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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
a931ed0a-2157-43af-8747-b3be6ac8830c | genq-automated-question-generation-to-support | 2305.16809 | null | https://arxiv.org/abs/2305.16809v2 | https://arxiv.org/pdf/2305.16809v2.pdf | GenQ: Automated Question Generation to Support Caregivers While Reading Stories with Children | When caregivers ask open--ended questions to motivate dialogue with children, it facilitates the child's reading comprehension skills.Although there is scope for use of technological tools, referred here as "intelligent tutoring systems", to scaffold this process, it is currently unclear whether existing intelligent systems that generate human--language like questions is beneficial. Additionally, training data used in the development of these automated question generation systems is typically sourced without attention to demographics, but people with different cultural backgrounds may ask different questions. As a part of a broader project to design an intelligent reading support app for Latinx children, we crowdsourced questions from Latinx caregivers and noncaregivers as well as caregivers and noncaregivers from other demographics. We examine variations in question--asking within this dataset mediated by individual, cultural, and contextual factors. We then design a system that automatically extracts templates from this data to generate open--ended questions that are representative of those asked by Latinx caregivers. | ['Art Glenberg', 'M. Adelaida Restrepo', 'Chris Blais', 'Tri Nguyen', 'Martha Michelle Soto Fernandez', 'Ligia E. Gomez', 'Arun Balajiee Lekshmi Narayanan'] | 2023-05-26 | null | null | null | null | ['reading-comprehension', 'question-generation'] | ['natural-language-processing', 'natural-language-processing'] | [ 1.37406141e-01 1.11254096e+00 8.68850425e-02 -8.13678741e-01
-9.62080061e-01 -7.81588793e-01 3.25169683e-01 3.88679683e-01
-1.45560592e-01 9.53047216e-01 1.08973706e+00 -5.56956947e-01
-8.09346065e-02 -5.11317134e-01 -2.67126977e-01 2.00677902e-01
6.42717481e-01 6.27051353e-01 7.99361169e-02 -3.64168435e-01
2.46076271e-01 -3.73500697e-02 -1.49141383e+00 2.38311812e-01
1.74145842e+00 -1.90971300e-01 1.36708885e-01 9.04699981e-01
-6.32211328e-01 1.49704421e+00 -9.09084976e-01 -4.16290551e-01
-2.63582557e-01 -9.11168933e-01 -1.30028057e+00 -4.72719260e-02
8.42393398e-01 -9.97920752e-01 3.84490676e-02 4.88174975e-01
5.22266388e-01 2.23474000e-02 7.39768207e-01 -1.18086720e+00
-1.06587839e+00 8.20851207e-01 3.12090546e-01 -1.97169594e-02
1.38623619e+00 1.83613777e-01 4.47565109e-01 -2.99907357e-01
7.66256332e-01 1.10259748e+00 6.71848357e-01 1.27049065e+00
-9.60366607e-01 -5.52624881e-01 -1.15281895e-01 -2.35250622e-01
-4.42647159e-01 -6.68899179e-01 4.98142481e-01 -5.71767688e-01
7.38977551e-01 -2.67931353e-03 8.42267215e-01 1.42711067e+00
-4.03791010e-01 6.10686541e-01 1.27605045e+00 -7.51070678e-01
2.54345536e-01 5.81823230e-01 5.66347480e-01 5.19150734e-01
1.21391542e-01 -4.12448555e-01 -8.67554605e-01 -4.22630459e-01
3.91189963e-01 -4.25499767e-01 -5.50581813e-01 -7.13399276e-02
-9.60716307e-01 1.22568917e+00 -1.11063654e-02 3.69831741e-01
-9.52005908e-02 -4.42763746e-01 3.91148822e-03 5.84378123e-01
3.45851302e-01 9.23162043e-01 -5.26352286e-01 -8.66009772e-01
-2.74371594e-01 4.41250235e-01 1.85785699e+00 1.44512308e+00
3.89125168e-01 -4.04919893e-01 -3.71707290e-01 1.24623644e+00
3.49810392e-01 4.30781603e-01 3.00187290e-01 -1.30021095e+00
5.36015213e-01 9.70250249e-01 2.85303921e-01 -2.30350897e-01
-2.98908114e-01 5.12093246e-01 1.52479127e-01 9.41281766e-02
1.09832692e+00 -1.00040936e+00 -4.05767113e-01 1.59899998e+00
5.42146206e-01 -7.95726538e-01 1.74576789e-01 7.25088239e-01
1.48914683e+00 8.87819678e-02 4.46108550e-01 5.02153635e-02
1.39643276e+00 -7.63109505e-01 -8.45260859e-01 -9.91938189e-02
7.67013013e-01 -7.32870698e-01 1.20559895e+00 8.08795635e-03
-1.30697560e+00 6.34836107e-02 -6.72778547e-01 -3.67143959e-01
-3.51364940e-01 -4.55029011e-01 5.76276302e-01 1.21726096e+00
-1.36977494e+00 -1.07741719e-02 -3.25640261e-01 -1.29882765e+00
3.99785966e-01 1.83707356e-01 -2.53677487e-01 -5.92258796e-02
-8.24231982e-01 1.08429492e+00 -3.54096532e-01 -8.93847883e-01
-3.73347044e-01 -9.75625634e-01 -8.11494589e-01 -2.39946768e-01
8.13973397e-02 -5.48785031e-01 2.02785683e+00 -1.25192726e+00
-1.59571540e+00 9.82393801e-01 -9.27547365e-02 1.18285507e-01
3.80816609e-01 -3.37835044e-01 1.23388849e-01 5.44126272e-01
5.56611776e-01 1.24086821e+00 2.79792517e-01 -9.12084639e-01
-5.69371521e-01 -4.55523908e-01 5.07718384e-01 6.03131890e-01
-4.20733333e-01 6.63818181e-01 2.57152349e-01 -5.11064768e-01
-2.34429538e-01 -4.39713448e-01 2.01236665e-01 2.58645266e-01
5.96506409e-02 -6.40360534e-01 4.79550034e-01 -9.66036260e-01
7.10844338e-01 -1.46487021e+00 -6.31251454e-01 -2.52312303e-01
5.36365032e-01 2.44491771e-01 -9.57569033e-02 7.99020171e-01
2.71558642e-01 1.08103491e-01 1.64579228e-01 3.41912448e-01
2.28574246e-01 -7.57141933e-02 2.11748526e-01 -3.91935736e-01
3.71736139e-01 6.35365427e-01 -1.27288783e+00 -5.95068157e-01
-1.07180439e-01 2.51874089e-01 -4.69911247e-01 8.21062982e-01
-5.90502203e-01 3.22145641e-01 -4.61851746e-01 5.53789020e-01
2.29571864e-01 -6.91735595e-02 -1.58461198e-01 9.51128066e-01
-7.99797699e-02 7.09851623e-01 -3.97428185e-01 1.25667572e+00
-1.62581384e-01 6.89786553e-01 4.72492129e-01 -5.37641197e-02
1.08736479e+00 4.75798160e-01 9.06899646e-02 -3.93698305e-01
3.11726630e-01 2.19497338e-01 -4.03388403e-02 -1.17270827e+00
3.76168549e-01 2.34946176e-01 3.54652017e-01 1.03171468e+00
-9.66796558e-03 -5.22556186e-01 -1.38244592e-02 2.48677075e-01
1.67184901e+00 2.70707875e-01 6.17026091e-02 -3.14592838e-01
3.00006731e-03 4.87549543e-01 -6.90374523e-02 7.85974622e-01
-4.99751002e-01 6.57073915e-01 4.32419807e-01 -2.33959537e-02
-7.94367850e-01 -8.94703746e-01 1.35897577e-01 1.30924845e+00
-5.97652435e-01 -8.90850723e-02 -1.68292499e+00 -7.08633959e-01
-2.57222235e-01 1.09316528e+00 -2.41593465e-01 1.40976831e-01
-2.64358938e-01 1.81268185e-01 8.20323586e-01 3.93904626e-01
3.70739192e-01 -1.22119236e+00 -1.17573845e+00 5.40554345e-01
-3.22464228e-01 -8.32580566e-01 -6.55480862e-01 -1.04335681e-01
-4.69094187e-01 -1.37093413e+00 -1.00643253e+00 -1.26710057e+00
6.05366170e-01 2.25907043e-01 1.36706007e+00 2.39842594e-01
1.56878248e-01 1.23423839e+00 -1.01278818e+00 -8.47098291e-01
-7.43269563e-01 3.45335007e-01 -5.63978076e-01 -8.18832994e-01
1.23714077e+00 -4.75814521e-01 -4.43909913e-01 8.64775926e-02
-5.89788973e-01 2.26907060e-01 6.72885254e-02 4.15883720e-01
-6.96686327e-01 -1.12465858e+00 1.03548205e+00 -9.14571643e-01
1.34850359e+00 -7.09471881e-01 -1.04029156e-01 4.83324409e-01
-2.40872473e-01 -1.58419088e-01 3.79255652e-01 -8.36190820e-01
-1.07984185e+00 -2.58868515e-01 -3.30225676e-01 3.26379895e-01
-9.31074142e-01 -1.42870516e-01 -1.16546310e-01 -1.43903717e-01
9.28230941e-01 -4.74457324e-01 5.20997822e-01 -2.13133037e-01
1.62064791e-01 1.23148644e+00 4.87707555e-01 -8.33523035e-01
4.84144062e-01 -5.44846237e-01 -9.09522235e-01 -1.02602923e+00
-2.97340006e-01 4.22336869e-02 -9.09703970e-02 -5.87235749e-01
1.19314229e+00 -6.95401192e-01 -8.28806579e-01 4.27514881e-01
-1.22417831e+00 -1.10906923e+00 -2.64935017e-01 4.54632998e-01
-4.15239751e-01 -3.11484277e-01 -4.24833953e-01 -9.35256600e-01
-5.34966946e-01 -9.16722715e-01 7.85320997e-01 8.80828083e-01
-1.20594609e+00 -1.08383524e+00 8.07183683e-02 1.08702099e+00
4.51248825e-01 9.89491194e-02 1.45080900e+00 -1.20995164e+00
-4.19365317e-01 2.51689374e-01 -2.52114564e-01 -9.84013900e-02
3.86057198e-02 -1.46514714e-01 -7.89526105e-01 4.01199877e-01
1.92432560e-03 -9.66487646e-01 -5.04251599e-01 -6.80178180e-02
1.23484291e-01 -6.22802794e-01 1.36552349e-01 -1.04539849e-01
8.07050288e-01 3.40211213e-01 2.38279015e-01 4.52354550e-02
3.83743614e-01 1.44082355e+00 4.16632444e-01 1.29760936e-01
1.37989068e+00 9.59936306e-02 -2.53204465e-01 4.28157598e-01
-9.48948637e-02 -6.87123597e-01 4.22336191e-01 6.88482285e-01
5.83120346e-01 4.06767381e-03 -1.18539333e+00 9.77240801e-01
-1.22420895e+00 -6.74047351e-01 -4.87833202e-01 1.55187929e+00
1.14024782e+00 -5.07998705e-01 2.88307369e-01 -8.48198384e-02
5.37938714e-01 -2.64158785e-01 -6.11728251e-01 -7.48483658e-01
1.61437422e-01 5.44920802e-01 -1.06012128e-01 7.03485370e-01
-1.00600526e-01 6.19197607e-01 6.61151648e+00 -1.18880183e-01
-4.93962944e-01 2.15191245e-02 7.12964654e-01 1.16293132e-01
-7.00000644e-01 -4.43708487e-02 -4.67082322e-01 2.62563974e-01
9.56722915e-01 -1.22264214e-01 5.42198122e-01 5.51270485e-01
2.12683380e-01 -5.71597755e-01 -1.42698920e+00 5.53880036e-01
5.09668767e-01 -6.82686090e-01 -6.35581732e-01 -3.21141303e-01
6.94252908e-01 -3.04892689e-01 -6.22298479e-01 4.43281502e-01
7.82639086e-01 -1.06309187e+00 5.61629236e-01 6.05513394e-01
4.64065313e-01 -2.58512229e-01 3.33148867e-01 4.32425112e-01
-2.86404282e-01 6.87677264e-02 2.84352481e-01 -4.76097375e-01
-1.05553471e-01 -2.33585492e-01 -1.26941419e+00 -4.53382611e-01
6.87744200e-01 -2.54036546e-01 -7.57115483e-01 8.28310966e-01
-2.58689791e-01 7.87889898e-01 -2.82796919e-01 -1.12593234e+00
-1.14197686e-01 -2.62992561e-01 1.14453748e-01 6.48838043e-01
3.31995636e-01 8.89053106e-01 -2.20767558e-01 7.03950882e-01
-5.75129986e-02 5.34688890e-01 -8.58877897e-01 -1.15829259e-01
8.77061069e-01 9.22344506e-01 -1.84791505e-01 -7.30782673e-02
-8.80794168e-01 6.04861140e-01 5.17234206e-01 3.91640574e-01
-2.49539047e-01 -6.00947261e-01 5.73205054e-01 4.22944397e-01
-2.63429791e-01 1.66601986e-01 -5.02670884e-01 -7.53373623e-01
6.88242093e-02 -1.77073014e+00 -6.20424561e-02 -1.36761534e+00
-1.27693534e+00 1.58370987e-01 1.17326051e-01 -4.40626979e-01
-8.54572058e-01 -3.70342076e-01 -7.71880507e-01 9.29819822e-01
-6.09266579e-01 -8.15657139e-01 -8.61434400e-01 3.44671518e-01
6.24206662e-01 -1.33285612e-01 8.72039795e-01 -1.99138805e-01
-9.19341221e-02 7.28057384e-01 -3.36608499e-01 6.76126257e-02
9.24130619e-01 -1.41916680e+00 1.68131553e-02 2.87182242e-01
-6.38322830e-01 5.19935787e-01 5.95240474e-01 -8.54929149e-01
-1.30230117e+00 -6.52634978e-01 1.41619706e+00 -7.68086851e-01
4.62892562e-01 -2.52525836e-01 -1.02012014e+00 3.32173735e-01
9.93653357e-01 -1.03999591e+00 1.10651684e+00 -9.28927809e-02
-6.97427765e-02 3.12315613e-01 -1.69068778e+00 7.89114416e-01
1.22277296e+00 -7.44139135e-01 -1.19347334e+00 3.37166071e-01
9.86320853e-01 -2.56853759e-01 -1.06764519e+00 -1.25115007e-01
4.54097390e-01 -1.00586987e+00 9.95375216e-02 -4.17775154e-01
7.15188980e-01 3.88328522e-01 2.02551663e-01 -1.09248936e+00
2.48109281e-01 -1.22776759e+00 2.83602715e-01 1.92326176e+00
3.69223595e-01 -8.16492200e-01 9.15566683e-01 1.84980583e+00
4.79003564e-02 -6.40055418e-01 -4.85974252e-01 2.21323952e-01
5.21990418e-01 -1.60336703e-01 1.04010081e+00 9.41682220e-01
5.00941098e-01 6.43321633e-01 5.03049076e-01 -1.37077540e-01
1.47917345e-01 -7.70931900e-01 9.58641827e-01 -1.08707476e+00
1.53469548e-01 -2.93340921e-01 1.23808660e-01 -7.32976735e-01
2.27369756e-01 -4.18113858e-01 2.10707784e-01 -1.68562400e+00
-6.48092702e-02 -2.07515165e-01 1.21439803e+00 4.58055258e-01
-5.81327319e-01 -6.23097420e-01 1.94917366e-01 -3.50930899e-01
-2.10255846e-01 4.18062359e-01 1.12613463e+00 3.47230405e-01
-4.60693955e-01 2.84205619e-02 -1.40418863e+00 7.60757506e-01
8.22869539e-01 -1.67269289e-01 -8.89972210e-01 -8.29532146e-01
1.58440173e-02 5.42361438e-01 1.42438725e-01 -1.04537249e+00
3.14573437e-01 -2.75498956e-01 3.02171171e-01 -3.86401176e-01
-2.38474146e-01 -5.54167032e-01 -2.47912124e-01 -1.10377356e-01
-7.29069829e-01 1.72153473e-01 -1.08723179e-01 -4.27120507e-01
2.42947027e-01 -1.04791713e+00 2.84512162e-01 -9.87449475e-03
4.64522332e-01 -1.27172858e-01 -1.10913253e+00 8.11627507e-01
9.05343056e-01 -3.22971880e-01 -6.37783527e-01 -1.26287091e+00
-4.09048572e-02 8.57232571e-01 9.66406345e-01 3.77285302e-01
4.09029365e-01 -9.90555882e-01 -7.00059056e-01 1.30308062e-01
1.19216062e-01 3.01363528e-01 -3.44019502e-01 1.14011452e-01
-8.31148148e-01 4.26752031e-01 -9.90716293e-02 1.85211107e-01
-1.28148842e+00 -1.91359833e-01 2.63753235e-01 3.28491211e-01
-2.03179494e-01 5.98853409e-01 -3.09743911e-01 -8.98074865e-01
5.50299525e-01 -3.53971034e-01 -5.35684288e-01 2.00121567e-01
7.93156028e-01 6.17407680e-01 -3.97929013e-01 -3.79980713e-01
2.99035877e-01 3.34359035e-02 8.59747753e-02 -6.59106076e-01
9.67985570e-01 -1.14304215e-01 1.14526376e-01 2.37334773e-01
7.54521310e-01 1.24962784e-01 -1.00696731e+00 1.35960445e-01
1.16178617e-01 -2.28775874e-01 -9.14654315e-01 -1.05145299e+00
-1.78775460e-01 5.80765128e-01 3.72679532e-01 4.95555073e-01
9.32415724e-01 2.49706477e-01 6.79391742e-01 6.05616689e-01
3.69640946e-01 -1.09101522e+00 1.25559017e-01 7.25109100e-01
1.04826605e+00 -1.24576974e+00 -5.43447375e-01 -2.57360399e-01
-6.39389694e-01 9.80620563e-01 1.33344615e+00 3.68555546e-01
2.58173525e-01 9.91920158e-02 7.38147080e-01 -1.40210250e-02
-1.02883744e+00 -2.91616470e-02 -2.09023565e-01 1.56432974e+00
9.82050598e-01 -8.12624581e-03 -6.78773046e-01 7.75797665e-01
-8.80951405e-01 2.62622952e-01 8.85092080e-01 1.20581460e+00
-4.55320895e-01 -1.31028867e+00 -7.78966784e-01 8.57470334e-01
-3.59141707e-01 -5.68228774e-02 -1.24221778e+00 6.86549485e-01
-6.69078436e-03 1.92054594e+00 -1.29767135e-01 -7.12359920e-02
3.49755347e-01 4.93091673e-01 4.87053573e-01 -9.55698013e-01
-1.35219657e+00 -7.17782080e-01 9.27012503e-01 -1.91333592e-02
-1.38128981e-01 -1.07394779e+00 -1.13209605e+00 -2.65528411e-01
-1.68693215e-01 1.44342348e-01 4.34049070e-01 1.08715868e+00
1.10714108e-01 -1.20432355e-01 3.07575762e-01 -2.77128041e-01
-4.14742619e-01 -1.35700476e+00 1.07285686e-01 2.77004302e-01
2.51378775e-01 2.91444343e-02 -4.65726219e-02 -7.85510391e-02] | [12.103094100952148, 7.998887538909912] |
7a2b14b5-1792-4086-8d38-1bf7b11e4d48 | smac-simultaneous-mapping-and-clustering | null | null | https://icml.cc/Conferences/2018/Schedule?showEvent=1899 | http://proceedings.mlr.press/v80/bajaj18a/bajaj18a.pdf | SMAC: Simultaneous Mapping and Clustering Using Spectral Decompositions |
We introduce a principled approach for simultaneous mapping and clustering (SMAC) for establishing consistent maps across heterogeneous object collections (e.g., 2D images or 3D shapes). Our approach takes as input a heterogeneous object collection and a set of maps computed between some pairs of objects, and outputs a homogeneous object clustering together with a new set of maps possessing optimal intra- and inter-cluster consistency. Our approach is based on the spectral decomposition of a data matrix storing all pairwise maps in its blocks. We additionally provide tight theoretical guarantees on the exactness of SMAC under established noise models. We also demonstrate the usefulness of the approach on synthetic and real datasets.
| ['Qi-Xing Huang', 'Zihang He', 'Chandrajit Bajaj', 'Zhenxiao Liang', 'Tingran Gao'] | 2018-07-01 | null | null | null | icml-2018-7 | ['smac-1', 'smac'] | ['playing-games', 'playing-games'] | [-2.35723052e-03 -4.28257696e-02 1.05408065e-01 -2.65361369e-01
-1.06787038e+00 -8.63734961e-01 5.32385409e-01 3.42282534e-01
-4.21595946e-03 1.77809790e-01 -5.25717810e-02 1.87599123e-01
-7.20270693e-01 -6.64476037e-01 -8.68634582e-01 -8.47122192e-01
-5.65106690e-01 1.04405153e+00 5.32750547e-01 3.19849700e-01
2.84090817e-01 5.96466601e-01 -1.49005044e+00 1.58153087e-01
6.49989426e-01 9.50061262e-01 4.15355504e-01 7.95817792e-01
-8.16333666e-03 3.05329621e-01 -2.53929287e-01 -2.66954750e-01
8.23545992e-01 -1.82212129e-01 -7.63770401e-01 4.97642279e-01
4.39726949e-01 3.80204171e-01 4.47899587e-02 1.41967165e+00
3.29337001e-01 -1.04942136e-01 7.16453910e-01 -1.59415495e+00
-6.92434192e-01 4.98028129e-01 -1.02041340e+00 -1.70678869e-01
9.49964523e-02 -2.92175561e-01 8.56097281e-01 -1.11490476e+00
8.83567035e-01 1.08185983e+00 7.66774476e-01 1.58795998e-01
-1.92212141e+00 -2.83627748e-01 -1.35687947e-01 3.57687846e-02
-1.94266605e+00 -3.92947435e-01 7.63848484e-01 -7.05233514e-01
3.92580748e-01 4.95389938e-01 4.21269149e-01 2.59119987e-01
-1.92976132e-01 2.88329244e-01 1.05443728e+00 -3.92878860e-01
7.22965956e-01 1.43602252e-01 1.22807682e-01 3.73667479e-01
6.08394861e-01 -2.77290612e-01 -6.26314700e-01 -5.81169605e-01
5.82236230e-01 -1.63407102e-01 -1.73916668e-02 -1.15014493e+00
-1.31588089e+00 3.62366289e-01 4.12764311e-01 1.78032711e-01
-3.77550989e-01 9.28186774e-02 6.25495911e-02 6.33180961e-02
2.85608858e-01 2.57144421e-01 -6.09232970e-02 1.77312121e-01
-7.39445865e-01 1.65585294e-01 6.30901694e-01 1.51513207e+00
1.11325586e+00 -6.21948540e-01 5.60383856e-01 7.54447460e-01
8.32004398e-02 8.45145762e-01 -1.27221331e-01 -1.21232903e+00
5.78717709e-01 6.26804531e-01 2.55145460e-01 -1.71058524e+00
-4.49201167e-01 -9.02378261e-02 -1.06466985e+00 1.43922657e-01
3.99291217e-02 4.28732067e-01 -7.06013858e-01 1.86227107e+00
5.00657797e-01 3.27889055e-01 -4.14511226e-02 8.54605258e-01
3.04117590e-01 3.96299034e-01 -3.81927431e-01 -3.11524510e-01
9.04175580e-01 -3.28353941e-01 -4.88701373e-01 1.60032779e-01
2.60320097e-01 -5.83764970e-01 4.88269627e-01 1.81293413e-01
-1.34233057e+00 -5.94215393e-01 -9.86538708e-01 9.02327374e-02
-4.06019896e-01 -1.98211029e-01 2.08904296e-02 4.92061287e-01
-1.45433271e+00 3.08138609e-01 -9.15659547e-01 -5.42310834e-01
2.25092605e-01 6.37515545e-01 -7.35912621e-01 2.64338776e-02
-3.55188817e-01 5.20482540e-01 6.18962467e-01 -2.28799134e-02
-4.34441984e-01 -8.93953621e-01 -6.27466738e-01 -2.21388787e-01
1.66702494e-01 -4.81561095e-01 4.20825303e-01 -5.71894765e-01
-6.22705162e-01 1.03659964e+00 -2.04337627e-01 -1.62181109e-01
3.44617426e-01 3.55776966e-01 -3.18769068e-01 3.62433016e-01
6.29793942e-01 6.69928253e-01 5.11759698e-01 -2.09055042e+00
-7.01048672e-01 -7.16865659e-01 -4.83629912e-01 2.22187042e-01
-1.18164986e-01 -1.01020873e-01 -9.64620590e-01 -4.07682240e-01
8.53793740e-01 -1.17426383e+00 -2.83386081e-01 6.45580441e-02
-6.91430032e-01 2.32881024e-01 7.16083646e-01 -5.11053443e-01
1.01265585e+00 -2.42186594e+00 4.98224318e-01 1.17139697e+00
4.77931470e-01 -4.78838205e-01 -2.74857938e-01 4.40968305e-01
-1.00696974e-01 1.38036817e-01 -5.32899618e-01 -5.25758147e-01
8.84158313e-02 1.41561940e-01 2.65276581e-02 9.57583725e-01
-2.27036122e-02 6.16980791e-01 -8.21322918e-01 -7.71894336e-01
1.05432555e-01 2.21901517e-02 -3.31224263e-01 3.02345920e-02
3.18671614e-01 1.40498385e-01 -7.47793391e-02 6.10218585e-01
1.19855082e+00 -4.28405553e-01 5.09926140e-01 -2.66876280e-01
-3.84276100e-02 -3.74270290e-01 -1.74333143e+00 1.63461590e+00
1.38887271e-01 2.84869522e-01 4.57722604e-01 -8.67489994e-01
9.25578594e-01 1.68914676e-01 1.07893467e+00 -3.89529169e-01
-2.57932603e-01 3.34467709e-01 -1.65678397e-01 -2.26680085e-01
4.35837865e-01 1.14945680e-01 -3.23688239e-01 8.75025332e-01
-8.92820507e-02 -5.59456535e-02 2.57560201e-02 5.41907191e-01
1.18678212e+00 -5.35188615e-01 2.11695656e-01 -9.94618356e-01
2.20446408e-01 1.20703824e-01 4.14266914e-01 8.27732384e-01
-1.98265553e-01 1.04587686e+00 2.66248316e-01 -2.24944279e-01
-1.47031784e+00 -1.53464341e+00 -1.98631421e-01 6.18703723e-01
6.85833514e-01 -3.82683814e-01 -7.19632626e-01 -3.45687002e-01
3.72289896e-01 -1.27937302e-01 -1.07426178e+00 2.06019841e-02
-3.39788616e-01 -7.35734701e-01 1.26999155e-01 5.24854302e-01
3.36562067e-01 -3.62690896e-01 -2.11314574e-01 -9.75079834e-02
-2.36876935e-01 -1.13543499e+00 -5.48614979e-01 6.26693591e-02
-6.80191815e-01 -1.12008238e+00 -2.27856994e-01 -1.04974675e+00
1.16307116e+00 7.47241437e-01 1.05589139e+00 2.11223587e-01
-1.17234595e-01 4.59950238e-01 -7.26498291e-02 6.12585545e-02
-2.75200278e-01 7.09434226e-03 4.67904955e-01 3.55517507e-01
1.52795345e-01 -8.59234512e-01 -1.85800344e-01 6.26981258e-01
-7.95545101e-01 4.01194170e-02 3.92833829e-01 4.83124465e-01
1.09152842e+00 3.47771078e-01 2.86951363e-01 -7.84259439e-01
3.45589012e-01 -7.10712790e-01 -9.27754819e-01 6.46290720e-01
-4.82942462e-01 -1.57554045e-01 2.03226104e-01 -1.60266757e-01
-7.93501318e-01 5.13579428e-01 8.04654539e-01 -5.68452001e-01
-2.77688857e-02 4.21118915e-01 -4.94672716e-01 -1.57858089e-01
7.62152612e-01 7.86665678e-02 1.12384163e-01 -4.66181517e-01
4.79583859e-01 6.52350843e-01 1.23831522e+00 -8.37609470e-01
1.16099012e+00 1.01068437e+00 2.56217182e-01 -5.24689376e-01
-1.92554027e-01 -9.09789324e-01 -1.31693339e+00 -2.49434531e-01
6.68779135e-01 -8.80639911e-01 -5.79068661e-01 2.84724742e-01
-9.57900226e-01 -1.00326650e-01 -1.45882085e-01 2.52367795e-01
-9.13688064e-01 1.36364371e-01 -2.47180402e-01 -6.08952820e-01
2.48103365e-01 -8.04513991e-01 1.10402119e+00 -3.16263080e-01
-1.37979180e-01 -9.59576905e-01 3.84241343e-01 -3.01759169e-02
9.88747254e-02 3.56119663e-01 8.24652553e-01 -5.24278104e-01
-8.84511411e-01 -3.17262769e-01 -4.20336008e-01 -2.57185996e-01
3.04559231e-01 1.12067230e-01 -7.08914161e-01 -3.02760065e-01
-1.84231237e-01 -2.57644244e-03 3.52616757e-01 4.71054137e-01
9.35960948e-01 -1.71138048e-01 -6.64738476e-01 6.11928463e-01
1.74266088e+00 1.38887390e-01 5.39369047e-01 1.56737804e-01
7.42953837e-01 8.33135426e-01 4.30698544e-01 3.78403485e-01
3.67006987e-01 8.87410104e-01 1.13371052e-01 -4.82763126e-02
4.74773496e-02 -6.60657063e-02 -2.48256743e-01 1.07550263e+00
-4.56211567e-02 4.95123528e-02 -1.09658992e+00 8.57035160e-01
-2.15681863e+00 -9.54396665e-01 -2.31069580e-01 2.42182183e+00
5.85097253e-01 -2.92245299e-01 3.43479842e-01 -8.47708210e-02
1.34966063e+00 -3.48286390e-01 -5.02525449e-01 1.97179914e-01
-4.32423353e-01 -2.29190126e-01 6.90199375e-01 6.69242322e-01
-1.13359869e+00 3.09088945e-01 7.04937267e+00 3.03236514e-01
-3.59827071e-01 4.41240855e-02 4.39202935e-01 -1.68250367e-01
-5.38384736e-01 -1.12952799e-01 -3.76993895e-01 5.46512842e-01
5.99966586e-01 -4.34868783e-01 3.89054000e-01 6.80094302e-01
-2.13865086e-01 -1.38799056e-01 -1.10256469e+00 9.99999762e-01
-3.83614227e-02 -1.55035901e+00 -2.03003541e-01 4.10455495e-01
1.25209808e+00 -4.41932417e-02 4.61098291e-02 -5.48178136e-01
8.87861192e-01 -5.88962317e-01 9.22511280e-01 3.29571158e-01
8.25515866e-01 -7.14398205e-01 5.18279135e-01 7.31520355e-02
-1.38221288e+00 3.14613074e-01 -5.75833499e-01 4.76075441e-01
-5.05705290e-02 6.11362636e-01 -6.37673974e-01 8.10409307e-01
9.46232080e-01 7.37924993e-01 -8.06476057e-01 9.64793921e-01
5.87225974e-01 9.05966014e-02 -7.02114880e-01 3.62300813e-01
-2.73876101e-01 -3.96609932e-01 4.56530631e-01 1.08083963e+00
3.02932292e-01 1.91018194e-01 3.43170434e-01 7.49665737e-01
7.44256005e-03 2.13838816e-02 -7.60345817e-01 4.97437656e-01
1.17467165e+00 1.20773005e+00 -1.28977311e+00 -3.89153868e-01
-1.86651036e-01 8.43383551e-01 5.01928508e-01 3.22814167e-01
-5.40391922e-01 -1.32678360e-01 6.94653034e-01 -7.44050816e-02
4.54724878e-01 -1.81932956e-01 -8.58166575e-01 -8.56490731e-01
3.46886128e-01 -4.88720626e-01 5.21050394e-01 -5.86499810e-01
-1.46727753e+00 3.05495828e-01 2.10388631e-01 -1.39205074e+00
1.70827433e-02 -2.16946468e-01 -2.85036147e-01 7.86959350e-01
-5.24037063e-01 -1.00008702e+00 -4.05597240e-01 6.71930194e-01
-1.69909507e-01 -8.93585905e-02 5.76419294e-01 2.15444490e-01
-2.67440259e-01 2.17981458e-01 4.35643345e-01 -2.98066437e-03
4.40391183e-01 -1.47307026e+00 4.41104949e-01 1.08672810e+00
5.00320867e-02 7.12050080e-01 6.27835453e-01 -7.38093972e-01
-1.44483817e+00 -1.23915553e+00 5.31867862e-01 -6.50997400e-01
7.54388452e-01 -6.60357177e-01 -9.17850375e-01 7.96728373e-01
1.30450185e-02 2.08153650e-01 6.83836997e-01 1.95912808e-01
-5.24641454e-01 -2.28701204e-01 -1.29062665e+00 3.63149703e-01
1.23604500e+00 -6.81383848e-01 -1.82951260e-02 3.17631960e-01
5.25872946e-01 -2.10653618e-01 -1.12813187e+00 2.44632602e-01
4.58050102e-01 -9.23437238e-01 1.06289601e+00 -4.34007853e-01
2.70823259e-02 -9.11638021e-01 -8.31232965e-01 -1.18355060e+00
-8.75118732e-01 -3.49483192e-01 3.71008426e-01 1.24587286e+00
3.99120688e-01 -4.64215606e-01 7.40911007e-01 7.71245062e-01
-1.33210391e-01 -1.83466315e-01 -9.71382141e-01 -1.18997669e+00
-2.08595976e-01 -4.56659377e-01 7.16020763e-01 1.05240405e+00
1.46652609e-01 -6.12472147e-02 -1.79193184e-01 8.16122532e-01
1.19502079e+00 3.31645072e-01 8.71740460e-01 -1.43005633e+00
-1.54559193e-02 -2.51986891e-01 -9.37555432e-01 -2.78504670e-01
-2.30541527e-02 -8.98149490e-01 4.54915941e-01 -1.22154689e+00
7.25144684e-01 -1.08200872e+00 -9.94141996e-02 1.73729360e-01
-4.13859263e-02 6.07054770e-01 2.08665311e-01 7.81116903e-01
-1.01396132e+00 8.17016438e-02 5.60983956e-01 -3.48011516e-02
-1.25949770e-01 -3.05749327e-01 -6.44602954e-01 3.05628449e-01
6.16143227e-01 -4.62103605e-01 -3.81026596e-01 -2.83837795e-01
2.19290450e-01 -1.69353515e-01 4.37521219e-01 -9.54053402e-01
6.17249012e-01 -2.37356633e-01 2.62059093e-01 -8.33388925e-01
1.97947666e-01 -1.14392745e+00 1.01153719e+00 2.33255953e-01
-3.73287082e-01 4.19995070e-01 2.37030555e-02 7.45696664e-01
-1.01412848e-01 1.84014574e-01 8.45727503e-01 1.31149694e-01
-5.83749413e-01 1.83723509e-01 1.88146278e-01 -1.43213868e-01
1.31598377e+00 -2.06611440e-01 -3.68155748e-01 -1.81789026e-01
-7.78243899e-01 2.95737505e-01 1.05807984e+00 1.78503960e-01
5.64504325e-01 -1.86196899e+00 -8.02162349e-01 3.85714769e-01
3.81102115e-01 1.91300422e-01 1.90072224e-01 9.11932528e-01
-4.32983428e-01 2.22061664e-01 -2.45020300e-01 -1.15799999e+00
-1.34970784e+00 1.00499868e+00 1.47788867e-01 2.65409589e-01
-4.65214729e-01 5.90053558e-01 3.72856587e-01 -5.57800472e-01
7.18879029e-02 6.40979111e-02 7.04676092e-01 -1.25613987e-01
5.17900467e-01 5.14062345e-01 2.75517285e-01 -8.95256042e-01
-6.45813465e-01 7.03929543e-01 1.95675179e-01 -5.92815518e-01
1.25442910e+00 -4.92244720e-01 -6.65141582e-01 6.21077597e-01
1.50954151e+00 2.52132677e-02 -1.13017762e+00 -3.15489590e-01
3.64362955e-01 -7.76920557e-01 -4.36818391e-01 -3.43043953e-01
-1.03551817e+00 1.59932047e-01 5.13777435e-01 5.29653311e-01
1.19672894e+00 7.90393949e-01 -1.51045462e-02 2.57801741e-01
7.21923709e-01 -1.04563963e+00 -8.74544755e-02 -2.47262955e-01
7.96111524e-01 -1.09115314e+00 8.58461261e-02 -6.46037817e-01
-3.35598499e-01 7.09487379e-01 1.98047623e-01 -1.18544564e-01
9.42496717e-01 6.11509800e-01 2.23898720e-02 -4.51803476e-01
-7.64755845e-01 -2.31374785e-01 2.38673627e-01 9.81951118e-01
-3.25270027e-01 3.49244237e-01 2.30700031e-01 6.04047999e-02
-1.14231512e-01 -4.20949936e-01 3.98786366e-01 9.12502408e-01
-4.25914615e-01 -6.47636235e-01 -7.20703363e-01 3.14870626e-01
1.37546539e-01 3.40951294e-01 -6.83992505e-01 8.13679636e-01
1.42514497e-01 7.54747331e-01 5.17023385e-01 -7.10377872e-01
2.14828879e-01 -3.56847584e-01 4.28068250e-01 -4.25225407e-01
-8.61030258e-03 2.47816026e-01 -2.95230508e-01 -5.11643827e-01
-7.71253824e-01 -1.10286045e+00 -1.14164388e+00 -5.87027371e-01
-1.60792857e-01 1.70448750e-01 5.69711149e-01 5.35718024e-01
6.75695896e-01 -2.83105463e-01 1.05199885e+00 -6.64258182e-01
3.25622619e-03 -5.80175877e-01 -9.10655916e-01 6.81936562e-01
2.70888448e-01 -5.74850559e-01 -3.48703772e-01 4.84699249e-01] | [7.690502643585205, 4.513182163238525] |
0d3b1cbc-7d98-4b07-8742-f9754f462d44 | text-video-retrieval-with-disentangled | 2305.12218 | null | https://arxiv.org/abs/2305.12218v1 | https://arxiv.org/pdf/2305.12218v1.pdf | Text-Video Retrieval with Disentangled Conceptualization and Set-to-Set Alignment | Text-video retrieval is a challenging cross-modal task, which aims to align visual entities with natural language descriptions. Current methods either fail to leverage the local details or are computationally expensive. What's worse, they fail to leverage the heterogeneous concepts in data. In this paper, we propose the Disentangled Conceptualization and Set-to-set Alignment (DiCoSA) to simulate the conceptualizing and reasoning process of human beings. For disentangled conceptualization, we divide the coarse feature into multiple latent factors related to semantic concepts. For set-to-set alignment, where a set of visual concepts correspond to a set of textual concepts, we propose an adaptive pooling method to aggregate semantic concepts to address the partial matching. In particular, since we encode concepts independently in only a few dimensions, DiCoSA is superior at efficiency and granularity, ensuring fine-grained interactions using a similar computational complexity as coarse-grained alignment. Extensive experiments on five datasets, including MSR-VTT, LSMDC, MSVD, ActivityNet, and DiDeMo, demonstrate that our method outperforms the existing state-of-the-art methods. | ['Jie Chen', 'Chang Liu', 'Li Yuan', 'Zhennan Wang', 'Jinfa Huang', 'Zesen Cheng', 'Hao Li', 'Peng Jin'] | 2023-05-20 | null | null | null | null | ['video-retrieval'] | ['computer-vision'] | [-1.27151787e-01 -4.07971203e-01 -2.49378994e-01 -3.78390491e-01
-5.51826894e-01 -6.97541177e-01 7.77467847e-01 3.29967588e-01
-2.73689628e-01 2.95637518e-01 6.19556546e-01 2.45055303e-01
-3.25688452e-01 -6.28427327e-01 -2.55651385e-01 -6.20799363e-01
2.23402053e-01 4.53026742e-01 9.69410613e-02 -6.96059763e-02
2.42713064e-01 9.75428969e-02 -1.67450643e+00 6.44549131e-01
7.26002157e-01 8.07155490e-01 1.17688425e-01 1.80046216e-01
-2.82295287e-01 6.13095701e-01 -3.55761111e-01 -6.62865877e-01
9.08084661e-02 -1.75405115e-01 -5.10320485e-01 7.49210715e-02
8.17345560e-01 -2.86247790e-01 -3.66337836e-01 1.10805559e+00
4.86491948e-01 1.18830144e-01 6.79040313e-01 -1.73353291e+00
-1.02884877e+00 3.63803983e-01 -8.26813936e-01 -4.96707903e-03
4.72918749e-01 -9.89416614e-02 1.38837945e+00 -1.23233712e+00
6.47800446e-01 1.59006023e+00 1.42532647e-01 4.08858716e-01
-1.25603104e+00 -9.70832229e-01 5.99909067e-01 3.94865692e-01
-1.69909871e+00 -2.18491524e-01 5.91930568e-01 -5.95582366e-01
8.57047915e-01 2.92786926e-01 8.01854372e-01 1.11198556e+00
-1.43588498e-01 8.66751492e-01 8.40980113e-01 -5.85425384e-02
2.49234512e-01 -4.41450700e-02 2.77240187e-01 7.91827500e-01
7.29209840e-01 -3.40917379e-01 -1.00657797e+00 -3.97932440e-01
7.06101775e-01 3.26649427e-01 -3.40696096e-01 -8.09278250e-01
-1.73073053e+00 8.42812896e-01 2.32671902e-01 1.74980000e-01
-2.81692117e-01 2.18949974e-01 3.54121774e-01 -1.85391288e-02
4.21190053e-01 5.08686960e-01 -1.93276152e-01 1.58633813e-02
-9.19774950e-01 4.51048255e-01 5.37291944e-01 1.24792731e+00
8.41458499e-01 -3.98478210e-01 -5.03650486e-01 6.70561016e-01
4.08734560e-01 5.67037940e-01 4.20219034e-01 -8.02058339e-01
7.27177382e-01 8.94560456e-01 3.29020359e-02 -1.44172740e+00
-8.27566069e-03 -2.26162612e-01 -9.01447415e-01 -1.61408693e-01
4.03721817e-02 3.31765205e-01 -7.56370127e-01 1.92110932e+00
1.53279319e-01 2.58416504e-01 6.05831109e-02 1.12840927e+00
8.40259671e-01 5.29724598e-01 3.04724276e-01 7.30636762e-03
1.90525842e+00 -1.10048056e+00 -8.74430537e-01 -3.56017619e-01
3.37923288e-01 -6.75837100e-01 1.14151907e+00 4.07447554e-02
-9.66302812e-01 -3.74683648e-01 -1.22046125e+00 -4.06277597e-01
-5.77189922e-01 1.28747672e-01 6.12493575e-01 2.37749487e-01
-7.24384844e-01 1.52060285e-01 -5.58978856e-01 -4.17144150e-01
4.71687675e-01 -6.65630624e-02 -6.86614752e-01 -2.01992691e-01
-1.28714001e+00 5.81965864e-01 4.45092738e-01 -2.10007921e-01
-5.44587433e-01 -7.40055501e-01 -8.07245076e-01 4.47611421e-01
6.06717348e-01 -1.15291286e+00 7.66212463e-01 -4.83731925e-01
-1.07930374e+00 1.07467246e+00 -2.47570232e-01 -9.37079862e-02
2.87169278e-01 -4.92843628e-01 -3.07004273e-01 4.43721384e-01
2.84261674e-01 8.28662217e-01 8.05504084e-01 -1.17621326e+00
-3.92177343e-01 -5.00202298e-01 3.41648698e-01 6.42020106e-01
-6.86103821e-01 -3.25043657e-04 -7.93890893e-01 -8.86628091e-01
3.81520301e-01 -7.27910399e-01 1.54270962e-01 4.93188918e-01
-2.68505365e-01 -1.46152988e-01 7.92099893e-01 -5.33147752e-01
1.31796050e+00 -2.19528127e+00 6.62154078e-01 -5.27040288e-03
8.40383053e-01 -4.28333282e-02 -2.41866633e-01 5.98256826e-01
-1.06170043e-01 2.46837407e-01 2.75270883e-02 -6.65984571e-01
4.15426314e-01 9.00740176e-02 -3.75597566e-01 3.43633890e-01
7.45328963e-02 1.09995556e+00 -1.08723617e+00 -8.27698648e-01
-8.18129703e-02 4.35738862e-01 -6.51399016e-01 2.75489628e-01
-1.75471038e-01 -1.03668772e-01 -7.51801074e-01 6.02536321e-01
5.92917323e-01 -5.52541673e-01 2.56244898e-01 -5.74639738e-01
1.56688198e-01 -1.27569392e-01 -1.00668120e+00 2.25137496e+00
-1.66842729e-01 6.80510044e-01 -3.83932322e-01 -6.99071229e-01
7.84858286e-01 2.93512285e-01 5.36662042e-01 -6.72730565e-01
-1.77936986e-01 -1.93333820e-01 -3.90600801e-01 -5.32211125e-01
6.42906249e-01 1.94253728e-01 -4.09938157e-01 5.34127593e-01
6.94189891e-02 5.10058142e-02 6.10831566e-02 6.84417069e-01
8.94943893e-01 1.70798764e-01 5.83961248e-01 -3.24007988e-01
3.05736274e-01 -1.01045050e-01 6.20520353e-01 5.75775743e-01
-1.22052766e-01 6.86427653e-01 6.36740744e-01 -2.83451468e-01
-9.99982059e-01 -1.34872723e+00 3.89198720e-01 1.12011158e+00
6.51866317e-01 -9.92810845e-01 -6.94226742e-01 -2.89292246e-01
4.67239730e-02 5.97491145e-01 -5.85690320e-01 -1.77400097e-01
-2.80407339e-01 -5.37103415e-01 4.62120354e-01 6.23450816e-01
7.11664557e-01 -6.26228750e-01 -6.37843430e-01 -1.53661236e-01
-5.00902057e-01 -1.28861058e+00 -8.84298325e-01 -5.95662713e-01
-5.20158947e-01 -1.13478196e+00 -6.28940463e-01 -4.59161818e-01
6.74303412e-01 7.34941542e-01 1.35538316e+00 3.33419442e-02
-4.08097237e-01 4.63823110e-01 -5.24851739e-01 -1.91501588e-01
4.52806681e-01 -2.47200042e-01 1.03633672e-01 1.78188592e-01
7.45796263e-01 -4.74932164e-01 -8.33184540e-01 4.06076550e-01
-8.98855150e-01 5.37781119e-01 5.15629411e-01 8.43340933e-01
5.76609790e-01 -2.23112315e-01 -3.35203744e-02 -5.48322141e-01
8.31066608e-01 -5.28618336e-01 -2.90563196e-01 7.68587053e-01
-4.48240012e-01 2.34637827e-01 7.00610727e-02 -6.63139343e-01
-1.06150091e+00 -2.01144814e-01 7.63948679e-01 -8.93387258e-01
-9.16853249e-02 5.18315852e-01 -4.14324641e-01 4.20506388e-01
3.02354544e-01 1.00712940e-01 -1.70792714e-01 -3.90587628e-01
8.21879447e-01 4.15587097e-01 4.46498930e-01 -9.61126566e-01
9.02383983e-01 6.78719997e-01 -2.52291024e-01 -4.12196994e-01
-9.13054764e-01 -6.58956468e-01 -6.84996963e-01 -8.16172883e-02
1.27660382e+00 -1.41707611e+00 -7.34191298e-01 2.51316249e-01
-1.35922158e+00 2.87472755e-01 3.05877943e-02 4.18428659e-01
-4.78787184e-01 3.84371877e-01 -2.98260331e-01 -3.79076868e-01
-3.64088267e-01 -9.24438715e-01 1.41963196e+00 1.99736714e-01
-3.26473564e-01 -7.53721595e-01 -5.39583713e-02 6.04113221e-01
3.05991530e-01 2.11345926e-01 1.07098472e+00 -6.50983453e-01
-7.99190998e-01 4.91478853e-02 -6.46577001e-01 -2.57787853e-01
1.00952402e-01 -1.12958308e-02 -6.83023274e-01 -4.25131291e-01
-2.48956144e-01 -4.16769058e-01 7.77470291e-01 -1.30283505e-01
1.18821239e+00 -3.41960520e-01 -4.91336793e-01 4.51370358e-01
1.29934609e+00 -1.02489047e-01 5.12306213e-01 1.74470082e-01
8.89767349e-01 6.75812960e-01 8.11685085e-01 5.00218213e-01
6.54250026e-01 8.13876987e-01 2.70312518e-01 -8.35805610e-02
-1.55778557e-01 -4.09080893e-01 6.04140796e-02 9.54446971e-01
2.86157383e-03 -3.88788790e-01 -9.03952003e-01 5.30492544e-01
-2.28086162e+00 -1.24876106e+00 1.57608166e-01 1.99371803e+00
5.47970295e-01 -2.05386132e-01 -2.28280693e-01 -4.35282707e-01
8.29116166e-01 3.85450155e-01 -4.17473435e-01 2.01280296e-01
-2.14427054e-01 -1.48791537e-01 2.73455586e-02 1.95711568e-01
-8.60135674e-01 8.67733240e-01 5.36578465e+00 9.34874058e-01
-7.21982777e-01 2.70175070e-01 1.43710330e-01 -4.91595656e-01
-5.82425833e-01 7.02110678e-02 -5.66087127e-01 3.34795266e-01
2.71036893e-01 -4.28359807e-01 2.98882216e-01 5.76891541e-01
-1.28611848e-01 1.13573059e-01 -1.46026719e+00 1.40552318e+00
3.91751856e-01 -1.27682126e+00 5.56797624e-01 -3.60098146e-02
5.86147249e-01 -3.68365794e-01 2.10138217e-01 1.33838668e-01
2.69459546e-01 -1.00645232e+00 9.41535056e-01 7.16558576e-01
6.42816186e-01 -6.22787356e-01 5.16970932e-01 1.23133801e-01
-1.63676143e+00 1.72145113e-01 -3.56409132e-01 1.62008002e-01
-7.87077297e-04 2.51530379e-01 -9.05865803e-02 7.69762814e-01
7.51936436e-01 7.99524188e-01 -6.60784781e-01 6.70027435e-01
-1.28321961e-01 -1.25363722e-01 1.64898783e-02 -1.89829215e-01
1.76226914e-01 -1.69684500e-01 4.57398325e-01 1.08564770e+00
3.49934816e-01 4.37124908e-01 2.29524121e-01 1.13458657e+00
-3.60512882e-02 9.09433514e-02 -3.21182370e-01 -4.11719739e-01
8.44376743e-01 1.16987252e+00 -5.57545483e-01 -5.26170909e-01
-6.09302759e-01 1.18380868e+00 6.03395820e-01 5.82748830e-01
-9.07956898e-01 -3.02692086e-01 1.13090062e+00 -1.14015318e-01
1.58144459e-01 -3.26693356e-01 1.48690827e-02 -1.73348391e+00
1.32672921e-01 -8.19483399e-01 5.99213123e-01 -1.14813280e+00
-1.60488975e+00 5.71384549e-01 5.56414545e-01 -1.38407898e+00
6.49876297e-02 -4.82658833e-01 -3.41746897e-01 7.71995902e-01
-1.16997159e+00 -1.39024353e+00 -8.43605101e-01 7.34420240e-01
6.17986321e-01 -2.82237262e-01 9.48944867e-01 2.88136244e-01
-4.12767857e-01 4.06632781e-01 -9.41573754e-02 1.21896483e-01
9.58878994e-01 -9.77778971e-01 4.43847001e-01 6.84562087e-01
3.15764397e-01 1.04296029e+00 7.27360964e-01 -5.23634434e-01
-1.31599450e+00 -8.24815989e-01 8.28824520e-01 -4.57644284e-01
7.32273757e-01 -7.50524163e-01 -8.92647326e-01 6.22716069e-01
4.86053884e-01 2.15666041e-05 1.01855481e+00 1.43012404e-01
-1.04020894e+00 4.93410192e-02 -8.71438920e-01 9.27572966e-01
1.33292162e+00 -9.14990544e-01 -1.10742188e+00 2.68398076e-01
1.05828297e+00 -2.85984069e-01 -7.74193466e-01 1.42337635e-01
7.94717431e-01 -8.74344647e-01 1.24857235e+00 -8.95217776e-01
5.41448593e-01 -4.52417731e-01 -5.63559175e-01 -1.20810449e+00
-5.37799954e-01 -2.96350092e-01 -2.68240005e-01 1.33230424e+00
6.16958402e-02 -4.85416502e-01 7.07766712e-01 7.43296802e-01
4.02735710e-01 -6.38548732e-01 -6.71214819e-01 -5.85222065e-01
-2.59066045e-01 -9.11984295e-02 9.25934374e-01 1.30543971e+00
1.11738987e-01 3.05238515e-01 -4.19141412e-01 2.77570665e-01
9.49213803e-01 5.41848719e-01 6.03384733e-01 -1.24573684e+00
-2.42443547e-01 -3.41987163e-01 -5.69322109e-01 -1.00821865e+00
2.41877764e-01 -5.93529642e-01 -3.58312875e-01 -1.52106583e+00
8.59673262e-01 -1.14325285e-01 -5.09824932e-01 6.18929386e-01
-3.65540266e-01 9.98704582e-02 4.57550377e-01 3.99815649e-01
-9.24018145e-01 7.14216590e-01 1.15147126e+00 -4.09072995e-01
1.80810317e-01 -7.46975482e-01 -7.97015250e-01 5.44883370e-01
4.93902445e-01 -5.10817409e-01 -7.37473190e-01 -7.44503379e-01
3.93049300e-01 5.84499836e-02 6.66945755e-01 -8.10731053e-01
5.39579809e-01 -3.59180689e-01 2.20144868e-01 -7.67161191e-01
7.31528461e-01 -7.77922153e-01 3.10992539e-01 1.83714271e-01
-4.42548215e-01 4.62823272e-01 8.08256492e-02 7.96983242e-01
-4.67204720e-01 1.33340716e-01 3.01401168e-01 -1.67548344e-01
-9.73898768e-01 5.75588644e-01 -8.43242705e-02 3.57472934e-02
1.04504180e+00 -1.29933402e-01 -7.04288244e-01 -2.97030717e-01
-5.32410681e-01 5.40904164e-01 4.88147020e-01 8.17864895e-01
8.76945496e-01 -1.78048968e+00 -6.09092295e-01 -1.31303817e-01
7.09270120e-01 -1.54679090e-01 4.59156752e-01 4.88007367e-01
-2.75806159e-01 5.36651254e-01 -4.77677464e-01 -5.51144958e-01
-1.55141473e+00 7.26344168e-01 6.89082965e-02 -1.30661666e-01
-4.01713639e-01 8.40661168e-01 7.17457235e-01 -1.28257468e-01
1.17412508e-01 -8.65962915e-03 -2.27796778e-01 3.61340761e-01
4.77816373e-01 1.86311528e-01 -4.41318423e-01 -7.36580789e-01
-4.95797783e-01 9.01793778e-01 -2.04139516e-01 -1.01000637e-01
1.13629639e+00 -3.09386224e-01 -3.32196742e-01 3.59820276e-01
1.27652586e+00 -1.29517987e-01 -1.08716345e+00 -5.56403756e-01
-1.18137248e-01 -8.85251343e-01 -6.31304458e-02 -4.50936943e-01
-1.01132321e+00 1.00581539e+00 4.83445525e-01 -1.94758192e-01
1.11001611e+00 1.20150864e-01 3.72451574e-01 4.85827625e-01
3.55033815e-01 -8.99223089e-01 6.40359640e-01 6.10022694e-02
1.01970255e+00 -1.01063931e+00 3.49934399e-01 -6.60617173e-01
-8.94816041e-01 1.00200963e+00 9.95120347e-01 1.31510049e-01
3.92081171e-01 -1.12935476e-01 -2.44887635e-01 -5.58716595e-01
-1.09543240e+00 -1.59472629e-01 6.39477193e-01 2.69311309e-01
2.44823456e-01 1.60917222e-01 -2.09811538e-01 5.82133353e-01
1.30545050e-01 -2.62432486e-01 1.35254815e-01 7.49994099e-01
-2.82337308e-01 -9.39478099e-01 -1.62176013e-01 1.56153247e-01
2.22222935e-02 -2.97435552e-01 -4.12838966e-01 8.74323070e-01
9.58447456e-02 7.30842590e-01 3.77841562e-01 -3.67970854e-01
3.33210766e-01 -5.00033312e-02 3.70618433e-01 -5.63706219e-01
-1.07854150e-01 -1.31578192e-01 -3.26784737e-02 -9.32819366e-01
-6.21176481e-01 -5.24147689e-01 -1.14573109e+00 -3.84034455e-01
-1.27521351e-01 5.75081520e-02 2.83481985e-01 9.74061489e-01
5.99892139e-01 3.76400858e-01 1.15680397e-02 -6.14614546e-01
-3.73042792e-01 -5.80795050e-01 -5.06864011e-01 9.10115361e-01
7.91197643e-02 -9.62874711e-01 -1.37591571e-01 6.60537183e-02] | [10.451278686523438, 1.1522945165634155] |
97f726cb-bbfc-4116-bbb2-23ee996befc0 | progressive-graph-convolution-network-for-eeg | 2112.09069 | null | https://arxiv.org/abs/2112.09069v1 | https://arxiv.org/pdf/2112.09069v1.pdf | Progressive Graph Convolution Network for EEG Emotion Recognition | Studies in the area of neuroscience have revealed the relationship between emotional patterns and brain functional regions, demonstrating that dynamic relationships between different brain regions are an essential factor affecting emotion recognition determined through electroencephalography (EEG). Moreover, in EEG emotion recognition, we can observe that clearer boundaries exist between coarse-grained emotions than those between fine-grained emotions, based on the same EEG data; this indicates the concurrence of large coarse- and small fine-grained emotion variations. Thus, the progressive classification process from coarse- to fine-grained categories may be helpful for EEG emotion recognition. Consequently, in this study, we propose a progressive graph convolution network (PGCN) for capturing this inherent characteristic in EEG emotional signals and progressively learning the discriminative EEG features. To fit different EEG patterns, we constructed a dual-graph module to characterize the intrinsic relationship between different EEG channels, containing the dynamic functional connections and static spatial proximity information of brain regions from neuroscience research. Moreover, motivated by the observation of the relationship between coarse- and fine-grained emotions, we adopt a dual-head module that enables the PGCN to progressively learn more discriminative EEG features, from coarse-grained (easy) to fine-grained categories (difficult), referring to the hierarchical characteristic of emotion. To verify the performance of our model, extensive experiments were conducted on two public datasets: SEED-IV and multi-modal physiological emotion database (MPED). | ['Rui Cheng', 'Yuanfang Chen', 'Lijian Zhang', 'Wenming Zheng', 'Guangming Shi', 'Youshuo Ji', 'Yang Li', 'Fu Li', 'Yijin Zhou'] | 2021-12-14 | null | null | null | null | ['eeg-emotion-recognition'] | ['miscellaneous'] | [-2.53824413e-01 -4.81743157e-01 4.48000342e-01 -6.34758830e-01
1.56596452e-01 -4.94778514e-01 2.26723790e-01 1.30718827e-01
-2.74955422e-01 7.85684168e-01 1.95869938e-01 2.19281539e-01
-6.74148023e-01 -6.82938993e-01 -4.63203609e-01 -7.14549363e-01
-5.73302209e-01 -1.00142516e-01 -3.16924006e-01 -1.65689155e-01
3.22319329e-01 5.63687503e-01 -1.40400875e+00 3.91274929e-01
1.10702586e+00 1.35106266e+00 9.14642885e-02 1.34538293e-01
-3.23929191e-02 3.02161783e-01 -6.69681072e-01 -2.28215843e-01
-1.79456651e-01 -5.60402215e-01 -5.32640159e-01 4.42202687e-02
-2.78793573e-01 4.48165536e-01 -2.71859705e-01 1.30948353e+00
5.35274684e-01 2.35574782e-01 8.01764011e-01 -1.41116273e+00
-8.52259815e-01 4.20094639e-01 -5.87054372e-01 5.59358120e-01
4.13221210e-01 -4.01228815e-02 9.68517721e-01 -8.40213954e-01
2.54111648e-01 9.25830007e-01 4.95499492e-01 1.19054735e-01
-1.00203156e+00 -1.04885745e+00 4.42304015e-01 5.51357388e-01
-1.72501504e+00 -2.72745602e-02 1.09092081e+00 -6.13320053e-01
7.44756281e-01 2.32859030e-01 1.28929555e+00 1.36678374e+00
7.42599249e-01 2.60356903e-01 1.72466493e+00 1.06978744e-01
1.76989257e-01 1.14203461e-01 4.22429353e-01 4.37095106e-01
4.76325452e-02 1.32438876e-02 -7.20597148e-01 -1.37246147e-01
5.99910438e-01 1.05015680e-01 -7.25269496e-01 -8.10215846e-02
-1.08735812e+00 4.82225239e-01 6.15595758e-01 8.49789858e-01
-6.84907973e-01 -1.62898734e-01 4.93895113e-01 4.92449999e-01
3.79334033e-01 5.12883425e-01 -4.09169734e-01 -3.13082874e-01
-7.72541285e-01 -4.54028994e-01 5.54525912e-01 6.70481145e-01
9.71898854e-01 1.63613874e-02 -1.06827527e-01 8.17458212e-01
1.67468220e-01 7.33460439e-03 8.85106683e-01 -4.07934859e-02
1.18433006e-01 8.83691132e-01 -4.32254285e-01 -1.60424638e+00
-8.59152913e-01 -7.30503559e-01 -1.53303182e+00 -1.93902716e-01
-4.89492863e-02 -1.95267245e-01 -3.59912336e-01 1.90312445e+00
2.65523256e-03 3.75949293e-01 -1.46569848e-01 1.08222091e+00
7.94848025e-01 3.58621895e-01 2.54319996e-01 -3.67550731e-01
1.58871925e+00 -2.48476371e-01 -6.83374822e-01 3.58123109e-02
2.40324199e-01 -5.78308776e-02 1.16452348e+00 4.28954929e-01
-4.91498590e-01 -8.23920310e-01 -9.25313830e-01 5.51125824e-01
-8.91981006e-01 7.23274499e-02 8.90695810e-01 5.34885466e-01
-1.09307921e+00 5.66654921e-01 -3.80351365e-01 -2.78175473e-01
5.07950902e-01 4.27730978e-01 -6.27959073e-01 2.65081823e-01
-1.63429630e+00 6.64575815e-01 4.77705598e-01 5.82223773e-01
-5.18537641e-01 -6.70118570e-01 -5.46384513e-01 4.72322285e-01
-1.40081510e-01 -4.29535806e-01 1.10485420e-01 -1.22069597e+00
-1.24374127e+00 5.24371743e-01 1.16740055e-01 1.73836857e-01
2.16488298e-02 2.83827662e-01 -8.94163370e-01 2.19685525e-01
-2.18419760e-01 3.65633100e-01 7.71563351e-01 -9.94366109e-01
-2.61512786e-01 -7.99118698e-01 -1.74958751e-01 2.17334226e-01
-6.43719316e-01 -4.19750474e-02 -4.38336320e-02 -6.26997292e-01
3.12631354e-02 -5.70416331e-01 2.22352326e-01 -5.50454319e-01
-3.61669689e-01 -4.78409916e-01 5.85242927e-01 -2.99111545e-01
1.39739275e+00 -2.41777754e+00 4.50064957e-01 6.05572879e-01
5.98831892e-01 -3.46915662e-01 -2.20958859e-01 1.17240116e-01
-5.44012785e-01 3.68960313e-02 -2.40523472e-01 3.81221652e-01
2.79910684e-01 -9.80195180e-02 -3.94875929e-02 5.32698572e-01
3.02164972e-01 1.14036083e+00 -7.75044024e-01 -2.81096548e-01
4.10741232e-02 3.82129163e-01 -1.99290827e-01 2.24544927e-01
4.40308034e-01 7.62552321e-01 -6.53081715e-01 5.89086890e-01
6.37142658e-01 -2.55433440e-01 4.49196026e-02 -6.76632464e-01
-5.35449805e-03 -3.28631312e-01 -9.20762539e-01 1.60416234e+00
-3.37174982e-01 5.74409604e-01 -1.87094584e-02 -1.29652405e+00
1.16585851e+00 3.92662585e-01 6.55971169e-01 -7.76987553e-01
4.76327062e-01 -3.80451791e-02 2.92122811e-01 -5.66620409e-01
1.50277898e-01 -2.45906293e-01 -2.09540486e-01 3.66647035e-01
3.56410742e-01 1.38621032e-01 -9.58184153e-02 -1.30782202e-01
1.07210720e+00 -3.20195049e-01 3.92906070e-01 -7.39887714e-01
5.10235369e-01 -6.11912966e-01 4.66286004e-01 4.53561515e-01
-3.53376508e-01 2.11464390e-01 7.04495132e-01 -3.67150843e-01
-2.10518003e-01 -7.65141308e-01 -5.98757207e-01 1.01633942e+00
3.78960699e-01 -3.79764259e-01 -6.62233353e-01 -5.22531331e-01
-1.00782953e-01 1.47621378e-01 -1.04366744e+00 -6.21584356e-01
2.45319083e-02 -1.27417386e+00 6.51511967e-01 4.74987656e-01
6.86346114e-01 -1.39588153e+00 -6.01895332e-01 1.13955989e-01
-2.88031269e-02 -9.11187887e-01 -2.06934214e-01 4.96560156e-01
-4.89337713e-01 -1.04462886e+00 -4.96972054e-01 -5.96027076e-01
5.98143518e-01 -1.96256474e-01 1.18718088e+00 6.29450902e-02
-4.04002845e-01 4.12466675e-01 -5.09827852e-01 -1.32589832e-01
4.39220637e-01 -1.13045223e-01 1.74057797e-01 5.38551390e-01
5.67804515e-01 -1.04159784e+00 -6.89968765e-01 3.18193257e-01
-9.89193082e-01 -1.76253051e-01 7.74848759e-01 9.38699961e-01
5.52398443e-01 5.36336899e-01 8.36945117e-01 -3.87054533e-01
1.23127627e+00 -7.23839223e-01 -1.60849974e-01 3.50134134e-01
-3.67213994e-01 -1.44494563e-01 9.23959613e-01 -6.82056844e-01
-8.32466245e-01 -2.99436808e-01 -2.36725271e-01 -4.30053800e-01
-6.81532383e-01 7.46058941e-01 -4.09839928e-01 -3.80240649e-01
2.33683556e-01 6.22581899e-01 -5.87897658e-01 -2.36912929e-02
2.09398657e-01 6.71840489e-01 2.71584064e-01 -7.96515703e-01
2.71201491e-01 6.47753403e-02 -1.66449681e-01 -7.26136327e-01
-3.72316241e-01 -2.32161343e-01 -7.64024854e-01 -3.21405023e-01
1.11810601e+00 -6.31974399e-01 -1.15788090e+00 4.89869475e-01
-9.34560061e-01 -7.00512677e-02 -5.25484905e-02 6.91216886e-01
-3.54928464e-01 2.59725600e-01 -7.56094396e-01 -4.35924053e-01
-2.96513259e-01 -1.17366743e+00 9.19331849e-01 3.07057381e-01
-4.49411511e-01 -1.07248437e+00 4.86736782e-02 -4.34128821e-01
2.98601091e-01 2.65544683e-01 1.16866386e+00 -4.76410657e-01
7.24602910e-03 5.97808175e-02 -4.44187015e-01 2.76136920e-02
2.20535591e-01 -2.32186019e-01 -8.57308328e-01 -1.47974342e-01
1.90144941e-01 -2.92340428e-01 5.56086779e-01 2.73129195e-01
1.75359797e+00 2.59762891e-02 -1.86451346e-01 8.54379117e-01
1.31314492e+00 4.17533010e-01 7.44794667e-01 5.48556931e-02
6.78785563e-01 5.77336609e-01 1.94294363e-01 5.57920933e-01
2.89774001e-01 4.80645359e-01 2.48071000e-01 -2.00532749e-01
4.33672518e-01 1.30288810e-01 8.20066929e-02 1.10954428e+00
-3.48255962e-01 5.68385795e-03 -5.16604066e-01 1.73172340e-01
-1.57414997e+00 -8.50136697e-01 5.53442165e-02 1.61764574e+00
6.50913000e-01 -2.05894008e-01 -1.97952226e-01 -5.57165779e-02
6.92571878e-01 1.78338200e-01 -6.05110824e-01 -4.51969534e-01
-2.77860194e-01 3.25392038e-01 -2.42121041e-01 -1.56854376e-01
-8.04926872e-01 7.41816401e-01 5.66066217e+00 9.65098143e-01
-1.46632385e+00 1.03183568e-01 7.65080094e-01 8.99816304e-02
-3.28842431e-01 -5.78858495e-01 -2.44506016e-01 8.05592477e-01
7.75345623e-01 -2.84354389e-01 6.93648696e-01 4.55358982e-01
5.95511794e-02 1.89397424e-01 -1.08988416e+00 1.46385694e+00
5.11859283e-02 -9.25506294e-01 9.92166996e-03 -1.74358618e-02
3.25530708e-01 -1.72873303e-01 -9.73327085e-02 4.61059898e-01
-9.71586481e-02 -1.26249349e+00 5.54808438e-01 9.08275306e-01
9.21763420e-01 -7.58008599e-01 7.52151012e-01 1.34796605e-01
-1.62759805e+00 -1.17739335e-01 -5.51094830e-01 -9.11439583e-02
-1.97469413e-01 7.11099207e-01 1.16918229e-01 7.63243377e-01
1.06754482e+00 1.08129275e+00 -6.51972115e-01 8.80658209e-01
-1.47006273e-01 3.53192896e-01 1.73536003e-01 -2.06947818e-01
4.28534113e-02 -6.54575884e-01 1.39145046e-01 1.42282820e+00
5.33008754e-01 6.25220180e-01 -2.65915871e-01 1.19258726e+00
-1.83901384e-01 3.15119594e-01 -5.27616739e-01 -1.62408605e-01
2.60006875e-01 1.72524393e+00 -9.43771541e-01 -1.86225355e-01
-3.37936997e-01 1.14185286e+00 6.46077991e-01 5.98354876e-01
-9.76514041e-01 -5.82001925e-01 8.06522012e-01 -5.51954567e-01
-1.06243394e-01 -3.74406576e-02 -2.11395577e-01 -1.38683450e+00
-5.41305281e-02 -8.36266398e-01 2.22473770e-01 -8.99471939e-01
-1.72117865e+00 1.26747608e+00 -3.07898194e-01 -9.56528366e-01
4.39325869e-01 -6.90224826e-01 -8.12667191e-01 9.10260260e-01
-1.23634613e+00 -5.58042467e-01 -7.16682315e-01 1.18470454e+00
1.20121643e-01 1.44306242e-01 1.07294595e+00 3.45929503e-01
-6.71519399e-01 5.64625084e-01 -1.75311327e-01 1.41676828e-01
4.64464903e-01 -9.43459749e-01 -5.17812312e-01 3.82535905e-01
2.31080145e-01 6.73298001e-01 9.98920575e-02 -2.80772507e-01
-1.37519407e+00 -9.32445824e-01 3.57201815e-01 -1.68831363e-01
9.48030114e-01 -6.20441437e-01 -1.10166883e+00 2.27893427e-01
3.28834206e-01 -2.80727241e-02 9.55538630e-01 3.77758861e-01
-2.26516664e-01 -3.30000728e-01 -9.69422877e-01 3.82818788e-01
1.04461944e+00 -8.66255760e-01 -6.72709048e-01 8.69932920e-02
3.88954818e-01 1.41577259e-01 -1.36051464e+00 4.84292358e-01
6.36603713e-01 -1.06407344e+00 6.84123218e-01 -4.49843466e-01
3.28508228e-01 -9.42527652e-02 -5.28906733e-02 -2.03329682e+00
-7.68388987e-01 -3.95368002e-02 2.06788361e-01 9.87257063e-01
1.43874750e-01 -9.59217906e-01 2.19107121e-01 3.76898855e-01
-2.40232751e-01 -1.07530177e+00 -8.91668975e-01 -4.77683365e-01
-5.87440748e-03 -4.37930495e-01 9.67304468e-01 1.26123524e+00
6.13127410e-01 3.05846751e-01 -1.15041070e-01 1.84487715e-01
6.44028932e-02 4.47638959e-01 1.12306088e-01 -1.37535155e+00
-1.82410643e-01 -8.76948833e-01 -8.10775518e-01 -5.44885576e-01
5.66606939e-01 -9.82066989e-01 -1.32140100e-01 -1.25415170e+00
3.74731302e-01 -3.81024867e-01 -1.08881223e+00 3.89812559e-01
-2.15199575e-01 3.93823564e-01 -1.66958734e-01 -1.37560125e-02
-7.03965843e-01 8.57068121e-01 1.26570380e+00 -1.05017819e-01
-1.04668409e-01 -4.84079987e-01 -8.79622519e-01 5.95405161e-01
6.80696070e-01 -2.10489020e-01 -3.99238229e-01 -5.52123412e-02
3.06224167e-01 1.00730553e-01 3.95518452e-01 -1.18093336e+00
1.84404626e-01 2.55570076e-02 8.44504595e-01 -1.92644775e-01
1.15658633e-01 -1.01985741e+00 2.67888159e-01 1.50511324e-01
-2.25863293e-01 2.03481123e-01 3.75890374e-01 4.84537959e-01
-5.94103396e-01 2.67928243e-01 4.00141984e-01 3.27114239e-02
-9.37273800e-01 6.52853131e-01 -5.75751007e-01 -1.77837908e-02
1.03630638e+00 -4.02451336e-01 -1.72732979e-01 -1.55917734e-01
-1.12770402e+00 -3.36167887e-02 -2.99626701e-02 4.68271375e-01
7.91857421e-01 -1.40537536e+00 -4.06868994e-01 5.83758175e-01
2.36550167e-01 -6.57425880e-01 5.95379055e-01 1.25208569e+00
1.95226714e-01 2.54036278e-01 -8.31607163e-01 -4.45853323e-01
-9.08505917e-01 5.58723152e-01 4.35230792e-01 -8.68280306e-02
-5.35768390e-01 8.14616442e-01 7.55283117e-01 -3.41781199e-01
-8.69897660e-04 -3.74948055e-01 -6.48157060e-01 4.65633541e-01
3.08306932e-01 1.38886601e-01 8.34286734e-02 -6.57188237e-01
-5.26900709e-01 7.34717786e-01 3.87781113e-01 4.46163118e-01
1.44841647e+00 -2.95793355e-01 -6.63584173e-01 5.84577918e-01
1.27021623e+00 -1.24977529e-01 -9.59692478e-01 1.62085176e-01
-1.78611025e-01 -2.30204210e-01 -1.97366789e-01 -7.81533182e-01
-1.29995906e+00 1.04742062e+00 6.06950819e-01 4.31627035e-01
1.69844317e+00 1.11920476e-01 2.37056568e-01 1.94086403e-01
7.77728021e-01 -9.17326212e-01 9.85353813e-02 3.45090985e-01
9.77044463e-01 -7.94580519e-01 -4.21953022e-01 -1.92799732e-01
-6.64851904e-01 1.21581292e+00 6.72469020e-01 -1.73771545e-01
1.00183463e+00 2.05776885e-01 -3.07020515e-01 -8.51407707e-01
-4.73752260e-01 1.55097991e-01 6.87893152e-01 4.29847538e-01
4.33350652e-01 3.94586295e-01 -3.67465675e-01 1.32719815e+00
-3.38633806e-01 -6.91150548e-03 -9.21382569e-03 1.43445238e-01
-2.07791910e-01 -6.74810767e-01 -1.16711631e-01 7.56798267e-01
-1.67030931e-01 -3.13073963e-01 -4.37895387e-01 7.20809519e-01
3.82535428e-01 8.39350462e-01 2.05733418e-01 -7.66056359e-01
3.53793830e-01 1.85515106e-01 5.07653475e-01 -2.02327505e-01
-7.13043094e-01 -1.49463177e-01 -3.21815312e-01 -7.63524890e-01
-3.70039642e-01 -2.16560960e-01 -1.22986722e+00 -4.30381373e-02
-1.55702055e-01 5.36792636e-01 2.43835926e-01 1.10350013e+00
6.44818783e-01 9.78335679e-01 7.00743556e-01 -8.25588822e-01
8.51853862e-02 -1.09272122e+00 -1.30606425e+00 6.49800122e-01
-1.15542121e-01 -9.06576335e-01 -4.36008364e-01 -3.01301032e-01] | [13.114912033081055, 3.4905002117156982] |
aa99597d-6ef3-401d-b718-2269524cd5bc | improved-skin-lesion-recognition-by-a-self | 2112.12086 | null | https://arxiv.org/abs/2112.12086v1 | https://arxiv.org/pdf/2112.12086v1.pdf | Improved skin lesion recognition by a Self-Supervised Curricular Deep Learning approach | State-of-the-art deep learning approaches for skin lesion recognition often require pretraining on larger and more varied datasets, to overcome the generalization limitations derived from the reduced size of the skin lesion imaging datasets. ImageNet is often used as the pretraining dataset, but its transferring potential is hindered by the domain gap between the source dataset and the target dermatoscopic scenario. In this work, we introduce a novel pretraining approach that sequentially trains a series of Self-Supervised Learning pretext tasks and only requires the unlabeled skin lesion imaging data. We present a simple methodology to establish an ordering that defines a pretext task curriculum. For the multi-class skin lesion classification problem, and ISIC-2019 dataset, we provide experimental evidence showing that: i) a model pretrained by a curriculum of pretext tasks outperforms models pretrained by individual pretext tasks, and ii) a model pretrained by the optimal pretext task curriculum outperforms a model pretrained on ImageNet. We demonstrate that this performance gain is related to the fact that the curriculum of pretext tasks better focuses the attention of the final model on the skin lesion. Beyond performance improvement, this strategy allows for a large reduction in the training time with respect to ImageNet pretraining, which is especially advantageous for network architectures tailored for a specific problem. | ['Juan Carlos SanMiguel', 'Pablo Carballeira', 'Marcos Escudero Viñolo', 'Kirill Sirotkin'] | 2021-12-22 | null | null | null | null | ['skin-lesion-classification'] | ['medical'] | [ 9.64913726e-01 1.98731706e-01 -3.41738760e-01 -3.71128529e-01
-6.15048170e-01 -5.34142494e-01 4.67327356e-01 1.08624801e-01
-7.42430151e-01 5.83645344e-01 -2.25366846e-01 -2.72733182e-01
-4.29455996e-01 -6.28232241e-01 -7.13964105e-01 -8.32046866e-01
1.51223868e-01 3.89520407e-01 3.23925227e-01 3.98001187e-02
-1.20851509e-01 4.86623347e-01 -1.30148792e+00 7.01069295e-01
1.13472760e+00 9.93947864e-01 4.79953766e-01 5.24970293e-01
-1.83512103e-02 5.62015355e-01 -4.62074369e-01 -3.29934448e-01
3.45983863e-01 -6.04741991e-01 -1.07558846e+00 2.97139317e-01
8.27885330e-01 -1.45286396e-01 -2.91997761e-01 1.10282981e+00
4.18133020e-01 -1.74447164e-01 8.39153647e-01 -8.82860780e-01
-1.90071583e-01 3.76815349e-01 -4.27179068e-01 1.41083568e-01
-9.56380144e-02 1.45035177e-01 7.23290980e-01 -2.95619071e-01
8.83270621e-01 6.82468653e-01 7.91738868e-01 8.60384643e-01
-1.26807213e+00 -2.75147706e-01 1.60893217e-01 2.78351247e-01
-1.36739826e+00 1.28741950e-01 6.72427416e-01 -4.21459347e-01
6.08334541e-01 4.00327921e-01 6.05966628e-01 1.41276824e+00
1.05692133e-01 7.95276940e-01 1.46448267e+00 -5.37466109e-01
1.65145263e-01 3.90732408e-01 8.25059563e-02 5.54224908e-01
1.81048781e-01 2.07811400e-01 -4.16634902e-02 1.13506116e-01
1.03952038e+00 -1.71854734e-01 -2.69167244e-01 -5.26399791e-01
-6.12806201e-01 7.16625988e-01 6.90886855e-01 6.32936060e-01
-5.19709051e-01 -1.88590698e-02 4.75898832e-01 3.86930555e-01
2.39262074e-01 5.63451350e-01 -3.50146383e-01 3.34776103e-01
-9.93738234e-01 -1.46283463e-01 7.35352695e-01 3.72272819e-01
6.33008063e-01 -1.35188326e-01 7.99497515e-02 9.44516540e-01
-2.21232489e-01 1.25436291e-01 5.95124900e-01 -4.00950432e-01
1.74194336e-01 6.47162974e-01 -5.13316810e-01 -2.73998469e-01
-3.77864897e-01 -6.34206414e-01 -7.80174255e-01 2.95486748e-01
7.29770184e-01 -3.60055596e-01 -1.47173977e+00 1.58740056e+00
1.50061235e-01 2.05966353e-01 1.02915458e-01 8.47054660e-01
6.56781375e-01 1.43408507e-01 4.50884253e-01 5.09013496e-02
1.12674320e+00 -8.65261257e-01 -2.60945499e-01 -3.08614194e-01
8.07027996e-01 -5.39944947e-01 8.96537662e-01 5.51097035e-01
-8.25745523e-01 -6.16322041e-01 -9.37802672e-01 2.89599448e-01
-4.77785736e-01 2.46720165e-01 6.04267359e-01 6.38424873e-01
-1.23780072e+00 5.77394664e-01 -5.56954801e-01 -8.25175047e-01
6.11513317e-01 6.17123961e-01 -5.06941199e-01 -2.85509557e-01
-1.00461662e+00 1.27166498e+00 5.56288242e-01 -9.21990126e-02
-8.57980490e-01 -8.71211827e-01 -5.68791509e-01 2.85931379e-02
4.41093653e-01 -4.91560847e-01 9.74496305e-01 -1.65886390e+00
-1.29419565e+00 1.23049545e+00 3.18282247e-01 -6.46442294e-01
7.38484800e-01 1.10307500e-01 -3.72210026e-01 5.54705441e-01
-2.64780432e-01 7.29488969e-01 1.01570451e+00 -1.36974704e+00
-5.89909434e-01 -2.70266712e-01 1.75275758e-01 3.19751650e-01
-4.96939212e-01 -2.58967727e-01 -3.93136442e-01 -4.17452425e-01
-3.88564885e-01 -9.23358738e-01 -6.17456615e-01 1.51001466e-02
-4.09247518e-01 -9.21332613e-02 7.14315414e-01 -4.84656632e-01
9.11823690e-01 -2.19010496e+00 1.49045110e-01 6.14605904e-01
2.23039791e-01 7.73416102e-01 -6.10978007e-01 4.13734525e-01
-6.31653726e-01 9.15989503e-02 -2.40017712e-01 1.96986459e-02
-4.08341885e-01 3.01870793e-01 3.32227722e-02 3.22918773e-01
3.40782195e-01 7.52878189e-01 -6.23731017e-01 -6.36465073e-01
4.66356635e-01 4.16653961e-01 -3.35581183e-01 1.09227449e-01
-1.42626375e-01 4.85009313e-01 -4.12334502e-01 3.79666090e-01
5.94964921e-01 -4.62924033e-01 4.87988353e-01 -4.22801375e-01
2.12217420e-01 -1.87259346e-01 -9.27755773e-01 1.73104870e+00
-5.42888343e-01 4.90424961e-01 8.70547295e-02 -1.19248581e+00
6.31269634e-01 4.13247168e-01 8.15681458e-01 -7.15221941e-01
2.88008541e-01 2.32662588e-01 3.36689740e-01 -7.97147334e-01
-4.40256596e-02 -3.10907573e-01 3.82538140e-01 2.17743874e-01
4.95256543e-01 -1.08449616e-01 4.03306931e-01 -1.07677154e-01
1.11833465e+00 -1.16731405e-01 2.56620646e-01 -4.21653420e-01
6.34783745e-01 1.98193535e-01 1.10992782e-01 7.03886926e-01
-1.65426984e-01 5.57696640e-01 5.16001165e-01 -4.53751385e-01
-9.28398132e-01 -1.10471678e+00 -5.58171034e-01 8.58206749e-01
6.95729256e-02 -3.64121646e-02 -1.11140275e+00 -1.23935497e+00
2.50813700e-02 3.54929626e-01 -1.12144649e+00 -1.07086055e-01
-5.95260918e-01 -8.71722519e-01 5.01674473e-01 6.38734996e-01
4.55230176e-01 -9.89676058e-01 -4.76095289e-01 -2.69323494e-02
2.09543392e-01 -9.77991641e-01 -2.02443540e-01 4.92319018e-01
-8.95447135e-01 -1.51090968e+00 -8.93902838e-01 -1.05040324e+00
1.20551622e+00 -4.26266789e-02 7.18464494e-01 3.32224071e-01
-7.24212646e-01 4.71770465e-01 -2.80704737e-01 -2.63627946e-01
-5.48986673e-01 3.42213094e-01 -3.40267122e-01 1.77395254e-01
3.35354954e-01 -3.41652840e-01 -5.72085500e-01 1.27123132e-01
-1.28611088e+00 8.33165422e-02 1.05315423e+00 1.04699469e+00
5.49439073e-01 6.16168454e-02 5.55600584e-01 -1.25618124e+00
5.16855121e-01 -3.56659055e-01 -2.69165099e-01 5.54397643e-01
-4.36666846e-01 -1.51057258e-01 5.82727849e-01 -7.54919052e-01
-1.01237035e+00 4.75116044e-01 -2.56117493e-01 -3.90759557e-01
-4.80200946e-01 6.07190967e-01 2.70862699e-01 -5.66708386e-01
1.06489253e+00 1.74978673e-01 2.23060116e-01 -3.48177880e-01
7.53501579e-02 3.99096668e-01 4.47026551e-01 -4.87908870e-01
6.70506418e-01 4.04378533e-01 8.78657922e-02 -8.49679291e-01
-8.42143536e-01 -4.51857597e-01 -1.03799176e+00 -3.64108503e-01
9.24285293e-01 -5.33199906e-01 -3.34985465e-01 4.96407449e-01
-7.30083287e-01 -7.00162947e-01 -5.38915873e-01 4.01036829e-01
-4.54202116e-01 3.76632154e-01 -4.80708331e-01 -4.02598381e-01
-1.88045576e-01 -1.14614785e+00 6.55954361e-01 3.20201397e-01
-1.53615788e-01 -1.57232594e+00 -4.44931686e-02 2.16346130e-01
4.32651132e-01 4.38166738e-01 1.13438177e+00 -9.64497149e-01
-1.07133612e-01 -4.39589053e-01 -3.28527391e-01 7.43494511e-01
3.05558264e-01 -1.61589026e-01 -9.80709314e-01 -5.15368581e-01
-2.95072272e-02 -5.67194581e-01 1.16641438e+00 4.02958214e-01
1.49523079e+00 7.65903592e-02 -3.69001865e-01 6.66507840e-01
1.81495249e+00 1.68127730e-01 6.34639859e-01 2.69650429e-01
5.82374036e-01 8.50711703e-01 2.07375586e-01 -3.12383413e-01
4.45576617e-03 1.64567813e-01 5.35553694e-01 -8.34544182e-01
-5.19818723e-01 -1.34086996e-01 -8.40526968e-02 3.69278699e-01
-1.47805944e-01 -7.92095363e-02 -9.47602868e-01 7.25924850e-01
-1.55523038e+00 -6.08601987e-01 1.13613904e-01 2.16281056e+00
9.51772749e-01 8.96689370e-02 1.35193586e-01 1.43701538e-01
7.51038492e-01 3.83630767e-02 -6.27093375e-01 -3.50458860e-01
-1.31557425e-02 8.11717987e-01 6.69132829e-01 2.80622154e-01
-1.26386952e+00 8.45957458e-01 6.98209953e+00 1.00439918e+00
-1.64904952e+00 -6.75413460e-02 5.24103224e-01 1.63035274e-01
4.51041907e-02 -4.67121750e-01 -4.48728383e-01 2.54422337e-01
8.24403167e-01 1.17679741e-02 1.55456707e-01 8.79126549e-01
-3.10838282e-01 -2.21776858e-01 -1.34802067e+00 7.21562266e-01
1.34975761e-01 -1.36880755e+00 1.84208468e-01 2.68282145e-01
8.69882703e-01 -6.63906261e-02 3.44336629e-01 1.51866555e-01
1.24385774e-01 -1.26521230e+00 -4.13173111e-04 2.14284196e-01
1.05855811e+00 -5.04564941e-01 8.31766963e-01 1.10609896e-01
-7.31211305e-01 -8.80930051e-02 -1.75191745e-01 3.67332876e-01
-2.62882560e-01 3.92321765e-01 -1.18794811e+00 6.08150244e-01
3.45065743e-01 4.71223116e-01 -8.86678994e-01 1.39133561e+00
-2.17331544e-01 7.28036761e-01 -1.55928180e-01 -8.62994194e-02
5.05723476e-01 1.33141316e-02 1.23045594e-01 1.37133396e+00
-6.05557077e-02 -1.38232142e-01 2.21052587e-01 5.43244123e-01
1.29947081e-01 1.98681191e-01 -4.68067139e-01 -1.38515934e-01
-3.42064202e-02 1.31555152e+00 -6.14651561e-01 -2.64112562e-01
-2.53887177e-01 8.75350535e-01 3.09861153e-01 5.47375619e-01
-5.80023170e-01 -3.52406859e-01 2.53647178e-01 3.45762730e-01
3.08053404e-01 2.48831123e-01 -3.22152674e-01 -9.11459863e-01
-1.89148664e-01 -8.72429073e-01 5.31574786e-01 -4.38222796e-01
-1.46078563e+00 9.08816695e-01 -5.55833764e-02 -1.07045138e+00
-3.05772096e-01 -1.03832889e+00 -8.24802816e-01 9.12931025e-01
-1.67727423e+00 -1.29357123e+00 -4.84712362e-01 9.00587380e-01
2.66321421e-01 -5.66321984e-02 1.03784275e+00 2.88738936e-01
-5.18753707e-01 6.99913025e-01 -9.04431790e-02 2.75594741e-01
7.69479930e-01 -1.40959966e+00 -2.03556851e-01 6.29035115e-01
-9.90234464e-02 3.40158433e-01 3.26099843e-01 -4.50527310e-01
-9.30539906e-01 -1.18280971e+00 4.66319740e-01 -1.74585566e-01
7.69107997e-01 -2.70844735e-02 -9.64859128e-01 5.29370129e-01
3.35438728e-01 -4.75172997e-02 9.83560324e-01 1.79908574e-01
-2.61554211e-01 -2.87147969e-01 -1.33903646e+00 4.81395394e-01
7.19166577e-01 -5.88210285e-01 -4.88975793e-01 6.24362946e-01
1.29611984e-01 -5.11961281e-01 -1.02846122e+00 5.88392019e-01
3.49935025e-01 -5.48521698e-01 8.10523868e-01 -7.82964528e-01
5.96136808e-01 2.31307477e-01 2.36887619e-01 -1.39287162e+00
-1.46120951e-01 -1.86304778e-01 2.27444112e-01 6.99019909e-01
3.66443455e-01 -5.10861516e-01 1.12536395e+00 3.60030144e-01
3.76221426e-02 -1.01182771e+00 -7.76643097e-01 -7.31432676e-01
4.47566360e-01 -4.88435365e-02 8.90659988e-02 1.10142505e+00
-2.14931130e-01 8.44978169e-02 -8.18903595e-02 -3.61032672e-02
6.63210094e-01 -1.34835631e-01 3.56743634e-01 -1.15502644e+00
-2.25983515e-01 -3.98019433e-01 -3.84988993e-01 -7.31681466e-01
-6.62905946e-02 -9.96647656e-01 -5.90644171e-03 -1.44034171e+00
2.62822479e-01 -6.50847673e-01 -6.88820481e-01 6.95759773e-01
-9.66948718e-02 3.15533668e-01 1.47370026e-01 -7.36633912e-02
-2.87142754e-01 -2.30285242e-01 1.67830873e+00 -1.17279507e-01
-1.37454465e-01 7.40064532e-02 -5.62718451e-01 6.67253256e-01
8.09853017e-01 -2.64892608e-01 -6.34401977e-01 -4.20401186e-01
-3.39383036e-01 -1.01882242e-01 5.59699178e-01 -1.09922540e+00
4.10306603e-01 -1.74309120e-01 5.80303252e-01 -2.76315898e-01
2.97281414e-01 -8.23233426e-01 -1.05535395e-01 7.71385193e-01
-5.01999557e-01 -6.04567826e-01 4.19391781e-01 2.86522478e-01
-2.67414510e-01 -6.19387329e-01 1.13990855e+00 -2.50208229e-01
-8.53986204e-01 3.72414082e-01 -3.54351223e-01 -1.75002143e-01
1.16182363e+00 -4.34509218e-01 -3.62936378e-01 3.24453078e-02
-1.13134205e+00 8.45070183e-02 3.30136776e-01 2.69637287e-01
3.85558903e-01 -1.04627597e+00 -6.55674815e-01 2.58889645e-01
1.52552582e-03 -1.18232191e-01 5.29109955e-01 9.87339795e-01
-5.14791369e-01 4.54525113e-01 -6.44917488e-01 -7.80917764e-01
-1.38679087e+00 4.39129323e-01 7.03786552e-01 -5.80326855e-01
-4.18592036e-01 9.05745685e-01 4.59452778e-01 -4.07686800e-01
3.45943809e-01 -1.78625837e-01 -1.59482241e-01 -1.71975300e-01
3.89243633e-01 -5.10510467e-02 2.26923168e-01 -3.23899537e-01
-6.37232140e-02 5.53784430e-01 -5.12941957e-01 2.00893611e-01
1.19762266e+00 4.23518211e-01 1.38500750e-01 -1.57455996e-01
1.22996342e+00 -4.29004133e-01 -1.17936945e+00 -2.83743739e-01
-2.99503393e-02 -2.50773907e-01 4.38756458e-02 -1.35529304e+00
-1.17835462e+00 9.47492599e-01 8.42630744e-01 1.02519512e-01
1.46835172e+00 -1.43269435e-01 5.00086963e-01 2.73190856e-01
1.17244944e-01 -1.18387878e+00 2.43294537e-01 1.49962410e-01
8.57451558e-01 -1.03586924e+00 -2.15538144e-01 -7.02841640e-01
-7.49634624e-01 1.12964332e+00 8.99479270e-01 -4.55820680e-01
6.21390641e-01 1.63710922e-01 2.30887949e-01 -2.28008524e-01
-4.61546361e-01 -5.01952887e-01 5.27185857e-01 8.01832438e-01
1.50458112e-01 -2.21760906e-02 -4.23247129e-01 4.28598434e-01
9.04436186e-02 7.70858154e-02 2.72047967e-01 7.73055434e-01
-2.67950445e-01 -1.40837336e+00 -1.59945488e-02 7.06655324e-01
-4.61327583e-01 1.90045051e-02 -7.14809179e-01 1.20913398e+00
4.82086539e-01 5.22808850e-01 2.89970469e-02 -4.45184499e-01
3.08063179e-01 -4.62910831e-02 9.08496499e-01 -7.51636326e-01
-9.42992985e-01 4.91318889e-02 1.23457886e-01 -3.77245456e-01
-4.05829549e-01 -4.14356261e-01 -1.05986631e+00 7.70740882e-02
-2.71802217e-01 1.00124497e-02 6.73060656e-01 1.01395154e+00
-7.77074881e-03 6.87342405e-01 5.74784517e-01 -5.45655787e-01
-7.59940922e-01 -8.29675317e-01 -5.71581244e-01 6.19854152e-01
1.38192028e-01 -4.19685423e-01 -1.12713061e-01 1.09458901e-01] | [15.374006271362305, -2.7290492057800293] |
9af59514-c878-4607-b523-15c83e2803bd | fid-light-efficient-and-effective-retrieval | 2209.14290 | null | https://arxiv.org/abs/2209.14290v1 | https://arxiv.org/pdf/2209.14290v1.pdf | FiD-Light: Efficient and Effective Retrieval-Augmented Text Generation | Retrieval-augmented generation models offer many benefits over standalone language models: besides a textual answer to a given query they provide provenance items retrieved from an updateable knowledge base. However, they are also more complex systems and need to handle long inputs. In this work, we introduce FiD-Light to strongly increase the efficiency of the state-of-the-art retrieval-augmented FiD model, while maintaining the same level of effectiveness. Our FiD-Light model constrains the information flow from the encoder (which encodes passages separately) to the decoder (using concatenated encoded representations). Furthermore, we adapt FiD-Light with re-ranking capabilities through textual source pointers, to improve the top-ranked provenance precision. Our experiments on a diverse set of seven knowledge intensive tasks (KILT) show FiD-Light consistently improves the Pareto frontier between query latency and effectiveness. FiD-Light with source pointing sets substantial new state-of-the-art results on six KILT tasks for combined text generation and provenance retrieval evaluation, while maintaining reasonable efficiency. | ['Hamed Zamani', 'Karthik Raman', 'Jiecao Chen', 'Sebastian Hofstätter'] | 2022-09-28 | null | null | null | null | ['open-domain-question-answering'] | ['natural-language-processing'] | [-3.39180827e-02 -1.28687724e-01 -5.40732861e-01 1.05864264e-01
-1.47755527e+00 -8.08690786e-01 9.74178791e-01 6.05215192e-01
-5.83155215e-01 6.99062467e-01 5.55662811e-01 -2.27464169e-01
-1.52896181e-01 -9.01008844e-01 -8.92593324e-01 1.45273684e-02
-2.09790587e-01 6.20806992e-01 6.36678040e-01 -4.77653474e-01
4.80770409e-01 -4.24071029e-02 -1.57426286e+00 7.34348834e-01
9.01155531e-01 7.27729321e-01 -1.65444076e-01 8.50353062e-01
-2.04155698e-01 1.26203191e+00 -8.38550746e-01 -4.64994252e-01
-1.96772236e-02 7.34298453e-02 -1.21090627e+00 -1.13339150e+00
7.33697116e-01 -5.67308962e-01 -4.07440871e-01 4.57804352e-01
7.31075287e-01 -4.31169271e-02 4.97093827e-01 -1.26318836e+00
-1.29410386e+00 1.21018672e+00 -3.10958803e-01 2.35401347e-01
7.36254096e-01 7.76478723e-02 1.09266245e+00 -9.18949842e-01
1.05258584e+00 1.12462544e+00 4.43482161e-01 2.91007131e-01
-9.31781173e-01 -4.30981427e-01 -5.79902157e-02 2.45906487e-01
-1.37719750e+00 -7.86070347e-01 -1.33136615e-01 -7.94016644e-02
1.80984628e+00 4.24350798e-01 4.31404948e-01 7.41919577e-01
2.32644394e-01 8.70314479e-01 5.43938816e-01 -4.88661557e-01
7.48185962e-02 -4.34413329e-02 3.84637445e-01 7.19132721e-01
5.07869601e-01 -9.34866369e-02 -1.37005854e+00 -4.19955194e-01
6.13425784e-02 -4.68404084e-01 -3.26040179e-01 -8.55661333e-02
-1.32943475e+00 3.98215294e-01 3.50055248e-01 9.06672105e-02
-2.88912207e-01 8.92948687e-01 6.66578650e-01 5.10912895e-01
5.23517966e-01 7.16029584e-01 -3.94410342e-01 -4.46806401e-01
-1.08390808e+00 5.42342424e-01 1.02858949e+00 1.38727379e+00
5.80418646e-01 -2.94796705e-01 -1.08368897e+00 4.49053198e-01
3.27206910e-01 7.78766811e-01 3.43099833e-01 -8.64653647e-01
6.96414292e-01 5.19170463e-01 2.87623763e-01 -4.79062617e-01
5.29931076e-02 -6.45065665e-01 -1.72391027e-01 -3.08751166e-01
1.06984027e-01 1.33046463e-01 -8.80187571e-01 1.59232914e+00
1.38766766e-01 -1.75109461e-01 3.35956126e-01 6.86186016e-01
8.47756147e-01 9.05228913e-01 8.32670853e-02 -2.42917053e-02
1.41343081e+00 -8.33490193e-01 -6.12196088e-01 -2.89725691e-01
9.16499138e-01 -9.17315304e-01 1.14254200e+00 2.29496032e-01
-1.27356660e+00 -2.31702417e-01 -1.04200566e+00 -6.31241918e-01
-4.53748167e-01 9.01581999e-03 7.05097735e-01 4.40694332e-01
-1.59370708e+00 4.65500176e-01 -7.36584544e-01 -3.54969084e-01
1.27844110e-01 -1.34478003e-01 -1.91504687e-01 -1.93509534e-01
-1.47521174e+00 8.56592476e-01 5.56024909e-01 -2.71607697e-01
-8.83745253e-01 -1.16005123e+00 -5.27253211e-01 1.64268270e-01
6.59112036e-01 -1.19507527e+00 1.73285365e+00 1.88116595e-01
-1.22708178e+00 4.71076846e-01 -1.95377588e-01 -4.62284982e-01
3.63707691e-01 -8.29147637e-01 -2.73142576e-01 2.38023743e-01
2.65896201e-01 6.59952998e-01 6.06600940e-01 -1.00513601e+00
-5.00996768e-01 -4.00953889e-02 1.39302492e-01 8.55880827e-02
-4.55544382e-01 1.08851328e-01 -1.00391400e+00 -5.95773578e-01
-4.48238492e-01 -7.73708642e-01 4.41134334e-01 -1.03032619e-01
-3.91443819e-01 -4.09304231e-01 6.41641200e-01 -5.79304814e-01
1.81597662e+00 -1.83796585e+00 9.27296653e-02 -8.17289483e-03
1.50764033e-01 2.77651489e-01 -2.90191710e-01 1.06445861e+00
5.05626798e-01 7.06476152e-01 1.09856218e-01 -2.83803225e-01
2.29757115e-01 1.39218450e-01 -7.87981093e-01 -1.38763756e-01
1.53286144e-01 1.39220297e+00 -1.21732962e+00 -7.78266668e-01
-5.65568686e-01 3.52193117e-01 -5.81857800e-01 8.24682713e-02
-9.80426431e-01 -3.15564036e-01 -4.75267410e-01 6.38887405e-01
1.49046406e-01 -4.96830195e-01 -2.48540625e-01 2.26013418e-02
-1.58229116e-02 7.66855001e-01 -8.02973390e-01 2.19968486e+00
-5.68652630e-01 5.08537471e-01 -3.59726965e-01 2.91383564e-01
4.20558155e-01 4.71920460e-01 2.42526501e-01 -9.26380336e-01
-5.09840488e-01 3.50863338e-01 -5.74039102e-01 -5.12072384e-01
1.52059805e+00 5.33096671e-01 -3.21786135e-01 1.13044190e+00
-3.70489955e-02 -1.82659864e-01 5.75999916e-01 1.06596708e+00
1.40361619e+00 3.63339871e-01 -1.96223572e-01 1.05552137e-01
-8.29412714e-02 2.25715235e-01 -7.95275047e-02 1.20663118e+00
6.60154819e-01 2.43353829e-01 4.14360821e-01 8.66471231e-02
-8.46544504e-01 -8.62690032e-01 1.54982388e-01 1.27755022e+00
1.48947686e-01 -1.24676752e+00 -4.85900283e-01 -6.40274286e-01
3.53645205e-01 1.13260353e+00 -4.63542968e-01 -5.25341272e-01
-6.40905082e-01 -3.82656783e-01 1.21880817e+00 6.65960550e-01
3.17567587e-01 -8.43590260e-01 -9.86717701e-01 1.17533542e-01
-5.36876917e-01 -6.70285761e-01 -5.88863373e-01 -7.57659450e-02
-6.54963613e-01 -7.13029563e-01 -5.86340904e-01 2.19407603e-02
1.58062175e-01 3.45125526e-01 1.61495757e+00 3.76833797e-01
-2.33503148e-01 6.39755011e-01 -7.56505787e-01 -4.21164274e-01
-5.23533821e-01 5.78453302e-01 -3.20418298e-01 -7.55307019e-01
1.22352559e-02 -1.83033183e-01 -5.53965688e-01 6.14711195e-02
-1.29845083e+00 -8.33739564e-02 5.79672873e-01 6.74555004e-01
5.55105388e-01 -3.17142189e-01 5.56011617e-01 -7.69055665e-01
1.02510512e+00 -5.23928285e-01 -4.75063890e-01 9.19403076e-01
-1.21348011e+00 8.00135255e-01 3.15387130e-01 -1.28660947e-01
-1.21011639e+00 -7.67771423e-01 2.21831158e-01 -3.09805185e-01
7.84000576e-01 1.11526895e+00 4.50156212e-01 2.81103760e-01
1.05735445e+00 2.39637792e-01 -3.53398532e-01 -6.13634586e-01
1.07699370e+00 5.02600610e-01 7.29944408e-01 -1.02871335e+00
5.88339567e-01 8.46139789e-02 -1.89405292e-01 -9.02081802e-02
-6.23826027e-01 -4.59066659e-01 -3.22933137e-01 -1.05848953e-01
2.95386910e-01 -8.94411981e-01 -3.90924692e-01 3.30930859e-01
-1.29650474e+00 -3.51120651e-01 -6.07636154e-01 4.13963571e-02
-2.70291984e-01 2.76700467e-01 -7.08735406e-01 -6.63440347e-01
-1.17104089e+00 -7.70369709e-01 1.59082699e+00 -1.22859672e-01
-2.61148721e-01 -5.36693156e-01 4.12286311e-01 1.78639755e-01
7.67132938e-01 -2.33945534e-01 1.23017764e+00 -2.98576295e-01
-1.15799689e+00 -3.26644540e-01 -3.18340033e-01 -1.78526640e-01
-3.22614312e-01 2.00594530e-01 -1.01410651e+00 -3.08745742e-01
-7.61503398e-01 -7.49960423e-01 1.15838969e+00 -3.78291547e-01
8.12306941e-01 -5.84452093e-01 -5.10057747e-01 4.40678269e-01
1.42788720e+00 -8.05606321e-02 6.45926535e-01 3.33802432e-01
3.48929495e-01 1.75258577e-01 7.36806214e-01 5.04348695e-01
4.93778050e-01 7.16897488e-01 3.40615481e-01 5.54925263e-01
-6.31475449e-01 -6.61345899e-01 5.08124292e-01 6.97464108e-01
1.34802461e-01 -9.86357212e-01 -9.67536569e-01 7.27699876e-01
-2.09842706e+00 -1.02261913e+00 -1.37375921e-01 2.31743264e+00
1.29239547e+00 1.61565319e-02 -2.43883669e-01 -6.37834072e-02
6.88938191e-03 1.84662819e-01 -5.06484866e-01 -2.59763002e-01
1.41221564e-02 3.53106111e-01 5.84698379e-01 4.36491817e-01
-3.33045274e-01 1.06803155e+00 6.88591528e+00 9.95926678e-01
-1.06532729e+00 2.71944463e-01 -1.50165856e-01 -5.30079961e-01
-8.99987400e-01 2.06068262e-01 -1.00025642e+00 3.53350192e-01
1.11855650e+00 -7.08896935e-01 4.40363675e-01 5.23857296e-01
-4.68592882e-01 -4.02100623e-01 -1.37853098e+00 5.41434169e-01
1.95340499e-01 -1.56416726e+00 4.05243725e-01 -1.40284672e-01
3.63144219e-01 2.35756174e-01 1.07799932e-01 4.55016792e-01
6.21772170e-01 -8.52170289e-01 1.23819053e+00 1.00324130e+00
1.12740326e+00 -4.08649951e-01 5.37162423e-01 1.66646048e-01
-1.03057754e+00 2.56799608e-01 -1.40529960e-01 3.38880718e-01
3.21966022e-01 6.03602052e-01 -1.21217358e+00 1.07017231e+00
7.29793429e-01 4.21605527e-01 -1.04010475e+00 1.01991510e+00
-3.30435067e-01 4.54301834e-01 -5.16474187e-01 -4.33789253e-01
-2.43378505e-02 7.85585403e-01 5.53248584e-01 1.35493588e+00
5.23613095e-01 -2.48159349e-01 -1.77941069e-01 7.97517240e-01
-3.74508083e-01 -1.35143623e-01 -5.94431043e-01 -2.40478992e-01
9.33566272e-01 8.60495806e-01 -2.06848174e-01 -6.69138908e-01
8.38381425e-02 8.18688631e-01 4.10279930e-01 3.01446527e-01
-7.19761550e-01 -5.56179821e-01 3.63143921e-01 1.81436345e-01
1.84308782e-01 -1.62929565e-01 6.35636374e-02 -1.20881212e+00
4.24205184e-01 -8.42610180e-01 5.41972756e-01 -1.32915008e+00
-8.93803954e-01 5.93300283e-01 5.31305254e-01 -8.16762507e-01
-7.58463323e-01 2.82838028e-02 -7.74383778e-03 1.09182227e+00
-1.72915149e+00 -1.04013705e+00 -2.12069318e-01 5.12868941e-01
1.09834209e-01 2.48464480e-01 1.01085663e+00 4.19514596e-01
-2.39167020e-01 8.07468116e-01 2.25668401e-01 -1.34791613e-01
1.12746048e+00 -1.25059736e+00 7.37222910e-01 1.07298267e+00
1.72174260e-01 1.26389086e+00 2.58270293e-01 -1.00814223e+00
-2.03851414e+00 -1.02436495e+00 1.19293594e+00 -9.07492757e-01
6.54199898e-01 -2.09670514e-01 -8.14996004e-01 4.31738794e-01
4.93468791e-01 -1.85394838e-01 5.03856480e-01 2.44077295e-01
-9.43218768e-01 -1.39948830e-01 -5.97035229e-01 5.79708219e-01
1.05948341e+00 -9.80661929e-01 -5.39524317e-01 4.28442836e-01
1.29767346e+00 -4.77850288e-01 -1.06565893e+00 3.08010042e-01
7.93357313e-01 -4.28931087e-01 1.03231275e+00 -4.72092271e-01
5.21248460e-01 -5.31729341e-01 -5.67740388e-02 -1.20913970e+00
-1.01560690e-01 -9.01008427e-01 -8.07865202e-01 1.27395022e+00
6.12850785e-01 -4.54886198e-01 4.50686067e-01 5.80916643e-01
-2.46702313e-01 -6.13205492e-01 -7.86486447e-01 -9.51551378e-01
-1.45401835e-01 -6.29654229e-01 8.52768004e-01 3.53056759e-01
2.54321605e-01 5.59281766e-01 -1.96175158e-01 -8.94209594e-02
3.87022763e-01 5.93212359e-02 7.68117905e-01 -9.45622861e-01
-3.96589875e-01 -2.09363982e-01 3.17229509e-01 -1.09591138e+00
-2.06386700e-01 -1.16021919e+00 6.22749366e-02 -1.96121991e+00
1.88708961e-01 -8.19091916e-01 -1.06395639e-01 1.00262523e+00
-2.91000009e-01 -8.10289904e-02 1.93919420e-01 7.18466520e-01
-1.01147139e+00 4.91343230e-01 8.66163492e-01 -3.63771796e-01
-1.30532007e-03 -4.02410924e-01 -9.77109671e-01 -1.81692407e-01
4.09263790e-01 -7.04568207e-01 -9.45841610e-01 -1.33579910e+00
1.12334418e+00 3.39678168e-01 3.89684051e-01 -7.84463406e-01
8.30807865e-01 1.49747729e-01 -4.98516038e-02 -6.82725966e-01
4.32534754e-01 -2.78245449e-01 2.51220882e-01 3.02031130e-01
-7.37035751e-01 4.92781669e-01 3.83624643e-01 5.58090329e-01
-1.11460350e-01 -3.59978527e-01 -8.05217102e-02 -7.34505942e-03
-6.47799790e-01 1.56125233e-01 -1.42750829e-01 4.72033262e-01
4.61305678e-01 1.55786291e-01 -1.42955172e+00 -3.41691554e-01
1.50613517e-01 4.82075095e-01 5.72661996e-01 5.47499835e-01
6.37896836e-01 -1.07098854e+00 -7.92297482e-01 -1.37853637e-01
7.04942167e-01 -2.91766822e-02 -8.17499682e-02 4.91783917e-01
-5.90203762e-01 9.87396896e-01 1.59564093e-01 -3.65314245e-01
-1.11514938e+00 6.02474213e-01 -6.59862980e-02 -7.15465903e-01
-3.58245552e-01 9.56870317e-01 -6.41620517e-01 -9.79196355e-02
3.00978541e-01 -4.80878145e-01 1.87011391e-01 1.00446455e-01
6.92579329e-01 5.28506756e-01 5.87454557e-01 6.81723058e-02
-3.98210198e-01 1.52441293e-01 -3.50963473e-01 -6.34733677e-01
9.30471241e-01 1.32877007e-01 -4.92609292e-01 1.61444694e-01
1.00952423e+00 1.96214616e-01 -5.93517661e-01 -4.29645956e-01
1.92767128e-01 -4.98752147e-01 2.04523295e-01 -1.55165839e+00
-7.05835938e-01 6.07784629e-01 1.85071960e-01 8.46484229e-02
9.88761008e-01 -9.61195752e-02 9.84184444e-01 9.32899892e-01
8.42140615e-01 -7.48498797e-01 2.48956293e-01 6.68360531e-01
8.47193599e-01 -6.00182593e-01 2.31211215e-01 -7.93540627e-02
-2.85454810e-01 7.50918150e-01 4.28553432e-01 5.79927862e-01
3.10590804e-01 2.50649124e-01 -1.78188428e-01 -4.25692409e-01
-1.54930067e+00 2.27934979e-02 3.76879573e-01 4.65744734e-01
6.29161179e-01 -3.63734841e-01 -1.66572914e-01 2.21599728e-01
-1.91485763e-01 3.81947786e-01 3.65423143e-01 1.39431000e+00
-3.08769345e-01 -1.36348128e+00 -1.94348291e-01 4.83200014e-01
-5.95907211e-01 -7.04252243e-01 -4.60868388e-01 7.29222178e-01
-3.39356750e-01 1.02036357e+00 -2.45266765e-01 -4.74889696e-01
3.23681653e-01 2.41574362e-01 5.13615310e-01 -7.70507097e-01
-1.04912031e+00 -2.69021809e-01 5.01929045e-01 -6.67022705e-01
-1.43996000e-01 -3.76026034e-01 -1.09885883e+00 -5.95265567e-01
-3.65299344e-01 5.22828877e-01 6.19505882e-01 3.90657991e-01
1.02250874e+00 4.98110622e-01 -6.33192807e-02 4.23479639e-03
-6.62594378e-01 -9.98548567e-01 3.95578705e-02 -1.13885157e-01
2.47543260e-01 -1.80235729e-01 1.22203991e-01 4.02631275e-02] | [11.3593111038208, 7.829767227172852] |
e052bd07-7b5a-488e-8d6c-91f8717010b7 | specmar-fast-heart-rate-estimation-from-ppg | 1810.06196 | null | http://arxiv.org/abs/1810.06196v2 | http://arxiv.org/pdf/1810.06196v2.pdf | SPECMAR: Fast Heart Rate Estimation from PPG Signal using a Modified Spectral Subtraction Scheme with Composite Motion Artifacts Reference Generation | The task of heart rate estimation using photoplethysmographic (PPG) signal is
challenging due to the presence of various motion artifacts in the recorded
signals. In this paper, a fast algorithm for heart rate estimation based on
modified SPEctral subtraction scheme utilizing Composite Motion Artifacts
Reference generation (SPECMAR) is proposed using two-channel PPG and three-axis
accelerometer signals. First, the preliminary noise reduction is obtained by
filtering unwanted frequency components from the recorded signals. Next, a
composite motion artifacts reference generation method is developed to be
employed in the proposed SPECMAR algorithm for motion artifacts reduction. The
heart rate is then computed from the noise and motion artifacts reduced PPG
signal. Finally, a heart rate tracking algorithm is proposed considering
neighboring estimates. The performance of the SPECMAR algorithm has been tested
on publicly available PPG database. The average heart rate estimation error is
found to be 2.09 BPM on 23 recordings. The Pearson correlation is 0.9907. Due
to low computational complexity, the method is faster than the comparing
methods. The low estimation error, smooth and fast heart rate tracking makes
SPECMAR an ideal choice to be implemented in wearable devices. | ['Shaikh Anowarul Fattah', 'Sk. Tanvir Ahmed', 'Mohammad Tariqul Islam', 'Celia Shahnaz'] | 2018-10-15 | null | null | null | null | ['heart-rate-estimation'] | ['medical'] | [ 4.64514911e-01 -3.84886116e-01 3.54100525e-01 1.06689304e-01
-5.16409993e-01 -2.11078808e-01 9.52173918e-02 -1.85775697e-01
-2.77279109e-01 8.16646934e-01 1.05035909e-01 1.71455249e-01
9.04425457e-02 -2.25592121e-01 -3.79213528e-03 -7.75400996e-01
-4.52165082e-02 -5.27218163e-01 -5.39389327e-02 3.16728711e-01
2.64423102e-01 2.43500113e-01 -9.52968836e-01 -5.18958092e-01
8.62921596e-01 1.05730951e+00 -1.25270989e-02 1.18888044e+00
5.21997154e-01 3.22904855e-01 -8.48173976e-01 1.40664786e-01
2.99268842e-01 -1.04235554e+00 1.81383407e-03 -5.05205840e-02
1.05255179e-01 -3.64411771e-01 1.97520524e-01 8.67067814e-01
1.01098287e+00 8.54158625e-02 2.18254521e-01 -6.47208035e-01
1.82017744e-01 1.59103036e-01 -7.14810491e-01 4.66776848e-01
4.22680676e-01 2.03982508e-03 1.92931503e-01 -8.33311260e-01
4.81032938e-01 4.32778955e-01 1.22811389e+00 2.10475817e-01
-1.43637371e+00 -5.77976763e-01 -8.64767730e-01 1.38109744e-01
-1.45411575e+00 -3.15832168e-01 1.21435797e+00 -4.14559662e-01
5.51050663e-01 7.08723962e-01 1.05924726e+00 5.53536654e-01
5.99940002e-01 -2.20057979e-01 1.57887173e+00 -4.56024438e-01
2.04950750e-01 1.94960847e-01 2.93870211e-01 4.67761993e-01
7.05012441e-01 1.50779812e-02 -5.78590155e-01 -4.49784577e-01
8.51061702e-01 -8.45268369e-02 -6.36286139e-01 9.62798670e-02
-1.29545546e+00 3.56419176e-01 -3.63264084e-01 4.17555094e-01
-6.45737112e-01 1.74372673e-01 5.07528543e-01 1.32772952e-01
2.02991307e-01 2.65276223e-01 -1.54284582e-01 -7.07283258e-01
-1.17801011e+00 1.11734206e-02 9.37715709e-01 4.12553519e-01
2.30989560e-01 3.94293815e-01 6.61861822e-02 6.17099762e-01
5.13158500e-01 6.87588513e-01 7.05509186e-01 -9.95648026e-01
-1.80656873e-02 2.88348824e-01 4.55628097e-01 -1.21655571e+00
-6.24546170e-01 -3.67148370e-01 -9.50775743e-01 -1.65936127e-02
4.46643621e-01 -3.72572839e-01 -3.32801849e-01 1.20301902e+00
6.63185537e-01 7.00410724e-01 3.74156348e-02 1.08572912e+00
7.72752941e-01 6.29528165e-01 8.30203444e-02 -8.21000874e-01
1.47450101e+00 -1.29677281e-01 -1.28402531e+00 6.63295090e-02
2.34785452e-01 -9.47810590e-01 5.55906892e-01 7.16636539e-01
-1.02197671e+00 -8.31213176e-01 -1.23834312e+00 2.71652877e-01
4.42166090e-01 4.24017608e-01 6.03816472e-02 1.19064713e+00
-6.03021801e-01 8.90408874e-01 -9.65672433e-01 -3.80093396e-01
-2.27991909e-01 1.41087338e-01 -4.08062935e-02 5.80518603e-01
-1.05701280e+00 7.31667757e-01 6.45219982e-02 4.12453622e-01
2.44558807e-02 -6.48464680e-01 -7.36778021e-01 -1.94422826e-01
-2.72104561e-01 -5.49667835e-01 9.29859579e-01 -6.84805274e-01
-2.19298387e+00 4.23337579e-01 -2.62692630e-01 -3.58254015e-01
6.87553465e-01 -2.52848536e-01 -6.46327138e-01 3.13751221e-01
-2.99799979e-01 -4.52397943e-01 1.16548443e+00 -4.78298873e-01
-3.07361130e-02 -3.01406205e-01 -1.08253455e+00 2.88217276e-01
1.50155500e-01 -5.50996214e-02 -1.26152918e-01 -6.62997127e-01
6.73639476e-01 -9.20109332e-01 2.19872706e-02 -3.98204863e-01
3.24229337e-03 5.52009106e-01 4.54711527e-01 -1.18144906e+00
1.59615672e+00 -2.24818826e+00 -2.73207426e-01 4.93962884e-01
8.21118578e-02 4.11088675e-01 5.58458924e-01 1.86417684e-01
7.50977430e-04 -2.43880898e-01 -1.64201304e-01 2.72385292e-02
-4.84129995e-01 -3.06600928e-01 6.58517331e-02 9.62905169e-01
-2.19649419e-01 3.71244609e-01 -5.31211138e-01 -2.58188933e-01
7.01180577e-01 8.53973806e-01 7.15453736e-03 2.08425164e-01
8.20669889e-01 8.15110028e-01 -8.99536014e-02 4.66341138e-01
9.09921050e-01 2.57027745e-01 3.39566797e-01 -6.96535707e-01
-3.05282921e-01 2.01783046e-01 -1.54840326e+00 1.40039349e+00
-4.02967125e-01 7.23413825e-01 8.37758649e-03 -4.68751192e-01
1.47327542e+00 6.90159857e-01 6.06158674e-01 -3.75555009e-01
3.06458980e-01 2.53907979e-01 2.55493224e-01 -7.93958187e-01
2.23088160e-01 -4.84971493e-01 3.36416960e-01 3.80719602e-01
-2.63146579e-01 -2.37358660e-01 4.03748453e-02 -1.72315136e-01
9.33975875e-01 3.44793856e-01 8.70261729e-01 -4.50621158e-01
8.21901500e-01 -3.70425433e-01 6.75189555e-01 4.58962351e-01
-6.76209569e-01 7.09934235e-01 8.36512968e-02 -5.63026845e-01
-6.60470426e-01 -7.79340386e-01 -1.67414844e-01 -6.41078427e-02
6.72706738e-02 -5.02937436e-01 -7.08671391e-01 1.92677021e-01
-2.45985687e-01 2.21866265e-01 -2.93215543e-01 7.33394101e-02
-6.83533370e-01 -8.79869044e-01 5.10389745e-01 3.03234637e-01
6.21234894e-01 -7.63441801e-01 -1.11723721e+00 7.79989123e-01
-2.93127894e-01 -9.40187275e-01 -3.49299490e-01 -3.52761596e-01
-1.38648069e+00 -1.02335298e+00 -7.70990849e-01 -1.92641914e-01
7.96874687e-02 1.16747841e-02 6.93977833e-01 -2.71426082e-01
-8.00498188e-01 4.01003808e-01 -1.09732680e-01 -4.67429072e-01
-1.80126578e-01 -6.01278663e-01 3.18207324e-01 2.83285797e-01
4.33753461e-01 -6.16465330e-01 -1.00982547e+00 1.22501291e-01
-1.49769098e-01 -1.55460611e-01 2.57248998e-01 5.08622229e-01
5.12117028e-01 -2.09804043e-01 5.80106378e-01 -5.03757358e-01
9.29992199e-01 -4.43738289e-02 -8.38127017e-01 -4.22717124e-01
-6.24702871e-01 -5.45421600e-01 5.70967197e-01 -5.54167926e-01
-9.43102419e-01 2.59069711e-01 2.42781155e-02 -4.30899598e-02
3.68174701e-03 2.10408151e-01 2.54100919e-01 -7.74162859e-02
7.26456761e-01 3.77961516e-01 2.68248558e-01 -2.73330510e-01
-1.63687348e-01 4.43071932e-01 8.34235489e-01 -6.56605437e-02
5.24852633e-01 2.70652056e-01 3.04357916e-01 -1.39416015e+00
1.73203903e-03 -5.62082052e-01 -2.97848880e-01 -4.91215795e-01
9.62227941e-01 -9.16004717e-01 -8.22770715e-01 7.76407480e-01
-1.07784486e+00 5.19395433e-02 -2.11183447e-02 1.26433897e+00
-3.97334486e-01 9.37352657e-01 -6.02323353e-01 -1.40954614e+00
-1.00112998e+00 -5.92332482e-01 1.96542889e-01 5.30744016e-01
-6.86581135e-01 -8.02687287e-01 3.87119919e-01 3.07234317e-01
5.12859583e-01 8.24135482e-01 -7.57058337e-02 1.95850000e-01
-5.15008532e-02 -6.10755444e-01 2.67266035e-01 4.83157605e-01
4.51406807e-01 3.62129463e-03 -9.11043823e-01 -1.74585313e-01
8.43906283e-01 3.76000851e-01 1.29347682e-01 8.59793127e-01
5.12452841e-01 -5.10320663e-02 4.86181537e-03 6.01999044e-01
1.64233172e+00 3.50496739e-01 9.25140142e-01 5.31444289e-02
4.46147054e-01 2.10650608e-01 6.36539042e-01 8.26655865e-01
-1.37851298e-01 3.28355134e-01 -1.93353683e-01 -2.79477000e-01
1.46640502e-02 2.26370245e-01 4.65515584e-01 1.02377832e+00
-5.35207212e-01 4.05787766e-01 -5.69096923e-01 2.17607290e-01
-1.42709661e+00 -9.41472530e-01 -8.92512143e-01 2.54321837e+00
7.63406813e-01 -1.18182920e-01 1.48444042e-01 5.91483593e-01
8.53526652e-01 -4.27881582e-03 -3.86235565e-01 -7.36603856e-01
1.60447136e-01 6.03397548e-01 5.97162724e-01 4.64787483e-01
-8.25799763e-01 -4.09116969e-02 5.69128227e+00 -1.53361812e-01
-1.28406310e+00 1.09299086e-01 2.92849451e-01 -3.05142384e-02
4.76786375e-01 -3.37650627e-02 -5.49512684e-01 8.03239942e-01
1.27611279e+00 -3.71729821e-01 9.50641781e-02 6.09564662e-01
9.09921765e-01 -6.70490801e-01 -4.74268526e-01 1.61983514e+00
1.76967189e-01 -7.94662237e-01 -8.40788722e-01 -2.26498440e-01
3.88962388e-01 -5.37228405e-01 -4.45398420e-01 -3.91844243e-01
-8.20973516e-01 -4.71645355e-01 1.64427951e-01 8.00034046e-01
7.58553982e-01 -6.31455600e-01 7.00795889e-01 1.03742778e-01
-1.31594634e+00 2.93170333e-01 -2.50781804e-01 -3.97796601e-01
2.46844500e-01 1.07640791e+00 -7.64452040e-01 4.95643020e-01
1.59099191e-01 5.61380625e-01 -1.52429253e-01 1.28837454e+00
-1.46327242e-01 9.90957856e-01 -6.53726220e-01 1.08128019e-01
-5.24338841e-01 -5.33781946e-01 7.03817308e-01 1.09071386e+00
6.83884442e-01 6.16168737e-01 -3.65643680e-01 7.61416852e-01
4.03158188e-01 4.67221081e-01 -4.73930031e-01 1.51699513e-01
3.71169388e-01 1.32715356e+00 -5.91944873e-01 -4.22998369e-01
-4.59211469e-01 6.35424733e-01 -7.68044174e-01 2.02570986e-02
-9.97204065e-01 -9.35183585e-01 3.47510517e-01 2.68926114e-01
-3.12632978e-01 -2.54562438e-01 -6.06730759e-01 -9.96632218e-01
1.87152371e-01 -7.08905935e-01 2.97900587e-01 -6.38943732e-01
-6.13739252e-01 4.84732807e-01 -3.76511008e-01 -1.55571461e+00
-3.05197150e-01 5.44886179e-02 -9.43063021e-01 1.42035472e+00
-1.11187983e+00 -7.12671429e-02 -5.83535135e-01 4.74748194e-01
4.35319871e-01 2.80636817e-01 8.28853130e-01 4.32736158e-01
-5.56186736e-01 3.00752908e-01 -2.78052717e-01 -3.51715624e-01
7.28515804e-01 -9.83043849e-01 8.15529004e-03 1.25110626e+00
-4.62896883e-01 7.48230696e-01 8.36733043e-01 -8.06020260e-01
-1.52888870e+00 -7.05239892e-01 1.04603410e+00 -2.78220754e-02
4.10078824e-01 -8.20923480e-04 -8.96173894e-01 3.14452387e-02
1.10555761e-01 2.14361191e-01 7.76228786e-01 -6.87162340e-01
3.53966683e-01 -4.44511473e-01 -1.29905784e+00 1.51756153e-01
4.89593707e-02 -3.41715455e-01 -6.38765514e-01 -2.78258443e-01
-2.09806874e-01 -5.30011237e-01 -1.46823800e+00 2.46201456e-01
1.13412380e+00 -8.60255420e-01 5.96478045e-01 4.11057889e-01
-2.22878247e-01 -6.76397920e-01 4.73577857e-01 -8.54786813e-01
-2.27129325e-01 -1.54311800e+00 -1.71595111e-01 1.24007630e+00
-1.19799972e-02 -8.54591608e-01 5.89437306e-01 8.12818170e-01
3.25142503e-01 -8.47942084e-02 -7.09963441e-01 -5.53919613e-01
-9.00105417e-01 -2.87146866e-01 -2.13081613e-01 8.68059754e-01
3.69202912e-01 2.09866136e-01 -9.18106496e-01 1.71551839e-01
9.61682618e-01 -6.70322031e-02 7.27675855e-01 -1.24041128e+00
-3.38752002e-01 5.05488038e-01 -4.35159177e-01 -3.51448178e-01
-8.18550587e-01 -1.82430625e-01 -1.15439199e-01 -1.19117522e+00
-1.15659811e-01 4.08125296e-02 -5.48627377e-01 -2.04115376e-01
-4.74427462e-01 7.74931729e-01 1.94405943e-01 7.65765011e-02
1.78067252e-01 3.05296201e-02 7.69992828e-01 4.99904335e-01
-1.04186559e+00 2.59322554e-01 -1.26601741e-01 6.87707484e-01
9.72017765e-01 -4.37355816e-01 -3.91937673e-01 5.03033400e-01
-2.35752668e-02 5.48917532e-01 1.37053430e-01 -1.50633693e+00
-3.90119888e-02 3.87491822e-01 6.85940623e-01 -7.28861690e-01
3.90106767e-01 -6.37453318e-01 1.06012154e+00 9.75672781e-01
2.16005415e-01 1.44653887e-01 1.57856554e-01 2.55194694e-01
-1.39282912e-01 -4.83705588e-02 1.06613886e+00 -2.28926823e-01
-2.05794722e-01 -2.28336141e-01 -6.87134027e-01 -1.75464943e-01
6.08634055e-01 -8.73124182e-01 1.66876204e-02 -2.73541629e-01
-9.30675328e-01 -4.76383865e-01 6.07083328e-02 -1.51161954e-01
7.50884235e-01 -1.27856028e+00 -6.24556959e-01 3.29992175e-01
-2.27317974e-01 -5.91533661e-01 6.24317050e-01 1.54751217e+00
-9.23858225e-01 4.63629179e-02 -4.17489916e-01 -5.96706748e-01
-1.59155321e+00 3.09451018e-02 4.41176593e-01 -2.47024768e-03
-9.08109725e-01 2.68680304e-01 -5.02896726e-01 7.67254531e-01
-1.75730661e-01 -4.54222322e-01 -2.94602811e-01 -7.67967012e-03
8.14548075e-01 9.48325336e-01 1.20707661e-01 -4.01169568e-01
-4.61766750e-01 9.67120469e-01 5.11746883e-01 -1.94508716e-01
9.45333540e-01 -5.35274744e-01 -1.48874447e-02 4.97214556e-01
9.73223567e-01 3.39577436e-01 -8.19256365e-01 2.60347247e-01
-1.10500090e-01 -6.42079830e-01 5.66517748e-02 -6.06797695e-01
-8.81059229e-01 6.34906590e-01 1.22894382e+00 1.03332326e-01
1.53214157e+00 -1.14279950e+00 7.20708668e-01 -8.41876790e-02
7.84735195e-03 -1.25620675e+00 -2.29375705e-01 -1.79303437e-01
5.12878835e-01 -5.58465481e-01 4.78506923e-01 -4.15003091e-01
-4.53886747e-01 1.22881866e+00 1.17735952e-01 -3.73155922e-01
4.82541054e-01 1.89863726e-01 4.17001992e-01 3.67806226e-01
-2.93037027e-01 1.46651894e-01 1.37870256e-02 6.43655062e-01
8.75148535e-01 -8.20860714e-02 -1.45702779e+00 3.01478773e-01
6.60166740e-02 6.58955693e-01 1.10973442e+00 7.17041016e-01
-2.64080882e-01 -7.65492976e-01 -7.11725712e-01 2.72829235e-01
-1.05751240e+00 6.75533488e-02 -3.89913544e-02 5.32163799e-01
-2.04887286e-01 1.19911873e+00 -3.09318155e-01 -8.67110118e-02
3.52697313e-01 3.30929250e-01 4.84535635e-01 -1.96472034e-01
-8.58674645e-01 6.26272917e-01 6.25007227e-02 -7.00289905e-01
-4.94254768e-01 -6.25067353e-01 -1.12007594e+00 9.51376855e-02
-3.50235879e-01 1.84924155e-01 1.07067370e+00 2.52263427e-01
4.26364630e-01 3.13489288e-01 6.90524042e-01 -5.26254594e-01
-3.48187834e-01 -9.95719492e-01 -9.52802837e-01 4.33764160e-01
3.34238231e-01 -2.67378926e-01 -5.20102441e-01 3.10795099e-01] | [13.970719337463379, 3.0059893131256104] |
04826ea3-af28-410b-b1db-e4e7d83fd5f2 | puffin-a-path-unifying-feed-forward | 2307.02903 | null | https://arxiv.org/abs/2307.02903v2 | https://arxiv.org/pdf/2307.02903v2.pdf | PUFFIN: A Path-Unifying Feed-Forward Interfaced Network for Vapor Pressure Prediction | Accurately predicting vapor pressure is vital for various industrial and environmental applications. However, obtaining accurate measurements for all compounds of interest is not possible due to the resource and labor intensity of experiments. The demand for resources and labor further multiplies when a temperature-dependent relationship for predicting vapor pressure is desired. In this paper, we propose PUFFIN (Path-Unifying Feed-Forward Interfaced Network), a machine learning framework that combines transfer learning with a new inductive bias node inspired by domain knowledge (the Antoine equation) to improve vapor pressure prediction. By leveraging inductive bias and transfer learning using graph embeddings, PUFFIN outperforms alternative strategies that do not use inductive bias or that use generic descriptors of compounds. The framework's incorporation of domain-specific knowledge to overcome the limitation of poor data availability shows its potential for broader applications in chemical compound analysis, including the prediction of other physicochemical properties. Importantly, our proposed machine learning framework is partially interpretable, because the inductive Antoine node yields network-derived Antoine equation coefficients. It would then be possible to directly incorporate the obtained analytical expression in process design software for better prediction and control of processes occurring in industry and the environment. | ['Nadia Shardt', 'Idelfonso B. R. Nogueira', 'Ana Mafalda Ribeiro', 'Luana P. Queiroz', 'Carine Menezes Rebello', 'Vinicius Viena Santana'] | 2023-07-06 | null | null | null | null | ['transfer-learning'] | ['miscellaneous'] | [ 3.79277617e-01 -3.26221474e-02 -4.42930400e-01 -2.50229299e-01
2.06352949e-01 -7.08408654e-01 3.48932534e-01 8.17990303e-01
-6.68013468e-02 7.13600338e-01 -4.15185004e-01 -1.00638664e+00
-5.50395429e-01 -1.28861094e+00 -7.24722564e-01 -6.40438855e-01
-1.93875089e-01 2.04738304e-01 6.49477616e-02 -2.92679280e-01
3.74805957e-01 9.03595150e-01 -1.05024099e+00 1.17451221e-01
1.01100588e+00 1.13933945e+00 2.35775948e-01 4.48464215e-01
-2.07587853e-01 5.99571168e-01 -1.17928475e-01 4.56027985e-02
1.85006931e-01 -3.62805307e-01 -4.44474250e-01 -5.17545819e-01
1.35739699e-01 8.53555202e-02 6.30851695e-03 6.93043292e-01
1.52175397e-01 3.68267983e-01 1.11517644e+00 -1.24815333e+00
-7.71204650e-01 4.88508165e-01 -2.70182282e-01 -1.49229184e-01
2.15528429e-01 1.97792947e-01 1.02461565e+00 -7.20669150e-01
2.80324727e-01 8.28697741e-01 6.41606748e-01 1.95633128e-01
-1.41276908e+00 -7.69568682e-01 1.95435464e-01 2.70834655e-01
-1.29567313e+00 2.28371128e-01 7.91826427e-01 -6.80504501e-01
1.06819785e+00 2.44830355e-01 6.31092608e-01 8.31930995e-01
5.24967372e-01 1.74226806e-01 9.36874390e-01 -2.86083728e-01
5.75648487e-01 5.93731940e-01 -4.63111252e-02 4.67011839e-01
5.52699625e-01 3.56185466e-01 -2.96187580e-01 -1.29607311e-02
8.05102527e-01 2.47944877e-01 -7.68972561e-02 -6.65829480e-01
-5.43298185e-01 9.92046714e-01 8.87492537e-01 3.26806068e-01
-4.67213452e-01 -8.02423339e-03 1.90586567e-01 3.77829283e-01
3.94134462e-01 1.15778708e+00 -8.52263391e-01 1.89531609e-01
-8.85032415e-01 -1.29522029e-02 1.13858664e+00 8.72191370e-01
1.11777425e+00 2.57789761e-01 2.35541016e-01 3.87341231e-01
4.44589794e-01 2.84418792e-01 2.93856770e-01 -6.02547348e-01
1.17719330e-01 8.94802809e-01 6.11237176e-02 -1.09308672e+00
-5.14403045e-01 -4.11298096e-01 -7.00818658e-01 4.49228108e-01
3.32793921e-01 -1.99799076e-01 -7.58604348e-01 1.26887727e+00
2.63068408e-01 -2.47457381e-02 2.36463889e-01 6.80635214e-01
5.82203805e-01 7.82359064e-01 5.70591152e-01 -1.37172893e-01
1.04137385e+00 -8.59050691e-01 -4.77181405e-01 4.28995211e-03
7.09651887e-01 -4.53732163e-01 7.40267754e-01 4.44294989e-01
-4.92061615e-01 -6.14002466e-01 -1.16534293e+00 -2.07130741e-02
-1.19594026e+00 -2.02812135e-01 9.09509003e-01 6.34908557e-01
-5.62261045e-01 1.35001135e+00 -5.10741532e-01 -2.11680949e-01
2.19958499e-01 8.95663679e-01 -2.83742875e-01 1.29053548e-01
-1.42046404e+00 1.04905832e+00 5.56526423e-01 2.68257976e-01
-4.55827981e-01 -1.09389555e+00 -9.97545302e-01 1.62808731e-01
3.02714020e-01 -4.96193111e-01 9.06510174e-01 -6.40650690e-01
-1.76518440e+00 -1.89132690e-01 1.48314670e-01 -4.93542254e-01
3.85712624e-01 -1.86772689e-01 -4.61627811e-01 -3.62457596e-02
-4.32978481e-01 4.04487669e-01 6.38552725e-01 -1.03402090e+00
-2.59102970e-01 -2.85546601e-01 1.67487487e-02 -1.67966604e-01
-7.16560066e-01 -7.21602738e-01 3.88580948e-01 -2.04152092e-02
-2.29727522e-01 -7.38731027e-01 -4.17504132e-01 1.20638482e-01
-1.77930504e-01 -3.89892131e-01 1.08065689e+00 -4.79021877e-01
1.00612497e+00 -1.90359581e+00 -5.65656926e-03 7.22229660e-01
1.46817580e-01 5.26765764e-01 -1.13566555e-01 9.36848938e-01
-3.73352915e-01 2.81037211e-01 -2.70388991e-01 5.96159995e-01
-9.60754454e-02 -4.00589183e-02 4.35050279e-02 2.56387085e-01
6.77588999e-01 7.31108487e-01 -1.21045101e+00 -2.51488030e-01
8.13819110e-01 6.32616580e-01 -5.03771484e-01 2.31687889e-01
-3.94758999e-01 4.50998694e-01 -6.31307483e-01 4.13660437e-01
5.46912611e-01 -1.55475080e-01 3.75680238e-01 -6.46123171e-01
-3.50496739e-01 1.10259093e-02 -9.11540270e-01 1.36608934e+00
-9.49990153e-01 3.35118353e-01 -1.70465380e-01 -8.93760562e-01
1.37551713e+00 2.52458274e-01 6.20291114e-01 -4.48847771e-01
2.89544463e-01 2.55453318e-01 2.24177420e-01 -5.46308696e-01
3.14358085e-01 -3.15204501e-01 3.61959904e-01 2.19696742e-02
9.85205695e-02 -5.51434159e-01 1.54345199e-01 -1.76579192e-01
7.87732303e-01 5.01631618e-01 4.97060001e-01 -4.74920660e-01
9.29457068e-01 -1.35537326e-01 1.58996165e-01 1.73653826e-01
-6.92768320e-02 -1.77809268e-01 3.35565299e-01 -4.46870685e-01
-9.89740133e-01 -6.97611928e-01 -3.02211583e-01 9.86465812e-01
5.90868965e-02 -2.27754310e-01 -2.84761637e-01 -5.32006264e-01
5.33806801e-01 9.13209558e-01 -6.23090625e-01 -4.56700057e-01
-1.49279788e-01 -3.58370483e-01 1.26471341e-01 8.31445575e-01
-1.77675392e-02 -8.12640250e-01 -9.23352465e-02 4.52561826e-01
6.55159891e-01 -8.52952719e-01 4.61695604e-02 7.33711243e-01
-9.68212485e-01 -1.27612436e+00 -2.99044490e-01 -5.60499132e-01
4.52132612e-01 -2.90902201e-02 6.90300643e-01 -2.68014610e-01
-2.47144058e-01 7.28269070e-02 -2.03245848e-01 -6.73988342e-01
-5.94264746e-01 3.84538859e-01 -5.40194511e-02 -9.24450904e-02
6.34623945e-01 -7.45789707e-01 -7.62522101e-01 1.99797094e-01
-8.64094317e-01 -2.29099780e-01 6.75519407e-01 5.60470283e-01
3.60625327e-01 1.05161063e-01 8.38109434e-01 -1.07913506e+00
7.14073181e-01 -7.55021155e-01 -6.90073550e-01 1.36962235e-01
-1.19954443e+00 3.16914409e-01 1.00964570e+00 -4.30020511e-01
-9.38738763e-01 1.34084985e-01 4.46884446e-02 -2.86382824e-01
-1.37387112e-01 6.10901594e-01 -4.34591919e-02 -2.84618586e-01
7.23110020e-01 -1.04906060e-01 7.06417114e-02 -3.04540962e-01
3.82995337e-01 6.84855878e-01 7.59016648e-02 -5.13977230e-01
8.62349749e-01 -1.41634211e-01 6.35335565e-01 -1.04375851e+00
-2.02554390e-01 -5.07031977e-01 -7.14149117e-01 -1.76502392e-01
8.21856916e-01 -6.82534635e-01 -1.18646896e+00 -9.64728072e-02
-9.70440030e-01 -2.42625713e-01 -3.14386189e-01 6.61296964e-01
-3.44393373e-01 2.51955390e-01 -4.62638825e-01 -1.05205464e+00
-4.85173166e-01 -8.91046166e-01 5.77663660e-01 3.39608073e-01
-4.24583226e-01 -1.38153505e+00 -8.12201351e-02 -1.00095030e-02
5.62614202e-01 3.77096355e-01 1.20845234e+00 -8.43924403e-01
-4.44006950e-01 -2.87251413e-01 -2.11317882e-01 5.67032754e-01
7.39529014e-01 4.08091158e-01 -9.76263463e-01 -2.02737227e-01
-2.91064918e-01 -1.10103078e-01 7.40733147e-01 2.57659823e-01
1.09104812e+00 5.09404726e-02 -2.16782838e-01 3.17338437e-01
1.63674545e+00 6.04740024e-01 4.02835935e-01 1.96448043e-01
8.12823355e-01 7.89692342e-01 4.72075641e-01 2.93179601e-01
-5.71831614e-02 1.80934772e-01 6.33733392e-01 -3.74170989e-01
1.93427652e-01 -4.64748740e-01 2.45535657e-01 5.79954267e-01
-4.32201564e-01 -2.22178519e-01 -6.83985591e-01 2.67114490e-01
-1.52149141e+00 -5.96218765e-01 -3.07242393e-01 2.19313669e+00
8.07038844e-01 -5.96202202e-02 -2.82487087e-02 3.30949157e-01
4.68967170e-01 -1.03311941e-01 -8.45050931e-01 -1.12012589e+00
4.56499308e-01 4.53329921e-01 8.44125628e-01 5.43289006e-01
-9.13731277e-01 7.58885324e-01 5.94076252e+00 3.63618016e-01
-1.38405180e+00 -4.33253556e-01 3.39521140e-01 4.06919599e-01
-2.32540548e-01 -7.09881410e-02 -6.63316846e-01 6.21877946e-02
1.37693667e+00 -3.15490097e-01 3.94925505e-01 1.02048409e+00
4.14624780e-01 5.78177981e-02 -1.62469649e+00 3.59040707e-01
-3.13224286e-01 -1.09808099e+00 -4.13276441e-02 7.13043213e-02
4.81377721e-01 -3.88950378e-01 1.10907473e-01 1.69149056e-01
2.59571493e-01 -9.14738476e-01 6.31059036e-02 4.13841575e-01
5.58188796e-01 -6.97066247e-01 6.94732428e-01 2.31454577e-02
-1.22728145e+00 -2.60916770e-01 -3.74473691e-01 -1.93646431e-01
-2.56543189e-01 5.75378656e-01 -1.39184678e+00 1.00870025e+00
2.09309250e-01 8.28867853e-01 -3.10977101e-01 7.78417230e-01
-1.45490602e-01 4.67893571e-01 -3.26797694e-01 -4.91786420e-01
2.23367348e-01 -6.02386475e-01 2.04611067e-02 1.29455996e+00
4.09883946e-01 -3.06049079e-01 -7.78048811e-03 1.07208323e+00
5.82255609e-02 5.75849593e-01 -1.05997133e+00 -4.61369514e-01
3.96597415e-01 1.19630671e+00 -4.99621898e-01 -7.98574463e-02
-5.06282687e-01 6.40157580e-01 -2.99897138e-02 3.65441054e-01
-6.66047096e-01 -5.44058859e-01 6.42743826e-01 2.64315933e-01
4.52691287e-01 -3.46968859e-01 -1.67963669e-01 -4.87914294e-01
-5.29957533e-01 -4.49495971e-01 5.18601127e-02 -4.89464730e-01
-1.42875075e+00 3.98113251e-01 -2.69303452e-02 -1.14367187e+00
-2.47553661e-01 -1.26160085e+00 -4.54982817e-01 9.72831309e-01
-1.96206594e+00 -8.95555317e-01 -2.63779223e-01 3.85430813e-01
3.09693605e-01 3.34117748e-02 1.03511989e+00 1.41957507e-01
-6.13227308e-01 2.19623849e-01 2.29272902e-01 -2.58692145e-01
8.64262998e-01 -1.45765281e+00 -9.36720148e-02 1.95951626e-01
-2.59212762e-01 6.60900652e-01 7.57806778e-01 -5.16408265e-01
-1.55508971e+00 -1.44811559e+00 5.46930075e-01 -1.27057984e-01
1.02341735e+00 -2.08391771e-01 -9.22146201e-01 4.05952394e-01
1.14554301e-01 -9.66461673e-02 1.08717680e+00 1.48358628e-01
-2.65687943e-01 -4.97610599e-01 -1.13773727e+00 3.12002629e-01
4.61626053e-01 -4.15582031e-01 -3.90703529e-01 4.79270101e-01
7.47888505e-01 5.16600534e-02 -1.52884603e+00 2.79324979e-01
5.77526093e-01 -4.28601593e-01 8.75707805e-01 -7.69015193e-01
6.67375207e-01 -3.13775897e-01 6.46808892e-02 -1.35534990e+00
-5.20707190e-01 -2.49972731e-01 -2.55643606e-01 1.04031444e+00
9.21431243e-01 -8.21880758e-01 7.58532465e-01 1.00267637e+00
1.65511202e-02 -7.37681329e-01 -4.00387377e-01 -1.06554866e+00
4.13048059e-01 -1.23139732e-01 7.27601767e-01 8.72424185e-01
4.02113736e-01 4.67851520e-01 -2.23512426e-01 1.86742589e-01
1.13503054e-01 6.92638904e-02 5.01667559e-01 -1.55935037e+00
-1.72988251e-01 -1.60637185e-01 -4.35508907e-01 -6.85382843e-01
3.51666734e-02 -9.33558404e-01 -1.40959978e-01 -1.49144745e+00
-3.60056430e-01 -5.94309270e-01 -8.31267655e-01 4.48071748e-01
2.07022727e-01 -5.92425615e-02 1.78242058e-01 -2.13919103e-01
1.67648897e-01 5.31943679e-01 1.54138875e+00 -3.91297311e-01
-6.41858399e-01 -3.23150717e-02 -5.37661254e-01 4.48047459e-01
1.03476954e+00 -3.25511158e-01 -6.72400534e-01 7.16731772e-02
2.26462752e-01 -1.78959936e-01 6.08740263e-02 -9.83866274e-01
2.52191961e-01 -4.11754161e-01 6.23501003e-01 -1.70990989e-01
3.55437011e-01 -1.45457196e+00 2.05226928e-01 6.59427881e-01
-2.83817798e-01 -1.66702926e-01 3.62340540e-01 7.54161716e-01
-5.47061674e-02 -3.75353217e-01 3.14275652e-01 7.48814596e-03
-5.93594015e-01 4.09927249e-01 -4.59010631e-01 -8.46800029e-01
1.35914075e+00 -3.51957321e-01 -1.41170070e-01 -9.93640423e-02
-5.92309177e-01 1.58842757e-01 1.95336431e-01 4.14446503e-01
3.47383678e-01 -9.92242813e-01 -3.34528625e-01 3.08630258e-01
2.26402674e-02 -5.21523580e-02 3.58702466e-02 5.79057097e-01
-6.19067848e-01 6.12895370e-01 -1.22568280e-01 -3.59679610e-01
-8.65514040e-01 9.07794893e-01 3.49317551e-01 -3.32678348e-01
-1.28985077e-01 4.28281605e-01 5.52966073e-02 -3.75468731e-01
-1.86794579e-01 -6.78159416e-01 -2.27374449e-01 9.88066867e-02
1.73780635e-01 5.48755527e-01 3.20405930e-01 -1.68056905e-01
-3.17654103e-01 4.73464072e-01 6.45477185e-03 5.46671569e-01
1.62356925e+00 4.40498739e-01 2.16158833e-02 5.09017587e-01
1.18058264e+00 -2.30914280e-01 -1.15665257e+00 -2.60393135e-02
1.41613940e-02 -1.15446582e-01 2.86061198e-01 -8.96041214e-01
-9.89088118e-01 1.03596866e+00 5.74734390e-01 3.09666455e-01
9.84121501e-01 -5.75390995e-01 4.43232924e-01 6.88555479e-01
1.48057714e-01 -1.05293679e+00 -1.19930498e-01 1.90101564e-01
7.79363751e-01 -1.06691825e+00 3.37723613e-01 -9.28320587e-01
-2.98983961e-01 1.59691894e+00 5.94815493e-01 -5.69546642e-03
9.22090411e-01 1.85124964e-01 -9.49875787e-02 -2.64126249e-02
-4.36512589e-01 5.18432669e-02 1.29181087e-01 8.98883581e-01
7.00015128e-01 1.13386549e-01 -4.26792532e-01 3.01226497e-01
5.57534909e-03 2.22454682e-01 2.40022913e-01 9.44356799e-01
-3.42819154e-01 -1.56409502e+00 -1.02122813e-01 5.27793527e-01
-4.04805876e-02 -2.17609823e-01 -4.27850783e-01 8.36751044e-01
2.40566537e-01 1.13044107e+00 -1.18357301e-01 -5.12071550e-01
4.91233647e-01 1.77264661e-01 4.22421962e-01 -6.00931287e-01
-6.22528970e-01 -2.95658797e-01 -3.81479226e-02 -2.42125675e-01
-3.73421580e-01 -2.72561461e-01 -1.19661355e+00 -3.34791869e-01
-6.45556390e-01 3.56279194e-01 1.06744695e+00 7.26147234e-01
4.39364284e-01 7.80757010e-01 8.29795241e-01 -8.05149198e-01
-5.30454636e-01 -9.41039920e-01 -7.49419451e-01 1.01613589e-01
1.50838852e-01 -7.28028476e-01 -3.33232939e-01 8.16157535e-02] | [5.138703346252441, 5.664463520050049] |
4b4d39da-863c-4281-a927-d3ebdcce3e2f | green-cws-extreme-distillation-and-efficient-1 | 2111.09078 | null | https://arxiv.org/abs/2111.09078v1 | https://arxiv.org/pdf/2111.09078v1.pdf | Green CWS: Extreme Distillation and Efficient Decode Method Towards Industrial Application | Benefiting from the strong ability of the pre-trained model, the research on Chinese Word Segmentation (CWS) has made great progress in recent years. However, due to massive computation, large and complex models are incapable of empowering their ability for industrial use. On the other hand, for low-resource scenarios, the prevalent decode method, such as Conditional Random Field (CRF), fails to exploit the full information of the training data. This work proposes a fast and accurate CWS framework that incorporates a light-weighted model and an upgraded decode method (PCRF) towards industrially low-resource CWS scenarios. First, we distill a Transformer-based student model as an encoder, which not only accelerates the inference speed but also combines open knowledge and domain-specific knowledge. Second, the perplexity score to evaluate the language model is fused into the CRF module to better identify the word boundaries. Experiments show that our work obtains relatively high performance on multiple datasets with as low as 14\% of time consumption compared with the original BERT-based model. Moreover, under the low-resource setting, we get superior results in comparison with the traditional decoding methods. | ['Yong liu', 'Yulan Hu'] | 2021-11-17 | green-cws-extreme-distillation-and-efficient | https://openreview.net/forum?id=9poQ2m0R-- | https://openreview.net/pdf?id=9poQ2m0R-- | null | ['chinese-word-segmentation'] | ['natural-language-processing'] | [ 2.62341142e-01 -1.82692438e-01 -3.17942262e-01 -3.82642418e-01
-8.90550077e-01 -4.80292231e-01 2.63854623e-01 -3.92517522e-02
-6.32619262e-01 7.05668211e-01 -6.04991987e-02 -9.15174425e-01
4.67994839e-01 -7.91382194e-01 -4.80219901e-01 -5.07021904e-01
6.12052083e-01 3.57831597e-01 4.46482748e-01 -4.69304174e-02
1.11009210e-01 -7.41702840e-02 -1.09416819e+00 -1.31105363e-01
1.64376390e+00 8.24216664e-01 9.80986178e-01 3.31936717e-01
-5.35020530e-01 3.66756827e-01 -4.01757032e-01 -4.58344221e-01
2.30220389e-02 -3.20590734e-01 -6.76613748e-01 -2.73850352e-01
-1.33032262e-01 -3.51533622e-01 -1.56794623e-01 1.09848559e+00
4.84532773e-01 -1.19439520e-01 2.44468644e-01 -6.10527158e-01
-4.14814174e-01 8.66412163e-01 -7.07047343e-01 5.46857901e-02
2.45311677e-01 5.43498248e-02 7.92277455e-01 -6.41486049e-01
4.71802354e-01 1.00261569e+00 4.51078624e-01 4.01036501e-01
-7.60614455e-01 -8.75581622e-01 4.48537737e-01 3.79023664e-02
-1.20992219e+00 -3.29448044e-01 5.18913746e-01 -5.02475016e-02
9.91151989e-01 2.42175348e-02 4.62545633e-01 9.15449977e-01
5.32116443e-02 1.13692868e+00 1.08727539e+00 -6.76727295e-01
4.04273160e-02 2.90108602e-02 2.41035521e-01 7.21120238e-01
2.35346705e-01 -4.14000154e-01 -4.03093427e-01 2.79818267e-01
5.27611971e-01 -6.98406771e-02 -1.72764480e-01 4.23789263e-01
-1.10539222e+00 6.79768026e-01 -4.14378755e-02 6.17267907e-01
-2.20110178e-01 -1.73571497e-01 1.27935097e-01 -1.40070602e-01
6.65059566e-01 2.12929368e-01 -7.12776899e-01 -5.97434282e-01
-1.22635138e+00 -5.02757505e-02 8.66543591e-01 1.21437001e+00
6.84685707e-01 1.13658354e-01 -2.08474338e-01 7.76818216e-01
4.29610908e-01 6.50829315e-01 6.15624905e-01 -5.01585662e-01
8.96441936e-01 5.49981833e-01 -1.11773074e-01 -5.57892621e-01
-7.06491992e-02 -6.25295877e-01 -7.82021046e-01 -5.03068626e-01
4.15869623e-01 -4.11319047e-01 -1.12903559e+00 1.69317961e+00
2.75317043e-01 4.10652250e-01 -5.73989414e-02 6.13249302e-01
4.92896110e-01 9.28363860e-01 3.40105832e-01 -2.21186146e-01
1.51007783e+00 -1.03787291e+00 -9.46987271e-01 -5.36099374e-01
6.66728854e-01 -1.06099391e+00 9.64013040e-01 3.75951052e-01
-9.75364327e-01 -5.83271146e-01 -1.15915740e+00 -6.90650940e-02
-3.56282920e-01 4.20811415e-01 8.91773283e-01 9.14937735e-01
-6.38230741e-01 4.73942637e-01 -1.02754307e+00 -2.42700167e-02
2.74560601e-01 2.42246762e-01 6.30659834e-02 -4.99593914e-01
-1.43969226e+00 7.37746000e-01 6.34321690e-01 4.10134882e-01
-4.83218610e-01 -5.58885038e-01 -9.47068095e-01 1.21826520e-02
5.43771327e-01 -3.71529728e-01 1.27249467e+00 -6.69057250e-01
-1.78286374e+00 1.94454774e-01 -4.94195938e-01 -1.95542604e-01
3.65831375e-01 -5.19175112e-01 -5.23290873e-01 2.87855044e-02
9.42032039e-02 3.67411673e-01 5.44172287e-01 -9.09529030e-01
-8.12865734e-01 -2.19364941e-01 -5.84282055e-02 2.11539879e-01
-4.93035078e-01 2.01706052e-01 -1.02205813e+00 -5.77440202e-01
8.25559869e-02 -7.53492236e-01 -4.54378098e-01 -5.47038853e-01
-5.97838581e-01 -2.00506344e-01 5.95903158e-01 -1.17996573e+00
1.73499763e+00 -2.18878126e+00 -1.54516414e-01 1.32047415e-01
-7.65180364e-02 7.78483510e-01 3.42733483e-03 3.10930163e-01
2.76181847e-01 2.47225970e-01 -4.53847855e-01 -3.70470345e-01
-3.51324514e-03 3.61800671e-01 -1.11690141e-01 6.40714765e-02
4.29615200e-01 8.47629786e-01 -9.13223922e-01 -7.70745933e-01
3.48190777e-02 2.61668742e-01 -4.05182391e-01 3.28718752e-01
-2.91239828e-01 4.05582875e-01 -7.80647218e-01 5.48857629e-01
9.11871552e-01 -3.81743163e-02 4.31801170e-01 2.42240310e-01
-3.46545637e-01 4.55923438e-01 -1.21923435e+00 1.98698747e+00
-7.77132869e-01 7.75352716e-02 4.73038442e-02 -8.92796576e-01
9.48995590e-01 2.90463388e-01 5.32533675e-02 -8.52678478e-01
-4.47425582e-02 4.83299285e-01 -1.91951431e-02 -3.35639715e-01
6.42083466e-01 4.85062376e-02 -2.27584556e-01 3.03395480e-01
1.12929031e-01 6.70919567e-02 3.33837897e-01 3.69222999e-01
8.45477104e-01 6.39014423e-01 8.98122191e-02 -2.06747428e-01
4.46413726e-01 -5.50647937e-02 9.93518770e-01 3.88916910e-01
5.71709871e-02 3.63360047e-01 2.62173355e-01 1.53612450e-01
-7.20672846e-01 -9.04520214e-01 -1.54042825e-01 8.68154287e-01
2.88459897e-01 -4.66294348e-01 -1.09520805e+00 -6.48136318e-01
-3.38182390e-01 8.91764641e-01 1.31188691e-01 1.77867580e-02
-8.24814796e-01 -9.77094471e-01 7.45646894e-01 7.11791575e-01
7.76168525e-01 -7.98380375e-01 -2.23424852e-01 4.70725805e-01
-4.53787237e-01 -1.27143013e+00 -5.15932679e-01 2.12881550e-01
-8.81364763e-01 -6.73843384e-01 -7.68163443e-01 -1.05825710e+00
5.70429742e-01 1.71631157e-01 8.74809444e-01 2.11017467e-02
-5.48130227e-03 -3.39192063e-01 -6.77989006e-01 -3.93137068e-01
-3.38303685e-01 5.68910539e-01 -3.03096980e-01 -2.63085663e-01
4.88931388e-01 -3.14454317e-01 -5.07506967e-01 1.13871738e-01
-7.13073254e-01 4.04712081e-01 9.62444603e-01 7.24329233e-01
3.79229009e-01 1.64195567e-01 7.26430714e-01 -1.08512104e+00
4.75157261e-01 -4.53866780e-01 -6.05766177e-01 4.89846706e-01
-9.68749762e-01 1.32614791e-01 7.61819184e-01 -3.81080061e-01
-1.75059664e+00 -1.57108605e-02 -6.38507903e-01 4.22766298e-01
-6.53964281e-02 7.19309688e-01 -3.90882313e-01 4.97449368e-01
-3.89172174e-02 3.48900139e-01 -3.14284027e-01 -9.57630336e-01
3.25929999e-01 1.10871959e+00 4.65018809e-01 -7.39681780e-01
8.08012128e-01 -3.05225980e-02 -6.01778448e-01 -6.66087627e-01
-8.64840746e-01 -5.23636520e-01 -6.55462384e-01 2.66166311e-02
9.81881559e-01 -1.13319659e+00 -5.15165746e-01 7.27011681e-01
-1.25698841e+00 -2.64961928e-01 1.13581039e-01 7.91563809e-01
1.34978816e-02 7.48518944e-01 -1.02154720e+00 -9.79687691e-01
-4.46672618e-01 -1.14098835e+00 1.00824130e+00 6.01989150e-01
1.64467871e-01 -7.40192473e-01 -2.89378405e-01 5.71177840e-01
3.58973056e-01 -1.31652057e-01 1.02909148e+00 -5.99407911e-01
-6.68595970e-01 -1.43809363e-01 -3.74885529e-01 4.48729366e-01
-1.97713926e-01 -1.69794932e-01 -8.65574598e-01 -5.91303781e-02
-1.45489901e-01 -8.69043469e-02 8.36169720e-01 2.92830970e-02
1.18407631e+00 1.16573520e-01 -3.11316013e-01 3.88684601e-01
1.43091452e+00 4.06573474e-01 6.52243197e-01 5.33542000e-02
7.78445542e-01 2.69342840e-01 8.39076400e-01 2.87122995e-01
6.05983853e-01 4.11462039e-01 1.00836441e-01 -2.87871122e-01
-9.86422002e-02 -5.73808789e-01 4.28859472e-01 1.73323894e+00
-1.28165826e-01 -3.83706540e-01 -9.28439558e-01 5.00592470e-01
-1.60798812e+00 -4.70739156e-01 -3.58128011e-01 2.10486746e+00
1.29110634e+00 3.60231459e-01 -3.98634613e-01 4.92087565e-02
8.06129694e-01 1.81053385e-01 -4.93906587e-01 -4.39297497e-01
9.10420939e-02 5.40868878e-01 6.41514957e-01 3.27519804e-01
-8.32341492e-01 1.28396618e+00 5.62102938e+00 1.36462796e+00
-7.98211098e-01 2.61898369e-01 5.82172573e-01 5.14605880e-01
-6.13401175e-01 1.88907966e-01 -1.24751854e+00 7.77575433e-01
1.15147376e+00 9.74313542e-02 3.65278631e-01 7.90355504e-01
1.17782548e-01 -3.85482550e-01 -5.28633773e-01 5.10468125e-01
-1.88818082e-01 -8.87917399e-01 -3.23307276e-01 1.59212008e-01
5.63336372e-01 9.50107258e-03 -1.54494986e-01 5.42816401e-01
5.04580140e-01 -7.62838006e-01 7.66808987e-01 3.58025134e-01
1.05812526e+00 -7.52061546e-01 1.02697766e+00 6.60385489e-01
-1.47294629e+00 1.48581535e-01 -4.36286122e-01 -5.53743578e-02
4.68911946e-01 9.78770912e-01 -6.86756015e-01 1.05019104e+00
2.95339674e-01 4.86703396e-01 -3.57531607e-01 7.61741996e-01
-5.63959360e-01 1.23550785e+00 -3.44325870e-01 -1.25504717e-01
3.98831874e-01 -4.31962997e-01 -1.60949305e-01 1.50240767e+00
4.48441565e-01 1.16150379e-01 3.88837755e-01 6.02057874e-01
-1.64802223e-02 3.04368705e-01 -5.68032712e-02 -2.36859664e-01
5.18999934e-01 1.21846581e+00 -7.94377863e-01 -3.68648231e-01
-6.17953241e-01 9.76766348e-01 2.34135792e-01 2.55370498e-01
-9.05751944e-01 -5.94415963e-01 3.38985145e-01 -1.83313042e-01
4.70780611e-01 -5.46418548e-01 -3.07635367e-01 -1.26793444e+00
1.22967169e-01 -7.42846966e-01 1.17148258e-01 -4.34811532e-01
-9.82665718e-01 5.21209478e-01 -2.01534092e-01 -9.17115390e-01
-3.18689495e-01 -4.71381217e-01 -5.78105867e-01 1.35964727e+00
-2.08529329e+00 -1.25481641e+00 -4.15492943e-03 3.05436790e-01
6.63115680e-01 1.48019612e-01 9.17230129e-01 6.44372582e-01
-1.04864454e+00 6.60969853e-01 2.70689964e-01 3.66596073e-01
6.65610909e-01 -1.22172379e+00 5.96527576e-01 1.23921514e+00
1.80141181e-02 8.17069173e-01 3.44099522e-01 -8.53423953e-01
-1.46897590e+00 -9.20817912e-01 1.25609696e+00 8.85006338e-02
5.05924463e-01 -4.87376839e-01 -7.74407625e-01 3.92569274e-01
1.93828478e-01 -3.34286004e-01 7.17595577e-01 2.37842441e-01
-1.95027992e-01 -9.27726552e-02 -7.58227706e-01 4.83832896e-01
9.76615191e-01 -3.60136092e-01 -5.77688992e-01 6.79959059e-02
9.95448411e-01 -4.75216508e-01 -8.76082540e-01 2.04189539e-01
5.28496027e-01 -4.93190974e-01 5.08488953e-01 -2.62201369e-01
4.11140680e-01 -2.23922983e-01 -7.80126005e-02 -1.08779895e+00
-8.09496082e-03 -6.48448646e-01 1.17968430e-03 1.68419504e+00
7.52454937e-01 -6.68880403e-01 5.58686972e-01 5.64671457e-01
-3.72376502e-01 -7.93157697e-01 -7.35072732e-01 -7.45851099e-01
5.21534979e-02 -6.36328399e-01 7.12805510e-01 6.73398614e-01
-2.68472973e-02 4.13958609e-01 -2.49943912e-01 2.93490179e-02
2.98108190e-01 4.08233732e-01 4.37406749e-01 -1.10471559e+00
-4.44853365e-01 7.11852079e-03 1.67633846e-01 -1.73704565e+00
1.32067978e-01 -8.24417174e-01 4.38337266e-01 -1.61457205e+00
2.14097738e-01 -8.81236374e-01 -2.64411569e-01 3.99794787e-01
-5.62190890e-01 2.21290346e-02 2.50904441e-01 -9.22988653e-02
-5.56943476e-01 3.86722296e-01 1.32710946e+00 1.68073729e-01
5.93687035e-02 -4.18471778e-03 -8.68150115e-01 7.64384389e-01
8.50382745e-01 -4.50412452e-01 -1.76145658e-01 -6.96813822e-01
5.19220829e-02 2.03564048e-01 -3.42015386e-01 -7.33846843e-01
2.79275477e-01 -8.76899511e-02 2.58725315e-01 -5.75499892e-01
1.23443231e-02 -6.17897511e-01 -3.13148648e-01 4.32996005e-01
9.03881863e-02 -1.79108884e-02 -3.22268680e-02 6.47020519e-01
-2.70363808e-01 -5.49659312e-01 4.08010244e-01 -1.79409906e-01
-8.12402368e-01 2.74978876e-01 -3.74241978e-01 1.99085847e-01
7.72186935e-01 -1.25460461e-01 -9.63213593e-02 -7.13228956e-02
-2.09372088e-01 2.82783389e-01 1.29537046e-01 2.80943751e-01
2.90439039e-01 -8.72787535e-01 -5.32313466e-01 2.74628013e-01
-1.32437482e-01 3.40317070e-01 3.24694127e-01 7.17890203e-01
-5.41443229e-01 4.51244324e-01 2.32746929e-01 -3.12970787e-01
-8.11389089e-01 3.18327188e-01 -3.83354306e-01 -7.50735223e-01
-4.24394608e-01 6.42780364e-01 -1.47700652e-01 -2.83349633e-01
8.25690106e-02 -3.69362921e-01 -1.95456669e-01 -5.00841178e-02
4.64550138e-01 2.32687995e-01 9.17176679e-02 -4.35473800e-01
-2.63076097e-01 4.03848678e-01 -1.25227928e-01 -8.00342634e-02
1.32381356e+00 -1.72904551e-01 -1.18978703e-02 2.87074178e-01
9.10004556e-01 1.56774029e-01 -1.22956538e+00 -2.92506754e-01
1.93198815e-01 -2.24006072e-01 2.81075537e-01 -8.87868166e-01
-9.85037327e-01 1.17410815e+00 1.96469292e-01 -1.59865126e-01
1.16539848e+00 -3.45471263e-01 1.51324964e+00 1.44164413e-01
4.48129416e-01 -1.53648221e+00 -3.79327834e-01 5.54693758e-01
-1.54467717e-01 -1.19165421e+00 -1.30569696e-01 -7.96452463e-01
-7.01050699e-01 1.13039339e+00 4.37466264e-01 2.00974107e-01
5.74067175e-01 6.56460702e-01 8.54781270e-03 3.51001829e-01
-3.66649896e-01 -4.87727821e-01 5.88456988e-02 4.32755142e-01
5.38980067e-01 1.77764013e-01 -6.97882533e-01 9.78009999e-01
-4.92388662e-03 3.81164253e-02 2.07003206e-01 8.51027489e-01
-5.06428778e-01 -1.47967350e+00 4.26791646e-02 3.91488671e-01
-7.39333034e-01 -4.93590385e-01 2.78813004e-01 5.22097051e-01
2.90130258e-01 1.21078122e+00 -1.12350903e-01 -3.29847842e-01
1.00535423e-01 2.65487343e-01 1.30727068e-01 -6.75804079e-01
-5.14159739e-01 2.32982785e-01 2.12740645e-01 -4.42655772e-01
-3.76757532e-01 -6.01925850e-01 -1.37559855e+00 -3.07583213e-01
-8.22151899e-01 3.46305847e-01 9.61432457e-01 1.24112582e+00
1.98779240e-01 6.11799359e-01 5.00548065e-01 -3.02837700e-01
-6.57469690e-01 -1.02634466e+00 -6.11717999e-01 -6.02338649e-02
-3.56696784e-01 -3.74581873e-01 1.00794256e-01 1.06629301e-02] | [9.989754676818848, 10.0977144241333] |
41be110a-cb89-43d3-91d0-54cd2d0cd589 | atlas-automate-online-service-configuration | 2210.16902 | null | https://arxiv.org/abs/2210.16902v1 | https://arxiv.org/pdf/2210.16902v1.pdf | Atlas: Automate Online Service Configuration in Network Slicing | Network slicing achieves cost-efficient slice customization to support heterogeneous applications and services. Configuring cross-domain resources to end-to-end slices based on service-level agreements, however, is challenging, due to the complicated underlying correlations and the simulation-to-reality discrepancy between simulators and real networks. In this paper, we propose Atlas, an online network slicing system, which automates the service configuration of slices via safe and sample-efficient learn-to-configure approaches in three interrelated stages. First, we design a learning-based simulator to reduce the sim-to-real discrepancy, which is accomplished by a new parameter searching method based on Bayesian optimization. Second, we offline train the policy in the augmented simulator via a novel offline algorithm with a Bayesian neural network and parallel Thompson sampling. Third, we online learn the policy in real networks with a novel online algorithm with safe exploration and Gaussian process regression. We implement Atlas on an end-to-end network prototype based on OpenAirInterface RAN, OpenDayLight SDN transport, OpenAir-CN core network, and Docker-based edge server. Experimental results show that, compared to state-of-the-art solutions, Atlas achieves 63.9% and 85.7% regret reduction on resource usage and slice quality of experience during the online learning stage, respectively. | ['Tao Han', 'Nakjung Choi', 'Qiang Liu'] | 2022-10-30 | null | null | null | null | ['thompson-sampling', 'safe-exploration'] | ['methodology', 'robots'] | [-5.56502879e-01 -1.70166761e-01 -2.09510028e-01 -6.13876939e-01
-7.75961876e-01 -7.57478237e-01 -3.42621952e-01 -8.58892381e-01
-1.05212897e-01 1.12425625e+00 -2.42341235e-01 -8.85677218e-01
-4.30319965e-01 -5.43116927e-01 -4.87664670e-01 -5.85093081e-01
-5.38353801e-01 1.05272746e+00 1.86701730e-01 3.05950195e-01
-2.13541985e-01 4.73913372e-01 -7.65078723e-01 -2.21630380e-01
8.60729575e-01 1.50673163e+00 1.81742653e-01 3.98883283e-01
-1.51043981e-01 5.65588653e-01 -3.02147418e-01 -3.63572210e-01
7.69479990e-01 -3.69382501e-01 -4.68424171e-01 2.26601735e-01
-4.52616900e-01 -4.23082680e-01 -2.48618975e-01 9.67434824e-01
7.18828440e-01 9.91625478e-04 -1.85045317e-01 -1.58490753e+00
2.71593392e-01 1.05623043e+00 -4.35923100e-01 2.32041344e-01
-1.75988615e-01 6.67442918e-01 1.00846100e+00 -4.02014792e-01
4.95851904e-01 1.05201423e+00 4.34869528e-01 1.26140624e-01
-1.52473509e+00 -1.02571356e+00 5.34320474e-01 2.71004826e-01
-1.56690705e+00 -4.97838050e-01 3.39171231e-01 -1.25721574e-01
6.17915273e-01 1.64734945e-01 1.02312315e+00 8.92501175e-01
2.88264416e-02 3.22602510e-01 9.72421706e-01 1.70900505e-02
7.84953356e-01 2.91701476e-03 -4.42412555e-01 4.23737973e-01
-7.11593404e-02 4.93944108e-01 -2.42605448e-01 -4.15703773e-01
9.85000372e-01 -2.04940569e-02 -2.74824709e-01 -5.37264287e-01
-9.62891281e-01 3.73847246e-01 1.79459855e-01 -5.13873458e-01
-8.73107970e-01 2.59311050e-01 4.17107284e-01 3.43372673e-01
3.61778617e-01 1.13064654e-01 -8.55728984e-01 -6.61645234e-01
-1.41798282e+00 7.45887905e-02 1.40691316e+00 1.32268667e+00
3.90943617e-01 2.80755639e-01 -2.71837533e-01 3.75931889e-01
5.60316563e-01 6.76080763e-01 -2.99914241e-01 -1.43223810e+00
4.54803735e-01 -2.65270382e-01 3.18828732e-01 -1.25701621e-01
-4.14605498e-01 -1.18356597e+00 -7.85400271e-01 2.00893685e-01
2.15586752e-01 -9.97687101e-01 -4.43491191e-01 1.67646158e+00
4.82941121e-01 9.16825593e-01 -3.23185772e-01 1.35664940e+00
-2.62436986e-01 5.67301393e-01 -1.86163262e-01 -5.00414431e-01
1.12429583e+00 -1.34244370e+00 -6.62401199e-01 8.58492479e-02
2.15090901e-01 -6.92292809e-01 7.22069621e-01 6.34514689e-01
-1.33719158e+00 -3.20795923e-02 -8.55795562e-01 9.63392913e-01
4.04923439e-01 -1.70417339e-01 5.49101830e-01 8.96976769e-01
-9.37465608e-01 6.63342059e-01 -7.71339715e-01 -4.56260033e-02
5.62334716e-01 3.83770138e-01 3.85825157e-01 -3.63454074e-02
-9.23056841e-01 3.57208371e-01 2.80865639e-01 1.89886749e-01
-1.21366048e+00 -1.48534477e+00 -1.14466660e-01 6.57955408e-01
1.19262207e+00 -9.61906493e-01 1.45202768e+00 -8.33507895e-01
-2.35236263e+00 -1.96448043e-02 2.47132570e-01 -5.77616930e-01
7.59626150e-01 -3.36267464e-02 -8.38230133e-01 7.94034675e-02
-2.70514339e-01 -4.86568622e-02 5.22213221e-01 -1.11606526e+00
-6.92142606e-01 -6.48247823e-02 1.30435184e-01 -6.53613731e-02
4.00498331e-01 2.81512439e-01 -5.96380293e-01 -2.71815062e-01
1.23428807e-01 -1.01553917e+00 -7.41984010e-01 1.18742168e-01
-3.05521846e-01 4.19092655e-01 5.16694486e-01 -4.43332881e-01
1.11202097e+00 -1.84344232e+00 -2.31650978e-01 9.18274939e-01
2.91539878e-01 -1.19180456e-02 8.27917531e-02 3.27849329e-01
2.43036389e-01 2.34602571e-01 3.52316320e-01 -2.58390784e-01
2.08419055e-01 1.94451153e-01 -2.19730705e-01 2.66822755e-01
-4.53509629e-01 3.92747968e-01 -8.49860907e-01 -4.24606889e-01
7.27659240e-02 9.02100950e-02 -8.77180874e-01 3.60547513e-01
-4.02915895e-01 5.85281491e-01 -3.88383299e-01 7.73949921e-01
9.98824835e-01 -4.55386609e-01 6.90788865e-01 -2.61282384e-01
-7.45783001e-02 1.49201185e-01 -1.44388914e+00 1.75049472e+00
-1.01367807e+00 -3.63612548e-02 7.18755901e-01 -6.42906249e-01
6.23478472e-01 4.71285135e-01 5.60246766e-01 -5.15977085e-01
1.33257911e-01 2.21980050e-01 7.35463053e-02 -2.78111935e-01
-1.96184695e-01 -4.19867516e-01 1.77449331e-01 7.00833201e-01
1.90133438e-01 1.31404579e-01 -1.44068941e-01 -6.04813024e-02
1.14423609e+00 2.23881245e-01 -1.16205372e-01 -2.49136597e-01
2.64572710e-01 -5.49954712e-01 1.40140259e+00 6.75903916e-01
-5.69425941e-01 1.44941881e-01 1.01709580e+00 -3.08105201e-01
-8.05391073e-01 -1.75911880e+00 1.05415411e-01 7.70646513e-01
2.23055016e-02 -1.90143645e-01 -7.15497494e-01 -6.25946820e-01
-1.92137703e-01 1.04954648e+00 8.00409243e-02 1.37317136e-01
-1.47304675e-02 -5.43855786e-01 2.41203412e-01 2.63558000e-01
4.82569396e-01 -5.31821609e-01 -1.84091046e-01 6.06455445e-01
8.51339251e-02 -1.57847691e+00 -6.79881275e-01 1.48016930e-01
-9.38186765e-01 -5.90773582e-01 -6.92386627e-02 6.64394796e-02
3.14290226e-01 1.38980970e-01 1.32795501e+00 -4.06033754e-01
-1.48911223e-01 7.05307126e-02 3.11874263e-02 2.27526560e-01
3.31232585e-02 1.62911326e-01 1.89948037e-01 2.09374726e-01
-1.39640331e-01 -1.30337095e+00 -7.63820946e-01 7.88975060e-01
-3.47602606e-01 4.03939076e-02 6.35980427e-01 7.56706715e-01
8.48416805e-01 -1.92408469e-02 6.36427939e-01 -9.11697268e-01
2.53757328e-01 -7.34345078e-01 -1.36465836e+00 2.71986037e-01
-8.84390950e-01 -7.00343074e-03 8.62931907e-01 -1.24011725e-01
-1.18545485e+00 -1.91611141e-01 9.98137295e-02 -8.99967790e-01
3.69624227e-01 4.87376958e-01 -6.19182587e-01 -8.34449604e-02
2.93552995e-01 -1.18454091e-01 1.09468892e-01 -2.58967310e-01
2.49714434e-01 5.81535935e-01 6.41973838e-02 -9.72626865e-01
9.81209397e-01 4.12329972e-01 8.29660073e-02 -2.15037889e-03
-1.03828788e+00 -2.85975993e-01 -5.07561900e-02 -4.73365068e-01
3.03291321e-01 -1.11642468e+00 -1.24144065e+00 -9.55924541e-02
-6.93923473e-01 -6.61941886e-01 -3.52703989e-01 8.11986685e-01
-9.45161700e-01 -4.99485843e-02 -6.14780188e-01 -7.09832430e-01
-5.39332032e-01 -1.22712958e+00 3.50105673e-01 4.22026277e-01
4.29044574e-01 -7.10190892e-01 -1.26730695e-01 3.74878287e-01
9.14792180e-01 -9.73554403e-02 4.06881243e-01 -4.71442997e-01
-1.34045470e+00 1.55888259e-01 -5.65626621e-01 5.09975672e-01
-3.89133245e-01 1.58976510e-01 -5.69451630e-01 -5.48737824e-01
-9.23832282e-02 -1.38634201e-02 -4.07846309e-02 6.18725181e-01
1.73067462e+00 -3.71394485e-01 -1.94058165e-01 1.36115575e+00
1.78438473e+00 1.21699341e-01 5.04696131e-01 8.34669098e-02
2.51937062e-01 5.03303707e-02 7.87865877e-01 1.31358051e+00
4.07025069e-01 6.26630008e-01 8.37027431e-01 1.67455330e-01
4.23772871e-01 -2.13778228e-01 3.32109481e-01 6.85646474e-01
4.85361964e-02 -1.36815920e-01 -6.04773402e-01 3.48731615e-02
-1.89932561e+00 -9.64365721e-01 2.08801910e-01 2.46581006e+00
5.08012891e-01 7.32223630e-01 1.74296364e-01 -5.81148386e-01
5.27841151e-01 -1.47839338e-01 -7.72650599e-01 -1.37161925e-01
3.45761478e-01 8.36736038e-02 9.90758538e-01 3.34968686e-01
-3.79146010e-01 9.27647829e-01 5.10250378e+00 1.00969160e+00
-1.04026544e+00 4.21071231e-01 7.23107457e-01 -6.27249360e-01
-1.15563385e-01 5.57218492e-01 -5.71809530e-01 9.24341917e-01
1.78041673e+00 -5.70582807e-01 1.25038338e+00 1.08762598e+00
1.14873624e+00 1.21292127e-02 -9.61539745e-01 9.64307666e-01
-6.39998794e-01 -1.56959283e+00 -5.55445313e-01 1.51763380e-01
7.13879585e-01 4.25631672e-01 -3.78880471e-01 5.43640375e-01
7.27044642e-01 -6.60601318e-01 7.68813670e-01 7.68743038e-01
7.12783456e-01 -1.23749876e+00 1.01066148e+00 1.81746572e-01
-9.91157055e-01 -2.53564924e-01 -1.29093066e-01 1.91824004e-01
1.03923917e+00 1.24836385e+00 -7.36288488e-01 5.24533629e-01
7.05196202e-01 2.09635735e-01 3.91079277e-01 1.61062622e+00
-9.74980891e-02 9.28190291e-01 -4.04000401e-01 3.22213531e-01
1.29787028e-01 -6.90422893e-01 6.01225674e-01 7.39532053e-01
3.67949754e-01 1.44580156e-02 8.31290543e-01 1.35447037e+00
-1.59710541e-01 -2.90142685e-01 3.29341918e-01 2.84034044e-01
9.84333992e-01 1.74670184e+00 -5.68279028e-01 -2.79961705e-01
-2.81781882e-01 7.01274276e-01 -2.52600014e-01 7.47734487e-01
-1.42431414e+00 -2.96458215e-01 1.35254335e+00 2.28281155e-01
4.99805421e-01 -2.97055513e-01 -3.37521732e-01 -9.50114250e-01
6.30684197e-02 -7.94236481e-01 -5.68921082e-02 -6.75990105e-01
-1.27404451e+00 5.55322051e-01 -4.01309878e-01 -1.36471856e+00
-8.82566795e-02 -1.40228018e-01 -7.36655414e-01 1.07691789e+00
-1.66927850e+00 -7.56215811e-01 -2.24719658e-01 4.11160201e-01
2.98342526e-01 -2.50241578e-01 6.41735017e-01 6.01088405e-01
-8.28770459e-01 5.73586285e-01 3.32119167e-01 -2.53507018e-01
6.18991554e-01 -1.02033961e+00 2.55322993e-01 7.01949477e-01
-4.09249961e-01 1.05239265e-01 7.60126352e-01 -2.30018780e-01
-1.36594093e+00 -1.09688997e+00 3.71264108e-02 6.41937912e-01
1.03827155e+00 -4.80978131e-01 -3.66446227e-01 6.27669036e-01
8.40560645e-02 6.94070101e-01 9.30460691e-01 2.91443318e-01
-1.25565082e-01 -6.65079832e-01 -1.33221829e+00 7.11860120e-01
1.24372160e+00 -3.12332511e-01 6.03818178e-01 5.01881063e-01
8.83993864e-01 -6.09629691e-01 -1.01383126e+00 2.38432232e-02
3.41161311e-01 -1.00362372e+00 6.59220219e-01 -7.26754725e-01
-3.75272036e-01 -4.42562461e-01 -2.26631865e-01 -1.57123375e+00
-2.88474828e-01 -1.42299271e+00 -3.79242003e-01 1.27209079e+00
4.58466023e-01 -7.43790925e-01 8.71661425e-01 4.36386347e-01
-1.71536326e-01 -1.00473380e+00 -1.21360993e+00 -1.16839981e+00
-6.10793650e-01 -4.71941650e-01 1.10963070e+00 4.37906712e-01
-3.25310677e-01 1.37088686e-01 -1.23791352e-01 6.31685376e-01
1.06567430e+00 3.18825334e-01 6.01388037e-01 -1.06604719e+00
-1.01155949e+00 -3.75203788e-01 -1.11047260e-01 -1.10484397e+00
3.41132849e-01 -5.00410199e-01 -1.63258873e-02 -9.54074383e-01
5.47242053e-02 -8.97770941e-01 -3.64554286e-01 -1.38938874e-01
3.05375159e-01 -6.98206067e-01 3.91322911e-01 -9.34842080e-02
-1.14240396e+00 8.21382463e-01 1.15765333e+00 4.74777877e-01
-7.69984126e-02 8.16133440e-01 -4.49055165e-01 4.01207477e-01
1.01687598e+00 -6.45991623e-01 -5.81028402e-01 -2.52229124e-01
1.30066648e-01 9.31296587e-01 3.25168520e-01 -1.07402897e+00
2.28886962e-01 -4.92959857e-01 -1.85424417e-01 -4.87579167e-01
8.39327499e-02 -1.28318202e+00 4.05343324e-01 2.23921388e-01
1.65295601e-01 9.27324519e-02 -5.47416769e-02 9.26012158e-01
1.80967227e-01 -4.82439902e-03 8.01398337e-01 -8.45599547e-03
-4.50142413e-01 9.03521061e-01 -4.66049761e-01 4.64381099e-01
1.05122375e+00 2.02703908e-01 7.19573349e-02 -7.12210536e-01
-1.01627350e+00 9.39515829e-01 1.48207247e-01 -7.27608427e-02
1.91572323e-01 -1.08944535e+00 -5.25588930e-01 -8.79340470e-02
-4.59466577e-01 -2.54771322e-01 6.71908855e-01 1.04466534e+00
-5.62386930e-01 7.82805011e-02 -2.08270058e-01 -6.32564485e-01
-6.81848884e-01 3.56232017e-01 7.71359742e-01 -7.85090625e-01
-2.44796619e-01 5.61168671e-01 -4.68470991e-01 -5.52463591e-01
3.71506363e-01 -1.38469845e-01 9.72625732e-01 -3.55703354e-01
2.72581756e-01 6.45454764e-01 7.99667016e-02 4.93197024e-01
-1.06625587e-01 -1.89730376e-01 1.26926005e-01 -4.04819131e-01
1.35217893e+00 -5.83615541e-01 -1.42065361e-01 2.40038902e-01
7.38342643e-01 -1.99164808e-01 -1.75670218e+00 -4.88804162e-01
-7.11588189e-02 -7.13819385e-01 6.03869140e-01 -1.11389923e+00
-1.64158726e+00 4.68243062e-01 6.84550703e-01 -1.12216197e-01
1.15432060e+00 -4.64491695e-01 9.14719045e-01 6.49006143e-02
8.93854439e-01 -1.24911404e+00 -5.09316027e-01 3.71854007e-01
2.29220524e-01 -1.06912768e+00 -1.64738432e-01 -4.12719995e-01
-5.19161224e-01 1.09107208e+00 7.75086105e-01 -2.24083774e-02
1.31212938e+00 5.97527504e-01 1.78070832e-02 1.09269716e-01
-1.31571543e+00 7.05889193e-03 -5.92375219e-01 2.43010476e-01
-1.49523959e-01 6.13444686e-01 1.05262533e-01 9.63430464e-01
9.13576707e-02 4.36143667e-01 6.52032852e-01 4.35881943e-01
-1.35788277e-01 -1.06963539e+00 -1.97705463e-01 4.79827225e-01
-2.97529578e-01 -1.76225543e-01 8.46125484e-01 3.09890747e-01
-2.16251701e-01 6.89829946e-01 1.15229674e-01 -2.78101325e-01
1.63561851e-01 -1.73619285e-01 7.44171143e-02 -3.38258415e-01
-5.90074778e-01 1.45989627e-01 3.51782650e-01 -1.18286312e+00
4.08385575e-01 -6.70199513e-01 -1.18207335e+00 -8.35713685e-01
-2.81124145e-01 2.05944240e-01 8.38585198e-01 8.90521348e-01
7.72054493e-01 1.00766885e+00 1.22708213e+00 -8.45932901e-01
-1.24015892e+00 -6.03233695e-01 -1.08213866e+00 -5.76267540e-01
-1.81585789e-01 -5.80843389e-01 -4.18364018e-01 -5.65584838e-01] | [5.854942321777344, 1.7323272228240967] |
d33ccded-849a-44e3-8e07-c2b0278340bb | decentralized-learning-for-wireless | 1503.08855 | null | http://arxiv.org/abs/1503.08855v1 | http://arxiv.org/pdf/1503.08855v1.pdf | Decentralized learning for wireless communications and networking | This chapter deals with decentralized learning algorithms for in-network
processing of graph-valued data. A generic learning problem is formulated and
recast into a separable form, which is iteratively minimized using the
alternating-direction method of multipliers (ADMM) so as to gain the desired
degree of parallelization. Without exchanging elements from the distributed
training sets and keeping inter-node communications at affordable levels, the
local (per-node) learners consent to the desired quantity inferred globally,
meaning the one obtained if the entire training data set were centrally
available. Impact of the decentralized learning framework to contemporary
wireless communications and networking tasks is illustrated through case
studies including target tracking using wireless sensor networks, unveiling
Internet traffic anomalies, power system state estimation, as well as spectrum
cartography for wireless cognitive radio networks. | ['Ioannis D. Schizas', 'Qing Ling', 'Gonzalo Mateos', 'Georgios B. Giannakis', 'Hao Zhu'] | 2015-03-30 | null | null | null | null | ['spectrum-cartography'] | ['computer-vision'] | [ 4.75823402e-01 6.07189536e-01 -6.93929017e-01 -3.25675070e-01
-4.57753837e-01 -3.60377103e-01 1.74368739e-01 1.84724525e-01
-5.64244807e-01 1.09520912e+00 -3.19575340e-01 -6.42320454e-01
-7.51686275e-01 -8.68217230e-01 -4.81707484e-01 -8.90930057e-01
-1.19214153e+00 4.63895798e-01 -4.66049016e-01 1.85388595e-01
-1.42775714e-01 5.91900349e-01 -6.76167548e-01 -2.08191782e-01
2.83153802e-01 1.29024637e+00 1.19284637e-01 8.26476097e-01
1.43257841e-01 1.04079759e+00 -5.13917506e-01 -2.10426360e-01
6.87695682e-01 -3.24109793e-01 -5.24713099e-01 4.60362643e-01
1.04564190e-01 -5.56763224e-02 -3.10291708e-01 1.23438716e+00
4.90656346e-01 -1.74122453e-01 4.12660331e-01 -1.57203662e+00
8.07986502e-03 1.13927805e+00 -8.72894287e-01 1.04078606e-01
-1.22868136e-01 -2.47018054e-01 9.88318503e-01 -4.00845945e-01
4.05593544e-01 1.10029185e+00 8.30062389e-01 1.99705109e-01
-1.44991946e+00 -9.37798083e-01 4.13264245e-01 5.12811728e-02
-1.29867280e+00 -6.05623782e-01 8.99548590e-01 -1.27631754e-01
6.14699721e-01 1.77182376e-01 6.38856530e-01 6.88359320e-01
4.12014842e-01 6.14416838e-01 6.61205709e-01 -5.57247043e-01
6.08535409e-01 -8.52749497e-03 -3.32773030e-01 9.55354691e-01
4.78017807e-01 2.74728954e-01 -8.15037787e-01 -5.49468637e-01
5.53550661e-01 -1.68001086e-01 1.13729216e-01 -6.90356731e-01
-1.26731956e+00 1.09319341e+00 5.32551110e-01 4.76989150e-02
-8.20502102e-01 4.61424679e-01 3.27303559e-01 9.54805195e-01
7.26655304e-01 1.55956492e-01 -8.51365864e-01 7.41371214e-01
-1.06985545e+00 -1.14772998e-01 1.10095036e+00 1.03339028e+00
1.08324504e+00 6.08467996e-01 3.90963703e-01 2.15528339e-01
7.94680834e-01 7.48329580e-01 -6.32911697e-02 -1.00224280e+00
4.97843146e-01 2.23364100e-01 -2.89945573e-01 -1.00791645e+00
-1.01386976e+00 -9.00733829e-01 -1.48837745e+00 3.23243201e-01
3.01412553e-01 -1.11110485e+00 -4.49121356e-01 1.83474863e+00
4.49235171e-01 1.37190014e-01 1.54416069e-01 6.30654514e-01
5.12686484e-02 7.30211794e-01 2.34465376e-01 -8.66877317e-01
9.10708845e-01 -6.46157920e-01 -5.84322572e-01 -4.75535274e-01
6.25788927e-01 -2.84700841e-01 -3.24124426e-01 6.58933640e-01
-1.01809835e+00 -1.51285142e-01 -1.14076197e+00 6.72433674e-01
-3.09077561e-01 -1.48947239e-01 1.02779520e+00 9.13184404e-01
-1.36046553e+00 3.37230861e-01 -8.13250840e-01 -1.04275875e-01
6.37690604e-01 6.85215294e-01 1.37380302e-01 6.81567267e-02
-1.14591956e+00 6.97208166e-01 5.18610537e-01 1.94575325e-01
-1.24448490e+00 -9.90081966e-01 -6.21316731e-01 -9.93326083e-02
5.42991400e-01 -7.53355324e-01 9.78274941e-01 -1.16861582e+00
-1.37272990e+00 3.23787004e-01 3.14902633e-01 -7.37967312e-01
4.65210497e-01 5.97209871e-01 -5.40011346e-01 5.15478179e-02
4.33877436e-03 4.59860653e-01 1.28667617e+00 -8.97717178e-01
-7.49008477e-01 -3.60154152e-01 -5.74287213e-02 1.68714255e-01
-6.23785496e-01 -4.52787101e-01 2.00498715e-01 -6.07400596e-01
1.43781826e-01 -8.05256367e-01 -7.41707444e-01 5.01784027e-01
-4.52113032e-01 -8.65037441e-02 1.19639254e+00 -6.60555542e-01
7.58028209e-01 -1.62464142e+00 3.28084171e-01 1.06311238e+00
4.12343740e-01 -3.53436083e-01 -2.86882967e-01 3.99462491e-01
-9.94574949e-02 -3.14537019e-01 -1.34597257e-01 -9.98271629e-02
-1.22443676e-01 2.99064785e-01 2.42458254e-01 8.58184457e-01
-3.32455099e-01 5.46462178e-01 -1.08246517e+00 -4.27875727e-01
-1.40039036e-02 -8.08699057e-03 -3.81060004e-01 -3.23423535e-01
-3.55297506e-01 5.03872156e-01 -8.64682198e-01 4.19131964e-01
5.62474608e-01 -3.04368168e-01 8.12098384e-01 -2.74490207e-01
1.97088882e-01 -2.09968328e-01 -1.44781375e+00 1.91375268e+00
-5.17449737e-01 4.75710779e-01 1.19803870e+00 -1.68806756e+00
7.42657721e-01 4.74913687e-01 1.41608512e+00 -7.03375220e-01
-7.23202154e-02 -1.11198254e-01 -2.02300563e-01 -9.65509266e-02
-2.06424072e-01 2.45508179e-02 -1.35302350e-01 9.61241543e-01
2.73708224e-01 1.45796269e-01 2.28337467e-01 5.95668674e-01
1.16663539e+00 -5.38719594e-01 3.86569768e-01 -7.56795824e-01
4.98641670e-01 -5.85056953e-02 5.19748449e-01 9.39467013e-01
-2.31213495e-02 -7.12370992e-01 2.98336834e-01 -3.99004698e-01
-8.72586250e-01 -1.30608296e+00 2.59985834e-01 1.33063543e+00
-3.10008079e-01 -1.60357848e-01 -3.83565992e-01 -7.20073462e-01
1.29986763e-01 2.20178574e-01 -1.05319209e-01 1.31300632e-02
-8.80953819e-02 -9.98050034e-01 3.58862191e-01 1.32641392e-02
4.01435286e-01 -5.49043775e-01 -4.05943155e-01 4.42948073e-01
8.78866911e-02 -8.32151771e-01 -2.87910551e-01 6.94289327e-01
-9.48098242e-01 -1.12731183e+00 -2.67312527e-01 -8.75385582e-01
8.71470928e-01 6.09527081e-02 9.95234549e-01 -3.67697254e-02
-5.00921190e-01 1.04278100e+00 5.34104630e-02 -7.11081862e-01
-3.02622467e-01 2.93488801e-01 2.53342390e-01 4.55616862e-02
-1.31083488e-01 -8.04340839e-01 -3.43529344e-01 -3.35994437e-02
-7.08361149e-01 -6.60511628e-02 6.51660681e-01 4.58069652e-01
1.81046754e-01 5.29396534e-01 1.06806338e+00 -8.94132912e-01
6.26890004e-01 -7.98765361e-01 -1.10393560e+00 4.56773996e-01
-1.05365694e+00 1.06043808e-01 5.93437135e-01 -2.92485654e-01
-7.56741226e-01 4.73983765e-01 7.86893547e-01 1.82742149e-01
3.96810770e-01 8.35550725e-01 -8.02246109e-02 -5.85714579e-01
6.98029459e-01 -5.35067432e-02 3.40329647e-01 9.94379893e-02
6.39508784e-01 4.77220178e-01 1.79165065e-01 -6.36871040e-01
1.19484532e+00 4.92324769e-01 6.00672662e-01 -1.17576098e+00
-8.18764627e-01 -1.90985814e-01 -5.69768906e-01 -5.99655271e-01
4.58502620e-01 -1.28266764e+00 -9.47858512e-01 1.81512251e-01
-8.51019681e-01 -3.81210417e-01 -5.50248504e-01 6.98648989e-01
-4.27188456e-01 1.89999998e-01 -3.32256466e-01 -8.60720098e-01
-4.58827168e-01 -4.83526856e-01 4.12083268e-01 -8.88640061e-02
2.21620068e-01 -1.69456124e+00 6.31497130e-02 4.06338237e-02
6.99466705e-01 3.75555575e-01 8.65564585e-01 -4.46071029e-01
-6.45293474e-01 -1.22055940e-01 -4.10841824e-03 1.53651416e-01
3.29157501e-01 -6.70013428e-01 -7.90641665e-01 -1.00989687e+00
-2.12807860e-02 -5.58159471e-01 6.15565956e-01 6.05221391e-01
1.08121443e+00 -7.16887951e-01 -3.83666217e-01 5.30483842e-01
1.50193858e+00 -1.06818058e-01 -1.54892713e-01 -5.41199557e-02
5.56468964e-01 5.91922998e-01 -2.68990137e-02 1.07038713e+00
2.00803623e-01 -6.67954534e-02 5.65997362e-01 -1.07371442e-01
2.48765722e-01 1.49115250e-01 4.54498470e-01 6.91039443e-01
3.40144694e-01 -2.31104150e-01 -6.84792876e-01 2.16692075e-01
-1.87115729e+00 -9.25508916e-01 2.79150993e-01 1.97818756e+00
5.80248296e-01 -1.73287746e-02 9.79421660e-02 2.41185613e-02
7.59450376e-01 4.12325650e-01 -1.05886364e+00 -4.13039699e-02
1.22485839e-01 1.74193129e-01 1.27226341e+00 6.36664152e-01
-1.04119027e+00 1.94009006e-01 7.08491945e+00 7.33647168e-01
-8.87379885e-01 1.58518061e-01 6.45785630e-01 -7.79610649e-02
-3.03593010e-01 -1.13371149e-01 -2.57054746e-01 5.60414104e-04
1.24555802e+00 -4.58652377e-01 1.01715875e+00 6.58929408e-01
4.82696682e-01 3.01428884e-02 -1.04469347e+00 1.02056730e+00
-1.32122874e-01 -1.36097276e+00 -3.86423737e-01 2.70362049e-01
9.83932793e-01 4.40315068e-01 -5.37360646e-02 1.10631138e-02
8.67035508e-01 -6.54608965e-01 4.28034365e-01 3.74652654e-01
5.15160978e-01 -9.40275848e-01 1.17586434e-01 5.25100827e-01
-1.38738513e+00 -5.40625989e-01 -4.33098227e-01 -1.14753842e-01
-1.61975026e-02 1.29859567e+00 -8.40359867e-01 8.56504142e-01
3.01282078e-01 9.29328084e-01 -2.80300647e-01 9.05382097e-01
1.17419034e-01 6.93069637e-01 -7.85116255e-01 6.91720322e-02
2.81379521e-01 -2.74570793e-01 6.61697686e-01 9.20493245e-01
6.35744259e-02 -2.38620758e-01 6.93840981e-01 5.07447839e-01
-4.17697161e-01 -2.51676083e-01 -4.33794022e-01 5.81512265e-02
6.72170281e-01 1.69911742e+00 -6.53753579e-01 -6.70942441e-02
-5.69681108e-01 3.53686869e-01 1.43164009e-01 7.15692937e-01
-4.18484420e-01 -1.90374672e-01 4.84843314e-01 -3.06078821e-01
2.10829988e-01 -3.42835873e-01 -2.17788797e-02 -7.12385476e-01
-3.19745451e-01 -6.78748131e-01 7.82818615e-01 -8.63719657e-02
-1.65861738e+00 -1.57682881e-01 7.22615048e-02 -1.02503479e+00
-4.87696856e-01 -5.08500218e-01 -6.70756280e-01 5.69207788e-01
-1.58211625e+00 -1.07786489e+00 -1.21270129e-02 1.13300574e+00
1.62949711e-01 -6.62905395e-01 7.86296606e-01 3.71019572e-01
-3.56717408e-01 5.30660748e-01 3.72295797e-01 3.71503234e-02
3.16242307e-01 -9.86098409e-01 -1.11705780e-01 8.35144639e-01
1.19610526e-01 -1.39697641e-01 5.40749907e-01 -6.83594525e-01
-1.65376329e+00 -1.38211751e+00 4.31865245e-01 3.43552411e-01
1.14649725e+00 -2.88829625e-01 -7.14205801e-02 7.15340853e-01
2.90705144e-01 2.74779707e-01 6.43302441e-01 7.33466670e-02
-9.26766619e-02 -5.93047321e-01 -1.34862280e+00 8.84141847e-02
8.25623035e-01 -3.02426606e-01 3.58243734e-01 9.86699879e-01
2.92766869e-01 9.95448381e-02 -9.95062470e-01 -3.14681679e-02
8.30935091e-02 -6.46020472e-02 9.86205876e-01 -6.16801500e-01
-5.74507058e-01 -5.61499931e-02 -1.20515257e-01 -1.41712427e+00
-5.70291877e-01 -1.17680335e+00 -5.22086501e-01 8.46497357e-01
6.61225021e-01 -6.18466198e-01 1.28629422e+00 1.72008991e-01
3.44084918e-01 -1.46385744e-01 -1.17881358e+00 -7.16028035e-01
-1.49095282e-01 -4.71126378e-01 -3.73542798e-03 1.09105468e+00
2.12112382e-01 4.39589977e-01 -6.81546867e-01 4.06322658e-01
1.68577397e+00 -3.97563964e-01 3.19930255e-01 -1.56331933e+00
-2.37916261e-01 -6.72415122e-02 -5.24179637e-01 -8.65145922e-01
4.08093721e-01 -1.18230975e+00 -3.06618363e-01 -1.16435492e+00
-1.38549179e-01 -7.21839547e-01 -4.14401561e-01 5.67110717e-01
6.32437527e-01 -1.01799577e-01 1.30724460e-02 -2.25212593e-02
-7.45716155e-01 1.07365780e-01 9.23865736e-01 -7.24589109e-01
-7.69149959e-02 4.97901589e-01 -6.89427912e-01 5.25315285e-01
9.58116531e-01 -6.65530503e-01 -9.42041159e-01 -4.05704439e-01
4.70076084e-01 6.09018385e-01 3.17072779e-01 -1.01587892e+00
9.84587073e-01 -3.16721320e-01 7.00042248e-01 -3.59698147e-01
9.31249261e-02 -1.46125543e+00 -1.02941645e-03 9.06078577e-01
-6.98164940e-01 -9.88636389e-02 -2.08254874e-01 1.16567397e+00
2.22057313e-01 -4.47358489e-02 6.51985526e-01 -5.19717298e-02
-7.40784943e-01 6.02768064e-01 -7.76601255e-01 -5.63027710e-03
1.22311282e+00 2.54513681e-01 3.38073052e-03 -8.56222212e-01
-9.68195140e-01 9.04401839e-01 -4.49544281e-01 3.64845432e-02
4.90480334e-01 -1.23390746e+00 -9.73819971e-01 1.06139250e-01
-4.34336156e-01 -1.95880115e-01 -6.22026101e-02 8.66921365e-01
8.60815048e-02 2.35863432e-01 -9.58571061e-02 -4.42462951e-01
-7.65785575e-01 4.88431305e-01 4.42279220e-01 -4.41564023e-01
-3.17055136e-01 6.02052271e-01 -5.39978862e-01 -6.44261897e-01
6.27718568e-01 1.43644467e-01 2.14966342e-01 1.68967336e-01
3.25510502e-01 6.92375541e-01 5.46393059e-02 1.42193064e-01
-3.90004486e-01 1.92895800e-01 1.32559225e-01 -7.56931752e-02
1.44900393e+00 -6.39767051e-01 -3.84966969e-01 8.10373724e-02
1.20539629e+00 -4.38365638e-01 -1.25949895e+00 -7.47335017e-01
3.82433474e-01 2.63834983e-01 5.79587758e-01 -6.27110779e-01
-1.72968149e+00 4.50903736e-02 7.02222884e-01 4.54148203e-01
1.15066206e+00 -1.59491777e-01 2.79279560e-01 1.28700066e+00
6.70973480e-01 -1.37491548e+00 -1.22280791e-01 3.59436214e-01
1.59233615e-01 -1.00932395e+00 8.13534319e-01 -1.86758991e-02
-1.56132713e-01 1.51078439e+00 1.42797440e-01 -1.07234873e-01
1.47866106e+00 6.32497609e-01 -1.13268621e-01 -3.00320983e-01
-7.67144263e-01 1.96263909e-01 -8.97950009e-02 9.14609313e-01
2.38505825e-02 2.10636929e-02 2.29813293e-01 -1.71687514e-01
1.28491759e-01 -3.91435862e-01 3.76554161e-01 1.12495518e+00
-6.85379386e-01 -9.95788932e-01 -2.65792549e-01 7.63194740e-01
-2.00531408e-01 2.45064288e-01 -1.29130244e-01 6.05598330e-01
-1.88172355e-01 1.26924455e+00 -4.51827832e-02 -1.01779364e-01
-2.35983700e-01 -2.50748247e-01 3.04509342e-01 -3.53659779e-01
-1.41904265e-01 8.23253989e-02 2.03193620e-01 -7.23979712e-01
-9.89909053e-01 -4.83599156e-01 -1.10424662e+00 -2.79148161e-01
-2.76729226e-01 4.31577533e-01 9.57404196e-01 1.06981516e+00
-1.34301409e-01 7.03864098e-01 1.14619970e+00 -9.84013081e-01
-8.06077123e-01 -5.70464790e-01 -8.82659793e-01 -4.67431992e-01
6.19292140e-01 -1.05771646e-01 -4.48562950e-01 5.25351763e-02] | [6.182238578796387, 5.012168884277344] |
7c956b2c-a0a2-46aa-8623-81bf3a29856a | using-consumer-behavior-data-to-reduce-energy | 1510.00165 | null | http://arxiv.org/abs/1510.00165v1 | http://arxiv.org/pdf/1510.00165v1.pdf | Using consumer behavior data to reduce energy consumption in smart homes | This paper discusses how usage patterns and preferences of inhabitants can be
learned efficiently to allow smart homes to autonomously achieve energy
savings. We propose a frequent sequential pattern mining algorithm suitable for
real-life smart home event data. The performance of the proposed algorithm is
compared to existing algorithms regarding completeness/correctness of the
results, run times as well as memory consumption and elaborates on the
shortcomings of the different solutions. We also present a recommender system
based on the developed algorithm that provides recommendations to the users to
reduce their energy consumption. The recommender system was deployed to a set
of test homes. The test participants rated the impact of the recommendations on
their comfort. We used this feedback to adjust the system parameters and make
it more accurate during a second test phase. | ['Hans-Friedrich Witschel', 'Daniel Schweizer', 'Holger Wache', 'Miguel Rodriguez', 'Michael Zehnder', 'Danilo Zanatta'] | 2015-10-01 | null | null | null | null | ['sequential-pattern-mining'] | ['natural-language-processing'] | [-6.68286532e-02 2.59225935e-01 1.11972988e-01 -7.71728694e-01
1.41609013e-01 -2.59849340e-01 1.28450930e-01 2.40320936e-01
-2.18501434e-01 8.56385589e-01 5.61982393e-01 -1.85624212e-01
-6.14426613e-01 -1.07276821e+00 8.05192068e-02 -7.14832485e-01
-2.32426256e-01 5.95437467e-01 3.35546523e-01 -1.40962988e-01
1.33125603e-01 4.28445935e-01 -2.31895375e+00 4.21725959e-01
1.00593853e+00 6.80364609e-01 4.70441550e-01 4.74293321e-01
3.35448086e-01 4.31886196e-01 -4.93572742e-01 3.84600431e-01
2.72554219e-01 -1.43498808e-01 -4.61207777e-01 8.03120956e-02
-4.14963216e-01 -4.69696045e-01 1.26694560e-01 4.74673808e-01
6.29436672e-01 7.18769968e-01 5.11196315e-01 -1.43221521e+00
4.16493304e-02 5.45127034e-01 3.43613267e-01 8.55981782e-02
9.79313254e-01 -1.81649506e-01 6.88817143e-01 -2.68154174e-01
2.79879689e-01 5.91653466e-01 4.52560514e-01 2.82283008e-01
-1.00430155e+00 -5.87700784e-01 2.95250207e-01 9.43278193e-01
-1.53931487e+00 -5.31215847e-01 5.55406272e-01 -1.25064015e-01
1.62028396e+00 8.57049942e-01 1.15489721e+00 6.58972502e-01
-7.24271014e-02 2.35440522e-01 9.40276802e-01 -6.54089272e-01
9.54525352e-01 5.55219352e-01 4.25645232e-01 4.47286993e-01
6.21383965e-01 1.85386911e-01 1.08399227e-01 -4.11329657e-01
-1.39889151e-01 3.61481875e-01 -3.20666313e-01 -1.56706005e-01
-5.60170412e-01 6.83681846e-01 -1.73259177e-03 7.11085498e-01
-6.36794508e-01 -5.75845242e-01 1.04393110e-01 1.78166956e-01
-1.04888141e-01 8.98203030e-02 -7.67841518e-01 -1.74768254e-01
-7.05668867e-01 1.34907410e-01 1.33096671e+00 1.07929134e+00
5.40414393e-01 -4.07922208e-01 -9.58257914e-02 6.49662197e-01
6.80221915e-01 3.52681667e-01 6.67854130e-01 -4.96232957e-01
-1.71830803e-01 9.39913630e-01 5.28112411e-01 -7.24383116e-01
-8.28769863e-01 2.67071515e-01 -5.26702702e-01 2.60646671e-01
4.49135862e-02 -4.92870390e-01 -4.87584084e-01 1.17869473e+00
4.96347457e-01 -3.18137966e-02 -5.85522838e-02 5.69147408e-01
4.77166235e-01 7.45095253e-01 2.51925260e-01 -8.43568504e-01
1.27817345e+00 -5.06745517e-01 -9.39956665e-01 3.05569649e-01
5.15299857e-01 -4.27757055e-01 8.76757205e-01 4.59377587e-01
-5.85740745e-01 -3.05441499e-01 -1.07010233e+00 6.55020952e-01
-4.81330067e-01 7.42073581e-02 4.48594928e-01 1.27362239e+00
-1.02132571e+00 7.45049834e-01 -8.85744035e-01 -1.33980322e+00
-1.11727498e-01 8.43334734e-01 2.10539252e-01 3.06281239e-01
-8.01342428e-01 1.02086687e+00 4.97910351e-01 -3.36244971e-01
5.76507812e-03 -4.38696951e-01 -5.78074217e-01 2.90593877e-02
-2.96464022e-02 -6.89377129e-01 1.20183754e+00 -3.48846138e-01
-1.61080933e+00 2.80280650e-01 -1.83540910e-01 -3.70407909e-01
3.16494495e-01 -1.03226295e-02 -1.08409715e+00 -3.54964763e-01
-7.27212504e-02 -3.20823751e-02 -7.86841512e-02 -8.04458380e-01
-1.34623146e+00 -2.71245062e-01 -1.97523758e-01 4.03277203e-02
-3.09352577e-01 -3.46846312e-01 2.86842793e-01 -2.17650354e-01
-3.08833867e-01 -8.71109605e-01 -3.90172601e-01 -7.56867528e-01
-9.72380489e-03 -5.05672693e-01 9.63974714e-01 -4.73470569e-01
1.55746424e+00 -1.80191350e+00 -6.32984698e-01 7.94722319e-01
-6.83367372e-01 -1.02502622e-01 4.34861600e-01 7.62479246e-01
3.49908844e-02 -3.33817482e-01 8.66226926e-02 5.85735440e-02
2.66015381e-01 5.48657417e-01 1.85881957e-01 3.06551933e-01
-6.85527027e-01 9.72435623e-02 -6.76907718e-01 -1.77966982e-01
9.99792695e-01 4.81593430e-01 -5.06288946e-01 5.63566625e-01
-1.69428527e-01 1.46817386e-01 -4.52080339e-01 5.57588577e-01
2.20337093e-01 2.33271286e-01 6.42053187e-01 -4.16829258e-01
-3.79443169e-01 5.34998357e-01 -1.61196387e+00 9.60129619e-01
-5.96594632e-01 7.90876895e-02 -4.06278789e-01 -7.14888692e-01
8.74232352e-01 5.68395376e-01 9.99709368e-01 -8.18033159e-01
4.08184826e-01 -6.11221455e-02 -3.26688319e-01 -1.00837564e+00
2.52078354e-01 2.69308835e-01 1.41111210e-01 7.90662408e-01
-4.11769599e-01 5.27954638e-01 4.41879779e-01 -4.17688847e-01
1.30093467e+00 -1.09277092e-01 9.97285962e-01 -6.98486865e-01
7.04664230e-01 -8.32942054e-02 4.73145664e-01 2.69044846e-01
-1.34291485e-01 -2.47130528e-01 -6.47307456e-01 -8.14848125e-01
-8.24419022e-01 -6.05263412e-01 1.24549888e-01 1.21655738e+00
-7.18767270e-02 -5.57380795e-01 -6.62586033e-01 -5.65576911e-01
-2.03166232e-01 1.52409816e+00 -4.26000476e-01 1.74632668e-01
-2.84964800e-01 -1.01114869e+00 -4.30257171e-01 2.04329178e-01
4.67759192e-01 -1.21018803e+00 -1.64325070e+00 3.92471761e-01
-1.77392140e-01 -7.46593535e-01 -5.35251610e-02 4.15037125e-01
-6.74772561e-01 -1.22774804e+00 3.41006279e-01 -5.34410834e-01
7.63271511e-01 2.50937492e-01 8.54835391e-01 2.14346975e-01
-2.07122728e-01 6.73090518e-01 -8.09146404e-01 -4.48718667e-01
-1.72715887e-01 1.15543008e-01 3.03400218e-01 -2.62953460e-01
8.48219573e-01 -1.09722245e+00 -8.82573366e-01 6.21813774e-01
-5.03434241e-01 -2.26715297e-01 2.70801276e-01 1.24405287e-01
4.65848595e-01 1.03404820e+00 3.17542136e-01 -8.44136178e-01
7.43776441e-01 -6.95460618e-01 -6.01527095e-01 3.38691920e-01
-1.25199521e+00 2.72610933e-01 5.98405838e-01 -2.77893156e-01
-1.01343060e+00 4.92087096e-01 -2.60979921e-01 7.04497814e-01
-7.83495665e-01 -1.96679235e-01 -7.09056318e-01 1.38111591e-01
4.39531773e-01 -3.49712782e-02 -5.86196065e-01 -4.98327523e-01
5.68380691e-02 7.67673254e-01 7.56249502e-02 -2.27684170e-01
5.22649407e-01 1.48795307e-01 -4.51999366e-01 -8.46461058e-01
-1.85926184e-01 -6.69605732e-01 -4.55940604e-01 -3.80843818e-01
5.04751563e-01 -3.81800234e-01 -1.07477820e+00 -1.86143517e-01
-6.66729867e-01 -4.42136437e-01 -4.70486253e-01 2.43039399e-01
-5.31785190e-01 -1.01415314e-01 1.49250850e-01 -1.23593628e+00
-7.58669138e-01 -6.35283530e-01 2.69127607e-01 4.46700901e-01
-1.02353311e+00 -8.16092730e-01 4.43502516e-01 3.97428833e-02
4.57447410e-01 2.82597572e-01 7.59372652e-01 -7.92849183e-01
-6.16482794e-02 -1.82584941e-01 4.72133338e-01 -3.22608948e-01
7.45405257e-01 -4.91298623e-02 -9.42116678e-01 -2.95224935e-01
5.21154739e-02 5.74459434e-01 3.54524888e-02 3.53722423e-01
7.51876891e-01 -9.63075578e-01 -7.41047144e-01 9.65007916e-02
1.54037535e+00 9.28334177e-01 7.54960418e-01 6.21082485e-01
3.53153050e-02 5.40050566e-01 6.89917207e-01 9.36774313e-01
7.11137056e-01 8.20639431e-01 2.85758197e-01 3.16427499e-01
3.20192873e-01 3.54217030e-02 5.26021659e-01 5.42681575e-01
-3.34394008e-01 -2.56953061e-01 -2.80896485e-01 5.44524252e-01
-2.20532131e+00 -1.17021513e+00 -2.93379109e-02 2.17834282e+00
2.01089740e-01 -1.02979228e-01 9.07360733e-01 8.43055844e-01
5.19492209e-01 -3.41366261e-01 -2.59779006e-01 -6.63430810e-01
4.71043706e-01 2.48712674e-01 5.65480351e-01 5.40928960e-01
-8.55295599e-01 1.14874914e-01 6.81894350e+00 -1.81956530e-01
-5.20119488e-01 1.21994630e-01 2.11348742e-01 -2.55018950e-01
-1.06056057e-01 -3.27658534e-01 -6.25636101e-01 7.14094400e-01
1.39707911e+00 -3.35824698e-01 6.16828620e-01 1.21506536e+00
8.03071201e-01 -4.30168420e-01 -1.19910216e+00 9.94478583e-01
-1.00219443e-01 -8.90875638e-01 -2.67412901e-01 1.32814661e-01
6.17285132e-01 -1.37347549e-01 -8.01278651e-01 -8.41145217e-02
7.42157519e-01 -6.01302207e-01 3.66955191e-01 5.79317808e-01
4.46254462e-02 -1.06324160e+00 9.37961638e-01 4.28161353e-01
-1.41129291e+00 -5.74221671e-01 -1.97211564e-01 -4.96646732e-01
3.09756279e-01 5.06547928e-01 -1.33866382e+00 3.47826302e-01
1.02348351e+00 1.11889824e-01 -3.77736270e-01 1.21083128e+00
-1.31869808e-01 3.90908122e-01 -9.01713252e-01 -6.15565002e-01
-4.28871870e-01 -3.50066364e-01 1.13780648e-01 1.22275102e+00
8.48241031e-01 7.81341791e-01 4.20519412e-02 3.65472049e-01
7.51220465e-01 4.04610425e-01 -5.86401522e-01 6.77891135e-01
7.82887757e-01 1.21195745e+00 -8.51663113e-01 -2.36886591e-01
-3.05628300e-01 9.48773384e-01 -2.29497075e-01 4.25444543e-03
-4.27155256e-01 -1.24188639e-01 6.90788388e-01 5.55866420e-01
4.06372577e-01 1.62334159e-01 -2.24435985e-01 -5.81247270e-01
-8.04165304e-02 -4.92713779e-01 8.63995254e-01 -3.82162035e-01
-9.09431279e-01 5.75630128e-01 3.36781800e-01 -8.53154778e-01
-5.12900710e-01 -2.08364740e-01 -7.10732102e-01 1.54216096e-01
-7.46403813e-01 -6.50879622e-01 -6.91907883e-01 8.46782029e-01
5.28032124e-01 -1.28732249e-01 1.47158837e+00 4.79365736e-01
-4.50223088e-01 4.39513296e-01 -1.08591743e-01 -7.44047344e-01
1.61118999e-01 -1.11905956e+00 -6.24471419e-02 7.51773775e-01
-1.42314225e-01 3.27508539e-01 1.25401103e+00 -5.92907131e-01
-9.83155906e-01 -1.05563819e+00 9.09541726e-01 -1.28736675e-01
1.64314374e-01 -1.83415726e-01 -4.35857892e-01 4.87657100e-01
3.16752881e-01 -6.38288736e-01 1.31360447e+00 7.33874366e-02
4.51049209e-01 -5.47800183e-01 -1.93401468e+00 4.12791282e-01
1.00954616e+00 2.51616836e-01 -5.13253152e-01 1.89223513e-01
1.61515728e-01 3.78094494e-01 -9.11436975e-01 2.00218171e-01
8.32403183e-01 -1.25199032e+00 4.06844109e-01 5.04375882e-02
-8.15695345e-01 -5.84951401e-01 -5.17090738e-01 -1.25438619e+00
-7.49310553e-01 -5.94379485e-01 -2.43037418e-01 1.20304668e+00
2.97747463e-01 -7.03546941e-01 7.74975300e-01 1.15391135e+00
1.30783468e-01 -3.47514659e-01 -7.68499255e-01 -2.29904458e-01
-9.86668110e-01 -4.51407164e-01 1.16974747e+00 5.58944941e-01
5.67174852e-01 1.95374727e-01 -2.01618701e-01 3.78307074e-01
6.08963609e-01 -1.78878941e-02 4.69761163e-01 -1.55658269e+00
-1.71626940e-01 -7.31783882e-02 -4.06777054e-01 -1.47920504e-01
-4.33258563e-01 -3.08958054e-01 -6.00802228e-02 -1.81764197e+00
-5.95380776e-02 -4.26902771e-01 -4.24277693e-01 5.72883725e-01
3.72234553e-01 -1.23038851e-01 -2.31227309e-01 -2.42876768e-01
-4.11401778e-01 9.18726102e-02 1.94468740e-02 -7.59275332e-02
-9.47618008e-01 5.79552889e-01 -6.50901735e-01 6.07490063e-01
1.17613804e+00 -3.01099479e-01 -8.46052468e-01 2.37656862e-01
1.36824474e-01 -3.94282699e-01 -1.50835171e-01 -1.45953381e+00
3.44861090e-01 -3.91694963e-01 4.32508498e-01 -8.33393812e-01
-1.95469633e-01 -1.87084997e+00 1.12123513e+00 8.69579256e-01
1.27676666e-01 1.36547193e-01 -5.12351375e-03 2.57636964e-01
6.91381454e-01 -6.09597713e-02 5.26778042e-01 1.34278029e-01
-7.61626184e-01 -9.45596918e-02 -8.48574221e-01 -1.07721722e+00
1.60220885e+00 -5.68721533e-01 2.28486627e-01 -2.72803038e-01
-8.86138082e-01 2.26252943e-01 5.95626831e-01 2.71181583e-01
3.82075757e-01 -1.42478919e+00 -7.01628402e-02 2.60709941e-01
8.67802277e-02 -4.80397940e-01 1.18011273e-01 3.92904043e-01
-2.84420848e-01 4.01492208e-01 -4.63996887e-01 -3.45539190e-02
-1.69756079e+00 7.11050093e-01 3.22667420e-01 -1.63175002e-01
-8.61391723e-01 -7.48594999e-02 -6.47766471e-01 -2.28223160e-01
5.30439138e-01 -7.05032229e-01 -6.84039116e-01 1.59292128e-02
7.78531373e-01 9.86754775e-01 2.96956748e-01 -2.61464387e-01
-7.83981740e-01 3.83178115e-01 2.66699374e-01 1.06467143e-01
1.80303955e+00 -5.30304492e-01 5.28784320e-02 4.18250740e-01
6.12517595e-01 -6.75488776e-03 -7.99230635e-01 3.53305995e-01
3.18109006e-01 -1.46291167e-01 -1.10171154e-01 -9.78512585e-01
-8.56502473e-01 -3.68317097e-01 1.47902811e+00 4.88787144e-01
1.62488747e+00 -2.51090765e-01 8.13270867e-01 6.58772707e-01
8.32200527e-01 -1.53943729e+00 -6.76936567e-01 -1.54834315e-01
4.38722432e-01 -9.73191917e-01 3.31444114e-01 -3.64554822e-01
-2.61719197e-01 1.15192711e+00 2.32651472e-01 -8.55695978e-02
1.15571654e+00 5.24690449e-01 -3.32321942e-01 -9.96714458e-02
-8.19586694e-01 -1.60462871e-01 -1.36723071e-01 1.11670923e+00
1.13317490e-01 6.05252266e-01 -6.45667076e-01 9.17456388e-01
-5.91296315e-01 4.16364461e-01 3.10919553e-01 9.96184468e-01
-8.92374873e-01 -1.28633642e+00 -4.74392235e-01 6.71538651e-01
1.18484404e-02 6.02598488e-01 -2.15043291e-01 5.39570034e-01
6.04128897e-01 1.64999056e+00 -1.25508696e-01 -8.59403789e-01
1.06635809e+00 -8.24722089e-03 4.00513768e-01 -3.87186557e-01
-5.33385336e-01 -2.23197967e-01 5.17188430e-01 -7.62575924e-01
-6.47104263e-01 -9.81995344e-01 -1.30943620e+00 -5.72518945e-01
-9.94135216e-02 5.78967690e-01 5.24375498e-01 9.44127381e-01
2.83356071e-01 6.33355081e-01 9.13565040e-01 -5.83966076e-01
