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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
ffe90c6e-cb6f-441e-aba7-6ddc593daf9f | zero-shot-language-transfer-vs-iterative-back | 2104.00106 | null | https://arxiv.org/abs/2104.00106v1 | https://arxiv.org/pdf/2104.00106v1.pdf | Zero-Shot Language Transfer vs Iterative Back Translation for Unsupervised Machine Translation | This work focuses on comparing different solutions for machine translation on low resource language pairs, namely, with zero-shot transfer learning and unsupervised machine translation. We discuss how the data size affects the performance of both unsupervised MT and transfer learning. Additionally we also look at how the domain of the data affects the result of unsupervised MT. The code to all the experiments performed in this project are accessible on Github. | ['Har Simrat Singh', 'Chengzhi Huang', 'Aviral Joshi'] | 2021-03-31 | null | null | null | null | ['unsupervised-machine-translation'] | ['natural-language-processing'] | [ 9.39082280e-02 1.63619936e-01 -6.65115654e-01 -3.73807549e-01
-1.37091327e+00 -6.09978080e-01 8.74787569e-01 -1.10152485e-02
-3.85551721e-01 1.19278693e+00 3.57368469e-01 -6.73093617e-01
2.12896228e-01 -4.51473981e-01 -6.70684338e-01 -3.60996485e-01
3.54586095e-01 9.22613025e-01 -2.67763045e-02 -5.32877445e-01
2.01464832e-01 1.74382459e-02 -9.94659066e-01 3.83519351e-01
8.95757377e-01 3.07838500e-01 2.24235371e-01 6.88663661e-01
-2.46778190e-01 3.90281856e-01 -4.10282075e-01 -6.93504035e-01
4.01996970e-01 -6.71980321e-01 -1.16066682e+00 -4.21814710e-01
3.44055861e-01 -7.38498643e-02 -8.77668634e-02 9.87315238e-01
8.93602550e-01 7.30921179e-02 7.70704389e-01 -1.05862534e+00
-6.50039375e-01 7.64578581e-01 -8.84624943e-02 5.17907500e-01
6.16878986e-01 1.38738155e-01 7.39129484e-01 -1.42711985e+00
8.62386525e-01 1.09964967e+00 4.52429891e-01 5.03759086e-01
-1.02406919e+00 -6.99999154e-01 -6.00962341e-01 2.33601451e-01
-1.29292166e+00 -1.03227258e+00 3.63785356e-01 -3.85337889e-01
1.19061077e+00 4.70779650e-02 1.35156810e-01 1.32186413e+00
3.23693901e-01 4.27352160e-01 1.40462101e+00 -1.08969390e+00
-9.65321809e-02 4.78261739e-01 1.26559315e-02 5.99437833e-01
4.69372012e-02 1.34973958e-01 -6.08447254e-01 -3.01897407e-01
5.77344179e-01 -5.28699994e-01 9.16454718e-02 -1.06305838e-01
-1.48034525e+00 8.32759440e-01 -7.97008872e-02 7.21853018e-01
-9.88038536e-03 -7.87722245e-02 5.96414387e-01 9.87061083e-01
8.86936605e-01 6.03829622e-01 -7.70399094e-01 -5.98897338e-01
-7.44208872e-01 -4.46758360e-01 1.07813799e+00 1.32221603e+00
1.15513217e+00 -2.83871770e-01 -2.28476197e-01 1.01652431e+00
-1.37381509e-01 5.58063388e-01 6.19897842e-01 -6.34008467e-01
9.92902458e-01 2.06777692e-01 1.45649329e-01 -2.20391095e-01
-3.88263230e-04 1.30875915e-01 -2.69802958e-01 -5.05748272e-01
4.52852786e-01 -7.38748550e-01 -8.14122677e-01 1.39910996e+00
3.71067710e-02 -1.18460067e-01 2.65049040e-01 7.85293758e-01
8.91745806e-01 6.71880603e-01 -1.33104309e-01 -4.94707584e-01
1.03059089e+00 -1.34395337e+00 -9.44442034e-01 -2.32329518e-01
1.20141387e+00 -1.42580807e+00 1.10061622e+00 -3.32984686e-01
-1.09573174e+00 -3.95158619e-01 -6.43924773e-01 -4.60092843e-01
-5.73354781e-01 8.10971260e-02 3.41137469e-01 5.15579522e-01
-1.14657581e+00 6.50944471e-01 -6.80583239e-01 -1.15316916e+00
-2.21128371e-02 5.63064575e-01 -5.33819735e-01 7.61197787e-03
-1.45167446e+00 1.54657423e+00 2.23813161e-01 -3.60765994e-01
-3.89458209e-01 -3.17184538e-01 -5.74236691e-01 -1.68736279e-01
-1.30849751e-02 -5.86681843e-01 1.46674502e+00 -1.27450073e+00
-1.74716365e+00 1.13543320e+00 -3.53273958e-01 -1.67332634e-01
5.73390543e-01 -2.12799758e-01 -1.67941034e-01 -1.91299647e-01
3.09952050e-01 6.46554947e-01 4.48437512e-01 -7.57427514e-01
-3.73499870e-01 -1.50551513e-01 -3.13284159e-01 3.69902879e-01
-3.36032599e-01 6.50384367e-01 -3.50920081e-01 -5.04811704e-01
-3.86588126e-01 -1.00005329e+00 4.58569005e-02 -5.54281831e-01
6.55868575e-02 -3.05480212e-01 6.64541960e-01 -9.09205437e-01
8.19782317e-01 -2.04121208e+00 4.80004102e-01 -2.70298570e-01
-5.68832099e-01 1.41213492e-01 -5.40030062e-01 1.02840519e+00
-6.63687512e-02 3.54093432e-01 -7.74285942e-02 -2.43708923e-01
-1.95887998e-01 1.12158306e-01 -2.61677295e-01 4.05366272e-01
1.22925945e-01 1.31340730e+00 -1.05675769e+00 -6.87543392e-01
-1.22022584e-01 1.38306782e-01 2.39776983e-03 4.26106840e-01
-5.14233932e-02 6.75308943e-01 -3.12082052e-01 6.68776274e-01
3.18889201e-01 6.06636815e-02 3.27798784e-01 3.68075490e-01
-3.13592374e-01 7.67666638e-01 -1.23538673e-01 1.92303443e+00
-5.43657720e-01 8.93928409e-01 -3.43883932e-01 -7.74408340e-01
1.05474460e+00 8.46717715e-01 3.16185504e-01 -8.18976164e-01
7.23776296e-02 6.66326642e-01 2.00505242e-01 -3.82799178e-01
6.09931409e-01 -2.37311170e-01 1.35357395e-01 9.31421697e-01
3.71332139e-01 -1.92001626e-01 2.56471694e-01 -8.06128234e-02
8.12095523e-01 3.83382559e-01 3.67064595e-01 -5.81336319e-01
1.81735680e-01 4.43523169e-01 3.30065459e-01 4.55158561e-01
-1.40357569e-01 3.46341908e-01 1.85880631e-01 -3.22141021e-01
-1.35629106e+00 -1.09232759e+00 -5.06931916e-02 1.29454195e+00
-8.86913761e-02 -3.27430159e-01 -6.22036397e-01 -7.06508934e-01
-4.63055521e-01 6.48269773e-01 -4.03856844e-01 -1.39389172e-01
-6.03017867e-01 -4.99027938e-01 4.57120180e-01 2.45768487e-01
1.78162381e-01 -1.06790793e+00 -6.08651824e-02 -1.03937171e-01
-8.57445955e-01 -9.92124379e-01 -5.36138117e-01 2.42799535e-01
-1.29107904e+00 -4.53771114e-01 -6.96916640e-01 -1.30957806e+00
4.41845983e-01 3.47427428e-01 1.30425775e+00 4.43978496e-02
1.01586558e-01 3.01345676e-01 -8.08631241e-01 -5.36360800e-01
-6.14131033e-01 9.01244998e-01 2.81852216e-01 -5.64480305e-01
8.75448942e-01 -5.57458580e-01 -1.11987725e-01 2.65648633e-01
-3.73828888e-01 8.74126554e-02 7.28458405e-01 7.16672242e-01
1.93490863e-01 -7.26806223e-01 4.33086812e-01 -8.01486850e-01
8.56115520e-01 -7.01272547e-01 -3.05492043e-01 5.99186718e-01
-6.04399025e-01 1.13446899e-01 3.58687192e-01 -5.02931178e-01
-8.25704396e-01 -1.54119208e-01 2.26926520e-01 -2.98364103e-01
1.61195546e-02 3.04641724e-01 3.50980312e-01 -2.52956189e-02
6.43433928e-01 8.31336752e-02 4.67606150e-02 -5.76776266e-01
3.17509711e-01 1.10507894e+00 -1.06622733e-01 -6.82096362e-01
7.74233043e-01 2.12182906e-02 -4.73353237e-01 -7.19375789e-01
-2.79942125e-01 -5.89917779e-01 -1.09744620e+00 -5.78582026e-02
8.34718823e-01 -9.27213192e-01 3.21514130e-01 4.93040383e-02
-1.17442167e+00 -7.07375169e-01 -1.64185897e-01 8.26226652e-01
-9.60805297e-01 1.82276443e-01 -8.40729892e-01 -4.00306374e-01
-7.01439202e-01 -1.22737861e+00 1.00146627e+00 -1.06143720e-01
-4.83533561e-01 -1.22044241e+00 5.90595603e-01 4.36445624e-01
4.01656091e-01 -4.65955108e-01 1.03327322e+00 -8.63819003e-01
-3.29096287e-01 1.32230714e-01 -8.55091661e-02 -7.15855286e-02
4.65826601e-01 -1.10828817e-01 -6.66228592e-01 -3.46933544e-01
-1.75001875e-01 -6.81567729e-01 3.72198671e-01 1.26704946e-01
-5.18432632e-03 -2.30805546e-01 -4.07980442e-01 2.10458338e-01
1.16212904e+00 -1.29939601e-01 7.77503431e-01 4.71120447e-01
4.96822178e-01 9.95174348e-01 1.09675455e+00 -2.95150504e-02
2.27929428e-01 9.30046082e-01 -5.10653019e-01 -2.60990530e-01
-9.90424305e-02 -1.81855410e-01 7.82543898e-01 1.42514575e+00
-2.32992560e-01 -9.63015482e-02 -1.26464343e+00 5.52572250e-01
-2.02718377e+00 -6.88871980e-01 -1.51255846e-01 2.23325086e+00
1.08682656e+00 -3.88251901e-01 1.94166064e-01 -5.20456612e-01
9.10236776e-01 -2.11333215e-01 5.40329628e-02 -1.17618990e+00
3.45488675e-02 4.96979505e-01 6.11399889e-01 7.45426893e-01
-7.44274855e-01 1.55052209e+00 7.53621054e+00 6.61529422e-01
-1.26379657e+00 6.56175733e-01 4.15800154e-01 -2.04827130e-01
7.63387594e-04 2.88392186e-01 -5.00792027e-01 4.03505981e-01
1.44566131e+00 -6.02072001e-01 4.87680644e-01 3.73290449e-01
4.27855998e-01 4.53057997e-02 -1.32210565e+00 5.45397878e-01
-4.26972732e-02 -1.04068267e+00 2.05024369e-02 2.12769225e-01
9.85249341e-01 7.05315828e-01 -7.39939138e-02 4.99491811e-01
4.89540190e-01 -8.16420197e-01 2.53821373e-01 2.23176524e-01
1.02903724e+00 -5.07134378e-01 7.24065781e-01 5.05687952e-01
-6.32337153e-01 3.09142798e-01 -5.27794003e-01 -3.23407412e-01
-1.37786949e-02 1.99547276e-01 -9.97202218e-01 7.18389630e-01
5.07894218e-01 5.96014440e-01 -4.77251917e-01 7.06000090e-01
-3.44963938e-01 5.69654942e-01 -1.04631938e-01 -1.08783230e-01
1.16948046e-01 -5.43352187e-01 3.95044625e-01 1.38815379e+00
5.26671231e-01 -3.57463732e-02 1.67235911e-01 3.05529594e-01
-1.39077112e-01 6.65402174e-01 -9.77912068e-01 -3.91831189e-01
4.48089540e-01 9.78837371e-01 -3.71629775e-01 -4.46013719e-01
-6.29745781e-01 1.11265707e+00 6.73145235e-01 4.70559001e-01
-5.34331501e-01 -1.61092877e-01 4.62191939e-01 1.16719395e-01
1.01638101e-01 -4.70256716e-01 -3.03236216e-01 -1.40778732e+00
2.08089291e-03 -1.09767723e+00 3.37871730e-01 -6.85599148e-01
-1.27139342e+00 4.77455407e-01 8.77498910e-02 -1.47135115e+00
-6.17643237e-01 -2.82925904e-01 -6.90279663e-01 1.16777492e+00
-1.18116462e+00 -1.00945020e+00 3.45552385e-01 2.79940665e-01
8.38811874e-01 -3.68634105e-01 1.10437894e+00 2.58738875e-01
-4.15104330e-01 7.01733232e-01 3.72854531e-01 1.83308363e-01
1.47432661e+00 -8.73390198e-01 8.07232857e-01 5.92691779e-01
2.97862053e-01 6.06486440e-01 6.39216006e-01 -7.35781908e-01
-1.58589554e+00 -8.78597379e-01 1.56727338e+00 -7.13821232e-01
8.72806966e-01 -4.77916956e-01 -6.11975133e-01 1.00945270e+00
8.39743376e-01 -3.67208868e-01 8.15116704e-01 4.00880277e-01
-2.71688372e-01 3.24918777e-01 -9.39711988e-01 6.41134620e-01
9.33991134e-01 -8.60810161e-01 -7.49402702e-01 8.25519264e-01
7.65313387e-01 -2.45783105e-01 -1.05473089e+00 3.41607183e-01
3.54030460e-01 -3.60086322e-01 4.64561462e-01 -9.15543437e-01
8.38055789e-01 2.69465953e-01 -1.08419508e-01 -1.50943744e+00
-4.69376892e-01 -7.79709756e-01 6.73521236e-02 1.05889225e+00
9.81300473e-01 -8.32462430e-01 3.95574629e-01 3.69868666e-01
1.96477801e-01 -5.60423493e-01 -9.67684388e-01 -1.04094851e+00
7.38886714e-01 2.32774571e-01 1.89157903e-01 1.34230244e+00
4.66200918e-01 9.08620894e-01 -3.60406727e-01 -2.63958126e-01
8.22794586e-02 1.40930191e-01 9.67773736e-01 -7.06108451e-01
-3.62204224e-01 -1.37880802e-01 -1.30393311e-01 -6.67304218e-01
2.98879266e-01 -1.13028669e+00 6.88306540e-02 -1.41361392e+00
3.82338643e-01 -2.01471001e-01 -9.68456268e-02 4.99438733e-01
-3.22454095e-01 4.54958439e-01 1.25320911e-01 7.32947409e-01
-3.29018593e-01 3.91214222e-01 9.89099503e-01 -6.40556365e-02
-3.40178281e-01 -2.82639503e-01 -2.12433740e-01 8.01838785e-02
1.24308848e+00 -7.43388772e-01 -1.02774791e-01 -9.58803654e-01
-4.54859845e-02 1.35039449e-01 -3.83177936e-01 -4.26278532e-01
1.17981173e-01 -2.12461248e-01 -3.44009581e-03 -1.14708520e-01
1.92976549e-01 -4.67735946e-01 -8.27231780e-02 4.74733651e-01
-4.02958065e-01 5.14151871e-01 2.73996353e-01 -2.57693082e-02
-2.58843035e-01 -3.44310462e-01 4.68317509e-01 -3.18806648e-01
-3.94626468e-01 7.16570169e-02 -5.06342351e-01 -3.72367948e-02
8.33909452e-01 -1.37440558e-03 -2.68278867e-01 -5.66149056e-01
-4.76994514e-01 1.18688121e-01 7.32668877e-01 8.36094677e-01
5.55132963e-02 -1.35769558e+00 -1.04266381e+00 -1.98298469e-01
3.29139441e-01 -8.89012158e-01 -6.51652217e-01 1.13738191e+00
-2.47902140e-01 6.55608833e-01 -4.47541028e-01 -4.51673388e-01
-1.41055477e+00 3.69566023e-01 2.24587426e-01 -2.51221895e-01
-2.70605534e-01 4.20011669e-01 -3.21668506e-01 -8.49622071e-01
-2.59117782e-01 4.63004231e-01 1.14542007e-01 -2.23187543e-02
1.06168509e-01 4.39698368e-01 1.55892611e-01 -6.58732116e-01
-3.56944382e-01 4.52275068e-01 -3.11450094e-01 -6.53851509e-01
1.07559025e+00 -2.86722064e-01 -5.73936164e-01 9.61615503e-01
1.29417348e+00 -1.07550636e-01 -3.77706796e-01 -3.83732349e-01
2.49504730e-01 -3.40494573e-01 -1.65911347e-01 -6.82085752e-01
-3.56064022e-01 9.12774622e-01 5.48240185e-01 -3.89369369e-01
7.90861726e-01 2.70379912e-02 7.45623410e-01 6.85585797e-01
6.26160145e-01 -1.33557379e+00 -1.93453550e-01 9.93151724e-01
6.55373156e-01 -1.57302272e+00 2.48959810e-02 -4.75697666e-01
-3.80651504e-01 1.18297863e+00 4.33747441e-01 1.41739041e-01
2.47488618e-01 1.91923752e-01 5.37603319e-01 1.17955461e-01
-9.71389055e-01 -2.83046484e-01 1.29084840e-01 5.74999213e-01
1.02874935e+00 5.92998937e-02 -9.48867619e-01 -2.02916577e-01
-4.18650061e-01 1.67829007e-01 3.44677538e-01 1.07406175e+00
-2.53965765e-01 -1.70565224e+00 -2.58168101e-01 1.99742690e-01
-5.10244787e-01 -4.81237203e-01 -1.27823210e+00 6.26386702e-01
-2.62575418e-01 1.15668869e+00 4.90549393e-02 -4.56552386e-01
-1.59450490e-02 4.52199578e-01 7.73335457e-01 -1.16163099e+00
-7.65227079e-01 -1.32983029e-01 5.75789750e-01 -1.20731905e-01
-3.04031789e-01 -5.22638023e-01 -8.84586394e-01 -4.19811755e-01
-5.37203014e-01 5.37376106e-01 7.99805522e-01 1.11268568e+00
4.59105700e-01 -3.06992441e-01 6.56159639e-01 -6.07976556e-01
-4.94897038e-01 -1.55408192e+00 2.14911953e-01 1.92251831e-01
-2.33859736e-02 -2.78101712e-01 -2.12820128e-01 1.28361925e-01] | [11.505293846130371, 10.27646541595459] |
d4e8abdb-6987-42e9-b267-c1819761ad6f | simoap-improve-coherence-and-consistency-in | 2305.11130 | null | https://arxiv.org/abs/2305.11130v2 | https://arxiv.org/pdf/2305.11130v2.pdf | SimOAP: Improve Coherence and Consistency in Persona-based Dialogue Generation via Over-sampling and Post-evaluation | Language models trained on large-scale corpora can generate remarkably fluent results in open-domain dialogue. However, for the persona-based dialogue generation task, consistency and coherence are also key factors, which are great challenges for language models. Existing works mainly focus on valuable data filtering, model structure modifying, or objective function designing, while their improvements are limited and hard to generalize to all types of pre-trained language models. However, we find that language models can produce consistent and coherent responses if we consider enough generations. Thus, the problems lay in large-scale response generation and target response selection. In this work, a simple but effective two-stage SimOAP strategy is proposed, i.e., over-sampling and post-evaluation. The over-sampling stage takes large-scale responses from existing trained models efficiently via off-the-shelf distilling and compressing methods, and the post-evaluation stage selects a good response based on multiple well-designed evaluation metrics from large-scale candidates. Experimental results show that the proposed plug-in SimOAP strategy improves the backbone models and outperforms the baseline strategies in both automatic and human evaluations. | ['Xueqi Cheng', 'HuaWei Shen', 'Liang Pang', 'Junkai Zhou'] | 2023-05-18 | null | null | null | null | ['dialogue-generation', 'response-generation', 'dialogue-generation'] | ['natural-language-processing', 'natural-language-processing', 'speech'] | [ 2.28652898e-02 6.55327961e-02 -1.07820690e-01 -6.49929881e-01
-1.28549457e+00 -5.04883647e-01 7.10436165e-01 6.69323504e-02
-6.63037598e-01 1.11523914e+00 4.44726914e-01 4.19671796e-02
2.51198918e-01 -7.25231051e-01 -1.27115414e-01 -3.65262717e-01
3.91586363e-01 1.18243814e+00 2.32452422e-01 -7.63065994e-01
2.22566575e-01 -1.59959033e-01 -1.03710473e+00 5.10559261e-01
1.22961378e+00 7.47412920e-01 2.64105976e-01 8.50812018e-01
-4.12950486e-01 7.18115211e-01 -1.03508472e+00 -6.58151984e-01
3.94577254e-03 -8.43562186e-01 -1.17151248e+00 -5.00695407e-02
-1.00011185e-01 -4.89893794e-01 -3.60436998e-02 7.07056344e-01
1.13263607e+00 4.49986428e-01 3.69008631e-01 -7.55047381e-01
-5.56924045e-01 9.41669166e-01 -2.24310681e-02 -2.01652661e-01
7.04924822e-01 3.72625023e-01 9.32285130e-01 -9.71674442e-01
5.14064372e-01 1.74548709e+00 3.32520366e-01 1.09678638e+00
-1.17208362e+00 -4.19163883e-01 6.20366149e-02 -3.05281848e-01
-1.02210462e+00 -6.22078598e-01 6.06695175e-01 -2.11342424e-01
8.24790776e-01 4.14873302e-01 3.22877437e-01 1.22427404e+00
-2.21938297e-01 9.68659043e-01 1.16391408e+00 -4.13457394e-01
1.20727286e-01 5.50285161e-01 1.56341285e-01 3.75879794e-01
-3.24459612e-01 -2.66856045e-01 -5.34738541e-01 -5.41558266e-01
5.58236659e-01 -2.91363686e-01 -3.29358935e-01 2.89463729e-01
-1.31178665e+00 1.15301108e+00 3.46923582e-02 1.21912822e-01
-4.52237219e-01 -3.08011144e-01 5.12355745e-01 5.66479981e-01
7.19216526e-01 9.17009771e-01 -5.75859368e-01 -3.75980228e-01
-8.42538118e-01 7.95382261e-01 1.09630907e+00 8.58345151e-01
6.74091220e-01 -6.50569946e-02 -7.75957167e-01 1.40146196e+00
1.54927090e-01 4.85617131e-01 8.20519805e-01 -6.93425655e-01
7.88961053e-01 6.66017950e-01 3.91077667e-01 -8.27521861e-01
-2.39522904e-01 -2.30726033e-01 -9.12149012e-01 -5.79570174e-01
3.81486118e-01 -7.02527046e-01 -4.34542209e-01 1.84195042e+00
5.37991166e-01 -4.81200069e-01 2.48020560e-01 1.01621437e+00
1.18904543e+00 9.23432946e-01 4.82985824e-02 -6.12381876e-01
1.11588597e+00 -1.35634971e+00 -8.39684784e-01 -2.76439428e-01
6.66864872e-01 -1.00768423e+00 1.41509342e+00 1.62663966e-01
-1.56213760e+00 -7.08975136e-01 -5.47016025e-01 -1.08177863e-01
9.59611759e-02 3.22031140e-01 6.40444398e-01 3.17419648e-01
-7.72042930e-01 3.56947601e-01 -2.02652365e-01 -1.79406643e-01
-2.16073513e-01 3.68190199e-01 -5.49471900e-02 6.38279691e-02
-1.73137069e+00 9.03145671e-01 3.39548022e-01 9.53278169e-02
-7.75981963e-01 -2.63232917e-01 -6.39221489e-01 -1.45253494e-01
4.55149800e-01 -8.68979037e-01 1.75932300e+00 -8.79591644e-01
-2.04425502e+00 6.29548788e-01 -2.40017354e-01 -3.33335906e-01
7.23400414e-01 -2.56459534e-01 -2.89193153e-01 -7.69729018e-02
-9.02918652e-02 7.42345035e-01 4.88920391e-01 -9.51497376e-01
-5.91298401e-01 -1.65436473e-02 2.32499376e-01 5.79507887e-01
-5.82079887e-01 3.51217210e-01 -4.71995562e-01 -5.35678685e-01
-2.73975283e-01 -8.08603227e-01 -5.81353545e-01 -6.47309899e-01
-3.83673638e-01 -7.15180099e-01 1.33903325e-01 -7.56711364e-01
1.71661007e+00 -1.57535172e+00 1.26364514e-01 -2.82498840e-02
8.93931016e-02 4.84947801e-01 -4.07821864e-01 6.92015707e-01
4.78664398e-01 5.29888272e-02 1.30372301e-01 -4.11686093e-01
1.45762086e-01 -1.77956596e-01 -3.72024149e-01 -2.25719035e-01
1.60096094e-01 8.30257058e-01 -1.07219470e+00 -6.84936523e-01
-2.25139081e-01 -5.74296974e-02 -6.93577170e-01 8.77308428e-01
-5.80327928e-01 4.08492208e-01 -6.53798044e-01 2.46079087e-01
3.84986222e-01 -2.19888195e-01 9.41996649e-02 8.05975497e-02
1.36921465e-01 8.89921904e-01 -1.12924385e+00 1.84040868e+00
-7.06807911e-01 9.80699342e-03 1.47334114e-02 -5.11826158e-01
1.31310093e+00 4.12352473e-01 3.48990709e-02 -5.94700754e-01
1.25965476e-01 4.81907994e-01 2.21415758e-01 -4.76425201e-01
1.00579917e+00 -7.49403313e-02 -4.75239068e-01 8.78727973e-01
1.00585103e-01 -3.46639782e-01 6.13969564e-01 4.06736463e-01
8.72676730e-01 -1.44119442e-01 3.08393806e-01 -5.26023768e-02
8.22967649e-01 1.34478867e-01 6.29464686e-01 7.61844575e-01
5.34303114e-02 5.22384465e-01 4.34177428e-01 -2.80458748e-01
-8.43840182e-01 -4.53399271e-01 3.33600849e-01 1.43780923e+00
-1.10985227e-01 -4.83413547e-01 -9.72280085e-01 -7.39232004e-01
-3.82766128e-01 5.58754504e-01 -2.05380157e-01 1.73235796e-02
-7.30716884e-01 -7.46252060e-01 5.62927365e-01 2.34298676e-01
5.53579569e-01 -1.34073472e+00 -1.34606630e-01 6.15665913e-01
-6.86436057e-01 -8.11995268e-01 -8.04914176e-01 -2.08554104e-01
-6.12827241e-01 -5.71851671e-01 -9.12170529e-01 -7.72236884e-01
6.38906121e-01 1.21393606e-01 1.37040281e+00 2.61512160e-01
2.64581591e-01 -2.44305506e-01 -4.55214858e-01 -2.02351689e-01
-8.27353358e-01 4.28160489e-01 -8.31123255e-03 -4.51441258e-02
3.09528202e-01 -5.96209727e-02 -5.94542384e-01 5.68702757e-01
-5.72635233e-01 3.50302875e-01 6.07347906e-01 1.29383755e+00
3.50271314e-01 -3.75399858e-01 1.04780078e+00 -1.14336729e+00
1.62085068e+00 -2.40892783e-01 -1.91564605e-01 5.63014209e-01
-6.83851838e-01 1.84490174e-01 9.93789077e-01 -6.68911815e-01
-1.33253026e+00 -2.96500117e-01 -2.50382155e-01 9.12455693e-02
2.06866600e-02 6.21117294e-01 -1.55332252e-01 3.90603155e-01
7.57109284e-01 2.59644300e-01 6.41553551e-02 -3.70202601e-01
3.23144644e-01 1.10767984e+00 1.96531579e-01 -8.44796002e-01
4.82053190e-01 -2.95458585e-01 -7.62285948e-01 -4.79776561e-01
-8.21703553e-01 -4.19843704e-01 -2.56754249e-01 -6.95379078e-02
4.35369253e-01 -8.56426418e-01 -4.33902770e-01 4.73355532e-01
-1.43286312e+00 -4.75376457e-01 -1.58953220e-01 3.70008558e-01
-3.57659131e-01 1.52912602e-01 -8.63225460e-01 -9.37516928e-01
-8.67006421e-01 -1.12153292e+00 9.07961249e-01 4.51243222e-01
-6.79157615e-01 -1.01832318e+00 4.08658206e-01 6.70890272e-01
5.45901418e-01 -2.90766180e-01 6.85246468e-01 -1.18548167e+00
-2.54600525e-01 -3.63715678e-01 1.18361823e-01 3.32031190e-01
-1.26501054e-01 -2.46558353e-01 -7.89825559e-01 -3.16841424e-01
-7.46647536e-04 -1.12695742e+00 5.38230717e-01 -1.07245566e-02
8.32902908e-01 -6.86499298e-01 3.33549306e-02 1.54367730e-01
6.26874745e-01 1.11533888e-01 4.03264940e-01 -5.22323847e-02
3.69255781e-01 9.39658642e-01 1.04476488e+00 6.90726936e-01
5.39464951e-01 7.84429252e-01 -3.16024840e-01 -2.02058300e-01
1.21925429e-01 -5.21988690e-01 5.68178833e-01 1.30622220e+00
1.58323765e-01 -4.97154683e-01 -6.24932528e-01 4.68915820e-01
-2.04511523e+00 -8.84356976e-01 2.86751240e-02 2.24346733e+00
1.65952051e+00 4.48530428e-02 3.15424770e-01 -3.78005594e-01
8.01301718e-01 1.79842651e-01 -3.73660237e-01 -6.12016201e-01
-1.11664146e-01 1.90411136e-01 -2.42093861e-01 6.88507318e-01
-6.74482822e-01 1.18878162e+00 5.83272409e+00 1.18477321e+00
-9.55332816e-01 2.70411789e-01 8.11795294e-01 -1.82115197e-01
-5.89373350e-01 9.69669297e-02 -1.11322773e+00 4.48530853e-01
1.01002538e+00 -4.38497484e-01 3.61220479e-01 8.37733984e-01
5.05064249e-01 2.04825208e-01 -1.00393271e+00 6.90678239e-01
1.17782235e-01 -1.08934033e+00 1.47022054e-01 -2.73731321e-01
7.96344697e-01 -4.86706972e-01 -3.91953498e-01 8.19886684e-01
7.00340867e-01 -9.28682625e-01 4.56719607e-01 2.50711441e-01
6.12153411e-01 -7.36584783e-01 8.94799888e-01 7.66137660e-01
-9.12357450e-01 2.01675355e-01 -5.80542445e-01 -3.12135573e-02
4.16342258e-01 3.50174606e-01 -1.05842042e+00 3.66061866e-01
1.09552750e-02 1.96741611e-01 -4.35586005e-01 6.39115036e-01
-4.20957774e-01 5.76849639e-01 -1.14234611e-01 -6.62971973e-01
3.23568881e-01 -1.14500374e-01 2.34211728e-01 1.28946769e+00
1.22295663e-01 1.50582194e-01 6.30444109e-01 7.97455430e-01
-1.43633276e-01 3.99508893e-01 -1.01381190e-01 -1.22003995e-01
7.12223947e-01 1.49345756e+00 -6.57972470e-02 -4.67937440e-01
-1.10298507e-01 8.05490375e-01 4.31126982e-01 2.15641513e-01
-7.18974829e-01 -4.99719501e-01 1.55586585e-01 9.12806168e-02
-3.85628819e-01 1.97252035e-01 -8.56179968e-02 -1.24611723e+00
-2.96250861e-02 -1.50943136e+00 5.00407040e-01 -3.85616869e-01
-1.30117273e+00 1.00447369e+00 -1.42111674e-01 -1.15788972e+00
-9.81191039e-01 -1.20456629e-01 -6.40917718e-01 1.19163418e+00
-1.21189082e+00 -1.11973536e+00 -1.93827674e-01 6.47805870e-01
1.04099905e+00 -3.37265968e-01 1.08907926e+00 2.10190564e-01
-6.71399117e-01 9.19457853e-01 -1.15900286e-01 4.90172878e-02
9.99022126e-01 -1.14159918e+00 4.10172820e-01 6.63232803e-01
-2.54823357e-01 9.38644707e-01 6.02419853e-01 -6.36647701e-01
-1.07944036e+00 -8.45912397e-01 1.38705575e+00 -1.56809166e-01
4.45100367e-01 -4.66688305e-01 -8.60649943e-01 7.09337145e-02
4.30886567e-01 -4.58414704e-01 6.78790867e-01 2.11459592e-01
1.94507927e-01 4.60157804e-02 -9.33038771e-01 7.04978287e-01
7.28560209e-01 -2.43564382e-01 -5.98084569e-01 6.24169827e-01
7.21402287e-01 -6.85346961e-01 -7.13506341e-01 2.51976788e-01
3.41641128e-01 -5.99792302e-01 7.15800107e-01 -8.51780832e-01
6.22210383e-01 7.03032240e-02 2.39190385e-01 -1.62476373e+00
-1.14243507e-01 -1.42399430e+00 8.60103220e-02 1.61616433e+00
7.68476725e-01 -5.04130423e-01 5.25641203e-01 9.89085495e-01
-1.19586311e-01 -8.60553086e-01 -3.41428965e-01 -3.94338667e-01
1.53699089e-02 7.35047013e-02 6.92227662e-01 7.28634834e-01
2.28830367e-01 9.91235852e-01 -7.89987326e-01 -5.32081664e-01
1.28581509e-01 3.42598051e-01 1.33958662e+00 -9.02283788e-01
-6.79011941e-01 -3.45643461e-01 5.33071160e-01 -1.73541212e+00
1.42627493e-01 -4.85531479e-01 3.72330010e-01 -1.48003268e+00
2.46550635e-01 -6.05243206e-01 1.40329540e-01 1.60566226e-01
-7.24792480e-01 -3.14980149e-01 9.10681561e-02 3.57589513e-01
-8.48641932e-01 8.61634076e-01 1.56464255e+00 -3.44926375e-03
-4.63587642e-01 3.76864761e-01 -7.69413233e-01 4.96777534e-01
7.46872604e-01 -4.12726521e-01 -5.55737376e-01 -4.60109144e-01
1.03846781e-01 6.14435375e-01 -1.79563314e-01 -4.68441248e-01
3.31208646e-01 -5.59920371e-01 -3.65236588e-02 -4.44664061e-01
3.85734499e-01 -2.48417556e-02 -2.98227876e-01 3.14301133e-01
-1.04128206e+00 3.26769322e-01 -1.88499063e-01 2.85189062e-01
-4.89031434e-01 -5.05425155e-01 7.00055599e-01 -5.00031888e-01
-3.64547133e-01 1.90631762e-01 -2.08728984e-01 5.97307920e-01
5.65883577e-01 1.19497828e-01 -4.12889093e-01 -7.54775107e-01
-3.06628525e-01 5.98000169e-01 1.17035404e-01 5.13106465e-01
5.91918111e-01 -1.34393001e+00 -1.18727124e+00 -4.47126403e-02
8.45918506e-02 2.70407379e-01 2.70829022e-01 5.81572294e-01
-2.86847502e-01 5.33485591e-01 1.76692158e-01 -2.49094516e-01
-1.30195498e+00 -4.36290987e-02 1.38314694e-01 -1.08431602e+00
9.19437706e-02 1.19676936e+00 5.41922487e-02 -8.62214804e-01
3.03035885e-01 4.82885577e-02 -3.42889160e-01 3.12400889e-02
7.05147684e-01 1.51941568e-01 -7.68575445e-02 -4.57220167e-01
-2.17214040e-02 5.82509069e-03 -4.38284844e-01 -4.14478600e-01
9.71180320e-01 -1.38120636e-01 -1.71925932e-01 3.26678783e-01
7.43895948e-01 7.72476047e-02 -8.53925526e-01 -5.91635704e-01
-2.51386315e-01 -4.26376969e-01 -2.77032912e-01 -9.82454956e-01
-6.40238881e-01 9.21932101e-01 -7.82404989e-02 3.25119108e-01
9.13682282e-01 -3.10731024e-01 1.14849055e+00 7.04680443e-01
4.18812364e-01 -1.45925498e+00 4.62075323e-01 8.37150335e-01
1.07158136e+00 -1.26341689e+00 -2.02656999e-01 -2.51970530e-01
-1.17952776e+00 9.29338038e-01 1.16622877e+00 1.98715001e-01
-3.15058418e-02 -2.83953641e-02 2.33773157e-01 1.87511846e-01
-1.40408707e+00 5.00070043e-02 3.13709915e-01 2.39026904e-01
8.88971984e-01 1.11651763e-01 -7.82822788e-01 1.03851068e+00
-3.80405843e-01 -1.34205744e-01 3.14108878e-01 6.32869244e-01
-4.66704041e-01 -1.54190421e+00 -1.62312463e-01 3.91647875e-01
-4.14956182e-01 -2.05723286e-01 -8.44726980e-01 4.69213694e-01
-3.05913687e-01 1.29824328e+00 -3.73656690e-01 -6.70067251e-01
4.26497102e-01 9.14620459e-02 1.11263566e-01 -9.27289963e-01
-1.25132167e+00 2.56468594e-01 7.39703715e-01 -2.32550919e-01
-1.96303740e-01 -4.12547499e-01 -1.07469916e+00 -4.47089523e-01
-8.46939385e-01 7.29354262e-01 3.35207611e-01 8.38946998e-01
3.23527604e-01 2.22417623e-01 1.02175605e+00 -4.58751589e-01
-1.39383984e+00 -1.55378246e+00 -1.48974612e-01 6.08399153e-01
-2.32063144e-01 -2.58997709e-01 -2.42683634e-01 -1.64557397e-01] | [12.576828002929688, 8.248071670532227] |
61f2b6d7-b815-4f9f-9242-8cb705167cf5 | decoding-kinetic-features-of-hand-motor | null | null | https://onlinelibrary.wiley.com/doi/abs/10.1111/ejn.14936 | https://onlinelibrary.wiley.com/doi/epdf/10.1111/ejn.14936 | Decoding kinetic features of hand motor preparation from single‐trial EEG using convolutional neural networks | Building accurate movement decoding models from brain signals is crucial for many biomedical applications. Predicting specific movement features, such as speed and force, before movement execution may provide additional useful information at the expense of increasing the complexity of the decoding problem. Recent attempts to predict movement speed and force from the electroencephalogram (EEG) achieved classification accuracies at or slightly above chance levels, highlighting the need for more accurate prediction strategies. Thus, the aims of this study were to accurately predict hand movement speed and force from single‐trial EEG signals and to decode neurophysiological information of motor preparation from the prediction strategies. To these ends, a decoding model based on convolutional neural networks (ConvNets) was implemented and compared against other state‐of‐the‐art prediction strategies, such as support vector machines and decision trees. ConvNets outperformed the other prediction strategies, achieving an overall accuracy of 84% in the classification of two different levels of speed and force (four‐class classification) from pre‐movement single‐trial EEG (100 ms and up to 1,600 ms prior to movement execution). Furthermore, an analysis of the ConvNet architectures suggests that the network performs a complex spatiotemporal integration of EEG data to optimize classification accuracy. These results show that movement speed and force can be accurately predicted from single‐trial EEG, and that the prediction strategies may provide useful neurophysiological information about motor preparation. | ['José Biurrun Manresa', 'Mads Jochumsen', 'Luciano Schiaffino', 'Yanina Atum', 'Ramiro Gatti'] | 2020-08-11 | null | null | null | null | ['eeg-decoding', 'eeg-decoding'] | ['medical', 'time-series'] | [ 3.13148797e-01 -3.14053297e-01 -2.64206320e-01 -2.44664162e-01
-3.34780753e-01 -1.27864540e-01 4.56240803e-01 5.53614274e-02
-6.88783526e-01 8.03165913e-01 2.02198237e-01 -4.17531043e-01
-5.28141737e-01 -4.58304077e-01 -5.06910205e-01 -4.94979143e-01
-4.49984938e-01 2.49918476e-01 8.44079703e-02 -3.28909636e-01
6.24045610e-01 6.99649990e-01 -1.66615999e+00 6.31146252e-01
7.09671736e-01 1.19672453e+00 6.20518446e-01 6.76708221e-01
3.78287971e-01 7.54882276e-01 -6.50720716e-01 3.05599660e-01
-1.85832545e-01 -4.94189769e-01 -7.35979557e-01 -5.87398231e-01
-1.32231489e-01 -3.53624761e-01 -4.78607357e-01 4.91668820e-01
6.11631632e-01 1.69501990e-01 7.47108102e-01 -1.07043302e+00
-3.73483807e-01 7.24739611e-01 9.76101309e-03 7.71380663e-01
4.74110067e-01 3.34444553e-01 6.46412313e-01 -4.45949823e-01
4.04500514e-01 3.69987488e-01 7.18809187e-01 4.61007446e-01
-1.07919657e+00 -9.93797779e-01 2.96235438e-02 9.49041605e-01
-1.23302579e+00 -2.84572870e-01 7.57802308e-01 -6.91082060e-01
1.64315057e+00 3.32490623e-01 1.30397308e+00 1.46082020e+00
1.10987377e+00 8.04977953e-01 1.08743978e+00 -2.16112584e-01
1.46316245e-01 -3.32349807e-01 2.75238007e-01 9.73469112e-04
1.01834334e-01 2.38555938e-01 -9.76792991e-01 3.60624135e-01
5.83198607e-01 -9.51072350e-02 -5.61865330e-01 3.82867813e-01
-1.33154726e+00 3.44801664e-01 4.87019181e-01 8.04688692e-01
-8.52724493e-01 1.97326213e-01 3.59866291e-01 2.59952158e-01
3.03105056e-01 7.25325882e-01 -6.77417219e-01 -8.79137754e-01
-1.37197840e+00 2.36863643e-01 3.83769125e-01 4.78536010e-01
2.66889542e-01 1.70274839e-01 -2.30479956e-01 4.90573257e-01
7.67834932e-02 2.68881232e-01 1.02574301e+00 -3.94908398e-01
6.63341403e-01 6.37147546e-01 -1.18389599e-01 -9.14707720e-01
-1.18079269e+00 -4.08616006e-01 -6.83017373e-01 2.16999650e-01
5.13987958e-01 -2.35152394e-02 -7.43523657e-01 1.41574502e+00
-4.97459054e-01 -1.55707709e-02 -3.89571756e-01 9.84363973e-01
4.83552843e-01 2.36455590e-01 3.54700983e-01 -4.50837649e-02
1.12704551e+00 -3.40868026e-01 -8.17994237e-01 -6.48553073e-01
8.52586091e-01 -2.23805949e-01 8.52033496e-01 7.40953684e-01
-8.82594645e-01 -7.32470155e-01 -1.25830150e+00 3.76044035e-01
-2.41513550e-01 5.01076579e-01 8.25350761e-01 4.87276256e-01
-7.17960656e-01 1.21100783e+00 -1.58089054e+00 -2.59145778e-02
5.09460986e-01 9.80685771e-01 -5.75673699e-01 3.83695751e-01
-1.19616294e+00 1.53384376e+00 4.12678957e-01 6.32006943e-01
-3.73444259e-01 -5.90856194e-01 -6.30529404e-01 2.24098247e-02
-4.64818031e-01 -1.92879662e-01 6.59929097e-01 -6.99150026e-01
-1.58229220e+00 4.84936059e-01 -1.77225560e-01 -3.45316321e-01
2.08097473e-01 -3.57451946e-01 -5.55787265e-01 -3.75710651e-02
-1.52154535e-01 4.50622171e-01 4.78798091e-01 -1.99378565e-01
-3.82502347e-01 -6.37263775e-01 -4.60145622e-01 3.53104770e-02
-2.03209117e-01 -5.01014553e-02 3.44858289e-01 -6.16848290e-01
1.77651361e-01 -1.08730042e+00 3.07954818e-01 -5.56173384e-01
-1.70312658e-01 -2.81269699e-01 2.61954308e-01 -1.09774089e+00
1.33580267e+00 -1.86180425e+00 4.43360388e-01 2.63960451e-01
1.74525514e-01 6.54581413e-02 -1.34810433e-01 3.79943520e-01
-4.61068362e-01 -2.58242905e-01 1.49699226e-02 3.38284433e-01
-1.15227394e-01 -1.47839531e-01 6.89915866e-02 6.85669303e-01
1.13247380e-01 1.23880184e+00 -6.05417967e-01 2.80192792e-01
5.42296231e-01 4.10172611e-01 -3.66049528e-01 1.17846794e-01
2.59003580e-01 7.57661819e-01 1.21275768e-01 3.75531465e-01
2.68825740e-02 2.74026431e-02 2.00451821e-01 -3.34888339e-01
-3.88844684e-02 7.08771765e-01 -6.74868464e-01 1.68458128e+00
-4.10942048e-01 1.44601583e+00 -4.70383674e-01 -1.28997231e+00
8.45409870e-01 4.58983213e-01 8.99094462e-01 -1.21600366e+00
7.51960814e-01 1.82136253e-01 7.89204955e-01 -6.95402503e-01
-5.79461604e-02 3.28697264e-02 7.60367662e-02 4.16576773e-01
4.21319045e-02 9.91518572e-02 8.74476135e-02 -6.50362492e-01
1.29238832e+00 3.21814626e-01 1.08548678e-01 -1.99560523e-01
1.93793893e-01 1.51603758e-01 2.25183070e-01 4.84046906e-01
-1.59259677e-01 2.11881623e-01 1.81568101e-01 -5.79813957e-01
-6.04657471e-01 -8.17195952e-01 -1.46563143e-01 9.74466681e-01
5.45495115e-02 -4.61375743e-01 -7.06473589e-01 -1.99234411e-02
-1.95524827e-01 8.10545206e-01 -8.47233951e-01 -5.58487117e-01
-8.39938819e-01 -8.76549602e-01 3.71282816e-01 1.15308559e+00
2.27177307e-01 -1.44584262e+00 -1.15253258e+00 6.13068402e-01
-4.50830042e-01 -9.92989540e-01 2.24200070e-01 6.99174285e-01
-8.73851955e-01 -1.37133574e+00 -3.86789858e-01 -5.78033268e-01
2.53575951e-01 -3.70524764e-01 3.94080162e-01 -8.78226459e-02
-2.70037681e-01 -5.63857034e-02 -5.10509729e-01 -4.08380598e-01
6.16364321e-03 3.40802252e-01 2.54152149e-01 -4.62417006e-01
7.40791023e-01 -9.57444847e-01 -7.39430428e-01 2.68458605e-01
-4.48707610e-01 4.23851877e-01 8.81616831e-01 8.07092905e-01
4.29538429e-01 -1.62537396e-01 6.20911658e-01 2.42677256e-02
9.55551744e-01 -3.37055475e-01 -1.50153423e-02 4.52948585e-02
-6.96823478e-01 -6.22933246e-02 7.08878577e-01 -8.93746197e-01
-4.79021192e-01 4.84623946e-02 -4.74588007e-01 7.16219470e-02
-3.08722317e-01 5.65943122e-01 1.80307791e-01 -2.13058248e-01
7.67568886e-01 5.92875421e-01 3.55819725e-02 -1.35709196e-01
-2.08005205e-01 7.61609912e-01 6.09387398e-01 -1.76890746e-01
-2.09731564e-01 8.59256610e-02 -4.09054607e-02 -6.54279470e-01
-6.24512210e-02 -1.77528590e-01 -1.13561356e+00 -5.36160111e-01
8.17524910e-01 -5.45238316e-01 -7.95103550e-01 7.14404881e-01
-9.91461933e-01 -7.63006330e-01 1.51298776e-01 1.36888635e+00
-9.13630962e-01 -5.17387167e-02 -6.54806674e-01 -4.86402929e-01
-3.86403948e-01 -1.35632038e+00 6.56973839e-01 -5.27980328e-02
-1.04635954e+00 -6.11459315e-01 -2.45773807e-01 1.71480983e-01
5.12748182e-01 2.75759548e-01 7.94457912e-01 -7.96789587e-01
-5.17214499e-02 -6.02863193e-01 2.09771186e-01 6.74356669e-02
8.78547430e-02 -5.29710352e-01 -5.96210241e-01 -1.68715328e-01
1.13936812e-01 -3.21171395e-02 2.99144030e-01 7.49084830e-01
1.30179310e+00 1.51151374e-01 -5.08615196e-01 5.88025808e-01
1.03881752e+00 5.96190751e-01 9.65406835e-01 6.77063227e-01
4.76502389e-01 4.32833731e-01 2.24119037e-01 1.82999194e-01
1.20321169e-01 8.34252834e-01 2.83210188e-01 3.95707667e-01
-5.41302748e-02 2.05576107e-01 2.43099168e-01 7.64741123e-01
-7.57930636e-01 1.99534580e-01 -1.06260121e+00 2.96525151e-01
-1.67374611e+00 -1.12450778e+00 -4.37108129e-01 2.02056289e+00
4.87104774e-01 2.59913266e-01 -1.05620117e-03 7.90559590e-01
2.54424632e-01 -1.57422706e-01 -7.15272665e-01 -4.78337258e-01
2.21536189e-01 5.61905146e-01 4.32961613e-01 3.40590328e-02
-5.98941267e-01 6.61651969e-01 7.02238941e+00 4.46450561e-01
-1.68365884e+00 1.04937188e-01 5.48738129e-02 -4.58801568e-01
4.00983959e-01 -5.31124890e-01 -6.05469644e-01 8.00735772e-01
1.42195165e+00 2.61263475e-02 8.37208986e-01 5.01586556e-01
6.87474906e-01 -3.39683950e-01 -1.40397537e+00 1.08663332e+00
1.04427926e-01 -1.32802963e+00 -5.96976519e-01 7.96600208e-02
2.31966451e-01 4.40945059e-01 -3.70061159e-01 2.77708799e-01
-4.66088027e-01 -1.38822782e+00 8.46522987e-01 7.43234873e-01
8.92664850e-01 -5.84975600e-01 7.86451459e-01 7.65522659e-01
-1.23587906e+00 -3.57225388e-01 5.32106347e-02 -7.15532601e-01
2.27796033e-01 -2.79447194e-02 -5.35805345e-01 1.47440717e-01
7.55942583e-01 9.20376003e-01 -5.34005880e-01 9.65100884e-01
-4.24222201e-01 5.83029151e-01 -3.57593477e-01 -4.62734193e-01
8.63697603e-02 3.73424798e-01 1.24195166e-01 1.10234773e+00
5.65886259e-01 1.73012301e-01 -6.17300332e-01 7.94947088e-01
4.48681265e-01 -2.31514256e-02 -3.68947506e-01 -6.03998862e-02
3.21637839e-01 6.53030992e-01 -8.30731213e-01 9.29917544e-02
-1.83459371e-01 9.75320399e-01 3.96047413e-01 2.53433287e-01
-8.82346869e-01 -5.52919865e-01 4.80354607e-01 1.01912834e-01
-1.35610446e-01 -5.32822013e-01 -8.02489579e-01 -8.17168117e-01
2.55035222e-01 -5.57919860e-01 -1.60250887e-01 -9.69506443e-01
-6.30926967e-01 7.19336927e-01 8.91532078e-02 -1.30034459e+00
-6.14811897e-01 -1.14112365e+00 -6.49175406e-01 9.66265321e-01
-9.68379617e-01 -7.03788996e-01 -3.56640935e-01 6.14131689e-01
5.20127058e-01 1.32319573e-02 9.62375998e-01 1.38264328e-01
-4.07082886e-01 3.49510998e-01 -3.48364823e-02 1.76045030e-01
2.07233265e-01 -8.20968986e-01 1.81644619e-01 5.32413900e-01
1.74624100e-01 5.82581282e-01 5.79413712e-01 -5.22355914e-01
-1.37444127e+00 -5.45614064e-01 8.95682752e-01 -4.60708052e-01
5.61676323e-01 -2.50147730e-01 -7.64098883e-01 6.53444707e-01
-1.15634903e-01 -3.85468036e-01 9.58628953e-01 2.84695737e-02
4.04164284e-01 1.09414414e-01 -7.52654076e-01 4.17697340e-01
1.05142725e+00 -5.08578479e-01 -1.08651769e+00 1.93971708e-01
-4.47138697e-01 -4.23682302e-01 -1.02866733e+00 4.51040596e-01
1.24877918e+00 -8.17072332e-01 8.67761493e-01 -5.49627841e-01
4.92655843e-01 2.32168093e-01 7.25926980e-02 -1.75319099e+00
-6.24490440e-01 -3.39902006e-02 -3.62733006e-01 2.15113968e-01
3.39031845e-01 -4.84072089e-01 7.22296596e-01 6.98103368e-01
-3.36231947e-01 -1.03968716e+00 -1.12043643e+00 -8.42036188e-01
6.69196919e-02 -1.26461923e+00 4.64255869e-01 3.80450279e-01
7.70189881e-01 2.32990086e-01 -2.93191969e-01 -1.55180573e-01
-5.16014360e-02 -2.33651698e-01 2.88617134e-01 -1.27414739e+00
8.71949866e-02 -9.01097238e-01 -9.59261775e-01 -8.52354586e-01
4.05784428e-01 -1.18476462e+00 7.20820576e-02 -2.07563400e+00
-2.37055123e-01 -4.01715785e-02 -4.62416440e-01 7.61205912e-01
1.04715243e-01 2.92733312e-01 7.44537190e-02 3.44162375e-01
1.14432767e-01 3.60848218e-01 1.08181131e+00 -2.00475335e-01
-5.57676017e-01 8.93429220e-02 -3.63640606e-01 5.09176314e-01
9.99904454e-01 -4.18091118e-01 -3.48908961e-01 -5.02604246e-01
5.70224114e-02 2.96187967e-01 2.30895117e-01 -1.60421002e+00
3.59204859e-01 -3.98194082e-02 1.02278304e+00 -6.92724705e-01
4.78895932e-01 -6.41026497e-01 3.74378741e-01 7.66138196e-01
-3.63599032e-01 2.18256209e-02 4.90994126e-01 2.49645352e-01
1.64684430e-02 1.28062800e-01 3.49999249e-01 2.88124949e-01
-9.29423511e-01 7.26887733e-02 -1.04290128e+00 -5.23636281e-01
1.23738790e+00 -9.37001228e-01 -1.82745650e-01 9.21726972e-02
-1.10932064e+00 -1.35488853e-01 -1.51968896e-01 7.41882145e-01
6.52634203e-01 -1.33775043e+00 -4.53244716e-01 5.57336926e-01
2.47050785e-02 -7.14860082e-01 3.84464592e-01 1.43538558e+00
-4.73192573e-01 8.12613487e-01 -1.08519006e+00 -6.10976517e-01
-1.06903625e+00 6.30432740e-02 3.47218037e-01 1.22298216e-02
-9.97465730e-01 7.88216889e-01 -6.02302372e-01 1.33421287e-01
2.25182757e-01 -6.04562163e-01 -6.26092196e-01 -5.30938022e-02
7.28880286e-01 3.74376148e-01 5.42420983e-01 -6.75469637e-01
-7.05311954e-01 3.43263328e-01 3.00978303e-01 6.10900074e-02
1.62653601e+00 2.21614540e-01 1.37035310e-01 5.14067292e-01
1.03747165e+00 -5.45235217e-01 -1.19138086e+00 5.85083961e-01
3.26495059e-02 -2.22035065e-01 2.99902409e-01 -1.15596664e+00
-9.29357648e-01 1.07462013e+00 8.25826705e-01 -1.33919507e-01
1.19282794e+00 -3.20160151e-01 7.84639657e-01 3.51108700e-01
5.90550661e-01 -1.36308527e+00 -2.33155310e-01 1.80122569e-01
9.96504605e-01 -9.31412876e-01 5.47097027e-02 9.59782377e-02
-3.78019810e-01 1.54313493e+00 7.20554531e-01 -3.28483820e-01
7.03782797e-01 3.84811223e-01 -3.91045868e-01 -2.59567380e-01
-4.90363121e-01 1.10249713e-01 7.81487584e-01 6.97509766e-01
7.28615344e-01 3.55471522e-01 -7.86409676e-01 1.25465190e+00
-5.69853961e-01 5.76731026e-01 9.33383405e-02 9.95137036e-01
-3.37592900e-01 -7.35797286e-01 -2.65256852e-01 1.20406282e+00
-3.55882436e-01 7.14243501e-02 -5.30477706e-03 9.62079883e-01
1.53912187e-01 9.14305091e-01 1.67379946e-01 -1.07379234e+00
4.11272228e-01 2.48630509e-01 8.97092938e-01 -4.53345299e-01
-7.98260629e-01 -3.81124228e-01 -2.64062230e-02 -8.60807717e-01
-2.61361569e-01 -6.72698855e-01 -1.40926218e+00 -1.80241063e-01
-3.50655824e-01 -2.37706766e-01 9.72987294e-01 1.45609260e+00
1.35232389e-01 9.44280744e-01 -2.19293963e-02 -1.51373374e+00
-1.44273400e-01 -1.37082422e+00 -5.28633773e-01 1.72171980e-01
-3.65029275e-02 -9.88505423e-01 -2.61556834e-01 -1.10556036e-01] | [12.990964889526367, 3.4084746837615967] |
f6fb58f5-42e2-463d-bfab-78f4fe38e5eb | tinyml-tools-applications-challenges-and | 2303.13569 | null | https://arxiv.org/abs/2303.13569v1 | https://arxiv.org/pdf/2303.13569v1.pdf | TinyML: Tools, Applications, Challenges, and Future Research Directions | In recent years, Artificial Intelligence (AI) and Machine learning (ML) have gained significant interest from both, industry and academia. Notably, conventional ML techniques require enormous amounts of power to meet the desired accuracy, which has limited their use mainly to high-capability devices such as network nodes. However, with many advancements in technologies such as the Internet of Things (IoT) and edge computing, it is desirable to incorporate ML techniques into resource-constrained embedded devices for distributed and ubiquitous intelligence. This has motivated the emergence of the TinyML paradigm which is an embedded ML technique that enables ML applications on multiple cheap, resource- and power-constrained devices. However, during this transition towards appropriate implementation of the TinyML technology, multiple challenges such as processing capacity optimization, improved reliability, and maintenance of learning models' accuracy require timely solutions. In this article, various avenues available for TinyML implementation are reviewed. Firstly, a background of TinyML is provided, followed by detailed discussions on various tools supporting TinyML. Then, state-of-art applications of TinyML using advanced technologies are detailed. Lastly, various research challenges and future directions are identified. | ['Onel L. A. López', 'Sridhar Iyer', 'Prasoon Raghuwanshi', 'Krishna Pai', 'Rakhee Kallimani'] | 2023-03-23 | null | null | null | null | ['edge-computing'] | ['time-series'] | [-1.06856890e-01 1.19335242e-01 -3.73516738e-01 -5.09807050e-01
-2.38355637e-01 -2.71609873e-01 3.31057906e-01 2.23467544e-01
-1.68912664e-01 6.55298829e-01 -1.06901646e-01 -1.51421770e-01
9.54095926e-03 -1.04423070e+00 -4.00642663e-01 -2.90159851e-01
-6.25805408e-02 2.78010547e-01 5.68294153e-02 3.14300925e-01
-7.35531822e-02 4.86064851e-01 -1.96142328e+00 2.36571208e-01
9.27518427e-01 1.39246976e+00 1.88838094e-01 5.46483397e-01
-4.70268846e-01 1.00774038e+00 -7.34379649e-01 -4.04649198e-01
2.09625930e-01 -2.46827260e-01 -3.48128825e-01 -1.42467931e-01
1.72612995e-01 -2.54485965e-01 -6.74187616e-02 9.62384820e-01
6.42365813e-01 -1.43897444e-01 -9.98990089e-02 -1.50431943e+00
-2.98429608e-01 6.74140394e-01 -6.57684961e-03 -1.84003666e-01
4.27511036e-01 1.44741818e-01 5.68143308e-01 -6.63889050e-01
4.95243192e-01 1.03710258e+00 6.99738681e-01 5.13921797e-01
-4.05865371e-01 -5.63379347e-01 -5.52657209e-02 2.44882211e-01
-1.26397681e+00 -7.99025953e-01 6.98817551e-01 3.53102200e-02
1.37384248e+00 5.42976737e-01 9.64261532e-01 6.39230549e-01
5.03853679e-01 9.74233389e-01 8.08857799e-01 -5.89531362e-01
7.01888263e-01 4.01869148e-01 -2.67224073e-01 8.00313532e-01
4.86326337e-01 -5.32198191e-01 -5.39557874e-01 -1.43942870e-02
5.77268124e-01 1.36293441e-01 3.79855245e-01 -3.00338626e-01
-1.24720550e+00 4.10412878e-01 2.63444215e-01 5.82705259e-01
-4.26275820e-01 4.28624392e-01 5.68859339e-01 2.13352188e-01
6.23769701e-01 5.97911701e-02 -7.45768845e-01 -5.89207113e-01
-7.23165393e-01 -1.69830650e-01 1.16195261e+00 1.21080542e+00
6.17416203e-01 3.07263017e-01 4.88194644e-01 6.71831489e-01
7.22448707e-01 4.33560103e-01 4.81452525e-01 -1.12762058e+00
4.61585581e-01 1.10562444e+00 -3.06469977e-01 -9.74805593e-01
-4.78714317e-01 -2.51744211e-01 -1.04849470e+00 1.75518245e-01
-1.19342804e-01 -1.89596608e-01 -5.80957651e-01 1.07155573e+00
7.29224324e-01 1.95684642e-01 1.63693517e-01 4.04356360e-01
1.08555722e+00 5.82625926e-01 1.65131956e-01 -1.70848653e-01
1.11679673e+00 -9.28263426e-01 -9.74351168e-01 -4.46885854e-01
8.44578922e-01 -4.02005106e-01 1.06255221e+00 3.22307795e-01
-1.21834731e+00 -3.55871886e-01 -1.26021802e+00 -1.18863650e-01
-7.15707362e-01 -2.21424535e-01 1.25453806e+00 9.84598994e-01
-8.23386014e-01 4.82717186e-01 -1.19976044e+00 -7.48535156e-01
6.56442523e-01 6.57029510e-01 -3.79788578e-02 -3.57764274e-01
-7.52723813e-01 7.72620559e-01 1.72555670e-01 -2.35338993e-02
-3.54903311e-01 -7.89171636e-01 -7.97320247e-01 -1.16301201e-01
2.89285220e-02 -8.68759155e-01 1.09430480e+00 -7.73075879e-01
-1.88353479e+00 5.76774120e-01 1.03547871e-01 -3.65077794e-01
3.49370003e-01 -1.73537597e-01 -7.26518154e-01 -1.27690405e-01
-2.17764631e-01 6.08562350e-01 4.26405281e-01 -7.50458002e-01
-8.61365557e-01 -4.24304545e-01 1.06319614e-01 2.19709828e-01
-5.99257529e-01 2.06582084e-01 -4.01762277e-01 -1.45709634e-01
-1.72624849e-02 -7.18180180e-01 -3.85117352e-01 3.85717839e-01
-5.55839203e-02 -3.31897229e-01 1.31194413e+00 -1.68133378e-01
1.10950077e+00 -2.11736202e+00 -2.10059986e-01 -1.20365299e-01
3.10649991e-01 9.81254131e-02 1.52696609e-01 3.74768972e-01
7.74601460e-01 2.29019463e-01 4.97295037e-02 -2.50557631e-01
1.62047550e-01 4.39730674e-01 4.12228882e-01 4.95182753e-01
-3.24665934e-01 1.04138279e+00 -7.78691351e-01 -8.15645814e-01
8.30831528e-01 7.64376938e-01 -3.09274077e-01 -8.55112169e-03
-4.52658892e-01 3.35536927e-01 -7.19874144e-01 1.24325418e+00
3.32130969e-01 -3.23512077e-01 2.48321787e-01 -2.06480235e-01
-2.98980147e-01 1.73964173e-01 -1.27477896e+00 1.86323631e+00
-8.20413172e-01 4.17568922e-01 3.56499761e-01 -7.01892793e-01
9.01151061e-01 5.83932102e-01 8.90152037e-01 -9.06370997e-01
1.74656510e-01 5.78581393e-01 -1.37295634e-01 -6.07008576e-01
-3.53483446e-02 2.08074883e-01 -1.40387803e-01 5.97671509e-01
-2.00675771e-01 -5.44649661e-01 2.96902269e-01 7.35527202e-02
1.29613602e+00 1.69484511e-01 4.77204949e-01 -4.15322110e-02
2.59072214e-01 -1.04880974e-01 6.44615948e-01 3.88642341e-01
-2.82106221e-01 -5.30307144e-02 -2.07522899e-01 -8.89530301e-01
-8.16516936e-01 -6.73615634e-01 -4.89402339e-02 8.14814985e-01
2.67049368e-03 -7.28794992e-01 -6.48653924e-01 -5.07035553e-01
-8.96674544e-02 4.76023734e-01 8.12537596e-02 1.47602409e-01
-3.66979659e-01 -8.72188807e-01 3.45257103e-01 4.48481917e-01
1.07123721e+00 -1.22329676e+00 -1.34357107e+00 3.48637819e-01
2.51139253e-01 -1.32849801e+00 4.66574669e-01 6.65798634e-02
-1.34796250e+00 -5.56473792e-01 8.61338302e-02 -5.94336987e-01
5.41632533e-01 2.42820159e-02 1.11405766e+00 3.65016818e-01
-5.17582953e-01 6.98292315e-01 -4.28243041e-01 -8.83182287e-01
-1.90403506e-01 -1.07070273e-02 2.94297457e-01 -4.32824761e-01
3.53660345e-01 -7.73492754e-01 -5.02066851e-01 -3.28498706e-02
-8.36279571e-01 3.16080511e-01 8.19361567e-01 -2.19742185e-03
5.60786366e-01 5.14166057e-01 7.05261946e-01 -5.95187902e-01
4.59697336e-01 -5.21935880e-01 -5.37622571e-01 3.57826024e-01
-7.22043395e-01 -2.15668023e-01 7.86759079e-01 -3.33182305e-01
-9.14578915e-01 1.23146780e-01 -2.39278331e-01 7.26743415e-02
-2.19953254e-01 5.51128149e-01 -7.39454031e-01 -2.38588139e-01
2.43743271e-01 -1.88978299e-01 -1.68845668e-01 -4.27789629e-01
4.53087777e-01 1.14117563e+00 9.75964665e-02 -2.94054657e-01
2.81578481e-01 4.18212563e-01 -1.83062121e-01 -1.02765167e+00
-7.10713983e-01 -2.26030387e-02 -5.62558889e-01 -3.36762846e-01
4.62452054e-01 -6.95086956e-01 -5.87812901e-01 3.49998862e-01
-9.34766650e-01 -5.18239081e-01 -6.78975463e-01 4.20786351e-01
-3.34814787e-01 -7.72157684e-02 -6.12502158e-01 -7.17484653e-01
-7.61308491e-01 -1.16194415e+00 8.40992630e-01 4.80787158e-01
-3.60941142e-01 -1.08597660e+00 -1.49585500e-01 4.78701800e-01
7.48892367e-01 2.60581613e-01 6.71742916e-01 -3.19519311e-01
-7.93860078e-01 -7.54006982e-01 4.54998622e-03 7.96274841e-02
3.29019040e-01 1.20844156e-01 -9.73619521e-01 -1.89378306e-01
1.10779889e-01 -3.85392368e-01 -2.54034698e-01 2.57191747e-01
1.31924713e+00 -4.71767217e-01 -6.84184670e-01 7.28867769e-01
1.51146138e+00 3.49105954e-01 4.84064430e-01 3.95087212e-01
7.42936671e-01 2.74895877e-01 5.41015148e-01 6.15629077e-01
5.27810991e-01 3.85463774e-01 4.70395803e-01 2.82748416e-02
-1.97806731e-01 4.21382822e-02 1.08591162e-01 1.55898774e+00
-5.96910715e-04 -2.28573978e-01 -8.33315670e-01 1.97738424e-01
-1.81770825e+00 -6.30593240e-01 7.36172274e-02 2.06287956e+00
6.48581147e-01 7.53056780e-02 -3.37771147e-01 4.18306231e-01
1.36513084e-01 1.57816350e-01 -9.43654537e-01 -5.82025111e-01
9.22407806e-02 2.27848627e-02 1.73059195e-01 7.33386725e-02
-8.38338673e-01 7.92946100e-01 6.66780186e+00 5.91776669e-01
-1.20310163e+00 3.50911438e-01 5.97276449e-01 -1.47088513e-01
-1.45843074e-01 -2.50147432e-01 -6.28192067e-01 5.26547670e-01
1.43987024e+00 -2.34670013e-01 5.20264387e-01 1.18662918e+00
2.19058096e-01 -2.04524428e-01 -1.10592341e+00 1.21630943e+00
-9.61017460e-02 -1.37164199e+00 -3.02961707e-01 -3.66984457e-02
6.96307063e-01 2.91101336e-01 -1.38027966e-01 1.60892606e-01
3.39290947e-01 -9.11625326e-01 5.92311859e-01 1.52597606e-01
9.32037532e-01 -8.32653940e-01 7.77453125e-01 5.10659039e-01
-1.45855188e+00 -1.93258449e-01 -5.76533735e-01 -4.90469098e-01
-2.44209561e-02 1.18443680e+00 -5.49906909e-01 3.83341223e-01
1.04842710e+00 8.23470354e-01 -4.08373803e-01 9.97883022e-01
-6.44959733e-02 3.00327390e-01 -5.78426898e-01 -2.95556098e-01
-3.40623021e-01 -1.28431842e-01 8.31823274e-02 9.89482939e-01
3.85236025e-01 1.24037594e-01 1.79851025e-01 5.67949533e-01
-3.17213774e-01 6.21399768e-02 -7.20893443e-01 -2.11829022e-01
9.69308734e-01 1.66557705e+00 -1.01139092e+00 -4.86401230e-01
-7.72991896e-01 8.39979410e-01 1.05117194e-01 -9.63294804e-02
-7.40971684e-01 -3.12211722e-01 7.19439149e-01 -1.39059722e-01
-3.56227517e-01 -3.27393085e-01 -6.77661955e-01 -9.30853367e-01
-2.41846628e-02 -9.96585011e-01 5.84858097e-02 -5.79656541e-01
-9.07040000e-01 3.25154483e-01 -3.38180661e-01 -1.11909819e+00
-5.14955893e-02 -2.20860049e-01 -4.63303804e-01 6.87386468e-02
-1.03131998e+00 -1.30813468e+00 -5.19122481e-01 2.76884943e-01
8.57385457e-01 1.43131916e-03 1.18079197e+00 5.63284099e-01
-7.24494338e-01 1.80924356e-01 2.36245722e-01 -2.79601365e-01
1.79234877e-01 -8.11094224e-01 3.25051188e-01 5.86083651e-01
2.32726395e-01 3.45858693e-01 5.58797061e-01 -6.42642438e-01
-2.29674673e+00 -1.12515604e+00 8.62399638e-01 -3.60102415e-01
5.41509748e-01 -5.64004302e-01 -4.06434745e-01 9.54205751e-01
1.04141505e-02 6.40923619e-01 6.37574673e-01 -3.22538495e-01
1.60994247e-01 -6.49984539e-01 -1.56387842e+00 5.69487333e-01
9.37645555e-01 -3.34547341e-01 8.58333930e-02 5.78891933e-01
8.32343876e-01 -2.67339796e-01 -1.25199640e+00 4.21843231e-01
6.39191210e-01 -9.19533432e-01 8.83404851e-01 5.96511066e-02
-1.10970944e-01 -1.77206397e-01 -2.79821783e-01 -9.52340484e-01
1.06274202e-01 -5.64642608e-01 -7.43502975e-01 1.32831180e+00
7.05766976e-02 -6.52423739e-01 9.50259745e-01 1.07797670e+00
-2.63090163e-01 -8.93916905e-01 -9.82217133e-01 -5.95090866e-01
-4.74963099e-01 -8.30329120e-01 8.36915135e-01 8.07890296e-01
4.75790024e-01 2.07026273e-01 -7.76252672e-02 -1.98074639e-01
6.77042723e-01 -2.87182331e-02 6.79143310e-01 -1.47927511e+00
1.68224707e-01 -2.58661639e-02 -6.95398927e-01 -9.01109874e-01
-2.22902149e-01 -5.31972468e-01 -1.63662404e-01 -2.07771492e+00
-3.25507745e-02 -9.47953343e-01 -6.76823631e-02 4.80653703e-01
4.40347940e-01 3.51367623e-01 1.15741491e-01 1.39068991e-01
-1.23248005e+00 2.94891238e-01 1.02913821e+00 -1.37174219e-01
-2.17243880e-01 6.46988675e-03 -3.67550880e-01 9.91640151e-01
1.06253314e+00 -3.46680880e-01 -5.47124445e-01 -8.11135709e-01
6.51989579e-01 -1.78186268e-01 -1.27478376e-01 -1.62429738e+00
6.03934050e-01 -1.09431453e-01 5.25233448e-01 -2.60392368e-01
4.65072036e-01 -1.36843777e+00 4.64290738e-01 6.43569887e-01
1.98712945e-01 1.52435496e-01 1.30381227e-01 1.63182586e-01
7.03379139e-02 -2.43019722e-02 5.32790601e-01 -3.16624075e-01
-8.61130536e-01 2.47292846e-01 -5.65147817e-01 -4.88964349e-01
1.50491452e+00 -3.75651538e-01 -2.67376989e-01 -1.39550701e-01
-1.30984202e-01 -1.40544884e-02 6.16442561e-01 3.64174753e-01
6.94290936e-01 -1.04613042e+00 -2.53174454e-01 3.05709302e-01
-1.92218021e-01 3.94211352e-01 -5.46618067e-02 7.68366635e-01
-7.30390668e-01 5.47405720e-01 -1.18762210e-01 -4.74112213e-01
-1.29091215e+00 4.31400627e-01 1.37356147e-01 9.83944461e-02
-8.68473411e-01 6.18127823e-01 -6.49310708e-01 -7.43685901e-01
4.98379469e-01 -5.00481844e-01 1.65057540e-01 -4.75761503e-01
6.78577542e-01 7.42780328e-01 3.49333942e-01 -1.85531929e-01
-5.51395416e-01 5.00317216e-01 5.34355879e-01 1.95387110e-01
1.45006335e+00 -4.54079568e-01 -3.67356062e-01 7.68172145e-01
7.92288363e-01 8.62353202e-03 -8.67166162e-01 2.64487088e-01
2.10187286e-01 -2.24171922e-01 2.00905070e-01 -8.33766401e-01
-1.11768842e+00 7.22361624e-01 6.92481577e-01 2.75115132e-01
1.06180418e+00 8.29134583e-02 1.05712211e+00 3.52851838e-01
1.07851803e+00 -1.15761280e+00 -1.05489746e-01 2.00617731e-01
3.07678550e-01 -1.20162070e+00 3.39116424e-01 -3.92147422e-01
-5.57541512e-02 9.80090618e-01 5.02453566e-01 3.76424968e-01
7.53207326e-01 1.03476799e+00 1.24598682e-01 -2.90050834e-01
-8.36718976e-01 3.23549688e-01 -1.80954143e-01 8.10558975e-01
6.83686078e-01 1.24439389e-01 -6.63397908e-02 3.65575105e-01
-1.55959517e-01 3.68281543e-01 1.27416715e-01 1.46771896e+00
-6.55126393e-01 -1.11416852e+00 -2.67379344e-01 7.08443344e-01
-4.61845994e-01 1.93452522e-01 -2.50713348e-01 6.54915214e-01
4.09612328e-01 1.22993851e+00 2.92326301e-01 -4.10105973e-01
2.95120757e-02 -8.88114274e-02 4.40857679e-01 -6.38233900e-01
-5.61152995e-01 -3.23053241e-01 6.05729446e-02 -7.35725343e-01
-4.64713365e-01 -2.32600495e-01 -1.64273822e+00 -6.73348188e-01
-2.44109571e-01 -6.50936365e-03 1.48512793e+00 1.00546610e+00
6.42572880e-01 5.36809385e-01 1.24831818e-01 -7.93863237e-01
2.51262193e-03 -8.24058294e-01 -4.19865251e-01 -2.46068850e-01
-2.64627516e-01 -1.59369409e-01 -1.53061539e-01 -2.02492042e-03] | [8.015889167785645, 2.64483642578125] |
ddc58d10-3c9f-45b5-8bb0-47ffdff8338d | pybibx-a-python-library-for-bibliometric-and | 2304.14516 | null | https://arxiv.org/abs/2304.14516v1 | https://arxiv.org/pdf/2304.14516v1.pdf | pyBibX -- A Python Library for Bibliometric and Scientometric Analysis Powered with Artificial Intelligence Tools | Bibliometric and Scientometric analyses offer invaluable perspectives on the complex research terrain and collaborative dynamics spanning diverse academic disciplines. This paper presents pyBibX, a python library devised to conduct comprehensive bibliometric and scientometric analyses on raw data files sourced from Scopus, Web of Science, and PubMed, seamlessly integrating state of the art AI capabilities into its core functionality. The library executes a comprehensive EDA, presenting outcomes via visually appealing graphical illustrations. Network capabilities have been deftly integrated, encompassing Citation, Collaboration, and Similarity Analysis. Furthermore, the library incorporates AI capabilities, including Embedding vectors, Topic Modeling, Text Summarization, and other general Natural Language Processing tasks, employing models such as Sentence-BERT, BerTopic, BERT, chatGPT, and PEGASUS. As a demonstration, we have analyzed 184 documents associated with multiple-criteria decision analysis published between 1984 and 2023. The EDA emphasized a growing fascination with decision-making and fuzzy logic methodologies. Next, Network Analysis further accentuated the significance of central authors and intra-continental collaboration, identifying Canada and China as crucial collaboration hubs. Finally, AI Analysis distinguished two primary topics and chatGPT preeminence in Text Summarization. It also proved to be an indispensable instrument for interpreting results, as our library enables researchers to pose inquiries to chatGPT regarding bibliometric outcomes. Even so, data homogeneity remains a daunting challenge due to database inconsistencies. PyBibX is the first application integrating cutting-edge AI capabilities for analyzing scientific publications, enabling researchers to examine and interpret these outcomes more effectively. | ['Carlos Henrique Tarjano Santos', 'Marcio Pereira Basilio', 'Valdecy Pereira'] | 2023-04-27 | null | null | null | null | ['text-summarization'] | ['natural-language-processing'] | [-5.54645717e-01 -2.06771240e-01 -6.16919339e-01 4.29756969e-01
-3.99390429e-01 -8.50657582e-01 9.32130039e-01 7.40651369e-01
-3.55311841e-01 5.76008439e-01 6.00094259e-01 -7.96236992e-01
-8.63191605e-01 -7.70534992e-01 -2.09188998e-01 -2.03219593e-01
-1.82652399e-01 6.10005617e-01 -6.28723800e-01 -2.69024342e-01
1.02252448e+00 4.87040132e-01 -1.39182484e+00 -1.22650407e-01
1.15616131e+00 6.04313374e-01 7.29807541e-02 2.96772122e-01
-5.43372989e-01 5.48060238e-01 -8.78710806e-01 -7.36104488e-01
-1.33024603e-01 1.44343348e-02 -6.64842725e-01 -6.69073999e-01
1.10466123e-01 2.17624590e-01 -3.56914103e-01 9.69336748e-01
5.90601206e-01 -7.67044127e-02 9.46800292e-01 -1.27902102e+00
-9.37092960e-01 7.53117859e-01 -4.14749235e-01 6.67182863e-01
9.69810545e-01 2.71189660e-01 1.31513441e+00 -8.65384698e-01
9.50739324e-01 1.30424678e+00 5.73691249e-01 -3.41224968e-01
-1.11752856e+00 -7.62410343e-01 -1.56927899e-01 4.93922561e-01
-1.21509635e+00 -1.65349141e-01 9.29303110e-01 -9.31118786e-01
9.04903829e-01 5.17296314e-01 1.15960157e+00 1.04306328e+00
4.53655273e-01 4.62207764e-01 1.08069551e+00 -4.94426370e-01
3.01343709e-01 -3.02627869e-02 2.90862858e-01 2.88511664e-01
8.52875054e-01 -4.60271478e-01 -6.82177067e-01 -3.80232215e-01
2.96857476e-01 1.56048596e-01 -1.67937353e-01 4.72341418e-01
-1.59531677e+00 5.75442374e-01 2.52373815e-01 5.73468804e-01
-7.26952493e-01 -3.49524468e-01 5.21862149e-01 4.03570294e-01
4.15724277e-01 1.18839085e+00 7.36438408e-02 -4.05300170e-01
-1.38398147e+00 2.55189091e-01 1.21712077e+00 7.91923642e-01
2.80082285e-01 -7.68453926e-02 -2.82145828e-01 4.36446518e-01
1.65645227e-01 3.18371505e-01 3.56447786e-01 -1.37579143e+00
3.79389465e-01 1.19638669e+00 -2.01696828e-01 -1.47638130e+00
-3.57867420e-01 -5.04539132e-01 -9.12143946e-01 2.08795127e-02
3.63140330e-02 -4.80331331e-02 -3.30112010e-01 9.78340089e-01
-1.44657463e-01 -4.84557450e-01 -1.01251699e-01 7.18825340e-01
1.17171276e+00 6.04589641e-01 2.08105341e-01 -3.24700713e-01
1.52016568e+00 -4.77167696e-01 -1.15706408e+00 3.21352214e-01
4.94556695e-01 -7.58153081e-01 9.42217350e-01 4.68494833e-01
-1.40428138e+00 1.77722275e-02 -9.81199265e-01 -6.73477128e-02
-9.21367884e-01 -6.22718185e-02 7.19038665e-01 2.67063498e-01
-1.25587451e+00 5.86910665e-01 -3.83410543e-01 -3.80303234e-01
7.65227556e-01 1.23398535e-01 -2.28657648e-01 5.74825890e-02
-1.29532123e+00 1.23572385e+00 3.43873620e-01 -1.80977553e-01
1.43137604e-01 -1.11035359e+00 -6.10053003e-01 3.99066508e-01
3.58727068e-01 -1.04662836e+00 6.54096484e-01 -3.32544416e-01
-1.23229170e+00 7.59379327e-01 -2.02135265e-01 -2.75506496e-01
4.26834196e-01 1.90533414e-01 -5.94112039e-01 3.56116056e-01
4.92269993e-01 6.38289601e-02 -1.11546367e-01 -7.86230028e-01
-2.44049802e-01 -4.71275389e-01 -3.03079247e-01 4.86132950e-01
-6.51263773e-01 4.30721968e-01 -5.45970917e-01 -7.61822164e-01
-1.64174169e-01 -1.48061663e-01 -6.80026114e-02 -2.30208397e-01
-5.01200914e-01 -6.43516481e-01 4.39783663e-01 -8.16392720e-01
1.82777703e+00 -1.55649149e+00 2.62905747e-01 5.26921928e-01
7.65554130e-01 -1.67486086e-01 4.89329457e-01 8.70888174e-01
2.88202167e-01 7.54897952e-01 2.55819876e-02 1.55096129e-01
2.25302756e-01 -3.25375587e-01 -3.47316489e-02 2.89824635e-01
6.10043146e-02 1.10154510e+00 -9.05784011e-01 -8.59286189e-01
2.87862062e-01 1.58022940e-02 -3.20766002e-01 -3.59675854e-01
1.09094605e-01 -1.10032737e-01 -5.96773088e-01 1.18234503e+00
2.19223350e-01 -4.26391304e-01 3.63514185e-01 -6.73604459e-02
-8.10330212e-01 3.25034976e-01 -8.63425493e-01 1.49775720e+00
-1.82263047e-01 1.03836071e+00 4.25748639e-02 -8.57259631e-01
8.57352018e-01 2.47265309e-01 6.66135728e-01 -8.32065940e-01
3.17544818e-01 2.04882503e-01 1.28597513e-01 -4.44750875e-01
7.17725635e-01 4.42190945e-01 -1.43849134e-01 4.73148853e-01
9.18239206e-02 -3.89260411e-01 7.08375275e-01 7.41946518e-01
9.91510212e-01 -3.00866663e-01 5.00273705e-01 -6.85971737e-01
3.17851722e-01 5.46290517e-01 2.14481190e-01 5.77442586e-01
2.32512038e-02 1.91566423e-01 8.41619074e-01 -9.93868932e-02
-1.14729476e+00 -9.76551771e-01 -4.00593966e-01 6.25402987e-01
-5.84030636e-02 -6.95117533e-01 -3.25512975e-01 3.31023753e-01
3.39989662e-01 8.00010324e-01 -4.08289313e-01 1.10808730e-01
-9.20444205e-02 -6.67780697e-01 5.79294622e-01 5.29740006e-02
2.19880387e-01 -9.40701663e-01 -4.40877885e-01 1.41156867e-01
-6.91324100e-02 -7.30267525e-01 2.11743340e-02 -2.62767434e-01
-8.20813298e-01 -1.14387655e+00 -8.53452682e-01 -5.42706609e-01
2.92545259e-01 -1.50782526e-01 1.22820747e+00 -2.81095147e-01
-5.28934181e-01 3.84990007e-01 1.74812645e-01 -5.65672576e-01
-2.83703119e-01 1.80466831e-01 2.92888641e-01 -8.89048994e-01
4.61311013e-01 -6.10780597e-01 -5.80344141e-01 -9.97738242e-02
-7.05047429e-01 -8.85952637e-02 5.18485010e-01 5.76438010e-01
2.24488139e-01 -1.13474712e-01 8.92163754e-01 -4.50680465e-01
1.39565718e+00 -1.02120411e+00 -4.74553823e-01 3.33775192e-01
-1.21346879e+00 -2.21408367e-01 4.73942816e-01 2.33052280e-02
-7.85455525e-01 -9.72217262e-01 4.55586404e-01 -2.27213338e-01
1.88375786e-01 1.27410090e+00 1.06899925e-01 1.83675304e-01
5.83821297e-01 -6.02731779e-02 3.12984407e-01 -3.95844489e-01
3.94376665e-01 1.00419307e+00 7.45387495e-01 -6.80749714e-01
4.94656950e-01 2.93851137e-01 -1.67616412e-01 -7.77148306e-01
1.75589100e-02 -5.81794798e-01 -3.35703582e-01 -6.14981711e-01
4.68618393e-01 -9.12380815e-01 -1.53692675e+00 -1.04791887e-01
-1.01820683e+00 5.26382327e-01 -2.68422276e-01 7.32216001e-01
5.63548319e-02 2.86970615e-01 -5.19803703e-01 -5.46553075e-01
-7.03081489e-01 -9.34234798e-01 2.03217924e-01 5.74622095e-01
-7.29515433e-01 -1.14183259e+00 9.49908271e-02 4.59580332e-01
5.63208938e-01 3.12207699e-01 1.16189420e+00 -8.75367939e-01
-2.20154181e-01 -2.86930203e-01 -3.16996217e-01 -2.76235938e-01
-1.78438518e-02 6.96659625e-01 -5.18436968e-01 9.78952050e-02
-4.02823895e-01 2.25794986e-01 4.83945727e-01 4.85128373e-01
1.08304870e+00 -5.73341787e-01 -4.71229851e-01 2.82493323e-01
1.21548808e+00 2.90133178e-01 4.40744996e-01 1.01557600e+00
4.70501572e-01 7.04750299e-01 1.34473756e-01 5.21805942e-01
7.20273972e-01 2.98009068e-01 6.97303116e-02 1.62746102e-01
2.16740206e-01 1.82698250e-01 8.25370010e-03 1.31690967e+00
-3.88768941e-01 -1.41666770e-01 -1.52102041e+00 7.06961215e-01
-1.62254643e+00 -1.05621541e+00 -2.76373088e-01 1.85904026e+00
9.64479029e-01 2.33443603e-01 -2.27493723e-03 2.79803332e-02
6.26110911e-01 5.61874807e-02 -3.16953182e-01 -7.22225428e-01
-5.28590620e-01 1.20267831e-01 4.82012272e-01 1.79473817e-01
-5.01063824e-01 7.49019384e-01 6.58387661e+00 7.48147488e-01
-8.71719062e-01 -3.37036371e-01 4.47366357e-01 -3.78457099e-01
-8.47536087e-01 7.19996542e-02 -1.34530753e-01 6.20166957e-01
9.41399038e-01 -1.40227151e+00 1.63635582e-01 4.72749650e-01
4.21924859e-01 -3.43245834e-01 -6.92830801e-01 7.71758258e-01
-1.11390464e-01 -2.11859608e+00 2.56772995e-01 1.12666257e-01
7.35686243e-01 -6.12931792e-03 5.87390624e-02 5.22586703e-02
4.06412631e-01 -1.14141178e+00 4.75644886e-01 8.53890717e-01
8.97467732e-01 -5.62595248e-01 6.51019037e-01 -9.50490907e-02
-6.66941583e-01 -1.06021821e-01 -1.17479026e-01 -5.02584159e-01
3.19896266e-02 9.02432859e-01 -5.50968647e-01 1.24976385e+00
8.30925405e-01 9.66845214e-01 -5.48012972e-01 1.21031880e+00
2.77233988e-01 6.32159948e-01 -2.30394453e-01 -1.97681576e-01
-3.27578224e-02 -6.69092774e-01 9.66581464e-01 1.36141527e+00
3.10954928e-01 2.60207076e-02 -3.76926124e-01 1.08264184e+00
-5.13105810e-01 3.59310687e-01 -8.03685725e-01 -6.05349541e-01
1.07817543e+00 1.10634100e+00 -6.71733320e-01 -4.11727726e-01
-2.79512256e-01 2.47191936e-01 -5.45934588e-02 4.99732643e-01
-1.65760681e-01 -7.50221193e-01 5.38801312e-01 1.50857931e-02
-4.72063988e-01 -2.06168279e-01 -1.12856781e+00 -1.00504041e+00
-5.37648648e-02 -8.08716416e-01 3.50142986e-01 -7.41733789e-01
-1.16420484e+00 4.72525731e-02 7.50751942e-02 -9.04222012e-01
-1.14985093e-01 -4.26060349e-01 -1.03334248e+00 1.04864335e+00
-1.01935112e+00 -5.30734301e-01 -5.36619462e-02 1.68511435e-01
1.28209561e-01 -4.67037141e-01 6.09714568e-01 3.08445126e-01
-9.39969838e-01 2.36472994e-01 6.59366786e-01 -2.00868756e-01
6.14175141e-01 -1.22122991e+00 1.66208789e-01 4.30995345e-01
-2.28201538e-01 8.92384291e-01 6.23412192e-01 -9.33507502e-01
-1.62599099e+00 -3.78543645e-01 1.10466659e+00 -4.57877219e-01
1.19331145e+00 1.11572184e-01 -8.67559791e-01 2.72762924e-01
6.23753250e-01 -8.93325686e-01 7.54068792e-01 3.38544160e-01
7.71258846e-02 3.93364839e-02 -6.84326530e-01 1.07601023e+00
7.69619942e-01 -4.37845558e-01 -1.00416636e+00 3.15220505e-01
5.52532971e-01 -1.04625218e-01 -1.48159587e+00 1.27201691e-01
6.80706918e-01 -3.46072048e-01 1.02602983e+00 -4.66714650e-01
8.50723743e-01 2.55436599e-02 3.95763278e-01 -1.09072399e+00
-4.33612555e-01 -7.64242589e-01 1.98855661e-02 1.46670663e+00
3.88531029e-01 -5.76010942e-01 3.69513839e-01 7.14288950e-01
-2.96035528e-01 -7.10642457e-01 -8.56944799e-01 -4.33768481e-01
3.55009884e-01 -3.09444368e-01 4.70811874e-01 1.46732664e+00
7.39840269e-01 1.92685965e-02 4.40139055e-01 -5.04897952e-01
5.37310302e-01 9.68527794e-02 4.45592135e-01 -1.85321367e+00
6.24294221e-01 -1.47694361e+00 -3.59059155e-01 -5.42471223e-02
7.67745152e-02 -1.26569235e+00 -9.81156170e-01 -2.13309813e+00
1.91722125e-01 -3.23699713e-02 -1.86006531e-01 1.86658651e-01
-5.50918803e-02 -2.74657905e-01 2.39933491e-01 7.78130293e-01
-3.36137742e-01 2.96524525e-01 1.17531669e+00 -3.09728295e-01
-2.74224639e-01 -4.25610453e-01 -1.13582659e+00 6.25392735e-01
7.75875568e-01 -3.20432544e-01 1.33040130e-01 -9.77649391e-02
7.78552294e-01 5.25300801e-02 3.81050974e-01 -6.92999542e-01
9.26081300e-01 -2.73099452e-01 3.23307216e-01 -6.25500143e-01
-1.65500835e-01 -4.64321971e-01 2.55700499e-01 3.20040911e-01
-2.96917081e-01 5.20635545e-01 5.82172871e-01 2.47944921e-01
-3.69317740e-01 1.69277832e-01 -1.97904423e-01 3.46319862e-02
-3.90063077e-01 -2.05376506e-01 -6.33148551e-01 1.27426356e-01
8.51887107e-01 -4.20394152e-01 -8.09144676e-01 -2.35856667e-01
-4.68195945e-01 8.87175202e-01 6.04567111e-01 3.34087431e-01
4.01501954e-01 -1.06809592e+00 -8.44362438e-01 -4.25584674e-01
-6.43800795e-02 -8.50995779e-02 1.72814339e-01 9.91434693e-01
-6.74381495e-01 6.43706024e-01 -4.44013506e-01 -2.23649591e-01
-8.40341866e-01 3.65288377e-01 -2.98381925e-01 -2.03756541e-01
-6.58267438e-01 2.04006284e-01 -4.39731002e-01 -2.62398243e-01
2.64544666e-01 -4.17846665e-02 -6.15236163e-01 8.20502818e-01
1.05859883e-01 1.03436077e+00 1.21863186e-02 -2.66217828e-01
-5.15802205e-01 4.12146538e-01 9.24680978e-02 -3.28791022e-01
1.24228859e+00 -1.34517312e-01 -6.34700239e-01 7.45261312e-01
9.65069413e-01 2.20186114e-02 -2.55136758e-01 5.39599471e-02
4.93834794e-01 -3.21823475e-03 1.67977661e-01 -1.08731806e+00
-4.74610746e-01 5.02601445e-01 -1.55687243e-01 2.91514248e-01
8.13802481e-01 -9.38995257e-02 1.87979534e-01 4.44367647e-01
-2.21781388e-01 -1.37548661e+00 -4.23402220e-01 3.60881150e-01
9.92303908e-01 -9.33591783e-01 6.12296224e-01 1.39777318e-01
-6.12512827e-01 1.19584072e+00 2.36744359e-01 2.49216214e-01
5.05238235e-01 1.97637945e-01 -1.39869943e-01 -6.55146241e-01
-6.83645904e-01 1.47037238e-01 6.48924053e-01 -9.74242948e-03
5.33091486e-01 1.18473239e-01 -9.80647981e-01 7.65444100e-01
-5.91492951e-01 1.72945231e-01 6.56007886e-01 6.95563138e-01
5.94110647e-03 -6.43791735e-01 -7.81648219e-01 9.69551802e-01
-8.17864180e-01 -3.72295916e-01 -6.09488845e-01 9.70926046e-01
-5.83679557e-01 7.44656563e-01 2.97318310e-01 -2.72912264e-01
1.63501710e-01 8.94356966e-02 -1.67602912e-01 -6.61210641e-02
-9.87855911e-01 -1.39714479e-01 2.39429995e-02 -1.92076966e-01
-2.48002052e-01 -7.41257966e-01 -1.19037449e+00 -8.57217610e-01
-1.61627203e-01 5.08198798e-01 1.15446675e+00 7.81372666e-01
7.24135995e-01 8.94085884e-01 3.04207087e-01 -7.08757222e-01
-3.38371359e-02 -9.36091840e-01 -3.02701265e-01 -1.15088753e-01
-1.73412785e-02 -4.61018711e-01 -4.88400847e-01 -3.78454477e-01] | [9.60920238494873, 8.214582443237305] |
4b01185d-4017-4491-9242-6fda6e38ba6d | dmix-adaptive-distance-aware-interpolative | null | null | https://aclanthology.org/2022.acl-short.67 | https://aclanthology.org/2022.acl-short.67.pdf | DMix: Adaptive Distance-aware Interpolative Mixup | Interpolation-based regularisation methods such as Mixup, which generate virtual training samples, have proven to be effective for various tasks and modalities.We extend Mixup and propose DMix, an adaptive distance-aware interpolative Mixup that selects samples based on their diversity in the embedding space. DMix leverages the hyperbolic space as a similarity measure among input samples for a richer encoded representation.DMix achieves state-of-the-art results on sentence classification over existing data augmentation methods on 8 benchmark datasets across English, Arabic, Turkish, and Hindi languages while achieving benchmark F1 scores in 3 times less number of iterations.We probe the effectiveness of DMix in conjunction with various similarity measures and qualitatively analyze the different components.DMix being generalizable, can be applied to various tasks, models and modalities. | ['Lucie Flek', 'Diyi Yang', 'Di Jin', 'Ritesh Soun', 'Shrey Pandit', 'Megh Thakkar', 'Ramit Sawhney'] | null | null | null | null | acl-2022-5 | ['sentence-classification'] | ['natural-language-processing'] | [ 1.90441310e-01 -1.62917115e-02 -2.69770771e-01 -4.69731331e-01
-1.02312922e+00 -6.63785040e-01 1.04465401e+00 6.33886337e-01
-6.50157332e-01 4.26718742e-01 7.06862748e-01 -2.63620913e-01
1.01328857e-01 -5.27717292e-01 -4.86546099e-01 -4.27061498e-01
7.02610472e-03 2.10180208e-01 -2.33194754e-01 -3.04476619e-01
8.77465904e-02 1.88897446e-01 -1.30060554e+00 6.93226278e-01
9.11345780e-01 8.64085257e-01 -3.31349015e-01 7.49640465e-01
3.69997211e-02 6.21666849e-01 -3.20972145e-01 -7.03750908e-01
2.38196164e-01 -4.86171752e-01 -9.89467740e-01 -9.37367156e-02
7.36556172e-01 -2.73862351e-02 -5.56906164e-01 5.76640844e-01
8.38321209e-01 4.55435425e-01 7.77170539e-01 -1.11350322e+00
-1.16067982e+00 8.80386651e-01 -3.96744698e-01 1.69628218e-01
4.72016782e-01 3.20397504e-03 1.06350577e+00 -1.46527660e+00
6.70816898e-01 1.20976019e+00 1.09354997e+00 7.40791917e-01
-1.61266279e+00 -2.51630396e-01 2.27077808e-02 1.37151316e-01
-1.18376017e+00 -3.84145141e-01 7.87860394e-01 -3.18611652e-01
8.15887988e-01 5.03316462e-01 2.45714039e-01 1.45436144e+00
-2.41846472e-01 1.33484852e+00 1.20374763e+00 -7.56628990e-01
1.55589908e-01 3.98482352e-01 4.04409885e-01 5.06153941e-01
-2.77397335e-01 -2.38338977e-01 -8.34524274e-01 -1.99000597e-01
1.81568012e-01 -3.73564661e-02 -3.65926534e-01 -1.70847535e-01
-1.44210231e+00 9.69503522e-01 5.04311979e-01 1.00841776e-01
-1.26314446e-01 -9.34581012e-02 8.59911203e-01 3.38625818e-01
8.17929149e-01 7.08821833e-01 -4.42081600e-01 -3.33662897e-01
-8.44361901e-01 4.38515186e-01 4.34644580e-01 6.90778255e-01
3.82745087e-01 -2.98874788e-02 -7.12349892e-01 1.29877174e+00
9.41495970e-02 2.76724160e-01 6.15606487e-01 -7.86259711e-01
7.57144749e-01 6.76431417e-01 -2.09725723e-01 -6.38590813e-01
-4.71724987e-01 -5.36111772e-01 -9.57388043e-01 3.93534591e-03
3.91210556e-01 -1.24376722e-01 -7.02962637e-01 1.72371638e+00
3.94217581e-01 1.19943926e-02 1.81290627e-01 7.74525404e-01
9.53426659e-01 6.25700772e-01 1.23504937e-01 2.97107905e-01
1.24747908e+00 -1.14786565e+00 -4.24333423e-01 -2.06949562e-01
8.73482049e-01 -6.31743908e-01 1.73117304e+00 3.29475552e-01
-1.01070821e+00 -6.28306925e-01 -1.12935781e+00 -3.70860845e-01
-4.34324235e-01 3.19446027e-01 5.43963730e-01 7.03085423e-01
-8.66899729e-01 7.26499557e-01 -7.07091451e-01 -1.98035702e-01
4.95429844e-01 2.75110621e-02 -6.65011883e-01 -4.01259720e-01
-1.07957649e+00 9.14934933e-01 3.09879512e-01 -1.71313599e-01
-2.48339027e-01 -9.84801531e-01 -1.18440068e+00 -2.56057143e-01
-2.94422716e-01 -5.43321252e-01 9.86197352e-01 -6.63653314e-01
-1.35855532e+00 9.05411601e-01 -1.08420387e-01 -6.36382937e-01
5.79692066e-01 -2.43277401e-01 -3.47014189e-01 7.35155866e-02
-2.42417559e-01 8.75035524e-01 7.16272175e-01 -1.04422235e+00
-1.99826747e-01 -2.44428560e-01 -1.12632185e-01 4.04159456e-01
-9.12125468e-01 -2.04282567e-01 -2.91959852e-01 -9.57303703e-01
-5.40260300e-02 -8.33517373e-01 -1.42237142e-01 1.08806007e-01
-4.24722195e-01 -2.82374293e-01 7.50806570e-01 -6.97139859e-01
1.21770287e+00 -2.20847845e+00 4.94773954e-01 -3.76803800e-02
6.28929958e-02 2.16212422e-01 -5.66966057e-01 4.69380409e-01
7.97149912e-02 -1.68403313e-02 -3.74010026e-01 -1.02447283e+00
3.47892284e-01 3.54060046e-02 -2.12707803e-01 3.31794560e-01
4.11608517e-01 9.00597751e-01 -8.78704250e-01 -2.21657440e-01
2.64593422e-01 8.26156318e-01 -6.19653761e-01 5.71199059e-02
9.46756452e-03 2.46879369e-01 1.60498470e-01 4.61912394e-01
7.17100382e-01 -2.24906176e-01 -1.23359554e-01 -3.41736287e-01
3.03064197e-01 4.27900612e-01 -9.74252760e-01 1.78837979e+00
-5.54634213e-01 1.09726286e+00 -3.67007792e-01 -8.51028740e-01
8.57950866e-01 2.21012533e-01 1.73784941e-01 -5.23065150e-01
1.38617039e-01 9.19886827e-02 -3.71526301e-01 -3.68394434e-01
9.61080909e-01 1.11547947e-01 -8.30585808e-02 5.16182184e-01
3.27416211e-01 4.06536907e-02 2.73927450e-01 3.47902387e-01
9.90110457e-01 8.43561813e-02 9.34282467e-02 -1.46510914e-01
3.12929779e-01 -3.62813532e-01 2.65217815e-02 7.33602464e-01
-9.83319581e-02 1.05802834e+00 1.90144002e-01 -1.85030267e-01
-1.20897758e+00 -1.10777068e+00 -5.66245079e-01 1.51095974e+00
-2.55913764e-01 -5.21367669e-01 -6.14062548e-01 -9.95179772e-01
1.56982750e-01 7.67335296e-01 -1.07932162e+00 -2.55404294e-01
-2.91003644e-01 -1.01892018e+00 7.17553318e-01 8.11094701e-01
3.81772280e-01 -9.28784549e-01 -1.71124145e-01 7.69837499e-02
-2.01459229e-01 -1.23731518e+00 -5.59124827e-01 2.49602184e-01
