2.11658716e-01 4.29973425e-03 8.70598972e-01 -7.80378580e-02
-2.77905017e-01 9.43652317e-02 -3.17887701e-02 -1.47598431e-01
7.87249506e-01 -4.81001616e-01 -4.84100133e-01 -6.26567900e-01
-2.65612960e-01 -2.98848838e-01 1.28051627e+00 1.32669806e-01
3.32845181e-01 -1.26952827e+00 -5.35031021e-01 2.49779671e-01
-1.10735288e-02 5.21771461e-02 3.28512281e-01 6.53038204e-01
-8.48785639e-01 -1.61882327e-03 1.20879963e-01 -5.35769165e-01
-1.12321103e+00 2.00150803e-01 2.83996046e-01 -2.18236297e-01
-6.14801705e-01 4.65953588e-01 4.30237889e-01 -4.51274633e-01
-1.30946279e-01 -3.54394764e-01 5.75590320e-02 -6.40129894e-02
3.68387103e-01 5.75117946e-01 2.09884971e-01 -3.54491740e-01
-3.56143750e-02 8.93737912e-01 6.20508969e-01 -2.03356609e-01
1.12323225e+00 -3.80201727e-01 -2.65756249e-01 4.63087261e-01
8.30319762e-01 9.59318638e-01 -1.68766022e+00 1.73796356e-01
-7.92804718e-01 -8.75009954e-01 1.12727277e-01 -1.03596210e+00
-9.71617162e-01 8.22681189e-01 6.83464766e-01 3.31812531e-01
1.50559795e+00 -3.52738619e-01 7.46612251e-01 -4.21930730e-01
6.47523999e-01 -9.86431360e-01 -2.36101657e-01 2.65677094e-01
8.88973773e-01 -7.02745795e-01 9.88278463e-02 -5.58448136e-01
-4.15620357e-01 1.32981157e+00 2.72030890e-01 8.31891745e-02
3.13871950e-01 6.86333060e-01 2.87221611e-01 -4.39848714e-02
-2.07469508e-01 -1.47404149e-01 3.22438508e-01 6.93670928e-01
5.58549404e-01 1.34772927e-01 -1.81469128e-01 -1.29277959e-01
-3.64432037e-01 1.91704053e-02 6.06227100e-01 8.34633172e-01
-1.69833884e-01 -1.33090925e+00 -7.51581728e-01 9.45817828e-02
-2.28953093e-01 -4.88738298e-01 -8.49449188e-02 4.21222955e-01
3.71147484e-01 9.61385846e-01 -2.85982639e-02 1.03310995e-01
3.43473226e-01 -2.49559909e-01 7.43143916e-01 -4.17484134e-01
-6.60142481e-01 2.60304034e-01 1.25845611e-01 -7.54725158e-01
-5.13063014e-01 -4.14676636e-01 -5.84045589e-01 -3.20529729e-01
-1.96748748e-01 -2.97504336e-01 9.02345777e-01 4.28677380e-01
2.42125496e-01 7.05797911e-01 8.23321104e-01 -1.44664860e+00
2.02802584e-01 -6.07038915e-01 -5.42633712e-01 1.31992117e-01
2.98947722e-01 -5.16594350e-01 -4.31787014e-01 1.64251685e-01] | [9.748034477233887, -3.056702136993408] |
2c49b5e0-21a0-4d4f-af5e-0b2d238b82fd | multi-armed-bandit-learning-on-a-graph | 2209.09419 | null | https://arxiv.org/abs/2209.09419v4 | https://arxiv.org/pdf/2209.09419v4.pdf | Multi-armed Bandit Learning on a Graph | The multi-armed bandit(MAB) problem is a simple yet powerful framework that has been extensively studied in the context of decision-making under uncertainty. In many real-world applications, such as robotic applications, selecting an arm corresponds to a physical action that constrains the choices of the next available arms (actions). Motivated by this, we study an extension of MAB called the graph bandit, where an agent travels over a graph to maximize the reward collected from different nodes. The graph defines the agent's freedom in selecting the next available nodes at each step. We assume the graph structure is fully available, but the reward distributions are unknown. Built upon an offline graph-based planning algorithm and the principle of optimism, we design a learning algorithm, G-UCB, that balances long-term exploration-exploitation using the principle of optimism. We show that our proposed algorithm achieves $O(\sqrt{|S|T\log(T)}+D|S|\log T)$ learning regret, where $|S|$ is the number of nodes and $D$ is the diameter of the graph, which matches the theoretical lower bound $\Omega(\sqrt{|S|T})$ up to logarithmic factors. To our knowledge, this result is among the first tight regret bounds in non-episodic, un-discounted learning problems with known deterministic transitions. Numerical experiments confirm that our algorithm outperforms several benchmarks. | ['Na Li', 'Kasper Johansson', 'Tianpeng Zhang'] | 2022-09-20 | null | null | null | null | ['decision-making-under-uncertainty', 'decision-making-under-uncertainty'] | ['medical', 'reasoning'] | [ 6.07446842e-02 5.68835318e-01 -8.24392140e-01 -9.72291231e-02
-7.31268406e-01 -7.39667892e-01 1.20627895e-01 2.87003636e-01
-6.49111927e-01 1.10001373e+00 -3.29066545e-01 -6.05572522e-01
-8.99616241e-01 -1.06801295e+00 -1.21265733e+00 -9.08784807e-01
-6.61821425e-01 9.94688392e-01 3.60750267e-03 -1.36552095e-01
2.61884779e-01 4.45529997e-01 -9.58038867e-01 -5.43760777e-01
7.15445101e-01 1.46734679e+00 1.99342698e-01 4.62649941e-01
-3.46316919e-02 4.96444136e-01 -2.07378045e-01 -5.06535172e-01
5.34705400e-01 -4.91521537e-01 -9.03636932e-01 2.38828272e-01
-4.26369578e-01 -2.17762470e-01 -6.12841010e-01 1.14023757e+00
1.54805973e-01 5.72226167e-01 2.89840072e-01 -1.16580033e+00
-1.08910285e-01 1.28324330e+00 -9.53920901e-01 1.88586697e-01
-1.57387964e-02 1.54967070e-01 1.04424798e+00 2.90696561e-01
5.14621496e-01 1.24706781e+00 -1.51581047e-02 5.16912222e-01
-1.06888914e+00 -4.17456865e-01 7.58818686e-01 2.30579767e-02
-8.52728486e-01 4.18735966e-02 8.44476402e-01 -9.35975686e-02
6.30100667e-01 4.04531270e-01 8.77853870e-01 5.92542350e-01
2.59461224e-01 1.00521481e+00 1.03056967e+00 -5.64201772e-01
8.21192801e-01 -3.52254391e-01 6.28286600e-02 7.94456482e-01
5.28878212e-01 4.59101349e-01 -5.43004036e-01 -1.45250842e-01
6.15594983e-01 2.20841959e-01 1.30268214e-02 -7.06008434e-01
-7.84455240e-01 1.11003935e+00 5.49313664e-01 -7.90339634e-02
-5.41539133e-01 7.18965352e-01 -5.88516658e-03 4.17025089e-01
1.87896505e-01 2.88339972e-01 -2.74075210e-01 -1.79993525e-01
-5.11006713e-01 1.41706228e-01 7.54799545e-01 1.06397080e+00
8.17106128e-01 -1.22083709e-01 -1.95143044e-01 3.56241494e-01
1.95309564e-01 5.84718049e-01 -7.89419189e-03 -1.08349347e+00
9.62583125e-01 3.96061659e-01 7.29736388e-01 -5.96646011e-01
-4.03184325e-01 -6.31648839e-01 -5.88707566e-01 2.32022434e-01
5.26378453e-01 -5.53030670e-01 -9.10779238e-01 1.85418093e+00
3.59856904e-01 -2.22818106e-01 -1.42260671e-01 8.91315043e-01
-2.49443546e-01 4.83651370e-01 -3.29365343e-01 -7.58152485e-01
7.81805933e-01 -8.71471703e-01 -4.99364018e-01 -7.18366265e-01
2.90772587e-01 -5.40208779e-02 6.78757250e-01 5.78434289e-01
-1.22822762e+00 2.33361825e-01 -9.45802391e-01 7.53448904e-01
-1.83983278e-02 -4.83883351e-01 1.17649341e+00 9.80366468e-01
-5.90627313e-01 7.46579349e-01 -1.09157217e+00 9.26604494e-02
4.44337726e-01 4.72797811e-01 5.47748571e-03 -3.13477755e-01
-7.69571960e-01 6.40884221e-01 4.04305935e-01 3.36262822e-01
-1.40020525e+00 1.58146825e-02 -5.22879422e-01 -6.17827848e-02
1.37348592e+00 -4.22125608e-01 1.33078396e+00 -5.75455487e-01
-1.67950749e+00 2.20196724e-01 6.72453418e-02 -8.23973536e-01
7.42259979e-01 2.29321886e-02 2.32903600e-01 -9.24033895e-02
-1.02308616e-01 1.61087796e-01 7.58130848e-01 -9.20180619e-01
-7.60969818e-01 -8.98633778e-01 5.00093997e-01 3.37913871e-01
-1.00438252e-01 -4.98485357e-01 -3.11507255e-01 -1.45664513e-01
2.58298308e-01 -1.46681428e+00 -8.51151884e-01 -4.62219298e-01
-6.04715407e-01 -1.96432233e-01 5.08266315e-02 2.67553590e-02
9.59729671e-01 -1.79990184e+00 3.80142778e-01 5.20871758e-01
-1.41417608e-01 -5.10590196e-01 8.51042941e-02 6.46125078e-01
5.62833011e-01 5.96829802e-02 -2.04815745e-01 -2.54315138e-01
3.13618749e-01 5.18722594e-01 -2.62366712e-01 6.43303931e-01
-7.26290524e-01 7.04300880e-01 -9.08915639e-01 -1.22744329e-01
2.53806245e-02 -4.60605174e-01 -2.19225466e-01 -6.93209097e-03
-7.56281197e-01 1.74700633e-01 -1.13624573e+00 5.21938145e-01
5.22796810e-01 -3.18422794e-01 5.25715113e-01 7.02885985e-01
5.77120632e-02 3.64627242e-02 -1.38405430e+00 1.74416411e+00
-3.43262970e-01 7.26804277e-03 2.46482283e-01 -1.24664044e+00
5.95976532e-01 -2.13983238e-01 5.65212369e-01 -5.87367058e-01
1.77737862e-01 2.17557102e-01 -8.63027051e-02 -5.24908863e-02
2.50279814e-01 -1.77448764e-01 -4.44840997e-01 6.93823338e-01
-2.39186674e-01 -2.01854110e-01 3.59510094e-01 2.17050001e-01
1.30292976e+00 -4.58940193e-02 1.85747862e-01 -1.88711584e-02
-2.05830887e-01 1.27800927e-01 6.38002217e-01 1.23901057e+00
-1.53148130e-01 -1.69853598e-01 8.68719697e-01 -4.94356811e-01
-4.50592846e-01 -8.82220328e-01 4.11587268e-01 1.28164315e+00
6.56977296e-01 1.46636978e-01 -6.20690882e-01 -7.63912857e-01
3.01928222e-01 8.79060805e-01 -9.14832354e-01 -1.41982967e-02
-3.55811805e-01 -8.53132129e-01 -1.93748385e-01 2.22454086e-01
3.39518607e-01 -9.98869300e-01 -8.91815603e-01 3.66974294e-01
1.41488671e-01 -6.28986359e-01 -5.31943858e-01 6.20172501e-01
-1.02953959e+00 -1.02405643e+00 -3.77811342e-01 -1.76818594e-01
7.36091137e-01 1.97218999e-01 7.34633029e-01 -3.74331802e-01
8.71834345e-03 5.82790434e-01 -2.90734172e-01 -6.07742310e-01
1.20506659e-01 1.07754126e-01 -1.17894396e-01 -1.08018689e-01
-1.77473575e-01 -4.14566338e-01 -8.23322892e-01 8.90824497e-02
-6.62775874e-01 1.78314606e-03 8.36403728e-01 6.91195250e-01
9.79674280e-01 4.32822347e-01 3.61572653e-01 -8.49134445e-01
4.55122113e-01 -4.82795745e-01 -1.30688906e+00 4.64051217e-01
-6.26472116e-01 4.42068756e-01 4.03474241e-01 -3.97450328e-01
-9.91721451e-01 2.91263044e-01 6.16430163e-01 -1.52837113e-01
3.79202992e-01 6.76968277e-01 -1.10963359e-03 -1.65255591e-01
3.18698585e-01 1.85875103e-01 -1.37338772e-01 -3.13624799e-01
4.71955687e-01 2.27045521e-01 1.29284576e-01 -8.82023335e-01
3.53231937e-01 4.10812467e-01 5.73461890e-01 -1.65768653e-01
-9.47616756e-01 4.41619270e-02 2.95289885e-02 -2.50046194e-01
2.74206758e-01 -4.63620186e-01 -1.48543072e+00 -3.95614840e-02
-6.48149192e-01 -5.84379494e-01 -4.29201931e-01 5.48465073e-01
-1.12631285e+00 9.80400518e-02 -1.81465149e-01 -1.68222427e+00
-2.26887167e-01 -9.57623720e-01 5.63573897e-01 4.78803694e-01
5.41399300e-01 -5.27147651e-01 -9.05548558e-02 5.24208724e-01
1.18720904e-01 4.74333346e-01 8.83002639e-01 -4.13746744e-01
-1.16586185e+00 -5.97163700e-02 4.24777120e-01 -1.77237347e-01
-1.61265343e-01 -7.06838012e-01 -1.44849822e-01 -6.34062827e-01
-8.18911493e-02 -3.32077891e-01 8.54176819e-01 6.88618541e-01
1.17215669e+00 -7.91899323e-01 -7.22184181e-01 2.14103207e-01
1.52795613e+00 7.30863094e-01 2.26380393e-01 5.19626915e-01
1.31548703e-01 2.64956564e-01 9.13354993e-01 9.58037615e-01
1.66164145e-01 5.06624281e-01 1.23316813e+00 6.32107079e-01
5.99623501e-01 -2.46403471e-01 2.38409907e-01 -1.91293821e-01
-1.02657877e-01 -6.04944110e-01 -6.88084543e-01 6.58407450e-01
-2.27257085e+00 -6.55424058e-01 3.65987659e-01 2.73926854e+00
7.22266793e-01 6.14084601e-01 2.83747852e-01 -1.39819324e-01
7.52504408e-01 1.29941478e-01 -1.28966570e+00 -5.77346802e-01
2.05618948e-01 1.50219142e-01 1.30291355e+00 6.47667646e-01
-7.29135990e-01 6.94314897e-01 5.43928909e+00 1.04562128e+00
-6.78200424e-01 2.68560722e-02 9.31251168e-01 -7.55258501e-01
-2.81035960e-01 4.61549871e-02 -7.17270255e-01 6.54996634e-01
8.16280723e-01 -3.52476805e-01 1.17897999e+00 1.05806398e+00
3.79047729e-02 -5.67360163e-01 -1.10360610e+00 6.22071564e-01
-5.13974726e-01 -1.25537515e+00 -4.27121788e-01 3.38637620e-01
9.52408016e-01 3.59885693e-02 1.16586484e-01 8.15456063e-02
1.07488096e+00 -8.74998987e-01 7.70486474e-01 4.17252243e-01
4.46036309e-01 -1.40742779e+00 5.55193126e-01 7.88502693e-01
-9.15136218e-01 -6.91236317e-01 -3.29214752e-01 -4.51463945e-02
2.90597171e-01 6.91214442e-01 -7.72606611e-01 7.20992804e-01
5.65909863e-01 -6.52316734e-02 2.72877425e-01 1.00620556e+00
-2.77956039e-01 3.23233843e-01 -6.94707751e-01 -5.72278261e-01
6.30380809e-01 -5.18045664e-01 5.91811359e-01 4.55419242e-01
3.16832364e-01 3.85280579e-01 4.83552963e-01 5.26687860e-01
-9.19774771e-02 -1.91460207e-01 -3.53703618e-01 -3.03762138e-01
6.63931131e-01 8.29115868e-01 -9.13979411e-01 3.78985479e-02
1.33347824e-01 6.86599195e-01 5.56434870e-01 1.70999765e-01
-8.11458886e-01 -2.46164247e-01 3.50320280e-01 -2.14799881e-01
5.57677329e-01 -3.10183942e-01 -2.73946464e-01 -4.03095245e-01
1.67499140e-01 -3.13456774e-01 7.46376872e-01 -2.64821202e-01
-1.03952777e+00 3.17191482e-01 -3.93982790e-03 -6.90181077e-01
-3.59601706e-01 -3.19133341e-01 -1.74835652e-01 4.42138970e-01
-1.33861637e+00 -7.41333902e-01 2.13304698e-01 4.84508395e-01
3.52135032e-01 -5.70801124e-02 5.38954914e-01 -6.17650867e-01
-5.85563660e-01 3.54935110e-01 4.57514226e-01 -3.95409107e-01
-1.81853622e-01 -1.30649936e+00 -8.85626208e-03 6.71025157e-01
-7.49760643e-02 1.96332619e-01 8.50896418e-01 -7.03270853e-01
-2.02766609e+00 -7.86056459e-01 4.28308919e-02 8.97753909e-02
6.42473459e-01 -1.13154918e-01 -4.57822494e-02 9.07043278e-01
-2.23960844e-03 5.23562878e-02 3.67969304e-01 1.27574340e-01
9.40713510e-02 -4.50985044e-01 -1.31611490e+00 5.32450259e-01
1.30874097e+00 1.66682318e-01 -1.17487319e-01 6.12244904e-01
8.12866986e-01 -7.17791677e-01 -7.04753935e-01 2.48000994e-01
4.37328428e-01 -8.75237942e-01 5.65530658e-01 -5.86127877e-01
-1.26298994e-01 1.70262977e-01 -1.76747173e-01 -1.39428437e+00
-1.86120734e-01 -1.33188820e+00 -4.88823742e-01 5.95062077e-01
5.96314907e-01 -7.56565511e-01 1.22587657e+00 8.59004319e-01
1.64946597e-02 -9.98844504e-01 -1.59988129e+00 -1.02641392e+00
7.10363388e-02 -2.91918844e-01 8.19715798e-01 3.10865164e-01
2.32123181e-01 -6.30113930e-02 -5.48837006e-01 1.86159939e-01
9.64606643e-01 6.58128381e-01 3.94949883e-01 -9.24742222e-01
-8.14020216e-01 -3.72432381e-01 2.50079185e-01 -1.18865049e+00
-4.84054722e-02 -4.19879347e-01 1.34566069e-01 -1.59892046e+00
4.09908235e-01 -9.33934867e-01 -5.16204298e-01 4.73212302e-01
2.83768117e-01 -6.36088312e-01 3.59137952e-01 -1.51663512e-01
-1.02394760e+00 4.93572831e-01 1.43962729e+00 -2.48902842e-01
-3.88351709e-01 6.66649342e-01 -7.83065915e-01 3.80333573e-01
8.21490467e-01 -5.30678689e-01 -5.26912451e-01 -5.30469000e-01
5.42576194e-01 9.69851315e-01 -2.55805761e-01 -5.06441832e-01
3.32115978e-01 -9.90423620e-01 -1.22472323e-01 -5.97922802e-01
4.84571576e-01 -9.70718205e-01 2.56341666e-01 8.58369231e-01
-5.56952834e-01 -1.87332109e-01 -1.29617929e-01 1.33419907e+00
4.75458860e-01 -4.40814942e-01 5.19176781e-01 -3.86899859e-01
-3.11074913e-01 6.57930255e-01 -1.33991480e-01 -2.25748628e-01
1.34268975e+00 8.69505107e-02 -4.03034389e-01 -6.97778940e-01
-6.96793795e-01 6.71176910e-01 1.26644641e-01 1.52307674e-02
3.00613701e-01 -1.00813150e+00 -2.49636948e-01 -3.62053871e-01
-2.38161862e-01 3.35889161e-01 2.97145128e-01 6.33215725e-01
8.51431414e-02 5.47092378e-01 1.92576554e-02 -2.40607530e-01
-6.41607702e-01 8.84957731e-01 3.05916935e-01 -6.16362631e-01
-2.40435094e-01 8.81060123e-01 -2.61586159e-01 1.04736395e-01
4.50117588e-01 -1.86730430e-01 2.80514568e-01 -1.43872589e-01
2.02273071e-01 7.40118325e-01 -1.23206668e-01 2.15974331e-01
-2.21477345e-01 1.27330825e-01 -1.46358132e-01 -5.62391102e-01
1.50965786e+00 -2.78568447e-01 9.19669960e-03 2.59825021e-01
4.21009541e-01 -2.09277958e-01 -1.55991554e+00 -3.38320881e-01
4.91631664e-02 -6.73222721e-01 1.06715776e-01 -9.28284228e-01
-1.17526424e+00 4.16402161e-01 4.61688489e-01 6.53059602e-01
9.40312088e-01 1.71558499e-01 5.54843843e-01 7.01647401e-01
1.31598055e+00 -1.41241682e+00 1.33302342e-02 3.65394711e-01
6.48243606e-01 -9.82774436e-01 -1.22208958e-02 -1.52876377e-01
-5.99362552e-01 7.74515808e-01 4.73432571e-01 -1.22760214e-01
3.56565177e-01 -2.49639470e-02 -7.26796746e-01 4.24552802e-03
-8.13414931e-01 -2.74711877e-01 -2.36463621e-01 3.40591878e-01
-3.53320539e-01 7.00528026e-01 -4.66771573e-01 6.35060906e-01
-2.37268195e-01 -9.81160700e-02 5.04564166e-01 1.11116314e+00
-9.56779540e-01 -1.08270597e+00 -3.76122415e-01 6.24865770e-01
-3.79320681e-01 4.31064159e-01 -1.90741688e-01 6.24029219e-01
-2.52870589e-01 1.21588242e+00 -2.76453253e-02 7.64494389e-02
1.40572011e-01 -3.07482719e-01 8.92631114e-01 -3.68520081e-01
-5.72626740e-02 3.62341702e-02 1.70968100e-01 -6.68943703e-01
-1.78925410e-01 -6.25523508e-01 -1.28399265e+00 -4.65795904e-01
-3.60943317e-01 5.66371560e-01 8.44106376e-01 1.00925040e+00
2.62117505e-01 3.40772808e-01 1.00790799e+00 -5.43130338e-01
-9.36827719e-01 -7.64689207e-01 -8.92223954e-01 -2.73702264e-01
2.48995975e-01 -8.50436032e-01 -3.49000037e-01 -7.06611633e-01] | [4.489352226257324, 3.15429949760437] |
e91c9335-9bb0-4aeb-8580-655e558e5463 | methods-for-spoken-language-identification | null | null | http://cs229.stanford.edu/proj2017/final-reports/5239784.pdf | http://cs229.stanford.edu/proj2017/final-reports/5239784.pdf | Methods for Spoken Language Identification | In this paper, we explore several machine learning techniques for classifying spoken language. In particular, we construct algorithms which utilize various spectral features derived from English and Mandarin Chinese phone call audio to predict the language to
which the phone call belongs. We investigate multiple feature sets and modeling approaches, and find that Gaussian Mixture Models,
combined with shifted delta cepstra (SDC) features, achieve the best performance. | ['Justin Pyron', 'Andrew Deveau', 'Julien Boussard'] | 2017-12-16 | null | null | null | null | ['spoken-language-identification'] | ['speech'] | [-1.40394002e-01 -6.61326408e-01 -2.82936394e-01 -6.07156336e-01
-1.06128323e+00 -6.05281711e-01 5.54255247e-01 -1.63451150e-01
-4.25334364e-01 6.40915692e-01 2.63845026e-01 -3.71892840e-01
2.74136905e-02 -3.79389524e-01 1.25065312e-01 -5.27681470e-01
-1.07603550e-01 2.37864792e-01 8.75064358e-02 -1.12722732e-01
5.78938007e-01 5.35450161e-01 -1.59005296e+00 3.07511121e-01
8.87732744e-01 9.95902002e-01 1.66766658e-01 1.02122831e+00
-5.61610222e-01 6.09081805e-01 -9.73348498e-01 1.68889016e-01
-3.09683502e-01 -6.88152969e-01 -6.98450446e-01 1.02838188e-01
-4.28161889e-01 1.53140739e-01 -9.87960398e-02 8.48268092e-01
4.81133074e-01 3.38896364e-01 9.53215361e-01 -1.18759704e+00
-2.15268388e-01 7.81388640e-01 -2.70858914e-01 6.56575620e-01
7.85738707e-01 -5.62388837e-01 6.21531487e-01 -1.06812978e+00
-1.74509417e-02 1.10681725e+00 6.46222353e-01 3.22016060e-01
-9.04395521e-01 -6.79775953e-01 -2.89191622e-02 1.95907757e-01
-1.85788870e+00 -6.94641292e-01 1.02344573e+00 -3.31052542e-01
1.09857082e+00 3.46633017e-01 3.67845118e-01 7.18357921e-01
2.61937737e-01 8.34601045e-01 1.14793074e+00 -9.15769517e-01
2.26188481e-01 4.61347908e-01 3.75229806e-01 5.16512036e-01
-5.03942192e-01 -1.63936257e-01 -5.44308424e-01 -7.65891433e-01
1.86371356e-01 -3.21745157e-01 -4.48289067e-01 4.01149064e-01
-5.94227195e-01 1.07698989e+00 -5.40772021e-01 7.50876665e-01
-4.26217616e-01 -1.13767184e-01 1.07043326e-01 3.11782688e-01
6.08312547e-01 1.60694838e-01 -5.71228802e-01 -5.73285460e-01
-9.58960354e-01 -7.15372488e-02 1.11800075e+00 7.72849500e-01
7.64097631e-01 2.82609880e-01 4.68653888e-01 1.29404187e+00
4.51658309e-01 6.41406059e-01 8.16374779e-01 -9.31924522e-01
4.96250652e-02 -1.18943192e-01 2.17401665e-02 -8.08280945e-01
-3.90990138e-01 4.66056243e-02 -3.83261234e-01 -6.25481784e-01
-1.04724867e-02 -4.95793819e-01 -4.57328528e-01 1.14171147e+00
9.55551341e-02 2.30557844e-01 5.10251045e-01 4.02114183e-01
7.18354285e-01 1.03008270e+00 -2.66023338e-01 -6.60933256e-01
1.09114051e+00 -6.44869089e-01 -1.01736784e+00 1.19050592e-01
4.20849770e-01 -1.16057253e+00 8.78039837e-01 7.71999598e-01
-7.18108952e-01 -6.27784848e-01 -7.29038537e-01 5.86302340e-01
-3.74352247e-01 1.23523213e-01 4.54415381e-01 1.09829760e+00
-8.65208030e-01 4.33539331e-01 -7.49174356e-01 -3.32621932e-02
-5.45792997e-01 2.54595608e-01 -8.95903725e-03 2.45148569e-01
-1.26608217e+00 5.73230922e-01 2.21587837e-01 -6.53898120e-02
-2.72377521e-01 -1.99095473e-01 -8.39227617e-01 -9.52925533e-02
-1.32557705e-01 3.51281643e-01 1.62123871e+00 -9.25807118e-01
-1.93855393e+00 2.69472808e-01 -5.86101711e-01 -3.78379196e-01
-3.73926103e-01 -3.35222669e-02 -1.25257361e+00 4.94972736e-01
-3.22187632e-01 1.87824070e-01 1.14487171e+00 -1.08643687e+00
-9.32126880e-01 2.20460333e-02 -4.96191084e-01 1.03738159e-01
-3.80553305e-01 7.94341505e-01 -1.62584767e-01 -6.14122272e-01
1.87321246e-01 -7.54973531e-01 9.19991955e-02 -1.02181900e+00
-2.48810157e-01 -6.49509132e-01 9.52077568e-01 -6.37280822e-01
1.67278516e+00 -2.42090559e+00 -1.78941905e-01 5.99700212e-01
-3.28405321e-01 1.11647047e-01 1.68013483e-01 5.12766719e-01
2.31816739e-01 -1.42585356e-02 -3.37061435e-02 -2.30890319e-01
-7.21340775e-02 3.61847669e-01 -5.91803908e-01 2.26089731e-01
2.65118986e-01 3.13145667e-01 -5.70742905e-01 -5.26398897e-01
6.48241043e-02 5.93015730e-01 -2.54898906e-01 1.49895802e-01
2.97634572e-01 2.30931252e-01 -3.21139008e-01 5.39610147e-01
4.58779782e-01 2.68008709e-01 -9.09713376e-03 1.62421182e-01
-3.68074656e-01 6.28080070e-01 -1.18330717e+00 9.64456618e-01
-4.84436095e-01 6.67196870e-01 2.89364547e-01 -9.25816596e-01
1.21170461e+00 7.75167584e-01 4.68582064e-01 -1.11534886e-01
1.39263630e-01 4.07261550e-01 1.14360796e-02 -6.95635796e-01
4.53715175e-01 -1.67753771e-01 -2.01857328e-01 3.98206383e-01
3.42481494e-01 -4.20616657e-01 -5.09063788e-02 -3.17326158e-01
6.48310006e-01 -4.61777389e-01 5.61217666e-01 -3.80870104e-01
7.80283928e-01 -2.23725721e-01 4.34287131e-01 5.86281598e-01
-3.83799314e-01 4.57411408e-01 1.83914855e-01 6.55614911e-03
-3.19047183e-01 -8.30029905e-01 -1.56596005e-01 1.46984506e+00
-3.67955327e-01 -4.18672889e-01 -7.29990244e-01 -2.15637445e-01
-2.71790028e-01 8.92607152e-01 2.70608161e-02 -1.03007384e-01
-4.72346902e-01 -6.28463209e-01 9.23361540e-01 4.12891299e-01
2.73425151e-02 -9.02157128e-01 -2.95793772e-01 4.34776872e-01
-3.32894534e-01 -1.05293977e+00 -9.23456788e-01 5.16693592e-01
-6.28362834e-01 -7.03624964e-01 -7.24171579e-01 -9.37038064e-01
-1.29941031e-01 2.75789320e-01 5.17822206e-01 -4.97278631e-01
-8.01458880e-02 8.35122287e-01 -8.74151349e-01 -5.74764967e-01
-5.07544279e-01 -6.42613694e-02 3.95381749e-01 3.13144267e-01
6.63821042e-01 -3.97317082e-01 1.06723204e-01 1.76187754e-01
-6.51134014e-01 -8.38572621e-01 8.12170953e-02 4.86427963e-01
4.09824461e-01 5.15584648e-01 6.95384741e-01 -5.42151332e-01
1.13222003e+00 -5.13296664e-01 -3.49277556e-01 1.82854593e-01
-4.82777655e-01 -2.34731063e-01 6.84985638e-01 -6.43533587e-01
-9.59164679e-01 1.43923342e-01 -4.89752978e-01 -3.07049334e-01
-5.63976884e-01 7.55728662e-01 -8.03198218e-02 -2.09577814e-01
4.51482296e-01 8.41854453e-01 -9.94902551e-02 -6.35725319e-01
1.57023035e-02 1.38898599e+00 4.94758517e-01 -3.17945331e-01
1.52316883e-01 4.40990329e-02 -7.83256054e-01 -1.66927373e+00
-4.74694699e-01 -9.87642050e-01 -3.69447798e-01 -2.55562186e-01
7.94857323e-01 -6.70292139e-01 -3.53603452e-01 6.45702720e-01
-1.13477707e+00 2.28753969e-01 2.84098417e-01 1.16383660e+00
-6.10817611e-01 4.42921519e-01 -7.60541379e-01 -1.64922059e+00
-1.70566052e-01 -8.13282549e-01 9.60087299e-01 3.04242909e-01
-5.34210205e-01 -9.11977530e-01 3.00818413e-01 -1.31722152e-01
3.45777720e-01 -5.13142705e-01 7.19043195e-01 -1.19048536e+00
3.19391012e-01 -2.23867089e-01 4.66452330e-01 5.32010674e-01
4.98985618e-01 5.30114889e-01 -1.14877594e+00 2.99251955e-02
3.94473672e-01 -9.76619497e-02 5.70297837e-01 6.81312263e-01
1.14905310e+00 -1.55260876e-01 -2.01595187e-01 4.14402068e-01
7.49945343e-01 9.90296543e-01 2.51218170e-01 -2.86494195e-01
8.15941766e-02 5.80046415e-01 5.01121879e-01 5.42064011e-01
4.00662571e-01 3.99011880e-01 -5.34923732e-01 2.98448175e-01
4.92686421e-01 -5.05329855e-02 6.35353923e-01 1.60946727e+00
1.01891428e-01 -2.37540215e-01 -1.02667487e+00 2.76055723e-01
-1.28036177e+00 -1.04716659e+00 -3.00820395e-02 1.87027276e+00
8.21439028e-01 6.64328858e-02 4.08013076e-01 7.10216820e-01
7.10816622e-01 -6.98180422e-02 1.05043925e-01 -5.83315611e-01
-2.26394370e-01 5.72978795e-01 1.63515583e-01 7.94650197e-01
-1.32408047e+00 8.07767153e-01 8.32006168e+00 9.93812382e-01
-1.28669596e+00 2.34674010e-02 2.84754813e-01 4.17895705e-01
-1.82566822e-01 -2.95593500e-01 -8.95258248e-01 4.44504887e-01
1.51242745e+00 -3.51846278e-01 4.90497649e-01 7.34638274e-01
3.35538149e-01 -2.03496858e-01 -6.65142298e-01 1.09249735e+00
3.32116336e-01 -8.20809126e-01 -2.77391076e-01 -1.38034374e-01
2.72760630e-01 -4.65168282e-02 5.54490499e-02 2.95956552e-01
-4.84045688e-03 -7.30197191e-01 5.36388934e-01 7.15989351e-01
4.04301494e-01 -1.11108768e+00 6.32344246e-01 6.36472404e-01
-1.43750870e+00 2.83723921e-02 -1.82577282e-01 -1.71838738e-02
1.44925445e-01 4.21887457e-01 -9.32664752e-01 4.36077476e-01
6.85132563e-01 2.69051969e-01 -1.30691350e-01 9.96500075e-01
2.18839556e-01 1.17535031e+00 -4.38603550e-01 -4.34709728e-01
2.63223737e-01 -1.89928770e-01 3.87805104e-01 1.56824005e+00
6.06023610e-01 4.06127781e-01 4.03269976e-01 4.37852204e-01
5.68827093e-01 4.23396081e-01 -5.42424381e-01 -4.99474406e-01
8.24775815e-01 8.18163216e-01 -8.69994760e-01 -3.03216934e-01
-5.80916524e-01 8.37723136e-01 -3.06526601e-01 4.90049034e-01
-6.34591937e-01 -8.45507741e-01 6.12813234e-01 -4.19143111e-01
2.44978130e-01 -5.79280198e-01 3.30606312e-01 -1.01069677e+00
-4.14728224e-01 -7.49286115e-01 2.78251946e-01 -4.08771098e-01
-1.29461253e+00 8.02254021e-01 8.91272500e-02 -1.17338192e+00
-8.84862959e-01 -4.77827311e-01 -7.67613113e-01 9.79790688e-01
-1.34284210e+00 -3.41806471e-01 4.14855689e-01 7.02027440e-01
7.16383755e-01 -6.16585314e-01 1.22680736e+00 2.79150069e-01
-2.40190536e-01 5.01930118e-01 3.90989363e-01 1.81202114e-01
5.24345040e-01 -1.08515167e+00 5.08278981e-02 2.47456580e-01
6.37841463e-01 5.75675368e-01 7.32709587e-01 -2.53520578e-01
-1.15533805e+00 -6.91492379e-01 1.25881588e+00 -9.07729045e-02
6.13536358e-01 -1.30440414e-01 -9.54322636e-01 2.42574528e-01
1.33268103e-01 -4.19240147e-01 1.39127624e+00 8.42960104e-02
-8.05402622e-02 3.41038704e-02 -9.68157887e-01 -9.84259322e-02
1.58106297e-01 -1.03886962e+00 -8.12654376e-01 1.69715304e-02
7.40477145e-01 1.07308887e-01 -7.97078490e-01 3.44937108e-02
3.93765837e-01 -8.10900927e-01 6.72146380e-01 -5.41708529e-01
-3.70066822e-01 -2.12283790e-01 -6.48743391e-01 -1.45153189e+00
-1.60284027e-01 -7.92559564e-01 -1.14370612e-02 1.36153173e+00
6.03923202e-01 -9.63734329e-01 3.96008968e-01 1.27843812e-01
-1.37963161e-01 -3.06025863e-01 -1.03789818e+00 -8.50596666e-01
-4.95211259e-02 -7.77166545e-01 4.62720037e-01 7.89673030e-01
5.12366712e-01 3.90667230e-01 -3.51326048e-01 1.00936994e-01
1.53135166e-01 1.92282230e-01 1.24832235e-01 -1.12787950e+00
-3.80362988e-01 -5.89876413e-01 -1.28420487e-01 -1.11225832e+00
4.92299139e-01 -4.49830860e-01 4.70185161e-01 -7.32842624e-01
-3.12342912e-01 -2.25294277e-01 -4.06187475e-01 9.11797658e-02
-9.85307470e-02 -1.37018293e-01 -2.58710496e-02 6.16181903e-02
-3.07351232e-01 5.97560823e-01 3.21652859e-01 -5.93280494e-02
-5.97457349e-01 8.29215288e-01 -3.68110269e-01 9.43241417e-01
9.22332764e-01 -3.10190290e-01 -3.27776521e-01 -3.79928686e-02
-4.72736001e-01 3.36938590e-01 -2.68121362e-01 -8.55354428e-01
3.31495404e-01 -1.82605669e-01 9.05643329e-02 -6.92872465e-01
5.95397949e-01 -6.50802791e-01 -1.17801107e-01 3.72152448e-01
-3.88461828e-01 1.54113770e-01 9.48636010e-02 4.66599435e-01
-9.45651054e-01 -5.97314715e-01 6.63404524e-01 1.26544327e-01
-5.62056959e-01 -1.82262480e-01 -1.33777452e+00 -2.77109325e-01
5.88341713e-01 -4.96533923e-02 3.49876761e-01 -8.02717805e-01
-7.64582396e-01 -1.19643040e-01 -1.54785812e-01 3.92254889e-01
7.90626884e-01 -1.24826217e+00 -4.39224243e-01 5.01300573e-01
-8.29130411e-02 -8.15030277e-01 -4.41793799e-02 8.09087992e-01
-2.82780856e-01 8.18469882e-01 2.68198490e-01 -4.65594411e-01
-1.35159791e+00 1.88939437e-01 2.40474418e-01 3.97858411e-01
2.86603589e-02 9.30620492e-01 -5.70393205e-01 -5.42908847e-01
3.89634162e-01 -3.06075037e-01 -5.03035486e-01 2.10226297e-01
6.85596824e-01 4.04021204e-01 5.47218435e-02 -8.49588454e-01
-6.15400255e-01 5.04639030e-01 3.36609125e-01 -6.49281919e-01
8.83569658e-01 -2.91469961e-01 -1.25697866e-01 1.13079214e+00
1.30089784e+00 3.90362680e-01 -5.31435907e-01 -2.19239905e-01
4.94114429e-01 -2.60823905e-01 4.46490720e-02 -2.97918171e-01
-5.17592371e-01 7.72122264e-01 5.62642097e-01 7.67659605e-01
1.38416421e+00 6.10386990e-02 6.38128877e-01 4.56670821e-01
4.36428905e-01 -1.49082625e+00 -3.23969573e-01 9.48337674e-01
3.90935421e-01 -9.89696801e-01 -5.98948658e-01 -3.96296203e-01
-8.32481205e-01 1.43864870e+00 2.05056816e-01 -1.71039000e-01
1.48269355e+00 5.90541363e-01 3.96041602e-01 3.43176484e-01
-8.63627672e-01 -3.23751092e-01 4.24729466e-01 6.70485437e-01
7.62413561e-01 2.48029545e-01 -4.39627439e-01 9.60579276e-01
-5.25199950e-01 -3.18432927e-01 4.33304429e-01 9.59392667e-01
-1.08665359e+00 -1.10319543e+00 -8.48614573e-01 5.60695469e-01
-4.85636920e-01 -2.08535623e-02 -6.39650404e-01 3.44417155e-01
-1.11116663e-01 1.54012644e+00 1.46138653e-01 -6.78700805e-01
1.38071343e-01 8.05282712e-01 1.69042081e-01 -6.23537660e-01
-3.31449240e-01 1.02900565e+00 1.17793314e-01 -1.33002684e-01
-6.88649237e-01 -8.29233944e-01 -1.45870173e+00 2.75539532e-02
-5.32169640e-01 9.28638935e-01 6.82424247e-01 1.09631324e+00
-4.01215963e-02 1.31366655e-01 1.19284618e+00 -6.70770586e-01
-7.02725887e-01 -1.26618314e+00 -1.11831427e+00 -3.41577888e-01
4.62078273e-01 -3.10344458e-01 -4.47305620e-01 1.34567171e-01] | [14.684094429016113, 5.975832462310791] |
a6859e4c-b357-4832-aaf8-bdcc9a53aff0 | vocalset-a-singing-voice-dataset | null | null | http://ismir2018.ircam.fr/pages/events-main-program.html | http://ismir2018.ircam.fr/doc/pdfs/114_Paper.pdf | VocalSet: A Singing Voice Dataset | We present VocalSet, a singing voice dataset of a capella singing. Existing singing voice datasets either do not capture a large range of vocal techniques, have very few singers, or are single-pitch and devoid of musical context. VocalSet captures not only a range of vowels, but also a diverse set of voices on many different vocal techniques, sung in contexts of scales, arpeggios, long tones, and excerpts. VocalSet has recordings of 10.1 hours of 20 professional singers (11 male, 9 female) performing 17 different different vocal techniques. This data will facilitate the development of new machine learning models for singer identification, vocal technique identification, singing generation and other related applications. To illustrate this, we establish baseline results on vocal technique classification and singer identification by training convolutional network classifiers on VocalSet to perform these tasks. | ['Bryan Pardo', 'Alison Wahl', 'Prem Seetharaman', 'Julia Wilkins'] | 2018-09-25 | null | null | null | international-society-for-music-information | ['singer-identification', 'vocal-technique-classification'] | ['music', 'music'] | [-3.94937098e-02 -6.58532798e-01 -4.34879698e-02 -1.98776528e-01
-7.22211063e-01 -1.44657028e+00 3.26878846e-01 -3.75881076e-01
1.22281045e-01 9.20313373e-02 2.21016333e-01 -7.15868473e-02
-1.37328148e-01 -2.49630645e-01 -1.56798586e-01 -5.16154885e-01
-2.19389409e-01 3.14680457e-01 -1.61071837e-01 -3.53575468e-01
4.38349061e-02 6.30340993e-01 -1.66755044e+00 3.43525350e-01
-6.16300814e-02 9.47604477e-01 6.30676523e-02 1.26381385e+00
-1.30449021e-02 3.92484605e-01 -1.20205462e+00 -2.41280466e-01
5.51660992e-02 -5.53201795e-01 -5.93853712e-01 3.15256324e-03
1.05499697e+00 -2.24203095e-01 -1.58942327e-01 5.31198144e-01
9.46123123e-01 3.94132733e-01 4.78943139e-01 -8.53575647e-01
-4.68503207e-01 1.11946487e+00 -9.76421908e-02 5.23504198e-01
3.99387330e-01 4.12540823e-01 1.50719237e+00 -8.71893883e-01
4.16009396e-01 1.26084638e+00 1.05333853e+00 9.47912693e-01
-1.13040245e+00 -1.13162494e+00 -7.48257339e-01 -1.25473753e-01
-1.03778982e+00 -8.08098793e-01 9.15294528e-01 -4.18277174e-01
7.59313107e-01 6.88623607e-01 9.90204573e-01 1.33045793e+00
-3.63066077e-01 7.32164085e-01 8.45321834e-01 -1.97510287e-01
-1.35644287e-01 -2.68059641e-01 3.64820287e-02 3.95980805e-01
-7.50206053e-01 4.02568132e-01 -1.00934434e+00 -3.72449368e-01
7.36342669e-01 -3.24955881e-01 -4.44303006e-01 4.99603868e-01
-1.33006346e+00 5.26093602e-01 1.09801270e-01 6.07400119e-01
1.49551988e-01 4.09255117e-01 8.40032339e-01 6.71853960e-01
2.39169784e-02 1.20537114e+00 -3.33232373e-01 -6.71918094e-01
-1.06485403e+00 7.22825289e-01 9.80028808e-01 7.94495344e-01
1.84534401e-01 7.60627687e-01 -1.63467899e-01 1.40077651e+00
-3.40805769e-01 4.83500063e-01 5.72597146e-01 -1.37780201e+00
2.04060540e-01 -1.01255395e-01 -2.41790429e-01 -4.17257458e-01
-3.95400047e-01 -6.27022803e-01 -3.48493606e-01 -1.56265378e-01
4.70325440e-01 -7.68578099e-03 -5.26633978e-01 1.56599271e+00
-2.55081773e-01 3.55314225e-01 -4.08159763e-01 9.21823502e-01
1.39526677e+00 5.76334953e-01 -3.71219754e-01 -1.67982265e-01
1.19335616e+00 -1.14744055e+00 -8.14225197e-01 3.35231811e-01
-7.05427006e-02 -1.27576160e+00 1.65798879e+00 7.69824266e-01
-1.39101636e+00 -9.29090261e-01 -9.03898358e-01 -1.87933490e-01
-5.26063628e-02 1.75855637e-01 5.58116078e-01 8.33373666e-01
-7.01651335e-01 9.11452651e-01 -2.31569543e-01 3.27655189e-02
2.19917431e-01 2.50771254e-01 -2.01271400e-01 8.46970558e-01
-8.92638326e-01 2.18439117e-01 -1.22987017e-01 -2.49739483e-01
-1.29293811e+00 -1.25216687e+00 -4.20060396e-01 1.46075770e-01
1.25008345e-01 -3.64758939e-01 2.14404392e+00 -7.35286057e-01
-1.75044715e+00 9.37722385e-01 1.74555138e-01 -2.80226558e-01
-6.11941591e-02 -2.37968490e-01 -8.12936306e-01 -3.41713317e-02
-1.54571727e-01 1.23529240e-01 1.31176198e+00 -8.03154349e-01
-3.76524150e-01 2.98342295e-02 -1.80013418e-01 -4.65065390e-02
-5.37624419e-01 5.98787546e-01 5.90902455e-02 -9.22992587e-01
-2.75143176e-01 -9.38245475e-01 5.36568224e-01 -4.75908935e-01
-5.07367134e-01 -3.18942457e-01 9.07494009e-01 -4.83615488e-01
1.53322375e+00 -2.54771399e+00 3.36521417e-01 -1.43674284e-01
2.77749747e-01 4.43962365e-01 -2.77075082e-01 6.78842425e-01
-1.13923587e-01 3.39820206e-01 1.37640573e-02 -4.42426622e-01
1.04677789e-02 7.43236579e-03 -7.52947152e-01 1.34061992e-01
-3.03519845e-01 6.65249705e-01 -9.52083349e-01 -8.04785937e-02
1.23208568e-01 1.16754748e-01 -4.54403579e-01 5.65008342e-01
-5.98688088e-02 5.09079337e-01 4.28422354e-02 1.09728181e+00
1.05523497e-01 4.47162986e-01 -2.09300995e-01 -1.39708117e-01
-1.97061256e-01 8.00701022e-01 -8.30236256e-01 1.57777095e+00
-8.14196467e-01 8.24627280e-01 5.42874515e-01 -1.69419050e-01
7.94908583e-01 5.66061735e-01 3.13235521e-01 7.33386800e-02
1.12164691e-01 7.11550832e-01 5.99306643e-01 -4.77364272e-01
6.63075328e-01 -4.88906652e-01 -1.06775193e-02 4.69342649e-01
6.54204726e-01 -7.38751709e-01 1.59993157e-01 -3.14929754e-01
1.15532815e+00 -4.16214854e-01 3.77836227e-02 -1.37393549e-01
2.45787844e-01 -6.15392268e-01 3.20108384e-01 6.44091308e-01
-1.17225409e-01 9.60704863e-01 1.24851868e-01 -2.60059386e-01
-8.40351760e-01 -1.32883573e+00 -3.51015925e-01 1.66737986e+00
-6.62595987e-01 -9.22810137e-01 -5.36882401e-01 -2.11213633e-01
1.24323107e-01 4.88802373e-01 -2.90683538e-01 -7.12680444e-02
-8.10515165e-01 1.41197652e-01 1.47881973e+00 6.07438684e-01
4.36952263e-02 -1.75160503e+00 -1.02888860e-01 2.33383566e-01
1.28370926e-01 -7.19845176e-01 -1.38172567e+00 2.19179437e-01
-5.14172733e-01 -9.83974814e-01 -5.45197129e-01 -9.70340371e-01
-4.71691102e-01 2.23740786e-01 1.45046711e+00 -7.87504688e-02
-2.88443178e-01 3.09155077e-01 -1.78142443e-01 -6.38015687e-01
-8.43958974e-01 2.72050321e-01 6.29530430e-01 -2.47247100e-01
2.99451314e-03 -9.90924716e-01 -2.25064352e-01 3.81518245e-01
-4.12572116e-01 -5.99777758e-01 -3.49312415e-03 1.08293474e+00
3.97254527e-01 -1.00473896e-01 1.03089082e+00 -6.49565279e-01
1.03493488e+00 -3.30317527e-01 -2.96093255e-01 -1.34353563e-01
1.66589171e-02 -7.81932175e-01 9.28205490e-01 -1.00322223e+00
-4.88491625e-01 -3.53183329e-01 -5.28774023e-01 -9.83016968e-01
-2.60854989e-01 2.88107097e-01 -5.46855927e-02 -4.49876487e-03
1.02507615e+00 2.01479360e-01 -1.20432027e-01 -1.03013730e+00
3.60097736e-01 1.12990081e+00 1.04447865e+00 -6.60636127e-01
8.22446823e-01 -1.98726058e-01 -3.13849568e-01 -1.21128917e+00
-8.90026808e-01 -8.25006723e-01 -4.00195003e-01 -3.87675554e-01
5.52690685e-01 -4.37560171e-01 -1.04894662e+00 6.64917052e-01
-8.17005813e-01 -2.55791903e-01 -6.05563223e-01 2.76096046e-01
-7.90825605e-01 6.66887388e-02 -1.03348768e+00 -8.74920130e-01
-8.15427721e-01 -8.07494521e-01 1.13979697e+00 2.40388051e-01
-7.32194781e-01 -6.92358315e-01 1.45080850e-01 3.61777782e-01
4.45720017e-01 -7.71542732e-03 7.08772719e-01 -7.00584471e-01
-3.43440217e-03 -1.88345369e-02 4.97910142e-01 5.87191880e-01
6.48605585e-01 3.22816938e-01 -1.30301344e+00 -2.75759101e-01
-7.26362981e-04 -7.34234750e-01 7.59269118e-01 1.90820545e-01
1.41340768e+00 -2.61004031e-01 3.70400876e-01 8.48750353e-01
3.36714119e-01 2.34031722e-01 -1.06532304e-02 -2.43579820e-01
8.10599685e-01 5.53239346e-01 4.57177490e-01 1.92840904e-01
-3.70942563e-01 9.19836342e-01 1.82607844e-01 1.88482136e-01
-5.73135555e-01 -3.91246289e-01 4.95864153e-01 1.30898249e+00
-3.43007088e-01 -1.81795899e-02 -7.49698699e-01 8.81002784e-01
-9.97983098e-01 -1.26652932e+00 -4.71715108e-02 2.03004956e+00
1.25828946e+00 -2.88683116e-01 8.75599623e-01 6.28905237e-01
6.29739940e-01 4.31013256e-01 -5.01456857e-01 -9.32308078e-01
3.35061476e-02 8.44108403e-01 -9.96154696e-02 2.54057050e-02
-1.10057151e+00 9.94956970e-01 7.44966030e+00 9.22500074e-01
-1.40839803e+00 -1.67670287e-02 -1.67139307e-01 -6.15014434e-01
-1.25609323e-01 -3.34739566e-01 -7.39276409e-01 3.74512494e-01
1.17715096e+00 -2.04910375e-02 1.19214058e+00 8.14021409e-01
7.30238184e-02 8.23556423e-01 -1.24952865e+00 1.19435942e+00
9.91565408e-04 -1.45859325e+00 -9.03735310e-02 -1.94015577e-02
6.58096075e-01 -2.45772861e-03 4.82747525e-01 6.04357064e-01
-2.46346340e-01 -1.37045908e+00 8.99936140e-01 1.96525186e-01
1.16018879e+00 -7.61877358e-01 2.59887755e-01 1.93906888e-01
-1.27871418e+00 -2.38169074e-01 -1.72300152e-02 6.70851767e-02
6.60335645e-02 -6.95076659e-02 -9.66380060e-01 1.74215496e-01
7.97658145e-01 8.93458843e-01 -2.25551471e-01 1.00196803e+00
6.31436259e-02 1.44835520e+00 -2.76690006e-01 7.04163918e-03
-7.49306604e-02 3.33655588e-02 1.17200184e+00 1.30970705e+00
3.83152574e-01 -2.63320625e-01 -3.79661135e-02 8.79040480e-01
-4.08363938e-01 2.44145706e-01 -4.75971699e-01 -8.54030490e-01
7.49717295e-01 1.54141927e+00 -9.25496370e-02 6.05756082e-02
-1.52077004e-01 4.23817337e-01 -2.02263907e-01 9.29029882e-02
-5.32491446e-01 -4.01102781e-01 8.46949041e-01 3.37958813e-01
-8.33288999e-04 -4.64515865e-01 5.36010452e-02 -6.81231260e-01
-2.40254998e-01 -1.41512561e+00 1.94158733e-01 -7.13044167e-01
-1.69188190e+00 6.00348890e-01 -3.99689376e-01 -1.45770979e+00
-6.12291157e-01 -5.64456046e-01 -1.23166645e+00 7.86706984e-01
-9.35469747e-01 -1.40111291e+00 -1.93002149e-01 3.67550969e-01
1.07231128e+00 -8.59827101e-01 1.18785667e+00 4.79832083e-01
-1.30403861e-01 7.46892035e-01 -1.87197819e-01 8.43227282e-02
1.09302580e+00 -1.70844364e+00 6.05898440e-01 -1.43849611e-01
8.07326674e-01 8.48253429e-01 6.31666303e-01 5.76478653e-02
-1.43628120e+00 -9.08751905e-01 5.82339942e-01 -6.48717523e-01
9.77212906e-01 -5.09533942e-01 -8.87116015e-01 4.17546660e-01
1.34585515e-01 -9.75716636e-02 1.12240660e+00 6.93007588e-01
-4.13239121e-01 -2.32690185e-01 -6.74990535e-01 4.24158514e-01
1.20561969e+00 -9.75691020e-01 -7.08844185e-01 2.78567463e-01
6.15727305e-01 -4.59969759e-01 -1.12243330e+00 2.41084874e-01
9.86948669e-01 -8.48367691e-01 8.30913782e-01 -8.23928058e-01
3.97752255e-01 -1.16553657e-01 2.75068339e-02 -1.43637645e+00
-1.91313863e-01 -1.17762530e+00 -4.27847177e-01 1.62987530e+00
2.20318735e-01 -2.39075273e-02 2.00137883e-01 -6.15516186e-01
-6.64052725e-01 -4.06777114e-01 -7.99469054e-01 -1.09577131e+00
2.19269097e-01 -5.59880733e-01 8.00615728e-01 1.13685811e+00
-1.11805938e-01 5.90531945e-01 -6.38304532e-01 -4.74982738e-01
1.71327502e-01 4.75729108e-01 9.39199746e-01 -1.22266161e+00
-9.47410583e-01 -7.72772610e-01 -1.32085785e-01 -7.42635489e-01
3.49287122e-01 -1.06926084e+00 -1.18668988e-01 -5.61216593e-01
-2.51378059e-01 -2.99887955e-01 -4.50339288e-01 4.88287300e-01
1.26001939e-01 4.64642912e-01 2.60424078e-01 3.83135676e-01
-8.77121240e-02 3.62335980e-01 1.38661754e+00 -2.76794165e-01
-2.42896497e-01 8.01266372e-01 -3.73411536e-01 9.08430576e-01
8.90044451e-01 -1.71755031e-01 -3.39849502e-01 -1.45923957e-01
-1.11329347e-01 3.34709823e-01 1.61117390e-01 -1.03911030e+00
-1.79800361e-01 2.70118825e-02 -9.07735825e-02 -4.58484173e-01
8.00768316e-01 -1.22994542e-01 3.78560275e-02 1.35160601e-02
-5.88172555e-01 -2.60634184e-01 2.33717978e-01 1.51506081e-01
-4.36265320e-01 -3.70623380e-01 8.10868025e-01 -8.22714195e-02
-2.39693016e-01 2.62811035e-01 -3.99980962e-01 5.54743297e-02
2.08167344e-01 -1.76163130e-02 -1.47664547e-01 -4.79523122e-01
-8.79031420e-01 -3.77841145e-01 2.38594800e-01 9.70174551e-01
3.58892858e-01 -1.34761000e+00 -7.48245418e-01 1.39813408e-01
3.19780767e-01 -4.84724551e-01 4.87625040e-02 4.57718670e-01
-3.43986839e-01 6.37501478e-02 -2.15419754e-01 -4.12083328e-01
-1.77479160e+00 2.44721234e-01 6.08644545e-01 3.19918185e-01
-5.17987132e-01 1.12000954e+00 1.24421008e-02 -9.13265884e-01
3.01808745e-01 -2.48974085e-01 -4.03679937e-01 1.49443716e-01
4.52883393e-01 5.52248359e-01 -1.13735743e-01 -7.41213322e-01
-7.96074420e-02 4.67729539e-01 1.88616589e-01 -9.56273898e-02
9.59884226e-01 4.19763386e-01 -1.07308418e-01 1.20160234e+00
1.18904650e+00 9.23788249e-01 -5.42401254e-01 1.74890712e-01
-1.46938697e-01 -5.47364473e-01 6.00403622e-02 -7.97285080e-01
-8.52303922e-01 9.71288264e-01 3.67426932e-01 6.58903837e-01
1.03019750e+00 1.29674271e-01 1.24414253e+00 3.37592483e-01
1.75393552e-01 -8.72822046e-01 2.89336354e-01 6.29764259e-01
1.48853946e+00 -8.92089069e-01 -3.82392198e-01 -2.77339727e-01
-4.14454550e-01 1.33513463e+00 5.47673941e-01 -1.19009316e-01
4.92510676e-01 3.62888515e-01 2.40522176e-01 -2.01553643e-01
-7.35579431e-01 -1.36691034e-01 6.35725856e-01 3.60355675e-01
8.10223162e-01 2.21025378e-01 -2.19604105e-01 1.04340351e+00
-1.18887651e+00 -4.44481701e-01 2.84837425e-01 2.88755745e-01
-1.99332938e-01 -1.24168897e+00 -2.87960917e-01 5.19629359e-01
-8.74549031e-01 -2.59571135e-01 -9.81138051e-01 4.40465093e-01
3.40369731e-01 1.08685076e+00 -3.08843534e-02 -7.10404396e-01
4.84345585e-01 1.08212575e-01 5.69050074e-01 -9.14163589e-01
-1.51861095e+00 3.41996491e-01 4.36119944e-01 -3.01536713e-02
-3.33691165e-02 -7.20238745e-01 -9.77654457e-01 -2.87089705e-01
-4.95841593e-01 2.58529276e-01 6.28840506e-01 3.84702176e-01
-2.80720834e-02 7.96740651e-01 1.02059710e+00 -1.02858281e+00
-8.39543700e-01 -1.48909259e+00 -1.22174489e+00 3.93555790e-01
5.46369135e-01 -4.54162419e-01 -6.32762194e-01 4.32080999e-02] | [15.490644454956055, 6.014828681945801] |
3dcf909c-27ac-4afe-b823-34dcabebc35f | qtrojan-a-circuit-backdoor-against-quantum | 2302.08090 | null | https://arxiv.org/abs/2302.08090v1 | https://arxiv.org/pdf/2302.08090v1.pdf | QTrojan: A Circuit Backdoor Against Quantum Neural Networks | We propose a circuit-level backdoor attack, \textit{QTrojan}, against Quantum Neural Networks (QNNs) in this paper. QTrojan is implemented by few quantum gates inserted into the variational quantum circuit of the victim QNN. QTrojan is much stealthier than a prior Data-Poisoning-based Backdoor Attack (DPBA), since it does not embed any trigger in the inputs of the victim QNN or require the access to original training datasets. Compared to a DPBA, QTrojan improves the clean data accuracy by 21\% and the attack success rate by 19.9\%. | ['Fan Chen', 'Martin Swany', 'Lei Jiang', 'Cheng Chu'] | 2023-02-16 | null | null | null | null | ['data-poisoning'] | ['adversarial'] | [ 3.77077937e-01 4.76540834e-01 5.34431124e-03 2.58948773e-01
-6.16067588e-01 -8.66241455e-01 5.11937797e-01 1.40641602e-02
-7.21358657e-01 8.85354757e-01 -5.00650167e-01 -8.61458242e-01
1.72815755e-01 -1.18304729e+00 -1.10387528e+00 -1.23378062e+00
2.85850227e-01 -1.96105093e-01 3.30593169e-01 -4.84669715e-01
1.86236069e-01 5.18354356e-01 -6.46948814e-01 -1.02685735e-01
6.30281687e-01 4.63376671e-01 -4.32605088e-01 6.53608620e-01
4.66489166e-01 7.34593391e-01 -7.82538354e-01 -1.03544044e+00
4.02618945e-01 -6.31534159e-01 -5.73916256e-01 -1.29442596e+00
4.63925779e-01 -5.19361675e-01 -1.47039425e+00 1.91656935e+00
3.27867091e-01 -1.30320430e-01 4.21083450e-01 -1.14245725e+00
-7.96210945e-01 1.16749191e+00 -1.01782694e-01 3.26988757e-01
-3.94048750e-01 2.66627222e-01 8.78703952e-01 -2.57490188e-01
5.58694720e-01 8.27356994e-01 4.80580539e-01 1.27943349e+00
-1.46050835e+00 -1.46903574e+00 -8.28007698e-01 2.98288226e-01
-1.46715879e+00 -4.41023976e-01 5.97237766e-01 1.96828842e-01
9.75051403e-01 3.24365109e-01 3.73706609e-01 1.50921583e+00
6.96493566e-01 2.26631671e-01 8.57250094e-01 -2.70691514e-01
4.93469357e-01 1.28494129e-02 5.02392054e-01 8.19665194e-01
9.39728022e-01 1.03616965e+00 -5.52597940e-01 -5.39327502e-01
7.49613762e-01 -1.36856670e-02 4.82196175e-02 5.72999418e-02
-6.54140711e-01 1.09122193e+00 7.50723422e-01 8.27680305e-02
1.02432288e-01 1.04508150e+00 3.00969660e-01 5.53076208e-01
-3.71634096e-01 6.19551122e-01 -2.83697173e-02 1.11118101e-01
-5.26477516e-01 3.08344424e-01 7.16931880e-01 7.64342427e-01
7.75448561e-01 3.69039416e-01 -1.54173121e-01 -4.56762820e-01
1.70685634e-01 1.09915125e+00 -5.48057035e-02 -7.84280062e-01
4.73049641e-01 1.03565417e-02 -4.88818079e-01 -5.18713593e-01
-2.53061652e-01 -3.53620082e-01 -8.96964788e-01 8.24703276e-02
3.81328940e-01 -6.29022539e-01 -1.11380446e+00 1.97757304e+00
1.92445666e-02 9.47552472e-02 6.20047152e-01 4.50284123e-01
9.48873937e-01 6.52024806e-01 -9.23974440e-02 2.30229795e-02
1.16302526e+00 -4.29174155e-01 -8.97592723e-01 1.95505053e-01
7.81167507e-01 -8.63545313e-02 1.76112771e-01 4.75719392e-01
-6.80278897e-01 1.91070233e-02 -1.42835796e+00 -4.78938483e-02
-4.48406339e-01 -6.23053908e-01 7.87768722e-01 1.62166226e+00
-6.97931111e-01 6.80513203e-01 -1.04130220e+00 3.23349983e-01
7.20013916e-01 7.92888045e-01 -4.97655451e-01 5.99753670e-03
-1.72942030e+00 7.32519448e-01 8.26464713e-01 -1.34276703e-01
-1.59771109e+00 -4.84270126e-01 -6.62658215e-01 1.97660606e-02
3.44458282e-01 -6.60830855e-01 8.78377140e-01 -9.07073021e-02
-1.71867442e+00 3.85932833e-01 1.50496110e-01 -1.22277629e+00
7.38316402e-02 5.09589724e-02 -6.13704503e-01 4.79079068e-01
-3.56393099e-01 2.41221011e-01 8.39874804e-01 -6.86475217e-01
5.77109307e-02 -4.50714260e-01 2.14013353e-01 -6.27221763e-01
-2.09280565e-01 -1.64058313e-01 1.21129796e-01 -3.74740958e-01
-6.04930557e-02 -1.12475777e+00 -4.75387216e-01 -6.24025941e-01
-1.09211075e+00 -4.36789058e-02 8.23630810e-01 -1.20141003e-02
9.40302730e-01 -2.20273614e+00 -2.80578844e-02 4.88628149e-01
6.63810253e-01 7.61645198e-01 -1.37492180e-01 6.91431999e-01
-3.65266725e-02 2.67391086e-01 -2.60766417e-01 3.88613679e-02
9.75707769e-02 4.19923306e-01 -8.85812104e-01 7.50703573e-01
-5.60491011e-02 1.17913890e+00 -8.87322962e-01 1.19272083e-01
-9.66373533e-02 4.64925468e-01 -8.46755624e-01 -3.95689934e-01
-8.02366212e-02 2.01475665e-01 -4.63346899e-01 5.00723839e-01
8.34654331e-01 1.84615124e-02 -8.43975171e-02 -1.48544714e-04
2.46926948e-01 2.97290683e-01 -5.21375000e-01 1.37871921e+00
2.11004257e-01 5.96027136e-01 -6.37220562e-01 -5.66121101e-01
6.79106891e-01 5.10762990e-01 -1.19854249e-01 -7.87349463e-01
6.66152954e-01 3.24099869e-01 3.55738789e-01 -7.17919543e-02
3.88030976e-01 -5.19779921e-01 -4.17593896e-01 3.14362466e-01
5.07174969e-01 4.45401222e-02 -3.10475320e-01 5.62525868e-01
1.55332780e+00 -4.91970748e-01 -6.73114583e-02 -2.99608558e-01
1.74011439e-01 -2.22812474e-01 6.51731431e-01 1.41145587e+00
-3.52855802e-01 9.96556729e-02 7.58622944e-01 -1.69325128e-01
-9.23766077e-01 -1.46803939e+00 -3.78725469e-01 4.60599780e-01
4.22444373e-01 -7.29080617e-01 -1.08009124e+00 -6.78419769e-01
-2.74256561e-02 8.22130919e-01 -6.33299828e-01 -9.75769758e-01
-1.33458339e-02 -8.01574349e-01 1.81946170e+00 8.10173620e-03
9.31925774e-01 -7.73890316e-01 -3.65266323e-01 -2.52947379e-02
3.08090031e-01 -1.26818895e+00 -1.69304669e-01 8.18182886e-01
-9.60531473e-01 -1.06479132e+00 1.79875091e-01 -2.36833885e-01
5.42666197e-01 -1.01296738e-01 2.85069913e-01 -2.75317341e-01
-1.34350032e-01 -4.53080654e-01 -1.60391450e-01 -6.05331600e-01
-5.95137835e-01 7.22190319e-03 4.86625165e-01 -3.40787321e-01
7.84528553e-01 -4.62988704e-01 -2.98826754e-01 -2.51075059e-01
-1.25292456e+00 -4.43213373e-01 4.26857620e-01 9.34563160e-01
5.13893843e-01 4.01447475e-01 3.62524688e-01 -1.26087594e+00
2.66118884e-01 -2.31789798e-01 -9.37511086e-01 3.24862264e-02
-4.26414788e-01 5.28294623e-01 1.13363135e+00 -3.54693353e-01
-2.78608054e-01 -1.73768014e-01 -5.95964730e-01 -6.16291881e-01
1.09810069e-01 1.51069075e-01 -1.08531989e-01 -6.28822148e-01
1.09682453e+00 2.67062336e-01 -3.79473150e-01 -2.37349302e-01
4.25203741e-01 3.30948621e-01 8.54519904e-01 -1.40859768e-01
1.56229150e+00 7.04456508e-01 5.71747959e-01 -1.02133751e+00
-6.74181938e-01 3.16175729e-01 -1.62005827e-01 5.32117546e-01
1.02670062e+00 -8.25529099e-01 -1.22800243e+00 6.70062840e-01
-1.14860976e+00 -1.37397975e-01 -3.47286224e-01 5.28748333e-01
-8.58210474e-02 2.45542869e-01 -7.82005727e-01 -8.88729095e-01
-5.21576941e-01 -1.04694712e+00 3.66185218e-01 3.59613985e-01
5.27900577e-01 -8.21014583e-01 -1.19889993e-02 2.91080385e-01
4.66541015e-03 5.77787578e-01 9.00364935e-01 -8.19272101e-01
-7.63082922e-01 -6.05536819e-01 6.99928328e-02 4.42517579e-01
-2.62172520e-01 -1.22344077e-01 -1.45344436e+00 -4.45362121e-01
2.37403616e-01 -4.53146875e-01 1.24900937e+00 4.27959524e-02
1.04711282e+00 -5.59515953e-01 -1.38199165e-01 9.96779382e-01
1.48852134e+00 4.10039246e-01 9.07570899e-01 -8.93102214e-03
9.67733145e-01 -2.13929042e-01 -3.86296272e-01 6.70778081e-02
5.43295965e-03 2.34589741e-01 8.98886502e-01 2.96682090e-01
3.64694118e-01 -6.34796083e-01 7.91538298e-01 6.27608001e-01
1.02571562e-01 -7.77995437e-02 -6.25205755e-01 -3.10268700e-02
-1.35766661e+00 -1.13217580e+00 -5.14159441e-01 2.04506636e+00
8.48254204e-01 4.55397576e-01 -2.43385598e-01 7.34648556e-02
2.89033413e-01 1.83156077e-02 -7.17267990e-01 -7.17472255e-01
-2.21923769e-01 8.03888440e-01 1.31051731e+00 4.54511523e-01
-1.11028957e+00 1.31008708e+00 6.77484608e+00 1.12907553e+00
-9.12577212e-01 4.43845272e-01 1.02023698e-01 -2.27900878e-01
-5.78320801e-01 3.90277833e-01 -9.40338135e-01 2.93579668e-01
1.45242822e+00 -8.37648585e-02 8.23049843e-01 3.29438120e-01
-6.10663116e-01 1.28464296e-01 -1.12954319e+00 9.44643974e-01
-2.97094285e-01 -1.70260453e+00 1.42913610e-01 6.22004569e-01
7.25523293e-01 3.80220890e-01 4.61880952e-01 3.90671551e-01
5.46844542e-01 -1.23396766e+00 5.49894035e-01 3.62836838e-01
9.73999023e-01 -1.48499036e+00 7.19926417e-01 5.00996530e-01
-5.44941545e-01 -2.14511245e-01 -9.47388709e-01 2.08363198e-02
-5.08774936e-01 3.41715604e-01 -2.54571855e-01 6.08508348e-01
4.25603926e-01 -8.96810293e-02 -4.57124621e-01 3.79688233e-01
-5.28191686e-01 1.01942527e+00 -3.78105611e-01 -2.39228889e-01
5.96632302e-01 -2.43376672e-01 7.09179878e-01 6.41726494e-01
-2.39356160e-01 3.25200289e-01 -4.75916922e-01 1.33387113e+00
-6.60038173e-01 -5.42322755e-01 -1.00187159e+00 -4.60765123e-01
6.49401784e-01 8.81016493e-01 -2.29685038e-01 -1.29961193e-01
1.98972076e-01 1.01935017e+00 3.57551314e-02 2.46125534e-01
-8.30438554e-01 -1.21776581e+00 6.84119701e-01 -6.14520550e-01
2.94105291e-01 -3.71004753e-02 -2.49055743e-01 -1.20561922e+00
-4.73023027e-01 -7.30631173e-01 2.08103240e-01 -1.36247739e-01
-1.00207961e+00 4.83439684e-01 -4.94803578e-01 -6.22236192e-01
3.37864041e-01 -6.93292737e-01 -6.44025266e-01 7.81846941e-01
-1.30674231e+00 -6.28494620e-01 5.36101580e-01 7.77956128e-01
-7.84521937e-01 -3.19803238e-01 1.21387804e+00 7.26137310e-02
-1.08124959e+00 1.17248654e+00 6.12961769e-01 6.73723519e-01
-2.75072176e-03 -1.08602762e+00 7.00194061e-01 1.18376696e+00
4.60723519e-01 1.19167924e+00 9.05243933e-01 -4.99505579e-01
-1.87781346e+00 -1.06223154e+00 6.38199210e-01 -7.94975042e-01
7.85987735e-01 -7.54063845e-01 -5.95761478e-01 5.60441017e-01
4.76083383e-02 1.11840345e-01 1.01622891e+00 -2.93482870e-01
-9.93394673e-01 -1.62311032e-01 -1.43555927e+00 7.28995740e-01
7.31948316e-01 -1.36502171e+00 -6.38218522e-01 2.77262270e-01
1.14387810e+00 -2.91217715e-01 -6.83585763e-01 -2.07370315e-02
4.61115122e-01 -6.29911125e-01 9.24895406e-01 -8.14785361e-01
6.69497922e-02 -3.67151909e-02 -3.28703046e-01 -7.10759044e-01
-3.31448354e-02 -1.29078233e+00 -4.96680915e-01 5.16284764e-01
3.22275609e-01 -9.26847756e-01 9.08480585e-01 3.06792796e-01
1.99464276e-01 -2.42115986e-02 -1.60124457e+00 -9.89821911e-01
7.44765222e-01 -5.77831745e-01 6.17747128e-01 7.84010470e-01
1.50937840e-01 2.56337643e-01 -4.28964913e-01 6.20402336e-01
1.15147364e+00 -4.75755543e-01 6.30387187e-01 -9.21225190e-01
-2.18028128e-01 -1.80173501e-01 -6.41296387e-01 -5.76432586e-01
-7.27482811e-02 -1.46495581e+00 -2.30866224e-01 -6.64620221e-01
1.30077228e-01 4.63935025e-02 -8.93225372e-01 5.36509991e-01
2.66701039e-02 7.34552264e-01 1.83133379e-01 -2.23769527e-02
-4.17971879e-01 8.68380964e-01 1.01849031e+00 -2.00335339e-01
-5.61506115e-02 -2.74261888e-02 -7.80998230e-01 3.14535677e-01
7.92156935e-01 -1.28515148e+00 -3.32356691e-01 -9.13390599e-04
5.05318105e-01 2.07682058e-01 7.21399546e-01 -1.16902518e+00
3.83875430e-01 1.87531695e-01 -4.96684276e-02 -4.50822055e-01
2.40758643e-01 -6.35362566e-01 2.30242476e-01 1.34094703e+00
-6.71127364e-02 -3.71600121e-01 2.51483828e-01 8.09043169e-01
3.50986123e-01 -4.23415750e-01 8.73495877e-01 2.16770887e-01
-2.26437226e-01 5.88934898e-01 -5.11006773e-01 -4.75958027e-02
8.46929073e-01 -5.87804317e-02 -9.14858520e-01 2.21670970e-01
-4.95696336e-01 -8.92829075e-02 2.09701508e-01 -2.74220407e-02
7.46144652e-01 -1.12409532e+00 -2.24478692e-01 5.52218795e-01
1.31952465e-01 -2.81839252e-01 2.92958051e-01 5.21752119e-01
-7.19232321e-01 8.14900339e-01 -2.16001213e-01 -1.03205003e-01
-7.97402442e-01 6.58426106e-01 7.22568095e-01 5.23518175e-02
-7.37681389e-01 1.07699835e+00 1.71248198e-01 -4.86018717e-01
1.79535225e-01 -2.83404738e-01 1.83462366e-01 -4.72953260e-01
6.68753862e-01 2.82894611e-01 -2.82558706e-03 -3.50145817e-01
-2.90877998e-01 -5.51462136e-02 -2.26117998e-01 -1.75839111e-01
9.52823579e-01 5.44355273e-01 -1.59981638e-01 1.17880434e-01
1.29519355e+00 -2.49436125e-02 -1.04291928e+00 -3.60483289e-01
-1.86155230e-01 7.22102225e-02 3.74284983e-01 -6.31445587e-01
-9.23123956e-01 9.78795469e-01 6.59057379e-01 1.73080876e-01
6.76777303e-01 -1.54752940e-01 1.11969554e+00 1.40861714e+00
6.85998619e-01 -8.64868999e-01 -1.56033918e-01 6.21829033e-01
-8.86100307e-02 -6.86204731e-01 -2.86576122e-01 1.08255267e-01
-2.38048539e-01 1.00364602e+00 2.63345629e-01 -5.52936316e-01
7.88657486e-01 6.31763116e-02 -1.55213758e-01 -4.73918647e-01
-6.58396721e-01 5.76755889e-02 -1.36748075e-01 4.93977010e-01
-5.74956596e-01 1.56175971e-01 3.86401303e-02 6.59195960e-01
-4.84229535e-01 -2.12344959e-01 8.79044414e-01 6.88601315e-01
-4.56331372e-01 -1.16648436e+00 -2.33072713e-01 2.26162374e-01
-8.77728224e-01 -4.26156163e-01 -3.77564311e-01 5.07690430e-01
2.76613712e-01 1.00945306e+00 -3.31006646e-01 -1.04554498e+00
6.43400177e-02 2.23573279e-02 6.04166448e-01 -4.84280407e-01
-6.98338091e-01 -3.00224096e-01 -4.13220257e-01 -6.52897120e-01
1.60555094e-01 -2.62039751e-01 -1.66005635e+00 -9.53439355e-01
-9.07226324e-01 3.88214797e-01 5.72258353e-01 1.00758314e+00
9.01759118e-02 5.49187660e-01 8.29287708e-01 7.59819373e-02
-9.32157278e-01 -3.89812589e-01 -1.02101862e+00 -2.09295034e-01
7.64260709e-01 -2.31044337e-01 -4.11416471e-01 -6.26611650e-01] | [5.574103355407715, 5.159966468811035] |
9b0e7240-c2a1-4b9b-bb70-f8847bf9d285 | augmenting-librispeech-with-french | 1802.03142 | null | http://arxiv.org/abs/1802.03142v1 | http://arxiv.org/pdf/1802.03142v1.pdf | Augmenting Librispeech with French Translations: A Multimodal Corpus for Direct Speech Translation Evaluation | Recent works in spoken language translation (SLT) have attempted to build
end-to-end speech-to-text translation without using source language
transcription during learning or decoding. However, while large quantities of
parallel texts (such as Europarl, OpenSubtitles) are available for training
machine translation systems, there are no large (100h) and open source parallel
corpora that include speech in a source language aligned to text in a target