8.98313522e-02 -1.07931554e+00 -5.99528730e-01 3.99525911e-01
-1.05255812e-01 -6.95162952e-01 -3.14883202e-01 2.50822276e-01] | [5.961394786834717, 2.550546169281006] |
64501468-85e4-442e-b5f4-8c0c120859af | kinematic-structure-correspondences-via | null | null | http://openaccess.thecvf.com/content_cvpr_2016/html/Chang_Kinematic_Structure_Correspondences_CVPR_2016_paper.html | http://openaccess.thecvf.com/content_cvpr_2016/papers/Chang_Kinematic_Structure_Correspondences_CVPR_2016_paper.pdf | Kinematic Structure Correspondences via Hypergraph Matching | In this paper, we present a novel framework for finding the kinematic structure correspondence between two objects in videos via hypergraph matching. In contrast to prior appearance and graph alignment based matching methods which have been applied among two similar static images, the proposed method finds correspondences between two dynamic kinematic structures of heterogeneous objects in videos. Our main contributions can be summarised as follows: (i) casting the kinematic structure correspondence problem into a hypergraph matching problem, incorporating multi-order similarities with normalising weights, (ii) a structural topology similarity measure by a new topology constrained subgraph isomorphism aggregation, (iii) a kinematic correlation measure between pairwise nodes, and (iv) a combinatorial local motion similarity measure using geodesic distance on the Riemannian manifold. We demonstrate the robustness and accuracy of our method through a number of experiments on complex articulated synthetic and real data. | ['Yiannis Demiris', 'Martina Zambelli', 'Hyung Jin Chang', 'Tobias Fischer', 'Maxime Petit'] | 2016-06-01 | null | null | null | cvpr-2016-6 | ['hypergraph-matching'] | ['graphs'] | [ 1.86038762e-01 2.80954223e-02 2.16113627e-01 -1.51666060e-01
-2.56690651e-01 -5.63081861e-01 8.01701188e-01 1.28821328e-01
-1.26283243e-01 2.05423251e-01 6.04133196e-02 1.26249000e-01
-7.85132587e-01 -4.92252886e-01 -5.49273849e-01 -4.84678358e-01
-5.99462628e-01 4.93177772e-01 5.23590267e-01 -1.36156395e-01
3.22692096e-01 6.77443385e-01 -1.30033302e+00 -4.31820154e-01
5.96343100e-01 7.00792968e-01 2.03526989e-01 7.61316538e-01
2.43076459e-01 4.61909413e-01 -6.48446605e-02 -5.37515879e-01
5.47726393e-01 -7.34912336e-01 -1.31587934e+00 6.42938018e-01
1.04940343e+00 2.93261223e-02 -5.53633988e-01 1.08693266e+00
2.40312219e-01 6.16986930e-01 6.73516452e-01 -1.44461632e+00
-1.96481392e-01 8.62501040e-02 -5.09974778e-01 3.56711328e-01
7.94542611e-01 -1.27081499e-01 1.22007453e+00 -7.74404585e-01
1.25501156e+00 1.46448541e+00 5.99572897e-01 1.24619059e-01
-1.30359960e+00 1.32596314e-01 -7.56426062e-03 4.73307431e-01
-1.28315341e+00 -2.44483262e-01 9.93648112e-01 -6.69274151e-01
6.74146473e-01 2.35218167e-01 8.92179668e-01 6.21840954e-01
3.26806813e-01 2.98369348e-01 9.02475476e-01 -4.06123459e-01
1.39103131e-02 -4.76808459e-01 -4.54454906e-02 1.20746505e+00
1.67803302e-01 -2.55622715e-02 -3.44189227e-01 -3.65748912e-01
1.04324472e+00 -1.79467529e-01 -1.86850533e-01 -1.09133124e+00
-1.64653957e+00 7.79134870e-01 3.57788771e-01 4.98003513e-01
-1.89225391e-01 3.48355263e-01 2.13492483e-01 4.38549250e-01
1.84718668e-01 4.11040306e-01 -2.72334204e-03 -9.33190510e-02
-4.76307929e-01 2.33302772e-01 9.12625790e-01 1.09837294e+00
9.72106278e-01 -1.11308232e-01 4.25838172e-01 5.71420908e-01
4.11633462e-01 4.21574026e-01 1.26877099e-01 -1.35397065e+00
3.75844181e-01 4.37900424e-01 -2.62858927e-01 -1.78779352e+00
-5.05015016e-01 1.13746710e-01 -5.59870660e-01 3.53412703e-03
5.50017416e-01 2.53747076e-01 -3.40212941e-01 1.82170260e+00
6.47587121e-01 4.80740488e-01 -2.97245055e-01 1.02403629e+00
5.86405575e-01 3.02187055e-01 -3.28709036e-01 -1.74102291e-01
1.16652119e+00 -8.61950517e-01 -6.15746021e-01 3.04667681e-01
4.98045862e-01 -9.72672999e-01 5.77132523e-01 -2.34522998e-01
-1.39751315e+00 -5.30996799e-01 -9.77210402e-01 8.47321227e-02
-1.27769619e-01 -4.90400106e-01 4.05719191e-01 2.54354954e-01
-1.30278885e+00 1.02067900e+00 -9.30957258e-01 -8.74593198e-01
-3.50846887e-01 3.90320748e-01 -6.64515734e-01 1.52255952e-01
-7.47823477e-01 8.95675778e-01 1.98293746e-01 8.02037492e-02
-4.14178729e-01 -4.50007379e-01 -1.07230508e+00 -3.39689344e-01
5.23187220e-01 -1.07760751e+00 7.19148993e-01 -8.09417546e-01
-1.66825891e+00 1.05041885e+00 7.19125122e-02 5.19394204e-02
6.10888302e-01 1.64800525e-01 -2.56756395e-01 5.29179454e-01
-3.13534425e-03 4.20430541e-01 5.56379914e-01 -1.33044040e+00
-4.77389276e-01 -2.85339892e-01 1.64638564e-01 4.96208698e-01
2.55312622e-01 3.80218066e-02 -4.81585324e-01 -7.84556031e-01
6.87726676e-01 -1.27514338e+00 -2.55216390e-01 1.58715710e-01
-2.99938500e-01 -2.55374014e-01 9.13806498e-01 -7.70780742e-01
8.80911112e-01 -1.83753169e+00 1.07353437e+00 7.58257031e-01
4.13702697e-01 -1.34672835e-01 -2.91290283e-01 5.82461178e-01
-7.99213126e-02 -1.06489442e-01 -2.76905596e-01 1.84223831e-01
-6.05017208e-02 2.08019629e-01 3.34853202e-01 1.09033346e+00
-7.79808387e-02 8.50144625e-01 -1.27227974e+00 -7.35803664e-01
2.93882459e-01 2.56427139e-01 -2.98587501e-01 1.97672933e-01
3.37313771e-01 3.98249179e-01 -4.97312635e-01 4.50840026e-01
4.19080764e-01 9.86939413e-04 3.48920524e-01 -5.66468596e-01
2.85972171e-02 -2.29368918e-02 -1.62225521e+00 1.90530372e+00
-2.36348435e-02 4.72861916e-01 3.78720254e-01 -1.10377407e+00
8.88118744e-01 2.30033502e-01 1.00404453e+00 -1.99331820e-01
2.69414812e-01 4.90066037e-02 1.80920541e-01 -6.15335584e-01
4.68212098e-01 9.60143656e-02 -1.32490499e-02 5.12890697e-01
2.42980570e-01 -1.79049045e-01 3.83053124e-01 3.90315980e-01
1.22160375e+00 2.60864913e-01 2.09791601e-01 -7.10864067e-01
8.57819140e-01 -2.74167508e-01 2.92944968e-01 2.34378815e-01
-1.40581891e-01 5.75646460e-01 2.55636454e-01 -6.08964443e-01
-1.30702686e+00 -1.49370992e+00 4.06928137e-02 3.23077321e-01
7.30464816e-01 -5.79250515e-01 -8.41873050e-01 -5.40337503e-01
1.35085449e-01 -2.57635146e-01 -5.19788921e-01 1.38239404e-02
-8.60728920e-01 -2.55676091e-01 1.32669851e-01 2.58398265e-01
4.73176062e-01 -6.77448273e-01 -5.67777753e-01 4.19122726e-03
-4.50305134e-01 -1.32851923e+00 -9.92149293e-01 -4.10376459e-01
-1.23111475e+00 -1.54649854e+00 -5.17823100e-01 -1.07192659e+00
8.06417704e-01 6.15687490e-01 1.00441062e+00 4.41889226e-01
-5.75486720e-01 1.30049312e+00 -2.23164856e-01 5.33115149e-01
-4.03323233e-01 -1.19515449e-01 3.86597455e-01 2.28947341e-01
-3.21734905e-01 -7.83408463e-01 -5.76031387e-01 7.87313700e-01
-7.61675179e-01 -2.97172572e-02 2.40087092e-01 5.71377039e-01
6.44783020e-01 -2.17837572e-01 -7.46785477e-02 -3.59207898e-01
2.98925132e-01 -3.03332478e-01 -5.75631082e-01 3.40714693e-01
-3.67986202e-01 7.78347775e-02 1.62145838e-01 -3.88913363e-01
-7.79529333e-01 3.14664423e-01 4.52189654e-01 -4.64877754e-01
1.50221318e-01 3.93837214e-01 -2.29596004e-01 -7.14080095e-01
2.22045675e-01 1.04512982e-02 4.22700971e-01 -3.35255891e-01
5.86937249e-01 -8.02324805e-03 7.28009999e-01 -7.25558519e-01
1.23572755e+00 7.90352345e-01 6.60663247e-01 -1.08542848e+00
-7.20255747e-02 -8.57217312e-01 -1.52229404e+00 -6.47469401e-01
1.01069236e+00 -4.24934983e-01 -7.42441535e-01 3.06109756e-01
-1.02738857e+00 3.61618446e-03 -1.72327623e-01 7.63299167e-01
-1.09560752e+00 1.27392161e+00 -6.23081982e-01 -3.63086253e-01
2.21175596e-01 -9.37826216e-01 9.19679463e-01 -1.44866392e-01
-3.64363164e-01 -1.46112800e+00 6.20682299e-01 3.54888588e-01
-1.71403915e-01 7.97170222e-01 7.01129854e-01 -2.66226590e-01
-8.31777215e-01 -9.33950841e-02 1.18456213e-02 -1.60202280e-01
3.34362656e-01 3.79754961e-01 -1.08233199e-01 -5.38802564e-01
-9.37556773e-02 2.00039908e-01 3.19588035e-01 2.56523550e-01
3.37415367e-01 -1.71409085e-01 -4.67113823e-01 6.76642001e-01
1.37511683e+00 1.16079614e-01 5.32881677e-01 4.23087716e-01
9.99465287e-01 9.36067700e-01 4.89223510e-01 9.36230198e-02
1.75274879e-01 1.12834418e+00 4.07374084e-01 -1.32918045e-01
-3.42297673e-01 -1.45933628e-01 2.71313518e-01 1.38620245e+00
-6.96399093e-01 1.73859507e-01 -6.46480560e-01 5.76600671e-01
-2.23541212e+00 -1.14656937e+00 -4.26190674e-01 2.22905636e+00
9.99563560e-02 -3.60168278e-01 4.38352495e-01 -1.99832201e-01
1.09519041e+00 6.05767630e-02 -1.53642878e-01 -2.58586496e-01
-1.72658250e-01 -2.09494397e-01 3.53215963e-01 6.98158741e-01
-1.06450891e+00 6.07611299e-01 6.45477390e+00 4.44730163e-01
-4.69734967e-01 1.03970073e-01 -1.48717746e-01 2.84311891e-01
-1.23321213e-01 2.96774775e-01 -9.35799703e-02 2.76071787e-01
3.84364247e-01 -2.64359534e-01 4.00908113e-01 6.60611033e-01
-8.37294757e-02 -2.25369935e-03 -1.09155118e+00 9.92747247e-01
3.17237943e-01 -1.26623213e+00 8.13046619e-02 3.65957946e-01
7.77075112e-01 -2.73662865e-01 -1.92744449e-01 -6.81643426e-01
1.42050058e-01 -4.75717932e-01 6.54457867e-01 6.54885828e-01
2.45838881e-01 -5.50072968e-01 3.60016257e-01 -3.00787508e-01
-1.81884038e+00 3.60443056e-01 -3.12700659e-01 1.75858691e-01
4.16396946e-01 1.07710078e-01 -4.66014236e-01 1.13066602e+00
5.35779297e-01 8.78845215e-01 -4.69921172e-01 1.35476196e+00
3.57874446e-02 1.15025043e-02 -2.12625608e-01 1.66941792e-01
2.84421176e-01 -1.00974238e+00 1.10093296e+00 1.05012715e+00
2.08550230e-01 1.58631235e-01 2.56584316e-01 5.78549385e-01
2.84967721e-01 2.42146105e-01 -8.48875105e-01 3.19622368e-01
-2.70570312e-02 1.49651849e+00 -1.24523008e+00 -1.22553408e-01
-5.02402186e-01 1.20588422e+00 1.19324200e-01 1.81455150e-01
-5.52071571e-01 -2.28718609e-01 7.30073929e-01 1.45256117e-01
1.78469017e-01 -6.58593118e-01 4.24095422e-01 -1.11200798e+00
1.59749895e-01 -5.06881177e-01 4.93911505e-01 -6.09228194e-01
-1.19739640e+00 3.99148226e-01 4.50516373e-01 -1.40231287e+00
-2.95485109e-01 -5.44695020e-01 -6.44535065e-01 4.17023331e-01
-7.17578292e-01 -1.03645134e+00 -3.01659167e-01 7.14499176e-01
2.83141941e-01 1.30540282e-02 5.11977136e-01 2.16340959e-01
-2.23821178e-01 1.30848184e-01 1.79312378e-01 1.64022535e-01
3.98409456e-01 -1.30487740e+00 5.23411095e-01 8.48535299e-01
3.11924964e-01 5.96616387e-01 6.84513628e-01 -7.67975330e-01
-1.85949671e+00 -7.96549916e-01 7.40844250e-01 -4.89795387e-01
1.01284659e+00 -2.44674638e-01 -9.62134063e-01 6.78744972e-01
2.07787037e-01 3.34915286e-03 3.33738595e-01 -2.52927095e-01
-1.42702222e-01 1.20565653e-01 -1.06863260e+00 5.88054001e-01
1.69037163e+00 -4.23587829e-01 -6.22268558e-01 5.22469163e-01
3.09477597e-01 -4.34759736e-01 -1.39268899e+00 3.28937888e-01
6.97164595e-01 -7.57543325e-01 1.11381292e+00 -7.48937070e-01
-3.46048810e-02 -5.60693145e-01 -1.43892199e-01 -1.15813518e+00
-6.17332995e-01 -1.10465634e+00 1.90014452e-01 1.00967240e+00
9.62020364e-03 -4.30136025e-01 5.72690725e-01 3.37985218e-01
-1.49674311e-01 -5.06893396e-01 -1.04704154e+00 -1.23120570e+00
-3.97526264e-01 1.37825176e-01 1.74279168e-01 1.08741868e+00
1.67247802e-01 2.09990248e-01 -3.08905751e-01 1.23305567e-01
9.19801295e-01 1.67031050e-01 8.77592921e-01 -1.32638073e+00
-2.30391994e-01 -5.34899294e-01 -1.26525939e+00 -8.42246354e-01
2.89821178e-01 -1.08058655e+00 1.99842732e-02 -1.32384217e+00
4.06509489e-01 -9.26753134e-02 1.79478183e-01 -1.39181510e-01
2.98560947e-01 2.23558664e-01 3.57694596e-01 4.03168827e-01
-7.17220724e-01 4.82332051e-01 1.33123648e+00 3.64250620e-03
3.76416892e-02 -2.11508706e-01 1.88727811e-01 8.78720224e-01
3.52746516e-01 -2.70157874e-01 -4.24783111e-01 -8.29262659e-02
1.75325513e-01 2.17159659e-01 5.09269595e-01 -8.97692919e-01
1.74252406e-01 -1.93576425e-01 -1.33306384e-01 -3.50703537e-01
3.19131941e-01 -9.33127761e-01 6.43611908e-01 7.93976486e-01
2.49282643e-02 7.06686735e-01 -2.20042795e-01 9.29552257e-01
-9.50369611e-02 -2.64635742e-01 7.66744673e-01 -9.85564366e-02
-9.74322855e-01 4.85515296e-01 -3.31215650e-01 3.73921208e-02
1.17569482e+00 -7.67004848e-01 1.23534184e-02 -3.82920444e-01
-9.29591715e-01 1.19072668e-01 9.19426084e-01 5.23301780e-01
8.14037025e-01 -1.66975796e+00 -5.01772404e-01 -1.58265233e-01
1.47819035e-02 -4.45372164e-01 7.35389367e-02 1.10975945e+00
-7.64035285e-01 1.89982116e-01 -4.04388785e-01 -8.78018618e-01
-1.83983731e+00 4.80395019e-01 4.88562316e-01 9.74975377e-02
-7.31853902e-01 3.82770091e-01 8.26994628e-02 -4.32408780e-01
-1.03163077e-02 -1.92558721e-01 1.05619624e-01 -2.01945961e-01
3.42848673e-02 1.03923500e+00 -9.01358798e-02 -1.42018616e+00
-5.26195407e-01 1.39163077e+00 3.68809938e-01 -2.32359841e-01
1.09700012e+00 -5.35967767e-01 -4.19005245e-01 4.13968533e-01
1.51675665e+00 -2.53762603e-01 -1.14241147e+00 -2.53069669e-01
2.21774295e-01 -7.78200150e-01 -3.58196974e-01 8.52351934e-02
-1.13341832e+00 3.68332148e-01 5.33328414e-01 4.22097623e-01
7.50945807e-01 4.29095060e-01 7.12145448e-01 4.59309965e-01
3.22282046e-01 -1.14366376e+00 4.65059012e-01 3.20338875e-01
1.00823951e+00 -8.06235790e-01 2.56067663e-01 -9.06856894e-01
-2.89807051e-01 1.31435525e+00 3.60064894e-01 -4.36095834e-01
7.19684362e-01 -1.98526874e-01 -1.59340084e-01 -5.76187968e-01
-3.29249322e-01 -2.78641015e-01 7.77312517e-01 7.29362845e-01
2.25857347e-01 -2.97206998e-01 -5.44257998e-01 -5.69324017e-01
-1.64764628e-01 -6.64463460e-01 5.43342292e-01 9.59268153e-01
-2.96009511e-01 -1.13877571e+00 -3.23232710e-01 7.82207474e-02
-2.50877649e-01 4.59143132e-01 -6.34638727e-01 1.08619428e+00
-2.45165661e-01 8.94900084e-01 1.70571789e-01 -5.31165898e-01
5.11615455e-01 -3.35652560e-01 1.04361987e+00 -3.99367839e-01
-4.03515667e-01 9.67411771e-02 4.49560434e-02 -9.65285957e-01
-1.08075500e+00 -1.09826517e+00 -9.77666318e-01 -4.07086074e-01
-2.99513131e-01 7.69878328e-02 6.38093233e-01 9.06673789e-01
1.96439281e-01 5.10120615e-02 7.60894477e-01 -1.01678586e+00
-3.42623919e-01 -5.33495426e-01 -7.65950739e-01 9.68714833e-01
5.15270121e-02 -1.01457584e+00 -4.22904730e-01 3.71978700e-01] | [8.15243911743164, -1.9572609663009644] |
9ddfccec-5f44-452e-a655-4348e3231364 | autoregressive-image-generation-using | 2203.01941 | null | https://arxiv.org/abs/2203.01941v2 | https://arxiv.org/pdf/2203.01941v2.pdf | Autoregressive Image Generation using Residual Quantization | For autoregressive (AR) modeling of high-resolution images, vector quantization (VQ) represents an image as a sequence of discrete codes. A short sequence length is important for an AR model to reduce its computational costs to consider long-range interactions of codes. However, we postulate that previous VQ cannot shorten the code sequence and generate high-fidelity images together in terms of the rate-distortion trade-off. In this study, we propose the two-stage framework, which consists of Residual-Quantized VAE (RQ-VAE) and RQ-Transformer, to effectively generate high-resolution images. Given a fixed codebook size, RQ-VAE can precisely approximate a feature map of an image and represent the image as a stacked map of discrete codes. Then, RQ-Transformer learns to predict the quantized feature vector at the next position by predicting the next stack of codes. Thanks to the precise approximation of RQ-VAE, we can represent a 256$\times$256 image as 8$\times$8 resolution of the feature map, and RQ-Transformer can efficiently reduce the computational costs. Consequently, our framework outperforms the existing AR models on various benchmarks of unconditional and conditional image generation. Our approach also has a significantly faster sampling speed than previous AR models to generate high-quality images. | ['Wook-Shin Han', 'Minsu Cho', 'Saehoon Kim', 'Chiheon Kim', 'Doyup Lee'] | 2022-03-03 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Lee_Autoregressive_Image_Generation_Using_Residual_Quantization_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Lee_Autoregressive_Image_Generation_Using_Residual_Quantization_CVPR_2022_paper.pdf | cvpr-2022-1 | ['conditional-image-generation'] | ['computer-vision'] | [ 3.13895881e-01 -4.87835817e-02 -2.70368513e-02 -8.22728649e-02
-1.15478742e+00 -3.38280678e-01 6.51828885e-01 -3.22372466e-01
-3.83734219e-02 5.17009437e-01 3.08448672e-01 -2.24171072e-01
9.53699350e-02 -1.12600720e+00 -1.10460806e+00 -6.67805552e-01
-7.32481927e-02 8.22139382e-02 1.25086168e-02 -2.14325547e-01
3.43814284e-01 3.71087968e-01 -1.84810865e+00 3.24854672e-01
5.64088523e-01 1.06489277e+00 6.17029607e-01 1.04757559e+00
-9.72751081e-02 8.92486572e-01 -3.24915171e-01 -3.11861068e-01
3.44369054e-01 -6.51700258e-01 -5.60188234e-01 8.49296749e-02
2.47219786e-01 -4.75748688e-01 -6.76610708e-01 1.00511849e+00
1.53088972e-01 1.52776673e-01 8.94124210e-01 -9.37132120e-01
-1.00682902e+00 5.19767225e-01 -6.19165063e-01 -3.97836939e-02
3.29737604e-01 3.13573405e-02 9.81476068e-01 -1.21645081e+00
6.35994971e-01 1.28565025e+00 3.66076440e-01 5.45950174e-01
-1.45551693e+00 -7.11090922e-01 -8.94032046e-02 1.41972423e-01
-1.90127301e+00 -4.79160547e-01 5.78655243e-01 -4.89039749e-01
9.55044627e-01 2.72049367e-01 5.88774145e-01 8.94453406e-01
4.21873540e-01 3.32843602e-01 8.15953910e-01 -4.70907778e-01
3.97416770e-01 -2.12436259e-01 -3.09870392e-01 8.02520752e-01
-3.17760438e-01 2.88769811e-01 -6.59354150e-01 -1.40779838e-01
1.33307564e+00 7.82006606e-02 -2.49259457e-01 -2.71541148e-01
-1.22835529e+00 1.03072786e+00 5.39618254e-01 1.02853350e-01
-5.68140388e-01 5.58075309e-01 1.26128597e-02 9.51516926e-02
-5.07269390e-02 1.81974187e-01 8.22814181e-02 -1.46568373e-01
-9.94560242e-01 1.65322825e-01 1.74042672e-01 1.00299656e+00
9.15485859e-01 3.60167295e-01 -2.62326658e-01 8.80653560e-01
2.58625060e-01 5.69459081e-01 4.35893446e-01 -1.37740159e+00
3.40138555e-01 2.61628889e-02 1.20679989e-01 -1.08052742e+00
2.84045458e-01 -2.63832688e-01 -1.38468254e+00 2.31108129e-01
-2.20558286e-01 3.85273874e-01 -1.05242252e+00 1.69273472e+00
-1.51005208e-01 5.27378321e-01 1.45275041e-01 7.18027711e-01
5.76446414e-01 1.26641023e+00 -2.57544845e-01 -5.12084126e-01
1.28227770e+00 -8.02052259e-01 -7.38925338e-01 7.47987479e-02
2.88183212e-01 -6.64385736e-01 9.87244785e-01 4.03700948e-01
-1.41338241e+00 -1.02295244e+00 -1.12642860e+00 -1.69298023e-01
4.17017043e-02 1.55151263e-01 3.52695465e-01 4.41170931e-01
-1.29915500e+00 4.20567870e-01 -6.50134146e-01 2.81808138e-01
3.88723850e-01 1.68860331e-01 -2.15531200e-01 -3.12058419e-01
-1.15453517e+00 3.95901531e-01 2.84812599e-02 -7.71335140e-02
-1.04915583e+00 -6.51759982e-01 -1.18285251e+00 1.42404452e-01
-1.46266952e-01 -7.13093758e-01 8.94404829e-01 -5.40341735e-01
-1.43567705e+00 5.39673746e-01 -6.28902435e-01 -7.24674523e-01
1.97914466e-01 -2.18308438e-02 -3.70332092e-01 2.08701968e-01
1.11139864e-01 1.14689577e+00 1.23983908e+00 -1.36080444e+00
-4.78845447e-01 -7.77424574e-02 -1.21032342e-01 4.25620526e-02
-1.00711256e-01 -2.24320874e-01 -5.42014539e-01 -7.81068683e-01
2.09439293e-01 -8.44818115e-01 -5.02731681e-01 -6.99017867e-02
-1.28926665e-01 1.01636566e-01 5.21732450e-01 -3.65254819e-01
1.44019020e+00 -2.32819724e+00 2.36428112e-01 1.26859933e-01
2.89859921e-01 1.03359699e-01 -3.11324924e-01 2.43672520e-01
-1.36222407e-01 3.79966825e-01 -2.75129378e-01 -3.07812333e-01
-3.37899327e-01 4.04698163e-01 -8.60030472e-01 1.70843154e-01
3.43542039e-01 1.04902065e+00 -8.11769664e-01 -5.72704315e-01
2.95466572e-01 9.45653081e-01 -7.94543087e-01 2.67701507e-01
-7.02065229e-02 3.69940817e-01 -2.95468420e-01 2.65209556e-01
7.58092880e-01 -5.13593912e-01 -1.35791332e-01 -2.26038069e-01
-1.41557828e-01 -3.21286395e-02 -1.02786016e+00 1.70780051e+00
-7.02447832e-01 6.79857135e-01 -4.71350968e-01 -8.49053025e-01
1.21974003e+00 1.62800089e-01 3.76165658e-01 -8.93837512e-01
-1.99362561e-01 -3.22635956e-02 -4.58173364e-01 5.88914156e-02
7.54782856e-01 4.24705632e-02 -1.53202251e-01 2.83119619e-01
-1.61921605e-01 -3.18005592e-01 -1.05270885e-01 3.08609962e-01
9.74061787e-01 8.58788788e-02 3.16180378e-01 2.46727884e-01
5.69765747e-01 -3.97740901e-01 3.71940583e-01 1.00390637e+00
2.44461671e-01 1.01515079e+00 2.67313093e-01 -4.79031771e-01
-1.49356306e+00 -1.21518683e+00 -1.05523869e-01 7.52130926e-01
4.95023420e-03 -7.51052141e-01 -7.03495264e-01 -1.32998288e-01
-4.11918432e-01 7.59817660e-01 -5.31398952e-01 -1.62135571e-01
-6.65953100e-01 -2.89092749e-01 4.27425504e-01 2.96889573e-01
5.31802297e-01 -9.71093297e-01 -7.10555136e-01 2.63426751e-01
-3.57048541e-01 -1.05797791e+00 -6.36277974e-01 -1.07061677e-01
-6.71892881e-01 -5.28350949e-01 -7.80130625e-01 -6.98658049e-01
6.00146711e-01 3.90237302e-01 1.11724174e+00 3.70347090e-02
-2.32196227e-01 1.39280036e-01 -4.51419055e-01 7.17377588e-02
-5.80643237e-01 -3.66414487e-01 2.46393681e-02 4.37837578e-02
-2.41534393e-02 -6.63886845e-01 -6.27990603e-01 8.51205215e-02
-1.02254081e+00 2.77767658e-01 6.25886261e-01 1.07830942e+00
1.24591064e+00 2.61389226e-01 1.79382235e-01 -4.15929496e-01
3.83241534e-01 -3.05904746e-01 -7.09911585e-01 3.25787254e-02
-4.58783746e-01 2.92523324e-01 7.71983266e-01 -3.13568473e-01
-6.95350647e-01 2.91210532e-01 -3.34897012e-01 -9.89647031e-01
2.96383828e-01 3.32752049e-01 1.44047424e-01 9.06062573e-02
6.07462704e-01 9.40129578e-01 -8.60316083e-02 -2.29592741e-01
6.97848022e-01 5.38339853e-01 8.19387972e-01 -3.56005102e-01
9.40368652e-01 4.30475265e-01 2.11349636e-01 -7.73737729e-01
-5.40325582e-01 -1.16377221e-02 -5.80974877e-01 -1.47393420e-02
9.31114018e-01 -1.24627912e+00 -6.65368795e-01 2.59179354e-01
-1.22365212e+00 -1.37913331e-01 -3.40777189e-01 2.59505332e-01
-1.05094302e+00 2.78258771e-01 -5.78785539e-01 -8.39234471e-01
-2.07618624e-01 -1.47413850e+00 1.31020534e+00 -5.25935227e-03
8.51264000e-02 -4.94076610e-01 -1.36676461e-01 4.26116660e-02
3.02844793e-01 7.35350251e-02 9.26010191e-01 3.73360813e-01
-1.04534543e+00 1.18143864e-01 -2.67398536e-01 3.15575838e-01
5.55150285e-02 1.26967905e-02 -7.80323863e-01 -2.91340947e-01
-1.04318038e-01 -3.33519399e-01 1.00784576e+00 5.57115912e-01
1.59153914e+00 -5.76861560e-01 -1.32604584e-01 7.88489938e-01
1.53509450e+00 2.44675770e-01 1.26022398e+00 9.24199894e-02
6.13859892e-01 5.29398210e-02 8.24113011e-01 7.54778385e-01
4.32708800e-01 8.69804382e-01 5.20033598e-01 1.05603673e-01
-2.05324918e-01 -7.24351168e-01 5.04870951e-01 1.01856339e+00
-8.35250244e-02 -4.12104800e-02 -7.08327889e-01 3.85164887e-01
-1.55216539e+00 -1.12638295e+00 1.00220941e-01 2.32918262e+00
8.08448970e-01 4.91040526e-03 -2.45135501e-01 1.65209249e-01
6.50724053e-01 3.70021462e-01 -4.76756185e-01 -4.63264763e-01
-1.57413810e-01 3.53512764e-01 4.06527132e-01 6.79977655e-01
-8.45162451e-01 8.97307992e-01 7.04113626e+00 9.88322794e-01
-9.71256137e-01 -1.22631691e-01 8.62000644e-01 3.27620655e-02
-5.36820889e-01 -1.37414753e-01 -1.00508463e+00 5.58906376e-01
1.29306221e+00 -2.97341764e-01 7.55448401e-01 8.22304547e-01
5.79687729e-02 2.76348591e-01 -8.93158674e-01 1.35256839e+00
6.14793040e-02 -1.87679482e+00 6.77255988e-01 3.71755451e-01
7.34792054e-01 -3.07696491e-01 6.33007109e-01 3.32113594e-01
3.06438506e-01 -1.45245528e+00 7.57129133e-01 7.07903206e-01
1.42721570e+00 -9.77475286e-01 3.70904028e-01 3.08395267e-01
-1.66051912e+00 -2.59755105e-01 -8.66634607e-01 1.70227304e-01
1.64084196e-01 4.09645408e-01 -5.63788235e-01 4.27948505e-01
6.76907659e-01 7.08045721e-01 -3.38849157e-01 3.70934933e-01
3.40461060e-02 3.36394399e-01 2.86509320e-02 4.57425445e-01
1.98328391e-01 -3.16850573e-01 2.60542095e-01 9.67256844e-01
8.88580203e-01 5.06723762e-01 6.07652329e-02 9.58024383e-01
-3.28719802e-02 -1.75239190e-01 -6.79691374e-01 2.30663102e-02
8.35021079e-01 8.04559410e-01 -3.43225509e-01 -3.62290770e-01
-3.26429337e-01 1.15471005e+00 1.73517972e-01 2.82260656e-01
-9.52626705e-01 -1.75298139e-01 8.20270717e-01 1.33083150e-01
8.01311076e-01 -3.36417198e-01 -5.86809404e-03 -9.43709910e-01
-1.93806052e-01 -9.96705711e-01 1.45153806e-01 -1.22690511e+00
-6.75262213e-01 8.35213602e-01 -3.48671556e-01 -1.71300304e+00
-6.87445402e-01 -1.15087546e-01 3.73539655e-03 9.19379115e-01
-1.53433311e+00 -1.05609918e+00 -3.22122216e-01 7.41288960e-01
7.52327621e-01 -2.37732306e-01 1.12035835e+00 4.85784980e-03
-1.17256597e-01 6.76394820e-01 1.58314139e-01 8.63852501e-02
2.43551180e-01 -8.88684690e-01 6.24392092e-01 7.93463469e-01
4.75566596e-01 6.52146876e-01 6.98181570e-01 -4.38092858e-01
-1.59746766e+00 -1.37271369e+00 5.53076148e-01 -3.95536482e-01
3.02033395e-01 -3.34139794e-01 -1.04869044e+00 6.27709687e-01
-5.21850660e-02 3.92627418e-01 4.14683223e-01 -6.62584841e-01
-6.48291707e-01 -9.85820964e-02 -8.93146813e-01 5.97211003e-01
1.10297811e+00 -7.82949686e-01 -2.46586576e-01 -1.59063973e-02
1.12304163e+00 -5.00968337e-01 -1.15545714e+00 1.87951401e-01
4.83916730e-01 -1.04875982e+00 1.40915775e+00 -5.54625615e-02
1.05038142e+00 -4.28004473e-01 -6.09468341e-01 -1.16954744e+00
-7.21351624e-01 -4.91188079e-01 -3.35708976e-01 1.01842916e+00
7.84191266e-02 -1.69711187e-01 5.77647984e-01 9.76764038e-02
1.26829445e-01 -7.50645041e-01 -1.04220414e+00 -5.95834553e-01
4.05742899e-02 -4.88119215e-01 7.69731760e-01 4.83539164e-01
-4.93251771e-01 1.37483954e-01 -7.61798024e-01 4.71374661e-01
8.65966201e-01 1.67714760e-01 7.55565107e-01 -8.57398391e-01
-2.39622205e-01 -2.40762144e-01 -5.80050826e-01 -1.54574502e+00
1.74155176e-01 -6.78619683e-01 1.01502314e-01 -1.21545529e+00
2.45479777e-01 -4.82789010e-01 -2.60855675e-01 1.23788714e-01
-4.80136052e-02 5.75550020e-01 2.68647760e-01 4.78922486e-01
-4.19544905e-01 8.33639383e-01 1.24649060e+00 -1.72837362e-01
-1.03706397e-01 -3.27165425e-01 -5.49769938e-01 3.71223062e-01
3.74355108e-01 -3.37445945e-01 -5.62774599e-01 -3.13184679e-01
3.52928638e-01 6.57786250e-01 3.48175228e-01 -9.56674933e-01
2.34858394e-01 -1.58532634e-01 6.04659200e-01 -8.63885701e-01
7.41222143e-01 -6.00902319e-01 6.13324106e-01 3.73924404e-01
-5.04522741e-01 3.05081606e-02 -1.72050506e-01 6.70323968e-01
-3.83369625e-01 -1.86424349e-02 9.21095490e-01 -2.71126240e-01
-6.40475392e-01 4.99788404e-01 -3.91777128e-01 -3.40686738e-01
8.38422239e-01 -3.34261864e-01 7.52865225e-02 -6.45872116e-01
-5.49671113e-01 -1.82540834e-01 5.84464848e-01 5.49466789e-01
1.28778386e+00 -1.65045285e+00 -7.78471053e-01 6.11086607e-01
1.02834910e-01 9.49814767e-02 5.03590107e-01 8.39255378e-02
-2.97785878e-01 5.04453063e-01 -2.71608591e-01 -8.33711982e-01
-1.21561372e+00 7.90808678e-01 2.15528667e-01 -3.39268237e-01
-7.57766843e-01 8.72583330e-01 4.67867941e-01 4.16337401e-02
-7.24753439e-02 1.03960093e-02 -4.08249199e-01 -2.07081974e-01
1.02114952e+00 2.23424807e-02 -3.18237096e-01 -1.00638473e+00
1.39108039e-02 1.02902961e+00 -1.16706088e-01 -3.25543642e-01
1.20669389e+00 -4.12413836e-01 -8.12387168e-02 3.36827844e-01
1.33993638e+00 -7.84014240e-02 -1.43421555e+00 -2.78045475e-01
-5.72561085e-01 -8.58348906e-01 1.82822257e-01 -3.51856649e-02
-1.02835393e+00 8.19402039e-01 6.34832561e-01 -1.38634533e-01
1.22674918e+00 8.23088735e-02 6.94648206e-01 1.69463933e-01
7.03766823e-01 -5.94632626e-01 4.86208349e-01 4.05488372e-01
1.13466585e+00 -7.97213614e-01 -6.85169175e-02 -4.50231582e-01
-7.24732459e-01 9.57036018e-01 2.60744601e-01 -1.48522258e-01
5.38271904e-01 2.17794210e-01 -2.13198543e-01 1.40853047e-01
-1.02813625e+00 2.90037561e-02 8.10178965e-02 8.28512192e-01
1.57530904e-01 4.25794572e-02 7.57073611e-02 4.03135061e-01
-6.34161711e-01 1.64962053e-01 7.97583520e-01 5.24269640e-01
-3.84764075e-01 -9.62227404e-01 -4.71450061e-01 3.94277781e-01
-3.27547342e-01 -2.88572311e-01 4.68882769e-01 2.56843150e-01
-3.04795001e-02 7.97389746e-01 3.67790669e-01 -6.53516471e-01
2.46957336e-02 -2.45205864e-01 3.76199067e-01 -2.80371785e-01
5.51850796e-02 1.09608322e-01 -5.18457055e-01 -7.69058228e-01
-2.77007014e-01 -5.01562357e-01 -1.28960836e+00 -4.14354771e-01
4.04138342e-02 4.84214388e-02 6.14246547e-01 3.81764829e-01
6.22588694e-01 5.70167959e-01 1.09668934e+00 -7.64703214e-01
-4.97463822e-01 -4.74569023e-01 -5.22098541e-01 2.57179409e-01
5.88350236e-01 -4.67048615e-01 -2.75371224e-01 4.88794625e-01] | [11.17631721496582, -1.2405914068222046] |
77feda30-bb89-4f95-97d4-dc303970c5cb | diffbev-conditional-diffusion-model-for-bird | 2303.08333 | null | https://arxiv.org/abs/2303.08333v1 | https://arxiv.org/pdf/2303.08333v1.pdf | DiffBEV: Conditional Diffusion Model for Bird's Eye View Perception | BEV perception is of great importance in the field of autonomous driving, serving as the cornerstone of planning, controlling, and motion prediction. The quality of the BEV feature highly affects the performance of BEV perception. However, taking the noises in camera parameters and LiDAR scans into consideration, we usually obtain BEV representation with harmful noises. Diffusion models naturally have the ability to denoise noisy samples to the ideal data, which motivates us to utilize the diffusion model to get a better BEV representation. In this work, we propose an end-to-end framework, named DiffBEV, to exploit the potential of diffusion model to generate a more comprehensive BEV representation. To the best of our knowledge, we are the first to apply diffusion model to BEV perception. In practice, we design three types of conditions to guide the training of the diffusion model which denoises the coarse samples and refines the semantic feature in a progressive way. What's more, a cross-attention module is leveraged to fuse the context of BEV feature and the semantic content of conditional diffusion model. DiffBEV achieves a 25.9% mIoU on the nuScenes dataset, which is 6.2% higher than the best-performing existing approach. Quantitative and qualitative results on multiple benchmarks demonstrate the effectiveness of DiffBEV in BEV semantic segmentation and 3D object detection tasks. The code will be available soon. | ['Xingang Wang', 'Yun Ye', 'Zheng Zhu', 'Jiayu Zou'] | 2023-03-15 | null | null | null | null | ['motion-prediction'] | ['computer-vision'] | [-9.27105919e-03 -3.91031645e-04 -2.57580519e-01 -4.79134530e-01
-3.93980622e-01 -3.66245627e-01 8.05585623e-01 3.54893208e-02
-5.61150968e-01 2.51798749e-01 2.27238685e-01 -2.68659085e-01
1.18901975e-01 -9.54971135e-01 -7.67028868e-01 -5.73523104e-01
5.42129397e-01 2.37756357e-01 6.72299683e-01 -4.76012439e-01
1.83275029e-01 3.69625062e-01 -1.64358866e+00 1.51231149e-02
1.05544174e+00 1.20379508e+00 7.06744790e-01 3.60809624e-01
-2.48191431e-01 4.80939180e-01 -3.40988636e-01 -2.80059665e-01
2.39359602e-01 -1.37283608e-01 -3.99730623e-01 1.22200169e-01
3.26520920e-01 -4.12642956e-01 -4.24388170e-01 1.22059464e+00
3.39617670e-01 4.25335675e-01 7.59624004e-01 -1.06180906e+00
-4.67186838e-01 3.77141118e-01 -6.17266238e-01 3.47652048e-01
-3.59364860e-02 3.85780364e-01 8.97572875e-01 -9.43626523e-01
6.25888526e-01 1.49152374e+00 3.68448228e-01 5.84217012e-01
-9.81954932e-01 -7.80273378e-01 3.93540114e-01 3.96083742e-01
-1.36812770e+00 -4.35308874e-01 9.04179990e-01 -4.89564031e-01
7.34280288e-01 8.16057846e-02 6.90624416e-01 1.04787755e+00
2.64863312e-01 1.09571624e+00 9.63083863e-01 1.32835023e-02
3.47221166e-01 -2.89065577e-02 1.97254583e-01 5.73553503e-01
1.34081379e-01 3.02242339e-01 -3.93561035e-01 3.56868029e-01
4.37835932e-01 6.18606433e-02 -2.11187825e-01 -3.96383405e-01
-8.73169661e-01 9.34958279e-01 7.89114058e-01 5.29767238e-02
-5.73439419e-01 3.38606954e-01 1.40324995e-01 -8.79215077e-02
4.81722176e-01 6.82724938e-02 -1.28165320e-01 -1.67921931e-01
-6.31179094e-01 2.96562999e-01 2.56820977e-01 8.49775910e-01
7.98282981e-01 1.39750943e-01 -3.54758769e-01 9.13058579e-01
5.05450845e-01 7.08915412e-01 3.49300802e-01 -1.11598802e+00
3.89415592e-01 3.98491830e-01 1.24439783e-01 -8.43267620e-01
-7.30389282e-02 -3.25561881e-01 -6.52484059e-01 3.98091197e-01
1.26874065e-02 7.48193040e-02 -1.26097500e+00 1.61721349e+00
4.58674669e-01 3.95539492e-01 1.64052472e-01 1.21946585e+00
8.54392588e-01 7.23981500e-01 5.54343835e-02 2.45785639e-01
1.07346499e+00 -1.01323974e+00 -7.68587351e-01 -6.31666958e-01
3.72217923e-01 -5.36829770e-01 8.97248805e-01 1.60232097e-01
-7.86108613e-01 -8.22317004e-01 -1.20636153e+00 -2.25175858e-01
-1.68651432e-01 -1.12942941e-01 4.95868206e-01 4.19612020e-01
-7.65095234e-01 3.75854343e-01 -9.60579872e-01 -3.14687818e-01
7.61469483e-01 -8.95769596e-02 -4.37245592e-02 -5.07891178e-01
-1.14950013e+00 8.08412254e-01 5.34480155e-01 2.76917130e-01
-1.17672658e+00 -4.54635352e-01 -9.62805331e-01 -1.97691277e-01
5.78388810e-01 -7.40076602e-01 1.23456562e+00 -5.49113631e-01
-1.33813453e+00 3.39890271e-01 -4.01688039e-01 -6.59637630e-01
6.94654107e-01 -2.83010691e-01 -9.93901417e-02 -4.93305586e-02
2.93527424e-01 9.95597243e-01 8.18944693e-01 -1.45009243e+00
-8.75905156e-01 -3.74533147e-01 1.01353906e-01 1.92550242e-01
3.45599692e-05 -6.19324446e-01 -8.74970198e-01 -5.55745065e-01
2.27446422e-01 -8.95975709e-01 -3.52861673e-01 3.92510630e-02
-1.35605946e-01 -3.90238881e-01 8.49343002e-01 -4.22610700e-01
9.25127208e-01 -2.49732590e+00 2.39786729e-01 1.98794641e-02
2.69034356e-01 4.04517531e-01 -1.83920100e-01 -8.16844124e-03
4.54742372e-01 3.84949967e-02 -3.90953928e-01 -5.51713943e-01
1.73421517e-01 5.16411483e-01 -2.94408262e-01 3.62161607e-01
4.20528173e-01 1.14951372e+00 -8.91061306e-01 -3.04016918e-01
4.87368792e-01 5.68588853e-01 -6.29691303e-01 9.55466554e-02
-3.60277265e-01 4.37279016e-01 -7.54476249e-01 4.07904148e-01
7.87604630e-01 1.91702679e-01 -4.62346703e-01 -1.76645622e-01
-2.39216071e-02 8.21537152e-02 -1.04840922e+00 1.68512952e+00
-4.10067856e-01 5.86889088e-01 2.68316180e-01 -7.67498851e-01
1.13037670e+00 -2.28812829e-01 2.57567853e-01 -9.71629858e-01
3.73587221e-01 1.78729996e-01 3.29637788e-02 -3.06978911e-01
6.67733669e-01 -1.87569507e-03 1.46334797e-01 -8.53941590e-02
-2.47501627e-01 -5.38801789e-01 8.60637277e-02 2.82183349e-01
7.30874419e-01 9.98604223e-02 3.01707489e-03 -8.13558549e-02
3.73099893e-01 -3.87821393e-03 7.76942253e-01 6.71740711e-01
-5.18220365e-01 6.33852124e-01 2.33411536e-01 -1.56853721e-01
-7.96015799e-01 -1.01669633e+00 -1.95553020e-01 5.98266900e-01
6.76413476e-01 -2.76509017e-01 -6.31793618e-01 -6.34606719e-01
2.36865610e-01 1.09764278e+00 -5.87674141e-01 -5.20990729e-01
-3.95349860e-01 -4.08426523e-01 2.41164103e-01 7.82003343e-01
7.94440627e-01 -8.55661869e-01 -5.64788580e-01 2.78977603e-01
-1.53821364e-01 -1.45606899e+00 -4.10182387e-01 1.21703945e-01
-6.61085069e-01 -8.17741752e-01 -5.06080210e-01 -4.87230361e-01
2.68515676e-01 5.89430869e-01 7.00058997e-01 2.59558894e-02
1.34553000e-01 2.18020424e-01 -4.77121949e-01 -5.92183173e-01
-3.96696091e-01 6.04769289e-02 -1.55549228e-01 1.34974971e-01
3.46262187e-01 -2.84411222e-01 -6.96081579e-01 2.87716985e-01
-9.68452454e-01 1.10926777e-01 5.96435666e-01 7.56200850e-01
8.35559368e-01 1.91618308e-01 3.38951200e-01 -6.29519343e-01
4.58967388e-01 -4.81077820e-01 -5.70726097e-01 -3.39745075e-01
-6.15796149e-01 -1.07798614e-02 3.32121372e-01 -3.05195212e-01
-1.04058242e+00 1.24389514e-01 -5.61521351e-01 -7.78359294e-01
-1.76242545e-01 4.07114327e-01 -3.41419727e-01 1.76213115e-01
3.75754118e-01 2.16753796e-01 -3.28768604e-02 -4.55775291e-01
5.64851940e-01 6.32850826e-01 5.18954039e-01 -3.40416253e-01
7.11756170e-01 6.47979438e-01 -1.12058781e-01 -9.28879499e-01
-7.89104819e-01 -5.96615314e-01 -3.07355821e-01 -2.67260104e-01
1.04479992e+00 -9.58647430e-01 -1.11152224e-01 5.27661502e-01
-1.08827960e+00 -4.87587005e-01 -2.85412073e-01 4.17728871e-01
-4.05747294e-01 3.98617089e-01 -3.45451713e-01 -7.77042627e-01
2.37431228e-02 -1.49033105e+00 1.25296748e+00 1.88478693e-01
9.87570062e-02 -6.85460389e-01 -3.05821270e-01 5.28444290e-01
2.19587937e-01 1.68426763e-02 5.48375845e-01 -2.72566199e-01
-6.87749863e-01 -6.20007107e-04 -4.13937747e-01 6.01775646e-01
-6.27178401e-02 -2.34674841e-01 -1.12035823e+00 -1.38035983e-01
5.51883802e-02 -9.67939571e-02 1.51347446e+00 4.34221715e-01
1.09381318e+00 2.54836917e-01 -4.07880217e-01 6.89026296e-01
1.18073463e+00 2.18893424e-01 6.38702810e-01 2.48981431e-01
9.47577775e-01 5.98182440e-01 9.92647469e-01 2.65010238e-01
6.78659081e-01 6.32340491e-01 7.86245525e-01 2.89185741e-03
-4.82193947e-01 -3.83596331e-01 2.78244764e-01 5.43321311e-01
1.34029835e-01 -3.72987062e-01 -9.39251304e-01 6.97376311e-01
-1.70715582e+00 -6.69651747e-01 -2.78391182e-01 1.76895177e+00
4.13127273e-01 5.54112971e-01 -2.39613414e-01 -1.74633712e-02
5.06309211e-01 4.71846998e-01 -6.40834868e-01 -2.04176009e-01
3.13157253e-02 -1.42462566e-01 5.97831666e-01 6.61084533e-01
-1.07591844e+00 1.31173289e+00 4.54917908e+00 1.10355937e+00
-1.19168186e+00 1.78019270e-01 5.44344544e-01 7.96759203e-02
-5.58459401e-01 -6.69747740e-02 -9.88107026e-01 5.08169949e-01
4.89861101e-01 1.29299909e-01 2.70787865e-01 7.84797847e-01
4.78874981e-01 -2.38982886e-01 -8.53711367e-01 9.54706430e-01
-1.30294621e-01 -1.13271630e+00 -3.47230956e-02 9.74561423e-02
7.64603078e-01 3.96440387e-01 1.46552190e-01 3.37479949e-01
4.02836472e-01 -8.89236391e-01 1.12624431e+00 5.02658963e-01
3.78861099e-01 -7.74179518e-01 8.28210831e-01 7.20797062e-01
-1.32070291e+00 -2.33550724e-02 -5.32402039e-01 1.32695362e-01
3.99037123e-01 6.94455266e-01 -6.61522150e-01 4.96131778e-01
7.19840884e-01 1.00293148e+00 -6.41823709e-01 9.71285820e-01
-3.69882554e-01 6.15567982e-01 -3.42934430e-01 -1.60417035e-02
5.23936093e-01 -1.72422603e-01 7.71889806e-01 9.96482551e-01
3.67214680e-01 1.77724838e-01 4.05585557e-01 9.62493777e-01
2.96314768e-02 -1.45377189e-01 -5.95841348e-01 -2.06507612e-02
4.31464106e-01 1.01897514e+00 -6.15383625e-01 -3.53656828e-01
-3.44754487e-01 9.48761523e-01 2.21440062e-01 4.53777313e-01
-8.92642736e-01 -3.73835005e-02 8.40022624e-01 -2.22562303e-04
8.48686516e-01 -3.84785920e-01 -3.79326999e-01 -9.35108423e-01
1.57526806e-02 -5.67514896e-01 4.22659284e-03 -6.38176382e-01
-1.20542049e+00 6.13718152e-01 2.83597261e-02 -9.88681972e-01
5.00614231e-04 -3.40381801e-01 -4.02002156e-01 8.85987937e-01
-1.86050105e+00 -1.05739021e+00 -6.17440939e-01 2.05327809e-01
9.08739567e-01 1.30353525e-01 2.11622268e-01 1.50406554e-01
-5.22685826e-01 8.59230086e-02 -7.95479342e-02 -1.58789143e-01
4.17909682e-01 -1.06476867e+00 6.21703088e-01 9.70975578e-01
9.69334096e-02 2.12344050e-01 7.83994675e-01 -7.51930118e-01
-1.23011327e+00 -1.49442613e+00 4.89707857e-01 -4.39504385e-01
4.48115587e-01 -2.11527973e-01 -1.16607273e+00 3.20628405e-01
1.75432563e-02 1.12904824e-01 6.71566417e-03 -3.72499555e-01
-1.60023615e-01 -2.58503854e-01 -8.62578392e-01 6.51043832e-01
1.23258746e+00 -3.76618803e-01 -5.78283966e-01 -1.28661290e-01
9.07173872e-01 -4.88087595e-01 -5.36967754e-01 6.62146449e-01
2.95163184e-01 -1.00250626e+00 9.98890996e-01 -1.76194366e-02
2.91198611e-01 -5.28731942e-01 -3.82399023e-01 -1.41962802e+00
-2.70426124e-01 -9.00934935e-02 -3.76895890e-02 1.27452052e+00
2.34076023e-01 -6.68161809e-01 7.24425316e-01 3.79387468e-01
-4.07245159e-01 -8.69282246e-01 -8.91897440e-01 -6.51431262e-01
1.31625637e-01 -9.84456122e-01 6.68463349e-01 5.40121138e-01
-9.00919676e-01 4.63110209e-01 -1.01555072e-01 3.04811984e-01
4.56083715e-01 1.62168480e-02 8.86215925e-01 -1.09516239e+00
-8.02334324e-02 -6.74690187e-01 -4.46743667e-01 -1.74177861e+00
2.77294248e-01 -7.95841455e-01 4.06819552e-01 -1.82713234e+00
-2.05472857e-01 -4.95818049e-01 -1.80858016e-01 9.86296982e-02
-4.39749807e-01 1.88223690e-01 3.72328758e-01 4.21648026e-02
-4.83480811e-01 1.16038013e+00 1.66422153e+00 -5.11018515e-01
-2.51712739e-01 -3.17396596e-02 -6.34117663e-01 7.37855911e-01
7.88431942e-01 -4.06559855e-01 -6.99088335e-01 -6.57703459e-01
-1.90421671e-01 -2.33746216e-01 4.23291743e-01 -9.29970920e-01
1.62575245e-01 -3.21606338e-01 2.40138486e-01 -8.53329778e-01
5.56962252e-01 -7.93382823e-01 -1.54158682e-01 3.71238112e-01
-4.54305634e-02 -2.03638464e-01 1.86199099e-01 8.63546968e-01
-3.54965091e-01 -2.01180026e-01 8.72736275e-01 5.75230457e-03
-1.38775551e+00 5.09098470e-01 -3.44890445e-01 3.08181364e-02
1.08881950e+00 -2.84058630e-01 -1.29421368e-01 -4.27391142e-01
-2.78615117e-01 5.74731827e-01 5.94092369e-01 7.35437930e-01
9.22790945e-01 -1.22696984e+00 -7.17607439e-01 3.75504464e-01
2.24589899e-01 5.91346085e-01 3.33689123e-01 8.29871237e-01
-2.75110900e-01 1.35745957e-01 1.76328659e-01 -8.99840653e-01
-8.89322579e-01 5.27064145e-01 3.09878558e-01 1.87974349e-01
-7.82463491e-01 7.86714017e-01 5.00119746e-01 -1.71071663e-01
9.68923867e-02 -5.96021056e-01 -3.84689033e-01 4.27043252e-02
3.56881708e-01 2.51145840e-01 -4.78125922e-02 -9.48797762e-01
-3.22233081e-01 5.95632672e-01 -6.52919561e-02 -1.17143899e-01
1.11717677e+00 -4.11141843e-01 2.68076271e-01 3.86821657e-01
1.09825706e+00 5.15372679e-02 -1.75595140e+00 -2.24209577e-01
-2.32706010e-01 -5.24961591e-01 5.42878866e-01 -4.67234671e-01
-1.26997662e+00 1.03910327e+00 6.61830544e-01 1.59281474e-02
9.20760214e-01 2.45850772e-01 1.09015024e+00 2.13647649e-01
3.34016949e-01 -9.24657285e-01 6.11905158e-02 5.95784485e-01
8.51253748e-01 -1.43174541e+00 -2.97455788e-01 -6.74703658e-01
-9.09035861e-01 6.18868887e-01 6.80184245e-01 -1.53683096e-01
5.03359497e-01 4.78964597e-02 1.95235968e-01 -1.77576184e-01
-6.59229100e-01 -5.64786732e-01 5.18586218e-01 5.84289849e-01
-2.43869349e-01 8.89173076e-02 -1.93887368e-01 6.68859422e-01
-2.12477237e-01 -9.25671831e-02 2.48879716e-01 7.69684374e-01
-6.38056099e-01 -8.27836394e-01 -1.74362451e-01 2.60960042e-01
1.51520651e-02 2.08633281e-02 -2.26858810e-01 5.80928385e-01
3.09954882e-01 1.17413163e+00 2.19367892e-02 -5.31719804e-01
6.06873512e-01 -2.73342103e-01 1.73930675e-01 -5.27361274e-01
1.15510505e-02 1.97415426e-01 1.55725017e-01 -7.25638747e-01
-2.78116375e-01 -5.97980797e-01 -1.53547132e+00 -2.43514702e-01
-1.81218058e-01 -1.66661367e-01 6.68527067e-01 1.04245698e+00
3.02323610e-01 7.31432438e-01 5.13123155e-01 -7.95834363e-01
-5.31633973e-01 -8.89620960e-01 -3.39132726e-01 5.19100726e-01
5.13558507e-01 -1.05008411e+00 -2.23860115e-01 -2.99530596e-01] | [8.172842025756836, -2.5296506881713867] |
0d248ebb-8a0b-4e2f-a62d-12615c5b127f | self-attention-based-deep-feature-fusion-for | null | null | https://ieeexplore.ieee.org/abstract/document/8982033 | https://ieeexplore.ieee.org/abstract/document/8982033 | Self-Attention-Based Deep Feature Fusion for Remote Sensing Scene Classification | Remote sensing scene classification aims to assign automatically each aerial image a specific sematic label. In this letter, we propose a new method, called self-attention-based deep feature fusion (SAFF), to aggregate deep layer features and emphasize the weights of the complex objects of remote sensing scene images for remote sensing scene classification. First, the pretrained convolutional neural network (CNN) model is applied to extract the abstract multilayer feature maps from the original aerial imagery. Then, a nonparametric self-attention layer is proposed for spatial-wise and channel-wise weightings, which enhances the effects of the spatial responses of the representative objects and uses the infrequently occurring features more sufficiently. Thus, it can extract more discriminative features. Finally, the aggregated features are fed into a support vector machine (SVM) for classification. The proposed method is experimented on several data sets, and the results prove the effectiveness and efficiency of the scheme for remote sensing scene classification. | ['Ran Cao; Leyuan Fang; Ting Lu; Nanjun He'] | 2022-05-30 | null | null | null | ieee-geoscience-and-remote-sensing-letters-7 | ['remote-sensing-image-classification'] | ['miscellaneous'] | [ 4.98823732e-01 -4.81744051e-01 1.06218070e-01 -6.07912123e-01
-3.77675205e-01 -7.66768828e-02 4.17118073e-01 2.67162055e-01
-4.63930845e-01 4.99136955e-01 2.11439893e-01 -1.09310113e-01
-4.56319094e-01 -9.07454908e-01 -2.82688141e-01 -9.66037452e-01
-1.32939786e-01 -4.50075984e-01 -8.22707117e-02 3.29809003e-02
1.55706927e-01 7.55694389e-01 -1.82748258e+00 3.16613942e-01
1.04068100e+00 1.33113742e+00 6.58943772e-01 4.41673011e-01
-6.69538975e-02 8.45295429e-01 -1.98661372e-01 2.93533444e-01
1.57710478e-01 -2.47679114e-01 -5.17073214e-01 4.09531862e-01
3.81097108e-01 -4.95806336e-01 -3.44063371e-01 1.27021241e+00
4.27900881e-01 3.80800486e-01 6.02904975e-01 -6.90927804e-01
-6.85672760e-01 4.92012739e-01 -5.23252189e-01 6.60226762e-01
-3.29907149e-01 2.08119407e-01 1.05284750e+00 -1.13676298e+00
-1.00877890e-02 9.95646775e-01 4.19926494e-01 -5.55113405e-02
-7.61274636e-01 -4.61620718e-01 6.37734890e-01 2.72047669e-01
-1.48869216e+00 -2.17067018e-01 1.01130903e+00 -4.93478388e-01
6.08162403e-01 3.59274387e-01 6.64093077e-01 4.48269516e-01
1.35957912e-01 7.60164082e-01 9.79332447e-01 -1.40339687e-01
7.88981542e-02 2.09875464e-01 4.69682664e-01 4.17420775e-01
5.89961000e-02 -1.18335746e-01 1.21045567e-01 6.05535740e-03
6.32893384e-01 8.20202529e-01 -4.34311241e-01 1.12551950e-01
-1.02983630e+00 8.17973435e-01 1.15042913e+00 6.26852870e-01
-9.76700783e-01 -1.16783522e-01 2.33756915e-01 -7.52395659e-04
8.14948738e-01 2.49066070e-01 -4.45958495e-01 7.38939345e-01
-8.79578114e-01 9.93243232e-02 -1.22204624e-01 3.79512161e-01
1.22165620e+00 1.38621688e-01 -2.94655830e-01 9.63312924e-01
3.64681840e-01 4.83199954e-01 6.38502300e-01 -4.29813772e-01
1.62410751e-01 9.77204025e-01 8.41997471e-03 -1.41162145e+00
-5.33523917e-01 -7.90770829e-01 -1.22629631e+00 1.49851456e-01
-5.41962504e-01 -2.33917236e-01 -8.84121895e-01 1.27144468e+00
1.97507069e-01 2.12987497e-01 9.98003855e-02 1.23453510e+00
1.00520623e+00 1.06603634e+00 4.74282175e-01 1.21198982e-01
1.09766126e+00 -6.93849564e-01 -4.41929102e-01 -1.91906214e-01
2.72681952e-01 -3.28319341e-01 1.04231787e+00 -4.78142612e-02
-4.56067085e-01 -8.93686771e-01 -1.07362854e+00 2.12633952e-01
-4.97473150e-01 7.31164873e-01 7.58588493e-01 3.34678739e-02
-5.76870739e-01 5.29485583e-01 -4.45909351e-01 -2.26359725e-01
8.34218740e-01 1.69980720e-01 -1.90204501e-01 -5.73257357e-02
-1.03498721e+00 5.57136714e-01 8.19405854e-01 6.38696373e-01
-8.20776999e-01 -5.94163597e-01 -9.55370426e-01 4.47544664e-01
1.55955395e-02 -2.11559400e-01 7.01807439e-01 -1.42909396e+00
-1.13539851e+00 4.48171228e-01 1.19442590e-01 2.46778056e-02
-2.87450496e-02 -1.42135963e-01 -5.64160287e-01 1.50661394e-01
4.22671847e-02 4.23942953e-01 8.81779432e-01 -1.14715517e+00
-1.01060212e+00 -4.18659776e-01 1.20903753e-01 3.60963017e-01
-5.39299726e-01 -2.88846549e-02 2.55120039e-01 -6.73261464e-01
3.51940393e-01 -2.68105119e-01 -4.50302750e-01 -9.69767347e-02
-2.14981318e-01 -1.98880121e-01 9.46402729e-01 -6.06471360e-01
1.01666987e+00 -2.60937142e+00 8.37459788e-02 3.57221514e-01
2.62493938e-01 4.41913605e-01 -4.04138654e-01 1.51567325e-01
-2.25478098e-01 -6.02942109e-02 -4.96080369e-01 2.39378378e-01
-4.34730440e-01 -1.29390135e-01 -3.31284106e-01 4.37277704e-01
6.52505100e-01 7.08460569e-01 -7.65234292e-01 -3.34174305e-01
6.01056755e-01 3.85565758e-01 -3.24973732e-01 2.69041538e-01
-3.76118384e-02 4.70536590e-01 -8.34490776e-01 5.56184113e-01
9.03763592e-01 -2.72885770e-01 -1.60827175e-01 -5.09404480e-01
-3.64465475e-01 -1.33467391e-01 -9.90978718e-01 1.24815643e+00
-5.24863780e-01 4.25295919e-01 -1.27070293e-01 -1.02972877e+00
1.07328665e+00 -1.88682511e-01 4.12334442e-01 -5.75012505e-01
3.62977862e-01 5.36292307e-02 -1.57343075e-01 -7.63700962e-01
4.53413874e-01 1.68951795e-01 6.57835603e-02 6.98774382e-02
6.35798052e-02 2.75207698e-01 -4.63019043e-01 -2.81321973e-01
5.66582859e-01 -1.30805112e-02 4.24921423e-01 -4.53463942e-01
7.79454052e-01 7.26518407e-02 2.85833657e-01 4.46526170e-01
-7.46093458e-03 1.20186985e-01 -3.77673775e-01 -9.90550518e-01
-7.84832656e-01 -5.96136510e-01 -3.54391694e-01 1.33980834e+00
3.65648866e-01 1.95704937e-01 -4.35107380e-01 -6.57819271e-01
1.37981519e-01 4.36581999e-01 -7.19873250e-01 -3.41241032e-01
-2.39598155e-02 -9.79488552e-01 7.79526010e-02 5.45540988e-01
9.92676973e-01 -1.21245372e+00 -7.51034975e-01 3.88213545e-01
3.35121863e-02 -5.47832072e-01 3.98037173e-02 1.37156248e-01
-7.58226931e-01 -7.29838371e-01 -7.41110504e-01 -6.98129117e-01
6.17578685e-01 7.75223255e-01 3.15465957e-01 9.26339924e-02
-2.78221101e-01 -1.29150331e-01 -4.90309685e-01 -2.32818127e-01
2.59690672e-01 1.32804245e-01 -2.66096890e-01 7.06567287e-01
4.38279301e-01 -6.27328873e-01 -7.04101741e-01 -1.11901142e-01
-9.20436382e-01 -2.29156744e-02 9.65863407e-01 9.47237015e-01
4.84958023e-01 4.11603689e-01 4.73892927e-01 -6.67959869e-01
4.43980306e-01 -6.98186696e-01 -3.48469019e-01 1.54466808e-01
-1.85971692e-01 -1.90305039e-01 7.57694483e-01 -2.68214732e-01
-1.13370907e+00 4.11194414e-02 -2.14907471e-02 -3.21098596e-01
-5.43684721e-01 9.92722213e-01 -3.35753918e-01 -1.98699579e-01
5.88608265e-01 5.51791966e-01 -2.81974941e-01 -5.52555084e-01
1.75431088e-01 1.14981496e+00 3.27874571e-01 -9.19649228e-02
5.36649883e-01 2.83468872e-01 -3.77239138e-01 -9.67240036e-01
-1.14848888e+00 -4.63881016e-01 -4.22681659e-01 -2.25619867e-01
9.42436755e-01 -1.17963207e+00 -5.31060576e-01 7.33683288e-01
-8.33263099e-01 -9.00648758e-02 -2.81173706e-01 7.35826433e-01
-1.38387933e-01 1.05020732e-01 -2.93669701e-01 -9.89019096e-01
-3.73031735e-01 -8.45462024e-01 1.07610154e+00 6.17424548e-01
2.70894438e-01 -5.91301560e-01 -2.15244994e-01 -1.70762062e-01
5.57496369e-01 1.78203300e-01 8.55461299e-01 -5.72077394e-01
-4.17439550e-01 -1.41918987e-01 -8.46492589e-01 5.36710978e-01
4.66603011e-01 1.70136198e-01 -1.14652479e+00 -3.91759984e-02
-1.24833062e-02 -2.72443861e-01 1.31633425e+00 5.77382803e-01
1.60194147e+00 -5.13959348e-01 -1.21706523e-01 8.21608245e-01
1.80185795e+00 1.92456901e-01 4.83930975e-01 2.17860535e-01
9.35975492e-01 7.37119913e-01 5.10712385e-01 5.60077131e-01
2.34459594e-01 2.15801120e-01 6.22229993e-01 -5.44632494e-01
3.93255144e-01 1.97262708e-02 -2.80399863e-02 4.47703809e-01
-2.97380179e-01 6.86909258e-02 -6.21279478e-01 4.49337453e-01
-1.75505006e+00 -1.12121153e+00 -4.66627581e-03 1.63365781e+00
3.19785506e-01 -3.19506049e-01 -2.89316863e-01 2.15451643e-01
8.37115347e-01 5.27473390e-01 -6.60550654e-01 2.03121468e-01
-3.18701953e-01 4.61618751e-02 4.57443565e-01 2.40229771e-01
-1.68454754e+00 8.91012907e-01 5.41286993e+00 8.55771244e-01
-1.61258638e+00 -1.09872350e-03 6.87086761e-01 2.66943067e-01
-2.72694498e-01 -2.72034168e-01 -5.31898499e-01 3.87898803e-01
4.42363352e-01 4.65355478e-02 2.48609230e-01 9.78528678e-01
1.99318111e-01 3.86824720e-02 -3.03547770e-01 8.65479290e-01
-7.29803890e-02 -1.22461164e+00 5.29489100e-01 -1.49438098e-01
6.21674538e-01 1.45127982e-01 2.01834276e-01 1.74203053e-01
2.96499848e-01 -7.39224136e-01 6.01659834e-01 8.49046409e-01
6.68712616e-01 -7.98397481e-01 1.00945616e+00 3.15358073e-01
-1.38506854e+00 -6.27032518e-01 -7.07698166e-01 -1.39109954e-01
-3.91813278e-01 6.34838402e-01 -2.19827488e-01 8.62863243e-01
7.06834018e-01 1.25623214e+00 -6.79325044e-01 9.87419784e-01
1.98136970e-01 4.97485071e-01 -1.36510447e-01 -1.14826597e-01
6.82564020e-01 -1.72558248e-01 2.42799997e-01 1.13851261e+00
3.73818547e-01 4.17602628e-01 4.08626944e-01 7.52177417e-01
1.85417995e-01 3.63189369e-01 -5.56125343e-01 -1.30193472e-01
2.58524895e-01 1.56159306e+00 -4.92029101e-01 -3.85629654e-01
-2.40768850e-01 8.85666072e-01 1.93890527e-01 5.20536244e-01
-5.43287337e-01 -6.16375387e-01 5.17976999e-01 -3.18194538e-01
4.66326237e-01 5.93285002e-02 -1.20165654e-01 -1.24244404e+00
-2.11849913e-01 -5.78588843e-01 3.55647892e-01 -1.04578829e+00
-1.36207068e+00 7.78980315e-01 -2.45896146e-01 -1.32933247e+00
4.54865843e-01 -5.61114907e-01 -8.20529521e-01 1.12571919e+00
-1.96543157e+00 -1.26833308e+00 -9.28061485e-01 6.54375732e-01
4.17090416e-01 -2.57066458e-01 7.16951847e-01 1.74411044e-01
-6.66948497e-01 1.07234262e-01 2.10292473e-01 2.23664641e-01
1.21358894e-01 -8.79085362e-01 -8.29952881e-02 9.34622586e-01
-1.89728454e-01 3.01706076e-01 1.83162794e-01 -3.50153536e-01
-9.41115320e-01 -1.78309250e+00 6.78800285e-01 3.50444645e-01
3.48470658e-01 5.25536649e-02 -1.03621471e+00 4.65637088e-01
-1.48113415e-01 2.30030730e-01 8.17205846e-01 -1.80944741e-01
-2.90372491e-01 -5.48497558e-01 -1.08637667e+00 2.86484569e-01
6.47501886e-01 -5.01816332e-01 -5.29330492e-01 3.35025877e-01
7.51108706e-01 1.24706984e-01 -7.35099018e-01 5.06320000e-01
3.89934868e-01 -6.68904603e-01 8.38573277e-01 -7.27337182e-01
6.72786236e-01 -5.04065812e-01 -6.25522316e-01 -1.36664903e+00
-1.00878847e+00 2.64621228e-01 6.20331466e-01 1.00031602e+00
1.71506658e-01 -5.99691868e-01 1.87576294e-01 -8.37655067e-02
-3.29458743e-01 -4.89678174e-01 -5.52546322e-01 -1.95205793e-01
-2.69727677e-01 -3.69926691e-02 9.83699739e-01 1.25363743e+00
-4.76252794e-01 2.98559099e-01 -2.54470706e-01 5.06285012e-01
3.63011748e-01 4.16174501e-01 3.01038861e-01 -1.45283353e+00
9.56600383e-02 -5.85519850e-01 -5.98739803e-01 -7.64132917e-01
1.54514879e-01 -9.33727503e-01 1.41993714e-02 -1.41060543e+00
2.17508316e-01 -4.84647512e-01 -7.62760162e-01 6.54371738e-01
-6.12248421e-01 2.10961804e-01 6.18810654e-02 2.91402012e-01
-4.44169998e-01 9.14636910e-01 1.11552632e+00 -5.18982530e-01
-2.48572513e-01 9.54793021e-02 -7.95520067e-01 5.91225922e-01
7.90999472e-01 -1.30433276e-01 -1.46946624e-01 -5.38136482e-01
-2.75878936e-01 -3.01287621e-01 7.51989543e-01 -1.19840813e+00
5.97952493e-02 -3.29661101e-01 8.04003477e-01 -7.17446506e-01
-3.84959072e-04 -1.09470212e+00 8.03823769e-03 5.08427918e-01
-3.77474874e-01 -4.93077755e-01 2.40752891e-01 4.56420898e-01
-3.44605118e-01 -4.99131717e-02 7.31635869e-01 -1.31520107e-01
-1.13191223e+00 6.72689140e-01 -3.63057405e-01 -6.03874803e-01
9.83208001e-01 -1.26756519e-01 -5.92139959e-02 9.93000865e-02
-6.84508383e-01 1.56041488e-01 -1.57575663e-02 1.53649911e-01
7.79330552e-01 -1.45713270e+00 -1.01434565e+00 6.92800701e-01
6.18332386e-01 -5.21349944e-02 8.97876740e-01 5.40509880e-01
-4.28072691e-01 5.46664894e-02 -4.15295243e-01 -6.76675558e-01
-8.16067994e-01 4.08012301e-01 5.64506769e-01 8.19356367e-02
-5.82508326e-01 7.26089180e-01 2.49041483e-01 -3.34501505e-01
-2.33716086e-01 -4.08178419e-01 -7.59665191e-01 2.44233534e-01
8.09709013e-01 3.34777348e-02 -6.33996679e-03 -8.37604523e-01
-4.11090046e-01 6.50342941e-01 -1.54383155e-02 3.55508775e-01
1.66881335e+00 -1.12031996e-01 -1.74750447e-01 3.78756911e-01
1.10009348e+00 -2.69433558e-01 -1.37044299e+00 -6.74838543e-01
-4.33138043e-01 -4.29379344e-01 5.65875232e-01 -6.34168565e-01
-1.21424329e+00 1.08709764e+00 8.29395473e-01 1.16252191e-01
1.38594377e+00 -2.38803580e-01 3.26620579e-01 5.46843827e-01
-1.11573189e-02 -6.01420462e-01 -4.83636945e-01 5.20184934e-01
7.58524716e-01 -1.24001873e+00 -1.16597645e-01 5.87187819e-02
-5.56943834e-01 1.07448137e+00 4.70107645e-01 -5.21042824e-01
9.47828293e-01 -1.55652121e-01 1.18224472e-02 -2.92671889e-01
-2.47129694e-01 -5.75543702e-01 3.64265352e-01 4.34726477e-01
9.37103853e-02 3.34097415e-01 1.56421494e-02 8.13584864e-01
2.00626016e-01 -4.23001498e-03 -5.28970771e-02 6.85148716e-01
-9.51852798e-01 -1.95694640e-01 -3.23528945e-02 7.40762234e-01
-2.15863839e-01 -2.87683815e-01 -2.16174737e-01 3.69106382e-01
3.34885627e-01 7.74251461e-01 3.49283248e-01 -7.70444691e-01
2.53993541e-01 -1.57346681e-01 -5.39306886e-02 -7.02692807e-01
-7.34602749e-01 -4.32021022e-02 -4.60058540e-01 -3.19468170e-01
-7.35546887e-01 -3.39255691e-01 -8.45692635e-01 -8.08026269e-02
-3.81366163e-01 1.98502943e-01 5.32101095e-01 9.77791607e-01
3.70973855e-01 8.11409652e-01 1.27586210e+00 -1.24070883e+00
-5.11727333e-01 -1.19572711e+00 -8.21269333e-01 3.37744147e-01
5.75925589e-01 -5.75222909e-01 -5.89472294e-01 -1.54672638e-02] | [9.862548828125, -1.4758330583572388] |
6d294856-e8f5-44b2-8be0-f6a8bf00b8d7 | community-recovery-in-hypergraphs | 1709.03670 | null | http://arxiv.org/abs/1709.03670v1 | http://arxiv.org/pdf/1709.03670v1.pdf | Community Recovery in Hypergraphs | Community recovery is a central problem that arises in a wide variety of
applications such as network clustering, motion segmentation, face clustering
and protein complex detection. The objective of the problem is to cluster data
points into distinct communities based on a set of measurements, each of which
is associated with the values of a certain number of data points. While most of
the prior works focus on a setting in which the number of data points involved
in a measurement is two, this work explores a generalized setting in which the
number can be more than two. Motivated by applications particularly in machine
learning and channel coding, we consider two types of measurements: (1)
homogeneity measurement which indicates whether or not the associated data
points belong to the same community; (2) parity measurement which denotes the
modulo-2 sum of the values of the data points. Such measurements are possibly
corrupted by Bernoulli noise. We characterize the fundamental limits on the
number of measurements required to reconstruct the communities for the
considered models. | ['Kwangjun Ahn', 'Kangwook Lee', 'Changho Suh'] | 2017-09-12 | null | null | null | null | ['face-clustering'] | ['computer-vision'] | [ 7.45473921e-01 6.39845580e-02 -1.40870795e-01 -6.39102235e-02
-2.11671099e-01 -4.16073769e-01 6.12303495e-01 7.85133243e-01
-5.44662178e-01 4.99506712e-01 -2.49495819e-01 -2.24813849e-01
-1.21292077e-01 -7.59100318e-01 -5.74520051e-01 -1.25138140e+00
-3.38049978e-01 7.65409648e-01 7.49568939e-02 1.72011971e-01
2.69350857e-01 4.97429550e-01 -1.20686722e+00 -3.56463164e-01
3.96363884e-01 7.81060874e-01 -6.53712526e-02 6.74021661e-01
3.46858683e-03 6.77494183e-02 -8.00491154e-01 7.15643466e-02
2.75106847e-01 -5.13242126e-01 -4.82073069e-01 6.23600304e-01
-2.72307873e-01 4.07532632e-01 1.55897677e-01 1.53720927e+00
3.15039158e-01 -1.76325992e-01 6.72560930e-01 -1.50493848e+00
-1.51173295e-02 6.20229602e-01 -9.04621303e-01 1.37349382e-01
9.94351730e-02 -3.88659120e-01 6.73108935e-01 -2.39308223e-01
4.70146805e-01 8.56112778e-01 2.12530911e-01 1.71016455e-01
-1.55221915e+00 -4.57350194e-01 -4.34850633e-01 1.64712802e-01
-1.62037873e+00 -3.76169771e-01 6.09421194e-01 -7.96344161e-01
7.61179328e-02 1.33833691e-01 3.34205985e-01 5.46381772e-01
-1.92423254e-01 1.32663576e-02 1.07257986e+00 -7.17459917e-01
7.47315228e-01 2.63912007e-02 -3.64065426e-03 2.47681767e-01
8.00504684e-01 -3.01670283e-01 -2.02450588e-01 -4.98609811e-01
4.84117001e-01 7.47157484e-02 -3.16373378e-01 -6.94338918e-01
-1.43335593e+00 1.06256449e+00 -2.13301834e-02 4.61773515e-01
-4.10203546e-01 -1.89474765e-02 1.34967759e-01 4.85157818e-01
2.21264958e-02 6.49439543e-02 9.08052921e-02 2.99924791e-01
-8.58305693e-01 -1.02361649e-01 7.81586766e-01 8.28079760e-01
1.01285219e+00 -4.71911699e-01 3.27490121e-01 3.18626285e-01
1.82753265e-01 5.40865004e-01 -4.22118455e-02 -8.53345990e-01
4.14644301e-01 4.56103593e-01 1.78315535e-01 -1.10611212e+00
-1.29152372e-01 -2.55274653e-01 -1.28050792e+00 2.18861565e-01
7.82470703e-01 -1.57828227e-01 -3.75767916e-01 1.87922585e+00
5.00660419e-01 2.93029189e-01 1.52055725e-01 8.29053879e-01
1.65019378e-01 1.62866041e-01 -4.35281754e-01 -6.95005834e-01
1.09019434e+00 -2.26537466e-01 -2.86359727e-01 1.34933472e-01
3.35374057e-01 -6.67635798e-01 -1.98518319e-04 5.75977802e-01
-8.08427393e-01 -2.24532768e-01 -1.04804850e+00 5.57776511e-01
-2.21146028e-02 9.91757810e-02 1.09960236e-01 8.16562772e-01
-8.48783910e-01 3.48406166e-01 -6.71031177e-01 -4.11140561e-01
-6.05612844e-02 3.49140346e-01 -4.91136104e-01 -2.38221854e-01
-5.96386075e-01 2.40846723e-01 2.25348547e-01 1.56834051e-01
-3.69701356e-01 1.63454220e-01 -5.42393744e-01 3.17647569e-02
4.90023881e-01 -3.60684365e-01 5.07332563e-01 -8.01495969e-01
-7.22130239e-01 9.96321917e-01 -3.34900111e-01 -3.80803823e-01
5.11137009e-01 5.87263763e-01 -1.43604204e-01 3.54195148e-01
2.68878788e-01 1.53296620e-01 9.00175035e-01 -8.97675574e-01
-3.48098516e-01 -6.95535839e-01 -2.25869015e-01 -1.54561296e-01
-2.87222881e-02 -1.16014434e-02 -2.56963521e-01 -9.73285586e-02
5.93943119e-01 -1.22564363e+00 -4.17494237e-01 -3.08915507e-02
-6.88865006e-01 8.66418630e-02 7.27062762e-01 -1.37146577e-01
6.66866779e-01 -2.10966015e+00 6.90479696e-01 6.20009065e-01
4.34978634e-01 -1.69564760e-03 1.97448693e-02 6.81748271e-01
-6.41249940e-02 2.15890273e-01 -4.41598833e-01 -4.36481744e-01
-1.82599232e-01 2.11004362e-01 1.73757344e-01 1.09531975e+00
-6.42232150e-02 8.61232579e-02 -6.59824073e-01 -2.37615198e-01
4.39465903e-02 3.58402669e-01 -2.04406261e-01 -1.07857093e-01
1.42017469e-01 8.41013193e-01 -5.05160809e-01 1.19463213e-01
8.60413373e-01 -4.69592363e-01 6.21550024e-01 1.43809661e-01
-3.15518752e-02 -8.57044533e-02 -1.73366344e+00 1.19825482e+00
3.40124190e-01 4.82716084e-01 3.68993998e-01 -1.34497380e+00
7.36538410e-01 5.65736115e-01 8.25533032e-01 3.65063362e-02
2.75367200e-01 6.70173839e-02 2.70985365e-01 -2.34763473e-01
2.53371239e-01 -1.34989709e-01 -1.46692246e-01 6.11122370e-01
-3.82743359e-01 3.07132035e-01 3.71266365e-01 1.56701118e-01
1.42017353e+00 -9.60576952e-01 5.46861947e-01 -1.94943622e-01
6.35015845e-01 -5.36336005e-01 7.77965128e-01 5.85066974e-01
-2.13150069e-01 5.75327456e-01 9.07506883e-01 3.72651994e-01
-1.20119631e+00 -7.15111196e-01 -2.21763924e-01 2.83199191e-01
4.43936467e-01 -2.32312709e-01 -6.49841368e-01 -1.82944268e-01
-1.74403249e-03 -2.14747596e-03 -6.59742534e-01 5.06356694e-02
-3.65364365e-02 -9.39172626e-01 2.24618524e-01 -8.46235156e-02
4.95584548e-01 -6.77751184e-01 -7.25067675e-01 -1.13335242e-02
-2.79763103e-01 -1.22221863e+00 -2.99994439e-01 1.05750479e-01
-8.04110825e-01 -1.46278799e+00 -4.62283045e-01 -5.39621651e-01
9.37776983e-01 4.08140302e-01 5.40971398e-01 2.73327589e-01
-1.37433320e-01 5.13395131e-01 -5.67947924e-01 -1.87763661e-01
-6.38285339e-01 -1.06102906e-01 8.59064087e-02 5.54175138e-01
4.08248246e-01 -4.91565347e-01 -2.80786008e-01 2.94177234e-01
-1.00855637e+00 -3.20140630e-01 2.74361789e-01 4.22716737e-01
5.03856242e-01 5.55946827e-01 1.97591558e-01 -7.07370222e-01
3.87661934e-01 -7.78225303e-01 -5.64863384e-01 1.06758684e-01
-2.40151271e-01 8.89810324e-02 2.94568300e-01 -4.28588092e-01
4.10949029e-02 1.87096596e-01 4.96546775e-01 -1.61557421e-01
-2.65872121e-01 6.41701162e-01 -2.22836733e-01 -2.43236527e-01
2.47354478e-01 1.91706106e-01 2.36931160e-01 -3.13047647e-01
-2.96217743e-02 7.41668224e-01 2.97241986e-01 -3.29040289e-01
6.01998031e-01 5.52331984e-01 7.87506402e-01 -1.30994654e+00
4.52535301e-02 -8.68879795e-01 -7.32491076e-01 -7.38297552e-02
7.65865862e-01 -5.65950692e-01 -9.50970769e-01 4.40343976e-01
-9.09295082e-01 -1.95607580e-02 3.24313454e-02 5.45446932e-01
-2.95963883e-01 8.17164183e-01 -3.17126662e-01 -1.01461554e+00
1.24532908e-01 -1.24707639e+00 8.23656261e-01 -7.52132833e-02
7.24473521e-02 -7.97457635e-01 8.50706622e-02 1.30712181e-01
7.01956451e-02 6.28728688e-01 9.34820533e-01 -5.98690510e-01
-5.19501626e-01 -5.33583999e-01 1.07103243e-01 1.46118835e-01
1.28280908e-01 -1.44385740e-01 -3.34426731e-01 -5.17993808e-01
2.07405299e-01 1.44301832e-01 5.04239678e-01 3.80601496e-01
6.45193338e-01 -8.84234682e-02 -5.25282741e-01 9.19388831e-02
1.20400286e+00 1.40823156e-01 5.83744526e-01 -1.40464813e-01
4.82897937e-01 6.16100609e-01 1.71297222e-01 5.48287153e-01
3.97313610e-02 8.49082887e-01 5.22725523e-01 1.72551244e-01
5.03849447e-01 3.32025617e-01 -5.99243306e-02 6.34785473e-01
4.94552813e-02 -5.15747905e-01 -6.37361228e-01 5.71166158e-01
-1.90227854e+00 -8.40791702e-01 -4.49465930e-01 2.74200630e+00
4.59080577e-01 -1.46394178e-01 3.03502113e-01 7.48072684e-01
1.39095902e+00 -1.31403115e-02 -4.20736372e-01 1.50484636e-01
-1.42097741e-01 -1.19346818e-02 7.13850021e-01 3.16206157e-01
-8.03404033e-01 1.94857866e-02 5.28565025e+00 3.09648901e-01
-1.13001788e+00 -1.94052756e-02 3.59415174e-01 1.98305458e-01
1.06802061e-02 3.46646965e-01 -5.44881046e-01 7.67301440e-01
7.72744834e-01 1.19971540e-02 3.39801013e-01 2.59815574e-01
3.47190872e-02 -5.38579047e-01 -1.06163669e+00 8.34984243e-01
-1.28352139e-02 -8.85060132e-01 -2.52822220e-01 7.98844337e-01
6.80735111e-01 -1.24533378e-01 -1.74229406e-02 -5.86440086e-01
1.19192429e-01 -7.26965964e-01 4.67547834e-01 3.52414906e-01
7.48303771e-01 -5.76173663e-01 5.25229573e-01 7.38331974e-01
-1.05635512e+00 -7.22589344e-02 -3.54033768e-01 -9.20674950e-02
1.80819944e-01 9.83810961e-01 -8.39698076e-01 3.96507978e-01
1.09635472e-01 4.30257857e-01 -1.18275411e-01 1.32989120e+00
-2.54081213e-03 8.09420407e-01 -4.75798398e-01 6.33941889e-02
-1.75909385e-01 -7.26035953e-01 7.15366006e-01 6.07142389e-01
5.61124921e-01 1.45364121e-01 4.18229491e-01 6.11651957e-01
-2.17558444e-01 3.69577482e-02 -4.92803842e-01 -1.24462813e-01
7.70023763e-01 9.94728744e-01 -1.05422533e+00 -1.69560403e-01
-4.14356887e-01 8.64873230e-01 6.18955232e-02 2.32765198e-01
-1.65718123e-01 -1.99824840e-01 5.01685143e-01 3.68817300e-01
6.17121875e-01 -5.20734370e-01 -1.70025870e-01 -1.09076488e+00
5.53900115e-02 -4.86792475e-01 6.85499907e-02 -1.87854469e-01
-9.65319276e-01 1.63041335e-02 -9.77699757e-02 -1.04015553e+00
-3.55038404e-01 -1.36235669e-01 -5.49424231e-01 8.03572774e-01
-1.20909977e+00 -5.73431373e-01 -1.66712254e-01 4.41362441e-01
-4.03567046e-01 -1.14778042e-01 8.06781411e-01 2.29936257e-01
-5.28380275e-01 1.16551705e-01 5.71190894e-01 3.14844191e-01
2.69555569e-01 -1.03898585e+00 1.29954860e-01 9.11182404e-01
2.21881807e-01 6.14529490e-01 8.46515656e-01 -6.37182951e-01
-1.01161456e+00 -6.42750144e-01 1.07167780e+00 5.76745644e-02
2.91875452e-01 -3.73707533e-01 -6.27222657e-01 4.56616610e-01
-2.54196227e-01 1.73295945e-01 7.08129108e-01 -2.52689689e-01
-2.07885236e-01 1.81478888e-01 -1.37890327e+00 1.54043466e-01
4.46952581e-01 -3.26818019e-01 1.15992054e-01 2.23212883e-01
-1.07503697e-01 4.84152138e-02 -9.03786004e-01 2.38732830e-01
1.43233582e-01 -1.03484118e+00 7.07663774e-01 -3.03370267e-01
-3.75002921e-02 -7.47655690e-01 -2.96385854e-01 -9.11278903e-01
-8.69153067e-02 -5.42520940e-01 1.46901518e-01 8.52098584e-01
-5.32901026e-02 -5.99599898e-01 9.39479232e-01 2.14996263e-01
6.63081110e-01 -2.52784252e-01 -1.35086524e+00 -4.35730308e-01
-1.98549792e-01 -6.41212538e-02 5.38285732e-01 1.10804534e+00
2.26856582e-02 3.57493877e-01 -3.05208534e-01 3.89248252e-01
9.83843446e-01 9.78153292e-03 8.39997888e-01 -1.79083467e+00
-4.89468306e-01 -1.64218083e-01 -9.73096192e-01 -8.31574798e-01
-3.49256918e-02 -7.43855655e-01 4.58944254e-02 -1.00527203e+00
4.18919027e-01 -7.42986500e-01 1.80268466e-01 -9.04869381e-03
1.64186999e-01 2.51760662e-01 1.55240089e-01 3.17300469e-01
-4.59312469e-01 3.18905301e-02 6.95121884e-01 1.21813044e-01
-1.25654668e-01 5.21783352e-01 -5.14838099e-01 3.89444917e-01
5.97449183e-01 -5.70030510e-01 -3.03295106e-01 8.10494274e-02
3.55059594e-01 6.21305943e-01 4.11044091e-01 -1.12279844e+00
4.46036309e-01 8.14916845e-03 -5.71488291e-02 -2.20154539e-01
3.84483695e-01 -9.72035885e-01 5.74455202e-01 4.60719258e-01
-3.22740108e-01 -1.39262229e-01 -4.29868072e-01 1.03260517e+00
-5.41781411e-02 -5.81607163e-01 8.54797840e-01 -3.85993794e-02
-1.08869992e-01 1.58468083e-01 -4.53002453e-01 -2.77135104e-01
1.26310992e+00 -3.43374431e-01 -9.10568833e-02 -7.00822890e-01
-9.14655805e-01 -6.95265830e-02 6.51054919e-01 -1.08321458e-01
2.46730626e-01 -1.10603905e+00 -8.03159893e-01 3.93692911e-01
1.20641269e-01 4.01226878e-02 -4.85553443e-02 1.15213370e+00
-1.75267100e-01 2.99775302e-01 5.89368083e-02 -8.86058450e-01
-1.68570685e+00 5.47158957e-01 1.03861883e-01 2.50254683e-02
-2.13606864e-01 2.33732745e-01 -1.61392704e-01 6.00932874e-02
3.32391635e-02 5.38152270e-02 -1.30993381e-01 -9.62586422e-03
4.84005809e-01 5.18444777e-01 1.22192129e-01 -1.16989470e+00
-9.07290354e-02 6.76190794e-01 7.90433362e-02 -3.59014422e-01
1.00785100e+00 -2.45383158e-01 -5.05055368e-01 5.34039795e-01
1.16573715e+00 1.56130120e-01 -8.56781304e-01 -5.04957855e-01
3.52736898e-02 -3.20559293e-01 -2.69092709e-01 -2.93545015e-02
-9.05887544e-01 7.27358222e-01 4.09375131e-01 5.36782324e-01
1.00132465e+00 1.72561273e-01 2.97879487e-01 2.42088750e-01
6.61756098e-01 -6.45248950e-01 -1.49365276e-01 1.72016591e-01
2.32005388e-01 -9.89426434e-01 -1.52452514e-01 -4.16693807e-01
-1.20336592e-01 1.05774295e+00 -2.61214733e-01 -1.42787009e-01
7.14637220e-01 9.38662589e-02 -4.08687323e-01 -8.77731293e-02
-2.50936300e-01 -2.99310088e-01 -6.91923965e-03 6.77096069e-01
6.95241466e-02 2.41750702e-01 -4.85112399e-01 -1.77826472e-02
2.13116053e-02 -8.83945003e-02 9.34601545e-01 8.95996094e-01
-5.51177979e-01 -1.30054021e+00 -7.31096029e-01 3.47898275e-01
-4.04379547e-01 2.20190704e-01 -4.29200351e-01 3.57652843e-01
1.78292945e-01 1.19916666e+00 8.39027613e-02 -1.52762040e-01
-9.90687236e-02 -1.85598716e-01 4.66097564e-01 -7.45802641e-01
4.61626090e-02 1.83223590e-01 -3.18320632e-01 1.78540871e-02
-8.39772999e-01 -9.77929294e-01 -1.08434319e+00 -5.40011048e-01
-3.83995801e-01 5.54559529e-01 8.21843207e-01 1.12063742e+00
1.28258020e-01 -4.79955263e-02 7.31772065e-01 -5.54313004e-01
-4.54105288e-01 -9.25097883e-01 -1.19368339e+00 3.71661782e-01
6.24703288e-01 -4.71051037e-01 -5.86341202e-01 5.87321632e-02] | [6.982907295227051, 5.1568098068237305] |
e4745659-d11d-4512-84ba-7cc1a3969811 | differentiable-multi-fidelity-fusion | 2306.06904 | null | https://arxiv.org/abs/2306.06904v1 | https://arxiv.org/pdf/2306.06904v1.pdf | Differentiable Multi-Fidelity Fusion: Efficient Learning of Physics Simulations with Neural Architecture Search and Transfer Learning | With rapid progress in deep learning, neural networks have been widely used in scientific research and engineering applications as surrogate models. Despite the great success of neural networks in fitting complex systems, two major challenges still remain: i) the lack of generalization on different problems/datasets, and ii) the demand for large amounts of simulation data that are computationally expensive. To resolve these challenges, we propose the differentiable \mf (DMF) model, which leverages neural architecture search (NAS) to automatically search the suitable model architecture for different problems, and transfer learning to transfer the learned knowledge from low-fidelity (fast but inaccurate) data to high-fidelity (slow but accurate) model. Novel and latest machine learning techniques such as hyperparameters search and alternate learning are used to improve the efficiency and robustness of DMF. As a result, DMF can efficiently learn the physics simulations with only a few high-fidelity training samples, and outperform the state-of-the-art methods with a significant margin (with up to 58$\%$ improvement in RMSE) based on a variety of synthetic and practical benchmark problems. | ['Wei W. Xing', 'Wang Kang', 'Yuwen Deng'] | 2023-06-12 | null | null | null | null | ['architecture-search'] | ['methodology'] | [-6.24636188e-03 -5.87021410e-01 -5.36043420e-02 -3.33937049e-01
-8.98205101e-01 -2.88128942e-01 3.83847624e-01 -3.48890498e-02
-3.67397070e-01 1.06460738e+00 -4.11893666e-01 -4.48296010e-01
-4.35882419e-01 -7.32339323e-01 -1.01476729e+00 -7.80513406e-01
-7.36867785e-02 5.17918229e-01 -4.93582115e-02 -2.19380334e-01
9.53850895e-02 7.44319856e-01 -1.29718912e+00 -3.25075805e-01
1.24034178e+00 1.17988622e+00 2.46006951e-01 2.78046191e-01
1.25672340e-01 4.34230626e-01 -3.54440957e-01 -2.30058841e-02
2.85652995e-01 -2.76901513e-01 -5.09049773e-01 -6.44141436e-01
1.63468510e-01 -2.42453396e-01 -6.83733106e-01 9.61327434e-01
7.71026015e-01 4.64392602e-01 6.55940950e-01 -9.74889219e-01
-5.30379951e-01 2.41159052e-01 -4.09752905e-01 1.93813890e-01
-2.29935974e-01 6.47378087e-01 4.92166638e-01 -9.30133104e-01
1.64533466e-01 1.01999021e+00 1.03700328e+00 4.62182850e-01
-1.37358892e+00 -1.11173868e+00 -1.08374983e-01 2.41998807e-01
-1.50453699e+00 -3.99881482e-01 7.00656950e-01 -3.77025366e-01
9.27879095e-01 9.87947080e-03 5.73658645e-01 1.22793639e+00
3.78581733e-01 1.83747232e-01 1.15901351e+00 -2.27237605e-02
3.09300959e-01 6.03088029e-02 1.03628285e-01 4.83660609e-01
2.03854248e-01 6.05432868e-01 -2.48487204e-01 -2.62173086e-01
1.05660546e+00 -2.39443317e-01 -3.61469805e-01 -2.36599043e-01
-1.14884984e+00 7.37935245e-01 6.33696139e-01 -1.31725222e-01
-4.20514405e-01 3.88569564e-01 3.33703816e-01 2.76957780e-01
1.03737466e-01 9.92733359e-01 -5.97035408e-01 -2.02675447e-01
-8.75080168e-01 5.93104601e-01 7.16377199e-01 6.42449975e-01
6.27963841e-01 5.95529735e-01 -1.51643101e-02 7.85483837e-01
2.50500888e-01 7.33497500e-01 2.45113909e-01 -9.61024284e-01
4.28600192e-01 3.84139061e-01 3.67794067e-01 -9.16655362e-01
-4.92069155e-01 -9.51369405e-01 -1.33230877e+00 2.99067289e-01
2.57487684e-01 -2.07106814e-01 -8.88830006e-01 1.83378172e+00
3.64629418e-01 4.56120461e-01 -1.48941442e-01 8.87181342e-01
6.44585490e-01 9.17428970e-01 1.09054126e-01 -7.96284750e-02
8.08947325e-01 -9.16711569e-01 -2.94342607e-01 -1.60707772e-01
3.06056261e-01 -5.40830851e-01 1.23316455e+00 3.29865664e-01
-1.25323319e+00 -7.56023705e-01 -1.26003313e+00 1.19008318e-01
-2.55091876e-01 6.76698908e-02 6.02340221e-01 3.60099107e-01
-7.49699950e-01 1.16324627e+00 -8.82309020e-01 -1.86687335e-02
4.79438156e-01 6.91795826e-01 7.63907507e-02 8.83499309e-02
-1.36512887e+00 9.10869837e-01 3.68505388e-01 2.84795761e-01
-1.10342646e+00 -1.44286132e+00 -4.58282828e-01 2.43115246e-01
1.78232461e-01 -8.73656511e-01 1.07466984e+00 -7.14599073e-01
-1.91806591e+00 1.61708429e-01 1.76239952e-01 -4.06988114e-01
4.91044670e-01 -2.20409036e-01 -6.24727428e-01 -7.61874840e-02
-2.83014953e-01 4.11642939e-01 7.44510591e-01 -1.22126079e+00
-1.53459027e-01 -3.07216085e-02 -3.35877419e-01 -9.27382559e-02
-2.18400329e-01 -2.30784804e-01 -1.41811326e-01 -6.41378403e-01
-4.67461459e-02 -1.01757967e+00 -2.78994918e-01 1.36591166e-01
-5.42441495e-02 -1.64453045e-01 7.03400612e-01 -5.91836154e-01
1.14963996e+00 -1.71415746e+00 1.50405601e-01 2.45284483e-01
2.27762356e-01 6.20748460e-01 -8.24715197e-02 2.75222123e-01
-2.07429975e-01 1.72699671e-02 -2.79235601e-01 -3.12224887e-02
-9.07248259e-02 9.90834460e-02 -5.34296572e-01 3.75560522e-01
2.51341283e-01 9.72742319e-01 -8.47903132e-01 -1.40360117e-01
3.43112111e-01 6.65251851e-01 -4.29325461e-01 3.35836768e-01
-1.88945472e-01 1.02680480e+00 -6.56511664e-01 4.16168660e-01
7.25323141e-01 -7.65695035e-01 -2.75076568e-01 -4.55163777e-01
-1.48988366e-02 2.45011687e-01 -1.22981334e+00 1.54728723e+00
-5.92418849e-01 3.35823625e-01 1.55992165e-01 -1.27925324e+00
9.87192571e-01 2.18115479e-01 4.65519547e-01 -8.93862188e-01
2.83132225e-01 4.91162956e-01 2.21480250e-01 -5.28681219e-01
3.86915244e-02 -9.82441306e-02 8.24467372e-03 2.67946154e-01
-2.80849189e-02 -2.30457410e-01 -1.20250061e-01 -3.74425769e-01
7.89585471e-01 2.60579616e-01 -8.96731764e-02 -5.79870701e-01
5.18674076e-01 -9.70867202e-02 6.67892516e-01 7.61674404e-01
-5.91534935e-02 3.77713233e-01 6.17357790e-02 -6.28929317e-01
-1.35187984e+00 -1.00835204e+00 -2.73745328e-01 7.46279359e-01
1.25182033e-01 6.12351522e-02 -6.18648171e-01 -7.73663968e-02
2.10809976e-01 6.80396199e-01 -3.87649328e-01 -5.98994613e-01
-9.28055167e-01 -9.60060179e-01 5.39254010e-01 7.34813154e-01
5.97390413e-01 -1.01227677e+00 -3.52698058e-01 3.18013668e-01
3.18188429e-01 -1.05866909e+00 -1.79526165e-01 2.59376913e-01
-9.49172556e-01 -8.36579978e-01 -5.77612102e-01 -5.33996046e-01
4.55770284e-01 -7.40180239e-02 1.14950800e+00 1.26427904e-01
-2.48707116e-01 -2.47407198e-01 2.58992314e-01 3.33183333e-02
-4.31718588e-01 2.00635701e-01 4.09143478e-01 -3.23348641e-01
-3.40371192e-01 -8.07938218e-01 -7.45830953e-01 3.85389566e-01
-6.99035287e-01 2.16677990e-02 7.98814476e-01 1.19984055e+00
7.03275383e-01 -4.88190958e-03 9.62652087e-01 -5.43902874e-01
6.65110290e-01 -6.33887947e-01 -1.02668178e+00 2.41800934e-01
-1.10500956e+00 3.01715970e-01 1.28591084e+00 -6.02783144e-01
-7.97140837e-01 -3.54296297e-01 -1.77235708e-01 -8.27175856e-01
1.32244810e-01 7.22244322e-01 5.74851558e-02 -6.09546483e-01
8.21323752e-01 1.54001430e-01 -6.28892705e-02 -5.16430974e-01
-3.96602554e-04 3.05217415e-01 5.44938922e-01 -1.00603318e+00
8.05186152e-01 -4.75299805e-02 5.12879729e-01 -6.02392435e-01
-8.95755768e-01 9.37835351e-02 -3.26061726e-01 -3.87296155e-02
4.54289049e-01 -7.82824397e-01 -9.44722652e-01 6.42053246e-01
-7.60380685e-01 -5.54544806e-01 -1.09139457e-01 6.39353037e-01
-4.19612139e-01 1.18713900e-01 -5.44826448e-01 -5.37917614e-01
-6.78352952e-01 -1.38355446e+00 5.90288818e-01 6.00027502e-01
-8.00907463e-02 -8.55910242e-01 -1.28630891e-01 8.94452408e-02
8.93614769e-01 3.81946802e-01 1.18085682e+00 -2.14874148e-01
-6.81056619e-01 -2.17068523e-01 -3.57711375e-01 3.95512849e-01
-1.69475198e-01 1.02891468e-01 -8.16234767e-01 -6.75942540e-01
-6.14131475e-03 -5.77090859e-01 5.22413969e-01 4.71268386e-01
1.77577507e+00 -2.28413939e-02 -4.22429979e-01 1.11574936e+00
1.27591038e+00 2.87043065e-01 3.33389431e-01 9.50479433e-02
6.52489245e-01 2.39987429e-02 2.10847914e-01 1.90270156e-01
1.55623585e-01 5.23524821e-01 2.47395188e-01 -5.26258424e-02
5.02507910e-02 -2.66304255e-01 2.37752404e-02 1.13344097e+00
-1.37680406e-02 -6.98696598e-02 -1.14092457e+00 1.98239252e-01
-1.75300753e+00 -5.90625167e-01 1.85205057e-01 2.38912582e+00
8.33572447e-01 3.45716029e-01 -1.94641694e-01 -1.23368531e-01
5.21529794e-01 -1.05997577e-01 -1.23429036e+00 -1.95193052e-01
6.71091229e-02 4.46077108e-01 4.60592687e-01 2.56309628e-01
-8.89557600e-01 5.82990825e-01 7.02104902e+00 1.00885963e+00
-1.64836586e+00 -1.15341753e-01 9.03562248e-01 -1.35220975e-01
-6.98958635e-02 -1.49002314e-01 -7.20092058e-01 6.58125103e-01
1.44048846e+00 -1.59172639e-01 7.57575393e-01 6.89782858e-01
2.94068933e-01 1.83667764e-01 -9.97176647e-01 1.06399035e+00
-3.86717081e-01 -1.67513847e+00 -8.63408446e-02 -2.12094426e-01
9.19951022e-01 2.67580926e-01 2.55850852e-01 6.83179677e-01
8.65666047e-02 -1.52158749e+00 4.37614828e-01 8.06640089e-01
1.00835192e+00 -7.19455779e-01 5.51428616e-01 4.79602665e-01
-1.10669851e+00 2.80276462e-02 -4.55837876e-01 4.05297428e-02
-3.17434631e-02 4.99889731e-01 -4.04766023e-01 4.60771233e-01
8.42813373e-01 6.01140380e-01 -1.57253355e-01 1.17096317e+00
5.43124005e-02 8.59102666e-01 -5.69974661e-01 -1.92281529e-01
2.58227795e-01 -4.02797431e-01 4.41398114e-01 7.91163266e-01
4.61422890e-01 -2.59292051e-02 1.77786842e-01 1.45660210e+00
-2.00556248e-01 -4.17515874e-01 -6.19161390e-02 -1.66108698e-01
7.22165883e-01 1.23809874e+00 -3.57841045e-01 -1.25846967e-01
-1.42415807e-01 2.92270333e-01 3.58434051e-01 5.51783264e-01
-1.12891698e+00 -3.28120351e-01 7.02356815e-01 1.74706265e-01
2.10298255e-01 -4.14501816e-01 -3.78056139e-01 -7.33224452e-01
-3.67628857e-02 -9.00606751e-01 -8.16782862e-02 -8.13874960e-01
-1.46961248e+00 5.78900218e-01 -8.65357667e-02 -1.18929493e+00
-3.09978306e-01 -7.29008317e-01 -6.96053028e-01 1.25174940e+00
-1.63392437e+00 -8.67984712e-01 -4.97454047e-01 3.54635119e-01
1.42358139e-01 -2.94105828e-01 9.25800323e-01 4.51053262e-01
-7.99415290e-01 7.59680986e-01 6.83649123e-01 -1.16524830e-01
6.05211616e-01 -9.68961120e-01 5.35558879e-01 3.02383453e-01
-5.27340412e-01 7.31115878e-01 7.97549844e-01 -3.85601997e-01
-1.81895781e+00 -1.13645160e+00 2.45350301e-01 -9.67240110e-02
7.35391140e-01 -1.94643363e-01 -1.25259197e+00 1.86700344e-01
-1.23918317e-01 3.95302832e-01 4.03683126e-01 -7.29465112e-02
-2.00173825e-01 -4.24273938e-01 -1.10273516e+00 6.23563051e-01
8.33567321e-01 -3.90782773e-01 -9.73451212e-02 2.22536936e-01
5.72059274e-01 -7.61538625e-01 -1.20782185e+00 8.92299891e-01
5.14958441e-01 -5.23484528e-01 1.20125306e+00 -7.28016555e-01
4.52603966e-01 -3.14650685e-01 1.19076103e-01 -1.39074945e+00
-3.27266306e-01 -8.16020846e-01 -4.44143236e-01 8.29865873e-01
4.03323412e-01 -7.51413584e-01 5.72973192e-01 6.58822596e-01
-2.56885558e-01 -1.39472926e+00 -1.05807400e+00 -8.61177623e-01
5.52572310e-01 -1.53695792e-01 7.44424939e-01 9.04065847e-01
-5.46769559e-01 3.60285908e-01 -3.93897802e-01 2.09698737e-01
7.04195857e-01 3.10005635e-01 6.10631585e-01 -1.23423302e+00
-4.14339095e-01 -6.10592604e-01 -1.12621523e-02 -1.18534136e+00
2.40873203e-01 -7.21682310e-01 -5.23934700e-02 -1.13478839e+00
-7.69049004e-02 -8.50623012e-01 -5.49486279e-01 2.19312429e-01
-2.06027225e-01 -1.33266494e-01 -1.28156364e-01 3.73879135e-01
-2.94251144e-01 1.01738691e+00 1.35581338e+00 1.55045930e-02
3.49863693e-02 1.09579489e-02 -5.35799682e-01 5.22580564e-01
7.52828002e-01 -4.49243635e-01 -3.25683236e-01 -4.72455740e-01
2.76022583e-01 3.21980268e-01 6.29379749e-01 -1.45716381e+00
2.62040436e-01 -2.90111870e-01 7.36971498e-01 -4.80620772e-01
4.08600241e-01 -7.51765192e-01 2.98222393e-01 5.15338421e-01
-3.10479909e-01 2.28269413e-01 6.26260221e-01 5.05159676e-01
2.77068280e-02 -7.18075261e-02 1.13797283e+00 1.07269235e-01
-3.44172508e-01 6.13885522e-01 8.58044624e-02 2.23570183e-01
7.31456995e-01 1.02647707e-01 -3.49650562e-01 -2.25282252e-01
-3.38104874e-01 4.86863971e-01 3.68910462e-01 2.79171288e-01
2.87981302e-01 -1.30618548e+00 -5.21160841e-01 3.63737911e-01
-2.50753045e-01 2.69273698e-01 3.82702857e-01 7.82109022e-01
-4.76112187e-01 4.11730886e-01 -2.81232268e-01 -7.73572505e-01
-4.01510179e-01 4.15777028e-01 7.13619709e-01 -2.01527774e-01
-5.66008925e-01 6.52536631e-01 -1.76163763e-01 -6.14114761e-01
1.16484776e-01 -2.62303859e-01 3.41941535e-01 -7.23765671e-01
2.16176778e-01 6.31767035e-01 2.19208866e-01 -3.53953719e-01
-1.36866987e-01 7.16811776e-01 -7.35190958e-02 3.17904830e-01
1.48679793e+00 2.79219449e-01 1.85775384e-01 3.46067399e-01
1.20164347e+00 -4.22439098e-01 -1.76198471e+00 -2.72982210e-01
-2.61289030e-01 -1.34494439e-01 3.11467171e-01 -9.98841763e-01
-1.08520162e+00 9.58457768e-01 7.92595923e-01 1.94978565e-02
9.09326077e-01 -4.29015189e-01 1.02707636e+00 6.45739019e-01
2.35277295e-01 -1.19147432e+00 -2.07601227e-02 6.91220641e-01
9.00462151e-01 -1.23760104e+00 1.50501683e-01 -2.93159811e-03
-3.20037127e-01 1.17575312e+00 7.15619862e-01 -4.64003712e-01
8.76092196e-01 3.83649319e-01 -1.66063711e-01 -7.03808814e-02
-5.83358824e-01 6.22993410e-01 4.44938064e-01 2.69562662e-01
3.30822021e-01 -1.22825116e-01 1.25961408e-01 7.08503425e-01
-6.18315078e-02 1.82813108e-01 -9.10900608e-02 6.88950419e-01
-2.03829542e-01 -9.26926494e-01 -1.27165884e-01 7.12293684e-01
-8.59662965e-02 8.18975344e-02 2.97344804e-01 9.09143031e-01
-2.02355653e-01 5.60028553e-01 -2.03545347e-01 -2.61900395e-01
3.43573302e-01 -1.13143504e-01 3.98174226e-01 -1.33848652e-01
-7.46378601e-01 -1.45487517e-01 -3.24072778e-01 -6.01059139e-01
-6.00580312e-02 -3.66771311e-01 -1.26228905e+00 -6.50058985e-01
-3.12845230e-01 5.95795810e-02 4.89863753e-01 1.03715265e+00
6.80614352e-01 8.25484276e-01 6.27784550e-01 -1.05476320e+00
-1.27264750e+00 -9.37530816e-01 -2.52522945e-01 2.60370284e-01
3.45177978e-01 -9.99212146e-01 -3.43568891e-01 -4.00438398e-01] | [6.428447723388672, 3.500135660171509] |
1384e3d1-a9d4-4f07-9bf1-e21819201766 | syntax-guided-domain-adaptation-for-aspect | 2211.05457 | null | https://arxiv.org/abs/2211.05457v1 | https://arxiv.org/pdf/2211.05457v1.pdf | Syntax-Guided Domain Adaptation for Aspect-based Sentiment Analysis | Aspect-based sentiment analysis (ABSA) aims at extracting opinionated aspect terms in review texts and determining their sentiment polarities, which is widely studied in both academia and industry. As a fine-grained classification task, the annotation cost is extremely high. Domain adaptation is a popular solution to alleviate the data deficiency issue in new domains by transferring common knowledge across domains. Most cross-domain ABSA studies are based on structure correspondence learning (SCL), and use pivot features to construct auxiliary tasks for narrowing down the gap between domains. However, their pivot-based auxiliary tasks can only transfer knowledge of aspect terms but not sentiment, limiting the performance of existing models. In this work, we propose a novel Syntax-guided Domain Adaptation Model, named SDAM, for more effective cross-domain ABSA. SDAM exploits syntactic structure similarities for building pseudo training instances, during which aspect terms of target domain are explicitly related to sentiment polarities. Besides, we propose a syntax-based BERT mask language model for further capturing domain-invariant features. Finally, to alleviate the sentiment inconsistency issue in multi-gram aspect terms, we introduce a span-based joint aspect term and sentiment analysis module into the cross-domain End2End ABSA. Experiments on five benchmark datasets show that our model consistently outperforms the state-of-the-art baselines with respect to Micro-F1 metric for the cross-domain End2End ABSA task. | ['Jing Xiao', 'Lei Wang', 'Xuan Wang', 'Qing Liao', 'Yan Jia', 'Cuiyun Gao', 'Anguo Dong'] | 2022-11-10 | null | null | null | null | ['aspect-based-sentiment-analysis'] | ['natural-language-processing'] | [ 5.27585074e-02 -3.17410618e-01 -5.06187558e-01 -7.77893245e-01
-9.24333274e-01 -8.63033235e-01 6.21838748e-01 1.73307717e-01
-2.04192847e-01 5.62649071e-01 2.82795787e-01 -2.88741708e-01
8.92643780e-02 -8.39352727e-01 -5.14673650e-01 -5.94491243e-01
7.76105642e-01 4.56688046e-01 7.03268945e-02 -6.87945187e-01
1.63325340e-01 -1.26311004e-01 -9.75814879e-01 5.59716225e-01
1.23839498e+00 1.09818208e+00 -4.81906012e-02 -2.02330485e-01
-7.50763237e-01 1.95165575e-01 -6.53568804e-01 -9.33099091e-01
1.81758076e-01 -3.69097978e-01 -7.40496218e-01 1.79497436e-01
1.27555087e-01 1.95537448e-01 4.52335864e-01 1.03763926e+00
3.31597000e-01 -5.81533164e-02 7.50723600e-01 -1.08172941e+00
-7.88007796e-01 3.96815002e-01 -7.99237192e-01 -1.13029428e-01
3.33207101e-01 -1.37069553e-01 1.47783637e+00 -1.13528514e+00
4.48120415e-01 1.18700635e+00 7.20151842e-01 4.35058802e-01
-9.67592597e-01 -9.00045872e-01 7.09739208e-01 7.97640160e-02
-8.91821861e-01 -9.47823450e-02 1.10162008e+00 -3.60511124e-01
9.72834766e-01 5.88190369e-02 5.88216126e-01 1.08807659e+00
9.38134547e-03 1.04896879e+00 1.25037432e+00 -3.23664725e-01
2.15662912e-01 4.20263797e-01 3.62330854e-01 1.88372776e-01
3.86657804e-01 -5.65065980e-01 -6.60087109e-01 -2.22656503e-01
2.11846367e-01 -2.57445984e-02 7.26820603e-02 -4.82071280e-01
-1.00715911e+00 9.22771335e-01 -1.49271348e-02 1.91742688e-01
-3.08559626e-01 -8.34131122e-01 8.58505726e-01 4.71688867e-01
8.48177314e-01 6.54022813e-01 -1.31510222e+00 -6.58861250e-02
-5.37313879e-01 3.41021389e-01 8.02723050e-01 1.05675316e+00
8.78912628e-01 -1.31291077e-01 -1.08052135e-01 1.21528625e+00
3.27693015e-01 7.77229011e-01 7.47243583e-01 -3.28312635e-01
8.24675143e-01 1.28648722e+00 -2.01089665e-01 -7.78416634e-01
-2.09548637e-01 -3.86080861e-01 -7.20887601e-01 -2.42533654e-01
1.39012694e-01 -2.22076550e-01 -8.30083132e-01 1.57020497e+00
5.90993524e-01 -3.04898202e-01 3.82057011e-01 6.20938659e-01
8.41434836e-01 5.45750320e-01 9.44871902e-02 -8.98911953e-02
1.92135191e+00 -1.12969685e+00 -5.56655705e-01 -7.33975053e-01
8.12279105e-01 -1.09338939e+00 1.50865579e+00 4.39782977e-01
-5.87526083e-01 -4.65724587e-01 -1.03601575e+00 -2.36159787e-01
-6.67534113e-01 2.62941420e-01 7.61245668e-01 6.36926413e-01
-3.43057156e-01 5.79168610e-02 -4.71466273e-01 -4.33802865e-02
4.72466588e-01 2.14295521e-01 -4.53809291e-01 -2.43835375e-01
-1.49156845e+00 5.66833794e-01 2.08357558e-01 -2.46988446e-01
-2.03328803e-01 -9.48553801e-01 -1.16585088e+00 -1.60644427e-01
4.08128232e-01 -6.93976581e-01 1.25254142e+00 -1.27067363e+00
-1.50031853e+00 1.10329759e+00 -5.61876178e-01 -1.01251371e-01
-1.09043211e-01 -4.91138667e-01 -7.29512393e-01 -2.76921570e-01
5.55609286e-01 1.58909455e-01 8.88856947e-01 -9.99699831e-01
-8.30646157e-01 -5.91056347e-01 3.77545029e-01 4.96006161e-01
-6.76965773e-01 5.94394282e-02 -5.42054296e-01 -9.54044282e-01
-1.04356475e-01 -9.16422546e-01 -2.15764001e-01 -5.06781161e-01
-1.50561303e-01 -4.96727437e-01 8.42679203e-01 -4.82748926e-01
1.31794131e+00 -2.21531630e+00 3.87314484e-02 7.83352405e-02
-1.42461538e-01 5.05603075e-01 -3.91482204e-01 1.26915768e-01
-1.84080362e-01 -1.54481307e-01 -4.57398593e-01 -3.25038403e-01
7.23489979e-03 1.20459557e-01 -6.16822958e-01 -1.09842278e-01
4.34782386e-01 7.89406002e-01 -7.42073894e-01 -3.21012884e-01
-1.09758407e-01 2.00462505e-01 -6.38715923e-01 1.66170582e-01
-4.26576316e-01 3.19731116e-01 -7.61183262e-01 8.01454961e-01
1.00613642e+00 -2.22486556e-01 2.26149753e-01 -3.57408792e-01
9.60588977e-02 8.85999858e-01 -8.35699201e-01 1.83781374e+00
-1.05492854e+00 1.90565214e-01 -1.57804489e-01 -1.13615990e+00
1.29502618e+00 8.81508514e-02 3.19658250e-01 -8.29193532e-01
1.65058091e-01 3.07712704e-01 -1.06431305e-01 -2.50293732e-01
6.11041188e-01 -6.01705492e-01 -6.55974150e-01 3.78192395e-01
8.67526531e-02 -4.29527313e-01 3.63143742e-01 -4.30310238e-03
7.06029177e-01 1.09289452e-01 6.54235125e-01 -3.52479964e-01
1.02839398e+00 2.22876936e-01 9.17986572e-01 -2.16876213e-02
-6.26047924e-02 5.60765445e-01 7.10550666e-01 -3.76623005e-01
-7.31835425e-01 -7.28387177e-01 -2.31776237e-01 1.33317089e+00
3.27943027e-01 -6.64612532e-01 -5.42546928e-01 -1.22495079e+00
-4.71400097e-02 6.22636735e-01 -7.15179801e-01 -1.73006386e-01
-4.03530926e-01 -9.54685688e-01 1.13274843e-01 5.72966754e-01
6.38573647e-01 -9.45253670e-01 2.98398048e-01 1.64312080e-01
-3.89580458e-01 -1.33643413e+00 -6.47170007e-01 1.94136247e-01
-8.21012139e-01 -9.64013398e-01 -7.99351811e-01 -9.56966043e-01
5.56827664e-01 4.03214961e-01 1.36407864e+00 -5.31148493e-01
3.69811654e-01 -4.46951278e-02 -7.44182467e-01 -6.10985219e-01
-5.80928065e-02 4.96563703e-01 -5.83404042e-02 7.09810480e-02
1.29702568e+00 -5.42418003e-01 -6.66665256e-01 4.89808440e-01
-8.88272762e-01 -1.78873762e-01 8.45824957e-01 8.47644985e-01
8.46670926e-01 -1.47828653e-01 7.68803537e-01 -1.45703149e+00
9.58507538e-01 -5.74882925e-01 -4.72039640e-01 2.01878414e-01
-7.29804039e-01 2.16795020e-02 1.01044261e+00 -3.58276933e-01
-1.36132503e+00 -1.35886997e-01 -1.61726102e-01 7.56408870e-02
-1.02457497e-03 6.80049956e-01 -7.67094970e-01 3.35196316e-01
5.85632145e-01 3.90885293e-01 -1.64276809e-01 -4.45298642e-01
2.33869642e-01 8.32992792e-01 -1.15682033e-03 -6.95951521e-01
6.52177930e-01 4.37245965e-01 -2.53389895e-01 -5.52026212e-01
-1.54196656e+00 -8.51483464e-01 -5.72181702e-01 4.84764844e-01
5.48885584e-01 -1.46921265e+00 -5.08387573e-02 6.99674666e-01
-1.11712646e+00 6.75432831e-02 -2.30907500e-01 2.52149880e-01
-2.00094521e-01 4.45019037e-01 -2.41259366e-01 -1.07466787e-01
-6.51855886e-01 -1.15053129e+00 1.28654313e+00 4.11626816e-01
-4.33421642e-01 -1.19168675e+00 4.00953859e-01 7.05295205e-01
1.98109210e-01 -1.74082294e-01 9.77365553e-01 -1.07161200e+00
-2.75147660e-03 -1.78862646e-01 -3.18035901e-01 9.47763920e-01
5.88031709e-01 -4.64337289e-01 -8.55736554e-01 -3.89211774e-02
2.25614995e-01 -2.71389395e-01 6.30445063e-01 1.37365237e-01
8.48414660e-01 -2.00712189e-01 -9.91284028e-02 5.82431316e-01
1.14690375e+00 2.54652533e-03 3.38813126e-01 7.82706380e-01
8.10783803e-01 6.41118646e-01 1.31867146e+00 4.77803349e-01
7.49145687e-01 4.29135680e-01 -1.58658177e-01 -3.17126475e-02
1.39526688e-02 -8.97600725e-02 6.65115356e-01 1.18903446e+00
2.15866446e-01 6.75776750e-02 -7.29919136e-01 1.00112796e+00
-1.57358921e+00 -4.01375353e-01 -8.25283825e-02 1.85082257e+00
1.24570310e+00 3.40434790e-01 8.19920972e-02 -6.61457852e-02
4.00470197e-01 3.89655888e-01 -7.18296707e-01 -6.66564465e-01
-3.04388255e-01 3.49642754e-01 6.23404980e-02 1.61523357e-01
-1.31579411e+00 1.15334153e+00 4.46486759e+00 1.17353880e+00
-1.00291800e+00 1.57920241e-01 4.71657813e-01 1.50899723e-01
-7.24668622e-01 7.25492239e-02 -1.14735973e+00 4.42909151e-01
4.64752913e-01 -3.60245228e-01 -1.20870583e-01 1.43272388e+00
-6.50785863e-02 1.71283305e-01 -9.25431073e-01 7.90840209e-01
3.04294020e-01 -9.45208788e-01 3.96549493e-01 -1.41862944e-01
1.10835695e+00 -1.08596399e-01 2.50491768e-01 7.85584033e-01
2.17886776e-01 -5.84251583e-01 2.11254537e-01 -2.28012294e-01
7.50317216e-01 -9.55382884e-01 1.06723642e+00 -4.26689237e-02
-1.35310590e+00 2.46252015e-01 -3.80920917e-01 1.29579185e-02
8.25289562e-02 1.02035809e+00 -5.46804190e-01 7.08854556e-01
6.26388490e-01 1.17126667e+00 -4.57440466e-01 4.25787061e-01
-6.14007056e-01 5.39487541e-01 6.14332175e-03 -7.09357187e-02
4.14455980e-01 -4.06522095e-01 4.49321240e-01 1.27544999e+00
1.38126165e-01 -1.07912295e-01 2.74642580e-03 4.11496341e-01
-1.75652876e-01 5.20397782e-01 -5.17219126e-01 -6.28877878e-02
2.53969043e-01 1.41787922e+00 -3.28568220e-01 -3.58470738e-01
-9.10114706e-01 1.00168395e+00 1.04451410e-01 2.39619181e-01
-5.38900256e-01 -6.91285729e-01 1.28002906e+00 1.48822293e-02
6.36395991e-01 1.54190697e-02 -7.12310672e-01 -1.60598564e+00
4.44647104e-01 -1.48831522e+00 4.71059144e-01 -3.24477345e-01
-1.79465127e+00 6.61237776e-01 -2.19173700e-01 -1.66743052e+00
-8.44663829e-02 -8.26779127e-01 -5.64488828e-01 1.01438189e+00
-2.05775023e+00 -1.44586992e+00 5.08759497e-03 7.74810910e-01
9.16564345e-01 -4.87000138e-01 7.87422597e-01 3.78415018e-01
-3.70037496e-01 8.59869838e-01 1.56207323e-01 1.95537463e-01
1.26648402e+00 -1.16411233e+00 6.34609401e-01 6.94694459e-01
-1.11476779e-01 9.35604692e-01 4.43409801e-01 -5.80094576e-01
-1.17222226e+00 -1.21190178e+00 1.06174076e+00 -6.46196067e-01
8.66481304e-01 -4.34378862e-01 -1.16031039e+00 6.08112752e-01
1.53612360e-01 -1.58881277e-01 1.17746437e+00 8.17134202e-01
-8.14440310e-01 -4.56715137e-01 -8.01036000e-01 5.57042897e-01
6.70349300e-01 -7.03091800e-01 -1.03292716e+00 1.28431201e-01
9.57508683e-01 -2.20769241e-01 -7.21777678e-01 6.44037843e-01
4.44651127e-01 -7.86466897e-01 7.59979486e-01 -5.61965823e-01
7.62855053e-01 -3.67857903e-01 -1.13348849e-01 -1.61584759e+00
-8.08135718e-02 -2.37963945e-01 7.07714260e-02 1.63019872e+00
6.99541688e-01 -7.06001699e-01 6.87306881e-01 4.05994177e-01
5.09305065e-03 -8.09306562e-01 -6.15385294e-01 -7.90136158e-01
3.94547731e-01 -4.19388920e-01 8.70882154e-01 1.13144064e+00
6.16245717e-02 9.80686724e-01 -9.07140691e-03 3.36305574e-02
2.27008834e-01 6.55877531e-01 9.32865262e-01 -1.06743443e+00
-2.14334428e-01 -4.40995604e-01 -1.36391178e-01 -1.18587232e+00
4.29018259e-01 -7.56970763e-01 -1.36309475e-01 -1.29004765e+00
2.22610086e-01 -4.96108025e-01 -4.39064413e-01 4.77702051e-01
-5.47249258e-01 2.30361169e-04 -1.52628466e-01 -3.38110812e-02
-7.37982333e-01 8.08367729e-01 1.36145973e+00 -2.86001474e-01
-2.83795148e-01 3.04391772e-01 -1.36064506e+00 1.03647339e+00
8.71564031e-01 -5.02505183e-01 -6.30024493e-01 -4.76230830e-01
6.46276593e-01 -6.01416111e-01 -3.74394029e-01 -5.09706795e-01
-1.36181876e-01 -2.77600706e-01 4.33520451e-02 -6.26098156e-01
8.26945528e-02 -8.22550595e-01 -6.92856669e-01 -8.60356092e-02
-5.74784130e-02 6.93413541e-02 4.72966433e-01 4.52832639e-01
-8.91347408e-01 -1.77230269e-01 5.53947270e-01 9.37201828e-03
-7.91181505e-01 2.41703987e-01 -4.57763933e-02 5.99904835e-01
8.21899593e-01 1.31073624e-01 -1.97472274e-01 3.62808630e-02
-1.29498556e-01 2.31281891e-01 4.46934611e-01 6.95986986e-01
3.61165285e-01 -1.28712690e+00 -5.80497563e-01 2.68833697e-01
7.06091285e-01 4.88591433e-01 4.22563195e-01 5.41689396e-01
-6.83232099e-02 4.17707950e-01 -2.40667374e-03 -3.90196621e-01
-1.30277860e+00 3.76155794e-01 -1.08520851e-01 -9.22871470e-01
-5.97389936e-02 9.54731941e-01 7.70815551e-01 -9.82645094e-01
-3.86683762e-01 -2.32181534e-01 -3.20032984e-01 3.94287407e-01
5.27700901e-01 -1.58570305e-01 3.63711417e-01 -4.85566705e-01
-5.63091338e-01 8.82859051e-01 -6.90268993e-01 7.51633495e-02
1.31362963e+00 -3.21136981e-01 -2.79997945e-01 3.53466660e-01
1.18296778e+00 3.94345760e-01 -8.64113271e-01 -6.42847538e-01
-9.38756298e-03 -2.65387326e-01 -2.52287298e-01 -1.00111341e+00
-8.97406995e-01 1.03889275e+00 2.25540996e-02 -1.12895131e-01
1.42678010e+00 5.38214892e-02 1.19295001e+00 3.67427588e-01
2.69927159e-02 -1.25448155e+00 9.22447443e-02 1.00437343e+00
7.37835705e-01 -1.58882189e+00 1.52672902e-01 -4.93097395e-01
-1.23190451e+00 8.12642097e-01 1.00531018e+00 -9.67932492e-02
7.26007402e-01 7.14872107e-02 3.90319675e-01 -7.20843375e-02
-6.98823869e-01 -1.21303789e-01 4.51820970e-01 6.31187022e-01
4.64450240e-01 3.01838275e-02 -6.33162379e-01 1.48233902e+00
-3.24689358e-01 -2.26883397e-01 3.81977893e-02 7.78688252e-01
-1.69980988e-01 -1.63289917e+00 -1.07737958e-01 2.46078953e-01
-6.06615365e-01 -5.20298839e-01 -4.45403367e-01 6.61965132e-01
1.59994498e-01 7.46098459e-01 -3.10159355e-01 -2.93487787e-01
6.92462325e-01 7.91186914e-02 5.24910167e-02 -9.27648127e-01
-7.01708615e-01 2.38919389e-02 2.59808213e-01 -2.50000060e-01
-5.60421705e-01 -7.23432958e-01 -1.01661968e+00 1.10788755e-01
-1.82328120e-01 2.44142234e-01 6.84004843e-01 1.25669312e+00
6.98163033e-01 4.48913991e-01 8.54950547e-01 -1.10820226e-01
-3.01770240e-01 -9.95664001e-01 -5.52065372e-01 5.75673044e-01
2.28778467e-01 -5.11487901e-01 -1.90323994e-01 8.78154859e-03] | [11.399372100830078, 6.695052623748779] |
655bf48f-e7b9-4133-a50b-3e04ffb19ad8 | utts-unsupervised-tts-with-conditional | 2206.02512 | null | https://arxiv.org/abs/2206.02512v3 | https://arxiv.org/pdf/2206.02512v3.pdf | UTTS: Unsupervised TTS with Conditional Disentangled Sequential Variational Auto-encoder | In this paper, we propose a novel unsupervised text-to-speech (UTTS) framework which does not require text-audio pairs for the TTS acoustic modeling (AM). UTTS is a multi-speaker speech synthesizer that supports zero-shot voice cloning, it is developed from a perspective of disentangled speech representation learning. The framework offers a flexible choice of a speaker's duration model, timbre feature (identity) and content for TTS inference. We leverage recent advancements in self-supervised speech representation learning as well as speech synthesis front-end techniques for system development. Specifically, we employ our recently formulated Conditional Disentangled Sequential Variational Auto-encoder (C-DSVAE) as the backbone UTTS AM, which offers well-structured content representations given unsupervised alignment (UA) as condition during training. For UTTS inference, we utilize a lexicon to map input text to the phoneme sequence, which is expanded to the frame-level forced alignment (FA) with a speaker-dependent duration model. Then, we develop an alignment mapping module that converts FA to UA. Finally, the C-DSVAE, serving as the self-supervised TTS AM, takes the predicted UA and a target speaker embedding to generate the mel spectrogram, which is ultimately converted to waveform with a neural vocoder. We show how our method enables speech synthesis without using a paired TTS corpus. Experiments demonstrate that UTTS can synthesize speech of high naturalness and intelligibility measured by human and objective evaluations. Audio samples are available at our demo page https://neurtts.github.io/utts_demo. | ['Dong Yu', 'Gopala Krishna Anumanchipalli', 'Chunlei Zhang', 'Jiachen Lian'] | 2022-06-06 | null | null | null | null | ['voice-cloning'] | ['speech'] | [ 3.17367017e-01 1.86557278e-01 -1.97718814e-01 -2.44872838e-01
-1.31608391e+00 -4.96139199e-01 4.70277816e-01 -6.07983589e-01
1.72750413e-01 4.68793273e-01 4.94037986e-01 -4.48277503e-01
2.89210528e-01 -5.26684165e-01 -5.97657502e-01 -8.24311912e-01
4.12099779e-01 3.62316668e-01 -2.30629086e-01 -1.67249396e-01
-4.70072865e-01 1.12715513e-01 -1.57589495e+00 1.46512002e-01
8.49225044e-01 8.33898187e-01 5.69419086e-01 1.23760498e+00
2.79924404e-02 5.93069732e-01 -6.78673804e-01 -4.55002874e-01
1.21890649e-01 -7.92138517e-01 -4.10298735e-01 1.06271300e-02
-4.43400554e-02 -4.68673617e-01 -4.42854553e-01 6.55050099e-01
7.36442924e-01 1.43367559e-01 7.34436512e-01 -1.16675580e+00
-6.04425311e-01 9.53046560e-01 6.68371022e-02 -9.08291861e-02
3.27242285e-01 1.24353610e-01 1.27810609e+00 -1.15970242e+00
2.72249609e-01 1.34401619e+00 2.34426305e-01 7.58952081e-01
-1.26302218e+00 -7.83698082e-01 -1.31635934e-01 5.64452782e-02
-1.28928614e+00 -1.14591038e+00 9.71287668e-01 -4.15187180e-01
9.36969519e-01 6.49470091e-01 5.67358077e-01 1.62796903e+00
-1.82675898e-01 9.80178237e-01 7.38100946e-01 -7.11237133e-01
2.71641880e-01 1.06250115e-01 -2.78310716e-01 5.28380990e-01
-6.90593600e-01 3.74293715e-01 -8.37551117e-01 7.27592930e-02
5.75544477e-01 -4.59465146e-01 -3.04161370e-01 1.07864559e-01
-1.17130888e+00 7.26168692e-01 -2.65104830e-01 2.07608864e-01
-2.23353788e-01 4.86390106e-02 3.21875572e-01 3.98580402e-01
5.20262361e-01 -1.63532328e-03 -4.37664449e-01 -3.33108842e-01
-1.17963767e+00 -4.48418781e-02 8.06141853e-01 1.00351465e+00
5.32354355e-01 8.67341638e-01 -2.85615414e-01 1.23820782e+00
5.00678539e-01 9.62586880e-01 7.98138380e-01 -9.26049769e-01
4.48538125e-01 -1.54932693e-01 -2.18499750e-01 -3.44368488e-01
2.07908019e-01 -5.51697135e-01 -6.81134999e-01 5.54570854e-02
-1.97275952e-01 -2.04338461e-01 -7.96652615e-01 2.07086086e+00
2.83018261e-01 5.05180836e-01 5.32969832e-01 6.55636072e-01
7.04250991e-01 1.07529116e+00 -4.19415653e-01 -5.64274013e-01
1.25229800e+00 -1.10657346e+00 -1.18224823e+00 -1.67538062e-01
3.96997780e-01 -8.45620632e-01 1.38097167e+00 4.36057806e-01
-1.31659353e+00 -7.66650200e-01 -1.21004367e+00 -1.82859972e-01
-1.98699385e-01 5.53007782e-01 -2.00691611e-01 9.96687412e-01
-8.56325388e-01 4.44185078e-01 -9.24638093e-01 8.67695883e-02
-1.42770171e-01 5.29678427e-02 -9.15505961e-02 4.06116605e-01
-1.43574429e+00 7.10210621e-01 1.43267944e-01 -4.64451574e-02
-1.31814420e+00 -6.28074825e-01 -1.18653119e+00 4.77224290e-02
1.42131120e-01 -7.52723992e-01 1.61044896e+00 -6.12511158e-01
-2.45998049e+00 5.41207850e-01 -5.47339261e-01 -4.40212637e-01
1.78494826e-01 -1.49082482e-01 -8.90041888e-01 1.12076484e-01
4.19000462e-02 4.08598393e-01 1.40372396e+00 -1.14180326e+00
-4.41431075e-01 2.86215097e-02 -6.61035657e-01 3.22112501e-01
-5.26204526e-01 6.85214698e-02 -2.36778736e-01 -1.02626228e+00
-5.62289096e-02 -8.73096347e-01 1.53430864e-01 -4.84886795e-01
-8.25331032e-01 -5.09127937e-02 7.66867459e-01 -1.00032938e+00
1.51069438e+00 -2.30509329e+00 6.38880551e-01 -1.62471578e-01
-5.47846779e-02 2.29036376e-01 -1.01383455e-01 5.92812598e-01
-2.12552220e-01 1.08788669e-01 -3.83931786e-01 -1.27482939e+00
3.93165290e-01 2.61830151e-01 -8.55098307e-01 1.78910658e-01
3.45610410e-01 6.67765558e-01 -6.95683122e-01 -4.04093951e-01
4.84544694e-01 5.04759669e-01 -5.99553525e-01 7.77184308e-01
-3.61524284e-01 7.36253798e-01 -1.00529879e-01 3.60550106e-01
2.59377003e-01 3.51999342e-01 1.23624571e-01 -8.22733268e-02
-3.60212147e-01 7.49321818e-01 -9.40101504e-01 1.96232700e+00
-8.53012502e-01 5.57788730e-01 8.46452117e-02 -7.05424130e-01
1.00755727e+00 9.23463345e-01 1.17304839e-01 -3.04909140e-01
1.32316187e-01 3.62124085e-01 -7.96654224e-02 -4.98747051e-01
3.53655577e-01 -2.71351337e-01 -3.48664448e-02 4.66627985e-01
5.40537298e-01 -6.50136948e-01 -1.23341702e-01 1.19191231e-02
8.19347978e-01 1.80928960e-01 1.95130229e-01 2.06655890e-01
5.10847390e-01 -6.19455516e-01 4.82889235e-01 2.01485425e-01
1.84742510e-01 9.37375188e-01 2.55498648e-01 4.37449694e-01
-1.14232326e+00 -1.68884671e+00 -3.38769294e-02 1.13890517e+00
-5.72938383e-01 -5.77317119e-01 -1.00849366e+00 1.72889866e-02
-3.52851063e-01 1.25823796e+00 -2.76082009e-01 -2.02724308e-01
-5.09623766e-01 2.91853286e-02 9.26968098e-01 2.53176808e-01
-8.24918672e-02 -1.02941310e+00 1.21786103e-01 2.58106619e-01
-3.31121653e-01 -1.16822696e+00 -8.31086457e-01 1.85912684e-01
-4.78801697e-01 -2.37604186e-01 -8.64349484e-01 -7.69683719e-01
6.35163710e-02 -6.73311949e-02 5.84199369e-01 -6.35262966e-01
2.26820990e-01 2.85926938e-01 -3.69560033e-01 -3.64410996e-01
-1.05847204e+00 3.51308696e-02 7.01454878e-01 3.13622713e-01
-1.44340068e-01 -1.01879179e+00 -2.54992604e-01 2.02471912e-01
-8.00154865e-01 4.10956949e-01 3.47443312e-01 7.99667656e-01
6.21535599e-01 -1.94297045e-01 8.08423877e-01 -4.18533057e-01
6.66969121e-01 -5.22654891e-01 -4.38211054e-01 1.54167667e-01
-3.40362400e-01 2.79501360e-02 9.18992579e-01 -6.35809541e-01
-1.23623037e+00 -1.27101019e-01 -5.53485990e-01 -9.22012448e-01
-1.28134757e-01 3.67235273e-01 -6.02586925e-01 6.62526846e-01
6.26304030e-01 6.34611845e-01 1.61710799e-01 -5.45110047e-01
7.31330037e-01 1.35866106e+00 7.54594743e-01 -5.26634514e-01
9.36701655e-01 1.88986808e-02 -7.57505119e-01 -1.07312405e+00
-6.35295391e-01 2.02982742e-02 -4.43102568e-01 -6.17915653e-02
8.61613095e-01 -1.08435309e+00 -5.06663322e-01 3.67437392e-01
-1.26471913e+00 -4.53792393e-01 -3.67697030e-01 7.05390096e-01
-9.94111538e-01 2.40045995e-01 -7.03906894e-01 -1.19163823e+00
-4.44469422e-01 -1.32462752e+00 1.13419080e+00 -1.08579315e-01
-3.86802614e-01 -8.06108356e-01 2.96125710e-01 6.10503852e-01
2.60641903e-01 -3.55245881e-02 9.06062901e-01 -5.62893331e-01
-2.82989591e-01 1.79185882e-01 5.68161190e-01 8.75535667e-01
4.01093900e-01 8.05666447e-02 -1.47118783e+00 -2.16485679e-01
2.29365900e-01 -2.32385695e-01 4.08961266e-01 4.73311931e-01
8.67540419e-01 -4.37653571e-01 8.47638994e-02 8.32814574e-01
6.94023132e-01 3.81925315e-01 3.97653103e-01 -3.91131759e-01
7.53575683e-01 5.72610617e-01 2.75854915e-01 4.50298727e-01
3.63162309e-01 7.04643011e-01 -1.60696357e-02 6.77574128e-02
-5.92287302e-01 -6.58795595e-01 9.96663094e-01 2.10145879e+00
2.29127333e-01 -4.99837816e-01 -6.43508852e-01 4.22658235e-01
-1.45038521e+00 -7.98692286e-01 2.83477068e-01 2.23345017e+00
1.18729711e+00 3.50002423e-02 5.21826930e-02 4.71080154e-01
7.51408994e-01 2.79525012e-01 -5.99515915e-01 -5.00424862e-01
-2.94925589e-02 5.23979723e-01 -9.07507688e-02 7.61247218e-01
-5.50035894e-01 1.07881105e+00 5.33337641e+00 1.20199919e+00
-1.07947862e+00 3.81670624e-01 2.33502254e-01 -2.19466895e-01
-6.95632160e-01 -1.32302597e-01 -6.22227073e-01 5.13580561e-01
1.57033682e+00 -3.42341185e-01 8.57123196e-01 4.94929731e-01
5.99501073e-01 6.64856136e-01 -1.28804278e+00 1.04627085e+00
-3.76063623e-02 -1.10699129e+00 7.44121000e-02 -1.13835908e-01
3.65697920e-01 -1.68709517e-01 4.94095743e-01 4.60592896e-01
1.14863724e-01 -1.10086143e+00 1.28141057e+00 3.44863236e-01
1.52546728e+00 -6.61217511e-01 1.42966080e-02 6.08980834e-01
-1.29401898e+00 7.47967586e-02 8.89246985e-02 1.54038906e-01
4.12590951e-01 2.58524626e-01 -1.14377606e+00 6.98150635e-01
1.06263928e-01 5.16102850e-01 1.68332383e-01 3.17054182e-01
-4.62759733e-01 1.20053399e+00 -1.30526781e-01 1.47770047e-01
-2.18017980e-01 -1.64204761e-01 9.39438283e-01 1.08194745e+00
5.90021431e-01 2.68457793e-02 -1.04780041e-01 1.08969092e+00
6.19375966e-02 9.09513906e-02 -4.66246009e-01 -4.84607726e-01
9.38696802e-01 8.68292630e-01 -1.45775229e-01 -1.64055809e-01
-1.82940081e-01 1.17627621e+00 1.51134163e-01 6.99963570e-01
-8.29847038e-01 -3.16823751e-01 8.39986920e-01 -2.25919276e-01
2.82172352e-01 -3.61239791e-01 -5.64938560e-02 -1.25825202e+00
-1.80628762e-01 -9.84335184e-01 -1.90454140e-01 -1.04171848e+00
-1.12161839e+00 9.29403126e-01 -4.66975719e-02 -1.47553182e+00
-8.13730657e-01 -4.38407242e-01 -6.58994377e-01 1.18366098e+00
-1.19596219e+00 -1.25533211e+00 3.57194364e-01 5.40264547e-01
1.02161837e+00 -5.40090501e-01 1.25628805e+00 2.44770452e-01
-9.16345060e-01 7.78394282e-01 7.22275451e-02 4.21909802e-02
5.13341546e-01 -1.21573138e+00 7.15703845e-01 8.99698853e-01
4.22156811e-01 6.29136384e-01 8.12957108e-01 -4.49783087e-01
-1.29538500e+00 -1.07953143e+00 7.67588019e-01 -3.49193662e-01
7.68048644e-01 -8.31261694e-01 -6.97654545e-01 7.73202062e-01
3.95973504e-01 -2.06236839e-01 1.08908761e+00 -1.65239245e-01
-2.05916271e-01 -3.58548313e-02 -6.99691355e-01 7.92922497e-01
8.05674016e-01 -1.11443567e+00 -7.68826902e-01 1.03589803e-01
1.58169532e+00 -5.91008425e-01 -8.49842131e-01 6.51678443e-02
5.07619500e-01 -7.83754408e-01 8.84489000e-01 -2.88962454e-01
3.19847912e-01 -2.65128762e-01 -5.44409037e-01 -1.63978994e+00
-2.83373855e-02 -1.28106630e+00 -3.85057747e-01 1.61435020e+00
5.97655058e-01 -5.71799457e-01 3.38224143e-01 1.01306883e-03
-7.38498569e-01 -5.94638467e-01 -1.20399857e+00 -8.54536831e-01
9.65133160e-02 -1.01336432e+00 7.21158862e-01 7.70384133e-01
-8.65935721e-03 5.15002251e-01 -6.93879187e-01 4.91189480e-01
4.85501438e-01 -6.33418038e-02 5.69789231e-01 -7.26248443e-01
-5.78099370e-01 -2.63767272e-01 1.66279584e-01 -1.18526816e+00
3.89392167e-01 -9.93974090e-01 2.04962030e-01 -1.08608019e+00
-5.01975894e-01 -1.80243343e-01 -1.14576437e-01 1.98591754e-01
8.51818025e-02 -2.72399843e-01 1.81757227e-01 -1.94696728e-02
1.07435957e-01 1.19064760e+00 1.14295316e+00 7.00429231e-02
-3.99950743e-01 3.91156286e-01 -2.08833173e-01 4.49905902e-01
8.05350304e-01 -5.37270665e-01 -7.33342946e-01 -3.58911544e-01
-2.61308908e-01 7.74382889e-01 6.19742386e-02 -9.04634297e-01
1.41974702e-01 4.75581698e-02 -1.29223287e-01 -5.64479411e-01
9.58162963e-01 -4.73188907e-01 3.19262117e-01 2.36402929e-01
-4.89358515e-01 -3.81002277e-01 1.06782004e-01 3.72870177e-01
-3.23242456e-01 -1.66537270e-01 5.15686452e-01 2.72534251e-01
-5.50482385e-02 3.61403584e-01 -4.94551390e-01 -1.05678342e-01
5.88832140e-01 -1.75144047e-01 7.64650118e-04 -5.11058211e-01
-9.57911909e-01 -3.77681479e-02 2.05790792e-02 4.83918011e-01
7.61145175e-01 -1.60454071e+00 -8.99643600e-01 5.33502996e-01
-1.94883514e-02 -8.81643221e-02 3.96062702e-01 5.31372607e-01
7.26960525e-02 3.16902876e-01 2.57815897e-01 -4.60465074e-01
-1.02670896e+00 3.28360796e-01 2.28453636e-01 2.64111042e-01
-4.14434701e-01 8.44317853e-01 2.15562522e-01 -6.74810469e-01
2.70440429e-01 -4.78569448e-01 -5.80318570e-02 4.01185378e-02
3.93361717e-01 3.41493547e-01 -9.38365236e-02 -8.97895515e-01
-7.12190792e-02 2.59354055e-01 3.68466824e-01 -9.76736903e-01
1.13385844e+00 -4.54492688e-01 2.38400146e-01 1.09576583e+00
1.39352465e+00 4.99972016e-01 -1.23001158e+00 -2.45083138e-01
-4.22339648e-01 -6.72346577e-02 1.60572916e-01 -5.32688141e-01
-7.16701806e-01 1.17800474e+00 3.81620407e-01 2.97555655e-01
1.00090075e+00 -3.37373763e-02 1.28597510e+00 -2.84186993e-02
9.56522450e-02 -8.83701921e-01 1.34569541e-01 5.81945956e-01
1.13466358e+00 -7.86301613e-01 -6.93200111e-01 -3.26630414e-01
-8.56826723e-01 1.11568439e+00 2.23979637e-01 1.48273975e-01
5.74483097e-01 5.59226811e-01 1.76687375e-01 3.09345990e-01
-9.58177507e-01 -2.36113101e-01 3.94598126e-01 6.72603726e-01
2.78556526e-01 3.71897221e-01 3.28634560e-01 1.01985383e+00
-1.06181610e+00 -4.85798597e-01 3.83449733e-01 3.02061498e-01
-3.80701900e-01 -1.14781213e+00 -3.78132999e-01 2.44648252e-02
-2.12179169e-01 -4.20837492e-01 -8.66993070e-02 1.23616382e-01
-6.34655282e-02 1.19295394e+00 3.72601557e-04 -8.09042215e-01
2.88645506e-01 4.98114288e-01 2.50660986e-01 -9.62839544e-01
-2.77830005e-01 5.00305593e-01 1.34416938e-01 -1.86380967e-01
2.23695524e-02 -6.16338015e-01 -1.40771174e+00 -2.17618444e-03
-3.99519652e-01 2.41793901e-01 7.38519847e-01 1.00210929e+00
1.00495994e-01 9.12422001e-01 1.06271577e+00 -9.06889439e-01
-6.04833603e-01 -1.14240324e+00 -6.06843591e-01 -6.17563389e-02
8.49545598e-01 -3.72926891e-01 -6.82955861e-01 2.65851706e-01] | [14.995444297790527, 6.544651031494141] |
2be43fc3-52bc-411d-9e1a-b8fdc0ea80e7 | attentional-aggregation-of-deep-feature-sets | 1808.00758 | null | https://arxiv.org/abs/1808.00758v2 | https://arxiv.org/pdf/1808.00758v2.pdf | Robust Attentional Aggregation of Deep Feature Sets for Multi-view 3D Reconstruction | We study the problem of recovering an underlying 3D shape from a set of images. Existing learning based approaches usually resort to recurrent neural nets, e.g., GRU, or intuitive pooling operations, e.g., max/mean poolings, to fuse multiple deep features encoded from input images. However, GRU based approaches are unable to consistently estimate 3D shapes given different permutations of the same set of input images as the recurrent unit is permutation variant. It is also unlikely to refine the 3D shape given more images due to the long-term memory loss of GRU. Commonly used pooling approaches are limited to capturing partial information, e.g., max/mean values, ignoring other valuable features. In this paper, we present a new feed-forward neural module, named AttSets, together with a dedicated training algorithm, named FASet, to attentively aggregate an arbitrarily sized deep feature set for multi-view 3D reconstruction. The AttSets module is permutation invariant, computationally efficient and flexible to implement, while the FASet algorithm enables the AttSets based network to be remarkably robust and generalize to an arbitrary number of input images. We thoroughly evaluate FASet and the properties of AttSets on multiple large public datasets. Extensive experiments show that AttSets together with FASet algorithm significantly outperforms existing aggregation approaches. | ['Sen Wang', 'Andrew Markham', 'Niki Trigoni', 'Bo Yang'] | 2018-08-02 | null | null | null | null | ['3d-object-reconstruction'] | ['computer-vision'] | [ 6.79105008e-03 -3.22260223e-02 3.73011865e-02 -3.50668877e-01
-9.18499231e-01 -8.14880013e-01 6.01985335e-01 -1.59597024e-01
-5.00206463e-02 4.62146699e-01 3.77821386e-01 6.92700371e-02
-3.41388196e-01 -8.64955246e-01 -1.21033525e+00 -8.70542049e-01
-7.61893988e-02 4.26264852e-01 -7.23853931e-02 1.17236711e-02
2.21477076e-01 8.82335365e-01 -1.68296111e+00 5.18942058e-01
3.18241447e-01 1.38428843e+00 3.09490323e-01 4.16758627e-01
-1.32461578e-01 1.99601576e-01 -3.18798184e-01 -3.06614816e-01
7.36515164e-01 -8.82674828e-02 -7.21636593e-01 2.77087927e-01
6.15260005e-01 -4.12068158e-01 -2.42521122e-01 8.56404364e-01
6.28961146e-01 2.88383782e-01 7.01509297e-01 -1.03845406e+00
-1.01158512e+00 4.92059231e-01 -6.44342422e-01 5.32367453e-02
2.79903382e-01 -4.48853932e-02 1.15817106e+00 -1.28920400e+00
6.90928519e-01 1.28441560e+00 6.90594673e-01 2.90878117e-01
-1.19026840e+00 -3.01268965e-01 2.17751801e-01 -1.07069582e-01
-1.42777717e+00 -3.37326348e-01 7.57718325e-01 -6.37652278e-02
1.35047174e+00 3.43792200e-01 5.46563327e-01 9.26014364e-01
2.98089564e-01 9.42037046e-01 1.07842362e+00 -1.03956603e-01
3.87644842e-02 -2.72766441e-01 5.03266193e-02 8.74671578e-01
3.36373091e-01 -7.18581304e-03 -5.44006705e-01 -2.19526574e-01
9.77556944e-01 3.46269995e-01 -1.93665326e-01 -5.68822086e-01
-1.30882537e+00 6.89674854e-01 9.64950085e-01 1.14902146e-01
-5.71149886e-01 1.41426966e-01 1.75318897e-01 3.60510945e-01
4.92359459e-01 4.77877706e-01 -6.82454586e-01 4.27126437e-01
-6.10978842e-01 3.96601886e-01 5.41166961e-01 1.01229227e+00
1.03413272e+00 -2.42668375e-01 -1.31259128e-01 7.95269668e-01
5.69529012e-02 2.07376480e-01 4.17086184e-01 -9.08464313e-01
4.35258746e-01 5.64207137e-01 -9.64485854e-02 -8.44264269e-01
-5.61238527e-01 -2.47803867e-01 -1.18943906e+00 5.60376868e-02
1.30247593e-01 1.89786460e-02 -1.18306601e+00 1.65221536e+00
4.79369313e-02 1.69229925e-01 6.63180128e-02 8.21093202e-01
1.10636055e+00 6.19066715e-01 -4.67734337e-01 1.19138107e-01
1.21561015e+00 -7.62908280e-01 -7.97686353e-03 -1.09027326e-01
2.68542588e-01 -7.29587138e-01 7.49611139e-01 2.87232995e-01
-1.36949956e+00 -5.21131396e-01 -9.13830996e-01 -1.26496077e-01
-5.99167347e-01 1.10499948e-01 6.50192976e-01 3.70181352e-01
-1.40717018e+00 7.95529962e-01 -7.58928239e-01 -2.32498065e-01
7.60402977e-01 7.18661308e-01 -8.21839392e-01 -3.12136054e-01
-6.60925865e-01 7.13444769e-01 5.54128945e-01 1.74999118e-01
-5.27813494e-01 -6.75322115e-01 -1.00272155e+00 2.14338616e-01
3.70043159e-01 -9.42154288e-01 9.80021894e-01 -7.32239366e-01
-1.40372717e+00 8.78084183e-01 -1.91807911e-01 -4.18394804e-01
1.62052840e-01 -3.97155285e-02 1.01282112e-01 9.42136943e-02
1.51414901e-03 1.09238052e+00 1.01436031e+00 -1.35369754e+00
-3.66369665e-01 -4.21305239e-01 8.13795403e-02 2.62197822e-01
-9.99341831e-02 -1.58423096e-01 -4.44681585e-01 -6.74099088e-01
5.65797865e-01 -9.44903791e-01 -3.62268329e-01 -5.86147532e-02
-5.55527925e-01 -2.50958860e-01 8.08842421e-01 -2.60016918e-01
5.14624655e-01 -2.01941681e+00 3.49619865e-01 3.28495115e-01
3.33895057e-01 4.40295078e-02 -4.42887574e-01 3.39983433e-01
-2.92346656e-01 2.52109230e-01 -3.93293023e-01 -3.65847975e-01
2.83212662e-02 4.61749494e-01 -3.21671307e-01 3.48998725e-01
5.35665512e-01 1.24587584e+00 -5.90607285e-01 -4.52465490e-02
2.74448723e-01 4.37264323e-01 -8.31811607e-01 8.66912454e-02
-1.41298130e-01 3.08015734e-01 -4.09511060e-01 7.16344476e-01
8.68082881e-01 -5.31829357e-01 -5.53362891e-02 -3.74316633e-01
9.08021070e-03 1.91950589e-01 -1.11560285e+00 1.88307178e+00
-3.77336234e-01 1.70208916e-01 -4.00407165e-01 -1.13660622e+00
1.19286704e+00 1.82078987e-01 5.85546374e-01 -3.77702624e-01
1.37185171e-01 3.33492845e-01 -2.40673229e-01 -2.75111854e-01
5.89273691e-01 -1.60092469e-02 -2.37689823e-01 6.59640372e-01
3.99654835e-01 -1.69949457e-01 7.73963099e-03 -1.19742349e-01
1.04635334e+00 1.56932071e-01 3.90942454e-01 -1.02968998e-01
3.62105519e-01 -5.07009029e-01 4.31449264e-01 8.27551067e-01
2.04820737e-01 1.13410199e+00 3.88254762e-01 -8.67630661e-01
-1.09495699e+00 -1.26766694e+00 -6.82943240e-02 8.46210539e-01
1.13409512e-01 -1.50092915e-01 -3.81327540e-01 -8.75490427e-01
1.59621581e-01 1.13258123e-01 -6.64108872e-01 9.69014317e-03
-6.34676754e-01 -9.65992808e-01 3.02219719e-01 6.50945485e-01
6.98451161e-01 -1.08509707e+00 -5.11724114e-01 1.36315808e-01
-6.53095990e-02 -1.05273354e+00 -5.63992858e-01 2.32176691e-01
-1.00922227e+00 -8.16361606e-01 -7.99652219e-01 -8.33094895e-01
8.42451155e-01 5.02953231e-01 1.08700848e+00 -4.86743934e-02
-7.39680529e-02 2.10702091e-01 -2.83248425e-01 -2.47407630e-01
5.27227111e-02 1.60262316e-01 5.64190559e-02 1.77214324e-01
2.52262235e-01 -8.23949516e-01 -5.62323928e-01 4.07330692e-01
-9.79629755e-01 2.81062368e-02 9.39827621e-01 1.06594133e+00
1.02298105e+00 -5.49738333e-02 6.08369172e-01 -6.82022572e-01
4.18145388e-01 -4.77232128e-01 -3.38506132e-01 2.65380293e-01
-3.51906791e-02 3.01551431e-01 6.26968980e-01 -2.99079716e-01
-7.95690894e-01 2.59721190e-01 -2.11683586e-01 -7.41499603e-01
-1.65891618e-01 5.78278482e-01 -2.70261735e-01 -2.38478154e-01
5.25192022e-01 2.78627425e-01 -2.57901289e-02 -5.94295204e-01
3.67575854e-01 3.09132129e-01 5.86689949e-01 -4.31255639e-01
5.87732017e-01 5.53201497e-01 9.88368988e-02 -5.76358378e-01
-8.42831969e-01 -2.10077763e-01 -7.74435461e-01 1.22607641e-01
6.76756024e-01 -9.09493268e-01 -5.06410301e-01 5.33614933e-01
-1.14034963e+00 -4.08993512e-02 -2.92612046e-01 3.54843289e-01
-9.28250074e-01 1.70880020e-01 -5.02772868e-01 -2.74544418e-01
-4.59693164e-01 -1.19432139e+00 1.37066674e+00 1.34101510e-01
-1.88887387e-03 -7.49077737e-01 -2.50723839e-01 1.95810527e-01
1.94289595e-01 4.04681385e-01 9.71223772e-01 -6.46948516e-01
-9.73085761e-01 -2.83758759e-01 -5.45355439e-01 3.87051761e-01
2.87556887e-01 -3.44706059e-01 -1.05859029e+00 -3.64322335e-01
-1.65418789e-01 -4.39764291e-01 1.19401932e+00 7.25187600e-01
1.63369310e+00 -5.02459586e-01 -1.32430792e-01 7.67773986e-01
1.40703964e+00 -1.68371245e-01 7.49646544e-01 6.13762587e-02
6.40787661e-01 3.01552653e-01 1.11110970e-01 3.08018059e-01
3.53405714e-01 4.55452651e-01 4.56668049e-01 2.35825684e-03
-1.20905451e-01 -1.19948946e-01 3.43320012e-01 8.28669548e-01
-4.00629401e-01 -3.73023242e-01 -6.14861488e-01 3.30124825e-01
-1.68337083e+00 -8.43133032e-01 5.31944454e-01 2.08144093e+00
3.86706263e-01 8.51625577e-02 -1.00638255e-01 -1.01578414e-01
6.21886730e-01 3.61925721e-01 -5.76875508e-01 -4.59531456e-01
-4.20787752e-01 3.80612373e-01 5.21173954e-01 8.95970613e-02
-1.17884481e+00 6.75846875e-01 5.96365452e+00 6.94236934e-01
-1.00856686e+00 -9.77399871e-02 7.09597111e-01 -2.27127269e-01
-5.33173740e-01 -3.54744256e-01 -6.62878394e-01 1.09575950e-01
3.53484720e-01 8.39752033e-02 5.00928640e-01 4.29278076e-01
-1.71211466e-01 1.58826932e-01 -1.11504555e+00 1.13809597e+00
4.33674157e-01 -1.76688802e+00 4.58721012e-01 9.57398638e-02
9.63562012e-01 3.83806378e-01 1.86326027e-01 5.10737784e-02
3.40509921e-01 -1.08730447e+00 6.86296642e-01 6.52184904e-01
8.17072988e-01 -8.73866498e-01 8.33007991e-01 9.73195434e-02
-1.15549684e+00 -1.70905292e-01 -5.27626038e-01 5.18783331e-02
3.05290818e-02 4.54873145e-01 -5.95179141e-01 9.11635697e-01
9.14836109e-01 1.00196767e+00 -4.71637785e-01 1.07058847e+00
1.06638566e-01 -1.52076095e-01 -5.70937634e-01 8.86001289e-02
5.53783596e-01 -9.97005403e-02 6.40398562e-01 7.37260997e-01
6.78374350e-01 1.48821235e-01 2.07196921e-01 7.46168971e-01
-5.10195613e-01 -1.39349073e-01 -9.43435907e-01 1.42840490e-01
2.26022959e-01 1.29057014e+00 -8.47147644e-01 -1.37486637e-01
-4.59547877e-01 9.04757798e-01 5.76019645e-01 3.87415737e-01
-4.38606590e-01 -1.69800311e-01 6.41534507e-01 -1.18581541e-01
8.89314830e-01 -2.98117071e-01 -3.47814888e-01 -1.09661543e+00
1.87566191e-01 -6.11625075e-01 4.98761594e-01 -8.27089787e-01
-1.50344574e+00 7.69773781e-01 6.04563057e-02 -1.35551834e+00
-1.35467678e-01 -7.50731468e-01 -6.33087277e-01 8.26626897e-01
-1.51181054e+00 -1.24479437e+00 -8.71852338e-02 5.48610032e-01
4.91455704e-01 -2.58381039e-01 7.96603739e-01 1.61316562e-02
-4.33846653e-01 4.58137780e-01 -9.41311121e-02 5.68823256e-02
4.96909529e-01 -1.15583277e+00 5.83753288e-01 6.04174733e-01
3.97752672e-01 5.14360785e-01 3.08001220e-01 -3.41070205e-01
-1.64238632e+00 -1.31233978e+00 7.44499087e-01 -3.97736430e-01
3.34673285e-01 -3.76437217e-01 -9.28116083e-01 8.56519759e-01
4.10565920e-02 5.79715967e-01 5.39569914e-01 -1.27493534e-02
-4.67042267e-01 4.30197157e-02 -1.21276283e+00 4.30039704e-01
1.28776574e+00 -4.08889472e-01 -7.71288574e-01 1.61036476e-01
6.51903749e-01 -6.76422775e-01 -9.39553142e-01 5.06902695e-01
6.43488169e-01 -1.00748324e+00 1.35112047e+00 -6.22862041e-01
5.44978380e-01 -2.11899132e-01 -5.36693394e-01 -1.34875298e+00
-5.02184868e-01 -3.36414903e-01 -1.39156565e-01 8.16862881e-01
4.62115347e-01 -8.43401968e-01 7.46885061e-01 1.54510647e-01
-3.26910973e-01 -1.04140365e+00 -1.09734225e+00 -8.14432025e-01
-7.33909663e-03 -3.48603815e-01 1.11823869e+00 6.61422789e-01
-5.47047198e-01 7.08867386e-02 -3.00123870e-01 3.80783707e-01
5.24278581e-01 6.40499651e-01 6.54080749e-01 -1.17399275e+00
-2.61503339e-01 -5.99379778e-01 -5.91103733e-01 -1.06151044e+00
1.61756441e-01 -1.32444453e+00 -5.01395166e-02 -1.49967837e+00
2.96209812e-01 -5.40395021e-01 -4.88350362e-01 7.64679790e-01
2.16905680e-02 7.11800396e-01 2.10811138e-01 8.05075616e-02
-4.97498572e-01 6.66984558e-01 1.46081007e+00 -2.24372745e-01
-4.00719285e-01 6.07640930e-02 -7.81007290e-01 7.53882229e-01
8.20070982e-01 -2.53737032e-01 -1.60767004e-01 -6.54563546e-01
2.18102872e-01 -2.03495786e-01 6.67345703e-01 -5.91934979e-01
4.50532548e-02 2.48175576e-01 8.41398001e-01 -8.46037447e-01
4.28848505e-01 -5.45386136e-01 1.19923182e-01 1.18308157e-01
-6.80628652e-03 1.48989618e-01 3.76655430e-01 4.46754485e-01
-2.28209451e-01 3.70049402e-02 5.08192837e-01 -3.92471164e-01
-7.31206119e-01 7.98776150e-01 -3.64538096e-02 -2.56740600e-01
6.27112925e-01 -4.62888986e-01 -1.78453818e-01 -2.44683281e-01
-7.37959623e-01 1.13482345e-02 2.51161098e-01 5.47046661e-01
9.05025899e-01 -1.58579469e+00 -6.80068016e-01 5.88677585e-01
1.58016369e-01 2.77816564e-01 4.12793577e-01 6.88336492e-01
-1.27822399e-01 3.32551926e-01 -5.18894851e-01 -6.86776698e-01
-1.02098429e+00 4.55411285e-01 3.37383837e-01 -3.17007154e-01
-8.96319747e-01 9.07645762e-01 4.46687251e-01 -8.42765987e-01
-1.46614850e-01 -6.30128860e-01 -1.00119762e-01 1.56646863e-01
5.01119733e-01 -1.23248799e-02 3.87402713e-01 -5.48565745e-01
-2.48426527e-01 8.43615770e-01 -1.53364301e-01 2.24772662e-01
1.77427959e+00 -1.00221030e-01 -1.78390220e-01 1.88078120e-01
1.46179521e+00 -3.25967997e-01 -1.39538884e+00 -5.15832841e-01
-2.91137278e-01 -4.93162930e-01 3.36344354e-02 -5.56453526e-01
-1.32713544e+00 5.86842537e-01 2.38569975e-01 1.62566230e-01
1.38496590e+00 2.72016168e-01 8.36562932e-01 6.76896632e-01
4.25210118e-01 -6.48545861e-01 3.38225514e-02 5.73983133e-01
1.33952510e+00 -1.23984599e+00 -1.43636661e-02 -2.80050695e-01
-4.61875707e-01 1.17155564e+00 4.29011226e-01 -4.21294034e-01
7.63456285e-01 1.35233864e-01 -3.69228691e-01 -2.47628972e-01
-7.17052758e-01 -2.42694229e-01 5.58954597e-01 4.06226903e-01
-1.80633757e-02 7.00873733e-02 2.16597721e-01 5.98196805e-01
-3.45561594e-01 -4.58850935e-02 3.45442504e-01 7.56118000e-01
-2.66738981e-01 -9.72139955e-01 -4.29047406e-01 8.90654624e-01
-1.68303221e-01 2.34713908e-02 -2.16068864e-01 6.89027548e-01
2.21668318e-01 3.22095156e-01 4.74558830e-01 -4.52926129e-01
3.21934432e-01 -8.47705454e-02 8.11470687e-01 -5.40783167e-01
-4.95139390e-01 2.05436572e-02 -1.91209078e-01 -5.95165193e-01
-6.20090246e-01 -9.30494189e-01 -1.05197155e+00 -3.46557200e-01
-1.35013655e-01 -4.19231266e-01 2.25490063e-01 1.00388443e+00
5.69989085e-01 2.17372268e-01 9.07155573e-01 -1.31469584e+00
-2.63851523e-01 -6.84946358e-01 -5.42252481e-01 1.60415247e-01
4.55337554e-01 -7.85485923e-01 -7.00449049e-02 -2.27196977e-01] | [8.153120040893555, -3.5992496013641357] |
58caeb88-697f-483f-be25-3d28983670de | troubleshooting-blind-image-quality-models-in | 2105.06747 | null | https://arxiv.org/abs/2105.06747v1 | https://arxiv.org/pdf/2105.06747v1.pdf | Troubleshooting Blind Image Quality Models in the Wild | Recently, the group maximum differentiation competition (gMAD) has been used to improve blind image quality assessment (BIQA) models, with the help of full-reference metrics. When applying this type of approach to troubleshoot "best-performing" BIQA models in the wild, we are faced with a practical challenge: it is highly nontrivial to obtain stronger competing models for efficient failure-spotting. Inspired by recent findings that difficult samples of deep models may be exposed through network pruning, we construct a set of "self-competitors," as random ensembles of pruned versions of the target model to be improved. Diverse failures can then be efficiently identified via self-gMAD competition. Next, we fine-tune both the target and its pruned variants on the human-rated gMAD set. This allows all models to learn from their respective failures, preparing themselves for the next round of self-gMAD competition. Experimental results demonstrate that our method efficiently troubleshoots BIQA models in the wild with improved generalizability. | ['Kede Ma', 'Zhangyang Wang', 'Tianlong Chen', 'Haotao Wang', 'Zhihua Wang'] | 2021-05-14 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Wang_Troubleshooting_Blind_Image_Quality_Models_in_the_Wild_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Wang_Troubleshooting_Blind_Image_Quality_Models_in_the_Wild_CVPR_2021_paper.pdf | cvpr-2021-1 | ['blind-image-quality-assessment'] | ['computer-vision'] | [ 5.29027581e-02 -7.44139180e-02 2.57334024e-01 -3.31875384e-01
-1.28209376e+00 -6.37234569e-01 2.43962929e-01 -3.21547687e-01
-4.49456155e-01 6.32116318e-01 1.83862612e-01 -2.31145307e-01
-4.60729271e-01 -1.95003837e-01 -5.67031205e-01 -7.25706816e-01
-9.16416124e-02 6.45842910e-01 2.17082575e-01 -2.53020674e-01
2.97441959e-01 4.24510807e-01 -1.54857016e+00 5.33649862e-01
1.20291066e+00 6.47662222e-01 2.36414209e-01 8.32741857e-01
5.08139610e-01 7.49223828e-01 -1.03059769e+00 -9.48825240e-01
4.47320998e-01 -3.79833966e-01 -9.48705077e-01 -5.82934022e-02
8.82424116e-01 -3.41140896e-01 -2.55626917e-01 1.23098111e+00
1.07848179e+00 1.80991814e-01 4.15634632e-01 -1.05979311e+00
-5.50894976e-01 6.37481451e-01 -4.09869015e-01 6.77516818e-01
1.54520303e-01 7.15877533e-01 1.25965178e+00 -7.93550372e-01
3.69337708e-01 1.24315190e+00 6.17748916e-01 9.23528731e-01
-1.58309317e+00 -7.08535492e-01 2.28910461e-01 5.26203215e-01
-1.41666269e+00 -6.37342215e-01 5.42253971e-01 -3.25791687e-01
9.83925581e-01 4.16245013e-01 2.47946844e-01 1.38323891e+00
-1.71238989e-01 6.74287498e-01 1.20083547e+00 -3.22143853e-01
3.16722363e-01 2.54259389e-02 3.50521170e-02 6.90616846e-01
-2.84853810e-03 3.90225112e-01 -7.49043405e-01 -3.57639343e-01
4.59015071e-01 -6.56720400e-01 -6.35817647e-01 -3.14601123e-01
-9.68760133e-01 6.21182501e-01 6.25559866e-01 1.78526029e-01
-3.61034751e-01 -9.93839800e-02 5.23643158e-02 4.77323025e-01
1.98897421e-01 1.11043930e+00 -5.59515834e-01 -3.90119888e-02
-1.09046149e+00 3.86811525e-01 6.37295783e-01 3.33097905e-01
6.66294694e-01 -3.52308691e-01 -5.82735062e-01 1.20292842e+00
-1.59587525e-02 2.27773592e-01 3.94195884e-01 -1.23662007e+00
4.05747771e-01 4.06652659e-01 2.48340026e-01 -5.98337114e-01
-2.75711864e-01 -8.21967065e-01 -8.01820636e-01 5.75320005e-01
5.28883219e-01 -3.60314287e-02 -1.13324034e+00 1.86446059e+00
4.17573228e-02 1.97312176e-01 -2.59050518e-01 1.14135849e+00
4.41499770e-01 1.86058551e-01 1.27195567e-01 -8.01284686e-02
9.47724462e-01 -1.07973862e+00 -7.24957557e-03 -1.70498043e-01
5.73178947e-01 -4.81014073e-01 1.40762150e+00 1.05188107e+00
-1.16390789e+00 -6.08864725e-01 -9.43087041e-01 2.52273768e-01
1.47026286e-01 -1.70279052e-02 2.80035347e-01 7.37827301e-01
-1.62873220e+00 9.55627322e-01 -6.01519465e-01 -3.18166148e-03
7.87446856e-01 7.07027495e-01 -2.32093900e-01 -2.55799115e-01
-9.14155960e-01 1.09570384e+00 4.24233228e-02 1.58206299e-01
-1.65837073e+00 -7.40432739e-01 -2.10881293e-01 -5.76077886e-02
3.93190712e-01 -1.14958358e+00 1.26804972e+00 -1.22528052e+00
-1.32462513e+00 9.25474763e-01 -2.34629288e-01 -5.14722466e-01
5.92813730e-01 -2.25277603e-01 -5.32090664e-01 2.52890766e-01
6.22466803e-02 7.20609426e-01 1.09444726e+00 -1.27317131e+00
-4.92088735e-01 -3.27593863e-01 3.40216279e-01 3.57153296e-01
-9.72894058e-02 2.73344159e-01 -3.54609013e-01 -5.27379751e-01
-1.11769987e-02 -6.91087842e-01 -3.93223703e-01 -1.26305729e-01
-7.87597060e-01 -3.21467221e-01 -1.42571390e-01 -6.46796107e-01
1.12394357e+00 -2.14287329e+00 3.38746130e-01 4.14020896e-01
7.76008606e-01 3.80209655e-01 -6.51684701e-01 -9.70494375e-02
-1.73353195e-01 4.04615998e-01 -8.99245366e-02 -6.52685285e-01
-1.54381335e-01 3.34394886e-03 -7.77306454e-03 2.61626542e-01
4.38655198e-01 6.68386161e-01 -8.00703049e-01 -2.13145897e-01
-1.42728731e-01 2.72302359e-01 -8.50666046e-01 3.13645899e-01
-1.79095179e-01 7.08630264e-01 -1.48207396e-01 4.76580977e-01
6.51660800e-01 -5.51132619e-01 -1.25300169e-01 -1.29060924e-01
4.70145166e-01 5.31663001e-01 -1.00855052e+00 1.48158622e+00
-3.20585191e-01 3.29763442e-01 -1.36463404e-01 -8.19325984e-01
6.14430010e-01 2.07031116e-01 1.44823208e-01 -8.45063329e-01
-7.88089037e-02 1.28802270e-01 3.56999606e-01 -4.00251091e-01
2.43048698e-01 -6.49220869e-02 4.24228668e-01 3.04294884e-01
4.30119306e-01 2.89368451e-01 2.12029487e-01 3.27907443e-01
1.54222286e+00 -2.39718199e-01 -3.09908211e-01 -1.19724922e-01
3.67248327e-01 -1.28048286e-01 5.25914073e-01 1.32048488e+00
-4.42676127e-01 1.21050096e+00 4.91394848e-01 -4.24182445e-01
-1.08458328e+00 -1.49117041e+00 2.38245204e-02 1.16838288e+00
-6.00077249e-02 -3.21127713e-01 -7.58718193e-01 -1.04699636e+00
-3.07501972e-01 5.33474028e-01 -6.90523267e-01 -3.57755601e-01
-4.45697039e-01 -1.02506745e+00 5.80638647e-01 1.56516150e-01
2.53252327e-01 -1.13916266e+00 -1.46373764e-01 3.64030339e-02
-2.13960022e-01 -6.57062590e-01 -2.37095907e-01 1.65458977e-01
-7.61875570e-01 -1.14945710e+00 -8.97166371e-01 -5.81025422e-01
6.50686800e-01 1.35508075e-01 1.64348769e+00 5.22802591e-01
-1.12825125e-01 4.25260141e-02 -3.61098349e-01 3.35860997e-02
-6.78795755e-01 5.00206500e-02 1.17850356e-01 1.18188433e-01
2.72343129e-01 -6.48401856e-01 -9.47822630e-01 5.30566812e-01
-6.41933322e-01 -1.83308780e-01 7.51188695e-01 9.84932005e-01
5.40319681e-01 1.78607494e-01 7.58286536e-01 -6.94771349e-01
7.48392224e-01 -2.77699739e-01 -5.24169087e-01 4.50699657e-01
-7.12428629e-01 2.13685483e-01 4.82614249e-01 -7.28874087e-01
-6.50962710e-01 -3.95723343e-01 -1.33272275e-01 -6.03108764e-01
-1.37984768e-01 2.70930171e-01 -2.42128074e-01 -3.34879279e-01
1.27696347e+00 -7.98541605e-02 -1.21281460e-01 -8.04544687e-01
4.13870394e-01 3.10850590e-01 7.47960627e-01 -6.15959048e-01
7.58039117e-01 8.40097293e-02 -5.36141336e-01 -2.01411545e-01
-1.00302243e+00 -4.16186929e-01 -1.79511771e-01 -1.19585678e-01
5.64018071e-01 -8.78192782e-01 -6.00846648e-01 5.03341734e-01
-1.02424169e+00 -4.64088082e-01 -3.17384899e-01 1.84362710e-01
-3.53988796e-01 5.40905237e-01 -4.32423651e-01 -5.44020116e-01
-2.23505527e-01 -1.41221702e+00 7.71871924e-01 3.03319901e-01
-1.62954107e-01 -4.34827834e-01 2.79006183e-01 7.10749626e-01
2.99234837e-01 -4.57362056e-01 1.01253390e+00 -7.49462664e-01
-5.33337653e-01 9.43781212e-02 -2.42506593e-01 8.60327959e-01
-3.07995737e-01 -2.65854508e-01 -1.27790129e+00 -6.05970740e-01
-8.93180445e-02 -4.80690747e-01 1.01584101e+00 3.10333699e-01
1.34140253e+00 -3.00599754e-01 -4.64330316e-02 6.41810775e-01
1.15153360e+00 -2.43356526e-01 7.73126602e-01 4.66615379e-01
3.16389829e-01 4.16809589e-01 9.40396413e-02 1.98816374e-01
2.02949271e-01 5.94367683e-01 5.08249044e-01 3.96269634e-02
-3.87874246e-01 -6.35463074e-02 3.61134470e-01 5.35243571e-01
-1.03742972e-01 -3.64423275e-01 -9.43068445e-01 7.53982425e-01
-1.42208469e+00 -7.47310042e-01 1.91813692e-01 2.35227084e+00
9.49377298e-01 3.55427742e-01 3.52668703e-01 7.79216439e-02
7.92858839e-01 -1.14338741e-01 -6.00172341e-01 -1.27394702e-02
-4.56924409e-01 6.05977476e-01 1.34211108e-01 4.07742381e-01
-1.01634681e+00 7.55920112e-01 6.73740101e+00 8.76898646e-01
-7.19393969e-01 3.49537849e-01 9.47884619e-01 -5.15568614e-01
-1.89090073e-01 1.64954245e-01 -6.68739259e-01 5.95511854e-01
1.16653073e+00 2.48257414e-01 8.05918932e-01 4.31095809e-01
1.47775680e-01 -1.31982967e-01 -1.02323830e+00 9.59310293e-01
-3.31167132e-02 -1.08676195e+00 2.29907691e-01 -1.69969931e-01
9.62038398e-01 4.89589036e-01 3.16971153e-01 2.12940678e-01
6.89824343e-01 -1.29715550e+00 4.55731064e-01 6.98630393e-01
6.49329603e-01 -6.32364631e-01 7.00725436e-01 1.13258719e-01
-3.18215072e-01 -4.88439620e-01 -6.91508830e-01 2.53879607e-01
3.30120586e-02 7.50211835e-01 -8.98008466e-01 4.20566082e-01
9.78826106e-01 1.85793325e-01 -1.17141533e+00 1.80131459e+00
-2.18390241e-01 7.88685143e-01 -2.36265957e-01 3.47324103e-01
-4.34365533e-02 2.90411115e-01 7.85645127e-01 5.92870295e-01
1.46851286e-01 -1.96069971e-01 -1.79287925e-01 1.04199874e+00
-1.97959006e-01 -2.32154787e-01 -1.11930132e-01 2.42661640e-01
3.43177050e-01 9.88546073e-01 -1.88642725e-01 -2.06675932e-01
-1.54194519e-01 1.02440441e+00 6.66198194e-01 5.03364027e-01
-5.25603890e-01 6.44454616e-04 5.34347773e-01 1.24573417e-01
2.63559163e-01 2.00010628e-01 -2.07318157e-01 -1.25419533e+00
6.34681955e-02 -1.58662677e+00 5.18479586e-01 -1.06406510e+00
-1.80124795e+00 1.10531473e+00 -4.47935224e-01 -1.27971911e+00
1.69730990e-03 -4.97364104e-01 -5.71347713e-01 1.14961743e+00
-1.47455156e+00 -7.29223609e-01 -1.64445266e-01 8.31040919e-01
2.73255765e-01 -3.06312978e-01 7.19241440e-01 4.64693964e-01
-6.30158365e-01 1.10679877e+00 -5.15269376e-02 -1.17735557e-01
9.11886334e-01 -1.26624811e+00 3.68637413e-01 1.17407608e+00
5.34952343e-01 5.10675848e-01 7.89134860e-01 -4.93467033e-01
-6.21147156e-01 -8.51064026e-01 6.83144212e-01 -8.76319289e-01
5.65597236e-01 -1.31030828e-01 -1.15048134e+00 2.45415986e-01
-6.51168525e-02 -8.82875465e-04 4.48721081e-01 4.88704562e-01
-5.36445975e-01 -1.42385304e-01 -1.05217564e+00 6.20804250e-01
1.16682124e+00 -8.30968261e-01 -4.90822196e-01 4.41498697e-01
6.84228837e-01 -1.20354056e-01 -5.09164035e-01 5.44028103e-01
1.63222611e-01 -1.25762939e+00 9.50429320e-01 -9.03138995e-01
1.44061342e-01 -3.41862351e-01 3.32936198e-02 -1.51242185e+00
-4.39586669e-01 -8.11544001e-01 5.17292283e-02 1.00228333e+00
6.98518336e-01 -2.54158795e-01 8.30070376e-01 4.50023592e-01
-1.63575664e-01 -5.53211689e-01 -1.10669506e+00 -6.95592165e-01
3.58513966e-02 -3.92103076e-01 4.64023948e-01 5.55113673e-01
-5.00054598e-01 2.16291219e-01 -4.79218543e-01 6.38234138e-01
6.38878167e-01 -3.38602006e-01 6.12136543e-01 -1.34686708e+00
-9.35886979e-01 -6.64387047e-01 -3.54406387e-01 -1.12337089e+00
-2.91578740e-01 -8.46977949e-01 2.13517919e-01 -1.13306868e+00
3.60947818e-01 -6.04702890e-01 -7.60290146e-01 2.79811502e-01
-5.88484764e-01 4.38700050e-01 1.26871139e-01 4.69325721e-01
-1.04017305e+00 3.99962276e-01 1.13126731e+00 -1.91795975e-01
-2.44029105e-01 2.11241335e-01 -9.30700898e-01 3.17730844e-01
6.07830584e-01 -6.72098219e-01 -2.76391625e-01 -6.95231974e-01
5.35244226e-01 -1.62304938e-01 8.13037336e-01 -1.30734038e+00
1.63627073e-01 2.79737890e-01 2.08692446e-01 -3.28373760e-01
2.48749480e-01 -2.93481469e-01 1.24917808e-03 1.68696985e-01
-5.45725167e-01 1.51359811e-01 3.02927122e-02 4.36735213e-01
8.65895599e-02 -4.08791512e-01 8.74833226e-01 -1.29393548e-01
-5.84490836e-01 3.35993320e-01 -2.88632140e-02 2.02890098e-01
2.71119416e-01 1.61754657e-02 -6.40334725e-01 -3.04443955e-01
-1.26634192e+00 1.80760637e-01 3.86554599e-01 2.18952984e-01
6.18543327e-01 -8.60684633e-01 -9.11678374e-01 2.65383989e-01
1.53645635e-01 2.12679803e-02 6.48383796e-01 1.01194835e+00
-2.64214873e-01 -5.34301549e-02 -2.35638633e-01 -6.16442978e-01
-9.33749557e-01 5.26673257e-01 9.20812905e-01 -4.92654294e-01
-4.90683019e-01 1.40783334e+00 3.11325282e-01 -2.22117841e-01
3.49187344e-01 1.96973622e-01 -1.61252797e-01 -3.71381700e-01
6.97463274e-01 3.35036546e-01 4.73565698e-01 -2.50371575e-01
-2.16569796e-01 1.72436517e-02 -3.94441485e-01 -2.95823455e-01
1.06114566e+00 -1.86223045e-01 -8.55604708e-02 -4.18507904e-02
8.48333061e-01 -1.19930143e-02 -1.30684483e+00 -3.48030388e-01
1.05757847e-01 -5.57407737e-01 7.37189725e-02 -1.61654174e+00
-1.07947683e+00 9.03906167e-01 9.84662771e-01 8.93966705e-02
1.40003586e+00 1.63239583e-01 3.29599440e-01 3.71923625e-01
3.98104519e-01 -7.90968597e-01 3.42786044e-01 1.14490062e-01
9.29213107e-01 -1.15470862e+00 -4.17046875e-01 1.91148013e-01
-5.96878290e-01 5.80565155e-01 7.14070261e-01 -2.95437545e-01
2.81624436e-01 -1.33461863e-01 3.25100005e-01 -3.09369057e-01
-9.69610214e-01 -9.49868783e-02 5.83582342e-01 9.37300026e-01
-1.14030585e-01 7.66252130e-02 1.46122843e-01 7.34903336e-01
-3.46698016e-01 -2.11361334e-01 4.79151309e-01 2.38249838e-01
-3.26581687e-01 -1.11188257e+00 -2.66221106e-01 8.70010674e-01
-5.91063619e-01 -4.53710437e-01 -3.26912403e-01 2.89928705e-01
1.55849218e-01 1.18166208e+00 -2.37919971e-01 -7.28091896e-01
2.81928271e-01 3.36253010e-02 4.27206546e-01 -5.85711181e-01
-8.48320007e-01 -1.44212738e-01 1.43005475e-01 -8.89657080e-01
-2.28585631e-01 -3.75021785e-01 -4.23678368e-01 -3.14997345e-01
-4.54346716e-01 1.30630404e-01 2.39028767e-01 1.00888431e+00
6.14767015e-01 4.75012273e-01 6.65838659e-01 -7.16425478e-01
-9.21396255e-01 -1.05887318e+00 -3.35437119e-01 4.77663815e-01
5.64961553e-01 -7.71608651e-01 -5.84601343e-01 -3.19941074e-01] | [11.892057418823242, -1.8052126169204712] |
b5072041-9adf-4ae7-b8d6-277b7eb4fd2f | automated-essay-scoring-in-argumentative | 2307.04276 | null | https://arxiv.org/abs/2307.04276v1 | https://arxiv.org/pdf/2307.04276v1.pdf | Automated Essay Scoring in Argumentative Writing: DeBERTeachingAssistant | Automated Essay scoring has been explored as a research and industry problem for over 50 years. It has drawn a lot of attention from the NLP community because of its clear educational value as a research area that can engender the creation of valuable time-saving tools for educators around the world. Yet, these tools are generally focused on detecting good grammar, spelling mistakes, and organization quality but tend to fail at incorporating persuasiveness features in their final assessment. The responsibility to give actionable feedback to the student to improve the strength of their arguments is left solely on the teacher's shoulders. In this work, we present a transformer-based architecture capable of achieving above-human accuracy in annotating argumentative writing discourse elements for their persuasiveness quality and we expand on planned future work investigating the explainability of our model so that actionable feedback can be offered to the student and thus potentially enable a partnership between the teacher's advice and the machine's advice. | ['Choong Hee Kim', 'Karan Jha', 'Tonghua Tian', 'Yann Hicke'] | 2023-07-09 | null | null | null | null | ['automated-essay-scoring'] | ['natural-language-processing'] | [ 3.32463056e-01 6.81708932e-01 -2.18693897e-01 -5.56843996e-01
-7.31480896e-01 -6.94682240e-01 4.33036655e-01 7.72803426e-01
-1.60731092e-01 5.74297845e-01 6.17692411e-01 -9.22662079e-01
-4.23613280e-01 -6.46909952e-01 -2.97937304e-01 -2.87244856e-01
8.17239463e-01 5.62500298e-01 3.61284494e-01 -3.60227078e-01
8.26753914e-01 2.37975806e-01 -1.61883652e+00 4.27444428e-01
1.18366623e+00 2.62246132e-01 5.00255339e-02 6.37904763e-01
-2.87262917e-01 1.56507182e+00 -8.42309058e-01 -7.99255133e-01
-4.50450003e-01 -6.23721600e-01 -1.16653657e+00 -7.74393380e-02
7.80498147e-01 -2.02233061e-01 4.27828074e-01 1.10736871e+00
3.66175026e-01 -6.20291755e-02 4.34038341e-01 -1.03387976e+00
-5.39084077e-01 8.04905117e-01 -2.19839707e-01 1.51203737e-01
6.14847302e-01 -5.78222387e-02 1.26820588e+00 -3.11792165e-01
6.07244790e-01 1.15980411e+00 2.96916932e-01 4.45843101e-01
-9.13552999e-01 -7.10603476e-01 4.27230857e-02 5.13477623e-01
-1.95422098e-01 -2.69685715e-01 8.27327669e-01 -7.04862952e-01
6.47231102e-01 4.23626751e-01 8.04951191e-01 7.06890225e-01
2.94528976e-02 8.91585052e-01 1.27563345e+00 -8.66490662e-01
1.24149345e-01 4.37251002e-01 3.05881500e-01 8.22427452e-01
2.18312576e-01 -4.57574934e-01 -6.37898445e-01 5.65469339e-02
2.58596629e-01 -3.85809094e-01 -2.01494619e-01 -7.79200569e-02
-8.51777077e-01 1.05828810e+00 4.37928028e-02 6.06942058e-01
-1.30501792e-01 -7.70044699e-02 4.03221905e-01 4.56692249e-01
3.10798615e-01 8.87736142e-01 -3.63269061e-01 -9.43420649e-01
-8.49975526e-01 4.38401431e-01 1.00034237e+00 3.11928481e-01
2.33600065e-01 -1.91004574e-01 -6.94891289e-02 8.60302269e-01
4.63669479e-01 1.09882183e-01 4.26600993e-01 -1.27974010e+00
3.52549464e-01 1.38156688e+00 1.01018369e-01 -1.07472026e+00
-1.31036798e-02 -4.31393772e-01 6.43556267e-02 4.34091806e-01
6.92814231e-01 7.06193149e-02 -3.19454312e-01 1.48569894e+00
4.06818509e-01 -1.75501645e-01 -1.80784732e-01 8.12988043e-01
9.00749147e-01 1.92026675e-01 1.41344622e-01 8.30942914e-02
1.55462801e+00 -7.26926327e-01 -6.92971528e-01 -1.22412115e-01
1.09461772e+00 -1.26680219e+00 1.24432552e+00 6.02760255e-01
-1.30536115e+00 -1.50931969e-01 -9.06900406e-01 -2.58064687e-01
-7.44033605e-02 2.13969991e-01 3.28953266e-01 8.66458595e-01
-7.11619377e-01 6.34500444e-01 -6.49671614e-01 -2.45893300e-01
4.51041996e-01 1.86623782e-01 -7.67699331e-02 -1.68222710e-01
-8.77202809e-01 1.28686523e+00 -2.78533310e-01 -1.34768367e-01
-4.38489616e-02 -7.16098189e-01 -6.12739205e-01 1.30839631e-01
5.30292094e-01 -4.62934703e-01 1.57017779e+00 -7.97261119e-01
-1.52989078e+00 9.52871442e-01 8.07947367e-02 -9.43987370e-02
5.41202068e-01 -2.39097983e-01 5.03158867e-02 -2.68722400e-02
2.11953595e-01 3.44979733e-01 3.12362105e-01 -4.97417718e-01
-7.99626529e-01 -5.01180768e-01 1.79695249e-01 3.84366274e-01
-6.62852049e-01 4.67261642e-01 3.01134527e-01 -5.17353415e-01
5.21541089e-02 -9.38863277e-01 1.12790272e-01 -1.70644701e-01
3.32607031e-02 -8.96102607e-01 8.71662796e-01 -7.37803578e-01
1.36680138e+00 -1.68629980e+00 -1.36471123e-01 1.09142233e-02
2.63021022e-01 6.83233023e-01 1.44063741e-01 6.17380857e-01
1.08355567e-01 2.52113730e-01 3.25512886e-01 2.86192745e-01
1.96285486e-01 7.74215087e-02 -1.52118862e-01 1.53894544e-01
1.98943555e-01 6.64875209e-01 -1.16394341e+00 -4.91704196e-01
1.68010160e-01 3.08720380e-01 -4.19891119e-01 4.31396455e-01
-1.39274850e-01 -1.59870498e-02 -6.20702982e-01 2.76567906e-01
-3.53483297e-02 -2.72585839e-01 2.28138342e-01 4.67920870e-01
-3.12873572e-01 1.34965134e+00 -1.01549804e+00 8.83654594e-01
-5.26338875e-01 1.05102324e+00 1.70333251e-01 -7.73692608e-01
1.00373280e+00 4.75410968e-01 9.55265686e-02 -5.44554591e-01
1.94193080e-01 3.42221856e-01 6.11319184e-01 -6.50982618e-01
5.63486993e-01 -1.22866206e-01 2.64107764e-01 1.13763154e+00
-3.73280078e-01 -3.04341763e-01 4.61629629e-01 4.51395661e-01
1.16621256e+00 2.38015383e-01 2.09641621e-01 -1.89967260e-01
6.77730680e-01 1.48744673e-01 1.86181307e-01 3.87551755e-01
-2.20798507e-01 2.77832355e-02 6.69443429e-01 -4.51572239e-01
-1.03975105e+00 -2.24557832e-01 4.49713841e-02 1.31354558e+00
-5.93011200e-01 -3.21379989e-01 -8.58884394e-01 -7.44490802e-01
-1.56632498e-01 9.48713660e-01 -4.08309311e-01 7.37933069e-02
-6.06517255e-01 3.15272026e-02 4.63355452e-01 4.55174029e-01
3.34979594e-01 -1.14610672e+00 -1.18259072e+00 5.76129317e-01
-3.05175692e-01 -7.76854336e-01 -1.61100030e-01 5.39773032e-02
-5.59909880e-01 -1.31888092e+00 -1.91986233e-01 -7.33170629e-01
6.04517937e-01 2.61853963e-01 9.75363135e-01 7.29784548e-01
-4.19086628e-02 3.55968982e-01 -4.58466917e-01 -8.96950901e-01
-1.08896065e+00 6.27475977e-02 -5.00745475e-01 -7.18452513e-01
6.53875351e-01 -4.22826141e-01 -2.83224165e-01 2.90517777e-01
-6.28645360e-01 4.59555566e-01 3.22636724e-01 6.61145270e-01
-9.19379145e-02 -6.68973476e-02 7.19714284e-01 -1.23140073e+00
1.08853900e+00 2.01021209e-02 -4.96200055e-01 3.10968190e-01
-9.90442455e-01 1.19333036e-01 3.76167774e-01 -3.68113011e-01
-1.08394635e+00 -5.73813260e-01 -3.40544492e-01 3.53601933e-01
-1.61912575e-01 4.63609636e-01 3.62925641e-02 -2.68465519e-01
6.72699153e-01 -2.50614583e-01 1.38659477e-01 -3.21459174e-01
7.10582808e-02 7.10888803e-01 2.41551414e-01 -9.07895505e-01
5.34871995e-01 -2.21578732e-01 3.82223614e-02 -6.89312816e-01
-1.18314564e+00 -5.49843669e-01 -2.67006457e-01 -5.61076701e-01
3.67560118e-01 -5.52731693e-01 -1.10291886e+00 -8.59825462e-02
-9.65935230e-01 -2.58165956e-01 -1.12717889e-01 3.51520121e-01
-2.57889867e-01 2.79174924e-01 -4.75330651e-01 -7.76100338e-01
-2.23255098e-01 -1.16639078e+00 4.70594913e-01 5.76523721e-01
-9.84589159e-01 -1.09110391e+00 5.47866933e-02 1.41319275e+00
2.94077933e-01 -9.06229764e-02 1.20124388e+00 -1.02368963e+00
-2.24411517e-01 -2.10353047e-01 -8.78740624e-02 4.89800960e-01
-2.51410037e-01 2.95385450e-01 -8.41104507e-01 2.51466244e-01
3.09713297e-02 -4.55842108e-01 2.22152948e-01 1.29476823e-02
7.57905424e-01 -7.06096172e-01 1.69209331e-01 -2.74119973e-01
8.98711562e-01 1.37645034e-02 3.43576998e-01 6.47363424e-01
3.99787158e-01 1.01888454e+00 6.51628077e-01 1.27886562e-02
6.98922813e-01 5.56210458e-01 2.46288359e-01 1.32357806e-01
-2.71741718e-01 -2.95961231e-01 3.29591334e-01 7.98282206e-01
-8.08071941e-02 -1.54450461e-01 -1.04084241e+00 7.13194251e-01
-1.86340022e+00 -1.23532355e+00 -7.38907158e-01 1.92932904e+00
1.03518915e+00 2.99274921e-01 9.33000892e-02 5.60438514e-01
3.39576691e-01 -5.09699620e-02 4.98329327e-02 -1.12526691e+00
4.05329704e-01 2.40869850e-01 -4.04200852e-02 6.23736620e-01
-4.55279887e-01 8.64290655e-01 4.83735180e+00 4.03068572e-01
-9.48335290e-01 -2.55570292e-01 5.80570340e-01 2.52263516e-01
-6.41076446e-01 1.60936505e-01 -8.66388500e-01 3.56099367e-01
7.77601480e-01 -3.44416827e-01 1.92781001e-01 7.44072020e-01
4.42684561e-01 -2.67137825e-01 -9.21042562e-01 1.62817329e-01
-4.31567021e-02 -1.12928021e+00 -4.07992572e-01 2.03975528e-01
7.32272685e-01 -3.81930470e-01 -6.67520287e-03 2.10190594e-01
7.15728283e-01 -8.78998399e-01 6.13370895e-01 7.73238689e-02
2.67131650e-03 -6.14810407e-01 8.09889078e-01 6.92038000e-01
-4.36893642e-01 -1.44738797e-03 3.21429819e-02 -8.73006165e-01
-3.46364230e-01 3.60731423e-01 -1.64482629e+00 -7.85646886e-02
2.84157187e-01 2.11038962e-01 -6.50888681e-01 8.99923325e-01
-9.75035667e-01 1.23616445e+00 -1.23952150e-01 -7.40836740e-01
3.50854546e-01 -1.49353772e-01 3.44619483e-01 1.04206729e+00
1.55031115e-01 4.66593415e-01 1.12328060e-01 6.44117951e-01
-1.21880330e-01 4.74815369e-01 -2.46800676e-01 -4.86263275e-01
7.06732929e-01 1.38647211e+00 -7.74299383e-01 -1.46725953e-01
-2.79460430e-01 2.01516315e-01 5.14339507e-01 -9.05695036e-02
-1.39111310e-01 -2.52004534e-01 4.79396403e-01 5.94268918e-01
9.19075757e-02 1.42034844e-01 -6.89854503e-01 -5.44594646e-01
-8.79098326e-02 -1.18286514e+00 2.13702306e-01 -8.09366226e-01
-8.66361082e-01 -1.16768599e-01 -4.53553200e-01 -5.84100306e-01
-5.57720900e-01 -7.36849487e-01 -8.57544720e-01 9.58695292e-01
-1.37532949e+00 -1.03220510e+00 -2.25675866e-01 -1.01297535e-01
4.08124775e-01 1.21438250e-01 8.28564584e-01 -1.10194266e-01
-2.25654870e-01 4.26268697e-01 -3.56446326e-01 1.55791223e-01
8.09549570e-01 -1.44001842e+00 1.90214310e-02 7.85301149e-01
4.76904213e-01 6.94232583e-01 1.19458640e+00 -4.62122917e-01
-1.09772742e+00 -5.13026237e-01 1.73382723e+00 -6.37839019e-01
9.22128141e-01 3.36764991e-01 -1.16528738e+00 3.70056301e-01
4.51109827e-01 -7.91896105e-01 8.88391376e-01 3.45470339e-01
-3.10082138e-01 1.90887541e-01 -9.56039190e-01 4.25540954e-01
4.18978065e-01 -4.41135436e-01 -1.03873062e+00 1.92999020e-01
1.17264584e-01 -4.87712592e-01 -9.11777198e-01 -1.75901912e-02
4.93858039e-01 -1.04049540e+00 3.54374737e-01 -5.15918672e-01
9.63413894e-01 -1.19180851e-01 5.92066348e-01 -1.11643314e+00
-1.21910155e-01 -4.15005624e-01 2.27413848e-01 1.39342630e+00
3.72913390e-01 -3.05701524e-01 9.83610153e-01 8.35667849e-01
-3.38532031e-01 -1.01829910e+00 -4.80394989e-01 -1.36596739e-01
1.42364427e-01 -5.57032824e-01 5.11691570e-01 1.01147842e+00
5.60001671e-01 6.13940299e-01 2.76855320e-01 -2.27432325e-01
2.88375467e-01 2.81278223e-01 9.28236783e-01 -1.61498463e+00
-5.79900742e-02 -9.45170224e-01 -2.16420844e-01 -5.76487124e-01
1.16673194e-01 -7.51369417e-01 -7.88919702e-02 -1.89390922e+00
1.77254274e-01 -3.06631893e-01 1.63543969e-01 7.27651596e-01
-3.59746605e-01 9.89110768e-02 1.35189295e-01 -1.78236187e-01
-5.54684341e-01 -1.06603548e-01 1.43920004e+00 1.99609175e-01
-2.28236374e-02 1.51897132e-01 -1.26561356e+00 9.52601731e-01
6.50943398e-01 -5.58839679e-01 -5.09665906e-01 -3.12787861e-01
8.09036791e-01 -5.42734601e-02 2.62371480e-01 -6.66311860e-01
4.15566027e-01 -5.43539345e-01 7.93440454e-03 -1.29807249e-01
-2.44420558e-01 -5.51001430e-01 -2.80212939e-01 1.79297239e-01
-7.93358564e-01 1.35856226e-01 -4.60523702e-02 -1.83873072e-01
-2.83759117e-01 -5.96263230e-01 5.57533681e-01 -1.77102480e-02
-5.70019037e-02 -3.91281962e-01 -3.73303801e-01 2.61340380e-01
8.49314213e-01 -2.37994418e-01 -5.97684383e-01 -5.31893432e-01
-1.80455565e-01 2.31289819e-01 6.41512156e-01 3.70022237e-01
3.71897161e-01 -8.15725327e-01 -8.25651526e-01 -7.35095143e-02
-1.26509950e-01 -2.48943985e-01 -1.45699680e-01 7.06049502e-01
-4.73538786e-01 6.74294591e-01 -1.54311052e-02 -2.33062655e-01
-1.95003104e+00 -8.64330754e-02 -4.34130803e-03 -4.99028027e-01
-4.80102688e-01 9.79444385e-01 -4.76382643e-01 -2.24554539e-01
2.13094249e-01 -1.74799636e-01 -5.47706962e-01 1.81583807e-01
6.38095915e-01 6.10178649e-01 2.73760170e-01 -4.72657740e-01
8.13032612e-02 2.39603035e-02 -2.66676039e-01 -6.93488419e-02
1.62807071e+00 2.16982409e-01 -1.59997657e-01 3.68365824e-01
5.23129523e-01 4.83780652e-01 -7.43099928e-01 5.55953644e-02
4.03961271e-01 -5.81372917e-01 1.86738834e-01 -1.43045449e+00
-3.85618180e-01 1.06116569e+00 -2.19856761e-02 5.23827553e-01
6.80473328e-01 -3.50605953e-03 5.62849700e-01 2.12464288e-01
-6.77101240e-02 -1.23421180e+00 2.28206620e-01 3.61153036e-01
5.85855663e-01 -1.17391121e+00 2.97771096e-01 -2.79509455e-01
-5.97583652e-01 1.24234891e+00 6.61319017e-01 2.82425404e-01
-8.12559426e-02 1.50339842e-01 2.76183426e-01 -3.97148788e-01
-1.10614204e+00 1.21706598e-01 5.20700932e-01 2.43819788e-01
1.34327257e+00 2.31425583e-01 -9.03561234e-01 3.56218845e-01
-6.55525148e-01 -9.18912888e-02 9.97222126e-01 7.51165390e-01
-9.36798453e-01 -1.55504096e+00 -4.25618559e-01 5.58235884e-01
-7.56986439e-01 -4.08787914e-02 -9.35976148e-01 6.71617746e-01
9.24966633e-02 1.13969183e+00 -4.32565778e-01 4.46918942e-02
3.17079782e-01 4.00244772e-01 5.88519096e-01 -9.93377209e-01
-1.18985653e+00 -2.25887060e-01 5.37547827e-01 -7.01340362e-02
-3.88795763e-01 -8.30054462e-01 -1.21897113e+00 -5.25387406e-01
-4.70873594e-01 4.97189611e-01 8.92394423e-01 1.26553750e+00
1.05832830e-01 4.89320457e-01 5.49874939e-02 1.55341074e-01
-9.19999123e-01 -9.89035606e-01 3.36892009e-02 2.78261542e-01
-8.26236382e-02 -4.12258953e-01 -9.59243402e-02 -1.30629539e-01] | [11.265480041503906, 9.141261100769043] |
0d3a40b0-4b2b-4bc7-b191-5edc907ac8cf | composite-triggered-intermittent-control-for | 2305.19644 | null | https://arxiv.org/abs/2305.19644v1 | https://arxiv.org/pdf/2305.19644v1.pdf | Composite Triggered Intermittent Control for Constrained Spacecraft Attitude Tracking | This paper focuses on the spacecraft attitude control problem with intermittent actuator activation, taking into account the attitude rotation rate limitation and input saturation issue simultaneously. To address this problem, we first propose a composite event-trigger mechanism, which composed of two state-dependent trigger that governing the activation and deactivation of actuators. Subsequently, by introducing the cascaded decomposition of Backstepping control philosophy, the designed trigger mechanism is then applied to the decomposed dynamical subsystem, providing a layered intermittent stabilization strategy. Further, the basic intermittent attitude controller is extended to a "constrained version" by introducing a strictly bounded virtual control law and an input saturation compensation auxiliary system. By analyzing the local boundedness of the system on each inter-event time interval, a uniformly, strictly decreasing upper boundary of the lumped system is further characterized, thereby completing the proof of the system's uniformly ultimately boundedness (UUB). Finally, numerical simulation results are illustrated to demonstrate the effectiveness of the proposed scheme. | ['Shujian Sun', 'Weijia Wang', 'Kun Wang', 'Tao Meng', 'Jiakun Lei'] | 2023-05-31 | null | null | null | null | ['philosophy'] | ['miscellaneous'] | [-6.05137609e-02 4.07578528e-01 -2.82435268e-01 4.16463107e-01
1.41072765e-01 -8.53898168e-01 3.81248057e-01 1.79905239e-02
-6.32588491e-02 1.17515528e+00 -1.20641671e-01 -4.09099221e-01
-3.73271435e-01 -4.69388932e-01 -6.37121558e-01 -1.16653299e+00
2.69257784e-01 -5.55855874e-03 -1.22846533e-02 -4.97416049e-01
-2.29614720e-01 2.17257231e-01 -1.08113647e+00 -8.32052946e-01
9.89228845e-01 9.56280172e-01 -8.91543031e-02 3.31267893e-01
6.22174919e-01 4.68637943e-01 -3.70701760e-01 6.45553946e-01
7.75178149e-02 -5.11328340e-01 -3.41506600e-01 8.44044745e-01
-3.63299072e-01 -4.45156813e-01 4.09642346e-02 1.15762973e+00
3.00290078e-01 5.77685297e-01 3.89221668e-01 -1.14621985e+00
-9.68489572e-02 1.35420278e-01 -3.68862033e-01 -4.56282832e-02
2.10100487e-01 1.80616647e-01 4.70097959e-01 -8.17711592e-01
3.27907711e-01 9.86186564e-01 3.03421170e-01 3.86171877e-01
-1.12989855e+00 -3.29111248e-01 5.43442965e-01 -4.44693416e-01
-1.14868867e+00 -2.21106321e-01 6.50455415e-01 -5.24845779e-01
4.07767534e-01 4.73071575e-01 9.45886135e-01 6.05033636e-01
3.78003061e-01 3.35886717e-01 7.93401778e-01 -1.69956669e-01
3.26842606e-01 1.08456291e-01 2.73562670e-01 6.66692019e-01
7.34945953e-01 1.83081299e-01 3.70947987e-01 -3.20818335e-01
1.06237900e+00 -9.68083665e-02 -5.69559932e-01 -4.41086918e-01
-1.05804527e+00 4.44780499e-01 2.87490964e-01 4.70088273e-02
-7.20377266e-01 -2.22102240e-01 3.97589147e-01 5.63524544e-01
4.02561724e-01 1.40974447e-01 -1.76332042e-01 4.92224097e-01
-1.88103557e-01 2.61532962e-01 7.78621733e-01 1.17657554e+00
1.75876617e-01 8.55432630e-01 7.96883330e-02 4.19564359e-02
3.91747147e-01 5.45046926e-01 -2.83223167e-02 -6.55482829e-01
1.05723506e-02 7.96680510e-01 1.04565835e+00 -5.61936259e-01
-3.97907495e-01 -7.49724269e-01 -9.86037195e-01 1.76630318e-01
1.80057451e-01 -8.06788564e-01 -4.10574228e-01 1.94312346e+00
8.99959624e-01 -1.23563476e-01 3.28757226e-01 1.19542956e+00
-1.31707817e-01 1.08621681e+00 -2.44032755e-01 -1.32596290e+00
1.28016472e+00 -3.81611168e-01 -1.32801449e+00 9.16255936e-02
1.50413886e-01 -1.13770835e-01 7.99051225e-01 1.33255690e-01
-1.21805429e+00 -2.71815598e-01 -1.40422022e+00 7.02539027e-01
1.26978993e-01 4.23898160e-01 -5.54852374e-02 -1.73980758e-01
-7.09436357e-01 -5.37708141e-02 -8.28611255e-01 -3.92162800e-01
-7.78339148e-01 4.25597847e-01 1.68327466e-02 8.15863311e-01
-1.38633049e+00 7.07628846e-01 7.23245442e-01 4.58677948e-01
-6.81910276e-01 -5.89326024e-01 -6.05362892e-01 -2.34809324e-01
7.76706755e-01 -1.06492472e+00 1.12733138e+00 -1.05829740e+00
-1.90452170e+00 2.31447995e-01 -1.86907783e-01 -3.13892007e-01
5.71334302e-01 -1.44070819e-01 -2.55759835e-01 9.67524871e-02
-1.26846164e-01 -2.00306445e-01 9.75004673e-01 -1.26963770e+00
-5.70010245e-01 -2.32959092e-01 8.92517716e-02 6.04465187e-01
-5.38405299e-01 -2.32791275e-01 -5.32833859e-02 -3.91814828e-01
1.04225457e-01 -8.88140082e-01 -3.21541041e-01 -2.10708752e-01
-2.67021835e-01 -1.89426482e-01 9.30941045e-01 -3.64172280e-01
1.61082113e+00 -2.32211781e+00 9.26986098e-01 1.60597280e-01
-3.00704658e-01 5.63532561e-02 5.58875799e-01 6.85115278e-01
-9.10658166e-02 -3.94791752e-01 -3.29241693e-01 -8.84904526e-03
-3.05350184e-01 8.30180496e-02 -7.84294307e-01 7.97843933e-01
3.70341331e-01 8.93152356e-02 -8.44252110e-01 5.26752323e-02
3.19097519e-01 2.03431502e-01 -2.75511265e-01 1.21822551e-01
-3.09435964e-01 8.35563123e-01 -9.03644323e-01 5.97972155e-01
4.30629373e-01 -4.24204841e-02 4.93842542e-01 -1.38055906e-01
-9.30773854e-01 -4.69490677e-01 -1.40603638e+00 7.70007849e-01
-1.60577893e-01 -6.99834973e-02 1.24931705e+00 -9.74251270e-01
9.66771185e-01 6.84425831e-01 4.54041541e-01 -1.22812279e-01
8.62958014e-01 3.24226707e-01 -2.47756302e-01 -1.11147724e-01
3.27692658e-01 -3.74247372e-01 -1.14387073e-01 -1.80774108e-01
-2.52921641e-01 -1.24397278e-01 -8.86458606e-02 -1.14718610e-02
5.07469475e-01 -1.77443735e-02 5.83217025e-01 -9.10743415e-01
1.35371995e+00 3.28520030e-01 8.25676560e-01 -1.71927094e-01
-3.17880601e-01 -4.37704235e-01 3.95006835e-01 -6.71881959e-02
-9.82813299e-01 -5.10911465e-01 -1.68937892e-01 5.86106658e-01
7.42098987e-01 2.21406862e-01 -4.89132673e-01 7.48889223e-02
2.57432729e-01 6.00733817e-01 -5.46435237e-01 -5.03333151e-01
-6.22995794e-01 -8.15716326e-01 -1.93803936e-01 3.19202662e-01
5.24903953e-01 -1.49661452e-01 -7.11979151e-01 4.46820676e-01
3.51989991e-03 -7.37227917e-01 -3.16973776e-01 1.05956960e-02
-9.42463219e-01 -9.87332225e-01 -4.28118646e-01 -8.60507905e-01
1.16433346e+00 1.26070142e-01 1.53942332e-01 -3.29276919e-02
3.18333715e-01 6.48814976e-01 1.65728748e-01 -3.29670191e-01
-1.96521774e-01 -3.84848773e-01 9.16948438e-01 3.82672101e-01
-6.21956706e-01 -1.51080176e-01 -4.56467777e-01 5.72141171e-01
-8.50127220e-01 -3.39625739e-02 2.30136275e-01 9.98373628e-01
6.28828883e-01 1.48166105e-01 8.92432570e-01 2.26225313e-02
8.05220485e-01 -4.79151279e-01 -1.38400054e+00 -1.11542121e-01
-2.72932291e-01 -3.05418611e-01 1.00121367e+00 -6.94158435e-01
-1.28796983e+00 2.17326418e-01 4.00567621e-01 -5.42566240e-01
4.12690759e-01 5.28845966e-01 -3.30647707e-01 2.17197444e-02
5.66403866e-02 5.33944190e-01 8.51854742e-01 -1.61379114e-01
2.48186260e-01 5.55198610e-01 7.63910234e-01 -5.51866770e-01
8.95552635e-01 6.91504896e-01 3.40921551e-01 -8.69082212e-01
-3.60811740e-01 -2.63729900e-01 -3.12931418e-01 -6.04301512e-01
6.01297796e-01 -1.08483911e+00 -1.15006256e+00 4.54716891e-01
-8.58173370e-01 -5.10822348e-02 -2.60593623e-01 6.48074389e-01
-6.36254191e-01 1.62229970e-01 -7.39427090e-01 -1.48127425e+00
-3.58140409e-01 -7.64884412e-01 6.24274254e-01 3.52675080e-01
-6.25130013e-02 -1.09269142e+00 1.62994668e-01 -2.80153006e-01
2.30095968e-01 6.48834229e-01 5.89643240e-01 -1.86085105e-01
-4.13456768e-01 -5.68171203e-01 3.25151116e-01 1.83348015e-01
1.22792110e-01 -1.78775966e-01 -1.12707332e-01 -1.14114749e+00
7.45082140e-01 -1.43533945e-01 -7.44228112e-03 4.46505338e-01
-1.04329586e-01 -6.16143107e-01 -5.51678896e-01 2.43555143e-01
1.72694921e+00 4.60812151e-01 1.21543109e-01 5.46632230e-01
2.38826811e-01 6.61175311e-01 1.07532382e+00 6.64732933e-01
1.37456596e-01 2.92253643e-01 8.87985528e-01 -1.13257810e-01
7.15627372e-01 2.19572023e-01 6.56457305e-01 9.42808807e-01
-1.68676808e-01 -1.63309023e-01 -3.95073324e-01 7.24550068e-01
-1.90917754e+00 -4.59049881e-01 -3.64497513e-01 2.48069119e+00
6.77979767e-01 -1.77146345e-02 1.11578904e-01 1.74912080e-01
1.16148520e+00 -2.22622126e-01 -6.78045154e-01 -1.97771594e-01
-2.28841379e-01 -8.17447484e-01 5.77937484e-01 7.74511814e-01
-9.89012599e-01 3.11306447e-01 5.79540825e+00 2.64220476e-01
-1.37219000e+00 -3.54070902e-01 6.86334819e-02 1.27867103e-01
5.50199449e-02 2.81986147e-01 -8.82868528e-01 5.58880627e-01
8.68287742e-01 -7.69831121e-01 9.03846323e-02 7.49477446e-01
8.99450183e-01 -1.22033022e-01 -4.03065979e-01 2.33534053e-01
-2.58516729e-01 -6.92295611e-01 -6.36036992e-02 -5.18082194e-02
7.12316990e-01 -8.80508900e-01 -8.25667009e-03 7.06083551e-02
-2.98894942e-02 1.37828752e-01 8.85373116e-01 5.11107683e-01
5.66329181e-01 -7.28231490e-01 4.83682036e-01 6.95181370e-01
-1.38815832e+00 -7.12703943e-01 -3.08186617e-02 -5.29911757e-01
6.82817817e-01 2.71236718e-01 -1.80920288e-01 1.13599992e+00
-1.50283679e-01 5.45325816e-01 1.82224035e-01 6.66460931e-01
-1.32607505e-01 4.21134055e-01 -5.44025481e-01 -2.81984955e-01
2.48677209e-01 -7.52089739e-01 1.35424507e+00 3.66950244e-01
2.10357338e-01 7.77186036e-01 3.43489319e-01 7.65870571e-01
5.70075572e-01 -1.70349374e-01 -5.87282658e-01 3.49457711e-01
3.64766359e-01 1.14428806e+00 -3.91448826e-01 -5.63645482e-01
-1.54220849e-01 6.54933393e-01 -3.15340966e-01 4.67184782e-01
-1.29406011e+00 -5.25049925e-01 6.52233183e-01 -4.69930470e-02
-2.46348158e-02 -5.45073271e-01 -6.83019534e-02 -1.05337369e+00
1.91692755e-01 -3.64199489e-01 1.61350206e-01 -5.09549260e-01
-7.32377946e-01 3.25801373e-01 1.68227941e-01 -2.04960227e+00
-3.28043520e-01 -2.37611309e-01 -8.72243762e-01 9.34598505e-01
-9.94343638e-01 -8.22326481e-01 8.18847120e-03 5.16729295e-01
3.81731331e-01 2.14005485e-01 3.42994064e-01 2.81773746e-01
-1.09440303e+00 -2.50417471e-01 5.13259292e-01 -4.57633257e-01
4.77662176e-01 -8.66346240e-01 -6.86281502e-01 1.21852195e+00
-1.69527698e+00 8.26459587e-01 1.18708634e+00 -8.20921302e-01
-2.07253337e+00 -1.39614749e+00 3.75068814e-01 4.05067392e-02
1.04811752e+00 -5.10247366e-04 -9.77671981e-01 8.99938703e-01
4.14289445e-01 -1.87300742e-01 -3.47848505e-01 -8.40429485e-01
7.40673602e-01 -6.20572902e-02 -1.03847301e+00 6.98308229e-01
4.66542065e-01 3.41679417e-02 -6.80029094e-01 1.19793087e-01
9.97064412e-01 -6.16029322e-01 -9.47228193e-01 7.61652291e-01
5.77657036e-02 1.17977276e-01 7.09233105e-01 -6.33887172e-01
-2.02091336e-01 -1.03760612e+00 1.64122775e-01 -1.27342474e+00
-4.19302851e-01 -1.08120155e+00 -3.96037340e-01 1.20992589e+00
2.50777379e-02 -1.00532246e+00 1.46869853e-01 2.06509799e-01
-4.61428225e-01 -6.60385847e-01 -9.54872310e-01 -1.00423145e+00
-2.17710868e-01 5.89002907e-01 -5.29134413e-03 1.05797923e+00
5.84185123e-01 5.49479365e-01 -4.88612980e-01 1.02847457e+00
6.14793837e-01 6.88061267e-02 5.80383718e-01 -8.26179028e-01
1.94207598e-02 -7.21021742e-02 2.71200925e-01 -8.58939111e-01
-1.40495613e-01 -1.25099748e-01 2.04847097e-01 -1.32077801e+00
-4.15276051e-01 1.64877370e-01 -1.43423036e-01 -4.45600413e-03
-3.02011520e-01 -4.29186076e-01 -3.48020285e-01 3.85231555e-01
-2.39136711e-01 1.02362061e+00 1.24439406e+00 -6.18789978e-02
-7.38359392e-01 1.97317302e-01 -2.13973336e-02 5.68039954e-01
5.26670814e-01 3.54970127e-01 -9.13913786e-01 -1.69467688e-01
-1.39442846e-01 8.19067955e-01 3.37457895e-01 -1.02282417e+00
2.64514297e-01 -4.35008585e-01 -2.82680899e-01 -6.95210040e-01
2.10682064e-01 -9.58623767e-01 5.35545826e-01 1.11341870e+00
1.59149826e-01 9.96129140e-02 3.86239499e-01 9.64390516e-01
-4.03285593e-01 3.87539923e-01 1.04066682e+00 4.12320584e-01
-3.65030915e-01 9.31089371e-03 -9.38118994e-01 -5.56972325e-01
1.46997738e+00 -7.78384581e-02 -1.62222430e-01 -1.53259903e-01
-9.43553150e-01 8.88752401e-01 3.35629880e-01 7.73222893e-02
9.38989893e-02 -1.22364855e+00 -2.76281536e-01 1.87876999e-01
-3.60019147e-01 -3.58057804e-02 4.48692352e-01 1.37650943e+00
-3.71890217e-02 6.11226916e-01 -2.57556021e-01 -5.30519187e-01
-1.06572521e+00 6.45693243e-01 5.85771680e-01 3.64664555e-01
-4.37840700e-01 3.24405879e-01 1.37098342e-01 7.73842586e-03
-1.72761120e-02 -5.70646465e-01 -3.06266636e-01 1.54790714e-01
1.51993960e-01 2.78737992e-01 -2.10122168e-01 -3.53550494e-01
-2.58072942e-01 4.75174725e-01 5.12953401e-01 -2.67944813e-01
9.68705773e-01 -7.47467756e-01 -3.33479971e-01 5.92101157e-01
3.93840998e-01 -2.04486340e-01 -1.41333377e+00 9.90069136e-02
-4.25464988e-01 2.60953426e-01 -2.06766129e-01 -1.66379929e-01
-4.87477541e-01 8.66066292e-02 2.44108722e-01 6.11353517e-01
1.27632439e+00 -6.67068303e-01 3.28447253e-01 1.68387607e-01
4.15979117e-01 -1.01585519e+00 -2.21446529e-01 8.09909761e-01
1.02622390e+00 -5.53551733e-01 1.33824712e-02 -6.81744039e-01
-3.90024185e-01 9.87750590e-01 1.04553938e+00 -6.04699612e-01
3.95726532e-01 3.15456390e-01 -4.29850608e-01 2.26085961e-01
-1.04088795e+00 1.17045455e-01 -9.06136930e-02 -2.17227101e-01
1.83156073e-01 -1.03857666e-01 -1.26992393e+00 8.10209394e-01
7.64446437e-01 2.89305188e-02 7.93744326e-01 1.22656131e+00