-7.78032541e-01 -7.84584641e-01 -7.90548980e-01 -5.41312695e-01
7.72830427e-01 -3.47685954e-03 1.20046163e+00 -9.98925194e-02
-5.57028176e-03 5.81537068e-01 -7.48905718e-01 -2.41424650e-01
-5.45988262e-01 3.45014811e-01 9.32350010e-02 7.40386471e-02
3.55408579e-01 -1.13756932e-01 -5.30665517e-01 -4.21550795e-02
-9.78603363e-01 6.75926134e-02 2.45574579e-01 1.18211401e+00
3.74658793e-01 -6.56634331e-01 5.19286215e-01 -1.04319739e+00
9.30424929e-01 -6.65501654e-01 1.19492061e-01 1.82004958e-01
-6.42315090e-01 1.80191725e-01 6.61575615e-01 -7.23046720e-01
-6.40377879e-01 -1.63023606e-01 -3.99867654e-01 -7.46066198e-02
3.17435991e-03 6.28197312e-01 1.58689320e-01 4.97890934e-02
1.08719802e+00 2.40872979e-01 1.42050609e-01 -5.42735338e-01
7.58455217e-01 7.84680009e-01 6.06847107e-01 -5.46496153e-01
6.95205331e-01 2.74031490e-01 -1.79787636e-01 -8.69252443e-01
-8.44428122e-01 -3.29578787e-01 -6.43187046e-01 8.34468678e-02
5.24469197e-01 -8.18573117e-01 -3.84651870e-01 4.15407687e-01
-9.22107399e-01 -4.03804839e-01 -5.52761376e-01 5.45663893e-01
-1.96813568e-01 3.87133271e-01 -8.39981794e-01 -6.38868093e-01
-5.50959349e-01 -8.22841108e-01 1.01098859e+00 1.45076364e-01
-4.60121214e-01 -1.40041721e+00 1.29575402e-01 2.77979970e-01
4.96687829e-01 2.71865696e-01 6.77893281e-01 -9.67401981e-01
1.67208299e-01 -4.54541475e-01 -1.43507034e-01 6.21748924e-01
1.04080036e-01 5.90482987e-02 -1.28532898e+00 -4.91589367e-01
-4.94954854e-01 -6.57148600e-01 1.03922653e+00 1.32707611e-01
1.11140633e+00 -4.28167999e-01 1.18623167e-01 5.31932950e-01
1.09116948e+00 -4.81788039e-01 5.01057684e-01 5.10983348e-01
6.21638834e-01 4.91834790e-01 4.13516968e-01 5.62499344e-01
4.64614630e-01 7.53635228e-01 2.26644963e-01 -3.09363484e-01
-1.95140958e-01 -1.72734931e-01 3.41278195e-01 8.87458563e-01
2.54716992e-01 -1.58507586e-01 -7.44607151e-01 5.95511496e-01
-1.65639496e+00 -9.89496052e-01 2.86165662e-02 2.07889915e+00
1.23186171e+00 1.36675969e-01 4.87700492e-01 5.34604788e-01
3.54953408e-01 9.99801904e-02 -3.04044127e-01 -4.96566772e-01
-4.53440577e-01 3.71988654e-01 2.38533661e-01 5.69167972e-01
-1.26831985e+00 4.38757420e-01 6.97090864e+00 7.17501044e-01
-1.16622365e+00 2.00204372e-01 8.40969026e-01 -1.71564788e-01
-5.79083145e-01 -4.74024922e-01 -7.10942745e-01 4.97608662e-01
9.07453597e-01 1.42493486e-01 3.72470796e-01 7.33982444e-01
-1.89308092e-01 2.70269722e-01 -1.22115183e+00 9.24089670e-01
3.84727150e-01 -1.68105888e+00 -1.01663336e-01 -3.22774172e-01
9.18663740e-01 2.52097577e-01 5.21813869e-01 5.29212177e-01
1.94335610e-01 -1.09653902e+00 7.67433286e-01 2.60470241e-01
7.26760209e-01 -7.47623324e-01 8.21961641e-01 1.42044038e-01
-9.02988017e-01 -2.32799463e-02 -2.50488520e-01 4.60088626e-02
-8.55749324e-02 4.04820263e-01 -8.17423403e-01 5.51256359e-01
5.21689117e-01 7.25654900e-01 -1.11176407e+00 8.64817679e-01
-1.71129052e-02 8.21917355e-01 -2.92048335e-01 -1.71321273e-01
2.00287923e-01 6.09192587e-02 3.38568032e-01 1.64476609e+00
4.52766232e-02 -4.34186637e-01 -9.59910899e-02 4.87159312e-01
-2.56843179e-01 2.58955717e-01 -4.17743325e-01 -4.47955467e-02
8.15624237e-01 1.17700493e+00 -3.39344800e-01 -2.98772842e-01
-3.54725540e-01 1.14216638e+00 3.96224320e-01 1.95547715e-01
-6.40051544e-01 -5.98796844e-01 5.97583592e-01 -1.97384674e-02
9.29984003e-02 -2.23092422e-01 -6.28917158e-01 -1.10917139e+00
1.01532742e-01 -1.11665797e+00 5.06695867e-01 -4.91888523e-01
-1.69283712e+00 9.29919064e-01 -6.26107529e-02 -1.45526266e+00
-2.80219376e-01 -7.45839596e-01 -4.32595462e-01 7.25531816e-01
-1.29504371e+00 -1.30411994e+00 -4.02233481e-01 3.66352350e-01
5.56385100e-01 -2.97551334e-01 1.12267220e+00 2.82134593e-01
-5.68701446e-01 1.17565978e+00 4.54981953e-01 3.91077399e-01
7.56768942e-01 -1.52879393e+00 6.47144377e-01 7.58294642e-01
5.29844522e-01 5.80872059e-01 6.61591530e-01 1.89384539e-02
-1.20477974e+00 -1.05465722e+00 8.04004073e-01 -8.02446663e-01
3.95837575e-01 -6.50533676e-01 -9.73622799e-01 6.20889664e-01
5.52501142e-01 3.89641166e-01 9.49924350e-01 4.06995833e-01
-8.38483572e-01 -6.76523298e-02 -1.22513378e+00 7.19579041e-01
6.62151754e-01 -8.43630612e-01 -3.09205323e-01 3.29594731e-01
6.61350071e-01 -5.79205155e-01 -1.07833743e+00 2.95259714e-01
4.56590772e-01 -7.66601026e-01 1.09999394e+00 -7.90320218e-01
6.84826732e-01 6.57145381e-02 -4.18506235e-01 -1.54828942e+00
-2.90368319e-01 -6.12725914e-01 -4.26702708e-01 1.40270197e+00
6.40812755e-01 -5.21419287e-01 7.59829223e-01 6.03809237e-01
-8.63464400e-02 -1.19102204e+00 -9.63173449e-01 -6.30564213e-01
3.63308221e-01 -6.32215858e-01 5.52534282e-01 1.17190075e+00
3.59018505e-01 4.80743229e-01 -3.79712075e-01 -2.95856923e-01
5.18543899e-01 -3.05473417e-01 7.58857310e-01 -8.79757881e-01
-3.44080687e-01 -5.52717268e-01 -3.65638465e-01 -9.99368072e-01
1.09935530e-01 -1.27837253e+00 -2.43169755e-01 -1.59117949e+00
7.88660422e-02 -5.99023938e-01 -3.27481508e-01 6.06257498e-01
-5.09386718e-01 7.04999149e-01 3.52698922e-01 -4.91625480e-02
-4.74684417e-01 7.73873806e-01 8.59544218e-01 -3.05249542e-01
-2.21121565e-01 -1.78048223e-01 -7.87056863e-01 3.77716780e-01
7.30735838e-01 -1.31513104e-01 -3.09178770e-01 -5.81075430e-01
1.90521315e-01 -5.46836317e-01 1.36950895e-01 -1.02760875e+00
-1.31081015e-01 1.85524642e-01 6.03415370e-01 -3.33677232e-01
5.33795536e-01 -4.49308544e-01 -4.39258307e-01 2.97551900e-01
-8.94234598e-01 2.11749688e-01 2.68302083e-01 3.05888295e-01
-1.31482303e-01 -2.36199006e-01 9.35088456e-01 4.55329329e-01
-4.47075725e-01 2.21059486e-01 -1.56672105e-01 4.59306479e-01
7.28080630e-01 -1.15455776e-01 -4.57457393e-01 -2.97733188e-01
-6.59456730e-01 5.24597615e-02 3.26045603e-01 5.99195659e-01
6.97853148e-01 -1.83947682e+00 -1.04514718e+00 2.03726172e-01
3.50289464e-01 -2.44991973e-01 8.74900222e-02 8.09299886e-01
-3.49083930e-01 9.04064551e-02 1.02453016e-01 -7.07841277e-01
-1.39732420e+00 3.50765258e-01 1.23970665e-01 -2.22484559e-01
-5.66130877e-01 9.88445401e-01 -4.86789972e-01 -9.58880603e-01
2.93331861e-01 -3.53754401e-01 -1.14992768e-01 2.26865441e-01
6.47463620e-01 3.15593749e-01 2.48266965e-01 -4.70871240e-01
-3.06729019e-01 2.28168696e-01 -1.98016629e-01 -2.62024879e-01
1.28217375e+00 5.84821776e-02 2.91359007e-01 5.72437882e-01
1.56450975e+00 1.50794432e-01 -1.32363021e+00 -6.08801484e-01
-7.41958339e-03 -5.04654527e-01 -2.24714354e-01 -8.72072816e-01
-7.42589772e-01 8.98446381e-01 6.06966257e-01 2.61098832e-01
1.01559019e+00 -9.97853950e-02 7.09955454e-01 2.52188385e-01
-1.90188482e-01 -1.06938732e+00 3.29807132e-01 5.60053229e-01
8.94851685e-01 -1.47245395e+00 9.36797634e-02 -2.21029036e-02
-1.03118634e+00 1.02588165e+00 4.83515829e-01 -2.54434291e-02
4.89176482e-01 1.46929592e-01 1.83339685e-01 4.31362003e-01
-7.79829919e-01 1.33791402e-01 6.35767817e-01 6.79756045e-01
8.55554402e-01 -7.81073496e-02 -2.01694652e-01 5.63807786e-01
-2.06384003e-01 -3.72098923e-01 3.57717305e-01 8.91492426e-01
1.28387744e-02 -1.21734297e+00 -1.60406560e-01 5.14783740e-01
-3.44955385e-01 -1.92986459e-01 -2.07899407e-01 7.06379354e-01
-4.59854752e-02 7.35266268e-01 1.33336455e-01 -6.33798242e-01
3.53744835e-01 1.70196220e-01 6.66303396e-01 -4.44268554e-01
-8.22181165e-01 -4.67767715e-01 1.79462746e-01 -8.01487043e-02
-2.31921941e-01 -7.89436877e-01 -1.15901744e+00 -3.07764888e-01
-1.19586587e-01 7.18721300e-02 7.18598545e-01 9.26243663e-01
5.04102230e-01 3.17812532e-01 8.05294037e-01 -8.30468953e-01
-7.76600420e-01 -1.36146200e+00 -1.60083354e-01 9.20298815e-01
5.16554773e-01 -3.46117646e-01 -2.59104729e-01 -1.17269181e-01] | [10.76498031616211, 8.40109634399414] |
6ee667c6-d6f3-4cfc-aa7a-066d08ffe087 | understanding-the-effect-of-the-long-tail-on | 2306.06238 | null | https://arxiv.org/abs/2306.06238v3 | https://arxiv.org/pdf/2306.06238v3.pdf | Understanding the Effect of the Long Tail on Neural Network Compression | Network compression is now a mature sub-field of neural network research: over the last decade, significant progress has been made towards reducing the size of models and speeding up inference, while maintaining the classification accuracy. However, many works have observed that focusing on just the overall accuracy can be misguided. E.g., it has been shown that mismatches between the full and compressed models can be biased towards under-represented classes. This raises the important research question, can we achieve network compression while maintaining "semantic equivalence" with the original network? In this work, we study this question in the context of the "long tail" phenomenon in computer vision datasets observed by Feldman, et al. They argue that memorization of certain inputs (appropriately defined) is essential to achieving good generalization. As compression limits the capacity of a network (and hence also its ability to memorize), we study the question: are mismatches between the full and compressed models correlated with the memorized training data? We present positive evidence in this direction for image classification tasks, by considering different base architectures and compression schemes. | ['Ganesh Gopalakrishnan', 'Michael Garland', 'Saurav Muralidharan', 'Aditya Bhaskara', 'Vinu Joseph', 'Harvey Dam'] | 2023-06-09 | null | null | null | null | ['neural-network-compression', 'neural-network-compression', 'memorization'] | ['methodology', 'miscellaneous', 'natural-language-processing'] | [ 8.65128875e-01 2.45224237e-01 -2.77609766e-01 -4.32650387e-01
1.27822235e-01 -2.52309322e-01 5.87382317e-01 4.18776572e-01
-7.85567224e-01 7.62122631e-01 8.87369365e-02 -3.23377848e-01
-4.97349381e-01 -9.20214295e-01 -8.83690536e-01 -8.01662266e-01
8.89183953e-02 3.69410396e-01 2.28457171e-02 5.12334555e-02
4.87323076e-01 5.72969675e-01 -1.97904432e+00 4.39409733e-01
6.40700102e-01 1.03216612e+00 3.77027720e-01 5.74592412e-01
8.14992264e-02 9.74807203e-01 -5.80106318e-01 -5.22449732e-01
1.12424165e-01 -6.12991393e-01 -1.11107349e+00 -9.64147076e-02
7.29690492e-01 -9.89320874e-02 -4.82756466e-01 1.35415995e+00
1.00824311e-01 1.07020080e-01 4.12850767e-01 -1.20749962e+00
-7.77959406e-01 8.14146459e-01 -5.40879630e-02 5.73709249e-01
-3.20901573e-01 -1.63471758e-01 8.50349128e-01 -3.79069120e-01
5.48399687e-01 1.04765606e+00 6.14136755e-01 5.96073151e-01
-1.43346119e+00 -5.63211203e-01 7.46432692e-02 5.32446325e-01
-1.32741535e+00 -6.60889506e-01 6.29874051e-01 -2.32665241e-01
1.02639043e+00 5.05913615e-01 8.44038725e-01 9.06402588e-01
2.04326376e-01 5.61165273e-01 9.82160211e-01 -6.96781874e-01
4.09539998e-01 3.98633838e-01 3.06492567e-01 5.71175814e-01
7.70267069e-01 6.21212795e-02 -6.98559999e-01 2.06525549e-01
7.07683146e-01 -3.74094509e-02 -5.22015989e-01 -1.79984763e-01
-7.22309232e-01 9.11070943e-01 5.37039399e-01 7.11400032e-01
-3.94170225e-01 1.18712708e-01 3.48379344e-01 7.23248303e-01
4.67119515e-01 7.53938675e-01 -3.68567944e-01 -8.61383006e-02
-1.17871451e+00 1.71909764e-01 8.67275894e-01 6.94203973e-01
5.20120859e-01 1.28916904e-01 2.03578144e-01 7.53876865e-01
-2.09820122e-02 5.54819219e-02 9.66899633e-01 -1.07610035e+00
3.88094306e-01 4.52641755e-01 -4.64966774e-01 -1.26184118e+00
-2.43008062e-01 -1.06883693e+00 -1.24414754e+00 -1.91336852e-02
4.52991158e-01 2.23063007e-01 -4.97627348e-01 2.15696049e+00
-3.20535302e-01 -1.05592027e-01 1.34294495e-01 5.16441166e-01
3.29869956e-01 3.21468234e-01 1.41087726e-01 -3.07092786e-01
9.36794162e-01 -7.60532975e-01 -5.68865895e-01 -6.05743825e-01
7.70105124e-01 -2.92761654e-01 9.03757870e-01 4.03966814e-01
-1.40554774e+00 -6.90058112e-01 -1.33700383e+00 1.02331489e-01
-4.77735698e-01 -2.29369357e-01 4.94151950e-01 7.24944532e-01
-1.26012552e+00 1.16229665e+00 -5.84415376e-01 -4.59085882e-01
6.93360507e-01 4.29857224e-01 -3.98374736e-01 -4.10905093e-01
-1.09927964e+00 1.27350736e+00 6.16422057e-01 -1.08622126e-01
-5.04521310e-01 -5.52883744e-01 -5.48456430e-01 4.72595870e-01
1.21138439e-01 -7.70680308e-01 1.02832234e+00 -1.31088281e+00
-8.59504104e-01 8.96018326e-01 -1.25829831e-01 -1.08896422e+00
2.57780641e-01 -7.08653918e-03 -2.20333427e-01 1.55672997e-01
-4.75247681e-01 8.15120876e-01 6.87944591e-01 -1.12066114e+00
-5.27644038e-01 -8.28467071e-01 -3.15610855e-03 1.27703875e-01
-8.20592403e-01 -3.58740836e-01 3.06618717e-02 -6.86408758e-01
3.29462469e-01 -8.99317682e-01 1.65313277e-02 5.30277528e-02
-3.09975762e-02 -1.22246392e-01 5.50771356e-01 -4.67048466e-01
1.18604493e+00 -2.14214373e+00 1.47925779e-01 2.31858030e-01
3.73767912e-01 5.16505599e-01 -2.35843748e-01 1.53595626e-01
-2.54530281e-01 4.12229747e-01 -3.99275273e-01 -3.82748574e-01
-2.01781541e-01 6.13009930e-01 -4.34144288e-01 4.89773482e-01
2.71334890e-02 7.50327528e-01 -4.97032106e-01 -3.57939571e-01
-7.70345554e-02 6.12095058e-01 -5.52671015e-01 -1.38658434e-01
7.74682909e-02 6.36747852e-03 1.28299788e-01 1.06894463e-01
5.35273314e-01 -3.75800282e-01 2.43620306e-01 -1.16738796e-01
1.42579585e-01 3.29266995e-01 -8.32959771e-01 1.38799584e+00
-3.07348162e-01 1.15986919e+00 -1.32734373e-01 -1.66469407e+00
7.26804733e-01 3.43614012e-01 1.48852795e-01 -1.02171564e+00
2.00259551e-01 9.84777883e-02 4.31638151e-01 -2.47049257e-01
5.60136259e-01 -2.73464799e-01 5.26330948e-01 5.41115582e-01
1.57655805e-01 6.32834435e-02 3.13289106e-01 -3.74993263e-03
8.31134081e-01 -4.45379138e-01 2.42951080e-01 -2.19847172e-01
2.14957878e-01 -4.81333118e-03 1.32412434e-01 1.08067453e+00
-1.15457140e-01 3.63287300e-01 4.01847094e-01 -2.73529768e-01
-1.21965075e+00 -7.37560332e-01 -3.34189355e-01 1.01688552e+00
-1.31287932e-01 -2.42342055e-01 -1.06105494e+00 -1.38960153e-01
-1.32660151e-01 8.44226539e-01 -8.56667340e-01 -7.90061831e-01
-4.83535349e-01 -8.06503415e-01 8.82661700e-01 6.26249015e-01
7.70507336e-01 -1.01099694e+00 -1.09373355e+00 4.05357331e-02
-1.43854231e-01 -9.86382544e-01 2.04006016e-01 4.84339058e-01
-1.59328854e+00 -9.54709172e-01 -4.75209832e-01 -7.15402186e-01
8.44810128e-01 2.90057153e-01 1.25459838e+00 6.94933116e-01
1.60747096e-02 1.83561109e-02 -2.37561032e-01 -5.48103094e-01
-6.49312615e-01 3.09657156e-01 -1.11343704e-01 -2.64939606e-01
4.22500461e-01 -8.02913487e-01 -4.16789770e-01 -2.55184025e-02
-1.36971366e+00 2.59540260e-01 8.21207285e-01 6.86515033e-01
2.47505277e-01 3.56397808e-01 6.54799044e-01 -9.31820154e-01
8.17203343e-01 -3.29727739e-01 -2.45001301e-01 1.92598835e-01
-1.15331173e+00 3.11167061e-01 7.33000934e-01 -4.24350590e-01
-6.38472021e-01 -3.15946281e-01 -1.92926675e-01 -1.33111417e-01
-1.99880362e-01 6.56448543e-01 1.06796026e-01 -2.09615335e-01
6.45836353e-01 6.15974605e-01 4.08304483e-01 -4.35169637e-01
1.27276257e-01 4.44506049e-01 6.10640585e-01 -2.76994288e-01
3.60542476e-01 4.28447813e-01 1.24370404e-01 -8.87326896e-01
-8.12567949e-01 -7.75421634e-02 -6.07503772e-01 -2.59897485e-02
3.54784548e-01 -4.78925258e-01 -4.63241518e-01 1.75449327e-01
-1.21149015e+00 -8.17559361e-02 -6.54600143e-01 4.20541584e-01
-5.75586438e-01 1.79573178e-01 -5.89634299e-01 -5.71140110e-01
-2.17755660e-01 -8.97042394e-01 2.86711574e-01 8.19060877e-02
-4.61889565e-01 -1.15970469e+00 -2.33410329e-01 8.35922360e-02
8.26128364e-01 -2.31437311e-01 1.37388015e+00 -8.43558490e-01
-2.45779499e-01 -3.33577693e-01 -3.55717182e-01 7.57307708e-01
-2.99425572e-01 -4.32296157e-01 -1.19042695e+00 -2.81730860e-01
5.22922158e-01 -2.92642921e-01 1.20337081e+00 3.45825821e-01
1.46029651e+00 -7.04850256e-01 -1.45879284e-01 3.71417284e-01
1.51375782e+00 1.84280977e-01 7.61202574e-01 3.69424820e-01
1.80820763e-01 8.27672184e-01 -4.99566421e-02 1.31622121e-01
-2.64484528e-03 3.23861182e-01 5.09551764e-01 2.67967463e-01
-2.67881811e-01 -3.36444914e-01 -1.06445082e-01 9.31942344e-01
-1.38866439e-01 -4.08254564e-01 -6.67022705e-01 4.93735611e-01
-1.57098842e+00 -1.08920979e+00 2.42413566e-01 2.14464307e+00
8.35651815e-01 3.14977527e-01 -3.32555979e-01 9.07962441e-01
5.04564583e-01 1.81252256e-01 -4.67941999e-01 -3.64320219e-01
-3.07719827e-01 3.15651834e-01 4.95285928e-01 4.44271564e-01
-5.65299630e-01 5.36627650e-01 6.58078384e+00 8.72579575e-01
-1.30998504e+00 1.61195412e-01 8.11330914e-01 -1.53890237e-01
-2.06327066e-01 -1.16572402e-01 -5.63115180e-01 3.97970498e-01
1.39101660e+00 -3.15730065e-01 4.98799741e-01 6.66681230e-01
-2.19432876e-01 -1.93259284e-01 -1.34787929e+00 9.37418222e-01
3.92452359e-01 -1.53220749e+00 4.36257899e-01 1.34434193e-01
6.35151327e-01 -3.95568367e-03 3.47142547e-01 1.61212072e-01
-1.18239857e-01 -1.37793112e+00 6.96398199e-01 5.40519357e-01
4.39153969e-01 -6.67955637e-01 7.94410944e-01 7.94105768e-01
-6.08049214e-01 -2.94984668e-01 -7.79431462e-01 -3.40127528e-01
-3.02681416e-01 6.56080306e-01 -6.10434294e-01 1.45401791e-01
5.25348365e-01 5.32312155e-01 -8.91945124e-01 1.04546607e+00
1.13008954e-01 7.82283604e-01 -2.32427418e-01 -4.82296012e-02
8.88800845e-02 8.99993256e-02 2.13553101e-01 1.00315714e+00
2.42428154e-01 -2.37587422e-01 -4.15485859e-01 7.49455929e-01
-3.45844895e-01 -8.68737996e-02 -6.98593915e-01 -1.77667439e-01
5.25275230e-01 6.75430834e-01 -8.61931682e-01 -5.61341882e-01
-5.69455978e-03 6.96794629e-01 4.62263852e-01 2.27900550e-01
-3.36078972e-01 -1.87879086e-01 1.63081184e-01 1.34280905e-01
1.31710470e-02 -8.53131041e-02 -8.34375620e-01 -7.89058983e-01
-1.73742627e-03 -8.38174760e-01 2.75431663e-01 -5.89640617e-01
-7.92935073e-01 6.07751012e-01 1.79250911e-02 -6.96369290e-01
-1.79164872e-01 -5.67162037e-01 -3.08098018e-01 6.59580052e-01
-1.41772139e+00 -4.39465612e-01 -2.16159865e-01 3.76236677e-01
5.55602074e-01 -8.40529352e-02 1.01196694e+00 4.14208978e-01
-3.87837261e-01 5.85027635e-01 2.85037875e-01 -2.93499902e-02
1.95091695e-01 -7.29702175e-01 -2.88980315e-03 6.46293461e-01
5.07054150e-01 7.44293749e-01 8.84520113e-01 -3.52787822e-01
-1.16906989e+00 -1.03841150e+00 1.21807420e+00 -1.93827704e-01
1.68640018e-01 1.85258761e-02 -1.19595790e+00 6.73196375e-01
2.02898011e-01 -2.81563282e-01 6.89352930e-01 -5.08246869e-02
-4.47382003e-01 -1.06552348e-01 -1.22686064e+00 4.33647871e-01
1.07722044e+00 -5.83040833e-01 -6.32376373e-01 2.22689524e-01
5.92537463e-01 1.37001485e-01 -6.65075123e-01 3.24613124e-01
5.68900228e-01 -1.12119269e+00 8.80275667e-01 -8.04663837e-01
7.00851858e-01 3.06615412e-01 -3.88177872e-01 -1.29219377e+00
-1.85608983e-01 2.38574833e-01 -3.29755485e-01 7.89214969e-01
2.57769585e-01 -6.63105607e-01 1.09540033e+00 5.36582649e-01
5.62729239e-02 -9.16741073e-01 -1.04980981e+00 -8.54882121e-01
2.50396401e-01 -5.66347003e-01 5.17275512e-01 9.62066770e-01
-4.78758290e-02 2.78068155e-01 -1.35173291e-01 -2.43868694e-01
5.24818957e-01 -1.79357499e-01 1.36048153e-01 -1.57835364e+00
-2.28482142e-01 -8.02369177e-01 -5.66966474e-01 -1.06850171e+00
1.60357490e-01 -1.25864577e+00 -2.28207737e-01 -1.23807955e+00
2.92354882e-01 -3.15851748e-01 -4.06660706e-01 2.23449245e-01
3.61681551e-01 4.60410684e-01 4.19287354e-01 5.00773013e-01
-4.19133723e-01 1.80213153e-01 1.03050518e+00 -9.81095955e-02
3.35200906e-01 -5.34870774e-02 -1.04655468e+00 7.77732253e-01
1.01739597e+00 -6.16984487e-01 -6.28515899e-01 -7.27021575e-01
4.76001173e-01 -1.20301619e-01 5.93425751e-01 -1.47535646e+00
5.77959836e-01 1.59995809e-01 4.41525489e-01 -2.43458629e-01
4.27030087e-01 -9.50471640e-01 -1.94392912e-03 1.01013601e+00
-1.01326835e+00 3.61408710e-01 8.94736648e-02 6.22919977e-01
-3.29697102e-01 -7.95886338e-01 9.85001802e-01 -3.02911758e-01
-3.39816719e-01 1.73511371e-01 -2.62194216e-01 4.45101373e-02
6.84391618e-01 -4.27686185e-01 -4.46647882e-01 -3.16182256e-01
-6.65790081e-01 -3.24534655e-01 3.91040295e-01 3.05063009e-01
7.61773586e-01 -9.77917016e-01 -5.14226615e-01 3.45882028e-01
-1.55115470e-01 -1.48282766e-01 1.46656439e-01 7.33112991e-01
-4.33857709e-01 7.81150281e-01 -2.80296385e-01 -3.65655392e-01
-1.37735546e+00 4.96097863e-01 3.38530570e-01 -1.66420400e-01
-4.09496933e-01 8.91292632e-01 -2.88827926e-01 5.86621128e-02
6.10162377e-01 -3.11768055e-01 -2.10051522e-01 5.40240891e-02
5.96247971e-01 4.16831255e-01 2.59354591e-01 -3.95440310e-01
7.98856393e-02 1.94428071e-01 -2.64251083e-01 8.24764557e-03
1.43652701e+00 1.98898483e-02 -2.76343822e-01 5.33934593e-01
1.44684100e+00 -5.54841459e-01 -7.87327766e-01 -4.47491676e-01
1.79854482e-01 -4.32762355e-01 9.09241438e-02 -5.38417816e-01
-1.08991432e+00 1.23237479e+00 6.12619102e-01 6.79152727e-01
1.09559178e+00 -1.27225757e-01 4.58547533e-01 7.59034038e-01
2.31528580e-01 -1.10515344e+00 2.63018049e-02 5.88360846e-01
8.29535365e-01 -8.87074828e-01 -4.58257608e-02 -1.33503318e-01
-1.84156150e-01 1.07520735e+00 4.24982131e-01 -1.87407225e-01
6.10179901e-01 -5.48472442e-02 -4.95818675e-01 -1.73962802e-01
-9.81650949e-01 2.05037549e-01 -6.50059571e-03 5.48224866e-01
4.46531594e-01 -1.64661989e-01 -2.58655280e-01 3.71632189e-01
-7.36056685e-01 2.01781332e-01 5.46941280e-01 1.01929140e+00
-7.84037530e-01 -7.67019212e-01 -2.38126293e-01 9.14669275e-01
-5.24759054e-01 -3.26912820e-01 -3.00879478e-01 7.11358786e-01
2.11016655e-01 9.33432281e-01 4.15586025e-01 -4.40839738e-01
7.14882761e-02 3.61073941e-01 7.33509600e-01 -4.60028738e-01
-4.10823405e-01 -7.77800381e-01 -1.26785636e-01 -2.04389066e-01
-6.01509333e-01 -3.39300036e-01 -7.64754236e-01 -6.56446338e-01
-3.12721282e-01 3.64407599e-01 7.66036391e-01 1.22226799e+00
2.26527184e-01 6.28607988e-01 1.46891013e-01 -5.16851008e-01
-9.59453225e-01 -9.74245906e-01 -4.59440768e-01 5.02551317e-01
2.84471124e-01 -3.48788023e-01 -6.12615645e-01 2.51100808e-02] | [8.469149589538574, 3.453639268875122] |
7a0ae689-015e-4687-b9a0-8cd38948e842 | a-multi-level-contextual-model-for-person | null | null | http://openaccess.thecvf.com/content_cvpr_2016/html/Li_A_Multi-Level_Contextual_CVPR_2016_paper.html | http://openaccess.thecvf.com/content_cvpr_2016/papers/Li_A_Multi-Level_Contextual_CVPR_2016_paper.pdf | A Multi-Level Contextual Model For Person Recognition in Photo Albums | In this work, we present a new framework for person recognition in photo albums that exploits contextual cues at multiple levels, spanning individual persons, individual photos, and photo groups. Through experiments, we show that the information available at each of these distinct contextual levels provides complementary cues as to person identities. At the person level, we leverage clothing and body appearance in addition to facial appearance, and to compensate for instances where the faces are not visible. At the photo level we leverage a learned prior on the joint distribution of identities on the same photo to guide the identity assignments. Going beyond a single photo, we are able to infer natural groupings of photos with shared context in an unsupervised manner. By exploiting this shared contextual information, we are able to reduce the identity search space and exploit higher intra-personal appearance consistency within photo groups. Our new framework enables efficient use of these complementary multi-level contextual cues to improve overall recognition rates on the photo album person recognition task, as demonstrated through state-of-the-art results on a challenging public dataset. Our results outperform competing methods by a significant margin, while being computationally efficient and practical in a real world application. | ['Xiaohui Shen', 'Haoxiang Li', 'Zhe Lin', 'Jonathan Brandt', 'Gang Hua'] | 2016-06-01 | null | null | null | cvpr-2016-6 | ['person-recognition'] | ['computer-vision'] | [ 4.48732108e-01 -3.29037845e-01 -5.12163676e-02 -5.91704667e-01
-6.16615891e-01 -7.57876158e-01 8.55785787e-01 8.33679065e-02
-2.67617136e-01 4.78039503e-01 4.56526875e-01 5.84321618e-01
6.96020201e-02 -5.74605048e-01 -5.96913278e-01 -6.85873508e-01
2.15354949e-01 1.76182181e-01 -2.32053310e-01 5.25920950e-02
8.62698257e-03 5.21708786e-01 -1.95569968e+00 3.55367422e-01
4.54163253e-01 1.01690578e+00 -2.77395666e-01 4.37015116e-01
3.75392675e-01 5.00187218e-01 -2.76088834e-01 -8.30483854e-01
7.42109895e-01 -2.71384031e-01 -5.56899190e-01 7.59715915e-01
1.56092322e+00 -3.36630791e-01 -3.39818180e-01 1.04137945e+00
5.29736578e-01 2.90349007e-01 5.62002659e-01 -1.12477589e+00
-6.46097064e-01 7.10103363e-02 -7.17954040e-01 -2.05622032e-01
7.37412930e-01 2.11328059e-01 1.07400990e+00 -9.00083363e-01
6.74174726e-01 1.27724159e+00 1.01524067e+00 5.06663263e-01
-1.59764850e+00 -4.03376251e-01 5.31840503e-01 1.97051942e-01
-1.68727911e+00 -8.66186500e-01 6.83506608e-01 -5.30851364e-01
6.00905061e-01 3.14128280e-01 5.91009498e-01 1.08902669e+00
-5.82904518e-01 7.07203388e-01 1.25375259e+00 -6.15176141e-01
-1.07022226e-01 9.28510875e-02 1.10117465e-01 6.71360373e-01
1.38879955e-01 -4.62202216e-03 -8.17618906e-01 -2.54558653e-01
5.82124352e-01 3.34543169e-01 -1.86892584e-01 -4.38092798e-01
-9.85116303e-01 3.65050584e-01 6.10703826e-01 7.15110973e-02
-3.77153486e-01 -3.35639305e-02 3.31602804e-02 -6.35991246e-02
4.66049522e-01 1.93931684e-01 4.10731323e-02 2.08552003e-01
-9.68241870e-01 1.31757423e-01 5.79747796e-01 8.39250088e-01
1.18424451e+00 -2.68644661e-01 -3.82324994e-01 1.24871862e+00
7.57369325e-02 6.67108178e-01 9.05399844e-02 -9.13521409e-01
3.42070103e-01 6.74607992e-01 3.09341729e-01 -1.17150581e+00
-1.18666030e-01 -3.28781933e-01 -7.43367493e-01 1.79409131e-01
6.28858864e-01 4.75341193e-02 -1.11688209e+00 2.00446057e+00
5.09912550e-01 4.01636064e-01 -1.77436605e-01 9.72924411e-01
5.57123065e-01 6.29769936e-02 1.47142172e-01 1.54599622e-01
1.71160352e+00 -8.67214859e-01 -1.95409924e-01 -5.53297877e-01
2.20777228e-01 -7.43309498e-01 6.41394079e-01 1.81701034e-01
-9.29036081e-01 -7.59618998e-01 -9.73826766e-01 -1.34277046e-01
-2.67272294e-01 6.67032719e-01 4.91735905e-01 8.96003783e-01
-1.26150203e+00 4.58965272e-01 -4.84637797e-01 -7.41395831e-01
4.67663258e-01 4.46800023e-01 -7.59049594e-01 -5.49814641e-01
-7.39911079e-01 5.65131307e-01 -1.77454919e-01 1.84963286e-01
-5.11746109e-01 -6.70268118e-01 -1.09466279e+00 -1.91936195e-01
2.90525317e-01 -1.00123286e+00 8.24765384e-01 -1.30268478e+00
-1.24219966e+00 1.25565457e+00 -5.08490920e-01 -3.80416751e-01
6.01701975e-01 -1.69811606e-01 -3.47748101e-01 3.82284820e-01
1.51776791e-01 7.34057903e-01 9.38934505e-01 -1.52585900e+00
-6.01852655e-01 -6.67753935e-01 7.50478730e-02 4.52148169e-01
-5.01693904e-01 3.86945084e-02 -8.08870375e-01 -5.80686808e-01
4.23717797e-02 -1.32034016e+00 5.70951379e-05 1.79320380e-01
-3.69288236e-01 1.28416913e-02 2.93837667e-01 -8.18971097e-01
7.47493625e-01 -2.13805103e+00 1.64796814e-01 4.27007169e-01
1.43912479e-01 8.77267867e-02 -3.22827622e-02 5.19444108e-01
1.43491447e-01 -1.83956951e-01 -1.84253678e-01 -1.07499254e+00
1.45667404e-01 -7.37291947e-03 -1.13953790e-03 5.51055312e-01
1.35155633e-01 6.86513782e-01 -7.30973721e-01 -3.38405043e-01
2.71063685e-01 7.52318740e-01 -3.64661723e-01 1.78478211e-01
2.60510206e-01 4.36065018e-01 -2.45740488e-02 9.10434961e-01
7.36790597e-01 -1.82325870e-01 4.78278548e-01 -2.67878324e-01
2.27716476e-01 -3.54454555e-02 -1.42857969e+00 1.57949948e+00
-3.36211771e-01 5.50902903e-01 2.68624455e-01 -5.91939688e-01
7.99856424e-01 1.68417543e-01 4.68124598e-01 -3.40712070e-01
-1.73044205e-01 -8.81790668e-02 -4.15087044e-01 -2.35718042e-01
5.64029396e-01 -1.50765723e-03 7.15437829e-02 3.47330600e-01
3.71048674e-02 5.28406978e-01 1.83764711e-01 4.44477759e-02
7.91029274e-01 1.59581989e-01 3.67552131e-01 1.19922226e-02
5.80874801e-01 -3.38169336e-01 5.12360215e-01 8.69873583e-01
-1.69002727e-01 1.03536201e+00 -1.11307248e-01 -5.25026441e-01
-1.05642509e+00 -9.95838284e-01 -7.13867461e-03 1.15180993e+00
2.91681439e-01 -4.36778367e-01 -7.01544940e-01 -5.38732648e-01
3.84024322e-01 -5.19908257e-02 -9.47737753e-01 2.76982248e-01
-4.45507735e-01 -6.05588675e-01 4.07454640e-01 5.04010856e-01
5.58679104e-01 -6.73275471e-01 8.75738114e-02 -1.67472258e-01
-1.26351073e-01 -1.62135172e+00 -7.98702538e-01 -6.67808592e-01
-3.63111585e-01 -1.18638563e+00 -8.50407779e-01 -6.11474514e-01
1.09855413e+00 5.57188869e-01 9.97973561e-01 2.80598760e-01
-6.87189102e-01 9.76469874e-01 -1.66410029e-01 8.61893296e-02
1.10720299e-01 -2.15010598e-01 2.61790812e-01 1.03609252e+00
3.60022545e-01 -6.96321368e-01 -1.00027966e+00 4.30087000e-01
-5.30042350e-01 -4.25544567e-02 5.29933512e-01 7.56480932e-01
3.91068608e-01 -3.21400404e-01 1.31404340e-01 -8.10359597e-01
5.41099310e-02 -2.89702833e-01 -3.50770146e-01 5.79270542e-01
-2.60325968e-01 -2.63047636e-01 3.09048712e-01 -4.27884221e-01
-1.22396052e+00 4.43996578e-01 2.31621638e-01 -3.65129203e-01
-5.16136408e-01 5.55818602e-02 -4.03112561e-01 -5.12987971e-01
6.16572440e-01 2.01261118e-01 -4.34053130e-02 -5.43663681e-01
3.07982624e-01 5.43425560e-01 7.90244997e-01 -7.55466282e-01
8.37944567e-01 8.25519621e-01 1.75002947e-01 -8.72596562e-01
-8.95421624e-01 -9.06321883e-01 -1.08502364e+00 -2.92196304e-01
7.38338113e-01 -1.38933957e+00 -7.70122230e-01 5.85304320e-01
-7.80125916e-01 -1.78804666e-01 1.10586509e-01 2.80405670e-01
-1.46405697e-01 7.41953790e-01 -4.65925366e-01 -1.00501823e+00
-1.34950310e-01 -8.00672889e-01 1.36928880e+00 3.16279054e-01
-1.62351981e-01 -9.92323399e-01 -1.93868458e-01 7.69541681e-01
-1.31814629e-01 4.53521132e-01 1.19511776e-01 -3.34184617e-01
-7.77935207e-01 -4.24642980e-01 -3.34532350e-01 2.55952597e-01
2.91181684e-01 -2.00560495e-01 -1.25428522e+00 -3.90918702e-01
-5.41243136e-01 -2.06271723e-01 1.03028786e+00 4.29152548e-02
7.44047701e-01 -2.35372692e-01 -3.48676890e-01 7.47570872e-01
1.39361024e+00 -5.14598370e-01 5.20237267e-01 1.49353236e-01
1.05108917e+00 8.76943231e-01 3.27754498e-01 4.86635089e-01
5.74045062e-01 1.04670668e+00 5.07760346e-02 -2.48003975e-01
-4.73211914e-01 -3.28148663e-01 3.37493330e-01 2.25392822e-02
-3.68885010e-01 4.83446941e-02 -7.83189356e-01 7.20319510e-01
-2.01913953e+00 -1.29396176e+00 1.45021200e-01 2.56211281e+00
5.73857784e-01 -4.51602519e-01 4.55020308e-01 -2.34391004e-01
1.06902933e+00 9.15929228e-02 -4.03191507e-01 2.52769321e-01
-4.45165545e-01 -1.78684294e-01 5.80075979e-01 5.30719161e-01
-1.38121057e+00 7.40442634e-01 6.25360250e+00 4.43911821e-01
-6.17912591e-01 -1.59962565e-01 5.74754834e-01 -2.39188656e-01
2.50163954e-02 -6.93940222e-02 -1.12570381e+00 4.13308948e-01
2.98510104e-01 2.29434490e-01 5.88757217e-01 6.56380177e-01
-1.72905132e-01 -1.09176442e-01 -1.43656218e+00 1.38303423e+00
6.04383767e-01 -1.13113260e+00 -9.85406414e-02 3.93898547e-01
1.09596109e+00 -2.84195870e-01 1.71031401e-01 -2.16953263e-01
2.44085968e-01 -9.50575590e-01 5.76226056e-01 6.79791152e-01
7.25479066e-01 -5.83258748e-01 5.96322000e-01 -9.28753242e-02
-1.37434483e+00 -1.80420473e-01 -2.27706701e-01 -1.80121630e-01
7.19870552e-02 4.91849005e-01 -7.82737136e-01 8.36563170e-01
6.02131903e-01 9.10714090e-01 -7.89508224e-01 9.82280254e-01
-2.32121259e-01 1.13716990e-01 -4.13742423e-01 5.22241354e-01
-1.27171829e-01 -1.45373121e-01 3.36762160e-01 1.22390306e+00
3.41195405e-01 1.98982731e-01 6.78205192e-01 4.27414924e-01
-2.82377720e-01 1.71605833e-02 -3.95133823e-01 3.73639494e-01
5.53915858e-01 1.38934219e+00 -3.60926121e-01 -3.27808410e-01
-4.66621011e-01 1.35592413e+00 4.65209663e-01 4.63996291e-01
-3.32224429e-01 2.21123457e-01 1.18793082e+00 1.85090736e-01
2.98319638e-01 -2.52744973e-01 -1.37455046e-01 -1.36742151e+00
3.90975505e-01 -7.75556862e-01 6.45469010e-01 -5.66161215e-01
-1.72408640e+00 3.64484727e-01 -1.26423985e-01 -1.15528858e+00
-2.69837022e-01 -6.46424294e-01 -3.41553122e-01 1.09715664e+00
-1.55600417e+00 -1.81674016e+00 -6.24859273e-01 6.92038596e-01
9.95530412e-02 -2.80800164e-01 8.84264708e-01 3.11741561e-01
-6.49340570e-01 8.93356442e-01 5.24293445e-02 4.90253299e-01
1.16698921e+00 -1.24211824e+00 3.03657740e-01 1.16124666e+00
5.39890409e-01 9.47631836e-01 4.71373081e-01 -5.35369813e-01
-1.30090857e+00 -8.66330802e-01 7.78264225e-01 -6.01795554e-01
2.33621091e-01 -5.47545075e-01 -5.51834762e-01 6.24347925e-01
-1.44109521e-02 3.18403333e-01 1.11115134e+00 6.60403132e-01
-8.21931958e-01 -3.29520136e-01 -1.13426340e+00 6.38172448e-01
1.33467555e+00 -9.23297703e-01 -1.85499892e-01 3.27137589e-01
1.14550330e-02 -1.92109823e-01 -1.03732550e+00 2.52766728e-01
9.20482874e-01 -1.15285146e+00 1.36524665e+00 -2.80664414e-01
1.68402940e-01 -5.41195214e-01 -2.09190309e-01 -1.00727987e+00
-3.17211896e-01 -6.93731964e-01 2.46728379e-02 1.68441558e+00
3.21042687e-02 -6.72565877e-01 8.68579507e-01 1.07716799e+00
1.83037743e-01 -3.00226718e-01 -6.86593592e-01 -6.77960932e-01
-5.30994773e-01 -9.20125097e-02 7.07773089e-01 8.57902229e-01
-2.84211010e-01 1.40907422e-01 -8.16487551e-01 5.36489487e-01
9.92231250e-01 3.98417979e-01 1.08557534e+00 -1.13299537e+00
-4.67802763e-01 -3.06246758e-01 -8.19071233e-01 -9.78518128e-01
2.42339551e-01 -7.13803351e-01 -3.59750800e-02 -1.29643476e+00
5.76794505e-01 -4.46377844e-01 -3.42149198e-01 6.50963902e-01
-5.83631217e-01 1.07753217e+00 5.73788583e-01 4.87106949e-01
-5.44334292e-01 2.07310200e-01 6.60749435e-01 -2.12016538e-01
7.96374977e-02 -1.40161917e-01 -8.22223842e-01 5.91252387e-01
5.21666408e-01 1.59794802e-03 -1.08078331e-01 -4.99664187e-01
-1.76910460e-01 -5.07833540e-01 8.06686461e-01 -1.17513633e+00
3.85830998e-01 8.30264613e-02 8.90005350e-01 -1.93404689e-01
9.17407334e-01 -6.97677195e-01 1.99997127e-01 -1.62039831e-01
-2.05359206e-01 -3.27037424e-01 1.12149697e-02 7.96948969e-01
-1.42687142e-01 1.40254095e-01 6.73694015e-01 -1.64343968e-01
-8.60400379e-01 5.31134665e-01 3.67890060e-01 -2.33680308e-01
9.10986722e-01 -4.77676839e-01 -2.62312770e-01 -4.51342970e-01
-8.44291747e-01 1.64301679e-01 1.04719162e+00 3.28060836e-01
2.55188346e-01 -1.47588360e+00 -7.61440814e-01 2.55939066e-01
4.45890456e-01 -6.09009087e-01 6.92601442e-01 6.79181874e-01
-6.19705357e-02 1.08159997e-01 -3.59903038e-01 -8.10408354e-01
-1.88213813e+00 3.73105288e-01 3.58286172e-01 -4.37940434e-02
-4.25178140e-01 8.52727830e-01 4.11655337e-01 -3.04990798e-01
1.60379946e-01 4.77533907e-01 -8.19876567e-02 2.15660200e-01
8.31250310e-01 1.37263834e-01 -7.64880404e-02 -1.18748116e+00
-4.60955530e-01 1.01032650e+00 1.57868359e-02 -1.58016339e-01
1.16738009e+00 -4.87056822e-01 -8.21637362e-02 1.84458986e-01
1.12496686e+00 3.66566598e-01 -1.74411738e+00 -6.20369017e-01
-2.08207428e-01 -9.95728314e-01 -3.86883706e-01 -8.91449392e-01
-7.25806236e-01 6.29172325e-01 4.83313113e-01 1.42092481e-01
1.15847552e+00 -1.41305298e-01 5.88589787e-01 5.39902747e-02
4.19641167e-01 -1.04532433e+00 -4.14106157e-03 4.93362099e-02
6.21591508e-01 -1.52458644e+00 4.45092261e-01 -7.84983575e-01
-7.67041504e-01 1.00126743e+00 2.95306474e-01 6.97802193e-03
2.65278935e-01 -2.86766678e-01 1.37975201e-01 1.38354182e-01
-2.97644198e-01 -6.62603378e-01 8.06843221e-01 8.08303654e-01
2.26096585e-01 1.57514781e-01 2.50572860e-01 2.26784855e-01
-1.03501379e-01 -4.72552627e-01 1.23254940e-01 6.34371102e-01
-1.05491020e-01 -1.48539197e+00 -6.89434350e-01 6.56010881e-02
-4.98719424e-01 -7.24657997e-02 -8.55588317e-01 6.01964593e-01
2.69152105e-01 1.05996120e+00 4.72894944e-02 -2.55742699e-01
9.94715765e-02 2.20154926e-01 8.54753792e-01 -6.61571681e-01
-5.71897864e-01 -1.09707564e-01 4.21108872e-01 -4.72837687e-01
-7.89912701e-01 -1.13691950e+00 -4.99723315e-01 -3.48224491e-01
2.34395117e-01 -2.92068332e-01 4.77760613e-01 9.40394163e-01
6.01754367e-01 -1.23795956e-01 7.10042834e-01 -1.21983278e+00
-2.03067794e-01 -7.42821515e-01 -4.91033971e-01 1.01483166e+00
4.09194201e-01 -6.46158218e-01 -1.36753827e-01 5.15764892e-01] | [14.515748977661133, 0.9530503153800964] |
cd635de9-7a9f-4085-956e-76ca5f6fea7e | uit-hwdb-using-transferring-method-to | 2211.05407 | null | https://arxiv.org/abs/2211.05407v1 | https://arxiv.org/pdf/2211.05407v1.pdf | UIT-HWDB: Using Transferring Method to Construct A Novel Benchmark for Evaluating Unconstrained Handwriting Image Recognition in Vietnamese | Recognizing handwriting images is challenging due to the vast variation in writing style across many people and distinct linguistic aspects of writing languages. In Vietnamese, besides the modern Latin characters, there are accent and letter marks together with characters that draw confusion to state-of-the-art handwriting recognition methods. Moreover, as a low-resource language, there are not many datasets for researching handwriting recognition in Vietnamese, which makes handwriting recognition in this language have a barrier for researchers to approach. Recent works evaluated offline handwriting recognition methods in Vietnamese using images from an online handwriting dataset constructed by connecting pen stroke coordinates without further processing. This approach obviously can not measure the ability of recognition methods effectively, as it is trivial and may be lack of features that are essential in offline handwriting images. Therefore, in this paper, we propose the Transferring method to construct a handwriting image dataset that associates crucial natural attributes required for offline handwriting images. Using our method, we provide a first high-quality synthetic dataset which is complex and natural for efficiently evaluating handwriting recognition methods. In addition, we conduct experiments with various state-of-the-art methods to figure out the challenge to reach the solution for handwriting recognition in Vietnamese. | ['Kiet Van Nguyen', 'Duong T. D. Vo', 'Nghia Hieu Nguyen'] | 2022-11-10 | null | null | null | null | ['handwriting-recognition'] | ['computer-vision'] | [ 1.21732458e-01 -7.22838640e-01 -2.02623438e-02 -3.13425094e-01
-1.58990100e-01 -8.83670747e-01 6.35439873e-01 -6.54510796e-01
-4.24627244e-01 6.61495566e-01 -5.10038510e-02 -2.58001715e-01
3.60287800e-02 -6.95058763e-01 -3.59386623e-01 -8.08026671e-01
5.34816802e-01 4.21072811e-01 5.61414771e-02 -3.73050958e-01
6.98106527e-01 1.10338199e+00 -1.27629042e+00 2.93814808e-01
8.72477531e-01 3.54880571e-01 4.13392395e-01 7.01611161e-01
-3.75155896e-01 7.76463568e-01 -9.97209251e-01 -5.86767137e-01
3.19908440e-01 -5.82923055e-01 -5.71150243e-01 7.50623941e-01
5.35703063e-01 -6.23534799e-01 -7.56450117e-01 9.78388727e-01
6.48749471e-01 -1.69136599e-01 7.14924276e-01 -8.25235009e-01
-1.19446647e+00 2.74275690e-01 -4.20499057e-01 -1.15656946e-02
3.41547817e-01 3.81448805e-01 5.13847411e-01 -1.08687186e+00
8.39385390e-01 8.95342350e-01 4.19871151e-01 5.70659339e-01
-6.32122576e-01 -4.37416971e-01 -7.04168826e-02 2.28600830e-01
-1.26816428e+00 -1.87774926e-01 8.46040905e-01 -4.42804396e-01
6.64358735e-01 3.07032764e-01 5.72054207e-01 1.21082425e+00
-1.03168979e-01 9.53054130e-01 1.37900817e+00 -7.06643224e-01
-2.86609918e-01 2.78849334e-01 2.28129417e-01 7.96438038e-01
3.39176913e-04 -1.64842054e-01 -4.77695823e-01 3.83681208e-01
1.04006040e+00 5.77525645e-02 -1.62975416e-01 1.99036494e-01
-1.43809414e+00 3.06275815e-01 7.61045441e-02 4.70050156e-01
2.60091238e-02 -4.79220241e-01 3.20397049e-01 3.91844392e-01
-2.59605348e-01 4.07892704e-01 1.83234841e-01 -4.98470426e-01
-9.16979015e-01 2.67712921e-01 8.69492650e-01 1.35129249e+00
4.10351038e-01 1.49096414e-01 -2.27971151e-01 1.16714251e+00
-2.05777511e-01 8.39729548e-01 5.88238657e-01 -3.91254932e-01
8.30909133e-01 7.27897406e-01 1.01620397e-02 -1.18967140e+00
3.35373580e-02 1.89097539e-01 -9.62868690e-01 1.53407708e-01
8.86386812e-01 1.44097835e-01 -9.24459279e-01 7.90640473e-01
-3.94781262e-01 -6.36650980e-01 1.17651694e-01 1.21635640e+00
5.71724296e-01 5.84890604e-01 -4.77572918e-01 -1.53113618e-01
1.11542773e+00 -1.23974705e+00 -9.83895004e-01 1.15282893e-01
8.67186785e-02 -1.22986352e+00 1.62614536e+00 5.22332311e-01
-8.23677182e-01 -7.13389516e-01 -1.19264781e+00 -3.73782292e-02
-4.27329093e-01 8.03182781e-01 5.80541134e-01 9.59853530e-01
-4.84768271e-01 4.35623348e-01 -7.49473393e-01 -5.07788718e-01
3.58386338e-01 -3.61403152e-02 -5.05267620e-01 -1.08058944e-01
-7.72977531e-01 9.88005280e-01 3.91995192e-01 5.23594320e-01
-6.42339349e-01 -4.25347947e-02 -3.25312555e-01 -3.60386372e-01
2.99097478e-01 4.06948805e-01 1.01298165e+00 -1.01695716e+00
-1.71444535e+00 7.73912609e-01 -1.48933768e-01 2.25480601e-01
1.16257584e+00 2.87297279e-01 -7.15635717e-01 4.34317440e-02
-2.38624632e-01 1.17010094e-01 9.82149005e-01 -1.14125025e+00
-4.62677926e-01 -4.12013799e-01 -3.07701558e-01 1.24221757e-01
-5.46491981e-01 3.91641587e-01 -7.76633501e-01 -8.44581485e-01
8.96307006e-02 -8.29898000e-01 3.04958224e-01 4.90966961e-02
-3.66489470e-01 -1.04272187e-01 1.18363094e+00 -9.69954848e-01
1.21382678e+00 -2.17954373e+00 -2.27711290e-01 3.65605980e-01
-1.87200114e-01 6.21081710e-01 -3.29265207e-01 7.30630040e-01
4.76385266e-01 -4.16222140e-02 -1.83439314e-01 1.39194682e-01
2.88265441e-02 4.86887902e-01 -6.46446407e-01 3.46848786e-01
4.17621732e-01 8.06840360e-01 -8.15369248e-01 -5.83493412e-01
-1.47158193e-04 2.29972318e-01 1.84057325e-01 2.59590387e-01
9.58291069e-02 2.12579161e-01 -5.00979424e-01 1.33415341e+00
6.69606507e-01 2.38839030e-01 3.87678057e-01 6.97857067e-02
-3.77498746e-01 -3.09759229e-01 -1.13705707e+00 1.16846669e+00
-2.05025449e-01 8.92014742e-01 -4.05535668e-01 -5.55090904e-01
1.46632564e+00 1.00955091e-01 -2.12368771e-01 -7.49170065e-01
-5.86877167e-02 6.42856598e-01 4.02191915e-02 -9.72658753e-01
7.64983296e-01 2.36398250e-01 2.23109171e-01 4.97281998e-01
-8.23159218e-02 -3.38121116e-01 6.38539970e-01 -2.51602799e-01
7.14474201e-01 1.95681646e-01 -2.14587182e-01 -1.59687370e-01
7.52641976e-01 2.48194441e-01 3.89231443e-01 8.57038200e-01
-1.65315941e-01 1.01814675e+00 4.96414185e-01 -6.31751180e-01
-1.43498039e+00 -8.65202188e-01 -2.02881902e-01 6.32604837e-01
2.52934061e-02 -1.23666592e-01 -8.70346427e-01 -6.64621174e-01
-1.18835494e-01 2.02117756e-01 -3.11678857e-01 4.17491019e-01
-8.99509013e-01 -6.09137356e-01 1.19831705e+00 8.93083215e-01
1.07184732e+00 -1.40647173e+00 -3.77022296e-01 9.94234383e-02
1.42899424e-01 -1.04746079e+00 -8.18690002e-01 -1.62547037e-01
-5.74872732e-01 -1.09972656e+00 -1.35318792e+00 -1.20565379e+00
1.23061121e+00 1.52316958e-01 5.04314840e-01 2.83055007e-01
-4.05455053e-01 -7.80474171e-02 -3.90924603e-01 -4.54507977e-01
-6.06563747e-01 9.47364941e-02 1.77429661e-01 1.41144648e-01
5.00860751e-01 8.47072080e-02 -8.56960863e-02 7.17144489e-01
-8.53225112e-01 -8.99269478e-04 8.13012481e-01 1.10281873e+00
3.66541743e-01 1.24304116e-01 2.11200595e-01 -7.79808879e-01
9.99348104e-01 2.80081093e-01 -7.50729918e-01 8.74773562e-01
-3.09662104e-01 -1.54626906e-01 1.05867195e+00 -7.48215497e-01
-1.03177750e+00 8.18021148e-02 1.33554190e-01 -2.65066385e-01
-1.41018555e-01 4.46350425e-01 -1.74633294e-01 -2.97555178e-01
4.88361478e-01 8.95540178e-01 3.54742557e-01 -4.29521650e-01
-1.36045977e-01 1.42190337e+00 7.50913918e-01 -8.06567013e-01
6.03997588e-01 2.41026744e-01 -1.45822033e-01 -1.37217748e+00
-2.93277532e-01 -2.34298974e-01 -1.14858317e+00 -3.06927890e-01
5.54328203e-01 -2.86897480e-01 -8.32217634e-01 1.23367369e+00
-1.23052251e+00 -1.21972091e-01 3.22906300e-02 3.48857164e-01
-3.96143466e-01 7.07317412e-01 -7.24179208e-01 -9.61693108e-01
4.76221815e-02 -1.35275173e+00 7.84803450e-01 2.98228085e-01
4.14956212e-02 -7.12770700e-01 -2.22034484e-01 4.16041970e-01
3.55057806e-01 -1.19234510e-01 8.01370442e-01 -2.68424243e-01
-5.85114360e-01 -3.33123654e-01 -4.95618343e-01 7.59132981e-01
5.00702739e-01 6.28283441e-01 -7.89039969e-01 -2.22317591e-01
-2.18930423e-01 -4.45787936e-01 6.02544904e-01 -4.76166755e-01
1.30364108e+00 -2.61667609e-01 2.39244699e-01 4.73067760e-01
1.14102495e+00 6.54667556e-01 1.10841501e+00 4.00777072e-01
8.18341136e-01 5.42018652e-01 7.18920171e-01 3.54362369e-01
2.34645624e-02 5.06871402e-01 -5.37210763e-01 -1.68140352e-01
-1.28069311e-01 -4.31535631e-01 4.66057241e-01 1.26414824e+00
-8.69165242e-01 -1.01216324e-01 -1.18466771e+00 3.47191662e-01
-1.57026863e+00 -1.03980768e+00 -8.20543319e-02 2.07168198e+00
1.12178206e+00 -2.57530153e-01 1.94050714e-01 4.23453242e-01
5.91734231e-01 -2.07851417e-02 -3.44027489e-01 -3.88392955e-01
-6.10860467e-01 -6.70087188e-02 5.42581975e-01 1.64931804e-01
-1.06319189e+00 1.11608672e+00 6.38820267e+00 8.69056821e-01
-1.55272102e+00 -5.84897339e-01 4.20018345e-01 5.18187404e-01
1.33259505e-01 -2.76385218e-01 -1.03011024e+00 6.17156386e-01
2.75776982e-01 -6.26559108e-02 6.67556882e-01 5.70929646e-01
6.48963090e-04 1.57336891e-01 -1.03928876e+00 1.24271691e+00
5.29766679e-01 -1.25599527e+00 4.28587437e-01 1.91750620e-02
8.21971893e-01 -5.86780548e-01 -5.32700047e-02 -1.09877042e-01
-4.50648844e-01 -1.24829960e+00 7.95184553e-01 7.06905186e-01
9.35700893e-01 -5.95970511e-01 7.24196374e-01 4.96650785e-01
-7.21830308e-01 1.00028053e-01 -6.67210639e-01 -1.08066685e-02
-1.49498329e-01 -2.46709790e-02 -8.31245482e-01 4.52695340e-01
3.00603598e-01 6.22593820e-01 -8.14026535e-01 7.64614522e-01
-3.68026704e-01 5.50845385e-01 -1.28161862e-01 -7.22276092e-01
4.30027664e-01 -6.07125401e-01 1.34159610e-01 1.42147851e+00
5.15901625e-01 1.35021150e-01 6.27394244e-02 8.96672904e-01
-4.76126671e-02 2.39048854e-01 -7.37626433e-01 -8.58387053e-01
3.07296276e-01 8.93714905e-01 -9.07451868e-01 -2.10853636e-01
-5.48428476e-01 1.30857742e+00 6.81437105e-02 3.93439412e-01
-4.74376112e-01 -8.71203125e-01 9.23149958e-02 8.65991265e-02
-6.05292581e-02 -7.59053051e-01 -3.53159398e-01 -1.31578183e+00
4.35408324e-01 -1.39787638e+00 -2.21878722e-01 -4.45438862e-01
-1.41667247e+00 6.60414755e-01 -3.70687664e-01 -1.33829093e+00
-9.80850682e-02 -1.22917914e+00 -6.17881119e-01 1.06773365e+00
-1.13491094e+00 -1.41854668e+00 -6.62913680e-01 6.24721467e-01
8.58673692e-01 -9.44883883e-01 8.49564135e-01 2.71798402e-01
-5.66645384e-01 8.75989258e-01 2.40454927e-01 7.54374325e-01
7.45918751e-01 -1.07106423e+00 4.50033516e-01 7.67261207e-01
2.85169274e-01 7.31137335e-01 1.36203617e-01 -7.03362644e-01
-1.92783678e+00 -6.89144075e-01 9.95425045e-01 -5.56716144e-01
6.89148903e-01 -3.79811108e-01 -9.28241491e-01 4.28625852e-01
1.62033085e-02 -1.98891938e-01 4.46183890e-01 -2.52276540e-01
-4.51232344e-02 1.02109499e-02 -9.02294993e-01 8.13575983e-01
1.02199399e+00 -6.65120780e-01 -5.73011875e-01 4.00993407e-01
-2.98492193e-01 -3.63359123e-01 -6.83745265e-01 3.58670019e-02
9.62508678e-01 -6.21556461e-01 4.89411414e-01 -5.02523541e-01
5.55008054e-01 -4.33851749e-01 -2.76367031e-02 -9.55934703e-01
-3.60287540e-02 -5.83405674e-01 3.17456335e-01 1.53169882e+00
3.41020703e-01 -5.60881913e-01 9.02660549e-01 5.82630098e-01
2.59204417e-01 -5.32511294e-01 -3.74724716e-01 -1.22882521e+00
1.78480849e-01 3.72500569e-02 7.21900940e-01 1.07900488e+00
-1.92212909e-01 -1.63290828e-01 -5.04987895e-01 -2.25324497e-01
4.30961967e-01 2.18472391e-01 9.39479232e-01 -6.40864313e-01
-6.25214577e-02 -6.36759579e-01 -5.76023161e-01 -9.60688531e-01
1.63852900e-01 -7.84303784e-01 -9.81816947e-02 -1.16175556e+00
2.72368997e-01 -5.33180952e-01 2.17940480e-01 5.55781901e-01
-4.61248728e-03 1.86007366e-01 4.61189389e-01 5.70899129e-01
8.37009102e-02 3.54010761e-01 1.80412793e+00 -6.30701900e-01
-6.91023469e-02 -2.29006428e-02 -1.30040124e-01 4.95463192e-01
7.83485651e-01 1.66509852e-01 -2.19541773e-01 -6.13714337e-01
-3.60178500e-01 -9.83029138e-03 1.07763425e-01 -7.08169103e-01
4.70110267e-01 -4.59394425e-01 5.95115423e-01 -6.80416405e-01
-7.47015327e-02 -6.84383273e-01 -1.05457440e-01 2.55660504e-01
-1.18184283e-01 2.87039429e-01 -1.04365788e-01 -9.98909175e-02
-6.56680703e-01 -5.27531803e-01 6.10509992e-01 -1.81538507e-01
-1.03559303e+00 -4.79796116e-04 -5.23335338e-01 -3.26913804e-01
9.44486022e-01 -7.29264021e-01 -4.38529730e-01 -9.85089988e-02
-2.28290007e-01 -5.50633073e-02 6.01774037e-01 4.65409964e-01
7.85248399e-01 -1.23510253e+00 -8.69097173e-01 7.90433645e-01
2.34871447e-01 -4.25166160e-01 -3.19957472e-02 5.29158175e-01
-1.51227868e+00 4.86123204e-01 -8.66943896e-01 -3.20668042e-01
-1.40728104e+00 3.38980854e-01 1.51758313e-01 1.12365514e-01
-6.07341468e-01 3.15029085e-01 -4.67433393e-01 -5.85038781e-01
4.11222368e-01 -1.58539414e-01 -4.58402187e-02 -8.59375000e-02
7.19178259e-01 4.82123524e-01 9.57623571e-02 -5.70658565e-01
-1.44933417e-01 7.33215213e-01 -1.94616079e-01 -9.99625996e-02
9.43085849e-01 4.43349779e-01 -2.70467997e-01 5.04615903e-01
7.82610297e-01 2.35439554e-01 -1.08324516e+00 1.19677596e-01
1.56542376e-01 -1.08342993e+00 -4.93894637e-01 -9.15058553e-01
-8.66923273e-01 9.73629594e-01 4.18346763e-01 -1.16761327e-01
1.00336957e+00 -5.81848145e-01 5.37509382e-01 1.21357131e+00
6.78823173e-01 -1.75762188e+00 2.42680032e-02 8.52912784e-01
1.37036419e+00 -1.24078572e+00 -1.01416960e-01 -3.02542895e-01
-9.33568120e-01 1.77462912e+00 6.18166268e-01 5.56569733e-02
2.07324341e-01 6.06265306e-01 6.68839037e-01 2.69769758e-01
-1.17588736e-01 1.24765128e-01 2.02546924e-01 6.08305871e-01
5.15095830e-01 -2.66581215e-02 -6.25021935e-01 4.51419115e-01
-1.87213749e-01 1.07214384e-01 6.35706723e-01 1.10077631e+00
-9.07539651e-02 -1.55496931e+00 -5.32559037e-01 2.68953651e-01
-2.16594070e-01 2.81730950e-01 -8.68130624e-01 9.26339984e-01
-1.28852695e-01 5.14595032e-01 -4.58931327e-02 -3.29233795e-01
6.39373422e-01 9.79534816e-03 8.28867614e-01 -1.30993202e-01
-5.38070917e-01 -1.96338087e-01 -1.28486142e-01 6.96248561e-02
-2.27053508e-01 -5.82866013e-01 -1.00660300e+00 -5.27351379e-01
-1.91838890e-01 -1.39955461e-01 8.04550886e-01 6.33581579e-01
-2.83961445e-01 6.20417707e-02 7.21476734e-01 -7.86118865e-01
-9.28221703e-01 -9.57432985e-01 -9.87350881e-01 5.37972987e-01
-1.48290336e-01 -2.98223168e-01 3.06520946e-02 3.85518014e-01] | [11.865046501159668, 2.5406997203826904] |
36232faf-c9ff-4b51-91f5-2f86533041e2 | ctap-complementary-temporal-action-proposal | 1807.04821 | null | http://arxiv.org/abs/1807.04821v2 | http://arxiv.org/pdf/1807.04821v2.pdf | CTAP: Complementary Temporal Action Proposal Generation | Temporal action proposal generation is an important task, akin to object
proposals, temporal action proposals are intended to capture "clips" or
temporal intervals in videos that are likely to contain an action. Previous
methods can be divided to two groups: sliding window ranking and actionness
score grouping. Sliding windows uniformly cover all segments in videos, but the
temporal boundaries are imprecise; grouping based method may have more precise
boundaries but it may omit some proposals when the quality of actionness score
is low. Based on the complementary characteristics of these two methods, we
propose a novel Complementary Temporal Action Proposal (CTAP) generator.
Specifically, we apply a Proposal-level Actionness Trustworthiness Estimator
(PATE) on the sliding windows proposals to generate the probabilities
indicating whether the actions can be correctly detected by actionness scores,
the windows with high scores are collected. The collected sliding windows and
actionness proposals are then processed by a temporal convolutional neural
network for proposal ranking and boundary adjustment. CTAP outperforms
state-of-the-art methods on average recall (AR) by a large margin on THUMOS-14
and ActivityNet 1.3 datasets. We further apply CTAP as a proposal generation
method in an existing action detector, and show consistent significant
improvements. | ['Jiyang Gao', 'Ram Nevatia', 'Kan Chen'] | 2018-07-12 | ctap-complementary-temporal-action-proposal-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Jiyang_Gao_CTAP_Complementary_Temporal_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Jiyang_Gao_CTAP_Complementary_Temporal_ECCV_2018_paper.pdf | eccv-2018-9 | ['temporal-action-proposal-generation'] | ['computer-vision'] | [ 5.82261086e-01 7.77751021e-03 -6.97284520e-01 -3.96194637e-01
-9.57062125e-01 -3.44629079e-01 7.81748295e-01 -6.43813331e-03
-4.80522454e-01 4.79342878e-01 5.66857517e-01 2.91347057e-01
-6.09828904e-02 -5.33297956e-01 -6.15690351e-01 -6.15822315e-01
-3.67463440e-01 1.19640650e-02 1.31331694e+00 5.53024374e-02
3.95202041e-01 2.75356382e-01 -1.61555278e+00 8.34501326e-01
4.71756041e-01 1.28476179e+00 2.21132371e-03 4.79014963e-01
2.04788059e-01 9.03934836e-01 -7.17179954e-01 -1.38505578e-01
4.14842963e-01 -6.08226478e-01 -6.95080936e-01 -3.21601331e-02
2.70157278e-01 -6.57537162e-01 -3.17243278e-01 8.56029987e-01
1.90606996e-01 3.71037483e-01 5.38581431e-01 -1.48192263e+00
-1.72974765e-01 8.20698142e-01 -5.62909544e-01 5.32262802e-01
5.96021891e-01 2.69370526e-01 1.01386893e+00 -1.05043721e+00
8.69938374e-01 1.27105415e+00 5.37718654e-01 6.04140162e-01
-6.73494995e-01 -7.93070912e-01 4.70894039e-01 5.62490523e-01
-1.32625079e+00 -4.05685335e-01 5.23832500e-01 -2.61165589e-01
8.28844965e-01 1.84331506e-01 7.61235595e-01 1.22537148e+00
4.28018093e-01 1.27099395e+00 5.82380354e-01 -1.59477573e-02
4.29194689e-01 -3.71391535e-01 -1.78602070e-01 2.21668974e-01
-5.23215085e-02 -1.90723408e-02 -7.31968880e-01 -2.40487140e-02
6.66760862e-01 2.13674590e-01 -9.13098454e-02 -1.15929857e-01
-1.63258934e+00 5.95690072e-01 2.80606002e-01 2.66351968e-01
-6.34401143e-01 2.99267232e-01 6.40575588e-01 9.37017947e-02
3.49876612e-01 3.22161019e-01 -3.97783190e-01 -3.65738392e-01
-1.17015135e+00 5.68015277e-01 1.40775457e-01 9.37136352e-01
3.03207338e-01 -1.92358404e-01 -1.08346140e+00 5.26097298e-01
3.55675012e-01 1.43032074e-01 5.89104712e-01 -1.15268385e+00
5.99242926e-01 7.21904039e-01 3.63533050e-01 -7.70583034e-01
-2.09581971e-01 8.59939083e-02 -4.23213124e-01 1.45410091e-01
2.73953974e-01 2.04514593e-01 -9.32563961e-01 1.41955936e+00
3.93544912e-01 4.11336780e-01 -2.97403157e-01 1.07915092e+00
7.81128526e-01 8.41841817e-01 3.59380126e-01 -4.91690844e-01
1.41638970e+00 -1.03478396e+00 -7.83369660e-01 6.20129285e-03
4.94937927e-01 -7.52140343e-01 7.54778504e-01 6.06205523e-01
-1.12459445e+00 -7.84567118e-01 -1.18079019e+00 2.41834715e-01
-2.12511364e-02 4.71762508e-01 4.67838168e-01 1.99965402e-01
-7.55064189e-01 7.89743304e-01 -1.04549658e+00 -3.10476124e-01
6.83942854e-01 4.59261276e-02 -1.87567249e-01 2.56982923e-01
-1.24413991e+00 7.12815702e-01 7.53305078e-01 -1.12017013e-01
-1.23970580e+00 -4.20785397e-01 -6.54513419e-01 -1.31101534e-01
6.80508375e-01 -1.97513431e-01 1.48403323e+00 -1.09048605e+00
-1.39373899e+00 4.48524773e-01 -1.07869066e-01 -9.48406577e-01
8.22852075e-01 -2.55548209e-01 -6.64409161e-01 5.07999778e-01
3.01700890e-01 1.14393497e+00 9.70774889e-01 -5.56876481e-01
-1.23617220e+00 1.54735357e-01 2.45142765e-02 2.18756020e-01
-3.60417999e-02 4.29017037e-01 -7.12441146e-01 -9.61807311e-01
3.48025143e-01 -8.10066402e-01 -2.26014763e-01 2.55845338e-01
-2.73750991e-01 -7.21423447e-01 8.26822758e-01 -4.61180240e-01
1.64045370e+00 -2.01322389e+00 -2.83448219e-01 3.39060053e-02
-8.10435191e-02 3.54115248e-01 -2.87454456e-01 1.86194822e-01
1.90743357e-02 1.41665675e-02 1.88827086e-02 3.51808108e-02
1.38998469e-02 -1.37921602e-01 -4.17061776e-01 3.26243490e-01
4.26328868e-01 6.55246794e-01 -1.11049545e+00 -7.91692317e-01
2.33593166e-01 8.79918486e-02 -4.16719228e-01 2.03985721e-01
-3.78711015e-01 1.79530859e-01 -6.96394682e-01 7.59917855e-01
3.89167994e-01 3.33436430e-02 2.35139411e-02 -3.00325692e-01
-9.91190225e-02 6.07395113e-01 -1.46263719e+00 1.52399492e+00
1.68135792e-01 5.35871804e-01 -6.33565247e-01 -6.59040153e-01
8.99844110e-01 5.04637837e-01 9.13290977e-01 -6.73233092e-01
-9.78809744e-02 5.92669062e-02 -1.15092471e-01 -5.79762578e-01
7.61237621e-01 4.30577509e-02 -1.18620746e-01 5.29705465e-01
-1.66298561e-02 2.54919022e-01 4.65166956e-01 3.33992749e-01
1.55612314e+00 7.19494402e-01 6.85518920e-01 2.44651973e-01
4.59077895e-01 -1.64574668e-01 1.00308216e+00 7.79034317e-01
-6.85980856e-01 9.31417108e-01 6.75314724e-01 -6.86168551e-01
-9.52120125e-01 -1.16182077e+00 4.71010283e-02 1.28730023e+00
2.73636788e-01 -5.20845413e-01 -5.29645026e-01 -1.33432531e+00
-3.15734506e-01 4.63901281e-01 -8.24783683e-01 -2.58012056e-01
-7.22177267e-01 -3.27509493e-01 5.10697961e-01 9.49101031e-01
6.70548081e-01 -1.62202394e+00 -9.75238144e-01 4.24579680e-01
-3.12870115e-01 -1.13831437e+00 -7.62973011e-01 -2.44338617e-01
-8.70580614e-01 -1.23409438e+00 -6.71890259e-01 -3.75948608e-01
4.54920292e-01 2.83779860e-01 9.88405704e-01 -5.69616444e-02
-4.88158874e-02 -1.26585048e-02 -9.24257815e-01 -4.01937783e-01
-4.31332976e-01 -3.09953153e-01 1.53192624e-01 1.76346555e-01
6.11423314e-01 -1.74968347e-01 -9.62221265e-01 8.13505232e-01
-9.87134278e-01 -8.15437213e-02 7.78995872e-01 4.04447943e-01
8.03794920e-01 -4.00526114e-02 5.99358380e-01 -3.62347424e-01
2.67372549e-01 -4.18890387e-01 -3.81677300e-01 2.47648239e-01
-3.31882298e-01 -1.35370195e-01 2.06835568e-01 -7.99138784e-01
-1.10988533e+00 2.34557852e-01 -1.67733300e-02 -5.67527890e-01
-1.67227015e-01 -1.38990944e-02 -1.36279957e-02 4.69109386e-01
7.42439151e-01 1.54443532e-01 -3.87341201e-01 -2.23538682e-01
2.06144810e-01 3.60315621e-01 4.02225882e-01 -1.01586647e-01
3.35180581e-01 6.65980935e-01 -3.67391676e-01 -2.29473725e-01
-8.16152215e-01 -7.14232385e-01 -5.61824620e-01 -5.28464139e-01
9.89900529e-01 -8.08945477e-01 -5.05565882e-01 2.84610689e-01
-1.18574131e+00 -5.34306504e-02 -3.52947831e-01 7.97742307e-01
-6.47556663e-01 4.23579395e-01 -3.84473026e-01 -8.48768115e-01
-3.75533462e-01 -1.07981968e+00 1.25937128e+00 1.25591695e-01
-6.71912849e-01 -2.02381045e-01 -6.79698959e-02 8.57500881e-02
-6.94887247e-03 3.23450565e-01 1.93302184e-01 -8.03666592e-01
-6.40670061e-01 -4.25949454e-01 -4.92035747e-02 3.29015195e-01
-1.17634274e-02 2.75136113e-01 -7.39668667e-01 -6.15714118e-02
-1.71494991e-01 -2.38971710e-01 1.10804784e+00 7.69734025e-01
1.24494565e+00 -4.38308924e-01 -5.29864132e-01 1.56767145e-01
1.03571749e+00 5.45344532e-01 9.72658813e-01 3.10286194e-01
1.79907337e-01 3.98526579e-01 1.62143147e+00 8.13028336e-01
-9.89953503e-02 1.11783516e+00 4.91407961e-01 2.08624452e-01
-1.01391375e-01 -3.21184933e-01 8.56699347e-01 2.09891737e-01
-3.24075997e-01 -2.73056895e-01 -4.08263206e-01 7.04889774e-01
-2.11185431e+00 -1.46970773e+00 -1.38867190e-02 2.28041244e+00
6.86618090e-01 6.63479030e-01 5.74572265e-01 2.04364434e-01
8.19144547e-01 4.15616393e-01 -4.72444445e-01 -2.59510148e-02
7.58273974e-02 -3.57553698e-02 3.01256657e-01 -1.61562011e-01
-1.51597726e+00 9.18670535e-01 5.92564487e+00 1.19553471e+00
-7.00434864e-01 1.80346996e-01 5.02765596e-01 -2.97969729e-01
1.84316218e-01 3.94836403e-02 -1.02833152e+00 6.56190634e-01
5.97702503e-01 4.65306304e-02 -3.00150931e-01 9.40348387e-01
6.92584753e-01 -4.78095531e-01 -1.25094306e+00 6.71330273e-01
9.32201073e-02 -1.28400338e+00 1.83609918e-01 -4.05235618e-01
7.82913089e-01 -8.08017030e-02 -1.45451695e-01 4.35325652e-01
2.07813546e-01 -5.92753530e-01 1.06476319e+00 6.00501001e-01
5.91314495e-01 -5.07118523e-01 7.89340675e-01 7.61339590e-02
-1.54630220e+00 -2.76868463e-01 -3.07816774e-01 1.69833302e-01
4.54225510e-01 3.31262976e-01 -9.18384910e-01 2.96968162e-01
9.06778574e-01 1.03808498e+00 -5.62039852e-01 1.38176954e+00
-3.75003934e-01 7.90550828e-01 -2.38202348e-01 -1.64862216e-01
2.96231717e-01 1.12527139e-01 6.06804907e-01 1.16345656e+00
4.75151300e-01 2.58197516e-01 2.19027624e-01 6.68578744e-01
9.60815400e-02 -1.29294500e-01 -2.60293365e-01 1.35660395e-01
5.87752819e-01 1.15999079e+00 -1.12140167e+00 -7.86808491e-01
-2.73606837e-01 7.87419975e-01 -2.95410812e-01 9.87467468e-02
-1.39719117e+00 -2.76323944e-01 3.89755726e-01 2.52373755e-01
4.95705128e-01 1.67357266e-01 -5.34231737e-02 -8.96862447e-01
3.57186586e-01 -7.69521356e-01 7.72953749e-01 -8.68414283e-01
-9.03526545e-01 3.91047388e-01 2.91007429e-01 -2.28053069e+00
-1.51316434e-01 -2.19453245e-01 -7.96523571e-01 5.09853810e-02
-9.21785891e-01 -1.03406918e+00 -2.33545482e-01 3.33090484e-01
1.19676542e+00 1.11173904e-02 1.72605932e-01 1.81247890e-01
-2.52906740e-01 4.20613825e-01 -5.09367824e-01 1.74502075e-01
7.24110603e-01 -8.93835843e-01 4.14457411e-01 1.06315112e+00
2.25236148e-01 3.39605272e-01 6.83406234e-01 -8.86471868e-01
-7.27803588e-01 -1.33529496e+00 8.91453505e-01 -6.32895410e-01
5.86538613e-01 -3.35686356e-02 -8.19495559e-01 5.73970318e-01
-1.86208516e-01 1.21503554e-01 1.57217026e-01 -4.17354405e-01
-2.59702355e-01 -1.95142686e-01 -9.66458142e-01 6.28982782e-01
1.27592206e+00 -5.49370013e-02 -8.13176513e-01 3.70874792e-01
7.70735919e-01 -2.85530835e-01 -6.52588964e-01 6.39800310e-01
8.20611179e-01 -1.24553883e+00 9.05134916e-01 -4.40242797e-01
5.37932873e-01 -6.12429202e-01 1.44578889e-01 -6.68242455e-01
-3.39719445e-01 -6.38854861e-01 -2.40701303e-01 1.05799365e+00
2.61027902e-01 3.61233167e-02 8.13349366e-01 2.53368020e-01
-3.14345717e-01 -8.17519665e-01 -1.19148457e+00 -9.72783208e-01
-5.92184424e-01 -8.11765313e-01 7.20847785e-01 5.86042225e-01
1.82614863e-01 -1.15957439e-01 -3.98955643e-01 -1.32601961e-01
2.41382092e-01 -2.55018055e-01 5.55079639e-01 -8.15284610e-01
-1.55310944e-01 -7.08520293e-01 -5.89889884e-01 -1.20091212e+00
-2.91360170e-01 -3.26061487e-01 4.50356811e-01 -1.36978078e+00
2.77052432e-01 -1.30765930e-01 -5.99373639e-01 7.53768265e-01
-1.47944853e-01 4.19423521e-01 1.23727292e-01 3.26896161e-01
-1.39893246e+00 6.07461989e-01 1.10781384e+00 -1.23446032e-01
-4.98274118e-01 3.66846383e-01 1.14961334e-01 9.13613379e-01
4.47252035e-01 -5.84574342e-01 -3.66795987e-01 1.61136433e-01
1.51067272e-01 9.37689021e-02 2.67122060e-01 -1.41849315e+00
1.71313211e-02 -4.19742703e-01 3.43796551e-01 -1.03397584e+00
2.56288290e-01 -6.02934182e-01 5.60420826e-02 5.02589881e-01
-7.06266880e-01 -1.40002295e-01 -1.61515772e-01 7.01108873e-01
-1.74767956e-01 -3.86054926e-02 7.51658380e-01 -6.13719644e-03
-1.11262524e+00 4.04963523e-01 -5.73182881e-01 -1.15586348e-01
1.27696490e+00 -6.34920537e-01 -9.95387211e-02 -2.21077904e-01
-6.15351439e-01 9.83751193e-02 9.61101651e-02 8.33879769e-01
9.87847507e-01 -1.61536300e+00 -6.00760639e-01 -5.52826375e-02
5.15011251e-01 -2.41324559e-01 2.69065559e-01 1.13626730e+00
-2.28151217e-01 2.62542069e-01 -1.64912477e-01 -6.13150179e-01
-1.40936124e+00 5.53569853e-01 3.38274986e-02 -2.19451025e-01
-7.65859365e-01 8.52590024e-01 2.27704093e-01 3.56823236e-01
4.68987286e-01 -8.40369761e-01 -5.65711558e-01 1.49302855e-01
8.95027339e-01 5.01995623e-01 -1.87791437e-01 -5.91450930e-01
-4.45522517e-01 4.13043827e-01 -6.91522658e-02 -2.16779485e-01
1.02544200e+00 1.77501917e-01 3.93366098e-01 3.08887035e-01
7.53690541e-01 -3.99904549e-01 -1.74982929e+00 -2.34085679e-01
2.34399766e-01 -5.54690778e-01 -3.26199293e-01 -6.44395769e-01
-8.63461316e-01 5.38052797e-01 6.51589334e-01 1.94866419e-01
1.29191983e+00 2.28159964e-01 8.59938800e-01 9.37804729e-02
3.91689748e-01 -1.57533205e+00 5.72438240e-01 2.35092238e-01
9.59641099e-01 -1.21614420e+00 3.46704751e-01 -3.85006309e-01
-9.92499411e-01 1.00107634e+00 7.89883733e-01 -4.92720492e-02
3.56951565e-01 -9.24358666e-02 -2.98714161e-01 -1.52240157e-01
-9.34965193e-01 -7.84790441e-02 5.79573631e-01 4.23487544e-01
2.29245022e-01 1.24821952e-02 -8.32184494e-01 6.48615122e-01
3.59722644e-01 3.02048385e-01 4.80289072e-01 8.64774048e-01
-7.42077112e-01 -6.92580581e-01 -2.73873836e-01 5.13400972e-01
-5.16141713e-01 2.48240933e-01 -2.38645449e-01 6.11649632e-01
4.19418812e-01 8.21149647e-01 2.42652223e-01 -5.56078672e-01
3.15285325e-01 -1.82562619e-01 1.72621861e-01 -5.17930269e-01
-5.15816391e-01 2.41905972e-01 1.79759920e-01 -1.11672390e+00
-7.46615946e-01 -8.04718673e-01 -1.50697196e+00 2.80494273e-01
-3.87684792e-01 2.28675865e-02 1.47577271e-01 1.10622585e+00
2.58810312e-01 3.92294854e-01 6.16746485e-01 -8.31001580e-01
-5.24210811e-01 -1.29008877e+00 -3.64434153e-01 5.92520475e-01
-1.30845038e-02 -7.75959492e-01 -1.39848650e-01 3.47286195e-01] | [8.353755950927734, 0.43458056449890137] |
dd68563b-cac9-4e81-9b26-d862e62859d8 | approaches-toward-physical-and-general-video | 2112.07661 | null | https://arxiv.org/abs/2112.07661v1 | https://arxiv.org/pdf/2112.07661v1.pdf | Approaches Toward Physical and General Video Anomaly Detection | In recent years, many works have addressed the problem of finding never-seen-before anomalies in videos. Yet, most work has been focused on detecting anomalous frames in surveillance videos taken from security cameras. Meanwhile, the task of anomaly detection (AD) in videos exhibiting anomalous mechanical behavior, has been mostly overlooked. Anomaly detection in such videos is both of academic and practical interest, as they may enable automatic detection of malfunctions in many manufacturing, maintenance, and real-life settings. To assess the potential of the different approaches to detect such anomalies, we evaluate two simple baseline approaches: (i) Temporal-pooled image AD techniques. (ii) Density estimation of videos represented with features pretrained for video-classification. Development of such methods calls for new benchmarks to allow evaluation of different possible approaches. We introduce the Physical Anomalous Trajectory or Motion (PHANTOM) dataset, which contains six different video classes. Each class consists of normal and anomalous videos. The classes differ in the presented phenomena, the normal class variability, and the kind of anomalies in the videos. We also suggest an even harder benchmark where anomalous activities should be spotted on highly variable scenes. | ['Niv Cohen', 'Laura Kart'] | 2021-12-14 | null | null | null | null | ['physical-video-anomaly-detection', 'general-action-video-anomaly-detection'] | ['computer-vision', 'computer-vision'] | [ 4.38230574e-01 -2.05627248e-01 1.24115370e-01 -9.89917200e-03
-3.04453999e-01 -5.43953061e-01 9.36341107e-01 2.78928727e-01
7.96811655e-02 4.34926480e-01 -2.10182443e-01 -1.19202226e-01
-3.45221043e-01 -4.17897403e-01 -8.20346594e-01 -1.04379463e+00
-5.85751891e-01 -4.60567549e-02 6.53081834e-01 -1.75750209e-03
4.52330500e-01 8.36208880e-01 -2.13693786e+00 5.46159804e-01
3.09617192e-01 1.33005989e+00 -1.07861936e-01 8.58824611e-01
4.91373762e-02 9.61047709e-01 -8.91060591e-01 -3.26505750e-02
1.86443448e-01 -5.41391253e-01 -5.88583827e-01 5.70997655e-01
6.03332102e-01 -2.21484154e-01 -2.66605794e-01 1.06026971e+00
-4.58303988e-02 1.24521628e-01 7.12352514e-01 -1.53922939e+00
-2.00841844e-01 -8.81934240e-02 -4.50693071e-01 8.62086594e-01
9.73341405e-01 3.91522229e-01 4.65844303e-01 -5.44882417e-01
6.33467495e-01 1.01882231e+00 3.73459160e-01 3.31737131e-01
-8.67202699e-01 -9.38421488e-02 2.19423100e-01 7.10252583e-01
-1.01878762e+00 -2.18264490e-01 9.61745381e-01 -8.12406063e-01
7.59342134e-01 4.17084426e-01 5.81859291e-01 1.59884262e+00
5.09575605e-01 7.22973287e-01 7.28959322e-01 -2.66191542e-01
3.79228264e-01 -4.09752987e-02 -7.77582526e-02 3.68407875e-01
4.60130036e-01 2.31699534e-02 -2.49539435e-01 -3.61242324e-01
3.46757174e-01 2.44867459e-01 -4.96197999e-01 -3.44207793e-01
-1.32822251e+00 3.97416651e-01 -3.77370775e-01 7.32537687e-01
-4.88100559e-01 -5.37211336e-02 7.43191421e-01 7.64624476e-01
4.65611607e-01 5.63247323e-01 -3.36911023e-01 -3.37140501e-01
-7.94429541e-01 3.85683298e-01 6.97673082e-01 7.92958856e-01
3.83030355e-01 3.60292643e-01 -1.29147887e-01 4.61709470e-01
-1.36058614e-01 2.42301628e-01 4.34261829e-01 -7.95619547e-01
1.94639742e-01 4.95885491e-01 1.59604490e-01 -1.32238579e+00
-2.78184444e-01 1.00579366e-01 -5.47670603e-01 1.16955340e-01
6.16331100e-01 1.14091270e-01 -6.41162694e-01 1.13750374e+00
2.04917744e-01 5.07190406e-01 6.59115463e-02 7.33905077e-01
2.22708091e-01 7.99505532e-01 -3.39833587e-01 -5.77826321e-01
9.70464289e-01 -4.76645142e-01 -8.44163120e-01 2.59467900e-01
6.04614854e-01 -8.59980464e-01 7.55494237e-01 1.04830301e+00
-9.10598576e-01 -4.65115905e-01 -8.99853051e-01 7.68889844e-01
-5.58814526e-01 -1.25187173e-01 1.02098033e-01 6.48832619e-01
-7.17975497e-01 7.87342012e-01 -9.05766070e-01 -6.70602441e-01
1.50429368e-01 9.92809087e-02 -4.11578596e-01 -3.14329825e-02
-6.71414673e-01 5.89632928e-01 3.44468862e-01 6.73799068e-02
-1.11141491e+00 -3.14131349e-01 -8.87929499e-01 -2.87660867e-01
7.81407058e-01 2.15236228e-02 8.11148584e-01 -1.34704244e+00
-1.06597900e+00 7.54854798e-01 6.31409511e-02 -4.24157202e-01
5.96520483e-01 -5.37197173e-01 -1.07565916e+00 2.84495920e-01
-2.28279322e-01 -7.50738382e-02 1.39238107e+00 -1.20084739e+00
-6.91664338e-01 -2.72102326e-01 9.71974283e-02 -5.49290597e-01
-2.75435001e-01 1.36520565e-01 3.49306539e-02 -9.72221673e-01
-6.15196675e-02 -9.46917653e-01 8.75061005e-02 -1.37841061e-01
-5.13258874e-01 -2.02796206e-01 1.49422264e+00 -5.72001517e-01
1.19847488e+00 -2.43371749e+00 1.20086506e-01 2.18930960e-01
-2.54069120e-01 1.97659671e-01 5.80559447e-02 5.06063819e-01
-3.45102638e-01 -1.81472674e-01 -3.09740067e-01 2.81158805e-01
-3.78824532e-01 3.08574468e-01 -4.34082240e-01 8.00449014e-01
6.07167482e-01 5.19673228e-02 -8.56829703e-01 -2.60371387e-01
3.54883850e-01 3.32353041e-02 -4.08723116e-01 3.84747386e-01
-3.07215184e-01 7.73910880e-01 -4.11757022e-01 1.02685428e+00
4.47610766e-01 2.83552259e-01 -2.13322848e-01 -1.33527726e-01
-1.72925219e-01 -4.46998298e-01 -1.17526317e+00 1.31086051e+00
1.90231621e-01 8.61724317e-01 -2.55549520e-01 -1.34956682e+00
6.74349427e-01 5.14631152e-01 1.04337120e+00 -4.17701304e-01
6.77413717e-02 2.57482499e-01 -3.86531055e-02 -1.27599204e+00
5.10789752e-01 3.83449733e-01 1.87636435e-01 -3.89159992e-02
1.77993581e-01 1.79133564e-01 5.23075521e-01 -2.07711682e-01
1.47618210e+00 2.83273548e-01 1.47968784e-01 -3.11187208e-01
9.75237846e-01 -7.83137977e-03 2.99244076e-01 5.34813344e-01
-2.00460598e-01 6.52518511e-01 8.99694026e-01 -8.79562378e-01
-9.82819259e-01 -1.04346442e+00 -3.83387357e-02 5.84319830e-01
2.83262908e-01 -2.08774984e-01 -6.67716444e-01 -8.09269190e-01
9.07462370e-03 6.15233302e-01 -4.53664482e-01 -3.04801643e-01
-6.79337919e-01 -8.81418049e-01 3.86074752e-01 3.09510499e-01
-3.99435684e-02 -9.87070799e-01 -8.06992292e-01 -5.58693819e-02
7.83305615e-02 -1.44184136e+00 1.07382797e-01 -5.73255494e-02
-9.38460052e-01 -1.53120744e+00 -5.04665732e-01 -3.75728697e-01
6.15002811e-01 5.62768579e-02 1.07406282e+00 2.05782443e-01
-6.18409038e-01 1.09265816e+00 -7.70760596e-01 -3.39133501e-01
-7.38676786e-01 -5.00728011e-01 4.43494737e-01 6.57347023e-01
4.74299252e-01 -4.27317232e-01 -4.28099960e-01 7.42753446e-01
-1.22130346e+00 -7.81929493e-01 3.89598995e-01 4.12533134e-01
5.74633420e-01 2.65512526e-01 1.60643026e-01 -5.67616522e-01
2.64483929e-01 -8.70623708e-01 -6.07524812e-01 8.39778185e-02
-1.63015574e-01 -4.01897095e-02 8.76974285e-01 -5.76916158e-01
-7.04978108e-01 7.97201842e-02 -7.67553672e-02 -8.78071308e-01
-9.06331241e-01 3.45697701e-02 -1.91307608e-02 6.90853745e-02
5.68580210e-01 2.80561507e-01 2.77560111e-02 -3.16699654e-01
-2.47221097e-01 3.00590336e-01 6.25150859e-01 -5.28877378e-01
8.13371241e-01 5.19699693e-01 1.08855158e-01 -1.56836998e+00
-2.89491385e-01 -6.40911520e-01 -5.23387253e-01 -6.41857147e-01
9.96906757e-01 -5.12120306e-01 -2.49091208e-01 6.75421536e-01
-1.03109086e+00 6.66624308e-02 -5.25058627e-01 5.41864872e-01
-7.30019152e-01 8.98797631e-01 -3.21883470e-01 -9.94901240e-01
4.40415621e-01 -1.14586532e+00 1.12972951e+00 -1.08690016e-01
-2.08814040e-01 -8.30827415e-01 1.09017923e-01 -8.38617235e-02
2.25336015e-01 8.22756946e-01 8.62369955e-01 -9.45259809e-01
-6.86350882e-01 -5.30277550e-01 3.51131618e-01 6.46679699e-01
2.99459279e-01 5.73687136e-01 -1.10119104e+00 -3.84274781e-01
3.05655092e-01 1.62196830e-02 5.84123492e-01 4.66823667e-01
1.54245591e+00 -6.17634691e-02 -2.67790407e-01 3.04102898e-01
1.23086286e+00 5.25610328e-01 8.59023511e-01 2.07763627e-01
4.86659050e-01 7.52828479e-01 8.83850455e-01 5.23109674e-01
-4.79323179e-01 8.77696037e-01 9.42347944e-01 3.53387445e-01
3.40521842e-01 4.66684788e-01 9.28584754e-01 5.57915866e-01
-5.99005520e-01 -5.19306540e-01 -7.29798675e-01 5.05780518e-01
-1.78031123e+00 -1.41416836e+00 -3.97670269e-01 2.31999779e+00
-2.50181675e-01 3.20510387e-01 3.89196754e-01 6.47111893e-01
7.95969844e-01 3.09491575e-01 -4.69643772e-01 -3.73142123e-01
-1.75715625e-01 -3.77726972e-01 1.26577049e-01 -2.09494933e-01
-1.43986714e+00 2.85760332e-02 5.92828226e+00 6.53698683e-01
-1.03151155e+00 -1.04466945e-01 6.17879689e-01 -1.08677186e-01
1.92125455e-01 -3.63872439e-01 -4.49480265e-01 7.72892416e-01
1.09265983e+00 3.31350327e-01 -3.60510796e-02 9.78724241e-01
1.88123465e-01 -3.00222605e-01 -1.29269552e+00 1.00664544e+00
4.89214152e-01 -9.45647418e-01 1.80085361e-01 1.03549048e-01
5.09963810e-01 -3.89762193e-01 -6.41451031e-02 3.56557891e-02
-6.07295692e-01 -5.57944953e-01 6.83429658e-01 7.62293756e-01
1.65608644e-01 -7.84760594e-01 9.91559565e-01 1.19776614e-01
-1.14899004e+00 -4.26183581e-01 -3.15878659e-01 -1.41986916e-02
2.45253131e-01 6.46900058e-01 -2.14060798e-01 7.61960387e-01
9.65038896e-01 9.91978884e-01 -7.47087896e-01 1.03296006e+00
3.78329068e-01 6.09381258e-01 -2.68726140e-01 1.65087670e-01
2.78248608e-01 -2.78239548e-01 1.01439166e+00 1.15691996e+00
7.66140640e-01 -4.22178447e-01 2.83143729e-01 5.76529741e-01
6.40813112e-01 -5.29708527e-02 -1.21947920e+00 -2.36503273e-01
-2.46136487e-01 1.00577688e+00 -9.99037623e-01 -2.14531310e-02
-7.33388543e-01 1.01921535e+00 -4.99740183e-01 1.65424630e-01
-1.01361239e+00 -1.14932306e-01 7.10324287e-01 4.95959491e-01
2.30251685e-01 -2.60799438e-01 5.47057450e-01 -1.18963099e+00
4.73882735e-01 -7.95062184e-01 5.30542076e-01 -5.64476788e-01
-1.39622629e+00 6.33862257e-01 4.02106971e-01 -1.81505537e+00
-5.81876278e-01 -1.11772990e+00 -7.19151795e-01 4.86891204e-03
-9.65158463e-01 -6.45660400e-01 -5.37944853e-01 7.35093892e-01
1.00316417e+00 -5.27148545e-01 3.99300039e-01 4.01602894e-01
-6.25050187e-01 2.01923251e-02 -3.85177955e-02 -1.36120422e-02
5.40875852e-01 -1.27374852e+00 2.00429186e-01 1.15909231e+00
1.15027130e-01 -1.69964626e-01 1.09885263e+00 -5.34939408e-01
-1.41288507e+00 -1.07185447e+00 8.00595284e-02 -8.50865245e-01
8.56626093e-01 -1.27856240e-01 -1.14750457e+00 5.57881892e-01
1.44153923e-01 2.17731521e-01 3.08349669e-01 -4.59163785e-01
1.62631780e-01 4.41951826e-02 -9.58667994e-01 3.17926645e-01
8.99081826e-01 -2.80445844e-01 -5.22283971e-01 5.05139351e-01
3.54195744e-01 -1.03623442e-01 -8.18872452e-01 5.27998805e-01
3.43560994e-01 -1.40798271e+00 9.79005098e-01 -6.58804119e-01
5.17292738e-01 -4.69822466e-01 -3.15054834e-01 -1.14158285e+00
2.71616485e-02 -5.44982255e-01 -5.91501117e-01 9.34785008e-01
4.06247377e-02 -4.43039209e-01 6.40640974e-01 -2.31722221e-02
-3.46836179e-01 -6.78794682e-01 -8.91005337e-01 -1.35332227e+00
-4.44087029e-01 -6.77173436e-01 2.71535069e-01 9.45562482e-01
-2.09175900e-01 -3.82116526e-01 -5.38654983e-01 4.20598686e-01
2.14860544e-01 -2.38419652e-01 7.75707185e-01 -1.26540124e+00
-2.10680783e-01 -4.17448163e-01 -1.32816172e+00 -4.14428502e-01
-7.59023800e-02 -7.49360099e-02 -3.81513722e-02 -8.11704099e-01
-2.33899474e-01 2.47163385e-01 -3.23330313e-01 -1.36132419e-01
1.01891123e-01 1.16255887e-01 -2.10627183e-01 1.41741395e-01
-5.23634434e-01 2.71237969e-01 6.24280810e-01 -1.70685083e-01
8.87864828e-02 2.51460850e-01 3.80058616e-01 1.04777110e+00
6.35280669e-01 -2.90596038e-01 -4.42220718e-01 8.71802941e-02
3.94533992e-01 4.34428528e-02 7.21138835e-01 -1.69758391e+00
-1.55102879e-01 -2.35901773e-01 8.17065984e-02 -7.11585820e-01
1.43038392e-01 -1.19886887e+00 1.91469163e-01 5.92282236e-01
6.91273585e-02 7.87543237e-01 3.43616791e-02 9.77417231e-01
-5.91847301e-01 -3.74160767e-01 5.76719463e-01 -9.17274505e-02
-1.07573247e+00 1.25437722e-01 -8.85414958e-01 -6.99924752e-02