language. This paper tries to fill this gap by augmenting an existing
(monolingual) corpus: LibriSpeech. This corpus, used for automatic speech
recognition, is derived from read audiobooks from the LibriVox project, and has
been carefully segmented and aligned. After gathering French e-books
corresponding to the English audio-books from LibriSpeech, we align speech
segments at the sentence level with their respective translations and obtain
236h of usable parallel data. This paper presents the details of the processing
as well as a manual evaluation conducted on a small subset of the corpus. This
evaluation shows that the automatic alignments scores are reasonably correlated
with the human judgments of the bilingual alignment quality. We believe that
this corpus (which is made available online) is useful for replicable
experiments in direct speech translation or more general spoken language
translation experiments. | ['Laurent Besacier', 'Ali Can Kocabiyikoglu', 'Olivier Kraif'] | 2018-02-09 | augmenting-librispeech-with-french-1 | https://aclanthology.org/L18-1001 | https://aclanthology.org/L18-1001.pdf | lrec-2018-5 | ['speech-to-text-translation'] | ['natural-language-processing'] | [ 1.58761546e-01 2.69666556e-02 -7.42387921e-02 -5.23119748e-01
-1.78512192e+00 -7.66979992e-01 5.58569193e-01 -3.16879936e-02
-4.94529009e-01 9.32113588e-01 5.86721003e-01 -6.44511521e-01
3.03579926e-01 -2.61767119e-01 -6.20438576e-01 -3.91135812e-01
2.08860070e-01 9.18106616e-01 -1.16948346e-02 -5.72794974e-01
1.65740818e-01 1.58289537e-01 -9.13307369e-01 6.29504025e-01
9.10734653e-01 2.04953834e-01 5.61460435e-01 6.99556649e-01
-1.81204617e-01 2.10324422e-01 -7.52700567e-01 -5.21598756e-01
2.69948661e-01 -1.00266695e+00 -1.07494295e+00 4.10479344e-02
3.45356286e-01 -7.36171082e-02 -9.43356305e-02 8.22531164e-01
7.45976985e-01 -2.54349679e-01 3.82451087e-01 -7.03551948e-01
-5.24750233e-01 9.56722260e-01 6.70427550e-03 4.26960409e-01
6.63774848e-01 -5.36529161e-02 9.91447091e-01 -1.13595366e+00
7.69958556e-01 1.18657041e+00 3.76302928e-01 4.12187248e-01
-1.15318263e+00 -5.66868722e-01 -4.48146313e-01 -1.34058460e-03
-1.54558039e+00 -1.15088940e+00 2.74130017e-01 -1.86437607e-01
1.34257770e+00 4.11686152e-01 5.76105773e-01 1.27775216e+00
1.82693437e-01 5.92302322e-01 1.22397673e+00 -1.12220275e+00
-1.27761252e-02 4.33871806e-01 -1.33608416e-01 2.95423061e-01
-4.45318788e-01 9.53084305e-02 -8.29350650e-01 -1.31251529e-01
4.30323333e-01 -9.61316288e-01 -3.69324446e-01 2.60434508e-01
-1.74655390e+00 5.49707711e-01 -1.91273227e-01 7.35989571e-01
-3.22342813e-01 -5.21455169e-01 6.47010148e-01 1.02751970e+00
6.36440694e-01 3.87113094e-01 -4.93079782e-01 -6.00501537e-01
-1.06370497e+00 7.55716488e-02 1.10294914e+00 1.35337198e+00
5.47530293e-01 7.92197511e-02 6.54982328e-02 1.26474142e+00
2.29693338e-01 1.04125774e+00 7.68937707e-01 -6.57916725e-01
1.08459866e+00 -4.73258048e-02 4.34519872e-02 -5.42537630e-01
3.03431898e-02 1.27629610e-02 -1.87389076e-01 -2.99068034e-01
4.13548410e-01 -3.95038188e-01 -5.96916199e-01 1.49948561e+00
1.03489056e-01 -6.15775585e-01 6.29052043e-01 7.77169228e-01
5.86610973e-01 1.17587578e+00 -3.93285900e-01 -6.24245405e-01
1.16665483e+00 -1.06084168e+00 -7.67791152e-01 -3.34650636e-01
5.75600863e-01 -1.65265274e+00 1.27257061e+00 2.86188632e-01
-1.35942256e+00 -5.22266328e-01 -1.02359557e+00 4.66415472e-02
4.33691852e-02 1.95575833e-01 -9.91925895e-02 7.72626877e-01
-1.32922316e+00 1.36067405e-01 -7.75738776e-01 -8.08848917e-01
-4.65998024e-01 2.88528532e-01 -5.25440335e-01 5.95169980e-03
-1.37664211e+00 1.15460038e+00 1.10220365e-01 -1.31639332e-01
-5.95693946e-01 2.02405810e-01 -6.85303390e-01 -3.10208321e-01
-2.35282294e-02 -1.68003842e-01 1.74977958e+00 -1.20427597e+00
-1.89846683e+00 9.58316386e-01 -3.88155669e-01 -3.20886344e-01
2.75489867e-01 8.85604545e-02 -7.25145340e-01 1.33561030e-01
1.93055630e-01 5.01097977e-01 5.84070742e-01 -7.43844151e-01
-6.83669090e-01 -3.42657775e-01 -4.18958753e-01 5.42015791e-01
-2.03319982e-01 8.53869200e-01 -3.99526507e-01 -6.50774181e-01
1.01692609e-01 -1.07806110e+00 6.95569590e-02 -7.28136718e-01
-1.37441427e-01 -1.90359447e-02 3.23490322e-01 -1.21216547e+00
1.17979848e+00 -2.13069463e+00 2.49446601e-01 1.50930762e-01
-5.44859886e-01 2.09743083e-01 -4.27786618e-01 1.08511674e+00
6.60586655e-02 1.78363025e-02 -6.37655258e-02 -3.08124095e-01
-7.74241015e-02 3.21344137e-01 -3.86489630e-01 4.72825706e-01
-2.00668126e-01 7.49016285e-01 -6.65869236e-01 -5.51446915e-01
-8.78675934e-03 1.81743041e-01 -1.31220803e-01 2.27028251e-01
8.13060030e-02 5.54430008e-01 -1.39346153e-01 4.44693893e-01
-1.33549888e-03 5.91649294e-01 1.09881066e-01 3.59279037e-01
-4.91854817e-01 1.22663319e+00 -5.87755084e-01 1.76417577e+00
-6.76293254e-01 7.68165946e-01 2.08615676e-01 -7.07760513e-01
1.15693486e+00 9.91052926e-01 2.01240867e-01 -7.58228958e-01
-3.33557874e-02 8.51095438e-01 2.85186708e-01 -3.83540928e-01
6.37142479e-01 -1.95656672e-01 -1.70310557e-01 6.86415195e-01
3.08670700e-01 -5.80394566e-01 4.20760751e-01 -9.09534767e-02
7.50223041e-01 -2.68545039e-02 3.46567184e-01 -4.07615036e-01
5.43619335e-01 5.17938793e-01 1.85609654e-01 1.51122704e-01
-9.03352723e-02 7.40916371e-01 3.00762355e-02 -5.91163896e-03
-1.46624875e+00 -8.53139818e-01 -1.41206652e-01 9.60018039e-01
-4.74970847e-01 -7.09294319e-01 -1.10651350e+00 -2.23413527e-01
-8.58773589e-01 8.06600869e-01 2.73077816e-01 4.01504874e-01
-9.37961936e-01 -3.01617384e-01 8.55289161e-01 1.00923456e-01
7.75568113e-02 -1.23617363e+00 2.63654083e-01 4.68144357e-01
-8.14141273e-01 -1.22393537e+00 -9.98851836e-01 6.11964390e-02
-6.08707011e-01 -6.20126426e-01 -8.13768208e-01 -1.27855992e+00
3.34460944e-01 2.41078719e-01 9.67145920e-01 -3.73302907e-01
4.94178534e-01 1.79397851e-01 -5.73274970e-01 -3.34538907e-01
-1.45956731e+00 5.56383908e-01 3.43695879e-01 -3.20554703e-01
6.57614768e-01 -4.24435049e-01 1.67944402e-01 5.91639280e-01
-4.76993620e-01 1.68156251e-01 5.97883582e-01 7.83455431e-01
4.96263176e-01 -6.00719571e-01 7.07677305e-01 -3.94402444e-01
8.82355034e-01 -1.20410666e-01 -5.08929610e-01 2.33715370e-01
-3.96110207e-01 -1.93635121e-01 6.79811120e-01 -2.22571954e-01
-8.46723795e-01 -3.13422084e-02 -7.75689244e-01 3.40048447e-02
-3.12691152e-01 6.44371271e-01 -1.76896200e-01 1.28058642e-01
8.79240334e-01 4.30499613e-01 7.50379562e-02 -3.47249627e-01
1.23030581e-01 1.63158023e+00 6.24735475e-01 -5.48016012e-01
6.49789512e-01 -4.91232365e-01 -8.06363761e-01 -1.14920282e+00
-1.82388514e-01 -5.09036541e-01 -7.95059800e-01 -8.57994258e-02
6.71487153e-01 -1.02032232e+00 6.77946284e-02 1.77189723e-01
-1.24503636e+00 -4.51011419e-01 -1.31238088e-01 8.94233584e-01
-8.73564899e-01 1.89079732e-01 -7.20136225e-01 -4.80428219e-01
-4.67851013e-01 -1.28539658e+00 1.10212231e+00 -3.09151679e-01
-7.44972050e-01 -7.59030282e-01 4.32159811e-01 5.66392839e-01
3.41822118e-01 -6.15904331e-01 7.66276658e-01 -8.79697919e-01
-2.46415883e-01 -2.39658773e-01 4.18658108e-01 6.05167210e-01
2.02152491e-01 -3.07324007e-02 -6.76823437e-01 -4.37455684e-01
1.66083351e-01 -4.09753084e-01 3.00494190e-02 1.69299950e-03
-1.49604067e-01 -5.02655625e-01 1.19481206e-01 2.37781238e-02
8.34059596e-01 4.23723549e-01 5.44007778e-01 2.22602725e-01
2.61804551e-01 7.07549453e-01 6.41738534e-01 -7.72505552e-02
4.74617600e-01 1.01042295e+00 -4.95610178e-01 1.30326957e-01
-2.31783763e-01 -3.11842293e-01 1.19803655e+00 2.17597246e+00
1.86922342e-01 -2.54760325e-01 -1.16638339e+00 7.61775196e-01
-1.45572126e+00 -7.57849276e-01 -2.63118118e-01 2.25945807e+00
1.24053919e+00 -1.12913780e-01 2.30960310e-01 5.81990406e-02
7.56656468e-01 -3.42152677e-02 1.19013883e-01 -7.58195639e-01
-1.54266879e-01 2.43864790e-01 3.26996863e-01 1.00326574e+00
-5.79442680e-01 1.23323679e+00 6.98180151e+00 8.47387552e-01
-1.17137825e+00 4.03729200e-01 3.14295053e-01 -6.83330814e-04
-2.58953899e-01 7.60246590e-02 -7.94576108e-01 2.47495919e-01
1.82939446e+00 -5.66967189e-01 7.64388382e-01 4.84895885e-01
5.06862760e-01 8.03519338e-02 -1.28849769e+00 8.55439186e-01
1.85770601e-01 -1.02150202e+00 -1.47759214e-01 -3.86379212e-02
7.91052818e-01 6.49110794e-01 -3.80492955e-01 3.93110603e-01
-1.52020771e-02 -7.65156686e-01 9.79291737e-01 -9.78946686e-02
1.06008196e+00 -7.47914493e-01 7.37751245e-01 6.91525996e-01
-9.33480740e-01 4.99524921e-01 -3.52341861e-01 -4.32805270e-02
3.44151497e-01 1.11918142e-02 -1.26292408e+00 6.46935642e-01
2.78718323e-01 4.42591161e-01 -3.83198678e-01 5.97173512e-01
-3.12995702e-01 1.06935263e+00 -2.98330247e-01 -2.05756441e-01
3.22500437e-01 -4.29330140e-01 7.79364944e-01 1.35200083e+00
6.18222654e-01 -1.69973984e-01 3.19063753e-01 3.34677994e-01
5.13036959e-02 8.74030530e-01 -7.02640712e-01 -3.60963225e-01
4.05407786e-01 8.10916007e-01 -5.02967358e-01 -3.00210446e-01
-6.90118313e-01 1.11208022e+00 -6.48443699e-02 3.92934859e-01
-3.08598727e-01 -4.94804144e-01 3.87108088e-01 7.18174577e-02
-1.26886398e-01 -5.73342323e-01 -1.68335378e-01 -1.10127735e+00
2.28394583e-01 -1.64228714e+00 -2.44972929e-01 -7.25455165e-01
-1.18689513e+00 1.31916857e+00 1.12709757e-02 -1.33379555e+00
-9.32799101e-01 -2.59157568e-01 -1.87430337e-01 1.38269913e+00
-8.02959263e-01 -9.94561076e-01 3.94024044e-01 4.80200320e-01
1.09864390e+00 -6.56157255e-01 1.22460699e+00 5.15570819e-01
-2.06464857e-01 4.86753821e-01 1.99061438e-01 1.18433908e-01
1.10195172e+00 -9.72219408e-01 8.96659970e-01 8.34591687e-01
6.33156776e-01 6.06233418e-01 8.08373690e-01 -6.13098621e-01
-1.19468975e+00 -7.65852332e-01 1.72475553e+00 -5.13773263e-01
7.71566391e-01 -5.23113668e-01 -7.79218674e-01 8.00633550e-01
9.21256840e-01 -6.96897745e-01 6.74661696e-01 -1.18120173e-02
2.31110498e-01 -1.16022468e-01 -6.91267252e-01 6.98161125e-01
8.46605599e-01 -8.41731966e-01 -8.87057483e-01 6.22149348e-01
7.07933426e-01 -6.06923223e-01 -8.57914627e-01 -4.72293086e-02
3.77522856e-01 -6.12743318e-01 3.10693026e-01 -2.06290379e-01
2.75465220e-01 -2.05627963e-01 -4.50284988e-01 -1.73594201e+00
1.65996730e-01 -1.04061973e+00 8.11635911e-01 1.36076224e+00
8.88538241e-01 -6.32402003e-01 3.00117433e-01 1.86363503e-01
-5.60060382e-01 -3.68411511e-01 -1.16737330e+00 -8.71018291e-01
3.70571166e-01 -3.34451377e-01 3.52391869e-01 8.17094803e-01
5.97771347e-01 9.64266241e-01 -2.47301862e-01 1.03778793e-02
3.03475298e-02 -7.78521299e-02 9.00292635e-01 -7.00565517e-01
-2.81427681e-01 -2.37640753e-01 -1.21022359e-01 -1.11093986e+00
2.41926059e-01 -1.14198887e+00 4.23005611e-01 -1.43136549e+00
-1.90875113e-01 -3.10938954e-01 4.35225457e-01 5.04666805e-01
1.41888395e-01 2.54248619e-01 1.01851657e-01 4.84907269e-01
8.61438438e-02 5.03661633e-01 1.00021040e+00 -5.38499542e-02
-3.46544594e-01 1.43448755e-01 -2.85901159e-01 3.31638396e-01
9.52308714e-01 -6.64560497e-01 -2.78789163e-01 -5.49163222e-01
-1.35669291e-01 5.51540196e-01 -3.97087276e-01 -8.62188160e-01
7.32855946e-02 -4.65258360e-02 -1.37697160e-01 -4.84536618e-01
1.29649252e-01 -5.77176630e-01 1.94090351e-01 1.28264099e-01
-3.16962093e-01 5.90659499e-01 2.44463310e-01 -2.31220558e-01
-7.27017641e-01 -4.27518278e-01 6.02797449e-01 -7.73746893e-02
-2.47880027e-01 -1.30689934e-01 -9.55184340e-01 5.23934364e-02
6.60303771e-01 -9.94025916e-02 -5.83460890e-02 -6.53871953e-01
-4.55091029e-01 -1.08000807e-01 5.63897967e-01 6.28482342e-01
2.72266120e-01 -1.24785435e+00 -1.17062235e+00 5.11531770e-01
1.64751202e-01 -5.23066103e-01 -4.18083996e-01 9.36898708e-01
-7.09415615e-01 8.03178787e-01 -3.35781515e-01 -5.94684660e-01
-1.40292776e+00 1.73466697e-01 6.72811717e-02 4.78229076e-02
-5.97962618e-01 4.15605217e-01 -3.58804971e-01 -7.09938824e-01
8.74128938e-02 -1.91460893e-01 2.14090496e-01 -2.51684040e-01
4.51560795e-01 6.80917650e-02 4.63516414e-01 -1.07742286e+00
-1.90788001e-01 1.91664189e-01 -1.93985831e-02 -1.07216227e+00
9.77075577e-01 -5.63779593e-01 -2.62421876e-01 9.07716811e-01
1.05279374e+00 6.24446571e-01 -4.62962806e-01 -1.72678113e-01
1.82213038e-01 -2.21909061e-01 -3.42248887e-01 -7.08613932e-01
-3.89217287e-01 8.18496525e-01 2.45280579e-01 2.97870021e-02
9.71307218e-01 -8.33096132e-02 1.01610017e+00 7.04547942e-01
6.96666181e-01 -1.18810344e+00 -3.82135659e-01 1.02625787e+00
1.17602158e+00 -1.02889383e+00 -4.48307216e-01 -2.74723440e-01
-7.21551657e-01 1.04097855e+00 1.65010706e-01 2.06226826e-01
1.77292511e-01 2.95174122e-01 8.37720752e-01 3.50159645e-01
-6.81396246e-01 6.55520633e-02 1.36909381e-01 5.21137178e-01
8.36804330e-01 5.55474758e-02 -8.50569129e-01 2.50562042e-01
-8.89104605e-01 -2.81508297e-01 5.69347560e-01 7.35334992e-01
-4.29247797e-01 -1.79236472e+00 -6.57944322e-01 -6.15475839e-03
-6.23972476e-01 -4.65159029e-01 -7.46333301e-01 6.67108119e-01
-3.84173870e-01 1.25694227e+00 -1.57475814e-01 -2.57610202e-01
4.16406602e-01 4.49364930e-01 5.16107440e-01 -9.20489788e-01
-7.51451313e-01 6.71051264e-01 6.44944787e-01 -2.51763046e-01
-3.64249766e-01 -7.71895051e-01 -1.03981054e+00 -2.99610555e-01
-2.17086375e-01 7.51460552e-01 9.68995690e-01 1.03910780e+00
-1.48893595e-01 1.52501345e-01 6.28383696e-01 -8.98415208e-01
-5.74139118e-01 -1.43690372e+00 -4.26850259e-01 -1.37244418e-01
1.38231307e-01 2.64646739e-01 -4.20237780e-02 3.40922654e-01] | [14.40749740600586, 7.165937900543213] |
779922d8-fb7f-4a30-bb09-7edb842be626 | secure-and-robust-mimo-transceiver-for | 2012.05829 | null | https://arxiv.org/abs/2012.05829v1 | https://arxiv.org/pdf/2012.05829v1.pdf | Secure and Robust MIMO Transceiver for Multicast Mission Critical Communications | Mission-critical communications (MCC) involve all communications between people in charge of the safety of the civil society. MCC have unique requirements that include improved reliability, security and group communication support. In this paper, we propose a secure and robust Multiple-Input-Multiple-Output (MIMO) transceiver, designed for multiple Base Stations supporting multicast MCC in presence of multiple eavesdroppers. We formulate minimization problems with the Sum-Mean-Square-Error (SMSE) at legitimate users as an objective function, and a lower bound for the MSE at eavesdroppers as a constraint. Security is achieved thanks to physical layer security mechanisms, namely MIMO beamforming and Artificial Noise (AN). Reliability is achieved by designing a system which is robust to Channel State Information errors. They can be of known distribution or norm-bounded. In the former case, a coordinate descent algorithm is proposed to solve the problem. In the latter case, we propose an worst-case based iterative algorithm. Numerical results at physical layer and system level reveal the crucial role of robust designs for reliable MCC. We show the interest of both robust design and AN to improve the security gap. We show that full BS cooperation in preferred for highly secured and reliable MCC but dynamic clustering allows to trade-off security and reliability against capacity. | ['Marceau Coupechoux', 'Deepa Jagyasi'] | 2020-12-10 | null | null | null | null | ['robust-design'] | ['miscellaneous'] | [ 3.50986749e-01 3.16209316e-01 2.86579937e-01 8.71082693e-02
-8.69330883e-01 -6.63745046e-01 7.81830177e-02 4.83671017e-02
-5.33423722e-01 9.64009166e-01 -1.15899049e-01 -6.83902025e-01
-9.83734667e-01 -5.12973785e-01 -3.22376162e-01 -1.53221023e+00
-5.28149962e-01 -4.77567583e-01 -8.41316432e-02 -2.43428260e-01
6.53957576e-02 7.25810826e-01 -6.60914361e-01 -1.45919114e-01
5.73227167e-01 1.48372447e+00 1.88014835e-01 1.05872810e+00
6.52060986e-01 5.95047593e-01 -7.09076107e-01 -3.25542182e-01
4.16976333e-01 -2.22367808e-01 -2.27784485e-01 2.77373433e-01
-7.26042628e-01 2.37686351e-01 -4.01086748e-01 8.74786258e-01
7.80902386e-01 -1.98616475e-01 7.06280828e-01 -1.61903632e+00
-2.38844052e-01 4.38781440e-01 -6.58577442e-01 -3.64677571e-02
8.00678805e-02 -2.70264328e-01 6.48987293e-01 -6.86605394e-01
2.80377343e-02 6.45419776e-01 5.34188271e-01 5.11759758e-01
-7.14130163e-01 -5.38639486e-01 -1.84935629e-01 2.26914864e-02
-1.68071640e+00 -7.12322593e-01 3.15798670e-01 -6.52440218e-03
2.19998375e-01 9.66169119e-01 1.74435049e-01 2.89765507e-01
4.90148664e-01 9.11495090e-02 6.79698527e-01 -6.34203255e-01
4.65923727e-01 1.93852529e-01 -3.16098422e-01 4.48001891e-01
7.15795338e-01 -6.70709535e-02 -6.52108565e-02 -4.14834112e-01
2.47726768e-01 -3.64026725e-01 -6.08583748e-01 -2.70992696e-01
-1.12222052e+00 6.98231280e-01 2.99695581e-01 5.00742614e-01
-5.70170760e-01 4.01858091e-01 -4.94706221e-02 6.77016318e-01
7.77490884e-02 3.31427038e-01 -2.23018050e-01 3.77606779e-01
-6.46901190e-01 -3.17251235e-01 9.17135775e-01 1.20930684e+00
-2.97355533e-01 2.43447140e-01 -1.21244520e-01 3.51409078e-01
5.30969322e-01 9.63886440e-01 -3.88827890e-01 -8.75078797e-01
7.47803390e-01 -1.28881142e-01 1.50911018e-01 -1.17396224e+00
-7.61311352e-01 -1.26500118e+00 -1.29910970e+00 1.19861111e-01
-4.87892888e-02 -7.54797518e-01 -1.10427782e-01 1.65852094e+00
2.67448843e-01 -3.39137286e-01 6.77240610e-01 7.12589145e-01
1.44068822e-01 5.96324146e-01 -6.29861414e-01 -9.89465356e-01
9.26285505e-01 -2.33205900e-01 -8.95053685e-01 1.16546117e-02
5.47935843e-01 -7.47182429e-01 -1.20434582e-01 5.51701248e-01
-1.34262323e+00 2.85306185e-01 -1.37053096e+00 9.73872483e-01
2.16969460e-01 3.78726304e-01 2.10052207e-02 1.30530250e+00
-1.18552673e+00 -8.74583498e-02 -6.25441313e-01 -2.28304602e-02
6.31919429e-02 8.67735624e-01 -2.52563208e-01 -3.17082666e-02
-7.28645384e-01 7.36644864e-01 7.70377070e-02 6.18680477e-01
-3.35079938e-01 -2.17743576e-01 -7.27324128e-01 -3.03618493e-03
4.10254061e-01 -6.23300076e-01 7.73332357e-01 -4.13939416e-01
-1.05417299e+00 7.65162930e-02 6.92344606e-02 -4.25442696e-01
4.65782583e-01 3.07464898e-01 -4.80835021e-01 4.70875978e-01
-1.47300661e-01 -3.56618077e-01 6.84516191e-01 -1.64893901e+00
-7.54621148e-01 -2.12146401e-01 3.16403620e-02 -2.32647896e-01
-4.10045832e-01 1.98715195e-01 1.20632574e-02 -4.65692937e-01
4.19726968e-01 -1.15995395e+00 -4.68623847e-01 -1.90384299e-01
-7.84194231e-01 7.03876674e-01 7.25633323e-01 -7.08679080e-01
1.40074289e+00 -2.18314409e+00 3.38439554e-01 7.64478266e-01
8.18493366e-02 3.51021998e-02 9.94753391e-02 8.62084746e-01
6.18957430e-02 2.42966324e-01 -2.61698067e-01 -4.09140885e-01
-3.25442135e-01 -3.08555039e-03 1.90788701e-01 1.14428759e+00
-6.40589325e-03 1.72655359e-01 -4.10470366e-01 -1.68585956e-01
-1.18193388e-01 4.07373250e-01 -3.55927467e-01 3.15316319e-02
6.89623356e-01 7.28214264e-01 -7.08766162e-01 5.62477171e-01
8.28595936e-01 1.08910084e-01 1.52720898e-01 -2.34714746e-01
-3.24116558e-01 -5.69021523e-01 -1.51521158e+00 8.61675918e-01
-6.80324256e-01 4.47836965e-01 1.31130290e+00 -1.14515865e+00
3.99895042e-01 7.79905319e-01 5.69582701e-01 -2.51025379e-01
5.38054824e-01 3.41920316e-01 1.78372294e-01 -3.62469703e-01
1.41179934e-01 -2.20277473e-01 -4.57602262e-01 3.85262311e-01
-4.71920699e-01 6.76365197e-02 -2.46515021e-01 5.46970785e-01
1.25847054e+00 -1.07625985e+00 2.97498107e-01 -4.80220288e-01
1.01001370e+00 -5.56673110e-01 5.35382509e-01 6.68965161e-01
1.70514420e-01 2.70241946e-01 4.36276168e-01 5.87865412e-01
-7.25856662e-01 -6.65584683e-01 -4.22243960e-02 5.02368808e-01
2.31828064e-01 5.78639563e-03 -6.00939274e-01 -2.15883538e-01
-1.59522370e-01 7.92248309e-01 -3.97966713e-01 -4.04276848e-01
-2.94625163e-01 -8.71411681e-01 5.29141188e-01 -3.67987379e-02
3.43247175e-01 2.07929745e-01 -4.48023260e-01 3.23770046e-01
-1.52311688e-02 -1.41866457e+00 -3.10589045e-01 4.08741713e-01
-2.31158331e-01 -1.09656942e+00 -5.65258503e-01 -3.97575140e-01
8.43760669e-01 6.43435538e-01 2.30453104e-01 4.30393487e-01
-1.13985434e-01 9.42955792e-01 -3.93099040e-01 -3.86587679e-01
-6.22128844e-01 -2.00993195e-01 5.38638473e-01 4.88439858e-01
-9.17169511e-01 -4.23945636e-01 -4.89580750e-01 5.09458482e-01
-8.64240944e-01 -2.47138187e-01 6.31712317e-01 5.90521395e-01
-1.04594789e-02 5.34389555e-01 9.35778797e-01 -1.34307146e-02
5.98156810e-01 -6.88277304e-01 -5.01599073e-01 4.67011899e-01
-2.78478235e-01 -2.26738855e-01 6.34092033e-01 9.51493979e-02
-8.08748782e-01 -2.15208605e-02 -1.55759245e-01 3.42267931e-01
4.71223235e-01 5.76486349e-01 -3.76762182e-01 -8.45470130e-01
2.67643601e-01 2.11270675e-01 -1.10620894e-01 4.82606664e-02
1.02037892e-01 1.13018203e+00 2.11897865e-01 -2.99023956e-01
1.29278696e+00 6.25120163e-01 8.53605747e-01 -1.19431353e+00
-6.35541499e-01 -3.80455166e-01 -5.86528420e-01 -4.98266876e-01
4.55317825e-01 -9.33484554e-01 -1.00738621e+00 8.14908072e-02
-1.06492150e+00 3.27109873e-01 2.45186627e-01 7.39927888e-01
-4.01869774e-01 4.03657109e-01 -1.84166819e-01 -1.60836792e+00
-4.28262383e-01 -8.02227437e-01 5.81887782e-01 -1.24278255e-01
4.25685138e-01 -8.48685682e-01 -7.14932978e-01 3.38741571e-01
7.50461996e-01 7.09367931e-01 4.33099777e-01 -3.48239452e-01
-4.81650561e-01 -5.85159183e-01 -1.11209497e-01 4.56157118e-01
8.77169818e-02 -2.29231983e-01 -5.47056794e-01 -9.32270706e-01
3.70313197e-01 1.22996666e-01 2.98749954e-01 4.12047088e-01
8.86250079e-01 -7.40591228e-01 -3.92449021e-01 5.11207283e-01
1.50203490e+00 2.14672744e-01 3.05152953e-01 -4.81381752e-02
2.03241110e-01 5.06152511e-01 7.65151143e-01 9.23603952e-01
3.34030807e-01 4.05744851e-01 8.70773315e-01 2.95766853e-02
6.06883943e-01 9.01250541e-01 1.75809547e-01 1.00292277e+00
-7.11090118e-02 -7.39139140e-01 -5.97603798e-01 2.09924951e-01
-1.69950712e+00 -6.78077042e-01 -5.17002881e-01 2.43045402e+00
4.23277557e-01 1.30241290e-01 -5.06309681e-02 8.49621773e-01
8.04316819e-01 -4.18743551e-01 -3.77761200e-02 -2.06605956e-01
-3.37904394e-01 -4.27507848e-01 1.14453983e+00 5.47544897e-01
-8.65623534e-01 -1.33940712e-01 4.81110239e+00 9.67787743e-01
-9.41419601e-01 3.40442389e-01 4.81397092e-01 -2.18834162e-01
-4.19307500e-02 -2.55131155e-01 -4.74880487e-01 2.24863142e-01
9.01666284e-01 -2.52916843e-01 1.26876429e-01 2.45109439e-01
6.10428870e-01 -3.50032926e-01 -5.46440005e-01 7.72841394e-01
1.01683699e-01 -1.15108514e+00 -7.49290526e-01 4.65156764e-01
6.35622084e-01 -4.85582918e-01 5.21850064e-02 -3.36303413e-01
-2.49242052e-01 -7.10515261e-01 8.34400356e-01 6.85855210e-01
7.63700008e-01 -1.10436559e+00 8.62271011e-01 6.69357240e-01
-1.28934169e+00 -8.24756026e-01 -1.53746098e-01 1.31609783e-01
3.29379231e-01 7.83734679e-01 -3.14268768e-01 1.14362335e+00
6.76179305e-02 -1.84493169e-01 -2.55456716e-02 1.38235569e+00
9.64563247e-03 3.38643581e-01 -4.94377524e-01 -3.43644023e-01
3.24744046e-01 1.68225002e-02 8.75934362e-01 1.17217982e+00
8.91445994e-01 4.98108029e-01 9.78489220e-02 -8.29940289e-02
3.10973704e-01 9.64032039e-02 -3.55526716e-01 3.75109106e-01
8.21528614e-01 1.29974079e+00 -5.44746161e-01 3.23627800e-01
-2.05002651e-01 6.91189706e-01 -4.17485267e-01 3.60349268e-01
-7.18950272e-01 -6.63511395e-01 5.39558709e-01 -4.55427356e-02
2.96519190e-01 -8.27238083e-01 -4.87212956e-01 -7.05811381e-01
1.67819858e-01 -5.77963829e-01 2.70527989e-01 -1.88516602e-01
-7.79075801e-01 4.92638558e-01 -3.62155497e-01 -1.38479340e+00
-8.36039260e-02 -2.09174842e-01 -6.24418557e-01 7.00828850e-01
-1.38925123e+00 -1.21228361e+00 8.22043568e-02 7.45668769e-01
-1.90206558e-01 -2.78105676e-01 5.04177690e-01 3.62235606e-01
-6.06773794e-01 6.40047193e-01 7.97789633e-01 -1.13316394e-01
8.11858624e-02 -7.66879141e-01 -6.46485090e-01 1.12181163e+00
-5.55649757e-01 5.29907584e-01 9.75092709e-01 -3.09976339e-01
-1.94100809e+00 -1.05588329e+00 6.40386581e-01 1.74521253e-01
5.82378864e-01 -5.24442315e-01 9.89848468e-03 -8.82779888e-04
2.03170121e-01 -7.35753253e-02 1.02893257e+00 -6.37071013e-01
2.69713670e-01 -1.78622022e-01 -1.43555677e+00 3.81825238e-01
6.32882953e-01 -1.19809015e-02 3.14926386e-01 5.14350951e-01
5.34662068e-01 -7.42384121e-02 -9.91594434e-01 4.63174522e-01
3.67303342e-01 -5.35870194e-01 9.25213575e-01 -2.16709614e-01
-4.42809343e-01 -3.95867884e-01 -7.28628814e-01 -1.27521527e+00
-2.77094513e-01 -1.07670569e+00 -2.36347914e-02 1.14203846e+00
7.88151741e-01 -4.49815243e-01 5.43341756e-01 4.23088729e-01
-4.64759544e-02 -9.06189144e-01 -1.56111205e+00 -9.97727871e-01
-1.46388471e-01 -5.88886440e-01 6.69091865e-02 4.39767718e-01
1.98066562e-01 2.22450510e-01 -6.80709720e-01 9.14324760e-01
8.88877451e-01 -5.26657224e-01 5.12808502e-01 -9.33255196e-01
-3.98139656e-01 3.51625346e-02 -3.13185453e-01 -7.06888318e-01
-2.26501897e-01 -4.73624110e-01 -7.22123235e-02 -1.59808052e+00
-3.53351951e-01 -7.11465776e-01 -3.91897745e-02 -9.15951803e-02
3.38492036e-01 1.34375662e-01 4.92739946e-01 -1.93669617e-01
-7.04595029e-01 4.97800887e-01 8.47049117e-01 2.05550373e-01
6.14931062e-02 9.60199654e-01 -9.17242348e-01 5.22810400e-01
9.50858414e-01 -6.02902770e-01 -2.99599648e-01 -1.39881581e-01
5.97049773e-01 8.32495749e-01 2.70760179e-01 -1.26913118e+00
5.92578232e-01 -1.55000076e-01 2.18681432e-02 -5.13175488e-01
7.46514916e-01 -1.59372008e+00 2.15526134e-01 9.61239934e-01
-1.24695122e-01 -2.12031424e-01 -2.81829268e-01 1.02060056e+00
2.88656980e-01 -4.71209228e-01 1.03191221e+00 4.94421661e-01
2.87383735e-01 5.95886409e-02 -8.32464814e-01 -4.66876030e-01
1.52857411e+00 -1.26104340e-01 -2.02462137e-01 -1.11193681e+00
-9.04086709e-01 6.02325916e-01 -1.82144970e-01 -1.83488548e-01
5.84340513e-01 -9.74231541e-01 -1.01076937e+00 -2.20472470e-01
-1.57685280e-01 -6.74710751e-01 1.47442833e-01 1.41921687e+00
-1.00585066e-01 6.63571477e-01 3.99559915e-01 -2.84805536e-01
-1.78024495e+00 1.75764576e-01 3.60209167e-01 4.42182831e-02
2.42890552e-01 9.60970998e-01 -4.63442296e-01 3.70644517e-02
4.17934865e-01 2.08449006e-01 -6.12882152e-02 -9.90216434e-02
5.04477680e-01 5.35308480e-01 3.19139570e-01 -8.10482740e-01
-3.80522609e-01 5.05962312e-01 4.73825157e-01 -3.28424305e-01
1.25702643e+00 -1.01545393e+00 -2.48817801e-01 -1.43777758e-01
1.19038689e+00 6.32904053e-01 -6.07928216e-01 -2.21453667e-01
8.62825960e-02 -5.60163796e-01 2.44387984e-01 -4.73435849e-01
-1.25183821e+00 4.31346178e-01 4.38018173e-01 8.18271279e-01
1.30436146e+00 -1.86783805e-01 4.49920058e-01 7.33681679e-01
7.31413782e-01 -1.05696428e+00 -1.54521659e-01 3.81789684e-01
1.01912487e+00 -7.82897830e-01 9.89987403e-02 -9.03768718e-01
-3.54134500e-01 1.18757188e+00 -2.35452726e-02 1.64327830e-01
1.00367153e+00 6.17796898e-01 -2.70999104e-01 1.26818255e-01
-6.67528987e-01 -3.03658336e-01 9.43206102e-02 5.57288289e-01
2.19615549e-01 7.03078434e-02 -7.13010550e-01 9.12434399e-01
-9.03091580e-02 -7.14436114e-01 8.61535609e-01 1.20049620e+00
-7.81265497e-01 -1.05319762e+00 -9.60422516e-01 4.21377778e-01
-8.95364702e-01 5.51846363e-02 -1.06476016e-01 5.18630266e-01
-3.22263986e-02 1.97581327e+00 -5.20573556e-01 -4.35531080e-01
3.72639745e-01 -7.46304631e-01 2.14674637e-01 -2.85177290e-01
-3.59868973e-01 9.25596952e-02 6.59749508e-01 2.99590779e-03
-2.59073526e-01 -6.47751927e-01 -1.12155581e+00 -2.77447909e-01
-8.19121003e-01 6.84140801e-01 1.16832078e+00 1.03882635e+00
2.44408026e-01 7.26433814e-01 1.52219844e+00 -2.20842898e-01
-7.77111888e-01 -4.84099001e-01 -8.47247601e-01 -6.08115137e-01
6.37253284e-01 -2.24790782e-01 -5.98839879e-01 -4.72275019e-01] | [6.15503454208374, 1.4308875799179077] |
53da3792-ac08-44c4-a4b7-573065ca76e3 | z-bert-a-a-zero-shot-pipeline-for-unknown | 2208.07084 | null | https://arxiv.org/abs/2208.07084v2 | https://arxiv.org/pdf/2208.07084v2.pdf | Z-BERT-A: a zero-shot Pipeline for Unknown Intent detection | Intent discovery is a fundamental task in NLP, and it is increasingly relevant for a variety of industrial applications (Quarteroni 2018). The main challenge resides in the need to identify from input utterances novel unseen in-tents. Herein, we propose Z-BERT-A, a two-stage method for intent discovery relying on a Transformer architecture (Vaswani et al. 2017; Devlin et al. 2018), fine-tuned with Adapters (Pfeiffer et al. 2020), initially trained for Natural Language Inference (NLI), and later applied for unknown in-tent classification in a zero-shot setting. In our evaluation, we firstly analyze the quality of the model after adaptive fine-tuning on known classes. Secondly, we evaluate its performance casting intent classification as an NLI task. Lastly, we test the zero-shot performance of the model on unseen classes, showing how Z-BERT-A can effectively perform in-tent discovery by generating intents that are semantically similar, if not equal, to the ground truth ones. Our experiments show how Z-BERT-A is outperforming a wide variety of baselines in two zero-shot settings: known intents classification and unseen intent discovery. The proposed pipeline holds the potential to be widely applied in a variety of application for customer care. It enables automated dynamic triage using a lightweight model that, unlike large language models, can be easily deployed and scaled in a wide variety of business scenarios. Especially when considering a setting with limited hardware availability and performance whereon-premise or low resource cloud deployments are imperative. Z-BERT-A, predicting novel intents from a single utterance, represents an innovative approach for intent discovery, enabling online generation of novel intents. The pipeline is available as an installable python package at the following link: https://github.com/GT4SD/zberta. | ['Matteo Manica', 'Pier Francesco Piazza', 'Dimitrios Christofidellis', 'Daniele Comi'] | 2022-08-15 | null | null | null | null | ['intent-discovery', 'intent-classification'] | ['natural-language-processing', 'natural-language-processing'] | [ 3.61762762e-01 1.91884503e-01 -6.16931878e-02 -4.64769065e-01
-1.05087757e+00 -6.48968816e-01 5.80923557e-01 8.55209753e-02
1.82373181e-03 4.65997934e-01 2.66922861e-01 -4.74968493e-01
-6.88341558e-02 -6.08208001e-01 -5.68072677e-01 -3.56491357e-01
1.40664786e-01 8.63950372e-01 -1.38689712e-01 -3.42098475e-01
1.81252792e-01 9.07883346e-02 -1.76608789e+00 5.56078315e-01
6.16915584e-01 1.21792567e+00 4.38733071e-01 6.56716883e-01
-3.87660414e-02 8.44955683e-01 -5.12631416e-01 -4.41859037e-01
1.95381626e-01 -5.30580878e-02 -9.58326757e-01 -2.34826609e-01
1.61789775e-01 -5.03513634e-01 1.55207828e-01 5.65566540e-01
6.42644942e-01 1.18762963e-01 2.89932668e-01 -1.42003989e+00
-4.96110171e-01 8.19409609e-01 9.45060700e-02 1.24436922e-01
9.31880236e-01 4.73043025e-01 1.13965130e+00 -1.04419351e+00
4.77683127e-01 1.00861359e+00 7.16745079e-01 6.20771587e-01
-1.16323018e+00 -9.11903322e-01 3.95712256e-02 3.95070106e-01
-1.25844049e+00 -6.28759444e-01 6.03853881e-01 -3.34634990e-01
1.34465849e+00 5.32856405e-01 8.17653760e-02 1.61924434e+00
1.41075596e-01 1.16949606e+00 7.99282610e-01 -4.03267264e-01
5.49434125e-01 1.81449085e-01 3.37667346e-01 3.33751827e-01
-2.15342313e-01 7.81016191e-03 -6.66695714e-01 -3.27817827e-01
7.43632838e-02 2.24186093e-01 -1.26376092e-01 1.40485182e-01
-1.33372843e+00 8.02942574e-01 4.87651080e-02 3.70769501e-01
-6.55206382e-01 4.54219012e-03 5.70638239e-01 3.12136352e-01
5.67683637e-01 6.45917475e-01 -7.37103462e-01 -6.65436804e-01
-8.76464307e-01 2.88252503e-01 1.09528387e+00 1.34132636e+00
6.79012835e-01 9.30626877e-03 -4.91967261e-01 7.45863020e-01
1.07311523e-02 2.22566113e-01 9.48010266e-01 -6.19136631e-01
3.50359887e-01 4.15450424e-01 -1.17989324e-01 -5.28329432e-01
-3.50225955e-01 -4.05413032e-01 -6.32930398e-01 -2.83035487e-01
-6.57867640e-02 -8.58755708e-02 -8.61046374e-01 1.58500671e+00
2.40724877e-01 5.94736636e-01 1.83411464e-01 6.42477095e-01
8.71631622e-01 6.60343111e-01 8.03207755e-02 -1.76195577e-01
1.60954475e+00 -8.77828479e-01 -5.18229485e-01 -3.85134459e-01
7.05541968e-01 -9.07007933e-01 1.38003421e+00 6.05826914e-01
-7.44556963e-01 -6.08949602e-01 -6.68803751e-01 2.85054259e-02
-3.53222162e-01 6.74164519e-02 8.41108978e-01 6.93981588e-01
-9.93088543e-01 3.24178845e-01 -6.55937195e-01 -6.04820192e-01
2.47131228e-01 3.22924763e-01 -9.01588649e-02 -1.47225648e-01
-1.32349348e+00 5.31060755e-01 5.76876938e-01 -2.38900483e-02
-9.51567769e-01 -9.76104677e-01 -8.64439726e-01 4.75547276e-02
6.12122297e-01 -7.91471839e-01 1.84741056e+00 -6.34471893e-01
-1.61854744e+00 5.22932351e-01 -1.73450470e-01 -7.27316618e-01
2.58487701e-01 -2.39939243e-01 -7.06197917e-01 -1.17998891e-01
3.66929770e-01 5.75273812e-01 9.49185133e-01 -8.06186736e-01
-9.02137280e-01 -2.68852264e-01 3.04258496e-01 -3.28979790e-02
-4.00088161e-01 5.02821282e-02 -1.34835288e-01 -5.54431319e-01
-1.85277298e-01 -7.74925709e-01 -2.95720194e-02 -6.48041189e-01
-4.66080576e-01 -4.24810678e-01 9.61605012e-01 -4.36383724e-01
1.15464473e+00 -2.18589544e+00 -4.33378965e-01 -1.36078045e-01
7.31058046e-03 2.73443729e-01 -2.52980441e-01 6.95527375e-01
-1.66235626e-01 1.39897510e-01 -3.58663052e-02 -4.15392280e-01
3.46447259e-01 6.99244067e-02 -8.04533660e-01 -5.44705763e-02
3.81836206e-01 1.09779751e+00 -1.08909464e+00 -1.19409978e-01
4.39705491e-01 2.88737476e-01 -5.68015337e-01 5.35567820e-01
-4.75515336e-01 3.94108981e-01 -1.57012492e-01 7.72954524e-01
3.38178992e-01 -2.87082195e-01 5.29127233e-02 -2.53041804e-01
-6.72769221e-03 7.44329035e-01 -9.36927319e-01 1.72823012e+00
-1.09325111e+00 3.11895788e-01 -3.01143944e-01 -8.00376356e-01
9.51023519e-01 6.26837969e-01 4.13410038e-01 -5.19208848e-01
1.77685842e-01 2.24136710e-01 -1.54300541e-01 -5.31743526e-01
5.58011532e-01 -2.80682921e-01 -5.14470875e-01 6.14504933e-01
4.30902690e-01 -2.11705163e-01 2.87449867e-01 1.39807567e-01
1.50667679e+00 -9.64435637e-02 6.05961263e-01 8.90705958e-02
3.92513037e-01 -7.33139813e-02 4.12787348e-01 9.57698524e-01
-2.61296686e-02 5.92493474e-01 4.73104529e-02 -4.83926117e-01
-5.98441660e-01 -9.17650819e-01 -1.22987263e-01 1.34269750e+00
-1.36305571e-01 -5.17084479e-01 -3.44512224e-01 -8.49856317e-01
-1.56418398e-01 1.51740909e+00 -2.63396919e-01 -3.65386337e-01
-3.62485290e-01 -3.32693368e-01 5.71928144e-01 2.47107103e-01
2.80272961e-01 -1.43867743e+00 -8.48613977e-01 4.64595199e-01
-5.40200055e-01 -1.29631269e+00 -4.00404871e-01 3.46236974e-01
-4.09002692e-01 -7.24906623e-01 -2.00757921e-01 -6.20876253e-01
1.30440369e-01 2.33658969e-01 1.31119394e+00 -3.44463050e-01
-3.78454477e-01 6.49053931e-01 -7.41127312e-01 -6.17971241e-01
-5.99529326e-01 2.57817447e-01 2.28530645e-01 1.14177160e-01
6.80791557e-01 -7.33964801e-01 -3.68170977e-01 2.85470098e-01
-8.37324202e-01 1.74673602e-01 6.11554265e-01 7.53763795e-01
4.05018955e-01 1.41239405e-01 8.07466149e-01 -8.85316372e-01
7.42249072e-01 -9.44386721e-01 -2.89404660e-01 1.21561237e-01
-5.92364609e-01 -1.65687427e-01 9.11062896e-01 -5.27572751e-01
-1.05309618e+00 -1.24722198e-01 -6.21839941e-01 -5.53343892e-01
-5.55216789e-01 4.33219701e-01 -4.22852375e-02 3.56658787e-01
4.79486346e-01 3.41388553e-01 -4.51382279e-01 -5.74904799e-01
4.64009106e-01 1.10705996e+00 4.64477956e-01 -5.83198071e-01
5.96160829e-01 1.96099028e-01 -6.02749348e-01 -6.75673187e-01
-9.63643193e-01 -8.69892061e-01 -3.52249920e-01 -3.84949408e-02
4.34330136e-01 -8.51740360e-01 -7.93229401e-01 1.84134424e-01
-1.13282466e+00 -3.99110049e-01 -6.05105758e-01 2.37904340e-01
-7.26428747e-01 1.80250719e-01 -4.08537537e-01 -9.81091857e-01
-8.50132525e-01 -1.06094742e+00 1.57609153e+00 8.24404589e-04
-8.44985187e-01 -8.59633148e-01 -7.15008676e-02 4.98026878e-01
3.18496525e-01 -1.82888687e-01 7.95691013e-01 -1.37177432e+00
-4.23427731e-01 -3.09384972e-01 1.25132605e-01 3.48530978e-01
2.57488757e-01 -4.81164038e-01 -1.19179487e+00 -1.92875683e-01
3.31945777e-01 -5.16435862e-01 2.71160603e-01 -6.71895444e-02
1.16922760e+00 -6.36535048e-01 -2.10713252e-01 3.74868155e-01
1.02523148e+00 2.88806915e-01 4.73930866e-01 1.95646942e-01
4.15929794e-01 3.60296756e-01 1.05359435e+00 6.39268398e-01
3.89068127e-01 8.86622608e-01 3.78355622e-01 3.06254476e-01
2.57975957e-03 -3.05964619e-01 4.97218549e-01 8.17237675e-01
4.11841720e-01 -3.77360880e-01 -1.09753418e+00 6.37654543e-01
-1.81217086e+00 -1.10292959e+00 1.80581227e-01 2.09543991e+00
8.31193328e-01 1.29484862e-01 3.91872264e-02 5.61206602e-02
4.93651181e-01 -1.74281701e-01 -5.27821064e-01 -6.45902336e-01
4.33010191e-01 5.20758092e-01 -1.43653423e-01 2.13583246e-01
-9.38728034e-01 8.06453288e-01 5.48075724e+00 8.61845851e-01
-1.16962469e+00 5.42000413e-01 6.05431855e-01 -1.75623998e-01
-2.37362623e-01 -6.73893699e-03 -1.04804528e+00 6.66083694e-01
1.37252927e+00 -4.15031314e-01 4.06976581e-01 1.21405852e+00
1.04855672e-01 2.53357738e-01 -1.45046675e+00 1.03453362e+00
2.21064061e-01 -1.32332873e+00 -1.43636391e-01 -2.29614332e-01
4.08313215e-01 1.50147498e-01 9.07271951e-02 9.78374064e-01
2.50677407e-01 -6.50569618e-01 8.05471122e-01 1.65502161e-01
8.23161483e-01 -5.15989363e-01 8.40542793e-01 7.14917183e-01
-1.10576284e+00 -3.67051989e-01 3.52410898e-02 -8.56959894e-02
4.67947364e-01 7.34453559e-01 -1.49378669e+00 5.84347844e-01
7.40062177e-01 5.03291368e-01 -1.99961990e-01 6.63972974e-01
-1.00683756e-01 7.68685579e-01 -3.46571177e-01 -9.14308429e-03
1.07910305e-01 2.53851801e-01 6.50798500e-01 1.37689829e+00
4.40489799e-01 -4.98571135e-02 3.16431284e-01 9.27243352e-01
-1.45609463e-02 3.40718143e-02 -6.32241726e-01 -1.04593381e-01
4.99913365e-01 1.41108644e+00 -4.81449753e-01 -3.74559283e-01
-2.41787389e-01 1.09277892e+00 6.55383170e-02 8.29141065e-02
-9.92542505e-01 -2.45408773e-01 7.02215135e-01 4.03745025e-02
3.02010328e-01 2.69390196e-01 -3.95539068e-02 -1.11884868e+00
9.54110548e-02 -1.22160864e+00 4.95253712e-01 -8.25873077e-01
-1.44390607e+00 7.51718998e-01 2.15495318e-01 -1.30490649e+00
-8.15967798e-01 -4.56762850e-01 -6.80189788e-01 6.02524042e-01
-1.35708773e+00 -1.35280716e+00 -3.07662249e-01 2.97972172e-01
1.17653453e+00 -9.68379974e-02 1.16884255e+00 2.65656769e-01
-5.45732021e-01 6.89234197e-01 -3.10143679e-01 -2.90707678e-01
5.12756169e-01 -1.00013471e+00 7.65563369e-01 7.83578277e-01
3.04103494e-01 7.97835886e-01 7.23987043e-01 -5.98581970e-01
-1.53440893e+00 -1.53325045e+00 1.08067536e+00 -5.34332693e-01
7.71035075e-01 -7.91122615e-01 -7.68827140e-01 9.82256353e-01
2.44063623e-02 -1.79300755e-01 8.74783754e-01 1.85281113e-01
-2.57532746e-01 -8.19102079e-02 -1.20459867e+00 4.99309391e-01
1.00766969e+00 -5.26588440e-01 -6.01562262e-01 7.50231445e-01
1.16184294e+00 -3.84856641e-01 -8.37023675e-01 5.05065620e-01
4.83493954e-01 -1.07890379e+00 9.25668061e-01 -4.49581951e-01
2.68182039e-01 3.73852067e-02 -3.05004060e-01 -9.10648644e-01
-2.18714952e-01 -1.07665002e+00 -3.96018177e-01 1.37804127e+00
3.48923206e-01 -8.22998881e-01 4.52348322e-01 5.84293127e-01
-3.55118573e-01 -9.03468847e-01 -9.05272126e-01 -8.96176875e-01
-5.17015219e-01 -1.03179395e+00 9.12922859e-01 6.25301898e-01
1.10870719e-01 6.43481612e-01 -2.90377438e-01 7.84573182e-02
2.59186953e-01 1.81142956e-01 8.49909127e-01 -1.02583945e+00
-6.64016724e-01 -1.30348787e-01 -4.57576454e-01 -9.11849558e-01
2.62192070e-01 -1.06272650e+00 2.69780070e-01 -1.28968048e+00
-9.94891524e-02 -4.64205533e-01 -2.97994733e-01 8.98161590e-01
-2.91694142e-02 1.90531880e-01 1.49883151e-01 1.92367241e-01
-6.59273267e-01 4.64575857e-01 3.72760773e-01 -1.96291998e-01
-2.69464523e-01 3.81321669e-01 -8.34036589e-01 6.41556382e-01
9.47288811e-01 -3.86072636e-01 -6.42300487e-01 -5.00616655e-02
9.76871923e-02 -2.49024138e-01 3.24364990e-01 -1.13335204e+00
1.02255292e-01 -6.92207413e-03 -2.70734429e-01 -6.32831633e-01
4.64346796e-01 -8.85301232e-01 2.48268634e-01 2.89876044e-01
-2.36644790e-01 -4.50668931e-02 2.33064637e-01 3.85115176e-01
-5.61846327e-03 -4.76872146e-01 3.14146668e-01 -7.13316202e-02
-1.07214212e+00 2.45248467e-01 -3.46839875e-01 3.91810127e-02
1.13861787e+00 -1.21729560e-01 -1.48047373e-01 -4.82693762e-01
-3.43790412e-01 -3.20451558e-02 8.53877664e-02 8.05395186e-01
7.10435510e-01 -1.17589605e+00 -5.27026117e-01 4.26976353e-01
6.28226697e-01 -9.22614112e-02 2.48090386e-01 8.35039735e-01
3.07557825e-02 5.64328253e-01 3.44416142e-01 -6.69606566e-01
-9.84037340e-01 6.73431814e-01 -1.96029007e-01 -4.13997501e-01
-7.02345729e-01 7.52549171e-01 2.91801482e-01 -7.26664960e-01
1.82727888e-01 -5.01063228e-01 1.67140067e-01 -3.01482260e-01
8.43039393e-01 2.54316866e-01 3.93031031e-01 -4.48116392e-01
-5.06028175e-01 -9.48541537e-02 -2.85091430e-01 9.62853357e-02
1.28759122e+00 -1.26658380e-01 5.27579412e-02 7.13238001e-01
1.16688073e+00 -2.86798507e-01 -6.33208692e-01 -3.40969823e-02
4.57523018e-02 -2.94583172e-01 -1.55671872e-02 -9.82527614e-01
-4.65434790e-01 6.20029867e-01 5.20008624e-01 3.17587048e-01
1.23392916e+00 1.36964753e-01 1.15746057e+00 5.71432531e-01
7.18298137e-01 -7.66895235e-01 1.48770139e-01 5.65623820e-01
9.44468260e-01 -1.21789002e+00 -4.44329977e-01 -3.24691415e-01
-6.63522661e-01 9.38897371e-01 3.88933539e-01 3.31294239e-01
3.78792852e-01 4.62755322e-01 5.07426150e-02 -1.35097653e-01
-1.16724837e+00 -2.52520412e-01 1.12182774e-01 6.77587032e-01
4.14916366e-01 1.50973871e-01 1.38414130e-01 8.84881794e-01
-4.57493931e-01 1.39347225e-01 2.87103921e-01 7.80196309e-01
-6.75336272e-02 -1.11998689e+00 -2.02061757e-01 6.69401646e-01
-2.61945128e-01 -4.70941007e-01 -5.76508604e-03 4.34845984e-01
3.07791919e-01 1.35980737e+00 -1.07414365e-01 -6.87234700e-01
4.19862151e-01 4.11007226e-01 4.70335633e-02 -1.22005558e+00
-7.88048089e-01 -2.75341392e-01 2.24174291e-01 -7.12537408e-01
6.46836832e-02 -6.56175435e-01 -1.07929444e+00 -7.02541769e-02
-4.66494590e-01 3.52935791e-02 7.50374317e-01 1.02882957e+00
7.84799457e-01 5.66788137e-01 6.29506886e-01 -7.54390180e-01
-8.23121369e-01 -9.91768181e-01 -1.64245754e-01 4.33085769e-01
1.20951973e-01 -5.01842856e-01 -4.45237249e-01 5.71317673e-02] | [12.236568450927734, 7.586358070373535] |
b01a0664-bd38-4508-9f1c-ce94051c52a8 | timeline-extraction-using-distant-supervision | null | null | https://aclanthology.org/D16-1200 | https://aclanthology.org/D16-1200.pdf | Timeline extraction using distant supervision and joint inference | null | ['Savelie Cornegruta', 'Andreas Vlachos'] | 2016-11-01 | null | null | null | emnlp-2016-11 | ['temporal-information-extraction'] | ['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] | [-7.432382106781006, 3.664335250854492] |
ea73f335-be66-468f-b63e-a6947b53b041 | tps-attention-enhanced-thin-plate-spline-for | 2305.05322 | null | https://arxiv.org/abs/2305.05322v1 | https://arxiv.org/pdf/2305.05322v1.pdf | TPS++: Attention-Enhanced Thin-Plate Spline for Scene Text Recognition | Text irregularities pose significant challenges to scene text recognizers. Thin-Plate Spline (TPS)-based rectification is widely regarded as an effective means to deal with them. Currently, the calculation of TPS transformation parameters purely depends on the quality of regressed text borders. It ignores the text content and often leads to unsatisfactory rectified results for severely distorted text. In this work, we introduce TPS++, an attention-enhanced TPS transformation that incorporates the attention mechanism to text rectification for the first time. TPS++ formulates the parameter calculation as a joint process of foreground control point regression and content-based attention score estimation, which is computed by a dedicated designed gated-attention block. TPS++ builds a more flexible content-aware rectifier, generating a natural text correction that is easier to read by the subsequent recognizer. Moreover, TPS++ shares the feature backbone with the recognizer in part and implements the rectification at feature-level rather than image-level, incurring only a small overhead in terms of parameters and inference time. Experiments on public benchmarks show that TPS++ consistently improves the recognition and achieves state-of-the-art accuracy. Meanwhile, it generalizes well on different backbones and recognizers. Code is at https://github.com/simplify23/TPS_PP. | ['Yu-Gang Jiang', 'Hongtao Xie', 'Jinfeng Bai', 'Zhineng Chen', 'Tianlun Zheng'] | 2023-05-09 | null | null | null | null | ['scene-text-recognition'] | ['computer-vision'] | [ 4.67756301e-01 -2.54463702e-01 5.40230237e-02 -3.03373903e-01
-9.78801370e-01 -3.06317389e-01 5.51390231e-01 -2.38314644e-01
-2.66597658e-01 1.34360284e-01 2.27811843e-01 -2.04738542e-01
4.49803144e-01 -7.07848310e-01 -8.52286041e-01 -8.62802446e-01
9.92267966e-01 3.96208644e-01 5.48220932e-01 -1.38695449e-01
4.89831448e-01 3.48389983e-01 -1.22005272e+00 3.52093190e-01
1.25746894e+00 9.11139607e-01 3.42891157e-01 5.98232627e-01
-5.83133459e-01 5.90414107e-01 -5.33066213e-01 -7.53162444e-01
1.79657266e-01 -1.29622638e-01 -4.87982333e-01 3.23194355e-01
7.30872154e-01 -3.37416261e-01 -5.07247686e-01 9.43605661e-01
5.54592252e-01 9.80584919e-02 7.28489459e-01 -8.40974569e-01
-1.10910761e+00 4.62409317e-01 -1.03573751e+00 1.79405913e-01
2.29449403e-02 3.83303851e-01 8.76210213e-01 -1.42565703e+00
1.58320591e-01 1.00289571e+00 8.17598701e-01 3.63475829e-01
-1.03338945e+00 -4.91019100e-01 4.02637362e-01 2.07388818e-01
-1.62825561e+00 -4.35603589e-01 6.76572382e-01 -2.31392831e-01
1.05181777e+00 4.68344420e-01 4.05619621e-01 8.03288043e-01
1.16333356e-02 1.05180991e+00 5.03211081e-01 -2.81379491e-01
-2.07582518e-01 -1.65114760e-01 1.80161029e-01 5.82205832e-01
8.79350379e-02 -4.38590676e-01 -3.83890003e-01 3.44447672e-01
8.57661963e-01 1.58698201e-01 -6.59489691e-01 -4.87993397e-02
-1.09446144e+00 4.71465826e-01 3.64572763e-01 1.31269723e-01
-2.12375924e-01 1.52765512e-01 3.23357493e-01 -1.65281385e-01
7.17880845e-01 4.61505614e-02 -4.16307330e-01 -8.58246014e-02
-1.00111854e+00 -4.81655747e-02 3.22655082e-01 1.01255667e+00
4.99197423e-01 4.13867384e-01 -4.85174090e-01 1.14784551e+00
1.83090985e-01 8.28297794e-01 3.98391217e-01 -2.31814995e-01
7.82154679e-01 7.41329432e-01 -6.29232600e-02 -1.05648875e+00
-2.15961725e-01 -5.13964951e-01 -9.65119541e-01 -1.74101159e-01
3.13065290e-01 2.08749294e-01 -1.01998317e+00 1.14713299e+00
3.34754795e-01 3.71979654e-01 -2.34195426e-01 1.09135425e+00
9.63659048e-01 1.00786114e+00 -8.45370069e-03 1.02810048e-01
1.39703095e+00 -1.58250833e+00 -8.11785042e-01 -2.48132482e-01
5.72953999e-01 -1.02704418e+00 1.48694646e+00 5.15940368e-01
-1.11464024e+00 -5.05643487e-01 -8.60406578e-01 -6.02162719e-01
-2.05277085e-01 5.95341563e-01 2.92421877e-01 5.34068108e-01
-9.79340613e-01 3.83493513e-01 -7.55846918e-01 -1.72603279e-01
6.18603468e-01 3.93747896e-01 -2.82900291e-03 -4.42823432e-02
-6.60227180e-01 7.67722011e-01 -8.06834996e-02 3.69361281e-01
-3.74358863e-01 -7.89145112e-01 -7.73066819e-01 2.26579770e-01
6.11025870e-01 -4.79891270e-01 1.13970661e+00 -1.02920151e+00
-1.80006766e+00 6.78166568e-01 -4.02336866e-01 -1.62812531e-01
7.97292233e-01 -4.71170396e-01 -2.12013215e-01 -8.86208788e-02
-1.34163186e-01 4.24187183e-01 1.28732383e+00 -1.04648852e+00
-4.26019132e-01 -2.55876660e-01 -4.51399177e-01 4.81322527e-01
-3.81219596e-01 -4.28240784e-02 -1.23671377e+00 -1.14769673e+00
8.56526047e-02 -6.16867006e-01 9.66586545e-02 2.62469351e-01
-5.22107244e-01 -1.59395218e-01 9.81903076e-01 -9.88975465e-01
1.28520381e+00 -2.17187238e+00 9.11468714e-02 -1.45633593e-01
2.52067715e-01 5.93000412e-01 -2.88380325e-01 1.06581420e-01
-1.74231213e-02 1.66730136e-01 -4.76816803e-01 -6.73115730e-01
1.86996497e-02 -7.86842629e-02 -5.57504296e-01 5.34835398e-01
4.77596641e-01 1.12189448e+00 -3.80076379e-01 -4.77400661e-01
4.79170352e-01 8.09706688e-01 -6.63105369e-01 2.54461961e-03
-3.62857014e-01 1.33448109e-01 -3.84731203e-01 7.01972485e-01
1.01727784e+00 -3.10286790e-01 -1.90015331e-01 -3.88349980e-01
-2.47212052e-01 2.38310203e-01 -1.03981376e+00 1.51006067e+00
-4.07702267e-01 7.34948337e-01 -3.67610306e-02 -7.02188551e-01
1.02935719e+00 -4.78844866e-02 1.44094422e-01 -6.90836132e-01
3.92779440e-01 -6.11380488e-03 -4.45378244e-01 -4.88624215e-01
8.22089434e-01 3.39568257e-01 1.65421098e-01 3.38549614e-01
-2.52115428e-01 -2.47854486e-01 -2.32263878e-01 -8.47498253e-02
8.17282557e-01 5.04410982e-01 1.10030845e-01 -1.12688772e-01
8.07237566e-01 -1.00937754e-01 5.96360862e-01 5.53845286e-01
2.37942152e-02 1.07465959e+00 1.78142697e-01 -4.36138064e-01
-1.08818328e+00 -8.37177575e-01 -1.25257805e-01 1.20541108e+00
4.16107178e-01 -3.05557877e-01 -8.99366379e-01 -5.83704233e-01
-2.38325104e-01 6.73465073e-01 -5.32976985e-01 -1.94838736e-02
-6.47450864e-01 -8.78021181e-01 5.62977076e-01 7.28744090e-01
8.62993956e-01 -8.47154737e-01 -6.34462982e-02 -3.66261229e-02
-2.84336239e-01 -1.02394629e+00 -1.24419594e+00 -2.83106808e-02
-8.09802055e-01 -7.11131215e-01 -9.47210729e-01 -8.77327561e-01
9.47604358e-01 6.33276641e-01 8.18196654e-01 5.02572238e-01
-1.82634771e-01 7.88526013e-02 -4.66717988e-01 -1.67437479e-01
5.36529627e-03 1.44110158e-01 -4.83783662e-01 3.45255733e-01
2.13413611e-01 -1.78069040e-01 -5.09309888e-01 5.37574828e-01
-9.17524874e-01 3.74928743e-01 6.87285602e-01 9.19749975e-01
7.65079916e-01 -4.28170636e-02 1.38265312e-01 -8.46206784e-01
1.93938315e-01 -2.34471597e-02 -6.76229239e-01 3.36919785e-01
-3.02332729e-01 -1.24685459e-01 7.61333048e-01 -6.33030355e-01
-1.26238525e+00 2.01603815e-01 -3.12890947e-01 -4.49769884e-01
6.70424663e-03 5.35017513e-02 -5.17437220e-01 -2.25880459e-01
3.10851008e-01 7.21020341e-01 -3.84313941e-01 -6.18491054e-01
2.72480011e-01 7.44860768e-01 6.05111063e-01 -3.78327221e-01
1.08153033e+00 3.86936367e-01 -5.27257204e-01 -1.03906476e+00
-8.14364314e-01 -3.24349791e-01 -6.00459099e-01 2.21892935e-03
7.97396779e-01 -8.42227578e-01 -6.63658261e-01 9.05476749e-01
-1.21157026e+00 -6.61621571e-01 -1.57033280e-01 6.63100481e-02
-2.28534222e-01 5.77949405e-01 -5.68072975e-01 -5.43582499e-01
-7.50411451e-01 -1.24077237e+00 1.34451354e+00 2.76296735e-01
2.50987828e-01 -6.58949018e-01 -3.50828975e-01 6.42410994e-01
4.97626275e-01 -3.53157729e-01 7.35781610e-01 -5.47697306e-01
-7.79182971e-01 -1.07727386e-01 -6.88795924e-01 2.49539047e-01
-4.75799851e-02 2.50211507e-01 -1.11678290e+00 -6.28174515e-03
-8.21412355e-02 3.11176389e-01 1.07428539e+00 3.14285547e-01
1.51956511e+00 -3.28637600e-01 -9.56555307e-02 1.03392744e+00
1.27326083e+00 8.32973123e-02 1.03314543e+00 1.96295232e-01
1.24309528e+00 1.48048490e-01 5.06312251e-01 3.51433843e-01
5.56579947e-01 9.39172745e-01 3.20630252e-01 -4.42411989e-01
-4.30571169e-01 -1.42076984e-01 4.49338257e-01 1.04532278e+00
2.94796973e-02 -5.17689705e-01 -9.09160912e-01 2.47409299e-01
-1.71108663e+00 -7.79409647e-01 -5.40643096e-01 2.24067664e+00
8.47825766e-01 3.02800499e-02 -9.06843171e-02 1.13842912e-01
1.06500018e+00 1.18745044e-01 -5.70694983e-01 -2.18231335e-01
-2.31664300e-01 1.32601589e-01 5.92176914e-01 6.82181895e-01
-1.07612455e+00 1.41151607e+00 5.23367929e+00 1.23279786e+00
-1.36963451e+00 7.29339495e-02 6.72410071e-01 7.53822625e-02
-1.92475453e-01 -1.15936100e-01 -8.71125758e-01 5.82476556e-01
3.05504978e-01 4.86076958e-02 4.59086508e-01 7.11260557e-01
3.12450677e-01 5.36606424e-02 -6.12421274e-01 1.02427280e+00
3.32303405e-01 -1.28310752e+00 3.80548120e-01 -2.66262025e-01
6.29965782e-01 3.58181968e-02 1.64587662e-01 2.39774063e-01
-1.25190496e-01 -8.38731587e-01 9.73700762e-01 4.76382405e-01
9.48824167e-01 -5.96462727e-01 7.26811409e-01 1.04216166e-01
-1.41507411e+00 9.77902040e-02 -4.84092504e-01 1.96592376e-01
2.13054344e-02 5.57528913e-01 -7.01430976e-01 4.33484018e-01
5.45708895e-01 8.84464622e-01 -6.75847888e-01 1.15676486e+00
-3.76507491e-01 6.85325325e-01 -2.26769954e-01 -3.72480392e-03
-7.08866641e-02 -3.34640473e-01 4.58111614e-01 1.48246169e+00
2.67248958e-01 1.34107992e-01 3.02448757e-02 9.03275132e-01
-2.22216249e-01 3.62735093e-01 -5.70702888e-02 1.27898470e-01
3.52456242e-01 1.25805855e+00 -8.94148648e-01 -3.98728639e-01
-5.42833090e-01 1.34644735e+00 2.35194162e-01 3.71619344e-01
-1.25709069e+00 -4.98903990e-01 2.56342292e-01 2.41167650e-01
6.56046748e-01 -6.34118989e-02 -6.79430127e-01 -1.55133176e+00
1.60436541e-01 -9.04043376e-01 6.09082431e-02 -1.03113890e+00
-1.17783988e+00 6.32816613e-01 -6.31998241e-01 -1.11450148e+00
6.57952845e-01 -5.64077854e-01 -7.32842326e-01 8.86707664e-01
-1.71009314e+00 -1.40625060e+00 -7.55211115e-01 5.71509242e-01
1.12107611e+00 3.02307308e-01 4.57551271e-01 3.95979673e-01
-1.18116176e+00 8.89993429e-01 -4.14353125e-02 2.86549002e-01
8.21349442e-01 -1.08243811e+00 8.39342713e-01 1.06765580e+00
-7.65002146e-02 2.80324996e-01 4.50359523e-01 -7.26843596e-01
-1.55210805e+00 -1.32420599e+00 7.02728510e-01 -3.75693172e-01
5.57705522e-01 -3.57766390e-01 -1.33975613e+00 5.30226588e-01
8.91355518e-03 -3.27564366e-02 2.48938650e-01 -4.05911773e-01
-4.12632883e-01 -1.99128479e-01 -9.33885157e-01 7.92028666e-01
9.60637927e-01 -4.29463476e-01 -4.09807116e-01 3.20706546e-01
6.82027638e-01 -7.67786384e-01 -5.85112989e-01 5.83563596e-02
3.15017611e-01 -7.32123613e-01 8.31072032e-01 5.82329333e-02
4.47633564e-01 -5.21802723e-01 5.67964464e-02 -8.58872652e-01
-3.85105669e-01 -7.10746706e-01 1.00357598e-02 1.53664982e+00
3.66879463e-01 -7.33564079e-01 6.90392673e-01 5.62539637e-01
-5.30165792e-01 -6.98714018e-01 -6.36626482e-01 -5.08053601e-01
-5.49248159e-02 -4.93002981e-01 6.36855006e-01 8.78295481e-01
-4.13876832e-01 2.82346576e-01 -4.27433997e-01 3.52847725e-01
3.82118702e-01 1.34974299e-03 8.44899356e-01 -1.05582190e+00
-1.98585004e-01 -6.25011623e-01 -1.78046882e-01 -1.59845257e+00
-6.40935227e-02 -8.34284365e-01 3.23791504e-01 -1.42741823e+00
2.13870361e-01 -4.22673017e-01 1.62759379e-01 6.43140197e-01
-4.79749024e-01 4.19542670e-01 2.78806210e-01 2.46656403e-01
-6.13825023e-01 7.95156777e-01 1.38445699e+00 -2.87692368e-01
-2.50961930e-01 7.11164903e-03 -6.49919212e-01 8.64487171e-01
8.83803308e-01 -4.44623023e-01 -1.56807844e-02 -9.77406800e-01
4.27260669e-03 -2.40712434e-01 4.17001098e-01 -7.98479259e-01
3.02836925e-01 -8.20404887e-02 3.91327024e-01 -8.67458701e-01
2.90396094e-01 -7.07762837e-01 -5.26935048e-02 1.21492408e-01
-2.11795405e-01 6.29780143e-02 3.13006997e-01 4.54743654e-01
7.82443210e-02 -2.24158973e-01 8.92157495e-01 3.00040752e-01
-5.03688693e-01 5.11616945e-01 -1.27649963e-01 7.59078115e-02
7.01710463e-01 -3.19128573e-01 -5.77429473e-01 -1.50080442e-01
-3.06834906e-01 1.33924603e-01 6.98477745e-01 3.44481885e-01
7.48109996e-01 -1.02975285e+00 -7.53603637e-01 1.52227014e-01
-1.33028135e-01 3.90771896e-01 4.39683467e-01 1.11652637e+00
-7.07427561e-01 2.01114774e-01 2.78812796e-01 -6.54368758e-01
-1.17099404e+00 4.33791697e-01 4.40887243e-01 -1.92034215e-01
-1.06964445e+00 8.73812675e-01 5.99714279e-01 -2.21494898e-01
3.47625405e-01 -4.92515862e-01 -8.43308866e-03 -1.48519546e-01
7.96718895e-01 3.70603949e-01 1.95976034e-01 -7.40952134e-01
-2.15846106e-01 1.11926162e+00 -3.11024994e-01 2.49723747e-01
1.38427484e+00 -2.93981403e-01 1.70830600e-02 4.64092419e-02
8.01266849e-01 9.56459641e-02 -1.57608807e+00 -3.63212407e-01
-2.74030745e-01 -6.40185893e-01 3.72697055e-01 -8.19753885e-01
-1.51044035e+00 1.01925933e+00 2.81117380e-01 -2.45804086e-01
1.26813889e+00 -5.02464652e-01 9.05987978e-01 2.58232474e-01
-9.80654433e-02 -9.25853908e-01 1.74942762e-01 5.87551832e-01
1.05672860e+00 -1.10734117e+00 4.11336906e-02 -7.89049208e-01
-8.16098690e-01 1.09160185e+00 7.57554293e-01 -6.90401495e-02
2.40770757e-01 5.25726974e-01 -1.54174762e-02 1.30389467e-01
-5.55965424e-01 -1.74417777e-03 4.53946263e-01 5.65019965e-01
4.05306250e-01 -2.51021475e-01 3.21037252e-03 7.93008685e-01
1.38619974e-01 -1.24078773e-01 4.79749233e-01 5.69196701e-01
-4.27813590e-01 -8.02640617e-01 -4.98854697e-01 4.19588417e-01
-4.00602132e-01 -6.03173196e-01 -2.93065667e-01 5.36688268e-01
-1.77737270e-02 6.74781322e-01 1.14088260e-01 -2.31491432e-01
4.45843369e-01 -1.44693449e-01 2.06573576e-01 -4.72213179e-01
-8.40278804e-01 3.63106191e-01 -2.67812103e-01 -4.32689756e-01
7.59033561e-02 -4.89599824e-01 -1.40136671e+00 -4.94511634e-01
-8.26156139e-01 -3.21698427e-01 5.73527038e-01 8.34073663e-01
4.73605990e-01 7.20240057e-01 5.52182496e-01 -1.06853259e+00
-4.01171833e-01 -8.68158340e-01 -2.71549791e-01 1.38871089e-01
2.64835089e-01 -4.68835622e-01 -2.63509512e-01 3.36634040e-01] | [11.982915878295898, 2.1447219848632812] |
878752fd-c519-44e0-9802-ad088861b54f | learned-spectral-super-resolution | 1703.09470 | null | http://arxiv.org/abs/1703.09470v1 | http://arxiv.org/pdf/1703.09470v1.pdf | Learned Spectral Super-Resolution | We describe a novel method for blind, single-image spectral super-resolution.