-7.82391131e-01 -4.71714526e-01 -8.68749797e-01 -3.55584294e-01
-2.73120571e-02 5.54311633e-01 -1.80711478e-01 1.18567002e+00
-3.11683804e-01 4.88687009e-01 2.57959902e-01 -4.51878421e-02
8.14168096e-01 1.09910518e-01 -6.62807822e-02 -3.10954005e-01
-2.46768132e-01 4.83642101e-01 2.84230728e-02 -1.76724106e-01
-1.23855166e-01 -3.86174411e-01 -1.40262079e+00 -7.59955915e-03
-7.93354869e-01 5.73349714e-01 2.54583389e-01 6.16728187e-01
3.29478055e-01 8.07240009e-01 8.98543537e-01 -6.81686997e-01
-1.12645376e+00 -1.04224455e+00 -4.31791276e-01 -1.72138572e-01
7.62033105e-01 -9.49740648e-01 -9.36576009e-01 5.79902120e-02] | [5.244969367980957, 2.4337925910949707] |
449353ba-125c-4bfd-b121-82eaaf86eef9 | discovery-radiomics-with-clear-dr | 1710.10675 | null | http://arxiv.org/abs/1710.10675v1 | http://arxiv.org/pdf/1710.10675v1.pdf | Discovery Radiomics with CLEAR-DR: Interpretable Computer Aided Diagnosis of Diabetic Retinopathy | Objective: Radiomics-driven Computer Aided Diagnosis (CAD) has shown
considerable promise in recent years as a potential tool for improving clinical
decision support in medical oncology, particularly those based around the
concept of Discovery Radiomics, where radiomic sequencers are discovered
through the analysis of medical imaging data. One of the main limitations with
current CAD approaches is that it is very difficult to gain insight or
rationale as to how decisions are made, thus limiting their utility to
clinicians. Methods: In this study, we propose CLEAR-DR, a novel interpretable
CAD system based on the notion of CLass-Enhanced Attentive Response Discovery
Radiomics for the purpose of clinical decision support for diabetic
retinopathy. Results: In addition to disease grading via the discovered deep
radiomic sequencer, the CLEAR-DR system also produces a visual interpretation
of the decision-making process to provide better insight and understanding into
the decision-making process of the system. Conclusion: We demonstrate the
effectiveness and utility of the proposed CLEAR-DR system of enhancing the
interpretability of diagnostic grading results for the application of diabetic
retinopathy grading. Significance: CLEAR-DR can act as a potential powerful
tool to address the uninterpretability issue of current CAD systems, thus
improving their utility to clinicians. | ['Alexander Wong', 'Graham W. Taylor', 'Devinder Kumar'] | 2017-10-29 | null | null | null | null | ['diabetic-retinopathy-grading'] | ['medical'] | [ 6.51196539e-01 6.68297112e-02 -4.07842815e-01 -4.45384383e-01
-5.16033828e-01 -4.93682146e-01 2.88305998e-01 4.08736497e-01
-1.04820348e-01 5.82105935e-01 4.66444075e-01 -8.22316349e-01
-4.85107541e-01 -6.20054483e-01 -2.44315655e-04 -9.91927505e-01
3.09962630e-01 6.94212079e-01 -7.62673765e-02 -2.33151302e-01
4.74134356e-01 8.54611814e-01 -1.47637069e+00 7.03284383e-01
1.04293823e+00 8.72295201e-01 3.02163094e-01 8.35268080e-01
-4.37108427e-02 1.47951353e+00 -4.65132892e-01 3.99132445e-03
1.39300840e-03 -7.89924622e-01 -6.17893636e-01 -1.15512572e-01
-2.70976257e-02 -5.71911693e-01 -5.10802418e-02 8.10326338e-01
8.47229481e-01 -3.47167164e-01 7.87025094e-01 -5.98418236e-01
-5.30334651e-01 3.54515672e-01 -4.80688334e-01 4.08680618e-01
2.75165021e-01 3.16486418e-01 5.83943069e-01 -5.46680689e-01
4.14907902e-01 9.38643813e-01 3.36596549e-01 6.97771549e-01
-9.24599767e-01 -3.37999433e-01 -1.97772399e-01 4.48485672e-01
-1.10297525e+00 -4.64145422e-01 3.30459446e-01 -9.03952479e-01
5.88046670e-01 6.02526605e-01 8.65782082e-01 5.75047195e-01
7.28046358e-01 5.43585300e-01 1.23721969e+00 -6.00717723e-01
7.63529360e-01 -1.29743010e-01 2.01797560e-01 9.06489193e-01
4.67332333e-01 4.68581587e-01 -2.54563838e-01 -3.24277788e-01
7.72767663e-01 3.70171778e-02 -4.76279169e-01 -2.92319655e-01
-7.81324983e-01 9.28830206e-01 5.25232852e-01 2.76459157e-01
-4.99713659e-01 -1.29293604e-02 3.82842690e-01 -1.87293157e-01
4.04309779e-01 5.54228425e-01 3.68985459e-02 8.76042023e-02
-7.19316304e-01 -8.29207376e-02 1.62995070e-01 5.86826622e-01
9.93832946e-02 -8.63153040e-02 -4.52339709e-01 5.15942514e-01
6.61897421e-01 3.67427886e-01 8.38912964e-01 -6.41024172e-01
-1.45999148e-01 1.17240465e+00 2.69491561e-02 -5.10924578e-01
-8.51245344e-01 -7.39695370e-01 -9.19748068e-01 6.55752659e-01
1.90609664e-01 6.36146963e-02 -1.31034219e+00 1.05429828e+00
3.27892229e-02 -2.55486578e-01 3.36818993e-01 1.27397740e+00
9.36637402e-01 -8.98376331e-02 1.76143542e-01 -1.16418503e-01
1.68610561e+00 -5.06525576e-01 -5.95104516e-01 -6.47277385e-03
1.23848271e+00 -5.96441686e-01 7.14658022e-01 4.10449386e-01
-7.34767497e-01 -7.69434199e-02 -1.07927859e+00 -4.45511825e-02
-3.53045762e-02 3.70995075e-01 9.11171079e-01 7.28876889e-01
-9.79603231e-01 -9.56144836e-03 -8.98854017e-01 -5.36597133e-01
8.87089372e-01 3.12479675e-01 -1.06834419e-01 -4.68419462e-01
-7.80581415e-01 1.31107831e+00 2.01732423e-02 3.48617017e-01
-6.72149122e-01 -8.24253201e-01 -5.20672262e-01 -1.97187021e-01
8.89666602e-02 -1.20376015e+00 1.37868583e+00 -6.50051236e-01
-1.53256834e+00 1.00433016e+00 -1.66836813e-01 -6.08454585e-01
6.80518150e-01 7.56398290e-02 -4.10746306e-01 3.82658660e-01
4.42952365e-02 4.11449432e-01 2.26166472e-01 -7.57192135e-01
-7.99964786e-01 -7.43317842e-01 -1.21869370e-01 3.41840059e-01
2.66162366e-01 -9.51840132e-02 -2.20920369e-02 -4.18311417e-01
1.09067947e-01 -8.17518115e-01 -6.62698388e-01 4.43485945e-01
-2.22151190e-01 -4.36937883e-02 4.56502378e-01 -2.46484265e-01
1.22913814e+00 -2.01783419e+00 -2.12622091e-01 6.29945621e-02
7.70093083e-01 5.52745819e-01 3.26213092e-01 3.33192013e-02
-2.37292930e-01 7.83277377e-02 -2.73062289e-01 5.16256154e-01
-6.44034684e-01 1.61367673e-02 1.28465071e-01 5.96889973e-01
4.79908139e-02 1.05344033e+00 -9.77476478e-01 -1.69950739e-01
7.13513076e-01 3.42954040e-01 -2.44644672e-01 1.17900498e-01
-1.05465993e-01 6.75531924e-01 -7.61370242e-01 6.56165063e-01
4.83981848e-01 -5.68269610e-01 2.50694245e-01 -3.63850623e-01
-2.54018307e-01 1.14732534e-01 -4.91820872e-01 1.37641394e+00
-2.49051347e-01 7.69781232e-01 -4.76222634e-01 -6.91625118e-01
8.69623601e-01 3.77380073e-01 4.46098685e-01 -9.31338310e-01
2.71943599e-01 1.05170429e-01 4.48005736e-01 -6.99890494e-01
2.15593409e-02 -7.53542662e-01 4.27504063e-01 2.78700262e-01
-5.24497628e-01 -1.70896738e-03 -1.34474948e-01 1.72613502e-01
1.31254590e+00 -1.01114981e-01 8.39893579e-01 -2.06135258e-01
5.30124724e-01 5.41588604e-01 2.99167931e-01 6.68688059e-01
-1.74439773e-01 7.79811919e-01 4.13283110e-01 -5.59015691e-01
-8.56636882e-01 -7.74725795e-01 -5.10174394e-01 4.21379417e-01
-4.40012105e-02 1.31149232e-01 -3.66331011e-01 -4.48856145e-01
-1.81625634e-01 9.01539385e-01 -8.87715936e-01 -3.36511791e-01
-2.13985029e-03 -1.21709311e+00 5.34897327e-01 5.95366061e-01
2.46953085e-01 -5.19388616e-01 -9.81296241e-01 3.90480042e-01
3.02071613e-03 -7.24263072e-01 8.73545632e-02 1.63008735e-01
-1.07775557e+00 -1.57885003e+00 -1.01739240e+00 -5.43871701e-01
9.35097158e-01 4.75132346e-01 7.98545837e-01 1.28393903e-01
-7.80635178e-01 2.00653434e-01 -4.32925582e-01 -6.78679347e-01
-8.61097276e-01 -5.48765898e-01 -2.58480400e-01 -1.13577932e-01
4.37237084e-01 -9.06635523e-02 -1.15264404e+00 4.67698961e-01
-8.71430933e-01 3.46592546e-01 9.74053741e-01 6.75312638e-01
5.65685391e-01 -2.99717672e-02 5.38087487e-01 -1.23418105e+00
4.47722733e-01 -2.84425259e-01 -6.28574252e-01 2.33458608e-01
-8.97234917e-01 2.07541659e-01 4.49846536e-01 1.97153822e-01
-1.10381138e+00 -5.24429558e-03 -3.73282796e-03 -9.87203345e-02
-1.16853729e-01 5.87057829e-01 5.57771586e-02 -1.41732171e-01
1.17750061e+00 8.35382193e-02 3.98976713e-01 -4.17294532e-01
2.82127619e-01 8.87395918e-01 4.79548544e-01 8.99495557e-03
9.40368846e-02 5.64698040e-01 4.30015564e-01 -5.68252921e-01
-7.98291206e-01 -9.00645554e-01 -2.55580485e-01 -1.93434790e-01
8.48499477e-01 -9.09445047e-01 -7.58714914e-01 2.90650994e-01
-8.35491240e-01 -9.36583355e-02 -3.10887098e-01 6.35497212e-01
-5.18990636e-01 3.06171149e-01 -2.44335294e-01 -5.45940757e-01
-6.19904637e-01 -1.50920832e+00 6.94846630e-01 3.23667794e-01
-4.05885726e-01 -1.03501630e+00 1.02305859e-01 6.45357311e-01
4.61489290e-01 4.02676493e-01 1.49727345e+00 -4.93637264e-01
-4.70649093e-01 -2.21987888e-01 -3.91439945e-01 9.58408341e-02
6.28280342e-01 -8.91676079e-03 -9.51653361e-01 6.51132241e-02
3.27215083e-02 6.63506165e-02 5.84919810e-01 7.30134308e-01
9.07758117e-01 6.93685189e-02 -4.12954181e-01 3.04474771e-01
1.50774872e+00 6.98969245e-01 1.05513477e+00 2.75888413e-01
3.44252050e-01 3.35372686e-01 7.15572000e-01 3.32615197e-01
2.47904360e-01 6.24047756e-01 4.50750977e-01 -6.37535691e-01
-4.14126873e-01 3.26415896e-02 -1.63810000e-01 7.04247784e-03
-2.19088092e-01 -3.35038275e-01 -9.87390697e-01 3.40919852e-01
-1.66740930e+00 -6.18227184e-01 -5.55948794e-01 2.02693105e+00
4.97151941e-01 -1.79317445e-01 -1.71384260e-01 5.14509045e-02
7.22707450e-01 -5.17425060e-01 -7.32274175e-01 -4.69658494e-01
-2.44216546e-02 1.71460703e-01 5.06527364e-01 2.55379915e-01
-4.18947756e-01 4.24489856e-01 6.43081236e+00 3.32153827e-01
-1.42791796e+00 -1.83923036e-01 5.97863317e-01 1.88855633e-01
-1.21442042e-01 -1.39905559e-02 -5.87088346e-01 3.86260092e-01
7.62690902e-01 -1.48062453e-01 -1.15407124e-01 6.10953808e-01
8.87505054e-01 -7.17459440e-01 -1.34861779e+00 7.24336386e-01
-1.89486369e-01 -1.82940936e+00 3.75206292e-01 1.94313839e-01
3.45375359e-01 -1.72527984e-01 1.05091080e-01 -3.88963908e-01
4.57730561e-01 -1.07252145e+00 1.89665005e-01 7.48471916e-01
1.25325572e+00 -4.74046201e-01 1.12039089e+00 -4.77551594e-02
-4.50155109e-01 -4.06214558e-02 -2.06004694e-01 9.79266390e-02
-1.41231120e-01 7.48307586e-01 -1.84090185e+00 5.92592001e-01
4.18245159e-02 5.52325547e-01 -4.44116801e-01 1.33776259e+00
-2.23693028e-01 6.19466543e-01 1.38725504e-01 6.16008900e-02
-7.97663350e-03 1.08779967e-01 3.83856535e-01 1.03603125e+00
1.50035888e-01 6.59267485e-01 -3.76908064e-01 4.65775996e-01
3.15618426e-01 3.83139700e-02 -3.83662969e-01 -8.11464190e-02
1.50715977e-01 9.33887422e-01 -8.82972836e-01 -3.71873647e-01
-3.04866850e-01 6.53747499e-01 -2.54514962e-01 5.87830245e-02
-3.57480705e-01 -5.59239183e-04 4.17315662e-01 4.44999397e-01
2.97111180e-03 3.74617428e-01 -7.30460346e-01 -6.66037977e-01
-5.29358506e-01 -9.07314599e-01 6.08710408e-01 -1.12104559e+00
-9.84133363e-01 4.87502426e-01 -3.88019592e-01 -1.63791955e+00
-2.25493416e-01 -6.70221567e-01 -2.10453942e-01 9.90954459e-01
-1.54827178e+00 -9.94835138e-01 -5.11963308e-01 1.63185164e-01
4.13656980e-01 -3.80523562e-01 1.08382356e+00 -2.85908431e-02
-4.04857218e-01 3.49020451e-01 3.48737746e-01 -2.55358517e-01
6.61670208e-01 -1.11695278e+00 -2.69842565e-01 5.30709624e-01
-5.05075991e-01 3.26573551e-01 6.79805934e-01 -5.42865813e-01
-1.23243034e+00 -1.09799814e+00 3.97116154e-01 -5.45181036e-01
2.84231275e-01 4.11680818e-01 -5.51325798e-01 2.94957459e-01
-4.55278367e-01 1.97758265e-02 1.25382125e+00 -2.38505676e-01
1.07316725e-01 -5.62301911e-02 -1.49557137e+00 5.34284711e-01
4.74698842e-01 -1.59286216e-01 -4.91726458e-01 4.08216923e-01
1.67913407e-01 -3.99208128e-01 -7.57898510e-01 4.02627856e-01
6.55896842e-01 -8.51799488e-01 6.15786493e-01 -6.35764182e-01
6.01112425e-01 -7.03025460e-01 1.14162816e-02 -1.20352530e+00
-5.91446042e-01 -2.33031496e-01 1.43882364e-01 5.64031184e-01
3.74029547e-01 -6.59211993e-01 9.16281998e-01 6.13565743e-01
-2.52342224e-01 -8.23402405e-01 -8.27397227e-01 3.90241183e-02
-3.16450775e-01 -1.64395899e-01 3.28879774e-01 6.86455965e-01
2.62013264e-03 2.57162720e-01 1.29066035e-01 3.59749854e-01
3.45569223e-01 -1.35332122e-01 3.07470858e-01 -1.25971437e+00
-1.63330987e-01 -2.80269772e-01 -9.91132498e-01 -6.73810542e-01
-7.87398696e-01 -9.35794294e-01 -1.20958857e-01 -2.00491118e+00
3.18475634e-01 -6.99754179e-01 -3.95125955e-01 3.49041522e-01
-1.63494661e-01 2.67334461e-01 -2.38729283e-01 3.67710233e-01
-6.36413842e-02 7.98680335e-02 1.47339976e+00 -4.28910181e-02
-3.66910785e-01 2.47581542e-01 -1.44674802e+00 6.33898139e-01
5.02462983e-01 -5.33693016e-01 -6.35312319e-01 -1.36548996e-01
2.57524461e-01 3.24679226e-01 5.45096636e-01 -7.18982041e-01
3.68572026e-01 -1.06912792e-01 4.75723565e-01 -5.64829469e-01
-2.32717797e-01 -7.16069043e-01 2.25103498e-01 9.14021969e-01
-3.13052714e-01 -4.34109628e-01 2.00531319e-01 6.03880465e-01
-1.85982380e-02 -3.92073803e-02 1.06885469e+00 -1.69487998e-01
-8.05029154e-01 -1.57023802e-01 -9.00284290e-01 -2.25749567e-01
1.03913617e+00 -4.93083477e-01 -5.10594726e-01 -5.84453642e-02
-6.61236227e-01 7.39345178e-02 4.41340357e-01 6.52509630e-02
8.17477524e-01 -9.17824149e-01 -8.45462024e-01 -3.29451892e-03
5.39705992e-01 3.80785793e-01 6.15466535e-01 1.05972648e+00
-1.02131283e+00 7.69750476e-01 -2.30902180e-01 -8.78638387e-01
-1.51849771e+00 3.29982549e-01 8.25762570e-01 -1.75241381e-01
-8.27321231e-01 5.38709342e-01 6.47116959e-01 2.79565752e-01
-9.76061597e-02 -4.33610588e-01 -4.83791947e-01 -1.09086167e-02
8.87987554e-01 4.41363454e-01 6.79475129e-01 -2.18153059e-01
-4.76722896e-01 4.44559842e-01 -6.16904199e-01 2.41443992e-01
1.25740814e+00 -2.44704500e-01 -4.35822085e-03 4.23305109e-02
5.71299255e-01 -3.36297393e-01 -8.20419073e-01 -2.77845170e-02
-3.14903647e-01 -3.22054207e-01 6.63408458e-01 -1.62713170e+00
-7.30987728e-01 7.36900568e-01 1.22532988e+00 -3.73480588e-01
1.29996073e+00 -1.46796659e-01 8.80701095e-02 1.04134485e-01
2.50768363e-01 -5.08089602e-01 -3.05630386e-01 -2.07663402e-01
7.57602870e-01 -1.14429665e+00 3.63840818e-01 -5.50826073e-01
-6.91246629e-01 1.37071764e+00 2.52719700e-01 4.76461411e-01
2.06688166e-01 1.49794176e-01 4.39917743e-01 -4.82723415e-01
-5.26297987e-01 -4.72704954e-02 2.64062971e-01 8.00440490e-01
6.07203126e-01 2.60786921e-01 -5.80592930e-01 3.53297830e-01
1.86556801e-01 7.19748497e-01 8.36005092e-01 1.04121614e+00
-5.85585833e-01 -9.61814702e-01 -3.15644860e-01 6.11229360e-01
-2.11261168e-01 -1.64214790e-01 -3.84087503e-01 5.03850102e-01
-7.18563236e-03 9.92878854e-01 -2.39539832e-01 -4.13044244e-02
4.01094675e-01 -2.14618623e-01 4.54133511e-01 -7.64366329e-01
-1.84856132e-01 -9.05989632e-02 2.00452209e-01 -2.88603634e-01
-3.69445324e-01 -1.97530627e-01 -1.39901161e+00 -3.19558941e-02
-3.81614327e-01 9.65484530e-02 7.89218009e-01 1.15483797e+00
5.91761351e-01 9.11738575e-01 7.11278379e-01 4.80213538e-02
-3.36848438e-01 -6.64082229e-01 -4.61044610e-01 -1.95061773e-01
5.28932095e-01 -5.48782170e-01 -4.35933610e-03 7.90229216e-02] | [15.217050552368164, -2.631983757019043] |
9b8597bf-5def-4193-8604-e5b7a296acff | mask-fpan-semi-supervised-face-parsing-in-the | 2212.09098 | null | https://arxiv.org/abs/2212.09098v5 | https://arxiv.org/pdf/2212.09098v5.pdf | Mask-FPAN: Semi-Supervised Face Parsing in the Wild With De-Occlusion and UV GAN | Fine-grained semantic segmentation of a person's face and head, including facial parts and head components, has progressed a great deal in recent years. However, it remains a challenging task, whereby considering ambiguous occlusions and large pose variations are particularly difficult. To overcome these difficulties, we propose a novel framework termed Mask-FPAN. It uses a de-occlusion module that learns to parse occluded faces in a semi-supervised way. In particular, face landmark localization, face occlusionstimations, and detected head poses are taken into account. A 3D morphable face model combined with the UV GAN improves the robustness of 2D face parsing. In addition, we introduce two new datasets named FaceOccMask-HQ and CelebAMaskOcc-HQ for face paring work. The proposed Mask-FPAN framework addresses the face parsing problem in the wild and shows significant performance improvements with MIOU from 0.7353 to 0.9013 compared to the state-of-the-art on challenging face datasets. | ['Xikun Jiang', 'Zhongfeng Kang', 'Tianfang Zhang', 'Lei LI'] | 2022-12-18 | null | null | null | null | ['face-model', 'face-parsing'] | ['computer-vision', 'computer-vision'] | [ 1.91463754e-02 3.60485524e-01 2.18043163e-01 -8.20305943e-01
-9.04689670e-01 -5.86217582e-01 3.65049362e-01 -7.12722898e-01
-1.41092405e-01 4.21805263e-01 -1.09726883e-01 3.20475012e-01
2.15976745e-01 -4.98386443e-01 -7.80900478e-01 -7.10461795e-01
2.26367533e-01 6.98532760e-01 -2.78637987e-02 1.15281701e-01
-1.33512482e-01 6.47767663e-01 -1.59321344e+00 2.02889997e-03
7.04263687e-01 1.04167819e+00 -4.08281796e-02 3.54436696e-01
-7.36354515e-02 6.32705837e-02 -5.25490403e-01 -7.87822187e-01
3.07772011e-01 -2.65075505e-01 -7.22968519e-01 4.97192383e-01
9.38892126e-01 -3.79193038e-01 3.15072760e-02 1.02333152e+00
7.11257815e-01 -5.26595674e-03 3.75231832e-01 -1.18662798e+00
-1.07126683e-01 1.79821357e-01 -1.13825774e+00 -2.66645074e-01
3.38443667e-01 -5.54987080e-02 5.18305421e-01 -9.27119255e-01
5.28933346e-01 1.80554056e+00 6.03706479e-01 9.86691952e-01
-1.03348827e+00 -9.88904655e-01 2.78760701e-01 -1.51143685e-01
-1.36175454e+00 -8.16576600e-01 7.97652543e-01 -2.79151082e-01
5.41049421e-01 1.20510250e-01 2.05292955e-01 8.79686356e-01
-2.90299892e-01 5.79901814e-01 1.13284051e+00 -3.97388041e-01
7.02674687e-02 -2.36186966e-01 -6.28524274e-02 1.17042506e+00
1.41266704e-01 -2.28721514e-01 -5.53542435e-01 7.82744735e-02
5.98732233e-01 -2.58271068e-01 -2.61933170e-02 -1.38831660e-01
-3.95653933e-01 5.84900200e-01 2.67954022e-01 -1.21696949e-01
-8.37330595e-02 -7.74073452e-02 2.24694222e-01 -3.02264988e-01
6.18993998e-01 -1.49887204e-01 -5.12215257e-01 2.61729270e-01
-9.00689721e-01 2.58392721e-01 5.00707746e-01 8.95901263e-01
7.43787825e-01 6.31769300e-02 -9.85659212e-02 1.17590153e+00
5.54251611e-01 6.69603169e-01 8.11782330e-02 -9.71310973e-01
5.70002258e-01 5.52571654e-01 -2.45658651e-01 -7.94940412e-01
-4.72652525e-01 -1.93595961e-01 -4.80670244e-01 6.33033961e-02
4.23013628e-01 -3.45340282e-01 -1.41149819e+00 1.90352595e+00
9.62772071e-01 3.11383992e-01 -1.66301087e-01 7.45472372e-01
1.07962561e+00 4.71891165e-01 2.96020120e-01 -2.32399970e-01
1.66889131e+00 -1.00664687e+00 -7.34633029e-01 -5.92046559e-01
8.27676803e-02 -7.51280010e-01 6.87203586e-01 1.15593903e-01
-1.14554071e+00 -5.44406056e-01 -6.27476513e-01 -2.25945890e-01
-1.11692421e-01 5.08573771e-01 5.21927297e-01 9.68923390e-01
-1.01380432e+00 2.98550069e-01 -7.98047125e-01 -3.63055348e-01
9.68268096e-01 7.60184228e-01 -7.88067222e-01 -2.41220027e-01
-5.52802861e-01 4.36277002e-01 1.13811783e-01 4.76606667e-01
-8.69680226e-01 -3.55368376e-01 -1.14820957e+00 -5.48478998e-02
5.84367812e-01 -4.32469130e-01 1.18362188e+00 -6.55850887e-01
-1.68679619e+00 1.27622998e+00 -4.56464350e-01 1.64651647e-01
5.54709494e-01 -3.90558481e-01 -2.78601259e-01 2.51785189e-01
2.32963160e-01 9.75615025e-01 1.17439115e+00 -1.14913011e+00
-3.76217723e-01 -1.07525194e+00 -1.74814850e-01 1.37326092e-01
1.09233307e-02 4.50980514e-01 -1.20722294e+00 -6.12760127e-01
1.30021885e-01 -9.09343243e-01 4.59503755e-02 4.21019606e-02
-6.58133030e-01 -3.27835023e-01 9.31943476e-01 -1.08149695e+00
6.23972356e-01 -2.09826374e+00 2.32223108e-01 1.67299315e-01
-1.00608736e-01 3.63636225e-01 -3.34826827e-01 -3.34733874e-01
-4.01996374e-02 -4.43162397e-03 -4.97780621e-01 -1.06348813e+00
4.02193554e-02 1.67021170e-01 2.29785562e-01 5.30544698e-01
4.96709883e-01 7.83078849e-01 -2.72374213e-01 -7.11948931e-01
6.64682537e-02 6.90218389e-01 -6.78934515e-01 4.48652208e-01
-2.97837406e-01 7.26452827e-01 -4.56197202e-01 1.12022674e+00
1.31225789e+00 1.44557282e-01 2.54793495e-01 -8.54562446e-02
1.90536320e-01 -2.18896657e-01 -1.19007850e+00 1.81542647e+00
-1.81090876e-01 -1.30820991e-02 6.85098410e-01 -5.42356491e-01
7.77075291e-01 3.01452845e-01 2.51843244e-01 -5.78794599e-01
3.74780327e-01 1.13108475e-02 -2.93239117e-01 -5.38534880e-01
-1.08057931e-01 -1.28800526e-01 1.16200984e-01 1.86142504e-01
3.13123345e-01 -1.22844223e-02 1.97108030e-01 -2.62516528e-01
6.09922767e-01 5.28319657e-01 -7.18641207e-02 -1.96926519e-01
7.87760019e-01 -5.59547126e-01 1.08124602e+00 -8.76891464e-02
-2.79113024e-01 8.89435470e-01 5.45029223e-01 -1.95547625e-01
-4.62018162e-01 -7.93914497e-01 -2.27788404e-01 1.05054271e+00
-2.17513785e-01 -4.03936297e-01 -1.73948967e+00 -1.16216159e+00
-3.82522754e-02 1.70256615e-01 -5.93425691e-01 2.61329889e-01
-8.03757012e-01 -8.85973334e-01 4.57858652e-01 5.13487399e-01
8.49340022e-01 -1.14144933e+00 -4.36768234e-02 -1.61168873e-01
-1.41522497e-01 -1.60842037e+00 -7.00617909e-01 -2.63350248e-01
-5.59491158e-01 -1.26560152e+00 -7.22772896e-01 -9.54195321e-01
9.82060611e-01 -1.37926146e-01 1.00142562e+00 2.55204868e-02
-5.98751187e-01 9.95755568e-02 -8.11684877e-02 -3.53639215e-01
-1.76050644e-02 -1.91727784e-02 4.62775379e-02 3.35688680e-01
3.23398709e-01 -2.51266330e-01 -4.63639408e-01 4.69965309e-01
-5.62501729e-01 -3.32635492e-01 3.88747007e-01 5.90012491e-01
7.19511032e-01 6.92462027e-02 4.95396733e-01 -1.11236942e+00
-1.58728361e-01 -1.28237322e-01 -8.59574378e-01 3.39433670e-01
-9.82937440e-02 -1.42021298e-01 3.38972896e-01 1.18715800e-02
-1.52648687e+00 5.93737364e-01 -6.11541986e-01 -2.46649623e-01
-5.31738281e-01 -3.69132608e-01 -1.20822573e+00 -3.10392529e-01
1.19085632e-01 -2.72331506e-01 -3.52479927e-02 -8.25685143e-01
3.07787597e-01 6.08262181e-01 7.14764595e-01 -7.41628408e-01
9.66957271e-01 5.22201657e-01 1.68979108e-01 -9.19028819e-01
-9.44549859e-01 -7.42940754e-02 -9.43162382e-01 -3.84224243e-02
1.14985037e+00 -9.60586190e-01 -7.12722480e-01 7.93976843e-01
-1.18347263e+00 -1.19026512e-01 1.18649468e-01 1.54866669e-02
-3.47610831e-01 3.85861784e-01 -5.64017475e-01 -8.38833153e-01
-5.24638176e-01 -1.30125654e+00 1.55233681e+00 5.54841280e-01
5.07392921e-02 -5.27091920e-01 -3.31927419e-01 8.82017970e-01
-1.92801550e-01 4.93207663e-01 6.09957695e-01 -2.97120392e-01
-5.50208569e-01 -1.72056928e-02 -2.04990193e-01 3.82361919e-01
1.51770204e-01 -9.72045660e-02 -1.48856163e+00 -3.36926728e-01
-8.03559050e-02 -3.15369934e-01 6.28596246e-01 3.77166748e-01
1.30120778e+00 -1.63825482e-01 -3.49687338e-01 9.01393473e-01
9.42172766e-01 1.55401826e-01 4.99942690e-01 -2.60926634e-01
9.37555134e-01 1.05186534e+00 5.53844571e-01 2.92066962e-01
4.51145291e-01 8.10504675e-01 3.74654263e-01 -4.48132530e-02
-2.98101753e-01 -3.05944681e-01 1.86874062e-01 2.99793065e-01
-2.87742913e-02 -8.38799551e-02 -6.23153448e-01 1.64974704e-01
-1.33655214e+00 -4.13396686e-01 1.27235308e-01 1.98026729e+00
6.42167687e-01 -1.68454513e-01 1.47032693e-01 3.94220538e-02
1.04076409e+00 1.69286549e-01 -5.52708685e-01 -6.46827221e-02
-1.48740131e-02 5.82845867e-01 -1.15969786e-02 5.80486894e-01
-1.25940657e+00 1.31520832e+00 5.24709845e+00 6.21266484e-01
-9.34057653e-01 3.75228375e-01 9.81957972e-01 1.41160354e-01
1.61671385e-01 -3.59109551e-01 -1.24800658e+00 4.99908477e-01
4.56425071e-01 7.71200836e-01 4.38749492e-01 7.48924375e-01
-1.25768930e-01 -3.32404971e-02 -9.94400799e-01 1.28668332e+00
5.06445825e-01 -5.81078827e-01 -2.08291054e-01 2.95839887e-02
5.29839456e-01 -4.83738065e-01 1.29406273e-01 2.11277515e-01
-1.09621525e-01 -1.19384754e+00 6.31224930e-01 1.17032796e-01
1.03458166e+00 -9.17850494e-01 6.03832483e-01 1.18741125e-01
-1.32788467e+00 7.50029907e-02 -1.76802710e-01 4.18289065e-01
1.95399970e-01 4.78786439e-01 -5.59436977e-01 4.49729055e-01
8.54394913e-01 3.03651184e-01 -5.42631149e-01 6.75768971e-01
-6.35164380e-01 4.48566854e-01 -3.90104234e-01 6.78828895e-01
-2.31224462e-01 -3.35490815e-02 2.78409421e-01 9.52424824e-01
2.88534790e-01 2.94653744e-01 1.10610344e-01 6.82113588e-01
-4.81526524e-01 6.71870559e-02 -1.33998811e-01 1.10553391e-01
3.61370265e-01 1.56373560e+00 -7.68640816e-01 1.22851826e-01
-2.99161404e-01 1.22371101e+00 3.24024826e-01 1.85024589e-01
-8.34201217e-01 8.95749778e-03 7.79281199e-01 1.16547443e-01
3.94157231e-01 1.00728706e-01 -1.90240555e-02 -1.02416492e+00
3.44860762e-01 -8.43085647e-01 4.81433153e-01 -4.12474722e-01
-9.39356089e-01 8.07692230e-01 -1.42553389e-01 -4.49155301e-01
-1.56676412e-01 -6.58552647e-01 -5.37610769e-01 7.84819543e-01
-1.57633114e+00 -1.59787059e+00 -4.99953032e-01 4.84661788e-01
7.26621926e-01 -8.05754587e-02 7.95269370e-01 4.04882848e-01
-1.12812054e+00 1.01408243e+00 -4.67073202e-01 3.75457168e-01
6.89486265e-01 -1.00847042e+00 6.96713090e-01 9.08590496e-01
-2.98402756e-02 4.02785718e-01 3.10345381e-01 -7.01316297e-01
-1.31895244e+00 -1.33112419e+00 7.60024071e-01 -3.96737933e-01
-3.25907916e-01 -7.73153782e-01 -6.55813992e-01 7.37485647e-01
-7.12627592e-03 5.24351597e-01 4.13507849e-01 -1.94945540e-02
-4.74677056e-01 -3.83192629e-01 -1.72907460e+00 2.46299714e-01
1.29243255e+00 -3.48219246e-01 -2.58767247e-01 3.48742247e-01
4.05247658e-01 -6.63277209e-01 -5.40403605e-01 5.05656958e-01
3.86488408e-01 -8.28531682e-01 9.31996107e-01 -3.12564462e-01
-8.61218646e-02 -3.02141935e-01 -8.44103992e-02 -9.96437609e-01
4.86016423e-02 -8.55631948e-01 -1.07826680e-01 1.79212916e+00
-1.65387336e-02 -5.31716943e-01 1.08245420e+00 6.34551406e-01
3.61232124e-02 -4.96738464e-01 -9.85876620e-01 -4.43810344e-01
-1.12690609e-02 -6.95676217e-03 8.46392751e-01 5.35064757e-01
-6.34058774e-01 2.97378570e-01 -2.31642082e-01 4.35628623e-01
1.06296468e+00 3.95274013e-02 8.09798956e-01 -1.34828389e+00
-4.36136723e-02 -9.93513968e-03 -3.62384617e-01 -7.78253078e-01
6.55056357e-01 -6.13763452e-01 7.81080127e-02 -1.20564866e+00
1.52485043e-01 -3.05748582e-01 1.97627530e-01 7.52765775e-01
-3.26858670e-01 5.77630222e-01 1.76444501e-01 -3.10677111e-01
-3.20387632e-01 5.08967519e-01 1.12940097e+00 4.36236821e-02
7.65568316e-02 2.26542819e-02 -6.25383615e-01 9.51250732e-01
6.41733170e-01 -4.04670447e-01 -2.38943651e-01 -5.79141378e-01
-5.02797246e-01 -9.12570767e-03 2.50224382e-01 -8.55588436e-01
-4.97914329e-02 1.98473245e-01 7.19958782e-01 -5.86666584e-01
7.07005382e-01 -7.07421422e-01 1.54451191e-01 1.45270631e-01
1.72964513e-01 8.57792702e-03 3.21963757e-01 3.57122809e-01
-5.75067624e-02 -1.31717697e-01 1.11782813e+00 -2.07204610e-01
-5.62402546e-01 6.60512149e-01 4.00140435e-01 2.38300070e-01
1.21803463e+00 2.05618795e-03 -1.87906176e-01 5.09879366e-02
-7.54996121e-01 2.42274970e-01 2.87930250e-01 4.66616541e-01
4.33883339e-01 -1.04254949e+00 -7.12026775e-01 6.38817728e-01
-1.59652397e-01 5.77142477e-01 5.76125324e-01 4.74785984e-01
-4.50493962e-01 2.48135284e-01 -5.08686602e-02 -5.67118347e-01
-1.66185784e+00 2.47389838e-01 4.18918788e-01 1.97487786e-01
-3.75560999e-01 1.24822545e+00 4.36956018e-01 -6.72574341e-01
3.86187315e-01 1.14172120e-02 -2.85827518e-01 1.24798983e-01
6.20494962e-01 3.35368186e-01 2.07244083e-01 -1.15014350e+00
-6.45777643e-01 1.21547174e+00 -1.22799009e-01 2.41582736e-01
1.37251770e+00 -1.13742173e-01 -3.40176523e-01 -4.66962963e-01
1.13324797e+00 1.53176740e-01 -1.53830969e+00 -6.04709908e-02
-1.02773696e-01 -5.71531415e-01 -1.45053968e-01 -9.17675853e-01
-1.64348280e+00 9.41802680e-01 7.79297888e-01 -5.60581982e-01
1.16850507e+00 2.21362725e-01 7.99698174e-01 -9.45749953e-02
4.43941385e-01 -1.01065910e+00 -5.95566742e-02 2.70876646e-01
8.72249484e-01 -1.35542595e+00 -1.51010022e-01 -1.18038189e+00
-4.27545488e-01 9.18817401e-01 9.24624801e-01 2.01053306e-01
5.78745008e-01 2.47523069e-01 1.91470310e-01 -7.73537681e-02
-7.49278292e-02 -2.44511724e-01 3.89496803e-01 5.60530603e-01
3.39771181e-01 1.16095471e-03 7.26752207e-02 7.81201363e-01
-2.52741873e-01 -2.09714219e-01 -2.00936794e-01 8.93939197e-01
-1.31734550e-01 -1.28299546e+00 -6.89910293e-01 2.10005511e-02
-8.32706869e-01 2.83185691e-01 -3.78240764e-01 6.75473034e-01
4.30939794e-01 1.03435135e+00 1.38869127e-02 -5.76772243e-02
3.57101887e-01 2.63713717e-01 6.95504785e-01 -8.66773546e-01
-3.81170869e-01 4.25157100e-01 2.59840880e-02 -6.97157145e-01
-3.67166996e-01 -7.23652959e-01 -1.43226981e+00 -3.03624272e-02
-3.05810004e-01 -1.92283187e-02 7.93625832e-01 9.98099148e-01
3.53929788e-01 3.63265038e-01 6.04847431e-01 -1.01462317e+00
-3.87071818e-01 -8.97542834e-01 -5.83105206e-01 3.08922470e-01
1.80418789e-01 -8.78805339e-01 -1.29770160e-01 1.51775815e-02] | [13.42341423034668, 0.5694437026977539] |
9787874d-5f65-4c86-8807-b93959dd4501 | msn-efficient-online-mask-selection-network | 2106.10452 | null | https://arxiv.org/abs/2106.10452v1 | https://arxiv.org/pdf/2106.10452v1.pdf | MSN: Efficient Online Mask Selection Network for Video Instance Segmentation | In this work we present a novel solution for Video Instance Segmentation(VIS), that is automatically generating instance level segmentation masks along with object class and tracking them in a video. Our method improves the masks from segmentation and propagation branches in an online manner using the Mask Selection Network (MSN) hence limiting the noise accumulation during mask tracking. We propose an effective design of MSN by using patch-based convolutional neural network. The network is able to distinguish between very subtle differences between the masks and choose the better masks out of the associated masks accurately. Further, we make use of temporal consistency and process the video sequences in both forward and reverse manner as a post processing step to recover lost objects. The proposed method can be used to adapt any video object segmentation method for the task of VIS. Our method achieves a score of 49.1 mAP on 2021 YouTube-VIS Challenge and was ranked third place among more than 30 global teams. Our code will be available at https://github.com/SHI-Labs/Mask-Selection-Networks. | ['Humphrey Shi', 'Harsh Maheshwari', 'Shubhika Garg', 'Jiachen Li', 'Vidit Goel'] | 2021-06-19 | null | null | null | null | ['video-instance-segmentation'] | ['computer-vision'] | [ 3.53277802e-01 4.60978523e-02 -4.68251854e-02 -2.77578354e-01
-6.34803951e-01 -7.28663683e-01 2.60232687e-01 -2.78325945e-01
-5.61960399e-01 4.95429754e-01 -1.59582257e-01 -8.68367925e-02
1.04207151e-01 -6.04512930e-01 -8.92535269e-01 -4.73791778e-01
-5.42476922e-02 6.20414674e-01 1.06581151e+00 -1.62814513e-01
3.28889668e-01 5.43138385e-01 -1.55928826e+00 6.21106923e-01
6.70232117e-01 1.10900819e+00 4.90671486e-01 7.61890769e-01
-1.30260691e-01 5.97660840e-01 -4.83422071e-01 -2.01154113e-01
8.81907403e-01 -4.04167622e-01 -9.42623854e-01 2.48195902e-01
7.67959833e-01 -3.14731181e-01 -2.90149152e-01 1.06593359e+00
3.19559127e-01 6.35355189e-02 2.14016914e-01 -1.20546055e+00
2.50938088e-01 9.69681263e-01 -7.20315456e-01 4.79342252e-01
-7.08987191e-02 3.29107642e-01 6.88376606e-01 -8.85670066e-01
8.63058627e-01 8.76838863e-01 7.35598087e-01 6.30777955e-01
-9.13668752e-01 -6.21029139e-01 3.96506816e-01 2.39359543e-01
-1.42075562e+00 -5.97275555e-01 5.65332830e-01 -6.86186373e-01
6.67425811e-01 3.22055846e-01 7.15835869e-01 6.91587508e-01
-1.38439581e-01 9.11741734e-01 7.33251393e-01 -1.79402620e-01
1.82149690e-02 -5.66932149e-02 2.46155206e-02 7.83159733e-01
-2.09975126e-03 -1.58357099e-01 -4.83144999e-01 2.17513636e-01
8.34622681e-01 -1.43507630e-01 -3.82453591e-01 -3.10141176e-01
-1.36359227e+00 3.08285445e-01 4.73955452e-01 3.86435896e-01
-4.70269978e-01 4.43117291e-01 3.10405612e-01 1.34774297e-01
4.73368227e-01 3.30578506e-01 -6.20253444e-01 -5.18430993e-02
-1.70329225e+00 2.26252556e-01 3.88346404e-01 7.65780628e-01
8.86760831e-01 -9.30151865e-02 -5.19658625e-01 6.03252649e-01
1.09047718e-01 -3.48487534e-02 1.12186112e-01 -1.36594796e+00
2.68538386e-01 4.32716012e-01 1.51171342e-01 -8.56275737e-01
-4.26725388e-01 -4.26248312e-01 -3.95115554e-01 4.08178985e-01
6.15516901e-01 -3.35289121e-01 -1.33843791e+00 1.44509363e+00
5.71407855e-01 5.14914870e-01 -5.84528387e-01 9.83974040e-01
6.58403218e-01 5.01767695e-01 -1.01491258e-01 2.25660559e-02
1.24997973e+00 -1.34028912e+00 -4.97910053e-01 -2.15607867e-01
4.59830493e-01 -7.49384284e-01 5.30656695e-01 6.47950113e-01
-1.40002716e+00 -7.42460907e-01 -1.04464173e+00 9.24828276e-02
-2.38089338e-01 2.15890065e-01 3.49659264e-01 5.29711545e-01
-1.40746987e+00 9.86743510e-01 -1.04958272e+00 -2.89789170e-01
8.53820205e-01 7.21471667e-01 -4.61389497e-02 2.25581974e-01
-7.41447866e-01 4.42514390e-01 6.01417422e-01 3.05375278e-01
-9.50032353e-01 -6.09130263e-01 -5.29242873e-01 -1.03681535e-01
7.41858542e-01 -4.96126294e-01 1.17309093e+00 -1.55098617e+00
-1.18002057e+00 7.91315973e-01 -1.81401908e-01 -6.89773202e-01
8.91581833e-01 -2.08838314e-01 -1.18421175e-01 3.33519965e-01
2.24967711e-02 1.20544410e+00 9.74736035e-01 -1.22507310e+00
-1.00018740e+00 4.24558222e-02 -7.33273998e-02 5.19688167e-02
4.52500470e-02 1.41426712e-01 -1.10267806e+00 -5.92886806e-01
1.77566379e-01 -7.36922741e-01 -3.38469148e-01 1.66575357e-01
-5.07168949e-01 2.01285854e-02 1.04169154e+00 -8.83428752e-01
1.22718728e+00 -2.08134437e+00 2.12113604e-01 3.19553167e-01
4.39329624e-01 4.71348524e-01 -1.36033952e-01 2.10590288e-02
7.82853141e-02 1.16468757e-01 -4.36563402e-01 -4.90591913e-01
-3.20169836e-01 -9.56731960e-02 3.95800136e-02 4.84969884e-01
3.54696184e-01 7.78014660e-01 -6.97963059e-01 -5.35606682e-01
3.58674228e-01 3.98684472e-01 -5.46995342e-01 -3.74360234e-02
-5.23114860e-01 5.23391426e-01 -1.80267021e-01 6.84735417e-01
9.19348776e-01 -8.76938775e-02 5.37165105e-02 -2.72374332e-01
-3.49130720e-01 1.59022853e-01 -1.41475773e+00 1.69720435e+00
2.25164771e-01 8.76001894e-01 4.92063224e-01 -9.20988262e-01
6.94641769e-01 1.80697173e-01 7.95391500e-01 -3.25577170e-01
3.36474538e-01 2.25294590e-01 1.02981523e-01 -3.81383151e-01
5.65044045e-01 4.34282482e-01 2.48862371e-01 2.13045180e-01
1.12937696e-01 2.64808357e-01 6.54633284e-01 2.39954025e-01
1.16232181e+00 5.60707748e-01 -1.55219644e-01 -4.08542961e-01
5.23503065e-01 2.11512938e-01 7.67225802e-01 9.18713689e-01
-5.04043281e-01 1.05114377e+00 4.48477805e-01 -5.29397428e-01
-8.85904908e-01 -6.88238144e-01 2.10940540e-02 9.85808492e-01
2.96400994e-01 -4.26895499e-01 -1.09098327e+00 -9.38575447e-01
-2.22811282e-01 3.28336626e-01 -5.75006604e-01 2.84550697e-01
-1.02178824e+00 -3.62914413e-01 4.48973864e-01 4.15578812e-01
6.35754466e-01 -1.29238117e+00 -6.42593622e-01 2.72760838e-01
-2.71645933e-01 -1.02972710e+00 -6.37295127e-01 2.65599102e-01
-6.66237772e-01 -1.16795468e+00 -7.06985116e-01 -8.76149952e-01
5.47307909e-01 3.08160216e-01 1.00161910e+00 4.40159619e-01
-2.44274899e-01 3.64523567e-02 -2.80386388e-01 -1.99555680e-01
-3.52598906e-01 2.20140845e-01 -2.15082049e-01 3.06063145e-01
1.67263582e-01 -3.52243960e-01 -8.52705121e-01 5.08177817e-01
-8.27378809e-01 2.99036503e-01 2.54503131e-01 2.89750665e-01
7.08078563e-01 -4.93814424e-02 3.20556670e-01 -1.05032742e+00
-8.06365982e-02 -3.87871027e-01 -9.27967370e-01 8.28208327e-02
-1.82857037e-01 -2.76461810e-01 2.46495485e-01 -2.31066048e-01
-6.83210373e-01 4.67271686e-01 -2.39834875e-01 -4.66476232e-01
-2.38348573e-01 -1.25534117e-01 -9.25548822e-02 -2.60723770e-01
5.30403376e-01 -1.25436142e-01 -1.64048344e-01 -5.76890707e-01
2.23302364e-01 4.38694954e-01 4.48233396e-01 -2.81515598e-01
7.82139421e-01 5.87496936e-01 -3.01449239e-01 -4.85089183e-01
-7.03165710e-01 -6.80569589e-01 -8.67811680e-01 -6.49249256e-01
1.02294099e+00 -8.38398039e-01 -3.15112948e-01 5.79883039e-01
-1.13944948e+00 -7.69218385e-01 -2.12405697e-01 4.86541614e-02
-4.19804007e-01 2.02622920e-01 -5.85017622e-01 -5.59524179e-01
-1.66342944e-01 -1.31477559e+00 1.04449368e+00 3.93182158e-01
-1.45712972e-01 -4.91031528e-01 -2.03776449e-01 4.29068208e-01
3.84594381e-01 2.56445229e-01 1.69701859e-01 -5.64931989e-01
-1.10053992e+00 1.39141247e-01 -3.08376342e-01 3.51683706e-01
-2.21687853e-02 4.07236934e-01 -9.44229186e-01 -2.24786967e-01
-2.99101919e-01 8.98830295e-02 1.38861334e+00 7.94209599e-01
1.45921016e+00 5.13200415e-04 -5.58387220e-01 8.46288800e-01
1.47233009e+00 2.54000217e-01 7.61261225e-01 4.73582417e-01
1.01189888e+00 4.71023977e-01 5.23291469e-01 2.03620523e-01
4.27150074e-03 7.93012202e-01 5.73415041e-01 -2.19800100e-01
-5.42350292e-01 1.21262558e-01 2.33189687e-01 3.45008969e-01
-1.35963380e-01 -4.76492107e-01 -9.48339105e-01 7.38135993e-01
-2.01016593e+00 -8.85011613e-01 -3.05440962e-01 2.15969038e+00
5.67614436e-01 4.77527022e-01 4.20664668e-01 -3.44073921e-02
1.04152226e+00 1.49540633e-01 -4.04141456e-01 -1.47191465e-01
-7.54189193e-02 3.15087408e-01 8.26265574e-01 5.82070470e-01
-1.46507597e+00 1.20590651e+00 5.29021406e+00 8.81454110e-01
-1.00150979e+00 1.91101715e-01 8.12352359e-01 -1.98446721e-01
1.45176396e-01 1.55122206e-02 -8.45163822e-01 5.01081109e-01
6.60339236e-01 4.13630038e-01 4.08138037e-01 4.98011261e-01
3.59140575e-01 -2.88582027e-01 -9.75341856e-01 8.33479941e-01
-1.58889234e-01 -1.78705144e+00 -1.90466642e-01 -1.93840802e-01
9.24558997e-01 3.69488418e-01 -1.78858653e-01 -1.39741218e-02
-7.97084421e-02 -8.11829448e-01 1.15991414e+00 4.99888390e-01
5.69780171e-01 -6.43940330e-01 6.12679899e-01 1.32062465e-01
-1.22175193e+00 5.16682789e-02 -3.51044722e-02 2.58404273e-03
3.27989161e-01 4.73853678e-01 -7.57657230e-01 3.84021610e-01
9.94193017e-01 7.20774889e-01 -6.30423188e-01 1.59709573e+00
1.21401198e-01 7.55500913e-01 -4.56238329e-01 3.77463043e-01
3.08454484e-01 -7.37313777e-02 7.17832327e-01 1.42048728e+00
2.10778654e-01 -1.10081308e-01 3.82268518e-01 7.50421584e-01
-2.23710701e-01 -1.68257177e-01 -1.41686276e-01 8.28196034e-02
3.14570725e-01 1.32247794e+00 -1.54593801e+00 -4.26930010e-01
-8.36488381e-02 1.18067276e+00 1.53553024e-01 2.63041943e-01
-9.71453130e-01 -1.62723437e-01 6.17964208e-01 4.24938977e-01
8.95527899e-01 -2.31799603e-01 -3.75419378e-01 -8.27460170e-01
7.19370469e-02 -7.50444114e-01 2.71557748e-01 -5.62510610e-01
-8.02650630e-01 7.34862626e-01 -1.07943870e-01 -1.12165582e+00
2.88855582e-01 -5.24114490e-01 -5.88786244e-01 5.06513000e-01
-1.30846334e+00 -7.91091263e-01 -3.17162007e-01 4.01005596e-01
8.84621918e-01 1.28618643e-01 6.21814094e-02 6.66443169e-01
-6.66055679e-01 3.85043681e-01 -1.42860502e-01 2.44881213e-01
6.57255650e-01 -1.13321495e+00 5.84150553e-01 1.26947033e+00
2.59868503e-01 4.76380177e-02 5.95463932e-01 -8.03721011e-01
-7.33073413e-01 -1.33555686e+00 7.12186515e-01 -3.86628985e-01
3.90953630e-01 -5.50449669e-01 -7.52337873e-01 6.42063439e-01
3.96865785e-01 1.07481182e-02 5.74198775e-02 -5.07846594e-01
5.59956357e-02 -7.72946626e-02 -1.13424551e+00 4.69110459e-01
1.20147455e+00 1.22658305e-01 6.18417822e-02 5.10536253e-01
8.21309626e-01 -7.13980436e-01 -2.21643135e-01 4.55894768e-01
3.11489701e-01 -1.25510848e+00 8.04457545e-01 -2.12706044e-01
3.38659257e-01 -8.42985630e-01 1.73244774e-01 -8.04504693e-01
-3.68072510e-01 -9.62901413e-01 -1.20715629e-02 1.17788577e+00
5.25515854e-01 -2.31941983e-01 1.08339620e+00 2.75729924e-01
-3.25358391e-01 -7.46570528e-01 -8.77535880e-01 -5.22597194e-01
-4.20820594e-01 -4.97624695e-01 4.08458561e-01 5.90436518e-01
-7.41874516e-01 -2.14463487e-01 -4.32851672e-01 2.73941368e-01
5.49081326e-01 -1.68259993e-01 6.33638799e-01 -1.02543366e+00
-3.71979296e-01 -7.31222391e-01 -3.49169910e-01 -1.29080188e+00
-3.08457345e-01 -7.37778306e-01 3.08523983e-01 -1.56632340e+00
-8.59136507e-02 -5.84252238e-01 -3.80041778e-01 4.86509621e-01
-6.75511882e-02 8.55949879e-01 5.12339771e-01 9.19697583e-02
-9.25210357e-01 -6.34471104e-02 1.10539544e+00 -6.78684935e-02
-3.31897885e-01 1.44578904e-01 -4.85086411e-01 7.84036458e-01
1.12255526e+00 -6.75859690e-01 -8.55009779e-02 -4.36828673e-01
3.69506627e-02 -3.21317583e-01 3.78236562e-01 -1.28321409e+00
1.47561044e-01 -9.14276857e-03 4.87414539e-01 -8.76138031e-01
2.44957253e-01 -8.59092772e-01 3.30559671e-01 5.26572943e-01
-1.23900235e-01 -3.43517512e-02 3.27487737e-01 4.05852854e-01
2.42170468e-02 -3.15012455e-01 8.12804818e-01 -3.04491818e-01
-1.06968629e+00 4.48372126e-01 -3.31358373e-01 9.36096814e-03
1.11184430e+00 -4.65309590e-01 -1.24902077e-01 -1.22913398e-01
-8.23805988e-01 3.93753290e-01 4.86978889e-01 3.24834049e-01
4.26716805e-01 -9.72229719e-01 -6.85343266e-01 1.42648160e-01
-2.93491751e-01 1.17232449e-01 3.38426501e-01 1.11287880e+00
-1.05454910e+00 -3.37654687e-02 -1.97969958e-01 -7.78177798e-01
-1.34354091e+00 3.26621771e-01 4.99760896e-01 -6.53576851e-02
-8.26632738e-01 1.32999194e+00 9.07670930e-02 -3.11933346e-02
4.62462395e-01 -3.30682963e-01 -1.49471313e-01 -4.09999788e-02
6.53263628e-01 3.77700746e-01 1.02339596e-01 -7.31067598e-01
-5.25973737e-01 5.11945128e-01 -1.35456264e-01 -6.13186955e-02
1.14496660e+00 -2.27586761e-01 -1.04537643e-01 2.83181548e-01
9.25872087e-01 -4.51840786e-03 -1.71591580e+00 -9.27949417e-03
1.05802581e-01 -3.88355106e-01 7.90783539e-02 -7.63693750e-01
-1.49609065e+00 4.19161916e-01 6.58726037e-01 2.04822645e-01
1.20410240e+00 2.53577717e-02 9.70962286e-01 -1.16019011e-01
2.17754140e-01 -1.18068385e+00 -1.36375129e-01 3.49595070e-01
5.71471930e-01 -1.07636046e+00 -1.31018758e-01 -6.33341074e-01
-3.80911142e-01 1.10239196e+00 7.28325784e-01 -4.28674310e-01
5.09709120e-01 4.57695454e-01 1.80571511e-01 -2.62726188e-01
-4.29416925e-01 -3.29073817e-01 4.04811263e-01 5.31808555e-01
2.37763450e-01 1.90315023e-03 -3.25530469e-01 1.98644906e-01
8.98485258e-02 -3.54323946e-02 5.06302476e-01 7.75986195e-01
-6.69216990e-01 -1.19148350e+00 -3.63497406e-01 5.13581872e-01
-5.49006641e-01 -1.19181186e-01 -3.21474224e-01 5.59277117e-01
7.46128619e-01 8.57991040e-01 1.01244777e-01 -3.62341017e-01
4.81854454e-02 4.55952138e-02 5.02279997e-01 -5.52659631e-01
-8.64770234e-01 2.55793393e-01 1.97951913e-01 -1.02267694e+00
-6.80346131e-01 -7.30131984e-01 -1.25583756e+00 -1.23295099e-01
-2.15434402e-01 -1.66108385e-01 5.35840392e-01 8.60662460e-01
4.60765958e-01 7.20829546e-01 2.92426199e-01 -1.32767308e+00
1.85452610e-01 -6.53609872e-01 -2.13015839e-01 2.15360671e-01
4.42286998e-01 -5.20560563e-01 -1.44725323e-01 2.99461246e-01] | [9.147354125976562, -0.18109728395938873] |
60a0276a-029c-46e1-a1e6-e8cec87e27ea | regularization-free-estimation-in-trace | 1504.06305 | null | http://arxiv.org/abs/1504.06305v1 | http://arxiv.org/pdf/1504.06305v1.pdf | Regularization-free estimation in trace regression with symmetric positive semidefinite matrices | Over the past few years, trace regression models have received considerable
attention in the context of matrix completion, quantum state tomography, and
compressed sensing. Estimation of the underlying matrix from
regularization-based approaches promoting low-rankedness, notably nuclear norm
regularization, have enjoyed great popularity. In the present paper, we argue
that such regularization may no longer be necessary if the underlying matrix is
symmetric positive semidefinite (\textsf{spd}) and the design satisfies certain
conditions. In this situation, simple least squares estimation subject to an
\textsf{spd} constraint may perform as well as regularization-based approaches
with a proper choice of the regularization parameter, which entails knowledge
of the noise level and/or tuning. By contrast, constrained least squares
estimation comes without any tuning parameter and may hence be preferred due to
its simplicity. | ['Matthias Hein', 'Martin Slawski', 'Ping Li'] | 2015-04-23 | regularization-free-estimation-in-trace-1 | http://papers.nips.cc/paper/5726-regularization-free-estimation-in-trace-regression-with-symmetric-positive-semidefinite-matrices | http://papers.nips.cc/paper/5726-regularization-free-estimation-in-trace-regression-with-symmetric-positive-semidefinite-matrices.pdf | neurips-2015-12 | ['quantum-state-tomography'] | ['medical'] | [ 6.37109339e-01 1.57925576e-01 -7.14394599e-02 -1.28438488e-01
-8.19863200e-01 -4.91277039e-01 4.94034082e-01 1.05118364e-01
-6.93979323e-01 8.04118335e-01 1.31882250e-01 -4.53700632e-01
-5.98158062e-01 -3.29340637e-01 -4.99058008e-01 -1.07523382e+00
5.02528474e-02 4.24192160e-01 -1.62178904e-01 -3.03653777e-01
2.96110600e-01 2.44526967e-01 -9.55188453e-01 -5.54489613e-01
8.94841015e-01 8.26378345e-01 -3.53120044e-02 2.47747436e-01
2.84572929e-01 5.58714628e-01 -6.78495988e-02 -4.60702389e-01
3.81823301e-01 -3.94679785e-01 -5.27919769e-01 2.03203082e-01
1.37539387e-01 3.75068560e-02 -6.17552161e-01 1.50554764e+00
3.45007181e-01 1.53129235e-01 6.28822625e-01 -7.99410820e-01
-1.13697484e-01 4.30198431e-01 -6.37368202e-01 1.32368803e-01
2.24401489e-01 -1.85930088e-01 1.25047970e+00 -9.98009324e-01
6.40849173e-01 5.76007724e-01 4.63649422e-01 1.46357611e-01
-1.82605231e+00 -3.46067905e-01 -1.32031903e-01 5.27142696e-02
-1.71758544e+00 -6.35290265e-01 9.11829412e-01 -3.40303332e-01
4.19291109e-01 4.22983468e-01 2.60098755e-01 9.76334989e-01
-1.24604352e-01 5.37573397e-01 1.22529936e+00 -6.50167584e-01
4.00891989e-01 1.96526885e-01 -1.39258401e-02 7.03137696e-01
6.81750953e-01 -1.34792998e-01 -5.46612501e-01 -4.17662531e-01
4.42673534e-01 -3.21105152e-01 -5.11456549e-01 -8.71353269e-01
-1.35827577e+00 9.47840154e-01 -1.03179522e-01 3.62490565e-01
-4.55878615e-01 1.19395852e-01 4.43965018e-01 3.18653286e-01
4.13546503e-01 5.36870122e-01 8.66689086e-02 -6.82151020e-02
-1.08730412e+00 1.89525604e-01 8.28527033e-01 8.45150709e-01
8.90774131e-01 2.04138890e-01 5.86360060e-02 6.95707798e-01
9.34993848e-02 5.94613433e-01 -3.24151069e-01 -1.05342269e+00
6.75877571e-01 1.14374548e-01 4.80613261e-01 -1.03764844e+00
-3.88272166e-01 -5.18794656e-01 -1.09425259e+00 -1.31735191e-01
6.07693851e-01 -2.33840719e-01 -4.08532739e-01 1.75858796e+00
1.08725250e-01 3.23070027e-02 3.76454592e-02 1.17948484e+00
6.55895174e-02 4.48602349e-01 -3.52002680e-01 -5.75192213e-01
9.70145524e-01 -1.85202777e-01 -8.18759441e-01 -3.61972809e-01
7.24536061e-01 -6.51555419e-01 8.88364434e-01 6.40578032e-01
-8.35605681e-01 2.14522868e-01 -1.07268631e+00 2.80303717e-01
2.53748566e-01 1.93714172e-01 5.94268322e-01 8.94793272e-01
-7.21395254e-01 4.65094566e-01 -6.72608733e-01 -1.63649693e-01
-1.07843637e-01 5.40697217e-01 -4.46752071e-01 -1.96485847e-01
-1.04581428e+00 6.59048676e-01 3.45767319e-01 4.91089135e-01
-4.83088613e-01 -2.97131747e-01 -6.58758163e-01 -6.06959760e-02
7.75377095e-01 -4.15519804e-01 7.36949384e-01 -5.80673277e-01
-1.47154975e+00 8.43026400e-01 1.02042267e-02 -4.14492965e-01
4.66399699e-01 -8.80718678e-02 -1.49573639e-01 2.54294455e-01
-1.03256240e-01 -3.17912310e-01 1.05311489e+00 -7.75313497e-01
1.54613376e-01 -3.94641548e-01 1.52719676e-01 1.05348155e-01
-1.57990739e-01 -5.60682686e-03 -2.97156215e-01 -3.32966179e-01
7.46777833e-01 -1.27260578e+00 -5.62799215e-01 -3.70025098e-01
-7.01361895e-01 1.97306097e-01 1.61264181e-01 -2.47515842e-01
1.23419428e+00 -2.10400891e+00 3.77851665e-01 7.65629947e-01
1.11636065e-01 2.67097540e-02 1.69075221e-01 6.74695313e-01
-7.97513053e-02 -1.49415106e-01 -4.60594237e-01 -2.36942410e-01
7.77631029e-02 6.76440820e-02 -3.17773782e-02 1.13511348e+00
1.24762040e-02 3.80949616e-01 -7.17841506e-01 -2.56104767e-01
2.46427417e-01 3.79542619e-01 -8.26869547e-01 -3.18074673e-01
-1.44377753e-01 7.70286620e-01 -8.49330544e-01 1.32270157e-01
6.59462750e-01 -4.12505031e-01 6.23212397e-01 -3.41454983e-01
-3.61191899e-01 2.95693278e-01 -1.78713930e+00 1.63985372e+00
-2.24025652e-01 4.35827315e-01 7.62848258e-01 -1.30964327e+00
6.35453343e-01 3.54819000e-01 7.84019411e-01 -6.35238528e-01
1.41697258e-01 3.59076649e-01 2.44131371e-01 -3.45248371e-01
5.30941010e-01 -5.11864364e-01 -8.80716071e-02 5.19819438e-01
-3.81235331e-01 -5.04849218e-02 2.44243786e-01 5.43295979e-01
1.12255836e+00 -3.72667275e-02 3.21729243e-01 -6.00158989e-01
6.62966371e-01 -1.53766572e-01 6.40876174e-01 7.47381330e-01
2.96219569e-02 5.42498469e-01 6.94887578e-01 1.04006454e-01
-1.29587746e+00 -8.09179544e-01 -4.69262838e-01 6.12464428e-01
5.95223643e-02 -5.90531230e-01 -4.83609170e-01 -6.53832704e-02
-2.42070571e-01 4.95128691e-01 -2.59896964e-01 -1.07516624e-01
-4.75660592e-01 -9.57307518e-01 3.60239327e-01 -6.86952919e-02
3.37647110e-01 -3.12939852e-01 -2.47657374e-01 1.93202764e-01
-1.68645635e-01 -1.09677505e+00 -4.86828417e-01 6.19831443e-01
-6.90372825e-01 -7.41392553e-01 -7.09543109e-01 -1.48234442e-01
8.11139822e-01 2.96976686e-01 6.74402118e-01 -2.88809329e-01
7.98172280e-02 4.64058012e-01 -1.78476840e-01 1.88381732e-01
-9.53290015e-02 -4.43776362e-02 3.32585573e-01 3.58497620e-01
5.26580140e-02 -7.30330765e-01 -5.04728377e-01 2.09035382e-01
-1.05104625e+00 -1.84225202e-01 5.12373865e-01 8.42262208e-01
4.78801250e-01 -2.73649450e-02 2.79099405e-01 -1.22636116e+00
4.95831192e-01 -3.19296360e-01 -8.00335050e-01 1.11312732e-01
-6.42680526e-01 4.41823781e-01 5.94883025e-01 -1.07939012e-01
-8.56561244e-01 4.00515413e-03 8.76702964e-02 -3.93524796e-01
2.43851885e-01 1.00949335e+00 -1.82647362e-01 -3.51637304e-01
7.60707498e-01 2.95839578e-01 -9.89010409e-02 -3.60900700e-01
1.96736217e-01 4.27492887e-01 2.38756716e-01 -7.17926204e-01
9.22142446e-01 5.89686155e-01 4.93476510e-01 -1.05214834e+00
-9.08997416e-01 -5.24487078e-01 -5.55178463e-01 1.62588675e-02
5.79553127e-01 -8.14458728e-01 -7.40650892e-01 -8.82187262e-02
-6.93071127e-01 5.65414988e-02 -1.65725604e-01 9.42437172e-01
-7.71125734e-01 7.47990966e-01 -4.16975349e-01 -1.22845817e+00
8.18972960e-02 -1.11753249e+00 6.29334331e-01 -3.29157233e-01
-1.08561389e-01 -1.00449169e+00 -2.38363847e-01 4.09052283e-01
4.13870513e-01 6.74865842e-02 7.67733574e-01 -5.02354443e-01
-7.13812411e-01 -4.49003309e-01 -1.67991668e-01 2.61178464e-01
-1.56736001e-01 -4.76827085e-01 -6.00336373e-01 -3.88800174e-01
2.83042312e-01 -2.70096093e-01 7.65721321e-01 2.92965055e-01
7.28481054e-01 -3.43379378e-01 4.51901890e-02 6.04991913e-01
1.33882368e+00 -2.07780823e-01 4.54508126e-01 1.91040173e-01
6.24908447e-01 3.80749851e-01 4.54296350e-01 7.73812711e-01
1.13583811e-01 8.27457786e-01 4.14561898e-01 2.95356750e-01
5.40139973e-01 -9.52382982e-02 2.16595873e-01 9.56185758e-01
-1.03791460e-01 2.67551132e-02 -9.06374037e-01 4.83592935e-02
-1.68145728e+00 -8.94951105e-01 -3.61454695e-01 2.74907756e+00
7.14334548e-01 2.27063894e-01 -9.21639428e-02 2.73554862e-01
6.46298409e-01 4.38430876e-01 -5.07268131e-01 -1.35434940e-01
-3.86603177e-01 5.56833623e-03 8.59553516e-01 6.75034344e-01
-7.46760368e-01 5.31321406e-01 5.68009090e+00 9.29733157e-01
-9.50200021e-01 1.61034912e-02 6.90344051e-02 1.10108756e-01
-7.37059414e-01 5.08149207e-01 -5.57036459e-01 3.70771021e-01
7.15185285e-01 -7.40703866e-02 5.70796490e-01 4.05814886e-01
5.01320302e-01 -5.01222849e-01 -9.56784964e-01 1.07434916e+00
7.39312731e-03 -1.04511070e+00 -3.72191340e-01 5.16682565e-01
6.78122640e-01 -1.03506170e-01 1.37872636e-01 -4.76326011e-02
-1.67305663e-01 -6.42903984e-01 5.18806756e-01 5.13756335e-01
8.07827294e-01 -6.55671895e-01 4.80571806e-01 6.77833855e-01
-7.05880046e-01 1.45589948e-01 -4.65118587e-01 -9.84140188e-02
2.86388993e-01 1.12581325e+00 -4.45390344e-01 4.90210623e-01
-2.54128408e-02 4.50070441e-01 -3.67811299e-03 8.48124683e-01
-9.43380222e-02 9.33239996e-01 -7.10539103e-01 1.42146572e-01
3.41677815e-01 -1.13960230e+00 9.56382632e-01 7.77108192e-01
3.00348908e-01 2.06074908e-01 4.07845765e-01 8.55459809e-01
7.10549280e-02 3.18496197e-01 -6.60305679e-01 -3.17813545e-01
3.33743662e-01 1.07006264e+00 -6.43374383e-01 1.33269295e-01
-4.88972962e-01 7.71762013e-01 1.91000938e-01 6.05723917e-01
-5.19454420e-01 -4.31064069e-02 3.32893431e-01 3.58008116e-01
2.36550257e-01 -6.26386583e-01 -2.42095545e-01 -1.42433035e+00
2.34112799e-01 -7.15258002e-01 2.10973322e-01 -2.96600878e-01
-9.26305473e-01 1.61663800e-01 -1.35567829e-01 -1.16255438e+00
-1.62462160e-01 -4.61760700e-01 -3.50057244e-01 6.36263251e-01
-1.27617741e+00 -6.28915608e-01 3.35796803e-01 6.31972432e-01
-2.78935134e-01 4.88108285e-02 6.21773720e-01 4.24627483e-01
-8.59602869e-01 2.27733910e-01 5.69917262e-01 -2.06313133e-01
4.94253129e-01 -9.81484473e-01 -4.61958587e-01 9.18688178e-01
2.13026807e-01 9.65036154e-01 1.28736079e+00 -5.45698404e-01
-1.83886826e+00 -6.81297719e-01 7.98934460e-01 2.29764730e-01
1.06281185e+00 -4.52834725e-01 -6.59206986e-01 5.67438424e-01
-8.55473727e-02 7.52790049e-02 6.73869014e-01 3.54540050e-01
-3.89372051e-01 -1.11365326e-01 -8.04380536e-01 6.17220342e-01
9.20833945e-01 -9.90645170e-01 -9.46785416e-03 6.58797085e-01
-7.77257904e-02 -1.61935166e-01 -8.30467284e-01 3.96582693e-01
1.86781347e-01 -9.10328448e-01 7.22326040e-01 -4.26785856e-01
-2.48916924e-01 -2.82631189e-01 -5.38707197e-01 -9.41034257e-01
-2.86127210e-01 -1.13124084e+00 1.91528231e-01 8.42488468e-01
5.25613785e-01 -4.71680492e-01 1.02728403e+00 8.96633685e-01
-2.25060154e-02 -5.22155881e-01 -1.18176293e+00 -7.17493534e-01
-8.50702748e-02 -6.62746072e-01 -1.83375388e-01 7.78739691e-01
2.84318507e-01 5.16080081e-01 -9.51841891e-01 2.74208575e-01
8.22156191e-01 1.48197785e-01 5.88881791e-01 -1.21262622e+00
-4.65384096e-01 -8.27725306e-02 -2.61630535e-01 -1.18889642e+00
2.77462929e-01 -9.80574369e-01 1.57834683e-02 -1.06023562e+00
2.79292077e-01 -6.87998176e-01 -1.20258883e-01 -3.00090741e-02
1.13921925e-01 1.71473712e-01 8.79073143e-02 3.24571848e-01
-7.83086658e-01 6.82756960e-01 1.03441358e+00 5.16919903e-02
-8.27793553e-02 5.21791995e-01 -6.00757241e-01 6.20146632e-01
6.54746413e-01 -4.62157488e-01 -3.64470840e-01 -1.20940350e-01
1.03331387e+00 4.62405771e-01 2.58401543e-01 -7.36862898e-01
3.66831809e-01 -1.14835210e-01 -2.70941705e-01 -7.75516480e-02
6.99789584e-01 -8.47496092e-01 3.75037074e-01 1.40052095e-01
-3.20395619e-01 -3.75531346e-01 -3.66931677e-01 8.10438156e-01
-1.03918925e-01 -6.48779094e-01 7.58632839e-01 8.45012534e-03
-2.80787259e-01 2.30029225e-01 -4.50384051e-01 1.22051574e-01
5.62671065e-01 -3.89524370e-01 1.76895618e-01 -7.53240228e-01
-8.59790325e-01 -9.25653353e-02 3.82786125e-01 -2.74775624e-01
4.97226834e-01 -1.12475550e+00 -6.86174095e-01 7.73519129e-02
1.35895595e-01 -3.88282359e-01 2.39671767e-01 1.60282946e+00
-1.04201481e-01 6.40717566e-01 3.85747284e-01 -4.76908565e-01
-8.00262272e-01 3.96062911e-01 1.80467904e-01 -1.75983384e-01
-6.51506841e-01 6.17068052e-01 1.20016277e-01 -2.11020589e-01
8.07631314e-02 1.98957816e-01 2.01279357e-01 4.58824374e-02
1.10062256e-01 3.44793886e-01 1.12514183e-01 -9.87801790e-01
-2.86902279e-01 3.65148097e-01 -8.72338843e-03 -2.89793819e-01
1.27050376e+00 -4.83553469e-01 -1.40721232e-01 3.86692613e-01
1.07750380e+00 5.70551693e-01 -1.13435602e+00 -6.02584600e-01
3.85334313e-01 -3.34270328e-01 3.50343436e-01 -9.33994129e-02
-9.35802460e-01 6.89691782e-01 1.04789929e-02 1.49581879e-01
8.59352767e-01 -1.51651919e-01 3.33625942e-01 7.57924259e-01
7.06523478e-01 -1.11729705e+00 -1.38081089e-01 6.10464036e-01
5.02252638e-01 -1.08706498e+00 2.86298722e-01 -5.31742156e-01
-6.01102591e-01 9.70082581e-01 -6.74170107e-02 -2.63115287e-01
6.45031452e-01 3.63494828e-02 -4.58411306e-01 -2.04643115e-01
-3.63509148e-01 -4.13916945e-01 2.68929720e-01 2.02292427e-01
4.83273208e-01 3.56557183e-02 -6.04183495e-01 1.82755575e-01
-5.10163493e-02 -2.62124866e-01 6.11803114e-01 6.65545166e-01
-4.37148273e-01 -1.22175682e+00 -3.16176772e-01 5.68027556e-01
-5.71483612e-01 -2.94724405e-01 -5.85785992e-02 5.20511746e-01
-2.69751966e-01 9.81446743e-01 -4.65937734e-01 2.77363765e-03
2.91345865e-01 3.39870062e-03 6.01345181e-01 -5.80036521e-01
-1.71815053e-01 5.24680674e-01 3.83395761e-01 -4.67560172e-01
-5.26968896e-01 -9.81202900e-01 -1.06684899e+00 -3.57572019e-01
-6.74402297e-01 5.77190578e-01 5.26769519e-01 1.11354828e+00
-2.90316008e-02 -1.34640515e-01 6.12789690e-01 -4.04486030e-01
-1.07854831e+00 -8.13480616e-01 -1.15233719e+00 1.91362351e-01
3.19984615e-01 -5.66545486e-01 -5.44179618e-01 -3.87124836e-01] | [6.753365516662598, 4.620147705078125] |
177f6d1e-c22b-4e5f-b800-f8b1792e3aae | kpeval-towards-fine-grained-semantic-based | 2303.15422 | null | https://arxiv.org/abs/2303.15422v1 | https://arxiv.org/pdf/2303.15422v1.pdf | KPEval: Towards Fine-grained Semantic-based Evaluation of Keyphrase Extraction and Generation Systems | Despite the significant advancements in keyphrase extraction and keyphrase generation methods, the predominant approach for evaluation only relies on exact matching with human references and disregards reference-free attributes. This scheme fails to recognize systems that generate keyphrases that are semantically equivalent to the references or keyphrases that have practical utility. To better understand the strengths and weaknesses of different keyphrase systems, we propose a comprehensive evaluation framework consisting of six critical dimensions: naturalness, faithfulness, saliency, coverage, diversity, and utility. For each dimension, we discuss the desiderata and design semantic-based metrics that align with the evaluation objectives. Rigorous meta-evaluation studies demonstrate that our evaluation strategy correlates better with human preferences compared to a range of previously used metrics. Using this framework, we re-evaluate 18 keyphrase systems and further discover that (1) the best model differs in different dimensions, with pre-trained language models achieving the best in most dimensions; (2) the utility in downstream tasks does not always correlate well with reference-based metrics; and (3) large language models exhibit a strong performance in reference-free evaluation. | ['Kai-Wei Chang', 'Da Yin', 'Di wu'] | 2023-03-27 | null | null | null | null | ['keyphrase-generation', 'keyphrase-extraction'] | ['natural-language-processing', 'natural-language-processing'] | [ 4.37875167e-02 -1.26748696e-01 -5.50559878e-01 1.52600154e-01
-9.92090940e-01 -9.77511227e-01 1.28005672e+00 7.76506543e-01
-7.10804164e-01 6.30070448e-01 7.63574541e-01 -2.14110136e-01
-3.47643375e-01 -5.32336950e-01 -2.96115130e-01 -8.62800982e-03
2.11008623e-01 2.25141525e-01 3.86280239e-01 -6.57057047e-01
7.41773963e-01 4.10484195e-01 -1.81404018e+00 3.12358260e-01
7.99521446e-01 1.01456463e+00 -6.31013215e-02 5.83053470e-01
-3.07832658e-01 9.28487957e-01 -7.77260542e-01 -5.95088542e-01
3.71933021e-02 -4.31160063e-01 -1.00396621e+00 -4.37838078e-01
7.63699949e-01 -2.10681081e-01 -2.25944772e-01 8.49944413e-01
4.79904056e-01 -9.87892300e-02 7.19668806e-01 -1.25608432e+00
-9.35061812e-01 7.74375200e-01 -5.85301667e-02 1.73504069e-01
7.52966225e-01 -1.12017795e-01 1.65001297e+00 -9.67425108e-01
9.06355381e-01 1.07718849e+00 6.46164179e-01 3.38089108e-01
-1.24571276e+00 -4.68424171e-01 1.40201479e-01 -7.13161379e-03
-1.36988986e+00 -4.26531434e-01 4.00248826e-01 -3.11010420e-01
9.88983989e-01 5.58565259e-01 6.84188962e-01 1.20548129e+00
2.80168205e-01 8.42659056e-01 1.25575662e+00 -5.85643709e-01
4.86344285e-02 3.95234972e-01 2.90350139e-01 3.40878785e-01
5.35811603e-01 1.39586136e-01 -7.86625862e-01 -2.79011190e-01
3.25011373e-01 -2.10037082e-01 -3.80777895e-01 -2.33959422e-01
-1.77598095e+00 7.12450802e-01 -1.25133665e-02 4.60947305e-01
-3.92545819e-01 -1.58806309e-01 4.27958012e-01 5.22211194e-01
2.26554364e-01 1.50326169e+00 -4.49844211e-01 -2.00404122e-01
-1.34496212e+00 7.77830005e-01 1.01375329e+00 8.45033705e-01
6.32420182e-01 -2.30557397e-01 -8.02323818e-01 6.42172098e-01
-6.70987889e-02 4.67714131e-01 6.16341114e-01 -8.21458697e-01
1.53675914e-01 7.76905298e-01 4.33201253e-01 -1.22071898e+00
-3.73682648e-01 -6.52974129e-01 -4.36915785e-01 -1.06006034e-01
1.31531626e-01 1.66451991e-01 -5.67856610e-01 1.63792384e+00
-2.48766944e-01 -6.16455734e-01 2.42947172e-02 5.36462188e-01
1.05316257e+00 3.85384768e-01 1.92229003e-01 -1.83956429e-01
1.41548991e+00 -8.65570068e-01 -7.33335078e-01 -4.59062234e-02
5.19190907e-01 -1.22630095e+00 1.60702610e+00 3.09338927e-01
-1.00890994e+00 -5.06212115e-01 -1.24331355e+00 -1.89923614e-01
-7.75184274e-01 2.05480799e-01 6.30360305e-01 5.56914389e-01
-1.10319936e+00 4.67186809e-01 -1.45185113e-01 -5.30863404e-01
-2.28865296e-01 -1.27184227e-01 -1.27584249e-01 4.22329813e-01
-1.32564187e+00 1.23644400e+00 5.45837760e-01 -7.69901752e-01
-7.99901843e-01 -8.54355991e-01 -5.36209702e-01 -8.68766978e-02
5.06353617e-01 -7.94415414e-01 1.33759165e+00 -6.14945948e-01
-1.15452623e+00 8.74558449e-01 -1.82107762e-02 -4.43463057e-01
4.22809958e-01 -4.76119816e-01 -7.02822030e-01 2.01047480e-01
3.53349954e-01 8.02693725e-01 6.58705890e-01 -1.16329885e+00
-1.08804107e+00 2.39388123e-01 4.90304142e-01 3.40501934e-01
-7.63137758e-01 1.04154922e-01 -2.98149288e-01 -1.13311696e+00
-1.11113749e-01 -6.21370852e-01 1.04918398e-01 -3.72944593e-01
-5.33884943e-01 -2.88081437e-01 4.56773520e-01 -3.25128764e-01
2.06895685e+00 -1.71617424e+00 -1.10926792e-01 1.29898917e-02
4.17542458e-01 1.23595530e-02 -1.71304047e-01 9.25366342e-01
2.15345010e-01 3.78926307e-01 1.20843388e-01 1.41415313e-01
3.38883430e-01 -3.13438118e-01 -8.59870613e-01 -1.63558900e-01
4.51935269e-02 1.12465882e+00 -1.11072469e+00 -6.45353556e-01
-6.57330155e-02 2.43043184e-01 -3.49349320e-01 8.34245607e-02
-3.69294256e-01 -1.43580481e-01 -4.63163912e-01 7.61701822e-01
2.62472890e-02 -4.25171405e-01 -2.30544314e-01 -4.82781023e-01
-3.02557796e-01 7.26308763e-01 -1.10801125e+00 1.48579013e+00
-4.54988778e-01 5.20873785e-01 -3.64201516e-01 -2.40930915e-01
7.45048046e-01 2.16911286e-01 3.85635376e-01 -8.37895751e-01
-3.06816369e-01 4.33524668e-01 -3.17253292e-01 -1.48147181e-01
1.37958896e+00 2.96111882e-01 -3.17878187e-01 8.12534153e-01
1.07308686e-01 -4.04488891e-01 5.55200517e-01 5.61649799e-01
1.00601768e+00 8.17415267e-02 6.14297330e-01 -5.77149332e-01
4.16172564e-01 2.90479988e-01 -1.47490744e-02 1.00857687e+00
-1.33293178e-02 5.38876176e-01 3.97243351e-01 -3.28639030e-01
-9.47977185e-01 -9.43247020e-01 3.77693251e-02 1.42363989e+00
3.63977283e-01 -1.12310433e+00 -6.66496336e-01 -6.93762422e-01
2.43706182e-02 1.00970113e+00 -7.03712285e-01 -1.33321032e-01
-1.63936406e-01 -6.15156770e-01 8.26106071e-01 2.83119202e-01
1.71773568e-01 -8.82891953e-01 -9.99625444e-01 1.15987040e-01
-3.91068369e-01 -1.01655960e+00 -6.59983218e-01 2.72699278e-02
-4.96458262e-01 -9.10212696e-01 -8.89468551e-01 -4.12319958e-01
3.47925156e-01 4.60769206e-01 1.74474156e+00 1.62664160e-01
1.34688497e-01 7.35707879e-01 -6.24720335e-01 -5.73754311e-01
-4.22106713e-01 4.15535539e-01 -6.40971493e-03 -4.61711109e-01
5.68827033e-01 -1.94374561e-01 -6.84694588e-01 2.49352008e-01
-9.40083623e-01 -3.15192081e-02 7.68017173e-01 6.51977241e-01
4.65859562e-01 -7.16364533e-02 4.47991252e-01 -6.42966747e-01
1.41789210e+00 -1.59623280e-01 -2.03327030e-01 6.21521950e-01
-1.39911747e+00 3.41649413e-01 4.63620812e-01 -3.68241459e-01
-5.96877813e-01 -5.27162969e-01 1.93224803e-01 4.63168137e-02
3.60480435e-02 4.97180969e-01 2.57934451e-01 7.88492262e-02
1.06839263e+00 4.12295103e-01 -3.20225894e-01 -4.08974022e-01
7.35668123e-01 3.53436470e-01 4.22901243e-01 -8.88688087e-01
8.07121038e-01 3.15509439e-01 -2.83350974e-01 -7.17846930e-01
-9.80374038e-01 -4.46442157e-01 -4.22812015e-01 -4.37377170e-02
4.85372782e-01 -9.40477848e-01 -4.77894753e-01 7.87663683e-02
-1.00020707e+00 1.71591192e-01 -5.11880755e-01 4.08603430e-01
-3.66052687e-01 5.15473843e-01 -4.32216674e-01 -6.13248348e-01
-6.40813291e-01 -1.07980275e+00 1.18209016e+00 9.44168791e-02
-9.85222340e-01 -7.84088850e-01 -3.84600237e-02 -6.25206605e-02
6.16422594e-01 2.43026540e-01 1.06752896e+00 -9.09505367e-01
-3.91956300e-01 -3.24008584e-01 -2.31599659e-01 5.38957417e-02
9.45954174e-02 1.09551899e-01 -6.77519202e-01 -1.09877676e-01
-3.14064622e-01 -4.81347263e-01 9.43067074e-01 7.24428371e-02
7.18809009e-01 -4.68535036e-01 -1.39399335e-01 2.73333043e-01
1.23684835e+00 -4.07824107e-02 3.26712608e-01 7.51051128e-01
5.09581029e-01 8.65654826e-01 6.62838101e-01 2.94633090e-01
7.08587885e-01 6.56007946e-01 -1.40299322e-02 7.35028684e-02
-2.01833114e-01 -8.18131149e-01 5.06870806e-01 6.57231867e-01
2.67022364e-02 -1.72632307e-01 -7.96969354e-01 6.52358949e-01
-1.61212528e+00 -9.36349988e-01 1.81147352e-01 2.22396874e+00
1.15234911e+00 4.91503417e-01 3.68662268e-01 1.59004688e-01
2.76664883e-01 4.34525371e-01 -6.12668507e-02 -3.35061491e-01
-5.30781448e-01 2.91344821e-01 4.50494081e-01 3.66656214e-01
-8.46394122e-01 1.09104931e+00 7.69265509e+00 9.42175627e-01
-1.03387296e+00 -1.77628011e-01 4.41460103e-01 1.39497101e-01
-8.62287879e-01 1.28092036e-01 -1.02550662e+00 2.05473468e-01
7.39933074e-01 -6.45323277e-01 2.56177783e-01 7.64968514e-01
-1.16130687e-01 9.67658535e-02 -1.42584670e+00 7.83814013e-01
1.16378881e-01 -1.45173919e+00 6.08860672e-01 -5.71449101e-02
6.31170332e-01 -3.21980357e-01 2.36006930e-01 3.09828550e-01
4.43231940e-01 -1.03958797e+00 1.19391525e+00 4.55305129e-01
8.08252513e-01 -6.41164184e-01 3.83099049e-01 5.02206646e-02
-9.20466781e-01 1.95663825e-01 -4.31653894e-02 -2.40891557e-02
-1.47138968e-01 4.68721896e-01 -5.92821479e-01 4.42031145e-01
4.61981475e-01 3.54561687e-01 -8.55312109e-01 5.49838305e-01
-4.23906684e-01 3.68400037e-01 -1.07980110e-01 -4.38060939e-01
4.79240924e-01 1.35833889e-01 6.82128251e-01 1.84362984e+00
3.14623654e-01 -2.49458894e-01 1.78402811e-01 8.81518364e-01
6.86607230e-03 5.01665652e-01 -4.72883373e-01 -5.50779223e-01
7.97937334e-01 1.27151179e+00 -8.68130982e-01 -4.90553975e-01
-3.96142155e-01 8.12749684e-01 -1.58382550e-01 3.71338546e-01
-4.18417275e-01 -6.44768596e-01 5.83277404e-01 1.93917602e-01
-1.55563012e-03 -2.24189430e-01 -4.56724316e-01 -1.11115789e+00
6.96452036e-02 -1.34056234e+00 5.69725394e-01 -6.49103761e-01
-1.15940356e+00 8.99969280e-01 2.77557641e-01 -1.24844861e+00
-5.29078126e-01 -4.51861084e-01 -2.43229643e-01 8.44637752e-01
-1.46937406e+00 -1.24775684e+00 -2.20617920e-01 3.18586469e-01
4.01150554e-01 -2.46316478e-01 1.07211971e+00 5.07904179e-02
-5.78924343e-02 7.08876908e-01 -2.98440635e-01 -6.89097494e-02
9.29162681e-01 -1.46339536e+00 6.88248813e-01 9.71625984e-01
4.71657455e-01 9.93264616e-01 9.76137161e-01 -5.47499835e-01
-1.28790390e+00 -6.19776070e-01 1.31691635e+00 -8.97439241e-01
8.45819116e-01 -9.20710713e-02 -5.99664152e-01 1.88329875e-01
3.64140749e-01 -4.83676106e-01 6.81475580e-01 2.37985566e-01
-8.49758863e-01 6.13789745e-02 -7.17325568e-01 9.51263428e-01
9.37833011e-01 -8.18187296e-01 -9.39102054e-01 1.04081020e-01
9.91451800e-01 -2.98792988e-01 -7.89505124e-01 5.13655663e-01
7.96550989e-01 -9.24599290e-01 1.11567652e+00 -6.04464233e-01
5.67383707e-01 -3.58440936e-01 -4.65712517e-01 -1.14908278e+00
-4.20679480e-01 -8.61661911e-01 -4.48393434e-01 1.24461222e+00
6.30802870e-01 -4.41800922e-01 3.27521443e-01 5.01611531e-01
1.89004913e-01 -7.61982143e-01 -2.96421349e-01 -7.52007067e-01
4.24278527e-02 -3.53430122e-01 7.66086757e-01 1.06466079e+00
1.15093999e-01 7.16658890e-01 -2.46609643e-01 -3.73811424e-01
3.12932700e-01 3.22563142e-01 6.56775057e-01 -1.10103381e+00
-1.67389840e-01 -1.05945539e+00 -9.43551958e-02 -1.08160305e+00
-1.14355423e-01 -8.00330639e-01 -4.31709468e-01 -1.58890975e+00
2.07625955e-01 -2.06673339e-01 -6.07682049e-01 3.51914793e-01
-4.94132459e-01 1.04385406e-01 2.32161954e-01 5.31493783e-01
-7.87699521e-01 2.15047687e-01 9.99248743e-01 -8.50475878e-02
-2.56386071e-01 -2.03364849e-01 -1.49200904e+00 6.07421517e-01
4.37933594e-01 -1.87339008e-01 -6.06460333e-01 -3.03543776e-01
7.01157510e-01 -3.33720744e-01 2.95944810e-01 -9.37878668e-01
2.21079558e-01 -4.46857691e-01 2.18242049e-01 -5.39618552e-01
5.63235022e-03 -5.67986906e-01 -1.17293678e-01 3.44421446e-01
-8.09156299e-01 5.19533873e-01 1.01132132e-01 2.37176180e-01
-2.92188376e-01 -1.40612975e-01 4.56320882e-01 -2.10343972e-01
-1.00802827e+00 7.93566331e-02 -2.58640826e-01 5.25067329e-01
4.32370812e-01 -2.31828302e-01 -3.36957902e-01 -4.69204932e-01
-1.95857827e-02 1.03247697e-02 6.59979165e-01 8.87548447e-01
4.32785541e-01 -1.36408305e+00 -9.54452813e-01 -2.45249614e-01
7.27188110e-01 -6.81222081e-01 -4.51898605e-01 5.77403903e-01
-3.55577618e-01 7.88716435e-01 -6.52704090e-02 -2.77409762e-01
-8.41254354e-01 4.35589373e-01 -6.33406825e-03 -4.58557039e-01
-4.55878526e-01 6.89220011e-01 2.52857935e-02 -2.21325886e-02
3.83873731e-01 -5.36201119e-01 -4.39167053e-01 5.67427874e-01
7.20824838e-01 3.27715695e-01 3.04891676e-01 -7.02302635e-01
-2.17635244e-01 3.53532940e-01 -2.97331601e-01 -4.59729820e-01
9.70736444e-01 -8.00569728e-02 -5.43444715e-02 5.37708104e-01
8.95984530e-01 3.40125054e-01 -6.35351896e-01 -4.47669804e-01
5.65291226e-01 -3.02302897e-01 9.30618644e-02 -1.13426912e+00
-3.26128393e-01 4.95591670e-01 2.99320638e-01 4.44734097e-01
1.10025001e+00 -1.63961232e-01 8.67376566e-01 5.40805876e-01
3.66797060e-01 -1.41306460e+00 1.92641884e-01 5.96942961e-01
9.65941846e-01 -9.32223380e-01 2.05363199e-01 -5.46392389e-02
-7.82689869e-01 1.15081549e+00 4.23593134e-01 3.16366106e-01
3.94894600e-01 1.05862722e-01 1.26039043e-01 -3.50100100e-01
-8.71287346e-01 -3.30343783e-01 8.44521403e-01 3.06180239e-01
8.81848395e-01 3.43900360e-02 -7.53366530e-01 6.57584786e-01
-7.25366950e-01 -2.12412208e-01 3.65840167e-01 7.93118834e-01
-5.73925912e-01 -1.00157750e+00 -2.08459407e-01 5.11661410e-01
-7.73620129e-01 -4.04784620e-01 -9.39516544e-01 8.58627617e-01
-2.01821715e-01 1.03694773e+00 -3.28560531e-01 -6.40646040e-01
5.19513309e-01 3.34189869e-02 3.31718832e-01 -6.61140621e-01
-7.92558551e-01 -1.31025255e-01 2.70544171e-01 -5.34136772e-01
-3.47950280e-01 -3.16725343e-01 -7.52780199e-01 -2.79473752e-01
-1.93905786e-01 5.12442887e-01 4.86658603e-01 6.91751540e-01
5.15240014e-01 2.68837839e-01 4.38170373e-01 -3.24217647e-01
-7.98383892e-01 -8.80644739e-01 -1.62207514e-01 5.81837952e-01
3.87116432e-01 -4.38340604e-01 -3.05338413e-01 5.24331070e-02] | [12.299357414245605, 8.899238586425781] |
4a8a83e3-a785-4ae0-b2b2-9cb11078841b | representations-of-time-expressions-for | null | null | https://aclanthology.org/W17-2341 | https://aclanthology.org/W17-2341.pdf | Representations of Time Expressions for Temporal Relation Extraction with Convolutional Neural Networks | Token sequences are often used as the input for Convolutional Neural Networks (CNNs) in natural language processing. However, they might not be an ideal representation for time expressions, which are long, highly varied, and semantically complex. We describe a method for representing time expressions with single pseudo-tokens for CNNs. With this method, we establish a new state-of-the-art result for a clinical temporal relation extraction task. | ['Guergana Savova', 'Dmitriy Dligach', 'Chen Lin', 'Timothy Miller', 'Steven Bethard'] | 2017-08-01 | null | null | null | ws-2017-8 | ['temporal-relation-extraction'] | ['natural-language-processing'] | [ 8.11939314e-03 -1.13695621e-01 -7.36727178e-01 -5.05815506e-01
-2.23372236e-01 -4.10598576e-01 6.42210066e-01 5.41947901e-01
-8.64706933e-01 7.11824596e-01 3.01709563e-01 -4.39674288e-01
2.34793052e-01 -8.75951231e-01 -2.58502722e-01 -4.64915007e-01
-5.64302862e-01 1.89755812e-01 -7.09799826e-02 -4.10710216e-01
-3.70218456e-01 6.87593877e-01 -7.55874217e-01 5.31607091e-01
3.81458879e-01 1.06273890e+00 -6.66396856e-01 4.13394898e-01
-5.64016700e-01 1.12355137e+00 -8.78373742e-01 -3.51047039e-01
-4.80288148e-01 -4.40756798e-01 -1.13632667e+00 -6.13242209e-01
-4.51572448e-01 3.02135292e-02 -7.39448786e-01 8.18876684e-01
3.61158729e-01 2.22027645e-01 4.64725792e-01 -1.01475537e+00
-4.39374477e-01 9.12262976e-01 -1.67150930e-01 5.58188856e-01
4.35794473e-01 9.53048095e-03 7.73967564e-01 -9.05456021e-02
1.03256714e+00 1.05040443e+00 8.52573991e-01 6.11893535e-01
-9.41038489e-01 -5.58343530e-01 1.15854308e-01 3.27128023e-01
-1.15333498e+00 -1.65940329e-01 8.15031469e-01 -2.54372567e-01
1.59069395e+00 1.45753175e-01 1.15592229e+00 1.60022497e+00
5.26807964e-01 1.05805397e+00 8.34361672e-01 -2.56172687e-01
1.95874691e-01 -4.16954547e-01 1.27765372e-01 6.90881193e-01
-5.20469069e-01 2.59060830e-01 -5.11782527e-01 -1.98842064e-01
7.29683518e-01 1.17582925e-01 -4.65109013e-02 5.92626035e-01
-1.70071483e+00 8.19678724e-01 7.71228611e-01 8.85229468e-01
-6.17822468e-01 6.76237762e-01 1.24320829e+00 3.17943752e-01
8.00103903e-01 4.39190477e-01 -6.63589537e-01 -7.66882658e-01
-9.32022274e-01 3.65079761e-01 5.49111187e-01 8.47576320e-01
-1.00936912e-01 8.95116776e-02 -7.81637967e-01 6.99043274e-01
-4.05683696e-01 -1.06159776e-01 8.07183027e-01 -3.05656552e-01
1.85442474e-02 4.91666824e-01 -1.87940463e-01 -7.01909304e-01
-9.77286398e-01 -2.76457071e-01 -1.22946739e+00 -4.56607550e-01
4.17672336e-01 -7.96055645e-02 -1.13029337e+00 2.03057003e+00
5.98750124e-03 1.99703604e-01 3.33541781e-02 4.57641095e-01
1.20622170e+00 5.48355639e-01 6.43812895e-01 -5.74712038e-01
1.85367966e+00 -6.66870356e-01 -1.52206600e+00 -2.69595403e-02
1.13476539e+00 -3.32437694e-01 4.43620235e-01 1.42422214e-01
-1.04463530e+00 -1.68800533e-01 -6.83568835e-01 -2.58265853e-01
-9.73828256e-01 -7.73490593e-02 1.35839880e+00 5.06619848e-02
-7.16184378e-01 9.19167936e-01 -1.17560852e+00 -2.73914844e-01
8.00866187e-01 2.10483745e-01 -3.89200985e-01 3.09235781e-01
-1.93628550e+00 1.28411794e+00 5.60074627e-01 2.76866227e-01
-7.00113773e-01 -7.81983733e-01 -1.10273659e+00 -1.27391279e-01
-2.56764621e-01 -6.64018035e-01 1.71910894e+00 -9.64893937e-01
-1.26708984e+00 1.11744213e+00 -2.43880734e-01 -7.56309867e-01
5.17336130e-01 1.46204934e-01 -8.70678008e-01 1.11864507e-03
-1.09273501e-01 5.71032345e-01 3.11019391e-01 -1.10460952e-01
-9.36782360e-02 -9.72015113e-02 -8.49535502e-03 -4.22178537e-01
4.09463383e-02 4.59135503e-01 -2.51617342e-01 -1.13835537e+00
1.87956709e-02 -7.70357370e-01 -7.39740610e-01 2.09242254e-01
-5.70782304e-01 -7.49401152e-01 4.67529118e-01 -4.41267699e-01
1.55477369e+00 -2.30325365e+00 -2.26968378e-01 -1.84579849e-01
4.17270839e-01 1.86980948e-01 1.45688638e-01 2.80273706e-01
-6.66265070e-01 3.38565648e-01 -1.49220556e-01 -2.40529105e-01
-1.06169529e-01 4.39369589e-01 -1.05342381e-01 5.30187666e-01
5.19310772e-01 1.44969225e+00 -1.51026583e+00 -6.83518171e-01
-6.56780321e-03 4.65559840e-01 -9.76408087e-03 1.42033184e-02
-4.48671073e-01 4.75004703e-01 -6.91649497e-01 7.40074098e-01
2.27761447e-01 -1.97215497e-01 -9.31833237e-02 -2.60939240e-01
-1.26524782e-02 6.00682855e-01 -1.56476751e-01 2.03243279e+00
-4.31253612e-01 8.19028676e-01 -8.19409728e-01 -1.12075639e+00
5.87052584e-01 8.65441382e-01 1.11433613e+00 -9.75735903e-01
4.47920531e-01 1.37358442e-01 6.22052401e-02 -6.92296743e-01
2.36715943e-01 -7.11161315e-01 -4.51081932e-01 2.08471954e-01
7.42592290e-02 -1.71064422e-01 2.97916412e-01 -8.41285512e-02
1.31761420e+00 -5.50647750e-02 7.60705352e-01 6.41279854e-03
7.97836483e-02 8.73615444e-02 8.83081615e-01 2.80526757e-01
-5.06458998e-01 3.16944003e-01 8.98448586e-01 -1.21723163e+00
-8.59468102e-01 -7.62587786e-01 -3.35400283e-01 6.68130755e-01
-5.44623315e-01 -6.69039488e-01 -3.41823936e-01 -7.41360068e-01
-2.41778463e-01 4.61179823e-01 -9.95930433e-01 -3.74952763e-01
-7.14371085e-01 -6.94566965e-01 1.28754687e+00 9.85858560e-01
1.34435073e-01 -1.37663925e+00 -9.10148978e-01 7.96205461e-01
-4.60513592e-01 -1.53612638e+00 -3.90909165e-01 6.42349601e-01
-1.01704776e+00 -8.51390839e-01 -1.02315128e+00 -7.51429319e-01
4.43884939e-01 -4.11083341e-01 1.40714800e+00 -1.05414562e-01
-4.87005711e-01 -3.04300189e-01 -3.74655694e-01 -8.77516150e-01
-1.60797223e-01 -3.96569772e-03 -8.04557465e-03 -4.40178514e-01
6.85846686e-01 -5.73628783e-01 -6.72631145e-01 -1.80205926e-01
-1.02518570e+00 1.00947879e-01 1.22126669e-01 8.37197542e-01
6.51868999e-01 -6.49891570e-02 3.64903897e-01 -9.63056028e-01
6.59581304e-01 -4.28744107e-01 -9.49365646e-02 9.50949416e-02
-3.29442769e-01 1.97954744e-01 8.68693709e-01 -9.99388278e-01
-5.52477598e-01 -1.93069745e-02 -3.11498791e-01 -3.91322345e-01
-9.37649533e-02 9.93434012e-01 6.66168213e-01 1.54114529e-01
8.02339733e-01 3.41783203e-02 -2.38168716e-01 -1.44932091e-01
5.59685588e-01 2.05928981e-01 6.14257455e-01 -6.56059444e-01
1.30801573e-01 6.64825261e-01 2.20242277e-01 -5.17471910e-01
-9.46670353e-01 -3.74125034e-01 -6.37105405e-01 -1.84396580e-02
1.11795914e+00 -7.11402953e-01 -7.12680221e-01 5.60524881e-01
-1.61771119e+00 -3.14683527e-01 -4.42001790e-01 6.64276361e-01
-6.67347491e-01 2.18530521e-02 -1.01481235e+00 -3.75758529e-01
-5.55379450e-01 -9.30598080e-01 9.42181528e-01 2.26246446e-01
-8.06316912e-01 -1.23413563e+00 1.78773984e-01 -5.43209970e-01
6.39952600e-01 1.14915812e+00 1.15098941e+00 -6.79376543e-01
1.45777166e-01 -3.72113526e-01 -9.36384201e-02 -1.95904970e-01
3.21068227e-01 2.33373642e-01 -9.79640961e-01 2.66000658e-01
-2.29283050e-01 -2.60976523e-01 7.47260332e-01 6.71907842e-01
1.79954803e+00 -3.73912901e-01 -6.80816650e-01 8.19038153e-01
1.04774344e+00 4.41425920e-01 8.71968687e-01 2.37943411e-01
4.15847689e-01 3.75750810e-01 4.59779054e-01 7.89293826e-01
2.61016250e-01 5.72737932e-01 1.42532334e-01 -4.39979523e-01
2.35040888e-01 3.22380155e-01 1.21156998e-01 7.74307787e-01
-2.65209287e-01 -1.56006485e-03 -1.18286371e+00 9.77977335e-01
-1.83352995e+00 -1.01163232e+00 -2.16511860e-01 1.46185613e+00
1.46182144e+00 3.26254994e-01 -4.88086902e-02 1.32210314e-01
1.80721745e-01 4.23809886e-01 -4.37992960e-01 -7.89773583e-01
-1.89493075e-01 8.15914512e-01 3.92083913e-01 -2.89009005e-01
-1.34987175e+00 1.12205839e+00 6.97796822e+00 7.34002173e-01
-1.48581123e+00 1.49606183e-01 6.49798036e-01 3.18798311e-02
-1.13637045e-01 -3.15858155e-01 6.68791728e-03 2.11001441e-01
1.55041051e+00 -4.26716208e-01 7.22754793e-03 5.85956633e-01
4.80407834e-01 2.52867907e-01 -1.45337188e+00 1.28496563e+00
-6.24899864e-01 -1.54072607e+00 -2.33880505e-01 -4.14188355e-01
5.44683158e-01 7.13170245e-02 -2.81057078e-02 3.09408695e-01
2.26663500e-01 -1.57459068e+00 7.07826555e-01 4.50821280e-01
1.01429021e+00 -5.62862158e-01 7.86879539e-01 -2.08423853e-01
-1.45076013e+00 3.51097316e-01 7.08851889e-02 6.03473634e-02
4.20883536e-01 7.97091842e-01 -6.27856255e-01 5.99331260e-01
5.74734092e-01 1.07943428e+00 -2.00682864e-01 9.48721886e-01
-3.88418376e-01 4.56799865e-01 -3.61591130e-01 -3.41819316e-01
7.71386862e-01 5.25381029e-01 1.35441050e-01 1.88343811e+00
3.68475169e-02 3.80366772e-01 -4.24003340e-02 9.10048127e-01
-2.79080033e-01 -7.34987929e-02 -9.77297723e-01 -8.06891024e-01
8.36505368e-02 1.05671000e+00 -9.14102554e-01 -5.16256452e-01
-1.67900026e-01 8.51653039e-01 1.11410730e-01 3.69963735e-01
-1.20578229e+00 -5.91839850e-01 8.05385768e-01 -3.44449908e-01
-2.66851425e-01 -4.53384042e-01 -2.35283196e-01 -1.04449081e+00
5.61023457e-03 -5.67640185e-01 7.76354551e-01 -7.88350165e-01
-1.32051075e+00 7.81765699e-01 1.33220971e-01 -1.60891795e+00
-4.74184543e-01 -7.53722131e-01 -5.83902478e-01 7.82716036e-01
-1.49554706e+00 -1.09064531e+00 -1.16516210e-01 7.92339861e-01
3.20557952e-01 1.97772160e-01 1.34336865e+00 4.00421441e-01
-2.31560320e-01 4.71872598e-01 -3.81349742e-01 6.90041125e-01
5.23285151e-01 -1.19166529e+00 5.40451765e-01 4.25494105e-01
2.57140715e-02 8.66459429e-01 5.05512297e-01 -2.11483344e-01
-1.22620583e+00 -1.17530572e+00 1.30566561e+00 -1.94573104e-01
8.49310875e-01 -1.85005683e-02 -7.73684084e-01 7.69464493e-01
1.50430679e-01 3.99113536e-01 9.00274634e-01 2.55932629e-01
-5.45219362e-01 1.36193782e-01 -1.08608818e+00 6.70901239e-01
9.16128099e-01 -9.77140069e-01 -5.33468187e-01 8.55085313e-01
9.91851449e-01 -9.23042238e-01 -1.36873150e+00 4.09823865e-01
5.14769733e-01 -4.89649661e-02 1.00075614e+00 -1.17887878e+00
7.68034697e-01 5.65168113e-02 2.84270763e-01 -1.28235924e+00
-9.82879177e-02 -8.28323007e-01 -2.66044021e-01 6.55856371e-01
3.92820865e-01 -3.95169944e-01 4.57680255e-01 3.68043303e-01
-1.93291232e-01 -1.11801720e+00 -1.24777448e+00 -8.85497868e-01
2.54980296e-01 -8.25615346e-01 6.19745076e-01 1.55143464e+00
6.05372846e-01 8.29710886e-02 2.06169132e-02 -4.27495658e-01
-4.56066877e-02 -1.15277722e-01 -2.04949364e-01 -8.90073180e-01
2.70792216e-01 -1.00337100e+00 -6.96519315e-01 -3.67614388e-01
4.27927613e-01 -1.01559782e+00 5.15081547e-02 -1.53767812e+00
-1.20109841e-01 -3.07180107e-01 -6.88289583e-01 1.01197362e+00
2.13683352e-01 -1.61460742e-01 -1.97056755e-01 -2.76045054e-01
-3.51599216e-01 4.59707618e-01 1.43734753e+00 -6.55781448e-01
-1.60506979e-01 -4.81553227e-02 -2.38884792e-01 4.44395155e-01
7.38486469e-01 -7.08546996e-01 -2.02820525e-01 -3.14007223e-01
3.58507901e-01 4.01937574e-01 1.77686438e-01 -5.59120119e-01
2.34984398e-01 -4.82697845e-01 3.74122262e-01 -5.90176940e-01
1.85580984e-01 -4.99823123e-01 2.41021633e-01 7.97976077e-01
-5.69766104e-01 5.20564914e-01 6.59192264e-01 1.09730534e-01
-6.63470030e-01 7.10317940e-02 8.31717432e-01 -5.77012777e-01
-5.40723026e-01 8.15771520e-01 -4.50806439e-01 1.51750389e-02
7.30691195e-01 2.36426413e-01 -3.53240520e-01 -1.84515759e-01
-8.91447067e-01 1.13275453e-01 -2.53590405e-01 6.04263127e-01
7.65959680e-01 -1.69494390e+00 -5.32140076e-01 -4.02214527e-01
3.64567280e-01 2.16873661e-01 1.67761475e-01 9.94563460e-01
-7.00963974e-01 6.57882750e-01 -2.64846832e-01 -4.93411541e-01
-9.52581763e-01 8.61585617e-01 5.91965139e-01 -6.35061204e-01
-9.52867508e-01 7.95672178e-01 -2.17956364e-01 -7.14190155e-02
2.34497845e-01 -1.12728846e+00 -4.56492424e-01 3.18223357e-01
6.25523925e-01 -5.80751419e-01 1.09614164e-01 -3.31358522e-01
-7.08991766e-01 2.85774589e-01 -2.73282584e-02 -1.08741693e-01
1.39947057e+00 6.70733869e-01 -4.44842339e-01 8.47184122e-01
1.51582837e+00 -8.58021140e-01 -1.98298305e-01 -3.97271365e-01
1.98659703e-01 1.37319565e-01 -9.89098772e-02 -5.55925488e-01
-1.18007946e+00 9.34451699e-01 4.38675702e-01 1.86573461e-01
1.23550713e+00 6.86722100e-02 1.04983997e+00 4.18506175e-01
3.05869222e-01 -9.54724252e-01 3.21527533e-02 9.19209123e-01
6.74031019e-01 -9.34420526e-01 -1.10510781e-01 -1.06616788e-01
-3.87645841e-01 1.64939260e+00 2.76691049e-01 2.38852501e-01
9.80208635e-01 3.22351545e-01 2.70301942e-02 -6.03091836e-01
-9.18106139e-01 -3.74242008e-01 2.25962549e-01 5.16712725e-01
1.08341980e+00 4.42162067e-01 -6.30301118e-01 7.18278170e-01
-2.50900447e-01 3.72425944e-01 3.44755679e-01 1.11061168e+00
4.40230012e-01 -9.38544691e-01 2.56959230e-01 3.17731023e-01
-1.00493252e+00 -3.01866442e-01 -1.05447270e-01 6.40814841e-01
-5.75710554e-03 5.69629312e-01 2.63010487e-02 -2.53988922e-01
4.02850419e-01 1.87813759e-01 5.87522328e-01 -4.55980986e-01
-9.22335625e-01 -1.92475855e-01 4.15555775e-01 -8.92213583e-01
-7.87382126e-01 -2.98323572e-01 -1.78895557e+00 -2.80928254e-01
1.69845775e-01 -1.18446089e-01 6.60499394e-01 1.05537486e+00
-1.18791964e-02 1.18769276e+00 3.32224369e-01 -2.50383794e-01
4.36583608e-02 -9.12084222e-01 -2.96124458e-01 7.25909233e-01
5.63828111e-01 -5.49356699e-01 1.63161173e-01 3.40620838e-02] | [8.549769401550293, 9.017213821411133] |
bca9f8f3-761f-4712-8256-fafd2d07ae69 | a-hybrid-mesh-neural-representation-for-3d | 2203.12613 | null | https://arxiv.org/abs/2203.12613v3 | https://arxiv.org/pdf/2203.12613v3.pdf | Hybrid Mesh-neural Representation for 3D Transparent Object Reconstruction | We propose a novel method to reconstruct the 3D shapes of transparent objects using hand-held captured images under natural light conditions. It combines the advantage of explicit mesh and multi-layer perceptron (MLP) network, a hybrid representation, to simplify the capture setting used in recent contributions. After obtaining an initial shape through the multi-view silhouettes, we introduce surface-based local MLPs to encode the vertex displacement field (VDF) for the reconstruction of surface details. The design of local MLPs allows to represent the VDF in a piece-wise manner using two layer MLP networks, which is beneficial to the optimization algorithm. Defining local MLPs on the surface instead of the volume also reduces the searching space. Such a hybrid representation enables us to relax the ray-pixel correspondences that represent the light path constraint to our designed ray-cell correspondences, which significantly simplifies the implementation of single-image based environment matting algorithm. We evaluate our representation and reconstruction algorithm on several transparent objects with ground truth models. Our experiments show that our method can produce high-quality reconstruction results superior to state-of-the-art methods using a simplified data acquisition setup. | ['Weiwei Xu', 'Hujun Bao', 'Zihan Zhu', 'Jiamin Xu'] | 2022-03-23 | null | null | null | null | ['transparent-objects', 'image-matting', 'object-reconstruction'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 3.37383479e-01 1.03010975e-01 1.96772963e-01 -3.26262265e-01
-4.04239208e-01 -1.45255789e-01 3.63236576e-01 5.77821285e-02
-4.38423246e-01 5.45884192e-01 -1.84790432e-01 1.11483604e-01
-2.93220598e-02 -1.15444052e+00 -1.11932695e+00 -7.66504288e-01
1.70534998e-01 7.65994728e-01 4.31222856e-01 1.37246266e-01
3.10823292e-01 1.18808234e+00 -1.71091664e+00 4.47337508e-01
4.03602749e-01 1.20499361e+00 4.85146970e-01 4.70402896e-01
-5.18224239e-01 4.74682689e-01 -9.80235711e-02 -3.21538985e-01
4.83787894e-01 2.44665578e-01 -4.66920167e-01 1.95342764e-01
7.47718573e-01 -5.24786413e-01 -2.96047945e-02 7.80369997e-01
4.94552195e-01 -1.32896200e-01 6.78400695e-01 -5.01252949e-01
-4.76776630e-01 5.30660003e-02 -6.47639096e-01 -6.51091278e-01
3.50412637e-01 3.25825438e-02 8.26901555e-01 -8.91512334e-01
8.24354053e-01 1.14008033e+00 9.16174591e-01 3.76835257e-01
-1.44833434e+00 -2.49309484e-02 -4.25653830e-02 -3.67932059e-02
-1.27745748e+00 -3.64325851e-01 1.23626614e+00 -5.69329321e-01
8.94137204e-01 5.04047811e-01 8.55438769e-01 6.22223198e-01
4.16632593e-01 2.88263321e-01 1.03132057e+00 -6.79643095e-01
1.55658796e-01 1.43102020e-01 -7.87631497e-02 1.05835187e+00
-4.29456271e-02 -1.69017620e-03 -2.87902087e-01 -3.38956058e-01
1.61355340e+00 1.27980277e-01 -5.35548151e-01 -7.26487100e-01
-1.29400933e+00 5.07543802e-01 5.61349332e-01 -5.37153110e-02
-4.72916037e-01 3.30939293e-01 -5.01757674e-02 4.47877906e-02
4.66161519e-01 3.74727130e-01 -3.13747466e-01 3.37906271e-01
-6.76333308e-01 2.57000718e-02 7.62943387e-01 6.34491503e-01
1.23934305e+00 1.47214932e-02 -1.39206916e-01 9.81594384e-01
6.38320982e-01 6.53287828e-01 3.73875909e-02 -1.25988781e+00
2.66367137e-01 7.39446163e-01 3.63174796e-01 -9.04600620e-01
-5.52826166e-01 -4.29882437e-01 -7.90950239e-01 6.58855438e-01
3.77017289e-01 2.69552261e-01 -8.44829142e-01 1.12412620e+00
5.63921392e-01 4.50858660e-02 -2.62246639e-01 8.94835293e-01
9.41230476e-01 6.04777575e-01 -8.38469207e-01 -2.90495843e-01
1.22762966e+00 -8.72231424e-01 -4.32827443e-01 3.22366729e-02
4.94404137e-02 -7.84972012e-01 1.15908873e+00 4.43988234e-01
-1.37420964e+00 -4.98268872e-01 -8.35135579e-01 -1.77651197e-01
1.09357350e-01 3.79534870e-01 4.80695546e-01 2.96735883e-01
-1.07473361e+00 7.32817590e-01 -8.30031872e-01 1.42763957e-01
2.55899400e-01 5.09044826e-01 -1.46834716e-01 1.56601757e-01
-4.48043704e-01 5.77856481e-01 -1.20275185e-01 2.42010996e-01
-5.88559151e-01 -8.30873668e-01 -6.78527296e-01 -5.75667135e-02
2.70959288e-01 -1.00578094e+00 9.18803751e-01 -7.02051878e-01
-2.16616988e+00 1.14801264e+00 -3.18492472e-01 3.61904241e-02
6.78531706e-01 -4.34385464e-02 1.51486501e-01 3.21194768e-01
-3.08657199e-01 4.59029466e-01 8.42840850e-01 -1.86428225e+00
-4.89748679e-02 -2.60519356e-01 1.50540814e-01 3.70719135e-02
1.17861561e-03 -3.33180219e-01 -6.90977454e-01 -4.03078020e-01
5.53553104e-01 -6.60789967e-01 -3.42250884e-01 7.61564016e-01
-4.33109760e-01 2.08448887e-01 5.67534387e-01 -4.59745705e-01
6.28059030e-01 -1.92387664e+00 2.26264924e-01 3.81974965e-01
2.41247877e-01 -1.45184934e-01 8.44731256e-02 4.20430601e-01
1.49347872e-01 -3.85486305e-01 -5.24940372e-01 -9.66409385e-01
-3.51024359e-01 3.34779501e-01 -1.86907813e-01 6.86805129e-01
-1.87508374e-01 7.91500509e-01 -4.00942475e-01 -5.52594960e-01
4.61427867e-01 7.71236002e-01 -6.78056538e-01 3.55432183e-01
-6.31660521e-01 6.00564539e-01 -4.29207116e-01 6.01534665e-01
8.30829918e-01 -2.69040376e-01 -1.06738411e-01 -4.59711999e-01
-4.78685081e-01 1.11252971e-01 -1.36864448e+00 1.98380339e+00
-7.41288960e-01 3.04693311e-01 4.77577448e-01 -5.67950010e-01
1.03888190e+00 2.69083053e-01 5.42131841e-01 -4.99794483e-01
1.01129249e-01 1.88161060e-01 -5.74614823e-01 -5.11466503e-01
1.93697989e-01 -3.20548028e-01 5.36356688e-01 4.51340765e-01
-3.44965100e-01 -4.29906309e-01 -4.88721788e-01 -4.78150010e-01
6.94675684e-01 5.62192380e-01 -1.17544234e-01 -1.61542743e-01
6.34102762e-01 -2.35407859e-01 4.92671698e-01 5.52852929e-01
5.48295736e-01 1.09814775e+00 1.22517750e-01 -7.68792033e-01
-1.04550290e+00 -1.15852988e+00 -4.11309838e-01 4.86497790e-01
3.63571763e-01 -6.99504316e-02 -5.46561837e-01 -2.47908551e-02
2.24614851e-02 3.30694318e-01 -3.22479665e-01 5.19666374e-01
-1.03229821e+00 -5.27387738e-01 -1.21836111e-01 2.93177634e-01
5.21145225e-01 -1.06270981e+00 -1.01328111e+00 2.08136111e-01
3.92976357e-03 -1.09269679e+00 4.21319306e-02 8.47183540e-02
-1.14542961e+00 -8.39056671e-01 -8.44381273e-01 -7.11654782e-01
9.60580707e-01 2.45824549e-02 1.11276340e+00 5.01583442e-02
-1.49755210e-01 4.83726174e-01 1.31810799e-01 -1.97275400e-01
-4.76692051e-01 -3.94341260e-01 -1.48810759e-01 1.97209015e-01
-4.64033991e-01 -1.08485579e+00 -5.89396119e-01 2.87783802e-01
-6.94197774e-01 5.84601223e-01 3.97333920e-01 4.70732182e-01
1.21334279e+00 -2.45240569e-01 -1.86733544e-01 -7.33263016e-01
3.41582373e-02 1.08314820e-01 -1.11353314e+00 2.33298764e-01
-1.09799244e-01 2.22314626e-01 7.54694939e-01 -3.63983274e-01
-1.00796032e+00 2.16582015e-01 -3.83934706e-01 -5.90348780e-01
-2.71439254e-01 1.21875100e-01 -7.08034188e-02 -5.40917099e-01
4.62226719e-01 3.42914462e-01 5.28011806e-02 -9.01267588e-01
1.45316973e-01 3.84318113e-01 3.20958704e-01 -6.67020798e-01
7.30359912e-01 1.01054573e+00 2.93162704e-01 -1.21106088e+00
-4.34505850e-01 -1.90351918e-01 -8.37920845e-01 -3.21952671e-01
8.71400476e-01 -6.48233771e-01 -1.03428423e+00 5.09121656e-01
-1.67000818e+00 -4.47822660e-01 -4.59182948e-01 5.21924496e-01
-7.07682610e-01 4.64033812e-01 -8.10985088e-01 -9.00387585e-01
-3.51214528e-01 -1.36086857e+00 1.43827248e+00 -1.54300734e-01
2.09323049e-01 -8.30266714e-01 1.94374874e-01 3.62232715e-01
2.82358527e-01 4.89494085e-01 1.03761995e+00 6.38846934e-01
-1.32931840e+00 1.56382069e-01 -8.58063847e-02 2.90366024e-01
4.56151180e-02 1.25629783e-01 -1.27624798e+00 -1.16536498e-01
2.46288612e-01 2.20319629e-02 6.69433832e-01 6.62358165e-01
1.43391395e+00 -1.96921021e-01 -3.46060842e-01 1.05094802e+00
1.87891352e+00 -4.21397528e-03 5.93040168e-01 3.38441312e-01
1.11901975e+00 5.98331273e-01 3.34500298e-02 4.45639551e-01
4.23102379e-01 1.08075321e+00 7.47744560e-01 -2.55747944e-01
-2.70489782e-01 5.37452623e-02 1.54248074e-01 8.58350277e-01
-4.67965424e-01 2.97049172e-02 -6.20698869e-01 1.16645440e-01
-1.59881413e+00 -5.71445465e-01 -4.42861676e-01 2.37848759e+00
7.92695045e-01 5.09590805e-02 -1.79863036e-01 -7.40249604e-02
2.75058359e-01 8.18138644e-02 -3.14359367e-01 -4.43840593e-01
-3.60530019e-02 2.41006866e-01 5.82924187e-01 9.57785368e-01
-6.27889395e-01 5.97435951e-01 6.10823679e+00 6.24617934e-01
-1.50069344e+00 1.08445518e-01 1.59754813e-01 -8.96882489e-02
-8.31787109e-01 -1.83529630e-01 -6.96430922e-01 2.69868881e-01
4.22754318e-01 6.92265689e-01 5.28478444e-01 5.42499840e-01
2.66869187e-01 -5.62150329e-02 -1.14191079e+00 1.26232076e+00
-1.13707535e-01 -1.64319086e+00 5.89715093e-02 -1.40697565e-02
5.87234080e-01 3.75386208e-01 -1.87383756e-01 -6.00241065e-01
-2.19324604e-01 -7.78445601e-01 1.05443442e+00 9.02945280e-01
9.70053792e-01 -2.66031057e-01 3.04023594e-01 5.73541343e-01
-1.05566859e+00 2.43647829e-01 -5.23698747e-01 1.17671397e-02
4.30172145e-01 1.07039487e+00 -6.99050188e-01 6.21946394e-01
6.77510500e-01 3.22958767e-01 -8.84342864e-02 1.00610054e+00
-3.37863602e-02 2.17915148e-01 -6.95224047e-01 1.48769006e-01
-3.95067781e-02 -5.37279248e-01 7.87709415e-01 8.80307555e-01
9.30276364e-02 -1.71514913e-01 1.10513896e-01 1.36437845e+00
2.33152881e-02 3.51495109e-02 -3.36350083e-01 5.73291004e-01
7.85882249e-02 1.11700571e+00 -8.05451453e-01 1.26686469e-01
-3.09070170e-01 8.29283357e-01 4.90570396e-01 4.37824339e-01
-5.42543828e-01 -2.64445841e-02 3.08462024e-01 6.99005365e-01
3.68083805e-01 -5.14147997e-01 -5.06040394e-01 -1.05158341e+00
5.66938162e-01 -3.32202315e-01 -3.76963735e-01 -9.70873296e-01
-1.10012400e+00 6.80327475e-01 -1.15427105e-02 -1.23102701e+00
2.12624580e-01 -7.42502689e-01 -5.57628989e-01 8.32157433e-01
-1.87273502e+00 -1.30642974e+00 -3.25273037e-01 4.30385888e-01
1.64654672e-01 1.84734359e-01 9.40833032e-01 2.49847695e-01
-2.72366732e-01 1.50930390e-01 1.25477195e-01 -2.44827986e-01
3.78770500e-01 -9.27244246e-01 8.22123811e-02 4.93001670e-01
1.00605533e-01 4.79847789e-01 4.68093097e-01 -5.65450311e-01
-1.61207151e+00 -8.35527122e-01 3.29828173e-01 -3.58190715e-01
-1.12167643e-02 -5.30297339e-01 -1.05898345e+00 6.28960133e-01
-5.80267720e-02 1.68946773e-01 2.67560720e-01 -2.22722813e-01
2.05799770e-02 -3.08085710e-01 -1.34789729e+00 4.86578315e-01
9.88297343e-01 -5.26778638e-01 -4.41602141e-01 2.75166988e-01
7.74893403e-01 -7.32318997e-01 -8.99261951e-01 3.47285539e-01
7.38946080e-01 -1.35591936e+00 1.31795359e+00 2.74973899e-01
4.33059752e-01 -4.29222256e-01 -1.24513894e-01 -9.97216821e-01
-4.13697630e-01 -6.86171889e-01 -1.95048451e-01 8.92916977e-01
2.32466802e-01 -8.40076089e-01 9.31878090e-01 4.42863375e-01
-4.00979549e-01 -1.09166491e+00 -1.01139653e+00 -3.41008425e-01
-3.62634450e-01 -4.50787961e-01 6.00480616e-01 7.63807952e-01
-8.61439288e-01 -2.02442165e-02 -7.49704316e-02 5.89472234e-01
9.58227038e-01 6.96182013e-01 6.02471411e-01 -1.44656229e+00
-6.82135284e-01 -2.89479882e-01 -1.40702188e-01 -1.47921598e+00
4.67312988e-03 -7.56762981e-01 5.08399738e-04 -1.73632658e+00
-1.25508130e-01 -9.78164792e-01 1.82251722e-01 3.05919886e-01
4.35011268e-01 3.85319591e-01 -1.41257301e-01 3.67401451e-01
-4.16793041e-02 7.14739442e-01 1.67947197e+00 9.81304124e-02
-4.36177731e-01 -5.11345007e-02 -1.30256847e-01 1.05773973e+00
4.77092475e-01 -3.99662882e-01 -1.00525647e-01 -1.08712256e+00
4.15758163e-01 3.12218741e-02 5.01136243e-01 -1.03041184e+00
1.35971561e-01 -1.91421717e-01 3.21983665e-01 -8.29832137e-01
9.12379622e-01 -1.18565643e+00 4.32217538e-01 2.70323902e-01
-7.27153420e-02 -3.22903812e-01 1.23566553e-01 4.70749736e-01
2.16617331e-01 -4.77758139e-01 9.00666058e-01 -4.39970940e-01
-4.54020724e-02 3.90976220e-01 9.40079987e-02 -5.76125503e-01
7.22783804e-01 -6.20345592e-01 1.65765718e-01 -3.44997421e-02
-4.95829493e-01 -1.25535727e-01 1.01569462e+00 -1.82717249e-01
9.11146879e-01 -1.21059120e+00 -5.04109621e-01 5.94637871e-01
-6.81221485e-02 6.17053092e-01 1.15299515e-01 6.42421067e-01
-1.24792790e+00 2.69506499e-02 -1.30113095e-01 -9.58558559e-01
-1.21119380e+00 1.96429729e-01 6.20406508e-01 -1.23995408e-01
-1.23521936e+00 8.95597816e-01 3.00788522e-01 -6.97864771e-01
1.69033632e-01 -8.00922453e-01 -7.80802569e-04 -5.40226281e-01
1.83317989e-01 3.69280815e-01 1.92844272e-01 -5.17243564e-01
-1.35845304e-01 1.38312674e+00 4.16579187e-01 -1.52761519e-01
1.65513968e+00 -8.42520297e-02 -4.15656030e-01 5.89946330e-01
1.07580197e+00 5.05543888e-01 -1.39759731e+00 -2.18908757e-01
-8.01734030e-01 -5.87070167e-01 3.77351671e-01 -5.42313516e-01
-1.09096527e+00 1.02648318e+00 4.54454124e-01 3.53142507e-02
7.96192706e-01 -3.32906796e-03 9.22663927e-01 4.88397092e-01
7.66556978e-01 -6.64475858e-01 -1.33676454e-01 1.87351122e-01
1.14545262e+00 -8.64055991e-01 2.14794993e-01 -8.40737343e-01
9.10096895e-03 1.30996466e+00 3.27902734e-01 -2.14257717e-01
6.82635367e-01 4.01064754e-01 -4.42504399e-02 -3.91827613e-01
-1.26682416e-01 2.56426036e-01 3.37240547e-01 4.79429334e-01
1.02281198e-01 2.26861611e-02 7.97841772e-02 4.61324491e-02
-4.71582403e-03 -6.84886500e-02 3.81816030e-01 8.11173260e-01
-4.09022093e-01 -1.23033130e+00 -6.99036300e-01 2.52946854e-01
-1.27929717e-01 1.28848895e-01 1.99598875e-02 4.86587256e-01
3.02169800e-01 1.96407437e-01 3.60220551e-01 1.19411927e-02
3.61485451e-01 -1.66377217e-01 1.00924075e+00 -7.30474472e-01
-4.04178172e-01 2.70103842e-01 -7.96439052e-02 -6.57577455e-01
-5.85847616e-01 -1.79362491e-01 -1.38259697e+00 4.83151823e-02
-2.80814111e-01 -2.74963439e-01 7.67460108e-01 6.96936011e-01
4.47711378e-01 3.34879428e-01 5.62009275e-01 -1.48444355e+00
-7.61654302e-02 -5.49823165e-01 -5.59262514e-01 1.10176884e-01
5.92830837e-01 -7.98439205e-01 -1.67571947e-01 4.29447256e-02] | [9.319685935974121, -3.190127372741699] |
f00ca646-ad70-42f6-a6e3-8f76fde9d8b6 | grad-fec-unequal-loss-protection-of-deep | 2307.01846 | null | https://arxiv.org/abs/2307.01846v1 | https://arxiv.org/pdf/2307.01846v1.pdf | Grad-FEC: Unequal Loss Protection of Deep Features in Collaborative Intelligence | Collaborative intelligence (CI) involves dividing an artificial intelligence (AI) model into two parts: front-end, to be deployed on an edge device, and back-end, to be deployed in the cloud. The deep feature tensors produced by the front-end are transmitted to the cloud through a communication channel, which may be subject to packet loss. To address this issue, in this paper, we propose a novel approach to enhance the resilience of the CI system in the presence of packet loss through Unequal Loss Protection (ULP). The proposed ULP approach involves a feature importance estimator, which estimates the importance of feature packets produced by the front-end, and then selectively applies Forward Error Correction (FEC) codes to protect important packets. Experimental results demonstrate that the proposed approach can significantly improve the reliability and robustness of the CI system in the presence of packet loss. | ['Ivan V. Bajić', 'S. Faegheh Yeganli', 'Korcan Uyanik'] | 2023-07-04 | null | null | null | null | ['feature-importance'] | ['methodology'] | [-2.35030912e-02 -7.48406351e-02 1.62438154e-01 4.66703363e-02
7.54874051e-02 -1.33860081e-01 3.34061310e-02 2.51448471e-02
-4.86471225e-03 4.97535348e-01 5.00377640e-02 -2.41944585e-02
-3.44879985e-01 -9.27787721e-01 -2.23677188e-01 -7.83456504e-01
-3.57459813e-01 -9.19358507e-02 1.12553097e-01 1.33101255e-01
3.33078563e-01 7.03746974e-01 -1.27899349e+00 1.67056039e-01
8.21544707e-01 1.84038281e+00 1.33630201e-01 5.45353413e-01
-1.22024892e-02 1.30315328e+00 -8.61181140e-01 -2.20385522e-01
2.67464280e-01 -6.71704020e-03 -4.48770046e-01 1.83472991e-01
-5.19747019e-01 -3.73257965e-01 -3.72007936e-01 1.04611087e+00
3.50133359e-01 -6.83273897e-02 3.11386347e-01 -1.90477955e+00
-5.03880203e-01 2.58359075e-01 -4.63851064e-01 4.05864209e-01
-5.52680157e-02 5.07052392e-02 4.91438419e-01 -8.96778762e-01
2.65567005e-01 8.48793089e-01 5.75552642e-01 8.22986886e-02
-4.54213172e-01 -9.33821142e-01 2.03540042e-01 6.20127559e-01
-1.39018524e+00 -6.41614497e-01 1.18617785e+00 -1.87612593e-01
3.58881265e-01 1.62649408e-01 2.52184659e-01 2.29909971e-01
5.20326376e-01 7.48964131e-01 4.51030672e-01 -1.99887037e-01
5.76972663e-01 1.91034243e-01 2.53330613e-03 3.16266596e-01
3.25200677e-01 2.26025611e-01 -5.27112067e-01 -4.60185021e-01
4.50054049e-01 3.95443738e-01 -3.96817684e-01 -9.75519046e-02
-7.57981896e-01 5.10933340e-01 6.12289250e-01 1.36711672e-01
-1.04411244e+00 -1.91820413e-01 5.85361183e-01 4.65242565e-01
6.80286050e-01 -1.52326971e-01 -4.63967115e-01 1.00334056e-01
-8.84870529e-01 -5.37727833e-01 8.80223036e-01 1.04053605e+00
2.11490005e-01 -3.93410586e-02 -4.69192229e-02 4.08922821e-01
4.15701985e-01 1.06119104e-01 2.59410262e-01 -1.02230620e+00
4.22901958e-01 3.58217686e-01 2.21927185e-02 -1.37983346e+00
-1.93909824e-01 -9.14959431e-01 -1.28493595e+00 4.59892094e-01
-3.69908541e-01 -5.62482953e-01 -9.63030457e-02 1.24640906e+00
4.16330487e-01 5.91626167e-01 2.96507388e-01 9.44537163e-01
3.29064995e-01 7.64563918e-01 -1.38133004e-01 -5.84967196e-01
8.85881841e-01 -8.64835203e-01 -6.71692967e-01 3.13020617e-01
2.79680967e-01 -7.29195058e-01 2.25915387e-02 3.88238668e-01
-1.21127987e+00 -6.31022453e-01 -9.31433618e-01 4.94049609e-01
1.79580394e-02 1.55332595e-01 1.56969845e-01 5.05906463e-01
-1.01004148e+00 4.74108189e-01 -3.50678980e-01 9.65958312e-02
6.01116836e-01 3.59907538e-01 -2.77070642e-01 -3.00577015e-01
-8.58178437e-01 5.12573346e-02 7.66042471e-02 1.00388296e-01
-6.39475167e-01 -9.10366237e-01 -4.82739866e-01 4.78174537e-01
6.66012466e-02 -8.57704878e-01 6.71628833e-01 -1.42610073e+00
-1.30601120e+00 -8.75025690e-02 3.80195901e-02 -4.51646417e-01
2.40350708e-01 -5.01186699e-02 -4.10482556e-01 3.95630926e-01
-1.11363024e-01 2.72422563e-02 9.37617302e-01 -1.13341665e+00
-9.98516321e-01 -4.95667428e-01 -7.10862875e-02 9.68540087e-02
-5.79283297e-01 2.83061862e-01 -3.59008610e-01 -8.66632760e-01
1.09529160e-01 -6.72412395e-01 -4.19527777e-02 4.86570179e-01
-2.02119648e-01 -5.52553050e-02 1.40545058e+00 -7.00513899e-01
8.70600402e-01 -2.46121693e+00 1.82408243e-02 2.79767722e-01
2.18885586e-01 4.76369709e-01 1.11627933e-02 2.59950638e-01
1.51606604e-01 -9.03038755e-02 7.29607046e-02 -3.52267653e-01
-6.22330725e-01 -5.11890575e-02 -3.95210475e-01 3.21318537e-01
3.66047211e-02 1.00237474e-01 -7.79935479e-01 -3.46022964e-01
3.95281427e-02 5.33206761e-01 -5.63245893e-01 4.19183701e-01
4.36560541e-01 3.84432197e-01 -5.80699146e-01 8.26374173e-01
1.03578138e+00 8.80410615e-03 -3.21777575e-02 -3.01417172e-01
-3.92146669e-02 -2.08156645e-01 -1.05253696e+00 1.01267302e+00
-5.57659328e-01 6.20825052e-01 5.57391465e-01 -9.37911212e-01
8.03606153e-01 5.43366432e-01 8.81675243e-01 -3.66681010e-01
2.32724696e-01 -1.14491142e-01 -2.33226538e-01 -2.65726238e-01
1.90949395e-01 2.55353302e-01 1.77707687e-01 3.31115752e-01
-3.53950888e-01 6.81870162e-01 -1.41271651e-01 1.46939635e-01
1.51407349e+00 -4.19974983e-01 -2.28074923e-01 7.96781778e-02
9.00238991e-01 -3.18566740e-01 1.02256048e+00 4.59823012e-01
-5.25592387e-01 2.85709411e-01 1.54167965e-01 -3.03226948e-01
-7.00602949e-01 -9.71395969e-01 1.85907528e-01 5.85146904e-01
4.64916170e-01 -2.40018547e-01 -4.44616079e-01 -7.53292918e-01
1.97550170e-02 4.78532284e-01 -1.86021119e-01 -6.16218686e-01
1.74858961e-02 -6.04375780e-01 1.16753407e-01 5.83213389e-01
1.00969744e+00 -8.47468257e-01 -4.26978767e-01 5.15883803e-01
-6.71026483e-02 -1.19465709e+00 -2.81294018e-01 -4.03878748e-01
-8.24844122e-01 -9.25435424e-01 -3.81544203e-01 -4.34929103e-01
8.12660933e-01 6.04730844e-01 7.03652680e-01 4.14976507e-01
-1.23496182e-01 4.33018446e-01 -7.06175566e-01 -4.55690742e-01
-1.31000653e-01 -4.81201947e-01 1.41259357e-01 6.81621194e-01
-9.44632068e-02 -6.80071652e-01 -6.38287544e-01 2.09097698e-01
-8.07476640e-01 -2.14620799e-01 5.65079272e-01 5.09471476e-01
4.59411412e-01 1.00107563e+00 6.62770271e-01 -4.65095043e-01
5.62588871e-01 -1.01189387e+00 -3.23715776e-01 -2.35908460e-02
-3.92801464e-01 -3.29052955e-01 1.04849184e+00 -1.08353466e-01
-8.91047895e-01 -2.07560812e-03 7.19991773e-02 -5.97471058e-01
1.34811789e-01 4.18090731e-01 -4.64705735e-01 -3.36890846e-01
-1.39080599e-01 8.65715593e-02 8.13451633e-02 -6.13671660e-01
-7.20145926e-02 1.19883931e+00 4.43163365e-01 6.95257932e-02
8.19949806e-01 6.29306138e-01 2.60860503e-01 -8.45141232e-01
-4.07375664e-01 -2.42579103e-01 -2.09216565e-01 -4.62337196e-01
2.58404285e-01 -1.02792382e+00 -8.05001140e-01 5.04864812e-01
-1.07806396e+00 2.28514820e-01 -6.07464276e-02 5.30365407e-01
-6.66118413e-02 2.56101191e-01 -6.23023391e-01 -9.41082656e-01
-8.57237577e-01 -8.51284206e-01 6.04426384e-01 5.07266223e-01
4.59113419e-01 -5.73527813e-01 -5.51270485e-01 3.91797781e-01
6.19970381e-01 7.55282566e-02 4.35210586e-01 -2.81002551e-01
-6.52617693e-01 -5.13898730e-01 -6.16734445e-01 5.36007226e-01
2.71621406e-01 1.56702355e-01 -6.82591677e-01 -4.03414339e-01
3.60322207e-01 4.24871117e-01 5.25578499e-01 2.28179976e-01
1.42632484e+00 -4.46195900e-01 -4.09589112e-01 7.68282294e-01
1.30499840e+00 4.24680263e-01 5.62524140e-01 -1.10127807e-01
3.00288558e-01 4.65746641e-01 8.46201897e-01 8.86553943e-01
4.82305259e-01 4.14711177e-01 8.05270016e-01 1.11149736e-01
-3.83458644e-01 3.90430748e-01 5.50580561e-01 8.32588255e-01
2.39526719e-01 -4.98999059e-01 -5.54988265e-01 3.23503196e-01
-1.67446184e+00 -8.41640115e-01 8.95235464e-02 2.17477512e+00
1.42327309e-01 2.37681225e-01 -1.59407616e-01 5.49699247e-01
7.78957546e-01 -2.66973644e-01 -5.67643404e-01 -1.91283852e-01
3.48806202e-01 -3.56267035e-01 4.42349464e-01 1.52867496e-01
-8.39138508e-01 2.69103229e-01 5.21319914e+00 5.33612609e-01
-1.22250640e+00 2.73378313e-01 5.65865457e-01 -1.03809305e-01
9.70078483e-02 -1.54122077e-02 -2.95286924e-01 1.23265839e+00
9.90998268e-01 -2.70224690e-01 2.58653164e-01 8.65763962e-01
4.57927346e-01 -4.54547331e-02 -5.22872329e-01 8.61598492e-01
-6.01772107e-02 -1.24518144e+00 -2.82043278e-01 -4.28324006e-02
2.08591565e-01 -2.93541905e-02 -4.89511862e-02 2.41859052e-02
-7.37788975e-02 -1.54141992e-01 4.57272500e-01 1.01287556e+00
3.02466691e-01 -1.25236595e+00 1.07686758e+00 4.71703589e-01
-1.37050617e+00 -6.14463747e-01 -4.20936197e-01 -2.14987546e-01
3.78684551e-02 1.28907728e+00 -4.98046607e-01 6.88309669e-01
8.98074687e-01 5.07286370e-01 -1.90274462e-01 1.28851056e+00
5.06119281e-02 4.91692245e-01 -2.25443155e-01 2.66904742e-01
-3.04292202e-01 -4.66068648e-02 8.99295807e-01 5.12331307e-01
4.06880498e-01 4.26262245e-02 2.67686248e-01 4.31066066e-01
-3.60138953e-01 -1.04026742e-01 -5.38016081e-01 3.01373512e-01
8.77289176e-01 1.23832858e+00 -5.41898787e-01 -3.00834596e-01
-4.62224036e-01 1.39782500e+00 1.84200183e-01 2.54812717e-01
-6.26385868e-01 -7.19900310e-01 9.64497924e-01 -2.41100699e-01
3.25640529e-01 -1.50585681e-01 -3.28266025e-01 -8.22380185e-01
3.30295444e-01 -2.88500845e-01 3.10528278e-01 -9.52442527e-01
-1.13061261e+00 6.06999278e-01 -7.05823302e-01 -1.38339639e+00
1.50971532e-01 -8.00684988e-02 -8.09476912e-01 5.92392027e-01
-1.62896013e+00 -9.13448811e-01 -4.43386525e-01 7.16454327e-01
4.15069461e-01 -3.54627103e-01 5.06357610e-01 7.11131394e-01
-9.13244903e-01 6.61086619e-01 3.96780491e-01 8.22844580e-02
2.85366982e-01 -4.80051041e-01 -1.32734925e-01 1.15115058e+00
-2.96804339e-01 2.74968147e-02 3.47620100e-01 -7.20337331e-01
-1.49113369e+00 -1.43873417e+00 7.55188048e-01 2.39096716e-01
2.79516429e-01 9.67984647e-02 -5.20260811e-01 2.06188709e-01
-1.01058677e-01 7.37594545e-01 8.21932375e-01 -3.35426241e-01
-1.56802952e-01 -5.88873386e-01 -1.84946632e+00 2.92715847e-01
7.55980492e-01 -3.34961057e-01 -4.31111194e-02 2.42107853e-01
7.55764008e-01 2.14559495e-01 -9.90116835e-01 3.31460118e-01
2.73302972e-01 -8.71581316e-01 9.24541712e-01 -1.48767218e-01
9.66537595e-02 -4.51332480e-01 -9.99024510e-02 -1.33097970e+00
-5.23012996e-01 -5.56635320e-01 -4.59472537e-01 1.53270471e+00
-2.03399315e-01 -6.42799795e-01 5.99329054e-01 5.95294535e-01
-1.58781484e-01 -5.33330441e-01 -1.00401223e+00 -7.23561943e-01
-4.93124574e-01 -4.09448236e-01 7.38014281e-01 5.02271056e-01
-3.41879390e-02 4.26246710e-02 -1.46062523e-01 8.52754414e-01
8.37324142e-01 -2.32865959e-02 4.29786116e-01 -1.31275892e+00
-1.57084942e-01 -1.43668586e-02 -7.73372650e-01 -4.26183999e-01
8.08914825e-02 -7.67532408e-01 -1.89022288e-01 -1.11281681e+00
-1.13575719e-01 -7.43356586e-01 -7.02905416e-01 2.14970261e-01
-1.62363037e-01 2.62913197e-01 6.14990234e-01 6.21998668e-01
-6.10574543e-01 6.56689286e-01 5.24718642e-01 5.72115071e-02
-8.48236084e-02 3.55720997e-01 -6.76643908e-01 7.77184904e-01
9.96798337e-01 -6.51481330e-01 -4.55885977e-01 -3.78822446e-01
-2.25767121e-01 5.01922667e-01 2.51482457e-01 -1.50366127e+00
7.12709129e-01 6.04607873e-02 5.88166952e-01 -4.19787288e-01
1.82372093e-01 -1.56498849e+00 1.54581651e-01 4.75701809e-01
9.68097374e-02 -7.76837021e-02 1.92537848e-02 6.77213907e-01
-2.11848795e-01 1.01308651e-01 6.05765402e-01 4.31941092e-01
-4.50285822e-01 5.77316523e-01 -5.68713665e-01 -6.02050185e-01
1.36919343e+00 9.46983099e-02 -1.22067459e-01 -2.98916847e-01
-4.06084239e-01 3.78222078e-01 3.22118968e-01 2.80937374e-01
8.64337504e-01 -1.19356441e+00 -7.03163207e-01 4.95348155e-01
-6.72985613e-02 -4.52852517e-01 4.38040107e-01 8.06724489e-01
-2.89131880e-01 -9.25608426e-02 -2.26010606e-01 -1.74679786e-01
-1.38322508e+00 9.78743017e-01 6.14513382e-02 -1.95881333e-02
-5.44668913e-01 7.62383580e-01 -1.32901281e-01 2.60795265e-01
3.37070048e-01 4.33544517e-01 -2.04669684e-01 -7.84138069e-02
1.00760102e+00 5.57279825e-01 1.19108953e-01 -7.65647471e-01
-3.81191581e-01 1.78796738e-01 -1.97106786e-02 2.85032064e-01
1.18800724e+00 -8.31963420e-01 -2.26092130e-01 -2.95822799e-01
1.20745587e+00 -1.10336415e-01 -1.13766742e+00 -4.54440266e-01
-3.11533868e-01 -6.05053782e-01 9.14558649e-01 -6.38612032e-01
-1.81315017e+00 3.11481833e-01 7.77838826e-01 2.66845524e-01
1.81248534e+00 -4.59088802e-01 1.17529643e+00 -6.35696650e-02
6.16825819e-01 -9.13677812e-01 -2.16958880e-01 2.24136248e-01
7.61560500e-01 -9.05986786e-01 -1.97837561e-01 -5.65124869e-01
-3.89530301e-01 1.12680328e+00 2.78829008e-01 -1.87872812e-01
1.26324224e+00 5.54814875e-01 -2.42649406e-01 -3.32797244e-02
-9.68406737e-01 4.57341075e-02 -1.83422193e-01 8.25101316e-01
-6.13593347e-02 2.82983445e-02 -1.81014299e-01 7.10667193e-01
3.10040951e-01 3.96266818e-01 5.04435360e-01 1.15479159e+00
-3.31473947e-01 -8.48185182e-01 -5.58467507e-01 5.28940976e-01
-5.64538240e-01 1.62578076e-02 -1.37280390e-01 -1.16378598e-01
4.10753310e-01 1.54149044e+00 3.56724083e-01 -7.85935402e-01
1.60296008e-01 -2.81401157e-01 -1.95752323e-01 -1.77956358e-01
-5.78963816e-01 -4.73889261e-01 -1.61580607e-01 -7.44318247e-01
-1.20106526e-01 -4.72785383e-01 -1.21846020e+00 -5.14907897e-01
-2.02023745e-01 2.64772594e-01 9.12612975e-01 8.64819527e-01
1.14998531e+00 7.90725768e-01 1.50541413e+00 -6.69984698e-01
-1.58626989e-01 -5.18212378e-01 -6.52916789e-01 1.07717253e-01
5.60030043e-01 -4.48665589e-01 -5.40846765e-01 -1.37874544e-01] | [5.99778938293457, 5.6057868003845215] |
8795ba58-bf7f-4620-82c3-2e291b927862 | discovering-novel-actions-in-an-open-world | 2305.16602 | null | https://arxiv.org/abs/2305.16602v1 | https://arxiv.org/pdf/2305.16602v1.pdf | Discovering Novel Actions in an Open World with Object-Grounded Visual Commonsense Reasoning | Learning to infer labels in an open world, i.e., in an environment where the target ``labels'' are unknown, is an important characteristic for achieving autonomy. Foundation models pre-trained on enormous amounts of data have shown remarkable generalization skills through prompting, particularly in zero-shot inference. However, their performance is restricted to the correctness of the target label's search space. In an open world where these labels are unknown, the search space can be exceptionally large. It can require reasoning over several combinations of elementary concepts to arrive at an inference, which severely restricts the performance of such models. To tackle this challenging problem, we propose a neuro-symbolic framework called ALGO - novel Action Learning with Grounded Object recognition that can use symbolic knowledge stored in large-scale knowledge bases to infer activities (verb-noun combinations) in egocentric videos with limited supervision using two steps. First, we propose a novel neuro-symbolic prompting approach that uses object-centric vision-language foundation models as a noisy oracle to ground objects in the video through evidence-based reasoning. Second, driven by prior commonsense knowledge, we discover plausible activities through an energy-based symbolic pattern theory framework and learn to ground knowledge-based action (verb) concepts in the video. Extensive experiments on two publicly available datasets (GTEA Gaze and GTEA Gaze Plus) demonstrate its performance on open-world activity inference and its generalization to unseen actions in an unknown search space. We show that ALGO can be extended to zero-shot settings and demonstrate its competitive performance to multimodal foundation models. | ['Shubham Trehan', 'Sanjoy Kundu', 'Sathyanarayanan N. Aakur'] | 2023-05-26 | null | null | null | null | ['object-recognition', 'visual-commonsense-reasoning'] | ['computer-vision', 'reasoning'] | [ 3.87261182e-01 2.54799157e-01 -3.03837985e-01 -3.11506957e-01
-5.12782335e-01 -4.19015288e-01 5.94665229e-01 -4.06296313e-01
-3.52681428e-01 7.96268463e-01 2.03422338e-01 5.10778762e-02
-3.42754096e-01 -6.72944725e-01 -1.24106526e+00 -6.13428295e-01
3.15982290e-02 6.17165565e-01 2.49157161e-01 -2.66045988e-01
3.98752019e-02 1.40269592e-01 -2.00576901e+00 4.47225988e-01
6.15727186e-01 1.06734788e+00 2.87071824e-01 5.00103176e-01
1.11159794e-01 1.24549210e+00 -2.33048752e-01 -2.39660770e-01
1.72590673e-01 -4.89848346e-01 -1.06803918e+00 2.48666048e-01
4.72751945e-01 -4.40493107e-01 -4.49398041e-01 1.23797309e+00
-1.85869429e-02 4.95907307e-01 6.23399079e-01 -1.56235898e+00
-7.40949214e-01 5.54182112e-01 8.07437152e-02 1.51278421e-01
7.97667503e-01 4.71808970e-01 1.18852186e+00 -7.96787262e-01
9.16121423e-01 1.31863332e+00 4.95866686e-01 7.46147990e-01
-1.22754824e+00 -4.16933924e-01 2.76093185e-01 8.28565717e-01
-1.34081364e+00 -4.68269378e-01 5.74864686e-01 -3.63697290e-01
1.14943695e+00 2.16764629e-01 1.10360646e+00 1.66380310e+00
1.06664607e-03 8.53771925e-01 1.11552644e+00 -3.80639195e-01
6.82015002e-01 -2.41469488e-01 -1.10545993e-01 1.06068420e+00
1.82880729e-01 1.57905251e-01 -1.07433820e+00 8.93533677e-02
5.84178388e-01 1.11647338e-01 -3.87381524e-01 -4.42228049e-01
-1.77319217e+00 6.43343806e-01 5.98342359e-01 1.78662166e-01
-3.89442027e-01 5.64569473e-01 1.28114209e-01 2.34276548e-01
-5.04175462e-02 5.86388171e-01 -2.27751300e-01 -3.09873253e-01
-6.95549786e-01 2.91880250e-01 9.57207441e-01 1.13434982e+00
8.15934300e-01 -6.38723671e-02 -1.24918334e-01 3.59303176e-01
1.94573477e-01 7.06323147e-01 5.82527518e-01 -1.51376402e+00
2.47866198e-01 5.87288260e-01 1.28296897e-01 -8.29032540e-01
-3.30227137e-01 2.95576174e-02 -3.38459820e-01 2.07907230e-01
5.99228024e-01 1.69274673e-01 -8.65822852e-01 1.91501164e+00
4.24681842e-01 6.93462133e-01 3.03250223e-01 1.11762941e+00
8.67099643e-01 2.56074995e-01 -1.21214777e-01 -9.20782089e-02
1.33083189e+00 -9.17699516e-01 -7.02001035e-01 -2.04320431e-01
7.25753188e-01 1.55788988e-01 1.20696163e+00 6.14856243e-01
-8.48457098e-01 -3.83767873e-01 -9.38956380e-01 -1.73866764e-01
-5.21488011e-01 -2.27818877e-01 8.60915899e-01 1.27537742e-01
-8.92115533e-01 5.72560191e-01 -9.48903322e-01 -6.13694370e-01
6.99516594e-01 4.47371006e-01 -6.11913502e-01 -2.28829309e-01
-1.08626819e+00 9.26797926e-01 8.66673291e-01 5.80562353e-02
-1.40864551e+00 -2.40918636e-01 -1.10679018e+00 6.17202446e-02
1.12945271e+00 -8.56408715e-01 1.31163180e+00 -1.07625437e+00
-1.45952463e+00 7.08555698e-01 -2.44051963e-02 -4.51014817e-01
3.78413290e-01 -1.06533527e-01 -2.32978418e-01 3.89334559e-01
2.34218806e-01 9.07682896e-01 8.85997176e-01 -9.13722336e-01
-6.48761809e-01 -3.68435740e-01 6.19527400e-01 3.71683180e-01
-2.78824031e-01 -3.34712803e-01 -3.04655194e-01 -2.87640810e-01
3.82103115e-01 -1.16999304e+00 4.61229943e-02 1.78682223e-01
-3.30949724e-01 -4.65634435e-01 5.56450725e-01 -3.36770982e-01
6.10210896e-01 -1.93114698e+00 6.27558887e-01 1.22456633e-01
1.38574138e-01 -2.24744320e-01 6.57005161e-02 6.31774813e-02
1.83905840e-01 -2.90814042e-01 -5.38376793e-02 -3.88019644e-02
1.87836185e-01 8.80185902e-01 -4.43895251e-01 4.64369506e-01
1.53732210e-01 1.22645462e+00 -1.40519786e+00 -6.72607541e-01
1.53260410e-01 2.08741561e-01 -7.18538165e-01 1.69034794e-01
-5.99956751e-01 4.86598969e-01 -4.59886879e-01 1.04611635e+00
-1.68480590e-01 -4.34031218e-01 1.50963649e-01 -1.64109603e-01
2.50244826e-01 -1.86505854e-01 -1.14720404e+00 2.23097730e+00
-1.56856135e-01 5.89932382e-01 -4.61540997e-01 -1.15075171e+00
3.79908293e-01 2.77389944e-01 4.03266966e-01 -4.74948168e-01
1.83715120e-01 1.96899801e-01 1.52493315e-02 -1.00607800e+00
2.28620768e-01 -3.54163349e-01 -1.84248611e-01 2.34679505e-01
6.26365125e-01 -2.54560530e-01 2.55670488e-01 2.97794908e-01
1.34835923e+00 7.58827865e-01 2.85486996e-01 1.98347151e-01
3.82374763e-01 2.90247440e-01 5.69678962e-01 9.09721494e-01
-2.08080396e-01 2.39006206e-01 5.41791081e-01 -4.30242717e-01
-6.14414692e-01 -1.04019284e+00 3.42525356e-02 1.28469920e+00
3.77509922e-01 -3.78640145e-01 -7.65323460e-01 -7.70761669e-01
-1.19924195e-01 8.60320926e-01 -8.52508068e-01 -3.17426443e-01
-3.18419695e-01 -1.26749277e-01 6.24501109e-01 7.15235710e-01
4.19121861e-01 -1.38848042e+00 -1.07075059e+00 -9.19495076e-02
-5.47508001e-01 -1.51192462e+00 -7.90032558e-03 2.99398810e-01
-7.42483854e-01 -1.33696437e+00 -3.20294917e-01 -4.47573096e-01
8.05594921e-01 -1.30183734e-02 8.75233233e-01 -6.10178187e-02
-1.57496393e-01 1.02528572e+00 -4.23508227e-01 -4.15457338e-01
-1.46942362e-01 -5.30074000e-01 4.82587725e-01 2.11965263e-01
4.76243913e-01 -5.44131398e-01 -3.23555112e-01 3.31077754e-01
-6.72002494e-01 1.70453966e-01 4.61097509e-01 1.01288545e+00
6.64208710e-01 -1.77950770e-01 3.54446948e-01 -4.04752225e-01
2.19849795e-01 -5.68357050e-01 -4.74682391e-01 3.43554080e-01
-2.98985571e-01 2.42158026e-01 4.18117255e-01 -7.39860773e-01
-9.52091694e-01 1.53666854e-01 3.65820140e-01 -9.28824663e-01
-2.52947658e-01 5.92079878e-01 -1.43691048e-01 -4.82074395e-02
8.26253831e-01 2.94799566e-01 -6.00766577e-02 -9.73258838e-02
5.06458223e-01 3.65376741e-01 1.00632083e+00 -8.47501099e-01
6.22483313e-01 9.33680534e-01 2.86091208e-01 -6.20355785e-01
-1.17091131e+00 -4.52168971e-01 -7.65238941e-01 -4.69283730e-01
1.08951473e+00 -1.05202663e+00 -1.04262877e+00 5.37848435e-02
-8.98646295e-01 -4.75429744e-01 -6.48826838e-01 7.42286682e-01
-1.29009640e+00 3.60724717e-01 -1.47927105e-01 -6.50771976e-01
2.24847853e-01 -1.06512427e+00 1.05871463e+00 4.90503274e-02
-3.38818222e-01 -8.15930068e-01 -9.40676257e-02 6.45171642e-01
-1.81630209e-01 3.31876069e-01 5.65196395e-01 -7.91020751e-01
-8.31658006e-01 -8.14090297e-03 6.31213561e-02 1.82625726e-01
-1.38668761e-01 -4.61647689e-01 -8.81032288e-01 1.40000470e-02
1.94117281e-04 -9.24795866e-01 7.38352239e-01 -7.38050714e-02
1.19346094e+00 -3.80470634e-01 -3.13722283e-01 6.92117691e-01
1.05818176e+00 -1.79095954e-01 4.44164336e-01 3.85524273e-01
6.82922959e-01 5.37433684e-01 8.58161151e-01 4.33496982e-01
4.78643268e-01 6.51234269e-01 8.09686303e-01 5.65986037e-01
-6.69169649e-02 -3.69186342e-01 6.75539851e-01 3.92829537e-01
-6.83767796e-01 4.08636257e-02 -7.88255990e-01 4.77501422e-01
-2.19147897e+00 -1.38984239e+00 2.24392146e-01 1.82648981e+00
8.38896275e-01 6.52772337e-02 -6.58634380e-02 1.63955837e-02
2.37399191e-01 -1.66890487e-01 -9.31661606e-01 -1.79589167e-02
-1.47001624e-01 -4.37718220e-02 1.00623369e-01 9.74918976e-02
-8.42817008e-01 1.03714299e+00 5.63278532e+00 5.67202747e-01
-8.21550310e-01 3.53963792e-01 -5.52713983e-02 -4.53415245e-01
-1.40453741e-01 1.55500650e-01 -6.12162292e-01 2.94666529e-01
8.81657958e-01 6.84221163e-02 8.41166914e-01 9.45001423e-01
-4.74131107e-02 -4.38432366e-01 -1.74100840e+00 1.40866554e+00
4.51446146e-01 -1.34726071e+00 -2.69814488e-02 4.11230810e-02
6.14301264e-01 2.06645176e-01 -2.52245277e-01 6.62201285e-01
4.26503986e-01 -1.01875603e+00 1.20720863e+00 8.09284270e-01
8.83096397e-01 -2.22917557e-01 3.94819975e-01 7.45780230e-01
-7.50443518e-01 -5.73666930e-01 -3.47496331e-01 -2.20208764e-01
1.14146948e-01 -9.36674476e-02 -7.41305470e-01 2.67309636e-01
8.41487825e-01 8.66158485e-01 -3.57833266e-01 6.05700731e-01
-6.60042942e-01 3.76879632e-01 -5.50025344e-01 -6.32881885e-03
3.35533619e-01 -6.65118620e-02 6.88868761e-01 7.19997048e-01
3.87422621e-01 4.63719815e-01 1.99465245e-01 1.09109902e+00
-1.08945213e-01 -3.05859298e-01 -8.35023105e-01 -4.00600806e-02
6.32466227e-02 9.82307792e-01 -7.10291266e-01 -5.58224857e-01
-4.64135289e-01 1.07524991e+00 4.85750765e-01 5.75697243e-01
-1.16234863e+00 4.20258828e-02 3.45560104e-01 -1.51153341e-01
3.62470478e-01 -9.90116596e-02 1.85350001e-01 -1.61536920e+00
7.32352585e-02 -8.55570972e-01 5.19652247e-01 -1.28125918e+00
-7.83903301e-01 9.32435095e-02 4.45280641e-01 -1.21461844e+00
-3.74482870e-01 -8.24029386e-01 -1.75393298e-01 2.33945802e-01
-1.37365460e+00 -1.31640112e+00 -6.98151290e-01 1.11541069e+00
6.82919979e-01 -1.58384353e-01 8.20356727e-01 -2.92715698e-01
-2.82937706e-01 1.47879571e-01 -2.15233728e-01 8.79876390e-02
4.09648657e-01 -1.22481346e+00 -2.82875597e-01 6.13666832e-01
7.13720202e-01 3.87343585e-01 6.17721677e-01 -6.73593879e-01
-1.80610061e+00 -8.62511098e-01 5.10792732e-01 -9.28896129e-01
8.12433839e-01 -4.15679067e-01 -7.42439628e-01 1.15498078e+00
-1.94617748e-01 5.52523196e-01 6.77121341e-01 6.37141764e-02
-4.89777476e-01 1.55992121e-01 -9.35677886e-01 7.04184771e-01
1.70006359e+00 -6.82645798e-01 -1.13089228e+00 6.46073878e-01
5.89490116e-01 -6.10609889e-01 -7.47698307e-01 4.16577578e-01
7.01980889e-01 -8.84179115e-01 8.53918791e-01 -1.05341041e+00
3.47477555e-01 -4.04472858e-01 -5.53146005e-01 -1.00683248e+00
2.25852057e-03 -6.16509736e-01 -7.53847599e-01 5.24183869e-01
-2.20237169e-02 -4.10714507e-01 5.81132829e-01 6.64428890e-01
-1.68595895e-01 -6.76882327e-01 -1.21451533e+00 -1.01916671e+00
-5.88659525e-01 -8.14115465e-01 4.92235184e-01 8.81443679e-01
3.90554190e-01 2.16745749e-01 -3.22071075e-01 9.56188589e-02
6.53648436e-01 2.04069838e-02 8.32818151e-01 -1.22915459e+00
-4.02957797e-01 -3.13185081e-02 -7.65900016e-01 -9.27329302e-01
6.98719323e-01 -1.04888821e+00 2.86575377e-01 -1.39805269e+00
3.00582469e-01 -7.54795820e-02 -4.12910044e-01 9.17539597e-01
2.33197168e-01 2.87127107e-01 2.32463211e-01 4.14794445e-01
-1.16726506e+00 6.50653720e-01 1.21481705e+00 -2.84735888e-01
-1.08456388e-02 -4.00229305e-01 -3.94848228e-01 1.13026977e+00
3.22866946e-01 -4.71749574e-01 -4.95283753e-01 -2.58072138e-01
6.34073436e-01 5.11877760e-02 9.39315557e-01 -1.26831377e+00
4.16029692e-01 -5.85781276e-01 2.27445662e-01 -1.89503163e-01
7.29771316e-01 -9.82933164e-01 5.41534722e-02 3.01937789e-01
-3.81322533e-01 -5.21524668e-01 -1.68019667e-01 1.12230885e+00
-9.16336940e-05 -2.15557069e-01 2.44866580e-01 -3.42732936e-01
-1.27005088e+00 1.09963961e-01 -1.24274492e-01 9.41008404e-02
1.31226170e+00 -6.89986467e-01 -2.14582279e-01 -3.45601290e-01
-1.19068623e+00 2.59497911e-01 4.37394381e-01 3.92892569e-01
7.63308525e-01 -1.35468614e+00 -3.23752016e-01 3.53448838e-02
4.93426919e-01 1.47908747e-01 1.20836332e-01 1.16627371e+00
-2.35120729e-01 4.14214671e-01 -3.45124841e-01 -9.10990000e-01
-9.54531491e-01 6.94188118e-01 3.10494244e-01 2.77038872e-01
-8.47806633e-01 8.05194139e-01 1.22505389e-01 -1.29000202e-01
3.46182644e-01 -6.88680112e-01 -1.60552248e-01 -9.88480523e-02
3.13523173e-01 2.64705539e-01 -2.87406415e-01 -8.90426874e-01
-3.56567293e-01 4.36917603e-01 4.30583060e-01 -1.03736058e-01
1.29999804e+00 2.83175502e-02 2.15477962e-02 8.13615561e-01
7.45803952e-01 -5.45689404e-01 -1.46356320e+00 -3.94794703e-01
-2.51256377e-02 -4.55464959e-01 1.50064193e-02 -6.60854042e-01
-5.34360707e-01 6.83340549e-01 3.25124025e-01 -2.35382438e-01
9.06926334e-01 4.47322756e-01 5.08518696e-01 1.30942833e+00
6.97311759e-01 -1.33562481e+00 6.45922661e-01 4.69217688e-01
9.05642211e-01 -1.55910802e+00 -4.32360359e-02 -9.10436064e-02
-6.84796691e-01 1.13705277e+00 7.30697989e-01 1.39767393e-01
3.46107006e-01 -2.91907817e-01 -4.06866521e-01 -5.21802187e-01
-8.07221293e-01 -4.70902592e-01 3.01174045e-01 6.75897717e-01
-3.96580964e-01 -9.86520667e-03 2.96687305e-01 5.46084583e-01
-1.01203524e-01 4.10286307e-01 5.63401818e-01 9.30935979e-01
-4.78036076e-01 -3.73236358e-01 -4.08439457e-01 3.28353226e-01
-1.36154667e-01 1.05162092e-01 -1.57929376e-01 6.63678229e-01
4.99741405e-01 8.49066734e-01 -6.63091093e-02 -2.48362750e-01
2.18119204e-01 3.51937741e-01 8.46341550e-01 -7.60164440e-01
-1.40144736e-01 -2.79439807e-01 4.10740562e-02 -1.21406400e+00
-8.97233725e-01 -7.72649288e-01 -1.61063659e+00 6.40818626e-02
-3.88171583e-01 -1.58838555e-01 5.40202022e-01 1.42559826e+00
1.59671813e-01 4.50587392e-01 -2.05874821e-04 -8.42306137e-01
-7.64437735e-01 -7.51428604e-01 -4.60374504e-01 6.44993603e-01
2.92928815e-01 -1.31658602e+00 -4.50639755e-01 4.61134195e-01] | [8.559459686279297, 0.818389892578125] |
dc716ce4-2a8a-477f-8d28-bb6ed905c99f | megan-multi-explanation-graph-attention | 2211.13236 | null | https://arxiv.org/abs/2211.13236v2 | https://arxiv.org/pdf/2211.13236v2.pdf | MEGAN: Multi-Explanation Graph Attention Network | We propose a multi-explanation graph attention network (MEGAN). Unlike existing graph explainability methods, our network can produce node and edge attributional explanations along multiple channels, the number of which is independent of task specifications. This proves crucial to improve the interpretability of graph regression predictions, as explanations can be split into positive and negative evidence w.r.t to a reference value. Additionally, our attention-based network is fully differentiable and explanations can actively be trained in an explanation-supervised manner. We first validate our model on a synthetic graph regression dataset with known ground-truth explanations. Our network outperforms existing baseline explainability methods for the single- as well as the multi-explanation case, achieving near-perfect explanation accuracy during explanation supervision. Finally, we demonstrate our model's capabilities on multiple real-world datasets. We find that our model produces sparse high-fidelity explanations consistent with human intuition about those tasks. | ['Pascal Friederich', 'Patrick Reiser', 'Luca Torresi', 'Jonas Teufel'] | 2022-11-23 | null | null | null | null | ['graph-regression'] | ['graphs'] | [ 5.29039681e-01 1.24102378e+00 -7.67625749e-01 -5.95923007e-01
-2.99008369e-01 -5.02876997e-01 8.06839406e-01 1.58784688e-01
3.23573232e-01 8.37729692e-01 6.42889202e-01 -7.99425066e-01
-3.14039916e-01 -5.58814585e-01 -1.04572058e+00 1.44363940e-01
7.29674548e-02 7.73350835e-01 -3.32738549e-01 -4.98460531e-02
3.42835784e-01 2.92852223e-02 -1.09813333e+00 2.69321352e-01
1.05510116e+00 4.01270628e-01 -2.62952596e-01 7.99513519e-01
9.20300186e-02 9.62248147e-01 -2.77709544e-01 -8.65087688e-01
-2.15974599e-02 -4.29359138e-01 -1.05589175e+00 2.28833944e-01
5.77385485e-01 -2.90484369e-01 -5.13616204e-01 7.84581304e-01
-2.68135756e-01 5.79228578e-03 9.39859867e-01 -1.92229342e+00
-1.58599806e+00 1.19776797e+00 -3.53176296e-01 7.20551759e-02
4.01989788e-01 2.32518852e-01 1.75244391e+00 -8.83045971e-01
6.62600279e-01 1.27832592e+00 6.77821219e-01 6.68759406e-01
-1.47335279e+00 -5.87938786e-01 6.12006366e-01 5.32352328e-02
-6.83650255e-01 -2.03274146e-01 5.83142161e-01 -2.44953007e-01
1.20884085e+00 2.72820950e-01 7.59130120e-01 1.43033803e+00
2.88735241e-01 5.10479808e-01 5.46705127e-01 -2.53023446e-01
-1.67964865e-02 -1.10347830e-01 5.43881595e-01 8.89837742e-01
8.14686716e-01 9.27109048e-02 -6.83647513e-01 3.33264954e-02
1.04044759e+00 1.74735680e-01 -5.19264698e-01 -3.64085019e-01
-1.48401403e+00 1.01198530e+00 1.01423788e+00 -1.17630675e-01
-4.39284652e-01 8.74633372e-01 -2.73377523e-02 3.88080448e-01
5.13920844e-01 9.39544916e-01 -5.70177197e-01 2.08298743e-01
-1.83580130e-01 1.15139231e-01 8.04916143e-01 1.12408102e+00
7.46783733e-01 3.07367027e-01 -2.15477273e-01 2.54644781e-01
4.77430671e-01 3.83128285e-01 8.85834545e-02 -1.04422915e+00
6.20189071e-01 8.19607198e-01 -1.30093955e-02 -1.10124683e+00
-4.19742942e-01 -7.21251190e-01 -9.37168658e-01 -9.52499136e-02
4.28536654e-01 -5.70314750e-02 -9.01497602e-01 1.84618127e+00
-1.42287254e-01 1.50388464e-01 1.46461591e-01 1.06800675e+00
1.01119339e+00 1.59528494e-01 1.39060751e-01 1.86687246e-01
1.10020030e+00 -1.45628786e+00 -8.36447716e-01 -7.91603625e-01
5.94768107e-01 -1.45436853e-01 1.28888559e+00 1.22409910e-02
-9.01903272e-01 -3.98692727e-01 -8.42060745e-01 -3.51221979e-01
-1.17758274e-01 -1.54593140e-01 1.40383637e+00 2.51392514e-01
-1.18103135e+00 5.94976187e-01 -4.81184334e-01 -1.65766060e-01
5.12427986e-01 6.26098037e-01 -7.21064925e-01 -1.82154953e-01
-1.12696159e+00 7.59709477e-01 8.50983635e-02 -7.76480809e-02
-7.77757943e-01 -7.18380690e-01 -1.12378287e+00 6.25665426e-01
3.90458852e-01 -1.29691041e+00 1.19757247e+00 -1.24040902e+00
-8.47528458e-01 5.09168744e-01 -4.64553952e-01 -4.85429317e-01
1.81644127e-01 -1.39797136e-01 -3.03388476e-01 -1.56731755e-01
3.67143363e-01 8.79914820e-01 5.62816203e-01 -1.39484406e+00
6.54514832e-03 -7.39404373e-03 5.63117921e-01 2.03211531e-01
-1.17574617e-01 -7.30917513e-01 -2.93891281e-01 -4.37645704e-01
2.93099970e-01 -9.07554567e-01 -3.19418848e-01 5.47148511e-02
-1.07334614e+00 -1.64236054e-01 2.81414956e-01 -3.73702347e-01
8.52101445e-01 -1.67672861e+00 2.76605725e-01 2.75249064e-01
9.94157016e-01 -4.60977495e-01 -2.69101143e-01 3.53464335e-01
-5.13974607e-01 7.58846343e-01 -1.10742673e-01 -6.32248819e-01
3.38527560e-01 4.19235110e-01 -4.29734349e-01 3.11234891e-01
4.13538396e-01 1.42874849e+00 -1.01347005e+00 -1.59314036e-01
-3.79804298e-02 3.95033807e-01 -8.95460188e-01 2.69071758e-01
-3.77945483e-01 4.44441855e-01 -4.05322313e-01 5.01233041e-01
1.14710011e-01 -1.15612161e+00 2.61588037e-01 5.51251434e-02
5.10009825e-01 5.93136489e-01 -5.73311448e-01 1.44292068e+00
-4.55048352e-01 1.06107068e+00 -4.28916782e-01 -5.37273765e-01
6.76004648e-01 3.05234402e-01 -1.78300172e-01 -2.72316873e-01
-5.71743958e-02 1.14898875e-01 3.32990587e-01 -4.65533920e-02
6.26439869e-01 -2.01105941e-02 1.45133764e-01 8.98591876e-01
5.87296374e-02 1.45727685e-02 -7.61734024e-02 8.25189829e-01
1.24754918e+00 -1.87885106e-01 5.66726983e-01 -2.03977630e-01
7.44982138e-02 1.04065679e-01 2.35062867e-01 1.06489646e+00
1.78117633e-01 6.54289842e-01 1.06667864e+00 -6.08039021e-01
-1.00924981e+00 -9.77699757e-01 2.96865225e-01 7.98399746e-01
3.86260420e-01 -4.47298884e-01 -5.41198909e-01 -1.12911355e+00
3.62534553e-01 1.13042188e+00 -1.07891846e+00 -3.62634182e-01
-1.12993516e-01 -6.89119026e-02 1.41525596e-01 6.79985404e-01
1.73013732e-01 -1.12237072e+00 3.37459780e-02 -1.55285761e-01
-1.88223138e-01 -1.03203237e+00 -7.11361945e-01 2.22848937e-01
-7.95065343e-01 -1.44669676e+00 1.50275618e-01 -3.92956585e-01
1.24105859e+00 5.34888327e-01 1.67876565e+00 1.13006413e+00
3.82342666e-01 5.95936060e-01 -5.47806025e-02 -2.46559069e-01
-3.59200716e-01 3.05482388e-01 -2.19816118e-01 -3.75599504e-01
3.07110071e-01 -6.79780066e-01 -4.14810419e-01 2.70827383e-01
-5.50374269e-01 5.12974143e-01 5.42288780e-01 9.91785884e-01
4.20454681e-01 -5.55009365e-01 6.80452585e-01 -1.54613316e+00
9.18679118e-01 -8.69325519e-01 -2.68698305e-01 1.79152131e-01
-1.22528553e+00 3.80516171e-01 6.58247590e-01 -3.60635370e-01
-7.11899400e-01 -1.33965805e-01 3.48217487e-01 -3.63682330e-01
5.53073063e-02 4.62225765e-01 5.66714704e-02 -6.32851645e-02
7.64402986e-01 -2.87203997e-01 -1.29659504e-01 1.33275807e-01
8.14135730e-01 -3.15292389e-03 5.45245826e-01 -2.57710814e-01
1.07244766e+00 1.82168067e-01 1.59576744e-01 -9.38995704e-02
-1.16171074e+00 2.30575446e-02 -5.54262400e-01 -2.10373312e-01
8.18974793e-01 -7.96290576e-01 -1.05016410e+00 -4.54840422e-01
-1.54736423e+00 -4.22658980e-01 -2.04092145e-01 4.49144632e-01
-3.75397414e-01 -5.11682518e-02 -5.34788787e-01 -5.25193989e-01
-7.59640560e-02 -1.14286411e+00 1.04945469e+00 8.79035965e-02
-8.37558508e-01 -1.65436602e+00 -2.26750746e-01 4.70714599e-01
4.01028305e-01 2.54103601e-01 1.00507498e+00 -9.71308768e-01
-1.14002931e+00 -9.99978632e-02 -5.58490694e-01 -5.53143680e-01
1.83820426e-02 -8.19104463e-02 -7.84590423e-01 1.27463132e-01
-6.94784820e-01 -3.84040982e-01 1.11621857e+00 4.07152921e-01
1.28751028e+00 -6.11665547e-01 -5.81643999e-01 5.70199072e-01
1.03579700e+00 -5.90427339e-01 4.41119105e-01 1.78791255e-01
1.12631524e+00 3.01307052e-01 2.54725575e-01 1.93800982e-02
7.83384442e-01 2.74797499e-01 1.19626367e+00 -5.69782078e-01
-1.95231348e-01 -7.64048755e-01 2.91010737e-02 3.11920553e-01
-1.69698015e-01 -8.82802367e-01 -7.33895779e-01 5.53347647e-01
-2.16187406e+00 -9.98274624e-01 -8.82581234e-01 1.77186465e+00
2.11405724e-01 2.78456420e-01 -3.11420292e-01 -5.46946004e-02
6.30079508e-01 2.28100941e-01 -9.51728046e-01 -5.83493590e-01
-4.85121794e-02 -1.27336577e-01 5.15383780e-01 1.05654871e+00
-6.02441609e-01 8.97479057e-01 7.01325941e+00 -5.95266782e-02
-5.39981663e-01 -1.90864801e-02 8.04534614e-01 1.39156461e-01
-1.54578388e+00 3.50663334e-01 -1.24415390e-01 -5.02427965e-02
7.24108994e-01 -2.15651125e-01 8.03894877e-01 9.48311567e-01
-1.49726242e-01 3.31194252e-01 -1.51839840e+00 6.15224361e-01
2.30204333e-02 -1.75072265e+00 2.66081691e-01 2.29767621e-01
8.67795825e-01 -7.91401714e-02 6.52061328e-02 2.78366476e-01
8.49477351e-01 -1.56153226e+00 6.83534563e-01 2.94997126e-01
8.04111362e-01 -4.41196293e-01 7.15361416e-01 1.14347145e-01
-1.03068948e+00 -2.09717453e-02 -2.69950211e-01 -5.24167538e-01
1.96297035e-01 5.69118381e-01 -1.13701367e+00 4.77825552e-01
5.50160110e-02 1.10991883e+00 -6.04727924e-01 3.14920932e-01
-1.16031981e+00 7.00037897e-01 2.01589957e-01 -4.67952080e-02
8.40157941e-02 -4.89952154e-02 3.69923890e-01 7.59287000e-01
1.51125297e-01 3.13870162e-01 -1.49811029e-01 1.44587207e+00
-6.34949386e-01 -3.68759245e-01 -9.57224786e-01 -2.61905730e-01
5.12784898e-01 1.21546173e+00 -5.82427144e-01 -2.62678593e-01
-3.78831387e-01 1.13364875e+00 8.31920028e-01 6.65511191e-01
-1.12376940e+00 1.20403782e-01 7.32836902e-01 1.79301277e-02
2.05041915e-02 -7.33840540e-02 -7.63387561e-01 -1.28407931e+00
-2.34045550e-01 -6.32048607e-01 3.22138876e-01 -1.26989806e+00
-1.34684145e+00 7.00006604e-01 -3.69343936e-01 -6.11106694e-01
-3.92407745e-01 -5.17888963e-01 -1.00642610e+00 8.29307675e-01
-1.55631685e+00 -1.33941591e+00 -5.91224849e-01 3.85684162e-01
3.62779349e-01 -1.31291315e-01 8.13885033e-01 -2.62540311e-01
-4.26998407e-01 5.95504165e-01 -7.18662679e-01 -1.12235107e-01
4.44370359e-01 -1.68209171e+00 1.04746199e+00 7.86263466e-01
5.21258771e-01 9.30761755e-01 8.89137805e-01 -8.28099847e-01
-1.39132679e+00 -1.15171921e+00 1.17662215e+00 -8.82563889e-01
7.40345359e-01 -1.91289335e-01 -1.06742823e+00 1.57001841e+00
3.99789661e-01 1.10585496e-01 5.63533008e-01 8.33494246e-01
-6.69129014e-01 4.63019639e-01 -8.61428499e-01 7.88395107e-01
1.67742109e+00 -5.20011067e-01 -4.59093094e-01 6.54631853e-01
1.30565131e+00 -4.79119867e-01 -5.94919622e-01 6.26788437e-02
4.27393585e-01 -1.05695629e+00 8.07669759e-01 -1.11779726e+00
1.18830502e+00 -1.34081587e-01 9.52669159e-02 -1.64441788e+00
-7.28492916e-01 -5.41657448e-01 -1.65122971e-01 8.76143813e-01
1.16250575e+00 -9.36878562e-01 7.75571108e-01 1.02859950e+00
-3.80157024e-01 -5.82366049e-01 -4.99991387e-01 -2.93556303e-01
-3.53459984e-01 -3.64406765e-01 1.06595898e+00 1.21399498e+00
2.90263474e-01 8.56469035e-01 -4.58640695e-01 4.01645184e-01
7.36013830e-01 2.27295682e-01 1.01094794e+00 -1.41531456e+00
-5.11756539e-01 -3.60936344e-01 -7.81808197e-02 -1.03258646e+00
8.51089180e-01 -1.22849071e+00 -2.32038677e-01 -2.07889700e+00
4.47153717e-01 -1.77203417e-01 7.64840245e-02 9.07436430e-01
-5.86142480e-01 2.27196515e-01 -1.09726779e-01 1.80221796e-01
-6.81754589e-01 5.57813287e-01 1.53428042e+00 -1.25943571e-01
2.12408915e-01 -1.98722079e-01 -1.40191841e+00 5.97982824e-01
8.53935063e-01 -6.60519838e-01 -8.76362860e-01 -7.40308166e-01
5.29513597e-01 2.32088342e-01 6.02393925e-01 -2.59579092e-01
1.08191468e-01 -3.42637658e-01 3.43689591e-01 3.21405120e-02
2.06101567e-01 -6.79550529e-01 4.45499897e-01 3.39891315e-01
-8.46149802e-01 4.48762119e-01 -4.16754857e-02 9.90140438e-01
5.26916981e-02 2.09901571e-01 7.74483308e-02 -1.39797227e-02
-3.40197891e-01 4.72630024e-01 4.15781960e-02 1.37313930e-02
6.10996902e-01 -2.24525943e-01 -1.00197077e+00 -1.03757346e+00
-5.14118731e-01 5.11424482e-01 6.52747035e-01 4.71464306e-01
6.92975700e-01 -1.47250414e+00 -5.26703537e-01 1.00910492e-01
3.44220579e-01 -1.15311772e-01 -1.29213512e-01 7.04968810e-01
-1.27960101e-01 4.46660131e-01 4.43034340e-03 -4.22176659e-01
-8.66387606e-01 6.10512912e-01 3.63349348e-01 -2.65502632e-01
-5.34595966e-01 5.82359970e-01 5.47269762e-01 -6.71042442e-01
-1.00443669e-01 -4.59140360e-01 -9.00766701e-02 -8.87787759e-01
1.76209316e-01 2.05308069e-02 -5.00633955e-01 -2.47804448e-01
-2.09657431e-01 1.81920573e-01 1.93062931e-01 4.79108244e-02
1.32799113e+00 -3.88822965e-02 -8.66402835e-02 3.11051488e-01
6.55539393e-01 -1.49243884e-02 -1.23110926e+00 1.27259403e-01
-2.11320028e-01 -5.27368128e-01 -2.01439232e-01 -8.88359368e-01
-1.11672938e+00 7.25575984e-01 -6.27956450e-01 5.18468022e-01
5.48062205e-01 3.23695719e-01 2.71882772e-01 4.11997020e-01
-1.19492309e-02 -2.02063084e-01 3.05783838e-01 2.07374364e-01
1.27955294e+00 -1.68741453e+00 1.44118756e-01 -7.86785901e-01
-8.93581629e-01 1.01375687e+00 9.17334676e-01 -1.56063735e-01
1.86472073e-01 -2.25857109e-01 -2.21855551e-01 -7.06313014e-01
-1.30097806e+00 7.17722028e-02 6.84526324e-01 6.90459251e-01
6.88380122e-01 2.43697107e-01 1.07655011e-01 7.60668933e-01
-3.20246875e-01 -4.14768755e-01 8.19143355e-01 -1.82741910e-01
-2.06004262e-01 -8.50643694e-01 -6.02458045e-02 6.73470497e-01
5.72201423e-03 -3.84414524e-01 -9.45562899e-01 1.19913626e+00
-4.91167188e-01 1.11581767e+00 1.78739382e-03 -3.96549195e-01
1.54032558e-01 -1.98619440e-01 1.41665295e-01 -7.12576747e-01
-6.82039380e-01 -5.51316559e-01 5.82853496e-01 -8.14692259e-01
-1.22798838e-01 -2.79092401e-01 -1.54595912e+00 -6.73769176e-01
-4.82453048e-01 2.65468985e-01 3.47075850e-01 1.05956757e+00
5.12677789e-01 9.61734116e-01 3.25375378e-01 -6.49986088e-01
-3.21781397e-01 -8.20147276e-01 -3.36824924e-01 6.82070494e-01
5.30896723e-01 -6.51511967e-01 -7.74287581e-01 -8.08195919e-02] | [8.3828706741333, 6.169857501983643] |
cea9e5de-a322-43b5-9a6e-67e88a08caed | clozer-adaptable-data-augmentation-for-cloze-1 | null | null | https://aclanthology.org/2022.repl4nlp-1.7 | https://aclanthology.org/2022.repl4nlp-1.7.pdf | Clozer”:" Adaptable Data Augmentation for Cloze-style Reading Comprehension | Task-adaptive pre-training (TAPT) alleviates the lack of labelled data and provides performance lift by adapting unlabelled data to downstream task. Unfortunately, existing adaptations mainly involve deterministic rules that cannot generalize well. Here, we propose Clozer, a sequence-tagging based cloze answer extraction method used in TAPT that is extendable for adaptation on any cloze-style machine reading comprehension (MRC) downstream tasks. We experiment on multiple-choice cloze-style MRC tasks, and show that Clozer performs significantly better compared to the oracle and state-of-the-art in escalating TAPT effectiveness in lifting model performance, and prove that Clozer is able to recognize the gold answers independently of any heuristics. | ['Pascale Fung', 'Dan Su', 'Samuel Cahyawijaya', 'Zeng Min', 'Willy Chung', 'Bryan Wilie', 'Holy Lovenia'] | null | null | null | null | repl4nlp-acl-2022-5 | ['machine-reading-comprehension'] | ['natural-language-processing'] | [ 5.87064147e-01 2.40379289e-01 5.98393977e-02 -4.23867732e-01
-1.42725825e+00 -8.49959195e-01 9.86972004e-02 2.32906029e-01
-6.20838165e-01 8.26276183e-01 2.22752139e-01 -1.03936946e+00
-2.81944156e-01 -6.66307628e-01 -7.92203248e-01 -2.64332473e-01
1.78670242e-01 8.08337212e-01 5.15163481e-01 -4.06778336e-01
2.84859180e-01 -8.18578154e-02 -1.33701694e+00 9.70412076e-01
1.31720340e+00 6.68796659e-01 3.53759378e-01 1.39994252e+00
8.00404605e-03 1.20201421e+00 -6.02820933e-01 -6.86285198e-01
7.40988832e-03 -5.31527221e-01 -1.48356009e+00 -4.77045685e-01
7.33698130e-01 -4.60131615e-02 -4.69652899e-02 5.75967848e-01
5.86849451e-01 3.81933123e-01 6.65877879e-01 -8.44426155e-01
-1.14703870e+00 9.48180020e-01 2.50850677e-01 4.49991882e-01
4.78319108e-01 4.06561136e-01 1.10362494e+00 -1.01459086e+00
5.59192538e-01 1.11692357e+00 8.61225247e-01 1.18558311e+00
-1.08274472e+00 -1.47692859e-01 2.64914095e-01 6.49772942e-01
-9.05917406e-01 -4.44302320e-01 1.03803612e-01 -1.10862345e-01
1.75577378e+00 6.31994128e-01 -5.87406419e-02 1.41158843e+00
-3.73947561e-01 1.18765521e+00 1.51334107e+00 -9.19105530e-01
1.83898330e-01 -3.29313844e-01 5.16180933e-01 6.91395342e-01
-1.81361079e-01 -1.02584817e-01 -4.95097280e-01 -1.12206295e-01
-3.16325799e-02 -5.56677759e-01 -4.62253720e-01 -1.66785285e-01
-1.13226521e+00 8.60440016e-01 1.16091497e-01 1.11535624e-01
-8.81966352e-02 -1.17475070e-01 5.15474677e-01 7.58395433e-01
3.66490126e-01 1.23001182e+00 -1.16727257e+00 -4.32499200e-01
-5.29422462e-01 4.45557564e-01 1.02830625e+00 1.31510961e+00
1.67301714e-01 -2.70620078e-01 -9.16827321e-01 8.46672535e-01
-3.17332447e-01 6.39101267e-01 8.33086431e-01 -1.27785754e+00
9.79270935e-01 3.92351449e-01 -4.01374474e-02 -2.23671347e-01
-5.94431877e-01 -5.17582238e-01 -5.04273415e-01 -2.65688539e-01
6.94164872e-01 -1.86772779e-01 -1.13010693e+00 1.89206016e+00
7.87314866e-03 -2.21528023e-01 3.32993627e-01 5.92414856e-01
8.71041298e-01 4.59754467e-01 4.34146672e-01 -2.13848576e-01
1.22238350e+00 -1.55605423e+00 -5.45787215e-01 -6.75119102e-01
1.24416542e+00 -3.09748173e-01 1.89278066e+00 7.04014122e-01
-1.21912193e+00 -5.21126926e-01 -9.54058647e-01 -4.41082269e-01
-4.16113347e-01 -2.88817108e-01 6.62639380e-01 6.49777353e-01
-1.07151675e+00 3.98728102e-01 -3.32162976e-01 -2.39955276e-01
2.82403231e-01 1.74375460e-01 -2.62974679e-01 -7.20434666e-01
-1.49472845e+00 1.38302398e+00 6.28291905e-01 -2.58765548e-01
-1.01177299e+00 -6.80470943e-01 -6.80908322e-01 1.97594970e-01
7.74634659e-01 -9.98244107e-01 1.95200038e+00 -5.28628409e-01
-1.67803705e+00 8.12035143e-01 -4.15354282e-01 -5.50663590e-01
5.25808156e-01 -5.23948014e-01 -1.90503359e-01 -3.01048718e-03
1.49534717e-01 6.19157493e-01 6.42028391e-01 -7.80434132e-01
-7.61158407e-01 -8.69135186e-02 2.00445250e-01 3.04888994e-01
-2.07420349e-01 -7.60539919e-02 -1.35330349e-01 -2.49837458e-01
-5.66438586e-02 -6.04972661e-01 -1.28824800e-01 -1.00781000e+00
-2.67607599e-01 -8.35701108e-01 9.92692336e-02 -6.37836933e-01
1.61443746e+00 -1.61140966e+00 3.60836267e-01 -1.19246319e-01
3.68194014e-01 5.76563001e-01 -7.57506609e-01 3.17820668e-01
1.08402133e-01 3.54424417e-01 -3.67995143e-01 -1.93220004e-01
2.52679810e-02 3.15018386e-01 -1.44256026e-01 -2.01904103e-01
3.28478724e-01 1.31550336e+00 -1.35806811e+00 -2.16645196e-01
-3.01669419e-01 -5.43630421e-01 -6.85630977e-01 6.40245259e-01
-8.19061100e-01 3.69526535e-01 -3.20455819e-01 6.25318646e-01
3.82351786e-01 -3.01907450e-01 1.14300802e-01 7.31266618e-01
3.43055040e-01 8.46963048e-01 -2.97206044e-01 1.79049540e+00
-7.37812996e-01 2.73154795e-01 -3.07684422e-01 -6.20836973e-01
5.20730197e-01 3.27988893e-01 -5.43570578e-01 -7.90405869e-01
1.65938661e-01 5.42316496e-01 2.77371556e-01 -1.11628377e+00
2.73974180e-01 -2.79091783e-02 -1.74595982e-01 3.54648083e-01
2.12702498e-01 -2.57619321e-01 3.53572339e-01 2.46104419e-01
1.74555993e+00 1.30553827e-01 4.34382319e-01 -6.03055619e-02
6.99460268e-01 4.20359194e-01 1.79540634e-01 1.40943742e+00
-2.63911635e-01 4.69955534e-01 2.25336701e-01 -7.13388324e-02
-9.54407573e-01 -8.50950062e-01 1.11739442e-01 2.11265421e+00
-3.68439198e-01 -4.12615895e-01 -1.17174244e+00 -1.24554408e+00
-1.27222389e-01 1.30447400e+00 -5.68983436e-01 -3.20849955e-01
-8.34212959e-01 -6.03075206e-01 9.55116689e-01 7.93456435e-01
4.90940571e-01 -1.33997202e+00 -5.77482879e-01 3.81315917e-01
-6.14410400e-01 -1.18931746e+00 -4.17342067e-01 6.56144500e-01
-9.74592865e-01 -1.02237761e+00 -3.24398309e-01 -8.05567980e-01
3.33171219e-01 -3.35006788e-02 1.64957833e+00 3.01881701e-01
3.63352388e-01 1.99196115e-01 -8.48858595e-01 -1.62712455e-01
-7.88828373e-01 9.73225594e-01 -2.42111310e-01 -6.57690287e-01
1.05917645e+00 -2.90637851e-01 -8.50526690e-02 6.37602136e-02
-5.55998564e-01 6.88938648e-02 7.46435761e-01 1.16033626e+00
2.92237580e-01 -5.23113430e-01 9.76945519e-01 -1.57266045e+00
1.21833801e+00 -5.98348558e-01 6.55731633e-02 6.21902466e-01
-8.01060140e-01 2.39675477e-01 8.40751052e-01 -5.98055601e-01
-1.16955221e+00 -1.21814750e-01 -4.37587112e-01 -3.73625755e-03
-5.58003724e-01 5.63739955e-01 -3.02921921e-01 2.19594955e-01
1.30981386e+00 -1.62811540e-02 -5.90973914e-01 -7.66085744e-01
8.80938351e-01 9.57384169e-01 1.01031291e+00 -9.58769917e-01
4.37665880e-01 -3.16712290e-01 -3.61155123e-01 -1.39782056e-01
-1.62308526e+00 -6.09157681e-01 -7.07212448e-01 4.32561040e-01
8.91253114e-01 -5.40299356e-01 -6.45250440e-01 4.40199822e-01
-1.27456641e+00 -9.34517086e-01 -1.16739340e-01 -7.08163381e-02
-7.42964983e-01 6.68787956e-02 -8.69568586e-01 -6.24539256e-01
-4.09581065e-01 -8.05291653e-01 6.67407990e-01 -2.61670146e-02
-5.30916989e-01 -1.05260158e+00 1.55100469e-02 8.05769861e-01
6.39810205e-01 -2.19788522e-01 1.60029757e+00 -1.17699349e+00
-4.36985075e-01 -5.08486107e-03 3.42633352e-02 5.31614125e-01
-5.11327267e-01 -7.82710850e-01 -1.23157799e+00 -1.72724679e-01
1.28579661e-01 -9.84034538e-01 1.02742958e+00 -1.50051024e-02
1.25909162e+00 -4.74667072e-01 -4.57977504e-02 6.87045395e-01
9.55011845e-01 -2.59543154e-02 5.56758165e-01 7.81653047e-01
7.70876884e-01 5.55869937e-01 5.62336445e-01 -3.31546634e-01
5.26347756e-01 3.91605675e-01 2.69102991e-01 4.83673662e-01
-2.87384063e-01 -6.89116001e-01 3.30065489e-01 1.01790738e+00
7.00187907e-02 -6.70033634e-01 -1.13905954e+00 7.02861309e-01
-1.82485592e+00 -6.42501652e-01 -5.97718716e-01 1.84556699e+00
1.34199381e+00 1.89410344e-01 -1.73472077e-01 2.07070500e-01
4.26583737e-01 -4.24170971e-01 -5.69208503e-01 -9.88065481e-01
-2.50009537e-01 7.71288514e-01 4.93342936e-01 7.90147483e-01
-1.13903129e+00 1.47178519e+00 6.87036896e+00 1.04935718e+00
-4.37254518e-01 8.84486854e-01 3.65382940e-01 1.83207318e-01
-2.58541286e-01 -1.57798216e-01 -7.23229051e-01 2.83541381e-02
1.18359149e+00 2.54296064e-01 5.38928568e-01 9.91462588e-01
-3.86314034e-01 8.84414613e-02 -1.47817516e+00 4.37074721e-01
2.22882345e-01 -9.15939629e-01 4.65532672e-03 -7.18598723e-01
9.75657701e-01 8.50679055e-02 1.57300249e-01 1.10870969e+00
4.86920029e-01 -1.52087295e+00 6.85108900e-01 1.95178077e-01
9.11241293e-01 -4.89644140e-01 9.15057600e-01 8.70111823e-01
-3.46332669e-01 -4.10872072e-01 -4.06783313e-01 -5.57908356e-01
-1.24728475e-02 6.84275404e-02 -1.19646561e+00 1.57601267e-01
5.35307586e-01 4.95678559e-02 -1.13224626e+00 9.49362040e-01
-1.11797976e+00 1.14643633e+00 3.46351415e-02 -3.27596188e-01
4.37990546e-01 6.28279448e-01 4.44519758e-01 1.40427196e+00
-1.22492295e-03 3.43937129e-01 8.19072053e-02 5.45404017e-01
-2.89747864e-01 5.31253666e-02 -4.14035320e-01 1.79705054e-01
7.59850264e-01 7.48925209e-01 3.50268465e-03 -5.17026544e-01
-1.26222461e-01 1.36052227e+00 1.00861073e+00 3.56860459e-01
-4.58245397e-01 -3.90567690e-01 1.17566288e-01 -6.12463430e-02
2.95278460e-01 4.45363717e-03 -6.82085037e-01 -1.10247660e+00
3.94507684e-02 -1.56369090e+00 6.99890733e-01 -7.83288360e-01
-1.78958547e+00 8.01464736e-01 1.19252848e-02 -6.41146779e-01
-5.06515801e-01 -8.87604058e-01 -3.55712295e-01 9.09786344e-01
-1.51438403e+00 -1.19056308e+00 -2.40167752e-01 3.80141735e-01
9.81159329e-01 -1.26562454e-02 9.88058329e-01 -1.16584480e-01
-1.89266130e-01 1.13311374e+00 -6.60562664e-02 -1.07258618e-01
7.15872467e-01 -1.81787109e+00 7.81662166e-01 9.65658665e-01
-1.01922862e-01 8.21781933e-01 5.84871531e-01 -5.71938396e-01
-9.65009511e-01 -1.07267845e+00 1.38886321e+00 -1.47115886e+00
6.06236994e-01 -3.99964124e-01 -1.20596123e+00 9.35605109e-01
1.30905643e-01 -4.22174186e-01 4.91166711e-01 4.87120807e-01
-6.60611928e-01 4.25965190e-01 -9.78337169e-01 5.56741297e-01
1.52696824e+00 -5.72682858e-01 -1.30025506e+00 4.99131083e-01
1.11627805e+00 -8.62253845e-01 -5.81292033e-01 6.21521950e-01
-1.84419211e-02 -8.16304922e-01 7.00775683e-01 -1.40590835e+00
5.09767354e-01 7.04641193e-02 -3.82609963e-02 -1.59028602e+00
-6.38884604e-01 -6.18794799e-01 -4.19707537e-01 1.03189158e+00
9.26048875e-01 -5.37487686e-01 4.52245802e-01 5.16050816e-01
-5.30265868e-01 -7.95073330e-01 -9.98161435e-01 -8.43375266e-01
8.05843472e-01 -6.25060201e-01 5.52627683e-01 7.45323241e-01
2.77717024e-01 1.16534078e+00 -1.22858278e-01 1.38064781e-02
2.31580019e-01 -2.54351795e-01 4.10872936e-01 -9.40363705e-01
-6.71098650e-01 -4.74118531e-01 2.78013021e-01 -1.50874162e+00
2.77696818e-01 -1.09764671e+00 4.51275438e-01 -1.35001647e+00
2.43936330e-01 -6.11511707e-01 -1.18128650e-01 8.63450050e-01
-9.42967713e-01 -3.31826396e-02 7.29219690e-02 -8.71086866e-02
-1.11641765e+00 5.53225242e-02 1.29655802e+00 -1.00944206e-01
-1.28094926e-01 6.17537312e-02 -8.17878902e-01 5.62679410e-01
8.49611521e-01 -6.49586260e-01 -6.79130375e-01 -9.56228197e-01
5.82435310e-01 -1.25548467e-01 -5.67320883e-02 -8.41795802e-01
1.50615931e-01 5.72926104e-02 -1.28685478e-02 5.95564060e-02
-3.13105553e-01 -1.99410230e-01 -7.82012343e-01 1.65015340e-01
-9.26379323e-01 2.15273082e-01 1.17754370e-01 4.90694106e-01
1.65233284e-01 -8.71157587e-01 4.93027598e-01 -4.53242123e-01
-7.95785725e-01 -1.33082882e-01 -4.55416650e-01 9.57470238e-01
4.79802102e-01 6.12763725e-02 -8.64960730e-01 -4.20930356e-01
-7.02840984e-01 5.71214020e-01 2.65408397e-01 5.37673533e-01
4.74219978e-01 -8.13503742e-01 -1.06371188e+00 -1.61052588e-02
5.25539577e-01 2.10533008e-01 1.73896119e-01 8.49761248e-01
-5.38535118e-01 8.07729900e-01 2.70973355e-01 -3.63277107e-01
-1.17595112e+00 8.73930156e-01 4.35268760e-01 -6.73349440e-01
-4.17639941e-01 1.44633710e+00 -1.96515813e-01 -8.06634545e-01
2.76492208e-01 -5.39126635e-01 -2.72145092e-01 -4.60459948e-01
5.41774273e-01 4.58713830e-01 6.12996101e-01 7.57612139e-02
-1.69737816e-01 1.36024907e-01 -1.42685026e-01 -1.83744933e-02
9.48547065e-01 -1.27112851e-01 4.29357775e-02 2.43168503e-01
7.60891736e-01 -1.76969226e-02 -9.75997210e-01 -2.32818604e-01
6.51738942e-01 -1.79158151e-01 -3.84229094e-01 -1.64632916e+00
-1.00715846e-01 1.16520798e+00 7.17792064e-02 2.76054107e-02
1.07338631e+00 1.02589369e-01 1.06429672e+00 9.78309810e-01
4.17666733e-01 -8.61705184e-01 -1.00901565e-02 1.20981252e+00
6.24781370e-01 -1.07245135e+00 -6.67307138e-01 -4.62653637e-01
-8.41786027e-01 8.16604435e-01 1.14750552e+00 2.41583437e-01
-1.71486169e-01 1.16786687e-03 2.53076851e-01 -6.98893815e-02
-1.27771831e+00 -4.71331716e-01 3.94222319e-01 1.05892885e+00
4.21230942e-01 1.43162623e-01 -2.29768217e-01 9.45103407e-01
-3.56870085e-01 -1.03263415e-01 4.12565172e-01 9.73735690e-01
-5.97253323e-01 -1.14938867e+00 -1.16696991e-01 4.05309767e-01
-4.40888166e-01 -7.91435778e-01 -6.75774813e-01 6.60632670e-01
4.35352981e-01 1.28756511e+00 -4.36733097e-01 -4.30913150e-01
5.31111360e-01 8.20937932e-01 8.78004849e-01 -1.12926650e+00
-9.45287764e-01 -6.64199114e-01 5.88127315e-01 -3.95062506e-01
1.45086884e-01 -2.78793752e-01 -1.13319850e+00 -1.34178519e-01
-3.68061781e-01 4.45835412e-01 -1.32957742e-01 1.30362868e+00
2.50647992e-01 5.24085999e-01 1.18156988e-02 -7.94316903e-02
-1.16467249e+00 -1.61906850e+00 1.25519961e-01 5.71860671e-01
2.88767010e-01 -3.35859239e-01 -4.85980213e-01 -3.55597176e-02] | [11.263666152954102, 8.114883422851562] |
beb68157-c214-4523-82da-94559e4cecb8 | odd-one-out-representation-learning | 2012.07966 | null | https://arxiv.org/abs/2012.07966v1 | https://arxiv.org/pdf/2012.07966v1.pdf | Odd-One-Out Representation Learning | The effective application of representation learning to real-world problems requires both techniques for learning useful representations, and also robust ways to evaluate properties of representations. Recent work in disentangled representation learning has shown that unsupervised representation learning approaches rely on fully supervised disentanglement metrics, which assume access to labels for ground-truth factors of variation. In many real-world cases ground-truth factors are expensive to collect, or difficult to model, such as for perception. Here we empirically show that a weakly-supervised downstream task based on odd-one-out observations is suitable for model selection by observing high correlation on a difficult downstream abstract visual reasoning task. We also show that a bespoke metric-learning VAE model which performs highly on this task also out-performs other standard unsupervised and a weakly-supervised disentanglement model across several metrics. | ['Bjørn Sand Jensen', 'Anders Kirk Uhrenholt', 'Salman Mohammadi'] | 2020-12-14 | null | null | null | null | ['odd-one-out'] | ['reasoning'] | [ 5.06474912e-01 2.76059002e-01 -4.57127601e-01 -4.73014534e-01
-1.12372327e+00 -6.33969069e-01 1.00210488e+00 2.94825613e-01
-2.02706650e-01 7.93843210e-01 5.50168574e-01 -3.53386909e-01
-6.33339643e-01 -5.88382900e-01 -4.25506055e-01 -6.69816017e-01
-1.18519783e-01 7.52693534e-01 -4.12382662e-01 -2.09174797e-01
2.64387399e-01 3.46980184e-01 -1.58879232e+00 3.21324885e-01
3.89296770e-01 6.13393366e-01 -2.45581672e-01 7.97556221e-01
3.16956908e-01 9.31799531e-01 -3.58724117e-01 -3.21301758e-01
3.93346071e-01 -5.63939571e-01 -6.05199575e-01 -1.93135500e-01
6.00899875e-01 -1.39082134e-01 -2.87351727e-01 9.40502167e-01
2.25515381e-01 3.25223893e-01 1.44386518e+00 -1.38789630e+00
-8.20060909e-01 8.07959199e-01 -4.76090670e-01 3.50151837e-01
3.60026658e-01 1.56392679e-01 1.80923355e+00 -7.24175930e-01
4.19793785e-01 1.35329235e+00 3.78802896e-01 5.07561028e-01
-2.09398198e+00 -7.68115938e-01 -9.58293024e-03 8.67967308e-02
-9.98644352e-01 -6.88974917e-01 7.07329571e-01 -8.93080413e-01
9.82548296e-01 4.24113035e-01 2.84450322e-01 1.50724542e+00
-1.21343791e-01 5.18303454e-01 1.42611396e+00 -4.65405166e-01
2.93497860e-01 1.22991968e-02 4.04319048e-01 7.97547460e-01
9.27579105e-01 6.19440317e-01 -6.99572027e-01 -2.38009766e-01
8.48317325e-01 -1.28533706e-01 -2.44441584e-01 -1.01717043e+00
-1.49659872e+00 1.01288462e+00 3.77407193e-01 1.65268376e-01
-3.20134126e-02 4.54432338e-01 3.13799709e-01 6.31872416e-01
2.86500901e-01 1.28576815e+00 -6.16247416e-01 -1.59624130e-01
-8.20008576e-01 2.99366027e-01 5.63873231e-01 8.70575666e-01
9.49602723e-01 2.30951399e-01 -2.16677204e-01 5.26513934e-01
4.30273890e-01 2.04984710e-01 4.83254105e-01 -9.57005322e-01
5.31911254e-01 5.16114712e-01 9.15445089e-02 -8.90108705e-01
-3.62019777e-01 -2.53732234e-01 -7.40621328e-01 5.59096038e-01
7.41658866e-01 9.01321247e-02 -1.04825258e+00 2.00897217e+00
-4.32656944e-01 -7.44850338e-02 6.08617105e-02 7.79350877e-01
4.97830361e-01 1.67600363e-01 8.06550011e-02 -2.33424410e-01
1.33091295e+00 -5.08644879e-01 -5.10863721e-01 -2.56046325e-01
8.67934585e-01 -3.06234717e-01 1.13542604e+00 6.06120110e-01
-9.23294067e-01 -4.62910324e-01 -1.44447410e+00 -4.14951533e-01
-3.10572356e-01 2.60944292e-02 1.22440803e+00 7.31888056e-01
-7.76556611e-01 6.33765757e-01 -8.82777750e-01 -1.61219493e-01
6.10101640e-01 4.22933608e-01 -7.34407544e-01 -1.45454988e-01
-7.95478046e-01 1.17366111e+00 2.74778694e-01 -2.12035775e-01
-1.32064557e+00 -6.65067732e-01 -1.13309300e+00 2.10141242e-01
3.57506692e-01 -9.34605777e-01 1.02316296e+00 -7.76220024e-01
-1.15313077e+00 9.53657806e-01 6.24032915e-02 -2.75496483e-01
2.67162770e-01 -1.01464562e-01 -2.48537347e-01 -6.93910047e-02
5.74206747e-02 2.53133118e-01 9.62495923e-01 -1.41820037e+00
1.54820383e-01 -5.96457899e-01 2.36300603e-01 1.34086698e-01
-1.11661792e-01 -1.93424582e-01 5.28544962e-01 -6.08820915e-01
3.14770281e-01 -7.35094309e-01 -3.77164185e-01 9.39064771e-02
-3.77743751e-01 1.70175824e-02 4.99565482e-01 -3.13231558e-01
8.63943875e-01 -1.96683967e+00 5.56452036e-01 3.55616689e-01
6.90150261e-01 -2.47806057e-01 -2.53560930e-01 4.07228678e-01
-6.29841506e-01 2.79818684e-01 -2.39386708e-01 -3.26029360e-01
3.17137778e-01 3.09750050e-01 -5.20268321e-01 7.78613865e-01
3.12763900e-01 8.06008875e-01 -9.83334303e-01 -2.67548144e-01
1.51835114e-01 1.05752811e-01 -6.65168941e-01 3.32263559e-01
-1.40596405e-01 3.62270892e-01 -1.71160847e-01 2.31987283e-01
1.78295776e-01 -2.96000957e-01 2.59453356e-01 8.31052568e-03
3.98279935e-01 7.22654402e-01 -1.04204881e+00 1.91047060e+00
-5.18290937e-01 8.51362705e-01 -3.20272624e-01 -1.07589877e+00
8.56467128e-01 2.76790202e-01 1.09969772e-01 -3.78943950e-01
1.45202070e-01 9.40684378e-02 4.80982482e-01 -4.22934502e-01
3.02151322e-01 -5.19542933e-01 -2.38255128e-01 7.45267034e-01
4.10734147e-01 -4.26468343e-01 7.50763267e-02 4.31451172e-01
1.28788829e+00 1.00915730e-01 6.48365557e-01 -4.03676778e-01
-4.80915681e-02 -1.83763131e-01 5.10111451e-01 6.72818124e-01
-4.70555983e-02 7.66639948e-01 8.52402329e-01 -1.48103520e-01
-8.91350508e-01 -1.35353720e+00 -3.86484042e-02 1.11467516e+00
-1.42309278e-01 -7.74566710e-01 -1.47851795e-01 -5.80401063e-01
4.71123382e-02 1.02624142e+00 -8.51282895e-01 -5.17243922e-01
-4.42669317e-02 -7.19239771e-01 5.18991768e-01 8.78870666e-01
-4.51542079e-01 -7.87293375e-01 -4.46764350e-01 -1.59719944e-01
3.14073682e-01 -7.90099025e-01 1.01886176e-01 7.15659916e-01
-9.83537078e-01 -1.23781621e+00 -2.66151428e-01 -3.87671173e-01
7.45314658e-01 3.23605657e-01 1.30833459e+00 9.95161899e-05
-3.00071836e-01 2.35773504e-01 -8.50169957e-02 -3.42673570e-01
-3.72712642e-01 -1.60526216e-01 1.35453656e-01 -3.33509088e-01
6.19613588e-01 -8.68836522e-01 -1.98142499e-01 1.39829949e-01
-6.92378104e-01 1.11760184e-01 6.51742280e-01 1.06558263e+00
2.75467098e-01 -4.07366097e-01 5.19368589e-01 -1.04872811e+00
9.24023449e-01 -5.38046181e-01 -4.99594569e-01 1.05358504e-01
-8.34859788e-01 7.26707399e-01 3.57527405e-01 -4.06670958e-01
-6.88753009e-01 -1.13671593e-01 6.20135427e-01 -5.03584504e-01
-1.10143408e-01 5.56212783e-01 -3.63840938e-01 4.48721200e-01
1.29130781e+00 -3.26258600e-01 1.18169509e-01 -2.77637392e-01
7.90083230e-01 2.86941856e-01 1.55253455e-01 -9.78275836e-01
1.03526402e+00 2.80019671e-01 2.43406966e-01 -5.79637110e-01
-1.05859077e+00 -3.20669144e-01 -7.35655725e-01 4.50435549e-01
8.19729209e-01 -1.07588303e+00 -5.88419378e-01 -3.57939690e-01
-9.12640750e-01 -3.93564969e-01 -5.68890035e-01 8.61924469e-01
-1.02090371e+00 6.44561723e-02 -2.27726087e-01 -8.51811588e-01
3.20469648e-01 -1.07101071e+00 1.14292443e+00 -3.07567805e-01
-7.13980854e-01 -9.91784453e-01 4.43233520e-01 4.81064647e-01
2.13454336e-01 3.80987525e-01 1.31194544e+00 -8.56236577e-01
-5.62783957e-01 -7.31501505e-02 -3.28684062e-01 1.61795288e-01
1.49167195e-01 -2.09463298e-01 -1.29381645e+00 -2.97457933e-01
-1.70119435e-01 -8.56931090e-01 1.37599266e+00 -3.75715978e-02
1.24348652e+00 -1.63457930e-01 -1.54506385e-01 5.60776532e-01
1.20796084e+00 -3.30416650e-01 5.29771030e-01 -1.91429868e-01
9.57213819e-01 7.77076125e-01 1.48770556e-01 7.21453503e-02
2.64722049e-01 5.37992060e-01 1.79928720e-01 5.79549484e-02
1.76969748e-02 -4.60637689e-01 3.18982095e-01 6.91797316e-01
-4.21511710e-01 -5.79778589e-02 -7.76446164e-01 5.46546400e-01
-1.93231308e+00 -1.21981907e+00 -1.02488428e-01 2.28845382e+00
8.56724024e-01 3.28217417e-01 7.28158206e-02 4.27300394e-01
1.89550295e-01 4.42171007e-01 -4.72736925e-01 -4.15608793e-01
-1.04341395e-01 5.33335924e-01 2.85739720e-01 5.08913457e-01
-8.10689807e-01 6.79445088e-01 6.79227448e+00 3.59735191e-01
-4.13120627e-01 9.08788666e-02 2.67198473e-01 -1.19145349e-01
-9.43403184e-01 3.58667463e-01 -1.50143042e-01 -1.48414582e-01
1.00458086e+00 -1.33370176e-01 5.32983482e-01 7.19926596e-01
-8.88544694e-03 7.10911676e-02 -2.12212491e+00 1.27350664e+00
2.00614840e-01 -1.18052900e+00 2.70775497e-01 2.87609130e-01
7.49180794e-01 -1.70136437e-01 1.53493389e-01 4.16091025e-01
9.13338602e-01 -1.83694601e+00 4.96655673e-01 4.82841700e-01
9.64275539e-01 -5.49950302e-01 3.33401382e-01 1.74999446e-01
-8.39489937e-01 -4.31372151e-02 -4.68982041e-01 -4.20269638e-01
-3.19932550e-01 4.69430447e-01 -6.26433909e-01 2.09817305e-01
-3.65435444e-02 8.48237514e-01 -5.74868917e-01 7.06836998e-01
-7.93031812e-01 5.42799294e-01 9.54288095e-02 1.25655085e-01
-1.57625571e-01 -4.88929525e-02 5.04758179e-01 1.06170058e+00
4.72639166e-02 -7.86198825e-02 -8.38102549e-02 1.09513414e+00
-1.04658574e-01 -2.60316372e-01 -1.13635218e+00 -4.51131344e-01
1.72864616e-01 1.04261088e+00 -3.75181168e-01 -2.41762891e-01
-2.34281495e-01 5.62304378e-01 6.45064890e-01 5.28017104e-01
-3.89649510e-01 -2.20729828e-01 1.19637299e+00 -1.31432772e-01
1.67443618e-01 -5.78432620e-01 -5.38354993e-01 -1.67786837e+00
-2.19493717e-01 -1.04272795e+00 4.59613711e-01 -8.97009909e-01
-1.68863928e+00 2.27229446e-01 2.73596078e-01 -1.11999500e+00
-6.91323280e-01 -1.02268744e+00 -5.53050458e-01 7.80935049e-01
-1.24362421e+00 -1.02303708e+00 -1.31837919e-01 7.05465078e-01
2.72460639e-01 -4.01794732e-01 1.29485357e+00 -3.37105036e-01
-4.70910817e-01 5.35861731e-01 -1.96019232e-01 1.66163892e-01
5.73061585e-01 -1.60423648e+00 1.21480808e-01 7.27891624e-01
1.08715677e+00 1.04978955e+00 9.07023251e-01 -1.75135791e-01
-1.50983858e+00 -6.28886282e-01 5.98883808e-01 -1.02877092e+00
8.62500489e-01 -8.26092899e-01 -6.41507089e-01 1.12326252e+00
1.23979144e-01 2.07324743e-01 1.09684253e+00 8.38450432e-01
-1.42593813e+00 1.48830786e-01 -9.90068316e-01 6.40017211e-01
1.26385391e+00 -1.10469353e+00 -1.12077224e+00 3.18629116e-01
5.53870440e-01 -9.38249193e-03 -6.78326607e-01 5.36642410e-02
6.38901353e-01 -8.72493505e-01 9.16248560e-01 -1.37093961e+00
6.93662405e-01 -2.52994746e-01 -6.00940704e-01 -1.64858997e+00
-4.40935701e-01 -7.06196964e-01 -4.00270760e-01 8.25529933e-01
6.74775064e-01 -5.61926007e-01 5.45702159e-01 7.41958559e-01
2.24104419e-01 -3.76554191e-01 -5.08069694e-01 -7.53792405e-01
1.88705266e-01 -7.68389702e-01 3.90831709e-01 1.26509905e+00
2.21887603e-01 8.67867231e-01 -8.78992826e-02 1.08933128e-01
8.03769469e-01 2.44125277e-01 8.16992819e-01 -1.63618219e+00
-6.41353428e-01 -5.53121567e-01 -9.64164555e-01 -7.60211706e-01
4.05254841e-01 -1.27737057e+00 -1.65113002e-01 -1.44182456e+00
3.74053061e-01 -2.90107310e-01 -5.28803527e-01 4.05089796e-01
2.09431108e-02 1.11236326e-01 1.18584991e-01 1.37595370e-01
-3.38562399e-01 5.13062179e-01 1.08278191e+00 -3.44495803e-01
1.59529284e-01 -3.03101748e-01 -1.27539849e+00 7.43287444e-01
5.62864244e-01 -7.11026371e-01 -9.06471193e-01 -3.74572575e-01
6.48044705e-01 -7.66927004e-03 6.66861892e-01 -6.44902885e-01
-2.56517947e-01 -4.74725157e-01 5.30989289e-01 1.42335035e-02
5.89787841e-01 -6.44158721e-01 -1.79839715e-01 -7.79712340e-03
-8.64639640e-01 4.59332429e-02 -1.26142725e-01 7.64715433e-01
9.15284455e-03 -2.29407445e-01 4.32973951e-01 -2.45807588e-01
-4.55903620e-01 1.20422289e-01 -1.06364816e-01 4.40454692e-01
7.79687643e-01 -2.37735331e-01 -6.29270017e-01 -4.76908267e-01
-7.91319191e-01 -1.36948124e-01 3.39960486e-01 3.30840468e-01
5.88139594e-01 -1.26919484e+00 -8.38192761e-01 2.64406145e-01
6.10979915e-01 -2.07361177e-01 -4.72115725e-01 6.23134851e-01
1.09131476e-02 2.07744509e-01 -3.19367796e-01 -3.70304137e-01
-9.08901632e-01 6.13610625e-01 7.21657276e-02 -2.60217935e-01
-3.68594676e-01 7.93774545e-01 5.61763525e-01 -3.42404574e-01
-5.93438148e-02 -6.43489897e-01 -6.66365325e-02 2.87324905e-01
4.64601308e-01 2.03535616e-01 -5.54067492e-02 -6.04934692e-01
-2.85987258e-01 2.93836385e-01 -6.00194819e-02 -4.39146340e-01
1.43821561e+00 2.57847667e-01 4.22159582e-02 9.75848734e-01
1.37505126e+00 8.18669200e-02 -1.08134568e+00 -2.49131676e-02
2.10545167e-01 -6.86534524e-01 6.94365576e-02 -5.91462612e-01
-7.23370552e-01 1.09968531e+00 2.17416584e-01 2.10584193e-01
6.13623500e-01 6.82441443e-02 -1.18219040e-01 7.47904956e-01
5.73139906e-01 -5.45913279e-01 4.07987684e-01 1.83121666e-01
1.11402166e+00 -1.30510938e+00 4.53048170e-01 -2.28040323e-01
-6.20730758e-01 1.02864861e+00 5.21792591e-01 -4.46783781e-01
5.54844975e-01 2.72513907e-02 -1.83903426e-01 -5.79499185e-01
-1.09114659e+00 -3.58196229e-01 6.76256776e-01 8.88616681e-01
4.86378074e-01 4.30647790e-01 8.10455978e-02 4.96953666e-01
-5.77574432e-01 -5.74276805e-01 7.99668312e-01 5.62894523e-01
-8.30841362e-02 -1.02168310e+00 -9.87200160e-03 7.21558750e-01
-1.00772291e-01 -1.25012577e-01 -5.81530631e-01 9.38242555e-01
-1.80203795e-01 9.24561679e-01 1.02626599e-01 -3.82115126e-01
7.04551442e-03 9.81233492e-02 9.49142396e-01 -1.28095877e+00
-2.48671025e-01 -3.06645870e-01 4.02132154e-01 -6.09411299e-01
-5.95233500e-01 -5.96754313e-01 -9.88058686e-01 5.36061563e-02
-3.53654921e-01 -2.66378261e-02 4.37931508e-01 1.06615365e+00
1.48327276e-01 4.02103126e-01 2.68227339e-01 -7.77529776e-01
-8.29116464e-01 -1.09379470e+00 -7.29630530e-01 6.88908160e-01
6.47814929e-01 -1.12642324e+00 -7.06384182e-01 -1.75396338e-01] | [9.261765480041504, 4.87069034576416] |
1e36ae8a-2535-4aca-b865-a8b4e1e89400 | extraction-of-clinical-information-from-the | 1606.01093 | null | http://arxiv.org/abs/1606.01093v1 | http://arxiv.org/pdf/1606.01093v1.pdf | Extraction of clinical information from the non-invasive fetal electrocardiogram | Estimation of the fetal heart rate (FHR) has gained interest in the last
century, low heart rate variability has been studied to identify intrauterine
growth restricted fetuses (prepartum), and abnormal FHR patterns have been
associated with fetal distress during delivery (intrapartum). Several
monitoring techniques have been proposed for FHR estimation, including
auscultation and Doppler ultrasound. This thesis focuses on the extraction of
the non-invasive fetal electrocardiogram (NI-FECG) recorded from a limited set
of abdominal sensors. The main challenge with NI-FECG extraction techniques is
the low signal-to-noise ratio of the FECG signal on the abdominal mixture
signal which consists of a dominant maternal ECG component, FECG and noise.