1.69899940e+00 -6.45074904e-01 -8.99209827e-02 -3.69633198e-01
-8.98936987e-01 -3.46885711e-01 5.36083221e-01 7.25363076e-01
6.18897557e-01 -1.07736826e+00 -5.17825544e-01 2.61206061e-01
4.18116510e-01 -2.82924592e-01 4.71602589e-01 1.18993986e+00
-7.99885690e-01 1.38628051e-01 -4.16894943e-01 -1.14598191e+00
-1.17387533e+00 1.10951853e+00 2.44129688e-01 3.15835290e-02
-6.87036693e-01 2.48075679e-01 -3.76895294e-02 3.59995842e-01
1.66037127e-01 -5.90938151e-01 -4.92378414e-01 1.01118445e-01
6.49720848e-01 7.42152512e-01 2.26821646e-01 -7.93892622e-01
-2.36012310e-01 5.38663685e-01 1.10539451e-01 6.38194680e-01
9.61870551e-01 1.33509398e-01 -2.48802509e-02 8.87932301e-01
9.13915992e-01 5.87335825e-02 -1.09564340e+00 3.72421920e-01
5.41927218e-01 -5.04217386e-01 -5.14692187e-01 -1.79096267e-01
-1.16672027e+00 6.52445734e-01 9.36403930e-01 1.06479204e+00
1.28519666e+00 -2.91521344e-02 4.80666101e-01 5.87550640e-01
5.34309864e-01 -9.43738401e-01 4.30818349e-01 2.18190417e-01
8.74555588e-01 -1.16394806e+00 -7.52559528e-02 -4.90980446e-01
-4.69953269e-01 1.42402232e+00 5.59834540e-01 -3.22363079e-01
5.87939262e-01 2.08035484e-01 -2.62572885e-01 -4.35634941e-01
-5.13310611e-01 4.73269075e-03 4.08205628e-01 6.33378565e-01
2.66406775e-01 -3.17138553e-01 3.31293754e-02 4.95427698e-02
3.18881810e-01 -2.62453943e-01 8.35459232e-01 1.19414532e+00
-3.34873885e-01 -6.65950954e-01 -7.17965245e-01 7.22575426e-01
-6.72894657e-01 5.53455114e-01 -3.97135675e-01 9.52383935e-01
3.19116950e-01 8.90510082e-01 3.26256096e-01 -4.01248246e-01
5.59628725e-01 2.08537847e-01 3.93281251e-01 -2.24004567e-01
-1.84595734e-01 -2.12415367e-01 3.12505290e-02 -1.03617239e+00
-7.77656496e-01 -9.95964408e-01 -6.05751336e-01 -9.96266399e-03
-2.16770679e-01 -5.40778926e-03 3.04515958e-01 1.09746480e+00
1.37430236e-01 5.80142558e-01 6.02443933e-01 -1.05302656e+00
-2.41578132e-01 -7.81467915e-01 -7.68392742e-01 1.01026595e+00
5.93394101e-01 -1.07611251e+00 -6.27831876e-01 3.40652466e-01] | [7.8846564292907715, 1.5533655881881714] |
9d267713-13f8-4891-aa15-eb27a0b05899 | semi-supervised-chinese-word-segmentation | null | null | https://aclanthology.org/D14-1010 | https://aclanthology.org/D14-1010.pdf | Semi-Supervised Chinese Word Segmentation Using Partial-Label Learning With Conditional Random Fields | null | ['Paul Vozila', 'Fan Yang'] | 2014-10-01 | null | null | null | emnlp-2014-10 | ['partial-label-learning'] | ['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] | [-7.340752601623535, 3.701204776763916] |
835daf30-7f76-4ce4-bce6-333aaf75f722 | feature-learning-for-fault-detection-in-high | 1810.05550 | null | https://arxiv.org/abs/1810.05550v2 | https://arxiv.org/pdf/1810.05550v2.pdf | Feature Learning for Fault Detection in High-Dimensional Condition-Monitoring Signals | Complex industrial systems are continuously monitored by a large number of heterogeneous sensors. The diversity of their operating conditions and the possible fault types make it impossible to collect enough data for learning all the possible fault patterns. The paper proposes an integrated automatic unsupervised feature learning and one-class classification for fault detection that uses data on healthy conditions only for its training. The approach is based on stacked Extreme Learning Machines (namely Hierarchical, or HELM) and comprises an autoencoder, performing unsupervised feature learning, stacked with a one-class classifier monitoring the distance of the test data to the training healthy class, thereby assessing the health of the system. This study provides a comprehensive evaluation of HELM fault detection capability compared to other machine learning approaches, such as stand-alone one-class classifiers (ELM and SVM), these same one-class classifiers combined with traditional dimensionality reduction methods (PCA) and a Deep Belief Network. The performance is first evaluated on a synthetic dataset that encompasses typical characteristics of condition monitoring data. Subsequently, the approach is evaluated on a real case study of a power plant fault. The proposed algorithm for fault detection, combining feature learning with the one-class classifier, demonstrates a better performance, particularly in cases where condition monitoring data contain several non-informative signals. | ['Thomas Palmé', 'Gabriel Michau', 'Yang Hu', 'Olga Fink'] | 2018-10-12 | null | null | null | null | ['one-class-classifier'] | ['methodology'] | [ 2.38596573e-01 4.60313521e-02 4.71462905e-01 -1.79104149e-01
-2.21528709e-01 -7.76169822e-02 5.50009131e-01 4.95628148e-01
1.12348080e-01 5.41644990e-01 -4.15884405e-01 -1.13656648e-01
-8.80642235e-01 -8.21151614e-01 -3.07555974e-01 -1.23674154e+00
-4.39204752e-01 8.83736014e-01 8.00136011e-03 -1.55870706e-01
2.90744841e-01 8.57681036e-01 -2.22445297e+00 3.89403045e-01
7.04807281e-01 1.46890211e+00 1.74971044e-01 5.70131898e-01
2.68360019e-01 7.52883554e-01 -1.05297732e+00 5.28305709e-01
1.55714795e-01 -2.45356709e-01 -5.88230371e-01 4.74281341e-01
5.31146526e-02 -1.96796998e-01 -8.76584277e-03 8.28823864e-01
3.47673625e-01 1.08039998e-01 8.09817374e-01 -1.35523570e+00
-9.81513560e-02 2.54084200e-01 1.37397513e-01 1.34141758e-01
6.09981298e-01 8.32904801e-02 5.40444911e-01 -7.74029195e-01
1.16427459e-01 1.01758254e+00 6.59232318e-01 -1.49453413e-02
-9.84482825e-01 -4.08838727e-02 -3.27651173e-01 6.85104609e-01
-1.09464920e+00 7.71169364e-02 9.23427582e-01 -7.80436218e-01
1.26382470e+00 2.38767654e-01 8.16747606e-01 8.85549009e-01
6.08035147e-01 3.07788253e-01 1.29729271e+00 -5.87668538e-01
8.41157675e-01 2.03094631e-01 5.29670894e-01 5.22497714e-01
4.91315454e-01 3.46579671e-01 -1.36595055e-01 -3.34736407e-01
2.48745188e-01 4.65895861e-01 -3.30342740e-01 -3.82789344e-01
-7.77207613e-01 7.47907102e-01 1.48775771e-01 8.74682546e-01
-8.70643318e-01 -5.39109111e-01 5.26525855e-01 8.18002999e-01
2.10417226e-01 5.60172439e-01 -9.08456445e-01 7.67606124e-02
-9.03568327e-01 -3.09164003e-02 1.22175086e+00 2.72376388e-01
7.81977534e-01 3.87206167e-01 1.18470892e-01 4.81686115e-01
2.97076106e-01 1.30293667e-01 1.15155637e+00 -4.60310906e-01
-1.61805287e-01 1.03183186e+00 -1.58553481e-01 -9.20081019e-01
-6.74089968e-01 -5.55742204e-01 -7.02278018e-01 7.06439137e-01
9.36420113e-02 -3.88333887e-01 -6.42531991e-01 8.05664062e-01
2.99255311e-01 -6.32159412e-02 2.97057688e-01 4.91149187e-01
4.11186635e-01 6.18435621e-01 -4.92385358e-01 -3.77452731e-01
1.20562255e+00 -5.96046686e-01 -8.91389728e-01 2.55352557e-01
5.12775600e-01 -2.92165130e-01 5.43210864e-01 1.29632878e+00
-5.60774326e-01 -7.66980708e-01 -1.46938741e+00 8.01666439e-01
-8.27092230e-01 1.58981845e-01 2.54964322e-01 7.78404236e-01
-6.79975629e-01 1.21952748e+00 -7.39107430e-01 -4.19856191e-01
1.44138902e-01 4.98754054e-01 -6.17152274e-01 -2.50395060e-01
-1.20322347e+00 1.17791271e+00 8.61362815e-01 2.42284641e-01
-1.16535473e+00 -2.17057690e-01 -8.29977870e-01 1.62526354e-01
1.26766771e-01 -2.60191411e-01 7.76662529e-01 -8.53888392e-01
-1.58761263e+00 1.60190701e-01 4.36178148e-01 -6.48630023e-01
3.83752555e-01 -3.48415852e-01 -7.16298819e-01 3.87128949e-01
-3.59474987e-01 -3.42673779e-01 1.15616179e+00 -1.37794518e+00
-5.53493679e-01 -7.32275128e-01 -4.54886526e-01 -3.28608125e-01
-5.04194438e-01 -3.24181348e-01 7.56631076e-01 -9.03538316e-02
2.91048288e-01 -5.44877112e-01 7.99226835e-02 -7.41911292e-01
-4.53767389e-01 -2.79447347e-01 1.43952084e+00 -7.38890409e-01
9.03143764e-01 -2.16664934e+00 1.41408563e-01 5.29740691e-01
-1.22403748e-01 1.25793740e-01 5.39625525e-01 8.32358778e-01
-6.07799888e-01 -6.11696005e-01 -5.73210120e-01 1.83771864e-01
-5.05612418e-02 4.57963467e-01 2.10082129e-01 8.01248252e-01
3.89299184e-01 7.33777881e-02 -6.54927790e-01 -1.77561715e-01
7.29609668e-01 4.33305919e-01 5.98221608e-02 4.63665307e-01
1.51695326e-01 2.59115040e-01 -1.54893517e-01 7.23859310e-01
3.73703659e-01 1.87175632e-01 3.49175520e-02 -3.57847184e-01
-9.73351076e-02 -4.23835814e-01 -1.49170017e+00 1.00244975e+00
-4.99024719e-01 3.57042044e-01 -8.20392445e-02 -1.73635685e+00
1.43441939e+00 9.66502190e-01 8.42709482e-01 -1.74779311e-01
4.08729255e-01 3.36004108e-01 1.12873025e-01 -1.11065149e+00
-2.19837829e-01 -2.34660104e-01 9.54669788e-02 -1.32336596e-03
6.61177754e-01 7.47010717e-03 2.61581063e-01 -6.77326024e-01
1.42008352e+00 -1.70952231e-02 3.03554803e-01 -4.36002582e-01
7.09980309e-01 -8.67623091e-02 3.06232333e-01 3.43382716e-01
-1.29361019e-01 3.60336840e-01 5.01496792e-01 -6.59999788e-01
-6.57653451e-01 -7.65841126e-01 -4.79258686e-01 4.09533739e-01
-1.55898541e-01 5.82198612e-03 -7.53005207e-01 -5.98892450e-01
3.80977750e-01 7.85656869e-01 -7.14978397e-01 -6.73857749e-01
-1.68283641e-01 -1.03191054e+00 2.70914227e-01 1.15966648e-01
2.52832741e-01 -1.21675098e+00 -1.15391469e+00 2.70773798e-01
5.41762114e-01 -5.41001618e-01 1.11127961e+00 9.92744803e-01
-1.16078389e+00 -1.49530721e+00 -2.63067096e-01 -8.47151697e-01
6.00855350e-01 -4.97855067e-01 9.50192153e-01 1.81785841e-02
-6.13455057e-01 3.54973227e-01 -5.82884312e-01 -5.03985941e-01
-5.40443420e-01 -4.12133574e-01 2.09842741e-01 2.81438142e-01
3.22452366e-01 -7.18078792e-01 -1.86145633e-01 1.20675303e-01
-9.82909620e-01 -9.37202096e-01 9.21945751e-01 1.10502994e+00
2.00807571e-01 1.02738905e+00 8.15151393e-01 -7.46741056e-01
6.97148860e-01 -7.64258504e-01 -3.33907604e-01 7.07875341e-02
-9.73699987e-01 -1.52125388e-01 8.58892977e-01 -5.53586222e-02
-9.18607831e-01 7.49189258e-02 -2.16028661e-01 -4.81987119e-01
-1.04671574e+00 4.98336971e-01 -4.70817864e-01 -2.06496082e-02
9.08679664e-01 2.55412906e-01 1.85135588e-01 -7.46136963e-01
-2.81184614e-01 9.49425876e-01 4.00387585e-01 -1.57114178e-01
5.40317237e-01 8.10724124e-02 1.71961129e-01 -1.07054698e+00
-1.84611455e-01 -4.52518404e-01 -9.56053913e-01 -4.49195445e-01
6.52682185e-01 -5.51179826e-01 -6.07633471e-01 7.17822373e-01
-7.08456516e-01 4.03554402e-02 -6.95057690e-01 6.74932063e-01
-6.47055030e-01 3.12764615e-01 -8.54328573e-01 -1.08714688e+00
-3.44463468e-01 -1.02283466e+00 8.87775779e-01 -2.52933018e-02
-7.56662246e-03 -1.00429535e+00 1.43894047e-01 3.96145210e-02
2.37945735e-01 7.49927461e-01 9.72359776e-01 -1.27635872e+00
2.50906289e-01 -9.78957534e-01 6.83120370e-01 9.90398884e-01
3.92722905e-01 1.61273986e-01 -1.10219896e+00 -5.46358883e-01
7.41577506e-01 -2.77581364e-01 7.16269612e-01 1.46047786e-01
8.34757388e-01 -6.31815661e-03 -1.50135949e-01 3.50334053e-03
1.68480945e+00 5.96027732e-01 5.08365333e-01 5.31402230e-01
5.03535867e-01 6.40551984e-01 6.59532189e-01 7.67493963e-01
-3.42364967e-01 2.17005581e-01 9.30539310e-01 5.31027205e-02
3.86827677e-01 5.94535947e-01 4.89231169e-01 7.15180516e-01
1.28515260e-02 8.82404968e-02 -8.24520111e-01 2.76546597e-01
-1.62973654e+00 -8.37870479e-01 -2.05157787e-01 2.14044476e+00
1.27766624e-01 3.15622449e-01 7.05074295e-02 1.60240507e+00
8.61748815e-01 -2.57902175e-01 -3.73299807e-01 -4.50214088e-01
-1.25461191e-01 1.70926601e-01 1.00870477e-02 1.18654050e-01
-1.34452057e+00 -3.04813385e-02 5.60552931e+00 2.58586109e-01
-1.04697371e+00 -3.50887817e-03 1.85077339e-01 1.55807823e-01
5.28446615e-01 -2.92502195e-01 -3.14936548e-01 4.96179819e-01
1.46339595e+00 4.92733210e-01 8.29811543e-02 9.78866458e-01
-1.70398317e-02 -2.23525509e-01 -1.28051817e+00 7.17208862e-01
2.89158225e-01 -5.23522794e-01 -2.15310410e-01 1.59805659e-02
5.05021572e-01 -1.33428186e-01 -3.57296050e-01 2.62761742e-01
-3.20993483e-01 -6.24801755e-01 5.23796618e-01 7.90627360e-01
-3.53355333e-02 -8.49719703e-01 1.41163564e+00 3.08765024e-01
-5.91037273e-01 -1.01441717e+00 -1.44441381e-01 -3.38479191e-01
9.64617580e-02 1.19102383e+00 -7.64726937e-01 1.17750823e+00
8.27979565e-01 7.90448725e-01 -6.82670951e-01 1.21001756e+00
3.98952402e-02 4.99159873e-01 -2.75035441e-01 1.37475029e-01
-2.73497067e-02 -4.26431783e-02 5.77380121e-01 9.83646929e-01
5.50267160e-01 -4.97462332e-01 3.94495517e-01 3.13812405e-01
7.71387517e-01 3.78723107e-02 -8.19036007e-01 2.30072230e-01
2.02070028e-01 1.24550986e+00 -7.28677571e-01 -5.51722407e-01
-3.00144404e-01 7.58134007e-01 -2.02582091e-01 5.82383014e-02
-3.10771286e-01 -7.20256388e-01 3.00642431e-01 -3.13711576e-02
4.62983578e-01 1.33052155e-01 -8.17027017e-02 -7.39377797e-01
-8.61540958e-02 -7.61666417e-01 5.59005022e-01 -5.23430884e-01
-1.55962646e+00 7.87704647e-01 7.19651282e-02 -1.46214247e+00
-5.61815143e-01 -1.09367406e+00 -7.05266416e-01 5.09013355e-01
-1.15327215e+00 -5.85708916e-01 -3.46942753e-01 6.96473539e-01
4.24625188e-01 -5.10659218e-01 1.11346912e+00 3.39587510e-01
-9.57707226e-01 -5.44937588e-02 4.23521131e-01 -1.40939742e-01
1.96451247e-01 -1.61408484e+00 -6.03845239e-01 6.57283902e-01
-1.58277303e-01 -5.08695617e-02 8.44616652e-01 -6.03414655e-01
-1.51790094e+00 -8.74994159e-01 3.73628229e-01 -1.96609527e-01
4.42163706e-01 1.65695965e-01 -1.06787717e+00 4.43275392e-01
2.21944824e-01 1.57906801e-01 7.73495257e-01 -9.61194634e-02
2.78081030e-01 -3.63940686e-01 -1.61840773e+00 -3.54661584e-01
2.43221279e-02 -2.40134493e-01 -1.05149913e+00 5.58745146e-01
9.88271460e-03 1.57995522e-01 -1.60903645e+00 7.01029658e-01
3.53930205e-01 -1.27913487e+00 6.35381818e-01 -3.25003058e-01
2.08959401e-01 -4.16917562e-01 -2.02355802e-01 -1.51691020e+00
-4.69673425e-01 3.65454592e-02 -6.00050747e-01 9.65061724e-01
8.21582079e-02 -8.30618978e-01 4.15535510e-01 7.08901361e-02
-4.58728939e-01 -9.18921709e-01 -9.06757295e-01 -8.43467474e-01
-3.41175109e-01 -1.07786708e-01 4.86003309e-01 1.07912123e+00
2.15182751e-01 1.25726685e-01 1.33344516e-01 4.04816657e-01
4.45327848e-01 1.49348348e-01 2.73449421e-01 -2.04051089e+00
-4.06568259e-01 -2.84373462e-01 -1.24507356e+00 3.92812639e-01
5.10961600e-02 -5.74738622e-01 1.44254163e-01 -1.45428646e+00
-5.38551509e-01 5.95245622e-02 -6.33026958e-01 4.38861102e-01
2.22150624e-01 -9.33464915e-02 -1.94998145e-01 5.01582846e-02
-1.53031200e-01 5.80277443e-01 7.44832218e-01 -1.27092257e-01
-3.12718935e-02 1.80446625e-01 -7.40947714e-03 7.16378927e-01
8.61969948e-01 -2.13153452e-01 -3.04064006e-01 2.31320590e-01
-2.54565716e-01 3.95625606e-02 3.68655473e-01 -1.84106684e+00
1.54355252e-02 4.68760520e-01 9.36910689e-01 -5.99122047e-01
1.06079191e-01 -1.43019390e+00 3.83744180e-01 1.12099600e+00
1.39320582e-01 1.89144805e-01 -1.05410367e-02 4.66974229e-01
-7.33889759e-01 -7.12183714e-01 5.84986866e-01 -3.31269115e-01
-7.98099279e-01 -2.63183475e-01 -6.14961982e-01 -8.20862412e-01
1.43760467e+00 -3.96168649e-01 -2.94172615e-02 9.37271491e-02
-1.23668373e+00 -1.57665953e-01 2.66239762e-01 1.07847653e-01
6.41205430e-01 -1.08364713e+00 -5.88899553e-01 6.42475367e-01
9.17628035e-02 2.97559071e-02 3.44387025e-01 8.11702490e-01
-5.16665697e-01 2.01770931e-01 -6.47450566e-01 -8.00894260e-01
-9.51358199e-01 9.64819789e-01 5.75540781e-01 -9.99414176e-02
-6.97737038e-01 3.26556921e-01 -7.19908595e-01 -2.73164511e-01
1.23730607e-01 -3.52064818e-01 -7.67605424e-01 4.08198595e-01
4.06009853e-01 8.55150759e-01 9.68913257e-01 -6.84688389e-01
-2.34216124e-01 4.17347372e-01 4.54333276e-01 3.29226255e-01
1.46491218e+00 3.20638239e-01 -4.06918585e-01 1.01951718e+00
1.25105214e+00 -8.16406667e-01 -1.05949426e+00 3.59964818e-01
3.38633597e-01 -5.34675717e-02 2.46854290e-01 -8.84860218e-01
-1.11442363e+00 8.03034425e-01 1.22438145e+00 1.07275462e+00
1.46881878e+00 -2.71363705e-01 1.97473004e-01 7.96402097e-01
3.42432827e-01 -1.09728253e+00 -9.93349403e-02 2.89836049e-01
8.58259857e-01 -1.01316810e+00 -4.31476198e-02 -1.96467802e-01
-1.59832969e-01 1.63312447e+00 3.52865696e-01 -5.61470330e-01
1.16146708e+00 2.70029992e-01 -1.06455147e-01 -5.84370911e-01
-7.55007923e-01 3.27583328e-02 -9.98413116e-02 7.43195057e-01
1.63243845e-01 1.16777688e-01 -4.96836230e-02 5.03473222e-01
-1.22964084e-01 -5.28665585e-03 3.96505624e-01 1.32233679e+00
-6.73708439e-01 -8.33983779e-01 -8.42435777e-01 8.14953446e-01
-3.80587369e-01 5.43409467e-01 -2.05150425e-01 7.24420071e-01
5.57609200e-01 1.23250711e+00 1.36375219e-01 -6.52555406e-01
7.36401498e-01 5.48471749e-01 5.03803790e-01 -4.20904547e-01
-8.62799704e-01 -2.49845326e-01 -1.29442662e-01 -3.84680718e-01
-2.02253580e-01 -8.74375701e-01 -1.08038867e+00 3.92599612e-01
-5.03336966e-01 3.81927639e-01 9.03063834e-01 1.21565771e+00
4.05947901e-02 1.09406149e+00 1.08856153e+00 -1.25292706e+00
-8.01564872e-01 -1.40200770e+00 -1.25513124e+00 5.91049314e-01
4.85604256e-01 -1.11078715e+00 -7.27091253e-01 8.62863958e-02] | [6.747382640838623, 2.4039876461029053] |
08afe1de-b485-4480-a30f-083a67aa343a | learning-to-selectively-learn-for-weakly-1 | null | null | https://aclanthology.org/2022.naacl-main.99 | https://aclanthology.org/2022.naacl-main.99.pdf | Learning to Selectively Learn for Weakly Supervised Paraphrase Generation with Model-based Reinforcement Learning | Paraphrase generation is an important language generation task attempting to interpret user intents and systematically generate new phrases of identical meanings to the given ones. However, the effectiveness of paraphrase generation is constrained by the access to the golden labeled data pairs where both the amount and the quality of the training data pairs are restricted. In this paper, we propose a new weakly supervised paraphrase generation approach that extends the success of a recent work that leverages reinforcement learning for effective model training with data selection. While data selection is privileged for the target task which has noisy data, developing a reinforced selective learning regime faces several unresolved challenges. In this paper, we carry on important discussions about the above problem and present a new model that could partially overcome the discussed issues with a model-based planning feature and a reward normalization feature. We perform extensive evaluation on four weakly supervised paraphrase generation tasks where the results show that our method could significantly improve the state-of-the-art performance on the evaluation datasets. | ['Ping Li', 'Dingcheng Li', 'Haiyan Yin'] | null | null | null | null | naacl-2022-7 | ['paraphrase-generation', 'paraphrase-generation'] | ['computer-code', 'natural-language-processing'] | [ 5.89337111e-01 3.48414451e-01 -6.27456009e-01 -3.32837939e-01
-1.06141996e+00 -5.54896057e-01 9.28599000e-01 1.06204234e-01
-3.50616932e-01 1.17723417e+00 7.23979175e-01 -2.00307831e-01
2.41269879e-02 -7.06080854e-01 -7.85366237e-01 -4.18017596e-01
4.33708012e-01 7.58558273e-01 -2.06641078e-01 -5.93297422e-01
5.51338851e-01 7.03371540e-02 -1.54257512e+00 5.21334767e-01
1.13167918e+00 5.49844086e-01 3.47572654e-01 4.84763682e-01
-3.71571369e-02 9.67596769e-01 -5.61004400e-01 -2.04758480e-01
3.50484461e-01 -7.41989613e-01 -1.10011697e+00 -6.38006181e-02
2.79427171e-01 -2.17726991e-01 -2.09522098e-02 6.77950263e-01
6.60006225e-01 2.25866660e-01 5.27759254e-01 -1.28405142e+00
-8.82774472e-01 7.68306017e-01 -2.74075538e-01 1.05295897e-01
7.79331148e-01 1.67510360e-01 1.24896240e+00 -9.43591237e-01
8.04911435e-01 9.55423594e-01 3.44049156e-01 9.57151234e-01
-1.23207533e+00 -7.11011589e-01 3.18798274e-02 1.68624207e-01
-9.59472835e-01 -6.09915733e-01 8.68996739e-01 -1.81517914e-01
1.03102112e+00 2.23851278e-01 6.40188873e-01 1.67049038e+00
-8.99398252e-02 9.36235309e-01 1.18898916e+00 -7.88409472e-01
4.70498145e-01 2.58290380e-01 1.93342254e-01 6.73355877e-01
5.98785542e-02 4.27023292e-01 -8.97803783e-01 -3.58413666e-01
4.52701628e-01 -2.60464609e-01 -2.36469284e-01 -6.32773876e-01
-1.23251855e+00 1.08626163e+00 4.18595850e-01 3.36731344e-01
-8.93613324e-02 1.35868594e-01 2.49154121e-01 5.62054276e-01
4.55744445e-01 1.25666904e+00 -3.00616294e-01 -3.79765034e-01
-9.82036710e-01 6.00434721e-01 8.76501501e-01 1.11648560e+00
6.73915982e-01 -1.19492404e-01 -7.17593253e-01 1.01807463e+00
3.51506509e-02 2.73450762e-01 7.97446430e-01 -9.50173497e-01
8.45627129e-01 5.23129761e-01 2.63824344e-01 -3.54375333e-01
-1.90990657e-01 -2.42610455e-01 -5.51587284e-01 -1.54298851e-02
4.64413196e-01 -2.07476869e-01 -8.57078195e-01 2.07925534e+00
-1.97147392e-02 6.78340197e-02 2.21315920e-01 7.23973751e-01
5.58718503e-01 5.14298260e-01 6.40605344e-03 -2.18653113e-01
1.00274849e+00 -1.31375587e+00 -4.79472488e-01 -5.79641640e-01
6.58490777e-01 -5.92489481e-01 1.52429473e+00 1.06606744e-01
-1.31874919e+00 -5.38482070e-01 -1.01735950e+00 -1.96939290e-01
-5.87689765e-02 -1.88543573e-02 7.27079332e-01 4.41898525e-01
-1.01724172e+00 6.70023143e-01 -4.12159801e-01 -5.58966398e-01
3.43060732e-01 2.51186699e-01 -1.48938313e-01 -1.50578633e-01
-1.40917587e+00 1.03469884e+00 4.37686890e-01 -3.37693989e-01
-8.06998849e-01 -6.21798158e-01 -7.37620533e-01 5.29052950e-02
3.57346684e-01 -1.16870844e+00 1.53060722e+00 -8.65650058e-01
-1.65528369e+00 8.80851567e-01 -1.53167874e-01 -5.94606578e-01
5.08488297e-01 -1.93501145e-01 1.90239400e-01 -1.56923711e-01
3.31461012e-01 8.62017632e-01 8.61743808e-01 -1.21462047e+00
-5.58611989e-01 -1.96484342e-01 2.80733943e-01 7.02017367e-01
-3.04572970e-01 -2.55871952e-01 -1.29525391e-02 -7.32955277e-01
-2.03763917e-01 -1.12383723e+00 -4.56826508e-01 -4.28172946e-01
-2.27232456e-01 -2.91608572e-01 2.72153080e-01 -3.80751520e-01
1.01230657e+00 -1.76725972e+00 6.00748837e-01 -6.47733882e-02
-1.31632507e-01 1.24936372e-01 -3.00167441e-01 7.11626053e-01
-3.04920096e-02 1.71066880e-01 -2.58233458e-01 -5.62932611e-01
-7.63206184e-02 2.10716844e-01 -7.16754317e-01 -2.65724778e-01
1.90981522e-01 1.13566232e+00 -1.22220123e+00 -3.43793809e-01
-1.59316570e-01 -2.22469807e-01 -7.43716240e-01 5.20543456e-01
-3.67867768e-01 4.46224183e-01 -6.01746380e-01 5.09205878e-01
2.32912585e-01 -1.74186975e-01 -4.53769155e-02 2.23477334e-01
1.70302272e-01 5.02886415e-01 -6.50736392e-01 2.22910142e+00
-7.68189847e-01 2.04702139e-01 -3.42714965e-01 -8.42313707e-01
8.70774865e-01 1.85256615e-01 1.56281725e-01 -7.73893595e-01
-2.92674422e-01 3.72932762e-01 -8.88941512e-02 -5.20767450e-01
8.34869385e-01 -4.39271033e-01 -2.42072776e-01 7.77323365e-01
-1.76367210e-03 -7.91989386e-01 2.92909056e-01 3.38706791e-01
1.06741321e+00 3.67632866e-01 5.19004107e-01 -2.17632383e-01
3.25750142e-01 1.02875300e-01 2.54198909e-01 1.32625473e+00
-1.62625104e-01 8.21958542e-01 3.96559477e-01 -1.49478719e-01
-1.14580393e+00 -9.89377320e-01 1.99776843e-01 1.25185299e+00
9.39835683e-02 -3.80860776e-01 -6.63779438e-01 -1.05455053e+00
-1.29992336e-01 1.16940200e+00 -6.01209521e-01 -4.90539402e-01
-5.93660951e-01 -6.88964784e-01 5.75601280e-01 6.70566797e-01
2.75553226e-01 -1.64619815e+00 -4.30355906e-01 1.71535730e-01
-5.12496769e-01 -8.46589327e-01 -4.83911663e-01 2.74548680e-01
-8.79109502e-01 -7.00776041e-01 -6.45476043e-01 -9.35006082e-01
6.06693447e-01 3.82068276e-01 1.30086482e+00 -1.78797413e-02
1.32803217e-01 1.28779918e-01 -6.56879723e-01 -4.20201182e-01
-8.36401284e-01 5.08145273e-01 1.09817006e-03 -5.16906559e-01
2.30072036e-01 -6.22777224e-01 -3.41624975e-01 1.73392463e-02
-8.04135263e-01 5.25059581e-01 6.25077128e-01 1.33290803e+00
2.18729287e-01 -5.92732489e-01 1.02907848e+00 -9.06862140e-01
1.30424476e+00 -4.98749107e-01 -3.99628915e-02 3.04518342e-01
-8.46736193e-01 5.62948167e-01 7.66941845e-01 -1.90097645e-01
-1.08734989e+00 -4.62522171e-02 -1.12356767e-01 -2.60993484e-02
-2.35181339e-02 5.00461638e-01 1.48027405e-01 1.93089083e-01
9.67232943e-01 3.55085552e-01 -4.68111504e-03 -2.29366973e-01
5.61599970e-01 7.85755932e-01 1.14975803e-01 -8.09032142e-01
6.42372906e-01 1.80148438e-01 -1.99422851e-01 -4.39791232e-01
-1.18348527e+00 -3.52682054e-01 -3.01671773e-01 2.52582997e-01
4.03986692e-01 -7.75716543e-01 2.75629144e-02 1.78284466e-01
-1.07816553e+00 -4.96417791e-01 -7.88960040e-01 1.12903260e-01
-1.14853024e+00 4.30356115e-01 -5.02107978e-01 -5.13980746e-01
-5.57882547e-01 -1.22821259e+00 1.17180920e+00 8.12539756e-02
-6.41458213e-01 -7.20570982e-01 3.93248230e-01 7.91121781e-01
5.06142616e-01 -1.46926403e-01 9.94326591e-01 -8.54005456e-01
-5.09377182e-01 5.97715788e-02 -7.72375017e-02 2.10710391e-01
2.74339706e-01 -5.58574736e-01 -8.53329062e-01 -3.01916629e-01
1.18908420e-01 -1.03413236e+00 9.30143178e-01 3.64708379e-02
1.04258358e+00 -3.02338809e-01 -1.80713788e-01 4.27822173e-01
9.83465910e-01 6.55868836e-03 4.05071914e-01 3.82020384e-01
5.16308129e-01 7.79871047e-01 8.80051196e-01 4.43329245e-01
2.39354268e-01 8.67497027e-01 2.00722456e-01 4.12467197e-02
-5.19536212e-02 -7.15237617e-01 3.52251858e-01 3.67765307e-01
3.24837156e-02 -2.97034591e-01 -6.48856044e-01 6.09660268e-01
-2.22308683e+00 -1.12190771e+00 3.62388641e-01 2.19301105e+00
1.33402860e+00 1.94421962e-01 3.46388258e-02 -4.08264324e-02
3.42914939e-01 3.86025935e-01 -5.25636792e-01 -5.66118956e-01
4.35867794e-02 5.36867142e-01 1.83709428e-01 6.32731378e-01
-7.86887586e-01 1.18820667e+00 6.62061739e+00 7.96308994e-01
-5.07616580e-01 -7.42560104e-02 5.02596438e-01 -3.02109450e-01
-4.54430670e-01 1.36654750e-01 -7.06802428e-01 4.57449377e-01
8.06894064e-01 -4.62142974e-01 7.43220508e-01 9.18582261e-01
4.14354742e-01 -3.75297964e-02 -1.54610467e+00 7.69130290e-01
2.14279309e-01 -1.32450223e+00 2.96301365e-01 -9.25026536e-02
9.14643824e-01 -6.41468391e-02 1.21029548e-01 6.16356134e-01
4.13656712e-01 -1.12546277e+00 5.49082696e-01 3.07384521e-01
5.60263038e-01 -6.96451128e-01 5.24198174e-01 6.89267159e-01
-6.67335570e-01 -2.79155612e-01 -3.04904312e-01 -3.90367895e-01
2.74865508e-01 2.76640385e-01 -1.14682662e+00 5.09312391e-01
2.17096984e-01 7.24979043e-01 -7.74558604e-01 6.17022574e-01
-5.55960119e-01 4.90438938e-01 4.14029211e-02 -2.68096954e-01
2.88555831e-01 -1.31600961e-01 5.31428397e-01 1.08656597e+00
2.59573340e-01 -1.48591638e-01 2.35789940e-01 1.08195698e+00
-2.14768916e-01 1.60966918e-01 -8.81790638e-01 -3.45437899e-02
4.38552469e-01 9.92138505e-01 -1.68910325e-01 -3.48276675e-01
-1.39080435e-01 1.14654160e+00 7.55682111e-01 2.47156978e-01
-6.55607641e-01 -1.13001779e-01 3.30191016e-01 7.88895711e-02
-6.16533644e-02 -4.81362566e-02 -5.68179429e-01 -1.40319240e+00
1.56317592e-01 -1.20854664e+00 2.37113014e-01 -7.46257007e-01
-1.43098080e+00 4.01752591e-01 2.54545510e-01 -1.13584661e+00
-8.62199664e-01 -1.57103360e-01 -6.48253202e-01 8.53707314e-01
-1.56031823e+00 -1.25364816e+00 -2.47327536e-01 2.61088222e-01
8.87328506e-01 -4.14465129e-01 9.04548705e-01 -2.49543697e-01
-2.89336979e-01 6.14847064e-01 3.64737678e-03 -3.76697749e-01
8.46440792e-01 -1.44406724e+00 6.98081851e-01 7.28622496e-01
2.55993187e-01 8.70596886e-01 7.89497733e-01 -6.16423965e-01
-9.47918117e-01 -8.73003602e-01 1.15448380e+00 -6.51448429e-01
5.02154350e-01 -5.24188995e-01 -6.09891415e-01 5.96795440e-01
4.42632228e-01 -4.77106869e-01 6.95255518e-01 1.96921065e-01
-1.41005874e-01 2.25344434e-01 -1.12989569e+00 8.98586631e-01
1.42151630e+00 -5.08001566e-01 -1.02230191e+00 5.34818470e-01
8.75691831e-01 -4.67003405e-01 -3.76447707e-01 2.32696623e-01
4.35592800e-01 -9.35985565e-01 8.68166447e-01 -1.00772476e+00
8.53554726e-01 5.59708439e-02 -4.17698780e-03 -1.72081089e+00
-2.84740388e-01 -6.73339844e-01 -1.77804634e-01 1.06213021e+00
5.60597599e-01 -3.75099152e-01 1.06743217e+00 7.70722687e-01
-1.50220767e-01 -8.35829198e-01 -8.28140914e-01 -6.70511961e-01
2.03618020e-01 -4.32263836e-02 5.63347161e-01 8.23284566e-01
5.78160942e-01 8.27247500e-01 -6.43146634e-01 -5.63118458e-01
4.23220962e-01 2.90561050e-01 7.76755929e-01 -8.50757718e-01
-6.35504901e-01 -2.01208815e-01 1.84905335e-01 -1.18503463e+00
5.08997798e-01 -1.15872753e+00 1.23865597e-01 -1.55577219e+00
4.57705975e-01 -4.19885606e-01 -1.83518127e-01 4.92042303e-01
-4.33397502e-01 -1.37458533e-01 3.53221208e-01 2.83926308e-01
-3.90181601e-01 8.03742588e-01 1.28910506e+00 -1.28473788e-01
-5.74035823e-01 3.31243694e-01 -1.15566766e+00 4.56265718e-01
9.57899868e-01 -3.74114960e-01 -9.35185015e-01 -4.38387454e-01
3.37784946e-01 1.59934804e-01 2.30998710e-01 -7.86132872e-01
1.35514647e-01 -4.50139165e-01 1.87518261e-02 -1.86857685e-01
3.76515895e-01 -5.03953815e-01 -2.61610597e-01 4.06026840e-01
-1.10391879e+00 1.40707508e-01 -4.34461050e-02 7.14353740e-01
3.77030596e-02 -6.71669960e-01 6.97252691e-01 -4.09201175e-01
-5.10064185e-01 -4.57406081e-02 -1.49350613e-01 3.97224903e-01
9.69321191e-01 -1.97519541e-01 -2.87999779e-01 -6.61316752e-01
-4.62357312e-01 2.39090532e-01 6.57075763e-01 8.42674196e-01
6.59368455e-01 -1.30100703e+00 -7.79496670e-01 2.75027692e-01
5.08569837e-01 -5.10378219e-02 -2.12918758e-01 1.61229104e-01
1.91855431e-02 5.64951897e-01 -2.58674383e-01 -1.38296589e-01
-1.03405380e+00 5.47024071e-01 2.10069478e-01 -8.73864114e-01
-2.52918422e-01 6.76557541e-01 -7.30294734e-02 -5.18190444e-01
5.65916933e-02 -3.86774629e-01 -2.90426135e-01 -1.35715399e-02
2.99852937e-01 2.89805621e-01 5.35177365e-02 -2.07165271e-01
-2.24220138e-02 1.37660831e-01 -5.56428552e-01 -4.47634667e-01
1.18372667e+00 -5.44166006e-02 2.52730340e-01 2.86666721e-01
7.81806469e-01 -1.02569617e-01 -9.72240627e-01 -3.04674566e-01
7.73273259e-02 -4.61508542e-01 -4.03493643e-01 -1.01757860e+00
-4.46826696e-01 5.33876538e-01 2.06073672e-01 3.99317853e-02
8.38229179e-01 -1.08303661e-02 8.95687103e-01 6.66938841e-01
5.99369049e-01 -1.22425747e+00 5.69732428e-01 6.73207462e-01
1.18343318e+00 -1.35466528e+00 -2.47289278e-02 -1.40917942e-01
-1.11982083e+00 7.09649801e-01 8.30589116e-01 -7.02015758e-02
9.93860187e-04 -1.98608994e-01 -2.62075871e-01 -1.17010903e-03
-9.91028190e-01 -7.69529268e-02 1.21725008e-01 6.33132398e-01
5.12806237e-01 -1.83802381e-01 -6.79228008e-01 4.85609770e-01
-5.47692358e-01 3.14646989e-01 7.13462651e-01 1.11377084e+00
-3.92319709e-01 -1.59879577e+00 1.05172498e-02 5.54917276e-01
-6.69000149e-02 -3.54266673e-01 -7.77477682e-01 5.07807374e-01
-2.34135747e-01 1.07694769e+00 -2.90185928e-01 -2.82342464e-01
5.13950884e-01 3.09189469e-01 5.68323493e-01 -1.14483523e+00
-8.06620657e-01 -6.17000043e-01 3.21567088e-01 -4.53663051e-01
-3.37291837e-01 -5.50015032e-01 -1.06849337e+00 1.10548481e-01
-1.63759828e-01 2.27189496e-01 2.90664494e-01 1.03040814e+00
5.13509452e-01 1.83997810e-01 7.90260077e-01 -8.34062338e-01
-1.35043299e+00 -1.00747979e+00 -3.65620136e-01 7.59674430e-01
2.18946859e-01 -5.09634256e-01 -2.82327801e-01 -1.18990831e-01] | [11.628988265991211, 9.060684204101562] |
68e2e7a6-7c92-4699-8573-4f172c0cb480 | qvhighlights-detecting-moments-and-highlights | 2107.09609 | null | https://arxiv.org/abs/2107.09609v2 | https://arxiv.org/pdf/2107.09609v2.pdf | QVHighlights: Detecting Moments and Highlights in Videos via Natural Language Queries | Detecting customized moments and highlights from videos given natural language (NL) user queries is an important but under-studied topic. One of the challenges in pursuing this direction is the lack of annotated data. To address this issue, we present the Query-based Video Highlights (QVHIGHLIGHTS) dataset. It consists of over 10,000 YouTube videos, covering a wide range of topics, from everyday activities and travel in lifestyle vlog videos to social and political activities in news videos. Each video in the dataset is annotated with: (1) a human-written free-form NL query, (2) relevant moments in the video w.r.t. the query, and (3) five-point scale saliency scores for all query-relevant clips. This comprehensive annotation enables us to develop and evaluate systems that detect relevant moments as well as salient highlights for diverse, flexible user queries. We also present a strong baseline for this task, Moment-DETR, a transformer encoder-decoder model that views moment retrieval as a direct set prediction problem, taking extracted video and query representations as inputs and predicting moment coordinates and saliency scores end-to-end. While our model does not utilize any human prior, we show that it performs competitively when compared to well-engineered architectures. With weakly supervised pretraining using ASR captions, MomentDETR substantially outperforms previous methods. Lastly, we present several ablations and visualizations of Moment-DETR. Data and code is publicly available at https://github.com/jayleicn/moment_detr | ['Mohit Bansal', 'Tamara L. Berg', 'Jie Lei'] | 2021-07-20 | null | null | null | null | ['highlight-detection', 'moment-retrieval'] | ['computer-vision', 'computer-vision'] | [ 2.36042425e-01 -1.60798728e-01 -4.54799861e-01 -3.74867648e-01
-1.34777701e+00 -6.46296978e-01 6.68836772e-01 -6.66363463e-02
-3.31255674e-01 4.12348151e-01 7.38606036e-01 2.06846088e-01
2.49579877e-01 -1.18629687e-01 -9.67235267e-01 -2.55566239e-01
-2.38326013e-01 2.51497626e-02 5.41716993e-01 -2.47254878e-01
4.50814724e-01 1.81646749e-01 -1.76633513e+00 5.82723439e-01
4.83975768e-01 1.22637320e+00 3.03805470e-01 8.51435602e-01
3.90985161e-01 8.95130634e-01 -3.94031227e-01 -4.37922150e-01
1.93315789e-01 -4.52640355e-01 -6.41355813e-01 6.39229268e-02
9.49082673e-01 -5.70291698e-01 -7.37029493e-01 8.33589375e-01
6.59397364e-01 4.04350817e-01 4.38735157e-01 -1.26864851e+00
-6.96827710e-01 3.10974598e-01 -6.84883952e-01 7.34098017e-01
7.59420514e-01 1.86554432e-01 1.39920890e+00 -1.28855038e+00
1.12621307e+00 1.10067081e+00 3.41128141e-01 3.36410820e-01
-9.43861067e-01 -3.24158520e-01 1.70925170e-01 4.29241747e-01
-1.34082997e+00 -7.11007774e-01 8.63139391e-01 -4.09351796e-01
9.79255319e-01 4.11923587e-01 7.92868435e-01 1.36629117e+00
4.24940176e-02 1.33914530e+00 3.97247076e-01 -1.77609444e-01
-4.15585823e-02 -1.02962613e-01 -2.88811833e-01 7.25852609e-01
-2.89941788e-01 -2.61117369e-01 -9.55336928e-01 -7.81697184e-02
6.45781934e-01 2.37968177e-01 -4.72791374e-01 -6.35772467e-01
-1.59786808e+00 7.49483645e-01 4.38044995e-01 1.21036723e-01
-3.99668664e-01 3.22026968e-01 5.54864705e-01 2.04121351e-01
6.39298856e-01 4.13468093e-01 -4.82416689e-01 -4.50339079e-01
-1.25583339e+00 4.38618213e-01 5.04115701e-01 1.18366194e+00
6.63595498e-01 -2.31547747e-02 -5.08830249e-01 6.69721603e-01
2.09379774e-02 6.26235008e-01 3.33803773e-01 -9.91539717e-01
4.49677289e-01 3.09972227e-01 2.24872276e-01 -1.27095866e+00
-1.23970345e-01 -2.52471447e-01 -1.04311675e-01 -3.60389084e-01
1.22457370e-01 -3.11078578e-02 -8.93577933e-01 1.75929713e+00
2.51155347e-01 2.81342179e-01 -2.51301765e-01 1.29958439e+00
1.07875633e+00 9.28834558e-01 -5.28179258e-02 -2.40130071e-02
1.61179495e+00 -1.18048143e+00 -5.60277045e-01 -1.61479369e-01
2.70955920e-01 -9.20098841e-01 1.33412051e+00 2.59732574e-01
-1.25889957e+00 -4.00021166e-01 -7.36499906e-01 -7.69033730e-01
-3.75775129e-01 2.67073423e-01 4.61263895e-01 -8.55102986e-02
-1.16504061e+00 2.85171419e-01 -8.28098118e-01 -6.57070637e-01
3.34402055e-01 3.29400005e-04 -2.68014193e-01 -8.57754145e-03
-1.21293879e+00 4.89893585e-01 2.88664680e-02 -2.21271724e-01
-1.02729654e+00 -7.38938093e-01 -1.17810714e+00 1.25635620e-02
7.85844743e-01 -5.45821309e-01 1.64403784e+00 -1.06040287e+00
-1.06670380e+00 1.02290356e+00 -5.46832204e-01 -5.71059346e-01
5.88900506e-01 -7.92308033e-01 -4.91263330e-01 7.18572259e-01
5.21278381e-01 1.15178192e+00 9.19810474e-01 -7.63462961e-01
-7.71105468e-01 6.54288009e-02 1.68662280e-01 3.89348954e-01
-3.34207892e-01 3.41141552e-01 -1.14998555e+00 -9.70898151e-01
-2.08924964e-01 -8.63777220e-01 5.06608896e-02 1.51307806e-01
-5.32013893e-01 -5.43046556e-02 1.11098766e+00 -8.30598772e-01
1.23825192e+00 -2.25444865e+00 1.95304230e-01 -1.37151793e-01
1.21538915e-01 -1.61292404e-01 -2.89439857e-01 5.88434279e-01
6.87611997e-02 -7.64254853e-02 4.66860421e-02 -2.52626151e-01
9.51158553e-02 -2.02222764e-01 -6.41619861e-01 4.31475282e-01
2.79431999e-01 1.13264358e+00 -1.11988497e+00 -6.90247715e-01
2.20056832e-01 7.61088610e-01 -7.80404389e-01 1.82246923e-01
-4.47078913e-01 1.82653591e-01 -3.94305110e-01 1.01308095e+00
4.64769118e-02 -4.55377638e-01 -3.41183871e-01 -6.36868775e-01
-1.36102930e-01 2.42937863e-01 -8.39595556e-01 2.07756996e+00
-6.98670223e-02 1.21723378e+00 -2.26313472e-01 -6.02060258e-01
4.90429491e-01 2.41996258e-01 8.15599144e-01 -7.92934895e-01
-1.55455843e-01 1.38415530e-01 -5.12787402e-01 -8.48415434e-01
9.83469248e-01 3.98627549e-01 -3.97895485e-01 2.46254876e-01
2.72073686e-01 3.53177398e-04 5.89321315e-01 6.01013243e-01
1.12936878e+00 4.17792857e-01 2.33572334e-01 1.14643795e-03
1.72838792e-01 4.78109494e-02 4.91101623e-01 5.86860061e-01
-2.65466928e-01 1.03725088e+00 6.07897222e-01 -4.39087242e-01
-1.00966299e+00 -1.10244131e+00 2.42109478e-01 1.71478927e+00
2.16868177e-01 -6.35800838e-01 -4.47418511e-01 -5.18033803e-01
2.80893613e-02 4.97099668e-01 -6.44107580e-01 1.06328063e-01
-5.87891221e-01 -2.15825792e-02 2.55152434e-01 3.74304503e-01
3.20333749e-01 -1.08917928e+00 -9.16671455e-01 -1.06128044e-01
-6.48144782e-01 -1.23722351e+00 -1.21837318e+00 -2.14005187e-01
-4.24600959e-01 -1.11695409e+00 -1.21848726e+00 -7.62023807e-01
4.28103358e-01 6.44396186e-01 1.37464011e+00 -1.88486308e-01
-4.75269437e-01 7.61397064e-01 -4.62111861e-01 -3.32823873e-01
2.23799869e-01 8.39381665e-02 -8.75758752e-02 6.22656010e-02
4.13492203e-01 -2.89384097e-01 -1.07570481e+00 3.68071437e-01
-9.74163711e-01 1.87285975e-01 4.86940503e-01 4.69639748e-01
8.17458510e-01 -5.79813898e-01 2.41159737e-01 -5.33952773e-01
5.01303136e-01 -5.80256581e-01 -3.69008332e-01 1.36454016e-01
5.02666086e-02 -2.93303758e-01 2.56313890e-01 -3.43881845e-01
-6.74047589e-01 2.46726170e-01 1.22966550e-01 -8.30416799e-01
4.42483872e-02 4.98682261e-01 7.08345398e-02 2.67668366e-01
7.02205062e-01 2.64821649e-01 -2.84608334e-01 -3.20757538e-01
4.46739823e-01 3.79262894e-01 9.39498246e-01 -3.27414155e-01
6.21955752e-01 6.15470529e-01 -3.96097451e-01 -1.02861583e+00
-1.16470230e+00 -8.84269595e-01 -4.50769573e-01 -4.90022570e-01
8.40516090e-01 -1.32869411e+00 -3.64689559e-01 -6.03108741e-02
-1.03733242e+00 -3.01241040e-01 -2.67317861e-01 4.66110855e-01
-7.77445793e-01 3.17250937e-01 -6.60427630e-01 -5.12075961e-01
-3.50203186e-01 -1.12170410e+00 1.63639724e+00 9.78367627e-02
-3.50452155e-01 -6.83412969e-01 -1.09228529e-01 2.92944610e-01
2.95595616e-01 6.48461103e-01 2.95744598e-01 -5.94112635e-01
-8.22520733e-01 -3.34678113e-01 -2.39323303e-01 -1.27488494e-01
-1.31772906e-01 1.51752055e-01 -7.75207698e-01 -2.91687429e-01
-3.26978445e-01 -6.24277532e-01 1.14033747e+00 4.97511953e-01
1.16767752e+00 -4.97615397e-01 -2.43227214e-01 6.06554329e-01
1.15036654e+00 -1.94795638e-01 5.13861358e-01 2.63662875e-01
6.66497648e-01 4.33749229e-01 8.50999177e-01 6.59027457e-01
5.48153520e-01 8.45327914e-01 3.83360863e-01 -1.80809245e-01
-3.42854485e-02 -5.40242255e-01 6.68985009e-01 4.58265185e-01
-1.26772493e-01 -3.02035987e-01 -7.26112008e-01 7.74959803e-01
-1.93323946e+00 -1.51445508e+00 1.27940685e-01 1.91737854e+00
6.77573264e-01 2.89877076e-02 5.08143544e-01 -3.83690476e-01
6.68792486e-01 7.11896718e-01 -9.07023668e-01 2.33240630e-02
-2.19086051e-01 -3.27337980e-01 2.42470846e-01 2.31099904e-01
-1.41622603e+00 9.84703422e-01 5.58030987e+00 6.53950155e-01
-1.26556969e+00 1.32134557e-01 6.64365351e-01 -7.63796449e-01
-3.22524846e-01 -1.53405517e-01 -5.25331438e-01 4.71247822e-01
7.21545696e-01 -1.87318310e-01 3.22899491e-01 1.07449841e+00
5.04623592e-01 -1.09513432e-01 -1.22036588e+00 1.19906855e+00
5.97351670e-01 -1.47179925e+00 8.89572948e-02 -1.58804893e-01
7.72043824e-01 4.04747695e-01 3.59968036e-01 3.88192505e-01
-2.17831954e-01 -6.39951587e-01 1.10470080e+00 5.66905737e-01
1.08079696e+00 -4.87735182e-01 2.97125965e-01 -2.81465761e-02
-1.29725671e+00 9.52949654e-03 -1.76817060e-01 1.30339578e-01
4.54128534e-01 4.53691602e-01 -6.10331416e-01 7.16822445e-02
9.60392416e-01 1.11717284e+00 -7.97793090e-01 1.19978309e+00
-1.67855933e-01 4.50953454e-01 -3.79002362e-01 -2.02473626e-02
3.40658933e-01 2.64541388e-01 6.69129670e-01 1.64281213e+00
5.47663748e-01 -2.70835999e-02 2.13787585e-01 6.63495362e-01
-2.31154695e-01 1.21706419e-01 -6.42505646e-01 -1.61544845e-01
2.84987628e-01 1.24023175e+00 -8.26232910e-01 -5.21012306e-01
-5.18152893e-01 1.30200851e+00 1.42327875e-01 6.41429186e-01
-1.08715534e+00 -5.47790766e-01 5.42210340e-01 3.27122241e-01
4.52134907e-01 -7.07039014e-02 4.98383909e-01 -1.53770852e+00
2.28266835e-01 -7.47847974e-01 6.06903195e-01 -1.22025192e+00
-1.02489197e+00 4.92009431e-01 1.27601936e-01 -1.45441163e+00
-5.54537475e-01 -3.14194769e-01 -5.06298780e-01 3.47924620e-01
-1.53911722e+00 -1.02898729e+00 -4.12802100e-01 8.25727940e-01
1.08249187e+00 7.10191234e-05 2.76895016e-01 4.26235229e-01
-3.15730900e-01 3.82085919e-01 -2.93798894e-02 1.86072707e-01
1.00468969e+00 -1.15415788e+00 5.05831957e-01 9.29340303e-01
4.73684251e-01 4.41506624e-01 9.42064762e-01 -6.17196798e-01
-1.52876997e+00 -1.13840759e+00 9.68394637e-01 -6.60652578e-01
6.74542904e-01 -4.84533191e-01 -5.65873802e-01 8.11353505e-01
3.54686856e-01 2.23085895e-01 6.16385758e-01 -1.48271471e-01
-2.60186255e-01 1.25010610e-01 -6.84969902e-01 8.44626069e-01
1.16209936e+00 -8.33866417e-01 -4.50086355e-01 6.98655248e-01
9.09779489e-01 -7.71501958e-01 -5.16271412e-01 1.44619524e-01
6.43830478e-01 -1.14376497e+00 1.09752023e+00 -4.09198880e-01
7.23114371e-01 -2.17389181e-01 -2.01919079e-01 -9.74258304e-01
-8.61609355e-02 -9.20985460e-01 -4.83338326e-01 8.40015173e-01
3.74066919e-01 7.94219524e-02 6.74173474e-01 6.29466116e-01
-3.83036286e-01 -9.43198144e-01 -6.16331279e-01 -4.75338876e-01
-6.63425088e-01 -5.33295810e-01 2.37846896e-01 7.90484428e-01
9.97654274e-02 3.90695065e-01 -7.10161507e-01 -3.73169631e-02
3.47414881e-01 3.81353498e-01 1.00480425e+00 -8.30031693e-01
-1.63662598e-01 -4.05663699e-01 -3.74599874e-01 -1.37502897e+00
-1.06091939e-01 -6.73515677e-01 1.67853199e-02 -1.66280091e+00
3.92299801e-01 4.45976675e-01 -2.67939478e-01 3.48096699e-01
1.50967587e-03 5.54547012e-01 3.41911167e-01 4.12723839e-01
-1.52284575e+00 4.35206562e-01 1.12832260e+00 4.10987958e-02
-2.59198129e-01 -2.31303588e-01 -6.08217120e-01 7.06794024e-01
5.22170842e-01 -1.57821536e-01 -4.15143758e-01 -5.22939205e-01
3.68381858e-01 1.69552937e-01 6.01402104e-01 -9.69367445e-01
2.26195768e-01 -8.57876092e-02 5.07239759e-01 -8.80192041e-01
6.51673138e-01 -4.38384533e-01 -1.69454664e-01 6.99738041e-02
-6.31692052e-01 4.48038161e-01 1.82444174e-02 7.01555550e-01
-3.59108716e-01 2.81115890e-01 3.94529104e-01 -1.21521465e-01
-1.14713752e+00 4.55119103e-01 -6.19677007e-02 3.29543799e-01
9.30460393e-01 -2.90919781e-01 -3.39769959e-01 -9.25678015e-01
-6.83726788e-01 3.99264246e-01 5.77545702e-01 7.11449206e-01
8.82623255e-01 -1.32414234e+00 -6.67382717e-01 -7.74710998e-02
3.59008491e-01 -1.67869687e-01 4.10403848e-01 9.28043783e-01
-4.58779544e-01 5.67997277e-01 1.86160468e-02 -6.75789714e-01
-1.14177501e+00 7.57936656e-01 -1.30972072e-01 1.07816093e-01
-7.06477046e-01 8.80251527e-01 3.27856183e-01 1.34962782e-01
4.10660356e-01 -4.79299873e-01 -1.30216584e-01 1.89710036e-01
7.13352919e-01 2.00734884e-01 -2.61180937e-01 -1.04586029e+00
-2.86009073e-01 4.39865738e-01 -1.06326409e-01 -1.88061744e-01
1.29779506e+00 -3.48694623e-01 3.63281429e-01 4.58039016e-01
1.61726737e+00 2.07119808e-01 -1.49577701e+00 -3.27364773e-01
1.27354208e-02 -6.19902194e-01 -1.87080249e-01 -5.25195003e-01
-1.04245496e+00 7.48571217e-01 3.88970762e-01 1.20938979e-01
1.11522996e+00 2.64974087e-01 1.10339630e+00 5.32662213e-01
9.99091193e-02 -1.19654858e+00 6.07967317e-01 3.56124520e-01
1.20191097e+00 -1.43666220e+00 1.32113874e-01 -2.12596819e-01
-1.07217538e+00 9.63868737e-01 5.40651739e-01 -2.18709737e-01
4.02022630e-01 -2.35661387e-01 -1.19490176e-02 -4.10920918e-01
-9.49180543e-01 -4.82976794e-01 7.29115725e-01 2.95891732e-01
5.38437903e-01 -2.60464758e-01 2.26847036e-03 3.11915517e-01
-3.41576248e-01 -1.55653777e-02 4.15779740e-01 8.53986084e-01
-5.53387046e-01 -4.02623147e-01 -2.16026649e-01 5.43758810e-01
-6.78060651e-01 -1.23501852e-01 -3.72575283e-01 7.22395241e-01
-2.97493070e-01 6.96375668e-01 1.50596172e-01 -2.24637926e-01
3.86199594e-01 -1.12658493e-01 6.07703589e-02 -5.85739434e-01
-2.31898427e-01 5.16215622e-01 8.83208066e-02 -9.91514981e-01
-4.65511590e-01 -7.84160376e-01 -1.13581514e+00 1.29169464e-01
6.91692159e-02 9.44671184e-02 3.95699382e-01 4.59363192e-01
7.23332644e-01 2.99816608e-01 3.56653571e-01 -1.38538516e+00
-1.18609853e-01 -6.93729639e-01 -3.32291812e-01 7.62104332e-01
5.85178792e-01 -7.26229131e-01 -3.39421004e-01 4.55324084e-01] | [10.131299018859863, 0.6682005524635315] |
597df21d-072f-46bd-8df1-1089eea3b718 | 0-1-deep-neural-networks-via-block-coordinate | 2206.09379 | null | https://arxiv.org/abs/2206.09379v1 | https://arxiv.org/pdf/2206.09379v1.pdf | 0/1 Deep Neural Networks via Block Coordinate Descent | The step function is one of the simplest and most natural activation functions for deep neural networks (DNNs). As it counts 1 for positive variables and 0 for others, its intrinsic characteristics (e.g., discontinuity and no viable information of subgradients) impede its development for several decades. Even if there is an impressive body of work on designing DNNs with continuous activation functions that can be deemed as surrogates of the step function, it is still in the possession of some advantageous properties, such as complete robustness to outliers and being capable of attaining the best learning-theoretic guarantee of predictive accuracy. Hence, in this paper, we aim to train DNNs with the step function used as an activation function (dubbed as 0/1 DNNs). We first reformulate 0/1 DNNs as an unconstrained optimization problem and then solve it by a block coordinate descend (BCD) method. Moreover, we acquire closed-form solutions for sub-problems of BCD as well as its convergence properties. Furthermore, we also integrate $\ell_{2,0}$-regularization into 0/1 DNN to accelerate the training process and compress the network scale. As a result, the proposed algorithm has a high performance on classifying MNIST and Fashion-MNIST datasets. | ['Naihua Xiu', 'Geoffrey Ye Li', 'Shenglong Zhou', 'HUI ZHANG'] | 2022-06-19 | null | null | null | null | ['rotated-mnist'] | ['computer-vision'] | [-3.06377187e-02 -7.20557645e-02 -4.01065618e-01 -4.78249609e-01
-2.63083994e-01 -1.39010459e-01 1.56576112e-01 1.66248158e-02
-7.44400501e-01 9.53822732e-01 -4.39633280e-01 -4.62825477e-01
-4.40103203e-01 -8.50762963e-01 -6.90299630e-01 -1.11481750e+00
5.62340580e-02 2.24001974e-01 -8.68374333e-02 -1.66629270e-01
9.67569873e-02 7.25939870e-01 -1.27461052e+00 -4.25006956e-01
1.28543079e+00 1.54069316e+00 1.46703541e-01 5.34105971e-02
-8.46714154e-02 8.15566361e-01 -5.45787215e-01 -4.03181016e-01
4.56488878e-01 -3.33869010e-01 -4.39357609e-01 -1.45056263e-01
2.14465261e-01 -2.32892215e-01 -5.46549678e-01 1.43946779e+00
5.09678066e-01 3.39043826e-01 5.03912270e-01 -1.34902334e+00
-7.09362984e-01 3.78487438e-01 -6.92507267e-01 1.69556066e-01
-3.96986634e-01 -2.38704626e-02 1.02001333e+00 -7.05770552e-01
2.31581643e-01 8.61512721e-01 5.38969219e-01 5.08169711e-01
-1.18708766e+00 -6.24615192e-01 3.11799496e-01 2.54856080e-01
-1.47361672e+00 -3.28840762e-01 8.99896026e-01 -2.71918058e-01
4.61609989e-01 3.18225533e-01 6.74267828e-01 8.96596491e-01
-2.29802225e-02 9.52306151e-01 6.95800900e-01 -1.41727254e-01
3.85476381e-01 1.16987467e-01 1.74362287e-01 5.83821595e-01
3.93471479e-01 -7.45754987e-02 -1.96208298e-01 9.13382322e-02
9.45935428e-01 1.69694662e-01 -4.12734360e-01 -2.18779534e-01
-8.93444121e-01 8.70574355e-01 7.01654792e-01 2.55078286e-01
-4.95462358e-01 1.36092126e-01 3.35887432e-01 2.85551429e-01
4.49956417e-01 3.24054062e-01 -2.30265856e-01 -1.91344898e-02
-8.95803332e-01 2.64464498e-01 4.64239359e-01 6.80903375e-01
5.72094560e-01 5.22620738e-01 -4.44797613e-02 1.06017578e+00
1.00022718e-01 2.47597024e-01 5.80310583e-01 -7.57194519e-01
6.01406455e-01 7.79064536e-01 -6.65545538e-02 -1.28961086e+00
-5.49377143e-01 -8.57785225e-01 -1.41497290e+00 1.87172800e-01
4.83111918e-01 -2.48768166e-01 -6.08982027e-01 1.84742451e+00
1.96838468e-01 5.78683540e-02 -2.92682293e-04 1.00577891e+00
6.75191700e-01 7.02700794e-01 -2.27245152e-01 -3.69348079e-01
8.22132707e-01 -7.04475045e-01 -7.18174219e-01 -1.71227217e-01
5.78049541e-01 -2.18501315e-01 1.06970286e+00 5.52079797e-01
-1.08077860e+00 -6.06134892e-01 -1.12589884e+00 -7.29742125e-02
-9.24865082e-02 4.26919222e-01 6.53163314e-01 4.43655849e-01
-9.88999546e-01 6.37628138e-01 -8.74159634e-01 2.25645766e-01
6.04552627e-01 6.80589437e-01 -1.41601712e-01 6.32722443e-03
-1.14401436e+00 6.37867808e-01 7.28912234e-01 6.65226460e-01
-5.93785822e-01 -5.83601654e-01 -5.74482918e-01 7.86562040e-02
4.46082205e-01 -1.77313223e-01 7.92809367e-01 -1.13184857e+00
-1.43917537e+00 5.19285381e-01 -3.18408497e-02 -5.68894029e-01
6.79511905e-01 -1.24881588e-01 -3.64143550e-01 -1.47197640e-03
-1.16871767e-01 5.25180995e-01 8.43231797e-01 -9.32222366e-01
-6.44219577e-01 -4.24992234e-01 1.40015140e-01 1.22080281e-01
-8.54083061e-01 -2.91942835e-01 -2.90413618e-01 -7.99867094e-01
4.26152945e-01 -7.33939111e-01 -1.83164433e-01 2.49717832e-01
-4.94016886e-01 -3.66562247e-01 6.14306867e-01 -6.47432685e-01
1.51278579e+00 -2.26862311e+00 1.43314525e-01 3.44063520e-01
2.66229182e-01 5.02831280e-01 -4.23040874e-02 -7.49968812e-02
-7.86466673e-02 -5.39742894e-02 -3.83102298e-01 -8.75951201e-02
-1.54189050e-01 4.63764757e-01 -7.30288774e-02 6.71503663e-01
2.58052260e-01 7.66337395e-01 -7.54382730e-01 -2.47648358e-01
6.62494302e-02 6.08602345e-01 -5.64842701e-01 -1.72873467e-01
-8.89924914e-02 3.84395361e-01 -5.57795763e-01 4.83194649e-01
7.43903518e-01 -1.26841292e-01 -3.35036516e-02 -1.30836144e-01
-2.06538096e-01 -1.70484364e-01 -1.38002706e+00 1.17080712e+00
-2.24967778e-01 6.05287075e-01 3.83421332e-01 -1.67584872e+00
1.31916809e+00 2.25590065e-01 7.21463978e-01 -7.83918560e-01
2.25154862e-01 3.75192046e-01 1.36021152e-01 -3.52054566e-01
2.16202706e-01 -1.73800617e-01 2.64344275e-01 2.68006027e-02
-1.71060935e-01 2.30443209e-01 3.55819821e-01 -1.86291322e-01
6.56329393e-01 -4.34475422e-01 4.64377031e-02 -4.45868820e-01
8.81498873e-01 -2.31945485e-01 8.87079060e-01 5.15824437e-01
-8.84474814e-02 5.00318646e-01 8.74760211e-01 -4.72908586e-01
-9.63673294e-01 -7.81658709e-01 -4.71245199e-01 8.21598649e-01
1.03750326e-01 1.44204974e-01 -6.41964912e-01 -2.78061539e-01
-8.41505155e-02 6.53385818e-01 -4.50787753e-01 -2.98029184e-01
-7.64504552e-01 -9.93165672e-01 5.52815497e-01 6.33955181e-01
9.88208055e-01 -7.97608435e-01 -2.53101498e-01 3.00242633e-01
4.07490395e-02 -8.56791198e-01 -2.41694525e-01 4.31466699e-01
-1.01548433e+00 -9.84805703e-01 -9.81412232e-01 -1.02889538e+00
9.14615989e-01 5.10922410e-02 7.72694409e-01 1.05102070e-01
2.29454506e-02 -3.16437811e-01 -5.02415188e-02 -2.60712862e-01
-1.18731819e-01 2.06130758e-01 2.66039431e-01 3.06914717e-01
2.09350307e-02 -5.66234350e-01 -5.48557818e-01 3.28626156e-01
-1.14402843e+00 -1.42322835e-02 5.35871327e-01 1.01894581e+00
8.26342821e-01 5.38834989e-01 7.85654783e-01 -7.15019464e-01
8.65394413e-01 -3.39478016e-01 -9.20879841e-01 1.66648224e-01
-6.60762072e-01 -5.99508062e-02 1.30769348e+00 -5.09628534e-01
-8.29601705e-01 -1.74478646e-02 -4.46082324e-01 -4.61910874e-01
1.17503539e-01 6.40444696e-01 -2.63822168e-01 -8.07490423e-02
5.47463059e-01 4.12882864e-01 4.83044945e-02 -4.93859947e-01
6.96110651e-02 4.19793665e-01 5.32875896e-01 -6.91742778e-01
6.80798054e-01 3.60692441e-01 2.56686866e-01 -9.30345237e-01
-9.77234602e-01 -1.19602941e-01 -4.69780952e-01 -9.17135626e-02
6.56115055e-01 -5.69011092e-01 -1.03700602e+00 5.76388180e-01
-8.34723175e-01 -1.55108735e-01 -3.46172899e-01 4.56320196e-01
-1.79525152e-01 2.14528352e-01 -5.84794581e-01 -8.69275093e-01
-2.66031504e-01 -1.17141318e+00 1.69761002e-01 3.50858212e-01
2.48173863e-01 -1.04448140e+00 -3.47432107e-01 9.05433744e-02
3.76899600e-01 4.58960205e-01 9.43115413e-01 -6.04579866e-01
-2.55809426e-01 -4.04068738e-01 -3.58810157e-01 9.01423395e-01
-4.00667191e-02 -2.69961078e-02 -7.20913351e-01 -4.40571308e-01
4.28637296e-01 -2.78705865e-01 8.98848712e-01 6.18862510e-01
1.57870626e+00 -6.99084222e-01 7.73778558e-03 1.01752126e+00
1.37322354e+00 5.67077994e-01 4.18549925e-01 4.64631677e-01
7.57353842e-01 3.90331686e-01 2.16638327e-01 4.11898255e-01
7.94417597e-03 3.50888371e-01 6.43313527e-01 -1.64450675e-01
4.63669360e-01 -2.04392355e-02 1.01112053e-01 9.45936382e-01
-2.04720348e-01 -5.42979836e-02 -7.03949571e-01 2.84079581e-01
-1.85428417e+00 -7.86768854e-01 -2.33519301e-01 2.28089762e+00
7.98256934e-01 3.55775207e-01 6.02693409e-02 5.81671059e-01
7.74162531e-01 1.93938106e-01 -1.07731509e+00 -1.38889924e-01
-4.48073715e-01 9.40873995e-02 4.76292223e-01 1.54794589e-01
-1.09246016e+00 4.82654870e-01 5.04003143e+00 9.67921257e-01
-1.47567475e+00 -1.81704924e-01 1.11139607e+00 -2.11509034e-01
-2.70309802e-02 -4.15299445e-01 -8.57257426e-01 6.68691158e-01
4.37304765e-01 5.67851448e-03 6.41341627e-01 1.01128542e+00
4.18029636e-01 1.73948482e-01 -7.77486145e-01 1.03671455e+00
-2.09251657e-01 -1.18544424e+00 -1.21153891e-01 1.14645166e-02
9.73350167e-01 -8.40998590e-02 2.71315247e-01 2.61383772e-01
-1.73194781e-01 -1.17312169e+00 8.21981847e-01 6.90039173e-02
7.71746874e-01 -1.06411862e+00 8.04289758e-01 5.61399877e-01
-9.60683405e-01 -3.53280991e-01 -7.79086053e-01 -9.96947959e-02
-9.99017060e-02 8.79179060e-01 -2.45094880e-01 5.44249356e-01
4.56564218e-01 7.83040464e-01 -2.95198321e-01 1.11914384e+00
-1.38327986e-01 5.53720415e-01 -5.16221941e-01 -1.85263053e-01
4.65151936e-01 -7.23601162e-01 2.62786210e-01 7.70892143e-01
3.93320024e-01 1.53031439e-01 1.37182130e-02 7.76448011e-01
-3.18837732e-01 2.08908558e-01 -2.84468889e-01 -2.32314989e-02
4.74834263e-01 9.80063915e-01 -7.45248437e-01 -1.29896492e-01
-3.51009578e-01 5.31355441e-01 4.70327944e-01 5.80302835e-01
-7.23682284e-01 -5.07174850e-01 7.10475683e-01 6.00698404e-02
1.51811689e-01 -2.36633539e-01 -6.97330832e-01 -1.11906183e+00
4.06911105e-01 -8.00480247e-01 3.76192331e-01 -2.14825898e-01
-1.18011153e+00 7.14042187e-01 -2.59282708e-01 -1.10104156e+00
3.95839140e-02 -8.94608498e-01 -5.41222513e-01 7.13471115e-01
-1.48998404e+00 -4.38247323e-01 -2.64566779e-01 7.71481037e-01
2.44252726e-01 -1.24240831e-01 3.32687408e-01 6.52080178e-01
-1.08619022e+00 6.65051758e-01 6.73579037e-01 3.62171978e-01
7.06253350e-02 -9.59106863e-01 -1.72490522e-01 7.46939659e-01
-1.91393614e-01 5.73981285e-01 5.30316949e-01 -1.78821683e-01
-1.22943687e+00 -1.06651592e+00 4.96993542e-01 4.00684655e-01
6.50384724e-01 -2.36424655e-01 -1.13123178e+00 5.16326189e-01
-4.49539930e-01 1.71049327e-01 1.22865111e-01 -9.81178135e-02
1.03390001e-01 -6.27783597e-01 -1.04333520e+00 6.39634073e-01
8.77827227e-01 -1.21737219e-01 -1.22104459e-01 3.30953836e-01
4.40207720e-01 -4.77861136e-01 -8.78558874e-01 5.19227087e-01
3.21105152e-01 -9.41138387e-01 1.06230938e+00 -4.65747029e-01
4.18539256e-01 -8.15856308e-02 4.57815081e-02 -1.14059067e+00
-2.02451184e-01 -6.31564796e-01 -3.21702480e-01 1.12520397e+00
2.82366216e-01 -9.98250663e-01 1.05781364e+00 6.24581158e-01
-1.72808766e-01 -1.34760106e+00 -1.24308729e+00 -8.67297649e-01
2.38039732e-01 -4.22708780e-01 3.06207836e-01 9.46992576e-01
-3.87011617e-01 8.94551203e-02 -3.32308322e-01 -1.55078396e-01
6.32804275e-01 -1.46229312e-01 3.54443818e-01 -1.40567183e+00
-3.98276635e-02 -8.23779404e-01 -2.89199412e-01 -1.33103991e+00
1.61717772e-01 -8.77741039e-01 -1.56166986e-01 -1.35059702e+00
-3.02221537e-01 -7.95904160e-01 -5.65110207e-01 4.38248724e-01
-6.51465580e-02 2.33464204e-02 -2.49404251e-03 3.06701899e-01
-1.08117498e-01 7.70776808e-01 1.32593286e+00 -9.05966163e-02
-3.94842654e-01 3.89422148e-01 -7.90524840e-01 7.64302135e-01
1.02649486e+00 -3.75808746e-01 -4.58244294e-01 -4.89421844e-01
2.27789626e-01 1.84353348e-02 2.34814644e-01 -1.04546702e+00
2.38312781e-01 -2.40510657e-01 4.25771177e-01 -5.17868578e-01
4.00285989e-01 -7.34989643e-01 -1.91400051e-01 5.81925988e-01
-4.28057969e-01 1.32939026e-01 -4.41175327e-02 3.46569300e-01
-4.71648067e-01 -4.96869743e-01 1.04033637e+00 -2.70913076e-02
-3.79050106e-01 5.71846962e-01 -1.39314011e-01 2.68187433e-01
8.72049332e-01 -3.10042113e-01 -8.37094486e-02 -1.97720990e-01
-3.16241294e-01 3.17079246e-01 1.57517925e-01 5.36788292e-02
6.13932550e-01 -1.41833103e+00 -6.68674231e-01 3.47385883e-01
-4.01879907e-01 5.64643621e-01 1.64676487e-01 9.35600996e-01
-6.98014796e-01 4.35730845e-01 -1.48360774e-01 -5.73660970e-01
-7.16233134e-01 3.90747845e-01 3.94844681e-01 -2.42038116e-01
-6.05581880e-01 9.58409727e-01 1.96301550e-01 -3.17854702e-01
6.66880727e-01 -2.88697153e-01 -1.28512353e-01 1.74889877e-03
3.80881578e-01 5.58931291e-01 1.36022255e-01 -3.83769184e-01
-3.12024146e-01 2.83363104e-01 1.10664077e-01 1.96674362e-01
1.66433227e+00 1.31335929e-01 -2.88577348e-01 2.18816444e-01
1.43629587e+00 -4.78434950e-01 -1.50931692e+00 -2.83267617e-01
-1.73337404e-02 -2.02376932e-01 2.71085113e-01 -3.50556791e-01
-1.55166197e+00 9.11947012e-01 4.64433968e-01 3.94107312e-01
1.26260102e+00 -3.99502397e-01 9.19751108e-01 6.18698537e-01
-3.56609784e-02 -1.37243080e+00 -4.02333960e-02 6.02558732e-01
7.79767096e-01 -1.04673886e+00 -1.15775606e-02 9.33476612e-02
-3.49203736e-01 1.35679710e+00 7.02627897e-01 -3.54348511e-01
5.39785087e-01 8.34155902e-02 5.82191497e-02 1.03077270e-01
-3.73631835e-01 1.60685837e-01 2.02537283e-01 2.39614457e-01
3.07636738e-01 -1.23391524e-01 -5.47182143e-01 4.71080482e-01
-2.75215447e-01 1.60809103e-02 1.79236766e-03 5.88331759e-01
-3.28774661e-01 -6.62175179e-01 -2.76572704e-01 7.37574100e-01
-4.55425143e-01 -1.34902019e-02 4.40964438e-02 8.98094714e-01
1.34079412e-01 6.93404675e-01 -5.80893010e-02 -2.61359602e-01
3.98888707e-01 -1.92248762e-01 1.91104636e-01 -9.39678028e-02
-4.48875546e-01 -2.32877970e-01 -3.35994601e-01 -2.42332488e-01
-6.00333847e-02 -3.79666567e-01 -1.28400135e+00 -4.18401122e-01
-3.65463912e-01 1.61242291e-01 7.08055913e-01 9.10184562e-01
-5.01246005e-02 4.91599351e-01 7.01384127e-01 -6.37734890e-01
-8.69468093e-01 -7.51242936e-01 -7.15920687e-01 3.20939779e-01
3.68629217e-01 -5.19223392e-01 -3.84629518e-01 -3.77590477e-01] | [8.005598068237305, 3.599343776702881] |
cc7afce9-2880-49db-89d8-06b4b1b3c367 | distilprotbert-a-distilled-protein-language | null | null | https://www.biorxiv.org/content/10.1101/2022.05.09.491157v1 | https://www.biorxiv.org/content/10.1101/2022.05.09.491157v1.full.pdf | DistilProtBert: A distilled protein language model used to distinguish between real proteins and their randomly shuffled counterparts | Recently, Deep Learning models, initially developed in the field of Natural Language Processing (NLP), were applied successfully to analyze protein sequences. A major drawback of these models is their size in terms of the number of parameters needed to be fitted and the amount of computational resources they require. Recently, "distilled" models using the concept of student and teacher networks have been widely used in NLP. Here, we adapted this concept to the problem of protein sequence analysis, by developing DistilProtBert, a distilled version of the successful ProtBert model. Implementing this approach, we reduced the size of the network and the running time by 50%, and the computational resources needed for pretraining by 98% relative to ProtBert model. Using two published tasks, we showed that the performance of the distilled model approaches that of the full model. We next tested the ability of DistilProtBert to distinguish between real and random protein sequences. The task is highly challenging if the composition is maintained on the level of singlet, doublet and triplet amino acids. Indeed, traditional machine learning algorithms have difficulties with this task. Here, we show that DistilProtBert preforms very well on singlet, doublet, and even triplet-shuffled versions of the human proteome, with AUC of 0.92, 0.91, and 0.87 respectively. Finally, we suggest that by examining the small number of false-positive classifications (i.e., shuffled sequences classified as proteins by DistilProtBert) we may be able to identify de-novo potential natural-like proteins based on random shuffling of amino acid sequences. | ['Ron Unger', 'Yanay Ofran', 'Yaron Geffen'] | 2022-05-10 | null | null | null | biorxiv-2022-5 | ['protein-language-model', 'protein-secondary-structure-prediction'] | ['medical', 'medical'] | [ 3.42396975e-01 2.43354797e-01 1.79731086e-01 -4.81521875e-01
-4.54772383e-01 -7.63285935e-01 3.35441917e-01 5.63634455e-01
-7.26402342e-01 1.30709028e+00 -5.13397694e-01 -5.51396906e-01
-4.09970060e-02 -5.56858540e-01 -1.11658335e+00 -9.39689636e-01
-1.19143657e-01 8.48457634e-01 3.30959201e-01 -2.83738464e-01
2.43160613e-02 6.55591965e-01 -1.41815448e+00 5.76026082e-01
9.44530249e-01 6.66111171e-01 3.34866226e-01 6.25689745e-01
-3.02285433e-01 5.37290156e-01 -6.40055180e-01 -3.92177194e-01
1.15002275e-01 -3.10589582e-01 -8.36200893e-01 -3.64900917e-01
2.14425504e-01 1.38804883e-01 1.39183238e-01 9.73548472e-01
5.04554868e-01 -4.45166864e-02 6.03066802e-01 -9.56397593e-01
-3.91250432e-01 5.58908403e-01 -3.60656381e-01 1.64049074e-01
2.89630234e-01 3.67839992e-01 7.40801752e-01 -8.12042713e-01
8.81112874e-01 1.10246217e+00 9.13968146e-01 3.47955883e-01
-1.64535284e+00 -5.13960183e-01 -1.39991254e-01 3.61692995e-01
-9.50773120e-01 -7.43369833e-02 2.81295508e-01 -6.23277366e-01
1.52177513e+00 2.39218436e-02 4.65815127e-01 1.08705056e+00
3.57191205e-01 4.82641190e-01 1.22179699e+00 -6.55155003e-01
2.00883389e-01 2.25914776e-01 3.65152627e-01 6.50415778e-01
2.72346795e-01 1.18757874e-01 -3.50886077e-01 -2.83024162e-01
2.64943242e-01 -2.36569330e-01 -1.91628784e-01 -7.11316109e-01
-1.06367850e+00 9.81734633e-01 2.25518093e-01 5.03157377e-01
-5.09338021e-01 -4.06452894e-01 5.66846073e-01 3.65519285e-01
3.38066965e-01 8.29650521e-01 -1.12400520e+00 -7.41756484e-02
-8.66214395e-01 1.45946994e-01 1.04690731e+00 5.76443911e-01
8.67281377e-01 -2.45003909e-01 3.38665843e-01 7.78521717e-01
-1.25059634e-01 4.07781936e-02 8.41926396e-01 -4.38392609e-01
-2.81258207e-02 5.19221544e-01 1.95969284e-01 -7.11960793e-01
-6.60656691e-01 -6.44680187e-02 -6.66246593e-01 1.63524508e-01
7.34258652e-01 -2.25699335e-01 -1.03303826e+00 1.79862547e+00
2.20498130e-01 -2.67474711e-01 2.91746259e-01 6.73343658e-01
5.81080616e-01 7.09639311e-01 1.76729947e-01 -3.12626630e-01
1.36771810e+00 -7.31477439e-01 -3.09353799e-01 5.32275401e-02
6.47782922e-01 -6.76634431e-01 8.86977613e-01 6.27416313e-01
-9.53739762e-01 -5.72362959e-01 -1.08636951e+00 -3.23285721e-02
-8.17279577e-01 8.07549525e-03 5.72965443e-01 4.61859733e-01
-9.29786265e-01 1.08079302e+00 -8.85491490e-01 -6.27040863e-01
1.81262791e-01 8.01235914e-01 -6.55305684e-01 1.90267727e-01
-1.27725220e+00 1.24266028e+00 1.01511061e+00 -1.60833329e-01
-6.45432711e-01 -5.56142569e-01 -4.14119810e-01 2.95332402e-01
1.36340946e-01 -3.26083034e-01 1.22390938e+00 -1.03229642e+00
-1.39291966e+00 1.03865147e+00 2.12276150e-02 -8.65308821e-01
3.55445266e-01 -8.60138163e-02 -1.30212575e-01 2.25001231e-01
-2.37248972e-01 7.09456027e-01 1.49074703e-01 -8.50784361e-01
-5.24769425e-01 -2.93788522e-01 -7.52453059e-02 -2.49416709e-01
1.59920394e-01 -3.32102664e-02 3.79264832e-01 -1.56955957e-01
-2.01003090e-01 -8.94941688e-01 -2.78945833e-01 -1.78378478e-01
-1.28828809e-01 -4.31958765e-01 3.82566571e-01 -5.18704653e-01
2.49699518e-01 -2.08964920e+00 4.15270813e-02 1.06033526e-01
1.94280609e-01 9.26323771e-01 -3.05645645e-01 7.51464725e-01
-6.14291370e-01 1.49685934e-01 -2.86545724e-01 4.53748971e-01
-8.17946419e-02 3.04867208e-01 -1.43676087e-01 3.62766743e-01
3.76832902e-01 8.69657576e-01 -8.72368932e-01 -1.15951173e-01
9.01398733e-02 5.84197342e-01 -2.76786268e-01 2.15722769e-01
-4.96396929e-01 1.61796734e-01 -6.09990843e-02 1.88391939e-01
7.54793644e-01 -4.86987740e-01 7.00022101e-01 -1.50169447e-01
-2.04868138e-01 3.36249441e-01 -4.92066085e-01 1.40333605e+00
-5.62664075e-03 5.76085389e-01 -9.76024047e-02 -1.51344979e+00
1.15278888e+00 3.56135458e-01 6.62897110e-01 -5.37425280e-01
2.86869239e-02 4.14821327e-01 4.72503185e-01 -4.65810061e-01
1.07834816e-01 -4.22292531e-01 4.80548441e-01 1.72861040e-01
3.77304226e-01 2.06076056e-01 4.90341783e-01 -9.06410292e-02
1.06887555e+00 4.53602493e-01 5.53506672e-01 -4.78058696e-01
6.64688408e-01 4.09652859e-01 6.58804536e-01 5.38918495e-01
-2.09627762e-01 2.51709700e-01 8.32149327e-01 -7.52504468e-01
-1.29315460e+00 -8.40897202e-01 -1.33965850e-01 1.01518333e+00
-3.31032395e-01 -2.08807856e-01 -7.27515936e-01 -5.62724590e-01
7.36525655e-02 5.39742589e-01 -3.46514374e-01 -1.71549246e-01
-5.92906773e-01 -1.31370378e+00 6.12788796e-01 1.35547727e-01
8.63000825e-02 -1.30937409e+00 -6.85521305e-01 5.49541354e-01
1.67271383e-02 -9.40188825e-01 2.63400465e-01 9.82300580e-01
-6.30942166e-01 -1.44832671e+00 -6.31568968e-01 -9.66701150e-01
2.77968347e-01 -1.05859391e-01 1.06070399e+00 -1.74022913e-01
-4.81684029e-01 -3.11378092e-01 -4.67632234e-01 -5.47644198e-01
-6.02416337e-01 8.82412940e-02 9.35337842e-02 -4.33761925e-01
8.41172218e-01 -5.79198539e-01 -3.92134964e-01 1.74763039e-01
-9.77551401e-01 7.59644583e-02 8.03938568e-01 1.21746588e+00
5.55395544e-01 -7.56738856e-02 8.80383909e-01 -9.65985119e-01
5.06867230e-01 -3.65821302e-01 -5.75281382e-01 3.52901667e-01
-6.40442848e-01 2.86860287e-01 8.72614145e-01 -6.80885196e-01
-6.78094625e-01 4.48769093e-01 -4.46967363e-01 -6.45647347e-02
-3.11299503e-01 6.35237753e-01 2.51833126e-02 -2.07227960e-01
7.79980302e-01 3.88408065e-01 4.23528582e-01 -6.18516445e-01
2.07684606e-01 5.11080027e-01 3.46589446e-01 -5.70293009e-01
2.66644150e-01 5.27857728e-02 -1.30374730e-02 -7.43530989e-01
-5.79823136e-01 -3.49773169e-01 -8.30313325e-01 4.07143116e-01
7.92230368e-01 -6.22538805e-01 -1.07839644e+00 4.18691278e-01
-1.15878379e+00 -3.23571682e-01 -2.95639277e-01 3.98454487e-01
-6.93208277e-01 8.14621866e-01 -7.11839080e-01 -4.87926424e-01
-3.81024867e-01 -1.23805726e+00 5.67400515e-01 -3.41994278e-02
-2.40936622e-01 -8.29533458e-01 1.97145134e-01 1.00406952e-01
2.67672628e-01 4.54480141e-01 1.47113550e+00 -1.45351791e+00
-1.25620618e-01 -3.79590951e-02 -2.59511173e-01 4.28183913e-01
9.52657908e-02 -1.11246146e-01 -7.22159743e-01 -2.34503508e-01
-2.19381750e-02 -6.44575238e-01 8.15212786e-01 1.15399353e-01
7.73498178e-01 -1.94530174e-01 -2.47112960e-01 2.46366680e-01
1.39878118e+00 5.42464972e-01 4.97770399e-01 5.56157112e-01
2.68817484e-01 8.87041926e-01 4.48037893e-01 2.22086608e-02
-2.06341892e-01 5.78710139e-01 2.29679719e-01 -1.86169639e-01
3.06453586e-01 -9.66073498e-02 3.59803289e-01 4.39529657e-01
1.38301253e-01 -4.23837185e-01 -9.50498462e-01 2.94932604e-01
-1.93788767e+00 -8.34432065e-01 -1.62594691e-01 1.99822271e+00
1.10530221e+00 7.71848485e-02 2.17236072e-01 2.75389384e-02
5.41385412e-01 -3.50656480e-01 -6.57525182e-01 -7.78998315e-01
-3.47259790e-01 5.98611653e-01 4.31599081e-01 3.13878804e-01
-9.85816956e-01 8.55516136e-01 6.82020521e+00 5.33566236e-01
-1.15731108e+00 -2.88286150e-01 5.85626602e-01 1.08723745e-01
2.83981234e-01 -4.84923124e-02 -7.03939497e-01 5.36358297e-01
1.48527408e+00 -9.51140076e-02 3.34143251e-01 8.22992921e-01
1.16663381e-01 2.19451729e-02 -1.25690389e+00 6.15848243e-01
-1.57227606e-01 -1.21591055e+00 -8.59076679e-02 -2.82468367e-02
3.61583412e-01 3.31873804e-01 -4.46972668e-01 3.78600240e-01
5.90623140e-01 -1.13918233e+00 1.23708203e-01 2.28068501e-01
3.77495944e-01 -7.23709166e-01 1.13443804e+00 6.09587073e-01
-5.38958788e-01 2.04494208e-01 -7.78121650e-01 -1.16244182e-01
-3.13720177e-03 7.16807246e-01 -1.33487904e+00 4.04445529e-01
5.23909569e-01 3.94275904e-01 -4.07072335e-01 8.20547938e-01
-3.90114747e-02 3.59941214e-01 -3.73304307e-01 -2.27202013e-01
3.18388551e-01 -2.27520630e-01 1.23673312e-01 1.32075298e+00
8.65033269e-02 -8.03880394e-03 3.58208209e-01 6.06564760e-01
1.07274972e-01 3.17987591e-01 -2.80705065e-01 -3.42832983e-01
8.67690542e-04 1.13156271e+00 -6.23216689e-01 -5.91647744e-01
-1.88209340e-01 7.08330154e-01 6.49985135e-01 2.54313290e-01
-5.72733164e-01 -4.82919991e-01 5.59488654e-01 -9.99721810e-02
4.54683691e-01 -1.31956609e-02 2.74155259e-01 -9.73894656e-01
-1.60661474e-01 -1.25768983e+00 1.83676690e-01 -8.08405340e-01
-1.55383146e+00 7.50379443e-01 -3.19435447e-02 -6.02528453e-01
-2.78306514e-01 -1.21744061e+00 -7.93585330e-02 1.04973352e+00
-1.35017693e+00 -7.45153487e-01 1.03081964e-01 1.57913476e-01
3.68729442e-01 -1.88977450e-01 1.05511343e+00 1.75108165e-01
-3.44750226e-01 3.91689628e-01 5.43271661e-01 -1.18955910e-01
7.89096832e-01 -1.28941298e+00 4.15653437e-01 4.38501775e-01
-8.41914397e-03 7.66659796e-01 9.68257129e-01 -4.65189844e-01
-9.76981342e-01 -8.19800258e-01 1.24317503e+00 -3.77640985e-02
6.81475699e-01 -4.35438603e-01 -1.23605418e+00 5.92394292e-01
2.13497147e-01 -3.50980014e-01 9.14791763e-01 1.90810338e-02
-3.74484181e-01 1.92624465e-01 -1.36115301e+00 2.03683794e-01
5.61105311e-01 -2.58119017e-01 -8.40398669e-01 7.79251814e-01
7.03943908e-01 -1.04553610e-01 -8.81072521e-01 5.13808191e-01
7.03936934e-01 -1.19010413e+00 9.16067779e-01 -1.03824079e+00
3.33774358e-01 -1.05430804e-01 6.73498586e-02 -1.27655005e+00
-4.00916845e-01 -3.11362028e-01 1.74596965e-01 8.11731815e-01
5.58159649e-01 -8.15786660e-01 8.78259659e-01 2.87542671e-01
-1.42939359e-01 -9.02057588e-01 -8.12407970e-01 -8.95556450e-01
3.53457868e-01 3.81059647e-01 4.81265187e-01 8.30124617e-01
2.63558894e-01 4.19559687e-01 -2.48119757e-01 -2.42621168e-01
6.47246093e-02 -3.15479860e-02 5.58369637e-01 -1.51081204e+00
-5.72976649e-01 -1.44500330e-01 -4.96419311e-01 -8.10664654e-01
1.47337168e-01 -9.82243896e-01 2.52133846e-01 -1.19502950e+00
3.20295066e-01 -1.29356787e-01 -4.84051526e-01 6.83815062e-01
-1.37401279e-03 -1.62777528e-02 -3.32919993e-02 2.05041125e-01
-1.43279240e-01 1.68438271e-01 7.45876610e-01 -1.19604550e-01
-1.35170773e-01 -2.35703260e-01 -4.77957249e-01 7.01936960e-01
1.08742630e+00 -4.22076732e-01 -1.83402017e-01 5.84611334e-02
2.50000417e-01 -1.32422000e-01 1.08089119e-01 -7.60210216e-01
-1.27172649e-01 -7.56556392e-02 4.25948203e-01 -6.45400226e-01
3.69365454e-01 -7.58261025e-01 3.21420133e-01 8.54360402e-01
-4.03557926e-01 3.12028885e-01 3.55201185e-01 3.11344862e-01
-9.99660790e-02 -4.24138069e-01 9.54807580e-01 -4.48712885e-01
-4.82595235e-01 2.91057788e-02 -6.94841027e-01 -3.77681851e-01
1.06584120e+00 -2.11050138e-01 -3.72466534e-01 2.49056756e-01
-1.00067902e+00 -3.20665911e-02 4.95509356e-01 1.08289666e-01
2.73907900e-01 -5.57572305e-01 -5.20946205e-01 2.69861519e-01
9.39314440e-02 -3.00572753e-01 -2.03222111e-02 8.14194500e-01
-1.13460279e+00 7.73510337e-01 -6.03485525e-01 -5.32921195e-01
-1.56513417e+00 1.00915337e+00 4.19845670e-01 -5.57540178e-01
-5.53820968e-01 6.07092440e-01 2.47747019e-01 -7.10112154e-01
-9.81666264e-04 -2.30051398e-01 -1.72484607e-01 -2.72560548e-02
2.97768712e-01 -1.70418136e-02 2.70375758e-01 -4.81220871e-01
-3.58945042e-01 3.03966135e-01 -4.70604956e-01 3.30781609e-01
1.52543187e+00 3.21745187e-01 -3.44617099e-01 3.02684993e-01
1.14527678e+00 -4.17859852e-01 -9.48368907e-01 1.32282935e-02
3.79595280e-01 3.38725280e-03 -6.66846991e-01 -1.11058080e+00
-4.45037186e-01 9.95907843e-01 8.59651983e-01 1.13117963e-01
8.82024109e-01 -3.29921156e-01 6.30733013e-01 7.57684708e-01
4.20947611e-01 -6.83202326e-01 -3.02268028e-01 6.85600936e-01
3.98260921e-01 -1.20746255e+00 -1.27733409e-01 -3.09001952e-01
-2.89361507e-01 1.19843185e+00 3.91970307e-01 -9.84005332e-02
2.96164840e-01 1.27596244e-01 1.16652884e-01 -2.41076380e-01
-9.66996610e-01 1.62156016e-01 -8.34783241e-02 7.74579823e-01
8.49828899e-01 1.16196033e-02 -8.33323836e-01 2.94065058e-01
-2.66234547e-01 1.57009318e-01 5.25352061e-01 8.38385820e-01
-7.84150004e-01 -1.38910937e+00 -6.28095120e-02 2.84643471e-01
-6.98223829e-01 -2.20001593e-01 -8.40986609e-01 8.41305435e-01
2.39005551e-01 8.37644815e-01 -2.51221925e-01 -2.05793262e-01
1.93168685e-01 6.62306249e-01 4.37308878e-01 -5.64734757e-01
-8.04677248e-01 -7.79650360e-02 1.58205450e-01 -2.64644533e-01
-5.01060188e-01 -3.49376708e-01 -1.23973131e+00 -1.74411818e-01
-3.02578211e-01 6.96001828e-01 6.59646451e-01 8.91348958e-01
4.46399182e-01 3.76169145e-01 1.61498204e-01 -7.06304908e-01
-7.81858504e-01 -1.10235763e+00 -5.15639305e-01 4.79080468e-01
1.59281969e-01 -5.36705375e-01 -2.52617985e-01 1.36083022e-01] | [4.753504753112793, 5.592456817626953] |
4b20e3d7-3eb0-4a03-91e5-ba0fd1e43b99 | a-sourceful-twist-emoji-prediction-based-on | 2103.07833 | null | https://arxiv.org/abs/2103.07833v1 | https://arxiv.org/pdf/2103.07833v1.pdf | A `Sourceful' Twist: Emoji Prediction Based on Sentiment, Hashtags and Application Source | We widely use emojis in social networking to heighten, mitigate or negate the sentiment of the text. Emoji suggestions already exist in many cross-platform applications but an emoji is predicted solely based a few prominent words instead of understanding the subject and substance of the text. Through this paper, we showcase the importance of using Twitter features to help the model understand the sentiment involved and hence to predict the most suitable emoji for the text. Hashtags and Application Sources like Android, etc. are two features which we found to be important yet underused in emoji prediction and Twitter sentiment analysis on the whole. To approach this shortcoming and to further understand emoji behavioral patterns, we propose a more balanced dataset by crawling additional Twitter data, including timestamp, hashtags, and application source acting as additional attributes to the tweet. Our data analysis and neural network model performance evaluations depict that using hashtags and application sources as features allows to encode different information and is effective in emoji prediction. | ['Patrick Dudas', 'Shomir Wilson', 'Kenneth Huang', 'Rahul Katiki', 'Chi-Yang Hsu', 'Zeba Karishma', 'Pranav Venkit'] | 2021-03-14 | null | null | null | null | ['twitter-sentiment-analysis'] | ['natural-language-processing'] | [-3.52977484e-01 4.95467670e-02 -6.85850158e-02 -5.41621268e-01
2.62659490e-01 -2.69909561e-01 5.29104769e-01 4.02628511e-01
-2.05285653e-01 6.08708978e-01 4.29514199e-01 -2.28502557e-01
2.43330039e-02 -8.64056647e-01 -6.31094798e-02 -1.86747894e-01
1.73979104e-01 -1.75502390e-01 2.50139654e-01 -7.60649621e-01
4.02550727e-01 1.22680716e-01 -1.72873735e+00 4.43085372e-01
6.30350649e-01 1.18918943e+00 -1.17884077e-01 6.40806496e-01
-5.69616497e-01 1.49668288e+00 -9.26394761e-01 -6.94056988e-01
-1.29549280e-01 -3.33385527e-01 -4.77319330e-01 -3.58413130e-01
1.15364185e-02 3.57508585e-02 -1.28890678e-01 6.74667001e-01
5.27173162e-01 2.40088254e-02 4.65193599e-01 -1.15083766e+00
-5.39806128e-01 6.50105596e-01 -3.83620560e-01 2.06755295e-01
4.90133733e-01 -4.81698543e-01 7.32631028e-01 -4.69634205e-01
5.46906054e-01 7.96600223e-01 8.83199573e-01 2.78600216e-01
-4.71665382e-01 -5.58009326e-01 -2.18962934e-02 2.28457540e-01
-9.37583506e-01 -4.01557803e-01 1.24981511e+00 -4.88451302e-01
6.51053250e-01 8.35896015e-01 5.00249445e-01 1.07197297e+00
2.10486203e-01 6.28412485e-01 1.21091652e+00 -5.97150624e-01
2.14277163e-01 1.02935576e+00 4.21341598e-01 6.41366720e-01
-1.40311256e-01 -5.66475272e-01 -5.29535055e-01 -4.02436733e-01