While conventional super-resolution aims to increase the spatial resolution of
an input image, our goal is to spectrally enhance the input, i.e., generate an
image with the same spatial resolution, but a greatly increased number of
narrow (hyper-spectral) wave-length bands. Just like the spatial statistics of
natural images has rich structure, which one can exploit as prior to predict
high-frequency content from a low resolution image, the same is also true in
the spectral domain: the materials and lighting conditions of the observed
world induce structure in the spectrum of wavelengths observed at a given
pixel. Surprisingly, very little work exists that attempts to use this
diagnosis and achieve blind spectral super-resolution from single images. We
start from the conjecture that, just like in the spatial domain, we can learn
the statistics of natural image spectra, and with its help generate finely
resolved hyper-spectral images from RGB input. Technically, we follow the
current best practice and implement a convolutional neural network (CNN), which
is trained to carry out the end-to-end mapping from an entire RGB image to the
corresponding hyperspectral image of equal size. We demonstrate spectral
super-resolution both for conventional RGB images and for multi-spectral
satellite data, outperforming the state-of-the-art. | ['Konrad Schindler', 'Silvano Galliani', 'Emmanuel Baltsavias', 'Charis Lanaras', 'Dimitrios Marmanis'] | 2017-03-28 | null | null | null | null | ['spectral-super-resolution'] | ['computer-vision'] | [ 9.78059292e-01 -2.32779786e-01 1.29153341e-01 -1.91843525e-01
-7.76339352e-01 -6.41680419e-01 2.59057134e-01 -5.31232595e-01
-3.60764146e-01 1.02639318e+00 2.69907236e-01 -1.22507803e-01
-3.09514821e-01 -1.11984932e+00 -7.23836124e-01 -1.05421662e+00
1.81465939e-01 1.44651487e-01 2.63993945e-02 -4.36801016e-01
-1.16272882e-01 6.99328065e-01 -1.96932900e+00 4.35259551e-01
1.09328604e+00 7.50328541e-01 5.24337530e-01 9.46198702e-01
8.17731768e-02 5.49251258e-01 -2.91147649e-01 4.10371087e-02
6.16956770e-01 -6.08178675e-01 -9.54312325e-01 3.71629059e-01
7.63550699e-01 -3.00764054e-01 -2.07746401e-01 1.51467538e+00
3.62883240e-01 1.15694255e-01 4.24887478e-01 -4.23074365e-01
-1.10080945e+00 4.04326022e-01 -6.77804589e-01 4.12658192e-02
2.28996187e-01 5.70199713e-02 7.54010856e-01 -5.82337797e-01
2.11156160e-01 8.28599691e-01 6.94874883e-01 3.81965548e-01
-1.50997174e+00 -4.21541750e-01 -5.23582816e-01 3.49009708e-02
-1.38544655e+00 -2.30779588e-01 5.97577214e-01 -2.44727179e-01
6.14295185e-01 2.65580893e-01 5.74957669e-01 7.61065006e-01
-3.32032740e-01 -9.95855257e-02 1.63053584e+00 -6.23121619e-01
3.29036415e-02 -2.89194416e-02 -4.14653301e-01 4.27888602e-01
1.53946668e-01 6.85932577e-01 -7.17700720e-01 1.64682671e-01
9.40242052e-01 -1.09385930e-01 -7.71409631e-01 1.41418353e-01
-1.20725286e+00 4.89612967e-01 8.63190651e-01 4.62756038e-01
-6.56054735e-01 -2.88502067e-01 -3.47101837e-01 1.54352576e-01
5.56902349e-01 7.39689410e-01 -5.16987979e-01 4.30602312e-01
-1.18234909e+00 -3.00025046e-02 2.66248167e-01 3.11207414e-01
1.31674278e+00 1.72417164e-01 2.52617985e-01 8.63268852e-01
-2.00474560e-01 7.43273914e-01 4.19651091e-01 -1.18709612e+00
-1.23820543e-01 2.79145837e-01 2.58905709e-01 -3.82871687e-01
-5.44231057e-01 -6.58318698e-01 -1.09729469e+00 6.97585523e-01
4.22885776e-01 -1.02465101e-01 -1.19572222e+00 1.57495046e+00
7.14641362e-02 1.29722878e-01 3.15798640e-01 1.37132382e+00
5.04646659e-01 7.79098868e-01 -2.51553744e-01 -3.90936852e-01
1.30060470e+00 -4.44182187e-01 -3.17712039e-01 -4.56915706e-01
-1.11603148e-01 -7.51514375e-01 1.23589170e+00 3.39314014e-01
-9.41074848e-01 -7.21147716e-01 -1.01634908e+00 1.39094936e-02
-5.92123389e-01 9.07106921e-02 4.25950915e-01 6.20780528e-01
-1.19006884e+00 8.76989484e-01 -3.39917034e-01 -3.63389909e-01
3.22836787e-01 7.08757415e-02 -4.35088545e-01 -9.99802127e-02
-1.13987684e+00 8.11972439e-01 8.35464180e-01 -9.93467495e-02
-4.44449395e-01 -1.12953579e+00 -4.19698745e-01 -1.90169569e-02
3.13947141e-01 -5.64420044e-01 9.19577360e-01 -1.24130225e+00
-1.38128114e+00 9.71840799e-01 -2.78454155e-01 -2.86327124e-01
-9.07212775e-03 1.73629478e-01 -5.66277742e-01 3.23842734e-01
-5.26980497e-03 2.80824482e-01 9.03515339e-01 -1.58536851e+00
-7.80020773e-01 -6.23390317e-01 8.26013088e-02 3.38218361e-01
-3.75854313e-01 -6.33524209e-02 8.15358572e-03 -3.37854773e-01
2.71752059e-01 -8.20023417e-01 -2.25469530e-01 -1.74809396e-01
-3.65570575e-01 6.19797766e-01 6.28576219e-01 -7.72673309e-01
4.77272213e-01 -2.23837662e+00 1.15845673e-01 9.33871195e-02
2.14771017e-01 4.39775795e-01 -1.32266447e-01 1.59200996e-01
-4.41509902e-01 7.94243589e-02 -7.74212182e-01 1.72703072e-01
-4.75615233e-01 -5.83544970e-02 -1.95758983e-01 4.83966500e-01
9.63883474e-02 5.62416852e-01 -8.52364242e-01 -9.84988082e-03
3.73820513e-01 8.65172565e-01 -2.28749007e-01 1.54954135e-01
-2.16411740e-01 6.73727334e-01 7.19573945e-02 3.41853857e-01
1.22268963e+00 -4.71484303e-01 -2.95216236e-02 -5.90583682e-01
-5.27862668e-01 -2.37501450e-02 -1.19148147e+00 1.65999281e+00
-5.35737395e-01 5.33132136e-01 3.27098578e-01 -6.55854404e-01
6.79003835e-01 1.56300306e-01 4.28067803e-01 -1.10500991e+00
-4.93803054e-01 3.29777539e-01 -3.90205830e-01 -3.86808306e-01
6.60988748e-01 -6.33677959e-01 5.21272719e-01 3.77648503e-01
-4.42283638e-02 -3.52586180e-01 -1.51580364e-01 -2.98554152e-01
5.75357318e-01 9.89383087e-02 2.16253027e-01 -7.45915323e-02
5.39891541e-01 2.41695121e-01 7.01222271e-02 7.31425941e-01
2.98954219e-01 8.31881762e-01 -7.71035478e-02 -3.41067493e-01
-1.52846384e+00 -1.25131714e+00 -3.59965056e-01 1.22763991e+00
4.93573099e-02 2.56150335e-01 -6.78057075e-01 1.45754088e-02
-7.84215406e-02 4.64927554e-01 -6.56796098e-01 2.16876082e-02
-2.01492503e-01 -1.29958999e+00 3.90142530e-01 1.82239443e-01
1.15613174e+00 -9.29779947e-01 -7.18978226e-01 4.71104868e-02
-1.48907900e-01 -1.28533268e+00 3.87040302e-02 3.00743312e-01
-5.83295465e-01 -1.01924407e+00 -6.50288701e-01 -4.73611474e-01
2.35956103e-01 6.74065590e-01 1.11981344e+00 -2.40969375e-01
-7.39010036e-01 4.44927104e-02 -3.89217257e-01 -2.89469838e-01
-4.10908848e-01 -2.14044899e-01 -1.18528411e-01 1.98910370e-01
2.17235684e-01 -7.99876392e-01 -6.14952445e-01 -9.44521949e-02
-1.18872821e+00 2.48218045e-01 7.99642920e-01 8.32027376e-01
6.95222020e-01 7.26842582e-01 3.02025020e-01 -7.58524776e-01
3.65384132e-01 -1.15855843e-01 -6.96642756e-01 1.79349467e-01
-6.36342764e-01 1.64881468e-01 8.54436338e-01 -5.08909002e-02
-1.49907386e+00 4.30661470e-01 -2.10074447e-02 -7.16255978e-02
-7.11956620e-01 5.79861462e-01 -2.12732889e-02 -4.82339501e-01
1.41432035e+00 6.75992787e-01 -1.34635279e-02 -5.12991190e-01
5.93069315e-01 9.10151839e-01 1.27562106e+00 -7.17424005e-02
1.07145834e+00 1.01758218e+00 3.23807359e-01 -1.26843953e+00
-1.17153621e+00 -6.35910153e-01 -8.77683222e-01 -3.01582646e-02
1.03668225e+00 -1.06632280e+00 -4.97219026e-01 4.35017973e-01
-8.15783381e-01 -4.01141584e-01 -3.88335615e-01 4.11999881e-01
-4.52733546e-01 2.21791729e-01 -2.51568407e-01 -9.11312580e-01
-1.27070546e-01 -7.54506409e-01 1.17014015e+00 4.59009290e-01
4.68902647e-01 -8.01254034e-01 5.33022033e-03 4.27975953e-01
7.88153827e-01 3.38521868e-01 8.54303420e-01 2.26450354e-01
-5.69992661e-01 1.76755741e-01 -8.82446706e-01 4.50741500e-01
5.71608424e-01 -2.93784946e-01 -1.54063153e+00 -1.48695469e-01
4.52197231e-02 -2.49911517e-01 1.12879634e+00 5.24487674e-01
1.05446410e+00 -1.55473620e-01 2.48667508e-01 1.06393480e+00
2.03249669e+00 -1.26361743e-01 7.29464114e-01 4.21681792e-01
7.06407845e-01 7.59705663e-01 1.68507144e-01 2.76534915e-01
-1.42672613e-01 5.67567348e-01 8.02734256e-01 -7.06730843e-01
-4.81474549e-01 1.56355038e-01 -1.06148638e-01 -3.38764876e-01
-5.73032379e-01 2.11165905e-01 -6.61510170e-01 4.94783044e-01
-1.40019119e+00 -1.34564853e+00 -3.63872588e-01 2.44746923e+00
1.01177931e+00 -5.28624058e-01 1.10214204e-01 2.53505886e-01
8.15077066e-01 3.76907080e-01 -6.34547651e-01 2.95403093e-01
-8.00871313e-01 5.83014369e-01 9.95990455e-01 7.51170933e-01
-1.07910991e+00 8.52831960e-01 6.47284126e+00 4.31265026e-01
-1.37631559e+00 5.22920378e-02 3.26903522e-01 -1.75528049e-01
-2.16429651e-01 -1.99318558e-01 -1.86256751e-01 1.55185044e-01
1.00738502e+00 -5.50368913e-02 1.40451443e+00 2.70543307e-01
3.19029719e-01 -1.94400311e-01 -4.85614061e-01 1.15084219e+00
-4.80541959e-02 -1.38510907e+00 -4.51491075e-03 1.92807317e-01
9.86301482e-01 2.41068989e-01 3.20456684e-01 -4.79070038e-01
3.44777495e-01 -1.35434139e+00 4.00734663e-01 5.97100317e-01
1.23765659e+00 -8.39728177e-01 4.67069149e-01 2.93745667e-01
-1.14770246e+00 -3.34631592e-01 -7.26287484e-01 8.19584429e-02
-1.83083057e-01 9.07252550e-01 -7.05274343e-01 8.39189231e-01
9.61994767e-01 3.86070460e-01 -5.26532352e-01 9.32311594e-01
-2.29514316e-01 3.83453399e-01 -8.44501480e-02 6.74536347e-01
1.40732944e-01 -4.36149359e-01 3.97150993e-01 1.11373365e+00
4.61719126e-01 2.94742197e-01 -1.33041680e-01 1.16673744e+00
-8.26101601e-02 -1.95845515e-01 -5.23939908e-01 -1.43459424e-01
2.80308664e-01 1.35747278e+00 -3.49034786e-01 -4.72938791e-02
-5.23443222e-01 1.24288976e+00 -9.76855382e-02 6.99124992e-01
-3.10537308e-01 -1.23101510e-01 7.07963943e-01 9.04221758e-02
2.53767192e-01 -9.38593745e-02 -5.36537230e-01 -8.82956803e-01
-4.69304882e-02 -1.01396048e+00 2.80521989e-01 -1.48824787e+00
-1.15494502e+00 6.33003771e-01 -4.05800760e-01 -1.10312712e+00
-1.02973908e-01 -8.58125508e-01 -1.84361219e-01 1.67973721e+00
-2.06321788e+00 -1.31812501e+00 -9.02967811e-01 9.82364178e-01
2.04327673e-01 1.84970573e-01 1.00579405e+00 -4.43795472e-02
2.14569294e-03 -2.35068500e-01 4.04517174e-01 -8.76799077e-02
7.16938853e-01 -1.50247514e+00 7.09137246e-02 1.11675656e+00
-6.95338892e-03 2.22425014e-01 9.70330596e-01 -5.35259843e-01
-1.25378907e+00 -1.28577471e+00 3.60902220e-01 -3.01432963e-02
5.85195482e-01 1.30403176e-01 -1.10667348e+00 2.42495254e-01
1.29072040e-01 1.12959012e-01 5.92466712e-01 -2.46426892e-02
-5.46299636e-01 -2.25028649e-01 -1.16906893e+00 3.17016512e-01
9.12156820e-01 -1.14063799e+00 -2.37063363e-01 6.12413585e-01
7.07942009e-01 -2.92481869e-01 -8.32368135e-01 4.99423265e-01
5.36824048e-01 -1.46723294e+00 1.34931552e+00 -4.68851566e-01
5.91554523e-01 -5.48864007e-01 -3.92667204e-01 -1.71639764e+00
-7.17603147e-01 -1.92433983e-01 3.53399634e-01 7.15477228e-01
6.35155380e-01 -4.75362152e-01 9.00096595e-01 3.05843651e-01
-1.15354843e-01 2.57776622e-02 -6.36773705e-01 -8.53588820e-01
8.97678826e-03 -1.72237113e-01 9.36731994e-01 9.70304072e-01
-5.39108932e-01 9.03883055e-02 -3.87012541e-01 1.08056557e+00
1.13631761e+00 4.66533929e-01 3.20673406e-01 -1.52758944e+00
-2.64727473e-01 -3.84284049e-01 9.58203152e-02 -2.73286164e-01
-4.78626788e-02 -6.25048220e-01 3.00533116e-01 -1.64639974e+00
2.51637995e-01 6.03424013e-02 -2.16159731e-01 2.99083382e-01
-2.65325516e-01 7.01001406e-01 1.97999366e-02 3.79052162e-01
-6.03339029e-03 6.60658404e-02 1.43148112e+00 -1.10662803e-01
-3.90859276e-01 -2.32137576e-01 -8.74498069e-01 5.12077093e-01
8.74506891e-01 -7.36446455e-02 -2.39789799e-01 -3.16921890e-01
5.37580490e-01 1.46847710e-01 9.14904416e-01 -1.14143610e+00
2.22974479e-01 -3.75847548e-01 6.39024436e-01 -2.01000869e-01
4.83668178e-01 -8.79428566e-01 3.29850763e-01 4.11131419e-02
-2.32218862e-01 -7.13657856e-01 7.83046857e-02 2.30722263e-01
-8.63019675e-02 -1.01367839e-01 1.23130715e+00 -5.56716144e-01
-9.73456681e-01 3.67730528e-01 -3.48867327e-02 -3.44701141e-01
3.74507546e-01 -2.72398055e-01 -8.30305696e-01 -3.35075945e-01
-6.47938490e-01 -5.29075801e-01 8.69777501e-01 -1.22477368e-01
5.23553610e-01 -1.06936932e+00 -9.66397285e-01 2.60331184e-01
2.30085477e-01 -2.74832826e-02 5.65946758e-01 4.57743257e-01
-4.17198122e-01 1.74656451e-01 -4.39438909e-01 -3.54574889e-01
-1.16151655e+00 5.69258034e-01 8.72070551e-01 1.55571163e-01
-6.04675055e-01 8.09693515e-01 1.91125929e-01 -3.32684457e-01
-4.16860431e-01 1.40607014e-01 -2.13617429e-01 -1.85887158e-01
9.02032256e-01 1.69322237e-01 2.63383508e-01 -7.46755302e-01
4.72478289e-03 6.89353168e-01 4.76790726e-01 -1.82503164e-01
1.59765697e+00 -2.32545078e-01 -5.92193246e-01 3.00519377e-01
1.08252704e+00 6.37212545e-02 -1.42669141e+00 -5.64425230e-01
-4.86686319e-01 -7.71388531e-01 5.70669174e-01 -1.11821496e+00
-1.09441948e+00 7.27825522e-01 8.48116577e-01 7.56084800e-01
1.73538744e+00 8.27092957e-03 3.03275198e-01 2.47832179e-01
3.58333379e-01 -1.00238013e+00 -1.02198340e-01 3.39895010e-01
7.88010299e-01 -1.26675296e+00 3.45679075e-02 -4.33206052e-01
-4.07490104e-01 1.28446543e+00 2.04032183e-01 1.44188210e-01
2.91894376e-01 2.50344425e-01 1.21419489e-01 -2.67114401e-01
-6.07502013e-02 -8.93038213e-01 2.59431213e-01 1.04414523e+00
3.74184102e-01 1.56685680e-01 5.32296360e-01 -7.41553754e-02
-5.07514417e-01 1.69520065e-01 7.16087699e-01 2.81015247e-01
-8.81996930e-01 -6.81574047e-01 -9.06793237e-01 4.26438332e-01
-3.06439549e-01 -4.79254186e-01 -3.94516170e-01 2.45910779e-01
3.95563632e-01 1.09849823e+00 -7.14012086e-02 -4.39375728e-01
1.72463074e-01 1.94059327e-01 4.58920062e-01 -5.64813614e-01
3.51819135e-02 -8.34259391e-02 -1.23713560e-01 -6.99205220e-01
-9.33258295e-01 -4.49736714e-01 -9.57880020e-01 -3.58437151e-01
-1.90334730e-02 -2.36483179e-02 7.92451859e-01 8.68261278e-01
-4.94282693e-02 6.45774126e-01 7.09692836e-01 -1.18754673e+00
-1.30181298e-01 -9.51312840e-01 -1.19599104e+00 3.22603375e-01
9.95647073e-01 -2.54911840e-01 -6.49408221e-01 3.57853830e-01] | [10.225908279418945, -2.0316128730773926] |
b853246d-4b08-4014-8a48-b673bfcb02d9 | gauche-a-library-for-gaussian-processes-in | 2212.04450 | null | https://arxiv.org/abs/2212.04450v2 | https://arxiv.org/pdf/2212.04450v2.pdf | GAUCHE: A Library for Gaussian Processes in Chemistry | We introduce GAUCHE, a library for GAUssian processes in CHEmistry. Gaussian processes have long been a cornerstone of probabilistic machine learning, affording particular advantages for uncertainty quantification and Bayesian optimisation. Extending Gaussian processes to chemical representations, however, is nontrivial, necessitating kernels defined over structured inputs such as graphs, strings and bit vectors. By defining such kernels in GAUCHE, we seek to open the door to powerful tools for uncertainty quantification and Bayesian optimisation in chemistry. Motivated by scenarios frequently encountered in experimental chemistry, we showcase applications for GAUCHE in molecular discovery and chemical reaction optimisation. The codebase is made available at https://github.com/leojklarner/gauche | ['Alán Aspuru-Guzik', 'Bingqing Cheng', 'Felix Strieth-Kalthoff', 'Saudamini Chaurasia', 'Johannes Durholt', 'Chengzhi Guo', 'Jacob Moss', 'Alex Chan', 'Anthony Bourached', 'Simon Frieder', 'Gregory Kell', 'Austin Tripp', 'Julius Schwartz', 'Aryan Deshwal', 'Arian Jamasb', 'Yuanqi Du', 'Gary Tom', 'Samuel Stanton', 'Jian Tang', 'Philippe Schwaller', 'Alpha A. Lee', 'Bojana Rankovic', 'Sang Truong', 'Aditya Ravuri', 'Henry B. Moss', 'Leo Klarner', 'Ryan-Rhys Griffiths'] | 2022-12-06 | null | null | null | null | ['bayesian-optimisation'] | ['methodology'] | [ 1.34482697e-01 -3.67041454e-02 4.10410911e-02 -1.31606847e-01
-8.14713955e-01 -9.92856562e-01 9.86232936e-01 7.17588305e-01
-2.91759312e-01 9.57156956e-01 -6.28367960e-02 -8.18424761e-01
-3.32370639e-01 -8.75693858e-01 -6.34682834e-01 -1.15006804e+00
-1.88035190e-01 6.88601971e-01 -3.25290523e-02 4.21142042e-01
4.09081161e-01 8.25686216e-01 -1.03569388e+00 -1.16548270e-01
5.87899804e-01 7.19792604e-01 -9.15370649e-04 9.66403127e-01
-3.48351270e-01 3.18742305e-01 -2.13509887e-01 -6.89475417e-01
-1.54909968e-01 -1.49098665e-01 -6.31061018e-01 -7.59826958e-01
-3.20245832e-01 6.74073994e-01 -3.13996464e-01 1.11336720e+00
7.39248574e-01 4.56801951e-01 1.30770195e+00 -9.41501558e-01
-4.32929754e-01 6.15697563e-01 -1.90253884e-01 1.02189817e-01
3.86926442e-01 4.52481151e-01 9.49591339e-01 -9.68412280e-01
2.86561459e-01 1.52321661e+00 5.14473319e-01 3.69200498e-01
-1.67842388e+00 -4.82559204e-01 -1.72640383e-01 -5.13477735e-02
-1.47944653e+00 -7.24498391e-01 4.49697524e-01 -6.37714922e-01
9.34971511e-01 3.45656246e-01 2.08525836e-01 1.07239556e+00
7.34809816e-01 6.50429010e-01 1.01265836e+00 -3.27909708e-01
9.38140213e-01 -1.80844977e-01 1.45147266e-02 3.88805270e-01
3.47498924e-01 3.48286659e-01 -5.81298172e-01 -7.34946847e-01
6.14533126e-01 1.78739168e-02 -2.07540020e-02 -6.91651762e-01
-1.20336115e+00 1.08945978e+00 -3.38572380e-03 -2.76220351e-01
-4.66817886e-01 6.40003860e-01 3.45703721e-01 -1.99342504e-01
3.18131834e-01 6.37562871e-01 -4.68965650e-01 -4.20070529e-01
-4.86008763e-01 4.09893751e-01 1.19287360e+00 9.84762967e-01
4.73932654e-01 -1.64664045e-01 -1.40240461e-01 4.56960350e-01
8.91601741e-01 5.51488280e-01 -2.42095515e-01 -8.03378522e-01
-9.15894508e-02 -7.06122071e-02 2.20128715e-01 -3.56282532e-01
-2.59687275e-01 2.90462654e-02 -6.83804333e-01 2.94936597e-01
5.74446082e-01 -2.70453602e-01 -1.07966173e+00 1.39480853e+00
5.50173879e-01 -2.88176626e-01 2.04434380e-01 3.86728168e-01
7.12959349e-01 8.20565283e-01 7.11187005e-01 -2.03695357e-01
1.43723190e+00 -2.29054511e-01 -5.71855545e-01 2.49471441e-01
4.62365091e-01 -1.11966193e+00 4.02137399e-01 6.94228470e-01
-9.70892966e-01 9.13281441e-02 -7.58759797e-01 -1.09057114e-01
-7.32292831e-01 -4.65811133e-01 1.46118486e+00 1.15542722e+00
-6.47699356e-01 9.56105053e-01 -1.03442264e+00 -1.42354667e-01
6.46010756e-01 4.61504102e-01 -2.29685053e-01 -4.44901288e-02
-1.12251735e+00 7.70358503e-01 6.67956471e-01 1.93418920e-01
-1.02086723e+00 -7.94564068e-01 -9.39474702e-01 -8.40739384e-02
4.39737976e-01 -7.56145477e-01 1.48792672e+00 2.07144812e-01
-1.96406913e+00 2.30565503e-01 -2.25245610e-01 -3.28368932e-01
3.43831569e-01 1.52316168e-01 -2.72576988e-01 -1.78022832e-01
-3.00940454e-01 4.15995300e-01 5.78899503e-01 -7.37787843e-01
-7.46347383e-02 -2.61632115e-01 -2.50744641e-01 5.97573668e-02
8.05614352e-01 3.06284636e-01 -1.71649940e-02 -3.27161610e-01
6.20222948e-02 -6.65645599e-01 -7.94256032e-01 -2.68243492e-01
-5.94501436e-01 -4.25228953e-01 2.21197546e-01 -1.86743349e-01
8.31833839e-01 -1.68133295e+00 2.18771636e-01 7.01921701e-01
1.64805546e-01 -7.95350969e-02 1.92566320e-01 7.48742402e-01
-3.59443456e-01 3.79338264e-01 -5.43022752e-01 -1.04882739e-01
4.64748085e-01 7.62910470e-02 -1.31006226e-01 6.95954025e-01
3.07370305e-01 1.01637781e+00 -1.24089897e+00 -1.16592675e-01
3.91605705e-01 6.13652050e-01 -2.26406574e-01 -1.97880398e-02
-7.10162103e-01 4.54408348e-01 -6.04645610e-01 9.70102191e-01
5.81212759e-01 -1.45356432e-01 4.37440723e-03 8.00136104e-02
-3.46485257e-01 4.84827757e-01 -1.50287259e+00 1.57510257e+00
-1.60821617e-01 -9.57931131e-02 -1.14782855e-01 -4.46581453e-01
7.21652806e-01 2.86718816e-01 4.27497506e-01 2.17808753e-01
1.83498994e-01 4.96027358e-02 9.56248417e-02 1.67127758e-01
2.38773584e-01 -4.63183850e-01 -1.35517925e-01 3.58526409e-01
2.60126919e-01 -1.16587985e+00 3.46893609e-01 3.05173337e-01
9.04801786e-01 4.04125422e-01 5.90937912e-01 -4.39095318e-01
2.59427518e-01 -1.28023863e-01 1.80288702e-01 8.50935161e-01
-1.42086044e-01 2.98410922e-01 6.50314689e-01 -3.04418057e-01
-9.06238496e-01 -1.48197031e+00 -6.13066196e-01 8.19350839e-01
-4.21204090e-01 -8.19960773e-01 -4.13656652e-01 -2.70377159e-01
2.66168743e-01 7.45638549e-01 -3.34325939e-01 -3.39033455e-02
3.56849283e-01 -1.36440849e+00 5.54026246e-01 2.11075306e-01
-2.34364763e-01 -7.72407353e-01 1.77915454e-01 5.35425484e-01
7.04610765e-01 -5.94898880e-01 -4.54740942e-01 8.56991649e-01
-6.36929095e-01 -1.06791246e+00 -5.09864450e-01 6.05174378e-02
2.71212488e-01 -2.00816974e-01 9.24760640e-01 -8.20562184e-01
-7.44242311e-01 4.38967168e-01 1.80203483e-01 -1.13079584e+00
-6.41451657e-01 -9.42628533e-02 1.88651204e-01 -5.14314473e-01
5.15199780e-01 -6.66023970e-01 -5.98271430e-01 -9.03889015e-02
-8.70961785e-01 -5.18563330e-01 3.73285979e-01 5.09250760e-01
7.97587454e-01 -2.38211211e-02 2.80349344e-01 -1.04913461e+00
7.78990626e-01 -6.22644782e-01 -1.08950889e+00 -3.56510654e-02
-6.41847610e-01 4.47548956e-01 8.97159800e-02 -2.83033252e-01
-7.62020767e-01 3.13204318e-01 -3.25663388e-01 -5.69034517e-02
-4.40761924e-01 7.51302481e-01 -4.33559269e-01 -6.74539506e-02
7.95038402e-01 -5.74298613e-02 8.04310897e-04 -2.41531610e-01
9.50540900e-01 3.63203287e-01 1.93267524e-01 -1.15505087e+00
4.73653525e-01 2.15506315e-01 5.75779974e-01 -8.83328497e-01
-3.96837652e-01 -6.39550149e-01 -5.26969314e-01 2.02191144e-01
9.71495092e-01 -7.87299871e-01 -1.20383704e+00 1.80092081e-01
-1.12904620e+00 -1.85384065e-01 -4.08075929e-01 5.66045165e-01
-7.40859330e-01 4.82789785e-01 -5.56139529e-01 -1.25492275e+00
-2.51539201e-01 -1.43346274e+00 8.49047065e-01 2.22359821e-01
-5.00117481e-01 -1.29387820e+00 2.77167529e-01 -1.89452395e-01
1.73746899e-01 2.64753819e-01 9.99928713e-01 -8.07346344e-01
-6.47159219e-01 -2.72403210e-01 1.12916999e-01 -1.58899277e-02
1.68081731e-01 4.08702612e-01 -1.09693420e+00 1.74038306e-01
-1.88158512e-01 -1.08417556e-01 9.30305839e-01 5.47087133e-01
1.24235642e+00 1.06228575e-01 -4.12012726e-01 6.12213135e-01
1.08319139e+00 3.02440852e-01 6.73255146e-01 1.25408927e-02
6.60231054e-01 4.00797963e-01 7.51310885e-02 6.12150133e-01
-3.32170762e-02 2.23911315e-01 1.63460284e-01 2.74592370e-01
4.05377269e-01 -2.75776476e-01 2.71553934e-01 3.86576772e-01
-1.99320629e-01 -1.76444069e-01 -1.03056073e+00 6.03594221e-02
-1.72313023e+00 -1.04748952e+00 -4.19028401e-01 2.44792604e+00
1.21311903e+00 -4.87021916e-02 3.32324170e-02 -1.25350818e-01
6.13186896e-01 9.03803185e-02 -5.32941699e-01 -6.60364330e-01
1.60281640e-03 7.23267674e-01 9.75258112e-01 8.21573973e-01
-1.30921376e+00 8.98336887e-01 7.28962088e+00 1.13952708e+00
-6.42773092e-01 5.22286929e-02 6.76167786e-01 2.44162157e-02
-4.74341214e-01 3.43510509e-01 -1.08289564e+00 5.12027204e-01
1.36602020e+00 -3.67271036e-01 3.98606896e-01 6.63952529e-01
3.66385311e-01 -3.44998181e-01 -1.25872982e+00 9.23469782e-01
-7.30934083e-01 -1.47667170e+00 -2.22005695e-02 1.37353912e-01
5.34636736e-01 1.15127144e-02 4.33074236e-02 1.97694466e-01
1.07282734e+00 -1.44682193e+00 2.14664966e-01 5.04278004e-01
6.04217291e-01 -9.74602878e-01 4.33403492e-01 -1.04980975e-01
-7.46753633e-01 5.14489591e-01 -5.74931443e-01 2.89696634e-01
1.79943517e-01 1.07947028e+00 -1.08841264e+00 6.05051935e-01
1.69302642e-01 3.97544742e-01 1.83493234e-02 1.34198070e+00
-3.60968769e-01 5.91885626e-01 -6.62859201e-01 -2.67897695e-01
1.96957514e-01 -7.08115637e-01 5.26056051e-01 1.40027344e+00
2.45095953e-01 -2.74493881e-02 -6.23650998e-02 1.20157206e+00
3.71086895e-02 1.05808087e-01 -7.18034685e-01 -7.20859885e-01
4.81508166e-01 1.17243063e+00 -7.99002886e-01 -1.67424623e-02
-1.59294337e-01 5.62632501e-01 -1.85505256e-01 5.80903590e-01
-6.36285484e-01 -5.41871727e-01 1.02701807e+00 -2.92778403e-01
1.27781272e-01 -5.54311037e-01 -2.25125507e-01 -7.53587902e-01
-6.56079054e-01 -6.81992114e-01 3.25339288e-01 -5.24559259e-01
-1.19672847e+00 -1.02586955e-01 2.05003813e-01 -3.55801344e-01
-1.65663838e-01 -1.17250061e+00 -5.75034916e-01 1.48091173e+00
-1.11399913e+00 -7.13430941e-01 5.23952782e-01 3.05178791e-01
1.43435225e-01 3.24055813e-02 1.14283907e+00 -3.04634273e-01
-5.43315113e-01 -5.21881729e-02 5.98686814e-01 -6.28360689e-01
6.93829536e-01 -1.71904755e+00 8.10042381e-01 5.66798151e-01
1.90939844e-01 1.15732980e+00 1.05279338e+00 -7.96897471e-01
-1.83966470e+00 -9.18050468e-01 2.50541002e-01 -8.45042706e-01
1.17403579e+00 -5.94526291e-01 -7.72959113e-01 3.45976144e-01
-1.89706177e-01 -2.06329048e-01 1.12575972e+00 3.07203770e-01
-3.70800018e-01 4.93712157e-01 -1.11933112e+00 6.69847369e-01
7.26037860e-01 -7.07563937e-01 -1.65862724e-01 7.89686322e-01
6.59987926e-01 -4.63990957e-01 -1.40971339e+00 6.96627647e-02
3.62749219e-01 -3.25669676e-01 1.17309272e+00 -7.66871929e-01
-3.26466886e-03 -5.52822828e-01 -1.63034633e-01 -1.47897553e+00
-2.64142007e-01 -1.30836630e+00 -4.18930441e-01 8.62584293e-01
7.06050515e-01 -8.95958960e-01 6.85800076e-01 1.01644242e+00
-9.19779241e-02 -6.63997829e-01 -8.77083182e-01 -7.76616454e-01
5.00176072e-01 -7.08494484e-01 7.16565013e-01 4.36130702e-01
2.08471209e-01 2.22467944e-01 1.84484974e-01 3.07923555e-01
8.37039948e-01 -2.98267603e-01 4.29508328e-01 -1.22850037e+00
-3.86404455e-01 -7.00777709e-01 -4.57104862e-01 -7.38437235e-01
-1.09599814e-01 -9.94082153e-01 2.65949339e-01 -1.36419404e+00
9.92377922e-02 -4.78914499e-01 -2.58084536e-01 1.56876087e-01
-2.28985399e-01 -1.70134485e-01 -1.41689166e-01 -9.72457007e-02
-3.88690591e-01 6.16193771e-01 9.61641550e-01 -2.64548630e-01
-1.34244218e-01 3.97158116e-01 -8.56575072e-01 4.25584555e-01
1.02798617e+00 -5.55964470e-01 -3.18642646e-01 4.21983570e-01
4.67919201e-01 -1.72077343e-01 1.45655453e-01 -6.20898247e-01
2.83165514e-01 -4.31423962e-01 5.62953293e-01 -4.46074635e-01
4.15541977e-01 -2.71300018e-01 6.68767750e-01 1.42081127e-01
-2.79944479e-01 -8.11320916e-02 4.31576878e-01 8.54384005e-01
2.03162953e-01 -6.13110721e-01 6.69713259e-01 -4.24772501e-01
-3.10849845e-01 7.25235760e-01 -6.47272289e-01 -2.50921607e-01
8.27276349e-01 3.28495204e-02 -2.37387568e-01 -1.68300837e-01
-1.10895789e+00 1.89501002e-01 4.82741565e-01 -7.49519989e-02
4.31774765e-01 -9.31468189e-01 -3.38999599e-01 -1.41604692e-01
2.17545122e-01 3.07605922e-01 8.15786049e-02 6.71855330e-01
-3.91760081e-01 6.87952936e-01 3.22940916e-01 -4.21191245e-01
-8.94503236e-01 7.61664808e-01 3.26875687e-01 1.45055344e-02
2.00761389e-02 9.61637080e-01 1.90899983e-01 -4.33756083e-01
6.74065426e-02 -4.89939064e-01 2.94215024e-01 -1.85551107e-01
6.30633354e-01 2.69271731e-01 1.67248473e-01 -9.86244436e-03
-3.85682434e-01 9.84904394e-02 -1.38334721e-01 -2.15496376e-01
1.26151216e+00 1.89266860e-01 -1.93388909e-01 6.19588614e-01
6.93761945e-01 1.12714835e-01 -1.23213458e+00 1.33673906e-01
3.77047002e-01 -7.89709911e-02 1.25552222e-01 -6.54107571e-01
-1.46072671e-01 9.59899664e-01 2.99603015e-01 -7.44871423e-02
1.08075015e-01 1.28171906e-01 1.43392701e-02 7.32950509e-01
3.66724372e-01 -6.31160080e-01 -3.90618414e-01 6.66554153e-01
7.70535529e-01 -1.15522730e+00 3.04157257e-01 -5.26548803e-01
-4.94193763e-01 1.21539390e+00 -2.96011031e-01 3.08292568e-01
8.49934697e-01 3.73816907e-01 -3.08253199e-01 -3.34568113e-01
-7.45478034e-01 -2.46511146e-01 7.72721618e-02 9.13298726e-01
7.87782371e-01 3.99718851e-01 9.78106484e-02 4.10152435e-01
5.03380895e-02 -1.95262060e-01 5.82840025e-01 1.05578244e+00
-4.82365221e-01 -1.63242900e+00 -6.02817774e-01 4.84969497e-01
-6.48880064e-01 -6.35498047e-01 -1.42627776e-01 3.19888920e-01
-3.64847720e-01 8.01680624e-01 -4.00011241e-01 3.77479166e-01
1.19320475e-01 5.54643333e-01 6.79184914e-01 -9.43939209e-01
-9.45206657e-02 5.03412969e-02 3.37421298e-01 -3.58888745e-01
-1.90275107e-02 -9.36300755e-01 -1.11410606e+00 -4.75885302e-01
-1.48834929e-01 6.45844758e-01 1.10073996e+00 5.06260037e-01
3.97249848e-01 3.42406839e-01 1.91373155e-02 -9.00439262e-01
-7.25758851e-01 -8.44998419e-01 -8.32907617e-01 -3.56519908e-01
2.28883941e-02 -7.79643476e-01 -3.51238370e-01 2.31290143e-02] | [6.3898539543151855, 4.293345928192139] |
5936a411-78e1-4ead-bdb0-6217904a8505 | a-dense-cnn-approach-for-skin-lesion | 1807.06416 | null | http://arxiv.org/abs/1807.06416v2 | http://arxiv.org/pdf/1807.06416v2.pdf | A Dense CNN approach for skin lesion classification | This article presents a Deep CNN, based on the DenseNet architecture jointly
with a highly discriminating learning methodology, in order to classify seven
kinds of skin lesions: Melanoma, Melanocytic nevus, Basal cell carcinoma,
Actinic keratosis / Bowen's disease, Benign keratosis, Dermatofibroma, Vascular
lesion. In particular a 61 layers DenseNet, pre-trained on IMAGENET dataset,
has been fine-tuned on ISIC 2018 Task 3 Challenge Dataset exploiting a Center
Loss function. | ['Pierluigi Carcagnì', 'Cosimo Distante', 'Andrea Cuna'] | 2018-07-17 | null | null | null | null | ['skin-lesion-classification'] | ['medical'] | [ 3.79956990e-01 1.29343435e-01 -1.40774667e-01 -1.06549725e-01
-1.81210503e-01 -2.44567111e-01 9.07805800e-01 -1.06182888e-01
-6.16512299e-01 9.52125847e-01 2.73101270e-01 -4.69326913e-01
-3.70397955e-01 -5.30230343e-01 -2.39836469e-01 -8.10329378e-01
-1.01766303e-01 1.99088052e-01 -2.71775424e-02 -5.45277409e-02
7.02567697e-02 7.57337868e-01 -1.16057837e+00 3.23335856e-01
8.97152722e-01 1.19301724e+00 -3.24106663e-01 1.02976990e+00
2.73760915e-01 1.07042229e+00 -2.79796839e-01 -6.66868985e-01
2.15378612e-01 -4.84266691e-02 -9.32442665e-01 4.63383347e-02
9.91179883e-01 -7.38271698e-02 -2.98467427e-01 1.01939011e+00
5.58777630e-01 -4.05002862e-01 1.13336599e+00 -7.45362818e-01
-5.25282621e-01 1.26195565e-01 -9.03219819e-01 3.11430305e-01
1.76059246e-01 5.85375667e-01 6.02615476e-01 -3.55888814e-01
1.00892913e+00 9.93917584e-01 1.21244836e+00 8.23781848e-01
-7.06303358e-01 -5.59775293e-01 -2.66140997e-01 4.08597887e-01
-1.33744037e+00 1.41157001e-01 8.21314678e-02 -3.69742692e-01
1.01319361e+00 3.65920335e-01 9.90825415e-01 1.80049860e+00
6.45693481e-01 5.28582454e-01 1.59537911e+00 -2.73642898e-01
1.01611741e-01 -6.74958229e-02 6.44385815e-02 9.38035548e-01
2.08501160e-01 4.24644500e-01 9.90186036e-02 1.92410842e-01
8.45121324e-01 -1.49603831e-02 -8.39241073e-02 1.50793314e-01
-6.43981695e-01 6.73635364e-01 7.72846043e-01 1.25086114e-01
-5.43740988e-01 9.90745798e-02 5.92358947e-01 4.76095796e-01
1.34409294e-01 1.18951552e-01 -2.98464775e-01 1.53859854e-01
-6.55737102e-01 1.33930877e-01 8.96476507e-01 2.39312395e-01
5.35246313e-01 -3.77610981e-01 -2.83332914e-01 9.21464682e-01
2.93872207e-01 -3.98341641e-02 7.14158595e-01 -9.06086341e-02
-1.58319533e-01 8.69309664e-01 -7.67824173e-01 -2.12206364e-01
-5.85049510e-01 -6.73082054e-01 -1.48502100e+00 5.48677504e-01
2.05017850e-01 -3.55831474e-01 -1.95041132e+00 1.16814268e+00
2.15189874e-01 5.90960979e-01 2.23895803e-01 1.14966106e+00
1.10901618e+00 -1.39824703e-01 4.21493351e-01 5.10625303e-01
1.38513291e+00 -1.21598887e+00 -4.42195296e-01 2.79522777e-01
3.85973752e-01 -7.56962180e-01 7.51888335e-01 6.33457303e-01
-8.64919364e-01 -1.87737092e-01 -8.99548531e-01 -3.72740716e-01
-7.76818156e-01 1.10159591e-01 1.02707255e+00 5.43261409e-01
-1.33303201e+00 2.97195256e-01 -6.84740245e-01 -1.07632327e+00
9.14098680e-01 3.56925279e-01 -8.34129572e-01 -7.81459063e-02
-1.02224863e+00 1.01367486e+00 5.47371149e-01 2.10575312e-01
-9.84411240e-01 -1.00279582e+00 -5.34347236e-01 -1.65130988e-01
-1.12955339e-01 -1.19338405e+00 8.71478975e-01 -1.42019606e+00
-1.51364934e+00 1.44556749e+00 4.44620669e-01 -1.01992512e+00
7.65365601e-01 1.24672957e-01 -1.05994689e+00 1.16732061e-01
-2.71622926e-01 6.99227870e-01 6.17013395e-01 -5.90842068e-01
-8.33527505e-01 -3.41424644e-01 1.83559299e-01 1.14170108e-02
5.23835048e-02 -3.18102241e-01 -2.56513238e-01 -4.09666091e-01
-7.49964416e-01 -7.37136245e-01 -5.60964406e-01 1.41806811e-01
-1.27270555e+00 -3.27654779e-01 7.47738063e-01 -5.78714490e-01
8.50511909e-01 -2.06247711e+00 9.22005698e-02 5.80882370e-01
3.76955152e-01 7.10908473e-01 -4.24435079e-01 3.71477574e-01
-1.43928453e-01 3.52680355e-01 -1.98829085e-01 4.10681069e-02
-1.11212600e-02 -7.52711147e-02 5.66746831e-01 3.98707241e-01
5.32164454e-01 1.10596836e+00 -5.47159076e-01 -3.59932214e-01
5.80616593e-01 7.08108723e-01 1.56292822e-02 -1.82668820e-01
-7.41085336e-02 -1.13157861e-01 1.44281834e-01 1.19662070e+00
8.91487062e-01 -5.19443274e-01 -5.27227037e-02 -3.97263110e-01
1.13445275e-01 -3.92724872e-01 -6.37780368e-01 1.74623883e+00
-7.64192104e-01 8.27874303e-01 5.73243320e-01 -4.67369825e-01
4.17052478e-01 -1.37650976e-02 8.02937225e-02 -8.92536402e-01
3.11879098e-01 1.68674201e-01 2.89765775e-01 -8.57103348e-01
-2.50424683e-01 -1.53110981e-01 5.81525922e-01 -2.74214000e-01
3.60408634e-01 2.43287846e-01 5.91901004e-01 1.15330495e-01
1.56949341e+00 -3.83616537e-01 4.98975843e-01 -3.11272264e-01
8.87079060e-01 -9.22263190e-02 1.42438933e-01 3.70121688e-01
-4.11463141e-01 7.31481671e-01 8.31255794e-01 -8.23737562e-01
-9.86187577e-01 -1.39021409e+00 -6.33686423e-01 3.88532281e-01
-2.53637195e-01 5.03519736e-03 -5.47216773e-01 -9.75281715e-01
4.21671212e-01 1.31553188e-02 -1.19870996e+00 -3.55665907e-02
-6.90626577e-02 -8.91079903e-01 8.31318915e-01 3.60808939e-01
1.05997968e+00 -9.24928844e-01 3.22245248e-02 3.11536947e-04
7.57777154e-01 -7.16024816e-01 4.80207801e-02 3.88275295e-01
-1.28812030e-01 -1.96755195e+00 -8.29582691e-01 -1.00016117e+00
7.31258869e-01 -6.06471241e-01 1.01147044e+00 -1.17450356e-01
-1.39325464e+00 1.40451923e-01 -1.76619291e-01 -2.94538110e-01
-2.42032811e-01 2.96882600e-01 -2.84739763e-01 2.90664643e-01
8.20435226e-01 -4.90824997e-01 -9.13637757e-01 -1.64028436e-01
-1.03650427e+00 3.03543396e-02 1.40873075e+00 1.02433991e+00
5.89561999e-01 8.79646093e-02 3.01039308e-01 -1.49529076e+00
7.24711180e-01 -7.96958268e-01 -2.42376491e-01 1.32785738e-01
-3.66991103e-01 -4.95158851e-01 6.08387768e-01 -2.62152135e-01
-9.50615168e-01 3.06891382e-01 -7.20247090e-01 -2.86169380e-01
-6.12708271e-01 6.58550501e-01 3.33768934e-01 -6.15684628e-01
1.16267359e+00 8.94172564e-02 2.10686177e-01 -2.97889531e-01
9.71892253e-02 7.77409792e-01 8.76218617e-01 2.57256597e-01
5.16280830e-01 6.22054875e-01 4.15882289e-01 -9.53596830e-01
-9.28000212e-01 -6.30048990e-01 -4.11278456e-01 4.12446670e-02
1.11836767e+00 -1.12524021e+00 -4.36421454e-01 1.02608895e+00
-7.25253701e-01 -3.89830709e-01 -3.57960403e-01 3.70219171e-01
-1.29503474e-01 9.12542567e-02 -8.58694673e-01 -7.06988946e-02
-6.13885760e-01 -5.72275341e-01 1.04288983e+00 7.78943479e-01
-6.04323260e-02 -1.73029745e+00 4.15432632e-01 1.86424211e-01
7.10572183e-01 9.62854803e-01 9.41821635e-01 -5.96457124e-01
-2.43090034e-01 -4.05286580e-01 -7.28159547e-01 8.18785548e-01
1.62251785e-01 4.93327916e-01 -8.64630759e-01 -3.82360786e-01
-9.32375312e-01 -8.34520102e-01 1.65749824e+00 6.32887959e-01
1.39180279e+00 -1.83945641e-01 -3.07196915e-01 1.56349766e+00
2.08758044e+00 -8.75292122e-02 8.57494950e-01 3.94560874e-01
7.49305248e-01 3.87246981e-02 -2.78375238e-01 1.15488544e-01
4.35367674e-02 -3.83052826e-02 7.91497409e-01 -8.07188749e-01
-6.85745835e-01 6.91854954e-02 -5.07502794e-01 -9.50944796e-03
-4.71222728e-01 -3.70271862e-01 -9.02160048e-01 3.83583784e-01
-1.12830126e+00 -4.59650397e-01 -7.04610124e-02 1.48902619e+00
5.42014241e-01 -6.11980520e-02 -2.04795003e-01 -9.19454396e-02
5.98384380e-01 1.32056683e-01 -7.55989075e-01 -7.02577055e-01
-5.63162625e-01 8.92635703e-01 8.89154553e-01 2.56521702e-01
-1.42960632e+00 1.02385676e+00 6.21906948e+00 9.29171443e-01
-1.90794325e+00 -9.01237801e-02 5.79326272e-01 1.02846012e-01
-1.34775743e-01 -6.02628291e-01 -5.30828476e-01 4.42121446e-01
7.83110976e-01 -1.48347795e-01 6.32523745e-02 6.10561252e-01
-3.24709445e-01 -9.34882239e-02 -8.43634903e-01 1.02093089e+00
4.53370959e-02 -1.75322735e+00 2.56022304e-01 1.81732938e-01
9.68750060e-01 3.58195007e-01 4.72239584e-01 4.07868683e-01
6.20273113e-01 -1.54906619e+00 -4.70661253e-01 8.12380552e-01
1.30777156e+00 -7.99242377e-01 1.17734802e+00 -1.38106614e-01
-4.62303758e-01 -5.81023283e-02 -3.40000063e-01 1.50832340e-01
-2.29454666e-01 6.18079484e-01 -1.01483059e+00 3.63339126e-01
5.71809769e-01 1.09916711e+00 -9.91894841e-01 1.55633831e+00
-2.72156864e-01 8.11373472e-01 -1.81967571e-01 -1.43966466e-01
9.86586392e-01 2.28157878e-01 3.53439718e-01 1.49023283e+00
2.08350290e-02 -5.11304319e-01 -1.60760164e-01 6.16630018e-01
-1.08282238e-01 1.86331525e-01 -5.64908683e-01 -9.35179293e-02
-3.19845751e-02 1.96152747e+00 -3.79039466e-01 5.37812011e-03
-1.84245020e-01 8.30908656e-01 5.72104678e-02 6.01282299e-01
-4.88870293e-01 -7.88961589e-01 6.28794611e-01 1.36584669e-01
2.30258733e-01 6.71956480e-01 -3.72600779e-02 -1.03127646e+00
-4.75216359e-01 -5.82208395e-01 4.20364827e-01 -6.12260044e-01
-1.76254904e+00 6.81687176e-01 -7.31429756e-01 -8.08224201e-01
-2.81294525e-01 -1.22020626e+00 -1.30696559e+00 1.05456269e+00
-2.00954843e+00 -1.77217710e+00 -7.89703846e-01 9.81154799e-01
1.08079344e-01 -7.98291326e-01 7.87421525e-01 1.45639330e-01
-7.23625124e-01 6.56707764e-01 7.57970512e-02 1.76118359e-01
7.68574476e-01 -1.88860595e+00 1.75587490e-01 3.21589410e-01
-6.99944377e-01 3.82049143e-01 -4.08934094e-02 -3.02247375e-01
-1.08683825e+00 -1.65563214e+00 2.91366398e-01 2.61192560e-01
9.80406344e-01 -7.83951283e-02 -4.32926446e-01 3.34666401e-01
7.15245008e-01 4.73084718e-01 1.14221144e+00 2.18254879e-01
-4.68941480e-01 -1.89373165e-01 -1.47654331e+00 3.59620661e-01
6.36628687e-01 -2.67685801e-01 -1.58229649e-01 9.54528987e-01
1.70177475e-01 -5.91911137e-01 -1.11356449e+00 4.99033928e-01
5.85647523e-01 -1.02174413e+00 9.18297768e-01 -9.79221463e-01
9.78010952e-01 2.59469748e-01 2.97558159e-01 -1.38378692e+00
-7.11233854e-01 -5.93899727e-01 8.34726077e-03 6.27789795e-01
3.53682518e-01 -5.36859870e-01 1.19180965e+00 -2.98646241e-01
-4.13782209e-01 -1.10225439e+00 -9.49493587e-01 -5.91621995e-01
3.74979526e-01 3.12104315e-01 3.06445122e-01 7.76623666e-01
-7.57013023e-01 4.49756294e-01 -2.95281820e-02 -2.89999127e-01
6.89942539e-01 -6.74456298e-01 5.22158146e-01 -1.40937126e+00
2.25837439e-01 -8.80776525e-01 -1.13815713e+00 -1.45216197e-01
2.41309181e-01 -9.40503895e-01 -8.01769853e-01 -1.73398149e+00
-1.81441829e-02 -2.33935624e-01 -8.31170499e-01 4.05032843e-01
2.55553514e-01 6.80533707e-01 -2.62433171e-01 -3.60926777e-01
-4.27266955e-01 -3.05376470e-01 1.56980348e+00 -5.04140913e-01
2.48702422e-01 2.22971365e-01 -7.77946830e-01 8.96917820e-01
4.09923971e-01 4.58649099e-01 -2.31512293e-01 -1.50079966e-01
2.14413553e-01 -2.49185845e-01 7.42397726e-01 -1.31910694e+00
5.07872820e-01 -7.93201327e-02 8.92189205e-01 -6.59765124e-01
1.83614880e-01 -6.66966498e-01 -5.99121209e-03 7.87321091e-01
-4.58109200e-01 -8.16779613e-01 2.83008575e-01 5.73120952e-01
-6.38274252e-01 -2.27566268e-02 1.27543092e+00 -1.35689944e-01
-1.17423463e+00 8.51061702e-01 -4.76860732e-01 -2.54589528e-01
1.49690473e+00 -2.54827529e-01 -6.05625629e-01 2.63903320e-01
-1.05837977e+00 3.54669183e-01 2.84976959e-01 1.13488279e-01
6.25800252e-01 -1.33067656e+00 -9.63366210e-01 3.20707083e-01
3.94899964e-01 1.24435380e-01 8.76719415e-01 9.65794027e-01
-1.24964643e+00 5.01964331e-01 -8.71446013e-01 -5.12330532e-01
-1.17390811e+00 1.69391918e-03 8.19401205e-01 -5.76012790e-01
-4.74719137e-01 1.45405388e+00 -2.55323499e-01 -4.38254416e-01
4.75356102e-01 -3.91678870e-01 -4.55935329e-01 -1.49969265e-01
3.72825712e-01 2.23504096e-01 1.38281241e-01 -8.20096284e-02
-3.02145243e-01 4.93414551e-01 -8.30265403e-01 4.44395870e-01
9.91845131e-01 5.61672568e-01 -3.05292726e-01 -2.39093363e-01
1.46366727e+00 -5.75374663e-01 -1.09810710e+00 -1.20889738e-01
-2.80215681e-01 -1.21850044e-01 3.90820503e-01 -1.89395952e+00
-1.17505062e+00 6.47138894e-01 1.00345147e+00 1.75118044e-01
1.50885808e+00 -4.56631124e-01 6.55569792e-01 4.42735285e-01
-2.70189017e-01 -1.06749034e+00 -2.51391172e-01 4.71607238e-01
7.06913888e-01 -1.47136712e+00 -9.59722102e-02 -3.09646487e-01
-4.24331337e-01 1.59355044e+00 8.87344658e-01 -7.66212225e-01
9.55769837e-01 4.62386996e-01 2.04389811e-01 -2.95746535e-01
-1.26255405e+00 -5.63497424e-01 3.51860017e-01 9.37655747e-01
3.45798969e-01 2.38723680e-01 -4.08281296e-01 2.02237517e-01
-4.81040329e-02 4.60942596e-01 6.24419272e-01 4.91348028e-01
-3.39726284e-02 -7.55322516e-01 3.88700962e-01 1.20460117e+00
-6.86656535e-01 -1.58198729e-01 -9.72953737e-01 1.43500710e+00
5.47548413e-01 -4.56219576e-02 4.30270255e-01 -1.87391073e-01
1.19431540e-01 -5.48031926e-01 4.14838731e-01 -6.94981635e-01
-1.05653477e+00 -9.89427641e-02 2.11922243e-01 -4.14212584e-01
-3.73828709e-01 -2.06506744e-01 -5.51110625e-01 -2.74171650e-01
2.42658481e-01 -4.12398994e-01 9.86612320e-01 5.66137850e-01
-7.67352954e-02 5.23136437e-01 4.40427452e-01 -2.12206692e-02
-5.05499840e-01 -1.26961219e+00 -9.40705240e-01 6.09728754e-01
5.08955300e-01 -3.18238825e-01 -5.06081939e-01 -1.19779512e-01] | [15.700451850891113, -3.0154011249542236] |
fbbd1562-d2c2-4af5-acfa-b6716f151a2e | predicting-grammaticality-on-an-ordinal-scale | null | null | https://aclanthology.info/papers/P14-2029/p14-2029 | https://www.aclweb.org/anthology/P14-2029 | Predicting Grammaticality on an Ordinal Scale | null | ['Melissa Lopez', 'Nitin Madnani', 'Matthew Mulholland', 'Joel Tetreault', 'Aoife Cahill', 'Michael Heilman'] | 2014-06-01 | null | https://aclanthology.org/P14-2029 | https://aclanthology.org/P14-2029.pdf | acl-2014-6 | ['automated-essay-scoring'] | ['natural-language-processing'] | [-2.44508207e-01 3.89024585e-01 -2.65282035e-01 -2.15905145e-01
-8.60921741e-02 -7.76765764e-01 4.48510379e-01 -7.23253429e-01
-5.48377395e-01 1.31954515e+00 3.66348401e-02 -9.49533224e-01
-2.40340635e-01 -1.05564880e+00 -8.44053447e-01 -8.75781775e-01
-7.42435038e-01 6.86515033e-01 1.44298598e-01 -6.52004302e-01
8.47113907e-01 6.29996777e-01 -1.62287033e+00 5.77558100e-01
6.81926727e-01 5.49597681e-01 5.66466339e-02 1.04480565e+00
6.13576770e-02 1.59847176e+00 -6.05450153e-01 -4.56729174e-01
1.43710867e-01 -1.88022718e-01 -5.35770595e-01 -3.36825520e-01
7.91536197e-02 -5.88986814e-01 -3.10633808e-01 6.07356608e-01
1.14270830e+00 -5.53557873e-02 1.07025254e+00 -1.51067197e+00
-7.97060013e-01 5.44234335e-01 2.30399013e-01 1.20632313e-01
6.10446513e-01 -1.79472014e-01 3.51212233e-01 -1.36621380e+00
6.32680655e-01 8.64870071e-01 9.47046995e-01 5.99273622e-01
-1.20647264e+00 -5.60552299e-01 -7.17137158e-01 -2.57835120e-01
-1.59567785e+00 -6.30042374e-01 6.23617955e-02 -3.34735900e-01
1.69142139e+00 6.72587395e-01 1.35647905e+00 1.47341585e+00
1.36649036e+00 4.72517729e-01 1.13904178e+00 -3.16340059e-01
3.65351558e-01 6.43754780e-01 -2.29148138e-02 5.51889002e-01
1.19686353e+00 6.76312149e-01 -5.60917914e-01 -7.92779744e-01
1.11365557e+00 -3.12949389e-01 2.60771513e-01 -7.06989706e-01
-1.06566191e+00 7.88466871e-01 -1.33808568e-01 2.20882341e-01
-3.46203089e-01 -9.92989019e-02 2.24548295e-01 4.12961632e-01
-2.94082850e-01 4.77862865e-01 -8.89795244e-01 -1.87686309e-01
-1.01472771e+00 1.91863477e-01 1.30615628e+00 1.45811689e+00
4.72722203e-03 3.70100170e-01 1.66493103e-01 2.89523482e-01
5.29429734e-01 1.05927479e+00 4.36951250e-01 -1.33365822e+00
-9.55119133e-02 4.20878716e-02 6.12505674e-01 -1.13473701e+00
-7.12930143e-01 -2.41300568e-01 -9.54724610e-01 5.18771529e-01
-1.59270123e-01 -4.59180593e-01 -7.59686172e-01 4.79171664e-01
-1.81984559e-01 -3.19489211e-01 5.45619786e-01 2.69257158e-01
7.28577793e-01 1.45816103e-01 9.25893057e-03 -4.73371416e-01
8.38296890e-01 -1.26899445e+00 -1.11426139e+00 2.33304992e-01
6.82289064e-01 -1.11704338e+00 4.06794518e-01 4.29345578e-01
-1.87200487e+00 -9.69799384e-02 -1.08317137e+00 1.19022481e-01
-7.97166348e-01 -2.57194549e-01 6.66624963e-01 1.49922979e+00
-1.63972652e+00 6.47176981e-01 -6.05021298e-01 5.40335439e-02
-4.73218322e-01 9.07614470e-01 -3.47142309e-01 2.90655226e-01
-1.28751957e+00 9.52449858e-01 -1.04219764e-01 3.03568155e-01
-3.30324143e-01 -1.84501112e-01 -1.03385210e+00 -6.24127567e-01
-2.59417385e-01 -1.07312346e+00 1.36302292e+00 1.09815992e-01
-1.30379498e+00 1.00962746e+00 -4.14315462e-01 -1.14109404e-01
6.61359608e-01 -1.52096570e-01 -8.18174660e-01 1.58321753e-01
2.87593510e-02 3.47163439e-01 5.30214727e-01 -9.69232559e-01
-6.31655395e-01 -1.00153096e-01 -2.54188716e-01 2.48816490e-01
1.43904146e-03 2.05205917e-01 6.38460100e-01 -9.53604504e-02
1.91062525e-01 -8.61961663e-01 -3.21419090e-01 -7.61404037e-01
-1.77353188e-01 -3.84736449e-01 5.59762537e-01 -2.92163283e-01
1.33848476e+00 -1.84365213e+00 -4.54683900e-01 2.94456817e-03
2.53527164e-01 3.89016122e-02 2.33101144e-01 1.23831403e+00
-2.43849114e-01 1.06667292e+00 3.98064762e-01 9.52541176e-03
3.13382268e-01 5.07273376e-01 2.86674555e-02 2.51206756e-01
9.43330750e-02 1.15280068e+00 -1.09214199e+00 -5.20940363e-01
5.57267785e-01 4.60968643e-01 -3.84588838e-01 6.24593735e-01
9.30012882e-01 2.99172718e-02 -8.94021094e-02 1.33241999e+00
1.05377018e+00 -2.90812284e-01 -2.46628653e-02 3.44544262e-01
-6.44541264e-01 6.00222833e-02 -7.41804183e-01 9.49429095e-01
1.31349964e-02 1.99757457e-01 1.74998537e-01 -6.80728197e-01
3.94654661e-01 1.10545933e+00 4.45659161e-01 -9.24297392e-01
-1.69569388e-01 9.00310338e-01 1.03920102e-01 -9.62983370e-01
6.40755475e-01 -1.72503099e-01 7.69825354e-02 1.40852645e-01
-3.61913115e-01 -3.84120375e-01 5.52458428e-02 4.19574350e-01
1.00224769e+00 -1.33707494e-01 5.48685491e-01 -8.72328460e-01
6.88946426e-01 -2.01589778e-01 3.00597936e-01 1.23978651e+00
-2.90862411e-01 9.21478033e-01 3.19022417e-01 -5.37574828e-01
-5.72780192e-01 -1.20098352e+00 -4.80556756e-01 9.86326516e-01
-1.58388391e-01 -3.87671381e-01 -7.10155964e-01 -2.63155133e-01
-1.85071185e-01 4.93483543e-01 -5.93591571e-01 5.23159266e-01
-5.14705896e-01 -6.85349941e-01 7.65277922e-01 3.17972392e-01
6.49624616e-02 -1.49228525e+00 -5.70813596e-01 3.34754556e-01
-1.07132711e-01 -5.39811492e-01 -1.57590687e-01 6.12205923e-01
-1.19366693e+00 -1.01526415e+00 -8.75530541e-01 -1.03603637e+00
7.18889177e-01 2.44572356e-01 1.18463767e+00 2.84042627e-01
-2.87709564e-01 6.49904788e-01 -1.31030381e-02 -2.67939389e-01
-3.16364132e-02 -5.78750819e-02 5.01899779e-01 -1.07851994e+00
6.46393359e-01 -3.92214775e-01 -1.00622165e+00 5.35368681e-01
-3.85821253e-01 -2.04611585e-01 7.85597622e-01 1.04281425e+00
1.42742008e-01 -1.33182362e-01 2.41971418e-01 -1.16735983e+00
1.11498392e+00 -4.85333532e-01 1.30341025e-02 2.62960047e-01
-8.53362501e-01 -4.53584343e-01 2.27881029e-01 -1.33037493e-02
-6.33760452e-01 -2.51423597e-01 -4.14545417e-01 3.22794706e-01
-4.06866640e-01 3.34393084e-02 2.24873364e-01 -5.26000679e-01
5.21005154e-01 1.50258854e-01 -4.07556817e-02 1.21735156e-01
6.30975887e-02 3.75316888e-01 3.67786407e-01 -7.01224983e-01
5.19809723e-01 1.93325520e-01 -3.00995037e-02 -1.28791058e+00
2.66786069e-01 -1.64763972e-01 -6.60750508e-01 -2.69126564e-01
7.75121748e-01 -9.50198710e-01 -6.35705650e-01 5.21708906e-01
-9.54387486e-01 -1.36340424e-01 -5.25335252e-01 6.30965471e-01
-7.02474535e-01 2.94757709e-02 -4.17852998e-01 -1.17976546e+00
-4.78847146e-01 -1.38869667e+00 9.59160328e-01 3.85633826e-01
-2.05811903e-01 -1.19790471e+00 3.38975668e-01 8.99848714e-02
4.29175526e-01 7.39965662e-02 5.03531873e-01 -2.80004859e-01
-2.98304528e-01 -2.85470933e-01 1.59573574e-02 -1.65876746e-01
-3.09365243e-02 4.80770379e-01 -5.48391879e-01 -4.37962860e-01
2.44738594e-01 -4.86500829e-01 -7.51581416e-02 7.11018145e-01
4.09077644e-01 -6.02920949e-01 -7.12544680e-01 5.56552589e-01
1.27709007e+00 4.72497940e-01 5.88340998e-01 6.77552521e-01
9.28869769e-02 5.17871320e-01 4.48484391e-01 2.08113909e-01
1.07506551e-01 1.08483166e-01 1.14217520e-01 -2.17137560e-01
9.45900455e-02 -1.75644696e-01 4.44304377e-01 1.09980536e+00
-8.23837519e-01 -2.15223029e-01 -4.88283783e-01 6.68056428e-01
-1.69156313e+00 -1.31570101e+00 -6.70152605e-01 9.76623476e-01
3.78074080e-01 3.38884830e-01 1.58041969e-01 1.21232189e-01
5.47049284e-01 -4.76693302e-01 -1.27305150e-01 -9.16352808e-01
-3.82233024e-01 4.21136379e-01 7.14838982e-01 6.51479006e-01
-6.10602558e-01 5.08361042e-01 1.25505590e+01 5.49717486e-01
-5.10860123e-02 1.27314463e-01 3.14815581e-01 1.99914098e-01
-2.24020526e-01 1.62632689e-01 -1.04165065e+00 9.26906094e-02
1.64754760e+00 -4.36735392e-01 3.93212438e-01 6.44444764e-01
3.94769877e-01 -1.96429595e-01 -9.78620768e-01 5.21999180e-01
1.12709433e-01 -1.50397491e+00 -3.76861334e-01 9.64888573e-01
5.80093563e-01 -1.81895152e-01 6.81941509e-01 4.33209062e-01
7.39425182e-01 -1.14519978e+00 7.88427055e-01 3.18626672e-01
8.87266219e-01 -8.18826377e-01 1.07315242e+00 2.76328743e-01
-1.04355454e+00 1.15487054e-02 -5.90443075e-01 -9.22581375e-01
1.62244976e-01 2.02188566e-01 -5.61583102e-01 4.06869203e-01
6.19966745e-01 2.15250537e-01 -8.53338003e-01 1.45581341e+00
-4.29231405e-01 4.22687940e-02 -3.00976187e-01 -4.39922899e-01
4.83290881e-01 5.28384298e-02 3.26606274e-01 1.02295005e+00
2.89785862e-01 5.45581102e-01 -2.62208551e-01 2.92286038e-01
6.63503349e-01 5.84383309e-03 -1.38496268e+00 -8.68683681e-02
3.54620427e-01 9.69217539e-01 -3.99494410e-01 -3.56748998e-02
-6.11456513e-01 6.62115753e-01 -1.66942239e-01 5.75146377e-01
-2.92761147e-01 -8.89406264e-01 4.33969021e-01 -2.30117049e-02
-2.88469017e-01 -1.35338321e-01 -3.23372036e-01 -1.02459288e+00
-6.06416404e-01 -3.90122980e-01 7.32179061e-02 -8.63867998e-01
-1.79313409e+00 6.85844719e-01 -3.50327820e-01 -1.03178740e+00
-9.29034173e-01 -9.72984850e-01 -2.90242791e-01 6.67755604e-01
-6.59234166e-01 -1.30476952e+00 1.77773401e-01 4.65253919e-01
1.82401046e-01 -5.88519633e-01 1.09945750e+00 -1.42670095e-01
-3.18711132e-01 1.08882773e+00 3.51253241e-01 -3.05974871e-01
8.32965672e-01 -1.16271269e+00 4.23785180e-01 1.09189451e-02
-6.59259140e-01 1.28460062e+00 6.81860268e-01 -6.81785107e-01
-1.16639841e+00 -1.52893141e-01 1.38110602e+00 -9.15031374e-01
5.27874053e-01 -1.02810413e-01 1.46300867e-01 6.57190859e-01
9.07351732e-01 -5.09461224e-01 1.08935249e+00 -1.13616258e-01
4.93123859e-01 8.23448122e-01 -1.38805819e+00 5.99272728e-01
1.21222687e+00 -4.81854081e-01 -8.27020168e-01 3.28937650e-01
9.94227231e-01 -5.60786903e-01 -1.34066856e+00 6.99696779e-01
1.14132881e+00 -1.04883564e+00 1.54237401e+00 -1.16912270e+00
4.08816934e-01 1.70770600e-01 -2.18455836e-01 -6.52935803e-01
-5.24122715e-01 -8.07741940e-01 -3.97186369e-01 3.40213925e-01
8.64144087e-01 -1.16800272e+00 5.53630650e-01 1.39166510e+00
-3.58625412e-01 -4.51715857e-01 -8.63314152e-01 -1.06149745e+00
-4.66870656e-03 1.81816339e-01 3.85523647e-01 8.18638563e-01
7.09820271e-01 2.02331603e-01 -2.05201268e-01 -4.37674336e-02
6.59316301e-01 -4.87512946e-01 5.82342803e-01 -8.74011815e-01
2.49149442e-01 -3.85976404e-01 -2.94489831e-01 -2.90531307e-01
-3.26795012e-01 -4.87107635e-01 -8.59293997e-01 -1.53084815e+00
8.49664882e-02 1.97367743e-01 -2.33256146e-01 3.44385266e-01
6.77823246e-01 2.45789096e-01 -2.27382466e-01 1.17054820e-01
-1.86927155e-01 -9.06788632e-02 9.18564737e-01 1.10557920e-03
-2.17724726e-01 3.92412305e-01 -1.01716614e+00 9.60534811e-01
7.85404518e-02 -6.09580100e-01 -5.51487744e-01 1.64939865e-01
5.71885824e-01 3.57775450e-01 4.49405432e-01 -9.70509052e-01
9.87445056e-01 -4.32846308e-01 6.12257063e-01 -1.43385470e+00
-2.65485227e-01 -9.00021493e-01 3.61826897e-01 6.29400074e-01
4.08042461e-01 3.53796333e-01 2.26811334e-01 -9.13986787e-02
-2.01378226e-01 -6.44658387e-01 4.43001747e-01 -7.45609999e-01
-4.55086857e-01 -3.90635371e-01 -1.02313221e+00 -9.67486948e-03
7.88939476e-01 -4.61360693e-01 -6.47767961e-01 -1.95199717e-02
-1.44608676e+00 7.30553791e-02 8.33678424e-01 1.34732248e-02
6.98022604e-01 -1.44444084e+00 -1.11748077e-01 6.18802965e-01
-5.42432606e-01 -4.98194307e-01 2.20674574e-01 1.06156552e+00
-1.31528914e+00 1.26915324e+00 -5.21896303e-01 -2.05424115e-01
-8.96585643e-01 7.19095051e-01 3.15837234e-01 1.75098896e-01
-3.93049836e-01 7.00151443e-01 1.12239346e-02 -6.58594191e-01
1.24011397e-01 2.97559589e-01 -4.87763733e-01 -2.52128989e-01
9.22277749e-01 1.17918742e+00 1.32526740e-01 -3.91715407e-01
-5.58191121e-01 4.39446121e-01 2.08604142e-01 -6.12734079e-01
1.38662541e+00 -2.20989168e-01 -7.44191706e-01 5.00318348e-01
7.50003338e-01 -7.33495504e-02 -3.28966863e-02 1.27011871e+00
4.80256975e-03 -5.47239244e-01 -3.11336160e-01 -7.82959402e-01
-3.42039734e-01 5.93214929e-01 6.17733955e-01 4.59080935e-01
9.69997942e-01 -6.47576571e-01 7.70057857e-01 7.64313757e-01
7.83133745e-01 -1.22997761e+00 -6.36951804e-01 3.69483382e-01
9.39836800e-01 -9.30032432e-01 1.30400375e-01 -3.14221263e-01
-5.55534840e-01 1.08689797e+00 5.38173854e-01 -1.62142932e-01
1.17091513e+00 6.96656525e-01 -1.75611794e-01 -4.53250527e-01
-1.04549897e+00 9.83817875e-02 2.10727617e-01 1.32608259e+00
8.26061428e-01 1.41275764e-01 -1.40426433e+00 8.46834600e-01
-6.34404242e-01 4.76996809e-01 9.47268784e-01 1.55018985e+00
-5.12773991e-01 -1.41692734e+00 -4.41162825e-01 2.31142953e-01
-2.82854587e-01 -3.49930972e-01 -8.62966061e-01 1.09929097e+00
1.97083559e-02 1.37330377e+00 -4.48758155e-01 -4.63731140e-01
8.27405751e-01 3.19600284e-01 4.41484600e-01 -1.25142381e-01
-8.94688129e-01 6.48750722e-01 3.84649158e-01 -1.05672014e+00
-7.94483483e-01 -1.27049911e+00 -1.08188879e+00 -1.16740942e+00
-3.07217389e-01 4.65777248e-01 2.31674016e-01 3.19533348e-01
-2.09303293e-02 1.56313553e-03 6.67447627e-01 -1.00111675e+00
-1.97397545e-01 -6.07395887e-01 -1.20775104e+00 -5.56946039e-01
3.94425273e-01 -5.85869551e-01 -1.15223050e+00 -6.24734610e-02] | [-1.5392177104949951, 15.869220733642578] |
2bfa51b4-4a96-40ab-a53d-0d7c10af1472 | a-dive-into-sam-prior-in-image-restoration | 2305.13620 | null | https://arxiv.org/abs/2305.13620v1 | https://arxiv.org/pdf/2305.13620v1.pdf | A Dive into SAM Prior in Image Restoration | The goal of image restoration (IR), a fundamental issue in computer vision, is to restore a high-quality (HQ) image from its degraded low-quality (LQ) observation. Multiple HQ solutions may correspond to an LQ input in this poorly posed problem, creating an ambiguous solution space. This motivates the investigation and incorporation of prior knowledge in order to effectively constrain the solution space and enhance the quality of the restored images. In spite of the pervasive use of hand-crafted and learned priors in IR, limited attention has been paid to the incorporation of knowledge from large-scale foundation models. In this paper, we for the first time leverage the prior knowledge of the state-of-the-art segment anything model (SAM) to boost the performance of existing IR networks in an parameter-efficient tuning manner. In particular, the choice of SAM is based on its robustness to image degradations, such that HQ semantic masks can be extracted from it. In order to leverage semantic priors and enhance restoration quality, we propose a lightweight SAM prior tuning (SPT) unit. This plug-and-play component allows us to effectively integrate semantic priors into existing IR networks, resulting in significant improvements in restoration quality. As the only trainable module in our method, the SPT unit has the potential to improve both efficiency and scalability. We demonstrate the effectiveness of the proposed method in enhancing a variety of methods across multiple tasks, such as image super-resolution and color image denoising. | ['Zhiwei Xiong', 'Zhihe Lu', 'Jiawang Bai', 'Zeyu Xiao'] | 2023-05-23 | null | null | null | null | ['color-image-denoising', 'image-super-resolution', 'image-restoration'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 6.10321403e-01 -1.42013952e-01 1.26851067e-01 -1.47200912e-01
-1.02686584e+00 -4.80976939e-01 3.73240888e-01 -3.02607059e-01