However the NI-FECG offers many advantages over the alternative fetal
monitoring techniques, the most important one being the opportunity to enable
morphological analysis of the FECG which is vital for determining whether an
observed FHR event is normal or pathological. In order to advance the field of
NI-FECG signal processing, the development of standardised public databases and
benchmarking of a number of published and novel algorithms was necessary. | ['Joachim Behar'] | 2016-05-27 | null | null | null | null | ['heart-rate-variability'] | ['medical'] | [ 5.18150985e-01 1.29617840e-01 2.03331217e-01 -2.49486968e-01
-1.57116666e-01 -6.03903234e-01 -1.52885139e-01 2.44832352e-01
6.64206874e-03 6.63922369e-01 -2.01161131e-01 -4.23970997e-01
-4.42656815e-01 -6.04277670e-01 -2.69892663e-01 -7.10587442e-01
-6.45450950e-01 1.81426689e-01 6.13948330e-02 2.42245629e-01
-4.79217060e-02 7.30278075e-01 -1.31796515e+00 1.50904506e-01
7.23561585e-01 1.14104474e+00 1.58975795e-01 1.25164175e+00
2.26023927e-01 5.35171151e-01 -7.55224526e-01 -2.29657531e-01
1.84038818e-01 -1.10262752e+00 -4.02817369e-01 -6.06909811e-01
8.03490728e-03 -3.42571080e-01 1.98624179e-01 6.97781205e-01
1.05636990e+00 -2.45406106e-01 7.13584483e-01 -8.13969254e-01
3.82743478e-01 5.90844989e-01 -7.37256706e-01 6.86675012e-01
5.05361736e-01 -4.65855300e-01 1.83999479e-01 -8.01351070e-01
6.76305115e-01 5.19059837e-01 1.16369367e+00 2.70647079e-01
-1.15012264e+00 -7.91301489e-01 -9.95663941e-01 6.05643056e-02
-1.36582398e+00 -4.67685223e-01 9.20822680e-01 -4.78080720e-01
6.57209277e-01 7.01075494e-01 1.06591332e+00 5.15171289e-01
3.29902589e-01 1.03786057e-02 9.80144441e-01 -5.95034301e-01
-8.81103799e-02 -2.60875095e-02 -4.80516255e-01 7.89839029e-01
6.40754819e-01 2.08559763e-02 -3.50944787e-01 -3.03641528e-01
1.00804305e+00 -4.42942828e-01 -5.09697676e-01 1.28392830e-01
-6.81604683e-01 4.95648772e-01 -5.29375434e-01 9.14512634e-01
-5.13584316e-01 3.73119973e-02 5.38289130e-01 4.50705707e-01
1.50364578e-01 6.70430660e-01 -1.90861166e-01 -6.39807045e-01
-1.35194170e+00 -1.50560319e-01 9.55472231e-01 4.41652596e-01
4.32324737e-01 1.99906081e-02 7.25068375e-02 7.53666282e-01
5.10923505e-01 5.10184765e-01 4.35787439e-01 -1.33561826e+00
2.65134066e-01 2.30360836e-01 -1.71148762e-01 -1.44780123e+00
-6.06951773e-01 -3.73333514e-01 -8.44520450e-01 -2.06514657e-01
5.88525653e-01 -7.75528371e-01 -3.77038509e-01 1.34313631e+00
5.34157395e-01 2.82916486e-01 -6.41779304e-02 5.90346217e-01
1.16079736e+00 4.08729225e-01 -1.99831605e-01 -1.03049803e+00
1.05544960e+00 2.23180130e-01 -9.34340000e-01 1.50897458e-01
5.42239547e-01 -8.19009602e-01 -6.55897632e-02 4.77701008e-01
-1.14449263e+00 -2.55722761e-01 -9.22005355e-01 4.90391582e-01
8.97481814e-02 1.51392013e-01 3.70946705e-01 1.09978282e+00
-7.96180964e-01 7.32161283e-01 -1.10848141e+00 -1.98267654e-01
8.11315700e-02 1.61552429e-01 -6.05057180e-01 2.02573940e-01
-1.14687037e+00 1.14264989e+00 2.96539366e-01 3.16191614e-01
-1.96607187e-01 -7.13023186e-01 -9.37791765e-01 -2.90602688e-02
3.59472446e-02 -1.56373940e-02 7.15184271e-01 -6.99904442e-01
-1.38603902e+00 7.16655135e-01 -2.96754166e-02 -1.13862269e-01
6.49625599e-01 1.32267177e-01 -5.35558999e-01 6.45374477e-01
-1.14397295e-01 -3.65106553e-01 8.26900303e-01 -7.36292362e-01
-2.33190805e-01 -3.44481558e-01 -8.38670850e-01 -1.13651425e-01
3.08953971e-01 3.87687445e-01 6.46064505e-02 -5.69059253e-01
6.80861771e-01 -5.87273717e-01 1.94633350e-01 -4.31567222e-01
2.50353634e-01 3.69432539e-01 4.70811009e-01 -9.86857653e-01
1.46239698e+00 -2.05019403e+00 -4.29040581e-01 6.85325623e-01
2.07140535e-01 6.68628514e-01 4.34362978e-01 6.21365130e-01
-2.39730150e-01 -2.62678172e-02 -7.40293562e-02 3.61080974e-01
-5.53939641e-01 1.35637239e-01 2.58815825e-01 1.08826566e+00
-8.42669606e-02 6.12701237e-01 -1.00159001e+00 -7.50712633e-01
2.42939278e-01 4.78721827e-01 -1.20472990e-01 4.43709254e-01
7.45216608e-01 5.42259276e-01 -1.76342696e-01 4.62367475e-01
6.92311227e-01 3.38532120e-01 4.93379563e-01 1.38083138e-02
-2.41129979e-01 1.31906390e-01 -1.23912334e+00 1.39496684e+00
1.52590528e-01 8.65143836e-01 3.58109713e-01 -1.24695814e+00
1.22147346e+00 1.15299642e+00 7.23827302e-01 -5.01658499e-01
4.86485124e-01 5.15499294e-01 5.77788234e-01 -1.16976976e+00
-2.59426504e-01 -4.53771889e-01 4.68771070e-01 4.23174560e-01
2.28345811e-01 1.31047340e-02 3.82189363e-01 -2.41957083e-01
1.02957988e+00 1.84570938e-01 8.46908987e-01 -5.14033616e-01
5.91790318e-01 -6.67337537e-01 7.70186484e-01 6.40800714e-01
-1.89425528e-01 9.34763849e-01 8.89461994e-01 -4.18910116e-01
-4.16502565e-01 -5.15357614e-01 -4.29824710e-01 3.53781223e-01
-2.43073404e-01 -4.18711156e-02 -4.73512650e-01 -4.46144231e-02
6.10341281e-02 2.89549708e-01 -7.15152264e-01 1.16315052e-01
-8.67630720e-01 -8.04532647e-01 1.03085589e+00 3.36805254e-01
-1.42314774e-03 -8.49285245e-01 -1.41275990e+00 7.08596051e-01
-1.77267238e-01 -6.35562181e-01 3.02734852e-01 1.89009532e-01
-8.49868655e-01 -1.16775942e+00 -9.06004965e-01 -2.83384353e-01
4.16581303e-01 -4.03517634e-01 9.49122667e-01 1.93983048e-01
-8.26305509e-01 2.48253301e-01 -4.57819521e-01 -8.27703893e-01
-7.32071877e-01 -2.49531761e-01 -9.49364677e-02 1.78282723e-01
-1.92815483e-01 -9.36189234e-01 -7.52099633e-01 1.60854369e-01
-3.91828507e-01 -5.23245446e-02 5.90110540e-01 1.71538934e-01
3.37157667e-01 -4.50907826e-01 9.17942643e-01 -9.89179790e-01
4.62150902e-01 -7.89934933e-01 -3.40626091e-01 2.74480768e-02
-7.01136589e-01 -6.76185846e-01 3.93079191e-01 -3.62336874e-01
-9.01387930e-01 -3.90831590e-01 -4.15018797e-01 -1.46236718e-01
-4.02470291e-01 7.05445528e-01 1.80108830e-01 -1.34058192e-01
7.72371888e-01 1.46984905e-01 2.34333664e-01 -2.93112963e-01
-2.77006269e-01 5.01075029e-01 8.90453279e-01 -4.37439084e-01
2.99299598e-01 -2.20641121e-03 5.91488779e-01 -1.25109816e+00
-2.28574201e-01 -7.00958788e-01 -3.91692042e-01 -5.39162457e-01
6.59885406e-01 -2.54420400e-01 -5.92295825e-01 2.47039288e-01
-1.08114874e+00 9.01535749e-02 -2.23362908e-01 9.96015966e-01
-3.15926611e-01 5.49119949e-01 -5.65183163e-01 -1.30352521e+00
-7.08154202e-01 -6.78804517e-01 -2.74619106e-02 3.38245779e-01
-5.61486900e-01 -7.72368133e-01 4.57170010e-01 3.76133509e-02
6.17644489e-01 1.16837025e+00 5.81928074e-01 -5.03142476e-01
4.80710059e-01 -6.51723802e-01 1.01093166e-01 1.48860902e-01
3.43734473e-01 3.06393653e-01 -9.48072135e-01 -1.20117925e-01
3.95666420e-01 2.85747200e-01 1.08513325e-01 6.53940439e-01
5.57569444e-01 -5.12761511e-02 -1.33865237e-01 7.62817383e-01
1.36319661e+00 8.11434209e-01 7.04342484e-01 -1.92453071e-01
2.00922117e-01 4.70252603e-01 5.96108794e-01 5.78456819e-01
-1.23434134e-01 -2.85736546e-02 3.49149942e-01 -1.58219844e-01
-1.20810896e-01 2.78498709e-01 -2.26788834e-01 6.70075536e-01
-1.11804199e+00 -7.69212693e-02 -1.04801857e+00 4.45491582e-01
-1.35689783e+00 -1.03773773e+00 -6.29423499e-01 2.27550888e+00
7.95730174e-01 -2.00067788e-01 1.29940301e-01 6.82377458e-01
6.64010286e-01 -2.28215739e-01 3.69652957e-02 -7.61899590e-01
1.47338182e-01 5.16467929e-01 2.35325903e-01 1.39304429e-01
-6.69671178e-01 -4.27311271e-01 6.47816992e+00 -5.02690524e-02
-1.45735312e+00 1.82131585e-02 3.59321833e-01 2.78850555e-01
7.95327052e-02 -3.18697333e-01 -2.78094798e-01 5.90530634e-01
1.14726305e+00 -1.12079524e-01 -8.60416144e-02 4.66061711e-01
1.42282754e-01 -4.29056615e-01 -9.17135596e-01 9.91781414e-01
-3.18784470e-04 -1.17233694e+00 -8.85678887e-01 -1.39253482e-01
3.50682586e-01 -2.88713992e-01 -4.98664051e-01 -3.03008229e-01
-1.13429153e+00 -9.75362182e-01 3.72902572e-01 6.30810678e-01
1.47291327e+00 -6.13798022e-01 9.92342770e-01 3.48146737e-01
-1.11269045e+00 2.70248830e-01 -1.34738281e-01 -8.14052746e-02
-1.08637586e-01 7.39752948e-01 -1.09276843e+00 7.29978263e-01
4.51695532e-01 3.53460044e-01 -1.34271607e-01 1.32254386e+00
-1.30831212e-01 1.15264869e+00 -5.66581964e-01 3.11740607e-01
-3.28870207e-01 -1.34701908e-01 7.51706660e-01 1.41708171e+00
5.04891157e-01 6.01005673e-01 -6.11517549e-01 8.29794347e-01
3.60571533e-01 3.62497896e-01 -6.38679266e-01 5.87812252e-02
4.73931909e-01 1.29027319e+00 -9.75046515e-01 -1.11466415e-01
-5.66408753e-01 1.62094057e-01 -3.47891241e-01 7.23454580e-02
-5.62231362e-01 -8.75357091e-01 -5.26237227e-02 6.21350527e-01
-1.84772298e-01 3.89059156e-01 -2.45424092e-01 -5.53068399e-01
-1.72330752e-01 -5.97804010e-01 4.55523103e-01 -2.50230521e-01
-1.78910255e-01 3.62555802e-01 1.52770177e-01 -9.92182791e-01
-7.15421557e-01 -5.95672242e-02 -9.87468004e-01 1.17588067e+00
-9.91640806e-01 -4.52115357e-01 -2.06654534e-01 1.87996745e-01
-1.28416836e-01 2.19718695e-01 1.01945829e+00 6.50383651e-01
-1.86196417e-01 5.35778105e-01 -4.51292753e-01 1.30614132e-01
2.87240475e-01 -1.33093166e+00 -2.78369784e-01 9.73168015e-01
-5.71244359e-01 5.95007777e-01 8.60158741e-01 -7.57157147e-01
-1.37338996e+00 -5.62796116e-01 9.33099627e-01 2.01829784e-02
-5.09078801e-02 2.10384309e-01 -8.88741314e-01 3.26488823e-01
-1.89012378e-01 4.80080575e-01 9.63603377e-01 -5.75794280e-01
3.28442276e-01 -2.49927789e-01 -1.43312883e+00 -2.84742296e-01
2.84294397e-01 1.03261478e-01 -6.28444433e-01 -4.59833235e-01
-3.68456453e-01 -6.75997734e-01 -1.47186065e+00 8.20693016e-01
1.32789958e+00 -1.16739869e+00 5.35266697e-01 -9.55313444e-03
1.41499922e-01 -1.43356264e-01 5.05441844e-01 -7.43740976e-01
-2.61650681e-01 -1.23175144e+00 -1.43394306e-01 1.44577181e+00
3.52002054e-01 -8.78455400e-01 6.10429049e-01 6.33760035e-01
7.01844916e-02 -7.28889048e-01 -9.24966633e-01 -2.78922319e-01
-6.10531807e-01 -3.29663336e-01 1.20971613e-01 9.79242265e-01
3.06168467e-01 -1.35568559e-01 -2.52532780e-01 5.09284288e-02
3.76261592e-01 4.88651767e-02 1.57703727e-01 -1.41873384e+00
-2.84302682e-01 -1.30371347e-01 -5.81434011e-01 1.00222096e-01
-8.03656995e-01 -4.10813689e-01 5.11767827e-02 -1.47025585e+00
-3.04504007e-01 -4.37423766e-01 -4.62940663e-01 7.89848715e-02
-2.15838149e-01 4.68280524e-01 -1.01935990e-01 -3.89335036e-01
1.83160082e-01 -5.20944655e-01 1.10790455e+00 6.38386786e-01
-5.25968552e-01 4.39102292e-01 -1.81097120e-01 9.33243573e-01
5.74464321e-01 -8.21636677e-01 -4.21842277e-01 2.30854973e-01
6.76516473e-01 1.03643215e+00 -3.37745488e-01 -8.58740449e-01
-1.95193477e-02 9.82564613e-02 7.71901727e-01 -6.81815684e-01
-3.55343670e-02 -7.37035453e-01 8.66820097e-01 7.20763743e-01
7.59956017e-02 -6.72625229e-02 3.67999589e-03 -1.51073679e-01
-1.64579853e-01 -6.74137831e-01 1.03020120e+00 -4.45027739e-01
1.32578552e-01 -2.53844019e-02 -8.65678906e-01 2.49894515e-01
8.58513355e-01 -5.36600292e-01 1.44579470e-01 -2.46973902e-01
-8.47231269e-01 -2.87644446e-01 -4.94721271e-02 -3.06485087e-01
6.93575561e-01 -7.30714023e-01 -9.01180387e-01 5.31570435e-01
-2.63624668e-01 6.74151927e-02 1.56527519e-01 1.68338597e+00
-1.35722005e+00 5.78338094e-02 -3.19835663e-01 -4.04689312e-01
-1.37848222e+00 -3.88566218e-02 5.35106540e-01 7.18790516e-02
-6.92036688e-01 9.05491829e-01 -4.13608760e-01 4.61733192e-01
3.86225805e-02 -2.15442747e-01 -7.04667151e-01 4.17195797e-01
6.68061614e-01 1.08492398e+00 2.46993750e-01 -5.33076584e-01
-5.22264719e-01 4.65682596e-01 6.00571215e-01 9.02493149e-02
1.25463653e+00 -2.10323557e-01 -5.01252353e-01 5.79293907e-01
9.15134966e-01 2.64597595e-01 -4.25924480e-01 3.45773906e-01
-1.81764543e-01 -5.53153753e-01 -1.99737430e-01 -6.04827106e-01
-1.07835245e+00 7.40331769e-01 5.91594875e-01 7.94358253e-01
1.37043655e+00 -3.48570168e-01 3.61353159e-01 -6.42540827e-02
2.95557261e-01 -8.96527350e-01 -6.64155662e-01 -1.95384517e-01
9.18824971e-01 -6.40563428e-01 1.65601358e-01 -3.95987302e-01
-1.57750502e-01 1.41807604e+00 -1.30415475e-02 -1.37501284e-01
7.27169037e-01 6.47365987e-01 4.11101997e-01 -1.40750453e-01
-3.18085253e-01 1.75329164e-01 4.28888589e-01 6.37959540e-01
9.42443788e-01 4.82864790e-02 -1.10281622e+00 5.53402722e-01
-1.37328267e-01 4.20157582e-01 5.49522579e-01 1.00648654e+00
-3.76247287e-01 -5.43477416e-01 -3.83113652e-01 9.64746833e-01
-1.61416674e+00 2.70185262e-01 -1.38863465e-02 6.57571614e-01
5.17797351e-01 8.97866726e-01 -3.77390742e-01 2.92026192e-01
2.91141957e-01 3.79118502e-01 8.03533196e-01 -5.78127623e-01
-7.05157876e-01 5.18531680e-01 2.26312101e-01 -5.62266052e-01
-4.89428937e-01 -9.07452285e-01 -1.16109467e+00 2.61803448e-01
-6.18983686e-01 4.28284436e-01 9.49450970e-01 7.66114891e-01
4.03079540e-02 4.32044953e-01 4.55648333e-01 -6.30400360e-01
6.52790368e-02 -1.15539122e+00 -1.24901497e+00 4.57308516e-02
5.71283400e-01 -3.86296928e-01 -2.71379650e-01 2.26393789e-01] | [14.122944831848145, 3.101020574569702] |
e239ea92-c3be-4bc4-a17e-9447490dc1b1 | feature-level-collaboration-joint | null | null | http://openaccess.thecvf.com//content/CVPR2021/html/Chi_Feature-Level_Collaboration_Joint_Unsupervised_Learning_of_Optical_Flow_Stereo_Depth_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Chi_Feature-Level_Collaboration_Joint_Unsupervised_Learning_of_Optical_Flow_Stereo_Depth_CVPR_2021_paper.pdf | Feature-Level Collaboration: Joint Unsupervised Learning of Optical Flow, Stereo Depth and Camera Motion | Precise estimation of optical flow, stereo depth and camera motion are important for the real-world 3D scene understanding and visual perception. Since the three tasks are tightly coupled with the inherent 3D geometric constraints, current studies have demonstrated that the three tasks can be improved through jointly optimizing geometric loss functions of several individual networks. In this paper, we show that effective feature-level collaboration of the networks for the three respective tasks could achieve much greater performance improvement for all three tasks than only loss-level joint optimization. Specifically, we propose a single network to combine and improve the three tasks. The network extracts the features of two consecutive stereo images, and simultaneously estimates optical flow, stereo depth and camera motion. The whole network mainly contains four parts: (I) a feature-sharing encoder to extract features of input images, which can enhance features' representation ability; (II) a pooled decoder to estimate both optical flow and stereo depth; (III) a camera pose estimation module which fuses optical flow and stereo depth information; (IV) a cost volume complement module to improve the performance of optical flow in static and occluded regions. Our method achieves state-of-the-art performance among the joint unsupervised methods, including optical flow and stereo depth estimation on KITTI 2012 and 2015 benchmarks, and camera motion estimation on KITTI VO dataset. | ['Xin Yang', 'Peng Guo', 'Tianyu Hao', 'Qingjie Wang', 'Cheng Chi'] | 2021-06-19 | null | null | null | cvpr-2021-1 | ['stereo-depth-estimation', 'depth-and-camera-motion'] | ['computer-vision', 'computer-vision'] | [-1.10879630e-01 -4.30071771e-01 -4.08571273e-01 -2.58422077e-01
-5.13095319e-01 -4.24949974e-01 4.20434594e-01 -6.49136543e-01
-4.36532974e-01 5.07009506e-01 3.20803612e-01 -7.39407614e-02
1.20841749e-01 -5.30391872e-01 -6.20640993e-01 -6.12627566e-01
-2.79187635e-02 -6.14215806e-02 4.31020498e-01 1.84965387e-01
5.18714368e-01 6.24128282e-01 -1.54032326e+00 -9.11238138e-03
8.32616985e-01 1.38528895e+00 3.70224923e-01 8.99129510e-01
-1.39498115e-01 1.31518471e+00 -1.25389546e-01 1.93540417e-02
5.18635035e-01 -1.45595267e-01 -7.06601739e-01 3.98032814e-01
1.02535629e+00 -9.86354649e-01 -9.17651176e-01 9.46441650e-01
5.20600975e-01 3.75331759e-01 5.42321980e-01 -1.26990116e+00
-4.15304333e-01 -5.98137900e-02 -8.50928247e-01 3.45992029e-01
4.24664497e-01 6.02190852e-01 8.02562654e-01 -9.95220482e-01
6.25374198e-01 1.36545801e+00 3.36518645e-01 3.93096358e-01
-8.39054704e-01 -6.73142374e-01 3.46823603e-01 3.05579096e-01
-1.27932906e+00 -5.79993427e-01 8.98008585e-01 -7.26348221e-01
1.19346499e+00 -3.07650119e-01 8.33404362e-01 7.40081668e-01
2.19713748e-01 9.46466446e-01 6.24334216e-01 2.66743675e-02
-2.62734890e-01 -2.46070139e-02 -1.09267056e-01 1.04400682e+00
9.55382511e-02 4.33404833e-01 -6.73938692e-01 3.59695494e-01
1.23163319e+00 -4.10559103e-02 -5.67232668e-01 -4.56341982e-01
-1.29011738e+00 7.20175922e-01 6.20208263e-01 -2.02694863e-01
-1.13063954e-01 3.78430128e-01 2.26341307e-01 -1.04701724e-02
4.82740372e-01 5.85141554e-02 -4.21129137e-01 -1.30052194e-01
-7.53200829e-01 1.36539429e-01 6.48852646e-01 1.10382581e+00
1.25478935e+00 1.28890514e-01 -1.73936531e-01 6.61939800e-01
6.38110399e-01 6.37483239e-01 2.47845158e-01 -1.42116547e+00
9.84465659e-01 4.61971402e-01 2.34621186e-02 -1.22251117e+00
-3.19324732e-01 -1.92829370e-01 -8.18341017e-01 1.58232480e-01
4.63633627e-01 -2.37604797e-01 -7.43489861e-01 1.71505058e+00
3.47885132e-01 5.76510489e-01 -2.18159571e-01 1.23190677e+00
1.00109112e+00 6.65386915e-01 -3.22921306e-01 -1.38329983e-01
9.05579209e-01 -1.36051285e+00 -6.42494798e-01 -5.59168100e-01
5.09275079e-01 -1.00996757e+00 5.28936923e-01 8.60437006e-02
-1.23407257e+00 -8.60602438e-01 -1.12320375e+00 -7.15196192e-01
-5.84293604e-02 1.56534210e-01 7.36860931e-01 3.32754165e-01
-1.08835113e+00 3.43410552e-01 -6.97223961e-01 2.86997911e-02
6.60035849e-01 4.17953610e-01 -4.19401020e-01 -3.07103634e-01
-9.26308990e-01 7.59656072e-01 1.05399691e-01 3.90244305e-01
-8.55839312e-01 -7.26153135e-01 -1.42961907e+00 -1.18624948e-01
3.62332225e-01 -1.21322131e+00 8.28578055e-01 -5.81366241e-01
-1.71035922e+00 8.35595191e-01 -4.60594296e-01 -3.56757380e-02
6.53375745e-01 -3.89812678e-01 1.26995251e-01 3.35666329e-01
1.96731478e-01 1.06674445e+00 9.31528091e-01 -8.65673780e-01
-1.05112123e+00 -4.03045446e-01 6.56030998e-02 4.54319090e-01
-8.55307207e-02 -3.86298120e-01 -8.62847149e-01 -3.38904530e-01
1.70271710e-01 -7.80337453e-01 -2.46904686e-01 4.10922289e-01
-3.80583167e-01 -8.42614546e-02 8.06323290e-01 -5.23711622e-01
1.02371848e+00 -2.04018092e+00 3.51675183e-01 -1.51047885e-01
5.23848891e-01 2.93622732e-01 -1.55045986e-01 -2.37425476e-01
2.24711984e-01 -1.68544173e-01 -4.98433784e-02 -6.76335871e-01
-2.42077217e-01 6.77180141e-02 -1.08086511e-01 7.86163390e-01
3.30642134e-01 9.95851994e-01 -9.75930035e-01 -6.79029465e-01
9.17930186e-01 5.27482271e-01 -8.50380182e-01 3.84238094e-01
1.54733822e-01 6.04066312e-01 -5.24975002e-01 6.14945829e-01
9.81826127e-01 -2.72464484e-01 -3.54965925e-01 -3.53426158e-01
-1.81476846e-01 2.81183451e-01 -1.44099998e+00 1.96651566e+00
-3.71194571e-01 8.61177802e-01 -7.38925338e-02 -6.83822095e-01
7.99125135e-01 5.79179125e-03 6.97135270e-01 -6.01757586e-01
3.11342418e-01 1.90454900e-01 -3.22735012e-01 -6.66838944e-01
3.82095426e-01 2.65991420e-01 1.73055872e-01 2.51171529e-01
2.81737506e-01 -5.17104328e-01 1.21407636e-01 1.82411760e-01
8.50076199e-01 3.67202431e-01 1.91987485e-01 -5.39474450e-02
9.80272651e-01 -3.96780998e-01 9.26547706e-01 4.70932066e-01
-6.96480453e-01 7.36465216e-01 3.50111544e-01 -6.21747017e-01
-1.00371182e+00 -1.16334128e+00 -1.90479830e-02 3.25498581e-01
6.94428861e-01 -2.64304191e-01 -2.84650892e-01 -7.38516152e-01
2.61318594e-01 -5.47199138e-02 -5.67061245e-01 -1.97720960e-01
-6.46363676e-01 -2.44182587e-01 3.36614847e-01 5.63114107e-01
9.60938632e-01 -6.61140442e-01 -4.95772123e-01 5.82930185e-02
-5.27461469e-01 -1.65990317e+00 -1.13214791e+00 -1.21076725e-01
-8.69187057e-01 -1.26905596e+00 -6.00324512e-01 -6.63442194e-01
5.64249158e-01 8.04203093e-01 9.11717713e-01 -3.23071256e-02
-2.76587814e-01 2.85748005e-01 -3.62202600e-02 -6.52547777e-02
1.86251923e-01 -4.36495282e-02 -1.18329391e-01 1.34075567e-01
1.92482814e-01 -6.45784378e-01 -9.72522199e-01 4.40061569e-01
-7.20679045e-01 1.46976694e-01 4.97699916e-01 6.37615204e-01
3.55412960e-01 -3.73305887e-01 -6.06507249e-02 -4.82306361e-01
-5.98575100e-02 -1.96291625e-01 -7.48635888e-01 -1.21096738e-01
-3.12922001e-01 -9.01332349e-02 4.78817433e-01 -1.93096444e-01
-1.19128287e+00 4.91454959e-01 -4.19111326e-02 -8.22379589e-01
6.88511506e-02 -6.30052760e-02 -2.47767463e-01 -4.53657418e-01
3.43789339e-01 2.91905940e-01 1.20018773e-01 -1.48782089e-01
4.17867690e-01 5.13767540e-01 5.90992689e-01 -2.55835056e-01
9.11383331e-01 8.79624546e-01 3.15855145e-01 -7.56961763e-01
-1.12486422e+00 -8.21607471e-01 -8.78406823e-01 -3.75055403e-01
1.06525183e+00 -1.30867445e+00 -9.44809079e-01 9.52473938e-01
-1.43508661e+00 -1.88678056e-01 1.03855774e-01 9.03192461e-01
-7.64158130e-01 6.90407753e-01 -7.60735214e-01 -4.72064614e-01
-1.33368596e-01 -1.60022521e+00 1.20443678e+00 4.47067946e-01
3.02822441e-01 -1.22689104e+00 -3.12446523e-02 5.51185250e-01
1.42160267e-01 1.04207166e-01 2.50611365e-01 3.66200179e-01
-1.19395196e+00 1.40892103e-01 -7.28080809e-01 7.95462906e-01
2.86218315e-01 1.95566073e-01 -1.17613268e+00 -2.48911902e-01
1.15488125e-02 -2.35247031e-01 1.10202658e+00 8.43036711e-01
1.05285490e+00 1.04535006e-01 -1.57901481e-01 1.55265677e+00
1.51188540e+00 6.54072985e-02 7.61306167e-01 1.83491886e-01
1.35736716e+00 6.16225302e-01 4.16350722e-01 4.33338225e-01
6.63746417e-01 6.40763402e-01 6.18370831e-01 -1.21168219e-01
-4.23422247e-01 -1.58127204e-01 5.52959085e-01 8.57070804e-01
-1.71511889e-01 1.25071509e-02 -4.55480933e-01 4.82693672e-01
-1.91751528e+00 -9.66666043e-01 -1.19423524e-01 1.91517270e+00
5.24002790e-01 -6.11620536e-03 -1.01091258e-01 -1.40937477e-01
5.95177352e-01 6.33413494e-01 -7.49981225e-01 2.43661832e-02
-2.15006962e-01 -2.61739105e-01 7.28579044e-01 8.17985892e-01
-1.21326113e+00 9.79299307e-01 6.03235912e+00 5.42892098e-01
-1.00769818e+00 -2.00160712e-01 5.42498946e-01 -3.60696197e-01
-1.02590971e-01 1.03289314e-01 -1.09301889e+00 5.27008712e-01
1.87467948e-01 3.07622924e-02 4.40016806e-01 5.43426991e-01
4.26380068e-01 -2.48960525e-01 -1.05545044e+00 1.39076734e+00
2.25520968e-01 -1.51144886e+00 1.40651688e-01 4.07908052e-01
9.79991734e-01 2.31503561e-01 -8.86953399e-02 -8.41392577e-02
9.84700099e-02 -8.93107653e-01 6.12003922e-01 6.62818551e-01
7.71821558e-01 -7.06903756e-01 7.89788187e-01 1.99644804e-01
-1.48260343e+00 -1.51293710e-01 -3.67370218e-01 -2.42550492e-01
4.62187141e-01 8.30378354e-01 -7.85442442e-02 6.80906296e-01
7.74152637e-01 1.59348321e+00 -3.97774190e-01 1.20641327e+00
-3.64315122e-01 -8.49109367e-02 -2.11649477e-01 2.95506597e-01
4.89019483e-01 -2.64061183e-01 7.29209125e-01 1.10269189e+00
1.25834733e-01 -1.18627533e-01 2.60901839e-01 8.02975118e-01
-8.91542528e-03 -2.34081328e-01 -7.27863371e-01 4.29920197e-01
3.53326052e-01 1.22692001e+00 -3.45286101e-01 -3.34868878e-01
-7.29196846e-01 9.34501886e-01 3.81511986e-01 3.76139045e-01
-8.43725204e-01 -4.75196779e-01 1.18399429e+00 -1.55625641e-01
3.58757973e-01 -4.56045687e-01 -2.52931684e-01 -1.55842769e+00
1.09182760e-01 -2.92338014e-01 2.43690610e-01 -8.17238152e-01
-1.08210802e+00 2.83384651e-01 -2.86473036e-01 -1.43823862e+00
-2.76353747e-01 -7.73335576e-01 -5.06355643e-01 1.06490934e+00
-2.12551618e+00 -7.75485039e-01 -8.45736623e-01 9.50619698e-01
6.17272854e-01 1.65719146e-04 -2.14901809e-02 6.19148910e-01
-8.01816106e-01 2.88821995e-01 -2.61047184e-01 3.51576626e-01
8.86591315e-01 -1.08325624e+00 3.69565070e-01 1.06431162e+00
-7.21190870e-02 3.48439038e-01 5.80127537e-02 -3.31536293e-01
-1.47929788e+00 -1.15868747e+00 8.44736934e-01 -5.21009743e-01
4.17136192e-01 -2.40420163e-01 -5.85819125e-01 6.79938674e-01
-9.49681252e-02 4.64774549e-01 1.17652796e-01 -4.95259523e-01
-1.99536368e-01 -2.19004273e-01 -7.68720627e-01 3.08357030e-01
1.42054403e+00 -6.85415566e-01 -1.83724955e-01 1.39110416e-01
7.78683245e-01 -7.93333471e-01 -6.81641281e-01 2.93544590e-01
5.83362520e-01 -1.39238179e+00 1.31868637e+00 -2.85751998e-01
7.79644489e-01 -5.07879555e-01 -1.91528559e-01 -9.78654325e-01
-2.25785717e-01 -7.08279848e-01 -3.81589264e-01 1.02352858e+00
-6.38999511e-03 -6.01138413e-01 7.99457490e-01 4.19691414e-01
-2.96866924e-01 -6.79223955e-01 -7.59575009e-01 -5.88374436e-01
-1.69455066e-01 -5.55278540e-01 2.99178690e-01 7.72623062e-01
-4.62434292e-01 4.77355868e-01 -5.65013707e-01 5.13988547e-02
7.99938023e-01 -1.14174463e-01 1.10319996e+00 -9.34583187e-01
-1.83525562e-01 -6.63039625e-01 -6.85073197e-01 -1.95112681e+00
2.94480294e-01 -7.23060668e-01 1.57718554e-01 -1.41729891e+00
2.14804158e-01 -5.33268191e-02 5.46001978e-02 -2.86895894e-02
-2.98431069e-01 2.21254587e-01 2.59320229e-01 2.96816200e-01
-5.26854038e-01 7.48292625e-01 2.05880809e+00 1.16027044e-02
-3.05377185e-01 -9.80935842e-02 -4.67586815e-01 8.36724997e-01
2.34800160e-01 -1.84837818e-01 -6.37115717e-01 -8.43269467e-01
-6.22035563e-02 2.00457796e-01 5.87668717e-01 -9.35345650e-01
5.04881799e-01 -2.30541691e-01 5.94750464e-01 -6.66762292e-01
3.87192339e-01 -6.42283142e-01 -5.14721572e-01 3.12556773e-01
-1.79058425e-02 -5.37715815e-02 8.91388729e-02 6.13373756e-01
-4.56037194e-01 9.83325318e-02 9.01487231e-01 -1.64524853e-01
-1.10381305e+00 9.43780899e-01 -9.00247842e-02 1.84918985e-01
9.22656238e-01 -5.62693477e-01 -5.44035017e-01 -4.85715955e-01
-2.75268465e-01 5.32964468e-01 3.45218837e-01 6.02750063e-01
8.72028649e-01 -1.32331347e+00 -6.34830415e-01 4.01404530e-01
-8.31703246e-02 5.56895435e-01 4.64444518e-01 9.49460983e-01
-8.83488655e-01 5.26763320e-01 -3.47933263e-01 -1.01377845e+00
-8.28251243e-01 3.02145243e-01 6.64325416e-01 -1.09499447e-01
-4.99290913e-01 9.69136596e-01 6.33409381e-01 -3.16578031e-01
3.83633614e-01 -3.33419472e-01 -2.15586841e-01 -1.90716553e-02
5.69371939e-01 5.56264400e-01 -3.06488365e-01 -7.63804853e-01
-4.04539496e-01 1.24452591e+00 2.96023507e-02 9.13765803e-02
1.04807305e+00 -7.23437071e-01 -2.41679344e-02 2.03372389e-01
1.76346278e+00 -3.12547475e-01 -2.07041574e+00 -3.89295936e-01
-6.50049448e-01 -1.07784438e+00 4.43043470e-01 -2.36606717e-01
-1.49469018e+00 1.22186565e+00 2.88726240e-01 -2.91109413e-01
1.24965203e+00 -2.27978349e-01 1.07161295e+00 1.12881504e-01
1.82676151e-01 -9.36088383e-01 3.17135662e-01 7.58396447e-01
4.29077953e-01 -1.52844870e+00 1.11644998e-01 -7.17475891e-01
-4.46126401e-01 1.16558707e+00 8.65320206e-01 -2.55610496e-01
7.31997907e-01 3.12513020e-03 6.87749907e-02 2.05896422e-02
-8.09666932e-01 -4.14909989e-01 6.35972083e-01 6.63830757e-01
1.96965367e-01 -4.62500483e-01 2.42194697e-01 7.10418867e-03
2.53989659e-02 7.36142918e-02 4.27527040e-01 5.64225197e-01
-4.14085656e-01 -6.73770607e-01 -7.75260990e-03 2.62750894e-01
-2.15541065e-01 -4.01227027e-02 -1.34498745e-01 8.94169509e-01
1.88420087e-01 8.71141076e-01 2.73231268e-01 -4.98560220e-01
3.29647571e-01 -6.01695597e-01 6.46621883e-01 -3.70331764e-01
-1.86872214e-01 4.67698053e-02 -4.86141182e-02 -1.10053384e+00
-6.27457440e-01 -5.93336523e-01 -9.75154936e-01 -6.01866186e-01
-1.47351727e-01 -4.44439292e-01 4.20268685e-01 1.10515893e+00
2.76053578e-01 3.79413903e-01 9.62087810e-01 -1.24489331e+00
-1.77188873e-01 -6.88749850e-01 -5.45025945e-01 3.71165186e-01
7.67773330e-01 -9.04066622e-01 -8.47532451e-01 1.04962610e-01] | [8.705829620361328, -1.9925874471664429] |
870ec49f-669a-4d0c-9faa-56da6cb608f7 | font-acknowledgment-and-character-extraction | 1305.4064 | null | http://arxiv.org/abs/1305.4064v1 | http://arxiv.org/pdf/1305.4064v1.pdf | Font Acknowledgment and Character Extraction of Digital and Scanned Images | The font recognition and character extraction is of immense importance as
these are many scenarios where data are in such a form, which cannot be
processed like in image form or as a hard copy. So the procedure developed in
this paper is basically related to identifying the font (Times New Roman, Arial
and Comic Sans MS) and afterwards recovering the text using simple correlation
based method where the binary templates are correlated to the input image text
characters. All of this extraction is done in the presence of a little noise as
images may have noisy patterns due to photocopying. The significance of this
method exists in extraction of data from various monitoring (Surveillance)
camera footages or even more. The method is developed on Matlab\c{opyright}
which takes input image and recovers text and font information from it in a
text file. | ['Syed Muhammad Arsalan Bashir'] | 2013-05-17 | null | null | null | null | ['font-recognition'] | ['computer-vision'] | [ 8.69066477e-01 -3.07562739e-01 4.66004223e-01 -3.32035720e-02
8.75051394e-02 -1.03362978e+00 6.32571459e-01 1.78649977e-01
-4.00779098e-01 7.98878133e-01 -1.15465716e-01 -6.07903361e-01
-1.18037052e-01 -7.09151983e-01 -2.83529669e-01 -5.88818908e-01
3.59321237e-01 1.68934166e-01 4.19291615e-01 -1.03640452e-01
5.99760592e-01 6.96084619e-01 -1.36735630e+00 2.96884686e-01
5.79117179e-01 5.91875196e-01 2.07922980e-01 1.45200658e+00
-3.19232881e-01 1.01189435e+00 -8.44691336e-01 -3.59341800e-01
6.42262220e-01 -3.82104725e-01 -4.87566441e-01 4.87926692e-01
1.46210790e-01 -6.54560626e-01 -1.47819832e-01 1.08724070e+00
1.80317283e-01 -2.58697867e-01 1.13803756e+00 -6.96265757e-01
-3.64133537e-01 -7.74812102e-02 -8.59546006e-01 3.45474839e-01
3.55078667e-01 1.22056656e-01 6.12141825e-02 -7.92574525e-01
4.59327430e-01 8.32565427e-01 4.33571100e-01 -1.06815994e-01
-9.36759651e-01 -3.67039621e-01 -6.14001393e-01 -1.22951351e-01
-1.06686151e+00 -2.15001568e-01 5.88359118e-01 -7.12422907e-01
6.83612466e-01 4.70770121e-01 4.32521939e-01 6.64940059e-01
3.46544534e-01 3.16818058e-01 1.75336051e+00 -8.74777913e-01
-1.91126034e-01 6.35977566e-01 3.92583132e-01 5.87933540e-01
5.34720540e-01 -1.45150140e-01 -1.22984836e-03 1.13243066e-01
8.69190454e-01 1.17312841e-01 -1.58593312e-01 3.97637725e-01
-1.01009858e+00 4.55666989e-01 -4.31424111e-01 4.81698036e-01
-4.15796041e-01 -5.40257990e-01 1.85691252e-01 4.47195113e-01
-1.08619332e-01 2.30708141e-02 -8.65902752e-02 -3.50020558e-01
-1.23373783e+00 -1.93283781e-01 7.79904902e-01 9.18768167e-01
3.48218173e-01 1.93853840e-01 3.48157734e-01 6.93054557e-01
5.89227378e-02 1.02003396e+00 3.47895980e-01 -1.86524004e-01
7.41740823e-01 6.78282976e-01 2.42881462e-01 -1.16482866e+00
-1.90500021e-01 9.87749696e-02 -7.24864423e-01 7.30214894e-01
5.34667015e-01 -5.22835612e-01 -1.00265539e+00 6.81833327e-01
1.38650641e-01 -3.47208023e-01 9.04371068e-02 4.80944157e-01
5.22944689e-01 8.20530295e-01 -2.08545998e-01 -2.71349669e-01
1.62655222e+00 -4.29295391e-01 -9.79417622e-01 7.15800375e-02
-1.36076272e-01 -1.58394289e+00 5.97855866e-01 1.03222489e+00
-7.50285387e-01 -4.94024336e-01 -1.18914199e+00 2.20841914e-01
-8.45463693e-01 7.06054151e-01 2.57572308e-02 1.03116977e+00
-6.06161177e-01 4.28991467e-01 -3.53104770e-01 -5.40470719e-01
1.43443411e-02 4.22325313e-01 -4.20269489e-01 2.66083747e-01
-4.24221843e-01 1.02195203e+00 4.11186129e-01 2.82063663e-01
1.55582204e-02 4.09609646e-01 -3.39731693e-01 -3.36441308e-01
2.07502961e-01 6.25116304e-02 5.72383583e-01 -1.29578948e+00
-1.26319695e+00 7.44625747e-01 8.40697214e-02 -2.49518916e-01
9.42807198e-01 7.65213072e-02 -9.07796264e-01 3.99596363e-01
-2.65119523e-01 3.11477855e-02 1.49003780e+00 -8.05834413e-01
-5.00370562e-01 -4.72708166e-01 -4.54118460e-01 -2.86874231e-02
-2.77770728e-01 3.93822759e-01 -1.90475330e-01 -8.22841227e-01
1.12979293e-01 -6.46076739e-01 3.38944197e-01 -2.11617917e-01
-4.58087802e-01 3.08963180e-01 1.27889049e+00 -1.38831377e+00
1.26096785e+00 -1.77478349e+00 -3.95774245e-01 6.48426712e-01
-1.25708610e-01 6.73176229e-01 3.42285067e-01 8.48739743e-01
-2.21277222e-01 -8.61575380e-02 -2.37690851e-01 3.94885540e-01
-4.11237091e-01 -3.47365215e-02 -1.77247003e-01 4.55864489e-01
3.77959311e-01 6.04947098e-02 -2.24464774e-01 -8.76104593e-01
4.57286507e-01 4.26520884e-01 3.92166197e-01 -2.92603336e-02
1.83469921e-01 2.47806028e-01 -3.82767707e-01 7.57516742e-01
1.00236332e+00 1.65693462e-01 2.53679156e-01 -2.53247410e-01
-4.06091541e-01 -3.49100381e-01 -1.65727890e+00 5.00979841e-01
-2.25114282e-02 1.06321597e+00 1.43670784e-02 -7.78107166e-01
1.18096614e+00 3.74464691e-01 1.68676838e-01 -4.04314339e-01
4.74519163e-01 8.96300673e-02 -1.82966262e-01 -1.00214827e+00
9.83541012e-01 3.68969709e-01 5.17678618e-01 5.53707242e-01
-4.00266290e-01 -5.51240693e-04 6.66509569e-01 -5.50517477e-02
7.68086612e-01 1.20809115e-01 5.14643550e-01 -3.76035832e-02
8.22446406e-01 1.96999416e-01 -1.17365278e-01 3.42502922e-01
1.14760481e-01 4.47452515e-01 5.61925590e-01 -2.02394184e-02
-1.57242584e+00 -6.70304954e-01 -4.37724769e-01 1.62756175e-01
-3.28146815e-01 -4.88297231e-02 -7.34457791e-01 -2.42262289e-01
-3.74577045e-01 1.88024983e-01 -4.60556984e-01 6.37416601e-01
-6.37715816e-01 -5.67030668e-01 3.66461813e-01 1.23875767e-01
6.09495938e-01 -1.12989438e+00 -8.10419381e-01 1.65129349e-01
3.78202319e-01 -1.10618389e+00 -9.49009880e-02 1.26422450e-01
-1.09344947e+00 -1.37309659e+00 -9.04702842e-01 -6.50241137e-01
1.07299173e+00 2.88072854e-01 5.47875285e-01 1.01266533e-01
-6.56012714e-01 2.03833476e-01 -5.59687793e-01 -6.57735407e-01
-6.08556569e-01 -2.78789043e-01 -4.23904240e-01 1.32347958e-03
6.26624346e-01 -3.05500180e-01 -3.95939708e-01 4.23809886e-02
-1.16930056e+00 -1.19166739e-01 6.96968615e-01 6.75101995e-01
1.89167976e-01 4.97964948e-01 7.24954996e-03 -1.07626021e+00
9.67148542e-01 -2.67938077e-01 -1.14800048e+00 3.16879809e-01
-4.79822636e-01 -1.11367635e-01 9.30093348e-01 -1.99598014e-01
-7.77047396e-01 4.40962195e-01 2.43951038e-01 3.98671068e-02
-4.14468020e-01 6.36973977e-02 1.87296167e-01 -1.54493535e-02
5.10276675e-01 4.95205045e-01 2.40504116e-01 -5.91307938e-01
-2.68174738e-01 1.32879138e+00 6.85320497e-01 -1.38800293e-01
1.26505697e+00 3.06137800e-01 2.86694318e-01 -1.31582594e+00
8.00961182e-02 -4.13630247e-01 -8.53726149e-01 -5.06958544e-01
9.32822526e-01 -3.30490172e-01 -5.62451303e-01 6.50483847e-01
-1.14264452e+00 3.41319978e-01 2.88498521e-01 4.31890815e-01
-6.14657998e-02 8.63548517e-01 -3.11351866e-01 -1.45907652e+00
-3.53646457e-01 -8.81461620e-01 4.56073552e-01 5.26015520e-01
7.07267448e-02 -7.55991817e-01 -9.93614867e-02 2.08653241e-01
9.78400782e-02 4.31857586e-01 8.20683777e-01 -4.05821681e-01
-4.10649091e-01 -6.88299298e-01 -4.99535024e-01 5.62498331e-01
3.68923068e-01 6.63992465e-01 -7.48175979e-01 1.77745044e-01
8.54854509e-02 1.21947825e-01 3.44575614e-01 5.22306152e-02
5.23633182e-01 -2.55061448e-01 1.93723917e-01 3.20139229e-01
1.99331248e+00 7.16752112e-01 8.75720799e-01 3.86901528e-01
2.75599271e-01 4.50208664e-01 5.75109363e-01 5.56244016e-01
-2.95234740e-01 2.92410374e-01 2.35884041e-01 -1.94419384e-01
1.78064927e-01 6.56218380e-02 3.07662904e-01 5.25397480e-01
-2.33708560e-01 -2.56574363e-01 -8.20113480e-01 2.48021781e-01
-1.32366288e+00 -1.18101740e+00 -9.05928910e-01 2.43629074e+00
5.01403153e-01 6.20283820e-02 2.53419280e-01 6.39804661e-01
9.56982374e-01 -1.13970987e-01 -1.19363412e-01 -8.28641474e-01
-3.57959241e-01 4.08521384e-01 9.59653616e-01 3.34360689e-01
-1.05142879e+00 4.67252076e-01 5.97593355e+00 4.53415364e-01
-1.31801641e+00 -3.02151173e-01 5.47274053e-01 3.82796347e-01
4.10189927e-01 -2.31272746e-02 -6.91951394e-01 8.52790773e-01
7.71501899e-01 3.08720887e-01 5.12283862e-01 5.05506217e-01
5.62018275e-01 -9.66631889e-01 -4.58872050e-01 1.16668546e+00
1.62337765e-01 -9.32302415e-01 -1.48054257e-01 2.57364929e-01
6.74149156e-01 -5.86183608e-01 2.25989074e-02 -5.08913755e-01
2.78327037e-02 -1.01976633e+00 5.87745607e-01 8.19860876e-01
7.27213562e-01 -5.69413900e-01 7.54745305e-01 3.13299656e-01
-9.55298901e-01 -8.43836740e-02 -5.21598220e-01 1.27011538e-01
-1.57977839e-03 3.95820171e-01 -1.10885835e+00 4.84815836e-01
3.39326799e-01 2.39226669e-01 -8.86186063e-01 1.14200151e+00
1.24715105e-01 5.95992148e-01 -5.48471510e-01 -2.87060589e-01
1.75033092e-01 -8.18189800e-01 4.86276358e-01 1.57282364e+00
4.37333554e-01 1.11830562e-01 -4.36316580e-01 4.35577035e-01
3.38152379e-01 4.94856238e-01 -7.90363371e-01 -3.83107126e-01
2.01268122e-01 1.32490015e+00 -1.32296550e+00 -4.81280029e-01
-5.62896609e-01 9.91884470e-01 -5.21476448e-01 2.12888867e-01
-5.32669604e-01 -9.45807993e-01 -3.27907860e-01 1.90449372e-01
3.90638590e-01 -4.11498308e-01 -4.10695136e-01 -8.03583205e-01
3.93351078e-01 -1.14279830e+00 1.22390792e-01 -8.14614177e-01
-7.85466731e-01 7.49446273e-01 -7.16864318e-02 -1.58143878e+00
-1.76075727e-01 -1.10236049e+00 -7.09525466e-01 1.06840849e+00
-9.00469005e-01 -8.27777207e-01 -5.03548265e-01 7.25983679e-01
6.50700152e-01 -6.13504708e-01 4.45775419e-01 1.56957895e-01
-4.99487400e-01 1.67895779e-01 7.16264725e-01 5.54219007e-01
5.40036082e-01 -1.24150836e+00 -1.38811320e-01 1.27906621e+00
-8.57743397e-02 4.59694237e-01 8.72764111e-01 -9.68692601e-01
-1.25233340e+00 -3.57603908e-01 7.02242315e-01 -4.45842594e-01
6.05765462e-01 -2.45343044e-01 -4.40727651e-01 1.71649173e-01
6.98020399e-01 -5.60626864e-01 3.14840376e-01 -8.51658285e-01
-3.33215087e-03 -6.09255806e-02 -1.29981530e+00 3.60646933e-01
5.85729145e-02 -3.81930143e-01 -6.36871994e-01 3.83044988e-01
-4.31391507e-01 6.02702145e-03 -7.09512770e-01 -5.02456844e-01
6.18955851e-01 -1.13414693e+00 5.68334341e-01 -2.63746530e-01
3.43796104e-01 -6.35276198e-01 1.66459549e-02 -3.81680369e-01
4.00924742e-01 -8.42802405e-01 5.50966859e-01 1.43512809e+00
4.34240490e-01 -6.00613594e-01 6.77828789e-01 4.63969588e-01
5.58823287e-01 9.74734593e-03 -4.35527295e-01 -5.69242299e-01
-4.64925498e-01 5.37402369e-02 2.54580051e-01 8.66331041e-01
-2.19834819e-01 1.39724612e-01 -5.94588637e-01 -2.35409196e-02
7.11047351e-01 -1.70332253e-01 6.74225748e-01 -1.25237119e+00
-2.66981989e-01 -7.18130544e-02 -5.04864693e-01 -3.03929269e-01
-5.85163772e-01 -1.71719670e-01 -3.81530404e-01 -1.27547264e+00
-2.71791983e-02 -1.11154974e-01 4.09951210e-01 4.09088917e-02
1.81172222e-01 2.78817654e-01 4.31076050e-01 4.05396312e-01
2.15365738e-01 -3.52879375e-01 8.83259356e-01 3.60911600e-02
-6.91262782e-02 2.13588327e-01 -1.29360035e-01 6.15470767e-01
7.17692494e-01 -3.64439189e-01 -3.77829462e-01 -1.23329453e-01
4.56812024e-01 2.15547457e-01 2.26242334e-01 -1.14851677e+00
1.96673870e-01 -1.52444243e-02 8.67534935e-01 -1.16239047e+00
1.39061630e-01 -1.49660718e+00 4.33192134e-01 3.17870080e-01
1.93207830e-01 7.89568484e-01 -1.50415733e-01 4.97859478e-01
8.21283739e-03 -1.06059110e+00 8.59812737e-01 -3.67470205e-01
-2.82105744e-01 -3.75945151e-01 -8.37626994e-01 -4.21174735e-01
1.03722906e+00 -1.12015390e+00 -4.02545363e-01 -4.10821676e-01
-4.59729165e-01 -3.83087009e-01 6.08933330e-01 -1.43379703e-01
2.98184454e-01 -7.16864467e-01 -5.27514696e-01 2.55004883e-01
-3.55824947e-01 -5.20574510e-01 1.79376349e-01 8.56971145e-01
-1.52096868e+00 6.02644801e-01 -7.95221210e-01 -1.85850456e-01
-1.82036543e+00 6.23874128e-01 -5.28923459e-02 -1.22471824e-01
-5.33046186e-01 -3.56391221e-02 -6.58374071e-01 3.23699772e-01
-2.53966573e-04 -1.91966876e-01 -8.56770456e-01 1.40082210e-01
6.79924428e-01 6.37371898e-01 1.24377124e-01 -7.80195534e-01
4.95716147e-02 1.01100171e+00 -1.13096252e-01 -4.20447141e-01
1.34760308e+00 -1.00922637e-01 -2.86991626e-01 1.11089483e-01
9.50762331e-01 6.10654235e-01 -1.08224022e+00 1.38347581e-01
9.84031260e-02 -7.08197713e-01 -2.94599980e-01 -7.33660221e-01
-7.47198403e-01 8.48228872e-01 1.00188136e+00 4.67591465e-01
1.32303774e+00 -8.76495183e-01 2.97606915e-01 5.96281052e-01
-1.32551447e-01 -1.83425546e+00 -3.44666064e-01 2.54861057e-01
5.66813171e-01 -1.09546065e+00 4.62202489e-01 -3.06589991e-01
-7.51606226e-01 1.98427355e+00 1.44248709e-01 -2.00720742e-01
6.88808322e-01 5.88016689e-01 2.69884579e-02 -2.05752477e-02
-3.69696021e-01 -2.34972581e-01 -1.17892928e-01 9.16181386e-01
5.61910748e-01 -1.36629924e-01 -8.28037143e-01 6.85422821e-03
-1.41727924e-01 8.59084800e-02 1.22033119e+00 1.25161505e+00
-5.51530719e-01 -1.19379425e+00 -1.13154089e+00 8.17691267e-01
-9.31930900e-01 -1.92196861e-01 -8.03085327e-01 8.86701763e-01
1.14306219e-01 1.02608216e+00 -3.68782692e-02 -3.41135591e-01
3.27840261e-02 1.94803402e-01 3.21520597e-01 -1.22284070e-02
-9.03203726e-01 2.97336340e-01 1.81335256e-01 3.16016376e-01
-2.79794216e-01 -8.00436735e-01 -8.30046356e-01 -4.32393283e-01
-2.62238353e-01 -6.23722747e-03 1.28115571e+00 6.33805931e-01
-4.21059430e-01 6.50980324e-02 6.33461237e-01 -2.45842606e-01
-3.27420503e-01 -1.03131723e+00 -5.28926849e-01 2.60908842e-01
2.23767504e-01 -9.62101668e-02 -5.94879240e-02 6.77467227e-01] | [11.855216026306152, 2.544970750808716] |
b97a3f2c-3156-463c-bc88-7512c8c20aff | effects-of-spectral-normalization-in-multi | 2212.05331 | null | https://arxiv.org/abs/2212.05331v2 | https://arxiv.org/pdf/2212.05331v2.pdf | Effects of Spectral Normalization in Multi-agent Reinforcement Learning | A reliable critic is central to on-policy actor-critic learning. But it becomes challenging to learn a reliable critic in a multi-agent sparse reward scenario due to two factors: 1) The joint action space grows exponentially with the number of agents 2) This, combined with the reward sparseness and environment noise, leads to large sample requirements for accurate learning. We show that regularising the critic with spectral normalization (SN) enables it to learn more robustly, even in multi-agent on-policy sparse reward scenarios. Our experiments show that the regularised critic is quickly able to learn from the sparse rewarding experience in the complex SMAC and RWARE domains. These findings highlight the importance of regularisation in the critic for stable learning. | ['Pawan Kumar', 'Anuj Mahajan', 'Kinal Mehta'] | 2022-12-10 | null | null | null | null | ['smac-1', 'smac'] | ['playing-games', 'playing-games'] | [-6.68249130e-02 4.88651767e-02 -2.46809751e-01 3.42284799e-01
-1.13921428e+00 -5.07589161e-01 5.38147688e-01 2.59185466e-03
-7.88918197e-01 1.27558756e+00 3.09291422e-01 -5.21528833e-02
-4.16720688e-01 -3.46416533e-02 -4.52653885e-01 -9.05313849e-01
-5.55930912e-01 5.08059025e-01 1.78219214e-01 -3.20458323e-01
3.67102288e-02 3.64326626e-01 -1.26930523e+00 -3.87703866e-01
6.53112352e-01 7.18983233e-01 1.96882874e-01 9.89260614e-01
3.87003660e-01 1.35007739e+00 -8.27898264e-01 3.01621050e-01
7.28919387e-01 -7.03789532e-01 -4.03493404e-01 8.17767065e-03
9.36233476e-02 -5.08241236e-01 -7.82006904e-02 9.05966461e-01
7.84967780e-01 6.59276187e-01 5.70543826e-01 -1.07533836e+00
-2.55532954e-02 6.61999881e-01 -7.55868852e-01 1.85365021e-01
1.86445147e-01 5.58154762e-01 9.52726424e-01 -4.12421644e-01
4.43683714e-01 1.71589875e+00 6.56781018e-01 6.19053125e-01
-1.28085363e+00 -5.15250564e-01 5.72820723e-01 -2.64017105e-01
-7.81774759e-01 -6.13612592e-01 4.06035602e-01 -1.65077627e-01
9.91806149e-01 -3.49567570e-02 9.79462504e-01 1.16873372e+00
5.46744019e-02 8.82796049e-01 1.20052433e+00 -1.12089954e-01
7.55399406e-01 -1.73935011e-01 -6.26294494e-01 4.84782100e-01
2.94152528e-01 6.98724926e-01 -2.54680544e-01 -3.80832940e-01
1.21395755e+00 -1.44533440e-01 1.37011737e-01 -4.83965039e-01
-1.19225204e+00 9.92036521e-01 4.69388485e-01 3.33313918e-04
-7.02689350e-01 8.53278935e-01 6.97358310e-01 8.58833969e-01
2.76470274e-01 9.24659789e-01 -5.75578272e-01 -6.88926578e-01
-5.38994074e-01 6.61450088e-01 8.39630425e-01 7.52360642e-01
3.09120238e-01 9.73143816e-01 1.54276595e-01 7.63059378e-01
2.09340096e-01 9.66574371e-01 5.30356288e-01 -1.66314995e+00
5.34325659e-01 -3.39332297e-02 4.14323360e-01 -4.36043024e-01
-5.42491198e-01 -6.43353224e-01 -6.38204336e-01 1.14355242e+00
8.91413689e-01 -7.93296635e-01 -5.64133108e-01 1.74405336e+00
2.86668450e-01 8.35248381e-02 4.60528851e-01 9.77001071e-01
6.98016360e-02 1.98987350e-01 1.48613736e-01 -5.24011195e-01
7.74442375e-01 -8.27421248e-01 -7.16223836e-01 -5.99205494e-01
4.65769976e-01 -5.62650681e-01 8.65055084e-01 3.86198997e-01
-1.29030704e+00 -1.53713688e-01 -8.62599969e-01 6.13803566e-01
1.02726914e-01 -2.28678375e-01 7.21804500e-01 2.28457361e-01
-7.89400518e-01 7.17488468e-01 -9.70905542e-01 -8.76820907e-02
5.22041380e-01 4.75561291e-01 -1.39827253e-02 5.63003197e-02
-9.64705944e-01 1.44923675e+00 3.17500144e-01 -2.61559606e-01
-1.38003957e+00 -3.46393913e-01 -7.08382487e-01 -2.90477537e-02
7.13976085e-01 -4.27110583e-01 1.73299789e+00 -1.57412887e+00
-2.13872194e+00 -2.63937533e-01 3.42849493e-01 -4.72374260e-01
7.03580320e-01 -2.06009030e-01 1.11128397e-01 1.40089601e-01
-3.34630795e-02 2.76490301e-01 1.35394573e+00 -1.16176009e+00
-7.03925729e-01 -4.96906377e-02 5.45646325e-02 8.24979484e-01
2.28934184e-01 -3.77608724e-02 3.71120214e-01 -5.79481542e-01
-4.19230014e-01 -9.57299948e-01 -9.35855687e-01 -2.84014076e-01
3.10204118e-01 -1.23455770e-01 6.76566899e-01 -3.23753983e-01
5.38181126e-01 -2.02671027e+00 3.44413072e-01 4.13818151e-01
1.16506346e-01 2.83017337e-01 -6.38173163e-01 4.18491125e-01
-2.98222080e-02 -4.15781468e-01 2.14609519e-01 -1.46093056e-01
-4.40296978e-02 5.25327623e-01 -2.85582840e-01 8.98024499e-01
2.63197720e-01 7.48567343e-01 -1.41898048e+00 -2.17802003e-01
1.70723677e-01 2.54472226e-01 -4.54607666e-01 2.40648985e-01
-3.56938392e-01 6.61517322e-01 -8.07705939e-01 3.69001836e-01
5.20116054e-02 -1.20053843e-01 3.46613795e-01 7.53802896e-01
-3.05993948e-02 6.61014169e-02 -1.59399176e+00 1.44286406e+00
-4.07322705e-01 2.49845043e-01 6.24233186e-01 -9.77994621e-01
6.66839123e-01 5.81901848e-01 8.03328872e-01 -9.08965051e-01
6.38000742e-02 5.13694942e-01 1.89852133e-01 -2.15145335e-01
3.37981671e-01 -3.56812328e-01 -6.55797347e-02 7.47238040e-01
7.14661181e-02 -5.24025977e-01 2.86068171e-01 1.92002907e-01
1.33199513e+00 2.33020201e-01 4.91296023e-01 -2.03685716e-01
1.25718340e-02 3.55653814e-03 7.68032730e-01 9.97712791e-01
-3.01968694e-01 1.23588562e-01 7.11195409e-01 -4.23262864e-01
-1.15862906e+00 -6.74324274e-01 3.75872642e-01 1.34122729e+00
-2.22760364e-01 -1.87189341e-01 -1.38623029e-01 -6.91020966e-01
3.30751985e-01 2.30708450e-01 -3.61649632e-01 2.63174605e-02
-7.09085405e-01 -6.03519082e-01 3.08570206e-01 4.98203188e-01
-2.27989163e-02 -1.15687394e+00 -1.04212475e+00 8.33545268e-01
4.79535848e-01 -6.83028996e-01 -4.14893091e-01 6.63012862e-01
-9.04150963e-01 -1.18234026e+00 -1.00500512e+00 -3.99591058e-01
6.02816522e-01 1.29586682e-01 1.02005768e+00 -5.68259172e-02
-2.08230242e-02 8.30467701e-01 -2.94599771e-01 -5.08600175e-01
-5.61591744e-01 -2.21425563e-01 4.90690172e-01 -5.45190036e-01
-4.10063684e-01 -6.07114851e-01 -2.11361557e-01 1.19677462e-01
-6.09695733e-01 -4.74251390e-01 8.45608771e-01 1.09931374e+00
3.51267487e-01 -2.94849947e-02 9.92855072e-01 -6.38159096e-01
1.28580737e+00 -4.59998012e-01 -1.13939214e+00 -1.70634359e-01
-6.22205675e-01 3.93999010e-01 9.14532363e-01 -1.02415740e+00
-7.15692163e-01 4.18019891e-01 2.07197756e-01 -3.73949766e-01
1.96050912e-01 3.38880688e-01 6.91516042e-01 -3.09558779e-01
9.91462350e-01 -2.08972350e-01 5.74546635e-01 -2.34259978e-01
5.17102182e-01 1.51179671e-01 2.39789132e-02 -7.63206303e-01
8.67767513e-01 1.43528387e-01 3.55015635e-01 -6.83961570e-01
-6.53982878e-01 -2.79727131e-01 -1.17447607e-01 -2.81324267e-01
2.78724909e-01 -1.28398955e+00 -1.10034287e+00 6.16847351e-02
-8.22374046e-01 -1.04899371e+00 -1.01355350e+00 8.61480296e-01
-1.10401332e+00 1.48904920e-01 -5.03500104e-01 -1.31946158e+00
-2.67342124e-02 -1.12715292e+00 6.14917099e-01 1.87186629e-01
-1.24134913e-01 -1.14193821e+00 5.57914972e-01 -4.11338568e-01
7.13071227e-01 3.68336767e-01 2.07628876e-01 -5.32875776e-01
-3.26144397e-01 3.33353840e-02 1.88927248e-01 3.57033581e-01
-7.98332989e-02 -4.09505159e-01 -6.10032976e-01 -8.07173491e-01
-3.72115634e-02 -9.99196768e-01 5.95676780e-01 4.39171910e-01
2.08111256e-01 -6.94226623e-01 3.03059906e-01 1.53490514e-01
1.27746785e+00 -5.44517338e-02 7.68316314e-02 4.56978083e-01
5.07299602e-01 1.58232078e-01 7.88024426e-01 8.86602163e-01
6.68502003e-02 4.35544640e-01 6.23156428e-01 1.29086487e-02
-1.18641350e-02 -7.90196359e-02 9.32626486e-01 7.03441978e-01
-4.24905241e-01 3.56310695e-01 -6.29607201e-01 2.95470655e-01
-2.42373037e+00 -1.27503073e+00 4.15573925e-01 2.10644770e+00
1.02553797e+00 -6.24889396e-02 6.90831482e-01 -1.63384363e-01
1.56257868e-01 2.15400398e-01 -1.03353739e+00 -4.06198353e-01
-1.73388347e-01 2.00645402e-02 9.26739216e-01 7.27840185e-01
-7.71935225e-01 7.90776849e-01 7.75924015e+00 7.58113444e-01
-9.83049810e-01 1.42205596e-01 2.44316891e-01 -6.39810205e-01
1.24407396e-01 -1.25266746e-01 -4.88622397e-01 2.65584260e-01
9.73233879e-01 -2.70952702e-01 1.21980178e+00 1.03034997e+00
4.93141919e-01 -3.75988573e-01 -8.19337010e-01 8.72580647e-01
-2.14750677e-01 -9.10857677e-01 -6.35736287e-01 6.37306795e-02
1.03272581e+00 5.81941545e-01 9.83657241e-02 6.68031752e-01
1.36500239e+00 -1.13298345e+00 8.20751727e-01 2.03679636e-01
6.35265708e-01 -9.04991090e-01 5.50304830e-01 6.14136934e-01
-1.01445782e+00 -6.03881896e-01 -5.88818789e-01 -4.73564059e-01
-5.33677712e-02 1.91064134e-01 -8.47369671e-01 1.88550308e-01
1.66397974e-01 1.00111556e+00 -2.24462390e-01 1.05428696e+00
-4.16938424e-01 3.70588750e-01 -4.93590564e-01 -1.06465980e-01
6.93696856e-01 -1.93163738e-01 7.66580522e-01 7.35357642e-01
1.51189014e-01 -2.58590113e-02 6.29561424e-01 3.31815511e-01
3.44338626e-01 -1.01534426e-01 -6.57365322e-01 -5.49740903e-02
2.43179828e-01 1.10992527e+00 -4.00807172e-01 -1.99681535e-01
-2.83105224e-01 3.41129839e-01 6.89789712e-01 7.33524740e-01
-4.08217967e-01 8.75158235e-02 7.00351000e-01 -2.77681112e-01
3.69503915e-01 -5.25694668e-01 2.55175501e-01 -1.04408526e+00
-3.41975391e-01 -1.46318471e+00 4.12487596e-01 -2.02585310e-01
-1.39435458e+00 3.06311756e-01 -1.29084259e-01 -1.34358442e+00
-9.01081026e-01 -2.94566214e-01 -3.85065138e-01 6.76231980e-01
-1.69455099e+00 -6.19254768e-01 4.40491945e-01 7.97164619e-01
5.37812173e-01 -6.11625433e-01 1.00977206e+00 -9.05795544e-02
-3.64836156e-01 1.58670172e-01 6.34999812e-01 -3.66212785e-01
6.02373719e-01 -1.53523159e+00 -3.84165272e-02 4.84568506e-01
-1.85794622e-01 2.04429194e-01 9.15385365e-01 -5.60862303e-01
-1.71734190e+00 -8.41300786e-01 -9.58722234e-02 -3.00935209e-01
1.01629448e+00 -1.48262093e-02 -4.94989753e-01 5.89262724e-01
2.08037853e-01 3.82609129e-01 2.17258930e-01 2.30832666e-01
-2.19736114e-01 2.02217679e-02 -1.03246605e+00 4.90318239e-01
4.59508330e-01 -2.09578097e-01 -3.31842810e-01 4.32147205e-01
2.99049139e-01 -5.53949416e-01 -8.20451558e-01 -2.33724937e-01
4.45114940e-01 -5.85938752e-01 8.92856300e-01 -7.36869156e-01
-1.01857670e-01 -2.96721995e-01 2.09605053e-01 -1.96468532e+00
-3.78583819e-01 -1.21841288e+00 -4.91345882e-01 4.35926616e-01
2.02257529e-01 -7.66553700e-01 5.36150098e-01 4.79473948e-01
2.76739120e-01 -4.78352129e-01 -1.28078389e+00 -1.15612376e+00
1.69756085e-01 -2.00856850e-01 1.49530932e-01 7.20936775e-01
2.66479224e-01 3.87570500e-01 -7.82817602e-01 -2.21958384e-01
8.07025611e-01 -3.37783068e-01 6.95290983e-01 -1.20528221e+00
-7.08757281e-01 -4.68922794e-01 -2.66362950e-02 -9.14656103e-01
3.05820048e-01 -4.13634926e-01 2.90936112e-01 -1.17029250e+00
-1.95864365e-02 -8.11233401e-01 -2.54519165e-01 4.89923954e-01
-2.57258534e-01 -1.00896925e-01 6.40559733e-01 3.15857023e-01
-1.19742155e+00 8.07829797e-01 1.54282081e+00 5.17729251e-03
-4.28945065e-01 1.94896877e-01 -5.50489366e-01 7.23369241e-01
9.11914051e-01 -7.97173560e-01 -5.26242077e-01 -2.70143509e-01
5.06160855e-01 4.07748759e-01 1.89210400e-01 -8.84992480e-01
-2.25135144e-02 -7.25193083e-01 3.53877604e-01 2.17056096e-01
4.03391093e-01 -9.65490460e-01 -1.65360615e-01 6.20075881e-01
-6.44904435e-01 2.41843879e-01 2.32351750e-01 9.27543402e-01
7.24202469e-02 -3.25236052e-01 9.35451865e-01 -5.86155295e-01
-2.53756702e-01 1.21015489e-01 -6.76147878e-01 5.65448403e-01
9.80900049e-01 3.08141768e-01 -1.83723405e-01 -8.58650029e-01
-4.96534348e-01 4.86437678e-01 3.43110055e-01 6.63946941e-02
4.86154616e-01 -1.36064827e+00 -7.21433699e-01 -1.35981152e-02
-4.06569839e-01 3.71227711e-02 -3.32877576e-01 8.60329151e-01
2.44976394e-02 3.90540548e-02 -4.26122993e-01 -1.64258465e-01
-1.03259110e+00 3.03326905e-01 3.47936779e-01 -6.08879209e-01
-8.62249732e-01 4.92231190e-01 -4.06505346e-01 -1.21639274e-01
5.96240520e-01 -1.22106083e-01 -1.87641442e-01 -1.21213887e-02
7.03416824e-01 5.29646158e-01 -5.50642967e-01 -3.28058541e-01
-1.14170806e-02 3.24631602e-01 4.69441675e-02 -5.15921474e-01
1.68787396e+00 1.18119717e-01 3.58041376e-01 4.28670764e-01
6.20065629e-01 -1.64676920e-01 -2.16785598e+00 -2.68050730e-01
-2.13457234e-02 -5.30288100e-01 8.36085975e-02 -6.54984236e-01
-9.11224544e-01 3.23813975e-01 3.85006875e-01 2.69732028e-01
5.22559643e-01 -3.08559179e-01 3.25507969e-01 8.43549371e-01
3.46401960e-01 -1.55922771e+00 4.92536038e-01 9.86362457e-01
1.02436316e+00 -1.27374113e+00 5.28989673e-01 5.72758079e-01
-1.26949835e+00 1.16227555e+00 2.63015896e-01 -5.82631350e-01
2.23716736e-01 4.57727492e-01 2.37395704e-01 -9.49165672e-02
-1.05080247e+00 -4.27009344e-01 -1.86617851e-01 8.26450109e-01
-7.30239972e-02 3.26410681e-02 2.54476577e-01 -2.88396347e-02
2.55564392e-01 -2.46207759e-01 7.01553702e-01 1.08263600e+00
-6.59328103e-01 -1.13190532e+00 -4.07581866e-01 4.22798425e-01
-3.77009422e-01 2.78972685e-01 -9.56496596e-02 6.71103895e-01
-5.78571975e-01 9.24493253e-01 -3.22373003e-01 2.93077976e-01
3.39572519e-01 -3.41586322e-01 5.64179540e-01 -6.54579401e-01
-9.42296922e-01 4.42297786e-01 1.46653086e-01 -1.02002645e+00
-5.36576033e-01 -9.03009355e-01 -1.37873614e+00 -1.78680718e-01
-1.99557051e-01 2.45808840e-01 6.10196769e-01 8.51243377e-01
1.47022307e-01 6.53785586e-01 7.04417408e-01 -1.19857609e+00
-1.53637135e+00 -8.81217122e-01 -8.62222433e-01 2.84676969e-01
9.60169494e-01 -8.19038570e-01 -6.22768760e-01 -4.63720858e-01] | [3.923616886138916, 2.0741233825683594] |
c4d97546-a232-4788-8492-aa0ca1b1129e | micro-expression-action-unit-detection | 1907.05023 | null | https://arxiv.org/abs/1907.05023v2 | https://arxiv.org/pdf/1907.05023v2.pdf | Micro-expression Action Unit Detection with Spatio-temporal Adaptive Pooling | Action Unit (AU) detection plays an important role for facial expression recognition. To the best of our knowledge, there is little research about AU analysis for micro-expressions. In this paper, we focus on AU detection in micro-expressions. Microexpression AU detection is challenging due to the small quantity of micro-expression databases, low intensity, short duration of facial muscle change, and class imbalance. In order to alleviate the problems, we propose a novel Spatio-Temporal Adaptive Pooling (STAP) network for AU detection in micro-expressions. Firstly, STAP is aggregated by a series of convolutional filters of different sizes. In this way, STAP can obtain multi-scale information on spatial and temporal domains. On the other hand, STAP contains less parameters, thus it has less computational cost and is suitable for micro-expression AU detection on very small databases. Furthermore, STAP module is designed to pool discriminative information for micro-expression AUs on spatial and temporal domains.Finally, Focal loss is employed to prevent the vast number of negatives from overwhelming the microexpression AU detector. In experiments, we firstly polish the AU annotations on three commonly used databases. We conduct intensive experiments on three micro-expression databases, and provide several baseline results on micro-expression AU detection. The results show that our proposed approach outperforms the basic Inflated inception-v1 (I3D) in terms of an average of F1- score. We also evaluate the performance of our proposed method on cross-database protocol. It demonstrates that our proposed approach is feasible for cross-database micro-expression AU detection. Importantly, the results on three micro-expression databases and cross-database protocol provide extensive baseline results for future research on micro-expression AU detection. | ['Guoying Zhao', 'Xiaohua Huang', 'Yante Li'] | 2019-07-11 | null | null | null | null | ['action-unit-detection'] | ['computer-vision'] | [ 2.49562636e-01 -4.22776997e-01 -2.05787435e-01 -4.10487860e-01
-8.34218144e-01 -2.69881755e-01 2.56635159e-01 -8.49407986e-02
-4.82827127e-01 5.00274718e-01 -1.52535051e-01 3.42633039e-01
3.16915274e-01 -7.18103409e-01 -5.21082222e-01 -1.10474849e+00
-1.69761851e-01 -4.81411278e-01 6.03949428e-02 -5.11329591e-01
-1.77193806e-01 5.91391206e-01 -1.70084774e+00 5.87198853e-01
5.37973821e-01 1.58296347e+00 -8.18824545e-02 3.39589298e-01
6.89487606e-02 6.87943399e-01 -6.81169748e-01 -4.57017303e-01
9.79136601e-02 -7.58623362e-01 -3.81165773e-01 4.24093977e-02
2.69920468e-01 -2.91597784e-01 -8.03904310e-02 1.19230282e+00
8.83470535e-01 -9.31208953e-02 6.51465893e-01 -1.16920245e+00
-2.14278772e-01 -6.83114305e-02 -1.02723253e+00 2.30685845e-01
4.05403435e-01 4.38634902e-02 9.62671101e-01 -9.03194904e-01
5.73001981e-01 1.09108949e+00 6.43283486e-01 5.80341041e-01
-1.08642197e+00 -9.92714643e-01 -5.07964604e-02 8.65630656e-02
-1.67176485e+00 -5.10602593e-01 5.77746809e-01 -2.96049833e-01
9.91029859e-01 3.16513151e-01 4.80096459e-01 9.63290274e-01
2.27329731e-01 1.09632838e+00 1.18749142e+00 -3.66574675e-01
-4.06275280e-02 1.32174328e-01 -4.56207916e-02 8.87789249e-01
-5.80457628e-01 -1.82548255e-01 -4.85057324e-01 -2.74346203e-01
6.25272751e-01 -1.17913701e-01 -2.21166342e-01 2.72894353e-01
-6.08503163e-01 5.22404671e-01 3.05229664e-01 5.85689783e-01
-3.85512739e-01 -8.80630240e-02 9.05209064e-01 5.16008615e-01
6.78071916e-01 -1.68821514e-02 -3.01919162e-01 -3.64962280e-01
-5.30525923e-01 1.01559766e-01 3.17017972e-01 8.10529113e-01
9.53200817e-01 -3.02909259e-02 -3.28678668e-01 1.42004800e+00
-2.05545977e-01 2.80831605e-01 5.52911520e-01 -4.25152302e-01
1.97919175e-01 6.68699920e-01 -2.72004783e-01 -1.24210882e+00
-5.78374684e-01 -1.14355169e-01 -9.14185464e-01 1.79514959e-01
2.63284475e-01 -4.32485700e-01 -4.89427447e-01 2.02667570e+00
2.75806129e-01 2.03737020e-01 -6.79573324e-03 1.02963471e+00
8.89878869e-01 7.22375095e-01 1.53024152e-01 -6.04987264e-01
1.52306020e+00 -5.79787850e-01 -8.61349761e-01 2.45039374e-01
1.04851961e+00 -7.05592394e-01 1.08831811e+00 3.26401204e-01
-7.19270885e-01 -5.14436424e-01 -8.56823683e-01 2.06778973e-01
-2.72666752e-01 5.93258440e-01 6.98590398e-01 6.19112074e-01
-7.01662540e-01 2.74115711e-01 -5.63314497e-01 -4.36820209e-01
5.73218942e-01 5.98977685e-01 -8.05769086e-01 2.72402942e-01
-1.40574682e+00 6.45306468e-01 8.51021707e-02 4.15064305e-01
-4.37211961e-01 -4.73171353e-01 -7.50781476e-01 -2.09032357e-01
1.83956578e-01 1.60407946e-02 1.06881368e+00 -1.66683245e+00
-1.68153071e+00 1.12811911e+00 -1.73081651e-01 6.71042725e-02
2.19373241e-01 1.53759299e-02 -6.64437413e-01 1.80127949e-01
-1.55072764e-01 4.25065488e-01 7.27074564e-01 -6.00280583e-01
-7.08415627e-01 -5.84893942e-01 1.49801485e-02 1.17864013e-01
-5.63283265e-01 6.25368536e-01 -5.59827447e-01 -6.95349991e-01
-1.66632622e-01 -7.98460007e-01 2.11821303e-01 1.95733860e-01
4.22341190e-02 -4.78426188e-01 7.08845079e-01 -4.49435443e-01
1.41281784e+00 -2.51347280e+00 -8.04395154e-02 2.77433306e-01
-1.58248916e-02 3.79933506e-01 -2.38892436e-01 1.09863333e-01
-2.31443003e-01 -8.99821594e-02 -1.53996393e-01 -2.03167528e-01
-1.35701045e-01 1.20140523e-01 5.80165274e-02 6.14901543e-01
4.10017550e-01 8.32119882e-01 -5.56487083e-01 -6.71816707e-01
-9.49742645e-02 3.03005397e-01 -3.10058475e-01 2.97644734e-01
9.91708711e-02 2.23091543e-02 -5.45912802e-01 1.13028908e+00
8.90430450e-01 2.93723524e-01 -1.22437514e-02 -4.51679379e-01
3.73138748e-02 -4.71151918e-01 -9.79962885e-01 1.46234810e+00
-5.54402709e-01 5.20681739e-01 4.43076879e-01 -1.02860785e+00
1.10665429e+00 3.97524089e-01 5.45553744e-01 -9.57604468e-01
5.50066769e-01 3.12152654e-01 -1.25440240e-01 -7.27135181e-01
5.84642077e-03 -4.13126796e-01 -1.07077718e-01 3.21649283e-01
1.28880367e-01 4.25822139e-01 2.21247822e-01 -1.50057971e-01
9.96742249e-01 -1.84212497e-03 3.70411932e-01 -1.44621268e-01
7.46406138e-01 -3.14898580e-01 8.76013517e-01 1.60056517e-01
-6.19459331e-01 4.36028391e-01 8.42544317e-01 -3.39244604e-01
-3.64144176e-01 -8.01972210e-01 -3.92841786e-01 1.32823932e+00
-7.74971768e-02 -4.94352043e-01 -8.17759037e-01 -7.73295641e-01
-1.13633543e-01 -7.72473812e-02 -1.03666687e+00 -9.44649726e-02
-3.06672513e-01 -1.27550089e+00 9.87729490e-01 5.19257247e-01
6.80307984e-01 -1.24760807e+00 -3.46448690e-01 5.18958941e-02
-1.18701674e-01 -1.01120234e+00 -4.64698642e-01 -3.13706622e-02
-4.67625082e-01 -1.06052172e+00 -7.54176080e-01 -9.73731816e-01
5.84914923e-01 4.00944948e-02 6.43462718e-01 -8.88120607e-02
-5.19802153e-01 -3.46092507e-02 -3.79772514e-01 -4.64963853e-01
-1.66354537e-01 -2.27875948e-01 1.60270631e-01 6.52312100e-01
9.08954382e-01 -4.63188589e-01 -5.03869236e-01 4.33661819e-01
-9.37878370e-01 -4.41534400e-01 7.39342749e-01 9.38585460e-01
7.82027602e-01 -1.30240455e-01 7.20887423e-01 -6.43415213e-01
8.64917576e-01 -3.83071870e-01 -4.16723520e-01 2.10966364e-01
-1.05837062e-01 -4.85583335e-01 5.25174379e-01 -6.73929632e-01
-1.18194366e+00 2.32615694e-01 -5.38484871e-01 -3.93773854e-01
-9.87223536e-02 4.18001741e-01 -3.47450614e-01 -2.77482688e-01
6.86392069e-01 2.78469056e-01 4.31864142e-01 -2.64808059e-01
-2.95225028e-02 1.07267439e+00 3.92902523e-01 -4.01335359e-01
-5.77684455e-02 4.79863793e-01 -3.68836299e-02 -1.18506682e+00
-4.97008055e-01 -4.58844841e-01 -4.30657893e-01 -2.19932050e-01
8.35838318e-01 -1.14240861e+00 -8.09207022e-01 8.99141848e-01
-8.70706856e-01 -1.28149450e-01 3.54992598e-01 2.73604870e-01
-4.06765670e-01 4.69105303e-01 -8.86465013e-01 -7.29511738e-01
-5.91495812e-01 -1.14132631e+00 1.20254958e+00 2.02797353e-01
-4.00360674e-01 -5.39561033e-01 8.15137625e-02 7.17574805e-02
1.06262878e-01 4.44486797e-01 5.05644560e-01 -1.37483031e-01
8.23765546e-02 -5.12958467e-01 -4.35247421e-01 6.90728426e-01
3.38316500e-01 1.82682380e-01 -1.24681199e+00 -2.63946950e-02
-1.22487925e-01 -6.97443306e-01 6.98114455e-01 2.64554381e-01
1.23044217e+00 -8.60618651e-02 -1.30727440e-01 5.62259018e-01
1.06856048e+00 2.90539354e-01 8.25202584e-01 1.77018955e-01
4.43839490e-01 7.88901627e-01 1.15941966e+00 6.57579124e-01
-2.56648093e-01 1.02265525e+00 1.79371655e-01 -4.68693227e-01
2.83214331e-01 1.34673432e-01 5.74199140e-01 2.70546198e-01
-2.33784169e-01 1.45811558e-01 -2.68926233e-01 4.87549961e-01
-1.72872436e+00 -9.83295560e-01 -5.50193526e-02 1.76497257e+00
9.75386918e-01 -2.54111797e-01 4.84443516e-01 -8.24462399e-02
3.98598850e-01 1.41158581e-01 -3.00572366e-01 -8.05145502e-01
-5.48287809e-01 3.32770079e-01 6.23007901e-02 4.46702875e-02
-1.27359807e+00 9.30000603e-01 5.81675720e+00 1.39690900e+00
-1.47324765e+00 2.23868549e-01 8.11329246e-01 -2.76908845e-01
2.68879801e-01 -6.36195540e-01 -6.85993552e-01 5.28527677e-01
4.98677462e-01 5.89683056e-02 -1.60262913e-01 9.96036649e-01
2.82186419e-01 -1.62361696e-01 -7.65816391e-01 1.35891366e+00
9.81585160e-02 -9.99149024e-01 -2.75041312e-01 -1.73284456e-01
5.05864263e-01 -1.59353226e-01 -1.18048042e-01 2.96118379e-01
-3.67885798e-01 -9.80929613e-01 2.67529875e-01 9.13483351e-02
1.05825722e+00 -1.06149530e+00 1.18265128e+00 5.17485291e-02
-1.15641022e+00 1.10625431e-01 -5.43448091e-01 -1.12207770e-01
-1.19002461e-01 3.98010641e-01 -4.06610936e-01 3.24924082e-01
8.46362114e-01 7.45830595e-01 -2.23990619e-01 2.45327607e-01
-1.30400002e-01 4.74712104e-01 -5.48288941e-01 -2.95412630e-01
2.71072268e-01 -3.65800381e-01 -4.77137836e-03 1.68091559e+00
2.05572650e-01 4.42599148e-01 -1.21678114e-01 6.16159022e-01
8.56420142e-04 7.82077551e-01 -6.10267997e-01 -7.02614058e-03
-1.95159033e-01 1.48962855e+00 -2.24124923e-01 -9.13799554e-02
-7.13684678e-01 1.24901450e+00 4.24023360e-01 2.12164029e-01
-8.47609878e-01 -5.09595156e-01 1.11451113e+00 -2.00249374e-01
-7.66106241e-04 2.95734316e-01 1.82106748e-01 -1.05217898e+00
1.92927510e-01 -9.80202794e-01 6.58538401e-01 -4.97688681e-01
-1.25225413e+00 7.56872952e-01 -2.63703078e-01 -1.31494093e+00
-9.68270674e-02 -7.52416313e-01 -5.82573175e-01 8.14997315e-01
-1.19582558e+00 -1.18561697e+00 -3.13961685e-01 8.04350555e-01
2.79979110e-01 -1.20836042e-01 1.12102902e+00 5.49180210e-01
-1.01455200e+00 1.20804369e+00 -9.36115459e-02 4.28860813e-01
7.92777359e-01 -6.44149661e-01 -3.69373024e-01 7.17845738e-01
-6.15495183e-02 4.48424727e-01 3.92526984e-01 -2.91402996e-01
-1.24522567e+00 -8.72465193e-01 5.98623395e-01 3.00519541e-02
7.04805017e-01 -5.35892904e-01 -8.91417384e-01 3.95366400e-01
-1.19094409e-01 3.00307661e-01 9.47828114e-01 5.39864525e-02
-2.66021997e-01 -5.32424867e-01 -1.25346756e+00 5.30403614e-01
8.34253848e-01 -5.32166421e-01 3.09448186e-02 2.47274056e-01
4.17757407e-02 -2.41005883e-01 -1.14626288e+00 5.51593244e-01
1.00289702e+00 -1.12071931e+00 5.51361322e-01 -4.50350016e-01
4.36211765e-01 -2.20510915e-01 -2.76820451e-01 -1.05251729e+00
-1.25262141e-01 -5.92466414e-01 2.57541418e-01 1.41285396e+00
3.46784711e-01 -5.49937010e-01 7.76929617e-01 3.64711136e-01
1.05884902e-01 -1.13264394e+00 -1.00595534e+00 -6.54448211e-01
-2.81740010e-01 -4.61984873e-01 3.48806053e-01 8.72040868e-01
4.06354219e-01 2.66819268e-01 -5.03707945e-01 -1.20609470e-01
1.80914909e-01 3.31890792e-01 9.09584761e-01 -4.90376770e-01
-2.14021221e-01 -5.33654273e-01 -8.69271934e-01 -8.31717968e-01
3.44609886e-01 -6.62500918e-01 6.45256490e-02 -7.51699507e-01
2.53167421e-01 -2.47541174e-01 -4.88869399e-01 6.81398034e-01
-1.97194919e-01 7.18650758e-01 -2.68772542e-01 -5.55967055e-02
-3.15267235e-01 7.18421578e-01 1.09915686e+00 -2.18920819e-02
-2.98902273e-01 -6.42108843e-02 -5.06316721e-01 5.96878946e-01
9.22045112e-01 -1.00995250e-01 -2.60490298e-01 -3.68953794e-02
-1.14201657e-01 -2.19617009e-01 -1.17864475e-01 -5.72138548e-01
-2.36140937e-01 -1.40872106e-01 2.58711994e-01 -2.36953259e-01
4.41471428e-01 -5.46902716e-01 -2.65955955e-01 1.93840593e-01
1.23290926e-01 -2.65802965e-02 5.67360699e-01 2.47117415e-01
-7.24951029e-01 1.08208187e-01 1.04512393e+00 5.26097119e-02
-1.20173132e+00 4.82368410e-01 -5.51522970e-01 -3.19711626e-01
1.25540709e+00 -3.78570199e-01 6.77510574e-02 -3.32602918e-01
-5.70022881e-01 1.51259139e-01 2.50203907e-01 4.22316581e-01
5.71055532e-01 -1.42213774e+00 -7.12741792e-01 3.76569539e-01
7.48122871e-01 -4.34307277e-01 5.81012666e-01 1.22698319e+00
-5.46101391e-01 -9.53880325e-02 -5.40475070e-01 -6.13038957e-01
-2.02673197e+00 2.94562131e-01 5.53574562e-01 4.01404612e-02
-1.93220600e-01 1.15696311e+00 6.43952191e-01 -2.63928682e-01
1.31421700e-01 -1.38142880e-03 -2.78452933e-01 3.73014987e-01
9.10588622e-01 7.45762587e-02 1.27764091e-01 -1.04062808e+00
-4.38557744e-01 7.37787366e-01 -7.98767656e-02 3.21128398e-01
1.18132472e+00 7.69258803e-03 -5.34533799e-01 2.23852202e-01
1.77080584e+00 1.45026185e-02 -8.99615526e-01 -6.28753826e-02
-3.25590074e-01 -4.30907249e-01 -8.39672014e-02 -3.90898168e-01
-1.18378317e+00 8.68515134e-01 8.46625865e-01 -3.74268860e-01
1.68249738e+00 3.51922214e-02 6.23671353e-01 1.64444029e-01
1.42530695e-01 -1.26922274e+00 -6.85975328e-03 5.67420162e-02
9.61870968e-01 -1.17207658e+00 -2.11236253e-01 -6.60834372e-01
-7.79666781e-01 1.04234040e+00 1.04958582e+00 9.37934872e-03
6.62218571e-01 4.25806403e-01 5.13003588e-01 -2.85495907e-01
-5.46345413e-01 -2.60754108e-01 -9.88049246e-03 3.71777534e-01
6.23489439e-01 7.79481530e-02 -5.74556231e-01 7.14993656e-01
1.14984542e-01 2.05632895e-01 1.01119488e-01 8.64573479e-01
-1.74867004e-01 -9.90548253e-01 -1.54151872e-01 4.53552991e-01
-9.55057919e-01 1.71249583e-01 -5.94076037e-01 8.64053667e-01
2.64049143e-01 8.12979281e-01 1.01268999e-01 -6.36458278e-01
6.38468206e-01 -3.26734455e-03 5.26293397e-01 -2.02002794e-01
-5.48854053e-01 4.14273649e-01 2.33309433e-01 -8.70580316e-01
-5.72184205e-01 -5.28189898e-01 -1.23898995e+00 -2.50785381e-01
-2.75244206e-01 -4.79789935e-02 1.34451568e-01 6.48310125e-01
3.61888349e-01 4.30301651e-02 7.11080134e-01 -3.95497948e-01
-2.61163414e-01 -1.25122523e+00 -1.07480729e+00 8.19920003e-01
1.66512765e-02 -7.47575402e-01 -2.55026400e-01 -1.86579704e-01] | [13.627649307250977, 1.7319908142089844] |
7d4c9fcb-fbf2-4568-979f-2103de12093f | understanding-quantum-machine-learning-also | 2306.13461 | null | https://arxiv.org/abs/2306.13461v1 | https://arxiv.org/pdf/2306.13461v1.pdf | Understanding quantum machine learning also requires rethinking generalization | Quantum machine learning models have shown successful generalization performance even when trained with few data. In this work, through systematic randomization experiments, we show that traditional approaches to understanding generalization fail to explain the behavior of such quantum models. Our experiments reveal that state-of-the-art quantum neural networks accurately fit random states and random labeling of training data. This ability to memorize random data defies current notions of small generalization error, problematizing approaches that build on complexity measures such as the VC dimension, the Rademacher complexity, and all their uniform relatives. We complement our empirical results with a theoretical construction showing that quantum neural networks can fit arbitrary labels to quantum states, hinting at their memorization ability. Our results do not preclude the possibility of good generalization with few training data but rather rule out any possible guarantees based only on the properties of the model family. These findings expose a fundamental challenge in the conventional understanding of generalization in quantum machine learning and highlight the need for a paradigm shift in the design of quantum models for machine learning tasks. | ['Carlos Bravo-Prieto', 'Jens Eisert', 'Elies Gil-Fuster'] | 2023-06-23 | null | null | null | null | ['memorization'] | ['natural-language-processing'] | [ 5.06813765e-01 1.99718237e-01 -2.14110896e-01 -3.16059738e-01
-4.17414695e-01 -6.72531486e-01 8.50056946e-01 2.76830196e-01
-5.79139233e-01 7.90008545e-01 -2.05338478e-01 -7.79463053e-01
-3.70361209e-01 -1.19198632e+00 -6.19975686e-01 -9.70939398e-01
-2.68575698e-02 4.96514380e-01 1.01479180e-01 -4.80945200e-01
5.09299636e-01 3.45908582e-01 -1.43653727e+00 -2.18177643e-02
8.11453581e-01 6.82753801e-01 -3.98230523e-01 7.00999439e-01
1.27287328e-01 5.10108173e-01 -1.82040870e-01 -4.56548303e-01
3.96857828e-01 -8.61547351e-01 -9.33840930e-01 -5.11335850e-01
4.82248127e-01 -1.50072183e-02 -7.89970875e-01 1.41890633e+00
2.07190305e-01 3.02089363e-01 7.83592284e-01 -9.23642397e-01
-9.33744729e-01 6.48011446e-01 2.44186953e-01 4.70265538e-01
-1.22882605e-01 1.63799480e-01 1.39888942e+00 -1.16690367e-01
6.18036747e-01 7.66527593e-01 6.04099691e-01 1.00133705e+00
-1.47986245e+00 -5.54722607e-01 -5.33187807e-01 3.97525400e-01
-1.31725597e+00 -3.87837887e-01 2.79779434e-01 -3.13549489e-01
9.75632787e-01 9.56301913e-02 4.67797756e-01 8.14777017e-01
3.36235434e-01 -5.04665971e-02 1.35298860e+00 -8.57554615e-01
5.24780095e-01 -4.94165234e-02 4.92448509e-01 8.81423712e-01
7.44064748e-01 6.88250601e-01 -5.47514617e-01 -8.01263899e-02
6.45036757e-01 -2.39772230e-01 -2.85103858e-01 -4.58757609e-01
-1.15507174e+00 1.05357730e+00 7.14213550e-01 5.11534512e-01
1.04291543e-01 4.33202803e-01 3.64488512e-01 4.16255981e-01
-4.39266264e-02 6.57939374e-01 -2.94296265e-01 -1.32668414e-03
-5.63927054e-01 1.75141647e-01 6.33277953e-01 5.86754680e-01
1.03580403e+00 -1.54801548e-01 2.76396573e-01 2.07454730e-02
-7.62990266e-02 6.00120008e-01 4.46840972e-01 -7.75971651e-01
1.68603852e-01 1.60005108e-01 -1.07650701e-02 -3.24183673e-01
-5.70331037e-01 -6.38715267e-01 -8.40864658e-01 1.73179824e-02
4.85861659e-01 1.83154136e-01 -7.93180645e-01 2.03443837e+00
-2.53047705e-01 1.04306199e-01 4.12702858e-01 5.28334379e-01
3.11128110e-01 4.02013451e-01 3.38030979e-02 -2.70657986e-01
7.15068817e-01 -4.03640389e-01 -2.76793808e-01 1.69262886e-01
1.24817729e+00 -2.14048505e-01 9.95742857e-01 3.34050745e-01
-7.57481277e-01 -3.56663525e-01 -1.40419173e+00 2.40713060e-01
-3.70128661e-01 -5.30533969e-01 1.21144807e+00 1.28332412e+00
-9.32546496e-01 1.40363014e+00 -9.05900121e-01 -4.73271102e-01
2.46425033e-01 6.04606926e-01 -4.38176215e-01 -5.38429022e-02
-1.15551198e+00 1.01810074e+00 7.70200133e-01 -1.62966717e-02
-3.10930967e-01 -2.33002827e-02 -4.76442695e-01 8.77005327e-03
-4.64246087e-02 -8.30578506e-01 1.23679578e+00 -5.83360791e-01
-1.47483993e+00 8.05844665e-01 -1.41933501e-01 -7.58812726e-01
-7.19083548e-02 4.23320502e-01 -5.31422794e-01 3.65398675e-02
-3.17320347e-01 4.35915232e-01 2.51707911e-01 -7.56709039e-01
-2.19111100e-01 -3.58585536e-01 2.10040003e-01 -3.11259508e-01
-2.06272945e-01 -7.05316424e-01 6.93309605e-01 2.25315839e-01
7.15465784e-01 -1.11927521e+00 -2.56235540e-01 -6.78098619e-01
-1.18152089e-01 -3.39917332e-01 -8.19981024e-02 3.87353271e-01
1.16811788e+00 -1.88923287e+00 6.77094832e-02 4.15593058e-01
1.20782025e-01 2.39838541e-01 -1.80321530e-01 5.91470599e-01
-2.10872710e-01 4.01004344e-01 -9.26134586e-02 2.02972382e-01
4.58545238e-01 4.19973135e-01 -5.84158421e-01 7.36265242e-01
1.86223947e-02 1.12934923e+00 -9.68636394e-01 -2.67658057e-03
1.88505992e-01 -5.99105516e-03 -8.84202003e-01 -4.00356412e-01
-2.13267714e-01 4.68831748e-01 -3.62332344e-01 -2.76508611e-02
5.44989824e-01 -6.06179655e-01 3.12897921e-01 -7.62117803e-02
-5.22533357e-02 9.61924076e-01 -8.33593786e-01 1.40778184e+00
-1.79560944e-01 5.23758590e-01 -6.15040898e-01 -1.29580510e+00
5.07957757e-01 7.32032955e-02 2.70684715e-02 -8.77622783e-01
3.13231170e-01 5.94568253e-01 6.31956100e-01 -2.95497864e-01
4.09494162e-01 -8.21679354e-01 -2.50974268e-01 7.99508572e-01
4.12086099e-01 -2.22436592e-01 1.47354722e-01 7.16654062e-02
9.90264297e-01 -7.32204467e-02 1.62628472e-01 -4.41296488e-01
2.91764170e-01 -1.20672099e-01 2.03066990e-01 1.45792317e+00
-2.04444349e-01 2.78245389e-01 3.67675275e-01 -6.73100770e-01
-1.30062938e+00 -1.17102778e+00 -6.51654243e-01 8.52494061e-01
3.13325763e-01 -7.00636744e-01 -7.86618173e-01 -2.83800483e-01
-4.24526244e-01 6.96298599e-01 -8.81685078e-01 -6.24388456e-01
-3.41174811e-01 -1.26252186e+00 6.53958857e-01 2.53361370e-02
2.31069311e-01 -7.60393202e-01 -5.05287349e-01 -8.60793218e-02
2.92897701e-01 -1.03840268e+00 4.26041245e-01 6.69587374e-01
-1.10835946e+00 -9.42722440e-01 -7.61046410e-02 -4.42280740e-01
5.49163461e-01 -6.08284697e-02 8.25636089e-01 2.27972060e-01
-7.39005432e-02 1.18603319e-01 -2.91312963e-01 5.87432608e-02
-1.06868494e+00 3.54739100e-01 6.16315901e-01 -6.07068419e-01
7.68077791e-01 -7.60596633e-01 -5.51828861e-01 -1.80234667e-02
-9.45632279e-01 -1.87301993e-01 5.41137338e-01 9.71635759e-01
1.19002044e-01 2.09676046e-02 4.43880171e-01 -7.22869456e-01
4.77912366e-01 -2.87358075e-01 -4.36895192e-01 3.53439182e-01
-8.09157252e-01 8.05638969e-01 7.27092087e-01 -2.13675320e-01
-2.98394501e-01 -4.21054810e-01 -1.76826045e-01 3.74764442e-01
-1.96167618e-01 4.46669579e-01 4.84687477e-01 -5.45638442e-01
1.23220384e+00 3.82977277e-01 -1.76076382e-01 1.07173428e-01
4.95519459e-01 5.61195791e-01 2.84983635e-01 -7.77950764e-01
7.05480039e-01 3.70259762e-01 6.98922098e-01 -7.91630566e-01
-1.07269645e+00 -8.98833200e-02 -8.08047771e-01 3.51117581e-01
7.38905132e-01 -5.05736232e-01 -1.01704752e+00 1.03169389e-01
-8.31918895e-01 -2.40107536e-01 -3.76012415e-01 6.90550089e-01
-8.29812109e-01 5.95153451e-01 -7.81172454e-01 -8.32399189e-01
-9.78033915e-02 -9.03322339e-01 4.90503550e-01 3.01642567e-01
-5.97897582e-02 -9.35640335e-01 2.25109845e-01 1.27408430e-01
6.15973055e-01 -1.90385297e-01 1.32034957e+00 -7.72448957e-01
-8.51761580e-01 -3.39004576e-01 -6.29402995e-02 1.75591975e-01
-1.39951959e-01 -1.29047215e-01 -1.00780499e+00 -2.63243347e-01
-1.18659109e-01 -5.92945039e-01 1.21238875e+00 2.94632688e-02
8.54436874e-01 -4.75571044e-02 -1.12480968e-02 6.26690388e-01
1.51871049e+00 -1.97188333e-01 6.30680561e-01 2.91899025e-01
4.00593579e-01 1.76706955e-01 -3.25900853e-01 9.29682776e-02
8.51679742e-02 3.90703291e-01 3.49255443e-01 6.46749198e-01
2.79310405e-01 -2.96388805e-01 4.73713130e-02 1.12276912e+00
-4.30119574e-01 -2.40376797e-02 -8.39397788e-01 -2.04898734e-02
-1.50365949e+00 -1.17556477e+00 -1.10656023e-01 2.67616844e+00
6.56432986e-01 5.60663164e-01 -2.48701200e-02 2.38409683e-01
4.33389872e-01 -8.83488730e-02 -3.05958629e-01 -6.32007062e-01
-3.00120771e-01 9.43171620e-01 7.02158213e-01 5.66880167e-01
-8.91974688e-01 1.03557885e+00 6.96049881e+00 7.58375823e-01
-1.16148973e+00 2.74077713e-01 2.20612302e-01 1.51138932e-01
-5.56091666e-01 4.46684510e-01 -6.89287424e-01 1.41836479e-01
1.46780550e+00 -1.37351856e-01 7.27328062e-01 2.95992553e-01
-4.28183526e-01 -9.65173692e-02 -1.30712640e+00 8.87236178e-01
-2.07916096e-01 -1.63562703e+00 1.63068190e-01 3.07451725e-01
9.02004063e-01 4.47222561e-01 1.60099849e-01 4.95676666e-01
1.08578227e-01 -1.33718204e+00 5.78300536e-01 2.86325306e-01
6.32393956e-01 -5.10939240e-01 6.23514593e-01 5.03015339e-01
-4.33585465e-01 -1.48567662e-01 -7.81928897e-01 -4.89209265e-01
-1.21970326e-01 5.19434325e-02 -4.43983167e-01 3.14852417e-01
2.88006682e-02 1.72460049e-01 -5.67552090e-01 8.61627340e-01
-1.19573347e-01 6.66779220e-01 -5.50933480e-01 -5.45941174e-01
4.09686208e-01 -4.72605288e-01 6.45499900e-02 8.57720017e-01
3.63431066e-01 3.18992883e-01 -3.47275972e-01 8.43785048e-01
1.48818363e-03 -5.88348834e-03 -6.32274568e-01 -5.12999058e-01
3.74050260e-01 6.82792842e-01 -8.96245837e-01 -2.31269673e-02
-2.79093683e-01 8.12452614e-01 4.57748950e-01 1.79413974e-01
-5.15839040e-01 -2.30737537e-01 1.42061278e-01 -7.57986158e-02
3.18023473e-01 -7.16516435e-01 -4.38209087e-01 -1.56095421e+00
-4.88070697e-02 -4.64562625e-01 2.63468232e-02 -1.98603928e-01
-1.14204776e+00 5.38512528e-01 -2.01457113e-01 -8.99228632e-01
-3.42361510e-01 -1.01090133e+00 -5.23308933e-01 6.86312318e-01
-1.11329138e+00 -7.15126514e-01 2.49130875e-01 1.65981784e-01
-4.67099071e-01 -8.42100382e-02 1.42861784e+00 -1.01168223e-01
-3.60016257e-01 6.50635660e-01 5.60257554e-01 -1.38438288e-02
2.74946034e-01 -1.00299764e+00 4.99288887e-01 6.63152695e-01
7.35852718e-01 1.10705566e+00 9.76428032e-01 -6.64270222e-02
-1.37637413e+00 -2.82543540e-01 8.92921984e-01 -6.88273013e-01
9.26053882e-01 -4.03180569e-01 -7.08875239e-01 4.50884014e-01
-1.21313743e-01 2.34045699e-01 9.32742119e-01 6.34379804e-01
-7.44920194e-01 2.24804088e-01 -8.30226779e-01 4.62780833e-01
1.16065860e+00 -1.07642436e+00 -5.57661176e-01 4.86047596e-01
4.25862610e-01 1.93825178e-02 -5.85417032e-01 5.35596728e-01
6.43581986e-01 -1.21369994e+00 6.71002686e-01 -1.11159301e+00
3.56280833e-01 1.52776642e-02 -5.26029289e-01 -1.00503421e+00
-2.81186968e-01 -5.90262651e-01 9.21777114e-02 3.60637546e-01
5.64629495e-01 -7.97918677e-01 9.02260721e-01 6.27007127e-01
1.38540104e-01 -6.17990732e-01 -1.26003623e+00 -9.24410999e-01
8.81771863e-01 -4.92023677e-01 2.79815048e-01 9.97165978e-01
4.53479946e-01 4.94311988e-01 1.04316138e-01 3.31754051e-02
6.99054480e-01 1.84096023e-01 3.13759476e-01 -1.38558590e+00
-5.36593914e-01 -7.42684424e-01 -1.04710448e+00 -9.63641226e-01
3.05893570e-01 -1.29463947e+00 -2.39611045e-01 -9.98798013e-01
3.26677740e-01 -6.91100538e-01 -5.74935794e-01 -3.61302256e-04
9.39405188e-02 4.41795886e-01 -9.81244594e-02 2.37503603e-01
-8.06204855e-01 4.48217809e-01 8.60056818e-01 1.94657966e-01
-3.04087084e-02 4.37037498e-02 -6.02975607e-01 4.77084696e-01
9.51058567e-01 -5.60827613e-01 -2.52792627e-01 -2.00357944e-01
9.29605305e-01 -2.76287347e-01 5.65003991e-01 -1.34049618e+00
2.57940352e-01 -8.52996781e-02 1.19936749e-01 6.69625178e-02
1.77783221e-01 -1.93161681e-01 -2.25738496e-01 7.12261021e-01
-4.43481743e-01 -2.58182019e-01 -1.45353109e-01 5.29207945e-01
2.51067281e-01 -7.75404155e-01 8.87087166e-01 -1.89350381e-01
-4.91106749e-01 2.81662107e-01 -3.10090274e-01 4.75315005e-02
6.01669550e-01 -9.15459096e-02 -6.98729575e-01 -2.43293419e-01
-9.20834780e-01 -4.24510241e-01 7.54054964e-01 -6.84904680e-02
6.05057254e-02 -1.12598395e+00 -3.62602532e-01 2.58279145e-01
2.82634974e-01 -6.08231723e-01 2.14423940e-01 6.90950096e-01
-6.48272932e-01 9.04036701e-01 -2.36780301e-01 -4.95982289e-01
-4.82630044e-01 6.66944027e-01 5.92131197e-01 8.68922994e-02
-2.73740828e-01 7.63997912e-01 -1.42104566e-01 -4.39751297e-01
-2.27990806e-01 -3.74032140e-01 4.36481416e-01 -4.43108857e-01
1.72675282e-01 2.68972814e-02 1.46083042e-01 -6.56409085e-01
-1.51405126e-01 5.87878585e-01 1.55438073e-02 -2.01864913e-01
1.08699608e+00 1.08638711e-01 -1.79771259e-01 4.88892555e-01
1.16017008e+00 -1.48132995e-01 -6.49166405e-01 -3.97428900e-01
4.34434712e-02 2.44847126e-02 -1.09146431e-01 -3.26833218e-01
-9.33072194e-02 1.32855165e+00 5.53367913e-01 6.66655719e-01
5.90329289e-01 2.19929039e-01 6.65903986e-01 1.22250056e+00
9.56086636e-01 -9.06278789e-01 -2.47490361e-01 7.76815593e-01
-2.21222639e-02 -1.24833381e+00 -1.15101233e-01 -3.09152976e-02
1.03273600e-01 1.45718145e+00 1.41310915e-01 -3.26149702e-01
5.16870797e-01 -3.18854809e-01 -2.55697906e-01 -1.26569614e-01
-7.08708823e-01 -4.67281938e-01 1.00384541e-01 3.13113481e-01
5.04641652e-01 3.29669863e-01 -3.40939641e-01 3.88974130e-01
-7.28001356e-01 -1.70529008e-01 7.63405561e-01 6.98925853e-01
-9.69881833e-01 -1.53206861e+00 1.56808913e-01 2.67004192e-01
-3.96832913e-01 -3.12858194e-01 -1.13455281e-01 5.99602222e-01
2.90145576e-01 6.75778270e-01 -6.09160215e-02 -7.51067519e-01
-1.89115316e-01 4.76032078e-01 1.37800670e+00 -6.66995645e-01
-7.77124912e-02 -5.97571254e-01 -2.26792291e-01 -6.47137910e-02
-3.86805713e-01 -4.40778404e-01 -1.21014869e+00 -7.86522388e-01
-6.45139575e-01 5.29194176e-01 6.94062829e-01 1.28621340e+00
1.34825110e-01 -5.40803969e-02 3.38595331e-01 -6.16985798e-01
-1.31360054e+00 -8.08637261e-01 -5.53101957e-01 3.50964129e-01
4.55394179e-01 -4.31304544e-01 -4.86891866e-01 -3.49365294e-01] | [5.60796594619751, 4.979931354522705] |
9d087765-19dd-4e68-86e8-5923790f59fe | unsupervised-adversarial-domain-adaptation-3 | 2003.02244 | null | https://arxiv.org/abs/2003.02244v2 | https://arxiv.org/pdf/2003.02244v2.pdf | Unsupervised Adversarial Domain Adaptation for Implicit Discourse Relation Classification | Implicit discourse relations are not only more challenging to classify, but also to annotate, than their explicit counterparts. We tackle situations where training data for implicit relations are lacking, and exploit domain adaptation from explicit relations (Ji et al., 2015). We present an unsupervised adversarial domain adaptive network equipped with a reconstruction component. Our system outperforms prior works and other adversarial benchmarks for unsupervised domain adaptation. Additionally, we extend our system to take advantage of labeled data if some are available. | ['Hsin-Ping Huang', 'Junyi Jessy Li'] | 2020-03-04 | unsupervised-adversarial-domain-adaptation-2 | https://aclanthology.org/K19-1064 | https://aclanthology.org/K19-1064.pdf | conll-2019-11 | ['implicit-discourse-relation-classification', 'implicit-relations'] | ['natural-language-processing', 'natural-language-processing'] | [ 3.20827991e-01 9.34015930e-01 -5.35707474e-01 -2.64089048e-01
-4.90150601e-01 -9.94565248e-01 8.97421658e-01 9.96709913e-02
-3.99418116e-01 1.44099057e+00 4.98260796e-01 -3.29167426e-01
1.09615646e-01 -8.95653367e-01 -5.28436542e-01 -3.08745861e-01
2.13423118e-01 9.90497231e-01 2.10241139e-01 -7.48143256e-01
-2.27398589e-01 2.40403295e-01 -7.38548994e-01 2.74963409e-01
1.05120218e+00 6.00139856e-01 -3.89675915e-01 5.96380889e-01
1.71470061e-01 1.46790814e+00 -8.24429452e-01 -6.53568149e-01
3.84790008e-03 -2.44349375e-01 -1.56698275e+00 -1.31433010e-01
8.64408687e-02 -4.99917686e-01 -6.42186463e-01 8.08983684e-01
4.70801473e-01 4.51741934e-01 1.07244754e+00 -1.13888848e+00
-1.18092334e+00 8.22090685e-01 1.61407683e-02 6.17288277e-02
5.36364615e-01 1.34708583e-01 9.89494860e-01 -7.30313301e-01
1.23328292e+00 1.09249783e+00 6.30933285e-01 8.30742180e-01
-1.48728681e+00 -6.65764689e-01 3.51401806e-01 5.85181355e-01
-1.09991896e+00 -6.22016966e-01 9.97222900e-01 -5.48219323e-01
1.10427535e+00 9.20336843e-02 1.82463564e-02 1.81411326e+00
-6.33724809e-01 6.08315587e-01 1.08659065e+00 -4.91800398e-01
2.81576719e-02 1.54111102e-01 2.35936865e-01 1.87257096e-01
-1.24867439e-01 3.57169092e-01 -1.55086786e-01 -5.94192185e-02
5.91086268e-01 -5.40857613e-01 -2.44497851e-01 -2.67837644e-01
-8.63346457e-01 9.83531594e-01 6.78690612e-01 2.56021649e-01
-1.47551924e-01 -2.85583854e-01 7.08949447e-01 8.46811831e-01
8.03620338e-01 1.02463365e+00 -5.86203992e-01 -2.16695204e-01
-4.32012260e-01 3.60671848e-01 1.24168098e+00 1.21661055e+00
7.18783319e-01 1.84984468e-02 -2.03052565e-01 8.96844208e-01
-1.52241156e-01 1.83780462e-01 4.13540542e-01 -1.24468052e+00
6.63218260e-01 4.94648337e-01 1.88364625e-01 -8.60060930e-01
-3.75157177e-01 -2.49039140e-02 -9.16310787e-01 3.41615453e-02
7.23942637e-01 -3.78139108e-01 -6.85065210e-01 1.74553013e+00
1.75646335e-01 3.14367443e-01 4.83577192e-01 9.63491023e-01
1.35392499e+00 2.28994876e-01 3.51485521e-01 -1.47735432e-01
9.26529527e-01 -1.05323470e+00 -9.67106044e-01 -3.56638908e-01
6.73774123e-01 -5.40321827e-01 8.42009544e-01 1.43149152e-01
-8.38442266e-01 -2.75189191e-01 -8.85140181e-01 -3.58109921e-01
-4.33812857e-01 -2.96503752e-01 5.80425978e-01 2.42339700e-01
-5.92222571e-01 5.50716281e-01 -6.33351028e-01 -1.41701445e-01
4.39934343e-01 2.73516297e-01 -4.98114377e-01 1.25522524e-01
-1.79370785e+00 1.44780135e+00 7.60119379e-01 -2.70722240e-01
-7.28631198e-01 -4.89258617e-01 -1.25079572e+00 -3.67816359e-01
6.82510078e-01 -4.97092038e-01 1.41252446e+00 -1.26081467e+00
-1.76102269e+00 1.10001051e+00 4.70504090e-02 -8.13738585e-01
6.00243568e-01 -5.53127646e-01 -4.15172070e-01 1.92251250e-01
-1.66162979e-02 3.28371763e-01 5.54446459e-01 -1.35585701e+00
5.28002381e-02 9.15718079e-02 8.34694684e-01 2.84930348e-01
-3.42111290e-01 3.28286178e-02 1.99350223e-01 -9.30771053e-01
-3.95219207e-01 -8.46038103e-01 -2.69132525e-01 -2.41416886e-01
-7.44295120e-01 -4.22553718e-01 1.07791698e+00 -8.94323051e-01
9.97011483e-01 -1.86331820e+00 5.30043304e-01 -4.84466925e-02
3.88345271e-01 7.07995415e-01 -5.81726693e-02 2.20135659e-01
-1.88844353e-01 2.54044801e-01 -4.45772260e-01 -2.63623625e-01
1.24003187e-01 7.62148619e-01 -3.85272801e-01 2.85170943e-01
7.75744975e-01 8.99138808e-01 -1.08844376e+00 -6.26401722e-01
2.87590958e-02 7.91900158e-02 -6.37698948e-01 5.98022282e-01
-6.12490475e-01 9.96263266e-01 -4.59360361e-01 4.25071806e-01
4.51683193e-01 -1.90315634e-01 6.38705373e-01 -3.96686010e-02
3.96345347e-01 8.33819151e-01 -7.83275664e-01 1.43814206e+00
-4.64458466e-01 9.47693646e-01 8.31221342e-02 -1.60923624e+00
1.05321777e+00 7.84903646e-01 1.72744676e-01 -4.32522446e-01
2.62460291e-01 5.30560315e-02 6.31323904e-02 -6.05597496e-01
6.33212745e-01 -3.91812027e-01 -3.89928907e-01 4.49001580e-01
3.93491894e-01 -3.67898911e-01 2.68015474e-01 2.99838156e-01
1.23774517e+00 3.83010477e-01 7.05060303e-01 1.01553679e-01
6.10393167e-01 1.57512784e-01 6.73988163e-01 5.27190924e-01
-3.01261872e-01 5.35655975e-01 8.07217002e-01 -2.35811964e-01
-1.07299817e+00 -1.10046184e+00 -2.47329295e-01 1.14211559e+00
1.56379387e-01 -2.16643512e-01 -3.73330861e-01 -1.30130219e+00
3.24344635e-03 7.93482602e-01 -8.89099598e-01 -6.40067533e-02
-8.75347972e-01 -8.84807929e-02 8.29492569e-01 7.56321132e-01
3.70429456e-01 -1.35699189e+00 3.40916477e-02 1.85753316e-01
-4.38290536e-01 -1.47146714e+00 -9.17059407e-02 4.65232611e-01
-6.09446645e-01 -1.29561639e+00 -3.36213768e-01 -8.53412986e-01
6.16765678e-01 -5.90905070e-01 1.62061286e+00 -4.08510379e-02
4.16141480e-01 3.06926757e-01 -7.42810547e-01 -2.83193290e-01
-9.42395866e-01 5.11907697e-01 4.94299084e-02 -6.23503625e-01
4.26367015e-01 -8.06967556e-01 -2.53613479e-02 3.90172213e-01
-6.55251920e-01 -1.91386104e-01 1.35662913e-01 1.38287354e+00
4.12692666e-01 -2.91709751e-01 9.67877090e-01 -1.71567738e+00
7.39811420e-01 -7.41319120e-01 -7.71694109e-02 2.37153517e-03
-3.67701530e-01 2.48358343e-02 8.23450506e-01 -6.99692607e-01
-1.33244491e+00 -2.63659328e-01 -1.89381391e-01 -4.32159841e-01
-5.11677802e-01 5.21620214e-01 -2.84472793e-01 9.36216563e-02
1.28938651e+00 -5.40986717e-01 -1.29767463e-01 -4.23255235e-01
8.17247510e-01 5.78312278e-01 6.92544401e-01 -1.11762273e+00
1.19973123e+00 4.95484322e-02 -4.17380989e-01 -3.86396199e-01
-1.23196769e+00 -2.27864385e-01 -1.16612923e+00 2.29428411e-01
7.92562783e-01 -8.99222016e-01 -1.97344959e-01 -1.62711404e-02
-1.46974242e+00 -8.54981780e-01 -6.76993191e-01 2.52231181e-01
-7.08282650e-01 2.55942017e-01 -8.75090837e-01 -4.73850518e-01
-1.10773802e-01 -7.91699946e-01 3.13985676e-01 2.10280176e-02
-9.49954271e-01 -1.45549738e+00 1.57414854e-01 3.17546159e-01
2.97603041e-01 6.93048656e-01 5.27244270e-01 -1.32420361e+00
-3.76523100e-02 1.19253516e-01 -2.58541137e-01 5.48949838e-01
3.51422459e-01 -2.70014167e-01 -1.00965345e+00 5.54444008e-02
-1.98929295e-01 -1.11281073e+00 6.31675184e-01 -2.00094551e-01
9.57125604e-01 -9.11489487e-01 -2.87339032e-01 5.44209063e-01
7.95048416e-01 -1.51229054e-01 6.00980997e-01 7.17910409e-01
7.87351966e-01 8.00950229e-01 8.68769169e-01 1.09958053e-01
5.41511595e-01 7.05319703e-01 2.63589561e-01 -1.37335271e-01
-2.02460393e-01 -4.03105617e-02 2.41516829e-01 6.68565929e-01
-4.36366647e-01 -1.98327228e-01 -1.02149868e+00 7.47353315e-01
-1.78562295e+00 -1.10021019e+00 -7.32027367e-02 1.36250496e+00
1.72589946e+00 2.94128895e-01 -9.28587280e-03 3.93303633e-02
5.65828919e-01 1.78659528e-01 -6.88781023e-01 -7.21004009e-01
-3.55803758e-01 9.05010343e-01 2.50860035e-01 5.67547023e-01
-1.70326161e+00 1.14974451e+00 6.60076952e+00 4.12309945e-01
-7.72547245e-01 3.41068536e-01 3.47835243e-01 1.61078796e-01
-3.48296285e-01 -2.64268816e-02 -1.87358588e-01 1.57936215e-01
8.31646025e-01 -1.64673701e-01 2.85302490e-01 9.48039174e-01
-5.23204505e-01 3.40603918e-01 -1.40299141e+00 2.48592764e-01
-1.35296270e-01 -1.10990429e+00 -5.03718674e-01 -2.35127091e-01
1.01115775e+00 -1.01882532e-01 -3.10320593e-02 7.28140831e-01
1.21790254e+00 -1.43259192e+00 3.87953609e-01 2.15042815e-01
9.14669394e-01 -4.69847739e-01 8.78460526e-01 4.39426661e-01
-7.67342329e-01 1.31991789e-01 -1.08954370e-01 -4.97467458e-01
9.50089842e-02 2.37261981e-01 -1.09088957e+00 6.97703958e-01
4.67510194e-01 1.02476203e+00 -3.03088576e-01 2.04605088e-01
-1.09661710e+00 7.77084827e-01 -2.95812279e-01 2.49858007e-01
6.94096927e-03 8.32317397e-02 6.55335426e-01 1.27556121e+00
-3.33720475e-01 5.22369623e-01 3.18461359e-01 8.88776541e-01
-4.17797118e-01 -1.75894901e-01 -8.52834046e-01 -1.13315927e-02
7.35138118e-01 9.57163572e-01 -5.53987734e-03 -3.96234214e-01
-3.24244767e-01 8.83998513e-01 8.88042331e-01 6.33675456e-01
-5.81207573e-01 -2.17110023e-01 6.48926318e-01 -1.95872322e-01
9.51371416e-02 -3.14237684e-01 -4.63421017e-01 -1.22504461e+00
-2.48541594e-01 -1.19500649e+00 6.26810312e-01 -4.85401124e-01
-1.79310870e+00 6.00999236e-01 1.30084187e-01 -1.24519074e+00
-7.85605371e-01 -6.32251561e-01 -4.90129918e-01 9.70180631e-01
-1.65915740e+00 -1.33992994e+00 -2.46080950e-01 6.51693642e-01
2.30088815e-01 -3.63781780e-01 1.32342494e+00 2.41807684e-01
-4.25715983e-01 9.20670509e-01 -6.13290034e-02 8.49872172e-01
1.23834872e+00 -1.42516875e+00 3.74327719e-01 5.27629137e-01
-3.04870550e-02 5.93755305e-01 8.32887232e-01 -5.17198205e-01
-7.25603282e-01 -1.18494821e+00 6.06511295e-01 -8.36759806e-01
1.08758748e+00 -1.92783907e-01 -1.36773789e+00 1.29379737e+00
5.21794021e-01 1.50835469e-01 7.93386042e-01 6.17683709e-01
-6.99081182e-01 4.49442059e-01 -1.32146907e+00 4.62189198e-01
1.14139915e+00 -8.88819039e-01 -1.23675811e+00 3.79719675e-01
1.04528427e+00 -9.87548590e-01 -1.35563350e+00 5.70219815e-01
6.48105145e-02 -4.36990589e-01 1.20597649e+00 -1.04398429e+00
9.89705384e-01 4.89324406e-02 6.01045527e-02 -1.60536361e+00
-8.96129981e-02 -4.73755360e-01 -6.74065173e-01 1.58542061e+00
5.89867890e-01 -8.00721586e-01 7.52057493e-01 7.11673379e-01
-7.94512331e-02 -2.60220617e-01 -8.21987450e-01 -9.08913493e-01
6.88582778e-01 -1.30344257e-01 3.79469424e-01 1.73179209e+00
2.97677517e-01 6.56232417e-01 -2.75134623e-01 2.19981205e-02
2.47983649e-01 1.07636238e-02 9.44724202e-01 -1.10221410e+00
-4.66907620e-01 -3.00443977e-01 -2.57586211e-01 -1.00217474e+00
9.14230525e-01 -1.04408836e+00 4.54889424e-03 -1.30413961e+00
-2.99682438e-01 -8.40415716e-01 -2.04593450e-01 7.19647408e-01
-4.23128337e-01 4.35010046e-01 4.45346870e-02 1.76513985e-01
-6.56459153e-01 6.69924200e-01 1.29633868e+00 -5.03117800e-01
-6.41004220e-02 -1.90021470e-01 -6.51725233e-01 9.25810218e-01
9.22517121e-01 -4.12129253e-01 -3.56269598e-01 -3.77822638e-01
-8.19181949e-02 4.35488857e-02 3.28419298e-01 -6.74540520e-01
-5.51480218e-04 -4.20742601e-01 1.51519507e-01 -1.48854330e-01
4.21529830e-01 -6.76499188e-01 -2.12249339e-01 8.18985999e-02
-7.52544284e-01 -4.89245176e-01 2.38513485e-01 3.86632055e-01
-5.14730871e-01 -4.48494107e-01 6.21640861e-01 -2.19268855e-02
-8.19012582e-01 9.65826884e-02 -2.27577165e-01 6.85177982e-01
1.12970829e+00 2.43005499e-01 -7.98615575e-01 -5.56253672e-01
-1.26926410e+00 3.81270736e-01 4.70367521e-01 2.77098119e-01
3.01595777e-01 -1.50665712e+00 -8.92763615e-01 -4.42620248e-01
1.36854425e-01 5.85538387e-01 -2.40609273e-01 5.41303217e-01
-4.00492519e-01 3.04960869e-02 -2.23849058e-01 -1.91581458e-01
-1.10121953e+00 7.45665371e-01 2.80746073e-01 -7.99848616e-01
-5.68541110e-01 8.88086259e-01 2.30165109e-01 -9.10011232e-01
2.87790447e-02 9.71562341e-02 -5.38081110e-01 1.13439679e-01
1.67162001e-01 1.91234220e-02 -2.71392435e-01 -7.04176605e-01
-2.26697937e-01 2.32414044e-02 -1.82012349e-01 2.61526778e-02
1.43729651e+00 5.35127483e-02 -1.10816415e-02 3.73829663e-01
9.10730779e-01 1.31165572e-02 -1.18098378e+00 -9.61917579e-01
1.62197843e-01 -2.35182673e-01 -5.84339321e-01 -9.21666801e-01
-7.15149641e-01 8.22663605e-01 -3.20812128e-02 4.63317364e-01
1.02374601e+00 2.23516583e-01 5.30025303e-01 5.77323616e-01
7.83276930e-02 -1.16810787e+00 3.79746169e-01 1.20862472e+00
1.12671232e+00 -1.45118082e+00 2.78500259e-01 -7.02250957e-01
-8.39512765e-01 1.03186142e+00 1.11569464e+00 -4.86693591e-01
5.04441619e-01 3.92782897e-01 2.13765308e-01 4.04693708e-02
-7.40792632e-01 -3.65829617e-01 3.00243944e-01 1.42164981e+00
7.06569493e-01 2.12547109e-01 -2.52073467e-01 9.39038336e-01
-4.07613039e-01 -1.73798993e-01 4.01237309e-01 7.00140119e-01
1.36148080e-01 -1.58998299e+00 -2.51923651e-01 2.34982848e-01
-4.38621283e-01 -1.39250502e-01 -7.83522248e-01 9.41935062e-01
4.96153012e-02 9.94257748e-01 -2.28320230e-02 -3.41840267e-01
4.66215372e-01 1.15007006e-01 4.85072523e-01 -9.67941999e-01
-6.41334772e-01 -7.18056858e-01 1.01524258e+00 -1.62422240e-01
-7.71414578e-01 -5.21506310e-01 -1.16576147e+00 -4.50225741e-01
-2.45232478e-01 1.61154404e-01 -8.04322585e-02 1.26485586e+00
-6.20631315e-02 9.25563931e-01 4.73082870e-01 -7.30615020e-01
-4.12421703e-01 -1.31815803e+00 -3.97811048e-02 9.89494860e-01
3.69923651e-01 -1.05177820e+00 -1.41851187e-01 4.24154133e-01] | [10.724157333374023, 9.20433521270752] |
440fcb9d-aa59-48a2-aa0c-a01cf8d28b84 | neural-network-augmented-compartmental | 2212.08481 | null | https://arxiv.org/abs/2212.08481v1 | https://arxiv.org/pdf/2212.08481v1.pdf | Neural Network Augmented Compartmental Pandemic Models | Compartmental models are a tool commonly used in epidemiology for the mathematical modelling of the spread of infectious diseases, with their most popular representative being the Susceptible-Infected-Removed (SIR) model and its derivatives. However, current SIR models are bounded in their capabilities to model government policies in the form of non-pharmaceutical interventions (NPIs) and weather effects and offer limited predictive power. More capable alternatives such as agent based models (ABMs) are computationally expensive and require specialized hardware. We introduce a neural network augmented SIR model that can be run on commodity hardware, takes NPIs and weather effects into account and offers improved predictive power as well as counterfactual analysis capabilities. We demonstrate our models improvement of the state-of-the-art modeling COVID-19 in Austria during the 03.2020 to 03.2021 period and provide an outlook for the future up to 01.2024. | ['Kevin Sidak', 'Lorenz Kummer'] | 2022-12-15 | null | null | null | null | ['epidemiology'] | ['medical'] | [ 3.02928835e-02 -1.77637532e-01 -5.03866494e-01 1.23014584e-01
1.50622860e-01 -4.22038764e-01 9.12796021e-01 2.12579533e-01
-6.42355025e-01 1.04599369e+00 5.45928106e-02 -1.02452004e+00
-6.08210087e-01 -7.62320101e-01 -7.22496986e-01 -7.34180391e-01
-6.31641269e-01 7.80657470e-01 -1.76411301e-01 -1.58953071e-01
-3.67874652e-01 5.62667668e-01 -1.01063728e+00 -8.71123821e-02
7.54948854e-01 3.88150126e-01 8.04800764e-02 6.21816576e-01
1.36795938e-01 5.76549113e-01 -5.86515069e-01 -1.22447446e-01
1.76449761e-01 -5.32208122e-02 1.38515253e-02 -9.73406851e-01
-5.45853674e-01 -6.26450777e-01 -3.82220626e-01 4.50052351e-01
7.47224152e-01 -3.01253319e-01 8.36984098e-01 -1.12980747e+00
-4.97251153e-01 5.83058774e-01 -4.40023035e-01 3.17871720e-01
6.54439405e-02 2.47086406e-01 2.08617169e-02 -1.62361741e-01
6.50195122e-01 1.22378075e+00 1.01453948e+00 6.96737826e-01
-1.44341171e+00 -5.91365814e-01 7.57697970e-02 -1.60151243e-01
-1.10623765e+00 -1.18845955e-01 2.35727429e-01 -7.19886065e-01
1.31292248e+00 4.76176769e-01 7.55963266e-01 1.41402745e+00
8.62929404e-01 4.77807373e-01 1.20834172e+00 -1.10084556e-01
3.88537288e-01 1.85649335e-01 -5.30534238e-03 7.82445148e-02
7.46678591e-01 7.18923092e-01 7.15112463e-02 -7.50931203e-01
1.04253650e+00 3.83679509e-01 -8.04827735e-02 5.06509878e-02
-9.39174533e-01 1.12975144e+00 2.58622289e-01 1.69058621e-01
-1.16300333e+00 3.19434106e-01 3.04275811e-01 -1.60470739e-01
1.00186980e+00 2.40642712e-01 -8.73379290e-01 4.56769109e-01
-9.61466551e-01 7.25086153e-01 6.34650648e-01 3.36236417e-01
-1.39960289e-01 3.74884933e-01 -2.53240854e-01 4.00872529e-01
3.02161723e-01 9.01797950e-01 4.56235930e-02 -8.57637346e-01
-1.12112857e-01 1.08046383e-01 5.01629114e-01 -5.94004333e-01
-9.67152476e-01 -6.92453563e-01 -1.37103307e+00 4.07360047e-02
2.95945078e-01 -7.29033947e-01 -1.03652608e+00 1.58741856e+00
3.52539152e-01 2.41654292e-01 6.47579357e-02 3.38666975e-01
4.26047742e-01 1.06902862e+00 5.14186680e-01 -6.96041107e-01
1.17619205e+00 -2.03944340e-01 -7.87530184e-01 2.97410637e-01
5.81289053e-01 -2.60323733e-01 2.84141954e-02 4.07393992e-01
-1.08020926e+00 1.16818748e-01 -4.37621146e-01 8.94382596e-01
-8.01034391e-01 -2.24187046e-01 8.41319263e-01 8.32837224e-01
-1.12395668e+00 6.32832944e-01 -1.02324367e+00 -2.67986804e-01
2.74342924e-01 5.17433345e-01 2.58125901e-01 2.25380823e-01
-1.56738317e+00 1.11249626e+00 1.48501113e-01 1.00342119e-02
-9.44353282e-01 -1.34707880e+00 -5.74524403e-01 8.64561200e-02
9.00955424e-02 -9.92940009e-01 1.03924358e+00 -4.84903336e-01
-1.27814734e+00 2.18492284e-01 3.30248445e-01 -1.12946272e+00
5.99939883e-01 2.74804950e-01 -4.11266834e-01 -1.15751483e-01
-3.21352631e-01 2.76428163e-01 7.59301931e-02 -9.95626807e-01
-2.97743052e-01 -3.86306524e-01 -2.47676551e-01 -2.58060098e-01
1.60085827e-01 6.41405463e-01 2.74156719e-01 -8.20504427e-01
-8.61669838e-01 -9.79753256e-01 -7.63559639e-01 -3.38949561e-01
-2.96207964e-01 1.29623208e-02 2.44750544e-01 -1.32629800e+00
1.27204823e+00 -1.30267656e+00 -1.91602528e-01 1.30899474e-01
-1.02245308e-01 6.18678331e-01 2.46374551e-02 6.65990055e-01
-2.23994568e-01 2.08410352e-01 -2.82919347e-01 -9.77271125e-02
-1.16791263e-01 9.09134001e-02 -8.48082900e-02 5.63103437e-01
7.78596103e-02 7.88144469e-01 -6.66138589e-01 5.40994853e-02
5.26723921e-01 9.17631209e-01 -4.00985807e-01 -1.17392652e-01
-3.43930364e-01 4.04972255e-01 -4.19113249e-01 1.75053790e-01
8.79096687e-01 -8.16332623e-02 2.56168962e-01 4.54938382e-01
-4.50661957e-01 8.37980881e-02 -6.77018702e-01 8.47752452e-01
-5.17546058e-01 1.61621466e-01 1.64847881e-01 -9.95990872e-01
3.70346248e-01 6.90251231e-01 8.04986358e-01 -4.13765669e-01
1.00672998e-01 6.47976324e-02 3.68239194e-01 -3.45486552e-01
6.19535223e-02 -4.02744740e-01 2.67480910e-01 3.24078739e-01
-3.30635697e-01 3.87241215e-01 3.30602452e-02 -7.25671425e-02
9.04547274e-01 1.91992626e-01 5.71414530e-01 -6.76273704e-01
1.42205907e-02 1.18347712e-01 4.29222077e-01 6.38573170e-01
5.31603843e-02 -9.99346003e-02 5.19171357e-01 -6.35459840e-01
-1.13634801e+00 -9.46634114e-01 -3.73281270e-01 5.72227478e-01
-6.33875132e-01 4.47977036e-01 -8.64408970e-01 -1.95664376e-01
3.20455104e-01 1.05519450e+00 -7.88400888e-01 1.18945599e-01
-5.69461763e-01 -1.63561141e+00 6.51773155e-01 2.29031250e-01
3.08525339e-02 -7.29000151e-01 -7.93413341e-01 5.15844762e-01
4.01742786e-01 -5.17233074e-01 3.14774990e-01 1.11780852e-01
-7.72787273e-01 -1.06403470e+00 -1.36423814e+00 -1.41543197e-02
3.48626256e-01 -2.02747330e-01 7.81536758e-01 -4.18733239e-01
-1.71685457e-01 3.24645229e-02 2.75106907e-01 -1.10084653e+00
-6.01944208e-01 -4.36843008e-01 5.88593364e-01 -4.71809417e-01
1.87025145e-01 -1.90135270e-01 -8.62827182e-01 -5.20536490e-02
-8.13642502e-01 2.23320406e-02 4.04889256e-01 5.40350735e-01
1.99662402e-01 3.18567753e-02 9.75557685e-01 -9.35467899e-01
6.07202649e-01 -8.32586348e-01 -1.05677891e+00 2.55901039e-01
-8.77044201e-01 -2.86437064e-01 6.96227610e-01 -4.65776950e-01
-1.22428584e+00 -1.42256334e-01 -1.34641221e-02 -2.06570491e-01
-9.98442397e-02 9.32093203e-01 4.85145599e-01 3.46376777e-01
4.55949873e-01 1.15672782e-01 3.10926288e-01 -7.05200255e-01
5.19200303e-02 6.80240571e-01 2.29341283e-01 -2.49697119e-01
2.95184642e-01 5.80189168e-01 2.75374591e-01 -8.43118906e-01
1.31945208e-01 -2.05840692e-01 1.33719340e-01 -1.57244295e-01
7.20489144e-01 -1.13071823e+00 -9.24932778e-01 6.45092130e-01
-1.31537986e+00 -5.40060043e-01 5.53082861e-02 8.50026727e-01
-4.83317941e-01 -2.34686121e-01 -8.71698260e-01 -1.49524605e+00
-4.98202413e-01 -9.47920442e-01 4.80690658e-01 4.24862392e-02
-1.55145451e-01 -1.31227934e+00 4.71325845e-01 -2.04000384e-01
9.99675989e-01 7.29202032e-01 1.09420097e+00 -7.63301015e-01
-9.95529145e-02 -3.39605629e-01 3.22645307e-02 -5.62146977e-02
1.42458389e-02 3.39584321e-01 -5.95875561e-01 -2.66319036e-01
2.82103773e-02 7.37295210e-01 5.44235468e-01 1.31874788e+00
6.73308909e-01 -8.88534486e-01 -7.19632447e-01 2.06774056e-01
1.58451569e+00 7.83718944e-01 5.57173789e-01 2.25807726e-01
1.10754438e-01 9.29889977e-01 2.13563219e-01 4.78944838e-01
1.11086667e-01 6.89094841e-01 3.57101858e-01 -4.26236004e-01
4.47588563e-01 1.70356557e-02 9.12397206e-02 2.27560714e-01
-5.54362237e-01 -4.48912352e-01 -1.26298344e+00 5.65849364e-01
-1.72295594e+00 -1.00266767e+00 -4.81634170e-01 2.53163886e+00
7.18360782e-01 -2.35308819e-02 5.87536752e-01 -3.34884018e-01
6.82205379e-01 -2.36410171e-01 -3.13207775e-01 -7.13277698e-01
-4.06619385e-02 1.92090809e-01 1.07633185e+00 2.77452946e-01
-9.03077483e-01 2.49215156e-01 7.42461681e+00 6.97294474e-01
-8.68036091e-01 3.52576345e-01 1.03017497e+00 -2.68736750e-01
-8.51607621e-02 -3.81218284e-01 -4.70882475e-01 6.82598412e-01
1.83861399e+00 -2.01192498e-01 5.32095909e-01 4.72933620e-01
9.81364906e-01 -1.45171285e-01 -5.98961174e-01 2.96164334e-01
-3.27568710e-01 -1.51665926e+00 -9.72519368e-02 4.49068427e-01
7.44855881e-01 4.56399232e-01 2.68852562e-01 2.68499017e-01
5.52559793e-01 -8.56455982e-01 2.24352017e-01 7.90799320e-01
7.91836441e-01 -9.53031719e-01 1.08222854e+00 5.47130466e-01
-3.91788661e-01 3.11479568e-02 -2.48463258e-01 -2.17380837e-01
4.45490181e-01 5.76636672e-01 -6.96087539e-01 5.51920593e-01
5.21114230e-01 2.33030066e-01 2.09691361e-01 1.01978970e+00
3.66637051e-01 7.40435839e-01 -5.10942400e-01 -1.61545113e-01
3.01306456e-01 -2.69980669e-01 4.73084331e-01 1.12762976e+00
4.75049078e-01 -6.29315898e-02 -5.18641472e-01 7.64645576e-01
5.07465720e-01 -1.73370197e-01 -9.17816639e-01 -9.97011662e-02
1.93230048e-01 5.61553538e-01 -4.14231360e-01 -3.67821544e-01
-1.63194373e-01 3.82293195e-01 -5.72013259e-01 6.86816812e-01
-9.98475552e-01 9.73760709e-02 7.65544772e-01 4.26962256e-01
-3.76732647e-02 7.66951740e-02 -6.55061677e-02 -6.20817602e-01
-6.44506216e-01 -7.72500455e-01 3.67206514e-01 -7.82475710e-01
-9.54123735e-01 2.52339929e-01 5.85581541e-01 -7.45719373e-01
-6.60190046e-01 -9.18454707e-01 -3.25361907e-01 1.16574013e+00
-1.33887923e+00 -9.47903514e-01 6.69817507e-01 6.26741126e-02
2.36334935e-01 2.29513105e-02 1.02062559e+00 2.68334508e-01
-7.26178110e-01 -1.69347692e-02 9.72288013e-01 -4.78488654e-01
1.26769394e-01 -9.32295442e-01 4.74856019e-01 5.13174832e-01
-7.26146579e-01 8.93728971e-01 1.09881055e+00 -1.06322551e+00
-1.07336855e+00 -1.18370950e+00 8.86153162e-01 -2.22267970e-01
7.25118995e-01 -1.87764138e-01 -4.37218934e-01 6.32539511e-01
2.89727271e-01 -6.73177958e-01 6.17712498e-01 -1.33997157e-01
4.09459323e-02 6.45719245e-02 -1.32094622e+00 6.95901036e-01
3.59470338e-01 -2.78898608e-02 -7.98761621e-02 6.28416598e-01
7.22689688e-01 -9.21714008e-02 -8.91155124e-01 4.63431597e-01
6.98143482e-01 -5.98401487e-01 1.30263746e+00 -1.00437963e+00
2.27172390e-01 8.63339901e-02 1.44012600e-01 -1.46900415e+00
-2.93452173e-01 -9.26188409e-01 -2.20325112e-01 7.70256639e-01
4.77489442e-01 -1.12964141e+00 3.09580714e-01 5.48465192e-01
3.69194329e-01 -8.84489954e-01 -1.29383314e+00 -1.19328129e+00
3.72580439e-01 -1.57422483e-01 8.64972711e-01 9.18323636e-01
-2.10174605e-01 -1.60307735e-01 -7.85131633e-01 2.02963799e-01
6.60657048e-01 -5.61924934e-01 2.18075722e-01 -1.34354305e+00
-2.99459577e-01 -3.67115110e-01 -4.06001657e-02 -2.25411654e-02
-1.47384286e-01 -2.10297674e-01 -4.14084852e-01 -1.71442747e+00
3.29802722e-01 -3.43400478e-01 -3.46916974e-01 2.64374644e-01
4.08410281e-02 -1.78590342e-01 -5.44455759e-02 3.78661044e-02
4.41223323e-01 1.34037957e-01 4.82797176e-01 -5.15800808e-03
-3.15945864e-01 2.49526054e-01 -2.51866490e-01 6.95179105e-01
1.04121327e+00 -6.34038210e-01 -2.18501776e-01 -4.11795825e-02
2.69362479e-01 4.26880002e-01 6.45879269e-01 -4.71562445e-01
-2.05545306e-01 -5.01432300e-01 3.13040644e-01 -7.45607138e-01
1.35434330e-01 -9.48833525e-01 1.14834416e+00 1.36118793e+00
-7.45145902e-02 2.64712214e-01 5.95536828e-01 7.03745186e-01
6.09759331e-01 1.25460848e-01 4.53478634e-01 -1.21740393e-01
2.92905569e-01 3.51062745e-01 -8.81169319e-01 -6.09616697e-01
1.09318018e+00 1.17989428e-01 -8.05779457e-01 -3.75907272e-01
-6.65379584e-01 2.23026797e-01 2.20675170e-01 1.04896359e-01
8.07690248e-02 -8.48781824e-01 -9.18729365e-01 -3.48012894e-01
-3.61263990e-01 -6.17104769e-01 5.21621644e-01 9.99113917e-01
-8.09167683e-01 1.20459223e+00 -1.19513758e-01 -1.35013282e-01
-9.47298646e-01 1.32673562e+00 5.96446812e-01 -7.53955424e-01
-2.02780887e-01 1.79264545e-01 4.39648539e-01 -4.74005163e-01
-3.36764641e-02 -1.11134596e-01 -2.29775399e-01 -4.40804996e-02
7.08035231e-01 8.45671952e-01 -3.02273452e-01 -5.06906509e-01
-6.18470430e-01 -2.85653435e-02 2.82257289e-01 -1.54623752e-02
1.76311219e+00 7.76143074e-02 -8.34042430e-02 3.15088660e-01
7.16816247e-01 -5.93533337e-01 -1.07544124e+00 2.09353894e-01
-2.07498267e-01 3.86494577e-01 1.72338605e-01 -1.30002320e+00
-6.80882633e-01 6.16556704e-01 8.04295003e-01 4.21722084e-01
9.66586769e-01 -4.70669180e-01 2.78921336e-01 -5.12654707e-02
3.43802601e-01 -7.09102273e-01 -1.07480979e+00 -7.18927607e-02
8.87124419e-01 -5.41212678e-01 2.08179832e-01 -2.52800077e-01
-2.21020266e-01 4.41253155e-01 -1.93377823e-01 -7.12902918e-02
8.15275252e-01 4.39199477e-01 -1.24826044e-01 -9.15143564e-02
-7.67019391e-01 9.63081643e-02 5.62740341e-02 1.17906702e+00
2.20673859e-01 6.55108035e-01 -8.57660592e-01 6.79842114e-01
4.15259510e-01 4.32653695e-01 5.20953000e-01 6.71486020e-01
2.44674146e-01 -7.13620901e-01 -4.60564703e-01 6.47658050e-01
-8.81932020e-01 -4.85893846e-01 -1.07336909e-01 1.12552476e+00
2.92417072e-02 8.33164573e-01 7.59727210e-02 6.74676150e-02
1.74382731e-01 -1.70148499e-02 1.96718976e-01 -2.06506550e-01
-9.30507481e-01 2.79761404e-01 3.00080776e-01 -3.22914898e-01
-4.30967748e-01 -6.85978413e-01 -7.12234735e-01 -7.52878487e-01
-3.45584810e-01 -8.22513774e-02 1.18192041e+00 6.60407424e-01
5.82311273e-01 7.04112887e-01 5.00051081e-01 -8.82648766e-01
-7.16810048e-01 -1.09699249e+00 -6.60034657e-01 -5.85338712e-01
4.32599932e-01 -5.61029375e-01 -5.16500175e-01 -3.84821951e-01] | [6.019142150878906, 4.362648010253906] |
0e2b3aed-1735-4cd4-8049-aed0195103bb | anomalous-event-recognition-in-videos-based | null | null | https://doi.org/10.3390/app11031344 | https://doi.org/10.3390/app11031344 | Anomalous Event Recognition in Videos Based on Joint Learningof Motion and Appearance with Multiple Ranking Measures | Given the scarcity of annotated datasets, learning the context-dependency of anomalous events as well as mitigating false alarms represent challenges in the task of anomalous activity detection. We propose a framework, Deep-network with Multiple Ranking Measures(DMRMs), which addresses context-dependency using a joint learning technique for motion and appearance features. In DMRMs, the spatial-time-dependent features are extracted from a video using a 3Dresidual network(ResNet), and deep motion features are extracted by integrating the Motionflow maps’ information with the 3D ResNet. Afterward, the extracted features are fused for joint learning. This data fusion is then passed through a deep neural network for deep multiple instance learning(DMIL)to learn the context-dependency in a weakly-supervised manner using the proposed multiple ranking measures(MRMs). These MRMs consider multiple measures of false alarms, and the network is trained with both normal and anomalous events, thus lowering the false alarm rate. Meanwhile, in the inference phase, the network predicts each frame’s abnormality score along with the localization of moving objects using motion flow maps. A higher abnormality score indicates the presence of an anomalous event. Experimental results on two recent and challenging datasets demonstrate that our proposed framework improves the area under the curve(AUC)score by6.5%compared to the state-of-the-art method on the UCF-Crime dataset and shows AUC of68.5%on theShanghaiTech dataset. | ['Moongu Jeon', 'Jeonghwan Gwak', 'Abhijeet Boragule', 'Shikha Dubey'] | 2021-02-02 | null | null | null | journal-2021-2 | ['anomaly-detection-in-surveillance-videos', 'anomaly-detection-in-surveillance-videos'] | ['computer-vision', 'methodology'] | [ 1.82605788e-01 -3.11066747e-01 -1.36287436e-01 -4.85178471e-01
-8.93765688e-01 -2.48488747e-02 4.46591824e-01 2.72884607e-01
-5.98912477e-01 5.00411689e-01 7.46614560e-02 1.73683427e-02
-2.65296578e-01 -5.70297062e-01 -5.88660836e-01 -7.78798163e-01
-4.66431886e-01 -1.70532629e-01 5.96997440e-01 1.02522880e-01
2.89818674e-01 5.37841022e-01 -1.53023016e+00 5.86968899e-01
7.41281629e-01 1.43652058e+00 -2.27895468e-01 7.71444738e-01
7.13590309e-02 1.33334041e+00 -8.13836813e-01 6.50715232e-02
4.87049669e-02 -9.14138928e-02 -6.34913027e-01 8.65149349e-02
3.05732101e-01 -6.85969949e-01 -4.80087191e-01 5.44206679e-01
3.23975027e-01 3.51796150e-01 3.52654099e-01 -1.22763729e+00
-4.00535911e-02 4.64393869e-02 -6.31901801e-01 1.16332936e+00
1.07984036e-01 2.12910205e-01 8.11305344e-01 -1.04104233e+00
1.79205477e-01 1.02965319e+00 5.31155348e-01 1.77008733e-01
-7.22666144e-01 -5.74963212e-01 3.86224747e-01 6.44596219e-01
-1.26864314e+00 -2.59860665e-01 7.13518381e-01 -4.72938895e-01
1.10335326e+00 1.33136868e-01 5.65398455e-01 1.19869959e+00
2.96972513e-01 9.11939383e-01 4.78071153e-01 8.44329298e-02
6.24684095e-02 -4.10144389e-01 7.39435926e-02 7.34730721e-01
1.24401614e-01 6.75177798e-02 -6.39085352e-01 -7.40807131e-02
6.43673062e-01 4.37840432e-01 7.66406059e-02 1.42008722e-01
-1.15112340e+00 5.30040205e-01 3.29282403e-01 3.69405746e-01
-5.16014040e-01 4.69718613e-02 6.27056479e-01 1.75732464e-01
7.20654190e-01 8.90656710e-02 -5.30767202e-01 -1.41447559e-01
-1.02873397e+00 1.75409332e-01 2.50652403e-01 2.10405514e-01
4.99753475e-01 2.42283925e-01 -4.08712089e-01 7.36936986e-01
3.42531413e-01 3.30828905e-01 4.36285973e-01 -8.19730401e-01
6.66438878e-01 9.03903961e-01 2.23640725e-02 -1.40228152e+00
-7.30171025e-01 -6.38391554e-01 -9.09281552e-01 1.75789192e-01
3.19588155e-01 -1.89780325e-01 -6.66488528e-01 1.57114625e+00
3.14624637e-01 1.05122674e+00 4.12836522e-02 9.02595580e-01
5.95578313e-01 7.21476674e-01 2.43743271e-01 -3.39088857e-01
1.05855596e+00 -7.49035716e-01 -5.90838730e-01 -1.37681425e-01
8.56526971e-01 -3.16301763e-01 5.04619658e-01 5.14784038e-01
-8.40258181e-01 -8.14703226e-01 -1.01780248e+00 3.71545643e-01
-1.37865111e-01 1.23729274e-01 2.48711646e-01 2.21836325e-02
-6.55800819e-01 7.46909916e-01 -1.18765843e+00 -5.97115755e-02
6.74342096e-01 1.44253135e-01 -3.73481512e-01 -1.25771798e-02
-1.24874783e+00 7.81367004e-01 5.68634808e-01 3.83587390e-01
-1.11785460e+00 -6.44415379e-01 -8.52912128e-01 -7.20971897e-02
3.55300397e-01 -3.37702811e-01 7.80874908e-01 -7.77530670e-01
-8.52082729e-01 5.25174022e-01 -6.03479184e-02 -6.91138566e-01
6.90987289e-01 -5.18815696e-01 -9.31514919e-01 4.96437311e-01
2.39997104e-01 3.47442627e-01 9.89340603e-01 -6.46885335e-01
-1.25798702e+00 -3.59015614e-01 5.19443080e-02 2.17981860e-01
-2.27357581e-01 -3.71205732e-02 -3.76344413e-01 -7.67142177e-01
1.19097427e-01 -5.11905909e-01 -1.22661114e-01 -1.21652462e-01
-2.30367824e-01 -2.72814721e-01 1.07943904e+00 -7.46181846e-01
1.60494566e+00 -2.37540817e+00 -1.59115193e-03 1.61483213e-01
2.69699842e-01 5.41569889e-01 -7.25710839e-02 -8.18815157e-02
-2.58750975e-01 -1.87825769e-01 -2.90354729e-01 -2.44983241e-01
-4.11349714e-01 1.69279173e-01 -1.31768689e-01 6.52523637e-01
6.81114137e-01 6.25407338e-01 -1.11225545e+00 -5.18328130e-01
6.23130918e-01 4.71915990e-01 -4.62707520e-01 3.30154568e-01
1.88447028e-01 7.27658272e-01 -5.28550923e-01 6.65815651e-01
4.52227414e-01 -1.94569156e-01 -2.77685106e-01 -1.90172076e-01
-8.50314423e-02 -9.07440782e-02 -1.21643436e+00 1.61036134e+00
-3.58463109e-01 4.21856314e-01 -4.62310433e-01 -1.28886974e+00
8.63405228e-01 4.83439326e-01 9.29251552e-01 -7.23431289e-01
-6.93213716e-02 2.00743318e-01 -1.56727523e-01 -9.39780712e-01
3.50408256e-02 3.22511375e-01 -7.83950761e-02 1.24412134e-01
1.88769726e-03 8.20283592e-01 1.53637663e-01 4.83887196e-02
1.58777535e+00 1.72160596e-01 7.58757815e-02 8.45654160e-02
9.82424080e-01 -2.64291435e-01 9.54581141e-01 8.14420283e-01
-3.98166090e-01 4.14684057e-01 4.25819695e-01 -9.43793297e-01
-6.51306391e-01 -1.20865047e+00 -7.79900700e-02 8.74460995e-01
1.99861810e-01 -2.30754197e-01 -4.23780233e-01 -9.92735922e-01
-3.13187353e-02 5.36260724e-01 -6.42606437e-01 -5.14874518e-01
-7.52964616e-01 -1.07422340e+00 6.30831182e-01 8.42726171e-01
7.00165093e-01 -1.12878716e+00 -1.00399518e+00 3.05319011e-01
-3.54241937e-01 -1.34153962e+00 -1.18952647e-01 -1.80277660e-01
-8.41397524e-01 -1.18328631e+00 -3.20936233e-01 -3.41924846e-01
5.38798213e-01 1.17759302e-01 7.44265914e-01 2.69773453e-01
-4.43422884e-01 2.52606213e-01 -4.16036725e-01 -5.48834912e-02
-8.92164856e-02 -2.83235848e-01 5.92788793e-02 5.97269058e-01
4.09568280e-01 -6.41222596e-01 -9.01688099e-01 2.39004672e-01
-9.28290784e-01 -2.73739934e-01 6.23446405e-01 7.71408260e-01
6.58867538e-01 1.35173827e-01 9.57513869e-01 -5.15780091e-01
2.63836652e-01 -7.20745623e-01 -5.64544201e-01 8.64672195e-03
-2.89818585e-01 -2.12475702e-01 5.08418679e-01 -3.50579381e-01
-1.04610205e+00 -1.00944206e-01 -2.52849329e-02 -6.99057877e-01
-3.94823343e-01 2.98757046e-01 1.36388438e-02 3.63200814e-01
5.48458397e-01 9.86649245e-02 -3.38431329e-01 -1.87829077e-01
-5.59596382e-02 3.38124335e-01 7.86249936e-01 -7.66590908e-02
5.97859979e-01 7.42910802e-01 2.45670721e-01 -5.79221010e-01
-1.32192433e+00 -7.44191706e-01 -7.51916170e-01 -3.89320582e-01
1.27036929e+00 -9.88006413e-01 -5.88357508e-01 7.49900281e-01
-1.12528634e+00 1.51641652e-01 -1.37819231e-01 7.29362845e-01
-3.78829837e-01 2.76644111e-01 -5.14680684e-01 -8.96907151e-01
-2.55345672e-01 -9.45577860e-01 1.01514602e+00 2.20991805e-01
-1.11552842e-01 -9.65846896e-01 -3.24641988e-02 4.04125780e-01
1.29068926e-01 7.46450841e-01 6.00471914e-01 -1.01045942e+00
-5.80219984e-01 -5.53221822e-01 -3.12090427e-01 4.75606054e-01
2.08304346e-01 -1.77298397e-01 -1.03254437e+00 7.96080288e-03
-6.43174397e-04 9.57622230e-02 1.07554221e+00 5.87505400e-01
1.44665492e+00 -1.81684807e-01 -2.23516598e-01 6.33242071e-01
1.15989482e+00 3.52869779e-01 5.20299137e-01 4.38152194e-01
9.29250956e-01 1.68135330e-01 8.41947734e-01 9.61162746e-01
4.88968104e-01 4.54612285e-01 7.21282244e-01 -1.44999415e-01
2.02257946e-01 1.87798679e-01 3.80106628e-01 4.77117121e-01
-2.01687738e-01 -4.19983380e-02 -8.50431919e-01 5.99463463e-01
-2.22241402e+00 -1.26329982e+00 -2.22569242e-01 2.22908425e+00
1.62635922e-01 4.67656732e-01 1.20291851e-01 4.66523021e-01
7.39757836e-01 4.22506422e-01 -7.38831699e-01 -3.87465619e-02
-1.61483988e-01 -7.09721372e-02 1.36798918e-01 2.47891054e-01
-1.62505364e+00 5.93313634e-01 4.92199755e+00 5.59283555e-01
-9.29319263e-01 2.15265810e-01 8.44177485e-01 -3.23734611e-01
4.84314770e-01 -1.88625440e-01 -6.32103503e-01 6.09212637e-01
1.16132987e+00 2.31754914e-01 1.49285316e-01 4.95197624e-01
5.73522925e-01 -1.24115802e-01 -7.13288426e-01 9.52054560e-01
2.39466935e-01 -1.22185719e+00 -1.63877979e-01 -2.60227650e-01
4.26002651e-01 9.52833965e-02 -1.43169820e-01 4.57947642e-01
-2.74719745e-01 -6.87313378e-01 4.47630107e-01 9.86221492e-01
2.90562749e-01 -1.09990370e+00 9.51641381e-01 3.44374090e-01
-1.28035724e+00 -6.13182247e-01 -1.52330115e-01 1.61129281e-01
3.46680880e-01 8.27629328e-01 -6.05694056e-01 6.24305189e-01
9.84022617e-01 1.20448291e+00 -5.67017198e-01 9.46388602e-01
-1.32676125e-01 6.01754487e-01 -3.03239860e-02 5.97644150e-01
2.46062323e-01 1.94723144e-01 6.93053484e-01 1.22830820e+00
3.57625604e-01 1.45647570e-01 4.98010099e-01 4.60313559e-01
2.12466717e-01 -9.55714956e-02 -4.39618647e-01 2.29226142e-01
1.94407403e-01 1.41807556e+00 -6.67042494e-01 -4.84990805e-01
-5.01186192e-01 8.78039658e-01 2.12678194e-01 2.09585518e-01
-1.14361107e+00 -3.94746482e-01 4.52167541e-01 -1.36796638e-01
3.44890177e-01 -3.19035053e-02 -1.04687817e-01 -1.26725805e+00
2.14901298e-01 -5.18039107e-01 9.18878794e-01 -4.75176334e-01
-1.17676044e+00 6.03693008e-01 1.31749613e-02 -1.66330671e+00
-2.71374047e-01 -4.28701788e-01 -9.94990289e-01 4.91871506e-01
-1.66004050e+00 -9.86870348e-01 -4.54417557e-01 6.62239850e-01
6.94814205e-01 -3.15151632e-01 4.82869446e-01 4.94252086e-01
-1.04495823e+00 3.88404995e-01 -4.74070609e-01 3.47469926e-01
5.57778060e-01 -1.06856096e+00 3.24509770e-01 1.24151480e+00
-1.21012934e-01 -1.83663014e-02 2.32482404e-01 -6.64146662e-01
-8.68241251e-01 -1.55164909e+00 6.65525138e-01 -4.98035669e-01
6.66553140e-01 4.43206243e-02 -1.25990880e+00 3.98020178e-01
-4.09280002e-01 7.35978484e-01 6.20649815e-01 -2.84753352e-01
-2.91247249e-01 -2.12425336e-01 -1.09795475e+00 8.28159451e-02
1.06577206e+00 -3.87548774e-01 -5.63391685e-01 2.94905305e-01
5.89428365e-01 -3.33941221e-01 -9.98761892e-01 8.37075710e-01
3.53957117e-01 -1.11119378e+00 9.89510953e-01 -6.91553354e-01
2.95137018e-01 -4.06755656e-01 -2.23282874e-01 -1.08623230e+00
-2.82971144e-01 -1.51256159e-01 -7.66639531e-01 9.40695584e-01
2.66947210e-01 -3.68564337e-01 3.93788487e-01 2.04098001e-01
-4.03258711e-01 -9.21764135e-01 -1.11764312e+00 -6.05063915e-01
-4.89087343e-01 -7.49754131e-01 4.36062425e-01 9.25799668e-01
-2.10347265e-01 1.47095263e-01 -5.25231540e-01 6.67346597e-01
4.60662931e-01 -3.01648468e-01 3.04393828e-01 -1.33568573e+00
-1.70425922e-01 -1.90011725e-01 -8.54666591e-01 -4.70434934e-01
2.16396004e-01 -7.84784973e-01 -1.47300795e-01 -1.35023713e+00
-3.27207632e-02 1.30148351e-01 -9.98038173e-01 5.84134877e-01
-4.73406464e-01 1.33127034e-01 -8.55762288e-02 4.85254452e-02
-1.10206783e+00 3.90779883e-01 9.18731272e-01 2.82388367e-02
-2.70333886e-01 2.32035458e-01 -1.07486196e-01 9.74310577e-01
6.70772135e-01 -4.67492908e-01 -3.92660916e-01 -4.04383063e-01
-1.75983861e-01 2.97569931e-01 6.87036097e-01 -1.45910966e+00
7.68866166e-02 -4.44638319e-02 9.51963305e-01 -9.27503586e-01
3.62075061e-01 -7.67408490e-01 -4.11024541e-01 5.67359984e-01
-3.31203550e-01 2.97122061e-01 2.01351449e-01 8.57285440e-01
-2.56353676e-01 1.88644767e-01 7.31879115e-01 6.08526319e-02
-9.49093461e-01 6.54161870e-01 -3.76169533e-01 1.35656819e-01
1.11864746e+00 -1.62567735e-01 -3.32934737e-01 -1.61098108e-01
-1.04183996e+00 3.22910428e-01 -3.50123376e-01 6.70941770e-01
9.60250735e-01 -1.40520573e+00 -7.52387524e-01 3.26497674e-01
2.89334297e-01 2.25050561e-02 7.55225658e-01 1.23558664e+00
-1.55995548e-01 6.84459358e-02 -1.77529469e-01 -1.04339433e+00
-8.95662010e-01 2.84816891e-01 5.39880514e-01 -4.43666458e-01
-7.74476051e-01 5.57305872e-01 1.16332352e-01 4.73736525e-02
1.51418805e-01 -2.68284261e-01 -6.77460313e-01 1.54297635e-01
9.44488049e-01 7.13281572e-01 2.40998149e-01 -8.71885002e-01
-6.53062344e-01 3.18094820e-01 -2.65769124e-01 1.67914644e-01
1.35427487e+00 4.67551686e-03 2.37086594e-01 4.13317978e-01
1.22502851e+00 -4.97580975e-01 -1.54747725e+00 -2.15865225e-01
1.89765960e-01 -4.60079163e-01 1.43487647e-01 -5.90788662e-01
-1.28751278e+00 8.77997220e-01 1.06606936e+00 -2.58892328e-02
1.40396214e+00 -2.37055540e-01 8.49529147e-01 3.35674286e-01
-7.11815134e-02 -1.06791472e+00 4.43281174e-01 5.45567751e-01
5.17091990e-01 -1.33156228e+00 -2.11813495e-01 2.16781825e-01
-8.55163991e-01 1.01833212e+00 9.29445803e-01 -3.29298377e-01
7.50398457e-01 -8.06861743e-02 6.32133335e-02 -2.09815800e-01
-8.95203054e-01 1.73136452e-03 4.69071537e-01 3.65514249e-01
7.87990242e-02 -1.74632087e-01 5.73910736e-02 6.24329090e-01
4.97813553e-01 -2.33963449e-02 2.03091428e-01 1.00575757e+00
-6.02259338e-01 -3.77226800e-01 -1.06021702e-01 5.50198078e-01
-6.04913652e-01 2.90243328e-01 1.22305699e-01 4.85641778e-01
4.66984123e-01 1.14829218e+00 4.68463928e-01 -6.40574157e-01
4.03691083e-01 -5.99197985e-04 -1.01682074e-01 -3.18660825e-01
-3.86368006e-01 4.19188924e-02 9.64562129e-03 -9.53979492e-01
-4.88602132e-01 -6.76257610e-01 -1.56580544e+00 1.85046017e-01
-1.16073959e-01 -5.74546568e-02 2.63714284e-01 1.42752397e+00
4.92156178e-01 9.80304122e-01 8.67493987e-01 -7.23112404e-01
-1.58415899e-01 -7.84967661e-01 -2.41666511e-01 6.41278088e-01
7.47293651e-01 -8.19731534e-01 -4.15045679e-01 -9.83080864e-02] | [7.8430094718933105, 1.6003310680389404] |
0a56cf15-7403-4da6-b22f-9a44b5cb2624 | deepsdf-x-sim-3-extending-deepsdf-for | 2004.09048 | null | https://arxiv.org/abs/2004.09048v3 | https://arxiv.org/pdf/2004.09048v3.pdf | Extending DeepSDF for automatic 3D shape retrieval and similarity transform estimation | Recent advances in computer graphics and computer vision have found successful application of deep neural network models for 3D shapes based on signed distance functions (SDFs) that are useful for shape representation, retrieval, and completion. However, this approach has been limited by the need to have query shapes in the same canonical scale and pose as those observed during training, restricting its effectiveness on real world scenes. We present a formulation to overcome this issue by jointly estimating shape and similarity transform parameters. We conduct experiments to demonstrate the effectiveness of this formulation on synthetic and real datasets and report favorable comparisons to the state of the art. Finally, we also emphasize the viability of this approach as a form of 3D model compression. | ['S. Shankar Sastry', 'Allen Y. Yang', 'Oladapo Afolabi'] | 2020-04-20 | null | null | null | null | ['3d-shape-retrieval'] | ['computer-vision'] | [ 5.43851443e-02 -1.93478942e-01 1.68716595e-01 -3.63813937e-01
-5.86801946e-01 -6.10338926e-01 9.31612372e-01 8.43316466e-02
-3.61897916e-01 2.71550179e-01 3.00050899e-02 -2.81874359e-01
-1.13364682e-01 -8.75071406e-01 -6.77754045e-01 -3.02350193e-01
-7.22318441e-02 7.63299048e-01 6.56730607e-02 -1.09180868e-01
4.54620361e-01 1.39831877e+00 -1.36378527e+00 1.14214063e-01
3.00173014e-01 1.13883185e+00 -4.17308472e-02 2.21926585e-01
-2.39502922e-01 -1.93787217e-01 -4.87865806e-01 -3.86431009e-01
5.65238833e-01 3.08577716e-01 -3.47375095e-01 2.11124003e-01
8.43585968e-01 -7.09305286e-01 -4.87455010e-01 7.36768365e-01
7.55807698e-01 2.04861552e-01 9.97577846e-01 -9.99444067e-01
-6.85508311e-01 -4.48933542e-01 -6.52145088e-01 -1.38086304e-01
2.41637528e-01 -1.40754774e-01 8.46733987e-01 -1.44566071e+00
8.83490443e-01 1.35812283e+00 8.24644804e-01 3.43035698e-01
-1.20799565e+00 -4.29779351e-01 -2.47837529e-01 -1.69845313e-01
-1.42967820e+00 -6.20713711e-01 9.41908181e-01 -2.82142520e-01
1.34400046e+00 1.66489020e-01 7.66201794e-01 6.03812873e-01
-7.92896152e-02 8.05066705e-01 3.83766055e-01 -4.24054056e-01
2.92241454e-01 -2.90147871e-01 -3.14022809e-01 6.44177616e-01
3.74314070e-01 3.77695978e-01 -2.39205062e-01 -3.60907257e-01
1.44286013e+00 -2.89700329e-01 -7.90587366e-02 -9.78865802e-01
-9.37381625e-01 9.05810118e-01 5.87058008e-01 3.37091833e-01
-4.52874303e-01 2.74383008e-01 3.50719362e-01 1.61566641e-02
7.62055874e-01 2.98339099e-01 -3.06804806e-01 1.03079043e-01
-1.04323387e+00 5.70906103e-01 7.92334378e-01 1.02711308e+00
3.34464639e-01 4.16631281e-01 5.39332442e-02 9.08115685e-01
3.34576070e-01 6.06126308e-01 1.27799347e-01 -1.17540979e+00
1.57973394e-01 3.91778350e-01 3.63029204e-02 -1.42727256e+00
-3.48482430e-01 -3.86681527e-01 -7.17798054e-01 4.27420020e-01
2.35322401e-01 4.03864443e-01 -1.02693450e+00 1.43265820e+00
3.79565448e-01 2.25919887e-01 -2.53052235e-01 8.35404873e-01
9.20045674e-01 4.02321339e-01 -2.76908129e-01 2.76762456e-01
7.51212060e-01 -4.57007974e-01 -2.94588894e-01 1.22814909e-01
4.26269382e-01 -9.65598464e-01 7.82592475e-01 1.93691313e-01
-1.36691773e+00 -5.40125012e-01 -1.04240787e+00 -2.36044049e-01
-2.59032130e-01 7.46504664e-02 6.74186528e-01 6.10683262e-01
-1.31998563e+00 7.76470065e-01 -8.32163572e-01 -3.10453147e-01
7.60066450e-01 6.04380250e-01 -3.58763933e-01 3.24330432e-03
-5.42638063e-01 7.47117519e-01 9.88280401e-02 7.04720020e-02
-4.72620547e-01 -5.77113807e-01 -9.24397111e-01 2.18284830e-01
-1.06283188e-01 -8.31590772e-01 1.18338645e+00 -4.63659793e-01
-1.20663214e+00 1.09686863e+00 -2.46350527e-01 -4.13863420e-01
4.74141598e-01 -2.06425294e-01 9.03611258e-02 1.69339120e-01
-2.84233540e-01 9.29576039e-01 7.76307642e-01 -1.36472225e+00
-1.02093197e-01 -4.41229641e-01 -8.07291195e-02 2.00905427e-01
-2.03788862e-01 -1.95182905e-01 -6.50691390e-01 -6.51549578e-01
6.55072391e-01 -8.61773312e-01 -2.06997216e-01 7.43239343e-01
8.65479186e-02 -2.34124586e-01 9.96215701e-01 -4.47405756e-01
6.61853790e-01 -2.18653345e+00 -1.38731360e-01 5.45317829e-01
1.54411614e-01 5.04844248e-01 -4.52321619e-01 4.44730431e-01
3.98983918e-02 9.63705033e-02 -1.75936222e-01 -7.45127320e-01
3.05544809e-02 2.47066423e-01 -6.79390967e-01 5.75106323e-01
1.86677560e-01 1.09838712e+00 -4.81639773e-01 -2.13524476e-01
3.98043782e-01 9.13093746e-01 -5.97657442e-01 4.82009910e-02
-1.40476555e-01 2.70221859e-01 -3.14180970e-01 6.68151796e-01
9.08657670e-01 -1.61431521e-01 -4.56281975e-02 -4.03972536e-01
3.29235718e-02 1.66617557e-01 -1.07498324e+00 1.70569825e+00
-2.66319573e-01 6.62004292e-01 7.07738772e-02 -8.97506654e-01
1.20268607e+00 2.27967992e-01 7.98623025e-01 -5.77271700e-01
2.01889411e-01 3.88010830e-01 -2.13802338e-01 -1.08652301e-01
5.47776282e-01 -1.29782230e-01 5.45489788e-01 6.27954245e-01
-2.96802640e-01 -9.17572677e-01 -1.25430450e-01 -6.13775179e-02
4.75920290e-01 3.58976930e-01 7.37275705e-02 -9.37461331e-02
3.41226935e-01 -2.35480562e-01 1.06182247e-01 5.42098522e-01
-8.44735801e-02 9.70682800e-01 -2.32997220e-02 -6.69120133e-01
-1.44772625e+00 -1.13549328e+00 -2.55980253e-01 5.79783082e-01
4.95655052e-02 -9.79811773e-02 -4.44902182e-01 -2.61055857e-01
2.86480635e-01 5.27662694e-01 -4.01923209e-01 9.77634192e-02
-8.96148622e-01 -2.79276103e-01 5.63973963e-01 8.28728795e-01
2.59455413e-01 -1.02338314e+00 -6.59353197e-01 1.25844836e-01
2.57682204e-01 -1.15440941e+00 -5.91949880e-01 -3.26855779e-01
-1.24638879e+00 -9.10632432e-01 -8.69519174e-01 -9.12144959e-01
7.33121455e-01 4.99034137e-01 1.22948718e+00 3.31135839e-01
-3.65584105e-01 6.61534369e-01 1.77473426e-02 -4.31742400e-01
-2.75139004e-01 -1.08761378e-01 7.00189173e-02 -3.00735116e-01
1.31651938e-01 -7.61120081e-01 -6.31110430e-01 1.22614600e-01
-1.12531269e+00 -1.00807987e-01 3.13544452e-01 7.76275873e-01
6.50311947e-01 -3.88449103e-01 4.68498111e-01 -4.44295704e-01
7.66159177e-01 8.94361660e-02 -7.30504453e-01 5.84666729e-02
-5.67407548e-01 -5.11421300e-02 1.95417643e-01 -3.30571294e-01
-7.32160330e-01 1.80833161e-01 -3.12895626e-01 -8.02694023e-01
-1.50517806e-01 4.13502514e-01 1.24510691e-01 -4.97623473e-01
4.39970016e-01 3.88231218e-01 2.89643347e-01 -5.86702585e-01
3.48997265e-01 3.06416482e-01 4.68964607e-01 -6.94656909e-01
7.62867391e-01 9.30519283e-01 4.16338801e-01 -1.11891258e+00
-2.83105850e-01 -3.01612139e-01 -7.70303488e-01 6.24560751e-02
3.97073656e-01 -6.41469419e-01 -6.82862163e-01 3.95551383e-01
-1.49348056e+00 1.91309657e-02 -3.70892048e-01 3.36939931e-01
-9.33108389e-01 6.18877888e-01 -5.81309855e-01 -7.58743823e-01
-5.39983094e-01 -9.17300403e-01 1.38910794e+00 -8.29270259e-02
-1.57818899e-01 -1.09727299e+00 -2.57361867e-03 -1.23661667e-01
6.39737248e-01 2.56397396e-01 1.12187612e+00 -5.09461641e-01
-7.19614148e-01 -5.17353415e-01 -5.36725640e-01 2.98061371e-02
3.49349193e-02 4.06159721e-02 -9.29198802e-01 -5.86335123e-01
6.78843632e-02 -2.92237014e-01 6.25713170e-01 6.76931977e-01
1.12300324e+00 2.22945705e-01 -3.24330807e-01 8.36096823e-01
1.38783062e+00 1.81592047e-01 4.72370565e-01 -2.48621274e-02
4.31626081e-01 5.36628962e-01 2.05639377e-01 4.50725704e-01
1.37845635e-01 7.67653048e-01 4.46512252e-01 -1.84487775e-01
-3.26860726e-01 -2.73151696e-01 -3.02944213e-01 7.52283514e-01
-1.57281384e-01 -1.26116812e-01 -8.24284971e-01 3.99520665e-01
-1.52485561e+00 -6.77436888e-01 1.45753339e-01 2.36639571e+00
2.44861796e-01 1.44733399e-01 -1.65584281e-01 -2.69531384e-02
4.82962638e-01 1.23642594e-01 -6.02379620e-01 -3.53819817e-01
-3.60331982e-01 4.66192096e-01 2.25766659e-01 4.52506751e-01
-1.01932907e+00 8.12916934e-01 7.37797308e+00 7.55095124e-01
-1.36042368e+00 -3.10226798e-01 5.80801249e-01 1.54534742e-01
-4.97554481e-01 -2.61366785e-01 -4.09576654e-01 -2.21377134e-01
3.07570547e-01 -2.08260223e-01 3.40672880e-01 6.55560255e-01
1.41878635e-01 1.00616649e-01 -1.32337606e+00 1.28067172e+00
1.72065169e-01 -1.50940585e+00 3.15489113e-01 2.84048140e-01
6.34135962e-01 1.14584215e-01 2.95112461e-01 -1.53760929e-02
-1.55990571e-01 -1.11186945e+00 6.59111202e-01 3.60022902e-01
1.12449336e+00 -5.32509208e-01 3.09533864e-01 2.66041547e-01
-1.24437261e+00 3.01093131e-01 -6.39563918e-01 1.07772149e-01
1.98862597e-01 1.72417119e-01 -1.19610703e+00 4.66214657e-01
2.20665038e-01 6.31745160e-01 -3.67544144e-01 1.31525409e+00
3.76630694e-01 8.05857256e-02 -7.23849714e-01 7.66114972e-04
2.53748566e-01 -2.99200058e-01 6.47011101e-01 9.48644221e-01
6.72638655e-01 8.78162235e-02 8.58884379e-02 1.00059736e+00
-3.15280482e-02 1.33067608e-01 -1.13790834e+00 2.51833256e-02
4.74127799e-01 7.58556306e-01 -9.51903403e-01 -2.25571856e-01
-3.10747802e-01 7.39961445e-01 2.56523430e-01 3.91805112e-01
-3.44933748e-01 4.95536402e-02 4.85997289e-01 2.96287835e-01
5.79127431e-01 -8.24276209e-01 -6.86486840e-01 -8.40979457e-01
8.72875080e-02 -4.78110343e-01 3.12409773e-02 -7.68314958e-01
-1.38013816e+00 6.41851306e-01 1.30047083e-01 -1.45849991e+00
-3.45629752e-01 -7.43343949e-01 -4.64688033e-01 8.32283795e-01
-1.26513207e+00 -1.31813216e+00 1.95923410e-02 3.39010835e-01
5.38008630e-01 -1.83295503e-01 1.01171350e+00 3.77399415e-01
2.34081537e-01 5.77293634e-01 2.24599347e-01 -8.80059320e-03
3.25779021e-01 -8.60063374e-01 1.04926836e+00 3.40876043e-01
4.64129776e-01 5.86526036e-01 5.00383198e-01 -6.10927165e-01
-1.49279130e+00 -7.11934209e-01 5.85622370e-01 -2.44064659e-01
9.99633968e-02 -1.65857688e-01 -9.22849417e-01 4.88159657e-01
-1.49785653e-01 4.39205281e-02 3.09738666e-01 -1.03295669e-01
-2.95195222e-01 1.82937846e-01 -1.34885168e+00 5.86640716e-01
1.13853788e+00 -4.77559626e-01 -3.50064188e-01 1.04737885e-01
5.48546731e-01 -7.05263495e-01 -5.29839575e-01 6.79616809e-01
8.73936951e-01 -1.04344416e+00 1.34533966e+00 -6.07997477e-01
1.81786388e-01 -2.14693900e-02 -4.06087130e-01 -1.17399442e+00
-2.71111995e-01 -1.96031153e-01 -7.84353390e-02 5.17104983e-01
8.02045390e-02 -4.50655043e-01 1.08937871e+00 8.61253560e-01
-1.43563479e-01 -1.01189256e+00 -1.14282620e+00 -8.25709879e-01
3.61355096e-01 -6.30189300e-01 7.28977263e-01 7.68589079e-01
-6.37087882e-01 3.48340534e-03 -1.35908499e-01 -1.09062232e-01
6.12551153e-01 4.14786041e-01 7.98632324e-01 -1.35467482e+00
3.97036336e-02 -7.08503604e-01 -5.49755931e-01 -1.33702970e+00
1.76032722e-01 -1.04389894e+00 -1.87252909e-01 -1.52489281e+00
-1.94757044e-01 -6.67402685e-01 1.90118775e-01 1.45687535e-01
5.11430442e-01 5.81340194e-01 3.64003509e-01 2.09063157e-01
1.14032611e-01 7.60850370e-01 1.47158253e+00 -1.03165850e-01
-1.88177750e-01 -4.51642424e-02 -2.40728304e-01 7.71266520e-01
4.74110425e-01 -8.68749842e-02 -3.88328016e-01 -8.88446629e-01
8.35262425e-03 8.39103162e-02 3.21189493e-01 -7.46025324e-01
2.65064359e-01 8.85515884e-02 6.62661254e-01 -1.03508806e+00
8.84392500e-01 -1.09540009e+00 6.45232573e-02 4.57085103e-01
-2.17681170e-01 3.95373225e-01 4.83029038e-01 4.69599247e-01
-4.84841727e-02 -2.35045359e-01 6.40828371e-01 2.17241459e-02
-2.76415348e-01 6.28871977e-01 1.89871788e-01 -2.35397279e-01
7.38271236e-01 -5.84436178e-01 1.71175659e-01 -6.98440433e-01
-7.22360671e-01 -2.84731805e-01 5.36536276e-01 3.21714222e-01
1.19914818e+00 -1.68817282e+00 -8.04479480e-01 7.43126512e-01
-2.04927996e-01 1.14241160e-01 -2.00556498e-02 3.25384200e-01
-7.94253588e-01 6.57856703e-01 -1.76257238e-01 -8.31009150e-01
-1.23798478e+00 4.49801087e-01 3.05454105e-01 -3.77872549e-02
-6.85026169e-01 6.11892283e-01 3.63077462e-01 -5.40197730e-01
3.81021738e-01 -2.48329222e-01 8.22704658e-02 -3.50794464e-01
2.43370593e-01 2.28756279e-01 4.14338112e-01 -8.68874729e-01
-2.10337296e-01 8.16075623e-01 1.40592366e-01 -2.76081711e-01
1.52461946e+00 1.39962509e-01 1.74054280e-02 4.46980782e-02
1.35146534e+00 -7.36863464e-02 -1.21501064e+00 -3.23429018e-01
-1.39219984e-02 -8.24504614e-01 -1.20739847e-01 -3.40210915e-01
-1.20682061e+00 9.97476161e-01 6.43342674e-01 1.51031360e-01
8.24866116e-01 -1.56287819e-01 1.03449094e+00 6.21433914e-01
4.32443649e-01 -6.77847266e-01 4.92651574e-03 7.65048444e-01
1.35683441e+00 -9.56249893e-01 1.49575338e-01 -4.84407455e-01
-5.52516095e-02 1.42236137e+00 2.62357086e-01 -4.81669426e-01
1.05510759e+00 2.87950754e-01 1.06534166e-02 -3.81625056e-01
-4.82611299e-01 -1.31157376e-02 6.37513399e-01 6.97238147e-01
5.20229518e-01 -9.57387835e-02 -2.13178024e-01 -9.70955566e-02
-1.52699679e-01 3.54328491e-02 1.65243104e-01 1.06176853e+00
-2.82182902e-01 -1.13711011e+00 -4.37078208e-01 4.30409640e-01
-2.11642429e-01 -3.37185934e-02 -2.80304104e-01 6.88446403e-01
-4.17248785e-01 4.87779975e-01 3.20642233e-01 -9.63340625e-02
5.11446953e-01 -6.78064600e-02 9.64813113e-01 -4.67723221e-01
-2.91623384e-01 4.41084802e-01 -2.19779164e-01 -3.92449915e-01
-3.05980265e-01 -5.01949787e-01 -1.10703921e+00 -2.59553939e-01
-2.54672647e-01 -3.65693480e-01 8.17655861e-01 6.97058856e-01
5.81935167e-01 -1.24358930e-01 6.34073973e-01 -1.41614771e+00
-8.22398245e-01 -7.68367767e-01 -4.70283657e-01 6.16598904e-01
1.91702858e-01 -7.65252054e-01 -7.54082110e-03 -9.99576971e-03] | [8.358115196228027, -3.6227128505706787] |
a07628d8-71c9-4ab5-992d-8387f1dc935d | plpca-persistent-laplacian-enhanced-pca-for | 2306.06292 | null | https://arxiv.org/abs/2306.06292v1 | https://arxiv.org/pdf/2306.06292v1.pdf | PLPCA: Persistent Laplacian Enhanced-PCA for Microarray Data Analysis | Over the years, Principal Component Analysis (PCA) has served as the baseline approach for dimensionality reduction in gene expression data analysis. It primary objective is to identify a subset of disease-causing genes from a vast pool of thousands of genes. However, PCA possesses inherent limitations that hinder its interpretability, introduce classification ambiguity, and fail to capture complex geometric structures in the data. Although these limitations have been partially addressed in the literature by incorporating various regularizers such as graph Laplacian regularization, existing improved PCA methods still face challenges related to multiscale analysis and capturing higher-order interactions in the data. To address these challenges, we propose a novel approach called Persistent Laplacian-enhanced Principal Component Analysis (PLPCA). PLPCA amalgamates the advantages of earlier regularized PCA methods with persistent spectral graph theory, specifically persistent Laplacians derived from algebraic topology. In contrast to graph Laplacians, persistent Laplacians enable multiscale analysis through filtration and incorporate higher-order simplicial complexes to capture higher-order interactions in the data. We evaluate and validate the performance of PLPCA using benchmark microarray datasets that involve normal tissue samples and four different cancer tissues. Our extensive studies demonstrate that PLPCA outperforms all other state-of-the-art models for classification tasks after dimensionality reduction. | ['GuoWei Wei', 'Rui Wang', 'Sean Cottrell'] | 2023-06-09 | null | null | null | null | ['dimensionality-reduction'] | ['methodology'] | [ 2.17999578e-01 -2.93078959e-01 -6.73815608e-02 1.45136997e-01
-6.82911515e-01 -6.24024451e-01 4.77434933e-01 2.07616255e-01
7.96206445e-02 4.74473685e-01 2.38161355e-01 -6.72420114e-02
-5.85791469e-01 -5.59845567e-01 -2.47497678e-01 -1.18455184e+00
-3.76756161e-01 3.64937365e-01 -2.44028226e-01 -1.05158105e-01
-1.37515273e-02 7.81365752e-01 -1.19192648e+00 1.88214809e-01
8.94448638e-01 5.54703236e-01 -2.31831282e-01 4.88962382e-01
-2.62585521e-01 -2.24466324e-02 7.86173120e-02 -6.46118447e-02
1.22402519e-01 -2.53654420e-01 -3.89722735e-01 3.34856182e-01
2.19525605e-01 4.67872024e-01 -4.40920621e-01 1.30908275e+00
4.23443019e-01 -2.77908564e-01 7.93016195e-01 -1.36376297e+00
-7.07349896e-01 1.08390480e-01 -1.00881624e+00 1.12824574e-01
1.26115352e-01 -5.60475886e-02 1.11710930e+00 -1.26490009e+00
6.29376113e-01 1.52273965e+00 7.53948987e-01 4.79325652e-01
-1.76912820e+00 -2.35499755e-01 -4.07380499e-02 4.39889403e-03
-1.64182508e+00 -3.41125876e-01 1.23870146e+00 -6.40064836e-01
5.69705069e-01 5.06721199e-01 6.44146621e-01 9.09353614e-01
4.22870815e-01 6.19275749e-01 1.20747232e+00 -2.49788493e-01
2.21956074e-01 -4.78323191e-01 5.32387078e-01 9.85726953e-01
5.38822711e-01 -3.45189393e-01 -2.81429410e-01 -8.07844877e-01
6.36806726e-01 2.19989389e-01 -4.14107054e-01 -8.16945374e-01
-1.35971344e+00 8.01938593e-01 -1.11009879e-02 3.94760072e-01
-6.26825988e-01 -2.67043244e-02 3.00012916e-01 -1.26899451e-01
4.93941516e-01 3.75877559e-01 -2.23222524e-01 -1.40339986e-03
-4.93855238e-01 3.76068726e-02 6.06202006e-01 5.90346813e-01
8.90177190e-01 -7.92247579e-02 -6.77244514e-02 8.45661223e-01
4.29185241e-01 4.11087871e-01 4.05829906e-01 -8.10334921e-01
1.97619557e-01 1.16120112e+00 -5.03341198e-01 -1.55945933e+00
-7.73773789e-01 -5.73898911e-01 -1.50551283e+00 -1.97625220e-01
4.29014534e-01 1.07229836e-01 -7.66503572e-01 1.53423882e+00
3.26152951e-01 3.31943005e-01 -1.43211726e-02 7.73087084e-01
4.55139101e-01 2.58030772e-01 1.00711174e-01 -3.91034573e-01
1.50330448e+00 -4.98608947e-01 -8.94679129e-01 1.67720094e-01
6.09362721e-01 -5.45840204e-01 1.08729017e+00 3.86296004e-01
-6.84029043e-01 -1.36259288e-01 -8.95772696e-01 -2.91796657e-03
-3.05650622e-01 3.31755966e-01 1.01691687e+00 6.73621178e-01
-1.02473700e+00 3.38336647e-01 -1.20817101e+00 -4.76497054e-01
6.06693327e-01 3.35149556e-01 -6.56269431e-01 -1.81436673e-01
-8.18521142e-01 2.31033474e-01 -2.48417646e-01 2.64015704e-01
-4.41243261e-01 -9.59431112e-01 -7.57193744e-01 4.24494967e-02
1.30418807e-01 -7.99447298e-01 3.32084715e-01 -4.30599779e-01
-1.29853535e+00 6.89998448e-01 -4.79007632e-01 1.32846013e-01
2.38666371e-01 1.68127581e-01 -2.80247211e-01 2.10660055e-01
-2.08957344e-02 2.74364620e-01 5.09405971e-01 -1.04598510e+00
-1.86910272e-01 -7.35134482e-01 -3.99318993e-01 6.59853742e-02
-6.89943552e-01 -3.98094565e-01 -6.44380927e-01 -6.45300627e-01
7.41167843e-01 -1.30379772e+00 -4.97394204e-01 5.67576550e-02
-4.23865795e-01 -1.58297136e-01 1.08848715e+00 -3.40837985e-01
1.23053133e+00 -2.33009410e+00 6.27437413e-01 3.88589591e-01
6.60805941e-01 2.05684323e-02 -3.07374179e-01 4.15154040e-01
-3.30120355e-01 3.42052639e-01 -4.76601124e-01 -2.24696666e-01
-1.81345731e-01 2.18967110e-01 2.89695822e-02 8.42924953e-01
3.89684528e-01 8.55954587e-01 -9.96996582e-01 -4.05436695e-01
3.58020552e-02 8.55781138e-01 -5.04713714e-01 -4.16419268e-01
7.85930306e-02 4.40604448e-01 -6.57008231e-01 8.84953678e-01
8.56418729e-01 -5.04600823e-01 2.67288387e-01 -6.28771961e-01
1.54179439e-01 -3.84797543e-01 -1.14113796e+00 1.76899827e+00
8.52080658e-02 4.52423155e-01 4.29978251e-01 -1.05320036e+00
7.24002600e-01 1.96905047e-01 1.10104728e+00 2.72987522e-02
1.65740978e-02 9.76233631e-02 1.04736038e-01 -4.85639244e-01
-9.82972383e-02 -1.02120675e-02 -1.89438406e-02 6.87859720e-03
-2.62854248e-01 2.07295716e-01 3.41396630e-01 2.31467962e-01
1.49376130e+00 -2.50311971e-01 2.36638427e-01 -7.00319469e-01
8.59538257e-01 1.16203323e-01 8.22518349e-01 2.18929034e-02
-3.18053931e-01 6.01386011e-01 1.01814497e+00 -2.67257214e-01
-6.51090980e-01 -9.62523222e-01 -3.11971962e-01 5.78475833e-01
-2.26964429e-01 -6.46981657e-01 -5.95336139e-01 -6.68727040e-01
1.00268021e-01 1.79315489e-02 -6.95705712e-01 -4.32622209e-02
-3.32212061e-01 -1.67801952e+00 6.45965397e-01 2.33294412e-01
3.35242748e-01 -1.49553433e-01 1.87896371e-01 2.26640506e-04
-1.74031958e-01 -1.10975921e+00 -5.12568355e-01 -1.20340295e-01
-1.09309196e+00 -1.53503704e+00 -4.76265311e-01 -7.38768280e-01
1.23764396e+00 4.73334193e-01 6.09721363e-01 -1.84480086e-01
-7.41838515e-01 5.77504992e-01 -5.54670505e-02 -2.20760763e-01
-2.26859868e-01 -1.45177156e-01 3.23477715e-01 3.67788821e-01
5.24153054e-01 -7.02289641e-01 -4.51070160e-01 1.54872864e-01
-1.10002494e+00 -2.40963623e-02 8.95810544e-01 1.02040493e+00
1.16637480e+00 3.83231848e-01 4.01355088e-01 -9.68754947e-01
9.19769168e-01 -3.83763343e-01 -4.50676143e-01 1.83877409e-01
-6.27912343e-01 8.82330984e-02 4.51199621e-01 -3.45786452e-01
-6.92496955e-01 5.36342502e-01 1.24628738e-01 -5.73881924e-01
1.80136214e-03 9.25867617e-01 -3.90514255e-01 -3.30382854e-01
5.09073734e-01 3.00711751e-01 4.58636969e-01 -5.32993674e-01
3.03585708e-01 2.71871656e-01 1.61030635e-01 -4.15927559e-01
6.51625633e-01 8.91756773e-01 9.41037238e-01 -1.22758651e+00
-3.04574162e-01 -5.72783828e-01 -9.51179981e-01 1.44122079e-01
6.95493400e-01 -8.07677984e-01 -6.98076308e-01 4.81139362e-01
-8.01675737e-01 3.13964665e-01 7.55112395e-02 4.22847629e-01
-2.59169221e-01 8.95381272e-01 -7.55717337e-01 -5.15180409e-01
-2.97452748e-01 -1.13573956e+00 1.05269003e+00 -2.60463089e-01
-8.14871863e-02 -1.22863269e+00 2.76038945e-01 1.17218278e-01
3.62966001e-01 6.73762739e-01 1.50508773e+00 -3.16733956e-01
-2.25500807e-01 -4.68872190e-01 -1.27717838e-01 1.07724994e-01
5.99972904e-01 2.07152426e-01 -6.74468040e-01 -3.01700830e-01
-6.77881837e-02 3.67735416e-01 7.92094648e-01 3.69609118e-01
1.21609390e+00 -1.90232739e-01 -8.55956078e-01 6.26533508e-01
1.35269606e+00 -2.89193511e-01 4.52800155e-01 -1.33010909e-01
1.06903660e+00 6.12774670e-01 1.96266443e-01 1.67444348e-01
1.81739807e-01 4.93430853e-01 2.30445802e-01 -1.69273734e-01
-1.05568312e-01 5.60650826e-02 1.55227199e-01 1.18530297e+00
-7.08677992e-02 2.42747471e-01 -1.19150114e+00 4.09122765e-01
-1.89073277e+00 -5.51504612e-01 -7.77989447e-01 1.82345891e+00
6.07211769e-01 -2.17756972e-01 9.36712325e-03 2.86311746e-01
6.36710763e-01 -2.63290629e-02 -4.72231984e-01 2.54760176e-01
-5.25012910e-01 -2.90122867e-01 3.55930805e-01 4.28368777e-01
-1.19780934e+00 6.32623851e-01 6.62647581e+00 7.31759846e-01
-1.05422115e+00 -6.79657981e-02 5.73108912e-01 1.99172884e-01
-2.78641731e-01 -7.51781315e-02 -6.43837452e-01 1.64246574e-01
5.58841765e-01 -2.51540780e-01 2.67390490e-01 8.02758932e-01
4.00436968e-01 2.67469853e-01 -1.05726910e+00 1.03371823e+00
-1.83669496e-02 -1.18598664e+00 2.15527624e-01 5.27627051e-01
8.06743681e-01 -6.37972951e-02 2.52474755e-01 -5.38775697e-02
-6.51106536e-02 -9.13763404e-01 -3.33445728e-01 6.81266904e-01
6.23688638e-01 -5.89751363e-01 5.65868378e-01 1.60799846e-01
-1.36623120e+00 9.88426059e-03 -4.72270966e-01 2.60109995e-02
-9.90938023e-02 1.18455410e+00 -7.51242757e-01 6.28131390e-01
3.02892655e-01 9.31295872e-01 -6.44427955e-01 8.75737965e-01
8.89602825e-02 5.94748735e-01 -2.01436356e-01 1.83275461e-01
-7.47951958e-03 -6.17943525e-01 8.41709554e-01 1.25799870e+00
3.28823209e-01 2.25715473e-01 1.52443945e-01 7.62182176e-01
-1.27416002e-02 3.32366914e-01 -5.48953474e-01 -4.02380645e-01
1.87107056e-01 1.72269142e+00 -8.86048079e-01 5.89946285e-02
-6.75342381e-01 7.48180091e-01 2.36596748e-01 5.03190041e-01
-3.90222341e-01 -1.37930155e-01 1.04153156e+00 1.91473067e-01
-1.33110628e-01 -6.63255751e-01 -2.91884154e-01 -1.25514638e+00
-6.59805313e-02 -9.58492577e-01 4.80253667e-01 -1.92571625e-01
-1.72374427e+00 2.69578725e-01 -2.41212636e-01 -1.13429070e+00
2.74144858e-01 -9.58187699e-01 -4.86269504e-01 7.17085421e-01
-1.41632640e+00 -1.14617527e+00 -2.86687106e-01 4.59913611e-01
2.33806401e-01 -1.01887390e-01 1.22115338e+00 3.97527605e-01
-9.54091609e-01 7.03902021e-02 3.77174973e-01 1.14956968e-01
4.87975895e-01 -1.25830829e+00 1.46972872e-02 6.97382390e-01
-1.19995058e-01 9.18358088e-01 4.35960084e-01 -7.27486253e-01
-2.30820036e+00 -1.21975148e+00 4.79118109e-01 -4.88275826e-01
9.73942101e-01 -4.74905103e-01 -1.02310789e+00 4.83691841e-01
-2.89372116e-01 3.87534350e-01 1.08272326e+00 5.76288141e-02
-2.31135771e-01 -1.62787303e-01 -8.24280262e-01 6.21818185e-01
9.90234971e-01 -4.32585597e-01 -5.57139292e-02 4.12689030e-01
2.17857212e-01 2.34642439e-02 -1.12824774e+00 5.37578404e-01
3.59908164e-01 -4.55978751e-01 1.03710687e+00 -6.54098332e-01
-3.62472832e-02 -5.35275519e-01 -8.92377645e-02 -1.35170317e+00
-8.03981602e-01 -6.96996629e-01 -2.18296051e-01 1.13142622e+00
2.38859057e-01 -6.47157252e-01 9.40223575e-01 5.67867100e-01
-9.88331363e-02 -8.36451292e-01 -9.95628059e-01 -6.59748793e-01
5.05539775e-02 -1.90561384e-01 3.11896175e-01 1.28922474e+00
4.05509621e-01 3.60580921e-01 -6.65928200e-02 2.89476097e-01
8.91957402e-01 -9.99261290e-02 5.58021188e-01 -1.56884301e+00
-3.70125927e-04 -7.38776147e-01 -6.88750327e-01 -5.91004848e-01
2.02576801e-01 -1.17224419e+00 -2.64677703e-01 -1.50904655e+00
3.89170021e-01 -2.82597035e-01 -3.33164752e-01 5.15431583e-01
-3.05523694e-01 1.84970617e-01 -2.82815367e-01 3.83336544e-01
-4.03837800e-01 4.82430667e-01 1.28325784e+00 -3.05803806e-01
-3.82410884e-01 -3.11431497e-01 -8.44253182e-01 8.04493368e-01
6.25769973e-01 -2.11523622e-01 -2.95757055e-01 -1.95294723e-01
2.15019777e-01 -2.18995169e-01 2.80291826e-01 -8.02181125e-01
2.79096574e-01 -1.36929542e-01 4.10458863e-01 -2.78627515e-01
3.05418253e-01 -7.88645208e-01 4.94000375e-01 4.89504606e-01
3.41061130e-02 9.42018628e-02 3.33865821e-01 1.00528312e+00
-2.42680743e-01 4.37460691e-01 5.23449481e-01 1.23508438e-01
-4.59637046e-01 4.45715278e-01 -5.85431993e-01 -3.71463060e-01
9.79022264e-01 7.42338151e-02 -3.95731300e-01 9.50775743e-02
-9.47631299e-01 1.35956645e-01 4.61354017e-01 2.77130395e-01
7.15415955e-01 -1.49583340e+00 -8.17288578e-01 3.32009345e-01
1.78104565e-01 -1.24402002e-01 3.20242643e-01 1.45638287e+00
-4.35800970e-01 6.30587041e-01 -1.05524592e-01 -8.05977941e-01
-1.44825649e+00 6.48568869e-01 1.00485995e-01 -3.41399163e-01
-4.53245133e-01 5.16673863e-01 4.26407099e-01 -2.08387613e-01
-9.08332616e-02 -1.88770264e-01 -2.65929878e-01 6.47694394e-02
2.21668676e-01 6.61645949e-01 4.50019119e-03 -7.99008191e-01
-5.23755550e-01 9.06296134e-01 2.07197681e-01 2.63717651e-01
1.36393631e+00 -2.36846104e-01 -7.22908914e-01 3.53731841e-01
1.40601313e+00 1.93023294e-01 -9.49248612e-01 -7.65795186e-02
2.16228828e-01 -3.14457089e-01 2.30314985e-01 -2.99445927e-01
-1.02033639e+00 9.87028718e-01 4.32183772e-01 2.53794611e-01
1.19204664e+00 -1.49431512e-01 2.87427217e-01 3.97271901e-01
1.12743035e-01 -7.00495064e-01 1.41448125e-01 3.36403131e-01
9.28631008e-01 -8.76501679e-01 8.30216631e-02 -1.07113373e+00
-3.83078873e-01 1.16707981e+00 3.76948029e-01 -1.37795329e-01
1.00621271e+00 1.30342573e-01 -5.28005064e-02 -5.55434585e-01
-6.13635600e-01 2.04496309e-02 5.45532465e-01 4.30704445e-01
6.42267764e-01 5.93212582e-02 -6.38312876e-01 7.37430632e-01
2.65829116e-01 -1.97314993e-01 3.64328712e-01 7.93137908e-01
-1.55121341e-01 -1.26104856e+00 -5.63292801e-01 5.95261693e-01
-6.08839095e-01 5.79900742e-02 -6.05010867e-01 6.53089106e-01
-1.70201316e-01 7.49511480e-01 -1.58855081e-01 -2.56724536e-01
2.28312194e-01 2.12937802e-01 2.52361625e-01 -5.34224272e-01
2.18955666e-01 6.64196193e-01 -1.75754979e-01 -5.20877540e-01
-4.70740527e-01 -1.04148877e+00 -1.31098270e+00 3.02380789e-02
-2.36521214e-01 2.43665844e-01 7.36648440e-01 5.72639823e-01
9.66048956e-01 6.50559187e-01 3.62194240e-01 -6.07580364e-01
-4.03541237e-01 -7.81742454e-01 -8.54664028e-01 4.14918631e-01
2.67400295e-01 -6.47941947e-01 -4.42941159e-01 2.08342478e-01] | [7.086864471435547, 4.975013732910156] |
f63bb61a-78ba-4a5c-8f91-8fb913fe406b | magnetic-resonance-fingerprinting | 1807.06356 | null | http://arxiv.org/abs/1807.06356v2 | http://arxiv.org/pdf/1807.06356v2.pdf | Magnetic Resonance Fingerprinting Reconstruction via Spatiotemporal Convolutional Neural Networks | Magnetic resonance fingerprinting (MRF) quantifies multiple nuclear magnetic
resonance parameters in a single and fast acquisition. Standard MRF
reconstructs parametric maps using dictionary matching, which lacks scalability
due to computational inefficiency. We propose to perform MRF map reconstruction
using a spatiotemporal convolutional neural network, which exploits the
relationship between neighboring MRF signal evolutions to replace the
dictionary matching. We evaluate our method on multiparametric brain scans and
compare it to three recent MRF reconstruction approaches. Our method achieves
state-of-the-art reconstruction accuracy and yields qualitatively more
appealing maps compared to other reconstruction methods. In addition, the
reconstruction time is significantly reduced compared to a dictionary-based
approach. | ['Sairam Geethanath', 'Olivier Scheidegger', 'Shivaprasad Chikop', 'Vimal Chandran', 'Fabian Balsiger', 'Amaresha Shridhar Konar', 'Mauricio Reyes'] | 2018-07-17 | null | null | null | null | ['magnetic-resonance-fingerprinting'] | ['medical'] | [ 2.26299420e-01 -5.17103635e-02 -1.23197235e-01 -3.76423806e-01
-9.19542789e-01 -2.94985145e-01 4.40069795e-01 8.68544579e-02
-5.58184862e-01 7.36668706e-01 3.06566149e-01 -3.01068127e-02
-5.02768874e-01 -6.41715705e-01 -7.41234660e-01 -6.18777812e-01
-3.25431556e-01 7.23307788e-01 4.60717142e-01 -6.00453243e-02
1.12898529e-01 6.76672697e-01 -7.61408806e-01 6.22013547e-02
6.71333313e-01 9.48521972e-01 5.42454004e-01 1.55123711e-01
2.19091430e-01 8.44140828e-01 6.44177496e-02 1.24839090e-01
5.61171696e-02 -2.06926167e-01 -8.34770262e-01 -5.02967298e-01
5.32004595e-01 -6.65504456e-01 -1.00682151e+00 1.06061316e+00
7.61045873e-01 2.79346526e-01 5.10690451e-01 -4.34400141e-01
-6.68955922e-01 7.29578853e-01 -4.96684670e-01 9.70997930e-01
9.89479348e-02 -5.80169335e-02 3.78966689e-01 -9.89278853e-01
1.05687964e+00 5.65911531e-01 1.05546892e+00 3.18958193e-01
-1.70366216e+00 -4.57440495e-01 -4.31644738e-01 -4.42156830e-04
-1.30118454e+00 -4.90727842e-01 7.49752581e-01 -4.82387304e-01
8.36115718e-01 -1.69419035e-01 6.98377073e-01 8.55364919e-01
7.27663279e-01 4.36138332e-01 1.39396405e+00 -5.57379834e-02
3.24095860e-02 -6.36516511e-01 1.29102394e-01 9.46371734e-01
-2.87679322e-02 4.94461507e-01 -4.92412478e-01 -3.06079477e-01
1.45782161e+00 -5.72211519e-02 -5.64520061e-01 -4.96218562e-01
-1.96865892e+00 7.59132087e-01 2.94862449e-01 6.66954100e-01
-6.22185230e-01 4.89615291e-01 3.04860234e-01 1.91405132e-01
2.18639970e-01 5.00855923e-01 1.74522609e-01 -4.17462848e-02
-1.33027840e+00 1.51726261e-01 1.96878210e-01 1.99947476e-01
4.07801419e-01 2.17956573e-01 -3.15422088e-01 8.55696797e-01
5.87052889e-02 5.05067945e-01 6.94608748e-01 -1.35128260e+00
-7.51355961e-02 -2.88308650e-01 4.51450311e-02 -1.08432472e+00
-1.01423907e+00 -5.89587331e-01 -8.48768413e-01 -8.90687183e-02
3.65904272e-01 1.63072839e-01 -7.17555761e-01 1.74750173e+00
2.49483541e-01 4.87244219e-01 -5.44253469e-01 1.27503979e+00
9.34888065e-01 1.89378425e-01 -2.91999847e-01 -3.20221037e-01
1.03177106e+00 -6.76381469e-01 -9.47840333e-01 2.38317862e-01
3.04917842e-01 -2.91155815e-01 7.81591058e-01 3.12202990e-01
-1.20623839e+00 -2.42651165e-01 -9.28462029e-01 4.58445167e-03
1.76610991e-01 -1.43309817e-01 8.74865592e-01 4.48830962e-01
-1.37255287e+00 9.80565667e-01 -1.16630852e+00 1.39927164e-01
4.48833168e-01 5.61546206e-01 -6.21548951e-01 -5.81774414e-02
-1.41625726e+00 1.09428680e+00 2.56511658e-01 -2.00110674e-01
-1.11301661e+00 -1.50026023e+00 -8.09539378e-01 -4.20646340e-01
-3.76924453e-03 -5.94194651e-01 1.17304385e+00 -8.05871114e-02
-1.54981112e+00 8.25335562e-01 -9.11422670e-02 -6.00998461e-01
5.26755631e-01 3.29009444e-01 -5.97310543e-01 7.91096628e-01
3.05487990e-01 8.33842576e-01 6.04546010e-01 -7.52351880e-01
1.94335431e-01 -3.47212106e-01 -1.55323863e-01 -5.81306517e-02
2.99258322e-01 -1.31741032e-01 -5.38067967e-02 -8.09214771e-01
6.57974124e-01 -7.20981538e-01 -4.00624067e-01 1.10625945e-01
-4.67134863e-02 4.15437967e-01 4.00000572e-01 -8.24459255e-01
9.21990514e-01 -1.92681575e+00 3.71554792e-02 2.75817752e-01
8.83422077e-01 -3.21466267e-01 -7.46749043e-02 -2.02024952e-01
-3.34496260e-01 -3.28089893e-01 -4.54635590e-01 5.77129051e-03
-1.97521016e-01 3.37663479e-02 -1.57751396e-01 1.14669538e+00
-2.87408829e-01 1.27691627e+00 -1.15998828e+00 -4.25493836e-01
2.71657616e-01 7.08880424e-01 -3.48211735e-01 -2.91113406e-01
3.95924002e-01 1.04414856e+00 -2.50344008e-01 5.81552684e-01
5.64462245e-01 -3.36210698e-01 3.74325871e-01 -6.59649611e-01
-2.48804584e-01 1.95275858e-01 -6.78556502e-01 2.37126517e+00
-3.32311243e-01 6.20372236e-01 1.18782975e-01 -1.32489324e+00
8.57177794e-01 4.85728532e-01 1.19692957e+00 -1.15079844e+00
1.72067553e-01 4.84323382e-01 4.46150154e-02 -2.24328533e-01
1.99967578e-01 -4.96418983e-01 2.59517133e-01 7.98999071e-01
2.52089292e-01 4.13532695e-03 -4.11008932e-02 5.73359877e-02
1.30794930e+00 8.94705504e-02 3.95438671e-02 -8.99527133e-01
1.43115282e-01 -2.99931288e-01 3.25525373e-01 1.04892576e+00
-4.83359903e-01 6.90010488e-01 1.12801440e-01 -7.53240883e-01
-1.03742623e+00 -1.39318204e+00 -8.46777618e-01 5.25165737e-01
2.21425951e-01 2.69640926e-02 -5.42144120e-01 -3.08144510e-01
1.29001588e-01 1.91481858e-01 -9.67181802e-01 -3.14262658e-02
-8.97296906e-01 -9.30329621e-01 5.82435489e-01 4.91029292e-01
4.70340312e-01 -8.16515625e-01 -7.14894295e-01 6.66370690e-01
-4.01182860e-01 -1.18900275e+00 -6.21864259e-01 1.52450055e-01
-1.21097100e+00 -6.49184406e-01 -9.55747306e-01 -6.38252854e-01
2.77368695e-01 2.26566210e-01 1.06740642e+00 -2.77304888e-01
-3.65162969e-01 3.57490331e-01 1.40553474e-01 3.10520947e-01
-2.35744506e-01 -9.28817503e-03 2.69651175e-01 -1.20831758e-01
-8.29807743e-02 -1.14237750e+00 -8.62935841e-01 2.61007428e-01
-7.90726125e-01 -9.71931666e-02 5.47453582e-01 9.12745476e-01
1.24488556e+00 -2.53380328e-01 6.88821733e-01 -7.76371002e-01
5.16655684e-01 -5.51891327e-01 -5.14900208e-01 1.08299673e-01
-6.28831744e-01 2.84323305e-01 3.73454064e-01 -4.96971756e-01
-7.11639285e-01 3.96210477e-02 5.72971813e-02 -6.02535963e-01
1.42747641e-01 5.76166511e-01 6.31162167e-01 -7.91640580e-01
6.22619390e-01 6.13203704e-01 3.85533899e-01 -3.80609095e-01
1.11005522e-01 -1.95621967e-01 1.14202321e+00 -6.55827940e-01
3.64159435e-01 8.81456912e-01 2.74287432e-01 -5.72910368e-01
-3.84911954e-01 -2.31653377e-01 -8.57961059e-01 -5.71261168e-01
6.66764557e-01 -5.17613232e-01 -7.79048920e-01 1.56623974e-01
-1.02688754e+00 -4.28372830e-01 -2.84475565e-01 9.31050897e-01
-8.50396514e-01 5.17970741e-01 -8.49942505e-01 -7.52724856e-02
-3.74881536e-01 -1.43147910e+00 9.03647423e-01 -2.96348184e-01
-4.55858968e-02 -1.12181628e+00 4.22712058e-01 -5.11064753e-03
8.69469941e-01 6.14201128e-01 7.40189016e-01 -1.31599888e-01
-3.45281571e-01 1.22313671e-01 -2.37353414e-01 -4.27994996e-01
-2.52972804e-02 -8.48609746e-01 -6.27833903e-01 -3.24034721e-01
2.02564150e-01 -1.04847513e-01 8.86169910e-01 1.04180992e+00
1.31817734e+00 2.26056889e-01 -2.58082032e-01 9.47561085e-01
1.31514812e+00 3.07730466e-01 7.59770274e-01 2.76964843e-01
5.32543540e-01 2.24135384e-01 1.08136879e-02 2.78219342e-01
2.97766536e-01 8.22329283e-01 -1.14199016e-02 -1.32842198e-01
-4.44071233e-01 -8.47157240e-02 -2.80465223e-02 1.02178800e+00
-1.27387747e-01 5.78271985e-01 -1.06199121e+00 5.71723521e-01
-1.62914395e+00 -9.26942647e-01 -8.55666548e-02 1.99308252e+00
7.79463828e-01 -1.67595029e-01 1.27022326e-01 -8.01170021e-02
6.34881556e-01 3.35703582e-01 -8.18715096e-01 2.41403237e-01
-1.40715137e-01 6.03970766e-01 6.35149717e-01 6.36114538e-01
-9.82519686e-01 3.72380733e-01 8.03041935e+00 5.54830968e-01
-1.40077400e+00 9.25607026e-01 4.42543298e-01 -2.64988780e-01
-6.72199428e-01 -3.77468348e-01 -4.47862875e-03 3.37544739e-01
9.59073722e-01 -4.68359403e-02 9.20700550e-01 2.37568453e-01
2.57991701e-01 -5.16316853e-02 -8.53648663e-01 1.37263024e+00
-1.24440774e-01 -1.95445371e+00 -1.84956625e-01 7.14724362e-02
5.17085254e-01 5.54587245e-01 1.69998676e-01 -1.91030771e-01
1.21306457e-01 -1.21472764e+00 7.77299881e-01 9.33910668e-01
1.18321204e+00 -6.16017461e-01 4.33582872e-01 -6.58127442e-02
-1.07088220e+00 3.41490418e-01 -3.33927691e-01 3.58022034e-01
3.64622951e-01 9.76495564e-01 -4.97551143e-01 4.20181751e-01
7.29130805e-01 6.55467570e-01 -1.39879361e-01 8.33494484e-01
1.96985558e-01 3.73442203e-01 -2.23986641e-01 5.47615469e-01
2.50274599e-01 -2.64867425e-01 6.57505810e-01 1.02149105e+00
2.41474345e-01 2.62190193e-01 1.71270937e-01 1.24700999e+00
1.12986825e-02 5.67385554e-02 -6.64933264e-01 8.14322531e-02
4.91541058e-01 1.08896506e+00 -9.09773231e-01 -1.81649223e-01
-2.68454671e-01 7.61437654e-01 3.32521856e-01 1.93927258e-01
-7.40348339e-01 -2.72087365e-01 1.08612806e-01 5.71124792e-01
7.72538707e-02 -4.18634564e-01 -2.62448519e-01 -1.18469834e+00
-1.39811441e-01 -5.08504629e-01 1.03777140e-01 -7.36081183e-01
-1.07432234e+00 6.93006635e-01 6.64232150e-02 -7.35742033e-01
-1.59144357e-01 -2.66039014e-01 -1.95850983e-01 7.80953169e-01
-1.58992934e+00 -6.75393403e-01 -2.47898206e-01 7.74205327e-01
-8.67186934e-02 -5.19293398e-02 7.64588714e-01 6.68266296e-01
-1.19753398e-01 4.04654354e-01 2.11341783e-01 8.45120698e-02
4.59870219e-01 -1.23359895e+00 2.96749353e-01 6.62799120e-01
-1.81872621e-01 7.49320507e-01 2.65965879e-01 -6.24743760e-01
-1.52650201e+00 -8.02826345e-01 5.10280967e-01 -2.58099645e-01
8.40473592e-01 3.58135663e-02 -1.06009531e+00 6.08148515e-01
-1.20155640e-01 8.18461895e-01 5.18357158e-01 -3.58899236e-01
-1.89877898e-01 -1.04805440e-01 -1.45958185e+00 6.86957538e-02
9.83429730e-01 -8.57036412e-01 -5.87436020e-01 2.05771953e-01
5.05990446e-01 -6.83710694e-01 -1.53447843e+00 3.47289979e-01
9.53375518e-01 -9.05696034e-01 1.11949492e+00 -1.71136275e-01
2.35921144e-01 -2.79693484e-01 -2.24653095e-01 -1.24698341e+00
-7.65089869e-01 -6.19524598e-01 -2.01428428e-01 2.47818664e-01
4.63367291e-02 -8.54774058e-01 5.95463395e-01 3.90675485e-01
-3.42892170e-01 -7.17501104e-01 -1.35599315e+00 -8.52314651e-01
2.38971740e-01 -3.78756940e-01 6.44258559e-01 1.26242316e+00
-2.23800130e-02 -2.58827239e-01 -3.30686569e-01 8.06557387e-03
1.05874348e+00 1.73058584e-01 -1.99365884e-01 -1.13101995e+00
-3.58173043e-01 -4.95581627e-01 -3.30383569e-01 -8.41199338e-01
2.74020225e-01 -1.50851357e+00 -1.20318405e-01 -1.26878893e+00
3.39754760e-01 -5.89593470e-01 -6.81692660e-01 3.55655044e-01
4.18051720e-01 7.37269700e-01 -9.46032777e-02 4.28621382e-01
-3.38324755e-01 3.44566017e-01 1.75613320e+00 -2.10173070e-01
7.96625838e-02 -7.76077390e-01 -2.85852194e-01 3.65113080e-01
5.91400802e-01 -5.28594732e-01 -4.03169960e-01 -4.33011502e-01
-3.57161574e-02 8.41463625e-01 5.88763714e-01 -1.08077443e+00
2.39200696e-01 1.12714805e-01 5.79831123e-01 -5.15585244e-01
1.42300412e-01 -3.12089860e-01 5.85216880e-01 6.32555723e-01
-3.79551142e-01 2.22279146e-01 1.52818978e-01 2.95626312e-01
-1.81123972e-01 -1.19043589e-01 1.14653742e+00 -3.57289284e-01
-4.35456216e-01 7.69225657e-01 -4.61682320e-01 2.74516344e-02
4.22643632e-01 -6.57196715e-02 -8.14888924e-02 -1.62947461e-01
-1.02559543e+00 -4.89353806e-01 1.81681842e-01 -6.20341413e-02
8.10065031e-01 -1.65249395e+00 -6.06140912e-01 1.31176651e-01
-3.31784964e-01 -5.20440817e-01 5.94065785e-01 1.66254699e+00
-7.67674685e-01 6.15419030e-01 -5.95174432e-01 -8.00525486e-01
-3.33297104e-01 4.47042733e-01 1.02421713e+00 -4.90837365e-01
-1.27999032e+00 3.79182011e-01 1.61721155e-01 -6.47942126e-01
-2.40125343e-01 -3.42046797e-01 -8.90016779e-02 -1.80899099e-01
7.32566178e-01 3.14170718e-01 3.00225526e-01 -5.50735950e-01
-5.83745599e-01 5.29467583e-01 -3.79275493e-02 -3.46166611e-01
1.52289796e+00 -7.03443512e-02 -1.61859944e-01 2.91567057e-01
1.21593523e+00 -2.61656255e-01 -1.19164634e+00 -5.37999868e-01
-1.59273297e-01 -3.91369998e-01 9.42093670e-01 -8.78550708e-01
-1.53192878e+00 6.90586746e-01 9.45939362e-01 -2.97216505e-01
8.27661335e-01 -9.41300206e-03 1.14722276e+00 1.56853035e-01
8.31361294e-01 -6.62525952e-01 -6.01176880e-02 2.98019350e-01
9.35391724e-01 -1.03519034e+00 -4.67981212e-02 -1.15410298e-01
-1.87034070e-01 1.26093554e+00 9.09009855e-03 -2.84755170e-01
8.13672781e-01 5.69434106e-01 -3.10153008e-01 -6.58640444e-01
-1.52744353e-01 3.90902519e-01 3.74216497e-01 7.19094813e-01
3.10561717e-01 1.69004425e-01 -3.90678406e-01 5.41205525e-01
-4.81444299e-02 3.47060442e-01 3.90688658e-01 6.32973969e-01
-2.23879397e-01 -8.18075478e-01 -1.73842609e-01 6.46037579e-01
-5.13418257e-01 -2.24656299e-01 3.41839790e-01 6.93397582e-01
-1.89021006e-01 2.28488490e-01 7.72835687e-02 -1.15800381e-01
1.31939679e-01 -1.35102227e-01 1.12255311e+00 -1.23887196e-01
-3.78103256e-01 1.39754429e-01 -1.85700119e-01 -1.09977686e+00
-6.21047854e-01 -9.00534987e-01 -1.49231672e+00 -3.41079205e-01
2.30114326e-01 -1.29512444e-01 5.45840502e-01 9.26165640e-01
3.42928737e-01 5.49936354e-01 4.50353682e-01 -9.15667057e-01
-2.31821328e-01 -6.00209951e-01 -1.02536666e+00 1.70642778e-01
4.23072696e-01 -1.20471931e+00 1.21558629e-01 -4.40117717e-01] | [13.531961441040039, -2.4053547382354736] |
f181d33c-941b-4651-abd3-dfbf8a7eb5a7 | a-hierarchical-regression-chain-framework-for | 2303.08027 | null | https://arxiv.org/abs/2303.08027v1 | https://arxiv.org/pdf/2303.08027v1.pdf | A Hierarchical Regression Chain Framework for Affective Vocal Burst Recognition | As a common way of emotion signaling via non-linguistic vocalizations, vocal burst (VB) plays an important role in daily social interaction. Understanding and modeling human vocal bursts are indispensable for developing robust and general artificial intelligence. Exploring computational approaches for understanding vocal bursts is attracting increasing research attention. In this work, we propose a hierarchical framework, based on chain regression models, for affective recognition from VBs, that explicitly considers multiple relationships: (i) between emotional states and diverse cultures; (ii) between low-dimensional (arousal & valence) and high-dimensional (10 emotion classes) emotion spaces; and (iii) between various emotion classes within the high-dimensional space. To address the challenge of data sparsity, we also use self-supervised learning (SSL) representations with layer-wise and temporal aggregation modules. The proposed systems participated in the ACII Affective Vocal Burst (A-VB) Challenge 2022 and ranked first in the "TWO'' and "CULTURE'' tasks. Experimental results based on the ACII Challenge 2022 dataset demonstrate the superior performance of the proposed system and the effectiveness of considering multiple relationships using hierarchical regression chain models. | ['Helen Meng', 'Xunying Liu', 'Dongsheng Li', 'Kaitao Song', 'Xixin Wu', 'Jinchao Li'] | 2023-03-14 | null | null | null | null | ['a-vb-culture', 'a-vb-high', 'culture', 'a-vb-two'] | ['speech', 'speech', 'speech', 'speech'] | [-6.70893639e-02 -3.71623278e-01 -1.19727649e-01 -5.66720605e-01
-2.99404681e-01 -3.00990164e-01 2.16693014e-01 -1.18154541e-01
-1.91737473e-01 6.18484974e-01 4.47179973e-01 3.38067055e-01
-8.34048092e-02 -2.62223363e-01 -2.27442756e-01 -6.12472296e-01
-3.51394594e-01 1.03034250e-01 -4.89652991e-01 -2.70089060e-01
-1.48987114e-01 5.24036050e-01 -2.16572189e+00 5.77457130e-01
6.01525903e-01 1.50567079e+00 -2.09529534e-01 7.35640883e-01
-3.45953763e-01 1.02216792e+00 -5.35350621e-01 -2.16525257e-01
-7.06153512e-02 -7.01694787e-01 -6.21208549e-01 8.46740603e-02
-8.56542438e-02 3.23545456e-01 1.81233227e-01 7.38509417e-01
4.16343123e-01 3.95471364e-01 6.36896729e-01 -1.45891225e+00
-3.68158042e-01 6.38430357e-01 -4.41782326e-01 4.60132062e-02
3.31973821e-01 -1.21865258e-01 1.29734707e+00 -9.11835909e-01
3.10062706e-01 1.20425129e+00 7.69647837e-01 7.22328782e-01
-1.03202140e+00 -9.94905949e-01 3.65734220e-01 3.84319037e-01
-1.16719031e+00 -2.80384213e-01 9.68401492e-01 -6.56643212e-01
9.88015771e-01 5.91617763e-01 1.05450726e+00 1.30698216e+00
-9.46241524e-03 8.07701409e-01 1.38630223e+00 -8.76806453e-02
6.72037378e-02 3.82133126e-01 3.40617895e-01 5.86987019e-01
-5.93538880e-01 4.44800453e-03 -9.73784149e-01 -2.39102185e-01
1.10468738e-01 -2.53954649e-01 9.16196704e-02 3.13664168e-01
-8.62814486e-01 9.16545033e-01 2.97316134e-01 5.12440383e-01
-6.23915136e-01 1.67444244e-01 7.16346383e-01 2.71146238e-01
8.41557324e-01 5.65791309e-01 -4.80569899e-01 -4.45912838e-01
-7.47866213e-01 -1.54618189e-01 7.62465537e-01 7.73858190e-01
5.98528266e-01 5.66617012e-01 1.19266780e-02 1.56348944e+00
2.83662647e-01 1.67764723e-01 6.91225946e-01 -7.29098618e-01
-4.73797694e-02 4.98356313e-01 -2.25026533e-01 -1.14920926e+00
-8.52585673e-01 -6.30404353e-01 -9.46658969e-01 -3.75658095e-01
-2.65914619e-01 -3.45027864e-01 -3.97256553e-01 2.05442834e+00
2.69708455e-01 4.14166749e-01 9.04850438e-02 8.61702561e-01
1.27320957e+00 1.14282525e+00 2.69809067e-01 -9.54062045e-01
1.25937164e+00 -8.70590925e-01 -1.04259109e+00 -5.72606362e-02
2.46610448e-01 -6.37182355e-01 1.21749270e+00 7.15400100e-01
-9.64397967e-01 -5.97558260e-01 -8.53884041e-01 1.49119154e-01
-4.78234023e-01 3.68257880e-01 1.02708030e+00 4.84101146e-01
-8.80766094e-01 2.98399091e-01 -3.33686262e-01 -2.56223202e-01
-5.73551580e-02 5.61326563e-01 -1.32469624e-01 5.59575140e-01
-1.44072151e+00 5.85238278e-01 -1.14309341e-01 4.81925279e-01
-4.49221969e-01 -6.22515738e-01 -7.17452824e-01 4.31246273e-02
-3.10090948e-02 -6.22073896e-02 9.25839126e-01 -1.27416599e+00
-1.77235329e+00 6.46957457e-01 7.79100358e-02 -1.11192055e-01
-1.37855202e-01 -3.54881525e-01 -8.41938853e-01 8.86396915e-02
-3.06518108e-01 4.45924282e-01 8.74218345e-01 -1.30760133e+00
-3.82842094e-01 -3.37087333e-01 -5.26035309e-01 2.57977962e-01
-8.00435066e-01 3.60883564e-01 -1.60630912e-01 -6.20231032e-01
-8.12887773e-02 -1.30499947e+00 -5.21506742e-02 -5.67373097e-01
-3.01280797e-01 -6.38519824e-01 7.16374159e-01 -5.13691962e-01
1.67876577e+00 -2.21173668e+00 4.56663936e-01 2.45670080e-01
7.53658637e-02 1.01219676e-02 -1.68265402e-01 3.73544246e-01
-2.26248845e-01 1.34955151e-02 8.24734122e-02 -6.20163441e-01
6.67245910e-02 2.33353615e-01 -3.44231337e-01 3.25572014e-01
1.64307892e-01 3.73986840e-01 -7.53746986e-01 -5.38797498e-01
-9.03967172e-02 3.65965903e-01 -6.43459022e-01 6.91867411e-01
-1.39592379e-01 4.06084538e-01 -1.69743016e-01 7.72422671e-01
1.94125131e-01 1.88129842e-02 1.70743376e-01 -5.89783490e-01
-4.51359659e-01 -1.07484370e-01 -7.13657737e-01 1.40119457e+00
-7.29199827e-01 6.24538422e-01 2.46254981e-01 -8.14926207e-01
1.06628442e+00 4.38215315e-01 1.06798947e+00 -3.79428834e-01
4.46756512e-01 9.08875540e-02 -1.85405537e-02 -7.98551083e-01
5.20091534e-01 -6.24143481e-01 -5.81691325e-01 3.05095583e-01
4.76677626e-01 -1.99043572e-01 9.22374893e-03 -3.30764055e-02
5.61697066e-01 -8.65523890e-02 1.48093238e-01 -1.93706840e-01
4.65467989e-01 -4.61027145e-01 8.60224783e-01 6.19642362e-02
-2.88852602e-01 2.16239735e-01 6.25905275e-01 -3.76193643e-01
-4.00717705e-01 -6.88510001e-01 -8.09579249e-03 1.59290516e+00
-3.84998396e-02 -5.04967749e-01 -4.69788194e-01 -3.40228945e-01
-9.45634395e-02 5.89532673e-01 -9.78480756e-01 -3.08096498e-01
-2.14394093e-01 -8.01980793e-01 6.72523558e-01 4.09768999e-01
8.45347792e-02 -1.32896483e+00 -2.56108940e-01 1.92705438e-01
-4.69531924e-01 -1.13750064e+00 -3.80906075e-01 4.49264854e-01
-3.18443060e-01 -4.96253073e-01 -6.56736121e-02 -7.41976738e-01
7.15703815e-02 -1.61022708e-01 8.79080534e-01 -1.88097969e-01
-5.81029177e-01 6.89818144e-01 -5.55971622e-01 -4.62740719e-01
-1.62308499e-01 -1.13304302e-01 5.80831826e-01 6.29657924e-01
1.66270852e-01 -7.71105707e-01 -3.17967296e-01 2.13970616e-01
-6.49158895e-01 -1.25912234e-01 3.91823560e-01 7.77236998e-01
5.32679081e-01 -3.19854468e-01 1.21887505e+00 -7.89246798e-01
1.03415966e+00 -1.05579460e+00 -7.36936778e-02 -2.58512353e-03
-3.44258785e-01 -4.77323800e-01 7.54299521e-01 -7.67782271e-01
-1.00292277e+00 3.52922603e-02 -3.89204085e-01 -7.04126954e-01
7.05429763e-02 6.61536992e-01 7.93506950e-02 2.59165496e-01
3.29494089e-01 -3.38477045e-02 -7.97306597e-02 -2.85538107e-01
4.90093589e-01 9.93333220e-01 3.42098653e-01 -6.16995752e-01
1.37091428e-01 4.13117111e-02 -1.70875937e-01 -1.35389411e+00
-9.16260540e-01 -7.54814446e-01 -3.92638862e-01 -9.02846992e-01
1.20676160e+00 -9.92021263e-01 -1.10393977e+00 2.98266292e-01
-9.08676863e-01 -4.93147857e-02 -1.62920430e-01 6.55061245e-01
-6.51105881e-01 4.93704937e-02 -8.76064420e-01 -1.17388499e+00
-4.45824742e-01 -9.58521724e-01 9.15270627e-01 2.70622641e-01
-6.92513943e-01 -6.81627572e-01 3.29496741e-01 5.37095904e-01
3.37418526e-01 3.88050705e-01 9.76457298e-01 -6.33165359e-01
3.18819582e-01 6.00093342e-02 1.47744924e-01 7.02073932e-01
2.18792647e-01 1.74466997e-01 -1.10834217e+00 3.62006500e-02
6.36201277e-02 -1.08646595e+00 5.84812939e-01 1.28555536e-01
1.35566783e+00 -4.95620728e-01 2.44593412e-01 5.26630521e-01
9.72147167e-01 3.56982380e-01 1.99822128e-01 -4.62755263e-01
6.47588909e-01 1.09425497e+00 7.21014857e-01 8.65707874e-01
4.16666120e-01 5.73521256e-01 4.71901745e-01 -2.19955623e-01
8.08297023e-02 1.02906421e-01 5.86360574e-01 1.98825598e+00
-1.16186194e-01 1.41436514e-02 -5.04551053e-01 5.15038729e-01
-1.77389002e+00 -1.09541368e+00 -1.47742867e-01 1.56398976e+00
9.87997591e-01 -4.44384128e-01 4.68812943e-01 3.83341871e-02
4.42287475e-01 1.85611516e-01 -6.42530203e-01 -1.27145016e+00
-1.17162265e-01 1.29450679e-01 -3.86714548e-01 3.15739393e-01
-1.01334929e+00 1.04810524e+00 5.67702484e+00 6.87109828e-01
-1.47178423e+00 5.18421456e-02 6.19853377e-01 -4.22913909e-01
-1.75572500e-01 -6.58966184e-01 -6.44934475e-01 4.83934999e-01
1.12527645e+00 -4.61584926e-02 7.41988301e-01 8.64101946e-01
2.00475499e-01 3.43602151e-01 -8.55641842e-01 1.20746160e+00
2.93362200e-01 -8.71475160e-01 -1.97212026e-01 -2.13102147e-01
7.70970106e-01 -1.22742891e-01 2.52921760e-01 8.80355120e-01
1.95709784e-02 -1.14019346e+00 6.47078097e-01 7.01194286e-01
8.75670850e-01 -1.01563883e+00 4.64641154e-01 1.94926992e-01
-1.47288263e+00 -3.45993489e-01 -7.09675997e-02 9.97092500e-02
-1.20782435e-01 2.64693975e-01 -6.63580775e-01 2.91414380e-01
9.47569072e-01 9.85811412e-01 -1.12790331e-01 3.43639582e-01
2.26292789e-01 9.33029711e-01 -1.87873170e-01 -5.74686289e-01
3.29398692e-01 -5.63311636e-01 4.71884280e-01 1.80407894e+00
2.95062423e-01 4.30930465e-01 9.32528749e-02 6.42126501e-01
-1.62529677e-01 5.17935932e-01 -5.33120096e-01 -5.82550943e-01
2.13895410e-01 1.57619703e+00 -5.04273474e-01 -7.38551989e-02
-3.37100774e-01 7.32016087e-01 3.14410388e-01 2.00640082e-01
-1.11718488e+00 -1.47521466e-01 1.06480968e+00 -3.64856184e-01
2.17149537e-02 -1.50621608e-01 1.56883653e-02 -9.63513374e-01
-4.56175327e-01 -1.02361929e+00 3.49493921e-01 -6.86195791e-01
-1.61304021e+00 1.18712378e+00 -2.71519423e-01 -1.47661126e+00
-3.04148525e-01 -3.19356322e-01 -2.48897612e-01 4.15325522e-01
-1.10537636e+00 -1.25498343e+00 -3.04122657e-01 7.63851643e-01
7.46213675e-01 -1.97386250e-01 1.10384548e+00 4.41253006e-01
-6.87436342e-01 5.27956665e-01 -1.62442788e-01 -3.58090639e-01
5.79119980e-01 -1.05035293e+00 -7.36518264e-01 1.39388189e-01
7.27225095e-02 5.29415250e-01 5.93713582e-01 -3.33506316e-01
-1.49529195e+00 -1.15359950e+00 6.68227971e-01 1.28008515e-01
8.06205869e-01 -6.23071015e-01 -7.83465087e-01 3.49337608e-01
3.08472365e-01 -9.91872102e-02 1.55849695e+00 4.90492076e-01
-2.07517356e-01 -4.05023128e-01 -8.70706499e-01 4.14565921e-01
7.30698466e-01 -5.96246839e-01 -1.72149628e-01 2.99104691e-01
8.28675866e-01 1.80559710e-01 -1.08977056e+00 4.92576569e-01
8.31846058e-01 -8.68951917e-01 7.25575686e-01 -8.91282499e-01
5.32337904e-01 3.97361182e-02 -4.36517477e-01 -1.44720650e+00
-1.58869565e-01 -6.62946999e-01 -7.14958683e-02 1.48467612e+00
1.89473331e-01 -1.30025968e-01 4.74896610e-01 2.17478871e-01
-2.70994991e-01 -1.28321099e+00 -1.08344579e+00 -4.68086004e-01
-3.24066281e-01 -7.11860359e-01 3.20300460e-01 1.08359039e+00
4.11959708e-01 6.91422462e-01 -1.07426119e+00 -8.94950479e-02
1.19164921e-01 3.08077693e-01 5.29623330e-01 -1.21067560e+00
-1.10573292e-01 -5.75021267e-01 -2.00252205e-01 -5.82136810e-01
5.57383776e-01 -7.21996248e-01 4.02539551e-01 -1.11620307e+00
1.77188218e-02 -3.56807411e-01 -5.73523283e-01 3.87565047e-01
-4.41113161e-03 3.02049041e-01 1.96279779e-01 5.54120392e-02
-6.99391365e-01 9.70623910e-01 7.14032471e-01 2.06293851e-01
-2.76618004e-01 1.11496486e-01 -5.66401601e-01 9.31428611e-01
7.39084840e-01 -2.79102594e-01 -4.90149349e-01 3.90367806e-02
4.54590529e-01 4.25859392e-01 -2.98711777e-01 -7.65707672e-01
1.10089391e-01 -3.03596944e-01 7.24489763e-02 -5.49635768e-01
1.14444935e+00 -7.96549976e-01 3.76145877e-02 1.81614697e-01
-5.62325418e-01 -1.36775877e-02 2.98641741e-01 5.06614685e-01
-4.92196411e-01 2.01309398e-01 9.81247544e-01 5.43676130e-02
-4.03634578e-01 3.27437431e-01 -7.78914154e-01 -1.95129097e-01
9.13558006e-01 1.62571132e-01 8.31867903e-02 -6.28878593e-01
-1.09362483e+00 2.07667962e-01 -3.81881356e-01 7.03455627e-01
7.38426745e-01 -1.47066033e+00 -5.46419024e-01 1.24251679e-01
7.85693899e-02 -7.33461916e-01 4.93373573e-01 8.16652358e-01
-4.05501993e-03 3.67279872e-02 -3.67218375e-01 -5.35632849e-01
-1.61292505e+00 2.50601172e-01 1.75190777e-01 -3.51862043e-01
1.86492965e-01 1.26722538e+00 1.92402855e-01 -5.50287724e-01
4.43833470e-01 -2.72854239e-01 -7.09585249e-01 8.40462923e-01
1.87765136e-01 4.14971828e-01 -1.91943511e-01 -1.02263594e+00
-4.09866542e-01 5.48366785e-01 2.73486346e-01 -1.97892040e-02
1.65187013e+00 -1.02646798e-01 -4.54917878e-01 1.21169293e+00
1.52909756e+00 2.01560259e-01 -7.32275426e-01 -7.74906948e-02
1.44884577e-02 4.62947637e-02 6.88902941e-03 -8.59360635e-01
-1.07163358e+00 1.03114843e+00 5.10188520e-01 5.31233549e-01
1.41394305e+00 -3.05905603e-02 8.81753147e-01 1.40811116e-01
-8.29822384e-03 -1.41688192e+00 6.20716929e-01 5.53535819e-01
1.37799788e+00 -9.36767459e-01 -1.78831175e-01 -2.61058897e-01
-1.14011455e+00 1.08341348e+00 8.57178211e-01 3.13417725e-02
7.20049679e-01 1.88366115e-01 1.88178733e-01 -2.67663360e-01
-1.20515776e+00 -1.89200148e-01 5.74021101e-01 7.74462596e-02
6.40673220e-01 3.04725200e-01 -4.08298910e-01 1.45708740e+00
-3.30815166e-01 -3.16354841e-01 1.27374351e-01 4.32175457e-01
-2.46679246e-01 -7.86138058e-01 -8.26544911e-02 4.94918168e-01
-4.04140949e-01 5.01724593e-02 -7.24427342e-01 4.34490234e-01
4.09609884e-01 1.21625280e+00 -8.51293430e-02 -1.00817001e+00
3.05505633e-01 1.03215158e-01 1.63624376e-01 -6.11695230e-01
-1.17085850e+00 4.10095870e-01 4.30900872e-01 -7.02447116e-01
-5.51627815e-01 -7.60855496e-01 -1.23966026e+00 1.39430180e-01
-2.22494975e-01 5.06261647e-01 9.67748046e-01 7.10015833e-01
2.73473829e-01 7.07385540e-01 1.30008030e+00 -8.12079847e-01
-2.86100000e-01 -1.15571713e+00 -8.43804896e-01 4.41158444e-01
2.24186510e-01 -6.45411909e-01 -5.76973855e-01 1.00576743e-01] | [13.472016334533691, 5.603561878204346] |
6f9cd454-a79d-4fc2-b4d0-9b668cd0fcc5 | mutual-supervised-feature-modulation-network | 2010.10744 | null | https://arxiv.org/abs/2010.10744v1 | https://arxiv.org/pdf/2010.10744v1.pdf | Mutual-Supervised Feature Modulation Network for Occluded Pedestrian Detection | State-of-the-art pedestrian detectors have achieved significant progress on non-occluded pedestrians, yet they are still struggling under heavy occlusions. The recent occlusion handling strategy of popular two-stage approaches is to build a two-branch architecture with the help of additional visible body annotations. Nonetheless, these methods still have some weaknesses. Either the two branches are trained independently with only score-level fusion, which cannot guarantee the detectors to learn robust enough pedestrian features. Or the attention mechanisms are exploited to only emphasize on the visible body features. However, the visible body features of heavily occluded pedestrians are concentrated on a relatively small area, which will easily cause missing detections. To address the above issues, we propose in this paper a novel Mutual-Supervised Feature Modulation (MSFM) network, to better handle occluded pedestrian detection. The key MSFM module in our network calculates the similarity loss of full body boxes and visible body boxes corresponding to the same pedestrian so that the full-body detector could learn more complete and robust pedestrian features with the assist of contextual features from the occluding parts. To facilitate the MSFM module, we also propose a novel two-branch architecture, consisting of a standard full body detection branch and an extra visible body classification branch. These two branches are trained in a mutual-supervised way with full body annotations and visible body annotations, respectively. To verify the effectiveness of our proposed method, extensive experiments are conducted on two challenging pedestrian datasets: Caltech and CityPersons, and our approach achieves superior performance compared to other state-of-the-art methods on both datasets, especially in heavy occlusion case. | ['Xu-Cheng Yin', 'Chao Zhu', 'Ye He'] | 2020-10-21 | null | null | null | null | ['body-detection', 'occlusion-handling'] | ['computer-vision', 'computer-vision'] | [-2.43911415e-01 1.15166247e-01 -1.06487840e-01 -2.58482724e-01