-3.98137718e-01 2.71329939e-01 2.89041430e-01 4.27852899e-01
-3.13098788e-01 9.09791231e-01 1.99548632e-01 7.12086618e-01
-5.40892720e-01 -2.23928764e-02 8.45894694e-01 3.19164395e-01
9.57333744e-01 4.83254254e-01 -9.29039419e-01 -4.86896545e-01
-5.89466035e-01 -2.77382642e-01 1.32930708e+00 2.01375663e-01
1.07591212e+00 1.61442116e-01 1.18556805e-02 1.07938600e+00
6.31787479e-02 4.21973079e-01 6.97668672e-01 -5.08035064e-01
3.90663669e-02 9.62550521e-01 1.08777858e-01 -1.77217150e+00
-4.73595887e-01 -1.48713201e-01 -2.88349330e-01 -2.90853560e-01
7.40406513e-02 -7.38832414e-01 -6.19921207e-01 1.32148457e+00
3.65654111e-01 1.69068098e-01 -5.19252837e-01 4.66948658e-01
1.18199694e+00 2.39859536e-01 4.10662532e-01 9.22664255e-02
1.52181649e+00 -7.18511820e-01 -8.24163079e-01 -1.63238674e-01
1.16568267e+00 -9.21408892e-01 1.04356325e+00 4.81600203e-02
-6.38053536e-01 -3.37829381e-01 -7.90263474e-01 9.84336734e-02
-1.41741300e+00 3.74856323e-01 1.00815618e+00 1.20864284e+00
-8.32515240e-01 7.79514313e-01 -3.05796653e-01 -3.86377454e-01
1.76483914e-01 5.20004928e-01 -2.28544742e-01 4.12306279e-01
-1.62391078e+00 1.19300997e+00 -2.44807914e-01 -1.84960112e-01
2.61272073e-01 -5.09742916e-01 -6.93152249e-01 9.94813889e-02
1.31653443e-01 -1.62288174e-01 8.07045817e-01 -1.37604737e+00
-1.70184517e+00 6.76548958e-01 -1.89126983e-01 -2.05487370e-01
5.83291911e-02 -1.26384525e-02 -9.56908882e-01 1.73398666e-02
5.68223745e-02 2.23484356e-03 9.48662043e-01 -9.87307787e-01
-5.08182049e-01 -3.48563701e-01 8.34083185e-02 -2.67235219e-01
-1.10654700e+00 3.85651886e-01 -2.19330475e-01 -3.06264937e-01
-7.36167669e-01 -9.89053130e-01 -5.89698516e-02 -8.37747693e-01
-5.21282971e-01 -2.24397764e-01 1.13991690e+00 -9.90183353e-01
1.68992960e+00 -1.95475447e+00 -3.03231448e-01 4.15751576e-01
2.39448085e-01 3.81187767e-01 2.76891202e-01 6.18756652e-01
9.30548981e-02 6.07354999e-01 4.43281710e-01 -5.68633020e-01
-4.09485139e-02 1.49181336e-01 -2.84635991e-01 1.31772652e-01
-8.68416280e-02 6.54280484e-01 -6.11464143e-01 -4.98924613e-01
3.64747643e-01 6.94059074e-01 -3.54681492e-01 -4.74456213e-02
1.46060556e-01 1.46023750e-01 -7.06140935e-01 3.85653436e-01
3.91718030e-01 -2.13194013e-01 1.05958112e-01 -3.55259955e-01
-1.75821841e-01 3.73059392e-01 -5.57492971e-01 6.00164354e-01
-1.02528310e+00 8.61871958e-01 -2.74727613e-01 -4.90943491e-01
1.20441663e+00 1.30451441e-01 5.98720253e-01 -6.55583799e-01
8.99072111e-01 9.79633722e-03 -1.27640456e-01 -7.26853788e-01
6.81892455e-01 1.53348237e-01 -2.27201596e-01 4.21047747e-01
-2.34261855e-01 2.01807395e-01 2.33658835e-01 2.26281896e-01
1.08907747e+00 -2.55750179e-01 2.09577769e-01 -2.03742877e-01
7.49158204e-01 -1.45878270e-01 -1.30053535e-02 7.59120762e-01
-1.58781588e-01 2.36541405e-01 5.86234808e-01 -4.42752868e-01
-6.59718990e-01 -7.94949532e-02 1.92824870e-01 1.59285378e+00
-1.13595881e-01 -7.34814942e-01 -7.40529835e-01 -9.63074327e-01
-9.15397257e-02 7.82903552e-01 -1.21536505e+00 -1.31236717e-01
-9.27217826e-02 -7.12828279e-01 3.78358454e-01 -7.11587295e-02
3.51596236e-01 -9.21589732e-01 -4.33875918e-01 3.14203613e-02
-6.17911620e-03 -8.83894324e-01 1.97881833e-03 2.54709363e-01
-5.68405628e-01 -8.51644158e-01 -2.41676643e-01 -1.90806732e-01
4.36564505e-01 3.24683249e-01 1.09298038e+00 5.69519401e-01
-1.30095124e-01 5.41452765e-01 -7.26054788e-01 -6.54645026e-01
-4.08351451e-01 3.77966702e-01 -3.72128375e-02 2.43322596e-01
7.99649894e-01 -6.24337375e-01 -4.06849742e-01 3.24610621e-01
-8.12509298e-01 -3.31752390e-01 -1.17556177e-01 3.21196169e-01
-6.33582652e-01 -1.22814491e-01 6.03862643e-01 -1.47767293e+00
8.41212451e-01 -1.12444460e+00 1.74064115e-01 1.32518802e-02
-4.78644550e-01 -4.29310799e-01 9.56007063e-01 -4.02249366e-01
-1.18901014e+00 -1.06955990e-01 -3.76773030e-01 1.93484694e-01
-2.26921842e-01 6.65000856e-01 2.80188084e-01 -4.78817552e-01
7.17583239e-01 -5.22638448e-02 -3.17315832e-02 -6.25486910e-01
1.70306280e-01 8.95875335e-01 -5.70949376e-01 -5.05335443e-02
3.37987483e-01 5.37468076e-01 -1.76938057e-01 -1.16888547e+00
-9.59294856e-01 -6.04130745e-01 -3.29589039e-01 -3.83924544e-01
5.40235221e-01 -5.07616580e-01 -1.16733050e+00 8.01723227e-02
-9.39630389e-01 5.86374365e-02 3.36726457e-01 7.16835931e-02
1.27027676e-01 4.59407836e-01 -6.36868417e-01 -1.11514497e+00
-3.69743079e-01 -8.55856657e-01 5.98796070e-01 2.89580464e-01
-6.97600007e-01 -1.45821106e+00 -5.74258808e-03 4.16553527e-01
9.70845819e-01 1.15316391e-01 6.84042394e-01 -1.10541511e+00
4.32880998e-01 -8.16565752e-01 -2.59497136e-01 2.02304989e-01
4.48894948e-01 3.88760477e-01 -1.16783774e+00 5.18940568e-01
1.47530377e-01 3.01047079e-02 5.08868873e-01 -3.82400374e-03
1.19662595e+00 -4.11293298e-01 -1.33137614e-01 3.30405608e-02
1.26359606e+00 2.10658014e-01 8.28396440e-01 5.31004906e-01
7.72356868e-01 1.00381768e+00 2.97085583e-01 8.57112110e-01
7.12461174e-01 4.51632380e-01 4.85507011e-01 -1.79464400e-01
4.35069911e-02 -2.17336956e-02 3.41255039e-01 1.08335888e+00
1.13419294e-01 -1.51883528e-01 -6.12920702e-01 3.73827875e-01
-1.58606005e+00 -6.89364076e-01 -6.52728677e-01 1.58216131e+00
6.95870936e-01 5.29518574e-02 2.72106141e-01 1.06312975e-01
5.00746548e-01 3.09323162e-01 -6.22124746e-02 -9.99677598e-01
-8.55192915e-02 1.67697370e-01 6.10379159e-01 5.09318471e-01
-9.54838753e-01 9.14657533e-01 5.69952917e+00 7.40719259e-01
-1.48043716e+00 2.10498124e-01 9.12741601e-01 2.98486143e-01
-3.34898651e-01 -5.51521480e-02 -7.41912246e-01 1.00004053e+00
1.29937851e+00 1.34791017e-01 4.41317439e-01 1.18821764e+00
2.98061341e-01 -5.35117388e-01 -5.06011844e-01 8.09078276e-01
3.99218023e-01 -1.21193397e+00 -3.43887061e-01 -2.19784498e-01
4.68815684e-01 1.81169346e-01 1.33618683e-01 4.48078036e-01
1.38528675e-01 -5.18577158e-01 2.52734095e-01 4.96686459e-01
3.00599765e-02 -4.16589051e-01 9.45235491e-01 1.19644580e-02
-6.63444698e-01 1.23393293e-02 -1.90194607e-01 -7.09996939e-01
-1.83283448e-01 5.86721897e-01 -8.89364064e-01 1.15569241e-01
7.01585412e-01 8.01361382e-01 -1.10592461e+00 5.26728451e-01
1.31735504e-01 7.78914392e-01 -3.68843913e-01 -8.08795452e-01
1.54969767e-01 -1.08585022e-01 9.76989865e-02 1.51400816e+00
2.43558139e-01 -2.25741550e-01 -2.12725744e-01 1.79827005e-01
-5.99800050e-02 7.31984854e-01 -7.82362282e-01 -3.43555838e-01
1.90629050e-01 1.71022928e+00 -7.95215249e-01 -3.55610877e-01
-5.57553709e-01 9.72893298e-01 2.05338925e-01 1.68343380e-01
-7.85479546e-01 -5.97110033e-01 6.72292829e-01 3.74518372e-02
1.16467737e-01 -9.11428710e-04 -6.82945788e-01 -1.07471335e+00
-4.39760000e-01 -5.80685019e-01 1.27022281e-01 -8.33738446e-01
-1.24469352e+00 7.52274215e-01 -6.50100946e-01 -9.33925807e-01
1.63502712e-02 -7.72902429e-01 -9.90560293e-01 7.97303796e-01
-1.51284564e+00 -1.10212159e+00 -2.61550158e-01 2.76152611e-01
1.45833135e-01 -1.38259158e-01 9.74391401e-01 6.40296400e-01
-5.36276639e-01 4.80600446e-01 -4.06993087e-03 1.14071943e-01
5.75002670e-01 -1.00118828e+00 1.24192471e-02 1.27027765e-01
-9.38794240e-02 9.77766097e-01 1.13195384e+00 -7.54385412e-01
-1.37262964e+00 -5.46747386e-01 1.19213259e+00 -9.16177928e-01
1.22012866e+00 -3.63384485e-02 -5.21976590e-01 2.93861747e-01
4.94223714e-01 -3.80986691e-01 1.40656698e+00 4.27792400e-01
-1.74344376e-01 -1.52898505e-01 -1.17702615e+00 8.70425165e-01
2.56476879e-01 -7.04797328e-01 -9.53249335e-02 3.42477620e-01
4.21246856e-01 2.70414203e-01 -8.17487121e-01 -9.01806578e-02
5.94964743e-01 -1.22212636e+00 6.16074920e-01 -9.26061094e-01
8.41477096e-01 1.74530476e-01 -3.03652994e-02 -1.28664494e+00
1.11205615e-01 -5.06250918e-01 -2.04036683e-01 1.49742973e+00
6.78236425e-01 -5.67109466e-01 1.18515420e+00 1.21785939e+00
2.59415329e-01 -7.29290187e-01 -4.50259626e-01 -1.24153425e-03
-3.82900447e-01 -6.47735000e-01 4.72304314e-01 1.21929944e+00
3.47541660e-01 4.41481441e-01 -1.10116518e+00 -3.76051396e-01
-1.94448642e-02 -1.22277893e-01 8.10180545e-01 -1.21974790e+00
5.41539416e-02 -3.61232340e-01 -2.39537224e-01 -3.38836044e-01
1.52804703e-02 -5.14109433e-01 -4.77379441e-01 -1.15167487e+00
-2.25579381e-01 -5.47363937e-01 -2.37910509e-01 3.72291088e-01
-9.03796107e-02 6.45120084e-01 3.69142979e-01 -1.50534853e-01
-8.11847925e-01 2.06664160e-01 7.70844758e-01 -7.79844401e-03
-3.29831332e-01 1.84294000e-01 -9.85382795e-01 1.09802473e+00
9.43389177e-01 -5.06949186e-01 -3.66913415e-02 -5.18888570e-02
9.91411865e-01 -1.69725120e-01 -1.23589739e-01 -5.98527849e-01
9.40675810e-02 2.56170966e-02 2.91990936e-01 -3.13106626e-01
6.67724252e-01 -1.24985671e+00 -2.15988785e-01 3.11595082e-01
-3.15901965e-01 -1.48553237e-01 1.23891942e-01 3.02633643e-01
-3.52741741e-02 -5.99989355e-01 1.26108542e-01 -1.68406248e-01
-7.01935828e-01 -1.64070111e-02 -6.99206948e-01 -2.14576229e-01
5.52955687e-01 -2.37597972e-01 -2.65723675e-01 -8.96556199e-01
-7.56373882e-01 -8.94735381e-02 1.96320951e-01 7.61615276e-01
2.36294255e-01 -8.63454998e-01 -3.07884514e-02 -1.83614537e-01
-1.77422725e-02 -1.50128090e+00 4.71828699e-01 8.88022006e-01
-2.56381810e-01 8.73706639e-02 -1.48552418e-01 9.05422717e-02
-1.48585248e+00 2.94687331e-01 -5.19175790e-02 3.92308347e-02
2.36774504e-01 6.87321305e-01 -4.28419471e-01 -5.30406415e-01
-9.06708930e-03 9.39562768e-02 -1.09675813e+00 6.89956605e-01
5.90015054e-01 7.13511586e-01 2.00130910e-01 -8.20831835e-01
-3.51198286e-01 2.93361276e-01 1.19583607e-02 3.79961014e-01
1.40192747e+00 -5.00092506e-01 -5.89408815e-01 7.81773090e-01
1.52332926e+00 7.28026509e-01 -4.27106768e-01 7.91743174e-02
3.15249175e-01 -5.81913292e-01 7.17283040e-02 -1.01441991e+00
-1.12245893e+00 8.17294061e-01 4.50144380e-01 1.16513288e+00
7.46225417e-01 -2.76774168e-01 1.09354317e+00 4.37971979e-01
2.62947772e-02 -1.41650820e+00 4.05786820e-02 7.87308872e-01
4.50474590e-01 -1.44815445e+00 -3.65152955e-02 -4.91748989e-01
-1.04772437e+00 1.23296809e+00 5.65612078e-01 -9.32533145e-02
1.29454005e+00 1.18601076e-01 3.46344888e-01 -3.28842133e-01
-4.43091512e-01 -2.62689382e-01 1.49067447e-01 3.23544055e-01
6.71222210e-01 -2.85864398e-02 -6.66684926e-01 9.20540631e-01
-6.62456080e-02 -1.38979685e-02 7.64140785e-01 7.67658532e-01
-3.34322274e-01 -1.08024418e+00 -7.15115294e-02 9.07229483e-01
-1.11510754e+00 -2.56806403e-01 -7.85519958e-01 2.78450370e-01
2.08038315e-01 1.24344945e+00 -2.79236227e-01 -1.11802065e+00
3.47003415e-02 2.59099185e-01 -3.40658009e-01 -5.15943944e-01
-1.40133560e+00 -4.98601645e-01 7.39831448e-01 -2.28896573e-01
-3.66519272e-01 -1.79056063e-01 -7.85908580e-01 -5.73916137e-01
-4.64614123e-01 3.25973064e-01 1.34886324e+00 1.20088351e+00
7.94353545e-01 4.44840550e-01 9.88076031e-01 -8.05185437e-01
6.63223863e-02 -1.09770787e+00 -3.56881022e-01 6.63042009e-01
6.76315576e-02 -6.12655520e-01 -6.20123506e-01 -1.19153820e-01] | [11.288226127624512, 6.913079738616943] |
45b17767-e00f-4ec7-8015-22f7ddff9105 | time-contrastive-networks-self-supervised | 1704.06888 | null | http://arxiv.org/abs/1704.06888v3 | http://arxiv.org/pdf/1704.06888v3.pdf | Time-Contrastive Networks: Self-Supervised Learning from Video | We propose a self-supervised approach for learning representations and
robotic behaviors entirely from unlabeled videos recorded from multiple
viewpoints, and study how this representation can be used in two robotic
imitation settings: imitating object interactions from videos of humans, and
imitating human poses. Imitation of human behavior requires a
viewpoint-invariant representation that captures the relationships between
end-effectors (hands or robot grippers) and the environment, object attributes,
and body pose. We train our representations using a metric learning loss, where
multiple simultaneous viewpoints of the same observation are attracted in the
embedding space, while being repelled from temporal neighbors which are often
visually similar but functionally different. In other words, the model
simultaneously learns to recognize what is common between different-looking
images, and what is different between similar-looking images. This signal
causes our model to discover attributes that do not change across viewpoint,
but do change across time, while ignoring nuisance variables such as
occlusions, motion blur, lighting and background. We demonstrate that this
representation can be used by a robot to directly mimic human poses without an
explicit correspondence, and that it can be used as a reward function within a
reinforcement learning algorithm. While representations are learned from an
unlabeled collection of task-related videos, robot behaviors such as pouring
are learned by watching a single 3rd-person demonstration by a human. Reward
functions obtained by following the human demonstrations under the learned
representation enable efficient reinforcement learning that is practical for
real-world robotic systems. Video results, open-source code and dataset are
available at https://sermanet.github.io/imitate | ['Yevgen Chebotar', 'Stefan Schaal', 'Eric Jang', 'Sergey Levine', 'Corey Lynch', 'Pierre Sermanet', 'Jasmine Hsu'] | 2017-04-23 | null | null | null | null | ['video-alignment'] | ['computer-vision'] | [-1.50662825e-01 3.63699608e-02 -1.30036235e-01 -1.25517502e-01
-1.99373156e-01 -8.34114373e-01 4.50730026e-01 -5.94749212e-01
-4.03315604e-01 7.35077441e-01 3.97247598e-02 5.56834757e-01
8.58806521e-02 -8.31346214e-02 -1.24806798e+00 -7.53822625e-01
-4.17226911e-01 5.27824581e-01 8.08691606e-02 -1.92028731e-02
-6.28638417e-02 6.45855427e-01 -1.43813074e+00 6.80907145e-02
1.88139454e-01 5.46392202e-01 6.25680804e-01 8.96620452e-01
6.32446408e-01 1.14029908e+00 -5.26254833e-01 2.80103862e-01
6.87811613e-01 -4.97650146e-01 -6.27460420e-01 4.57041502e-01
3.48881185e-01 -7.92469621e-01 -1.04104102e+00 8.22406530e-01
1.95884794e-01 4.15767372e-01 9.04203355e-01 -1.54926956e+00
-7.10621715e-01 4.33328189e-02 -3.65430653e-01 -2.76888549e-01
5.79543471e-01 7.53151417e-01 8.05880308e-01 -5.79061151e-01
9.38656509e-01 1.49909294e+00 2.21772805e-01 1.00208330e+00
-1.34605575e+00 -4.35550511e-01 2.56895632e-01 2.04998940e-01
-9.44353282e-01 -2.54586667e-01 6.92680538e-01 -6.45439923e-01
6.86454415e-01 -4.85974997e-02 8.51847410e-01 1.60027015e+00
3.57514173e-01 9.40575480e-01 6.89791977e-01 -9.14836302e-03
1.79680824e-01 -6.68861493e-02 -3.20629001e-01 8.33125591e-01
-4.58003879e-02 5.10484219e-01 -3.91781926e-01 -7.90327862e-02
1.26633704e+00 4.67627466e-01 -6.08948946e-01 -1.36794996e+00
-1.57506096e+00 5.69347203e-01 6.32041574e-01 6.50903806e-02
-5.55325747e-01 6.34245813e-01 1.07876025e-04 6.37516856e-01
-3.73276830e-01 6.55890584e-01 -6.36279821e-01 -7.71151558e-02
1.06118567e-01 4.25382793e-01 7.49991536e-01 1.30333853e+00
7.28348136e-01 6.51822314e-02 1.16194375e-01 4.24716502e-01
2.45013177e-01 7.99791813e-01 5.92281342e-01 -1.61477900e+00
2.34306127e-01 4.08957094e-01 6.07029498e-01 -8.10672402e-01
-3.85983080e-01 1.77026525e-01 -1.82786539e-01 7.89096117e-01
6.22172654e-01 -2.36900836e-01 -7.82382250e-01 2.18249726e+00
3.22628528e-01 7.06021115e-02 1.22522280e-01 1.32476926e+00
3.58247459e-01 4.89844471e-01 -2.00333372e-01 6.07619211e-02
1.10944104e+00 -1.05329597e+00 -3.93126845e-01 -4.64515150e-01
4.19904292e-01 -4.62530285e-01 1.00404477e+00 2.50804573e-01
-9.35666502e-01 -5.95658720e-01 -9.08536971e-01 4.38546995e-03
5.57298167e-03 1.78928733e-01 3.10941219e-01 -2.54527032e-01
-8.19052160e-01 8.09429228e-01 -1.17270517e+00 -7.72405863e-01
-2.84755230e-02 4.94251668e-01 -6.80892408e-01 -2.55374406e-02
-7.05414057e-01 1.14192092e+00 5.13407290e-02 -1.06862679e-01
-1.75911427e+00 -1.85524866e-01 -1.05922616e+00 -3.03323358e-01
5.85933328e-01 -6.73357010e-01 1.46389878e+00 -1.22076845e+00
-1.69880629e+00 7.96053171e-01 -1.18459025e-02 -2.01929003e-01
5.55470765e-01 -4.75995690e-01 6.33395016e-02 4.49648142e-01
2.79989801e-02 8.85979414e-01 1.24079728e+00 -1.40487897e+00
-9.39594209e-02 -6.34008646e-01 3.24277550e-01 6.43249810e-01
5.60084581e-02 -2.79397517e-01 -2.56755501e-01 -5.19587338e-01
-6.53582364e-02 -1.44713485e+00 -1.17469788e-01 7.82337368e-01
-2.21173048e-01 7.81856291e-03 1.03591812e+00 -3.96726429e-01
-1.53168840e-02 -2.26891971e+00 9.21027601e-01 -2.01472148e-01
-3.56392711e-02 -1.55312151e-01 -4.61916208e-01 4.39611971e-01
9.50426012e-02 -4.71491396e-01 -3.95616964e-02 -3.44515331e-02
-1.33982450e-01 5.13112128e-01 -1.35423556e-01 8.53147805e-01
2.09910408e-01 1.02693713e+00 -1.34335124e+00 -7.62823522e-02
3.18906575e-01 4.57027704e-01 -4.23745900e-01 7.38549769e-01
-4.40071911e-01 1.17159581e+00 -5.83150327e-01 4.59805727e-01
4.45834287e-02 -3.04334700e-01 3.74853820e-01 -1.62876979e-01
2.94606239e-01 -2.88242344e-02 -9.79219377e-01 1.90879774e+00
-3.68478715e-01 5.60032368e-01 3.35195780e-01 -1.10078955e+00
6.04209960e-01 3.97938728e-01 5.52324951e-01 -3.58202457e-01
2.81540066e-01 -1.00741908e-02 1.70038521e-01 -9.13915396e-01
7.02459738e-02 1.51244327e-01 -7.15096518e-02 5.24578333e-01
4.00269300e-01 -3.77120465e-01 -1.81329548e-01 7.54959360e-02
1.16879737e+00 7.52186298e-01 1.92619607e-01 1.80795088e-01
-1.93503760e-02 1.35158673e-01 4.73011166e-01 5.40792763e-01
-4.06571329e-01 4.87757027e-01 2.21626624e-01 -4.65065598e-01
-1.24056184e+00 -1.46560729e+00 3.52316797e-01 9.63800848e-01
6.48315549e-01 2.33288214e-01 -3.81133467e-01 -5.99608779e-01
3.23491931e-01 3.52129072e-01 -7.79116631e-01 -3.71217936e-01
-7.93373525e-01 1.56415209e-01 2.14844197e-02 5.85526288e-01
6.38563707e-02 -1.53752637e+00 -1.27812326e+00 4.07281965e-02
-2.68093739e-02 -1.05468476e+00 -6.41548157e-01 2.27711409e-01
-7.78465509e-01 -1.41274059e+00 -9.94980574e-01 -9.44546223e-01
8.24912250e-01 4.22949284e-01 8.47640753e-01 -1.83333516e-01
-4.86570507e-01 1.32075000e+00 -2.74148047e-01 -8.49302765e-03
-4.30164874e-01 -6.07574224e-01 5.21599472e-01 -5.77622987e-02
-4.01725397e-02 -6.87225103e-01 -6.75137818e-01 5.94170094e-01
-5.55870950e-01 -1.98865369e-01 4.96718526e-01 1.03104949e+00
3.02115947e-01 -6.17481947e-01 1.50242746e-01 -1.53981999e-01
2.41536364e-01 -4.01499242e-01 -4.70285445e-01 1.62731111e-01
1.89239174e-01 1.79316536e-01 6.75006151e-01 -1.17743289e+00
-6.55871809e-01 4.78274107e-01 7.10115612e-01 -1.11172795e+00
-3.08825582e-01 -1.38808042e-01 -2.90739164e-02 1.10662822e-02
7.14119017e-01 1.93654746e-01 5.45755744e-01 -2.74054766e-01
5.27049899e-01 3.37562919e-01 6.78858399e-01 -7.16128290e-01
9.07288969e-01 6.61742985e-01 -1.75126672e-01 -7.26665437e-01
-1.44455031e-01 -3.45284641e-01 -7.26216614e-01 -3.73125970e-01
9.59646881e-01 -9.64765191e-01 -1.11800826e+00 4.37370509e-01
-1.17334211e+00 -6.88297868e-01 -4.25349146e-01 8.96265030e-01
-1.18505740e+00 1.98999062e-01 -6.26309872e-01 -7.26990700e-01
3.26669008e-01 -1.22896826e+00 1.09660208e+00 2.24785522e-01
-4.01790977e-01 -5.92649579e-01 6.10846616e-02 2.69318130e-02
-6.26494586e-02 1.61697894e-01 6.58787906e-01 -4.95107800e-01
-7.61302769e-01 3.51689570e-03 2.43322462e-01 3.95569116e-01
4.65366781e-01 -2.95954466e-01 -5.81888258e-01 -6.62749231e-01
-3.58745601e-04 -9.08877492e-01 3.21142763e-01 3.37599099e-01
7.85788059e-01 -3.83762747e-01 -3.93420309e-01 4.20657605e-01
8.64851713e-01 4.78973299e-01 2.70005614e-01 1.12760015e-01
6.51126146e-01 7.40173221e-01 5.95872045e-01 2.36973315e-01
1.59347758e-01 8.24805975e-01 7.47824371e-01 2.55252838e-01
-2.85924040e-02 -4.12236869e-01 8.26951921e-01 3.14234465e-01
-5.93318827e-02 2.99085770e-02 -4.51545030e-01 5.17619848e-01
-2.01552606e+00 -9.54279244e-01 4.94026214e-01 2.23165631e+00
4.20684338e-01 -2.36782953e-01 2.89202154e-01 -5.12122214e-01
6.24202251e-01 -5.39012067e-02 -1.44185269e+00 -6.17268346e-02
1.75895706e-01 -3.41417879e-01 3.80531102e-01 2.52506673e-01
-9.42840457e-01 8.19220424e-01 5.93107986e+00 -1.49978593e-01
-1.17746210e+00 -2.77602732e-01 7.60011151e-02 -3.02103579e-01
1.09121501e-01 -1.06446944e-01 -2.77279876e-02 2.55327731e-01
2.82268643e-01 -2.53265332e-02 9.33215976e-01 1.02895808e+00
2.79281676e-01 4.26668562e-02 -1.79441643e+00 9.29326594e-01
1.19350351e-01 -6.92513585e-01 -1.58948570e-01 7.53322393e-02
5.23090124e-01 2.40273103e-01 1.85540125e-01 3.56173038e-01
6.35017574e-01 -8.19112062e-01 7.16049492e-01 4.93890971e-01
4.74809140e-01 -2.96011176e-02 9.79381874e-02 5.64829886e-01
-9.03488874e-01 -4.00580078e-01 -3.50181192e-01 -1.72290981e-01
6.82355165e-02 -4.11186248e-01 -6.96316361e-01 7.36178160e-02
6.37528658e-01 1.00418067e+00 5.21443263e-02 6.39584363e-01
-3.85022998e-01 -5.69322556e-02 -1.73298478e-01 -1.66639015e-01
4.37741354e-02 -2.35448673e-01 9.57002103e-01 4.55578923e-01
1.00992747e-01 1.57371283e-01 3.85168403e-01 1.07197189e+00
3.26704718e-02 -3.43501836e-01 -1.08210409e+00 -1.04940593e-01
2.47293741e-01 8.99177969e-01 -2.77756482e-01 -2.12868407e-01
-3.53347629e-01 1.36975324e+00 5.23030341e-01 8.52429390e-01
-8.66954207e-01 3.02611478e-03 1.04407263e+00 -2.45799407e-01
4.70674366e-01 -5.45651019e-01 5.56181192e-01 -1.47097778e+00
1.76240981e-01 -9.81080234e-01 -3.57195996e-02 -1.16508281e+00
-1.28113735e+00 4.60043311e-01 1.20103538e-01 -1.82256424e+00
-6.26178443e-01 -8.99327993e-01 -4.00099069e-01 4.45214510e-01
-1.02339435e+00 -7.98377454e-01 -3.14131618e-01 7.73047745e-01
8.00605774e-01 -2.10150570e-01 7.64551818e-01 -3.57104599e-01
-1.03707798e-01 2.78711289e-01 8.98457542e-02 1.91829041e-01
7.89892077e-01 -1.05096066e+00 2.28227954e-02 2.10889742e-01
3.72426778e-01 5.90131819e-01 6.45024598e-01 -3.80949348e-01
-1.85989988e+00 -7.61611402e-01 -1.03898212e-01 -6.06194854e-01
6.95332944e-01 -3.67900491e-01 -6.97239459e-01 9.41363692e-01
5.46540022e-02 3.63214463e-01 3.72451879e-02 -3.35465610e-01
-5.59944153e-01 1.02978192e-01 -1.10734117e+00 8.90329361e-01
1.25644779e+00 -5.34014761e-01 -8.37235451e-01 4.40423161e-01
6.17070854e-01 -4.30363357e-01 -5.27322471e-01 2.71009713e-01
9.53409016e-01 -6.28134429e-01 1.01067531e+00 -1.17289948e+00
4.41215813e-01 -3.65322024e-01 -2.04041675e-01 -1.58059716e+00
-2.80603588e-01 -6.33075237e-01 -2.24343121e-01 2.74690300e-01
1.15561619e-01 -5.53376853e-01 5.92924178e-01 5.14141679e-01
1.71161726e-01 -5.67635655e-01 -8.12149465e-01 -1.28548121e+00
1.02719963e-01 2.32334331e-01 6.89372867e-02 7.56685615e-01
2.90445507e-01 2.79494822e-01 -7.06732690e-01 3.38904440e-01
5.90574563e-01 3.20667565e-01 1.19392312e+00 -9.11987305e-01
-5.81910670e-01 -7.33882263e-02 -6.08274579e-01 -1.44976330e+00
4.78595614e-01 -6.68394625e-01 3.91796380e-01 -1.17118251e+00
3.71424347e-01 -7.31697828e-02 -8.72249007e-02 4.50019568e-01
1.60512760e-01 -1.38149381e-01 5.73098004e-01 5.88510990e-01
-4.86296713e-01 8.66550803e-01 1.64022076e+00 -3.80654693e-01
-1.71934471e-01 -1.58408880e-01 4.40905131e-02 8.47709060e-01
7.73797274e-01 -4.96962845e-01 -5.14875412e-01 -7.21776009e-01
-4.21526998e-01 4.26267475e-01 8.61358523e-01 -9.85301912e-01
9.97099355e-02 -3.42382669e-01 5.05065441e-01 8.54523480e-02
8.19027662e-01 -1.16098273e+00 1.88160717e-01 7.78695583e-01
-4.88297403e-01 1.73530325e-01 1.24081429e-02 9.66191351e-01
1.85005724e-01 -1.32207274e-01 6.28112495e-01 -5.54482460e-01
-6.61930680e-01 3.39169502e-01 -4.74663585e-01 2.02083215e-02
1.33899307e+00 -1.40833795e-01 -1.23525582e-01 -8.44926834e-01
-9.68276441e-01 3.73112977e-01 9.03754354e-01 7.20809460e-01
6.19641125e-01 -1.25885773e+00 -4.29779053e-01 1.37684792e-01
1.57917082e-01 -2.43795320e-01 9.77345631e-02 5.33358872e-01
-2.15325132e-01 2.69683003e-02 -5.93561828e-01 -8.93952072e-01
-1.04960406e+00 7.73967981e-01 4.75652158e-01 2.79821873e-01
-8.80205810e-01 5.70330203e-01 6.60503864e-01 -7.85109699e-01
4.97937262e-01 -3.26857775e-01 1.35830164e-01 -6.37456477e-01
2.08403885e-01 1.72396302e-01 -7.53227174e-01 -6.93647027e-01
-2.05928564e-01 8.09035480e-01 4.06801626e-02 -1.69516623e-01
1.16760218e+00 -4.20328043e-02 3.33349586e-01 7.52750993e-01
1.36414242e+00 -3.51347625e-01 -2.00391173e+00 -2.72296309e-01
-3.96262586e-01 -4.39479768e-01 -6.81940496e-01 -6.14265621e-01
-1.00612330e+00 7.94946313e-01 5.74938416e-01 -2.01439098e-01
7.06911147e-01 3.91718745e-01 4.38683927e-01 9.55359697e-01
8.22234273e-01 -9.19685602e-01 9.18738306e-01 4.93863821e-01
1.35712445e+00 -1.38402987e+00 -2.24383637e-01 1.42362649e-02
-9.27703202e-01 8.99959087e-01 8.97258997e-01 -6.33897901e-01
4.02479917e-01 1.56652465e-01 1.46771148e-01 -1.54194504e-01
-7.47363389e-01 -8.41806829e-02 6.72515631e-02 8.92294705e-01
-1.62985489e-01 1.26875639e-01 2.93559045e-01 9.23641771e-02
1.42674163e-01 -2.01062635e-01 5.17323792e-01 1.13531816e+00
-1.75194249e-01 -7.85640061e-01 -1.78329706e-01 1.27987042e-01
1.98037893e-01 5.32809675e-01 -5.49761832e-01 9.14161026e-01
-1.84529856e-01 6.36770248e-01 1.51205078e-01 -4.71615285e-01
4.56980467e-01 -1.76202774e-01 9.94881094e-01 -7.46267498e-01
-7.08670989e-02 -7.43782669e-02 -3.26636344e-01 -9.97480392e-01
-5.14246643e-01 -8.50304067e-01 -1.39157617e+00 2.30083123e-01
8.98712203e-02 -6.75435215e-02 3.71258199e-01 7.05873609e-01
3.32181960e-01 2.89993376e-01 7.79385090e-01 -1.45539355e+00
-9.50114667e-01 -8.99239838e-01 -6.20653093e-01 7.55179524e-01
7.51592517e-01 -1.22126555e+00 -6.25823498e-01 3.04315358e-01] | [4.625287055969238, 0.7261697053909302] |
46929d78-f949-43dc-9640-3bf42bbf404f | series-photo-selection-via-multi-view-graph | 2203.09736 | null | https://arxiv.org/abs/2203.09736v1 | https://arxiv.org/pdf/2203.09736v1.pdf | Series Photo Selection via Multi-view Graph Learning | Series photo selection (SPS) is an important branch of the image aesthetics quality assessment, which focuses on finding the best one from a series of nearly identical photos. While a great progress has been observed, most of the existing SPS approaches concentrate solely on extracting features from the original image, neglecting that multiple views, e.g, saturation level, color histogram and depth of field of the image, will be of benefit to successfully reflecting the subtle aesthetic changes. Taken multi-view into consideration, we leverage a graph neural network to construct the relationships between multi-view features. Besides, multiple views are aggregated with an adaptive-weight self-attention module to verify the significance of each view. Finally, a siamese network is proposed to select the best one from a series of nearly identical photos. Experimental results demonstrate that our model accomplish the highest success rates compared with competitive methods. | ['Yilong Yin', 'Xiushan Nie', 'Jian Zhang', 'Yongshun Gong', 'Lu Zhang', 'Jin Huang'] | 2022-03-18 | null | null | null | null | ['aesthetics-quality-assessment'] | ['computer-vision'] | [ 1.98235020e-01 -2.42904186e-01 -6.93760961e-02 -3.04380417e-01
-6.79657221e-01 -1.80036008e-01 2.87837386e-01 7.53160641e-02
-1.54996673e-02 2.09017709e-01 2.62575299e-01 3.52976799e-01
-2.10806310e-01 -6.05715156e-01 -5.82374513e-01 -7.73025513e-01
1.82747126e-01 -1.77583754e-01 2.45870203e-01 -3.13882440e-01
5.79175591e-01 4.03584599e-01 -1.54835403e+00 1.65055737e-01
8.58459234e-01 1.28408408e+00 2.69725323e-01 1.61966562e-01
1.00343920e-01 3.72876287e-01 -3.68980676e-01 -7.64663279e-01
4.35401052e-01 -5.45347035e-01 -4.11400765e-01 6.31911695e-01
7.49436617e-01 -3.85369182e-01 -2.97823846e-01 1.43529534e+00
6.42000794e-01 2.27897033e-01 3.96221608e-01 -1.25651002e+00
-9.78826463e-01 5.13932943e-01 -9.84122097e-01 2.71381021e-01
3.95730674e-01 2.98922867e-01 1.42707348e+00 -1.00113595e+00
5.62498033e-01 1.24909794e+00 1.64242759e-01 1.21757630e-02
-1.15602911e+00 -4.51320827e-01 3.80105615e-01 6.14373982e-01
-1.26100099e+00 -2.97061414e-01 1.43308365e+00 -9.84706432e-02
5.02304792e-01 2.72388339e-01 1.07527041e+00 8.02421629e-01
2.47865245e-01 9.66514230e-01 1.17467260e+00 -1.40088305e-01
1.36012122e-01 1.12234779e-01 -3.25929165e-01 7.79596388e-01
5.49678802e-02 -2.25610390e-01 -6.33985817e-01 1.15962125e-01
5.85567832e-01 3.65933537e-01 -5.12363613e-01 -7.74681091e-01
-1.07382965e+00 5.75404406e-01 9.02766764e-01 2.35242799e-01
-4.94556606e-01 -1.21382065e-01 2.66302139e-01 1.38250813e-01
2.97587842e-01 5.45958340e-01 -1.53785273e-01 2.57272333e-01
-6.51126325e-01 -8.16249624e-02 1.36878297e-01 8.33490372e-01
9.01821077e-01 7.87556618e-02 -2.69088387e-01 9.10562098e-01
9.55015123e-02 4.71691579e-01 3.51759046e-01 -9.40929472e-01
4.87730533e-01 1.07288897e+00 -1.27569839e-01 -1.60982502e+00
-2.64875293e-01 -4.79895592e-01 -9.76700842e-01 2.99457073e-01
-2.94690561e-02 2.23785356e-01 -6.61485136e-01 1.33837783e+00
3.87708157e-01 -1.50083169e-01 -2.95312285e-01 1.17941165e+00
6.50325477e-01 4.65597630e-01 -1.02216005e-01 -3.31353515e-01
1.20766211e+00 -1.12930441e+00 -4.70385343e-01 -3.32062036e-01
-1.87765822e-01 -8.81262600e-01 1.21143842e+00 4.45647597e-01
-1.34108078e+00 -5.84766984e-01 -1.10964060e+00 -1.02181643e-01
-2.75962681e-01 2.65800565e-01 4.35248494e-01 3.41860801e-01
-9.01772261e-01 7.63621271e-01 -3.69149119e-01 -3.65572751e-01
4.75046933e-01 2.18027771e-01 -3.90722632e-01 -3.52695405e-01
-8.56311858e-01 5.76289773e-01 2.03575701e-01 1.97203487e-01
-5.53126335e-01 -4.22485620e-01 -7.69117177e-01 3.20014894e-01
5.74179411e-01 -7.32457101e-01 6.54769182e-01 -1.35712695e+00
-1.50836384e+00 7.44816244e-01 -2.40788888e-02 -1.18759513e-01
4.31954026e-01 4.56043631e-02 -4.67590690e-01 4.40376103e-01
6.73119798e-02 4.51193988e-01 1.12018466e+00 -1.50157666e+00
-7.71964133e-01 -5.39735973e-01 2.17905998e-01 5.47752917e-01
-8.07901919e-01 -2.23281994e-01 -9.27329123e-01 -5.67410648e-01
3.75762373e-01 -8.49244118e-01 -3.41547698e-01 1.11348875e-01
-6.15908861e-01 -6.54578730e-02 4.68429297e-01 -6.07157946e-01
1.31847870e+00 -2.07853365e+00 4.62796211e-01 3.51377040e-01
2.92219371e-01 1.57611480e-03 -2.68245459e-01 4.95775908e-01
-7.66709745e-02 1.85669124e-01 -2.22586319e-02 -1.84951872e-01
-1.32662579e-01 -1.92805424e-01 4.56800796e-02 4.56580162e-01
3.64647746e-01 7.65712023e-01 -8.18816960e-01 -5.10453224e-01
3.19908261e-01 3.47188264e-01 -4.53158379e-01 1.20824210e-01
1.53211206e-01 3.51000577e-01 -6.23556256e-01 6.90415084e-01
8.13683093e-01 -5.38517475e-01 8.84914100e-02 -6.67520285e-01
1.08540125e-01 -2.03678668e-01 -1.11553037e+00 1.75833440e+00
-3.96871269e-01 2.11694896e-01 -2.94869423e-01 -7.54484117e-01
8.23536336e-01 -9.15485919e-02 7.96199024e-01 -9.50405121e-01
2.46266887e-01 -1.10751130e-01 -5.38759343e-02 -6.55430198e-01
4.27658945e-01 5.96703514e-02 8.29397738e-02 2.67652422e-01
-1.98742151e-01 -5.94971068e-02 2.89015800e-01 2.03058198e-01
5.70075095e-01 -5.17185703e-02 5.47174394e-01 2.35290304e-01
7.57275224e-01 -3.30124348e-01 6.40679300e-01 3.04301262e-01
-4.16365534e-01 9.69752252e-01 5.36209822e-01 -4.96007919e-01
-1.16816843e+00 -1.01867449e+00 2.33446896e-01 8.70635629e-01
6.98681891e-01 -4.39060032e-01 -5.26312768e-01 -7.58226037e-01
-1.04740053e-01 4.10569668e-01 -8.03949714e-01 -3.37450057e-01
-3.69216889e-01 -5.77374756e-01 -3.25185567e-01 2.47589767e-01
6.50658369e-01 -1.03784096e+00 -4.74822998e-01 -4.64828014e-02
-1.84662640e-01 -1.03353393e+00 -7.84494042e-01 -2.95921803e-01
-6.97786927e-01 -1.17121875e+00 -7.36459792e-01 -6.89911604e-01
8.37074220e-01 9.64102745e-01 1.03275478e+00 2.35150903e-01
-2.10389093e-01 4.00209814e-01 -5.61917007e-01 -4.37957011e-02
9.80768912e-03 -1.36347920e-01 -3.15237850e-01 6.33898497e-01
1.20931054e-02 -6.13629878e-01 -1.10745466e+00 2.92156577e-01
-9.04839277e-01 1.18332669e-01 9.42445397e-01 8.27610970e-01
8.20666254e-01 2.71568865e-01 1.10393316e-01 -6.60690546e-01
5.69925487e-01 -2.91374654e-01 -2.24590912e-01 5.19060194e-01
-8.03085387e-01 -3.42834920e-01 7.95558214e-01 -2.66881585e-01
-9.58629072e-01 5.82170114e-02 1.71159863e-01 -6.96553648e-01
8.32999721e-02 2.81908870e-01 -4.57201481e-01 -2.26787090e-01
2.51191258e-01 5.18734932e-01 6.49897978e-02 -7.06150010e-02
4.42363054e-01 2.18025595e-01 1.13703124e-01 -1.51707223e-02
7.08545983e-01 5.99194884e-01 -1.01496026e-01 -6.54163480e-01
-9.23694611e-01 -5.53823769e-01 -6.11055553e-01 -3.95975113e-01
6.53019488e-01 -7.40851820e-01 -6.56043470e-01 2.71752000e-01
-7.14969158e-01 3.48601371e-01 -1.43476978e-01 2.11531550e-01
-4.48777109e-01 6.32794917e-01 -1.85874462e-01 -6.88936472e-01
-3.57950896e-01 -1.34395719e+00 1.20144820e+00 6.10570490e-01
2.02560320e-01 -5.82097292e-01 -2.97064334e-01 5.49447060e-01
2.64834136e-01 1.51185364e-01 9.02980745e-01 -2.56165296e-01
-7.38309681e-01 -3.70899856e-01 -3.94586056e-01 4.36116606e-01
1.75433040e-01 2.51054257e-01 -6.55374646e-01 -3.93176347e-01
-8.42412412e-02 -1.83197081e-01 7.95318663e-01 4.63541210e-01
1.47208929e+00 -2.75505871e-01 -2.04334915e-01 7.30687559e-01
1.63012934e+00 1.86018988e-01 7.04467058e-01 3.59066874e-01
7.99959540e-01 5.88735580e-01 4.45702463e-01 5.57761014e-01
5.23846388e-01 6.63110733e-01 8.07919562e-01 -2.24386275e-01
-1.31501570e-01 -3.17553610e-01 2.21956506e-01 7.42803752e-01
-6.59688413e-02 -2.81176746e-01 -4.40453976e-01 4.36451048e-01
-1.61854208e+00 -1.11641920e+00 9.37832072e-02 2.25611210e+00
2.11754709e-01 9.01062563e-02 1.57456204e-01 5.60534634e-02
7.80851543e-01 5.46847880e-01 -9.17314351e-01 8.33945535e-03
-4.25216407e-01 -3.17696899e-01 2.14439765e-01 -2.34480370e-02
-1.14668667e+00 6.59172297e-01 5.69358492e+00 9.94761884e-01
-1.22741663e+00 -3.29025030e-01 1.05816793e+00 -2.69318312e-01
-5.83902299e-01 4.87372987e-02 -3.14955801e-01 6.30910933e-01
-3.67059074e-02 -2.02082008e-01 6.55902147e-01 6.90280735e-01
3.48589867e-01 -1.43873066e-01 -6.43059194e-01 1.22973430e+00
6.52695358e-01 -9.74384308e-01 2.84518838e-01 -7.99876004e-02
9.56884921e-01 -2.99668908e-01 3.74927342e-01 -1.22722752e-01
2.03555543e-02 -6.81602538e-01 5.30579746e-01 5.94222486e-01
4.68437880e-01 -9.45249557e-01 5.56602538e-01 -4.23528068e-02
-1.26143146e+00 -4.71839815e-01 -4.55120772e-01 4.11684513e-01
2.08948955e-01 6.02686346e-01 -3.64736199e-01 7.40170836e-01
9.89755988e-01 1.04100609e+00 -1.07816565e+00 1.22692406e+00
-1.71927989e-01 1.74563512e-01 1.00389071e-01 -1.73926011e-01
1.37965143e-01 -5.47411323e-01 5.73920786e-01 5.57271421e-01
4.38863307e-01 1.73082888e-01 2.73222804e-01 6.38581038e-01
-4.44875918e-02 5.93920410e-01 -6.48649156e-01 1.45255104e-01
5.65131642e-02 1.69216955e+00 -7.41765261e-01 -2.86916375e-01
-4.63115424e-01 1.29783523e+00 5.01437426e-01 4.08054084e-01
-5.97304404e-01 -7.32360333e-02 4.22503978e-01 9.01333764e-02
6.38557494e-01 9.10042226e-02 -2.77263641e-01 -1.16301072e+00
2.19786942e-01 -9.05665934e-01 5.45173585e-01 -1.13802946e+00
-1.62692952e+00 6.57298267e-01 -5.40467799e-01 -1.85517395e+00
2.97365963e-01 -1.44188628e-01 -8.29217374e-01 4.80149060e-01
-1.60943949e+00 -1.17778337e+00 -6.05735362e-01 4.09766316e-01
6.95110798e-01 -7.43375123e-02 3.29070210e-01 1.57907009e-01
-7.28145421e-01 5.45887172e-01 1.02770098e-01 -9.60691497e-02
7.08447456e-01 -1.28706253e+00 1.53106317e-01 8.75812590e-01
1.19493119e-01 5.62574327e-01 5.63458025e-01 -4.12996024e-01
-1.51480675e+00 -8.22553396e-01 4.07324284e-01 1.05778545e-01
6.62014067e-01 1.03269860e-01 -7.56169558e-01 7.35212713e-02
4.89018232e-01 5.24687022e-02 6.11238718e-01 -1.20750457e-01
-1.94549814e-01 -4.89117116e-01 -8.29251289e-01 8.86884451e-01
1.04008126e+00 -2.97691554e-01 -3.03286742e-02 2.66045958e-01
4.03091043e-01 -2.71877885e-01 -9.25745070e-01 1.72442317e-01
5.86755872e-01 -1.57226253e+00 1.15345109e+00 -4.57757950e-01
8.30580652e-01 -3.39586794e-01 2.57621463e-02 -1.64728856e+00
-5.66749156e-01 -4.37767684e-01 3.02873313e-01 1.24435425e+00
2.42916629e-01 -3.87417883e-01 7.15812922e-01 4.26629245e-01
1.76721632e-01 -1.05136931e+00 -6.98139906e-01 -3.44549477e-01
-4.35483187e-01 1.00726478e-01 6.69883132e-01 7.06872225e-01
-1.28319994e-01 6.18456721e-01 -7.05258369e-01 1.96615949e-01
6.83682501e-01 7.94133961e-01 7.84111202e-01 -1.16053665e+00
-1.38272122e-01 -7.60780752e-01 -4.72445130e-01 -7.67564416e-01
-1.92263842e-01 -6.80141866e-01 -1.27169117e-01 -1.62624347e+00
4.74898130e-01 -1.21041700e-01 -6.15481973e-01 1.41315997e-01
-6.76772892e-01 3.64747226e-01 3.48804653e-01 9.63085815e-02
-8.91805887e-01 9.87402022e-01 1.80899990e+00 -4.29282546e-01
-1.05057888e-01 3.02370358e-02 -1.03640556e+00 7.14318514e-01
5.97479403e-01 -2.61845171e-01 -5.29627860e-01 -4.14256692e-01
3.43102843e-01 1.60803586e-01 3.22562695e-01 -1.04980326e+00
2.52284169e-01 -4.45682079e-01 6.68643355e-01 -3.71273339e-01
3.15685183e-01 -1.06645632e+00 -1.03007682e-01 2.17801481e-01
-2.89737433e-01 3.86411577e-01 -2.86811560e-01 9.47145402e-01
-3.89570147e-01 9.83935148e-02 8.11034441e-01 -2.47315973e-01
-9.85337496e-01 7.38764524e-01 2.57434279e-01 -7.07911029e-02
1.08612871e+00 -4.28267002e-01 -1.63427055e-01 -5.05283892e-01
-5.27069390e-01 2.98305303e-01 6.68367207e-01 6.15274370e-01
8.77210438e-01 -1.52256393e+00 -6.46866202e-01 1.67381659e-01
4.56206530e-01 -2.14899927e-01 7.70570517e-01 8.74463141e-01
-2.42766812e-01 -9.54745263e-02 -4.39145297e-01 -4.74322319e-01
-1.33332717e+00 7.82043457e-01 1.60034254e-01 -5.52136339e-02
-6.66908503e-01 6.31614566e-01 2.58251935e-01 1.66232333e-01
5.37093133e-02 7.68423602e-02 -6.43103659e-01 2.49960124e-01
4.67116803e-01 2.89216787e-01 -7.24416822e-02 -7.94667840e-01
-2.09402561e-01 9.72933352e-01 -2.26970203e-02 1.58742234e-01
1.59429932e+00 -5.40906489e-01 -9.63388383e-02 2.28516191e-01
1.33754873e+00 1.72619477e-01 -1.42802393e+00 -2.94835836e-01
-5.03084540e-01 -9.70938504e-01 1.14585787e-01 -5.59562266e-01
-1.59044659e+00 8.14962804e-01 6.80835962e-01 2.83947915e-01
1.55364406e+00 -2.96574086e-01 8.54050040e-01 5.42642437e-02
2.55045891e-01 -1.16460884e+00 6.86206996e-01 -1.99608713e-01
1.19697487e+00 -1.61492813e+00 3.49435151e-01 -3.24460775e-01
-1.09922838e+00 1.07660282e+00 8.18849206e-01 -2.71476150e-01
4.79906440e-01 -3.57145846e-01 7.07498938e-02 -3.99653614e-01
-4.59766388e-01 -3.16420734e-01 6.22744501e-01 4.33684170e-01
1.19818643e-01 -1.65039077e-01 -2.74055362e-01 2.60049313e-01
-4.65698987e-02 -4.80819613e-01 4.18982774e-01 4.89302754e-01
-3.78174514e-01 -7.07581460e-01 -7.95622990e-02 5.46256840e-01
-2.36696020e-01 -1.30442917e-01 -5.02283335e-01 4.45234805e-01
6.91263452e-02 8.52140427e-01 -7.19963089e-02 -5.24203479e-01
6.00093424e-01 -3.95340800e-01 3.00843924e-01 -2.46892363e-01
-5.54886043e-01 2.29961306e-01 -4.05439407e-01 -7.12047577e-01
-4.54103947e-01 -5.41540325e-01 -8.19749117e-01 -1.91511586e-01
-2.98152536e-01 -2.29149997e-01 4.98630881e-01 5.70252597e-01
3.87610525e-01 5.59797704e-01 1.27627873e+00 -9.86144662e-01
-3.79403383e-01 -5.39244354e-01 -8.09043884e-01 8.43390703e-01
1.64020553e-01 -5.75260520e-01 -3.63502145e-01 -2.16920301e-02] | [11.472626686096191, -0.9995842576026917] |
dea82dae-659d-41ac-b5a3-bf237b64f125 | semantically-tied-paired-cycle-consistency | 1903.03372 | null | http://arxiv.org/abs/1903.03372v1 | http://arxiv.org/pdf/1903.03372v1.pdf | Semantically Tied Paired Cycle Consistency for Zero-Shot Sketch-based Image Retrieval | Zero-shot sketch-based image retrieval (SBIR) is an emerging task in computer
vision, allowing to retrieve natural images relevant to sketch queries that
might not been seen in the training phase. Existing works either require
aligned sketch-image pairs or inefficient memory fusion layer for mapping the
visual information to a semantic space. In this work, we propose a semantically
aligned paired cycle-consistent generative (SEM-PCYC) model for zero-shot SBIR,
where each branch maps the visual information to a common semantic space via an
adversarial training. Each of these branches maintains a cycle consistency that
only requires supervision at category levels, and avoids the need of
highly-priced aligned sketch-image pairs. A classification criteria on the
generators' outputs ensures the visual to semantic space mapping to be
discriminating. Furthermore, we propose to combine textual and hierarchical
side information via a feature selection auto-encoder that selects
discriminating side information within a same end-to-end model. Our results
demonstrate a significant boost in zero-shot SBIR performance over the
state-of-the-art on the challenging Sketchy and TU-Berlin datasets. | ['Zeynep Akata', 'Anjan Dutta'] | 2019-03-08 | semantically-tied-paired-cycle-consistency-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Dutta_Semantically_Tied_Paired_Cycle_Consistency_for_Zero-Shot_Sketch-Based_Image_Retrieval_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Dutta_Semantically_Tied_Paired_Cycle_Consistency_for_Zero-Shot_Sketch-Based_Image_Retrieval_CVPR_2019_paper.pdf | cvpr-2019-6 | ['sketch-based-image-retrieval'] | ['computer-vision'] | [ 3.93662632e-01 -2.18350172e-01 -1.04067259e-01 -3.11128914e-01
-1.10807753e+00 -6.28298163e-01 1.14023471e+00 -2.46253058e-01
-6.79580197e-02 3.56580287e-01 2.83732936e-02 1.14830844e-01
-1.01586483e-01 -8.41106415e-01 -8.13733935e-01 -5.75057089e-01
3.83795321e-01 5.62428415e-01 3.33203286e-01 -2.36429960e-01
3.08804184e-01 4.06328708e-01 -1.71979403e+00 5.48177183e-01
4.41454589e-01 1.16780889e+00 5.93494952e-01 6.91494167e-01
-2.90414840e-01 6.79658115e-01 -4.17026341e-01 -6.63276494e-01
4.01063353e-01 -5.81776619e-01 -6.60287380e-01 -2.98783183e-02
9.85417008e-01 -4.88480479e-01 -6.23130023e-01 9.20718372e-01
5.11050880e-01 2.12403253e-01 9.19879377e-01 -1.57163751e+00
-1.14384508e+00 1.14215270e-01 -1.83984399e-01 -1.74011841e-01
2.60758787e-01 8.47816989e-02 1.15672648e+00 -1.33181262e+00
1.34644198e+00 1.33678019e+00 1.85945436e-01 7.73790658e-01
-1.31963241e+00 -7.28281915e-01 3.26493494e-02 3.58383596e-01
-1.69428837e+00 -4.94528949e-01 9.84089792e-01 -4.14927483e-01
7.24193931e-01 2.91543692e-01 5.00955701e-01 1.30407190e+00
-6.42039487e-03 8.22726846e-01 7.06591785e-01 -3.12757790e-01
4.86344337e-01 1.39336601e-01 -2.60975301e-01 6.10806882e-01
-1.37223288e-01 1.79749057e-01 -9.01139736e-01 -1.59985393e-01
1.02652621e+00 4.11933124e-01 7.71644637e-02 -9.35165286e-01
-1.02801216e+00 1.03906500e+00 7.04194725e-01 3.98082972e-01
-2.70799905e-01 2.90438890e-01 2.84814686e-01 5.59603155e-01
2.45896548e-01 4.30280715e-01 1.09811038e-01 5.41104317e-01
-1.19740546e+00 3.09239060e-01 5.02388895e-01 1.30887890e+00
7.24121630e-01 -8.05473849e-02 -7.97833741e-01 8.47189367e-01
2.21879423e-01 6.70929670e-01 2.44401202e-01 -8.63915622e-01
2.71288335e-01 3.48708957e-01 -3.99154760e-02 -8.86624932e-01
5.49643457e-01 -2.37729773e-01 -8.95916283e-01 3.22779119e-01
-1.09603107e-01 7.79988289e-01 -1.12690985e+00 1.80852890e+00
-5.09994440e-02 1.22334033e-01 9.34099555e-02 1.12306523e+00
9.66105282e-01 6.62059605e-01 2.09465563e-01 1.39357373e-01
1.41095996e+00 -1.05420065e+00 -5.46844065e-01 -3.52583826e-01
-2.11900532e-01 -7.82185316e-01 1.36052644e+00 4.23371345e-02
-1.23281717e+00 -7.51232862e-01 -1.29859018e+00 -5.17332017e-01
-6.37102902e-01 4.30500433e-02 2.69704312e-01 2.04284266e-01
-1.11783242e+00 5.10601699e-01 -3.55785489e-01 -3.12858611e-01
4.48532641e-01 -8.39645788e-02 -4.89010483e-01 -3.55616152e-01
-1.21533406e+00 7.51763284e-01 1.69261232e-01 -3.49576503e-01
-1.17791283e+00 -7.35247612e-01 -9.04206157e-01 2.65963167e-01
3.27331990e-01 -9.81409371e-01 9.34002161e-01 -1.00246572e+00
-1.21084630e+00 9.55417037e-01 -1.31679446e-01 -3.03014755e-01
5.90640008e-01 8.10225084e-02 -6.35256693e-02 4.60801780e-01
3.90130073e-01 1.27225709e+00 1.53536296e+00 -1.44988418e+00
-3.29903096e-01 -4.64976966e-01 -1.69471443e-01 1.74518898e-01
-2.84161001e-01 -1.95318669e-01 -8.86476219e-01 -1.03134692e+00
5.17740026e-02 -8.89574826e-01 -3.57304281e-03 4.42062169e-01
-8.70059505e-02 -2.13040933e-01 9.81618524e-01 -6.40728772e-01
9.42138672e-01 -2.13332462e+00 5.46057284e-01 1.43482640e-01
-2.81585678e-02 1.31356031e-01 -4.64243025e-01 5.90384126e-01
1.09319463e-01 -3.10981292e-02 -2.56578147e-01 -7.07227647e-01
1.66369244e-01 1.35764420e-01 -8.73626053e-01 1.25609338e-01
3.92266363e-01 1.44301832e+00 -1.01753950e+00 -5.81521511e-01
4.66100514e-01 6.58009112e-01 -4.32209730e-01 5.13753474e-01
-3.04125398e-01 1.62850529e-01 -1.87660232e-01 6.68057740e-01
4.84769940e-01 -3.10209304e-01 3.53971086e-02 -2.07155943e-01
3.48245263e-01 -1.64509684e-01 -8.35439622e-01 2.47925711e+00
-5.63247085e-01 5.32984018e-01 -1.86795980e-01 -6.47148788e-01
1.10183311e+00 1.66780785e-01 2.77962182e-02 -1.07754731e+00
-2.14999393e-01 1.69086859e-01 -7.24986613e-01 1.87434331e-02
5.64519525e-01 -1.98682055e-01 -3.76555592e-01 3.20612848e-01
6.06939495e-01 -4.06262696e-01 3.14606838e-02 5.88501513e-01
7.81311691e-01 1.75839752e-01 1.11148551e-01 -4.46747616e-02
5.20721972e-01 -2.83080101e-01 -1.70841049e-02 9.70244110e-01
1.40117005e-01 9.76190090e-01 9.76012275e-02 -4.48947579e-01
-1.43558931e+00 -1.51348174e+00 1.60147160e-01 1.24786806e+00
4.59695071e-01 -2.82039583e-01 -4.34470564e-01 -5.02501011e-01
6.75027668e-02 8.21657777e-01 -7.33050644e-01 -4.32684600e-01
-1.88747257e-01 2.98765779e-01 4.54025209e-01 4.84436989e-01
3.65742952e-01 -1.32155180e+00 -5.52800834e-01 2.97923340e-03
-6.88714236e-02 -9.09085691e-01 -7.88934529e-01 -1.10860109e-01
-4.89770710e-01 -7.39040554e-01 -1.07081556e+00 -1.01828432e+00
6.96411729e-01 5.83199680e-01 1.24832988e+00 -4.78599742e-02
-6.84670091e-01 6.43238008e-01 -2.93173939e-01 -2.40545627e-02
-3.92882168e-01 -4.34195325e-02 -4.66389835e-01 1.56269759e-01
9.16031525e-02 -5.35486579e-01 -8.74163032e-01 2.42337152e-01
-1.05631602e+00 1.85479626e-01 6.36174560e-01 1.19051361e+00
7.98134089e-01 -6.79275513e-01 6.13387644e-01 -3.33494365e-01
4.93977338e-01 -2.78680146e-01 -4.27611828e-01 7.10506916e-01
-6.86986804e-01 3.50088984e-01 4.47854847e-01 -4.66411352e-01
-8.10741246e-01 2.44417772e-01 3.30156863e-01 -1.36318874e+00
4.60802801e-02 -1.06092475e-01 -2.72322565e-01 -2.63147312e-03
5.00606537e-01 5.96288741e-01 1.71182588e-01 -2.38322288e-01
9.63159442e-01 4.80971485e-01 8.65278780e-01 -5.17885268e-01
7.95322895e-01 5.42299926e-01 -1.65500287e-02 -6.12007499e-01
-5.76291919e-01 -5.31197429e-01 -5.64556897e-01 -1.21113546e-01
1.01937842e+00 -1.08443761e+00 -3.90204608e-01 9.09605175e-02
-1.23465872e+00 1.06932186e-01 -5.31943858e-01 -2.35165134e-01
-8.63231719e-01 2.83794999e-01 -3.29346985e-01 -7.82965183e-01
-9.23407137e-01 -1.10428298e+00 1.70255458e+00 4.19808105e-02
-1.73458606e-01 -5.84233880e-01 2.24312898e-02 1.73618704e-01
4.45819348e-01 -4.80261482e-02 8.16919208e-01 -4.44809645e-01
-1.02563727e+00 -1.87543988e-01 -4.04004812e-01 2.29008153e-01
-2.16949373e-01 -3.09894115e-01 -1.08999979e+00 -4.97325301e-01
-3.36122990e-01 -6.75124645e-01 1.20646930e+00 -1.47062670e-02
1.14561558e+00 -3.29592645e-01 -1.90510884e-01 5.24611592e-01
1.63986659e+00 8.61693695e-02 8.69054735e-01 -9.73026082e-02
5.99738479e-01 4.20506209e-01 7.22378790e-01 1.69262856e-01
1.29461750e-01 9.09202278e-01 3.53932858e-01 -4.25655618e-02
-5.86736202e-01 -9.69872952e-01 1.29386291e-01 2.72935301e-01
4.11859334e-01 -1.07710786e-01 -4.22559887e-01 7.48870850e-01
-1.86535299e+00 -1.27905631e+00 6.19737327e-01 2.35576725e+00
7.16500282e-01 -2.68166184e-01 -1.90712735e-01 -1.99771330e-01
6.49065673e-01 3.49335700e-01 -5.20294547e-01 -1.20183565e-01
-1.19581506e-01 4.81336206e-01 1.81706771e-01 5.22499442e-01
-1.02280354e+00 1.17679441e+00 5.25397682e+00 1.29826951e+00
-1.01222491e+00 2.90069669e-01 5.41365862e-01 -1.74073920e-01
-5.27129710e-01 6.77689984e-02 -4.11080897e-01 3.18619639e-01
3.85444075e-01 -1.78715140e-01 7.08382487e-01 1.03222120e+00
-5.86062551e-01 8.91327187e-02 -1.36018288e+00 1.33487976e+00
4.49864417e-01 -1.61264563e+00 6.46986127e-01 -1.26759991e-01
7.18902946e-01 -3.37063164e-01 2.95497417e-01 3.16909313e-01
7.50606358e-02 -1.12443912e+00 9.86855507e-01 8.31849873e-01
1.37220585e+00 -7.64240265e-01 1.23022124e-01 1.54395774e-01
-1.39261365e+00 -1.02053117e-02 -5.39694130e-01 5.50155818e-01
-4.24085706e-02 -1.09447129e-01 -8.69005263e-01 6.69374466e-01
4.94364530e-01 5.31665266e-01 -7.09975600e-01 5.22244453e-01
-1.69769853e-01 -1.90098375e-01 1.13492921e-01 6.17517810e-03
2.06438780e-01 -1.29648715e-01 5.49072504e-01 9.89244521e-01
4.27772492e-01 -1.67415082e-01 7.30339438e-02 1.39090598e+00
-5.57745956e-02 -8.15271288e-02 -1.08591998e+00 -1.32838011e-01
6.91806197e-01 1.25133395e+00 -5.17499626e-01 -6.36484027e-01
-9.77579206e-02 1.65071654e+00 3.13925982e-01 2.34624296e-01
-5.33221304e-01 -2.96426445e-01 5.78213692e-01 -3.34612131e-02
6.67683363e-01 4.65640277e-02 -3.00248265e-01 -1.14010906e+00
1.97886657e-02 -5.66749871e-01 3.25389773e-01 -1.04505348e+00
-1.48779309e+00 5.06553292e-01 -5.14381379e-02 -1.43882823e+00
-5.21031082e-01 -1.50734961e-01 -6.17186606e-01 9.38620508e-01
-1.26034868e+00 -1.72155309e+00 -3.88758093e-01 8.54015052e-01
9.62321460e-01 -4.97201443e-01 1.08209062e+00 1.67075112e-01
1.37242317e-01 8.38486552e-01 -1.50122389e-01 -3.15171517e-02
9.02228296e-01 -1.04461896e+00 6.37387812e-01 6.84691370e-01
5.95544636e-01 5.73195934e-01 4.56334412e-01 -7.22273290e-01
-1.59895778e+00 -1.18335581e+00 9.23221886e-01 -4.66053605e-01
2.97256261e-01 -7.23648131e-01 -8.64593446e-01 2.50494599e-01
2.35491887e-01 1.57007456e-01 3.08483452e-01 -2.94658720e-01
-9.59409356e-01 -1.34606451e-01 -1.07191026e+00 7.24334002e-01
1.24110878e+00 -1.10918534e+00 -5.77548802e-01 3.05427946e-02
7.94560909e-01 9.95351747e-03 -7.12979615e-01 5.47057614e-02
8.19394529e-01 -7.72940278e-01 1.43279588e+00 -6.75686598e-01
7.41232336e-01 -3.34907562e-01 -4.30934429e-01 -1.00471961e+00
-4.44229037e-01 -2.44125158e-01 4.19158535e-03 1.10990322e+00
6.22483389e-03 -2.77377348e-02 5.02711892e-01 4.48045403e-01
1.61117151e-01 -4.46019709e-01 -9.58320677e-01 -9.17661846e-01
-1.10453054e-01 -8.40322971e-02 4.61370468e-01 7.90792584e-01
-2.38900825e-01 7.00302422e-01 -5.59085488e-01 -1.39115676e-01
8.45927358e-01 5.75583100e-01 9.00236666e-01 -1.12111330e+00
-2.03649178e-01 -5.13294637e-01 -5.92869103e-01 -6.69361293e-01
2.54205346e-01 -1.25406218e+00 9.96078365e-03 -1.47665441e+00
4.75636542e-01 -1.49249390e-01 -3.64328116e-01 4.98673737e-01
-5.93105629e-02 4.38818008e-01 6.76945686e-01 3.20871025e-01
-7.71506608e-01 8.27341080e-01 1.15636075e+00 -5.77240527e-01
1.09449826e-01 -4.86111343e-01 -4.76489514e-01 -1.23950336e-02
2.82883555e-01 -2.27709591e-01 -7.60825276e-01 -2.12880865e-01
-9.98727903e-02 3.05750370e-01 1.00255680e+00 -7.72708952e-01
4.40597683e-01 3.28079872e-02 5.81000388e-01 -6.67423308e-01
7.19427347e-01 -8.60732198e-01 3.99033338e-01 3.34341347e-01
-5.69410145e-01 -2.23263338e-01 -9.89127606e-02 8.37262928e-01
-2.96556592e-01 -2.52604485e-03 8.09888244e-01 -1.93781242e-01
-9.13836300e-01 4.99450266e-01 3.98753256e-01 -1.97538599e-01
1.08863151e+00 -1.25512540e-01 -2.88433313e-01 -4.87638086e-01
-5.68881333e-01 7.94554874e-02 6.87227488e-01 9.28757727e-01
1.09252954e+00 -1.78956258e+00 -6.51867449e-01 5.04923940e-01
5.92848659e-01 -2.44496584e-01 5.26750565e-01 2.38431916e-01
-1.25052378e-01 5.00431538e-01 -6.12831473e-01 -6.39310837e-01
-1.26518595e+00 1.02336049e+00 1.83264185e-02 -1.80195481e-01
-6.86836660e-01 8.37517977e-01 5.33263981e-01 -2.00885698e-01
2.97563434e-01 3.44509095e-01 2.75442213e-01 1.05839416e-01
5.92056870e-01 1.34318694e-01 6.34742202e-03 -4.82966900e-01
-2.82843918e-01 5.17159522e-01 -5.61295673e-02 -5.76668084e-01
1.08824563e+00 2.77544502e-02 7.36146942e-02 4.66629118e-01
1.36687279e+00 -5.34088194e-01 -1.24538577e+00 -4.18464094e-01
-2.22283557e-01 -7.38816321e-01 -8.02626610e-02 -7.88417995e-01
-9.97365296e-01 9.59148526e-01 7.21784592e-01 -1.12624258e-01
9.75728869e-01 3.22965622e-01 6.69378161e-01 3.35519165e-01
5.61745703e-01 -9.61290956e-01 4.67070192e-01 2.05531955e-01
1.62012899e+00 -1.19264650e+00 -2.65825957e-01 -1.09350890e-01
-9.00243282e-01 9.12720382e-01 3.98988962e-01 -4.52991068e-01
4.10901457e-01 -1.40198156e-01 -3.33059996e-01 -1.64507225e-01
-1.00025451e+00 -2.94826746e-01 6.81771696e-01 5.11924624e-01
-1.46066785e-01 7.01675639e-02 -1.71459645e-01 5.75116396e-01
1.11168526e-01 -6.17211461e-02 -2.28685722e-01 8.02370846e-01
-2.42702201e-01 -1.14132380e+00 -1.78118467e-01 2.41118416e-01
1.06786020e-01 -2.34327078e-01 -7.63726592e-01 4.80725050e-01
-1.65167943e-01 5.97067535e-01 3.18794608e-01 -2.71219283e-01
2.34215289e-01 3.91492844e-01 7.30331242e-01 -4.31939214e-01
-4.13537145e-01 9.03474241e-02 -4.08615261e-01 -6.40120208e-01
3.97871658e-02 -2.87254900e-01 -8.08958232e-01 5.84066547e-02
-7.07422476e-03 -1.48753360e-01 7.58614957e-01 5.40649533e-01
6.74864709e-01 2.91163564e-01 6.48555756e-01 -1.05994570e+00
-6.03169858e-01 -8.02690804e-01 -4.58106697e-01 7.71604836e-01
1.39689043e-01 -7.59601116e-01 -1.23248346e-01 1.59517631e-01] | [11.611011505126953, 0.6764500141143799] |
68258055-af86-45c7-9980-6d787c05cd7a | jseegraph-joint-structured-event-extraction | 2306.14633 | null | https://arxiv.org/abs/2306.14633v1 | https://arxiv.org/pdf/2306.14633v1.pdf | JSEEGraph: Joint Structured Event Extraction as Graph Parsing | We propose a graph-based event extraction framework JSEEGraph that approaches the task of event extraction as general graph parsing in the tradition of Meaning Representation Parsing. It explicitly encodes entities and events in a single semantic graph, and further has the flexibility to encode a wider range of additional IE relations and jointly infer individual tasks. JSEEGraph performs in an end-to-end manner via general graph parsing: (1) instead of flat sequence labelling, nested structures between entities/triggers are efficiently encoded as separate nodes in the graph, allowing for nested and overlapping entities and triggers; (2) both entities, relations, and events can be encoded in the same graph, where entities and event triggers are represented as nodes and entity relations and event arguments are constructed via edges; (3) joint inference avoids error propagation and enhances the interpolation of different IE tasks. We experiment on two benchmark datasets of varying structural complexities; ACE05 and Rich ERE, covering three languages: English, Chinese, and Spanish. Experimental results show that JSEEGraph can handle nested event structures, that it is beneficial to solve different IE tasks jointly, and that event argument extraction in particular benefits from entity extraction. Our code and models are released as open-source. | ['Lilja Øvrelid', 'Samia Touileb', 'Huiling You'] | 2023-06-26 | null | null | null | null | ['event-extraction'] | ['natural-language-processing'] | [ 2.43115798e-01 6.99493110e-01 -2.83513993e-01 -3.92306328e-01
-6.37264848e-01 -8.88857782e-01 6.80404544e-01 7.58556485e-01
-3.61311674e-01 7.34430075e-01 5.71420789e-01 -3.31304610e-01
-4.95566353e-02 -1.08736265e+00 -7.91262090e-01 -1.37215897e-01
-5.73037624e-01 5.14004171e-01 6.60038769e-01 1.10105425e-01
-3.41983080e-01 3.36993635e-01 -1.33903611e+00 3.92202139e-01
4.63863254e-01 1.01303577e+00 -1.17464073e-01 6.61938906e-01
-4.34944510e-01 1.41946912e+00 -6.69150293e-01 -7.48340487e-01
-1.27576217e-01 -2.59120643e-01 -1.19337296e+00 -4.90087271e-02
-2.64856786e-01 -2.91090012e-02 -2.95169175e-01 6.35556340e-01
2.82860339e-01 3.22822332e-02 2.92845339e-01 -1.51256824e+00
-3.12926501e-01 1.12686706e+00 -3.35858971e-01 1.63775936e-01
6.14285052e-01 -2.43455350e-01 1.47225332e+00 -5.91754794e-01
1.02362835e+00 1.28644693e+00 4.70880777e-01 1.29619136e-01
-1.13243425e+00 -4.60325629e-01 5.58070600e-01 -1.00987200e-02
-1.10585630e+00 -2.86155254e-01 5.06197274e-01 -1.12972274e-01
1.58850658e+00 3.64161283e-01 5.33892989e-01 1.15892839e+00
4.10245135e-02 9.16622758e-01 4.34308738e-01 -2.24534929e-01
2.95085400e-01 -4.39470500e-01 5.81453264e-01 8.81925285e-01
3.66614848e-01 -7.62251243e-02 -6.92968965e-01 -3.40253890e-01
5.52630484e-01 -1.10879354e-01 -1.07152320e-01 -1.30783901e-01
-1.51441741e+00 7.78281212e-01 1.78609967e-01 2.33713076e-01
-4.31475610e-01 4.44428831e-01 8.91623259e-01 1.45245880e-01
2.77365655e-01 1.59318417e-01 -8.06906044e-01 -8.90422016e-02
-2.64897406e-01 3.82988751e-01 1.34763253e+00 1.25801229e+00
6.61916554e-01 -2.71932036e-01 -3.47733587e-01 5.87516308e-01
9.60746557e-02 1.16096713e-01 6.62150532e-02 -7.86731184e-01
7.95879245e-01 8.82100046e-01 5.43855540e-02 -9.23220873e-01
-6.90269530e-01 -2.49790564e-01 -4.44503456e-01 -3.63816619e-01
2.13534296e-01 -4.50130790e-01 -7.88441837e-01 1.99345517e+00
4.49034244e-01 4.88870651e-01 2.25409478e-01 3.76323640e-01
1.26784241e+00 6.75703108e-01 6.23425663e-01 -1.94996119e-01
2.07699752e+00 -8.54135394e-01 -1.05815125e+00 -6.80195689e-01
7.88421154e-01 -3.33768606e-01 6.61193252e-01 -2.51780869e-03
-1.24913013e+00 -1.31881654e-01 -8.09681594e-01 -3.15018594e-01
-4.62783366e-01 -2.63222277e-01 1.15300262e+00 1.72764555e-01
-7.45725572e-01 2.74627835e-01 -1.13082445e+00 -1.56793833e-01
1.92875728e-01 2.55908728e-01 -4.41235989e-01 2.31493279e-01
-1.58317602e+00 5.97895622e-01 1.15364254e+00 -1.65266618e-01
-4.95524198e-01 -5.73490620e-01 -1.57134962e+00 4.45837677e-01
1.08306563e+00 -6.92777038e-01 1.36548746e+00 -4.29080009e-01
-9.86053765e-01 8.34344923e-01 -2.85148919e-01 -4.39529419e-01
-4.75325398e-02 -2.27104515e-01 -7.08629012e-01 2.34260842e-01
3.68318111e-01 4.44519818e-01 2.18388364e-01 -8.58125865e-01
-9.58338261e-01 -3.19219917e-01 3.19838136e-01 1.42766505e-01
9.65767801e-02 5.63395321e-01 -1.04857147e+00 -7.46424854e-01
-6.68087155e-02 -7.15862691e-01 -2.61383504e-01 -4.65597659e-01
-6.81993365e-01 -5.99983037e-01 6.01464212e-01 -7.71729410e-01
1.68572593e+00 -2.08268976e+00 1.44209325e-01 1.52008474e-01
5.25619805e-01 -3.81443113e-01 7.94294626e-02 6.68094993e-01
-2.67939240e-01 2.31005117e-01 -4.70327288e-01 -2.34470993e-01
2.92478353e-01 6.11288190e-01 -1.92801714e-01 -1.68647543e-01
5.54332197e-01 1.22422326e+00 -1.16827953e+00 -6.06797755e-01
-6.68335259e-02 1.60115317e-01 -4.94880915e-01 1.75205350e-01
-4.77216899e-01 -1.35595752e-02 -7.45856166e-01 6.39342904e-01
2.40798429e-01 -6.81175292e-01 7.28111982e-01 -1.64071158e-01
7.23232180e-02 8.88779104e-01 -1.70683873e+00 1.62804854e+00
-2.72833616e-01 3.31810564e-01 1.33893769e-02 -9.85116959e-01
6.82666540e-01 6.07366323e-01 4.59988952e-01 -4.36468720e-01
-7.25753158e-02 -6.68840408e-02 -4.43284422e-01 -3.69796038e-01
4.76268589e-01 1.78506061e-01 -9.17931259e-01 4.24495041e-01
3.02120984e-01 1.91988751e-01 7.20189154e-01 6.60149276e-01
1.55037630e+00 1.99439526e-01 7.55680680e-01 7.73028210e-02
2.13002861e-01 -3.17507349e-02 9.15870726e-01 6.67970896e-01
2.83238471e-01 4.02333476e-02 1.19021606e+00 -3.74052405e-01
-4.31285322e-01 -1.44048488e+00 -7.27707194e-03 1.26990163e+00
7.10256472e-02 -1.16109300e+00 -3.49638104e-01 -1.10333443e+00
-1.30666584e-01 7.73955584e-01 -4.50301498e-01 1.70986086e-01
-9.69894767e-01 -7.10908890e-01 6.88179135e-01 9.65660512e-01
3.19117844e-01 -1.31003821e+00 -6.76122189e-01 7.81114697e-01
-5.92644393e-01 -1.78846085e+00 -3.52043301e-01 6.82809651e-01
-5.85142672e-01 -1.42407632e+00 5.17990962e-02 -8.74049366e-01
2.80277520e-01 -3.85419279e-01 1.89871073e+00 -3.80201526e-02
-9.20488760e-02 2.69633293e-01 -4.66584235e-01 -5.31414509e-01
-2.03831449e-01 -7.55056366e-03 -7.11967111e-01 -3.10290843e-01
3.07722807e-01 -5.09123743e-01 -2.68524885e-01 1.95282638e-01
-1.02106404e+00 1.41694754e-01 1.62227422e-01 5.66311955e-01
6.61674321e-01 2.64374137e-01 6.32324219e-01 -1.36547375e+00
4.06570435e-01 -7.57890224e-01 -6.32483900e-01 3.96507949e-01
-2.27471054e-01 2.12653443e-01 4.41386819e-01 -9.25584063e-02
-1.30392873e+00 -5.39844744e-02 -1.05976589e-01 1.95074543e-01
-2.40362868e-01 9.54021096e-01 -4.21550959e-01 7.65687287e-01
3.56796622e-01 -1.50928795e-01 -5.40459931e-01 -2.95248389e-01
5.71294844e-01 3.21287960e-02 8.42857480e-01 -9.49526131e-01
3.33784729e-01 3.40065986e-01 -2.20554671e-03 -2.84549177e-01
-9.06592906e-01 -2.98606992e-01 -2.79100209e-01 2.44794488e-01
1.07652497e+00 -1.09798884e+00 -7.18306422e-01 1.63642734e-01
-1.31992614e+00 -4.03095186e-01 -4.61016297e-01 3.18620712e-01
-2.17903420e-01 4.27118003e-01 -1.06516182e+00 -4.36381966e-01
-2.87551731e-01 -8.30375254e-01 1.44413888e+00 1.05086647e-01
-4.45763677e-01 -1.26554251e+00 -1.74204022e-01 -2.22023785e-01
-2.41574407e-01 6.74850523e-01 1.08206785e+00 -9.31145549e-01
-5.96066594e-01 -8.23778287e-02 -2.55902022e-01 -3.14838678e-01
1.59424189e-02 -2.13532463e-01 -5.47376812e-01 7.68408030e-02
-3.06940913e-01 -2.07350284e-01 6.91065252e-01 1.97666660e-01
1.09535706e+00 -2.66040802e-01 -6.43822968e-01 4.90559101e-01
1.18638182e+00 2.66075283e-01 6.04173303e-01 1.26277536e-01
6.23678327e-01 6.69656992e-01 3.68519455e-01 5.29387295e-01
8.71550500e-01 5.31125486e-01 3.28777373e-01 -1.58203036e-01
-3.62870917e-02 -4.39361006e-01 3.48754197e-01 5.58214486e-01
1.19744308e-01 -7.64081359e-01 -8.23736787e-01 6.37850642e-01
-2.15211535e+00 -8.20971549e-01 -4.58426505e-01 1.80710566e+00
8.92483413e-01 2.03227520e-01 1.83228076e-01 6.02879189e-02
6.86859310e-01 3.03077191e-01 -3.03187579e-01 -2.74887383e-01
-1.83838442e-01 6.71898544e-01 3.96088511e-01 2.01817155e-01
-1.10650110e+00 1.04973125e+00 6.01791954e+00 5.32600462e-01
-5.05346060e-01 5.02130762e-02 3.63697201e-01 1.96715564e-01
-4.50792015e-01 3.34155262e-01 -8.80302608e-01 3.21837664e-01
9.68028486e-01 -1.76732093e-01 1.69394702e-01 5.49482286e-01
-3.99089038e-01 -1.05428658e-02 -1.35736001e+00 7.30422378e-01
-4.35373574e-01 -1.41259444e+00 -6.14505373e-02 -1.38831541e-01
2.42109954e-01 -3.17184553e-02 -8.21263790e-01 5.31258941e-01
7.99524844e-01 -6.85941875e-01 8.55041683e-01 1.04314476e-01
6.47797346e-01 -6.83149159e-01 5.34988284e-01 1.04210250e-01
-1.93202496e+00 3.36114578e-02 1.62943780e-01 3.91706713e-02
7.42184103e-01 7.87778139e-01 -6.19755983e-01 1.16902149e+00
6.73674643e-01 7.43138254e-01 -3.71907204e-01 5.18677950e-01
-7.42777944e-01 6.96355939e-01 -4.18511689e-01 2.13015124e-01
1.62818477e-01 -3.17432769e-02 5.84228814e-01 1.64492190e+00
8.72170646e-03 2.85566300e-01 4.89964753e-01 8.07561398e-01
-2.85379618e-01 -1.71492323e-01 -5.15366197e-01 -3.63836348e-01
7.26572931e-01 1.08634520e+00 -1.14213145e+00 -6.21818006e-01
-7.29408503e-01 7.99110353e-01 5.07528782e-01 4.11127537e-01
-1.15583587e+00 -6.72474980e-01 5.10496020e-01 -1.95913374e-01
5.57019651e-01 -4.47428152e-02 -1.21075168e-01 -1.34283304e+00
1.56216323e-01 -6.54302359e-01 1.29950202e+00 -7.21223474e-01
-1.03916192e+00 5.18996000e-01 2.73692131e-01 -5.17061234e-01
-4.97285426e-01 -4.51148391e-01 -5.72463155e-01 4.72924471e-01
-1.30132449e+00 -9.00488138e-01 -1.55207381e-01 6.96143568e-01
3.37243855e-01 4.31757301e-01 9.72245872e-01 4.29441065e-01
-7.48111606e-01 3.31687242e-01 -9.10657942e-01 5.73756456e-01
3.22660744e-01 -1.56584811e+00 9.64220107e-01 9.58143532e-01
3.57727677e-01 5.00032008e-01 3.00934136e-01 -9.51704025e-01
-1.39874816e+00 -1.23993421e+00 1.40371335e+00 -4.51000750e-01
8.31949353e-01 -5.51111519e-01 -9.15044785e-01 1.37449360e+00
1.85682312e-01 2.22097501e-01 5.90337694e-01 4.81631219e-01
-4.63741124e-01 3.22852463e-01 -8.17115963e-01 5.04494965e-01
1.57593286e+00 -5.71672499e-01 -7.97841966e-01 3.80371630e-01
1.22314787e+00 -6.74807906e-01 -1.16133964e+00 5.14390886e-01
9.88670588e-02 -5.41298568e-01 1.20446479e+00 -7.47559249e-01
2.47195363e-01 -3.12280774e-01 4.58203144e-02 -8.57536972e-01
-2.00318515e-01 -6.85637712e-01 -7.05009878e-01 1.52488387e+00
6.83681548e-01 -7.38523185e-01 3.25710356e-01 6.24827147e-01
-2.21319929e-01 -5.66143215e-01 -5.61120331e-01 -6.17155492e-01
-7.83208251e-01 -8.70426595e-01 8.60223413e-01 1.01523983e+00
2.04275325e-02 8.30588460e-01 -6.92684501e-02 4.13091630e-01
4.74845678e-01 4.04075384e-01 3.40609610e-01 -1.40868425e+00
-5.54588854e-01 -1.84272408e-01 1.14592845e-02 -7.90363729e-01
3.80612820e-01 -1.17204833e+00 -1.79055691e-01 -1.91715372e+00
6.04429841e-02 -4.01838630e-01 -2.04243302e-01 1.06527412e+00
-4.45055634e-01 -2.67361522e-01 -4.39630076e-02 -2.62772530e-01
-9.62244332e-01 1.34957507e-01 9.44119811e-01 1.44670933e-01
-1.92125067e-01 -2.88080722e-01 -7.71219075e-01 8.48556042e-01
4.52327937e-01 -6.05380416e-01 -4.77727711e-01 -4.40575182e-01
6.64067924e-01 7.40929842e-01 4.64708894e-01 -4.81940001e-01
1.11626893e-01 3.20353657e-02 1.28649414e-01 -4.57537711e-01
-7.58368596e-02 -7.33913541e-01 4.43798274e-01 1.07486531e-01
-2.19671294e-01 1.77319869e-01 3.33838791e-01 7.85544395e-01
-3.87351662e-01 -6.60359487e-02 1.20364271e-01 -1.74424365e-01
-1.06554437e+00 2.93660372e-01 -1.40709072e-01 7.11512685e-01
8.87286246e-01 1.02571234e-01 -2.84017503e-01 -1.55814394e-01
-1.02695537e+00 4.98007715e-01 6.05236143e-02 5.01758993e-01
2.75418043e-01 -1.19751394e+00 -5.69629073e-01 -1.22068241e-01
2.28395134e-01 4.32337642e-01 4.02662903e-02 6.15820348e-01
-1.23158894e-01 3.41896638e-02 3.30340624e-01 -3.95985246e-01
-1.09074295e+00 5.46859980e-01 -1.85495261e-02 -9.37613606e-01
-9.35919404e-01 7.28602111e-01 4.57919300e-01 -5.38990796e-01
2.83926815e-01 -7.41210461e-01 -2.06611410e-01 9.13469940e-02
5.21554112e-01 1.91291451e-01 7.73725286e-02 -9.48730111e-02
-4.95599896e-01 -3.24948579e-02 1.07732445e-01 -2.55611446e-02
1.33397222e+00 3.57137471e-02 -3.25176954e-01 1.36587515e-01
8.75526428e-01 2.08863541e-01 -9.82381105e-01 -3.74623388e-01
6.93518639e-01 1.19638570e-01 -2.30190933e-01 -8.96623015e-01
-9.00694251e-01 2.45034859e-01 -3.66998464e-01 5.87182760e-01
1.24979591e+00 5.11638045e-01 1.01457155e+00 2.03773379e-01
4.25356299e-01 -7.60249794e-01 -1.94273040e-01 6.18941605e-01
5.93029380e-01 -6.91867709e-01 -2.37234116e-01 -1.06468117e+00
-7.73546815e-01 9.27166939e-01 2.98077822e-01 6.90386221e-02
3.44165206e-01 9.25295591e-01 -6.31039500e-01 -6.06180429e-01
-9.91054773e-01 -4.94798750e-01 3.22168529e-01 2.40605250e-01
4.42755222e-01 2.14700460e-01 -3.08157235e-01 1.25112402e+00
7.54811764e-02 7.19407946e-02 1.06354930e-01 1.23449242e+00
-4.67256233e-02 -1.35283625e+00 6.55477121e-02 3.57129216e-01
-8.18315506e-01 -2.33196035e-01 -2.54970670e-01 1.05513692e+00
1.20542698e-01 1.03011215e+00 2.18413383e-01 5.72620891e-02
5.76094091e-01 2.32628465e-01 3.36504251e-01 -8.32504392e-01
-7.92318165e-01 8.87837261e-02 1.00132751e+00 -8.79281461e-01
-2.02626824e-01 -5.34894764e-01 -2.09700298e+00 2.17950374e-01
-1.16383597e-01 3.83861274e-01 2.31610984e-01 9.46041942e-01
6.70820415e-01 1.01995301e+00 1.25321239e-01 -2.35285595e-01
3.58109176e-02 -4.98288542e-01 -4.49826717e-01 6.87678397e-01
-5.31330444e-02 -6.24806821e-01 -1.88916832e-01 2.05173090e-01] | [9.067686080932617, 9.164521217346191] |
55b51326-b311-41eb-82e6-599ff31cce68 | improving-human-robot-collaboration-via | 2303.11425 | null | https://arxiv.org/abs/2303.11425v1 | https://arxiv.org/pdf/2303.11425v1.pdf | Improving Human-Robot Collaboration via Computational Design | When robots entered our day-to-day life, the shared space surrounding humans and robots is critical for effective Human-Robot collaboration. The design of shared space should satisfy humans' preferences and robots' efficiency. This work uses kitchen design as an example to illustrate the importance of good space design in facilitating such collaboration. Given the kitchen boundary, counters, and recipes, the proposed method computes the optimal placement of counters that meet the requirement of kitchen design rules and improve Human-Robot collaboration. The key technical challenge is that the optimization method usually evaluates thousands of designs and the computational cost of motion planning, which is part of the evaluation function, is expensive. We use a decentralized motion planner that can solve multi-agent motion planning efficiently. Our results indicate that optimized kitchen designs can provide noticeable performance improvement to Human-Robot collaboration. | ['Jyh-Ming Lien', 'Jixuan Zhi'] | 2023-03-20 | null | null | null | null | ['motion-planning'] | ['robots'] | [-4.90834415e-01 3.22390586e-01 -1.29967883e-01 -1.50094822e-01
-1.05506115e-01 -3.68901342e-01 7.85226822e-02 -1.50259927e-01
-5.33405542e-01 8.50563645e-01 1.80485249e-01 -1.85410097e-01
-3.02373707e-01 -6.17245257e-01 -2.32724234e-01 -3.10348004e-01
-2.00777650e-01 8.39852333e-01 5.18425889e-02 -3.94283891e-01
2.48583078e-01 5.01529694e-01 -1.15286350e+00 -3.09266061e-01
6.66744411e-01 2.88316727e-01 1.04782867e+00 8.01237762e-01
2.35290214e-01 7.71663368e-01 -6.94296956e-01 3.53816390e-01
5.18010795e-01 -3.09010655e-01 -1.11813283e+00 1.40829250e-01
-7.02993393e-01 -7.17960894e-01 -2.07767248e-01 5.64011335e-01
4.79189187e-01 5.81102252e-01 6.88763916e-01 -1.93219244e+00
-2.76976913e-01 8.55594397e-01 -3.24544132e-01 -4.66300040e-01
5.89369655e-01 5.75461030e-01 8.28429997e-01 -3.40791792e-01
6.21981978e-01 1.44549906e+00 4.59691554e-01 3.63164335e-01
-7.81915843e-01 -2.68394738e-01 -2.34535392e-02 1.78947121e-01
-1.52139926e+00 -9.73201543e-02 3.97096515e-01 -3.39863479e-01
1.26881504e+00 1.61496788e-01 8.15274596e-01 4.83975142e-01
3.89534473e-01 4.67817158e-01 3.70956331e-01 -4.16208148e-01
4.85469341e-01 -4.22230782e-03 -2.00916380e-01 5.30675054e-01
6.43118978e-01 -2.03924328e-01 -2.57481277e-01 -3.52243751e-01
1.10816228e+00 6.35693893e-02 9.99999326e-03 -7.78603911e-01
-1.60865188e+00 6.78499997e-01 4.30770963e-01 3.47218901e-01
-5.59885859e-01 7.55188584e-01 1.61674991e-01 -1.33180037e-01
-6.70161307e-01 9.88109767e-01 -4.89576340e-01 -2.30099887e-01
6.45393953e-02 6.67621076e-01 1.03835583e+00 1.35577583e+00
7.99026489e-01 -6.14540875e-01 1.62200615e-01 5.81433237e-01
5.82663536e-01 5.20152330e-01 2.75932415e-03 -1.63149691e+00
6.52406216e-01 6.05440259e-01 9.96694565e-01 -1.12256658e+00
-9.05307353e-01 3.95050585e-01 -5.98545969e-01 7.55980387e-02
3.20788682e-01 -5.74008882e-01 -1.07460134e-01 1.28066337e+00
6.22180223e-01 -6.76237404e-01 1.42124534e-01 1.12900472e+00
8.51762444e-02 7.08851933e-01 8.68958309e-02 4.08601984e-02
1.32196522e+00 -1.25350487e+00 -9.33063686e-01 -3.93304110e-01
1.17138350e+00 -4.81801808e-01 8.38172317e-01 -2.55439430e-03
-8.32548261e-01 -1.63223550e-01 -1.09243846e+00 3.33584808e-02
9.91122574e-02 2.98839629e-01 8.38027358e-01 3.06294799e-01
-9.03140604e-01 5.59179366e-01 -7.23906696e-01 -8.60536635e-01
-2.27620795e-01 5.55858254e-01 -1.79113284e-01 -9.87324864e-02
-7.98064828e-01 1.46127391e+00 2.69042313e-01 6.01051841e-03
-8.44949126e-01 5.53035960e-02 -8.02070022e-01 8.81894007e-02
6.88415527e-01 -7.89016187e-01 1.76803577e+00 -1.50381297e-01
-1.59134948e+00 2.13597104e-01 5.01139581e-01 -9.73338336e-02
4.65665221e-01 -2.23408565e-01 3.60039830e-01 8.73845443e-02
5.46262801e-01 8.26212943e-01 1.41208455e-01 -1.25560093e+00
-6.94934309e-01 4.67961915e-02 -5.79042435e-02 7.84907639e-01
8.97583738e-02 7.81796351e-02 -7.71902725e-02 -7.65599012e-02
-1.76303256e-02 -1.36522281e+00 -1.07267523e+00 -7.10738897e-02
-1.35654971e-01 -4.23596293e-01 3.10509264e-01 -3.33976746e-01
9.38528657e-01 -2.10366297e+00 8.49442109e-02 2.70528167e-01
-9.61955562e-02 -4.85191733e-01 -1.90238491e-01 9.06562567e-01
5.68616629e-01 -1.57664210e-01 1.30863339e-01 -2.23439977e-01
5.29235244e-01 3.36915344e-01 2.75883824e-01 6.97262526e-01
-5.55941641e-01 7.37931550e-01 -1.12759078e+00 -5.55082917e-01
3.73599142e-01 -1.94188163e-01 -5.56866944e-01 3.54246944e-01
-1.33047432e-01 2.52384245e-01 -6.68939531e-01 4.74551052e-01
2.76136935e-01 -2.73980081e-01 7.88257599e-01 4.34588790e-01
-2.53696173e-01 9.74866599e-02 -1.28090119e+00 1.54414093e+00
-2.04810008e-01 2.35984668e-01 5.15115201e-01 -2.10766211e-01
9.41869080e-01 2.13388413e-01 7.91769743e-01 -3.21436703e-01
3.65195185e-01 3.23869973e-01 -8.35389644e-02 -5.39209247e-01
8.75669420e-01 3.15139621e-01 -7.17664421e-01 9.14718211e-01
-7.19056785e-01 -6.13912463e-01 -1.67279050e-01 8.12656581e-02
1.45370853e+00 -1.66835412e-02 8.69330049e-01 -3.97612274e-01
-2.69942963e-03 5.63808501e-01 6.69575036e-01 6.07472003e-01
-7.49057531e-01 -8.10040385e-02 1.97386757e-01 -4.28837925e-01
-1.34057438e+00 -5.90170860e-01 9.25035059e-01 1.26459897e+00
8.15620184e-01 -2.45995924e-01 -7.97822833e-01 -1.85680240e-01
1.20813370e-01 7.38418639e-01 -7.06787407e-02 1.94573924e-01
-8.92418027e-01 -1.24524146e-01 1.68012619e-01 5.41568398e-01
4.43683296e-01 -1.11821282e+00 -1.78899467e+00 3.03248525e-01
-4.29621339e-01 -9.07543719e-01 -8.27037871e-01 1.69033036e-01
-3.42508495e-01 -1.22538316e+00 -5.88603437e-01 -1.14189041e+00
1.07164478e+00 9.10475194e-01 3.93961817e-01 4.11246687e-01
-3.84885013e-01 6.64142370e-01 -4.30003077e-01 -2.27726221e-01
-1.05636328e-01 5.21593541e-02 4.33242530e-01 -9.80214596e-01
-9.29514244e-02 -4.05873388e-01 -6.18209779e-01 9.21560526e-01
-4.18044291e-02 4.48286802e-01 6.83700621e-01 5.44647813e-01
-1.03496820e-01 4.67770845e-01 4.72932190e-01 1.04519658e-01
8.13320041e-01 -4.59662676e-01 -5.26611626e-01 2.24253744e-01
-2.41842911e-01 1.40201952e-02 2.63650537e-01 -3.61992538e-01
-9.44283545e-01 5.41161954e-01 9.37632799e-01 1.58376485e-01
-8.30647126e-02 4.53852825e-02 -3.57014805e-01 1.80553168e-01
5.64562440e-01 -6.42503381e-01 2.17835918e-01 -1.74177408e-01
4.77289796e-01 8.01423073e-01 4.85487103e-01 -7.74545789e-01
3.71519208e-01 1.32645205e-01 -5.99801242e-02 -5.71082413e-01
2.67494678e-01 -6.02977276e-01 -6.63533330e-01 -3.76726151e-01
9.34329212e-01 -7.38311887e-01 -1.40425265e+00 1.02136813e-01
-1.72089589e+00 -9.44474220e-01 4.61218366e-03 5.75409412e-01
-9.86344337e-01 3.50762963e-01 -5.62964499e-01 -1.30580950e+00
-9.48333368e-03 -1.18244696e+00 7.29604602e-01 1.14053547e-01
-8.35314572e-01 -3.30272406e-01 1.43330887e-01 3.74649346e-01
2.57220745e-01 1.78479820e-01 6.03472769e-01 -1.68931901e-01
-9.33766961e-01 1.39109090e-01 -9.54696722e-03 -5.39984167e-01
2.01295599e-01 -3.70686650e-01 -6.65593371e-02 -2.26703227e-01
-3.12429816e-01 -4.10005562e-02 -3.05696636e-01 4.35465306e-01
2.38002509e-01 -5.91914952e-01 -7.87904859e-01 -2.69981056e-01
1.00884891e+00 5.76862812e-01 5.03213763e-01 7.57799268e-01
3.58531207e-01 1.13334215e+00 1.34493148e+00 1.01301122e+00
9.24079597e-01 7.62895346e-01 3.16910625e-01 3.77095938e-01
5.56768179e-01 -2.06452191e-01 5.11801004e-01 4.39975619e-01
2.26144329e-01 -9.88666266e-02 -1.20016229e+00 8.39737535e-01
-2.66334605e+00 -7.10283756e-01 -8.63945633e-02 1.75315487e+00
3.42656285e-01 -4.23712939e-01 3.45012605e-01 1.63410157e-02
9.79843915e-01 -4.61137056e-01 -2.41138101e-01 -4.64583427e-01
5.64890385e-01 -9.03023720e-01 9.76691008e-01 4.33496892e-01
-8.14101756e-01 8.90415668e-01 7.27676487e+00 1.82693720e-01
-1.17296800e-01 8.50547850e-02 1.21019222e-01 -1.01930693e-01
-1.40770141e-03 2.25853249e-01 -4.43280637e-01 2.96849042e-01
5.89233756e-01 -2.78559893e-01 7.72772729e-01 1.31560349e+00
5.47284961e-01 -7.67003179e-01 -1.29198074e+00 1.26773167e+00
-3.91878456e-01 -8.97991002e-01 -7.90864825e-01 3.54235470e-01
6.28815949e-01 -3.70322883e-01 -3.87927622e-01 1.08506374e-01
1.09786582e+00 -9.58475053e-01 1.07947505e+00 2.00984150e-01
1.20753996e-01 -8.35475445e-01 7.63833046e-01 8.56608629e-01
-1.30524862e+00 -4.27538425e-01 -3.95596117e-01 -6.94938779e-01
7.65781701e-01 6.81279078e-02 -1.57102883e+00 1.08616866e-01
6.54085040e-01 1.10522658e-02 1.10129192e-01 9.75677729e-01
-2.00029179e-01 -7.02865899e-01 -4.93512958e-01 -8.52725029e-01
1.86998397e-01 -2.74628520e-01 3.60266298e-01 7.03482091e-01