-2.80407637e-01 4.17854041e-01 2.83429742e-01 -1.28735334e-01
-1.95059270e-01 -6.60199940e-01 -7.82660484e-01 -8.76563430e-01
4.74521697e-01 -2.07521573e-01 1.12446815e-01 -3.53203118e-01
2.30569899e-01 5.42883337e-01 -1.62513292e+00 1.77709162e-01
1.25686955e+00 9.68090117e-01 6.85403228e-01 4.18869466e-01
1.31046131e-01 4.63251233e-01 -4.41853613e-01 -1.86394632e-01
4.61646914e-01 -3.88930798e-01 -6.08016431e-01 2.52188444e-01
5.37403822e-01 -3.46358567e-01 -1.30304694e-01 1.12391186e+00
6.12626493e-01 2.98373908e-01 3.40528220e-01 -8.52421761e-01
-7.37853289e-01 2.55371541e-01 -6.87061608e-01 7.88664520e-02
1.16560452e-01 3.27995151e-01 9.89694059e-01 -8.04330826e-01
4.62874800e-01 1.18664670e+00 5.10394156e-01 3.68770063e-01
-1.49368405e+00 -5.01857400e-01 2.07542434e-01 1.95026159e-01
-1.11104822e+00 -7.32079268e-01 9.53856826e-01 -1.94408953e-01
7.05396771e-01 1.72329471e-01 5.41815639e-01 9.22124624e-01
6.11383431e-02 8.24396014e-01 1.21414256e+00 -4.78874803e-01
2.56044835e-01 -1.07536493e-02 -1.06315561e-01 3.60423833e-01
3.75476070e-02 6.31983578e-02 -4.69716668e-01 1.57776415e-01
9.55564141e-01 -1.45073891e-01 -5.36113799e-01 -1.97308064e-01
-8.86184752e-01 5.92105806e-01 7.31255651e-01 1.81017637e-01
-3.54255289e-01 1.82568386e-01 -7.01229870e-02 9.04080868e-02
5.83922803e-01 5.00971973e-01 -3.50289255e-01 2.37582967e-01
-9.11405921e-01 -2.73862258e-02 3.57615650e-01 3.99623007e-01
8.20510447e-01 9.22213495e-02 -2.97684252e-01 1.23107767e+00
1.81395814e-01 3.82972926e-01 1.25803590e-01 -1.22192216e+00
3.60726595e-01 4.79180574e-01 2.64437646e-01 -8.51417720e-01
-2.61303574e-01 -7.74869144e-01 -8.12594295e-01 5.31049967e-01
1.27171263e-01 2.20862582e-01 -1.16362846e+00 1.75837648e+00
4.55236435e-01 2.70111561e-01 6.41591474e-02 1.17360771e+00
6.32402599e-01 5.93417466e-01 -6.82622716e-02 -1.30745664e-01
1.23162150e+00 -8.54356527e-01 -5.53024530e-01 -5.69798470e-01
2.17010640e-02 -9.58253384e-01 1.32110190e+00 5.43232679e-01
-1.00155079e+00 -7.71750033e-01 -1.10200322e+00 -2.89203942e-01
3.50081958e-02 1.66206434e-01 4.05806184e-01 4.70344365e-01
-1.10956216e+00 6.35736525e-01 -7.37884700e-01 -2.75915295e-01
5.58451295e-01 1.84544787e-01 -3.42651606e-01 -6.35002971e-01
-9.72721338e-01 8.61544490e-01 1.83242872e-01 4.80019599e-01
-7.88698256e-01 -6.55228972e-01 -6.75526679e-01 1.03846565e-02
5.37120759e-01 -8.96881998e-01 8.50374460e-01 -1.14932227e+00
-1.57957768e+00 6.17081940e-01 -1.08099766e-01 -1.67008460e-01
3.71535629e-01 -3.72184455e-01 -2.99556702e-01 3.50747555e-01
-2.32790783e-02 6.87720120e-01 1.24188471e+00 -1.58958364e+00
-5.40300429e-01 -2.78943360e-01 2.26953179e-01 3.77128005e-01
-3.17378491e-01 -7.92151876e-03 -8.22519779e-01 -8.38266253e-01
1.77014068e-01 -7.88764656e-01 -8.10958445e-02 8.64024237e-02
-1.98055953e-01 1.72755644e-01 5.80001652e-01 -9.96783912e-01
8.28627765e-01 -2.26553059e+00 3.09941113e-01 3.00990641e-02
2.69380659e-02 3.56632143e-01 -4.45072740e-01 1.98914245e-01
1.04100183e-01 3.18120308e-02 -4.71688479e-01 -4.56533134e-01
-9.86575857e-02 4.29725379e-01 -2.84049898e-01 5.60181677e-01
3.98575544e-01 7.37114251e-01 -6.73502564e-01 -3.15015823e-01
4.30774152e-01 8.95221651e-01 -5.69504261e-01 3.63802642e-01
-3.21097165e-01 8.17332089e-01 -3.43708277e-01 5.21163046e-01
8.74070585e-01 -2.14971691e-01 -5.54428473e-02 -6.19261146e-01
-1.29059672e-01 1.08375549e-01 -1.36280870e+00 1.78222907e+00
-6.60385668e-01 3.73016626e-01 3.98889810e-01 -9.09020662e-01
8.28079164e-01 2.41094485e-01 4.72347140e-01 -1.01301455e+00
-8.92196670e-02 1.21814027e-01 -1.72541991e-01 -4.33232337e-01
3.87558073e-01 -1.20451987e-01 4.13600594e-01 2.04325855e-01
-1.12717405e-01 -8.00006166e-02 1.79139182e-01 3.25302593e-02
9.06984389e-01 4.22977567e-01 7.62523785e-02 -4.57922090e-03
5.70468009e-01 -1.96565956e-01 7.19023108e-01 6.83242023e-01
1.65159442e-02 1.08038580e+00 1.45453393e-01 -6.81815520e-02
-8.30410779e-01 -1.06225848e+00 -2.86114454e-01 1.04888713e+00
3.11510772e-01 5.40671917e-03 -6.45672560e-01 -3.12963694e-01
-1.93712771e-01 5.69366395e-01 -3.50523412e-01 -2.08371207e-01
-7.14114010e-01 -9.02154207e-01 5.87082244e-02 3.71132880e-01
6.81799829e-01 -1.17718768e+00 -5.51263869e-01 2.89566904e-01
-4.83150125e-01 -1.19928956e+00 -4.08273816e-01 1.14519276e-01
-8.63487422e-01 -9.65311289e-01 -6.19522452e-01 -5.09328604e-01
5.70335031e-01 5.79063237e-01 1.01859868e+00 1.11833923e-01
-3.05608541e-01 3.98977876e-01 -4.55500782e-01 8.65763277e-02
-3.00482959e-01 -2.53085941e-01 -2.79132634e-01 3.21836591e-01
-2.74076343e-01 -6.67739093e-01 -1.02638900e+00 3.56851608e-01
-1.38440669e+00 2.59937376e-01 7.83274591e-01 8.63564372e-01
6.34578466e-01 2.84659714e-01 4.46030676e-01 -5.06452382e-01
5.11986494e-01 -5.42002507e-02 -6.71359181e-01 3.40082288e-01
-6.22184694e-01 1.42119870e-01 6.89405918e-01 -4.03796375e-01
-1.40538716e+00 2.95329150e-02 -2.09912539e-01 -3.70052785e-01
-7.47681931e-02 5.68021655e-01 -4.38123018e-01 -4.02380794e-01
5.14172018e-01 1.51475325e-01 8.71499926e-02 -7.87203848e-01
6.90997422e-01 4.23595101e-01 8.65021467e-01 -6.39175892e-01
7.69845128e-01 5.27030349e-01 -4.78038751e-02 -6.18006885e-01
-1.00530994e+00 -5.45388341e-01 -4.39096093e-01 -1.98913783e-01
7.62115896e-01 -1.15061712e+00 -3.52586716e-01 4.64378625e-01
-8.52169573e-01 -3.58664274e-01 -1.91691428e-01 1.37259826e-01
-4.73168731e-01 4.75376904e-01 -5.81724942e-01 -6.22043371e-01
-3.95083934e-01 -1.34038818e+00 1.07763815e+00 3.62545907e-01
3.07150185e-01 -7.69761026e-01 -1.68158904e-01 7.56772161e-01
5.18977046e-01 4.67077382e-02 7.73535967e-01 8.90361443e-02
-7.76113868e-01 -3.52407619e-02 -5.19804895e-01 7.74017155e-01
1.20423637e-01 -2.40788698e-01 -1.17776167e+00 -4.94803041e-01
1.06416196e-01 -1.70093164e-01 1.17249095e+00 5.94458520e-01
9.61427093e-01 -1.46682635e-01 4.81040105e-02 8.73702109e-01
1.70966184e+00 -1.57163382e-01 8.69334757e-01 5.78416169e-01
7.60930538e-01 5.55974603e-01 6.02268159e-01 2.45989546e-01
2.24250153e-01 8.81428003e-01 6.29376590e-01 -5.64043760e-01
-7.20056117e-01 2.75613144e-02 4.76214141e-01 3.30761045e-01
-7.84788206e-02 -2.24508256e-01 -6.37227535e-01 4.47399855e-01
-1.87413573e+00 -6.36273444e-01 1.07056685e-01 2.23238277e+00
1.09282029e+00 -8.56656134e-02 -2.73922145e-01 4.93591763e-02
5.73795974e-01 2.40559146e-01 -6.00246906e-01 9.87482741e-02
-2.56931156e-01 2.20939681e-01 3.28814805e-01 5.75021923e-01
-1.05040669e+00 9.34881926e-01 5.27869081e+00 9.45159912e-01
-1.10062599e+00 1.63741428e-02 6.31825030e-01 4.69029462e-03
-2.62525082e-01 1.01584636e-01 -5.81018567e-01 2.59682477e-01
5.16198993e-01 3.45015347e-01 8.12861621e-01 4.13146257e-01
5.92906773e-01 -2.80423880e-01 -8.76632750e-01 8.61916006e-01
6.65259883e-02 -1.02917945e+00 -1.50509000e-01 -3.80716361e-02
7.71044970e-01 -5.50487339e-02 2.76640713e-01 -2.33571846e-02
1.58633709e-01 -9.91583169e-01 7.09829390e-01 4.87876058e-01
7.45621622e-01 -6.57857656e-01 6.16547465e-01 1.20045513e-01
-1.03780544e+00 -2.72265166e-01 -4.07059580e-01 1.87327459e-01
3.80796880e-01 8.94109130e-01 -4.58830267e-01 6.54990077e-01
9.20598030e-01 7.25988805e-01 -5.49298346e-01 1.16669595e+00
-4.80860412e-01 5.58737457e-01 -2.74964541e-01 8.71915996e-01
-3.76832969e-02 -4.77976501e-01 5.25414586e-01 9.04350162e-01
1.81930318e-01 1.80763036e-01 2.67583013e-01 9.98725176e-01
-2.01579314e-04 -1.27339780e-01 -3.46154608e-02 2.31104165e-01
1.46341100e-01 1.43839812e+00 -6.38558269e-01 2.42965333e-02
-4.40981239e-01 1.05360031e+00 1.82695836e-01 7.23522544e-01
-6.45359397e-01 8.18530694e-02 7.33669043e-01 7.90898651e-02
5.55015624e-01 -1.82409897e-01 -4.68769401e-01 -1.06641543e+00
1.67902291e-01 -1.12323546e+00 2.81595379e-01 -9.30963099e-01
-1.40091300e+00 5.85957825e-01 -2.55944610e-01 -1.18847477e+00
1.30995423e-01 -3.94290030e-01 -3.59615386e-01 1.14864099e+00
-1.97673190e+00 -1.40039778e+00 -5.20447493e-01 6.75964773e-01
4.37952548e-01 2.04867795e-01 4.20589089e-01 3.53757203e-01
-7.18332052e-01 2.13987559e-01 1.08306170e-01 -2.58015186e-01
9.57392335e-01 -1.06984603e+00 1.37307391e-01 1.30978775e+00
5.26966229e-02 6.15715384e-01 9.09428537e-01 -5.71242929e-01
-1.40218389e+00 -1.17333615e+00 1.58070177e-01 -1.59063131e-01
3.62813979e-01 -1.52066261e-01 -1.07331705e+00 2.59155989e-01
3.31933517e-03 2.07144722e-01 3.24200332e-01 -5.34596890e-02
-4.70947534e-01 -4.57139760e-01 -1.04319918e+00 6.57056332e-01
8.05232286e-01 -5.20188153e-01 -3.70021760e-01 7.86214992e-02
7.35062599e-01 -2.32771814e-01 -8.11693370e-01 5.40555358e-01
4.08154339e-01 -1.13468313e+00 1.54257309e+00 -3.48479487e-02
4.89894867e-01 -7.17943311e-01 -1.63976282e-01 -1.26514709e+00
-4.30446863e-01 -3.03264737e-01 -4.86408658e-02 1.32933545e+00
1.04527496e-01 -6.09199524e-01 3.68865520e-01 7.49898493e-01
-3.49394947e-01 -4.53977853e-01 -8.11535656e-01 -5.11715174e-01
-1.84149191e-01 -3.65442574e-01 3.01505297e-01 6.56222522e-01
-7.22392201e-01 1.06824942e-01 -5.41630387e-01 5.50982654e-01
8.12237799e-01 2.59804875e-01 6.77160203e-01 -1.10743928e+00
-5.09782493e-01 -3.26900572e-01 3.82424854e-02 -1.12766874e+00
-6.72408491e-02 -6.56210303e-01 3.78073484e-01 -1.83960330e+00
2.01870844e-01 -4.99609709e-01 -4.97878253e-01 6.07764244e-01
-5.09792864e-01 5.93807042e-01 3.83670211e-01 3.22473168e-01
-4.90322262e-01 7.49276102e-01 1.41732264e+00 -1.16418637e-01
-3.16287130e-01 -9.31781009e-02 -9.42718387e-01 4.76059735e-01
7.49831557e-01 -4.35646802e-01 -4.34371382e-01 -6.20732248e-01
2.87085891e-01 -4.19563986e-02 6.54262304e-01 -9.11624253e-01
1.36513427e-01 -1.19722232e-01 4.52166438e-01 -2.58815795e-01
4.67716426e-01 -8.68217945e-01 2.32460171e-01 6.99759787e-03
-5.72956316e-02 -4.77305532e-01 1.92177624e-01 6.62479818e-01
-3.35570365e-01 -7.25510120e-02 9.59181547e-01 -1.44323096e-01
-8.39362025e-01 1.41841069e-01 1.42384499e-01 -1.74847305e-01
5.09582937e-01 -3.53443325e-01 -3.73400688e-01 -2.91446090e-01
-4.10828680e-01 6.63217977e-02 6.61965966e-01 4.08609390e-01
7.27416873e-01 -8.93667340e-01 -7.12841034e-01 1.62880719e-01
5.09890765e-02 1.03443123e-01 5.59420824e-01 9.60598409e-01
-2.47799411e-01 -4.41458300e-02 -2.40910530e-01 -6.31224036e-01
-1.09415340e+00 4.21829283e-01 2.90396810e-01 -2.19255760e-01
-1.04995024e+00 6.76358521e-01 3.50546837e-01 -1.89020932e-01
2.34803617e-01 -1.00142360e-01 -2.65648246e-01 -2.11285472e-01
5.74510157e-01 2.47448951e-01 1.99095055e-01 -4.38171476e-01
-8.34675282e-02 7.33348846e-01 1.42808584e-02 -9.15708914e-02
1.54192972e+00 -5.09346724e-01 -2.58041114e-01 4.05859165e-02
9.02531981e-01 -7.48323873e-02 -1.76242077e+00 -2.66252995e-01
-2.57785618e-01 -6.19611919e-01 5.96485734e-01 -1.11944735e+00
-1.31818855e+00 6.50554776e-01 8.11404526e-01 -1.33767590e-01
1.65482116e+00 -1.92449078e-01 8.32861423e-01 -5.78828752e-02
3.22130889e-01 -9.99147475e-01 2.63561338e-01 8.96519795e-02
1.00853443e+00 -1.37265885e+00 1.36561334e-01 -5.38196683e-01
-5.62269568e-01 1.00610828e+00 3.66604447e-01 3.66533734e-02
1.84556335e-01 1.50495902e-01 2.82199144e-01 3.06989867e-02
-4.37442780e-01 -4.13002431e-01 6.10542655e-01 6.09983683e-01
3.24792057e-01 -2.33024016e-01 -6.00675568e-02 4.01793778e-01
1.85759917e-01 1.19542629e-02 3.74697089e-01 6.82509124e-01
-5.11610270e-01 -1.14001632e+00 -6.15836143e-01 1.52657926e-01
-3.50379527e-01 -2.81005740e-01 1.03615560e-01 3.73150319e-01
1.50718406e-01 1.32536149e+00 -3.32506925e-01 -4.19965945e-02
2.71259606e-01 -3.76037985e-01 5.06908417e-01 -3.52571815e-01
-5.43231726e-01 4.65308756e-01 -1.89598389e-02 -8.34762990e-01
-5.43036222e-01 -3.43963295e-01 -9.92958605e-01 1.24885179e-01
-2.06845596e-01 -3.68198097e-01 7.08043873e-01 9.29620326e-01
3.83886576e-01 7.01036155e-01 5.08084714e-01 -9.71804023e-01
-3.82979959e-01 -8.33829820e-01 -5.67018151e-01 5.75190902e-01
5.21429539e-01 -6.11482978e-01 -4.17862564e-01 1.53382197e-01] | [10.912443161010742, -2.4700136184692383] |
fe5c210b-b7fe-4776-829a-4060d6a1a2da | evcenternet-uncertainty-estimation-for-object | 2303.03037 | null | https://arxiv.org/abs/2303.03037v1 | https://arxiv.org/pdf/2303.03037v1.pdf | EvCenterNet: Uncertainty Estimation for Object Detection using Evidential Learning | Uncertainty estimation is crucial in safety-critical settings such as automated driving as it provides valuable information for several downstream tasks including high-level decision-making and path planning. In this work, we propose EvCenterNet, a novel uncertainty-aware 2D object detection framework utilizing evidential learning to directly estimate both classification and regression uncertainties. To employ evidential learning for object detection, we devise a combination of evidential and focal loss functions for the sparse heatmap inputs. We introduce class-balanced weighting for regression and heatmap prediction to tackle the class imbalance encountered by evidential learning. Moreover, we propose a learning scheme to actively utilize the predicted heatmap uncertainties to improve the detection performance by focusing on the most uncertain points. We train our model on the KITTI dataset and evaluate it on challenging out-of-distribution datasets including BDD100K and nuImages. Our experiments demonstrate that our approach improves the precision and minimizes the execution time loss in relation to the base model. | ['Abhinav Valada', 'Chih-Hong Cheng', 'Wolfram Burgard', 'Paulo L. J. Drews-Jr', 'Kshitij Sirohi', 'Monish R. Nallapareddy'] | 2023-03-06 | null | null | null | null | ['2d-object-detection'] | ['computer-vision'] | [-4.82665747e-02 2.32016966e-01 -2.74307638e-01 -6.67193651e-01
-1.25873423e+00 -2.63060451e-01 6.60396874e-01 4.02769536e-01
-2.29787380e-01 6.72830582e-01 -1.21504873e-01 -4.67124462e-01
-5.82212031e-01 -5.74339271e-01 -8.66810322e-01 -6.47648990e-01
-1.96023509e-02 6.02190793e-01 3.36174846e-01 1.77457631e-01
4.33497220e-01 7.72558451e-01 -1.62769330e+00 6.34469539e-02
9.91122246e-01 1.64632750e+00 7.22325221e-02 3.51222575e-01
9.77674127e-02 9.35660541e-01 -3.27609479e-01 -2.09324360e-01
3.93805832e-01 2.90080547e-01 -1.55197352e-01 -1.65458679e-01
3.78238261e-01 -3.21759701e-01 -7.60818645e-02 7.72852242e-01
5.83983541e-01 1.25340253e-01 1.08064377e+00 -1.67312670e+00
3.13255608e-01 3.72081637e-01 -8.15874815e-01 2.81605333e-01
-2.99242765e-01 4.34946477e-01 6.04854882e-01 -1.08916974e+00
8.37956667e-02 9.09659982e-01 8.88830841e-01 1.73676312e-01
-8.33031833e-01 -9.17737842e-01 2.14529559e-01 4.86792207e-01
-1.44006181e+00 -5.38735330e-01 8.97257507e-01 -6.79865301e-01
6.51143134e-01 1.64332286e-01 2.49543235e-01 7.52352774e-01
4.55225855e-01 9.66004491e-01 1.10068095e+00 -2.68786550e-01
5.64373434e-01 1.34636074e-01 1.37116715e-01 7.04292238e-01
4.57090944e-01 5.22292733e-01 -7.14798450e-01 -7.17269629e-02
1.62577853e-01 -1.65042967e-01 1.50468171e-01 -5.44747531e-01
-7.43744195e-01 6.91413701e-01 4.30899799e-01 -4.41107780e-01
-4.55629557e-01 4.50877964e-01 1.39299661e-01 -3.42924088e-01
7.84566164e-01 3.87914300e-01 -3.61493826e-01 -1.48867309e-01
-9.27366853e-01 2.77426004e-01 3.52534592e-01 9.72128212e-01
9.19876397e-01 -1.55203044e-01 -6.33691192e-01 6.29169941e-01
6.26689970e-01 6.52910948e-01 -2.21117228e-01 -9.49313104e-01
7.85800755e-01 5.70439219e-01 2.30874151e-01 -7.13340163e-01
-6.61489546e-01 -7.09637344e-01 -5.86506188e-01 7.79687822e-01
1.41145885e-01 -1.58928126e-01 -1.19400167e+00 1.32908249e+00
5.57491541e-01 5.49077272e-01 -4.98464666e-02 9.09224987e-01
4.94399250e-01 5.37505209e-01 1.83911204e-01 1.78029805e-01
1.03193593e+00 -6.81211054e-01 -4.98073220e-01 -4.42186296e-01
5.47764301e-01 -4.11747754e-01 6.30489826e-01 4.38307494e-01
-7.89902806e-01 -4.45876151e-01 -1.40819252e+00 5.68507090e-02
-2.75309622e-01 3.95409971e-01 5.45987010e-01 5.29928148e-01
-3.69960874e-01 5.11858165e-01 -1.00813448e+00 3.04298759e-01
8.58732581e-01 4.12132367e-02 5.89558519e-02 -2.29645327e-01
-9.16377187e-01 1.13411903e+00 4.99692500e-01 4.65085745e-01
-1.27064860e+00 -1.17421579e+00 -1.14473343e+00 -8.63533989e-02
6.40395522e-01 -4.28059459e-01 1.06507146e+00 3.07904303e-01
-1.18172812e+00 2.54542798e-01 1.81343853e-01 -8.61952364e-01
7.68119574e-01 -4.16402668e-01 -1.66638494e-01 -7.74266720e-02
-1.03919534e-02 6.04134619e-01 1.07395184e+00 -1.33246100e+00
-8.32635224e-01 -3.27652663e-01 -1.85094848e-01 1.28157347e-01
1.46575958e-01 -6.07948780e-01 -2.70800203e-01 -2.53270864e-01
2.54348427e-01 -8.51823747e-01 -4.81125772e-01 1.64211914e-01
-6.23636365e-01 -1.23205394e-01 8.54460001e-01 -5.06765842e-01
1.03794479e+00 -2.23305655e+00 -2.97714621e-01 4.88073498e-01
2.64253318e-01 2.72139758e-02 3.15793574e-01 -1.29910791e-03
2.93701142e-01 -3.69377971e-01 -6.06237888e-01 -6.15920544e-01
3.27103198e-01 1.60909399e-01 -3.18216652e-01 6.67068720e-01
7.27515697e-01 8.37842464e-01 -7.86045432e-01 -5.30316651e-01
6.71583831e-01 3.21052849e-01 -2.68217146e-01 2.26205081e-01
-3.54724616e-01 1.46993577e-01 -5.77143252e-01 5.73234797e-01
9.57387447e-01 9.14245918e-02 -5.04654109e-01 -2.16854930e-01
-1.40074044e-01 8.71169344e-02 -1.24075639e+00 1.71546102e+00
-7.54221082e-01 6.46732032e-01 -1.47521242e-01 -7.30701745e-01
9.66759026e-01 -2.61306226e-01 2.96875417e-01 -6.57932997e-01
1.13160066e-01 1.29796281e-01 -3.88567984e-01 -3.78253281e-01
6.71645403e-01 -4.63289283e-02 -4.25660998e-01 1.45112425e-01
-2.51317561e-01 -4.73351687e-01 -2.49581456e-01 2.65321821e-01
1.00142109e+00 2.47035280e-01 -2.09999606e-02 -2.32537448e-01
4.78833318e-01 1.62025437e-01 6.22460246e-01 6.75989330e-01
-2.27424800e-01 5.48219264e-01 5.95150590e-01 -2.06093878e-01
-6.39858902e-01 -1.10840964e+00 -3.47147465e-01 5.77493966e-01
4.37530369e-01 6.15472198e-02 -3.22266430e-01 -8.81774545e-01
5.41223943e-01 1.36328125e+00 -6.11974299e-01 -5.39004445e-01
-1.77119002e-01 -8.60215306e-01 2.84991562e-01 7.49697506e-01
3.26223433e-01 -4.89196151e-01 -7.72968590e-01 -1.21918069e-02
2.34401330e-01 -1.06487143e+00 6.97805919e-03 5.32460213e-01
-4.82892185e-01 -1.12324643e+00 -2.48834476e-01 -3.05756170e-04
6.91520333e-01 -3.10148597e-02 1.04100704e+00 -3.36254925e-01
-5.76249659e-01 3.52398038e-01 -5.28610609e-02 -9.86406446e-01
-2.96562016e-01 5.92003055e-02 -8.98681507e-02 -4.81251888e-02
1.68168351e-01 -1.67180941e-01 -5.83307505e-01 3.90894353e-01
-5.63943326e-01 -1.05748132e-01 7.25796580e-01 7.63470113e-01
8.57811689e-01 2.18431026e-01 7.21865892e-01 -6.93341196e-01
3.35772395e-01 -7.65451908e-01 -1.17829859e+00 1.38408184e-01
-7.87407458e-01 4.42169279e-01 5.85231557e-03 4.10613231e-02
-1.44018924e+00 3.40860546e-01 -1.37292385e-01 -5.75181127e-01
4.50120494e-02 5.20692408e-01 -6.73725083e-02 -1.95540920e-01
7.97948539e-01 -2.77020991e-01 -2.56455868e-01 -3.25311422e-01
3.62138540e-01 6.83849752e-01 6.33226752e-01 -7.50438631e-01
9.29992914e-01 4.38863844e-01 5.45284152e-01 -5.12045562e-01
-1.07368696e+00 -6.51544750e-01 -4.08983946e-01 -4.68565583e-01
5.48983157e-01 -1.15979064e+00 -6.42293036e-01 3.11900675e-01
-8.73315156e-01 -4.07449037e-01 -4.38850641e-01 7.74427772e-01
-5.92651606e-01 4.09801565e-02 2.56717782e-02 -1.22552204e+00
-2.25177780e-01 -1.11819911e+00 1.40977621e+00 -4.76863421e-02
1.85438558e-01 -6.20616734e-01 5.83358966e-02 3.26244205e-01
3.34355742e-01 4.71270621e-01 6.10464215e-01 -3.58991921e-01
-8.86487067e-01 -3.04765850e-01 -4.52321291e-01 3.50909770e-01
-3.58468503e-01 -1.86150864e-01 -1.28857207e+00 -4.24686819e-02
-2.88574606e-01 -3.44691455e-01 1.23219180e+00 4.02574718e-01
1.39215529e+00 1.86524704e-01 -4.84536827e-01 4.83784765e-01
1.37833905e+00 -6.03079125e-02 4.07599092e-01 1.15794644e-01
5.44903278e-01 7.49800265e-01 1.20406747e+00 6.10453188e-01
4.07125056e-01 5.73374867e-01 8.70356858e-01 4.57427874e-02
-1.33503959e-01 -2.34451294e-01 6.72005937e-02 1.22404538e-01
3.97494793e-01 -1.25957295e-01 -1.02654040e+00 6.58231974e-01
-2.01304126e+00 -6.81463659e-01 -1.04877166e-01 2.24095798e+00
5.18270195e-01 6.04125023e-01 -3.15836519e-01 1.53798103e-01
5.52199185e-01 6.81353966e-04 -1.03140008e+00 -5.50777800e-02
2.99740016e-01 -1.31842494e-01 9.14860427e-01 4.44581628e-01
-1.18658650e+00 4.11872953e-01 5.53309584e+00 1.14615202e+00
-8.91470253e-01 1.65430814e-01 8.39104056e-01 -3.24433833e-01
-3.41068894e-01 -2.80770123e-01 -1.17697740e+00 4.29684758e-01
9.91267025e-01 7.22248182e-02 -9.94667560e-02 1.07092488e+00
7.62653649e-02 -4.60131586e-01 -1.08152819e+00 8.68160665e-01
-2.19766408e-01 -1.45278871e+00 -6.06706381e-01 -7.30029568e-02
7.14331448e-01 3.26948971e-01 2.28193879e-01 4.06467468e-01
4.41934437e-01 -8.83670866e-01 1.01215661e+00 6.07302666e-01
7.04607010e-01 -1.05289257e+00 7.68347085e-01 5.90227127e-01
-9.91485655e-01 -1.95987388e-01 -3.27243239e-01 1.76318616e-01
4.18808788e-01 1.26328945e+00 -1.40817571e+00 7.20065653e-01
4.76244599e-01 5.12975872e-01 -5.46756923e-01 1.32534206e+00
-4.52411473e-01 5.95964134e-01 -6.79059446e-01 1.36569872e-01
2.11121961e-01 1.85630515e-01 5.73223829e-01 1.01128900e+00
4.46591765e-01 -1.12685770e-01 -6.78958446e-02 1.02546954e+00
1.23133332e-01 -4.46345747e-01 -4.88739014e-01 4.44413245e-01
6.03352726e-01 1.26118135e+00 -5.52346110e-01 -7.22838715e-02
-8.25136155e-03 2.70651817e-01 2.98562348e-01 1.34596944e-01
-1.07429576e+00 -2.74464548e-01 6.69532239e-01 1.45606086e-01
2.82242656e-01 -2.49892786e-01 -6.31044090e-01 -5.51571786e-01
1.65028349e-01 2.07961500e-02 3.55325341e-01 -7.79868722e-01
-1.22080171e+00 4.82208252e-01 6.02474511e-01 -1.31226695e+00
-3.93985391e-01 -6.63243830e-01 -5.52164078e-01 8.28409076e-01
-2.11044812e+00 -1.07883060e+00 -6.25258803e-01 1.11707792e-01
3.54585886e-01 -5.08128814e-02 4.37621623e-02 3.13083768e-01
-6.47769332e-01 4.33613628e-01 -2.34689983e-03 -4.61074680e-01
6.14380181e-01 -1.15800273e+00 4.18071866e-01 8.63000572e-01
-2.00060621e-01 -8.53066519e-02 8.41980278e-01 -6.67242348e-01
-1.20128429e+00 -1.58567405e+00 2.71761805e-01 -5.61204493e-01
5.10638356e-01 -3.77009451e-01 -6.14789724e-01 3.84615362e-01
-3.57777804e-01 3.89463603e-01 2.96438456e-01 -7.49626681e-02
-8.27624202e-02 -4.29859519e-01 -1.42143691e+00 1.21141784e-01
8.57535779e-01 -4.05793250e-01 -3.16962183e-01 2.44932085e-01
8.01529467e-01 -6.74152553e-01 -7.04870045e-01 9.01940763e-01
5.21372736e-01 -7.99114525e-01 1.01172578e+00 -2.79834270e-01
3.36721659e-01 -6.10996187e-01 -5.67630902e-02 -1.42779052e+00
-7.69739375e-02 -1.86232030e-01 -5.29271841e-01 1.06733084e+00
6.99256837e-01 -6.90250754e-01 9.78973389e-01 7.15130866e-01
-7.00949490e-01 -8.73832762e-01 -1.27524531e+00 -7.29948699e-01
-2.78057009e-01 -1.19289327e+00 7.52654850e-01 3.33377540e-01
-4.12387043e-01 -4.35437337e-02 -2.26944938e-01 6.98536873e-01
1.13501501e+00 -1.55072045e-02 7.72371233e-01 -1.16278982e+00
-1.63355157e-01 -7.55022913e-02 -6.10593021e-01 -8.05439293e-01
1.40615344e-01 -5.66406608e-01 6.96691394e-01 -1.41008937e+00
-1.45391926e-01 -1.15924323e+00 -5.54156959e-01 4.03313488e-01
-1.70287341e-01 1.37755603e-01 -6.71767071e-02 -1.47481039e-01
-7.39359081e-01 7.99868345e-01 8.28115761e-01 -4.68032271e-01
-1.17398631e-02 3.56017679e-01 -4.46268976e-01 5.70946872e-01
8.24534535e-01 -6.56816363e-01 -7.37624347e-01 -2.87243813e-01
4.72385645e-01 5.07136658e-02 5.53653121e-01 -1.32344317e+00
4.47875977e-01 -1.24914207e-01 4.72643912e-01 -1.44652390e+00
4.63881344e-01 -9.80388820e-01 7.50473514e-02 2.58361071e-01
-3.77164245e-01 -3.88642967e-01 6.19248748e-01 9.52892244e-01
-3.54374386e-02 -3.36232334e-01 5.81372321e-01 4.42219347e-01
-9.55214083e-01 3.72904778e-01 -9.29114893e-02 -1.74345493e-01
1.38775802e+00 2.15623111e-01 -4.23002839e-01 -6.47086650e-02
-3.48125190e-01 7.62936354e-01 1.15026928e-01 2.20886841e-01
7.26676702e-01 -1.30796516e+00 -8.21748734e-01 2.53248543e-01
6.61930263e-01 6.39834404e-01 3.71752024e-01 7.76099563e-01
-2.73283392e-01 3.23829651e-01 8.96224603e-02 -9.32823420e-01
-8.00689876e-01 4.26269054e-01 2.10673407e-01 -3.08305562e-01
-3.22578967e-01 9.17756915e-01 1.04052342e-01 -4.44233209e-01
3.33697766e-01 -5.89324236e-01 -1.72640234e-01 -2.89935488e-02
4.20300424e-01 7.87342906e-01 5.46317041e-01 -2.58264661e-01
-4.25442159e-01 2.33222350e-01 1.29551873e-01 5.55709451e-02
1.22497678e+00 -2.25568861e-01 4.03509289e-01 3.55946869e-01
1.02527070e+00 -2.43683010e-01 -1.98963737e+00 -1.31316721e-01
1.79037690e-01 -4.07190889e-01 6.48500979e-01 -9.54156518e-01
-9.65097904e-01 9.55800056e-01 8.23457062e-01 -2.28336081e-01
8.38930249e-01 6.33998737e-02 5.39597750e-01 4.35787320e-01
4.20386642e-01 -1.12833452e+00 -2.08905622e-01 1.48458764e-01
8.88488173e-01 -1.55121005e+00 4.44957227e-01 -4.73746955e-01
-6.37254894e-01 7.47669041e-01 5.40405571e-01 1.93662122e-01
8.30488503e-01 4.95406419e-01 9.09699779e-03 -4.07053709e-01
-8.40797484e-01 -2.09423080e-01 6.57634556e-01 4.79491234e-01
-4.16189551e-01 6.93108812e-02 7.99572244e-02 5.11114895e-01
2.12132961e-01 7.45610669e-02 2.25943133e-01 9.79042768e-01
-5.36092520e-01 -5.55443406e-01 -4.64093655e-01 6.56813800e-01
1.77724455e-02 2.61723220e-01 2.85084456e-01 5.11316895e-01
1.39077559e-01 7.99501181e-01 1.28764793e-01 -5.01931250e-01
4.30802703e-01 -1.14521332e-01 3.04805994e-01 -4.78669763e-01
1.49709091e-01 -4.59269971e-01 3.23563635e-01 -6.74461663e-01
-1.77309945e-01 -6.64834082e-01 -1.15882623e+00 1.41699448e-01
-4.62485790e-01 1.34717241e-01 1.25752056e+00 1.01605737e+00
5.70634425e-01 7.86551595e-01 7.34095156e-01 -1.08406317e+00
-1.00591302e+00 -8.68032217e-01 -4.94801670e-01 -1.82404757e-01
5.66850662e-01 -1.16358888e+00 -5.09732127e-01 -5.09560466e-01] | [7.684136390686035, -1.2045000791549683] |
4e88607c-4773-400f-b3f3-54af4e792cd0 | deep-transfer-learning-for-single-channel | 1904.05945 | null | https://arxiv.org/abs/1904.05945v2 | https://arxiv.org/pdf/1904.05945v2.pdf | Deep Transfer Learning for Single-Channel Automatic Sleep Staging with Channel Mismatch | Many sleep studies suffer from the problem of insufficient data to fully utilize deep neural networks as different labs use different recordings set ups, leading to the need of training automated algorithms on rather small databases, whereas large annotated databases are around but cannot be directly included into these studies for data compensation due to channel mismatch. This work presents a deep transfer learning approach to overcome the channel mismatch problem and transfer knowledge from a large dataset to a small cohort to study automatic sleep staging with single-channel input. We employ the state-of-the-art SeqSleepNet and train the network in the source domain, i.e. the large dataset. Afterwards, the pretrained network is finetuned in the target domain, i.e. the small cohort, to complete knowledge transfer. We study two transfer learning scenarios with slight and heavy channel mismatch between the source and target domains. We also investigate whether, and if so, how finetuning entirely or partially the pretrained network would affect the performance of sleep staging on the target domain. Using the Montreal Archive of Sleep Studies (MASS) database consisting of 200 subjects as the source domain and the Sleep-EDF Expanded database consisting of 20 subjects as the target domain in this study, our experimental results show significant performance improvement on sleep staging achieved with the proposed deep transfer learning approach. Furthermore, these results also reveal the essential of finetuning the feature-learning parts of the pretrained network to be able to bypass the channel mismatch problem. | ['Alfred Mertins', 'Oliver Y. Chén', 'Philipp Koch', 'Huy Phan', 'Maarten De Vos'] | 2019-04-11 | null | null | null | null | ['sleep-staging'] | ['medical'] | [ 1.18274547e-01 6.31235167e-02 -1.20971039e-01 -5.18514454e-01
-5.30492485e-01 -2.71162957e-01 -7.57005624e-03 -1.97832078e-01
-8.13465655e-01 9.97264028e-01 1.21871538e-01 -2.71076504e-02
1.00348383e-01 -5.91693878e-01 -7.36879826e-01 -6.55376732e-01
2.13264391e-01 4.12482768e-01 4.06055570e-01 -1.59534574e-01
-8.18516761e-02 9.18848254e-03 -1.27836239e+00 1.99685693e-01
1.02389431e+00 1.01106346e+00 2.79618025e-01 2.34983161e-01
2.48101473e-01 2.67545074e-01 -8.35724413e-01 -2.10304663e-01
4.31138068e-01 -5.98727345e-01 -5.85983098e-01 -2.42288366e-01
3.53366971e-01 -2.19202861e-01 -3.67024034e-01 1.00598228e+00
9.13564503e-01 1.76816702e-01 2.38385871e-01 -1.16093659e+00
-4.18235093e-01 4.97177392e-01 -2.11016715e-01 7.32778370e-01
-5.58237322e-02 2.19969064e-01 6.87649012e-01 -5.10617793e-01
3.41232449e-01 4.98849630e-01 8.61300349e-01 8.33887577e-01
-1.33300805e+00 -1.11567211e+00 -3.31540376e-01 2.53142089e-01
-1.47084737e+00 -6.16389930e-01 6.34904504e-01 -2.04548955e-01
8.08354378e-01 -2.27068588e-01 8.98169041e-01 1.26838827e+00
2.11980686e-01 1.76720336e-01 1.21838915e+00 -2.99524069e-01
3.73586506e-01 5.48236787e-01 2.12220281e-01 5.84232509e-01
1.93712324e-01 4.70609218e-02 -7.21427679e-01 1.97269455e-01
5.49865782e-01 -4.95513864e-02 -1.69348270e-01 -1.42303690e-01
-1.00442350e+00 6.88546836e-01 6.44850135e-01 5.04848123e-01
-9.87681746e-02 -1.89345211e-01 6.25208497e-01 4.95179027e-01
3.42582792e-01 4.98994857e-01 -7.14693725e-01 -2.11005583e-01
-1.37162936e+00 -2.00957239e-01 8.09943676e-01 6.92123771e-01
8.14532101e-01 -4.34785150e-02 -2.57175863e-01 8.27874064e-01
9.88818333e-02 3.86611432e-01 9.52742755e-01 -6.78920507e-01
5.87911189e-01 7.94404209e-01 -2.85850793e-01 -3.09765905e-01
-8.39126468e-01 -6.63329244e-01 -8.47468376e-01 -4.70544323e-02
6.62021935e-01 -3.39335412e-01 -8.51518810e-01 1.95950389e+00
-3.90223451e-02 2.92504698e-01 1.41244873e-01 9.67965961e-01
8.61762941e-01 2.47480154e-01 -1.07928459e-02 -1.49278313e-01
1.34235227e+00 -9.31145787e-01 -4.68394190e-01 -4.01406139e-01
5.48204005e-01 -3.55013669e-01 1.41462755e+00 4.68786955e-01
-9.65731084e-01 -7.67977893e-01 -1.29102242e+00 -2.13668510e-01
-3.86554569e-01 3.08877438e-01 4.06340212e-01 7.42817879e-01
-1.07194531e+00 6.56664670e-01 -9.29861546e-01 -6.08958602e-01
5.98986983e-01 9.54070926e-01 -3.66278052e-01 1.10292234e-01
-1.39156997e+00 7.71465540e-01 2.46131852e-01 9.88798663e-02
-9.77292240e-01 -9.13030028e-01 -4.99504685e-01 2.92069972e-01
1.68570325e-01 -8.69350255e-01 8.54638219e-01 -1.16276324e+00
-1.37015939e+00 9.50498521e-01 2.46289279e-02 -5.99861801e-01
2.81485736e-01 -2.49075759e-02 -4.43633765e-01 -2.67504975e-02
1.90041348e-01 7.30570912e-01 6.70551360e-01 -7.17222154e-01
-4.32035774e-01 -5.33082902e-01 3.35908830e-02 6.78540543e-02
-6.15709603e-01 -3.60583335e-01 -4.09705490e-01 -4.71848816e-01
-4.67470706e-01 -1.10520768e+00 9.95350704e-02 -1.82081968e-01
-3.08118433e-01 5.28802909e-02 3.94400716e-01 -5.20745873e-01
1.13670933e+00 -2.40899801e+00 9.96850431e-02 1.49196267e-01
3.11007589e-01 2.69162387e-01 -1.51581079e-01 9.97110084e-02
-1.15227304e-01 -1.44106582e-01 -8.31519961e-02 -5.07066429e-01
-1.79178849e-01 3.94410998e-01 3.54923546e-01 5.40985644e-01
-1.70628384e-01 6.79054379e-01 -5.36867559e-01 -3.57697248e-01
-1.04769543e-01 3.50562274e-01 -7.47973680e-01 2.62327820e-01
2.60865510e-01 8.57574761e-01 -3.35732847e-01 2.89823949e-01
3.32698733e-01 -2.12852985e-01 -9.95016471e-03 -3.73409092e-01
-2.82469857e-02 4.64503258e-01 -6.81401312e-01 2.06781149e+00
-7.54557073e-01 6.63673043e-01 -1.82541076e-03 -1.07268739e+00
9.16236162e-01 1.98864356e-01 5.30379355e-01 -1.07389462e+00
3.64784509e-01 1.96606815e-01 5.90219736e-01 -4.07342166e-01
2.42168158e-01 -5.66961706e-01 -2.08825350e-01 3.88026774e-01
3.21052432e-01 3.52365851e-01 5.05223610e-02 -3.15985680e-02
1.22640872e+00 -3.70552897e-01 1.33007526e-01 -3.45650345e-01
3.75398457e-01 -2.47878522e-01 8.68024051e-01 6.10818744e-01
-4.05466735e-01 5.76189101e-01 4.75462139e-01 -1.90940753e-01
-7.92460144e-01 -1.03762448e+00 -1.63673088e-01 1.16748857e+00
1.32279381e-01 -5.10727763e-01 -8.16294789e-01 -7.74123192e-01
-1.55734792e-01 3.71081293e-01 -9.52363193e-01 -7.66500056e-01
-2.10166037e-01 -9.01551545e-01 9.38239276e-01 7.34829366e-01
8.36273193e-01 -1.06646621e+00 -5.99370003e-01 -8.07094797e-02
-1.18193023e-01 -9.86371577e-01 -5.01833200e-01 6.07857108e-01
-8.56122911e-01 -1.01758552e+00 -5.76236069e-01 -5.47106266e-01
4.78938490e-01 -1.71812147e-01 9.03672636e-01 1.64817587e-01
-1.57770753e-01 2.66692080e-02 -2.07835853e-01 -3.87920499e-01
-2.47921541e-01 5.92508912e-01 1.57896310e-01 6.36879280e-02
4.84443843e-01 -9.01883006e-01 -8.60330045e-01 4.44711536e-01
-5.64945996e-01 -1.99024037e-01 8.48623991e-01 9.48276281e-01
4.21421677e-01 2.80004530e-03 1.00323284e+00 -9.06022787e-01
5.17151892e-01 -6.31088793e-01 -3.17239970e-01 -1.09289289e-02
-7.98207343e-01 8.73893425e-02 7.02718258e-01 -4.30820107e-01
-8.15316677e-01 1.03273857e-02 -3.16285670e-01 -3.82386893e-01
-1.14040397e-01 3.13135117e-01 -4.08966362e-01 1.14165366e-01
7.82105029e-01 2.18474165e-01 1.06680028e-01 -4.93448168e-01
-8.26705918e-02 6.97385728e-01 4.16306257e-01 -2.60343671e-01
5.62261760e-01 2.87487715e-01 -2.08027467e-01 -6.61423624e-01
-7.97881842e-01 -4.87834960e-01 -9.01743531e-01 2.60751158e-01
1.07139266e+00 -1.00866032e+00 -4.26304311e-01 3.60757917e-01
-5.86372077e-01 -8.54343474e-01 -4.82988179e-01 4.89110380e-01
-3.92274201e-01 -5.30649126e-02 -4.26446497e-01 -8.77969805e-03
-5.69336474e-01 -1.21661460e+00 9.15405095e-01 3.56140137e-01
-1.94884062e-01 -1.01727724e+00 2.48051971e-01 4.79931384e-01
4.62085545e-01 -2.56221712e-01 9.11977112e-01 -1.00315237e+00
-1.06834635e-01 -1.40683129e-02 -1.98379263e-01 5.96703649e-01
1.61544129e-01 -6.37295365e-01 -1.18003285e+00 -3.20868969e-01
1.45218903e-02 -3.95592660e-01 9.44437206e-01 3.36696714e-01
9.77345049e-01 2.11191416e-01 -2.48517036e-01 7.67075479e-01
1.07210445e+00 -4.58389614e-03 8.30965757e-01 4.07989979e-01
4.26764011e-01 2.38394827e-01 2.35566154e-01 3.25333655e-01
5.36183536e-01 7.69959092e-01 5.57611026e-02 -1.44854784e-01
-3.72200698e-01 -4.35498916e-02 4.02533740e-01 7.54620314e-01
7.93095678e-03 -7.74825886e-02 -5.57995617e-01 3.62360507e-01
-1.40398133e+00 -5.79855919e-01 2.67608225e-01 2.27403975e+00
1.03380895e+00 3.42172921e-01 2.99855262e-01 1.18113115e-01
4.13665891e-01 -2.49308199e-01 -7.49064386e-01 -2.61872470e-01
2.47282177e-01 4.91483659e-01 4.11042184e-01 -2.72992142e-02
-7.36335874e-01 7.04389274e-01 5.50477648e+00 6.11213803e-01
-1.46893966e+00 5.64037204e-01 2.57456601e-01 -5.16198039e-01
3.32108259e-01 -2.20413417e-01 -8.58177900e-01 8.52160811e-01
1.56999588e+00 6.95114508e-02 6.05483234e-01 6.59076273e-01
3.08605790e-01 -2.73998022e-01 -1.30786502e+00 1.00770187e+00
-4.97201830e-02 -9.18077946e-01 -5.77832043e-01 2.34916717e-01
4.22645062e-01 3.83011848e-01 2.41268538e-02 7.15306699e-01
-4.31617647e-01 -9.41255212e-01 4.96234089e-01 5.04010677e-01
9.52226937e-01 -5.22658825e-01 1.07799840e+00 3.64533365e-01
-9.20943022e-01 -1.77821606e-01 -4.19329166e-01 2.03081537e-02
-1.15469255e-01 4.37948167e-01 -8.49534690e-01 2.12067142e-01
9.45418119e-01 7.23718464e-01 -1.11683869e+00 1.11235058e+00
-4.19090129e-02 9.09105778e-01 -3.57285976e-01 1.57319590e-01
-2.34725371e-01 1.76253822e-02 -3.28024067e-02 9.77592766e-01
3.16363037e-01 -1.12168223e-01 -2.83173714e-02 9.87557590e-01
-5.03440142e-01 -1.71628252e-01 -3.88188332e-01 8.77769738e-02
3.57051671e-01 1.30347872e+00 -7.50027120e-01 -1.03290416e-01
-5.82984209e-01 1.02407992e+00 3.78419816e-01 1.40627652e-01
-8.82030666e-01 -3.51903617e-01 4.36144054e-01 3.55646282e-01
3.35363120e-01 7.86904916e-02 -4.03491080e-01 -1.15996826e+00
-2.53674030e-01 -6.70177460e-01 3.11229736e-01 -5.58233678e-01
-1.38178372e+00 7.06466198e-01 -7.70585313e-02 -1.03249729e+00
9.28986520e-02 -2.98758805e-01 -8.31406593e-01 8.63608062e-01
-1.32039332e+00 -9.80394840e-01 -4.59636718e-01 9.44409430e-01
3.91518950e-01 -2.16250673e-01 6.12886131e-01 9.90485430e-01
-8.33951294e-01 1.13311994e+00 -7.58962631e-02 -3.06245685e-03
1.08783305e+00 -1.24866664e+00 -1.25223249e-01 4.14178014e-01
-3.15512568e-02 7.36669600e-01 3.12910616e-01 -4.99416232e-01
-1.07261550e+00 -1.02779675e+00 5.39702356e-01 -4.93267566e-01
4.14220303e-01 -7.78436363e-01 -9.12829936e-01 6.00543320e-01
1.23576276e-01 1.79045141e-01 9.27746236e-01 2.72083759e-01
9.99937728e-02 -5.22431552e-01 -1.12913013e+00 7.95804411e-02
8.85232270e-01 -4.74571913e-01 -6.64557993e-01 -5.43039255e-02
3.24978113e-01 -3.30603927e-01 -9.65984821e-01 1.59736738e-01
6.10494971e-01 -9.82154667e-01 5.87644100e-01 -2.50843972e-01
1.22558087e-01 3.79102281e-03 -1.28632807e-03 -1.48734975e+00
-5.61003983e-02 -1.08243652e-01 2.49244824e-01 1.39218628e+00
5.76744020e-01 -5.08984268e-01 9.41184580e-01 5.37736416e-01
-4.92284656e-01 -7.30354965e-01 -1.13519526e+00 -5.76038122e-01
2.61952490e-01 -1.93915859e-01 4.08377618e-01 5.04339814e-01
-1.36658177e-01 8.38558078e-01 -2.78033942e-01 -7.12939724e-02
-9.74381529e-03 -1.89958829e-02 7.44215965e-01 -1.30837977e+00
-3.42300951e-01 -6.18986823e-02 -2.90604621e-01 -6.99838758e-01
2.16117784e-01 -1.08024263e+00 1.48011968e-01 -1.25739586e+00
3.05821121e-01 -5.89464903e-01 -5.84244311e-01 6.21334434e-01
3.53703499e-02 5.81291437e-01 7.84829706e-02 1.69025362e-01
-6.40227139e-01 6.49689019e-01 1.15301800e+00 4.75307070e-02
-4.71643567e-01 3.03059310e-01 -9.00549829e-01 6.07041657e-01
8.56976986e-01 -6.49396241e-01 -7.31455505e-01 -1.28403053e-01
1.80182263e-01 -1.93688840e-01 4.14320916e-01 -1.22356033e+00
1.53048009e-01 5.02425194e-01 6.95414126e-01 -4.99374159e-02
4.45050716e-01 -8.08429301e-01 -9.87609848e-02 3.64883333e-01
-3.15352231e-01 -1.98092014e-01 4.89382952e-01 5.04874825e-01
1.56723280e-02 -1.82931319e-01 8.91205966e-01 4.17826581e-04
-6.27411008e-01 6.22060895e-02 -4.21630800e-01 3.27598393e-01
6.95764422e-01 -2.57067204e-01 -3.04903924e-01 -6.96474314e-02
-1.01399815e+00 1.60730273e-01 4.14703518e-01 3.62991422e-01
2.66274482e-01 -1.00491381e+00 -1.92852646e-01 6.52592599e-01
3.23995738e-03 -2.68180221e-01 3.09210211e-01 1.37565041e+00
-6.81669787e-02 2.53967792e-01 -6.41959906e-01 -5.02626419e-01
-1.02773297e+00 3.83652538e-01 6.11451030e-01 -3.48654866e-01
-5.52142322e-01 6.83028698e-01 3.68392467e-01 -4.32950228e-01
2.60235727e-01 -6.35112047e-01 -8.24037120e-02 9.13471133e-02
1.73845068e-01 2.48120725e-01 5.04893720e-01 -3.69467288e-01
-4.45519477e-01 4.13190186e-01 6.27599424e-03 1.79808825e-01
1.24117935e+00 -3.10347527e-01 1.70128256e-01 3.64968538e-01
1.29929543e+00 -6.71390723e-03 -1.09868002e+00 -3.13182808e-02
-2.66003221e-01 9.63290781e-02 1.38060600e-01 -1.02723980e+00
-1.38416135e+00 1.08088565e+00 1.15104651e+00 -2.26652011e-01
1.34133065e+00 1.35812223e-01 1.02337551e+00 3.75922561e-01
2.90905356e-01 -9.81189549e-01 8.12524557e-02 1.66407138e-01
5.17454207e-01 -1.24484313e+00 -1.31597340e-01 2.32271180e-02
-7.24803984e-01 8.97024095e-01 8.56416643e-01 -9.60150808e-02
8.11683357e-01 1.30926967e-01 -6.90202266e-02 -3.99437994e-01
-5.02366006e-01 -1.61219299e-01 2.43351460e-01 5.65469503e-01
1.99069068e-01 -2.17037544e-01 -1.57338142e-01 1.37363946e+00
-3.95028472e-01 4.54134166e-01 4.41439003e-01 6.00945830e-01
-2.52162158e-01 -1.08180690e+00 -6.31621182e-02 8.19185138e-01
-5.64661562e-01 -2.32203886e-01 -2.79499471e-01 9.58992839e-01
7.38534391e-01 8.22615921e-01 1.17968328e-01 -6.06269836e-01
5.84000528e-01 1.61871225e-01 4.54095125e-01 -1.00433671e+00
-8.83797646e-01 -1.08183205e-01 -4.11426164e-02 -4.93405133e-01
-4.03647035e-01 -4.68427539e-01 -1.44213057e+00 -9.65033919e-02
-4.16536808e-01 2.88081139e-01 2.90918648e-01 1.03274953e+00
5.15474141e-01 7.82915354e-01 1.62936434e-01 -8.28472912e-01
-5.09203732e-01 -1.24581432e+00 -8.41298461e-01 2.69997835e-01
3.79548758e-01 -8.52091789e-01 -1.15704104e-01 5.86022288e-02] | [13.452412605285645, 3.5177111625671387] |
d39eba28-2727-47ba-9c9a-192ee5875260 | words-with-consistent-diachronic-usage | null | null | https://onlinelibrary.wiley.com/doi/full/10.1111/cogs.12963 | https://onlinelibrary.wiley.com/doi/epdf/10.1111/cogs.12963 | Words with Consistent Diachronic Usage Patterns are Learned Earlier: A Computational Analysis Using Temporally Aligned Word Embeddings | In this study, we use temporally aligned word embeddings and a large diachronic corpus of English to quantify language change in a data-driven, scalable way, which is grounded in language use. We show a unique and reliable relation between measures of language change and age of acquisition (AoA) while controlling for frequency, contextual diversity, concreteness, length, dominant part of speech, orthographic neighborhood density, and diachronic frequency variation. We analyze measures of language change tackling both the change in lexical representations and the change in the relation between lexical representations and the words with the most similar usage patterns, showing that they capture different aspects of language change. Our results show a unique relation between language change and AoA, which is stronger when considering neighborhood-level measures of language change: Words with more coherent diachronic usage patterns tend to be acquired earlier. The results support theories positing a link between ontogenetic and ethnogenetic processes in language. | ['Marco Marelli', 'Federico Bianchi', 'Giovanni Cassani'] | 2021-04-20 | null | null | null | cognitive-science-2021-4 | ['diachronic-word-embeddings'] | ['natural-language-processing'] | [-3.88519824e-01 -4.27134007e-01 -4.72137839e-01 -1.81544468e-01
-5.63062727e-02 -7.21130908e-01 1.07665956e+00 8.24834108e-01
-1.04933012e+00 3.41154605e-01 1.24587119e+00 -2.61331320e-01
-2.64807314e-01 -1.06286263e+00 -5.68311036e-01 -4.71771628e-01
-3.25680561e-02 5.32553159e-02 -1.51007175e-01 -4.96807247e-01
2.55918652e-01 2.70626962e-01 -1.45266080e+00 -2.12080076e-01
7.28764832e-01 7.96142891e-02 2.15906486e-01 2.25621507e-01
-4.25360471e-01 6.94713742e-02 -3.58280949e-02 -2.62895733e-01
1.16638280e-02 -4.90242362e-01 -5.77310026e-01 -3.49147469e-01
5.43616176e-01 6.25692829e-02 -3.70238155e-01 9.56798375e-01
4.35720712e-01 3.07825822e-02 9.62477326e-01 -2.22099736e-01
-1.08932948e+00 1.09567022e+00 -2.41249949e-01 8.44245851e-01
4.48649824e-01 1.29456982e-01 1.17249942e+00 -1.10424292e+00
1.12260222e+00 1.40679276e+00 7.10895896e-01 3.78205448e-01
-1.14206600e+00 -4.23942596e-01 2.95647234e-01 3.51291560e-02
-1.30960643e+00 -4.60455626e-01 5.90533912e-01 -9.22645211e-01
8.56804192e-01 -1.68313712e-01 1.29542446e+00 1.12821770e+00
4.63960320e-01 1.28648400e-01 1.38798237e+00 -9.25067127e-01
1.55429602e-01 1.56930879e-01 4.93723869e-01 4.31511045e-01
6.32984161e-01 3.57561111e-01 -7.50178695e-01 1.93391815e-01
5.15304148e-01 2.18200758e-01 -4.26584817e-02 -1.96165778e-02
-1.34386325e+00 9.83695805e-01 1.93418890e-01 1.25646734e+00
-3.97487164e-01 1.59578577e-01 3.67858618e-01 4.72126842e-01
7.11944103e-01 5.93603909e-01 -5.98785460e-01 -6.95526659e-01
-7.03836441e-01 1.51492342e-01 3.26766700e-01 2.06840336e-01
8.09895337e-01 1.07186101e-02 1.91337511e-01 1.01591659e+00
3.77082735e-01 6.34743929e-01 1.21951759e+00 -3.96040529e-01
-1.76325832e-02 6.58830285e-01 -5.79433680e-01 -8.04325521e-01
-2.89103866e-01 -6.84648659e-03 3.17060649e-02 -1.04609653e-01
5.02987921e-01 1.66714378e-02 -6.02322578e-01 2.27319312e+00
2.60306209e-01 -4.22521204e-01 -1.57579571e-01 3.66665721e-01
4.74936932e-01 6.05765283e-01 3.61737341e-01 -3.66721630e-01
1.54973483e+00 -2.99209386e-01 -4.99039561e-01 -2.67987698e-01
7.57836282e-01 -5.95407665e-01 1.50517285e+00 -1.28738314e-01
-9.14962709e-01 -6.44481003e-01 -8.75786245e-01 -2.10732773e-01
-8.38317692e-01 -6.91502512e-01 4.69788700e-01 7.92958915e-01
-9.11144912e-01 6.91614568e-01 -8.43178988e-01 -9.92089212e-01
-1.69133112e-01 -2.10241422e-01 -4.16841328e-01 1.01253316e-01
-1.12841904e+00 1.05612016e+00 2.88307965e-01 -5.54449141e-01
-3.98102671e-01 -9.49836612e-01 -9.12741780e-01 -3.93110961e-01
-2.81696677e-01 -2.07768559e-01 8.87175858e-01 -1.00228143e+00
-1.24724042e+00 1.14453983e+00 -5.58723509e-02 -6.40748218e-02
4.25728969e-02 -3.88494194e-01 -7.60326505e-01 -3.46036851e-01
1.28706783e-01 3.96825284e-01 4.94499534e-01 -7.21033037e-01
-3.55309039e-01 -7.62828887e-01 -3.99991781e-01 1.53313100e-01
-8.36140931e-01 6.65775612e-02 2.21581444e-01 -7.92524159e-01
-3.91925238e-02 -6.08747959e-01 1.13139264e-01 -1.13632575e-01
4.86622155e-01 -5.40919006e-01 2.97565609e-02 -9.48255777e-01
1.66650033e+00 -2.46772814e+00 3.65605801e-01 4.29077297e-02
1.39718935e-01 -7.30661079e-02 -1.45376161e-01 8.98645043e-01
1.49387062e-01 4.28367764e-01 -3.32092829e-02 6.13824017e-02
1.71850234e-01 2.66882211e-01 -9.17778984e-02 7.92695642e-01
-1.00025639e-01 1.01796496e+00 -1.18565750e+00 -3.45413685e-01
2.10922942e-01 3.96129906e-01 -6.60258472e-01 -3.44273925e-01
8.22791234e-02 7.99471214e-02 -6.90095350e-02 3.09926927e-01
1.84747159e-01 4.44182456e-01 3.46214682e-01 3.97400767e-01
-9.34471309e-01 6.15068853e-01 -4.75391656e-01 1.66150033e+00
-8.06859076e-01 9.46392298e-01 -3.60817969e-01 -5.13908327e-01
7.12069154e-01 -8.04749429e-02 2.22997982e-02 -1.09694767e+00
2.43478626e-01 4.35846925e-01 7.18322337e-01 -4.63695347e-01
7.45509505e-01 -4.41763937e-01 -2.96813667e-01 5.57978153e-01
1.18206419e-01 -1.74099907e-01 1.64320379e-01 -2.22448990e-01
9.51923728e-01 -2.49618141e-04 8.36801767e-01 -1.03483856e+00
4.71840873e-02 -2.58108765e-01 4.23148274e-01 2.45614082e-01
-1.53964356e-01 8.84917825e-02 2.24672988e-01 -3.69560212e-01
-1.20670915e+00 -1.55060565e+00 -6.76870048e-01 1.34125650e+00
-1.34806708e-02 -4.37534362e-01 -4.34965044e-01 -7.59288073e-02
1.73355848e-01 1.02224731e+00 -9.93558288e-01 -4.81669724e-01
-7.33867288e-01 -4.17716622e-01 2.19479859e-01 3.34073335e-01
-1.91437334e-01 -1.20492268e+00 -6.45819485e-01 1.70678943e-01
4.20185506e-01 -5.24670362e-01 -5.34673035e-01 -1.20980352e-01
-1.08766663e+00 -6.29432857e-01 -4.80215400e-01 -7.81856656e-01
3.41630757e-01 -4.33689803e-02 1.09689462e+00 8.75757262e-02
-3.27539206e-01 5.82129002e-01 -5.16967595e-01 -4.69705760e-01
-6.30961001e-01 2.10970700e-01 5.23949265e-01 -5.67869842e-01
7.40915000e-01 -6.60550058e-01 -5.02862811e-01 -4.07083750e-01
-8.23840976e-01 -5.80161989e-01 2.95376778e-01 3.58507842e-01
3.12982768e-01 -3.92451584e-01 4.09915954e-01 -5.48860669e-01
7.59526491e-01 -8.24611068e-01 -1.37233979e-03 -9.70502477e-03
-1.09335732e+00 2.39607900e-01 3.69043738e-01 -6.37894511e-01
-7.46028960e-01 -7.50605404e-01 -1.16665237e-01 3.62435639e-01
-2.21409798e-01 6.38607800e-01 3.02575916e-01 5.92939258e-01
8.11744392e-01 2.06319362e-01 1.13132082e-01 -5.04968524e-01
6.95937872e-01 5.84166765e-01 1.66039497e-01 -6.65018737e-01
7.00591564e-01 3.26346368e-01 -6.92057610e-01 -1.41143942e+00
-3.61287445e-01 -3.80639762e-01 -1.01165915e+00 -1.00388534e-01
1.08540606e+00 -9.34434772e-01 -8.38458985e-02 4.39156711e-01
-7.63560593e-01 -4.49421257e-01 -6.51243329e-01 9.49245811e-01
-1.99891865e-01 2.11394593e-01 -3.46699953e-01 -4.65896487e-01
-1.15651965e-01 -5.94136715e-01 7.56581783e-01 9.52323601e-02
-8.86976063e-01 -1.43948567e+00 1.06327701e+00 -2.68682420e-01
3.76861840e-01 2.87878931e-01 1.34722936e+00 -6.74584627e-01
1.01843271e-02 2.99311608e-01 3.05127740e-01 6.34342656e-02
5.93191087e-01 1.73395708e-01 -3.94367903e-01 -2.40887210e-01
-2.70798206e-01 1.77875996e-01 7.54908323e-01 3.36076319e-01
1.00361489e-01 -1.64079756e-01 -1.67757630e-01 3.38017613e-01
1.53753901e+00 3.03378761e-01 2.52455384e-01 3.53762984e-01
6.37608588e-01 6.37256622e-01 1.01520374e-01 1.19487233e-01
3.51330161e-01 5.82150578e-01 -5.35384893e-01 4.99421656e-01
-4.36995447e-01 -5.00341415e-01 9.67252314e-01 1.75776458e+00
3.72705199e-02 1.64282113e-01 -1.24134254e+00 1.28067958e+00
-1.11636043e+00 -1.00550759e+00 -1.06472857e-01 2.24771738e+00
1.16118324e+00 -1.16631677e-02 3.36194545e-01 -4.02655266e-02
3.81260097e-01 4.03408647e-01 -2.02209353e-01 -1.00126636e+00
9.47908610e-02 7.14712262e-01 3.64365906e-01 7.04239309e-01
-1.49307609e-01 1.14618433e+00 7.18823814e+00 4.40051556e-01
-1.24170160e+00 2.96141595e-01 2.00886354e-02 -3.19726259e-01
-9.93259072e-01 -1.85243443e-01 -4.80717033e-01 5.60936868e-01
1.34305012e+00 -5.60973704e-01 3.32985580e-01 4.02993292e-01
1.58351660e-01 -6.42330050e-02 -1.25240707e+00 5.19209087e-01
6.14719987e-02 -9.50628817e-01 -5.36951162e-02 2.10275605e-01
6.47177994e-01 2.55813271e-01 1.17265165e-01 2.56449580e-01
3.45979661e-01 -8.65373194e-01 1.06619883e+00 4.73482311e-01
8.63329530e-01 -4.40945566e-01 2.10637748e-01 -6.61064312e-02
-1.06606364e+00 -4.65759672e-02 -1.04329303e-01 -6.28292978e-01
1.00302249e-01 5.04992545e-01 -5.12047768e-01 -1.70895401e-02
6.44502461e-01 7.61253953e-01 -7.24374533e-01 3.51864636e-01
-2.32307121e-01 7.39262044e-01 -1.80966049e-01 -2.52034694e-01
6.62692562e-02 -3.83718789e-01 4.47062343e-01 1.26602328e+00
4.78341609e-01 -2.03083709e-01 -3.92439693e-01 6.46658778e-01
2.16352701e-01 6.59846008e-01 -9.64561164e-01 -7.39179730e-01
8.95510852e-01 7.95859635e-01 -8.54541719e-01 4.85939847e-04
-6.56570971e-01 8.23008478e-01 4.23720479e-01 -3.50262262e-02
-3.46163541e-01 -5.04306220e-02 1.16136754e+00 3.92177761e-01
1.88202694e-01 -7.16832221e-01 -3.48353237e-01 -9.45805609e-01
-1.40960246e-01 -5.94150722e-01 3.38376537e-02 1.44207440e-02
-1.24182463e+00 3.00034672e-01 1.58668175e-01 -4.92533475e-01
-2.70916909e-01 -6.08873188e-01 -5.87936223e-01 6.92233443e-01
-1.01998091e+00 -7.45584548e-01 3.04260224e-01 2.41349265e-01
5.56082010e-01 -1.14342391e-01 8.09056580e-01 1.72550693e-01
-5.05695045e-01 5.94183385e-01 3.88984710e-01 -7.44252205e-02
5.49364746e-01 -1.17632902e+00 6.77452743e-01 7.36830473e-01
5.06169736e-01 9.78708327e-01 5.22100747e-01 -8.42611074e-01
-1.06908941e+00 -6.35761023e-01 1.36948240e+00 -7.49521911e-01
1.14990592e+00 -5.71153700e-01 -5.62515497e-01 6.32551193e-01
3.95077080e-01 -6.60486460e-01 9.64034498e-01 7.25189924e-01
-5.00696898e-01 1.46892995e-01 -6.75999701e-01 9.64436471e-01
1.59237611e+00 -9.94446516e-01 -1.13425469e+00 -1.13965176e-01
1.11992264e+00 4.24875438e-01 -1.22097468e+00 -2.82118898e-02
8.57468963e-01 -6.74868405e-01 6.50811613e-01 -3.52747321e-01
5.32088220e-01 2.86763936e-01 -2.87021577e-01 -1.52812219e+00
-6.77314937e-01 -1.81336701e-01 4.09630150e-01 1.25827360e+00
5.11699438e-01 -6.34601116e-01 6.13489412e-02 -3.46233174e-02
-1.71873897e-01 -5.12888730e-01 -1.19281089e+00 -7.73724854e-01
9.37406242e-01 -1.56794101e-01 6.14250779e-01 1.17699730e+00
1.38987914e-01 3.40500593e-01 2.37427652e-01 -3.81444931e-01
-9.55963731e-02 -2.29999363e-01 2.15883270e-01 -1.46798408e+00
2.52746791e-03 -8.50032210e-01 -5.76570868e-01 -3.27745587e-01
2.59440482e-01 -1.16042900e+00 -1.13707803e-01 -1.37334168e+00
2.31672358e-02 1.40663413e-02 -1.72698349e-01 -9.59804729e-02
-2.02712025e-02 -2.88020253e-01 2.39787757e-01 3.58116627e-02
3.69285271e-02 5.21399856e-01 7.83602595e-01 1.90122470e-01
-6.05967045e-01 -7.23361969e-01 -7.18806505e-01 8.11168492e-01
6.31034791e-01 -4.52511907e-01 -3.26751947e-01 -6.54651046e-01
5.28642237e-01 -6.88955009e-01 -1.35855973e-01 -8.43981564e-01
-3.49419534e-01 -3.01257491e-01 1.75939143e-01 2.61099692e-02
-8.55158120e-02 -6.08082652e-01 9.41729993e-02 8.81171346e-01
-2.80679464e-01 7.06675291e-01 2.27616817e-01 2.44212806e-01
7.29918033e-02 -2.43977278e-01 6.72947049e-01 1.13084316e-01
-9.12837565e-01 1.95511788e-01 -8.16322386e-01 2.88989514e-01
9.80311215e-01 -5.40026367e-01 -3.76295671e-03 2.26682305e-01
-4.43873405e-01 -3.71573418e-01 8.81896794e-01 9.60329831e-01
1.93322524e-01 -1.59971333e+00 -6.92093432e-01 1.83122501e-01
3.50213945e-01 -9.06475306e-01 -1.98032428e-03 5.18831611e-01
-4.81887430e-01 2.14639083e-02 -2.24198997e-01 -4.55435887e-02
-9.06612992e-01 5.94698966e-01 1.08643733e-01 1.20242640e-01
-5.99610448e-01 6.69185519e-01 1.22042008e-01 -6.58442736e-01
-3.15093905e-01 -6.80505037e-01 -2.75705516e-01 7.86317170e-01
2.59263575e-01 6.18830860e-01 -3.10552299e-01 -8.92951846e-01
-5.43913007e-01 9.17759180e-01 -4.67157997e-02 -4.41015273e-01
1.47622228e+00 -4.29164171e-01 -3.85807335e-01 1.44730198e+00
1.22987986e+00 6.51555300e-01 -6.24676108e-01 -2.61313558e-01
2.40660951e-01 -2.55051315e-01 -1.79414570e-01 -4.10215020e-01
-5.01524568e-01 7.27615118e-01 9.64570820e-01 1.20025925e-01
4.47745919e-01 3.89509141e-01 6.87889695e-01 -9.31495503e-02
1.97031930e-01 -1.38705027e+00 -2.87845079e-02 6.48198307e-01
8.05174768e-01 -6.81512356e-01 -2.69909538e-02 2.70820975e-01
-1.90944999e-01 8.17075372e-01 3.99796635e-01 -2.07610205e-01
1.07561266e+00 -1.98991075e-01 -6.03390858e-03 -4.59306151e-01
-4.42545563e-01 -3.26653212e-01 3.06491047e-01 4.73894060e-01
9.73715544e-01 4.82110202e-01 -1.27482629e+00 3.75072002e-01
-9.70880806e-01 -4.78425473e-01 4.10161108e-01 6.12248361e-01
-5.95764637e-01 -1.27560925e+00 -8.62318575e-02 3.16479951e-01
-2.88743854e-01 -4.49029773e-01 -5.25478721e-01 1.22444749e+00
4.97957855e-01 4.36445415e-01 8.43785584e-01 -3.62025499e-01
2.70315826e-01 4.99041975e-01 8.24450195e-01 -9.86038268e-01
-6.68661833e-01 -2.47226730e-01 -4.59623756e-03 -2.20476359e-01
-3.43901604e-01 -1.21270895e+00 -1.14493191e+00 -4.91609067e-01
4.41922396e-02 -9.13011804e-02 5.83558321e-01 9.43273485e-01
1.62401527e-01 3.77861232e-01 2.72671819e-01 -4.32197422e-01
-1.72111988e-01 -1.19778430e+00 -8.03874552e-01 4.90234852e-01
1.83120966e-01 -4.65682894e-01 -2.88642675e-01 -2.49215066e-01] | [10.189831733703613, 8.995393753051758] |
c8fcea18-5540-4b5e-9410-b1fb9b42763d | seizure-prediction-with-long-term-ieeg | 2201.04137 | null | https://arxiv.org/abs/2201.04137v1 | https://arxiv.org/pdf/2201.04137v1.pdf | Seizure prediction with long-term iEEG recordings: What can we learn from data nonstationarity? | Repeated epileptic seizures impair around 65 million people worldwide and a successful prediction of seizures could significantly help patients suffering from refractory epilepsy. For two dogs with yearlong intracranial electroencephalography (iEEG) recordings, we studied the influence of time series nonstationarity on the performance of seizure prediction using in-house developed machine learning algorithms. We observed a long-term evolution on the scale of weeks or months in iEEG time series that may be represented as switching between certain meta-states. To better predict impending seizures, retraining of prediction algorithms is therefore necessary and the retraining schedule should be adjusted to the change in meta-states. There is evidence that the nature of seizure-free interictal clips also changes with the transition between meta-states, accwhich has been shown relevant for seizure prediction. | ['Ronald Tetzlaff', 'Jens Müller', 'Matthias Eberlein', 'Hongliu Yang'] | 2022-01-11 | null | null | null | null | ['seizure-prediction'] | ['medical'] | [ 1.63547739e-01 -3.68537247e-01 1.73090756e-01 -2.55755752e-01
-3.71552467e-01 -2.93727219e-01 6.05894327e-01 5.25940731e-02
-2.54230380e-01 1.08108878e+00 7.37298802e-02 -4.09219772e-01
-5.78065634e-01 -1.15993761e-01 -5.40680110e-01 -8.22575331e-01