-3.15898538e-01 4.35622633e-02 4.67851132e-01 9.89592597e-02
-3.72117400e-01 6.41931236e-01 6.52205050e-02 1.53268546e-01
5.44683158e-01 -7.67741501e-01 -5.50841630e-01 -8.15561831e-01
1.48359880e-01 1.80641264e-01 1.10959578e+00 -1.91915721e-01
-2.22629577e-01 6.77697286e-02 -1.74250472e+00 2.65864104e-01
9.30699408e-01 9.16474521e-01 8.68525356e-02 3.36489230e-01
-8.69422257e-02 3.72556627e-01 -4.41270411e-01 -6.06792092e-01
5.45060813e-01 -8.94240439e-02 -1.50148213e-01 3.64079475e-01
5.96330345e-01 -4.71004397e-01 -3.95021290e-01 9.58872557e-01
6.81906998e-01 5.62969130e-03 5.24819911e-01 -1.41326463e+00
-1.44826233e-01 1.84376374e-01 -1.09477556e+00 2.90852785e-01
3.86896610e-01 5.23906350e-01 6.69214427e-01 -8.62578869e-01
2.66880363e-01 1.56513548e+00 7.47875571e-01 4.62936491e-01
-8.67821753e-01 -7.39085853e-01 7.05997050e-01 2.09379703e-01
-1.31431687e+00 -4.31975752e-01 7.92341113e-01 -5.06590664e-01
4.81074721e-01 6.33703992e-02 9.19257402e-01 7.89577663e-01
3.93704250e-02 1.10649598e+00 1.02177203e+00 -2.69789517e-01
-6.81915730e-02 3.69371504e-01 3.17154527e-01 9.44171607e-01
6.77129984e-01 3.10753673e-01 -1.27424851e-01 5.37913665e-03
5.33806562e-01 2.18083963e-01 -9.31159109e-02 -7.96976805e-01
-7.49577761e-01 7.15813458e-01 7.89772153e-01 7.06946105e-02
-2.43994091e-02 -2.43335769e-01 5.24090052e-01 -1.89834014e-01
2.70373732e-01 -6.25952542e-01 -2.17477009e-01 5.90707898e-01
-9.04992819e-01 2.58917034e-01 5.00839055e-01 1.00566661e+00
7.27870107e-01 -8.82658213e-02 -4.21558976e-01 4.88772094e-01
9.32006836e-01 5.02070427e-01 2.05988094e-01 -1.20502613e-01
8.45398068e-01 8.79802585e-01 1.47366703e-01 -1.03318024e+00
-5.32845497e-01 -6.54263556e-01 -8.44488204e-01 5.43955445e-01
7.09993958e-01 -5.59305213e-02 -1.01398563e+00 1.56508541e+00
9.65499401e-01 -6.53905049e-02 -1.78347781e-01 1.24720991e+00
1.23073721e+00 4.26450133e-01 5.53279579e-01 2.55262498e-02
1.77194953e+00 -1.28448308e+00 -4.24736291e-01 -5.25288761e-01
4.79302257e-01 -7.97042131e-01 6.50162220e-01 1.20613473e-02
-8.69083226e-01 -1.08330560e+00 -1.26297712e+00 -1.24386460e-01
-3.24719906e-01 6.44390285e-01 5.62173724e-01 9.63054955e-01
-5.67873716e-01 2.28770792e-01 -7.27220118e-01 -4.90341872e-01
5.61394274e-01 2.83503592e-01 -4.40992177e-01 2.86054760e-02
-9.64686334e-01 8.43020380e-01 3.53723675e-01 4.46238875e-01
-8.90016854e-01 -1.27042934e-01 -9.92003739e-01 -1.71781164e-02
3.53493810e-01 -9.12042201e-01 7.26091683e-01 -7.83785164e-01
-1.13059139e+00 9.44515586e-01 -1.71714231e-01 -3.32283109e-01
8.65478158e-01 -4.40515041e-01 -3.22956443e-01 -2.00309962e-01
3.25967014e-01 7.57645011e-01 9.29153919e-01 -1.27064240e+00
-1.08306396e+00 -6.48562729e-01 -1.22443862e-01 2.05786958e-01
-2.55054861e-01 6.65878579e-02 -5.82981706e-01 -3.97649914e-01
2.54511029e-01 -7.06914842e-01 -3.24704617e-01 4.15217251e-01
-5.37991464e-01 -3.60918522e-01 9.86586034e-01 -6.97578967e-01
1.01386547e+00 -2.14475870e+00 -1.48258999e-01 -4.59658355e-02
3.17562848e-01 6.47144496e-01 9.65459868e-02 -8.01093429e-02
9.96057913e-02 -2.72820085e-01 7.29694739e-02 -7.01456964e-01
-1.50374383e-01 4.46182303e-02 1.06119096e-01 7.05168009e-01
3.07342052e-01 5.76458514e-01 -6.73059225e-01 -1.03640258e+00
4.87835824e-01 4.37964976e-01 -2.69133717e-01 2.07422704e-01
1.90420896e-01 5.26391447e-01 -5.57329655e-01 9.00190890e-01
1.07042181e+00 1.75897077e-01 -2.25778833e-01 -3.68657470e-01
-3.33627850e-01 -1.55345306e-01 -1.62986231e+00 1.33964562e+00
6.31429255e-02 1.85343876e-01 2.54003853e-01 -9.39936757e-01
9.54518378e-01 2.69882739e-01 1.95384189e-01 -4.36721444e-01
3.41850132e-01 5.29918782e-02 1.66654676e-01 -4.82450068e-01
7.97030851e-02 1.87369213e-01 1.69879586e-01 -1.91528603e-01
-3.20507102e-02 5.77014923e-01 4.14611995e-01 6.33730069e-02
7.14117646e-01 3.99761379e-01 3.31565022e-01 -1.61970809e-01
1.11449540e+00 -2.30450377e-01 9.63844836e-01 6.92605257e-01
-8.86752605e-01 5.53009748e-01 1.29609659e-01 -7.92135656e-01
-8.46601367e-01 -1.01440835e+00 -1.47527725e-01 1.12109113e+00
5.56397855e-01 -4.17575836e-01 -6.59051955e-01 -9.70600784e-01
4.72506210e-02 -7.34920874e-02 -4.18900102e-01 3.14135440e-02
-7.79140949e-01 -9.26459074e-01 2.96789408e-01 7.97547519e-01
9.56982672e-01 -8.31495702e-01 -7.90788531e-01 2.03531027e-01
-1.56337276e-01 -1.11708081e+00 -3.38352054e-01 -1.68743040e-02
-6.31065845e-01 -1.13159156e+00 -9.36109245e-01 -8.51270556e-01
7.28471279e-01 6.14864051e-01 7.38656521e-01 3.21625441e-01
-4.77001190e-01 -1.90219715e-01 -2.29953110e-01 -4.29092228e-01
2.27692425e-02 3.43676307e-03 6.18729368e-02 2.05419540e-01
3.80988061e-01 -4.57057446e-01 -9.29522812e-01 6.69320762e-01
-2.10235223e-01 8.16188455e-02 8.36161911e-01 8.12090218e-01
4.41451669e-01 3.65775079e-02 3.21321756e-01 -3.46411109e-01
-1.21675469e-01 -2.79157996e-01 -6.36068702e-01 1.18906744e-01
-9.73782912e-02 -1.76154912e-01 5.06096900e-01 -4.09427702e-01
-1.25557661e+00 4.63000685e-01 -3.35391998e-01 -9.89248827e-02
-5.11565506e-01 -2.91539043e-01 -7.91481018e-01 -1.78772360e-01
5.36652267e-01 -9.56868846e-03 -1.94729403e-01 -5.82940817e-01
3.33141506e-01 5.65098643e-01 5.98765850e-01 -5.34555554e-01
1.03762925e+00 6.04860127e-01 6.13988303e-02 -4.79380250e-01
-7.73302019e-01 -9.18456316e-01 -8.92852128e-01 -3.35742921e-01
9.29059267e-01 -1.13969934e+00 -5.54833412e-01 4.11104023e-01
-1.09807396e+00 2.34378040e-01 -1.39666587e-01 4.04658586e-01
-7.71909282e-02 7.75232375e-01 -5.27551413e-01 -1.14804220e+00
-4.72939938e-01 -1.21053040e+00 1.36012590e+00 6.21082783e-01
2.39633426e-01 -3.95905793e-01 -2.09963873e-01 3.80688310e-01
1.00455008e-01 2.83547997e-01 4.15703148e-01 -4.68905568e-01
-5.68915308e-01 -4.75528359e-01 -5.38876057e-01 1.44615680e-01
-9.44041312e-02 -6.10474274e-02 -1.12166357e+00 -4.70301837e-01
-2.74534494e-01 -1.44188136e-01 1.13279057e+00 4.19966072e-01
5.19090950e-01 5.76847605e-02 -8.83279204e-01 4.44990665e-01
1.16804838e+00 -8.04679543e-02 5.74801445e-01 3.97576392e-01
6.75218940e-01 6.37109220e-01 6.14059806e-01 3.40479463e-01
4.46887136e-01 8.67076516e-01 5.12438118e-01 -5.36138237e-01
-4.74948317e-01 -3.62452030e-01 4.87947464e-01 4.16674256e-01
-2.45265588e-01 2.45061032e-02 -4.71955985e-01 4.21134025e-01
-2.02726221e+00 -8.46450329e-01 -5.38868308e-01 2.32558370e+00
3.46903443e-01 5.78704834e-01 6.20591164e-01 4.30099964e-01
8.89842868e-01 7.42320120e-02 -3.43731761e-01 2.05497086e-01
-1.10659346e-01 -2.65170872e-01 3.71800959e-01 6.23426363e-02
-1.82448220e+00 7.52375543e-01 5.21644402e+00 7.80673146e-01
-6.62674129e-01 2.97498077e-01 4.44970578e-01 2.27069572e-01
6.52406752e-01 1.21493340e-01 -1.48472989e+00 6.59779966e-01
8.81678760e-02 5.44118226e-01 -5.23060024e-01 1.25499320e+00
1.10863477e-01 -3.11527312e-01 -7.92866230e-01 8.73440266e-01
8.96979868e-03 -5.74504495e-01 -2.37320036e-01 -9.60347578e-02
3.85686964e-01 -2.58049339e-01 -1.10156551e-01 4.95291680e-01
9.29150954e-02 -4.78396147e-01 6.84333503e-01 3.20495784e-01
3.05317938e-01 -4.61109310e-01 1.07015967e+00 6.29799008e-01
-1.98523414e+00 -2.17555761e-01 -7.18376279e-01 -1.92361474e-01
4.04650956e-01 5.94899535e-01 -2.63093263e-01 7.21848667e-01
8.72483015e-01 5.19242465e-01 -9.32243705e-01 1.61152685e+00
-4.38204795e-01 2.82570153e-01 -4.96821046e-01 8.91653076e-02
1.74548570e-02 1.08236104e-01 6.32027388e-01 1.25168371e+00
-2.00838055e-02 -4.10765633e-02 7.75066376e-01 6.37166262e-01
4.88423139e-01 4.16396260e-01 -4.01050895e-01 9.09718752e-01
8.98311809e-02 1.51390207e+00 -7.85497129e-01 -5.68262041e-01
-7.53061652e-01 7.54805803e-01 3.25640291e-01 5.88179268e-02
-9.23997462e-01 -1.31798357e-01 3.36148560e-01 5.27126670e-01
4.12561655e-01 -7.61073316e-03 -1.04376674e-01 -1.29776025e+00
3.47605407e-01 -5.67810655e-01 5.60304880e-01 -3.15826476e-01
-1.34596312e+00 3.94513279e-01 -2.73006801e-02 -1.31601202e+00
2.38736704e-01 -5.98678589e-01 -8.72894466e-01 7.40997016e-01
-1.48242712e+00 -1.65633225e+00 -7.28749633e-01 6.82421267e-01
6.26658916e-01 -6.85418844e-02 3.25462848e-01 6.59411907e-01
-1.00854719e+00 7.79619873e-01 -4.30149347e-01 3.72118324e-01
6.98940158e-01 -9.36486959e-01 2.40993783e-01 1.13289261e+00
-2.54927754e-01 3.78786474e-01 6.64855838e-01 -8.64707947e-01
-8.40218604e-01 -1.18057060e+00 6.05711401e-01 -3.48630607e-01
2.12932155e-02 -3.63765299e-01 -7.05371976e-01 4.35409814e-01
-7.33684823e-02 4.37394768e-01 3.88466001e-01 3.15402411e-02
-2.56382674e-01 -1.97539374e-01 -1.13232362e+00 3.93248588e-01
1.26372969e+00 2.64162868e-01 -7.72292376e-01 2.56997049e-01
2.59945661e-01 -2.87993699e-01 -4.03114885e-01 7.01034188e-01
6.73769057e-01 -1.10899842e+00 1.26738250e+00 -3.82024974e-01
2.04239953e-02 -9.44794655e-01 3.28861140e-02 -7.75182188e-01
-4.38896596e-01 -9.14622843e-02 -3.30849499e-01 1.43407762e+00
-1.15922324e-01 -5.06091058e-01 9.19647396e-01 3.90058190e-01
-1.33770943e-01 -6.41285598e-01 -1.05032587e+00 -7.79279053e-01
-1.97474599e-01 -4.08730470e-03 3.70133340e-01 2.06225276e-01
-9.41419527e-02 4.10008490e-01 -5.95439315e-01 2.71084398e-01
9.57220495e-01 4.22453955e-02 1.30556440e+00 -1.42548370e+00
-1.83420733e-01 -2.80738920e-01 -9.51743901e-01 -1.31793416e+00
-2.17749804e-01 -4.89690691e-01 1.45507812e-01 -1.50286329e+00
5.70935786e-01 -4.57686126e-01 -1.38284728e-01 3.71723324e-01
-6.28492117e-01 2.51986563e-01 2.98920691e-01 1.53590381e-01
-7.72531271e-01 6.32252276e-01 1.04425240e+00 -2.63622105e-01
-1.62922591e-01 3.03127229e-01 -3.61755371e-01 1.06872559e+00
4.23852593e-01 -3.29949319e-01 -1.89180255e-01 -9.37330872e-02
-4.91256058e-01 -2.94873536e-01 7.56284356e-01 -1.57241154e+00
3.82447958e-01 2.44093359e-01 1.03653574e+00 -1.06933451e+00
3.59937489e-01 -6.89160287e-01 -7.71273747e-02 6.87531531e-01
3.33565861e-01 -2.10735932e-01 1.48097038e-01 6.26358688e-01
-6.19375408e-02 -2.41816454e-02 9.31620538e-01 -2.22198769e-01
-7.78475761e-01 3.31132501e-01 -2.05303848e-01 -2.35924095e-01
1.26247561e+00 -3.78213644e-01 -3.20757002e-01 6.69053495e-02
-8.00947249e-01 4.25411135e-01 2.65623033e-01 3.94099355e-01
5.46736777e-01 -1.25214028e+00 -7.65509903e-01 4.82567310e-01
1.35917485e-01 -7.63827702e-03 3.62133682e-01 1.03534055e+00
-1.65185094e-01 4.11690235e-01 -3.08272094e-01 -8.31883490e-01
-1.47804153e+00 8.48064125e-01 3.48779440e-01 -4.26576287e-01
-8.94982696e-01 6.02707446e-01 6.00460589e-01 -1.40133366e-01
3.52796227e-01 -1.90785095e-01 -4.81047571e-01 -4.31950837e-02
6.43049896e-01 3.60184103e-01 -1.70777231e-01 -1.13745844e+00
-4.34409916e-01 8.18814278e-01 -9.90094468e-02 2.77452528e-01
5.98864853e-01 -2.64541477e-01 3.21516186e-01 -1.65432349e-01
7.27772057e-01 -2.32960023e-02 -1.32774448e+00 -3.27169091e-01
-3.79867822e-01 -6.08018100e-01 -2.48566747e-01 -2.54220366e-01
-1.16670060e+00 1.04640770e+00 1.12838650e+00 -1.31982028e-01
8.92363369e-01 -7.78772235e-02 8.96261454e-01 -5.81133179e-02
5.15888453e-01 -9.94283736e-01 1.64436717e-02 7.44247809e-02
4.25881267e-01 -1.53910589e+00 3.00264031e-01 -9.18187201e-01
-2.61252999e-01 8.83357644e-01 1.19713461e+00 -1.27367362e-01
5.92846453e-01 1.20793596e-01 -1.41121462e-01 1.17744301e-02
-2.68895417e-01 -7.56040096e-01 3.62033367e-01 8.52124631e-01
2.22071856e-01 2.63678730e-02 -6.51250958e-01 8.36422086e-01
1.34474441e-01 -2.30128422e-01 -1.01574026e-01 9.12864685e-01
-8.25356662e-01 -1.05039740e+00 -8.49165678e-01 2.74064869e-01
-2.12037638e-01 2.40459278e-01 -8.03855956e-02 9.28822756e-01
8.16682041e-01 1.06947064e+00 -2.89982736e-01 -2.72862762e-01
3.94982457e-01 2.01206505e-02 3.22502017e-01 -4.53414589e-01
-4.09636348e-01 2.92263180e-01 -3.07901837e-02 -4.82168704e-01
-5.37106931e-01 -7.88113832e-01 -9.63543475e-01 -7.14086816e-02
-8.00347745e-01 -1.04980424e-01 2.99593985e-01 9.40894246e-01
-2.06290055e-02 4.44146216e-01 2.47687086e-01 -1.11162126e+00
-4.47158486e-01 -1.07592213e+00 -2.50678837e-01 5.32723308e-01
1.01399608e-01 -1.28890061e+00 -1.36829033e-01 -2.64519360e-02] | [8.019067764282227, -0.6034425497055054] |
f3b0267b-6373-44b3-94a4-7421f49bd54d | accelerating-multiframe-blind-deconvolution | 2306.12078 | null | https://arxiv.org/abs/2306.12078v1 | https://arxiv.org/pdf/2306.12078v1.pdf | Accelerating Multiframe Blind Deconvolution via Deep Learning | Ground-based solar image restoration is a computationally expensive procedure that involves nonlinear optimization techniques. The presence of atmospheric turbulence produces perturbations in individual images that make it necessary to apply blind deconvolution techniques. These techniques rely on the observation of many short exposure frames that are used to simultaneously infer the instantaneous state of the atmosphere and the unperturbed object. We have recently explored the use of machine learning to accelerate this process, with promising results. We build upon this previous work to propose several interesting improvements that lead to better models. As well, we propose a new method to accelerate the restoration based on algorithm unrolling. In this method, the image restoration problem is solved with a gradient descent method that is unrolled and accelerated aided by a few small neural networks. The role of the neural networks is to correct the estimation of the solution at each iterative step. The model is trained to perform the optimization in a small fixed number of steps with a curated dataset. Our findings demonstrate that both methods significantly reduce the restoration time compared to the standard optimization procedure. Furthermore, we showcase that these models can be trained in an unsupervised manner using observed images from three different instruments. Remarkably, they also exhibit robust generalization capabilities when applied to new datasets. To foster further research and collaboration, we openly provide the trained models, along with the corresponding training and evaluation code, as well as the training dataset, to the scientific community. | ['C. Kuckein', 'S. Esteban Pozuelo', 'A. Asensio Ramos'] | 2023-06-21 | null | null | null | null | ['image-restoration'] | ['computer-vision'] | [ 5.68761408e-01 -3.28776002e-01 3.39902312e-01 -2.48540476e-01
-4.16161537e-01 -6.27109051e-01 6.03999496e-01 -9.31006446e-02
-4.75090772e-01 6.71940684e-01 -9.46199149e-02 -4.73525584e-01
3.07630078e-04 -5.11919141e-01 -8.07325780e-01 -1.01211059e+00
2.16777593e-01 2.55323887e-01 6.20373599e-02 -1.77009836e-01
5.80934584e-01 7.11310029e-01 -1.71631014e+00 -2.62830239e-02
1.10505056e+00 9.73787546e-01 4.73705173e-01 9.01486695e-01
1.44747660e-01 8.64890516e-01 -4.99369115e-01 -4.63471822e-02
3.99987489e-01 -4.98055190e-01 -4.91261005e-01 4.88446414e-01
5.38380384e-01 -4.67914522e-01 -7.76278302e-02 1.01573205e+00
5.63852668e-01 4.31776732e-01 4.37817484e-01 -5.10028183e-01
-2.54429400e-01 -1.73002034e-01 -3.10241073e-01 3.23856682e-01
2.63063461e-01 2.72278368e-01 4.86729741e-01 -8.94739091e-01
4.10107613e-01 6.77857935e-01 7.69481897e-01 1.60420403e-01
-1.27653658e+00 -2.91500151e-01 -1.86160833e-01 4.36251402e-01
-9.98086393e-01 -7.29854941e-01 7.79275417e-01 -4.56990361e-01
9.22316134e-01 3.94101948e-01 7.37977207e-01 5.83670616e-01
-1.37672260e-01 4.13209833e-02 1.45616889e+00 -8.78159463e-01
4.53399450e-01 7.34921992e-02 2.84739248e-02 7.01770246e-01
1.32679552e-01 4.75795478e-01 -5.22451818e-01 -7.72373304e-02
5.24409592e-01 -1.92146972e-01 -7.42403030e-01 -2.23866090e-01
-9.55721974e-01 7.13800490e-01 4.78526652e-01 1.61188066e-01
-5.35142124e-01 -1.54093400e-01 1.26527799e-02 2.90898591e-01
6.74518406e-01 5.58188140e-01 -1.82001948e-01 1.78673834e-01
-1.21244252e+00 1.17441177e-01 9.01925981e-01 1.02573246e-01
1.03617013e+00 1.04463592e-01 2.04550818e-01 7.54740596e-01
3.28102797e-01 5.70193946e-01 5.49116850e-01 -1.20503354e+00
-8.72189552e-03 3.86617064e-01 3.44320834e-01 -1.00328386e+00
-3.22179109e-01 -4.88251477e-01 -7.77641773e-01 7.59627044e-01
4.88688290e-01 -1.25655919e-01 -9.34977233e-01 1.27620924e+00
5.25854111e-01 3.66381854e-01 3.44364643e-01 9.38277662e-01
4.98901695e-01 7.09203243e-01 -5.63368022e-01 -6.04862690e-01
9.63391185e-01 -1.16141582e+00 -6.51828945e-01 -2.79644638e-01
2.11693421e-01 -1.10218835e+00 9.17756617e-01 5.49090981e-01
-1.03984237e+00 -4.70049679e-01 -1.29036188e+00 8.27439800e-02
-3.06537211e-01 3.43492270e-01 4.32678074e-01 6.14351809e-01
-1.20736897e+00 1.16642964e+00 -8.90690327e-01 -6.39790952e-01
2.01361582e-01 2.13523433e-01 -1.21326819e-01 -9.78844240e-02
-5.44634223e-01 1.01583862e+00 1.70937181e-01 5.42653203e-01
-8.85080695e-01 -4.38435227e-01 -5.48145592e-01 -2.57386655e-01
2.18689829e-01 -7.49952674e-01 1.15700006e+00 -1.14090693e+00
-1.97488749e+00 8.23563695e-01 -5.32831192e-01 -4.21003819e-01
4.29492146e-01 -1.02166161e-01 -2.03974798e-01 4.23611790e-01
-1.83523551e-01 8.48028529e-03 1.09617376e+00 -1.41270006e+00
-3.17466289e-01 -1.96812496e-01 -1.35109751e-02 1.85622871e-01
-3.26571286e-01 -7.31781824e-03 -3.98125440e-01 -5.01763225e-01
3.73712689e-01 -8.91250789e-01 -2.64891028e-01 5.37061505e-02
-2.20697150e-01 3.72504979e-01 5.39821446e-01 -1.00303805e+00
6.75890803e-01 -2.13254642e+00 1.86400622e-01 2.30633505e-02
-6.87661394e-02 4.60087746e-01 4.06084917e-02 4.14124459e-01
-2.53343105e-01 -2.35760868e-01 -6.88384771e-01 -5.60034513e-01
-3.99154663e-01 1.67128071e-01 -4.34828848e-01 7.58006036e-01
-1.15810424e-01 5.42402625e-01 -6.37585044e-01 1.37924487e-02
4.76054221e-01 4.22897816e-01 -3.77409428e-01 4.99589950e-01
-9.45923254e-02 8.80425274e-01 1.96678601e-02 3.76158744e-01
8.61453712e-01 -2.50180006e-01 1.17737651e-01 -4.55816239e-01
-5.39006174e-01 2.51206696e-01 -1.18789089e+00 1.49744987e+00
-3.36875677e-01 7.03709722e-01 2.84485012e-01 -1.32251585e+00
6.17809355e-01 2.54276901e-01 2.32670739e-01 -6.05949461e-01
1.89242922e-02 4.00350183e-01 -2.71428168e-01 -7.05573201e-01
3.01007867e-01 -4.41233844e-01 8.03152382e-01 5.17097592e-01
-7.67156780e-02 -3.72609317e-01 2.58067071e-01 -3.12210969e-03
7.85843313e-01 2.83976644e-01 3.12065214e-01 -5.19008636e-02
7.18974292e-01 1.73259676e-01 3.67304265e-01 8.83774757e-01
-2.64821667e-02 6.82159185e-01 -9.30828005e-02 -5.70084155e-01
-9.24408615e-01 -6.95519030e-01 -1.92374811e-01 6.04573607e-01
8.44540820e-02 -1.04335889e-01 -8.05619836e-01 -1.76505476e-01
-2.67738193e-01 6.05103493e-01 -3.96091193e-01 -1.63051616e-02
-4.81547505e-01 -1.27106619e+00 5.53022735e-02 1.25671789e-01
5.42433083e-01 -9.43198383e-01 -6.65547788e-01 7.70777613e-02
-3.05066437e-01 -1.21880460e+00 -5.10010263e-03 2.93316483e-01
-1.14357328e+00 -1.16362989e+00 -6.24917150e-01 -5.35122812e-01
8.57055247e-01 5.35671949e-01 9.84287441e-01 3.73348355e-01
-2.97346979e-01 3.72475535e-01 -3.93996507e-01 -2.92595834e-01
-5.00577986e-01 -5.09021342e-01 2.13697255e-01 1.60901636e-01
-2.10700020e-01 -8.40544820e-01 -5.02944946e-01 2.07461476e-01
-8.18781435e-01 9.20756981e-02 3.66257668e-01 7.70849109e-01
6.33997619e-01 1.49541080e-01 -1.12805916e-02 -5.67278445e-01
2.61562645e-01 -4.25090268e-02 -1.14397681e+00 8.05314854e-02
-8.09826195e-01 4.54053320e-02 5.89582980e-01 -1.92497462e-01
-1.22948456e+00 1.47132769e-01 -3.07065602e-02 -2.73255914e-01
-1.83190376e-01 6.64386034e-01 1.47444174e-01 -7.70519972e-01
8.18888545e-01 5.15467584e-01 2.32896432e-01 -6.32676244e-01
2.28415042e-01 5.55970907e-01 8.66919637e-01 -2.31988072e-01
1.13836122e+00 8.44126940e-01 1.50515437e-01 -1.17813706e+00
-8.41402531e-01 -4.50120687e-01 -5.41153491e-01 -4.10194337e-01
6.83069587e-01 -9.68041837e-01 -5.70138574e-01 7.68659413e-01
-1.08011889e+00 -5.85854411e-01 -1.75842270e-01 6.16527021e-01
-3.71134162e-01 8.27040970e-01 -3.67964476e-01 -8.44463646e-01
-2.79857844e-01 -8.70043874e-01 7.65721381e-01 5.53184569e-01
3.21012914e-01 -1.01005435e+00 2.79419422e-01 6.78921342e-01
5.48600554e-01 9.20206234e-02 5.26028812e-01 -1.38003364e-01
-7.72189379e-01 -1.67108461e-01 -1.41708046e-01 7.09570050e-01
1.60053849e-01 2.37032603e-02 -1.27845919e+00 -4.63877439e-01
7.38219202e-01 -1.63073897e-01 1.05358315e+00 5.85069239e-01
1.04559839e+00 -1.24469705e-01 -3.38621326e-02 1.14270425e+00
1.57871902e+00 -2.55957302e-02 7.46521294e-01 6.49386108e-01
4.86282021e-01 5.60816526e-01 3.44265401e-01 2.69398212e-01
7.16545433e-02 5.24763584e-01 7.24773467e-01 -4.05071050e-01
-2.14502290e-01 3.44598144e-01 4.51616645e-01 7.41212666e-01
-6.50476456e-01 -1.39764309e-01 -5.42974949e-01 3.29662472e-01
-1.63936520e+00 -9.40041423e-01 -3.95831585e-01 2.59260392e+00
5.25373638e-01 -2.34875470e-01 -3.11195046e-01 2.10454196e-01
4.20280248e-01 1.97920457e-01 -4.19062227e-01 -1.01203397e-01
-2.95999706e-01 2.51943231e-01 4.93179798e-01 8.99747849e-01
-9.65575755e-01 7.61874199e-01 7.07475042e+00 2.61965364e-01
-1.46965110e+00 6.93283901e-02 3.00190032e-01 -7.89715201e-02
7.23045971e-03 3.43085915e-01 -3.42719465e-01 3.17781419e-01
1.00511837e+00 4.21275906e-02 1.07917547e+00 3.15220475e-01
6.65179729e-01 -5.82525134e-01 -6.43164277e-01 8.84832442e-01
3.43452007e-01 -1.39705050e+00 -3.66874278e-01 9.54756886e-03
8.02409172e-01 2.95752287e-01 -3.79766554e-01 -2.94871807e-01
-6.96275532e-02 -7.28126526e-01 3.52608263e-01 8.26415777e-01
4.57326323e-01 -3.65606278e-01 5.92153311e-01 3.45447004e-01
-6.72686160e-01 -1.24576434e-01 -4.98535007e-01 -5.69608748e-01
2.50043899e-01 9.86145437e-01 -6.62354469e-01 8.91600013e-01
8.42597425e-01 7.38480270e-01 -5.80575883e-01 1.23726106e+00
-5.32552719e-01 7.64637351e-01 -3.78067970e-01 3.73077154e-01
-1.97116181e-01 -6.97271526e-01 7.11916566e-01 6.68235362e-01
5.57235837e-01 8.08091089e-02 -7.25972727e-02 7.11180627e-01
2.49255031e-01 -4.04698253e-02 -3.38243067e-01 -6.72970191e-02
-2.18608856e-01 1.48814797e+00 -6.24693453e-01 -3.91626805e-01
-3.26361954e-01 1.24717283e+00 2.49556229e-01 6.22627795e-01
-4.42845821e-01 -3.54942195e-02 3.95631552e-01 -8.38605836e-02
4.84968692e-01 -3.71893704e-01 -2.31971100e-01 -1.37688696e+00
2.71287769e-01 -8.64193141e-01 9.26573724e-02 -1.22028279e+00
-1.08622074e+00 5.33937633e-01 -4.05789763e-01 -1.06292868e+00
-1.61983103e-01 -7.03780234e-01 -8.81254911e-01 1.19159889e+00
-2.00470996e+00 -6.92198277e-01 -7.94696629e-01 4.33736652e-01
1.56907827e-01 -1.22762308e-01 9.20216322e-01 2.26143986e-01
-7.26254225e-01 -1.71683401e-01 6.19853377e-01 -3.35988045e-01
7.72330225e-01 -1.28844535e+00 1.99201666e-02 1.36941957e+00
1.20727286e-01 3.57354075e-01 1.08136582e+00 -4.83241379e-01
-1.23114896e+00 -7.67487943e-01 9.45199430e-01 -2.53798902e-01
5.51236212e-01 -4.07562703e-02 -9.19430137e-01 4.78644371e-01
4.32486296e-01 2.01474443e-01 6.04444742e-01 -1.67503923e-01
1.91859737e-01 -3.09693575e-01 -9.52683985e-01 2.73888677e-01
5.03136694e-01 -6.57159746e-01 -4.76501793e-01 7.66782463e-01
3.74399304e-01 -5.09761989e-01 -4.45304126e-01 2.24819154e-01
2.78702945e-01 -1.34769690e+00 9.57936227e-01 -1.04870409e-01
4.25207108e-01 -6.85778677e-01 -1.53860658e-01 -1.25652111e+00
-6.82086274e-02 -7.44115591e-01 -2.87604421e-01 9.66175437e-01
2.24151075e-01 -9.21231151e-01 5.37716985e-01 3.11445475e-01
-2.16705605e-01 -3.25340688e-01 -7.41941988e-01 -7.47250259e-01
-3.86095047e-01 -2.46004030e-01 1.49915427e-01 6.87741637e-01
-4.53141391e-01 8.69579017e-02 -4.42628711e-01 7.74554372e-01
8.79736423e-01 4.20279860e-01 8.33802342e-01 -1.17934585e+00
-4.91824895e-01 -6.56224135e-03 -5.94867729e-02 -8.40721667e-01
-2.96126883e-02 -7.53143013e-01 2.98617363e-01 -1.39666963e+00
1.72043145e-02 -6.96040243e-02 -1.58369973e-01 2.59068489e-01
-3.45088035e-01 5.29500306e-01 -3.79663520e-02 6.73006833e-01
2.74347272e-02 4.99616534e-01 1.09956574e+00 5.88195287e-02
-1.98836356e-01 8.78159553e-02 -4.95158404e-01 7.10046768e-01
6.67471528e-01 -3.87415111e-01 -8.38592574e-02 -4.77892369e-01
1.79974258e-01 -1.75220788e-01 8.52893353e-01 -1.28190851e+00
4.03274894e-01 1.46190181e-01 4.32226032e-01 -3.35543454e-01
4.82996970e-01 -6.89407110e-01 8.36922973e-02 4.44427073e-01
1.61129579e-01 -2.95689702e-01 2.02776432e-01 4.96771038e-01
-2.72263139e-01 -5.73178291e-01 9.14823294e-01 -1.94158018e-01
-6.16457641e-01 -7.34152868e-02 -4.13618594e-01 -6.25301719e-01
6.59079969e-01 -1.04050487e-01 -1.12091720e-01 -4.53621656e-01
-7.82871246e-01 -1.88424230e-01 7.56045401e-01 -3.95529211e-01
3.06212276e-01 -6.97400033e-01 -6.22464240e-01 2.47638166e-01
-2.80569315e-01 -3.84856090e-02 2.45526060e-01 1.09682167e+00
-8.64108145e-01 7.52827451e-02 6.09396026e-02 -7.00690627e-01
-1.23584068e+00 5.87290525e-01 7.96428978e-01 -7.12731853e-02
-6.55308783e-01 6.19230866e-01 -2.46927768e-01 -4.58810955e-01
-1.26315862e-01 -1.32336378e-01 -1.91219822e-01 -1.70264747e-02
8.07807088e-01 3.60593647e-01 3.70348573e-01 -5.05120516e-01
-1.13456778e-01 7.65163720e-01 2.58664280e-01 -9.46732536e-02
1.53544056e+00 -3.90591204e-01 -5.80455720e-01 2.64494658e-01
8.70371640e-01 1.63681194e-01 -1.47293854e+00 -1.69583529e-01
-3.99239331e-01 -4.65025127e-01 5.39277971e-01 -9.36975598e-01
-1.03330326e+00 7.25138664e-01 9.50050652e-01 2.52624989e-01
1.61101758e+00 -4.39438611e-01 3.20814490e-01 5.42277396e-01
-7.24251270e-02 -9.94987190e-01 -2.57885884e-02 3.63357097e-01
8.40815365e-01 -1.47211730e+00 3.31021637e-01 -4.48563963e-01
-6.94861216e-03 1.33973634e+00 3.10394969e-02 4.48550507e-02
3.98744822e-01 -5.69279166e-03 4.61422890e-01 -2.15279654e-01
-2.88480282e-01 -2.76393890e-01 1.10029206e-01 4.30584371e-01
1.80909440e-01 -2.73047358e-01 -2.05339432e-01 -3.21201831e-02
-9.04328004e-02 3.00454706e-01 8.26341450e-01 7.99630046e-01
-5.49449086e-01 -1.18103075e+00 -7.88420141e-01 1.15410052e-01
-5.98685965e-02 -2.46677458e-01 -1.52598262e-01 2.72731274e-01
-4.20580944e-03 1.30658066e+00 -2.89675921e-01 5.36809266e-02
1.11063585e-01 7.64942020e-02 4.87913907e-01 -3.26074362e-01
-3.56183857e-01 -3.18712252e-03 -2.16229595e-02 -5.59348404e-01
-8.78894210e-01 -7.83546388e-01 -8.89101505e-01 -1.20534845e-01
-4.70238060e-01 8.65078568e-02 8.67788613e-01 1.35430527e+00
2.57502288e-01 5.41802704e-01 7.93300152e-01 -1.47000921e+00
-4.00049061e-01 -8.79478931e-01 -5.65284491e-01 1.99678749e-01
6.73657358e-01 -4.11847979e-01 -9.12124872e-01 4.08826411e-01] | [11.633167266845703, -2.6131534576416016] |
ce59be92-1fda-48ca-a703-dc7849b0ce61 | feeder-microgrid-management-on-an-active | 2208.10712 | null | https://arxiv.org/abs/2208.10712v1 | https://arxiv.org/pdf/2208.10712v1.pdf | Feeder Microgrid Management on an Active Distribution System during a Severe Outage | Forming a microgrid on a distribution system with large scale outage after a severe weather event is emerging as a viable solution to improve resiliency at the distribution level. This option becomes more attractive when the distribution system has high levels of distributed PV. The management of such feeder-level microgrid has however many challenges, such as limited resources that can be deployed on the feeder quickly, and the limited real-time monitoring and control on the distribution system. Effective use of the distributed PV is also challenging as they are not monitored and controlled. To handle these challenges, the paper proposes a 2-stage hierarchical energy management scheme to securely operate these feeder level micorgrids. The first stage of the scheme solves a sequential rolling optimization problem to optimally schedule the main resources (such as a mobile diesel generator and battery storage unit). The second stage adopts a dispatching scheme for the main resources to adjust the stage-1 set-points closer to real- time. The proposed scheme has unique features to assure that the scheme is robust under highly varying operating conditions with limited system observability: (i) an innovative PV forecast error adjustment and a dynamic reserve adjustment scheme to handle the extreme uncertainty on PV power output, and (ii) an intelligent fuel management scheme to assure that the resources are utilized optimally over the multiple days of the restoration period. The proposed algorithm is tested on sample system with real-time data. The results show that the proposed scheme performs well in maximizing service to loads by effective use of all the resources and by properly taking into account the challenging operating conditions. | ['Wenyuan Tang', 'David Lubkeman', 'Ning Lu', 'Mesut Baran', 'Yiyan Li', 'Bei Xu', 'Victor Paduani', 'Rongxing Hu', 'Ashwin Shirsat', 'Valliappan Muthukaruppan'] | 2022-08-23 | null | null | null | null | ['energy-management'] | ['time-series'] | [-3.38175356e-01 -1.36244908e-01 1.00901134e-01 2.18854249e-01
-6.85248077e-02 -9.55004692e-01 3.67209643e-01 3.29587847e-01
3.22151393e-01 1.14453697e+00 -1.59823909e-01 -1.68127641e-01
-5.30813277e-01 -9.19836819e-01 -1.92472965e-01 -1.24511445e+00
-2.36350894e-01 5.05137622e-01 1.72237188e-01 -2.49091104e-01
2.45781556e-01 8.27615321e-01 -1.42188478e+00 -5.09717643e-01
1.37893760e+00 1.29313493e+00 4.91781384e-01 2.44032696e-01
4.52968329e-01 1.82931960e-01 -9.32368755e-01 5.93443871e-01
4.85182554e-01 -2.07123071e-01 -1.34999976e-02 2.25098938e-01
-8.85332584e-01 -4.98369783e-01 7.33635277e-02 1.11475384e+00
7.26537049e-01 4.85788584e-01 5.83754122e-01 -1.69826639e+00
-1.04912147e-01 1.53673559e-01 -7.23545372e-01 3.59471589e-01
9.78317782e-02 2.96367258e-01 4.72141236e-01 -5.53220928e-01
2.26970557e-02 7.31823146e-01 5.87124676e-02 -2.72456229e-01
-1.02876425e+00 -3.30205709e-01 1.88974217e-01 5.96889377e-01
-1.50893903e+00 -6.38029352e-02 6.15960300e-01 -3.49477768e-01
1.18060994e+00 3.07570249e-01 9.60234642e-01 -1.42149702e-01
4.63805616e-01 -6.72605773e-03 1.21611059e+00 -3.97770703e-01
5.66090643e-01 2.56687343e-01 -6.18515015e-02 -2.06527606e-01
5.11605501e-01 -2.48863384e-01 1.57089934e-01 -6.49887472e-02
8.96047279e-02 -1.85717538e-01 -7.38281012e-01 6.64259889e-04
-4.85324532e-01 6.09599888e-01 3.41765970e-01 5.78217208e-01
-6.80247009e-01 -4.92034316e-01 2.12721035e-01 1.28776863e-01
3.00802916e-01 3.96097377e-02 -4.46982354e-01 1.85325593e-01
-9.13252175e-01 -5.99610396e-02 9.08176601e-01 9.61639285e-01
2.51917392e-01 9.01395261e-01 1.27405592e-03 4.41637456e-01
2.37459719e-01 6.64694905e-01 5.18084109e-01 -4.26638156e-01
2.87896693e-01 5.12183607e-01 5.26101947e-01 -5.87509215e-01
-4.08535331e-01 -6.83900893e-01 -8.56974304e-01 6.87718689e-01
8.85032937e-02 -4.92354095e-01 -5.28144777e-01 1.41269803e+00
5.09185791e-01 -2.84102529e-01 1.16170801e-01 5.62025368e-01
-9.78264064e-02 1.23003161e+00 -2.31530175e-01 -1.08691609e+00
1.49432898e+00 -4.29607391e-01 -1.03071833e+00 2.58162349e-01
4.23395298e-02 -7.77008951e-01 2.22207978e-01 3.37088197e-01
-1.34142113e+00 -2.24797815e-01 -1.41282094e+00 6.57560408e-01
-6.04925275e-01 2.54287571e-01 -2.87017137e-01 3.41123730e-01
-1.06047523e+00 4.28184181e-01 -5.72548687e-01 -3.40333670e-01
-5.91416508e-02 4.07557786e-01 5.38933389e-02 2.63771653e-01
-1.24787831e+00 1.43883967e+00 8.08219612e-01 7.27182150e-01
-5.88506520e-01 -6.23488188e-01 -6.81385338e-01 5.70790052e-01
3.61752301e-01 -3.47239941e-01 9.96270657e-01 -7.10859358e-01
-1.58474338e+00 -2.36127391e-01 1.09857835e-01 -4.61768836e-01
4.81785029e-01 3.97420555e-01 -5.01419485e-01 1.84446767e-01
-1.76487520e-01 -3.60031813e-01 4.18786764e-01 -8.58013034e-01
-8.74611139e-01 -3.24086010e-01 -4.33808595e-01 6.54643118e-01
-4.01291937e-01 -2.07107976e-01 5.42269766e-01 -4.65367675e-01
-2.51960754e-01 -7.46197343e-01 -2.20131040e-01 -5.02218008e-01
-2.26646066e-01 -4.86130476e-01 1.18389058e+00 -9.97424722e-01
1.11897838e+00 -1.73200607e+00 1.47208244e-01 5.40139079e-01
-5.88330746e-01 3.32622617e-01 5.63866973e-01 1.06211460e+00
-2.67324835e-01 -2.98369914e-01 -1.90592304e-01 2.50818312e-01
1.24974228e-01 4.68817234e-01 -9.66992676e-02 4.53505993e-01
-7.36232027e-02 8.68100971e-02 -5.54533839e-01 1.85279489e-01
3.94577563e-01 2.65992433e-01 1.72832012e-01 4.31427836e-01
-1.84964612e-01 3.47099513e-01 -2.68490791e-01 5.91686845e-01
9.58927631e-01 1.26639754e-01 2.59504676e-01 -2.79344678e-01
-4.88979846e-01 -3.07878375e-01 -1.72507715e+00 8.18286717e-01
-4.02878702e-01 3.21314275e-01 6.47377193e-01 -1.39366591e+00
8.23797703e-01 5.45137525e-01 6.97881222e-01 -6.95421517e-01
-9.62838978e-02 4.62481201e-01 1.40556738e-01 -4.59711373e-01
1.63675696e-01 -7.53917322e-02 2.96137869e-01 3.68956417e-01
-7.44735748e-02 -2.36248121e-01 6.36246502e-01 -1.90452915e-02
5.37486136e-01 -9.58185419e-02 6.51052713e-01 -8.06350589e-01
9.83968675e-01 -1.60982892e-01 1.16311753e+00 -3.30759645e-01
1.18127324e-01 -3.52093071e-01 2.95777142e-01 1.43567577e-01
-7.58112550e-01 -5.63920200e-01 -2.54336655e-01 1.36170089e-01
1.93966925e-01 3.74443024e-01 -3.03005338e-01 -1.60801932e-01
2.50413209e-01 1.35368240e+00 -1.49325477e-02 1.15493923e-01
-4.94354457e-01 -1.13434887e+00 -4.24216807e-01 2.19622493e-01
6.21471226e-01 -6.47778690e-01 -7.67736554e-01 4.25188422e-01
2.89592534e-01 -6.77582085e-01 -3.85496348e-01 5.19794226e-01
-4.89520937e-01 -1.07708287e+00 -6.66873336e-01 -7.11300671e-01
1.02381849e+00 3.25653851e-02 6.34235561e-01 9.81865749e-02
-1.26049206e-01 1.44645199e-01 -3.21323834e-02 -3.80390674e-01
-1.90967515e-01 -2.41536871e-01 3.06263685e-01 -2.04676241e-01
-4.03147906e-01 -7.74536788e-01 -7.27695942e-01 4.11232710e-01
-8.57309520e-01 -1.06794618e-01 2.11297125e-01 8.47502589e-01
4.67207164e-01 1.24629760e+00 1.57823074e+00 -2.60255933e-01
6.72992110e-01 -8.45073700e-01 -1.38327503e+00 4.15371120e-01
-1.12832141e+00 -3.83253515e-01 1.11838377e+00 -1.49329051e-01
-1.28657699e+00 -1.59271002e-01 9.36011150e-02 1.43351525e-01
1.41993940e-01 4.12580132e-01 -6.37273073e-01 -2.23647103e-01
-3.03923190e-01 2.52058834e-01 1.29262045e-01 -4.71326739e-01
-3.30365933e-02 6.47746623e-01 2.86364794e-01 -2.51987904e-01
1.28619671e+00 -5.35640344e-02 6.07374489e-01 -4.80713665e-01
-6.37764717e-03 -2.75228202e-01 -4.28431243e-01 -3.87598664e-01
3.46280634e-01 -9.05512750e-01 -1.19014978e+00 6.22440815e-01
-6.97835505e-01 -6.79524168e-02 -1.26201019e-01 3.57077450e-01
-2.33766615e-01 5.63243628e-01 -3.39908332e-01 -1.08344781e+00
-7.49858677e-01 -1.09161663e+00 -2.83567272e-02 9.52462733e-01
4.06097114e-01 -1.16142523e+00 -2.43608654e-01 -1.36016726e-01
6.70618713e-01 8.47691715e-01 9.85605299e-01 -3.65914196e-01
-6.86857760e-01 7.76660964e-02 2.67842472e-01 7.08120167e-01
5.24836600e-01 6.34519160e-02 -3.52785438e-01 -1.03856122e+00
2.24172398e-01 1.84983313e-01 -2.91907042e-01 2.77693957e-01
7.22075820e-01 -6.37320578e-01 -2.34633267e-01 3.38745594e-01
2.19672036e+00 9.70012188e-01 2.63144702e-01 4.76234376e-01
-1.37717901e-02 5.36115468e-01 1.01423073e+00 7.32299149e-01
4.33723003e-01 7.15999722e-01 7.90224791e-01 2.86743101e-02
3.77155006e-01 2.73080647e-01 4.23212439e-01 8.29219341e-01
2.24502802e-01 -5.58622539e-01 -3.35715920e-01 7.25374520e-01
-1.85469520e+00 -8.77781570e-01 -7.12930113e-02 2.42728019e+00
5.71890354e-01 -4.24120538e-02 -1.73688214e-02 4.62775618e-01
7.41587222e-01 -1.90200374e-01 -7.19795346e-01 -6.25246346e-01
-3.59920293e-01 -1.01968557e-01 5.21975160e-01 3.61974031e-01
-2.96688855e-01 -2.63889670e-01 4.38614178e+00 6.76752746e-01
-1.00840330e+00 3.18749668e-03 5.62024713e-01 5.37724271e-02
-6.78750053e-02 1.22421481e-01 -6.40411079e-01 1.21612680e+00
1.04186106e+00 -9.08453107e-01 6.42709017e-01 6.14614606e-01
9.87000465e-01 -8.54782045e-01 -7.14349449e-01 2.33784705e-01
-1.11254640e-01 -8.14508319e-01 -4.70327944e-01 1.14720166e-01
9.73638833e-01 -1.26029477e-01 -6.53463423e-01 -9.12958756e-02
-6.86666071e-02 -4.67333972e-01 7.35005498e-01 5.29735386e-01
2.48760507e-01 -1.25023913e+00 8.71396422e-01 6.41380191e-01
-1.37881613e+00 -6.34306431e-01 -1.91298947e-01 -1.33923486e-01
8.72022092e-01 7.87843049e-01 -7.68569529e-01 1.14679658e+00
7.42397130e-01 1.34640351e-01 -1.34366497e-01 1.20088899e+00
-3.95054132e-01 3.07509631e-01 -6.58165872e-01 1.03256926e-01
-1.29210278e-01 -6.95631802e-01 6.05515063e-01 5.41246116e-01
7.69167006e-01 4.54520732e-01 4.32940423e-01 4.19831485e-01
4.32156950e-01 -3.51774506e-02 -3.91075015e-01 4.23799694e-01
8.02851915e-01 1.63540661e+00 -6.77642047e-01 -4.27794158e-01
-2.00527370e-01 4.42048222e-01 -2.19457135e-01 5.55326164e-01
-5.35440087e-01 -5.19893467e-01 5.20067751e-01 1.71787012e-02
1.18730284e-01 -2.36688808e-01 -3.02463919e-01 -7.48068333e-01
3.49205434e-01 -3.49858463e-01 7.59230673e-01 -9.31161821e-01
-1.25300527e+00 3.78277987e-01 1.77708432e-01 -1.33019435e+00
-7.24046350e-01 -1.86908379e-01 -1.15537441e+00 1.41486812e+00
-1.97984171e+00 -8.28100324e-01 -1.82104364e-01 5.88051617e-01
6.56527996e-01 -2.05122232e-01 5.92962265e-01 3.67597103e-01
-1.04868555e+00 8.44500680e-03 7.75728047e-01 -7.03619301e-01
4.76661734e-02 -1.38643408e+00 -8.33605587e-01 1.52334189e+00
-9.07817125e-01 -2.78103352e-02 7.50599444e-01 -6.23777747e-01
-1.62533700e+00 -6.16858006e-01 4.77450907e-01 6.36071324e-01
6.97976708e-01 -1.95349827e-02 -9.19351459e-01 3.76722008e-01
1.14996779e+00 -3.63384932e-01 3.91890794e-01 -8.53915453e-01
7.45317161e-01 -5.42171717e-01 -1.71525609e+00 1.19001053e-01
-1.89350188e-01 1.71206817e-01 -3.59566450e-01 6.08801305e-01
4.50208448e-02 -4.62790728e-01 -1.31572068e+00 4.27873462e-01
7.43853599e-02 -2.98152924e-01 4.40979779e-01 1.58937603e-01
-6.79823875e-01 -9.99947965e-01 7.48351812e-02 -2.11235976e+00
-1.32224187e-01 -9.97760952e-01 -2.86989540e-01 1.68744922e+00
7.87218884e-02 -1.37964022e+00 4.70983237e-02 7.05462277e-01
-2.66353428e-01 -7.50247777e-01 -1.36073470e+00 -6.71128511e-01
-2.40031645e-01 6.56158864e-01 8.01417291e-01 8.17008257e-01
3.92569989e-01 -3.80420424e-02 6.34292662e-02 9.30870831e-01
8.35194230e-01 3.24270695e-01 2.36284196e-01 -9.33656156e-01
-9.76279825e-02 -1.40750840e-01 -2.17359271e-02 -7.74874240e-02
-1.24136001e-01 -4.33940977e-01 -7.00868964e-02 -2.00074601e+00
-3.31962556e-01 -3.09728831e-01 -3.16318780e-01 5.15632987e-01
-2.39587482e-02 -2.90214390e-01 3.21872830e-01 2.35179648e-01
5.20418227e-01 7.62867272e-01 7.70516992e-01 -8.70820042e-03
-1.29229233e-01 4.89165545e-01 -1.67672515e-01 4.33252543e-01
1.06097543e+00 -4.75265384e-02 -8.57906878e-01 -1.15139455e-01
-2.60922581e-01 6.23823285e-01 -1.83500767e-01 -8.76155376e-01
3.38354170e-01 -3.98751348e-01 4.22966033e-01 -9.64277685e-01
-4.09552269e-03 -1.63095534e+00 9.01077390e-01 9.41866517e-01
6.26455367e-01 6.26483083e-01 1.41387001e-01 4.46277827e-01
-1.10179044e-01 -3.66405368e-01 9.16538596e-01 1.92580193e-01
-6.73225820e-01 -1.29930303e-01 -5.45299113e-01 -6.70740306e-01
1.85679173e+00 -1.78624645e-01 -7.24263430e-01 -3.85097116e-01
-6.90086246e-01 1.22789776e+00 3.71643543e-01 2.69217223e-01
-6.19409941e-02 -1.09929311e+00 -5.31896174e-01 2.23024306e-03
-5.09514272e-01 -8.40706825e-02 3.71161461e-01 7.57497370e-01
-4.89852607e-01 3.57530057e-01 -4.02851105e-01 -4.21935439e-01
-9.01081800e-01 4.69062835e-01 5.20367742e-01 -2.97415942e-01
-3.21308821e-01 -8.59449729e-02 -4.25473452e-01 3.80350083e-01
-7.17506632e-02 -2.23472208e-01 -3.87647718e-01 6.40574455e-01
3.45194131e-01 7.67921746e-01 4.50482816e-01 -4.59079742e-01
-3.03576440e-01 3.27267855e-01 4.91502881e-01 9.92663652e-02
1.53026545e+00 -5.63662112e-01 -4.85837370e-01 1.03045486e-01
6.09079719e-01 -5.86570725e-02 -1.30005717e+00 2.43175834e-01
-1.09757453e-01 -3.08664083e-01 8.92214552e-02 -1.13440037e+00
-1.25089872e+00 1.68375388e-01 4.23461080e-01 8.18995416e-01
1.74142170e+00 -1.04112756e+00 4.65718091e-01 -1.92153618e-01
4.87274379e-01 -1.50964022e+00 -5.60872078e-01 6.58136383e-02
8.89992297e-01 -5.34888685e-01 3.58381033e-01 -2.27587312e-01
-3.62711191e-01 1.28483176e+00 5.15666485e-01 -4.48464043e-02
7.79941142e-01 3.83979112e-01 -4.99531627e-01 2.86546290e-01
-7.45619833e-01 8.95315856e-02 -1.78651050e-01 3.29583704e-01
-1.82898208e-01 2.78727800e-01 -9.60514545e-01 3.35869163e-01
2.09615126e-01 -1.21891975e-01 1.02258277e+00 1.03186727e+00
-4.42671895e-01 -8.48890901e-01 -6.92295492e-01 4.19686109e-01
-5.37109137e-01 5.23219109e-01 8.55870128e-01 7.73013949e-01
3.27779919e-01 1.17404306e+00 -5.18114008e-02 5.57603419e-01
6.48445487e-01 -1.31425351e-01 -6.76239654e-02 -3.18308562e-01
-5.93568265e-01 2.24768519e-01 9.88695920e-02 1.21215381e-01
-5.61156422e-02 -8.36405456e-01 -1.47469199e+00 -1.38224334e-01
-6.21810973e-01 8.26403201e-01 1.08384895e+00 9.76067841e-01
1.49671957e-01 5.21241605e-01 1.47862649e+00 -9.04405773e-01
-1.00973535e+00 -1.00159848e+00 -1.06483102e+00 -3.13536584e-01
3.27949114e-02 -6.42428100e-01 -8.45370412e-01 -3.07438940e-01] | [5.6724162101745605, 2.540416717529297] |
e764c025-058f-49cf-85e7-9eeea829ea33 | adversarial-learning-for-discourse-rhetorical | null | null | https://aclanthology.org/2021.acl-long.305 | https://aclanthology.org/2021.acl-long.305.pdf | Adversarial Learning for Discourse Rhetorical Structure Parsing | Text-level discourse rhetorical structure (DRS) parsing is known to be challenging due to the notorious lack of training data. Although recent top-down DRS parsers can better leverage global document context and have achieved certain success, the performance is still far from perfect. To our knowledge, all previous DRS parsers make local decisions for either bottom-up node composition or top-down split point ranking at each time step, and largely ignore DRS parsing from the global view point. Obviously, it is not sufficient to build an entire DRS tree only through these local decisions. In this work, we present our insight on evaluating the pros and cons of the entire DRS tree for global optimization. Specifically, based on recent well-performing top-down frameworks, we introduce a novel method to transform both gold standard and predicted constituency trees into tree diagrams with two color channels. After that, we learn an adversarial bot between gold and fake tree diagrams to estimate the generated DRS trees from a global perspective. We perform experiments on both RST-DT and CDTB corpora and use the original Parseval for performance evaluation. The experimental results show that our parser can substantially improve the performance when compared with previous state-of-the-art parsers. | ['Guodong Zhou', 'Fang Kong', 'Longyin Zhang'] | 2021-08-01 | null | null | null | acl-2021-5 | ['drs-parsing'] | ['natural-language-processing'] | [ 3.36255819e-01 5.89503944e-01 -2.40139529e-01 -4.07390088e-01
-1.48820436e+00 -9.92497683e-01 5.70389569e-01 2.01202363e-01
-7.37342089e-02 4.85303849e-01 5.73454738e-01 -8.69573832e-01
5.02273917e-01 -8.40542734e-01 -6.40697122e-01 -4.82540756e-01
2.12051123e-01 6.18771851e-01 3.77610773e-01 -5.99826634e-01
3.02391887e-01 1.39757678e-01 -8.47741246e-01 7.36281753e-01
1.03044498e+00 4.55736190e-01 1.17365897e-01 8.62315416e-01
-4.47201669e-01 1.29023194e+00 -1.12635660e+00 -9.54384625e-01
2.82366239e-02 -3.14761341e-01 -1.23859859e+00 -3.07779402e-01
5.82558572e-01 -4.55604315e-01 -3.76421236e-03 1.16057718e+00
4.42605138e-01 -3.13136131e-01 3.02262694e-01 -6.08877480e-01
-7.22874343e-01 1.27526486e+00 -7.24615455e-01 1.87094554e-01
5.93057692e-01 -9.01028048e-03 1.72620308e+00 -5.66557407e-01
1.05050778e+00 1.65176678e+00 4.81070817e-01 5.73263466e-01
-1.11460412e+00 -3.40451658e-01 6.28461123e-01 2.41077945e-01
-6.87612593e-01 -2.98083007e-01 1.12934279e+00 -4.41108346e-01
1.04915011e+00 4.06148911e-01 1.56540379e-01 1.36030817e+00
1.24989249e-01 1.08787835e+00 1.25287950e+00 -6.59235120e-01
-7.22102262e-03 -2.61838078e-01 4.01472449e-01 9.10706758e-01
-6.26872033e-02 -1.26237288e-01 -3.43763113e-01 -8.24993551e-02
9.47067142e-02 -8.11676562e-01 -6.66270927e-02 -2.71550175e-02
-9.64889467e-01 1.16447973e+00 5.19010127e-01 5.89634001e-01
8.25836211e-02 -2.10890546e-02 4.97922152e-01 2.17641741e-01
4.92917895e-01 5.61442196e-01 -2.53653884e-01 -2.81802297e-01
-7.03859985e-01 3.29591572e-01 1.02632725e+00 6.55420423e-01
3.90157938e-01 -1.50516137e-01 -4.55169290e-01 6.08475149e-01
5.20965457e-02 2.26545900e-01 1.56927288e-01 -1.16169715e+00
1.09633017e+00 5.67764401e-01 -9.89744663e-02 -1.41867149e+00
-3.69685441e-01 -3.34803224e-01 -4.48132515e-01 9.58123356e-02
7.75053859e-01 -1.24062233e-01 -6.41128600e-01 1.85470963e+00
5.12369692e-01 -4.33242291e-01 1.82328045e-01 7.26221800e-01
1.00981438e+00 7.35030770e-01 1.36422545e-01 -3.44373398e-02
1.56674516e+00 -1.23155427e+00 -8.62627983e-01 -6.07122540e-01
1.12981761e+00 -7.74816513e-01 1.27510583e+00 3.34120691e-01
-1.13220048e+00 -2.59507686e-01 -1.09231603e+00 -5.24249315e-01
-4.89123017e-02 2.13735476e-01 4.34855968e-01 6.99475110e-01
-8.21276963e-01 5.74289978e-01 -1.01138544e+00 1.18717328e-02
2.58631676e-01 -1.03760697e-01 -1.68029249e-01 4.99630161e-02
-1.25607657e+00 1.21368694e+00 2.68979907e-01 2.18046367e-01
-5.41422546e-01 -3.27078760e-01 -8.97276223e-01 -3.31456997e-02
6.70913458e-01 -4.53048378e-01 1.66703057e+00 -7.10372031e-01
-1.81606245e+00 1.00914145e+00 -3.15233022e-01 -5.03661633e-01
5.55743933e-01 -4.35497016e-01 1.41006932e-01 8.70135203e-02
2.96398759e-01 2.88825840e-01 4.44378614e-01 -1.33378041e+00
-5.91586173e-01 -4.29061025e-01 5.29644132e-01 3.73384327e-01
-2.34920178e-02 3.69862437e-01 -2.99177617e-01 -8.14990103e-01
1.94998473e-01 -7.42650986e-01 -1.99684232e-01 -4.63276118e-01
-9.26946402e-01 -4.22585666e-01 7.31398523e-01 -1.21832263e+00
1.56790948e+00 -1.76576877e+00 5.08565366e-01 -2.30047703e-01
2.71505415e-01 2.84383029e-01 2.71197930e-02 2.76403785e-01
1.66987792e-01 4.50250268e-01 -4.89149123e-01 -6.32703781e-01
-9.34158918e-03 2.64540404e-01 -3.36197466e-01 1.42261803e-01
2.57955104e-01 9.73886311e-01 -1.02387547e+00 -7.47663498e-01
1.30671844e-01 2.22749524e-02 -6.39168561e-01 2.78241634e-01
-4.65148419e-01 4.66595203e-01 -6.64893508e-01 6.68977201e-01
5.47119498e-01 -1.62586331e-01 8.79199862e-01 -4.56744246e-02
1.82847884e-02 1.01998961e+00 -8.03813815e-01 1.69470596e+00
-4.21753526e-01 6.43533945e-01 2.88138270e-01 -1.29719722e+00
8.17836821e-01 -5.06604426e-02 -8.65417346e-03 -6.61853194e-01
1.39803365e-01 2.69970477e-01 1.57748550e-01 -3.26661974e-01
7.22516954e-01 -1.62824556e-01 -6.63022220e-01 3.98442656e-01
-2.15390250e-01 -1.78857148e-01 1.22319497e-01 4.51797962e-01
1.46275520e+00 3.30587715e-01 4.19677317e-01 2.51911301e-02
6.13166869e-01 4.25124645e-01 5.55407166e-01 7.54020452e-01
-1.10775694e-01 6.41636491e-01 1.03997707e+00 -2.93260366e-01
-8.28286707e-01 -9.62681234e-01 2.12795481e-01 1.33122492e+00
2.68140826e-02 -7.03545213e-01 -1.15849757e+00 -1.22871602e+00
-4.25179720e-01 1.14380050e+00 -5.01264453e-01 2.70594448e-01
-1.56688917e+00 -6.96668565e-01 8.09928656e-01 4.48229879e-01
3.90344679e-01 -1.11595607e+00 -4.70785558e-01 4.19696897e-01
-4.41592395e-01 -1.51825929e+00 -1.12122260e-01 -2.67856149e-03
-5.70582390e-01 -1.05957413e+00 -1.75233960e-01 -8.37294459e-01
3.27907622e-01 5.03117824e-03 1.32332027e+00 2.03167766e-01
1.67796969e-01 -2.62335718e-01 -5.42372704e-01 -1.86479315e-01
-1.17076695e+00 4.07594323e-01 -6.71681643e-01 -5.18205225e-01
9.88611355e-02 -2.69051671e-01 -2.66411096e-01 -2.85131596e-02
-3.51304859e-01 3.60214680e-01 2.48306632e-01 8.31530154e-01
3.36883366e-01 -2.62469172e-01 1.11036032e-01 -1.48878777e+00
7.52060235e-01 -1.55231535e-01 -5.38686216e-01 3.99932921e-01
-1.42703727e-01 2.38131732e-01 8.28965902e-01 1.96037263e-01
-1.34614372e+00 -2.63214201e-01 -5.74730814e-01 2.17893168e-01
7.11617172e-02 5.69546223e-01 -2.77872175e-01 4.94647682e-01
7.98652828e-01 -3.58998597e-01 -3.86658192e-01 -6.04281723e-01
7.67131329e-01 5.87612391e-01 6.44588828e-01 -9.90879118e-01
8.02910626e-01 2.90790349e-01 -2.97735989e-01 -3.09781611e-01
-1.37017822e+00 -7.45107010e-02 -5.89747429e-01 7.04393014e-02
1.11883831e+00 -6.55179799e-01 -4.05947953e-01 3.53959382e-01
-1.70041919e+00 -3.62890482e-01 -1.85432702e-01 -3.13948900e-01
-3.45069498e-01 7.84059465e-01 -8.55864644e-01 -6.61491096e-01
-5.27412653e-01 -1.41420674e+00 1.24319351e+00 -1.39018759e-01
-3.95656079e-01 -9.69478190e-01 1.29727662e-01 8.25877726e-01
1.89784262e-02 6.35191560e-01 1.21868742e+00 -6.34966314e-01
-5.04068971e-01 1.33922279e-01 -1.26681075e-01 1.98638350e-01
5.62550686e-02 1.55805409e-01 -9.12121773e-01 1.98609028e-02
7.49010518e-02 -3.20269704e-01 7.56593108e-01 2.09712490e-01
9.17943001e-01 -5.32103598e-01 -3.59953731e-01 3.00237507e-01
1.07021558e+00 -7.54300579e-02 5.22403419e-01 7.24682748e-01
9.81640577e-01 8.34286392e-01 8.53430390e-01 -1.62417570e-03
7.83097565e-01 9.27324474e-01 4.77651834e-01 2.47569811e-02
-2.93045759e-01 -4.98804450e-01 6.44903362e-01 1.04893196e+00
1.97334349e-01 -5.13342381e-01 -1.06957364e+00 4.27093327e-01
-1.81918061e+00 -9.09736276e-01 -2.23819822e-01 1.60321069e+00
1.04700327e+00 5.15984654e-01 -1.49141625e-01 6.50172308e-02
7.98022330e-01 7.46435523e-01 -2.26852402e-01 -9.06089485e-01
-2.49458209e-01 3.75984877e-01 1.18343882e-01 8.21904600e-01
-1.18682790e+00 1.56960714e+00 6.24301910e+00 9.25072670e-01
-9.24976408e-01 2.33621672e-01 7.74194717e-01 9.57358778e-02
-5.00163138e-01 4.09793288e-01 -9.27369773e-01 1.98388562e-01
7.71310329e-01 2.54953086e-01 1.73457265e-01 1.02562785e+00
-2.38455497e-02 1.60022065e-01 -8.79092991e-01 4.39938247e-01
-1.78934291e-01 -1.55003560e+00 -9.12093371e-02 -1.64318323e-01
4.55771208e-01 -3.67958695e-02 -2.04683080e-01 4.89343554e-01
1.00281417e+00 -1.08076417e+00 9.77103353e-01 -1.26131743e-01
6.15459085e-01 -4.56320882e-01 6.21839643e-01 6.03610873e-01
-1.10418034e+00 -7.40726143e-02 -5.03901914e-02 -2.97691431e-02
4.51736927e-01 4.06468183e-01 -9.54558969e-01 6.45893812e-01
5.81124604e-01 5.59110701e-01 -5.88170409e-01 4.22798395e-02
-9.56278503e-01 8.13654244e-01 -3.12961906e-01 -1.35645196e-01
4.03018326e-01 -7.51137733e-02 8.40799034e-01 1.24730551e+00
1.34402946e-01 1.62903920e-01 2.50605613e-01 7.22069085e-01
-3.18776041e-01 1.74208909e-01 -3.54323298e-01 1.21024527e-01
6.48372710e-01 1.29455340e+00 -4.26987201e-01 -2.68652737e-01
-2.47678593e-01 7.26098061e-01 1.01861203e+00 -4.21384312e-02
-8.35477650e-01 9.47602689e-02 5.01685977e-01 -2.87873717e-03
3.66242945e-01 -4.06178504e-01 -6.84634387e-01 -1.21790302e+00
1.95506364e-01 -1.29481518e+00 6.38604224e-01 -5.35128355e-01
-1.10222673e+00 6.31155789e-01 -8.00171196e-02 -7.86595404e-01
-4.04217422e-01 -7.32722640e-01 -7.93250084e-01 5.45277655e-01
-1.51853299e+00 -1.29763460e+00 9.55231488e-02 -2.06629932e-03
7.16296673e-01 2.29202230e-02 9.31362689e-01 -2.00701118e-01
-7.87497342e-01 7.87276804e-01 2.10033692e-02 4.66686785e-01
5.14814854e-01 -1.61162603e+00 9.35232341e-01 1.16466498e+00
2.68682718e-01 5.22284627e-01 8.81948531e-01 -6.41203761e-01
-1.25897360e+00 -6.30222201e-01 9.75822330e-01 -7.68184125e-01
9.63541090e-01 -5.37693083e-01 -9.70768631e-01 7.17417121e-01
3.15713823e-01 -3.47146899e-01 2.51572818e-01 3.92573297e-01
-4.30443168e-01 3.18190187e-01 -9.24975395e-01 8.51064026e-01
1.13436556e+00 -3.85588676e-01 -9.26215172e-01 3.76977623e-01
9.24879849e-01 -1.07102692e+00 -8.53074729e-01 3.19747090e-01
3.35380882e-01 -1.09568036e+00 8.54006350e-01 -6.11123979e-01
9.87699747e-01 -9.22379196e-02 -2.96357125e-01 -1.03764975e+00
-1.04031503e-01 -8.44472349e-01 -2.27879539e-01 1.41110218e+00
5.52306771e-01 -4.24444795e-01 9.83591437e-01 4.34155017e-01
-4.18517262e-01 -8.54203761e-01 -1.19820893e+00 -3.22579175e-01
5.73016047e-01 -4.11595553e-01 5.26928484e-01 7.65112162e-01
-1.27693042e-02 7.84321070e-01 -2.47328922e-01 3.78693976e-02
3.96450460e-01 5.11641204e-01 9.78420615e-01 -8.35627973e-01
-6.11041784e-01 -6.06279492e-01 1.14974715e-01 -1.28608823e+00
3.99113387e-01 -1.05204952e+00 1.29942968e-01 -1.78480065e+00
-5.65022184e-03 -5.43122768e-01 1.64180532e-01 4.00555432e-01
-5.73498905e-01 -2.88702518e-01 5.13355255e-01 1.21837631e-01
-4.21463400e-01 3.50220263e-01 1.37019300e+00 -2.60195285e-01
-9.99796167e-02 -2.43545651e-01 -1.05383444e+00 8.97615850e-01
8.57758641e-01 -6.75398111e-01 -2.39927396e-02 -8.14738750e-01
9.51920971e-02 1.86677203e-01 1.35630161e-01 -4.18647915e-01
-1.17059119e-01 -1.69376969e-01 -2.18743488e-01 -6.29286051e-01
2.85988688e-01 -1.37155741e-01 -4.76780623e-01 3.62240613e-01
-4.40986037e-01 1.06838770e-01 -5.87694310e-02 4.33280647e-01
-2.09104955e-01 -4.68312055e-01 6.15490973e-01 -2.78649747e-01
-5.11756957e-01 -3.20905507e-01 -2.10860670e-01 4.98722494e-01
7.23867059e-01 -9.53823254e-02 -8.78442287e-01 -2.97987014e-01
-5.33001363e-01 1.21228635e-01 4.11205471e-01 3.11614156e-01
1.18581913e-01 -9.41035986e-01 -8.10101807e-01 -4.24310207e-01
-2.39914566e-01 3.56180727e-01 2.75240862e-03 4.45797503e-01
-7.14094877e-01 2.29861677e-01 2.30978146e-01 -5.94404399e-01
-1.57334733e+00 3.30377817e-01 1.65931970e-01 -1.28464925e+00
-7.39758372e-01 9.73735452e-01 2.65890330e-01 -6.87827408e-01
-7.68535435e-02 -3.24658394e-01 -4.54139501e-01 -5.37852868e-02
2.15070784e-01 3.02644167e-02 1.03887215e-01 -8.02739620e-01
-3.04297656e-01 6.12003148e-01 -2.12894008e-01 -1.78388119e-01
1.26581490e+00 2.19204817e-02 -1.71785846e-01 5.77760749e-02
1.03489530e+00 4.72055554e-01 -9.38315094e-01 -2.51305670e-01
4.07657087e-01 -2.54568994e-01 -1.75454587e-01 -6.95662618e-01
-7.58024395e-01 1.17856467e+00 -8.04202929e-02 3.98346364e-01
8.53841722e-01 9.65868458e-02 1.03905451e+00 3.48388195e-01
1.44951567e-01 -1.15737486e+00 1.75105169e-01 8.27790976e-01
9.57708836e-01 -1.12886727e+00 1.57326907e-01 -7.50309885e-01
-8.66027415e-01 1.14755809e+00 5.40698886e-01 -1.79459989e-01
1.15269646e-01 4.10029650e-01 1.87395379e-01 -1.31659642e-01
-6.91862345e-01 1.59782499e-01 8.19293559e-02 4.14806217e-01
6.29413545e-01 2.02537298e-01 -3.89622808e-01 6.96413457e-01
-8.60787988e-01 -6.46778941e-01 5.20018935e-01 1.08744109e+00
-4.74752963e-01 -1.58891714e+00 -3.54795396e-01 7.11915344e-02
-8.81119549e-01 -2.08292902e-01 -6.59250498e-01 9.03659642e-01
-2.89907873e-01 1.08055568e+00 -2.57938713e-01 -3.52086753e-01
4.80086833e-01 -3.00965961e-02 5.04766345e-01 -7.43972242e-01
-9.75488424e-01 -1.32001042e-01 1.03242922e+00 -4.84965116e-01
-3.19380313e-01 -7.69077361e-01 -1.47032058e+00 -5.67885697e-01
-2.97077984e-01 -6.57843798e-02 5.25457799e-01 1.11125040e+00
1.66826859e-01 5.86235583e-01 4.25318986e-01 -6.03106558e-01
-9.36317027e-01 -8.99470448e-01 2.34722674e-01 3.12518358e-01
5.09403050e-02 -4.14963931e-01 -1.85238257e-01 -1.58375233e-01] | [10.691483497619629, 9.428243637084961] |
c43d340c-1a32-4934-95dc-6b3069c7d5df | exploring-evolution-based-free-protein | 2206.06583 | null | https://arxiv.org/abs/2206.06583v2 | https://arxiv.org/pdf/2206.06583v2.pdf | Exploring evolution-aware & -free protein language models as protein function predictors | Large-scale Protein Language Models (PLMs) have improved performance in protein prediction tasks, ranging from 3D structure prediction to various function predictions. In particular, AlphaFold, a ground-breaking AI system, could potentially reshape structural biology. However, the utility of the PLM module in AlphaFold, Evoformer, has not been explored beyond structure prediction. In this paper, we investigate the representation ability of three popular PLMs: ESM-1b (single sequence), MSA-Transformer (multiple sequence alignment) and Evoformer (structural), with a special focus on Evoformer. Specifically, we aim to answer the following key questions: (i) Does the Evoformer trained as part of AlphaFold produce representations amenable to predicting protein function? (ii) If yes, can Evoformer replace ESM-1b and MSA-Transformer? (ii) How much do these PLMs rely on evolution-related protein data? In this regard, are they complementary to each other? We compare these models by empirical study along with new insights and conclusions. All code and datasets for reproducibility are available at https://github.com/elttaes/Revisiting-PLMs. | ['Qiuyang Ding', 'Fei Yang', 'Hui Wang', 'Jin Su', 'Fusong Ju', 'Kevin K. Yang', 'Fajie Yuan', 'Mingyang Hu'] | 2022-06-14 | null | null | null | null | ['multiple-sequence-alignment'] | ['medical'] | [ 3.20982426e-01 3.11113477e-01 -8.62744078e-02 -8.41917023e-02
-2.83882141e-01 -8.07905078e-01 3.23183566e-01 5.61443865e-01
-2.32949093e-01 1.07313144e+00 2.06252486e-01 -6.39276743e-01
-2.52153575e-02 -3.61210406e-01 -1.04623878e+00 -7.78934777e-01
1.24376165e-02 6.42785668e-01 1.97321355e-01 -5.06204128e-01
1.04020089e-01 5.33894718e-01 -1.31570017e+00 6.21289134e-01
1.08696425e+00 5.21464407e-01 6.22041881e-01 4.84348089e-01
-1.12957999e-01 4.64973897e-01 -2.65892267e-01 -4.00163382e-01
1.32568413e-02 -3.88970256e-01 -1.00676966e+00 -3.99903059e-01
-2.05411799e-02 2.12503955e-01 -5.13908453e-02 7.10028887e-01
3.74011517e-01 -2.04866901e-01 6.92240536e-01 -8.91913891e-01
-8.20016026e-01 3.26675028e-01 -3.01037461e-01 5.87248355e-02
4.68114436e-01 5.74596465e-01 9.72505748e-01 -7.96766520e-01
1.05375028e+00 1.26152790e+00 7.71132767e-01 5.12761414e-01
-1.59125221e+00 -2.63257951e-01 -5.38293384e-02 3.07910502e-01
-1.16528225e+00 -6.90315589e-02 2.99929917e-01 -6.88173831e-01
1.52906632e+00 2.30723426e-01 5.41802108e-01 1.18357372e+00
7.30285406e-01 7.11896777e-01 1.10467434e+00 -7.81519935e-02
1.63932264e-01 -1.58986330e-01 2.06702575e-01 7.56768584e-01
2.27771655e-01 -5.67242652e-02 -5.95199347e-01 -5.83063185e-01
3.21604609e-01 -1.46632325e-02 -4.57590371e-01 -2.69915789e-01
-1.18291450e+00 7.02163815e-01 2.95414090e-01 4.53316152e-01
-5.63191831e-01 -2.77967483e-01 2.94406354e-01 4.19712842e-01
3.20102543e-01 9.43517625e-01 -9.74826515e-01 5.56203648e-02
-4.80952799e-01 2.92756319e-01 8.13396275e-01 5.19961178e-01
8.79243433e-01 -9.52003449e-02 1.07348815e-01 7.45350897e-01
1.49984911e-01 1.81440338e-01 8.27855885e-01 -5.11816502e-01
-1.34504467e-01 6.61741436e-01 5.36678126e-03 -7.58840740e-01
-7.48338580e-01 -2.50376761e-01 -7.01146066e-01 4.36809808e-02
3.67149353e-01 1.40668526e-01 -5.77771068e-01 1.76877046e+00
2.64485925e-01 1.57416426e-02 1.59476474e-01 8.10348392e-01
6.34023905e-01 7.88713217e-01 3.64522725e-01 -1.98618293e-01
1.43742919e+00 -6.86772466e-01 -1.36586562e-01 -1.17984548e-01
8.44908357e-01 -7.59680510e-01 8.92545879e-01 3.56665790e-01
-8.92532229e-01 -5.15804708e-01 -1.00253665e+00 -1.05373532e-01
-5.17288089e-01 4.05047461e-02 5.54851711e-01 8.33831355e-02
-1.00691390e+00 9.41180408e-01 -7.90703654e-01 -7.52712667e-01
-1.19863348e-02 3.77572894e-01 -7.60637462e-01 1.79378510e-01
-1.33240879e+00 1.26036215e+00 5.71228683e-01 -3.85703146e-01
-6.37690902e-01 -7.77786732e-01 -5.80505252e-01 6.15567341e-02
1.11469164e-01 -9.57947016e-01 1.05180097e+00 -9.24841523e-01
-1.31444073e+00 1.01624024e+00 -5.48367560e-01 -8.00199032e-01
-1.26963817e-02 -2.13677719e-01 -1.17195629e-01 1.52015314e-01
-1.16749026e-01 6.80152535e-01 2.60896891e-01 -1.08142328e+00
-7.79134184e-02 -4.47110981e-01 -1.87990502e-01 -7.78158605e-02
3.98659110e-01 1.02096215e-01 2.67164558e-01 -6.79448068e-01
-1.71804816e-01 -9.67340946e-01 -4.52147663e-01 1.21279255e-01
-1.74823269e-01 -1.73479289e-01 2.80097306e-01 -8.84770751e-01
9.93267357e-01 -1.86369491e+00 5.68487108e-01 -2.08066497e-02
4.28205162e-01 7.80235767e-01 -3.91978502e-01 9.71991897e-01
-4.19475555e-01 1.66121840e-01 -4.59388286e-01 3.49895328e-01
-2.03265950e-01 1.05532497e-01 -1.81562692e-01 4.28751618e-01
4.86605823e-01 1.26564384e+00 -7.25917041e-01 -3.34893763e-02
1.46683842e-01 5.04195631e-01 -6.32324934e-01 1.65720984e-01
-7.47345328e-01 6.79505706e-01 -4.40540433e-01 5.58156610e-01
7.19159901e-01 -6.16498470e-01 6.85898006e-01 -3.04688096e-01
-2.76970983e-01 2.70326406e-01 -3.27309251e-01 1.48393691e+00
1.09407760e-01 1.88918889e-01 -9.57425460e-02 -9.92143571e-01
9.62861836e-01 1.08569078e-01 6.63770318e-01 -5.90953887e-01
-1.63438410e-01 2.21008718e-01 2.46316180e-01 -2.84171909e-01
2.12435573e-01 -2.84837902e-01 2.12059394e-01 2.55902678e-01
2.29574323e-01 3.96499008e-01 2.17477620e-01 1.20140128e-01
1.39749897e+00 5.95542908e-01 8.65700722e-01 -3.81893754e-01
5.23215890e-01 2.29124367e-01 7.45806277e-01 1.68233171e-01
-1.78531870e-01 5.04836738e-01 6.32583261e-01 -4.88128603e-01
-1.25767374e+00 -1.02333426e+00 -1.77459165e-01 1.04072857e+00
7.71121159e-02 -7.79926956e-01 -6.90535903e-01 -3.81831020e-01
2.64770359e-01 6.96490705e-01 -3.89887989e-01 -4.29472476e-01
-6.30274475e-01 -1.01052868e+00 6.07053995e-01 9.61278751e-02
-5.61206900e-02 -1.10277474e+00 -5.72362423e-01 4.96014982e-01
-2.38149047e-01 -5.98982573e-01 -2.93224096e-01 5.65878272e-01
-7.52037406e-01 -1.32373524e+00 -7.41588116e-01 -6.83563650e-01
3.63543481e-01 1.44635275e-01 1.08261883e+00 4.95032640e-04
-2.51754224e-01 -3.60901169e-02 -4.70589280e-01 -2.84812570e-01
-8.26562405e-01 1.26653630e-02 3.50362211e-01 -2.77261823e-01
6.16546154e-01 -8.72671723e-01 -5.59612691e-01 3.53606284e-01
-1.01347399e+00 4.43629742e-01 9.52434242e-01 7.97179222e-01
6.87741756e-01 -6.84886158e-01 7.05756783e-01 -1.01107502e+00
6.25998139e-01 -7.03699887e-01 -3.70967597e-01 4.56942320e-01
-5.98083496e-01 5.01397192e-01 8.46343875e-01 -4.12274241e-01
-7.21198380e-01 1.49921447e-01 -6.61228955e-01 -3.29014868e-01
-1.43298253e-01 5.96083522e-01 -2.53282458e-01 2.17014223e-01
7.43851423e-01 6.56525195e-01 3.44791234e-01 -9.47774112e-01
2.13726088e-01 4.28497374e-01 3.15459192e-01 -7.11957633e-01
4.52224910e-01 8.31052214e-02 5.02747484e-03 -8.59624088e-01
-3.83825153e-01 -4.00640130e-01 -7.41604507e-01 2.59925514e-01
8.83890808e-01 -7.00591862e-01 -8.51760387e-01 2.91413635e-01
-1.19648373e+00 -2.46378496e-01 2.02640705e-02 1.93662643e-01
-7.86000252e-01 8.14238071e-01 -8.02570403e-01 -3.36342454e-01
-5.66094816e-01 -1.22442067e+00 9.76093709e-01 2.29423456e-02
-6.00168288e-01 -6.54387057e-01 3.58600676e-01 4.14724231e-01
2.93468237e-01 2.75917888e-01 1.63793719e+00 -9.94320452e-01
-2.50695676e-01 3.79368603e-01 -1.17639080e-02 1.62473708e-01
1.18508525e-01 -1.59382537e-01 -4.99695927e-01 -3.56009215e-01
-2.29811266e-01 -4.25427824e-01 8.91923547e-01 7.66000301e-02
8.83064508e-01 -5.16546488e-01 -4.52433109e-01 5.28817058e-01
1.15606999e+00 2.80681878e-01 5.88257313e-01 4.90597934e-01
2.57892013e-01 6.42693520e-01 5.46160400e-01 2.60003775e-01
6.05800711e-02 7.21174300e-01 3.60559344e-01 1.51084557e-01
-1.55014265e-02 -4.71930563e-01 5.30275404e-01 8.76840949e-01
-5.84609285e-02 -4.11429375e-01 -9.27951276e-01 2.74868459e-01
-1.97405863e+00 -7.64296055e-01 -6.73817992e-02 1.83420706e+00
1.03754508e+00 -2.54104644e-01 -2.59479545e-02 -4.06362802e-01
3.93574178e-01 7.63627738e-02 -9.28269863e-01 -5.54858267e-01
-5.99469900e-01 2.79040784e-01 2.28994921e-01 3.06423724e-01
-7.68226147e-01 9.82003093e-01 5.78593683e+00 6.76283121e-01
-1.11046946e+00 -1.11958809e-01 5.87262213e-01 1.49412602e-01
-3.10561597e-01 3.49890649e-01 -7.67254531e-01 5.62927485e-01
1.17397392e+00 -2.02303723e-01 3.78377080e-01 7.69340873e-01
2.34451264e-01 3.14198256e-01 -1.08926058e+00 5.50526440e-01
-1.97884604e-01 -1.59837830e+00 3.39540035e-01 1.48985371e-01
2.31642097e-01 2.52923369e-01 -3.20243448e-01 3.50411534e-01
2.89202034e-01 -1.18180346e+00 3.65498245e-01 7.78438926e-01
6.57285035e-01 -4.98897314e-01 5.82241535e-01 4.77334112e-01
-9.00852382e-01 3.08306247e-01 -6.22458339e-01 2.67821908e-01
1.32152840e-01 4.77295905e-01 -8.50082695e-01 7.55830050e-01
4.42540109e-01 6.60467505e-01 -5.43023646e-01 7.82458842e-01
-3.43362428e-02 6.17051840e-01 -8.07291120e-02 5.61812147e-02
8.77826810e-02 -3.91898990e-01 7.00003743e-01 1.17763269e+00
7.27625415e-02 1.09904200e-01 1.62742496e-01 8.47869456e-01
3.46078561e-03 3.01857859e-01 -3.11976850e-01 -4.65002835e-01
1.40934214e-01 8.98790896e-01 -1.86451077e-01 -7.73995146e-02
-3.58694673e-01 1.15605342e+00 7.08634317e-01 2.23894462e-01
-6.88150644e-01 -1.98584255e-02 1.34192157e+00 4.89883214e-01
2.01029658e-01 -6.63809255e-02 3.96860272e-01 -1.19688857e+00
-3.09700072e-01 -1.30795240e+00 2.61553109e-01 -1.05701363e+00
-1.51335192e+00 5.15335023e-01 -2.52896070e-01 -8.18425357e-01
-1.93116367e-01 -1.00765944e+00 -3.28010559e-01 8.04562867e-01
-1.09654427e+00 -1.12012076e+00 4.36121792e-01 -3.19788186e-03
5.08419037e-01 -2.19561413e-01 1.05490065e+00 3.33577320e-02
-4.48397517e-01 3.94475967e-01 6.47015512e-01 -4.53008622e-01
7.87144601e-01 -1.01539969e+00 5.86711586e-01 3.47053826e-01
-2.13033214e-01 9.44831014e-01 1.02755630e+00 -9.38182533e-01
-1.49572754e+00 -1.24193418e+00 9.50927436e-01 -4.97540146e-01
5.84007382e-01 -2.70440489e-01 -1.41551912e+00 6.96866930e-01
3.22717661e-03 -5.51061332e-01 9.65880036e-01 -1.65073320e-01
-5.30555487e-01 5.49933016e-01 -1.14930236e+00 4.86405849e-01
1.10389495e+00 -2.61520207e-01 -6.91613376e-01 3.65580976e-01
1.03062773e+00 -1.20093636e-02 -1.10876155e+00 5.96488714e-01
6.63354754e-01 -8.64184976e-01 1.02796447e+00 -1.24674118e+00
6.02723479e-01 -4.07992989e-01 -1.75784424e-01 -1.28976691e+00
-6.95700347e-01 -3.84971142e-01 -2.15744406e-01 8.10082793e-01
4.81647134e-01 -9.14599061e-01 6.00504696e-01 4.18031290e-02
-2.72013396e-01 -1.23227346e+00 -9.95710850e-01 -7.74933398e-01
4.94402856e-01 1.09937258e-01 4.85156745e-01 8.95156145e-01
2.47689024e-01 4.92767632e-01 -4.84869897e-01 -2.07525864e-01
-4.93635871e-02 2.96051085e-01 6.94632769e-01 -1.26067436e+00
-7.30808914e-01 -2.75433332e-01 -4.07234013e-01 -1.08021343e+00
1.98806718e-01 -1.30715430e+00 -3.39251757e-01 -1.48344028e+00
3.95250797e-01 1.49506971e-01 -2.96488732e-01 6.65276170e-01
-1.72773004e-01 -2.19969973e-01 1.12065785e-01 5.58263659e-01
-4.44549322e-01 7.07822382e-01 9.14953709e-01 -6.22098334e-02
6.02413639e-02 -2.08098158e-01 -8.12078953e-01 5.38835824e-01
1.01253033e+00 -3.49077731e-01 -7.59674311e-02 5.36428541e-02
1.96484104e-01 4.69156913e-02 1.57639340e-01 -6.71080351e-01
-1.90786168e-01 -2.96085626e-01 3.57464701e-01 -3.74413699e-01
3.75705421e-01 -4.15056616e-01 5.87549508e-01 8.82368445e-01
-1.49147376e-01 3.81820679e-01 3.79864216e-01 6.08611405e-01
9.63066332e-03 -9.95825604e-02 8.51653039e-01 -4.31468725e-01
-5.90843320e-01 1.47473156e-01 -1.05089247e+00 -3.30593526e-01
9.02284682e-01 -1.21255361e-01 -4.74842012e-01 3.13300118e-02
-7.31570780e-01 4.21694666e-02 9.50182498e-01 4.49622571e-01
4.67106432e-01 -7.93600321e-01 -7.02707887e-01 8.59620273e-02
3.66947681e-01 -5.59055090e-01 1.56507790e-01 7.56805658e-01
-6.47769392e-01 9.36368287e-01 -4.41972464e-01 -3.33125710e-01
-1.41972661e+00 9.23136890e-01 3.60772699e-01 -4.97772127e-01
-4.37019408e-01 4.91902590e-01 6.70571446e-01 -8.13864589e-01
-3.18004876e-01 -2.60086451e-02 -7.24714696e-02 -3.21656734e-01
2.81879067e-01 3.91867608e-02 -4.74140160e-02 -8.98282290e-01
-4.31562155e-01 2.18969882e-01 -3.41439188e-01 6.67666137e-01
1.54381955e+00 1.30514428e-01 -4.13896859e-01 3.07729274e-01
1.08507872e+00 -2.19698057e-01 -9.00000393e-01 -2.89053053e-01
2.95105457e-01 2.82790400e-02 -7.69272208e-01 -1.13155591e+00
-3.27906340e-01 6.76463902e-01 5.84813595e-01 -5.42124689e-01
9.02066767e-01 2.01462969e-01 8.52393270e-01 4.80811954e-01
6.01225972e-01 -4.34267312e-01 -1.66021675e-01 6.82228208e-01
9.55894232e-01 -8.66383016e-01 -1.71912283e-01 -1.91855147e-01
-6.50456965e-01 8.54710162e-01 6.60761297e-01 1.40354931e-01
3.87355864e-01 -1.85686387e-02 -3.16310138e-01 -2.55040795e-01
-1.22317970e+00 -8.46596882e-02 1.85509115e-01 5.53078055e-01
8.03002834e-01 2.92339846e-02 -7.31237233e-01 9.35235620e-01
-2.32067078e-01 -4.54184227e-02 2.39689261e-01 1.01469994e+00
-7.91385233e-01 -1.48209369e+00 -1.33179545e-01 4.83041257e-01
-4.34504718e-01 -2.46802613e-01 -1.10661113e+00 4.69135016e-01
1.78962871e-01 6.69717729e-01 -3.88646692e-01 -4.87095684e-01
1.66041926e-01 5.23669183e-01 3.52062911e-01 -5.52641392e-01
-7.39842713e-01 -2.07741201e-01 8.99437517e-02 -4.10392940e-01
-6.34405091e-02 -5.85493565e-01 -1.27154660e+00 -3.95388126e-01
-4.34089750e-02 4.73269999e-01 3.77272695e-01 8.02232921e-01
1.06318021e+00 3.07359248e-01 -3.69958095e-02 -4.77319270e-01
-5.36269724e-01 -1.08505511e+00 -3.58813345e-01 3.38364631e-01
6.27958998e-02 -4.57774669e-01 2.93262489e-02 4.37339991e-02] | [4.6611456871032715, 5.649188041687012] |
2deb3b9e-d587-4e9b-876f-19d77b988e5d | multi-level-cross-view-contrastive-learning | 2204.08807 | null | https://arxiv.org/abs/2204.08807v1 | https://arxiv.org/pdf/2204.08807v1.pdf | Multi-level Cross-view Contrastive Learning for Knowledge-aware Recommender System | Knowledge graph (KG) plays an increasingly important role in recommender systems. Recently, graph neural networks (GNNs) based model has gradually become the theme of knowledge-aware recommendation (KGR). However, there is a natural deficiency for GNN-based KGR models, that is, the sparse supervised signal problem, which may make their actual performance drop to some extent. Inspired by the recent success of contrastive learning in mining supervised signals from data itself, in this paper, we focus on exploring the contrastive learning in KG-aware recommendation and propose a novel multi-level cross-view contrastive learning mechanism, named MCCLK. Different from traditional contrastive learning methods which generate two graph views by uniform data augmentation schemes such as corruption or dropping, we comprehensively consider three different graph views for KG-aware recommendation, including global-level structural view, local-level collaborative and semantic views. Specifically, we consider the user-item graph as a collaborative view, the item-entity graph as a semantic view, and the user-item-entity graph as a structural view. MCCLK hence performs contrastive learning across three views on both local and global levels, mining comprehensive graph feature and structure information in a self-supervised manner. Besides, in semantic view, a k-Nearest-Neighbor (kNN) item-item semantic graph construction module is proposed, to capture the important item-item semantic relation which is usually ignored by previous work. Extensive experiments conducted on three benchmark datasets show the superior performance of our proposed method over the state-of-the-arts. The implementations are available at: https://github.com/CCIIPLab/MCCLK. | ['Xin Cao', 'Feida Zhu', 'Minghui Qiu', 'Ziyang Wang', 'Xian-Ling Mao', 'Wei Wei', 'Ding Zou'] | 2022-04-19 | null | null | null | null | ['knowledge-aware-recommendation'] | ['miscellaneous'] | [-7.50377774e-02 -4.79479469e-02 -7.31023669e-01 -2.58904964e-01
-3.26286107e-01 -2.58949161e-01 1.62739247e-01 3.05511151e-02
8.19570199e-02 4.30869043e-01 4.39096123e-01 -1.15408719e-01
-7.22484469e-01 -1.07058883e+00 -5.71195841e-01 -5.97550929e-01
-8.79748464e-02 2.05430120e-01 -5.25154807e-02 -4.57676351e-01
5.07512875e-02 1.13194302e-01 -1.16729283e+00 2.94573456e-01
8.78749728e-01 8.48145425e-01 1.32693693e-01 2.06993431e-01
-1.45400807e-01 9.51371908e-01 -9.55902413e-02 -5.84240556e-01
3.07892680e-01 -4.68111455e-01 -4.89204168e-01 1.78781733e-01
2.67458826e-01 6.52087256e-02 -4.98291194e-01 1.22513676e+00
5.14399230e-01 3.36976975e-01 3.80962640e-01 -1.29821134e+00
-1.25836647e+00 1.08209217e+00 -8.72538149e-01 5.67104556e-02
3.60079646e-01 -2.76599765e-01 1.53736043e+00 -1.03603601e+00
4.98694718e-01 1.06956196e+00 6.76313996e-01 2.75044799e-01
-8.96079898e-01 -5.98356307e-01 8.08338761e-01 4.37985092e-01
-1.38305950e+00 4.22240309e-02 1.23637104e+00 -8.96079019e-02
6.43029153e-01 2.17508838e-01 7.67239869e-01 9.47069526e-01
-2.38491222e-01 9.48023200e-01 1.14085579e+00 -1.43378943e-01
-4.72433720e-04 -1.59604009e-02 3.77876312e-01 8.04495573e-01
6.34631455e-01 -1.33090019e-01 -6.68577611e-01 -1.83939904e-01
8.66458714e-01 4.24250931e-01 -5.12028992e-01 -5.83492100e-01
-9.54765201e-01 8.41332376e-01 6.92353427e-01 2.67024279e-01
-3.74905825e-01 -7.36757964e-02 3.10860068e-01 5.09652078e-01
6.10400975e-01 1.15808144e-01 -5.91284573e-01 4.08508271e-01
-5.53826869e-01 -1.10737570e-01 7.92473435e-01 1.00188649e+00
7.40075052e-01 3.28054637e-01 1.78946450e-01 9.69932973e-01
5.83530545e-01 2.53907740e-01 4.69533622e-01 -1.97280258e-01
6.44740880e-01 8.82135570e-01 -4.62171495e-01 -1.50151253e+00
-3.63246739e-01 -1.02518904e+00 -1.13495159e+00 -5.19535720e-01
4.07564491e-02 -9.77074057e-02 -6.86941326e-01 1.43626463e+00
3.97919863e-01 6.82170570e-01 1.23980172e-01 1.14139903e+00
1.31623387e+00 4.11623061e-01 -1.15125932e-01 -4.02197093e-01
1.11824644e+00 -1.21855724e+00 -6.96579456e-01 4.94164117e-02
6.93353593e-01 -3.94619524e-01 9.11139727e-01 5.65026641e-01
-7.87504971e-01 -6.26897514e-01 -9.89493608e-01 1.75221995e-01
-4.32742894e-01 7.36384839e-02 9.60702240e-01 3.33263367e-01
-8.34581733e-01 4.76173252e-01 -3.13014328e-01 -2.84412056e-01
3.87302309e-01 2.15750262e-01 -4.88768965e-01 -4.42542285e-01
-1.38234472e+00 2.99212247e-01 5.28971493e-01 2.58196563e-01
-4.24244702e-01 -6.68543518e-01 -7.32077301e-01 1.95549011e-01
9.71546233e-01 -8.63393366e-01 6.07144475e-01 -9.65868175e-01
-1.23382723e+00 3.48486245e-01 2.78005928e-01 -2.22660899e-01
-2.42640097e-02 -1.45954788e-01 -1.05565727e+00 -2.15783510e-02
-2.28743926e-01 -1.61884859e-01 8.28660071e-01 -1.48650539e+00
-5.19935369e-01 -6.79160655e-01 2.76259601e-01 5.99712074e-01
-5.10053396e-01 -3.16929042e-01 -8.01567793e-01 -1.06801927e+00
3.35218191e-01 -7.10873365e-01 -2.43213013e-01 -6.89553022e-01
-3.47676098e-01 -2.00004026e-01 7.19912648e-01 -6.25559568e-01
1.58583283e+00 -2.01439929e+00 2.02989265e-01 5.81636727e-01
5.41158319e-01 2.48941988e-01 -2.71956414e-01 6.02701902e-01
-2.93994039e-01 -1.09674081e-01 2.17213556e-01 -2.22480819e-02
-2.13085547e-01 2.84166485e-01 -1.07107937e-01 3.19473624e-01
-4.09741849e-01 1.18366444e+00 -1.13784838e+00 -1.75136223e-01
-2.84658913e-02 3.99662107e-01 -5.62117577e-01 -2.91728638e-02
-5.34861609e-02 2.38271639e-01 -7.38292396e-01 8.29191327e-01
7.03236163e-01 -7.48988390e-01 5.89492023e-01 -6.89755261e-01
3.53550106e-01 2.51903564e-01 -1.45746565e+00 1.73344970e+00
-3.38963181e-01 -2.94840068e-01 -1.43222548e-02 -1.45207870e+00
1.01648235e+00 1.84832692e-01 5.12009859e-01 -6.23774827e-01
-1.35651454e-02 7.33257458e-02 9.27209407e-02 -2.42830142e-01
5.33169031e-01 3.75369638e-02 -6.67196186e-03 3.94172609e-01
3.72536361e-01 7.45926619e-01 -2.07606666e-02 6.18740618e-01
9.51924026e-01 6.44180179e-02 5.16373992e-01 -2.37685382e-01
5.90421081e-01 -1.76450372e-01 5.77539206e-01 5.63909888e-01
2.58415371e-01 4.91234690e-01 1.90856755e-01 -3.65177751e-01
-2.37686217e-01 -8.27222288e-01 4.32468504e-01 1.20353532e+00
4.43331629e-01 -8.33933115e-01 -2.26507336e-01 -1.14945805e+00
9.79584679e-02 4.26298410e-01 -4.29736167e-01 -3.31981570e-01
-4.92404401e-01 -8.16000283e-01 4.07123864e-02 5.29372871e-01
4.08013880e-01 -9.43706810e-01 4.84088361e-01 2.59618461e-01
-1.26987830e-01 -9.95456100e-01 -6.46867335e-01 -2.06844006e-02
-9.32615221e-01 -1.19657934e+00 -5.00087738e-01 -7.17296541e-01
7.64726818e-01 9.60538745e-01 1.17584562e+00 2.43495926e-01
1.35335207e-01 7.68450081e-01 -9.52743471e-01 -1.15679108e-01
1.24009982e-01 -4.25304808e-02 1.77055225e-01 4.08282578e-01
3.99300992e-01 -9.12090957e-01 -7.70390809e-01 4.22054470e-01
-6.66103780e-01 1.27374843e-01 6.82767451e-01 7.03659475e-01
9.24777806e-01 3.59166622e-01 1.05848074e+00 -1.60787654e+00
7.61115491e-01 -8.58453512e-01 -2.48885304e-01 3.93154979e-01
-9.96897697e-01 -3.05070847e-01 9.40797746e-01 -3.71166408e-01
-9.47829068e-01 -2.26549268e-01 1.07064480e-02 -6.40343845e-01
3.19256447e-02 1.22335756e+00 -5.21703243e-01 -3.61692011e-02
3.12026262e-01 3.93636376e-01 -3.58070314e-01 -6.82937801e-01
6.95247769e-01 4.90316242e-01 2.38769218e-01 -3.91946763e-01
9.65152681e-01 3.18256170e-01 -1.23128973e-01 -5.18310785e-01
-1.09666884e+00 -8.46893549e-01 -3.87361348e-01 -2.32103840e-01
3.51981193e-01 -1.24398649e+00 -5.18194377e-01 2.72431433e-01
-4.16186243e-01 9.18022841e-02 -3.14238936e-01 6.36715353e-01
-1.74384877e-01 4.83452588e-01 -3.88965130e-01 -4.59708214e-01
-5.78651369e-01 -6.68327034e-01 7.01976955e-01 1.36476874e-01
2.91734606e-01 -1.20609164e+00 -9.86979604e-02 6.91621363e-01
1.81564733e-01 -1.08258240e-01 6.91131294e-01 -1.05721068e+00
-5.25267541e-01 3.01305467e-04 -3.38164091e-01 3.74821335e-01
4.23253506e-01 -6.50578439e-01 -6.04520321e-01 -5.10011852e-01
-1.87957987e-01 -1.39676124e-01 8.78036618e-01 2.33728588e-01
1.21410763e+00 -4.53684986e-01 -3.59502852e-01 6.59411609e-01
1.67008770e+00 4.98935878e-02 3.13124597e-01 -8.29274356e-02
1.48397255e+00 3.01228136e-01 6.37447953e-01 3.71328712e-01
8.01808357e-01 5.93514740e-01 4.36972529e-01 -1.53666198e-01
-1.78147152e-01 -6.76116407e-01 2.61682659e-01 1.66565228e+00
-4.40019071e-01 -5.47258139e-01 -4.17683691e-01 3.08076978e-01
-2.30714679e+00 -8.93698394e-01 -3.54170591e-01 2.09126163e+00
3.36646110e-01 -2.83386018e-02 2.32128084e-01 -2.58160438e-02
7.42406785e-01 3.80668163e-01 -5.34945309e-01 2.93611991e-03
-1.80045903e-01 -6.40644133e-02 4.56374019e-01 9.10467952e-02
-1.05168366e+00 8.42425108e-01 4.28145409e+00 1.10540020e+00
-9.16935503e-01 1.74024969e-01 2.57070035e-01 4.78482917e-02
-5.44533312e-01 1.26818111e-02 -7.42762685e-01 4.76596057e-01
4.97596145e-01 -2.20205486e-01 6.96937084e-01 9.70638633e-01
7.80279264e-02 3.61833125e-01 -8.04327011e-01 1.25732100e+00
4.15793985e-01 -1.33189988e+00 4.08333778e-01 1.45414799e-01
8.86055887e-01 -1.19027132e-02 1.47195291e-02 4.84632045e-01
4.57240462e-01 -6.52351141e-01 1.36376768e-01 6.08358085e-01
5.03025830e-01 -9.45697367e-01 6.22234523e-01 4.18842286e-01
-1.76776278e+00 -5.58228903e-02 -4.24261600e-01 3.87036055e-02
9.08901691e-02 7.43136883e-01 -2.40185216e-01 1.65821755e+00
8.30371380e-01 1.48041737e+00 -4.14119989e-01 7.38301635e-01
-2.74047643e-01 8.86534512e-01 8.08184072e-02 2.36962631e-01
5.69613725e-02 -7.14674175e-01 5.17497003e-01 9.04961407e-01
2.47096166e-01 4.04316097e-01 5.46352446e-01 3.63578677e-01
-4.07802224e-01 4.64518905e-01 -6.96932137e-01 -1.55642731e-02
4.10418451e-01 1.63191724e+00 -6.74985707e-01 -2.46110827e-01
-8.80377471e-01 8.67939651e-01 4.77782756e-01 4.57257241e-01
-6.95685446e-01 -1.51000813e-01 2.59893537e-01 9.79976133e-02
6.38854146e-01 4.98691313e-02 1.55218273e-01 -1.55849218e+00
-6.80952445e-02 -1.03540111e+00 8.76212001e-01 -5.60415328e-01
-1.86373818e+00 1.75678790e-01 -1.55806646e-01 -1.42095447e+00
3.29583049e-01 -3.87519807e-01 -5.39133191e-01 6.20639324e-01
-1.59712112e+00 -1.53599119e+00 -2.79052466e-01 9.61550534e-01
3.04755956e-01 -3.84512544e-01 4.80855644e-01 6.65217578e-01
-4.04112607e-01 8.73775721e-01 1.39572248e-02 1.85835212e-01
5.83354592e-01 -1.13925111e+00 9.92261246e-02 6.76383495e-01
7.21672118e-01 8.06869268e-01 9.21632573e-02 -8.02220821e-01
-1.88170922e+00 -1.52009821e+00 2.97346413e-01 -2.18634784e-01
8.11499417e-01 -2.66047388e-01 -9.95345235e-01 8.34220886e-01
-1.36843070e-01 2.96687573e-01 9.81140673e-01 5.79655170e-01
-5.66339016e-01 -2.58864194e-01 -8.72786283e-01 5.30026913e-01
1.62415767e+00 -4.03197795e-01 -4.44408536e-01 3.40214998e-01
8.56655836e-01 -2.11747736e-01 -1.15438807e+00 6.45499468e-01
4.64755714e-01 -8.81223023e-01 1.06702125e+00 -6.33567095e-01
9.28601474e-02 -5.10579705e-01 -2.22610965e-01 -1.46600997e+00
-7.12165952e-01 -3.02930415e-01 -6.80029929e-01 1.41055942e+00
2.21273333e-01 -8.35768700e-01 8.62218559e-01 -2.25708723e-01
-2.69311368e-01 -1.07074654e+00 -4.26288158e-01 -7.82474279e-01
-5.24058223e-01 -3.96527797e-01 6.28400564e-01 1.55792534e+00
7.58698657e-02 7.80459106e-01 -7.13536203e-01 3.75433773e-01
6.04124665e-01 6.16696894e-01 6.86999917e-01 -1.36190426e+00
-7.39186823e-01 -1.76041588e-01 -3.78895223e-01 -1.14766562e+00
-1.03124313e-03 -1.45740831e+00 -6.15946770e-01 -1.75596333e+00
3.82585406e-01 -4.45303500e-01 -1.02341688e+00 4.70954418e-01
-2.38645688e-01 1.80049971e-01 1.33743986e-01 2.98532099e-01
-1.05977416e+00 5.76845109e-01 1.41575778e+00 -4.42112088e-02
-3.22234809e-01 4.03730236e-02 -1.36168933e+00 7.98383236e-01
7.82421231e-01 -3.28439832e-01 -9.61298525e-01 -2.45630071e-01
5.75943053e-01 2.84817386e-02 1.67963386e-01 -6.02594912e-01
3.43495637e-01 -8.19392055e-02 1.93306684e-01 -5.65270603e-01
5.87186888e-02 -8.86256635e-01 3.52738768e-01 6.14241175e-02
-1.05810560e-01 -1.99348051e-02 -2.70070821e-01 1.09391260e+00
-3.53252798e-01 1.71644539e-01 3.31674546e-01 -2.90318578e-01
-9.71480012e-01 8.96631360e-01 3.24459702e-01 8.95603672e-02
6.29663408e-01 -2.93371409e-01 -4.49434698e-01 -3.76732737e-01
-8.66198063e-01 2.98545569e-01 1.32572815e-01 6.49180353e-01
7.96282649e-01 -1.49804938e+00 -5.76476932e-01 4.19642217e-03
4.68003750e-01 -1.38411090e-01 6.85704350e-01 9.49616551e-01
1.83600828e-01 1.42937467e-01 2.02912688e-01 -8.27462077e-02
-1.18132329e+00 8.65523160e-01 -5.14184758e-02 -6.25804424e-01
-7.33553410e-01 8.04297805e-01 4.30981159e-01 -6.59701407e-01
3.37331003e-05 1.34676024e-01 -7.02939689e-01 1.57146960e-01
3.08895856e-01 4.09386188e-01 1.85936857e-02 -7.41822958e-01
-1.68522045e-01 5.34479678e-01 -2.77238190e-01 6.63657188e-01
1.45270789e+00 -2.29987457e-01 -1.31821334e-01 4.16071683e-01
9.28031266e-01 3.15940797e-01 -5.60227752e-01 -7.15616822e-01
-1.38004199e-01 -3.22403312e-01 1.66535050e-01 -8.51249397e-01
-1.49299872e+00 4.32243675e-01 2.22932771e-01 3.13776910e-01
1.28024256e+00 6.01732917e-02 8.41447830e-01 3.42634410e-01
6.88096106e-01 -9.56884384e-01 1.01294674e-01 3.17239791e-01
7.62258351e-01 -1.23634577e+00 3.62570435e-01 -7.32191563e-01
-6.79931045e-01 6.86565638e-01 7.68691540e-01 -3.39707077e-01
1.07077503e+00 -1.89968765e-01 -2.22214013e-01 -5.10750711e-01
-6.22901142e-01 -3.85536224e-01 6.94289386e-01 5.70327878e-01
2.27571830e-01 2.73225158e-01 -2.50897497e-01 1.15289199e+00
1.15970120e-01 -8.66413675e-03 2.28292480e-01 7.04656243e-01
-2.15012059e-01 -1.16228843e+00 1.68903649e-01 9.67392385e-01
-4.25952882e-01 -3.82762671e-01 -1.76647425e-01 8.07378471e-01
1.04792424e-01 9.39765930e-01 -4.79101241e-01 -7.60503054e-01
3.57813746e-01 -3.51697803e-01 3.49376202e-01 -8.08735847e-01
-5.01748681e-01 1.65116802e-01 1.71572611e-01 -5.90602219e-01
-6.68736458e-01 -3.09886307e-01 -1.17447650e+00 -1.16626285e-01
-6.11775637e-01 3.35883707e-01 1.36167020e-01 8.53968859e-01
6.55861735e-01 5.75715899e-01 7.14998722e-01 -4.71390069e-01
-2.26416722e-01 -7.35369742e-01 -1.13448906e+00 4.77202922e-01
-1.18128739e-01 -6.84047878e-01 -2.74594754e-01 -1.36746287e-01] | [10.226811408996582, 5.62482213973999] |
f97c6739-2ef9-4bab-8ee2-889441364964 | a-comprehensive-survey-on-pose-invariant-face | 1502.04383 | null | http://arxiv.org/abs/1502.04383v3 | http://arxiv.org/pdf/1502.04383v3.pdf | A Comprehensive Survey on Pose-Invariant Face Recognition | The capacity to recognize faces under varied poses is a fundamental human
ability that presents a unique challenge for computer vision systems. Compared
to frontal face recognition, which has been intensively studied and has
gradually matured in the past few decades, pose-invariant face recognition
(PIFR) remains a largely unsolved problem. However, PIFR is crucial to
realizing the full potential of face recognition for real-world applications,
since face recognition is intrinsically a passive biometric technology for
recognizing uncooperative subjects. In this paper, we discuss the inherent
difficulties in PIFR and present a comprehensive review of established
techniques. Existing PIFR methods can be grouped into four categories, i.e.,
pose-robust feature extraction approaches, multi-view subspace learning
approaches, face synthesis approaches, and hybrid approaches. The motivations,
strategies, pros/cons, and performance of representative approaches are
described and compared. Moreover, promising directions for future research are
discussed. | ['DaCheng Tao', 'Changxing Ding'] | 2015-02-15 | null | null | null | null | ['robust-face-recognition'] | ['computer-vision'] | [ 4.55711275e-01 -4.27308768e-01 -1.76966578e-01 -4.29052293e-01
-6.11170411e-01 -5.82502902e-01 6.08339489e-01 -9.54924226e-01
-9.12635177e-02 5.08583784e-01 1.38402849e-01 2.23332077e-01
-3.23100507e-01 -3.03664237e-01 -5.68572022e-02 -1.10324442e+00
2.49424595e-02 8.99355337e-02 -3.39274824e-01 -1.82237271e-02
1.09153889e-01 1.38754308e+00 -2.04150724e+00 -1.19677000e-02
3.76288027e-01 1.19141531e+00 -2.40560621e-01 2.96080559e-01
1.14349172e-01 2.31561452e-01 -4.45247471e-01 -6.12546623e-01
3.39032799e-01 -4.63574708e-01 -4.27360415e-01 4.44491297e-01
8.43092084e-01 -2.99823225e-01 -5.68938971e-01 1.08923769e+00
6.79721534e-01 3.14697176e-02 6.78632915e-01 -1.20709753e+00
-6.00940526e-01 -1.32748991e-01 -7.58129239e-01 1.38808191e-01
6.67586923e-01 -3.34736496e-01 4.13255215e-01 -1.42203116e+00
3.33656162e-01 1.51459301e+00 4.36419070e-01 1.06525469e+00
-1.17105746e+00 -7.94527113e-01 8.64786133e-02 3.17917645e-01
-1.57221091e+00 -1.12263918e+00 1.09768772e+00 -3.15153778e-01
5.96124232e-01 5.86024702e-01 2.66511679e-01 1.12976730e+00
9.58195850e-02 6.57014668e-01 1.32493436e+00 -2.92922437e-01
-6.79316074e-02 5.09653203e-02 -5.72813824e-02 5.15354037e-01
2.83342123e-01 1.60107508e-01 -8.03869903e-01 -2.52750486e-01
9.91186559e-01 1.96318179e-01 -4.67970937e-01 -7.90122449e-01
-9.50634062e-01 4.61083770e-01 -8.91046152e-02 2.85375595e-01
-1.86857179e-01 -3.94662052e-01 5.13502732e-02 1.35812640e-01
2.08140016e-01 8.67054164e-02 -2.42150173e-01 3.10947653e-02
-1.03522325e+00 1.62504435e-01 5.89579225e-01 7.53279686e-01
4.74164099e-01 5.29655457e-01 9.08277482e-02 1.07299340e+00
3.77628624e-01 9.60111916e-01 2.27568880e-01 -7.43867338e-01
-2.67940592e-02 5.61818361e-01 -9.27819461e-02 -1.06401277e+00
-2.39979461e-01 -2.28289768e-01 -8.32601190e-01 9.10018906e-02
3.11599463e-01 1.32741094e-01 -5.53258538e-01 1.35431337e+00
3.20080370e-01 -1.80253778e-02 7.77171105e-02 6.91920877e-01
9.85112250e-01 6.56259134e-02 -3.08765829e-01 -7.21953034e-01
1.49268758e+00 -5.80999672e-01 -8.85236442e-01 -8.74068514e-02
-4.65487838e-01 -1.02280700e+00 4.09511536e-01 4.74566549e-01
-9.75010157e-01 -4.76587474e-01 -8.10524166e-01 2.10935682e-01
-4.35944125e-02 3.92093599e-01 6.73641205e-01 1.16776824e+00
-1.03576255e+00 2.72177547e-01 -8.07175636e-01 -5.74568093e-01
4.62573767e-01 7.39567220e-01 -8.25860798e-01 -2.80377388e-01
-6.52812541e-01 6.92749739e-01 -7.66156241e-02 4.72742945e-01
-6.84818208e-01 -3.37640464e-01 -8.12119246e-01 -1.50012180e-01
5.20625651e-01 -2.43235543e-01 1.00892341e+00 -7.34685421e-01
-1.63415647e+00 1.00971639e+00 -8.16056013e-01 1.73919216e-01
2.52211064e-01 -1.44359514e-01 -8.87286842e-01 2.34864846e-01
-3.43339145e-01 2.02603981e-01 1.28656614e+00 -1.22603762e+00
-2.94598460e-01 -1.10391915e+00 -1.74431518e-01 1.39522433e-01
-4.07794327e-01 5.96138060e-01 -5.67096651e-01 -5.69341183e-01
4.91898447e-01 -1.05537140e+00 1.76675975e-01 1.78841457e-01
-9.65793803e-02 -2.70503968e-01 1.16357851e+00 -4.49579328e-01
1.09246111e+00 -2.18541265e+00 3.28314990e-01 5.92071265e-02
8.45536962e-02 3.75395298e-01 -5.26301861e-02 4.22699720e-01
-3.27497810e-01 -3.14157277e-01 -9.29383487e-02 -1.78802032e-02
-2.78789222e-01 -1.52476251e-01 -4.07185197e-01 8.53187919e-01
1.14246897e-01 7.06471980e-01 -5.49171329e-01 -1.94917649e-01
2.93538898e-01 8.35678935e-01 -1.89829558e-01 1.22842066e-01
4.11834031e-01 5.25788426e-01 -4.95994538e-01 1.29667807e+00
9.48174238e-01 1.53278530e-01 2.76762247e-01 -4.66386586e-01
-1.14781119e-01 -4.49505478e-01 -1.33008015e+00 1.05622733e+00
-2.52914298e-02 4.05636787e-01 3.67073715e-01 -7.51593471e-01
1.20955193e+00 5.69756567e-01 6.47173643e-01 -3.68900478e-01
6.56074146e-03 2.82132864e-01 -8.45995322e-02 -4.79955018e-01
2.13956639e-01 -2.56397843e-01 2.39467159e-01 4.55478318e-02
9.77561176e-02 2.57361561e-01 -4.41902950e-02 -3.02621216e-01
5.91778994e-01 1.71628147e-01 6.59610331e-01 -1.73374027e-01
1.26810348e+00 -5.16658425e-01 6.25874400e-01 1.02298148e-01
-5.49969852e-01 6.36588871e-01 2.34442413e-01 -5.70881248e-01
-3.40781033e-01 -1.22738326e+00 -5.08095562e-01 6.82900131e-01
1.02081984e-01 -1.09797634e-01 -7.08972931e-01 -6.05589628e-01
6.35511875e-02 -2.70358603e-02 -5.01147091e-01 6.42736703e-02
-6.23084962e-01 -7.35243857e-01 3.79631668e-01 6.15700245e-01
7.76762307e-01 -7.87736595e-01 -2.60407239e-01 -2.06235856e-01
-5.27850315e-02 -1.22709501e+00 -4.05042201e-01 -4.73873615e-01
-8.09068143e-01 -1.28332651e+00 -9.52969909e-01 -8.26887012e-01
7.70948768e-01 9.49668348e-01 7.88960338e-01 -1.08839206e-01
-4.63392943e-01 9.91304755e-01 -5.77583276e-02 -2.50832736e-01
1.38739780e-01 -4.35323626e-01 6.50400817e-01 4.38208401e-01
7.92533517e-01 -3.73472869e-01 -7.15324044e-01 7.21317708e-01
-6.44720554e-01 -5.96333027e-01 5.88017344e-01 8.76991451e-01
6.91385746e-01 -1.26384065e-01 6.87593341e-01 -6.92102849e-01
3.09212387e-01 1.02126531e-01 -5.86487710e-01 5.08699238e-01
-3.08593690e-01 -4.38499361e-01 3.85037601e-01 -1.85836092e-01
-1.30791128e+00 3.94798160e-01 -1.75248504e-01 -5.35838425e-01
-4.91980493e-01 1.47512168e-01 -8.52813303e-01 -7.38611817e-01
5.70939422e-01 6.36008620e-01 4.13584203e-01 -5.06561220e-01
9.99722406e-02 8.11357796e-01 5.84111571e-01 -4.75996047e-01
1.06816339e+00 6.91838205e-01 1.33992493e-01 -1.59502852e+00
-6.33089125e-01 -7.10206449e-01 -8.60389888e-01 -4.58339661e-01
5.90954721e-01 -7.92365074e-01 -9.14185405e-01 6.73228323e-01
-8.70093822e-01 6.37837768e-01 1.38951227e-01 3.38709593e-01
-4.19814497e-01 6.26873076e-01 -1.41013950e-01 -1.06769335e+00
-2.43920818e-01 -1.26344776e+00 1.20835042e+00 4.55657393e-01
6.99321553e-02 -8.07803690e-01 -1.00252777e-01 7.54598379e-01
3.91919196e-01 2.11126238e-01 4.57277745e-01 -3.06391269e-01
-5.86636066e-01 -3.94060135e-01 -9.59011838e-02 3.17155540e-01
5.24631381e-01 5.26025146e-02 -1.48797607e+00 -7.90190995e-01
3.40939164e-01 -2.44299993e-01 6.16402924e-01 2.10618541e-01
1.26834059e+00 -1.00903437e-01 -4.80773300e-01 6.47487104e-01
1.17811859e+00 5.76875269e-01 7.33988643e-01 -2.27818087e-01
5.11624038e-01 7.79428601e-01 3.01648587e-01 1.99148983e-01
7.64013454e-02 9.21224415e-01 2.15188667e-01 2.07423568e-01
-1.56115353e-01 4.10046242e-02 4.55628991e-01 6.69010580e-01
-4.70207781e-01 1.63531795e-01 -6.09229207e-01 -6.77530020e-02
-1.33256233e+00 -1.24694014e+00 2.92597979e-01 2.21943116e+00
1.33706003e-01 -5.81632376e-01 2.15482920e-01 3.66888523e-01
7.51763463e-01 1.67046800e-01 -7.46382594e-01 -1.97579768e-02
-2.86690891e-01 2.81317353e-01 4.37869094e-02 7.96587244e-02
-1.32993174e+00 7.75524914e-01 6.81399488e+00 6.95515990e-01
-1.39501834e+00 -1.49171874e-01 3.70132595e-01 2.00220704e-01
2.38874219e-02 -3.88694406e-01 -1.12024951e+00 3.88273671e-02
4.70243692e-01 -1.30170807e-01 6.50706887e-01 9.65283394e-01
-1.68394074e-01 2.76482925e-02 -1.19646406e+00 1.59493768e+00
8.85399699e-01 -8.68741512e-01 2.25884050e-01 2.45061859e-01
4.90103006e-01 -5.08332670e-01 5.61858833e-01 1.16526335e-01
-6.66502357e-01 -1.11324775e+00 1.28302902e-01 3.62747043e-01
1.13314641e+00 -7.68380404e-01 4.06960160e-01 4.88745756e-02
-1.45723999e+00 -1.71701610e-01 -5.01085162e-01 1.26502007e-01
-1.76128760e-01 3.56007159e-01 -1.92117736e-01 8.10874224e-01
5.00453293e-01 7.77410567e-01 -5.68464994e-01 9.74596500e-01
-9.27602686e-03 2.90512979e-01 1.01226926e-01 2.85869867e-01
-4.00762677e-01 -4.50921297e-01 6.04222655e-01 7.11526036e-01
3.36872935e-01 2.33699173e-01 8.15933347e-02 4.01234180e-01
1.08527867e-02 1.02342770e-01 -7.92932212e-01 -2.16612235e-01
4.30906892e-01 1.50948286e+00 -6.98700190e-01 2.92198300e-01
-8.69775295e-01 8.65098357e-01 8.13311040e-02 4.32597846e-01
-3.56727242e-01 -1.43489987e-01 8.84231746e-01 2.22978443e-02
1.12602435e-01 -3.28274190e-01 3.29850353e-02 -1.48342586e+00
1.72218606e-01 -1.25825083e+00 3.42022270e-01 -1.01986676e-01
-1.18349540e+00 7.76444197e-01 -2.81002442e-03 -1.45472121e+00
-1.08599171e-01 -9.99336839e-01 -2.76873231e-01 7.61554599e-01
-1.54864013e+00 -1.28246045e+00 -3.90143603e-01 7.16867030e-01
5.49678683e-01 -5.85349679e-01 1.22068536e+00 2.42928669e-01
-8.23074639e-01 9.66027915e-01 1.14743218e-01 1.36170089e-01
6.45271003e-01 -6.35431111e-01 8.33884627e-03 8.14888716e-01
2.55728900e-01 1.01686943e+00 3.89767855e-01 -3.32387716e-01
-2.25753331e+00 -9.42464411e-01 6.40327990e-01 -4.26058352e-01
2.08383828e-01 -2.80349374e-01 -8.97706211e-01 5.54433942e-01
-1.37969732e-01 2.28379071e-01 1.06398022e+00 1.79781541e-01
-5.93619227e-01 -5.27471602e-01 -1.37624228e+00 6.15076661e-01
1.15701878e+00 -6.37689948e-01 -3.21018070e-01 1.47449479e-01
-2.77830482e-01 -4.17653769e-01 -8.32410932e-01 5.35484016e-01
1.07453358e+00 -1.00773191e+00 1.21910036e+00 -2.21120030e-01
-2.54707962e-01 -3.02351058e-01 -3.93865496e-01 -8.25660408e-01
-3.78656983e-01 -5.26535749e-01 -2.81274289e-01 1.28235614e+00
-1.52514711e-01 -8.02019596e-01 9.99930143e-01 5.84946573e-01
2.62662888e-01 -7.29122162e-01 -1.13483632e+00 -1.09633327e+00
-9.91977975e-02 5.74425384e-02 4.93822634e-01 6.80198550e-01
-1.57948270e-01 1.90924704e-01 -5.03981709e-01 2.37755105e-01
1.02122331e+00 3.22297186e-01 6.92751944e-01 -1.49294674e+00
5.32541424e-02 -3.15079272e-01 -7.79957354e-01 -8.38145196e-01
4.74147737e-01 -7.70857871e-01 -3.63339186e-01 -1.09471750e+00
4.93918687e-01 1.26357973e-02 -1.59537762e-01 2.32908204e-01
-4.81801815e-02 5.09048045e-01 1.40842751e-01 2.81937391e-01
-1.50502980e-01 4.77451473e-01 1.35524595e+00 -1.65309608e-02
6.18234277e-02 3.40764403e-01 -8.91734421e-01 7.32484400e-01
6.96086049e-01 2.73820043e-01 -4.80138063e-01 -1.16536766e-01
-3.12084615e-01 2.17435986e-01 1.24596007e-01 -1.01131952e+00
2.09847942e-01 -2.28557035e-01 8.63336384e-01 -6.28738880e-01
8.49342048e-01 -1.01941752e+00 3.29867303e-01 2.53509551e-01
2.55910188e-01 5.40155470e-02 1.01224504e-01 4.96878058e-01
-4.63953853e-01 8.91075209e-02 1.01917088e+00 -9.07320976e-02
-7.61871040e-01 5.54510117e-01 -2.40230545e-01 -3.50880474e-01
1.26206625e+00 -6.90872669e-01 -1.07565917e-01 -2.95926481e-01
-6.47513866e-01 -8.65313187e-02 4.06561792e-01 5.59505820e-01
1.05079007e+00 -1.31850851e+00 -5.26972532e-01 6.80436254e-01
1.55815512e-01 -5.60227573e-01 6.13368630e-01 5.92833638e-01
-2.18545385e-02 7.09652662e-01 -3.62325400e-01 -5.85315645e-01
-1.81504130e+00 7.28064239e-01 3.25769037e-01 3.96244645e-01
-2.77570009e-01 6.88438058e-01 3.55259001e-01 -2.96998203e-01
2.53128469e-01 6.62519872e-01 -6.06619954e-01 1.27808541e-01
9.34685826e-01 5.50320208e-01 2.70626932e-01 -1.34071314e+00
-6.33830845e-01 9.81329024e-01 -1.85624644e-01 3.05631250e-01
1.20406449e+00 -6.34407774e-02 -3.53934050e-01 2.27002501e-01
9.53822553e-01 -2.61249840e-02 -9.82663572e-01 -2.81470329e-01
-3.33699817e-03 -7.72527874e-01 -1.87840298e-01 -5.41586578e-01
-1.20426130e+00 9.92811561e-01 7.90862322e-01 -3.47520560e-01
1.34899724e+00 -9.88791957e-02 2.41124213e-01 3.84729564e-01
8.82168353e-01 -6.83817625e-01 2.03342482e-01 3.98630649e-01
1.35821271e+00 -1.12686157e+00 1.80704638e-01 -9.37109292e-01
-3.10897529e-01 1.33130002e+00 7.69018590e-01 1.71378091e-01
9.12963271e-01 1.70354053e-01 1.17833421e-01 3.35556641e-02
-2.47869641e-01 -6.79521635e-02 5.96408069e-01 7.79849589e-01
8.03257883e-01 -8.04171637e-02 -7.17519745e-02 3.42627376e-01
-3.77307273e-02 -1.45464137e-01 7.23916516e-02 9.10401225e-01
-3.50717515e-01 -1.27354980e+00 -8.14303577e-01 4.97756541e-01
-6.44931555e-01 4.32501227e-01 -4.78267580e-01 6.10003293e-01
-1.36872351e-01 8.90409887e-01 -2.53045976e-01 -3.84418249e-01
5.25058866e-01 3.04080904e-01 9.71972704e-01 -5.52974522e-01
-1.68214172e-01 3.34491640e-01 -3.70923162e-01 -5.17761528e-01
-7.49194622e-01 -1.12165105e+00 -5.82065821e-01 -1.25418022e-01
-3.02623570e-01 -1.38457805e-01 4.56968546e-01 7.98694193e-01
2.86256582e-01 7.75779737e-03 9.14682090e-01 -1.11413002e+00
-6.21255815e-01 -5.72806895e-01 -7.69449949e-01 3.72098759e-02
2.52518475e-01 -1.06109917e+00 -2.07127109e-01 7.39341602e-03] | [13.173998832702637, 0.5372790694236755] |
86b6b950-8b07-4598-b2d0-cd2308062444 | principal-uncertainty-quantification-with | 2305.10124 | null | https://arxiv.org/abs/2305.10124v1 | https://arxiv.org/pdf/2305.10124v1.pdf | Principal Uncertainty Quantification with Spatial Correlation for Image Restoration Problems | Uncertainty quantification for inverse problems in imaging has drawn much attention lately. Existing approaches towards this task define uncertainty regions based on probable values per pixel, while ignoring spatial correlations within the image, resulting in an exaggerated volume of uncertainty. In this paper, we propose PUQ (Principal Uncertainty Quantification) -- a novel definition and corresponding analysis of uncertainty regions that takes into account spatial relationships within the image, thus providing reduced volume regions. Using recent advancements in stochastic generative models, we derive uncertainty intervals around principal components of the empirical posterior distribution, forming an ambiguity region that guarantees the inclusion of true unseen values with a user confidence probability. To improve computational efficiency and interpretability, we also guarantee the recovery of true unseen values using only a few principal directions, resulting in ultimately more informative uncertainty regions. Our approach is verified through experiments on image colorization, super-resolution, and inpainting; its effectiveness is shown through comparison to baseline methods, demonstrating significantly tighter uncertainty regions. | ['Michael Elad', 'Ehud Rivlin', 'Daniel Freedman', 'Yaniv Romano', 'Omer Belhasin'] | 2023-05-17 | null | null | null | null | ['colorization'] | ['computer-vision'] | [ 5.98830342e-01 3.64783525e-01 2.01624811e-01 -2.31582910e-01
-1.06154823e+00 -5.94466507e-01 4.36494082e-01 -1.05411597e-01
-2.45208010e-01 1.00939155e+00 2.87768155e-01 4.88058068e-02
-4.53905970e-01 -7.03552067e-01 -6.12578809e-01 -9.15822268e-01
-4.29420024e-02 2.99926251e-01 1.97685897e-01 2.37915844e-01
1.80674300e-01 4.79273111e-01 -1.31118393e+00 8.38545710e-02
9.78141427e-01 8.96179020e-01 1.17298089e-01 4.97429878e-01
1.39700351e-02 4.62686658e-01 -6.04706585e-01 -2.89006829e-01
2.56181985e-01 -3.71147662e-01 -5.52110612e-01 2.73224115e-01
2.29621828e-01 -5.71716011e-01 9.65993777e-02 1.27363896e+00
1.95436507e-01 1.40602186e-01 8.30537558e-01 -8.02911818e-01
-6.61191225e-01 5.30366957e-01 -9.62543786e-01 -8.54383968e-03
2.24217251e-01 -3.23714018e-01 8.78613472e-01 -8.34738553e-01
6.52187407e-01 1.24906909e+00 7.29584038e-01 4.13242191e-01
-1.66997075e+00 -2.60566890e-01 -8.07763413e-02 -1.16179194e-02
-1.63354921e+00 -3.28483313e-01 8.49647880e-01 -2.75049001e-01
3.70217353e-01 5.28153181e-01 3.14485431e-01 7.13714898e-01
2.90448517e-01 5.60383856e-01 1.42594206e+00 -5.66873968e-01
4.16980326e-01 2.17483371e-01 -1.07778274e-01 3.79862845e-01
3.55502069e-01 1.80834740e-01 -3.78614277e-01 -2.44039491e-01
1.26878977e+00 -2.31382519e-01 -6.31391883e-01 -6.69600129e-01
-1.07123303e+00 6.60534859e-01 4.11332428e-01 9.00169164e-02
-3.02985370e-01 1.70692325e-01 -1.23504668e-01 -3.29064608e-01
6.72151804e-01 2.64088929e-01 1.06782556e-01 -3.77638154e-02
-1.13043082e+00 1.06772974e-01 3.94923151e-01 9.22267139e-01
7.50265777e-01 6.79377615e-02 -3.15753102e-01 8.26893568e-01
3.46645653e-01 6.13759279e-01 -6.86615333e-02 -1.47805774e+00
7.70943016e-02 1.70659855e-01 5.06834388e-01 -1.09542835e+00
-1.61999539e-01 -5.59631050e-01 -7.36221552e-01 7.70766377e-01
4.24258292e-01 -6.02866113e-02 -9.91406024e-01 1.84657586e+00
3.65998149e-01 2.01908171e-01 1.06199067e-02 9.47202742e-01
1.36490941e-01 6.09588265e-01 -2.01722875e-01 -3.98436517e-01
1.31947446e+00 -2.82707363e-01 -8.92856598e-01 -4.94394749e-02
-1.83950245e-01 -5.87286234e-01 8.15051794e-01 4.95814532e-01
-1.02556896e+00 -2.30889574e-01 -9.56735432e-01 -4.73843515e-02
9.64478552e-02 -1.02409899e-01 2.95740217e-01 9.35081184e-01
-1.18864381e+00 7.56347656e-01 -7.59053469e-01 7.46871158e-02
4.88640934e-01 8.38177204e-02 -2.06426799e-01 -2.46700998e-02
-8.28328729e-01 7.14461803e-01 2.61401355e-01 2.04056472e-01
-6.60998940e-01 -9.14793849e-01 -9.21007216e-01 -1.03382111e-01
4.40759480e-01 -5.21208048e-01 8.99015009e-01 -5.78312159e-01
-1.40680265e+00 4.51972067e-01 -2.18379721e-01 -4.31223810e-01
8.67273986e-01 -3.58578026e-01 -1.21057369e-01 3.94796699e-01
1.39541313e-01 6.25767946e-01 1.22994244e+00 -1.82547534e+00
-2.18959719e-01 -4.22685176e-01 -1.02620609e-01 2.02471256e-01
3.80971730e-02 -2.86872894e-01 -5.89411318e-01 -7.77659953e-01
5.64578891e-01 -7.67673969e-01 -3.95118356e-01 1.78950444e-01
-4.68848020e-01 3.73719633e-01 5.95776379e-01 -7.67982066e-01
1.01809096e+00 -2.18981576e+00 2.80808389e-01 4.63186592e-01
5.12417853e-01 -3.07920873e-01 2.39738807e-01 -1.35104805e-01
-8.83509498e-03 1.43204451e-01 -8.94023359e-01 -3.22840363e-01
-9.22743604e-02 5.42216957e-01 -4.27552760e-01 5.12003303e-01
5.08641303e-01 6.52911842e-01 -1.04615593e+00 -6.89975858e-01
4.75257069e-01 7.75089800e-01 -3.75752211e-01 -1.32943332e-01
-1.65513381e-01 5.17158568e-01 -4.52682585e-01 5.63315809e-01
1.08377314e+00 -7.79080018e-02 -8.25877562e-02 -3.90336782e-01
-5.61802462e-02 -5.66516101e-01 -1.45670092e+00 1.65469754e+00
-4.12554413e-01 5.13916016e-01 2.89018780e-01 -4.14875209e-01
7.46489763e-01 9.82088074e-02 6.56919062e-01 -2.41936460e-01
-5.10535687e-02 1.55610796e-02 -3.65039498e-01 -6.88825846e-02
6.96793079e-01 -2.35175088e-01 8.25416744e-02 4.00576293e-01
-1.43100902e-01 -4.57241118e-01 5.54651991e-02 3.46182942e-01
6.44918144e-01 5.53752363e-01 1.91433489e-01 -3.62962723e-01
2.16347933e-01 -1.23429812e-01 5.67022979e-01 9.56016660e-01
-3.44989538e-01 1.24472940e+00 5.33809781e-01 1.73248559e-01
-1.08161306e+00 -1.64181888e+00 -7.42465019e-01 2.35095724e-01
3.09864104e-01 -2.23941177e-01 -1.06184697e+00 -3.65109175e-01
-8.33474845e-02 9.81964350e-01 -8.49666893e-01 2.04764962e-01
-2.45570749e-01 -9.22119975e-01 9.86857191e-02 3.51114333e-01
3.78325969e-01 -6.43563569e-01 -7.38763273e-01 1.39747873e-01
-3.26793164e-01 -1.17614114e+00 -2.46979639e-01 -1.07567392e-01
-9.18906569e-01 -8.94285202e-01 -1.12074924e+00 8.90547410e-02
8.52659762e-01 9.25681219e-02 1.09289360e+00 -5.82394719e-01
-4.82942849e-01 6.78906262e-01 -2.54313976e-01 -3.23929220e-01
-4.20679569e-01 -8.20813298e-01 -1.35183990e-01 7.83543885e-02
-2.50331312e-01 -5.84877074e-01 -5.57115734e-01 1.23277538e-01
-1.01222885e+00 -3.69608179e-02 5.43166637e-01 8.38705540e-01
8.88476551e-01 2.09048748e-01 1.91603884e-01 -8.47348154e-01
4.26656365e-01 -1.31341442e-01 -8.23763430e-01 2.17989326e-01
-6.91130340e-01 4.80131060e-01 6.56303167e-02 -8.86520147e-02
-1.67544234e+00 -1.61396280e-01 2.27030501e-01 -5.03329456e-01
-2.91999653e-02 2.03829169e-01 1.89639293e-02 -2.26011798e-01
8.10768783e-01 2.07426041e-01 8.55804905e-02 -3.76510859e-01
7.97181547e-01 2.99407721e-01 8.55062425e-01 -7.65701532e-01
6.44074559e-01 1.00923502e+00 1.94684282e-01 -8.89701962e-01
-8.73576880e-01 -3.43653888e-01 -4.71612126e-01 -3.79780263e-01
7.65547574e-01 -6.71674430e-01 -3.70761424e-01 7.89055824e-02
-1.06781077e+00 2.52228886e-01 -6.42716169e-01 5.05422950e-01
-7.03501463e-01 8.01819324e-01 -4.03690040e-01 -1.22149813e+00
-1.88028049e-02 -1.23897529e+00 1.32709873e+00 2.48142794e-01
-2.33745798e-01 -8.16273451e-01 -8.89188517e-03 2.36961752e-01
2.69550830e-01 5.11558175e-01 7.19416738e-01 2.53378451e-01
-7.52925873e-01 5.95626049e-02 -4.93454427e-01 5.08073092e-01
2.85270102e-02 -4.45066728e-02 -1.15831912e+00 4.55769598e-02
2.12401733e-01 1.28301218e-01 8.73899996e-01 8.87930870e-01
1.16728771e+00 -5.86887449e-03 -2.01560691e-01 4.38738793e-01
1.67147684e+00 -1.15542412e-02 8.95411968e-01 4.09907587e-02
4.44415301e-01 7.93362260e-01 4.73456979e-01 6.72135174e-01
-1.30067915e-01 5.74029326e-01 5.26368797e-01 1.99174610e-04
-3.17829043e-01 9.90678445e-02 6.33891225e-02 9.03504789e-02
-2.04011887e-01 -9.81091410e-02 -5.98232508e-01 7.15179145e-01
-1.66434944e+00 -7.69851089e-01 -1.57935277e-01 2.46904993e+00
8.01693916e-01 -1.90950055e-02 -2.67883033e-01 3.39032337e-02
7.70752013e-01 1.54298618e-01 -4.96942729e-01 -3.22915167e-02
-1.79067686e-01 1.60145029e-01 6.22746170e-01 9.33230817e-01
-8.58205795e-01 4.37426805e-01 7.06451941e+00 9.01983500e-01
-5.62321782e-01 3.77639346e-02 7.12501407e-01 -6.22722358e-02
-7.45298505e-01 1.89685505e-02 -4.79483515e-01 3.14327002e-01
3.95456672e-01 -1.06567070e-01 3.02221686e-01 5.87171137e-01
2.45678768e-01 -5.92184305e-01 -7.77811885e-01 1.01517439e+00
-2.89710388e-02 -1.13066339e+00 -2.71806307e-02 5.94168529e-02
9.20924842e-01 -1.78495586e-01 1.94478437e-01 -3.65317136e-01
3.29478651e-01 -7.33037949e-01 9.69844639e-01 8.83288383e-01
7.86294281e-01 -8.64783823e-01 5.18726528e-01 8.10438544e-02
-7.23941505e-01 2.58733243e-01 -5.08426964e-01 3.51501584e-01
4.13455367e-01 1.38751292e+00 -7.10703611e-01 8.54920864e-01
7.78414011e-01 3.54692131e-01 -2.24808946e-01 8.32565069e-01
-3.10936004e-01 3.27935219e-01 -5.73828638e-01 3.36113930e-01
-5.08969314e-02 -7.09723711e-01 9.53487992e-01 1.02207947e+00
4.50509757e-01 2.36446202e-01 -2.70024776e-01 1.60058415e+00
2.34468102e-01 -2.16078177e-01 -3.67521495e-01 2.08049566e-01
3.17974806e-01 1.15325344e+00 -9.89132524e-01 -3.16802040e-02
-1.12209901e-01 1.22108471e+00 -3.42826284e-02 5.89466274e-01
-8.69997859e-01 -2.34845623e-01 4.39407200e-01 -6.85795620e-02
9.95142385e-02 -1.87023982e-01 -6.36538267e-01 -9.72070694e-01
1.71788260e-01 -3.80520761e-01 1.82238400e-01 -1.04420638e+00
-1.12432790e+00 5.23812234e-01 4.21857744e-01 -1.15796041e+00
-3.87550771e-01 -6.25778139e-01 -1.86977029e-01 1.24486005e+00
-1.34078836e+00 -9.82004881e-01 -1.83826774e-01 1.96624801e-01
2.42303967e-01 4.02353793e-01 7.97104597e-01 -1.08995810e-01
-3.18516701e-01 9.94198322e-02 5.30512273e-01 -2.59392828e-01
5.58046043e-01 -1.46270633e+00 -1.97351780e-02 1.15971363e+00
8.88786912e-02 6.08903408e-01 1.15994787e+00 -6.79706573e-01
-1.03318560e+00 -8.27782512e-01 1.74111217e-01 -6.80527568e-01
5.32640457e-01 -1.90946683e-01 -8.95505607e-01 3.91915560e-01
-4.85794879e-02 5.08737527e-02 4.12517995e-01 -8.29791352e-02
-2.53697425e-01 8.06930661e-02 -1.51514280e+00 6.43114448e-01
7.07138956e-01 -5.03127337e-01 -3.56650591e-01 1.59220859e-01
6.21557236e-01 -5.27875721e-01 -9.05001104e-01 4.34767634e-01
5.84154725e-01 -1.29419899e+00 1.16335058e+00 1.94753736e-01
4.54644054e-01 -7.36853957e-01 -1.81528479e-01 -1.21896553e+00
-2.69068152e-01 -5.19703209e-01 -1.69897184e-01 1.10978019e+00
2.65488237e-01 -4.44278419e-01 7.72644222e-01 8.22616279e-01
1.31727513e-02 -4.60050315e-01 -1.05951989e+00 -7.48284757e-01
-1.55509874e-01 -7.91748941e-01 2.47039437e-01 6.32816017e-01
-1.90682739e-01 -2.35391051e-01 -4.10736918e-01 4.68796313e-01
1.36932898e+00 1.64878387e-02 1.93266407e-01 -1.04290450e+00
-4.84695971e-01 -2.58325279e-01 -3.71688813e-01 -6.77125096e-01
-1.99636206e-01 -2.83985168e-01 2.96926916e-01 -1.39955068e+00
3.29737991e-01 -4.17598873e-01 -1.11547329e-01 8.08205828e-02
-2.00071216e-01 4.99643683e-01 -8.20144936e-02 2.41323024e-01
-2.56939977e-01 5.59385240e-01 1.27297020e+00 2.57744044e-02
3.25140022e-02 -1.50638923e-01 -6.65636778e-01 7.49384463e-01
4.33747500e-01 -3.82940859e-01 -6.63315177e-01 -3.86226475e-01
1.37058482e-01 9.75069962e-03 5.92949986e-01 -9.86223638e-01
-2.34616041e-01 -1.34074111e-02 6.13008022e-01 -6.24552846e-01
5.28281510e-01 -9.43097115e-01 4.31579292e-01 9.29687992e-02
-9.83447731e-02 -5.10838628e-01 3.23848575e-01 9.31071997e-01
-6.93815053e-02 -5.13779998e-01 9.02374089e-01 -6.89723790e-02
-8.82298887e-01 -1.03635348e-01 -7.56224468e-02 -3.13462138e-01
1.00805271e+00 -2.95817763e-01 5.35522215e-02 -5.52446067e-01
-1.02048326e+00 -2.02524811e-01 5.53493559e-01 -2.60528438e-02
6.92964792e-01 -1.20297062e+00 -7.91112721e-01 1.65774986e-01
1.27962586e-02 6.15999848e-02 6.24394655e-01 6.36749625e-01
-5.44782519e-01 1.25129893e-01 -1.91078428e-02 -9.43404317e-01
-9.60274220e-01 3.48013341e-01 3.97022277e-01 -8.04206505e-02
-6.52478695e-01 9.56938922e-01 3.83316755e-01 -5.05677722e-02
-2.12970674e-01 -3.46549720e-01 -3.27102281e-02 -1.84821129e-01
6.72109306e-01 5.43924630e-01 -1.05616525e-01 -5.46620190e-01
-2.09712327e-01 7.00769007e-01 6.12802543e-02 -6.78074121e-01
1.10291374e+00 -5.30473471e-01 -2.68006802e-01 3.22636813e-01
8.65059853e-01 3.60332191e-01 -1.77099705e+00 -2.09150732e-01
2.93477997e-03 -7.34717309e-01 4.87602592e-01 -8.72979999e-01
-7.90825725e-01 7.10068464e-01 8.58750403e-01 1.41379297e-01
1.15794182e+00 5.52505217e-02 1.78379685e-01 -2.21212760e-01
4.70358968e-01 -9.24830437e-01 -3.12321663e-01 -9.91922570e-04
1.09779537e+00 -1.18995047e+00 3.83509427e-01 -8.58458221e-01
-4.85982776e-01 9.96154308e-01 1.25420287e-01 5.08977622e-02
5.58088362e-01 5.21403193e-01 -4.82597165e-02 -1.27727352e-02
-1.98964383e-02 -1.60444677e-02 4.29455698e-01 8.60511661e-01
2.91107059e-01 1.47934482e-01 -1.48992285e-01 3.78868520e-01
1.78660214e-01 2.84439996e-02 6.52769148e-01 8.21901500e-01
-4.22391653e-01 -9.30341542e-01 -9.94568765e-01 2.44840577e-01
-3.80017161e-01 -2.19236359e-01 5.00062928e-02 6.00517213e-01
-3.14671472e-02 8.28986824e-01 1.41750753e-01 1.50316313e-01
1.21378109e-01 -2.89643586e-01 7.94930637e-01 -3.86535555e-01
3.98395121e-01 4.77657348e-01 -6.32918552e-02 -7.46896863e-01
-3.62840921e-01 -7.51407206e-01 -1.29011512e+00 8.45957994e-02
-4.50948626e-01 1.73532188e-01 7.37247407e-01 7.24796414e-01
3.36414695e-01 5.59855580e-01 3.16225469e-01 -8.89167666e-01
-6.09947920e-01 -7.87902236e-01 -8.84953320e-01 1.50494248e-01
4.02045488e-01 -7.30686843e-01 -5.02547860e-01 2.05230653e-01] | [11.582904815673828, -2.042029619216919] |
09ef715d-b355-4415-836a-446a02ebd879 | incorporating-uncertainty-from-speaker | 2302.11763 | null | https://arxiv.org/abs/2302.11763v1 | https://arxiv.org/pdf/2302.11763v1.pdf | Incorporating Uncertainty from Speaker Embedding Estimation to Speaker Verification | Speech utterances recorded under differing conditions exhibit varying degrees of confidence in their embedding estimates, i.e., uncertainty, even if they are extracted using the same neural network. This paper aims to incorporate the uncertainty estimate produced in the xi-vector network front-end with a probabilistic linear discriminant analysis (PLDA) back-end scoring for speaker verification. To achieve this we derive a posterior covariance matrix, which measures the uncertainty, from the frame-wise precisions to the embedding space. We propose a log-likelihood ratio function for the PLDA scoring with the uncertainty propagation. We also propose to replace the length normalization pre-processing technique with a length scaling technique for the application of uncertainty propagation in the back-end. Experimental results on the VoxCeleb-1, SITW test sets as well as a domain-mismatched CNCeleb1-E set show the effectiveness of the proposed techniques with 14.5%-41.3% EER reductions and 4.6%-25.3% minDCF reductions. | ['Tianchi Liu', 'Kong Aik Lee', 'Qiongqiong Wang'] | 2023-02-23 | null | null | null | null | ['speaker-verification'] | ['speech'] | [-8.85093734e-02 1.03225028e-02 4.01239783e-01 -9.04491544e-01
-1.27702606e+00 -4.85617965e-01 2.36288100e-01 -6.11865055e-03
-4.80838567e-01 6.52374208e-01 3.35810333e-01 -3.34269911e-01
-5.20153232e-02 -1.08117178e-01 -4.74897027e-01 -7.12411702e-01
-1.26378655e-01 5.37561402e-02 -2.53876418e-01 3.71433526e-01
5.36115468e-02 2.73190290e-01 -1.11280453e+00 3.93085629e-02
5.87950051e-01 1.19412732e+00 1.60578992e-02 8.87548864e-01
3.24050546e-01 3.40177029e-01 -9.89312112e-01 -5.27236402e-01
1.74597606e-01 -2.07038224e-01 -7.35812783e-02 -9.48883295e-02
4.57810432e-01 -4.86823767e-01 -4.37921613e-01 1.28192270e+00
8.26109529e-01 2.16387168e-01 8.94993246e-01 -1.23655713e+00
-2.62184054e-01 9.54559922e-01 -3.38274628e-01 1.40307888e-01
1.68801561e-01 -1.23282224e-01 9.52474117e-01 -1.15908968e+00
2.38875881e-01 1.29052711e+00 6.34483516e-01 4.06246215e-01
-8.82919371e-01 -9.79249597e-01 -2.18058154e-01 4.01993960e-01
-1.76658750e+00 -1.02474582e+00 9.08006132e-01 -3.31471354e-01
1.04019630e+00 1.07839547e-01 -2.36765862e-01 9.05742764e-01
1.29727170e-01 6.20622873e-01 8.89309227e-01 -5.77544391e-01
2.69050181e-01 4.08245832e-01 3.19962859e-01 3.99025232e-01
1.55728618e-02 4.00934458e-01 -8.92423034e-01 -1.91715002e-01
1.50893509e-01 -6.19076967e-01 -4.34810579e-01 2.10996598e-01
-6.92730308e-01 5.95513225e-01 -1.14029676e-01 2.11661626e-02
-1.66760072e-01 2.55184080e-02 5.21348953e-01 2.47616559e-01
4.66747016e-01 -1.00046962e-01 -6.61369503e-01 -4.12916958e-01
-1.20895648e+00 -1.42459974e-01 7.66065359e-01 1.16988540e+00
1.95713967e-01 5.60346425e-01 -3.04449797e-01 1.10680544e+00
1.07383907e+00 6.99598491e-01 4.89992499e-01 -6.59702778e-01
7.55149007e-01 -1.91558838e-01 -1.36912232e-02 -7.66742229e-01
-1.33067176e-01 -5.07032692e-01 -7.12639749e-01 4.46586758e-01
2.44615674e-01 -4.71224040e-01 -8.52590561e-01 2.07363486e+00
1.81803629e-01 2.24741161e-01 4.03746873e-01 7.72458136e-01
4.64172214e-01 1.01652122e+00 -1.14848197e-01 -3.51691335e-01
1.31581986e+00 -7.36065030e-01 -1.07433271e+00 -6.32193983e-02
2.47121051e-01 -8.81186187e-01 6.50286853e-01 5.02650678e-01
-9.93093550e-01 -6.18174553e-01 -1.57130682e+00 2.36894593e-01
7.71177188e-02 5.51963925e-01 -4.52127844e-01 1.13598180e+00
-7.72681594e-01 5.95818162e-01 -5.78233778e-01 2.13002145e-01
-4.92414124e-02 3.83016825e-01 -2.82114089e-01 2.71174431e-01
-1.32372987e+00 8.11124384e-01 5.25548697e-01 2.21204743e-01
-7.41496384e-01 -5.52968025e-01 -8.48476648e-01 2.55245507e-01
1.79795418e-02 -5.73031120e-02 1.37288570e+00 -2.64151245e-01
-1.84737360e+00 1.94099292e-01 -3.00010085e-01 -4.48345453e-01
3.84437561e-01 -1.97262958e-01 -7.56160200e-01 -2.97112882e-01
-3.97365957e-01 3.71039689e-01 9.42022860e-01 -8.18425894e-01
-5.84551156e-01 -3.80432129e-01 -6.61204159e-01 1.98565364e-01
-3.47405970e-01 2.07736343e-01 -3.71723771e-01 -8.45942974e-01
3.02733541e-01 -8.44890535e-01 2.85114765e-01 -1.49336606e-01
-6.20866418e-01 -2.31492847e-01 8.74226332e-01 -1.25401807e+00
1.43319714e+00 -2.48080516e+00 3.00635658e-02 5.20129681e-01
-1.11626387e-01 3.15680385e-01 1.57651883e-02 3.95857394e-02
-3.11589204e-02 6.20110659e-03 -2.37043321e-01 -7.75547743e-01
4.68080431e-01 -1.48985997e-01 -2.24440411e-01 6.54765606e-01
3.02851856e-01 1.59656122e-01 -3.65778446e-01 -3.87495130e-01
2.44462222e-01 6.81218445e-01 -4.24483895e-01 2.44108096e-01
1.59345999e-01 -6.70469403e-02 1.52326062e-01 3.90134960e-01
9.20985460e-01 5.03230810e-01 3.24957483e-02 -3.32664609e-01
1.26367599e-01 5.72740018e-01 -1.52417207e+00 1.57304204e+00
-3.95681471e-01 1.03023446e+00 1.17771804e-01 -2.65273780e-01
1.04715955e+00 6.08317554e-01 -2.32975870e-01 -1.29392326e-01
4.63268846e-01 9.28625986e-02 1.74440727e-01 9.52357948e-02
5.84057689e-01 -1.78440809e-01 -1.49572507e-01 2.12209016e-01
4.39907312e-01 7.58790746e-02 -1.63062215e-01 -4.77250386e-03
7.33909070e-01 -2.58904129e-01 2.57221431e-01 -9.44632143e-02
7.20179200e-01 -9.24436212e-01 7.84885466e-01 4.09477055e-01
-5.74552536e-01 6.47728801e-01 3.03641587e-01 4.53011483e-01
-1.01707292e+00 -1.24261808e+00 -4.28880036e-01 6.57224417e-01
-3.21500868e-01 -3.72653604e-01 -8.15316498e-01 -4.14350301e-01
-4.80224118e-02 1.18837726e+00 -1.02692902e-01 -2.23346978e-01
-2.77869999e-01 -2.84448475e-01 1.01599693e+00 6.21485174e-01
2.74477154e-01 -4.52028662e-01 8.55528656e-03 2.72359937e-01
-3.53156030e-02 -1.22348821e+00 -7.67553091e-01 3.18907887e-01
-4.41522837e-01 -2.95717180e-01 -6.13660872e-01 -4.62432414e-01
3.84832352e-01 -3.41136873e-01 5.09856939e-01 -5.93738258e-01
2.11909741e-01 1.64289661e-02 -2.41062105e-01 -4.47168678e-01
-7.73995578e-01 -3.12935472e-01 7.85517693e-01 -3.92256342e-02
4.25329596e-01 -3.01908344e-01 -7.95157701e-02 4.97733563e-01
-4.64127868e-01 -4.78060484e-01 3.67355198e-01 8.80328834e-01
4.09229487e-01 3.36307824e-01 7.22595513e-01 -3.74549143e-02
7.65197217e-01 -1.60929516e-01 -9.68484163e-01 3.20209950e-01
-8.85249913e-01 3.59108090e-01 2.18497872e-01 -5.59532642e-01
-1.20736694e+00 -7.07302243e-02 -4.46740776e-01 -6.28006518e-01
5.56024797e-02 3.96589935e-01 -6.13484919e-01 1.85588166e-01
5.56185424e-01 -4.80486937e-02 -1.41818985e-01 -4.21511710e-01
3.59637171e-01 1.42753088e+00 6.45532489e-01 -3.56668681e-01
7.56373048e-01 -2.83607125e-01 -4.43049222e-01 -8.77421975e-01
-1.95059597e-01 -3.88512462e-01 -1.05302759e-01 -1.33318514e-01
6.94050789e-01 -9.43220735e-01 -6.92495167e-01 5.31850040e-01
-1.40471101e+00 3.77025545e-01 -8.25477093e-02 1.06018889e+00
-2.82911241e-01 5.64453006e-01 -6.74677551e-01 -1.37233543e+00
-6.92830205e-01 -1.24693596e+00 7.36872315e-01 2.47618973e-01
-3.60831082e-01 -4.65260804e-01 -5.26984222e-03 1.52902797e-01
4.49218482e-01 -3.62071604e-01 6.19108081e-01 -8.29594970e-01
-2.28086084e-01 -3.59740555e-01 -1.18481927e-01 9.95752871e-01
-1.21209979e-01 4.02341597e-02 -1.49582601e+00 -2.77922183e-01
3.00246477e-01 1.01602502e-01 5.40211558e-01 4.80441362e-01
5.97726583e-01 -3.37991536e-01 1.92948822e-02 3.66669297e-01
1.11441648e+00 6.60181344e-01 4.72932011e-01 -3.52678537e-01
2.27452472e-01 3.56851250e-01 4.89566058e-01 6.41429842e-01
-4.73092534e-02 9.02340829e-01 -7.66565949e-02 7.04017282e-01
-2.91219503e-01 -1.32516742e-01 8.90468776e-01 1.51332271e+00
5.04433215e-01 -6.72168612e-01 -7.27247775e-01 1.98033497e-01
-1.30876601e+00 -7.76656032e-01 9.50149298e-02 2.26027989e+00
8.13179433e-01 3.19360465e-01 -2.37867877e-01 5.11395097e-01
8.93291354e-01 9.16613080e-03 -5.04345238e-01 -6.48107767e-01
-4.92148511e-02 -1.08268641e-01 4.18518037e-01 8.61214101e-01
-9.46053624e-01 4.53202397e-01 5.66354227e+00 1.12104034e+00
-9.65010285e-01 1.27137959e-01 5.65736830e-01 -1.24162517e-01
-8.23901687e-03 -3.50948691e-01 -1.20125365e+00 5.66918433e-01
1.50357449e+00 -2.04274595e-01 3.06629241e-01 8.33105147e-01
1.51281178e-01 1.00823946e-01 -1.16598880e+00 1.19523406e+00
2.60036111e-01 -7.20657706e-01 -5.28310239e-01 -2.11949423e-02
3.24396223e-01 4.27758088e-03 3.53365988e-01 3.27182740e-01
2.59106085e-02 -6.19477332e-01 1.04894567e+00 3.84424895e-01
1.03814840e+00 -9.19192374e-01 8.92710567e-01 2.88795799e-01
-9.58759069e-01 -2.92235874e-02 -3.97757053e-01 4.90366608e-01
4.36092824e-01 7.70316541e-01 -1.43881679e+00 2.58623779e-01
3.73237252e-01 -1.40470251e-01 1.13553610e-02 9.26316381e-01
-4.52186406e-01 8.78259242e-01 -6.21628761e-01 -2.62430370e-01
-1.84774667e-01 2.70628221e-02 8.48118842e-01 1.43162572e+00
7.07466364e-01 -1.81973040e-01 -7.26874411e-01 8.11114132e-01
-2.76769459e-01 -1.25216737e-01 -1.48303539e-01 -1.65878966e-01
7.81319499e-01 7.84725308e-01 -9.80895013e-02 -1.50202334e-01
-1.18170522e-01 9.70100462e-01 8.95329863e-02 3.06591421e-01
-9.78310347e-01 -7.94978142e-01 7.50816107e-01 -5.52849054e-01
3.97954226e-01 -3.05500746e-01 -2.08400011e-01 -9.16741431e-01
1.40096307e-01 -7.74564743e-01 2.30004303e-02 -8.62013757e-01
-1.23796868e+00 9.96841371e-01 4.97902483e-02 -1.26339471e+00
-7.76329815e-01 -5.99597633e-01 -2.71668345e-01 1.26603913e+00
-1.25391936e+00 -5.33096135e-01 3.17056477e-01 1.90979153e-01
6.69151664e-01 -5.64005375e-01 8.40746999e-01 5.05033076e-01
-6.66532099e-01 1.39031839e+00 4.40763265e-01 1.91940501e-01
7.60716558e-01 -9.41335857e-01 3.54624331e-01 1.15948641e+00
2.18419492e-01 4.93996292e-01 9.06115055e-01 -3.90396118e-01
-1.05874228e+00 -8.63741815e-01 1.08682394e+00 -1.76659137e-01
4.26554263e-01 -5.08612156e-01 -8.51949275e-01 3.11794013e-01
3.22704107e-01 -2.06309602e-01 7.68588841e-01 1.78008333e-01
-6.11798108e-01 -1.41744673e-01 -1.49203968e+00 3.43534112e-01
5.11406541e-01 -8.57003093e-01 -8.01554143e-01 -2.97659844e-01
9.37918365e-01 -4.30698514e-01 -7.99848318e-01 3.51950526e-01
7.82783866e-01 -5.92284739e-01 7.11938679e-01 -1.54044196e-01
-1.52164891e-01 -6.14231825e-01 -9.18797076e-01 -1.38353181e+00
-1.01615161e-01 -5.22100449e-01 -1.81589648e-01 1.80467927e+00
8.10236454e-01 -3.81267935e-01 5.05360007e-01 8.35603118e-01
-2.15145037e-01 -1.90829977e-01 -1.59186053e+00 -9.17075217e-01
-2.13213325e-01 -1.06634772e+00 3.95398438e-01 4.88943160e-01
1.45131633e-01 3.43501151e-01 -5.29050529e-01 6.37559950e-01
7.57063270e-01 -7.80633569e-01 3.85215640e-01 -8.42916667e-01
-5.60601234e-01 -2.57325739e-01 -8.12077284e-01 -9.69777107e-01
1.74255729e-01 -6.13767207e-01 4.24925357e-01 -8.13771248e-01
-2.06196949e-01 -3.93142700e-02 -6.17480874e-01 4.49027643e-02
9.93195102e-02 -2.29391918e-01 3.37815672e-01 -1.49207652e-01
-1.37071133e-01 6.69288874e-01 3.34566027e-01 -2.16628864e-01
-2.77071178e-01 8.73639509e-02 -2.73417264e-01 6.24446750e-01
8.97488773e-01 -7.41588712e-01 -4.49682206e-01 -2.59057105e-01
-1.63236141e-01 3.23136359e-01 -1.27385393e-01 -1.27893507e+00
3.28794211e-01 5.40597200e-01 2.32831895e-01 -8.97280455e-01
7.01509893e-01 -9.19039786e-01 1.17344782e-01 6.55360445e-02
-4.74548280e-01 -2.17716649e-01 5.34914613e-01 6.85919523e-01
-2.37391427e-01 -5.88623941e-01 7.03452945e-01 7.70413876e-01
-2.43919268e-01 -2.33789578e-01 -5.12854099e-01 -3.47444683e-01
7.44955122e-01 3.50713059e-02 -1.17188305e-01 -5.30616581e-01
-6.82818055e-01 -5.22210263e-02 -2.48552799e-01 2.82631606e-01
7.79196680e-01 -1.40900481e+00 -1.04566646e+00 4.16286528e-01
1.21324092e-01 -4.59286630e-01 3.60258698e-01 3.24784845e-01
-2.65023839e-02 5.07933974e-01 2.27560416e-01 -5.76210022e-01
-1.64521444e+00 1.25024572e-01 1.93448082e-01 5.83842164e-03
4.40353602e-02 1.23507094e+00 -3.25965345e-01 -3.00678432e-01
7.89041460e-01 -5.55636823e-01 1.06737070e-01 -1.28486499e-01
6.07680023e-01 5.44587553e-01 2.68267572e-01 -8.16068053e-01
-6.22707844e-01 2.74489880e-01 -9.97163579e-02 -7.72266448e-01
7.63867319e-01 -3.64657968e-01 4.35363412e-01 4.71662253e-01
1.57659972e+00 2.53195673e-01 -9.97709751e-01 -4.37631667e-01
3.80176008e-02 -3.18103403e-01 4.13039714e-01 -9.59100842e-01
-7.84326851e-01 1.08208907e+00 1.15563917e+00 -2.28801087e-01
1.00077546e+00 -1.40923172e-01 5.20780861e-01 2.77249098e-01
1.16935909e-01 -1.13323379e+00 -4.24184114e-01 4.78092462e-01
1.16241252e+00 -1.03432548e+00 -1.24678172e-01 5.96109182e-02
-6.74572229e-01 1.05275261e+00 1.57527283e-01 4.26364720e-01
8.95739019e-01 6.45468771e-01 6.72424287e-02 4.35983330e-01
-7.72244990e-01 3.84599686e-01 4.16371167e-01 4.39323753e-01
3.58813286e-01 2.71003395e-01 -1.23341873e-01 9.10286427e-01
-3.54375154e-01 -2.67826557e-01 3.28624696e-01 5.11368155e-01
-3.83278728e-01 -9.23870087e-01 -6.65596187e-01 1.71624765e-01
-4.13037241e-01 -1.84218362e-01 -6.51658326e-02 1.52960584e-01
-1.83952227e-01 1.41500485e+00 -1.93647146e-02 -9.28075254e-01
3.14312518e-01 6.19812429e-01 7.25550503e-02 -3.71405669e-02
-2.20881611e-01 4.13757056e-01 2.59640276e-01 -5.74177802e-02
9.24918428e-02 -8.18998218e-01 -1.20450401e+00 -1.34323701e-01
-8.81578743e-01 2.07690194e-01 1.46637976e+00 6.61569595e-01
3.24710608e-01 6.40463293e-01 7.12828577e-01 -4.31431890e-01
-1.15315402e+00 -1.43410718e+00 -7.28116274e-01 -7.03981593e-02
4.77529436e-01 -4.81486976e-01 -8.81904781e-01 -4.66489000e-03] | [14.405673027038574, 6.12004280090332] |
643fdd26-38ed-45f9-826b-f61082d0af72 | sparse-array-selection-across-arbitrary | 2004.11637 | null | https://arxiv.org/abs/2004.11637v2 | https://arxiv.org/pdf/2004.11637v2.pdf | Sparse Array Selection Across Arbitrary Sensor Geometries with Deep Transfer Learning | Sparse sensor array selection arises in many engineering applications, where it is imperative to obtain maximum spatial resolution from a limited number of array elements. Recent research shows that computational complexity of array selection is reduced by replacing the conventional optimization and greedy search methods with a deep learning network. However, in practice, sufficient and well-calibrated labeled training data are unavailable and, more so, for arbitrary array configurations. To address this, we adopt a deep transfer learning (TL) approach, wherein we train a deep convolutional neural network (CNN) with data of a source sensor array for which calibrated data are readily available and reuse this pre-trained CNN for a different, data-insufficient target array geometry to perform sparse array selection. Numerical experiments with uniform rectangular and circular arrays demonstrate enhanced performance of TL-CNN on the target model than the CNN trained with insufficient data from the same model. In particular, our TL framework provides approximately 20% higher sensor selection accuracy and 10% improvement in the direction-of-arrival estimation error. | ['Ahmet M. Elbir', 'Kumar Vijay Mishra'] | 2020-04-24 | null | null | null | null | ['direction-of-arrival-estimation'] | ['audio'] | [ 3.05893779e-01 -2.19391376e-01 2.68503517e-01 -3.88746351e-01
-9.98246014e-01 -3.70721638e-01 -2.94531763e-01 1.75867409e-01
-3.58643174e-01 6.96588814e-01 1.13282517e-01 -1.99190125e-01
-3.76083344e-01 -9.07577217e-01 -9.60037410e-01 -8.94034863e-01
-2.14677244e-01 2.32647061e-01 -2.15110287e-01 -5.31610772e-02
1.22299962e-01 7.96587288e-01 -1.33322334e+00 1.40674040e-02
5.97576916e-01 1.90289938e+00 2.52185524e-01 3.89809936e-01
2.96245039e-01 4.07219350e-01 -4.29061085e-01 2.85050303e-01
5.25500774e-01 -9.67024788e-02 6.42332435e-02 -3.27639021e-02
4.80439901e-01 -3.20553780e-01 -2.08645359e-01 8.87219846e-01
1.00035107e+00 -8.27917531e-02 2.71585822e-01 -5.98756492e-01
-3.55139017e-01 6.74866021e-01 -5.53257287e-01 6.23553917e-02
1.72158003e-01 5.73211312e-02 4.94184852e-01 -1.23544204e+00
-1.68467119e-01 8.71531069e-01 1.00030637e+00 -1.52238041e-01
-8.04841518e-01 -1.09261394e+00 -4.29879278e-02 -4.40268099e-01
-1.63759112e+00 -6.16150081e-01 1.03047252e+00 -1.83582500e-01
7.03535080e-01 1.49992391e-01 6.41831636e-01 7.56631255e-01
1.98315948e-01 2.16683447e-01 3.89491946e-01 -4.36144382e-01
5.27751267e-01 -3.25514317e-01 -2.97445625e-01 5.63874364e-01
5.41859567e-01 2.21059576e-01 -4.85066444e-01 -9.85238105e-02
1.24889612e+00 4.45885770e-02 -2.45756701e-01 -6.10648513e-01
-1.27531695e+00 7.11325645e-01 8.23800623e-01 3.31741571e-01
-7.07473457e-01 2.14588240e-01 1.02062620e-01 3.64662558e-01
2.49848187e-01 1.05872309e+00 -3.38639349e-01 2.62926966e-01
-8.60513449e-01 -2.76481416e-02 7.78003573e-01 1.02391171e+00
9.57772672e-01 7.57402182e-01 1.45648345e-01 9.00623083e-01
1.17404155e-01 1.02870226e+00 1.99947879e-01 -9.06888902e-01
4.15977657e-01 5.13578832e-01 3.16917866e-01 -1.34046042e+00
-5.17692745e-01 -1.16678095e+00 -1.26861286e+00 1.71016842e-01
1.03874713e-01 -7.85733044e-01 -8.96013379e-01 1.62387836e+00
2.23123357e-01 3.03096980e-01 8.35133716e-02 1.14519060e+00
5.83831191e-01 6.35062099e-01 -2.89880931e-01 -1.83802024e-01
7.39844501e-01 -8.10388863e-01 -2.70279408e-01 -6.51093006e-01
3.14722747e-01 -7.72095621e-01 5.63584685e-01 3.22087079e-01
-9.34144795e-01 -6.23885572e-01 -1.59230685e+00 6.73854232e-01
-5.54018952e-02 3.96756470e-01 6.08140767e-01 4.11580741e-01
-8.50687385e-01 9.06255767e-02 -6.88480020e-01 -7.77640566e-02
4.02502447e-01 6.88388169e-01 -2.79577047e-01 -3.41199249e-01
-9.38028693e-01 3.27659726e-01 -2.36527156e-02 4.30491418e-01
-9.72276568e-01 -9.79330659e-01 -1.11904609e+00 1.26006365e-01
1.66636556e-01 -6.25358284e-01 8.70704293e-01 -9.34135377e-01
-1.52098525e+00 2.98113108e-01 4.43425849e-02 -4.54224914e-01
-1.49794668e-01 -4.89063740e-01 -3.44558626e-01 -1.04931556e-02
3.05995464e-01 1.78561717e-01 1.11881948e+00 -1.43411601e+00
-5.58771968e-01 -3.58729988e-01 -1.79920688e-01 1.01199120e-01
-3.79604250e-01 -4.25208509e-01 -2.70148188e-01 -5.97041011e-01
5.09556532e-01 -8.50423396e-01 -9.09483492e-01 -6.65976480e-02
-3.77341241e-01 3.21326464e-01 4.31151181e-01 -2.90230840e-01
1.22496426e+00 -2.34979677e+00 7.01505342e-04 7.00960279e-01
2.84736007e-01 1.39435500e-01 -3.30269396e-01 3.11750114e-01
-1.68372676e-01 -3.66408587e-01 -4.03501153e-01 -1.09578900e-01
-3.45076382e-01 -1.93009555e-01 -1.59580842e-01 5.13165772e-01
1.74521595e-01 5.37999332e-01 -7.29601502e-01 -9.63301733e-02
3.54286522e-01 2.25151047e-01 -5.88771105e-01 4.85889584e-01
-2.70646494e-02 7.26051390e-01 -8.01925063e-01 9.63962674e-01
7.10852742e-01 -6.30988359e-01 4.14848328e-03 -5.46243548e-01
-3.02311778e-01 -2.06875965e-01 -1.39329982e+00 1.68839419e+00
-8.54332507e-01 4.56819713e-01 3.44171464e-01 -1.32202590e+00
1.13486576e+00 3.70027535e-02 9.41966414e-01 -9.68228161e-01
3.46762478e-01 5.30411184e-01 4.77218479e-02 -3.13924402e-01
3.89507897e-02 4.08377796e-01 -3.04944485e-01 2.96904832e-01
-4.39818859e-01 -1.14421643e-01 -4.32319194e-01 -4.20548588e-01
1.39217854e+00 -6.04596376e-01 3.42790067e-01 -3.79957259e-01
4.49740082e-01 5.55971870e-03 5.82967579e-01 8.65869343e-01
3.67071033e-01 7.20987856e-01 -3.02956253e-01 -7.46546924e-01
-1.08894646e+00 -9.47416425e-01 -1.64435610e-01 9.52461481e-01
2.93657452e-01 1.68722510e-01 -3.85985017e-01 8.89769346e-02
1.61007285e-01 3.61898899e-01 -4.14759994e-01 -6.66276366e-02
-1.21750844e+00 -5.20577669e-01 1.60425618e-01 6.49347425e-01
7.44301498e-01 -8.66910756e-01 -1.15396631e+00 4.97481614e-01
3.13371450e-01 -1.19740748e+00 -1.29960820e-01 3.65076661e-01
-6.41201377e-01 -8.99904609e-01 -6.42326355e-01 -1.02065277e+00
8.76327515e-01 2.44793132e-01 1.20559359e+00 -2.30754279e-02
-1.48058563e-01 3.72027874e-01 -3.25959235e-01 -5.68978667e-01
3.01182032e-01 3.02243859e-01 5.38846366e-02 1.76778808e-01
1.23585343e-01 -7.52643347e-01 -8.53497267e-01 2.60159701e-01
-5.35403728e-01 -1.99987397e-01 9.93105292e-01 8.76767993e-01
6.90697670e-01 -7.14501785e-03 9.12358940e-01 -4.16066825e-01
3.51257622e-01 -7.06372380e-01 -9.30424750e-01 -3.03628240e-02
-3.44839215e-01 -8.52123871e-02 6.62064493e-01 -3.57753187e-01
-8.39121997e-01 5.54868042e-01 -2.75631428e-01 -6.46539807e-01
9.93174687e-02 9.51612890e-01 -1.21621557e-01 -6.39013588e-01
7.20664620e-01 5.74978068e-02 -7.94988032e-03 -1.62870735e-01
-2.30685890e-01 5.06888866e-01 4.52206969e-01 -4.00637448e-01
8.13966334e-01 4.64680523e-01 1.89775065e-01 -8.99660707e-01
-8.42663229e-01 -2.96288580e-01 -2.40388826e-01 -1.14215493e-01
4.99344766e-01 -1.20328498e+00 -3.42513204e-01 5.01577795e-01
-8.75972509e-01 -5.51946223e-01 -9.41826105e-02 6.39008760e-01
-1.07339196e-01 -9.68780518e-02 7.21187517e-03 -7.30169475e-01
-5.40939927e-01 -1.10856175e+00 1.48606014e+00 2.00503290e-01
9.36780944e-02 -8.87858689e-01 -2.73926079e-01 -2.67344981e-01
9.24696267e-01 4.82433170e-01 7.55990684e-01 -5.09708941e-01
-9.18756902e-01 -7.09589124e-01 -1.31699517e-01 1.32355109e-01
3.87609273e-01 -5.65202594e-01 -6.57187581e-01 -5.64502776e-01
2.59926707e-01 -2.50380635e-01 5.87637067e-01 9.03100193e-01
1.34692311e+00 -6.48116991e-02 -6.32198691e-01 9.03696954e-01
1.76062679e+00 4.03214067e-01 3.43448073e-01 9.00025293e-02
7.71622658e-01 1.58985332e-01 6.75011814e-01 8.37354362e-01
-4.61901017e-02 3.62290949e-01 6.39331579e-01 -3.15559477e-01
2.19927415e-01 6.19545653e-02 -2.78729796e-01 8.41477811e-01
4.54537004e-01 -3.69215190e-01 -9.40888941e-01 7.20462382e-01
-1.67999840e+00 -2.24791870e-01 5.48899770e-01 2.39182544e+00
5.88021278e-01 1.73883066e-02 -4.60404813e-01 3.57899209e-03
6.21782303e-01 1.87416375e-01 -9.51312363e-01 2.55277663e-01
-1.99100643e-01 3.77846062e-01 8.34622979e-01 4.91754502e-01
-1.20703077e+00 4.97995406e-01 6.37472725e+00 4.96179104e-01
-1.71250618e+00 -2.91300654e-01 6.65386379e-01 -8.97420570e-02
-3.05029720e-01 -5.01000881e-01 -7.09909558e-01 3.51011455e-01
6.85495973e-01 8.03173929e-02 2.26240858e-01 8.98505211e-01
-1.74469069e-01 -1.17089525e-01 -1.40521395e+00 1.52426100e+00
1.17994584e-01 -1.57788384e+00 -3.63749743e-01 -1.68284193e-01
7.63367176e-01 1.86474267e-02 3.25441539e-01 6.72428310e-02
1.53156132e-01 -1.31830323e+00 3.08074474e-01 3.35555941e-01
8.01025689e-01 -4.93901163e-01 9.40552652e-01 2.05719650e-01
-1.37602103e+00 -4.39270467e-01 -4.16806102e-01 -1.59732744e-01
1.74453110e-02 1.08240938e+00 -6.86626792e-01 4.32283700e-01
8.56118917e-01 6.64160490e-01 -2.48423845e-01 1.11148000e+00
3.77901196e-01 5.75538814e-01 -7.73213744e-01 -1.34843677e-01
2.61202306e-01 -2.32175961e-01 6.52184665e-01 7.97806084e-01
9.87365186e-01 1.89967498e-01 3.94531578e-01 5.52390575e-01
2.14343406e-02 -2.06743240e-01 -8.75690579e-01 2.85683274e-01
1.12537503e+00 1.05901814e+00 -3.14140975e-01 6.79564998e-02
-4.10872191e-01 3.13573360e-01 1.60070181e-01 5.99649787e-01
-5.68916261e-01 -3.18331122e-01 5.57581067e-01 1.08362690e-01
4.99871403e-01 -3.46110642e-01 -7.36925662e-01 -5.52653849e-01
2.61059571e-02 -8.83210897e-01 1.42821282e-01 -6.16843581e-01
-1.32063460e+00 6.05719686e-01 -1.40318319e-01 -1.41532266e+00
-3.90452772e-01 -5.64749360e-01 -4.09622103e-01 8.63430738e-01
-1.34530854e+00 -1.03764892e+00 -7.27149427e-01 3.28072071e-01
6.26083732e-01 -5.49571931e-01 7.37293959e-01 7.30898559e-01
-4.44865197e-01 7.14516819e-01 -3.90410013e-02 3.00480366e-01
4.22085673e-01 -7.48305798e-01 7.99374655e-02 8.75728786e-01
-4.26328212e-01 4.73270893e-01 5.74238420e-01 -4.66062009e-01
-1.95885265e+00 -1.38371289e+00 1.27858251e-01 -1.36455327e-01
3.91650051e-01 -5.64915538e-02 -6.86480105e-01 5.94621837e-01
2.02120602e-01 2.55910188e-01 5.26175678e-01 -2.13037372e-01
4.35640551e-02 -5.69984794e-01 -8.73488784e-01 4.48128194e-01
1.07139695e+00 -1.32173449e-01 -1.95723977e-02 3.12904328e-01
7.95846164e-01 -7.25572467e-01 -9.51803505e-01 1.04707384e+00
6.01311862e-01 -7.85321534e-01 1.31735396e+00 4.22286168e-02
3.42760980e-01 -2.05486178e-01 -6.67518079e-01 -1.47163999e+00
-4.17474955e-01 -2.85154521e-01 1.69988647e-01 6.04239702e-01
5.77481210e-01 -8.36141586e-01 1.03053284e+00 3.24309468e-01
-6.98567033e-01 -1.06612360e+00 -9.73969698e-01 -4.72941697e-01
-4.96906713e-02 -3.56945515e-01 1.01833582e+00 7.86708891e-01
-5.77356339e-01 3.79491597e-01 -2.43540093e-01 9.25591469e-01
7.48638213e-01 5.47774494e-01 7.67624497e-01 -1.01550221e+00
2.57464759e-02 -2.64305770e-01 -1.31811589e-01 -1.45233941e+00
2.62230039e-02 -2.30046213e-01 3.83545518e-01 -1.42361939e+00
-6.97073817e-01 -1.15134799e+00 -5.49360633e-01 1.23623289e-01
8.67167115e-02 1.86524376e-01 -4.17182475e-01 6.35826960e-02
-5.96752763e-01 5.46844065e-01 1.17700613e+00 -3.02378356e-01
-1.26026571e-01 1.75212100e-01 -6.58661962e-01 4.67184514e-01
7.60714412e-01 -1.40374735e-01 -3.91106218e-01 -1.09805620e+00
5.13991833e-01 3.16142023e-01 1.88394457e-01 -1.73535097e+00
6.09947741e-01 9.50740948e-02 7.31725574e-01 -5.89572310e-01
2.97332525e-01 -1.18864119e+00 3.45749885e-01 4.66478020e-01
-1.84653178e-01 -5.87627962e-02 3.76538277e-01 4.74058479e-01
-4.00469124e-01 2.39805132e-02 6.75653636e-01 -2.93247774e-03
-9.11356688e-01 4.84087199e-01 -1.35670125e-01 -2.68711835e-01
7.64595389e-01 -3.15171391e-01 1.11615457e-01 -4.52489913e-01
-3.99261862e-01 2.15603098e-01 1.56564429e-01 1.03326134e-01
1.02015185e+00 -1.52601635e+00 -7.40157783e-01 8.06452811e-01
1.92480072e-01 6.24760032e-01 8.69372338e-02 5.83563924e-01
-6.33452177e-01 3.69472325e-01 -1.50131226e-01 -8.86650026e-01
-4.51501459e-01 3.63031477e-01 5.73539853e-01 2.39915490e-01
-2.71758437e-01 1.13771248e+00 1.98304281e-01 -4.70257849e-01
2.66284287e-01 -3.70013863e-01 -2.19031852e-02 -4.91288245e-01
2.58560717e-01 1.78333431e-01 2.17755377e-01 -2.56529033e-01
-4.45305467e-01 9.80994821e-01 4.65185702e-01 2.71684647e-01
1.34679008e+00 -3.18623781e-02 -1.54122233e-01 1.07861824e-01
1.21376050e+00 3.64652336e-01 -1.34056640e+00 -4.94182169e-01
-3.20378929e-01 -4.48246807e-01 2.02107444e-01 -3.52078348e-01
-1.40703738e+00 5.15376389e-01 9.30648804e-01 2.04206586e-01
1.40469420e+00 -1.24748521e-01 8.11309516e-01 6.94820225e-01
7.18058825e-01 -6.89452529e-01 2.09024101e-01 4.27465558e-01
1.17401588e+00 -1.54354370e+00 -9.25206840e-02 -4.71718818e-01
-6.98259771e-02 1.05866337e+00 8.88866007e-01 -5.55727899e-01
1.14886963e+00 7.74159968e-01 1.61840305e-01 -4.28187579e-01
-2.40994230e-01 1.25781149e-01 9.18422341e-02 5.07620096e-01
2.80109435e-01 2.09798831e-02 4.81160223e-01 5.68327010e-01
-2.45145634e-01 -2.01804847e-01 -1.55390963e-01 1.00854850e+00
-5.98469436e-01 -5.30063868e-01 -3.84590000e-01 8.64210963e-01
-1.35212749e-01 -2.06264779e-01 2.55650520e-01 4.84629035e-01
-5.50712831e-02 9.43095446e-01 3.80699545e-01 -5.24169147e-01
4.53784555e-01 -6.56859457e-01 2.23540619e-01 -7.38589466e-01
-2.56590575e-01 1.94448411e-01 4.19990458e-02 -6.04603171e-01
-2.27045640e-01 -3.46787542e-01 -1.02126074e+00 1.67781323e-01
-1.38303980e-01 2.41870806e-01 5.63622713e-01 7.93391466e-01
6.81075215e-01 8.13685477e-01 9.71915245e-01 -1.15211928e+00
-5.59840143e-01 -6.65385127e-01 -4.00149167e-01 -1.56535372e-01
7.46507406e-01 -6.20582581e-01 -2.63432831e-01 -1.81189999e-01] | [6.568282604217529, 1.1548293828964233] |
d06fbae5-6c00-4506-944c-465d0dbcacdb | an-end-to-end-visual-audio-attention-network | 2003.00832 | null | https://arxiv.org/abs/2003.00832v1 | https://arxiv.org/pdf/2003.00832v1.pdf | An End-to-End Visual-Audio Attention Network for Emotion Recognition in User-Generated Videos | Emotion recognition in user-generated videos plays an important role in human-centered computing. Existing methods mainly employ traditional two-stage shallow pipeline, i.e. extracting visual and/or audio features and training classifiers. In this paper, we propose to recognize video emotions in an end-to-end manner based on convolutional neural networks (CNNs). Specifically, we develop a deep Visual-Audio Attention Network (VAANet), a novel architecture that integrates spatial, channel-wise, and temporal attentions into a visual 3D CNN and temporal attentions into an audio 2D CNN. Further, we design a special classification loss, i.e. polarity-consistent cross-entropy loss, based on the polarity-emotion hierarchy constraint to guide the attention generation. Extensive experiments conducted on the challenging VideoEmotion-8 and Ekman-6 datasets demonstrate that the proposed VAANet outperforms the state-of-the-art approaches for video emotion recognition. Our source code is released at: https://github.com/maysonma/VAANet. | ['Tengfei Xing', 'Kurt Keutzer', 'Hua Chai', 'Yang Gu', 'Sicheng Zhao', 'Runbo Hu', 'Pengfei Xu', 'Jufeng Yang', 'Yunsheng Ma'] | 2020-02-12 | null | null | null | null | ['video-emotion-recognition'] | ['computer-vision'] | [-1.19627126e-01 -4.63104516e-01 -4.52517159e-02 -6.00449026e-01
-7.33346522e-01 -1.71934262e-01 2.98053771e-01 -9.51604918e-02
-3.88263643e-01 3.05479020e-01 3.03523690e-01 -4.16507907e-02
3.11278731e-01 -2.04605475e-01 -6.83558047e-01 -5.30070364e-01
-1.09400496e-01 -3.49629462e-01 -3.42413843e-01 6.05956046e-03
2.02687800e-01 1.44447148e-01 -1.50568068e+00 4.99521434e-01
6.63159430e-01 1.78410494e+00 -2.45746836e-01 8.88230562e-01
8.00841898e-02 1.24325442e+00 -2.86213398e-01 -2.76850730e-01
-5.13462648e-02 -5.62891185e-01 -7.50783026e-01 5.80985211e-02
1.32591322e-01 -2.61982322e-01 -3.71024579e-01 8.49115014e-01
7.11603105e-01 3.75782400e-01 4.89638329e-01 -1.65969288e+00
-8.12685192e-01 1.99739695e-01 -6.80447698e-01 2.72591412e-01
3.93162042e-01 2.13549167e-01 1.01996481e+00 -1.17198992e+00
2.58225650e-01 1.23551536e+00 6.53634310e-01 5.52149892e-01
-7.88690507e-01 -8.25436115e-01 4.41369921e-01 6.86572194e-01
-1.49233603e+00 -5.47944546e-01 1.10934293e+00 -5.66239357e-01
9.42259312e-01 1.72480956e-01 7.82123268e-01 1.43713629e+00
1.39750540e-01 1.09863460e+00 7.16999710e-01 -1.68561324e-01
1.20795600e-01 3.55134194e-04 -1.67663291e-01 6.16928875e-01
-5.27585328e-01 -1.97424874e-01 -7.82862484e-01 -2.89673172e-03
3.87418330e-01 1.41621064e-02 -3.78573179e-01 -1.14660226e-01
-9.78185773e-01 7.41235077e-01 4.63062525e-01 1.75721608e-02
-6.55648947e-01 3.82676005e-01 9.79720116e-01 1.74632341e-01
8.41929138e-01 -4.97794785e-02 -4.92496669e-01 -5.21614671e-01
-8.38398516e-01 -5.11044301e-02 3.43358725e-01 7.52699375e-01
3.13700557e-01 1.74034685e-01 -4.88996834e-01 7.79059529e-01
4.81749892e-01 5.63179776e-02 5.33971846e-01 -8.02516520e-01
2.56953299e-01 3.36309522e-01 -1.47099659e-01 -1.09841800e+00
-2.67047137e-01 -1.91751227e-01 -9.20025706e-01 3.18253599e-02
-3.09419662e-01 -5.10112047e-01 -7.79445291e-01 1.81326628e+00
2.61551410e-01 4.95887816e-01 -3.01585775e-02 1.30186832e+00
1.15443289e+00 9.55479980e-01 2.76287854e-01 -2.25292981e-01
1.33086109e+00 -1.30441773e+00 -9.16190863e-01 -7.04693869e-02
2.36354306e-01 -6.25107229e-01 1.16862106e+00 5.19231737e-01
-1.17627919e+00 -6.73865438e-01 -9.07571495e-01 -3.41513664e-01
-2.94491589e-01 4.44670528e-01 4.06064481e-01 3.00765634e-01
-9.86180365e-01 2.59599328e-01 -7.32977390e-01 -1.42390579e-01
6.51501894e-01 1.28930897e-01 -2.13080734e-01 3.54422897e-01
-1.28501654e+00 4.24572706e-01 1.06862389e-01 6.24008298e-01
-1.08627570e+00 -5.42245924e-01 -1.01246274e+00 2.83501923e-01
3.39893162e-01 -5.01748800e-01 1.39464974e+00 -1.57730424e+00
-1.78432167e+00 7.22798347e-01 -2.81371355e-01 -2.32753754e-01
3.77000153e-01 -5.19959152e-01 -4.07266825e-01 3.22910041e-01
-1.40337169e-01 7.09426105e-01 8.98434043e-01 -1.05670595e+00
-4.47116256e-01 -1.70415282e-01 -8.00260231e-02 3.45027745e-01
-5.78602433e-01 4.54817891e-01 -6.63314521e-01 -7.36868560e-01
-3.54569852e-01 -7.76078641e-01 -5.14212102e-02 1.04045331e-01
-5.71838975e-01 -2.35677361e-01 9.68175709e-01 -7.68659472e-01
1.35163319e+00 -2.46510005e+00 2.50135392e-01 1.00184433e-01
3.38527590e-01 2.10028261e-01 -1.85406238e-01 1.73748523e-01
-4.08984154e-01 1.66127205e-01 4.94080707e-02 -6.11898243e-01
2.17788070e-01 -4.09644485e-01 -1.26426846e-01 4.31696951e-01
6.04507387e-01 8.76799583e-01 -7.92770565e-01 -5.68330705e-01
2.22024143e-01 9.37082052e-01 -6.95125222e-01 4.83278632e-01
-1.80910751e-02 5.30433774e-01 -2.78521597e-01 7.93459773e-01
4.59908813e-01 -2.54741043e-01 -1.11589618e-01 -2.65662611e-01
-2.86915272e-01 2.73457497e-01 -7.76005328e-01 1.77032793e+00
-5.71515441e-01 1.08154440e+00 1.92629263e-01 -9.96428311e-01
8.56168866e-01 7.59204268e-01 6.78040385e-01 -5.96812785e-01
5.87528706e-01 -1.15173236e-01 -3.48412901e-01 -8.64953101e-01
5.20288527e-01 6.86964914e-02 -1.60988763e-01 1.44557744e-01
2.66765177e-01 4.13853645e-01 -1.29510298e-01 4.74618077e-02
8.12317848e-01 1.69333279e-01 2.21555904e-01 8.52414668e-02
6.26786292e-01 -3.76042575e-01 8.69773924e-01 1.96040988e-01
-5.60666025e-01 7.85592973e-01 7.45705009e-01 -2.99601793e-01
-7.13998377e-01 -6.07311308e-01 1.80874705e-01 1.27336621e+00
8.07263609e-03 -6.25660837e-01 -7.17217445e-01 -6.66320324e-01
-2.93793172e-01 2.69565254e-01 -8.47440302e-01 -2.44812980e-01
-7.73503035e-02 -2.94977099e-01 4.76444423e-01 7.02375710e-01
6.61216378e-01 -1.34904957e+00 -8.21179986e-01 1.06238112e-01
-4.88515645e-01 -1.04991412e+00 -7.08352149e-01 7.87497312e-02
-3.02392215e-01 -8.61055613e-01 -8.24232161e-01 -7.97125399e-01
2.86563843e-01 1.75770856e-02 9.17627215e-01 -1.24114640e-01
-2.74376571e-01 5.84224284e-01 -6.72822535e-01 -4.87023711e-01
4.25293863e-01 -4.64793667e-03 -1.33846641e-01 5.54661036e-01
5.31459332e-01 -5.33569634e-01 -8.43385875e-01 -1.59733258e-02
-6.11300170e-01 2.76230555e-02 3.73482764e-01 9.16885018e-01
6.40621960e-01 -9.45722982e-02 6.07972980e-01 -3.40795696e-01
7.32389331e-01 -7.76843190e-01 -1.93833187e-01 7.89505690e-02
-1.77812427e-01 -3.03481489e-01 6.50377393e-01 -6.34574115e-01
-1.13276470e+00 2.36945242e-01 -3.40488523e-01 -1.14467978e+00
-2.13981912e-01 7.03384459e-01 -3.59786034e-01 1.22607186e-01
2.18472838e-01 1.62538052e-01 -2.82089710e-01 -2.17155248e-01
2.38470718e-01 9.01054144e-01 4.61467117e-01 -3.17068905e-01
2.58528799e-01 2.16399670e-01 -4.95337814e-01 -7.06013143e-01
-6.35020375e-01 -5.40051281e-01 -2.84127504e-01 -6.07316434e-01
1.19302404e+00 -1.27401173e+00 -9.90192056e-01 6.13742113e-01
-1.25296581e+00 -4.26614791e-01 4.14581299e-02 6.13014340e-01
-6.52019382e-01 9.86715928e-02 -7.48995841e-01 -1.08991730e+00
-6.15178466e-01 -1.18892634e+00 1.26888025e+00 3.02793205e-01
-2.37348840e-01 -7.11244881e-01 2.68480415e-03 2.56258726e-01
2.95952648e-01 4.62997586e-01 3.84409934e-01 -3.09324056e-01
-2.21802428e-01 2.61949096e-03 -3.52891028e-01 5.28872848e-01
-1.83567032e-01 3.17238510e-01 -1.25726402e+00 -1.07879482e-01
-1.66846544e-01 -7.68980443e-01 9.86414313e-01 5.19078314e-01
1.74152493e+00 -3.43335032e-01 1.00280978e-01 8.51395190e-01
1.18565905e+00 3.71921390e-01 7.65405655e-01 1.58171222e-01
8.42539668e-01 4.33798492e-01 7.66616285e-01 9.90149975e-01
6.22408450e-01 5.22249222e-01 5.56087971e-01 -3.03429544e-01
2.97545433e-01 -7.84976333e-02 5.28404117e-01 9.16343808e-01
-7.65187517e-02 -4.29399461e-01 -7.28386521e-01 5.55161715e-01
-2.01292491e+00 -9.98897970e-01 6.09536134e-02 1.65435231e+00
6.95506811e-01 -1.13759309e-01 1.27232060e-01 1.96164444e-01
6.62797213e-01 3.97375464e-01 -6.17842197e-01 -7.15179741e-01
1.70474976e-01 1.19735099e-01 -1.15695976e-01 1.85871035e-01
-1.43405771e+00 7.40948439e-01 4.96157598e+00 6.14134252e-01
-1.63518775e+00 1.85831159e-01 9.49573159e-01 -4.08619165e-01
-1.65373161e-02 -4.06441540e-01 -1.76637396e-01 6.21289790e-01
9.08452988e-01 -9.65670943e-02 3.61646444e-01 9.27283645e-01
5.66464901e-01 1.84746027e-01 -8.74697804e-01 1.41867423e+00
2.78340220e-01 -9.01320398e-01 -2.31687248e-01 -2.58457303e-01
4.20254529e-01 -2.27610767e-01 2.78951854e-01 4.39870059e-01
-3.84378910e-01 -9.56564784e-01 1.10150158e+00 6.17226601e-01
9.04567778e-01 -9.48980689e-01 6.60323501e-01 -2.44395852e-01
-1.62033331e+00 -1.92384303e-01 5.17198294e-02 -5.50302444e-04
2.27085248e-01 4.08431500e-01 -1.59984872e-01 4.00775433e-01
1.20895994e+00 1.10540175e+00 -1.84800133e-01 9.56131339e-01
-1.49105594e-01 6.95628583e-01 1.04620099e-01 -4.79799658e-02
3.92201424e-01 6.51217476e-02 4.13421154e-01 1.52989042e+00
3.59614193e-01 3.08316618e-01 -4.16414253e-02 7.55906463e-01
-3.81147146e-01 2.28123114e-01 -3.53743851e-01 -2.43731812e-01
2.35650733e-01 1.42843604e+00 -3.43506515e-01 -3.41359913e-01
-5.05592942e-01 1.28012848e+00 2.89056897e-01 4.82944846e-01
-1.39492702e+00 -8.83127868e-01 9.27412510e-01 -4.51747745e-01
5.25084436e-01 9.38124210e-02 2.00499911e-02 -1.28545821e+00
2.11965621e-01 -8.87017727e-01 3.39886427e-01 -1.15347183e+00
-1.19868767e+00 8.85485113e-01 -4.44021434e-01 -1.39083374e+00
-2.22657919e-02 -5.11855721e-01 -9.44535136e-01 6.32340252e-01
-1.48899674e+00 -9.33581471e-01 -7.36144543e-01 8.33004653e-01
7.18827009e-01 9.87256914e-02 5.63244581e-01 6.92793190e-01
-9.58248556e-01 7.75556028e-01 -2.50983149e-01 2.90330648e-01
8.24142456e-01 -9.41740453e-01 1.24253668e-02 7.34540820e-01
-2.07655355e-01 1.70627952e-01 4.02441174e-01 -1.69164822e-01
-1.34395683e+00 -1.29555261e+00 7.78049707e-01 6.07393794e-02
4.79054630e-01 -4.77035105e-01 -8.08662653e-01 6.86750054e-01
6.98455870e-01 2.14944258e-01 8.00832748e-01 2.31424961e-02
-4.06086087e-01 -1.62128910e-01 -9.23342347e-01 4.19620156e-01
8.75779569e-01 -7.48037219e-01 -9.44166034e-02 -4.04138267e-02
7.28290737e-01 -3.39603394e-01 -7.85463572e-01 4.88906771e-01
8.13762486e-01 -9.88569260e-01 7.38899350e-01 -5.97958386e-01
7.56605208e-01 -1.62066221e-01 -5.03724553e-02 -1.33830094e+00
-3.46682966e-01 -7.04941094e-01 -2.23884657e-01 1.32630610e+00
3.00109267e-01 -2.83368170e-01 3.37025076e-01 4.21846598e-01
-3.18934023e-01 -9.96632695e-01 -7.68049836e-01 -3.58894140e-01
-4.02670890e-01 -6.09812319e-01 3.09835941e-01 1.08912098e+00
2.34902278e-01 4.41644073e-01 -6.94227338e-01 2.14100536e-02
2.35012949e-01 -1.61464036e-01 4.36832458e-01 -8.90955150e-01
-1.79939829e-02 -6.35509849e-01 -5.06719232e-01 -8.98671150e-01
3.39862406e-01 -4.48255271e-01 4.32437867e-01 -1.32278609e+00
3.00176203e-01 7.29821473e-02 -6.13030732e-01 5.62176168e-01
-1.22438140e-01 3.37378532e-01 1.85206354e-01 -3.06396425e-01
-1.11991644e+00 1.16499007e+00 8.37316751e-01 -1.86173171e-01
-2.09359765e-01 -4.57810193e-01 -6.19077861e-01 5.92813194e-01
1.05616879e+00 -1.90840214e-01 -4.22911525e-01 -3.85820955e-01
1.33006111e-01 4.89152186e-02 5.19440591e-01 -1.05934334e+00
1.32860407e-01 -4.22348920e-03 5.10175765e-01 -6.95825517e-01
5.92565656e-01 -7.24995494e-01 -1.45428002e-01 6.24029078e-02
-5.80667913e-01 2.25366384e-01 4.12387133e-01 4.47225392e-01
-6.72168493e-01 3.08694333e-01 6.18550003e-01 1.54450729e-01
-8.16390932e-01 6.36264622e-01 -6.35492742e-01 7.66682550e-02
1.05264974e+00 8.61624703e-02 -1.33876592e-01 -6.60325706e-01
-7.02164888e-01 4.38713521e-01 -5.56451716e-02 6.02536321e-01
9.94746029e-01 -1.72255373e+00 -6.05088711e-01 1.12262033e-01
2.42056400e-01 -2.45161623e-01 7.04419851e-01 1.11219180e+00
-2.48405173e-01 2.81525135e-01 -3.03483844e-01 -5.14495909e-01
-1.51101589e+00 4.68576640e-01 4.42309618e-01 7.11021721e-02
-2.36306205e-01 1.07152510e+00 2.43310988e-01 -1.34316087e-01
6.68372989e-01 -3.41076314e-01 -3.17328572e-01 2.08634630e-01
5.96281350e-01 1.87173530e-01 -2.19921067e-01 -7.96848059e-01
-5.95173001e-01 4.63433325e-01 1.56877369e-01 -1.43571228e-01
1.31402218e+00 -2.06010148e-01 8.43547732e-02 5.88826895e-01
1.58259487e+00 -3.77963722e-01 -1.38955498e+00 -2.24847365e-02
-4.57785398e-01 -3.88542891e-01 3.73051435e-01 -5.63846707e-01
-1.52034271e+00 1.28031373e+00 7.64615297e-01 4.71049994e-02
1.67913032e+00 -1.61134228e-01 9.24391806e-01 3.27169187e-02
-2.75827050e-01 -1.16563642e+00 3.69837016e-01 5.47009110e-01
1.14890695e+00 -1.31514335e+00 -4.05847192e-01 -6.02394193e-02
-1.02339184e+00 9.71141577e-01 9.34372127e-01 9.47157145e-02
8.02344680e-01 -1.33092806e-03 1.58554971e-01 -1.76561654e-01
-1.11837173e+00 -3.02550465e-01 4.43962544e-01 2.03307003e-01
7.60088384e-01 -7.72505179e-02 -9.23506171e-02 1.04016542e+00
2.41597325e-01 2.74791211e-01 1.09768555e-01 9.05916810e-01
-1.44261410e-02 -5.30032635e-01 -1.59814090e-01 1.48535728e-01
-6.17199719e-01 -2.74755895e-01 -4.66338694e-01 4.22594070e-01
2.46010259e-01 9.84443188e-01 2.50862867e-01 -8.62585366e-01
2.23969370e-01 1.65548339e-01 1.77340254e-01 -2.29731217e-01
-7.24489808e-01 3.14297348e-01 5.14864996e-02 -8.52614641e-01
-5.55086434e-01 -5.78403950e-01 -1.22153521e+00 -1.78401709e-01
-9.46325585e-02 1.64663225e-01 5.27757645e-01 5.18805861e-01
7.56732047e-01 8.47948849e-01 9.08401132e-01 -1.19797277e+00
4.03264835e-02 -1.06211734e+00 -3.56962353e-01 3.68342549e-01
5.89425266e-01 -5.81643522e-01 -3.21644157e-01 2.00491622e-01] | [13.291775703430176, 4.967169761657715] |
8320c0c6-bb8d-4440-b5a7-4e447fc895ca | hybrid-message-passing-with-performance | null | null | http://openaccess.thecvf.com//content/CVPR2021/html/Song_Hybrid_Message_Passing_With_Performance-Driven_Structures_for_Facial_Action_Unit_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Song_Hybrid_Message_Passing_With_Performance-Driven_Structures_for_Facial_Action_Unit_CVPR_2021_paper.pdf | Hybrid Message Passing With Performance-Driven Structures for Facial Action Unit Detection | Message passing neural network has been an effective method to represent dependencies among nodes by propagating messages. However, most of message passing algorithms focus on one structure and the messages are estimated by one single approach. For the real-world data, like facial action units (AUs), the dependencies may vary in terms of different expressions and individuals. In this paper, we propose a novel hybrid message passing neural network with performance-driven structures (HMP-PS), which combines complementary message passing methods and captures more possible structures in a Bayesian manner. Particularly, a performance-driven Monte Carlo Markov Chain sampling method is proposed for generating high performance graph structures. Besides, the hybrid message passing is proposed to combine different types of messages, which provide the complementary information. The contribution of each type of message is adaptively adjusted along with different inputs. The experiments on two widely used benchmark datasets, i.e., BP4D and DISFA, validate that our proposed method can achieve the state-of-the-art performance. | ['Qiang Ji', 'Wenming Zheng', 'Zijun Cui', 'Tengfei Song'] | 2021-06-19 | null | null | null | cvpr-2021-1 | ['action-unit-detection', 'facial-action-unit-detection'] | ['computer-vision', 'computer-vision'] | [ 7.28303790e-02 3.16674225e-02 -2.01873258e-01 -6.20636225e-01
-2.09957644e-01 3.47522587e-01 7.66622722e-01 1.66590333e-01
-2.56418526e-01 8.00104022e-01 -6.04103692e-02 5.07534444e-02
-9.59435254e-02 -1.11949694e+00 -7.22028673e-01 -8.03480387e-01
-2.55465448e-01 3.72926742e-01 5.85104406e-01 -1.87127724e-01
-1.90580145e-01 2.55334973e-01 -1.51293612e+00 1.52345240e-01
4.60802495e-01 7.16710627e-01 -6.48288503e-02 5.80283046e-01
-4.28363621e-01 9.34904873e-01 -6.60271585e-01 -5.84485471e-01
-2.31923565e-01 -5.01712859e-01 -4.90271240e-01 1.05983997e-02
-9.46103483e-02 -5.47920108e-01 -5.45325696e-01 1.05458069e+00
3.61777276e-01 -1.52661517e-01 6.35069668e-01 -1.71652639e+00
-2.32470453e-01 1.11312544e+00 -8.70088339e-01 -1.19583897e-01
1.37378663e-01 9.83193144e-02 6.58293724e-01 -4.60690439e-01
4.95369136e-01 1.86643708e+00 3.27710181e-01 4.86674786e-01
-1.22852278e+00 -1.04076624e+00 5.66439331e-01 3.92677844e-01
-1.67697716e+00 -2.48381764e-01 9.98713851e-01 -1.68422714e-01
4.22172815e-01 -1.30898906e-02 7.69570589e-01 1.27360058e+00
4.78351027e-01 9.51168597e-01 1.02244020e+00 -1.17131405e-01
3.35891813e-01 -9.28937197e-02 2.35739470e-01 9.02474880e-01
3.44040155e-01 -2.69369464e-02 -5.94980896e-01 -3.65066141e-01
6.64889753e-01 1.75157413e-02 -1.84353247e-01 7.30232224e-02
-8.78423631e-01 7.75663793e-01 3.44029099e-01 1.22344308e-01
-2.63379663e-01 4.96885777e-01 4.32309180e-01 -1.33962587e-01
2.42152974e-01 -7.15750158e-01 -2.54642606e-01 -8.06339364e-03
-8.49174321e-01 1.79652825e-01 1.05003428e+00 9.32954967e-01
9.65485990e-01 1.02034397e-01 -5.33351600e-01 7.70683765e-01
9.36172485e-01 3.76010209e-01 2.95540750e-01 -3.39818478e-01
1.86161950e-01 7.22530067e-01 -8.57117027e-02 -1.50651789e+00
-6.42706454e-01 -4.02720720e-01 -1.33739662e+00 3.28632928e-02
8.98059011e-02 -4.28398371e-01 -1.13151217e+00 2.07310581e+00
7.79756665e-01 7.15419233e-01 -1.32032618e-01 5.83018303e-01
1.16950011e+00 1.08850086e+00 4.12802070e-01 -2.34672680e-01
1.36350775e+00 -6.08259439e-01 -7.40602255e-01 1.31344065e-01
3.07906061e-01 -6.78099811e-01 4.32848096e-01 3.78250450e-01
-8.15453529e-01 -6.27273381e-01 -1.20086265e+00 3.13640863e-01
-7.15200603e-02 2.86651284e-01 7.71585941e-01 7.08916605e-01
-8.24280083e-01 4.11812723e-01 -1.01626039e+00 -1.10905580e-01
5.52627385e-01 4.52600062e-01 -1.33411005e-01 9.46147740e-02
-1.53171968e+00 1.66633308e-01 8.12171400e-01 3.60290915e-01
-1.28168440e+00 -1.15874410e-01 -6.98834121e-01 2.12678462e-01
3.02528024e-01 -5.73014140e-01 1.07703853e+00 -9.28153932e-01
-1.78700197e+00 2.21515387e-01 -1.42224692e-02 -2.24238083e-01
4.79426622e-01 2.55313247e-01 -3.91345292e-01 1.29285768e-01
-5.33769429e-01 8.25364351e-01 9.98369157e-01 -1.24635446e+00
-6.10693455e-01 -2.51376867e-01 3.18227857e-01 8.96064043e-02
-5.22167444e-01 -3.25102312e-03 -5.96703649e-01 -2.75362700e-01
-1.71625808e-01 -7.91656017e-01 -1.73810750e-01 4.72146422e-02
-5.61784089e-01 -5.79960823e-01 8.45514119e-01 -2.25718111e-01
1.28233349e+00 -1.98666024e+00 7.89527819e-02 3.65466952e-01
4.17638063e-01 6.39710873e-02 -1.29124358e-01 4.92140740e-01
2.21986040e-01 2.17217598e-02 -9.74562988e-02 -4.75254953e-01
1.83418451e-03 5.47112286e-01 1.37526363e-01 4.28551704e-01
3.79680723e-01 5.61180294e-01 -7.57642567e-01 -9.39792097e-01
1.13715266e-03 8.52521539e-01 -2.05505535e-01 1.69919893e-01
-4.11598593e-01 3.12029302e-01 -5.89667320e-01 3.67637247e-01
1.07678759e+00 -3.97928566e-01 3.51048380e-01 -1.29517794e-01
2.61094898e-01 -3.41794252e-01 -1.59589100e+00 1.40577865e+00
-2.43950829e-01 1.93236306e-01 1.28470913e-01 -6.73226833e-01
1.28383982e+00 2.36414552e-01 2.04918057e-01 -2.15326816e-01
7.06044197e-01 -3.23003054e-01 7.95463920e-02 -3.67783487e-01
3.08502436e-01 1.27820615e-02 2.16373280e-01 4.05994236e-01
1.93583354e-01 3.19337457e-01 3.63836914e-01 3.52558225e-01
1.04417372e+00 7.58604705e-02 1.38343096e-01 -7.03652576e-02
8.12941611e-01 -6.19006217e-01 8.57123494e-01 7.42161751e-01
-1.26320317e-01 3.15726787e-01 8.62555265e-01 -1.65415302e-01
-5.63655674e-01 -8.01564097e-01 -1.17199443e-01 9.53059971e-01
3.79797161e-01 -5.45916378e-01 -8.62112820e-01 -5.53125143e-01
-2.84579664e-01 3.93537819e-01 -7.13987887e-01 -2.87341416e-01
-1.50403380e-01 -1.22273362e+00 6.50043845e-01 3.13331783e-01
7.61232555e-01 -9.58956718e-01 -3.26783657e-01 4.31186944e-01
9.80834197e-03 -9.17331040e-01 1.68927759e-02 -3.12219322e-01
-7.05767334e-01 -8.57348263e-01 -6.83973491e-01 -5.92855453e-01
8.88971686e-01 -1.59622788e-01 8.84302855e-01 4.79616284e-01
-1.56639278e-01 1.70567706e-02 -3.73320103e-01 -4.74711508e-01
-5.09285212e-01 2.72443444e-01 -2.03782067e-01 8.28717530e-01
3.50850523e-01 -6.34052217e-01 -6.05849683e-01 2.14010805e-01
-1.18677175e+00 3.50212634e-01 8.63339961e-01 8.52774978e-01
2.94908106e-01 2.89721161e-01 4.94408339e-01 -1.02930391e+00
7.76233315e-01 -7.54484236e-01 -5.03093421e-01 2.03063235e-01
-3.27184856e-01 1.00543045e-01 6.24095082e-01 -6.92316711e-01
-1.50685894e+00 -5.09636551e-02 -3.76435459e-01 -1.20382205e-01
-3.68385226e-01 6.07042015e-01 -2.02100039e-01 -3.92322317e-02
3.76628488e-01 1.46005705e-01 -3.28686237e-02 -1.97303936e-01
1.98641792e-01 6.07226491e-01 5.13340719e-02 -9.22094822e-01
5.63941121e-01 4.60346043e-01 3.99921715e-01 -7.30467677e-01
-2.62473404e-01 1.74711943e-01 -2.10469589e-01 -5.28628767e-01
6.36378109e-01 -7.33315408e-01 -1.13918793e+00 1.14721155e+00
-1.43405187e+00 -7.98716918e-02 4.00373936e-01 5.85318148e-01
-1.78093255e-01 2.47468159e-01 -1.01997662e+00 -1.01236248e+00
-2.24264741e-01 -1.30915189e+00 8.35005939e-01 8.56591165e-01
1.77418053e-01 -8.36428642e-01 1.00989670e-01 -2.63527006e-01
2.70607740e-01 3.25547010e-01 8.45844865e-01 -5.06095350e-01
-7.60039330e-01 -7.81999677e-02 -4.79286194e-01 1.20644070e-01
1.57548524e-02 6.21620476e-01 -8.01244497e-01 -2.74145871e-01
-4.18490589e-01 -4.13162321e-01 6.09680176e-01 1.45738468e-01
1.26700342e+00 -2.12191343e-01 -3.98029774e-01 3.17798913e-01
1.31264150e+00 2.12000757e-01 6.97142184e-01 -2.45409578e-01
6.19495869e-01 4.73152608e-01 2.19935641e-01 7.74115145e-01
5.18056452e-01 3.55561137e-01 5.82287312e-01 -1.34215549e-01
8.25106204e-02 -4.77368146e-01 1.64324299e-01 1.04341781e+00
6.47483990e-02 -4.81097192e-01 -6.94913507e-01 2.43977413e-01
-2.14655805e+00 -6.37097776e-01 -1.93665817e-01 1.87868321e+00
1.04442501e+00 4.95516747e-01 -1.18765436e-01 6.49037063e-02
9.34066951e-01 4.39230233e-01 -5.15764713e-01 -1.94508508e-01
1.74548090e-01 1.45803839e-01 2.95642257e-01 3.32478166e-01
-7.15573549e-01 6.64128780e-01 5.65179825e+00 1.24958718e+00
-9.40474987e-01 6.51054010e-02 8.81742418e-01 2.58975655e-01
-2.76333630e-01 1.18005693e-01 -1.16291213e+00 7.57451177e-01
8.94317031e-01 -1.85536429e-01 4.67302389e-02 6.08880997e-01
1.53101563e-01 -1.92347988e-01 -8.43667448e-01 9.27443802e-01
7.85120577e-03 -1.12726045e+00 3.62863988e-01 -2.10211888e-01
5.31327426e-01 -2.63493717e-01 -3.23276699e-01 2.85382181e-01
6.54395580e-01 -9.28889990e-01 4.58656460e-01 7.87554145e-01
3.56406093e-01 -9.73551273e-01 7.98165739e-01 5.19752324e-01
-1.36232984e+00 1.40149862e-01 -4.22315478e-01 -9.42634195e-02
2.47149259e-01 7.56378174e-01 -5.09387732e-01 8.45516026e-01
6.04599297e-01 5.75375795e-01 -5.17132938e-01 9.21818852e-01
-4.82541561e-01 9.98375833e-01 -6.20780170e-01 -7.98202753e-01
2.17170805e-01 -5.63975191e-03 3.29775304e-01 1.13231003e+00
7.99891539e-03 7.14221373e-02 2.05743745e-01 8.53568256e-01
-2.07030386e-01 2.95508772e-01 -1.71719164e-01 1.25977144e-01
5.79259694e-01 1.51549542e+00 -8.03811610e-01 -4.19029534e-01
-5.00633776e-01 4.45150077e-01 4.53169733e-01 3.87172103e-01
-1.04659188e+00 -5.90558290e-01 2.86290884e-01 -1.15050606e-01
1.64228946e-01 -2.67345339e-01 3.08665901e-01 -8.37974072e-01
-2.63149291e-01 -6.33536994e-01 4.09231901e-01 -6.03491247e-01
-1.25509644e+00 5.34645438e-01 3.86967003e-01 -8.59863758e-01
1.01039067e-01 -4.14424598e-01 -8.21266174e-01 6.59912407e-01
-1.38625860e+00 -1.23042750e+00 -5.67002892e-01 4.82695371e-01
3.01067352e-01 5.92616678e-04 5.48948407e-01 4.04398501e-01
-9.62041914e-01 7.21606076e-01 -1.89353228e-01 2.32466489e-01
4.00220454e-01 -6.05844498e-01 1.16710454e-01 6.48724198e-01
4.04986031e-02 6.02718711e-01 5.25198758e-01 -7.65246570e-01
-1.25367606e+00 -1.02966154e+00 3.61461550e-01 3.75861585e-01
5.16662955e-01 -3.30648839e-01 -9.80923057e-01 6.14739358e-01
3.58306646e-01 -1.74403876e-01 3.71604532e-01 4.96501513e-02
6.94768354e-02 -4.68478918e-01 -1.03883886e+00 7.74810493e-01
8.69120717e-01 7.08819628e-02 -9.20857266e-02 6.80352151e-02
6.87252581e-01 -6.37400031e-01 -7.83014059e-01 3.73189777e-01
5.23304701e-01 -9.41911578e-01 5.44235170e-01 -4.61199582e-01
4.47431296e-01 -5.53197205e-01 2.81696022e-01 -1.15894246e+00
-2.97106683e-01 -5.85048735e-01 -2.39800587e-01 1.61680055e+00
1.16608977e-01 -7.60127068e-01 9.58096504e-01 4.30575639e-01
4.80327606e-01 -1.07944369e+00 -9.31641102e-01 -3.79234403e-01
-3.30602676e-01 -3.70056838e-01 1.09556401e+00 5.54265559e-01
-4.51308608e-01 3.19511384e-01 -5.71121752e-01 1.27174124e-01
8.23596954e-01 -4.05211419e-01 9.88687992e-01 -1.23953080e+00
-4.47589874e-01 -1.85805902e-01 -7.61660635e-01 -1.27980435e+00
2.70654202e-01 -7.17732430e-01 1.19280085e-01 -1.38068449e+00
4.04967070e-01 -4.82326359e-01 -4.55529094e-01 4.28070813e-01
-2.16748476e-01 -1.41632199e-01 -1.00585505e-01 -1.17655359e-01
-4.42989379e-01 8.80337417e-01 1.10031199e+00 -2.88159639e-01
3.57391648e-02 -1.42576963e-01 -4.23163027e-01 9.40569282e-01
8.54924679e-01 -7.34916091e-01 -7.56776094e-01 -4.22359467e-01
3.22418749e-01 1.13312513e-01 4.51692164e-01 -1.06115437e+00
4.98079777e-01 -3.59223545e-01 4.17730719e-01 -8.21695864e-01
4.22259539e-01 -7.31832027e-01 4.53183502e-01 4.53649282e-01
-2.93745726e-01 -1.42444968e-01 8.66922364e-03 9.32361364e-01
-1.17848791e-01 -2.37911642e-01 7.91176677e-01 -3.86134274e-02
-4.68287647e-01 7.35026479e-01 -3.13069224e-01 -3.09765711e-02
1.01200449e+00 8.44345465e-02 -3.75217199e-01 -4.93150145e-01
-3.89410645e-01 5.60811520e-01 -8.32070187e-02 3.23232681e-01
7.19041109e-01 -1.45530391e+00 -9.69891727e-01 1.55099958e-01
1.72511265e-01 4.07047430e-03 5.47222912e-01 5.47826350e-01
-4.66040432e-01 -2.29574054e-01 -1.97878242e-01 -8.17494690e-01
-1.33551776e+00 -3.12258638e-02 2.62252122e-01 -2.70804495e-01
-3.75424862e-01 8.65099907e-01 1.15244566e-02 -1.32914081e-01
3.73308420e-01 -7.09477514e-02 -4.99985427e-01 -1.45056888e-01
6.37797773e-01 2.81073362e-01 -3.51155967e-01 -6.20822787e-01
-1.75417677e-01 3.48056495e-01 -3.95758599e-01 -1.63159609e-01
9.88938570e-01 -2.83551146e-03 -2.67605484e-01 5.57156146e-01
1.04207194e+00 -4.03659254e-01 -1.23928261e+00 -3.60511392e-01
-4.45215821e-01 -5.02515554e-01 -4.86886390e-02 -3.34940225e-01
-1.33511877e+00 8.27335298e-01 4.59753275e-01 2.62505949e-01
9.63317871e-01 -1.14392899e-01 8.63956928e-01 2.63636470e-01
4.60564822e-01 -1.00423133e+00 -1.65534258e-01 1.93248108e-01
5.24863720e-01 -9.31353033e-01 4.57273647e-02 -4.43170428e-01
-1.32657722e-01 1.21207035e+00 9.82984483e-01 1.23955458e-02
9.89308357e-01 3.49419832e-01 -1.25297502e-01 -1.78089261e-01
-9.68102872e-01 1.66775867e-01 -1.94377258e-01 3.46830666e-01
2.86476821e-01 2.09968448e-01 -8.67414296e-01 6.21989429e-01
1.27334058e-01 3.36796433e-01 5.13253927e-01 9.26747739e-01
-3.15820664e-01 -1.25726736e+00 -2.70129919e-01 5.95897555e-01
-1.79823071e-01 1.84968382e-01 -8.94664302e-02 6.26777589e-01
2.49817923e-01 9.79125917e-01 3.70241627e-02 -4.97046709e-01
-5.59334904e-02 -1.54291570e-01 4.26124096e-01 -3.36437225e-01
-2.44831651e-01 -2.66459644e-01 1.16334960e-01 -3.47553402e-01
-5.59562027e-01 -2.80522823e-01 -1.42032969e+00 -4.92083967e-01
-3.43445212e-01 3.62163819e-02 6.53245389e-01 8.72727931e-01
3.31720233e-01 6.91422582e-01 6.77467048e-01 -6.26383960e-01
-3.28019470e-01 -9.53082085e-01 -6.13399744e-01 3.88437301e-01
-5.11979498e-02 -8.58305097e-01 -2.80352205e-01 -1.48888350e-01] | [7.375450134277344, 6.155650615692139] |
bd73ef24-2ee3-45e4-b6b5-3753f7163b18 | improved-techniques-for-training-single-image | 2003.11512 | null | https://arxiv.org/abs/2003.11512v2 | https://arxiv.org/pdf/2003.11512v2.pdf | Improved Techniques for Training Single-Image GANs | Recently there has been an interest in the potential of learning generative models from a single image, as opposed to from a large dataset. This task is of practical significance, as it means that generative models can be used in domains where collecting a large dataset is not feasible. However, training a model capable of generating realistic images from only a single sample is a difficult problem. In this work, we conduct a number of experiments to understand the challenges of training these methods and propose some best practices that we found allowed us to generate improved results over previous work in this space. One key piece is that unlike prior single image generation methods, we concurrently train several stages in a sequential multi-stage manner, allowing us to learn models with fewer stages of increasing image resolution. Compared to a recent state of the art baseline, our model is up to six times faster to train, has fewer parameters, and can better capture the global structure of images. | ['Tobias Hinz', 'Stefan Wermter', 'Matthew Fisher', 'Oliver Wang'] | 2020-03-25 | null | null | null | null | ['single-image-generation'] | ['computer-vision'] | [ 5.18953145e-01 2.87458479e-01 1.45064220e-01 -2.77230918e-01
-9.31402385e-01 -5.86040676e-01 9.07147586e-01 -3.94423366e-01
-3.51787746e-01 6.78071618e-01 2.21080229e-01 -2.31709898e-01
2.57901073e-01 -7.77348340e-01 -8.01320851e-01 -5.35913169e-01
1.76428795e-01 7.07615137e-01 3.54523510e-01 7.43784104e-03
2.41525397e-01 4.02926028e-01 -1.67278922e+00 3.07248086e-01
5.37576675e-01 4.98136222e-01 5.91089606e-01 7.76595354e-01
1.96957082e-01 7.21257985e-01 -6.61929727e-01 -2.87295818e-01
2.40703344e-01 -7.42932081e-01 -8.96895885e-01 4.85648304e-01
8.11317623e-01 -5.75470388e-01 -2.07613707e-02 6.75849199e-01
5.84759712e-01 6.72722086e-02 4.87863481e-01 -1.05561459e+00
-5.30781269e-01 2.62255669e-01 -5.53377748e-01 1.49586156e-01
1.79536194e-01 2.65987813e-01 6.93927109e-01 -5.90009749e-01
8.16230953e-01 1.26234865e+00 3.84448975e-01 7.95360863e-01
-1.59765458e+00 -5.93256474e-01 5.05498685e-02 -3.22279811e-01
-1.10533488e+00 -5.99809885e-01 4.96523768e-01 -4.10581738e-01
9.60406959e-01 1.01925281e-03 7.63143420e-01 1.29550707e+00
4.84189391e-02 7.79889584e-01 1.25766289e+00 -5.80635846e-01
1.93971395e-02 1.64695591e-01 -5.29661596e-01 6.18272007e-01
2.41518646e-01 1.65772110e-01 -3.23036104e-01 -1.63708304e-04
1.20511293e+00 -1.82916328e-01 -1.61211506e-01 -5.14524877e-01
-1.27062535e+00 8.74265075e-01 3.91484827e-01 2.16815412e-01
-2.90787995e-01 4.56004828e-01 -5.81910461e-02 5.91901802e-02
4.64229882e-01 7.36645520e-01 -3.18082310e-02 -2.94480950e-01
-1.25722206e+00 4.60244477e-01 6.97960198e-01 8.56294274e-01
6.95761859e-01 1.51735216e-01 1.03292681e-01 6.59880102e-01
3.08637679e-01 3.35518211e-01 3.56526256e-01 -1.27420747e+00
1.83411971e-01 3.51666033e-01 9.83534753e-02 -6.83662295e-01
-4.59352322e-02 -3.46878082e-01 -6.74854636e-01 5.89997590e-01
4.26240146e-01 -1.51225463e-01 -1.14705575e+00 1.72428107e+00
2.22712934e-01 2.15614989e-01 -7.22221509e-02 6.22829199e-01
4.06751513e-01 7.41291404e-01 -1.43249556e-01 4.95153479e-02
1.16472328e+00 -1.00016689e+00 -2.91239738e-01 -5.28282762e-01
3.54277492e-01 -1.13913715e+00 1.03680873e+00 4.89336729e-01
-1.37243354e+00 -6.69521987e-01 -9.39341068e-01 -1.46324053e-01
-1.93733543e-01 -1.39551923e-01 9.04400408e-01 6.85946226e-01
-1.42761505e+00 5.65588236e-01 -9.74679410e-01 -6.29303098e-01
5.39834738e-01 3.26359540e-01 -3.49020839e-01 -2.38832548e-01
-6.32689178e-01 8.93593311e-01 4.43332136e-01 -1.13495849e-01
-1.04057765e+00 -6.74924254e-01 -9.17731762e-01 -6.95605427e-02
1.61327541e-01 -8.89271140e-01 1.35943806e+00 -9.31163192e-01
-1.34569573e+00 6.83164835e-01 -3.43226373e-01 -3.79594535e-01
6.08877957e-01 -8.39104056e-02 3.73115540e-02 1.08380035e-01
-3.15187834e-02 1.26882124e+00 7.87314832e-01 -1.52510989e+00
-4.91307139e-01 1.96635742e-02 2.78772682e-01 1.82945803e-01
-5.71401976e-02 -1.32317832e-02 -5.03003240e-01 -4.09177989e-01
-2.86695868e-01 -1.15714860e+00 -4.86798972e-01 -1.86404556e-01
-1.79585803e-03 5.73232621e-02 8.42004776e-01 -4.20710504e-01
8.68394911e-01 -2.06769180e+00 -9.37549397e-03 -5.30405752e-02
2.07352132e-01 4.25522715e-01 -2.58939445e-01 5.83106697e-01
1.07272461e-01 3.94655526e-01 -2.13643774e-01 -5.82992554e-01
-2.31226876e-01 3.08275372e-01 -3.31140518e-01 1.02043161e-02
4.45514590e-01 1.07574713e+00 -8.41618121e-01 -5.38448095e-01
2.83811629e-01 6.71315014e-01 -6.73509479e-01 1.85889855e-01
-4.12118435e-01 5.39579988e-01 -1.16149545e-01 3.50268036e-01
5.12336135e-01 -5.99134386e-01 1.74710929e-01 4.48099524e-02
-5.40417898e-03 2.33005732e-01 -1.27266800e+00 1.64472532e+00
-5.26796341e-01 7.39581108e-01 -1.79896757e-01 -7.98891187e-01
7.15983808e-01 3.99837315e-01 2.26832911e-01 -5.91664791e-01
-3.10562085e-03 2.57939883e-02 1.09785557e-01 -2.81218857e-01
4.31533158e-01 -2.48083368e-01 1.51964381e-01 8.73453796e-01
1.50317296e-01 -5.85131109e-01 6.35970831e-01 2.62349218e-01
8.78870845e-01 3.50781769e-01 1.73931181e-01 -1.00613907e-01
-3.93514223e-02 6.88701943e-02 2.97021300e-01 8.45678329e-01
2.30837330e-01 1.08152640e+00 3.14898014e-01 -5.35223961e-01
-1.47181070e+00 -9.85586166e-01 6.92019761e-02 5.88872194e-01
-2.93590017e-02 -3.79391998e-01 -7.35936403e-01 -3.87457699e-01
-3.82960767e-01 5.94341278e-01 -6.83314323e-01 1.13045454e-01
-7.29530871e-01 -8.99764538e-01 3.69134337e-01 7.08176374e-01
4.37202573e-01 -1.14777887e+00 -8.48981321e-01 1.40372843e-01
-3.81673910e-02 -1.25694406e+00 -3.83700609e-01 -6.07167855e-02
-1.15662646e+00 -8.03852677e-01 -7.23088562e-01 -7.68342853e-01
8.94413769e-01 4.50980932e-01 1.61574042e+00 1.78381443e-01
-4.79081392e-01 3.17112148e-01 -1.61501274e-01 -5.26225150e-01
-6.23457015e-01 1.90628365e-01 -4.45072591e-01 -3.36393744e-01
1.19722523e-01 -5.64278364e-01 -6.10452056e-01 2.07123652e-01
-1.37715864e+00 6.07549667e-01 8.86486173e-01 9.08294857e-01
5.71493328e-01 -4.27420775e-04 4.72145498e-01 -1.21515810e+00
6.24507308e-01 -3.38177383e-01 -6.95683062e-01 8.12379792e-02
-7.29014277e-01 2.41420805e-01 4.76370096e-01 -4.51528162e-01
-1.25853109e+00 1.87177524e-01 -1.86752602e-02 -1.20992035e-01
-3.85566354e-01 3.04135889e-01 1.08216226e-01 -2.91107073e-02
6.88061893e-01 2.05019623e-01 2.41599500e-01 -3.47497255e-01
3.90372217e-01 3.09786737e-01 2.85528749e-01 -5.24555326e-01
8.16718042e-01 5.17107487e-01 1.01483718e-01 -6.78449929e-01
-6.86302066e-01 -2.85940796e-01 -6.19428396e-01 -1.66380554e-02
6.58755541e-01 -9.35946226e-01 -1.34855479e-01 4.54388708e-01
-1.00480640e+00 -6.70912862e-01 -3.33818257e-01 3.84776592e-01
-6.10719442e-01 1.95767328e-01 -5.70434272e-01 -4.80744481e-01
-2.95938700e-02 -1.23807502e+00 1.22936487e+00 4.43974406e-01
-3.27538282e-01 -1.14459062e+00 1.77247137e-01 4.94399786e-01
6.71306074e-01 2.94605225e-01 6.54814065e-01 -1.42293274e-01
-9.98420417e-01 -8.71250257e-02 -1.31633773e-01 2.93152064e-01
2.09310606e-01 1.83449879e-01 -9.51427460e-01 -4.88721132e-01
-1.44408941e-01 -6.39751196e-01 7.55327106e-01 3.28961790e-01
9.63762224e-01 -2.97099464e-02 -3.27583939e-01 5.35055041e-01
1.54550397e+00 1.05593771e-01 9.15793300e-01 2.87770003e-01
5.05677581e-01 4.11700666e-01 3.34722877e-01 1.32949159e-01
4.10316050e-01 5.71477294e-01 2.23362729e-01 -5.27866662e-01
-3.80986094e-01 -2.73584217e-01 1.11949839e-01 4.92288023e-01
-2.79884309e-01 -3.70957822e-01 -7.74732053e-01 7.56556392e-01
-1.68059552e+00 -1.16752946e+00 1.74005136e-01 2.20707393e+00
8.78742397e-01 9.08594206e-02 2.70751536e-01 -2.69500315e-01
4.65334147e-01 1.29926369e-01 -4.61635232e-01 -2.26778686e-01
-5.19041577e-03 4.73030955e-01 2.43907914e-01 5.01704395e-01
-8.32120657e-01 9.29599643e-01 7.78409481e+00 5.22331178e-01
-1.27256870e+00 -1.60102800e-01 8.71982336e-01 -3.96912962e-01
-3.60123813e-01 3.00207168e-01 -8.10412705e-01 3.79119515e-01
1.02071798e+00 2.30576899e-02 5.09215236e-01 8.02531004e-01
1.42527312e-01 -4.32849735e-01 -1.15357471e+00 7.63632655e-01
3.01856190e-01 -1.40822887e+00 1.24258824e-01 4.95013893e-01
9.79912996e-01 9.66422558e-02 1.00279085e-01 1.30667817e-02
6.18593276e-01 -1.34192848e+00 6.07713521e-01 3.41421247e-01
5.65999508e-01 -6.56745970e-01 5.61579168e-01 5.73271751e-01
-9.54315364e-01 3.57744306e-01 -3.90011013e-01 -4.46623005e-02
4.00820225e-01 4.54611868e-01 -1.14367592e+00 2.98925728e-01
5.99989474e-01 3.94425869e-01 -8.22394073e-01 1.12056267e+00
-2.13003248e-01 6.68590248e-01 -3.90112907e-01 2.16804340e-01
1.54714480e-01 2.02892534e-02 8.75666738e-02 1.09350121e+00
4.98896807e-01 -1.55964270e-01 1.64873719e-01 9.48042214e-01
9.44532803e-04 -3.19268078e-01 -7.74983644e-01 -2.02025369e-01
2.29771003e-01 1.42465067e+00 -9.60667074e-01 -3.12946260e-01
-4.73884374e-01 9.33700979e-01 3.05984020e-01 2.79614955e-01
-6.78721189e-01 -8.63634795e-02 3.03576112e-01 3.78841221e-01
4.57933754e-01 -4.66841251e-01 -6.14618398e-02 -1.20922601e+00
-1.28401011e-01 -1.11138856e+00 8.29851925e-02 -9.02579546e-01
-8.88683736e-01 7.14156508e-01 1.78673133e-01 -9.56908345e-01
-9.18930292e-01 -4.19887602e-01 -5.41125953e-01 1.01390731e+00
-1.47873294e+00 -1.40008116e+00 -4.98824984e-01 2.82985389e-01
6.89676702e-01 1.87184304e-01 8.95207465e-01 1.16707049e-01
-2.95102626e-01 3.62045288e-01 -1.68614805e-01 -9.28394720e-02
6.98803663e-01 -1.26640821e+00 8.61331880e-01 1.04009044e+00
6.04414582e-01 8.35346103e-01 7.59317994e-01 -5.44393718e-01
-1.10556638e+00 -7.07302332e-01 7.51725852e-01 -6.39867067e-01
2.42476150e-01 -5.41798472e-01 -7.90026307e-01 8.37626576e-01
6.14894450e-01 -4.66526821e-02 6.94398761e-01 -2.76937988e-02
-1.58142477e-01 1.77385688e-01 -9.66379821e-01 5.97108006e-01
9.40291882e-01 -2.39334837e-01 -1.48373455e-01 9.79849473e-02
3.75242114e-01 -4.95259821e-01 -7.16743410e-01 2.71235108e-01
5.87658942e-01 -1.09283841e+00 1.01017249e+00 -2.80501068e-01
5.88150263e-01 -4.73147660e-01 1.59262300e-01 -1.45841455e+00
-3.97945046e-01 -6.60095274e-01 -1.63571343e-01 1.28933728e+00
5.16007602e-01 -4.70702201e-01 8.41421008e-01 6.28566325e-01
1.36610568e-01 -6.79042935e-01 -3.56823027e-01 -7.66651988e-01
1.32241964e-01 -1.34700850e-01 5.64043581e-01 5.42097270e-01
-5.01553357e-01 5.73884130e-01 -5.91214895e-01 -1.55615598e-01
6.40489399e-01 3.38673651e-01 1.12865758e+00 -1.23311365e+00
-5.06794751e-01 -2.86222935e-01 -3.32123578e-01 -1.14770687e+00
-1.24206565e-01 -5.35115957e-01 8.47428814e-02 -1.78967381e+00
4.73125041e-01 -3.85129213e-01 5.60336299e-02 4.05329257e-01
-3.41091126e-01 7.37208545e-01 2.94056565e-01 2.83075750e-01
-3.35436374e-01 1.66524485e-01 1.46826315e+00 2.15534642e-01
8.59103501e-02 -2.70854801e-01 -1.07258642e+00 5.92348158e-01
7.66859114e-01 -3.65093589e-01 -8.69647264e-01 -5.62951565e-01
1.46557897e-01 -4.06801015e-01 4.12384123e-01 -1.14961183e+00
7.53256157e-02 -2.65369862e-01 7.59260535e-01 -4.34245318e-01
5.17749548e-01 -4.80165869e-01 4.58431542e-01 2.10776463e-01
-1.03145689e-01 1.91565678e-02 2.73281395e-01 3.04400593e-01
-1.83935314e-01 -3.76191080e-01 9.24849391e-01 -6.27179205e-01
-7.63807356e-01 9.68076736e-02 -3.57083321e-01 1.56404391e-01
9.95526135e-01 -2.91181773e-01 -2.42979810e-01 -5.09810507e-01
-2.73303300e-01 -2.47928593e-02 9.53156471e-01 3.55065733e-01
4.54042107e-01 -1.17384088e+00 -6.96215272e-01 2.88043022e-01
-4.69910949e-02 3.52277786e-01 1.64357603e-01 3.52698743e-01
-6.59471810e-01 3.40113491e-01 -2.29089662e-01 -6.95885777e-01
-1.33598888e+00 4.17546570e-01 1.20218009e-01 -4.36938435e-01
-5.88614047e-01 7.22674191e-01 3.92185479e-01 -2.69632876e-01
-1.68904856e-01 3.60943526e-02 3.20666246e-02 -1.75527424e-01
6.87711477e-01 6.56943992e-02 -4.72735018e-02 -6.28522575e-01
3.18784900e-02 4.94660765e-01 -3.30001920e-01 -5.55453718e-01
1.55726659e+00 -2.62343902e-02 1.24060221e-01 2.28843451e-01
8.86900485e-01 -7.86806121e-02 -1.73578179e+00 -3.34230997e-02
-3.64888072e-01 -6.86789751e-01 5.52521050e-02 -8.02281976e-01
-9.60736275e-01 9.26807284e-01 4.48459983e-01 3.37784022e-01
1.14998150e+00 2.56386817e-01 8.32391441e-01 1.03608459e-01
4.11159664e-01 -9.47363377e-01 3.07538062e-01 2.95433491e-01
7.35095978e-01 -1.33202446e+00 1.05170920e-01 -1.97437719e-01
-7.14590907e-01 1.03498447e+00 6.35016561e-01 -1.94570974e-01
2.16636896e-01 6.10105693e-01 1.84825167e-01 -1.66072950e-01
-9.59604681e-01 -1.34271055e-01 2.83903211e-01 6.98341072e-01
7.05909133e-01 -3.01432014e-01 1.03571877e-01 -2.84236133e-01
-2.50785857e-01 2.76383251e-01 7.05988109e-01 1.13111794e+00
-3.77255738e-01 -1.69588172e+00 -2.39903778e-01 4.30019021e-01
-4.77885813e-01 -1.74669832e-01 -3.17302972e-01 8.69170666e-01
-5.68703283e-03 8.76373112e-01 7.47567862e-02 -1.40144333e-01
4.90627699e-02 2.26042569e-02 8.88023019e-01 -9.44027245e-01
-3.21087658e-01 2.71143228e-01 -6.91816732e-02 -5.62135041e-01
-6.63510144e-01 -8.13200176e-01 -7.95055509e-01 -2.30909333e-01
-3.87735873e-01 -6.41073063e-02 6.60210490e-01 9.09893453e-01
5.46939433e-01 4.38736945e-01 3.89015079e-01 -1.13963044e+00
-1.08700633e-01 -9.59764838e-01 -2.89653152e-01 4.62225318e-01
1.98459968e-01 -4.38090593e-01 -1.42298862e-01 5.72112143e-01] | [11.233120918273926, -0.22748635709285736] |
6ce0eaa2-6996-4fdf-9ff8-d488d3909e29 | detecting-uncertainty-cues-in-hungarian | null | null | https://aclanthology.org/W16-5002 | https://aclanthology.org/W16-5002.pdf | Detecting Uncertainty Cues in Hungarian Social Media Texts | In this paper, we aim at identifying uncertainty cues in Hungarian social media texts. We present our machine learning based uncertainty detector which is based on a rich features set including lexical, morphological, syntactic, semantic and discourse-based features, and we evaluate our system on a small set of manually annotated social media texts. We also carry out cross-domain and domain adaptation experiments using an annotated corpus of standard Hungarian texts and show that domain differences significantly affect machine learning. Furthermore, we argue that differences among uncertainty cue types may also affect the efficiency of uncertainty detection. | ['Veronika Vincze'] | 2016-12-01 | null | null | null | ws-2016-12 | ['instance-search'] | ['computer-vision'] | [-1.65186644e-01 6.22835577e-01 -1.69604532e-02 -7.61836112e-01