4.53886062e-01 5.37743926e-01 5.41370749e-01 6.82838321e-01
5.31093419e-01 3.13859172e-02 -9.01593685e-01 -1.16576545e-01
1.04930806e+00 1.00991905e+00 -9.77016151e-01 -1.19535476e-02
1.21136315e-01 8.73439252e-01 1.55782640e-01 5.07200928e-03
-8.81944597e-01 -4.24697787e-01 8.70162189e-01 -2.46728063e-01
5.16425781e-02 -9.02276278e-01 -7.31966853e-01 -1.95536390e-01
-1.28365904e-01 -5.35420299e-01 -3.40675004e-02 -8.67630184e-01
-6.49180830e-01 8.26485083e-03 2.06927493e-01 -1.09373701e+00
-4.79239881e-01 -2.71273017e-01 -3.46522510e-01 3.67850423e-01
-5.56069374e-01 -8.98705363e-01 -3.76437455e-01 1.50164083e-01
3.81751359e-01 -1.76858008e-01 5.32254696e-01 -3.20648327e-02
-4.18477297e-01 9.68670771e-02 -2.90033758e-01 -4.01908606e-01
5.72344840e-01 -7.86099017e-01 4.81771588e-01 4.88821864e-01
-7.90620327e-01 6.70625448e-01 9.97666478e-01 -1.10396624e+00
-1.83961225e+00 -6.50029778e-01 8.17005455e-01 -5.41319132e-01
3.82716477e-01 -1.82490632e-01 -1.26624525e-01 7.06882298e-01
1.61830172e-01 -6.08763576e-01 4.81377929e-01 -9.96068195e-02
5.22531092e-01 4.11212325e-01 -1.45087969e+00 1.01597142e+00
1.34579611e+00 2.96660941e-02 -5.19262910e-01 3.10594350e-01
1.02012408e+00 -3.68307322e-01 -5.88252127e-01 -2.32534651e-02
5.02133906e-01 -5.06340802e-01 6.44867420e-01 2.47196406e-02
-1.32563069e-01 -6.58995032e-01 -4.13719207e-01 -1.27717638e+00
-6.63504183e-01 -1.03281569e+00 4.14469838e-01 8.24468970e-01
2.50404716e-01 -4.63193715e-01 7.22558200e-01 1.54255009e+00
-1.48594499e-01 -3.96908224e-01 -6.49943590e-01 -8.55690122e-01
-6.49867833e-01 -1.26708969e-01 9.95860338e-01 6.93822563e-01
7.83251286e-01 2.60022253e-01 -5.23562312e-01 2.90964216e-01
5.06241739e-01 -3.65596294e-01 1.40794146e+00 -8.42478633e-01
-2.83918589e-01 -2.86504269e-01 -3.93632203e-02 -9.80772316e-01
-9.74586457e-02 -2.20629558e-01 7.48268068e-01 -2.16435266e+00
5.90157881e-02 -9.50717509e-01 6.30723298e-01 5.59301436e-01
3.61190885e-01 -7.88287759e-01 7.71389961e-01 3.17403644e-01
-1.01026404e+00 5.47361851e-01 1.20465350e+00 -1.28995311e-02
-6.81762278e-01 -3.43345881e-01 -5.77509701e-01 6.71293855e-01
9.24785137e-01 -1.78226486e-01 -3.67199361e-01 -5.10819197e-01
2.71372259e-01 3.81880194e-01 6.60715625e-02 -9.95482683e-01
7.69573748e-01 -1.00128388e+00 -1.33961797e-01 -6.05284333e-01
3.81738931e-01 -1.21614635e+00 7.00910151e-01 9.81404781e-01
-1.11304455e-01 3.51256222e-01 -3.18838298e-01 4.49322611e-01
5.98684132e-01 -3.11352044e-01 3.60259265e-01 -2.88210601e-01
-8.78350139e-01 -3.04266870e-01 -8.87583971e-01 -6.91890717e-01
1.77488005e+00 -1.68104231e-01 -3.43181521e-01 -4.16339099e-01
-2.92070895e-01 1.02923477e+00 9.17393565e-01 2.55946279e-01
5.25099754e-01 -1.40059197e+00 -2.16357365e-01 -4.64133948e-01
-2.29811281e-01 3.37350577e-01 1.42852768e-01 4.67296571e-01
-1.10343003e+00 5.46656311e-01 -4.77185071e-01 -3.21121037e-01
-1.19992757e+00 6.14523113e-01 1.14319183e-01 -1.17453255e-01
-7.15231538e-01 4.73008215e-01 1.65170684e-01 -7.33771443e-01
4.90176737e-01 -5.74373662e-01 8.85121971e-02 -5.30893743e-01
4.25053596e-01 1.06260777e+00 -7.44391203e-01 -3.11861366e-01
-6.69550836e-01 3.37123215e-01 6.43092513e-01 -5.75144827e-01
1.16179597e+00 -7.23074019e-01 -2.36194462e-01 1.34950608e-01
5.22859693e-01 -4.53874856e-01 -1.22502410e+00 2.78972059e-01
1.70313492e-01 -5.66341758e-01 -5.71926117e-01 -5.12793183e-01
-2.44054109e-01 1.69578847e-02 1.44867674e-01 -1.39459208e-01
4.79275703e-01 -5.59790656e-02 9.01940167e-01 9.99698639e-01
1.60788536e+00 -1.69768023e+00 4.31455016e-01 5.63445985e-01
9.34183598e-01 -9.52654719e-01 2.48462170e-01 -5.22783279e-01
-1.04227519e+00 5.18948317e-01 1.03106081e+00 -1.67388786e-02
1.63426295e-01 2.09750399e-01 -2.38229021e-01 -8.93366933e-02
-7.39285111e-01 6.65628091e-02 -6.91381514e-01 8.36323261e-01
-1.87741369e-01 7.54855692e-01 -3.56970966e-01 7.41449654e-01
-4.58544910e-01 -2.01880768e-01 7.95709610e-01 1.41326010e+00
-1.17296708e+00 -9.26843464e-01 -1.01026869e+00 -2.31185891e-02
5.28711855e-01 9.12099540e-01 -4.54908222e-01 5.53117216e-01
5.77071160e-02 1.40422535e+00 1.71457808e-02 -3.30267429e-01
4.25836474e-01 -3.90879035e-01 3.86141866e-01 -6.82223499e-01
-5.36672235e-01 -1.82949394e-01 5.14598191e-01 -8.09699893e-01
-5.57788052e-02 -5.43064713e-01 -2.09418011e+00 -5.84565282e-01
-1.72711238e-01 3.38947892e-01 8.80148709e-01 8.13355923e-01
3.96972626e-01 2.73148924e-01 6.50722146e-01 -1.35895288e+00
-5.89520872e-01 -8.10027182e-01 -6.50476396e-01 -3.15032043e-02
1.14847302e-01 -9.79457915e-01 -6.19160570e-02 -3.23364317e-01] | [4.892285346984863, 1.15488600730896] |
7f5ecc67-9d7a-4ed1-848e-7c600a29b2f3 | character-aware-neural-networks-for-arabic | null | null | https://aclanthology.org/W16-3703 | https://aclanthology.org/W16-3703.pdf | Character-Aware Neural Networks for Arabic Named Entity Recognition for Social Media | Named Entity Recognition (NER) is the task of classifying or labelling atomic elements in the text into categories such as Person, Location or Organisation. For Arabic language, recognizing named entities is a challenging task because of the complexity and the unique characteristics of this language. In addition, most of the previous work focuses on Modern Standard Arabic (MSA), however, recognizing named entities in social media is becoming more interesting these days. Dialectal Arabic (DA) and MSA are both used in social media, which is deemed as another challenging task. Most state-of-the-art Arabic NER systems count heavily on handcrafted engineering features and lexicons which is time consuming. In this paper, we introduce a novel neural network architecture which benefits both from character- and word-level representations automatically, by using combination of bidirectional LSTM and Conditional Random Field (CRF), eliminating the need for most feature engineering. Moreover, our model relies on unsupervised word representations learned from unannotated corpora. Experimental results demonstrate that our model achieves state-of-the-art performance on publicly available benchmark for Arabic NER for social media and surpassing the previous system by a large margin. | ['Mourad Gridach'] | 2016-12-01 | null | null | null | ws-2016-12 | ['text-clustering'] | ['natural-language-processing'] | [-2.61120081e-01 -2.39406317e-01 1.41960770e-01 -3.51028621e-01
-5.00123620e-01 -6.24901116e-01 7.15968966e-01 3.35781544e-01
-9.91778672e-01 8.85532618e-01 3.27960610e-01 -2.51680702e-01
2.75483787e-01 -1.08030272e+00 -4.36935604e-01 -5.43435276e-01
-2.17666611e-01 4.20330465e-01 2.23157242e-01 -7.43041754e-01
2.65510529e-01 7.52749920e-01 -1.01972508e+00 7.16944411e-02
1.12436593e+00 7.90769696e-01 -5.36793321e-02 1.89625278e-01
-6.59609616e-01 8.20718765e-01 -5.95618844e-01 -7.53008962e-01
-1.31338373e-01 -3.58312368e-01 -1.05289638e+00 -7.69055933e-02
-1.54874116e-01 -1.02103055e-01 -4.90153991e-02 1.03094876e+00
5.19068837e-01 3.94159198e-01 7.12719262e-01 -6.93310320e-01
-9.02550399e-01 8.76683056e-01 -5.30751705e-01 -1.41154286e-02
6.50748163e-02 -3.98328096e-01 9.20054913e-01 -1.10760021e+00
5.17780423e-01 8.53684723e-01 7.53726721e-01 3.09545279e-01
-5.31126380e-01 -5.67038417e-01 2.05910802e-02 2.08221689e-01
-1.58886790e+00 -4.21594501e-01 3.84870321e-01 -1.84719145e-01
9.76201713e-01 -2.47734696e-01 1.74449801e-01 8.30140769e-01
1.31379858e-01 5.86028934e-01 1.13359833e+00 -7.53983855e-01
1.77822739e-01 4.10106480e-02 2.87682235e-01 7.09736824e-01
1.23417392e-01 -6.53790057e-01 -1.82395801e-01 -5.88481799e-02
5.20962834e-01 -7.18715712e-02 -1.16455562e-01 1.08130701e-01
-1.28163040e+00 7.33171701e-01 4.69430387e-01 7.64601231e-01
-6.24124110e-01 -3.31706226e-01 3.37622523e-01 -1.54340966e-03
2.69485712e-01 2.67751038e-01 -6.45051718e-01 -3.01965594e-01
-7.51585901e-01 -3.56600404e-01 1.07508683e+00 9.31900740e-01
6.44054890e-01 2.27746055e-01 2.57433832e-01 1.01585841e+00
4.73194450e-01 5.80617964e-01 8.34476411e-01 4.01132256e-02
3.56990844e-01 1.04637599e+00 2.38146156e-01 -1.14533603e+00
-5.94485521e-01 -3.67983520e-01 -9.85726774e-01 -1.95515499e-01
4.89183128e-01 -6.04929864e-01 -1.16050160e+00 1.36603081e+00
3.54614824e-01 4.92364839e-02 3.03177267e-01 6.78950846e-01
8.61873627e-01 9.75176454e-01 3.49919140e-01 -1.14480384e-01
1.57028663e+00 -8.53776872e-01 -7.85855055e-01 -1.54664725e-01
6.36165321e-01 -8.75954270e-01 6.39469206e-01 9.23786089e-02
-6.94163322e-01 -1.70868263e-01 -8.95102322e-01 -6.71650395e-02
-1.06506419e+00 4.77224022e-01 5.83114147e-01 8.90414357e-01
-6.87846601e-01 3.10582250e-01 -9.00853932e-01 -6.67256892e-01
2.35511452e-01 5.54898918e-01 -6.47603571e-01 1.23021677e-01
-1.52032375e+00 1.11591470e+00 6.96722746e-01 4.99965042e-01
-3.41144651e-01 3.05372477e-02 -1.01803088e+00 9.92989168e-02
2.64428437e-01 6.17931224e-02 8.62098277e-01 -9.99766469e-01
-1.61920643e+00 5.68158090e-01 -3.62933986e-02 -3.22009116e-01
-4.53779101e-02 -3.68793905e-01 -1.10520363e+00 -1.68861672e-01
-3.99962403e-02 3.70939553e-01 3.92672330e-01 -1.06777048e+00
-7.70841599e-01 -2.88160235e-01 7.49544725e-02 1.39638290e-01
-7.62086928e-01 5.89262366e-01 -1.97969466e-01 -7.27700770e-01
7.40724131e-02 -8.44300330e-01 -2.28860080e-01 -7.32629716e-01
-4.65418398e-01 -4.95089501e-01 5.40778160e-01 -1.04723859e+00
1.52146506e+00 -1.92962694e+00 -1.73249349e-01 2.40340069e-01
-1.16589457e-01 5.45600533e-01 -3.72029245e-02 7.29413152e-01
9.02731493e-02 2.17779905e-01 -3.58543485e-01 1.99779630e-01
5.48057035e-02 -7.62964189e-02 -2.04334721e-01 3.97121519e-01
7.18995571e-01 6.52748048e-01 -8.06005359e-01 -6.34744227e-01
-2.02436030e-01 6.97043896e-01 -5.94450161e-02 6.57178387e-02
1.07981250e-01 4.39819723e-01 -3.69813651e-01 7.32783556e-01
7.00205743e-01 -1.68786570e-01 2.93013304e-01 -1.86642930e-01
-4.46137160e-01 2.23046303e-01 -1.44078112e+00 1.50550914e+00
-5.64277887e-01 9.70519781e-02 -3.53619874e-01 -8.27919364e-01
1.07140684e+00 5.43630362e-01 7.93179381e-04 -3.08771819e-01
5.17128289e-01 5.94781160e-01 6.15765294e-03 -2.90532619e-01
7.32826889e-01 6.07975200e-02 -3.21736276e-01 5.51404536e-01
3.53483230e-01 5.77634037e-01 4.58077788e-01 3.89041267e-02
7.62624860e-01 2.63277858e-01 7.45018601e-01 -1.27441362e-01
8.12661350e-01 2.28981627e-03 6.01455629e-01 4.12198603e-01
-1.55649275e-01 2.88886279e-01 1.36359245e-01 -3.99401903e-01
-8.12712491e-01 -8.42542171e-01 -2.54444301e-01 1.35136425e+00
-3.48953940e-02 -6.00836910e-02 -7.63133824e-01 -1.02880335e+00
-5.94629169e-01 5.84153295e-01 -4.29940134e-01 3.57369661e-01
-9.34009433e-01 -1.09140515e+00 8.26151848e-01 4.51967776e-01
7.28850305e-01 -1.43235028e+00 -1.79179713e-01 4.98857111e-01
-2.14633361e-01 -1.13789058e+00 -3.55620593e-01 1.19786449e-01
-5.82102716e-01 -7.37235367e-01 -1.00334716e+00 -1.31567383e+00
7.27451980e-01 -1.34065479e-01 1.07358444e+00 -1.43472925e-01
2.46662870e-01 9.39345267e-03 -9.05358255e-01 -4.56504911e-01
-1.93112910e-01 6.31089747e-01 7.12921545e-02 5.45332491e-01
8.16482186e-01 -3.48671883e-01 -1.65138185e-01 2.65554845e-01
-1.02262115e+00 -3.24900836e-01 9.00265336e-01 8.96306455e-01
2.12652564e-01 -7.85859153e-02 9.62029815e-01 -9.89676297e-01
4.73542303e-01 -8.21770370e-01 -2.58523971e-01 5.74668169e-01
-1.43032998e-01 9.82613415e-02 7.86603928e-01 -3.53327423e-01
-1.39897442e+00 3.01409960e-01 -3.70973527e-01 7.15935946e-01
-4.93190855e-01 9.66180146e-01 -3.19761992e-01 1.44492656e-01
5.65032065e-01 3.48969311e-01 -4.34542894e-01 -4.05845344e-01
4.33086067e-01 1.37380242e+00 2.89630026e-01 -4.50447768e-01
5.33408225e-01 2.80131876e-01 -3.04992735e-01 -1.04261792e+00
-8.17127764e-01 -4.51685756e-01 -1.05977225e+00 -1.68715883e-02
9.98158514e-01 -9.29942966e-01 -6.45045519e-01 7.50622213e-01
-1.23135948e+00 2.63359696e-01 2.07083821e-01 5.95758379e-01
1.24906518e-01 4.64647442e-01 -8.25998724e-01 -8.76715541e-01
-5.16385496e-01 -9.17660356e-01 6.25033975e-01 6.32750630e-01
4.86753024e-02 -9.51160371e-01 2.49518431e-03 4.95968424e-02
4.93394285e-01 2.63576269e-01 7.88321733e-01 -1.34161735e+00
-1.07036352e-01 -3.73221099e-01 -3.97710115e-01 1.99594751e-01
3.63234669e-01 -1.14707336e-01 -7.55698740e-01 -4.93693762e-02
-2.51102895e-01 -3.71603161e-01 5.35028398e-01 -2.15522245e-01
3.84058356e-01 -2.87878484e-01 5.65867946e-02 6.31712517e-03
1.39201391e+00 3.27426821e-01 5.45338154e-01 6.50540650e-01
8.19776595e-01 6.35049164e-01 4.82333541e-01 5.16303301e-01
6.66082323e-01 2.48890191e-01 1.78348169e-01 -9.86279026e-02
3.40100169e-01 8.71274471e-02 5.32400668e-01 1.41533291e+00
-2.31125832e-01 -2.46455833e-01 -1.39230895e+00 8.72124016e-01
-1.65616548e+00 -6.70439065e-01 -2.54879534e-01 1.91187012e+00
1.26322043e+00 -1.75519496e-01 -4.49479334e-02 2.26816200e-02
1.09986460e+00 -6.81179296e-03 -1.57239079e-01 -3.94605368e-01
-2.47446761e-01 5.41141570e-01 4.81457025e-01 8.78207833e-02
-1.45671856e+00 1.23698246e+00 4.55437279e+00 9.28018153e-01
-1.13139033e+00 1.52162805e-01 4.63905782e-01 8.23854208e-01
2.70504355e-01 -2.13802129e-01 -1.15661740e+00 3.79890382e-01
1.25689805e+00 1.87675521e-01 1.52564526e-01 6.34399652e-01
-1.20995477e-01 1.59071416e-01 -5.14183581e-01 5.75872898e-01
2.46656552e-01 -8.38481128e-01 1.39356405e-01 -1.73277408e-01
8.00954282e-01 1.18440218e-01 -3.91456366e-01 4.43917304e-01
5.24876773e-01 -1.04579306e+00 4.78204817e-01 5.92887998e-01
6.55879259e-01 -1.14391148e+00 1.37625730e+00 2.21650973e-01
-1.19473243e+00 8.66221339e-02 -2.92897403e-01 1.47783071e-01
3.04744929e-01 4.50373024e-01 -6.39441609e-01 6.26686752e-01
6.51227415e-01 5.12177646e-01 -5.29627621e-01 1.03255355e+00
-2.91562438e-01 8.22677791e-01 -3.75474870e-01 -3.45875412e-01
4.97224778e-01 -2.02822044e-01 7.82130100e-03 1.55470300e+00
4.15034652e-01 1.01390950e-01 1.59165427e-01 1.61781698e-01
-3.77588063e-01 8.89346600e-01 -2.98209876e-01 -4.49277371e-01
3.94329786e-01 1.40135503e+00 -1.10007858e+00 -1.90009058e-01
-5.30203879e-01 9.37357485e-01 5.09792626e-01 1.49886593e-01
-6.48801863e-01 -1.24836659e+00 3.45269740e-02 -3.00440222e-01
5.01938879e-01 -6.08485103e-01 -1.06804781e-01 -1.26473188e+00
-1.46631926e-01 -8.27688873e-01 3.31555784e-01 -1.85137302e-01
-1.64364123e+00 1.12003458e+00 -6.38585806e-01 -1.29630816e+00
-1.02403119e-01 -8.30764294e-01 -2.63471186e-01 8.11180413e-01
-1.60103881e+00 -1.46131837e+00 -1.18178628e-01 5.92923641e-01
3.12688708e-01 -3.90372366e-01 1.17855012e+00 5.60525477e-01
-6.68357730e-01 6.35883987e-01 3.14052790e-01 9.93562341e-01
9.70522344e-01 -1.15841997e+00 4.51690346e-01 9.06710863e-01
2.79505908e-01 8.69679630e-01 1.09190390e-01 -5.55289745e-01
-1.15018690e+00 -9.26122844e-01 1.55165565e+00 -3.53519350e-01
1.01113260e+00 -2.19345957e-01 -7.93822467e-01 6.80124044e-01
5.45463026e-01 -6.95215957e-03 1.14310408e+00 1.48748949e-01
-2.73803443e-01 6.41734675e-02 -1.02738285e+00 6.87490702e-01
6.05919421e-01 -2.62554884e-01 -7.59353042e-01 2.22433329e-01
4.95839566e-01 -2.02805698e-01 -1.11207700e+00 2.10105255e-01
4.53291565e-01 -5.77103913e-01 5.19242167e-01 -8.76676917e-01
2.01043844e-01 -4.60393846e-01 -1.19081251e-01 -1.26760805e+00
-4.48260047e-02 -3.86646569e-01 1.83731467e-01 1.77642465e+00
7.06831992e-01 -7.37876415e-01 4.18194652e-01 3.05192232e-01
-1.13021629e-02 -4.38307673e-01 -7.34212637e-01 -4.63898867e-01
1.65460646e-01 -9.88874212e-02 7.75705099e-01 1.50441849e+00
1.15646400e-01 5.26618779e-01 -4.35597897e-01 4.07695115e-01
1.68770567e-01 -1.89695463e-01 2.15227038e-01 -1.21997762e+00
2.62717664e-01 -1.31561860e-01 -4.98803318e-01 -9.07967627e-01
2.28043497e-01 -7.16207922e-01 2.86351323e-01 -1.66152298e+00
-3.19895566e-01 -5.77051401e-01 -5.03321230e-01 7.58859158e-01
-1.02789812e-01 4.04718488e-01 1.52137503e-01 1.27364352e-01
-6.93111241e-01 5.11781812e-01 6.58888996e-01 -4.85802814e-02
-3.64841133e-01 -3.21905524e-01 -4.68312711e-01 7.90132701e-01
9.22106504e-01 -6.34707272e-01 3.13705355e-01 -6.17259502e-01
5.29104173e-01 -2.14847252e-01 -4.30135161e-01 -1.01194274e+00
4.52911735e-01 -2.14464918e-01 3.51466089e-01 -3.72366220e-01
-4.75538708e-02 -8.69376421e-01 -4.39243615e-01 1.69301704e-01
4.93412130e-02 3.20267677e-01 4.01762240e-02 3.18104029e-01
-4.21092957e-01 -5.49227238e-01 7.20964313e-01 -5.68157136e-02
-1.06321275e+00 1.10246696e-01 -6.72085166e-01 1.42352492e-01
8.42171550e-01 9.13053229e-02 -2.39409238e-01 -1.74745210e-02
-7.04624712e-01 -5.59251569e-03 4.19272669e-02 4.49988067e-01
2.97536612e-01 -1.28745830e+00 -9.30345595e-01 -4.58470695e-02
1.16045780e-01 -1.73207372e-01 1.03679530e-01 6.48639679e-01
-9.21299815e-01 3.31918746e-01 -3.97957891e-01 -5.50274663e-02
-7.68069565e-01 -4.48715836e-02 -2.69935708e-02 -4.07321781e-01
-9.49774235e-02 6.72486007e-01 -4.58601564e-01 -9.19847429e-01
-2.36917157e-02 1.41292334e-01 -1.09950066e+00 4.12814438e-01
5.12100816e-01 3.57340932e-01 2.05633387e-01 -1.31648588e+00
-3.95403773e-01 3.43849957e-01 -3.07524472e-01 2.63798404e-02
1.45779920e+00 -4.45021056e-02 -5.43302059e-01 4.46624309e-01
8.39333475e-01 4.14747983e-01 -4.46269631e-01 -3.11146379e-01
5.44601083e-01 1.38353273e-01 -2.19667152e-01 -9.42660630e-01
-7.87426114e-01 8.14126313e-01 5.82717836e-01 1.15460768e-01
1.00021148e+00 -3.99691492e-01 1.17651963e+00 9.12549555e-01
5.66356957e-01 -1.22930801e+00 -3.74835104e-01 1.29070878e+00
5.01227677e-01 -1.24201083e+00 -3.00726980e-01 -1.84136763e-01
-6.93762541e-01 1.31886721e+00 4.53131765e-01 -1.52909368e-01
9.81086195e-01 2.83878054e-02 4.55835730e-01 2.54592389e-01
6.48032948e-02 -4.03068870e-01 2.50241458e-01 4.29541230e-01
8.46845031e-01 -1.09636039e-01 -7.71375418e-01 1.06108689e+00
-1.57323420e-01 -1.85969636e-01 7.70675540e-01 1.31026566e+00
-5.25386930e-01 -1.27435899e+00 -3.30483854e-01 3.91007960e-01
-9.52625334e-01 -4.88711685e-01 -1.27749845e-01 6.85462892e-01
2.48496741e-01 1.23019707e+00 1.73910260e-02 -1.90503165e-01
2.20266238e-01 3.20972413e-01 -1.10161342e-01 -6.80890262e-01
-9.17042196e-01 -2.42831171e-01 2.14827910e-01 1.21937163e-01
-7.31857896e-01 -4.66425329e-01 -1.77857780e+00 -1.39784619e-01
-7.41852820e-01 4.61180359e-01 9.75270212e-01 1.21982706e+00
4.78873104e-01 8.46496969e-02 6.49732172e-01 -5.87023199e-01
-3.85910153e-01 -1.31273997e+00 -6.95234060e-01 2.00220168e-01
-1.16143689e-01 -4.94409263e-01 -6.68511018e-02 2.74236709e-01] | [9.810142517089844, 9.799256324768066] |
a485fcae-54b5-4198-ad3a-03b6998b4829 | joint-feature-distribution-alignment-learning | 2204.11434 | null | https://arxiv.org/abs/2204.11434v1 | https://arxiv.org/pdf/2204.11434v1.pdf | Joint Feature Distribution Alignment Learning for NIR-VIS and VIS-VIS Face Recognition | Face recognition for visible light (VIS) images achieve high accuracy thanks to the recent development of deep learning. However, heterogeneous face recognition (HFR), which is a face matching in different domains, is still a difficult task due to the domain discrepancy and lack of large HFR dataset. Several methods have attempted to reduce the domain discrepancy by means of fine-tuning, which causes significant degradation of the performance in the VIS domain because it loses the highly discriminative VIS representation. To overcome this problem, we propose joint feature distribution alignment learning (JFDAL) which is a joint learning approach utilizing knowledge distillation. It enables us to achieve high HFR performance with retaining the original performance for the VIS domain. Extensive experiments demonstrate that our proposed method delivers statistically significantly better performances compared with the conventional fine-tuning approach on a public HFR dataset Oulu-CASIA NIR&VIS and popular verification datasets in VIS domain such as FLW, CFP, AgeDB. Furthermore, comparative experiments with existing state-of-the-art HFR methods show that our method achieves a comparable HFR performance on the Oulu-CASIA NIR&VIS dataset with less degradation of VIS performance. | ['Hitoshi Imaoka', 'Akinori F. Ebihara', 'Akihiro Hayasaka', 'Hiroshi Hashimoto', 'Takaya Miyamoto'] | 2022-04-25 | null | null | null | null | ['heterogeneous-face-recognition'] | ['computer-vision'] | [ 2.55173799e-02 -6.61190510e-01 -3.34525853e-02 -4.66727108e-01
-1.04979968e+00 -1.81255832e-01 7.04558134e-01 -6.62793398e-01
-5.51823005e-02 8.16953242e-01 9.14098769e-02 8.76247808e-02
-3.05072665e-01 -5.82877100e-01 -5.33924758e-01 -9.44236577e-01
4.04006511e-01 3.32304835e-01 -2.19963774e-01 -3.52426052e-01
-5.42542674e-02 8.08625340e-01 -1.93816400e+00 3.99421394e-01
6.25800490e-01 1.23758161e+00 -7.42399693e-02 1.73016816e-01
6.22409135e-02 4.97814327e-01 -5.23740053e-01 -6.74867570e-01
6.88362539e-01 -3.59603941e-01 -4.22312409e-01 -6.51499853e-02
1.15000784e+00 -2.67337382e-01 -7.01843560e-01 1.01905775e+00
1.01746547e+00 2.26351358e-02 7.86325753e-01 -1.46756041e+00
-9.47224677e-01 7.74891078e-02 -9.65703070e-01 -8.75368491e-02
5.01963735e-01 -2.52753105e-02 6.85823977e-01 -1.15252173e+00
5.54910958e-01 1.54951537e+00 9.51452792e-01 7.89044440e-01
-9.62240934e-01 -1.26675367e+00 -3.07396024e-01 5.00613272e-01
-1.74944425e+00 -7.59246171e-01 8.14051986e-01 -3.96023065e-01
7.95834363e-01 1.20066933e-01 1.64591193e-01 1.16379619e+00
-1.12762101e-01 5.01491487e-01 1.40268672e+00 -3.81056279e-01
-2.53126293e-01 1.52347773e-01 -7.18322843e-02 6.12657964e-01
9.91217792e-02 4.21234757e-01 -8.53311479e-01 -7.32338428e-02
5.01403451e-01 5.54240420e-02 -3.92749310e-01 -1.40128776e-01
-8.42944086e-01 5.79783320e-01 1.85722873e-01 7.69002959e-02
-1.76088750e-01 -3.49279344e-01 2.67661959e-01 6.33026004e-01
5.21701634e-01 -1.65274799e-01 -4.45447475e-01 3.76156598e-01
-9.77222145e-01 7.97792524e-02 6.29895627e-01 7.33851194e-01
7.80080259e-01 1.51393488e-01 -4.23312187e-01 1.33319521e+00
4.64712650e-01 8.53603363e-01 3.41047555e-01 -5.14392078e-01
4.69090194e-01 4.11672235e-01 -1.30928174e-01 -7.77712762e-01
-6.46525249e-02 -3.53981853e-01 -9.89581704e-01 3.69067013e-01
3.17242205e-01 1.08946741e-01 -1.10919619e+00 1.72654617e+00
3.87194127e-01 3.60844821e-01 3.55148882e-01 7.19132662e-01
1.45750165e+00 3.83573920e-01 -6.66233152e-02 -2.85192132e-01
1.18559730e+00 -8.84975553e-01 -7.61887789e-01 1.50139675e-01
1.41436979e-01 -1.14862192e+00 6.30167305e-01 3.09067667e-01
-6.33120596e-01 -8.72420430e-01 -9.67263937e-01 6.79757744e-02
-3.19652736e-01 5.56898475e-01 5.91541886e-01 1.08301294e+00
-9.83838558e-01 2.65650570e-01 -7.87155256e-02 -3.90509367e-01
9.22426403e-01 5.34649014e-01 -9.95141566e-01 -6.50166035e-01
-1.03062057e+00 7.62085199e-01 9.28647295e-02 1.87186435e-01
-1.06834543e+00 -1.04448175e+00 -8.24532568e-01 -1.51794627e-01
2.83499748e-01 -3.26595366e-01 7.42526889e-01 -8.85732889e-01
-1.50874615e+00 1.16742969e+00 -2.14043111e-01 -3.51735950e-02
6.46133006e-01 -6.32238910e-02 -8.68915141e-01 -7.15677291e-02
-2.64337391e-01 4.17990655e-01 1.16372323e+00 -1.21733594e+00
-3.58982176e-01 -7.92407274e-01 -2.74301767e-01 -6.76319152e-02
-4.32067305e-01 1.99311778e-01 -5.01953781e-01 -4.33202326e-01
-9.48587358e-02 -8.81583989e-01 4.74348783e-01 -5.33834472e-02
-4.40341309e-02 -5.10472536e-01 1.27977335e+00 -9.89152253e-01
7.75687039e-01 -2.27396059e+00 -2.07000792e-01 1.43060178e-01
-7.81378523e-03 7.61648238e-01 -4.95805234e-01 3.18339944e-01
-4.81666476e-01 -4.64381546e-01 -5.38824275e-02 -3.54896486e-01
-1.29427791e-01 -1.37013525e-01 -1.37545496e-01 8.93431067e-01
-6.65934896e-03 8.33825350e-01 -4.11218703e-01 -5.00330567e-01
2.41102651e-01 9.13346529e-01 -3.33281904e-01 3.85630816e-01
1.16574660e-01 6.45828366e-01 -6.91135004e-02 1.22222698e+00
1.34034526e+00 7.65746906e-02 8.12404379e-02 -6.53512955e-01
1.34945467e-01 -3.48400086e-01 -1.32306063e+00 1.51389813e+00
-3.34930897e-01 6.14014685e-01 2.51780637e-02 -1.13058650e+00
1.44504452e+00 4.42766845e-01 5.83726883e-01 -9.53027904e-01
1.30562901e-01 4.26903635e-01 -1.46316037e-01 -5.29325008e-01
8.92762989e-02 -2.33361840e-01 4.06389952e-01 7.59040844e-03
2.69928098e-01 2.57934928e-01 -1.57802567e-01 -2.90980816e-01
7.21030235e-01 2.87741572e-01 1.70964032e-01 -1.13116957e-01
1.07670164e+00 -4.61238921e-01 8.03621531e-01 4.66336042e-01
-4.64121044e-01 7.15435445e-01 -7.00564906e-02 -2.70163655e-01
-7.19526470e-01 -1.00511026e+00 -5.24990618e-01 7.78723538e-01
9.03020650e-02 -1.41707256e-01 -4.20670450e-01 -8.32622051e-01
5.21260977e-01 2.31481548e-02 -6.68064117e-01 -1.00634724e-01
-5.71417034e-01 -8.25134695e-01 5.82269073e-01 4.44266051e-01
1.20469892e+00 -8.97695303e-01 3.83769959e-01 -3.81474078e-01
-9.42618325e-02 -1.34801328e+00 -2.47399777e-01 -5.32993853e-01
-3.04888666e-01 -1.22605813e+00 -1.05862963e+00 -9.76831138e-01
4.42801148e-01 5.34623146e-01 9.25958753e-01 1.92312971e-02
-6.96280122e-01 4.86598641e-01 -1.95673004e-01 -2.75168568e-01
-3.20950925e-01 -3.30049038e-01 2.69010141e-02 4.47226822e-01
8.15340221e-01 -2.64284670e-01 -6.42955422e-01 5.64253211e-01
-5.78137636e-01 -3.75194281e-01 5.46148837e-01 1.01322126e+00
3.51997823e-01 -2.53984611e-02 8.80985320e-01 -8.29248726e-01
2.81587392e-01 -3.87574434e-01 -7.52730250e-01 6.92280352e-01
-6.02580845e-01 -3.02071035e-01 3.08037400e-01 -2.77454913e-01
-1.47034442e+00 3.40366811e-02 -1.90649509e-01 -6.60320699e-01
-2.55091459e-01 4.11451161e-02 -4.29858536e-01 -6.71949685e-01
6.78843081e-01 2.03700438e-01 2.22807258e-01 -6.11174762e-01
7.19446242e-02 1.22185922e+00 6.09138668e-01 -5.74373007e-01
1.05421937e+00 4.68871415e-01 2.63950646e-01 -9.01419044e-01
-7.20863163e-01 -5.98170936e-01 -4.48937237e-01 -2.96322882e-01
5.96370816e-01 -1.40834308e+00 -1.00458646e+00 8.45071971e-01
-7.54679024e-01 1.17034435e-01 1.30531743e-01 5.89257896e-01
-2.91790694e-01 3.89756650e-01 -2.93748558e-01 -8.77607286e-01
-4.05548751e-01 -1.16564214e+00 1.30085015e+00 3.50944787e-01
5.01494467e-01 -5.36623538e-01 4.28141765e-02 8.31103981e-01
5.33869267e-01 2.52870142e-01 5.48064590e-01 -4.45583254e-01
-4.69164521e-01 -1.88310668e-01 -6.65514469e-01 6.41917288e-01
3.70725185e-01 -1.30503207e-01 -1.40542269e+00 -7.88440228e-01
-3.16806883e-01 -5.11672616e-01 1.11713648e+00 7.81461298e-02
1.29849541e+00 1.96958154e-01 -3.11279148e-01 7.64487147e-01
1.64604759e+00 2.36215964e-01 8.85927439e-01 1.54176176e-01
6.45112038e-01 6.11313224e-01 7.42336154e-01 4.60290164e-01
1.11144103e-01 9.03183162e-01 2.19107047e-01 -1.23374775e-01
-8.00071061e-01 -1.54194692e-02 3.93742085e-01 2.31561318e-01
-2.79306859e-01 7.51312301e-02 -6.69310749e-01 4.68821526e-01
-1.72041154e+00 -1.07161963e+00 1.22482114e-01 2.32650161e+00
6.10710919e-01 -6.73926711e-01 9.50130820e-02 1.95785537e-01
8.76708388e-01 1.19321018e-01 -4.40439463e-01 1.15733936e-01
-4.98042792e-01 4.02609348e-01 3.11435908e-01 1.83031261e-01
-9.59793508e-01 9.30804431e-01 5.88586473e+00 1.08461785e+00
-1.23988581e+00 3.48430365e-01 4.00354147e-01 1.56356484e-01
1.48465231e-01 -4.46993738e-01 -1.24226439e+00 2.60844558e-01
5.91206551e-01 1.96354464e-01 5.58883011e-01 7.47878134e-01
-1.68895081e-01 2.30923608e-01 -1.08316600e+00 1.78371000e+00
6.31813288e-01 -1.24417460e+00 -1.27574997e-02 -6.71015028e-03
1.09312117e+00 -1.07305333e-01 3.27747226e-01 5.01801610e-01
1.44667202e-03 -1.36683023e+00 1.86534062e-01 4.15917188e-01
1.22382820e+00 -8.04155946e-01 8.04216087e-01 -1.31696805e-01
-1.28719854e+00 -3.16139013e-01 -6.81786299e-01 4.54024106e-01
-4.17520881e-01 4.96679276e-01 -7.23289907e-01 8.67280960e-01
8.71354997e-01 8.34713459e-01 -5.46741009e-01 1.05540168e+00
4.71335948e-02 3.51198137e-01 -1.35322791e-02 4.87014890e-01
-4.48187023e-01 -1.41459286e-01 2.81760514e-01 9.93899167e-01
4.60473180e-01 -1.89818010e-01 3.75542380e-02 5.55656552e-01
-5.58485210e-01 1.58953682e-01 -6.83081746e-01 -8.06343704e-02
4.17474151e-01 1.26171494e+00 9.43785533e-02 -1.06011741e-01
-6.83296442e-01 8.21603298e-01 1.63389489e-01 3.87761980e-01
-8.26672018e-01 -1.46867707e-01 8.72648001e-01 -1.35895506e-01
3.22107643e-01 2.43416913e-02 1.75454438e-01 -1.13490820e+00
1.01035438e-01 -1.21832144e+00 6.54060483e-01 -6.33086860e-01
-1.60649896e+00 6.89662814e-01 -2.97535866e-01 -1.24625075e+00
1.43011466e-01 -7.72373617e-01 -1.49650902e-01 1.07098579e+00
-2.28229618e+00 -1.69302106e+00 -7.85364747e-01 1.22438312e+00
3.81383955e-01 -1.02907372e+00 7.44957566e-01 8.11920047e-01
-6.25797808e-01 1.27184117e+00 3.05843949e-01 1.73760951e-01
1.25291407e+00 -6.61586642e-01 -3.14338893e-01 6.56658113e-01
7.20117986e-02 3.31397474e-01 3.53531033e-01 -5.59634209e-01
-1.72741544e+00 -1.09015203e+00 5.74418843e-01 -2.44259939e-01
6.39484525e-02 -2.07554907e-01 -7.96765268e-01 4.64871079e-01
1.41174689e-01 4.55641687e-01 9.27233040e-01 1.34754181e-01
-1.10799086e+00 -7.48286545e-01 -1.76143372e+00 1.39958143e-01
1.20160151e+00 -8.60250294e-01 -4.86433327e-01 4.86946583e-01
2.95188606e-01 -9.46668833e-02 -1.01848495e+00 8.81187022e-01
7.92074561e-01 -1.06172931e+00 1.18584311e+00 -4.24245656e-01
8.36225078e-02 -4.09568578e-01 -6.48982465e-01 -1.09350443e+00
-2.25155145e-01 -6.04642451e-01 -4.21899706e-02 1.65495884e+00
-5.68499453e-02 -8.63543749e-01 6.86947644e-01 1.43325642e-01
2.40443960e-01 -2.35433817e-01 -9.58909273e-01 -1.01720154e+00
-1.60995185e-01 -5.92643395e-02 9.32606518e-01 1.12139690e+00
-7.28006303e-01 5.59350513e-02 -5.99101841e-01 4.10409510e-01
1.05407941e+00 4.24962133e-01 9.21437681e-01 -1.44916332e+00
-2.50668406e-01 -1.71541125e-02 -4.74551022e-01 -4.68894839e-01
6.12672806e-01 -9.80436683e-01 -2.36404032e-01 -1.26253736e+00
6.03010535e-01 -3.92556250e-01 -4.91126209e-01 6.29756868e-01
-8.54773505e-04 8.25065553e-01 1.54488698e-01 2.89263725e-01
-3.29188764e-01 6.77028894e-01 1.34282351e+00 -2.27494955e-01
1.43841460e-01 -1.75851122e-01 -6.00725114e-01 2.91535318e-01
5.14501810e-01 -6.68526292e-02 -1.51304662e-01 -2.65946865e-01
-3.42909366e-01 -2.05629226e-02 2.29293287e-01 -1.16019809e+00
8.64362046e-02 -7.14053959e-02 8.21241498e-01 -6.43012166e-01
5.14017582e-01 -8.46727312e-01 2.82953739e-01 2.30964348e-01
1.15280114e-02 -3.42632860e-01 2.83878356e-01 3.93137634e-01
-3.66686374e-01 2.69735426e-01 1.30667377e+00 2.35178798e-01
-9.11516011e-01 6.97910190e-01 3.23060542e-01 -3.27601731e-01
1.10097492e+00 -2.54295379e-01 -6.06776536e-01 -6.36942536e-02
-3.66319031e-01 -7.19768405e-02 1.47319496e-01 8.15557539e-01
6.53996587e-01 -1.49875832e+00 -1.11210191e+00 6.76547647e-01
4.13912475e-01 -5.35014808e-01 6.66791618e-01 7.64550090e-01
-1.69485465e-01 4.53045428e-01 -5.15684009e-01 -5.29962003e-01
-1.94945228e+00 3.65126371e-01 3.99803311e-01 1.83671936e-02
-4.47074354e-01 9.40387905e-01 2.36757040e-01 -6.11305654e-01
2.74590135e-01 6.91133618e-01 -4.33576435e-01 1.40003070e-01
7.59251833e-01 3.92216712e-01 4.00342762e-01 -1.08949649e+00
-5.72515488e-01 9.45789278e-01 -7.32345730e-02 4.31843787e-01
1.44665325e+00 -1.31821170e-01 -2.03157663e-01 -4.03170258e-01
1.60101497e+00 6.15290254e-02 -1.03104031e+00 -4.03651625e-01
-4.91424352e-01 -1.06639016e+00 9.63068232e-02 -1.01716769e+00
-1.36225557e+00 8.52960765e-01 1.26652002e+00 -4.33321595e-01
1.30080271e+00 -6.71645477e-02 7.49863684e-01 2.26307377e-01
3.95582944e-01 -9.07724917e-01 2.70495504e-01 2.41173714e-01
9.70425785e-01 -1.66799748e+00 -1.72581952e-02 -5.87668240e-01
-2.58657753e-01 1.16647577e+00 8.07788789e-01 2.63719827e-01
8.29480529e-01 1.37923528e-02 1.43417120e-01 -2.46844124e-02
-2.84709722e-01 -3.65500867e-01 3.79438430e-01 8.80933821e-01
3.23986024e-01 -9.82164964e-02 -2.04696432e-01 1.79929450e-01
1.10563925e-02 2.05665007e-01 -6.31632134e-02 6.58172786e-01
-1.23765230e-01 -1.32801521e+00 -5.71377635e-01 2.44235441e-01
-6.17345333e-01 2.67296880e-01 -2.00812235e-01 7.98457086e-01
1.73110425e-01 1.04953945e+00 -3.52289557e-01 -3.93914044e-01
1.83454350e-01 7.67076463e-02 9.03696477e-01 -2.44719654e-01
-3.95406723e-01 -1.30589619e-01 -8.64835680e-02 -5.45641124e-01
-6.07864976e-01 -4.42066282e-01 -7.21384108e-01 -4.30257618e-01
-3.20450455e-01 -4.41908613e-02 8.08669627e-01 9.65705812e-01
4.09078747e-01 2.97377318e-01 1.03220046e+00 -4.51722264e-01
-6.52621746e-01 -8.49327922e-01 -8.09005380e-01 7.25679040e-01
3.95178348e-01 -1.01380587e+00 -2.74816155e-01 -1.98219880e-01] | [13.160563468933105, 0.5355408191680908] |
e15ea447-3d17-44b9-8ccf-71e9c54f2c88 | pymicetracking-an-open-source-toolbox-for | null | null | http://openaccess.thecvf.com//content/CVPR2022/html/Menezes_PyMiceTracking_An_Open-Source_Toolbox_for_Real-Time_Behavioral_Neuroscience_Experiments_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Menezes_PyMiceTracking_An_Open-Source_Toolbox_for_Real-Time_Behavioral_Neuroscience_Experiments_CVPR_2022_paper.pdf | PyMiceTracking: An Open-Source Toolbox for Real-Time Behavioral Neuroscience Experiments | The development of computational tools allows the advancement of research in behavioral neuroscience and elevates the limits of experiment design. Many behavioral experiments need to determine the animal's position from its tracking, which is crucial for real-time decision-making and further analysis of experimental data. Modern experimental designs usually generate the recording of a large amount of data, requiring the development of automatic computational tools and intelligent algorithms for timely data acquisition and processing. The proposed tool in this study initially operates with the acquisition of images. Then the animal tracking step begins with background subtraction, followed by the animal contour detection and morphological operations to remove noise in the detected shapes. Finally, in the final stage of the algorithm, the principal components analysis (PCA) is applied in the obtained shape, resulting in the animal's gaze direction. | ['Helton Maia', 'Aron de Miranda', 'Richardson Menezes'] | 2022-01-01 | null | null | null | cvpr-2022-1 | ['contour-detection'] | ['computer-vision'] | [ 4.95824814e-01 -6.95933342e-01 2.86146462e-01 -2.54854679e-01
4.15600091e-01 -4.75413352e-01 1.39399469e-01 5.65348923e-01
-9.61708128e-01 4.70227152e-01 -5.61740756e-01 -3.35659087e-01
-2.56515015e-02 -4.60641712e-01 -1.99133024e-01 -8.63160014e-01
-7.91379064e-03 3.29000279e-02 3.49547088e-01 2.43571356e-01
6.86586380e-01 9.68679130e-01 -1.76395738e+00 -3.59792590e-01
5.75482965e-01 8.39155316e-01 3.79365474e-01 7.13468432e-01
-1.68061271e-01 2.52851099e-01 -4.41968560e-01 -2.76225079e-02
1.86054438e-01 -6.56534791e-01 -1.02159142e-01 1.87380686e-01
-4.79334556e-02 -2.96455830e-01 4.38081294e-01 1.27593005e+00
3.59198034e-01 2.65405297e-01 4.67760891e-01 -9.47530806e-01
-2.11060137e-01 -1.14747137e-01 -7.20587671e-01 6.02524281e-01
2.51092583e-01 2.26778284e-01 2.22329438e-01 -7.47187018e-01
4.68197167e-01 9.28448796e-01 2.61229962e-01 3.33563924e-01
-1.42500746e+00 -6.56661868e-01 -3.48746419e-01 2.99150646e-01
-1.38037193e+00 -4.40312684e-01 9.07287598e-01 -7.45174110e-01
1.99739024e-01 8.01389739e-02 8.31148148e-01 5.95072925e-01
2.77774274e-01 4.05621603e-02 1.39771128e+00 -5.97062469e-01
2.91517138e-01 1.95674419e-01 4.51625943e-01 1.82577595e-01
7.39360273e-01 1.15260638e-01 -6.20275289e-02 1.51226595e-01
8.62737596e-01 6.79505765e-02 -2.34349936e-01 -3.85607183e-01
-8.65131259e-01 4.65514034e-01 -2.55285278e-02 7.68898845e-01
-6.92983568e-01 -4.47836965e-01 1.68753177e-01 -2.87678465e-02
4.48803641e-02 2.60813177e-01 2.75687911e-02 2.69218348e-02
-1.24294138e+00 1.99606031e-01 2.16697887e-01 4.08377498e-01
3.78578484e-01 -5.97643405e-02 7.66755128e-03 2.74287164e-01
4.90547597e-01 4.19429809e-01 7.21764445e-01 -5.84015906e-01
-2.49402858e-02 9.95169878e-01 1.96453825e-01 -1.06999457e+00
-5.99937320e-01 -2.88313955e-01 -4.75141346e-01 7.70807087e-01
6.93684220e-01 -3.03808421e-01 -7.59697795e-01 1.20057225e+00
8.06424141e-01 -2.50468612e-01 -1.66962042e-01 6.99913502e-01
3.53864461e-01 4.52606529e-01 3.75582069e-01 -5.52504718e-01
1.51420152e+00 -5.68402633e-02 -9.58801091e-01 -1.01870507e-01
2.58197039e-01 -8.52449238e-01 7.82693923e-01 4.10083562e-01
-6.85523570e-01 -6.55932963e-01 -1.02940285e+00 1.32161722e-01
-4.83968705e-01 6.03557289e-01 3.53398800e-01 4.33351308e-01
-7.83419311e-01 4.70554084e-01 -8.01240325e-01 -5.32056272e-01
4.04912710e-01 3.92734349e-01 -4.59678829e-01 3.73317689e-01
-2.91648448e-01 9.18786108e-01 5.16335487e-01 7.03039348e-01
-2.69158632e-01 -2.64855683e-01 -5.43196678e-01 -2.71288510e-02
6.76448271e-02 -2.64541090e-01 9.29893732e-01 -1.01695228e+00
-1.35729468e+00 1.05602860e+00 -1.77268609e-01 -3.61013293e-01
4.54012811e-01 -1.63345307e-01 -2.41939634e-01 2.62824595e-01
-8.90777260e-02 3.99352551e-01 1.06209600e+00 -9.98714685e-01
-7.00714231e-01 -1.00911784e+00 -4.46056098e-01 1.21787809e-01
6.03942163e-02 4.07722622e-01 -1.35528192e-01 -3.34237903e-01
2.09707886e-01 -7.31558502e-01 -1.85023770e-01 1.18573405e-01
1.30581945e-01 -6.97886348e-02 9.66892123e-01 -7.54805088e-01
1.23057377e+00 -2.48828506e+00 7.38166571e-02 1.47069722e-01
2.07974732e-01 3.61673683e-01 2.69912422e-01 8.50093663e-02
-2.61706054e-01 -4.58142489e-01 -6.78512827e-02 -1.00344449e-01
-4.25865322e-01 -2.81453222e-01 9.16208997e-02 7.10947931e-01
9.28285643e-02 1.39069170e-01 -3.77048582e-01 -7.09524035e-01
4.01618779e-01 2.19806999e-01 -3.31306040e-01 3.93308163e-01
8.14825520e-02 7.46722877e-01 -4.79339212e-01 3.40341747e-01
5.98616123e-01 2.68107861e-01 9.75695327e-02 -1.48549482e-01
-6.89496040e-01 -2.84558326e-01 -1.00552201e+00 9.10862923e-01
9.67533514e-02 7.11398542e-01 2.75869012e-01 -1.05501330e+00
9.68299508e-01 1.99774042e-01 6.11523807e-01 -6.81026459e-01
7.38509059e-01 9.86225232e-02 4.64538634e-01 -7.18356490e-01
2.65173137e-01 1.86253354e-01 5.44921339e-01 2.47860268e-01
-2.87232071e-01 1.92583352e-01 5.52797854e-01 -2.91487604e-01
3.63312185e-01 3.00965309e-01 5.50820649e-01 -2.54314393e-01
5.93607903e-01 -5.70282526e-03 3.40783656e-01 -1.17397189e-01
-4.58089590e-01 -1.74900249e-01 4.10298556e-01 -4.72320318e-01
-9.33171630e-01 -4.00551647e-01 -3.44140679e-01 6.19567215e-01
9.06520560e-02 4.04072076e-01 -1.10744536e+00 -7.43807182e-02
-1.30974889e-01 5.77367008e-01 -5.43936670e-01 1.75008029e-02
-4.34497416e-01 -7.77286351e-01 -1.58540145e-01 1.51438592e-02
4.17489648e-01 -1.49647963e+00 -1.63246787e+00 3.03831816e-01
3.15947354e-01 -8.40633154e-01 -5.31065697e-03 1.31145224e-01
-1.26415896e+00 -1.15190434e+00 -5.62619865e-01 -6.44130230e-01
1.07367015e+00 3.73476416e-01 9.87744927e-02 1.94399610e-01
-6.42371714e-01 -7.13843554e-02 -3.46523315e-01 -5.07009625e-01
-2.04097345e-01 -3.79758745e-01 4.94232774e-02 6.68551251e-02
6.02590978e-01 -5.75344563e-01 -4.68076915e-01 1.48918271e-01
-8.21042657e-01 -8.84831697e-03 6.88314557e-01 1.21919416e-01
5.65429389e-01 3.02003294e-01 3.67959619e-01 -3.30478400e-01
5.85341036e-01 -1.42366201e-01 -1.27918577e+00 9.30187628e-02
-3.76961499e-01 -1.99213717e-02 7.28202105e-01 -4.61194247e-01
-8.12187254e-01 1.08400039e-01 1.06543407e-01 -2.84625664e-02
-7.51800656e-01 3.65422547e-01 -1.08074874e-01 -2.59290695e-01
5.09296179e-01 2.82410622e-01 2.40090370e-01 -5.15255868e-01
-1.84048459e-01 5.64300358e-01 5.21705687e-01 1.98854506e-01
5.88202536e-01 2.45849460e-01 1.44740984e-01 -1.12889349e+00
-3.22694093e-01 -3.80800486e-01 -1.05036581e+00 -6.10949993e-01
1.14375854e+00 -6.54414296e-02 -9.37160254e-01 6.94382787e-01
-1.06784582e+00 6.45420402e-02 2.62810677e-01 9.59155321e-01
-1.49895877e-01 2.32830599e-01 -6.73076212e-02 -1.10039890e+00
-4.60886985e-01 -1.10797095e+00 3.57895553e-01 7.92037308e-01
-2.65673399e-01 -5.12480438e-01 -2.07213257e-02 2.63936967e-01
9.37090889e-02 -3.31719331e-02 8.60113025e-01 -4.81520802e-01
-5.63753188e-01 -5.84770322e-01 -7.13228136e-02 1.99167401e-01
1.04122922e-01 5.63493252e-01 -7.51837492e-01 -2.45191529e-02
4.47415084e-01 3.24399263e-01 1.01964988e-01 8.94507647e-01
7.70897925e-01 8.15769136e-02 -2.89655268e-01 3.76203448e-01
1.25667036e+00 1.03474355e+00 4.82515126e-01 2.41819561e-01
2.50572622e-01 8.60351562e-01 8.17773461e-01 3.58075500e-01
-1.06224105e-01 6.09209359e-01 3.68422985e-01 1.53148929e-02
1.57758266e-01 5.69134951e-02 1.44008830e-01 5.07568717e-01
-2.92018205e-01 1.69333979e-01 -5.32693326e-01 3.69082481e-01
-1.47010660e+00 -9.89517093e-01 -4.45419222e-01 2.57371211e+00
4.52100873e-01 4.18227166e-02 3.31391007e-01 4.73028570e-01
8.43394578e-01 -3.73344868e-01 -3.15306842e-01 -3.61533552e-01
3.11830431e-01 5.34109510e-02 4.48406994e-01 2.98266739e-01
-9.27149594e-01 5.98685622e-01 6.28637362e+00 2.29849041e-01
-1.64027750e+00 -5.91534436e-01 7.59763941e-02 1.17318444e-01
3.05686682e-01 -3.04008573e-01 -9.05671895e-01 5.83090067e-01
5.68710148e-01 -3.84755939e-01 4.53369647e-01 8.27355146e-01
7.09644556e-01 -9.65789020e-01 -6.32888198e-01 8.80560577e-01
-7.54613429e-02 -4.97822255e-01 -2.75252372e-01 1.41215831e-01
2.32188366e-02 -6.37218475e-01 -6.59744209e-03 -3.10536683e-01
-3.08343768e-01 -5.21852076e-01 6.82095945e-01 4.95752186e-01
2.82088250e-01 -5.97030520e-01 5.41695654e-01 4.49543059e-01
-8.76604378e-01 -1.17412269e-01 -3.25609177e-01 -1.80354863e-01
-9.84872878e-02 3.18338096e-01 -9.93151009e-01 9.21891928e-02
4.86529380e-01 2.24343836e-01 -6.50736213e-01 1.56705701e+00
-5.46267852e-02 4.74389017e-01 -2.87527442e-01 -2.45833158e-01
-1.18877053e-01 -7.77087212e-01 7.17066586e-01 6.89015090e-01
4.28620487e-01 1.92951977e-01 -2.21893266e-01 8.39674115e-01
3.56364578e-01 5.35387695e-01 -4.28309947e-01 -3.26487690e-01
3.52724433e-01 1.41723955e+00 -1.44079340e+00 -6.84559792e-02
-3.03916156e-01 6.20618761e-01 8.80996957e-02 2.75693327e-01
-5.99902332e-01 -3.99288356e-01 3.12933594e-01 4.32321429e-01
2.80905455e-01 -5.76495767e-01 -4.10580754e-01 -6.19793057e-01
1.20692827e-01 -6.74154878e-01 1.56520084e-01 -5.72425902e-01
-5.09346306e-01 3.40044826e-01 1.08531535e-01 -9.74538684e-01
-1.21241599e-01 -6.16613269e-01 -6.92359090e-01 9.60529506e-01
-9.79378104e-01 -5.37580669e-01 -1.47983700e-01 4.33086425e-01
3.85615617e-01 1.31235942e-01 7.43435442e-01 4.17230457e-01
-8.77826273e-01 -1.05645649e-01 -3.40301506e-02 8.24202821e-02
4.79720414e-01 -7.71847725e-01 -2.24070668e-01 1.16003442e+00
-1.22271381e-01 5.57314157e-01 1.04521036e+00 -7.80441642e-01
-1.01738727e+00 -2.49734849e-01 7.50050128e-01 1.24451883e-01
6.06175125e-01 -1.92233250e-01 -8.69504571e-01 4.35976297e-01
-6.40362129e-02 -2.48470992e-01 6.44141138e-01 -2.50661910e-01
5.25924981e-01 -2.02019736e-01 -1.11937523e+00 8.07966411e-01
1.96446225e-01 -6.67378902e-02 -7.40667105e-01 -3.22598934e-01
-6.02953173e-02 -2.05713883e-02 -4.79361355e-01 4.70447876e-02
8.49713445e-01 -9.13427055e-01 5.39139390e-01 -3.99053723e-01
-5.39987758e-02 -6.19633436e-01 5.53212702e-01 -8.51428211e-01
-2.68519998e-01 -1.59805149e-01 6.60307109e-01 1.09140241e+00
7.15149119e-02 -3.48675907e-01 5.81152499e-01 7.22838104e-01
2.03062117e-01 -1.06300578e-01 -6.59132957e-01 -1.53066978e-01
-6.74792469e-01 -1.85828075e-01 6.06579445e-02 5.17347813e-01
-1.99894235e-02 7.38919750e-02 7.89690688e-02 2.39862695e-01
1.02701795e+00 2.98368126e-01 6.91722393e-01 -1.48337793e+00
3.88769627e-01 -5.01940787e-01 -7.55820990e-01 -4.03055042e-01
-3.35127562e-01 -2.27788702e-01 3.34043391e-02 -1.19386303e+00
-2.00142026e-01 1.22776695e-01 2.71245791e-03 9.73458588e-02
-2.59309143e-01 -1.18766604e-02 -2.80813016e-02 6.19991794e-02
-1.12947039e-01 1.89885139e-01 1.35985374e+00 5.31941116e-01
-5.69110572e-01 4.24464256e-01 -2.91379213e-01 9.07972455e-01
7.40499258e-01 -5.85397482e-01 -3.61714840e-01 -2.36228984e-02
-4.03434932e-02 -5.13330027e-02 4.87205684e-01 -1.28632367e+00
2.88817793e-01 -1.15534931e-01 6.63638294e-01 -8.29297841e-01
2.91863028e-02 -1.32667053e+00 2.59781480e-01 6.55735731e-01
1.87233835e-02 1.65746376e-01 3.39913368e-01 1.58752352e-01
3.52943949e-02 -6.02083027e-01 1.09607649e+00 2.41573259e-01
-6.16095245e-01 -4.60586436e-02 -7.50485182e-01 -4.15678978e-01
1.64230144e+00 -6.22335434e-01 -1.67613532e-02 1.70627624e-01
-8.58450592e-01 -9.68121663e-02 5.00590146e-01 -4.44602519e-02
3.18025887e-01 -9.10013318e-01 -2.63543725e-01 5.42593598e-01
-3.50629896e-01 -1.41302168e-01 5.93152225e-01 1.31912529e+00
-1.00136459e+00 3.09619278e-01 -9.10735428e-01 -3.35547447e-01
-1.85617387e+00 8.67006481e-01 -1.83919573e-03 5.31196222e-02
-4.44410712e-01 3.23246866e-01 -2.84590106e-02 6.77731574e-01
1.23671576e-01 -5.49825251e-01 -1.06771553e+00 4.54994947e-01
6.52140856e-01 4.70768422e-01 -1.37654424e-01 -7.72818983e-01
-2.49473572e-01 7.52119958e-01 1.41190037e-01 -1.57090008e-01
1.19163179e+00 -2.42884994e-01 -2.59277374e-01 5.57233632e-01
7.07083464e-01 1.67572275e-01 -1.08921933e+00 2.61831105e-01
1.06218278e-01 -5.20188212e-01 1.66917056e-01 -5.00125408e-01
-6.95673943e-01 1.04785872e+00 7.52834857e-01 4.01990712e-01
1.13253260e+00 -6.40193701e-01 2.16191173e-01 9.90625322e-02
2.59697795e-01 -1.06109226e+00 -5.36130905e-01 -8.33847299e-02
6.35973632e-01 -1.03300071e+00 6.52371272e-02 -1.36999086e-01
-4.26725686e-01 1.17831874e+00 3.32632452e-01 -7.73162860e-03
5.94556272e-01 4.24206227e-01 1.65166572e-01 -2.78219908e-01
-2.68659770e-01 -3.39089185e-01 4.42860276e-01 4.38243598e-01
3.42152894e-01 -5.39896600e-02 -8.62294734e-01 6.16138041e-01
-1.04067095e-01 4.78601336e-01 3.01354885e-01 1.03864813e+00
-9.45608974e-01 -8.27025771e-01 -8.37266028e-01 2.56441236e-01
-7.11718261e-01 9.21685249e-02 -3.99264187e-01 6.20817363e-01
1.82421073e-01 7.60875523e-01 1.84144869e-01 1.61940247e-01
2.83461690e-01 1.49518415e-01 5.69705546e-01 -3.68281096e-01
-2.22205877e-01 3.85150850e-01 -4.29507583e-01 -1.40498936e-01
-5.43227673e-01 -9.17721093e-01 -1.36908734e+00 -5.58268577e-02
-2.21298173e-01 2.84684807e-01 1.19331956e+00 8.82977486e-01
1.10834166e-01 2.93759853e-01 4.61509854e-01 -7.96874404e-01
-8.45395401e-02 -9.21023726e-01 -5.99990666e-01 3.16599011e-01
9.49769914e-02 -7.66278863e-01 -4.13568854e-01 3.45096320e-01] | [12.935050964355469, 0.2744884490966797] |
cf183ad3-58ce-4510-8bce-1a722601c816 | a-unified-approach-to-discourse-relation | null | null | https://aclanthology.org/2021.disrpt-1.5 | https://aclanthology.org/2021.disrpt-1.5.pdf | A Unified Approach to Discourse Relation Classification in nine Languages | This paper presents efforts to solve the shared task on discourse relation classification (disrpt task 3). The intricate prediction task aims to predict a large number of classes from the Rhetorical Structure Theory (RST) framework for nine target languages. Labels include discourse relations such as background, condition, contrast and elaboration. We present an approach using euclidean distance between sentence embeddings that were extracted using multlingual sentence BERT (sBERT) and directionality as features. The data was combined into five classes which were used for initial prediction. The second classification step predicts the target classes. We observe a substantial difference in results depending on the number of occurrences of the target label in the training data. We achieve the best results on Chinese, where our system achieves 70 % accuracy on 20 labels. | ['Franziska Pannach', 'Hanna Varachkina'] | null | null | null | null | emnlp-disrpt-2021-11 | ['relation-classification'] | ['natural-language-processing'] | [ 2.43445233e-01 6.69594705e-01 -5.07033646e-01 -4.67758745e-01
-7.45257020e-01 -6.69725955e-01 1.30007112e+00 6.93794906e-01
-4.68009114e-01 7.53506780e-01 1.04136479e+00 -8.22433829e-01
3.51719595e-02 -3.67909342e-01 -1.95551485e-01 -3.94479573e-01
-6.45390525e-02 5.42248905e-01 2.01132759e-01 -7.76769161e-01
5.38989305e-01 1.28839105e-01 -1.09101510e+00 9.57857132e-01
6.19595587e-01 8.51224303e-01 2.11932346e-01 9.26650524e-01
-3.32358837e-01 1.69199538e+00 -9.60123003e-01 -4.20718580e-01
-4.01241153e-01 -5.01971066e-01 -1.54550672e+00 -2.53918648e-01
3.56942296e-01 1.40944853e-01 -1.26022413e-01 4.79976356e-01
3.55636984e-01 -5.35077937e-02 1.08384693e+00 -7.56796420e-01
-8.21136773e-01 1.04177022e+00 -2.03652099e-01 7.05214858e-01
7.72037566e-01 -4.41173732e-01 1.63309431e+00 -1.00458109e+00
1.10303938e+00 1.51049447e+00 5.34677029e-01 3.28343242e-01
-1.16888595e+00 -9.62641314e-02 9.42571741e-03 6.46973610e-01
-7.61918843e-01 -4.48399425e-01 7.88554132e-01 -7.87546396e-01
1.34772563e+00 6.07695162e-01 3.37807834e-01 1.24392796e+00
3.05727273e-01 5.00220597e-01 1.27914143e+00 -8.44837427e-01
-1.46975353e-01 4.12777483e-01 6.13815486e-01 3.91810685e-01
-4.28430974e-01 -2.77692944e-01 -5.02037764e-01 -1.82295591e-01
-1.98863428e-02 -9.04924810e-01 -2.75253981e-01 3.73114765e-01
-1.34495211e+00 1.19799304e+00 5.55043399e-01 9.02822852e-01
-2.07397938e-02 -4.59153622e-01 6.51174009e-01 6.00934446e-01
9.07662988e-01 8.53039503e-01 -5.96642375e-01 -1.55797347e-01
-2.17916176e-01 3.35536420e-01 1.17195868e+00 5.48463047e-01
3.48263592e-01 -3.66982609e-01 -4.90221530e-01 1.16651177e+00
-3.05132344e-02 -9.58228856e-02 6.80027604e-01 -8.01331043e-01
9.48030114e-01 6.90631092e-01 -1.24547765e-01 -1.20949829e+00
-5.80493033e-01 -1.42719254e-01 -2.56885678e-01 -3.04615229e-01
2.80443579e-01 -3.18916589e-01 -7.16173649e-02 1.48704529e+00
1.69869319e-01 -4.96341854e-01 4.67341423e-01 7.01257229e-01
1.24208200e+00 8.64595354e-01 1.48853779e-01 -3.85877907e-01
1.41739881e+00 -1.17500961e+00 -9.38285351e-01 -1.87586755e-01
1.48310471e+00 -1.12895763e+00 9.43343937e-01 -5.37253395e-02
-8.18434238e-01 -4.78828609e-01 -1.22704804e+00 -5.05498648e-01
-3.90671611e-01 1.26579404e-01 4.20367032e-01 2.72772342e-01
-7.83956647e-01 4.43533391e-01 -2.92637318e-01 -8.52346867e-02
1.58353642e-01 -2.33789477e-02 -2.60463625e-01 3.55782330e-01
-1.18955934e+00 1.77109790e+00 5.44622958e-01 -3.84138197e-01
-2.00558186e-01 -5.72772145e-01 -9.98505116e-01 -1.11455292e-01
-3.43603417e-02 2.72164531e-02 1.27060521e+00 -1.03068209e+00
-1.45947826e+00 1.49454200e+00 -1.42194942e-01 -5.34972012e-01
1.72978267e-01 -6.48601949e-02 -4.68160659e-01 -8.52555260e-02
1.86284319e-01 2.58748025e-01 3.31664264e-01 -1.12209928e+00
-5.94138205e-01 1.83103643e-02 1.90026820e-01 4.24856246e-01
-4.29540902e-01 6.82347715e-01 6.94772542e-01 -5.63486457e-01
6.38823658e-02 -6.67767704e-01 2.41396368e-01 -6.02724314e-01
-5.14049470e-01 -1.03874946e+00 8.46461236e-01 -1.02503967e+00
1.42866635e+00 -1.94212246e+00 5.87142467e-01 -4.76778448e-01
2.31023952e-01 1.38462454e-01 -9.37176198e-02 6.51493430e-01
-4.67355162e-01 2.23098144e-01 4.36780006e-02 -1.74034178e-01
-1.34765524e-02 1.41121268e-01 -3.94207656e-01 2.69028187e-01
4.84502852e-01 8.33623588e-01 -7.63316572e-01 -6.46928787e-01
-4.34271358e-02 9.09326449e-02 -2.88614064e-01 3.43257278e-01
-3.50031972e-01 3.70724291e-01 -2.25656822e-01 1.26502559e-01
2.93875262e-02 -1.18729360e-01 7.80318081e-01 -1.29117802e-01
-3.03652704e-01 1.35049212e+00 -4.43179220e-01 1.08673060e+00
-5.41963518e-01 1.36974943e+00 -1.72891140e-01 -1.11782384e+00
1.13945425e+00 5.54787040e-01 1.12839721e-01 -6.71430826e-01
3.29000294e-01 1.82000980e-01 7.43420839e-01 -7.71080911e-01
7.06182957e-01 -1.47954360e-01 -4.33654100e-01 5.03960133e-01
6.57991692e-02 1.32501582e-02 2.70235270e-01 7.41362646e-02
1.02956164e+00 -1.47739172e-01 7.05672860e-01 -6.43719614e-01
7.78987825e-01 2.89447606e-01 3.28948915e-01 -8.88623204e-03
-4.46461402e-02 2.60831177e-01 1.19713318e+00 -5.68882644e-01
-9.07147586e-01 -8.06615412e-01 -4.64365810e-01 1.55431187e+00
-2.04737052e-01 -6.73782051e-01 -3.99984360e-01 -8.67820859e-01
-1.81181595e-01 1.13076043e+00 -7.99619257e-01 3.13745141e-01
-1.14284003e+00 -5.72853327e-01 4.44599688e-01 4.08668667e-01
8.31959769e-02 -1.12540340e+00 -5.31973302e-01 9.18657631e-02
-4.56141412e-01 -1.15049970e+00 1.33603826e-01 2.97728240e-01
-4.11842972e-01 -1.35300124e+00 -7.03653181e-03 -1.34450448e+00
2.75288165e-01 -1.72261015e-01 1.13276398e+00 1.56236425e-01
7.93373287e-02 -1.17978521e-01 -5.35718799e-01 -2.76707917e-01
-8.40419412e-01 2.33921990e-01 -2.76357472e-01 -5.57225287e-01
1.90363541e-01 -1.50234953e-01 1.36474997e-01 -2.69698113e-01
-3.08953583e-01 1.67240426e-01 -1.35118023e-01 1.01236677e+00
3.46195288e-02 -4.50270146e-01 5.42497098e-01 -1.11364055e+00
1.16792119e+00 -6.01668000e-01 3.09216455e-02 3.66499066e-01
-1.67733803e-01 1.87198222e-02 4.77949321e-01 -4.12655115e-01
-1.09080863e+00 -8.16447675e-01 -2.82678753e-01 5.18291295e-01
1.35946110e-01 7.48415232e-01 1.60804898e-01 5.75419247e-01
1.00234675e+00 -4.83224809e-01 2.65568718e-02 -3.80921811e-01
4.68987197e-01 9.73376095e-01 7.84730092e-02 -5.11121988e-01
5.07230684e-02 -3.43312860e-01 -2.43570402e-01 -9.60293233e-01
-1.29554367e+00 -1.51909113e-01 -8.31687331e-01 -1.49859339e-01
1.17797112e+00 -6.66038394e-01 -3.82550210e-01 -2.43220389e-01
-1.65400672e+00 -3.33170921e-01 -2.16373354e-01 3.88746589e-01
-4.68539745e-01 -2.01089997e-02 -9.00041699e-01 -6.03749037e-01
-1.93603039e-01 -8.97514641e-01 5.79648376e-01 -2.65686989e-01
-9.65787888e-01 -1.54674006e+00 2.78975099e-01 4.48779434e-01
2.34430850e-01 5.80112875e-01 1.52151477e+00 -1.28004003e+00
-9.56518576e-03 2.94339776e-01 -1.25514627e-01 2.65507817e-01
3.48141879e-01 -6.93974942e-02 -8.62513602e-01 2.52912622e-02
3.97642553e-01 -5.97601950e-01 7.25695193e-01 3.01084332e-02
6.97070539e-01 -5.09760618e-01 -1.88252315e-01 3.07546444e-02
1.11128485e+00 1.17439970e-01 2.86990583e-01 6.91352487e-01
5.83566904e-01 1.00773716e+00 7.18517065e-01 7.46021420e-02
5.90458095e-01 8.09582591e-01 -1.65063608e-02 3.13461751e-01
-3.93464476e-01 1.06194228e-01 4.35585022e-01 1.35943854e+00
6.53367713e-02 -2.91503519e-01 -1.35857248e+00 4.94316429e-01
-1.63244414e+00 -1.00859380e+00 -5.46212792e-01 1.49348378e+00
1.26028836e+00 2.18090683e-01 -2.86879782e-02 2.96784908e-01
4.23561096e-01 6.78337872e-01 2.90949196e-01 -1.07200778e+00
-4.77618784e-01 1.72738954e-01 -3.22258249e-02 9.82123733e-01
-1.17433155e+00 9.44868565e-01 6.81242895e+00 5.36401987e-01
-9.65620577e-01 4.31067795e-01 6.22904480e-01 3.80216151e-01
-3.00385088e-01 6.15166761e-02 -7.01132953e-01 1.26870230e-01
1.25131440e+00 -2.78279990e-01 -8.93007591e-02 4.63393390e-01
-2.35981420e-01 -2.12426409e-01 -1.28920734e+00 3.62648189e-01
4.80649501e-01 -1.67281413e+00 -2.24358231e-01 -1.57239169e-01
6.76416159e-01 -1.06207810e-01 -1.08510129e-01 4.26344037e-01
2.06059784e-01 -1.09243608e+00 5.97632289e-01 1.78073213e-01
4.90780443e-01 -4.02130455e-01 7.98675656e-01 5.40708303e-01
-6.58987403e-01 -1.74690679e-01 -8.38686526e-02 -7.79502213e-01
1.25129983e-01 7.20179603e-02 -1.34540164e+00 4.11663830e-01
2.10243076e-01 8.16585004e-01 -7.40898490e-01 1.68350250e-01
-4.79414821e-01 7.83973217e-01 7.81694278e-02 -5.56570590e-01
1.35935441e-01 6.12001233e-02 6.56550169e-01 1.45883155e+00
-1.92333311e-01 6.60963878e-02 1.63887218e-01 3.40983748e-01
-3.88308875e-02 4.68930304e-01 -5.52287459e-01 6.44439012e-02
5.09600222e-01 9.62606907e-01 -5.18834829e-01 -4.80318725e-01
-2.65058458e-01 5.22356331e-01 9.04803336e-01 -5.49053848e-02
-6.73888087e-01 -6.80334494e-02 3.26110840e-01 -1.57301575e-01
2.11700603e-01 -2.59369433e-01 -5.22110105e-01 -8.32725525e-01
-6.50452450e-02 -8.64114225e-01 3.28039557e-01 -3.49810302e-01
-1.42156184e+00 8.40180576e-01 1.72969729e-01 -8.32688034e-01
-4.21581089e-01 -9.29277241e-01 -6.92978799e-01 9.15167809e-01
-1.22590244e+00 -1.19775355e+00 2.65189648e-01 -4.25417162e-02
8.14338207e-01 -4.26288843e-01 1.30406713e+00 -2.11610273e-02
-3.93683493e-01 3.09340119e-01 -2.13900348e-03 4.23481852e-01
6.83515131e-01 -1.49860930e+00 -5.12897931e-02 1.35368317e-01
2.31516123e-01 2.19420061e-01 7.63551772e-01 -3.99248511e-01
-5.51550925e-01 -6.63736761e-01 1.99058974e+00 -9.32366073e-01
1.21158898e+00 -2.41700932e-01 -7.40330517e-01 9.11794901e-01
9.37190354e-01 -4.39714760e-01 1.05479395e+00 7.22983241e-01
-4.73858833e-01 3.98671925e-01 -6.34334505e-01 5.77183604e-01
6.91457510e-01 -7.98061013e-01 -1.45866323e+00 8.16818237e-01
7.78985202e-01 -6.12215459e-01 -1.23190153e+00 1.51878521e-01
1.27701126e-02 -7.57701457e-01 6.24495685e-01 -9.82076049e-01
1.19153166e+00 2.89385736e-01 -5.48817039e-01 -1.25122142e+00
-2.95605928e-01 -9.10007134e-02 -7.12491199e-02 1.29739356e+00
8.59782040e-01 -5.20013869e-01 1.34968787e-01 2.05174819e-01
-2.46104330e-01 -9.96106327e-01 -1.16286504e+00 -5.17606974e-01
6.43742621e-01 -4.27807868e-02 2.40131572e-01 1.35961175e+00
6.86051965e-01 1.32960224e+00 3.92391123e-02 -4.76562977e-01
-1.43009841e-01 3.16192925e-01 4.71379250e-01 -1.23043108e+00
-9.23033655e-02 -5.99184811e-01 -3.52295548e-01 -7.98821330e-01
7.49188840e-01 -1.36093771e+00 -2.71654814e-01 -1.63478339e+00
9.80082378e-02 -4.35035378e-01 7.60352686e-02 1.15249872e-01
-2.13879794e-01 -2.85331041e-01 3.92619610e-01 2.07421243e-01
-3.94791692e-01 6.45124316e-01 1.02274609e+00 -2.13521540e-01
-1.78549916e-01 -3.12936753e-01 -6.30566537e-01 7.89702177e-01
8.61256480e-01 -3.05244744e-01 -1.53756440e-01 -6.77603602e-01
4.33474183e-02 2.97470659e-01 2.13058069e-02 -5.70611775e-01
-1.58034667e-01 -6.44134581e-02 1.49005190e-01 -5.72310448e-01
5.03301978e-01 -4.17670369e-01 -4.34500873e-01 3.68147582e-01
-1.10300493e+00 3.79224092e-01 -7.44953100e-03 -1.47530176e-02
-4.34129685e-01 -3.51014763e-01 4.62897331e-01 3.10904622e-01
-3.50261927e-01 -7.16750741e-01 -4.85052854e-01 3.42822999e-01
1.05241776e+00 1.83575794e-01 -8.58860433e-01 -1.04261301e-01
-1.01651025e+00 2.03900561e-01 -4.91449907e-02 7.36231148e-01
3.64304841e-01 -1.49595666e+00 -1.14841008e+00 -2.42673025e-01
-2.07910426e-02 -4.18422073e-01 -3.95919591e-01 8.46168995e-01
-5.03532529e-01 5.43504477e-01 -7.51717202e-03 -4.53809321e-01
-1.72920597e+00 3.83388966e-01 1.73935607e-01 -5.91930389e-01
-4.83349174e-01 9.06383574e-01 -3.21124464e-01 -4.75927800e-01
-1.13452271e-01 -2.39069372e-01 -9.59480405e-01 7.10139334e-01
3.90922934e-01 3.30108553e-01 -2.70258393e-02 -1.20634234e+00
-2.43820667e-01 1.29410654e-01 -2.12957799e-01 -2.77814597e-01
1.39826524e+00 2.24484522e-02 -6.16815865e-01 1.09690571e+00
1.47710061e+00 1.46927550e-01 -5.28603494e-01 -2.26850003e-01
7.24729657e-01 -1.57716751e-01 -2.10409164e-01 -5.96107423e-01
-1.87746406e-01 8.14378381e-01 1.69875368e-01 7.24235713e-01
4.78039503e-01 4.23977733e-01 3.46157253e-01 3.67199719e-01
-1.58212811e-01 -1.23447025e+00 1.45491034e-01 1.11956930e+00
1.38669848e+00 -9.88354087e-01 2.29736552e-01 -5.48686087e-01
-8.60090017e-01 1.26648128e+00 6.00771129e-01 -3.40934664e-01
7.46462107e-01 6.60752356e-02 2.66948789e-01 -4.56659973e-01
-1.14081526e+00 2.01460674e-01 3.14242810e-01 5.40295064e-01
1.28177524e+00 4.08780068e-01 -9.90336537e-01 4.61114228e-01
-8.19713831e-01 -1.04075956e+00 4.84104544e-01 7.77985752e-01
-5.22521853e-01 -1.23347342e+00 -1.88601911e-01 5.05599856e-01
-5.28427601e-01 -4.25669998e-02 -8.01964700e-01 9.83743608e-01
1.52754724e-01 1.10298908e+00 4.10175711e-01 -4.34905320e-01
3.15660447e-01 1.95069984e-01 4.90443558e-01 -8.78732443e-01
-8.48753810e-01 -2.79549599e-01 1.17957318e+00 -1.35843888e-01
-7.35258102e-01 -8.73026788e-01 -1.12860894e+00 -1.37892634e-01
-2.27427810e-01 2.66163409e-01 3.25706750e-01 1.29985809e+00
-2.20528305e-01 7.25507617e-01 7.26512015e-01 -6.36332870e-01
-5.42171180e-01 -1.54550087e+00 -4.90201674e-02 5.47419548e-01
3.55563283e-01 -5.78944266e-01 -2.97820359e-01 9.18091610e-02] | [10.821365356445312, 9.30653190612793] |
19174371-4dde-443a-91b1-e3d58f495244 | incorporating-uncertain-segmentation | 2004.06384 | null | https://arxiv.org/abs/2004.06384v2 | https://arxiv.org/pdf/2004.06384v2.pdf | Incorporating Uncertain Segmentation Information into Chinese NER for Social Media Text | Chinese word segmentation is necessary to provide word-level information for Chinese named entity recognition (NER) systems. However, segmentation error propagation is a challenge for Chinese NER while processing colloquial data like social media text. In this paper, we propose a model (UIcwsNN) that specializes in identifying entities from Chinese social media text, especially by leveraging ambiguous information of word segmentation. Such uncertain information contains all the potential segmentation states of a sentence that provides a channel for the model to infer deep word-level characteristics. We propose a trilogy (i.e., candidate position embedding -> position selective attention -> adaptive word convolution) to encode uncertain word segmentation information and acquire appropriate word-level representation. Experiments results on the social media corpus show that our model alleviates the segmentation error cascading trouble effectively, and achieves a significant performance improvement of more than 2% over previous state-of-the-art methods. | ['Shengbin Jia', 'Yang Xiang', 'Xiaojun Chen', 'Shijia E', 'Ling Ding'] | 2020-04-14 | incorporating-uncertain-segmentation-1 | https://aclanthology.org/2020.socialnlp-1.7 | https://aclanthology.org/2020.socialnlp-1.7.pdf | ws-2020-7 | ['chinese-named-entity-recognition'] | ['natural-language-processing'] | [-1.09891044e-02 4.43332680e-02 -2.10789099e-01 -4.79695290e-01
-7.41984129e-01 -5.24405658e-01 -3.68715562e-02 2.42583573e-01
-1.10377932e+00 5.54790378e-01 4.54277158e-01 -6.70100987e-01
4.38057303e-01 -8.90084803e-01 -3.71266186e-01 -3.90705675e-01
2.41232380e-01 2.85621643e-01 3.24421436e-01 -1.20228752e-01
4.34115231e-01 9.13728848e-02 -6.39327943e-01 2.74737686e-01
1.56080556e+00 6.23021901e-01 5.46778798e-01 5.35856009e-01
-1.07426536e+00 3.72585952e-01 -7.61215270e-01 -4.49013203e-01
-2.93863893e-01 -2.51162678e-01 -1.19251418e+00 -9.76973698e-02
-2.36013204e-01 -2.05588881e-02 -3.89695354e-02 1.37723613e+00
5.71143568e-01 7.43234158e-02 3.26016605e-01 -3.74372900e-01
-1.02790606e+00 1.36892104e+00 -4.49099392e-01 6.09765112e-01
-1.27971387e-02 -2.55164146e-01 1.26611090e+00 -9.77898240e-01
4.93317127e-01 9.91178274e-01 7.07410097e-01 8.84185791e-01
-5.71104527e-01 -5.92151165e-01 6.45925462e-01 4.55929488e-02
-1.38941312e+00 6.07825369e-02 4.83669430e-01 -5.37083820e-02
1.16059136e+00 2.01422423e-01 4.78857696e-01 8.06097150e-01
-3.00100143e-03 1.51162267e+00 5.24627984e-01 -4.02966619e-01
1.40368948e-02 -1.73539266e-01 8.20933580e-01 3.81123722e-01
4.10809010e-01 -7.78198719e-01 -1.24859072e-01 1.46632895e-01
5.20413876e-01 1.36726443e-02 -1.61269948e-01 8.02793980e-01
-9.95336890e-01 6.12185717e-01 4.26678389e-01 7.26393759e-01
-4.49783564e-01 7.57190660e-02 1.45235747e-01 -1.77955106e-01
7.07579851e-01 5.70413709e-01 -1.06945181e+00 -2.82735169e-01
-8.29114079e-01 -3.33919436e-01 6.38811588e-01 1.12492359e+00
6.81922972e-01 1.85932353e-01 -3.66468012e-01 7.08267927e-01
3.98606837e-01 4.43554789e-01 8.76865387e-01 -2.50286430e-01
7.99853265e-01 7.37717688e-01 6.74640164e-02 -7.17953444e-01
-6.21353686e-01 -5.31550288e-01 -6.62479997e-01 -8.37151229e-01
1.35067597e-01 -1.10013890e+00 -1.29987800e+00 1.44989538e+00
2.60605156e-01 4.46019292e-01 1.04487754e-01 7.40346670e-01
9.13463175e-01 1.07467985e+00 4.74476665e-01 8.44272077e-02
1.62162387e+00 -1.00703704e+00 -1.02482331e+00 -6.09131634e-01
8.60938907e-01 -6.73237503e-01 9.44541991e-01 4.12986204e-02
-7.48793900e-01 -3.19739550e-01 -8.10492814e-01 -1.12853877e-01
-6.30481839e-01 2.60792643e-01 6.09085321e-01 8.38976622e-01
-6.48761511e-01 4.38605636e-01 -1.07943070e+00 -2.94416491e-02
3.49045277e-01 3.69961798e-01 1.42753720e-01 6.32339418e-02
-1.81777060e+00 5.54363549e-01 8.96772444e-01 5.98769546e-01
3.38086374e-02 -6.45567358e-01 -1.05211878e+00 3.40318412e-01
2.88440377e-01 -1.41138416e-02 1.25517154e+00 -5.44285893e-01
-1.26297700e+00 4.43530470e-01 -4.80611205e-01 -2.19483376e-01
1.24760196e-01 -5.71117759e-01 -9.07849073e-01 -6.45631924e-02
2.61028737e-01 6.25921547e-01 2.29852691e-01 -9.39179897e-01
-8.23986351e-01 -1.87490255e-01 -3.86453241e-01 2.17557535e-01
-6.15607560e-01 3.26846093e-01 -9.39840257e-01 -7.52296329e-01
2.64278322e-01 -5.74175835e-01 -8.37185144e-01 -8.12358737e-01
-9.36241269e-01 -7.33732760e-01 4.52935666e-01 -8.38975847e-01
1.96458030e+00 -1.92390168e+00 -4.04221892e-01 1.60314575e-01
2.11673543e-01 6.82308674e-01 -1.77100599e-01 2.79905587e-01
1.80635735e-01 1.02643490e+00 -3.40200335e-01 -4.31973577e-01
4.90002856e-02 1.31541595e-01 -5.34977876e-02 1.03058644e-01
7.74648130e-01 1.22578061e+00 -1.13413239e+00 -7.01717377e-01
-2.11469159e-01 2.45347604e-01 -2.39368543e-01 1.24340311e-01
-1.46280482e-01 1.11673862e-01 -1.10373878e+00 6.17228448e-01
9.45303738e-01 -2.76591331e-01 2.00436279e-01 9.50753167e-02
-3.81903470e-01 5.11093140e-01 -1.14123046e+00 1.47758031e+00
-2.82118261e-01 3.73314232e-01 -2.19862953e-01 -5.85185528e-01
7.52314568e-01 3.26947123e-01 1.49071157e-01 -4.75576192e-01
3.94830942e-01 2.75797814e-01 -5.81897050e-02 -4.59067017e-01
9.69443917e-01 2.60597259e-01 -6.64299548e-01 2.23970160e-01
1.36396572e-01 3.25913876e-01 2.89343238e-01 2.49780938e-01
9.00712967e-01 -2.69625694e-01 -5.43935187e-02 -4.47759420e-01
5.12833714e-01 1.38312951e-01 1.15145767e+00 7.84777880e-01
-3.67038339e-01 6.82063103e-01 4.83710974e-01 -1.28421262e-01
-8.65227520e-01 -7.29750216e-01 -9.93643925e-02 1.09783602e+00
3.88188392e-01 -4.49119836e-01 -1.21168768e+00 -9.02372777e-01
-5.19018829e-01 7.92332530e-01 -5.54499209e-01 2.00856060e-01
-1.01132119e+00 -1.14684486e+00 9.23527122e-01 1.02732503e+00
5.43406487e-01 -1.33651912e+00 9.47689861e-02 5.65041423e-01
-2.61696219e-01 -1.27406514e+00 -8.96205723e-01 2.42449060e-01
-6.77813470e-01 -6.54104888e-01 -7.36471713e-01 -1.29066503e+00
8.69289756e-01 -9.59545281e-03 9.13740695e-01 3.41608942e-01
-1.15258219e-04 -9.19511020e-02 -7.30903089e-01 -3.99315834e-01
1.11490197e-01 7.36862898e-01 -1.18036330e-01 -5.12684025e-02
9.31464612e-01 1.94167405e-01 -6.82371616e-01 1.44813851e-01
-1.10433590e+00 -6.49455488e-02 7.13691115e-01 8.77316773e-01
5.72302341e-01 2.38053184e-02 7.98894167e-01 -1.43914843e+00
8.87732685e-01 -7.52495110e-01 -4.18776870e-01 3.98118943e-01
-6.65353179e-01 1.75994679e-01 8.22858214e-01 -5.39325655e-01
-1.49237347e+00 1.00834712e-01 -7.71522403e-01 4.29734349e-01
-2.64178634e-01 1.10644126e+00 -4.81607884e-01 5.00548184e-01
2.59619445e-01 2.59669989e-01 -7.70106673e-01 -9.03449416e-01
4.38946307e-01 1.10334742e+00 5.11083782e-01 -5.07432342e-01
3.99201959e-01 4.19759490e-02 -9.61087346e-01 -8.07035923e-01
-1.10177720e+00 -9.19557154e-01 -7.62067139e-01 2.24708453e-01
1.33038199e+00 -8.26185942e-01 -4.38547462e-01 6.00132108e-01
-1.51101720e+00 -1.36191010e-01 2.50886083e-01 5.44816554e-01
5.40206730e-01 4.87977266e-01 -1.28090227e+00 -7.05850422e-01
-5.54625332e-01 -9.50572550e-01 6.99756861e-01 9.26092684e-01
1.18128695e-01 -1.12104297e+00 -1.39688864e-01 1.96780786e-01
3.12442988e-01 -2.33907968e-01 5.49121678e-01 -1.20851636e+00
-4.33047712e-01 -2.68463284e-01 -4.42502052e-01 2.49692410e-01
-1.29232123e-01 -1.19182467e-01 -8.16471100e-01 2.21843973e-01
-3.83046895e-01 1.18290864e-01 1.13052309e+00 2.05511957e-01
1.29661119e+00 -2.33297616e-01 -3.77086490e-01 3.36500674e-01
1.34917355e+00 1.65475219e-01 5.96782804e-01 1.63834780e-01
1.07172716e+00 2.43637368e-01 5.75780272e-01 3.56322229e-01
6.59384847e-01 -2.31504351e-01 1.78532228e-01 -2.96352476e-01
2.95511276e-01 -2.93478429e-01 2.05044329e-01 1.47943497e+00
3.28394860e-01 -7.52188802e-01 -1.18883729e+00 1.06657529e+00
-1.62598789e+00 -5.68787813e-01 -4.97658014e-01 1.37367976e+00
1.17686093e+00 3.37400734e-01 -5.38263679e-01 -2.62196809e-01
1.32404399e+00 2.84997046e-01 -6.26011014e-01 -4.88901943e-01
-9.14987773e-02 3.59026253e-01 7.23885119e-01 5.52492380e-01
-1.29132164e+00 1.70174575e+00 5.32435751e+00 1.16829097e+00
-9.34948206e-01 3.18009406e-02 7.59247243e-01 8.32992852e-01
-6.93155646e-01 -1.73502177e-01 -1.42197371e+00 6.11190200e-01
9.18506742e-01 -7.39106014e-02 -3.88800874e-02 7.23534882e-01
-4.25663292e-02 2.13462710e-01 -3.95559251e-01 4.75223005e-01
-2.93308180e-02 -1.36469603e+00 -7.32861459e-02 -1.86263114e-01
9.32484388e-01 4.78827268e-01 -1.96034297e-01 5.03061116e-01
5.86034358e-01 -8.06074679e-01 4.89210129e-01 2.03511164e-01
6.70772433e-01 -7.92744398e-01 1.18756723e+00 5.00414789e-01
-1.42139292e+00 1.97410360e-01 -4.03171718e-01 2.30440095e-01
5.96889615e-01 6.93747103e-01 -6.84142113e-01 5.15387833e-01
4.79399353e-01 5.10663569e-01 -3.90959412e-01 1.09874153e+00
-6.71434879e-01 1.22891927e+00 -3.55208308e-01 -7.38946736e-01
7.94903874e-01 -2.47140750e-02 1.59417972e-01 1.77404165e+00
2.80825466e-01 4.29135561e-01 8.04652646e-02 7.51122594e-01
-5.03350675e-01 3.13055992e-01 3.21487099e-01 -4.87061501e-01
7.58343756e-01 1.29156506e+00 -1.10567391e+00 -3.06003928e-01
-4.10607010e-01 1.30019081e+00 6.07577085e-01 3.83025646e-01
-7.51352549e-01 -1.17584431e+00 6.39275193e-01 -5.99083781e-01
6.85198665e-01 -3.90480906e-01 -5.26318133e-01 -1.42796922e+00
-8.52146819e-02 -2.19407529e-01 4.83990490e-01 -3.08964580e-01
-1.40579462e+00 8.86948049e-01 -7.06759453e-01 -7.78777122e-01
1.19409032e-01 -8.13682854e-01 -1.04209054e+00 9.96557891e-01
-1.79306781e+00 -8.83943796e-01 9.06739458e-02 1.10532828e-01
7.16407955e-01 2.49974921e-01 6.54947162e-01 4.30785179e-01
-1.08393073e+00 8.49494994e-01 1.79863214e-01 1.08420300e+00
2.99135923e-01 -1.52726805e+00 1.11653829e+00 1.26031673e+00
2.07325324e-01 8.84703696e-01 2.29909495e-01 -9.39062178e-01
-1.04878747e+00 -1.24605346e+00 1.56032288e+00 -3.86743754e-01
7.35448480e-01 -4.29522693e-01 -9.58723307e-01 4.63136643e-01
2.15517089e-01 -3.98431830e-02 9.92615402e-01 1.98265359e-01
1.28762126e-01 3.20394367e-01 -8.15294504e-01 7.26650715e-01
8.43577743e-01 -3.83887917e-01 -9.46697295e-01 -1.03183329e-01
1.47650492e+00 -4.38721895e-01 -7.25038111e-01 1.11709155e-01
9.56580490e-02 -2.37197831e-01 5.91494381e-01 -7.76632607e-01
1.67564079e-01 -2.84845471e-01 2.51912415e-01 -1.29997003e+00
-3.32546234e-01 -6.44633889e-01 1.57365948e-01 1.57239234e+00
1.12582457e+00 -5.06724119e-01 8.87678802e-01 7.50533521e-01
-6.01869106e-01 -7.20700741e-01 -6.22326970e-01 -4.42597002e-01
3.47569942e-01 -8.03893268e-01 8.80534172e-01 1.11128533e+00
1.83949813e-01 3.33370298e-01 3.76283526e-02 4.51331884e-01
9.83229652e-02 -1.63354024e-01 -1.55573487e-01 -9.78116393e-01
3.25894624e-01 -3.74748141e-01 -4.68014218e-02 -1.76095617e+00
4.51008976e-01 -7.31040776e-01 5.46271205e-01 -1.78853011e+00
-8.93639401e-02 -5.70537210e-01 -7.94518173e-01 4.43998724e-01
-8.82693648e-01 2.61342879e-02 1.06631011e-01 -2.80603856e-01
-1.11058748e+00 4.03585702e-01 1.19899070e+00 3.62635404e-03
-2.62011230e-01 -2.88643129e-03 -9.84316230e-01 7.09037662e-01
9.34291184e-01 -6.65686429e-01 1.56520158e-01 -1.11817610e+00
5.82685828e-01 -2.15177804e-01 -5.05002797e-01 -5.44015586e-01
6.39413238e-01 -1.74743891e-01 4.30209249e-01 -9.20827627e-01
-1.94042310e-01 -5.17367840e-01 -7.80917764e-01 3.06991547e-01
-3.80649596e-01 5.49182929e-02 5.58030829e-02 6.36225045e-01
-2.38188103e-01 -5.91401935e-01 2.36566111e-01 -1.37288585e-01
-1.17672074e+00 5.75529754e-01 -7.29140997e-01 5.77577710e-01
4.43385929e-01 2.18223296e-02 -3.45576495e-01 6.88445717e-02
-6.54531538e-01 8.97557974e-01 -6.48212209e-02 4.57724601e-01
5.54846704e-01 -1.11960685e+00 -7.33454227e-01 1.43957004e-01
1.41030997e-02 3.25104713e-01 3.63545328e-01 3.07201654e-01
-7.39497960e-01 3.48631084e-01 3.41987520e-01 -1.64095178e-01
-6.49577498e-01 1.74234346e-01 4.56846282e-02 -3.03551793e-01
-2.86553651e-01 1.30518949e+00 -3.29450905e-01 -8.04718971e-01
-2.24128701e-02 -6.33280158e-01 -6.86346710e-01 1.99589450e-02
7.24018514e-01 1.58143491e-01 -3.37043256e-02 -5.65984964e-01
-3.85615796e-01 3.01224023e-01 -3.22225332e-01 7.22328899e-03
1.20358527e+00 -3.56004328e-01 -2.85095304e-01 2.62797803e-01
1.09924114e+00 3.10541034e-01 -9.87142086e-01 -4.61508542e-01
6.54854894e-01 3.14130746e-02 1.54021323e-01 -6.27144516e-01
-1.04600155e+00 1.04073012e+00 1.03030317e-01 2.84936607e-01
7.60303140e-01 6.60790801e-02 1.59079576e+00 5.88077068e-01
5.00425063e-02 -1.62854517e+00 -4.17652696e-01 1.18867445e+00
-9.06096250e-02 -1.37566733e+00 -5.18149972e-01 -4.15664047e-01
-8.24508965e-01 1.08770883e+00 8.96424770e-01 -4.15500998e-02
8.82748842e-01 2.09477708e-01 2.54305482e-01 4.14377078e-02
-2.02534586e-01 -6.14119112e-01 2.65717745e-01 2.90332288e-01
5.49531937e-01 1.50958195e-01 -8.58928263e-01 1.60258079e+00
1.03945926e-01 -5.32183766e-01 4.71839875e-01 7.90149629e-01
-7.22300649e-01 -1.22392523e+00 -1.51751675e-02 4.95975554e-01
-1.07497847e+00 -7.59589612e-01 -1.99760690e-01 2.93501675e-01
2.02844307e-01 1.12964678e+00 2.34626085e-01 -3.35755199e-01
2.10668612e-02 -3.34823020e-02 -4.62254226e-01 -8.68175149e-01
-7.72134602e-01 1.16000153e-01 2.04335153e-01 -1.87002957e-01
-3.26789767e-02 -5.23611307e-01 -2.07329631e+00 -1.30144786e-02
-8.89993906e-01 6.15279436e-01 7.30557561e-01 1.09152091e+00
3.79023284e-01 5.48811615e-01 6.43475115e-01 -4.19436216e-01
-2.32167989e-01 -1.00423074e+00 -5.26860893e-01 1.88270912e-01
1.05326228e-01 1.17160983e-01 -1.68947116e-01 -1.33933038e-01] | [9.847980499267578, 9.909340858459473] |
323ee2ca-35c5-4fed-9fcd-58e575e0e2c8 | constructing-self-motivated-pyramid | 1908.09547 | null | https://arxiv.org/abs/1908.09547v1 | https://arxiv.org/pdf/1908.09547v1.pdf | Constructing Self-motivated Pyramid Curriculums for Cross-Domain Semantic Segmentation: A Non-Adversarial Approach | We propose a new approach, called self-motivated pyramid curriculum domain adaptation (PyCDA), to facilitate the adaptation of semantic segmentation neural networks from synthetic source domains to real target domains. Our approach draws on an insight connecting two existing works: curriculum domain adaptation and self-training. Inspired by the former, PyCDA constructs a pyramid curriculum which contains various properties about the target domain. Those properties are mainly about the desired label distributions over the target domain images, image regions, and pixels. By enforcing the segmentation neural network to observe those properties, we can improve the network's generalization capability to the target domain. Motivated by the self-training, we infer this pyramid of properties by resorting to the semantic segmentation network itself. Unlike prior work, we do not need to maintain any additional models (e.g., logistic regression or discriminator networks) or to solve minmax problems which are often difficult to optimize. We report state-of-the-art results for the adaptation from both GTAV and SYNTHIA to Cityscapes, two popular settings in unsupervised domain adaptation for semantic segmentation. | ['Boqing Gong', 'Qing Lian', 'Lixin Duan', 'Fengmao Lv'] | 2019-08-26 | constructing-self-motivated-pyramid-1 | http://openaccess.thecvf.com/content_ICCV_2019/html/Lian_Constructing_Self-Motivated_Pyramid_Curriculums_for_Cross-Domain_Semantic_Segmentation_A_Non-Adversarial_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Lian_Constructing_Self-Motivated_Pyramid_Curriculums_for_Cross-Domain_Semantic_Segmentation_A_Non-Adversarial_ICCV_2019_paper.pdf | iccv-2019-10 | ['synthetic-to-real-translation'] | ['computer-vision'] | [ 5.71222305e-01 4.88277286e-01 -2.68442929e-01 -6.50128961e-01
-6.24924481e-01 -8.28674614e-01 4.41703767e-01 -6.99802116e-02
-5.50794244e-01 7.04486191e-01 -1.63861632e-01 -5.63224107e-02
2.22091287e-01 -9.58345115e-01 -1.03405488e+00 -6.53264046e-01