-9.93308187e-01 3.24314892e-01 1.28300279e-01 -1.45206422e-01
2.02790082e-01 5.84172726e-01 -1.32662177e+00 4.99763727e-01
5.65572977e-01 7.21951246e-01 3.69948953e-01 5.79668045e-01
2.00961262e-01 4.36683089e-01 -9.35272396e-01 4.69828665e-01
7.86280259e-02 -4.85063255e-01 -3.19860131e-01 1.15460552e-01
-4.77747887e-01 1.45582125e-01 -2.57525772e-01 6.94954872e-01
3.94534379e-01 -4.06370193e-01 5.98400235e-01 -1.08175528e+00
1.82040811e-01 5.55346012e-01 -2.32418478e-01 8.07248354e-01
3.32748115e-01 2.45839879e-01 -6.62910268e-02 -1.77789569e-01
4.40165967e-01 3.83648902e-01 7.73027897e-01 3.70901167e-01
-1.45351326e+00 -9.73934054e-01 -5.71260564e-02 5.63265324e-01
-1.25393343e+00 -3.59590173e-01 6.62780941e-01 -7.13321567e-01
1.20505250e+00 1.81463495e-01 1.53026533e+00 1.38124573e+00
1.32473373e+00 -1.14404801e-02 1.51459026e+00 -1.75837740e-01
3.65102172e-01 -3.39588910e-01 -8.59909505e-02 -2.96277612e-01
6.24459013e-02 5.34537375e-01 -6.71886206e-01 -3.85657012e-01
4.32199776e-01 -7.99863711e-02 -4.87360954e-01 4.32117246e-02
-1.12387133e+00 4.89467770e-01 -5.89160100e-02 7.86741793e-01
-7.61436999e-01 -7.85591900e-02 3.51097107e-01 7.66953468e-01
3.36488932e-01 7.83123136e-01 -9.37427104e-01 -5.52402556e-01
-1.21374667e+00 -6.97529167e-02 6.22681677e-01 3.82247895e-01
4.54290420e-01 6.22857660e-02 2.59144545e-01 4.47389483e-01
-3.60951036e-01 1.69584095e-01 1.20680380e+00 -3.20368528e-01
-1.40124083e-01 2.81046599e-01 5.06626889e-02 -4.93217707e-01
-8.12587619e-01 -6.43934369e-01 -7.77945280e-01 -7.18334243e-02
1.11053020e-01 -4.42303777e-01 -9.12520707e-01 1.54043281e+00
-3.38038206e-01 6.34366095e-01 1.06013075e-01 1.38763294e-01
6.82290271e-02 4.48189497e-01 1.84146851e-01 -9.37238157e-01
8.12343657e-01 1.20693862e-01 -7.19205678e-01 -3.68535280e-01
7.46641397e-01 -3.73248249e-01 2.44989708e-01 6.20674610e-01
-6.60216331e-01 7.82356709e-02 -1.04001188e+00 1.17219317e+00
-1.90620646e-01 -3.76656204e-01 6.17293954e-01 2.19564572e-01
-9.82472181e-01 9.69564557e-01 -1.43747950e+00 -3.96749228e-01
-4.86447196e-03 6.37144744e-01 -4.99710560e-01 7.10060358e-01
-1.21572495e+00 1.17986274e+00 4.39965248e-01 1.02814004e-01
-7.33112931e-01 -7.31790304e-01 -3.63203675e-01 -1.68322742e-01
-3.30873817e-01 -3.13905209e-01 7.82022476e-01 -7.63953328e-01
-1.32195210e+00 7.22817540e-01 -1.12049021e-01 -8.66120875e-01
4.59628910e-01 3.16679865e-01 -9.73154545e-01 -5.54795787e-02
-1.22581482e-01 3.57419491e-01 8.68083715e-01 -5.05725324e-01
-5.90510547e-01 -7.13452101e-01 -8.88607442e-01 -1.03310928e-01
-2.07866747e-02 5.95931783e-02 3.48277569e-01 -7.37796187e-01
2.28957593e-01 -1.10582614e+00 -1.78061619e-01 -9.45141256e-01
1.39012158e-01 -1.39591902e-01 7.65863955e-01 -6.54329538e-01
1.34067857e+00 -2.11894798e+00 -9.67472345e-02 1.03209235e-01
-9.37387198e-02 -1.08344287e-01 -7.51888379e-02 6.64981782e-01
-7.12889373e-01 -3.66912514e-01 -2.26944163e-01 3.39399308e-01
-5.64875484e-01 8.55786279e-02 -3.41955364e-01 7.58544266e-01
1.02253735e-01 5.72319329e-01 -5.27125299e-01 2.92829692e-01
2.32286140e-01 3.96369427e-01 -1.96450222e-02 1.70586899e-01
5.27375191e-02 9.43296731e-01 -2.30892941e-01 3.97465646e-01
3.01103055e-01 -2.16888756e-01 1.81903958e-01 3.33725929e-01
-4.06996071e-01 2.46139854e-01 -3.16906601e-01 1.46333265e+00
-2.40611583e-01 9.63875651e-01 -5.19736230e-01 -1.12274325e+00
9.27710235e-01 4.86561447e-01 8.91276240e-01 -6.98377669e-01
9.42863449e-02 4.13049489e-01 6.20622694e-01 -3.38047892e-01
-1.75734565e-01 -1.47988334e-01 1.79092437e-01 3.05764288e-01
-5.76615557e-02 -2.74189264e-01 1.15054110e-02 -4.12414372e-01
1.56153643e+00 -2.70302355e-01 3.42937976e-01 -7.00833380e-01
1.02180950e-01 4.98022214e-02 6.07763350e-01 4.79511201e-01
-5.78108653e-02 1.34253308e-01 5.03214896e-01 -5.95784247e-01
-7.46802628e-01 -6.32980227e-01 -8.28759253e-01 2.42084414e-01
-1.69558391e-01 -2.93129295e-01 -4.64922875e-01 2.99067181e-02
-2.27315530e-01 8.70924234e-01 -7.65473962e-01 -5.72119176e-01
-4.64973629e-01 -1.41948783e+00 1.80675060e-01 3.26463640e-01
1.69021070e-01 -1.19609821e+00 -1.20601320e+00 8.20423245e-01
1.48113310e-01 -7.93644190e-01 1.10711984e-01 1.02040577e+00
-1.07457662e+00 -9.52113390e-01 -6.31006300e-01 -6.02372050e-01
3.75742823e-01 -6.59649193e-01 7.46492624e-01 -7.19082132e-02
-4.26738173e-01 1.65125176e-01 -2.84904331e-01 -7.32665241e-01
-3.33040923e-01 -9.55768898e-02 3.63309145e-01 -3.47150952e-01
5.64169884e-01 -1.15727890e+00 -6.64527535e-01 1.46943152e-01
-6.10306382e-01 -5.79263419e-02 4.96524960e-01 7.09991097e-01
6.78358018e-01 3.27428102e-01 9.52315748e-01 -2.96021461e-01
6.45480573e-01 -5.40790200e-01 -7.66096711e-01 2.61628330e-01
-8.83962572e-01 -6.05171639e-03 4.54704106e-01 -8.76818538e-01
-4.50084835e-01 6.42485321e-02 2.65925884e-01 -2.34937027e-01
-3.08712900e-01 5.87576032e-01 1.98327333e-01 -4.66744006e-02
3.51749033e-01 5.99336147e-01 -1.20100223e-01 -6.94058910e-02
-6.71150446e-01 6.33563936e-01 3.68556798e-01 1.25980660e-01
2.71126300e-01 1.52216107e-01 -4.87605296e-03 -1.05895090e+00
1.80097707e-02 -2.09882528e-01 -4.10306662e-01 -2.27588534e-01
5.53011775e-01 -9.10650015e-01 -3.09906095e-01 8.16448271e-01
-7.84117520e-01 -6.27761304e-01 8.97191316e-02 9.27519619e-01
-7.93742597e-01 -2.62274057e-01 -3.61647725e-01 -5.32814205e-01
-2.59750754e-01 -1.01803744e+00 3.62669736e-01 -3.94021422e-02
-6.19089961e-01 -9.49153543e-01 5.68028092e-01 -5.29562056e-01
2.87517548e-01 3.56688052e-01 9.75811601e-01 -9.58658874e-01
-2.09688142e-01 -4.47423965e-01 4.89217877e-01 -9.32729542e-02
2.89986372e-01 -2.75155514e-01 -7.65969574e-01 -5.37384212e-01
4.83795017e-01 3.05193782e-01 4.23812479e-01 8.65628302e-01
1.01423943e+00 4.30227257e-02 -6.05286777e-01 6.97212338e-01
1.23329008e+00 1.00801766e+00 7.04426229e-01 7.32844710e-01
-3.98525566e-01 3.63249093e-01 1.39289007e-01 5.09277165e-01
-2.26091713e-01 3.99261862e-01 1.21103086e-01 4.52693135e-01
2.91616410e-01 8.42657238e-02 3.37662548e-01 6.63890123e-01
-5.07514402e-02 1.76190913e-01 -1.11541641e+00 8.23694885e-01
-1.41766608e+00 -9.46835935e-01 1.79643095e-01 2.28354025e+00
6.99943721e-01 2.60792345e-01 2.67637074e-02 1.13422088e-01
5.79365790e-01 -2.41453901e-01 -8.09250236e-01 -1.44719020e-01
-3.40107352e-01 2.50118822e-01 5.62417626e-01 9.30258110e-02
-5.01356006e-01 4.68397677e-01 7.78059196e+00 2.99215764e-01
-1.75152028e+00 -2.00164840e-01 4.36854601e-01 -1.75716370e-01
1.55202240e-01 6.19273707e-02 -3.98755431e-01 9.51787770e-01
1.54874671e+00 -8.58063698e-01 5.23858845e-01 2.62237459e-01
5.28640270e-01 -3.20399255e-01 -1.02722049e+00 1.12604237e+00
-7.08234310e-02 -1.09694684e+00 -3.19198668e-01 3.73081148e-01
5.72091162e-01 4.87809449e-01 -1.31771192e-01 8.66957679e-02
-8.72960910e-02 -9.25695539e-01 3.75424564e-01 7.38523781e-01
7.49331772e-01 -7.04937458e-01 7.10309446e-01 4.59997684e-01
-8.58167589e-01 -2.00532272e-01 -1.05886154e-01 -1.15050070e-01
1.18141778e-01 6.11235678e-01 -1.02523375e+00 1.03443019e-01
9.08701003e-01 8.99012327e-01 -6.72846317e-01 1.26964438e+00
1.17365988e-02 7.45915115e-01 -4.77183789e-01 1.94359183e-01
-1.18766151e-01 -1.43428594e-01 8.86241198e-01 6.34605110e-01
8.10985446e-01 2.51491785e-01 -5.10243773e-01 4.11635876e-01
6.81292593e-01 -2.95936078e-01 -5.34136951e-01 -2.95253724e-01
2.79868990e-01 5.82429647e-01 -9.00442719e-01 -2.68510822e-02
-1.61708564e-01 7.97264874e-01 -1.42486423e-01 3.42229992e-01
-5.71213067e-01 -3.16747427e-01 3.79186124e-01 3.49428445e-01
2.08940998e-01 3.40742841e-02 -1.83065236e-01 -1.27066779e+00
4.09691129e-03 -6.13028705e-01 2.32578084e-01 -7.91573107e-01
-9.92891490e-01 1.14401042e+00 1.41142339e-01 -1.27698672e+00
-8.80824625e-01 -1.12384751e-01 -9.16014969e-01 6.63745582e-01
-1.06660485e+00 -4.81584847e-01 3.00595552e-01 7.90969312e-01
4.78360265e-01 -2.78434813e-01 1.00559711e+00 -4.00998354e-01
-2.45243594e-01 6.10847808e-02 2.05883056e-01 -4.70447153e-01
5.33570349e-01 -9.72490549e-01 6.84059709e-02 5.86081803e-01
-1.22829504e-01 3.44682157e-01 1.30298305e+00 -1.01293778e+00
-8.91494930e-01 -8.42530906e-01 8.25469613e-01 -1.77077770e-01
1.21856630e+00 -3.07880063e-03 -1.05180609e+00 7.88102329e-01
2.96761841e-01 -4.19842094e-01 7.00076044e-01 -5.22692986e-02
4.85061795e-01 -4.58996743e-02 -8.88875306e-01 2.87763387e-01
5.55814385e-01 -3.62361491e-01 -1.02942395e+00 3.21530223e-01
-7.76297227e-02 -2.41497830e-02 -9.04012799e-01 7.89913118e-01
5.71706235e-01 -9.42357183e-01 2.39435717e-01 -4.89154190e-01
-4.94723618e-02 2.06700459e-01 2.40170583e-01 -1.86630404e+00
-1.76672623e-01 -7.62855649e-01 8.49934593e-02 3.93329442e-01
4.28919107e-01 -1.12600505e+00 7.22325265e-01 6.47515833e-01
-9.58296135e-02 -7.84009159e-01 -1.04217076e+00 -9.81028914e-01
2.28071570e-01 -1.99027196e-01 7.87030518e-01 5.88360369e-01
5.73665559e-01 -3.78563292e-02 -1.24507740e-01 1.96651652e-01
4.36504215e-01 1.65174589e-01 5.73054664e-02 -1.32925797e+00
1.81435458e-02 -1.43527448e-01 -9.61966217e-01 -1.68114856e-01
3.28325987e-01 -6.35580540e-01 -8.75720829e-02 -9.66516078e-01
2.35523228e-02 -1.47034585e-01 -5.75619519e-01 3.88659924e-01
1.98448211e-01 -6.55024871e-02 -2.30423003e-01 3.29962879e-01
-5.85622452e-02 2.78147191e-01 6.76716149e-01 8.73821229e-02
-6.95126355e-01 4.46340531e-01 -4.49954867e-02 6.67825699e-01
1.03889370e+00 -6.16971731e-01 -4.35957640e-01 -7.62681141e-02
1.24227926e-01 6.08857214e-01 4.76119146e-02 -1.42592418e+00
3.35091829e-01 -5.23029491e-02 6.23462677e-01 -8.04264188e-01
1.93607047e-01 -9.39704597e-01 1.08106244e+00 8.49385738e-01
-1.12104796e-01 3.47005308e-01 5.70762932e-01 6.22077763e-01
-3.20686430e-01 1.03167653e-01 6.51874006e-01 1.45327393e-02
-4.89444494e-01 3.35569173e-01 -1.32753491e+00 -4.60446686e-01
1.49278736e+00 -5.39014459e-01 2.36259043e-01 -2.33597055e-01
-1.35819447e+00 1.54986586e-02 4.25484687e-01 3.40695888e-01
3.78383726e-01 -8.75251114e-01 -5.59732318e-01 3.38488162e-01
5.92172407e-02 -8.38701487e-01 3.34694624e-01 1.16592109e+00
-2.64495611e-01 7.01151133e-01 -6.38054729e-01 -7.17975259e-01
-1.09608078e+00 2.58244306e-01 6.20881557e-01 -7.60529488e-02
-8.61222625e-01 4.59476411e-01 -2.52686888e-01 4.33044434e-01
-2.47447956e-02 -2.76983798e-01 -4.02171195e-01 2.02829763e-01
6.21983707e-01 4.56529707e-02 5.33801675e-01 -3.83740753e-01
-4.04882878e-01 1.67968109e-01 -1.39995456e-01 7.59467855e-02
1.91931808e+00 -8.31991583e-02 -3.39478523e-01 8.11342597e-01
8.61029148e-01 -5.75221300e-01 -1.20831537e+00 5.11005819e-01
2.59877652e-01 -7.12835193e-02 1.53079897e-01 -9.08263803e-01
-8.52465510e-01 3.45839918e-01 9.71631289e-01 3.03387880e-01
1.34532905e+00 -1.74562838e-02 4.95364636e-01 3.39077502e-01
9.99006093e-01 -7.27811158e-01 -5.17586410e-01 2.08951384e-01
9.04113650e-01 -7.38443136e-01 -1.56492680e-01 1.87453449e-01
-2.11944953e-01 1.21794689e+00 1.14044555e-01 -3.00678372e-01
1.09251273e+00 4.95766073e-01 -3.85994986e-02 -2.59442002e-01
-1.42966974e+00 3.35614532e-01 -7.96226114e-02 7.00817287e-01
1.39942572e-01 4.16296840e-01 -6.53326511e-01 7.27727532e-01
-6.16702318e-01 3.75539005e-01 7.78364658e-01 8.63825321e-01
-1.40171751e-01 -1.01641774e+00 -3.46297860e-01 1.06021965e+00
-4.68844742e-01 -1.64406791e-01 -1.71620548e-01 7.40804970e-01
-1.04655229e-01 9.16619599e-01 3.94900650e-01 -3.46549094e-01
3.58127020e-02 3.06037188e-01 5.77413559e-01 -4.52909529e-01
-6.20759666e-01 3.02238077e-01 -1.43872619e-01 -4.50489551e-01
-5.23304880e-01 -1.17516398e+00 -1.03908682e+00 1.06929049e-01
-5.18802047e-01 4.08361733e-01 7.01206148e-01 1.05444229e+00
4.38709080e-01 4.54846144e-01 5.24916410e-01 -4.50266182e-01
-8.10058415e-02 -1.25910425e+00 -1.02724254e+00 1.18094460e-04
5.13749003e-01 -5.70073605e-01 -1.01949453e+00 1.77050456e-01] | [13.230660438537598, 3.5234899520874023] |
052d3e35-f35e-42e7-a3e2-9e712179d647 | noise-injected-consistency-training-and | null | null | https://aclanthology.org/2022.coling-1.561 | https://aclanthology.org/2022.coling-1.561.pdf | Noise-injected Consistency Training and Entropy-constrained Pseudo Labeling for Semi-supervised Extractive Summarization | Labeling large amounts of extractive summarization data is often prohibitive expensive due to time, financial, and expertise constraints, which poses great challenges to incorporating summarization system in practical applications. This limitation can be overcome by semi-supervised approaches: consistency-training and pseudo-labeling to make full use of unlabeled data. Researches on the two, however, are conducted independently, and very few works try to connect them. In this paper, we first use the noise-injected consistency training paradigm to regularize model predictions. Subsequently, we propose a novel entropy-constrained pseudo labeling strategy to obtain high-confidence labels from unlabeled predictions, which can obtain high-confidence labels from unlabeled predictions by comparing the entropy of supervised and unsupervised predictions. By combining consistency training and pseudo-labeling, this framework enforce a low-density separation between classes, which decently improves the performance of supervised learning over an insufficient labeled extractive summarization dataset. | ['JianXin Li', 'Hongdong Zhu', 'Weifeng Jiang', 'Junnan Liu', 'Qianren Mao', 'Yiming Wang'] | null | null | null | null | coling-2022-10 | ['extractive-summarization'] | ['natural-language-processing'] | [ 3.64163399e-01 3.59219193e-01 -7.98021376e-01 -3.75499666e-01
-8.66515160e-01 -4.02909189e-01 4.42549199e-01 3.48005086e-01
-3.51369083e-01 1.10436893e+00 2.24767089e-01 -4.35624039e-03
1.75289378e-01 -5.87014854e-01 -4.57840502e-01 -6.09674156e-01
6.07435524e-01 5.00564814e-01 1.30328014e-01 1.75842568e-01
3.46103519e-01 4.07568440e-02 -1.56723392e+00 2.86475569e-01
1.50647545e+00 7.84899235e-01 1.95204541e-01 2.28580460e-01
-5.11533380e-01 7.92380929e-01 -4.93072510e-01 -2.19498575e-01
-6.16777409e-03 -9.13590848e-01 -7.99306750e-01 3.73479754e-01
-1.32015735e-01 -2.25551590e-01 1.55461058e-01 1.20334995e+00
3.00975382e-01 1.93610750e-02 1.11969090e+00 -1.04878509e+00
-5.78782678e-01 8.65131557e-01 -4.97982830e-01 -1.26747847e-01
2.77104735e-01 -2.36788198e-01 1.01056218e+00 -6.44816160e-01
4.49012399e-01 7.87967026e-01 4.79800314e-01 6.64124846e-01
-9.51103985e-01 -6.40674353e-01 2.60203630e-01 7.67749324e-02
-1.33919179e+00 -4.21901196e-01 9.44908500e-01 -2.53259778e-01
6.71078146e-01 3.01669329e-01 6.35911167e-01 9.92608368e-01
-3.25255752e-01 1.12725174e+00 1.15793097e+00 -6.67579591e-01
4.39415544e-01 5.18828273e-01 2.22900152e-01 5.14990926e-01
4.53193188e-01 -2.99863756e-01 -2.89485037e-01 -4.79354970e-02
3.39534283e-01 5.68500757e-02 -4.19637948e-01 -9.18151513e-02
-1.00502229e+00 6.16387129e-01 -2.14988440e-02 4.42807734e-01
-1.48516268e-01 -6.40733838e-01 4.42701340e-01 2.21457575e-02
8.40522826e-01 4.64474916e-01 -5.29997408e-01 -1.12097643e-01
-1.49924994e+00 -1.96035430e-01 8.30903053e-01 1.15378892e+00
9.18639183e-01 -1.03697060e-02 -1.41056150e-01 8.42882454e-01
3.78015846e-01 4.08890963e-01 7.97220647e-01 -7.01932669e-01
5.18422008e-01 8.72250617e-01 2.31975596e-03 -7.75005460e-01
-2.02146515e-01 -3.29895884e-01 -9.36848700e-01 -3.86657149e-01
3.45849693e-02 -3.05650443e-01 -8.46238732e-01 1.39085376e+00
3.68319988e-01 1.16580799e-01 3.43103677e-01 7.44157255e-01
9.04455364e-01 8.59733641e-01 2.05173209e-01 -1.06292236e+00
9.68000770e-01 -1.26377225e+00 -9.76535916e-01 -1.50694236e-01
8.96334648e-01 -6.69783056e-01 9.36062276e-01 1.82832688e-01
-7.92374253e-01 -5.04509568e-01 -1.19931507e+00 5.10467738e-02
-8.73090550e-02 5.13909161e-01 2.52328932e-01 6.00026727e-01
-5.12896836e-01 5.99227071e-01 -8.70135128e-01 -2.44070455e-01
2.65702993e-01 2.56367207e-01 -2.88053662e-01 9.71916765e-02
-1.01677597e+00 8.10940206e-01 1.05297554e+00 -2.30477247e-02
-3.22065771e-01 -2.50532269e-01 -8.48064184e-01 1.59233406e-01
6.08970523e-01 -3.38340521e-01 1.21299803e+00 -9.64296699e-01
-1.54490232e+00 6.83545709e-01 -3.03348005e-01 -2.05649227e-01
5.51316023e-01 -1.34107634e-01 -3.66416007e-01 9.37419012e-02
1.43008530e-01 4.77218509e-01 7.40758479e-01 -1.42321157e+00
-6.50743604e-01 -1.65720358e-01 -3.81053001e-01 5.59530675e-01
-7.40839005e-01 -2.07663432e-01 -4.44787920e-01 -4.51710343e-01
3.39549750e-01 -7.90872812e-01 -2.99894333e-01 -6.79271936e-01
-8.35774779e-01 -3.52250010e-01 7.18379617e-01 -6.16778076e-01
1.34488082e+00 -1.96375525e+00 -8.81535653e-03 1.50296958e-02
7.56713524e-02 5.80639422e-01 8.07933807e-02 1.94471478e-01
1.76942185e-01 4.16100025e-01 -5.63629925e-01 -4.35254067e-01
-1.17737919e-01 2.33882204e-01 -2.88618922e-01 1.42592071e-02
2.28421628e-01 6.48730457e-01 -9.90673602e-01 -1.22383928e+00
6.30993694e-02 -2.79972218e-02 -4.34270769e-01 4.46290195e-01
-3.70430350e-01 5.63940346e-01 -7.90665984e-01 6.31412923e-01
5.71826398e-01 -2.38714024e-01 4.59909528e-01 -1.70522943e-01
2.31858827e-02 7.34114721e-02 -1.04946601e+00 1.48829460e+00
-2.32303604e-01 1.78200170e-01 -2.84798801e-01 -1.34186268e+00
1.06611764e+00 3.19290310e-01 4.35952753e-01 -2.50255346e-01
3.62094462e-01 3.38923872e-01 -5.69639385e-01 -6.74721062e-01
6.72127724e-01 -2.67925292e-01 -3.94089296e-02 5.21116197e-01
3.11497618e-02 -1.52292788e-01 3.15815151e-01 2.04312325e-01
5.82103968e-01 3.27384382e-01 4.79771167e-01 6.38999511e-03
6.63224399e-01 1.46577865e-01 9.32755828e-01 4.93146122e-01
-5.99119179e-02 6.71937048e-01 3.72183323e-01 1.78395472e-02
-1.02184117e+00 -5.52080095e-01 -1.74524888e-01 6.78468645e-01
3.71580184e-01 -5.88703394e-01 -8.60896468e-01 -1.21020687e+00
-3.80629867e-01 7.88523078e-01 -1.84546039e-01 -3.00263584e-01
-3.48338932e-01 -8.94858778e-01 3.15545559e-01 4.39700365e-01
6.98845208e-01 -9.38455164e-01 -6.58597201e-02 1.73285708e-01
-6.14368141e-01 -1.08908343e+00 -3.24046820e-01 2.75016904e-01
-1.09710991e+00 -9.94271040e-01 -6.75120533e-01 -1.00053036e+00
9.99906659e-01 3.94290164e-02 8.06062281e-01 1.69561654e-01
2.87785619e-01 -1.02740362e-01 -6.03888988e-01 -3.32801968e-01
-8.04474175e-01 4.85075206e-01 2.28448629e-01 -2.08394527e-01
4.75804955e-01 -5.38582623e-01 -2.08299235e-01 2.92579800e-01
-9.40215886e-01 2.36847043e-01 8.00584614e-01 8.26576769e-01
8.53725255e-01 1.03961356e-01 9.57128048e-01 -1.24144530e+00
6.73364162e-01 -3.48343462e-01 -3.32403392e-01 6.51780844e-01
-1.06306577e+00 2.85947770e-01 1.00300455e+00 -3.94524306e-01
-1.34073031e+00 2.45498508e-01 -1.33763552e-01 -4.04618055e-01
-3.22714835e-01 7.10535169e-01 -4.17843908e-01 2.32010767e-01
5.15665770e-01 4.71544474e-01 -1.04523681e-01 -4.33511019e-01
3.78452837e-01 1.12752974e+00 4.50633824e-01 -5.21948814e-01
6.12348974e-01 3.01324576e-02 -2.99136579e-01 -6.43546104e-01
-1.36607611e+00 -4.53814089e-01 -8.66823971e-01 -1.02251932e-01
7.28637397e-01 -8.08453619e-01 -7.90429935e-02 1.38998225e-01
-9.42015946e-01 1.27915442e-01 -4.02718246e-01 7.42694616e-01
-4.55050260e-01 9.89687979e-01 -4.70183790e-01 -8.89730990e-01
-5.15996218e-01 -8.39700639e-01 8.08682203e-01 5.09401917e-01
-1.59858048e-01 -8.54382038e-01 1.46983564e-01 6.12525403e-01
-1.83235668e-02 -9.76247415e-02 7.65528738e-01 -1.19109678e+00
-2.44750142e-01 -5.32486796e-01 -1.00951843e-01 8.24094236e-01
2.46829823e-01 1.33629084e-01 -9.02108610e-01 -2.58659422e-02
2.65112579e-01 -7.41626978e-01 9.98519182e-01 2.54350245e-01
1.05447352e+00 -4.90442574e-01 -4.42204773e-01 2.13430524e-01
1.06055856e+00 4.62617315e-02 4.01119769e-01 1.04338825e-01
6.93501532e-01 5.54137111e-01 1.12224889e+00 5.14011085e-01
2.76467890e-01 1.70854762e-01 -5.12832515e-02 8.20131153e-02
2.29492143e-01 -5.99961579e-01 2.10792974e-01 1.45479095e+00
-1.44140571e-01 -4.06224549e-01 -6.02044165e-01 3.84092331e-01
-2.10206485e+00 -9.11300957e-01 8.38702321e-02 2.00309777e+00
1.29008532e+00 2.16416910e-01 -7.44671077e-02 3.55648279e-01
9.76759017e-01 4.46687937e-02 -5.10717034e-01 1.09383082e-02
1.42761981e-02 -3.09150010e-01 1.17900886e-01 2.05761895e-01
-1.22817922e+00 8.49348903e-01 6.31088591e+00 1.35082054e+00
-1.01332247e+00 3.90663706e-02 8.11115205e-01 1.01840474e-01
-5.09193122e-01 1.90886468e-01 -9.70011771e-01 7.09157467e-01
8.15896571e-01 -2.04770476e-01 -3.09204366e-02 1.08137894e+00
8.24806839e-02 -1.50052518e-01 -1.07097781e+00 8.43563795e-01
3.40034574e-01 -1.10688877e+00 1.56249344e-01 -1.15228467e-01
1.07867777e+00 -4.31668043e-01 -2.80492336e-01 5.19853890e-01
8.27301368e-02 -6.60146892e-01 4.11568016e-01 3.24883461e-01
7.92366326e-01 -6.72043979e-01 1.00714421e+00 1.08618617e+00
-9.51843560e-01 1.26286924e-01 -6.14592373e-01 -2.18139589e-02
2.31535971e-01 9.79673803e-01 -8.52928460e-01 1.03105271e+00
1.85650527e-01 8.42522740e-01 -3.16409916e-01 8.32473814e-01
-5.05460918e-01 8.30632329e-01 -3.19886327e-01 -1.88707605e-01
8.03803932e-03 -5.26168883e-01 1.84647039e-01 1.08071649e+00
2.82959014e-01 2.53620833e-01 7.02706754e-01 7.07081676e-01
-2.76604176e-01 5.93733668e-01 -4.08644825e-01 -1.31208777e-01
6.09932244e-01 1.32223499e+00 -9.60907638e-01 -6.86464846e-01
-2.47630179e-01 8.07741404e-01 4.05922472e-01 9.20201167e-02
-7.67552853e-01 -2.43356273e-01 -4.63234752e-01 -1.69782430e-01
-4.92707677e-02 5.99366575e-02 -4.41735506e-01 -1.57534349e+00
6.23628646e-02 -6.09725535e-01 3.64044338e-01 -4.92083251e-01
-1.23387122e+00 4.51864749e-01 1.22439325e-01 -1.60541558e+00
-3.36137384e-01 6.86537754e-03 -5.46985626e-01 5.56326389e-01
-1.60419619e+00 -9.96324182e-01 -2.90149748e-01 2.47382179e-01
6.36522412e-01 -1.87887266e-01 8.19291890e-01 2.21776456e-01
-9.15910065e-01 5.52431881e-01 3.01285863e-01 -6.58107772e-02
8.64910007e-01 -1.04267049e+00 -2.91109741e-01 8.20833087e-01
1.56466678e-01 4.21204180e-01 5.59128046e-01 -8.36883366e-01
-6.58982456e-01 -1.20670319e+00 1.00687122e+00 -1.87374964e-01
1.38396174e-01 1.18006282e-01 -1.01014864e+00 5.52631974e-01
1.51519313e-01 -1.71880305e-01 1.05145824e+00 1.10676914e-01
-1.01312637e-01 6.60858080e-02 -1.08389831e+00 4.60483402e-01
7.70240188e-01 -3.79867852e-01 -9.52132761e-01 6.90339088e-01
7.18811214e-01 -4.49304208e-02 -7.61472166e-01 7.08878160e-01
1.53956577e-01 -7.96590388e-01 4.58347350e-01 -4.61678654e-01
6.35898054e-01 -3.57679933e-01 1.15263894e-01 -1.20985317e+00
6.72512501e-03 -2.08962336e-01 -1.70233414e-01 1.70342624e+00
6.18466377e-01 -3.74542713e-01 9.63324308e-01 5.91059685e-01
-2.42150560e-01 -7.82631218e-01 -5.65644383e-01 -8.24255645e-01
-3.42977121e-02 -7.19571635e-02 3.37722242e-01 1.23719537e+00
5.69455087e-01 5.52882731e-01 -5.34040987e-01 -8.02528635e-02
4.09748912e-01 3.88803810e-01 6.12132192e-01 -1.19410098e+00
-5.95623963e-02 -1.67378053e-01 -1.20422162e-01 -1.23764980e+00
4.44833040e-01 -8.92923951e-01 3.20565492e-01 -1.56307793e+00
6.56886101e-01 -4.87337291e-01 -1.03621081e-01 6.98906541e-01
-4.70691770e-01 1.91987872e-01 -1.41316012e-01 8.07831764e-01
-8.95110607e-01 9.23931897e-01 1.18721128e+00 -1.47296131e-01
-3.12298775e-01 9.81726199e-02 -6.98510826e-01 8.99955273e-01
9.10575330e-01 -5.46242476e-01 -6.52366102e-01 -5.36124073e-02
-3.30162868e-02 1.78739145e-01 -4.10274684e-01 -1.08087683e+00
2.74780303e-01 -3.64923805e-01 2.25638002e-01 -6.64034069e-01
3.47850993e-02 -7.01354980e-01 -8.61358047e-02 2.04362333e-01
-5.86109042e-01 -4.75563258e-01 -1.87753767e-01 7.69547820e-01
-5.19788384e-01 -7.35289693e-01 7.09760547e-01 -2.50769556e-01
-4.04857039e-01 7.13067725e-02 -2.59233028e-01 2.51628369e-01
1.09102738e+00 -2.73092777e-01 -1.42604992e-01 -2.61662692e-01
-6.42179787e-01 3.35096627e-01 4.67198133e-01 7.08175078e-03
4.16255027e-01 -1.14273489e+00 -6.87836766e-01 8.03482160e-02
1.68575700e-02 3.59436721e-01 3.04554075e-01 7.67386556e-01
-3.99528801e-01 2.76787937e-01 -3.36937122e-02 -5.61705172e-01
-1.06400955e+00 6.98957086e-01 -2.63472319e-01 -5.57882726e-01
-3.23724210e-01 5.72554708e-01 -1.62484735e-01 -5.20118177e-01
1.55154258e-01 -9.19389948e-02 -5.10337770e-01 1.38012990e-01
2.69293308e-01 2.45231912e-01 -3.22034140e-03 -4.22559559e-01
-3.37449387e-02 5.10982275e-01 -4.03813571e-01 -3.33752343e-03
8.96803975e-01 -2.00983316e-01 -3.69533151e-02 4.67289984e-01
9.86545324e-01 1.25641271e-01 -9.79911923e-01 -2.70924807e-01
6.76161349e-02 -1.62396207e-01 -1.70188546e-01 -5.49020052e-01
-8.10524166e-01 6.89590931e-01 2.83259843e-02 2.71696150e-01
1.07781243e+00 -7.46275857e-02 7.74353623e-01 6.75876975e-01
1.06507488e-01 -1.41554272e+00 8.67230520e-02 2.91930079e-01
2.96308339e-01 -1.45327973e+00 3.41406763e-01 -7.45242953e-01
-8.40656281e-01 9.67799783e-01 7.71249533e-01 2.30788857e-01
5.00452578e-01 4.24415581e-02 -4.80153337e-02 3.14867795e-01
-4.94607896e-01 -4.67794351e-02 2.80682087e-01 3.57532889e-01
4.44297224e-01 -3.72963399e-02 -8.28370452e-01 1.01447999e+00
1.36629725e-03 7.32450001e-03 5.22831142e-01 1.06232166e+00
-7.55955577e-01 -1.16498411e+00 -1.50218382e-01 7.23818719e-01
-3.89480352e-01 -5.33347018e-02 -4.61121291e-01 4.40808415e-01
2.94343308e-02 1.05523610e+00 -2.88872331e-01 -5.55564880e-01
-7.33430125e-03 3.15189391e-01 8.37533996e-02 -8.88216078e-01
-2.12210342e-01 3.44225138e-01 1.74878567e-01 1.13652766e-01
-9.16880906e-01 -4.95158851e-01 -1.35367048e+00 3.29711213e-04
-1.02458167e+00 7.49581099e-01 3.18089545e-01 1.20021105e+00
2.03414857e-01 2.03379825e-01 8.05739045e-01 -6.86006784e-01
-8.67558539e-01 -1.18719363e+00 -7.16295779e-01 4.02438074e-01
-1.30971089e-01 -4.14802253e-01 -5.71917117e-01 3.05488795e-01] | [9.554106712341309, 4.086165428161621] |
d6f5828e-ca71-4e1c-ae10-739a53d8f036 | improved-algorithms-for-multi-period-multi | 2301.13791 | null | https://arxiv.org/abs/2301.13791v2 | https://arxiv.org/pdf/2301.13791v2.pdf | Improved Algorithms for Multi-period Multi-class Packing Problems with Bandit Feedback | We consider the linear contextual multi-class multi-period packing problem (LMMP) where the goal is to pack items such that the total vector of consumption is below a given budget vector and the total value is as large as possible. We consider the setting where the reward and the consumption vector associated with each action is a class-dependent linear function of the context, and the decision-maker receives bandit feedback. LMMP includes linear contextual bandits with knapsacks and online revenue management as special cases. We establish a new estimator which guarantees a faster convergence rate, and consequently, a lower regret in such problems. We propose a bandit policy that is a closed-form function of said estimated parameters. When the contexts are non-degenerate, the regret of the proposed policy is sublinear in the context dimension, the number of classes, and the time horizon $T$ when the budget grows at least as $\sqrt{T}$. We also resolve an open problem posed by Agrawal & Devanur (2016) and extend the result to a multi-class setting. Our numerical experiments clearly demonstrate that the performance of our policy is superior to other benchmarks in the literature. | ['Assaf Zeevi', 'Garud Iyengar', 'Wonyoung Kim'] | 2023-01-31 | null | null | null | null | ['multi-armed-bandits'] | ['miscellaneous'] | [ 4.05477434e-02 1.29111171e-01 -9.57428098e-01 -1.07726946e-01
-8.03786397e-01 -8.06286871e-01 -4.08397853e-01 4.45581406e-01
-5.11736631e-01 1.11568558e+00 -1.63098559e-01 -4.37600046e-01
-6.92530215e-01 -7.65532672e-01 -1.32648432e+00 -1.02858555e+00
-2.17698440e-01 6.46755099e-01 -1.71390474e-01 4.78972128e-04
1.29490197e-01 1.52244285e-01 -1.00429499e+00 -1.65250450e-01
9.52695370e-01 1.78632224e+00 3.24088931e-01 6.24463856e-01
3.07402164e-01 3.82975042e-01 -4.47002470e-01 -4.47936147e-01
8.34852993e-01 -1.59330696e-01 -6.66156292e-01 5.43389857e-01
-2.85687409e-02 -3.86090994e-01 2.16858182e-03 1.15565670e+00
1.63516060e-01 3.56874794e-01 2.67122239e-01 -1.29369390e+00
-4.18902248e-01 8.70386183e-01 -9.88496780e-01 2.24241868e-01
-1.22867249e-01 -1.89649716e-01 1.24422657e+00 8.52953121e-02
2.13141724e-01 8.81677926e-01 1.30868703e-01 4.21636015e-01
-1.33822334e+00 -6.84463441e-01 7.01629162e-01 9.71208885e-02
-7.77150571e-01 2.28448845e-02 1.65684015e-01 -1.26828715e-01
6.66964233e-01 5.00280142e-01 8.69157135e-01 4.98194307e-01
1.25251502e-01 8.77648890e-01 8.45452607e-01 -4.10489827e-01
6.83869183e-01 3.39220688e-02 2.51557231e-01 1.78627133e-01
7.85410583e-01 -1.23404682e-01 -1.04002118e-01 -2.27619633e-01
4.54296529e-01 4.48736072e-01 -1.94345981e-01 -4.73669231e-01
-5.86137831e-01 1.09570742e+00 1.86202839e-01 -2.32706904e-01
-5.92217982e-01 2.96914637e-01 2.65182018e-01 3.93112510e-01
6.62099183e-01 3.24234009e-01 -4.88924950e-01 -8.39465111e-02
-7.87992179e-01 4.38054532e-01 9.50365186e-01 1.45570683e+00
4.04456437e-01 -4.34875876e-01 -3.72869909e-01 7.24749863e-01
-3.26932877e-01 7.60674596e-01 -7.69453794e-02 -8.80034924e-01
1.23085463e+00 1.68183059e-01 9.22759593e-01 -6.25618577e-01
-2.37671524e-01 -7.88670361e-01 -5.44376433e-01 -4.74806786e-01
4.78221148e-01 -3.12194049e-01 -5.46074510e-01 1.76120603e+00
2.85202444e-01 -3.21191162e-01 -1.72817811e-01 1.03470814e+00
-3.69626284e-01 9.62584019e-01 -2.62313873e-01 -9.44116950e-01
1.26546288e+00 -1.06908309e+00 -7.24059284e-01 -5.48477292e-01
3.16921979e-01 -5.85723221e-01 6.17466152e-01 5.89452863e-01
-1.28800988e+00 3.21963996e-01 -1.14558566e+00 5.56992769e-01
7.04303160e-02 -1.42081931e-01 6.80781662e-01 7.88470745e-01
-2.93352425e-01 3.84937614e-01 -6.02760851e-01 6.84048906e-02
4.18245614e-01 4.41763341e-01 1.74747825e-01 -4.37832505e-01
-6.53907418e-01 4.99394238e-01 7.24817991e-01 1.07658885e-01
-5.44791877e-01 -6.02864981e-01 -5.15001297e-01 2.82985300e-01
1.13621426e+00 -3.77793461e-01 1.49899626e+00 -7.92904079e-01
-1.12992465e+00 1.86479986e-01 2.37599507e-01 -7.63480902e-01
6.33480072e-01 -1.65024340e-01 1.67074829e-01 8.41654763e-02
2.39047438e-01 -1.77323222e-01 4.99339283e-01 -7.39708185e-01
-1.11391246e+00 -5.50304055e-01 6.13995194e-01 3.27392220e-01
-3.29172701e-01 -1.96451232e-01 -1.55719444e-01 -3.78629386e-01
-1.97758526e-01 -1.11910510e+00 -5.42819262e-01 -4.87053037e-01
-2.31811836e-01 2.15959046e-02 9.20192823e-02 -3.87885630e-01
1.14753938e+00 -2.01207042e+00 1.02245986e-01 1.57255009e-01
-5.30542612e-01 -1.11198209e-01 4.36386541e-02 5.62383950e-01
2.68714249e-01 1.93298742e-01 -6.61265999e-02 -1.93539500e-01
1.79040402e-01 5.56443334e-01 -3.37812960e-01 7.59726644e-01
-3.91161293e-01 4.54254270e-01 -7.34264195e-01 6.20419101e-04
-1.70592323e-01 -6.02678776e-01 -8.23625863e-01 1.84509337e-01
-5.53071141e-01 -2.67196119e-01 -6.67828560e-01 6.44929409e-01
9.98577476e-01 -3.94117326e-01 2.77384043e-01 3.40096354e-01
-6.75479025e-02 5.32974638e-02 -1.42028654e+00 1.20018768e+00
-6.04905486e-01 1.31392442e-02 3.59625429e-01 -1.42880058e+00
2.65225947e-01 -1.05394714e-01 5.77398658e-01 -5.84797382e-01
4.89204973e-02 7.49444589e-02 -3.02301973e-01 -3.41220379e-01
5.09505153e-01 -3.58848631e-01 -3.01090807e-01 3.52352440e-01
-3.61013710e-01 2.74710625e-01 5.45604706e-01 3.99758928e-02
9.16493118e-01 -4.60738391e-01 5.70836067e-01 -2.70108670e-01
-1.99530587e-01 -1.34022802e-01 8.98886681e-01 9.60765302e-01
5.03903665e-02 -3.07044778e-02 9.79007483e-01 -1.40142050e-02
-1.09141922e+00 -6.31200850e-01 -7.65127316e-02 1.33174658e+00
5.71348190e-01 6.05140142e-02 -3.66359890e-01 -3.88435155e-01
6.26021147e-01 6.91223502e-01 -8.09617043e-01 2.08945513e-01
-1.31169945e-01 -1.05847800e+00 -5.87081134e-01 4.41182822e-01
3.09563518e-01 -4.66918409e-01 -7.56878078e-01 4.93747383e-01
-1.46508947e-01 -1.30806422e+00 -9.18912530e-01 4.67597991e-01
-7.55420208e-01 -8.93345237e-01 -7.58458376e-01 -4.96741205e-01
6.76777661e-01 4.61676478e-01 5.41919827e-01 -6.33487761e-01
-9.70398169e-03 2.22240344e-01 -4.33844537e-01 -5.63577890e-01
3.36130649e-01 6.40746802e-02 -2.90534258e-01 1.12562604e-01
-2.59624779e-01 -8.98526385e-02 -8.35367739e-01 4.11404878e-01
-1.01386547e+00 1.58401802e-01 4.93878543e-01 9.09945428e-01
8.27610672e-01 1.52232766e-01 7.21826792e-01 -9.05733705e-01
5.38147807e-01 -7.34921813e-01 -1.13625181e+00 4.70290899e-01
-5.75451136e-01 4.74762619e-02 6.91340923e-01 -7.58954525e-01
-6.70593083e-01 -1.38320148e-01 4.42150176e-01 -1.93659082e-01
5.36577046e-01 4.12991196e-01 -1.74848765e-01 9.85652432e-02
-1.06941685e-01 1.92385152e-01 -2.35840812e-01 -3.48532408e-01
6.62232935e-02 4.97748554e-01 1.01364218e-01 -9.24145520e-01
2.78352350e-01 1.48514897e-01 6.98646232e-02 -3.60333920e-01
-1.29490542e+00 -5.65839767e-01 2.94967771e-01 -5.82125559e-02
2.19413355e-01 -7.40242541e-01 -1.46570003e+00 -1.39293879e-01
-7.94296563e-01 -3.80631775e-01 -4.41708803e-01 5.03260016e-01
-1.06581402e+00 1.33019850e-01 -4.35512036e-01 -1.57381463e+00
-2.84355134e-01 -8.89838874e-01 5.49672306e-01 3.59577924e-01
6.65637255e-01 -3.87370884e-01 -2.21278176e-01 6.18818820e-01
2.28116065e-01 3.34966958e-01 8.44975471e-01 -4.33922797e-01
-7.37194061e-01 -2.17158660e-01 -1.68804348e-01 2.97550231e-01
-1.00989260e-01 -6.68928027e-01 7.33619556e-02 -6.67936325e-01
3.72802317e-02 -2.39028722e-01 5.30690193e-01 7.24140584e-01
1.49063075e+00 -1.34706187e+00 -2.47874603e-01 4.98712391e-01
1.76948619e+00 6.56211674e-01 7.08016828e-02 5.90754151e-01
2.83797625e-02 3.52292567e-01 1.28658164e+00 1.01947129e+00
2.87027985e-01 7.15550900e-01 8.81474018e-01 3.59165281e-01
1.05778253e+00 -2.86280718e-02 9.41771828e-03 2.13105053e-01
4.36831675e-02 -6.04716063e-01 -3.57131571e-01 7.62951076e-01
-2.35099745e+00 -9.68969762e-01 4.45695221e-01 2.71400237e+00
8.90637577e-01 1.99524924e-01 7.19064832e-01 -1.03855692e-02
8.81646037e-01 -1.53559893e-01 -1.09593546e+00 -8.81765246e-01
-4.34641987e-02 1.94538967e-03 1.60369301e+00 1.77188456e-01
-9.44487512e-01 2.78400540e-01 5.35196829e+00 9.05680895e-01
-6.87344253e-01 1.71100572e-01 9.53149140e-01 -9.58584130e-01
-1.04615660e-02 -8.62179995e-02 -9.02670443e-01 9.21676874e-01
9.51118588e-01 -4.42973793e-01 9.78928924e-01 9.79475498e-01
2.05822840e-01 -4.58961368e-01 -1.01251626e+00 8.55495870e-01
-3.49304706e-01 -1.31798995e+00 -7.68970251e-01 6.53952658e-01
8.68414164e-01 -2.23307982e-01 1.49964035e-01 1.40791088e-01
2.97037691e-01 -5.31851947e-01 8.25237691e-01 1.55688390e-01
5.17666459e-01 -1.48298478e+00 7.15004504e-01 6.17647946e-01
-9.70551193e-01 -9.24760759e-01 -5.97979248e-01 -1.03110015e-01
2.29052454e-01 5.95503569e-01 -4.28042620e-01 5.18930078e-01
5.81178129e-01 2.25982264e-01 4.72029060e-01 1.57276523e+00
3.23424846e-01 5.38226306e-01 -5.63855529e-01 -4.79461759e-01
4.53888923e-01 -5.42680264e-01 2.13668615e-01 9.98608232e-01
1.79932222e-01 4.70481157e-01 6.24061763e-01 1.99842289e-01
-1.49880901e-01 3.76380920e-01 -1.00787682e-02 -2.52158999e-01
5.36750138e-01 9.21739578e-01 -7.76124954e-01 -1.17381968e-01
-2.55733907e-01 5.44049323e-01 3.46026838e-01 2.47432649e-01
-8.99552524e-01 -2.63825655e-01 8.29650104e-01 -2.89414860e-02
9.30408359e-01 6.71152025e-02 -3.18956047e-01 -8.51210713e-01
5.38000762e-01 -3.57400298e-01 6.43394172e-01 -7.61245713e-02
-1.28618705e+00 -1.79615155e-01 2.19303802e-01 -1.01897812e+00
-1.53844103e-01 -4.46171522e-01 -1.23246655e-01 4.74135101e-01
-1.49882627e+00 -4.86278504e-01 3.99423599e-01 2.00128973e-01
4.44179118e-01 2.83353478e-01 3.03726017e-01 3.49029154e-01
-8.12210262e-01 6.62097633e-01 8.09447765e-01 -4.06999558e-01
-4.62080911e-02 -1.14725399e+00 -4.42274690e-01 7.42494643e-01
-6.68672681e-01 1.97246611e-01 1.04284525e+00 -2.95201570e-01
-1.73015499e+00 -1.03482342e+00 1.64796233e-01 5.36120772e-01
8.75562906e-01 -5.12070358e-01 -2.52881557e-01 6.86783731e-01
-6.70911372e-02 7.21677169e-02 7.40473270e-01 7.64158666e-02
-1.45865781e-02 -6.44445598e-01 -1.36891663e+00 1.70674950e-01
9.68366086e-01 3.00354034e-01 2.84924451e-02 6.58219814e-01
8.97623062e-01 -6.28562987e-01 -1.01935112e+00 1.24414548e-01
6.38305545e-01 -5.30132830e-01 5.55131197e-01 -8.87568116e-01
2.04226345e-01 4.14877474e-01 -4.62301552e-01 -1.25450778e+00
-2.79719591e-01 -7.95768559e-01 -3.07166785e-01 8.73183310e-01
4.64720726e-01 -7.94637024e-01 8.15639555e-01 8.22111487e-01
1.26514658e-01 -1.14836562e+00 -1.32660139e+00 -1.15781260e+00
1.31034732e-01 -4.40382659e-02 6.14309847e-01 3.35725129e-01
4.13283050e-01 -1.02370251e-02 -7.51180530e-01 1.20162025e-01
7.28741288e-01 7.71782041e-01 3.86969537e-01 -6.93951905e-01
-8.13799202e-01 -2.41157457e-01 6.63076043e-02 -1.30786383e+00
-1.18131973e-01 -4.88087088e-01 1.50648892e-01 -1.34869766e+00
7.00998843e-01 -7.20143378e-01 -4.50520635e-01 4.35555607e-01
2.29330167e-01 -5.07570922e-01 5.48344612e-01 -1.34194404e-01
-9.78739917e-01 3.94785434e-01 1.38965094e+00 -1.39951259e-01
-2.11943567e-01 7.80772507e-01 -1.15643644e+00 -6.08138219e-02
7.23044813e-01 -5.52684009e-01 -3.42365116e-01 -1.93678290e-01
4.27171111e-01 9.04665530e-01 -2.43979976e-01 -2.60740787e-01
-9.69986245e-03 -1.04146218e+00 -1.29007511e-02 -8.61366451e-01
3.75598252e-01 -1.10068774e+00 6.04518466e-02 5.71096539e-01
-5.44548929e-01 -5.31865433e-02 2.86959633e-02 1.09064579e+00
4.47869092e-01 -4.22394931e-01 6.77723169e-01 9.48567837e-02
1.56798303e-01 6.40129626e-01 -6.13836609e-02 2.26067469e-01
1.44430554e+00 -1.00388467e-01 -4.69783396e-01 -4.05682057e-01
-4.41531777e-01 8.50973248e-01 1.99020788e-01 1.41512781e-01
7.98344836e-02 -1.13772690e+00 -3.38944793e-01 -3.22763443e-01
1.18917099e-03 3.04034408e-02 3.94757211e-01 9.13902223e-01
3.98651399e-02 7.43005931e-01 -1.04573697e-01 -3.35967451e-01
-9.68618810e-01 8.41937184e-01 1.14179209e-01 -5.49271166e-01
-2.13154644e-01 6.42718732e-01 7.94094801e-02 4.03636098e-01
4.34469938e-01 -5.29575169e-01 3.61971736e-01 1.34614632e-01
4.96827513e-01 4.55017716e-01 1.47044435e-01 3.65201533e-02
-1.65013537e-01 -1.67804689e-03 -2.93720394e-01 -4.08155508e-02
1.77763307e+00 -2.10547835e-01 1.80952162e-01 6.96468502e-02
1.18541956e+00 -2.32263312e-01 -1.44336987e+00 -4.21032041e-01
-1.97379470e-01 -7.88922966e-01 -9.73494723e-02 -8.60291541e-01
-1.17107320e+00 1.42376006e-01 3.45513225e-01 7.60503173e-01
1.38369429e+00 3.48885916e-02 6.61787033e-01 1.87989205e-01
7.77161777e-01 -1.45319033e+00 -2.95319855e-01 1.26408428e-01
6.67438090e-01 -8.83777261e-01 1.59230903e-01 -2.57884294e-01
-6.04166269e-01 7.96354949e-01 4.37756151e-01 -2.64329821e-01
3.69220048e-01 6.73710108e-02 -9.97708023e-01 5.01549423e-01
-1.00863302e+00 -2.08033532e-01 4.32312116e-02 1.35307563e-02
-1.22653164e-01 8.72564971e-01 -9.20888782e-01 9.64308858e-01
-2.57841706e-01 -1.61963746e-01 7.00471699e-01 1.11866462e+00
-6.89910889e-01 -8.52789164e-01 -4.79858160e-01 8.53567541e-01
-8.02671015e-01 2.87493467e-01 2.31535047e-01 5.42717278e-01
-1.17770568e-01 1.08986294e+00 2.93706119e-01 1.96401939e-01
3.41237009e-01 -4.80414301e-01 8.47016037e-01 -5.66380739e-01
-1.80893525e-01 3.23072463e-01 1.75304294e-01 -4.07395273e-01
-2.62269229e-01 -6.57026529e-01 -7.07429111e-01 -4.75044101e-01
-6.98292732e-01 4.61117089e-01 8.44834328e-01 1.14575946e+00
-7.86216557e-03 1.74808890e-01 1.15701199e+00 -4.18587089e-01
-9.64422762e-01 -7.23540545e-01 -9.90786254e-01 -3.07043046e-02
5.09754241e-01 -5.96654058e-01 -1.91412166e-01 -6.09833241e-01] | [4.570013046264648, 3.3292136192321777] |
4b913be5-b635-440a-8999-751d17d0c1da | explicit-visual-prompting-for-universal | 2305.18476 | null | https://arxiv.org/abs/2305.18476v1 | https://arxiv.org/pdf/2305.18476v1.pdf | Explicit Visual Prompting for Universal Foreground Segmentations | Foreground segmentation is a fundamental problem in computer vision, which includes salient object detection, forgery detection, defocus blur detection, shadow detection, and camouflage object detection. Previous works have typically relied on domain-specific solutions to address accuracy and robustness issues in those applications. In this paper, we present a unified framework for a number of foreground segmentation tasks without any task-specific designs. We take inspiration from the widely-used pre-training and then prompt tuning protocols in NLP and propose a new visual prompting model, named Explicit Visual Prompting (EVP). Different from the previous visual prompting which is typically a dataset-level implicit embedding, our key insight is to enforce the tunable parameters focusing on the explicit visual content from each individual image, i.e., the features from frozen patch embeddings and high-frequency components. Our method freezes a pre-trained model and then learns task-specific knowledge using a few extra parameters. Despite introducing only a small number of tunable parameters, EVP achieves superior performance than full fine-tuning and other parameter-efficient fine-tuning methods. Experiments in fourteen datasets across five tasks show the proposed method outperforms other task-specific methods while being considerably simple. The proposed method demonstrates the scalability in different architectures, pre-trained weights, and tasks. The code is available at: https://github.com/NiFangBaAGe/Explicit-Visual-Prompt. | ['Xiaodong Cun', 'Chi-Man Pun', 'Xi Shen', 'Weihuang Liu'] | 2023-05-29 | null | null | null | null | ['image-manipulation-detection', 'defocus-blur-detection', 'visual-prompting', 'camouflaged-object-segmentation', 'foreground-segmentation', 'shadow-detection', 'salient-object-detection-1'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [ 4.03496355e-01 -2.09639043e-01 -2.42403038e-02 -4.60039526e-01
-7.33770788e-01 -6.57864034e-01 6.56143725e-01 -2.58025080e-01
-4.56275046e-01 5.67486882e-01 -1.95657797e-02 -2.12711677e-01
6.54059574e-02 -2.34910652e-01 -8.89642060e-01 -1.07266283e+00
4.07545358e-01 -5.89919537e-02 7.91294754e-01 2.96463907e-01
4.69683498e-01 3.16938162e-01 -1.30270457e+00 8.86576027e-02
9.45559561e-01 9.72275436e-01 6.66788161e-01 8.47521722e-01
-1.14829339e-01 4.90832180e-01 -6.17399096e-01 -1.29735723e-01
1.48875564e-01 -4.52669673e-02 -6.58168256e-01 2.23662749e-01
6.51087046e-01 -4.60429251e-01 -1.32457227e-01 1.01016653e+00
6.33009195e-01 1.62769705e-01 5.16932249e-01 -1.35600233e+00
-1.08830035e+00 1.08215623e-01 -7.07077384e-01 4.53016937e-01
-1.50646254e-01 5.88934720e-01 7.17433333e-01 -9.35579062e-01
5.24968266e-01 1.03972542e+00 6.22457683e-01 8.79371524e-01
-1.26128793e+00 -7.05319762e-01 5.74833035e-01 2.34610721e-01
-1.14383626e+00 -3.55408669e-01 8.54859769e-01 -6.97343171e-01
7.47338474e-01 1.76037091e-03 4.06128049e-01 1.29745126e+00
2.89583043e-03 8.62475514e-01 1.00752389e+00 -1.97811604e-01
2.40203798e-01 1.29387066e-01 3.50320905e-01 6.80538058e-01
3.39190394e-01 1.94829807e-01 -5.21779716e-01 -2.25550294e-01
8.95383537e-01 -3.65074985e-02 -6.39157891e-01 -5.01375198e-01
-1.27407563e+00 7.70946145e-01 4.29710269e-01 1.23610981e-01
1.97387603e-03 2.92541325e-01 3.35818559e-01 -1.22529790e-01
3.64660889e-01 3.49675685e-01 -5.78086436e-01 4.72772494e-02
-8.46876562e-01 1.04450807e-01 5.44390082e-01 9.19500709e-01
8.41397583e-01 2.24942669e-01 -5.86767197e-01 7.26332128e-01
3.03545892e-01 5.40669084e-01 4.24786359e-01 -1.02866149e+00
1.09450832e-01 3.51691097e-01 3.27295423e-01 -8.21601331e-01
-3.75639886e-01 -4.66499537e-01 -5.67281127e-01 2.30445683e-01
4.90837842e-01 -2.80813396e-01 -1.39363813e+00 1.82428527e+00
6.27215624e-01 7.34038591e-01 -2.03987613e-01 1.29633713e+00
1.05652010e+00 6.79500520e-01 1.24016099e-01 4.47314866e-02
1.36402380e+00 -1.28535378e+00 -6.77710295e-01 -4.09075379e-01
1.49687275e-01 -9.34887588e-01 1.35786414e+00 2.60784894e-01
-7.03267038e-01 -6.17885828e-01 -1.04868484e+00 -3.90776753e-01
-3.89974624e-01 3.47938687e-01 7.61888802e-01 6.39770269e-01
-1.09483945e+00 2.40932569e-01 -5.69901645e-01 -3.12167376e-01
6.06326163e-01 2.48477101e-01 -2.33947206e-02 6.14710897e-02
-8.90963435e-01 5.83873689e-01 2.67786771e-01 1.97401956e-01
-1.18648481e+00 -1.08942151e+00 -7.32953608e-01 -8.53317603e-02
6.67876899e-01 -8.36521566e-01 1.26591086e+00 -8.66024911e-01
-1.58386970e+00 1.00834203e+00 -2.76290417e-01 -2.30217636e-01
5.44155777e-01 -4.44452524e-01 -5.64163402e-02 2.47409999e-01
-1.86063834e-02 8.79288554e-01 1.51220322e+00 -1.38332498e+00
-6.04278326e-01 1.43030047e-01 1.31748542e-01 6.08761329e-03
-2.76720226e-01 5.98363690e-02 -7.86724031e-01 -7.70262301e-01
-3.04901809e-01 -7.89991856e-01 -7.65904486e-02 3.95518422e-01
-3.40641350e-01 -1.80743277e-01 1.15292907e+00 -4.44538176e-01
7.97064841e-01 -2.51192307e+00 -1.24619879e-01 -3.70145380e-01
5.13081372e-01 6.63690865e-01 -3.21588725e-01 -7.63187408e-02
2.00008422e-01 -3.57425511e-02 -4.04268086e-01 -2.82853693e-01
7.94120654e-02 3.01174581e-01 -4.41325784e-01 5.63409925e-01
5.10369837e-01 1.05354881e+00 -1.00487399e+00 -5.47286928e-01
4.79649961e-01 7.36727595e-01 -3.95580709e-01 3.10009629e-01
-4.63869274e-01 4.65436637e-01 -5.42463303e-01 7.28549778e-01
9.68833506e-01 -6.36411071e-01 -1.45589560e-01 -3.94218117e-01
-7.87672102e-02 -1.20918546e-02 -1.03561866e+00 1.61472476e+00
-7.02264905e-02 7.89905071e-01 4.08430189e-01 -9.88577008e-01
6.77074373e-01 1.18058898e-01 1.34285256e-01 -5.50511718e-01
2.41458923e-01 1.03363141e-01 -3.40367883e-01 -6.77736998e-01
4.17572916e-01 2.68206596e-01 2.35388115e-01 2.70113230e-01
2.17575803e-01 -7.98477158e-02 2.84653846e-02 2.68086642e-01
1.00786638e+00 2.62015522e-01 1.84079200e-01 -2.36066297e-01
4.33540642e-01 -8.58126208e-02 8.08442116e-01 8.66262972e-01
-6.30449772e-01 8.33656073e-01 2.85980970e-01 -3.44341636e-01
-7.64495850e-01 -1.02344728e+00 -2.25959152e-01 1.36936557e+00
4.77729678e-01 -1.64274588e-01 -8.45726252e-01 -6.88458264e-01
2.04466432e-01 4.06964958e-01 -8.85087550e-01 4.97192191e-03
-6.25361383e-01 -6.52884424e-01 3.25889587e-01 4.83747393e-01
6.58711374e-01 -1.16459835e+00 -7.58036911e-01 -6.49731308e-02
-2.05733970e-01 -1.39663422e+00 -8.61107945e-01 2.54862040e-01
-6.63835704e-01 -1.12011218e+00 -8.74002099e-01 -7.99552381e-01
6.86722517e-01 6.82119370e-01 9.35743749e-01 1.71828642e-01
-4.74880785e-01 5.62087595e-01 -2.41932601e-01 -3.62835616e-01
1.76562399e-01 -9.23272036e-03 -2.67207950e-01 1.89044863e-01
2.31077686e-01 -3.18307787e-01 -8.21467638e-01 2.82908946e-01
-9.93146718e-01 -1.04265455e-02 7.25477993e-01 9.29386497e-01
6.47949517e-01 -4.32143390e-01 3.64579260e-01 -1.07314885e+00
4.77413058e-01 -2.02929080e-01 -8.19212854e-01 2.62475461e-01
-3.70360374e-01 -5.75891417e-03 4.65742439e-01 -9.34079409e-01
-1.23995030e+00 4.60319407e-03 2.04998851e-01 -7.70280600e-01
-2.97837347e-01 -9.16295871e-02 -2.06742555e-01 -4.50911731e-01
6.69214129e-01 2.71583915e-01 -3.93925488e-01 -4.51459855e-01
5.39646745e-01 2.93381691e-01 7.62115300e-01 -6.64405107e-01
9.96451199e-01 6.15496397e-01 -3.08779150e-01 -8.91068339e-01
-9.59293723e-01 -5.93163490e-01 -3.90915543e-01 -1.68136984e-01
8.60214651e-01 -8.33530068e-01 -7.68091381e-01 6.54017210e-01
-1.28012848e+00 -7.86331892e-01 -1.48940802e-01 1.76511660e-01
-3.60434920e-01 4.84517246e-01 -4.02466089e-01 -6.61615133e-01
-3.62321526e-01 -1.12317002e+00 1.24509823e+00 4.56931531e-01
1.12805352e-01 -9.17950571e-01 -1.48556545e-01 3.33884120e-01
5.61060786e-01 2.82914996e-01 5.69182396e-01 -3.37799281e-01
-1.06110263e+00 4.34473723e-01 -7.16752172e-01 3.58247161e-01
1.69959903e-01 6.11561351e-02 -1.30900991e+00 -2.94945031e-01
8.79949555e-02 -3.64469141e-01 1.16255653e+00 7.04840064e-01
1.28048754e+00 -1.27479285e-01 -5.03507435e-01 1.12593973e+00
1.34796977e+00 7.16256574e-02 3.99753034e-01 3.21405292e-01
8.91357958e-01 3.51060182e-01 6.14657879e-01 2.29292825e-01
2.38805041e-01 4.63035911e-01 4.76654798e-01 -2.24476099e-01
-3.72066915e-01 4.43956219e-02 3.43035430e-01 1.88775018e-01
2.14415267e-01 -1.86045438e-01 -6.17773175e-01 6.93627775e-01
-1.95922852e+00 -6.76618278e-01 -9.07904729e-02 1.81037593e+00
9.84719038e-01 6.36978447e-02 -1.32668465e-01 -2.47979149e-01
7.26960063e-01 4.45310473e-01 -1.05863535e+00 -2.44524032e-01
-5.45367040e-02 4.27314691e-04 5.59774578e-01 6.13639712e-01
-1.32352829e+00 1.23812914e+00 6.09348869e+00 8.42073619e-01
-1.36612654e+00 4.22076017e-01 3.31151217e-01 -3.04966658e-01
-3.19946736e-01 -1.11841761e-01 -9.92972195e-01 6.68699801e-01
3.87623519e-01 1.37492448e-01 4.25044209e-01 7.44321585e-01
1.29309192e-01 -1.63927644e-01 -9.61239517e-01 1.07933629e+00
8.10583830e-02 -1.49034631e+00 -1.28494665e-01 -2.45583996e-01
6.51899934e-01 1.49436504e-01 2.75527060e-01 1.12882510e-01
1.27442136e-01 -7.39857733e-01 8.04369688e-01 2.65054107e-01
7.76372850e-01 -2.13380873e-01 3.39677662e-01 1.09740652e-01
-1.05047560e+00 -5.21659814e-02 -4.66352433e-01 1.79207847e-01
1.42872781e-01 8.04308951e-01 -6.41947150e-01 1.58444315e-01
9.69109535e-01 6.28499389e-01 -5.50389528e-01 1.17031586e+00
-4.49819475e-01 8.40012431e-01 -1.93577364e-01 1.36340717e-02
2.74845123e-01 3.56191844e-02 5.65533578e-01 1.54459965e+00
-1.99545156e-02 9.51964036e-02 1.09700203e-01 9.96781647e-01
3.86252790e-03 -3.03204089e-01 -2.64310718e-01 1.65719882e-01
4.15886015e-01 1.51998544e+00 -6.89210474e-01 -2.57521272e-01
-4.97184545e-01 1.01106107e+00 3.35172176e-01 7.67505467e-01
-1.09070706e+00 -4.51856315e-01 7.81225681e-01 -4.56692837e-02
7.89218962e-01 -1.96152017e-01 -2.87553728e-01 -1.15882492e+00
1.86496019e-01 -6.88759267e-01 2.08626479e-01 -9.37787414e-01
-1.24003696e+00 3.28514546e-01 -4.33060080e-02 -8.48913729e-01
1.67232037e-01 -8.82002294e-01 -7.30476201e-01 7.48839915e-01
-2.08987045e+00 -1.26453948e+00 -5.78529656e-01 7.90125310e-01
5.48902571e-01 1.47047564e-01 3.84596616e-01 1.54912412e-01
-6.78513885e-01 6.28860116e-01 -1.75611172e-02 2.66576465e-02
1.05089486e+00 -1.38112652e+00 3.82196546e-01 1.03321588e+00
-1.65542006e-01 6.24303222e-01 6.99819267e-01 -5.37901163e-01
-1.17777932e+00 -1.08922756e+00 5.41050315e-01 -4.27969843e-01
6.72814965e-01 -6.09568238e-01 -1.17361915e+00 5.63084543e-01
4.26542521e-01 3.61551166e-01 4.03086066e-01 -1.43158749e-01
-4.97285515e-01 -1.76928222e-01 -1.05412245e+00 5.05272210e-01
1.18645680e+00 -4.16304588e-01 -4.54424709e-01 4.02311623e-01
1.09662580e+00 -6.04374826e-01 -2.88114816e-01 2.78287500e-01
3.00572991e-01 -8.00681114e-01 1.09237909e+00 -2.51190811e-01
3.57146151e-02 -5.89850128e-01 3.46474312e-02 -1.02043879e+00
-5.15383899e-01 -9.02609408e-01 -2.80329347e-01 1.32200158e+00
1.31118789e-01 -8.40624869e-01 6.75050020e-01 5.94145179e-01
-3.13121259e-01 -6.97182357e-01 -7.01688766e-01 -6.12145841e-01
-1.43210486e-01 -2.72113949e-01 3.65656495e-01 8.81138206e-01
-7.25299478e-01 2.74578005e-01 -4.55912083e-01 5.47765851e-01
8.02549243e-01 3.68252128e-01 7.96450257e-01 -1.03860295e+00
-2.70943373e-01 -3.85091037e-01 -4.99567278e-02 -1.26103795e+00
-5.53606115e-02 -5.71724772e-01 3.05671215e-01 -1.41036999e+00
3.12318683e-01 -2.97878653e-01 -4.16405410e-01 7.26760864e-01
-5.75362444e-01 1.84979424e-01 2.02365488e-01 1.55168280e-01
-8.26528668e-01 4.11996245e-01 1.47163868e+00 -1.80908650e-01
-9.49986652e-02 -1.98931873e-01 -8.55043471e-01 6.95582747e-01
8.98846745e-01 -4.67328697e-01 -5.46352208e-01 -7.98045039e-01
-9.71128792e-02 -5.51372468e-01 8.40910375e-01 -6.80152893e-01
3.66815269e-01 -3.53524685e-01 4.74772304e-01 -4.27365750e-01
3.67905498e-01 -6.59083128e-01 -2.22569555e-01 2.04887018e-01
-8.24597478e-02 -4.23677564e-01 4.32701141e-01 7.22530842e-01
1.02759130e-01 -2.38627866e-01 9.65515256e-01 2.18566004e-02
-1.18289053e+00 3.33101541e-01 -2.19989836e-01 3.78531605e-01
1.02919877e+00 -3.19374323e-01 -8.31315875e-01 2.16126181e-02
-5.34928083e-01 3.43549609e-01 4.61795747e-01 4.74384516e-01
7.83019722e-01 -1.03034079e+00 -4.39472467e-01 1.96625844e-01
1.25977397e-01 1.71088353e-01 3.92017633e-01 8.13722610e-01
-1.70973614e-01 2.79064685e-01 -1.10164523e-01 -8.92051816e-01
-1.19862747e+00 6.05046630e-01 3.04205656e-01 2.85212640e-02
-7.35831201e-01 1.19281709e+00 7.82431304e-01 -2.37780586e-01
4.80360359e-01 -6.08916938e-01 1.21287048e-01 -2.25529581e-01
5.93420148e-01 7.64748976e-02 -2.37513945e-01 -2.01733992e-01
-5.67569077e-01 9.20710087e-01 -2.49352098e-01 2.33460993e-01
1.21741319e+00 -1.01214409e-01 2.36229986e-01 2.82418013e-01
1.03656256e+00 -1.08900577e-01 -2.05490828e+00 -4.64445710e-01
-2.59877026e-01 -6.71864390e-01 2.14170098e-01 -8.58868301e-01
-1.29638612e+00 1.04257333e+00 6.49314046e-01 -6.37141988e-02
1.26890695e+00 8.17092508e-02 7.81821847e-01 2.92298764e-01
1.28888458e-01 -9.77006674e-01 3.93579036e-01 2.87083566e-01
8.00698400e-01 -1.40778732e+00 -7.33214989e-02 -4.95611221e-01
-7.28430927e-01 6.41579390e-01 7.96750188e-01 -6.14211075e-02
6.19181216e-01 3.81804317e-01 3.39086026e-01 -1.04795314e-01
-7.00777948e-01 -3.25380445e-01 3.64250571e-01 9.51827586e-01
9.94250551e-02 -1.35770380e-01 6.11019582e-02 6.12260580e-01
3.39535862e-01 9.71190706e-02 2.10424498e-01 8.93760681e-01
-6.17806077e-01 -8.74733984e-01 -3.24192166e-01 2.73640811e-01
-3.64908069e-01 -9.04432163e-02 -4.14710134e-01 7.14983702e-01
3.46024185e-01 8.99170578e-01 -1.43218473e-01 -5.07357232e-02
1.28474087e-01 -1.95651874e-01 4.82905567e-01 -4.67003137e-01
-5.71626306e-01 3.80961150e-02 -3.53134662e-01 -7.70115137e-01
-5.40219426e-01 -4.61326271e-01 -1.18411660e+00 -1.72592387e-01
-3.29770923e-01 -1.92440554e-01 4.18315709e-01 7.97746480e-01
5.75043797e-01 6.69386327e-01 1.63243860e-01 -1.18537009e+00
-2.70367354e-01 -7.64161706e-01 -3.27706695e-01 4.61109936e-01
8.52084517e-01 -9.93387282e-01 -4.37181860e-01 2.71602511e-01] | [9.729228019714355, 0.24048329889774323] |
b19adaba-4570-4885-9a6e-9c21a7469531 | scaling-up-the-self-optimization-model-by | 2211.01698 | null | https://arxiv.org/abs/2211.01698v1 | https://arxiv.org/pdf/2211.01698v1.pdf | Scaling up the self-optimization model by means of on-the-fly computation of weights | The Self-Optimization (SO) model is a useful computational model for investigating self-organization in "soft" Artificial life (ALife) as it has been shown to be general enough to model various complex adaptive systems. So far, existing work has been done on relatively small network sizes, precluding the investigation of novel phenomena that might emerge from the complexity arising from large numbers of nodes interacting in interconnected networks. This work introduces a novel implementation of the SO model that scales as $\mathcal{O}\left(N^{2}\right)$ with respect to the number of nodes $N$, and demonstrates the applicability of the SO model to networks with system sizes several orders of magnitude higher than previously was investigated. Removing the prohibitive computational cost of the naive $\mathcal{O}\left(N^{3}\right)$ algorithm, our on-the-fly computation paves the way for investigating substantially larger system sizes, allowing for more variety and complexity in future studies. | ['Tom Froese', 'Werner Koch', 'Natalya Weber'] | 2022-11-03 | null | null | null | null | ['artificial-life'] | ['miscellaneous'] | [ 3.02617937e-01 3.14705342e-01 3.14798206e-01 -2.67373826e-02
4.37164783e-01 -5.13871908e-01 4.84593838e-01 3.71740192e-01
-5.90670586e-01 8.86410415e-01 -3.10158879e-01 -5.83165765e-01