-1.09385777e+00 -6.80303633e-01 1.01052499e+00 9.87105548e-01
-8.64300907e-01 1.12127376e+00 8.01814973e-01 2.91789602e-02
-8.78222957e-02 -6.03734672e-01 -4.58556950e-01 -2.65162196e-02
-1.16456293e-01 9.82035697e-01 4.67415512e-01 -4.56287205e-01
5.16663015e-01 -5.72444797e-02 -1.22429955e+00 3.87005597e-01
9.16900694e-01 7.87758350e-01 5.33404015e-02 2.86881864e-01
-3.87723148e-01 5.98569930e-01 -9.23312783e-01 -7.72403002e-01
-2.48177245e-01 -1.23381160e-01 -1.18988490e+00 -2.15683773e-01
1.02053232e-01 -4.80221286e-02 -1.62760299e-02 8.98713887e-01
6.21456802e-01 3.44749928e-01 1.09561014e+00 -7.27552772e-01
-2.38145098e-01 1.67115355e+00 2.17650067e-02 2.56173819e-01
8.61800194e-01 -3.94230545e-01 1.04455435e+00 -6.47707999e-01
1.31647718e+00 1.47591949e+00 4.54678833e-01 5.25431335e-01
-8.93454015e-01 -4.06706095e-01 1.16351284e-01 2.88698465e-01
-1.34181130e+00 -6.37119949e-01 1.03485441e+00 -3.72766465e-01
1.11266148e+00 -1.16470277e-01 2.98268825e-01 1.55190504e+00
1.62012100e-01 8.51642013e-01 1.00022960e+00 -7.86815047e-01
7.22392023e-01 2.91967750e-01 3.16016108e-01 2.65745521e-01
2.10495830e-01 -1.87723264e-01 -5.43687046e-01 -5.52711785e-01
1.83641717e-01 -9.69042242e-01 -7.44451210e-02 -8.90696347e-02
-1.02203715e+00 1.19106793e+00 -2.14991793e-01 7.35541642e-01
4.91818115e-02 -3.33143324e-01 7.48243272e-01 4.51785117e-01
7.06700623e-01 6.92009747e-01 -8.08009207e-01 -3.23414117e-01
-5.35162151e-01 2.75500178e-01 1.23737240e+00 9.58908141e-01
4.46331024e-01 -2.15852901e-01 7.43147954e-02 9.98523772e-01
4.83283758e-01 3.28788131e-01 7.18694329e-01 -8.23487461e-01
6.48653984e-01 2.05673024e-01 1.95348233e-01 -1.00450420e+00
-7.49373138e-01 4.00131732e-01 4.44904007e-02 -3.08555335e-01
6.33064926e-01 -6.30476773e-01 -3.53736430e-01 1.77545702e+00
2.15267241e-01 -4.20962989e-01 5.75447202e-01 5.23010075e-01
1.14672422e+00 4.52398360e-01 4.04468864e-01 -5.62128365e-01
1.22766125e+00 -3.31066668e-01 -9.43485320e-01 -3.12809229e-01
7.84795344e-01 -8.44802082e-01 8.86978388e-01 3.48910630e-01
-9.54132736e-01 -9.42267198e-03 -9.19119716e-01 8.76695216e-02
-6.11271381e-01 -2.27498040e-01 5.23891211e-01 6.30340695e-01
-3.95152658e-01 5.41307926e-01 -3.58118951e-01 -5.79747140e-01
-1.95614278e-01 2.64645010e-01 -4.36554700e-01 2.58441001e-01
-1.90734410e+00 1.12247276e+00 1.04234159e+00 -6.23757362e-01
-9.32699516e-02 4.33900990e-02 -1.41892612e+00 -2.34667391e-01
6.20828152e-01 -2.05887690e-01 1.39893818e+00 -8.92850935e-01
-1.67984521e+00 9.51178372e-01 5.68627156e-02 -5.12172639e-01
4.74716723e-01 -1.60472870e-01 -8.60124767e-01 8.65883678e-02
1.46903053e-01 3.52352142e-01 8.33210707e-01 -9.42888200e-01
-5.41984200e-01 -2.75240928e-01 -2.56629884e-01 2.19334215e-01
-5.02556972e-02 5.46952248e-01 1.54967889e-01 -7.54137516e-01
7.14882687e-02 -7.31760681e-01 2.77547762e-02 -8.33929121e-01
-9.34190229e-02 -7.07905173e-01 4.96639997e-01 -5.09350955e-01
1.28889549e+00 -2.29511571e+00 1.46402523e-01 5.30944288e-01
-3.73888910e-02 -1.59620225e-01 1.39629170e-01 3.36395860e-01
2.64458954e-01 1.84479609e-01 -1.94725201e-01 -3.06198616e-02
3.93778384e-01 3.22096705e-01 -1.30631864e-01 3.50873053e-01
2.99880058e-01 6.20990276e-01 -9.81683016e-01 -1.04880488e+00
1.59625486e-01 -4.27630097e-02 -4.72740114e-01 -1.23629637e-01
-7.09529281e-01 1.55479133e-01 -5.77044666e-01 5.36309183e-01
2.83515722e-01 3.26841027e-01 4.96271759e-01 3.36860269e-01
2.84306873e-02 5.94839931e-01 -1.29025447e+00 1.74559665e+00
-1.77497581e-01 8.06236386e-01 -1.07958399e-01 -7.14560926e-01
8.73180568e-01 4.32792187e-01 5.72248846e-02 -2.11409390e-01
7.66650975e-01 3.18742782e-01 9.59892794e-02 -4.19382274e-01
8.29500854e-01 1.13801397e-02 -9.48096275e-01 4.01628017e-01
3.26953501e-01 -7.16865003e-01 2.54458666e-01 2.41604820e-01
8.38103890e-01 -1.51622534e-01 8.08213890e-01 -5.55964410e-01
3.51095557e-01 1.24143712e-01 5.59500992e-01 6.18937731e-01
-6.90382659e-01 3.96065533e-01 6.95453584e-01 -1.13716967e-01
-7.79718101e-01 -7.53932238e-01 -7.59025097e-01 1.23715162e+00
6.47114739e-02 -7.36624479e-01 -6.83686912e-01 -9.22920108e-01
-2.82983612e-02 1.28761053e+00 -6.72573566e-01 2.09745914e-01
-3.43844324e-01 -3.63703668e-01 5.14761925e-01 3.03194672e-01
1.26064852e-01 -1.09254003e+00 -4.18119073e-01 4.85667169e-01
-3.27750891e-01 -1.41499376e+00 -1.49622828e-01 2.80833185e-01
-5.02360582e-01 -9.13129330e-01 1.36197418e-01 -6.25587881e-01
2.05852017e-02 -5.55879831e-01 1.17176044e+00 -5.35239756e-01
1.73215970e-01 4.37467515e-01 -7.75104344e-01 -1.02187502e+00
-6.24510765e-01 4.31690626e-02 1.78811461e-01 -7.00333595e-01
7.78380036e-01 -9.42739770e-02 3.61535311e-01 9.79391038e-02
-8.86405468e-01 -5.78354418e-01 1.66743845e-01 7.33804345e-01
4.96726520e-02 -3.36162969e-02 5.86337268e-01 -1.10096753e+00
9.43564415e-01 -5.00625134e-01 -4.85661209e-01 9.89909396e-02
-3.05919319e-01 4.58358645e-01 -1.23962313e-02 -4.73323286e-01
-1.57639551e+00 -1.44084508e-03 -2.16147035e-01 4.01676297e-01
-4.47792590e-01 7.77972639e-01 -3.70811164e-01 1.88263535e-01
1.08050025e+00 -4.62314904e-01 -2.84525752e-01 -7.52827972e-02
5.37635922e-01 7.99079359e-01 1.98915109e-01 -9.56636667e-01
2.13552505e-01 -1.17831500e-02 -5.17637372e-01 -1.16385317e+00
-8.04349542e-01 -2.73945898e-01 -6.02482200e-01 -1.15893297e-01
6.83999240e-01 -7.85499454e-01 -4.74835262e-02 3.79383340e-02
-1.18183982e+00 -7.92207867e-02 -2.36337721e-01 7.89565384e-01
-6.73604190e-01 4.66213435e-01 -6.41496539e-01 -7.29660451e-01
-4.99180071e-02 -7.27812290e-01 8.70087743e-01 4.18923721e-02
-1.11970258e+00 -1.27563846e+00 4.05872673e-01 1.84881017e-01
3.71726528e-02 2.62879163e-01 1.06458545e+00 -1.61982250e+00
2.90466636e-01 3.23836133e-02 4.45399322e-02 -1.02706239e-01
-2.92877108e-01 2.20745370e-01 -8.06601346e-01 2.78071225e-01
-1.78982407e-01 -6.77836359e-01 8.17444265e-01 3.87420893e-01
3.08627903e-01 -1.22173429e-01 -2.62025535e-01 -1.06432281e-01
1.11322558e+00 -3.49214412e-02 3.35085094e-01 5.56448460e-01
5.73466383e-02 9.30386484e-01 9.86985683e-01 7.57214665e-01
3.24190021e-01 3.69286835e-01 -1.16666041e-01 7.29647458e-01
4.49056596e-01 3.39881629e-02 4.86664295e-01 6.17184997e-01
1.20797642e-01 -2.93569237e-01 -1.11598706e+00 5.53602993e-01
-2.05973864e+00 -8.90353680e-01 1.62671253e-01 1.77377105e+00
1.15816891e+00 4.32022482e-01 4.66984436e-02 9.60522667e-02
7.93662131e-01 1.38141602e-01 1.54512972e-01 -7.60128438e-01
-5.09710133e-01 1.74357980e-01 1.29380599e-01 8.15524399e-01
-1.42834210e+00 1.45121837e+00 7.60088491e+00 6.66762173e-01
-6.17536068e-01 -7.84571022e-02 2.09887087e-01 1.05336674e-01
-4.95596766e-01 -3.48022580e-01 -5.99978268e-01 2.39622086e-01
1.11161554e+00 -4.55330491e-01 -4.39782552e-02 9.88600433e-01
-1.89844400e-01 -4.07386750e-01 -1.15593183e+00 7.03361511e-01
1.43596873e-01 -1.04428661e+00 -4.16455328e-01 -4.33394998e-01
6.72985554e-01 9.01614875e-02 -3.15833420e-01 5.12298644e-01
5.21353364e-01 -5.32321572e-01 7.77447224e-01 -1.20852903e-01
2.53576010e-01 -9.24829066e-01 1.09581864e+00 8.57057422e-02
-4.94541228e-01 1.02893263e-01 -3.82001668e-01 -4.49164696e-02
3.64854373e-02 5.59063733e-01 -9.89602566e-01 4.05552298e-01
5.81287861e-01 2.81267732e-01 -5.02047598e-01 4.66282010e-01
-3.91313732e-01 4.77884769e-01 -5.20983875e-01 -5.23433268e-01
1.89807788e-01 1.59891769e-01 8.00730407e-01 1.67600262e+00
5.60968854e-02 3.38156015e-01 9.57153216e-02 6.30372167e-01
9.68777165e-02 6.86682105e-01 -7.57991314e-01 -2.53265917e-01
5.27273655e-01 8.44628155e-01 -7.47086227e-01 -3.17841142e-01
-4.55034465e-01 6.70901537e-01 5.01758754e-01 -2.67342217e-02
-5.43181956e-01 -6.15438104e-01 4.61494148e-01 -3.09118539e-01
8.17332864e-02 -1.74552798e-01 -1.90475851e-01 -1.51876760e+00
-3.23956162e-01 -6.92086577e-01 7.41940856e-01 -3.77718419e-01
-1.57998049e+00 5.87830782e-01 5.00294328e-01 -6.81452036e-01
-8.22462916e-01 -6.60603702e-01 -4.85975057e-01 4.90778953e-01
-1.06256604e+00 -7.24279344e-01 2.52594382e-01 5.41749716e-01
5.31854153e-01 -3.87459993e-01 8.84151042e-01 -2.39068255e-01
-3.63526285e-01 3.79152000e-01 9.02224928e-02 1.57910377e-01
1.41675711e+00 -1.38608813e+00 1.30542785e-01 4.54097539e-01
3.33103235e-03 4.31771249e-01 1.24005461e+00 -1.04754221e+00
-8.20681453e-01 -6.73857927e-01 1.34308708e+00 -8.04084003e-01
1.08000386e+00 -3.03680331e-01 -8.53039265e-01 7.37643480e-01
3.72476578e-01 -4.46561635e-01 1.06566167e+00 6.39812291e-01
-3.74410182e-01 6.62415326e-01 -1.62790453e+00 4.88871634e-01
8.49870563e-01 -4.36974764e-01 -1.62344027e+00 2.76311576e-01
8.11112761e-01 -5.10867417e-01 -8.57829332e-01 5.45672417e-01
3.93866748e-01 -6.89844549e-01 4.74807858e-01 -5.95011175e-01
3.75899494e-01 1.58134833e-01 -4.74447012e-01 -1.61483610e+00
-6.38621747e-02 -4.85009819e-01 1.25977859e-01 1.44168568e+00
8.20979059e-01 -6.04646802e-01 5.17714739e-01 1.03022790e+00
-7.15477988e-02 7.86986277e-02 -9.63268220e-01 -7.01954067e-01
3.92711908e-01 -8.80940080e-01 3.26499820e-01 1.15896344e+00
1.27200961e+00 6.35700583e-01 9.06426981e-02 -1.61397561e-01
3.60885710e-01 -1.48034081e-01 1.59471884e-01 -1.57025659e+00
-1.44752830e-01 -3.03859174e-01 -4.24612582e-01 -1.77660525e-01
8.06800425e-01 -3.83875728e-01 2.55859315e-01 -1.03103566e+00
-7.81423226e-02 -2.14491397e-01 7.17519596e-02 1.24631733e-01
1.56545807e-02 -3.94573092e-01 -2.10077822e-01 -7.13508204e-02
-8.72108042e-01 4.91114497e-01 6.18841112e-01 -1.50064856e-01
-5.12903273e-01 -2.94852704e-01 -5.27076840e-01 9.26079214e-01
9.61061239e-01 -3.45252335e-01 -1.81633323e-01 -1.24422878e-01
3.00813854e-01 1.30929928e-02 -4.22421128e-01 -6.74212396e-01
-2.56731547e-02 -4.74894673e-01 3.68211985e-01 -2.85128355e-01
2.77214676e-01 -7.45406568e-01 -3.79654676e-01 2.40867808e-01
-5.27793348e-01 -2.23030552e-01 5.53581297e-01 5.44947267e-01
-3.08414549e-01 -6.45443320e-01 5.70867360e-01 -2.12608635e-01
-1.00317132e+00 -3.08973819e-01 -8.89547110e-01 4.55044389e-01
1.20743573e+00 2.90287942e-01 -3.49285990e-01 -4.36473191e-01
-9.14648354e-01 -3.02858301e-03 5.95825613e-01 4.49527621e-01
3.29116166e-01 -1.15930331e+00 -7.69999981e-01 -9.57220048e-02
6.15569055e-01 -3.62035662e-01 -2.18115360e-01 1.02646396e-01
-3.62146497e-01 6.16716325e-01 -2.24172920e-01 -3.01045626e-01
-1.27704144e+00 4.99084502e-01 -3.04248571e-01 8.85484070e-02
-6.97864518e-02 8.14201355e-01 -5.51457226e-01 -4.65717256e-01
3.93284768e-01 -3.59068632e-01 -5.86661756e-01 2.81781316e-01
3.49727750e-01 2.03844637e-01 -1.28220648e-01 -8.37536216e-01
-5.77292085e-01 1.57832101e-01 7.78723583e-02 -7.83747017e-01
9.76225376e-01 -2.68565893e-01 9.27974470e-03 7.90748417e-01
7.70575225e-01 4.47901398e-01 -5.14031768e-01 -6.24373853e-01
7.04706550e-01 -2.52036691e-01 1.13191400e-02 -6.50164425e-01
2.69266181e-02 2.58972585e-01 2.24007040e-01 3.45509917e-01
6.59737945e-01 6.02684140e-01 3.51822078e-01 8.10494959e-01
4.43543404e-01 -1.79040396e+00 -1.92701966e-01 1.02538729e+00
8.45534146e-01 -1.69594884e+00 -3.08800917e-02 -6.34134769e-01
-1.09626484e+00 1.09703183e+00 4.95730013e-01 1.54239357e-01
7.42571890e-01 4.10870910e-01 1.57747105e-01 -9.61829424e-02
-8.37819815e-01 -3.85750800e-01 1.85666710e-01 7.95448184e-01
7.66163409e-01 2.41859764e-01 -9.60507214e-01 1.08846080e+00
-4.89737272e-01 -4.40625519e-01 7.26474285e-01 1.01781595e+00
-8.57180834e-01 -1.39766872e+00 -5.30006647e-01 2.38461062e-01
-5.78798652e-01 -1.91061907e-02 -9.77270484e-01 1.00357246e+00
-4.93735783e-02 1.10055208e+00 -2.22498924e-02 -3.95090193e-01
6.25631958e-02 4.88985062e-01 7.41921484e-01 -8.54991019e-01
-4.73938853e-01 1.74828365e-01 9.33694422e-01 -2.25663632e-01
-4.41479355e-01 -1.21743727e+00 -1.37950110e+00 -2.82343090e-01
-3.88314337e-01 6.13021731e-01 7.52471328e-01 1.31320596e+00
-5.42665869e-02 2.53259670e-02 1.76984519e-01 -5.43582082e-01
-4.71887589e-01 -1.33920753e+00 -6.51641607e-01 4.31087881e-01
-1.55257229e-02 -7.34211862e-01 -5.17906666e-01 -6.06780462e-02] | [10.29333209991455, 9.316666603088379] |
07e24a56-5d40-45fc-8d36-2d02884a75dc | a-survey-on-audio-synthesis-and-audio-visual | 2108.00443 | null | https://arxiv.org/abs/2108.00443v1 | https://arxiv.org/pdf/2108.00443v1.pdf | A Survey on Audio Synthesis and Audio-Visual Multimodal Processing | With the development of deep learning and artificial intelligence, audio synthesis has a pivotal role in the area of machine learning and shows strong applicability in the industry. Meanwhile, significant efforts have been dedicated by researchers to handle multimodal tasks at present such as audio-visual multimodal processing. In this paper, we conduct a survey on audio synthesis and audio-visual multimodal processing, which helps understand current research and future trends. This review focuses on text to speech(TTS), music generation and some tasks that combine visual and acoustic information. The corresponding technical methods are comprehensively classified and introduced, and their future development trends are prospected. This survey can provide some guidance for researchers who are interested in the areas like audio synthesis and audio-visual multimodal processing. | ['Zhaofeng Shi'] | 2021-08-01 | null | null | null | null | ['music-generation', 'music-generation'] | ['audio', 'music'] | [ 1.95917249e-01 -3.87238175e-01 -1.68348089e-01 -8.73859152e-02
-7.39500225e-01 -4.07833159e-01 3.97543043e-01 7.03447359e-03
-7.00231940e-02 3.98571312e-01 3.65850359e-01 -2.59529110e-02
1.30661547e-01 -4.28348482e-01 -2.45929152e-01 -9.78281856e-01
1.62454560e-01 -2.32799370e-02 -2.02793732e-01 -1.40040517e-01
2.31896386e-01 4.90890294e-01 -1.84463394e+00 7.34522820e-01
5.00683784e-01 1.02823937e+00 3.47387642e-01 8.23126733e-01
-3.41097027e-01 5.88986814e-01 -9.22753394e-01 -4.67229366e-01
-3.51671070e-01 -5.23371875e-01 -4.03540015e-01 1.72596455e-01
1.97914049e-01 -2.18481138e-01 -4.27754223e-01 9.44142759e-01
1.04061639e+00 3.18291098e-01 7.54815400e-01 -1.47099674e+00
-5.13335049e-01 6.06978655e-01 -6.24054015e-01 -1.33628204e-01
6.58876061e-01 1.27902627e-01 9.37786877e-01 -1.06244063e+00
1.71373218e-01 1.45409489e+00 2.48480842e-01 5.73033273e-01
-6.42652869e-01 -7.74070680e-01 8.09891000e-02 6.19551957e-01
-1.32889080e+00 -4.87532556e-01 9.69059765e-01 -4.08171833e-01
8.12195420e-01 3.97659510e-01 7.35146999e-01 9.18353498e-01
1.16302237e-01 1.26105940e+00 7.07619250e-01 -6.61281526e-01
-4.49956581e-02 1.16945431e-01 -3.85652483e-01 4.11789507e-01
-5.22024095e-01 -1.02189772e-01 -7.66174138e-01 2.28491083e-01
6.47983313e-01 -2.36689195e-01 -8.47131908e-02 -2.27667321e-03
-1.37386739e+00 7.88725913e-01 1.16871871e-01 4.94713783e-01
-2.53247082e-01 1.96065471e-01 7.39959598e-01 2.77356505e-01
2.00171366e-01 1.72099128e-01 5.57315648e-02 -2.50920147e-01
-1.03543890e+00 1.61146566e-01 3.32380384e-01 7.53712475e-01
1.31242052e-01 8.52268696e-01 -9.86529887e-02 1.32147181e+00
4.21788514e-01 7.73440778e-01 3.05249423e-01 -1.07589567e+00
3.83493036e-01 5.61311841e-02 -2.78446704e-01 -9.89669800e-01
-3.36160719e-01 6.66124374e-02 -9.85686839e-01 2.60701448e-01
-4.09031697e-02 -2.41202950e-01 -5.77558577e-01 1.24500215e+00
1.11663692e-01 -1.41789064e-01 6.75276620e-03 1.06382942e+00
1.83912659e+00 1.43787587e+00 6.81302175e-02 -3.59454840e-01
1.43782222e+00 -7.23963857e-01 -1.27191222e+00 2.56263241e-02
-2.19407469e-01 -1.38071835e+00 1.04957736e+00 5.63540041e-01
-1.33907628e+00 -8.04106772e-01 -1.05639672e+00 -9.10923779e-02
-2.38701329e-01 3.80862772e-01 5.03254116e-01 5.96525550e-01
-9.20621336e-01 4.11778055e-02 -5.94030023e-01 -3.92296851e-01
4.39667851e-02 3.75024050e-01 -1.37785032e-01 2.24024653e-01
-1.34439421e+00 6.35488272e-01 3.63662899e-01 1.48748696e-01
-9.39353406e-01 -2.95645863e-01 -9.66127515e-01 1.71559025e-02
2.43905306e-01 -6.52804255e-01 1.59024179e+00 -8.13204110e-01
-1.80993164e+00 7.07280993e-01 -3.21526766e-01 -1.51740126e-02
1.33505195e-01 6.41107336e-02 -7.76729226e-01 4.14498210e-01
-2.61519313e-01 9.74826574e-01 1.09011126e+00 -1.32830584e+00
-9.75190818e-01 -1.62987888e-01 -3.60230841e-02 4.54469085e-01
-4.27845359e-01 4.88423616e-01 -6.89257324e-01 -9.29390073e-01
1.64793935e-02 -6.83532000e-01 2.27989063e-01 -1.71651497e-01
-3.62933159e-01 -4.02472377e-01 9.04714763e-01 -5.66610575e-01
1.35515630e+00 -2.28678679e+00 3.38789612e-01 -1.27709970e-01
1.63838249e-02 3.41289699e-01 -2.72813559e-01 9.40940261e-01
-6.53433427e-02 -1.12078220e-01 6.20442536e-03 -4.25882399e-01
5.36434762e-02 -1.15019940e-01 -6.07736826e-01 2.36881405e-01
-5.36638405e-03 8.48308444e-01 -6.77034557e-01 -7.42261648e-01
6.65541708e-01 6.21256411e-01 -4.16417941e-02 1.62599042e-01
-9.33289081e-02 6.16984069e-01 -2.69973338e-01 1.01188052e+00
4.32619333e-01 1.18629076e-01 -1.11711182e-01 -3.32391143e-01
-3.34360898e-01 -5.24723567e-02 -1.10735655e+00 1.46942449e+00
-5.46687901e-01 1.14822793e+00 3.55584145e-01 -9.50350344e-01
8.68475020e-01 7.87601948e-01 4.61537689e-01 -8.41829956e-01
2.41718546e-01 2.15118513e-01 -1.59796670e-01 -8.69934916e-01
6.63552821e-01 -4.43967551e-01 -7.65871778e-02 2.40792319e-01
1.25310838e-03 -4.57785696e-01 3.72557670e-01 1.17027208e-01
1.30436853e-01 -2.09708065e-01 3.09614033e-01 5.40806115e-01
9.27121162e-01 -2.27417648e-01 1.55962139e-01 2.90104419e-01
-7.11249635e-02 5.66601574e-01 2.89962292e-01 -2.00567290e-01
-9.10638452e-01 -1.18944550e+00 1.18607737e-01 1.38273644e+00
2.14805350e-01 -4.54805821e-01 -6.16488516e-01 -1.14203952e-01
-2.99922168e-01 3.28907788e-01 -1.51774138e-01 5.19789057e-03
-4.67377573e-01 -3.83050710e-01 7.04761684e-01 5.92462242e-01
3.50993961e-01 -1.56716204e+00 -1.25930563e-01 3.25018330e-03
-6.03862643e-01 -9.89240944e-01 -1.64463773e-01 -3.43890697e-01
-7.18808949e-01 -7.24076450e-01 -1.37726963e+00 -1.25280392e+00
3.70251387e-02 3.81077021e-01 6.44635737e-01 -2.96556383e-01
-2.63915449e-01 7.47648954e-01 -4.39597368e-01 -7.04563320e-01
-5.04628241e-01 -1.27365187e-01 8.58084187e-02 7.51722455e-02
1.79630250e-01 -4.21685666e-01 -3.88525516e-01 1.95299655e-01
-1.08353496e+00 2.48826388e-02 6.13809407e-01 6.16067529e-01
5.17031252e-01 1.20302528e-01 8.06096315e-01 -8.14785808e-02
1.01022446e+00 -1.10898338e-01 -3.95449728e-01 2.00946569e-01
-2.78810374e-02 -5.15695632e-01 5.21569014e-01 -4.88276660e-01
-1.29082346e+00 -7.90626258e-02 -3.79492611e-01 -3.00386369e-01
-4.38704848e-01 7.50165582e-01 -4.77155209e-01 -7.78004527e-02
3.99231255e-01 2.71623582e-01 7.97679052e-02 -4.93556917e-01
3.63635451e-01 1.04299688e+00 5.92719972e-01 -2.35862479e-01
4.32619274e-01 4.09188658e-01 -8.39519501e-03 -1.54158020e+00
-2.36642554e-01 -3.74811441e-01 -1.89546838e-01 -7.78830886e-01
8.85564148e-01 -7.20390499e-01 -8.93918276e-01 7.47813821e-01
-1.35403049e+00 -2.50411201e-02 1.05093762e-01 7.57465124e-01
-6.24310493e-01 5.92685521e-01 -8.37510645e-01 -8.68172288e-01
-3.60825747e-01 -1.43284965e+00 1.22618520e+00 5.39715707e-01
-2.41770580e-01 -9.24674451e-01 -4.79278900e-02 5.48942268e-01
1.50908083e-01 -1.69612393e-01 9.56754863e-01 -1.72384337e-01
-3.58097523e-01 -1.19841509e-01 -1.40706003e-01 2.81825542e-01
-2.38281805e-02 2.63851255e-01 -1.24368250e+00 -1.52395964e-02
-3.15538675e-01 -3.58168602e-01 7.24933267e-01 7.99979329e-01
1.13538861e+00 1.14052467e-01 -1.40999734e-01 3.34826678e-01
8.30368817e-01 8.59816790e-01 5.74556530e-01 -4.01528440e-02
5.77395916e-01 8.67165029e-01 7.89020300e-01 5.40711164e-01
1.19303018e-01 7.02679574e-01 5.67828953e-01 -2.94951111e-01
-3.42063487e-01 -1.90778509e-01 4.63783354e-01 1.25383520e+00
-1.61145344e-01 -4.55552876e-01 -6.89837456e-01 3.22361946e-01
-1.59827280e+00 -1.01820064e+00 -1.16672367e-01 1.85885072e+00
5.50858140e-01 -4.01760191e-01 2.79168636e-01 6.82778537e-01
1.03780556e+00 4.14501548e-01 -2.51916856e-01 -5.62562943e-01
-3.36801559e-01 7.47082233e-02 -3.86341065e-01 3.79609287e-01
-1.24455416e+00 1.00577307e+00 6.96239901e+00 1.27185607e+00
-1.52559948e+00 -2.36299619e-01 3.04754436e-01 -2.24410146e-01
-9.39901769e-02 -5.98362505e-01 -4.79633600e-01 2.98448801e-01
6.71924531e-01 -1.39701907e-02 5.97100139e-01 4.37524885e-01
5.83429575e-01 -1.03340954e-01 -8.43670547e-01 1.65639842e+00
2.03602538e-01 -1.26718211e+00 3.39079380e-01 -1.90919921e-01
5.65017998e-01 -3.81345481e-01 6.21191502e-01 1.22014873e-01
-5.35854280e-01 -1.01595855e+00 6.93041384e-01 3.84667814e-01
1.00798500e+00 -1.11756063e+00 6.87308550e-01 8.61644968e-02
-1.54201949e+00 -4.18262556e-02 -1.42208606e-01 4.00509313e-02
5.10887563e-01 2.99472034e-01 -6.45558357e-01 5.37903249e-01
7.31444597e-01 7.20929623e-01 -8.17848966e-02 1.34211016e+00
-3.88780862e-01 5.12741983e-01 1.91627368e-01 -3.00425172e-01
1.25385851e-01 -1.80938900e-01 7.11843252e-01 1.44166076e+00
4.07010585e-01 6.31966442e-03 6.45037293e-02 4.50853169e-01
1.79413781e-01 5.36501348e-01 -4.60551262e-01 -5.87290466e-01
3.90726626e-01 1.22760546e+00 -7.11875975e-01 -4.10461456e-01
-5.89403570e-01 6.23919249e-01 -6.21332288e-01 5.68406999e-01
-8.83367300e-01 -8.20979595e-01 5.46867192e-01 -1.03726923e-01
-1.77600145e-01 -2.69948483e-01 -3.85280341e-01 -7.15071321e-01
-3.24214727e-01 -1.14027786e+00 3.33159685e-01 -1.24302638e+00
-1.03097391e+00 4.69280750e-01 7.41626397e-02 -1.37119877e+00
-4.41977739e-01 -6.55206203e-01 -6.27037883e-01 6.85131073e-01
-1.11418045e+00 -1.04123712e+00 -2.11663052e-01 5.55516839e-01
8.83332133e-01 -5.44537961e-01 6.76573694e-01 5.28161645e-01
-3.87761205e-01 4.85543102e-01 -2.75553693e-03 2.51340605e-02
9.76788938e-01 -7.24519134e-01 -1.03566058e-01 2.79403776e-01
3.83976281e-01 4.62490797e-01 7.41868317e-01 -3.06468040e-01
-1.34657228e+00 -6.70898318e-01 9.31428313e-01 3.35823357e-01
5.06283879e-01 -3.96803468e-02 -5.99033654e-01 1.10675812e-01
8.51185203e-01 -6.85926497e-01 9.06068444e-01 -3.44904333e-01
1.33737966e-01 -3.02606136e-01 -7.73742557e-01 9.33876634e-01
4.03352290e-01 -7.52366245e-01 -3.71545196e-01 1.75930411e-01
5.11305869e-01 -3.67823392e-01 -5.35259485e-01 3.99383962e-01
7.59649396e-01 -8.33882630e-01 1.00081646e+00 -3.37769777e-01
4.23155904e-01 -4.16213930e-01 -1.34767756e-01 -1.29461980e+00
-3.03149987e-02 -6.73379004e-01 1.06035866e-01 1.27330446e+00
1.53768212e-01 -1.20794654e-01 5.30721307e-01 -4.57040519e-02
-3.47648829e-01 -3.54525238e-01 -7.49589443e-01 -4.36360568e-01
-2.91480601e-01 -8.03213060e-01 3.53819191e-01 7.31849194e-01
4.32447523e-01 6.14227474e-01 -6.31683171e-01 -1.36265010e-01
2.48780921e-01 1.43688262e-01 7.82181025e-01 -1.04443562e+00
2.71113236e-02 -6.97667658e-01 -3.31482589e-01 -1.04313159e+00
2.91696578e-01 -6.86152041e-01 -1.30428523e-01 -1.81777894e+00
5.31865247e-02 4.61854845e-01 -1.11455776e-01 4.03891712e-01
2.72674829e-01 5.85125327e-01 4.12302643e-01 -9.72918794e-02
-4.96547550e-01 7.21204817e-01 1.56724632e+00 -5.00247896e-01
-3.02239954e-01 3.10830474e-01 -4.24083382e-01 7.65088677e-01
1.04741013e+00 8.68810192e-02 -3.52950275e-01 -3.43333751e-01
2.36092627e-01 4.28933829e-01 7.01909736e-02 -7.93030858e-01
2.35896915e-01 -1.42722756e-01 3.89973968e-01 -1.04600000e+00
8.97257268e-01 -6.68498933e-01 -5.41233197e-02 1.43814817e-01
-4.43085432e-01 8.81581977e-02 3.94210905e-01 2.00568259e-01
-9.05757546e-01 -2.48403370e-01 8.05634737e-01 -4.48248014e-02
-9.40698624e-01 9.95207578e-02 -1.19373214e+00 -3.72599453e-01
8.95733118e-01 -1.50281623e-01 3.30654159e-02 -1.11187744e+00
-1.04686773e+00 5.81629835e-02 -1.13541640e-01 5.97908556e-01
1.03698349e+00 -1.53731930e+00 -5.33650875e-01 1.71524996e-03
9.63354185e-02 -5.79841912e-01 7.04176664e-01 8.46267700e-01
-5.43351352e-01 7.76813686e-01 -2.46189162e-01 -6.76693678e-01
-1.79762840e+00 5.51709771e-01 1.77921895e-02 4.05696630e-01
-1.52318656e-01 5.94811201e-01 4.13922310e-01 4.73209061e-02
8.63983870e-01 -1.53220340e-03 -6.74502134e-01 2.44291171e-01
7.24327683e-01 6.78453982e-01 -2.64989305e-02 -8.38822842e-01
-2.44192228e-01 7.58861601e-01 3.31473947e-01 -5.72703421e-01
8.80824387e-01 -3.10259640e-01 -2.05145687e-01 7.05460787e-01
1.09025538e+00 8.88204351e-02 -6.63000524e-01 -3.78286210e-03
-4.29423630e-01 -1.87105983e-01 1.13213107e-01 -7.57863283e-01
-1.30278611e+00 1.79009521e+00 5.66938579e-01 3.10194969e-01
1.45818603e+00 8.21303204e-02 8.16107035e-01 1.85775459e-01
5.01524359e-02 -1.26529658e+00 5.36205351e-01 6.06978953e-01
1.15699422e+00 -1.03570950e+00 -2.93253213e-01 -3.44525784e-01
-1.01967800e+00 1.42492688e+00 3.44813347e-01 3.95521402e-01
3.97634745e-01 3.84673238e-01 3.94976258e-01 9.70038846e-02
-4.42677915e-01 -2.52520084e-01 6.60526514e-01 8.93735051e-01
8.33922863e-01 -3.58476080e-02 -1.89260438e-01 4.80463296e-01
-1.95802495e-01 -2.22221583e-01 1.72365323e-01 6.06502712e-01
-6.11515462e-01 -1.23655581e+00 -9.27303851e-01 -6.41009510e-02
-4.22679096e-01 -9.03445929e-02 -6.79383993e-01 5.02783895e-01
-4.06292640e-02 1.36189270e+00 8.83469135e-02 -5.48359156e-01
2.36888736e-01 9.39868763e-02 6.52771175e-01 -3.25537384e-01
-4.14731860e-01 9.44603264e-01 2.20139455e-02 -2.57595658e-01
-6.21809423e-01 -4.80833650e-01 -1.43132865e+00 -4.38002318e-01
5.51815098e-03 9.88058448e-02 9.28879619e-01 7.21171558e-01
5.57654016e-02 7.89373994e-01 3.85013640e-01 -1.16323221e+00
7.83211887e-02 -8.13080013e-01 -7.75499821e-01 -2.13796079e-01
2.89226949e-01 -5.83241701e-01 -1.74190566e-01 1.35995731e-01] | [14.327959060668945, 5.025004863739014] |
1c50c41f-d4b1-4cef-a3ef-dae568de0115 | mega-moving-average-equipped-gated-attention | 2209.10655 | null | https://arxiv.org/abs/2209.10655v3 | https://arxiv.org/pdf/2209.10655v3.pdf | Mega: Moving Average Equipped Gated Attention | The design choices in the Transformer attention mechanism, including weak inductive bias and quadratic computational complexity, have limited its application for modeling long sequences. In this paper, we introduce Mega, a simple, theoretically grounded, single-head gated attention mechanism equipped with (exponential) moving average to incorporate inductive bias of position-aware local dependencies into the position-agnostic attention mechanism. We further propose a variant of Mega that offers linear time and space complexity yet yields only minimal quality loss, by efficiently splitting the whole sequence into multiple chunks with fixed length. Extensive experiments on a wide range of sequence modeling benchmarks, including the Long Range Arena, neural machine translation, auto-regressive language modeling, and image and speech classification, show that Mega achieves significant improvements over other sequence models, including variants of Transformers and recent state space models. | ['Luke Zettlemoyer', 'Jonathan May', 'Graham Neubig', 'Liangke Gui', 'Junxian He', 'Xiang Kong', 'Chunting Zhou', 'Xuezhe Ma'] | 2022-09-21 | null | null | null | null | ['long-range-modeling'] | ['natural-language-processing'] | [ 3.22341859e-01 1.86509751e-02 -4.76521939e-01 -2.86936224e-01
-9.38233376e-01 -6.03355110e-01 6.97807014e-01 -9.17041227e-02
-5.27118444e-01 9.13437605e-01 3.04620832e-01 -8.92856777e-01
2.44620487e-01 -3.74189645e-01 -1.14575779e+00 -6.03259563e-01
-3.25913489e-01 7.89032161e-01 3.37947905e-01 -4.41588938e-01
3.29334080e-01 1.91969454e-01 -1.10510349e+00 2.62394041e-01
8.17109346e-01 7.36926138e-01 6.84992611e-01 1.15230536e+00
-1.96618989e-01 1.28981924e+00 -4.10714179e-01 -3.14213037e-01
3.30855474e-02 -5.00748992e-01 -1.12340629e+00 -3.67901355e-01
1.53759539e-01 -3.13298672e-01 -4.42569584e-01 6.07760906e-01
6.85776412e-01 3.07125241e-01 3.19450915e-01 -1.05746675e+00
-1.01622081e+00 1.13936651e+00 -5.53044856e-01 9.08633351e-01
5.67116141e-02 4.50794041e-01 1.30172837e+00 -7.50677228e-01
4.28229034e-01 1.29994166e+00 8.25849473e-01 5.58877587e-01
-1.24083352e+00 -4.36430156e-01 6.70271516e-01 4.68085349e-01
-1.03684378e+00 -4.11849618e-01 2.13174269e-01 -3.65338594e-01
1.99655592e+00 3.14911932e-01 3.89523298e-01 1.26287973e+00
5.30244291e-01 1.09682798e+00 6.98037207e-01 -4.29818690e-01
1.76899076e-01 -5.94534695e-01 5.11016965e-01 6.26690090e-01
-1.71348646e-01 -4.83298814e-03 -4.64976579e-01 -2.86241651e-01
6.06863320e-01 -2.09590234e-02 -1.37670591e-01 -1.59935504e-01
-1.19249701e+00 7.50360131e-01 3.30542624e-01 1.48797721e-01
-5.36841154e-01 6.18766248e-01 5.49136758e-01 3.00843835e-01
6.29301786e-01 2.42236763e-01 -9.70744908e-01 -6.71757281e-01
-7.17538595e-01 5.74595928e-02 6.80867255e-01 1.27718925e+00
6.03612542e-01 1.14708304e-01 -4.02946711e-01 5.65303504e-01
1.35650784e-02 4.80102539e-01 7.56044507e-01 -7.97443569e-01
5.45736969e-01 2.33966202e-01 3.92544828e-03 -3.65494996e-01
-4.62566733e-01 -5.40290058e-01 -7.08072841e-01 -3.62313330e-01
1.15294326e-02 -1.12322755e-02 -1.15544450e+00 2.14823532e+00
-1.02459699e-01 5.16786814e-01 -8.53806660e-02 8.37992609e-01
3.36656332e-01 1.03681076e+00 2.43984595e-01 -3.57120931e-01
1.09087443e+00 -1.44127631e+00 -4.78764802e-01 -6.75423324e-01
7.85638750e-01 -4.51506585e-01 1.28901768e+00 6.05024472e-02
-1.54717851e+00 -3.54599476e-01 -8.46227109e-01 -3.84143680e-01
-1.14464574e-01 -4.75072026e-01 7.01691329e-01 4.17899072e-01
-1.17665005e+00 5.34455121e-01 -1.17126584e+00 -3.46360385e-01
1.61283836e-01 4.55740064e-01 4.28840257e-02 1.74506515e-01
-1.36605763e+00 1.16413939e+00 2.58957326e-01 -1.88825518e-01
-1.13381112e+00 -8.82303774e-01 -8.39308381e-01 5.25224805e-01
3.20730269e-01 -8.95883501e-01 1.76888669e+00 -1.02147138e+00
-1.70476925e+00 3.54104608e-01 -4.23116744e-01 -1.10900021e+00
2.40287736e-01 -5.40522397e-01 1.44956186e-01 -2.82667339e-01
-1.22836947e-01 6.47898853e-01 4.05800611e-01 -6.37013793e-01
-5.07830441e-01 -2.79274434e-01 -2.85855005e-03 2.82350779e-01
-1.23116449e-01 3.50831091e-01 -5.19794881e-01 -6.59436464e-01
-5.51355720e-01 -1.11559367e+00 -5.76872647e-01 -6.42055035e-01
-1.52292654e-01 -2.30421767e-01 5.71699023e-01 -8.20946872e-01
1.46303666e+00 -1.62527740e+00 5.59091747e-01 -3.41135263e-01
-1.57478675e-01 2.38751069e-01 -5.30928075e-01 4.46841598e-01
-7.34840333e-02 1.91071883e-01 -3.71711195e-01 -3.81684065e-01
-1.13519177e-01 4.93931919e-01 -3.30090165e-01 2.50611365e-01
2.29971677e-01 1.50977218e+00 -1.01379263e+00 -3.15275788e-01
-4.84783240e-02 2.33020365e-01 -8.85180235e-01 1.12720534e-01
-7.05082238e-01 1.22259416e-01 -2.21862137e-01 3.02082956e-01
2.58305192e-01 -6.56508386e-01 2.39360675e-01 4.38816071e-01
1.60834268e-02 7.01346636e-01 -3.45053107e-01 1.89582419e+00
-6.51941538e-01 6.97660744e-01 -1.88871194e-02 -8.03337634e-01
4.83399361e-01 1.17204919e-01 3.00228059e-01 -7.26961255e-01
6.48783967e-02 -5.42506352e-02 1.61614507e-01 -3.01961690e-01
6.29941881e-01 2.92942692e-02 -1.11567356e-01 5.75264692e-01
1.48204908e-01 3.88582289e-01 1.48554444e-01 3.67786527e-01
1.37584543e+00 3.51143450e-01 3.05135816e-01 -3.53788346e-01
2.51345038e-01 -5.85124195e-02 6.45414352e-01 9.32349980e-01
-1.84313908e-01 4.79213774e-01 3.97096634e-01 -4.99002308e-01
-1.54510796e+00 -7.89926112e-01 4.84991759e-01 1.96182251e+00
1.19507462e-02 -4.22576427e-01 -7.47913837e-01 -5.38545907e-01
-1.99340358e-01 9.28330600e-01 -7.30167866e-01 -2.94248968e-01
-1.09944975e+00 -8.95741820e-01 7.02786326e-01 9.91531730e-01
3.85994464e-02 -1.15836298e+00 -7.33876884e-01 5.08560359e-01
-5.29968739e-01 -8.71691823e-01 -9.98673320e-01 4.97780204e-01
-8.21594596e-01 -5.71455717e-01 -8.24812472e-01 -7.42188334e-01
1.09906614e-01 1.37912109e-02 1.46636093e+00 -8.28069523e-02
-4.58588302e-02 7.06276521e-02 -1.98706374e-01 -2.57519960e-01
-4.95611608e-01 5.97581983e-01 -8.55847001e-02 -3.89123321e-01
7.78755993e-02 -4.99405265e-01 -4.42324370e-01 1.90897390e-01
-5.90064704e-01 1.31670207e-01 8.47655714e-01 1.30808580e+00
1.44204095e-01 -7.91918874e-01 6.31984174e-01 -6.74534798e-01
5.65838158e-01 -6.03148937e-01 -4.85260189e-01 3.62627268e-01
-5.92988968e-01 4.09151971e-01 6.66573763e-01 -5.86139977e-01
-1.12620473e+00 -1.85342550e-01 -3.02615970e-01 -5.13376176e-01
3.14104974e-01 6.14340127e-01 7.06544295e-02 2.59635091e-01
6.53948963e-01 8.38746250e-01 3.50972831e-01 -4.25045669e-01
4.89198267e-01 4.64156538e-01 4.70815301e-01 -6.41920805e-01
9.29259062e-02 1.53943477e-02 -2.41231784e-01 -6.16314948e-01
-6.35873973e-01 -4.24372703e-01 -3.84458154e-01 2.97557533e-01
6.62164927e-01 -8.07496905e-01 -8.39403272e-01 4.54318911e-01
-1.45014250e+00 -1.05388367e+00 -6.28528073e-02 2.62957215e-01
-1.09960282e+00 5.46995699e-01 -1.28482151e+00 -8.38141739e-01
-7.84035742e-01 -1.26039088e+00 1.16478097e+00 -9.01359543e-02
-2.90171564e-01 -7.60938406e-01 2.61011600e-01 1.62299812e-01
7.13052273e-01 -5.15443027e-01 1.23553503e+00 -8.45960677e-01
-7.68130541e-01 2.06836417e-01 4.99244779e-02 2.26196736e-01
-3.55765313e-01 -9.26459953e-02 -5.62201321e-01 -3.89967471e-01
-4.54951882e-01 -3.37842643e-01 9.42410469e-01 5.84682882e-01
1.02565527e+00 -4.58865285e-01 -3.75468194e-01 5.95306218e-01
1.01620138e+00 3.35066348e-01 6.37590647e-01 4.62223649e-01
5.25900483e-01 3.00136283e-02 4.62710321e-01 4.74950850e-01
4.15427923e-01 7.04069436e-01 3.73459339e-01 1.71449631e-01
2.36326978e-02 -1.90638408e-01 6.27529383e-01 1.07356071e+00
-1.93099737e-01 -7.24298179e-01 -9.18989956e-01 7.29888380e-01
-2.17121649e+00 -1.23573935e+00 2.17237025e-01 1.96917355e+00
8.15708876e-01 1.68916777e-01 1.75461218e-01 -3.26462328e-01
6.93581522e-01 1.73442394e-01 -6.20394051e-01 -8.54871452e-01
-1.10993162e-01 1.91994891e-01 8.37504327e-01 5.90567589e-01
-8.77782524e-01 1.19512987e+00 7.64193630e+00 8.51471484e-01
-1.10063422e+00 2.97472060e-01 8.29399228e-01 -3.83688629e-01
-1.62314877e-01 -1.67464286e-01 -7.66442418e-01 6.16756260e-01
1.60679853e+00 -3.93895537e-01 5.86645186e-01 8.17472160e-01
1.68560773e-01 4.30384487e-01 -1.06902754e+00 4.26914483e-01
-4.38010655e-02 -1.57963932e+00 1.24041885e-01 -4.44572717e-02
7.51824141e-01 6.55980110e-01 1.64799854e-01 5.60246885e-01
7.46821940e-01 -9.23123240e-01 7.45429277e-01 2.56172776e-01
5.55883288e-01 -9.36642408e-01 6.52626157e-01 6.38527453e-01
-9.21214759e-01 -5.33529878e-01 -3.72191817e-01 -3.70533317e-01
5.13352811e-01 3.16861421e-02 -8.29537392e-01 4.28131938e-01
8.36367905e-01 4.91803616e-01 -1.53362617e-01 6.75232649e-01
1.61108077e-01 1.00408340e+00 -2.88693309e-01 -2.60550529e-01
7.58858919e-01 1.17753260e-01 6.44992769e-01 1.55352318e+00
2.05336645e-01 2.34549120e-01 1.31235629e-01 5.50232410e-01
-5.55403531e-02 1.25035224e-02 -3.87467980e-01 7.95139652e-03
4.21051502e-01 7.13379025e-01 -4.54939455e-01 -6.44038320e-01
-4.93807167e-01 9.73658144e-01 5.50127983e-01 3.05831224e-01
-1.27450275e+00 -2.17734780e-02 9.12137687e-01 -2.77984198e-02
7.58458495e-01 -3.22429299e-01 -3.72944117e-01 -9.87094760e-01
-2.85677880e-01 -1.12175024e+00 5.06286502e-01 -8.65440667e-01
-8.46866727e-01 8.03244054e-01 -1.40617043e-01 -5.39573967e-01
-5.71472704e-01 -4.20307815e-01 -5.91465950e-01 9.62207675e-01
-1.29735899e+00 -1.07345390e+00 1.69708341e-01 2.94727147e-01
1.11682105e+00 1.08038830e-02 6.41453266e-01 3.24511677e-01
-6.46327615e-01 8.27514946e-01 2.46000081e-01 -7.13077486e-02
3.31767768e-01 -1.19364381e+00 1.27742481e+00 7.85887301e-01
-9.46597755e-02 7.79245496e-01 7.61729717e-01 -6.26034617e-01
-1.53661168e+00 -1.18098450e+00 1.03635895e+00 -5.48515499e-01
7.71289468e-01 -3.81361634e-01 -1.06370127e+00 1.37829912e+00
5.76707304e-01 -2.50151962e-01 4.12887245e-01 1.88511848e-01
-3.66819769e-01 4.02548909e-01 -6.46161914e-01 8.36714089e-01
1.45333838e+00 -2.49659628e-01 -6.60429895e-01 2.04802766e-01
1.33763754e+00 -3.79870147e-01 -7.11743772e-01 5.00505090e-01
5.08296430e-01 -6.08384490e-01 1.01714516e+00 -1.21108353e+00
5.06766260e-01 1.28133968e-01 -7.72829950e-02 -1.36539078e+00
-1.02910471e+00 -9.96804297e-01 -4.46780235e-01 8.06735992e-01
7.30158091e-01 -7.02090025e-01 8.98911476e-01 3.32678407e-01
-6.04409635e-01 -1.00570655e+00 -8.34453702e-01 -1.06073105e+00
3.30658019e-01 -2.93112963e-01 6.54626131e-01 6.09147727e-01
1.62307292e-01 8.10908318e-01 -6.93916261e-01 2.54538283e-03
3.37045938e-01 1.54644623e-01 5.28576434e-01 -6.13033235e-01
-6.86159372e-01 -7.73018599e-01 -2.75797397e-01 -1.71265578e+00
2.44756907e-01 -8.72623324e-01 3.03041101e-01 -1.38123178e+00
5.19221306e-01 -2.30327427e-01 -3.56947869e-01 4.93726373e-01
-4.17770743e-01 -1.25401199e-01 1.92743808e-01 1.23817306e-02
-8.52808058e-01 7.11294472e-01 9.04969394e-01 -9.68241319e-02
1.71386659e-01 -1.19998224e-01 -5.08626878e-01 4.73064691e-01
7.60774314e-01 -2.84876674e-01 -3.84914875e-01 -6.81964040e-01
1.27529666e-01 1.23517476e-01 -6.96953312e-02 -5.24287522e-01
2.32722744e-01 -3.49412322e-01 1.08895786e-02 -5.86794853e-01
4.05753523e-01 -3.84881794e-01 3.24221747e-03 6.75617993e-01
-8.50856245e-01 6.78389311e-01 3.23638499e-01 5.32652855e-01
1.28425807e-01 1.35152027e-01 6.64510608e-01 -2.48369873e-01
-8.15281212e-01 5.01803041e-01 -6.36088192e-01 1.92370325e-01
1.00746775e+00 8.07703286e-02 -4.00459468e-01 -5.61851621e-01
-6.73636854e-01 4.83820945e-01 2.90621161e-01 6.00208580e-01
1.58724129e-01 -9.94551599e-01 -8.78700137e-01 -1.65740192e-01
-2.29739565e-02 -2.35506058e-01 3.51424217e-01 8.90573442e-01
-3.49900723e-01 9.34460402e-01 -7.98035339e-02 -6.95817769e-01
-1.34158003e+00 1.12681520e+00 2.06295833e-01 -6.60018206e-01
-6.05011702e-01 9.54610944e-01 3.99306595e-01 -5.74703634e-01
1.83020562e-01 -5.16740203e-01 3.52457821e-01 -5.38161099e-01
6.28965676e-01 5.26882291e-01 1.92814860e-02 -5.61200678e-01
-5.59612572e-01 1.60055369e-01 -2.09673673e-01 -3.02435993e-03
1.24875116e+00 -2.57994384e-01 2.27253395e-03 4.97004956e-01
1.12180293e+00 -5.30676782e-01 -1.38210928e+00 -2.96106547e-01
1.32259563e-01 -1.73577275e-02 -2.31231734e-01 -9.10615504e-01
-6.57359123e-01 9.64811206e-01 1.30215898e-01 1.64600253e-01
9.68679667e-01 4.99063246e-02 1.03814936e+00 4.69587922e-01
4.12703872e-01 -7.12779343e-01 -6.38774559e-02 1.13088334e+00
6.77164614e-01 -8.63298357e-01 -4.60581392e-01 7.21603353e-03
-7.10510969e-01 6.38258994e-01 7.50266492e-01 1.06661089e-01
1.81443512e-01 7.19722629e-01 -1.39017314e-01 2.56248534e-01
-1.74370253e+00 -2.23069564e-01 -1.65855378e-01 3.67881119e-01
6.10250175e-01 -6.34895451e-03 -2.31289193e-01 5.27733028e-01
-3.39784995e-02 1.02024429e-01 3.88168067e-01 1.03514624e+00
-6.60054505e-01 -9.15872335e-01 -9.21922997e-02 2.89441556e-01
-7.74053693e-01 -6.27431870e-01 -9.06350985e-02 5.32457650e-01
-4.26152259e-01 4.84126151e-01 4.23585325e-01 -1.95177361e-01
1.07782982e-01 4.13707048e-01 5.64393997e-01 -3.62774938e-01
-8.92255843e-01 -5.76455668e-02 2.40992174e-01 -5.84893227e-01
1.05295308e-01 -4.89595085e-01 -1.21439552e+00 -5.59903264e-01
-3.51305366e-01 2.09120780e-01 4.29385960e-01 8.88596892e-01
9.50142086e-01 5.96644223e-01 3.20313513e-01 -9.39265728e-01
-1.08693337e+00 -1.27825820e+00 1.43350199e-01 1.74858019e-01
4.58640158e-01 -3.72220635e-01 -2.23464202e-02 6.06225543e-02] | [10.854175567626953, 6.99728536605835] |
af17c3d1-c6ff-491a-87e4-0eccbdd2a4f1 | learning-diverse-tone-styles-for-image | 2207.05430 | null | https://arxiv.org/abs/2207.05430v2 | https://arxiv.org/pdf/2207.05430v2.pdf | Learning Diverse Tone Styles for Image Retouching | Image retouching, aiming to regenerate the visually pleasing renditions of given images, is a subjective task where the users are with different aesthetic sensations. Most existing methods deploy a deterministic model to learn the retouching style from a specific expert, making it less flexible to meet diverse subjective preferences. Besides, the intrinsic diversity of an expert due to the targeted processing on different images is also deficiently described. To circumvent such issues, we propose to learn diverse image retouching with normalizing flow-based architectures. Unlike current flow-based methods which directly generate the output image, we argue that learning in a style domain could (i) disentangle the retouching styles from the image content, (ii) lead to a stable style presentation form, and (iii) avoid the spatial disharmony effects. For obtaining meaningful image tone style representations, a joint-training pipeline is delicately designed, which is composed of a style encoder, a conditional RetouchNet, and the image tone style normalizing flow (TSFlow) module. In particular, the style encoder predicts the target style representation of an input image, which serves as the conditional information in the RetouchNet for retouching, while the TSFlow maps the style representation vector into a Gaussian distribution in the forward pass. After training, the TSFlow can generate diverse image tone style vectors by sampling from the Gaussian distribution. Extensive experiments on MIT-Adobe FiveK and PPR10K datasets show that our proposed method performs favorably against state-of-the-art methods and is effective in generating diverse results to satisfy different human aesthetic preferences. Source code and pre-trained models are publicly available at https://github.com/SSRHeart/TSFlow. | ['WangMeng Zuo', 'Xiaohe Wu', 'Ming Liu', 'Jiawei Zhang', 'Haolin Wang'] | 2022-07-12 | null | null | null | null | ['image-retouching'] | ['computer-vision'] | [ 3.35654914e-01 -1.97354227e-01 8.31005350e-03 -3.58740509e-01
-4.83857602e-01 -7.09024429e-01 4.01098043e-01 -4.56971139e-01
-7.30105862e-02 4.60693538e-01 3.03561240e-01 -1.66609243e-01
3.08702528e-01 -7.57988274e-01 -7.10920691e-01 -6.89290762e-01
8.46820056e-01 -2.04945914e-02 -4.22127210e-02 -3.59348238e-01
3.15521628e-01 2.58661419e-01 -1.39572895e+00 4.67786282e-01
1.03696179e+00 1.12784076e+00 4.23508912e-01 8.36890280e-01
-2.91122466e-01 6.89339101e-01 -7.18439579e-01 -6.68480396e-01
2.57177413e-01 -7.07197070e-01 -5.11752725e-01 2.94970810e-01
5.09578049e-01 -3.72053266e-01 -3.26500386e-01 1.29754114e+00
5.41105986e-01 2.07813270e-02 7.35104799e-01 -1.03922558e+00
-1.28153300e+00 3.98694396e-01 -7.18939662e-01 -2.57990986e-01
2.99110413e-01 6.21065557e-01 1.00872529e+00 -7.88648427e-01
5.51151812e-01 1.39052439e+00 2.50989020e-01 8.39771807e-01
-1.59260428e+00 -8.63038957e-01 3.56955156e-02 9.38861817e-02
-1.40220225e+00 -3.60515445e-01 1.09457564e+00 -3.85479331e-01
1.45734161e-01 5.05411446e-01 8.19024980e-01 1.49672413e+00
2.01875821e-01 9.62451875e-01 1.09554589e+00 -1.59604847e-01
3.45843583e-01 3.12705100e-01 -4.65970188e-01 2.67269313e-01
-4.66479659e-02 8.79697278e-02 -4.97500837e-01 1.84188187e-01
1.13601089e+00 -4.74344045e-02 -4.64851230e-01 -3.84725153e-01
-1.06032240e+00 7.07579315e-01 5.29862583e-01 8.74773860e-02
-1.50883466e-01 1.41505897e-01 2.94225276e-01 1.71677619e-01
2.98486382e-01 5.53712785e-01 -2.31145918e-01 -5.13690971e-02
-8.19673479e-01 2.66694605e-01 5.98825097e-01 1.05104887e+00
8.41429114e-01 2.41860479e-01 -6.33580506e-01 9.32824671e-01
1.50592625e-01 5.85265756e-01 5.34418166e-01 -9.83593881e-01
3.27188373e-01 4.57907170e-01 1.31340772e-01 -1.01273727e+00
1.11470468e-01 -3.50697964e-01 -9.00245786e-01 5.39766908e-01
2.54619718e-01 -2.00018838e-01 -1.01589262e+00 1.72695458e+00
-2.60849320e-03 -7.80964792e-02 -6.62375614e-02 1.13905489e+00
8.23447049e-01 8.35980356e-01 2.04138368e-01 1.57066807e-01
1.30803394e+00 -8.90377998e-01 -7.87986815e-01 -3.93218398e-01
2.37979770e-01 -1.13695657e+00 1.75200546e+00 4.33638036e-01
-1.16897011e+00 -8.87621701e-01 -1.08628428e+00 -2.85578489e-01
-1.25734612e-01 4.42607671e-01 3.64290297e-01 6.08127892e-01
-1.07030106e+00 4.63924050e-01 -3.21025342e-01 -2.17470512e-01
4.92241263e-01 -1.28101110e-01 -1.20981306e-01 -4.65682261e-02
-1.03544867e+00 6.69293940e-01 3.22526187e-01 4.42316495e-02
-6.84440494e-01 -7.25650012e-01 -9.63973701e-01 1.24765329e-01
2.43523508e-01 -8.27976584e-01 1.18856120e+00 -1.60446775e+00
-1.98321605e+00 7.09410071e-01 -2.93720048e-02 1.34153768e-01
7.27494180e-01 -5.98889776e-02 -5.95085740e-01 -3.85909826e-02
-1.70008261e-02 8.95203769e-01 1.22912812e+00 -1.58896613e+00
-4.56673115e-01 6.69847131e-02 -7.35081080e-03 3.95175695e-01
-3.49563658e-01 -3.51743191e-01 -5.57576954e-01 -1.11753643e+00
-2.43545279e-01 -7.97393024e-01 -5.86996600e-02 2.12462530e-01
-6.01949155e-01 2.46036544e-01 7.83387780e-01 -4.98620510e-01
1.32370508e+00 -2.39800334e+00 2.47449532e-01 3.50114256e-02
1.12560689e-01 2.93354094e-01 -3.55835974e-01 3.71129178e-02
-2.37572715e-01 1.46754682e-01 -1.49355188e-01 -3.48605931e-01
-2.66003218e-02 2.80357595e-03 -3.85674179e-01 8.91303644e-02
3.15524668e-01 1.01077282e+00 -9.51859355e-01 -2.24636242e-01
3.90052229e-01 4.68479007e-01 -6.31990492e-01 4.69728887e-01
-3.47461432e-01 5.09541333e-01 -3.71220142e-01 4.38215256e-01
9.17525887e-01 -1.96384013e-01 -2.39174236e-02 -6.58360541e-01
-2.38144938e-02 -1.49926573e-01 -1.24596357e+00 1.88297522e+00
-6.08491719e-01 5.60115874e-01 -1.57027453e-01 -3.90116662e-01
1.08793497e+00 1.04645640e-01 1.49005847e-02 -8.88141096e-01
3.46094370e-01 1.77763388e-01 -2.37441003e-01 -5.25012314e-01
7.06341207e-01 -3.21617365e-01 -2.06486210e-01 3.71910751e-01
6.18522353e-02 -4.32143450e-01 8.86383280e-02 6.62898198e-02
5.54044068e-01 2.30465144e-01 1.06748663e-01 -1.52268022e-01
4.97329742e-01 -3.47216636e-01 5.45338809e-01 4.60566014e-01
-1.08635873e-01 1.15127671e+00 4.93762851e-01 -4.87473547e-01
-1.25772500e+00 -1.45250559e+00 1.09832190e-01 1.06099594e+00
4.89236474e-01 -2.78931171e-01 -8.36625457e-01 -5.33543944e-01
-3.77372175e-01 9.75406647e-01 -7.00084865e-01 -3.74011993e-01
-2.54907966e-01 -2.95413643e-01 4.00892347e-01 2.82237142e-01
6.51042700e-01 -1.22407329e+00 -3.81088972e-01 2.62519941e-02
-2.93313354e-01 -7.74055481e-01 -1.17065144e+00 -1.28147170e-01
-4.68544543e-01 -7.77294338e-01 -1.07291341e+00 -6.88233018e-01
7.16144681e-01 3.48535448e-01 1.02635992e+00 -1.88953921e-01
-2.26392999e-01 1.38665047e-02 -4.05910015e-01 -2.37606928e-01
-5.11008799e-01 5.22318482e-02 -3.06400150e-01 3.78437936e-01
1.64285883e-01 -5.19110322e-01 -1.07288218e+00 3.00249398e-01
-1.43697906e+00 5.53709686e-01 6.09729588e-01 9.01121199e-01
4.94001895e-01 2.95205843e-02 3.42951745e-01 -9.24421787e-01
6.73997402e-01 -2.26096019e-01 -3.33569020e-01 2.90449262e-01
-2.89292902e-01 1.16456471e-01 7.77305484e-01 -7.93607771e-01
-1.47679269e+00 1.50721878e-01 -1.08923428e-01 -6.50044799e-01
-2.28585601e-01 -2.88603548e-02 -4.70448434e-01 5.19737303e-02
8.84366453e-01 2.55151719e-01 -4.37967554e-02 -3.66184384e-01
7.78436959e-01 6.63520813e-01 5.14637709e-01 -6.21459424e-01
9.00101066e-01 4.03677821e-01 -5.13340175e-01 -4.68915224e-01
-7.37121403e-01 1.02477558e-01 -3.15658778e-01 -3.42608362e-01
9.92323577e-01 -9.10738051e-01 -4.05612916e-01 5.84771693e-01
-1.09470534e+00 -4.96693403e-01 -4.52675015e-01 8.69638100e-02
-5.64414024e-01 2.10024998e-01 -5.05375445e-01 -5.58590114e-01
-2.46648252e-01 -1.39097154e+00 1.10017145e+00 7.37441301e-01
-4.18780386e-01 -8.27634096e-01 -4.46127690e-02 2.57008523e-01
6.40119374e-01 1.26849309e-01 9.40161228e-01 1.32313564e-01
-4.50700074e-01 4.37292755e-02 -4.88572657e-01 5.03311396e-01
4.72314805e-01 1.47900313e-01 -1.21013939e+00 -1.41388312e-01
-8.83729905e-02 -2.91161507e-01 6.29681587e-01 2.98506230e-01
1.42777181e+00 -4.14333791e-01 2.56740928e-01 8.92146289e-01
1.41803706e+00 2.88201839e-01 1.05600905e+00 2.14147583e-01
9.24575686e-01 5.54864407e-01 2.48212636e-01 4.52315986e-01
2.56714851e-01 5.45437038e-01 2.01785326e-01 -5.42859793e-01
-5.01438379e-01 -5.78748584e-01 4.06783581e-01 4.04937416e-01
2.22328499e-01 -3.13545883e-01 -2.61832446e-01 3.06058526e-01
-1.59998131e+00 -1.03708458e+00 2.91873217e-01 2.09288692e+00
1.17281485e+00 7.31724203e-02 2.49539409e-02 -5.55895008e-02
9.10847843e-01 3.14905286e-01 -8.24339092e-01 -6.66216016e-01
-2.96123952e-01 1.85034767e-01 3.32298011e-01 3.79128069e-01
-7.20870078e-01 1.07052386e+00 5.32603359e+00 1.17174351e+00
-1.58447158e+00 -1.74985155e-01 1.06596756e+00 -2.26171657e-01
-7.79434919e-01 -8.87809247e-02 -3.80278856e-01 6.93921804e-01
2.47387439e-01 -2.29771689e-01 7.32892334e-01 7.42452562e-01
3.44379693e-01 3.28075290e-02 -8.40645015e-01 1.36661220e+00
1.23409197e-01 -1.18774164e+00 4.14916962e-01 -2.59408295e-01
8.01728070e-01 -5.09496331e-01 5.78564346e-01 1.85470477e-01
2.46213779e-01 -1.01073194e+00 1.21381080e+00 6.17908418e-01
1.27788186e+00 -6.35160685e-01 2.34419242e-01 -1.68696836e-01
-1.13620043e+00 8.37656483e-02 -3.72846395e-01 1.43188566e-01
1.69040769e-01 5.84964395e-01 -4.68429774e-01 3.90852809e-01
6.50829494e-01 5.59319198e-01 -7.58653402e-01 9.56230819e-01
-4.24385339e-01 3.21366966e-01 2.07725093e-01 1.37530118e-01
-1.01742465e-02 -2.71191955e-01 4.54017580e-01 1.18298292e+00
4.11943406e-01 -6.89250007e-02 -1.26358539e-01 1.33096659e+00
-8.54515955e-02 1.62460670e-01 -3.71654838e-01 4.79193125e-03
2.63034284e-01 1.45139754e+00 -5.84002793e-01 -2.07818329e-01
-2.53191710e-01 1.45748281e+00 1.04268968e-01 7.28020787e-01
-8.20172429e-01 -5.66811562e-01 7.86506355e-01 7.46921897e-02
1.76021129e-01 1.23102047e-01 -7.75392771e-01 -1.39928293e+00
-8.48435517e-03 -9.98341143e-01 1.82649866e-01 -1.27992809e+00
-1.47188544e+00 9.22597289e-01 -3.24751884e-01 -1.53053355e+00
1.40247852e-01 -5.26500881e-01 -8.89396966e-01 1.17962933e+00
-1.35112178e+00 -1.19622433e+00 -5.46846807e-01 5.53883135e-01
7.79228866e-01 -3.57880034e-02 5.55724919e-01 2.47992188e-01
-5.94540119e-01 7.33162940e-01 -1.43086419e-01 -1.91112142e-02
1.03063428e+00 -1.25599229e+00 3.57761443e-01 8.56781781e-01
-7.74592683e-02 5.10698318e-01 8.58524740e-01 -5.73964596e-01
-1.11745012e+00 -1.15445530e+00 3.23248625e-01 -1.48003772e-01
4.34882164e-01 -4.61943150e-01 -9.20952141e-01 2.28671700e-01
3.22647989e-01 -2.16474861e-01 6.01395190e-01 -3.47384810e-01
-5.99917412e-01 -2.31198117e-01 -9.84353244e-01 1.14561737e+00
8.31420243e-01 -4.66888040e-01 -2.21818909e-01 -2.18262911e-01
6.08398438e-01 -4.59692955e-01 -4.48175281e-01 8.87761712e-02
6.85522497e-01 -1.08081651e+00 7.21379280e-01 -1.83543891e-01
7.77469873e-01 -5.59134424e-01 -7.53441527e-02 -1.60507810e+00
-6.06348932e-01 -6.94092095e-01 2.91151375e-01 1.49595845e+00
4.58645523e-01 -2.08261028e-01 6.14201128e-01 8.24188471e-01
9.69799757e-02 -5.85639000e-01 -3.88832629e-01 -4.28925157e-01
3.47536393e-02 -3.38790447e-01 8.40215206e-01 7.06278145e-01
-2.70535618e-01 4.34370637e-01 -5.93945384e-01 -1.27338499e-01
3.90013248e-01 1.15190998e-01 7.79079437e-01 -7.56582618e-01
-4.77493852e-01 -8.09330761e-01 -3.25575210e-02 -9.67469394e-01
3.38197537e-02 -6.53508842e-01 1.37012437e-01 -1.32215226e+00
1.61633477e-01 -3.67332816e-01 -8.36164728e-02 2.74925947e-01
-4.44597155e-01 3.93132925e-01 4.03335959e-01 1.23076864e-01
-3.87244254e-01 8.02028656e-01 1.87382162e+00 -2.80900180e-01
-3.56488109e-01 -9.86290276e-02 -1.32039058e+00 5.79027057e-01
7.80113935e-01 -5.05978093e-02 -7.59455800e-01 -6.60401583e-01
2.49866754e-01 -7.51402378e-02 3.26395124e-01 -7.64861822e-01
-9.55418646e-02 -4.03421849e-01 7.83188820e-01 -1.57436728e-01
2.95697570e-01 -5.74080229e-01 2.44131953e-01 1.84354052e-01
-5.27125001e-01 -6.68840259e-02 7.29238540e-02 4.26106751e-01
-1.08049363e-01 -9.11737308e-02 1.08704019e+00 -1.17124602e-01
-6.95878565e-01 3.79565537e-01 -1.57206669e-01 1.14785265e-02
7.64964342e-01 -2.85460025e-01 -2.92914778e-01 -5.47878504e-01
-4.25436884e-01 -5.98368160e-02 8.72236252e-01 7.00848877e-01
6.95549428e-01 -1.55782652e+00 -5.93138456e-01 3.45785886e-01
2.30684325e-01 3.14624272e-02 6.47425234e-01 2.31240526e-01
-4.92227793e-01 -2.79199660e-01 -4.34373736e-01 -3.76486689e-01
-8.40315342e-01 6.12800300e-01 2.73878425e-01 1.84179276e-01
-5.03989756e-01 7.51851439e-01 8.99943650e-01 -1.10344961e-01
1.75721869e-02 -2.04062581e-01 -1.08024947e-01 -5.72546422e-02
7.47992456e-01 1.31170765e-01 -1.04833446e-01 -5.89826941e-01
1.70420036e-01 5.84879100e-01 -1.97644114e-01 -2.51788437e-01
9.70343530e-01 -3.70650172e-01 1.82482660e-01 3.23269814e-01
1.33259523e+00 -2.01612525e-03 -1.75753713e+00 -1.33541241e-01
-4.56851810e-01 -7.32167363e-01 5.38103282e-03 -8.98181617e-01
-1.27419686e+00 8.37541938e-01 6.22601032e-01 5.69183752e-02
1.55864429e+00 -2.48326868e-01 8.65525603e-01 -1.95244789e-01
1.37185261e-01 -1.04315114e+00 4.39264804e-01 2.27296606e-01
1.10871387e+00 -1.02970386e+00 -4.79627907e-01 -3.31606358e-01
-1.25926089e+00 1.16480803e+00 7.98209190e-01 -1.88294739e-01
2.55139619e-01 9.91525874e-02 5.19014835e-01 5.73828742e-02
-4.80450392e-01 -6.79016560e-02 4.45800483e-01 6.32688999e-01
4.69318658e-01 1.50153324e-01 3.19326520e-02 7.80591846e-01
-4.40743536e-01 -7.09646940e-03 6.46122932e-01 3.58284950e-01
-1.93862632e-01 -1.13078320e+00 -3.73003215e-01 2.17476040e-01
-2.03876257e-01 -9.10977349e-02 -3.23528022e-01 2.76806235e-01
2.04496384e-01 8.92900825e-01 -1.45572936e-02 -6.09003961e-01
5.14745176e-01 -2.12962598e-01 3.77327114e-01 -5.14423609e-01
-3.83931607e-01 2.94943303e-01 -3.58002603e-01 -6.17051959e-01
5.26332371e-02 -3.01860064e-01 -7.66178906e-01 -3.78430039e-01
8.02580938e-02 -1.30159363e-01 3.84735286e-01 6.16397798e-01
3.81471694e-01 6.98372602e-01 8.47324371e-01 -1.01537740e+00
-1.72796816e-01 -8.40917110e-01 -7.23606467e-01 8.15482795e-01
2.43859425e-01 -4.87641901e-01 -3.30687314e-01 3.52986723e-01] | [11.5408296585083, -0.6135421991348267] |
e570e99e-9858-4c9e-a7ce-be01b983542c | se-bridge-speech-enhancement-with-consistent | 2305.13796 | null | https://arxiv.org/abs/2305.13796v1 | https://arxiv.org/pdf/2305.13796v1.pdf | SE-Bridge: Speech Enhancement with Consistent Brownian Bridge | We propose SE-Bridge, a novel method for speech enhancement (SE). After recently applying the diffusion models to speech enhancement, we can achieve speech enhancement by solving a stochastic differential equation (SDE). Each SDE corresponds to a probabilistic flow ordinary differential equation (PF-ODE), and the trajectory of the PF-ODE solution consists of the speech states at different moments. Our approach is based on consistency model that ensure any speech states on the same PF-ODE trajectory, correspond to the same initial state. By integrating the Brownian Bridge process, the model is able to generate high-intelligibility speech samples without adversarial training. This is the first attempt that applies the consistency models to SE task, achieving state-of-the-art results in several metrics while saving 15 x the time required for sampling compared to the diffusion-based baseline. Our experiments on multiple datasets demonstrate the effectiveness of SE-Bridge in SE. Furthermore, we show through extensive experiments on downstream tasks, including Automatic Speech Recognition (ASR) and Speaker Verification (SV), that SE-Bridge can effectively support multiple downstream tasks. | ['Hao Huang', 'Gulila Altenbek', 'Fuchun Sun', 'Mengfan Fu', 'Zhibin Qiu'] | 2023-05-23 | null | null | null | null | ['speech-enhancement', 'speaker-verification'] | ['speech', 'speech'] | [ 1.90425143e-01 1.87973425e-01 2.62782246e-01 5.39106987e-02
-1.16053438e+00 -2.11112484e-01 6.45989358e-01 -4.02536839e-01
-2.87049741e-01 5.92953801e-01 5.67196667e-01 -6.18356228e-01
1.14158413e-03 -4.54515070e-01 -5.39795756e-01 -8.66340578e-01
-5.98943532e-02 -4.34098803e-02 1.46072984e-01 -4.42763269e-01
-5.71131222e-02 4.71021831e-01 -1.24141455e+00 1.41455874e-01
8.76911879e-01 8.57279837e-01 2.26189345e-01 1.07974112e+00
-4.05930579e-02 5.69965541e-01 -8.58495772e-01 -5.96461833e-01
1.44687563e-01 -4.87258703e-01 -5.59953392e-01 1.05008088e-01
-2.04599574e-02 -2.21132129e-01 -7.48716831e-01 1.29156196e+00
9.28736508e-01 4.38046813e-01 7.30860591e-01 -1.24567533e+00
-1.07400274e+00 5.56189716e-01 -3.82456392e-01 2.89788753e-01
2.46777177e-01 1.63618296e-01 6.40007973e-01 -1.07926571e+00
5.07268071e-01 1.46072316e+00 8.44515502e-01 1.18942857e+00
-1.15793216e+00 -4.04051244e-01 9.65192765e-02 -2.89853284e-04
-1.07828164e+00 -9.02385950e-01 5.77024460e-01 -9.99977216e-02
7.77553558e-01 3.45558882e-01 2.85792261e-01 1.44838715e+00
-1.26271080e-02 1.08706260e+00 1.13953805e+00 -3.23573917e-01
4.91168946e-01 1.17777966e-01 2.11302161e-01 4.22372550e-01
-4.36790526e-01 5.81654370e-01 -7.55473614e-01 -2.30785012e-01
3.11191201e-01 -5.29189706e-01 -4.64674324e-01 2.64256835e-01
-8.49833965e-01 6.13207281e-01 -5.32358289e-02 2.43653595e-01
-4.64386135e-01 3.19029391e-02 3.01305860e-01 4.57246572e-01
8.23772669e-01 -1.69458792e-01 -2.28431210e-01 -3.96370590e-01
-1.13720584e+00 8.99086148e-02 8.15664172e-01 8.19139659e-01
2.03024462e-01 3.32871497e-01 -7.75770426e-01 8.35994542e-01
5.53098857e-01 9.90862906e-01 3.99484187e-01 -1.14689386e+00
5.14340699e-01 -4.98699754e-01 6.84454739e-02 -4.30604398e-01
2.62672901e-01 -5.43537557e-01 -1.09457874e+00 4.33126092e-01
2.45803118e-01 -5.15753508e-01 -9.17560339e-01 1.96808970e+00
2.82078832e-01 5.32612026e-01 4.54208225e-01 6.11101210e-01
4.69060272e-01 1.17508090e+00 7.61483982e-02 -4.34401929e-01
1.16775548e+00 -9.01166558e-01 -1.17985070e+00 1.04188390e-01
1.23515286e-01 -8.77363741e-01 1.02641797e+00 4.15316761e-01
-1.38620782e+00 -3.82480949e-01 -9.22188997e-01 2.95872122e-01
1.83563270e-02 -1.00961067e-01 -9.25399885e-02 1.05175889e+00
-1.59771013e+00 7.51525700e-01 -8.00514579e-01 6.18488640e-02
3.36527765e-01 3.37205194e-02 -1.74226418e-01 2.17775106e-01
-1.36094260e+00 6.93994284e-01 -4.05931354e-01 6.36502504e-02
-1.15267003e+00 -9.09520745e-01 -8.30945969e-01 2.37256855e-01
-1.61963269e-01 -5.88311017e-01 1.44765043e+00 -6.03808343e-01
-2.10039830e+00 4.10914481e-01 -7.15419471e-01 -7.35231340e-01
7.81125546e-01 -1.02528848e-01 -6.49633765e-01 7.56355822e-02
-3.06782961e-01 5.37591100e-01 1.36893809e+00 -1.29445422e+00
-4.59328353e-01 -1.03022046e-01 -5.30732214e-01 4.12734784e-02
-4.35330898e-01 4.80518878e-01 -2.37275392e-01 -8.73213589e-01
-1.25714913e-01 -9.20264065e-01 -2.09218040e-01 7.52116321e-03
-4.64768261e-01 -1.84546515e-01 1.11995173e+00 -1.24308062e+00
1.35428929e+00 -2.51952410e+00 1.65502414e-01 1.18527394e-02
-2.72660404e-02 6.96024060e-01 -2.19182536e-01 2.65313447e-01
8.34457427e-02 3.05227935e-01 -5.99709690e-01 -1.12141883e+00
3.15819234e-01 -4.04582964e-03 -8.62264156e-01 4.61910725e-01
1.83929861e-01 7.13543296e-01 -6.97589338e-01 -2.44348720e-01
1.29829124e-01 1.02277589e+00 -5.20586371e-01 1.96527526e-01
3.80793326e-02 4.25479412e-01 4.42827977e-02 5.50209172e-02
1.04111135e+00 3.32814336e-01 -4.04305220e-01 2.77092546e-01
-1.90177858e-02 3.35779279e-01 -1.24226344e+00 1.43845165e+00
-6.88211262e-01 8.92116725e-01 4.59993422e-01 -5.76151669e-01
8.51934254e-01 7.33688712e-01 8.93767476e-02 -3.31539363e-01
-1.98286027e-01 2.15132385e-01 -3.08575332e-01 -1.59969732e-01
4.76987123e-01 -2.43105739e-01 3.02805305e-01 6.19529188e-01
3.55314836e-02 -1.58938095e-01 -2.00088874e-01 3.37050170e-01
1.04273295e+00 -2.42644653e-01 -2.23610893e-01 -2.08185494e-01
7.23675966e-01 -7.31319845e-01 3.73587728e-01 7.71015465e-01
-7.44967580e-01 5.10519207e-01 2.76600927e-01 3.70336473e-01
-1.10109031e+00 -1.46737850e+00 -1.31326839e-01 6.68276370e-01
-9.43104178e-02 -8.38327631e-02 -1.26055503e+00 -4.50846910e-01
-3.22052956e-01 9.25728202e-01 -2.65172035e-01 -3.36081654e-01
-3.61740649e-01 -4.90471929e-01 9.32524979e-01 2.34096050e-01
6.60987079e-01 -1.13850343e+00 1.63477957e-02 3.06670398e-01
-2.58749396e-01 -9.47966933e-01 -9.85733390e-01 -2.24024937e-01
-7.14063227e-01 -1.12294488e-01 -1.31619084e+00 -6.50704324e-01
3.66560966e-01 1.61581799e-01 5.53139806e-01 -3.24857980e-01
4.77935597e-02 4.41551298e-01 -1.32382542e-01 -1.49175987e-01
-1.21243680e+00 -2.22079635e-01 1.91166908e-01 3.85923177e-01
-1.91422477e-02 -4.85963613e-01 -5.81561685e-01 3.00074071e-01
-7.52650678e-01 -4.17214096e-01 8.67671892e-02 8.02767992e-01
3.98313463e-01 3.31377052e-02 8.83483469e-01 -1.53355986e-01
1.17355192e+00 -2.92884529e-01 -3.01667958e-01 2.30002716e-01
-4.96149302e-01 3.11207831e-01 2.34015331e-01 -6.25391006e-01
-1.59855878e+00 -3.18754584e-01 -8.49924028e-01 -5.08219719e-01
-1.75467491e-01 1.36474937e-01 -9.38585401e-02 2.26923943e-01
6.40492380e-01 5.16469240e-01 2.95619339e-01 -4.63804126e-01
6.46533489e-01 1.11659026e+00 6.27539337e-01 -4.45543766e-01
7.10744739e-01 5.47293842e-01 -2.32969970e-01 -1.14552414e+00
-6.85430884e-01 -3.61603618e-01 -6.88914657e-02 -2.79623866e-01
8.55254531e-01 -7.67956674e-01 -7.23541796e-01 9.06602442e-01
-1.43299210e+00 -4.91681248e-01 -6.37066126e-01 3.90308470e-01
-5.73189199e-01 7.25882709e-01 -1.02543879e+00 -1.61533701e+00
-6.29744887e-01 -1.15531683e+00 1.16817844e+00 1.88046750e-02
-3.80301662e-02 -1.09086800e+00 2.26260856e-01 1.11810125e-01
7.42773294e-01 -4.26220208e-01 5.21392703e-01 -2.97883838e-01
-2.79766381e-01 5.08624762e-02 1.58132702e-01 8.75634134e-01
3.72916572e-02 1.05222493e-01 -1.33593917e+00 -3.71870726e-01
4.66323555e-01 1.38561517e-01 9.03968036e-01 7.64586449e-01
8.40035737e-01 -3.22006732e-01 -8.64204690e-02 5.35431683e-01
8.81039381e-01 2.43920684e-01 9.23862875e-01 -2.00372115e-01
6.26889616e-02 5.70826590e-01 4.00354773e-01 3.95375818e-01
2.70068739e-02 6.84224069e-01 -7.77974203e-02 -1.75646707e-01
-6.82925761e-01 -3.90732616e-01 1.10760617e+00 1.11544383e+00
2.65122384e-01 -2.66233563e-01 -5.49613714e-01 6.53497458e-01
-1.50849962e+00 -1.43579948e+00 -1.54557154e-01 2.27853322e+00
8.85227263e-01 -9.98234078e-02 2.04187319e-01 4.19524938e-01
1.17128253e+00 1.40332580e-01 -4.42124814e-01 -5.01221836e-01
-3.07878703e-01 3.89344901e-01 1.00745194e-01 1.10893822e+00
-8.30891371e-01 8.19487512e-01 6.94846582e+00 1.06971955e+00
-9.55154359e-01 5.36440253e-01 7.85048008e-01 1.05173439e-01
-3.22272003e-01 -4.08631891e-01 -9.10986304e-01 4.52351630e-01
1.22649491e+00 -3.80637050e-01 6.53248131e-01 4.93246704e-01
5.82329750e-01 3.43088657e-01 -5.46406686e-01 8.09592962e-01
-1.44673795e-01 -1.09873843e+00 -2.79980510e-01 2.17506379e-01
8.68216038e-01 3.34100947e-02 6.82126343e-01 2.87137389e-01
5.66223443e-01 -7.35714734e-01 8.99943590e-01 4.17840034e-01
8.92712653e-01 -6.26934290e-01 3.43570113e-01 4.27391112e-01
-1.18074763e+00 -1.02885284e-01 -3.14078391e-01 4.71609652e-01
7.39700794e-01 7.84764230e-01 -5.37708998e-01 3.53605062e-01
4.73524928e-01 2.93180496e-01 2.08496287e-01 8.51164758e-01
-3.85680616e-01 9.03627098e-01 -1.75359517e-01 3.22905071e-02
4.64955829e-02 -3.17756683e-01 1.17222905e+00 1.38401246e+00
7.87967443e-01 -1.40709922e-01 -5.16485870e-01 1.04167306e+00
3.43675576e-02 -1.07099779e-01 -5.71715653e-01 1.86891168e-01
5.32174945e-01 8.67388785e-01 -1.28694743e-01 -3.22983027e-01
-4.49817367e-02 1.38628042e+00 7.65588731e-02 6.63430035e-01
-9.86535668e-01 -3.89999121e-01 1.10196793e+00 -1.83692351e-01
4.69095469e-01 -2.82527268e-01 -1.71204641e-01 -1.00924051e+00
-5.34311729e-03 -1.00049591e+00 5.57429530e-02 -7.03290343e-01
-1.29763424e+00 9.01201487e-01 -3.75116378e-01 -8.92134190e-01
-1.50790066e-01 -3.24641049e-01 -6.95794940e-01 1.35379457e+00
-1.49607265e+00 -6.49251342e-01 1.90131575e-01 7.98811197e-01
6.67209744e-01 -2.20855713e-01 6.39948547e-01 4.39614534e-01
-5.38019538e-01 8.01447272e-01 2.33627930e-01 -2.45547548e-01
5.06631017e-01 -1.23813963e+00 9.93591428e-01 1.25404537e+00
2.06199929e-01 4.29124683e-01 8.89697969e-01 -6.50894165e-01
-9.13385391e-01 -1.04807389e+00 1.09467363e+00 -2.73258299e-01
6.05964899e-01 -2.71254450e-01 -9.02807236e-01 2.70711809e-01
4.56099987e-01 -1.87201366e-01 4.04589474e-01 -2.34779075e-01
-1.77636161e-01 1.29618689e-01 -1.39895451e+00 7.59030640e-01
1.32109964e+00 -9.19825017e-01 -4.61800039e-01 1.08664729e-01
1.13871300e+00 -2.87684321e-01 -7.04907715e-01 2.79091056e-02
5.14397339e-04 -7.97002196e-01 9.78163540e-01 -5.54168284e-01
2.98134297e-01 -2.80108720e-01 -2.26264134e-01 -1.71951056e+00
5.15300594e-02 -1.51009476e+00 -5.77202797e-01 1.68053198e+00
6.01345241e-01 -7.23785102e-01 3.66233528e-01 5.21350324e-01
-3.20110738e-01 -4.15174931e-01 -1.30354738e+00 -1.13725746e+00
2.77061880e-01 -8.71279001e-01 7.80346930e-01 3.41211349e-01
-2.19595864e-01 -1.95177086e-02 -4.26953465e-01 5.30972898e-01
7.47488678e-01 -6.64342999e-01 3.87320876e-01 -7.11807549e-01
-6.19319141e-01 -5.56918025e-01 -3.20013836e-02 -1.42023373e+00
4.82364982e-01 -8.77468407e-01 4.48591381e-01 -1.45300674e+00
-1.36082634e-01 -3.77021313e-01 -9.64700282e-02 -7.98099488e-02
-3.72786760e-01 -3.11669171e-01 3.28045160e-01 -7.94134662e-02
-9.35455933e-02 9.80029821e-01 1.12141848e+00 -1.77428380e-01
-3.93845856e-01 4.99464303e-01 -5.01765728e-01 3.46544862e-01
6.72750473e-01 -4.99239683e-01 -4.33221966e-01 -1.87627062e-01
-2.99302071e-01 2.78042346e-01 3.12241018e-01 -9.62120175e-01
3.68673205e-01 3.24841410e-01 -3.76703501e-01 -3.55274379e-01
5.37569463e-01 -4.93893921e-01 1.83688570e-02 6.58188045e-01
-5.23128510e-01 -2.44499102e-01 2.52275050e-01 8.24607611e-01
-2.69544810e-01 -3.67942423e-01 9.96761858e-01 5.79662621e-01
-2.64643192e-01 5.25781155e-01 -7.17323661e-01 1.42180115e-01
8.18358421e-01 2.69617826e-01 -2.44019240e-01 -8.62232149e-01
-1.05289710e+00 -3.05295885e-01 -4.71336432e-02 2.70261794e-01
7.83873737e-01 -1.32767713e+00 -9.87752795e-01 2.26707593e-01
-5.07965624e-01 -4.00066257e-01 4.66074973e-01 6.82817936e-01
-3.08002550e-02 -3.38350516e-03 5.40150166e-01 -5.24090767e-01
-1.36058128e+00 5.97275734e-01 6.20458305e-01 -3.91614854e-01
-7.03994155e-01 1.00083172e+00 -2.56741839e-03 -4.79056090e-01
4.57188994e-01 -1.46964192e-01 1.72062442e-01 -2.49595225e-01
7.75502563e-01 6.51238084e-01 3.30011547e-02 -5.72463512e-01
-1.60360813e-01 2.29689911e-01 2.81706810e-01 -1.28608644e+00
1.02453423e+00 -3.77657801e-01 1.23811163e-01 9.97135416e-02
1.17320120e+00 2.96631306e-01 -1.60323763e+00 -2.14805096e-01
-1.50525197e-01 -2.84687310e-01 3.34220052e-01 -6.85005128e-01
-9.69099104e-01 1.27187848e+00 9.11576331e-01 3.24682832e-01
9.93921578e-01 -1.19668275e-01 1.07714546e+00 2.65158005e-02
2.05956414e-01 -9.47722375e-01 2.61675306e-02 6.76763117e-01
1.06719410e+00 -7.30452180e-01 -8.51689696e-01 -4.03035641e-01
-8.93244505e-01 6.78582549e-01 -1.19203769e-01 2.01044112e-01
1.00943494e+00 6.69772565e-01 9.03371572e-02 3.48823607e-01
-7.55852997e-01 -1.67147696e-01 2.27981672e-01 9.75691199e-01
1.72165826e-01 1.10148385e-01 -8.49603303e-03 4.82740670e-01
-8.52589831e-02 -2.24068046e-01 2.97531128e-01 5.30757129e-01
-3.52453828e-01 -1.22366154e+00 -4.95700151e-01 -1.76438242e-01
-4.52551216e-01 -4.38503325e-01 -1.61082938e-01 2.03699455e-01
-5.90892732e-01 1.38263345e+00 1.00220842e-02 -2.15224236e-01
2.85206884e-01 2.76513785e-01 2.63402313e-01 -1.64104745e-01
-5.18880606e-01 -4.17696405e-03 1.16721414e-01 -2.17551172e-01
8.38345196e-03 -9.73506272e-01 -1.14474189e+00 -6.49498880e-01
-3.53525668e-01 5.06587178e-02 1.01335657e+00 8.38705778e-01
6.66696727e-01 5.84242880e-01 1.07063770e+00 -4.87856716e-01
-1.33150077e+00 -1.01297522e+00 -6.37777507e-01 4.80955273e-01
6.26909614e-01 -1.68731317e-01 -9.20500875e-01 1.19302064e-01] | [15.047028541564941, 6.051145553588867] |
3b0b000c-9850-45e9-b8a8-5a644745dedf | sanom-results-for-oaei-2019 | 2006.05219 | null | https://arxiv.org/abs/2006.05219v1 | https://arxiv.org/pdf/2006.05219v1.pdf | SANOM Results for OAEI 2019 | Simulated annealing-based ontology matching (SANOM) participates for the second time at the ontology alignment evaluation initiative (OAEI) 2019. This paper contains the configuration of SANOM and its results on the anatomy and conference tracks. In comparison to the OAEI 2017, SANOM has improved significantly, and its results are competitive with the state-of-the-art systems. In particular, SANOM has the highest recall rate among the participated systems in the conference track, and is competitive with AML, the best performing system, in terms of F-measure. SANOM is also competitive with LogMap on the anatomy track, which is the best performing system in this track with no usage of particular biomedical background knowledge. SANOM has been adapted to the HOBBIT platfrom and is now available for the registered users. | ['Yao-Hua Tan', 'Majid Mohammadi', 'Wout Hofman', 'Amir Ahooye Atashin'] | 2020-06-09 | null | null | null | null | ['ontology-matching'] | ['knowledge-base'] | [ 1.85257792e-01 7.14016736e-01 -3.49095166e-01 -2.54710093e-02
-6.24395072e-01 9.86887217e-02 3.19515109e-01 7.77379990e-01
-6.83521748e-01 5.17141223e-01 4.03795004e-01 6.74757212e-02
-8.23973119e-01 -7.06872702e-01 -1.73352018e-01 -2.26536632e-01
-1.59317479e-01 1.28482366e+00 4.69874501e-01 -5.62721550e-01
2.00123191e-02 2.62223899e-01 -1.78874886e+00 3.62439454e-01
8.23289096e-01 5.20042539e-01 -2.54922695e-02 2.02583387e-01
2.60491855e-02 3.74025822e-01 -3.81115168e-01 -6.31359816e-01
9.00757834e-02 -3.84982944e-01 -1.15817618e+00 -7.78570354e-01
6.96696281e-01 4.42842603e-01 -1.88838705e-01 1.03685653e+00
9.54455853e-01 -4.86810058e-02 1.44134298e-01 -1.36243761e+00
1.65093452e-01 7.81490445e-01 5.60660996e-02 2.01512665e-01
7.72571802e-01 -6.23431265e-01 8.68088305e-01 -5.17652392e-01
1.34949017e+00 1.03528643e+00 9.46079373e-01 6.43595159e-01
-7.19075322e-01 -6.13694370e-01 -3.34899724e-01 6.32747054e-01
-1.37719321e+00 -2.69493759e-01 9.45142955e-02 -1.30309939e-01
1.28926277e+00 6.01645708e-01 1.04793978e+00 5.77118218e-01
4.12978768e-01 5.22569597e-01 8.80211353e-01 -8.46114099e-01
3.72164398e-01 -6.87050745e-02 2.02806190e-01 9.52972770e-01
4.09432799e-01 7.79685378e-02 -6.28171802e-01 -4.44248140e-01
-5.46964537e-03 -4.93949801e-01 -5.74055165e-02 -3.70843440e-01
-1.28299594e+00 2.73809880e-01 4.34145123e-01 7.25441456e-01
-5.33470333e-01 -1.99646771e-01 4.99299794e-01 1.27909005e-01
1.19820412e-03 7.42067158e-01 -2.77111501e-01 -4.06939685e-01
-8.05267036e-01 3.96654874e-01 1.01909828e+00 9.35164034e-01
-6.26970977e-02 -7.50293732e-01 2.70633101e-01 9.14339900e-01
1.67479679e-01 2.08747499e-02 5.15545785e-01 -9.76129711e-01
5.49846329e-02 8.96771193e-01 -3.14679950e-01 -5.01738787e-01
-9.98423815e-01 -4.62175399e-01 -5.47379017e-01 1.07074372e-01
2.93384820e-01 2.10197493e-01 -8.94690096e-01 1.46390688e+00
3.04299653e-01 1.05025806e-01 1.91898093e-01 5.20950258e-01
1.22806537e+00 -6.58473372e-02 3.53176415e-01 -2.71781832e-01
1.88739014e+00 -9.84800756e-01 -1.07516015e+00 -1.19544476e-01
8.26597750e-01 -8.02942276e-01 3.56463790e-01 6.78173363e-01
-1.24962711e+00 -2.69248098e-01 -1.28049767e+00 1.72849834e-01
-7.13932216e-01 -4.31784213e-01 7.12744653e-01 9.05000150e-01
-1.04663944e+00 4.85611230e-01 -9.32692349e-01 -1.35818136e+00
4.27499324e-01 5.96816897e-01 -7.76980221e-01 -3.07339281e-01
-1.48665202e+00 1.79024553e+00 5.67727506e-01 -3.09639066e-01
-2.56371111e-01 -8.94780338e-01 -6.70267582e-01 -2.37975493e-01
2.09425479e-01 -9.56280053e-01 8.94684553e-01 -2.19773099e-01
-1.12911773e+00 1.37083375e+00 -7.79933929e-02 -7.73858309e-01
5.84249139e-01 -2.77450681e-02 -1.00269651e+00 1.43821597e-01
6.14992678e-02 8.93355191e-01 -5.49431324e-01 -4.71886873e-01
-7.30281591e-01 -3.17700982e-01 -6.44350052e-02 2.97598004e-01
-3.27988118e-01 1.89482436e-01 -5.73005259e-01 -1.95257664e-01
4.57641542e-01 -1.12996852e+00 -5.07294536e-01 2.12814612e-03
8.79956111e-02 -4.36471075e-01 1.90942049e-01 -6.41316950e-01
1.68898726e+00 -1.85192764e+00 1.96061656e-01 3.89541030e-01
-1.17416345e-01 2.09365934e-01 1.97468087e-01 7.82740593e-01
-4.50947851e-01 -3.60013545e-02 -1.26265258e-01 -1.90425143e-01
5.36997356e-02 4.25915271e-01 5.59451878e-01 3.47336054e-01
-4.57783103e-01 6.43169820e-01 -1.08830845e+00 -8.52234423e-01
1.71245620e-01 5.23282997e-02 -5.78677654e-01 -2.97530681e-01
1.96660474e-01 2.06984043e-01 -1.82926327e-01 5.61707199e-01
5.71063876e-01 -2.95504835e-02 6.71938717e-01 -3.10723603e-01
-9.73418076e-03 2.94350863e-01 -1.12091970e+00 2.38686562e+00
-2.42474064e-01 2.66675085e-01 -3.52143615e-01 -9.72359061e-01
7.94740081e-01 8.47526729e-01 1.11769331e+00 -7.20994353e-01
-3.21456604e-02 7.08115995e-01 3.33931625e-01 -6.49089336e-01
2.78362334e-01 -2.80162636e-02 1.78384781e-03 9.46004689e-02
2.31186435e-01 4.24896926e-02 5.32658219e-01 2.18991384e-01
1.46376717e+00 2.22618327e-01 9.09851611e-01 -7.84365654e-01
6.98822737e-01 2.07671896e-01 5.88311136e-01 5.77411234e-01
-3.57029200e-01 3.82436782e-01 1.64475646e-02 -6.19842291e-01
-7.78349817e-01 -9.67481017e-01 -8.18896949e-01 5.34698665e-01
2.98790395e-01 -1.02523339e+00 -9.29925025e-01 -4.93112326e-01
-1.54387861e-01 5.47979057e-01 -6.38914704e-01 -1.77537054e-01
-4.61295903e-01 -6.39960527e-01 6.79075122e-01 7.86365569e-02
5.57614326e-01 -1.28704321e+00 -7.32077539e-01 4.23449039e-01
-4.47680831e-01 -1.23268569e+00 4.39060517e-02 1.52193755e-01
-8.21448267e-01 -1.15223384e+00 -4.11977738e-01 -8.08085322e-01
4.51594949e-01 -6.56755149e-01 1.17807591e+00 2.08682239e-01
-6.00783587e-01 -2.00625714e-02 -2.77997643e-01 -8.82618010e-01
-4.60043848e-01 6.45972669e-01 -7.70462770e-03 -6.20689034e-01
6.95364177e-01 -4.81280118e-01 -3.45887840e-01 3.20486248e-01
-6.15755498e-01 4.71832789e-02 3.03335577e-01 6.83956504e-01
5.42613387e-01 2.01576035e-02 5.07327616e-01 -9.96173441e-01
1.01587296e-01 -2.96969891e-01 -2.34322235e-01 4.01237577e-01
-1.22272635e+00 -2.04471380e-01 -1.84517130e-01 2.06476048e-01
-6.43460214e-01 1.01401918e-01 -4.43151623e-01 1.91986769e-01
-2.06583157e-01 6.11814618e-01 -2.11009890e-01 -4.78698537e-02
9.00774777e-01 -4.13523942e-01 -9.18828696e-03 -5.73975921e-01
-2.05155954e-01 5.87768674e-01 5.90508759e-01 -4.26245630e-01
1.80517167e-01 4.76305127e-01 2.24371433e-01 -4.93129402e-01
-6.10281765e-01 -7.39681363e-01 -5.55012226e-01 -1.73150733e-01
1.04524565e+00 -6.05697095e-01 -5.85314035e-01 4.32013273e-01
-7.29439080e-01 2.56135792e-01 -3.88445050e-01 7.79584765e-01
-4.42166328e-01 3.41446213e-02 -2.94813782e-01 -5.91710269e-01
-5.54973900e-01 -1.05682671e+00 7.96307504e-01 3.36515784e-01
-8.73150349e-01 -8.91769290e-01 1.90567151e-01 6.44587338e-01
3.92987549e-01 4.82026815e-01 9.58267331e-01 -9.31193233e-01
1.05422018e-02 -4.81457710e-01 4.06136662e-01 -3.19732696e-01
-2.36073658e-02 -2.96378583e-01 -6.40369296e-01 -1.39424995e-01
-4.32344884e-01 3.83226633e-01 3.43442082e-01 3.71445149e-01
5.99720478e-01 3.20325404e-01 -9.74254131e-01 3.02149177e-01
1.34809065e+00 4.72884387e-01 1.04182565e+00 1.22099400e+00
-1.04551099e-01 8.05427074e-01 8.65409553e-01 8.29213206e-03
5.90306044e-01 1.23435819e+00 5.13646126e-01 -9.41864625e-02
-2.70065248e-01 7.33807832e-02 -4.41227481e-02 9.30618286e-01
-2.78986424e-01 3.90994027e-02 -1.42808115e+00 5.97831428e-01
-2.10695744e+00 -9.14633334e-01 -4.03109670e-01 2.13980746e+00
5.14410853e-01 3.42655450e-01 1.25237152e-01 1.82449967e-01
5.90624213e-01 -2.29501098e-01 -1.62161049e-03 -5.68127871e-01
-2.63249904e-01 9.99063790e-01 3.03343624e-01 3.47036868e-01
-9.30334449e-01 6.31466866e-01 6.88572884e+00 6.83932364e-01
-3.39686483e-01 5.96995771e-01 -1.85071707e-01 -1.26053944e-01
5.95978312e-02 2.66204476e-01 -5.52402914e-01 3.97469908e-01
1.33874857e+00 -5.86882532e-01 2.22484499e-01 5.28340638e-01
-3.58639210e-01 -3.10952902e-01 -1.09818244e+00 7.26276338e-01
1.58880576e-01 -1.53210306e+00 -3.57388586e-01 3.15450281e-01
4.20012593e-01 5.12288332e-01 -7.48924851e-01 2.22917259e-01
1.10473566e-01 -1.14324927e+00 4.22915906e-01 7.33155072e-01
3.61945152e-01 -6.88647807e-01 1.38729489e+00 6.67596459e-02
-1.09519207e+00 4.94908951e-02 7.20614791e-02 2.80526638e-01
2.08739087e-01 3.14503461e-01 -4.73638684e-01 1.54009390e+00
1.00798810e+00 6.06992960e-01 -4.40351635e-01 1.73461330e+00
1.11344926e-01 2.39875153e-01 -5.46554923e-01 1.11813121e-01
-2.35264972e-01 1.03413528e-02 7.93167651e-01 1.10367846e+00
1.69459969e-01 7.76468404e-03 2.49547154e-01 -5.03227040e-02
1.20621972e-01 3.81854773e-01 -1.15049526e-01 3.87744635e-01
6.39354587e-01 1.05662119e+00 -6.12995744e-01 -3.44700813e-01
-9.27564204e-02 6.60220325e-01 -1.54965281e-01 -4.07162964e-01
-6.83517933e-01 -5.77430606e-01 4.06261116e-01 1.54366508e-01
-2.24005222e-01 7.11265743e-01 -2.28149787e-01 -5.03045499e-01
8.56019482e-02 -1.05760849e+00 1.21060562e+00 -8.82755160e-01
-9.79116619e-01 1.00990355e+00 2.70537287e-01 -1.38153553e+00
1.81597117e-02 -3.80068868e-01 8.78496189e-03 7.33278215e-01
-1.05483866e+00 -1.09803748e+00 -4.28453654e-01 2.29583904e-01
-5.03641367e-02 -3.29128414e-01 1.72062111e+00 9.34687972e-01
-3.78500998e-01 6.99941814e-01 -2.26203039e-01 -1.23351984e-01
8.62526417e-01 -1.10612822e+00 3.55213225e-01 3.99107933e-01
6.49816543e-02 5.58399498e-01 7.72759497e-01 -5.09886444e-01
-9.99980032e-01 -4.93087918e-01 1.46025133e+00 -4.82367128e-01
4.72637028e-01 1.14486076e-01 -4.93841976e-01 4.05498415e-01
4.00139868e-01 -1.03191033e-01 8.45433056e-01 1.68819800e-01
4.25320491e-03 -2.55992979e-01 -1.39620662e+00 3.47690970e-01
1.44266117e+00 -2.58801550e-01 -8.56063843e-01 6.78176880e-01
2.43179038e-01 -9.45810914e-01 -1.66911447e+00 8.84981096e-01
9.72298503e-01 -7.39418089e-01 9.09050107e-01 -6.14315987e-01
6.34672120e-02 -3.53792995e-01 -9.36983079e-02 -9.27943468e-01
-2.48175994e-01 -7.06668854e-01 1.96933374e-01 9.90656674e-01
5.50569713e-01 -9.27297950e-01 9.17038977e-01 3.31458211e-01
-4.57835466e-01 -8.76640320e-01 -1.43614149e+00 -8.04775536e-01
-2.05834303e-02 -1.61335930e-01 8.84292066e-01 1.24490321e+00
6.54512882e-01 -4.96124066e-02 5.52237742e-02 2.42473707e-02
5.64104736e-01 -2.09316581e-01 5.66410780e-01 -1.69940186e+00
-5.54387011e-02 -5.57516634e-01 -1.16189241e+00 1.50815248e-01
-2.58603334e-01 -1.42810500e+00 -2.96498090e-01 -2.06635332e+00
2.43875578e-01 -3.80029619e-01 -7.00440943e-01 7.63316751e-01
-4.84947599e-02 2.99530804e-01 1.06511243e-01 1.46306604e-01
-4.21745569e-01 -1.51239902e-01 8.40375245e-01 7.58210272e-02
1.36373445e-01 -1.95027679e-01 -6.26258492e-01 5.25233924e-01
5.18078864e-01 -9.23051298e-01 1.64555628e-02 8.21751058e-02
2.15730324e-01 -5.82305305e-02 -1.71214744e-01 -1.54364491e+00
5.27994812e-01 3.41404647e-01 -1.12137504e-01 -5.89126229e-01
1.66077673e-01 -1.00957644e+00 1.10665560e+00 1.09129238e+00
-1.94086462e-01 2.28478611e-01 4.59817231e-01 -1.21462099e-01
-2.94753313e-01 -5.10280192e-01 6.14478350e-01 7.99388066e-02
-7.09385455e-01 1.18485764e-01 -3.48598957e-01 -9.10104811e-02
1.09214520e+00 -5.06329060e-01 -3.84432495e-01 1.50102153e-01
-1.18760645e+00 5.06184161e-01 3.88266951e-01 5.41494727e-01
-2.34489068e-02 -1.31216288e+00 -7.83648014e-01 -1.89718574e-01
6.21206880e-01 -4.86690700e-01 2.80756027e-01 1.36932135e+00
-7.10549891e-01 3.92668992e-01 -4.94354993e-01 -5.99534690e-01
-1.55740619e+00 2.01150894e-01 6.62272573e-01 -8.90884936e-01
-7.47166991e-01 5.14922559e-01 -6.35746479e-01 -7.99481690e-01
2.94746816e-01 2.10005388e-01 -3.47333342e-01 8.21867362e-02
3.80652964e-01 4.97204483e-01 8.65001798e-01 -3.51574928e-01
-9.16238248e-01 4.39916164e-01 6.90106442e-03 -1.38755232e-01
1.44921780e+00 3.12464148e-01 -5.67230105e-01 1.45581976e-01
7.83894539e-01 -1.49252504e-01 2.18332261e-01 -7.61787146e-02
6.72696710e-01 -1.05059877e-01 -4.57772650e-02 -1.34859514e+00
-9.84348774e-01 3.27590257e-01 1.15292382e+00 -2.67432809e-01
1.12352920e+00 -1.73133180e-01 7.23543644e-01 2.06043705e-01
7.82250643e-01 -1.05862164e+00 -6.55890942e-01 1.99710801e-01
6.10367119e-01 -6.92482412e-01 4.44011092e-01 -6.17931008e-01
-2.59074599e-01 1.07929027e+00 4.50287879e-01 1.73738584e-01
8.88165951e-01 4.24956620e-01 1.66200355e-01 -6.78232789e-01
-5.87666690e-01 -1.46192551e-01 4.22370255e-01 5.82528532e-01
6.20315611e-01 2.45370641e-02 -1.09806800e+00 2.81302780e-01
-5.25639057e-01 4.56599087e-01 -3.73677909e-02 1.03250754e+00
-9.75698307e-02 -1.74186981e+00 -1.09061599e-01 4.14609432e-01
-7.70975053e-01 -1.28860414e-01 -4.28564787e-01 1.11769342e+00
4.65042800e-01 5.80609262e-01 -7.84645975e-02 -2.73489207e-01
9.32128012e-01 2.83638775e-01 8.01060855e-01 -5.35296082e-01
-1.32583952e+00 -3.19851846e-01 8.74797404e-01 -8.45405698e-01
-6.32376790e-01 -1.01923001e+00 -1.51007819e+00 -2.11663350e-01
-4.43495154e-01 6.76904261e-01 8.14664662e-01 8.20016682e-01
3.04172963e-01 7.94825196e-01 -3.26008707e-01 -2.31684759e-01
2.42485672e-01 -9.45254803e-01 -1.49036273e-01 3.13054234e-01
-6.01778626e-01 -7.32354760e-01 2.35482246e-01 -3.05737495e-01] | [9.185012817382812, 8.065192222595215] |
47279e00-f7a7-4960-986c-87d374e53553 | coordvit-a-novel-method-of-improve-vision | null | null | https://ieeexplore.ieee.org/document/10049941 | https://ieeexplore.ieee.org/document/10049941 | CoordViT: A Novel Method of Improve Vision Transformer-Based Speech Emotion Recognition using Coordinate Information Concatenate | Recently, in speech emotion recognition, a Transformer-based method using spectrogram images instead of sound data showed improved accuracy than Convolutional Neural Networks (CNNs). Vision Transformer (ViT), a Transformer-based method, achieves high classification accuracy by using divided patches from the input image, but has a problem in that pixel position information is not retained due to embedding layers such as linear projection. Therefore, in this paper, we propose a novel method of improve ViT-based speech emotion recognition using coordinate information concatenate. Since the proposed method retains pixel position information by concatenating coordinate information to the input image, the accuracy of CREMA-D is greatly improved by 82.96% compared to the state-of-art about CREMA-D. As a result, it proved that the coordinate information concatenate proposed in this paper is effective not only for CNNs but also for Transformers. | ['Seung-Ho Lee', 'Jeongyoon Kim'] | 2023-03-10 | null | null | null | international-conference-on-electronics | ['speech-emotion-recognition'] | ['speech'] | [ 7.12970197e-02 -1.45468414e-01 4.35022384e-01 -1.13903679e-01
-4.04970884e-01 -5.99635132e-02 2.96091110e-01 -3.98748904e-01
-3.79399031e-01 4.48886961e-01 1.45833284e-01 -1.13020763e-01
2.80739427e-01 -8.62574995e-01 -5.30606508e-01 -9.21446383e-01
6.03396773e-01 -4.23828721e-01 1.04230352e-01 -1.46418408e-01
2.31390804e-01 4.41225767e-01 -1.74502850e+00 6.64994538e-01
6.64377451e-01 1.61396456e+00 3.39968562e-01 2.97706515e-01
-5.26269138e-01 1.02209401e+00 -8.29099119e-01 -3.48396182e-01
-1.60711989e-01 -4.57619399e-01 -4.30109560e-01 1.03089064e-02
1.76765099e-01 -1.53836906e-01 -5.56500912e-01 1.28012466e+00
6.74529970e-01 1.30463969e-02 5.60045958e-01 -1.27395868e+00
-8.77950728e-01 3.97468239e-01 -3.36585164e-01 -7.65307248e-02
7.49942437e-02 -3.87302488e-01 6.00118697e-01 -1.17651010e+00
1.68785796e-01 1.15997624e+00 9.97497797e-01 4.88467902e-01
-6.13469720e-01 -7.91736245e-01 -6.56510368e-02 7.76937962e-01
-1.34760880e+00 -3.07282269e-01 1.33661377e+00 -5.55809177e-02
1.45294464e+00 3.68710577e-01 1.12499166e+00 1.01749194e+00
4.06969696e-01 8.92130375e-01 1.17773557e+00 -5.17729342e-01
1.33345187e-01 3.14504385e-01 -1.75329611e-01 7.79427946e-01
-4.23412323e-01 3.32047082e-02 -6.14173710e-01 2.29245991e-01
7.41109371e-01 -2.59974636e-02 -4.55807358e-01 1.35410011e-01
-1.05103016e+00 6.02766275e-01 7.28085518e-01 6.53632343e-01
-4.37333345e-01 -3.43124196e-02 5.60404778e-01 3.68322998e-01
6.08994961e-01 1.08317249e-01 -3.58240604e-01 -4.57430273e-01
-7.85757601e-01 -3.59925151e-01 4.76517498e-01 5.62060535e-01
5.04308760e-01 6.83215439e-01 1.47305548e-01 1.18220961e+00
2.03723490e-01 5.21350205e-01 9.38485384e-01 -6.58059597e-01
3.34149778e-01 6.08807862e-01 -5.87351382e-01 -1.27597451e+00
-3.10455441e-01 -4.59998161e-01 -1.29913151e+00 3.10873955e-01
-6.30093440e-02 -4.15995270e-02 -1.01198065e+00 1.42478931e+00
-6.99721575e-02 1.38427258e-01 5.55397272e-01 9.30686593e-01
1.26830947e+00 1.11707139e+00 -3.20281088e-01 -3.13102990e-01
1.38657725e+00 -1.12007105e+00 -1.22327936e+00 1.23295635e-01
2.41234571e-01 -9.53146458e-01 1.12057483e+00 8.16469669e-01
-7.61622965e-01 -7.13156402e-01 -1.12873316e+00 -9.74424463e-03
-6.56014323e-01 6.57380104e-01 4.00757223e-01 8.99848282e-01
-1.31481934e+00 2.69742042e-01 -5.50311148e-01 -2.72743493e-01
3.63224834e-01 2.68436801e-02 -5.12577236e-01 2.99065709e-01
-1.17110741e+00 8.48203599e-01 -4.01235409e-02 3.08538377e-01
-5.17871022e-01 -3.37876439e-01 -8.64287078e-01 5.04450262e-01
-1.26355916e-01 -1.09389551e-01 1.23271656e+00 -1.08529413e+00
-2.09215403e+00 4.09766048e-01 -2.66738087e-01 -3.96608651e-01
7.98169374e-02 6.68646097e-02 -5.76200187e-01 4.80167568e-01
-3.88999522e-01 5.36661029e-01 1.02901471e+00 -8.85957360e-01
-4.76279795e-01 -3.18529904e-01 -1.97060138e-01 1.92330386e-02
-9.07216132e-01 -8.11531842e-02 -2.39095002e-01 -6.70821071e-01
5.97436666e-01 -5.06278634e-01 2.58598924e-01 1.74493894e-01
-7.84000680e-02 -3.24534863e-01 1.45315921e+00 -7.92007267e-01
9.76530135e-01 -2.24952483e+00 -3.02360982e-01 -1.18215606e-01
1.08409733e-01 6.19359314e-01 -1.63344406e-02 3.11139077e-01
-2.52368808e-01 2.66676117e-02 -1.12091891e-01 -3.57835591e-01
-1.14873305e-01 5.45826182e-02 -2.57082880e-01 2.00096294e-01
2.70311773e-01 7.32728124e-01 -4.29629594e-01 -4.48732853e-01
8.03317785e-01 9.33718622e-01 -3.18506360e-01 1.04901139e-02
3.18285078e-01 5.47812730e-02 -2.28264213e-01 5.97517848e-01
9.27484870e-01 2.75165588e-01 -2.17797846e-01 -7.37576127e-01
-4.09514278e-01 1.17374212e-01 -7.33863294e-01 1.51104808e+00
-6.90437675e-01 1.18043077e+00 -2.73688827e-02 -1.37447119e+00
1.60974598e+00 8.02996993e-01 4.65148658e-01 -1.17929828e+00
3.58081818e-01 6.30994365e-02 -4.43788499e-01 -8.52260172e-01
4.58407581e-01 -2.78234750e-01 2.68244594e-01 9.04666036e-02
1.46307617e-01 -3.24940413e-01 -6.58033252e-01 -2.90960610e-01
8.41713130e-01 -1.43312395e-01 -1.16162479e-01 1.05313092e-01
7.43314981e-01 -1.32869214e-01 6.65724576e-01 1.96034700e-01
-2.26340309e-01 8.03085327e-01 3.56242806e-01 -4.66711074e-01
-1.11368775e+00 -7.71901131e-01 -1.91764280e-01 1.37555480e-01
1.31234616e-01 -4.05957580e-01 -9.35524762e-01 -4.30440605e-01
-5.21781504e-01 4.21720386e-01 -5.45443714e-01 -3.51001322e-01
-2.58053631e-01 -4.36475366e-01 8.81238699e-01 5.67573607e-01
1.41482437e+00 -1.15861201e+00 -5.78677654e-01 3.03075075e-01
-5.92603743e-01 -1.17019057e+00 -2.33688220e-01 5.61961979e-02
-7.25620329e-01 -7.53801048e-01 -1.09319091e+00 -1.19656551e+00
5.06011367e-01 3.08413029e-01 5.46449125e-01 -1.93167746e-01
-1.09294645e-01 2.31115341e-01 -6.82703674e-01 -3.08646858e-01
3.46812569e-02 -3.12352002e-01 -1.67346269e-01 2.32599497e-01
3.00561011e-01 -6.72271311e-01 -4.60278988e-01 1.00784272e-01
-6.07336283e-01 2.83971816e-01 6.02918506e-01 1.00196290e+00
3.96364242e-01 2.98544139e-01 5.00495195e-01 -1.77424625e-01
6.90458536e-01 3.12775880e-01 -1.53455898e-01 2.91223228e-01
-4.33834702e-01 -3.16942334e-01 6.86332703e-01 -5.06859839e-01
-1.27929807e+00 -7.76197910e-02 -5.45588255e-01 -9.05880034e-01
-6.27451539e-02 5.25998414e-01 -2.25722760e-01 -2.27687627e-01
3.30133021e-01 6.46218956e-01 2.55844146e-01 -5.95191300e-01
-1.50342137e-01 1.20830750e+00 5.70692182e-01 -1.49085253e-01
5.51760048e-02 1.59987554e-01 -2.44412079e-01 -1.17826056e+00
-1.42706007e-01 -1.17769264e-01 -1.05961800e-01 -5.03712773e-01
9.25609171e-01 -9.58171844e-01 -1.08907962e+00 1.10807252e+00
-1.45472002e+00 1.27095878e-01 -5.13181137e-03 8.32953572e-01
-3.07655632e-01 4.24334407e-01 -7.88255811e-01 -1.09230065e+00
-5.22039056e-01 -1.17806125e+00 7.87748396e-01 3.56293797e-01
3.58709008e-01 -6.30631924e-01 -2.71981895e-01 2.26636127e-01
6.20509326e-01 6.53196946e-02 6.97096527e-01 -7.28895068e-02
-1.92700744e-01 -1.27251431e-01 -3.32607478e-01 8.52246523e-01
2.99076915e-01 3.07481587e-02 -1.40436769e+00 1.42927468e-01
4.75583762e-01 -2.29947880e-01 9.89332914e-01 3.80274028e-01
1.36850011e+00 -2.75383830e-01 -1.06414156e-02 5.26473999e-01
1.12347579e+00 7.26177871e-01 1.18201947e+00 3.64501119e-01
4.30470496e-01 3.80161971e-01 5.72698951e-01 3.54122430e-01
2.38883972e-01 6.10820532e-01 4.53668714e-01 -4.78061259e-01
-2.40302622e-01 -1.75708562e-01 5.33170938e-01 1.32312012e+00
4.06871475e-02 -2.14802787e-01 -4.64563668e-01 4.50108409e-01
-1.62986481e+00 -1.07138479e+00 -1.97502393e-02 1.73065877e+00
5.76277018e-01 -9.63005573e-02 -3.76497388e-01 7.42640495e-01
8.36624146e-01 9.27234888e-02 -2.32363194e-01 -7.64964104e-01
-4.37375993e-01 4.01152670e-01 -1.01784527e-01 2.32760921e-01
-8.62040222e-01 8.97749841e-01 5.90307903e+00 1.12708533e+00
-1.86206174e+00 1.43127546e-01 5.07810354e-01 2.57984936e-01
-8.13409537e-02 -5.79125524e-01 -5.74920028e-02 3.91365856e-01
5.85413873e-01 1.40305152e-02 3.22982311e-01 9.06540930e-01
2.44434312e-01 9.65672582e-02 -3.55864495e-01 1.53490663e+00
3.29981148e-01 -1.21369922e+00 8.33432972e-02 -1.40244797e-01
4.27336335e-01 -3.15102249e-01 3.27010453e-01 2.90460765e-01
-4.66558635e-01 -8.96183968e-01 7.59616494e-01 4.64801788e-01
9.13753688e-01 -9.09176469e-01 1.03454125e+00 1.15718357e-01
-1.41204774e+00 3.89311835e-02 -5.58062553e-01 -2.97420949e-01
-1.71349257e-01 6.79224372e-01 -7.89944112e-01 5.48675478e-01
1.07048225e+00 9.09801900e-01 -2.14749649e-01 9.44917381e-01
-1.54710919e-01 7.30690956e-01 -3.28562587e-01 -2.47901529e-01
3.54710937e-01 -1.36010811e-01 2.26634338e-01 1.17086303e+00
8.35759282e-01 1.81312963e-01 -5.53216875e-01 7.12393343e-01
9.83723029e-02 1.72251061e-01 -6.71112061e-01 -1.24617107e-01
3.60320956e-01 1.40234578e+00 -4.82044995e-01 -4.37126845e-01
-3.98903817e-01 1.01235318e+00 -2.83253398e-02 3.21172953e-01
-8.97775769e-01 -7.20571637e-01 4.60114658e-01 -4.95250612e-01
6.10164404e-01 -2.20598757e-01 -2.78110683e-01 -9.66144502e-01
2.04619676e-01 -6.94851875e-01 -1.27447471e-01 -1.48863471e+00
-7.36473441e-01 1.10695016e+00 -6.71288669e-01 -1.62386274e+00
2.13933960e-01 -8.77233267e-01 -8.15601110e-01 7.36737370e-01
-1.49503875e+00 -1.25633156e+00 -5.64839482e-01 8.48449767e-01
5.35005033e-01 -3.62953186e-01 9.94379342e-01 3.95913005e-01
-6.28712237e-01 7.43665934e-01 1.99557349e-01 2.00642928e-01
3.28738749e-01 -9.48735654e-01 6.51215436e-03 6.09702408e-01
3.00775111e-01 1.10754944e-01 3.93681884e-01 -1.35192260e-01
-1.15429616e+00 -8.02292645e-01 9.80544686e-01 3.88297528e-01
1.36320725e-01 -1.68190703e-01 -8.77623737e-01 2.43817598e-01
4.91628557e-01 -1.74097478e-01 3.66435379e-01 -2.31377959e-01
-3.35269064e-01 -4.70008820e-01 -1.39780879e+00 2.30868876e-01
4.84208167e-01 -9.14966643e-01 -6.79682434e-01 -2.37162054e-01
8.03067327e-01 -3.61018211e-01 -7.48055160e-01 4.57081258e-01
6.85879350e-01 -1.14593697e+00 8.03554952e-01 1.53870091e-01
4.68338013e-01 -3.38063926e-01 -2.79890597e-01 -1.47505176e+00
-6.64497092e-02 -1.62118614e-01 3.24104220e-01 1.39469016e+00
2.74463654e-01 -9.31913853e-01 6.75623417e-01 -1.04200896e-02
-5.23100734e-01 -7.15058029e-01 -1.31474853e+00 -5.38061082e-01
-1.34741649e-01 -6.03087246e-01 7.16680408e-01 9.74276304e-01
2.51431108e-01 1.77574426e-01 -4.12401706e-01 1.32721052e-01
2.28624910e-01 -8.49694461e-02 3.21204722e-01 -9.30296659e-01
2.42789209e-01 -3.86147082e-01 -6.84242308e-01 -1.01903367e+00
8.17276239e-02 -5.64923882e-01 -3.14836539e-02 -1.56285226e+00
-1.19095959e-01 -1.50314733e-01 -4.99950320e-01 6.74540401e-01
3.43805820e-01 5.44385791e-01 2.35636771e-01 -9.59732234e-02
-4.54447865e-02 1.21677411e+00 1.62104356e+00 -4.67179626e-01
1.25166550e-01 -2.91221440e-01 -2.80936807e-01 5.52162826e-01
9.66150403e-01 -1.22138873e-01 -3.39906484e-01 -3.65871012e-01
-1.92796364e-01 2.58154392e-01 5.80569446e-01 -1.37999284e+00
5.70756435e-01 3.08142513e-01 5.96610844e-01 -8.94531488e-01
9.44573283e-01 -9.56501245e-01 1.20048791e-01 4.21510220e-01
-9.33949649e-03 7.01001137e-02 4.88456994e-01 8.25597346e-02
-8.38138163e-01 -6.46795556e-02 6.05486870e-01 2.40225587e-02
-8.67663562e-01 -3.17642814e-03 -7.73057044e-01 -6.68333471e-01
6.19518399e-01 -4.87682670e-01 -4.90104347e-01 -5.21228015e-01
-6.66753232e-01 -4.29321498e-01 -2.30140947e-02 3.51380467e-01
1.13076675e+00 -1.58574665e+00 -2.53860891e-01 4.54030067e-01
-1.18999228e-01 -3.43281418e-01 6.09426677e-01 8.90808940e-01
-5.21693289e-01 4.68617082e-01 -4.85791028e-01 -6.17286682e-01
-1.61615074e+00 2.81018972e-01 6.07893944e-01 2.36828417e-01
-8.00511241e-01 8.04643393e-01 1.44026354e-01 -5.05494475e-01
3.56187522e-01 -5.36593974e-01 -4.35151249e-01 -6.40921071e-02
4.49825704e-01 1.02006644e-01 3.93389642e-01 -5.95025659e-01
-4.15895671e-01 8.90230119e-01 1.72132701e-01 -3.70522976e-01
1.35386372e+00 -4.82141450e-02 -1.40481904e-01 3.43009830e-01
1.46845567e+00 -2.42576942e-01 -9.58439827e-01 -2.21857265e-01
-7.39208758e-01 -3.84654224e-01 5.46088576e-01 -9.66001034e-01
-1.61697400e+00 1.49302745e+00 9.06527638e-01 2.07729533e-01
1.51135933e+00 -3.57541859e-01 9.67837334e-01 5.02220273e-01
2.09559590e-01 -1.26852989e+00 3.67409438e-01 5.70149064e-01
1.09268343e+00 -8.38746905e-01 -5.23620605e-01 -4.58435446e-01
-6.57625854e-01 1.44432068e+00 7.31577158e-01 1.74489543e-01
7.70712972e-01 3.44796926e-01 5.56748688e-01 -1.24785095e-01
-7.21765757e-01 -1.29093960e-01 -9.68084559e-02 7.64227867e-01
2.82954037e-01 5.05668111e-03 -5.35447448e-02 6.83850408e-01
-5.59635639e-01 -3.31490696e-03 2.25455984e-01 7.09785044e-01
-5.47833741e-01 -7.07930446e-01 -5.68791866e-01 9.56884250e-02
-1.95013732e-01 -1.43975034e-01 -4.64899957e-01 4.34738994e-01
2.03723595e-01 1.26454890e+00 3.23319316e-01 -1.03013098e+00
3.70970905e-01 3.69384065e-02 4.79210168e-01 1.07497513e-01
-5.18622100e-01 2.11971924e-01 -4.78413999e-02 -3.14078450e-01
-4.79875386e-01 -3.40183407e-01 -1.24463809e+00 -3.72172117e-01
-4.31168407e-01 3.38335484e-01 1.03277516e+00 8.63744259e-01
2.63884604e-01 8.96275520e-01 9.27187920e-01 -7.21963644e-01
-3.69154401e-02 -1.16958904e+00 -6.51152372e-01 -4.86966409e-02
3.51575047e-01 -6.50247037e-01 -5.03334105e-01 -6.64801970e-02] | [14.47581958770752, 5.763065338134766] |
1b292449-62ad-4f14-8588-2cb8ddbd429f | wav2shape-hearing-the-shape-of-a-drum-machine | 2007.10299 | null | https://arxiv.org/abs/2007.10299v1 | https://arxiv.org/pdf/2007.10299v1.pdf | wav2shape: Hearing the Shape of a Drum Machine | Disentangling and recovering physical attributes, such as shape and material, from a few waveform examples is a challenging inverse problem in audio signal processing, with numerous applications in musical acoustics as well as structural engineering. We propose to address this problem via a combination of time--frequency analysis and supervised machine learning. We start by synthesizing a dataset of sounds using the functional transformation method. Then, we represent each percussive sound in terms of its time-invariant scattering transform coefficients and formulate the parametric estimation of the resonator as multidimensional regression with a deep convolutional neural network. We interpolate scattering coefficients over the surface of the drum as a surrogate for potentially missing data, and study the response of the neural network to interpolated samples. Lastly, we resynthesize drum sounds from scattering coefficients, therefore paving the way towards a deep generative model of drum sounds whose latent variables are physically interpretable. | ['Vincent Lostanlen', 'Han Han'] | 2020-07-20 | null | null | null | null | ['audio-signal-processing'] | ['audio'] | [ 3.76436114e-01 -5.84703274e-02 7.31005728e-01 1.64178479e-02
-1.05821812e+00 -6.90150023e-01 3.05774868e-01 -2.32252672e-01
-4.21131365e-02 5.91682613e-01 2.64051199e-01 6.11206517e-02
-4.83840764e-01 -8.40803683e-01 -8.66781533e-01 -9.85043347e-01
-1.06825441e-01 5.28375626e-01 -3.49974066e-01 -2.00988248e-01
-2.23797768e-01 4.28081542e-01 -1.47707438e+00 2.01828614e-01
5.18487394e-01 9.11962152e-01 -2.32547775e-01 9.35822546e-01
2.78026700e-01 3.47450078e-01 -6.95591331e-01 -2.52439886e-01
9.08611938e-02 -4.72618103e-01 -3.59414309e-01 -5.39425611e-02
9.79965776e-02 -2.26982281e-01 -1.61060333e-01 6.36122286e-01
5.24502218e-01 3.83917034e-01 1.02725744e+00 -6.50880098e-01
-6.56579077e-01 7.42883563e-01 -2.18687013e-01 -2.91453600e-01
3.05985391e-01 -2.01195285e-01 1.19392812e+00 -8.65788877e-01
1.73733711e-01 1.06609845e+00 9.54857290e-01 1.11601271e-01
-1.81728721e+00 -5.05021811e-01 -6.37710929e-01 -2.45446071e-01
-1.27825248e+00 -4.21688080e-01 1.39641726e+00 -7.58673429e-01
2.29869887e-01 3.40714186e-01 6.36715829e-01 1.07754493e+00
-4.98444587e-02 4.84882981e-01 6.76066220e-01 -7.40069628e-01
9.83054489e-02 -2.41020709e-01 -1.71771884e-01 4.43171620e-01
-2.87840754e-01 3.73770028e-01 -6.98410094e-01 -3.62241268e-01
9.69444036e-01 -3.36449891e-01 -4.30480778e-01 -3.27297032e-01
-1.26477802e+00 7.77494729e-01 1.47375420e-01 7.64144138e-02
-5.08166313e-01 3.17722440e-01 1.61998674e-01 3.68681967e-01
6.55004382e-01 8.67832243e-01 -2.37781733e-01 7.83246830e-02
-9.60532546e-01 5.06783307e-01 7.61864305e-01 1.27308533e-01
5.18816829e-01 6.66947484e-01 1.50658444e-01 9.15298402e-01
4.39542085e-01 6.06025875e-01 1.33801267e-01 -1.14846849e+00
-6.02492020e-02 -3.23039621e-01 2.19414830e-01 -8.61280560e-01
-4.13541019e-01 -6.73373401e-01 -8.77469242e-01 2.44497180e-01
6.44185603e-01 -3.25234413e-01 -3.85018021e-01 1.79284751e+00
1.15729868e-01 5.89933693e-01 -3.92459631e-02 1.10776341e+00
6.95051730e-01 6.73513234e-01 -2.29008898e-01 -5.03674187e-02
1.12573314e+00 -3.55962217e-01 -4.97460425e-01 2.87417948e-01
1.77477747e-01 -9.75707829e-01 9.68697667e-01 8.69663596e-01
-1.31270671e+00 -9.68567729e-01 -9.57690835e-01 -5.27550764e-02
1.26871467e-01 3.40270787e-01 2.92702764e-01 4.12844896e-01
-6.61888361e-01 9.38785076e-01 -7.31907248e-01 7.10361600e-01
-1.27423778e-01 2.36778781e-01 -2.18720630e-01 4.91352916e-01
-1.30072963e+00 4.39804524e-01 -1.48780271e-01 3.91119242e-01
-7.41453707e-01 -1.25207794e+00 -7.40666389e-01 8.48979130e-02
1.55922696e-02 -8.41544449e-01 1.15423882e+00 -6.66404545e-01
-1.71570289e+00 4.22885358e-01 2.17498109e-01 -1.71023667e-01
2.94143885e-01 -3.36632937e-01 -6.52428925e-01 5.04486449e-03
-7.69997984e-02 1.10027753e-01 1.48375690e+00 -1.28170586e+00
-6.24504080e-03 -8.52554291e-02 -1.97920069e-01 -2.22803220e-01
-8.38224217e-02 -5.13990819e-01 3.27418178e-01 -1.02760148e+00
3.97407115e-01 -8.43330443e-01 -8.01064074e-02 -9.78840962e-02
-6.60515428e-01 3.62723470e-01 3.85420203e-01 -9.68752325e-01
8.26724052e-01 -2.34961128e+00 7.91863918e-01 4.59966987e-01
2.21375540e-01 -2.32102692e-01 -1.97731987e-01 4.05482948e-01
-4.29767996e-01 -3.14507365e-01 -4.86950040e-01 -5.45113444e-01
1.33948520e-01 -2.34278738e-01 -8.08945179e-01 4.33650523e-01
5.36627114e-01 6.77308559e-01 -6.40889227e-01 2.68693954e-01
1.83862984e-01 8.45660210e-01 -6.64469600e-01 2.43730441e-01
-4.56604302e-01 1.04245174e+00 -3.97148579e-01 1.89916700e-01
4.92335171e-01 2.62542963e-01 -2.44567826e-01 -4.28522468e-01
-9.44070220e-02 3.12780321e-01 -1.21082401e+00 1.88556647e+00
-1.06421471e+00 5.64354002e-01 4.15486217e-01 -1.19012296e+00
1.24980617e+00 5.73465049e-01 7.10314214e-01 -1.88916072e-01
1.50693301e-02 3.36938977e-01 7.98641741e-02 -6.55078292e-01
3.63726020e-01 -7.43907452e-01 -9.21446830e-02 7.05370963e-01
2.95750760e-02 -4.32460666e-01 -4.44306940e-01 -4.34792161e-01
8.63029480e-01 4.58046347e-01 -4.01537627e-01 -7.98623711e-02
4.75067586e-01 -3.63215953e-01 -1.33605432e-02 1.85482249e-01
6.40286386e-01 8.14893425e-01 3.30455035e-01 -4.98454750e-01
-1.10605228e+00 -1.29490757e+00 -1.34367332e-01 9.04024065e-01
-6.46664619e-01 -2.28563044e-02 -6.81787074e-01 3.18236053e-01
1.78099915e-01 7.03534126e-01 -4.57730383e-01 -5.08117914e-01
-7.65029550e-01 -5.84610343e-01 6.58345163e-01 2.50110030e-01
-1.28175840e-01 -1.08119071e+00 -4.13593620e-01 4.91126329e-01
-2.28977561e-01 -8.33204508e-01 -1.53249830e-01 2.63463289e-01
-4.95037377e-01 -6.91703916e-01 -6.69755340e-01 -5.41134834e-01
4.15495038e-02 -2.80638576e-01 1.10397720e+00 -1.52652591e-01
-6.39933825e-01 3.17722201e-01 -2.22139414e-02 -4.42902714e-01
-8.88786674e-01 5.11298329e-03 2.41009846e-01 5.84081650e-01
-2.94804633e-01 -1.17292202e+00 -4.69814420e-01 -1.45470202e-02
-9.69959497e-01 -1.29396310e-02 3.05115759e-01 6.05511844e-01
6.01059377e-01 1.73268214e-01 7.68522322e-01 -5.67998946e-01
9.70815778e-01 -3.95882905e-01 -3.92986208e-01 -1.69043303e-01
1.71829283e-01 3.11637491e-01 7.90637434e-01 -6.94080293e-01
-9.08920228e-01 1.87586486e-01 -4.45546240e-01 -5.44509709e-01
-1.18749507e-01 5.18140197e-01 -1.10026360e-01 2.24459052e-01
7.99138486e-01 9.52128097e-02 9.99415442e-02 -8.52561891e-01
5.29851913e-01 3.44127357e-01 9.09304082e-01 -9.85432386e-01
9.95200217e-01 3.73206139e-01 4.35222119e-01 -1.18241251e+00
-7.32292950e-01 -1.33436665e-01 -7.05485225e-01 -2.52799690e-01
7.88241565e-01 -7.63768971e-01 -8.76258492e-01 3.97885114e-01
-1.17046559e+00 -3.80137682e-01 -5.96978307e-01 8.59318674e-01
-8.71389031e-01 1.43771172e-01 -6.13935709e-01 -9.33802783e-01
-1.32033214e-01 -1.01692271e+00 1.32559204e+00 -4.25651938e-01
-4.88125563e-01 -9.40159142e-01 4.07776624e-01 1.68806136e-01
2.00093001e-01 5.29307544e-01 1.17419076e+00 -3.52067277e-02
-4.23179716e-01 -1.89771801e-01 3.06860328e-01 3.56299013e-01
7.90154263e-02 7.28384182e-02 -1.33675373e+00 -1.31410718e-01
3.63828927e-01 -1.56999692e-01 8.60974669e-01 8.08480501e-01
1.03791571e+00 9.40236822e-03 1.92445070e-01 7.48832643e-01
1.12706125e+00 -1.96398601e-01 3.36512715e-01 -1.81006700e-01
6.94897234e-01 8.50777090e-01 9.17652324e-02 6.42108202e-01
-2.10798562e-01 6.27372742e-01 3.00281078e-01 -4.00070194e-03
-2.23313391e-01 -2.58212328e-01 3.06508571e-01 1.09773159e+00
-3.42219174e-01 8.35453793e-02 -5.22217989e-01 1.87611744e-01
-1.25554609e+00 -7.93412864e-01 -3.43143016e-01 2.06937075e+00
7.07610011e-01 3.60592781e-03 2.37689972e-01 7.87523568e-01
5.19267321e-01 -1.17194943e-01 -5.16398728e-01 -3.22090894e-01
1.14961118e-01 8.50752294e-01 -9.19561908e-02 7.13885128e-01
-8.24314475e-01 3.62304002e-01 5.97848082e+00 4.60789889e-01
-1.19318485e+00 -2.09370419e-01 2.18241423e-01 -4.04559448e-02
-6.88763678e-01 -1.68034896e-01 -2.32471198e-01 1.69178993e-01
1.08486438e+00 2.04364017e-01 7.96687841e-01 1.00389861e-01
2.60488510e-01 5.34971297e-01 -1.14449799e+00 5.98725080e-01
-2.10738793e-01 -1.24794924e+00 -9.29381102e-02 -9.94706750e-02
3.17525268e-01 -5.05031288e-01 5.47997057e-01 -3.21887024e-02
-4.24962454e-02 -1.16879308e+00 1.13884258e+00 1.10899734e+00
9.84596670e-01 -6.12841845e-01 9.20600072e-02 4.88247722e-01
-1.04994297e+00 1.34206370e-01 -4.29346338e-02 -1.76314190e-01
2.24820048e-01 7.41327465e-01 -7.21480966e-01 5.02392471e-01
3.39341104e-01 5.88476598e-01 8.29134732e-02 7.86209524e-01
-1.50621921e-01 9.79248226e-01 -2.74323642e-01 4.45829958e-01
-3.61233205e-01 -6.11727893e-01 7.50371397e-01 7.18762755e-01
7.25695789e-01 5.07636741e-02 -1.97213709e-01 1.52608335e+00
-1.13005504e-01 -1.48444071e-01 -2.51761973e-01 -6.85909167e-02
1.87677696e-01 1.11979759e+00 -5.02392054e-01 2.42421985e-01
3.87400086e-03 4.62623894e-01 -2.52831697e-01 5.52599490e-01
-5.99406898e-01 -2.69953787e-01 5.19474924e-01 4.18343604e-01
2.01283842e-01 -4.54205662e-01 -1.20230652e-01 -7.70234764e-01
6.38886765e-02 -6.35345936e-01 -3.57832491e-01 -9.99844551e-01
-1.29015899e+00 4.73570645e-01 -1.74408779e-01 -1.38806558e+00
-5.58662176e-01 -5.53418100e-01 -7.30305791e-01 1.17252886e+00
-1.29293966e+00 -1.17615926e+00 2.40115330e-01 3.46635550e-01
2.10690364e-01 -7.80852363e-02 1.16924775e+00 1.84304327e-01
-1.81287140e-01 1.61122873e-01 3.09957355e-01 6.73288479e-02
5.93302488e-01 -1.14784241e+00 6.28587723e-01 1.98348880e-01
4.19197589e-01 4.69108909e-01 1.05617309e+00 -1.55328408e-01
-1.36347663e+00 -8.47276449e-01 2.36675426e-01 -2.26200312e-01
9.79872823e-01 -4.96749431e-01 -1.09786654e+00 2.37657070e-01
-2.78321117e-01 1.76753908e-01 7.89017618e-01 8.49289075e-02
-2.53305912e-01 -8.13321024e-02 -6.50755823e-01 2.80173212e-01
5.51063895e-01 -8.93954754e-01 -5.39705873e-01 3.21749985e-01
7.25403845e-01 -2.52513528e-01 -1.12874544e+00 3.61464322e-02
7.42028177e-01 -8.39096248e-01 1.49586034e+00 -6.38197362e-01
7.28243470e-01 -1.56775072e-01 -1.18447252e-01 -1.69236875e+00
-2.58163393e-01 -1.00726247e+00 -6.96167871e-02 1.14055157e+00
3.18578064e-01 -2.92276204e-01 6.51597679e-01 1.97609916e-01
-3.62351626e-01 -5.21112442e-01 -8.07844400e-01 -5.19867599e-01
4.31643724e-01 -6.81563079e-01 7.14346647e-01 7.61660337e-01
-4.22204912e-01 3.34056824e-01 -4.54487681e-01 4.48040634e-01
7.31760859e-01 3.94345462e-01 5.82474709e-01 -1.62084496e+00
-9.49666977e-01 -1.87304571e-01 8.23498666e-02 -8.57221663e-01
4.43485051e-01 -7.46109784e-01 3.01656812e-01 -1.04642892e+00
-4.27961618e-01 -6.15908086e-01 -9.17398632e-02 -1.88621864e-01
2.03079775e-01 4.52150077e-01 -1.11059263e-01 8.58151317e-02
4.10328209e-01 6.31206930e-01 1.46514642e+00 -1.35624379e-01
-1.83478072e-01 5.61893225e-01 -2.91397721e-01 9.04563487e-01
4.89112824e-01 -3.62289518e-01 -2.81172454e-01 -4.48694080e-01
5.82735479e-01 8.10201943e-01 7.32461870e-01 -9.58793044e-01
-4.79790755e-02 1.97610244e-01 2.91403264e-01 -2.66594678e-01
8.15588534e-01 -7.46412516e-01 4.83435005e-01 1.67537212e-01
-7.03587472e-01 -2.82151997e-01 1.72062904e-01 5.62606633e-01
-3.35875094e-01 -1.98718563e-01 5.50626338e-01 2.68523008e-01
1.31090984e-01 3.03867720e-02 -3.54839355e-01 2.95948777e-02
2.26210415e-01 2.47584172e-02 1.72908589e-01 -4.58900124e-01
-1.18396127e+00 -5.50804734e-01 -1.61891729e-01 6.58092424e-02
5.88147104e-01 -1.27173007e+00 -1.05596519e+00 8.10024738e-01
-3.66591096e-01 1.05947725e-01 4.10012811e-01 5.54922402e-01
-4.05872464e-01 1.26507625e-01 -3.63109671e-02 -5.26993454e-01
-8.98742914e-01 2.11681813e-01 4.98181432e-01 1.97435945e-01
-6.50629461e-01 8.61063182e-01 2.24181250e-01 -4.71170604e-01
-1.02830447e-01 -5.62476456e-01 -8.01091269e-02 2.53102332e-02
4.85394746e-02 2.20144257e-01 1.47302747e-01 -7.52821088e-01
5.13580590e-02 8.22629988e-01 7.27959752e-01 -5.18850923e-01
1.65447783e+00 1.77490920e-01 -1.51159421e-01 8.35216641e-01
1.41243649e+00 4.66233432e-01 -1.31230462e+00 -2.29481161e-01
-3.24711800e-01 -1.59364358e-01 2.14496240e-01 -3.73446822e-01
-1.00746155e+00 1.12190199e+00 5.66012003e-02 4.94673997e-01
9.36371207e-01 8.50036442e-02 8.71409893e-01 3.01964015e-01
-1.35052860e-01 -5.83348811e-01 3.25762331e-01 2.32061163e-01
1.14377749e+00 -6.35525227e-01 3.10275913e-03 -2.71192729e-01
-3.51324379e-02 1.42663693e+00 -2.72301525e-01 -6.60170019e-01
1.13844216e+00 4.71691251e-01 1.24981523e-01 -1.12593621e-01
-4.18005109e-01 3.53451908e-01 8.66149068e-01 4.28955525e-01
6.07746303e-01 2.15699926e-01 4.37551379e-01 1.05443978e+00
-8.87560189e-01 -2.04942927e-01 3.49329263e-01 1.83260337e-01
-2.20950812e-01 -1.18146586e+00 -6.27705574e-01 2.32085675e-01
-3.76396179e-01 -1.01590127e-01 -1.49693459e-01 3.22177619e-01
2.89291173e-01 5.63179016e-01 2.00513437e-01 -3.57394934e-01
4.65438724e-01 2.33907297e-01 5.90143085e-01 -5.44291258e-01
-2.59122163e-01 5.86404383e-01 -2.74045654e-02 1.91118687e-01
-3.34231317e-01 -7.55086064e-01 -1.19810891e+00 -3.03501692e-02
-4.61328894e-01 2.02826619e-01 8.18905354e-01 8.25386286e-01
-1.48652017e-01 1.11423934e+00 6.70384407e-01 -1.12965238e+00
-6.89422667e-01 -6.76833212e-01 -1.03473270e+00 2.11459607e-01
8.75017881e-01 -6.48366153e-01 -4.48975861e-01 2.91329861e-01] | [15.69200325012207, 5.867941856384277] |
cbf306d7-193b-41d1-9bdc-ea3957c089b0 | kelp-a-kernel-based-learning-platform-for | null | null | https://aclanthology.org/P15-4004 | https://aclanthology.org/P15-4004.pdf | KeLP: a Kernel-based Learning Platform for Natural Language Processing | null | ['Roberto Basili', 'Simone Filice', 'Giuseppe Castellucci', 'Danilo Croce'] | 2015-07-01 | kelp-a-kernel-based-learning-platform-for-1 | https://aclanthology.org/P15-4004 | https://aclanthology.org/P15-4004.pdf | ijcnlp-2015-7 | ['twitter-sentiment-analysis'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.265670299530029, 3.7381439208984375] |
43da5d2c-06b7-4310-bfe9-99198c624a42 | a-deep-neural-network-for-multi-species-fish | 2109.10664 | null | https://arxiv.org/abs/2109.10664v1 | https://arxiv.org/pdf/2109.10664v1.pdf | A deep neural network for multi-species fish detection using multiple acoustic cameras | Underwater acoustic cameras are high potential devices for many applications in ecology, notably for fisheries management and monitoring. However how to extract such data into high value information without a time-consuming entire dataset reading by an operator is still a challenge. Moreover the analysis of acoustic imaging, due to its low signal-to-noise ratio, is a perfect training ground for experimenting with new approaches, especially concerning Deep Learning techniques. We present hereby a novel approach that takes advantage of both CNN (Convolutional Neural Network) and classical CV (Computer Vision) techniques, able to detect a generic class ''fish'' in acoustic video streams. The pipeline pre-treats the acoustic images to extract 2 features, in order to localise the signals and improve the detection performances. To ensure the performances from an ecological point of view, we propose also a two-step validation, one to validate the results of the trainings and one to test the method on a real-world scenario. The YOLOv3-based model was trained with data of fish from multiple species recorded by the two common acoustic cameras, DIDSON and ARIS, including species of high ecological interest, as Atlantic salmon or European eels. The model we developed provides satisfying results detecting almost 80% of fish and minimizing the false positive rate, however the model is much less efficient for eel detections on ARIS videos. The first CNN pipeline for fish monitoring exploiting video data from two models of acoustic cameras satisfies most of the required features. Many challenges are still present, such as the automation of fish species identification through a multiclass model. 1 However the results point a new solution for dealing with complex data, such as sonar data, which can also be reapplied in other cases where the signal-to-noise ratio is a challenge. | ['Thomas Corpetti', 'Laurent Beaulaton', 'Marie Nevoux', 'François Martignac', 'Guglielmo Fernandez', 'Garcia Fernandez'] | 2021-09-22 | null | null | null | null | ['fish-detection'] | ['computer-vision'] | [ 7.60799497e-02 -1.02909461e-01 8.48848820e-01 -2.70234048e-01
-3.87803048e-01 -5.57306349e-01 4.96589899e-01 3.02110523e-01
-1.32741511e+00 3.25094759e-01 -4.58270371e-01 1.80774167e-01
-2.13415638e-01 -7.84837306e-01 -7.44303465e-01 -9.33482051e-01
-6.22522473e-01 4.06898171e-01 6.66601598e-01 -1.12426691e-01
3.66425067e-02 8.75182748e-01 -2.19557691e+00 7.84631148e-02
1.81686252e-01 7.94238985e-01 5.62092543e-01 8.43605697e-01
-7.85721317e-02 2.39133939e-01 -5.49505472e-01 -3.19717079e-01
3.30361694e-01 -2.81572312e-01 -1.28729746e-01 -2.54783154e-01
4.92799908e-01 -3.40784371e-01 3.41922462e-01 1.13419342e+00
4.58851397e-01 -5.79994768e-02 6.51935637e-01 -8.65150332e-01
5.40755451e-01 7.32623816e-01 -1.78655818e-01 2.33764693e-01
9.84120369e-02 6.26182497e-01 6.81482732e-01 -6.31476462e-01
3.04415911e-01 1.08604228e+00 8.50188613e-01 6.13425553e-01
-1.14526618e+00 -6.73559427e-01 -3.74454558e-01 2.88098097e-01
-1.20268643e+00 -3.31335545e-01 2.73235232e-01 -7.63665199e-01
6.03390872e-01 1.26947299e-01 9.89439487e-01 8.85462940e-01
-8.42587054e-02 2.93830335e-01 1.18480110e+00 -1.40045002e-01
2.79727757e-01 2.70954430e-01 -7.55648091e-02 2.57457078e-01
4.83675390e-01 2.09632427e-01 -1.92225829e-01 1.73314393e-01
4.77579802e-01 2.58623585e-02 -4.43252951e-01 1.54250376e-02
-7.77038395e-01 7.25422859e-01 2.92322159e-01 6.84420407e-01
-1.86808541e-01 1.72004491e-01 5.55729210e-01 4.21211481e-01
1.05402432e-01 4.75869060e-01 -5.52770615e-01 1.42446181e-04
-1.03063512e+00 1.09257601e-01 9.86762702e-01 3.30154032e-01
8.99913847e-01 2.51373827e-01 5.45929968e-01 6.81386411e-01
6.87684417e-01 8.93037856e-01 4.30474579e-01 -4.05783534e-01
-8.39907080e-02 5.32759249e-01 -1.53451726e-01 -8.07207763e-01
-7.64684141e-01 -5.70945561e-01 -5.80052793e-01 8.22202325e-01
4.59855348e-01 -1.73369378e-01 -6.77751184e-01 1.41568208e+00
2.39893019e-01 2.25125000e-01 2.92094469e-01 9.90704060e-01
1.07846987e+00 6.92140818e-01 5.83676882e-02 -2.38451883e-01
1.50331128e+00 -1.53044581e-01 -4.83291358e-01 -5.11005223e-02
5.92066944e-01 -6.71587765e-01 4.73863244e-01 6.63754821e-01
-6.34491026e-01 -5.53325295e-01 -1.04155743e+00 6.46482825e-01
-6.46852970e-01 3.90071958e-01 -1.35021746e-01 8.13234091e-01
-1.12924576e+00 5.92276394e-01 -8.45686197e-01 -4.49082226e-01
1.16786651e-01 5.73669076e-01 -8.35669518e-01 3.15660924e-01
-8.75443459e-01 9.16273296e-01 3.31671476e-01 7.08748758e-01
-1.38418698e+00 -4.65218365e-01 -8.43803406e-01 2.97618061e-01
5.78629114e-02 6.95361495e-02 1.06440365e+00 -1.10038114e+00
-1.50511611e+00 9.98016298e-01 5.84035337e-01 -6.91426039e-01
8.31346512e-01 -3.24708335e-02 -1.77073076e-01 3.10009360e-01
-1.21520162e-01 4.24609751e-01 6.95126772e-01 -1.28702927e+00
-9.47606146e-01 -2.66787857e-01 -6.93455413e-02 -1.70428544e-01
-2.39821464e-01 2.08551791e-02 -7.22666681e-02 -1.97833195e-01
-2.32250154e-01 -7.66933143e-01 -1.86213225e-01 1.08354755e-01
1.99653819e-01 4.25247401e-02 7.04665780e-01 -4.20665026e-01
4.67060953e-01 -2.30655146e+00 8.81992579e-02 2.49018192e-01
-7.08154589e-02 7.84664273e-01 -1.24829322e-01 5.33653378e-01
8.10815990e-02 -1.63612336e-01 -4.59366560e-01 -4.71392602e-01
-2.90986925e-01 3.03447932e-01 3.93517286e-01 8.59831333e-01
1.69554994e-01 -9.28818733e-02 -8.75127494e-01 -4.39192027e-01
4.75628763e-01 5.57115734e-01 -3.60211730e-01 6.43697143e-01
5.44055440e-02 5.64666033e-01 7.29315504e-02 3.09424609e-01
1.00033379e+00 7.11783707e-01 -4.43080477e-02 -2.11405322e-01
-6.81058228e-01 -5.58891296e-01 -1.48114491e+00 1.10674608e+00
-5.44927537e-01 7.69054115e-01 8.14008057e-01 -1.19462538e+00
1.22982228e+00 4.41116631e-01 2.82415032e-01 -1.98518455e-01
3.84517640e-01 6.16576612e-01 1.41999140e-01 -9.23216403e-01
1.28431812e-01 -4.61144269e-01 3.61112952e-01 -1.61219954e-01
4.56902236e-01 -3.92185263e-02 3.34243178e-01 -4.60044920e-01
7.29157746e-01 3.73120345e-02 1.75681878e-02 -6.08226836e-01
9.13509965e-01 -1.31450102e-01 3.33470494e-01 5.61787844e-01
-6.09484054e-02 4.82671350e-01 4.51242179e-01 -5.70855796e-01
-7.56997585e-01 -5.25755882e-01 -3.94517392e-01 7.91112065e-01
-4.06555980e-02 1.15629941e-01 -8.21108162e-01 -2.95028538e-01
-2.06926212e-01 7.82803372e-02 -6.59679890e-01 1.44538522e-01
-5.35212994e-01 -7.57599056e-01 7.86014140e-01 1.61226299e-02
3.87255818e-01 -1.17006767e+00 -1.38129961e+00 3.01494807e-01
3.87869626e-01 -1.09560692e+00 5.47823608e-01 4.49417114e-01
-7.92704046e-01 -1.11601138e+00 -7.97595143e-01 -7.64203429e-01
2.49058992e-01 4.78145368e-02 7.79553115e-01 2.06909671e-01
-4.34716970e-01 3.96775931e-01 -6.61867440e-01 -4.67904031e-01
-8.16763759e-01 -1.51118189e-01 2.72565484e-02 2.70978570e-01
4.36082482e-01 -5.24027288e-01 -5.42448997e-01 3.65003556e-01
-1.18339205e+00 -7.55152583e-01 8.88280213e-01 6.78067148e-01
3.16980667e-02 -3.33825439e-01 2.77506322e-01 -5.28659642e-01
-1.08814441e-01 -3.29948097e-01 -1.02232111e+00 -1.81796536e-01
7.57186264e-02 -1.56396061e-01 7.20812917e-01 -2.92265296e-01
-5.53701997e-01 6.16324067e-01 -9.03582990e-01 -2.81125724e-01
-8.30519557e-01 4.39056128e-01 -6.31294027e-02 -3.04149926e-01
5.97095728e-01 2.70559132e-01 2.68539757e-01 -8.90923858e-01
-3.75196636e-01 6.00206017e-01 3.42378914e-01 1.09892860e-01
8.43872011e-01 3.88237774e-01 2.98720449e-01 -1.45600748e+00
-3.05959824e-02 -7.00674295e-01 -5.43598711e-01 -7.26853788e-01
1.10771501e+00 -8.89598072e-01 -1.01619399e+00 8.18505347e-01
-1.22568858e+00 -9.72427428e-02 -2.31161654e-01 9.15125132e-01
-7.71530122e-02 5.30098498e-01 -3.92355293e-01 -1.18248737e+00
-2.66396731e-01 -1.60156572e+00 9.62834179e-01 4.54676360e-01
6.24152005e-01 -7.42201209e-01 3.35431576e-01 -2.80625224e-01
4.09995794e-01 1.47752434e-01 1.44877061e-01 -1.01978540e+00
-3.47294122e-01 -3.50324839e-01 6.86917081e-02 6.70500934e-01
-3.38512927e-01 4.64979172e-01 -1.27833927e+00 -2.81086177e-01
-1.41457208e-02 -1.92223668e-01 1.08319306e+00 2.44631276e-01
4.80377614e-01 2.09254980e-01 2.78518647e-02 4.84175295e-01
1.73771262e+00 1.17283039e-01 3.53834063e-01 3.72941077e-01
1.42761812e-01 1.06214833e+00 5.26662409e-01 5.40739357e-01
-1.42286718e-01 7.79668748e-01 1.16059124e+00 -1.76672056e-01
8.68166238e-02 2.60479420e-01 6.57920539e-01 2.53926456e-01
-3.70352805e-01 -9.00287256e-02 -7.81920135e-01 7.63106346e-01
-1.45121598e+00 -9.63531733e-01 -8.06972027e-01 2.42257214e+00
2.05922872e-01 7.07457885e-02 3.37237328e-01 4.65607703e-01
6.50818944e-01 -3.23336452e-01 2.74574786e-01 -3.89495671e-01
-1.54460177e-01 3.16990733e-01 7.50868857e-01 4.78835583e-01
-1.22093987e+00 4.10220504e-01 4.74876547e+00 5.37983894e-01
-1.33234930e+00 2.48755384e-02 -7.40115568e-02 2.72042304e-01
3.46583128e-01 -2.58531183e-01 -1.08955204e+00 4.09174412e-01
1.05654716e+00 6.25911593e-01 -4.56927568e-02 9.02314007e-01
3.66792589e-01 -2.74003208e-01 -8.58498871e-01 8.68751585e-01
2.02337950e-01 -8.88277531e-01 -1.46383882e-01 2.11858731e-02
1.32322103e-01 2.03387722e-01 -4.31827366e-01 1.09113850e-01
-2.39906326e-01 -7.31307328e-01 8.08827221e-01 6.05005443e-01
4.86275584e-01 -4.50178325e-01 1.52193630e+00 4.34424013e-01
-1.25154197e+00 -1.35866612e-01 -5.93649209e-01 -6.91564381e-02
2.23226428e-01 3.40273380e-01 -6.33444965e-01 4.09549356e-01
1.07650757e+00 5.65989256e-01 -6.64404094e-01 1.69205284e+00
6.87630400e-02 5.24136901e-01 -7.51516402e-01 -4.27659601e-01
5.07509708e-01 -2.20323026e-01 8.17565382e-01 1.90439963e+00
6.53403223e-01 -2.32125625e-01 -3.09074044e-01 5.41999400e-01
3.98084998e-01 4.15015548e-01 -7.20982194e-01 3.14013839e-01
-3.97786163e-02 1.49306393e+00 -9.15592849e-01 -1.74910069e-01
-2.80973345e-01 3.43770921e-01 -2.21786961e-01 -1.64280221e-01
-4.62161034e-01 -3.31145853e-01 6.05898142e-01 -7.73230344e-02
4.74983275e-01 -1.24986179e-01 4.42875713e-01 -9.50497448e-01
-3.34057301e-01 -5.36076367e-01 1.90639913e-01 -3.42308313e-01
-1.12489986e+00 8.73891413e-01 2.29829565e-01 -1.78429091e+00
-1.98469415e-01 -1.26794577e+00 -5.11211812e-01 6.80927873e-01
-1.67896485e+00 -1.03530049e+00 -6.01621270e-01 1.98014677e-01
5.13255358e-01 -1.79999813e-01 6.20688975e-01 5.53943396e-01
-2.68687844e-01 1.24385171e-01 2.27080479e-01 2.01598987e-01
5.67324221e-01 -1.31703842e+00 -3.39493841e-01 9.48749661e-01
1.01182044e-01 4.51905318e-02 1.08199334e+00 -2.20846027e-01
-1.32251132e+00 -6.37836814e-01 5.94533920e-01 5.29624596e-02
6.40154839e-01 -2.99538434e-01 -9.39570963e-01 1.97776124e-01
2.51877427e-01 1.77518316e-02 7.04484105e-01 -5.89786291e-01
1.77607015e-01 -4.87868458e-01 -1.03883171e+00 -5.12316488e-02
2.51716673e-01 8.83548632e-02 -5.79655886e-01 9.74759534e-02
-8.07943791e-02 -4.74346876e-02 -9.20676470e-01 3.82855117e-01
7.89417326e-01 -1.40488935e+00 6.71941757e-01 -1.31179839e-01
2.66854614e-01 -6.10680044e-01 1.40367210e-01 -1.30837274e+00
2.44745940e-01 -2.95530893e-02 9.00118947e-01 1.21673191e+00
4.01871115e-01 -5.73067725e-01 3.67695659e-01 -1.31570905e-01
-2.32970297e-01 2.14608908e-01 -1.31401622e+00 -7.76534915e-01
-1.70242399e-01 -4.02006209e-01 8.26642066e-02 5.29267371e-01
-3.50935966e-01 6.26616627e-02 -5.50412059e-01 6.85167789e-01
6.90664172e-01 -2.30951279e-01 1.08355844e+00 -1.68441141e+00
-5.02836704e-01 -6.21556818e-01 -1.17312968e+00 -5.53697705e-01
-1.03181049e-01 -2.80795693e-01 4.36121643e-01 -1.17236662e+00
-3.76622051e-01 1.34380292e-02 7.25847716e-03 2.72148013e-01
2.97251910e-01 7.39303291e-01 2.52075940e-01 -1.98123921e-02
-3.92663665e-02 2.63342351e-01 6.06389999e-01 5.55182695e-02
3.49235930e-03 3.42483491e-01 4.42716062e-01 7.42979348e-01
2.99192101e-01 -6.03262246e-01 2.57475108e-01 -3.56313854e-01
2.77192861e-01 3.56927589e-02 4.46172982e-01 -1.55979276e+00
4.41486478e-01 3.12750190e-01 -7.19254836e-02 -3.35128069e-01
3.97087663e-01 -1.50412977e+00 4.69328284e-01 1.15794933e+00
-9.58219320e-02 -6.13918826e-02 3.04458827e-01 5.81520975e-01
-5.09327948e-01 -1.16844893e+00 1.27567649e+00 -4.91366476e-01
-8.95093262e-01 -1.21135034e-01 -9.72444713e-01 -3.84641856e-01
9.51374590e-01 -3.29463363e-01 -1.22216843e-01 -1.62316114e-01
-8.44944656e-01 5.48321530e-02 2.26321280e-01 -4.22683246e-02
4.06784445e-01 -3.22675109e-01 -1.10095346e+00 3.57248902e-01
2.77391046e-01 -1.38609171e-01 5.77709019e-01 9.62452054e-01
-1.39477527e+00 -1.82734102e-01 -6.26104891e-01 -1.00675952e+00
-1.60294986e+00 3.53598267e-01 7.48225093e-01 2.22474441e-01
-4.09606934e-01 7.67404556e-01 1.32261589e-02 -8.50270167e-02
2.83814460e-01 -4.54007208e-01 -1.05097997e+00 5.45118272e-01
6.50841594e-01 3.40007126e-01 2.08327383e-01 -8.54260027e-01
-3.76856923e-01 1.03471351e+00 5.70231795e-01 -2.24238951e-02
1.73059487e+00 1.04641438e-01 -1.42498806e-01 4.48131442e-01
1.11805451e+00 8.11869875e-02 -1.36002398e+00 2.76124448e-01
-4.41877246e-02 -3.89358819e-01 7.24883750e-02 -3.76665026e-01
-1.20847678e+00 1.40009069e+00 1.16507554e+00 6.56516075e-01
9.64362919e-01 -1.42635822e-01 -3.30684111e-02 3.39018434e-01
2.40870699e-01 -7.02033043e-01 -4.87400472e-01 2.51981109e-01
7.03150988e-01 -1.27195871e+00 -2.19411328e-01 -3.42114493e-02
-3.44872564e-01 1.73354757e+00 3.47430944e-01 -2.12634653e-01
6.51568115e-01 6.50063992e-01 4.31303412e-01 -7.33008161e-02
-3.18821102e-01 -9.20852900e-01 -1.60218328e-01 7.63309956e-01
1.88843116e-01 -1.34947952e-02 -6.32313907e-01 4.38720524e-01
3.77927125e-02 -2.02751771e-01 8.21257591e-01 5.53974450e-01
-6.87257230e-01 -9.27165627e-01 -5.48779964e-01 3.15623917e-02
-5.67011893e-01 1.88012972e-01 -2.04052806e-01 9.37135637e-01
5.30610144e-01 8.35755050e-01 3.05536157e-03 -5.09987533e-01
6.85971200e-01 -1.32073671e-01 -8.08653459e-02 -4.83643591e-01
-1.17272735e+00 2.46732801e-01 -1.86701398e-02 -6.83022514e-02
-1.09677708e+00 -8.35374415e-01 -6.82103634e-01 1.69018969e-01
-4.45772380e-01 4.58054572e-01 1.16917217e+00 8.67766082e-01
-3.97556812e-01 2.08581179e-01 7.19337046e-01 -1.39536607e+00
-4.40478802e-01 -1.23894644e+00 -8.11423182e-01 2.94323832e-01
4.91475582e-01 -7.20999360e-01 -9.48253155e-01 2.07290843e-01] | [8.551814079284668, -1.1948878765106201] |
87233faa-f22e-40aa-88e1-5f78bdc61bb6 | a-codec-information-assisted-framework-for | 2210.08229 | null | https://arxiv.org/abs/2210.08229v1 | https://arxiv.org/pdf/2210.08229v1.pdf | A Codec Information Assisted Framework for Efficient Compressed Video Super-Resolution | Online processing of compressed videos to increase their resolutions attracts increasing and broad attention. Video Super-Resolution (VSR) using recurrent neural network architecture is a promising solution due to its efficient modeling of long-range temporal dependencies. However, state-of-the-art recurrent VSR models still require significant computation to obtain a good performance, mainly because of the complicated motion estimation for frame/feature alignment and the redundant processing of consecutive video frames. In this paper, considering the characteristics of compressed videos, we propose a Codec Information Assisted Framework (CIAF) to boost and accelerate recurrent VSR models for compressed videos. Firstly, the framework reuses the coded video information of Motion Vectors to model the temporal relationships between adjacent frames. Experiments demonstrate that the models with Motion Vector based alignment can significantly boost the performance with negligible additional computation, even comparable to those using more complex optical flow based alignment. Secondly, by further making use of the coded video information of Residuals, the framework can be informed to skip the computation on redundant pixels. Experiments demonstrate that the proposed framework can save up to 70% of the computation without performance drop on the REDS4 test videos encoded by H.264 when CRF is 23. | ['Li Song', 'Rong Xie', 'Youliang Yan', 'Jiaming Guo', 'Xueyi Zou', 'Hengsheng Zhang'] | 2022-10-15 | null | null | null | null | ['video-super-resolution'] | ['computer-vision'] | [ 3.25649053e-01 -5.53439081e-01 -2.82185644e-01 -2.34208956e-01
-4.80825722e-01 1.68687925e-02 2.08425641e-01 -2.22126499e-01
-4.19369996e-01 5.63419163e-01 3.35735828e-01 -1.27204254e-01
-4.09279298e-03 -5.27726769e-01 -6.52017891e-01 -7.00610816e-01
-2.77801573e-01 -5.16958773e-01 4.83284235e-01 -3.32272649e-01
4.47394937e-01 4.34600204e-01 -1.70011985e+00 5.03806293e-01
7.25868821e-01 1.13803732e+00 7.45259941e-01 7.43594825e-01
-7.69594014e-02 1.45063686e+00 -2.06646219e-01 5.06203063e-02
2.85135329e-01 -4.44356084e-01 -5.03928542e-01 4.04017754e-02
2.46436641e-01 -7.23393679e-01 -8.44632864e-01 9.78066146e-01
4.83863324e-01 3.84009808e-01 1.24547789e-02 -5.69738567e-01
-6.25053227e-01 6.20153785e-01 -7.56909132e-01 6.79028749e-01
3.21370602e-01 -1.44763052e-01 8.01086068e-01 -1.12682009e+00
7.39601016e-01 1.08402169e+00 3.09645355e-01 3.91430855e-01
-7.62639761e-01 -5.68220377e-01 2.09768087e-01 9.04190361e-01
-1.58892298e+00 -6.45647764e-01 7.39565492e-01 -9.95851457e-02
9.82786596e-01 1.99452519e-01 5.50649285e-01 6.68457925e-01
2.29530126e-01 5.59393704e-01 7.14426041e-01 -8.80643353e-02
5.71504459e-02 -3.66049826e-01 7.90869519e-02 5.83399594e-01
-1.49507955e-01 9.47771668e-02 -8.61321211e-01 3.75351697e-01
1.11707497e+00 1.98862910e-01 -6.07710779e-01 1.27607509e-01
-1.15272713e+00 5.64227343e-01 4.99619275e-01 3.63316894e-01
-5.90019286e-01 1.65790185e-01 4.30117458e-01 1.58891141e-01
3.30673993e-01 -1.92124054e-01 -1.11341685e-01 -2.33637720e-01
-1.33096290e+00 -1.51733845e-01 5.65698482e-02 9.44965661e-01
6.65788710e-01 6.62660062e-01 -1.83941007e-01 9.29494739e-01
2.03582615e-01 3.62746745e-01 6.25003934e-01 -1.43323684e+00
6.26413941e-01 3.12060803e-01 1.39684215e-01 -1.28385496e+00
-1.44248933e-01 -3.93625855e-01 -1.28661704e+00 -1.63735747e-02
-2.56197322e-02 1.19282916e-01 -6.53805614e-01 1.22195816e+00
7.08490312e-02 5.95864892e-01 2.69031107e-01 1.23646426e+00
7.82931387e-01 1.06053281e+00 -2.43915886e-01 -7.56251752e-01
1.16441882e+00 -1.02049446e+00 -1.04344261e+00 6.16266169e-02
3.80015045e-01 -8.95909011e-01 4.58554804e-01 2.57289022e-01
-1.32960796e+00 -9.89439011e-01 -1.19565904e+00 -2.34701142e-01
2.23925963e-01 1.98333696e-01 2.40718067e-01 1.39508739e-01
-1.24074721e+00 8.21760893e-01 -1.12103295e+00 2.29310527e-01
2.20354542e-01 3.63455415e-01 -1.05028585e-01 -3.03894579e-01
-1.21805394e+00 3.98666859e-01 3.10754955e-01 5.09603620e-01
-6.55216336e-01 -6.16084456e-01 -9.37281609e-01 2.49227554e-01
3.80567074e-01 -2.13554487e-01 7.78991640e-01 -9.99409437e-01
-1.48994076e+00 -1.79475266e-02 -6.40194237e-01 -7.30138242e-01
4.20974135e-01 -3.83336544e-01 -5.30811667e-01 5.17412364e-01
-3.52092385e-01 5.25794506e-01 1.03321552e+00 -7.29040205e-01
-7.47732520e-01 -1.12038702e-01 3.94025212e-03 3.17851514e-01
-3.13654989e-01 2.09254518e-01 -6.79298222e-01 -9.03288424e-01
1.42931849e-01 -9.40326571e-01 -3.62875760e-01 -1.38960466e-01
8.65607113e-02 1.82378203e-01 1.00320709e+00 -1.10276961e+00
1.58067262e+00 -2.35364318e+00 3.03232104e-01 -1.97646037e-01
1.62423819e-01 4.82484043e-01 -1.45601630e-01 -6.80455565e-02
-1.09855026e-01 -1.59548834e-01 -2.79995557e-02 -8.44145343e-02
-6.31448865e-01 8.13663453e-02 -3.17743033e-01 2.70025581e-01
1.41351461e-01 5.67704797e-01 -5.46868920e-01 -5.98820686e-01
4.62179989e-01 1.06635439e+00 -6.57739580e-01 2.84033835e-01
2.11733922e-01 5.08993089e-01 -3.32665175e-01 3.68613750e-01
7.22958386e-01 -3.28981161e-01 1.56847045e-01 -5.74113905e-01
-5.23270369e-01 5.33203706e-02 -1.33398139e+00 1.64413059e+00
-4.19614613e-01 7.56783962e-01 2.06637233e-02 -8.56577516e-01
9.12064493e-01 3.60370934e-01 5.76562405e-01 -8.58911276e-01
-4.41366388e-03 1.73524588e-01 -8.42336789e-02 -5.07776380e-01
8.97413135e-01 4.36497450e-01 5.74626386e-01 -2.18569338e-02
-3.48836690e-01 6.13253176e-01 3.89264435e-01 1.97780862e-01
9.28124607e-01 4.53602284e-01 2.58383065e-01 9.92765203e-02
1.06292033e+00 -4.46342647e-01 9.13074076e-01 2.58251905e-01
-1.41150191e-01 7.61892259e-01 -6.15134053e-02 -5.92381895e-01
-1.06189263e+00 -5.89357257e-01 8.03382024e-02 9.13286805e-01
2.98559129e-01 -5.03721714e-01 -5.46631753e-01 -2.08540191e-03
-6.17388129e-01 3.65810096e-01 -2.22508952e-01 1.83150232e-01
-1.25758982e+00 -6.38589621e-01 -4.96306531e-02 5.87307811e-01
8.32246363e-01 -9.17729735e-01 -9.05365050e-01 4.05309051e-01
-4.98960435e-01 -1.56525445e+00 -5.57239413e-01 -4.40013856e-01
-1.17528844e+00 -7.58269072e-01 -1.04038930e+00 -7.17526317e-01
4.10895348e-01 8.74937117e-01 6.49144113e-01 2.81049252e-01
-1.09657079e-01 -5.82863800e-02 -6.18004084e-01 6.40960336e-01
-4.61618185e-01 -2.30038837e-01 -1.14148259e-01 1.88313469e-01
-1.69619489e-02 -5.71824133e-01 -8.34106565e-01 4.53719079e-01
-1.00109816e+00 3.70050997e-01 5.05476773e-01 6.31996453e-01
7.37886012e-01 6.85009360e-02 3.17897469e-01 -5.10066688e-01
-4.98149591e-03 -2.71419823e-01 -6.97092116e-01 1.76500410e-01
-3.74567389e-01 9.12254304e-02 8.60014379e-01 -3.11990976e-01
-1.22825992e+00 -6.96906671e-02 -1.62028167e-02 -8.86003613e-01
2.98671633e-01 4.87359285e-01 2.08717808e-01 -1.07834712e-01
7.61578977e-02 5.71353614e-01 -1.68683693e-01 -4.50436622e-01
1.18255556e-01 5.92991233e-01 6.41353250e-01 -5.03572635e-02
5.78525364e-01 4.51230168e-01 1.83372974e-01 -1.06292200e+00
-3.79444897e-01 -3.65044504e-01 -5.09444475e-01 -3.01770896e-01
9.57749724e-01 -1.27628434e+00 -6.80035055e-01 3.43654811e-01
-1.18269920e+00 2.34314762e-02 1.88125640e-01 8.74117494e-01
-3.95803779e-01 7.59064496e-01 -1.16912925e+00 -6.86225176e-01
-4.40570056e-01 -1.48736823e+00 6.27246737e-01 2.99652010e-01
2.49505252e-01 -6.33334577e-01 -3.53639364e-01 4.57658112e-01
7.49400258e-01 -1.00151300e-01 5.69104791e-01 5.13526760e-02
-1.16204572e+00 1.89550504e-01 -4.86920834e-01 3.71008366e-01
2.16594413e-01 7.42730200e-02 -5.68797529e-01 -3.91669780e-01
1.39216304e-01 2.77005821e-01 9.09847796e-01 6.02026582e-01
1.17017162e+00 -3.22940737e-01 1.70980185e-01 8.05859625e-01
1.52482998e+00 4.56327558e-01 8.67449999e-01 3.56715143e-01
9.52215374e-01 4.40660387e-01 9.22283590e-01 7.14675009e-01
2.46058688e-01 8.74948442e-01 3.30045521e-01 1.23996489e-01
-3.14983875e-01 7.03285635e-02 7.13526428e-01 1.38173878e+00
-7.58256257e-01 -4.08400409e-02 -4.35078323e-01 3.14507723e-01
-1.84547162e+00 -1.23814225e+00 -1.15552142e-01 2.19503736e+00
5.72734058e-01 -9.98668596e-02 -3.00273716e-01 2.36874878e-01
1.04337549e+00 6.08152151e-01 -5.08160651e-01 -3.28353643e-01
-2.28571147e-01 -1.42210340e-02 5.27137935e-01 6.27236485e-01
-8.13349783e-01 8.29076350e-01 5.21421814e+00 9.00605500e-01
-1.14227092e+00 1.55263133e-02 6.87685609e-01 -3.04984599e-01
8.43852311e-02 -1.23088390e-01 -8.61200988e-01 4.97318983e-01
1.22041357e+00 2.96336375e-02 7.32971430e-01 5.80714941e-01
8.02885830e-01 5.65285683e-02 -6.68939650e-01 1.22953963e+00
1.25347584e-01 -1.51262760e+00 2.46934429e-01 -1.30092412e-01
6.97194874e-01 -3.20078433e-02 6.77412152e-02 -5.82663789e-02
-2.56260008e-01 -8.32732022e-01 6.54473543e-01 5.18419921e-01
8.43281925e-01 -9.20247793e-01 8.94616604e-01 -3.00224312e-02
-1.72266626e+00 -1.88403279e-01 -5.78981996e-01 4.70257364e-02
4.78823930e-01 3.84330511e-01 -1.87350661e-01 7.13145375e-01
8.93473983e-01 1.23070252e+00 -4.44164485e-01 6.96174145e-01
1.07747599e-01 4.50335592e-01 3.07381190e-02 4.76933062e-01
1.47768036e-01 -3.85322481e-01 5.50618887e-01 1.08008146e+00
5.92219353e-01 3.72319698e-01 -4.73637767e-02 3.62189382e-01
1.15091011e-01 1.71960250e-01 -1.45335391e-01 1.65183276e-01
2.83551365e-01 1.08616340e+00 -4.56121773e-01 -4.87411678e-01
-6.64845705e-01 1.06962848e+00 8.43716636e-02 5.29642284e-01
-1.00475335e+00 -7.15036392e-02 4.97581661e-01 6.24708943e-02
7.07205653e-01 -4.45153236e-01 2.59843379e-01 -1.26204181e+00
1.38047069e-01 -7.98313677e-01 2.26438105e-01 -8.43883395e-01
-5.44438660e-01 8.09367359e-01 -2.53172964e-01 -1.60769403e+00
-4.83858049e-01 -1.88844517e-01 -7.84377530e-02 6.42890871e-01
-1.91303861e+00 -7.22818911e-01 -4.80660945e-01 7.47437537e-01
1.11956835e+00 -1.32077575e-01 2.37752646e-01 5.79518735e-01
-7.16646314e-01 3.59231502e-01 1.00893229e-01 1.08724087e-01
5.52587032e-01 -5.65075457e-01 1.91020936e-01 1.24741173e+00
-3.89420092e-02 5.89609742e-01 6.68840885e-01 -5.38736999e-01
-1.61304212e+00 -1.13500416e+00 6.45651460e-01 3.66506457e-01
5.15078127e-01 2.51869828e-01 -1.23052859e+00 4.54205424e-01
1.65381387e-01 3.89491618e-01 3.56381804e-01 -6.18811369e-01
-1.74315542e-01 -2.80883938e-01 -6.82920098e-01 6.18644416e-01
1.03351057e+00 -4.47748095e-01 -1.97976753e-01 -1.63635910e-01
8.59916508e-01 -4.49781209e-01 -1.01604354e+00 5.24684548e-01
5.22340059e-01 -1.20830202e+00 1.06047571e+00 1.27666518e-02
8.50103796e-01 -6.73713028e-01 -5.05725443e-01 -6.45242572e-01
-3.98276508e-01 -6.50603831e-01 -4.61547703e-01 1.18000937e+00
-6.33266754e-03 -2.73903221e-01 5.69657505e-01 5.09256423e-01
1.80518478e-02 -6.21285796e-01 -9.17487025e-01 -4.84002918e-01
-6.99414492e-01 -3.17765534e-01 2.22496554e-01 7.50816762e-01
-3.42901826e-01 6.43769279e-02 -8.34280849e-01 3.44035059e-01
6.19625211e-01 1.79776058e-01 3.29610765e-01 -5.92215657e-01
-3.94219637e-01 -1.01219743e-01 -5.25137544e-01 -1.40607119e+00
5.47221769e-03 -5.25850296e-01 -1.33150190e-01 -1.25290561e+00
1.24863222e-01 -8.89310911e-02 -5.63878238e-01 5.26023805e-02
-2.32130826e-01 4.46252078e-01 5.87360024e-01 4.61889058e-01
-7.02934861e-01 6.82336330e-01 1.18297100e+00 9.11051705e-02
-3.07909757e-01 -1.81781441e-01 -1.63559526e-01 7.48705983e-01
5.91174901e-01 -2.08099648e-01 -3.73736620e-01 -5.61830521e-01
8.31930339e-02 7.69077957e-01 2.10594505e-01 -1.07400155e+00
3.80390108e-01 -7.81734213e-02 4.36257929e-01 -8.16351891e-01
4.77228791e-01 -8.43405843e-01 3.01351428e-01 4.61719036e-01
-4.90260243e-01 3.49505156e-01 5.45410998e-02 6.33150399e-01
-4.92593974e-01 -5.26525788e-02 9.21305776e-01 4.76378249e-03
-1.01184726e+00 3.08025718e-01 -4.88895506e-01 -3.64798486e-01
6.98387504e-01 -3.30383778e-01 -1.72160238e-01 -3.63686681e-01
-5.49910724e-01 -2.13208329e-02 3.75812531e-01 5.15881598e-01
1.09436560e+00 -1.11475539e+00 -8.48631084e-01 1.36664063e-01
-5.56385517e-01 -1.01440333e-01 8.78790915e-01 7.78775096e-01
-7.87893772e-01 4.87081409e-01 -2.74291933e-01 -6.34861290e-01
-1.55304909e+00 6.96850777e-01 -2.80341264e-02 -2.12301612e-01
-9.09917951e-01 4.12327856e-01 2.02683762e-01 6.33678615e-01
-3.57735530e-02 -2.19192669e-01 -6.77468061e-01 -9.66911912e-02
1.05957866e+00 8.27472746e-01 -1.69778436e-01 -1.01174736e+00
-2.10995808e-01 8.85827839e-01 -2.65826106e-01 1.17597125e-01
1.38462698e+00 -5.79589009e-01 -1.06989525e-01 1.78880379e-01
1.28161705e+00 3.96823976e-03 -1.52555370e+00 -3.26852202e-01
-1.76825330e-01 -6.90935075e-01 3.65861505e-01 -1.00181751e-01
-1.57557034e+00 8.89499247e-01 7.93054760e-01 -2.08542928e-01
1.46124089e+00 -6.20024264e-01 1.03360593e+00 1.76942870e-01
4.60357726e-01 -9.71206188e-01 -8.63905698e-02 4.54204857e-01
7.19360292e-01 -1.04426634e+00 2.66971052e-01 -5.35305381e-01
-5.93464732e-01 1.43426716e+00 3.05827111e-01 -2.10024342e-01
3.27978224e-01 1.70493990e-01 -9.77903977e-02 4.28638905e-01
-8.78386796e-01 9.70828161e-02 1.65909737e-01 1.69928595e-01
5.26545644e-01 -3.77192736e-01 -3.30822974e-01 1.78725660e-01
1.28450379e-01 2.71443963e-01 7.92178810e-01 7.29344249e-01
-3.98828298e-01 -7.63895035e-01 -2.84338593e-01 1.47173718e-01
-7.56460369e-01 -3.42598647e-01 5.68576038e-01 4.06913966e-01
-2.72744685e-01 1.03923786e+00 1.15562342e-01 -3.93599510e-01
-3.71540487e-02 -4.03113872e-01 2.29524612e-01 -4.33219224e-02
-2.74932206e-01 4.51810718e-01 -1.77932322e-01 -1.02631998e+00
-9.71603036e-01 -5.26007295e-01 -1.37871909e+00 -2.87865639e-01
-1.49378970e-01 -2.79615987e-02 5.31238377e-01 8.36767554e-01
5.30488431e-01 6.98701918e-01 8.45061362e-01 -1.10195494e+00
-2.26895154e-01 -7.33157754e-01 -4.45085019e-01 2.29028925e-01
5.68235397e-01 -3.69616449e-01 -4.15285230e-01 3.76572341e-01] | [11.080592155456543, -1.8377848863601685] |
a30d14d3-ce22-4efe-b4fb-5f8828dd3284 | conformal-loss-controlling-prediction | 2301.02424 | null | https://arxiv.org/abs/2301.02424v1 | https://arxiv.org/pdf/2301.02424v1.pdf | Conformal Loss-Controlling Prediction | Conformal prediction is a learning framework controlling prediction coverage of prediction sets, which can be built on any learning algorithm for point prediction. This work proposes a learning framework named conformal loss-controlling prediction, which extends conformal prediction to the situation where the value of a loss function needs to be controlled. Different from existing works about risk-controlling prediction sets and conformal risk control with the purpose of controlling the expected values of loss functions, the proposed approach in this paper focuses on the loss for any test object, which is an extension of conformal prediction from miscoverage loss to some general loss. The controlling guarantee is proved under the assumption of exchangeability of data in finite-sample cases and the framework is tested empirically for classification with a class-varying loss and statistical postprocessing of numerical weather forecasting applications, which are introduced as point-wise classification and point-wise regression problems. All theoretical analysis and experimental results confirm the effectiveness of our loss-controlling approach. | ['Hongyue Li', 'Xiaojun Yang', 'Zhong Ji', 'Ping Wang', 'Di Wang'] | 2023-01-06 | null | null | null | null | ['weather-forecasting'] | ['miscellaneous'] | [ 2.93698549e-01 5.16590118e-01 -2.78014779e-01 -6.81270063e-01
-9.11237597e-01 -2.79317468e-01 4.80277658e-01 5.67990005e-01
-2.19492719e-01 7.36589313e-01 -1.12956561e-01 -4.39363182e-01
-9.58462775e-01 -1.21922040e+00 -8.26039016e-01 -8.17467809e-01
-6.15289211e-01 6.03756726e-01 4.89563137e-01 -8.65506157e-02
5.46426892e-01 5.99502861e-01 -1.37318981e+00 6.53324500e-02
1.00155187e+00 1.17741776e+00 -2.74221689e-01 4.85311180e-01
2.29470849e-01 7.06587315e-01 -1.04466535e-01 -7.87593782e-01
4.49814290e-01 -2.17611298e-01 -4.50983316e-01 -5.88595271e-01
3.68404865e-01 1.65481210e-01 3.00861686e-01 8.50321233e-01
4.59672511e-01 5.92430197e-02 1.24669671e+00 -1.66010737e+00
-9.13380235e-02 3.64542753e-01 -2.28197545e-01 1.92007609e-02
-1.32645532e-01 -6.35140240e-01 1.00953734e+00 -7.01873362e-01
4.10828833e-03 9.12144899e-01 1.28077853e+00 2.84962386e-01
-1.05552542e+00 -6.71757936e-01 -9.23282802e-02 1.06823258e-01
-1.21339822e+00 2.93663383e-01 5.70123851e-01 -8.87891054e-01
3.07066470e-01 5.14616489e-01 4.39848036e-01 3.95375609e-01
9.13761497e-01 4.02159601e-01 8.50095093e-01 -6.70192182e-01
4.03875560e-01 4.31065619e-01 1.06018506e-01 3.23212981e-01
2.43259534e-01 7.89116442e-01 -3.42353899e-03 -3.35203260e-01
3.36673260e-01 1.32080719e-01 -2.79893875e-01 -8.23111117e-01
-7.24721193e-01 1.16998374e+00 3.14511806e-01 -1.59306332e-01
3.65803503e-02 -1.32816792e-01 3.42486560e-01 5.76624811e-01
1.09486961e+00 2.04732150e-01 -6.32295728e-01 3.41199130e-01
-9.72745240e-01 3.53734314e-01 9.03920949e-01 1.22332680e+00
2.43336126e-01 -3.15424323e-01 -5.75914323e-01 4.98456240e-01
3.10908228e-01 3.99265170e-01 -4.13848683e-02 -5.02554536e-01
4.95779574e-01 4.91601914e-01 1.32686123e-01 -8.68517399e-01
-5.37419260e-01 -6.96988404e-01 -8.04961681e-01 6.61696732e-01
2.42002428e-01 -2.06553429e-01 -2.20746294e-01 1.53067493e+00
4.17777359e-01 2.34882474e-01 -1.08538218e-01 4.82198119e-01
1.20059624e-01 6.48188472e-01 2.15593398e-01 -4.93579030e-01
6.59057319e-01 -6.22490466e-01 -1.82405114e-01 6.58845782e-01
8.70227277e-01 -4.47745740e-01 1.00878477e+00 4.59670126e-01
-8.73253286e-01 -3.90152752e-01 -1.09097338e+00 3.23268980e-01
-3.23647469e-01 -3.20222884e-01 1.84829354e-01 6.75803661e-01
-8.30297053e-01 8.45975220e-01 -4.14854348e-01 1.58604551e-02
5.82983911e-01 1.70410514e-01 -1.35566011e-01 5.31216264e-02
-1.20847678e+00 9.49076593e-01 1.12048328e-01 -4.14316773e-01
-5.93780100e-01 -1.68560207e+00 -5.46552479e-01 3.62918258e-01
-7.09342286e-02 -4.83906925e-01 8.56871843e-01 -7.93110728e-01
-9.80399370e-01 4.62176025e-01 3.52538675e-01 -7.27498829e-01
1.23622048e+00 -2.35926539e-01 -2.29554221e-01 -2.99069941e-01
1.04289122e-01 1.62965402e-01 7.88840175e-01 -1.04701507e+00
-7.64631867e-01 -3.31164777e-01 -1.36193290e-01 5.42675033e-02
-2.79818654e-01 -2.28560716e-01 2.75847942e-01 -8.58603716e-01
-1.80353910e-01 -5.82741857e-01 -4.03975070e-01 2.59874016e-01
-1.87788025e-01 -2.89423168e-01 6.37786150e-01 -3.81969661e-01
1.12437856e+00 -2.16008592e+00 8.44460353e-02 5.30585170e-01
-2.76800603e-01 -3.58805150e-01 1.38338998e-01 5.47747910e-01
-3.34625989e-01 2.00627282e-01 -7.62688875e-01 -1.87245041e-01
8.23705345e-02 -1.35272279e-01 -5.84452808e-01 7.47046709e-01
7.10802525e-02 3.19616616e-01 -4.99112546e-01 -5.98045170e-01
1.41925588e-01 5.14536321e-01 -7.57091641e-01 1.37301415e-01
-1.55584067e-01 1.67156726e-01 -4.89771068e-01 3.36868376e-01
7.89186835e-01 2.75128782e-01 -5.10896623e-01 3.83914918e-01
-1.27721533e-01 -2.29289204e-01 -9.28367078e-01 8.95637929e-01
-5.28086424e-01 2.01727703e-01 -4.49073642e-01 -1.18166661e+00
1.25669777e+00 1.20886855e-01 6.73795521e-01 -2.63424844e-01
-4.56133895e-02 -3.66706178e-02 -3.78379792e-01 -4.27734591e-02
-1.19218323e-02 -8.47646773e-01 -1.94390029e-01 1.46431953e-01
-2.22407192e-01 -2.83823252e-01 -4.29592788e-01 -1.76929519e-01
1.04153061e+00 9.39747766e-02 2.03300357e-01 -7.48538852e-01
6.86458886e-01 3.93307135e-02 5.65468550e-01 7.26697683e-01
3.25321183e-02 7.21332729e-01 7.91233242e-01 -3.71109247e-01
-1.16901314e+00 -1.11268210e+00 -1.13534188e+00 9.20206487e-01
5.79915904e-02 -1.64477542e-01 -6.06189191e-01 -9.35332477e-01
3.69723886e-01 1.25633299e+00 -8.60652268e-01 -2.78931558e-01
-5.82267046e-02 -9.29452181e-01 1.63406208e-01 4.30837929e-01
1.80894151e-01 -6.54588819e-01 -3.19183797e-01 -5.91407716e-02
4.83935595e-01 -3.25127482e-01 -1.90371782e-01 2.45519504e-01
-1.00566566e+00 -1.29287624e+00 -5.21851301e-01 -6.08595014e-01
5.17068446e-01 -6.01123929e-01 1.21232820e+00 -2.13671938e-01
1.28936972e-02 6.19058311e-01 -3.01602751e-01 -9.92104292e-01
-2.36822352e-01 7.53357112e-02 -1.44779637e-01 5.97527064e-02
5.55011965e-02 -5.33800840e-01 -5.05402386e-01 5.68851531e-01
-6.85313165e-01 -3.68282467e-01 2.19980747e-01 9.43828642e-01
8.62885475e-01 4.98143017e-01 8.20392966e-01 -9.90742505e-01
3.61895204e-01 -7.76020646e-01 -1.03993022e+00 5.47080636e-01
-1.03507304e+00 -4.42226715e-02 4.62807894e-01 -2.23574013e-01
-8.01304638e-01 -7.41897151e-02 5.76099157e-02 -2.79166222e-01
2.20506206e-01 2.19136372e-01 -3.25955331e-01 -2.60577202e-01
5.24355352e-01 -2.40709577e-02 -1.05678819e-01 -2.84754872e-01
-9.87855718e-02 2.26346508e-01 6.68356288e-03 -5.34615099e-01
8.87257934e-01 2.00820535e-01 7.87572026e-01 -3.58004451e-01
-8.61806750e-01 -1.00619249e-01 -8.20356667e-01 -3.64806592e-01
7.06743419e-01 -5.51851988e-01 -6.63281679e-01 1.82204559e-01
-6.91959321e-01 -3.08706939e-01 -7.14232624e-01 5.89229047e-01
-1.07078433e+00 -7.13570565e-02 -3.75942178e-02 -9.27775264e-01
-3.15050870e-01 -7.37069726e-01 9.79802966e-01 -3.40680480e-01
2.19518244e-01 -1.52278495e+00 4.68961418e-01 -1.82781756e-01
2.67036200e-01 5.19636154e-01 1.27324259e+00 -8.40468884e-01
-3.10844064e-01 -8.20697844e-01 1.88356325e-01 5.66340446e-01
-4.11607832e-01 -1.76237047e-01 -7.75078595e-01 -2.44411021e-01
2.48134434e-01 -7.04620630e-02 6.23177528e-01 5.94524682e-01
1.62096560e+00 -4.42164540e-01 -4.42426711e-01 6.41927719e-01
1.65350580e+00 1.28396213e-01 8.31355691e-01 3.16038549e-01
3.23029250e-01 9.67928529e-01 1.05344307e+00 5.05800664e-01
9.32967011e-03 6.83917582e-01 5.71278512e-01 8.89493376e-02
5.60473621e-01 -3.58201832e-01 4.23892438e-02 1.97786063e-01
1.07974425e-01 -2.03262735e-02 -8.61528277e-01 2.41634816e-01
-1.65475726e+00 -1.05540073e+00 -2.60058194e-01 2.69916129e+00
5.36217809e-01 9.55170691e-02 -6.64623082e-02 3.68252724e-01
7.84701526e-01 -2.25946411e-01 -4.76981401e-01 -5.13751149e-01
1.96518049e-01 8.96537453e-02 7.21701264e-01 8.66028965e-01
-1.20545673e+00 2.22793281e-01 6.46373224e+00 1.02715528e+00
-8.46123934e-01 2.25437269e-01 9.42375660e-01 -6.42432868e-02
-4.99201477e-01 -3.29727530e-02 -9.84603405e-01 6.77246153e-01
8.46010387e-01 -1.53831273e-01 -2.32675329e-01 1.16476023e+00
2.44813263e-02 2.77374715e-01 -1.22596943e+00 3.35939318e-01
-3.54725644e-02 -1.07126081e+00 -9.91690084e-02 3.57161611e-02
5.70723116e-01 -9.83475521e-02 2.28145376e-01 3.74638766e-01
5.74690364e-02 -1.07250619e+00 5.49011409e-01 1.18414855e+00
8.61776710e-01 -1.28140903e+00 1.00354981e+00 5.90067506e-01
-7.09848344e-01 -2.70138204e-01 -5.40586233e-01 5.84666617e-03
3.29888128e-02 1.00114262e+00 -5.22894561e-01 4.42268759e-01
8.20299864e-01 7.89312899e-01 -4.18365002e-01 1.34468186e+00
1.08428665e-01 9.01216507e-01 -2.60395348e-01 -6.52749613e-02
-1.27750531e-01 -2.90458649e-01 5.94057262e-01 8.90163064e-01
6.52094364e-01 9.22750160e-02 -1.16229929e-01 4.73933488e-01
4.44061667e-01 5.53949058e-01 -8.94957066e-01 9.68719959e-01
2.85404116e-01 9.09427345e-01 -2.40382791e-01 1.95358112e-01
-2.87677109e-01 3.11427861e-01 2.07867205e-01 -1.44498155e-01
-1.01210356e+00 -4.41338599e-01 2.48089597e-01 7.46488571e-01
3.33172113e-01 2.85943121e-01 -7.74714589e-01 -6.55065477e-01
1.35277361e-01 1.05936071e-02 6.15338027e-01 -4.82003301e-01
-1.73339117e+00 3.30463946e-01 5.36929071e-01 -1.83780098e+00
8.42790380e-02 -6.30960763e-01 -1.01510036e+00 7.44750500e-01
-1.42533660e+00 -1.06290138e+00 6.44495338e-02 5.07136941e-01
1.33610919e-01 -5.46380162e-01 7.06670284e-01 1.44292131e-01
5.00784069e-02 9.15213406e-01 3.61050010e-01 -4.15154189e-01
5.13211548e-01 -1.32246709e+00 -3.65672916e-01 4.66309398e-01
-4.19490218e-01 -1.93825468e-01 5.80100060e-01 -6.02609813e-01
-5.13480604e-01 -1.56733763e+00 1.20704126e+00 -7.75605798e-01
5.64077616e-01 -3.23131949e-01 -8.62919390e-01 4.80937719e-01
-2.76394963e-01 2.63592720e-01 9.34705079e-01 2.22602770e-01
-8.40052068e-02 -5.92591882e-01 -1.71179712e+00 1.11369245e-01
9.50620174e-01 6.98948950e-02 -4.30693656e-01 5.95037937e-01
7.42952526e-01 -1.25625849e-01 -1.36068451e+00 1.08085740e+00
5.01016021e-01 -9.43584859e-01 9.39792275e-01 -7.82552600e-01
6.60774708e-01 1.05047906e-02 -3.72565329e-01 -1.12927663e+00
-3.30244839e-01 -1.09683029e-01 3.82445931e-01 1.25701404e+00
5.46365619e-01 -7.59921610e-01 6.24450803e-01 5.61879635e-01
-3.25118303e-01 -1.07020295e+00 -1.30289936e+00 -7.91622043e-01
8.77662241e-01 -7.09793389e-01 5.85750639e-01 8.45523238e-01
-1.99094750e-02 -8.79491270e-02 -3.22025895e-01 4.20445979e-01
9.55748975e-01 -6.53323531e-02 4.15670097e-01 -1.67852104e+00
-4.17134106e-01 -2.08507806e-01 -8.28224123e-01 -1.90249026e-01
3.52432013e-01 -9.63322580e-01 -1.48064107e-01 -1.03828657e+00
-3.18932347e-02 -1.20834982e+00 -3.16278994e-01 2.02201441e-01
1.25075892e-01 -7.06899613e-02 -5.20629622e-02 6.69608787e-02
-9.89764556e-02 8.59406292e-01 1.14618814e+00 -9.22037289e-02
7.01645948e-03 8.73311818e-01 -4.93745506e-01 6.19355023e-01
7.74491906e-01 -6.42472208e-01 -9.31608319e-01 2.67541349e-01
3.77328098e-01 2.65763462e-01 5.93229771e-01 -1.18418181e+00
7.32828826e-02 -2.30372861e-01 4.42052096e-01 -6.39451385e-01
-6.98071197e-02 -1.26854789e+00 1.11681297e-01 5.80916286e-01
-7.40434110e-01 -1.15529131e-02 1.15751244e-01 8.56663108e-01
-1.65762052e-01 -4.91611719e-01 1.01823366e+00 5.84397376e-01
-4.58233878e-02 5.64181924e-01 1.20906666e-01 3.91343199e-02
1.63965428e+00 1.17197476e-01 -1.19264303e-02 -4.25823897e-01
-8.28296125e-01 2.93109685e-01 1.20213360e-01 1.75325617e-01
6.18975699e-01 -1.46737158e+00 -9.89599764e-01 2.83947468e-01
3.05405676e-01 7.19960779e-02 2.54702300e-01 7.38711774e-01
-3.68902415e-01 2.13868693e-01 9.75203067e-02 -6.47269368e-01
-9.25147951e-01 7.21550703e-01 6.27230048e-01 -2.36268252e-01
-6.67031586e-01 7.54634440e-01 3.69047821e-01 -7.10154057e-01
3.86210859e-01 -2.50050515e-01 -3.68446738e-01 -4.33508866e-02
4.62123275e-01 5.83520353e-01 2.16141224e-01 -2.31569514e-01
-1.16748929e-01 7.96033442e-01 1.49864450e-01 5.64188361e-02
1.35223675e+00 1.98023498e-01 -2.56148707e-02 6.87097490e-01
1.09846807e+00 -4.68585119e-02 -1.42852461e+00 1.77914321e-01
-8.33934844e-02 -4.18737501e-01 -2.27594823e-02 -9.10323620e-01
-8.60876381e-01 8.57947588e-01 9.20823276e-01 1.39292747e-01
1.11605942e+00 5.44476993e-02 -1.07787095e-03 -2.39030290e-02
5.42706072e-01 -8.51399660e-01 -4.88454551e-01 3.42868239e-01
1.27675271e+00 -1.24395061e+00 -2.14001462e-02 -4.79238421e-01
-4.66530293e-01 9.33788002e-01 4.65250671e-01 -4.43402737e-01
1.75522864e+00 4.51883733e-01 -4.75289404e-01 1.96312636e-01
-7.93537498e-01 4.90867853e-01 4.90177810e-01 7.04399288e-01
2.41658419e-01 1.61935329e-01 -6.35897279e-01 7.74310946e-01
-3.79944742e-01 -4.16999310e-02 4.39053923e-02 5.00016928e-01
-4.36792791e-01 -8.75849843e-01 -2.25881189e-01 6.85817361e-01
-3.66145492e-01 -1.36670604e-01 1.05876893e-01 1.07450330e+00
2.24537015e-01 5.17424583e-01 4.06037956e-01 -3.47851485e-01
4.89881247e-01 6.33798614e-02 1.92750916e-01 -3.47735584e-01
-3.47107291e-01 -5.21787226e-01 -2.89472282e-01 -2.28425875e-01
-1.89833909e-01 -9.24928308e-01 -9.49673116e-01 -3.04218620e-01
-2.60847002e-01 5.18583953e-01 4.53727961e-01 7.86872149e-01
-4.09845039e-02 2.88751215e-01 1.07731712e+00 -4.54473168e-01
-9.22376990e-01 -8.44195545e-01 -9.64079440e-01 7.40262866e-02
2.55089551e-01 -8.71263206e-01 -7.64394403e-01 -2.93050885e-01] | [7.978555679321289, 4.28490686416626] |
29576732-6f3c-4bb9-ab38-64b45d5d1aa1 | leftright-hand-segmentation-in-egocentric | 1607.06264 | null | http://arxiv.org/abs/1607.06264v1 | http://arxiv.org/pdf/1607.06264v1.pdf | Left/Right Hand Segmentation in Egocentric Videos | Wearable cameras allow people to record their daily activities from a
user-centered (First Person Vision) perspective. Due to their favorable
location, wearable cameras frequently capture the hands of the user, and may
thus represent a promising user-machine interaction tool for different
applications. Existent First Person Vision methods handle hand segmentation as
a background-foreground problem, ignoring two important facts: i) hands are not
a single "skin-like" moving element, but a pair of interacting cooperative
entities, ii) close hand interactions may lead to hand-to-hand occlusions and,
as a consequence, create a single hand-like segment. These facts complicate a
proper understanding of hand movements and interactions. Our approach extends
traditional background-foreground strategies, by including a
hand-identification step (left-right) based on a Maxwell distribution of angle
and position. Hand-to-hand occlusions are addressed by exploiting temporal
superpixels. The experimental results show that, in addition to a reliable
left/right hand-segmentation, our approach considerably improves the
traditional background-foreground hand-segmentation. | ['Emilia Barakova', 'Alejandro Betancourt', 'Lucio Marcenaro', 'Matthias Rauterberg', 'Carlo Regazzoni', 'Pietro Morerio'] | 2016-07-21 | null | null | null | null | ['hand-segmentation'] | ['computer-vision'] | [ 3.02424014e-01 -3.51246715e-01 -3.40916693e-01 -1.71548892e-02
-1.71896160e-01 -7.50743568e-01 5.40322006e-01 -6.51232898e-02
-4.73919660e-01 7.32686043e-01 -6.59486577e-02 -1.97473206e-02
1.13909267e-01 -3.25774491e-01 -4.90194350e-01 -1.00933516e+00
4.34048355e-01 5.04055023e-01 7.54947186e-01 1.64062411e-01
-5.65204881e-02 6.92588031e-01 -1.70589936e+00 9.76054668e-02
7.94255435e-01 5.32266676e-01 1.40915230e-01 1.00578201e+00
-7.65669495e-02 7.54768848e-01 -7.73743868e-01 -4.83401656e-01
1.02118336e-01 -4.72425967e-01 -6.14128470e-01 3.89080465e-01
5.44867635e-01 -4.85742986e-01 7.23164454e-02 1.07108998e+00
5.34285784e-01 2.01983243e-01 4.08409059e-01 -1.18181109e+00
1.00720748e-01 4.78087813e-02 -1.11163342e+00 1.97650343e-01
7.22740412e-01 1.45515054e-01 6.06727719e-01 -6.12804115e-01
7.23911524e-01 1.08290529e+00 5.90401947e-01 6.44912302e-01
-1.29785073e+00 -2.47107416e-01 7.14185119e-01 1.72581032e-01
-1.48579538e+00 -3.35645765e-01 8.85925055e-01 -7.41182506e-01
5.10830760e-01 8.08815837e-01 1.11395037e+00 1.31888330e+00
-1.27203241e-01 1.02632046e+00 1.01313758e+00 -6.38479114e-01
2.02691242e-01 2.96136677e-01 5.09013832e-01 4.23375100e-01
6.56691849e-01 -3.58707905e-01 -3.31267774e-01 -1.07246280e-01
1.09343731e+00 2.15043411e-01 -5.78321815e-01 -4.84527856e-01
-1.28526151e+00 7.37669021e-02 -4.27127210e-03 5.58998525e-01
-3.28471273e-01 1.91763714e-01 1.74310178e-01 -4.69095737e-01
2.81030536e-01 -3.83324236e-01 -2.21932828e-01 -2.10181326e-01
-1.24467695e+00 4.70133573e-01 1.04804802e+00 9.98399794e-01
3.59107167e-01 -3.73752922e-01 -4.20366645e-01 2.76725620e-01
4.31195349e-01 6.41728342e-01 -5.75218722e-03 -6.88178778e-01
4.73798841e-01 6.27904058e-01 6.86562479e-01 -1.01679182e+00
-3.51504117e-01 -1.87740520e-01 -7.36297905e-01 4.16820198e-01
1.12695944e+00 -1.62691444e-01 -6.16047382e-01 1.22479475e+00
7.32511938e-01 -3.98756228e-02 -5.64151406e-01 1.06601906e+00
5.40616870e-01 1.75626531e-01 2.58155167e-01 -5.57261527e-01
1.73819613e+00 -9.47137475e-01 -1.13353276e+00 -2.90358335e-01
9.81213450e-02 -7.64224529e-01 8.37711573e-01 7.23766208e-01
-1.15423346e+00 -5.65872431e-01 -6.01293743e-01 2.58177356e-03
-3.47306848e-01 3.65670264e-01 5.25738478e-01 1.11836684e+00
-7.84227610e-01 4.05905306e-01 -8.19367230e-01 -5.66468418e-01
2.64047027e-01 3.76965702e-01 -6.42709807e-02 1.80192679e-01
-4.60498363e-01 7.15939164e-01 9.30881947e-02 3.97226751e-01
-2.94747323e-01 -3.74669969e-01 -2.76751518e-01 -1.69495597e-01
7.30475485e-01 -7.10579157e-01 1.16580009e+00 -1.13719916e+00
-1.36480570e+00 9.85101819e-01 -5.94935715e-01 1.31289586e-01
1.30035412e+00 -5.15173674e-01 -2.70997971e-01 3.19157034e-01
-1.51621759e-01 7.32055679e-02 9.43233371e-01 -1.80791080e+00
-6.94878101e-01 -5.87598741e-01 -3.23945098e-02 2.56792694e-01
3.49525288e-02 3.07745248e-01 -9.99639452e-01 -4.76884961e-01
2.63449922e-02 -8.53689611e-01 -8.62527713e-02 1.11677855e-01
-6.26986146e-01 -1.88420430e-01 8.47195804e-01 -9.82052207e-01
1.54466152e+00 -1.86512220e+00 2.18962148e-01 2.55159110e-01
5.46978056e-01 5.13010263e-01 5.17049849e-01 5.95647320e-02
8.07433799e-02 -1.96577236e-01 9.07341689e-02 -6.34340227e-01
-1.43550292e-01 1.34917021e-01 2.25254759e-01 7.09603429e-01
-5.15428185e-01 7.21977711e-01 -1.04271770e+00 -8.58364701e-01
6.31966710e-01 7.03897417e-01 5.72913811e-02 1.01559632e-01
-1.32213071e-01 7.69567370e-01 -3.21239591e-01 9.03207004e-01
8.62575054e-01 -8.83057117e-02 3.82378012e-01 -2.24128053e-01
-2.73975134e-01 -4.76454556e-01 -1.59618843e+00 1.33448923e+00
5.21014035e-02 6.01434886e-01 4.34354454e-01 -4.90533739e-01
3.30768406e-01 4.20291513e-01 4.53502297e-01 -2.22150564e-01
3.64938706e-01 8.99088010e-03 -2.57402748e-01 -5.73042452e-01
2.16668114e-01 3.97117406e-01 4.14489776e-01 2.43829191e-01
-2.20554516e-01 3.37665349e-01 3.91077667e-01 4.43236018e-03
5.42552233e-01 4.93505001e-01 3.47700685e-01 -1.47076264e-01
6.19508684e-01 -1.81946144e-01 4.03263062e-01 8.86492372e-01
-4.51512516e-01 7.47515321e-01 3.14877748e-01 -4.45749819e-01
-5.11854947e-01 -8.77597451e-01 1.66406721e-01 1.05305409e+00
5.04698873e-01 -1.33062273e-01 -1.35648966e+00 -5.59691131e-01
-2.70073593e-01 4.51025993e-01 -4.94897336e-01 5.56364059e-01
-7.38638937e-01 -6.36303663e-01 1.74761713e-01 7.57979929e-01
3.71244192e-01 -7.87478328e-01 -1.04427767e+00 2.16168284e-01
-3.63730609e-01 -1.13818252e+00 -7.46769190e-01 -1.68625638e-02
-7.87685871e-01 -1.35849881e+00 -1.45051312e+00 -4.77071017e-01
7.19787061e-01 6.14783943e-01 1.03734612e+00 4.00583446e-02
-5.84267914e-01 8.66644204e-01 -9.90792736e-02 -2.92526811e-01
2.46132873e-02 -3.45950842e-01 1.95127726e-01 5.07154047e-01
4.26009566e-01 -6.44487798e-01 -9.51113820e-01 5.22731841e-01
-2.85825014e-01 6.19602017e-02 2.71080136e-01 2.46839762e-01
4.17913526e-01 -4.64322045e-02 -3.59491378e-01 -6.09626472e-01
3.87223214e-01 1.36482626e-01 -7.79825866e-01 7.04757035e-01
-7.42197111e-02 -5.62809944e-01 2.04390481e-01 -7.71900415e-01
-1.38642287e+00 5.14707327e-01 3.09563726e-01 -4.92141098e-01
-7.27807641e-01 -3.70948374e-01 -7.00460434e-01 -3.92038189e-02
5.16120076e-01 1.39757052e-01 -4.45300460e-01 -6.13371432e-01
5.41341484e-01 5.66587508e-01 7.15416610e-01 -5.48222065e-01
5.09528995e-01 8.38428795e-01 -1.86254978e-01 -1.31384432e+00
-5.11940300e-01 -9.52058852e-01 -1.43264186e+00 -8.21715593e-01
1.23957109e+00 -6.12649918e-01 -1.11202669e+00 7.46462047e-01
-1.61033726e+00 -1.49192289e-01 -2.73425937e-01 3.47283095e-01
-2.27254525e-01 7.58172631e-01 -3.37970704e-01 -1.67437017e+00
-1.29480720e-01 -8.86529624e-01 1.17703950e+00 5.12355089e-01
-5.39946437e-01 -1.05135548e+00 -1.67084679e-01 6.03021741e-01
-1.22409523e-01 3.99242222e-01 2.76059180e-01 -2.76834905e-01
-6.47444487e-01 -4.08657402e-01 -1.61333695e-01 7.10318834e-02
1.07820548e-01 2.15362608e-01 -1.24444664e+00 -6.64620697e-02
-1.23781860e-01 6.43091440e-01 4.65009630e-01 5.99627972e-01
7.49533713e-01 -2.17528298e-01 -7.80728877e-01 2.75803566e-01
1.05484307e+00 4.05831188e-01 5.70985436e-01 2.64460444e-02
1.12941456e+00 6.70320570e-01 2.75999308e-01 5.10389268e-01
2.62070775e-01 8.84487450e-01 2.44716212e-01 -3.01307112e-01
-3.32706332e-01 9.14626047e-02 1.55063212e-01 3.77365530e-01
-1.06909895e+00 -2.35940441e-01 -7.56680131e-01 4.57052320e-01
-2.06570506e+00 -1.13873756e+00 -8.60036194e-01 2.39122248e+00
5.34678757e-01 -6.49124943e-03 7.78285921e-01 4.06541348e-01
1.04568481e+00 -4.84636985e-02 -3.80884469e-01 2.61217862e-01
-5.87522015e-02 -1.61437884e-01 5.62378824e-01 4.08266753e-01
-1.14523864e+00 6.85934365e-01 6.00041103e+00 6.26069486e-01
-8.40894163e-01 3.13485861e-01 2.90969610e-01 -1.87680572e-01
7.39089921e-02 -1.65915400e-01 -1.19573426e+00 5.32864571e-01
4.17802185e-02 3.53500515e-01 3.69264781e-01 6.49168253e-01
3.49544913e-01 -7.53019392e-01 -1.14436364e+00 1.41687357e+00
7.41775557e-02 -7.42002547e-01 -1.89336374e-01 2.38592237e-01
4.35910255e-01 -8.68490815e-01 -3.24210703e-01 -4.67989504e-01
-9.05749723e-02 -6.19997203e-01 1.09369540e+00 7.86189914e-01
5.76676905e-01 -4.46909070e-01 3.73313457e-01 5.03871262e-01
-1.40345347e+00 1.09799184e-01 2.27282286e-01 -4.27311286e-02
5.03309071e-01 7.44277477e-01 -2.97228843e-01 4.58868593e-01
7.89089799e-01 2.34436810e-01 -3.99527907e-01 9.44071352e-01
-2.75281638e-01 2.67026752e-01 -3.42278868e-01 -1.41749501e-01
-3.21610957e-01 -4.03173387e-01 7.92004228e-01 1.45373988e+00
-1.66760191e-01 9.40061808e-02 2.29472935e-01 1.00730765e+00
4.68481719e-01 -5.23060784e-02 -9.19364095e-02 9.42827463e-02
8.78679827e-02 1.23672879e+00 -1.43190384e+00 -4.34600055e-01
-4.95683521e-01 1.56994903e+00 -3.77449214e-01 6.54998481e-01
-8.42664540e-01 -2.85440415e-01 4.43247795e-01 5.24901688e-01
2.04363406e-01 -2.91929215e-01 -3.40335190e-01 -1.31150186e+00
5.04801631e-01 -6.00992620e-01 2.21251659e-02 -6.18336022e-01
-8.99220765e-01 3.72101247e-01 2.15916798e-01 -1.06982219e+00
-8.53538979e-03 -6.36585891e-01 -5.96050203e-01 7.95565248e-01
-9.08982992e-01 -1.51314843e+00 -8.51857126e-01 8.65226388e-01
5.93075454e-01 2.45405465e-01 5.85841894e-01 4.60638285e-01
-8.04633617e-01 3.60219002e-01 -2.13025928e-01 1.45248830e-01
4.13804293e-01 -1.32086897e+00 1.04962714e-01 1.08020580e+00
8.89186785e-02 6.62208617e-01 8.14831913e-01 -9.14572716e-01
-1.29251993e+00 -5.45581579e-01 8.04744184e-01 -8.38649929e-01
2.70664066e-01 -6.11167669e-01 -6.22681797e-01 6.91925049e-01
1.13653488e-01 9.17772390e-03 5.48438668e-01 1.60885662e-01
1.37280494e-01 -3.61406319e-02 -9.54002857e-01 7.68379569e-01
1.28219187e+00 -2.93755859e-01 -4.34179395e-01 3.57579142e-01
4.34508035e-03 -5.28217256e-01 -2.74410218e-01 -1.50304049e-01
9.68454003e-01 -1.23164380e+00 1.22847080e+00 -1.49517402e-01
-4.13459092e-01 -5.61348855e-01 2.48949215e-01 -4.90072906e-01
-1.28338531e-01 -9.92539585e-01 -6.85242832e-01 1.38790548e+00
-2.82879055e-01 -3.68427306e-01 8.30606043e-01 1.03670633e+00
5.51381946e-01 -3.25564981e-01 -8.74216974e-01 -8.69023204e-01
-5.50421894e-01 -3.65323186e-01 2.43852288e-01 7.05364823e-01
-8.21063742e-02 7.92680010e-02 -4.22674745e-01 1.39058009e-01
1.03070736e+00 -1.16548367e-01 9.60560441e-01 -1.70768273e+00
-1.88988939e-01 -5.94471335e-01 -1.37018576e-01 -1.33720827e+00
-3.63838971e-01 -1.05888946e-02 2.89797280e-02 -1.39724004e+00
3.78613979e-01 -7.73791298e-02 1.28020048e-01 8.90666023e-02
-2.88782060e-01 1.55665934e-01 2.38332585e-01 1.25781894e-01
-7.17642665e-01 -3.28737020e-01 1.10325408e+00 5.19202836e-02
-5.22925615e-01 4.89723891e-01 -2.71660537e-01 1.20749593e+00
3.43455344e-01 -9.14074630e-02 -1.31428868e-01 -6.78029805e-02
1.07998714e-01 -4.03171359e-03 7.28205204e-01 -9.98157382e-01
5.33713341e-01 -2.68740386e-01 6.39640808e-01 -8.83677185e-01
4.34923589e-01 -1.08675563e+00 3.40706021e-01 3.68285090e-01
3.38215113e-01 -3.05640101e-01 2.27329042e-03 7.40239322e-01
2.45329648e-01 -3.66653860e-01 6.14609301e-01 -5.29836118e-01
-5.56092143e-01 -1.81262895e-01 -6.56317890e-01 -2.95700222e-01
1.39185727e+00 -9.04273152e-01 8.37332085e-02 -2.82246441e-01
-1.12248230e+00 -5.48659228e-02 3.78269762e-01 1.08908787e-01
1.74618870e-01 -8.63612533e-01 -1.12401649e-01 6.77076355e-02
-3.29484254e-01 1.70074522e-01 3.62472147e-01 1.14555669e+00
-5.75875163e-01 3.31861228e-01 8.39049667e-02 -8.33664834e-01
-1.97916293e+00 4.92779642e-01 3.98931563e-01 -8.80276561e-02
-6.36735201e-01 9.45390046e-01 1.82673067e-01 2.54405826e-01
7.96908855e-01 -4.73648340e-01 -1.77659169e-01 5.78210652e-01
8.52842152e-01 1.08699071e+00 -1.23583227e-01 -7.89687753e-01
-5.48906863e-01 9.43841398e-01 2.33597472e-01 -5.67649789e-02
7.58450210e-01 -4.42322403e-01 -1.87412091e-02 5.55802345e-01
4.85006303e-01 3.63956779e-01 -1.35726392e+00 -6.92669898e-02
-6.74355105e-02 -5.83688736e-01 -2.61513442e-01 -6.85639560e-01
-9.75230694e-01 1.13888812e+00 6.74696028e-01 2.60492951e-01
1.00174463e+00 -3.16538364e-02 5.31025171e-01 -2.16708556e-02
3.98491085e-01 -1.30019772e+00 -1.41633546e-03 1.15598775e-02
5.79457700e-01 -1.03200758e+00 2.72710979e-01 -9.22114491e-01
-6.14873350e-01 1.18391049e+00 5.11396408e-01 4.02115524e-01
4.29436088e-01 3.64630908e-01 3.09799872e-02 7.31884837e-02
1.96750119e-01 -5.87334037e-01 5.32374680e-01 8.15809786e-01
2.64405161e-01 1.30506054e-01 -2.45858207e-01 7.53384948e-01
3.37542534e-01 2.92412311e-01 1.84389889e-01 1.14180946e+00
-2.01509356e-01 -1.00093305e+00 -8.81275356e-01 -2.80408859e-02
-3.81915569e-01 3.70364398e-01 -5.97085655e-01 8.53913307e-01
4.92738992e-01 9.69055474e-01 4.40370068e-02 6.79993182e-02
5.10592461e-01 1.38157874e-01 9.37634528e-01 -2.91394293e-01
-5.90067089e-01 5.25169015e-01 -2.96971768e-01 -5.73485255e-01
-7.40298092e-01 -8.50666165e-01 -8.23114514e-01 -2.98311561e-01
-4.53453898e-01 -3.30854356e-01 5.02759933e-01 1.04547620e+00
-9.77728665e-02 4.88454491e-01 -4.70503084e-02 -1.62453210e+00
-1.05840474e-01 -6.15035772e-01 -8.05528402e-01 3.69826794e-01
4.30741221e-01 -7.06115484e-01 -3.25287074e-01 5.92743754e-01] | [6.6623945236206055, -0.7040334343910217] |
c6b121d1-9fcd-42e6-bc65-4e20c2d1c611 | a-theoretically-grounded-benchmark-for | 2203.12184 | null | https://arxiv.org/abs/2203.12184v2 | https://arxiv.org/pdf/2203.12184v2.pdf | A Theoretically Grounded Benchmark for Evaluating Machine Commonsense | Programming machines with commonsense reasoning (CSR) abilities is a longstanding challenge in the Artificial Intelligence community. Current CSR benchmarks use multiple-choice (and in relatively fewer cases, generative) question-answering instances to evaluate machine commonsense. Recent progress in transformer-based language representation models suggest that considerable progress has been made on existing benchmarks. However, although tens of CSR benchmarks currently exist, and are growing, it is not evident that the full suite of commonsense capabilities have been systematically evaluated. Furthermore, there are doubts about whether language models are 'fitting' to a benchmark dataset's training partition by picking up on subtle, but normatively irrelevant (at least for CSR), statistical features to achieve good performance on the testing partition. To address these challenges, we propose a benchmark called Theoretically-Grounded Commonsense Reasoning (TG-CSR) that is also based on discriminative question answering, but with questions designed to evaluate diverse aspects of commonsense, such as space, time, and world states. TG-CSR is based on a subset of commonsense categories first proposed as a viable theory of commonsense by Gordon and Hobbs. The benchmark is also designed to be few-shot (and in the future, zero-shot), with only a few training and validation examples provided. This report discusses the structure and construction of the benchmark. Preliminary results suggest that the benchmark is challenging even for advanced language representation models designed for discriminative CSR question answering tasks. Benchmark access and leaderboard: https://codalab.lisn.upsaclay.fr/competitions/3080 Benchmark website: https://usc-isi-i2.github.io/TGCSR/ | ['Mayank Kejriwal', 'Deborah L. McGuinness', 'Yasaman Razeghi', 'Alice M. Mulvehill', 'Ke Shen', 'Henrique Santos'] | 2022-03-23 | null | null | null | null | ['generative-question-answering'] | ['natural-language-processing'] | [ 2.78938681e-01 3.29402983e-01 -1.82522401e-01 -3.35514635e-01
-1.00764298e+00 -6.91009283e-01 8.75733018e-01 2.20157593e-01
-2.00174779e-01 6.65875733e-01 5.11046767e-01 -5.26596487e-01
-2.57662177e-01 -1.00024831e+00 -3.39062959e-01 -2.39641443e-01
3.92208040e-01 7.53675461e-01 1.56624354e-02 -9.05526221e-01
4.40178812e-01 6.49241805e-02 -1.47922301e+00 5.36283016e-01
1.15185177e+00 7.46930063e-01 1.95115715e-01 4.70790267e-01
-2.88564861e-01 1.64328575e+00 -7.47249544e-01 -7.34494865e-01
-7.45344982e-02 -8.12895417e-01 -1.36109626e+00 -4.93043631e-01
3.55097830e-01 -9.68886365e-04 -3.50462109e-01 1.08877695e+00
1.75695226e-01 3.68887872e-01 5.18994570e-01 -1.36785626e+00
-1.46439743e+00 9.90634263e-01 3.09631944e-01 5.04580021e-01
7.34926105e-01 2.36121535e-01 1.59564126e+00 -5.35744727e-01
8.03347170e-01 1.19763470e+00 5.80213845e-01 7.16679990e-01
-1.22141361e+00 -5.00311077e-01 -2.00325206e-01 7.74764121e-01
-1.09660780e+00 -3.63840699e-01 1.06031525e+00 -2.67416686e-01
1.35437727e+00 7.11802125e-01 6.32488966e-01 1.36648452e+00
5.41351438e-02 7.60416985e-01 1.55698180e+00 -4.76389557e-01
3.43855649e-01 1.24397255e-01 6.12549067e-01 5.61115444e-01
3.01384270e-01 -6.98919455e-03 -3.71385425e-01 -3.48132461e-01
3.97850573e-01 -3.40946794e-01 -2.02784047e-01 -1.72974661e-01
-1.31194234e+00 1.11607718e+00 4.47149366e-01 8.53951573e-01
-1.55772015e-01 2.38617510e-01 6.86927974e-01 5.99577606e-01
2.94529051e-01 8.61904085e-01 -1.97690710e-01 -4.32625979e-01
-9.82352376e-01 6.97804689e-01 9.32855010e-01 7.83083975e-01
4.32794392e-01 1.59541398e-01 -2.44940504e-01 1.00206447e+00
-6.06733523e-02 3.80123228e-01 7.96207488e-01 -1.17948210e+00
4.61750835e-01 6.67208731e-01 -2.47627854e-01 -1.25550234e+00
-3.70976090e-01 -1.52465716e-01 -5.26116550e-01 -2.54769593e-01
1.47891983e-01 4.82487410e-01 -5.68609238e-01 2.09997797e+00
-2.41858616e-01 -2.28378236e-01 2.71128297e-01 8.32579136e-01
1.16393685e+00 5.42276323e-01 1.89532578e-01 2.93937046e-02
1.26785660e+00 -5.35211265e-01 -5.99730015e-01 -4.78447586e-01
8.54960740e-01 -3.03424001e-01 1.65722692e+00 2.94905543e-01
-1.08614731e+00 -2.29836196e-01 -7.96366334e-01 -6.77693069e-01
-8.10443819e-01 -3.60154212e-01 1.00619268e+00 6.56161904e-01
-1.03064919e+00 3.08261484e-01 -2.74697781e-01 -4.75664645e-01
2.09387079e-01 -3.03418785e-01 -1.50294945e-01 -2.44305596e-01
-1.74685788e+00 1.51550698e+00 5.83047450e-01 -2.20271081e-01
-6.87727153e-01 -4.90480363e-01 -1.04528773e+00 -1.07597858e-01
2.50854462e-01 -6.01643682e-01 1.03984678e+00 -8.59886825e-01
-1.19525599e+00 1.50496709e+00 5.19630760e-02 -6.11592054e-01
3.57479393e-01 1.37898162e-01 -6.39870703e-01 6.36011586e-02
5.45559466e-01 3.80963594e-01 1.97707027e-01 -1.00388694e+00
-1.08113401e-01 -2.59600669e-01 6.49134934e-01 1.06293887e-01
-9.60127171e-03 2.89273530e-01 3.89603198e-01 -7.10046828e-01
1.07613049e-01 -7.59800255e-01 2.99682051e-01 -3.69613677e-01
-4.36106652e-01 -6.83110118e-01 3.27166349e-01 -7.41950929e-01
1.18816280e+00 -2.00539780e+00 -4.06622840e-03 -1.87362805e-01
9.87940431e-02 -1.74743444e-01 -2.17414483e-01 4.48934942e-01
-3.64129156e-01 3.48379582e-01 -5.00309110e-01 -1.12567237e-02
5.92264950e-01 3.82925421e-01 -8.45134079e-01 2.56318927e-01
1.19758122e-01 1.11385989e+00 -1.21205306e+00 -6.48341358e-01
1.21043265e-01 -1.09008234e-02 -6.47620797e-01 -1.68916032e-01
-4.47076023e-01 -1.45309061e-01 -1.83055803e-01 7.19458878e-01
2.14963973e-01 -2.95356125e-01 3.13716456e-02 1.23340614e-01
2.57037550e-01 9.15777624e-01 -7.24425375e-01 1.65957189e+00
-2.88065404e-01 8.33300591e-01 -5.34591913e-01 -1.22064757e+00
9.61778104e-01 1.67361230e-01 1.74619555e-02 -8.24202061e-01
8.76265019e-02 3.29237431e-01 2.94661939e-01 -4.15223062e-01
7.23041534e-01 -8.09138834e-01 -4.60523367e-01 3.43181700e-01
1.40295401e-01 -8.56962860e-01 5.04751265e-01 3.25006545e-01
1.29768741e+00 1.50559291e-01 5.60687840e-01 -3.75476241e-01
3.36748302e-01 4.90114063e-01 4.77708369e-01 6.27169788e-01
-4.94960278e-01 3.52986068e-01 5.71949720e-01 -2.78227955e-01
-9.18260574e-01 -1.29483843e+00 -3.02885383e-01 1.30023384e+00
1.17008910e-01 -5.10395110e-01 -4.14908588e-01 -3.14687103e-01
-2.59588718e-01 1.95210898e+00 -6.94380760e-01 -3.70113552e-01
-4.49242681e-01 -4.31317151e-01 1.08237112e+00 4.67009097e-01
3.83080781e-01 -1.46025527e+00 -9.37710166e-01 1.60902396e-01
-6.99883640e-01 -1.07319415e+00 1.69193417e-01 3.74796629e-01
-6.92924738e-01 -1.15652335e+00 -1.20977603e-01 -5.61092913e-01
-8.71087462e-02 1.34933367e-03 1.64739215e+00 -1.75659135e-02
-1.63438424e-01 4.80036527e-01 -4.91663158e-01 -4.06075895e-01
-6.57343745e-01 -2.77231455e-01 -2.94013083e-01 -7.68133521e-01
6.93009198e-01 -5.87156177e-01 -1.02294996e-01 -1.72623396e-01
-9.16416466e-01 -7.34327435e-02 -1.82603206e-03 9.65911448e-01
3.01074773e-01 -1.77356556e-01 6.70504808e-01 -7.84138381e-01
9.88637507e-01 -7.73293018e-01 1.76791844e-04 2.29436144e-01
-3.21386606e-01 -1.79687306e-01 6.64951921e-01 -1.97350621e-01
-8.69043350e-01 -9.78792131e-01 -8.59587416e-02 -3.37488472e-01
-1.65392935e-01 6.91805243e-01 8.30451101e-02 4.45737600e-01
1.08691037e+00 5.50742626e-01 -2.17481941e-01 -5.11487434e-03
5.72788358e-01 4.91609961e-01 5.10583639e-01 -1.09725845e+00
5.24875939e-01 2.34723538e-01 -3.04009557e-01 -7.70710468e-01
-1.12224936e+00 -2.25712016e-01 -3.88265364e-02 1.34911656e-01
7.29987323e-01 -6.72056854e-01 -3.73647422e-01 -1.09497398e-01
-1.21390116e+00 -4.07386929e-01 -7.36377001e-01 1.72423065e-01
-1.14880693e+00 1.72972485e-01 -6.77210152e-01 -6.96054399e-01
-3.63997370e-01 -7.87916839e-01 6.65161729e-01 -2.10141674e-01
-8.58365178e-01 -1.11812401e+00 1.53280422e-01 7.37422824e-01
5.17171621e-01 5.09862721e-01 1.47389126e+00 -1.19628513e+00
-2.39108264e-01 -1.08124293e-01 -5.64864874e-02 6.35688305e-01
-5.29279560e-02 -1.66702166e-01 -8.59802783e-01 2.84882516e-01
2.96875119e-01 -9.46983814e-01 7.59303212e-01 5.36520444e-02
1.16513884e+00 -2.46413872e-01 1.56247407e-01 2.24125579e-01
1.48719585e+00 3.39785255e-02 9.78473961e-01 5.95740736e-01
1.83005303e-01 4.30120736e-01 4.25809205e-01 1.74315751e-01
9.25976336e-01 4.05491501e-01 3.06249470e-01 5.39541960e-01
-2.32332647e-01 -2.28432536e-01 4.91733611e-01 9.09170389e-01
-2.55805105e-01 1.08742885e-01 -1.31093907e+00 7.66123831e-01
-1.37726653e+00 -1.67814016e+00 -1.56778008e-01 1.76099896e+00
1.25258934e+00 1.36726245e-01 -9.50286686e-02 1.85551345e-01
5.62781930e-01 4.39488381e-01 -4.18443799e-01 -9.09051836e-01
-4.94092703e-01 5.10836661e-01 -3.78961951e-01 4.50171471e-01
-6.85053051e-01 1.11423647e+00 6.32229710e+00 9.17294204e-01
-7.64740944e-01 3.18102270e-01 4.30813998e-01 -4.18434925e-02
-9.16255713e-01 1.28364876e-01 -2.33789578e-01 4.04223084e-01
9.95732486e-01 -6.18465543e-01 6.13360167e-01 1.03001773e+00
-2.91311026e-01 2.91665317e-03 -1.22019255e+00 1.07694769e+00
5.28346241e-01 -1.36039782e+00 1.57216772e-01 -2.52969891e-01
4.68079269e-01 2.48042300e-01 -1.79561362e-01 1.08406353e+00
4.20754820e-01 -1.15901804e+00 1.10196495e+00 4.46708828e-01
6.84097171e-01 -5.89760959e-01 5.26522219e-01 4.10215348e-01
-6.07683599e-01 -2.71254629e-01 -5.52545011e-01 -3.15336287e-01
-1.81054175e-01 5.46185374e-01 -2.92881697e-01 3.03394198e-01
4.80461448e-01 3.20764780e-01 -6.56218708e-01 3.93202156e-01
-4.04327273e-01 6.89236224e-01 -1.63344219e-01 -4.48611051e-01
3.40452373e-01 -1.36542186e-01 6.70260668e-01 1.29035640e+00
1.85614482e-01 3.59420329e-01 -2.15716902e-02 1.64793444e+00
-9.02375132e-02 1.34344092e-02 -7.95512855e-01 -3.13857645e-01
4.69575465e-01 8.66212785e-01 -3.70027423e-01 -3.90265435e-01
-5.50035499e-02 8.00606787e-01 5.69070697e-01 9.15271044e-02
-1.17323494e+00 -3.89892489e-01 6.30120993e-01 -9.79508758e-02
-2.21373186e-01 -3.40813585e-02 -4.56519842e-01 -1.37400424e+00
-6.27084775e-03 -1.18121755e+00 5.07350683e-01 -1.16428077e+00
-1.56105804e+00 4.93079811e-01 1.12946130e-01 -6.34309232e-01
-5.21470308e-01 -7.72602558e-01 -6.19177997e-01 8.37254167e-01
-1.29890203e+00 -1.20596135e+00 -2.36250266e-01 7.09065735e-01
4.56833422e-01 -5.38762957e-02 1.24859095e+00 -3.27922940e-01
-1.27648979e-01 3.18110764e-01 -1.05947927e-01 1.82500541e-01
4.73871499e-01 -1.45375049e+00 1.96126804e-01 4.53136116e-01
2.49005899e-01 9.36283529e-01 9.31405485e-01 -4.75440472e-01
-1.21848893e+00 -6.62255764e-01 1.25796485e+00 -9.62326348e-01
1.15348017e+00 -1.87470973e-01 -9.25452352e-01 9.70817924e-01
2.52450973e-01 -9.38404277e-02 1.01273286e+00 4.18803453e-01
-9.93202150e-01 1.82008579e-01 -1.26628852e+00 7.32556760e-01
1.11474931e+00 -1.03784668e+00 -1.65348816e+00 5.74612975e-01
6.87238157e-01 -2.52095908e-01 -7.67968595e-01 8.53793621e-02
1.60253316e-01 -1.08937383e+00 7.04765916e-01 -8.41392696e-01
1.05258274e+00 -6.59103617e-02 -8.29533279e-01 -1.33239341e+00
-4.37689424e-01 -3.17854375e-01 -6.17225170e-02 1.17834878e+00
1.43258989e-01 -7.78676331e-01 3.18717271e-01 6.73041344e-01
-2.46465519e-01 -6.26696110e-01 -1.26911545e+00 -1.07571411e+00
7.28075564e-01 -9.20846522e-01 4.65884507e-01 1.45910025e+00
6.25504792e-01 6.59647942e-01 3.20310652e-01 -4.50847089e-01
5.70793331e-01 4.93107170e-01 2.89194435e-01 -1.15520012e+00
-2.65055776e-01 -7.57617772e-01 -7.05764771e-01 -4.26391333e-01
5.97243667e-01 -1.52942252e+00 -1.37721822e-01 -1.83828092e+00
3.82366478e-01 -2.81180888e-01 -3.27185541e-01 6.10022902e-01
-1.74081191e-01 1.05721494e-02 3.61390710e-01 4.16089483e-02
-4.49929744e-01 4.45517331e-01 1.08881247e+00 -2.32574552e-01
3.42217743e-01 -6.27867222e-01 -1.07523346e+00 6.65064394e-01
1.20173299e+00 -1.93394437e-01 -5.75491488e-01 -3.22892159e-01
5.34036875e-01 -8.57875347e-02 1.01266229e+00 -8.97740781e-01
-5.67717245e-03 -4.82770234e-01 3.27902511e-02 -3.61960858e-01
5.14321625e-01 -3.89997274e-01 -4.90387231e-02 2.60897189e-01
-7.16110528e-01 1.50429547e-01 1.99933037e-01 7.75999203e-02
-2.88489908e-01 -4.42127764e-01 7.78983414e-01 -5.10585368e-01
-8.57508957e-01 -3.82650763e-01 -2.77125668e-02 8.84381831e-01
8.30313563e-01 -2.92616785e-01 -7.87265301e-01 -3.06000888e-01
-5.54347217e-01 -2.73045421e-01 4.00401264e-01 3.73848826e-01
5.57560027e-01 -1.30781484e+00 -9.35925424e-01 -4.80593890e-01
5.57341635e-01 -3.91155571e-01 2.01908126e-01 5.76511085e-01
-4.06120837e-01 6.56224549e-01 -1.93565056e-01 -8.27648714e-02
-7.45025039e-01 7.05623984e-01 3.15877378e-01 -3.27202469e-01
-5.73647797e-01 7.72394478e-01 -2.02874884e-01 -6.93745315e-01
-3.76118153e-01 -6.01953149e-01 7.31479302e-02 -3.31436866e-03
4.37591285e-01 3.80246341e-01 -1.39055550e-01 -6.32777750e-01
-4.93336886e-01 -6.11469336e-03 2.44540721e-01 -4.47353460e-02
1.41858625e+00 2.58160472e-01 -6.41952693e-01 9.92030621e-01
1.03533244e+00 -9.99899432e-02 -9.91650969e-02 -7.78352981e-03
1.36939481e-01 -2.07437873e-01 -1.03277422e-01 -1.17188334e+00
-3.36430788e-01 8.07400644e-01 2.57095806e-02 5.64224660e-01
7.40128040e-01 3.53099823e-01 4.90983933e-01 6.46439672e-01
5.68177223e-01 -1.16086054e+00 3.49283740e-02 9.96388137e-01
1.54460275e+00 -9.56594050e-01 -2.30780542e-01 -9.18403640e-02
-6.86190486e-01 9.58219409e-01 4.89480317e-01 -3.23578119e-01
2.85750806e-01 -7.67461509e-02 -2.15665638e-01 -3.91710669e-01
-8.63074780e-01 -3.12409610e-01 7.99435303e-02 6.45535707e-01
7.30919480e-01 4.41060036e-01 -4.12170500e-01 7.82803595e-01
-9.56818163e-01 -2.58609075e-02 5.63077748e-01 7.06061125e-01
-4.72003549e-01 -6.13156259e-01 -3.50077480e-01 6.76446855e-01
-1.92129925e-01 -6.64856136e-01 -5.88437736e-01 1.01539862e+00
8.45894516e-02 9.85782146e-01 -1.04009181e-01 -1.42319456e-01
3.39478254e-01 5.01461267e-01 6.80949867e-01 -7.88695216e-01
-6.16694689e-01 -8.22852075e-01 4.57793087e-01 -6.15284562e-01
-2.99614698e-01 -6.29943490e-01 -1.24400079e+00 -6.32657170e-01
-1.28354818e-01 5.96725568e-02 3.69977474e-01 1.06509793e+00
4.58854437e-02 5.06715238e-01 2.18998664e-03 -2.85343438e-01
-1.00142181e+00 -1.00128198e+00 -5.76692462e-01 7.44852543e-01
8.03913251e-02 -6.48297250e-01 -5.51270187e-01 -8.57734382e-02] | [10.057890892028809, 8.074295997619629] |
79e4a4e8-2679-4a06-aed6-cc61e899bc6d | baseline-needs-more-love-on-simple-word | 1805.09843 | null | http://arxiv.org/abs/1805.09843v1 | http://arxiv.org/pdf/1805.09843v1.pdf | Baseline Needs More Love: On Simple Word-Embedding-Based Models and Associated Pooling Mechanisms | Many deep learning architectures have been proposed to model the
compositionality in text sequences, requiring a substantial number of
parameters and expensive computations. However, there has not been a rigorous
evaluation regarding the added value of sophisticated compositional functions.