3.35986257e-01 8.15679371e-01 5.68057001e-01 -4.01161104e-01
-5.60578071e-02 5.51274382e-02 -1.37247765e+00 1.91564709e-01
1.30561495e+00 9.11169529e-01 4.14525479e-01 4.88555759e-01
-2.60513514e-01 4.85123485e-01 -6.07441902e-01 -2.18798473e-01
1.81326881e-01 -5.80054641e-01 -1.03219402e+00 4.17874008e-01
2.97492325e-01 -1.15152776e-01 -5.80976605e-02 1.36425817e+00
1.32674709e-01 3.54494840e-01 1.08713186e+00 -1.12710333e+00
-9.12414372e-01 6.49959743e-01 -4.85092491e-01 -2.23147422e-02
-1.63384259e-01 3.00003812e-02 9.12801623e-01 -6.02393746e-01
7.44144201e-01 1.25550210e+00 4.92323339e-01 7.84651101e-01
-1.38765848e+00 -6.51233196e-01 5.41155338e-01 4.54869717e-02
-1.13859367e+00 -5.57963364e-02 9.62578356e-01 -4.83325601e-01
3.72487187e-01 -2.02207401e-01 4.28959668e-01 1.49849391e+00
-5.04024804e-01 1.13867402e+00 1.16533601e+00 -5.36327124e-01
5.22675812e-01 6.11416280e-01 1.12216942e-01 3.62060010e-01
6.87237224e-03 -1.22100171e-02 -1.05684116e-01 1.17090642e-01
9.64312851e-01 -4.38106745e-01 -1.91326346e-02 -8.08191061e-01
-9.42205787e-01 1.05678499e+00 3.94440353e-01 1.53196052e-01
-6.65151030e-02 -8.97836164e-02 4.81043905e-01 3.09645653e-01
5.98695278e-01 5.21048903e-01 -7.98427045e-01 3.08779746e-01
-7.15581179e-01 2.78211325e-01 7.14609802e-01 1.42784870e+00
1.00266981e+00 1.28366441e-01 -1.67270154e-01 1.06422341e+00
1.34288102e-01 4.43805873e-01 7.45364964e-01 -8.83991420e-01
3.92602324e-01 4.51221734e-01 2.63048876e-02 -4.14458841e-01
-2.28096530e-01 -3.58599842e-01 -7.13057637e-01 3.15069072e-02
7.32438028e-01 -4.66219217e-01 -1.37664092e+00 1.99461985e+00
3.11398596e-01 3.46764684e-01 3.88144672e-01 8.42911065e-01
7.45573461e-01 5.03536940e-01 3.43622923e-01 2.17667848e-01
1.23450279e+00 -1.25217175e+00 -3.20931494e-01 -5.05503893e-01
5.17557383e-01 -3.39462608e-01 1.50659060e+00 2.03208715e-01
-8.11862171e-01 -9.09016907e-01 -9.65869248e-01 1.29601523e-01
-6.77460015e-01 -2.20817351e-03 4.24026310e-01 5.90543151e-01
-8.27126980e-01 4.98784125e-01 -7.75914848e-01 -5.36418378e-01
6.62034512e-01 2.23024458e-01 2.38886382e-02 9.37976912e-02
-1.32120824e+00 6.37903512e-01 8.99101913e-01 -6.44120932e-01
-1.10557997e+00 -7.18729794e-01 -1.06276453e+00 5.59617430e-02
5.10479033e-01 -6.75954223e-01 1.60921359e+00 -1.89129949e+00
-1.79751670e+00 1.07115650e+00 1.26151159e-01 -5.74201524e-01
4.73682314e-01 6.99590072e-02 -2.08968475e-01 1.50361314e-01
2.24645883e-01 1.16302729e+00 8.89477432e-01 -1.42746174e+00
-7.75828063e-01 -2.22168788e-01 5.31813540e-02 4.57889348e-01
-4.35962349e-01 -2.74273127e-01 -6.63626552e-01 -8.99347484e-01
-6.75174817e-02 -9.17155981e-01 -5.04881203e-01 -2.15081766e-01
-5.67413986e-01 -3.02078754e-01 8.13224196e-01 -4.25537050e-01
5.94451368e-01 -2.19471145e+00 2.39053011e-01 3.32126051e-01
-9.33300927e-02 1.54098377e-01 -3.28466624e-01 -2.91067511e-01
-1.13967180e-01 5.72048826e-03 -7.26455867e-01 -3.21485162e-01
2.07881570e-01 5.26509047e-01 -3.06020707e-01 1.54279724e-01
3.85455191e-01 7.86860764e-01 -9.61588383e-01 -5.33830166e-01
1.17693916e-02 1.57881021e-01 -6.96161985e-01 4.72064286e-01
-9.07328427e-01 9.04733181e-01 -7.17263103e-01 3.05767477e-01
6.74558759e-01 -2.76199251e-01 1.85471117e-01 2.45595127e-01
3.15909415e-01 1.94303185e-01 -1.06347108e+00 2.13767314e+00
-3.59895110e-01 4.03866529e-01 -3.54709513e-02 -1.61266041e+00
1.09356797e+00 1.61883309e-01 4.33379322e-01 -7.44423687e-01
1.33566126e-01 1.01178624e-01 -2.62998670e-01 -3.88636082e-01
3.74377847e-01 -1.57166377e-01 -3.94387335e-01 2.56059915e-01
4.96451557e-01 -3.90850216e-01 2.64278203e-01 -8.77169818e-02
5.08521616e-01 6.39264166e-01 1.35588974e-01 -4.97695923e-01
6.36878312e-01 5.24376988e-01 7.33457148e-01 7.09153831e-01
-1.51936755e-01 6.77519977e-01 5.44088423e-01 -5.10829166e-02
-1.21598196e+00 -1.28842056e+00 -4.08189222e-02 1.62063122e+00
3.28670651e-01 2.26698682e-01 -1.34462488e+00 -9.78043020e-01
-9.30208191e-02 7.39131927e-01 -8.25760484e-01 -1.60761818e-01
-5.28850019e-01 -5.55036366e-01 5.38605690e-01 7.44796813e-01
7.67897666e-01 -1.18839324e+00 -3.15173954e-01 2.64188409e-01
-1.58715352e-01 -1.46804607e+00 -5.60689628e-01 3.70923728e-01
-9.51748252e-01 -8.82712781e-01 -1.10230672e+00 -1.24657774e+00
8.50181699e-01 -1.80900365e-01 1.35382736e+00 -4.30430353e-01
1.09275021e-01 3.43312085e-01 -3.63616139e-01 -5.76304138e-01
-6.71212673e-01 5.11995733e-01 -2.41373494e-01 -7.54126981e-02
4.36076432e-01 -7.09311128e-01 -4.51599658e-01 3.44172627e-01
-9.37653780e-01 6.59388527e-02 5.50280035e-01 8.32906365e-01
9.20264423e-01 2.23604590e-02 6.14628255e-01 -1.51360846e+00
4.74880993e-01 -5.92344284e-01 -7.77980983e-01 2.18645602e-01
-4.64509964e-01 2.39279240e-01 9.58605170e-01 -7.53984392e-01
-1.34410810e+00 3.99554819e-01 -6.49084747e-02 -3.56071651e-01
-8.32143843e-01 2.13282093e-01 -3.32189202e-01 1.28049895e-01
9.34855819e-01 2.59899706e-01 -2.29557261e-01 -5.65267742e-01
6.87038243e-01 5.15394449e-01 8.92114460e-01 -1.00951433e+00
8.19339097e-01 3.94694626e-01 -5.46217799e-01 -6.51759028e-01
-9.86392558e-01 -4.85003352e-01 -9.55298007e-01 2.44250879e-01
1.14099133e+00 -9.78647470e-01 -2.11702317e-01 4.51169997e-01
-9.62009847e-01 -8.78537595e-01 -5.32592237e-01 1.34165272e-01
-9.56957221e-01 2.68969476e-01 -3.63173544e-01 -3.65190327e-01
-4.14975882e-02 -1.17858756e+00 9.18466091e-01 4.60301787e-01
-6.79833591e-02 -1.40201473e+00 8.32407624e-02 1.70508280e-01
3.06923956e-01 3.92033815e-01 1.03785348e+00 -9.48216021e-01
-2.14940801e-01 3.79992396e-01 -2.80195981e-01 6.36735678e-01
2.15587407e-01 -3.26914817e-01 -1.09321368e+00 -1.76796600e-01
-1.16495833e-01 -6.08259976e-01 8.32040489e-01 5.04469514e-01
1.56377852e+00 -1.87014431e-01 -3.34218681e-01 8.81102204e-01
1.38754177e+00 1.73593253e-01 4.30259198e-01 6.05713308e-01
5.88323593e-01 6.66734219e-01 7.48193383e-01 1.55276015e-01
4.90155578e-01 6.22428238e-01 2.71125078e-01 -6.01809382e-01
-1.19020820e-01 -4.49730009e-01 2.43585542e-01 3.18642795e-01
2.22002804e-01 -6.04720302e-02 -9.05185938e-01 8.48960817e-01
-1.73752034e+00 -3.38900924e-01 2.39837527e-01 1.88284755e+00
1.15646374e+00 2.33968511e-01 4.02772963e-01 -3.64190370e-01
7.24883199e-01 -5.82279116e-02 -9.90763307e-01 -4.18388337e-01
-1.75609633e-01 4.83280361e-01 6.91523731e-01 4.01426286e-01
-1.43675220e+00 1.50879812e+00 6.18370771e+00 8.92979980e-01
-1.02009857e+00 1.71370834e-01 7.69806743e-01 5.10654271e-01
-2.86780596e-01 -1.28672048e-01 -6.79260314e-01 3.90497655e-01
7.64498949e-01 -5.18729612e-02 3.94778311e-01 1.24125147e+00
-1.65924311e-01 1.04491599e-01 -1.03314292e+00 5.55170655e-01
-1.09463803e-01 -1.02486348e+00 3.17030698e-02 -3.81346494e-01
1.08854985e+00 7.93839023e-02 2.40523949e-01 6.18768573e-01
1.03266132e+00 -9.49036598e-01 7.10823834e-01 1.11425230e-02
9.34124351e-01 -6.68956161e-01 4.28288847e-01 4.40458447e-01
-8.78669858e-01 -1.87923003e-03 -6.08323038e-01 2.86535293e-01
-1.23305887e-01 2.67834723e-01 -9.26139116e-01 3.81892383e-01
8.07603776e-01 7.04344749e-01 -5.05784690e-01 6.62085295e-01
-6.03468418e-01 8.23026180e-01 -2.16035321e-01 2.28079140e-01
4.99080360e-01 -3.88048261e-01 3.30393940e-01 1.38451791e+00
5.13221091e-03 -1.32026657e-01 3.86005849e-01 1.25493467e+00
-1.11448243e-01 6.68949857e-02 -5.35588026e-01 1.09868966e-01
3.19519997e-01 9.89026010e-01 -8.35233688e-01 -6.45739079e-01
-4.43035275e-01 1.19774389e+00 3.31175119e-01 9.01315033e-01
-9.64309514e-01 -3.28762889e-01 7.67808378e-01 -1.90181032e-01
6.19635820e-01 -6.23874590e-02 -6.13019109e-01 -1.10590696e+00
-1.59167588e-01 -1.06829333e+00 5.00517368e-01 -5.71707368e-01
-1.28092813e+00 6.29033685e-01 2.06950918e-01 -1.08927536e+00
-2.30138376e-01 -6.89110994e-01 -5.63390613e-01 8.11353326e-01
-1.99534178e+00 -1.14160264e+00 -2.65028000e-01 1.00108528e+00
7.46055782e-01 -2.68956602e-01 8.12486708e-01 -1.53541402e-03
-4.18840140e-01 7.30819881e-01 2.73947835e-01 4.70719546e-01
8.26641142e-01 -1.60842371e+00 6.46858037e-01 6.24740541e-01
7.20968796e-03 6.05774522e-02 6.59832239e-01 -3.95555317e-01
-4.74816591e-01 -1.32835996e+00 1.94604591e-01 -3.79348308e-01
5.69513500e-01 -6.46893024e-01 -1.04803681e+00 9.16168213e-01
2.50105709e-01 -1.04009278e-01 5.57290673e-01 1.34735256e-01
-4.11056340e-01 1.28226383e-02 -1.16459203e+00 6.02286398e-01
9.67955410e-01 -2.57984191e-01 -6.64476037e-01 4.13258195e-01
1.01641464e+00 -6.42961442e-01 -8.36310804e-01 3.18491012e-01
4.74284254e-02 -6.65961802e-01 1.13917673e+00 -9.59660232e-01
4.12595034e-01 -3.83630842e-02 -2.06289142e-02 -1.63047874e+00
-2.54306465e-01 -1.77725479e-01 2.54666924e-01 1.24985218e+00
7.09161758e-01 -6.26757145e-01 1.21641576e+00 4.13312882e-01
-3.49267095e-01 -2.75182009e-01 -6.50494635e-01 -9.66895938e-01
7.39245594e-01 -2.41490513e-01 7.22464442e-01 1.21231365e+00
-3.55928808e-01 3.65456164e-01 9.74924862e-02 2.91679144e-01
6.43198907e-01 1.88496709e-01 8.24031532e-01 -1.21691632e+00
-4.77610856e-01 -4.65438157e-01 3.70993800e-02 -1.40993714e+00
4.94372696e-01 -8.78516555e-01 3.84226620e-01 -1.23630250e+00
3.82447913e-02 -7.28322327e-01 -3.63225043e-01 5.88406801e-01
-2.02499688e-01 -1.44123496e-03 -1.36598021e-01 6.14482015e-02
-6.15208983e-01 6.18855536e-01 1.67899573e+00 -4.44307178e-01
-4.62721407e-01 1.74058810e-01 -8.69377971e-01 8.34093094e-01
9.46550310e-01 -5.64265430e-01 -8.24714780e-01 -6.22763276e-01
-2.72733122e-01 -1.35579586e-01 1.69914111e-01 -8.45108151e-01
4.57412116e-02 -4.41040009e-01 2.90446460e-01 -1.21382415e-01
7.82822818e-02 -7.84935951e-01 -4.44982767e-01 1.75625414e-01
-4.53749657e-01 -4.67471033e-01 3.90988201e-01 4.77905601e-01
-4.29426402e-01 -4.18174684e-01 1.17490125e+00 -4.10575718e-01
-1.24282575e+00 2.28264332e-01 -3.57499093e-01 4.82635647e-01
1.02751362e+00 -3.32684398e-01 -8.46106559e-02 -2.73240030e-01
-1.02312565e+00 5.58870852e-01 5.52556992e-01 4.15810943e-01
2.89121896e-01 -1.16684079e+00 -6.76027715e-01 3.03613067e-01
2.57350087e-01 6.50677919e-01 6.82230368e-02 3.96568537e-01
-2.43742555e-01 2.28421658e-01 -4.10657614e-01 -8.33635926e-01
-8.24501574e-01 7.58978546e-01 4.43650067e-01 -3.44084740e-01
-5.29980898e-01 1.01220691e+00 7.26960897e-01 -9.76070821e-01
3.14493001e-01 -4.46692169e-01 -4.32033092e-01 -1.67709693e-01
3.51841748e-02 -1.58626467e-01 -2.77321339e-01 -2.70282865e-01
6.51928633e-02 6.47097468e-01 -1.77850068e-01 -1.39305159e-01
1.17714071e+00 -7.34226033e-02 3.85825694e-01 1.99020252e-01
1.00741601e+00 -5.35282969e-01 -1.84610105e+00 -7.06305742e-01
2.10922718e-01 -4.66978513e-02 -3.13974530e-01 -8.04503441e-01
-1.12947738e+00 8.86416972e-01 4.77478504e-01 7.23807961e-02
1.34089494e+00 1.40889227e-01 7.24567533e-01 3.80556732e-01
1.75925091e-01 -1.59471583e+00 2.69928098e-01 7.88226843e-01
5.27540326e-01 -1.36234987e+00 -5.09089768e-01 -5.85126519e-01
-9.95526612e-01 9.15255725e-01 9.44453537e-01 -4.06801552e-01
5.47131956e-01 2.07148660e-02 2.81671047e-01 8.35104957e-02
-3.45035136e-01 -4.86495316e-01 2.39703476e-01 1.15137887e+00
1.82928056e-01 9.13075730e-02 1.63186207e-01 7.62117743e-01
-2.85158664e-01 -1.80767566e-01 2.66833752e-01 5.25419116e-01
-5.20902395e-01 -1.28726101e+00 -2.93154269e-01 3.49447578e-02
-1.92505911e-01 2.00803876e-02 -3.97331625e-01 1.06090367e+00
2.09894240e-01 6.25555813e-01 2.92642206e-01 -6.79740161e-02
5.58534622e-01 1.33173734e-01 3.00575942e-01 -9.80150819e-01
-2.88554996e-01 9.73963812e-02 -1.18063770e-01 -2.23106831e-01
-4.32826847e-01 -5.77937663e-01 -1.39635253e+00 2.53701657e-01
-4.03189994e-02 2.00213268e-01 5.68048239e-01 1.05665135e+00
-1.51738869e-02 7.37825036e-01 5.39461374e-01 -5.36136270e-01
-5.27424157e-01 -1.01765895e+00 -5.70234776e-01 7.08359361e-01
1.94721267e-01 -6.63168013e-01 3.28801088e-02 5.07853985e-01] | [9.744067192077637, 1.364626169204712] |
bf63b985-fa65-4752-b887-342369d78c9d | divided-spectro-temporal-attention-for-sound | 2306.02591 | null | https://arxiv.org/abs/2306.02591v1 | https://arxiv.org/pdf/2306.02591v1.pdf | Divided spectro-temporal attention for sound event localization and detection in real scenes for DCASE2023 challenge | Localizing sounds and detecting events in different room environments is a difficult task, mainly due to the wide range of reflections and reverberations. When training neural network models with sounds recorded in only a few room environments, there is a tendency for the models to become overly specialized to those specific environments, resulting in overfitting. To address this overfitting issue, we propose divided spectro-temporal attention. In comparison to the baseline method, which utilizes a convolutional recurrent neural network (CRNN) followed by a temporal multi-head self-attention layer (MHSA), we introduce a separate spectral attention layer that aggregates spectral features prior to the temporal MHSA. To achieve efficient spectral attention, we reduce the frequency pooling size in the convolutional encoder of the baseline to obtain a 3D tensor that incorporates information about frequency, time, and channel. As a result, we can implement spectral attention with channel embeddings, which is not possible in the baseline method dealing with only temporal context in the RNN and MHSA layers. We demonstrate that the proposed divided spectro-temporal attention significantly improves the performance of sound event detection and localization scores for real test data from the STARSS23 development dataset. Additionally, we show that various data augmentations, such as frameshift, time masking, channel swapping, and moderate mix-up, along with the use of external data, contribute to the overall improvement in SELD performance. | ['Jung-Woo Choi', 'Byeong-Yun Ko', 'Yusun Shul'] | 2023-06-05 | null | null | null | null | ['sound-event-detection', 'sound-event-localization-and-detection'] | ['audio', 'audio'] | [ 3.12179446e-01 -4.54698503e-01 7.97795117e-01 -2.06056312e-01
-8.61146867e-01 -3.60824257e-01 3.07876587e-01 1.84050784e-01
-5.21862745e-01 2.54189253e-01 6.13612294e-01 -2.31383920e-01
9.95180160e-02 -5.67603648e-01 -7.16790438e-01 -6.34457886e-01
-7.40566179e-02 -6.93407118e-01 2.20222414e-01 -1.08418480e-01
-7.22997785e-02 2.59649694e-01 -1.65506816e+00 2.93886960e-01
6.71753764e-01 1.14020336e+00 3.13306838e-01 8.61679018e-01
1.69039443e-02 6.61584556e-01 -8.42543840e-01 6.36757538e-02
1.87735870e-01 -4.80113387e-01 -4.33335334e-01 -1.66220129e-01
4.31862861e-01 -1.58134907e-01 -3.82471651e-01 8.19876254e-01
8.88755322e-01 7.16590166e-01 1.77024648e-01 -7.15047240e-01
-5.87284982e-01 7.35039055e-01 -2.14276865e-01 3.07997912e-01
4.00005072e-01 5.29051153e-03 1.02666748e+00 -9.71110940e-01
-1.34953395e-01 1.12657368e+00 8.44410598e-01 2.51447648e-01
-1.17117190e+00 -6.37590945e-01 2.60297149e-01 4.03606623e-01
-1.33565450e+00 -5.01209855e-01 9.84331846e-01 -3.42706263e-01
1.20295703e+00 4.60149705e-01 4.29936886e-01 1.40609050e+00
-1.11672200e-01 4.38861698e-01 9.21378613e-01 -6.02455437e-01
1.99106202e-01 -1.10888489e-01 2.26483837e-01 2.64377326e-01
-2.92540789e-01 1.11205034e-01 -5.27619183e-01 6.33504940e-03
5.21043956e-01 8.81925225e-02 -5.70493162e-01 1.12062030e-01
-1.27896738e+00 3.76767993e-01 6.68572009e-01 3.30253720e-01
-4.37141836e-01 3.11148614e-01 5.52040637e-01 1.47828043e-01
4.47182983e-01 5.85336089e-01 -4.40678090e-01 -1.13900445e-01
-8.14690769e-01 5.43826260e-02 5.24394572e-01 4.16429341e-01
5.13272882e-01 4.15603161e-01 -3.54253143e-01 1.09301865e+00
1.41664356e-01 3.53560716e-01 5.54140151e-01 -7.64106452e-01
4.57128733e-01 4.10933886e-03 -2.87339315e-02 -8.43664169e-01
-6.19494677e-01 -8.84242535e-01 -7.02752173e-01 -1.35079011e-01
2.68183470e-01 -2.48517960e-01 -9.95519102e-01 2.01231074e+00
7.17062205e-02 6.20238304e-01 -1.82013772e-02 9.30613458e-01
7.48908639e-01 7.46554375e-01 1.98022053e-02 -7.77935237e-02
1.26100230e+00 -1.05960667e+00 -9.30243850e-01 -2.79710203e-01
4.25428957e-01 -9.64870274e-01 1.29453099e+00 2.08848283e-01
-8.43026519e-01 -8.86673570e-01 -1.20328140e+00 -2.74501830e-01
-4.31298405e-01 3.12657282e-02 3.82199347e-01 6.96146071e-01
-8.22498202e-01 5.14167488e-01 -7.68063068e-01 -2.73796916e-01
-4.35199142e-02 9.75624695e-02 -1.55787230e-01 5.00166342e-02
-1.43917406e+00 6.49623096e-01 8.41476396e-02 4.16702360e-01
-6.79194987e-01 -7.25424886e-01 -8.85560036e-01 3.54007959e-01
2.46283755e-01 -3.88052136e-01 1.13231122e+00 -6.10627055e-01
-1.53790224e+00 -1.00791380e-01 -1.50644511e-01 -5.02652466e-01
6.94629997e-02 -5.31144023e-01 -7.46729255e-01 -9.62903425e-02
-2.09759876e-01 2.46946305e-01 6.98165178e-01 -8.64026368e-01
-2.83384293e-01 4.65228483e-02 -4.97707054e-02 1.09988078e-01
-5.26373684e-01 1.35720327e-01 -2.95923382e-01 -9.49704230e-01
1.24136664e-01 -8.51800203e-01 -3.59131277e-01 -5.63999236e-01
-3.24855000e-01 1.16949290e-01 5.87105155e-01 -9.05221343e-01
1.46446776e+00 -2.65589714e+00 3.37044969e-02 8.68017673e-02
-2.12433502e-01 2.04814881e-01 -2.84743547e-01 2.19880879e-01
-3.47952336e-01 -8.93078744e-02 -1.68959677e-01 -6.86775923e-01
5.83641836e-03 -3.06906011e-02 -4.40791219e-01 2.90766299e-01
2.88281649e-01 3.82671118e-01 -7.55495429e-01 -4.70499098e-02
2.48248503e-01 8.31160128e-01 -8.58931541e-01 1.50857925e-01
9.10350382e-02 5.05767584e-01 8.31991062e-02 3.49757791e-01
4.60808843e-01 -1.23515725e-02 -2.40639508e-01 -2.79370755e-01
-3.56133610e-01 9.65143859e-01 -1.32526147e+00 1.70882940e+00
-9.14349139e-01 7.44746685e-01 -1.28721282e-01 -7.17080593e-01
7.66104281e-01 6.52692676e-01 3.28513920e-01 -6.99510813e-01
3.88613530e-02 1.93951711e-01 2.15716407e-01 -4.02001202e-01
5.27258456e-01 8.59860256e-02 -3.66746746e-02 3.45218390e-01
1.12320192e-01 1.83888137e-01 -7.79710636e-02 -1.64619058e-01
1.33538306e+00 1.90941524e-02 -2.12482855e-01 1.27744883e-01
4.05040205e-01 -7.66141236e-01 5.83510995e-01 7.49799013e-01
-2.89270319e-02 9.49773252e-01 6.21677227e-02 -1.86455980e-01
-7.81243503e-01 -1.07187963e+00 -4.77962345e-02 1.41759121e+00
-3.29996616e-01 -6.84320867e-01 -7.49247968e-01 -2.37732202e-01
-2.96813667e-01 7.88684428e-01 -5.57169199e-01 -3.44879419e-01
-6.49023592e-01 -7.63631284e-01 6.14186764e-01 7.12938368e-01
4.91826564e-01 -9.88365173e-01 -6.95500910e-01 5.21509409e-01
-4.44488585e-01 -1.20787275e+00 -6.98779225e-01 6.70442283e-01
-3.86505783e-01 -5.86597681e-01 -5.89586020e-01 -4.83574092e-01
1.15201078e-01 4.12623972e-01 6.21466875e-01 -3.11648756e-01
-2.82842636e-01 3.80383402e-01 -5.38403451e-01 -4.99883592e-01
-6.79180622e-02 4.56744358e-02 -2.49270955e-03 2.83757895e-01
3.92953344e-02 -7.66475976e-01 -6.70700312e-01 5.22971191e-02
-8.06039453e-01 -2.11958215e-01 4.59266633e-01 7.98787713e-01
1.93016931e-01 3.25676017e-02 6.77218616e-01 -4.57905382e-01
6.20132744e-01 -3.40807438e-01 -2.90075392e-01 -2.06572637e-02
-1.59188807e-01 7.21652806e-02 7.53337741e-01 -6.69318974e-01
-1.14558184e+00 -9.41793472e-02 -3.71630818e-01 -4.20229435e-01
-9.58240926e-02 4.67157811e-01 -1.50317773e-01 1.23024069e-01
7.95542479e-01 3.37299705e-02 -3.87056589e-01 -7.97836840e-01
8.11637938e-02 7.28609681e-01 4.40779984e-01 -3.17332357e-01
6.23879611e-01 7.56780282e-02 -2.28560835e-01 -9.91226375e-01
-9.99985337e-01 -4.91166413e-01 -3.28860551e-01 -3.41912061e-02
9.79771912e-01 -9.87657547e-01 -4.66019005e-01 4.23269004e-01
-1.26451302e+00 -3.20869535e-01 -1.94563150e-01 9.65405166e-01
-1.47976622e-01 1.03269443e-01 -6.42574728e-01 -9.53710139e-01
-1.01235181e-01 -1.13215065e+00 7.56502509e-01 -8.99539888e-02
-2.75566190e-01 -6.73761129e-01 7.35176653e-02 1.36478692e-01
7.56361008e-01 1.99110448e-01 7.38504469e-01 -4.91128862e-01
-4.12648350e-01 -1.15628809e-01 5.39208315e-02 8.17191243e-01
2.61586696e-01 -2.17228010e-01 -1.57443631e+00 -7.62553364e-02
2.36859217e-01 -5.51735759e-02 1.16153061e+00 3.87971818e-01
1.29437709e+00 -1.84676960e-01 1.26888126e-01 7.46743739e-01
1.04179943e+00 4.18228298e-01 6.13081872e-01 2.46190101e-01
8.46268594e-01 4.47306871e-01 2.36458287e-01 5.07657945e-01
2.63628215e-01 7.00626433e-01 3.24151635e-01 -3.53820652e-01
-3.55100572e-01 -4.87621985e-02 5.25435925e-01 1.08497763e+00
-1.21280655e-01 -1.32043421e-01 -7.12526977e-01 6.59484565e-01
-1.45541763e+00 -9.20997858e-01 3.70661058e-02 2.33979249e+00
8.19710255e-01 5.18272035e-02 -1.74062237e-01 6.36571467e-01
5.31408966e-01 4.12029296e-01 -2.31695011e-01 -6.22787297e-01
-6.26035482e-02 5.45208275e-01 3.56817514e-01 5.31033635e-01
-1.15777528e+00 5.19132257e-01 6.00005960e+00 4.72726673e-01
-1.40017033e+00 2.58505881e-01 2.20761463e-01 -3.84421617e-01
-1.65284395e-01 -1.88054994e-01 -4.74731863e-01 4.82392877e-01
1.37126064e+00 4.19040978e-01 8.40337396e-01 4.70932931e-01
3.48274499e-01 1.51441425e-01 -1.08952475e+00 8.84490728e-01
1.20819947e-02 -8.17908168e-01 -4.47893411e-01 -2.06175491e-01
3.95406187e-01 8.09073821e-02 2.64777720e-01 5.04653275e-01
-4.67831157e-02 -9.81929183e-01 9.52531993e-01 3.79517019e-01
4.06728625e-01 -7.29159713e-01 5.47992527e-01 -6.68010339e-02
-1.33015549e+00 -3.45145613e-01 -1.73582152e-01 -1.01781659e-01
7.67360702e-02 6.91107213e-01 -9.22216535e-01 3.89261752e-01
1.01648557e+00 3.45944405e-01 -4.85489368e-01 1.13997948e+00
-1.19289279e-01 9.10959184e-01 -3.70903313e-01 1.73931211e-01
6.06435686e-02 2.21771359e-01 5.07072151e-01 1.49071932e+00
5.08721113e-01 -1.79553494e-01 -9.24933106e-02 6.89865947e-01
-3.35221924e-03 -1.31252632e-02 -2.83389062e-01 9.48506370e-02
5.31437933e-01 1.08292770e+00 -2.83536881e-01 1.36727346e-02
-6.50495768e-01 9.85491097e-01 8.58622566e-02 7.47932374e-01
-1.07190919e+00 -7.60252714e-01 8.30355167e-01 -1.00711137e-02
5.10178387e-01 -3.98077369e-01 -1.05939493e-01 -8.81039262e-01
8.23400840e-02 -6.46317124e-01 2.01193362e-01 -7.24968672e-01
-9.35077369e-01 7.82323301e-01 -3.06707650e-01 -1.15962183e+00
-1.03826866e-01 -3.46654177e-01 -5.68699896e-01 1.07715082e+00
-1.50703979e+00 -8.33142638e-01 -2.27543294e-01 6.55752003e-01
4.98549074e-01 2.61110157e-01 9.63758290e-01 7.87951231e-01
-7.32039273e-01 8.09344172e-01 -1.22264579e-01 1.06350474e-01
1.00943756e+00 -1.18664742e+00 4.50636894e-01 1.10846603e+00
2.40688801e-01 7.74064004e-01 7.90539742e-01 -7.94663355e-02
-1.03528333e+00 -1.40907836e+00 7.85987437e-01 -6.40456975e-02
6.31763995e-01 -7.76849747e-01 -1.17972755e+00 6.42315030e-01
1.36090711e-01 2.87399650e-01 8.95442724e-01 3.09940398e-01
-5.72726429e-01 -3.62725377e-01 -5.84221840e-01 5.95726907e-01
9.00087833e-01 -9.00279403e-01 -5.37285924e-01 8.84194896e-02
1.10012269e+00 -3.62944424e-01 -7.19930112e-01 3.11565220e-01
3.84895861e-01 -8.53496671e-01 1.11261034e+00 -1.02099769e-01
7.05961734e-02 -4.45310473e-01 -4.38005149e-01 -1.48114681e+00
-4.66864318e-01 -6.90617740e-01 -2.33987859e-03 1.40191197e+00
4.15527731e-01 -6.62472844e-01 9.62054636e-03 2.89220363e-01
-6.43026292e-01 -3.49127054e-01 -9.24071074e-01 -7.58925796e-01
-2.53974795e-01 -8.53171527e-01 6.21479809e-01 8.03103924e-01
-2.35973179e-01 3.39267373e-01 -4.70853657e-01 5.32400787e-01
2.99012423e-01 -5.42158037e-02 4.37907845e-01 -1.06874156e+00
-5.60872734e-01 -1.97529659e-01 -7.28044510e-02 -9.82092738e-01
-5.00348322e-02 -6.48100615e-01 4.61234629e-01 -1.29662049e+00
-2.85373837e-01 -2.99359769e-01 -9.01432335e-01 4.99869257e-01
-2.35247448e-01 2.37078413e-01 2.23127306e-01 -2.36942083e-01
-3.91597569e-01 6.90770030e-01 9.49966133e-01 1.28178790e-01
-4.23775315e-01 -5.29824942e-02 -5.89489281e-01 6.66498423e-01
6.76315010e-01 -2.91806340e-01 -3.19502056e-01 -6.57883763e-01
9.11199600e-02 7.26303682e-02 6.08980477e-01 -1.45848382e+00
3.05277258e-01 2.53212869e-01 4.83736724e-01 -4.75572526e-01
6.29226506e-01 -7.21195817e-01 7.57151619e-02 1.45344466e-01
-4.32254523e-01 -1.35285482e-01 5.27502358e-01 4.57444489e-01
-3.35252792e-01 1.31437510e-01 5.49878180e-01 2.41515450e-02
-4.09601539e-01 -1.27988189e-01 -4.60721403e-01 -2.39438891e-01
3.66726458e-01 -7.92826936e-02 -1.84743598e-01 -2.09515303e-01
-7.62918651e-01 -1.97442189e-01 -2.81224884e-02 4.82777357e-01
4.18897510e-01 -1.43690169e+00 -4.85907227e-01 4.04842943e-01
-1.64276987e-01 -4.78802919e-02 6.02310896e-01 7.90021241e-01
-7.44622424e-02 3.40649784e-01 7.22333267e-02 -4.49703991e-01
-9.36622441e-01 4.94312853e-01 4.66748357e-01 4.21960056e-02
-5.29939055e-01 9.52422023e-01 3.96356612e-01 -2.36671120e-01
5.81786454e-01 -9.29329932e-01 -2.37624362e-01 4.55536768e-02
5.47686398e-01 3.77373517e-01 3.65965277e-01 -4.70185876e-01
-4.60389465e-01 4.08719867e-01 1.82640836e-01 -2.53528893e-01
1.33048332e+00 -2.44702995e-01 1.19109705e-01 7.51083970e-01
1.30955887e+00 4.34600711e-01 -1.27068508e+00 -3.22132766e-01
-1.40317678e-01 -1.74881920e-01 3.96206826e-01 -8.38709235e-01
-9.05578375e-01 1.09583139e+00 8.18261683e-01 3.75264913e-01
1.45219040e+00 -3.18010747e-01 9.09508288e-01 2.63075858e-01
1.10632624e-03 -1.07379436e+00 1.63580880e-01 8.07383597e-01
9.78184700e-01 -9.20391202e-01 -3.80356073e-01 -9.99031365e-02
-4.04060781e-01 9.63807106e-01 3.67334634e-01 -2.82492861e-02
7.02779710e-01 3.20096701e-01 5.68836182e-02 2.90108562e-01
-8.11250508e-01 -2.54802674e-01 3.22045475e-01 2.65598953e-01
6.66763008e-01 -9.21970308e-02 2.91159034e-01 8.08084130e-01
-4.58303124e-01 -4.72965747e-01 4.40059394e-01 8.40717494e-01
-3.45183283e-01 -8.93622816e-01 -6.90444052e-01 -1.13805411e-02
-7.47970819e-01 -4.25351530e-01 -2.08436027e-02 3.30130816e-01
2.81995803e-01 1.20865285e+00 2.19158947e-01 -5.87923586e-01
6.15992010e-01 3.60313863e-01 3.21493179e-01 -6.27260745e-01
-8.35265279e-01 5.11130631e-01 -4.15156074e-02 -5.64070225e-01
-2.43723482e-01 -5.63758910e-01 -1.25552332e+00 2.05392301e-01
-3.57750595e-01 2.36539468e-02 7.64653206e-01 7.23934233e-01
3.44206214e-01 1.50600195e+00 6.21210456e-01 -9.64913964e-01
-3.53230745e-01 -1.19516218e+00 -4.02279258e-01 3.46134961e-01
8.78682733e-01 -5.43489695e-01 -6.52175069e-01 -2.43283343e-02] | [15.18979263305664, 5.4107561111450195] |
b1e2ad5a-b0f5-452f-b0c4-4d42fc359ba0 | nanopublication-based-semantic-publishing-and | 2203.01608 | null | https://arxiv.org/abs/2203.01608v1 | https://arxiv.org/pdf/2203.01608v1.pdf | Nanopublication-Based Semantic Publishing and Reviewing: A Field Study with Formalization Papers | With the rapidly increasing amount of scientific literature,it is getting continuously more difficult for researchers in different disciplines to be updated with the recent findings in their field of study.Processing scientific articles in an automated fashion has been proposed as a solution to this problem,but the accuracy of such processing remains very poor for extraction tasks beyond the basic ones.Few approaches have tried to change how we publish scientific results in the first place,by making articles machine-interpretable by expressing them with formal semantics from the start.In the work presented here,we set out to demonstrate that we can formally publish high-level scientific claims in formal logic,and publish the results in a special issue of an existing journal.We use the concept and technology of nanopublications for this endeavor,and represent not just the submissions and final papers in this RDF-based format,but also the whole process in between,including reviews,responses,and decisions.We do this by performing a field study with what we call formalization papers,which contribute a novel formalization of a previously published claim.We received 15 submissions from 18 authors,who then went through the whole publication process leading to the publication of their contributions in the special issue.Our evaluation shows the technical and practical feasibility of our approach.The participating authors mostly showed high levels of interest and confidence,and mostly experienced the process as not very difficult,despite the technical nature of the current user interfaces.We believe that these results indicate that it is possible to publish scientific results from different fields with machine-interpretable semantics from the start,which in turn opens countless possibilities to radically improve in the future the effectiveness and efficiency of the scientific endeavor as a whole. | ['Jacco van Ossenbruggen', 'Davide Ceolin', 'Tobias Kuhn', 'Cristina-Iulia Bucur'] | 2022-03-03 | null | null | null | null | ['formal-logic'] | ['reasoning'] | [-8.81628022e-02 5.28361857e-01 -6.50898814e-02 -2.05248192e-01
-3.30963403e-01 -8.58120382e-01 7.54933357e-01 6.77894354e-01
-4.58638757e-01 1.02138269e+00 2.03863814e-01 -6.58101439e-01
-4.02816951e-01 -8.73807251e-01 -9.20778334e-01 1.52676746e-01
2.19273895e-01 5.53217292e-01 4.46075886e-01 -1.52103946e-01
7.60811746e-01 4.86301541e-01 -1.43289638e+00 2.21225649e-01
1.01272869e+00 5.77733874e-01 1.07417017e-01 1.48507506e-01
-6.36244535e-01 7.51667619e-01 -7.26678967e-01 -9.97621775e-01
-4.87656146e-02 -6.06448501e-02 -1.09623802e+00 -2.88644105e-01
1.66642398e-01 2.16886878e-01 2.06753314e-01 1.11183286e+00
8.57164618e-03 -3.92345548e-01 2.18850508e-01 -1.20878589e+00
-6.04227602e-01 1.05125535e+00 -3.27099919e-01 -9.60193872e-02
7.08325684e-01 -7.27414861e-02 9.41404998e-01 -6.62159920e-01
1.08995867e+00 1.03178394e+00 5.12768865e-01 3.61006886e-01
-9.92839754e-01 -5.84686041e-01 1.38096422e-01 2.72322953e-01
-1.11047757e+00 -3.62073362e-01 7.21970320e-01 -7.44161487e-01
8.36939156e-01 4.96790588e-01 7.04352975e-01 8.49790156e-01
1.96440116e-01 2.91149408e-01 1.35459054e+00 -7.41995275e-01
3.46450239e-01 5.68241060e-01 3.34747404e-01 4.08954799e-01
1.22400045e+00 -5.77959120e-01 -4.58845794e-01 -2.49052450e-01
2.85419554e-01 -5.49959987e-02 -2.91387707e-01 -5.86898364e-02
-1.36898434e+00 3.51957977e-01 -1.20612033e-01 9.42915618e-01
-2.68737912e-01 7.10311234e-02 3.50530744e-01 2.85982341e-01
3.69817048e-01 8.26439798e-01 -4.45544839e-01 -3.68098646e-01
-1.20462191e+00 4.85320956e-01 1.39463854e+00 8.75507951e-01
3.47087026e-01 -5.47081947e-01 5.75380698e-02 1.76431805e-01
3.67415994e-01 3.70788276e-01 -4.79106297e-04 -1.04014575e+00
1.98219016e-01 8.59126091e-01 4.59839165e-01 -1.24958944e+00
1.21783264e-01 -3.77992660e-01 -3.34406853e-01 2.80767441e-01
3.73205543e-01 1.01474807e-01 -3.24930668e-01 1.35109031e+00
5.86539656e-02 -3.57911229e-01 -4.62542512e-02 7.69488215e-01
8.46979260e-01 5.26986301e-01 1.52852014e-01 -3.93878311e-01
1.56941664e+00 -3.90898585e-01 -1.16120768e+00 3.49862695e-01
4.43200678e-01 -1.06660128e+00 9.06391084e-01 8.12713385e-01
-1.38011289e+00 -1.17906004e-01 -1.17436266e+00 -5.77573739e-02
-6.03730321e-01 -1.51568234e-01 6.67880952e-01 4.78022426e-01
-9.07485247e-01 7.80998409e-01 -5.97490251e-01 -4.63256121e-01
4.37003106e-01 1.26415297e-01 -3.31419170e-01 1.47247650e-02
-1.20045507e+00 1.03026497e+00 1.21319838e-01 3.26499902e-02
1.43938839e-01 -7.67336488e-01 -2.87751466e-01 2.55295988e-02
8.79631758e-01 -7.29504466e-01 1.01814330e+00 -6.00977182e-01
-1.17167711e+00 9.65674400e-01 -2.87244171e-01 -4.04437661e-01
7.53252447e-01 -1.26544446e-01 -5.27324617e-01 -7.36756437e-03
2.56082505e-01 8.59494805e-02 1.09351933e-01 -1.32037210e+00
-5.96606433e-01 -3.07996780e-01 4.16405022e-01 -5.19879520e-01
-3.22424054e-01 3.11068714e-01 -3.67271334e-01 -4.83198196e-01
-2.46255457e-01 -4.85376924e-01 2.20293161e-02 -6.22601472e-02
-1.74636289e-01 -4.85074729e-01 5.48612356e-01 -5.42748690e-01
1.52668107e+00 -1.77734053e+00 1.89893633e-01 2.93260396e-01
4.50211465e-01 2.64206767e-01 4.70888942e-01 7.65970290e-01
2.58376092e-01 1.02476108e+00 -1.94429502e-01 -1.19544826e-01
2.64801085e-01 9.93925780e-02 -6.94456995e-01 1.66024402e-01
1.93461150e-01 7.97810555e-01 -1.04476225e+00 -4.90557164e-01
-1.94371536e-01 2.37086847e-01 -2.18558684e-01 -2.20742509e-01
-5.13961971e-01 -5.31651378e-02 -5.92333496e-01 4.49230224e-01
6.97814703e-01 -2.95926452e-01 4.88549471e-01 -2.29580820e-01
-7.20869660e-01 3.70059818e-01 -1.37749851e+00 1.41031468e+00
-2.80690700e-01 5.23347259e-01 2.00326681e-01 -9.11770225e-01
8.73633444e-01 3.77576262e-01 4.54605490e-01 -5.88566482e-01
-2.29544923e-01 6.74309909e-01 -2.22458526e-01 -5.06937146e-01
6.38701439e-01 -1.82885066e-01 -9.73679721e-02 6.76669180e-01
-2.47840479e-01 -3.45313936e-01 8.32333207e-01 4.51592952e-01
8.96456778e-01 5.51499546e-01 2.01005712e-01 -3.68249506e-01
6.59722686e-01 2.93783844e-01 3.70301127e-01 6.81536138e-01
2.28986070e-01 3.31091553e-01 8.68537962e-01 -4.89397138e-01
-1.14903116e+00 -6.41112804e-01 -2.05693528e-01 1.80688620e-01
-1.26024812e-01 -1.06462646e+00 -6.16236866e-01 -5.61566651e-01
1.48221999e-02 7.13708997e-01 -5.25146842e-01 3.24151695e-01
-3.52646798e-01 -2.13863552e-01 3.63275677e-01 -1.74800176e-02
2.25119010e-01 -9.33336914e-01 -8.05149674e-01 2.60714918e-01
-2.28767097e-02 -1.15059268e+00 2.51534790e-01 -2.00744182e-01
-6.66439116e-01 -1.16822171e+00 -5.65029502e-01 -1.98429435e-01
5.61329305e-01 -2.43973225e-01 1.02927208e+00 4.25935358e-01
-1.35995105e-01 1.20103568e-01 -4.27879870e-01 -9.00058210e-01
-6.20757401e-01 -1.05890520e-01 -1.24263406e-01 -3.67486894e-01
3.22109610e-01 -3.53708684e-01 -1.91100493e-01 -5.44050448e-02
-1.30375648e+00 1.31953657e-01 3.50088924e-01 3.27028900e-01
4.08868313e-01 2.97064837e-02 6.02654278e-01 -1.21053278e+00
7.86468506e-01 -3.92344087e-01 -7.70165205e-01 5.81829667e-01
-1.08490372e+00 1.74353957e-01 6.78757966e-01 2.26705685e-01
-7.53520608e-01 -6.32737041e-01 6.66269287e-02 1.12553120e-01
-4.81734276e-02 1.07521594e+00 -1.00867264e-01 2.83611417e-01
4.28368002e-01 -3.11920881e-01 7.51205683e-02 -5.24166465e-01
2.31006309e-01 7.03895867e-01 3.04285824e-01 -8.00609350e-01
7.52193391e-01 4.18113232e-01 3.50151032e-01 -4.32488501e-01
-6.00046933e-01 -1.39128447e-01 -1.75557867e-01 -2.18063310e-01
4.27282155e-01 -4.30560499e-01 -9.33741808e-01 5.10681383e-02
-1.42528582e+00 1.36387780e-01 -6.29392862e-01 4.43090051e-01
-1.77297425e-02 5.25680125e-01 -1.32959098e-01 -7.99397588e-01
-2.48614445e-01 -9.65820730e-01 6.60161495e-01 1.25349641e-01
-5.96542418e-01 -9.87100661e-01 -1.91283654e-02 5.53923965e-01
4.38152313e-01 2.99681097e-01 9.35124218e-01 -5.71754694e-01
-5.92008233e-01 -2.21149847e-01 -3.28271300e-01 1.55479342e-01
-2.71806326e-02 6.21333182e-01 -7.54426301e-01 1.61267534e-01
-6.35368228e-02 1.78325668e-01 4.09932911e-01 -5.01712970e-02
1.05372560e+00 -3.95177275e-01 -3.20016056e-01 -1.34671196e-01
1.46442211e+00 1.44431228e-02 8.85298252e-01 7.66893446e-01
4.32702094e-01 1.02527821e+00 5.53464890e-01 2.30687335e-01
4.18253034e-01 7.39258051e-01 1.38751119e-01 2.22237572e-01
-1.56904638e-01 -1.26751080e-01 1.61930948e-01 8.09334517e-01
-5.28632522e-01 -2.56957531e-01 -1.10612881e+00 5.39519608e-01
-2.00735950e+00 -1.14482069e+00 -5.70499361e-01 2.20884323e+00
9.84231889e-01 4.19392169e-01 -1.00441813e-01 1.55506894e-01
4.76453573e-01 -1.27554700e-01 2.56125987e-01 -7.51635194e-01
-4.80324402e-02 3.55018586e-01 3.24139714e-01 5.14655054e-01
-4.34159756e-01 5.60509086e-01 5.95055819e+00 5.81656396e-01
-1.29440713e+00 -3.54248993e-02 2.25377589e-01 3.67632397e-02
-1.05027163e+00 7.21656978e-01 -5.89858234e-01 5.86597383e-01
1.15587723e+00 -6.52926683e-01 2.77637571e-01 3.56092840e-01
5.33158064e-01 -3.10801178e-01 -1.01005769e+00 4.86308157e-01
-2.05753729e-01 -1.98826492e+00 2.37175524e-01 1.10501267e-01
4.18982267e-01 -3.17162752e-01 -4.53713447e-01 -1.33003816e-01
5.32837845e-02 -1.09683490e+00 1.11549520e+00 1.11564195e+00
6.21882915e-01 -4.45020109e-01 9.69737411e-01 1.55355558e-01
-5.22333086e-01 1.97899044e-01 -3.19269188e-02 -4.78556514e-01
2.18994766e-01 1.19536996e+00 -8.20464373e-01 1.30406713e+00
7.28840113e-01 6.81407928e-01 -3.56751144e-01 1.08330870e+00
-1.59265235e-01 6.84904337e-01 -3.08204770e-01 -5.12335360e-01
-2.07846358e-01 -1.67675048e-01 6.29458964e-01 1.30127037e+00
3.73039603e-01 -1.42368317e-01 -2.04403237e-01 1.32230914e+00
-2.72681713e-01 2.34947294e-01 -4.66482550e-01 -5.36125541e-01
1.97553456e-01 1.10976171e+00 -8.59418452e-01 -3.72327000e-01
-3.09306294e-01 1.95915267e-01 -1.27674909e-02 1.71203613e-01
-4.14504379e-01 -6.59544826e-01 -5.91566116e-02 5.88400483e-01
2.64580220e-01 -2.19874844e-01 -6.74273312e-01 -1.24106097e+00
5.77356994e-01 -7.73398578e-01 1.41649246e-01 -1.01158917e+00
-1.08131552e+00 4.37650800e-01 2.80218422e-01 -8.96701396e-01
-9.44657996e-02 -5.78778625e-01 -2.90242761e-01 9.64024544e-01
-1.42325354e+00 -8.59794259e-01 1.30888388e-01 -2.18803629e-01
-1.22922555e-01 2.09474072e-01 7.52646148e-01 1.97483510e-01
-1.20840073e-01 -3.93228494e-02 -2.05858991e-01 -1.35472238e-01
8.65489900e-01 -1.05637550e+00 -8.32245424e-02 9.10068691e-01
2.06952184e-01 1.27350807e+00 1.05491793e+00 -8.39854598e-01
-1.58027184e+00 -4.15975839e-01 1.80991769e+00 -5.96191466e-01
1.11610973e+00 -3.00489992e-01 -9.63422358e-01 2.56216019e-01
5.25020182e-01 -4.19418514e-01 4.80388194e-01 2.07350329e-01
-2.32862711e-01 -3.10846120e-01 -1.03339517e+00 6.20322287e-01
8.05164576e-01 -3.56585175e-01 -1.10688245e+00 3.03566366e-01
6.36136591e-01 -1.38394564e-01 -1.04558647e+00 3.87004226e-01
5.34393072e-01 -7.07279205e-01 3.88263881e-01 -4.04231578e-01
6.79040611e-01 -6.86857522e-01 1.59834132e-01 -8.37511659e-01
6.03554808e-02 -7.66903520e-01 6.55294582e-02 1.40258372e+00
5.87808371e-01 -7.44728267e-01 2.37438783e-01 8.08146179e-01
-1.32059917e-01 -8.57619524e-01 -6.79987907e-01 -4.56043601e-01
1.61846384e-01 -6.59815609e-01 6.79924846e-01 9.64055419e-01
4.60947514e-01 1.40014559e-01 2.33841881e-01 -3.31814289e-01
4.41731393e-01 4.94377494e-01 6.86484158e-01 -1.76296043e+00
-3.25719640e-02 -5.89475632e-01 -3.23168576e-01 -2.50744164e-01
-2.68370546e-02 -1.01314843e+00 -5.68743408e-01 -2.21369410e+00
2.59013146e-01 -4.63667214e-01 -1.49062321e-01 5.80169261e-01
2.42127657e-01 -8.61296430e-03 1.56813249e-01 4.43192542e-01
-5.38207114e-01 -2.42218167e-01 1.25946641e+00 -1.41995847e-01
1.10041507e-01 -2.83843398e-01 -1.21445715e+00 5.35739958e-01
3.33237410e-01 -4.75228637e-01 -1.50724933e-01 -1.59716651e-01
9.88808572e-01 -3.51742029e-01 6.06449902e-01 -7.58618295e-01
4.33306634e-01 -3.58247191e-01 6.88829422e-02 -3.12223554e-01
-4.08834487e-01 -9.58758771e-01 6.46511853e-01 3.45894337e-01
-2.10604861e-01 -7.55533427e-02 3.12539697e-01 1.66266367e-01
-2.15707019e-01 -5.88680983e-01 -2.02221964e-02 -1.62228376e-01
-4.73237097e-01 -1.85015708e-01 -1.78786948e-01 4.96543683e-02
9.60617006e-01 -1.67782083e-01 -5.66565990e-01 -8.52591619e-02
-5.33520639e-01 -2.78408988e-03 8.54447484e-01 2.69293129e-01
2.39491612e-01 -5.90142250e-01 -6.60139740e-01 -2.58176237e-01
-7.27050230e-02 -4.43794690e-02 -1.73503548e-01 1.02239418e+00
-7.35423148e-01 6.92921340e-01 -2.22827822e-01 -1.84819415e-01
-9.65773284e-01 5.62228084e-01 -1.53568670e-01 -4.10864711e-01
-5.44113219e-01 2.90024839e-02 -5.66902339e-01 5.96246049e-02
1.31193576e-02 -6.00349844e-01 -4.15983856e-01 1.95933580e-01
4.61182445e-01 2.62244284e-01 3.74296933e-01 -3.39727312e-01
-5.36717594e-01 4.47714418e-01 -5.31143583e-02 -4.48420107e-01
1.66447771e+00 3.83765846e-02 -7.83475637e-01 7.24995792e-01
5.74266911e-01 9.14545000e-01 -5.28220594e-01 3.44062537e-01
2.44607925e-01 -3.97801787e-01 -1.00463115e-01 -1.25246131e+00
-6.03675783e-01 5.79909861e-01 -2.12767929e-01 7.27356017e-01
6.20828032e-01 1.34542063e-01 5.21438062e-01 2.57967621e-01
6.25534296e-01 -1.07133400e+00 -6.54653788e-01 2.72407562e-01
9.73088622e-01 -7.40552664e-01 6.04616046e-01 -5.94832242e-01
-2.19534069e-01 1.51593125e+00 -5.20693697e-02 1.89994037e-01
4.39473271e-01 5.45069695e-01 -1.29623353e-01 -4.73779798e-01
-6.45339906e-01 3.10043305e-01 1.61051482e-01 2.37294480e-01
7.35503078e-01 -1.85733251e-02 -1.22522581e+00 8.35163653e-01
-4.20091003e-02 8.01385283e-01 9.99478281e-01 1.21693993e+00
-2.67299533e-01 -1.73777056e+00 -4.30175066e-01 2.47385606e-01
-9.85101640e-01 6.29446805e-02 -7.13836193e-01 8.63797545e-01
2.38316134e-01 9.70292211e-01 -2.43641317e-01 -4.34520356e-02
6.66853666e-01 1.40632257e-01 5.52114725e-01 -6.17032945e-01
-7.98373938e-01 -2.93412060e-01 4.25257295e-01 -2.11639360e-01
-8.13856423e-01 -6.68814659e-01 -1.62629378e+00 -6.10209644e-01
-7.37325400e-02 8.37610483e-01 1.09230673e+00 1.04037416e+00
5.86500764e-01 5.18026114e-01 -5.34124933e-02 -1.85237378e-01
-2.97886193e-01 -5.64184606e-01 -2.70220876e-01 3.80127221e-01
2.83494219e-02 -6.74332976e-01 -3.64944905e-01 2.57733554e-01] | [9.429861068725586, 8.169365882873535] |
e9622451-dd54-44a7-9cff-9555f65c46e5 | local-and-global-contextual-features-fusion | 2305.01111 | null | https://arxiv.org/abs/2305.01111v1 | https://arxiv.org/pdf/2305.01111v1.pdf | Local and Global Contextual Features Fusion for Pedestrian Intention Prediction | Autonomous vehicles (AVs) are becoming an indispensable part of future transportation. However, safety challenges and lack of reliability limit their real-world deployment. Towards boosting the appearance of AVs on the roads, the interaction of AVs with pedestrians including "prediction of the pedestrian crossing intention" deserves extensive research. This is a highly challenging task as involves multiple non-linear parameters. In this direction, we extract and analyse spatio-temporal visual features of both pedestrian and traffic contexts. The pedestrian features include body pose and local context features that represent the pedestrian's behaviour. Additionally, to understand the global context, we utilise location, motion, and environmental information using scene parsing technology that represents the pedestrian's surroundings, and may affect the pedestrian's intention. Finally, these multi-modality features are intelligently fused for effective intention prediction learning. The experimental results of the proposed model on the JAAD dataset show a superior result on the combined AUC and F1-score compared to the state-of-the-art. | ['Chenghao Qian', 'Tanveer Hussain', 'Mahdi Rezaei', 'Mohsen Azarmi'] | 2023-05-01 | null | null | null | null | ['scene-parsing'] | ['computer-vision'] | [ 1.87301952e-02 -2.10208073e-01 -3.23283613e-01 -7.23818898e-01
-4.48803455e-01 -6.92681149e-02 8.90605390e-01 1.20143734e-01
-6.04038179e-01 5.65973043e-01 2.32625023e-01 -2.77139008e-01
2.03509361e-01 -7.92416513e-01 -5.69704533e-01 -7.69456685e-01
-1.91878881e-02 -1.73627958e-02 7.01544940e-01 -3.80855322e-01
2.59982526e-01 4.31425005e-01 -2.04948378e+00 2.23609373e-01
8.06490541e-01 1.13648558e+00 2.45413646e-01 5.76039135e-01
1.79334491e-01 5.31193435e-01 -2.94171106e-02 -5.48637509e-01
2.17555109e-02 1.81698650e-02 -5.05814791e-01 4.45705839e-02
5.66862345e-01 -2.17688233e-01 -3.53760689e-01 8.07260334e-01
3.62390727e-01 2.92198300e-01 5.90404630e-01 -1.47213292e+00
-1.12319797e-01 -2.56433427e-01 -3.72447848e-01 3.32151562e-01
4.56870347e-01 6.17467582e-01 6.46476924e-01 -5.81307590e-01
3.67853850e-01 1.31874681e+00 5.42200089e-01 3.71304005e-01
-7.80364931e-01 -4.52070087e-01 4.60160464e-01 9.86593306e-01
-1.20638847e+00 -6.14994407e-01 8.88609648e-01 -5.73090315e-01
5.94859600e-01 4.70434278e-01 7.81909287e-01 1.30769658e+00
2.26803869e-01 1.15780020e+00 9.92482662e-01 -2.42596313e-01
1.17001988e-01 3.78786087e-01 3.14918876e-01 6.26629710e-01
2.13329326e-02 2.92676955e-01 -5.07028878e-01 3.65723014e-01
-1.08601920e-01 -1.60378575e-01 2.90238768e-01 -3.90255809e-01
-9.31526601e-01 7.07256436e-01 5.25391519e-01 -2.01939762e-01
-5.13520420e-01 -2.58524586e-02 4.98632669e-01 -3.84335309e-01
3.42253506e-01 -5.53401113e-01 -3.56947839e-01 -2.42286086e-01
-4.73677933e-01 3.21194768e-01 2.21548095e-01 9.50976789e-01
7.29695976e-01 -1.89689413e-01 -4.28939648e-02 6.93503618e-01
8.16789687e-01 8.05975974e-01 -7.99016841e-03 -6.09099030e-01
6.03003919e-01 7.83108592e-01 6.54699802e-02 -1.16509390e+00
-6.26880109e-01 -2.87709951e-01 -7.51394331e-01 1.45970717e-01
5.16355515e-01 -9.86265913e-02 -8.16867650e-01 1.48058569e+00
8.35005105e-01 3.24955493e-01 -9.38200802e-02 9.98686016e-01
1.07478714e+00 6.05823994e-01 9.81180608e-01 -6.10786676e-02
1.77389848e+00 -9.63799655e-01 -6.27124667e-01 -7.09481716e-01
3.80549222e-01 -6.69869840e-01 6.21833205e-01 -2.67809276e-02
-8.33880603e-01 -1.08365202e+00 -7.68767118e-01 3.50362323e-02
-9.25298035e-01 3.76564533e-01 5.90642333e-01 8.69172156e-01
-6.28059149e-01 -1.55930310e-01 -6.50256574e-01 -6.98027551e-01
3.60275567e-01 3.75574589e-01 -3.68099898e-01 -2.22095549e-01
-1.42501760e+00 1.19116354e+00 2.55210757e-01 3.11630100e-01
-5.61523199e-01 -4.08764124e-01 -1.06467986e+00 -3.96351695e-01
3.36809427e-01 -7.62384236e-01 8.81810248e-01 -6.60947084e-01
-1.09295583e+00 7.18008876e-01 -5.27045846e-01 -4.49057668e-01
5.30231833e-01 -5.13143167e-02 -7.97304571e-01 -4.64711078e-02
2.54152000e-01 9.36838329e-01 7.19161212e-01 -1.32899904e+00
-1.37713122e+00 -5.20425975e-01 -5.33213280e-02 4.26865071e-01
1.11898176e-01 4.68516089e-02 -5.69119871e-01 6.09317869e-02
-2.58900315e-01 -1.01317108e+00 -2.21606985e-01 3.17988917e-02
-3.03529024e-01 -4.80707407e-01 1.23834085e+00 -8.89546216e-01
1.00854576e+00 -2.15409589e+00 -2.38323078e-01 1.91519454e-01
-2.49398034e-02 6.18378043e-01 4.45608310e-02 2.26140589e-01
3.02595198e-01 -2.06588268e-01 1.46786705e-01 -1.86346754e-01
7.93090612e-02 1.11327104e-01 3.80513072e-01 4.40888613e-01
2.69016296e-01 9.74456370e-01 -7.59922862e-01 -8.78001392e-01
8.98790836e-01 6.13970935e-01 -2.19764709e-01 1.28862187e-01
-3.71709354e-02 6.80194616e-01 -6.37676179e-01 5.80424488e-01
7.81536639e-01 3.39699984e-01 -1.57464504e-01 -3.56231511e-01
-4.13590491e-01 -4.92694490e-02 -1.05196929e+00 1.00593412e+00
-3.19096982e-01 6.28288150e-01 -8.37567672e-02 -9.54459429e-01
8.46143246e-01 1.75934359e-02 3.31693143e-01 -1.16663718e+00
1.29191175e-01 -1.45042568e-01 -5.71819432e-02 -1.10407650e+00
5.84199667e-01 2.74097532e-01 -4.93882112e-02 -4.64888424e-01
-3.67866397e-01 6.21616006e-01 2.98625976e-01 -1.08332254e-01
5.34703910e-01 3.41206998e-01 3.86504173e-01 9.04582143e-02
1.13415611e+00 4.89197336e-02 4.32654202e-01 4.71187413e-01
-7.92331517e-01 4.65111509e-02 1.61518574e-01 -5.98475635e-01
-7.44010389e-01 -9.18470621e-01 -1.31315172e-01 1.12392259e+00
5.49125731e-01 -1.59127582e-02 -6.00376546e-01 -7.84948528e-01
-1.05071098e-01 1.04130149e+00 -6.13041580e-01 -2.59109318e-01
-5.37913263e-01 -6.96320593e-01 2.42135733e-01 6.51740670e-01
8.81384015e-01 -8.48910987e-01 -5.89379132e-01 6.51054308e-02
-4.76638943e-01 -1.47313762e+00 -2.89468437e-01 -5.04615903e-01
-2.27781087e-01 -1.07720613e+00 -3.63451868e-01 -7.55566537e-01
3.24401200e-01 4.51963007e-01 7.02540815e-01 3.10526732e-02
-3.20885897e-01 5.62402666e-01 -2.39517465e-01 -6.31275475e-01
-3.07804346e-01 -3.49027552e-02 -9.04289335e-02 5.76624990e-01
7.51279354e-01 -1.81678310e-01 -9.88549113e-01 6.49582684e-01
-1.22633517e-01 3.37397337e-01 6.84219778e-01 3.83604705e-01
3.85100216e-01 2.53229916e-01 5.25740623e-01 -3.70584816e-01
7.25102872e-02 -6.94187284e-01 -3.48445207e-01 1.84196383e-01
-2.51177639e-01 -4.58409190e-01 2.96221435e-01 -2.17639908e-01
-1.60051906e+00 3.21209550e-01 -4.54135031e-01 2.99329430e-01
-1.05162907e+00 1.29442126e-01 -6.70658290e-01 8.72296542e-02
1.82387248e-01 3.39400440e-01 -2.34843660e-02 -2.29887560e-01
5.20045102e-01 8.76406133e-01 5.00270188e-01 -2.45746925e-01
5.00498295e-01 5.36119699e-01 3.22283983e-01 -1.37991059e+00
-5.96866429e-01 -9.53774750e-01 -8.79462719e-01 -9.45878506e-01
1.29791546e+00 -9.23081696e-01 -1.19987667e+00 5.39773464e-01
-1.17004514e+00 1.22463770e-01 2.39773914e-01 6.43084645e-01
-4.49280322e-01 4.07014728e-01 -8.19051079e-03 -1.34749269e+00
-3.68906781e-02 -1.29740763e+00 1.21136820e+00 5.55242121e-01
-4.98442166e-02 -9.87820923e-01 -3.23873311e-01 1.12696838e+00
2.86705464e-01 2.10978895e-01 6.23214483e-01 -4.07609344e-01
-7.37182200e-01 -4.32372451e-01 -5.00152707e-01 3.63375023e-02
-2.67043561e-01 1.02777500e-02 -9.00942147e-01 2.50528991e-01
-5.71466267e-01 2.39232510e-01 5.41112244e-01 6.63324773e-01
6.71192229e-01 -9.23797265e-02 -6.79451704e-01 8.80364701e-02
1.10030758e+00 4.15088475e-01 7.64537156e-01 4.67762381e-01
8.27100098e-01 1.09512222e+00 1.13101614e+00 3.65006655e-01
1.07169902e+00 9.24937725e-01 7.15071440e-01 -9.77919847e-02
-3.78631592e-01 -2.52640575e-01 2.98199385e-01 3.47989023e-01
-3.60998780e-01 -1.62133738e-01 -9.66416001e-01 6.64344907e-01
-2.05622196e+00 -1.26315391e+00 -1.02790129e+00 1.95057309e+00
7.29354843e-02 1.55183032e-01 5.18034160e-01 7.52228405e-03
8.00577819e-01 9.61717516e-02 -2.81086802e-01 -4.59071606e-01
7.96803162e-02 -4.64680433e-01 5.86656868e-01 3.85470778e-01
-1.59243762e+00 9.88730848e-01 5.20158386e+00 8.83267224e-01
-8.98805082e-01 9.62784141e-02 7.25147128e-01 4.38206643e-01
1.39979094e-01 -2.44065568e-01 -1.16987920e+00 7.81078875e-01
1.20027423e+00 4.37747449e-01 -2.32448708e-02 7.97369003e-01
7.68918633e-01 -5.25240719e-01 -7.94013500e-01 6.30066574e-01
3.36046889e-02 -8.79615903e-01 -3.09883237e-01 1.21321663e-01
3.65550935e-01 -1.61732718e-01 9.21054110e-02 4.90322620e-01
-1.66402191e-01 -6.88870251e-01 7.36665189e-01 9.94419754e-01
2.56037891e-01 -8.56285453e-01 9.34823632e-01 5.50482810e-01
-1.63960445e+00 -1.71441838e-01 -4.85134758e-02 1.16613157e-01
6.44374013e-01 5.84447384e-02 -7.07693160e-01 6.33267701e-01
7.12127924e-01 7.27245629e-01 -1.01669765e+00 1.26967680e+00
-3.91232483e-02 6.45618618e-01 -1.68409765e-01 -3.06677848e-01
3.47931027e-01 -2.81849891e-01 5.89631498e-01 1.40556502e+00
6.49481565e-02 5.56249134e-02 3.49470645e-01 3.01119506e-01
5.76897085e-01 3.14096272e-01 -8.37554634e-01 5.03373742e-01
8.71972591e-02 1.17741024e+00 -3.21257025e-01 -2.76495397e-01
-8.82839143e-01 4.17549670e-01 -5.95428348e-02 2.61359990e-01
-1.09347093e+00 1.67692810e-01 8.66037488e-01 2.65152484e-01
2.96524227e-01 -1.01757377e-01 -4.53273147e-01 -5.98659158e-01
1.45136207e-01 -4.79229063e-01 3.12575191e-01 -8.10311794e-01
-1.07142162e+00 2.34432846e-01 2.29413882e-01 -1.42166507e+00
-5.16730323e-02 -6.61366761e-01 -6.22811496e-01 7.32287109e-01
-1.63559175e+00 -2.03950882e+00 -4.09660071e-01 5.15945613e-01
8.52605224e-01 -2.07809553e-01 3.79702032e-01 5.37696242e-01
-6.89562976e-01 4.58498687e-01 -4.30228889e-01 -2.06581946e-03
2.42993444e-01 -8.16581786e-01 3.22897524e-01 9.01858807e-01
-4.05596495e-01 1.39113158e-01 8.72476399e-01 -7.44929433e-01
-1.30850756e+00 -1.26499271e+00 1.12134159e+00 -4.91282701e-01
4.65682626e-01 5.26449457e-03 -4.86062586e-01 4.75221992e-01
1.68605298e-01 3.11358708e-05 6.17733181e-01 7.05178529e-02
1.49322122e-01 -3.14765006e-01 -1.13757312e+00 8.64606500e-01
1.06452680e+00 -1.63653985e-01 -3.19601685e-01 1.16264090e-01
2.40551293e-01 -1.52934492e-01 -6.14470840e-01 5.10401368e-01
8.57720017e-01 -1.02990735e+00 1.40663373e+00 -7.00324237e-01
8.31162855e-02 -4.03513998e-01 -3.22882682e-01 -7.88665354e-01
-3.13064694e-01 4.88178134e-02 1.53917409e-02 1.35131228e+00
2.26022452e-01 -4.85429287e-01 7.64215410e-01 9.09723163e-01
-1.02373369e-01 -6.91265702e-01 -1.03202069e+00 -2.97215074e-01
-3.49395156e-01 -7.98033297e-01 4.22323406e-01 4.11785364e-01
-2.24439412e-01 5.05415797e-01 -6.55482233e-01 5.94014823e-01
8.37428331e-01 -3.09118330e-01 1.00408828e+00 -1.26816511e+00
4.35288996e-01 -4.22502369e-01 -1.03657186e+00 -1.04911101e+00
1.16544001e-01 -4.94738132e-01 7.94916824e-02 -1.55124819e+00
1.56910434e-01 -3.04258287e-01 -1.68264203e-04 1.50212944e-01
-4.03303266e-01 8.40377882e-02 1.24220528e-01 -3.09929520e-01
-8.40524852e-01 6.14698589e-01 1.25459886e+00 -3.70954007e-01
3.52847688e-02 5.42841077e-01 -3.03204209e-01 7.78621256e-01
8.03418994e-01 5.40901572e-02 -3.41295302e-01 9.80039462e-02
-2.63585091e-01 1.49800956e-01 7.58329868e-01 -1.08823001e+00
3.40527266e-01 -4.85647917e-01 3.99045885e-01 -1.13570058e+00
7.21887469e-01 -1.11610913e+00 1.92629895e-03 3.24568629e-01
-1.12875678e-01 -1.78321809e-01 1.28966063e-01 8.80916595e-01
1.27216754e-02 -9.29377787e-03 7.79702306e-01 1.23980336e-01
-1.39310288e+00 3.39251101e-01 -7.42954075e-01 -2.62565613e-01
1.55068183e+00 -5.08068264e-01 -3.23990524e-01 -3.22283864e-01
-6.89839840e-01 3.88352692e-01 -5.21074198e-02 7.37979889e-01
6.74040139e-01 -1.25824666e+00 -5.96549094e-01 2.80276626e-01
3.47312540e-01 -2.15803102e-01 7.13166475e-01 1.10920048e+00
-1.94652304e-01 6.39061332e-01 -2.61159629e-01 -9.19996619e-01
-1.74108112e+00 5.57803929e-01 1.89302310e-01 1.37127444e-01
-3.29741985e-01 4.39721793e-01 1.27667472e-01 -4.05801445e-01
1.46400213e-01 2.04358593e-01 -8.66068542e-01 9.46159437e-02
3.82335752e-01 7.83212543e-01 -2.05369979e-01 -1.61610198e+00
-5.80078065e-01 5.92604816e-01 7.63749704e-02 1.67114735e-01
9.27376032e-01 -7.88555682e-01 3.50571275e-01 1.88084468e-01
1.07671750e+00 -3.25043648e-01 -1.29828751e+00 -1.08264156e-01
9.50313359e-02 -5.38904846e-01 7.62207294e-03 -7.46504366e-01
-8.85251522e-01 1.11248231e+00 1.04697180e+00 9.93908942e-02
7.91612983e-01 9.94989127e-02 9.64777529e-01 -9.48551670e-02
2.85428643e-01 -1.28135633e+00 -3.61252189e-01 2.12039217e-01
5.49956083e-01 -1.80873251e+00 -4.03476469e-02 -6.60158575e-01
-1.06864643e+00 8.52279007e-01 8.58445764e-01 4.69259590e-01
7.92288303e-01 -3.34293097e-01 2.21883968e-01 -1.82754286e-02
-5.16178608e-01 -8.18845153e-01 8.04881752e-01 1.00768650e+00
1.28516760e-02 4.24206674e-01 -3.42885882e-01 5.58848500e-01
-8.13522711e-02 -2.97085553e-01 -1.07724115e-01 6.49827003e-01
-6.34080231e-01 -9.78210747e-01 -4.46578443e-01 4.47878748e-01
6.16071792e-03 3.54976147e-01 -5.46761006e-02 9.01363194e-01
7.02640653e-01 1.41277063e+00 -2.45552167e-01 -5.11687338e-01
6.28740311e-01 2.53128648e-01 1.11819040e-02 8.34511826e-04
-2.13021457e-01 -2.56625056e-01 5.20907700e-01 -5.26537299e-01
-6.76080346e-01 -1.16526377e+00 -9.39335883e-01 -4.74242300e-01
-2.20902190e-02 -4.14285064e-01 1.00555623e+00 1.30929935e+00
9.29214880e-02 3.11347574e-01 6.25597060e-01 -9.42892730e-01
3.36026065e-02 -6.85334444e-01 2.55237590e-03 5.91641009e-01
2.88105130e-01 -9.02900636e-01 6.49772286e-02 1.32942095e-01] | [7.717977523803711, -0.5460496544837952] |
7a8f76c1-abe8-4b6b-a7e8-8c8af5d05eb2 | deep-reinforcement-learning-framework-for | 1704.02532 | null | http://arxiv.org/abs/1704.02532v1 | http://arxiv.org/pdf/1704.02532v1.pdf | Deep Reinforcement Learning framework for Autonomous Driving | Reinforcement learning is considered to be a strong AI paradigm which can be
used to teach machines through interaction with the environment and learning
from their mistakes. Despite its perceived utility, it has not yet been
successfully applied in automotive applications. Motivated by the successful
demonstrations of learning of Atari games and Go by Google DeepMind, we propose
a framework for autonomous driving using deep reinforcement learning. This is
of particular relevance as it is difficult to pose autonomous driving as a
supervised learning problem due to strong interactions with the environment
including other vehicles, pedestrians and roadworks. As it is a relatively new
area of research for autonomous driving, we provide a short overview of deep
reinforcement learning and then describe our proposed framework. It
incorporates Recurrent Neural Networks for information integration, enabling
the car to handle partially observable scenarios. It also integrates the recent
work on attention models to focus on relevant information, thereby reducing the
computational complexity for deployment on embedded hardware. The framework was
tested in an open source 3D car racing simulator called TORCS. Our simulation
results demonstrate learning of autonomous maneuvering in a scenario of complex
road curvatures and simple interaction of other vehicles. | ['Senthil Yogamani', 'Mohammed Abdou', 'Etienne Perot', 'Ahmad El Sallab'] | 2017-04-08 | null | null | null | null | ['carracing-v0'] | ['playing-games'] | [-2.28412867e-01 3.75699013e-01 -7.39336163e-02 -3.69982868e-01
-2.83779740e-01 -2.18234748e-01 7.95463324e-01 -2.69929647e-01
-6.70121014e-01 7.04423845e-01 -2.38318488e-01 -7.32386529e-01
6.90098405e-02 -9.68650460e-01 -9.62689102e-01 -5.80422103e-01
-1.53411493e-01 5.21780312e-01 6.08490348e-01 -8.51340353e-01
2.70456195e-01 7.44099379e-01 -2.23219204e+00 -5.32542914e-02
6.19481862e-01 6.44372344e-01 7.17994750e-01 8.35534096e-01
2.28518203e-01 1.33840716e+00 -5.02195537e-01 -4.90976796e-02
1.88495308e-01 -3.96865010e-02 -6.65395498e-01 -8.14891756e-02
2.72825003e-01 -5.39989054e-01 -8.03385496e-01 8.75465274e-01
3.35535884e-01 3.02042037e-01 5.57935536e-01 -1.64144790e+00
-5.34878112e-02 2.99829185e-01 1.40753597e-01 4.20640677e-01
-1.57258332e-01 6.88596070e-01 6.31875873e-01 -7.10898042e-01
4.87832695e-01 1.30554211e+00 2.88885117e-01 6.32785320e-01
-5.24733543e-01 -6.54170275e-01 2.09020153e-01 1.06522155e+00
-8.70018780e-01 -2.14281887e-01 7.78414428e-01 -1.84180647e-01
1.31912804e+00 -1.54993340e-01 6.96777105e-01 1.29604459e+00
5.75484157e-01 1.17438483e+00 9.25462127e-01 -8.13213736e-02
5.35850048e-01 3.86628091e-01 1.04756303e-01 5.74338496e-01
-2.10481770e-02 8.75365198e-01 -8.48546252e-02 5.26337445e-01
3.25326294e-01 -1.58761382e-01 3.67250681e-01 -6.02529824e-01
-8.71373117e-01 9.86787438e-01 6.51659369e-01 1.03794765e-02
-4.29882586e-01 5.65221012e-01 5.35468638e-01 5.90622544e-01
-8.21634531e-02 7.42100552e-02 -2.77936757e-01 -5.06072104e-01
-4.02446747e-01 6.52103782e-01 7.55714655e-01 1.07087219e+00
9.10648167e-01 6.80520296e-01 2.03561738e-01 6.33243620e-01
3.82506728e-01 8.59579265e-01 3.22907329e-01 -1.03533959e+00
2.89957285e-01 3.45300972e-01 7.45471418e-02 -8.06486547e-01
-5.79324424e-01 -2.55891711e-01 -4.06775624e-01 1.04763329e+00
1.27004504e-01 -3.24158311e-01 -8.21210086e-01 1.56794167e+00
2.01137856e-01 4.56466883e-01 4.23230052e-01 9.58344162e-01
8.80291224e-01 5.82639337e-01 1.90596521e-01 3.44405860e-01
1.16047800e+00 -1.14246726e+00 -6.33238554e-01 -4.77864474e-01
6.46510482e-01 -4.07245636e-01 7.35413790e-01 4.42279011e-01
-8.88648927e-01 -9.93718505e-01 -1.39453995e+00 -1.59127582e-02
-9.48690891e-01 -2.95544267e-01 6.07392013e-01 5.03807843e-01
-1.19090915e+00 4.47502732e-01 -8.07843387e-01 -3.49830329e-01
4.14767027e-01 5.07226884e-01 -1.32380113e-01 -9.15733948e-02
-1.54329216e+00 1.54420209e+00 3.81963164e-01 8.21558312e-02
-1.72852635e+00 -3.35837930e-01 -1.05850351e+00 -1.27936736e-01
6.97564065e-01 -4.86637443e-01 1.54894376e+00 -6.70270622e-01
-1.79281366e+00 3.76977116e-01 5.03249057e-02 -1.04376972e+00
4.55308825e-01 -5.29386818e-01 -5.45433104e-01 -1.15392599e-02
-2.13192403e-01 9.55083132e-01 8.15592706e-01 -1.05484033e+00
-9.59136069e-01 -2.21840203e-01 4.37451959e-01 4.25099522e-01
2.66868830e-01 -4.89241481e-01 -1.05754979e-01 -2.18715426e-02
-6.95639610e-01 -1.22125459e+00 -5.48940837e-01 -3.77188236e-01
1.19031966e-01 -4.83489126e-01 1.57313776e+00 -7.96673596e-02
5.53644180e-01 -2.14284992e+00 -1.57740280e-01 1.52323440e-01
1.33671403e-01 5.96439779e-01 -1.22894071e-01 4.25565839e-01
1.60962358e-01 -4.63432252e-01 5.33663966e-02 1.06741562e-01
1.63313046e-01 7.03597605e-01 -2.49303222e-01 1.74661607e-01
5.91859996e-01 1.09502709e+00 -1.16476524e+00 -6.13689087e-02
8.30179870e-01 6.17271781e-01 -4.78810430e-01 3.33960086e-01
-2.05823451e-01 2.36540675e-01 -4.74493384e-01 2.66196132e-01
4.89283979e-01 4.12050009e-01 -2.89766878e-01 2.97054797e-01
-4.05181557e-01 3.47075135e-01 -1.16293871e+00 1.24261904e+00
-5.59092402e-01 1.04252613e+00 2.38844231e-02 -1.38405681e+00
1.06585348e+00 1.85305595e-01 1.88799173e-01 -1.13964379e+00
3.40732247e-01 1.64743423e-01 3.23326647e-01 -8.35373163e-01
6.95005715e-01 1.00621045e-01 -9.97254252e-02 1.24064513e-01
2.49436527e-01 -5.53303242e-01 1.79031625e-01 2.18121707e-01
1.20471621e+00 1.90846279e-01 -6.24777041e-02 -3.21761638e-01
6.06397212e-01 2.37038493e-01 2.86527932e-01 7.85689890e-01
-5.32313704e-01 -5.78667819e-02 2.51041532e-01 -6.72619998e-01
-9.71046984e-01 -1.00229526e+00 9.66689065e-02 1.01814914e+00
5.09385884e-01 -1.02751583e-01 -8.96596432e-01 -7.15550244e-01
2.50817332e-02 1.12957907e+00 -6.50298476e-01 -5.55803239e-01
-6.42984629e-01 -1.67376399e-01 2.76441067e-01 5.92912912e-01
7.56970048e-01 -1.65776086e+00 -1.17467082e+00 4.05407757e-01
4.22090560e-01 -1.23544657e+00 1.61833003e-01 4.90696877e-01
-4.88597661e-01 -1.16662550e+00 -3.02840710e-01 -9.17583525e-01
1.47969529e-01 4.73712623e-01 1.11748588e+00 1.81791000e-02
-2.93886751e-01 7.77093172e-01 -1.80923730e-01 -9.62893307e-01
-7.09294617e-01 4.93373722e-02 2.09387153e-01 -3.64925385e-01
8.26277614e-01 -3.16574037e-01 -5.99332035e-01 4.81019080e-01
-7.67445803e-01 -7.32608065e-02 7.02694416e-01 8.62959862e-01
1.40526742e-01 3.05052698e-01 8.36833179e-01 -7.36620843e-01
6.40682042e-01 -6.66289747e-01 -8.47749412e-01 -3.86489123e-01
-4.54041183e-01 5.97224310e-02 4.52118814e-01 -1.45065576e-01
-9.90034819e-01 1.35264164e-02 -5.71548462e-01 -1.05368108e-01
-7.31225908e-01 2.37720385e-01 -3.15311961e-02 2.03061551e-02
4.76417452e-01 6.68669865e-02 4.01999503e-01 1.57652184e-01
3.52857590e-01 5.85887194e-01 4.81921762e-01 -1.97244838e-01
7.14259624e-01 2.78486550e-01 1.45805821e-01 -1.19417572e+00
-2.83629447e-01 -2.50482649e-01 -5.98842144e-01 -6.14084363e-01
1.02038097e+00 -9.31930959e-01 -1.37337780e+00 4.27912533e-01
-8.50437522e-01 -8.03369045e-01 -2.77338415e-01 6.22128665e-01
-1.04151273e+00 -4.62324992e-02 -3.66178334e-01 -7.46268988e-01
1.35295674e-01 -1.61449683e+00 5.97856164e-01 4.23131704e-01
9.33651328e-02 -1.00623512e+00 1.31575102e-02 1.88898057e-01
6.95391834e-01 -1.17212832e-01 7.09284127e-01 -5.09689212e-01
-7.94328630e-01 -5.83092356e-03 6.00086823e-02 4.59681064e-01
-2.38786668e-01 -1.98286071e-01 -1.19708788e+00 -1.61250502e-01
-1.44426301e-01 -5.62856734e-01 7.85656452e-01 2.22176582e-01
9.46196735e-01 2.75524020e-01 -2.48860225e-01 7.97543302e-02
1.15469933e+00 4.59649056e-01 8.22542012e-01 4.83650982e-01
4.29793805e-01 7.72459805e-01 9.16462660e-01 1.08514436e-01
6.98312461e-01 6.82998240e-01 1.14655507e+00 -1.98528022e-01
-2.15829998e-01 -2.30671242e-01 7.09214747e-01 5.53501606e-01
6.44433722e-02 1.36145562e-01 -8.97722602e-01 6.27488852e-01
-1.98104620e+00 -1.13545215e+00 -1.33805633e-01 1.95994234e+00
8.79377425e-02 5.90996027e-01 1.96046174e-01 1.02302164e-01
2.02557042e-01 -1.84269965e-01 -9.23953593e-01 -1.02821767e+00
9.01154876e-02 1.49294943e-01 5.68187654e-01 6.69506788e-01
-1.02432215e+00 1.27507663e+00 6.49210405e+00 5.65335453e-01
-1.20119333e+00 -1.37718454e-01 3.74422640e-01 1.47262439e-01
3.21063884e-02 -3.00786048e-01 -9.76552844e-01 2.94819355e-01
1.50821996e+00 7.09591061e-02 3.35127622e-01 1.20226729e+00
3.52640510e-01 -3.53728175e-01 -9.34523940e-01 7.48398066e-01
-1.90653831e-01 -1.14850771e+00 -2.46670365e-01 2.95500597e-03
4.03456300e-01 6.85709476e-01 3.14931780e-01 1.12555540e+00
8.25935483e-01 -1.14136076e+00 5.77473938e-01 2.54654527e-01
1.28416359e-01 -1.33685827e+00 8.88989985e-01 5.86851835e-01
-8.68988991e-01 -4.23479795e-01 -4.84271884e-01 -2.11894184e-01
5.46608381e-02 -6.13338612e-02 -1.28303158e+00 1.37644574e-01
8.94633591e-01 7.38978446e-01 -5.29837072e-01 8.88564944e-01
-2.97891557e-01 6.51895642e-01 -5.48036732e-02 -5.11483252e-01
8.24504435e-01 -2.31296346e-02 5.36456287e-01 1.22867334e+00
1.24100201e-01 -1.05584718e-01 2.29534775e-01 6.41704142e-01
3.40277493e-01 -2.78800637e-01 -1.36122000e+00 3.86432767e-01
1.49282992e-01 1.16027629e+00 -4.12791133e-01 -3.77252877e-01
-6.38076365e-01 5.47439098e-01 3.45366627e-01 4.60678697e-01
-9.30884659e-01 -5.45200229e-01 9.99959230e-01 1.68175131e-01
6.21534944e-01 -4.30114806e-01 1.09992154e-01 -4.82446969e-01
-3.10502797e-01 -8.13820302e-01 -4.67774756e-02 -7.52212405e-01
-6.96673155e-01 7.51374602e-01 1.50262728e-01 -1.28058183e+00
-5.53913534e-01 -9.29833412e-01 -6.48974061e-01 4.59572107e-01
-2.07619047e+00 -7.53054142e-01 -3.37496877e-01 7.84113348e-01
9.04104054e-01 -6.22232258e-01 6.73347890e-01 1.98401868e-01
-2.56921262e-01 2.81206906e-01 -1.11875199e-01 -1.25263080e-01
3.35023850e-01 -1.36007154e+00 5.04968762e-01 4.48865533e-01
-1.08922705e-01 1.38362080e-01 1.02659965e+00 -3.18192065e-01
-1.58236408e+00 -1.04632688e+00 3.87779534e-01 -4.33604151e-01
5.24622858e-01 -5.12871146e-01 -8.71122301e-01 5.21815479e-01
7.30395555e-01 -8.51969272e-02 2.60634124e-01 -1.53218001e-01
4.91001606e-02 -3.15187663e-01 -8.95228326e-01 9.46984947e-01
7.54517198e-01 -3.92455280e-01 -6.23020768e-01 9.90076214e-02
5.63305497e-01 -3.57113689e-01 -3.82442802e-01 4.39683616e-01
2.13801354e-01 -1.07419825e+00 7.55929530e-01 -5.57321668e-01
1.57362014e-01 -4.56314117e-01 1.72269549e-02 -1.60693276e+00
-1.55022293e-01 -6.24070108e-01 -1.33804068e-01 5.35465062e-01
3.79840463e-01 -5.02116501e-01 8.59587550e-01 2.81115025e-01
-5.70602059e-01 -5.03948987e-01 -9.19721901e-01 -6.34846985e-01
2.66872883e-01 -8.74041855e-01 3.66066992e-01 3.36027592e-01
-6.48666872e-03 4.79998201e-01 -4.52688783e-01 1.32979676e-01
4.64347035e-01 -4.00511235e-01 1.19793677e+00 -1.05919170e+00
-2.52284817e-02 -3.98038030e-01 -9.01980340e-01 -1.12339795e+00
4.14338887e-01 -7.58297861e-01 4.72236305e-01 -1.33338559e+00
-3.86858821e-01 -3.20348859e-01 -3.76687706e-01 1.80973694e-01
1.20946981e-01 2.52430979e-02 1.04389608e-01 -4.41070914e-01
-7.51537979e-01 7.52624512e-01 1.19293559e+00 -2.97795773e-01
-2.24004332e-02 2.45732069e-01 -4.36320782e-01 6.77470624e-01
9.76725101e-01 -2.82388657e-01 -7.88283706e-01 -2.35704482e-01
1.00580327e-01 -7.10288063e-02 4.51843649e-01 -1.45685434e+00
3.84800524e-01 3.38792652e-02 1.06609359e-01 -8.76379669e-01
5.11446476e-01 -1.22516596e+00 -2.97755599e-01 7.47702599e-01
-2.51233369e-01 2.30286106e-01 6.97868407e-01 6.46686912e-01
-3.04968268e-01 -1.51868820e-01 7.56189406e-01 -1.78222284e-01
-1.54082012e+00 1.49722248e-01 -1.23923373e+00 -2.10108861e-01
1.47348058e+00 -1.99162632e-01 -7.72693902e-02 -6.57222569e-01
-7.31885552e-01 6.73512757e-01 -9.76656377e-03 1.06569207e+00
9.03553128e-01 -1.29905045e+00 -6.49788022e-01 5.04957438e-01
1.60905167e-01 -2.07674682e-01 2.84252822e-01 4.48344171e-01
-4.38746125e-01 6.13217115e-01 -6.53833270e-01 -8.10624301e-01
-1.12563944e+00 7.78310180e-01 3.48848403e-01 9.26561002e-03
-8.26323867e-01 1.74155325e-01 1.93243086e-01 -6.61060214e-01
3.45391452e-01 -2.50298709e-01 -6.54991567e-01 -3.52825910e-01
7.19455540e-01 3.24741095e-01 3.32262278e-01 -6.25823557e-01
-1.40859708e-01 2.06879899e-01 -2.53030390e-01 8.15847702e-03
1.18626225e+00 -1.63910612e-01 6.20094240e-01 5.70653856e-01
9.39432025e-01 -5.49810290e-01 -1.55959213e+00 -1.68521240e-01
1.31370142e-01 2.92278007e-02 1.85570598e-01 -5.99187195e-01
-9.76710379e-01 1.31792247e+00 9.47599709e-01 1.74336851e-01
5.14302731e-01 -2.16274768e-01 7.45249093e-01 9.19272840e-01
5.66016197e-01 -1.54612660e+00 2.16189608e-01 1.16735816e+00
8.46154213e-01 -1.66976607e+00 -5.16661882e-01 1.44286916e-01
-9.87550974e-01 1.08521307e+00 9.63250697e-01 -4.31523442e-01
1.01791990e+00 3.66093755e-01 3.81795317e-01 -2.52563298e-01
-1.12542808e+00 -5.68879843e-01 -2.68445872e-02 1.04685473e+00
3.36697921e-02 -3.95351611e-02 2.00879052e-01 -3.91105376e-02
-2.44376287e-01 -6.22928366e-02 6.18788183e-01 8.49206507e-01
-8.19456160e-01 -9.07589853e-01 -6.56746444e-04 1.83214650e-01
-1.43002942e-01 3.07844251e-01 3.03474776e-02 1.19636285e+00
4.73730862e-02 1.10333502e+00 1.24248102e-01 -4.26265180e-01
5.91646850e-01 -1.35303184e-01 1.65248498e-01 -4.30529803e-01
-5.07702947e-01 -4.83931303e-01 6.45959899e-02 -7.33387709e-01
-2.04832390e-01 -6.46815658e-01 -1.40614688e+00 -2.63619453e-01
1.65732294e-01 1.36624351e-01 9.48417544e-01 1.02079058e+00
4.15300369e-01 9.17056799e-01 6.33557260e-01 -9.43221509e-01
-4.39934433e-01 -7.54532456e-01 -3.62469822e-01 2.55802900e-01
5.68497598e-01 -1.12974203e+00 -4.63049747e-02 -3.61803770e-01] | [5.368793487548828, 1.140724539756775] |
4c88957d-7142-45b6-9769-9040b9939fac | upb-at-semeval-2021-task-5-virtual | 2104.08635 | null | https://arxiv.org/abs/2104.08635v1 | https://arxiv.org/pdf/2104.08635v1.pdf | UPB at SemEval-2021 Task 5: Virtual Adversarial Training for Toxic Spans Detection | The real-world impact of polarization and toxicity in the online sphere marked the end of 2020 and the beginning of this year in a negative way. Semeval-2021, Task 5 - Toxic Spans Detection is based on a novel annotation of a subset of the Jigsaw Unintended Bias dataset and is the first language toxicity detection task dedicated to identifying the toxicity-level spans. For this task, participants had to automatically detect character spans in short comments that render the message as toxic. Our model considers applying Virtual Adversarial Training in a semi-supervised setting during the fine-tuning process of several Transformer-based models (i.e., BERT and RoBERTa), in combination with Conditional Random Fields. Our approach leads to performance improvements and more robust models, enabling us to achieve an F1-score of 65.73% in the official submission and an F1-score of 66.13% after further tuning during post-evaluation. | ['Mihai Dascalu', 'Dumitru-Clementin Cercel', 'Andrei Paraschiv'] | 2021-04-17 | null | https://aclanthology.org/2021.semeval-1.26 | https://aclanthology.org/2021.semeval-1.26.pdf | semeval-2021 | ['toxic-spans-detection'] | ['natural-language-processing'] | [ 1.47330686e-01 2.01690838e-01 -5.09368181e-02 1.28728211e-01
-1.40550053e+00 -1.03828490e+00 9.65910316e-01 6.29271150e-01
-3.21267337e-01 1.03667998e+00 5.22549450e-01 -4.84713465e-01
2.26909611e-02 -7.04843998e-01 -8.18476260e-01 -3.12074542e-01
-9.15640965e-02 4.14700896e-01 1.88500941e-01 -1.90246642e-01
2.15955749e-01 2.61884600e-01 -9.07537520e-01 7.32793033e-01
8.22448730e-01 7.58926809e-01 -3.81600887e-01 7.00116456e-01
2.98591882e-01 9.18004096e-01 -9.30402160e-01 -1.21386111e+00
1.53772399e-01 -9.95601267e-02 -9.64893818e-01 -4.38492715e-01
4.71796781e-01 8.71088505e-02 -3.00210506e-01 1.07817817e+00
7.65450239e-01 -3.40780139e-01 6.07906163e-01 -1.15336299e+00
-2.98782885e-01 8.68107259e-01 -2.29306370e-01 3.21981877e-01
5.56170285e-01 5.63275278e-01 1.07757425e+00 -6.00413680e-01
8.88183415e-01 1.13658369e+00 9.87316132e-01 5.20260930e-01
-1.33884430e+00 -8.27252269e-01 -7.60855675e-02 1.63485214e-01
-1.13529050e+00 -1.94068864e-01 4.12551254e-01 -9.49297130e-01
7.14004099e-01 6.47072673e-01 2.51311153e-01 1.65140378e+00
2.17522368e-01 3.85388255e-01 1.37860155e+00 -1.08063214e-01
7.24641010e-02 3.85964483e-01 -6.87318817e-02 3.69503945e-01
-3.27664837e-02 2.10201457e-01 -6.67123020e-01 -6.73984408e-01
-3.66024673e-01 -6.13273323e-01 -2.58286655e-01 5.54792583e-01
-9.94592905e-01 9.70200062e-01 2.26196304e-01 1.49256825e-01
-2.86020964e-01 5.80339022e-02 7.65375853e-01 1.10336423e-01
9.35516059e-01 5.23911357e-01 -6.47308528e-01 -2.26270154e-01
-9.14218009e-01 6.38095856e-01 7.69188762e-01 6.05710089e-01
1.88053340e-01 -4.12514716e-01 -1.14620566e+00 5.63013375e-01
1.36735097e-01 4.79094744e-01 5.66250347e-02 -5.85103095e-01
8.55024993e-01 3.66180301e-01 2.91232169e-01 -6.87055707e-01
-5.72175503e-01 -6.78690195e-01 -4.84437138e-01 3.04912440e-02
6.32759035e-01 -5.89638948e-01 -7.27045298e-01 1.84981394e+00
2.86486506e-01 -1.73254624e-01 -2.81709522e-01 5.60408294e-01
6.18691623e-01 4.28589106e-01 5.70985138e-01 -6.85413927e-02
1.31260967e+00 -5.42514265e-01 -4.69385862e-01 1.17993146e-01
8.64158452e-01 -9.13442731e-01 1.20432317e+00 8.08616102e-01
-7.28386879e-01 -6.09689988e-02 -8.97377193e-01 2.64257073e-01
-5.10248244e-01 -1.92453489e-01 3.07773322e-01 1.04761803e+00
-7.39354193e-01 8.91618013e-01 -9.53830592e-03 -2.02503189e-01
4.86411065e-01 4.91908528e-02 -2.40972102e-01 1.38312191e-01
-1.79157841e+00 9.59625781e-01 2.35565994e-02 -4.83856171e-01
-1.14735806e+00 -1.34870601e+00 -3.37841630e-01 -2.85204321e-01
-7.97394023e-04 -5.17503917e-01 1.21741033e+00 -7.55012572e-01
-1.02840924e+00 9.68820214e-01 3.84643644e-01 -7.20495760e-01
1.21202564e+00 -3.29813570e-01 -5.54193377e-01 -4.87308912e-02
1.93803489e-01 1.53413072e-01 4.35947210e-01 -8.27726781e-01
-4.74533916e-01 -2.94430196e-01 1.79311484e-01 -2.41646931e-01
-5.78977048e-01 5.87365210e-01 -1.15336031e-01 -7.06022382e-01
-1.05991793e+00 -8.88321459e-01 -1.89561084e-01 -5.91450632e-01
-9.02302563e-01 -2.07269177e-01 4.75268990e-01 -1.17826617e+00
1.57345474e+00 -1.70925200e+00 -3.58201414e-01 1.40173122e-01
1.90738484e-01 1.98436469e-01 -1.40562251e-01 6.08660877e-01
-3.40487748e-01 5.57220519e-01 -1.16951928e-01 -2.72713423e-01
2.22490177e-01 -7.10882425e-01 -5.61579525e-01 6.49059713e-01
1.91380799e-01 7.19230473e-01 -1.04341125e+00 -1.76143348e-01
-2.18861386e-01 2.44279593e-01 -4.59992290e-01 8.51417854e-02
-4.83024240e-01 4.77907240e-01 -2.60924816e-01 4.98100936e-01
7.02321827e-01 2.47089177e-01 -2.05258932e-02 3.11492551e-02
-3.03351879e-01 6.75951183e-01 -4.54915524e-01 1.28691590e+00
-5.36609352e-01 5.77290177e-01 -2.45302334e-01 -3.25820565e-01
5.19714415e-01 4.25954849e-01 3.94441575e-01 -8.48511040e-01
3.37433629e-02 6.66225851e-02 -1.04065731e-01 -5.47456920e-01
4.19653952e-01 -3.21133912e-01 -6.00372016e-01 2.96054363e-01
-2.15665057e-01 1.19464658e-01 3.52669358e-01 5.20624101e-01
1.56342971e+00 -1.17643513e-01 -2.18271777e-01 -3.32831353e-01
5.15898883e-01 -1.50438204e-01 4.45037723e-01 5.28195083e-01
-9.35091078e-02 7.60483325e-01 9.00854349e-01 -2.67634392e-01
-1.03651845e+00 -8.88713479e-01 -1.18845150e-01 9.50734913e-01
-6.11961484e-01 -6.64670289e-01 -9.34989035e-01 -1.21016717e+00
-1.00945984e-03 1.29896331e+00 -9.30622458e-01 -2.40719080e-01
-9.84213501e-02 -1.16601431e+00 1.14698052e+00 2.45256834e-02
7.60525838e-02 -8.06510270e-01 -5.81574664e-02 2.52770931e-01
-5.04566550e-01 -1.11650062e+00 -5.55869281e-01 9.91032347e-02
-2.10501239e-01 -1.19129741e+00 -7.67951548e-01 2.59069148e-02
1.06541216e-01 -7.89392650e-01 1.18173528e+00 -3.11111450e-01
-2.03281209e-01 -9.24119726e-02 -2.54328340e-01 -7.07445800e-01
-8.03949893e-01 9.68281850e-02 -1.21309035e-01 1.10012248e-01
1.72567084e-01 -3.25750768e-01 -5.00516176e-01 2.02239349e-01
-8.05179536e-01 -2.87649870e-01 1.49397418e-01 5.39511800e-01
5.93894944e-02 -2.94090480e-01 6.22239590e-01 -1.39785969e+00
7.99906433e-01 -7.37886846e-01 -6.39644265e-01 1.48525983e-01
-7.09479213e-01 -2.95177013e-01 9.14558232e-01 -2.58040726e-01
-8.91139627e-01 -1.38232693e-01 -3.72890532e-01 9.21453014e-02
1.09763099e-02 5.10172367e-01 -3.73611301e-01 9.25118849e-02
1.13251805e+00 -2.65351325e-01 -5.97477555e-01 -4.27397341e-01
2.69148827e-01 9.07671988e-01 2.79338568e-01 -3.77495319e-01
1.03860712e+00 -9.29068215e-03 -1.27984613e-01 -2.86203265e-01
-1.19875252e+00 -3.36843520e-01 -3.07229370e-01 -4.09052104e-01
6.11253262e-01 -1.14893079e+00 -7.19361007e-01 6.36815548e-01
-1.33661294e+00 -5.22002995e-01 -4.14744951e-02 9.67743695e-02
-2.24268168e-01 4.22415584e-01 -5.16386509e-01 -7.20593512e-01
-5.34616828e-01 -6.24267280e-01 6.65226400e-01 -2.04476893e-01
-6.18071198e-01 -8.86654913e-01 4.45926458e-01 7.47556090e-01
2.25438550e-01 7.13649809e-01 9.84148681e-01 -8.79695475e-01
7.99210966e-02 -4.32306767e-01 -1.61404416e-01 2.80022711e-01
-4.63567406e-01 2.65953802e-02 -1.21198761e+00 -1.04042932e-01
-5.03423274e-01 -4.90712494e-01 8.42289686e-01 -3.41670588e-02
1.11219275e+00 -6.54544950e-01 -1.46613896e-01 1.01244546e-01