-5.59775710e-01 -9.51261461e-01 -3.83693874e-01 -9.50795531e-01
-8.86582077e-01 5.57645380e-01 5.62383354e-01 -6.81417048e-01
1.80633798e-01 4.47120965e-01 -1.85729241e+00 -4.08721447e-01
6.48170054e-01 8.71985793e-01 9.45937075e-03 8.35610926e-01
2.49238789e-01 2.86553651e-01 -5.80900073e-01 -1.34021983e-01
4.22386438e-01 -7.63213873e-01 -7.11793005e-01 -1.76025759e-02
-6.63832948e-02 2.78236777e-01 -3.96682084e-01 8.54885995e-01
5.73418975e-01 1.78878233e-01 4.85944897e-01 -1.17060101e+00
-2.35231072e-01 5.60747087e-01 -1.96832955e-01 4.89627242e-01
2.62204222e-02 2.85613537e-01 8.58758330e-01 -1.16962209e-01
6.35498583e-01 1.09810746e+00 5.80502868e-01 2.58529574e-01
-1.35971999e+00 -4.95242894e-01 1.68166578e-01 -3.34281921e-01
-1.31590712e+00 -2.95489222e-01 4.77226615e-01 -2.26678997e-01
1.11682963e+00 4.08401340e-01 9.64103818e-01 2.81464756e-01
3.47465873e-01 7.61232302e-02 1.26680887e+00 -5.32401919e-01
5.77289879e-01 1.15451984e-01 -8.79781041e-03 7.33051419e-01
7.72513032e-01 2.38745913e-01 -2.84071028e-01 -2.50955015e-01
8.53546560e-01 -3.89595836e-01 8.68277848e-02 -2.22166762e-01
-1.14770436e+00 7.15218484e-01 3.57573539e-01 6.60197973e-01
-2.17264801e-01 3.91316354e-01 2.73783743e-01 5.85864902e-01
4.51130927e-01 8.36023629e-01 -5.18627584e-01 -3.00261706e-01
-8.28527331e-01 2.12650537e-01 1.25023413e+00 7.11364627e-01
7.61222243e-01 1.22609340e-01 6.75666392e-01 6.02771997e-01
3.10508680e-04 2.92894781e-01 3.58848311e-02 -1.23324966e+00
-5.76760694e-02 9.61612999e-01 -1.25027662e-02 -9.83002543e-01
-7.96804667e-01 -7.57329106e-01 -1.13699973e+00 3.42089057e-01
7.25068152e-01 -4.52902377e-01 -2.14036122e-01 1.71072590e+00
3.67472708e-01 -3.73985797e-01 -8.44148397e-02 4.93666053e-01
1.10758670e-01 4.63415951e-01 -3.17770183e-01 -4.15177345e-01
1.26208758e+00 -6.35718226e-01 -1.36647895e-01 -1.72695935e-01
4.97228920e-01 -6.32741511e-01 8.37881386e-01 3.66708100e-01
-1.28691423e+00 -1.57565489e-01 -1.00443983e+00 5.85792124e-01
-5.48778772e-01 -6.05543494e-01 1.19988787e+00 9.06172037e-01
-1.42298126e+00 9.61027563e-01 -9.40355599e-01 -8.50522041e-01
2.30929792e-01 8.52832258e-01 -1.51938230e-01 2.78999478e-01
-9.65442061e-01 7.24393249e-01 1.94018453e-01 4.95186858e-02
-4.73504514e-01 -2.85427213e-01 -4.45264995e-01 1.40805393e-01
7.19128072e-01 -7.75848925e-01 5.55333138e-01 -9.01903152e-01
-1.44303775e+00 4.67447788e-01 -1.17877483e-01 -5.12694061e-01
3.94796729e-01 4.35571820e-01 -1.05423898e-01 7.97027349e-03
-2.24583998e-01 3.68898749e-01 2.87942827e-01 -1.11143923e+00
3.04064844e-02 -5.03852308e-01 3.82781059e-01 8.52014124e-03
-5.03349304e-01 1.17444918e-01 4.53052595e-02 -7.10378230e-01
3.47687721e-01 -1.36497116e+00 -6.83693647e-01 -1.44375578e-01
-1.47059843e-01 -1.95826277e-01 3.13499004e-01 3.65416557e-02
1.29959738e+00 -1.84690356e+00 2.96234548e-01 6.06057048e-01
4.29156125e-01 1.37111798e-01 7.36201135e-03 9.99271154e-01
9.93420705e-02 3.69214803e-01 -2.89587319e-01 4.25499454e-02
-1.09697215e-01 2.99069762e-01 5.35026848e-01 3.87760013e-01
-5.01596294e-02 5.12951314e-01 -8.27813148e-01 -3.76876980e-01
-1.00181945e-01 2.75880277e-01 -5.59034705e-01 -2.51849443e-01
-1.18210264e-01 2.30792999e-01 -3.70957047e-01 5.04360020e-01
2.47526988e-01 -7.32366145e-01 3.01879436e-01 2.32549131e-01
-2.93185949e-01 6.73427209e-02 -1.41515708e+00 1.07527554e+00
-1.27981931e-01 5.61136603e-01 3.78856868e-01 -9.48747933e-01
7.45300174e-01 1.97063595e-01 7.37288892e-01 -4.33882207e-01
4.11990196e-01 1.70871168e-01 7.04478025e-01 -1.42473400e-01
1.51694894e-01 -2.84519672e-01 -1.26673311e-01 9.85720515e-01
-1.38422996e-01 -1.79278925e-01 6.78279996e-01 2.48953313e-01
1.76760805e+00 -2.34870061e-01 2.10639030e-01 -8.90605628e-01
5.55987775e-01 -8.39829892e-02 3.27090710e-01 9.38792884e-01
-4.01138246e-01 -2.15459839e-01 7.89309859e-01 -5.72385550e-01
-1.29938066e+00 -7.28293419e-01 -1.35061026e-01 1.24853110e+00
1.95617229e-01 -4.99859363e-01 -9.96269643e-01 -1.56709671e-01
-1.98476128e-02 1.56851843e-01 -6.67880774e-01 3.16624790e-02
-6.18970037e-01 -1.10136890e+00 5.69094181e-01 1.26821652e-01
2.99058735e-01 -1.32299745e+00 -7.10484505e-01 4.18485135e-01
4.57519084e-01 -8.20060372e-01 -6.07230663e-02 4.56338823e-01
-9.94222999e-01 -8.70082974e-01 -3.21709126e-01 -4.13465798e-01
9.38222766e-01 1.11945316e-01 9.54198420e-01 6.65758014e-01
-6.15261376e-01 1.99614972e-01 -2.15030611e-01 -3.90111595e-01
-5.47097385e-01 3.85979503e-01 4.69532907e-01 -3.18711638e-01
8.22436959e-02 -1.05878723e+00 -7.48071134e-01 4.42005157e-01
-1.12539887e+00 -1.23008333e-01 4.85796392e-01 6.70694232e-01
2.26914823e-01 5.64741552e-01 8.69168341e-01 -7.36419022e-01
7.33234048e-01 -3.60319316e-01 -8.21751654e-01 -6.30213097e-02
-7.88360476e-01 2.48476923e-01 9.06334043e-01 -3.79267842e-01
-4.77257103e-01 -5.91606572e-02 6.33514225e-02 4.19367790e-01
-6.56415969e-02 3.03522319e-01 2.89810985e-01 -5.22903204e-01
4.79325980e-01 3.18936020e-01 3.63612175e-01 -1.48496479e-01
9.56610963e-02 4.37570572e-01 8.86426866e-02 -4.60039496e-01
8.75070989e-01 4.94877219e-01 5.98430336e-01 -1.05579138e+00
-2.95473546e-01 4.94230539e-02 -5.82133532e-01 -1.66915491e-01
3.83100450e-01 -3.83776635e-01 -1.25179338e+00 1.34281129e-01
-4.52196300e-01 -3.69951993e-01 -3.64668638e-01 1.99234232e-01
-4.70703602e-01 3.58861595e-01 -5.81447005e-01 -1.03712666e+00
-2.90965557e-01 -8.93086970e-01 2.87921846e-01 3.11096489e-01
-6.18139982e-01 -1.03818214e+00 8.39424878e-02 3.54597390e-01
8.74413788e-01 1.72098309e-01 1.05083442e+00 -5.93348086e-01
-6.55244470e-01 -3.63312185e-01 -2.54361313e-02 1.33087322e-01
2.99681388e-02 1.90714031e-01 -2.04817325e-01 -6.28761709e-01
-7.73718655e-02 -2.65828203e-02 5.86199880e-01 4.47444841e-02
8.10435236e-01 -3.77387255e-01 -2.86615729e-01 8.46941322e-02
1.40156007e+00 1.53010473e-01 4.71840650e-01 3.83177221e-01
4.37244810e-02 6.82873785e-01 7.20933601e-02 6.93825305e-01
4.04646933e-01 4.21709001e-01 5.29154837e-01 -2.12728903e-01
3.96325514e-02 4.15644795e-01 2.12355301e-01 7.90925264e-01
-3.31898361e-01 -2.68613398e-01 -9.74650443e-01 5.96270561e-01
-1.68251610e+00 -8.68083358e-01 3.57546769e-02 2.10961962e+00
5.94096899e-01 5.81185102e-01 4.00053799e-01 2.53433645e-01
5.82431436e-01 1.12790309e-01 -6.76785409e-01 -5.22091746e-01
-2.22594604e-01 5.49005866e-01 4.84440655e-01 4.58336979e-01
-5.31179607e-01 8.21002722e-01 6.31859636e+00 6.53469682e-01
-9.05102491e-01 -1.52768269e-01 4.00088817e-01 -3.69297624e-01
-1.19541779e-01 4.83215034e-01 -4.94944274e-01 5.67729592e-01
1.05681264e+00 -2.28652984e-01 8.63920212e-01 5.13582885e-01
3.90939772e-01 -6.54903829e-01 -9.37958062e-01 3.28084737e-01
-2.13054284e-01 -1.15296829e+00 -4.49999362e-01 5.07758081e-01
6.58219278e-01 1.64005533e-01 -2.70811558e-01 -1.66419536e-01
6.97441399e-01 -9.22878385e-01 3.93387347e-01 1.77864075e-01
3.15878272e-01 -4.81660903e-01 6.46468222e-01 6.13770783e-01
-1.10502708e+00 -1.97317526e-01 -1.94324523e-01 -6.08417511e-01
7.67832398e-02 7.04506516e-01 -5.41987181e-01 2.52919018e-01
9.75629091e-01 1.65429056e-01 -6.83889806e-01 7.34001219e-01
3.76453936e-01 4.11398888e-01 -1.02845895e+00 -6.24328732e-01
1.43501967e-01 -2.33814329e-01 7.93515146e-01 7.57656217e-01
2.34899342e-01 2.55136907e-01 5.09939063e-03 5.86036921e-01
4.93059196e-02 2.49532461e-02 -4.82512027e-01 -2.96563506e-01
4.68269497e-01 1.17863953e+00 -1.40676510e+00 -3.00689936e-01
3.41174267e-02 4.65727150e-01 2.40317598e-01 2.13877000e-02
-3.50215316e-01 -5.87575197e-01 3.65642250e-01 7.04392195e-01
3.72175068e-01 -6.64843559e-01 -3.52391690e-01 -9.38260436e-01
-4.36379351e-02 -6.42888784e-01 1.17451832e-01 -2.73051500e-01
-1.23785257e+00 7.90701330e-01 -6.01082705e-02 -7.81134009e-01
-1.13130765e-04 -3.74069721e-01 -3.88192922e-01 3.23171973e-01
-8.59203100e-01 -3.48285228e-01 -1.59714535e-01 3.70039433e-01
3.11603043e-02 -1.71259910e-01 8.46600771e-01 -3.48782279e-02
-7.94403732e-01 4.09356266e-01 4.48855370e-01 -4.42489266e-01
1.93665907e-01 -9.39556718e-01 1.01511911e-01 5.97483397e-01
-3.89798164e-01 8.58707011e-01 1.07320428e+00 -5.25943041e-01
-1.37324989e+00 -4.19029742e-01 8.24553490e-01 -3.40493888e-01
8.41462374e-01 -5.91258407e-01 -5.66943705e-01 2.10978985e-01
4.62737739e-01 -6.80820793e-02 7.43109345e-01 2.80955117e-02
8.24719742e-02 -2.51683593e-01 -1.22604382e+00 7.58667409e-01
1.32549715e+00 7.14082569e-02 -4.87553701e-02 2.34863728e-01
5.23561358e-01 2.94294477e-01 -1.31289804e+00 3.71466130e-01
5.73750257e-01 -1.11704373e+00 7.65667498e-01 -2.37388462e-01
1.47359058e-01 -4.51145023e-01 9.91204977e-02 -8.83072913e-01
-2.87517458e-01 -1.01394844e+00 -4.74021062e-02 9.89358068e-01
4.95730460e-01 -1.16120887e+00 8.24379861e-01 4.70495015e-01
2.95514554e-01 -9.84131515e-01 -1.15389633e+00 -7.87292063e-01
1.44114867e-01 4.08672839e-02 5.32438040e-01 8.03680420e-01
3.31703722e-01 2.51923293e-01 1.18226953e-01 -2.29177684e-01
5.96955240e-01 -7.98702985e-02 6.92620218e-01 -1.50263333e+00
-6.16602659e-01 -6.15963399e-01 -2.99001485e-01 -7.15691686e-01
-2.73388743e-01 -5.78820288e-01 -3.24567735e-01 -1.04859173e+00
1.77756712e-01 -9.97787416e-01 -3.07535648e-01 4.10304129e-01
1.65854007e-01 3.12901616e-01 2.02681169e-01 1.60778105e-01
-7.71106124e-01 3.36890638e-01 1.07456970e+00 3.47143918e-01
-2.10461110e-01 -9.60496664e-02 -9.36579466e-01 7.98722029e-01
8.76215219e-01 -6.24589324e-01 -2.35492572e-01 2.39281589e-03
6.83119178e-01 5.24244756e-02 1.58592716e-01 -1.04570448e+00
4.14583534e-01 -2.48167127e-01 -2.87536345e-02 -6.45653084e-02
3.04466724e-01 -7.28138626e-01 3.29417884e-01 6.62763298e-01
-1.83192581e-01 3.47366631e-01 7.08243325e-02 3.99016410e-01
1.84261218e-01 -1.87983513e-01 6.84501529e-01 -4.86097008e-01
8.78173783e-02 4.87132464e-03 -8.06913793e-01 -8.15693960e-02
1.25964308e+00 -4.88024533e-01 -3.90631676e-01 -3.42787236e-01
-8.04883838e-01 2.20832199e-01 8.72873724e-01 -6.01564441e-03
-1.12500057e-01 -6.99840665e-01 -3.22621644e-01 1.47499025e-01
-2.90010840e-01 -7.82413001e-04 1.17624357e-01 7.81269312e-01
-8.17816854e-01 4.46232021e-01 -2.61736631e-01 -3.60291064e-01
-1.15796387e+00 5.38184106e-01 4.61055964e-01 -4.70539391e-01
-2.07912296e-01 8.01935554e-01 -3.41729999e-01 -2.96471626e-01
-1.19397148e-01 2.41971597e-01 2.30527848e-01 -2.33295441e-01
-1.57481674e-02 8.47057283e-01 -1.92657605e-01 -4.44868743e-01
-3.91953290e-01 5.13475835e-01 -7.33324438e-02 -9.08246785e-02
1.71342444e+00 -2.89503068e-01 -8.38949740e-01 2.49331370e-01
6.11371100e-01 -1.58468947e-01 -1.03932369e+00 -1.50849372e-01
1.66141912e-02 -4.27637435e-02 -2.59138554e-01 -4.47491348e-01
-7.91453600e-01 6.27185822e-01 3.53670955e-01 9.40054774e-01
1.11937106e+00 2.23403290e-01 4.67971146e-01 5.18887639e-01
6.99192047e-01 -1.21601427e+00 8.86170417e-02 4.04179782e-01
4.35435236e-01 -8.05688739e-01 4.28250372e-01 -4.60179359e-01
-5.39516844e-02 9.40546036e-01 5.64932883e-01 -4.62361336e-01
8.64094615e-01 3.98211837e-01 -4.27947342e-01 -3.35792750e-01
-9.42624331e-01 -9.87378061e-02 -3.42418402e-01 2.37636194e-01
5.06573558e-01 6.86844885e-02 -7.50222266e-01 1.38219744e-01
-5.25468588e-01 -1.23886280e-01 8.20580542e-01 1.35249650e+00
-5.94670653e-01 -1.61780846e+00 -1.07789770e-01 6.18133426e-01
-4.65899527e-01 -2.12228037e-02 -6.27682328e-01 7.67575502e-01
2.77990937e-01 1.03596795e+00 -1.14065923e-01 -8.66706744e-02
5.10904789e-02 -4.31834124e-02 6.20539784e-01 -4.78632480e-01
-9.23606813e-01 -8.90338868e-02 1.41484812e-01 -2.75378764e-01
-3.69017750e-01 -7.47683465e-01 -1.24167395e+00 -8.55877638e-01
-4.26615655e-01 2.46413633e-01 4.79083687e-01 9.40601230e-01
6.30970895e-01 5.17481029e-01 4.38827872e-01 -9.04600859e-01
-3.19223434e-01 -7.69063532e-01 -5.95695376e-01 7.71314204e-02
9.16911289e-02 -6.76195025e-01 -4.86165166e-01 -2.23085254e-01] | [6.096345901489258, 4.588033199310303] |
01cd4691-6750-4f0c-b28e-68a53d3cd0f3 | heuristic-modularity-maximization-algorithms | 2302.14698 | null | https://arxiv.org/abs/2302.14698v3 | https://arxiv.org/pdf/2302.14698v3.pdf | Heuristic Modularity Maximization Algorithms for Community Detection Rarely Return an Optimal Partition or Anything Similar | Community detection is a fundamental problem in computational sciences with extensive applications in various fields. The most commonly used methods are the algorithms designed to maximize modularity over different partitions of the network nodes. Using 80 real and random networks from a wide range of contexts, we investigate the extent to which current heuristic modularity maximization algorithms succeed in returning maximum-modularity (optimal) partitions. We evaluate (1) the ratio of the algorithms' output modularity to the maximum modularity for each input graph, and (2) the maximum similarity between their output partition and any optimal partition of that graph. We compare eight existing heuristic algorithms against an exact integer programming method that globally maximizes modularity. The average modularity-based heuristic algorithm returns optimal partitions for only 19.4% of the 80 graphs considered. Additionally, results on adjusted mutual information reveal substantial dissimilarity between the sub-optimal partitions and any optimal partition of the networks in our experiments. More importantly, our results show that near-optimal partitions are often disproportionately dissimilar to any optimal partition. Taken together, our analysis points to a crucial limitation of commonly used modularity-based heuristics for discovering communities: they rarely produce an optimal partition or a partition resembling an optimal partition. If modularity is to be used for detecting communities, exact or approximate optimization algorithms are recommendable for a more methodologically sound usage of modularity within its applicability limits. | ['Hriday Chheda', 'Mahdi Mostajabdaveh', 'Samin Aref'] | 2023-02-28 | null | null | null | null | ['community-detection'] | ['graphs'] | [ 1.82995841e-01 2.76121587e-01 -1.65626749e-01 2.73813665e-01
-9.54106599e-02 -9.51153100e-01 1.32090986e-01 4.48784977e-01
-1.72680721e-01 6.21139526e-01 -2.15263620e-01 -4.66029108e-01
-7.69170761e-01 -1.16579926e+00 -1.74446911e-01 -4.63217020e-01
-8.83120954e-01 8.13943863e-01 5.19744873e-01 -2.22936030e-02
4.43589360e-01 4.04425502e-01 -1.41636443e+00 1.64535403e-01
1.09723318e+00 2.73848355e-01 2.87174344e-01 7.19839096e-01
-1.08532429e-01 8.21594968e-02 -5.74039459e-01 -5.64279437e-01
4.45903242e-01 -4.58801508e-01 -1.06647599e+00 4.33880866e-01
-2.29240924e-01 1.82785496e-01 -1.26091570e-01 1.29069901e+00
2.33960345e-01 -3.60418975e-01 7.36689687e-01 -1.56165063e+00
-2.82231569e-01 9.65493858e-01 -9.75721419e-01 3.84728432e-01
6.99968219e-01 1.52497470e-01 1.59032524e+00 -3.80402416e-01
8.11861813e-01 1.21629810e+00 6.98720992e-01 -2.55159616e-01
-1.87649906e+00 -5.61447442e-01 5.97192496e-02 -4.47001010e-02
-1.95490730e+00 -1.17366597e-01 5.91041446e-01 -6.40490174e-01
1.08306277e+00 5.72648942e-01 1.06845641e+00 1.79778025e-01
2.61136323e-01 2.66730130e-01 9.01308179e-01 -2.74441719e-01
1.52677089e-01 6.35463046e-03 -1.22145072e-01 6.25090182e-01
1.11578393e+00 -3.93804580e-01 7.86706433e-02 -7.36550570e-01
6.35272920e-01 -2.33162567e-01 -2.72181660e-01 -5.79673707e-01
-1.47809231e+00 8.50247860e-01 2.29322895e-01 4.05345708e-01
-3.81865203e-01 -5.50015233e-02 2.72107154e-01 4.60937023e-01
1.59810588e-01 8.40955257e-01 -2.83134937e-01 2.91810870e-01
-1.03933465e+00 2.01108143e-01 1.27679324e+00 8.34201753e-01
1.01539636e+00 -3.62553477e-01 2.66937703e-01 6.18402600e-01
4.31260973e-01 6.58265352e-02 3.17688398e-02 -1.01843214e+00
3.78553063e-01 1.17139566e+00 -1.13795668e-01 -1.61545444e+00
-3.63427907e-01 -5.67702472e-01 -8.84529352e-01 -1.66901797e-02
5.04926443e-01 4.99479733e-02 -8.15489069e-02 1.49838746e+00
1.35335296e-01 -4.27253336e-01 -3.15599918e-01 6.49022400e-01
3.88351947e-01 2.61888564e-01 -1.82319760e-01 -5.92358708e-01
1.35664546e+00 -3.43933523e-01 -1.79251164e-01 -1.96833596e-01
3.65984023e-01 -9.20461297e-01 4.32217509e-01 3.45771492e-01
-8.84235024e-01 7.43706003e-02 -1.11925113e+00 8.55926394e-01
3.74543220e-02 -3.76311988e-01 5.39211512e-01 7.79454172e-01
-1.47544575e+00 6.29695415e-01 -3.12252581e-01 -6.93519711e-01
-4.53465134e-02 6.38972342e-01 -4.28260416e-01 2.40465254e-01
-8.13047469e-01 6.23444438e-01 9.74848032e-01 -3.33628625e-01
-4.00669545e-01 -4.69635203e-02 -4.86140907e-01 4.38251168e-01
4.31850761e-01 -8.00047696e-01 5.67294359e-01 -1.30995858e+00
-6.05741441e-01 1.09500718e+00 1.23319373e-01 -3.11392456e-01
2.90312558e-01 8.58519137e-01 -1.32193297e-01 4.89307851e-01
4.10767496e-01 5.50691128e-01 5.02898157e-01 -1.25799477e+00
-4.70701665e-01 4.22324352e-02 2.68991798e-01 2.24244252e-01
-3.36409628e-01 7.17647024e-04 -1.99217901e-01 -3.48540455e-01
6.51377559e-01 -7.02792048e-01 -5.48604429e-01 -2.10205585e-01
-8.43746364e-01 -2.45134726e-01 7.88706243e-02 -4.19844776e-01
1.72732222e+00 -1.75326765e+00 1.80935562e-01 8.35631371e-01
7.32923508e-01 -3.23824435e-01 -2.58617967e-01 8.93823624e-01
-3.46396595e-01 6.47789240e-01 -3.63362312e-01 2.55195588e-01
-1.71469413e-02 1.10304162e-01 2.81422853e-01 8.06820214e-01
1.70342371e-01 3.40431273e-01 -1.02644253e+00 -8.23345363e-01
-8.00919458e-02 -2.46635526e-01 -7.49266744e-01 -1.91015363e-01
2.50339448e-01 -3.00118625e-01 -1.29509121e-01 6.58541441e-01
6.24397933e-01 -5.87456405e-01 1.12756586e+00 1.60090268e-01
-2.08627850e-01 8.82100239e-02 -1.60263193e+00 8.17490399e-01
4.77702707e-01 6.27781093e-01 2.31669560e-01 -1.01142955e+00
8.81828487e-01 4.41914946e-01 8.40719283e-01 -1.58481315e-01
-5.07470518e-02 1.50332212e-01 7.83048689e-01 -1.57825291e-01
5.05778313e-01 -1.72354221e-01 1.13327704e-01 1.02749014e+00
-6.71090335e-02 1.39697388e-01 8.17914546e-01 5.51870465e-01
1.57038081e+00 -7.51047254e-01 1.04230452e+00 -9.84576643e-01
4.67838466e-01 1.19725697e-01 6.21021628e-01 7.59784341e-01
-3.96887779e-01 6.00525618e-01 1.24366319e+00 -1.10378481e-01
-1.40161574e+00 -1.20967734e+00 -9.76481661e-03 7.52057612e-01
2.18992144e-01 -6.30063832e-01 -7.61303782e-01 -3.14974248e-01
2.69533455e-01 -4.60221171e-02 -3.56240034e-01 1.49319008e-01
-1.61126092e-01 -9.19051468e-01 3.73585433e-01 -2.11094052e-01
-5.06783538e-02 -8.12016010e-01 -4.73911524e-01 3.52127403e-01
-5.47222078e-01 -6.78621948e-01 -4.15654600e-01 -7.16429353e-02
-1.21798086e+00 -1.63009632e+00 -3.76986891e-01 -6.94071472e-01
1.03303957e+00 7.50370264e-01 1.27169168e+00 7.58406937e-01
-3.51054132e-01 3.14036936e-01 -1.02025874e-01 2.86007673e-01
-5.67071855e-01 2.41415083e-01 9.25584510e-02 -2.27760538e-01
7.29903355e-02 -9.27355826e-01 -4.95294362e-01 5.50102413e-01
-7.80754030e-01 -1.36520937e-01 5.52091122e-01 4.87569094e-01
1.75742954e-01 6.30019486e-01 5.03400147e-01 -5.31133235e-01
1.08760786e+00 -9.58498836e-01 -3.59765470e-01 2.75647968e-01
-7.98615217e-01 1.50331870e-01 3.41140538e-01 -3.72377813e-01
-2.31611446e-01 -3.53963710e-02 3.73001754e-01 7.35587534e-03
2.98848450e-01 7.28864133e-01 -2.52467413e-02 4.55221832e-02
6.28644049e-01 9.86859798e-02 1.31311581e-01 -6.75562546e-02
2.16839779e-02 4.36999083e-01 3.07627534e-03 -6.64297402e-01
7.14517593e-01 4.74552810e-01 -5.14396057e-02 -8.76805842e-01
6.45118356e-02 -8.56752634e-01 -6.64899111e-01 -4.31516439e-01
2.12646276e-01 -5.51707029e-01 -9.08183515e-01 -1.12890676e-01
-1.17749155e+00 2.45229572e-01 8.61304775e-02 1.17999420e-01
-5.25934935e-01 1.09047318e+00 -3.53116781e-01 -1.03638041e+00
-3.44336152e-01 -1.04156291e+00 4.93731588e-01 -7.59586617e-02
-7.58538544e-01 -9.72846627e-01 3.35113764e-01 -2.57939450e-03
-2.13969499e-02 3.85746241e-01 9.73828554e-01 -6.20320797e-01
-6.55664921e-01 1.56212775e-02 -4.63607460e-01 -2.36557230e-01
2.77515352e-01 7.17817068e-01 -3.94891948e-01 -6.76600993e-01
-5.16797781e-01 3.84831280e-01 5.54036319e-01 3.38771373e-01
2.81265497e-01 -5.86364806e-01 -6.94392979e-01 6.93541020e-02
1.71268845e+00 6.31244183e-02 5.01100719e-01 5.29087722e-01
5.31507917e-02 7.78603494e-01 1.19842924e-01 7.05436885e-01
2.43382588e-01 2.83907980e-01 6.22239172e-01 2.62707412e-01
4.17323530e-01 -5.29797524e-02 1.84928358e-01 1.00390434e+00
-1.94558024e-01 -2.57811487e-01 -1.21825349e+00 9.51859772e-01
-1.86989152e+00 -1.32816911e+00 -3.33786458e-01 2.35667014e+00
7.67852306e-01 4.19671565e-01 5.89972854e-01 3.58015805e-01
1.29570413e+00 -1.39216498e-01 -8.12306330e-02 -5.27112305e-01
-1.83637843e-01 -1.69802338e-01 3.04305017e-01 4.84506398e-01
-9.29346442e-01 3.63908291e-01 7.67072439e+00 6.40142679e-01
-4.01865900e-01 -1.78896755e-01 3.53083700e-01 -8.61438960e-02
-5.41848898e-01 4.06120747e-01 -1.05541326e-01 3.73051494e-01
7.01047063e-01 -8.29731882e-01 5.85828424e-01 6.22312129e-01
2.48297274e-01 -3.92462760e-01 -9.39834952e-01 5.62582910e-01
-4.68035966e-01 -1.04948807e+00 -9.54088420e-02 7.11058557e-01
8.67150724e-01 -1.75967783e-01 -3.97679061e-01 -2.43198395e-01
5.68128586e-01 -1.01646948e+00 6.74785733e-01 4.73292321e-01
5.90900064e-01 -9.31902885e-01 5.86815059e-01 4.18907225e-01
-1.56154191e+00 -9.01059657e-02 -4.90601212e-01 -5.51534653e-01
-1.40955225e-01 1.00663483e+00 -1.28265083e+00 4.63185132e-01
6.65761292e-01 3.80756289e-01 -5.24969757e-01 1.47801018e+00
8.49897880e-03 3.36628288e-01 -6.73935235e-01 -1.24983862e-01
1.39017731e-01 -4.63632643e-01 1.07635593e+00 1.41564119e+00
1.61016583e-01 -3.63537252e-01 1.83756664e-01 1.03233492e+00
5.39950132e-02 2.13220030e-01 -8.03972065e-01 -3.27914864e-01
8.37762415e-01 1.31126404e+00 -1.66132867e+00 -1.56869113e-01
-2.06538469e-01 6.43668890e-01 4.38530207e-01 2.32408211e-01
-4.86040264e-01 -5.24685502e-01 1.00274432e+00 3.62563342e-01
2.77778625e-01 -2.14512661e-01 -4.61029470e-01 -8.51718605e-01
-1.90676190e-02 -8.25671852e-01 6.43781841e-01 -2.71748573e-01
-1.27923608e+00 2.52380699e-01 5.06780595e-02 -1.24046278e+00
-1.39420807e-01 -1.97188810e-01 -7.69152880e-01 7.47344732e-01
-6.49477422e-01 -3.19436967e-01 1.84597775e-01 7.63834938e-02
3.75899784e-02 3.79489884e-02 5.50700605e-01 1.98732778e-01
-5.14479697e-01 1.24970496e-01 2.40505502e-01 8.42439979e-02
2.95383215e-01 -1.32205522e+00 3.05999815e-01 9.58371758e-01
-1.31575659e-01 1.01129651e+00 9.30073977e-01 -8.60846579e-01
-9.81332898e-01 -4.88268614e-01 1.09942663e+00 3.80012095e-02
9.57813203e-01 -7.76530728e-02 -5.57919681e-01 4.60624129e-01
4.30683821e-01 -6.62154377e-01 6.69873536e-01 2.86394507e-01
-7.34903589e-02 1.78447559e-01 -1.20854175e+00 8.50819349e-01
1.10151279e+00 -2.94090271e-01 -3.80253971e-01 8.22087377e-02
4.08227503e-01 4.97274131e-01 -1.03609335e+00 1.51437551e-01
6.55477166e-01 -1.36158574e+00 1.11960411e+00 -3.62353206e-01
2.92425185e-01 -5.03778458e-01 3.47819552e-02 -1.06237423e+00
-9.63272572e-01 -6.44333780e-01 1.84359819e-01 1.06139565e+00
4.08253461e-01 -7.50922918e-01 7.82781959e-01 7.67035186e-02
4.08150822e-01 -7.66382515e-01 -1.12599814e+00 -6.77332520e-01
-1.18400164e-01 -2.98331995e-02 4.81002480e-01 1.33145428e+00
6.02698803e-01 2.74284482e-01 2.29987890e-01 1.05245844e-01
1.03133261e+00 3.19670677e-01 5.56187451e-01 -1.68297648e+00
-2.50681251e-01 -1.07666457e+00 -7.80624390e-01 -4.34379011e-01
-9.29606706e-02 -1.08090281e+00 -2.83930093e-01 -1.72782695e+00
9.35779810e-01 -3.13631386e-01 8.25446248e-02 2.25498483e-01
-1.06496356e-01 1.77862287e-01 7.46624470e-02 4.40106243e-01
-7.34301388e-01 -5.68935610e-02 1.18366754e+00 -2.80356705e-01
-2.29885384e-01 -5.70437796e-02 -1.05819774e+00 5.79027116e-01
8.68052304e-01 -7.61336029e-01 -2.81232983e-01 2.49383211e-01
7.87319839e-01 4.86412756e-02 1.69054568e-01 -7.61735916e-01
4.15685892e-01 -3.50473583e-01 1.24120101e-01 -4.31206673e-01
-3.11646461e-01 -6.53338552e-01 8.28121245e-01 9.75235879e-01
1.79101720e-01 3.78737956e-01 -5.58383018e-02 5.80522895e-01
1.18900426e-01 -5.65619111e-01 5.83947122e-01 -4.54798341e-01
-3.43947649e-01 -1.82012722e-01 -1.04249072e+00 8.41493607e-02
1.17618895e+00 -9.31467950e-01 -2.33795777e-01 -3.54214549e-01
-7.16381609e-01 3.72398943e-01 8.79285693e-01 -3.17720287e-02
5.92120588e-01 -1.10107875e+00 -8.56111884e-01 -2.54695565e-01
7.09544495e-02 -4.84620512e-01 -2.77007371e-01 1.00194860e+00
-8.96432340e-01 2.98872739e-01 -3.15307230e-01 -6.55018330e-01
-1.45526814e+00 6.36628091e-01 3.59882593e-01 -5.75017214e-01
-3.46293896e-01 3.61628234e-01 7.93644041e-02 -4.92816120e-01
-2.83859551e-01 -1.27331996e-02 1.45334080e-01 3.58305693e-01
1.11071669e-01 7.81141579e-01 -3.35762113e-01 -5.05636990e-01
-5.44888854e-01 1.35829255e-01 2.76301593e-01 -1.17159307e-01
1.03236151e+00 -3.24312538e-01 -9.80765522e-01 1.11088969e-01
9.08082247e-01 -1.36153242e-02 -6.32309198e-01 1.05066739e-01
4.22448963e-01 -4.51860100e-01 -5.54597974e-01 -1.25253201e-01
-8.29566479e-01 1.93351492e-01 -8.72042105e-02 1.11820674e+00
9.85039771e-01 7.60986954e-02 -3.31407674e-02 2.68604547e-01
5.99568486e-01 -8.86625230e-01 -4.52145673e-02 2.82698065e-01
6.41517997e-01 -8.15577090e-01 6.08713150e-01 -6.68977737e-01
-2.67141759e-01 1.11109936e+00 2.62013733e-01 -2.56381482e-01
5.33931375e-01 8.42442065e-02 -6.24747753e-01 -3.66246641e-01
-8.26628685e-01 -3.17658126e-01 1.40568748e-01 5.43780744e-01
4.93344307e-01 4.08705175e-01 -7.66654491e-01 4.25175101e-01
-2.47400030e-01 -6.53461397e-01 8.10985625e-01 7.33618677e-01
-9.90472138e-01 -9.12520409e-01 -7.39230216e-01 7.27035463e-01
-1.62116259e-01 -1.18692212e-01 -9.59495783e-01 9.01132882e-01
4.54161689e-02 1.10356247e+00 1.36922494e-01 -4.27749455e-01
4.48429175e-02 -9.53137949e-02 4.29617405e-01 -6.65724218e-01
-7.04128444e-01 4.90988716e-02 2.29711726e-01 -2.86744326e-01
-2.75756150e-01 -8.10731828e-01 -1.03569031e+00 -1.00933611e+00
-6.91666245e-01 2.68454164e-01 -4.23359089e-02 6.42417789e-01
1.67651966e-01 2.80511230e-01 4.96363789e-01 -5.29140890e-01
-2.41532490e-01 -8.00643265e-01 -9.43903923e-01 7.20085427e-02
-1.78208992e-01 -5.01874447e-01 -5.74878395e-01 -2.32623860e-01] | [6.946511745452881, 5.266814231872559] |
d77936e5-c408-4b20-8d21-1e6e314c8d97 | occcasnet-occlusion-aware-cascade-cost-volume | 2305.17710 | null | https://arxiv.org/abs/2305.17710v1 | https://arxiv.org/pdf/2305.17710v1.pdf | OccCasNet: Occlusion-aware Cascade Cost Volume for Light Field Depth Estimation | Light field (LF) depth estimation is a crucial task with numerous practical applications. However, mainstream methods based on the multi-view stereo (MVS) are resource-intensive and time-consuming as they need to construct a finer cost volume. To address this issue and achieve a better trade-off between accuracy and efficiency, we propose an occlusion-aware cascade cost volume for LF depth (disparity) estimation. Our cascaded strategy reduces the sampling number while keeping the sampling interval constant during the construction of a finer cost volume. We also introduce occlusion maps to enhance accuracy in constructing the occlusion-aware cost volume. Specifically, we first obtain the coarse disparity map through the coarse disparity estimation network. Then, the sub-aperture images (SAIs) of side views are warped to the center view based on the initial disparity map. Next, we propose photo-consistency constraints between the warped SAIs and the center SAI to generate occlusion maps for each SAI. Finally, we introduce the coarse disparity map and occlusion maps to construct an occlusion-aware refined cost volume, enabling the refined disparity estimation network to yield a more precise disparity map. Extensive experiments demonstrate the effectiveness of our method. Compared with state-of-the-art methods, our method achieves a superior balance between accuracy and efficiency and ranks first in terms of MSE and Q25 metrics among published methods on the HCI 4D benchmark. The code and model of the proposed method are available at https://github.com/chaowentao/OccCasNet. | ['Guanghui Wang', 'Yingqian Wang', 'Xuechun Wang', 'Fuqing Duan', 'Wentao Chao'] | 2023-05-28 | null | null | null | null | ['disparity-estimation'] | ['computer-vision'] | [ 1.49131015e-01 -4.54854220e-01 -3.51699628e-02 -4.90860194e-01
-6.65776491e-01 -2.93058217e-01 2.25902140e-01 -2.79498845e-01
-2.12250441e-01 6.98485017e-01 2.73548424e-01 -1.02526201e-02
9.82953794e-03 -9.43366110e-01 -5.46152472e-01 -7.76104629e-01
6.05702400e-01 -1.89162567e-02 4.71829057e-01 7.30819181e-02
4.55442637e-01 3.60576957e-01 -1.67359960e+00 3.21661010e-02
1.21513247e+00 1.22002518e+00 4.41498697e-01 2.95506477e-01
1.24176003e-01 3.81861567e-01 -1.92205057e-01 -1.29275918e-01
5.36240697e-01 -2.24535778e-01 -4.67795044e-01 7.87131116e-02
6.22965693e-01 -8.59487653e-01 -2.33963862e-01 1.07154822e+00
6.97900534e-01 3.19500454e-02 2.07499817e-01 -8.94365966e-01
-1.22567959e-01 -9.65009853e-02 -9.85198021e-01 7.10465237e-02
3.80921632e-01 1.98915005e-01 7.06702590e-01 -1.14313996e+00
6.42331362e-01 1.14226961e+00 2.30967447e-01 2.53133297e-01
-1.10804510e+00 -8.65061462e-01 2.54799813e-01 2.43797764e-01
-1.32953835e+00 -4.90447372e-01 9.24300730e-01 -2.57882476e-01
5.16359031e-01 9.70707685e-02 7.47514606e-01 6.83098853e-01
1.90947905e-01 5.56062758e-01 1.25249779e+00 -2.12939307e-01
1.32632151e-01 -1.14947811e-01 -4.66728322e-02 7.60898709e-01
2.21426457e-01 2.69016355e-01 -6.42084062e-01 5.21515943e-02
1.20296121e+00 2.66419314e-02 -7.11465538e-01 -5.03076136e-01
-1.16026676e+00 6.86475158e-01 5.71724415e-01 2.98587941e-02
-3.75138342e-01 6.84914086e-03 1.95395976e-01 -1.64678812e-01
5.38067460e-01 9.95865092e-02 -2.44203329e-01 5.85330762e-02
-7.16498137e-01 1.81885540e-01 3.03069979e-01 9.03674006e-01
1.10495126e+00 -1.88350961e-01 -2.11917147e-01 9.55377519e-01
2.32274860e-01 3.29396456e-01 -2.12057885e-02 -1.30005050e+00
6.50745273e-01 7.80577362e-01 1.76095381e-01 -9.71536279e-01
-3.77562791e-01 -5.58965027e-01 -9.53420579e-01 3.15701008e-01
4.01425570e-01 -2.32289042e-02 -6.75650835e-01 1.48306584e+00
6.31448328e-01 1.70683667e-01 -2.27858320e-01 1.14613187e+00
6.82966292e-01 6.23345673e-01 -3.08568805e-01 -3.52947325e-01
1.22444856e+00 -9.65033412e-01 -4.77640688e-01 -2.31040806e-01
2.04410896e-01 -8.67905557e-01 1.02261031e+00 5.03486753e-01
-1.31339300e+00 -6.41055048e-01 -1.16820323e+00 -3.79156679e-01
2.77253479e-01 2.67653793e-01 5.56365550e-01 2.91988373e-01
-9.55864429e-01 3.21070254e-01 -5.78642309e-01 5.08163357e-03
4.47125465e-01 2.91866809e-01 -4.65855524e-02 -4.33911830e-01
-9.46766555e-01 5.06510019e-01 1.29444614e-01 1.60317272e-01
-6.93590641e-01 -8.98828745e-01 -9.05025780e-01 -1.85274020e-01
4.36722100e-01 -1.02167571e+00 9.59278226e-01 -5.94103277e-01
-1.48370314e+00 6.73154831e-01 -4.29491490e-01 1.26824211e-02
4.99973625e-01 -1.15839638e-01 -3.26228626e-02 2.82220989e-01
2.08782300e-01 8.72003019e-01 6.70376062e-01 -1.29592109e+00
-1.10184300e+00 -4.07525986e-01 3.06545228e-01 5.09764791e-01
-8.15616846e-02 -4.69104677e-01 -8.33732605e-01 -4.24217314e-01
5.19181669e-01 -6.34072661e-01 -1.09468132e-01 2.53609598e-01
-3.67802352e-01 5.81732299e-03 5.43276727e-01 -5.39279103e-01
1.37371886e+00 -2.23181987e+00 1.80953220e-01 9.01323706e-02
5.99156082e-01 2.32472215e-02 -1.93887670e-02 -5.65177351e-02
1.74146324e-01 -3.76964152e-01 -1.15251429e-01 -3.11638206e-01
-4.36541349e-01 -5.87749109e-02 1.07018970e-01 6.13259077e-01
-5.91954589e-02 5.82920849e-01 -9.36841905e-01 -4.23967510e-01
7.18368948e-01 5.86883724e-01 -8.80652726e-01 2.22486392e-01
5.54365979e-04 7.96531558e-01 -5.01950920e-01 7.82396197e-01
1.02216434e+00 -3.32701951e-01 -3.43287773e-02 -7.32716620e-01
-4.89142239e-01 1.74605027e-01 -1.33855963e+00 1.93733740e+00
-7.74890304e-01 4.36731279e-01 -3.23931910e-02 -5.83651960e-01
9.53709066e-01 -1.35109182e-02 5.03693342e-01 -9.17358279e-01
2.92475075e-01 4.40402687e-01 -2.61521399e-01 -2.28473097e-01
4.62015688e-01 -4.81574424e-02 2.66008109e-01 1.29020765e-01
-2.57358402e-01 -2.89063245e-01 1.02223761e-01 -8.81253406e-02
5.60967565e-01 2.51786739e-01 3.44269246e-01 -2.06681103e-01
9.18381989e-01 -4.84733492e-01 8.39251995e-01 1.85755387e-01
-1.95242777e-01 9.07123268e-01 3.26900244e-01 -5.58419764e-01
-1.00436151e+00 -1.04403210e+00 -3.17650616e-01 4.12554801e-01
6.96450830e-01 -1.18857443e-01 -6.74687207e-01 -2.68460214e-01
-1.91088796e-01 3.78188193e-01 -3.46459746e-01 2.04065204e-01
-5.96694469e-01 -4.95964110e-01 -1.47191346e-01 3.98048967e-01
1.08450675e+00 -5.46888411e-01 -6.93852365e-01 1.24026947e-01
-6.49692357e-01 -1.18711221e+00 -6.94230616e-01 -3.70872855e-01
-8.96045864e-01 -1.00675619e+00 -7.97543526e-01 -6.89734042e-01
6.76768780e-01 5.48519254e-01 9.95669246e-01 5.13703562e-02
-5.41490763e-02 -9.46689844e-02 -5.44065461e-02 -1.01429582e-01
1.43899381e-01 -6.43898249e-02 -1.97295874e-01 1.16935268e-01
5.21507151e-02 -8.46545875e-01 -1.34216905e+00 5.79431593e-01
-8.34366620e-01 5.84957242e-01 4.51995552e-01 7.50023067e-01
8.66995394e-01 -1.63924754e-01 3.35639119e-01 -5.24202228e-01
1.33466855e-01 -2.10416377e-01 -1.12874532e+00 -6.63075075e-02
-5.74525595e-01 -1.07712992e-01 3.95165861e-01 -1.81542501e-01
-1.18359089e+00 1.60497740e-01 -2.36758471e-01 -5.15024900e-01
1.16731778e-01 5.42279817e-02 -2.99404085e-01 -2.73245603e-01
3.57784361e-01 9.22993794e-02 -1.29088253e-01 -2.99460620e-01
1.10887922e-01 4.79293615e-01 4.50369805e-01 -3.70447814e-01
5.47565341e-01 8.29458296e-01 1.63893431e-01 -4.52345729e-01
-1.01341271e+00 -4.56360251e-01 -3.95689517e-01 -4.56551582e-01
8.54748428e-01 -1.14119780e+00 -8.39533508e-01 6.42290354e-01
-1.17860055e+00 -5.89300320e-02 7.06712529e-02 5.82344472e-01
-4.49779719e-01 3.98234427e-01 -5.26826978e-01 -4.36976463e-01
-3.65382731e-01 -1.48615348e+00 1.27357697e+00 4.27010596e-01
8.25924054e-02 -7.62807906e-01 -9.67101008e-02 5.16926408e-01
3.29456747e-01 1.87049672e-01 8.33328843e-01 4.17656839e-01
-1.20830965e+00 2.21094251e-01 -7.29740739e-01 2.94056416e-01
1.48114100e-01 -1.77874520e-01 -1.03115666e+00 -2.26763889e-01
1.44636303e-01 -1.19150300e-02 6.38893545e-01 9.24912393e-01
1.13826871e+00 1.55882044e-02 -2.06179574e-01 1.14008439e+00
1.86592162e+00 3.49682063e-01 7.96691775e-01 4.37926352e-01
7.94622898e-01 7.09159732e-01 8.05971563e-01 5.98412216e-01
6.36040211e-01 8.97577643e-01 4.53518927e-01 -1.50480464e-01
-2.54491478e-01 -2.76251763e-01 -6.81209937e-02 7.30739295e-01
-1.43712074e-01 -2.07389995e-01 -7.14445353e-01 4.12282556e-01
-1.54560757e+00 -5.68941653e-01 -1.43753842e-01 2.36918020e+00
6.33435369e-01 2.37242747e-02 -1.11402683e-01 6.68136775e-02
6.47895992e-01 3.21067750e-01 -7.43807673e-01 9.57075134e-02
-6.62054494e-02 6.53654486e-02 4.30719525e-01 7.74914563e-01
-7.84200132e-01 6.57058179e-01 5.32391834e+00 9.27877545e-01
-1.12096131e+00 1.28249407e-01 8.55340600e-01 -3.88541371e-01
-5.14229238e-01 3.41539942e-02 -1.00424981e+00 5.35194278e-01
2.83880055e-01 2.64304923e-03 4.72954780e-01 5.47830343e-01
3.84982765e-01 -7.21824944e-01 -1.02192473e+00 1.39243805e+00
-8.13839361e-02 -1.46752560e+00 -5.98575687e-03 2.16212153e-01
9.37239051e-01 2.17483509e-02 -4.57013249e-02 -1.34110570e-01
-1.18293911e-01 -5.56879461e-01 5.46458721e-01 3.45480859e-01
1.07059956e+00 -8.37350488e-01 5.80709219e-01 1.42914042e-01
-1.39240849e+00 -4.97939214e-02 -3.26422870e-01 -1.96054652e-02
4.04669464e-01 9.68981922e-01 -8.09162706e-02 6.95987701e-01
6.71608090e-01 8.43777359e-01 -2.95444965e-01 9.17868316e-01
-1.76380813e-01 -1.16534978e-01 -1.33076891e-01 2.95627087e-01
8.41738433e-02 -6.05402231e-01 5.02918959e-01 5.16162872e-01
4.05383676e-01 2.56460249e-01 3.07179298e-02 9.25529361e-01
-2.95761731e-02 6.05061883e-03 -3.84209305e-01 7.72845209e-01
5.63756227e-01 1.22559023e+00 -5.35738409e-01 -3.49773616e-01
-6.77403808e-01 9.14251268e-01 2.19053030e-01 4.34924841e-01
-8.16952407e-01 -2.87811428e-01 7.77817428e-01 2.56056339e-01
9.05069485e-02 -1.87328383e-02 -4.31638569e-01 -1.26325262e+00
3.69806081e-01 -7.40084529e-01 1.73652321e-01 -9.09615755e-01
-9.99722302e-01 6.98029816e-01 5.62471971e-02 -1.74316525e+00
-6.88330159e-02 -2.53137410e-01 -3.19692135e-01 1.09257317e+00
-1.94673395e+00 -8.71763945e-01 -9.31748390e-01 5.07733524e-01
7.17697799e-01 2.72204340e-01 2.62901783e-01 6.12654090e-01
-5.15583158e-01 3.26448381e-01 -1.90190479e-01 -2.68217862e-01
5.51709831e-01 -6.02263749e-01 2.66649306e-01 8.75427067e-01
-5.07290900e-01 3.17753941e-01 3.04775804e-01 -5.04639328e-01
-1.13795924e+00 -9.68987703e-01 6.01615965e-01 -1.99266031e-01
1.86274767e-01 -2.13107303e-01 -8.37094605e-01 2.81050473e-01
-1.70119256e-01 1.68626443e-01 1.12016797e-01 -3.14181775e-01
-8.71143937e-02 -4.81343001e-01 -1.07432437e+00 5.75182080e-01
1.29141998e+00 -4.54267055e-01 -2.65447050e-02 2.18830584e-03
7.75064826e-01 -8.32539558e-01 -7.30183899e-01 6.17968380e-01
8.18159580e-01 -1.58568466e+00 1.07860625e+00 5.49205482e-01
6.73147202e-01 -5.67163467e-01 -1.94642305e-01 -1.04948997e+00
-1.99085400e-01 -3.86389017e-01 1.54769599e-01 1.03262568e+00
1.43305296e-02 -8.94691050e-01 7.65697420e-01 3.60705078e-01
-2.22584531e-01 -1.17993736e+00 -9.09141541e-01 -5.23157060e-01
-3.51280332e-01 -3.33429277e-01 6.21120274e-01 6.66299760e-01
-4.60219562e-01 2.44248077e-01 -2.80708373e-01 1.88958362e-01
9.43536222e-01 3.75990808e-01 7.59010434e-01 -1.03983557e+00
-3.81284691e-02 -2.73036063e-01 -1.83684379e-01 -1.46799648e+00
-2.51241207e-01 -3.42405468e-01 -1.80043802e-02 -1.60461581e+00
3.32608938e-01 -5.94214499e-01 8.72999895e-03 -2.93773878e-02
-2.68768847e-01 3.77738863e-01 4.78197783e-02 2.74393290e-01
-2.51552284e-01 5.68892181e-01 1.71410704e+00 3.33992064e-01
-4.01596904e-01 -1.46421596e-01 -5.76737642e-01 8.42472970e-01
6.43514574e-01 -1.58739045e-01 -5.52419186e-01 -7.22653389e-01
1.30411372e-01 3.87676299e-01 3.30870807e-01 -1.03874016e+00
1.79657906e-01 -1.47242472e-01 4.34043586e-01 -9.05484915e-01
6.03113532e-01 -6.65380120e-01 2.67581433e-01 4.19963926e-01
1.60301656e-01 -1.43279612e-01 -2.72911717e-03 3.20236117e-01
-3.23603004e-01 1.35304689e-01 1.10002649e+00 1.26596550e-02
-6.68955922e-01 6.92694843e-01 2.16489345e-01 -1.00920290e-01
1.01940858e+00 -4.02496427e-01 -3.46037656e-01 -3.42076868e-01
-9.95432660e-02 2.86811382e-01 8.12392592e-01 3.01335692e-01
8.32391977e-01 -1.29928946e+00 -5.44153094e-01 5.11822522e-01
5.08569703e-02 3.95450532e-01 6.23909652e-01 9.46256757e-01
-7.38495529e-01 4.05554026e-01 -2.71186888e-01 -8.45937967e-01
-1.05201364e+00 2.63178170e-01 4.36257988e-01 -2.24088520e-01
-5.69971859e-01 7.49198854e-01 8.40717614e-01 -1.07689023e-01
2.08181843e-01 -5.79929173e-01 -9.52846482e-02 -9.82582122e-02
5.97464859e-01 6.02938652e-01 -4.16230969e-02 -4.06636834e-01
-3.09111655e-01 1.12758708e+00 2.05821872e-01 -8.40972364e-02
1.27863038e+00 -4.95496422e-01 -8.23351741e-02 2.30475869e-02
1.34546959e+00 1.15723073e-01 -1.55563056e+00 -2.17596427e-01
-6.00402534e-01 -1.03473294e+00 3.42221260e-01 -3.47652525e-01
-1.33226371e+00 8.37174356e-01 7.37895966e-01 -4.30949092e-01
1.69310427e+00 -1.96578890e-01 1.04060829e+00 -2.47173950e-01
5.33539474e-01 -8.13738406e-01 1.44768983e-01 2.12950543e-01
7.57047534e-01 -1.25144804e+00 2.33822793e-01 -8.56441796e-01
-3.72808903e-01 9.58631098e-01 8.72176528e-01 -1.47845075e-02
4.98073369e-01 1.66075408e-01 5.22202998e-03 -1.75819084e-01
-4.60546821e-01 -6.45008460e-02 3.24009210e-01 3.70581955e-01
2.77623951e-01 -2.66705036e-01 -4.20228422e-01 9.51911360e-02
9.35541168e-02 1.79779395e-01 3.86613697e-01 7.05906987e-01
-3.09682190e-01 -9.74320829e-01 -2.68298924e-01 2.91028410e-01
-1.61266565e-01 -2.11707562e-01 3.74980599e-01 5.53841531e-01
3.09785903e-01 8.41558993e-01 1.04250580e-01 -3.97080541e-01
5.12299240e-01 -5.40257692e-01 5.55702627e-01 -5.50674617e-01
-1.98714152e-01 2.73571074e-01 -2.64737755e-02 -9.70162988e-01
-6.59216642e-01 -4.08794433e-01 -9.63837266e-01 -4.76796091e-01
-3.76044840e-01 -3.79584044e-01 5.33467412e-01 6.17442966e-01
3.90521109e-01 3.28411192e-01 1.00771403e+00 -1.09603095e+00
6.95323795e-02 -6.61465883e-01 -4.14846331e-01 1.13532014e-01
4.46060270e-01 -7.37731338e-01 -4.99834985e-01 -2.07843199e-01] | [9.310364723205566, -2.477752447128296] |
0b29d9d4-8bca-424c-9b97-35ffebb9f991 | emaq-expected-max-q-learning-operator-for | 2007.11091 | null | https://arxiv.org/abs/2007.11091v2 | https://arxiv.org/pdf/2007.11091v2.pdf | EMaQ: Expected-Max Q-Learning Operator for Simple Yet Effective Offline and Online RL | Off-policy reinforcement learning holds the promise of sample-efficient learning of decision-making policies by leveraging past experience. However, in the offline RL setting -- where a fixed collection of interactions are provided and no further interactions are allowed -- it has been shown that standard off-policy RL methods can significantly underperform. Recently proposed methods often aim to address this shortcoming by constraining learned policies to remain close to the given dataset of interactions. In this work, we closely investigate an important simplification of BCQ -- a prior approach for offline RL -- which removes a heuristic design choice and naturally restricts extracted policies to remain exactly within the support of a given behavior policy. Importantly, in contrast to their original theoretical considerations, we derive this simplified algorithm through the introduction of a novel backup operator, Expected-Max Q-Learning (EMaQ), which is more closely related to the resulting practical algorithm. Specifically, in addition to the distribution support, EMaQ explicitly considers the number of samples and the proposal distribution, allowing us to derive new sub-optimality bounds which can serve as a novel measure of complexity for offline RL problems. In the offline RL setting -- the main focus of this work -- EMaQ matches and outperforms prior state-of-the-art in the D4RL benchmarks. In the online RL setting, we demonstrate that EMaQ is competitive with Soft Actor Critic. The key contributions of our empirical findings are demonstrating the importance of careful generative model design for estimating behavior policies, and an intuitive notion of complexity for offline RL problems. With its simple interpretation and fewer moving parts, such as no explicit function approximator representing the policy, EMaQ serves as a strong yet easy to implement baseline for future work. | ['Seyed Kamyar Seyed Ghasemipour', 'Dale Schuurmans', 'Shixiang Shane Gu'] | 2020-07-21 | null | https://openreview.net/forum?id=B8fp0LVMHa | https://openreview.net/pdf?id=B8fp0LVMHa | null | ['d4rl'] | ['robots'] | [-6.59069642e-02 3.21088284e-01 -7.40084410e-01 2.25905352e-03
-8.97299886e-01 -7.58993387e-01 5.34472525e-01 -1.54951811e-02
-7.17617333e-01 1.08261216e+00 1.25377253e-01 -6.69785619e-01
-3.14555645e-01 -4.06066626e-01 -8.91693056e-01 -8.86081457e-01
-2.29115501e-01 4.78600889e-01 -6.99611902e-02 -3.92289221e-01
5.74160889e-02 5.37589252e-01 -1.40096354e+00 -3.46282780e-01
7.17063129e-01 9.34568644e-01 6.39718845e-02 6.56532764e-01
1.98565841e-01 9.20729339e-01 -6.51968539e-01 -1.83238551e-01
5.42737901e-01 -6.73283815e-01 -5.13991058e-01 9.79224294e-02
8.99967626e-02 -7.49515116e-01 -2.21809283e-01 7.78666079e-01
6.93551242e-01 4.93075699e-01 4.30728942e-01 -1.31096017e+00
-2.92036235e-01 6.75605774e-01 -4.08193678e-01 3.18456292e-02
1.57952666e-01 4.74402964e-01 1.24515128e+00 -1.12375975e-01
4.82837915e-01 1.30393887e+00 4.22481239e-01 6.27276361e-01
-1.46117103e+00 -5.52804887e-01 6.24702632e-01 -1.66823700e-01
-8.23205292e-01 -3.68405312e-01 5.85455835e-01 -1.71193630e-01
8.91227663e-01 4.10233587e-02 7.23157406e-01 1.35017920e+00
-2.68623382e-02 1.18688846e+00 1.33960998e+00 -5.07907391e-01
6.13297760e-01 9.15467143e-02 -4.80412357e-02 5.79936445e-01
2.41759464e-01 5.23813307e-01 -1.28759295e-01 -2.85953075e-01
9.37594891e-01 -7.38366544e-02 -1.86049417e-01 -7.38930106e-01
-8.42275739e-01 9.72401083e-01 1.29763884e-02 1.86974593e-02
-2.42485419e-01 5.56637406e-01 4.86873031e-01 5.00575840e-01
3.48487139e-01 5.36869705e-01 -5.66357255e-01 -5.91206372e-01
-6.06921434e-01 6.83368087e-01 1.03667235e+00 9.27563310e-01
5.71728528e-01 1.59244061e-01 -5.72175443e-01 5.84060133e-01
2.39725821e-02 5.42432606e-01 2.51528770e-01 -1.28300726e+00
5.85746288e-01 2.47625168e-02 7.72370338e-01 -5.24967074e-01
-2.72225946e-01 -7.86040366e-01 -2.25333169e-01 4.20461923e-01
8.08192194e-01 -4.82659459e-01 -3.63516659e-01 2.20775247e+00
4.41148132e-01 7.14948550e-02 -3.98277454e-02 7.30436385e-01
-3.98146570e-01 3.34472924e-01 6.97380351e-03 -6.61151469e-01
8.93108726e-01 -8.85469258e-01 -6.67998254e-01 -1.67950969e-02
5.71607947e-01 -2.36735657e-01 1.48165023e+00 5.97502232e-01
-1.26926935e+00 -2.97654539e-01 -9.29138899e-01 2.45754555e-01
7.65605941e-02 8.73377696e-02 8.69362235e-01 6.51918173e-01
-1.10906088e+00 9.13802803e-01 -8.80340159e-01 -2.46164575e-01
2.97068655e-01 4.17926610e-01 1.80848777e-01 2.11105064e-01
-8.31418991e-01 8.26713681e-01 2.48441726e-01 -1.22490965e-01
-1.16378438e+00 -6.71169579e-01 -5.56920528e-01 1.09633468e-01
1.24308062e+00 -5.49833715e-01 2.05168462e+00 -1.25402164e+00
-2.26690674e+00 2.32591733e-01 -4.48166616e-02 -7.18280435e-01
9.67640519e-01 -3.34531039e-01 -3.55433375e-02 1.48888767e-01
-1.07624613e-01 2.70669699e-01 1.00759006e+00 -1.27530992e+00
-7.58113742e-01 -1.49501964e-01 6.85696721e-01 2.81492203e-01
-1.79484442e-01 -3.15372884e-01 -3.83985907e-01 -6.17857158e-01
-7.59775579e-01 -1.10731339e+00 -5.17007649e-01 -2.54652619e-01
-8.52845684e-02 -4.89092469e-01 4.12906140e-01 -2.37039447e-01
1.24316907e+00 -2.03843617e+00 1.11489296e-01 1.81393221e-01
2.63146609e-02 2.25664333e-01 -1.41069785e-01 5.95551312e-01
7.93614611e-02 4.23832657e-03 -1.89790335e-02 -5.12289166e-01
5.76557696e-01 4.77982134e-01 -7.53049135e-01 6.18302584e-01
-7.60852620e-02 1.04059136e+00 -1.25338650e+00 -2.03542739e-01
3.15883458e-01 3.08177099e-02 -7.14389145e-01 2.67716199e-01
-7.53035486e-01 4.79945153e-01 -6.94000483e-01 2.21802279e-01
3.40732336e-01 -2.09443182e-01 4.91140723e-01 3.39768231e-01
-6.58973604e-02 3.77087861e-01 -1.22398901e+00 1.70630682e+00
-6.34016693e-01 1.19666681e-01 2.26928070e-01 -1.04212785e+00
3.79856229e-01 1.57794222e-01 6.30984843e-01 -6.90469146e-01
2.38436073e-01 7.26768672e-02 -5.96691892e-02 -2.98208833e-01
3.24936181e-01 -8.96700919e-02 -1.96857765e-01 6.66617513e-01
5.62375449e-02 8.97714645e-02 2.81210810e-01 -8.25822167e-03
9.96393144e-01 7.38179982e-01 5.18373430e-01 -2.03354269e-01
1.96103483e-01 -1.76477224e-01 6.02131426e-01 1.19416976e+00
-2.79430240e-01 -7.12989941e-02 9.17017281e-01 -1.23949600e-02
-8.84299636e-01 -9.12751496e-01 1.23570569e-01 1.33247316e+00
-1.47411406e-01 -2.88545668e-01 -6.91683769e-01 -1.05741727e+00
4.12457675e-01 9.53597724e-01 -6.90554440e-01 2.55302005e-02
-6.43929660e-01 -4.69001263e-01 3.42110187e-01 4.94395167e-01
1.19753905e-01 -9.59989429e-01 -7.83822417e-01 4.42287892e-01
2.73457557e-01 -9.74170387e-01 -5.66556931e-01 3.46971571e-01
-8.31523538e-01 -1.00874865e+00 -6.63037658e-01 -6.97995946e-02
3.78563046e-01 1.77792594e-01 1.05750442e+00 -2.16251299e-01
5.68493716e-02 9.74382401e-01 -2.68111020e-01 -3.82333279e-01
-4.04636413e-01 2.87186205e-02 6.20832145e-02 3.13628907e-03
-1.67192087e-01 -6.98488057e-01 -7.21496761e-01 1.02258280e-01
-7.22598076e-01 -1.17996432e-01 6.22164369e-01 7.78182507e-01
4.92097527e-01 -1.65517375e-01 6.14936352e-01 -9.21685576e-01
7.95177579e-01 -5.27548432e-01 -1.06656241e+00 2.29022101e-01
-8.33230257e-01 5.64516544e-01 1.26268864e+00 -7.13237226e-01
-9.42966938e-01 -1.61079958e-01 1.22694232e-01 -6.05143905e-01
6.08283654e-02 2.51847901e-04 -1.09862037e-01 1.43749923e-01
3.50395232e-01 1.22561887e-01 2.27223396e-01 -4.23097551e-01
5.62570810e-01 2.52352953e-01 2.42899984e-01 -1.28199410e+00
6.26117527e-01 4.69755292e-01 2.20475376e-01 -5.04515886e-01
-1.05299354e+00 -2.97217935e-01 -4.83714826e-02 -7.73692429e-02
3.55180591e-01 -7.82783568e-01 -1.47356880e+00 -4.83892821e-02
-6.46707058e-01 -1.06875920e+00 -7.62255490e-01 4.68468875e-01
-1.14120162e+00 4.03309554e-01 -5.26985466e-01 -1.42621887e+00
-8.45149800e-04 -1.07078350e+00 8.76542330e-01 6.49491325e-02
5.20841070e-02 -1.10753238e+00 2.01752499e-01 -2.72962060e-02
1.90575197e-01 2.06370026e-01 7.83134043e-01 -5.86798131e-01
-4.44453895e-01 3.29136401e-01 1.61728323e-01 4.79975015e-01
-1.01036675e-01 -2.89345980e-01 -8.78049672e-01 -7.41931975e-01
-3.22992057e-02 -5.68735301e-01 7.33585417e-01 4.80616271e-01
1.29053843e+00 -4.70883340e-01 -5.43131828e-02 4.55910325e-01
1.42429256e+00 3.29337418e-01 2.02896684e-01 3.05535316e-01
2.90449202e-01 3.51877570e-01 8.91690254e-01 1.00748873e+00
2.81493187e-01 8.74494553e-01 3.76333058e-01 6.89494088e-02
2.81369150e-01 -6.91098452e-01 7.52289534e-01 3.30311567e-01
-7.39530548e-02 -3.09956372e-01 -1.94184557e-01 2.43816301e-01
-2.24026752e+00 -8.59703600e-01 5.82213938e-01 2.66485310e+00
9.66798961e-01 1.75927818e-01 7.14315474e-01 -1.28664777e-01
1.68269828e-01 7.64792934e-02 -1.03958666e+00 -4.56952691e-01
1.72628090e-01 6.43961489e-01 9.24641669e-01 7.58139372e-01
-8.86153936e-01 8.38342726e-01 6.21635962e+00 1.02823281e+00
-8.69959414e-01 3.03017288e-01 4.15848702e-01 -4.22811866e-01
-2.18590453e-01 2.76835918e-01 -1.05236113e+00 4.61732805e-01
1.05194116e+00 -3.50727141e-02 9.80317295e-01 9.51673687e-01
6.19355321e-01 -2.51616001e-01 -1.44834793e+00 7.16184139e-01
-2.45811090e-01 -9.17911232e-01 -3.65329176e-01 3.88390481e-01
6.00062490e-01 -2.67346859e-01 1.96687579e-01 8.43374491e-01
8.38247895e-01 -9.04565573e-01 9.82252598e-01 3.70640129e-01
7.47593641e-01 -9.95462954e-01 2.94317991e-01 5.30508161e-01
-9.51342762e-01 -5.90522766e-01 -2.40124330e-01 -1.98570445e-01
-7.66117349e-02 2.10351065e-01 -6.36739969e-01 5.31000078e-01
1.85523316e-01 6.08324528e-01 -1.84065439e-02 6.03312254e-01
-4.45364356e-01 8.06168497e-01 -3.58014315e-01 -1.78573698e-01
5.34061849e-01 -2.06399038e-01 4.16896790e-01 1.08996201e+00
2.54041012e-02 -3.19892652e-02 6.23690426e-01 8.66299927e-01
9.63767320e-02 3.18015888e-02 -3.76840323e-01 -1.80376306e-01
3.37590665e-01 1.03544259e+00 -3.42318684e-01 -1.66020900e-01
-2.96233624e-01 7.11431742e-01 6.97374761e-01 5.96181095e-01
-1.02255511e+00 -2.34256946e-02 7.60253370e-01 6.22074492e-02
5.59554756e-01 -4.65862244e-01 1.69901431e-01 -1.07309508e+00
3.49202380e-02 -1.14120162e+00 2.59258419e-01 -1.49086758e-01
-1.12049055e+00 -1.65024966e-01 3.55668247e-01 -1.07525218e+00
-7.14928448e-01 -5.87668180e-01 -3.08926731e-01 6.06210589e-01
-1.74452889e+00 -8.02906096e-01 3.90374839e-01 5.64421475e-01
5.70901930e-01 1.15721039e-01 5.34604371e-01 -7.31654689e-02
-7.10433960e-01 7.43173599e-01 5.10056317e-01 -3.19318175e-01
5.78945458e-01 -1.57355452e+00 1.77936093e-03 5.58112919e-01
6.99134022e-02 5.03321946e-01 7.84501314e-01 -3.90719742e-01
-1.84477150e+00 -8.28791857e-01 1.17654599e-01 -2.93623537e-01
8.47453535e-01 -4.93961990e-01 -4.65063095e-01 7.53839016e-01
-7.50430115e-03 5.82585111e-03 3.93999070e-01 1.85860664e-01
4.45245579e-02 -3.07187140e-01 -8.40840101e-01 8.84931266e-01
1.20786071e+00 -2.94742465e-01 -2.74244666e-01 2.81040579e-01
8.07705522e-01 -4.56837267e-01 -7.28182614e-01 4.87034135e-02
5.23403525e-01 -9.27476764e-01 8.40508342e-01 -7.71236718e-01
-2.92519238e-02 -4.89702411e-02 5.41904196e-02 -1.24566960e+00
-5.60542122e-02 -1.45147896e+00 -5.68748116e-01 1.02346814e+00
1.13461845e-01 -7.90388763e-01 5.99848092e-01 4.05043989e-01
4.76504043e-02 -1.09128273e+00 -7.39186287e-01 -1.11320853e+00
2.01435119e-01 -5.73397160e-01 4.08261269e-01 4.42449868e-01
-4.51167114e-02 1.50566086e-01 -6.08417809e-01 -3.45243216e-02
6.76292539e-01 9.36569795e-02 1.03686309e+00 -7.62228727e-01
-1.20979095e+00 -5.83048165e-01 3.92293155e-01 -1.63626707e+00
6.05754972e-01 -6.23700142e-01 -4.20382209e-02 -1.02601802e+00
-3.44035972e-04 -6.34019315e-01 -5.11290967e-01 4.29728597e-01
2.38280967e-02 -5.97400784e-01 4.33873564e-01 3.76581661e-02
-8.65778089e-01 9.34079230e-01 1.53859937e+00 3.68006796e-01
-4.92473364e-01 4.44849491e-01 -7.45694637e-01 4.83199060e-01
7.69391537e-01 -3.57882589e-01 -9.05774653e-01 5.85870072e-02
2.57909834e-01 4.90186006e-01 3.44368845e-01 -5.63704848e-01
-1.28035814e-01 -5.84295213e-01 -8.71651769e-02 -2.31077269e-01
2.51226813e-01 -8.15793991e-01 -3.37312579e-01 4.44093078e-01
-5.94454944e-01 -2.64418554e-02 1.92687675e-01 9.64038074e-01
4.38621163e-01 -2.51555830e-01 4.97718751e-01 -8.86385217e-02
-3.63813430e-01 3.11413974e-01 -3.31635833e-01 4.75292742e-01
1.07457709e+00 2.47370768e-02 -1.10499153e-03 -6.68011546e-01
-5.41483760e-01 4.59348053e-01 4.29864138e-01 6.85261413e-02
1.15861647e-01 -9.59125876e-01 -3.71861070e-01 -2.07374040e-02
-2.57899255e-01 -2.14612573e-01 -2.63952482e-02 8.84521723e-01
1.15478195e-01 5.03220975e-01 1.10920452e-01 -2.22864583e-01
-6.54065609e-01 7.18740225e-01 4.12346244e-01 -7.92005301e-01
-6.98667526e-01 2.64532477e-01 1.75273463e-01 -1.58757627e-01
4.90852207e-01 -7.35851765e-01 1.68994144e-01 -2.33025596e-01
2.31037244e-01 3.85220140e-01 -2.66739905e-01 1.64128363e-01
1.11061357e-01 1.60960481e-01 -4.70755436e-02 -5.06572366e-01
1.21967673e+00 -6.23691529e-02 4.23971713e-01 6.64782643e-01
8.14963281e-01 1.95305333e-01 -2.07183456e+00 -2.74810970e-01
-1.23323891e-02 -3.36916447e-01 -5.90955317e-02 -8.81991148e-01
-7.12749541e-01 6.26156092e-01 3.70604426e-01 5.18788099e-02
9.33507085e-01 -2.47284591e-01 5.65750957e-01 4.58211601e-01
6.16016209e-01 -1.34918725e+00 9.90659967e-02 3.98356080e-01
6.29341006e-01 -1.06587875e+00 4.07966897e-02 2.07095042e-01
-6.16763830e-01 8.88057649e-01 4.21377987e-01 -4.08761919e-01
2.16554880e-01 3.25785428e-01 -4.48222429e-01 3.56923461e-01
-9.46267664e-01 -3.81988615e-01 -6.64065257e-02 3.55825990e-01
1.39350787e-01 1.47030130e-01 -2.89650857e-01 6.20575011e-01
-9.08642858e-02 2.02395424e-01 2.59186000e-01 1.03606987e+00
-3.21259469e-01 -1.48172343e+00 -8.96979049e-02 2.31718197e-01
-5.47423959e-01 2.11842418e-01 -2.16537695e-02 1.18841755e+00
-1.95025638e-01 9.63403761e-01 -5.23159839e-02 1.48305640e-01
2.56262928e-01 -8.08674619e-02 9.13098693e-01 -4.42456603e-01
-6.66235805e-01 3.21580082e-01 7.57900923e-02 -9.54069316e-01
-3.32196683e-01 -6.46993101e-01 -1.14432693e+00 -2.31610596e-01
-1.91271201e-01 2.29735985e-01 4.18847591e-01 1.13992560e+00
3.22416216e-01 4.08385366e-01 7.95015275e-01 -7.80509174e-01
-1.64751410e+00 -7.06354439e-01 -7.58058548e-01 2.63749450e-01
6.73731923e-01 -8.92618716e-01 -4.59262520e-01 -5.15370131e-01] | [4.10744047164917, 2.2327263355255127] |
2949ed86-4c9c-4b4f-9323-6bdef96c23cd | abstractive-sentence-summarization-with | null | null | https://aclanthology.org/N16-1012 | https://aclanthology.org/N16-1012.pdf | Abstractive Sentence Summarization with Attentive Recurrent Neural Networks | null | ['er M.', 'Michael Auli', 'Alex Rush', 'Sumit Chopra'] | 2016-06-01 | null | null | null | naacl-2016-6 | ['summarization', 'abstractive-sentence-summarization'] | ['natural-language-processing', '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] | [-7.354478359222412, 3.6406850814819336] |