In this paper, we conduct a point-by-point comparative study between Simple
Word-Embedding-based Models (SWEMs), consisting of parameter-free pooling
operations, relative to word-embedding-based RNN/CNN models. Surprisingly,
SWEMs exhibit comparable or even superior performance in the majority of cases
considered. Based upon this understanding, we propose two additional pooling
strategies over learned word embeddings: (i) a max-pooling operation for
improved interpretability; and (ii) a hierarchical pooling operation, which
preserves spatial (n-gram) information within text sequences. We present
experiments on 17 datasets encompassing three tasks: (i) (long) document
classification; (ii) text sequence matching; and (iii) short text tasks,
including classification and tagging. The source code and datasets can be
obtained from https:// github.com/dinghanshen/SWEM. | ['Ricardo Henao', 'Wenlin Wang', 'Lawrence Carin', 'Yizhe Zhang', 'Qinliang Su', 'Martin Renqiang Min', 'Guoyin Wang', 'Dinghan Shen', 'Chunyuan Li'] | 2018-05-24 | baseline-needs-more-love-on-simple-word-1 | https://aclanthology.org/P18-1041 | https://aclanthology.org/P18-1041.pdf | acl-2018-7 | ['subjectivity-analysis'] | ['natural-language-processing'] | [ 4.53343928e-01 -1.65445969e-01 -2.18799114e-01 -4.01882231e-01
-6.79371893e-01 -6.68210566e-01 8.91350687e-01 5.15184820e-01
-7.54842103e-01 3.00321192e-01 5.10036528e-01 -6.93480730e-01
-1.53257204e-02 -4.49623883e-01 -3.91097337e-01 -6.60966098e-01
1.70814767e-02 1.44881830e-01 8.03387091e-02 -1.00314222e-01
4.99541283e-01 3.89786780e-01 -1.25934458e+00 5.11519730e-01
5.98547399e-01 9.28121984e-01 3.12914401e-01 1.00150561e+00
-4.76731569e-01 6.60537362e-01 -4.50120300e-01 -4.73576695e-01
-3.24351825e-02 -2.47208729e-01 -1.00028765e+00 -1.88736096e-01
2.85896391e-01 -2.68686712e-01 -4.94172275e-01 9.37160254e-01
6.28491223e-01 1.98713765e-01 5.31775057e-01 -1.12570906e+00
-1.21840346e+00 7.03240693e-01 -2.73117542e-01 2.38508627e-01
3.21429223e-01 1.89575806e-01 1.67307043e+00 -1.27464211e+00
4.66958970e-01 1.11082065e+00 8.81808519e-01 3.73934507e-01
-1.11012471e+00 -2.92919964e-01 7.25837946e-02 1.66641533e-01
-1.27449656e+00 -3.32059443e-01 5.02503455e-01 -4.49614793e-01
1.44164526e+00 3.49038422e-01 2.60655999e-01 1.16759849e+00
3.40244681e-01 7.87706137e-01 6.50015414e-01 -7.49245644e-01
-1.21184783e-02 1.33704498e-01 4.95546758e-01 6.34728789e-01
2.92400271e-01 -1.73438668e-01 -5.77088058e-01 -3.32538903e-01
3.24596792e-01 2.61870116e-01 -2.46532828e-01 -4.66589481e-02
-1.15800142e+00 9.76592600e-01 3.52226734e-01 6.43631577e-01
-2.39145488e-01 2.17206389e-01 7.15664208e-01 3.19060594e-01
6.46692157e-01 5.81273019e-01 -6.47438765e-01 -1.17039666e-01
-9.93027985e-01 2.24140450e-01 7.00223148e-01 8.32207322e-01
5.52973092e-01 -1.49662808e-01 -1.96876273e-01 1.01715016e+00
2.83834785e-01 -1.84883978e-02 1.03506684e+00 -3.24971467e-01
6.31081760e-01 5.67401946e-01 -1.01359747e-01 -1.00696397e+00
-5.41756570e-01 -1.59089461e-01 -7.29241788e-01 -2.01760218e-01
3.24542373e-01 -2.25465391e-02 -7.28739679e-01 1.55364537e+00
-1.64577827e-01 -1.96145117e-01 -4.21308912e-02 7.90009141e-01
7.91310310e-01 7.96879172e-01 6.03831187e-02 2.93394655e-01
1.80368483e+00 -1.11979580e+00 -7.03828454e-01 -2.75928319e-01
9.75026667e-01 -9.80315685e-01 1.22871840e+00 -8.36558416e-02
-9.42247689e-01 -4.31057036e-01 -1.01960838e+00 -5.59351623e-01
-9.38243151e-01 1.96949244e-01 4.67219234e-01 5.97835779e-01
-9.95142996e-01 5.30188680e-01 -6.61817968e-01 -5.92014313e-01
2.81146675e-01 3.02686661e-01 -3.95211369e-01 -4.28393856e-03
-1.42531157e+00 9.34042752e-01 4.54189748e-01 4.94710393e-02
-3.53286475e-01 -7.63083756e-01 -1.02267659e+00 3.00652653e-01
-3.97627167e-02 -5.18022954e-01 1.23493814e+00 -8.84009123e-01
-1.38725412e+00 9.42087531e-01 -4.15078193e-01 -4.45180893e-01
2.71125138e-01 -3.28193903e-01 -1.33204386e-01 1.77413579e-02
-2.18603104e-01 6.30464256e-01 6.02523744e-01 -8.97803843e-01
-3.55280668e-01 -3.15863580e-01 9.35461372e-02 1.06783517e-01
-7.58826375e-01 3.52106810e-01 -1.48648441e-01 -1.15743148e+00
-2.42240608e-01 -8.84192586e-01 -7.68665075e-02 2.07585394e-01
-4.17161554e-01 -4.77307618e-01 5.24146140e-01 -9.37702239e-01
1.54106379e+00 -2.09580255e+00 6.37360215e-02 -1.14446551e-01
4.94976901e-02 5.46263337e-01 -5.31695783e-01 9.54695106e-01
-2.31730908e-01 4.54474717e-01 -3.64019513e-01 -5.67316830e-01
3.86997372e-01 1.10492026e-02 -4.67007160e-01 3.53201568e-01
4.68846649e-01 1.15458000e+00 -7.83955991e-01 -3.84103775e-01
7.97229037e-02 4.55763161e-01 -5.25163829e-01 1.35354221e-01
-1.55931965e-01 -1.87117711e-01 -1.90143034e-01 5.19096792e-01
3.58197659e-01 -3.06540400e-01 3.49677414e-01 -1.23648262e-02
-1.27348021e-01 8.12701523e-01 -7.72266448e-01 1.72162306e+00
-5.79997838e-01 1.02068305e+00 -1.58069894e-01 -9.44528580e-01
7.21825898e-01 4.79611367e-01 2.46189982e-01 -3.48491311e-01
2.68401772e-01 2.15488613e-01 -3.54413933e-05 -5.46852171e-01
9.85682428e-01 -6.86804950e-02 -5.71238287e-02 6.99027181e-01
3.09194833e-01 1.56737626e-01 2.12650791e-01 4.00715731e-02
1.28384686e+00 -3.89712751e-02 4.94319886e-01 -4.99916852e-01
5.65254688e-01 -1.76882431e-01 9.78804603e-02 6.16586804e-01
-1.84024438e-01 6.28671646e-01 4.77339715e-01 -6.14145696e-01
-1.31590152e+00 -8.07850838e-01 -1.61467105e-01 1.52483225e+00
-1.28709361e-01 -5.40546179e-01 -7.66507864e-01 -3.45363259e-01
1.58739462e-01 4.93593097e-01 -6.96528792e-01 -7.73876682e-02
-7.29411006e-01 -9.27627385e-01 9.02039528e-01 8.32654297e-01
1.44145787e-01 -1.13351405e+00 -4.94408876e-01 1.55856460e-01
-2.65448213e-01 -8.93831611e-01 -1.01522791e+00 5.83832145e-01
-8.38764012e-01 -8.32626760e-01 -7.65660584e-01 -1.05515850e+00
5.64551115e-01 2.46007740e-01 9.00467098e-01 3.12662870e-01
-4.25188750e-01 1.25436187e-01 -5.63591003e-01 -3.25390726e-01
-1.22371778e-01 5.21304250e-01 -6.07920550e-02 -2.12922931e-01
7.48394549e-01 -2.45316565e-01 -4.93179798e-01 1.03114411e-01
-1.27614474e+00 -1.61581099e-01 6.82572126e-01 1.02994514e+00
1.39964432e-01 -3.03095073e-01 4.40759152e-01 -7.23528326e-01
1.10883534e+00 -4.23389822e-01 -2.42398307e-01 3.39658499e-01
-5.43668449e-01 2.59028226e-01 7.55764306e-01 -5.17315984e-01
-8.27627301e-01 -2.16210559e-01 -5.11041045e-01 8.86274427e-02
-9.68558937e-02 6.68394983e-01 1.81815356e-01 1.69058189e-01
5.68944156e-01 4.02758002e-01 7.22460402e-03 -6.53483987e-01
5.68373561e-01 1.03537023e+00 1.36087015e-01 -4.71795708e-01
5.83527029e-01 2.67941147e-01 -4.05541241e-01 -9.93238270e-01
-6.23705029e-01 -6.30079746e-01 -8.28309000e-01 2.78065205e-01
1.02010107e+00 -5.38847268e-01 -3.67141664e-01 4.16578323e-01
-1.38857508e+00 -3.88483733e-01 -1.96595155e-02 3.98335248e-01
-3.13888639e-01 5.99689484e-01 -1.06612170e+00 -6.24690413e-01
-5.79739869e-01 -9.73947108e-01 9.75708544e-01 -1.59815297e-01
-7.40878105e-01 -1.26846302e+00 1.01957612e-01 1.63836956e-01
4.79430914e-01 -1.63695991e-01 1.23569441e+00 -1.00849545e+00
-2.25244060e-01 -4.72820342e-01 -3.23035777e-01 3.68168473e-01
2.60822363e-02 1.19653344e-01 -1.06515098e+00 -2.98542172e-01
-3.41625571e-01 -2.59771377e-01 9.81529415e-01 2.50728250e-01
1.21837664e+00 -3.70399445e-01 -1.46858484e-01 6.27659559e-01
1.30212128e+00 1.41392842e-01 5.85776687e-01 4.49422300e-01
6.13276899e-01 6.44444823e-01 1.57703027e-01 3.14408749e-01
1.48774460e-01 6.70825958e-01 1.88356817e-01 9.99451354e-02
-1.55037299e-01 -3.33015889e-01 3.53888094e-01 1.18239951e+00
2.03163460e-01 -7.19104946e-01 -8.51401687e-01 6.27412796e-01
-1.85766828e+00 -7.40014136e-01 3.48283746e-03 1.87679601e+00
8.16216409e-01 5.07113561e-02 -4.29461673e-02 2.25504220e-01
6.75941885e-01 5.82044542e-01 -2.43828103e-01 -7.73288667e-01
-1.36368414e-02 3.26198339e-01 4.13399518e-01 4.48880315e-01
-1.08370364e+00 7.04963982e-01 6.32019949e+00 9.04607296e-01
-9.87929404e-01 1.30666181e-01 3.62257898e-01 -1.80733968e-02
-3.58165413e-01 -2.29068041e-01 -7.09565818e-01 5.36335647e-01
1.06987810e+00 6.28121942e-02 2.75385946e-01 6.32760227e-01
-1.53243616e-01 3.12704206e-01 -1.04872882e+00 7.25783169e-01
1.95451200e-01 -1.38153398e+00 1.64020494e-01 -1.76176876e-01
4.02809352e-01 1.05700187e-01 1.20345084e-03 5.45831770e-02
1.33172378e-01 -1.00417423e+00 7.32119381e-01 3.31943065e-01
7.17863441e-01 -4.42295104e-01 1.04815972e+00 1.49559021e-01
-1.30967879e+00 -5.16552618e-03 -2.81316549e-01 -8.81164074e-02
9.42972153e-02 5.24020910e-01 -4.31459606e-01 4.55393374e-01
5.89043498e-01 5.76575220e-01 -4.97985512e-01 7.23419368e-01
-3.39223981e-01 5.15587568e-01 5.40497452e-02 -5.80691040e-01
5.36857188e-01 -2.97794379e-02 2.76046962e-01 1.90015340e+00
2.59106308e-01 -2.01768145e-01 -3.13392133e-01 7.32029736e-01
-3.38733763e-01 2.04011828e-01 -4.40539986e-01 -5.05145311e-01
4.22859162e-01 1.25777054e+00 -7.48617113e-01 -2.49566674e-01
-6.31467462e-01 1.10699689e+00 4.01557297e-01 3.77025574e-01
-5.90676248e-01 -9.72859800e-01 9.98362780e-01 -1.20564148e-01
4.80900973e-01 -3.65438998e-01 -5.12515783e-01 -1.13110030e+00
1.62226200e-01 -7.14792550e-01 4.12602663e-01 -6.25091195e-01
-1.47347093e+00 5.76804519e-01 -2.69083977e-01 -7.89976418e-01
-1.22624755e-01 -1.14568448e+00 -6.72445536e-01 1.06354320e+00
-1.43027401e+00 -9.93864894e-01 -4.82308194e-02 6.66040275e-03
6.52286589e-01 -9.71096456e-02 9.94857967e-01 4.56456870e-01
-5.97006977e-01 7.51535833e-01 3.51842254e-01 4.94054586e-01
6.86092973e-01 -1.22963285e+00 6.52770400e-01 7.43797362e-01
1.38010085e-01 9.65993881e-01 4.38135952e-01 -2.82172441e-01
-1.34827840e+00 -1.07439089e+00 1.58114529e+00 -5.69856465e-01
9.23810005e-01 -7.85706580e-01 -1.17447674e+00 8.28608036e-01
4.58156407e-01 -6.24268837e-02 1.00059783e+00 7.96455666e-02
-6.04565144e-01 2.22628728e-01 -8.52674961e-01 7.08723724e-01
9.97744381e-01 -8.87924910e-01 -7.79238939e-01 2.30507717e-01
9.20192540e-01 -3.43874618e-02 -8.48807693e-01 1.16190743e-02
9.61135685e-01 -6.65088296e-01 1.04649401e+00 -8.68938684e-01
7.78165281e-01 -3.69125828e-02 -2.41157532e-01 -1.20209706e+00
-5.68104327e-01 -3.94861907e-01 1.92850213e-02 1.07142127e+00
5.11772096e-01 -1.02295887e+00 4.69036758e-01 3.51426244e-01
-2.65769422e-01 -1.04758024e+00 -6.61763668e-01 -8.70973825e-01
4.32796925e-01 -3.90336484e-01 6.10322177e-01 9.86349165e-01
3.62625211e-01 2.24305928e-01 -1.58477008e-01 -6.96990937e-02
1.24069052e-02 -3.39251906e-02 3.37848246e-01 -7.91236520e-01
-1.36549190e-01 -9.47626472e-01 -3.24315876e-01 -1.23339307e+00
2.59194762e-01 -1.08445919e+00 1.92995280e-01 -1.62083161e+00
2.72733033e-01 5.55891497e-03 -3.30588192e-01 5.21723866e-01
-3.71849537e-01 2.89498419e-01 9.42055136e-02 3.39278370e-01
-3.75000209e-01 4.43138689e-01 6.36353374e-01 -8.82575661e-02
1.73746884e-01 -3.90994966e-01 -7.02853501e-01 5.83209276e-01
1.09881413e+00 -4.85546917e-01 -1.42641172e-01 -6.78969502e-01
2.80226052e-01 -3.67228240e-01 1.88997686e-01 -4.40059483e-01
1.51970357e-01 5.97916096e-02 2.93003500e-01 -4.66878772e-01
1.70434117e-01 -5.33724785e-01 -3.75285089e-01 6.78902268e-01
-7.84355223e-01 3.50362659e-01 3.41668755e-01 3.44117284e-01
-2.83216238e-01 -7.38253236e-01 5.46270907e-01 5.60620539e-02
-6.25558317e-01 1.42499983e-01 -7.66460419e-01 -3.65799144e-02
5.82192063e-01 -1.34213999e-01 -5.30508101e-01 -1.95268199e-01
-2.74340600e-01 -5.09047471e-02 3.28365177e-01 6.13341749e-01
4.48579639e-01 -1.31017470e+00 -5.28146207e-01 1.45103619e-01
3.19222659e-01 -4.22734320e-01 -3.21544856e-02 7.04382837e-01
-6.55268192e-01 9.37792659e-01 7.55272657e-02 -9.51547399e-02
-1.27940822e+00 5.91332614e-01 1.13830514e-01 -5.18979430e-01
-3.39017838e-01 9.42159951e-01 1.25585586e-01 -8.18588614e-01
2.48804733e-01 -6.81499064e-01 1.93256214e-02 2.03013882e-01
6.78911269e-01 3.64023238e-01 1.67543367e-01 -5.31006813e-01
-4.90304619e-01 4.82483923e-01 -1.08336359e-01 8.43446180e-02
1.25082374e+00 -6.19623773e-02 -2.72238851e-01 6.25464082e-01
1.54254246e+00 -3.63907874e-01 -9.25702512e-01 -5.19790232e-01
4.55214918e-01 -3.15923542e-01 -9.71762165e-02 -5.42316139e-01
-5.43611825e-01 1.26546776e+00 2.71048486e-01 5.07744312e-01
9.77409840e-01 -1.41526073e-01 1.06816757e+00 3.40533286e-01
-1.82529628e-01 -1.00299191e+00 1.72048196e-01 7.82544553e-01
8.59761357e-01 -1.14970303e+00 -8.59959275e-02 -3.28015424e-02
-3.71144444e-01 1.47342145e+00 1.49194762e-01 -4.99353223e-02
6.99391246e-01 3.20699275e-01 -3.77959833e-02 -5.72077706e-02
-8.56238484e-01 -1.85197026e-01 5.02445519e-01 2.98372298e-01
9.72532451e-01 -2.87676677e-02 -4.60603148e-01 6.69902265e-01
-1.82549488e-02 -1.73942462e-01 7.46169090e-02 1.05285180e+00
-4.55606401e-01 -1.11708164e+00 -1.51200265e-01 4.80494827e-01
-6.21904254e-01 -5.66885233e-01 -5.22954166e-01 6.65727258e-01
-1.26845896e-01 7.68885434e-01 1.75216377e-01 -3.23327720e-01
1.42736599e-01 6.48341298e-01 2.61255175e-01 -6.43198967e-01
-8.09546649e-01 -4.17981058e-01 1.29277939e-02 -3.04074943e-01
-2.64516801e-01 -6.12593234e-01 -1.14563656e+00 -4.44372684e-01
-2.99192399e-01 -1.20510809e-01 7.35298872e-01 8.97040069e-01
3.53119850e-01 5.96014023e-01 1.94756180e-01 -8.92236173e-01
-8.12591136e-01 -1.36152840e+00 -4.82685119e-01 5.69658875e-01
3.09805512e-01 -2.34285176e-01 -4.65029866e-01 2.24551499e-01] | [10.69913101196289, 8.601259231567383] |
f8efab9c-cc0a-4c05-a368-a793a4019a6f | online-nonnegative-matrix-factorization-with-1 | 1604.02634 | null | http://arxiv.org/abs/1604.02634v2 | http://arxiv.org/pdf/1604.02634v2.pdf | Online Nonnegative Matrix Factorization with Outliers | We propose a unified and systematic framework for performing online
nonnegative matrix factorization in the presence of outliers. Our framework is
particularly suited to large-scale data. We propose two solvers based on
projected gradient descent and the alternating direction method of multipliers.
We prove that the sequence of objective values converges almost surely by
appealing to the quasi-martingale convergence theorem. We also show the
sequence of learned dictionaries converges to the set of stationary points of
the expected loss function almost surely. In addition, we extend our basic
problem formulation to various settings with different constraints and
regularizers. We also adapt the solvers and analyses to each setting. We
perform extensive experiments on both synthetic and real datasets. These
experiments demonstrate the computational efficiency and efficacy of our
algorithms on tasks such as (parts-based) basis learning, image denoising,
shadow removal and foreground-background separation. | ['Renbo Zhao', 'Vincent Y. F. Tan'] | 2016-04-10 | null | null | null | null | ['shadow-removal'] | ['computer-vision'] | [ 0.04690439 -0.33323166 0.06091622 -0.22466573 -0.90877026 -0.46608096
0.06297755 -0.2874176 -0.3874195 0.66834056 0.09090465 -0.3292489
-0.22185549 -0.16530623 -0.9023168 -0.92853427 -0.20856003 0.32646367
-0.29791966 0.02492271 0.11503962 0.4889313 -1.1843462 -0.10470826
0.95422596 0.8891002 0.05799958 0.9213043 0.10870861 0.60099936
-0.29431495 -0.24437219 0.79864174 -0.43797937 -0.41844022 0.8065896
0.53530025 -0.28132892 -0.00890067 1.2322437 0.5373921 0.21833955
0.2857177 -1.4143941 -0.41509104 0.11511749 -1.1548797 0.34088647
0.07186338 -0.01821639 0.7829536 -1.3405025 0.4195158 1.0968019
1.0415719 0.26004726 -1.3572065 -0.30097073 0.44151574 -0.18836993
-1.0487984 -0.52237755 0.5759139 -0.42511535 0.5312658 0.506624
0.5305201 0.7002073 -0.0972129 0.97868884 1.0456815 -0.5689418
0.22299404 0.00890466 0.12766643 0.8196113 0.4848142 -0.1514682
-0.6698432 -0.5545902 0.7039486 0.11915104 -0.44689235 -0.6488157
-1.2461189 0.90072525 -0.11920661 -0.16639538 -0.43003792 0.22893319
0.35504946 0.33961836 0.7891476 -0.02984649 -0.4362576 0.16470535
-0.93136024 0.1442272 0.9830774 0.8736906 0.57717466 0.41439408
0.09305079 0.8225694 0.3573048 1.0542214 0.11225646 -1.1607407
0.46566576 -0.01728517 0.50469375 -1.2408692 -0.29311743 -0.5613443
-0.96811795 0.08336817 0.58568066 -0.41948608 -0.74483943 1.7509273
0.58378565 0.60289735 -0.30824968 0.9927369 0.14121376 0.66088134
-0.48016664 -0.82031375 0.78411937 -0.852224 -0.86460215 -0.29482836
0.3575025 -0.8144384 0.9601205 0.5057389 -1.1773375 -0.2311531
-0.6710811 0.2766432 0.30266762 0.2819771 0.6609951 0.7406587
-0.9615559 0.5235167 -1.0094476 -0.11495045 0.25604525 0.3795908
-0.11770435 0.0286727 -0.4343497 0.36481816 -0.02438055 0.5415636
-0.82347757 -0.75278085 -0.76046443 -0.05744615 0.4543666 -0.7898945
1.0897799 -1.3057978 -1.317962 0.9395572 -0.4494136 -0.42868805
0.69452316 -0.622905 -0.04802245 -0.00992962 0.26210147 -0.19893117
1.2711613 -1.3068417 -0.5381313 -0.3714834 -0.18905638 0.1253904
-0.5201215 0.11834816 -0.5187127 -0.77545995 0.22458617 -1.0217549
-0.5848705 0.21861896 -0.28245184 0.44502985 0.44369876 -0.8880028
1.1071477 -2.2569053 0.22125994 0.44322598 0.22488528 -0.08358797
-0.02565116 0.3437151 -0.23628792 -0.2717624 -0.5649837 -0.60985535
-0.03233275 0.20272528 -0.63187915 1.1379462 -0.1645325 0.4411611
-0.90079314 -0.12665696 0.01872384 0.14943187 -0.54806215 0.05563608
0.14901656 0.21951252 -0.193436 0.49681053 0.9443722 -0.36176565
0.33314574 -0.06130487 0.01699979 -0.20631269 -1.8040628 1.4562259
-0.5970287 0.5738576 0.9166274 -1.1391703 0.45110977 0.12808661
0.73323154 -0.373382 -0.05640147 0.2612659 -0.45077568 -0.32649976
0.21639332 -0.46137443 0.3325515 0.7244166 -0.04431467 0.34945258
0.3425511 0.5316967 0.9643228 0.06245333 0.20381583 -0.5641375
0.44700688 -0.22043578 0.725795 0.79469824 -0.0233786 0.64753205
0.52173394 -0.4265124 -0.8646908 -1.0847063 -0.05177986 1.1913933
0.04510412 -0.38162035 -0.5277783 -0.6323407 0.30615646 0.34426877
-0.5773014 0.24304582 -0.5743276 -1.300471 -0.11056353 0.38795722
0.05495889 -0.51592207 -0.40054798 0.09921798 -0.22143224 -1.1497613
-0.7424858 0.15136643 -1.1265723 -1.177719 -0.72946656 -0.60185874
1.1074731 0.72043437 1.274122 0.09224073 -0.20478332 0.9513986
-0.06538229 -0.31097725 -0.08334112 -0.511151 0.2638982 0.5276712
-0.18812382 -0.48810443 -0.5349203 0.32055986 -0.9478464 -0.1696188
0.19034944 1.0421981 0.8930535 0.00773117 0.1335893 -1.0897992
0.6025188 -0.43323517 -1.0917137 0.20717539 -0.48805577 0.06940451
0.52702636 -0.36596242 -1.0514287 0.36585924 0.38545898 -0.5193323
0.37225074 0.4234428 -0.14359517 -0.20022784 0.52130437 0.12276036
0.029366 -0.41929555 0.5962221 0.27933395 0.5807233 -0.72381103
1.0332139 1.1674137 0.03456909 -1.0578246 -1.0575864 -0.8428864
-0.3244797 -0.30038953 0.3554676 -0.95246434 -0.6689367 0.36467877
-1.0086484 -0.38100296 -0.4438307 0.32070994 -0.70051324 0.8109213
-0.6115759 -1.1925547 -0.33614823 -0.7834026 1.0660343 -0.11857962
0.24272776 -1.1848835 0.37577775 0.25138217 0.05265656 0.34262347
0.18712026 0.01744086 -0.26105937 -0.08210092 -0.1695367 0.6574236
0.09977444 0.01405328 -0.8168767 -0.59423214 0.56042045 -0.11817223
1.1811774 0.7757777 0.9593649 -0.4320744 -0.20838456 0.9968463
1.5989437 -0.4239113 0.40293238 0.27982247 0.7311031 0.48172346
0.7145246 0.8101114 -0.08693647 0.40659088 0.42882594 -0.19769165
0.27611715 0.34870702 0.6838213 0.7542455 0.01321025 -0.05364846
-0.83823365 0.9010627 -2.3048518 -1.0557448 -0.36722326 2.437569
0.56934106 -0.0889281 0.35720915 0.07925908 0.7560807 0.03572006
-0.48775932 -0.2552251 -0.26667228 0.19746672 0.8972832 0.6480196
-1.2188052 0.5287989 6.677999 0.5482364 -0.8143084 0.23724177
0.5565229 -0.35499611 -0.1631304 -0.08043716 -0.50888765 0.3790038
0.7246191 -0.1005569 0.64348805 0.84541243 0.38418922 -0.13852502
-0.7750244 1.1900254 0.17316519 -1.1890339 -0.51567525 0.01743606
1.2628052 0.01012467 0.19347121 -0.31352392 0.4369928 -0.6164211
0.7569363 0.34439117 0.49704605 -0.80606115 0.3997667 0.17882957
-0.98706573 -0.286762 -0.47483093 -0.19231424 0.43083856 1.2107834
-0.37110436 0.52155435 0.65269125 0.7335131 -0.21158363 1.1519164
-0.23303927 0.7945239 -0.6972179 0.40207675 -0.05782281 -0.8058366
0.62849134 1.3112031 0.24752799 -0.02469177 0.2715032 0.466213
-0.09712642 0.14710496 -0.37320825 0.14232446 0.01178062 1.2915376
-1.0039974 -0.2590047 -0.72239274 1.133491 0.17480566 0.6993418
-0.81033254 -0.12907992 0.82994384 -0.03125839 0.6441828 -0.4315653
-0.412337 -1.5127345 0.41016784 -1.1396828 0.48675224 -0.3456393
-1.475309 0.06807858 -0.2155758 -0.870515 0.05796129 -0.70353734
-0.65803456 0.6040361 -1.2677071 -0.62525237 -0.19463564 0.7953008
0.29427627 0.05685104 0.4938401 0.36672026 -0.85449487 0.18521455
0.76087713 0.01473763 0.702278 -1.4346068 0.31106615 1.4328192
0.26207006 0.6264341 0.9433116 -0.56965595 -1.6860697 -0.87349784
0.4088282 -0.33937088 0.9702037 -0.42110103 -0.6366346 0.86048585
-0.04011991 0.25346857 0.7529422 0.3127979 -0.15256761 -0.26221675
-0.9368143 0.3872238 0.9788688 -0.37191454 0.03387112 0.9946876
0.2255011 -0.6026498 -0.37376672 0.1038873 0.42919296 -1.0285525
1.1065942 -0.89483494 0.04451252 -0.33487564 -0.40073755 -1.0242727
-0.3103132 -1.1445366 -0.5521541 0.9759819 0.26444054 -0.582127
0.85686094 0.578174 0.12890477 -0.6642269 -0.7390221 -0.97209036
-0.33831123 -0.6656382 -0.01953763 0.859515 -0.4371749 0.1504895
-0.8969236 0.4271829 1.096685 0.46081743 0.8892528 -0.8589106
-0.77972037 -0.04843055 0.01523849 -1.0414709 0.15527153 -0.7167167
0.08260182 -1.312285 0.4334706 -0.17885122 -0.33097818 0.10943336
-0.42335787 0.31088957 0.27207515 0.13728337 -0.8537014 0.52555007
0.81759214 -0.0366039 -0.30564672 0.2575181 -0.69993865 0.9254899
0.64880246 -0.5962325 -0.28165883 -0.69669694 0.69219744 -0.08758116
0.34349445 -0.529229 -0.06346557 -0.31008574 0.30210215 -0.2642953
0.18358581 -0.89906543 0.05448043 0.46579948 -0.04338476 0.08730211
0.16175315 0.9234246 0.0404258 -0.13256231 0.86286813 0.1464417
-0.26996738 0.2801868 -0.17146455 0.38595894 0.88620853 -0.1395317
0.23158295 -0.71658987 -0.9348589 0.36633834 0.4371363 -0.18895377
0.6420409 -1.2839279 -0.88615936 0.1485189 -0.3608909 -0.28345865
0.19599864 1.2880099 -0.50160086 -0.01800494 0.248164 -0.587546
-1.3654641 0.6164513 0.2777623 -0.30988023 -0.4777822 0.96199256
0.30028325 -0.26930273 0.292308 -0.21625343 0.50009197 0.0867906
0.6441465 0.9301981 0.03810198 -0.48041686 -0.3803291 0.3573234
0.16152814 -0.24065578 1.6643689 -0.35584056 -0.49606958 0.28634775
1.1182903 0.47386402 -1.4239209 -0.23230246 -0.02588937 -0.6997747
0.01221782 -0.22308676 -1.3597697 0.5722639 0.5743798 0.21931763
1.3898587 -0.37743744 0.69008183 0.45652232 0.1771506 -1.1644183
0.0302136 0.42344987 0.82923615 -1.2229606 0.44230106 -0.44711196
-0.39844108 1.048784 -0.01188519 -0.45077208 0.8043847 0.46368983
0.0765039 0.01704543 -0.4495946 -0.19780642 0.19077395 0.47889432
0.25785318 0.00995341 -0.25560948 0.1750739 0.15957607 -0.17030232
0.6370828 0.8689277 -0.3132463 -1.0369899 -0.80106544 0.48730466
-0.5490829 -0.12756196 -0.17914513 0.4604808 -0.27814186 0.93605924
-0.35969537 0.24647853 0.06443883 -0.09659716 0.70264935 -0.53182966
-0.18216377 0.4301333 -0.20234685 -0.7871735 -0.5476687 -0.98316234
-1.0260892 -0.40923166 -0.6642769 0.06794952 0.50296426 0.7581704
0.2112832 0.312498 0.73648304 -0.7520904 -0.7664648 -0.38130027
-0.80820334 0.5129044 0.5125625 -0.269937 -0.62847596 0.27684596] | [7.157547950744629, 4.497538089752197] |
71fa5a01-1b02-413d-a152-7d1eb7f0a359 | sensitivity-analysis-in-unconditional | 2303.14298 | null | https://arxiv.org/abs/2303.14298v2 | https://arxiv.org/pdf/2303.14298v2.pdf | Sensitivity Analysis in Unconditional Quantile Effects | This paper proposes a framework to analyze the effects of counterfactual policies on the unconditional quantiles of an outcome variable. For a given counterfactual policy, we obtain identified sets for the effect of both marginal and global changes in the proportion of treated individuals. To conduct a sensitivity analysis, we introduce the quantile breakdown frontier, a curve that (i) indicates whether a sensitivity analysis if possible or not, and (ii) when a sensitivity analysis is possible, quantifies the amount of selection bias consistent with a given conclusion of interest across different quantiles. To illustrate our method, we perform a sensitivity analysis on the effect of unionizing low income workers on the quantiles of the distribution of (log) wages. | ['Julian Martinez-Iriarte'] | 2023-03-24 | null | null | null | null | ['selection-bias'] | ['natural-language-processing'] | [ 7.15402961e-02 2.30749860e-01 -5.47263086e-01 -2.03174084e-01
-4.74006295e-01 -3.57923329e-01 6.58587456e-01 6.49573624e-01
-5.72453678e-01 1.00897706e+00 7.13980019e-01 -9.42482769e-01
-4.45508808e-01 -1.09093857e+00 -7.73957253e-01 -4.20953512e-01
9.26281512e-02 1.57619134e-01 -1.55182496e-01 2.69099861e-01
3.16844910e-01 4.13261950e-01 -1.06983054e+00 -1.43713161e-01
9.17617917e-01 3.78958881e-01 -3.28723580e-01 9.62828249e-02
6.43765703e-02 2.13478118e-01 -6.68574989e-01 -6.12732172e-01
4.31026787e-01 -3.55076671e-01 -5.37494004e-01 2.14354377e-02
1.88260838e-01 -3.16183746e-01 6.11550435e-02 9.34854984e-01
3.67736518e-01 2.42875695e-01 9.88675892e-01 -9.62250590e-01
-3.27067316e-01 8.99720073e-01 -6.32030725e-01 2.51571685e-01
3.18657577e-01 1.91049069e-01 7.18662858e-01 -5.29881334e-03
5.39624989e-01 1.43142486e+00 1.35023549e-01 4.41346020e-02
-1.49192584e+00 -8.54598343e-01 1.73106328e-01 -4.62661058e-01
-8.44445109e-01 -4.38080996e-01 3.83300632e-01 -6.30390525e-01
1.18556425e-01 1.69906408e-01 4.60348040e-01 5.07893622e-01
6.20265245e-01 -1.39801234e-01 1.21091247e+00 -4.56708908e-01
4.39068168e-01 1.69098955e-02 2.43483633e-01 5.07300720e-02
8.94897163e-01 3.04906338e-01 -1.67108163e-01 -6.48016393e-01
5.84755599e-01 -7.66278207e-02 -1.08295299e-01 -3.28664452e-01
-9.03257608e-01 6.91872537e-01 1.76533133e-01 -1.15184352e-01
-5.32882750e-01 1.06458917e-01 4.45979953e-01 1.68727607e-01
5.20933330e-01 4.07792717e-01 -2.51179904e-01 7.82680213e-02
-7.63229787e-01 6.92734897e-01 4.97274160e-01 -1.89288855e-01
5.87763608e-01 -1.81036919e-01 -6.98244333e-01 4.14104968e-01
-1.11098439e-01 6.35312617e-01 2.30778009e-01 -1.10927677e+00
6.98019326e-01 3.84928197e-01 7.81289101e-01 -3.40946615e-01
-1.55219853e-01 -4.08589989e-01 -2.66020119e-01 3.54177147e-01
9.36609209e-01 -5.89575589e-01 -5.76143563e-01 2.02481389e+00
1.64052665e-01 -2.64752060e-01 -1.01866320e-01 4.57417101e-01
-1.38164610e-01 4.25598472e-01 5.18472314e-01 -7.77061224e-01
1.10238850e+00 1.37960017e-01 -4.86553639e-01 7.93485716e-02
5.86191654e-01 -3.36257428e-01 8.65369558e-01 -2.68978894e-01
-1.03822744e+00 -3.93216044e-01 -3.53645325e-01 5.03970027e-01
3.12039088e-02 -4.87520248e-01 3.62682223e-01 8.30756903e-01
-4.01834458e-01 7.44586408e-01 -6.39931500e-01 -1.96265116e-01
5.18436283e-02 1.24067597e-01 -1.81592777e-02 1.62618518e-01
-9.55273151e-01 6.61553979e-01 9.68751870e-03 -3.74519020e-01
-4.72578913e-01 -8.38380218e-01 -5.26201010e-01 8.48779917e-01
4.24721777e-01 -7.35933721e-01 8.90797377e-01 -9.11625743e-01
-8.77510786e-01 3.87567371e-01 -8.35002810e-02 -4.91773099e-01
6.55710340e-01 1.33306175e-01 -1.84188396e-01 -2.82547593e-01
6.58681571e-01 5.27089089e-02 2.50077933e-01 -8.69914711e-01
-5.79542696e-01 -8.24511468e-01 1.82834893e-01 2.14113280e-01
3.00524998e-02 3.03150564e-01 3.76547128e-01 -5.93305469e-01
-3.53315949e-01 -6.73943400e-01 -3.84684175e-01 -5.27334929e-01
-5.27310312e-01 -1.61551848e-01 -3.89437303e-02 -3.46437752e-01
9.91105080e-01 -2.24815345e+00 -3.46614420e-01 4.20501351e-01
-6.03277981e-01 -3.32785755e-01 3.35327387e-01 2.45852694e-01
-2.40668148e-01 2.22515374e-01 -2.17973232e-01 8.08047801e-02
2.54113257e-01 -2.35091239e-01 -4.34381872e-01 6.63053513e-01
-2.32658125e-02 1.46890625e-01 -4.67135727e-01 -1.32291749e-01
1.47713333e-01 -2.11898565e-01 -4.76312459e-01 -1.53044879e-01
2.83073902e-01 1.23458706e-01 -1.38088837e-01 1.31542236e-01
5.64168930e-01 2.36823305e-01 6.34639978e-01 1.93979949e-01
-6.48187280e-01 6.26041055e-01 -1.01301277e+00 7.37260163e-01
-2.22489461e-01 3.18799555e-01 -1.24590524e-01 -9.80955780e-01
6.40354455e-01 1.76468179e-01 1.57248050e-01 -5.61467350e-01
1.27617911e-01 8.13963562e-02 2.48507947e-01 -8.78495798e-02
1.97400093e-01 -7.65531659e-01 -3.61662686e-01 7.79166758e-01
-3.51114661e-01 7.13013336e-02 3.83720517e-01 -2.76418068e-02
6.33319438e-01 -4.77123111e-02 5.05669296e-01 -7.81304836e-01
2.53578573e-02 -2.25159764e-01 6.72955930e-01 7.84407616e-01
-2.08899632e-01 4.99339178e-02 1.16472375e+00 -8.72112811e-02
-7.33355165e-01 -1.33918643e+00 -5.31958163e-01 8.22718680e-01
5.24753518e-02 4.42624837e-01 -4.59268808e-01 -5.04027188e-01
6.64060652e-01 1.23269081e+00 -7.59400427e-01 -1.42963335e-01
-2.78222859e-01 -1.06147182e+00 -5.71521819e-02 3.51600915e-01
5.06421447e-01 -7.02234089e-01 -1.09100127e+00 -3.76704842e-01
2.11092934e-01 -4.42877203e-01 -3.32825124e-01 -1.60660744e-01
-8.33205640e-01 -1.06940603e+00 -5.70202589e-01 -1.62283331e-01
4.72537130e-01 -1.96555145e-02 7.13568449e-01 -4.02083814e-01
4.25713390e-01 2.19736353e-01 6.69344738e-02 -7.06730187e-01
-2.45635703e-01 -5.87483868e-03 1.11795150e-01 -8.67389217e-02
1.71111301e-02 -2.54502147e-01 -4.64764088e-01 -2.47504294e-01
-5.60253263e-01 -4.18557525e-01 7.61030987e-02 2.69434452e-01
1.53470799e-01 1.13045871e-01 9.96511519e-01 -9.43635941e-01
8.86219442e-01 -5.35097003e-01 -8.37728858e-01 4.58933324e-01
-8.36626351e-01 -4.10484709e-02 3.18047196e-01 -2.30951920e-01
-1.51845407e+00 -4.54373002e-01 2.98404783e-01 4.00856584e-01
-3.55990797e-01 7.58868933e-01 -1.12434886e-01 6.08217239e-01
6.49386108e-01 -4.13727522e-01 -1.80269498e-02 -5.29887319e-01
6.43943846e-02 6.64235353e-01 6.21403158e-01 -9.19714570e-01
5.47348082e-01 5.52492261e-01 9.12096947e-02 -4.40102577e-01
-6.50122166e-01 -2.10382372e-01 -2.07681268e-01 -1.01455629e-01
6.65726244e-01 -7.57230699e-01 -8.19802701e-01 -1.15521237e-01
-4.47970480e-01 -5.97756565e-01 -5.87228179e-01 8.35489929e-01
-4.90262091e-01 1.51312768e-01 -1.16919786e-01 -1.15188575e+00
1.40974864e-01 -9.43811238e-01 2.26958364e-01 4.43288505e-01
-4.46676373e-01 -8.32387686e-01 1.43482342e-01 1.73069790e-01
-3.99158671e-02 7.53894806e-01 1.17757344e+00 -5.00789702e-01
-8.41205269e-02 -2.47390538e-01 -8.79607499e-02 -2.37553477e-01
2.75247455e-01 2.55620331e-01 -4.67122793e-01 -3.43304455e-01
-1.39323771e-01 5.95173799e-02 5.80531120e-01 1.06226802e+00
8.19457352e-01 -4.89362717e-01 -4.37788486e-01 1.63384750e-01
1.39074600e+00 3.50243151e-01 4.13988650e-01 5.11253953e-01
-7.41420016e-02 9.44407403e-01 5.37532091e-01 6.11853898e-01
2.36635670e-01 6.58361614e-01 1.40356600e-01 4.65677269e-02
3.01115990e-01 -2.21481696e-01 2.23055869e-01 -3.85692894e-01
-2.94995662e-02 -1.35706469e-01 -6.35457277e-01 9.43102837e-01
-1.36723077e+00 -1.06280351e+00 2.55223185e-01 2.64809775e+00
5.97930789e-01 4.22119707e-01 9.62842643e-01 1.31734237e-01
9.52512741e-01 6.56083897e-02 -3.19799751e-01 -8.85177076e-01
4.05780017e-01 1.98913589e-01 7.55600750e-01 5.47364652e-01
-4.68523175e-01 4.54358086e-02 7.02720499e+00 3.83231997e-01
-8.03044736e-01 -3.16661805e-01 9.59428906e-01 -3.08052927e-01
-4.77220356e-01 3.97248745e-01 -3.42642307e-01 5.95583022e-01
1.05725157e+00 -9.33493674e-01 -1.00509137e-01 3.14637244e-01
6.21985018e-01 -5.86567283e-01 -8.59236181e-01 -2.59580374e-01
-7.43462801e-01 -1.16014636e+00 -1.15053542e-02 5.45075297e-01
7.28378654e-01 -4.56585586e-01 3.41841698e-01 3.85305494e-01
6.20596528e-01 -7.33361840e-01 1.03335285e+00 3.49088937e-01
8.33162308e-01 -1.17444861e+00 8.24053824e-01 5.64535022e-01
-4.85832989e-01 -4.11228061e-01 -1.16316274e-01 -6.96283519e-01
2.19825376e-03 7.09931552e-01 -6.16406202e-01 5.60517311e-01
3.97801578e-01 -3.49440277e-01 -1.37362167e-01 1.05436575e+00
-6.56621065e-03 7.55303442e-01 -1.79759115e-01 2.29053080e-01
-3.43442708e-02 -3.16422284e-01 1.71319500e-01 9.25523877e-01
3.59043300e-01 1.76094651e-01 -2.23942146e-01 9.39333558e-01
-1.29948407e-01 2.00698122e-01 -5.84635019e-01 1.87631756e-01
7.44768977e-01 5.21008909e-01 -8.07252645e-01 -2.11770788e-01
-4.16529655e-01 2.93973237e-02 3.67050283e-02 5.28809428e-01
-2.82726407e-01 -4.01196986e-01 6.89405560e-01 5.27348399e-01
-2.01786309e-01 2.79394627e-01 -4.82518017e-01 -7.45244324e-01
-2.10737616e-01 -5.22620022e-01 6.81673348e-01 -3.33108902e-01
-7.21620202e-01 -7.19049096e-01 6.11922562e-01 -3.52160633e-01
-4.26822722e-01 -2.05547944e-01 -9.70431209e-01 1.15522897e+00
-1.25290549e+00 -2.01574996e-01 4.78190869e-01 -1.61171302e-01
1.20269004e-02 3.45613956e-01 3.81996512e-01 -2.75293589e-01
-6.66022718e-01 4.85393435e-01 2.36011848e-01 7.58140758e-02
5.81694067e-01 -1.08369458e+00 -3.12526822e-02 1.01705360e+00
-5.52744925e-01 7.65072823e-01 9.62556183e-01 -9.54962552e-01
-3.58148247e-01 -7.89260387e-01 8.51434350e-01 2.93961130e-02
1.39950588e-01 1.88819289e-01 -5.06936133e-01 9.30946946e-01
-2.26967916e-01 -6.41122460e-01 4.10846472e-01 5.92724383e-01
5.36215231e-02 -3.44090521e-01 -1.45656049e+00 7.19302714e-01
4.30893362e-01 -4.02587056e-01 -4.22470629e-01 1.82422832e-01
6.62206411e-01 -3.52174655e-04 -9.61458802e-01 5.29062629e-01
8.83235097e-01 -1.33771718e+00 7.94144332e-01 -5.92172861e-01
3.31875622e-01 6.82534799e-02 -1.81044206e-01 -1.31673527e+00
-4.51511085e-01 -2.01311707e-01 6.38954937e-01 1.17220128e+00
3.49142075e-01 -7.47039497e-01 3.60954851e-01 9.43965495e-01
2.28915557e-01 -7.22331703e-01 -1.09878755e+00 -8.45342934e-01
6.61634803e-01 8.33803117e-02 1.20110226e+00 1.05455732e+00
3.14765692e-01 2.35265382e-02 2.04318032e-01 7.45749027e-02
4.77821499e-01 5.37835777e-01 6.77242756e-01 -1.03163767e+00
-1.14622824e-01 -5.69845915e-01 5.83220646e-02 -1.93679873e-02
8.99324566e-02 -4.01988298e-01 -4.81823742e-01 -1.35000479e+00
6.23877347e-01 -7.76328146e-02 -8.27052444e-02 1.55353367e-01
-4.41979885e-01 -4.59986299e-01 3.63519281e-01 -2.53296271e-02
2.49711111e-01 2.10097313e-01 8.98094237e-01 2.14075670e-01
-2.89279550e-01 3.55037361e-01 -8.77325773e-01 4.70727652e-01
8.47659767e-01 -3.42047781e-01 -2.46911272e-01 1.32554650e-01
1.29863977e-01 6.94911540e-01 4.77029204e-01 -5.99299252e-01
-4.20294642e-01 -8.74226511e-01 2.94416845e-01 -6.05002046e-02
-1.76140234e-01 -3.75984758e-01 4.21224505e-01 9.03439641e-01
-7.82292902e-01 -2.65729446e-02 1.17751814e-01 3.96680683e-01
6.97998628e-02 -2.62069494e-01 7.85353065e-01 -9.07358900e-02
3.65637004e-01 -8.01812559e-02 -4.60939705e-01 9.12224315e-03
8.18313718e-01 -6.11477159e-02 -4.38686609e-01 -4.42168415e-01
-4.05743480e-01 2.39699870e-01 9.06456888e-01 -2.19266295e-01
-2.11557403e-01 -1.05365443e+00 -7.93809533e-01 -3.55018795e-01
-2.09096000e-01 -5.03230572e-01 -1.27188116e-01 5.07061005e-01
1.16155900e-01 4.93733525e-01 -2.28966281e-01 3.90573949e-01
-8.62749577e-01 5.41281223e-01 5.57785094e-01 -2.84530073e-01
-2.21043184e-01 2.04055622e-01 4.60516989e-01 6.47043064e-02
-1.63014427e-01 -1.95883974e-01 -1.26043603e-01 7.84258768e-02
2.46481925e-01 9.65825796e-01 -2.88997620e-01 -4.26573604e-01
-3.11649919e-01 3.78991105e-02 4.40650314e-01 -4.99551922e-01
9.11315680e-01 -3.17243785e-01 -8.32851678e-02 7.03925610e-01
6.23020053e-01 4.04451638e-01 -1.49464214e+00 2.04185694e-01
-1.48038432e-01 -6.05598450e-01 -2.62442529e-01 -7.07357228e-01
-5.47705889e-01 3.20298821e-01 4.68176067e-01 6.00423038e-01
1.00737798e+00 -1.96423784e-01 7.02908263e-03 -2.11008295e-01
4.47462536e-02 -9.98302281e-01 -5.65323830e-01 -2.67348886e-01
5.44805169e-01 -7.68354177e-01 4.64424610e-01 -2.24478990e-01
-2.33071849e-01 5.99730730e-01 2.57895887e-01 -1.09414704e-01
5.33803165e-01 -9.04273465e-02 -2.41801560e-01 -5.68993129e-02
-6.25655472e-01 -1.60431460e-01 8.79113004e-02 4.00566161e-01
4.43757117e-01 7.34622836e-01 -1.12392640e+00 3.29946905e-01
-4.94594812e-01 -1.25730515e-01 7.87826002e-01 4.16407734e-01
-3.39428544e-01 -5.83533645e-01 -7.32706726e-01 8.33719969e-01
-8.39564800e-01 3.07568014e-01 -2.19904169e-01 1.11553383e+00
3.38509172e-01 8.22240472e-01 6.16887033e-01 3.37260902e-01
5.23355186e-01 3.40420067e-01 4.10535723e-01 -6.46326602e-01
-1.91127375e-01 -1.61341578e-01 3.59387994e-01 3.05849463e-02
-3.82235825e-01 -1.11941361e+00 -1.11480224e+00 -5.21362543e-01
-3.79072875e-01 1.88034564e-01 2.71640390e-01 1.18072116e+00
-4.56065275e-02 2.42895454e-01 5.56166530e-01 -4.24993545e-01
-7.69814134e-01 -9.17708933e-01 -9.51183021e-01 3.51246476e-01
2.53056198e-01 -6.58309519e-01 -5.59870780e-01 -5.73398232e-01] | [8.021946907043457, 5.240142822265625] |
630017dc-cf51-4f0e-b85d-c577090aa44f | maskcl-semantic-mask-driven-contrastive | 2305.13600 | null | https://arxiv.org/abs/2305.13600v1 | https://arxiv.org/pdf/2305.13600v1.pdf | MaskCL: Semantic Mask-Driven Contrastive Learning for Unsupervised Person Re-Identification with Clothes Change | This paper considers a novel and challenging problem: unsupervised long-term person re-identification with clothes change. Unfortunately, conventional unsupervised person re-id methods are designed for short-term cases and thus fail to perceive clothes-independent patterns due to simply being driven by RGB prompt. To tackle with such a bottleneck, we propose a semantic mask-driven contrastive learning approach, in which silhouette masks are embedded into contrastive learning framework as the semantic prompts and cross-clothes invariance is learnt from hierarchically semantic neighbor structure by combining both RGB and semantic features in a two-branches network. Since such a challenging re-id task setting is investigated for the first time, we conducted extensive experiments to evaluate state-of-the-art unsupervised short-term person re-id methods on five widely-used clothes-change re-id datasets. Experimental results verify that our approach outperforms the unsupervised re-id competitors by a clear margin, remaining a narrow gap to the supervised baselines. | ['Jun Guo', 'Chun-Guang Li', 'Peng Xu', 'Mingkun Li'] | 2023-05-23 | null | null | null | null | ['person-re-identification', 'unsupervised-long-term-person-re', 'unsupervised-person-re-identification'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 3.71391624e-01 -1.84704781e-01 8.65823701e-02 -5.21627486e-01
-1.56041265e-01 -5.32725990e-01 7.57141113e-01 -1.99073091e-01
-6.92477465e-01 4.09980267e-01 4.98178184e-01 5.20059228e-01
-9.30706114e-02 -2.95218050e-01 -5.58754742e-01 -6.14455760e-01
2.93046206e-01 7.53020227e-01 -4.91037630e-02 -2.36852288e-01
-1.44473925e-01 2.52905339e-01 -1.86843920e+00 -6.47497624e-02
6.21874154e-01 7.59164512e-01 -8.79706442e-02 4.84343022e-01
1.43504620e-01 4.42287475e-01 -3.94285262e-01 -6.46380186e-01
5.74810863e-01 -4.83463258e-01 -8.74593914e-01 3.30691189e-01
1.11839318e+00 -2.93744594e-01 -3.94145846e-01 1.18637431e+00
7.56653070e-01 4.99553859e-01 4.74245250e-01 -1.14280295e+00
-1.01396036e+00 2.19724700e-01 -4.63673592e-01 9.35615525e-02
6.98313177e-01 3.28039616e-01 8.25785398e-01 -9.60161746e-01
4.27758783e-01 1.34647846e+00 1.10569298e+00 1.14189231e+00
-1.55771768e+00 -4.87539440e-01 6.46326900e-01 1.02751151e-01
-1.45726323e+00 -5.74647605e-01 1.14785445e+00 -3.54619294e-01
7.66906500e-01 3.29901874e-01 8.65392029e-01 1.65026855e+00
-5.78830183e-01 7.89536476e-01 1.72879374e+00 -3.32232207e-01
-9.78138149e-02 2.77214527e-01 3.21091950e-01 7.82377362e-01
3.13717455e-01 2.49096289e-01 -5.60582101e-01 2.08758280e-01
5.57402790e-01 3.72949511e-01 -6.23715147e-02 -4.78559256e-01
-1.46961308e+00 2.46454194e-01 6.18704736e-01 2.83294618e-01
-7.68393725e-02 1.21789984e-01 2.60889888e-01 3.03098202e-01
4.79900002e-01 2.73765355e-01 -2.96773374e-01 2.09725007e-01
-9.41068649e-01 4.93038088e-01 7.15262532e-01 8.43023837e-01
5.00752151e-01 -1.99839279e-01 -5.37436783e-01 9.42339540e-01
7.94318840e-02 8.66302371e-01 4.99876440e-01 -4.51457471e-01
2.86607236e-01 5.11407495e-01 3.02614123e-01 -8.96681309e-01
-6.22568965e-01 -5.78368247e-01 -1.03143489e+00 5.67220263e-02
8.73381019e-01 1.65115058e-01 -1.15370739e+00 1.85972631e+00
3.35131347e-01 1.55696109e-01 -1.92570016e-01 1.33074057e+00
9.71320689e-01 -4.17033583e-02 2.08494768e-01 2.31855065e-02
1.55680287e+00 -1.47578037e+00 -5.94161272e-01 -2.81852782e-01
-4.59106900e-02 -4.01597738e-01 1.42293000e+00 2.84146607e-01
-8.99705708e-01 -1.28499663e+00 -9.62321639e-01 -1.08500302e-01
-7.18966722e-01 1.53853163e-01 3.53861898e-01 7.30017066e-01
-1.09402859e+00 7.05570877e-01 -3.46922785e-01 -8.68150234e-01
2.10155889e-01 2.29225785e-01 -4.23642457e-01 -1.10429384e-01
-1.01011193e+00 6.64010227e-01 1.28801122e-01 3.18793505e-01
-8.25949788e-01 -4.47409123e-01 -6.90916121e-01 -1.83874965e-01
5.31403720e-01 -1.02117693e+00 7.13837206e-01 -1.26922834e+00
-1.53312361e+00 1.46113074e+00 -1.94935307e-01 -8.08947384e-02
8.63923311e-01 -2.47772306e-01 -6.19786203e-01 -5.73637001e-02
1.91224366e-01 6.53083920e-01 1.29071963e+00 -1.64230061e+00
-5.19390821e-01 -8.10934246e-01 4.11991030e-02 4.34614182e-01
-7.19614625e-01 -1.80001944e-01 -5.22497177e-01 -1.09000778e+00
-1.47954315e-01 -1.25620818e+00 -1.66492820e-01 -7.33551309e-02
-5.26906371e-01 -4.33446795e-01 4.34146166e-01 -7.75529206e-01
7.65617549e-01 -1.89799964e+00 4.41915303e-01 2.13314906e-01
4.11116362e-01 6.09345511e-02 -1.66918278e-01 -6.70129359e-02
-9.88432914e-02 -1.62885204e-01 1.45406909e-02 -1.09687912e+00
3.22916985e-01 -3.68321873e-02 -8.18147697e-03 6.52724266e-01
-2.38141164e-01 1.12653625e+00 -1.10392213e+00 -4.10423845e-01
4.38091159e-01 4.52890456e-01 -8.24929550e-02 3.98372591e-01
1.30055398e-01 8.22414875e-01 -3.00042215e-04 8.15050900e-01
5.82246304e-01 -2.82429069e-01 -8.92516449e-02 -5.52041113e-01
1.14873819e-01 -2.26021439e-01 -9.32859123e-01 2.04461455e+00
-7.59553090e-02 2.11280003e-01 -1.93095580e-01 -9.18159127e-01
7.66593695e-01 1.42671559e-02 6.00973427e-01 -8.87185693e-01
-2.01026071e-02 -2.60005537e-02 -6.18403375e-01 -3.59706759e-01
4.20114428e-01 -4.97047901e-02 -3.17200512e-01 1.89360127e-01
8.39347690e-02 4.66905892e-01 2.14663707e-02 -9.60547850e-02
6.44779563e-01 6.20977044e-01 3.69531326e-02 -3.68020624e-01
5.56596279e-01 -2.08798856e-01 3.66103858e-01 1.10989738e+00
-7.28000641e-01 8.70615900e-01 -4.26469952e-01 -7.46450603e-01
-9.23146307e-01 -1.16167331e+00 1.21214874e-01 1.34687734e+00
7.91372895e-01 -3.32367718e-01 -8.60952675e-01 -1.14173305e+00
1.32449359e-01 2.34599486e-01 -9.49981809e-01 4.81391251e-02
-4.48090494e-01 -6.69764519e-01 4.67161089e-01 6.75435305e-01
1.02840924e+00 -8.51539433e-01 -2.50780195e-01 -5.59644289e-02
-3.91929328e-01 -1.36585736e+00 -8.07081640e-01 -1.49148017e-01
-4.14425582e-01 -9.35094476e-01 -1.45842230e+00 -9.65014338e-01
7.03408122e-01 5.28215706e-01 1.16362309e+00 2.42759399e-02
-2.99478710e-01 1.04863739e+00 -2.79332757e-01 -3.85850668e-02
1.19903192e-01 3.99256535e-02 4.78216708e-01 4.52496558e-01
8.12212527e-01 -4.07941580e-01 -9.87676859e-01 4.88196760e-01
-1.59997955e-01 -2.41746269e-02 3.59189361e-01 8.09949636e-01
5.33751309e-01 -9.42097083e-02 4.10278767e-01 -5.06062865e-01
3.77975464e-01 1.30507424e-02 -2.49890581e-01 5.99978030e-01
-9.81219172e-01 -1.44920632e-01 6.32844210e-01 -6.34797692e-01
-1.15592158e+00 3.65144521e-01 9.97927189e-02 -3.46030533e-01
-6.00541472e-01 -2.89954573e-01 -2.89921165e-01 -2.32269734e-01
5.24410784e-01 3.33435804e-01 -8.49164426e-02 -7.51346707e-01
5.61880052e-01 5.33555567e-01 9.66479242e-01 -5.39612353e-01
1.07967401e+00 8.76376092e-01 -1.76687106e-01 -4.53507215e-01
-1.04780972e+00 -9.43118870e-01 -1.05159926e+00 -6.10478461e-01
1.23831880e+00 -1.26519883e+00 -8.79300475e-01 8.25696945e-01
-8.25219750e-01 -2.57145852e-01 -5.75866640e-01 4.96745370e-02
-3.35828602e-01 5.88334620e-01 -4.64922637e-01 -6.82693124e-01
-3.63617301e-01 -7.53209770e-01 1.26122379e+00 2.83837736e-01
-3.16276729e-01 -9.77925658e-01 1.44282252e-01 6.11410499e-01
3.40382427e-01 3.27877074e-01 1.03948601e-01 -4.82100993e-01
-9.54014361e-02 -7.85551891e-02 -4.15323645e-01 2.70665765e-01
4.44728941e-01 -1.01544392e+00 -1.39192939e+00 -6.38543010e-01
-2.55001992e-01 -3.80976349e-01 1.21875226e+00 -1.03337787e-01
1.23643720e+00 -3.08796853e-01 -4.02664542e-01 8.79758418e-01
1.35033166e+00 -5.40445149e-01 1.89621776e-01 4.08566803e-01
1.46258914e+00 8.72825205e-01 2.38991171e-01 3.03073913e-01
9.09577250e-01 9.70493019e-01 -3.78573611e-02 -5.73044240e-01
-7.52205431e-01 -4.67303842e-01 3.50470543e-01 5.01496911e-01
-6.50980115e-01 -6.02936721e-04 -5.87352037e-01 6.67785227e-01
-1.93040097e+00 -9.94150877e-01 1.41296715e-01 2.23196244e+00
6.46752775e-01 -9.81703848e-02 7.87761271e-01 3.96401025e-02
8.27013552e-01 2.21531719e-01 -7.29208350e-01 3.73675019e-01
-4.22811478e-01 -8.51282626e-02 7.26876378e-01 3.27727824e-01
-1.42280090e+00 1.03194654e+00 5.72404480e+00 5.39005280e-01
-7.34171808e-01 3.16998959e-01 3.69078994e-01 -8.46935287e-02
2.87644635e-03 -3.27081770e-01 -8.23931754e-01 5.53219676e-01
3.49614143e-01 4.57813442e-01 8.16498518e-01 6.38973057e-01
-7.22884238e-02 2.34034538e-01 -1.28412414e+00 1.84836602e+00
6.24609828e-01 -6.94620967e-01 -5.34072779e-02 -2.01867729e-01
1.00331783e+00 -4.44995105e-01 1.89298376e-01 1.82072744e-01
2.43773714e-01 -8.63618731e-01 9.14514780e-01 8.05232286e-01
9.60804403e-01 -3.77338171e-01 4.60743576e-01 -2.26242438e-01
-1.49482632e+00 -2.36240372e-01 -1.19098999e-01 -4.92416434e-02
1.13882929e-01 2.83951104e-01 -2.68903136e-01 4.92024332e-01
1.32718599e+00 1.16340530e+00 -9.20000494e-01 5.34947574e-01
-4.58678976e-02 2.91641831e-01 -1.08608857e-01 2.16830522e-01
-1.63580358e-01 7.21833110e-02 6.64667130e-01 1.40069258e+00
-5.12024723e-02 -1.46437019e-01 5.65385222e-01 8.36838126e-01
-3.58177684e-02 -1.73855633e-01 -3.70154262e-01 2.31231287e-01
-5.98971657e-02 1.22959101e+00 -8.02373528e-01 -4.81055945e-01
-3.72854382e-01 1.97855318e+00 2.45508522e-01 7.42461264e-01
-7.85382807e-01 -6.50852546e-02 5.30913949e-01 2.20939264e-01
2.38891080e-01 -2.11561516e-01 -2.21399277e-01 -1.49568164e+00
2.46359184e-01 -7.05089569e-01 5.56408823e-01 -5.73975742e-01
-2.03237295e+00 4.75797176e-01 -2.67549721e-03 -1.06312191e+00
2.35215724e-02 -6.75968528e-01 -1.21903099e-01 6.02978647e-01
-1.72954583e+00 -1.90611780e+00 -1.03089023e+00 1.09190118e+00
5.76405644e-01 -3.12769145e-01 8.12373936e-01 2.99233168e-01
-5.95590711e-01 1.08271706e+00 -5.29970117e-02 2.49700218e-01
1.06492424e+00 -1.59776938e+00 5.25627792e-01 9.05808091e-01
1.35000303e-01 7.32713878e-01 6.59711421e-01 -8.95847976e-01
-1.29348958e+00 -1.11503828e+00 7.75115728e-01 -8.14625382e-01
2.96557546e-01 -8.26358616e-01 -5.13110161e-01 6.57180250e-01
7.27571547e-02 2.60339022e-01 5.57029366e-01 3.72362673e-01
-6.34305954e-01 -5.70879281e-01 -1.10559916e+00 5.15613079e-01
1.88928568e+00 -6.92309856e-01 -5.58478117e-01 4.24555451e-01
4.94737417e-01 -5.08611947e-02 -8.62547874e-01 4.25139725e-01
8.42832267e-01 -7.78144479e-01 1.44386518e+00 -5.19616127e-01
-1.97307661e-01 -4.08209234e-01 3.88660803e-02 -1.18580782e+00
-6.54223561e-01 -7.56291926e-01 -1.15767784e-01 1.44684196e+00
-2.38352776e-01 -5.50764799e-01 7.12518632e-01 5.78742683e-01
4.66710389e-01 -8.92079398e-02 -6.09451890e-01 -1.09575260e+00
-3.30018938e-01 -9.94098466e-03 7.24940300e-01 9.82560933e-01
-2.67673999e-01 4.39462006e-01 -9.49024200e-01 -7.43117463e-03
1.33242416e+00 1.83179215e-01 9.81411278e-01 -1.36003041e+00
-3.73846322e-01 -4.20068920e-01 -3.28194737e-01 -1.38994229e+00
3.15606087e-01 -9.58572924e-01 1.80317968e-01 -1.28211653e+00
7.49477684e-01 -4.99865383e-01 -6.13353133e-01 3.18435073e-01
-6.28765106e-01 6.00330234e-01 3.93670946e-01 4.42706585e-01
-1.10799730e+00 5.93393624e-01 9.93366241e-01 -6.12798929e-01
-1.37280375e-01 -9.64411944e-02 -6.75643981e-01 7.12108135e-01
5.42040706e-01 -1.08308002e-01 -3.90834630e-01 -4.43074584e-01
-6.68171719e-02 -6.39463603e-01 1.18005300e+00 -1.23585367e+00
4.05543923e-01 1.57773405e-01 8.15407038e-01 -3.53406966e-01
3.92649204e-01 -8.60948682e-01 9.32102092e-04 1.18571222e-01
-4.54071999e-01 1.19575687e-01 -2.17691928e-01 8.07993829e-01
1.91981301e-01 4.05988812e-01 5.93632519e-01 -3.90555859e-01
-1.09615993e+00 5.21485388e-01 1.30751833e-01 4.99632619e-02
6.59353137e-01 -6.03763938e-01 -3.12742852e-02 -2.26210058e-01
-1.06369138e+00 2.37218570e-02 7.01494396e-01 6.92339420e-01
5.49743891e-01 -1.41764009e+00 -7.09268093e-01 1.94886476e-01
5.14468849e-01 -3.80562276e-01 4.48201984e-01 6.36130154e-01
1.91034973e-01 1.51242301e-01 -3.66044879e-01 -6.66857779e-01
-1.36242032e+00 9.61696148e-01 5.32117903e-01 -2.83063591e-01
-9.58045483e-01 8.86395872e-01 2.79087007e-01 -6.07165337e-01
6.40739262e-01 1.49825821e-02 -2.39830464e-01 -4.38378416e-02
5.46075344e-01 4.98137653e-01 -2.56879449e-01 -1.05104470e+00
-5.31890810e-01 1.12272966e+00 1.28694370e-01 -9.47956294e-02
9.19910848e-01 -7.64729500e-01 1.02333084e-01 5.40596426e-01
1.04157996e+00 -1.46789461e-01 -1.47688806e+00 -6.59057617e-01
-8.10600817e-02 -4.91203278e-01 -4.10487354e-01 -1.01161599e+00
-9.94518757e-01 5.31542122e-01 1.24458182e+00 -4.57376055e-02
1.13299763e+00 6.87371939e-02 1.06575787e+00 3.34409684e-01
4.08962935e-01 -1.50809598e+00 5.24984539e-01 1.31048188e-01
7.45703578e-01 -1.71104944e+00 -1.21625746e-02 -1.90532029e-01
-5.85471213e-01 7.80028880e-01 7.52257884e-01 -1.55459136e-01
5.01181364e-01 -1.99276045e-01 9.31896418e-02 -1.69878110e-01
1.99951306e-01 -7.61174738e-01 7.06610382e-01 8.83305848e-01
-1.15246154e-01 2.05306754e-01 1.92778125e-01 7.53452778e-01
-1.93243146e-01 -1.54320523e-01 -3.50453526e-01 5.24327815e-01
2.14948617e-02 -8.87417555e-01 -4.47038591e-01 -6.10532425e-02
-2.55219311e-01 1.60816506e-01 -6.71292841e-01 5.78832030e-01
5.15548646e-01 9.79178071e-01 -5.82705550e-02 -5.45451701e-01
4.35897231e-01 -1.58404157e-01 9.04716551e-01 -1.57914728e-01
-6.89624190e-01 -1.66353807e-01 -3.71773466e-02 -7.95450568e-01
-9.82252181e-01 -6.51548088e-01 -7.28672683e-01 -2.03237921e-01
1.18823208e-01 -3.16993743e-01 2.25759447e-01 1.06436503e+00
1.13518693e-01 4.49118704e-01 4.54961032e-01 -1.09453607e+00
-5.75818837e-01 -8.63987029e-01 -3.32704991e-01 1.39962518e+00
6.38620019e-01 -7.41141498e-01 -2.54606575e-01 3.69746804e-01] | [14.716933250427246, 0.9990857839584351] |
013c206a-c9f9-4a7b-8648-9505725bd92f | a-statistical-model-for-melody-reduction | 2105.05385 | null | https://arxiv.org/abs/2105.05385v1 | https://arxiv.org/pdf/2105.05385v1.pdf | A Statistical Model for Melody Reduction | A commonly-cited reason for the poor performance of automatic chord estimation (ACE) systems within music information retrieval (MIR) is that non-chord tones (i.e., notes outside the supporting harmony) contribute to error during the labeling process. Despite the prevalence of machine learning approaches in MIR, there are cases where alternative approaches provide a simpler alternative while allowing for insights into musicological practices. In this project, we present a statistical model for predicting chord tones based on music theory rules. Our model is currently focused on predicting chord tones in classical music, since composition in this style is highly constrained, theoretically making the placement of chord tones highly predictable. Indeed, music theorists have labeling systems for every variety of non-chord tone, primarily classified by the note's metric position and intervals of approach and departure. Using metric position, duration, and melodic intervals as predictors, we build a statistical model for predicting chord tones using the TAVERN dataset. While our probabilistic approach is similar to other efforts in the domain of automatic harmonic analysis, our focus is on melodic reduction rather than predicting harmony. However, we hope to pursue applications for ACE in the future. Finally, we implement our melody reduction model using an existing symbolic visualization tool, to assist with melody reduction and non-chord tone identification for computational musicology researchers and music theorists. | ['Claire Arthur', 'Tianxue Hu'] | 2021-05-12 | null | null | null | null | ['music-information-retrieval'] | ['music'] | [ 1.47285491e-01 -9.45879295e-02 -1.19298801e-01 1.12314142e-01
-8.00364614e-01 -1.10413802e+00 2.96992868e-01 2.84321666e-01
-1.45820394e-01 2.43039489e-01 5.97444654e-01 -2.40582138e-01
-6.07930899e-01 -6.13991022e-01 1.20248109e-01 -2.69444346e-01
-3.07640694e-02 5.48768044e-01 -2.82116383e-02 -5.35652101e-01
8.49734783e-01 2.75880814e-01 -1.57784307e+00 3.56831074e-01
3.05931926e-01 7.28678644e-01 -3.02776471e-02 7.87306130e-01
-3.81701104e-02 9.08956826e-01 -9.45241809e-01 -6.89021870e-02
2.97150314e-01 -9.41346347e-01 -8.31203103e-01 -3.85240763e-01
1.46408558e-01 1.76572323e-01 1.64173186e-01 7.17770219e-01
4.76327688e-01 1.93246171e-01 7.21959770e-01 -8.48029613e-01
-2.29206547e-01 1.40089190e+00 -2.03013584e-01 -1.23432644e-01
5.65184593e-01 -3.60721231e-01 1.52796721e+00 -6.32827103e-01
6.21573627e-01 8.20377886e-01 1.13872004e+00 1.60529479e-01
-1.43652499e+00 -6.93297029e-01 -4.33113515e-01 3.86988252e-01
-1.74798739e+00 -3.72018099e-01 1.17613685e+00 -8.06330621e-01
6.66975617e-01 7.69038558e-01 1.04676139e+00 4.07481909e-01
-1.00787036e-01 5.26075661e-01 8.69958937e-01 -8.93465579e-01
1.95916086e-01 2.30401754e-02 4.04500663e-02 -2.47546919e-02
-5.93693972e-01 1.22135952e-01 -7.51853466e-01 -3.94361973e-01
6.63790047e-01 -5.86706579e-01 -1.12971500e-01 -8.88058841e-02
-1.17043293e+00 8.24653804e-01 -1.08416140e-01 6.90029323e-01
-1.98986515e-01 -9.84086394e-02 5.77217937e-01 2.77708650e-01
3.32324095e-02 1.16230285e+00 -4.33911562e-01 -7.63452530e-01
-1.66209245e+00 8.63020003e-01 9.93097007e-01 4.55376297e-01
3.94175828e-01 2.61470735e-01 -7.09268898e-02 1.00144577e+00
1.38573736e-01 -1.56979620e-01 6.26138449e-01 -1.30893111e+00
-2.24520162e-01 2.66128033e-01 -5.63721023e-02 -1.11717665e+00
-3.49766165e-01 -3.64116699e-01 -2.61222303e-01 4.44811672e-01
6.75849855e-01 4.64159429e-01 -2.17146695e-01 1.61610508e+00
-1.38623238e-01 -8.15426409e-02 -3.20017368e-01 8.10544193e-01
4.76739496e-01 4.80168015e-01 -1.78951368e-01 -5.71838021e-01
1.21494341e+00 -4.43995744e-01 -7.76373446e-01 4.18114454e-01
4.74180579e-01 -1.39891088e+00 1.43553090e+00 1.11014116e+00
-1.23448336e+00 -6.07307255e-01 -1.07117105e+00 -1.74643382e-01
4.75878902e-02 1.99475996e-02 4.13139194e-01 5.60478330e-01
-3.76029819e-01 1.17833114e+00 -5.33878326e-01 1.47148147e-01
-1.70034304e-01 1.25931397e-01 1.69228956e-01 9.11459208e-01
-1.09091592e+00 8.49454343e-01 3.71922046e-01 -2.34581798e-01
-2.17789978e-01 -1.02174067e+00 -4.58850443e-01 -5.89576103e-02
8.81860256e-02 3.70065011e-02 1.62229121e+00 -8.31307054e-01
-1.51622975e+00 8.38778317e-01 2.34534696e-01 -3.47016782e-01
2.14355960e-01 -2.78736018e-02 -4.36868280e-01 -1.58510894e-01
7.17650503e-02 2.81361461e-01 5.35769999e-01 -9.06921268e-01
-4.29803193e-01 -2.40891948e-01 -1.31814033e-01 2.29947388e-01
-1.11856759e-01 2.32864767e-01 -6.28579035e-02 -1.06398463e+00
4.06653464e-01 -1.20019889e+00 1.00021653e-01 -4.80080068e-01
-4.39382970e-01 -3.81966174e-01 4.47581470e-01 -9.47867751e-01
2.08775735e+00 -2.31783080e+00 1.79960683e-01 4.90440398e-01
-1.75898895e-01 -8.26666504e-02 3.66956055e-01 5.77946484e-01
-1.99654981e-01 1.05208904e-01 -2.47871459e-01 7.32083023e-02
2.88247079e-01 -8.07492211e-02 -7.93156624e-01 4.08277288e-02
-4.15754616e-01 5.59448123e-01 -6.26567125e-01 -4.17539358e-01
2.07004622e-01 4.17249829e-01 -6.15836322e-01 -1.68824673e-01
-1.89511642e-01 4.38054532e-01 1.64292380e-01 6.06472611e-01
9.85246301e-02 4.83067244e-01 3.63244087e-01 -3.07539284e-01
-6.97270930e-01 1.05785739e+00 -1.50551319e+00 1.79018843e+00
-2.37542480e-01 7.47512341e-01 -3.75656664e-01 -4.64930087e-01
1.21197522e+00 4.39455062e-01 6.33275032e-01 -3.24473292e-01
-8.45972672e-02 3.47789317e-01 6.50292218e-01 -3.60077131e-03
1.03634727e+00 -2.96098620e-01 -2.76838750e-01 6.28328621e-01
-2.21425459e-01 -8.08862686e-01 2.88963705e-01 -1.53764650e-01
7.55584717e-01 5.97719669e-01 7.17273533e-01 -2.57782757e-01
2.45586470e-01 3.53124350e-01 5.61782122e-01 4.08447593e-01
-1.58886565e-03 8.89272749e-01 4.20284688e-01 -4.76345837e-01
-1.22675157e+00 -1.13529730e+00 -4.61030066e-01 1.29121459e+00
-3.70588601e-01 -1.40652895e+00 -6.51101947e-01 1.54127389e-01
-1.89699069e-01 9.64610457e-01 -3.81845146e-01 8.07033330e-02
-7.56625175e-01 -1.91436037e-01 9.74018335e-01 2.66961396e-01
-2.95626223e-01 -1.56615400e+00 -6.40374780e-01 6.03494346e-01
-3.31252784e-01 -4.20233488e-01 -4.41331238e-01 3.29487354e-01
-6.15864336e-01 -1.10227513e+00 -1.72455430e-01 -5.92807591e-01
-5.27247727e-01 -1.51185527e-01 1.22161710e+00 6.45888085e-03
-5.15778363e-01 1.82658359e-01 -4.61752266e-01 -8.78105998e-01
-5.73348403e-01 1.74045265e-01 4.96216379e-02 -6.04655325e-01
3.99813682e-01 -7.91162670e-01 -4.33419317e-01 2.56588191e-01
-7.03490257e-01 2.51884460e-02 -2.36660078e-01 3.81079435e-01
9.43974078e-01 2.13033751e-01 5.49779952e-01 -6.98005974e-01
8.26683879e-01 -1.23744920e-01 -2.62662351e-01 -1.45051360e-01
-7.30714679e-01 -4.54544187e-01 5.16031861e-01 -5.68977475e-01
-4.71807182e-01 -5.93839474e-02 -2.36223504e-01 -1.77908227e-01
4.23706546e-02 7.19449699e-01 1.39716208e-01 2.16774791e-01
1.08108401e+00 1.09180838e-01 -1.83696434e-01 -6.14305139e-01
2.37452120e-01 6.47410512e-01 7.49103725e-01 -6.38316154e-01
5.39862990e-01 3.89231183e-02 -1.03972241e-01 -8.09990764e-01
-8.36403012e-01 -3.83305937e-01 -7.09530056e-01 -3.21934253e-01
5.59188545e-01 -5.96309304e-01 -7.78560877e-01 -1.15300104e-01
-7.35668302e-01 -8.43385309e-02 -7.54601061e-01 5.96844196e-01
-1.06538534e+00 3.17955226e-01 -6.66693866e-01 -1.10072219e+00
-3.83015543e-01 -7.77429879e-01 5.66452324e-01 2.16620490e-02
-1.48986971e+00 -7.29006469e-01 5.60054123e-01 1.86866865e-01
2.14129835e-01 2.97749966e-01 1.15431201e+00 -5.72042048e-01
-1.15333624e-01 -7.07171932e-02 6.24233484e-01 2.42591351e-01
2.78732747e-01 3.28741848e-01 -8.96654487e-01 2.32000783e-01
1.28943220e-01 -2.12521866e-01 4.27482963e-01 4.13805544e-01
1.07799733e+00 -1.87699139e-01 2.65423864e-01 4.77867335e-01
9.54524577e-01 3.43786180e-01 6.30216479e-01 6.06044233e-01
5.01461267e-01 8.37738097e-01 9.43736374e-01 6.21279299e-01
1.60540238e-01 1.19216537e+00 -1.12158492e-01 4.63814139e-01
-1.63253486e-01 -4.45020854e-01 3.75859320e-01 1.28763402e+00
-4.05364215e-01 4.06960607e-01 -9.68647003e-01 5.88035643e-01
-1.79473960e+00 -1.36383212e+00 -4.96887594e-01 2.31507230e+00
1.19518554e+00 2.42352769e-01 5.94312012e-01 1.05816412e+00
3.64002466e-01 -3.70820612e-02 -9.07580256e-02 -7.12846398e-01
-1.53763652e-01 6.87350869e-01 6.46758601e-02 4.35900748e-01
-9.13193166e-01 9.66130137e-01 6.90577698e+00 9.86982644e-01
-9.33663785e-01 -3.22416097e-01 -8.91880691e-02 -3.38032544e-01
-4.61860716e-01 3.85977060e-01 -4.80142653e-01 3.10788542e-01
8.21986377e-01 -3.04962128e-01 8.67654443e-01 7.87539184e-01
4.62603688e-01 1.37287989e-01 -1.03630352e+00 9.47373688e-01
-4.16412354e-02 -1.40551829e+00 -1.16538309e-01 -7.89664090e-02
4.63524789e-01 -4.58624512e-01 -2.49076020e-02 2.97069967e-01
-2.33873963e-01 -1.09415042e+00 1.25751936e+00 6.23199105e-01
7.69160628e-01 -1.02323186e+00 1.30052835e-01 2.04373211e-01
-1.18866408e+00 8.62621889e-02 -1.60667524e-01 -5.16267061e-01
1.38856545e-01 1.98518202e-01 -8.63804042e-01 3.46499741e-01
6.05056286e-01 6.16974413e-01 -4.22145009e-01 1.06007230e+00
-1.81473479e-01 1.11380982e+00 -2.14459762e-01 4.53793108e-01
-1.80182546e-01 -4.03172344e-01 7.87664592e-01 1.19736969e+00
5.62052011e-01 9.34057683e-02 1.95615038e-01 9.35315192e-01
6.75231695e-01 5.92592239e-01 -2.55696297e-01 -2.38197133e-01
7.79621005e-01 1.17379189e+00 -7.78179526e-01 4.66132276e-02
1.34681642e-01 4.21047866e-01 -1.54547259e-01 -1.36372566e-01
-5.48963487e-01 -4.04079199e-01 6.14723563e-01 3.93366694e-01
8.16318616e-02 -3.56965721e-01 -8.93433630e-01 -6.19783044e-01
-3.02011967e-01 -1.28795803e+00 4.51087087e-01 -8.51206779e-01
-1.06407976e+00 5.03691673e-01 -2.25617945e-01 -1.48204947e+00
-8.64343464e-01 -3.01254600e-01 -5.31196475e-01 9.45720434e-01
-7.12806821e-01 -9.29793239e-01 2.52087235e-01 5.46603680e-01
4.74862337e-01 -2.67902255e-01 1.56970072e+00 1.46033406e-01
2.84244329e-01 4.91727859e-01 -2.23860949e-01 2.89927069e-02
1.04094303e+00 -1.63981664e+00 2.34089166e-01 1.11216046e-01
1.14296854e+00 6.17913902e-01 1.08338511e+00 -5.61018527e-01
-6.70108020e-01 -3.92110974e-01 9.58376229e-01 -5.93469620e-01
9.38684642e-01 -8.73631686e-02 -9.30592716e-01 3.57153028e-01
3.14766653e-02 -9.13688064e-01 1.47899163e+00 7.41390407e-01
-3.85611445e-01 1.34289443e-01 -4.83576447e-01 7.14752138e-01
6.97324455e-01 -9.91349220e-01 -1.08635724e+00 9.00031440e-03
3.56877655e-01 -2.39160463e-01 -1.12104154e+00 3.99570793e-01
1.00122619e+00 -9.86445963e-01 8.61626923e-01 -2.46719450e-01
2.89476544e-01 -8.16909313e-01 -2.01723933e-01 -1.21152782e+00
-5.69345415e-01 -1.01068652e+00 -3.38385850e-02 1.28995538e+00
2.86299109e-01 3.23570341e-01 8.07828605e-01 1.07015863e-01
-2.99266905e-01 -2.08104491e-01 -7.50206828e-01 -5.76329350e-01
1.20814264e-01 -1.12741816e+00 4.11213130e-01 1.13802254e+00
5.06915152e-01 3.58489007e-01 -5.36637664e-01 -4.14494306e-01
4.11777109e-01 5.36274970e-01 7.44117975e-01 -1.58816612e+00
-8.58919561e-01 -8.89347672e-01 -4.37661827e-01 -3.52015108e-01
-1.30774498e-01 -1.04874778e+00 -1.15669139e-01 -9.49841321e-01
-1.46197766e-01 -4.31800038e-01 -6.60606682e-01 4.47571903e-01
2.92643279e-01 8.53709579e-01 6.34152174e-01 6.27946556e-01
-2.47867103e-03 8.67784619e-02 9.52226341e-01 8.21614116e-02
-8.25319409e-01 3.19691986e-01 -6.37846828e-01 1.17564714e+00
9.25823867e-01 -6.39869332e-01 -2.45396778e-01 1.97955057e-01
5.54093421e-01 1.87241942e-01 1.37584180e-01 -1.07039142e+00
1.96228355e-01 -8.83934349e-02 2.64824659e-01 -9.44124877e-01
3.48028392e-01 -3.70694578e-01 6.93355203e-01 1.69591665e-01
-6.17148757e-01 1.28517807e-01 2.26367414e-01 -2.05516264e-01
-3.97983134e-01 -5.21149099e-01 4.12692755e-01 -1.06097115e-02
-3.04102719e-01 -3.30932468e-01 -6.32554293e-01 -7.09994584e-02
4.41788316e-01 -5.78560591e-01 3.22745889e-01 -4.48038042e-01
-1.30799389e+00 -5.79696476e-01 3.94104719e-01 5.15864253e-01
2.32475147e-01 -1.50355983e+00 -7.90812969e-01 -7.97183998e-03
3.15623850e-01 -6.24517500e-01 8.32534432e-02 6.29434049e-01
-5.59308469e-01 1.41172081e-01 -2.46707261e-01 -3.52609992e-01
-1.47409976e+00 2.54987091e-01 2.09554471e-02 -1.59152955e-01
-5.77352822e-01 5.09449303e-01 -1.52744174e-01 -4.24652845e-01
2.59894431e-01 -5.38346022e-02 -4.56077367e-01 4.78466690e-01
6.05176866e-01 4.73436832e-01 -5.29498719e-02 -6.86721623e-01
-1.77364036e-01 5.59014976e-01 1.22548662e-01 -6.94934070e-01
1.16611171e+00 -2.02311613e-02 -2.83858657e-01 1.37986648e+00
4.94876653e-01 9.02289510e-01 -6.83556974e-01 8.79415870e-02
5.03155649e-01 -3.41544300e-01 -1.24343999e-01 -1.11015427e+00
-6.92033097e-02 6.25953376e-01 3.39916527e-01 5.67483962e-01
1.00240350e+00 -2.49884337e-01 5.01944125e-01 1.94124445e-01
3.22782874e-01 -1.36831164e+00 -2.97075808e-01 7.63699770e-01
1.01243293e+00 -4.36925828e-01 1.15180366e-01 -1.09098487e-01
-6.96342766e-01 1.33642817e+00 1.26097992e-01 -3.19230616e-01
6.74779952e-01 3.22793573e-01 3.64211857e-01 1.03221219e-02
-5.16128957e-01 -5.08537777e-02 6.41475081e-01 2.50863254e-01
1.16395640e+00 3.74055505e-01 -8.02708268e-01 1.11775875e+00
-1.56946182e+00 -2.10536659e-01 6.00877047e-01 4.48421240e-01
-4.63083148e-01 -1.64085758e+00 -6.29098773e-01 3.57624814e-02
-8.31723750e-01 -5.19159317e-01 -7.35164106e-01 7.30230272e-01
5.19635856e-01 9.05113220e-01 1.05622157e-01 -6.10182226e-01
2.91193694e-01 2.80947089e-01 7.67506778e-01 -7.06301391e-01
-1.10170746e+00 7.32911706e-01 9.24523994e-02 -3.32166880e-01
-1.94761068e-01 -8.43791306e-01 -1.47155523e+00 -2.21190393e-01
-1.66661933e-01 6.14611924e-01 7.04540849e-01 6.64966702e-01
-1.21021323e-01 5.54657698e-01 4.63391989e-01 -7.54944324e-01
-4.16441828e-01 -1.09696531e+00 -1.10297191e+00 3.36914301e-01
-2.46602222e-01 -3.31020206e-01 -2.65360951e-01 1.75357237e-01] | [15.953985214233398, 5.343047618865967] |
d2a1df8f-80a2-4f6c-a4fb-815698aede94 | image-and-text-fusion-for-upmc-food-101-using | null | null | https://ieeexplore.ieee.org/abstract/document/9290622 | http://artelab.dista.uninsubria.it/res/research/papers/2020/2020-IVCNZ-Gallo-Food101.pdf | Image and Text fusion for UPMC Food-101 \\using BERT and CNNs | The modern digital world is becoming more and more multimodal. Looking on the internet, images are often associated with the text, so classification problems with these two modalities are very common.
In this paper, we examine multimodal classification using textual information and visual representations of the same concept.
We investigate two main basic methods to perform multimodal fusion and adapt them with stacking techniques to better handle this type of problem.
Here, we use UPMC Food-101, which is a difficult and noisy multimodal dataset that well represents this category of multimodal problems.
Our results show that the proposed early fusion technique combined with a stacking-based approach exceeds the state of the art on the dataset used. | ['and Riccardo La Grassa', 'Nicola Landro', 'Gianmarco Ria', 'Ignazio Gallo'] | 2020-12-17 | null | null | null | null | ['multimodal-text-and-image-classification'] | ['methodology'] | [ 1.19094633e-01 -5.80281973e-01 7.83525407e-03 -2.31248319e-01
-8.49756062e-01 -5.76429129e-01 1.08651102e+00 6.15931928e-01
-5.72767556e-01 8.94899189e-01 2.99967706e-01 -5.16037419e-02
2.56284848e-02 -4.89684880e-01 -3.14706802e-01 -6.36085033e-01
5.07764332e-03 4.25619781e-01 2.03791946e-01 -6.38606250e-01
4.07896221e-01 2.25095138e-01 -2.17360044e+00 1.00121081e+00
6.96650326e-01 1.04411638e+00 1.61708400e-01 8.23821902e-01
-5.31045496e-01 6.06523812e-01 -6.36336029e-01 -6.71881199e-01
-1.72874346e-01 -2.29477495e-01 -8.34250808e-01 2.37646848e-01
5.15076578e-01 5.12119383e-02 3.56139094e-02 9.02046859e-01
6.43028915e-01 1.90344825e-01 7.97340870e-01 -1.41547596e+00
-1.42550424e-01 5.39202929e-01 -8.06153178e-01 6.10948242e-02
1.07495165e+00 -4.08230096e-01 5.44723988e-01 -9.50588524e-01
7.45997488e-01 1.61624050e+00 5.34546018e-01 1.25389442e-01
-9.44736898e-01 -3.22527230e-01 -1.26649931e-01 4.35483485e-01
-1.43139040e+00 -2.99446493e-01 7.66937137e-01 -3.70318621e-01
6.11861050e-01 5.06670475e-01 4.28978235e-01 1.18506265e+00
1.69077203e-01 8.05162668e-01 1.33716536e+00 -7.74035513e-01
-3.34389089e-03 3.86613607e-01 5.65471947e-01 5.50065339e-01
1.44498169e-01 -3.84763777e-01 -5.67073286e-01 -2.41847873e-01
4.38088290e-02 6.32944331e-02 -1.97060645e-01 -1.14338271e-01
-1.33336771e+00 6.48031235e-01 6.13523014e-02 7.80959845e-01
-1.77853793e-01 -5.19019254e-02 5.29124498e-01 2.84267575e-01
1.47832274e-01 -7.75428042e-02 1.06596604e-01 -1.10229671e-01
-8.18792105e-01 1.85320586e-01 8.06460619e-01 8.98341060e-01
5.98119199e-01 -3.86315525e-01 4.26308550e-02 1.03373945e+00
6.95467412e-01 5.70306182e-01 6.77164257e-01 -5.96620381e-01
7.70430028e-01 8.61265600e-01 -7.01313838e-02 -1.14659333e+00
-5.67345917e-01 2.20605254e-01 -7.44047940e-01 2.20130846e-01
4.37301546e-01 -5.36343083e-02 -1.13334727e+00 1.21317708e+00
9.02609304e-02 -3.57026964e-01 3.23654354e-01 7.60509193e-01
1.37273872e+00 9.45791245e-01 2.94748276e-01 -2.10031033e-01
1.59967077e+00 -7.53443003e-01 -1.10030448e+00 1.74137890e-01
5.82986712e-01 -1.37417173e+00 4.15763885e-01 6.01941586e-01
-9.26433265e-01 -5.98492503e-01 -1.14520860e+00 -3.94499563e-02
-1.19761145e+00 1.97721958e-01 4.38828766e-01 8.88698220e-01
-9.57921147e-01 3.70712519e-01 -2.49033123e-01 -9.35359418e-01
-5.34008145e-02 3.03963572e-01 -8.86369646e-01 -4.38412696e-01
-1.11651540e+00 1.18762410e+00 7.59808183e-01 2.40013748e-02
-1.70940623e-01 7.76259229e-02 -8.81539583e-01 -3.64506513e-01
2.27659225e-01 -3.99490058e-01 9.52082276e-01 -8.43265772e-01
-1.20425391e+00 6.83818996e-01 -4.69315946e-02 2.18503503e-03
4.23906684e-01 7.32949525e-02 -7.68995583e-01 2.66881973e-01
-3.34064662e-01 7.26585686e-01 8.32717240e-01 -1.90346706e+00
-7.43206501e-01 -3.76094699e-01 -1.06927946e-01 3.62924159e-01
-2.91883975e-01 1.00807317e-01 -4.44109142e-01 -4.94755656e-01
3.42127047e-02 -7.14584410e-01 1.51897773e-01 -3.80595207e-01
-2.49667123e-01 -1.59754977e-01 1.19272482e+00 -8.45448852e-01
1.27259707e+00 -2.11911416e+00 4.92616236e-01 4.86785084e-01
4.16448265e-02 1.74849406e-01 -1.32683560e-01 9.80710983e-01
-1.11450158e-01 2.15852857e-01 -3.33979070e-01 -8.77982974e-01
4.60450491e-03 5.16262054e-01 1.17363371e-01 3.26877147e-01
-2.01528355e-01 4.30793434e-01 -5.82572997e-01 -1.05355108e+00
4.76467192e-01 6.56299531e-01 2.43794516e-01 1.31874219e-01
2.00083777e-01 1.79426253e-01 -1.93382517e-01 1.14888740e+00
7.77463913e-01 2.85215020e-01 3.31641346e-01 -6.84958279e-01
-2.02666238e-01 -8.40526402e-01 -1.40425515e+00 1.81007719e+00
-8.42186958e-02 7.18020558e-01 6.82325214e-02 -9.08964574e-01
7.49666154e-01 5.92415333e-01 1.80470645e-01 -5.07260323e-01
6.67891681e-01 3.13671947e-01 -1.63361594e-01 -8.40717018e-01
8.59674156e-01 -1.71020955e-01 -1.64387837e-01 -9.12296399e-02
4.44148839e-01 -9.70119908e-02 6.60998404e-01 2.24298224e-01
6.39422894e-01 7.21962750e-03 4.60133463e-01 -2.23857518e-02
8.99949908e-01 7.88542256e-02 -1.13520481e-01 5.50451040e-01
4.38911468e-02 7.72782028e-01 3.19572836e-01 -1.53402820e-01
-9.33121383e-01 -8.79588783e-01 -9.93679091e-02 9.60425079e-01
2.20584407e-01 -5.27704835e-01 -4.05390471e-01 -5.99934816e-01
-3.96470577e-02 2.96198934e-01 -5.20846903e-01 3.70773613e-01
-2.02695295e-01 -9.79518592e-01 4.17727470e-01 1.86433673e-01
4.88546908e-01 -5.31099200e-01 -4.57945764e-01 -3.42646912e-02
-3.54654282e-01 -1.10152340e+00 1.19094469e-01 -4.82185856e-02
-7.15241134e-01 -1.16944790e+00 -1.06685269e+00 -7.08359420e-01
5.68268716e-01 3.15980703e-01 9.84431922e-01 1.08462058e-01
-1.79073557e-01 8.53968084e-01 -9.40379918e-01 -3.72609735e-01
-6.30790114e-01 -4.71005477e-02 -3.51676881e-01 3.12908500e-01
-9.80181247e-03 -1.23280443e-01 -1.48650482e-01 1.02797337e-01
-1.38897431e+00 1.18185850e-02 6.71670377e-01 6.44915223e-01
3.94514985e-02 2.42502809e-01 2.06999481e-01 -5.54426730e-01
7.47450411e-01 -5.59610724e-01 -2.12296814e-01 5.55774748e-01
-4.39754128e-02 -1.32184187e-02 1.26742437e-01 -5.23537815e-01
-1.11827290e+00 2.70321190e-01 2.64819209e-02 -3.28436047e-01
-4.80445802e-01 8.27669322e-01 -1.92025438e-01 -3.93376887e-01
2.55981922e-01 -1.55336857e-01 1.19099475e-01 -5.91313601e-01
3.14153254e-01 9.36880827e-01 4.33846891e-01 -4.60606366e-01
4.05860990e-01 2.51957148e-01 3.01404685e-01 -1.08500552e+00
-1.73600260e-02 -6.72417164e-01 -6.75934315e-01 -8.00533652e-01
8.94929051e-01 -7.40288436e-01 -6.88440025e-01 5.65516353e-01
-1.29270065e+00 4.91394967e-01 2.68569797e-01 5.37217677e-01
-1.71738565e-01 7.96773911e-01 -3.58705789e-01 -1.12121212e+00
-8.46358296e-03 -1.22802424e+00 1.22356105e+00 2.56954759e-01
5.57155758e-02 -1.00696862e+00 1.19164608e-01 5.20789206e-01
3.67366046e-01 5.89163840e-01 8.61557424e-01 -6.36285067e-01
-4.01084870e-02 -4.63600218e-01 -2.55503863e-01 2.45357260e-01
-4.27554287e-02 3.47296745e-01 -9.61612880e-01 -8.73416960e-02
-3.84195864e-01 -2.42133036e-01 1.06676984e+00 -1.80574670e-01
6.53803587e-01 2.60519028e-01 -4.74117488e-01 -3.29129755e-01
1.58372974e+00 4.17209238e-01 7.74817765e-01 1.13832518e-01
6.43779457e-01 8.40891898e-01 5.03183067e-01 4.65672821e-01
5.14740646e-01 7.00122595e-01 4.43266809e-01 -1.27812281e-01
-7.28419982e-03 3.12134951e-01 1.71222344e-01 1.02716696e+00
-3.65076691e-01 -7.49988258e-01 -1.10606134e+00 2.80358464e-01
-2.18248916e+00 -9.56717312e-01 -4.00272310e-01 2.07184100e+00
4.90776122e-01 -3.06877762e-01 1.91055894e-01 6.50672495e-01
8.54220688e-01 -4.30967286e-02 6.09686434e-01 -4.01291877e-01
-5.18620193e-01 -6.86558560e-02 2.51037270e-01 4.22197700e-01
-1.41124260e+00 2.76617825e-01 6.77189970e+00 9.30014551e-01
-9.31536317e-01 1.96851537e-01 2.91138291e-01 3.47880453e-01
1.51487077e-02 -3.02740037e-01 -4.64160591e-01 5.79398274e-01
9.00747836e-01 2.19648615e-01 1.01866864e-01 1.13074012e-01
-4.40183163e-01 -8.16750109e-01 -1.08528256e+00 1.51582849e+00
7.72204399e-01 -9.30090666e-01 4.26250517e-01 -1.75689310e-02
5.29667974e-01 -5.99188864e-01 9.46725458e-02 1.37325987e-01
-3.17309946e-01 -1.04337728e+00 6.95410609e-01 1.06099975e+00
2.03719705e-01 -9.01840985e-01 1.47046471e+00 1.08048722e-01
-1.30367744e+00 5.60219586e-03 2.15751171e-01 3.11640769e-01
3.74855846e-01 2.82712102e-01 -3.41444999e-01 1.14118648e+00
5.80088437e-01 5.85048258e-01 -1.09632146e+00 1.32411778e+00
5.41945398e-01 -1.30222946e-01 -3.80313188e-01 -1.19276263e-01
7.00382739e-02 -2.61373520e-02 4.89233971e-01 1.52290666e+00
7.92339027e-01 7.86259305e-03 1.16296671e-01 -1.60221383e-02
1.22841820e-01 4.15947646e-01 -9.22037959e-01 -1.49894670e-01
-1.12030357e-01 1.35922420e+00 -8.81822526e-01 -5.68016529e-01
-5.60669601e-01 8.51897776e-01 -2.77673453e-01 1.82811499e-01
-7.26330221e-01 -2.66520768e-01 -3.56741510e-02 -3.56906176e-01
-8.68962631e-02 -3.36342871e-01 7.86276162e-02 -1.05601621e+00
-2.85582602e-01 -7.87467778e-01 6.40987813e-01 -1.12025321e+00
-1.43857586e+00 6.64456367e-01 5.81758738e-01 -1.55670238e+00
-1.51486233e-01 -9.61333334e-01 -2.05996245e-01 5.08237481e-01
-1.15326595e+00 -1.35666943e+00 -4.34196264e-01 7.01211035e-01
3.92125487e-01 -3.30634803e-01 8.06650519e-01 6.98893666e-01
-3.24095935e-01 2.62905091e-01 6.09301031e-02 -2.32563898e-01
9.55625057e-01 -1.13548338e+00 -7.61298060e-01 5.17066002e-01
5.03727933e-03 2.04372317e-01 8.36568594e-01 -5.25471449e-01
-1.61217499e+00 -4.02812034e-01 7.99303770e-01 -3.85901988e-01
5.39297640e-01 -1.40799358e-01 -8.05795670e-01 2.15088680e-01
1.13812327e+00 -4.79751468e-01 8.60709131e-01 -2.56637931e-01
-2.43497133e-01 -9.34510008e-02 -1.36162102e+00 3.70622933e-01
3.92927021e-01 -3.17324400e-01 -8.33839595e-01 2.24850759e-01
1.06106840e-01 -2.30387822e-01 -8.83146405e-01 5.86575627e-01
4.67739314e-01 -9.05799866e-01 8.15524578e-01 -2.93904580e-02
2.12829530e-01 -4.96593773e-01 -7.32598245e-01 -1.35962570e+00
3.58740509e-01 -9.12738740e-02 1.40192866e-01 1.49417543e+00
4.63897854e-01 -3.66013288e-01 2.58748144e-01 2.47918800e-01
1.42889008e-01 -1.07581533e-01 -1.12440360e+00 -3.84091496e-01
-3.55159789e-01 -1.98930115e-01 1.68822244e-01 1.00523341e+00
3.28680247e-01 4.23783273e-01 -4.46912050e-01 -1.05699085e-01
5.76070786e-01 -2.62134612e-01 5.17026186e-01 -1.33923399e+00
2.48639151e-01 -5.30262709e-01 -9.76299226e-01 -7.48390108e-02
-1.18274048e-01 -6.16125166e-01 -1.92831323e-01 -1.59402871e+00
2.99218655e-01 2.73474514e-01 -3.59190181e-02 3.57810229e-01
2.14471787e-01 6.37798548e-01 5.86143494e-01 -6.22271299e-02
-7.65149236e-01 3.44126463e-01 9.37397838e-01 -4.24731135e-01
2.07364127e-01 -5.52312851e-01 -7.01216385e-02 6.07284188e-01
5.77264786e-01 1.10791214e-02 -1.57454967e-01 -4.03282903e-02
4.02338386e-01 1.04738280e-01 3.46306413e-01 -1.25942862e+00
2.98473179e-01 1.85689613e-01 5.32500625e-01 -1.19779372e+00
8.77721667e-01 -1.42584956e+00 4.71840978e-01 1.41969144e-01
-4.27776985e-02 5.22895217e-01 3.68727684e-01 5.16496897e-01
-6.90692246e-01 -4.51114208e-01 5.92116237e-01 -5.05551994e-02
-1.02094030e+00 -5.76426685e-01 -8.25203955e-01 -7.09248185e-01
1.23419249e+00 -1.40341774e-01 -7.90564120e-01 -4.84395564e-01
-9.87052560e-01 2.68857300e-01 2.52502263e-01 5.09422660e-01
7.69020617e-01 -1.58539236e+00 -4.16237503e-01 -2.16072664e-01
3.40528369e-01 -7.03936696e-01 3.28040063e-01 1.04718769e+00
-6.19561017e-01 1.65152952e-01 -5.50255358e-01 -6.46890581e-01
-1.89643896e+00 6.64229929e-01 3.52486707e-02 2.12601628e-02
5.65630980e-02 2.48130202e-01 -4.26995575e-01 -2.51104772e-01
2.56002843e-01 -1.88701212e-01 -9.78141010e-01 8.68092597e-01
4.18315113e-01 6.17804170e-01 1.06274851e-01 -1.17406344e+00
-4.33594137e-01 8.00006688e-01 1.88667014e-01 -5.36412001e-01
8.06059062e-01 -4.08143789e-01 -4.47412699e-01 9.42953289e-01
1.31520033e+00 -2.11671025e-01 -5.43328412e-02 4.12578695e-02
-2.48785019e-02 -3.70929688e-01 -1.40426680e-01 -9.10091937e-01
-7.56347060e-01 9.03712153e-01 1.18682122e+00 6.44161880e-01
1.23049212e+00 -3.92793790e-02 3.21006387e-01 3.22566420e-01
3.48033965e-01 -1.22651887e+00 4.62946296e-02 4.61233675e-01
1.00596011e+00 -1.44352090e+00 2.52027601e-01 -4.94522274e-01
-8.01280320e-01 1.38846970e+00 2.64623314e-01 2.53544986e-01
7.35739827e-01 2.05191284e-01 9.60682053e-03 -7.54742473e-02
-6.23975992e-01 -5.49346924e-01 5.18966973e-01 4.41232145e-01
5.03207803e-01 5.20995185e-02 -7.26445436e-01 2.86119938e-01
2.91821033e-01 -1.62257642e-01 5.53322375e-01 1.44196022e+00
-4.44178492e-01 -1.43333650e+00 -1.17448521e+00 2.26093709e-01
-3.80411983e-01 1.38912305e-01 -6.29547238e-01 1.04743576e+00
4.57847029e-01 1.36074507e+00 -6.95265234e-02 -7.18839586e-01
3.39542150e-01 4.73916173e-01 8.44312727e-01 -8.19097757e-02
-6.82300687e-01 3.20383906e-01 4.59382266e-01 -2.48151615e-01
-1.29514658e+00 -6.89353645e-01 -9.38058436e-01 -4.89614904e-01
-1.66719407e-01 2.21585371e-02 1.06061745e+00 1.07364643e+00
-1.15440212e-01 3.40694308e-01 3.18955392e-01 -1.39827943e+00
1.06863439e-01 -9.95236099e-01 -5.08867264e-01 5.09006321e-01
2.72847801e-01 -9.52490151e-01 -1.18719019e-01 2.89750516e-01] | [13.119930267333984, 5.161306381225586] |
d775acdf-9770-4c2b-81fe-6d6b6fb701f3 | bert-got-a-date-introducing-transformers-to-1 | null | null | https://openreview.net/forum?id=9onUW-cjTl8 | https://openreview.net/pdf?id=9onUW-cjTl8 | BERT got a Date: Introducing Transformers to Temporal Tagging | Temporal expressions in text play a significant role in language understanding, and correctly identifying them is fundamental to various retrieval and natural language processing systems.
Previous works have slowly shifted from rule-based to neural architectures, capable of tagging expressions with higher accuracy. However, neural models cannot yet distinguish between different expression types at the same level as their rule-based counterparts.
In this work, we aim to identify the most suitable transformer architecture for joint temporal tagging and type classification as well as investigating the effect of semi-supervised training on the performance of these systems.
Based on our study of token classification variants and encoder-decoder architectures, we present a transformer encoder-decoder model using the RoBERTa language model as our best-performing system.
By supplementing training resources with weakly labeled data from rule-based systems, our model surpasses previous works in temporal tagging and type classification, especially on rare classes.
Our code and pre-trained experiments are available at: https://github.com/unknown_repo | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['temporal-tagging'] | ['natural-language-processing'] | [ 3.54018509e-02 -4.56815623e-02 -6.96203470e-01 -5.31952739e-01
-7.13155568e-01 -8.12845111e-01 7.74644673e-01 3.04405242e-01
-6.20586038e-01 7.36153722e-01 1.75668836e-01 -3.89002621e-01
1.03863530e-01 -6.49845481e-01 -4.53149378e-01 -3.94079775e-01
-2.58092761e-01 4.61745977e-01 4.13445860e-01 -2.84172386e-01
-1.18583418e-01 7.54545778e-02 -1.48071456e+00 7.81654179e-01
5.20572841e-01 1.36338580e+00 -1.57260299e-01 5.25101364e-01
-4.72501010e-01 1.41050971e+00 -3.32594633e-01 -6.09891236e-01
-2.09165335e-01 -2.78656483e-01 -1.09672272e+00 -5.02773464e-01
1.15613744e-01 3.43630612e-02 -3.28179836e-01 7.33634949e-01
3.19441468e-01 4.40028012e-02 5.43744683e-01 -1.04129601e+00
-7.81475723e-01 9.31155384e-01 4.26299721e-02 3.17182660e-01
3.52191031e-01 -3.39068770e-01 1.26755476e+00 -8.42176795e-01
7.27874219e-01 9.91472006e-01 9.34372902e-01 7.09282696e-01
-9.82999742e-01 -5.79633474e-01 2.10299700e-01 4.68545049e-01
-1.51766133e+00 -7.37502456e-01 5.48636019e-01 -4.44311976e-01
1.54162002e+00 -8.81064497e-03 6.56402886e-01 1.31558275e+00
1.95599869e-01 1.11955011e+00 1.09170127e+00 -5.34890294e-01
1.23250529e-01 -3.24500550e-04 2.23958775e-01 8.13282311e-01
-2.88913161e-01 9.91879329e-02 -8.00209522e-01 -6.02921098e-02
4.20508891e-01 -2.70406693e-01 2.33388133e-02 -1.37883916e-01
-1.24187124e+00 4.97085959e-01 1.53226912e-01 9.95785356e-01
-2.68984258e-01 2.46779963e-01 6.48572087e-01 4.32033926e-01
7.29028940e-01 3.19560915e-01 -1.01922905e+00 -5.49180031e-01
-9.51505482e-01 -1.06605269e-01 7.43413866e-01 9.49636519e-01
3.96896750e-01 -1.72387790e-02 -4.73420471e-01 1.12361956e+00
1.43981710e-01 2.41092041e-01 7.56540358e-01 -7.42110610e-01
9.64339003e-02 6.93822384e-01 -9.15154442e-03 -2.83678323e-01
-3.89437348e-01 -3.16387087e-01 -4.73000914e-01 -2.50067949e-01
6.82993650e-01 -1.62886247e-01 -9.41668808e-01 1.81280005e+00
-1.83467250e-02 3.93275507e-02 -1.14563676e-02 4.70032573e-01
6.04459524e-01 5.05643368e-01 4.71107006e-01 -2.24061027e-01
1.66119742e+00 -6.94838285e-01 -8.89751136e-01 -1.59550399e-01
1.25681651e+00 -7.49403834e-01 6.49514139e-01 3.25044304e-01
-9.84561861e-01 -2.01897338e-01 -6.58224225e-01 -2.62054056e-01
-7.28985608e-01 3.49523425e-01 8.33918571e-01 4.67162400e-01
-1.16297889e+00 6.49282694e-01 -1.14347184e+00 -6.30490482e-01
3.38255495e-01 3.17428321e-01 -4.31302160e-01 2.87118852e-01
-1.72562659e+00 1.22433484e+00 3.79286230e-01 2.44615927e-01
-6.29155457e-01 -6.04526699e-01 -9.24974501e-01 -2.20748007e-01
2.22063854e-01 -3.97667408e-01 1.94712174e+00 -1.19009173e+00
-1.47360110e+00 1.09664416e+00 -6.70104802e-01 -7.39282787e-01
9.45487693e-02 -1.37826562e-01 -6.24910474e-01 -1.57786056e-01
1.03793144e-02 5.41913033e-01 3.55733484e-01 -6.06501758e-01
-6.61386251e-01 -1.02656253e-01 8.09597373e-02 -1.60843641e-01
-3.87085885e-01 3.58991295e-01 -3.32497686e-01 -5.59973359e-01
-1.82776019e-01 -9.27921653e-01 4.13055503e-04 -1.42786384e-01
3.65545764e-03 -7.11152375e-01 5.13796806e-01 -5.12448430e-01
1.40430737e+00 -2.10729241e+00 -2.25317195e-01 -1.94938868e-01
-8.62358809e-02 3.66896242e-01 -3.80548388e-02 5.15865684e-01
-2.31833994e-01 1.79638878e-01 -1.47191316e-01 -4.29432005e-01
1.87913507e-01 4.10821080e-01 -3.03194761e-01 2.65886307e-01
3.26161712e-01 9.50676262e-01 -1.12138724e+00 -6.27868354e-01
-6.85404390e-02 4.39441442e-01 -2.44675219e-01 1.90926269e-01
-4.93080676e-01 2.48885989e-01 -5.05840898e-01 7.66232073e-01
1.25236943e-01 -1.80213109e-01 6.06594324e-01 -1.33768648e-01
-2.91849762e-01 1.08264172e+00 -4.78090256e-01 1.72628629e+00
-8.70035708e-01 7.65533745e-01 -2.58852839e-01 -1.10141766e+00
8.16862643e-01 8.23167264e-01 7.37335682e-01 -1.08359921e+00
1.24834955e-01 3.77200991e-01 1.26357339e-02 -5.45609236e-01
5.96074760e-01 -3.80228370e-01 -3.11151683e-01 3.75212491e-01
2.79483706e-01 3.29570770e-01 5.15278161e-01 1.96357872e-02
1.24148846e+00 4.22882766e-01 3.83155376e-01 -8.95813927e-02
2.63537318e-01 1.98795021e-01 8.69064391e-01 4.77716267e-01
-2.80654520e-01 2.01509014e-01 5.31619489e-01 -6.31002903e-01
-7.96185076e-01 -8.67747426e-01 -3.87830436e-01 1.38277018e+00
-3.26491475e-01 -6.38674498e-01 -3.37824941e-01 -7.15758324e-01
-2.43127078e-01 7.64707208e-01 -7.45529771e-01 -7.16145337e-02
-4.90800261e-01 -6.63401723e-01 1.14483488e+00 7.52199829e-01
3.57885092e-01 -1.18266821e+00 -2.93969065e-01 3.39176208e-01
-4.88929749e-01 -1.26765537e+00 -1.61131948e-01 6.55186355e-01
-7.27049708e-01 -9.20920134e-01 -4.99374300e-01 -8.05576622e-01
1.82671562e-01 -3.47672939e-01 1.40769863e+00 2.77757533e-02
2.04976693e-01 1.19751260e-01 -7.36266494e-01 -3.26516688e-01
-4.11394417e-01 3.96657825e-01 7.65563101e-02 1.28698042e-02
7.92098463e-01 -4.86657232e-01 -2.03718722e-01 2.29669422e-01
-5.97077787e-01 -1.40272379e-01 7.25656986e-01 7.15442300e-01
3.38328898e-01 -1.30271316e-01 2.48779893e-01 -1.01617360e+00
3.17487538e-01 -5.00876725e-01 -4.70648140e-01 4.20353621e-01
-6.96838081e-01 4.43487763e-01 4.99411553e-01 -4.79152799e-01
-1.25121713e+00 1.22836456e-01 -4.75459516e-01 -1.67595506e-01
-2.82821238e-01 7.44427681e-01 1.77095607e-01 2.30807751e-01
5.62493503e-01 2.64837086e-01 -3.89500737e-01 -6.03523612e-01
2.61350483e-01 4.64636117e-01 3.47048998e-01 -7.03270733e-01
2.81754851e-01 2.50325173e-01 -2.70456612e-01 -5.95729530e-01
-1.24480033e+00 -6.17942154e-01 -9.12184894e-01 -9.39791650e-02
8.67188871e-01 -9.36256230e-01 -5.01556039e-01 6.43163264e-01
-1.22160864e+00 -7.28819609e-01 -8.20924565e-02 4.04902667e-01
-5.01972079e-01 1.26979887e-01 -9.42871988e-01 -9.03011680e-01
-6.30636886e-02 -6.81300163e-01 1.13919950e+00 -1.72817722e-01
-5.54114401e-01 -1.26463413e+00 3.90814394e-01 1.52384952e-01
4.37420636e-01 -1.10447660e-01 8.37540865e-01 -7.76971579e-01
-3.35612833e-01 -1.20489307e-01 1.24324538e-01 4.67148842e-03
-8.40379894e-02 7.69299734e-03 -1.10506618e+00 1.31774694e-01
-3.48159343e-01 -5.32539487e-01 1.01025140e+00 2.16508925e-01
7.91727901e-01 -3.67089748e-01 -5.64978063e-01 4.99643743e-01
1.06463611e+00 3.69406462e-01 6.64928734e-01 3.08210850e-01
4.46643740e-01 5.17269909e-01 7.42100120e-01 2.46786907e-01
5.94966054e-01 9.33158219e-01 -1.57824978e-02 9.81620327e-02
-1.15292436e-02 -2.61604130e-01 7.15414584e-01 8.65914881e-01
-1.09642401e-01 -3.52614313e-01 -1.15232444e+00 8.32595527e-01
-1.87903905e+00 -1.17301631e+00 -5.02261259e-02 1.91999102e+00
1.29227626e+00 1.70137256e-01 1.08773924e-01 1.02352887e-01
3.71641576e-01 1.98202699e-01 -1.23521559e-01 -5.18398762e-01
-9.22468379e-02 3.52326334e-01 4.72683489e-01 3.22283268e-01
-1.22849202e+00 1.26194823e+00 6.59014511e+00 8.30271900e-01
-1.36127377e+00 3.33508462e-01 3.28477323e-01 -5.33145517e-02
-2.21381679e-01 1.97463319e-01 -8.53105128e-01 3.42146009e-01
1.49621212e+00 -5.55896796e-02 2.09997460e-01 5.17766893e-01
1.70964852e-01 -9.12283640e-03 -1.27342379e+00 6.46952748e-01
-1.69237807e-01 -1.24981594e+00 -2.33419433e-01 -1.80091143e-01
3.49773169e-01 2.80872881e-01 -9.97455195e-02 6.54557168e-01
5.73044181e-01 -8.60900104e-01 1.00832582e+00 5.20344079e-01
6.25196397e-01 -1.20289542e-01 8.37239444e-01 4.44788784e-02
-1.48662746e+00 2.85811238e-02 6.87910020e-02 -2.37848371e-01
8.81302431e-02 7.32750714e-01 -8.83558393e-01 5.49006164e-01
6.42675161e-01 1.08690989e+00 -4.46113139e-01 6.75986052e-01
-4.56609398e-01 8.60774040e-01 -2.60073692e-01 -2.96295166e-01
1.30163804e-01 2.96820581e-01 1.35793969e-01 1.57066190e+00
3.26000333e-01 1.75521582e-01 5.19440137e-02 5.34685373e-01
-3.71562950e-02 1.19473249e-01 -6.82348192e-01 -4.29821044e-01
6.13310158e-01 1.07522738e+00 -5.92945635e-01 -4.68115449e-01
-6.05479836e-01 7.15397060e-01 6.18086100e-01 1.97285742e-01
-8.84828091e-01 -5.88093661e-02 6.76598966e-01 1.55131102e-01
3.45834702e-01 -5.49940944e-01 -2.48476677e-03 -1.44267726e+00
-1.31546063e-02 -6.48335040e-01 7.53829420e-01 -7.61634469e-01
-1.12861681e+00 7.69938171e-01 2.86065284e-02 -1.25264454e+00
-6.41926467e-01 -8.01316679e-01 -2.78986663e-01 6.48557782e-01
-1.48183930e+00 -1.38531196e+00 2.54584432e-01 4.70195055e-01
2.68574625e-01 4.42887396e-02 1.12624860e+00 6.39230609e-01
-5.26504695e-01 7.82149971e-01 -3.92078385e-02 6.24126077e-01
8.59131217e-01 -1.17202866e+00 2.15826645e-01 9.59660709e-01
4.18575972e-01 7.31356323e-01 4.98450190e-01 -3.12817842e-01
-1.18915403e+00 -1.18442476e+00 1.73054206e+00 -6.64853573e-01
1.03315604e+00 -4.77689266e-01 -9.09551084e-01 9.34856951e-01
2.44932786e-01 -7.20711309e-04 7.14140415e-01 7.54656196e-01
-7.19333112e-01 -3.57855298e-03 -6.32034957e-01 3.49120915e-01
1.27703309e+00 -9.94847894e-01 -5.64637482e-01 2.31395453e-01
3.96570981e-01 -3.46419990e-01 -1.16798830e+00 3.30301434e-01
7.37013280e-01 -8.06128860e-01 5.80890119e-01 -6.21941686e-01
4.00090456e-01 -2.40607932e-01 -2.42233090e-02 -1.11655617e+00
-3.67913872e-01 -3.66375566e-01 -9.15999860e-02 1.34394777e+00
7.46918082e-01 -5.80835760e-01 6.23232126e-01 5.25064349e-01
-1.68164685e-01 -7.52016246e-01 -9.89674509e-01 -1.00726783e+00
2.05833167e-01 -7.71109045e-01 2.35887155e-01 1.22633851e+00
4.80931908e-01 4.15174007e-01 -3.34576070e-01 -3.47309895e-02
1.25606969e-01 2.02038452e-01 1.34952530e-01 -8.90852392e-01
-1.98959574e-01 -6.52858198e-01 -4.25934225e-01 -9.54875886e-01
6.56339705e-01 -1.21395099e+00 1.70622453e-01 -1.32701385e+00
-3.48597616e-02 -6.39342725e-01 -4.46911991e-01 1.40485275e+00
1.82625011e-01 3.69648904e-01 -2.33108535e-01 1.74392328e-01
-8.12796652e-01 5.06850958e-01 5.73419988e-01 -3.09397548e-01
1.35001123e-01 -2.38832623e-01 -3.45764488e-01 6.30742669e-01
7.77643263e-01 -6.87930226e-01 -1.37625471e-01 -6.31715238e-01
5.18258512e-01 6.14672676e-02 7.98867270e-02 -6.94077492e-01
2.40844280e-01 -1.26384407e-01 3.52745578e-02 -2.46292397e-01
4.10844088e-01 -5.04421711e-01 8.37117061e-02 3.84192973e-01
-5.81920862e-01 1.69461057e-01 3.26158106e-01 6.49791807e-02
-4.39991593e-01 -2.50507027e-01 5.83788633e-01 -1.43219128e-01
-1.26921415e+00 2.34708413e-01 -8.53400648e-01 1.17565222e-01
6.14576638e-01 9.53829437e-02 -3.49390298e-01 -2.85680354e-01
-8.43127549e-01 1.08155906e-01 3.17738861e-01 6.65222466e-01
2.58052945e-01 -1.32075047e+00 -5.17135024e-01 -1.29419923e-01
4.26387042e-01 -4.69979316e-01 -2.28899922e-02 1.12518394e+00
-1.22297570e-01 7.52060115e-01 -1.35224298e-01 -4.09534037e-01
-1.19807494e+00 5.27594030e-01 5.96423388e-01 -5.95656455e-01
-1.36871323e-01 8.11134160e-01 -1.56986281e-01 -3.83244008e-01
6.73570335e-02 -5.74568808e-01 -2.24962696e-01 1.85251802e-01
3.12786549e-01 -2.36575201e-01 3.42080206e-01 -6.09203994e-01
-7.21884906e-01 3.12857896e-01 -7.70049691e-02 -1.62986532e-01
1.17685831e+00 1.90162361e-01 -2.19746813e-01 8.51508737e-01
1.19913280e+00 -1.36951804e-01 -6.64831579e-01 -5.15991688e-01
4.17964906e-01 4.01228927e-02 5.43279350e-02 -7.37348437e-01
-9.06422615e-01 7.08156943e-01 3.71034622e-01 2.62485325e-01
1.07781565e+00 1.60275623e-01 7.75886595e-01 4.84369308e-01
5.28599024e-01 -1.00546038e+00 -2.83380210e-01 1.09759271e+00
3.83879572e-01 -1.10621929e+00 -3.77125829e-01 -1.88524470e-01
-5.05676329e-01 8.62837553e-01 4.28298235e-01 2.70968258e-01
7.53938615e-01 6.34320080e-01 3.78789783e-01 -1.71693534e-01
-1.54611385e+00 -5.10777414e-01 2.66112059e-01 4.93098199e-01
1.46618056e+00 6.00808077e-02 -3.57727379e-01 5.33814251e-01
-2.01844826e-01 3.74067545e-01 1.04359381e-01 1.04227257e+00
-4.97234836e-02 -1.64989316e+00 3.61733548e-02 3.77119690e-01
-8.37380528e-01 -4.62878615e-01 -4.80569214e-01 5.56966603e-01
1.31186828e-01 8.51268470e-01 2.00128257e-01 -4.95032936e-01
2.44802102e-01 6.24650300e-01 6.34303570e-01 -6.45557404e-01
-7.37010479e-01 -3.29338759e-01 8.52938533e-01 -5.11549413e-01
-7.85339952e-01 -9.21822846e-01 -1.20548296e+00 -1.69781700e-01
-8.17715153e-02 4.14214492e-01 1.99098721e-01 1.24333787e+00
3.03904444e-01 4.16284889e-01 1.58557206e-01 -3.10723424e-01
-2.40319535e-01 -1.09223473e+00 -3.53855699e-01 4.94061798e-01
1.00063734e-01 -7.75768936e-01 1.03954356e-02 4.21464473e-01] | [9.921918869018555, 9.214543342590332] |
fd475198-f89e-4da9-a1ef-474ec28da6fc | a-comparative-study-of-texture-attributes-for | 1812.08263 | null | http://arxiv.org/abs/1812.08263v1 | http://arxiv.org/pdf/1812.08263v1.pdf | A comparative study of texture attributes for characterizing subsurface structures in seismic volumes | In this paper, we explore how to computationally characterize subsurface
geological structures presented in seismic volumes using texture attributes.