1.23291647e+00 -3.48329902e-01 9.95717227e-01 5.37610471e-01
3.42968196e-01 6.66189492e-01 6.70580089e-01 8.43517423e-01
2.64765859e-01 7.88099527e-01 6.59336448e-01 4.14569266e-02
-2.75191933e-01 -5.65756202e-01 8.05955648e-01 2.71356463e-01
3.20598066e-01 -5.34363210e-01 -1.02198648e+00 7.57748246e-01
-1.42956591e+00 -1.06911314e+00 -7.17120349e-01 2.45428252e+00
1.27835500e+00 4.68677789e-01 4.71074849e-01 1.80766001e-01
7.77890563e-01 -2.70349300e-03 -1.75253838e-01 -7.35509336e-01
-3.04200977e-01 1.35978863e-01 8.94465387e-01 5.47028661e-01
-1.30650771e+00 7.86939263e-01 6.54849529e+00 1.18361020e+00
-8.60825121e-01 4.23347205e-01 9.77989078e-01 -1.25795648e-01
-6.66328967e-01 -7.31344568e-03 -6.64752483e-01 9.49141443e-01
1.57726085e+00 -3.68822157e-01 2.72222549e-01 2.42653877e-01
5.66446006e-01 2.94168591e-02 -7.71774292e-01 3.57439637e-01
1.44934103e-01 -1.26706290e+00 -1.46861896e-01 9.29790959e-02
8.47401917e-01 2.30977342e-01 3.23451906e-01 3.91589075e-01
4.86922592e-01 -1.23309338e+00 1.12348306e+00 6.39719248e-01
1.18971956e+00 -9.74756479e-01 7.62413859e-01 4.47664589e-01
-3.63788545e-01 -1.97589457e-01 1.04396090e-01 5.80063537e-02
-9.49961599e-03 1.29475462e+00 -1.10658610e+00 5.86357057e-01
6.26311481e-01 4.68408763e-01 -7.48409510e-01 1.18985629e+00
-4.69723344e-01 1.21190846e+00 -7.41016790e-02 -2.08806977e-01
1.50151253e-01 3.18747073e-01 8.08669567e-01 1.69434488e+00
2.26501301e-01 -5.14995337e-01 -2.27413937e-01 8.81246030e-01
-6.28414512e-01 1.85156584e-01 -4.04160827e-01 -2.40613833e-01
1.95959434e-01 1.44265461e+00 -1.63580850e-01 -1.72841605e-02
-8.74651223e-02 7.25404084e-01 1.59603223e-01 8.62392336e-02
-1.26361573e+00 -4.80286509e-01 2.80480146e-01 6.10113263e-01
-1.63958687e-03 3.61931592e-01 -5.38582265e-01 -6.52920544e-01
-3.19256246e-01 -9.72256839e-01 4.66293246e-01 -5.77009141e-01
-1.47395194e+00 5.12620628e-01 -3.88154417e-01 -1.30623806e+00
5.41742407e-02 -4.44197655e-01 -6.80568278e-01 1.14259350e+00
-1.18562102e+00 -1.26517332e+00 8.82319883e-02 2.89158285e-01
8.10834318e-02 -1.03741750e-01 8.84704113e-01 5.87985337e-01
-4.50115144e-01 1.04530072e+00 1.62632793e-01 1.36340454e-01
1.16066420e+00 -1.38602316e+00 4.35389817e-01 8.71114492e-01
-2.39404857e-01 1.61640123e-01 1.09839833e+00 -8.12780082e-01
-5.47772586e-01 -1.67747223e+00 1.60613227e+00 -1.04707599e+00
1.08376491e+00 -6.48153305e-01 -2.75385529e-01 1.93337664e-01
1.42243817e-01 -4.14166838e-01 8.70595276e-01 2.10544914e-01
-6.53601944e-01 1.72346264e-01 -1.19603848e+00 4.28929985e-01
6.67285085e-01 -7.04214394e-01 -2.24382564e-01 9.43380296e-01
7.47031987e-01 -3.31239909e-01 -9.27617788e-01 2.54635841e-01
2.31358305e-01 -7.67927885e-01 6.23087943e-01 -8.93581629e-01
8.78737092e-01 -7.67847970e-02 -9.73243341e-02 -1.29063797e+00
-3.62294316e-01 -1.04103303e+00 1.85261071e-01 1.53812027e+00
1.03037035e+00 -3.07246387e-01 6.12324059e-01 4.83694524e-01
-6.98297396e-02 -6.48890316e-01 -1.06047046e+00 -6.73878908e-01
4.47618932e-01 -7.26306438e-01 2.23158062e-01 7.74959564e-01
2.90786009e-02 3.00743252e-01 -4.45923597e-01 1.86652005e-01
6.30704522e-01 -2.76039630e-01 5.21653771e-01 -1.23019302e+00
-1.36522964e-01 -3.44051421e-01 -1.65620282e-01 -1.35291979e-01
2.04485267e-01 -1.12456083e+00 -1.25440836e-01 -1.07002401e+00
6.38527870e-01 -2.44868398e-01 -4.11537826e-01 5.40618837e-01
-3.16701204e-01 6.03760540e-01 2.57756591e-01 -1.78378925e-01
-6.52962446e-01 3.57076436e-01 6.95219517e-01 -5.57005525e-01
2.50526398e-01 2.96478748e-01 -1.02834094e+00 5.58573246e-01
8.53163600e-01 -9.04637575e-01 9.81959030e-02 1.69229791e-01
6.89449430e-01 -3.87540311e-01 6.19239926e-01 -7.94884801e-01
-2.22228423e-01 -1.61678001e-01 2.92577863e-01 -6.02252245e-01
-7.80613795e-02 -2.97657281e-01 2.26297855e-01 5.58509767e-01
-6.08480692e-01 -9.92968157e-02 4.05393779e-01 6.01620376e-01
1.37814373e-01 -2.90309817e-01 5.84878325e-01 -2.71530710e-02
6.68681413e-02 2.79140919e-01 -5.87716103e-01 5.38578629e-01
9.08579290e-01 3.76125276e-01 -5.96621811e-01 -4.23675448e-01
-6.52387857e-01 1.65011436e-01 2.72769839e-01 3.72131467e-01
-7.03595718e-03 -1.01051474e+00 -1.15281236e+00 -4.91391212e-01
3.08793336e-01 -7.40359604e-01 3.00415337e-01 1.00534427e+00
-3.38708550e-01 4.10965085e-01 2.12660674e-02 7.11744726e-02
-1.38901114e+00 3.90230119e-01 3.26065600e-01 -1.03618073e+00
-4.00045551e-02 9.39601660e-01 -8.78212824e-02 -6.06805563e-01
2.53257096e-01 2.21579507e-01 -2.72515476e-01 2.51757562e-01
6.54418945e-01 6.46150708e-01 5.58887780e-01 -6.22174680e-01
-4.85895127e-01 -4.63369861e-02 7.03147948e-02 -3.09104532e-01
1.11986554e+00 2.38326281e-01 -5.93964644e-02 2.54935831e-01
1.25267255e+00 7.42720366e-01 -8.25621903e-01 1.28384411e-01
2.85294503e-01 -1.84472650e-01 1.21752368e-02 -1.95620024e+00
-7.13340819e-01 8.30789089e-01 5.50054014e-01 1.69041194e-02
8.30675185e-01 -1.89107254e-01 7.35614717e-01 -1.56939700e-01
2.81972829e-02 -1.20242691e+00 5.57377422e-03 7.05913603e-01
1.15643609e+00 -7.11915970e-01 7.94617757e-02 -4.57494259e-01
-6.75144851e-01 7.76135087e-01 3.28917712e-01 3.72473836e-01
2.94400930e-01 1.62079811e-01 9.76270437e-02 -9.87294242e-02
-1.05768740e+00 6.27874509e-02 3.77645165e-01 6.74380124e-01
5.04381478e-01 2.52676398e-01 -6.22441649e-01 1.17514706e+00
-2.84340769e-01 -1.40859753e-01 6.89095974e-01 1.63484976e-01
-6.54607564e-02 -1.11002922e+00 -3.07101160e-01 4.72290814e-01
-1.10767412e+00 -4.00574267e-01 -9.19845998e-01 3.87189180e-01
4.10053313e-01 1.10516095e+00 -5.57035744e-01 -7.87966192e-01
5.96089840e-01 2.18245521e-01 7.60192350e-02 -3.98192585e-01
-1.36421227e+00 -2.31023818e-01 7.71443367e-01 -4.59338158e-01
-3.46749462e-02 -6.73944712e-01 -9.17856097e-01 -4.38043416e-01
-2.29406312e-01 3.02829266e-01 5.23457170e-01 7.09715068e-01
3.81013185e-01 5.45058548e-01 8.66905808e-01 -1.60594732e-01
-7.52652347e-01 -1.30475652e+00 -4.50752467e-01 7.74720311e-01
5.27741350e-02 -2.44185954e-01 -5.00968754e-01 1.57643706e-02] | [8.947410583496094, 10.62563419342041] |
869c333c-d62e-49e8-be2c-b7e105ef055c | residual-3d-scene-flow-learning-with-context | 2109.04685 | null | https://arxiv.org/abs/2109.04685v2 | https://arxiv.org/pdf/2109.04685v2.pdf | Residual 3D Scene Flow Learning with Context-Aware Feature Extraction | Scene flow estimation is the task to predict the point-wise or pixel-wise 3D displacement vector between two consecutive frames of point clouds or images, which has important application in fields such as service robots and autonomous driving. Although many previous works have explored greatly on scene flow estimation based on point clouds, there are two problems that have not been noticed or well solved before: 1) Points of adjacent frames in repetitive patterns may be wrongly associated due to similar spatial structure in their neighbourhoods; 2) Scene flow between adjacent frames of point clouds with long-distance movement may be inaccurately estimated. To solve the first problem, a novel context-aware set convolution layer is proposed in this paper to exploit contextual structure information of Euclidean space and learn soft aggregation weights for local point features. This design is inspired by human perception of contextual structure information during scene understanding with repetitive patterns. The context-aware set convolution layer is incorporated in a context-aware point feature pyramid module of 3D point clouds for scene flow estimation. For the second problem, an explicit residual flow learning structure is proposed in the residual flow refinement layer to cope with long-distance movement. The experiments and ablation study on FlyingThings3D and KITTI scene flow datasets demonstrate the effectiveness of each proposed component. The qualitative results show that the problems of ambiguous inter-frame association and long-distance movement estimation are well handled. Quantitative results on both FlyingThings3D and KITTI scene flow datasets show that the proposed method achieves state-of-the-art performance, surpassing all other previous works to the best of our knowledge by at least 25%. | ['Hesheng Wang', 'Xinrui Wu', 'Yunzhe Hu', 'Guangming Wang'] | 2021-09-10 | null | null | null | null | ['scene-flow-estimation'] | ['computer-vision'] | [ 4.37901840e-02 -4.26905364e-01 1.45548001e-01 -5.03664136e-01
2.96506472e-02 -3.00874352e-01 5.73177218e-01 -9.81284901e-02
-5.57735085e-01 5.03091931e-01 -1.03320241e-01 -1.31514579e-01
-2.50706136e-01 -7.93638825e-01 -7.34338999e-01 -6.18952394e-01
-2.91881561e-01 2.76492894e-01 6.88295484e-01 -3.94830018e-01
5.85511088e-01 8.87662768e-01 -1.94791603e+00 2.30194256e-01
8.79714310e-01 1.04277229e+00 5.76690495e-01 6.74126625e-01
-5.04837871e-01 9.49555397e-01 -4.18858409e-01 1.42660856e-01
6.17971718e-01 -5.48588000e-02 -8.28126609e-01 2.57781923e-01
7.78647184e-01 -4.10150468e-01 -4.23893869e-01 8.37852657e-01
1.56914458e-01 5.15218675e-01 4.17008609e-01 -1.41622651e+00
-1.79758847e-01 -1.81428626e-01 -8.05382311e-01 6.05757296e-01
2.29979336e-01 3.51389825e-01 5.50609291e-01 -9.90370750e-01
5.56371272e-01 1.27926505e+00 5.92645764e-01 1.27570331e-01
-6.98948383e-01 -6.81953490e-01 3.77679676e-01 4.86205220e-01
-1.31348228e+00 -2.46702470e-02 8.74793470e-01 -6.34707868e-01
1.13386869e+00 7.81195983e-02 8.16513777e-01 3.32165748e-01
1.61255464e-01 6.39252245e-01 6.68374658e-01 2.16538217e-02
5.20279333e-02 -1.37422219e-01 -4.35015522e-02 7.40461469e-01
8.38698074e-02 2.72706836e-01 -4.22469020e-01 3.38632256e-01
9.83328640e-01 2.87130982e-01 -4.15757626e-01 -6.27515674e-01
-1.44644856e+00 7.70409167e-01 9.31327343e-01 4.24017936e-01
-2.83135235e-01 2.80078888e-01 2.14085162e-01 1.36896685e-01
4.57194924e-01 2.56590545e-01 -4.64935780e-01 -4.21165563e-02
-9.26165044e-01 2.75864601e-01 5.42118251e-01 1.19378114e+00
1.23938131e+00 1.45230994e-01 1.42725870e-01 4.79019344e-01
2.28266716e-01 5.40749311e-01 2.29428872e-01 -1.09573555e+00
6.40489340e-01 8.56592298e-01 2.16838136e-01 -1.43488467e+00
-6.21250391e-01 -2.73015440e-01 -8.66211891e-01 5.38097799e-01
5.30664325e-01 1.49986893e-01 -8.08052540e-01 1.19699371e+00
6.22255266e-01 8.59908402e-01 -6.52027726e-02 1.37597680e+00
9.45186913e-01 8.83353651e-01 -1.24618009e-01 -2.44232323e-02
1.12094045e+00 -9.38973427e-01 -5.34138381e-01 -2.85032213e-01
6.52473748e-01 -8.18258584e-01 7.72350073e-01 7.39177614e-02
-9.12533283e-01 -1.02475047e+00 -9.30581629e-01 -2.77053326e-01
-4.72245753e-01 -6.04689345e-02 8.26934278e-01 2.27409124e-01
-8.02164495e-01 4.85278517e-01 -8.22642148e-01 -2.70136684e-01
5.12513518e-01 2.76617944e-01 -4.83591318e-01 -2.75770456e-01
-9.08184528e-01 8.16993713e-01 3.33492786e-01 4.45156664e-01
-4.68991250e-01 -1.14668179e+00 -8.91141772e-01 -7.57085010e-02
3.03443640e-01 -8.74518633e-01 8.79346132e-01 -7.08551824e-01
-1.13187110e+00 6.71186090e-01 -3.32099557e-01 -4.67704177e-01
6.11460447e-01 -3.78773630e-01 -2.01034874e-01 1.52671248e-01
2.10764766e-01 8.53463888e-01 7.98366189e-01 -1.18493259e+00
-1.24956656e+00 -2.67333776e-01 1.66749105e-01 4.52702135e-01
1.19286910e-01 -3.09859574e-01 -4.41018462e-01 -3.59504700e-01
3.68354887e-01 -8.35717440e-01 -4.13260400e-01 1.86330304e-01
-7.64398426e-02 -2.61053085e-01 1.33420730e+00 -1.11648433e-01
9.27263975e-01 -1.96990013e+00 4.93304022e-02 -7.72792548e-02
1.20803848e-01 4.12577480e-01 2.89219357e-02 2.39522494e-02
-1.27127720e-02 -2.06963003e-01 -4.25802171e-01 -2.02243894e-01
-3.95489395e-01 4.50174838e-01 -4.18216705e-01 7.23687351e-01
3.79314601e-01 6.65941954e-01 -1.09904397e+00 -3.83668423e-01
1.15057576e+00 7.10290849e-01 -6.76564455e-01 2.21384466e-01
3.86654399e-02 7.40064979e-01 -5.12651026e-01 3.61604333e-01
1.12921798e+00 -5.07263131e-02 -7.12958276e-01 -4.38841760e-01
-6.05385423e-01 6.93564564e-02 -1.46282196e+00 1.98289347e+00
-4.98922586e-01 8.12492013e-01 -2.10144371e-01 -9.56489682e-01
8.89031172e-01 -4.59662601e-02 7.83449709e-01 -5.26844919e-01
6.16558008e-02 1.74301490e-01 -9.35321674e-02 -6.81078911e-01
7.88183153e-01 6.53882101e-02 1.22240528e-01 -1.69004411e-01
-4.78292741e-02 -6.30112410e-01 1.73368081e-01 6.16563521e-02
9.58087206e-01 2.20925659e-01 1.16687387e-01 -2.30362922e-01
1.08651400e+00 3.05302173e-01 7.04480171e-01 5.89111865e-01
-3.10412556e-01 7.87706792e-01 9.30290204e-03 -9.09561276e-01
-8.39393973e-01 -7.72110164e-01 -7.99572766e-02 5.08704603e-01
8.77341747e-01 -2.65933633e-01 -2.30626404e-01 -5.01891732e-01
3.57534513e-02 4.31520969e-01 -4.49576318e-01 3.89628038e-02
-9.26297069e-01 -4.06115919e-01 8.27642530e-02 6.24914408e-01
9.15155530e-01 -9.31579947e-01 -1.18446374e+00 3.35180134e-01
-2.63446271e-01 -1.57852793e+00 -3.98619503e-01 -8.73826966e-02
-8.52416515e-01 -1.18908703e+00 -5.63862145e-01 -6.92531586e-01
5.91898203e-01 8.32278490e-01 1.00267053e+00 1.09911196e-01
-4.22702044e-01 3.50285530e-01 -3.34090322e-01 -2.57464975e-01
5.52486442e-02 -2.20229357e-01 -6.05226308e-02 1.20436564e-01
4.92616504e-01 -4.04516309e-01 -9.25348580e-01 6.18253171e-01
-7.83314764e-01 8.79264250e-02 2.99503595e-01 6.61685944e-01
5.20459831e-01 2.41206959e-01 -1.16273135e-01 -4.77697611e-01
3.50284241e-02 -2.37092480e-01 -8.19997370e-01 -1.56664565e-01
2.91940849e-03 -2.07695052e-01 4.07118738e-01 -1.85854867e-01
-1.01914227e+00 3.98554921e-01 -2.14113109e-02 -8.13869834e-01
-3.56832147e-01 -1.07766213e-02 2.07134429e-02 -2.54644990e-01
3.44697893e-01 3.92498709e-02 -1.97443500e-01 -5.59743680e-02
3.80978435e-01 2.53001362e-01 6.93876803e-01 -2.01948956e-01
8.77269447e-01 9.80793476e-01 3.21654618e-01 -1.08155859e+00
-6.75608277e-01 -1.02955008e+00 -1.00067031e+00 -4.51037437e-01
1.17410767e+00 -9.85740185e-01 -7.87703514e-01 5.08065104e-01
-1.36905396e+00 -1.65155649e-01 -3.23555142e-01 7.15460539e-01
-7.45621264e-01 4.65949863e-01 -2.51413286e-01 -6.36615634e-01
-4.55289073e-02 -1.44030452e+00 1.13328981e+00 3.44799697e-01
-1.75423194e-02 -1.07820451e+00 -1.10635109e-01 2.67709315e-01
2.69682676e-01 2.78752118e-01 2.91162848e-01 1.97649207e-02
-1.01562476e+00 8.08940828e-02 -3.91079307e-01 2.73640901e-01
2.48052508e-01 1.93884671e-02 -8.34612966e-01 -9.10921767e-02
1.86327130e-01 1.52993008e-01 8.60450625e-01 6.24160051e-01
1.06509256e+00 1.46535367e-01 -2.81963915e-01 1.02607894e+00
1.45708835e+00 2.85967857e-01 6.34480655e-01 3.46959978e-01
1.08606577e+00 8.27125072e-01 9.92296338e-01 4.45863217e-01
3.84174824e-01 7.38026261e-01 7.73200274e-01 -1.90075949e-01
-2.38583356e-01 -1.21271938e-01 6.54071346e-02 5.46336055e-01
-2.96377331e-01 -1.41617104e-01 -9.38197911e-01 6.45755589e-01
-1.96349776e+00 -1.10483408e+00 -6.52628720e-01 2.02194333e+00
-3.63648124e-02 5.59026562e-02 -1.90100372e-01 2.50644892e-01
5.59000015e-01 2.90660053e-01 -5.69912851e-01 -1.77622765e-01
-1.53741747e-01 -1.20919786e-01 7.11760342e-01 6.74847603e-01
-1.26539350e+00 1.17040062e+00 4.90611601e+00 6.62872851e-01
-1.22857261e+00 -8.80883262e-02 2.38053992e-01 -3.74683402e-02
9.23806429e-02 2.17612982e-01 -9.11035240e-01 4.38701361e-01
2.36974254e-01 9.68449339e-02 5.38354777e-02 6.25860274e-01
3.15549850e-01 -2.46553883e-01 -8.45745206e-01 1.26848209e+00
-1.97999831e-02 -1.45360935e+00 -6.99824467e-02 -1.71963945e-02
8.07419121e-01 2.17863590e-01 -1.65317371e-01 5.31963184e-02
3.99094783e-02 -7.23644316e-01 5.92442036e-01 5.16090512e-01
3.67656648e-01 -7.79709041e-01 7.63054907e-01 3.50253135e-01
-1.53386021e+00 -1.13380082e-01 -6.98677897e-01 -3.02651554e-01
5.38762629e-01 6.77102327e-01 -7.94727981e-01 6.97315097e-01
1.08092690e+00 1.35662639e+00 -3.81528199e-01 1.34005034e+00
-1.64484397e-01 2.12760806e-01 -6.26115322e-01 2.22755253e-01
7.39427686e-01 -3.21440399e-01 8.01293910e-01 1.11257637e+00
4.94358331e-01 2.25508943e-01 1.70237243e-01 9.31565642e-01
4.13555145e-01 -1.61823839e-01 -8.37290704e-01 6.74423754e-01
1.38820335e-01 1.30806708e+00 -8.65975440e-01 -3.42656195e-01
-6.23259485e-01 8.15370798e-01 -4.72272001e-02 4.21883732e-01
-8.49804759e-01 -2.99658328e-01 9.86317635e-01 2.76873678e-01
4.94130015e-01 -5.69525480e-01 -2.31539890e-01 -1.18937182e+00
-6.40988722e-02 -1.06729986e-02 2.32930601e-01 -9.83568788e-01
-1.07267272e+00 5.84081054e-01 5.51730767e-02 -1.84367609e+00
-1.14564806e-01 -5.40101051e-01 -6.41377449e-01 7.33581722e-01
-1.86496472e+00 -8.62006783e-01 -8.77992928e-01 9.03059065e-01
8.43497276e-01 1.05057925e-01 1.94754750e-01 3.08320850e-01
-2.63067603e-01 -1.10185854e-01 -3.17937791e-01 -3.83866876e-02
4.09712523e-01 -1.09048486e+00 4.07512307e-01 1.12362397e+00
3.93756665e-03 2.95742303e-01 6.58231318e-01 -5.19716203e-01
-1.21940088e+00 -1.17604160e+00 7.87516415e-01 -4.95132089e-01
4.22854632e-01 -1.97643980e-01 -1.11646557e+00 4.36990649e-01
-6.45760372e-02 5.28769016e-01 8.34241733e-02 -4.75161940e-01
1.97890885e-02 -2.18462870e-01 -1.05494332e+00 3.19204807e-01
1.41515672e+00 -2.07946688e-01 -5.18520534e-01 1.99651137e-01
7.18846023e-01 -6.84899509e-01 -7.29129672e-01 7.55675852e-01
1.96720660e-01 -1.25314033e+00 1.29770625e+00 -2.25488156e-01
3.78762186e-01 -8.86230588e-01 -1.65979385e-01 -1.16872370e+00
-2.18692407e-01 -4.56273466e-01 -7.50800818e-02 9.73162889e-01
-2.18498930e-01 -4.59027231e-01 9.00158048e-01 5.01228452e-01
-4.97187644e-01 -3.64462256e-01 -1.08788240e+00 -6.67924345e-01
-7.84665421e-02 -6.57502174e-01 6.62725747e-01 9.52141285e-01
-4.95972514e-01 2.95307428e-01 -3.15670699e-01 4.35698360e-01
5.19862592e-01 3.16122353e-01 1.15243363e+00 -1.13870513e+00
2.33261272e-01 -4.80910242e-01 -1.01799417e+00 -1.59132218e+00
1.23976089e-01 -6.93855166e-01 1.53771847e-01 -1.62870467e+00
-5.74103475e-01 -6.63154066e-01 1.63643174e-02 9.78651419e-02
-1.74886078e-01 2.37143561e-01 3.59219164e-01 2.04594493e-01
-4.88019943e-01 6.46914303e-01 1.53326881e+00 -1.43250585e-01
-3.73881459e-01 7.02431053e-02 -7.24763051e-02 8.35166752e-01
6.99382305e-01 -2.04037994e-01 -6.72883928e-01 -6.60383463e-01
1.61123183e-03 7.55220950e-02 6.39853954e-01 -1.46726990e+00
5.63707113e-01 -1.84791535e-01 3.44523758e-01 -1.19095886e+00
5.60204983e-01 -1.28446341e+00 4.98337746e-02 4.10871834e-01
1.06049366e-01 2.09769383e-01 3.08409095e-01 7.03147769e-01
-4.82100695e-01 2.04540715e-02 6.65150702e-01 -2.42492259e-01
-1.45399141e+00 6.41896844e-01 -3.36606175e-01 -8.59095380e-02
1.20534778e+00 -6.93007648e-01 -1.82495952e-01 -2.07698911e-01
-3.22586387e-01 3.98762107e-01 3.26178193e-01 8.24887156e-01
9.14458036e-01 -1.07288444e+00 -5.96678674e-01 6.36545718e-01
2.14832723e-01 7.73724139e-01 6.03658438e-01 9.48857725e-01
-8.53369296e-01 4.30018723e-01 -3.08110505e-01 -1.28899813e+00
-1.01672637e+00 5.30020833e-01 3.79365325e-01 1.38067901e-01
-9.99896169e-01 9.63147223e-01 6.00794375e-01 -4.20546591e-01
-2.31902003e-02 -8.61093283e-01 -3.41241032e-01 -1.84732884e-01
5.39392233e-01 5.06589651e-01 5.98964579e-02 -1.10827696e+00
-4.51385438e-01 1.21647704e+00 8.71562734e-02 3.69353771e-01
1.11305106e+00 -4.25003141e-01 8.07490721e-02 3.46119195e-01
1.34572423e+00 -3.09987903e-01 -1.63284004e+00 -1.16582379e-01
-1.78053260e-01 -9.83746767e-01 2.03215525e-01 -1.52452916e-01
-1.39739084e+00 1.20505941e+00 6.93267286e-01 -5.05406335e-02
1.08543789e+00 -2.23851830e-01 8.25774431e-01 2.54190296e-01
4.37227398e-01 -7.20661640e-01 -1.32103413e-01 7.25212634e-01
6.79894030e-01 -1.44290352e+00 4.38635722e-02 -6.82339847e-01
-5.20462871e-01 1.19576836e+00 8.46909583e-01 -4.39665526e-01
6.88678563e-01 6.21722601e-02 3.03466823e-02 -2.56794333e-01
-5.11857271e-01 -4.20183331e-01 3.07046026e-01 6.48108959e-01
1.46690998e-02 -3.64434987e-01 4.93605388e-03 -2.75072426e-01
-2.79611886e-01 4.38484140e-02 4.66374844e-01 9.50035095e-01
-4.99608517e-01 -5.93622983e-01 -3.99025887e-01 1.43754199e-01
6.57069832e-02 2.16202110e-01 1.02118678e-01 9.84643161e-01
4.10700440e-01 1.00048995e+00 6.49575651e-01 -3.62312138e-01
6.36572182e-01 -4.57378864e-01 4.23299104e-01 -2.20860407e-01
-5.43155849e-01 -8.44879895e-02 -3.41662198e-01 -9.80187297e-01
-1.08073175e+00 -6.16310000e-01 -1.49512529e+00 -3.18900377e-01
-1.77981287e-01 3.37733962e-02 6.09088182e-01 7.58358657e-01
3.50299537e-01 5.35880566e-01 5.56974232e-01 -1.28395188e+00
3.18676263e-01 -6.06127679e-01 -4.55129713e-01 6.38140976e-01
6.57996297e-01 -9.09356833e-01 -5.69371879e-01 1.07080981e-01] | [8.533295631408691, -1.9839069843292236] |
49128314-905d-485a-ad89-701c8afe5dcb | hyperspectral-compressive-wavefront-sensing | 2303.03555 | null | https://arxiv.org/abs/2303.03555v1 | https://arxiv.org/pdf/2303.03555v1.pdf | Hyperspectral Compressive Wavefront Sensing | Presented is a novel way to combine snapshot compressive imaging and lateral shearing interferometry in order to capture the spatio-spectral phase of an ultrashort laser pulse in a single shot. A deep unrolling algorithm is utilised for the snapshot compressive imaging reconstruction due to its parameter efficiency and superior speed relative to other methods, potentially allowing for online reconstruction. The algorithm's regularisation term is represented using neural network with 3D convolutional layers, to exploit the spatio-spectral correlations that exist in laser wavefronts. Compressed sensing is not typically applied to modulated signals, but we demonstrate its success here. Furthermore, we train a neural network to predict the wavefronts from a lateral shearing interferogram in terms of Zernike polynomials, which again increases the speed of our technique without sacrificing fidelity. This method is supported with simulation-based results. While applied to the example of lateral shearing interferometry, the methods presented here are generally applicable to a wide range of signals, including Shack-Hartmann-type sensors. The results may be of interest beyond the context of laser wavefront characterization, including within quantitative phase imaging. | ['Andreas Doepp', 'Peter Norreys', 'Robin H. W. Wang', 'Jannik Esslinger', 'Sunny Howard'] | 2023-03-06 | null | null | null | null | ['unrolling'] | ['computer-vision'] | [ 7.85342872e-01 -2.30140418e-01 4.18576658e-01 -1.98329672e-01
-6.93497419e-01 -3.48699003e-01 4.75506485e-01 -5.88070273e-01
-5.01383305e-01 6.88344598e-01 1.03913620e-01 -3.35858047e-01
-6.23050690e-01 -4.12307352e-01 -5.41884303e-01 -9.44434762e-01
-5.78208447e-01 3.47311169e-01 -2.34001756e-01 -1.41930908e-01
3.91147196e-01 8.43454003e-01 -1.59872210e+00 7.43527710e-02
3.31433654e-01 1.15634966e+00 2.68153340e-01 8.16075444e-01
7.02227592e-01 7.53423512e-01 -2.73162931e-01 2.62526184e-01
4.27274704e-01 -6.07469082e-01 -3.14782113e-01 8.35916847e-02
6.12372220e-01 -5.80715835e-01 -6.97126627e-01 7.78753221e-01
6.02247119e-01 9.57271680e-02 4.46466386e-01 -5.32094240e-01
-6.78891689e-02 1.82380080e-01 -3.15310568e-01 3.80622178e-01
4.52474743e-01 5.32469392e-01 7.52162516e-01 -9.28412735e-01
7.69500732e-01 6.80940151e-01 7.33469129e-01 9.94569436e-02
-1.17280602e+00 -4.87308532e-01 -9.05177236e-01 3.21055949e-01
-1.20052898e+00 -8.66756141e-01 8.26369584e-01 -4.97313261e-01
1.01139724e+00 9.70619991e-02 7.80746043e-01 5.79385221e-01
4.31113452e-01 2.86456019e-01 1.41207552e+00 -5.28549433e-01
1.47660151e-01 -6.15615785e-01 -2.93158203e-01 6.77211463e-01
2.03063950e-01 8.57183576e-01 -7.47842550e-01 8.30534697e-02
7.71420836e-01 -6.12441562e-02 -8.53902578e-01 -2.93240756e-01
-1.25564659e+00 5.59288383e-01 5.85335732e-01 4.15466219e-01
-6.57581091e-01 4.24986273e-01 -4.63880785e-02 6.00676298e-01
3.35555971e-01 1.00491059e+00 5.41069396e-02 -9.51370373e-02
-1.59406400e+00 3.06414455e-01 7.72384048e-01 4.89372492e-01
1.09933054e+00 2.53175795e-01 -7.41371736e-02 1.42037764e-01
2.07386300e-01 6.97064400e-01 3.04052114e-01 -1.23255002e+00
-1.93416625e-02 -8.89101699e-02 2.79297471e-01 -7.21434951e-01
-4.86831784e-01 -6.37767136e-01 -9.79609489e-01 5.79711795e-01
2.71557182e-01 -2.66176343e-01 -9.37354267e-01 1.35299444e+00
-4.08296883e-02 8.21757793e-01 6.30300166e-03 1.23766005e+00
1.82867795e-01 3.74830812e-01 -7.77313888e-01 -5.94388962e-01
1.08482277e+00 -2.36044884e-01 -6.33007348e-01 -4.09718424e-01
2.69564122e-01 -8.84305000e-01 2.57116616e-01 4.91966277e-01
-1.19537830e+00 -1.81332096e-01 -1.14393902e+00 4.00741324e-02
1.69867516e-01 -5.83586171e-02 3.83018017e-01 2.58166522e-01
-1.14809585e+00 1.11517501e+00 -8.97849083e-01 8.48379582e-02
5.50074279e-01 3.47946346e-01 -2.01147497e-01 -3.41810167e-01
-9.07140970e-01 8.99877667e-01 1.73429087e-01 2.71621943e-01
-5.51943362e-01 -8.66629601e-01 -5.76254427e-01 1.24899067e-01
1.20476909e-01 -8.34179759e-01 1.10655439e+00 -3.01692009e-01
-1.72371495e+00 6.77331328e-01 -1.74057767e-01 -7.80117273e-01
1.96895495e-01 1.41588539e-01 -8.58341455e-02 6.33238018e-01
-1.08246543e-01 4.57423925e-02 1.14089870e+00 -8.75865698e-01
-2.02762544e-01 -2.30296671e-01 -2.86950350e-01 8.98419693e-03
3.09901208e-01 -3.24225098e-01 1.01879850e-01 -3.15389901e-01
4.58100766e-01 -9.87905800e-01 -3.16220164e-01 -1.93156645e-01
-1.24427401e-01 5.35505295e-01 9.03101742e-01 -3.42474073e-01
9.28498983e-01 -1.95674515e+00 2.00002432e-01 1.56182731e-02
2.00282618e-01 3.92441988e-01 -1.18345387e-01 9.98317122e-01
-1.44079030e-01 -4.78298753e-01 -7.44446993e-01 -4.47423220e-01
-3.68272454e-01 -1.07870474e-01 -3.97025704e-01 8.61109197e-01
4.35241818e-01 9.82253611e-01 -8.41081083e-01 2.55575776e-01
3.17257762e-01 4.97410804e-01 -5.87184608e-01 3.44857752e-01
-2.61666328e-01 9.95766640e-01 -1.50355384e-01 4.26764935e-01
7.06915915e-01 -2.86201566e-01 -1.40219657e-02 -3.80092680e-01
-5.12394190e-01 2.42007360e-01 -9.25682724e-01 1.91712689e+00
-4.72899675e-01 1.16608167e+00 5.94809115e-01 -1.04023373e+00
6.95363104e-01 5.44519544e-01 5.32987118e-01 -8.18133891e-01
3.45701911e-02 5.09141088e-01 1.80609897e-01 -9.47128475e-01
5.16099095e-01 -8.66673768e-01 4.20551986e-01 7.82235444e-01
5.26874177e-02 -6.46989524e-01 2.80150995e-02 -1.67854894e-02
1.23336232e+00 1.89466298e-01 2.24760145e-01 -1.64483473e-01
1.44285530e-01 -1.38796061e-01 3.68447341e-02 6.83244586e-01
2.36776128e-01 7.56187737e-01 7.72462189e-02 -3.07113826e-01
-1.22158170e+00 -5.12514114e-01 -3.19073677e-01 1.34942725e-01
-9.12182778e-02 -3.05269286e-02 -2.07628280e-01 3.26015562e-01
-1.32987678e-01 3.95493478e-01 -1.82455927e-01 1.27885580e-01
-7.41464198e-01 -5.85764706e-01 3.18414122e-01 -3.93025428e-02
4.45710421e-01 -8.61898422e-01 -1.33352280e+00 3.64050984e-01
5.55079058e-02 -1.36485112e+00 1.02022067e-01 1.36912838e-01
-8.11549604e-01 -1.41627479e+00 -6.67190909e-01 -3.07833016e-01
2.05500156e-01 4.97599304e-01 8.29563558e-01 -1.55216008e-01
-4.57754016e-01 6.80104196e-01 -1.87738761e-01 -1.90994024e-01
-3.30537289e-01 -4.13204700e-01 1.13869473e-01 1.06434979e-01
9.41389129e-02 -1.17502391e+00 -1.00677049e+00 -2.17248172e-01
-1.09108818e+00 5.91821373e-02 9.04680133e-01 9.47228551e-01
2.01086402e-01 -1.76725257e-02 1.96226701e-01 -8.15831661e-01
5.12338281e-01 -3.86694282e-01 -7.78651774e-01 -5.07122219e-01
-6.29084468e-01 -7.51167238e-02 5.43586612e-01 -8.95011947e-02
-5.85374832e-01 -7.51214921e-02 -1.10668056e-01 -5.82514882e-01
-7.53703117e-02 1.00932181e+00 6.62702858e-01 -6.11997783e-01
7.35377014e-01 4.98245358e-01 5.88073492e-01 -2.42732897e-01
7.67349079e-02 5.13612151e-01 6.76622510e-01 6.33539334e-02
1.03224468e+00 7.17433095e-01 8.61497581e-01 -1.36594331e+00
-5.87314725e-01 -6.51293218e-01 -4.23542410e-01 -1.10172965e-01
3.88061732e-01 -8.53836417e-01 -1.01807201e+00 5.22777855e-01
-1.00944436e+00 -2.55198032e-01 -3.77661973e-01 9.06524539e-01
-8.90853167e-01 5.59415102e-01 -6.67461574e-01 -9.64520752e-01
-2.29762062e-01 -9.70483303e-01 1.05486286e+00 1.17182314e-01
2.68765599e-01 -7.21946001e-01 6.16552904e-02 1.38778389e-01
8.87541354e-01 3.51785511e-01 6.49358034e-01 3.10358033e-02
-1.22726274e+00 -5.01289070e-01 -1.16392180e-01 3.66300642e-01
-2.09830135e-01 -4.79734510e-01 -1.12779653e+00 -7.00612366e-01
6.15833879e-01 -3.42679709e-01 1.00444818e+00 6.42381608e-01
6.24089897e-01 -1.58461213e-01 -7.21532255e-02 8.24526191e-01
1.63500130e+00 -3.45369093e-02 6.89614594e-01 1.66042130e-02
3.62414300e-01 4.32118475e-01 2.23223165e-01 5.09938180e-01
-2.22930625e-01 6.82406425e-01 6.60156846e-01 1.17380552e-01
-2.69277066e-01 3.03610295e-01 1.40395552e-01 6.44826710e-01
-2.53006727e-01 -2.39060313e-01 -7.45096326e-01 3.62195849e-01
-1.38166106e+00 -1.29731071e+00 -1.74409673e-01 2.38069391e+00
2.19692796e-01 -2.63728291e-01 -4.06695932e-01 1.77751124e-01
2.12770998e-01 5.19489229e-01 -4.00453091e-01 2.22532555e-01
-1.20672375e-01 7.80568957e-01 4.63191450e-01 7.50065506e-01
-6.78296685e-01 3.69308293e-01 6.27224827e+00 2.30715364e-01
-1.64554667e+00 -9.39053521e-02 -1.02105960e-01 -1.65734529e-01
-2.44226202e-01 2.57385015e-01 -1.64639086e-01 4.34222788e-01
1.03174925e+00 -1.59727852e-03 9.24652874e-01 -2.20584959e-01
4.12187606e-01 -3.34164202e-01 -1.07541847e+00 1.09303856e+00
-1.19460419e-01 -1.75445056e+00 -4.68889952e-01 1.53405130e-01
5.00384748e-01 5.18298984e-01 2.06067786e-01 -1.67262226e-01
-4.84980732e-01 -1.12324917e+00 2.06737995e-01 8.15138042e-01
1.11058331e+00 -3.50877315e-01 5.81307709e-01 7.88071811e-01
-6.29816473e-01 -5.29894158e-02 -9.65059698e-02 -5.69736362e-01
4.98545438e-01 8.85300159e-01 -1.13012671e+00 6.89390481e-01
1.70926347e-01 8.64825904e-01 8.75650644e-02 1.07662261e+00
-1.47782400e-01 7.73952186e-01 -5.09958208e-01 2.27815732e-01
2.96596646e-01 -4.63043481e-01 1.09778154e+00 9.24112022e-01
9.05019641e-01 4.41229165e-01 -1.62905842e-01 8.31701756e-01
1.49048805e-01 -6.46936595e-01 -8.69443953e-01 -3.30847174e-01
3.10450703e-01 1.15702462e+00 -2.54908204e-01 2.13616528e-02
-2.87131518e-01 6.16475582e-01 -1.24818876e-01 6.56801045e-01
2.13731825e-02 -2.80230939e-01 4.96355385e-01 6.47908509e-01
5.49138606e-01 -6.85326755e-01 8.59147832e-02 -1.06029487e+00
-1.49555162e-01 -6.11899853e-01 -2.42494360e-01 -1.00484622e+00
-1.00496018e+00 4.18225229e-01 -1.81253571e-02 -1.35122156e+00
-5.08005559e-01 -6.54992461e-01 -6.25304043e-01 1.11309278e+00
-2.00624728e+00 -9.45520282e-01 -3.06435734e-01 2.66804755e-01
-3.91622484e-02 3.54052745e-02 9.00904953e-01 1.63593590e-01
-6.35927096e-02 -3.75014991e-01 2.05793038e-01 -3.21398854e-01
5.71690738e-01 -9.69919384e-01 1.47335604e-01 1.10088515e+00
-4.94090421e-03 7.29273617e-01 1.18168294e+00 -3.77014697e-01
-1.95492387e+00 -7.28830040e-01 6.49437308e-01 1.35542871e-02
6.20340705e-01 9.22393575e-02 -7.35605061e-01 6.14496529e-01
4.93665159e-01 3.97763789e-01 5.80546975e-01 -2.59606659e-01
-4.21963483e-02 3.60920765e-02 -1.04187918e+00 1.51007706e-02
7.60812223e-01 -7.16352701e-01 -3.53990495e-01 3.97739440e-01
1.61957607e-01 -6.92310691e-01 -5.88719726e-01 5.85580170e-01
6.25207722e-01 -1.43906128e+00 8.47761452e-01 -1.60577431e-01
7.88185418e-01 -4.45739746e-01 -1.86728999e-01 -1.54194283e+00
-4.68242109e-01 -1.30667472e+00 -3.86905104e-01 2.91863173e-01
-1.63651437e-01 -9.45435107e-01 9.46671009e-01 -1.65182486e-01
-4.89930332e-01 -7.48996079e-01 -1.27417779e+00 -5.73727310e-01
-1.60045281e-01 -4.50447828e-01 3.10219824e-01 7.11222947e-01
-2.27847695e-01 4.41007912e-01 -6.22252345e-01 6.04156256e-01
9.42292929e-01 4.92561668e-01 4.78596747e-01 -1.12988412e+00
-6.05587006e-01 -3.38439822e-01 -5.84530115e-01 -1.33266342e+00
-3.72699015e-02 -1.00007772e+00 2.92162687e-01 -1.08543205e+00
-2.74232566e-01 -4.64067727e-01 -9.82098877e-02 -1.92186669e-01
1.67725831e-01 6.33390307e-01 2.10564598e-01 3.87556255e-01
7.77587295e-02 5.01897156e-01 1.30939484e+00 1.03003561e-01
1.53545588e-02 2.38110110e-01 -2.41625354e-01 3.11181694e-01
4.82246399e-01 -3.14054251e-01 -3.32169056e-01 -4.59487110e-01
4.38801944e-01 5.93312979e-01 6.94291651e-01 -1.38827670e+00
7.95344770e-01 1.43323004e-01 9.45885181e-02 -2.84999222e-01
7.65663028e-01 -7.91422307e-01 4.43604380e-01 5.73819816e-01
1.05008669e-02 -3.84793788e-01 -1.15399547e-01 7.16935158e-01
-3.64923775e-01 -2.24270716e-01 9.84767377e-01 -2.81586021e-01
-4.83652860e-01 4.98803914e-01 -4.54525620e-01 -2.27579549e-01
6.45353496e-01 -2.93367714e-01 -1.72895223e-01 -5.47281086e-01
-4.28566843e-01 -1.31909043e-01 5.30960798e-01 -5.61016023e-01
7.43520498e-01 -9.96301055e-01 -8.62769842e-01 8.18334818e-01
-9.95811373e-02 -1.31123941e-02 4.26863790e-01 1.23580706e+00
-8.66507053e-01 5.40952802e-01 -7.61538744e-02 -7.31225371e-01
-8.64240229e-01 2.75820911e-01 4.83774483e-01 -1.63212325e-02
-8.01718116e-01 8.36834669e-01 -3.04969430e-01 7.78951123e-02
-3.49075586e-01 -2.51800716e-01 5.81293628e-02 -3.87304947e-02
4.72364694e-01 2.44216263e-01 2.18796939e-01 -5.36735415e-01
-1.95526555e-01 8.04166853e-01 3.60365182e-01 -2.09365904e-01
1.60053849e+00 -6.39724061e-02 -2.86385566e-01 3.33431035e-01
1.34998572e+00 2.03709736e-01 -1.53881359e+00 -3.14286143e-01
-2.37899527e-01 -5.18287599e-01 4.51529860e-01 -4.38733131e-01
-7.90615439e-01 9.77651775e-01 4.89870131e-01 4.51863527e-01
1.33072519e+00 -2.22412407e-01 8.46966147e-01 4.88406688e-01
5.24960935e-01 -2.89561212e-01 -8.30467865e-02 6.21607006e-01
6.28111362e-01 -9.95036066e-01 3.19794565e-01 -9.47378650e-02
1.07599646e-01 1.36441720e+00 -5.16839027e-01 -4.81511235e-01
5.91769993e-01 3.69070530e-01 -9.20108110e-02 -5.29033482e-01
-6.70729756e-01 -1.37886897e-01 1.61618024e-01 4.47776616e-01
3.94063175e-01 -1.26863986e-01 -1.47387743e-01 -4.90331024e-01
-1.03517011e-01 3.77185732e-01 8.93844068e-01 9.52341199e-01
-6.61620557e-01 -1.03259146e+00 -2.29041040e-01 5.75889885e-01
-4.34190631e-01 -1.27706349e-01 2.42415607e-01 4.57819611e-01
-6.65812567e-02 6.61142051e-01 -1.48325413e-02 -9.97240618e-02
1.63432643e-01 3.41887251e-02 9.58603144e-01 -7.19995022e-01
-2.00160399e-01 1.78838298e-01 -1.07427225e-01 -6.20872080e-01
-7.30860233e-01 -5.81130862e-01 -8.76785100e-01 -2.05841303e-01
-1.17836758e-01 1.35200247e-01 7.69119799e-01 1.01344526e+00
5.54860890e-01 2.13199601e-01 7.51893222e-01 -1.39389324e+00
-5.53399980e-01 -1.05370009e+00 -8.43560398e-01 9.31075960e-02
1.16116476e+00 -3.49233985e-01 -5.85411608e-01 -2.01066181e-01] | [11.256460189819336, -2.4292893409729004] |
78340be4-33ff-45f4-942f-13220c84e593 | video-compressive-sensing-for-dynamic-mri | 1401.7715 | null | http://arxiv.org/abs/1401.7715v2 | http://arxiv.org/pdf/1401.7715v2.pdf | Video Compressive Sensing for Dynamic MRI | We present a video compressive sensing framework, termed kt-CSLDS, to
accelerate the image acquisition process of dynamic magnetic resonance imaging
(MRI). We are inspired by a state-of-the-art model for video compressive
sensing that utilizes a linear dynamical system (LDS) to model the motion
manifold. Given compressive measurements, the state sequence of an LDS can be
first estimated using system identification techniques. We then reconstruct the
observation matrix using a joint structured sparsity assumption. In particular,
we minimize an objective function with a mixture of wavelet sparsity and joint
sparsity within the observation matrix. We derive an efficient convex
optimization algorithm through alternating direction method of multipliers
(ADMM), and provide a theoretical guarantee for global convergence. We
demonstrate the performance of our approach for video compressive sensing, in
terms of reconstruction accuracy. We also investigate the impact of various
sampling strategies. We apply this framework to accelerate the acquisition
process of dynamic MRI and show it achieves the best reconstruction accuracy
with the least computational time compared with existing algorithms in the
literature. | ['Wotao Yin', 'Jianing V. Shi', 'Richard G. Baraniuk', 'Aswin C. Sankaranarayanan'] | 2014-01-30 | null | null | null | null | ['video-compressive-sensing'] | ['computer-vision'] | [ 7.36401796e-01 -4.53787483e-02 -1.68688759e-01 2.02579126e-01
-7.13570952e-01 -2.51910329e-01 2.30080619e-01 -4.72592562e-01
-3.86168897e-01 3.83070141e-01 2.63839602e-01 -2.78559178e-01
-4.08284873e-01 -3.20837013e-02 -7.98220277e-01 -8.71273756e-01
-2.51535535e-01 1.25257269e-01 -2.12592736e-01 -3.19456495e-02
1.41115114e-01 1.62718266e-01 -7.90732980e-01 3.44921350e-02
6.26199424e-01 9.13004518e-01 5.69229186e-01 7.17781067e-01
7.14935362e-01 1.11628616e+00 8.50202218e-02 3.34907055e-01
5.52396417e-01 -7.06696153e-01 -7.42747724e-01 5.90994179e-01
6.38275892e-02 -6.08841717e-01 -8.63651216e-01 1.11647236e+00
5.57751238e-01 6.21907748e-02 3.09103310e-01 -8.47862959e-01
-2.91637272e-01 3.79503250e-01 -7.09488809e-01 4.59999293e-01
2.23399565e-01 -8.74133557e-02 3.88413280e-01 -1.10687530e+00
9.88584161e-01 7.53971398e-01 6.01608574e-01 4.42165822e-01
-1.32789445e+00 -2.14600727e-01 -9.98365805e-02 4.10780698e-01
-1.32820427e+00 -9.26934779e-01 8.54658067e-01 -4.50991303e-01
4.06348765e-01 2.50212014e-01 7.32038915e-01 7.68008769e-01
4.14577782e-01 9.20476019e-01 1.19708085e+00 -6.31749213e-01
4.56956595e-01 -5.08837759e-01 -7.57315606e-02 8.05618167e-01
1.83282077e-01 3.61564383e-02 -7.84343123e-01 -3.15259099e-01
1.10490620e+00 2.03203231e-01 -6.75556839e-01 -5.48475146e-01
-1.53775609e+00 7.87955046e-01 -4.42280956e-02 4.73859638e-01
-7.90681183e-01 3.11750263e-01 6.21008240e-02 3.87876213e-01
2.97393620e-01 -4.60918620e-02 2.07549661e-01 1.10843726e-01
-1.30921280e+00 3.18132713e-02 7.94413507e-01 5.67045093e-01
1.65018231e-01 3.22201669e-01 3.79121676e-02 6.21225357e-01
4.22376603e-01 7.35058427e-01 5.89002490e-01 -1.51502788e+00
2.24296466e-01 -3.63351226e-01 5.20659313e-02 -1.05292141e+00
-2.00662315e-01 -5.71897328e-01 -1.08320594e+00 -1.67016253e-01
-2.70184316e-02 -1.33559167e-01 -4.91303891e-01 1.71391559e+00
4.45298195e-01 8.54990005e-01 1.21827014e-01 1.00394475e+00
1.30967170e-01 6.82883620e-01 -6.83290839e-01 -1.03880644e+00
9.22103643e-01 -8.10833871e-01 -1.08503509e+00 -1.73699349e-01
4.24582481e-01 -6.17636919e-01 2.75103986e-01 5.39454579e-01
-1.49868095e+00 -1.27790615e-01 -1.09069037e+00 4.36428189e-01
7.89332926e-01 5.89958802e-02 1.98405042e-01 3.78706783e-01
-1.14204335e+00 6.33482695e-01 -1.59364259e+00 1.37339801e-01
1.04336739e-01 2.92569041e-01 -3.17307144e-01 -5.12527943e-01
-7.95068800e-01 7.63388216e-01 5.49285635e-02 1.60889447e-01
-1.18136239e+00 -1.01469016e+00 -7.71486580e-01 -1.91918582e-01
4.65804458e-01 -8.76330674e-01 1.02942073e+00 -7.17099547e-01
-1.43504727e+00 5.54724395e-01 -3.85038555e-01 -6.84808612e-01
5.50968230e-01 -2.57567853e-01 -1.48822084e-01 9.05632138e-01
2.33876822e-03 1.10615857e-01 1.41663039e+00 -1.09974182e+00
1.03896268e-01 -2.49402821e-01 -3.90450925e-01 1.30619913e-01
-2.97449887e-01 -1.77517474e-01 -2.42961884e-01 -9.13603842e-01
6.42016172e-01 -1.26535296e+00 -6.05110884e-01 -1.71175413e-02
-2.10365072e-01 7.27100074e-01 7.63059020e-01 -1.11907172e+00
1.49612832e+00 -2.12626624e+00 7.79757142e-01 2.38795668e-01
4.65441525e-01 8.52318704e-02 -2.31058616e-02 5.31434298e-01
-2.27157772e-01 -6.39003873e-01 -5.64245880e-01 -4.12777901e-01
-5.77370465e-01 2.03380287e-01 -2.32684538e-01 1.08955717e+00
-3.29632878e-01 7.00043023e-01 -8.98182452e-01 -3.51873517e-01
2.41110906e-01 5.45128763e-01 -8.39137435e-01 1.04190513e-01
3.28756988e-01 9.15937126e-01 -5.15030742e-01 3.34488451e-01
7.06517220e-01 -5.80732703e-01 6.90791547e-01 -4.77377385e-01
-2.57794261e-02 -3.62182111e-01 -1.39272439e+00 2.13744998e+00
-3.53115678e-01 3.59714031e-01 7.80666471e-01 -1.60661638e+00
1.84325054e-01 5.81738532e-01 1.32798684e+00 -3.41414481e-01
3.61438245e-02 3.88870418e-01 -1.31368563e-01 -7.68121123e-01
2.19202369e-01 -3.38998467e-01 3.90705854e-01 5.70085168e-01
-1.70840904e-01 1.00305572e-01 -6.30715024e-03 5.40090144e-01
1.39635146e+00 -1.31106660e-01 3.03147584e-01 -5.72627604e-01
5.36975563e-01 -2.56764948e-01 3.60662013e-01 7.61137605e-01
2.36402825e-02 5.32606483e-01 6.53496310e-02 -1.67453185e-01
-1.19541907e+00 -8.15142393e-01 -2.33334869e-01 3.78673553e-01
-4.21837196e-02 -2.78291970e-01 -7.89721489e-01 2.45285947e-02
-3.42898726e-01 2.09580198e-01 -4.10730988e-01 -1.07245492e-02
-8.80287111e-01 -8.34896505e-01 1.51646696e-02 -4.45511416e-02
2.26554155e-01 -4.40053493e-01 -9.57127929e-01 4.17704880e-01
-6.43476427e-01 -1.54341507e+00 -6.58150911e-01 -2.75570035e-01
-1.25528717e+00 -9.31948125e-01 -1.01283431e+00 -6.77686691e-01
8.90299141e-01 6.80680990e-01 5.97796500e-01 1.55072719e-01
-2.85794258e-01 9.37385023e-01 -2.56916553e-01 2.74990797e-01
-5.91182709e-01 -4.28407460e-01 4.51172888e-01 5.40379345e-01
-6.52372956e-01 -8.03344131e-01 -9.27979112e-01 7.04153627e-02
-1.36654460e+00 2.37851143e-01 6.00737572e-01 1.07746601e+00
8.12472701e-01 -1.33169249e-01 2.84480959e-01 -7.42257059e-01
3.13156724e-01 -7.39465415e-01 -4.61026490e-01 1.15552261e-01
-5.36025286e-01 1.35033056e-01 3.68086785e-01 -5.20272195e-01
-5.85573733e-01 2.92355090e-01 1.20508276e-01 -1.06009686e+00
5.98671317e-01 8.80063474e-01 4.50753361e-01 -5.82916498e-01
3.04803789e-01 8.42655599e-01 6.32687867e-01 -3.47189486e-01
2.17825487e-01 3.03450078e-01 7.44074702e-01 -4.74028289e-01
6.39835477e-01 1.10524464e+00 3.55692118e-01 -1.31906903e+00
-7.96116114e-01 -6.77146554e-01 -4.34767365e-01 -4.09051150e-01
5.28592527e-01 -1.03941154e+00 -5.57147026e-01 3.93092602e-01
-7.80332625e-01 -2.29437798e-01 -2.67839730e-01 9.74544108e-01
-1.02196419e+00 9.35842633e-01 -9.77367163e-01 -7.44269311e-01
-2.85989642e-01 -1.13988221e+00 8.57433856e-01 -5.85854769e-01
8.53647739e-02 -1.00458252e+00 3.03957611e-01 4.10716712e-01
4.87287462e-01 5.21906316e-01 4.47004765e-01 1.74629077e-01
-7.16439605e-01 7.38191232e-02 3.52410376e-01 3.82043064e-01
-1.17169708e-01 -6.71018302e-01 -2.89986551e-01 -9.63368952e-01
9.86026108e-01 -3.12821865e-02 7.54117846e-01 9.42830920e-01
1.05494583e+00 -6.66647792e-01 -2.73997962e-01 8.69380355e-01
1.71456099e+00 -7.60451555e-02 5.98051727e-01 1.77632838e-01
4.48052347e-01 2.03519449e-01 3.96653473e-01 8.69900703e-01
1.28887668e-01 5.59203088e-01 3.23948354e-01 2.82085150e-01
-9.24782362e-03 5.17243613e-03 4.90832627e-01 1.62422943e+00
-1.53992638e-01 2.56450802e-01 -6.47234261e-01 7.04942763e-01
-1.88386524e+00 -9.96970236e-01 3.68067957e-02 2.20294356e+00
5.90593994e-01 -4.67211127e-01 1.19738346e-02 3.77417356e-01
5.77930748e-01 2.07153350e-01 -6.28646255e-01 4.37150210e-01
2.78508775e-02 3.00887316e-01 6.67344749e-01 7.70175993e-01
-7.90568709e-01 3.70805293e-01 6.83695889e+00 5.87230265e-01
-1.11330497e+00 5.45643210e-01 3.01992148e-01 -3.99342269e-01
-1.58312216e-01 -1.75801211e-03 -1.34653002e-01 4.41814631e-01
9.95122015e-01 -1.50795698e-01 8.77872050e-01 3.50848019e-01
5.10267615e-01 4.81375381e-02 -7.18209386e-01 1.34950137e+00
3.54995281e-01 -1.68295670e+00 -2.01209292e-01 3.40354651e-01
1.09807026e+00 3.95581089e-02 1.16019182e-01 -5.07034838e-01
-1.80522352e-01 -5.97436488e-01 8.46924424e-01 5.16452253e-01
8.87544692e-01 -2.62522727e-01 1.73246652e-01 4.98774678e-01
-9.22890604e-01 -1.48265213e-01 -1.76787004e-01 9.13366154e-02
8.11826110e-01 8.42021286e-01 -2.34994665e-01 6.01423800e-01
2.81135648e-01 9.59602892e-01 2.38666862e-01 7.88708031e-01
2.02903524e-01 7.58934915e-01 -4.58026916e-01 6.36134744e-01
2.16167375e-01 -5.21487176e-01 1.02781701e+00 7.57862210e-01
6.37729585e-01 6.20995760e-01 3.60516578e-01 3.14657867e-01
1.97017446e-01 -8.28674287e-02 -4.75378036e-01 5.12111373e-02
3.67873549e-01 9.47876573e-01 -5.09325206e-01 -3.09143126e-01
-3.15948606e-01 1.04794836e+00 -1.32237568e-01 3.91095221e-01
-5.18460274e-01 5.56974649e-01 1.88125074e-01 2.32463777e-01
6.06167555e-01 -6.34360969e-01 1.24157548e-01 -1.73428679e+00
2.07155257e-01 -1.15250146e+00 2.90686727e-01 -5.25099099e-01
-7.84316599e-01 3.46595913e-01 1.05450541e-01 -1.47026992e+00
-2.57708877e-01 -2.32392162e-01 -2.32867241e-01 2.99885154e-01
-1.40102816e+00 -7.72670448e-01 -7.65059590e-02 9.72520709e-01
4.72212166e-01 -4.00215127e-02 4.29752707e-01 4.22038883e-01
-3.24462146e-01 1.45030573e-01 5.85968554e-01 -2.66538471e-01
2.18916342e-01 -6.46546006e-01 -2.47322574e-01 1.15108418e+00
-9.70510393e-02 6.58320069e-01 9.15444374e-01 -6.33475900e-01
-2.15322113e+00 -7.62803376e-01 3.85166526e-01 1.99713454e-01
9.92075980e-01 1.52660713e-01 -6.43365562e-01 8.85454774e-01
4.80435155e-02 2.11055055e-01 6.61397517e-01 -6.55581415e-01
1.02072291e-01 1.22943491e-01 -1.18197751e+00 2.56076843e-01
1.00104356e+00 -4.60019708e-01 -4.57656026e-01 6.32214308e-01
4.34352219e-01 -7.08680153e-01 -1.11202931e+00 1.67795762e-01
6.37625456e-01 -6.40793204e-01 1.26675415e+00 -4.01724607e-01
6.58839762e-01 -2.58263111e-01 -5.18879533e-01 -1.21651316e+00
-3.82045239e-01 -1.15780139e+00 -8.57161760e-01 1.69008717e-01
-1.48373589e-01 -5.04769385e-01 7.55572736e-01 1.53959572e-01
-1.25342309e-01 -8.37656379e-01 -1.26336098e+00 -8.36215615e-01
-2.18418092e-01 -3.97391766e-01 -2.91574329e-01 1.08378160e+00
2.00552449e-01 3.01544033e-02 -1.01177418e+00 3.19027692e-01
1.33736861e+00 1.19038433e-01 1.31778166e-01 -4.80165362e-01
-9.24532056e-01 1.58353299e-01 -3.45756799e-01 -1.29854667e+00
-1.37561923e-02 -9.51142311e-01 -1.07230872e-01 -1.11762893e+00
3.99070561e-01 -2.62334824e-01 -2.92006463e-01 -1.02638826e-01
-1.02275563e-02 2.44375780e-01 4.82191026e-01 7.41817594e-01
-5.44121802e-01 4.59949613e-01 1.33517754e+00 8.09722692e-02
4.51916531e-02 -2.12369815e-01 -4.55834091e-01 5.04519522e-01
2.85575658e-01 -5.49237669e-01 -4.89503205e-01 -6.70393407e-01
9.85658020e-02 9.67208564e-01 2.46812657e-01 -1.00905728e+00
4.81780499e-01 -7.13630542e-02 -1.54970199e-01 -3.27639848e-01
4.63217884e-01 -8.19716990e-01 5.48931003e-01 1.20674431e+00
-2.89487541e-01 4.54642400e-02 -3.20875823e-01 8.15539896e-01
-1.31616965e-01 -3.83344144e-01 1.00234783e+00 -1.29789829e-01
-2.68222868e-01 5.29143929e-01 -5.40031433e-01 -5.38126305e-02
8.23526382e-01 1.10470809e-01 2.26682216e-01 -5.29551625e-01
-1.16022670e+00 -7.42356246e-03 2.38202423e-01 -1.75030559e-01
9.23464894e-01 -1.41566825e+00 -9.70895469e-01 2.32880697e-01
-3.90894443e-01 -3.67275894e-01 7.17212081e-01 1.52753258e+00
-7.31213927e-01 2.56586641e-01 -1.78290475e-02 -8.44304144e-01
-1.04777670e+00 6.66519284e-01 2.94510007e-01 -4.80412513e-01
-9.42187846e-01 3.93712640e-01 -3.31571884e-02 7.39292651e-02
-1.88725621e-01 7.60703236e-02 1.67864457e-01 -4.07354355e-01
9.61609006e-01 5.96810758e-01 -4.43698019e-02 -6.26237750e-01
-2.44130924e-01 6.58282220e-01 1.49601236e-01 -5.25755346e-01
1.68219185e+00 -3.98054451e-01 -2.24623829e-01 1.63442835e-01
1.38592935e+00 -3.37645374e-02 -1.35469139e+00 -5.28679013e-01
-3.27909827e-01 -5.56666493e-01 4.97056305e-01 -5.15533537e-02
-1.32135952e+00 3.65081191e-01 6.31321967e-01 4.37159352e-02
1.22482848e+00 -1.80034623e-01 1.04920948e+00 2.86784828e-01
5.40059984e-01 -7.99820602e-01 1.69961631e-01 1.77557632e-01
8.62206697e-01 -9.63622630e-01 3.53815019e-01 -4.44200367e-01
-3.16406995e-01 9.76747096e-01 -4.25136209e-01 -6.66169822e-01
9.37749028e-01 3.97056103e-01 -3.59935105e-01 -2.45112047e-01
-6.66331947e-01 3.70597690e-01 1.20699398e-01 2.30472624e-01
1.04841717e-01 2.44858861e-02 -6.07930779e-01 -5.57618551e-02
1.71478108e-01 4.17320192e-01 8.64044726e-01 1.21565390e+00
-2.56164581e-01 -9.44007933e-01 -5.36594093e-01 3.59467387e-01
-7.14970887e-01 -1.03392161e-01 4.04318005e-01 1.95033520e-01
-5.77661097e-01 8.36698472e-01 -3.87202263e-01 -2.45016776e-02
-1.48903310e-01 -2.31199577e-01 1.05973768e+00 -2.63940960e-01
7.24164546e-02 4.52323407e-01 -3.24507535e-01 -7.45052457e-01
-9.34087455e-01 -1.04904115e+00 -7.80260205e-01 -3.99055690e-01
-1.76802963e-01 2.41114646e-01 6.90033078e-01 1.07200027e+00
3.44941944e-01 4.24007624e-01 9.77287292e-01 -8.53181124e-01
-9.21329498e-01 -6.63183630e-01 -6.80604637e-01 3.57642591e-01
5.74010193e-01 -2.83343524e-01 -2.67570496e-01 4.24988955e-01] | [11.713347434997559, -2.2345871925354004] |
75968e3d-b942-4160-8cf7-a851f63fe257 | a-review-of-deep-learning-for-video | 2304.11431 | null | https://arxiv.org/abs/2304.11431v1 | https://arxiv.org/pdf/2304.11431v1.pdf | A Review of Deep Learning for Video Captioning | Video captioning (VC) is a fast-moving, cross-disciplinary area of research that bridges work in the fields of computer vision, natural language processing (NLP), linguistics, and human-computer interaction. In essence, VC involves understanding a video and describing it with language. Captioning is used in a host of applications from creating more accessible interfaces (e.g., low-vision navigation) to video question answering (V-QA), video retrieval and content generation. This survey covers deep learning-based VC, including but, not limited to, attention-based architectures, graph networks, reinforcement learning, adversarial networks, dense video captioning (DVC), and more. We discuss the datasets and evaluation metrics used in the field, and limitations, applications, challenges, and future directions for VC. | ['Fatih Porikli', 'Erik Cambria', 'Abbas Khosravi', 'Abduallah Mohamed', 'Shuicheng Yan', 'Mohammad Ghavamzadeh', 'Daniel McDuff', 'Farhad Pourpanah', 'Swaraja Kuraparthi', 'Meenakshi Kollati', 'Moloud Abdar'] | 2023-04-22 | null | null | null | null | ['video-captioning', 'dense-video-captioning', 'video-question-answering', 'video-retrieval'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [ 4.49526310e-01 3.40736806e-01 -1.70704082e-01 -3.01616527e-02
-7.91407049e-01 -7.90468931e-01 6.19124472e-01 2.44294330e-01
-1.93373114e-01 6.49776638e-01 6.09620094e-01 -4.42304343e-01
1.64879799e-01 -5.10934174e-01 -9.10080492e-01 -1.43388122e-01
-5.25029562e-02 5.39589524e-01 1.98591575e-01 -3.79129231e-01
2.48067379e-01 1.20517202e-01 -1.41465437e+00 4.06899154e-01
5.76485634e-01 9.61527586e-01 1.51943579e-01 9.85999942e-01
-4.95275080e-01 1.34689152e+00 -5.12249768e-01 -7.62401581e-01
-4.48837459e-01 -5.50470948e-01 -1.16601896e+00 -8.76210630e-02
6.89896762e-01 -2.68492967e-01 -7.89950132e-01 9.30606723e-01
1.19822092e-01 3.88675630e-01 4.62937713e-01 -1.75677860e+00
-1.24112666e+00 3.30537766e-01 -2.81914324e-01 3.73352438e-01
9.76669550e-01 1.65435091e-01 8.94171417e-01 -7.50345588e-01
9.85898912e-01 1.36193454e+00 4.95332122e-01 9.50787485e-01
-5.49138606e-01 -2.47364596e-01 2.49219909e-01 7.03355849e-01
-9.77862239e-01 -9.39923227e-02 8.08570623e-01 -6.90231085e-01
6.33777499e-01 2.22400382e-01 5.84580958e-01 1.48698890e+00
-4.52630743e-02 1.23631132e+00 4.14595246e-01 -4.17835474e-01
3.15653920e-01 -7.94535205e-02 1.62219435e-01 7.05374241e-01
-3.70926298e-02 -1.61026865e-01 -5.50918221e-01 -7.44529590e-02
7.17637002e-01 -3.21023613e-01 -4.50613350e-01 -5.01702964e-01
-1.34280586e+00 1.15036428e+00 3.97433460e-01 1.37535691e-01
-4.30342823e-01 5.91750562e-01 6.88982248e-01 1.31100431e-01
1.52268782e-01 6.15768909e-01 -6.12671860e-02 -3.56297553e-01
-6.60242796e-01 4.67243344e-01 7.81821191e-01 1.29841983e+00
4.93336856e-01 3.49459410e-01 -6.85623348e-01 4.55229998e-01
2.10795075e-01 6.41273499e-01 4.11754906e-01 -1.31484878e+00
6.97266042e-01 2.21848637e-01 5.74341863e-02 -1.18587732e+00
-2.26060554e-01 3.28962684e-01 -6.99654996e-01 -3.54579598e-01
6.14694916e-02 -2.57704139e-01 -9.78245318e-01 1.69493711e+00
-1.67520687e-01 5.06583095e-01 1.08408295e-01 1.19495857e+00
1.57233405e+00 9.99138653e-01 4.49082822e-01 6.35516718e-02
1.44834507e+00 -1.40705335e+00 -9.79563177e-01 -6.83237970e-01
1.85958490e-01 -4.90899682e-01 9.21209991e-01 8.84685665e-03
-1.16301489e+00 -4.98693556e-01 -5.88014662e-01 -4.46461886e-01
-5.56696832e-01 -4.82432991e-01 6.16314471e-01 3.61008614e-01
-1.37030661e+00 1.47823662e-01 -4.57397997e-01 -6.30576253e-01
5.49892902e-01 6.66659996e-02 -3.26944679e-01 -3.60416442e-01
-1.61126959e+00 7.19224095e-01 2.20783323e-01 -7.79543594e-02
-1.00496507e+00 -5.60608685e-01 -1.31647098e+00 8.37815031e-02
4.62507874e-01 -8.24132860e-01 1.42789459e+00 -1.15437365e+00
-1.02386081e+00 8.77963185e-01 -1.08793564e-01 -9.66619849e-01
2.19341919e-01 -3.55611682e-01 -3.75431746e-01 6.73754573e-01
5.93511760e-02 1.29619217e+00 8.56115162e-01 -1.04782319e+00
-2.72786140e-01 -1.56560272e-01 4.07080233e-01 3.82370889e-01
-1.36271730e-01 3.69666442e-02 -8.90215039e-01 -4.80472267e-01
-6.73525870e-01 -8.40979934e-01 -1.52063161e-01 1.89693362e-01
-1.59997970e-01 -1.48340374e-01 1.15034354e+00 -9.00852799e-01
9.52032208e-01 -1.87753284e+00 3.22254151e-01 -1.63541958e-01
2.58201689e-01 6.39489233e-01 -4.46574926e-01 6.55135393e-01
1.51238352e-01 2.12755188e-01 -1.64037213e-01 -1.08085193e-01
-1.51225388e-01 1.14683352e-01 -3.57452840e-01 5.13310693e-02
3.32343131e-01 1.67874575e+00 -1.29710102e+00 -6.14882886e-01
4.98124063e-01 5.75908303e-01 -4.00263339e-01 2.55922437e-01
-7.58562565e-01 2.02995449e-01 -6.91378474e-01 6.21805251e-01
9.38798115e-02 -7.05229104e-01 -3.32797728e-02 -2.18658686e-01
3.88767034e-01 -2.75129259e-01 -4.59835380e-01 1.80855370e+00
-4.44600374e-01 1.47898102e+00 1.28075540e-01 -1.05083740e+00
6.14026427e-01 5.97198665e-01 4.62404281e-01 -8.20906103e-01
8.07010233e-02 -2.95670360e-01 -5.48917651e-01 -1.02590978e+00
7.81611562e-01 5.35776615e-01 1.12128891e-01 3.86652380e-01
9.27718282e-02 -1.04441335e-02 2.67727703e-01 6.26458049e-01
1.02580142e+00 1.55964941e-01 1.29458144e-01 2.48984978e-01
5.70641994e-01 4.45336282e-01 -2.37476408e-01 8.30038071e-01
-6.17943406e-01 8.80327165e-01 3.71854782e-01 -4.21466440e-01
-1.23256564e+00 -8.64089549e-01 5.40398359e-01 1.10268056e+00
5.24644911e-01 -3.13068688e-01 -1.17258346e+00 -4.29296315e-01
-1.62898928e-01 6.42614365e-01 -6.49214268e-01 -3.12918544e-01
-6.56128168e-01 1.89012602e-01 4.30291444e-01 6.34842455e-01
3.97330254e-01 -1.93306100e+00 -3.36623222e-01 8.95676017e-02
-6.68515921e-01 -1.70775294e+00 -6.96959853e-01 -5.79573810e-01
-5.29727519e-01 -1.03383577e+00 -1.01716590e+00 -1.25815606e+00
3.84121418e-01 4.58137155e-01 1.60732841e+00 1.59090042e-01
-9.02878866e-02 1.27438033e+00 -5.83139598e-01 -2.88860559e-01
-4.14281994e-01 -4.28486019e-02 -3.89346480e-01 -1.70524582e-01
4.03728157e-01 5.08050658e-02 -4.71326888e-01 -5.30599952e-02
-1.09654701e+00 1.08051479e-01 3.77670169e-01 7.08266616e-01
6.06759131e-01 -8.51213217e-01 4.78003889e-01 -9.11927938e-01
8.28474462e-01 -7.70629466e-01 -3.25377792e-01 5.05642533e-01
-7.90227577e-02 -2.18462020e-01 4.92737442e-01 -4.96609896e-01
-5.84339380e-01 -1.14280194e-01 -1.65275618e-01 -9.05395329e-01
-3.32854092e-01 6.09668851e-01 -5.25713414e-02 -1.96427763e-01
7.32739389e-01 2.95157880e-01 3.32312644e-01 6.63370788e-02
7.94877052e-01 5.64674497e-01 9.18287754e-01 -2.08590299e-01
8.08114648e-01 4.10038054e-01 -2.85069436e-01 -8.68991554e-01
-8.28666270e-01 -5.28736115e-01 -2.04515800e-01 -6.61323607e-01
1.46606195e+00 -7.75600553e-01 -7.16239512e-01 1.18334182e-01
-1.51988685e+00 -2.08818182e-01 -2.25245833e-01 1.61302954e-01
-8.82801890e-01 4.59658802e-01 -5.98476887e-01 -4.97272968e-01
-6.45146966e-01 -1.15925884e+00 1.28269756e+00 6.20036483e-01
-9.52317342e-02 -1.25943589e+00 -7.51217408e-03 9.94220078e-01
4.21049446e-01 5.07551074e-01 7.72828102e-01 -7.69512355e-01
-8.67986262e-01 -1.72192946e-01 -3.68080139e-01 3.23614627e-01
-2.75324255e-01 -1.99102879e-01 -6.94132626e-01 -1.75808668e-02
-5.94480217e-01 -7.09098339e-01 4.83035713e-01 5.43707132e-01
1.38650262e+00 -3.25942069e-01 -3.14508289e-01 4.23902571e-01
1.29030156e+00 3.41921240e-01 9.39213336e-01 3.99801701e-01
9.85210955e-01 6.48250043e-01 4.66601819e-01 -9.11443606e-02
6.01350963e-01 6.77833259e-01 8.75922382e-01 -6.02273978e-02
-2.86329687e-01 -4.62819099e-01 2.66622692e-01 5.21397829e-01
1.71042662e-02 -7.91839361e-01 -1.03127766e+00 6.99090600e-01
-1.96626031e+00 -1.23394215e+00 -2.26738244e-01 1.71856463e+00
2.00844586e-01 -1.95456415e-01 -3.94275598e-02 -3.07139158e-01
1.11724591e+00 2.36640468e-01 -6.26671433e-01 -5.76091409e-01
-8.96096602e-03 -6.95287362e-02 2.56222248e-01 3.91845971e-01
-1.17775750e+00 1.24656034e+00 6.53395939e+00 5.44641554e-01
-8.00556362e-01 9.47939828e-02 6.78104222e-01 4.38834488e-01
-5.48113465e-01 -4.08412874e-01 -3.23589116e-01 3.15228820e-01
1.12867856e+00 -3.66784066e-01 6.39104784e-01 1.07563496e+00
1.11001320e-01 2.15976343e-01 -1.01381421e+00 1.25126433e+00
7.29460955e-01 -1.91664279e+00 4.82756346e-01 -3.46415311e-01
7.15591848e-01 1.23349182e-01 1.58663660e-01 4.49338824e-01
-6.68546855e-02 -1.06422949e+00 6.50846362e-01 3.24272126e-01
8.33497703e-01 -4.99798059e-01 8.01505268e-01 -1.78025365e-01
-9.24368978e-01 7.65738543e-03 -9.98680890e-02 1.87068984e-01
5.12755394e-01 -7.17942342e-02 -4.87384379e-01 3.93335074e-01
7.41062343e-01 6.30184293e-01 -4.31453854e-01 1.13177645e+00
-4.07006592e-02 5.47533810e-01 3.15617949e-01 -3.98531556e-01
5.38643181e-01 -1.71349466e-01 5.19136190e-01 1.12314522e+00
-3.06533258e-02 2.67784208e-01 1.16101444e-01 7.37071216e-01
-4.74915773e-01 2.83865840e-03 -7.93201625e-01 -5.14593422e-01
3.93325061e-01 1.03780842e+00 -5.85369110e-01 -4.40208346e-01
-8.13199937e-01 1.03586578e+00 1.17592454e-01 5.98572850e-01
-1.08639741e+00 -4.47150052e-01 6.29614294e-01 1.91366136e-01
3.56518775e-01 -1.51508510e-01 2.37048924e-01 -1.12750685e+00
-9.04287621e-02 -1.08754790e+00 4.00669992e-01 -1.43679380e+00
-1.12456179e+00 9.00068343e-01 1.64313912e-01 -1.03192568e+00
-7.15702295e-01 -6.31502926e-01 -5.79069436e-01 4.33573633e-01
-1.60869229e+00 -1.11749232e+00 -8.38950098e-01 7.29912460e-01
7.91255593e-01 -2.45379403e-01 6.40320420e-01 3.57616484e-01
-1.53014764e-01 3.53072047e-01 -9.49608628e-03 5.24708629e-01
3.36807251e-01 -9.67463315e-01 9.40881371e-01 6.44311249e-01
4.75596130e-01 -6.24441355e-02 5.96363604e-01 -4.17049617e-01
-1.54904616e+00 -1.40392995e+00 7.72437453e-01 -3.88297826e-01
8.60949993e-01 -4.45623487e-01 -9.26769793e-01 7.39110827e-01
5.44084251e-01 5.43245375e-02 3.86362076e-01 -5.66750407e-01
-2.46626750e-01 3.74387562e-01 -9.61931527e-01 7.70615518e-01
9.07263458e-01 -8.16974878e-01 -3.96313012e-01 7.95593500e-01
1.25463748e+00 -5.91231346e-01 -5.06587744e-01 -1.71734005e-01
2.23560050e-01 -6.12145483e-01 1.16962922e+00 -1.10690963e+00
7.49467969e-01 6.87581077e-02 1.15832731e-01 -1.01328444e+00
-2.30243579e-01 -1.01942313e+00 -5.39443195e-01 1.04962885e+00
3.72204810e-01 1.30629307e-02 1.10741115e+00 7.17431486e-01
-5.47090650e-01 -5.70574880e-01 -4.92741346e-01 -4.77044255e-01
-1.04222104e-01 -4.86615270e-01 4.55949187e-01 8.52271736e-01
-2.90399700e-01 5.35755575e-01 -5.66639125e-01 -5.72486408e-02
4.28816170e-01 -9.89812538e-02 4.77587044e-01 -1.00494826e+00
8.25254340e-03 -2.51300812e-01 -6.47282004e-01 -1.21263576e+00
3.76060694e-01 -7.22304404e-01 8.24964195e-02 -2.21458054e+00
-1.27220720e-01 3.56990606e-01 1.43654659e-01 2.67055750e-01
-1.81689069e-01 4.34061170e-01 4.95312065e-01 5.98311052e-02
-1.14859712e+00 4.59453613e-01 1.49521649e+00 -4.61629838e-01
7.29126930e-02 -2.61873722e-01 -6.55198753e-01 4.61814791e-01
7.47412860e-01 -2.29595304e-02 -5.48094034e-01 -8.48307371e-01
2.82089055e-01 4.10069674e-01 5.52517951e-01 -8.69592011e-01
3.08129400e-01 -2.06545487e-01 3.30591053e-02 -2.24623755e-01
5.11933684e-01 -6.33965611e-01 -3.10167134e-01 2.11084723e-01
-5.67214072e-01 4.41599011e-01 2.35299259e-01 6.50502264e-01
-5.37699282e-01 -3.75431567e-01 4.36612874e-01 -3.67173702e-01
-1.32967818e+00 7.14437544e-01 -5.95303833e-01 7.29526341e-01
1.09112477e+00 -1.83585972e-01 -6.92122698e-01 -1.25143993e+00
-7.26340950e-01 6.88955307e-01 5.94435558e-02 1.00401700e+00
1.01152050e+00 -1.44105470e+00 -6.99161887e-01 -4.04485434e-01
3.44295561e-01 -1.77294269e-01 3.77946794e-01 2.30119169e-01
-9.10828531e-01 8.82798553e-01 -2.64743924e-01 -4.10069197e-01
-1.06666946e+00 8.61805916e-01 1.43192604e-01 -3.11660990e-02
-4.81166601e-01 7.54419208e-01 9.19795185e-02 -1.24942042e-01
5.24223089e-01 1.49200752e-01 -7.34209001e-01 -2.54686803e-01
7.31668174e-01 2.32832313e-01 -2.10707456e-01 -7.59691298e-01
-2.50374466e-01 4.51690942e-01 -7.88149461e-02 4.87737171e-02
8.52115333e-01 -1.79595932e-01 1.64480641e-01 9.20606405e-02
1.29923010e+00 -9.33554947e-01 -9.19490576e-01 -5.29115163e-02
3.90994437e-02 -3.40424627e-02 -7.79823512e-02 -5.70452332e-01
-1.07861400e+00 9.71003711e-01 5.70391178e-01 4.12490666e-01
8.61060977e-01 2.44721711e-01 1.20293283e+00 5.88243961e-01
1.48793995e-01 -9.42653775e-01 4.71585214e-01 5.36817312e-01
1.04118252e+00 -1.34192789e+00 -3.67765695e-01 -2.41305530e-01
-1.00098729e+00 9.05245066e-01 9.43372071e-01 -2.78886184e-02
5.77155173e-01 -2.52685934e-01 1.23262219e-01 -2.85566300e-01
-7.11211920e-01 -3.00050050e-01 4.36790079e-01 1.24780500e+00
2.05769494e-01 -3.69347930e-01 1.55799598e-01 1.97265565e-01
-1.14983628e-02 1.21737055e-01 7.62546599e-01 6.87206089e-01
-3.17792773e-01 -8.11249137e-01 -2.53899753e-01 4.59834456e-01
-5.52281201e-01 -4.01630074e-01 -5.08244276e-01 8.06688607e-01
-3.26915354e-01 1.05920553e+00 4.79097039e-01 -2.49668866e-01
2.25550964e-01 -1.03162095e-01 1.92643002e-01 -4.23124999e-01
-4.68444288e-01 -5.43866217e-01 1.21888168e-01 -7.25765765e-01
-5.72425783e-01 -3.47405612e-01 -1.14510620e+00 -1.66130930e-01
5.86399287e-02 4.42911744e-01 8.33402276e-01 8.97503078e-01
5.58184743e-01 4.03423697e-01 1.19255245e-01 -7.17859805e-01
-3.80898081e-02 -7.47565806e-01 1.39397368e-01 6.37155890e-01
4.05957609e-01 -3.91696781e-01 -1.85126647e-01 4.34179574e-01] | [10.492877960205078, 1.0178673267364502] |
48b1d0f2-be2b-444c-8222-5e2c7604f1c8 | determinantal-point-processes-implicitly | 2011.06964 | null | https://arxiv.org/abs/2011.06964v2 | https://arxiv.org/pdf/2011.06964v2.pdf | Determinantal Point Processes Implicitly Regularize Semi-parametric Regression Problems | Semi-parametric regression models are used in several applications which require comprehensibility without sacrificing accuracy. Typical examples are spline interpolation in geophysics, or non-linear time series problems, where the system includes a linear and non-linear component. We discuss here the use of a finite Determinantal Point Process (DPP) for approximating semi-parametric models. Recently, Barthelm\'e, Tremblay, Usevich, and Amblard introduced a novel representation of some finite DPPs. These authors formulated extended L-ensembles that can conveniently represent partial-projection DPPs and suggest their use for optimal interpolation. With the help of this formalism, we derive a key identity illustrating the implicit regularization effect of determinantal sampling for semi-parametric regression and interpolation. Also, a novel projected Nystr\"om approximation is defined and used to derive a bound on the expected risk for the corresponding approximation of semi-parametric regression. This work naturally extends similar results obtained for kernel ridge regression. | ['Johan A. K. Suykens', 'Joachim Schreurs', 'Michaël Fanuel'] | 2020-11-13 | null | null | null | null | ['geophysics'] | ['miscellaneous'] | [-6.72983751e-02 1.98767245e-01 1.19497173e-01 -5.08200288e-01
-1.00797951e+00 -1.00354098e-01 6.24269247e-01 -8.08807462e-02
-3.59770447e-01 1.03480232e+00 -2.01769490e-02 -1.86334372e-01
-2.86322385e-01 -7.29441226e-01 -7.73708999e-01 -8.77562463e-01
-1.23516046e-01 4.15027440e-01 -3.35707469e-03 -1.33759663e-01
3.00493747e-01 6.40132785e-01 -1.31573892e+00 -2.74152011e-01
1.25380969e+00 7.25083232e-01 -1.40792444e-01 4.88842458e-01
4.61567491e-02 1.89402342e-01 -2.30083480e-01 -2.77828276e-01
2.76669174e-01 -3.98883343e-01 -3.16877753e-01 -1.97404593e-01
6.11809865e-02 -6.06682897e-02 5.54295108e-02 8.81480038e-01
5.10389626e-01 4.65443313e-01 1.37671602e+00 -8.03694367e-01
-5.22299170e-01 3.84955853e-01 -7.01426089e-01 -5.36589585e-02
8.14826339e-02 -4.38086808e-01 5.84285617e-01 -1.12623644e+00
2.21385032e-01 1.19732130e+00 1.19743896e+00 2.19326854e-01
-1.83257115e+00 -4.92804080e-01 -6.03294313e-01 -2.71270126e-01
-1.60717380e+00 -3.46429616e-01 8.16740572e-01 -6.96991742e-01
2.16647536e-01 4.50558931e-01 3.19298714e-01 8.22711229e-01
3.08703748e-03 5.66228628e-01 1.42810392e+00 -7.71262109e-01
1.87508658e-01 3.29946041e-01 4.94096696e-01 5.79496324e-01
1.45973012e-01 3.86734039e-01 -1.42864704e-01 -9.28303897e-01
8.98299694e-01 -1.48603261e-01 -2.06772566e-01 -5.54676294e-01
-6.31224334e-01 1.14556170e+00 -1.73370659e-01 1.52434528e-01
-3.66003990e-01 2.82448661e-02 2.59867400e-01 6.05812930e-02
1.06993580e+00 3.50828171e-02 -6.63545132e-02 -3.81212868e-02
-1.22761071e+00 3.19535702e-01 9.70857084e-01 7.82134652e-01
8.05467129e-01 3.73082399e-01 3.08751110e-02 1.02158499e+00
3.19688648e-01 6.80872381e-01 -2.03695036e-02 -9.61947024e-01
1.67446494e-01 -1.53364509e-01 3.59936297e-01 -6.17945135e-01
-3.82749707e-01 -1.11627534e-01 -1.06924558e+00 2.98482984e-01
6.60739481e-01 -8.44906121e-02 -2.46978194e-01 1.42897081e+00
2.52679020e-01 3.67881656e-01 7.01106945e-03 5.34481287e-01
7.02825859e-02 8.26113105e-01 1.65280513e-02 -5.27437031e-01
8.56838167e-01 -1.99807748e-01 -5.53082645e-01 7.13666677e-01
3.15322340e-01 -7.30503976e-01 9.97492790e-01 4.15734500e-01
-1.08821309e+00 -4.95981842e-01 -7.27214754e-01 9.18247458e-03
9.67444927e-02 4.15329963e-01 3.42290848e-01 8.25164318e-01
-9.16305244e-01 1.02342319e+00 -8.03366005e-01 -6.70374557e-02
-1.20508233e-02 2.33089045e-01 -2.18462292e-03 4.69564795e-01
-9.84653771e-01 8.52462709e-01 -1.28058985e-01 2.72277832e-01
-3.63577515e-01 -9.19093370e-01 -6.76748753e-01 -1.58327758e-01
-1.63501486e-01 -3.46526176e-01 9.41513896e-01 -6.26172364e-01
-1.63870108e+00 6.67086184e-01 -3.42260420e-01 -5.74379027e-01
8.64328027e-01 -2.33578071e-01 -4.61134344e-01 -5.97262084e-02
-9.65855026e-04 8.58642459e-02 1.15651321e+00 -9.28793669e-01
3.04206833e-02 -4.44822252e-01 -5.44361055e-01 -7.44746104e-02
2.18018413e-01 1.65886745e-01 1.39877170e-01 -8.46720994e-01
4.37864959e-02 -9.07130301e-01 -4.66109842e-01 -1.55098632e-01
-4.13171023e-01 -1.27967969e-01 4.23814446e-01 -9.80163097e-01
1.04890335e+00 -2.28425813e+00 9.95755941e-02 6.47119403e-01
-2.96993136e-01 8.80501866e-02 2.93083102e-01 5.16640961e-01
-2.11960971e-01 1.61013395e-01 -8.50347400e-01 -3.73198956e-01
-5.51309064e-02 1.06593288e-01 -7.12949216e-01 7.63123631e-01
1.81210890e-01 4.38549995e-01 -4.87615019e-01 -2.93828011e-01
1.57405779e-01 7.93732285e-01 -2.61366218e-01 -1.75391123e-01
8.22886601e-02 7.35709548e-01 -3.75939250e-01 3.25138301e-01
7.40304947e-01 2.36728057e-01 -2.62445927e-01 9.34365988e-02
-6.36646092e-01 -1.83668852e-01 -1.05927920e+00 1.08421767e+00
-5.19388378e-01 6.19745076e-01 8.19159150e-02 -1.12369490e+00
1.46368289e+00 3.59793872e-01 3.67129475e-01 8.34967010e-03
-2.78868407e-01 5.57847679e-01 -5.05429924e-01 -2.05842093e-01
4.72338021e-01 -4.25626755e-01 1.57759041e-01 1.21860676e-01
-4.46208209e-01 -1.70889840e-01 -1.59559444e-01 -4.74911243e-01
4.23155457e-01 6.08825684e-01 5.06655753e-01 -9.01510656e-01
7.82353818e-01 -1.63207367e-01 3.93494099e-01 6.17205441e-01
3.19312871e-01 7.11645782e-01 6.33983433e-01 -3.84019911e-01
-1.28654957e+00 -1.00459003e+00 -1.00371075e+00 8.09740007e-01
-3.70762795e-01 1.70753837e-01 -4.85956877e-01 -4.68187816e-02
2.08927482e-01 1.09799433e+00 -7.13518739e-01 1.10094272e-01
-7.06080317e-01 -9.54282224e-01 8.31484616e-01 2.43252128e-01
3.12190145e-01 -6.70015335e-01 -2.85741299e-01 1.96795791e-01
9.49213877e-02 -5.63018441e-01 -3.02179903e-01 2.28661686e-01
-1.36291945e+00 -7.53139615e-01 -1.39412856e+00 -4.24937814e-01
5.67102075e-01 -2.46307299e-01 7.61830091e-01 -5.71035385e-01
-1.03991283e-02 3.22331697e-01 1.11160636e-01 -2.84750462e-01
-7.98973739e-01 -2.86506265e-01 1.40148148e-01 1.30602181e-01
2.44811893e-01 -8.11323345e-01 -3.53767306e-01 5.92369258e-01
-5.44636250e-01 -3.88981491e-01 2.00739279e-01 9.16290104e-01
9.07306969e-01 -4.53747958e-01 6.78605437e-01 -8.76684368e-01
7.44996250e-01 -4.30176377e-01 -1.06288862e+00 2.57141203e-01
-6.96173370e-01 6.03424311e-01 7.49888122e-01 -4.96782720e-01
-1.21742404e+00 1.34606749e-01 -2.07678005e-01 -4.97676671e-01
1.40560508e-01 4.65227574e-01 2.77470082e-01 -4.06019568e-01
9.70411777e-01 3.50366622e-01 1.91138297e-01 -7.64628649e-01
1.26305252e-01 5.06453991e-01 3.40250582e-01 -1.13209069e+00
7.07543671e-01 5.40625930e-01 5.86315334e-01 -1.49829197e+00
-3.32676411e-01 -4.80031610e-01 -5.95286131e-01 9.76967216e-02
5.61749041e-01 -6.23474896e-01 -5.48596978e-01 2.73579717e-01
-1.19073641e+00 -2.71356314e-01 -7.64797688e-01 6.57298207e-01
-1.00000715e+00 7.51924098e-01 -4.04117823e-01 -1.44424701e+00
-1.70439437e-01 -1.00189924e+00 7.85396278e-01 -1.14104182e-01
-2.48051789e-02 -1.19228208e+00 5.72806418e-01 -1.84544161e-01
3.01650584e-01 4.49677914e-01 8.51212025e-01 -5.69066107e-01
2.58813438e-04 -2.37029016e-01 -5.72919808e-02 5.37365079e-01
-2.56442457e-01 1.68275088e-01 -1.01822305e+00 1.09813921e-01
3.08249891e-01 2.27762535e-01 8.53064477e-01 8.53424430e-01
1.08741677e+00 -2.83899069e-01 -1.08066602e-02 7.86321104e-01
1.32175827e+00 -2.90985201e-02 7.89510190e-01 -5.61483242e-02
3.50011498e-01 7.96305418e-01 4.15576398e-01 6.33070171e-01
-1.73137367e-01 7.31445909e-01 -2.36159518e-01 1.77524492e-01
5.06889164e-01 -1.48885623e-01 4.02680188e-01 7.87691772e-01
-7.57938623e-01 5.60825765e-01 -7.08174229e-01 3.45864296e-01
-1.77429593e+00 -1.09036493e+00 -8.01294565e-01 2.75998235e+00
7.83942819e-01 -3.41346592e-01 3.34199399e-01 4.35946882e-02
8.89764309e-01 -1.24467015e-02 -3.58928442e-01 -6.32045031e-01
-2.43045717e-01 4.47609931e-01 8.17386150e-01 8.19041371e-01
-9.97223079e-01 3.66566300e-01 6.72386122e+00 1.04643381e+00
-7.32784569e-01 1.37992233e-01 5.53490460e-01 5.68431556e-01
-5.30791759e-01 1.95372134e-01 -9.76529837e-01 5.09722233e-01
1.11411893e+00 -2.94924736e-01 2.11935490e-01 8.01803648e-01
4.44923371e-01 -2.54586816e-01 -7.40206182e-01 6.55750632e-01
-2.64616132e-01 -9.57492888e-01 -3.78062427e-01 3.15021664e-01
6.96003020e-01 -2.68251896e-01 1.45629510e-01 3.24022204e-01
-5.88146150e-02 -9.92938459e-01 3.09916109e-01 9.95745778e-01
9.20900047e-01 -9.37875450e-01 3.14470321e-01 6.23773456e-01
-1.06509292e+00 3.71338278e-01 -7.40757108e-01 7.25708604e-02
2.82678962e-01 8.59053791e-01 -6.12349689e-01 6.61906540e-01
-6.48103803e-02 2.23351553e-01 -3.06641310e-02 1.03535223e+00
5.29316328e-02 8.09714079e-01 -6.73123896e-01 -9.25073549e-02
9.07217786e-02 -1.14915419e+00 8.43683958e-01 1.25687850e+00
8.83157909e-01 1.59155771e-01 -3.25542331e-01 9.68802094e-01
5.77367425e-01 4.35836047e-01 -6.60424829e-01 1.55154809e-01
1.33470893e-01 7.74757445e-01 -2.93674886e-01 6.99635670e-02
-3.66331428e-01 5.19679725e-01 -7.09950626e-02 6.99204803e-01
-7.58573174e-01 -3.25776905e-01 4.08901840e-01 4.10607100e-01
1.86505079e-01 -4.46528941e-01 -4.68294829e-01 -9.95077133e-01
-1.41481549e-01 -3.30116987e-01 1.66758031e-01 -4.73389745e-01
-1.22615647e+00 3.55053157e-01 4.85699594e-01 -1.26651239e+00
-5.73035181e-01 -4.49982792e-01 -6.98439240e-01 1.47925961e+00
-1.21409452e+00 -1.00123358e+00 3.03025305e-01 4.39519256e-01
1.82977170e-01 -1.43187314e-01 9.21618044e-01 1.04259968e-01
-2.95328677e-01 3.72405320e-01 7.36985743e-01 -3.28514814e-01
5.19825935e-01 -1.31888378e+00 2.55538642e-01 6.18010163e-01
-3.77833582e-02 6.05681062e-01 1.07913136e+00 -6.73499644e-01
-9.81444716e-01 -8.40639651e-01 8.08963299e-01 -2.75379390e-01
6.73164308e-01 5.10191955e-02 -1.29875386e+00 5.41426599e-01
-2.57081836e-01 2.39243861e-02 6.56512797e-01 1.10333428e-01
-1.40015438e-01 -6.00535721e-02 -1.13135958e+00 4.64308113e-01
4.01017517e-01 -5.23272097e-01 -5.17332137e-01 2.31342316e-01
3.07777166e-01 -1.19484723e-01 -9.67776477e-01 3.97702307e-01
6.75255418e-01 -1.03169882e+00 1.09965968e+00 -4.32309538e-01
2.04868108e-01 -1.66727170e-01 -2.70917803e-01 -1.12660599e+00
6.04338432e-03 -1.12245727e+00 4.63324301e-02 9.39761460e-01
2.51026332e-01 -7.84630597e-01 4.26520973e-01 6.85495853e-01
-8.84198621e-02 -7.03550756e-01 -1.10989785e+00 -1.08869588e+00
4.39094841e-01 -3.85977775e-01 2.13981166e-01 7.86349833e-01
-5.92589006e-02 -1.40177101e-01 -7.87736595e-01 8.73351619e-02
1.16903651e+00 -8.78910199e-02 6.19003236e-01 -1.65097654e+00
-5.47477484e-01 -1.10431388e-01 -1.75288796e-01 -9.87695873e-01
3.67247403e-01 -8.56212854e-01 -3.69400755e-02 -8.42266083e-01
-3.20010364e-01 -7.34557509e-01 6.17503822e-02 -2.74288327e-01
1.21377744e-01 4.53441218e-02 -2.99867898e-01 4.22300428e-01
3.74534309e-01 6.62004709e-01 9.52010930e-01 4.19304222e-01
-5.28237343e-01 8.56387854e-01 -2.25173324e-01 9.95234907e-01
7.04303503e-01 -4.11785662e-01 -2.81307250e-01 2.26600230e-01
3.89290661e-01 6.41786039e-01 5.83711326e-01 -7.17050135e-01
6.10256493e-02 -4.79459055e-02 2.41512358e-01 -6.33978367e-01
4.90636110e-01 -5.58585346e-01 6.65322959e-01 3.01861048e-01
-4.03509259e-01 -1.75268143e-01 -5.04458603e-03 8.17474961e-01
-1.23089023e-01 -7.74521887e-01 9.53739285e-01 2.19140008e-01
-9.64472741e-02 6.84711710e-02 -2.51511216e-01 -4.14182656e-02
7.72347927e-01 -1.31321579e-01 2.31417790e-01 -5.19789100e-01
-1.02331066e+00 -2.49968797e-01 2.67777264e-01 -5.17663538e-01
5.77352345e-01 -1.16440272e+00 -9.67196882e-01 2.60553092e-01
-2.89590269e-01 -2.48001367e-01 3.14044207e-01 1.27582669e+00
-5.74465156e-01 2.24631622e-01 1.19898831e-02 -5.66709518e-01
-8.97069216e-01 1.84553385e-01 1.91368029e-01 -2.00171471e-01
-5.78606308e-01 4.63265955e-01 1.85761452e-01 -3.97045672e-01
-6.19640835e-02 -3.30132067e-01 9.00239125e-02 1.31591529e-01
3.25281709e-01 9.18958843e-01 -2.48912379e-01 -6.06924474e-01
4.57265228e-02 7.51671195e-01 4.24539864e-01 -6.17973983e-01
1.14283121e+00 -1.87151041e-02 -1.83380589e-01 1.09813523e+00
1.19792938e+00 3.63586515e-01 -1.36728668e+00 -2.51880199e-01
4.34428677e-02 -1.03434131e-01 -3.23988527e-01 -8.71242285e-02
-2.97554731e-01 1.04104948e+00 2.35437915e-01 3.13511342e-01
8.67992163e-01 -4.19963390e-01 1.81961000e-01 3.49662811e-01
4.05019552e-01 -9.12186205e-01 -7.93367207e-01 4.98332620e-01
1.20157075e+00 -9.08412039e-01 2.64260292e-01 -4.60645050e-01
-5.60819685e-01 1.45133317e+00 -2.12126628e-01 -5.30992329e-01
9.77654159e-01 3.07484809e-02 -3.86663973e-01 2.78944075e-01
-1.70296952e-01 4.21159230e-02 5.41007340e-01 5.50619543e-01
5.70223808e-01 1.91143662e-01 -8.82127762e-01 5.09225667e-01
-9.74155366e-02 5.05209528e-02 4.89984632e-01 2.85942316e-01
-4.09867406e-01 -1.05658889e+00 -7.15565562e-01 4.45172817e-01
-4.46547657e-01 -1.62774935e-01 1.78403944e-01 9.09812212e-01
-4.91019189e-01 3.69456798e-01 -1.20992742e-01 1.91854626e-01
1.94952682e-01 4.88703549e-01 2.98044473e-01 -1.13192603e-01
-3.00522298e-01 4.87159908e-01 1.67933330e-01 -9.30143818e-02
-3.93771678e-01 -1.02879131e+00 -8.06642771e-01 -3.25078845e-01
-2.69421577e-01 6.07610703e-01 7.83107638e-01 6.69539332e-01
-1.91896651e-02 -4.73773070e-02 7.26663172e-01 -7.78138220e-01
-1.26942527e+00 -9.18453693e-01 -1.18529236e+00 -1.73256189e-01
2.16144711e-01 -4.31461394e-01 -7.09417343e-01 -5.50203323e-02] | [6.928389072418213, 3.986034393310547] |