26f4693a-ea68-4cdd-a242-1e62786c22cf | contrastive-principal-component-analysis | 1709.06716 | null | http://arxiv.org/abs/1709.06716v2 | http://arxiv.org/pdf/1709.06716v2.pdf | Contrastive Principal Component Analysis | We present a new technique called contrastive principal component analysis
(cPCA) that is designed to discover low-dimensional structure that is unique to
a dataset, or enriched in one dataset relative to other data. The technique is
a generalization of standard PCA, for the setting where multiple datasets are
available -- e.g. a treatment and a control group, or a mixed versus a
homogeneous population -- and the goal is to explore patterns that are specific
to one of the datasets. We conduct a wide variety of experiments in which cPCA
identifies important dataset-specific patterns that are missed by PCA,
demonstrating that it is useful for many applications: subgroup discovery,
visualizing trends, feature selection, denoising, and data-dependent
standardization. We provide geometrical interpretations of cPCA and show that
it satisfies desirable theoretical guarantees. We also extend cPCA to nonlinear
settings in the form of kernel cPCA. We have released our code as a python
package and documentation is on Github. | ['Martin J. Zhang', 'Abubakar Abid', 'James Zou', 'Vivek K. Bagaria'] | 2017-09-20 | null | null | null | null | ['subgroup-discovery'] | ['methodology'] | [ 4.35183555e-01 -2.33373135e-01 -3.17800134e-01 -3.12784284e-01
-8.66378844e-01 -6.17395699e-01 4.59116757e-01 -2.03753728e-03
5.52862138e-02 3.09861094e-01 7.45580316e-01 -1.60172701e-01
-8.29920411e-01 -2.40093663e-01 -3.46924305e-01 -1.15873861e+00
-6.05556250e-01 4.82036293e-01 -2.08395630e-01 1.91874012e-01
1.58321112e-01 5.96017718e-01 -1.28296196e+00 1.17136255e-01
3.58534276e-01 5.20251155e-01 -1.59607336e-01 5.54298759e-01
1.53818727e-01 8.52877721e-02 -4.77195270e-02 -1.70692950e-01
5.81952155e-01 -3.18116397e-01 -4.70875293e-01 2.66849577e-01
4.34562445e-01 3.89112830e-01 -5.38354553e-02 1.10921347e+00
5.10430396e-01 -6.64253905e-03 7.69927919e-01 -1.54007423e+00
-4.95963842e-01 3.80221307e-01 -9.85042155e-01 1.81706876e-01
1.36575699e-01 6.25367090e-02 8.93664896e-01 -8.34029138e-01
4.97599840e-01 1.46096158e+00 8.54482353e-01 4.07368481e-01
-1.89219141e+00 -5.55418313e-01 3.52196619e-02 -1.45049348e-01
-1.08885550e+00 -6.36235893e-01 9.97126997e-01 -8.77439022e-01
3.58342618e-01 6.21659517e-01 3.29416484e-01 1.26149333e+00
1.29728496e-01 7.41158724e-01 1.28209996e+00 -2.81505376e-01
3.91026467e-01 -1.67489693e-01 9.20603514e-01 2.45432973e-01
4.28110600e-01 1.85796231e-01 -5.12834191e-01 -1.08438659e+00
3.34480435e-01 -3.37807019e-03 -4.03991967e-01 -9.35435653e-01
-1.44303954e+00 9.51899469e-01 -2.53743529e-01 4.95186821e-02
-6.71951354e-01 -3.43759298e-01 3.79924238e-01 2.74465799e-01
5.49522460e-01 5.57776451e-01 -7.50563025e-01 -1.54567540e-01
-7.10950673e-01 3.14483285e-01 6.95043921e-01 5.24695098e-01
5.40999293e-01 -3.38925421e-01 5.46957329e-02 7.47890711e-01
4.47016731e-02 3.68903726e-01 8.36221516e-01 -1.10261977e+00
5.44693619e-02 4.30052847e-01 -1.09288126e-01 -1.11948681e+00
-7.96826124e-01 -2.55218983e-01 -9.81043279e-01 2.99132019e-02
3.37380379e-01 -2.62255341e-01 -6.78248346e-01 1.98583424e+00
4.08976644e-01 4.57785696e-01 1.60902143e-01 6.28344893e-01
4.38209802e-01 2.26543307e-01 8.65502283e-02 -6.30823851e-01
1.37402093e+00 -5.41534126e-01 -7.17090428e-01 -2.72573084e-01
5.60253799e-01 -5.40365875e-01 1.06681371e+00 4.37639207e-01
-7.56750762e-01 -9.02652740e-02 -6.71634912e-01 1.58910438e-01
-2.02308729e-01 1.32833108e-01 9.74410534e-01 8.33051383e-01
-9.28101897e-01 6.14086092e-01 -1.02538419e+00 -5.94396412e-01
3.43426138e-01 4.08609986e-01 -7.72597492e-01 -1.85651965e-02
-5.26068866e-01 3.15960914e-01 -2.69253161e-02 -1.04722686e-01
-3.81381482e-01 -1.25606453e+00 -7.45994747e-01 9.43967849e-02
1.01891972e-01 -7.49211848e-01 8.70960772e-01 -9.98880446e-01
-1.02941227e+00 7.45753467e-01 -4.15333658e-01 -6.92684948e-02
2.10848853e-01 6.73984215e-02 -5.00530005e-01 -8.37220401e-02
3.29425782e-01 1.66971669e-01 8.71532977e-01 -9.67414558e-01
-5.21576524e-01 -8.49205613e-01 -6.34723067e-01 9.95654836e-02
-3.27564508e-01 3.48775595e-01 -4.43577677e-01 -8.64927471e-01
3.74365509e-01 -1.26556611e+00 -4.67850238e-01 -2.27277488e-01
-4.80979174e-01 1.20849265e-02 9.85976577e-01 -7.64277577e-01
1.13200998e+00 -2.49397111e+00 5.91697931e-01 5.76641262e-01
3.76664996e-01 -3.33583355e-01 -2.86233783e-01 2.27656066e-01
-8.54619861e-01 1.74371913e-01 -7.32982218e-01 -3.14284623e-01
-1.35012418e-01 -4.41669412e-02 -1.84058383e-01 9.70231056e-01
1.92413718e-01 5.11043847e-01 -5.47373474e-01 -4.57370728e-02
-2.40664274e-01 3.68260354e-01 -4.90260184e-01 -1.35006756e-01
4.88283902e-01 3.89982879e-01 -3.39847714e-01 7.15863287e-01
9.99348760e-01 -1.46700099e-01 3.64366233e-01 -1.46347746e-01
-2.33392771e-02 -1.46614343e-01 -1.48202419e+00 1.35018241e+00
1.62231505e-01 6.34521067e-01 3.28081250e-01 -9.95997965e-01
5.55100083e-01 1.36780053e-01 8.29653680e-01 1.24223428e-02
-9.47855264e-02 -8.69000033e-02 1.69837803e-01 -4.25572693e-01
1.65890381e-02 -1.08814314e-01 -6.45691603e-02 6.53589606e-01
5.35153374e-02 3.49136829e-01 1.59333929e-01 1.49241490e-02
1.21524262e+00 -4.01183158e-01 5.87746084e-01 -7.35363722e-01
1.28359154e-01 -4.25136834e-02 9.11097050e-01 7.44624555e-01
-2.77666628e-01 6.45656407e-01 9.19753790e-01 -2.30300158e-01
-8.27481627e-01 -1.05606306e+00 -5.82022667e-01 8.95021558e-01
-4.50739115e-01 -5.17381847e-01 -3.39588702e-01 -4.33733195e-01
3.10763448e-01 2.38326281e-01 -9.48533654e-01 -1.88853115e-01
-2.81480372e-01 -1.73750412e+00 1.96629792e-01 4.41356957e-01
3.29106748e-02 -5.30904412e-01 -1.44531757e-01 -6.46968856e-02
2.62023807e-01 -8.14000368e-01 -5.18161952e-01 2.16583177e-01
-1.08251023e+00 -1.45074272e+00 -6.62737846e-01 -7.02070713e-01
6.34803951e-01 4.29133445e-01 8.18846345e-01 -5.82012713e-01
-1.34262830e-01 7.58594453e-01 -7.57323205e-02 -4.42334533e-01
-1.36936650e-01 -1.69722244e-01 5.70184290e-01 2.31010944e-01
7.80453503e-01 -7.93799281e-01 -2.55918711e-01 2.10019678e-01
-7.39975572e-01 -1.90610036e-01 5.76290071e-01 8.89475942e-01
6.56244874e-01 3.57094735e-01 2.09119976e-01 -1.03902090e+00
1.06281519e+00 -9.19100702e-01 -5.60039341e-01 1.96506962e-01
-6.37973607e-01 -6.03410117e-02 1.35801449e-01 -6.39995158e-01
-6.63880825e-01 4.44326192e-01 3.92764330e-01 -4.99759048e-01
-2.60265827e-01 8.35156739e-01 -3.90489727e-01 3.22555029e-03
7.43362904e-01 1.32007912e-01 4.67641026e-01 -8.57347369e-01
1.41848758e-01 4.44507867e-01 4.39508796e-01 -5.24914563e-01
5.85912704e-01 5.34666657e-01 3.13042402e-01 -1.06897724e+00
-1.82552993e-01 -8.62723708e-01 -7.95134246e-01 4.47651297e-01
5.21342695e-01 -9.93877649e-01 -6.71849608e-01 3.71799886e-01
-5.43050587e-01 -2.43969232e-01 -2.27923989e-01 7.89283931e-01
-4.84398276e-01 4.88250256e-01 -2.71069497e-01 -5.41721582e-01
-2.32128263e-01 -1.22975600e+00 8.91482472e-01 2.90930737e-02
-2.75034159e-01 -9.73683417e-01 6.05980217e-01 -3.91625538e-02
1.14288688e-01 2.80629903e-01 1.09523916e+00 -1.08558178e+00
2.53896415e-01 -3.58755141e-01 4.65713143e-02 1.28121704e-01
4.34495300e-01 1.28239676e-01 -1.00632918e+00 -4.05914783e-01
1.72814444e-01 4.16841388e-01 7.52395689e-01 9.76926923e-01
1.36456954e+00 -5.04646719e-01 -6.00084245e-01 7.05324113e-01
9.16811824e-01 2.22466588e-01 3.54983419e-01 1.35635450e-01
5.16195834e-01 9.41488743e-01 -1.28206955e-02 2.61158735e-01
1.15930520e-01 7.75004327e-01 -3.02839875e-01 -1.54629499e-01
2.03764588e-01 3.13853860e-01 3.45915347e-01 7.09509194e-01
5.19348271e-02 4.15954590e-01 -9.18400586e-01 5.74106812e-01
-1.97149408e+00 -1.03996634e+00 -3.66986871e-01 2.41477728e+00
6.14380300e-01 -2.31243879e-01 6.22361720e-01 2.51827259e-02
7.22522318e-01 -1.31976321e-01 -5.57932556e-01 -9.91632715e-02
-4.60213751e-01 1.69072643e-01 5.66509545e-01 3.45372885e-01
-1.45476091e+00 3.25887978e-01 7.46736288e+00 4.84577179e-01
-9.47980702e-01 -9.60412323e-02 6.57616198e-01 -4.83394749e-02
-7.52268806e-02 1.65618751e-02 -3.79509896e-01 3.50609809e-01
1.08760595e+00 -4.57342505e-01 3.64628613e-01 1.06127262e+00
4.50838298e-01 1.31258592e-01 -1.13808882e+00 1.07038271e+00
4.84909378e-02 -1.11444998e+00 -3.79042476e-01 5.47439098e-01
4.71670002e-01 -8.29860941e-02 2.23461866e-01 -2.00975742e-02
2.55284756e-01 -6.97644055e-01 5.02444208e-02 2.69356698e-01
4.20912921e-01 -6.14665687e-01 4.18213755e-01 -1.16887465e-01
-8.21713328e-01 -2.14707941e-01 -4.05373901e-01 1.39808983e-01
-2.04003200e-01 8.04039419e-01 -5.57071924e-01 3.91318560e-01
8.72414708e-01 1.07469058e+00 -8.52235913e-01 1.32694292e+00
2.36506000e-01 8.15759301e-01 -3.00965399e-01 6.43401384e-01
-3.62641007e-01 -5.05161285e-01 9.73573506e-01 1.12501216e+00
3.48078817e-01 1.63309172e-01 8.80372152e-02 5.27281404e-01
2.26717830e-01 2.67237335e-01 -7.05753088e-01 -3.15005258e-02
3.30563188e-01 1.40341914e+00 -5.75223625e-01 -7.71043748e-02
-5.92867017e-01 8.50431144e-01 -1.03680350e-01 3.98447633e-01
-4.32097726e-02 -2.95100927e-01 1.31554735e+00 -1.06346555e-01
1.46555632e-01 -7.73406774e-02 -3.16152871e-01 -1.39087033e+00
-1.03029877e-01 -1.21785522e+00 8.81915331e-01 -5.70532024e-01
-1.77053928e+00 6.82724416e-02 2.96778440e-01 -1.09326625e+00
-1.12713233e-01 -8.50950420e-01 -5.66848934e-01 8.32092226e-01
-9.47627723e-01 -9.08726275e-01 4.62114513e-02 7.04793036e-01
2.34220207e-01 -3.63278210e-01 9.97582078e-01 1.97893381e-01
-1.00311828e+00 3.32116425e-01 5.31527162e-01 -4.24130149e-02
8.16980779e-01 -1.37284970e+00 5.84200174e-02 1.05133295e+00
3.92028838e-02 9.76417005e-01 8.15729737e-01 -6.31161034e-01
-1.60689497e+00 -8.64786506e-01 5.35668492e-01 -4.18790430e-01
8.28306317e-01 -3.52711588e-01 -1.01732802e+00 1.07795751e+00
-1.40960008e-01 -1.45440206e-01 1.49599969e+00 7.02146173e-01
-3.48282635e-01 -2.61559729e-02 -1.13902044e+00 6.47249877e-01
8.38213682e-01 -3.51497501e-01 -3.98424268e-01 3.85876358e-01
3.87002826e-01 -6.73940927e-02 -1.01880503e+00 1.93195730e-01
6.44859314e-01 -6.16154015e-01 1.09324908e+00 -1.27496958e+00
1.91189777e-02 -3.34102273e-01 -3.13910514e-01 -1.52871192e+00
-8.22323680e-01 -7.51489282e-01 -7.84083158e-02 1.00506401e+00
3.91797900e-01 -6.65723860e-01 6.96940064e-01 1.04877031e+00
-1.12061054e-01 -1.55715778e-01 -9.48161066e-01 -6.81613982e-01
1.59755155e-01 -3.95142734e-01 6.97571099e-01 1.48256850e+00
2.70914346e-01 1.90555871e-01 -3.22118342e-01 3.96807581e-01
9.44141150e-01 1.48225382e-01 7.81050146e-01 -1.70756447e+00
-3.62748563e-01 -6.20073736e-01 -6.32676363e-01 -3.06637406e-01
2.15044141e-01 -8.60528111e-01 -2.79952228e-01 -8.62314343e-01
5.07697701e-01 -3.58633041e-01 -2.38299742e-01 6.19592488e-01
-3.38944912e-01 -8.33422244e-02 -2.59416434e-03 4.50221092e-01
8.28402638e-02 1.31997317e-01 6.48006737e-01 -1.04251921e-01
-7.52014220e-01 2.38854468e-01 -1.37496841e+00 5.72629213e-01
7.40092516e-01 -5.07771611e-01 -3.78599435e-01 1.57335505e-01
-7.99827054e-02 -1.36653006e-01 3.14621359e-01 -7.27510571e-01
-1.69198606e-02 -3.06118816e-01 5.93996525e-01 -1.35038272e-01
1.30927488e-01 -7.97648370e-01 5.67688048e-01 2.04448715e-01
-4.34377819e-01 2.04077154e-01 3.11613590e-01 7.06250846e-01
9.92481261e-02 4.71291989e-02 4.75195855e-01 1.54669166e-01
-6.29939795e-01 3.06331545e-01 -3.41601610e-01 -2.16091618e-01
9.24190283e-01 4.21954840e-02 -4.79347020e-01 -2.91548818e-01
-9.77243304e-01 7.63844401e-02 3.85808051e-01 2.33312726e-01
3.73496592e-01 -1.41557777e+00 -8.57045710e-01 4.59422141e-01
2.87880987e-01 -5.05166650e-01 2.74482876e-01 1.28497458e+00
8.28060731e-02 3.46275866e-01 -1.75966337e-01 -4.97358084e-01
-1.55504978e+00 8.97624552e-01 1.43063039e-01 4.34244126e-02
-8.12614918e-01 4.32492435e-01 5.69703877e-01 -4.56516564e-01
-1.04468308e-01 -1.35121435e-01 -1.78457811e-01 3.35139960e-01
9.46125090e-01 4.45936173e-01 4.69236001e-02 -7.00417280e-01
-4.98203039e-01 5.55538118e-01 9.85948890e-02 -7.81347975e-02
1.65322924e+00 -6.91014379e-02 -4.78542179e-01 4.76207495e-01
1.35504496e+00 2.85760313e-01 -1.06211329e+00 -1.57407865e-01
2.61545956e-01 -4.87670541e-01 2.82136887e-01 -3.83860052e-01
-1.15019619e+00 3.56306881e-01 8.04101110e-01 2.39164144e-01
1.36831605e+00 1.66406676e-01 -2.09003896e-01 2.26662174e-01
-2.35726878e-01 -6.78786993e-01 -4.75248754e-01 1.78287961e-02
1.09614420e+00 -1.13060188e+00 2.26389453e-01 -3.72771561e-01
-7.45528162e-01 1.24698031e+00 -3.66583243e-02 -1.20407648e-01
1.13587129e+00 1.67660639e-01 -1.35343283e-01 -4.14842755e-01
-7.02248335e-01 -2.99535450e-02 4.91724283e-01 9.38304901e-01
2.34158069e-01 2.38494888e-01 -4.03575599e-01 1.07160151e+00
-1.14689551e-01 -1.17972672e-01 6.89296961e-01 9.31026697e-01
1.43302143e-01 -1.15008748e+00 -6.94713533e-01 9.07085121e-01
-4.03339237e-01 -2.86425762e-02 -5.11877894e-01 7.62622774e-01
-3.43887806e-01 7.31040239e-01 2.68343061e-01 -3.35934877e-01
2.56633222e-01 4.31931973e-01 1.20468169e-01 -3.79394829e-01
-2.73895025e-01 6.32908940e-01 -3.67377549e-02 -7.49701977e-01
-6.80862904e-01 -1.41994739e+00 -5.17895818e-01 -3.43373388e-01
-1.79738842e-03 1.73513651e-01 5.23561239e-01 5.10034621e-01
7.18906164e-01 7.74758011e-02 5.36234319e-01 -6.70236230e-01
-1.61031291e-01 -7.95528173e-01 -9.94991839e-01 4.31009769e-01
6.54063165e-01 -7.32800066e-01 -6.33379221e-01 2.05706000e-01] | [7.440646648406982, 4.618546009063721] |
4614bc3a-7101-4fd1-abbb-2e83bae8f395 | learning-what-and-where-unsupervised | 2205.13349 | null | https://arxiv.org/abs/2205.13349v4 | https://arxiv.org/pdf/2205.13349v4.pdf | Learning What and Where: Disentangling Location and Identity Tracking Without Supervision | Our brain can almost effortlessly decompose visual data streams into background and salient objects. Moreover, it can anticipate object motion and interactions, which are crucial abilities for conceptual planning and reasoning. Recent object reasoning datasets, such as CATER, have revealed fundamental shortcomings of current vision-based AI systems, particularly when targeting explicit object representations, object permanence, and object reasoning. Here we introduce a self-supervised LOCation and Identity tracking system (Loci), which excels on the CATER tracking challenge. Inspired by the dorsal and ventral pathways in the brain, Loci tackles the binding problem by processing separate, slot-wise encodings of `what' and `where'. Loci's predictive coding-like processing encourages active error minimization, such that individual slots tend to encode individual objects. Interactions between objects and object dynamics are processed in the disentangled latent space. Truncated backpropagation through time combined with forward eligibility accumulation significantly speeds up learning and improves memory efficiency. Besides exhibiting superior performance in current benchmarks, Loci effectively extracts objects from video streams and separates them into location and Gestalt components. We believe that this separation offers a representation that will facilitate effective planning and reasoning on conceptual levels. | ['Martin V. Butz', 'Jannik Thümmel', 'Matthias Karlbauer', 'Tobias Menge', 'Sebastian Otte', 'Manuel Traub'] | 2022-05-26 | null | null | null | null | ['video-object-tracking'] | ['computer-vision'] | [ 1.78233892e-01 5.25157666e-03 -4.94112998e-01 -2.04825908e-01
-1.01038657e-01 -5.62900126e-01 8.27705204e-01 2.55363345e-01
-5.68486631e-01 1.88780740e-01 5.34227192e-01 -5.76444454e-02
-4.25558805e-01 -6.48805499e-01 -7.74106920e-01 -7.11930394e-01
-1.63211510e-01 6.11192107e-01 3.76076430e-01 1.10633075e-01
4.33261305e-01 5.08924663e-01 -1.95042169e+00 6.92747355e-01
4.62239832e-01 1.15081298e+00 5.04799247e-01 7.17063010e-01
-3.62026870e-01 1.31947613e+00 -4.12857682e-02 -2.94934183e-01
2.64749944e-01 -1.73701912e-01 -9.48162615e-01 -9.90565773e-03
6.56605661e-01 -3.95806432e-01 -6.36740804e-01 8.09459388e-01
1.13987938e-01 2.86195815e-01 5.78557312e-01 -1.42130196e+00
-1.19271243e+00 7.34813333e-01 -3.86935979e-01 5.19920111e-01
9.63505358e-02 6.60834968e-01 1.33334196e+00 -8.76397491e-01
7.36281276e-01 1.38820171e+00 4.30893987e-01 7.77900815e-01
-1.42028761e+00 -3.91618609e-01 7.70557404e-01 5.45911014e-01
-1.00758791e+00 -6.53561175e-01 7.29451120e-01 -6.80398285e-01
1.38386500e+00 3.16022396e-01 1.21758878e+00 1.18322647e+00
1.07125193e-01 1.60075176e+00 6.90771461e-01 1.80763248e-02
3.74252766e-01 -5.31866491e-01 2.85351098e-01 7.30367005e-01
2.11958468e-01 1.50330141e-01 -1.04389930e+00 2.94093519e-01
9.77706075e-01 3.10198039e-01 -1.87440023e-01 -5.98615110e-01
-1.93285775e+00 4.13862646e-01 7.42599428e-01 -1.35342525e-02
-6.68843567e-01 6.18912220e-01 2.20251024e-01 -1.11665249e-01
-1.31804213e-01 4.49493080e-01 -2.53275394e-01 -1.27626568e-01
-9.71434236e-01 4.41815078e-01 3.96192610e-01 1.22913516e+00
4.67226833e-01 9.52851698e-02 -7.18914926e-01 3.80118966e-01
6.45863652e-01 5.39256275e-01 3.99078101e-01 -1.28120387e+00
5.65177202e-01 8.42451394e-01 -4.68558557e-02 -8.45088422e-01
-5.20141482e-01 -2.17687652e-01 -6.67631924e-01 1.92066595e-01
4.11987036e-01 4.48327631e-01 -1.10327840e+00 1.88112509e+00
2.01273561e-02 1.02001481e-01 1.82948753e-01 1.31082869e+00
9.06109393e-01 5.74006796e-01 6.32049859e-01 -1.62664931e-02
1.68291473e+00 -1.27391112e+00 -6.98600292e-01 -7.59291947e-01
-2.39017550e-02 -4.08170104e-01 9.19731021e-01 2.37213656e-01
-1.73080397e+00 -5.44374287e-01 -7.64657259e-01 -4.66065228e-01
-2.09092647e-01 6.61021620e-02 1.30138588e+00 -6.03986345e-02
-1.02990663e+00 3.27604711e-01 -1.00354791e+00 -1.22594409e-01
1.27289689e+00 3.72651309e-01 -2.22848758e-01 -7.43739866e-03
-8.09946418e-01 7.18431771e-01 4.07492399e-01 2.28255570e-01
-1.04038084e+00 -7.66350150e-01 -8.43697131e-01 3.17399770e-01
2.96213597e-01 -1.03588927e+00 1.20652819e+00 -8.85609150e-01
-1.17293274e+00 1.00891149e+00 -4.27995533e-01 -6.62836432e-01
3.02943736e-01 -2.35870317e-01 9.67762470e-02 3.97072613e-01
5.98642565e-02 1.35560083e+00 1.10026634e+00 -9.02835011e-01
-5.55764079e-01 -3.95669401e-01 7.66912475e-02 3.89131725e-01
-1.43295117e-02 -1.44378632e-01 -7.51315355e-01 -7.48583317e-01
7.05060124e-01 -7.62668014e-01 2.94218101e-02 7.38396645e-01
3.65454927e-02 -4.37203139e-01 3.75438005e-01 -4.51361239e-01
9.36533213e-01 -2.05609703e+00 6.58662558e-01 -3.79953504e-01
6.31888986e-01 -9.98515543e-03 -2.25136474e-01 9.57702100e-02
-5.99877536e-03 -1.80697769e-01 -1.77316777e-02 -3.70625407e-01
1.97112173e-01 3.29067826e-01 -7.73142576e-01 4.61478949e-01
5.14482796e-01 1.64963090e+00 -1.02701414e+00 -5.08102179e-01
4.94122878e-02 2.07644179e-01 -8.07670355e-01 -3.12858671e-02
-7.68712282e-01 2.07180250e-02 -3.23302358e-01 9.55113649e-01
4.84273463e-01 -5.35728991e-01 3.31774764e-02 -2.49875262e-01
-1.51974604e-01 3.23685974e-01 -7.92174995e-01 2.00523043e+00
9.87848714e-02 1.03080261e+00 -5.82616776e-02 -9.21116292e-01
5.69101989e-01 8.13342407e-02 4.29968596e-01 -1.12064314e+00
6.26955042e-03 -3.53333473e-01 8.36930275e-02 -5.12725949e-01
3.96920234e-01 1.49489373e-01 2.73843527e-01 5.43386817e-01
2.28304610e-01 1.78141281e-01 1.48914710e-01 3.95964205e-01
9.89721477e-01 6.12416923e-01 3.76729444e-02 -1.33865446e-01
7.66646862e-02 2.38816932e-01 3.57008219e-01 7.69249022e-01
-6.07554674e-01 4.10255283e-01 4.46742117e-01 -6.63568139e-01
-8.97706509e-01 -1.48693991e+00 1.76594421e-01 1.48442340e+00
5.06587625e-01 -2.46868670e-01 -2.75677770e-01 -2.35991418e-01
1.79277003e-01 6.51220798e-01 -7.55223632e-01 -3.67149830e-01
-6.69513047e-01 -5.44277668e-01 4.13323522e-01 9.93562996e-01
3.64230365e-01 -1.56114423e+00 -1.27716720e+00 1.72843531e-01
-3.18600059e-01 -9.86876011e-01 -1.62778020e-01 3.92349809e-01
-8.66957009e-01 -1.14259529e+00 -4.98120517e-01 -7.78149545e-01
8.44496846e-01 5.60103893e-01 1.13522375e+00 6.12250976e-02
-4.83115286e-01 5.60044944e-01 -7.36770183e-02 -2.83126354e-01
2.05113739e-01 -2.20223337e-01 4.27016392e-02 -1.10864371e-01
4.94198054e-01 -3.56448203e-01 -7.45163620e-01 7.74517283e-02
-6.24481559e-01 6.30607247e-01 7.17203379e-01 6.89049423e-01
6.59354150e-01 -2.35036641e-01 1.11162327e-01 -4.24775958e-01
3.18587691e-01 -5.30523181e-01 -6.54112101e-01 3.47066045e-01
-4.70842779e-01 4.54344541e-01 1.85572088e-01 -6.26177669e-01
-1.26364815e+00 1.25570938e-01 3.43733519e-01 -6.64532661e-01
-2.04902560e-01 1.82288036e-01 1.57242686e-01 2.16850758e-01
5.58924556e-01 5.43631792e-01 -1.56779513e-01 -1.16132699e-01
5.92102826e-01 -1.16798997e-01 7.15546012e-01 -7.78438449e-01
4.56645072e-01 8.10188711e-01 -1.33595988e-01 -4.87850338e-01
-8.83979142e-01 -5.25867164e-01 -7.26738334e-01 -3.20906490e-01
1.19068027e+00 -1.14423096e+00 -1.46926486e+00 3.20945770e-01
-1.39723420e+00 -4.44121242e-01 -5.56511343e-01 4.30711508e-01
-9.67824817e-01 -1.14442296e-01 -5.95582724e-01 -8.39140534e-01
-4.79352325e-02 -9.70224202e-01 1.09212160e+00 3.65970284e-01
-4.59001511e-01 -7.47279704e-01 -2.23437294e-01 4.32286620e-01
3.72387975e-01 -1.55811489e-01 9.81970191e-01 -2.10407794e-01
-1.40069687e+00 2.52128720e-01 -5.36171854e-01 -4.80248302e-01
-4.48786944e-01 -1.57878309e-01 -1.01240408e+00 -2.11167112e-02
-2.03942060e-01 -4.05265152e-01 1.53513050e+00 3.65781516e-01
1.15004575e+00 -3.80538553e-01 -3.70967954e-01 7.80836940e-01
1.09794104e+00 2.50530064e-01 3.98313254e-01 3.85284275e-01
7.19203830e-01 6.54884696e-01 4.09267157e-01 3.50060463e-01
6.53157949e-01 4.38138753e-01 8.11754942e-01 3.38683158e-01
-4.08738911e-01 -2.89675087e-01 3.88393462e-01 4.01007742e-01
-3.31276685e-01 -1.12691522e-01 -1.13558710e+00 6.23336017e-01
-2.28815055e+00 -1.52767634e+00 -2.33273525e-02 1.62019551e+00
9.43201780e-01 2.43317008e-01 -6.15854450e-02 -2.37110019e-01
1.93183109e-01 4.11672711e-01 -1.02320361e+00 -2.57055834e-02
-3.03004205e-01 -4.90050726e-02 3.72437894e-01 1.31115615e-01
-1.13073349e+00 1.25327826e+00 6.57025480e+00 4.06873494e-01
-9.43109155e-01 -2.38254201e-02 1.50312886e-01 -5.09209394e-01
-4.13263649e-01 -3.12979408e-02 -8.79837632e-01 1.11115448e-01
6.78798616e-01 1.65864099e-02 7.75724053e-01 7.81994760e-01
-2.46342748e-01 -1.84516415e-01 -1.46776724e+00 1.12969482e+00
1.10726334e-01 -1.87973464e+00 3.00967902e-01 -1.98686659e-01
3.91493887e-01 1.42603263e-01 3.59407008e-01 3.05906355e-01
3.52457374e-01 -1.14334381e+00 1.43935478e+00 1.00464249e+00
2.63721704e-01 -1.33232459e-01 -5.25646210e-02 4.83693510e-01
-1.29834759e+00 -4.42448795e-01 -5.16874850e-01 -2.05265969e-01
1.89399458e-02 -1.41264135e-02 -3.43735844e-01 -6.12634234e-02
7.45782673e-01 1.08199787e+00 -5.76672137e-01 1.09516287e+00
-2.72952706e-01 3.49273622e-01 -2.16736402e-02 -9.61314961e-02
3.13190490e-01 1.10914312e-01 5.75465858e-01 1.02902794e+00
-1.88336864e-01 3.14393282e-01 -1.38932094e-03 1.43418431e+00
7.02819675e-02 -5.06089568e-01 -2.08539858e-01 -4.81931329e-01
4.51410741e-01 9.24640715e-01 -1.08598983e+00 -3.54632854e-01
-2.82257587e-01 8.79692614e-01 5.52593529e-01 5.83528697e-01
-9.82380927e-01 2.12844685e-01 9.48337555e-01 -9.31815431e-02
5.71507871e-01 -4.58045989e-01 -5.40476263e-01 -1.19961560e+00
-7.98485652e-02 -5.07214546e-01 2.80668437e-01 -9.88665104e-01
-9.86014843e-01 3.39630127e-01 -9.67029780e-02 -1.07289398e+00
8.56324807e-02 -9.09162343e-01 -2.02174857e-01 5.72911263e-01
-1.59274542e+00 -1.26176381e+00 -3.97915095e-01 7.36920655e-01
1.00321996e+00 -1.53131306e-01 6.94517374e-01 1.60885546e-02
-3.92014742e-01 1.47978097e-01 -4.22965765e-01 8.61444622e-02
2.92687386e-01 -1.11622798e+00 5.61800599e-01 6.54339373e-01
5.89482129e-01 8.69247675e-01 3.48188877e-01 -6.57785535e-01
-1.91062117e+00 -8.82120669e-01 7.05992758e-01 -8.06754649e-01
6.71532989e-01 -4.55576479e-01 -8.47267866e-01 8.74948621e-01
1.90446358e-02 1.27300248e-01 5.11982441e-01 1.53365701e-01
-8.21877658e-01 2.29641154e-01 -5.57389021e-01 9.39371288e-01
1.58175087e+00 -7.74071395e-01 -1.03932118e+00 3.64347994e-01
7.91803658e-01 -5.11162758e-01 -3.54654193e-01 -4.12735529e-02
8.04870069e-01 -9.98490810e-01 1.44519508e+00 -1.07081258e+00
5.74028075e-01 -2.71755546e-01 -1.85343638e-01 -5.88466525e-01
-8.71750176e-01 -3.36785257e-01 -7.72662878e-01 8.96001875e-01
4.55172993e-02 -1.36036962e-01 8.71975243e-01 8.42179775e-01
4.86613102e-02 -6.54602170e-01 -6.76009178e-01 -4.58331615e-01
-3.88419986e-01 -6.99202716e-01 3.46637905e-01 7.12503314e-01
-1.22867353e-01 2.59197652e-01 -6.14499934e-02 1.36111066e-01
7.87748277e-01 5.96838593e-01 1.64212838e-01 -1.35958469e+00
-9.10731927e-02 -1.00888193e+00 -3.25623512e-01 -1.27663374e+00
3.25816244e-01 -1.09871840e+00 8.85584056e-02 -1.74711180e+00
2.84287661e-01 -3.16136688e-01 -3.34070832e-01 8.60997140e-01
-2.81162784e-02 -3.91046706e-05 6.12994730e-01 6.36359990e-01
-1.09122956e+00 5.16707718e-01 1.37870431e+00 -3.89684051e-01
-5.21484651e-02 -2.54816502e-01 -6.84458017e-01 8.84972155e-01
4.69635904e-01 -2.35332176e-01 -4.85458434e-01 -9.65736866e-01
4.64116573e-01 2.17912346e-02 9.08354819e-01 -9.44896698e-01
6.94746852e-01 -5.11217475e-01 7.02834785e-01 -7.72904038e-01
5.69657862e-01 -9.51774001e-01 -1.34726226e-01 4.99805421e-01
-7.62635112e-01 -2.40277927e-02 4.08503920e-01 7.82281280e-01
-3.86339203e-02 1.40694752e-01 4.94577080e-01 -3.32704365e-01
-1.42083359e+00 5.29714525e-01 -3.94917220e-01 1.29210547e-01
1.16753829e+00 -4.96687233e-01 -7.92555690e-01 8.10930431e-02
-9.03341115e-01 5.46635747e-01 9.94307548e-02 6.41461551e-01
9.74870265e-01 -1.35624027e+00 -2.87453026e-01 4.40599620e-01
2.83683032e-01 1.34775847e-01 4.09380019e-01 9.31685925e-01
-4.33233380e-01 6.19967103e-01 -6.61810815e-01 -7.78374672e-01
-9.24620032e-01 8.66058171e-01 6.25083745e-02 3.34878147e-01
-7.99357057e-01 1.33272314e+00 3.97134006e-01 7.43462965e-02
5.04455805e-01 -5.03310561e-01 -3.13159615e-01 3.81416798e-01
6.27250850e-01 1.94011509e-01 -5.84776402e-01 -5.87889314e-01
-4.57348585e-01 4.51083720e-01 -4.51648310e-02 3.89010785e-03
1.46553671e+00 -1.00452378e-01 -2.80050814e-01 6.06565237e-01
5.94993055e-01 -4.03641403e-01 -1.99060416e+00 -2.67650962e-01
1.46874130e-01 -3.65275204e-01 8.86117592e-02 -7.63564765e-01
-8.15425456e-01 1.17399085e+00 1.99391365e-01 -1.34079047e-02
9.79960084e-01 3.22856545e-01 4.68392491e-01 5.54503620e-01
2.15608254e-01 -9.49687719e-01 4.59545583e-01 6.20862365e-01
9.14513290e-01 -1.22996664e+00 -1.06673419e-01 2.75390279e-02
-9.23100471e-01 1.13129604e+00 8.84374678e-01 -7.94639736e-02
3.16473156e-01 1.31242499e-01 -3.66929650e-01 -3.98271263e-01
-1.31212449e+00 -4.50823814e-01 5.92439651e-01 5.75225294e-01
1.38694167e-01 -6.20268546e-02 3.06816012e-01 6.39713287e-01
2.29540765e-02 -4.90860380e-02 -1.33304611e-01 9.30601537e-01
-6.67332649e-01 -3.96966726e-01 -2.17264518e-01 3.97593170e-01
2.23791361e-01 -2.57560790e-01 -1.09539494e-01 5.68065405e-01
3.19038421e-01 2.71317512e-01 5.05682349e-01 3.79124358e-02
5.46385795e-02 2.43566066e-01 7.03723907e-01 -4.89082813e-01
-3.56595755e-01 -9.48733389e-02 -4.39384311e-01 -1.06568730e+00
-6.46254539e-01 -8.01664710e-01 -1.82706225e+00 -1.49606347e-01
2.05006436e-01 -3.48405272e-01 4.05772805e-01 1.00433302e+00
3.30846220e-01 9.68639910e-01 -3.66654396e-01 -1.01381457e+00
-4.64055687e-01 -1.52939230e-01 -1.75147846e-01 3.28564197e-01
6.13013148e-01 -8.72613192e-01 8.62482265e-02 4.46449578e-01] | [9.953917503356934, 0.8956162333488464] |
0e451d69-dc90-454f-9ea5-0bce514a713e | distributionally-robust-neural-networks | null | null | https://openreview.net/forum?id=ryxGuJrFvS | https://openreview.net/pdf?id=ryxGuJrFvS | Distributionally Robust Neural Networks | Overparameterized neural networks can be highly accurate on average on an i.i.d. test set, yet consistently fail on atypical groups of the data (e.g., by learning spurious correlations that hold on average but not in such groups). Distributionally robust optimization (DRO) allows us to learn models that instead minimize the worst-case training loss over a set of pre-defined groups. However, we find that naively applying group DRO to overparameterized neural networks fails: these models can perfectly fit the training data, and any model with vanishing average training loss also already has vanishing worst-case training loss. Instead, the poor worst-case performance arises from poor generalization on some groups. By coupling group DRO models with increased regularization---stronger-than-typical L2 regularization or early stopping---we achieve substantially higher worst-group accuracies, with 10-40 percentage point improvements on a natural language inference task and two image tasks, while maintaining high average accuracies. Our results suggest that regularization is important for worst-group generalization in the overparameterized regime, even if it is not needed for average generalization. Finally, we introduce a stochastic optimization algorithm for the group DRO setting and provide convergence guarantees for the new algorithm.
| ['Pang Wei Koh*', 'Shiori Sagawa*', 'Tatsunori B. Hashimoto', 'Percy Liang'] | 2020-05-01 | null | null | null | iclr-2020-1 | ['l2-regularization'] | ['methodology'] | [ 7.25033656e-02 2.98123330e-01 -3.40267271e-01 -6.92098737e-01
-1.29303622e+00 -4.56816256e-01 3.19653869e-01 -9.20305029e-02
-8.03960860e-01 9.26747382e-01 -1.68387070e-01 -3.70952517e-01
-3.88689399e-01 -5.06829619e-01 -1.06119120e+00 -9.09517527e-01
-1.92831159e-01 4.55074400e-01 -1.91582412e-01 1.13024130e-01
-2.59969179e-02 4.39032823e-01 -1.32345414e+00 5.68924695e-02
1.04828703e+00 1.02270389e+00 -7.98595622e-02 6.25241876e-01
2.68851638e-01 6.00165427e-01 -5.08456886e-01 -4.49500203e-01
3.45513225e-01 -2.94284016e-01 -3.97457451e-01 -1.32331941e-02
1.32095265e+00 -8.51359293e-02 -2.43504405e-01 1.25702572e+00
6.11261308e-01 3.99209142e-01 9.63383377e-01 -1.10472691e+00
-8.65782022e-01 8.08563232e-01 -4.47802335e-01 1.88870013e-01
-1.80903643e-01 2.50473946e-01 1.44207919e+00 -8.74719203e-01
1.58815697e-01 1.28142178e+00 1.00423121e+00 7.57587254e-01
-1.51115024e+00 -7.82480896e-01 5.16842127e-01 -4.19471264e-01
-1.31395745e+00 -4.53555673e-01 3.46774429e-01 -3.79531205e-01
9.91915107e-01 1.79227188e-01 8.51157084e-02 1.14537609e+00
3.01400900e-01 7.05397367e-01 8.44098866e-01 -1.60891861e-01
1.48772627e-01 1.47087157e-01 5.08733630e-01 6.92416847e-01
5.89781642e-01 -1.49284936e-02 -3.05861235e-01 -1.76051438e-01
5.25682449e-01 -3.14825118e-01 -3.06282401e-01 -1.99365988e-01
-1.03507054e+00 8.79924536e-01 4.18442398e-01 1.50833353e-01
-1.65570125e-01 4.07193989e-01 4.73258913e-01 6.58232331e-01
8.57030332e-01 6.51139855e-01 -7.94795573e-01 3.45608622e-01
-1.00924087e+00 6.39205813e-01 8.28917980e-01 8.35313559e-01
7.65769899e-01 2.87954748e-01 -1.47863671e-01 1.18455350e+00
-2.22087540e-02 4.83551919e-01 5.58491588e-01 -1.06081188e+00
6.41766012e-01 1.04114302e-01 5.92306331e-02 -7.90187538e-01
-7.37381577e-01 -1.03561616e+00 -1.03300178e+00 -5.87145314e-02
9.02345300e-01 -3.84666353e-01 -8.11474085e-01 2.52304769e+00
-2.84310639e-01 -3.33313346e-02 -7.38284439e-02 6.83204710e-01
4.12546247e-01 4.16218907e-01 2.04807729e-01 -3.87446344e-01
7.50994086e-01 -6.15959167e-01 -4.70945686e-01 -5.11285961e-01
1.16817141e+00 -2.03216583e-01 1.56610298e+00 5.16278863e-01
-1.24150980e+00 -2.33984813e-01 -1.11631334e+00 5.91605045e-02
-1.67852029e-01 -1.14776894e-01 6.42785132e-01 7.80829847e-01
-1.15835023e+00 9.19969618e-01 -5.60529232e-01 -3.46120931e-02
5.74259102e-01 6.04625583e-01 -2.39801034e-01 -1.32193625e-01
-9.99502242e-01 7.80486465e-01 4.75279242e-01 4.30250820e-03
-7.29906738e-01 -1.13494253e+00 -6.98017955e-01 -9.29739028e-02
2.13602945e-01 -6.78090572e-01 1.10098553e+00 -1.20794964e+00
-1.16015267e+00 8.19266260e-01 -1.39784347e-02 -7.93609500e-01
8.21313083e-01 -4.63374108e-01 -1.74245045e-01 -3.42030466e-01
-7.34927552e-03 6.71675801e-01 5.60939133e-01 -1.24428308e+00
-2.97781169e-01 -4.01777178e-01 -1.66068867e-01 2.69777685e-01
-5.67827702e-01 -1.35032296e-01 -1.37374610e-01 -6.18000746e-01
1.58778638e-01 -1.11585438e+00 -2.97124714e-01 -1.35013282e-01
-4.56265360e-01 -3.19209278e-01 3.71117592e-01 -4.82194483e-01
1.28606105e+00 -2.01434755e+00 -5.77499643e-02 3.33165705e-01
2.25640550e-01 7.16043934e-02 -2.06188887e-01 -2.01031938e-01
-3.47783655e-01 5.65621555e-01 -4.30092901e-01 -6.52230322e-01
1.72691941e-01 4.29573119e-01 -3.91638577e-01 7.67107964e-01
1.71553925e-01 6.24550283e-01 -5.50132573e-01 -1.33958712e-01
-2.25674182e-01 1.16109505e-01 -1.05924416e+00 -1.65767461e-01
-4.00631726e-01 1.41747728e-01 -8.89670923e-02 1.82588086e-01
7.10798979e-01 -4.82066482e-01 -5.24021052e-02 1.21555321e-01
1.89827710e-01 4.68589738e-02 -1.07109344e+00 1.29361892e+00
-6.20867133e-01 7.47904360e-01 -1.54181356e-02 -1.31703591e+00
6.07639372e-01 -5.59396343e-03 2.04931110e-01 -3.59832823e-01
-5.22030070e-02 5.87772548e-01 2.43553787e-01 -3.68529201e-01
2.61817992e-01 -3.78445208e-01 -1.01448912e-02 5.41498542e-01
2.32275739e-01 1.10116288e-01 -1.09109968e-01 -1.57166913e-01
7.98000157e-01 -3.46220851e-01 2.12533195e-02 -6.13986254e-01
1.86784714e-01 -3.01947236e-01 6.86537981e-01 1.33668363e+00
-3.97294648e-02 7.72045434e-01 7.06135809e-01 -2.97885895e-01
-1.21780586e+00 -9.75276828e-01 -5.03502548e-01 1.47866809e+00
-3.02203685e-01 1.90492608e-02 -6.87370837e-01 -8.09012353e-01
1.99900314e-01 1.03145432e+00 -6.36638999e-01 -3.89622778e-01
-5.12969673e-01 -1.50324190e+00 8.40432823e-01 4.52388644e-01
3.71402085e-01 -7.14846730e-01 1.87483907e-01 -7.02840686e-02
2.66203165e-01 -8.88386309e-01 -3.80712807e-01 5.47615349e-01
-8.24805319e-01 -5.51270604e-01 -6.85814321e-01 -8.46522510e-01
6.78430557e-01 -1.94591731e-01 1.27365851e+00 8.68679211e-02
1.02225244e-01 -2.89629586e-03 1.91608161e-01 -4.45775270e-01
-2.96243459e-01 1.99094221e-01 5.43090343e-01 -1.94569632e-01
2.19264612e-01 -4.78884131e-01 -4.25227076e-01 4.16823238e-01
-6.85405970e-01 -4.41857189e-01 4.00130987e-01 1.14806056e+00
4.90279913e-01 9.98209417e-03 1.00013173e+00 -1.03628087e+00
7.53717244e-01 -6.98649466e-01 -4.97046083e-01 2.92335153e-01
-7.07321405e-01 2.27273732e-01 9.77411926e-01 -8.64818752e-01
-8.69683623e-01 -5.37010133e-01 -4.91402633e-02 -5.56886315e-01
6.65408969e-02 5.03704309e-01 -1.21184610e-01 -9.55386385e-02
1.19734848e+00 -1.35727540e-01 -3.90453893e-03 -4.38847393e-01
8.63583013e-02 4.91807699e-01 3.17006677e-01 -9.80963528e-01
4.98108238e-01 1.44214213e-01 -1.37429601e-02 -9.50640082e-01
-1.46506584e+00 2.62371618e-02 -3.45481992e-01 1.67890906e-01
5.52392125e-01 -1.04887307e+00 -4.94258255e-01 4.25004423e-01
-8.09037685e-01 -8.57997000e-01 -2.89678425e-01 7.15044320e-01
-6.62811756e-01 1.21375874e-01 -4.34520751e-01 -7.87860334e-01
-2.93900128e-02 -1.06641448e+00 7.01553702e-01 -2.79280115e-02
-4.74411622e-02 -1.53422272e+00 -2.18408182e-01 7.37872869e-02
2.27106184e-01 1.39191359e-01 1.06650162e+00 -1.28452361e+00
-2.10805103e-01 -2.33575195e-01 -2.95137197e-01 1.07324839e+00
-1.93282619e-01 3.38038057e-02 -1.14060271e+00 -4.84252214e-01
9.09629092e-02 -5.72710633e-01 1.19114363e+00 7.74325907e-01
1.68220174e+00 -4.46107507e-01 -1.24904282e-01 9.31055903e-01
1.37324584e+00 -1.69471040e-01 1.47420108e-01 -1.60591025e-02
7.38714695e-01 3.81551117e-01 1.70980603e-01 2.65109718e-01
4.72367518e-02 3.99621248e-01 4.97987308e-02 2.27555320e-01
8.06861818e-02 -1.77781537e-01 4.46185470e-01 8.11414421e-01
2.07292661e-01 -5.25373697e-01 -9.78278100e-01 4.60812420e-01
-1.79267800e+00 -8.18568289e-01 -1.57391373e-03 2.60698438e+00
9.92133141e-01 4.80278492e-01 1.15293458e-01 -1.99688762e-01
6.37494564e-01 5.09654395e-02 -7.81789005e-01 -4.50938076e-01
-4.77977902e-01 1.33548796e-01 9.30471480e-01 8.06201041e-01
-1.08758044e+00 7.66720891e-01 7.19220781e+00 1.22607684e+00
-1.08662784e+00 8.87853280e-02 1.09536791e+00 -3.52140218e-01
-4.87144351e-01 -3.02048564e-01 -1.12381411e+00 3.25493574e-01
1.12769198e+00 -1.62742987e-01 3.81289274e-01 9.12501931e-01
1.13944881e-01 8.38553254e-03 -1.38545263e+00 9.63912308e-01
7.91088119e-02 -1.16010344e+00 8.83682296e-02 2.20687538e-01
1.06472337e+00 2.77691245e-01 6.16947353e-01 4.76418734e-01
4.51013893e-01 -1.45626724e+00 5.50879776e-01 3.85502547e-01
8.55601132e-01 -9.59708750e-01 6.46745503e-01 7.46047795e-01
-2.94677198e-01 -2.25951388e-01 -8.03280354e-01 4.99044396e-02
-3.52823019e-01 8.96569669e-01 -5.00200450e-01 2.36672815e-02
4.05717552e-01 6.21552825e-01 -6.32736623e-01 1.03222346e+00
1.79961920e-01 8.73664200e-01 -7.57877052e-01 -1.48999378e-01
3.93551022e-01 -1.63339615e-01 5.69189310e-01 1.26423323e+00
3.42098027e-01 -1.00068770e-01 7.98843522e-03 6.82234526e-01
-3.59888017e-01 1.32089987e-01 -7.81984031e-01 2.06362501e-01
3.09856147e-01 6.89364791e-01 -1.84660837e-01 -1.71538711e-01
-2.83117980e-01 5.08957267e-01 7.49664724e-01 5.97041309e-01
-6.90529883e-01 -3.48469168e-01 6.73048139e-01 1.36787922e-03
-3.40984911e-02 -1.91898912e-01 -7.61709869e-01 -1.40700805e+00
7.89512023e-02 -8.54158223e-01 4.43586528e-01 -3.94264996e-01
-1.75623620e+00 3.67686659e-01 9.40591246e-02 -7.68965125e-01
-3.38041723e-01 -9.75460768e-01 -5.56393027e-01 8.49575758e-01
-1.09216976e+00 -7.05854416e-01 2.10111886e-01 4.21480745e-01
3.67095828e-01 -2.63351798e-01 6.29693925e-01 3.36627722e-01
-8.46261919e-01 1.33689678e+00 4.98166114e-01 6.00114390e-02
8.07566345e-01 -1.42919505e+00 4.77222838e-02 6.30024910e-01
1.10806577e-01 8.07158351e-01 9.40764189e-01 -3.09641689e-01
-8.74507129e-01 -1.29640090e+00 8.46586943e-01 -5.30922413e-01
8.18289459e-01 -3.75994772e-01 -1.21627986e+00 1.06779850e+00
-2.09802017e-01 1.76503330e-01 7.23687708e-01 6.87004507e-01
-6.33436322e-01 -1.26520917e-01 -1.21141779e+00 9.19948041e-01
1.29305828e+00 -5.25455594e-01 -4.03679788e-01 8.58124197e-01
6.86150670e-01 -1.15218773e-01 -9.22942281e-01 3.96852225e-01
4.40817535e-01 -9.59608316e-01 9.09716368e-01 -9.96875584e-01
2.48012036e-01 1.91967785e-01 -5.64879477e-01 -1.55302787e+00
-7.46585056e-02 -7.07829177e-01 2.87460476e-01 1.08472061e+00
8.68779182e-01 -8.68141830e-01 8.81325066e-01 7.56031930e-01
-1.34365186e-01 -8.60551894e-01 -8.85163546e-01 -1.09835410e+00
9.12076831e-01 -1.00478852e+00 8.38071704e-02 9.07002509e-01
-3.77274871e-01 1.06519386e-01 -3.11386973e-01 1.62616521e-01
1.04963899e+00 -4.44310635e-01 5.44773757e-01 -1.10930896e+00
-3.56695771e-01 -7.34797537e-01 -2.75966018e-01 -1.01827788e+00
7.33227372e-01 -1.09802949e+00 -2.93692797e-02 -8.47259879e-01
2.44463831e-02 -7.98032820e-01 -3.25001091e-01 2.67880768e-01
-3.27584624e-01 3.02523851e-01 -5.58500029e-02 3.04942541e-02
-5.69625795e-01 3.19446057e-01 1.13124871e+00 1.83594972e-01
-2.21661687e-01 2.49634624e-01 -8.71079147e-01 8.33465457e-01
8.97538602e-01 -4.37517881e-01 -4.27077472e-01 -4.99023318e-01
5.15867949e-01 -2.68297583e-01 4.63975787e-01 -1.14401543e+00
2.15232939e-01 -1.74796432e-01 4.20999289e-01 -1.32465154e-01
1.61411420e-01 -3.02089483e-01 -2.32038826e-01 8.31291750e-02
-9.37364101e-01 -1.50641769e-01 1.85034513e-01 5.07340789e-01
6.77586906e-03 -3.71543646e-01 1.31219792e+00 -5.30911423e-02
-1.65708899e-01 4.16212261e-01 -1.13923527e-01 7.77549326e-01
6.42712653e-01 -9.26073082e-03 -3.06131184e-01 -4.46342528e-01
-9.54584420e-01 3.37156385e-01 3.44805270e-01 1.06181223e-02
2.57556885e-01 -1.29164326e+00 -9.31450069e-01 2.00309500e-01
3.87411006e-02 1.62966788e-01 1.82875529e-01 8.08282197e-01
-2.86222368e-01 2.41379067e-01 1.63950160e-01 -7.07861483e-01
-7.89309442e-01 2.23712951e-01 8.22491169e-01 -1.95848867e-01
-4.12807941e-01 1.31438053e+00 6.25813544e-01 -6.36881888e-01
4.46056813e-01 -4.39627975e-01 3.35401148e-01 1.51139230e-01
3.26636016e-01 1.03109524e-01 9.60791484e-02 -3.20027620e-01
-7.60152787e-02 5.87899745e-01 -3.77149224e-01 -2.06798054e-02
1.27983201e+00 1.93079840e-02 1.57145217e-01 9.14830446e-01
1.64811790e+00 -1.19003460e-01 -1.47915030e+00 -3.43515992e-01
-2.35291198e-02 -1.66456506e-01 9.08087045e-02 -6.37532473e-01
-8.77712488e-01 9.85458732e-01 2.70432442e-01 2.57662147e-01
5.79509556e-01 -1.14238843e-01 2.81354338e-01 8.30260396e-01
1.12042665e-01 -1.34406555e+00 -1.28333926e-01 7.50558078e-01
8.45027626e-01 -1.37090325e+00 8.36178362e-02 7.12070242e-02
-5.97515702e-01 8.04214597e-01 5.42323411e-01 -6.14784539e-01
7.76474416e-01 5.00340462e-02 -1.80523589e-01 1.53861031e-01
-1.10812855e+00 2.23821744e-01 3.98695707e-01 4.52055573e-01
3.74206424e-01 -6.84122071e-02 3.48373875e-02 6.38963044e-01
-4.65588570e-01 -4.52942431e-01 3.40144962e-01 1.57933116e-01
-6.33763254e-01 -5.13970852e-01 -1.20734327e-01 1.01088071e+00
-7.38295794e-01 -1.21535905e-01 -1.03159256e-01 1.06657052e+00
-7.55424500e-02 6.37136579e-01 3.75794262e-01 -2.48144835e-01
1.40070394e-01 3.56949627e-01 4.67241377e-01 -5.42845845e-01
-5.84823251e-01 -1.57158315e-01 1.54782727e-01 -4.71994609e-01
-8.55986997e-02 -7.55110800e-01 -8.35456491e-01 -3.86861950e-01
-2.60771424e-01 -4.81891632e-02 4.59804267e-01 9.07974064e-01
-6.74353763e-02 2.29725018e-01 5.13433754e-01 -6.35549128e-01
-1.27849746e+00 -9.68356371e-01 -7.35226989e-01 4.18837965e-01
7.23157167e-01 -5.45294702e-01 -1.18650281e+00 -4.76945639e-01] | [8.091060638427734, 3.909313201904297] |
e7615352-d1ee-475e-853d-f89d7eee9830 | semantic-operator-prediction-and-applications | 2301.00399 | null | https://arxiv.org/abs/2301.00399v1 | https://arxiv.org/pdf/2301.00399v1.pdf | Semantic Operator Prediction and Applications | In the present paper, semantic parsing challenges are briefly introduced and QDMR formalism in semantic parsing is implemented using sequence to sequence model with attention but uses only part of speech(POS) as a representation of words of a sentence to make the training as simple and as fast as possible and also avoiding curse of dimensionality as well as overfitting. It is shown how semantic operator prediction could be augmented with other models like the CopyNet model or the recursive neural net model. | ['Farshad Noravesh'] | 2023-01-01 | null | null | null | null | ['semantic-parsing'] | ['natural-language-processing'] | [ 4.52992648e-01 9.93722856e-01 -5.66780530e-02 -6.97922885e-01
-3.24164748e-01 -5.77815950e-01 2.92963743e-01 3.37053806e-01
-4.95498121e-01 5.91922998e-01 4.64066744e-01 -6.37430668e-01
-4.74345055e-04 -9.21328664e-01 -6.52425647e-01 -2.20962241e-01
3.39931041e-01 4.95097041e-01 4.16341126e-01 -4.50788945e-01
3.54100734e-01 3.38674396e-01 -1.47479892e+00 4.12844092e-01
5.31378210e-01 7.53875732e-01 9.79747832e-01 7.93698430e-01
-7.85807073e-01 8.72186780e-01 -7.95845687e-01 -3.98756117e-01
1.40647203e-01 -4.51422334e-01 -1.35337877e+00 -1.33470923e-01
-8.64859968e-02 -8.16265941e-02 -2.17981011e-01 1.05299556e+00
1.52926922e-01 2.85871059e-01 3.51776332e-02 -6.53053820e-01
-9.17944789e-01 8.09124827e-01 1.06009595e-01 1.64198771e-01
3.59181494e-01 -3.31314296e-01 1.17443395e+00 -5.69233716e-01
9.45032358e-01 1.55731106e+00 4.43595737e-01 8.81265283e-01
-7.06836760e-01 -1.79862693e-01 3.32060695e-01 3.03869784e-01
-6.30371630e-01 -8.47594440e-02 7.25649297e-01 -1.13026556e-02
1.75435746e+00 9.70072672e-02 4.70880121e-01 9.26980197e-01
1.89617753e-01 9.51903820e-01 7.29237795e-01 -7.99582481e-01
2.09853366e-01 -3.89957398e-01 8.36450696e-01 9.15156960e-01
1.84766147e-02 1.45092178e-02 -3.71717155e-01 -4.12877798e-02
6.22840226e-01 -1.52975440e-01 1.94971412e-01 -2.30633527e-01
-8.50486815e-01 1.23176122e+00 3.24866652e-01 3.78272861e-01
-3.06301355e-01 2.03535616e-01 8.62209618e-01 4.27877009e-01
1.55509159e-01 4.81942683e-01 -9.96483564e-01 -2.36651257e-01
-3.05164099e-01 -3.23352516e-02 6.95953310e-01 1.09751320e+00
6.30727887e-01 1.27763614e-01 2.93226726e-02 9.38078880e-01
9.83087420e-02 6.14358000e-02 8.08851063e-01 -1.13208616e+00
7.29746759e-01 6.54358447e-01 -2.05169290e-01 -3.41235161e-01
-6.30434990e-01 -1.20933339e-01 -2.03095168e-01 -6.42098412e-02
3.53050202e-01 -1.50796935e-01 -1.17223382e+00 1.87854743e+00
1.96778830e-02 -5.96814714e-02 3.12099069e-01 6.56841278e-01
8.31998944e-01 6.25684083e-01 4.97908711e-01 -1.43759355e-01
1.78862429e+00 -1.08340049e+00 -7.21474886e-01 -6.17255390e-01
9.63438511e-01 -3.53313476e-01 1.03443670e+00 1.75685719e-01
-1.16571569e+00 -7.40596712e-01 -9.56918478e-01 -6.06599331e-01
-6.47197604e-01 -1.99118778e-01 1.15068567e+00 6.54912829e-01
-1.17849815e+00 9.32477951e-01 -1.06276417e+00 -7.51947999e-01
3.91034067e-01 5.11499166e-01 -4.25551206e-01 2.68521458e-02
-1.44499612e+00 1.17507410e+00 9.80642080e-01 7.44864270e-02
-3.19812000e-01 -2.84515709e-01 -1.18788207e+00 7.45205283e-02
1.28817469e-01 -7.94008136e-01 1.38717687e+00 -1.21805346e+00
-1.80185997e+00 7.57482886e-01 -5.66097021e-01 -7.26418912e-01
-3.05949569e-01 -2.25649804e-01 -3.07400823e-01 2.21448720e-01
-3.56021933e-02 8.28688860e-01 2.05799550e-01 -6.36602402e-01
-4.69549149e-01 -5.33933640e-01 2.95092285e-01 2.30182886e-01
1.63670838e-01 4.21914428e-01 -1.00291343e-02 -4.09904867e-01
5.84958851e-01 -6.73857212e-01 -4.94266540e-01 -5.57810068e-01
-1.42794326e-01 -4.05701965e-01 6.85315192e-01 -1.10914052e+00
9.02805507e-01 -1.86862397e+00 2.76864260e-01 -3.17537755e-01
-2.70558983e-01 4.23721075e-01 -3.68010998e-01 7.55287707e-01
-3.17977041e-01 1.56068996e-01 -3.75545382e-01 -6.63626790e-02
-9.01555493e-02 8.30616176e-01 -2.94589996e-01 -8.44545290e-02
6.45203054e-01 1.08640122e+00 -9.00238037e-01 -2.04612082e-03
2.71400809e-01 2.45335937e-01 -4.70603019e-01 2.94053033e-02
-3.91868442e-01 3.10462296e-01 -6.03369117e-01 3.60830277e-01
4.28067088e-01 1.73074342e-02 4.75549430e-01 1.89494684e-01
1.29640192e-01 1.13751996e+00 -7.62115002e-01 2.19162345e+00
-7.08678901e-01 3.58618557e-01 -1.93573967e-01 -1.46625757e+00
9.63436902e-01 4.81058091e-01 -8.05403516e-02 -9.68804479e-01
-6.06802553e-02 8.84669870e-02 4.20783162e-01 -3.42981279e-01
5.56419611e-01 -5.26130140e-01 -2.88512260e-01 2.03527063e-01
2.53951490e-01 3.02587360e-01 1.93677083e-01 -2.91331634e-02
1.33843386e+00 6.75102472e-01 4.64782238e-01 -2.65998244e-01
5.56321681e-01 2.92897582e-01 7.11765170e-01 5.50221920e-01
-8.81901607e-02 2.62327969e-01 7.21274197e-01 -4.08432811e-01
-1.06150305e+00 -9.08768177e-01 -1.45683914e-01 1.19514978e+00
-4.63197604e-02 -3.80836964e-01 -1.03622377e+00 -8.86809528e-01
-4.21150088e-01 1.04989541e+00 -5.44090688e-01 -1.78032033e-02
-9.05930996e-01 -6.18695557e-01 6.04587018e-01 9.53992248e-01
3.15945119e-01 -1.63975418e+00 -5.65605223e-01 3.34599853e-01
-9.02194604e-02 -1.15799654e+00 2.80361891e-01 8.15382481e-01
-1.21129644e+00 -7.99519777e-01 -2.34552383e-01 -1.13302934e+00
4.90366489e-01 -2.35122681e-01 9.61351037e-01 -2.12083310e-01
-1.40357614e-01 -1.41385898e-01 -6.23645008e-01 -1.92147464e-01
-6.80804849e-01 -2.90941782e-02 -5.87092698e-01 -6.29106760e-01
6.17979646e-01 -3.53733152e-01 -1.70168951e-01 -4.66819286e-01
-6.90572083e-01 -9.43402722e-02 3.67861450e-01 9.27603245e-01
2.59900272e-01 -1.04138978e-01 6.33836389e-01 -1.21884346e+00
4.02367771e-01 -3.40648741e-01 -5.31918943e-01 2.30278075e-01
-4.34228927e-01 3.18980396e-01 8.01408470e-01 3.70951742e-02
-1.28159344e+00 2.51196057e-01 -5.24330854e-01 8.15904737e-02
-4.24574435e-01 3.97668779e-01 -2.97854662e-01 5.55667102e-01
2.31756896e-01 2.43836448e-01 8.57365429e-02 -6.99634910e-01
4.12231565e-01 4.56443340e-01 1.57198831e-01 -4.42766100e-01
2.58908778e-01 2.05139503e-01 1.80990636e-01 -7.39471197e-01
-8.62288654e-01 -3.07963252e-01 -9.03309882e-01 6.31538808e-01
1.24296606e+00 -6.79010034e-01 -6.83751047e-01 7.44169503e-02
-1.55946434e+00 -1.18172742e-01 -3.87553722e-01 2.33464032e-01
-6.49541616e-01 6.19855642e-01 -1.01401520e+00 -8.41902077e-01
-3.10320497e-01 -1.05207098e+00 1.05874109e+00 1.99312031e-01
-2.90073752e-01 -1.26361954e+00 -3.11112702e-01 4.19237167e-01
1.46427140e-01 -8.86474401e-02 1.30966008e+00 -1.10024762e+00
-4.91601497e-01 -3.26925628e-02 -1.77324921e-01 4.80792314e-01
6.70048892e-02 -6.45618618e-01 -8.99755359e-01 1.15414098e-01
2.32309341e-01 -3.54099870e-02 9.23760533e-01 1.82444930e-01
1.08359373e+00 -1.03263110e-01 -2.39362329e-01 2.60464072e-01
1.55664194e+00 5.47453284e-01 7.56311655e-01 3.55332911e-01
6.09312952e-01 8.80233109e-01 5.93694150e-01 -1.13289721e-01
3.93393099e-01 3.00644070e-01 4.14287955e-01 2.63457507e-01
-2.18690544e-01 -4.76216674e-01 3.14367622e-01 7.99369156e-01
3.70175719e-01 -4.33250576e-01 -9.13551807e-01 5.08109272e-01
-1.79693091e+00 -7.86721647e-01 -1.82626218e-01 1.98244166e+00
1.99622512e-01 -7.66075999e-02 -1.25980854e-01 8.83598924e-02
8.94730091e-01 1.45907879e-01 -1.59641773e-01 -1.23706043e+00
-2.31456950e-01 9.22083497e-01 6.70677423e-01 8.32765341e-01
-8.82132590e-01 1.51844609e+00 7.00892925e+00 2.92060524e-01
-6.81121349e-01 4.02351469e-01 4.40679282e-01 5.52731276e-01
-1.24159299e-01 5.63999355e-01 -8.15527558e-01 3.54792386e-01
1.39195752e+00 5.78884542e-01 2.61491299e-01 7.63200462e-01
4.44743447e-02 -2.99992412e-02 -7.53108323e-01 4.80698466e-01
-1.13585509e-01 -1.32600069e+00 9.65388939e-02 -1.64653912e-01
-4.88820821e-02 2.96079606e-01 -6.07503533e-01 3.03451568e-01
4.52094793e-01 -9.41207886e-01 5.27799368e-01 2.24648401e-01
2.06316173e-01 -6.77708030e-01 9.01604831e-01 2.81272233e-01
-1.05075037e+00 -5.28344810e-01 -6.66886270e-01 -4.74463373e-01
3.75653088e-01 -6.40689686e-04 -6.54761136e-01 7.97478199e-01
3.06073397e-01 6.10959530e-01 -4.25725073e-01 3.49014163e-01
-3.70676339e-01 5.83138168e-01 -8.01440254e-02 -2.39113331e-01
4.97160286e-01 -1.50873750e-01 3.41492653e-01 1.33323240e+00
3.47237349e-01 2.50242651e-01 2.22363487e-01 6.72666430e-01
2.49735296e-01 1.76574111e-01 -4.93842512e-01 -3.69387001e-01
4.29514915e-01 7.70894110e-01 -1.01204693e+00 -3.52733970e-01
-5.63125789e-01 1.42442179e+00 5.52655697e-01 2.38799766e-01
-3.67860973e-01 -4.31128412e-01 6.13759518e-01 -8.64008069e-02
7.36617267e-01 -8.55803266e-02 -2.80976117e-01 -9.10636187e-01
1.24898087e-03 -3.61118674e-01 4.76037353e-01 -9.21332002e-01
-1.08565211e+00 6.53315425e-01 -1.23119302e-01 -5.61069071e-01
-4.36786890e-01 -1.09139633e+00 -5.02653182e-01 1.15698600e+00
-1.41758740e+00 -1.18265224e+00 3.50473374e-01 4.12618369e-02
8.55834901e-01 -1.44069344e-01 1.18380272e+00 -1.56982765e-01
-3.52210283e-01 1.97962031e-01 -3.05611581e-01 2.56872445e-01
-1.12289697e-01 -1.39227390e+00 1.00332260e+00 7.35125840e-01
6.18239492e-02 7.33424306e-01 5.24381161e-01 -5.16992033e-01
-1.30741513e+00 -7.94013560e-01 1.53698635e+00 -4.81764793e-01
7.99920321e-01 -4.78016555e-01 -9.80596900e-01 7.04935312e-01
9.55449492e-02 -2.22129449e-02 4.23087984e-01 2.93376476e-01
-3.90814215e-01 3.28073502e-01 -1.07325375e+00 3.50994438e-01
1.42043054e+00 -6.21506035e-01 -1.06598699e+00 3.61369163e-01
1.22040498e+00 -4.63316053e-01 -7.19116986e-01 1.19316906e-01
-3.35581182e-03 -8.18593264e-01 7.47631490e-01 -1.01421213e+00
2.77078807e-01 -1.20304506e-02 -4.77502614e-01 -9.86888289e-01
-4.68010545e-01 -4.29498494e-01 -8.67995024e-02 1.18449199e+00
6.05543375e-01 -7.37174928e-01 8.54576290e-01 2.43914679e-01
-6.72037601e-01 -4.09241766e-01 -1.13777900e+00 -8.50585163e-01
3.77865672e-01 -6.02355421e-01 6.70508325e-01 6.91267550e-01
3.27724606e-01 9.30059314e-01 1.18444848e-03 2.20800593e-01
6.18413873e-02 -6.52471259e-02 -4.32777964e-02 -1.29982698e+00
-4.78908807e-01 -2.58558989e-01 -7.66623318e-01 -1.11561394e+00
7.07293689e-01 -1.22351551e+00 4.84165810e-02 -1.89537382e+00
3.82267348e-02 -1.04760438e-01 -5.25376797e-01 6.02281570e-01
-4.98685651e-02 -2.24325582e-01 2.29046687e-01 -3.80247951e-01
-3.06729376e-01 2.75069445e-01 1.11449766e+00 2.79043525e-01
7.43751451e-02 -2.06851199e-01 -7.59355009e-01 5.57981312e-01
8.99632096e-01 -5.02983928e-01 -5.98270178e-01 -6.71684265e-01
1.11531526e-01 5.60237944e-01 2.49698371e-01 -4.78842735e-01
-2.29298145e-01 2.38866687e-01 5.99859133e-02 -3.89891922e-01
4.24569130e-01 -7.27046311e-01 -3.16515714e-01 6.50577962e-01
-4.74735707e-01 4.27181631e-01 2.67675787e-01 7.27182329e-01
-3.91547948e-01 -8.23806286e-01 4.04856056e-01 -4.91101801e-01
-9.38496232e-01 -2.54285783e-01 -3.11765999e-01 -1.98620930e-01
6.68414056e-01 -4.24545139e-01 -2.39007488e-01 5.79565810e-03
-1.09888577e+00 -9.43742618e-02 3.58880281e-01 7.60698795e-01
3.28510314e-01 -7.60544121e-01 -1.50326476e-01 2.29956031e-01
2.57268199e-03 -2.73458272e-01 3.01881701e-01 2.01540649e-01
-6.13458753e-01 9.98321235e-01 -1.30445555e-01 -1.40351251e-01
-1.00985360e+00 8.43320966e-01 1.61499679e-01 -3.87638122e-01
-9.31834996e-01 7.54693270e-01 2.21606612e-01 -9.46208596e-01
1.16792195e-01 -4.65574175e-01 -2.55398959e-01 -3.93345237e-01
3.16475749e-01 2.76114307e-02 1.53858691e-01 -6.89447999e-01
-5.72913229e-01 3.98034602e-01 -3.15720774e-02 -4.16163281e-02
1.30955255e+00 -2.01456174e-01 -2.32736975e-01 6.85432911e-01
9.84003484e-01 -5.38643062e-01 -1.06793451e+00 -5.76378182e-02
5.76682389e-01 1.32697999e-01 -8.99442956e-02 -1.04792511e+00
-5.61059415e-01 1.32035410e+00 5.25859773e-01 1.50541142e-01
7.66560137e-01 2.05154225e-01 1.24876273e+00 5.65712631e-01
3.07934761e-01 -1.27456558e+00 -4.38123435e-01 9.42009866e-01
3.90665919e-01 -9.03652608e-01 -4.42658037e-01 -6.12852871e-01
-9.01060581e-01 1.17163670e+00 2.91254669e-01 -4.11008298e-01
6.00924253e-01 8.05237591e-02 -2.08153352e-02 -8.59158859e-02
-9.24300373e-01 -5.17304242e-01 -1.03004634e-01 7.96866238e-01
8.16677213e-01 -1.26693562e-01 -6.61449432e-01 5.58043122e-01
-2.38773793e-01 -1.88001931e-01 3.61634582e-01 1.05409145e+00
-5.41643977e-01 -1.79656005e+00 1.12734266e-01 2.33930260e-01
-8.00119579e-01 -4.20739740e-01 -2.49662265e-01 7.91689336e-01
1.86559781e-01 8.81798267e-01 3.99794757e-01 4.63120565e-02
2.61715323e-01 7.55265415e-01 6.83057904e-01 -1.09066498e+00
-8.23228061e-01 -2.32074201e-01 5.22016168e-01 -5.89780867e-01
-2.91644633e-01 -6.01836979e-01 -1.96003902e+00 3.54953408e-01
-2.11952135e-01 1.59349114e-01 9.37691212e-01 1.20007527e+00
5.03433764e-01 5.42307973e-01 1.27200603e-01 -2.80178845e-01
-4.74308610e-01 -1.05264127e+00 -4.24803227e-01 3.52101624e-01
-8.26243460e-02 -3.89611572e-01 1.27978176e-01 -8.31833109e-02] | [10.42581844329834, 9.296927452087402] |