For this purpose, we conduct a comparative study of typical texture attributes
presented in the image processing literature. We focus on spatial attributes in
this study and examine them in a new application for seismic interpretation,
i.e., seismic volume labeling. For this application, a data volume is
automatically segmented into various structures, each assigned with its
corresponding label. If the labels are assigned with reasonable accuracy, such
volume labeling will help initiate an interpretation process in a more
effective manner. Our investigation proves the feasibility of accomplishing
this task using texture attributes. Through the study, we also identify
advantages and disadvantages associated with each attribute. | ['Suhail Al-Dharrab', 'Mohamed Deriche', 'Zhen Wang', 'Zhiling Long', 'Yuting Hu', 'Ghassan AlRegib', 'Haibin Di', 'Yazeed Alaudah', 'Muhammad Ali Qureshi', 'Motaz Alfarraj', 'Asjad Amin'] | 2018-12-19 | null | null | null | null | ['seismic-interpretation'] | ['miscellaneous'] | [ 4.00778860e-01 1.23755671e-01 2.56824315e-01 -5.13461649e-01
-5.42704642e-01 -3.50686580e-01 4.90776122e-01 3.59097540e-01
-2.78513789e-01 6.41388059e-01 6.04184903e-02 -2.83957541e-01
-4.27790284e-01 -1.36774421e+00 -2.54051238e-01 -7.73825169e-01
-3.51229846e-01 8.00631523e-01 2.32831553e-01 -1.47123143e-01
9.51816082e-01 8.53839874e-01 -1.64035773e+00 2.26858154e-01
8.94900620e-01 1.10907781e+00 4.30849791e-01 2.32013255e-01
-5.68496227e-01 5.07135749e-01 -5.42161524e-01 -8.01442787e-02
2.62973495e-02 -5.40791489e-02 -1.28130579e+00 3.42804909e-01
4.15702090e-02 -2.48066396e-01 3.17235470e-01 1.02634799e+00
8.14867318e-02 3.80429804e-01 1.27479637e+00 -7.67263591e-01
-7.32609853e-02 5.64245403e-01 -8.38593841e-01 2.53470123e-01
2.89194316e-01 -3.59662473e-01 8.82335067e-01 -1.10683930e+00
4.79637623e-01 1.19338465e+00 6.01866364e-01 1.94564927e-02
-1.17815399e+00 -1.85290188e-01 -5.39170299e-03 1.98781416e-02
-1.60112286e+00 -1.52848214e-01 1.13378084e+00 -6.21993899e-01
5.60739577e-01 3.50016683e-01 6.73748136e-01 2.73760617e-01
6.81011602e-02 6.93365872e-01 1.19889116e+00 -5.78133821e-01
4.66221452e-01 -3.95889491e-01 2.94153631e-01 5.65313637e-01
4.56486970e-01 -4.07346666e-01 -3.28433990e-01 -8.55935216e-02
8.20731461e-01 -5.12371182e-01 6.02338053e-02 -8.36721808e-02
-1.12870264e+00 7.72432923e-01 4.23457213e-02 4.36500669e-01
-3.31328779e-01 2.54973382e-01 4.01567668e-01 -3.28454524e-02
8.30484688e-01 4.24693555e-01 8.85273963e-02 2.78058555e-03
-8.48528802e-01 3.95018384e-02 5.33504307e-01 7.67196000e-01
1.05149972e+00 1.31134883e-01 2.01718345e-01 8.36154282e-01
3.77702028e-01 5.29427588e-01 -3.27634765e-03 -9.45535302e-01
4.70228523e-01 4.71025765e-01 1.05045959e-02 -1.23288393e+00
-5.40331423e-01 9.10345018e-02 -6.42906249e-01 2.82331318e-01
4.60223973e-01 2.71175414e-01 -1.03099287e+00 1.00218439e+00
3.17201376e-01 -9.34943333e-02 2.67907858e-01 5.57593167e-01
7.76857376e-01 6.92237079e-01 3.56001377e-01 4.62482460e-02
1.66751146e+00 -2.15345562e-01 -7.78982639e-01 6.17219210e-02
7.75068939e-01 -7.63646543e-01 1.12292302e+00 5.35417855e-01
-8.11872303e-01 -2.73738325e-01 -9.22292769e-01 2.35808372e-01
-2.66790092e-01 2.60259598e-01 7.83653021e-01 6.36980653e-01
-9.12345707e-01 4.73644644e-01 -9.84946012e-01 -1.89202085e-01
1.02887452e-01 3.41738999e-01 -2.04027489e-01 5.15390098e-01
-1.08069694e+00 7.79969275e-01 6.26945734e-01 3.02962899e-01
-4.97529387e-01 -5.72716631e-02 -8.24108124e-01 -8.45821500e-02
2.57042944e-01 -2.24748075e-01 1.09613323e+00 -4.88191545e-01
-9.96577084e-01 8.76192689e-01 -1.99946582e-01 -1.15728773e-01
2.78156877e-01 6.41485751e-02 -2.51698613e-01 5.37299514e-01
3.66618931e-01 2.51333356e-01 4.98553783e-01 -1.72952437e+00
-8.27614844e-01 -4.12703365e-01 3.18177082e-02 2.33305365e-01
-3.74538332e-01 -1.26910999e-01 -3.02376121e-01 -8.73264372e-01
8.89865398e-01 -5.98056376e-01 -2.81080365e-01 -3.29392701e-01
-4.20596987e-01 -2.20557377e-01 9.23422992e-01 -7.16921449e-01
1.18384886e+00 -2.18092155e+00 -1.62770763e-01 9.46958244e-01
3.66625130e-01 -3.52436453e-01 5.30532181e-01 4.15059388e-01
3.07997078e-01 4.30563062e-01 -8.67286444e-01 -3.05205081e-02
-3.25354427e-01 2.66829640e-01 -1.10110685e-01 4.27597612e-01
1.67374372e-01 2.66318321e-01 -7.88760185e-01 -1.15902650e+00
2.62649208e-01 1.25600830e-01 -2.53105968e-01 2.49160714e-02
1.24781378e-01 3.94929647e-01 -1.25817847e+00 1.04127514e+00
7.95829833e-01 7.85494149e-02 1.46172389e-01 -5.07308066e-01
-4.28890616e-01 -4.67129536e-02 -1.25605953e+00 9.88001943e-01
-5.25158167e-01 5.86959898e-01 -7.55209923e-02 -1.42596304e+00
1.58347881e+00 3.21575463e-01 7.37397611e-01 -5.76629400e-01
-3.86161283e-02 6.43091321e-01 -1.16289020e-01 -8.35758388e-01
1.14250171e+00 -2.27237836e-01 -2.58853942e-01 6.67729616e-01
-3.48267823e-01 -5.17017841e-01 2.76972711e-01 -5.08578382e-02
4.36470985e-01 7.44325891e-02 3.01876366e-01 -8.84457946e-01
5.02042353e-01 4.01083380e-01 1.86168015e-01 6.81767642e-01
1.43746257e-01 5.98731339e-01 4.53217596e-01 -6.56346738e-01
-1.26002455e+00 -1.01980066e+00 -6.02775753e-01 7.44332492e-01
3.39911401e-01 4.31123078e-02 -8.77071381e-01 -2.69243598e-01
-1.47479445e-01 3.58880043e-01 -6.54967964e-01 3.20784479e-01
-8.51735413e-01 -1.16152823e+00 5.90298474e-01 6.57032073e-01
7.00503945e-01 -1.13757050e+00 -1.02588022e+00 3.81071776e-01
-5.27542353e-01 -6.48757100e-01 8.25944990e-02 2.35711142e-01
-1.16483581e+00 -7.91336894e-01 -5.90682387e-01 -7.95651138e-01
1.06851864e+00 1.73788056e-01 9.64363337e-01 3.07735890e-01
-4.40993048e-02 4.04583752e-01 -7.99142957e-01 -2.23810807e-01
-2.40467146e-01 1.42878279e-01 -1.54347584e-01 1.91684157e-01
9.05283839e-02 -2.28416234e-01 -4.19211447e-01 3.51883471e-01
-1.19235277e+00 -6.98230043e-03 3.39755327e-01 5.89041531e-01
6.38677657e-01 2.87726223e-01 3.93590689e-01 -9.68946278e-01
7.20779181e-01 -3.63499761e-01 -4.30740893e-01 4.18003708e-01
-2.42932439e-01 3.14631552e-01 3.17864686e-01 -2.52417754e-02
-1.41522694e+00 2.17224117e-02 -2.21756876e-01 3.55914772e-01
-1.83310837e-01 1.14315510e+00 -1.76735044e-01 -4.08407390e-01
4.68324214e-01 1.18308954e-01 -8.97655562e-02 -4.44446951e-01
-4.24704291e-02 7.26197422e-01 4.65678871e-01 -1.12667823e+00
5.52892864e-01 7.10743308e-01 3.74518931e-01 -1.11938977e+00
-3.62122267e-01 -5.00857234e-01 -9.77338910e-01 -6.98462665e-01
8.19456637e-01 -2.30777085e-01 -6.89407349e-01 4.43780363e-01
-1.03808212e+00 -8.83636177e-02 7.19398493e-04 6.52053416e-01
-6.61054611e-01 6.07575595e-01 -3.26576352e-01 -1.10899663e+00
1.49355270e-04 -1.42180169e+00 1.10358226e+00 -2.71799136e-02
-3.73596847e-01 -1.27233315e+00 -3.25054228e-01 1.19618051e-01
1.40757307e-01 3.97076577e-01 9.76851463e-01 -4.54507172e-01
-4.51279640e-01 1.77792177e-01 -1.40001312e-01 -2.07689524e-01
2.95521259e-01 3.52455854e-01 -9.89860177e-01 9.32555124e-02
9.50034112e-02 8.11527520e-02 9.29474831e-01 4.35367554e-01
1.55297077e+00 -3.22422832e-02 -2.45060027e-01 5.80634713e-01
1.48355377e+00 7.56797254e-01 8.27874243e-01 9.71831858e-01
6.22964501e-01 1.21068394e+00 1.14573312e+00 5.94975114e-01
2.12044269e-01 5.27821839e-01 3.91835392e-01 -3.11523676e-01
3.26008439e-01 1.76353067e-01 -4.92778867e-01 6.87040567e-01
-6.95172012e-01 -3.61398429e-01 -1.56101131e+00 4.49079275e-01
-1.56638873e+00 -7.24688172e-01 -4.32213575e-01 1.98945355e+00
5.79701245e-01 3.85731682e-02 -2.41570309e-01 6.20111048e-01
8.01808536e-01 1.90776646e-01 1.15665920e-01 -4.37959015e-01
-2.80848652e-01 8.81686509e-02 6.18910789e-01 5.60308099e-01
-1.32334352e+00 7.97565222e-01 6.58655357e+00 9.86766756e-01
-1.06183445e+00 -3.93879503e-01 1.16905642e+00 9.00752187e-01
-8.29706132e-01 1.38106331e-01 -5.38605571e-01 2.40777299e-01
3.68657231e-01 -9.59923491e-02 -1.86405793e-01 4.88782585e-01
3.94881845e-01 -7.28066146e-01 -6.07330859e-01 5.77222288e-01
-1.13425471e-01 -1.31199288e+00 4.21612799e-01 7.19081387e-02
2.84034669e-01 -8.37559044e-01 -5.12515679e-02 -4.90499049e-01
-1.70117661e-01 -1.01635826e+00 1.00872147e+00 6.20185733e-01
9.09044147e-01 -8.41871500e-01 8.09844315e-01 -2.17212081e-01
-1.59827602e+00 1.50268734e-01 -1.52606800e-01 -7.29277655e-02
4.90328759e-01 6.20100379e-01 -9.46147084e-01 8.21791947e-01
6.37564123e-01 2.78827518e-01 -3.56478184e-01 1.03889954e+00
1.40142649e-01 5.68203509e-01 -3.28394711e-01 -1.14300035e-01
4.09222841e-01 -6.62763536e-01 2.80623078e-01 1.16350377e+00
7.42932558e-01 5.94747961e-01 2.58784920e-01 7.26134300e-01
4.81136024e-01 4.36714500e-01 -6.55897200e-01 2.55585521e-01
7.05066025e-01 8.82029891e-01 -1.61278439e+00 -4.57622230e-01
-2.72063673e-01 2.37726331e-01 -6.37732372e-02 2.17195794e-01
-4.17847455e-01 -4.76236343e-01 -1.81022547e-02 2.26296872e-01
-2.38159180e-01 -4.25717384e-01 -1.16591227e+00 -5.40746331e-01
1.92437768e-02 -2.73749202e-01 3.68141741e-01 -5.99933624e-01
-9.63937461e-01 6.60078645e-01 7.16385305e-01 -1.56164372e+00
-8.17328095e-02 -6.26311600e-01 -5.50384045e-01 7.70040870e-01
-1.12740731e+00 -1.01617932e+00 -4.00350928e-01 2.36109510e-01
3.70705813e-01 6.73611462e-02 5.73344648e-01 1.42960608e-01
-2.15253532e-01 1.03068270e-01 1.15230218e-01 2.48666570e-01
1.68931946e-01 -1.21870959e+00 3.98440599e-01 6.50442958e-01
-3.62677246e-01 3.65197480e-01 9.33600187e-01 -1.01934588e+00
-9.61757064e-01 -6.50989175e-01 8.48212063e-01 7.15453401e-02
7.01675236e-01 7.63592049e-02 -1.23483086e+00 4.08524156e-01
-4.33292598e-01 -3.65970731e-01 3.99589330e-01 -1.70456227e-02
2.75468707e-01 5.61786443e-02 -1.19111156e+00 5.62949777e-01
7.52108634e-01 -3.38459402e-01 -7.79155195e-01 -4.91777621e-02
7.47423470e-02 -2.72643477e-01 -1.13950980e+00 7.66708493e-01
4.04001862e-01 -5.93279183e-01 1.14238572e+00 2.09353790e-01
4.82281506e-01 -4.57222044e-01 -1.69923037e-01 -1.01756501e+00
-1.12250097e-01 1.30910829e-01 6.87261224e-01 1.02891314e+00
4.65265006e-01 -6.68907762e-01 9.23288703e-01 5.56947827e-01
-4.73270386e-01 -5.42579055e-01 -1.06642604e+00 -7.18546033e-01
-7.24742338e-02 -4.36235428e-01 8.28092933e-01 9.27604318e-01
2.78973114e-02 -3.14685494e-01 -1.98754087e-01 2.13444307e-01
8.86173844e-01 1.92446634e-01 2.53954917e-01 -1.42091000e+00
4.68783677e-01 -4.38062906e-01 -4.02876586e-01 -7.69355178e-01
-9.47704166e-02 -4.27841038e-01 2.63505280e-01 -1.64205980e+00
1.01365812e-01 -1.21261299e+00 1.97450578e-01 3.94429475e-01
-5.86555747e-04 2.86290437e-01 -2.11791620e-01 5.58839262e-01
2.99008526e-02 3.88616323e-01 1.30976200e+00 -2.87009239e-01
-1.98044538e-01 -4.53713797e-02 -2.64990300e-01 7.78039336e-01
1.22669029e+00 -3.05249840e-01 -2.55506307e-01 -5.39707065e-01
-8.88974126e-03 3.66371900e-01 1.63202807e-01 -7.90687323e-01
9.28230956e-02 -4.81868148e-01 6.43673837e-02 -8.78361106e-01
1.91265866e-01 -7.26923227e-01 2.86964267e-01 4.05896902e-01
-9.94722322e-02 2.14720100e-01 -6.20571859e-02 1.54868007e-01
-5.30233860e-01 -7.93006837e-01 5.78147650e-01 -1.93462983e-01
-1.33501256e+00 -6.65934607e-02 -8.00198793e-01 -2.90573448e-01
9.89536881e-01 -9.71964657e-01 -1.17397122e-01 -2.43163586e-01
-8.89510393e-01 -4.93747788e-03 4.03086782e-01 -2.94809222e-01
9.13654387e-01 -1.20420301e+00 -5.36206484e-01 5.22098280e-02
3.19587976e-01 1.30379885e-01 3.53200734e-01 7.12987483e-01
-1.28111231e+00 3.10285036e-02 -4.72554743e-01 -7.45435715e-01
-1.03784668e+00 1.49319712e-02 2.02838272e-01 2.69599594e-02
-6.87854826e-01 4.91120845e-01 5.77742696e-01 6.72974112e-03
-1.61312431e-01 -3.87307584e-01 -9.43276644e-01 3.39963973e-01
1.37814581e-01 5.59902966e-01 2.47213826e-01 -1.00784886e+00
-2.40190804e-01 9.99733746e-01 4.59365100e-01 -4.43576843e-01
1.20134354e+00 -2.68156379e-01 -3.24683249e-01 5.45880616e-01
7.43038356e-01 -1.09810717e-01 -1.04878378e+00 -1.62681052e-03
5.03275096e-01 -6.92944348e-01 3.17609906e-02 -1.28081664e-01
-1.03017330e+00 5.91049910e-01 2.57270426e-01 4.06129152e-01
1.25086713e+00 1.40847012e-01 2.81759024e-01 3.96864593e-01
7.16099083e-01 -1.30163407e+00 1.59864388e-02 6.06866658e-01
8.23722601e-01 -1.03010845e+00 1.87943533e-01 -9.57712293e-01
-6.44651413e-01 1.47086501e+00 4.07509685e-01 1.17911093e-01
7.01722443e-01 5.10620773e-01 -1.49616031e-02 -5.43914616e-01
9.07188132e-02 -3.52740586e-02 1.63207978e-01 5.51250875e-01
5.78651667e-01 3.28374654e-01 -7.16231108e-01 4.45184335e-02
-3.04880500e-01 -5.31348825e-01 8.58553231e-01 1.36582172e+00
-1.03993642e+00 -1.06452537e+00 -1.03904772e+00 6.78041041e-01
-4.29368496e-01 -1.59053981e-01 7.46603310e-02 8.63984406e-01
-1.79006636e-01 8.36091876e-01 4.60523605e-01 -3.26387465e-01
2.55523473e-01 -5.10358393e-01 1.21798486e-01 -4.99851733e-01
-1.21920511e-01 2.29045108e-01 4.59493190e-01 1.50566489e-01
-8.00290763e-01 -8.79233539e-01 -1.67778528e+00 -2.02859789e-01
-1.23142377e-01 4.60700303e-01 5.21180272e-01 1.02264738e+00
-5.94547987e-01 3.51670653e-01 6.11367822e-01 -9.78300512e-01
-6.72785863e-02 -8.39368165e-01 -9.83634591e-01 3.51653099e-01
-4.94381823e-02 -1.16245151e+00 -3.15717012e-01 4.36459720e-01] | [7.236192226409912, 2.0665860176086426] |
1278257d-747a-4ba2-aa63-123b212a991e | generative-goal-driven-user-simulation-for | null | null | https://aclanthology.org/D12-1007 | https://aclanthology.org/D12-1007.pdf | Generative Goal-Driven User Simulation for Dialog Management | null | ['Ben Allison', 'Mark Steedman', 'Aciel Eshky'] | 2012-07-01 | null | null | null | emnlp-2012-7 | ['user-simulation'] | ['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] | [-7.29979944229126, 3.8672454357147217] |
68d10b05-1822-41ce-a9fd-e04c37f5eb39 | dynamic-feature-pruning-and-consolidation-for | 2211.14742 | null | https://arxiv.org/abs/2211.14742v1 | https://arxiv.org/pdf/2211.14742v1.pdf | Dynamic Feature Pruning and Consolidation for Occluded Person Re-Identification | Occluded person re-identification (ReID) is a challenging problem due to contamination from occluders, and existing approaches address the issue with prior knowledge cues, eg human body key points, semantic segmentations and etc, which easily fails in the presents of heavy occlusion and other humans as occluders. In this paper, we propose a feature pruning and consolidation (FPC) framework to circumvent explicit human structure parse, which mainly consists of a sparse encoder, a global and local feature ranking module, and a feature consolidation decoder. Specifically, the sparse encoder drops less important image tokens (mostly related to background noise and occluders) solely according to correlation within the class token attention instead of relying on prior human shape information. Subsequently, the ranking stage relies on the preserved tokens produced by the sparse encoder to identify k-nearest neighbors from a pre-trained gallery memory by measuring the image and patch-level combined similarity. Finally, we use the feature consolidation module to compensate pruned features using identified neighbors for recovering essential information while disregarding disturbance from noise and occlusion. Experimental results demonstrate the effectiveness of our proposed framework on occluded, partial and holistic Re-ID datasets. In particular, our method outperforms state-of-the-art results by at least 8.6% mAP and 6.0% Rank-1 accuracy on the challenging Occluded-Duke dataset. | ['Wei Yang', 'Qiang Hu', 'Junqing Yu', 'Hang Zhou', 'Yuteng Ye'] | 2022-11-27 | null | null | null | null | ['person-re-identification'] | ['computer-vision'] | [ 9.54416841e-02 7.75335729e-03 8.67498815e-02 -1.87695861e-01
-7.37154186e-01 -1.67697176e-01 4.17497694e-01 1.71032384e-01
-4.75523889e-01 6.91210270e-01 3.49872380e-01 6.31344914e-01
-5.92941465e-03 -6.04073286e-01 -6.93516374e-01 -6.09952509e-01
2.59295911e-01 4.28817093e-01 1.48784965e-01 7.45300427e-02
1.50868967e-01 2.93778688e-01 -2.02383566e+00 1.92259103e-01
1.06277609e+00 1.01688576e+00 3.88890766e-02 1.99706048e-01
5.72647825e-02 4.80076134e-01 -6.95744514e-01 -6.38470650e-01
3.79692465e-01 -2.23551840e-01 -4.82288808e-01 2.33063355e-01
8.68727028e-01 -4.73774195e-01 -4.56397146e-01 1.08345687e+00
8.06569040e-01 3.30586493e-01 5.29297292e-01 -9.13931668e-01
-7.08368778e-01 1.72094733e-01 -9.61229801e-01 2.74500698e-01
5.75024664e-01 1.32591814e-01 5.49818993e-01 -1.12842381e+00
7.58263469e-01 1.33044767e+00 8.48460555e-01 4.47867334e-01
-1.13386858e+00 -7.61834204e-01 2.46583194e-01 6.08372450e-01
-1.75616074e+00 -6.41640425e-01 8.54441643e-01 -3.91383827e-01
8.09903800e-01 2.38682091e-01 5.44117093e-01 1.06157207e+00
-1.37644440e-01 7.68322647e-01 8.97839785e-01 -2.87681162e-01
-1.30249321e-01 4.27926630e-01 3.25460017e-01 7.71594822e-01
5.63780010e-01 1.10455438e-01 -5.19261539e-01 -1.12613633e-01
4.95581895e-01 6.83312416e-02 -1.84130892e-01 -3.65224302e-01
-8.27257752e-01 4.84599680e-01 5.22935927e-01 1.62825704e-01
-3.99861574e-01 -1.91201255e-01 2.97926933e-01 -5.80150820e-02
2.45463029e-01 -1.18640242e-02 -1.60913154e-01 3.46602619e-01
-1.07142782e+00 2.50533819e-01 5.87133944e-01 1.06091595e+00
8.54385555e-01 -1.32810727e-01 -5.19188464e-01 9.59094226e-01
2.67082065e-01 4.39460665e-01 5.08588076e-01 -6.15995526e-01
4.40312147e-01 7.36697435e-01 9.64952111e-02 -1.24758720e+00
-3.92445832e-01 -9.07726288e-01 -9.11268115e-01 -2.06015661e-01
2.96863854e-01 2.10647643e-01 -1.07374465e+00 1.70721328e+00
7.20752835e-01 3.95335704e-01 -3.42919618e-01 1.14254665e+00
1.14689708e+00 1.17331907e-01 2.47971863e-01 -1.76387057e-01
1.64375150e+00 -1.02521908e+00 -7.18489647e-01 -1.89819589e-01
3.26596558e-01 -6.88583374e-01 6.73041105e-01 8.90859812e-02
-9.18750763e-01 -1.04447281e+00 -9.91499245e-01 -3.61724287e-01
-3.58874023e-01 5.00542045e-01 2.57816374e-01 6.41557574e-01
-7.35571861e-01 4.47360933e-01 -4.60026860e-01 -5.00121951e-01
6.22409642e-01 4.84099776e-01 -6.05536997e-01 -3.18376750e-01
-9.75632429e-01 7.85162449e-01 1.72960207e-01 4.07434642e-01
-7.89677262e-01 -7.45819867e-01 -1.03936231e+00 -3.54970358e-02
3.55165392e-01 -8.99329603e-01 4.59252626e-01 -6.54891670e-01
-1.14105308e+00 9.49326158e-01 -5.11104524e-01 -3.14197749e-01
6.92854345e-01 -5.01917601e-01 -3.76612097e-01 2.57806838e-01
4.96896297e-01 5.77823222e-01 9.90137815e-01 -1.42278099e+00
-6.24851584e-01 -7.41410315e-01 -3.18080872e-01 3.72513592e-01
-1.09206200e-01 -1.06337592e-01 -7.52817690e-01 -7.21621454e-01
2.67613977e-01 -7.77128339e-01 3.71054858e-02 -1.18760780e-01
-5.62942684e-01 -2.45361418e-01 6.40645087e-01 -1.14852381e+00
1.03031182e+00 -2.20636582e+00 1.02193661e-01 3.15952599e-01
2.80104995e-01 1.46602452e-01 -3.05756107e-02 -4.71600890e-03
-1.99036375e-02 -6.43370077e-02 -1.86908077e-02 -6.85450613e-01
-1.15930513e-01 -1.66492417e-01 -4.89205420e-02 8.33476782e-01
1.37030482e-02 9.27519321e-01 -6.87788248e-01 -8.85498047e-01
3.45293343e-01 6.28662765e-01 -3.99259746e-01 8.36651027e-02
5.18863440e-01 6.13466382e-01 -3.03093553e-01 1.04420722e+00
8.81254852e-01 6.01270311e-02 -1.56339899e-01 -5.51925063e-01
-5.61440885e-02 -2.28239689e-02 -1.37048566e+00 1.84427154e+00
6.42940588e-03 1.13825299e-01 1.28549188e-01 -9.18403506e-01
8.22784185e-01 1.26218200e-01 3.95483881e-01 -8.05550098e-01
2.86888450e-01 2.06853002e-01 -3.29278439e-01 -4.70525622e-01
3.67430687e-01 3.80922884e-01 4.92098443e-02 7.62481615e-02
1.64796501e-01 6.81420147e-01 1.07283019e-01 2.73269147e-01
8.70725989e-01 1.88458994e-01 2.53368914e-01 -1.68564901e-01
7.79810429e-01 -1.80299357e-01 7.92094588e-01 8.16549540e-01
-5.33740044e-01 7.35893190e-01 -6.88104853e-02 -5.31751275e-01
-8.74674976e-01 -8.79578948e-01 -3.94281834e-01 1.00318205e+00
8.22517395e-01 -5.14338911e-01 -8.25314939e-01 -6.21897876e-01
1.67335615e-01 4.01293546e-01 -7.50655293e-01 -2.36129984e-01
-7.08922029e-01 -7.48593152e-01 4.48366076e-01 4.35195088e-01
9.63405073e-01 -9.17180300e-01 -2.18168259e-01 1.99266359e-01
-5.68974853e-01 -1.09316158e+00 -6.28307879e-01 -1.94841474e-01
-6.26871109e-01 -1.11938834e+00 -8.20037305e-01 -8.71408165e-01
9.26886976e-01 5.34499288e-01 6.91238403e-01 2.53689229e-01
-6.94027126e-01 3.86499137e-01 -2.22040430e-01 7.81836361e-02
2.58447737e-01 9.85423103e-02 1.29132614e-01 3.70945394e-01
5.44663727e-01 -6.22028828e-01 -9.35225487e-01 5.29980063e-01
-2.51306534e-01 -1.16427928e-01 7.30159163e-01 9.26226795e-01
9.49361444e-01 2.56754551e-02 4.45443213e-01 -6.66661143e-01
1.19312026e-01 -4.13921654e-01 -2.10855410e-01 3.82454425e-01
-3.87853622e-01 -1.45952359e-01 2.77389377e-01 -4.83852327e-01
-1.17107570e+00 3.90777797e-01 1.51608676e-01 -4.32012856e-01
-3.58971119e-01 -1.92164600e-01 -6.01927400e-01 -2.16992870e-01
5.51009655e-01 4.50500041e-01 -3.70365769e-01 -7.87258387e-01
3.11913133e-01 5.25114477e-01 9.93610382e-01 -6.49804115e-01
1.03033316e+00 7.00535655e-01 -2.41303548e-01 -6.43343031e-01
-7.51545727e-01 -7.04384387e-01 -8.86756718e-01 -2.24925175e-01
9.42238569e-01 -1.21572518e+00 -6.69050515e-01 3.50451648e-01
-1.15699112e+00 5.05157232e-01 -3.52943718e-01 3.59448284e-01
-6.34737983e-02 6.96273327e-01 -6.03231788e-01 -7.51198411e-01
-4.92826372e-01 -1.05283582e+00 1.30202973e+00 5.65685749e-01
-4.17943336e-02 -2.15162858e-01 -2.92983562e-01 7.16935396e-01
1.52641177e-01 3.23825449e-01 4.56162006e-01 -5.82307458e-01
-7.66538084e-01 -4.54795510e-01 -4.00411427e-01 5.29749505e-02
1.86667945e-02 -7.19950199e-01 -1.21054518e+00 -4.54918444e-01
-3.37985791e-02 -1.33370504e-01 1.16385281e+00 2.73974538e-01
1.01562631e+00 -3.29533845e-01 -7.18345404e-01 7.66454756e-01
1.19152117e+00 -1.53655469e-01 6.44876420e-01 2.06375539e-01
1.02267623e+00 8.10701787e-01 3.37948352e-01 4.17619973e-01
6.74351811e-01 7.84695923e-01 1.41219005e-01 -4.77463119e-02
-6.26092732e-01 -4.70095605e-01 1.44153759e-01 5.23384869e-01
-2.80807704e-01 1.76843181e-01 -4.08411324e-01 6.64508939e-01
-1.87542748e+00 -1.01882076e+00 -1.22663252e-01 2.43685913e+00
6.47209466e-01 -1.05491325e-01 6.37483411e-03 1.07332446e-01
1.17759764e+00 -6.49800822e-02 -7.16466427e-01 3.59443426e-01
-3.63153517e-01 4.56751660e-02 4.34951186e-01 3.15201759e-01
-1.36963964e+00 1.02769208e+00 5.03126240e+00 1.00172365e+00
-4.40058708e-01 3.84420931e-01 4.42112654e-01 -1.48680642e-01
8.71262550e-02 -9.05764177e-02 -1.08080029e+00 6.19558275e-01
2.17494532e-01 1.52765289e-01 3.64005566e-01 9.48814929e-01
-1.18928336e-01 -2.90522516e-01 -9.34879065e-01 1.32651234e+00
5.08354366e-01 -7.32371271e-01 1.26062408e-02 1.19842529e-01
6.06622756e-01 -4.90819752e-01 5.91029860e-02 3.37965399e-01
-2.17049092e-01 -9.43844557e-01 7.54988551e-01 6.97341859e-01
6.87866032e-01 -7.97654152e-01 8.29037845e-01 6.93706423e-02
-1.61349785e+00 -1.08642496e-01 -5.60609460e-01 3.22676778e-01
5.60395829e-02 5.90587854e-01 -4.04812247e-01 7.13539302e-01
8.96798432e-01 6.23009801e-01 -7.71294653e-01 1.19734085e+00
-1.86950073e-01 1.77689433e-01 -4.32692081e-01 5.74295223e-01
-4.52892482e-01 1.56780615e-01 6.44643307e-01 1.16582656e+00
3.12877782e-02 2.53261149e-01 4.07170594e-01 7.31607914e-01
-4.15411517e-02 2.40807325e-01 -3.39424253e-01 7.40392208e-01
6.31449223e-01 1.21997094e+00 -6.57277703e-01 -4.09744591e-01
-3.13126653e-01 1.37592840e+00 4.37411726e-01 5.32388031e-01
-9.26531017e-01 -3.67494166e-01 4.13396806e-01 2.99172461e-01
4.09318030e-01 5.64418845e-02 -3.25959861e-01 -1.29410136e+00
4.05738384e-01 -6.84275687e-01 4.89439696e-01 -3.75476688e-01
-1.36195421e+00 6.38256252e-01 -2.72989333e-01 -1.18488109e+00
9.67639610e-02 2.08736509e-02 -3.19048107e-01 8.94482076e-01
-1.55215800e+00 -1.52756846e+00 -6.82114244e-01 8.91488135e-01
4.99800533e-01 -1.67223319e-01 5.71231604e-01 7.12203383e-01
-1.00944591e+00 1.15564346e+00 -2.32501000e-01 1.93175375e-01
8.84203255e-01 -9.30317163e-01 -3.71791422e-02 9.69738781e-01
-2.02582274e-02 7.48478055e-01 4.98587877e-01 -1.07060885e+00
-1.23807907e+00 -1.20683146e+00 9.09575224e-01 -4.09964770e-01
-4.48148064e-02 -4.46629345e-01 -9.30574179e-01 4.76657063e-01
-2.22582892e-01 3.26430768e-01 4.96080220e-01 4.81769852e-02
-4.30496335e-01 -2.23202854e-01 -1.35352206e+00 2.76144147e-01
1.56455576e+00 -4.80106533e-01 -6.72941625e-01 1.97522655e-01
3.99301380e-01 -4.08915520e-01 -6.01824462e-01 5.17242670e-01
7.32585907e-01 -8.84642005e-01 1.26298273e+00 -2.81272650e-01
-9.08045247e-02 -5.99899530e-01 -6.61633760e-02 -6.92608416e-01
-6.55395746e-01 -3.75894278e-01 -2.54498452e-01 1.53265762e+00
-1.33174300e-01 -4.20625210e-01 6.06843889e-01 7.36371100e-01
-5.77735109e-03 -4.10383999e-01 -1.09244275e+00 -6.42442763e-01
-6.42911792e-01 -9.89762470e-02 5.20768285e-01 6.72181904e-01
-3.35635275e-01 1.97927296e-01 -7.23611534e-01 4.86439317e-01
1.10713935e+00 1.69924930e-01 8.46080780e-01 -1.25243413e+00
-7.59228244e-02 -7.24496171e-02 -6.01255119e-01 -1.05612588e+00
1.77372083e-01 -1.00250018e+00 2.96287891e-03 -1.36180902e+00
5.41574001e-01 -4.25013542e-01 -3.03698897e-01 4.26320523e-01
-5.68957031e-01 5.48911452e-01 1.91994280e-01 5.95619798e-01
-8.88877094e-01 7.66518235e-01 1.03088439e+00 -2.93186277e-01
-3.34606439e-01 -1.92746714e-01 -7.94504523e-01 6.64279342e-01
4.64896172e-01 -5.76974571e-01 -1.86756447e-01 -4.40135390e-01
-4.72498327e-01 -3.80791068e-01 8.45961034e-01 -1.28262711e+00
5.41837573e-01 3.83433908e-01 1.15292656e+00 -7.41375387e-01
4.44788337e-01 -6.51473284e-01 2.27922976e-01 2.87668794e-01
-3.50250006e-02 -2.04536453e-01 3.55027691e-02 8.03981364e-01
-4.34718132e-02 -4.85313945e-02 7.12982059e-01 -9.91715118e-02
-7.38211036e-01 3.44310015e-01 1.66572124e-01 -6.98255897e-02
9.49474871e-01 -4.91912842e-01 -3.30609977e-01 -2.91234590e-02
-8.79818261e-01 2.80876100e-01 3.26753050e-01 4.02954400e-01
6.78505182e-01 -1.33107841e+00 -8.38964164e-01 4.71541315e-01
6.07686490e-02 -1.52415082e-01 7.74968147e-01 8.61509562e-01
-5.82340583e-02 1.77593157e-01 -1.03954665e-01 -5.33339620e-01
-1.20648170e+00 7.01448560e-01 3.07873964e-01 -1.39047235e-01
-9.68280613e-01 8.39091301e-01 5.19648552e-01 -8.78412463e-03
5.03350079e-01 1.05829939e-01 -3.63199294e-01 2.16502026e-01
7.53489137e-01 6.31765902e-01 6.24673581e-03 -1.17604673e+00
-6.88900709e-01 9.13085341e-01 -2.42143437e-01 2.06976041e-01
9.27731812e-01 -4.14447129e-01 5.88141084e-02 -1.39324889e-01
1.06525123e+00 9.71048921e-02 -1.07734871e+00 -4.59243387e-01
-2.27988392e-01 -5.74752688e-01 -2.38524258e-01 -7.32120395e-01
-1.08465981e+00 6.13699377e-01 9.84551787e-01 -4.06306744e-01
9.95541453e-01 5.09778708e-02 1.01795614e+00 7.54070431e-02
4.95266825e-01 -1.10667431e+00 -6.62459657e-02 3.68789174e-02
9.05424237e-01 -1.27557278e+00 2.07123801e-01 -6.07703209e-01
-5.24028659e-01 6.13663316e-01 8.69770527e-01 -2.27311209e-01
6.28258944e-01 -1.18704222e-01 -2.63121098e-01 -5.30846417e-02
-8.99031982e-02 -6.19279742e-01 5.19750416e-01 8.00947070e-01
-5.18669486e-02 -3.74149121e-02 -4.19145733e-01 1.19439125e+00
-3.06664616e-01 -2.05754831e-01 -2.32005134e-01 5.78940988e-01
-3.60249668e-01 -8.06239009e-01 -5.42587459e-01 4.18562800e-01
-3.66001606e-01 -4.39848797e-03 -2.16961235e-01 7.13327050e-01
8.47731769e-01 9.45414245e-01 2.88596493e-03 -4.56788272e-01
4.24852490e-01 5.08624539e-02 4.92698371e-01 -4.87703741e-01
-7.22601831e-01 2.23405510e-01 1.67093449e-03 -7.20257342e-01
-3.21853101e-01 -8.42532158e-01 -1.01284873e+00 -1.25137612e-01
-5.35544217e-01 -4.56585772e-02 3.96019578e-01 8.65649462e-01
6.53469265e-01 3.76334459e-01 2.76825994e-01 -8.90159607e-01
-2.81006515e-01 -1.10525680e+00 -4.25890654e-01 7.20055282e-01
1.79018751e-01 -1.05390084e+00 -2.74575800e-01 1.06644392e-01] | [14.676971435546875, 0.8835920095443726] |
3812d3b2-93d7-4025-9867-029545e76c14 | taking-a-respite-from-representation-learning | 2209.13492 | null | https://arxiv.org/abs/2209.13492v3 | https://arxiv.org/pdf/2209.13492v3.pdf | Taking a Respite from Representation Learning for Molecular Property Prediction | Artificial intelligence (AI) has been widely applied in drug discovery with a major task as molecular property prediction. Despite booming techniques in molecular representation learning, fundamentals underlying molecular property prediction haven't been carefully examined yet. In this study, we conducted a systematic evaluation on a collection of representative models using various molecular representations. In addition to the commonly used MoleculeNet benchmark datasets, we also assembled a suite of opioids-related datasets from ChEMBL and two additional activity datasets from literature. To interrogate the basic predictive power, we also assembled a series of descriptors datasets with varying sizes to evaluate the models' performance. In total, we trained 62,820 models, including 50,220 models on fixed representations, 4,200 models on SMILES sequences and 8,400 models on molecular graphs. We first conducted dataset profiling and highlighted the activity-cliffs issue in the opioids-related datasets. We then conducted rigorous model evaluation and addressed key questions therein. Furthermore, we examined inter-/intra-scaffold chemical space generalization and found that activity cliffs significantly can impact prediction performance. Based on extensive experimentation and rigorous comparison, representation learning models still show limited performance in molecular property prediction in most datasets. Finally, we explored into potential causes why representation learning models fail and highlighted the importance of dataset size. By taking this respite, we reflected on the fundamentals underlying molecular property prediction, the awareness of which can, hopefully, bring better AI techniques in this field. | ['Fusheng Wang', 'Dimitris Samaras', 'Iwao Ojima', 'Hehe Wang', 'Zhibo Yang', 'Jianyuan Deng'] | 2022-09-26 | null | null | null | null | ['molecular-property-prediction'] | ['miscellaneous'] | [ 6.74355447e-01 -2.10083872e-01 -7.31272697e-01 -1.99635446e-01
-7.26257145e-01 -6.82603419e-01 4.57980633e-01 7.13515639e-01
-1.68263003e-01 1.38449335e+00 1.50339752e-01 -5.95750749e-01
-6.49930060e-01 -8.03915620e-01 -7.78580308e-01 -8.09798539e-01
-5.26254833e-01 1.41456097e-01 -1.37542158e-01 -2.50121266e-01
6.76812470e-01 7.65353322e-01 -1.30462837e+00 2.79424727e-01
1.07619429e+00 8.82131219e-01 1.41375384e-03 1.46026567e-01
1.34067774e-01 1.04995608e+00 -5.57469368e-01 -3.44070137e-01
-1.02635719e-01 -4.36400354e-01 -7.88294852e-01 -4.85089719e-01
2.89100826e-01 2.24937364e-01 -3.47912550e-01 6.87550128e-01
7.59064198e-01 2.97898144e-01 9.49653506e-01 -7.24519670e-01
-8.52327943e-01 4.55299884e-01 -2.95093924e-01 3.41318697e-01
7.07085490e-01 1.69225335e-01 1.06869364e+00 -7.76362896e-01
6.92455053e-01 9.84358311e-01 6.14010334e-01 4.14679468e-01
-1.37435663e+00 -9.37448144e-01 9.50045142e-05 4.25257772e-01
-1.36335754e+00 -1.88137352e-01 5.63407123e-01 -5.30216634e-01
1.22651231e+00 2.70263791e-01 4.46387649e-01 1.17248738e+00
4.71959025e-01 3.50065619e-01 1.06893885e+00 -1.24828078e-01
3.43829483e-01 -3.53565328e-02 2.01414555e-01 5.28206885e-01
5.15715003e-01 3.05576354e-01 -5.83127081e-01 -6.69865847e-01
3.44870985e-01 8.94417837e-02 -4.40729082e-01 -2.21107334e-01
-9.05499458e-01 1.04104066e+00 7.38418639e-01 1.97425902e-01
-3.60491395e-01 -7.51264207e-03 5.20630479e-01 1.25852272e-01
8.38752687e-02 1.06901300e+00 -7.28265941e-01 3.29520218e-02
-4.32465613e-01 1.42930046e-01 6.67055488e-01 6.40239656e-01
7.92712748e-01 2.68612001e-02 1.16407819e-01 6.74718857e-01
9.58557054e-03 -4.10256423e-02 3.97179127e-01 -1.57474533e-01
4.25177276e-01 7.18620777e-01 -2.00696737e-01 -1.10872221e+00
-6.15555167e-01 -4.21498895e-01 -7.69976616e-01 -1.26430735e-01
2.17240214e-01 3.67139205e-02 -7.49899507e-01 1.51925969e+00
1.66687388e-02 6.78157359e-02 1.48169041e-01 5.32972813e-01
1.01308548e+00 6.44236505e-01 8.07821512e-01 -3.31753254e-01
1.21641946e+00 -4.15051937e-01 -3.92739773e-01 4.67285246e-01
9.60456014e-01 -6.62693083e-01 6.05434597e-01 6.73253834e-01
-5.45717061e-01 -4.81902510e-01 -1.28007758e+00 2.67473817e-01
-8.49016964e-01 -2.06445456e-01 1.39747024e+00 7.78408110e-01
-4.59419847e-01 1.20618546e+00 -4.54880118e-01 -1.67182386e-01
6.26049578e-01 8.31959307e-01 -7.24629283e-01 -1.28073156e-01
-1.40593290e+00 1.09958351e+00 8.20962906e-01 -1.99635610e-01
-8.45633328e-01 -1.01195121e+00 -7.09395528e-01 -1.68544814e-01
2.28466928e-01 -5.95321238e-01 6.29486501e-01 -6.22987330e-01
-1.37257802e+00 2.84615964e-01 9.31379199e-02 -4.71978515e-01
4.55379859e-03 5.26701845e-02 -9.29627478e-01 8.89115334e-02
-1.91219717e-01 5.54267228e-01 1.83216646e-01 -1.04310453e+00
-2.07571521e-01 -4.69142824e-01 1.04826599e-01 1.61431562e-02
-9.48921964e-02 -2.25422889e-01 1.92943260e-01 -6.47997379e-01
-2.49470681e-01 -8.87846112e-01 -4.48729426e-01 -3.90738070e-01
-3.43117535e-01 -1.39106601e-01 3.51601034e-01 -2.49273673e-01
1.50591671e+00 -1.74177098e+00 2.65528321e-01 5.13210773e-01
1.05502285e-01 2.82458067e-01 -3.10497552e-01 1.09740019e+00
-7.35656738e-01 3.84679496e-01 -2.80058354e-01 4.78912354e-01
-5.12816906e-01 6.03070036e-02 -4.29189742e-01 6.12426579e-01
3.58851761e-01 8.66890907e-01 -8.91109824e-01 -2.00962529e-01
2.40567252e-01 7.23825634e-01 -7.81294584e-01 -8.78921896e-02
-4.26890403e-01 5.20811558e-01 -8.61287117e-01 1.10443151e+00
5.96205175e-01 -2.28532583e-01 4.44205016e-01 -2.86546022e-01
-4.49721962e-02 4.52776283e-01 -5.60979187e-01 1.60935056e+00
-1.21971883e-01 2.40306392e-01 -9.87539887e-01 -9.01209056e-01
1.05293083e+00 2.60618985e-01 6.92462325e-01 -9.41914558e-01
-8.92998204e-02 3.14319342e-01 5.67409456e-01 -2.12512806e-01
2.93940455e-01 -2.58579284e-01 3.26787651e-01 -4.48820367e-02
8.76251794e-03 1.00669533e-01 1.96966186e-01 -1.10451788e-01
1.04012978e+00 2.16817111e-01 7.27168024e-01 -2.77375847e-01
6.81822181e-01 1.76178545e-01 2.76522070e-01 4.39588785e-01
6.88753724e-02 3.82068843e-01 5.19531667e-01 -9.03855681e-01
-4.85540032e-01 -7.77886450e-01 -7.62120783e-01 1.09062660e+00
-1.38639314e-02 -1.04943693e+00 -1.94553897e-01 -5.26495218e-01
1.30500883e-01 5.04441261e-01 -1.03185618e+00 -4.68159437e-01
-3.50403160e-01 -1.31552279e+00 5.90925217e-01 3.21471602e-01
3.31637934e-02 -1.15237141e+00 -5.09559251e-02 4.95674014e-01
5.92204690e-01 -5.05400181e-01 2.11647116e-02 7.47471213e-01
-9.80259895e-01 -1.57073057e+00 -3.16987067e-01 -4.66734916e-01
3.73247415e-01 1.03334144e-01 9.62505043e-01 6.65824264e-02
-3.84930164e-01 -1.99312747e-01 -3.88686359e-01 -6.17616117e-01
-2.92459071e-01 1.46501333e-01 8.64052549e-02 -3.43155116e-01
2.80644536e-01 -7.82107532e-01 -8.52273166e-01 3.85946661e-01
-9.58366096e-01 -4.82918710e-01 7.81615198e-01 8.02017570e-01
7.24550545e-01 -1.80626631e-01 8.84315670e-01 -1.11223912e+00
6.51686311e-01 -8.43778491e-01 -3.11776400e-01 2.59288967e-01
-7.06958055e-01 1.65432766e-01 8.13663900e-01 -4.65637922e-01
-4.85888690e-01 1.01056257e-02 -2.22694635e-01 6.93757758e-02
7.23838583e-02 9.54260707e-01 -1.96371898e-01 -4.50664908e-01
9.52294767e-01 2.68107563e-01 -1.97810337e-01 -4.88606364e-01
1.51501417e-01 4.36799496e-01 -8.96826684e-02 -8.72376680e-01
2.83512861e-01 3.50443535e-02 4.23001081e-01 -9.65496361e-01
-5.17288983e-01 -2.55751103e-01 -4.25286204e-01 4.50557768e-01
6.60224676e-01 -8.49578083e-01 -1.04641461e+00 -2.04404160e-01
-9.70942914e-01 9.82920676e-02 3.69433165e-02 6.16327643e-01
-4.29308653e-01 5.24948776e-01 -3.61233652e-01 -4.62732613e-01
-3.46880227e-01 -1.32769227e+00 6.41583204e-01 1.98246539e-02
-4.70271081e-01 -1.03291094e+00 4.38495845e-01 2.86394298e-01
2.99277812e-01 6.34247065e-01 1.54342937e+00 -9.75216389e-01
-4.70987529e-01 -1.42808601e-01 -3.41241881e-02 -9.96208712e-02
4.33984846e-01 1.18551962e-01 -9.81816709e-01 -2.07012966e-01
-6.37707174e-01 -4.23027366e-01 9.23098803e-01 2.54094601e-01
1.67303920e+00 -1.36342600e-01 -4.88507569e-01 7.39330173e-01
1.35483718e+00 9.08384442e-01 9.75128174e-01 3.62019628e-01
6.41571641e-01 4.40884113e-01 4.28351671e-01 4.53781068e-01
-1.76288873e-01 6.48204505e-01 5.81880689e-01 -1.11389980e-01
2.19978690e-01 -5.16963184e-01 8.31826404e-02 3.59241128e-01
-6.95698798e-01 -2.57278472e-01 -7.16074705e-01 -9.37327594e-02
-1.43587458e+00 -9.69271183e-01 9.50731803e-03 2.21192861e+00
9.73077536e-01 -3.85367535e-02 2.32898876e-01 3.38432752e-02
3.99088115e-01 7.30889589e-02 -7.93575346e-01 -5.77236235e-01
-2.87462622e-01 7.34866500e-01 4.76479024e-01 1.78142592e-01
-9.55649078e-01 9.84210849e-01 7.12548971e+00 1.13251460e+00
-1.21953487e+00 -5.66340506e-01 7.91115582e-01 2.80504465e-01
-3.96537066e-01 2.20006313e-02 -6.47976220e-01 4.56568331e-01
1.13306808e+00 -1.34551734e-01 9.90523025e-02 8.45333457e-01
7.65043497e-02 4.75682653e-02 -1.30530322e+00 9.25928116e-01
-2.32436150e-01 -1.85844755e+00 7.46341825e-01 4.13135767e-01
7.03907728e-01 -7.31763318e-02 -1.65407415e-02 2.70718843e-01
-1.98279172e-01 -1.83459330e+00 -2.19506770e-02 3.61261040e-01
8.74147177e-01 -1.03680217e+00 4.30691898e-01 -1.03735782e-01
-1.05933535e+00 -6.29332755e-03 -6.18958116e-01 -1.58498555e-01
-3.51997793e-01 1.81189179e-01 -8.27504277e-01 1.06726587e+00
8.01537260e-02 1.13915920e+00 -7.26541460e-01 1.15542889e+00
-2.80228816e-03 5.38686216e-01 8.47795531e-02 -1.60587162e-01
2.84559757e-01 -4.24491376e-01 -3.61175425e-02 1.19126892e+00
-3.45810317e-02 1.72920898e-01 3.75413559e-02 5.24485946e-01
-1.02027230e-01 4.18044746e-01 -7.46904671e-01 -3.46711606e-01
3.89904171e-01 8.53914857e-01 -5.01660943e-01 -1.33590117e-01
-3.90639365e-01 4.86021280e-01 2.62567252e-01 4.22618091e-01
-9.02232587e-01 -2.30622411e-01 9.78412509e-01 1.41709343e-01
-3.92570421e-02 5.88137656e-02 1.09816738e-03 -9.04011905e-01
-5.48469007e-01 -1.08064008e+00 6.00154817e-01 -3.90939176e-01
-1.41992104e+00 6.13096893e-01 4.76230904e-02 -1.20909250e+00
2.60644495e-01 -9.94372904e-01 -5.64263999e-01 7.82164693e-01
-1.54437625e+00 -8.31631958e-01 1.97574839e-01 5.07643759e-01
2.83621997e-01 -2.44076982e-01 1.24296212e+00 4.08998072e-01
-7.46570289e-01 4.38032240e-01 2.00436279e-01 -4.32463557e-01
7.41116345e-01 -9.11241233e-01 8.27892274e-02 -1.81445196e-01
6.69196323e-02 1.22388172e+00 5.15429795e-01 -6.50315285e-01
-1.58580375e+00 -1.02735567e+00 4.02959853e-01 -5.10733306e-01
8.13466728e-01 -8.83583445e-03 -1.08369648e+00 3.78475934e-01
-4.86147664e-02 -9.28226635e-02 1.50368690e+00 1.25148460e-01
-4.69142824e-01 2.70107478e-01 -9.35683191e-01 4.07441854e-01
1.14357734e+00 -5.17337084e-01 -3.90356869e-01 4.78715181e-01
5.66847861e-01 -1.23908736e-01 -1.30277371e+00 7.34697938e-01
7.57473469e-01 -7.52590418e-01 1.31219447e+00 -1.36838853e+00
5.24147928e-01 -2.71287292e-01 -3.37811917e-01 -1.12077737e+00
-4.70870286e-01 -4.01935607e-01 -9.62647125e-02 6.22469544e-01
7.36529648e-01 -7.28414834e-01 7.39766002e-01 3.80259335e-01
-2.42837280e-01 -1.33118081e+00 -8.00837576e-01 -6.81481242e-01
4.02881682e-01 -2.54027426e-01 6.25508368e-01 1.13067150e+00
2.83085257e-01 5.11477649e-01 -4.64878380e-01 -1.83557212e-01
7.75205716e-02 8.26243088e-02 5.00500262e-01 -1.23957407e+00
-3.25513154e-01 -4.78116393e-01 -7.61455476e-01 -6.43427193e-01
1.58643931e-01 -1.20256400e+00 -8.87648642e-01 -1.18997276e+00
4.33055669e-01 -4.41558331e-01 -7.62408793e-01 4.56124425e-01
-1.00846648e-01 9.06599164e-02 -1.54644877e-01 2.57033616e-01
-4.25548881e-01 5.84208012e-01 1.39133477e+00 -3.32023174e-01
-3.29194367e-01 -5.13126813e-02 -1.06550777e+00 3.48133653e-01
8.98753107e-01 -3.40619981e-01 -5.49948931e-01 2.81005859e-01
3.59722674e-01 -8.21087658e-02 5.81818670e-02 -7.88947046e-01
-8.28939155e-02 -5.17821133e-01 7.97542572e-01 -5.00466347e-01
4.29211259e-01 -6.89746976e-01 5.11159539e-01 6.54276848e-01
-2.76348472e-01 -1.47381589e-01 4.43488955e-01 7.44828403e-01
-2.37745598e-01 -3.75606865e-02 5.86462319e-01 4.96941842e-02
-7.01258659e-01 5.97933531e-01 -2.71807402e-01 -3.04586142e-01
1.14258015e+00 -6.08988166e-01 -2.68247485e-01 1.24253325e-01
-6.90159619e-01 -1.38581768e-01 3.89823049e-01 2.96439469e-01
6.70547247e-01 -1.11236286e+00 -2.18290031e-01 -4.44490090e-02
5.64384580e-01 -4.29160684e-01 1.68494314e-01 6.30894661e-01
-6.20777011e-01 8.52633715e-01 -3.33407730e-01 -2.12632954e-01
-1.05488241e+00 8.70592237e-01 3.21615577e-01 -6.68300465e-02
-9.09932032e-02 4.66763973e-01 3.92543346e-01 -2.11328834e-01
1.23714954e-01 -3.50763917e-01 -4.23070610e-01 2.66937166e-01
4.23224121e-01 3.91591251e-01 2.30069593e-01 -4.87043470e-01
-7.49163210e-01 7.60499775e-01 -2.94395268e-01 7.81489968e-01
1.58305860e+00 6.70919597e-01 1.23528816e-01 1.05249710e-01
1.35865510e+00 -7.21247569e-02 -8.26958477e-01 2.50983715e-01
2.87822038e-01 -3.57421756e-01 -2.72688657e-01 -1.07099950e+00
-6.06954813e-01 6.87641799e-01 4.06644762e-01 -2.00764984e-01
8.61277282e-01 -1.68841645e-01 3.07874471e-01 7.29072332e-01
5.27824223e-01 -6.75784826e-01 1.27961233e-01 4.32917297e-01
9.45978045e-01 -1.33655679e+00 5.78695655e-01 -6.02089345e-01
-6.35826945e-01 1.25585806e+00 3.48028332e-01 5.22604920e-02
4.56805795e-01 -3.27423006e-01 -4.91310149e-01 -5.52784801e-01
-7.45226026e-01 1.53312776e-02 4.99639422e-01 7.08352745e-01
1.12847257e+00 8.26849267e-02 -5.89045405e-01 5.10979354e-01
-1.19566806e-01 -2.30500605e-02 2.13452518e-01 9.61959481e-01
-3.64768296e-01 -1.61767101e+00 -9.53538939e-02 3.67221147e-01
-6.08399451e-01 -4.58670437e-01 -8.41802418e-01 1.01976728e+00
1.44113898e-01 7.59245157e-01 -5.78881979e-01 -5.34313083e-01
4.10661101e-01 -1.35059506e-01 6.66296899e-01 -5.95658362e-01
-7.25827873e-01 -3.14659476e-01 2.40963772e-01 -2.79419631e-01
-4.76000577e-01 -1.09549850e-01 -1.20724952e+00 -3.18497181e-01
-4.72240359e-01 4.79614258e-01 5.74190021e-01 6.05409980e-01
5.54239094e-01 4.37978745e-01 5.18625736e-01 -6.46802902e-01
-3.99557382e-01 -9.23361480e-01 -5.79955459e-01 1.99007556e-01
6.94989935e-02 -7.28985369e-01 -1.36465237e-01 -1.74276069e-01] | [5.15069055557251, 5.809345722198486] |
2bac2874-2f35-40a9-bd2d-52b5515948a7 | 1st-solution-places-for-cvpr-2023-ug-textbf-2 | 2306.08963 | null | https://arxiv.org/abs/2306.08963v1 | https://arxiv.org/pdf/2306.08963v1.pdf | 1st Solution Places for CVPR 2023 UG$^{\textbf{2}}$+ Challenge Track 2.1-Text Recognition through Atmospheric Turbulence | In this technical report, we present the solution developed by our team VIELab-HUST for text recognition through atmospheric turbulence in Track 2.1 of the CVPR 2023 UG$^{2}$+ challenge. Our solution involves an efficient multi-stage framework that restores a high-quality image from distorted frames. Specifically, a frame selection algorithm based on sharpness is first utilized to select the sharpest set of distorted frames. Next, each frame in the selected frames is aligned to suppress geometric distortion through optical-flow-based image registration. Then, a region-based image fusion method with DT-CWT is utilized to mitigate the blur caused by the turbulence. Finally, a learning-based deartifacts method is applied to remove the artifacts in the fused image, generating a high-quality outuput. Our framework can handle both hot-air text dataset and turbulence text dataset provided in the final testing phase and achieved 1st place in text recognition accuracy. Our code will be available at https://github.com/xsqhust/Turbulence_Removal. | ['Luxin Yan', 'Yi Chang', 'Shuning Cao', 'Xueyao Xiao', 'Shengqi Xu'] | 2023-06-15 | null | null | null | null | ['image-registration'] | ['computer-vision'] | [ 2.74929136e-01 -8.34928334e-01 3.04209858e-01 -1.50398761e-01
-7.47906089e-01 -4.67377752e-01 4.71357107e-01 -2.99142569e-01
-2.74645150e-01 5.04273295e-01 3.57336104e-01 -2.03663275e-01
2.92333085e-02 -4.12356704e-01 -3.40201050e-01 -8.51821065e-01
5.01513004e-01 -1.43367663e-01 -2.27655154e-02 -1.60343960e-01
5.63283265e-01 3.72420788e-01 -1.38569665e+00 3.93919796e-01
9.19405282e-01 9.01715696e-01 2.51035213e-01 1.21426690e+00
-3.99504036e-01 9.49265838e-01 -4.84741211e-01 -1.59673348e-01
5.21219790e-01 -5.44042885e-01 -5.58730304e-01 3.55245560e-01
9.42674577e-01 -4.78589207e-01 -5.77557206e-01 9.48189735e-01
6.87527955e-01 5.14625490e-01 5.19588768e-01 -6.98037684e-01
-1.94344148e-01 2.02695027e-01 -7.68244565e-01 6.26433492e-01
6.13533020e-01 3.49313468e-01 6.65808380e-01 -1.20259476e+00
4.02057946e-01 1.12089217e+00 5.29774070e-01 2.66009331e-01
-8.14463735e-01 -7.16347873e-01 -4.35926802e-02 3.61683369e-01
-1.25925171e+00 -7.90866733e-01 6.73655927e-01 -5.28663158e-01
7.47166753e-01 6.41120851e-01 6.74523354e-01 9.26849484e-01
2.91695863e-01 6.30274236e-01 1.07553017e+00 -5.38839698e-01
-8.26652050e-02 -3.25805575e-01 2.51889557e-01 6.24648392e-01
2.13250779e-02 8.54219869e-02 -1.03170598e+00 9.58837271e-02
5.81663251e-01 -4.14524674e-02 -6.73353612e-01 3.37297857e-01
-1.24097610e+00 3.07688236e-01 1.16695069e-01 4.41467196e-01
-2.62906849e-01 -2.90187616e-02 2.59993345e-01 2.36651465e-01
7.53921807e-01 2.06416562e-01 -1.54354736e-01 -3.88496608e-01
-1.44535053e+00 1.24544196e-01 4.92829621e-01 5.43614209e-01
5.44397175e-01 3.55022699e-01 -4.72927094e-01 8.26373518e-01
4.11995828e-01 7.81080365e-01 5.38135827e-01 -9.17996705e-01
5.95955849e-01 2.64697164e-01 1.92064121e-01 -1.05124176e+00
-9.32519287e-02 -3.27813655e-01 -9.73282456e-01 8.99814814e-02
2.08570257e-01 -1.63318902e-01 -7.32296288e-01 8.60297263e-01
3.23301792e-01 7.61421084e-01 1.41403571e-01 1.19486892e+00
6.91924751e-01 7.36077666e-01 -4.70391482e-01 -3.33987892e-01
1.08767045e+00 -1.02053392e+00 -9.21771824e-01 1.56149521e-01
3.03355068e-01 -1.49164855e+00 8.16417575e-01 5.93242764e-01
-1.06067050e+00 -7.16403544e-01 -1.00822186e+00 -5.80079760e-03
-7.14535192e-02 1.59710690e-01 -1.65115315e-02 6.52229130e-01
-1.07493758e+00 4.82430190e-01 -7.90849328e-01 -3.05869579e-01
1.56280011e-01 -5.77702709e-02 -9.85668227e-02 -1.46019414e-01
-8.46185863e-01 6.82923973e-01 -6.48235232e-02 4.14086014e-01
-5.20042002e-01 -5.67048132e-01 -6.33657813e-01 -3.08360338e-01
9.36754271e-02 -6.36238098e-01 9.69251931e-01 -9.51776922e-01
-1.74841392e+00 5.30403674e-01 -5.84630847e-01 -4.78067309e-01
7.67616391e-01 -5.88713527e-01 -3.74859184e-01 3.52186918e-01
-5.36094308e-02 1.40770137e-01 1.40093255e+00 -9.26750481e-01
-6.10177994e-01 -7.96647221e-02 -4.53976661e-01 8.37955832e-01
-1.23691283e-01 2.14147776e-01 -6.84271693e-01 -9.38777745e-01
1.98789522e-01 -7.28219330e-01 1.42744631e-01 -2.78845459e-01
-4.77783680e-01 4.22743052e-01 1.18960106e+00 -1.12833285e+00
1.21562195e+00 -2.31451368e+00 1.71863109e-01 -1.91072859e-02
2.76664793e-01 4.35628682e-01 2.51196530e-02 1.31467968e-01
-1.52810561e-02 -1.13408148e-01 -1.42709240e-01 -6.49165273e-01
-2.82829016e-01 -3.71551186e-01 -5.12198150e-01 6.55604899e-01
-2.04016849e-01 5.00708401e-01 -6.07407570e-01 -3.07881624e-01
9.18979824e-01 5.52491188e-01 -1.97315812e-01 2.54538625e-01
3.52300964e-02 7.49288619e-01 -3.19385469e-01 5.25658011e-01
1.04160285e+00 3.67246754e-02 -3.19260448e-01 -3.35519850e-01
-4.49361056e-01 -2.11852174e-02 -1.32999134e+00 1.53945565e+00
-3.58022779e-01 9.23037887e-01 3.16611499e-01 -5.83503306e-01
9.19869542e-01 3.08998674e-01 8.38690221e-01 -5.16542614e-01
2.17781976e-01 -2.72300513e-03 -3.31543952e-01 -6.19190216e-01
7.85071790e-01 3.13431591e-01 4.37659949e-01 2.03631610e-01
-4.57849503e-01 -4.87470180e-01 1.43285543e-01 1.65558100e-01
1.09958625e+00 2.40692645e-01 -9.54074934e-02 -1.69152692e-01
9.83490050e-01 -1.62749410e-01 4.48336095e-01 7.89763153e-01
-4.85163003e-01 1.14079642e+00 -2.19158977e-01 -5.54832697e-01
-9.45710778e-01 -7.70100534e-01 -3.81273441e-02 5.31090856e-01
2.52630889e-01 -5.83676875e-01 -7.19973624e-01 -3.97234172e-01
-3.15988272e-01 4.04977173e-01 -3.05615485e-01 8.84258747e-02
-5.84608436e-01 -6.56438828e-01 5.52341759e-01 -3.90365906e-02
1.04870820e+00 -8.58248830e-01 -5.08599102e-01 -8.10606126e-03
-5.79954445e-01 -1.33776283e+00 -9.47858632e-01 -2.87374079e-01
-6.62539363e-01 -1.08974814e+00 -7.81513214e-01 -4.36298519e-01
4.56204861e-01 7.16410637e-01 5.70048451e-01 2.87243694e-01
-3.05708140e-01 3.86786133e-01 -6.41667306e-01 8.33142623e-02
-2.26337269e-01 -5.32971740e-01 5.49860671e-02 5.06610453e-01
8.14613923e-02 -6.68100268e-03 -6.76929176e-01 3.21810037e-01
-7.61213839e-01 3.55124891e-01 -1.85226258e-02 8.25839996e-01
5.05984545e-01 3.06812704e-01 -2.56908387e-01 -2.02899471e-01
4.85664994e-01 1.81354955e-01 -7.48557925e-01 3.50008602e-03
-2.10296318e-01 -2.99120039e-01 6.19144022e-01 -2.34086081e-01
-1.20091605e+00 1.25803798e-01 -1.13777015e-02 -6.62949920e-01
-1.94370523e-01 3.26353878e-01 2.02486545e-01 -2.05029517e-01
4.65226918e-01 5.67546666e-01 -1.18251547e-01 -3.27404231e-01
1.92523211e-01 8.59421194e-01 6.61087453e-01 -3.26603889e-01
1.01841092e+00 6.12887263e-01 -7.11759105e-02 -1.27199364e+00
-8.06437612e-01 -7.22938180e-01 -6.07303500e-01 -4.58865941e-01
1.07752120e+00 -1.09552228e+00 -4.23738539e-01 1.08880270e+00
-1.09509563e+00 -3.16808254e-01 -2.71240007e-02 8.08133960e-01
-2.19190598e-01 8.13371778e-01 -7.07237542e-01 -8.22856903e-01
-6.59102321e-01 -1.32627046e+00 1.07199562e+00 5.12487829e-01
1.25488102e-01 -6.97399437e-01 1.81715220e-01 8.34280491e-01
5.84628642e-01 -7.39662275e-02 1.59998536e-02 -1.12685420e-01
-6.48179054e-01 4.82065789e-02 -2.16739967e-01 5.45728385e-01
3.73058617e-01 2.90534884e-01 -9.67046440e-01 -4.53277022e-01
1.78074285e-01 1.61035568e-01 1.07650423e+00 6.97946250e-01
8.56768429e-01 1.25658467e-01 1.59476995e-01 9.32787359e-01
1.05569351e+00 1.57034487e-01 7.21720517e-01 4.42271173e-01
9.22143400e-01 7.11878315e-02 6.61350548e-01 5.26529193e-01
1.20214224e-01 8.40090692e-01 -1.66977905e-02 -2.61429638e-01
-6.99474514e-01 3.28709692e-01 6.55747533e-01 7.21519589e-01
2.50007827e-02 -4.19732243e-01 -7.82439530e-01 6.14400916e-02
-1.83355355e+00 -1.06745541e+00 -4.85680282e-01 2.12730861e+00
5.16291976e-01 -1.25685826e-01 -1.63371161e-01 1.98663473e-01
1.01987684e+00 3.36922169e-01 -3.14888895e-01 -1.76130250e-01
-4.66311991e-01 1.71137348e-01 2.77809650e-01 9.94140208e-01
-1.20975959e+00 1.15712106e+00 5.44508123e+00 6.94667280e-01
-1.39231896e+00 -4.99332659e-02 5.70347071e-01 -1.56295329e-01
2.81853616e-01 -1.35996670e-01 -6.14051282e-01 5.43088019e-01
7.75475502e-01 -9.66126770e-02 4.74084318e-01 5.10321930e-02
9.15801466e-01 -2.81343758e-01 -1.95803881e-01 1.22055519e+00
3.38635176e-01 -1.27303791e+00 -2.26866171e-01 -6.69642612e-02
6.11238062e-01 3.10190450e-02 4.89253066e-02 -1.86000317e-01
6.84282109e-02 -8.06416512e-01 6.20342374e-01 8.20763528e-01
8.84581029e-01 -4.99160022e-01 6.24249935e-01 2.51476038e-02
-1.44859302e+00 2.44803235e-01 -1.72797039e-01 1.03457652e-01
7.15043396e-02 8.95468175e-01 -6.42994761e-01 9.20564771e-01
8.34330142e-01 1.11753201e+00 -4.86679703e-01 1.15120041e+00
-2.63625570e-02 7.23245800e-01 -3.16809744e-01 3.88754308e-01
-1.88348636e-01 -5.30698121e-01 9.28794384e-01 1.49181902e+00
6.82709396e-01 2.02256039e-01 2.05840677e-01 3.36549103e-01
1.41725823e-01 1.21661171e-01 -4.33502614e-01 2.72165596e-01
1.23731732e-01 1.36473668e+00 -6.13771141e-01 -5.15352428e-01
-2.37936258e-01 1.36288619e+00 -2.55147606e-01 4.70833480e-01
-8.27444971e-01 -2.12765664e-01 6.05756640e-01 -5.30620515e-02
2.57912457e-01 -3.34188998e-01 -4.17806804e-01 -1.70314085e+00
1.49167562e-02 -9.59065497e-01 2.37821132e-01 -1.09147263e+00
-9.55075800e-01 5.97132802e-01 -4.18757290e-01 -1.43579185e+00
1.36242285e-01 -3.36163193e-01 -7.02420890e-01 1.18485808e+00
-1.55567074e+00 -8.06667626e-01 -7.42637098e-01 7.93451846e-01
9.61832166e-01 -3.70599002e-01 5.20087957e-01 3.72146130e-01
-9.53583062e-01 2.62946814e-01 2.80669034e-01 7.23589435e-02
1.16320837e+00 -1.12479126e+00 3.28432560e-01 1.59200561e+00
3.81957069e-02 3.14507693e-01 7.37099588e-01 -7.92537272e-01
-1.51896739e+00 -1.10196185e+00 7.63646126e-01 -1.87350735e-01
5.15622973e-01 -7.72964731e-02 -9.36135828e-01 2.89059371e-01
4.46328342e-01 5.33839054e-02 2.11390987e-01 -6.15066826e-01
-2.02205271e-01 -9.03749540e-02 -9.86737370e-01 4.64722216e-01
5.08282244e-01 -3.87214512e-01 -4.06896055e-01 2.57003009e-01
4.23887789e-01 -9.66511428e-01 -5.82064748e-01 1.92434475e-01
3.29653740e-01 -1.17931914e+00 8.22466433e-01 1.92811221e-01
2.25211486e-01 -8.62736404e-01 -2.84415692e-01 -1.17704475e+00
-1.67585492e-01 -1.03500915e+00 -1.18118757e-02 1.15266800e+00
-1.33655090e-02 -5.19239545e-01 6.47428095e-01 1.62090510e-01
-1.54577225e-01 -5.72549924e-02 -8.22525859e-01 -4.30674404e-01
-3.02292734e-01 -4.36373085e-01 1.02080695e-01 9.17010963e-01
-3.05257380e-01 8.30339342e-02 -5.82886279e-01 3.47268432e-01
8.06694567e-01 5.53785749e-02 8.46913755e-01 -7.30405152e-01
-2.40392104e-01 -3.15936178e-01 -2.32461303e-01 -9.93569493e-01
1.78328399e-02 -5.70410848e-01 1.35747358e-01 -1.40064752e+00
-2.89737191e-02 8.59631971e-02 -1.71224236e-01 1.79160222e-01
-4.15882438e-01 3.88975084e-01 5.56927085e-01 4.72310275e-01
-4.29100662e-01 4.02363092e-01 1.20915484e+00 -1.41172528e-01
-2.66405046e-01 -7.16397837e-02 -4.00934845e-01 4.62723821e-01
1.02882600e+00 -3.10776271e-02 -2.86300872e-02 -3.78502667e-01
-3.50350559e-01 1.54878512e-01 3.29083234e-01 -1.21456587e+00
2.66152591e-01 -1.51089475e-01 5.70126534e-01 -7.04918444e-01
4.71056193e-01 -6.62606537e-01 9.96382982e-02 4.52900946e-01
-1.42989784e-01 1.64379463e-01 2.94827729e-01 1.97552294e-01
-2.47883797e-01 -5.58986212e-04 1.00627732e+00 1.38414145e-01
-4.69480097e-01 2.42717028e-01 -4.61074620e-01 -2.71339476e-01
6.30435884e-01 -1.63561374e-01 -6.43221200e-01 -5.29500544e-01
-4.57537442e-01 8.90136436e-02 5.20310283e-01 2.94429064e-01
7.97989011e-01 -8.54160607e-01 -9.91937816e-01 2.56589562e-01
-3.29772800e-01 -1.85732499e-01 4.58972603e-01 1.07316983e+00
-7.52942264e-01 1.78285569e-01 1.72890857e-01 -7.42548525e-01
-1.58866775e+00 2.70264018e-02 5.60782135e-01 -2.24434465e-01
-1.00540709e+00 7.53944457e-01 -1.61260575e-01 2.62109600e-02
8.17181394e-02 -8.92794654e-02 -5.88931367e-02 -2.03909189e-01
8.74615073e-01 4.45086658e-01 3.04251671e-01 -9.92217004e-01
-3.19772065e-01 9.22825694e-01 -5.67931794e-02 -3.67681444e-01
1.03901660e+00 -6.18757784e-01 8.71211756e-03 1.68173313e-01
9.73275483e-01 3.25382382e-01 -1.56764114e+00 -9.48052406e-02
-4.28359300e-01 -7.88002312e-01 4.94741172e-01 -8.13028634e-01
-1.14663446e+00 6.66001499e-01 8.08537066e-01 5.62564768e-02
1.40203667e+00 -7.96821535e-01 7.85004735e-01 1.05321288e-01
-3.12426299e-01 -9.19133127e-01 1.39290243e-01 8.70330393e-01
7.59279609e-01 -9.69334602e-01 1.66741133e-01 -3.15516472e-01
-5.17520547e-01 1.40185571e+00 4.15759087e-01 1.33569483e-02
4.25771028e-01 5.02918839e-01 5.68483531e-01 1.14041246e-01
-5.05314708e-01 -2.49446914e-01 3.17204505e-01 3.45113009e-01
5.78580439e-01 -2.23677799e-01 -1.19849451e-01 -2.04756573e-01
-2.90607244e-01 9.06711724e-03 7.67146647e-01 8.75373840e-01
-4.97623503e-01 -9.57270980e-01 -9.68703806e-01 2.12360471e-01
-4.14833605e-01 -3.89238834e-01 -2.81685531e-01 1.64163843e-01
-8.09240937e-02 1.32936883e+00 -1.01104200e-01 -5.03709137e-01
2.83427179e-01 4.96608838e-02 4.23404723e-01 -2.11379871e-01
-6.61000788e-01 7.10360110e-01 -3.59265618e-02 -6.29885495e-01
-5.11995494e-01 -8.42049599e-01 -1.17294800e+00 -5.78750789e-01
-2.45787784e-01 5.48176505e-02 5.95595956e-01 8.66975904e-01
3.52991879e-01 4.92999941e-01 8.53955925e-01 -1.11448705e+00
1.34594455e-01 -8.36000621e-01 -4.61033881e-01 4.85459954e-01
6.70861244e-01 -2.15880156e-01 -7.84244239e-01 5.13321340e-01] | [11.121133804321289, -2.034360408782959] |
96debf53-60af-4e1f-8f84-cce95396bec7 | seeing-behind-objects-for-3d-multi-object | 2012.08197 | null | https://arxiv.org/abs/2012.08197v2 | https://arxiv.org/pdf/2012.08197v2.pdf | Seeing Behind Objects for 3D Multi-Object Tracking in RGB-D Sequences | Multi-object tracking from RGB-D video sequences is a challenging problem due to the combination of changing viewpoints, motion, and occlusions over time. We observe that having the complete geometry of objects aids in their tracking, and thus propose to jointly infer the complete geometry of objects as well as track them, for rigidly moving objects over time. Our key insight is that inferring the complete geometry of the objects significantly helps in tracking. By hallucinating unseen regions of objects, we can obtain additional correspondences between the same instance, thus providing robust tracking even under strong change of appearance. From a sequence of RGB-D frames, we detect objects in each frame and learn to predict their complete object geometry as well as a dense correspondence mapping into a canonical space. This allows us to derive 6DoF poses for the objects in each frame, along with their correspondence between frames, providing robust object tracking across the RGB-D sequence. Experiments on both synthetic and real-world RGB-D data demonstrate that we achieve state-of-the-art performance on dynamic object tracking. Furthermore, we show that our object completion significantly helps tracking, providing an improvement of $6.5\%$ in mean MOTA. | ['Matthias Nießner', 'Angela Dai', 'Niloy J. Mitra', 'Yu-Shiang Wong', 'Norman Müller'] | 2020-12-15 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Muller_Seeing_Behind_Objects_for_3D_Multi-Object_Tracking_in_RGB-D_Sequences_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Muller_Seeing_Behind_Objects_for_3D_Multi-Object_Tracking_in_RGB-D_Sequences_CVPR_2021_paper.pdf | cvpr-2021-1 | ['3d-multi-object-tracking'] | ['computer-vision'] | [-2.33942643e-01 -4.15299028e-01 1.11221984e-01 1.74104907e-02
-6.43475831e-01 -1.07838285e+00 2.88134664e-01 -1.79994985e-01
-2.49778807e-01 2.41953596e-01 -8.62928331e-02 2.45215386e-01
1.69321433e-01 -4.09593552e-01 -9.94010091e-01 -6.36551499e-01
-2.07030233e-02 7.24186718e-01 6.93010926e-01 2.61827447e-02
-9.22454000e-02 8.34887505e-01 -1.48187280e+00 -6.53520897e-02
2.08067790e-01 8.34441781e-01 2.65805125e-01 9.29125607e-01
1.62515625e-01 5.05163550e-01 -2.85886228e-01 -2.42554381e-01
6.21320605e-01 5.82502745e-02 -3.94975811e-01 7.71710217e-01
1.13372028e+00 -6.43791676e-01 -8.49696755e-01 1.01485157e+00
1.81102648e-01 2.70650029e-01 2.55363584e-01 -1.25138259e+00
-2.98184782e-01 -2.28674054e-01 -8.42799842e-01 1.22964136e-01
6.09798253e-01 3.14444184e-01 7.69633949e-01 -8.50675464e-01
8.81678820e-01 1.25373042e+00 6.76795185e-01 7.00636446e-01
-1.10692000e+00 -5.56881011e-01 4.34640050e-01 -5.44298664e-02
-1.39173651e+00 -5.27321756e-01 8.84908080e-01 -5.79486132e-01
5.23744106e-01 2.25101903e-01 1.05530775e+00 5.67636371e-01
7.26322532e-02 7.80331850e-01 4.39416647e-01 -1.60900325e-01
-8.76934975e-02 -1.87921554e-01 -7.19616711e-02 9.22572374e-01
4.01992679e-01 6.77379444e-02 -5.87237000e-01 -5.85267209e-02
9.70961571e-01 6.72654688e-01 -3.39766383e-01 -1.07623875e+00
-1.44501376e+00 2.70067632e-01 5.52596569e-01 -6.53695166e-02
-2.47078151e-01 4.96194363e-01 -1.04799680e-01 -1.51017249e-01
1.92602485e-01 -1.59246027e-01 -4.30079341e-01 1.08320005e-01
-4.57040906e-01 5.51281273e-01 4.58820164e-01 1.45812869e+00
7.57393658e-01 -4.56860363e-02 8.40406269e-02 7.76694119e-02
5.43986440e-01 1.16106963e+00 -2.11417124e-01 -1.50604153e+00
3.98026884e-01 5.29354811e-01 6.74427748e-01 -1.07101762e+00
-3.01122636e-01 -3.35936576e-01 -5.21137536e-01 4.11194652e-01
7.20046937e-01 -9.95852146e-03 -7.65522897e-01 1.83816278e+00
8.78104687e-01 4.86463308e-01 -2.18505263e-01 9.91872609e-01
3.44523013e-01 4.13469255e-01 -2.53049910e-01 -1.87669441e-01
1.54438925e+00 -7.50974298e-01 -4.89593923e-01 -3.61819178e-01
2.23742843e-01 -7.49521852e-01 4.11269516e-01 2.04334587e-01
-1.20672703e+00 -8.24685514e-01 -6.52591944e-01 2.30773948e-02
1.44844860e-01 -1.73376620e-01 5.60544729e-01 3.99838835e-01
-7.68299460e-01 4.00107652e-01 -1.33496261e+00 -1.96650416e-01
3.30817789e-01 4.07908320e-01 -5.55196822e-01 -3.28044057e-01
-3.93673837e-01 6.77439690e-01 1.86974421e-01 7.35603794e-02
-1.06806946e+00 -8.91159594e-01 -8.74332547e-01 -2.92767942e-01
6.80609047e-01 -9.27818775e-01 1.12239802e+00 -5.43236792e-01
-8.97921443e-01 7.86216915e-01 -5.47470272e-01 8.16520490e-03
6.23221338e-01 -5.54951131e-01 -1.47420034e-01 2.28144571e-01
-7.16424510e-02 6.25088930e-01 7.97548413e-01 -1.58425057e+00
-9.47434068e-01 -8.51581752e-01 3.23509239e-02 3.75464588e-01
-8.05489048e-02 -2.25955009e-01 -1.02334738e+00 -3.17581117e-01
5.85374177e-01 -1.34914732e+00 -1.69871122e-01 8.77231777e-01
-1.67498484e-01 -4.04279158e-02 1.32628691e+00 -5.84819376e-01
6.12055182e-01 -2.12020993e+00 5.27547002e-01 -1.47282809e-01
5.50061345e-01 -9.47970375e-02 6.65718019e-02 -2.27278218e-01
2.74938077e-01 -3.89558583e-01 1.98814884e-01 -6.52018547e-01
-1.89844325e-01 3.86946946e-01 -1.61414698e-01 9.18559909e-01
4.53955829e-02 9.82810616e-01 -1.12389636e+00 -3.39906156e-01
5.05782425e-01 7.42504954e-01 -4.97932851e-01 1.98167250e-01
-4.33485836e-01 8.38785350e-01 -6.48352981e-01 8.42599571e-01
7.90189266e-01 -3.90157789e-01 4.08591181e-02 -4.66647029e-01
-4.03870381e-02 -3.45271289e-01 -1.51942301e+00 2.01475763e+00
2.33917907e-02 6.19211972e-01 1.54241443e-01 -4.18991506e-01
4.28217471e-01 2.55066514e-01 1.03811955e+00 -3.20757568e-01
9.56941862e-03 -2.03073062e-02 -1.79285601e-01 -3.35417628e-01
4.79115665e-01 4.45560971e-03 1.91915661e-01 3.49801421e-01
-2.02176899e-01 -4.37726267e-02 9.95263010e-02 2.22751960e-01
1.06291103e+00 3.93780649e-01 -1.15903884e-01 9.13568437e-02
2.54426479e-01 -4.12131958e-02 6.96088314e-01 4.70466554e-01
-3.19110900e-01 5.77374458e-01 -2.53036529e-01 -6.36505067e-01
-1.21123838e+00 -1.44363511e+00 1.23063512e-01 6.97358608e-01
6.54410720e-01 -5.01890071e-02 -3.02326649e-01 -4.45050985e-01
4.30288672e-01 2.92871650e-02 -5.54246604e-01 6.70679137e-02
-1.03397322e+00 -2.66294092e-01 -5.35831079e-02 5.88559389e-01
2.00145170e-01 -4.05770212e-01 -1.11653543e+00 2.45629355e-01
-3.94665688e-01 -1.58613014e+00 -7.89514363e-01 -1.28858119e-01
-1.03066540e+00 -1.36982000e+00 -7.44471610e-01 -5.74849606e-01
7.80355632e-01 9.93838847e-01 1.01168489e+00 1.49196282e-01
-4.43914771e-01 7.40969837e-01 -8.43169317e-02 1.10389637e-02
-2.72481382e-01 -6.46361887e-01 3.67278755e-01 7.37307742e-02
4.68363687e-02 -3.71735662e-01 -7.96900868e-01 5.24558187e-01
-5.27257323e-01 -1.41189350e-02 2.11721510e-01 1.74851701e-01
9.04686868e-01 -7.81468302e-02 -4.16062534e-01 -2.21247733e-01
-6.03962660e-01 -4.26062196e-02 -9.66310084e-01 3.64785612e-01
1.12851754e-01 -1.00282580e-01 7.97237307e-02 -6.58087254e-01
-7.22221136e-01 5.78829169e-01 3.72583747e-01 -1.25125742e+00
-1.11095406e-01 -4.51624930e-01 -7.34100044e-02 -3.43158811e-01
2.82855928e-01 2.36268565e-01 3.18983458e-02 -5.47328711e-01
5.13071120e-01 -1.12026976e-02 7.35214233e-01 -6.19477749e-01
1.13468015e+00 1.10102797e+00 2.46464431e-01 -6.71495140e-01
-9.61857438e-01 -7.64086366e-01 -9.76306617e-01 -6.02715790e-01
1.04208171e+00 -1.22929811e+00 -1.14419734e+00 4.09310043e-01
-1.23118365e+00 -1.77635923e-01 -1.64418757e-01 5.60150146e-01
-5.81614912e-01 3.76748770e-01 -3.72261852e-01 -7.52997339e-01
2.65158862e-02 -1.23063374e+00 1.43497789e+00 2.61314094e-01
1.15895428e-01 -9.59983528e-01 9.56704561e-03 1.48741320e-01
-7.53544047e-02 5.80797791e-01 3.49387586e-01 2.21417129e-01
-1.29689574e+00 -2.31884003e-01 -7.00848270e-03 -2.47853562e-01
3.14152062e-01 2.31123846e-02 -7.43530393e-01 -6.50401235e-01
-2.60130651e-02 1.30704924e-01 5.05314112e-01 5.39481819e-01
7.55738139e-01 -2.81444173e-02 -7.11318493e-01 7.99222827e-01
1.53182685e+00 7.27705099e-03 1.46174788e-01 3.89574096e-02
1.11810029e+00 3.46822858e-01 6.73398018e-01 4.69330430e-01
3.87236238e-01 1.14522815e+00 6.47563636e-01 1.44428402e-01
-3.99854749e-01 -8.73315558e-02 3.24924350e-01 5.65019071e-01
-1.71575531e-01 -8.25388357e-02 -9.15441930e-01 3.69776130e-01
-1.81055307e+00 -8.60538125e-01 -2.96534389e-01 2.26304460e+00
5.70641816e-01 6.30726740e-02 3.55072379e-01 -3.22753906e-01
7.24068284e-01 -6.48707002e-02 -9.42051768e-01 8.43575358e-01
-7.56774843e-02 -1.68261424e-01 6.99536622e-01 5.39293230e-01
-9.92942333e-01 7.88285315e-01 5.74849033e+00 9.78687182e-02
-8.88634443e-01 5.61157763e-02 -1.31460291e-03 -4.62991446e-01
-7.99789280e-02 4.84528318e-02 -1.12164414e+00 2.49878287e-01
3.70833039e-01 -1.69808835e-01 3.20313215e-01 7.13943005e-01
7.32196146e-04 2.49580573e-03 -1.26257062e+00 1.06254923e+00
1.75225899e-01 -1.25369442e+00 -2.79064104e-02 1.48175374e-01
7.90580630e-01 1.88994661e-01 -6.61540627e-02 -2.35513940e-01
3.34684968e-01 -2.34837398e-01 1.19564605e+00 5.57158828e-01
5.17752290e-01 -4.05250579e-01 1.76648527e-01 3.20469648e-01
-1.69683671e+00 4.68849950e-02 -4.51218516e-01 1.11031219e-01
4.38108027e-01 2.75209457e-01 -3.57923448e-01 5.21698833e-01
7.80136287e-01 1.04182112e+00 -5.49916446e-01 1.13559079e+00
3.05695254e-02 -1.17146179e-01 -5.04381597e-01 3.99201870e-01
-2.09794655e-01 -1.08516425e-01 7.68280029e-01 7.38056600e-01
3.82631063e-01 4.71024811e-01 5.92004418e-01 6.76859856e-01
3.52377929e-02 -6.16250753e-01 -3.66730124e-01 2.91485697e-01
4.44102824e-01 1.13932455e+00 -8.64746809e-01 -3.29460919e-01
-4.49111640e-01 9.80004251e-01 1.28757462e-01 4.21589315e-01
-9.78802741e-01 2.67282754e-01 1.13910890e+00 1.27055600e-01
6.96016252e-01 -7.26502657e-01 1.24702401e-01 -1.40628481e+00
3.23814124e-01 -5.59881747e-01 2.45314181e-01 -9.29346561e-01
-8.52543056e-01 3.47603351e-01 -1.40818775e-01 -1.64488208e+00
4.93784323e-02 -5.97689748e-01 -6.98120743e-02 5.62549114e-01
-1.36479425e+00 -1.23975551e+00 -6.97103262e-01 8.45117927e-01
5.03913641e-01 3.67621303e-01 4.85192090e-01 4.01710808e-01
-1.43639609e-01 7.85302371e-02 1.01427905e-01 3.19544375e-01
5.60026586e-01 -1.06339347e+00 3.70995760e-01 8.85896504e-01
4.92296636e-01 6.06793761e-01 7.21532702e-01 -7.36957610e-01
-2.25964451e+00 -1.11704183e+00 1.94477394e-01 -1.13580406e+00
4.65464890e-01 -4.94266391e-01 -8.58320117e-01 1.01533270e+00
-2.70268708e-01 5.81590474e-01 2.10987717e-01 -2.18438700e-01
-5.01188755e-01 -8.65691304e-02 -8.50332379e-01 4.75234926e-01
1.37787735e+00 -4.00183827e-01 -3.96312028e-01 4.59846258e-01
8.67589235e-01 -1.05848670e+00 -8.36443961e-01 1.30900502e-01
8.04546773e-01 -8.20618570e-01 1.48649228e+00 -6.05270743e-01
-1.03269324e-01 -8.66389632e-01 -3.74394596e-01 -8.70732844e-01
-2.51325160e-01 -5.10835469e-01 -7.14373648e-01 8.57181966e-01
-2.64567286e-01 -8.32350552e-02 1.06776023e+00 8.48777294e-01
6.03113510e-03 -4.01587635e-01 -8.32917750e-01 -8.36335540e-01
-2.97866791e-01 -5.44464469e-01 2.23291501e-01 6.88339412e-01
-6.78669930e-01 -1.41904324e-01 -4.55605179e-01 6.94665015e-01
1.38563037e+00 4.71117616e-01 1.20095325e+00 -1.21298873e+00
-4.01907444e-01 -1.30419031e-01 -6.92759156e-01 -1.56599140e+00
2.71493383e-02 -4.37543094e-01 1.45757213e-01 -1.19919515e+00
4.49340582e-01 -5.73842049e-01 2.52632610e-02 3.46602947e-01
-3.69476885e-01 4.60479498e-01 5.93948722e-01 3.28166574e-01
-9.18937624e-01 3.78486514e-01 1.44143200e+00 -9.62984934e-02
-6.73500225e-02 9.34847444e-02 -2.67447054e-01 7.57908225e-01
1.29420102e-01 -5.36771178e-01 -1.06351040e-02 -8.63811076e-01
-1.69094339e-01 3.56445312e-01 7.84536779e-01 -1.07945025e+00
3.70131999e-01 -1.28533259e-01 7.87582636e-01 -9.26293314e-01
8.32196534e-01 -1.29293478e+00 6.15226984e-01 7.09140837e-01
1.36482209e-01 2.05712378e-01 3.79959524e-01 9.25981462e-01
2.38404274e-01 2.74630606e-01 7.69739509e-01 -2.88150609e-01
-9.21216488e-01 8.49671364e-01 3.62743258e-01 1.22776210e-01
1.26668727e+00 -4.68101382e-01 -3.12886126e-02 -1.99775010e-01
-9.82125521e-01 4.63397861e-01 9.17836726e-01 5.96170545e-01
6.72362506e-01 -1.43702126e+00 -5.92535734e-01 3.51093471e-01
8.04817379e-02 4.22476411e-01 4.11443919e-01 8.26184750e-01
-3.78380299e-01 1.66580543e-01 -2.25360289e-01 -1.37313938e+00
-1.65503836e+00 5.94519436e-01 4.27032709e-01 2.70680070e-01
-1.01542532e+00 8.51727724e-01 4.08755451e-01 2.89085843e-02
4.70597416e-01 -3.37849855e-01 4.23528910e-01 -1.97012156e-01
7.74696469e-01 3.64578009e-01 -2.24538207e-01 -8.89795125e-01
-5.24053037e-01 1.31658614e+00 2.65267119e-02 -1.09925598e-01
1.26596284e+00 -4.25909072e-01 1.55612096e-01 4.68606800e-01
1.16350222e+00 1.05738856e-01 -2.07010365e+00 -5.71951687e-01
-2.48738244e-01 -1.06304193e+00 -2.63250411e-01 -2.33385205e-01
-1.29716671e+00 7.35855639e-01 6.34426415e-01 -1.29774973e-01
7.30034888e-01 2.71528542e-01 7.33397067e-01 4.01258677e-01
7.40701497e-01 -3.64804983e-01 3.39800894e-01 4.10810947e-01
4.96443480e-01 -1.25558782e+00 2.24996850e-01 -4.68402505e-01
-2.91795909e-01 1.17561042e+00 4.74786460e-01 -1.24270953e-01
4.65557784e-01 2.62325168e-01 7.10297236e-03 -2.51445711e-01
-4.92382437e-01 -2.14301601e-01 3.46648663e-01 6.15605950e-01
2.74115894e-03 -1.47201240e-01 8.38406861e-01 -3.66585016e-01
1.87162727e-01 -2.15497419e-01 2.59211838e-01 1.10074282e+00
-4.96696174e-01 -9.51246560e-01 -6.91031635e-01 -8.14646930e-02
-1.92193076e-01 5.36107302e-01 -1.11841939e-01 9.91946280e-01
6.90094978e-02 6.69030905e-01 2.44332463e-01 -1.24293983e-01
5.98357916e-01 -2.04503760e-01 1.10082257e+00 -5.06619096e-01
-8.31731185e-02 4.02701199e-01 -4.67312872e-01 -7.98541188e-01
-7.39937067e-01 -1.08860826e+00 -1.34967577e+00 -3.99856776e-01
-3.97086650e-01 -1.23299949e-01 5.75343788e-01 8.37221742e-01
2.17900977e-01 6.14272892e-01 4.32271868e-01 -1.24990392e+00
-2.72952795e-01 -3.69775742e-01 -4.88879353e-01 7.32737064e-01
9.27350163e-01 -8.99404287e-01 -1.39669031e-01 5.00472784e-01] | [7.036330699920654, -2.281648874282837] |
5fa9857c-3f12-48eb-b6d8-a458918a2643 | efficiency-in-ambiguity-two-models-of | null | null | https://aclanthology.org/W15-0118 | https://aclanthology.org/W15-0118.pdf | Efficiency in Ambiguity: Two Models of Probabilistic Semantics for Natural Language | null | ['Daoud Clarke', 'Bill Keller'] | 2015-04-01 | null | null | null | ws-2015-4 | ['learning-semantic-representations'] | ['methodology'] | [-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] | [-7.392161846160889, 3.883859872817993] |
5abeb78e-b0ae-4a71-96c8-560133a5be01 | deep-image-matting | 1703.03872 | null | http://arxiv.org/abs/1703.03872v3 | http://arxiv.org/pdf/1703.03872v3.pdf | Deep Image Matting | Image matting is a fundamental computer vision problem and has many
applications. Previous algorithms have poor performance when an image has
similar foreground and background colors or complicated textures. The main
reasons are prior methods 1) only use low-level features and 2) lack high-level
context. In this paper, we propose a novel deep learning based algorithm that
can tackle both these problems. Our deep model has two parts. The first part is
a deep convolutional encoder-decoder network that takes an image and the
corresponding trimap as inputs and predict the alpha matte of the image. The
second part is a small convolutional network that refines the alpha matte
predictions of the first network to have more accurate alpha values and sharper
edges. In addition, we also create a large-scale image matting dataset
including 49300 training images and 1000 testing images. We evaluate our
algorithm on the image matting benchmark, our testing set, and a wide variety
of real images. Experimental results clearly demonstrate the superiority of our
algorithm over previous methods. | ['Brian Price', 'Scott Cohen', 'Ning Xu', 'Thomas Huang'] | 2017-03-10 | deep-image-matting-1 | http://openaccess.thecvf.com/content_cvpr_2017/html/Xu_Deep_Image_Matting_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Xu_Deep_Image_Matting_CVPR_2017_paper.pdf | cvpr-2017-7 | ['semantic-image-matting'] | ['computer-vision'] | [ 3.95371377e-01 -3.36647183e-01 -7.22174440e-03 -3.85365099e-01
-4.81852889e-01 -1.21256649e-01 3.41094524e-01 -2.75766194e-01
-2.01509416e-01 5.22093415e-01 1.00446217e-01 -2.17941374e-01
5.58737755e-01 -9.09354508e-01 -1.01569331e+00 -6.67712629e-01
2.90587544e-01 2.99428552e-01 6.03653848e-01 -4.91782986e-02
2.45935157e-01 1.98848575e-01 -1.28217542e+00 8.25107932e-01
9.68219399e-01 1.21437013e+00 3.71737659e-01 7.31819510e-01
-2.03889012e-01 1.06089807e+00 -4.84003276e-01 -5.15336573e-01
4.65549618e-01 -5.61920881e-01 -6.14712000e-01 4.43766803e-01
4.94306386e-01 -4.19912249e-01 -5.02696097e-01 1.30086184e+00
1.80574328e-01 -2.36088499e-01 4.44827735e-01 -1.29295361e+00
-7.88527310e-01 5.77948213e-01 -1.13326454e+00 1.42815113e-01
-6.58441111e-02 3.81828904e-01 6.07513845e-01 -8.46718132e-01
2.03532487e-01 1.37180865e+00 6.85837746e-01 1.93808645e-01
-1.13577247e+00 -6.86400771e-01 8.60840082e-02 3.38573635e-01
-1.22044170e+00 -1.32915929e-01 7.56631076e-01 -4.47830498e-01
5.36606789e-01 3.40990573e-02 7.07912087e-01 7.44460762e-01
4.63876307e-01 9.98005986e-01 1.39036083e+00 -2.90541261e-01
6.12794906e-02 -9.66423154e-02 -1.56938702e-01 9.38161314e-01
2.12687284e-01 -8.38260353e-02 -2.39887714e-01 9.69194695e-02
1.00230920e+00 2.00527862e-01 -2.52744108e-01 -1.94173113e-01
-1.50421643e+00 6.82465494e-01 8.15517545e-01 1.23722129e-01
-4.13187832e-01 3.37696403e-01 2.15459034e-01 2.64340550e-01
3.85178030e-01 9.51444358e-02 -1.86907336e-01 5.30807339e-02
-1.03602064e+00 3.21234055e-02 6.53962672e-01 7.40584910e-01
1.02000141e+00 1.22263655e-01 -1.40906900e-01 8.88605237e-01
2.42857575e-01 4.52748120e-01 4.56820577e-01 -8.09124231e-01
6.53790951e-01 7.47330725e-01 -4.84765507e-02 -1.19727015e+00
8.90683569e-03 -3.33982140e-01 -1.27212346e+00 4.86817777e-01
4.37881589e-01 -2.22639535e-02 -1.34930432e+00 1.35265672e+00
8.73898044e-02 3.76774937e-01 -1.81872696e-01 9.54721868e-01
6.85447395e-01 1.06388676e+00 -2.84626514e-01 1.19832642e-01
1.14335525e+00 -1.52857602e+00 -7.11455643e-01 -5.16017079e-01
2.06722334e-01 -1.06343734e+00 1.10907829e+00 4.85172987e-01
-1.27174783e+00 -7.43081152e-01 -1.25171626e+00 -2.20721826e-01
-2.95481831e-01 3.23283285e-01 5.58302343e-01 4.20435637e-01
-1.12272048e+00 4.43984628e-01 -7.15740263e-01 -1.43250361e-01
5.69022179e-01 2.99239367e-01 -8.15315321e-02 -2.16104597e-01
-8.67609203e-01 7.35163271e-01 5.55007637e-01 2.11265162e-01
-8.70708704e-01 -4.31586415e-01 -9.15109634e-01 1.55162975e-01
2.19604269e-01 -6.55569553e-01 1.12478280e+00 -1.67089283e+00
-1.40375197e+00 8.92788947e-01 -9.84548330e-02 -3.96147490e-01
6.41638756e-01 -2.65162319e-01 -3.04346889e-01 5.02255782e-02
9.40001234e-02 7.05800474e-01 1.04366040e+00 -1.56775010e+00
-7.23960817e-01 -4.41985168e-02 -1.29554138e-01 2.00417832e-01
-1.42799288e-01 -1.16130017e-01 -8.55916440e-01 -9.79335904e-01
1.14864148e-01 -7.13557661e-01 -2.55493194e-01 1.32796884e-01
-4.79698330e-01 3.00478250e-01 8.86625588e-01 -7.55652428e-01
1.02460074e+00 -2.20684242e+00 2.08340883e-01 2.47231796e-01
3.76134843e-01 4.05670345e-01 -2.91749328e-01 1.94455177e-01
-2.84200013e-01 -8.72107595e-02 -2.59133339e-01 -3.89657803e-02
-3.39771003e-01 1.45956933e-01 -4.15917248e-01 2.73680121e-01
2.33017430e-01 1.06937551e+00 -6.57012343e-01 -5.87969661e-01
3.24347019e-01 2.90873259e-01 -4.16927844e-01 3.87314886e-01
-2.80487776e-01 1.41972989e-01 -8.85283425e-02 6.15362048e-01
9.40464199e-01 -4.70235169e-01 -5.92646934e-02 -5.44472575e-01
-5.98742850e-02 1.13077294e-02 -1.02599895e+00 1.50863171e+00
-1.96519822e-01 8.33564878e-01 1.44652566e-02 -9.27341819e-01
9.29868519e-01 -1.29701152e-01 3.69141877e-01 -6.73822403e-01
2.92331517e-01 6.66780025e-02 1.59331918e-01 -5.08263946e-01
2.93562472e-01 4.51643728e-02 3.17711473e-01 4.95411485e-01
-2.71054327e-01 -4.93636467e-02 1.37898952e-01 2.00131074e-01
9.79053259e-01 -8.19684193e-02 2.50603743e-02 -1.61196038e-01
4.47821826e-01 -9.88492295e-02 6.70371592e-01 5.91168284e-01
-6.17916994e-02 9.76606846e-01 7.28426456e-01 -8.26667428e-01
-1.21825850e+00 -9.50577021e-01 2.74273396e-01 8.03821623e-01
5.95974803e-01 -3.60555619e-01 -8.52136791e-01 -5.33021748e-01
-1.24541961e-01 4.03473049e-01 -9.80994284e-01 -1.36226758e-01
-6.41202033e-01 -8.64188015e-01 2.88085848e-01 7.44829357e-01
1.13411188e+00 -1.00878704e+00 -4.08172429e-01 -6.46997839e-02
-2.91516393e-01 -1.03648841e+00 -6.90070748e-01 -8.44392851e-02
-8.56248260e-01 -1.05862153e+00 -7.02621043e-01 -1.18293691e+00
8.82918477e-01 4.78585660e-01 1.21055853e+00 4.53834653e-01
-1.35322422e-01 -2.93126136e-01 -2.93979049e-01 -3.40459406e-01
-4.82379645e-01 -1.16739966e-01 -3.91134441e-01 2.91469663e-01
2.07235977e-01 -3.47664803e-01 -6.05240762e-01 4.10160810e-01
-1.10329700e+00 7.73158908e-01 1.14811218e+00 1.03058779e+00
6.17769778e-01 1.40965372e-01 2.18880892e-01 -1.00143051e+00
3.33572239e-01 -2.31335059e-01 -6.29352629e-01 4.20960337e-01
-4.11821544e-01 6.29065111e-02 4.71386820e-01 -3.42482746e-01
-1.00298190e+00 2.42506266e-01 -1.86206788e-01 -4.87694085e-01
1.43947266e-02 5.16056299e-01 -3.33649844e-01 -5.96002936e-02
4.06707555e-01 3.17132443e-01 6.48146495e-02 -4.71892744e-01
1.80474684e-01 4.98439461e-01 7.55146205e-01 -4.28099632e-01
1.10842526e+00 3.94978493e-01 -2.87446320e-01 -6.18780434e-01
-1.00763714e+00 -7.82067403e-02 -6.89226389e-01 -1.59640148e-01
8.94086838e-01 -9.42105591e-01 -4.42090243e-01 8.88439417e-01
-1.11230242e+00 -6.87855601e-01 6.70630485e-03 2.67334849e-01
-4.73460734e-01 4.91379827e-01 -8.97869229e-01 -3.32418352e-01
-3.97725642e-01 -1.38875115e+00 9.41841245e-01 4.74559128e-01
3.95355582e-01 -7.01731563e-01 -8.16150233e-02 3.69891256e-01
4.03071851e-01 4.32361722e-01 9.34795320e-01 -5.20654535e-03
-8.50603282e-01 9.01778340e-02 -7.82180190e-01 3.96893829e-01
2.82778591e-01 2.69828767e-01 -8.43298435e-01 -2.70648211e-01
-1.75606176e-01 -3.81164342e-01 1.38121235e+00 3.08695942e-01
1.67603719e+00 -3.52669209e-01 -2.20381767e-01 9.40931141e-01
1.49153626e+00 1.94882676e-01 1.16515493e+00 5.20106137e-01
9.48334754e-01 1.04258128e-01 5.83208740e-01 1.37257755e-01
2.65774786e-01 5.22953212e-01 5.74168444e-01 -7.49010742e-01
-2.78159916e-01 -2.17347816e-01 4.22306389e-01 9.92538452e-01
-1.12818763e-01 -1.23136215e-01 -7.98109412e-01 4.45916176e-01
-2.02343249e+00 -9.66324270e-01 -1.46186188e-01 1.94891381e+00
9.58161116e-01 3.14854324e-01 9.88402292e-02 3.85197662e-02
8.09109509e-01 1.44918993e-01 -6.37237847e-01 -2.91032135e-01
-1.25134483e-01 1.15105733e-01 4.06005859e-01 2.71342486e-01
-1.19418025e+00 9.86892700e-01 6.50456715e+00 6.41901672e-01
-1.19375837e+00 -2.40487367e-01 1.17128193e+00 1.86003193e-01
-8.89181048e-02 -1.98985577e-01 -3.35844278e-01 6.72518849e-01
4.06425178e-01 4.63129394e-02 4.88355070e-01 7.51858413e-01
-8.49433914e-02 -1.66887224e-01 -9.49642718e-01 1.17122638e+00
2.39081115e-01 -1.47331262e+00 2.64960885e-01 -8.43082070e-02
9.46163118e-01 3.53379920e-03 2.13375345e-01 1.59750417e-01
3.48450512e-01 -1.25170064e+00 8.67523432e-01 5.05978167e-01
8.89753759e-01 -7.28258014e-01 9.46998537e-01 1.27758116e-01
-1.20327926e+00 8.29722062e-02 -5.92937887e-01 -5.03987633e-02
-7.12034851e-02 6.82611406e-01 -6.02503896e-01 3.21319699e-01
7.82152176e-01 8.64870667e-01 -8.99794698e-01 1.28541207e+00
-2.31955469e-01 6.54516935e-01 -1.01042330e-01 2.25805566e-01
2.83423305e-01 -3.43054056e-01 -5.12596965e-02 1.32932937e+00
1.02538973e-01 -2.80329920e-02 2.84445077e-01 9.33730662e-01
-3.19479853e-01 -1.55661240e-01 -4.37289774e-01 -5.03265066e-03
1.57732904e-01 1.36506045e+00 -9.75194514e-01 -6.30927324e-01
-6.02754951e-01 1.30620062e+00 3.15824330e-01 4.45051551e-01
-1.07122707e+00 -4.58898872e-01 6.32143497e-01 7.85143748e-02
3.68395030e-01 -2.37515077e-01 -5.36806762e-01 -1.37517226e+00
-1.81727365e-01 -1.17004526e+00 2.25323871e-01 -9.91338909e-01
-1.23127031e+00 6.77421629e-01 -2.88300335e-01 -1.14859581e+00
1.51492760e-01 -7.84517884e-01 -8.68898571e-01 7.30069280e-01
-1.43133783e+00 -1.28248465e+00 -8.55621219e-01 5.73459744e-01
7.44105756e-01 -1.09652199e-01 3.64719182e-01 3.67837638e-01
-7.49308765e-01 4.18164313e-01 2.03821018e-01 6.25523448e-01
7.57963955e-01 -1.40644217e+00 7.22457767e-01 1.11041915e+00
9.65271145e-02 3.02599698e-01 4.77996558e-01 -6.70605719e-01
-1.24632704e+00 -1.31583261e+00 4.90754366e-01 -2.51552999e-01
5.37884116e-01 -4.52146262e-01 -1.03467786e+00 8.21735144e-01
6.54224038e-01 -1.14123255e-01 2.98593730e-01 -3.09163451e-01
-3.93855751e-01 -2.99213797e-01 -7.93342352e-01 6.49192870e-01
6.90475881e-01 -1.42863154e-01 -4.24408555e-01 2.21237212e-01
5.89364886e-01 -6.94986999e-01 -4.60884243e-01 3.38134825e-01
5.01500189e-01 -1.13606405e+00 9.06705678e-01 -3.07833761e-01
9.01019275e-01 -6.03838325e-01 8.93412828e-02 -1.32932377e+00
-4.81508046e-01 -3.40310693e-01 1.86714269e-02 1.04643178e+00
3.04865688e-01 -3.32995862e-01 7.51938760e-01 3.26806396e-01
-1.09395288e-01 -9.50011075e-01 -3.02344501e-01 -4.12667066e-01
-1.27584517e-01 -9.33120996e-02 6.76534355e-01 9.30379570e-01
-5.12312889e-01 3.69510233e-01 -6.75610900e-01 1.17601091e-02
6.61924779e-01 6.01916313e-01 9.87007201e-01 -7.58484125e-01
-2.72176027e-01 -4.32680666e-01 -3.58411998e-01 -1.14760184e+00
-1.96304858e-01 -6.46353245e-01 2.47870475e-01 -1.62378550e+00
6.72594368e-01 -3.34267259e-01 -2.62937099e-01 7.13746667e-01
-4.63798523e-01 5.96968591e-01 8.80756006e-02 1.42331317e-01
-5.84282637e-01 5.57340324e-01 1.35212362e+00 -4.35594797e-01
4.11821567e-02 -1.62256479e-01 -5.05068719e-01 7.88489461e-01
9.11151171e-01 -2.84627289e-01 -2.17605293e-01 -8.13133121e-01
1.25140712e-01 -3.15522879e-01 4.39545125e-01 -1.05707347e+00
9.67856497e-02 -4.58434761e-01 1.01610029e+00 -6.88736975e-01
1.40705973e-01 -7.26118505e-01 1.61756650e-01 7.00013995e-01
-3.04351330e-01 3.37278336e-01 2.22828776e-01 3.88084352e-01
-2.06768900e-01 -4.47209589e-02 1.17060649e+00 -1.48914561e-01
-8.20668697e-01 4.44574296e-01 -2.01960668e-01 -8.96095708e-02
1.00105858e+00 -1.73775911e-01 -4.41339225e-01 -4.71506596e-01
-2.68733650e-01 1.73557505e-01 8.11586678e-01 4.84503359e-01
8.92730653e-01 -1.43842959e+00 -8.10496032e-01 4.25291806e-01
-6.42839000e-02 8.16426724e-02 -1.78247809e-01 6.97869301e-01
-1.09208083e+00 -1.80497944e-01 -6.65912807e-01 -6.38962507e-01
-1.21023667e+00 6.12698495e-01 3.63763094e-01 -8.55951458e-02
-7.15324819e-01 8.56628835e-01 6.74929857e-01 1.12779975e-01
2.24262670e-01 -4.38907892e-01 1.14407972e-01 -4.49408293e-01
7.60969341e-01 3.79181057e-02 -2.09749773e-01 -7.09352672e-01
-1.29018992e-01 4.35705423e-01 -4.40387249e-01 -1.39225852e-02
1.37951005e+00 5.90820052e-03 -4.07389671e-01 2.60750830e-01
1.10289371e+00 -2.34901667e-01 -1.46620309e+00 -2.60553628e-01
-1.89161316e-01 -9.77202415e-01 4.02147435e-02 -7.00727761e-01
-1.65889525e+00 1.02812195e+00 6.31441653e-01 3.83805111e-02
1.40826976e+00 -2.48475939e-01 1.05783319e+00 1.53301924e-01
-6.60978258e-03 -1.09334922e+00 6.02373421e-01 4.38817501e-01
8.59723151e-01 -1.32219946e+00 6.52183965e-02 -3.50132316e-01
-7.23349988e-01 1.20801437e+00 1.00693178e+00 -2.36499324e-01
3.05326104e-01 4.54302162e-01 3.85238945e-01 -1.87717512e-01
-7.46710598e-01 -1.85080260e-01 2.67838627e-01 4.38292712e-01
4.08854365e-01 -2.12140098e-01 2.35290453e-02 3.46958935e-01
-1.98504068e-02 8.80752653e-02 5.05598843e-01 8.12138557e-01
-6.58393979e-01 -9.67175961e-01 -4.28856045e-01 5.40276885e-01
-3.85371953e-01 -1.37662468e-02 -5.32764018e-01 4.64919299e-01
1.68281153e-01 6.97127938e-01 2.45649740e-02 -7.42475808e-01
2.13762261e-02 -3.55571866e-01 4.27233696e-01 -4.48657960e-01
-4.83487695e-01 1.02212802e-01 -3.64107788e-01 -5.63823879e-01
-3.31694365e-01 -2.99304545e-01 -1.01886749e+00 -5.85707188e-01
-2.86502212e-01 -1.11359090e-01 5.22742927e-01 7.85023153e-01
1.62583724e-01 5.87773323e-01 6.60889566e-01 -8.70351911e-01
-1.69426680e-01 -8.45046401e-01 -4.94748980e-01 6.76343501e-01
3.31165731e-01 -4.00089264e-01 4.26069647e-02 6.10044241e-01] | [10.653619766235352, -0.9164931774139404] |
e4b1224f-95e9-4f75-9cc1-6104297d52e0 | retrosynthesis-prediction-with-conditional-1 | 2001.01408 | null | https://arxiv.org/abs/2001.01408v1 | https://arxiv.org/pdf/2001.01408v1.pdf | Retrosynthesis Prediction with Conditional Graph Logic Network | Retrosynthesis is one of the fundamental problems in organic chemistry. The task is to identify reactants that can be used to synthesize a specified product molecule. Recently, computer-aided retrosynthesis is finding renewed interest from both chemistry and computer science communities. Most existing approaches rely on template-based models that define subgraph matching rules, but whether or not a chemical reaction can proceed is not defined by hard decision rules. In this work, we propose a new approach to this task using the Conditional Graph Logic Network, a conditional graphical model built upon graph neural networks that learns when rules from reaction templates should be applied, implicitly considering whether the resulting reaction would be both chemically feasible and strategic. We also propose an efficient hierarchical sampling to alleviate the computation cost. While achieving a significant improvement of $8.1\%$ over current state-of-the-art methods on the benchmark dataset, our model also offers interpretations for the prediction. | ['Connor W. Coley', 'Hanjun Dai', 'Le Song', 'Bo Dai', 'Chengtao Li'] | 2020-01-06 | retrosynthesis-prediction-with-conditional | http://papers.nips.cc/paper/9090-retrosynthesis-prediction-with-conditional-graph-logic-network | http://papers.nips.cc/paper/9090-retrosynthesis-prediction-with-conditional-graph-logic-network.pdf | neurips-2019-12 | ['retrosynthesis'] | ['medical'] | [ 7.22588658e-01 3.14534307e-01 -6.25942945e-01 -3.31435025e-01
-1.95005298e-01 -7.61644185e-01 6.89442337e-01 6.70318365e-01
-7.80229196e-02 8.75254154e-01 -1.38223946e-01 -8.16045582e-01
1.99069027e-02 -1.16181827e+00 -7.14449644e-01 -6.50626063e-01
1.12384446e-01 5.33663273e-01 2.85936385e-01 -6.30233660e-02
3.66846174e-01 7.61672735e-01 -1.27348495e+00 2.34354064e-01
8.77126038e-01 1.06262004e+00 7.15615377e-02 4.23944563e-01
-1.15996093e-01 1.01235950e+00 -1.67014018e-01 -4.45934147e-01
2.46640295e-01 -4.83668685e-01 -8.64312589e-01 -2.91934788e-01
-7.77426064e-02 2.35486865e-01 -1.87935263e-01 1.06229842e+00
1.71213225e-01 2.64395535e-01 1.01036370e+00 -9.75861430e-01
-1.60312727e-01 8.59576702e-01 -5.66584151e-03 -1.80338338e-01
3.60256463e-01 2.11104136e-02 1.27917027e+00 -7.41570473e-01
7.63063133e-01 1.07141018e+00 2.36621816e-02 3.28381062e-01
-1.57353854e+00 -7.99512386e-01 3.97781044e-01 2.46305898e-01
-1.35906351e+00 -3.76527190e-01 8.08519185e-01 -5.33920765e-01
1.21790671e+00 1.70866281e-01 7.96841383e-01 6.68895483e-01
4.77547556e-01 2.72679001e-01 9.57540810e-01 -3.76914442e-01
5.79617321e-01 -2.00534359e-01 -1.76109657e-01 9.80788350e-01
3.15803170e-01 2.37269729e-01 -5.01694679e-01 -6.72975034e-02
4.63953108e-01 7.93538541e-02 -1.26452148e-01 -5.14063835e-01
-1.01275051e+00 9.91349638e-01 5.59794605e-01 2.59619027e-01
-2.63389051e-01 2.09750861e-01 3.45285743e-01 -1.93063617e-01
3.01265299e-01 9.77704227e-01 -4.79031742e-01 3.60093385e-01
-1.01469386e+00 2.29439601e-01 1.09636068e+00 9.93871748e-01
1.00113130e+00 -8.54110792e-02 3.15257758e-02 1.28542438e-01
3.48623335e-01 -1.67173058e-01 -1.06268905e-01 -4.36473668e-01
2.74301022e-01 1.00655162e+00 -4.98705618e-02 -8.83522749e-01
-4.45205569e-01 -1.37213677e-01 -7.06610441e-01 7.67447278e-02
4.79212582e-01 -1.06870998e-02 -1.09438241e+00 1.50925004e+00
3.11368644e-01 -4.76132566e-03 5.06588630e-02 4.64735150e-01
7.17324495e-01 1.16884398e+00 2.53639281e-01 -6.01164579e-01
9.60371852e-01 -9.07485962e-01 -5.52774549e-01 2.20321771e-02
6.76682591e-01 -6.58192635e-01 3.71214360e-01 7.41779208e-01
-9.60780680e-01 -1.35016650e-01 -1.40001166e+00 1.02945731e-03
-6.84975028e-01 2.70950496e-02 9.80434060e-01 6.99418843e-01
-6.55551434e-01 9.74188387e-01 -6.70509934e-01 -1.46963745e-01
4.13850658e-02 7.66609967e-01 -2.89651215e-01 -5.33373989e-02
-1.07238185e+00 9.16661501e-01 8.27484131e-01 2.27648973e-01
-1.26288617e+00 -5.28938472e-01 -8.55753660e-01 2.45617181e-01
1.06190383e+00 -5.27850389e-01 1.24905300e+00 -6.05132639e-01
-1.74230552e+00 6.24442458e-01 -2.29592070e-01 -5.23328245e-01
3.17870706e-01 3.71288598e-01 -4.23675954e-01 -8.45206976e-02
-6.08255416e-02 5.68751931e-01 6.40588582e-01 -9.39208508e-01
-4.11313713e-01 -2.15204611e-01 2.45306000e-01 -1.40437290e-01
2.74665058e-01 -1.06036320e-01 -3.73154819e-01 -3.27906132e-01
1.73530176e-01 -1.04605532e+00 -7.90907204e-01 -2.47673318e-01
-9.02387500e-01 -6.16640985e-01 2.10878417e-01 -1.05474092e-01
1.41297805e+00 -1.36100006e+00 3.48380595e-01 6.58128917e-01
1.53560579e-01 5.78120425e-02 2.82774240e-01 7.82483578e-01
-2.91162282e-01 4.25941110e-01 -2.09263697e-01 4.66267258e-01
-4.79615368e-02 -8.07204917e-02 -3.16720933e-01 3.30619603e-01
2.13810667e-01 5.21284521e-01 -8.52820814e-01 -2.35597789e-01
1.89601779e-01 2.20753159e-02 -7.20115781e-01 2.69627631e-01
-9.27117825e-01 3.29441756e-01 -4.98093516e-01 5.78980625e-01
2.96847492e-01 -4.72304970e-01 1.14398456e+00 -3.41103464e-01
-3.73038560e-01 6.08385623e-01 -1.14526498e+00 1.66302371e+00
-3.45967352e-01 -1.11188572e-02 -3.24365377e-01 -1.09354782e+00
1.12641430e+00 3.91087502e-01 3.23602885e-01 -4.53367800e-01
2.87624985e-01 2.87379533e-01 2.80467123e-01 -6.91720694e-02
3.29617172e-01 -4.10686880e-01 -1.16483778e-01 1.81954041e-01
2.68557929e-02 -4.54211831e-01 8.39444399e-01 1.72266766e-01
8.73471975e-01 3.52096796e-01 9.07319427e-01 -4.94228423e-01
8.41211140e-01 9.83140096e-02 9.13128316e-01 5.33249557e-01
2.30819374e-01 -3.25585790e-02 8.40824723e-01 -7.15820968e-01
-9.20930326e-01 -7.31537044e-01 3.12661111e-01 8.39265645e-01
1.70834422e-01 -9.13819671e-01 -6.20813072e-01 -7.57530332e-01
-2.46357858e-01 6.81024015e-01 -5.42680562e-01 -1.37511000e-01
-3.18278760e-01 -3.46040279e-01 1.69765934e-01 4.08841997e-01
8.87287483e-02 -8.54573488e-01 -9.53667760e-02 5.65655768e-01
1.57122985e-01 -9.30610478e-01 -4.66468453e-01 6.43474460e-01
-6.98748648e-01 -1.46089137e+00 -1.65249020e-01 -6.66783631e-01
8.41499865e-01 5.56400158e-02 9.61674631e-01 6.90069273e-02
-3.80997032e-01 -3.58267874e-01 -3.95752341e-02 -5.11397719e-01
-6.69571579e-01 3.16651464e-01 -5.63694537e-02 -1.45305069e-02
7.55739957e-03 -5.13510406e-01 -6.37316465e-01 2.32459173e-01
-7.68669248e-01 3.60516459e-01 4.42289799e-01 6.35902762e-01
9.24466729e-01 2.30106145e-01 3.90534848e-01 -1.08167624e+00
3.12430173e-01 -3.17369521e-01 -1.11132181e+00 5.64174950e-01
-1.02751088e+00 6.44490957e-01 1.01895046e+00 -9.01677310e-02
-9.64904308e-01 7.25858629e-01 8.80776793e-02 2.57352181e-02
7.00487345e-02 7.36311615e-01 -4.92522776e-01 1.67465694e-02
5.06672263e-01 8.27844664e-02 -4.97720808e-01 -1.08552538e-01
4.57416147e-01 -2.90749855e-02 -6.25121454e-03 -8.27187777e-01
5.34480691e-01 1.76133767e-01 7.90748298e-01 -4.98774469e-01
-6.97589576e-01 -3.63382012e-01 -5.20232081e-01 -1.44495994e-01
9.29929018e-01 -6.74547315e-01 -1.19545841e+00 -2.44038135e-01
-1.12442780e+00 -3.78824323e-01 2.56925553e-01 3.18057477e-01
-5.93825161e-01 3.72035891e-01 -3.25393766e-01 -7.90600121e-01
-3.79790366e-01 -1.58010817e+00 5.80153108e-01 -2.06711721e-02
-3.01325440e-01 -7.92261302e-01 1.20420549e-02 1.27577409e-01
-1.73601747e-01 3.51683825e-01 1.37166023e+00 -7.15916932e-01
-9.75301325e-01 -9.71203223e-02 -9.04956236e-02 -1.07368469e-01
2.83219427e-01 2.79096156e-01 -5.24978995e-01 -6.59729987e-02
-5.70050955e-01 -1.92931250e-01 9.03980434e-01 2.12599203e-01
1.39083374e+00 -4.58840281e-01 -6.18803144e-01 3.52779031e-01
1.43106425e+00 8.00838590e-01 5.48439503e-01 -1.22641236e-01
6.16584480e-01 5.91218829e-01 6.07864261e-01 3.88393879e-01
1.10021971e-01 6.24236107e-01 7.14824080e-01 1.67631626e-01
3.16676766e-01 -7.41917253e-01 1.49190560e-01 2.52361447e-01
-4.15856510e-01 -5.94578981e-01 -9.69885826e-01 1.32819265e-01
-1.88427162e+00 -9.35797155e-01 -1.06157459e-01 2.29156232e+00
1.02094889e+00 2.81151056e-01 1.16905317e-01 3.47408831e-01
6.64089262e-01 1.43091053e-01 -4.51339573e-01 -5.19830287e-01
3.77355725e-01 6.67831361e-01 6.83629155e-01 5.86716771e-01
-1.03778458e+00 1.29932690e+00 6.63739872e+00 8.93149853e-01
-1.36219800e+00 -5.69197536e-01 7.06163883e-01 2.04099223e-01
-3.56603026e-01 5.66123486e-01 -8.80813837e-01 1.58110306e-01
8.15153420e-01 -3.60698998e-01 5.51762938e-01 7.88274586e-01
3.22342873e-01 1.36328498e-02 -1.59531641e+00 5.26096761e-01
-2.20004484e-01 -1.90212405e+00 3.11827660e-01 7.05063865e-02
7.30908573e-01 -6.06860697e-01 -3.05258512e-01 -2.62768026e-02
4.80222523e-01 -1.16476095e+00 7.62820303e-01 2.56249815e-01
7.41568685e-01 -9.87258196e-01 8.78093764e-02 2.63868719e-01
-1.54890025e+00 1.00266054e-01 -2.79741585e-02 -1.32401496e-01
-1.31151319e-01 5.60820282e-01 -1.58221567e+00 8.56314778e-01
2.87555661e-02 7.16731429e-01 -1.69256911e-01 7.18770146e-01
-8.76513958e-01 6.07419372e-01 -1.70699999e-01 -6.25745058e-01
3.37227255e-01 -4.96373624e-01 6.76994622e-02 1.14514840e+00
3.77999276e-01 9.88948867e-02 4.96009707e-01 1.05234289e+00
-2.91081607e-01 3.05877686e-01 -8.11405301e-01 -5.02448559e-01
3.45354706e-01 1.10214388e+00 -1.17760003e+00 -3.67986381e-01
-2.19034716e-01 5.94059825e-01 3.33071262e-01 3.53468865e-01
-7.55028725e-01 -4.23490554e-01 2.92335033e-01 1.08749166e-01
4.29862440e-01 -4.00498003e-01 -1.01697229e-01 -8.56092811e-01
-2.86878467e-01 -6.13097966e-01 3.60983700e-01 -3.92496556e-01
-6.95587814e-01 2.69209266e-01 -3.39259773e-01 -8.84999514e-01
-1.91778839e-01 -9.14079130e-01 -6.38420105e-01 5.87881386e-01
-1.38024223e+00 -8.31853747e-01 1.46568611e-01 2.50506788e-01
4.38691527e-01 6.99211657e-02 9.15528417e-01 -3.44955176e-02
-7.13138521e-01 1.67635486e-01 -3.14086139e-01 -2.94549584e-01
5.43052793e-01 -1.25318074e+00 1.09782808e-01 8.66452277e-01
1.18116908e-01 7.81679451e-01 7.56242812e-01 -6.56764507e-01
-1.70543063e+00 -1.27045333e+00 9.84489083e-01 8.09006542e-02
5.28916657e-01 -5.84789932e-01 -3.35241586e-01 5.99041104e-01
7.33012930e-02 -1.77633435e-01 8.11166584e-01 2.44312845e-02
-5.74358642e-01 -2.15529665e-01 -9.43209827e-01 9.00501907e-01
9.95830834e-01 -2.86627859e-01 -8.38660449e-02 4.50129986e-01
7.40888238e-01 -1.67832866e-01 -1.02977622e+00 3.52878749e-01
3.95227164e-01 -8.16751480e-01 8.68972778e-01 -7.19682813e-01
5.48444867e-01 -5.79315305e-01 -4.64079380e-02 -1.13615572e+00
-3.25418830e-01 -1.04615092e+00 -4.01731692e-02 7.03649759e-01
8.75845373e-01 -4.28083450e-01 9.86187100e-01 7.23013639e-01
-1.49094418e-01 -8.03154290e-01 -7.04091311e-01 -7.97352850e-01
-3.54664326e-02 -2.22630888e-01 7.61468887e-01 9.45407450e-01
4.86209065e-01 7.00148463e-01 -3.05671781e-01 -5.42141609e-02
3.47217470e-01 3.48883152e-01 5.31177759e-01 -1.30544150e+00
-1.78795263e-01 -5.62049329e-01 -2.07420722e-01 -9.68967557e-01
3.35905254e-01 -1.42659366e+00 5.95489480e-02 -1.63685000e+00
4.13621450e-03 -3.28991562e-01 -2.52921730e-01 5.67108989e-01
1.24931559e-01 -2.69965470e-01 9.41991061e-02 -3.25722367e-01
-5.40672600e-01 4.62764561e-01 1.21331882e+00 -4.12803173e-01
-3.19640249e-01 -2.46626455e-02 -7.12877095e-01 5.56042433e-01
9.05464709e-01 -5.19074440e-01 -4.47444052e-01 2.94822335e-01
8.01308215e-01 4.74439532e-01 -2.11572379e-01 -7.70994484e-01
4.18148220e-01 -6.77107155e-01 6.51494414e-02 -4.97864664e-01
1.47085562e-01 -8.76476109e-01 4.10941213e-01 8.98648262e-01
-4.53772724e-01 -1.78115144e-01 4.34344150e-02 7.64035463e-01
2.31381208e-02 -1.95287332e-01 6.13302767e-01 -3.02347183e-01
-4.46626782e-01 8.09833884e-01 -7.90637016e-01 -4.83799249e-01
1.15507841e+00 1.02550872e-01 -2.72296015e-02 -1.13602154e-01
-7.83127725e-01 1.04456693e-01 3.15498769e-01 4.79316339e-02
5.33176363e-01 -9.93799388e-01 -2.06259012e-01 -1.94365963e-01
3.17717582e-01 1.03064470e-01 -3.02493125e-01 4.45783168e-01
-6.94986105e-01 8.40402544e-01 2.40741909e-01 -1.61862358e-01
-1.13322163e+00 9.59848523e-01 2.77771235e-01 -6.48065805e-01
1.02306478e-01 6.51044905e-01 2.78153658e-01 -3.30979854e-01
1.28412724e-01 -6.94934547e-01 -2.81062182e-02 -1.44832358e-01
9.50672925e-02 3.73357683e-01 2.07300887e-01 -1.19331613e-01
-4.19321924e-01 3.14586937e-01 -2.53131360e-01 1.86314330e-01
1.15925789e+00 6.26305699e-01 -2.06200659e-01 2.43464839e-02
8.53890300e-01 -8.39525461e-02 -8.63070726e-01 -1.23565860e-01
1.63792849e-01 4.72186096e-02 5.27613498e-02 -9.72467601e-01
-7.95641243e-01 8.39432061e-01 -9.04389173e-02 2.11975560e-01
8.37760985e-01 -1.35058925e-01 3.08638126e-01 9.42699909e-01
3.51439685e-01 -1.06516862e+00 1.24823833e-02 3.86969268e-01
8.65779281e-01 -1.00045216e+00 3.65245193e-01 -9.55078363e-01
-2.11174086e-01 1.46757090e+00 3.65041733e-01 -9.59340259e-02
4.83900636e-01 -1.94613617e-02 -6.31926894e-01 -2.42299169e-01
-9.19723034e-01 -1.59891471e-01 1.96089640e-01 1.62073344e-01
7.71728516e-01 2.55236953e-01 -7.10717857e-01 4.82089698e-01
-4.47734706e-02 5.83885331e-03 2.78380454e-01 9.69771624e-01
-3.98668498e-01 -1.60010779e+00 6.59664348e-02 3.16756219e-01
-3.38814408e-01 -3.15561175e-01 -9.24076259e-01 5.70590198e-01
1.79166049e-01 1.11188996e+00 -4.38912839e-01 -3.34021598e-01
2.55001515e-01 1.10063516e-01 8.41895103e-01 -9.06880260e-01
-5.61257899e-01 4.45242189e-02 5.04600883e-01 -5.27851224e-01
-4.63501990e-01 -4.13826287e-01 -1.54569232e+00 -1.56969383e-01
-5.77496529e-01 5.23949444e-01 5.97278059e-01 8.37750912e-01
2.28770316e-01 5.08119166e-01 8.30633998e-01 -7.15276062e-01
-2.36979842e-01 -4.55181748e-01 -3.61261159e-01 -1.48295790e-01
-1.89142570e-01 -5.91453969e-01 -1.30776018e-01 4.35446203e-02] | [4.515755653381348, 6.095261573791504] |
1a48dd04-8686-4491-b3c3-98c03004bfaf | dynamic-perceiver-for-efficient-visual | 2306.11248 | null | https://arxiv.org/abs/2306.11248v1 | https://arxiv.org/pdf/2306.11248v1.pdf | Dynamic Perceiver for Efficient Visual Recognition | Early exiting has become a promising approach to improving the inference efficiency of deep networks. By structuring models with multiple classifiers (exits), predictions for ``easy'' samples can be generated at earlier exits, negating the need for executing deeper layers. Current multi-exit networks typically implement linear classifiers at intermediate layers, compelling low-level features to encapsulate high-level semantics. This sub-optimal design invariably undermines the performance of later exits. In this paper, we propose Dynamic Perceiver (Dyn-Perceiver) to decouple the feature extraction procedure and the early classification task with a novel dual-branch architecture. A feature branch serves to extract image features, while a classification branch processes a latent code assigned for classification tasks. Bi-directional cross-attention layers are established to progressively fuse the information of both branches. Early exits are placed exclusively within the classification branch, thus eliminating the need for linear separability in low-level features. Dyn-Perceiver constitutes a versatile and adaptable framework that can be built upon various architectures. Experiments on image classification, action recognition, and object detection demonstrate that our method significantly improves the inference efficiency of different backbones, outperforming numerous competitive approaches across a broad range of computational budgets. Evaluation on both CPU and GPU platforms substantiate the superior practical efficiency of Dyn-Perceiver. Code is available at https://www.github.com/LeapLabTHU/Dynamic_Perceiver. | ['Gao Huang', 'Shiji Song', 'Junlan Feng', 'Chao Deng', 'Yifan Pu', 'Xuran Pan', 'Yulin Wang', 'Zeyu Liu', 'Dongchen Han', 'Yizeng Han'] | 2023-06-20 | null | null | null | null | ['action-recognition-in-videos', 'classification-1'] | ['computer-vision', 'methodology'] | [ 2.59331495e-01 -8.46435204e-02 -3.38013411e-01 -6.70523286e-01
-5.23614347e-01 -6.22012913e-01 4.49285477e-01 8.47467035e-02
-5.12032330e-01 3.48017544e-01 -8.17093477e-02 -3.90878618e-01
1.02301121e-01 -7.66531825e-01 -6.79081023e-01 -6.28062129e-01
2.49429587e-02 -2.67340280e-02 3.42214584e-01 9.58608687e-02
1.45669654e-01 4.48848903e-01 -1.84142780e+00 1.04382718e+00
5.01633465e-01 1.40723121e+00 3.77937883e-01 8.31990361e-01
-1.07050706e-02 8.94943297e-01 -3.79165828e-01 -4.36635226e-01
2.93417931e-01 -4.92236130e-02 -5.11111081e-01 -1.88528061e-01
4.24191445e-01 -4.59427834e-01 -2.13808909e-01 7.66152859e-01
4.50827062e-01 -1.47386324e-02 3.58059168e-01 -1.38189292e+00
-4.84892696e-01 5.94050944e-01 -5.34526408e-01 4.79342937e-01
1.49659395e-01 4.41903561e-01 1.35747385e+00 -1.03262782e+00
9.35524032e-02 1.21363580e+00 7.55515039e-01 5.47926068e-01
-1.35174477e+00 -4.93960261e-01 5.34121692e-01 2.60962754e-01
-1.27423322e+00 -7.24030018e-01 5.71804941e-01 -4.64528382e-01
1.33337772e+00 4.78003681e-01 6.25639021e-01 1.11902928e+00
3.82007241e-01 1.19042516e+00 1.05944991e+00 -7.73030370e-02
3.97530943e-01 6.63427785e-02 4.77434933e-01 1.02342272e+00
6.89409748e-02 2.19122097e-01 -8.08994412e-01 -3.68124917e-02
6.22477710e-01 1.24789253e-01 -9.51119959e-02 -2.11412400e-01
-8.48003566e-01 6.87898278e-01 6.54030263e-01 -3.14754620e-02
-4.76543337e-01 3.84649545e-01 5.30961096e-01 2.72250950e-01
3.24695230e-01 1.69810504e-01 -5.63558459e-01 -2.45829821e-01
-7.63905287e-01 -1.72408782e-02 6.76583111e-01 8.67893875e-01
7.89456964e-01 -2.30129343e-02 -2.24210158e-01 7.57097542e-01
1.92467928e-01 8.25211629e-02 3.74129832e-01 -1.01074195e+00
3.81450117e-01 7.35317409e-01 -2.52435565e-01 -9.16520834e-01
-4.64432418e-01 -8.42259526e-01 -9.00669217e-01 2.72533715e-01
3.04748118e-01 -1.07790016e-01 -8.65701854e-01 1.61697745e+00
3.34430218e-01 -2.21597608e-02 -1.36467963e-01 7.82961786e-01
6.69380903e-01 5.04104376e-01 1.78476021e-01 2.46868849e-01
1.55888796e+00 -1.24515963e+00 -2.98899617e-02 -5.94010234e-01
5.66297233e-01 -3.37060124e-01 1.04310191e+00 5.85353494e-01
-9.53530371e-01 -9.52143192e-01 -1.07559037e+00 -3.77992868e-01
-3.17869186e-01 5.34816146e-01 1.10649610e+00 5.94917953e-01
-1.23308921e+00 5.40808737e-01 -1.22846031e+00 1.47349417e-01
7.49700189e-01 6.19876504e-01 -2.39894450e-01 -2.64734402e-02
-7.24694312e-01 5.19506812e-01 4.42619741e-01 3.63804460e-01
-8.67646992e-01 -6.62437737e-01 -6.91661656e-01 3.07787091e-01
2.88604677e-01 -8.35118413e-01 1.26536238e+00 -1.23315120e+00
-1.33388305e+00 6.79452062e-01 -3.67440671e-01 -5.91984451e-01
4.05279845e-01 -3.36819112e-01 -4.86794785e-02 3.71076353e-02
-1.63983822e-01 9.27327394e-01 9.47562099e-01 -9.62219417e-01
-9.09260213e-01 -3.64654452e-01 4.69795138e-01 2.96557009e-01
-3.87931287e-01 -2.34479636e-01 -4.99220729e-01 -3.40193391e-01
9.34050903e-02 -7.60090292e-01 -1.58383951e-01 2.94373423e-01
-2.45834202e-01 -4.00296152e-01 5.62012672e-01 -3.28843087e-01
1.13019967e+00 -2.34291697e+00 1.06052220e-01 -2.79986970e-02
6.11287236e-01 1.65642321e-01 -3.03024799e-01 7.25064054e-02
-1.04150265e-01 -1.60177320e-01 6.64687082e-02 -4.32084948e-01
-2.71061957e-02 2.28286132e-01 -2.86636174e-01 2.64836550e-01
5.01830935e-01 8.90987575e-01 -8.01892161e-01 -2.35302851e-01
1.39822945e-01 3.90424103e-01 -8.45442355e-01 9.45315510e-02
-2.25298256e-01 1.78010520e-02 -4.97602850e-01 8.37849259e-01
2.93160111e-01 -5.88058889e-01 1.87534854e-01 -3.54669869e-01
-1.17664970e-01 4.83024925e-01 -8.57529521e-01 1.60206437e+00
-6.42196298e-01 5.48806131e-01 2.32052088e-01 -1.13379478e+00
7.82158613e-01 -6.63529849e-03 2.14261502e-01 -6.22315109e-01
2.62362510e-01 1.14037268e-01 2.65210927e-01 -3.50030780e-01
3.59091312e-01 2.03629076e-01 -3.21218334e-02 3.06786090e-01
3.87729444e-02 5.64343929e-01 1.32360518e-01 -3.34413908e-02
1.29750991e+00 3.30945820e-01 2.91449994e-01 -1.52541608e-01
3.76920491e-01 -1.58728540e-01 7.13809669e-01 1.03099167e+00
-3.04209143e-01 1.89004973e-01 4.19999331e-01 -7.56336570e-01
-6.64085805e-01 -1.09644282e+00 -9.03675929e-02 1.61877954e+00
-7.51314983e-02 -5.44000685e-01 -6.36188924e-01 -7.62201250e-01
1.90309435e-01 5.83015084e-01 -5.71523190e-01 -2.07241222e-01
-4.23606128e-01 -8.09265018e-01 4.30056304e-01 9.49598968e-01
5.43462098e-01 -8.78066182e-01 -1.13414681e+00 2.53930151e-01
-8.72010142e-02 -9.89807427e-01 7.39528332e-03 6.25497341e-01
-7.92237580e-01 -8.46581221e-01 -1.03495993e-01 -3.49133760e-01
6.58044755e-01 2.54337609e-01 1.16864121e+00 1.77372813e-01
-4.61680859e-01 2.55075485e-01 -2.03541547e-01 -3.45208257e-01
-2.01699495e-01 4.52535629e-01 -1.37540586e-02 8.24857429e-02
6.27238750e-01 -7.11236179e-01 -6.94127560e-01 1.20273016e-01
-6.78907752e-01 4.57230896e-01 8.00812304e-01 8.41621459e-01
3.84915471e-01 -7.59951696e-02 5.30436635e-01 -5.81389546e-01
2.13800400e-01 -4.84777093e-01 -6.10541523e-01 1.40906453e-01
-4.19734120e-01 1.17829204e-01 8.53529394e-01 -3.86993766e-01
-8.15437913e-01 2.15094626e-01 -2.84041286e-01 -3.30438763e-01
-2.95337200e-01 4.33491945e-01 -8.88041109e-02 1.92440763e-01
6.37312353e-01 3.42465520e-01 -1.09927699e-01 -4.76905435e-01
2.33733416e-01 7.03534007e-01 4.13311362e-01 -7.05069065e-01
4.79169369e-01 4.28944886e-01 -1.94317877e-01 -5.87411284e-01
-1.18449533e+00 -2.59503931e-01 -5.86462915e-01 -3.17513257e-01
7.17610538e-01 -1.09030247e+00 -1.08802187e+00 6.01868749e-01
-1.06348979e+00 -5.40951788e-01 -2.70443976e-01 2.87553638e-01
-3.65683526e-01 -1.17347263e-01 -6.95278525e-01 -7.35250890e-01
-2.95132637e-01 -1.33128834e+00 1.19122517e+00 2.38073960e-01
-3.16320717e-01 -8.86071622e-01 -4.73046243e-01 5.27267933e-01
3.02909762e-01 -7.96516910e-02 8.95833492e-01 -6.49127245e-01
-8.81883085e-01 -2.34193658e-03 -3.98375034e-01 5.26692092e-01
1.73429679e-02 -1.14584453e-01 -1.33847356e+00 -4.59866554e-01
-8.68777558e-02 -7.07449079e-01 1.18536115e+00 1.13365158e-01
1.61183774e+00 -2.31026888e-01 -3.24031740e-01 7.61796534e-01
1.31277502e+00 1.48233250e-01 4.05178398e-01 1.96879134e-01
6.94340467e-01 3.69918376e-01 3.15072834e-01 5.06190300e-01
5.16700804e-01 4.22027498e-01 3.29725236e-01 -8.91037881e-02
-1.47189245e-01 4.58454154e-03 6.68366671e-01 7.88136542e-01
-8.09449051e-03 -2.10725605e-01 -9.73780453e-01 2.53535897e-01
-1.77746904e+00 -7.61271358e-01 1.11823782e-01 1.86113179e+00
9.60026443e-01 4.10679758e-01 -1.14675291e-01 1.15339145e-01
3.83140653e-01 2.09634662e-01 -9.36572075e-01 -4.84181553e-01
2.47780249e-01 2.24751696e-01 2.08247572e-01 4.44210529e-01
-1.05876696e+00 8.90120924e-01 5.79798460e+00 9.35204387e-01
-1.15282023e+00 6.39610812e-02 9.27957237e-01 -2.46849790e-01
-1.17421150e-01 6.23963773e-02 -1.25566924e+00 3.75974596e-01
9.58222985e-01 2.92884231e-01 3.84937972e-01 1.14354873e+00
-5.08592725e-02 -2.33499929e-01 -1.57091689e+00 9.02550161e-01
-1.12146914e-01 -1.46103644e+00 2.73140538e-02 -4.92053106e-02
2.60392517e-01 4.17008728e-01 6.54180124e-02 5.04546940e-01
2.51190275e-01 -7.56792367e-01 9.22174275e-01 5.26744127e-01
7.27313280e-01 -6.25845373e-01 4.10658538e-01 4.36038256e-01
-1.28782797e+00 -4.49922770e-01 -4.24737185e-01 -2.96221644e-01
-2.85359174e-01 5.47849953e-01 -6.91392779e-01 3.90459538e-01
7.21012533e-01 6.98111713e-01 -7.40829527e-01 6.14278078e-01
-1.21491745e-01 5.02854705e-01 -3.82960677e-01 -6.17309511e-02
3.08287621e-01 1.14638925e-01 4.08890307e-01 1.25145411e+00
-4.26522493e-02 -3.62504981e-02 3.33247334e-01 8.54619503e-01
-8.37244913e-02 -2.60979295e-01 -3.07826221e-01 1.38607845e-02
4.71507698e-01 1.39993060e+00 -8.78240943e-01 -5.52342415e-01
-5.09246230e-01 9.39763367e-01 6.43364072e-01 2.06802800e-01
-9.66497958e-01 -1.76216960e-01 6.62231982e-01 -1.00345604e-01
4.36311245e-01 -1.86916023e-01 -5.20162582e-01 -1.22200155e+00
1.27130017e-01 -9.93593335e-01 3.69065613e-01 -5.43214083e-01
-1.08498991e+00 6.31984055e-01 -2.22386032e-01 -8.58476281e-01
-7.50526115e-02 -9.66441035e-01 -2.87439853e-01 7.25972533e-01
-1.39454651e+00 -9.95288551e-01 -4.02657568e-01 4.20932055e-01
7.13322520e-01 -1.63596392e-01 8.34974766e-01 1.63753420e-01
-8.16324472e-01 7.79522479e-01 -5.68069257e-02 1.22758605e-01
1.73998609e-01 -1.13818574e+00 4.27062899e-01 8.08745563e-01
2.00128317e-01 7.59443104e-01 2.17069834e-01 -2.89194256e-01
-1.46078742e+00 -1.02546418e+00 5.59224963e-01 -4.40459281e-01
6.03513479e-01 -7.17334330e-01 -5.95861554e-01 6.74133837e-01
-1.36401206e-01 9.92650017e-02 9.16737497e-01 4.39325035e-01
-7.56835401e-01 -4.57778424e-01 -7.38810718e-01 6.63572013e-01
1.06987333e+00 -6.12501264e-01 -3.55945319e-01 7.76359364e-02
4.29602832e-01 -2.95500368e-01 -7.20989764e-01 3.84850234e-01
8.66370201e-01 -1.28249526e+00 8.77459884e-01 -5.23474514e-01
5.91391385e-01 -1.11166485e-01 -1.52558520e-01 -7.47400820e-01
-3.96019071e-01 -4.26183254e-01 -4.55199897e-01 9.60767984e-01
3.53296489e-01 -7.14842141e-01 8.65097225e-01 5.60678422e-01
-2.20876798e-01 -1.31770635e+00 -7.97191560e-01 -6.86486959e-01
-2.84327745e-01 -5.89518249e-01 3.72507364e-01 6.20635867e-01
-3.45561981e-01 6.67112231e-01 3.19190621e-02 1.50451258e-01
5.45568883e-01 3.56400192e-01 6.36067092e-01 -1.15665388e+00
-8.36533248e-01 -5.76209605e-01 -4.24280673e-01 -1.52836311e+00
7.31287226e-02 -1.18468118e+00 5.67544773e-02 -1.22540998e+00
3.70138288e-01 -6.43895745e-01 -4.77893323e-01 1.15793133e+00
-1.77566752e-01 2.49430761e-01 1.45364448e-01 9.56338644e-02
-5.66554487e-01 5.20136118e-01 9.83141303e-01 5.35564916e-03
-1.15473539e-01 5.60585335e-02 -7.86955237e-01 1.03181350e+00
9.11361754e-01 -3.92228931e-01 -6.04291141e-01 -7.49603570e-01
2.65354037e-01 -2.29795009e-01 6.40800595e-01 -1.27001238e+00
2.79498518e-01 1.28807098e-01 6.66372597e-01 -4.96610761e-01
5.52153349e-01 -5.99549174e-01 -2.73390353e-01 4.27045614e-01
-5.27024448e-01 1.59570634e-01 2.76085317e-01 5.03199160e-01
8.26068372e-02 -1.06326021e-01 8.43648612e-01 -1.88401178e-01
-7.51254201e-01 3.24214518e-01 -4.24246490e-01 -8.00658166e-02
7.95566976e-01 -3.71991605e-01 -3.29868644e-01 1.01890199e-01
-7.21033752e-01 1.16846785e-01 1.18546799e-01 5.37285924e-01
6.33980274e-01 -9.66100454e-01 -3.91497195e-01 3.60881448e-01
-4.91390266e-02 4.05449197e-02 2.02043578e-01 8.52376640e-01
-2.48562574e-01 3.16948026e-01 -1.90020591e-01 -8.20589662e-01
-1.14327765e+00 2.63916731e-01 3.43368351e-01 -2.54205644e-01
-5.78735054e-01 1.26348805e+00 4.93190259e-01 -2.88617700e-01
3.03993315e-01 -4.66067553e-01 1.95238411e-01 1.45836562e-01
5.91248095e-01 1.94756925e-01 3.51591632e-02 -3.03675234e-01
-3.57493997e-01 -4.79227155e-02 -4.58900779e-01 1.85768709e-01
1.45560253e+00 9.29647870e-03 -2.36986205e-01 5.76313794e-01
1.15323973e+00 -3.91944975e-01 -1.62889922e+00 -1.15864523e-01
5.97525388e-02 -2.29674637e-01 3.22733521e-01 -8.70970130e-01
-1.19211423e+00 9.83057261e-01 4.69316155e-01 5.77887110e-02
1.31068826e+00 -1.95739746e-01 6.36831284e-01 6.94143474e-01
4.44893509e-01 -9.30079877e-01 1.68623239e-01 4.96782124e-01
8.05069327e-01 -9.77748752e-01 -1.28065288e-01 -3.18322122e-01
-2.58053035e-01 1.11987495e+00 9.69057202e-01 -9.54586267e-02
4.85034108e-01 6.28042340e-01 -1.67711809e-01 -1.18806884e-01
-1.30573189e+00 1.42569086e-02 1.62725791e-01 2.21769065e-01
3.68895143e-01 -5.73476627e-02 8.26311111e-02 7.41212249e-01
-1.39685139e-01 -2.60016117e-02 1.27107009e-01 1.07446754e+00
-4.28872347e-01 -1.04344749e+00 -6.53478578e-02 5.66558301e-01
-3.83594841e-01 -4.13547844e-01 -1.56890079e-01 5.53056598e-01
4.62725371e-01 7.62843668e-01 3.87170583e-01 -5.86007118e-01
-5.06294966e-02 9.20134336e-02 4.35905695e-01 -6.91607177e-01
-7.82119691e-01 -2.23771520e-02 2.27137625e-01 -9.49716330e-01
-2.61075795e-01 -5.72986782e-01 -1.05278122e+00 -2.14626610e-01
-2.23689064e-01 -2.00958177e-01 5.29134572e-01 8.44159961e-01
6.78724229e-01 7.41146266e-01 4.96068805e-01 -7.40641117e-01
-8.34700882e-01 -5.93565047e-01 -1.49739131e-01 -2.42727265e-01
4.13080275e-01 -6.00884616e-01 -1.45104855e-01 9.95143428e-02] | [9.4284029006958, 1.443318247795105] |
b7ae2839-dcaa-48ab-8685-5a0ba41091da | meta-ordinal-regression-forest-for-learning | 2012.03480 | null | https://arxiv.org/abs/2012.03480v1 | https://arxiv.org/pdf/2012.03480v1.pdf | Meta Ordinal Regression Forest For Learning with Unsure Lung Nodules | Deep learning-based methods have achieved promising performance in early detection and classification of lung nodules, most of which discard unsure nodules and simply deal with a binary classification -- malignant vs benign. Recently, an unsure data model (UDM) was proposed to incorporate those unsure nodules by formulating this problem as an ordinal regression, showing better performance over traditional binary classification. To further explore the ordinal relationship for lung nodule classification, this paper proposes a meta ordinal regression forest (MORF), which improves upon the state-of-the-art ordinal regression method, deep ordinal regression forest (DORF), in three major ways. First, MORF can alleviate the biases of the predictions by making full use of deep features while DORF needs to fix the composition of decision trees before training. Second, MORF has a novel grouped feature selection (GFS) module to re-sample the split nodes of decision trees. Last, combined with GFS, MORF is equipped with a meta learning-based weighting scheme to map the features selected by GFS to tree-wise weights while DORF assigns equal weights for all trees. Experimental results on the LIDC-IDRI dataset demonstrate superior performance over existing methods, including the state-of-the-art DORF. | ['Hongming Shan', 'Junping Zhang', 'Haiping Zhu', 'Yiming Lei'] | 2020-12-07 | null | null | null | null | ['lung-nodule-classification'] | ['medical'] | [ 1.88181221e-01 1.68958455e-01 -7.46927440e-01 -5.19298136e-01
-6.26627445e-01 2.57435024e-01 5.78991771e-01 -9.96511951e-02
-1.28093407e-01 5.59421718e-01 2.14340776e-01 -4.70406085e-01
-4.67110962e-01 -1.04078531e+00 -2.76081860e-01 -8.99894476e-01
-6.35991246e-02 8.03504765e-01 3.50400239e-01 -1.76910292e-02
-1.33451596e-01 2.11654961e-01 -1.55099666e+00 8.60236406e-01
1.09491384e+00 1.19232702e+00 1.42330974e-01 3.38493794e-01
-1.37850672e-01 9.68966126e-01 -1.46023721e-01 -3.72916907e-01
4.02652584e-02 -1.96610674e-01 -7.87046313e-01 -3.59729454e-02
2.82461822e-01 -5.85767865e-01 -3.14781100e-01 6.88408315e-01
1.71496883e-01 -4.82639223e-01 1.20486891e+00 -1.21637714e+00
-4.23759907e-01 8.09096098e-01 -5.89571655e-01 -1.12187475e-01
-4.10400361e-01 -1.57294460e-02 1.09311163e+00 -8.94671321e-01
9.22714844e-02 9.93296385e-01 1.01772058e+00 4.94785428e-01
-9.14854527e-01 -9.33357239e-01 -6.50392771e-02 3.62195581e-01
-1.14747882e+00 -4.25897390e-02 3.61095428e-01 -7.00330794e-01
4.26082104e-01 5.52830577e-01 6.57221735e-01 1.05804944e+00
3.66896570e-01 6.31907165e-01 1.27921903e+00 -2.12033644e-01
-9.49265584e-02 6.42707795e-02 1.93412602e-01 1.34264338e+00
4.65727985e-01 2.18071625e-01 -1.61683604e-01 -2.78875202e-01
8.43593180e-01 5.59234619e-01 2.79505085e-02 -4.36961651e-01
-1.51417542e+00 1.03456056e+00 1.04253674e+00 2.35753357e-01
-4.48744208e-01 -8.25139694e-03 3.62698644e-01 1.11766510e-01
4.01742399e-01 3.39510053e-01 -3.15999597e-01 6.56672120e-01
-1.00743139e+00 -5.66169992e-02 5.56296647e-01 4.72843766e-01
6.12630785e-01 -3.16384286e-01 -8.61178875e-01 1.09047818e+00
2.97620088e-01 5.29440753e-02 1.03597450e+00 -6.73915207e-01
1.98007479e-01 8.94651890e-01 -2.18258888e-01 -6.68109477e-01
-6.39490426e-01 -9.55421567e-01 -1.62065363e+00 2.61464179e-01
8.49211439e-02 2.70084709e-01 -1.24707973e+00 1.18938339e+00
2.70711869e-01 1.88837618e-01 -1.56686932e-01 5.42520463e-01
1.27563024e+00 1.78639531e-01 8.93187672e-02 -2.24768713e-01
1.49789143e+00 -1.23496473e+00 -3.54422480e-01 1.18704371e-01
6.66546345e-01 -2.59217709e-01 8.52790654e-01 3.71119112e-01
-3.78895402e-01 -4.93834555e-01 -9.00571048e-01 1.76483378e-01
-1.95828557e-01 4.17408228e-01 8.46985817e-01 5.03771424e-01
-8.29188168e-01 5.87418735e-01 -8.65765452e-01 -2.68876329e-02
6.75099492e-01 6.82530880e-01 -3.68075490e-01 -1.45552903e-01
-1.19681132e+00 7.43052006e-01 2.83510178e-01 1.89971507e-01
-8.93320441e-01 -5.35525858e-01 -6.33360386e-01 5.18284366e-02
5.11354148e-01 -1.20935476e+00 1.45280457e+00 -7.61835754e-01
-1.18540871e+00 7.17806935e-01 -3.24376315e-01 -6.27091229e-01
6.84456944e-01 1.72034502e-01 -2.25003421e-01 -1.91807568e-01
2.50562876e-01 7.05811024e-01 1.08589399e+00 -8.98800969e-01
-8.57651055e-01 -2.41413280e-01 -3.41953784e-01 6.23179115e-02
-3.12071115e-01 -3.48875672e-01 -8.49063993e-02 -6.43194973e-01
4.81780946e-01 -9.56559956e-01 -5.15780091e-01 6.87135980e-02
-4.61121261e-01 -6.82042778e-01 7.71061063e-01 -3.24802011e-01
1.50980544e+00 -1.86979926e+00 -5.37876375e-02 1.41860574e-01
9.31352079e-01 1.60842001e-01 1.06084347e-01 -1.94646835e-01
-1.31670058e-01 3.08438957e-01 -3.38723898e-01 -1.88432410e-01
-2.68818706e-01 3.77304107e-01 1.73106343e-01 2.80881733e-01
4.10177946e-01 1.00751245e+00 -6.87897384e-01 -1.08651185e+00
3.05138260e-01 7.88123682e-02 -4.94514555e-01 9.37219113e-02
-5.69498204e-02 9.50145498e-02 -5.53205073e-01 9.89748001e-01
6.77296281e-01 -5.21589279e-01 6.57799020e-02 -2.01032192e-01
-2.13053823e-01 3.24765414e-01 -7.37363279e-01 9.40763712e-01
-5.17023683e-01 1.14535643e-02 -2.70628452e-01 -9.24554050e-01
9.16949928e-01 2.04646423e-01 7.19567478e-01 -2.96564996e-01
-1.52526936e-02 5.05104303e-01 3.36090922e-01 -4.27243441e-01
1.82345398e-02 -2.04408467e-01 -2.55227145e-02 9.59473848e-02
-1.35866091e-01 1.04434378e-01 -4.54180725e-02 -1.89639062e-01
1.35422897e+00 -1.29623324e-01 7.50108540e-01 -1.75132498e-01
6.34018362e-01 9.12024155e-02 6.85478091e-01 8.92226875e-01
-1.93896353e-01 6.26635671e-01 4.46045101e-01 -6.80785000e-01
-5.74564159e-01 -1.01713979e+00 -5.47583342e-01 1.11506557e+00
-9.10564587e-02 -1.96743220e-01 -3.08276147e-01 -1.39757764e+00
2.73613065e-01 4.01230097e-01 -1.21102679e+00 5.39299892e-03
-5.71266890e-01 -1.16936326e+00 2.13188052e-01 5.76998174e-01
6.28496945e-01 -1.07404268e+00 -3.21985573e-01 1.67509198e-01
-4.51108158e-01 -4.24557716e-01 -5.41221984e-02 9.10090387e-01
-1.09107542e+00 -1.15044761e+00 -6.40798032e-01 -8.85943234e-01
6.64199412e-01 5.72352171e-01 1.11172903e+00 2.92528242e-01
-3.10672492e-01 -4.90102559e-01 -5.39885163e-01 -3.09038967e-01
-6.35450900e-01 5.36483109e-01 -2.75378793e-01 4.13835868e-02
2.23307550e-01 -3.12172681e-01 -6.43833816e-01 7.34426975e-01
-6.34829342e-01 5.25669515e-01 1.45416212e+00 1.31933415e+00
8.20663691e-01 2.75865406e-01 5.72561264e-01 -1.34451640e+00
-8.84262770e-02 -8.13884258e-01 1.34795606e-01 1.30213022e-01
-6.78525329e-01 2.55032003e-01 6.15312815e-01 -4.72303897e-01
-7.67535031e-01 5.04178181e-02 -3.14593583e-01 -3.13077360e-01
1.03961974e-01 5.29506683e-01 2.05613434e-01 3.42686474e-02
7.94905424e-01 1.79525726e-02 1.80330858e-01 -3.13539356e-01
-1.32174879e-01 1.08257234e+00 3.45322311e-01 6.51820004e-03
8.94452751e-01 4.00257111e-01 2.26821214e-01 -3.42538625e-01
-1.21737087e+00 -7.06908762e-01 -7.92114735e-01 -1.33005738e-01
6.90861881e-01 -8.20331335e-01 -2.62356371e-01 4.49264109e-01
-5.51326573e-01 -2.14702964e-01 -2.59433985e-01 5.52007377e-01
-4.39792931e-01 1.16262436e-01 -4.17768568e-01 -5.50262570e-01
-4.64325398e-01 -1.17189169e+00 1.08288991e+00 -2.47176886e-02
-4.18434054e-01 -4.66482639e-01 -3.25449556e-01 3.83404315e-01
5.75851858e-01 2.68936872e-01 1.32740092e+00 -6.73922777e-01
-4.36046869e-01 -2.77404636e-01 -4.85240638e-01 1.75349817e-01
3.77107203e-01 5.37699834e-02 -7.96981633e-01 -1.18932568e-01
-7.65671283e-02 -2.75249749e-01 1.60977209e+00 6.19035721e-01
1.65032816e+00 -4.46706295e-01 -6.94064558e-01 8.41534853e-01
1.22232580e+00 -1.97953880e-01 4.18050379e-01 7.02982068e-01
7.31169224e-01 2.37195611e-01 7.39050090e-01 3.72806579e-01
4.34947133e-01 4.20856178e-01 1.01233411e+00 -3.58156890e-01
-3.66897970e-01 -1.21158414e-01 7.71708488e-02 6.70265198e-01
-1.45269498e-01 1.85149506e-01 -6.67829156e-01 1.92889243e-01
-1.84503365e+00 -8.24386418e-01 -4.60260391e-01 2.19661903e+00
7.87344158e-01 3.87386411e-01 5.96755631e-02 3.62112224e-01
6.08212292e-01 1.45280823e-01 -5.24602294e-01 -1.89989462e-01
1.92911044e-01 2.05960184e-01 5.42841792e-01 -1.04893543e-01
-1.68443763e+00 3.43399376e-01 5.48325682e+00 1.01358044e+00
-1.23675644e+00 1.10681646e-01 8.57133090e-01 2.72477388e-01
-1.27888650e-01 -2.61076421e-01 -7.85140216e-01 3.05736363e-01
6.07690036e-01 3.03755224e-01 -1.34650338e-02 1.05450010e+00
-8.80216062e-02 3.57432216e-02 -1.07249069e+00 5.91233671e-01
-2.49891683e-01 -1.20767760e+00 1.41507536e-01 2.21657306e-01
5.63023269e-01 2.79291630e-01 1.71257451e-01 8.16604495e-01
2.98905879e-01 -1.28294003e+00 3.87382120e-01 5.04881024e-01
1.08332145e+00 -3.87548625e-01 1.19175196e+00 5.58085680e-01
-1.17877936e+00 -5.55135965e-01 -4.45067704e-01 1.70774505e-01
-7.06431031e-01 8.47909451e-01 -1.43292749e+00 7.62017250e-01
8.82207632e-01 8.18197131e-01 -1.11015964e+00 1.14265954e+00
1.29998878e-01 8.09254766e-01 -1.72815487e-01 -8.93237367e-02
1.78438082e-01 1.73390418e-01 6.44146577e-02 1.13030064e+00
4.93807524e-01 -3.51709515e-01 1.97152302e-01 4.49686289e-01
3.75235379e-02 1.22047111e-01 -3.49912941e-01 3.43744695e-01
2.74435014e-01 1.48541272e+00 -4.90162790e-01 -1.89910412e-01
-3.38285208e-01 5.35678327e-01 3.01457763e-01 -3.59077543e-01
-9.21500802e-01 2.96449140e-02 6.04163706e-02 6.11362875e-01
2.93095171e-01 2.68326700e-01 -5.16150832e-01 -9.16259825e-01
-2.49983087e-01 -7.78480947e-01 7.05606997e-01 -4.92774099e-01
-1.38776791e+00 6.70446157e-01 -2.22739145e-01 -1.77815318e+00
-2.49572024e-01 -5.72342157e-01 -5.25526285e-01 5.87978184e-01
-1.74290764e+00 -1.51276004e+00 -7.47320056e-01 3.16072404e-01
5.47851861e-01 -2.88643807e-01 8.71600866e-01 -1.91384014e-02
-5.45048416e-01 6.09353960e-01 1.46825492e-01 2.11814493e-01
9.18262601e-01 -1.43326795e+00 5.19000180e-02 9.73187760e-02
-2.04840928e-01 2.30066210e-01 1.23086847e-01 -5.56864917e-01
-8.51670802e-01 -1.73084545e+00 8.16200972e-01 -2.90203035e-01
4.84327108e-01 7.54661933e-02 -9.64859188e-01 3.97060871e-01
-3.98697287e-01 2.87844986e-01 8.00711989e-01 3.06527346e-01
-3.09092879e-01 -3.74800384e-01 -1.09269047e+00 3.98555219e-01
8.32999825e-01 6.15388863e-02 -5.04681885e-01 3.87561589e-01
7.31921375e-01 -1.48184508e-01 -9.67781961e-01 1.28584981e+00
8.77305925e-01 -1.13122046e+00 9.23871994e-01 -3.38251293e-01
8.53817999e-01 -4.14563805e-01 -1.62155461e-02 -1.10191298e+00
-1.21684849e+00 1.31518260e-01 -2.15204731e-01 9.18101192e-01
5.73587418e-01 -5.72598815e-01 8.56977522e-01 -5.25199622e-02
-3.71568978e-01 -1.45269847e+00 -9.60895360e-01 -5.14017940e-01
1.40696880e-03 -9.94238406e-02 6.19039893e-01 7.56145716e-01
-6.75199866e-01 1.87448889e-01 -2.54977077e-01 -2.72192284e-02
4.85529542e-01 3.32473278e-01 6.00114405e-01 -1.79213262e+00
-4.23672915e-01 -6.22098744e-01 -2.30182424e-01 -4.23389733e-01
-1.70879424e-01 -1.24268162e+00 6.24409914e-02 -1.82156265e+00
6.13204718e-01 -8.17409694e-01 -6.96803391e-01 7.22025514e-01
-3.20077509e-01 2.03974783e-01 -1.40309930e-01 7.76889503e-01
-6.64869010e-01 3.74477297e-01 1.51042664e+00 -3.51162463e-01
-1.70932617e-02 8.07339251e-01 -9.10573781e-01 9.25490856e-01
6.48357511e-01 -7.61519969e-01 1.04048148e-01 -6.70524361e-03
-2.06376225e-01 1.47744372e-01 4.35572743e-01 -1.13154125e+00
2.42606085e-03 -3.61757278e-01 8.29990387e-01 -1.13608301e+00
1.66655369e-02 -8.70467365e-01 1.85120627e-01 1.07528532e+00
-2.47114152e-01 -4.41661716e-01 -4.30458188e-01 6.51571870e-01
-2.96575040e-01 -4.24147308e-01 1.03031933e+00 -2.28038445e-01
-5.09187102e-01 5.52076101e-01 -3.79926562e-01 -3.98979247e-01
1.00761080e+00 -3.44792664e-01 -2.64369190e-01 2.45532304e-01
-6.56791508e-01 3.03776979e-01 1.12256348e-01 1.68928489e-01
4.54123765e-01 -1.36061609e+00 -1.16052079e+00 3.85571212e-01
3.71511638e-01 4.96734321e-01 1.87612295e-01 1.11560225e+00
-3.30875188e-01 2.91775167e-01 -1.24375410e-02 -9.71505761e-01
-1.27965403e+00 3.29134643e-01 5.64367473e-01 -1.03736436e+00
-4.35577691e-01 7.72088349e-01 4.57453251e-01 -6.13939941e-01
1.95407674e-01 -5.87936461e-01 -4.73767579e-01 2.41370380e-01
1.03422724e-01 2.71387279e-01 3.54937404e-01 -1.09247901e-01
-3.43691319e-01 2.96312124e-01 -3.50770384e-01 4.88663256e-01
1.29818034e+00 4.10385966e-01 -3.24529111e-01 4.76164758e-01
1.00270176e+00 -2.36917734e-01 -7.73398757e-01 -3.88070822e-01
2.14431100e-02 -1.44591078e-01 1.11707062e-01 -9.86620486e-01
-7.98215508e-01 7.54109383e-01 6.68746173e-01 1.78658396e-01
1.22485864e+00 -3.23142111e-02 5.32015562e-01 4.18036163e-01
1.45864040e-01 -5.53472221e-01 4.92623672e-02 4.96994495e-01
8.31525147e-01 -1.60854435e+00 1.98813036e-01 -6.22633100e-01
-6.49841189e-01 1.35618067e+00 7.15437829e-01 -1.58208720e-02
8.99841487e-01 1.97402194e-01 -3.03561509e-01 2.09906608e-01
-1.05898905e+00 -3.18748116e-01 4.89626348e-01 4.34680939e-01
5.55716515e-01 5.61388671e-01 -2.85892099e-01 9.57360446e-01
-2.27586612e-01 3.78041893e-01 1.94496766e-01 5.74639320e-01
-8.00072849e-01 -1.07908332e+00 -3.02518755e-01 1.58794463e+00
-3.20018172e-01 -1.85839057e-01 -3.48843902e-01 8.83379161e-01
4.84596133e-01 5.59182525e-01 1.43166527e-01 -7.02353418e-01
1.90352887e-01 2.76357364e-02 8.47048536e-02 -7.95542300e-01
-9.04568911e-01 6.90544099e-02 6.66541755e-02 -1.22392565e-01
-2.14417353e-01 -6.09490752e-01 -1.03384256e+00 8.44488144e-02
-6.48744464e-01 -1.40533010e-02 3.88894498e-01 8.04764390e-01
-9.26478356e-02 9.25300241e-01 1.04790652e+00 -6.16618931e-01
-1.05254292e+00 -1.16188669e+00 -4.88534749e-01 9.26813260e-02
5.53851247e-01 -9.91324306e-01 -4.53271478e-01 -4.34636354e-01] | [15.340761184692383, -2.153601884841919] |
bc2533d2-9c4a-461d-90da-6c9a11f23de7 | joint-featurewise-weighting-and-lobal | 2007.12829 | null | https://arxiv.org/abs/2007.12829v1 | https://arxiv.org/pdf/2007.12829v1.pdf | Joint Featurewise Weighting and Lobal Structure Learning for Multi-view Subspace Clustering | Multi-view clustering integrates multiple feature sets, which reveal distinct aspects of the data and provide complementary information to each other, to improve the clustering performance. It remains challenging to effectively exploit complementary information across multiple views since the original data often contain noise and are highly redundant. Moreover, most existing multi-view clustering methods only aim to explore the consistency of all views while ignoring the local structure of each view. However, it is necessary to take the local structure of each view into consideration, because different views would present different geometric structures while admitting the same cluster structure. To address the above issues, we propose a novel multi-view subspace clustering method via simultaneously assigning weights for different features and capturing local information of data in view-specific self-representation feature spaces. Especially, a common cluster structure regularization is adopted to guarantee consistency among different views. An efficient algorithm based on an augmented Lagrangian multiplier is also developed to solve the associated optimization problem. Experiments conducted on several benchmark datasets demonstrate that the proposed method achieves state-of-the-art performance. We provide the Matlab code on https://github.com/Ekin102003/JFLMSC. | ['Shi-Xun Lina', 'Ting Shu', 'Guo Zhongb'] | 2020-07-25 | null | null | null | null | ['multi-view-subspace-clustering'] | ['computer-vision'] | [-3.49581003e-01 -6.08897150e-01 -1.10079125e-01 -1.89673543e-01
-6.75212085e-01 -7.53946424e-01 1.18604176e-01 -1.17074117e-01
9.76555347e-02 9.43790078e-02 3.65605146e-01 3.63493502e-01
-4.42500323e-01 -4.48814243e-01 -1.91436499e-01 -1.22003269e+00
3.41573983e-01 2.04030514e-01 -4.22425717e-02 7.02415481e-02
1.49305135e-01 1.43302098e-01 -1.35818648e+00 2.37761453e-01
9.74581242e-01 6.76991165e-01 4.39857364e-01 7.14735836e-02
1.82414219e-01 2.81306475e-01 -1.57602549e-01 5.02200611e-02
3.07497442e-01 -3.60208780e-01 -3.62642020e-01 7.08790600e-01
2.84446239e-01 4.16361317e-02 -2.35807866e-01 1.48618281e+00
3.68344367e-01 1.68811798e-01 3.75364572e-01 -1.23653650e+00
-6.76195443e-01 9.60342884e-02 -1.14389443e+00 2.67972089e-02
2.45156899e-01 -6.43848404e-02 1.00349176e+00 -1.00841081e+00
5.45548618e-01 1.00239027e+00 1.84800312e-01 9.34863612e-02
-1.31426013e+00 -5.39438784e-01 4.80051100e-01 2.92367369e-01
-1.66175783e+00 -3.90869081e-01 1.19195616e+00 -4.24754739e-01
7.14723542e-02 3.93105984e-01 6.08152747e-01 7.30154574e-01
-5.40212691e-02 7.91239500e-01 1.26042080e+00 8.78636762e-02
4.42116074e-02 2.15093300e-01 1.69182405e-01 5.94419420e-01
3.54548305e-01 -3.77247065e-01 -2.91972011e-01 -3.19620728e-01
4.64672863e-01 5.46362758e-01 -6.91081762e-01 -1.01956165e+00
-1.42364526e+00 8.17803979e-01 2.65781313e-01 4.39824969e-01
-2.74196684e-01 -4.65946823e-01 3.98378342e-01 3.08958697e-03
2.43319944e-01 -2.03612536e-01 -1.86751455e-01 2.40327775e-01
-5.87336540e-01 3.89487371e-02 3.40927184e-01 1.06332529e+00
9.58830714e-01 -8.23852643e-02 2.67791152e-01 8.56875837e-01
4.96571869e-01 5.00572741e-01 4.34529126e-01 -1.08773971e+00
6.35317683e-01 9.32939470e-01 -9.31998342e-02 -1.62305653e+00
-3.10896724e-01 -5.56623518e-01 -1.34602630e+00 -6.99219033e-02
1.67029530e-01 -5.69587899e-03 -3.10465008e-01 1.66341150e+00
6.09859407e-01 2.13793665e-01 1.24939069e-01 1.03963661e+00
8.54628444e-01 6.26300573e-01 -4.84216928e-01 -5.73543847e-01
1.26183558e+00 -8.49931121e-01 -6.87945247e-01 1.00205150e-02
3.55935305e-01 -8.99534523e-01 7.95298755e-01 4.35713887e-01
-9.06692743e-01 -5.33975840e-01 -1.01722419e+00 2.69019634e-01
-6.87752441e-02 3.55229318e-01 3.90037596e-01 3.14256400e-01
-7.35231400e-01 5.78682274e-02 -7.69240141e-01 -1.76211581e-01
1.42395645e-01 2.51124978e-01 -6.41876221e-01 -4.09822255e-01
-7.16016710e-01 2.02789232e-01 3.48036945e-01 2.87695259e-01
-3.16260248e-01 -4.63690788e-01 -7.72722423e-01 3.30246650e-02
7.19239950e-01 -5.34959674e-01 4.61556911e-01 -7.59261966e-01
-8.92712295e-01 5.54673672e-01 -4.97995168e-01 5.09733558e-01
3.24660718e-01 -5.32489605e-02 -5.44299543e-01 2.97472715e-01
4.20379102e-01 7.22706020e-02 6.88939691e-01 -1.82772017e+00
-4.66725022e-01 -7.61252642e-01 -1.95704222e-01 4.80090410e-01
-4.99601841e-01 -1.69393584e-01 -9.94875371e-01 -5.85991859e-01
7.99345255e-01 -9.49579000e-01 -2.59386986e-01 -1.89260364e-01
-5.01074672e-01 1.64911941e-01 1.12164068e+00 -5.06308496e-01
1.34398711e+00 -2.34121943e+00 6.50794923e-01 3.89313132e-01
4.32847887e-01 -4.71995361e-02 3.01532578e-02 4.76378798e-01
-1.24313161e-01 -2.24189952e-01 -3.44519407e-01 -1.47896990e-01
-3.91242594e-01 9.95854735e-02 1.51435629e-01 8.66447985e-01
-3.21005344e-01 4.23122197e-01 -8.24842513e-01 -5.99370658e-01
4.12014633e-01 4.53443021e-01 -5.06597638e-01 7.85427913e-02
4.09247816e-01 8.19667220e-01 -7.83363938e-01 6.27332568e-01
1.19062424e+00 -5.24848163e-01 3.73708218e-01 -5.99412441e-01
-7.48145580e-02 -5.03036320e-01 -1.85041273e+00 1.87350273e+00
-1.23839438e-01 -2.42793076e-02 5.32492816e-01 -1.20167458e+00
6.08552516e-01 4.02928144e-01 9.94957149e-01 -2.31959909e-01
1.19127765e-01 9.84498784e-02 2.38919668e-02 -4.50223118e-01
1.29600868e-01 -9.63517353e-02 1.01237558e-01 5.34583330e-01
-1.97427303e-01 3.75641465e-01 1.16805114e-01 3.04577827e-01
5.16217172e-01 -1.79559678e-01 4.65416789e-01 -3.60905588e-01
1.02074814e+00 -3.04498404e-01 1.03232145e+00 1.59986630e-01
-2.53339678e-01 8.99364114e-01 2.14302287e-01 -2.62036979e-01
-7.17176974e-01 -9.93751764e-01 -1.39853507e-01 4.65039760e-01
6.23663545e-01 -5.27669787e-01 -5.62886357e-01 -6.72103941e-01
-1.37798667e-01 2.10318476e-01 -4.89384651e-01 -7.41163865e-02
-3.96133035e-01 -8.92682612e-01 -1.47357062e-01 3.39513719e-01
5.83705783e-01 -3.45950305e-01 -2.60485619e-01 1.38744026e-01
-5.76388657e-01 -1.13210070e+00 -7.52748370e-01 -2.25708917e-01
-9.18316126e-01 -1.35719383e+00 -5.77232182e-01 -7.36845136e-01
9.75893974e-01 1.20448422e+00 7.07629740e-01 2.50106364e-01
-5.78983352e-02 5.74831426e-01 -4.42997545e-01 2.32117161e-01
8.58776420e-02 -1.10791229e-01 2.84510106e-01 5.47426522e-01
3.25976580e-01 -7.26383090e-01 -5.98193944e-01 5.97061217e-01
-9.90593791e-01 1.53508753e-01 3.28717053e-01 9.90391731e-01
8.84915352e-01 5.24393380e-01 3.29286546e-01 -7.32535422e-01
2.07222626e-01 -6.18146241e-01 -4.48826790e-01 4.43577886e-01
-4.08014864e-01 -3.49338502e-01 8.70898604e-01 -6.68728128e-02
-1.11403894e+00 2.59949535e-01 3.20927799e-01 -8.77569735e-01
-2.04373911e-01 4.86386478e-01 -7.93175042e-01 1.04341902e-01
-4.30976599e-02 6.69685483e-01 1.66139975e-01 -4.16871846e-01
3.92612308e-01 3.72297138e-01 2.80558646e-01 -5.07012308e-01
1.00746429e+00 9.36363399e-01 1.51898293e-02 -7.15530515e-01
-8.29249203e-01 -8.94911170e-01 -1.01746881e+00 -1.58452511e-01
6.98139548e-01 -1.16505373e+00 -5.90945780e-01 3.62009436e-01
-6.84480727e-01 5.74854910e-01 1.72290012e-01 6.61351740e-01
-3.53385687e-01 9.00045991e-01 -2.92399228e-01 -4.51026320e-01
-5.46669625e-02 -1.31417823e+00 8.79316449e-01 1.60563737e-01
2.38129035e-01 -1.05328178e+00 6.13495186e-02 5.74293196e-01
-1.03580065e-01 2.91994393e-01 6.32123768e-01 -3.54983479e-01
-5.49507737e-01 -6.15839511e-02 -1.77722558e-01 1.94666401e-01
6.62733316e-01 9.53013599e-02 -7.51642585e-01 -6.58340454e-01
5.41255116e-01 8.61510262e-02 7.08222330e-01 4.49266613e-01
1.14712632e+00 -1.78230062e-01 -4.14389193e-01 7.44036973e-01
1.63389170e+00 1.47932529e-01 1.53071284e-01 1.12925194e-01
1.14055479e+00 8.21504116e-01 6.21798575e-01 7.45151341e-01
5.18982351e-01 6.24884486e-01 3.46819788e-01 -1.69360843e-02
4.85675067e-01 6.07131720e-02 1.66165859e-01 1.45900214e+00
-1.42245740e-02 -2.64785383e-02 -7.22248316e-01 4.67305213e-01
-2.06392026e+00 -1.10387397e+00 -4.74733591e-01 2.21564245e+00
1.99265569e-01 -3.08253676e-01 1.16302311e-01 9.17763337e-02
1.07377589e+00 3.91353458e-01 -5.25363088e-01 3.78995061e-01
-2.75890261e-01 -6.47889256e-01 9.64879170e-02 2.98483968e-01
-1.23471522e+00 4.27793890e-01 4.67896748e+00 7.05680430e-01
-9.20522749e-01 1.03848822e-01 5.77548444e-01 -2.49555856e-01
-4.79027718e-01 1.27869129e-01 -4.27556604e-01 5.78018725e-01
1.67829394e-01 -3.01355243e-01 4.82850373e-01 6.66633427e-01
3.97583425e-01 -1.04649119e-01 -7.20254898e-01 1.17404509e+00
2.54330575e-01 -9.77418423e-01 6.29683658e-02 2.42462888e-01
9.31375921e-01 -2.07628727e-01 7.97282159e-02 -2.13506907e-01
-7.22744837e-02 -4.66587514e-01 4.30491358e-01 3.94874036e-01
4.88405317e-01 -1.05374897e+00 5.48320830e-01 4.07572299e-01
-1.75891507e+00 -4.39102091e-02 -5.73147297e-01 1.91664487e-01
1.10964835e-01 7.01454222e-01 6.79056793e-02 1.25150728e+00
7.78371572e-01 1.28638446e+00 -6.62930191e-01 7.34941840e-01
1.84972093e-01 8.01475495e-02 -1.61951825e-01 6.82004213e-01
2.02190354e-01 -8.94461870e-01 7.16868222e-01 5.30604362e-01
3.99178118e-01 2.87971735e-01 5.77771187e-01 6.79495394e-01
2.85927773e-01 3.15395772e-01 -7.36060202e-01 3.19780529e-01
4.77694571e-01 1.58311057e+00 -8.29751670e-01 -2.15879560e-01
-1.02265728e+00 1.01463211e+00 2.34267354e-01 4.19764638e-01
-6.26076162e-01 6.32031113e-02 6.77705050e-01 -1.26453638e-01
4.04646635e-01 -2.67265737e-01 -2.55097061e-01 -1.82251835e+00
3.72585863e-01 -1.01198828e+00 5.97870767e-01 -3.95090282e-01
-1.50721264e+00 3.31264734e-01 -1.50546864e-01 -2.00201821e+00
2.05243334e-01 -1.83720380e-01 -7.02836633e-01 6.60583913e-01
-1.16416657e+00 -1.06947517e+00 -4.80605692e-01 9.82588291e-01
3.87731165e-01 -2.06219345e-01 5.44622719e-01 3.70828092e-01
-8.84439170e-01 3.24429870e-01 7.22379267e-01 -2.66544726e-02
7.56108880e-01 -1.06387734e+00 -3.74092758e-01 9.05251026e-01
2.64236014e-02 9.13070381e-01 2.92552143e-01 -4.60619360e-01
-1.76679587e+00 -9.77728665e-01 2.30284810e-01 -2.00222120e-01
5.44266224e-01 -1.98008776e-01 -1.17373669e+00 5.81632197e-01
1.56835243e-01 1.81946158e-01 1.14265203e+00 5.91701865e-02
-2.69820839e-01 -1.87886953e-01 -9.25950468e-01 4.93513703e-01
7.80810773e-01 -3.92964810e-01 -2.95912862e-01 1.80345684e-01
2.94500321e-01 -1.70519724e-01 -1.04313946e+00 4.08349782e-01
4.58091617e-01 -1.39651322e+00 9.94372249e-01 -2.64510483e-01
2.97604173e-01 -8.50662291e-01 -5.01882076e-01 -1.29844260e+00
-5.59466004e-01 -1.49047717e-01 -7.00239316e-02 1.58261442e+00
-8.81912857e-02 -6.95927918e-01 6.68543518e-01 4.39235240e-01
5.22598950e-03 -8.06263924e-01 -8.81106853e-01 -7.76228964e-01
-1.31410450e-01 -1.77893275e-03 3.94924730e-01 1.37546635e+00
2.42045808e-05 3.01466316e-01 -4.71305430e-01 5.64293444e-01
1.05596292e+00 8.58744323e-01 7.30037093e-01 -1.26119733e+00
-2.51233160e-01 -2.77013034e-01 -3.90996814e-01 -7.04578757e-01
1.75110847e-01 -9.37623322e-01 -2.28954539e-01 -1.46242034e+00
8.91518712e-01 -3.04303139e-01 -4.76787001e-01 1.01567850e-01
-6.12088501e-01 1.15546184e-02 3.58112097e-01 6.39914036e-01
-7.57209301e-01 8.71107519e-01 1.37323630e+00 1.16192410e-02
-1.09098963e-01 -7.34037682e-02 -9.66783285e-01 8.76734555e-01
6.31730795e-01 -2.77923226e-01 -6.57750130e-01 -4.61100787e-01
-2.42067814e-01 2.20576406e-01 2.41350293e-01 -1.03724360e+00
3.79295290e-01 -2.65379488e-01 4.95248884e-01 -7.08230913e-01
3.24240237e-01 -1.24424493e+00 4.59496140e-01 3.28479141e-01
2.23694921e-01 2.09772587e-01 -1.34133920e-01 1.00770950e+00
-6.00515008e-01 3.73781323e-02 7.96298206e-01 -2.66433954e-01
-6.15183532e-01 5.27432621e-01 -6.47550747e-02 6.17927164e-02
1.21461093e+00 -1.64669067e-01 -1.19946226e-01 -3.15409541e-01
-7.51291156e-01 6.39053345e-01 9.10830617e-01 3.77015829e-01
6.34730756e-01 -1.75095880e+00 -6.57958031e-01 2.87608802e-01
3.85900468e-01 3.85197736e-02 9.05746698e-01 1.04612696e+00
-5.76131940e-02 2.78737783e-01 -2.56989509e-01 -8.86116982e-01
-1.52534914e+00 8.85333717e-01 6.94659278e-02 -1.59928977e-01
-6.44949317e-01 4.31289017e-01 7.07147181e-01 -5.55156589e-01
-2.65448302e-01 2.18423739e-01 -3.75193089e-01 3.41854393e-01
4.38471437e-01 5.37480533e-01 -2.31192395e-01 -1.11708462e+00
-4.90246892e-01 1.11280286e+00 -4.02402692e-02 1.60416618e-01
1.35677266e+00 -6.67437196e-01 -3.50197911e-01 5.50917685e-01
1.47073388e+00 2.39667207e-01 -1.18350554e+00 -5.34794688e-01
-3.29685777e-01 -8.26487243e-01 4.70156334e-02 -4.70237620e-02
-1.61675656e+00 8.32844853e-01 3.95454079e-01 5.87815512e-03
1.41515982e+00 -1.00883976e-01 5.40336549e-01 1.14574142e-01
4.32992399e-01 -9.39767957e-01 3.56496088e-02 7.19955489e-02
7.28381276e-01 -1.46483994e+00 4.31392759e-01 -7.96377599e-01
-8.70694816e-01 1.01877689e+00 7.20606863e-01 -5.49855791e-02
7.87133038e-01 -2.69623071e-01 3.04851178e-02 -4.68146592e-01
-4.62426424e-01 -4.05207425e-02 3.52179915e-01 2.68813103e-01
2.60868639e-01 7.26989880e-02 -2.70598173e-01 5.95024168e-01
3.70687693e-01 -5.21984279e-01 4.70447242e-01 8.95200312e-01
-1.02988072e-01 -1.14667249e+00 -6.54562414e-01 3.80339921e-01
-4.02992010e-01 1.96481720e-01 -2.14049086e-01 5.61093986e-01
5.77098988e-02 1.12085295e+00 -2.32917905e-01 -4.38232154e-01
1.47134975e-01 -2.18968511e-01 1.44511685e-01 -6.25464261e-01
-1.86450064e-01 7.26520121e-01 -4.89992082e-01 -5.82223594e-01
-8.03541124e-01 -1.09205794e+00 -1.13089550e+00 -2.82441556e-01
-3.31822067e-01 2.78171331e-01 1.27703935e-01 6.34677351e-01
4.62020099e-01 3.28854442e-01 1.17239535e+00 -6.53276384e-01
-2.12584123e-01 -4.79200125e-01 -1.03596461e+00 5.30749023e-01
1.44677192e-01 -8.50369155e-01 -3.65654707e-01 8.19138810e-02] | [8.239219665527344, 4.646291255950928] |
f40db8ef-a41d-4cb9-bece-4575e1d0ddf3 | mixbin-towards-budgeted-binarization | 2211.06739 | null | https://arxiv.org/abs/2211.06739v1 | https://arxiv.org/pdf/2211.06739v1.pdf | MixBin: Towards Budgeted Binarization | Binarization has proven to be amongst the most effective ways of neural network compression, reducing the FLOPs of the original model by a large extent. However, such levels of compression are often accompanied by a significant drop in the performance. There exist some approaches that reduce this performance drop by facilitating partial binarization of the network, however, a systematic approach to mix binary and full-precision parameters in a single network is still missing. In this paper, we propose a paradigm to perform partial binarization of neural networks in a controlled sense, thereby constructing budgeted binary neural network (B2NN). We present MixBin, an iterative search-based strategy that constructs B2NN through optimized mixing of the binary and full-precision components. MixBin allows to explicitly choose the approximate fraction of the network to be kept as binary, thereby presenting the flexibility to adapt the inference cost at a prescribed budget. We demonstrate through experiments that B2NNs obtained from our MixBin strategy are significantly better than those obtained from random selection of the network layers. To perform partial binarization in an effective manner, it is important that both the full-precision as well as the binary components of the B2NN are appropriately optimized. We also demonstrate that the choice of the activation function can have a significant effect on this process, and to circumvent this issue, we present BinReLU, that can be used as an effective activation function for the full-precision as well as the binary components of any B2NN. Experimental investigations reveal that BinReLU outperforms the other activation functions in all possible scenarios of B2NN: zero-, partial- as well as full binarization. Finally, we demonstrate the efficacy of MixBin on the tasks of classification and object tracking using benchmark datasets. | ['Deepak K. Gupta', 'Dilip K. Prasad', 'Neeraj Anand', 'Udbhav Bamba'] | 2022-11-12 | null | null | null | null | ['neural-network-compression', 'neural-network-compression'] | ['methodology', 'miscellaneous'] | [ 2.67839551e-01 -6.44133091e-02 -3.10363144e-01 -3.68413121e-01
-2.86340028e-01 -3.69836330e-01 3.69875044e-01 4.20226380e-02
-8.13186586e-01 8.39627564e-01 -4.44689065e-01 -4.96382475e-01
-4.50592816e-01 -9.53632772e-01 -1.04249144e+00 -9.24212158e-01
2.35207658e-02 4.89645541e-01 5.21961212e-01 1.56478751e-02
1.19407400e-01 9.23646688e-01 -1.75267577e+00 2.51457632e-01
5.87819517e-01 1.24628496e+00 1.32189140e-01 6.83220565e-01
-6.93394467e-02 5.77246547e-01 -1.08472455e+00 -5.13271332e-01
5.77680290e-01 -1.79384917e-01 -6.01240337e-01 -4.30259377e-01
5.15070796e-01 -3.84382635e-01 -1.49376750e-01 1.14144063e+00
4.86343205e-01 1.88202292e-01 5.26575208e-01 -1.09217751e+00
1.28489956e-01 1.00627887e+00 -1.81816697e-01 3.99713546e-01
-4.52941895e-01 1.93917036e-01 7.21301436e-01 -3.52118760e-01
2.99108833e-01 1.27513480e+00 7.73276269e-01 3.28965038e-01
-1.42850220e+00 -8.77663851e-01 -1.67056128e-01 9.97735336e-02
-1.69639623e+00 -4.77969199e-01 5.63831568e-01 -1.10167958e-01
8.98463786e-01 5.27018428e-01 7.65056193e-01 4.57161635e-01
-9.34510212e-03 3.84467244e-01 9.45584297e-01 -7.19331503e-01
3.42477053e-01 1.54085934e-01 4.61906433e-01 7.30806112e-01
6.77760303e-01 8.26697350e-02 -2.81107664e-01 -1.21325374e-01
7.92391062e-01 -2.30944574e-01 -3.16722542e-01 -2.57548153e-01
-6.21332228e-01 8.23171139e-01 6.34164095e-01 3.25738370e-01
-3.28525841e-01 6.52025163e-01 4.24664378e-01 1.56480640e-01
1.09783716e-01 6.65745437e-01 -4.08197790e-01 -8.97152349e-02
-1.21520019e+00 4.79044467e-01 9.61060166e-01 6.87771261e-01
8.39459419e-01 1.51991904e-01 -4.18739021e-01 9.94273186e-01
-7.53729790e-02 1.38490632e-01 5.83723187e-01 -9.52974260e-01
2.43221223e-01 5.95981419e-01 -3.52313936e-01 -9.23344851e-01
-1.96013182e-01 -4.73124772e-01 -1.02496767e+00 4.30311680e-01
6.35702789e-01 -1.67053908e-01 -1.14373744e+00 1.62602806e+00
2.83191592e-01 -2.05745131e-01 -1.52079642e-01 6.51155591e-01
6.00525796e-01 6.90622211e-01 -8.25855881e-02 -1.44463122e-01
1.27116120e+00 -9.17479455e-01 -5.32021940e-01 -1.05549887e-01
2.61478126e-01 -4.25453603e-01 7.62040436e-01 6.18477106e-01
-1.46527016e+00 -5.24925768e-01 -1.41341484e+00 1.63972691e-01
-5.55638194e-01 8.98806453e-02 5.62488616e-01 9.88286495e-01
-1.14055049e+00 1.03617406e+00 -8.67657483e-01 2.29231477e-01
5.09745300e-01 1.01952732e+00 -8.02307799e-02 7.09906146e-02
-1.02201009e+00 9.09723938e-01 1.04018235e+00 7.98560083e-02
-4.90298301e-01 -6.03979826e-01 -6.00629807e-01 5.36677778e-01
3.99100989e-01 -4.37109023e-01 1.18062294e+00 -9.48504388e-01
-1.18533957e+00 3.58475059e-01 1.66457355e-01 -1.09326780e+00
3.61977935e-01 2.19247401e-01 1.11146130e-01 2.06825763e-01
-6.19612098e-01 1.01407254e+00 8.94373059e-01 -9.12652075e-01
-5.44508874e-01 -1.29298046e-01 1.80701569e-01 2.80905627e-02
-6.12928271e-01 -1.83518216e-01 -6.46946490e-01 -5.73264658e-01
-3.98480184e-02 -9.23291683e-01 -2.04802856e-01 1.00630306e-01
-3.99466276e-01 -1.52323306e-01 6.17783427e-01 -4.00939256e-01
1.38475478e+00 -2.04195142e+00 -1.52345687e-01 5.33329666e-01
2.83140004e-01 7.88670003e-01 1.57074764e-01 -2.35226139e-01
-1.52552992e-01 3.67338657e-01 -1.68878973e-01 -3.69001687e-01
-5.11946268e-02 5.50026178e-01 6.32505566e-02 2.88776606e-01
2.04550132e-01 5.40414631e-01 -4.32408959e-01 -5.71752787e-01
5.99837303e-02 6.15182519e-01 -6.97784066e-01 1.01138316e-01
-2.41402730e-01 -3.60192776e-01 -9.30599123e-02 4.75728601e-01
7.42214203e-01 -9.69374850e-02 1.94769442e-01 -4.46098655e-01
1.95627231e-02 4.13354993e-01 -1.26576185e+00 7.18339384e-01
-6.98980242e-02 6.29558027e-01 1.58413455e-01 -1.12940228e+00
9.41306710e-01 2.73911078e-02 3.39524746e-01 -5.46353340e-01
3.35572183e-01 1.95560113e-01 3.31733882e-01 1.71609819e-01
6.85075939e-01 -6.54220656e-02 2.13600069e-01 2.14188561e-01
1.14109211e-01 -1.01440281e-01 6.99324250e-01 -1.26896322e-01
9.43804026e-01 -3.31366718e-01 3.23718339e-01 -2.39981309e-01
3.97550285e-01 -1.65221274e-01 3.53226870e-01 1.08933580e+00
-7.52257034e-02 5.84236562e-01 7.67000437e-01 -3.19350719e-01
-1.08508193e+00 -5.92103779e-01 -4.79164302e-01 9.41614091e-01
-1.63885653e-01 -2.23963171e-01 -1.09521163e+00 -5.74022532e-01
2.78311148e-02 5.73872805e-01 -4.42008138e-01 -3.34775716e-01
-7.31141806e-01 -1.07115245e+00 9.48353410e-01 3.41544122e-01
6.62398219e-01 -1.02662408e+00 -8.71267557e-01 1.64449010e-02
1.59721717e-01 -8.88356328e-01 -2.21911535e-01 1.01653183e+00
-1.03863060e+00 -8.29278886e-01 -6.53287172e-01 -4.27881777e-01
6.75435126e-01 1.15847783e-02 1.00113976e+00 4.96063739e-01
-1.01097256e-01 -3.34964156e-01 -9.18384939e-02 -2.84181774e-01
-6.63265169e-01 4.55103546e-01 -2.43580431e-01 -4.18979913e-01
8.14403743e-02 -5.11595130e-01 -4.05932963e-01 2.18451768e-01
-1.24902499e+00 5.20409271e-02 6.22778237e-01 8.17321241e-01
5.58590651e-01 4.98909771e-01 2.35513180e-01 -7.07348228e-01
6.56720161e-01 -8.75473320e-02 -9.41199899e-01 7.43973628e-02
-9.39050138e-01 4.07655180e-01 7.71601200e-01 -7.76792586e-01
-4.66797560e-01 -1.92676252e-03 -3.24030101e-01 -6.22037411e-01
6.85517676e-03 4.43550497e-01 -8.74483213e-02 -4.38387334e-01
7.19943762e-01 -7.52010271e-02 1.74874812e-01 -4.37805563e-01
6.60402700e-02 6.02312326e-01 4.27861273e-01 -5.25172174e-01
2.84672499e-01 7.90346861e-02 1.82646766e-01 -7.10115731e-01
-4.71095264e-01 -2.00309917e-01 -3.74149442e-01 -6.92019193e-03
4.63733137e-01 -3.59792680e-01 -8.08924019e-01 3.96911323e-01
-1.12818134e+00 -4.59533066e-01 -4.31674272e-01 1.44783422e-01
-9.26160589e-02 1.20277978e-01 -4.42423373e-01 -7.61671305e-01
-4.41586614e-01 -1.52032256e+00 6.06468618e-01 2.83363640e-01
-1.04500704e-01 -6.34725928e-01 -4.33469087e-01 2.52931076e-03
4.61405575e-01 -3.88870463e-02 1.03242540e+00 -9.80709136e-01
-5.74967980e-01 -2.50889808e-01 -4.16493297e-01 8.24556649e-01
-2.01575547e-01 1.52096853e-01 -7.45686531e-01 -1.92503288e-01
-2.17835372e-03 -1.37160942e-01 1.01742923e+00 5.74719131e-01
1.42159843e+00 -6.55426204e-01 -2.73051530e-01 8.32641602e-01
1.44309247e+00 3.93671572e-01 7.70481765e-01 4.54535902e-01
5.06468117e-01 2.64847934e-01 1.74653009e-01 2.65058786e-01
-9.98138115e-02 7.91420817e-01 6.37041032e-01 -6.87620556e-03
-2.52262354e-01 1.86648384e-01 5.29190078e-02 4.13147062e-01
6.45433664e-02 -2.86327779e-01 -7.21815944e-01 3.13997060e-01
-1.35910022e+00 -8.49497557e-01 1.83256999e-01 2.48320079e+00
1.26529360e+00 6.17483497e-01 2.33021140e-01 6.13051534e-01
7.67186284e-01 9.57481563e-02 -4.05226022e-01 -7.39064693e-01
-7.31539447e-03 4.15786117e-01 1.03316522e+00 4.64556843e-01
-1.05704749e+00 6.02718353e-01 6.56792068e+00 1.19914353e+00
-1.34023666e+00 -8.47234055e-02 8.23585093e-01 -3.84795070e-01
1.10296071e-01 -1.59210712e-01 -1.30585110e+00 6.99321091e-01
1.21657693e+00 2.38445625e-01 6.45826101e-01 9.48763728e-01
-2.56576031e-01 -2.25831985e-01 -9.02702689e-01 6.99906468e-01
-8.56162906e-02 -1.35857785e+00 4.10934277e-02 9.49677974e-02
4.94186819e-01 -1.65829867e-01 -2.10680261e-01 3.97024572e-01
1.07176170e-01 -1.01026773e+00 8.07606339e-01 1.78556338e-01
7.47241735e-01 -1.08582139e+00 7.80123889e-01 4.04858917e-01
-8.22790027e-01 -2.64408737e-01 -5.42939901e-01 7.51665458e-02
-8.40805471e-02 8.23873639e-01 -1.00364745e+00 -1.35086495e-02
8.39336812e-01 -1.31485954e-01 -6.54937685e-01 1.19439840e+00
1.59807220e-01 7.07803607e-01 -7.35416472e-01 -3.04310322e-01
1.57654807e-01 -9.96279418e-02 4.01314080e-01 1.32691085e+00
1.41484499e-01 -1.62451282e-01 -1.87425047e-01 8.58326554e-01
-1.36163771e-01 -2.24603653e-01 -2.46899262e-01 -1.01371318e-01
6.76379502e-01 1.04139233e+00 -9.37329352e-01 -4.81582910e-01
2.55101472e-01 4.00223881e-01 4.50961143e-01 9.61925089e-02
-9.41546321e-01 -5.09760559e-01 4.34873581e-01 1.08353697e-01
7.40279734e-01 3.23946513e-02 -4.44134653e-01 -4.30533171e-01
-4.35120910e-02 -1.08904052e+00 2.26341456e-01 -5.49534142e-01
-5.26989818e-01 7.87733316e-01 4.97959346e-01 -6.06915891e-01
-2.64122427e-01 -7.12532997e-01 -1.86427817e-01 8.12204063e-01
-1.28404808e+00 -5.98183393e-01 -2.29090288e-01 2.33649999e-01
1.81870997e-01 1.90610383e-02 4.21483517e-01 4.91093397e-01
-6.93922102e-01 9.46812868e-01 -1.81409586e-02 -5.64868934e-02
3.18985730e-01 -1.04050922e+00 1.65590085e-02 7.38288045e-01
-3.78012434e-02 8.92440796e-01 8.10253382e-01 -4.63976204e-01
-1.01440978e+00 -9.17046547e-01 6.16780877e-01 2.46621698e-01
1.61930218e-01 -2.02241644e-01 -9.78402138e-01 4.11692441e-01
-1.19064882e-01 -1.87647879e-01 1.79957241e-01 -8.20718631e-02
6.11665379e-03 -5.60661435e-01 -1.27644730e+00 7.41132021e-01
5.90562582e-01 -4.25698375e-03 -2.57421792e-01 1.55694336e-01
8.01189780e-01 -6.25803351e-01 -7.30823696e-01 6.82373405e-01
6.22592449e-01 -1.23729587e+00 1.20650506e+00 -9.91225243e-03
3.89300376e-01 -3.84452008e-02 -1.11294262e-01 -9.52922404e-01
-5.04061803e-02 -4.12166327e-01 -4.77302462e-01 1.11402249e+00
3.19881827e-01 -6.35272324e-01 1.04980898e+00 6.19854569e-01
-1.83480904e-02 -1.09707844e+00 -9.82388616e-01 -7.56886780e-01
-1.63323351e-03 -3.48541558e-01 7.62923837e-01 3.96456599e-01
-5.22604883e-01 -4.33951579e-02 -1.36802867e-01 -1.71736151e-01
2.57333130e-01 -2.43916556e-01 6.97749436e-01 -1.06383646e+00
-4.77096438e-01 -9.63887095e-01 -3.41452569e-01 -1.14355898e+00
-1.37277901e-01 -6.61026299e-01 2.39873096e-01 -1.01302576e+00
1.56604983e-02 -9.32389975e-01 -1.77382961e-01 6.09147787e-01
6.16365857e-02 3.87699515e-01 2.54233599e-01 1.57491520e-01
-3.72134969e-02 1.06470183e-01 1.03426385e+00 -3.81821245e-02
-2.90307462e-01 1.21922992e-01 -5.66740870e-01 6.73304975e-01
7.52200127e-01 -6.84334457e-01 -3.02974045e-01 -1.98839530e-01
5.49821071e-02 -4.69368398e-02 3.50399435e-01 -1.20766318e+00
3.96598846e-01 -1.90405026e-02 3.59619230e-01 -7.27614760e-01
5.67009509e-01 -9.52685297e-01 2.43960440e-01 5.86121559e-01
-4.25436407e-01 1.42308474e-01 3.69534612e-01 5.33469208e-02
-5.09773903e-02 -1.01491857e+00 1.06982315e+00 -5.11989333e-02
-1.67139217e-01 1.90747716e-02 -2.52569377e-01 -3.75894934e-01
6.66505754e-01 -5.38315415e-01 -2.96795696e-01 -1.21829607e-01
-2.88165331e-01 -1.16841443e-01 2.74511158e-01 -1.07450634e-01
2.54716545e-01 -9.99951839e-01 -1.59873873e-01 4.79245007e-01
-6.24544084e-01 3.59835565e-01 -7.01356605e-02 6.74689829e-01
-8.36407900e-01 5.21845102e-01 -2.24111944e-01 -5.02882957e-01
-1.45975447e+00 5.23903251e-01 5.92071712e-01 -6.05139077e-01
-3.33117515e-01 8.68560433e-01 -2.21973121e-01 -2.38783941e-01
7.33733177e-01 -6.80146217e-01 -2.20601276e-01 4.12734225e-02
4.84912962e-01 5.45982003e-01 4.12184864e-01 -3.29764485e-01
-8.28688741e-02 8.35760012e-02 -1.05279393e-01 -6.30555227e-02
1.06100345e+00 2.60053307e-01 -3.03261876e-01 9.12481919e-02
1.16377568e+00 -3.28130692e-01 -1.20664537e+00 1.47457868e-01
-2.25558177e-01 -2.40116462e-01 2.79161334e-01 -7.06429839e-01
-1.40254772e+00 6.92901850e-01 5.61083138e-01 5.44594467e-01
1.35129428e+00 -4.50018734e-01 6.69276655e-01 5.05547225e-01
1.71101585e-01 -8.50518465e-01 -2.24900991e-01 4.83169168e-01
5.79435945e-01 -6.97816193e-01 3.35540682e-01 -2.85326689e-01
-8.50869119e-02 1.01244724e+00 6.91402793e-01 -2.26220682e-01
5.61863363e-01 6.48990095e-01 -3.28596026e-01 -2.81692874e-02
-6.64895296e-01 1.28054172e-01 3.64407390e-01 2.45596290e-01
1.20325036e-01 -1.41014442e-01 -4.47160602e-01 4.64624524e-01
-4.52591777e-01 -3.26711163e-02 3.41284931e-01 8.78897846e-01
-4.82434958e-01 -1.11886859e+00 -7.26104200e-01 7.25646794e-01
-7.40369141e-01 -2.66387045e-01 -9.56241712e-02 8.84605050e-01
3.64856809e-01 5.21819115e-01 2.61502832e-01 -3.24278444e-01
1.00907370e-01 3.31438109e-02 5.34849703e-01 -3.11070234e-01
-9.96928573e-01 4.89672013e-02 1.84790596e-01 -3.51014465e-01
-6.79903924e-02 -1.95242345e-01 -1.09522426e+00 -4.51858848e-01
-7.36119270e-01 8.88623744e-02 1.06418872e+00 9.61008370e-01
4.95585240e-02 7.27068901e-01 7.46255890e-02 -1.13805056e+00
-9.49430823e-01 -8.25883746e-01 -4.02084887e-01 1.30170481e-02
4.53217924e-01 -6.46452665e-01 -5.53696513e-01 -1.34645835e-01] | [8.578322410583496, 3.1624035835266113] |
6b3acdf3-3050-4109-a373-65a086add788 | ssc-semantic-scan-context-for-large-scale | 2107.00382 | null | https://arxiv.org/abs/2107.00382v2 | https://arxiv.org/pdf/2107.00382v2.pdf | SSC: Semantic Scan Context for Large-Scale Place Recognition | Place recognition gives a SLAM system the ability to correct cumulative errors. Unlike images that contain rich texture features, point clouds are almost pure geometric information which makes place recognition based on point clouds challenging. Existing works usually encode low-level features such as coordinate, normal, reflection intensity, etc., as local or global descriptors to represent scenes. Besides, they often ignore the translation between point clouds when matching descriptors. Different from most existing methods, we explore the use of high-level features, namely semantics, to improve the descriptor's representation ability. Also, when matching descriptors, we try to correct the translation between point clouds to improve accuracy. Concretely, we propose a novel global descriptor, Semantic Scan Context, which explores semantic information to represent scenes more effectively. We also present a two-step global semantic ICP to obtain the 3D pose (x, y, yaw) used to align the point cloud to improve matching performance. Our experiments on the KITTI dataset show that our approach outperforms the state-of-the-art methods with a large margin. Our code is available at: https://github.com/lilin-hitcrt/SSC. | ['Yong liu', 'Tianxin Huang', 'Xiangrui Zhao', 'Xin Kong', 'Lin Li'] | 2021-07-01 | null | null | null | null | ['visual-place-recognition'] | ['computer-vision'] | [-7.27657601e-02 -5.29044390e-01 -2.07538918e-01 -6.32936180e-01
-7.26387799e-01 -4.11098003e-01 6.79893732e-01 2.96959281e-01
-2.44844347e-01 2.34573916e-01 -4.85754423e-02 1.33445501e-01
-5.36994450e-02 -1.10052884e+00 -7.89272547e-01 -5.14797807e-01
1.80050567e-01 5.24859011e-01 5.04668415e-01 -1.38688281e-01
7.76057780e-01 8.62216473e-01 -1.69540572e+00 -1.77687198e-01
7.12060630e-01 1.16363525e+00 3.15694928e-01 2.63342299e-02
-3.77677888e-01 3.70260000e-01 -2.48908818e-01 -5.01005352e-02
3.53946775e-01 5.04201453e-04 -5.81762969e-01 -1.37806134e-02
5.49336672e-01 -1.73796400e-01 -2.51884639e-01 1.16551149e+00
2.50140429e-01 1.51752397e-01 4.40093040e-01 -1.35265183e+00
-4.14740741e-01 -2.02186495e-01 -5.21620989e-01 -3.35542053e-01
7.91074514e-01 -7.15782791e-02 8.82480264e-01 -1.04681551e+00
7.53904760e-01 1.16817021e+00 8.00978124e-01 -5.77366017e-02
-8.53029788e-01 -8.60480785e-01 3.81456949e-02 3.98057252e-01
-1.81892824e+00 -2.68347383e-01 1.00865567e+00 -3.37957591e-01
6.84538186e-01 3.76944542e-01 7.16854692e-01 6.89592361e-01
8.14596750e-03 4.51716840e-01 9.51758742e-01 -3.63515615e-01
2.96479821e-01 -1.64265573e-01 -1.79956481e-01 6.20176196e-01
1.83314607e-01 -1.85845718e-02 -6.26620650e-01 -3.80480468e-01
9.35089886e-01 5.52173555e-01 -1.60297886e-01 -7.83311784e-01
-1.45936477e+00 8.05996835e-01 8.38596106e-01 1.52210072e-01
-3.97035748e-01 3.16001475e-01 -1.33369835e-02 -1.50539586e-02
2.54619539e-01 2.74041265e-01 -1.82980850e-01 -1.81017175e-01
-6.89630926e-01 3.97204727e-01 5.41228890e-01 1.26268423e+00
1.50940609e+00 -4.94455963e-01 2.40791187e-01 7.46347368e-01
5.06496072e-01 8.38413596e-01 3.29243988e-01 -1.02696216e+00
4.70422149e-01 7.41486192e-01 8.97447318e-02 -1.53397727e+00
-3.70076507e-01 3.01325452e-02 -5.77578187e-01 3.39860916e-02
-6.56147376e-02 6.25873923e-01 -8.44113767e-01 1.08984864e+00
5.20266473e-01 4.87622768e-01 -2.09374651e-01 1.16745746e+00
6.60423994e-01 4.94798869e-01 -3.01298529e-01 4.23984200e-01
1.26153505e+00 -7.79087186e-01 -4.13963825e-01 -4.25281763e-01
6.70184374e-01 -1.12930918e+00 6.47944033e-01 7.33609200e-02
-5.74134231e-01 -2.52300501e-01 -9.73166764e-01 -3.16129774e-01
-4.77693349e-01 -1.09674213e-02 6.38381124e-01 2.21535996e-01
-8.23341191e-01 5.96026480e-01 -9.47241902e-01 -5.89348257e-01
1.17953122e-01 2.48127088e-01 -5.11178434e-01 -2.80230075e-01
-9.76139665e-01 8.82982969e-01 2.45127857e-01 -1.02607049e-01
-2.60047197e-01 -5.24709940e-01 -1.12537587e+00 -2.64115095e-01
2.59410292e-01 -4.95256186e-01 1.02964616e+00 -4.58064109e-01
-1.40397787e+00 8.50548029e-01 -5.09164214e-01 7.70818628e-03
3.01562935e-01 4.92303520e-02 -4.35388274e-02 1.81843489e-01
3.74101073e-01 6.28106356e-01 4.16162282e-01 -1.40692675e+00
-6.36156678e-01 -5.48848271e-01 2.18885727e-02 3.24049592e-01
2.09416062e-01 -1.34000957e-01 -6.63531899e-01 -3.41487616e-01
1.05755758e+00 -1.15437353e+00 -2.11382404e-01 3.44921708e-01
-2.97884226e-01 -6.69285432e-02 8.96260977e-01 -3.10014337e-01
6.96567833e-01 -2.39031506e+00 -1.42933860e-01 6.25996113e-01
2.01732498e-02 -2.08334118e-01 2.14740448e-02 5.19593954e-01
1.71256155e-01 -5.50519638e-02 -2.44259015e-01 -5.01094162e-01
1.68716207e-01 4.48045433e-01 -1.35550022e-01 8.32302392e-01
1.87302437e-02 6.88978255e-01 -9.13100123e-01 -3.81112844e-01
5.90751767e-01 6.23847544e-01 -3.68844271e-01 -1.04908563e-01
9.61923674e-02 4.32637244e-01 -8.06500673e-01 9.21173096e-01
9.81959224e-01 -1.08241700e-01 -3.35760236e-01 -1.23319745e-01
-3.79158646e-01 4.55141842e-01 -1.37494659e+00 2.12780809e+00
-3.69255960e-01 4.31271553e-01 -2.20114946e-01 -6.03314281e-01
1.35970461e+00 -1.11092981e-02 5.74544668e-01 -7.95222461e-01
8.37399736e-02 4.61972564e-01 -6.08190417e-01 2.86942143e-02
6.86840117e-01 1.82788834e-01 -1.12584643e-01 -4.23432700e-02
-3.83066475e-01 -5.99663794e-01 -3.29207450e-01 -1.09245174e-01
9.25280094e-01 1.66326478e-01 3.73325735e-01 -1.92129463e-02
4.88865316e-01 2.69928902e-01 6.64471745e-01 5.47235668e-01
1.06119281e-02 1.05978370e+00 8.40856209e-02 -3.50053370e-01
-1.04010797e+00 -8.71245444e-01 -2.54118353e-01 4.50816035e-01
8.94127488e-01 -6.39100552e-01 -2.96380311e-01 -1.57080755e-01
2.97162294e-01 3.56335491e-01 -2.22557738e-01 3.99458408e-02
-4.43554521e-01 -1.99240297e-01 1.48546651e-01 4.90081459e-01
5.99138498e-01 -5.33568621e-01 -6.46143019e-01 3.83184180e-02
-2.73562878e-01 -1.14495957e+00 -2.60519564e-01 -7.14191422e-02
-9.52428341e-01 -9.36473787e-01 -4.22931254e-01 -6.18237734e-01
7.98464835e-01 7.37392128e-01 6.76292121e-01 2.62370974e-01
-5.65150902e-02 3.58050495e-01 -6.96262658e-01 -3.25329810e-01
3.22522670e-01 1.01759873e-01 2.89243460e-02 -2.43131578e-01
5.58890343e-01 -5.87627709e-01 -6.70632362e-01 6.39468670e-01
-6.96028173e-01 1.91344067e-01 5.11777759e-01 5.60289085e-01
1.05635190e+00 -3.28193873e-01 -3.48976046e-01 -4.60463703e-01
-6.41669929e-02 -3.61133546e-01 -6.76018715e-01 1.04984315e-02
-3.08769852e-01 -1.69257936e-03 1.78053036e-01 -1.01710185e-01
-5.07346153e-01 2.78907955e-01 -9.37659293e-02 -5.86526394e-01
-3.01649839e-01 5.11444688e-01 -1.39904097e-01 -6.10083401e-01
3.62651289e-01 3.39239687e-01 -3.90530750e-02 -5.99179327e-01
1.93735287e-01 5.80195427e-01 3.87946725e-01 -6.61475122e-01
1.03711545e+00 7.74756014e-01 2.31273577e-01 -7.15920269e-01
-5.12961447e-01 -9.95171726e-01 -9.45610166e-01 2.48528328e-02
6.28131926e-01 -9.80004489e-01 -5.95714629e-01 3.29556465e-01
-1.34477019e+00 1.40440598e-01 7.27393851e-02 6.92574918e-01
-7.58524776e-01 3.04146916e-01 -2.12310791e-01 -4.82194453e-01
-4.09410372e-02 -1.32798529e+00 1.60679555e+00 1.81724861e-01
-4.46226448e-03 -7.07861006e-01 1.85862735e-01 1.35185897e-01
3.93353164e-01 4.39452767e-01 3.55924308e-01 -3.36562067e-01
-1.09615719e+00 -3.74942571e-01 -3.05695653e-01 -1.83687821e-01
1.53864339e-01 4.75561321e-02 -8.03485334e-01 -1.37249872e-01
-4.98968475e-02 2.15968788e-01 5.09535074e-01 -5.63231036e-02
1.14270639e+00 1.13513637e-02 -5.97782433e-01 1.01008749e+00
1.61259961e+00 1.12863563e-01 8.19895983e-01 9.09329951e-01
8.39709818e-01 3.03939432e-01 9.72610414e-01 5.87717116e-01
6.78734481e-01 1.01301289e+00 6.09566033e-01 -1.66206267e-02
1.04902692e-01 -4.28877324e-01 3.21005248e-02 8.81468296e-01
-1.93769395e-01 2.34543279e-01 -1.25734437e+00 2.62575626e-01
-2.05216360e+00 -6.79420948e-01 -2.26821154e-01 2.37753582e+00
3.00784111e-01 -1.85606629e-01 -4.85250741e-01 -2.44251755e-03
6.63137257e-01 2.40827069e-01 -1.38313100e-01 2.88514215e-02
-5.98750897e-02 5.28576337e-02 9.76634085e-01 6.03732049e-01
-1.02150404e+00 1.10390937e+00 5.13318729e+00 7.14205563e-01
-1.40944815e+00 4.76890653e-02 -6.02733567e-02 1.41263872e-01
-2.73169905e-01 4.74721611e-01 -7.46839523e-01 4.34016138e-01
5.56546271e-01 -1.04837887e-01 3.17201912e-01 1.12102354e+00
-1.04799997e-02 -2.57339120e-01 -9.23956692e-01 1.41897011e+00
1.19843565e-01 -1.23999679e+00 -7.01820627e-02 1.99851796e-01
6.32323623e-01 2.78722227e-01 -1.30490839e-01 -3.57445478e-02
-1.36617914e-01 -7.03817606e-01 9.94873583e-01 6.27260029e-01
6.73089802e-01 -7.16064513e-01 7.13707209e-01 2.31995314e-01
-1.43607092e+00 3.46537769e-01 -6.98711455e-01 -1.90667287e-01
8.99301022e-02 5.90749085e-01 -7.92678535e-01 9.36403751e-01
7.11352408e-01 1.01602781e+00 -7.26484239e-01 1.34276021e+00
-2.81557530e-01 -8.40610191e-02 -7.43245423e-01 1.00182168e-01
3.55183423e-01 -3.56696218e-01 5.15115917e-01 7.77704477e-01
7.82787859e-01 2.17841908e-01 4.24486548e-01 7.25234151e-01
2.69970953e-01 1.98839962e-01 -7.92193413e-01 3.79695028e-01
8.23183656e-01 9.77180600e-01 -7.27066219e-01 -2.07584396e-01
-4.13035125e-01 9.37060177e-01 9.99553129e-02 1.92084759e-01
-6.49458528e-01 -4.99507129e-01 9.39375222e-01 2.70545036e-01
1.63552135e-01 -7.56592393e-01 -3.25674653e-01 -1.21485090e+00
3.13416928e-01 -3.79316807e-01 -1.36236221e-01 -9.56644893e-01
-9.58387196e-01 3.01986486e-01 2.61994544e-02 -1.75627518e+00
-1.28826037e-01 -5.41010797e-01 -3.34373921e-01 8.64182711e-01
-1.68643928e+00 -1.24583983e+00 -8.52926910e-01 6.47298872e-01
2.00187296e-01 2.89069235e-01 6.41370118e-01 4.27750468e-01
-2.36822274e-02 1.37984112e-01 2.68709391e-01 2.33127326e-01
7.52472162e-01 -8.84262145e-01 5.04940569e-01 6.00619197e-01
1.48298368e-01 7.06286252e-01 7.28649855e-01 -6.34501815e-01
-1.75406039e+00 -8.43087435e-01 8.57079387e-01 -4.70849276e-01
5.54264069e-01 -4.03154284e-01 -9.07265961e-01 8.10483694e-01
-4.93640721e-01 1.81372285e-01 3.02812219e-01 4.09822725e-02
-4.27938581e-01 -2.28980169e-01 -1.08539009e+00 4.12339330e-01
1.12481344e+00 -6.77184999e-01 -5.28052509e-01 3.31200242e-01
5.80215275e-01 -8.70450318e-01 -8.01626980e-01 4.17219996e-01
4.74735916e-01 -9.53270435e-01 1.12782443e+00 3.06301028e-01
7.05701038e-02 -6.71158850e-01 -5.26899099e-01 -1.14408052e+00
-3.63729835e-01 -8.29275846e-02 2.94055581e-01 1.11802280e+00
-1.49064407e-01 -9.08935845e-01 7.64830828e-01 5.05652249e-01
-1.22579709e-01 -4.49108362e-01 -1.21949923e+00 -9.08042610e-01
-3.53880048e-01 -5.57179987e-01 1.23183334e+00 9.90530610e-01
-1.54012814e-01 -2.15955347e-01 -1.10455543e-01 5.16684353e-01
6.74338520e-01 2.96404600e-01 1.02891409e+00 -1.40845513e+00
4.12563086e-01 -4.56918806e-01 -1.26388395e+00 -1.09266186e+00
1.13807023e-01 -9.22930717e-01 1.26839608e-01 -1.56122255e+00
-1.01194412e-01 -9.12893236e-01 -1.04387909e-01 5.35380602e-01
2.31644705e-01 2.70501852e-01 2.97401279e-01 6.28808796e-01
-6.11426592e-01 6.64153278e-01 1.07210171e+00 -5.06119207e-02
-1.39989089e-02 -2.00181857e-01 -2.04805672e-01 6.16476536e-01
8.11480403e-01 -6.82644010e-01 5.09944111e-02 -7.11576223e-01
3.70660014e-02 -1.41840756e-01 6.17872417e-01 -1.34200013e+00
5.94924331e-01 -4.39967692e-01 3.43341857e-01 -8.74068201e-01
6.80037737e-01 -1.04128361e+00 5.73966205e-01 2.75118977e-01
1.35606959e-01 1.59871042e-01 -1.06941111e-01 3.63248229e-01
-5.55140555e-01 -2.51589596e-01 4.43445534e-01 -1.39985219e-01
-9.28928256e-01 6.02985740e-01 2.43262857e-01 -5.98087311e-01
9.93276417e-01 -4.05429602e-01 -2.32459322e-01 -2.22894609e-01
-2.01698393e-01 1.91368744e-01 1.26583064e+00 4.98498708e-01
8.20731163e-01 -1.68948603e+00 -3.68783742e-01 3.34779263e-01
6.06514513e-01 3.97971451e-01 9.02763605e-02 9.01379347e-01
-1.08995080e+00 5.28945267e-01 -2.28366196e-01 -1.07997441e+00
-1.07409382e+00 3.07538748e-01 1.60019085e-01 4.36729729e-01
-7.57975459e-01 5.06703079e-01 1.63986415e-01 -6.48159385e-01
-1.42074853e-01 -4.28152353e-01 2.13337660e-01 -1.31415680e-01
3.80412668e-01 2.32376769e-01 2.33643115e-01 -1.03653026e+00
-8.00219893e-01 1.26777124e+00 2.00781301e-01 6.76909834e-03
1.21518087e+00 -2.04082504e-01 -3.29718143e-01 4.29777712e-01
1.46554196e+00 1.24841714e-02 -9.99959528e-01 -3.75865400e-01
1.57960027e-01 -1.23380661e+00 4.67456430e-02 -1.39485076e-01
-8.96700382e-01 7.74170339e-01 6.09439075e-01 -1.75746441e-01
8.50868762e-01 8.83554816e-02 6.22905314e-01 4.09834504e-01
1.08618271e+00 -7.93618202e-01 -2.88372397e-01 7.70778537e-01
8.50668013e-01 -1.41291618e+00 2.63915896e-01 -6.87671483e-01
-4.23920780e-01 1.22855806e+00 3.31932068e-01 -2.71528423e-01
4.62703824e-01 7.76331648e-02 1.43357307e-01 -2.42654771e-01
-1.31933376e-01 -2.82938361e-01 2.63785481e-01 5.71999550e-01
8.76379088e-02 1.50234863e-01 -8.61360729e-02 -5.05120754e-02
-4.71384734e-01 -1.19140178e-01 2.74355322e-01 1.34826922e+00
-7.31491506e-01 -1.26718569e+00 -8.11204016e-01 1.21637210e-01
1.71863005e-01 1.26317993e-01 -2.03220606e-01 6.85577571e-01
6.04242571e-02 5.79368412e-01 3.74772966e-01 -5.25491893e-01
4.22382265e-01 -1.69385925e-01 3.49646479e-01 -5.78626752e-01
-5.45411482e-02 -6.56831712e-02 -4.06123787e-01 -1.00215387e+00
-4.29121405e-01 -8.07814240e-01 -1.38155639e+00 -5.18036187e-01
-1.44921124e-01 7.78977154e-03 1.23852551e+00 7.89728343e-01
4.75902051e-01 -1.16183877e-01 6.50705099e-01 -1.30123353e+00
-1.84310332e-01 -5.86735070e-01 -5.82859039e-01 5.07765472e-01
3.68849367e-01 -1.05616462e+00 -3.30066949e-01 -3.15278292e-01] | [7.608547687530518, -2.468963861465454] |
1218f17e-3a0a-47c8-9cb5-e1308671167c | how-to-not-train-your-dragon-training-free | 2305.16925 | null | https://arxiv.org/abs/2305.16925v1 | https://arxiv.org/pdf/2305.16925v1.pdf | How To Not Train Your Dragon: Training-free Embodied Object Goal Navigation with Semantic Frontiers | Object goal navigation is an important problem in Embodied AI that involves guiding the agent to navigate to an instance of the object category in an unknown environment -- typically an indoor scene. Unfortunately, current state-of-the-art methods for this problem rely heavily on data-driven approaches, \eg, end-to-end reinforcement learning, imitation learning, and others. Moreover, such methods are typically costly to train and difficult to debug, leading to a lack of transferability and explainability. Inspired by recent successes in combining classical and learning methods, we present a modular and training-free solution, which embraces more classic approaches, to tackle the object goal navigation problem. Our method builds a structured scene representation based on the classic visual simultaneous localization and mapping (V-SLAM) framework. We then inject semantics into geometric-based frontier exploration to reason about promising areas to search for a goal object. Our structured scene representation comprises a 2D occupancy map, semantic point cloud, and spatial scene graph. Our method propagates semantics on the scene graphs based on language priors and scene statistics to introduce semantic knowledge to the geometric frontiers. With injected semantic priors, the agent can reason about the most promising frontier to explore. The proposed pipeline shows strong experimental performance for object goal navigation on the Gibson benchmark dataset, outperforming the previous state-of-the-art. We also perform comprehensive ablation studies to identify the current bottleneck in the object navigation task. | ['Fisher Yu', 'Bernard Ghanem', 'Suryansh Kumar', 'Guohao Li', 'Junting Chen'] | 2023-05-26 | null | null | null | null | ['simultaneous-localization-and-mapping', 'navigate'] | ['computer-vision', 'reasoning'] | [ 2.56362349e-01 1.10089175e-01 4.87168096e-02 -2.53328621e-01
-6.47990763e-01 -5.55696011e-01 7.50434577e-01 3.73921216e-01
-5.07884860e-01 5.18649578e-01 -3.19782533e-02 -2.22412273e-01
-3.88972223e-01 -8.91333699e-01 -1.04104757e+00 -4.61435467e-01
-3.12550396e-01 7.44684041e-01 5.12535691e-01 -4.98560220e-01
6.60838008e-01 3.18851143e-01 -1.67189884e+00 -5.20161167e-02
1.03353190e+00 9.87695873e-01 1.05085897e+00 5.30028462e-01
-4.15214449e-01 8.88405919e-01 -1.28882438e-01 2.08455056e-01
1.83875963e-01 -4.49260384e-01 -9.28765476e-01 -1.29708678e-01
2.19964206e-01 -2.00385321e-02 -1.79585591e-01 1.07049727e+00
2.06696272e-01 5.49594402e-01 4.91774142e-01 -1.37570393e+00
-6.92794442e-01 3.41912985e-01 -2.92279273e-01 2.44583953e-02
6.25793159e-01 4.00642842e-01 9.13550079e-01 -8.32482636e-01
8.28644693e-01 1.29931855e+00 3.79456758e-01 3.92281651e-01
-1.18291414e+00 -1.34713724e-01 7.65575409e-01 3.85186881e-01
-1.22272038e+00 -1.40155062e-01 8.17719400e-01 -3.47337246e-01
1.16648841e+00 -1.49775976e-02 9.59389627e-01 9.50428963e-01
2.13798285e-01 8.20848286e-01 1.16448593e+00 -3.05146307e-01
8.03356588e-01 -1.67094573e-01 -1.96935624e-01 1.07983267e+00
-5.34750009e-03 3.69431287e-01 -8.41552019e-01 1.46644026e-01
8.43387067e-01 -4.01063710e-02 -2.03797698e-01 -1.24915063e+00
-1.28951001e+00 8.35541487e-01 1.18250155e+00 1.59079269e-01
-5.01169145e-01 5.75538218e-01 -2.71182712e-02 -7.12472424e-02
6.28080145e-02 6.83389544e-01 -3.66891176e-01 -1.47508994e-01
-6.26942515e-01 5.98611116e-01 7.60125935e-01 1.25622427e+00
1.05866086e+00 -1.44496247e-01 -5.84543496e-03 3.77011508e-01
6.23873472e-01 3.86211187e-01 1.07411504e-01 -1.12044597e+00
3.58150303e-01 6.56860769e-01 2.27175519e-01 -8.52447093e-01
-5.99646211e-01 -4.00972515e-01 -3.88597578e-01 7.19826519e-01
3.92434269e-01 3.25928897e-01 -1.27112424e+00 1.74571633e+00
4.73852634e-01 1.79654092e-01 9.55147743e-02 1.15718758e+00
7.15035379e-01 3.79932910e-01 1.63957372e-01 3.58244032e-01
1.15786088e+00 -1.34251869e+00 -2.96585768e-01 -7.78671026e-01
6.23711169e-01 -1.78454548e-01 1.24456739e+00 1.92214727e-01
-7.82236874e-01 -5.58520257e-01 -9.13791955e-01 -1.90153807e-01
-6.21265769e-01 -3.02383095e-01 8.45386624e-01 1.92445427e-01
-1.31611228e+00 5.82115710e-01 -9.73972201e-01 -7.32090354e-01
6.31460607e-01 1.65168598e-01 -4.17039394e-01 -2.67741770e-01
-7.89275706e-01 1.06781948e+00 7.78082132e-01 2.65488643e-02
-1.50473118e+00 -5.82511008e-01 -1.24002945e+00 -2.19367564e-01
7.17550099e-01 -1.03458762e+00 1.12428629e+00 -5.51965594e-01
-1.62634850e+00 6.19655132e-01 -1.62855148e-01 -3.96577984e-01
3.21392626e-01 -4.03839320e-01 3.14169288e-01 -2.19941754e-02
4.42808747e-01 1.01377046e+00 5.78881621e-01 -1.62943077e+00
-8.70822251e-01 -4.50337529e-01 3.92888606e-01 6.32942438e-01
4.99331653e-01 -6.52715802e-01 -6.20629251e-01 -1.89044654e-01
6.72656596e-01 -9.86337125e-01 -6.97005451e-01 2.01024264e-02
-3.57298166e-01 -2.43432686e-01 5.24762154e-01 -3.82682592e-01
5.71615815e-01 -1.94427192e+00 6.38274550e-01 8.68509039e-02
1.53752044e-01 -4.70835358e-01 -1.04066320e-01 4.80778128e-01
3.50759298e-01 -1.06341958e-01 -2.79105842e-01 -4.14832741e-01
2.35445395e-01 2.64554530e-01 -3.86625290e-01 4.84485894e-01
-5.44928350e-02 1.16902268e+00 -1.47050810e+00 -1.98346540e-01
5.39105952e-01 3.05048615e-01 -7.95457363e-01 3.39552194e-01
-8.06642532e-01 9.16444778e-01 -7.92161465e-01 6.50315523e-01
3.76703322e-01 -2.29498640e-01 3.35041108e-03 1.24291264e-01
-3.93606961e-01 3.16794693e-01 -1.11131465e+00 2.85345817e+00
-4.99933183e-01 2.33286902e-01 -6.70674145e-02 -9.27840173e-01
7.47220516e-01 -4.28666711e-01 6.76303580e-02 -8.26130509e-01
-6.25212565e-02 3.25032830e-01 -3.00774932e-01 -3.04366380e-01
5.23858488e-01 1.87437624e-01 -2.58285820e-01 -4.54910584e-02
3.06420147e-01 -6.35404825e-01 -1.92882139e-02 2.09199145e-01
1.13427448e+00 1.01129627e+00 3.77530009e-01 -4.28959548e-01
2.81437695e-01 7.00942338e-01 1.92428470e-01 1.21882844e+00
-7.15158358e-02 4.16059524e-01 1.05525963e-01 -3.65307361e-01
-7.07397044e-01 -1.32714570e+00 3.69197994e-01 1.19263756e+00
9.31663096e-01 -3.87973040e-01 -5.21443188e-01 -6.08554780e-01
-1.31243363e-01 9.15010929e-01 -6.50727689e-01 -1.02673829e-01
-4.51594353e-01 -1.55445278e-01 -1.04753278e-01 4.60668504e-01
6.22108340e-01 -1.40248930e+00 -1.30784416e+00 1.78520918e-01
-5.17854467e-02 -9.08463717e-01 1.09622344e-01 4.17788535e-01
-7.20657468e-01 -1.08676076e+00 -3.92872542e-01 -8.19015205e-01
6.51039898e-01 2.36404598e-01 1.18164122e+00 -6.75965548e-02
-2.83375114e-01 6.98530078e-01 -4.15998518e-01 -3.15075129e-01
-1.43334880e-01 1.21763669e-01 -1.06810592e-02 -4.23856765e-01
1.40210558e-02 -5.62218904e-01 -7.86090136e-01 1.80280600e-02
-3.87129635e-01 4.28081840e-01 5.86830914e-01 6.68382168e-01
8.29231918e-01 -1.95398927e-03 2.32354492e-01 -4.51138109e-01
3.21357846e-01 -4.65858161e-01 -8.73716533e-01 1.25530303e-01
-5.68419456e-01 2.87939399e-01 3.81428450e-01 -5.29191606e-02
-8.67713690e-01 1.77066982e-01 3.94524820e-02 -2.94363499e-01
-4.34267193e-01 6.42142475e-01 -1.49703592e-01 -2.15683401e-01
6.01339817e-01 4.15538996e-01 -1.69056207e-01 -3.34801346e-01
9.11922693e-01 -1.95644706e-01 6.69731319e-01 -5.98165095e-01
6.45380735e-01 7.06313610e-01 1.35605350e-01 -4.61697608e-01
-6.89050972e-01 -4.84747499e-01 -6.36654794e-01 -2.15414807e-01
1.03720450e+00 -8.87983680e-01 -8.98882389e-01 1.89451694e-01
-1.06478095e+00 -8.15129220e-01 -3.67208481e-01 5.16383350e-01
-1.21475744e+00 1.55758448e-02 2.10390352e-02 -8.53482187e-01
1.80461958e-01 -1.32372093e+00 1.31807137e+00 3.75212163e-01
-1.12301879e-01 -8.89608979e-01 9.21314284e-02 4.90882993e-02
4.35236245e-01 2.65639722e-01 9.47267413e-01 -4.99903530e-01
-1.28966451e+00 3.02937508e-01 -3.35211575e-01 -6.09453022e-01
-5.00750281e-02 -8.67661655e-01 -7.22007573e-01 -2.98953921e-01
-2.00153127e-01 -4.43949074e-01 9.30950284e-01 3.72555912e-01
9.33600903e-01 1.27102509e-01 -8.23095977e-01 7.62682080e-01
1.60330677e+00 2.15821341e-01 3.60617727e-01 7.74403036e-01
8.00264835e-01 7.84312427e-01 7.88316727e-01 3.07274133e-01
8.28366220e-01 7.13282347e-01 1.15128243e+00 9.07528698e-02
-2.00683419e-02 -6.64102376e-01 1.01055600e-01 1.70175627e-01
9.54180509e-02 -2.21313745e-01 -1.15644050e+00 6.89207613e-01
-2.24186325e+00 -5.72193563e-01 2.30880007e-01 2.05250692e+00
3.18221360e-01 1.89529911e-01 -1.25943005e-01 -4.38376695e-01
1.92368284e-01 1.78959951e-01 -8.32315862e-01 9.66593325e-02
1.35811776e-01 1.15698960e-03 4.29513454e-01 6.61765695e-01
-1.06662476e+00 1.58194625e+00 5.45895147e+00 6.20438635e-01
-7.66400754e-01 1.20969743e-01 3.38516869e-02 3.45676132e-02
-3.52868050e-01 4.41345841e-01 -6.94577038e-01 -4.05118428e-02
3.95166874e-01 7.23234415e-02 9.19806838e-01 1.17568541e+00
-4.24820632e-02 -6.81738913e-01 -1.33958495e+00 1.08142304e+00
5.98174557e-02 -1.48586297e+00 -9.51997936e-02 -8.22580755e-02
4.66175884e-01 3.95812571e-01 2.51135845e-02 6.85375631e-01
6.80257976e-01 -1.14376342e+00 1.17542243e+00 5.92712402e-01
3.55556697e-01 -4.42041039e-01 2.53164262e-01 6.53803468e-01
-1.30084443e+00 -2.05709815e-01 -2.38878205e-01 -1.73940912e-01
3.51851761e-01 -1.35348551e-02 -7.67935336e-01 7.03355730e-01
9.39800680e-01 5.96565783e-01 -4.82038885e-01 1.17941236e+00
-5.22567034e-01 -1.26267612e-01 -4.48457181e-01 -3.10033739e-01
6.38713002e-01 -3.20994437e-01 6.88276947e-01 6.44622564e-01
3.35406750e-01 -8.09850916e-03 7.23759592e-01 1.32903540e+00
3.11624676e-01 -1.25238895e-01 -9.20576692e-01 1.77038327e-01
2.23790243e-01 9.70550537e-01 -1.05191696e+00 -9.05680209e-02
-1.43577188e-01 1.16114032e+00 8.17364812e-01 5.41090667e-01
-7.15795279e-01 -1.44020930e-01 4.03880894e-01 -6.41722903e-02
3.91826302e-01 -5.45290351e-01 -3.09421629e-01 -7.73037851e-01
6.51396886e-02 -5.05404115e-01 1.01423942e-01 -9.62622583e-01
-8.24086130e-01 5.77947378e-01 1.19320296e-01 -8.02023649e-01
-1.95561379e-01 -5.98131955e-01 -3.96290272e-01 9.73110735e-01
-1.77023304e+00 -1.28493476e+00 -6.95858538e-01 5.08254647e-01
7.06074953e-01 -1.37568057e-01 8.73447597e-01 -3.82345140e-01
3.66477631e-02 -1.48861989e-01 -3.08593720e-01 -3.61801267e-01
7.71996528e-02 -1.42089188e+00 5.81987917e-01 6.85727596e-01
3.62581134e-01 5.62394381e-01 7.68172562e-01 -8.81821573e-01
-1.67926216e+00 -9.05790567e-01 3.36031020e-01 -7.34128714e-01
5.44887066e-01 -6.75367832e-01 -7.15449989e-01 6.87655330e-01
6.31140321e-02 1.67075358e-02 1.23386122e-01 2.48432383e-01
-9.79866758e-02 3.62466544e-01 -8.09926867e-01 9.53984499e-01
1.62424421e+00 -4.17375684e-01 -7.75102794e-01 3.71261030e-01
8.66856396e-01 -6.59203053e-01 -2.26629063e-01 4.46744531e-01
3.27453434e-01 -1.11938071e+00 9.45325792e-01 -5.88574648e-01
4.23362590e-02 -7.14265525e-01 -4.41150248e-01 -1.53210902e+00
-4.77984548e-01 -6.05844378e-01 -5.36547527e-02 6.99911237e-01
3.21514517e-01 -4.01035011e-01 9.49581623e-01 3.07053357e-01
-3.84135723e-01 -7.03015864e-01 -9.44994092e-01 -6.73506975e-01
-3.15329552e-01 -6.28229856e-01 5.06293237e-01 4.71594989e-01
-1.47558898e-01 2.59498715e-01 4.18281294e-02 5.59718907e-01
8.16362143e-01 4.85094309e-01 9.50680256e-01 -1.08732176e+00
-3.89822721e-02 -5.96909225e-01 -3.93193156e-01 -1.46965623e+00
3.53971213e-01 -1.09105814e+00 7.42629349e-01 -2.25484395e+00
-1.00111313e-01 -7.38464892e-01 -2.55357921e-01 3.68857890e-01
-1.04448833e-01 -1.80409878e-01 3.28612924e-01 8.35471526e-02
-1.02238142e+00 8.96559894e-01 1.15578139e+00 -8.88104066e-02
-6.22982979e-01 -4.28779095e-01 -4.98133242e-01 8.02429318e-01
6.00751042e-01 -3.94408345e-01 -6.48476958e-01 -5.81145585e-01
3.48579437e-01 4.22680341e-02 8.14506173e-01 -1.22000945e+00
6.06143534e-01 -3.33926678e-01 1.59498990e-01 -7.11018145e-01
5.97071528e-01 -8.86414111e-01 -5.84270060e-02 5.61144769e-01
-1.59230456e-01 -5.43215685e-02 2.61404604e-01 1.01145673e+00
1.40143797e-01 2.40288563e-02 3.99765909e-01 -4.67401773e-01
-1.63650930e+00 3.35626453e-01 -1.62528187e-01 1.54448703e-01
1.08277678e+00 -2.94289351e-01 -2.62408882e-01 -2.13481128e-01
-8.76003504e-01 5.97639143e-01 6.65493667e-01 6.13685608e-01
8.70820343e-01 -1.15026796e+00 -3.31355184e-01 2.73208708e-01
4.03763592e-01 4.25679356e-01 2.40077019e-01 8.04603338e-01
-6.22604430e-01 4.08116102e-01 -2.89196670e-01 -1.01084495e+00
-4.89111513e-01 8.67640555e-01 3.45311105e-01 -4.81852144e-02
-8.13031018e-01 1.14305460e+00 6.85600281e-01 -7.41232216e-01
2.20843762e-01 -4.04075980e-01 -1.51993319e-01 -4.92653757e-01
1.30138487e-01 1.10747226e-01 -2.13213682e-01 -4.72017586e-01
-6.34584725e-01 6.29419148e-01 2.67755032e-01 -3.54088336e-01
1.10362446e+00 -2.99474090e-01 -2.00637616e-02 6.50200844e-01
6.76780581e-01 -3.75052452e-01 -1.48720312e+00 -1.41700670e-01
2.51349539e-01 -4.84423071e-01 9.62059423e-02 -8.23932946e-01
-3.91975135e-01 6.87111139e-01 5.62993526e-01 1.36809051e-02
7.15844452e-01 3.29714864e-01 1.65251523e-01 6.19919181e-01
1.12579024e+00 -8.91791701e-01 2.90556014e-01 7.47889817e-01
1.09539962e+00 -1.48881674e+00 -1.42150119e-01 -3.62905413e-01
-6.23252451e-01 8.50625157e-01 9.34415400e-01 -1.94482699e-01
5.79671979e-01 -1.57099962e-01 -6.55050157e-03 -5.47045827e-01
-5.12508869e-01 -6.13224626e-01 3.51879001e-01 7.89205611e-01
-1.48280710e-01 -2.30281606e-01 3.22270900e-01 3.03213835e-01
-2.95319736e-01 -1.37710437e-01 -8.13873038e-02 1.28315818e+00
-7.02498734e-01 -7.05989838e-01 -1.61968514e-01 1.19380675e-01
2.79721171e-01 -1.34658501e-01 -1.98147863e-01 8.32209945e-01
2.16243818e-01 8.34655046e-01 -5.24801807e-03 -1.32970810e-01
4.01911110e-01 -7.64094219e-02 7.65042126e-01 -8.94526422e-01
-1.96462333e-01 -1.00274704e-01 -2.18955651e-01 -1.18320620e+00
-3.98988694e-01 -4.73312736e-01 -1.62731874e+00 2.54683912e-01
-1.32128790e-01 2.53774505e-02 9.40614998e-01 1.07307732e+00
5.26603222e-01 6.80244207e-01 2.11833660e-02 -1.34611380e+00
-8.14038590e-02 -6.26231313e-01 -3.31300288e-01 2.72845596e-01
3.36498469e-01 -1.14854515e+00 1.12092914e-02 -2.81540751e-01] | [4.600912094116211, 0.5822535157203674] |
613bf636-1204-4f91-81b5-4497ce148fb2 | learning-semantic-representations-of-users | null | null | https://aclanthology.org/P15-1098 | https://aclanthology.org/P15-1098.pdf | Learning Semantic Representations of Users and Products for Document Level Sentiment Classification | null | ['Ting Liu', 'Duyu Tang', 'Bing Qin'] | 2015-07-01 | learning-semantic-representations-of-users-1 | https://aclanthology.org/P15-1098 | https://aclanthology.org/P15-1098.pdf | ijcnlp-2015-7 | ['learning-semantic-representations'] | ['methodology'] | [-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] | [-7.376448154449463, 3.78114914894104] |
e5065152-798a-4428-b6cd-d93ad712bad9 | dual-cross-polarized-gpr-measurement-method | 2205.08142 | null | https://arxiv.org/abs/2205.08142v1 | https://arxiv.org/pdf/2205.08142v1.pdf | Dual-Cross-Polarized GPR Measurement Method for Detection and Orientation Estimation of Shallowly Buried Elongated Object | Detecting a shallowly buried and elongated object and estimating its orientation using a commonly adopted co-polarized GPR system is challenging due to the presence of strong ground clutter that masks the target reflection. A cross-polarized configuration can be used to suppress ground clutter and reveal the object reflection, but it suffers from inconsistent detection capability which significantly varies with different object orientations. To address this issue, we propose a dual-cross-polarized detection (DCPD) method which utilizes two cross-polarized antennas with a special arrangement to detect the object. The signals reflected by the object and collected by the two antennas are combined in a rotationally invariant manner to ensure both effective ground clutter suppression and consistent detection irrespective of the object orientation. In addition, we present a dual-cross-polarized orientation estimation (DCPOE) algorithm to estimate the object orientation from the two cross-polarized data. The proposed DCPOE algorithm is less affected by environmental noise and performs robust and accurate azimuth angle estimation. The effectiveness of the proposed techniques in the detection and orientation estimation and their advantages over the existing method have been demonstrated using experimental data. Comparison results show that the maximum and average errors are 22.3{\deg} and 10.9{\deg} for the Alford rotation algorithm, while those are 4.9{\deg} and 1.8{\deg} for the proposed DCPOE algorithm in the demonstrated shallowly buried object cases. The proposed techniques can be unified in a framework to facilitate the investigation and mapping of shallowly buried and elongated targets. | ['Abdulkadir C. Yucel', 'Mohamed Lokman Mohd Yusof', 'Lai Fern Ow', 'Wenhao Luo', 'Yee Hui Lee', 'Hai-Han Sun'] | 2022-05-17 | null | null | null | null | ['gpr', 'gpr'] | ['computer-vision', 'miscellaneous'] | [ 4.21713769e-01 -1.76043242e-01 6.08622849e-01 7.25900382e-03
-7.09036231e-01 -5.16126275e-01 1.01363584e-01 -1.11827582e-01
-1.81870237e-01 7.79297352e-01 -1.32332876e-01 -2.01578960e-01
-5.99950910e-01 -7.45280743e-01 -3.60096037e-01 -1.41764808e+00
-2.31079742e-01 1.85446367e-01 3.18085462e-01 -1.67616352e-01
2.08111674e-01 7.34623969e-01 -1.63241434e+00 8.48166198e-02
8.58878374e-01 1.20262218e+00 4.97627735e-01 3.32425624e-01
3.28638047e-01 2.41554260e-01 -5.02928019e-01 4.26762700e-02
1.71229631e-01 1.59421898e-02 -6.48972839e-02 4.38807113e-03
8.97021890e-02 -1.98932905e-02 3.31839532e-01 1.10449696e+00
8.45733345e-01 -6.48066401e-02 8.52433801e-01 -7.18035758e-01
-1.62166238e-01 -1.08913511e-01 -1.26140785e+00 3.24368745e-01
3.79015803e-01 -5.16541839e-01 3.79922807e-01 -1.09165776e+00
1.01459008e-02 8.74742627e-01 9.14827526e-01 -5.00824228e-02
-8.21623504e-01 -6.69867933e-01 -2.19691679e-01 9.60162729e-02
-1.41011035e+00 -4.27852899e-01 9.72103179e-01 -4.78874862e-01
4.59868550e-01 5.22104383e-01 5.34653127e-01 5.57945430e-01
1.68450475e-01 1.77562267e-01 1.41189098e+00 -8.34995866e-01
-5.44193084e-04 3.36344428e-02 3.60424429e-01 6.35985613e-01
1.21591294e+00 2.45073467e-01 -1.90121196e-02 -3.15323502e-01
4.41406935e-01 -2.41596669e-01 -8.29918861e-01 -3.64429146e-01
-8.99213195e-01 4.02292877e-01 1.97228000e-01 3.72623742e-01
-8.83926451e-01 -5.79214096e-01 8.45055133e-02 -3.12649041e-01
3.51501971e-01 2.33865082e-01 -1.66580349e-01 2.58317947e-01
-5.55758417e-01 1.69761404e-01 7.51796722e-01 6.83960557e-01
2.29145557e-01 4.75853294e-01 5.99259138e-02 1.09680688e+00
6.73840880e-01 1.32502711e+00 -2.25250632e-01 -1.33772090e-01
4.75529611e-01 -1.10141091e-01 7.26898670e-01 -1.40829313e+00
-5.18607795e-01 -9.41092134e-01 -5.48818529e-01 2.89126765e-02
2.60494858e-01 -4.56041783e-01 -7.35372663e-01 1.29128671e+00
5.87721229e-01 -1.35998145e-01 6.39279842e-01 9.84511375e-01
7.81651378e-01 8.07187557e-01 -3.73681411e-02 -5.15320420e-01
1.64163423e+00 -3.51901025e-01 -8.25627863e-01 -6.39456451e-01
1.99261069e-01 -1.15854478e+00 1.04926765e-01 6.66526496e-01
-6.44124389e-01 -2.45883420e-01 -1.33978283e+00 1.11476028e+00
3.68299991e-01 5.32840669e-01 4.57714468e-01 8.54402483e-01
-3.24353844e-01 -2.23142758e-01 -7.34107971e-01 -2.58898824e-01
2.37554573e-02 -3.07022710e-03 -2.53588945e-01 -3.86851162e-01
-8.45100105e-01 7.68994212e-01 1.46242216e-01 8.22329640e-01
-2.94540644e-01 -3.09618235e-01 -5.95439434e-01 -1.36209011e-01
-1.43286958e-03 -1.04648404e-01 7.64157891e-01 -4.47623938e-01
-9.79459584e-01 5.52644968e-01 -2.91671269e-02 5.94083779e-02
1.24629937e-01 -3.55390042e-01 -9.94740546e-01 4.91475493e-01
4.22175944e-01 -6.58680022e-01 5.53918600e-01 -1.63923943e+00
-7.58049846e-01 -8.87040913e-01 -5.76602578e-01 3.16113353e-01
2.27059215e-01 -4.34182659e-02 1.65505603e-01 -3.33620161e-01
1.24448991e+00 -7.76526332e-01 2.87637338e-02 -5.54148316e-01
-3.65614384e-01 4.27737176e-01 1.06352651e+00 -8.73747289e-01
7.62632310e-01 -2.17012334e+00 -5.69348395e-01 3.43599916e-01
-2.92490959e-01 3.50172669e-01 2.11582422e-01 5.36719918e-01
-1.35130603e-02 -6.86165154e-01 -1.69125676e-01 3.89006138e-01
-4.52018321e-01 -1.73021063e-01 -4.07327898e-02 1.06414878e+00
2.21974384e-02 -6.73414692e-02 -7.15718746e-01 -6.11204542e-02
-1.55278936e-01 6.71803474e-01 -2.51896054e-01 7.15655610e-02
7.48236477e-01 4.20113951e-01 -7.41845846e-01 1.08211601e+00
1.62058973e+00 4.24126983e-01 3.43088180e-01 -7.51534283e-01
-5.58441877e-01 -1.77420601e-01 -1.65189588e+00 2.88842976e-01
-3.04823518e-01 2.03408852e-01 7.63580501e-01 -1.40743506e+00
1.64180183e+00 5.42012691e-01 4.54755276e-01 -8.88452470e-01
-1.08445017e-02 6.59253955e-01 4.25170474e-02 -8.44312549e-01
3.20834547e-01 -6.33064449e-01 2.09249005e-01 -6.25463277e-02
-3.47167194e-01 2.24529698e-01 -1.48121655e-01 -5.20832479e-01
6.55900002e-01 -7.65338466e-02 4.81105685e-01 -6.10738397e-01
6.05375469e-01 4.26959880e-02 8.60090137e-01 6.05105579e-01
1.17095411e-01 3.36391270e-01 -1.65772751e-01 -3.66669983e-01
-4.29611057e-01 -1.06690228e+00 -6.34100080e-01 4.34622407e-01
4.96816218e-01 5.24532497e-01 -2.31136069e-01 -9.69765708e-02
-7.06512257e-02 6.03982329e-01 1.60047431e-02 -1.80417094e-02
-4.62212563e-01 -1.39519131e+00 4.28217024e-01 3.46753389e-01
8.32249284e-01 -7.41429150e-01 -8.89512837e-01 2.74620146e-01
-5.55084646e-01 -1.17130876e+00 4.77061331e-01 3.70611250e-01
-7.54149377e-01 -1.09443951e+00 -6.64203107e-01 -9.45969284e-01
7.66225517e-01 9.22054946e-01 6.79512143e-01 -3.05362850e-01
-1.41075104e-01 5.87103665e-01 -6.38304055e-01 -6.85868084e-01
1.87917843e-01 -7.33911574e-01 1.44397080e-01 3.54701638e-01
4.42658305e-01 -6.02334201e-01 -6.30184233e-01 7.30598986e-01
-5.67658126e-01 -3.31721425e-01 7.21341372e-01 9.13906693e-01
3.89120221e-01 3.46585840e-01 7.11989820e-01 -5.05025566e-01
3.76003236e-01 -5.53692698e-01 -7.06439555e-01 -6.57247305e-02
-8.07318240e-02 -3.24807763e-01 1.26004606e-01 -1.37916178e-01
-1.54475975e+00 -3.02661866e-01 -9.85872597e-02 3.31378996e-01
-2.84619361e-01 7.60217965e-01 -3.54700029e-01 -2.83164382e-01
5.81091106e-01 3.88508081e-01 -3.84890705e-01 -5.70353150e-01
-5.05872428e-01 7.36982167e-01 6.05222762e-01 -6.18122637e-01
7.16146529e-01 5.84935248e-01 3.68760854e-01 -1.39713824e+00
-8.79069746e-01 -7.39082754e-01 1.64675917e-02 -1.98318005e-01
5.84703326e-01 -1.02165747e+00 -5.81816435e-01 6.25463903e-01
-1.09030890e+00 4.17077959e-01 4.19946104e-01 1.19903207e+00
-7.16388971e-02 5.68826318e-01 -4.58555929e-02 -1.59278083e+00
-6.39102519e-01 -6.88806176e-01 8.67678165e-01 2.30926201e-01
1.85699582e-01 -5.94473064e-01 -1.08691238e-01 3.34694266e-01
3.03104401e-01 3.53589326e-01 6.37629092e-01 -2.87099987e-01
-1.48948357e-01 -5.46823025e-01 -3.14890325e-01 1.86132938e-01
3.37551445e-01 -4.16292578e-01 -8.10400784e-01 -4.70876813e-01
4.91432130e-01 1.54713482e-01 3.50430250e-01 6.54649973e-01
2.55137801e-01 -2.30655536e-01 -7.34398842e-01 5.78770161e-01
1.82768714e+00 6.92515016e-01 7.55760193e-01 5.39340079e-01
3.46386522e-01 6.00816548e-01 1.24777091e+00 6.71472013e-01
-6.81362003e-02 4.79592770e-01 5.98285377e-01 -2.17599317e-01
3.74778688e-01 2.32409358e-01 -5.21540940e-02 5.71862221e-01
-5.94387412e-01 -3.51340681e-01 -7.47318864e-01 6.21504784e-01
-1.30911911e+00 -1.16623509e+00 -7.04420388e-01 2.26459026e+00
1.27799764e-01 -1.23205096e-01 -4.68042850e-01 5.65814614e-01
9.16974723e-01 -3.95918898e-02 2.53565252e-01 -1.93132892e-01
-3.57493013e-01 4.06570911e-01 7.77445555e-01 4.10015553e-01
-9.17570353e-01 1.38374642e-01 4.82227755e+00 3.14581186e-01
-1.44535863e+00 -1.50128350e-01 -1.81576043e-01 5.70605218e-01
-1.45387158e-01 -1.17939740e-01 -9.41117585e-01 2.43199095e-01
1.82492927e-01 3.61939788e-01 -4.17516381e-01 7.27097988e-01
1.67192742e-01 -6.53155446e-01 -1.75746262e-01 1.13455248e+00
-2.49409210e-02 -4.50107872e-01 -2.87248611e-01 6.00215932e-03
4.28081989e-01 -4.45357233e-01 -7.97125995e-02 8.81278142e-02
-3.79783541e-01 -3.11783463e-01 6.75702333e-01 4.49958235e-01
4.75450814e-01 -7.37653255e-01 1.00768626e+00 4.00459170e-01
-1.27413630e+00 -5.10652214e-02 -3.57477784e-01 -7.78035447e-02
2.60934025e-01 9.92258847e-01 -9.44909215e-01 9.96415019e-01
8.72849524e-01 3.29048820e-02 1.99698880e-01 1.21012855e+00
-2.06158087e-01 6.56644464e-01 -5.13077736e-01 -3.24607901e-02
1.67756528e-01 -4.62352633e-01 9.78716314e-01 1.19330585e+00
1.01860762e+00 5.17924309e-01 7.14812577e-02 4.43484783e-01
7.30398715e-01 2.72652566e-01 -5.26337087e-01 2.75690049e-01
5.82157075e-01 1.20791399e+00 -5.02147019e-01 6.45532236e-02
-4.20459658e-01 4.23370570e-01 -3.37329656e-01 5.95783114e-01
-8.64213824e-01 -6.06726766e-01 2.41943315e-01 5.38648307e-01
7.78452396e-01 -2.74514973e-01 -6.27297387e-02 -7.39214420e-01
4.25284714e-01 -5.36210001e-01 3.36213261e-01 -7.87472785e-01
-9.86118317e-01 7.63228953e-01 1.56604528e-01 -1.38801289e+00
2.77843416e-01 -7.39228308e-01 -5.01418829e-01 1.06967235e+00
-1.53244436e+00 -9.70479369e-01 -7.11555004e-01 3.69234025e-01
-8.11480209e-02 -1.78370923e-02 7.70700157e-01 4.71038908e-01
-2.66010344e-01 1.46500513e-01 2.85050124e-01 -1.99956059e-01
3.40076447e-01 -5.32086730e-01 -6.48453116e-01 9.59104002e-01
-4.50521082e-01 4.99088407e-01 1.06074786e+00 -6.69920623e-01
-1.51918352e+00 -8.86622965e-01 5.02266765e-01 4.31659460e-01
5.75531647e-02 4.22541387e-02 -6.71221256e-01 3.89455140e-01
-2.24686995e-01 3.83438200e-01 6.26518488e-01 -8.20355192e-02
-1.26269296e-01 -3.54814291e-01 -1.48032916e+00 2.02338591e-01
5.74754894e-01 2.21979290e-01 -5.68802536e-01 6.68205786e-03
-2.12000072e-01 -2.03154221e-01 -5.45428991e-01 1.24270606e+00
8.30857456e-01 -1.04049540e+00 9.07312036e-01 9.60799605e-02
-2.53189474e-01 -5.59176087e-01 -8.18709373e-01 -1.11472881e+00
-6.54741824e-01 -1.31700793e-02 3.77295107e-01 1.00832605e+00
2.03232914e-01 -1.25183463e+00 5.84333599e-01 -3.17473680e-01
-4.16846812e-01 -6.08071208e-01 -6.93272531e-01 -7.23092139e-01
-7.64982224e-01 -6.93023130e-02 8.62179995e-02 8.46693039e-01
-1.34375468e-01 2.69611508e-01 -3.00117016e-01 1.16982353e+00
1.07090974e+00 4.61327761e-01 4.19195682e-01 -1.40285456e+00
-2.14182124e-01 4.01866555e-01 -4.45517838e-01 -9.01029348e-01
-5.68768084e-01 -3.94498229e-01 2.68393219e-01 -1.68081868e+00
-6.35504574e-02 -8.31102669e-01 -3.07477176e-01 -2.21325215e-02
-2.73329415e-03 3.59298646e-01 -4.28959489e-01 -3.16091557e-03
6.76082447e-02 4.06007707e-01 1.06214714e+00 1.71366204e-02
-1.17952332e-01 4.11969244e-01 -6.85237110e-01 8.59515131e-01
7.36839473e-01 -4.80397791e-01 -8.56467560e-02 -4.73241866e-01
-4.16970812e-02 2.73323119e-01 4.90163833e-01 -1.24747503e+00
-9.73943099e-02 -4.99465549e-03 5.53063989e-01 -9.63344157e-01
4.98386204e-01 -9.77620423e-01 3.03888768e-01 5.42081833e-01
8.39643776e-01 -2.42129207e-01 3.85534495e-01 5.20463765e-01
-2.85397887e-01 -2.79261291e-01 1.14470458e+00 -7.75312036e-02
-6.52568042e-01 -3.37333947e-01 -5.07394731e-01 -3.77368480e-01
1.00240088e+00 -6.64711595e-01 -3.75730455e-01 -2.09146574e-01
-7.63166308e-01 -3.00913423e-01 -1.38671875e-01 -2.51641959e-01
6.36319399e-01 -1.26992929e+00 -7.93089151e-01 2.27700159e-01
1.55952334e-01 -2.69435763e-01 6.20114386e-01 1.25914049e+00
-6.05741441e-01 4.82529759e-01 -3.95288438e-01 -7.17326462e-01
-1.38576972e+00 1.43016018e-02 4.27660674e-01 -9.49652307e-03
-2.50480324e-01 6.88587368e-01 2.54206002e-01 -1.39989853e-01
-2.17352122e-01 8.17384198e-02 -4.72959846e-01 -7.99501464e-02
3.96932989e-01 6.41461015e-01 3.78067166e-01 -7.45626390e-01
-5.27753651e-01 8.90529335e-01 3.09129804e-01 4.59394492e-02
1.52691317e+00 -6.18698522e-02 -8.68240893e-02 -9.49162245e-02
8.52653503e-01 7.55914807e-01 -7.71081448e-01 -2.00434834e-01
-2.98766613e-01 -5.26533008e-01 -1.01422807e-02 -7.60598898e-01
-7.82112956e-01 5.52080870e-01 7.56441832e-01 2.38649234e-01
1.32062244e+00 -2.21509233e-01 2.65668273e-01 2.96779156e-01
6.80317223e-01 -6.30716622e-01 -1.91584736e-01 2.69423693e-01
1.01052761e+00 -6.48417115e-01 2.74265498e-01 -1.00991094e+00
-3.18639070e-01 1.05605268e+00 6.39122665e-01 -1.48043349e-01
5.75846553e-01 3.76167923e-01 1.36817962e-01 -5.01968324e-01
2.67913789e-02 -2.24260874e-02 -1.01639153e-02 1.02277696e+00
4.32597429e-01 1.48417532e-01 -7.73370564e-01 8.47024918e-01
7.66595826e-03 -3.54342073e-01 4.41629678e-01 1.38943708e+00
-9.13593471e-01 -4.72034603e-01 -1.23705721e+00 1.50344387e-01
-7.55256653e-01 1.72191098e-01 4.57756549e-01 7.92825460e-01
1.41283810e-01 1.10828805e+00 -1.94111675e-01 8.82160570e-03
8.27770591e-01 -2.05327690e-01 7.46299028e-01 -3.34647179e-01
2.31963992e-01 4.13955241e-01 6.00830078e-01 3.98767516e-02
-5.10918438e-01 -5.61315477e-01 -1.05166209e+00 4.15012360e-01
-7.79674828e-01 7.47675061e-01 7.60057867e-01 6.30840719e-01
5.43275811e-02 3.36249113e-01 6.40425324e-01 -6.96147144e-01
-6.10264599e-01 -1.08355188e+00 -1.26348531e+00 -3.63684334e-02
1.11157052e-01 -1.30223072e+00 -5.54200292e-01 -3.22512090e-01] | [6.803504943847656, 1.320444107055664] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.