54704740-07c3-4cdb-88d7-5902c91be461 | when-differential-privacy-meets-graph-neural | 2006.05535 | null | https://arxiv.org/abs/2006.05535v9 | https://arxiv.org/pdf/2006.05535v9.pdf | Locally Private Graph Neural Networks | Graph Neural Networks (GNNs) have demonstrated superior performance in learning node representations for various graph inference tasks. However, learning over graph data can raise privacy concerns when nodes represent people or human-related variables that involve sensitive or personal information. While numerous techniques have been proposed for privacy-preserving deep learning over non-relational data, there is less work addressing the privacy issues pertained to applying deep learning algorithms on graphs. In this paper, we study the problem of node data privacy, where graph nodes have potentially sensitive data that is kept private, but they could be beneficial for a central server for training a GNN over the graph. To address this problem, we develop a privacy-preserving, architecture-agnostic GNN learning algorithm with formal privacy guarantees based on Local Differential Privacy (LDP). Specifically, we propose an LDP encoder and an unbiased rectifier, by which the server can communicate with the graph nodes to privately collect their data and approximate the GNN's first layer. To further reduce the effect of the injected noise, we propose to prepend a simple graph convolution layer, called KProp, which is based on the multi-hop aggregation of the nodes' features acting as a denoising mechanism. Finally, we propose a robust training framework, in which we benefit from KProp's denoising capability to increase the accuracy of inference in the presence of noisy labels. Extensive experiments conducted over real-world datasets demonstrate that our method can maintain a satisfying level of accuracy with low privacy loss. | ['Daniel Gatica-Perez', 'Sina Sajadmanesh'] | 2020-06-09 | null | null | null | null | ['privacy-preserving-deep-learning', 'privacy-preserving-deep-learning'] | ['methodology', 'natural-language-processing'] | [ 2.20197514e-01 5.63241959e-01 -1.76660001e-01 -5.71006656e-01
-5.31105816e-01 -5.91181397e-01 2.00130582e-01 2.09620968e-01
-2.13521019e-01 6.46395326e-01 -4.38042395e-02 -2.98421860e-01
-9.20606926e-02 -1.26304924e+00 -7.95587301e-01 -9.78881121e-01
-2.09980577e-01 -2.38838866e-02 -2.46506393e-01 1.90644011e-01
-2.92937338e-01 6.05354548e-01 -8.48481715e-01 -6.28851801e-02
6.85687304e-01 1.03456843e+00 -4.89282072e-01 3.02735388e-01
1.53680280e-01 7.92842984e-01 -4.90127414e-01 -1.15243042e+00
7.08504736e-01 -9.53543559e-02 -7.10378885e-01 -3.95944178e-01
2.26868883e-01 -4.82805908e-01 -8.66747975e-01 1.46585548e+00
5.34058213e-01 -8.01542848e-02 1.90872699e-01 -1.74642932e+00
-9.21267152e-01 8.07117820e-01 -5.38344741e-01 -2.19185531e-01
6.36582300e-02 4.92986403e-02 1.04499924e+00 -1.07024945e-01
4.10788417e-01 1.12810671e+00 9.78036225e-01 8.14535499e-01
-1.32870185e+00 -1.07576954e+00 6.51668161e-02 -1.67325884e-01
-1.56691980e+00 -2.86010593e-01 9.08569515e-01 1.17249690e-01
3.98481101e-01 5.98062813e-01 4.08978224e-01 1.27322578e+00
1.75093949e-01 7.93129265e-01 7.01827049e-01 1.54571280e-01
2.30008140e-01 2.43000016e-01 -1.01188108e-01 6.55492008e-01
5.81489921e-01 7.27201402e-02 -3.05167168e-01 -6.11350715e-01
6.28728151e-01 3.57307047e-01 -4.12699074e-01 -6.70898259e-01
-5.08674204e-01 8.19510937e-01 8.61959815e-01 -9.41636264e-02
-3.40116948e-01 4.97715056e-01 6.44011617e-01 5.90923071e-01
6.51533842e-01 3.34685706e-02 -5.35202980e-01 5.81679285e-01
-5.60312390e-01 1.34904191e-01 1.18603277e+00 1.01863158e+00
7.89884686e-01 -3.56118411e-01 -2.90330052e-01 3.23121190e-01
3.43533784e-01 1.97673082e-01 1.29676042e-02 -8.16789031e-01
5.01042962e-01 7.03473151e-01 -4.22646254e-01 -1.50934362e+00
-1.22595988e-01 -4.09050822e-01 -1.53759444e+00 -1.28501520e-01
7.15521276e-02 -3.60011399e-01 -6.14762902e-01 2.11388755e+00
4.08194214e-01 2.82170326e-01 2.47211874e-01 8.86687756e-01
7.81572878e-01 4.24440980e-01 2.33069569e-01 -7.18511045e-02
1.20414352e+00 -6.11451089e-01 -6.83741271e-01 2.26056166e-02
6.14590764e-01 1.49718106e-01 5.46966434e-01 1.54031813e-02
-7.51342058e-01 1.08976789e-01 -7.95625746e-01 -4.22277093e-01
-5.25665104e-01 -3.47884715e-01 9.56872106e-01 1.01010084e+00
-1.18315387e+00 6.96147442e-01 -8.18537652e-01 -2.60595739e-01
9.76719618e-01 6.85313404e-01 -7.62682438e-01 -4.06359602e-03
-1.50019050e+00 2.27687374e-01 3.33590120e-01 3.26760024e-01
-7.94676304e-01 -8.01558793e-01 -1.04371691e+00 4.65988010e-01
5.24383962e-01 -8.67676616e-01 8.51098776e-01 -8.45671475e-01
-1.08912337e+00 8.24899375e-01 1.29346579e-01 -7.34664381e-01
6.36931360e-01 2.80064195e-01 -4.67734724e-01 9.66426358e-02
3.23934816e-02 1.89550236e-01 7.33858645e-01 -1.05560648e+00
-2.47610420e-01 -7.96804070e-01 2.32709855e-01 -2.34821998e-03
-6.59636557e-01 -2.35679954e-01 -5.10140479e-01 -6.11082375e-01
3.04769129e-02 -7.04160392e-01 -4.67880905e-01 4.39385146e-01
-6.97914124e-01 -1.47551879e-01 1.07502127e+00 -8.05968761e-01
1.05407417e+00 -2.35574579e+00 -2.40668148e-01 8.36399794e-01
7.61572063e-01 3.41924369e-01 -3.29550542e-02 3.37411106e-01
1.07495360e-01 3.48955065e-01 -3.57992858e-01 -5.59257686e-01
1.40326306e-01 3.97175461e-01 -1.68427244e-01 7.28792012e-01
-1.15466514e-03 1.13577020e+00 -7.52663970e-01 -3.52440923e-01
-1.34598181e-01 6.06057167e-01 -4.61534262e-01 2.10785493e-01
-1.63911402e-01 1.81432560e-01 -7.25749969e-01 5.39060712e-01
1.26481688e+00 -4.33882803e-01 4.71327692e-01 -9.12095979e-02
6.17388606e-01 6.93199271e-03 -1.06165659e+00 1.50263321e+00
-6.38458580e-02 2.32992768e-01 7.59805322e-01 -9.94197845e-01
9.17445421e-01 3.81060153e-01 4.01732981e-01 -4.43884909e-01
2.44784206e-01 -1.63249135e-01 -5.59820473e-01 -2.62775481e-01
6.15824349e-02 4.11879607e-02 -1.95053473e-01 4.71217781e-01
-1.23168871e-01 5.49845695e-01 -4.84072834e-01 4.28299814e-01
1.49185479e+00 -5.62485218e-01 4.66342270e-02 -1.57124043e-01
6.47625566e-01 -7.20774114e-01 7.75618315e-01 8.58257055e-01
-3.02188158e-01 4.10523474e-01 9.12554741e-01 -5.10693371e-01
-5.49921334e-01 -8.92394006e-01 1.12490021e-01 9.25180495e-01
1.03961617e-01 -5.66880405e-01 -7.95937836e-01 -1.09798706e+00
3.98191184e-01 3.14042002e-01 -6.85501575e-01 -5.46318233e-01
-3.09338123e-01 -7.70625651e-01 9.36247647e-01 3.31572533e-01
7.46855855e-01 -7.38292396e-01 1.07484702e-02 -1.36040017e-01
-3.84931490e-02 -9.67364550e-01 -5.86885691e-01 1.60085425e-01
-4.90069509e-01 -1.02954304e+00 -1.82684109e-01 -7.23947942e-01
9.30163920e-01 1.32637963e-01 7.38591611e-01 2.16086030e-01
6.02105893e-02 4.11121070e-01 6.59739748e-02 -3.59820843e-01
-3.23376298e-01 3.97140026e-01 -1.38688967e-01 4.10779297e-01
4.57946360e-01 -9.45570529e-01 -6.41909838e-01 -1.75382853e-01
-1.08404374e+00 -2.47520164e-01 4.67890531e-01 6.01483047e-01
4.81971264e-01 3.12756568e-01 4.52605456e-01 -1.52362692e+00
9.15480494e-01 -6.94741011e-01 -6.96008146e-01 3.48277271e-01
-7.94274986e-01 1.98459759e-01 1.03884280e+00 -1.34972885e-01
-8.26221228e-01 1.71624988e-01 -5.49097471e-02 -5.91414273e-01
1.94740400e-01 3.49884272e-01 -7.82402694e-01 -5.58337033e-01
3.33920062e-01 2.10914388e-02 3.64202768e-01 -3.50708067e-01
5.08223534e-01 6.50275588e-01 4.80739832e-01 -4.18088406e-01
9.48942959e-01 5.66624284e-01 5.00645041e-01 -2.54891247e-01
-5.17196655e-01 -7.31902048e-02 -7.18370229e-02 2.58673310e-01
6.86735630e-01 -1.01687777e+00 -1.36752212e+00 4.25870836e-01
-9.85374451e-01 9.54777226e-02 -2.66992688e-01 -3.48630128e-03
-1.67260200e-01 6.23099267e-01 -7.81024575e-01 -9.04719889e-01
-8.83216977e-01 -7.77278900e-01 8.56144845e-01 2.79454291e-01
1.53894812e-01 -1.10848296e+00 -3.44702870e-01 1.84464991e-01
4.12403762e-01 5.60739934e-01 7.32986510e-01 -9.66429234e-01
-8.83725107e-01 -4.48511571e-01 -4.40708965e-01 4.23365831e-01
6.29752278e-02 -5.57614028e-01 -1.17984509e+00 -6.51902080e-01
1.65692464e-01 -2.04008907e-01 8.07672501e-01 -1.75776090e-02
1.82880867e+00 -9.85141695e-01 -4.65829521e-01 1.23010767e+00
1.47188258e+00 -3.43590319e-01 5.26892066e-01 -1.32144973e-01
1.07561851e+00 5.29159069e-01 -1.33083880e-01 5.03459930e-01
5.44186294e-01 4.60720435e-02 7.89416313e-01 -2.01094121e-01
2.88061410e-01 -7.10675180e-01 6.97918534e-02 2.48318061e-01
3.84370208e-01 -5.27687371e-01 -4.10777867e-01 3.71190727e-01
-1.95469415e+00 -6.69265211e-01 1.62160620e-02 2.05406499e+00
7.39165246e-01 -3.53879422e-01 -1.03865936e-01 -2.12417722e-01
6.56382263e-01 4.44522262e-01 -8.06589365e-01 -3.38054985e-01
-1.06002986e-01 2.19898283e-01 1.12936270e+00 2.23504350e-01
-1.18113935e+00 5.89753628e-01 5.18502855e+00 6.66744411e-01
-9.09470141e-01 5.36365621e-02 8.32709789e-01 7.08946735e-02
-6.52114272e-01 -1.40879760e-02 -3.09032470e-01 5.04188180e-01
9.43953216e-01 -4.14836556e-01 7.40508854e-01 1.00355172e+00
-1.71747327e-01 6.48249924e-01 -1.17328823e+00 9.70303357e-01
-1.68678835e-01 -1.19275689e+00 -6.14531860e-02 4.03954625e-01
4.08061206e-01 -1.10956587e-01 8.12661201e-02 2.33959153e-01
6.96791351e-01 -1.18790841e+00 7.24360943e-02 5.45599699e-01
7.40291357e-01 -1.08132017e+00 6.95511818e-01 3.94050628e-01
-1.06723797e+00 -2.01914594e-01 -5.96933544e-01 1.66775703e-01
-2.75956064e-01 6.39151871e-01 -5.82754016e-01 9.35488880e-01
8.27943861e-01 6.42856896e-01 -4.21925575e-01 5.10914445e-01
-2.72012353e-01 4.94472295e-01 -4.97343957e-01 -8.47701263e-03
1.38838461e-03 -2.83608317e-01 3.22666824e-01 8.04681122e-01
1.43034175e-01 2.47880235e-01 -4.91008647e-02 1.11947072e+00
-9.08026457e-01 -9.93979815e-03 -8.69449019e-01 -2.48561669e-02
6.30673826e-01 1.33573782e+00 -1.64029270e-01 1.66618913e-01
-4.31157976e-01 1.24258780e+00 6.89415753e-01 4.86028224e-01
-5.70448995e-01 -4.62084413e-01 9.76560473e-01 4.20353636e-02
1.90079987e-01 3.44468057e-01 -2.44483665e-01 -1.10453749e+00
4.05698776e-01 -7.24838078e-01 8.56389284e-01 -2.08392158e-01
-1.74891627e+00 1.90949440e-01 -4.99245733e-01 -6.11251771e-01
1.90656051e-01 -2.50769526e-01 -6.10799432e-01 9.55617368e-01
-1.34682107e+00 -1.45069182e+00 -1.05275676e-01 9.54710305e-01
-7.43409574e-01 -8.97038057e-02 9.09510732e-01 3.23112726e-01
-6.35807097e-01 1.31015325e+00 1.76552311e-01 6.68914378e-01
3.49589884e-01 -1.12850523e+00 4.95339811e-01 1.01707065e+00
-5.22364378e-02 7.77511120e-01 2.47617081e-01 -6.96834564e-01
-1.77414191e+00 -1.54429746e+00 9.22832131e-01 -1.61612555e-01
3.43123525e-01 -8.30451190e-01 -1.14078248e+00 1.12401211e+00
-1.09246135e-01 6.24216676e-01 7.86761701e-01 4.54661325e-02
-5.26817679e-01 -6.83990300e-01 -1.99201632e+00 5.03298640e-01
1.17092216e+00 -9.13242638e-01 1.31428108e-01 4.15953428e-01
1.14304960e+00 -4.12738472e-01 -9.73083198e-01 2.67373264e-01
3.95989746e-01 -7.32776701e-01 9.56607640e-01 -6.11555934e-01
-1.24042876e-01 -2.15847075e-01 2.94157257e-03 -9.62841392e-01
-3.57996821e-01 -9.78471220e-01 -3.49754483e-01 1.56030071e+00
8.16949531e-02 -9.87435520e-01 1.36301911e+00 1.23125744e+00
5.19722521e-01 -5.28241336e-01 -1.08825445e+00 -4.16844934e-01
-1.22326121e-01 -2.00549364e-01 1.21547377e+00 1.07755792e+00
-2.12285861e-01 1.42887428e-01 -5.88255703e-01 6.74015760e-01
8.82720470e-01 -8.86260718e-02 7.96620011e-01 -1.24919987e+00
-9.54561234e-02 4.78524938e-02 -5.87644875e-01 -6.73080266e-01
6.76813304e-01 -1.29949510e+00 -3.46823424e-01 -1.24634540e+00
1.70912579e-01 -3.75598818e-01 -5.17504692e-01 7.55893528e-01
-7.98871219e-02 -7.45452642e-02 -3.59452739e-02 -1.06839836e-01
-5.78107655e-01 5.65374017e-01 9.62442935e-01 -1.95174992e-01
4.78084013e-02 3.80807579e-01 -1.29572475e+00 2.09257051e-01
7.45472372e-01 -5.99935472e-01 -7.03640461e-01 -3.54247779e-01
2.53903687e-01 6.02808148e-02 7.13666201e-01 -6.71887815e-01
5.10997057e-01 2.11153790e-01 3.10076863e-01 -2.50977516e-01
7.04763532e-02 -1.46012068e+00 5.43411314e-01 4.00767803e-01
-4.09445882e-01 -1.98830575e-01 -2.49904349e-01 9.83287036e-01
1.20645881e-01 2.46744439e-01 5.54014385e-01 -5.28152846e-02
-2.56849736e-01 1.03167224e+00 2.26540342e-01 -1.03808545e-01
1.09266603e+00 1.51717812e-01 -3.56244594e-01 -5.72217166e-01
-5.14254749e-01 6.64644122e-01 5.49165249e-01 4.31028344e-02
5.22519350e-01 -1.28338718e+00 -5.41565478e-01 6.16320848e-01
7.40704164e-02 2.95259923e-01 1.98610112e-01 4.02359813e-01
-2.41673633e-01 5.06318919e-02 1.15046322e-01 -2.24495996e-02
-1.05723441e+00 8.58139992e-01 4.02024627e-01 -3.21144432e-01
-6.36004448e-01 9.96877134e-01 2.01639101e-01 -8.00050855e-01
5.88552237e-01 -1.36551514e-01 2.64193296e-01 -2.49216869e-01
3.19939345e-01 2.83996344e-01 2.72711013e-02 -3.64066631e-01
-4.38842058e-01 -1.88801631e-01 -2.04729483e-01 4.90439147e-01
1.32459569e+00 -2.77314097e-01 -5.75965106e-01 -1.36444643e-01
1.55837142e+00 -1.02522083e-01 -1.20078301e+00 -4.64173079e-01
-1.66679576e-01 -4.65794742e-01 -9.40431952e-02 -5.09405434e-01
-1.71853888e+00 5.89684725e-01 3.60085487e-01 4.27969962e-01
1.42958081e+00 -1.12700777e-03 1.10783184e+00 2.24598050e-01
2.92329282e-01 -7.02313185e-01 -6.58379138e-01 2.70453729e-02
3.78355235e-01 -1.13393962e+00 -7.75782838e-02 -5.89689553e-01
-3.45701724e-01 7.62137055e-01 5.88203967e-01 -3.34328949e-03
8.86444807e-01 2.97978520e-01 -1.08603589e-01 -2.14424640e-01
-5.68458378e-01 3.24832648e-01 -6.98482320e-02 8.61493707e-01
-1.29603252e-01 3.01698089e-01 -7.16547668e-02 1.09950376e+00
-2.15371430e-01 -1.14506193e-01 1.85007110e-01 6.42664611e-01
3.04078907e-01 -1.15949726e+00 2.12512597e-01 6.19394660e-01
-7.88651109e-01 -2.96676215e-02 -6.42560124e-01 5.11999547e-01
7.67390952e-02 5.79333544e-01 -2.10620210e-01 -5.23085058e-01
1.83662668e-01 -2.06374750e-01 -8.49814117e-02 -2.45051220e-01
-8.69493544e-01 -5.98887205e-01 -7.68208653e-02 -7.69823909e-01
-1.34213731e-01 -4.98871148e-01 -1.09272683e+00 -8.95434678e-01
-7.56147206e-02 2.31429368e-01 3.13510954e-01 4.70350295e-01
6.75190449e-01 1.15632817e-01 8.66432607e-01 -7.81770609e-03
-7.38399446e-01 -3.30840558e-01 -1.02276194e+00 6.43400669e-01
5.59848785e-01 1.48605406e-01 -5.28960288e-01 -5.79056561e-01] | [6.024110317230225, 6.963780403137207] |
b4d62276-2b9a-41be-9a83-49c32079c8a6 | learning-from-ordered-sets-and-applications | 1408.0043 | null | http://arxiv.org/abs/1408.0043v1 | http://arxiv.org/pdf/1408.0043v1.pdf | Learning From Ordered Sets and Applications in Collaborative Ranking | Ranking over sets arise when users choose between groups of items. For
example, a group may be of those movies deemed $5$ stars to them, or a
customized tour package. It turns out, to model this data type properly, we
need to investigate the general combinatorics problem of partitioning a set and
ordering the subsets. Here we construct a probabilistic log-linear model over a
set of ordered subsets. Inference in this combinatorial space is highly
challenging: The space size approaches $(N!/2)6.93145^{N+1}$ as $N$ approaches
infinity. We propose a \texttt{split-and-merge} Metropolis-Hastings procedure
that can explore the state-space efficiently. For discovering hidden aspects in
the data, we enrich the model with latent binary variables so that the
posteriors can be efficiently evaluated. Finally, we evaluate the proposed
model on large-scale collaborative filtering tasks and demonstrate that it is
competitive against state-of-the-art methods. | ['Svetha Venkatesh', 'Truyen Tran', 'Dinh Phung'] | 2014-07-31 | null | null | null | null | ['collaborative-ranking'] | ['graphs'] | [ 9.89491716e-02 -1.31393477e-01 -2.64622837e-01 -5.49437404e-01
-1.02294481e+00 -6.87814355e-01 2.84377247e-01 1.89378679e-01
-6.17732167e-01 6.84817910e-01 -8.00865665e-02 -5.53229332e-01
-6.50716662e-01 -9.92339432e-01 -7.70247161e-01 -7.73534358e-01
-3.67056549e-01 1.08391929e+00 -1.80584472e-02 2.52595335e-01
1.43093839e-01 8.92840475e-02 -1.53993273e+00 1.53127700e-01
7.58904457e-01 9.10517037e-01 3.23788464e-01 5.27074516e-01
-1.18216410e-01 9.36406180e-02 -2.76593208e-01 -6.95783377e-01
3.95174354e-01 -1.88558757e-01 -7.36051083e-01 2.96463221e-01
1.35614976e-01 -1.03827581e-01 1.29325390e-02 1.23799646e+00
1.57379687e-01 3.51402134e-01 6.65001035e-01 -1.12016439e+00
-2.66115010e-01 8.53600085e-01 -6.65817797e-01 9.62486416e-02
2.33438183e-02 -1.13585047e-01 1.49034500e+00 -5.82977891e-01
6.12119734e-01 1.24945116e+00 -1.32120927e-04 9.77059230e-02
-1.59194970e+00 -7.17575073e-01 5.89564204e-01 6.97583146e-03
-1.26694500e+00 -8.92750174e-02 5.10930896e-01 -4.86035347e-01
4.63152736e-01 6.57304704e-01 5.45182288e-01 7.64416397e-01
-1.63356870e-01 9.37540054e-01 1.25230598e+00 -3.46662581e-01
5.15547454e-01 6.63843974e-02 4.91922468e-01 8.08224797e-01
4.15473282e-01 -3.15858424e-01 -5.80171406e-01 -7.05104947e-01
3.36869091e-01 4.35748845e-01 3.55209783e-02 -3.87345999e-01
-7.81916142e-01 8.75540137e-01 -1.00367926e-01 -1.76727220e-01
-3.18338782e-01 2.53291547e-01 -1.26197189e-01 2.84951329e-01
4.89547521e-01 3.25493306e-01 -7.48744726e-01 -2.30104893e-01
-1.11439741e+00 3.63088995e-01 8.34058166e-01 1.05666876e+00
7.26893306e-01 -7.27721870e-01 -2.09684670e-01 7.45340824e-01
5.17347455e-01 3.07022452e-01 -3.37940931e-01 -1.00617659e+00
5.14925539e-01 3.50323647e-01 5.74558377e-01 -4.49318260e-01
-1.10583752e-01 -6.11708105e-01 -8.50363851e-01 -1.84338734e-01
7.38821924e-01 -3.10458660e-01 -7.29592562e-01 1.79238951e+00
3.93498391e-01 -1.21822618e-01 -5.44101775e-01 6.25591099e-01
-4.51127589e-02 6.33494973e-01 6.21212609e-02 -6.07406139e-01
1.56206310e+00 -6.23932004e-01 -5.51469922e-01 -1.81841612e-01
2.95227557e-01 -5.73813140e-01 1.02116299e+00 8.87411237e-01
-1.32515001e+00 -2.30844259e-01 -4.87358302e-01 8.60314891e-02
-8.80391747e-02 1.62322864e-01 1.00554526e+00 6.82493925e-01
-8.00338209e-01 7.60643065e-01 -9.77208257e-01 5.24255587e-03
4.50722188e-01 6.03169203e-01 1.45468324e-01 -3.51023406e-01
-9.46230412e-01 8.82902965e-02 -1.83676798e-02 2.14977078e-02
-8.10124099e-01 -2.95207500e-01 -3.25821221e-01 3.50808680e-01
8.90111983e-01 -7.05667019e-01 1.07615817e+00 -3.93174171e-01
-1.18635786e+00 5.39955676e-01 -4.78963614e-01 -1.83117747e-01
3.81946146e-01 -1.38252020e-01 -4.03341316e-02 -2.52680063e-01
1.86156910e-02 1.62824795e-01 7.63908029e-01 -1.02527905e+00
-8.66601765e-01 -6.57692492e-01 2.84250230e-01 8.68280903e-02
-3.08750570e-01 2.36670360e-01 -9.15169716e-01 -3.14872146e-01
3.15937310e-01 -1.29441547e+00 -7.04050303e-01 -1.91443235e-01
-7.46620119e-01 -4.00222957e-01 -6.07391670e-02 -2.93874919e-01
1.42150259e+00 -2.10596061e+00 3.00239086e-01 5.53311050e-01
2.61561632e-01 -3.07332009e-01 2.18252972e-01 4.27388877e-01
2.59646773e-01 4.69041705e-01 3.13119330e-02 -6.31893337e-01
3.43692511e-01 9.41318944e-02 -5.82401231e-02 5.41960120e-01
-4.59843963e-01 5.11737227e-01 -5.61854184e-01 -3.42563003e-01
-2.35456154e-01 -3.88066918e-02 -8.32763076e-01 1.30965337e-02
-5.71788073e-01 1.50662556e-01 -7.60875940e-01 4.70469266e-01
7.98597157e-01 -6.41252220e-01 6.49667263e-01 2.20229849e-01
-9.52782705e-02 3.76066595e-01 -1.58661520e+00 1.47473121e+00
-1.79348990e-01 4.48446833e-02 1.70123383e-01 -8.15985978e-01
5.46484172e-01 2.65187956e-02 3.94037127e-01 -4.28521223e-02
1.09751672e-01 -8.59252270e-03 1.25499904e-01 -2.63127178e-01
2.95602500e-01 -2.11594313e-01 -3.86630237e-01 6.69443548e-01
-2.24970773e-01 2.18226120e-01 3.93748760e-01 2.30382010e-01
1.13669169e+00 -1.13439992e-01 -3.61987227e-03 -2.96922743e-01
1.72903121e-03 -3.41104597e-01 6.81950271e-01 1.24263394e+00
2.07249835e-01 2.03580961e-01 9.03683245e-01 -2.03984693e-01
-7.54392445e-01 -1.11855185e+00 -9.38152596e-02 1.27432394e+00
8.82370844e-02 -5.28556347e-01 -5.49264789e-01 -5.89375854e-01
-7.79392496e-02 7.91858613e-01 -5.33434689e-01 2.04181075e-01
-1.53639242e-01 -1.13859642e+00 -2.23005310e-01 1.79321811e-01
1.21877484e-01 -6.16166830e-01 -3.35283667e-01 2.25785300e-01
-2.21900761e-01 -6.78527534e-01 -5.25274754e-01 3.39272499e-01
-1.03934002e+00 -8.54053259e-01 -1.90578237e-01 -4.78324950e-01
7.81387269e-01 1.66198626e-01 1.04300058e+00 -1.28976524e-01
-2.16042757e-01 6.42686337e-02 -2.87698358e-01 -1.51809722e-01
1.10864565e-01 8.20872933e-02 1.29948437e-01 1.78459913e-01
4.79993910e-01 -5.63036442e-01 -7.33777523e-01 3.54893893e-01
-7.72371888e-01 -3.74318915e-03 5.64426303e-01 5.94961047e-01
8.44592929e-01 6.12555921e-01 1.02322280e-01 -1.19450140e+00
4.54656214e-01 -6.08580709e-01 -8.92881691e-01 3.87553871e-01
-5.76040268e-01 3.07610154e-01 4.96106267e-01 -5.00079513e-01
-9.23104882e-01 1.31885916e-01 2.88426485e-02 -1.23888239e-01
-6.33047968e-02 7.81283081e-01 -3.01201493e-01 5.21648884e-01
6.21172599e-02 1.23147212e-01 -2.47116163e-01 -9.73786414e-01
4.26119834e-01 6.08786285e-01 6.93747178e-02 -8.66217315e-01
7.26408720e-01 6.08875930e-01 -2.25505717e-02 -7.16150224e-01
-9.21895862e-01 -5.27883708e-01 -3.93642545e-01 2.42353939e-02
6.52315021e-01 -7.20058262e-01 -1.19404948e+00 6.05506338e-02
-6.90262139e-01 -3.04896295e-01 -2.06270084e-01 5.72911084e-01
-3.41698796e-01 3.65448117e-01 -6.16392910e-01 -1.35040748e+00
8.70100334e-02 -1.14853108e+00 8.15701663e-01 9.32665393e-02
-8.11799094e-02 -5.19284785e-01 -5.65442406e-02 6.21826291e-01
-1.40074501e-02 -2.22477272e-01 1.19926834e+00 -7.31134057e-01
-1.08787072e+00 -2.88280070e-01 4.94523309e-02 -1.10857323e-01
-1.74212113e-01 -3.54022719e-02 -3.50303948e-01 -3.93941402e-01
-1.19224101e-01 -9.69600156e-02 9.58603263e-01 5.68394363e-01
1.76842427e+00 -6.14704788e-01 -3.99391770e-01 4.21788007e-01
1.18141830e+00 2.27927685e-01 3.69927496e-01 4.87611862e-03
1.72238767e-01 5.28097749e-01 8.60226512e-01 8.75792563e-01
3.89256179e-01 6.65862322e-01 3.36911887e-01 1.49073794e-01
7.50664592e-01 -3.15732330e-01 1.27972975e-01 5.72169483e-01
-4.50903140e-02 -6.93568826e-01 -4.53029066e-01 3.03467661e-01
-1.83693171e+00 -8.02229524e-01 -1.35554120e-01 2.52068996e+00
9.44808364e-01 4.01266068e-01 7.39561766e-02 -1.61712244e-01
6.03957534e-01 -6.42562956e-02 -5.74524283e-01 -5.28643169e-02
1.41360655e-01 3.04365486e-01 6.07149601e-01 6.58280909e-01
-1.05859780e+00 8.04364562e-01 5.40765095e+00 1.17753482e+00
-3.38577062e-01 1.61459833e-01 8.97413552e-01 -5.85612595e-01
-5.02906740e-01 3.32976997e-01 -1.05159688e+00 7.07080126e-01
8.21097016e-01 1.18380703e-01 6.55134618e-01 7.11875260e-01
2.12094188e-01 -4.77156430e-01 -1.11155438e+00 8.38356972e-01
-1.95317790e-01 -9.84934807e-01 -1.84676111e-01 6.70421839e-01
7.24003613e-01 -2.43375421e-01 2.68482327e-01 2.96167493e-01
8.11445951e-01 -7.52575755e-01 5.63776433e-01 5.63163996e-01
5.56448579e-01 -7.59639263e-01 2.58073181e-01 8.11599493e-01
-8.90572548e-01 -3.46336216e-01 -3.59540612e-01 -4.52627130e-02
4.15380239e-01 7.81986773e-01 -3.49780291e-01 2.09386036e-01
8.66734922e-01 9.15361494e-02 -2.49664679e-01 1.09812462e+00
-6.13930337e-02 8.29347014e-01 -6.59890115e-01 -2.74923146e-01
1.69113707e-02 -6.94226623e-01 4.57187682e-01 7.07702994e-01
4.40462977e-01 4.98215526e-01 3.51130962e-01 6.98661745e-01
-3.56892407e-01 -4.01147716e-02 -1.19285196e-01 -2.57724047e-01
5.01683652e-01 1.18506062e+00 -9.83671248e-01 -2.44584054e-01
-2.50357181e-01 8.43556702e-01 3.92048925e-01 3.44389081e-01
-8.46818924e-01 -8.25783610e-03 6.80264473e-01 2.58521557e-01
7.48532474e-01 -4.27630633e-01 -3.26806977e-02 -1.23815119e+00
-3.60146575e-02 -7.09244370e-01 6.09877229e-01 -4.54484552e-01
-1.29069781e+00 2.96140432e-01 1.46384567e-01 -7.45893836e-01
1.37001589e-01 -3.39051694e-01 -2.32635602e-01 6.29881501e-01
-9.75280166e-01 -5.75341702e-01 1.85735464e-01 3.60142231e-01
5.54764330e-01 2.37915620e-01 5.31341970e-01 2.82860130e-01
-5.90116620e-01 5.94760537e-01 5.23253858e-01 -2.56080121e-01
3.13538045e-01 -1.22944069e+00 1.57929733e-01 7.32757866e-01
3.10634792e-01 1.07009351e+00 8.83948863e-01 -5.95808923e-01
-1.58214676e+00 -7.68666685e-01 9.78204250e-01 -4.32010323e-01
5.28849244e-01 -8.22496533e-01 -5.51420748e-01 6.79619551e-01
-9.83531848e-02 -2.92929739e-01 1.14119422e+00 7.86228716e-01
-1.79796457e-01 -1.06555626e-01 -9.19655740e-01 6.22296512e-01
1.23987508e+00 -2.72133499e-01 -1.26352921e-01 5.76935232e-01
6.12651944e-01 5.61470948e-02 -7.72423029e-01 1.36206508e-01
5.47734261e-01 -8.63191187e-01 9.12197828e-01 -9.05966222e-01
2.87369400e-01 -1.58731103e-01 -2.92935878e-01 -1.05598736e+00
-4.42716271e-01 -9.56158400e-01 -1.96337253e-01 1.10248280e+00
7.71524370e-01 -3.82903546e-01 1.00404775e+00 1.02882016e+00
4.35208410e-01 -9.40253019e-01 -9.06834006e-01 -5.35802722e-01
-1.57371372e-01 -4.89126712e-01 5.68071306e-01 5.03907740e-01
-1.21831179e-01 3.81212026e-01 -5.23646593e-01 2.53807694e-01
9.73141253e-01 5.80047905e-01 7.66378462e-01 -1.38377512e+00
-9.86221611e-01 -2.23847657e-01 3.16300243e-01 -1.65039384e+00
-2.52679318e-01 -6.38846993e-01 4.14427780e-02 -1.44272482e+00
7.10438848e-01 -8.77284646e-01 -3.15542758e-01 2.39928082e-01
-2.05630332e-01 -1.79915413e-01 1.93239227e-01 1.59742504e-01
-1.19823492e+00 4.85681355e-01 1.07084548e+00 1.97936557e-02
-2.51056731e-01 6.87484205e-01 -1.04673505e+00 6.28063142e-01
5.58575213e-01 -7.46697307e-01 -2.91824996e-01 -2.15996251e-01
6.72342718e-01 6.11685872e-01 -5.26632601e-03 -3.47999215e-01
2.91692108e-01 -4.52629030e-01 1.87381841e-02 -9.32561040e-01
5.18200636e-01 -7.89551497e-01 3.98029804e-01 2.46567145e-01
-6.24416173e-01 -2.85631210e-01 -4.25225556e-01 9.65958834e-01
2.99049944e-01 -3.03782254e-01 6.00673735e-01 -1.87317073e-01
1.99457780e-01 6.23914242e-01 -4.52785134e-01 9.67080891e-03
7.95551181e-01 2.23238453e-01 8.49388191e-04 -5.32641828e-01
-9.71619964e-01 3.95305157e-01 2.37737730e-01 5.77953011e-02
2.74990201e-01 -8.35399270e-01 -4.55195308e-01 5.01645915e-02
-9.05376375e-02 2.23683596e-01 4.81767684e-01 8.76583636e-01
3.87226976e-02 2.06801146e-01 4.19536710e-01 -3.59404594e-01
-1.43113422e+00 5.90099633e-01 -2.06716374e-01 -6.77202523e-01
-5.28395586e-02 1.10316265e+00 1.57713115e-01 -2.44860932e-01
2.69528508e-01 -2.43041441e-01 -1.51176199e-01 2.25470945e-01
2.86827296e-01 5.06525457e-01 -3.93809944e-01 -1.38624273e-02
-1.77718297e-01 8.84360373e-02 -3.45510006e-01 -4.61364090e-01
1.49999452e+00 -3.13149005e-01 -3.63579214e-01 4.91085082e-01
1.09800363e+00 1.76541358e-01 -1.34732676e+00 -4.47979718e-01
-3.31730135e-02 -6.72876596e-01 -1.24557309e-01 -6.17429376e-01
-8.67273927e-01 6.28789783e-01 3.60919684e-01 5.60406268e-01
9.42607999e-01 3.70846212e-01 5.62678516e-01 4.33419228e-01
6.08115196e-01 -1.00646961e+00 -1.72391564e-01 3.12207699e-01
2.76871085e-01 -9.76766109e-01 2.44884387e-01 -4.27566230e-01
-4.83988971e-01 5.50608337e-01 2.25815490e-01 -4.79324795e-02
1.16142106e+00 5.76470327e-03 -7.32741296e-01 -3.64909172e-01
-1.12065864e+00 -2.37630993e-01 1.76263973e-01 -1.78950489e-01
1.96772203e-01 4.49872226e-01 -5.92238963e-01 1.04432297e+00
-2.36358896e-01 -4.26798128e-02 3.37051392e-01 9.07361805e-01
-5.41945815e-01 -1.59630811e+00 -1.49631515e-01 1.05720437e+00
-7.83448875e-01 -2.55372912e-01 -6.40230998e-02 3.99823248e-01
1.52529776e-01 1.01158929e+00 3.74683142e-02 -1.20067559e-01
6.72516599e-02 6.89825267e-02 4.41387892e-01 -7.20078528e-01
-3.40095311e-01 6.18383110e-01 9.97847617e-02 -3.13827068e-01
-1.65927663e-01 -1.06748700e+00 -6.83616579e-01 -2.14873940e-01
-5.04447281e-01 5.43400347e-01 6.48318529e-01 9.98319209e-01
2.97353864e-01 1.62125036e-01 7.05120742e-01 -4.80808705e-01
-8.21144044e-01 -7.84804404e-01 -1.03358996e+00 1.20186388e-01
-2.09581503e-03 -7.98638046e-01 -4.92025107e-01 -3.53957824e-02] | [4.781757831573486, 3.39837908744812] |
b962f168-b1be-43f9-be3d-31fc70c699c1 | micro-expression-generation-with-thin-plate | null | null | https://dl.acm.org/doi/abs/10.1145/3503161.3551609 | https://dl.acm.org/doi/abs/10.1145/3503161.3551609 | Micro Expression Generation with Thin-plate Spline Motion Model and Face Parsing | Micro-expression generation aims at transfering the expression from the driving videos to the source images, which can be viewed as a motion transfer task. Recently, several works have been proposed to tackle this problem and achieve great performance. However, due to the intrinsic complexity of the face motion and different attributes of face regions, the task still remains challenging. In this paper, we propose an end-to-end unsupervised motion transfer network to tackle this challenge. As the motion of the face is non-rigid, we adopt an effective and flexible thin-plate spline motion estimation method to estimate the optical flow of the face motion. What's more, we find that several faces with eyeglasses show weird deformation in motion transfering. Thus, we introduce face parsing method to pay specific attention to the eyeglasses regions to ensure the reasonability of the deformation. We conduct several experiments on the provided datasets of the ACM MM 2022 micro-expression grand challenge (MEGC2022) and compare our method with several other typical methods. In comparison, our method shows the best performance. We (Team: USTC-IAT-United) also compare our method with other competitors' in MEGC2022, and the expert evaluation results show that our method performs best, which verifies the effectiveness of our method. Our code is available at https://github.com/HowToNameMe/micro-expression | ['Qiang Ling', 'Fang Gao', 'Peng He', 'Zhongpeng Cai', 'Guochen Xie', 'Jun Yu'] | 2022-10-10 | null | null | null | mm-22-proceedings-of-the-30th-acm | ['face-parsing'] | ['computer-vision'] | [-1.41302660e-01 -5.48002534e-02 -2.39751358e-02 -6.50491834e-01
-6.04542136e-01 -2.52876222e-01 3.31437796e-01 -1.00876236e+00
-2.38667890e-01 5.73188245e-01 2.75346905e-01 2.01742634e-01
2.29945794e-01 -3.43601644e-01 -7.57408500e-01 -8.51344764e-01
1.74073026e-01 4.57874723e-02 1.16645070e-02 -2.73870707e-01
2.18737759e-02 2.13843793e-01 -1.32120621e+00 2.65249878e-01
6.66177750e-01 8.02987874e-01 1.58848360e-01 4.38120365e-01
-5.73445112e-03 7.22804785e-01 -3.68866354e-01 -5.13996303e-01
2.23941207e-02 -6.57360554e-01 -9.56933141e-01 1.07061401e-01
5.25218248e-01 -4.45391774e-01 -1.56467736e-01 1.24578476e+00
6.70181572e-01 1.52998969e-01 5.64053297e-01 -1.57789874e+00
-5.81030309e-01 7.54315704e-02 -9.08232689e-01 8.21648464e-02
3.47492754e-01 1.08268037e-01 4.39906001e-01 -9.86887932e-01
8.62126112e-01 1.50488579e+00 4.82671618e-01 1.13955176e+00
-6.89531565e-01 -9.90789294e-01 1.92536116e-01 2.26046771e-01
-1.40305853e+00 -9.10223305e-01 8.83450627e-01 -4.53306615e-01
4.37521011e-01 5.41560426e-02 3.06420177e-01 1.39940560e+00
-8.68762732e-02 8.63346517e-01 9.64501619e-01 -1.91704333e-01
1.28301140e-02 1.10229775e-01 -1.64555028e-01 8.96934271e-01
-3.39273304e-01 -1.04774475e-01 -2.97897965e-01 1.06375903e-01
7.59500086e-01 -3.43498439e-01 -5.66744506e-01 1.99243110e-02
-8.90719175e-01 4.95041698e-01 2.44903401e-01 2.92618603e-01
-5.33884913e-02 3.58765960e-01 2.52796382e-01 6.63469210e-02
6.65146768e-01 -4.56554085e-01 -2.91812837e-01 -3.12446952e-01
-7.40877151e-01 1.88875526e-01 5.80985308e-01 1.00749969e+00
6.69178724e-01 -3.47271450e-02 -5.72228357e-02 9.61144567e-01
6.01096153e-01 3.66431087e-01 4.34847087e-01 -1.20212638e+00
6.18476808e-01 2.75040269e-01 6.59341067e-02 -1.17453516e+00
-1.73197672e-01 8.68839920e-02 -1.02555990e+00 1.46856010e-01
3.51719975e-01 -5.32239556e-01 -6.97276354e-01 2.15404296e+00
5.79446077e-01 6.36846423e-01 7.35286204e-03 1.15774560e+00
1.00708830e+00 9.21781301e-01 1.60550177e-01 -4.01284099e-01
1.24585247e+00 -1.06374788e+00 -9.64689255e-01 -3.76642533e-02
7.52854407e-01 -7.26165414e-01 9.42891836e-01 7.86785968e-03
-9.87115204e-01 -6.49078250e-01 -8.14220250e-01 -4.44306619e-03
5.02869338e-02 3.47269237e-01 4.35196996e-01 5.95854819e-01
-1.06885147e+00 5.38449049e-01 -8.59050453e-01 -3.08955520e-01
5.79707205e-01 4.93117660e-01 -6.60371780e-01 2.42314264e-02
-1.16218245e+00 5.99973202e-01 4.44591045e-02 5.55174589e-01
-6.09013796e-01 -4.41813827e-01 -7.09714234e-01 -2.57321984e-01
2.13661715e-01 -6.45219147e-01 1.16571808e+00 -1.34507811e+00
-1.82215714e+00 8.53135347e-01 -4.79862332e-01 2.17545465e-01
6.92757607e-01 -2.55443513e-01 -3.35282773e-01 4.97133285e-02
-1.53479576e-02 1.03369927e+00 7.65240192e-01 -9.56016541e-01
-4.70268786e-01 -4.12959427e-01 4.23199497e-02 1.66801319e-01
-4.02561814e-01 3.98598552e-01 -8.93728018e-01 -6.64344549e-01
-2.48465300e-01 -1.25214911e+00 1.89412221e-01 2.49098204e-02
-2.10897088e-01 -3.63852650e-01 1.06455970e+00 -6.34581983e-01
1.21680963e+00 -2.47517729e+00 2.77797639e-01 -2.18150750e-01
-8.59011039e-02 2.32751429e-01 -1.47242650e-01 -5.10855131e-02
-1.85681522e-01 1.80336744e-01 -3.62714410e-01 -5.74331403e-01
-1.29976198e-01 -1.10185996e-01 -3.63992527e-02 6.34210348e-01
2.54892707e-01 9.06562507e-01 -5.96853375e-01 -6.92126989e-01
-1.84407786e-01 7.24050760e-01 -4.71535474e-01 3.33987653e-01
-1.30170211e-01 8.32899332e-01 -5.92619240e-01 6.47873163e-01
1.04712796e+00 -6.68056607e-02 -6.99854083e-03 -2.47015059e-01
1.14079051e-01 -3.00585151e-01 -1.09582269e+00 1.77465522e+00
-2.81455457e-01 6.40441358e-01 3.11658233e-01 -6.98894680e-01
7.18032122e-01 4.29123163e-01 5.12557149e-01 -5.01882553e-01
2.13271797e-01 1.18084155e-01 -9.31303352e-02 -8.86449158e-01
6.59222752e-02 -2.01313078e-01 2.55062550e-01 2.06600562e-01
-1.01590142e-01 1.10800810e-01 -1.63456813e-01 -4.17544367e-03
7.32959986e-01 6.65342331e-01 -2.28089541e-01 -3.16570640e-01
7.75112748e-01 -4.34359193e-01 9.28132772e-01 -1.49032906e-01
-4.45154071e-01 7.41595864e-01 4.77558762e-01 -2.47594431e-01
-6.51092410e-01 -6.22677922e-01 1.27668440e-01 9.08758879e-01
1.88120827e-01 -3.88939083e-01 -1.24150300e+00 -8.12495649e-01
-3.33261162e-01 1.76972941e-01 -5.67394853e-01 -1.92250148e-03
-8.13956022e-01 -8.96725237e-01 5.77752829e-01 5.88820875e-01
8.63926053e-01 -1.21046579e+00 -2.28321388e-01 1.98444556e-02
-6.26859307e-01 -1.48146522e+00 -9.05215561e-01 -7.63503075e-01
-5.67734659e-01 -8.18700552e-01 -9.83569384e-01 -9.94318247e-01
6.62021637e-01 2.18399018e-01 6.48350060e-01 1.87778413e-01
-8.03609863e-02 1.62970498e-01 -3.07657987e-01 -1.23054042e-01
-2.67046362e-01 -4.64708358e-02 5.63108400e-02 4.39592481e-01
7.02549815e-02 -4.23404723e-01 -7.48167217e-01 5.30722082e-01
-7.74937212e-01 1.33607626e-01 2.70840734e-01 5.54111958e-01
2.63009548e-01 -1.55381337e-01 5.41241229e-01 -7.42866099e-01
4.51933533e-01 -4.85560209e-01 -5.94067812e-01 1.48884788e-01
-1.72395051e-01 -1.11379877e-01 3.49741131e-01 -4.03645515e-01
-1.35070741e+00 3.62171382e-01 -4.02780056e-01 -5.88055491e-01
-1.90327168e-01 7.34945014e-02 -7.34942436e-01 -9.67133790e-02
6.05193637e-02 -1.44654084e-02 1.63677752e-01 -3.87902200e-01
2.41013631e-01 7.79722691e-01 5.27861059e-01 -5.37318110e-01
5.95500231e-01 5.98441422e-01 6.19051903e-02 -7.72791028e-01
-5.49452782e-01 -8.90368596e-02 -4.29283112e-01 -4.63320106e-01
1.17447519e+00 -9.72987056e-01 -9.98050988e-01 8.37238729e-01
-1.49631906e+00 -3.94661009e-01 4.71052855e-01 4.07109767e-01
-5.85207224e-01 4.72479612e-01 -7.24101365e-01 -7.13730514e-01
-3.79308105e-01 -1.40823853e+00 1.07431614e+00 4.02844936e-01
-8.48869979e-02 -9.60094213e-01 5.20811044e-02 5.03311753e-01
3.26103330e-01 4.89332557e-01 5.64189494e-01 2.89263036e-02
-4.32066023e-01 1.89932749e-01 -1.50040314e-01 3.79487634e-01
1.12549409e-01 2.51141667e-01 -1.18223047e+00 -2.67054975e-01
1.24832354e-01 -1.98284194e-01 6.78864360e-01 3.80023897e-01
1.19858170e+00 -2.76694924e-01 -3.64958823e-01 8.76259565e-01
1.06983697e+00 2.66044587e-01 9.71545935e-01 1.05842784e-01
9.15930212e-01 1.10231972e+00 8.10537159e-01 2.46824801e-01
4.98735547e-01 9.78386760e-01 3.89255285e-01 -5.07528260e-02
-5.11768237e-02 -1.00004524e-01 8.77215147e-01 8.13953578e-01
-3.42100978e-01 -2.92207718e-01 -6.45738840e-01 2.94803083e-01
-1.94378662e+00 -9.31401014e-01 -3.91859859e-01 1.88916528e+00
6.16657257e-01 -2.28231832e-01 1.08016580e-01 -3.62271160e-01
9.01661992e-01 8.18654001e-02 -2.78884917e-01 -1.37781069e-01
-8.69073719e-02 -1.98636763e-02 -3.29782963e-02 5.55463076e-01
-1.01881003e+00 1.19507992e+00 5.27991009e+00 9.35009241e-01
-1.26700234e+00 3.79595339e-01 9.16476905e-01 -1.67087972e-01
-5.33458814e-02 -6.77604005e-02 -9.24459338e-01 8.22668195e-01
8.74948561e-01 8.96500498e-02 1.90280408e-01 4.08958763e-01
4.61279243e-01 1.02960333e-01 -9.73734438e-01 1.18286169e+00
5.59900217e-02 -1.05261087e+00 -3.13825369e-01 1.22508489e-01
5.45875549e-01 -3.61317992e-01 2.75598243e-02 2.40396813e-01
-2.07444251e-01 -1.05491590e+00 5.48387825e-01 5.11417568e-01
8.11781049e-01 -7.80285716e-01 6.06930852e-01 2.66274095e-01
-1.21045065e+00 2.47897938e-01 -2.99641043e-01 1.60174787e-01
2.47563779e-01 3.18549961e-01 -3.65319639e-01 5.16705751e-01
7.50028372e-01 6.60427570e-01 -3.25819075e-01 4.96114850e-01
-2.41695404e-01 5.84489107e-01 -1.47204116e-01 2.13335335e-01
6.06757924e-02 -3.71491015e-01 4.18934196e-01 1.17535090e+00
5.67470133e-01 2.52291471e-01 -3.10695976e-01 9.27810907e-01
-3.53095323e-01 2.87698001e-01 -5.59013128e-01 1.45094186e-01
9.78400707e-02 1.42559171e+00 -3.98815930e-01 -2.45812386e-01
-4.73316729e-01 1.23025405e+00 2.46491671e-01 3.20578635e-01
-1.26397765e+00 -1.45702779e-01 7.32457042e-01 7.19155073e-02
1.33082286e-01 -1.59962922e-02 4.40888673e-01 -1.20636737e+00
2.52935767e-01 -6.72499597e-01 1.14842571e-01 -9.35549021e-01
-1.01895869e+00 8.52909088e-01 -2.22588256e-02 -1.21440876e+00
-2.89799273e-01 -5.38983524e-01 -7.13552594e-01 7.33490407e-01
-1.21141446e+00 -1.19272399e+00 -5.90615749e-01 6.76976442e-01
5.09959519e-01 -9.85857248e-02 5.60021996e-01 6.88733160e-01
-8.87776613e-01 8.38689506e-01 -1.99620649e-01 2.89121807e-01
1.01174128e+00 -6.53567195e-01 3.95272523e-01 6.39353156e-01
-1.96619809e-01 4.26803470e-01 4.23400789e-01 -5.30779362e-01
-1.39616525e+00 -1.13764215e+00 6.27107680e-01 -3.56729388e-01
3.45435977e-01 -4.82271731e-01 -1.07497871e+00 6.63022876e-01
4.30616438e-01 2.49940395e-01 4.47683483e-01 -4.13487822e-01
-6.10340200e-03 -1.12329245e-01 -1.09387207e+00 5.03160894e-01
1.28056908e+00 -2.50179857e-01 -1.74979791e-01 1.75990880e-01
6.60566032e-01 -3.96600187e-01 -7.37788022e-01 6.32142484e-01
6.11816168e-01 -9.47296023e-01 6.15959227e-01 -5.40438414e-01
6.56816483e-01 -3.64746898e-01 -1.01423644e-01 -1.09994662e+00
-1.39855042e-01 -8.73141050e-01 1.49830028e-01 1.78192472e+00
1.28366187e-01 -5.19180834e-01 8.46988082e-01 6.92034006e-01
1.45230934e-01 -7.56030440e-01 -8.76861036e-01 -6.56154871e-01
1.40873432e-01 -2.10715801e-01 5.53361535e-01 1.13827729e+00
-1.79410711e-01 4.75970328e-01 -6.92352533e-01 9.29975882e-02
3.68716508e-01 4.20945808e-02 8.82158041e-01 -7.76151955e-01
-1.96548522e-01 -2.92407632e-01 -3.32138658e-01 -1.15867615e+00
6.86529875e-01 -8.31536531e-01 6.11594506e-02 -1.07836151e+00
2.81494081e-01 -1.53695583e-01 -1.52186584e-02 4.21575338e-01
-2.35641375e-01 2.59426951e-01 2.70191610e-01 2.40209892e-01
-3.30658495e-01 7.05522478e-01 1.47197282e+00 -5.08776419e-02
-1.14815101e-01 2.40483657e-02 -4.87396389e-01 8.14089417e-01
8.53625596e-01 -3.59658480e-01 -4.00595546e-01 -6.35804296e-01
-1.18205823e-01 2.40847796e-01 3.89059663e-01 -6.95233464e-01
1.37442276e-01 -2.12171227e-01 2.08673790e-01 -3.15747261e-01
4.59219307e-01 -6.04097307e-01 3.54565412e-01 2.12204859e-01
5.65263443e-02 1.53949544e-01 2.39053622e-01 1.84076130e-01
-3.03144336e-01 -1.65353671e-01 9.80657339e-01 2.66564488e-02
-7.48193681e-01 5.82314372e-01 -1.74886644e-01 1.03629977e-01
1.00135648e+00 -5.95122986e-02 -3.68075162e-01 -5.14707327e-01
-6.53869927e-01 1.65270865e-01 4.42279786e-01 6.15357101e-01
6.48778617e-01 -1.52757311e+00 -8.53676498e-01 8.85013491e-02
-9.33481231e-02 -1.64450295e-02 4.45865214e-01 1.13145697e+00
-4.64051783e-01 1.47892367e-02 -2.36049861e-01 -5.51761866e-01
-1.48097718e+00 2.69545913e-01 4.65495735e-01 1.90585837e-01
-4.78769422e-01 8.51313829e-01 5.13011992e-01 -1.94184035e-01
3.85529511e-02 1.24183167e-02 -2.13056773e-01 -1.10975159e-02
5.03202081e-01 2.95045197e-01 -2.18707323e-01 -1.08975434e+00
-6.14779592e-01 1.04397011e+00 4.28050756e-02 -1.22402683e-01
1.19897497e+00 -2.81811655e-01 -2.57898420e-01 4.85483669e-02
1.64966857e+00 3.41383815e-02 -1.34369206e+00 2.64845103e-01
-1.36715546e-01 -4.07031715e-01 -2.72208244e-01 -2.79766440e-01
-1.55990160e+00 1.02535093e+00 7.90139735e-01 -4.48023200e-01
1.27664542e+00 -1.16247363e-01 8.65470350e-01 -1.04900084e-01
2.29544103e-01 -9.41061676e-01 8.16760212e-02 3.71649623e-01
9.56118464e-01 -1.32212305e+00 -3.29690099e-01 -7.45703697e-01
-7.03531146e-01 9.83638763e-01 1.04241812e+00 1.20777182e-01
5.82517087e-01 2.15325490e-01 2.21999407e-01 7.10802749e-02
-7.28212118e-01 5.33075035e-02 9.40615088e-02 3.55365396e-01
5.70182085e-01 -1.83912247e-01 -2.57148832e-01 5.46629429e-01
-1.04637563e-01 2.98504353e-01 3.62298816e-01 7.26928294e-01
-8.32191184e-02 -1.07694483e+00 -3.00457269e-01 -2.70908922e-01
-6.21963263e-01 1.85314998e-01 -3.30731541e-01 7.99603581e-01
2.59691775e-01 8.47750545e-01 -8.94115865e-02 -4.04399663e-01
3.42534184e-01 6.56807348e-02 5.72802365e-01 -2.46473417e-01
-2.47749165e-01 3.76590520e-01 7.73331011e-03 -6.60449564e-01
-7.16180384e-01 -4.54595208e-01 -1.60368109e+00 -4.44859177e-01
-7.99343064e-02 6.01811819e-02 5.91905475e-01 7.88066328e-01
5.50923169e-01 3.05392206e-01 7.90342212e-01 -7.30360270e-01
-2.42196292e-01 -9.45365489e-01 -3.32063794e-01 7.67444551e-01
1.28362417e-01 -7.15910137e-01 -3.12415361e-01 3.70209843e-01] | [13.001837730407715, -0.19633445143699646] |
cf13b975-dcf7-4837-bd07-5f7bd250edf8 | recurrent-dynamic-embedding-for-video-object | 2205.03761 | null | https://arxiv.org/abs/2205.03761v1 | https://arxiv.org/pdf/2205.03761v1.pdf | Recurrent Dynamic Embedding for Video Object Segmentation | Space-time memory (STM) based video object segmentation (VOS) networks usually keep increasing memory bank every several frames, which shows excellent performance. However, 1) the hardware cannot withstand the ever-increasing memory requirements as the video length increases. 2) Storing lots of information inevitably introduces lots of noise, which is not conducive to reading the most important information from the memory bank. In this paper, we propose a Recurrent Dynamic Embedding (RDE) to build a memory bank of constant size. Specifically, we explicitly generate and update RDE by the proposed Spatio-temporal Aggregation Module (SAM), which exploits the cue of historical information. To avoid error accumulation owing to the recurrent usage of SAM, we propose an unbiased guidance loss during the training stage, which makes SAM more robust in long videos. Moreover, the predicted masks in the memory bank are inaccurate due to the inaccurate network inference, which affects the segmentation of the query frame. To address this problem, we design a novel self-correction strategy so that the network can repair the embeddings of masks with different qualities in the memory bank. Extensive experiments show our method achieves the best tradeoff between performance and speed. Code is available at https://github.com/Limingxing00/RDE-VOS-CVPR2022. | ['Dong Liu', 'Pan Pan', 'Bang Zhang', 'Zhiwei Xiong', 'Li Hu', 'Mingxing Li'] | 2022-05-08 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Li_Recurrent_Dynamic_Embedding_for_Video_Object_Segmentation_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Li_Recurrent_Dynamic_Embedding_for_Video_Object_Segmentation_CVPR_2022_paper.pdf | cvpr-2022-1 | ['semi-supervised-video-object-segmentation'] | ['computer-vision'] | [-1.37517396e-02 -2.22846597e-01 -1.60641402e-01 -8.91869515e-02
-3.06720316e-01 -9.57702845e-02 -6.19665980e-02 -2.38331288e-01
-4.81553823e-01 4.65180784e-01 3.03320028e-02 -1.83495685e-01
2.10004240e-01 -8.28043103e-01 -8.35720181e-01 -7.93174326e-01
3.57367665e-01 -2.57768154e-01 8.29355597e-01 2.41523460e-01
2.13364452e-01 2.47175157e-01 -1.34111118e+00 1.34511635e-01
8.49120200e-01 1.15216386e+00 7.21198380e-01 4.41842496e-01
-2.93566197e-01 7.62965024e-01 -6.64155424e-01 -1.38308123e-01
1.96857125e-01 -4.24190700e-01 -4.96218264e-01 -2.15679146e-02
1.86111435e-01 -6.73883379e-01 -9.12696958e-01 1.05421853e+00
3.92820656e-01 2.12404221e-01 1.37489527e-01 -1.09476304e+00
-6.18266404e-01 5.01308560e-01 -7.35205770e-01 5.24734080e-01
-1.79059178e-01 3.86721134e-01 5.24823189e-01 -8.40068400e-01
4.73450899e-01 9.41864371e-01 5.78531384e-01 7.40145326e-01
-6.78104758e-01 -8.11684132e-01 3.35565209e-01 4.81119186e-01
-1.66883051e+00 -5.59062719e-01 6.57057941e-01 -2.20853925e-01
7.10379481e-01 1.97122917e-01 7.09201097e-01 7.75092840e-01
1.26237586e-01 1.01867056e+00 4.53523785e-01 -9.18092858e-03
1.00571945e-01 8.70754346e-02 1.17256708e-01 7.36386895e-01
1.88256785e-01 -1.19329832e-01 -7.31833756e-01 2.37420470e-01
1.11032701e+00 2.50711948e-01 -4.72939789e-01 6.74565136e-02
-1.02199769e+00 4.67591882e-01 4.18472290e-01 3.21044177e-01
-2.40690544e-01 3.14678699e-01 4.97027397e-01 1.46811247e-01
2.97145844e-01 -3.66338231e-02 -3.20965976e-01 -2.28168994e-01
-1.16077256e+00 -2.79840320e-01 2.57417262e-01 1.06426096e+00
7.24264205e-01 3.49461228e-01 -6.99402839e-02 6.55235291e-01
4.07521605e-01 3.18218082e-01 6.44378901e-01 -9.36689436e-01
6.32896423e-01 5.89592457e-01 9.44372173e-03 -1.32704687e+00
-9.15795192e-02 -3.48725736e-01 -1.06445229e+00 -2.32362151e-01
1.81704938e-01 -9.06860158e-02 -8.59044075e-01 1.63530922e+00
3.50256890e-01 7.31378257e-01 -1.75256908e-01 1.02354264e+00
7.42246807e-01 1.11401284e+00 -1.00780785e-01 -5.28807700e-01
9.74597514e-01 -1.12796736e+00 -7.68106818e-01 -2.39442393e-01
4.65213954e-01 -5.69123805e-01 1.05981350e+00 1.18856274e-01
-1.08152592e+00 -8.09436142e-01 -1.29874432e+00 -6.72910362e-03
2.20585335e-02 2.19276428e-01 2.33381584e-01 3.27187538e-01
-1.03901339e+00 5.80670893e-01 -1.04107416e+00 1.85498483e-02
5.39659381e-01 2.80521512e-01 1.13259971e-01 7.49719441e-02
-1.18934262e+00 4.13104326e-01 4.37631339e-01 5.00589073e-01
-6.86454594e-01 -6.15960896e-01 -7.25697517e-01 3.56609514e-03
5.59914351e-01 -4.35598850e-01 9.19836044e-01 -1.01893210e+00
-1.34418201e+00 2.85705447e-01 -4.06288058e-01 -5.31039119e-01
6.41900539e-01 -2.24092886e-01 -4.68174785e-01 3.30146700e-01
-4.06727828e-02 6.97556198e-01 1.08996582e+00 -8.96229863e-01
-5.08718312e-01 -1.99405521e-01 -2.60434568e-01 -1.93556212e-02
-7.53573477e-01 -1.86248630e-01 -1.01227307e+00 -8.97597134e-01
2.10540950e-01 -6.60916507e-01 -2.29226843e-01 1.05282344e-01
-3.02665889e-01 -7.68589973e-03 1.04754281e+00 -7.96257913e-01
1.87533689e+00 -2.60563302e+00 3.13707441e-03 8.29950422e-02
2.37049699e-01 5.53562462e-01 -9.50469673e-02 -1.27744852e-02
2.23835066e-01 2.26590261e-01 -2.21827492e-01 -2.60026932e-01
-4.92647260e-01 2.98218459e-01 -4.79120731e-01 2.13071659e-01
1.32921383e-01 8.85428011e-01 -6.98564589e-01 -7.21227705e-01
1.27611160e-01 3.99748594e-01 -4.92631823e-01 1.97366148e-01
-1.97320774e-01 2.95103967e-01 -5.75503767e-01 4.47402209e-01
7.42495656e-01 -4.84121740e-01 6.87336400e-02 -4.41905200e-01
-2.22891182e-01 1.29992291e-01 -1.25874627e+00 1.49297690e+00
-2.77869731e-01 6.88747227e-01 -1.27706945e-01 -7.73619235e-01
7.34061301e-01 1.84468687e-01 3.09295923e-01 -8.17535579e-01
1.75994024e-01 9.73365754e-02 -3.70783955e-01 -4.54139113e-01
6.29976749e-01 3.96261305e-01 2.96970159e-01 3.34411412e-01
-2.17915133e-01 5.20295680e-01 3.57582085e-02 3.12101692e-01
9.41227019e-01 -6.44914657e-02 -1.54515192e-01 1.13213420e-01
5.82363784e-01 -2.41160780e-01 1.12925816e+00 4.56774414e-01
-2.84796596e-01 5.52348435e-01 3.75514120e-01 -4.19160455e-01
-1.08277774e+00 -8.03298831e-01 9.98684764e-02 7.26584613e-01
6.22983992e-01 -4.32789534e-01 -7.79787660e-01 -4.08409327e-01
-2.12613583e-01 3.70521873e-01 -3.22193682e-01 -4.53291088e-01
-8.92834783e-01 -5.67535400e-01 4.32887673e-01 6.88419640e-01
1.03466809e+00 -9.98763382e-01 -7.06922650e-01 3.76156300e-01
-3.31343025e-01 -1.03566515e+00 -9.54651535e-01 -5.20439506e-01
-1.10555983e+00 -8.40620637e-01 -7.80392468e-01 -7.86038637e-01
7.64303207e-01 6.44673526e-01 6.21408582e-01 5.36514878e-01
-3.96194030e-03 5.78089990e-02 -2.71039128e-01 8.89395401e-02
-2.89617449e-01 1.22092530e-01 4.95757684e-02 1.12742677e-01
2.41661415e-01 -5.83322883e-01 -8.86016011e-01 3.93014461e-01
-1.04991925e+00 3.50687623e-01 4.72277790e-01 7.62723088e-01
8.02058637e-01 1.74736425e-01 5.10782242e-01 -5.19931912e-01
2.23016635e-01 -4.64141071e-01 -7.68186629e-01 2.17102215e-01
-5.95723450e-01 -1.25538260e-01 6.17609978e-01 -6.70530379e-01
-8.83405030e-01 -6.75060749e-02 -1.55024752e-01 -8.46094787e-01
3.45345289e-01 2.91686058e-01 -2.72130013e-01 1.18245259e-01
1.65901072e-02 5.86871564e-01 4.59159799e-02 -4.32791024e-01
5.88930212e-02 6.73999250e-01 4.96007591e-01 -1.36945412e-01
5.49076021e-01 3.15008134e-01 -3.03128064e-01 -7.65103340e-01
-5.59372067e-01 -2.10803837e-01 -2.62420177e-01 -3.30056697e-01
7.52422631e-01 -9.03382719e-01 -5.99021316e-01 7.58956730e-01
-1.26974690e+00 -3.95656228e-01 -7.82942697e-02 2.79463381e-01
-2.17213929e-01 5.83957851e-01 -1.00674701e+00 -7.35256910e-01
-3.81868005e-01 -1.15064621e+00 6.54226840e-01 6.99921370e-01
1.05613619e-01 -5.25103390e-01 -3.94287020e-01 1.70185179e-01
2.73313671e-01 -2.36921273e-02 5.72634935e-01 -7.25844502e-02
-1.25008678e+00 2.51776606e-01 -3.94540697e-01 4.74089026e-01
-3.95532139e-03 1.67729989e-01 -6.93187177e-01 -2.96357214e-01
1.79227293e-01 2.10577607e-01 1.12438869e+00 3.64221543e-01
1.46947443e+00 -5.43944359e-01 -3.19118172e-01 6.56260192e-01
1.35298038e+00 5.25057673e-01 9.17447865e-01 2.98461795e-01
9.41791236e-01 1.77195996e-01 8.28019023e-01 4.62048382e-01
3.50307405e-01 5.57716489e-01 2.61920363e-01 1.45996571e-01
-1.15174145e-01 -4.00537819e-01 5.05253673e-01 1.26922154e+00
2.65595257e-01 -3.94887447e-01 -6.94848716e-01 6.37894273e-01
-1.98780704e+00 -8.80513787e-01 -4.61995006e-02 2.19200826e+00
9.86342371e-01 3.78480434e-01 -1.20348930e-01 4.89136502e-02
8.54757607e-01 4.91798788e-01 -7.82019317e-01 -1.08604915e-01
-6.96641505e-02 -2.41153553e-01 5.70810616e-01 3.19590896e-01
-8.15485001e-01 9.11674976e-01 5.26959229e+00 1.11790621e+00
-1.30069435e+00 3.17395061e-01 7.63284445e-01 -2.20540598e-01
-1.80151612e-01 -1.07031964e-01 -9.24421370e-01 1.19649005e+00
9.14787054e-01 6.55535907e-02 3.23665559e-01 6.40202403e-01
3.53318214e-01 -2.25202158e-01 -6.14894152e-01 1.08473384e+00
-9.04797614e-02 -1.68084204e+00 1.40244097e-01 -1.19979002e-01
5.37558317e-01 -1.47402465e-01 1.36141345e-01 -3.71079370e-02
-3.78356665e-01 -5.15468419e-01 7.35971749e-01 6.92842066e-01
8.60462844e-01 -7.38432586e-01 6.12338603e-01 2.87174851e-01
-1.40758097e+00 -1.55776203e-01 -5.00797451e-01 2.92256385e-01
3.71293187e-01 7.75061548e-01 -3.65529597e-01 2.97571599e-01
8.20128143e-01 7.49100924e-01 -4.69644755e-01 9.26728487e-01
-4.20487188e-02 7.62388110e-01 -3.35003346e-01 1.19588256e-01
1.34197712e-01 -1.69817045e-01 5.76751947e-01 9.66376901e-01
4.64128494e-01 2.49402210e-01 -8.83359462e-02 7.14151323e-01
-2.01085076e-01 -1.12700023e-01 -2.69922942e-01 -1.45727009e-01
9.16029334e-01 9.09136891e-01 -7.49764740e-01 -4.70037311e-01
-3.68637681e-01 1.19014227e+00 1.07942238e-01 3.36423784e-01
-1.09314823e+00 -3.30803573e-01 5.87707520e-01 1.62402600e-01
5.17864883e-01 -3.91746342e-01 -2.05431864e-01 -1.10474980e+00
4.13079053e-01 -5.80751240e-01 2.81635225e-01 -6.29692197e-01
-8.52788627e-01 5.26454151e-01 -4.25189793e-01 -1.41294038e+00
1.85784817e-01 -9.12959278e-02 -5.89494467e-01 5.23213744e-01
-1.32235789e+00 -6.85494900e-01 -3.70936543e-01 3.34313542e-01
6.95442855e-01 1.57866701e-02 1.84491709e-01 7.56121278e-01
-1.10603368e+00 8.81187379e-01 -2.15398166e-02 1.24244630e-01
5.74640632e-01 -5.32678604e-01 4.46538270e-01 1.09349692e+00
-3.62393446e-02 5.70792675e-01 4.63437885e-01 -8.13493907e-01
-1.29204774e+00 -1.39465964e+00 5.56333482e-01 1.34511199e-02
4.74642754e-01 -1.47224426e-01 -1.44817519e+00 4.42178756e-01
-1.59856647e-01 3.44908424e-02 3.49400580e-01 -5.78636289e-01
-2.00076491e-01 -3.73117357e-01 -8.25412750e-01 6.76724613e-01
1.05305350e+00 -5.71981192e-01 -3.09860289e-01 6.18926659e-02
1.14872670e+00 -4.75677937e-01 -6.65122807e-01 3.70429844e-01
4.60188448e-01 -8.81441534e-01 7.02033401e-01 -7.25998357e-02
4.02783155e-01 -6.64243340e-01 1.64940402e-01 -7.44805872e-01
-1.43821880e-01 -5.92029870e-01 -5.51757097e-01 1.39204597e+00
2.18022183e-01 -6.66017115e-01 8.52912307e-01 5.42880476e-01
-4.38943654e-02 -1.02147436e+00 -1.00002849e+00 -7.18845963e-01
-4.32689995e-01 -2.25412101e-01 6.56545758e-01 7.32050955e-01
-4.66528475e-01 5.86843975e-02 -6.30455017e-01 3.15457702e-01
2.63404042e-01 8.59727338e-02 5.48949778e-01 -6.06923580e-01
-2.52087265e-01 -3.48708481e-01 -3.70440453e-01 -1.68616009e+00
-1.64624736e-01 -3.24681252e-01 5.05146161e-02 -1.28008604e+00
1.68465748e-01 -6.05129957e-01 -4.33620602e-01 2.83576339e-01
-3.45607281e-01 2.62903243e-01 2.58799672e-01 4.75520045e-01
-8.49989355e-01 6.87558949e-01 1.21099651e+00 1.58754680e-02
-3.03138286e-01 -1.66144103e-01 -3.57098758e-01 5.96563101e-01
9.87957060e-01 -4.49727863e-01 -4.61835384e-01 -7.79533684e-01
1.34078279e-01 1.20053649e-01 3.74949574e-01 -1.09608853e+00
5.77205360e-01 -6.25133514e-02 3.90908509e-01 -8.42736542e-01
2.99921811e-01 -6.87836170e-01 3.33187550e-01 5.39063275e-01
6.97897822e-02 2.66918987e-01 2.51853108e-01 7.10636497e-01
-2.32172802e-01 -2.36024380e-01 7.03661442e-01 8.55256245e-02
-9.04340386e-01 6.03000998e-01 -3.43335509e-01 -1.00442722e-01
8.92933786e-01 -5.46424210e-01 -2.20882490e-01 -1.67454258e-01
-4.07004267e-01 5.50431132e-01 5.90105236e-01 4.61841434e-01
9.04888153e-01 -1.33954477e+00 -2.42149755e-01 2.36841187e-01
-3.07804883e-01 3.09318602e-01 9.16772068e-01 8.27852964e-01
-6.11052215e-01 6.37635812e-02 9.46815982e-02 -5.49208701e-01
-1.16286552e+00 6.25373662e-01 2.59995192e-01 3.11820419e-03
-7.30162024e-01 8.71829569e-01 9.26659554e-02 3.38709593e-01
3.02041352e-01 -1.34910926e-01 -4.81355116e-02 7.65044242e-02
7.88659811e-01 4.62541819e-01 -2.96805471e-01 -5.06316423e-01
-2.17610940e-01 4.44955349e-01 -2.81361908e-01 1.63899854e-01
1.11319840e+00 -5.47782242e-01 -6.90741017e-02 4.42933530e-01
1.20843792e+00 -1.61751702e-01 -1.61898935e+00 -2.82533497e-01
-1.56447247e-01 -6.72269583e-01 8.48524570e-02 -5.55534773e-02
-1.53016794e+00 8.38944972e-01 6.75368607e-01 -2.60497835e-02
1.24647141e+00 -2.93007433e-01 1.58311284e+00 8.34303722e-02
2.58914888e-01 -1.37057292e+00 2.39541069e-01 4.18277353e-01
4.42490131e-01 -8.84802401e-01 -7.63137341e-02 -3.32658350e-01
-5.75736344e-01 9.04183865e-01 9.17817891e-01 2.38387305e-02
5.90096772e-01 2.19103977e-01 -3.70835401e-02 7.70342946e-02
-6.59688354e-01 2.59130687e-01 -4.87398468e-02 1.42278045e-01
-1.28173321e-01 -1.53835535e-01 -2.59388715e-01 6.41774356e-01
1.58297211e-01 -1.65028777e-02 6.00794911e-01 9.46868896e-01
-5.05371332e-01 -8.14368010e-01 -9.28095654e-02 4.40549374e-01
-3.72968942e-01 -9.27971527e-02 7.56121501e-02 3.93393576e-01
1.86147496e-01 7.68593073e-01 2.56653219e-01 -6.34288967e-01
1.16271771e-01 -2.21685320e-01 1.01778679e-01 -2.35967964e-01
-1.42565414e-01 1.15326039e-01 -3.87300313e-01 -7.65370071e-01
-4.19030160e-01 -3.24467778e-01 -1.47806501e+00 -5.59935868e-01
-4.97253984e-01 1.81380305e-02 2.00830802e-01 6.96462452e-01
6.20025635e-01 6.56081557e-01 7.17631578e-01 -5.23094893e-01
-2.36331001e-01 -6.17431045e-01 -3.23631644e-01 -1.64146628e-02
3.52631301e-01 -4.82436568e-01 -2.50155598e-01 1.02307023e-02] | [9.19150161743164, -0.07980822026729584] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.