9cbf5c73-7edc-4144-86e1-1d181b6c5c9c | the-regretful-navigation-agent-for-vision-and | null | null | https://arxiv.org/abs/1903.01602 | https://arxiv.org/pdf/1903.01602.pdf | The Regretful Navigation Agent for Vision-and-Language Navigation | As deep learning continues to make progress for challenging perception tasks, there is increased interest in combining vision, language, and decision-making. Specifically, the Vision and Language Navigation (VLN) task involves navigating to a goal purely from language instructions and visual information without explicit knowledge of the goal. Recent successful approaches have made in-roads in achieving good success rates for this task but rely on beam search, which thoroughly explores a large number of trajectories and is unrealistic for applications such as robotics. In this paper, inspired by the intuition of viewing the problem as search on a navigation graph, we propose to use a progress monitor developed in prior work as a learnable heuristic for search. We then propose two modules incorporated into an end-to-end architecture: 1) A learned mechanism to perform backtracking, which decides whether to continue moving forward or roll back to a previous state (Regret Module) and 2) A mechanism to help the agent decide which direction to go next by showing directions that are visited and their associated progress estimate (Progress Marker). Combined, the proposed approach significantly outperforms current state-of-the-art methods using greedy action selection, with 5% absolute improvement on the test server in success rates, and more importantly 8% on success rates normalized by the path length. | ['Chih-Yao Ma', 'Zsolt Kira', 'Ghassan AlRegib', 'Caiming Xiong', 'Zuxuan Wu'] | 2019-03-05 | null | null | null | cvpr-2019-oral-2019-3 | ['vision-language-navigation'] | ['computer-vision'] | [ 6.37289807e-02 8.08300897e-02 -2.75861263e-01 -2.91215450e-01
-6.91149533e-01 -5.03664017e-01 7.28007853e-01 1.32473111e-01
-9.34558451e-01 7.18707025e-01 1.94707572e-01 -6.10342324e-01
-1.68466475e-02 -6.12705648e-01 -7.50782430e-01 -5.56722105e-01
-3.18176746e-01 4.53627050e-01 4.82576102e-01 -2.04455107e-01
4.50264573e-01 2.84573019e-01 -1.36566842e+00 -1.89320311e-01
8.57107162e-01 9.50376272e-01 5.31099200e-01 7.39808321e-01
-1.63335741e-01 1.04837620e+00 2.26691086e-02 1.04658902e-02
3.09285671e-01 -5.62297046e-01 -1.03853106e+00 -4.69897352e-02
3.06872070e-01 -2.67735183e-01 -3.11461091e-01 1.01814377e+00
4.01974052e-01 5.58485985e-01 2.87125617e-01 -1.12974513e+00
-3.13592941e-01 4.70560700e-01 -5.02640605e-01 1.72219664e-01
4.11815226e-01 7.71895885e-01 9.97228622e-01 -6.43985331e-01
8.89521420e-01 1.28255475e+00 3.31174225e-01 6.11297369e-01
-1.06052065e+00 -2.27430329e-01 6.91904247e-01 5.20771444e-01
-1.01591229e+00 -5.54777443e-01 4.48971987e-01 -3.68344635e-01
1.21043491e+00 -2.30272144e-01 7.49731779e-01 1.06391692e+00
2.89063454e-01 9.36291277e-01 9.81586516e-01 -3.57278496e-01
6.27678692e-01 -2.58739471e-01 -2.92313695e-02 1.17898273e+00
8.84887576e-02 2.38966137e-01 -6.49088204e-01 2.20644176e-01
5.85044622e-01 -2.71511346e-01 -3.18964183e-01 -9.29492772e-01
-1.28336215e+00 9.05130982e-01 8.39793026e-01 -7.32421875e-02
-7.73134172e-01 5.74662924e-01 2.97098398e-01 2.11440980e-01
-1.44200951e-01 3.30279827e-01 -2.70085335e-01 -4.16515172e-01
-6.77170694e-01 4.10573095e-01 6.64363503e-01 7.44768143e-01
7.10260391e-01 -9.13168713e-02 -3.98803890e-01 5.42642117e-01
5.04058361e-01 4.41508263e-01 3.17054093e-01 -1.04042292e+00
7.14406967e-01 6.56197190e-01 3.61670375e-01 -6.65006578e-01
-6.75244212e-01 -4.05569255e-01 -3.57306957e-01 6.39226496e-01
4.28155154e-01 -2.92788744e-01 -1.35294354e+00 1.94276941e+00
3.39854956e-01 -1.27682136e-02 1.89134642e-01 1.14944100e+00
3.64787132e-01 5.39725661e-01 1.32481888e-01 5.93566149e-02
9.67528224e-01 -1.55739748e+00 -3.65644008e-01 -8.18209946e-01
7.19615877e-01 -3.27965289e-01 1.11954117e+00 4.06183839e-01
-1.01729131e+00 -3.21815640e-01 -8.77809823e-01 3.44368815e-02
-2.19618082e-01 2.23918934e-03 7.03992844e-01 1.47365049e-01
-1.43536592e+00 3.90708089e-01 -1.12773466e+00 -7.58399546e-01
4.75243568e-01 2.30435744e-01 -3.34404826e-01 -2.84878284e-01
-7.59599984e-01 9.97262955e-01 5.16921282e-01 1.28486186e-01
-1.26258457e+00 -9.49066877e-03 -8.29877317e-01 -8.84968638e-02
9.02412772e-01 -9.53351200e-01 1.31935191e+00 -5.55691659e-01
-1.81900549e+00 4.87424850e-01 -1.90123215e-01 -8.77507627e-01
8.20608318e-01 -4.33874041e-01 5.82421608e-02 -9.97310206e-02
1.57276154e-01 9.82949018e-01 6.40983939e-01 -9.99219894e-01
-1.11601496e+00 -4.97408181e-01 3.44185174e-01 6.45526946e-01
2.54234910e-01 -4.78352755e-01 -1.08626878e+00 3.53058539e-02
3.97750497e-01 -1.29418588e+00 -6.94585383e-01 2.03163043e-01
-4.05255616e-01 -3.03942531e-01 4.16986853e-01 -4.69478309e-01
1.06768405e+00 -1.65043759e+00 5.00473976e-01 5.13532534e-02
-2.57214606e-02 1.30998403e-01 -3.46031189e-01 4.51469898e-01
5.92146933e-01 -1.27520651e-01 -5.48946485e-02 -4.40008610e-01
-3.99377495e-02 1.43132389e-01 -2.48128310e-01 4.04030949e-01
-1.67397663e-01 8.99532199e-01 -1.20098615e+00 -3.78552824e-02
1.99215457e-01 8.06632712e-02 -6.99874759e-01 3.55393551e-02
-4.46695179e-01 6.13761067e-01 -6.07909799e-01 5.47613800e-01
2.45059267e-01 -3.36347610e-01 1.38028771e-01 3.67435843e-01
-4.06103075e-01 5.17395139e-01 -1.15570962e+00 2.11181355e+00
-5.85005522e-01 4.63266939e-01 1.91566139e-01 -7.31141329e-01
6.96381092e-01 -3.70743036e-01 1.84996054e-03 -1.06184566e+00
-1.04798079e-01 3.56340818e-02 1.26349116e-02 -5.09354830e-01
4.36002403e-01 4.56587583e-01 1.30655468e-01 1.56219974e-01
-1.71514988e-01 1.09079354e-01 3.00055057e-01 9.76045206e-02
1.46579242e+00 5.81709862e-01 2.60302633e-01 2.66415160e-02
6.24339879e-01 4.58302140e-01 3.57127100e-01 1.18135035e+00
-4.78984982e-01 1.12079062e-01 3.39012504e-01 -5.08412063e-01
-6.78664565e-01 -9.47291017e-01 5.77879429e-01 1.24490404e+00
5.55001259e-01 -3.81824017e-01 -5.61556041e-01 -7.73913205e-01
-2.30599687e-01 9.20166135e-01 -5.47320426e-01 -7.72177503e-02
-6.75947249e-01 -2.22134843e-01 2.84910575e-02 4.96313483e-01
8.11818302e-01 -1.39052582e+00 -1.25158179e+00 2.42740154e-01
-1.96361989e-01 -9.93860543e-01 -2.80130833e-01 2.66564906e-01
-7.10050642e-01 -1.05852103e+00 -4.88823414e-01 -6.70357585e-01
6.16042256e-01 3.32790911e-01 9.24470842e-01 -7.90878907e-02
-8.23841058e-03 6.12478197e-01 -3.80832732e-01 -1.85671583e-01
-1.92195356e-01 1.63279176e-01 -6.97604716e-02 -3.11969757e-01
5.58638200e-02 -1.39376074e-01 -9.21314418e-01 1.04044855e-01
-2.79908359e-01 3.39335293e-01 7.66142011e-01 6.92612827e-01
6.21179223e-01 -2.02694759e-01 1.10657193e-01 -4.23462600e-01
6.51243985e-01 -3.61923635e-01 -9.00711179e-01 2.35443249e-01
-8.97156119e-01 4.53305840e-01 4.26040202e-01 -4.16313522e-02
-7.61655271e-01 2.54568666e-01 -9.43818390e-02 -2.73770839e-01
-1.51861668e-01 6.57986104e-01 2.71269262e-01 -1.59547448e-01
5.06274700e-01 5.14642477e-01 -9.83509794e-02 -2.71219254e-01
7.17118561e-01 -2.61221621e-02 5.23908377e-01 -2.88698941e-01
4.18383598e-01 4.55605030e-01 9.71290395e-02 -4.59918439e-01
-4.83652920e-01 -5.76030135e-01 -3.40349436e-01 -1.75620690e-01
8.84499490e-01 -8.62506270e-01 -9.96404231e-01 2.71036655e-01
-9.74544466e-01 -7.88470924e-01 4.22191881e-02 5.23958802e-01
-8.57761681e-01 2.33870938e-01 -1.66225955e-01 -9.99510765e-01
-2.21674204e-01 -1.33692336e+00 7.60698259e-01 4.54189628e-01
-3.82379964e-02 -8.26410055e-01 1.32764429e-01 1.95456460e-01
5.57302594e-01 5.26774712e-02 7.26979554e-01 -4.49207902e-01
-1.11211634e+00 6.76304623e-02 -1.65131792e-01 -2.74249732e-01
-3.01100016e-01 -5.93624234e-01 -5.01434863e-01 -5.69648623e-01
-3.34566981e-01 -3.93069893e-01 1.23224938e+00 4.38292563e-01
6.28262997e-01 -2.81574875e-01 -6.73091352e-01 6.32984936e-01
1.52569699e+00 4.49944705e-01 3.50494772e-01 8.54472160e-01
5.29058039e-01 5.51381230e-01 6.10128522e-01 3.38963270e-01
7.95874536e-01 8.46818805e-01 1.01413834e+00 2.44332507e-01
-8.90615508e-02 -5.13870955e-01 6.06384575e-01 3.52751762e-02
-1.41723435e-02 -5.23677289e-01 -1.13526511e+00 6.21841311e-01
-2.28296781e+00 -8.33107591e-01 3.81099343e-01 2.42977786e+00
3.96567196e-01 4.71000314e-01 1.61716402e-01 -4.02839065e-01
2.43230805e-01 2.24775076e-01 -1.07106483e+00 -2.94126987e-01
2.77792841e-01 -3.42552811e-01 6.94718778e-01 8.05424035e-01
-1.05672264e+00 1.49562001e+00 5.73755455e+00 4.44150060e-01
-1.34159088e+00 -1.64640844e-01 3.69100422e-01 -1.17849439e-01
1.14929331e-02 1.77575290e-01 -9.49281275e-01 1.16362110e-01
5.98735869e-01 6.23154864e-02 8.48077655e-01 9.41303432e-01
3.72455716e-01 -6.45119667e-01 -1.13966560e+00 9.42236304e-01
-1.43991396e-01 -1.24695337e+00 -7.80579373e-02 -2.88398378e-02
4.99510676e-01 5.67521334e-01 1.04616329e-01 5.46038210e-01
7.77814329e-01 -9.62817729e-01 8.41107130e-01 5.73305309e-01
3.35643888e-01 -5.86023390e-01 3.85822892e-01 7.09344149e-01
-1.06705678e+00 -3.87823790e-01 -1.24867454e-01 -2.00698301e-01
2.67192364e-01 -1.65927589e-01 -1.16647327e+00 3.73889297e-01
6.67281926e-01 4.12456691e-01 -4.70582336e-01 1.43248451e+00
-6.56085908e-01 4.45629388e-01 -2.38634229e-01 -3.50672573e-01
9.53845739e-01 -2.78333604e-01 8.14218462e-01 8.95270109e-01
2.18745410e-01 -1.33592784e-01 6.08992755e-01 6.84489191e-01
2.15988412e-01 -6.54280782e-02 -4.72880661e-01 -6.89349184e-03
3.43719661e-01 9.92133021e-01 -9.46910381e-01 -1.50872454e-01
-3.58717531e-01 1.17647076e+00 6.95777237e-01 6.72616005e-01
-6.30988955e-01 -1.77160740e-01 5.99768877e-01 -9.38424915e-02
6.82722092e-01 -5.44919431e-01 -1.66432247e-01 -7.86934793e-01
1.15547173e-01 -4.94181126e-01 2.08313301e-01 -7.47596681e-01
-5.25742352e-01 7.63635874e-01 -3.43672752e-01 -8.90864551e-01
-5.60497701e-01 -3.62859905e-01 -5.44157386e-01 9.25325513e-01
-1.73470104e+00 -1.01897073e+00 -4.49609488e-01 3.19105089e-01
9.01792765e-01 -1.36309057e-01 5.87615788e-01 -3.50864232e-01
-4.28628057e-01 3.03355545e-01 -1.67540852e-02 -1.00843348e-01
3.66887480e-01 -1.12273371e+00 6.56220555e-01 1.04767382e+00
1.97127372e-01 5.59664905e-01 7.79182434e-01 -6.44969463e-01
-1.66555405e+00 -8.17446113e-01 7.00873852e-01 -2.95664370e-01
5.44701099e-01 -1.36497200e-01 -6.03893936e-01 6.10188961e-01
8.29597041e-02 -7.27433041e-02 -8.09105486e-02 1.44085690e-01
-9.57223475e-02 1.98124185e-01 -8.88716102e-01 1.04316938e+00
1.33629143e+00 -1.33703217e-01 -3.73669237e-01 2.08296791e-01
6.49305284e-01 -6.95625007e-01 1.70645103e-01 1.23103388e-01
4.96806681e-01 -1.16659904e+00 7.74721086e-01 -5.79124808e-01
2.37149581e-01 -5.06195366e-01 -1.04386903e-01 -1.28654420e+00
-4.24989879e-01 -8.30974102e-01 -4.93347272e-02 5.48644364e-01
5.27585328e-01 -3.02645296e-01 1.02715731e+00 5.24018347e-01
-1.02111198e-01 -1.12380040e+00 -9.60670471e-01 -6.25334680e-01
-4.11648959e-01 -4.56553966e-01 8.29070136e-02 2.68924981e-01
-2.01247439e-01 5.04252493e-01 -2.98294365e-01 2.12149099e-01
4.14626181e-01 1.08880483e-01 8.34726334e-01 -8.95399451e-01
-1.16041258e-01 -6.95444167e-01 -2.25036860e-01 -1.65501380e+00
-3.65047753e-02 -7.95896292e-01 5.44031799e-01 -2.11906242e+00
-9.46704447e-02 -4.61553305e-01 -3.65837693e-01 6.09960020e-01
-6.67852256e-03 -3.21083665e-01 5.35434604e-01 2.38896489e-01
-1.21928740e+00 7.14038134e-01 1.21917188e+00 -8.21040869e-02
-7.10795760e-01 1.40108138e-01 -5.86041272e-01 8.44298422e-01
5.09071469e-01 -1.16999298e-01 -5.98582566e-01 -6.52425706e-01
3.57346684e-01 3.87352258e-01 3.18036824e-01 -1.22702718e+00
8.21262717e-01 -2.93481916e-01 3.50913964e-02 -6.37906790e-01
4.32703406e-01 -6.66889429e-01 -3.46088588e-01 7.82521129e-01
-5.91597915e-01 3.86301965e-01 2.36139178e-01 1.01751375e+00
1.05114676e-01 -5.72087876e-02 5.42010009e-01 -2.10925534e-01
-1.62730694e+00 2.69077003e-01 -4.16987121e-01 1.23278424e-02
1.04940510e+00 -1.83538869e-01 -2.61348605e-01 -5.25718749e-01
-8.16820085e-01 8.36761415e-01 4.89892572e-01 5.86332381e-01
6.72302842e-01 -1.04349363e+00 -4.54846650e-01 3.58144082e-02
1.01232767e-01 1.41212158e-02 1.07142664e-02 9.47689652e-01
-5.51505327e-01 7.93866634e-01 -1.51444986e-01 -6.03202820e-01
-9.91754830e-01 5.38724422e-01 3.38006586e-01 -3.68910819e-01
-7.05285966e-01 1.09301317e+00 1.09001145e-01 -2.67768204e-01
6.43291295e-01 -4.73632753e-01 -3.65359217e-01 -1.98958963e-01
4.62027401e-01 4.46096778e-01 -1.76130116e-01 -3.36303800e-01
-6.34435892e-01 6.10969365e-01 -1.46084994e-01 -3.82510692e-01
1.12396348e+00 -4.69187558e-01 2.61960506e-01 2.98616260e-01
6.41932189e-01 -1.83842197e-01 -1.79710841e+00 -3.08498025e-01
1.94463179e-01 -1.81701213e-01 1.88410789e-01 -1.35780668e+00
-6.59181297e-01 6.77670598e-01 7.27502823e-01 -6.51146770e-02
8.25330555e-01 8.56056530e-03 5.55663407e-01 9.30043995e-01
6.77336812e-01 -9.94651437e-01 9.79903415e-02 9.39321220e-01
8.61041427e-01 -1.43252945e+00 -1.77786633e-01 3.90990078e-02
-7.93253601e-01 9.57972169e-01 7.99615622e-01 6.50159940e-02
2.09168717e-01 -1.75902933e-01 1.42078042e-01 -1.33126155e-01
-1.10136747e+00 -6.13122702e-01 2.26269081e-01 6.61747038e-01
1.14980647e-02 -1.84009627e-01 -2.43216634e-01 1.51468813e-02
8.52051005e-03 1.47601292e-01 3.07773679e-01 1.03586769e+00
-9.42860663e-01 -7.52295136e-01 -2.06094272e-02 1.97054788e-01
1.07178949e-01 -3.77432890e-02 -1.78778037e-01 6.47798002e-01
-7.51219392e-02 9.36969817e-01 -2.68514622e-02 -3.23023707e-01
3.00703824e-01 -3.81948166e-02 3.63666505e-01 -4.53168839e-01
-1.38385892e-01 -1.23881385e-01 1.57771125e-01 -1.17240715e+00
-1.84050351e-01 -6.22691453e-01 -1.52279150e+00 4.60896567e-02
4.19699680e-03 -1.53725460e-01 8.79043221e-01 9.75675046e-01
6.53043032e-01 4.40197885e-01 9.34442803e-02 -8.96610200e-01
-7.46810913e-01 -5.27149856e-01 1.17642649e-01 -8.03508759e-02
5.25224090e-01 -7.24961877e-01 -6.93730563e-02 -3.47161382e-01] | [4.494775772094727, 0.5466209053993225] |
5794b8ed-2f89-42c0-bb89-128338c0b3f4 | aideveloper-deep-learning-image | null | null | https://www.biorxiv.org/content/10.1101/2020.03.03.975250v1 | https://www.biorxiv.org/content/10.1101/2020.03.03.975250v1.full.pdf | AIDeveloper: deep learning image classification in life science and beyond | Publications on artificial intelligence (AI)-based image analysis have increased drastically in recent years. However, all applications use individual solutions highly specialized for a particular task. Here, we present an easy-to-use, adaptable, open source software, called AIDeveloper (AID) to train neural nets (NN) for image classification without the need for programming. The software provides a variety of NN-architectures that can be simply selected for training. AID allows the user to apply trained models on new data, obtain metrics for classification performance, and export final models to different formats. The working principles of AID are first illustrated by training a convolutional neural net (CNN) on a large dataset consisting of images of different objects (CIFAR-10). We further explore the potential of AID by training a model to distinguish areas of differentiated and non-differentiated mesenchymal stem cells (MSCs) in culture. Additionally, we compare a conventional clinical whole blood cell count with a whole blood cell count performed by an NN-trained, using a dataset of more than 1.2 million images obtained by real-time deformability cytometry, delivering comparable results. Finally, we demonstrate how AID can be used for label-free classification of B- and T-cells derived from human blood, which currently requires costly and time-consuming sample preparation. Thus, AID can empower anyone to develop, train, and apply NNs for image classification. Moreover, models can be generated by non-programmers, exported, and used on different devices, which allows for an interdisciplinary use. | ['Thomas Krüger', 'Martin Kräter', 'Shada Abuhattum', 'Despina Soteriou', 'Angela Jacobi', 'Maik Herbig', 'Jochen Guck'] | 2020-03-05 | null | null | null | biorxiv-2020-3 | ['blood-cell-count'] | ['computer-vision'] | [ 2.16478676e-01 -2.83966213e-01 6.22473359e-02 -3.13537151e-01
-4.02390540e-01 -5.35223126e-01 1.89239547e-01 4.31566179e-01
-8.57513607e-01 6.94637239e-01 -3.85231495e-01 -3.79290760e-01
3.28836411e-01 -9.86218333e-01 -5.09328365e-01 -9.30916905e-01
1.23081012e-02 9.46049869e-01 8.71431977e-02 -9.51355398e-02
-9.87230390e-02 9.68463361e-01 -1.55769718e+00 5.70832253e-01
6.23789668e-01 1.15435112e+00 3.15940306e-02 8.32750797e-01
-2.37730369e-01 5.35701692e-01 -8.24765027e-01 -4.71970029e-02
1.20517790e-01 -2.27340698e-01 -4.50950950e-01 -1.98274776e-01
2.88115978e-01 -3.23510647e-01 4.54671048e-02 6.26037836e-01
7.85365999e-01 -1.73785403e-01 8.18022192e-01 -1.02825558e+00
-7.33014464e-01 2.14537501e-01 1.84007496e-01 2.66853303e-01
6.42944053e-02 4.61797208e-01 1.43168136e-01 -6.79652572e-01
5.83834350e-01 9.08948779e-01 6.90292180e-01 8.75903606e-01
-1.26853192e+00 -6.11051381e-01 -4.52221274e-01 -6.23044223e-02
-1.15555429e+00 -2.48270050e-01 3.60731810e-01 -7.29273736e-01
9.98766840e-01 3.49221855e-01 9.47380841e-01 1.10843813e+00
2.03080595e-01 7.33510792e-01 1.04160452e+00 -3.41168910e-01
3.86798501e-01 2.85609335e-01 1.63704157e-01 7.74966359e-01
2.15055808e-01 -2.00024266e-02 -1.74500763e-01 -2.15579327e-02
9.69994783e-01 1.01335861e-01 -2.38716736e-01 -4.24606949e-02
-1.15370405e+00 7.16862619e-01 2.73991972e-01 6.54434085e-01
-1.99030042e-01 5.31584024e-03 5.23331583e-01 1.40937045e-01
2.66982049e-01 4.44241256e-01 -5.66794515e-01 -2.02598684e-02
-6.70800447e-01 5.38573638e-02 8.63266826e-01 5.66299319e-01
5.51607490e-01 2.65200496e-01 -6.84285089e-02 9.45901573e-01
-1.56087205e-02 4.15517539e-01 9.57608640e-01 -1.09511709e+00
-4.98169929e-01 7.31896460e-01 -3.61688793e-01 -7.26848722e-01
-7.41794705e-01 -4.07799453e-01 -1.08613014e+00 5.24694502e-01
8.34284782e-01 -7.08408877e-02 -1.01090050e+00 1.37911928e+00
2.99658269e-01 -3.60018137e-04 1.02767885e-01 9.00216460e-01
1.20046186e+00 5.00153780e-01 1.59983680e-01 -7.57001061e-03
1.29407907e+00 -7.90899575e-01 -3.55900794e-01 1.39034569e-01
9.12576377e-01 -2.51175702e-01 1.08183932e+00 7.07247317e-01
-1.09551597e+00 -4.92523193e-01 -9.60915446e-01 -3.05736344e-02
-1.07128072e+00 -6.18080944e-02 7.56489933e-01 6.63711131e-01
-1.19044018e+00 7.29171872e-01 -7.71349192e-01 -5.98821700e-01
9.54028368e-01 7.72998631e-01 -3.87853980e-01 7.61949793e-02
-8.67480278e-01 8.38381469e-01 2.68769145e-01 -1.47432789e-01
-7.08377481e-01 -8.59820366e-01 -6.81636393e-01 -2.38336809e-02
-3.84844989e-01 -9.41730320e-01 9.21643317e-01 -1.26573110e+00
-1.69509780e+00 1.26445651e+00 2.16381431e-01 -4.89817917e-01
3.82995129e-01 3.16350818e-01 -3.48696679e-01 3.38477522e-01
-1.95163906e-01 9.37713325e-01 1.80844977e-01 -1.10615861e+00
-5.10769486e-01 -2.87219733e-01 -8.05195943e-02 -2.55974531e-01
-3.87567550e-01 5.23849577e-02 -3.85241002e-01 -7.00664163e-01
-4.25199956e-01 -6.81112468e-01 -1.31637007e-01 5.16777992e-01
-3.64808105e-02 -8.85542184e-02 6.63273752e-01 -3.80903214e-01
6.55525088e-01 -2.04400349e+00 -1.11693874e-01 1.11553617e-01
1.99612975e-01 6.86014891e-01 -3.53688091e-01 -8.82597826e-03
2.52178777e-02 3.13652933e-01 -2.37920552e-01 -1.69901643e-02
-2.75457323e-01 2.26489991e-01 3.50292087e-01 2.54185528e-01
1.34577975e-01 1.06401086e+00 -7.25935459e-01 -5.12787580e-01
3.43097419e-01 8.09971333e-01 -3.18544954e-01 1.40968591e-01
-1.36147588e-01 6.57828331e-01 -9.59870219e-03 7.96694756e-01
4.69583333e-01 -5.13854086e-01 7.65373334e-02 -1.66425437e-01
1.33176565e-01 -4.19495314e-01 -6.53152704e-01 1.46607983e+00
-5.07022440e-01 7.27067411e-01 8.37413892e-02 -1.05012250e+00
7.97956109e-01 1.07293673e-01 5.52885532e-01 -7.75393963e-01
6.46064639e-01 3.51558208e-01 -1.16326408e-02 -5.25719702e-01
-1.92909241e-01 -3.37757677e-01 2.54117101e-01 5.62314630e-01
3.36284757e-01 8.46109390e-02 5.39023399e-01 -2.09006876e-01
9.04432416e-01 -2.42980003e-01 -3.54809803e-03 -1.92220822e-01
6.44834936e-01 3.69329117e-02 2.48276979e-01 6.46515965e-01
-3.47799569e-01 7.59265661e-01 3.04180682e-01 -9.92707193e-01
-8.24676096e-01 -1.08593500e+00 -4.41145688e-01 1.21900356e+00
-2.94099122e-01 3.67612600e-01 -8.23724210e-01 -4.54082698e-01
1.47366717e-01 3.41318727e-01 -7.21328318e-01 1.76674739e-01
-6.42398059e-01 -7.86894858e-01 7.35716224e-01 5.62780201e-01
3.76190305e-01 -1.37510848e+00 -6.57307982e-01 2.85736620e-01
8.42479169e-02 -9.43015218e-01 4.29190919e-02 2.95965493e-01
-8.67305756e-01 -9.91973400e-01 -1.00690556e+00 -1.19580364e+00
8.81138027e-01 -2.20580980e-01 1.04035521e+00 5.68873584e-01
-5.29319465e-01 3.84863138e-01 -2.04797555e-02 -6.06209576e-01
-4.64491874e-01 3.99290621e-02 9.99309197e-02 -1.01828374e-01
4.72523361e-01 -4.83832121e-01 -6.54384673e-01 2.55320966e-01
-1.15904558e+00 6.32737204e-02 4.13407922e-01 9.01126742e-01
7.36618400e-01 -2.94532984e-01 9.56954896e-01 -9.85381186e-01
4.99794751e-01 -2.04551995e-01 -5.31518102e-01 2.95135289e-01
-5.27162373e-01 -2.47797132e-01 9.84540343e-01 -6.01573229e-01
-6.50321960e-01 8.66833329e-02 -4.59375560e-01 -2.92549998e-01
-6.81215823e-01 5.46185672e-01 2.23741964e-01 -3.86045486e-01
8.12372863e-01 2.92448699e-01 4.99836773e-01 -1.38589785e-01
-4.24560579e-03 8.77923667e-01 5.90789199e-01 -3.59537244e-01
1.86074480e-01 6.61975980e-01 3.08703091e-02 -7.43450999e-01
-4.54066455e-01 -5.34662381e-02 -4.58384544e-01 -3.39698523e-01
9.00661469e-01 -5.85461020e-01 -1.00846577e+00 8.95696104e-01
-8.67346644e-01 -9.27828729e-01 -3.58594805e-01 4.09597188e-01
-3.93634617e-01 -9.07296594e-03 -1.08602130e+00 -3.93100470e-01
-6.51991427e-01 -1.05040264e+00 7.09831476e-01 4.80510443e-01
-3.08216751e-01 -1.38528919e+00 1.12712495e-01 2.98781157e-01
8.17119181e-01 4.97222215e-01 1.07990646e+00 -9.49120283e-01
-3.52597356e-01 -4.52115476e-01 -1.54099524e-01 3.15699965e-01
8.34343135e-02 2.63981611e-01 -1.14754474e+00 -2.20067278e-01
-1.47861168e-01 -5.69953382e-01 7.15717614e-01 6.22973263e-01
1.54327512e+00 -2.05706373e-01 -5.40029883e-01 8.70875180e-01
1.35224986e+00 4.27290320e-01 7.54662633e-01 3.55333567e-01
4.93748099e-01 5.46499133e-01 2.37327926e-02 8.00021514e-02
9.65343267e-02 2.55554348e-01 1.38852000e-01 -5.86540520e-01
-1.00534059e-01 3.85973066e-01 -1.36972100e-01 7.21121907e-01
-1.05698645e-01 -3.59621197e-01 -9.27308023e-01 3.44299108e-01
-1.33086944e+00 -6.32662356e-01 -2.25736573e-02 2.08869672e+00
7.85698414e-01 -6.23341352e-02 2.65315175e-01 -5.41061722e-03
6.16645575e-01 -6.58950984e-01 -6.96065843e-01 -4.86411482e-01
-3.19148242e-01 6.50405765e-01 1.35913074e-01 3.01442802e-01
-8.55469406e-01 6.10156000e-01 7.16339493e+00 8.82128835e-01
-1.77198744e+00 -1.27940968e-01 1.20282423e+00 -3.55780691e-01
-9.34698433e-02 -5.05838633e-01 -3.13728809e-01 4.78336990e-01
1.04180789e+00 -6.03035279e-02 3.78479213e-01 5.69995999e-01
5.56561947e-02 -1.43105403e-01 -1.28184462e+00 1.01936686e+00
4.27808277e-02 -1.75332201e+00 1.05567582e-01 -4.11830433e-02
3.97075087e-01 1.66295871e-01 -8.10063034e-02 4.50390995e-01
1.30756527e-01 -1.24456608e+00 2.23281562e-01 6.44470990e-01
1.08962023e+00 -6.07471466e-01 1.10335279e+00 2.55446225e-01
-6.44817650e-01 2.68226117e-02 -3.21405172e-01 -2.99821347e-02
-9.78360884e-03 6.12834215e-01 -6.60920084e-01 1.76344171e-01
7.54758656e-01 4.64863360e-01 -6.02887034e-01 9.84566271e-01
3.84669632e-01 2.92400897e-01 -4.55777973e-01 -4.07176346e-01
-2.47946128e-01 -1.37594089e-01 -9.92873162e-02 1.45523572e+00
2.73559690e-01 8.83569419e-02 -1.67457759e-01 7.99962342e-01
-1.70482025e-02 1.67830735e-01 -4.60765123e-01 -7.04022273e-02
6.29303381e-02 1.63595343e+00 -1.15890217e+00 -5.49517989e-01
-2.99097538e-01 5.85111797e-01 4.12125647e-01 4.24201459e-01
-5.55271506e-01 -4.73550916e-01 2.87568450e-01 1.03895999e-01
1.67562947e-01 1.78207159e-01 -4.04972702e-01 -8.56481373e-01
-4.70021397e-01 -8.31198215e-01 4.67253119e-01 -9.12870824e-01
-1.57609129e+00 6.39545202e-01 -3.48491490e-01 -1.14269924e+00
-7.36579746e-02 -1.19435287e+00 -5.38332760e-01 6.95507407e-01
-1.24403298e+00 -8.24754000e-01 -5.30076087e-01 3.33516687e-01
1.89951003e-01 -3.23851317e-01 1.14698040e+00 5.40927351e-01
-7.51981020e-01 5.30731976e-01 -5.06740212e-02 4.76334214e-01
5.05099952e-01 -1.01182485e+00 -1.00925028e-01 1.39811501e-01
-2.08901256e-01 4.82602209e-01 2.99498588e-01 -2.96327978e-01
-1.01581860e+00 -1.12939334e+00 6.86906099e-01 -3.58098954e-01
3.58215868e-01 -2.61373311e-01 -9.16451812e-01 5.02360404e-01
2.24088281e-01 2.78105557e-01 1.34691954e+00 -3.44979376e-01
-2.84489039e-02 -3.21330756e-01 -1.55070972e+00 5.84413528e-01
6.49323463e-01 -1.98146030e-01 -1.63393058e-02 4.17681724e-01
3.20080996e-01 -5.52001059e-01 -1.15122902e+00 3.15635204e-01
9.32168901e-01 -1.11493969e+00 8.03101718e-01 -6.96675777e-01
3.71144414e-01 -1.50971830e-01 1.46870971e-01 -1.08289778e+00
-3.32560331e-01 -1.09422900e-01 -4.03979272e-02 9.19613361e-01
4.81957495e-01 -1.00309193e+00 9.08514559e-01 6.73956037e-01
-2.02732489e-01 -1.20590723e+00 -8.45476508e-01 -7.13488340e-01
5.43588758e-01 -2.44183525e-01 7.30730236e-01 1.03356767e+00
1.68515921e-01 1.53525740e-01 3.30820203e-01 -3.27370614e-01
2.08115056e-01 -1.01119541e-01 7.27304399e-01 -1.51769221e+00
-2.30108261e-01 -8.22170973e-01 -5.76383650e-01 -7.20128000e-01
3.60503606e-02 -1.14836299e+00 -1.47364393e-01 -1.62406266e+00
8.61760974e-02 -5.71095049e-01 -3.38308573e-01 6.78614676e-01
1.64204940e-01 5.77425241e-01 1.88242104e-02 2.88727552e-01
-3.38971525e-01 1.28531856e-02 1.30051136e+00 -2.51374692e-01
-7.02938884e-02 -2.86625504e-01 -5.90535223e-01 6.66061461e-01
9.53237057e-01 -5.18123806e-01 2.42063329e-02 -3.45646322e-01
6.00629561e-02 -1.48756698e-01 3.95562708e-01 -1.33725357e+00
1.99403301e-01 2.14068196e-03 9.02500629e-01 -3.76174629e-01
8.15878808e-02 -8.02044690e-01 3.01723093e-01 7.20075488e-01
-3.61994207e-01 -2.45440960e-01 4.38666016e-01 -4.02945699e-03
-2.67675659e-03 -3.51124227e-01 9.99226809e-01 -3.79345298e-01
-1.47343248e-01 3.20637643e-01 -8.40680778e-01 -1.72382101e-01
1.15827727e+00 -5.23568332e-01 -6.00831032e-01 -1.50340095e-01
-9.40341175e-01 7.22173899e-02 6.57722414e-01 -8.80366936e-02
5.55471778e-01 -1.39119565e+00 -5.51457465e-01 4.84607697e-01
1.00555167e-01 8.39396492e-02 4.38050985e-01 8.47249866e-01
-1.05710530e+00 2.51995683e-01 -4.89855796e-01 -9.46676314e-01
-9.12261069e-01 6.36757255e-01 7.20433652e-01 -2.78522342e-01
-2.99245656e-01 7.18791664e-01 1.18644200e-01 -7.75783956e-01
-2.80455872e-02 -3.72723013e-01 -3.94913524e-01 -7.38727227e-02
5.77568948e-01 2.06993178e-01 2.94213414e-01 -3.82467449e-01
-2.61597544e-01 4.76969957e-01 5.38584813e-02 1.04898602e-01
1.44085455e+00 2.84660310e-01 -1.50390357e-01 6.12401485e-01
1.32569265e+00 -4.16367590e-01 -1.03425014e+00 1.47685975e-01
-5.91588497e-01 -5.44149661e-04 -1.80449009e-01 -9.83699739e-01
-1.63466632e+00 8.34129632e-01 9.80820537e-01 2.94612408e-01
1.18967211e+00 -1.42565638e-01 8.41686308e-01 4.78410751e-01
3.34593803e-01 -1.03594089e+00 1.90744400e-01 3.55556756e-01
6.28367543e-01 -1.05982649e+00 -1.44209325e-01 -7.33265355e-02
-2.98601896e-01 1.48191476e+00 6.17363095e-01 -8.39501247e-02
8.23157191e-01 5.06993353e-01 6.54345095e-01 -2.66106993e-01
-6.43967509e-01 4.91094263e-03 -5.11943735e-03 1.15697479e+00
5.52247643e-01 -1.78010136e-01 -2.59722359e-02 5.75133502e-01
-1.96797386e-01 6.68166995e-01 5.81267536e-01 9.61271703e-01
-2.74018437e-01 -9.75859344e-01 -1.63876876e-01 8.66025686e-01
-6.04309618e-01 6.77728280e-02 -3.06489170e-01 7.36765683e-01
3.35808307e-01 6.66199863e-01 4.73137021e-01 -1.74308106e-01
1.43067062e-01 2.54494138e-02 5.29414773e-01 -5.55031180e-01
-8.56883287e-01 -6.29784241e-02 -1.13326833e-01 -2.30847448e-01
-4.55044001e-01 -3.44340295e-01 -1.47449124e+00 -3.28752667e-01
-1.08900614e-01 -2.51155227e-01 5.67107379e-01 8.64539385e-01
5.12688577e-01 4.40975696e-01 2.42990091e-01 -9.48516190e-01
2.79756874e-01 -9.03310955e-01 -6.31291389e-01 3.74655932e-01
1.92547336e-01 -3.95107716e-01 -5.76343656e-01 3.41020703e-01] | [14.806970596313477, -3.097543478012085] |
f374584a-c6fa-49ab-a86f-f8b9c725815e | nurse-care-activity-recognition-challenge | null | null | https://doi.org/10.1145/3341162.3345577 | http://delivery.acm.org/10.1145/3350000/3345577/p746-lago.pdf | Nurse care activity recognition challenge: summary and results | Although activity recognition has been studied for a long time now, research and applications have focused on physical activity recognition. Even if many application domains require the recognition of more complex activities, research on such activities has attracted less attention. One reason for this gap is the lack of datasets to evaluate and compare different methods. To promote research in such scenarios, we organized the Open Lab Nursing Activity Recognition Challenge focusing on the recognition of complex activities related to the nursing domain. Nursing domain is one of the domains that can benefit enormously from activity recognition but has not been researched due to lack of datasets. The competition used the CARE-COM Nurse Care Activity Dataset, featuring 7 activities performed by 8 subjects in a controlled environment with accelerometer sensors, motion capture and indoor location sensor. In this paper, we summarize the results of the competition. | ['http://delivery.acm.org/10.1145/3350000/3345577/p746-lago.pdf'] | 2019-09-09 | null | null | null | ubicompiswc-19-proceedings-of-the-2019-acm | ['multimodal-activity-recognition'] | ['computer-vision'] | [ 5.38942575e-01 2.05167532e-02 -6.22466683e-01 -4.80317444e-01
-3.62250417e-01 -6.65384233e-02 5.10690093e-01 2.85942644e-01
-5.67451298e-01 8.97667468e-01 9.10857022e-01 -1.50540918e-01
-4.23061907e-01 -5.13460577e-01 -1.76581681e-01 -7.42215335e-01
-2.09496468e-01 1.54378504e-01 -8.80387425e-02 1.75870925e-01
1.81688219e-01 3.05039197e-01 -1.57747102e+00 3.79995942e-01
4.77846444e-01 7.56581426e-01 -1.26069799e-01 5.78520834e-01
1.21762536e-01 1.01707149e+00 -7.22784996e-01 3.86415690e-01
4.92071435e-02 -1.11645114e+00 -8.43680441e-01 2.30882794e-01
4.55344059e-02 -3.17764170e-02 9.20821130e-02 3.94160867e-01
7.21539021e-01 2.96003133e-01 3.62625420e-01 -1.11321723e+00
-1.16477385e-02 2.85171151e-01 -9.28517431e-02 6.01311743e-01
1.06010509e+00 -1.68922082e-01 5.41167736e-01 -2.98729926e-01
4.28733051e-01 6.23001516e-01 4.82964545e-01 5.07309556e-01
-9.52137530e-01 -5.32810271e-01 -1.58720016e-01 4.35322374e-01
-1.09813857e+00 -3.23433906e-01 6.29170716e-01 -3.52598071e-01
1.19020450e+00 3.19146693e-01 1.25396359e+00 1.81269157e+00
2.24676669e-01 8.12730551e-01 1.20444846e+00 -4.64773387e-01
6.02270067e-01 -1.68799847e-01 3.73237103e-01 4.32179034e-01
5.83588481e-01 -2.31686130e-01 -5.78092754e-01 -1.57370254e-01
4.95998323e-01 6.90231025e-01 -4.50448185e-01 -4.03500080e-01
-1.58130562e+00 4.53438133e-01 -4.32244428e-02 8.72955143e-01
-5.66305816e-01 -1.93175599e-01 2.68737614e-01 1.22614972e-01
2.04222500e-01 4.37564373e-01 -4.27738070e-01 -1.33319855e+00
-8.67353320e-01 9.15059596e-02 1.36944604e+00 4.72063929e-01
3.28825623e-01 -3.09439123e-01 -2.35763595e-01 7.76244760e-01
4.09354925e-01 1.77885354e-01 8.60539675e-01 -6.96988046e-01
3.43106449e-01 1.32734036e+00 6.86880276e-02 -7.06200719e-01
-7.26815879e-01 -2.22345069e-01 -7.72793591e-01 1.03294306e-01
6.41141713e-01 -2.25220367e-01 -5.54391086e-01 1.26868141e+00
2.59554744e-01 3.76843274e-01 -5.99839315e-02 6.47206962e-01
5.08871913e-01 1.32542923e-01 1.45587951e-01 -4.21752661e-01
1.48347247e+00 -1.12073743e+00 -1.07765174e+00 -2.04567701e-01
1.00380826e+00 -3.58941704e-01 8.54868948e-01 5.15447497e-01
-8.78201663e-01 -3.39046150e-01 -1.19758463e+00 2.88593322e-01
-5.65519929e-01 -2.84767479e-01 6.46427333e-01 1.06145382e+00
-6.21907115e-01 5.25822639e-01 -1.49486542e+00 -8.46423149e-01
6.55513585e-01 2.48873919e-01 -6.30183101e-01 -1.47550240e-01
-7.64137924e-01 1.03567004e+00 2.36794934e-01 -2.19085410e-01
-5.54324210e-01 -5.38936794e-01 -9.21507835e-01 -1.31536052e-01
4.11364287e-01 -5.04383087e-01 1.14833629e+00 -8.67521524e-01
-1.33683372e+00 6.72520757e-01 4.06871438e-02 -5.03526628e-01
6.22694194e-01 -6.29899263e-01 -9.07937348e-01 -1.87629879e-01
6.68831728e-03 -1.61160186e-01 1.90524265e-01 -2.88770258e-01
-5.78579128e-01 -6.61264420e-01 -1.60312802e-01 2.74046183e-01
-3.79973382e-01 2.75318790e-02 -1.58438593e-01 -5.02935171e-01
-2.14388922e-01 -1.03014255e+00 -1.66268751e-01 -5.37727535e-01
9.72503424e-02 -8.29151496e-02 8.45911562e-01 -3.22500437e-01
1.55498672e+00 -2.12660217e+00 -2.42258757e-01 -6.38298988e-02
-2.98423357e-02 5.95491454e-02 4.55237925e-01 6.41832232e-01
-1.42590106e-01 -9.29341316e-02 -3.53231996e-01 -1.91020399e-01
-1.74938947e-01 7.77946174e-01 4.19852346e-01 6.48591995e-01
-2.21650466e-01 8.35892200e-01 -9.84469116e-01 -5.22290945e-01
4.85892057e-01 5.44665515e-01 -5.93969166e-01 2.03293517e-01
2.92836189e-01 5.84207594e-01 -6.03726685e-01 8.32353055e-01
1.09796211e-01 -4.00060624e-01 2.68674016e-01 1.77427769e-01
-5.42698875e-02 3.82105708e-01 -1.37554097e+00 2.03282094e+00
-1.54037133e-01 5.10952890e-01 -3.14658642e-01 -1.19947338e+00
4.96593326e-01 4.62476701e-01 1.46206391e+00 -8.09286058e-01
5.40590286e-02 -3.38269882e-02 1.29348516e-01 -8.38328779e-01
1.61271635e-02 3.49131897e-02 -1.68215290e-01 7.19602406e-01
-2.30692178e-01 5.89351833e-01 4.31749970e-01 -2.18788385e-01
2.03328443e+00 5.13840556e-01 1.01100183e+00 -1.37465417e-01
4.52852786e-01 1.12700425e-01 5.59328318e-01 5.25063872e-01
-6.47741318e-01 5.13206840e-01 1.05548336e-03 -4.79965448e-01
-2.64957875e-01 -9.79193211e-01 -7.91274309e-02 1.14556324e+00
-1.38940111e-01 -6.79203212e-01 -7.82763660e-01 -8.34273994e-01
-3.61206412e-01 2.24200770e-01 -9.17498767e-01 4.82365005e-02
-9.31932509e-01 -9.66893494e-01 7.10917413e-01 8.37860942e-01
7.50538647e-01 -1.39921010e+00 -1.42227149e+00 4.80037987e-01
-2.65977025e-01 -1.01216233e+00 -1.41663805e-01 3.99692982e-01
-1.34776235e+00 -1.68543851e+00 -5.81615508e-01 -6.43444300e-01
4.59982067e-01 7.31590986e-02 1.22696793e+00 -3.78221720e-02
-4.94451344e-01 8.61785352e-01 -5.54055214e-01 -6.37380123e-01
-1.51349187e-01 3.29808980e-01 3.94255295e-02 -2.55775861e-02
8.91792536e-01 -6.80991411e-01 -9.88290787e-01 6.08493745e-01
-9.06829834e-01 -1.27425224e-01 8.41782331e-01 5.22032499e-01
4.39208388e-01 -4.11161125e-01 3.92089516e-01 -6.79708779e-01
5.85015953e-01 -6.63936913e-01 1.29065802e-03 3.19689326e-02
-7.69648075e-01 5.92797436e-02 -7.34960064e-02 -4.52367604e-01
-8.15674841e-01 1.91408619e-01 -1.29391998e-01 4.39383030e-01
-7.49051273e-01 5.16878605e-01 -3.46124262e-01 3.92245114e-01
7.78711557e-01 3.39838117e-02 -3.16942446e-02 -6.20906413e-01
-5.54851949e-01 5.56828797e-01 2.11613238e-01 -4.13801402e-01
1.91576686e-02 5.48035443e-01 3.57311927e-02 -1.00797164e+00
-7.22453475e-01 -8.44635963e-01 -5.16996801e-01 -2.93104470e-01
1.11644673e+00 -7.33332157e-01 -6.89951181e-01 7.20524415e-02
-4.63211536e-01 -2.97521085e-01 -4.84880358e-01 1.02606750e+00
-5.55367112e-01 3.00274044e-01 -1.58672810e-01 -7.90843904e-01
-1.97819665e-01 -9.58490849e-01 7.85266280e-01 2.00368479e-01
-9.95481253e-01 -9.88267958e-01 7.04669058e-01 8.90571713e-01
4.96038646e-01 9.09229457e-01 2.97151536e-01 -7.36100614e-01
-2.74629444e-01 -4.37899411e-01 4.78836566e-01 2.35096276e-01
7.74001360e-01 -6.03570580e-01 -7.95327544e-01 -1.28256932e-01
3.12014461e-01 -4.76936214e-02 1.76520407e-01 2.72574276e-01
9.62904692e-01 -1.35601625e-01 -6.51692688e-01 2.02518210e-01
1.10306144e+00 4.70280826e-01 1.01590669e+00 6.09735787e-01
4.47447151e-01 2.64572620e-01 5.37662148e-01 5.48343003e-01
3.72661114e-01 6.88420713e-01 4.55157757e-01 4.63713221e-02
1.52394071e-01 1.75390586e-01 5.51125228e-01 7.39748776e-01
-7.35933363e-01 1.08177811e-01 -1.09728789e+00 5.46456993e-01
-1.95019603e+00 -1.19479823e+00 -9.43040922e-02 1.91230428e+00
5.23080945e-01 1.47316167e-02 4.13517058e-01 7.84007847e-01
-7.80404732e-02 1.93324864e-01 -2.92892814e-01 -2.24738285e-01
2.06453085e-01 3.61005753e-01 1.69742137e-01 -1.65464744e-01
-9.22046065e-01 -1.58715251e-04 6.57656527e+00 1.06536545e-01
-6.22159183e-01 2.34303907e-01 2.22929850e-01 -4.49822068e-01
6.19173527e-01 -2.60778725e-01 -5.69218934e-01 7.31058419e-01
1.23647678e+00 2.58622199e-01 -2.45511942e-02 1.16041481e+00
6.15287602e-01 -6.99652970e-01 -1.49323308e+00 1.20376909e+00
2.56951153e-01 -9.91885304e-01 -4.60188240e-01 5.49320221e-01
6.25591278e-01 6.44471496e-02 -5.78911483e-01 5.26847601e-01
-4.77323793e-02 -1.12853611e+00 -4.09710482e-02 7.08856404e-01
-3.17559130e-02 -3.82276416e-01 8.13231409e-01 5.91862321e-01
-1.08476293e+00 -1.99813977e-01 3.77380103e-01 -6.28389359e-01
1.50343403e-01 4.71255720e-01 -7.29671896e-01 4.15440738e-01
9.95491564e-01 9.97016609e-01 -6.11942530e-01 1.18986225e+00
6.18806891e-02 9.17846203e-01 -2.59844899e-01 -1.84433118e-01
1.87556282e-01 -4.31269556e-01 7.66094774e-02 1.42199373e+00
4.81170654e-01 1.30055577e-01 3.48054469e-01 3.14263254e-02
1.30659536e-01 1.09789550e-01 -8.42154026e-01 -2.24656835e-01
-1.50029108e-01 1.02648365e+00 -6.95349395e-01 -2.20459029e-01
-7.01626480e-01 8.03338706e-01 -6.84195310e-02 -5.67250438e-02
-8.60103369e-01 -8.87440331e-03 1.01604486e+00 6.00844204e-01
-3.27255398e-01 -3.12197924e-01 -2.66371518e-01 -1.03567457e+00
1.31228551e-01 -9.98961687e-01 7.39232659e-01 -4.50415313e-01
-6.74564779e-01 -4.66696266e-03 1.41901344e-01 -1.27838874e+00
-4.06272948e-01 -4.76361394e-01 -4.11300480e-01 3.08472514e-01
-1.00006700e+00 -6.19561195e-01 -8.83186877e-01 8.25642526e-01
7.92437255e-01 1.13693863e-01 1.19815826e+00 6.00067377e-01
-6.20188951e-01 4.83928248e-02 -2.56827533e-01 2.78189689e-01
6.85580611e-01 -1.12498701e+00 -2.60674655e-01 5.07191777e-01
1.70857742e-01 6.06619358e-01 4.65372622e-01 -5.38928807e-01
-1.48979032e+00 -6.86368167e-01 9.66416597e-01 -1.07545412e+00
2.53639042e-01 -2.53205687e-01 -6.32512152e-01 6.22195959e-01
3.17364067e-01 4.04391773e-02 1.48216724e+00 -6.74281120e-02
2.88552910e-01 -1.00568943e-01 -1.11153734e+00 6.06418788e-01
1.43083227e+00 -9.31322053e-02 -9.47987318e-01 6.44612033e-03
-1.10884808e-01 -3.91949087e-01 -1.19810295e+00 5.26501000e-01
8.13146949e-01 -1.03297067e+00 8.68596017e-01 -5.35762370e-01
2.54585356e-01 -2.11232603e-01 -1.44297376e-01 -1.04658699e+00
6.88868240e-02 -3.65851492e-01 -6.01802647e-01 7.32677877e-01
6.70844540e-02 -4.83760178e-01 1.25687075e+00 5.80122054e-01
-8.06725845e-02 -6.99511945e-01 -8.70921075e-01 -7.61943638e-01
-6.35338187e-01 -7.12594032e-01 6.21588349e-01 1.09956372e+00
3.78765583e-01 3.57224256e-01 -3.77324134e-01 -4.70695674e-01
1.24343120e-01 -1.55323341e-01 9.46541965e-01 -1.50022829e+00
-3.59124303e-01 -4.25609320e-01 -7.52742231e-01 -7.24329472e-01
-4.22761410e-01 -4.99333143e-01 -2.64909863e-01 -2.14772320e+00
1.04898535e-01 2.72318006e-01 -5.88895142e-01 7.13775516e-01
2.07257271e-01 1.19428240e-01 -3.14969569e-01 -1.23963676e-01
-9.15182233e-01 5.15777692e-02 1.03357077e+00 -4.16089036e-02
-3.88024151e-01 2.49581426e-01 -7.09968150e-01 7.46826708e-01
9.21912193e-01 -5.50848365e-01 -5.84214032e-01 8.88904557e-02
3.02144974e-01 -9.60681681e-03 1.01517454e-01 -1.68876636e+00
6.55218586e-02 -4.46298301e-01 5.86346865e-01 -5.95940769e-01
4.80187058e-01 -1.30159760e+00 3.88681173e-01 8.14129651e-01
-4.50722761e-02 2.47542962e-01 -1.25227153e-01 5.47199368e-01
-2.36509651e-01 1.91278070e-01 3.49030197e-01 -2.60136217e-01
-6.59384668e-01 1.52137712e-01 -7.84354746e-01 1.68603018e-01
1.26701844e+00 -9.83217180e-01 1.69476848e-02 -5.68584725e-02
-8.96727324e-01 -6.17249012e-02 2.87327766e-01 8.10475647e-01
3.99111092e-01 -1.26308465e+00 -3.69214237e-01 2.66259044e-01
5.41873991e-01 -1.80673316e-01 2.80097723e-02 1.28723502e+00
-4.27380085e-01 3.36854488e-01 -4.15063351e-01 -5.85261285e-01
-1.46496153e+00 3.38669389e-01 2.54364431e-01 -6.77031279e-01
-9.33905303e-01 -1.64364249e-01 -5.04303634e-01 -9.05682221e-02
5.37988305e-01 -7.05368459e-01 -5.78598201e-01 7.08502829e-02
8.44459653e-01 1.05539203e+00 2.96610713e-01 -3.72853190e-01
-8.42776418e-01 2.42467895e-01 4.60144609e-01 1.33203834e-01
1.42855024e+00 1.33720368e-01 3.27040642e-01 5.97117424e-01
1.03888929e+00 -2.33055487e-01 -7.96694398e-01 1.85416535e-01
8.02931547e-01 -3.22556973e-01 -3.36681187e-01 -9.98062968e-01
-6.56421006e-01 6.19003832e-01 1.17255449e+00 2.32577935e-01
1.09550536e+00 -2.90450454e-01 4.66042429e-01 5.86783230e-01
5.10671914e-01 -1.33099854e+00 5.64946651e-01 4.93354440e-01
4.71750468e-01 -1.35168278e+00 1.84976533e-02 1.04596466e-01
-4.77196932e-01 7.34565198e-01 4.43992198e-01 4.24426347e-02
8.38642359e-01 3.75241816e-01 3.46395709e-02 -2.58624494e-01
-2.78729469e-01 -1.71388850e-01 5.92605025e-02 8.36613894e-01
9.34500098e-01 1.22261316e-01 -5.04934788e-01 6.63539708e-01
5.59518971e-02 7.34289408e-01 4.13813919e-01 1.60550523e+00
-3.24706703e-01 -1.23359156e+00 -4.88492936e-01 6.78754330e-01
-9.28144038e-01 5.70773661e-01 -3.31625879e-01 8.65521073e-01
1.66364849e-01 1.22958648e+00 -2.65327692e-01 1.25297785e-01
7.59036243e-01 4.46005613e-01 6.05097711e-01 -7.69097686e-01
-9.36401010e-01 -3.53268027e-01 3.36399436e-01 -9.70390737e-01
-8.67358029e-01 -8.03435922e-01 -1.21998584e+00 -3.20626758e-02
3.70370388e-01 1.46995068e-01 7.79802442e-01 1.18934917e+00
3.27188760e-01 9.45721447e-01 2.74144728e-02 -5.79268396e-01
-8.29200074e-02 -1.26917207e+00 -7.38918245e-01 7.88743615e-01
7.89287239e-02 -8.12768340e-01 -4.07523848e-02 8.48282054e-02] | [7.393789291381836, 0.7194857001304626] |
44fd1279-786d-442b-b6fd-534be4c604b2 | a-pointer-network-architecture-for-joint | null | null | https://aclanthology.org/2020.findings-emnlp.391 | https://aclanthology.org/2020.findings-emnlp.391.pdf | A Pointer Network Architecture for Joint Morphological Segmentation and Tagging | Morphologically Rich Languages (MRLs) such as Arabic, Hebrew and Turkish often require Morphological Disambiguation (MD), i.e., the prediction of morphological decomposition of tokens into morphemes, early in the pipeline. Neural MD may be addressed as a simple pipeline, where segmentation is followed by sequence tagging, or as an end-to-end model, predicting morphemes from raw tokens. Both approaches are sub-optimal; the former is heavily prone to error propagation, and the latter does not enjoy explicit access to the basic processing units called morphemes. This paper offers MD architecture that combines the symbolic knowledge of morphemes with the learning capacity of neural end-to-end modeling. We propose a new, general and easy-to-implement Pointer Network model where the input is a morphological lattice and the output is a sequence of indices pointing at a single disambiguated path of morphemes. We demonstrate the efficacy of the model on segmentation and tagging, for Hebrew and Turkish texts, based on their respective Universal Dependencies (UD) treebanks. Our experiments show that with complete lattices, our model outperforms all shared-task results on segmenting and tagging these languages. On the SPMRL treebank, our model outperforms all previously reported results for Hebrew MD in realistic scenarios. | ['Reut Tsarfaty', 'Amit Seker'] | 2020-11-01 | null | null | null | findings-of-the-association-for-computational | ['morphological-disambiguation'] | ['natural-language-processing'] | [-1.06051922e-01 3.28792721e-01 1.50423851e-02 -3.32022130e-01
-7.19295025e-01 -1.16378355e+00 2.84537584e-01 4.60370272e-01
-7.85728455e-01 6.04403973e-01 1.50871336e-01 -9.28859949e-01
2.69915402e-01 -8.64473999e-01 -5.70003390e-01 -4.47761297e-01
-2.25384876e-01 8.33939254e-01 4.34789330e-01 -2.78609574e-01
6.99981814e-03 5.19299090e-01 -9.27728832e-01 3.45600963e-01
8.45740616e-01 6.18538380e-01 3.63772452e-01 5.06224155e-01
-4.96039689e-01 4.32373226e-01 -5.59207141e-01 -8.58963490e-01
2.17449680e-01 -3.49276885e-02 -1.22322237e+00 -2.70028234e-01
2.71549135e-01 8.69681388e-02 3.04156721e-01 1.20558429e+00
3.72216344e-01 8.93589226e-04 4.40682948e-01 -5.61682642e-01
-7.31008768e-01 1.27895033e+00 -4.17895913e-01 1.58158571e-01
1.31643459e-01 -8.59024599e-02 1.34336495e+00 -8.51385415e-01
7.05407858e-01 1.33894050e+00 7.42240191e-01 5.11476099e-01
-1.26240981e+00 -3.40535551e-01 3.01810980e-01 -2.96300347e-03
-1.33639419e+00 -3.95779967e-01 2.77522683e-01 -4.19798106e-01
1.59295166e+00 -1.17804624e-01 2.56288737e-01 3.16492081e-01
1.83903147e-02 7.77345002e-01 7.80163884e-01 -7.69073606e-01
1.73144177e-01 -1.96082801e-01 3.41973484e-01 1.02859879e+00
2.10815355e-01 4.83400971e-02 -3.84010345e-01 1.24391504e-01
5.98662496e-01 -6.41325891e-01 8.99468958e-02 1.79361641e-01
-9.27257240e-01 7.67495096e-01 2.75461257e-01 6.23311341e-01
-2.82771438e-01 -2.08019897e-01 6.22959971e-01 3.90577972e-01
5.11273503e-01 2.53831416e-01 -9.84886944e-01 7.52956048e-02
-8.65599632e-01 -1.21443830e-01 1.02037811e+00 7.87683308e-01
9.35373783e-01 1.09757826e-01 1.20504908e-01 8.74767423e-01
3.64448041e-01 2.35359326e-01 6.37635767e-01 -4.99635816e-01
5.57053030e-01 6.40494108e-01 -8.73572901e-02 -3.88132751e-01
-7.00142682e-01 2.16904525e-02 -3.46708864e-01 1.26041085e-01
8.38761449e-01 -4.77923423e-01 -1.06643784e+00 1.90315711e+00
4.14935946e-01 -1.66955560e-01 3.32066953e-01 6.86906695e-01
6.82226598e-01 7.45316327e-01 5.56409955e-01 -9.16510746e-02
1.70187759e+00 -7.93880939e-01 -5.01437962e-01 -5.22534728e-01
9.57313478e-01 -8.63178372e-01 9.82736111e-01 3.95463616e-01
-1.61080325e+00 -5.20639837e-01 -8.74075770e-01 -3.93968821e-01
-7.82679915e-01 1.52813032e-01 5.29058635e-01 9.11107838e-01
-1.29219055e+00 8.54143977e-01 -1.22743988e+00 -4.15827751e-01
3.96385863e-02 6.51527345e-01 -3.91598642e-01 1.74334943e-01
-1.14494896e+00 1.07360125e+00 1.04661405e+00 2.08421916e-01
-3.29550922e-01 -3.25164229e-01 -1.21135151e+00 2.36267643e-03
4.47534733e-02 -1.92730814e-01 1.39772964e+00 -1.07862020e+00
-1.58729815e+00 1.20644128e+00 -1.18877657e-01 -5.10648191e-01
-7.89360106e-02 -3.06274474e-01 -3.06475163e-01 2.16908678e-02
-8.31496716e-02 8.13637495e-01 2.17314452e-01 -9.91991401e-01
-8.34494114e-01 -3.95331442e-01 -6.40192851e-02 1.25431940e-01
-1.26467302e-01 6.01415336e-01 -4.71422374e-01 -5.56424797e-01
1.14697903e-01 -8.12325120e-01 -2.74353713e-01 -7.41041660e-01
-3.79659951e-01 -5.44739842e-01 3.49423558e-01 -1.21955025e+00
1.32218719e+00 -2.00341153e+00 1.63259834e-01 2.25547224e-01
-4.03437674e-01 7.56503224e-01 -3.46958995e-01 3.51449996e-01
-1.39448032e-01 3.10828060e-01 -3.90251219e-01 -4.67185229e-01
2.82190323e-01 6.60782397e-01 -3.37836482e-02 3.47799480e-01
7.39743412e-01 1.12777865e+00 -1.00363719e+00 -4.57837701e-01
7.00506382e-03 2.51836091e-01 -4.47341114e-01 1.24975257e-01
-4.51469302e-01 3.11334878e-01 1.14958316e-01 6.57346070e-01
5.52215934e-01 2.65655398e-01 7.59531438e-01 4.72682640e-02
-5.45417905e-01 1.05033958e+00 -1.14377046e+00 1.91325557e+00
-4.84921545e-01 1.67482376e-01 2.36260831e-01 -8.59292030e-01
9.01238382e-01 5.32269478e-01 7.75735170e-05 -5.63829243e-01
4.08230983e-02 6.69053733e-01 2.35585943e-01 -2.92292982e-01
6.55218303e-01 -3.47946346e-01 -4.41659510e-01 5.86275637e-01
4.52178538e-01 1.92022875e-01 6.11228585e-01 -1.51821882e-01
1.02632070e+00 5.35856068e-01 6.80059612e-01 -3.58565062e-01
5.19371033e-01 8.88384432e-02 7.96571434e-01 3.45417023e-01
1.67886138e-01 1.45523280e-01 5.58122158e-01 -4.23321426e-01
-7.98626900e-01 -1.18201709e+00 3.45927873e-03 1.66171944e+00
-2.82264918e-01 -4.15944099e-01 -9.34727669e-01 -9.25906181e-01
-3.19921553e-01 7.30823934e-01 -2.48584643e-01 2.79333174e-01
-1.40423930e+00 -8.05485964e-01 9.61395204e-01 6.52387977e-01
8.30242634e-02 -1.53783345e+00 -5.06189942e-01 6.57887995e-01
-3.34763527e-02 -1.14263105e+00 -3.08271825e-01 7.37545311e-01
-8.56605887e-01 -7.92088032e-01 -1.85160786e-01 -1.42330110e+00
4.97300863e-01 -3.62211496e-01 1.32506621e+00 1.62825912e-01
1.21820778e-01 -2.78992206e-01 -4.60276723e-01 -1.86040208e-01
-5.96218526e-01 4.29338992e-01 -1.61841556e-01 -3.08595419e-01
5.79314649e-01 -5.21176636e-01 -8.53790566e-02 -3.25421579e-02
-9.56672192e-01 -2.60104477e-01 6.95697308e-01 7.01964855e-01
6.36668324e-01 -2.67359287e-01 3.42619449e-01 -1.28870928e+00
2.24888608e-01 -5.04982054e-01 -7.51903415e-01 2.65176356e-01
-2.41176486e-01 4.22212481e-01 7.98012137e-01 -2.80005813e-01
-1.19921482e+00 4.85467881e-01 -6.88516796e-01 2.32604384e-01
-5.07962644e-01 9.24425781e-01 -5.74533939e-01 2.18028203e-01
5.39065838e-01 -1.27698630e-01 -4.15183783e-01 -1.02365291e+00
7.42036283e-01 5.50632119e-01 8.93354774e-01 -9.36883271e-01
6.13370359e-01 -9.51296091e-03 -1.67635068e-01 -7.24564731e-01
-7.06150174e-01 -4.79054779e-01 -1.31766403e+00 4.46010977e-01
8.98658097e-01 -6.80881381e-01 -3.25660110e-01 4.47108120e-01
-1.61114967e+00 -6.97359741e-01 -2.44019508e-01 8.95900726e-02
-2.93301135e-01 5.16722023e-01 -1.35077655e+00 -4.71803278e-01
-4.51353282e-01 -8.21355999e-01 9.98492897e-01 5.44346590e-03
-2.78548568e-01 -1.40903544e+00 1.12670094e-01 -2.38401562e-01
6.61213323e-03 -2.03212555e-02 1.55317414e+00 -1.13188696e+00
-2.89048195e-01 1.34114265e-01 -2.13267565e-01 2.41634056e-01
-1.20598739e-02 -1.60931945e-01 -7.23825753e-01 -2.11069345e-01
-3.55724096e-01 -4.22566801e-01 8.46447945e-01 1.23066485e-01
3.53378296e-01 -3.89376491e-01 -1.27554715e-01 5.83878994e-01
1.50901699e+00 2.20192641e-01 5.29691517e-01 3.21787119e-01
6.38531685e-01 8.81977379e-01 5.47920048e-01 4.83813286e-02
6.21278584e-01 3.60541761e-01 2.72513896e-01 -1.02841787e-01
-1.72774389e-01 -1.26999900e-01 8.62615585e-01 1.31362414e+00
-2.24561803e-02 -2.06693873e-01 -1.25453019e+00 7.12216556e-01
-1.60809374e+00 -5.44267833e-01 -1.90380260e-01 2.13068438e+00
1.22539985e+00 9.88451988e-02 1.26773641e-01 7.15310872e-02
6.65275574e-01 -5.20561002e-02 -7.90939927e-02 -8.97906899e-01
-9.02850255e-02 9.70989883e-01 5.14026105e-01 9.33133483e-01
-1.13296902e+00 1.76020622e+00 5.56650257e+00 7.51879334e-01
-1.06092513e+00 4.41525936e-01 1.80494100e-01 3.60812217e-01
-7.82496780e-02 1.74772829e-01 -1.22877800e+00 1.13043785e-01
1.43689275e+00 4.75618124e-01 3.23071897e-01 2.98515171e-01
-1.54022023e-01 -7.15760365e-02 -1.16180956e+00 3.68652642e-01
-2.61964202e-01 -1.00840712e+00 3.81115079e-02 -2.03880370e-01
2.65360475e-01 3.97980541e-01 -3.70374322e-01 3.29621404e-01
9.30268407e-01 -9.51902449e-01 1.07473660e+00 -5.98728321e-02
9.05200720e-01 -9.02978659e-01 7.64864147e-01 3.09661686e-01
-1.50476444e+00 2.07415402e-01 -5.77694654e-01 3.55307981e-02
3.41190308e-01 1.77698568e-01 -9.32682931e-01 6.26040161e-01
1.76721022e-01 4.84688282e-01 -5.87943673e-01 8.67181242e-01
-9.77225661e-01 9.25954044e-01 -2.95572162e-01 6.98691905e-02
5.49608946e-01 -2.13206396e-01 3.08827221e-01 2.12744689e+00
1.58946976e-01 1.62537277e-01 4.47881788e-01 5.22559464e-01
1.04290962e-01 3.88210297e-01 -6.73126057e-02 4.57589291e-02
6.07195914e-01 1.26190698e+00 -1.03137374e+00 -2.14057520e-01
-5.99916458e-01 9.98294294e-01 9.37018633e-01 1.77498907e-01
-3.98891479e-01 -6.09687269e-01 5.81149578e-01 -2.14912042e-01
7.00998604e-01 -5.52683294e-01 -2.79706210e-01 -9.20507133e-01
-2.20668837e-01 -7.12447941e-01 6.69367731e-01 -2.53311843e-01
-1.20377529e+00 7.87759304e-01 -3.52992922e-01 -5.59367120e-01
-4.56006765e-01 -1.21361196e+00 -4.83341306e-01 1.21891165e+00
-1.77789688e+00 -1.35256040e+00 5.75841069e-01 4.70072001e-01
3.44390452e-01 9.79155302e-02 1.08836794e+00 3.57921034e-01
-6.56649172e-01 6.13250732e-01 -1.60106599e-01 7.70298183e-01
4.55400199e-01 -1.69229138e+00 9.27580297e-01 1.16949809e+00
4.56558526e-01 5.84587514e-01 3.20027560e-01 -7.59113729e-01
-9.36872125e-01 -1.22268093e+00 1.56153595e+00 -3.01325709e-01
8.96278024e-01 -5.37909210e-01 -1.09742844e+00 1.03245032e+00
3.55453879e-01 -1.92707241e-01 8.97218049e-01 3.31743538e-01
-3.96122456e-01 3.28091294e-01 -8.16043317e-01 3.82115811e-01
9.83384073e-01 -4.12888616e-01 -9.51099098e-01 -6.86186403e-02
7.54592955e-01 -3.85579556e-01 -9.76220429e-01 -7.70819038e-02
3.57798308e-01 -6.66687489e-01 6.78583860e-01 -7.32712328e-01
4.53128703e-02 -2.31361404e-01 -2.13806421e-01 -1.45906460e+00
-4.13711518e-01 -6.66837275e-01 1.56953484e-01 1.53418624e+00
8.19497943e-01 -3.39587837e-01 5.84913194e-01 2.74105042e-01
-5.70205748e-01 -2.97998726e-01 -9.65017080e-01 -8.50432515e-01
5.29463828e-01 -5.28645575e-01 6.42346621e-01 9.28413093e-01
3.20866704e-01 5.86111426e-01 1.23548701e-01 3.77382070e-01
2.13481471e-01 2.63681486e-02 7.66294822e-02 -1.35186005e+00
-4.25127983e-01 -5.63497484e-01 -2.24717990e-01 -1.02357554e+00
6.43606782e-01 -1.40914547e+00 3.37017745e-01 -1.42243266e+00
-3.98679256e-01 -6.65290594e-01 -3.27687055e-01 1.12978399e+00
-1.73751965e-01 1.95584670e-01 2.45567173e-01 1.18806018e-02
-4.41989839e-01 -7.73319835e-03 6.03693843e-01 1.41356334e-01
-4.01276678e-01 -4.46208775e-01 -3.60058576e-01 1.01604438e+00
8.70752037e-01 -6.83455527e-01 2.91426957e-01 -9.45202291e-01
4.97568965e-01 1.68622091e-01 -1.90442950e-01 -6.20781958e-01
3.00700694e-01 5.52109778e-02 1.04764819e-01 -5.14919519e-01
-4.75125620e-04 -4.59128737e-01 -3.64924550e-01 5.76799750e-01
4.90298681e-02 4.16891605e-01 3.71336251e-01 -1.67616203e-01
-1.42684281e-01 -6.65697277e-01 8.46695483e-01 -3.87481809e-01
-9.01039660e-01 1.55872211e-01 -5.67451715e-01 2.65041679e-01
5.76974511e-01 -1.48659572e-01 -2.06222519e-01 4.70463872e-01
-9.21000183e-01 -9.17396788e-03 3.45068127e-01 7.53967017e-02
2.80758351e-01 -9.12968814e-01 -7.57403374e-01 2.90931791e-01
-3.27646762e-01 9.55160037e-02 -4.87601846e-01 6.42217994e-01
-6.80559993e-01 4.26987261e-01 -2.92356551e-01 -9.01852995e-02
-1.02906108e+00 5.67034364e-01 2.52323151e-01 -8.19877803e-01
-1.82610840e-01 1.09244168e+00 -9.42568406e-02 -8.96915555e-01
8.89762193e-02 -5.01751363e-01 -3.53922576e-01 3.45106363e-01
3.55294406e-01 2.41995856e-01 4.62737501e-01 -9.24277782e-01
-3.98393214e-01 3.74311030e-01 -1.05968483e-01 -2.55472928e-01
1.28040218e+00 -9.15447995e-02 -6.61374271e-01 3.94844085e-01
8.04071784e-01 4.11755890e-01 -1.01408839e+00 -4.27809119e-01
8.74691367e-01 2.81326771e-01 -3.20834190e-01 -1.02385390e+00
-9.31145310e-01 1.05582547e+00 -2.48964923e-03 1.18609689e-01
9.38939452e-01 1.24755770e-01 1.05364060e+00 4.64240879e-01
3.98125589e-01 -9.43128586e-01 -3.74079376e-01 1.29044259e+00
2.60177016e-01 -7.59262741e-01 -5.84195018e-01 -4.34889525e-01
-4.79763597e-01 1.22404540e+00 4.48120445e-01 -2.04218060e-01
5.07557571e-01 5.84390402e-01 2.77538091e-01 6.76814914e-02
-6.67948186e-01 -6.49700582e-01 2.18726873e-01 5.83443820e-01
9.39188957e-01 2.52765149e-01 -6.52306139e-01 9.25624907e-01
-4.53662246e-01 -5.70061147e-01 2.14947224e-01 9.15475488e-01
-6.53538465e-01 -1.69218278e+00 -3.17438096e-01 -1.96729451e-02
-9.02250350e-01 -6.99893951e-01 -3.33609313e-01 9.27132845e-01
6.47196949e-01 8.27137291e-01 3.02221835e-01 -3.18537392e-02
3.41364294e-01 5.21320641e-01 5.99685907e-01 -1.13063037e+00
-1.15770698e+00 1.54890224e-01 3.99548769e-01 -3.90495688e-01
-3.51164728e-01 -8.60477626e-01 -1.83251929e+00 -4.34317999e-02
-2.28634983e-01 3.52529109e-01 4.28559095e-01 1.10093391e+00
-1.37132062e-02 5.46502098e-02 6.91399276e-02 -9.81275320e-01
-3.14462364e-01 -9.70667481e-01 -7.17823982e-01 3.05401355e-01
2.82980292e-03 -1.42109364e-01 2.17003837e-01 3.11116219e-01] | [10.388566970825195, 10.06127643585205] |
140d2ca5-eb9b-4115-b9b0-2daee29fa041 | av-transpeech-audio-visual-robust-speech-to | 2305.15403 | null | https://arxiv.org/abs/2305.15403v1 | https://arxiv.org/pdf/2305.15403v1.pdf | AV-TranSpeech: Audio-Visual Robust Speech-to-Speech Translation | Direct speech-to-speech translation (S2ST) aims to convert speech from one language into another, and has demonstrated significant progress to date. Despite the recent success, current S2ST models still suffer from distinct degradation in noisy environments and fail to translate visual speech (i.e., the movement of lips and teeth). In this work, we present AV-TranSpeech, the first audio-visual speech-to-speech (AV-S2ST) translation model without relying on intermediate text. AV-TranSpeech complements the audio stream with visual information to promote system robustness and opens up a host of practical applications: dictation or dubbing archival films. To mitigate the data scarcity with limited parallel AV-S2ST data, we 1) explore self-supervised pre-training with unlabeled audio-visual data to learn contextual representation, and 2) introduce cross-modal distillation with S2ST models trained on the audio-only corpus to further reduce the requirements of visual data. Experimental results on two language pairs demonstrate that AV-TranSpeech outperforms audio-only models under all settings regardless of the type of noise. With low-resource audio-visual data (10h, 30h), cross-modal distillation yields an improvement of 7.6 BLEU on average compared with baselines. Audio samples are available at https://AV-TranSpeech.github.io | ['Zhou Zhao', 'Xiang Yin', 'Jinglin Liu', 'Lichao Zhang', 'Jinzheng He', 'Zhenhui Ye', 'Linjun Li', 'Yi Ren', 'Xize Cheng', 'Huadai Liu', 'Rongjie Huang'] | 2023-05-24 | null | null | null | null | ['speech-to-speech-translation'] | ['speech'] | [ 4.32516009e-01 2.52464935e-02 -1.96046144e-01 -1.10445678e-01
-1.61233878e+00 -7.11052716e-01 7.06717908e-01 -2.12096259e-01
-6.05321070e-03 3.25325787e-01 5.74675918e-01 -7.80697584e-01
7.09685326e-01 -4.28070650e-02 -7.49286532e-01 -5.74331999e-01
5.31395555e-01 2.27329835e-01 1.92171648e-01 -2.44618893e-01
-1.15594119e-01 8.21755975e-02 -1.37399757e+00 7.61634529e-01
6.22110367e-01 8.65553737e-01 5.52485943e-01 1.00038731e+00
-1.17514841e-01 3.52000773e-01 -5.44106543e-01 -1.48810714e-01
3.28764409e-01 -7.61990249e-01 -5.56027353e-01 2.95940563e-02
6.74290776e-01 -4.50932533e-01 -4.55335200e-01 7.72071779e-01
9.85114336e-01 1.61054675e-02 4.16431218e-01 -1.31645775e+00
-9.46128905e-01 3.48248154e-01 -5.10720074e-01 1.01470724e-01
5.63108921e-01 4.26991403e-01 1.16306818e+00 -1.47221780e+00
6.17562175e-01 1.46475208e+00 3.73318911e-01 8.28309298e-01
-1.41810429e+00 -7.48522162e-01 -1.00522406e-01 2.70181060e-01
-1.38503528e+00 -1.29842651e+00 5.69388151e-01 -2.93008745e-01
9.14287269e-01 5.46402633e-01 5.73742688e-01 1.62654674e+00
-2.64753401e-01 1.04424858e+00 1.20266616e+00 -6.08198285e-01
3.41594703e-02 1.27907276e-01 -4.94048566e-01 2.73091644e-01
-5.16377211e-01 2.82131106e-01 -1.16444075e+00 4.88895215e-02
3.82317483e-01 -5.38004577e-01 -4.38059598e-01 5.95601164e-02
-1.30374193e+00 5.28158128e-01 1.08590856e-01 8.84311497e-02
4.62217517e-02 2.28332486e-02 6.10207558e-01 6.44542813e-01
4.43424970e-01 -6.29574573e-03 -3.38411897e-01 -3.97506505e-01
-1.16202605e+00 -9.48255733e-02 5.02669811e-01 1.21543086e+00
2.30314359e-01 4.56110030e-01 -3.03939611e-01 1.16321266e+00
5.10808110e-01 1.06878781e+00 5.38606822e-01 -9.15103793e-01
1.08747780e+00 -2.20739469e-01 -1.26205862e-01 -4.02567834e-01
1.12338051e-01 -1.48830071e-01 -5.98258495e-01 7.27219582e-02
1.69526204e-01 -7.41791278e-02 -1.18620563e+00 1.51799417e+00
2.99586087e-01 5.49305044e-02 2.06703335e-01 9.96323586e-01
1.19691372e+00 1.22433746e+00 -2.83206284e-01 -3.50827038e-01
1.32576263e+00 -1.21007478e+00 -8.85361373e-01 -3.44549745e-01
4.11456823e-01 -1.29978585e+00 1.61164129e+00 3.04598957e-01
-1.09962809e+00 -6.18090391e-01 -1.02560353e+00 -2.50446260e-01
9.21591818e-02 2.10034445e-01 -2.31681958e-01 5.80319464e-01
-1.51342583e+00 -3.00149564e-02 -8.80852163e-01 -4.45182145e-01
4.08954732e-02 2.17003021e-02 -3.58372986e-01 -7.22166821e-02
-1.23026550e+00 5.64976573e-01 -8.33409727e-02 1.24748070e-02
-1.16917539e+00 -6.15213454e-01 -1.01561904e+00 -1.77926719e-01
5.03952205e-01 -5.80069184e-01 1.66475666e+00 -1.03676844e+00
-1.93282986e+00 6.71374381e-01 -6.05963051e-01 -2.30700165e-01
6.73915565e-01 -1.87451556e-01 -6.20746493e-01 4.24638540e-01
1.21244319e-01 8.89664590e-01 1.22804284e+00 -1.45792520e+00
-4.44812208e-01 -1.32106803e-02 -4.90143239e-01 4.56086546e-01
-3.12533677e-01 3.02228779e-01 -8.24861169e-01 -9.62240160e-01
1.56685531e-01 -1.10074675e+00 4.34061438e-01 2.31009707e-01
-6.08704984e-01 3.40322666e-02 1.10909820e+00 -1.01912487e+00
9.41513002e-01 -2.48029041e+00 -1.84289906e-02 -2.61960775e-01
-8.65988992e-03 4.81873274e-01 -5.90445697e-01 6.78094268e-01
1.63612962e-02 -2.94926809e-03 -7.10566491e-02 -9.58979607e-01
3.32480185e-02 1.22674771e-01 -4.70387489e-01 2.65419990e-01
6.20733872e-02 8.95689368e-01 -7.35935330e-01 -6.16836250e-01
2.30760470e-01 7.80958295e-01 -4.34302479e-01 2.65109420e-01
-2.58170497e-02 6.76188588e-01 1.03516378e-01 8.41691911e-01
4.82445717e-01 1.88765361e-03 -4.13642824e-02 -6.53788298e-02
-3.87548804e-02 7.91566133e-01 -7.50937641e-01 1.82319665e+00
-6.37476206e-01 1.18367648e+00 4.19124067e-01 -5.42795897e-01
8.71189833e-01 7.98889875e-01 2.21678853e-01 -1.00865436e+00
-7.23851994e-02 3.99753422e-01 -2.18185678e-01 -5.75295448e-01
5.39962411e-01 -2.58333355e-01 4.78063934e-02 1.75850704e-01
3.38436663e-02 -4.83382821e-01 -1.60809368e-01 2.38603473e-01
8.03889155e-01 -9.94090289e-02 -2.39636377e-02 2.77181119e-01
3.49653870e-01 -1.75548285e-01 5.15391707e-01 4.15362716e-01
-3.82304758e-01 1.14328861e+00 6.34475946e-02 3.29491884e-01
-1.17386675e+00 -1.45803988e+00 1.72944427e-01 1.29865885e+00
5.65972067e-02 -5.56252480e-01 -6.99739516e-01 -3.86509031e-01
-3.00406516e-01 7.94346869e-01 -9.58245248e-02 -3.68279889e-02
-3.72559994e-01 5.11672674e-03 8.81501794e-01 3.83248895e-01
2.38906056e-01 -8.53118956e-01 -2.90529728e-02 4.78550494e-02
-7.29066372e-01 -1.52095842e+00 -1.05936158e+00 -1.44276842e-01
-5.04331529e-01 -3.80508721e-01 -1.09398544e+00 -9.40509319e-01
2.39913598e-01 6.24487817e-01 8.02076638e-01 -3.39457810e-01
-4.63724881e-02 4.26800936e-01 -4.61998373e-01 -2.00229287e-01
-9.72580552e-01 -2.60641072e-02 1.96848214e-01 4.48676944e-02
4.53276820e-02 -5.44865668e-01 -5.14663517e-01 3.95045072e-01
-5.68163812e-01 3.07612181e-01 5.92008352e-01 9.97563362e-01
5.74646175e-01 -5.16301215e-01 5.20188510e-01 -1.78822786e-01
5.53423643e-01 -1.47789970e-01 -3.21746737e-01 6.32807380e-03
-5.87022245e-01 -3.14672738e-01 5.59936106e-01 -6.02711380e-01
-1.05449378e+00 -2.49899700e-01 -2.74506509e-01 -7.66103089e-01
-7.13021830e-02 3.19543779e-01 -3.05367589e-01 3.72950673e-01
5.49109280e-01 4.84018296e-01 4.95255701e-02 -6.25979722e-01
4.93969470e-01 1.37148309e+00 9.12302077e-01 -3.87500674e-01
7.92683840e-01 2.86623478e-01 -4.57919061e-01 -1.10889304e+00
-3.87750477e-01 -4.69800323e-01 -2.22524926e-01 -1.56789094e-01
7.59863615e-01 -1.29682291e+00 -3.34898651e-01 3.70507807e-01
-1.14028287e+00 -6.42244577e-01 -1.77474186e-01 5.95135927e-01
-5.35313308e-01 4.79191363e-01 -6.88679636e-01 -8.92119646e-01
-5.19409120e-01 -1.36402476e+00 1.48070395e+00 -2.64015853e-01
-1.64028093e-01 -4.67458010e-01 2.29420097e-04 9.06435192e-01
1.44174591e-01 -3.25029075e-01 5.06397188e-01 -4.63232309e-01
-4.08954412e-01 1.88235864e-01 -1.97836414e-01 4.88544375e-01
2.30663106e-01 1.85432471e-02 -1.21459222e+00 -5.82663417e-01
-3.05178344e-01 -4.41889197e-01 6.90626383e-01 1.93055078e-01
6.44980431e-01 -4.62184906e-01 2.39072777e-02 5.40256202e-01
7.48848855e-01 2.17254981e-01 4.43271756e-01 6.93247169e-02
7.15037584e-01 5.12863100e-01 7.36907005e-01 2.88156509e-01
5.90560436e-01 1.14147818e+00 2.02293590e-01 -2.16332778e-01
-1.04818118e+00 -7.29796231e-01 1.14345098e+00 1.56792080e+00
3.81916434e-01 -5.04480183e-01 -8.54786217e-01 8.03383291e-01
-1.38760114e+00 -7.00073242e-01 -1.19684823e-01 2.13089609e+00
1.01984644e+00 2.99071912e-02 1.87239647e-01 2.14374691e-01
8.75077903e-01 2.59700000e-01 -5.31782806e-01 -4.40015584e-01
-1.87068671e-01 1.30639132e-02 1.98002517e-01 8.34151924e-01
-7.28865862e-01 1.04715180e+00 5.72648382e+00 9.99704957e-01
-1.55816638e+00 5.03248513e-01 4.12796289e-01 -5.26223302e-01
-3.85484904e-01 -4.53201756e-02 -6.08986378e-01 6.12401426e-01
1.18256247e+00 3.96147370e-02 5.18671095e-01 3.52295637e-01
7.19884574e-01 2.33237728e-01 -1.03330410e+00 1.29900956e+00
1.29679382e-01 -1.09559047e+00 1.17071476e-02 9.18921642e-03
5.89611351e-01 4.62741464e-01 4.11417723e-01 2.10386306e-01
2.70217955e-02 -8.70183408e-01 1.24264348e+00 -1.37677073e-01
1.53481555e+00 -4.33858126e-01 2.30560303e-01 1.96716800e-01
-1.44664514e+00 1.48374006e-01 1.30534574e-01 4.32918906e-01
4.45342720e-01 1.94963649e-01 -1.19314218e+00 5.49039245e-01
8.04377973e-01 5.45069814e-01 -1.91831246e-01 5.96699119e-01
-4.52709347e-01 1.03411222e+00 -3.67244303e-01 2.64665455e-01
1.17215790e-01 2.43647426e-01 1.04452157e+00 1.36638868e+00
5.84537804e-01 -2.23378509e-01 1.84752584e-01 5.64938605e-01
-7.45512843e-02 2.95071363e-01 -5.85617483e-01 -2.26781964e-01
8.46602023e-01 7.74384379e-01 -2.73673594e-01 -3.26939404e-01
-5.74122906e-01 1.32983398e+00 -2.55494624e-01 5.59910953e-01
-6.93120003e-01 -3.24638188e-01 7.07074702e-01 2.73944914e-01
3.26167881e-01 -3.18338931e-01 -1.80289090e-01 -1.08786285e+00
2.93684661e-01 -1.17265582e+00 9.67763290e-02 -1.12026000e+00
-9.17358398e-01 7.35089540e-01 -2.21517846e-01 -1.55104423e+00
-4.36567605e-01 -3.07962358e-01 -1.97412446e-01 8.58260810e-01
-1.47588265e+00 -1.41719675e+00 -5.46857528e-02 6.76553249e-01
1.06809640e+00 -2.03182176e-01 5.57377338e-01 3.97860289e-01
-3.36562783e-01 1.19545281e+00 2.80019015e-01 4.95312139e-02
1.28514564e+00 -8.91830206e-01 6.97037816e-01 9.37407434e-01
3.71426135e-01 2.84066826e-01 9.10269618e-01 -4.80833620e-01
-1.54210401e+00 -1.05488801e+00 1.01354432e+00 -2.19839990e-01
6.13922894e-01 -7.74317920e-01 -7.23015726e-01 5.32011747e-01
4.23993200e-01 1.59244001e-01 6.44362986e-01 -3.12887639e-01
-6.63274705e-01 -2.11675420e-01 -8.58658493e-01 9.11810696e-01
1.14829946e+00 -1.17857754e+00 -4.39618438e-01 2.78804190e-02
1.29225838e+00 -4.97950822e-01 -6.63145781e-01 1.68226644e-01
5.17377436e-01 -6.48472548e-01 1.00752175e+00 -1.53612271e-01
2.62529999e-01 -4.62263465e-01 -5.42421758e-01 -1.42827737e+00
2.64838606e-01 -1.42722189e+00 -1.34708270e-01 1.43805373e+00
5.54253936e-01 -6.03694081e-01 3.77484918e-01 4.21312116e-02
-5.13098478e-01 -3.30898494e-01 -1.29650056e+00 -1.05021346e+00
-6.64226115e-02 -6.36867642e-01 3.27524900e-01 8.82987142e-01
1.46789372e-01 5.69535613e-01 -7.61958361e-01 2.74422765e-01
3.07176590e-01 -1.61611456e-02 9.86971498e-01 -5.75439036e-01
-5.88516414e-01 -2.80385166e-01 -1.74800932e-01 -1.64706612e+00
-3.31130214e-02 -9.16134715e-01 4.41436976e-01 -1.53649986e+00
-6.88377917e-02 -9.09450948e-02 3.73436250e-02 7.20506608e-01
-6.05388321e-02 5.27557194e-01 5.54822981e-01 3.93882781e-01
-2.67390728e-01 8.99416983e-01 1.38259053e+00 -3.98697585e-01
-4.48270231e-01 -2.11150885e-01 -5.57856381e-01 2.28279680e-01
5.83775938e-01 -3.71117324e-01 -6.13175631e-01 -6.08633399e-01
-2.64260560e-01 4.80001688e-01 2.42388457e-01 -6.29625976e-01
1.05952851e-01 -8.80239978e-02 -9.03077051e-02 -6.54653668e-01
8.58237267e-01 -5.77146471e-01 -8.70751403e-03 1.01347208e-01
-3.71036857e-01 2.20837183e-02 3.72776866e-01 5.43193042e-01
-3.95990312e-01 1.80087849e-01 6.65539384e-01 3.86791408e-01
-3.84336203e-01 2.03043297e-02 -6.47934139e-01 2.06145570e-01
5.91461360e-01 -1.28830314e-01 -4.24514860e-01 -8.80142450e-01
-8.62157404e-01 4.07637469e-02 3.82933885e-01 7.32572675e-01
7.77120829e-01 -1.54803133e+00 -1.03515005e+00 2.06958413e-01
3.49391609e-01 -1.11374274e-01 1.38647065e-01 9.50033903e-01
-2.00544506e-01 3.77963424e-01 2.43517295e-01 -1.01725578e+00
-1.79977405e+00 2.75192112e-01 -1.14665385e-02 2.98341662e-01
-8.78033757e-01 7.90315151e-01 3.21430713e-01 -3.74295056e-01
2.72481799e-01 -2.92519242e-01 4.02496964e-01 -6.22577667e-02
4.63867337e-01 2.43183926e-01 2.26510808e-01 -9.92896438e-01
-3.20910811e-01 2.20869839e-01 -4.55465876e-02 -9.40906823e-01
1.06631589e+00 -6.09100163e-01 4.37221110e-01 7.47804701e-01
1.29655182e+00 2.30538890e-01 -1.41178310e+00 -4.20948654e-01
-4.30535525e-01 -6.59473896e-01 8.83811191e-02 -8.39797795e-01
-9.75624025e-01 1.27183759e+00 5.79352677e-01 -5.13768084e-02
1.16712546e+00 1.73754543e-01 1.11985052e+00 1.55310839e-01
7.32839257e-02 -9.91549015e-01 3.83938849e-01 4.73829567e-01
1.25135863e+00 -1.13673699e+00 -4.23708558e-01 -3.54579955e-01
-8.87928903e-01 8.74793112e-01 3.47236544e-01 6.83235943e-01
2.54914314e-01 3.88508588e-01 7.09700882e-01 2.96207488e-01
-9.63309288e-01 -2.09919661e-01 4.11551565e-01 8.32129300e-01
3.77324104e-01 1.15936667e-01 2.81885892e-01 3.19355220e-01
-4.97196883e-01 -3.36959749e-01 2.70927817e-01 8.44463587e-01
-2.47503281e-01 -1.08864689e+00 -6.81267262e-01 -1.37208372e-01
-2.45492816e-01 -4.72094923e-01 -6.00974798e-01 4.10797060e-01
-2.72231787e-01 1.47926879e+00 -9.93839875e-02 -4.54477072e-01
3.13336909e-01 1.00622207e-01 3.52616251e-01 -4.49923187e-01
-2.47376621e-01 7.56987929e-01 2.83881694e-01 -4.92239743e-01
-4.88642529e-02 -8.08301806e-01 -1.21384299e+00 -3.53435874e-01
-2.65597314e-01 -1.29831970e-01 7.75504768e-01 7.71804988e-01
6.01101816e-01 4.90782350e-01 7.51878440e-01 -9.07061040e-01
-5.86809039e-01 -1.13272250e+00 -2.53770977e-01 1.63127959e-01
8.94999027e-01 -3.18892330e-01 -4.37630802e-01 2.99404681e-01] | [14.488015174865723, 5.402403354644775] |
d85c459b-0d7b-4e94-b0e4-78ddf2542cae | enhancing-real-world-adversarial-patches-with | 2102.05334 | null | https://arxiv.org/abs/2102.05334v2 | https://arxiv.org/pdf/2102.05334v2.pdf | Enhancing Real-World Adversarial Patches through 3D Modeling of Complex Target Scenes | Adversarial examples have proven to be a concerning threat to deep learning models, particularly in the image domain. However, while many studies have examined adversarial examples in the real world, most of them relied on 2D photos of the attack scene. As a result, the attacks proposed may have limited effectiveness when implemented in realistic environments with 3D objects or varied conditions. There are few studies on adversarial learning that use 3D objects, and in many cases, other researchers are unable to replicate the real-world evaluation process. In this study, we present a framework that uses 3D modeling to craft adversarial patches for an existing real-world scene. Our approach uses a 3D digital approximation of the scene as a simulation of the real world. With the ability to add and manipulate any element in the digital scene, our framework enables the attacker to improve the adversarial patch's impact in real-world settings. We use the framework to create a patch for an everyday scene and evaluate its performance using a novel evaluation process that ensures that our results are reproducible in both the digital space and the real world. Our evaluation results show that the framework can generate adversarial patches that are robust to different settings in the real world. | ['Yuval Elovici', 'Lior Rokach', 'Yael Mathov'] | 2021-02-10 | null | null | null | null | ['real-world-adversarial-attack'] | ['adversarial'] | [ 1.54857442e-01 -8.61692652e-02 4.34385240e-01 2.74406988e-02
-4.46917474e-01 -1.10615432e+00 8.07796538e-01 -3.98443878e-01
-4.84830767e-01 4.61334854e-01 -2.81215608e-01 -4.95854914e-01
2.32596606e-01 -1.16895354e+00 -1.05988479e+00 -5.89829087e-01
-2.17174992e-01 1.76825032e-01 4.73826289e-01 -5.13338506e-01
8.78164992e-02 9.15065169e-01 -1.29006588e+00 2.54854321e-01
4.13525790e-01 6.90696061e-01 -9.15206298e-02 9.00537670e-01
3.72000039e-01 6.21311307e-01 -1.54161298e+00 -7.22431898e-01
1.21861410e+00 -1.30507976e-01 -4.24030185e-01 -1.38307005e-01
7.61611342e-01 -7.36652970e-01 -7.02280581e-01 1.27296150e+00
8.73825312e-01 -1.37857959e-01 3.49780113e-01 -1.46196640e+00
-5.68581760e-01 3.67006540e-01 -2.22391009e-01 4.03605355e-03
4.37075108e-01 6.43047810e-01 1.36687234e-01 -8.59205052e-02
5.34716547e-01 1.49988973e+00 6.30673170e-01 7.06878543e-01
-1.07869160e+00 -1.09836936e+00 -5.43668270e-02 1.91144601e-01
-1.14447486e+00 -2.85439286e-02 1.06010640e+00 -2.06283689e-01
8.02055776e-01 3.18157315e-01 5.62150419e-01 1.76045644e+00
5.80210507e-01 4.00357902e-01 1.36654282e+00 -3.63850683e-01
3.15557420e-01 3.37360322e-01 -5.78704059e-01 1.44856274e-01
2.27895111e-01 7.72531211e-01 1.10388868e-01 -2.19719440e-01
7.47900307e-01 -1.42206624e-01 -2.03791216e-01 -3.87697279e-01
-7.87565351e-01 8.27080190e-01 6.89033031e-01 9.59373266e-02
-2.76082575e-01 3.71017694e-01 4.02494252e-01 6.32888973e-01
2.49678925e-01 8.60235810e-01 1.36441477e-02 1.57877102e-01
-5.17553866e-01 5.57893693e-01 8.59243333e-01 7.19420314e-01
3.64015281e-01 3.22184712e-01 2.46938005e-01 3.21236163e-01
8.58100578e-02 6.49584293e-01 2.36881360e-01 -9.18872356e-01
4.02416855e-01 1.31729990e-01 -2.32285485e-02 -1.36925972e+00
7.80131146e-02 -1.94433779e-01 -5.78178406e-01 1.20087600e+00
5.48888266e-01 -2.83332676e-01 -9.73911941e-01 1.69191921e+00
4.29029971e-01 5.67451477e-01 2.34550342e-01 8.65668833e-01
5.79024255e-01 4.70722437e-01 -7.25259036e-02 3.42483461e-01
7.16966689e-01 -5.04223704e-01 -3.27756375e-01 -2.10561171e-01
1.75088212e-01 -1.06226766e+00 1.10117459e+00 6.19504392e-01
-8.55040789e-01 -6.74953818e-01 -1.28049231e+00 6.25953615e-01
-7.18640745e-01 -7.93326795e-01 3.95440400e-01 1.23947048e+00
-8.24739754e-01 6.40085280e-01 -6.47511959e-01 -1.02859206e-01
5.09305775e-01 2.84740746e-01 -5.64435184e-01 -2.28334084e-01
-1.65480506e+00 1.21151519e+00 2.74279326e-01 -2.72607207e-01
-1.59787869e+00 -6.77923083e-01 -7.93080389e-01 -3.71522307e-01
2.78461993e-01 -4.62235898e-01 1.05749810e+00 -1.16666758e+00
-1.45241272e+00 9.29675400e-01 8.01953495e-01 -7.31882572e-01
9.85446692e-01 -2.66078353e-01 -3.76911581e-01 2.55766869e-01
-3.82312566e-01 4.17717844e-01 1.09358585e+00 -1.67151368e+00
-4.07254361e-02 -1.73115030e-01 9.44169462e-01 6.23249821e-02
-1.13466159e-01 -4.78555448e-03 -4.27709557e-02 -1.05645871e+00
-4.61099058e-01 -1.01609540e+00 -4.14242297e-01 2.20511958e-01
-3.63960832e-01 7.18155265e-01 1.28325033e+00 -3.28619897e-01
5.65589190e-01 -2.37620831e+00 -3.96592379e-01 3.01372081e-01
-3.65394913e-02 7.58733153e-01 -2.80580729e-01 5.90103090e-01
-3.74607295e-01 5.69798231e-01 -6.79983944e-02 -1.51888058e-01
-2.64049252e-03 3.55882138e-01 -7.24366069e-01 7.14678168e-01
1.34981632e-01 6.58036411e-01 -9.09426630e-01 -2.80690551e-01
5.02655029e-01 5.69168687e-01 -6.59589291e-01 4.54381585e-01
-1.01885535e-01 4.54084247e-01 -3.76373529e-01 3.79378349e-01
9.77063596e-01 7.03260005e-01 -2.73249090e-01 7.76492581e-02
4.38314855e-01 -2.02503636e-01 -1.30650127e+00 1.29381514e+00
-4.71563846e-01 6.92855954e-01 2.24110540e-02 -8.45059037e-01
8.10128570e-01 2.48937592e-01 8.38453881e-03 -6.02483809e-01
1.71615586e-01 -3.45150866e-02 2.52559423e-01 -6.15086794e-01
2.17270061e-01 -3.17829549e-01 -1.56040505e-01 4.38466161e-01
-2.81483442e-01 -9.03030753e-01 -3.57347190e-01 2.11904153e-01
1.52319598e+00 6.63883090e-02 9.98191610e-02 2.65313715e-01
2.87094980e-01 -7.11046383e-02 1.26839414e-01 1.09348762e+00
-4.36567843e-01 6.87409878e-01 4.12041843e-01 -5.44288635e-01
-1.25314581e+00 -1.44079030e+00 7.33804703e-02 3.17849576e-01
4.43610370e-01 -5.21450043e-02 -8.83068442e-01 -9.78574812e-01
1.93982348e-02 8.97567630e-01 -7.21202374e-01 -5.54218650e-01
-6.51244760e-01 -3.86420459e-01 1.12110007e+00 8.10626224e-02
7.86250472e-01 -1.24602664e+00 -7.68800914e-01 -6.48542717e-02
3.82813126e-01 -1.18109870e+00 -1.12255283e-01 -1.64387852e-01
-3.07724297e-01 -1.17399085e+00 -2.16006607e-01 -5.38830221e-01
5.99340796e-01 1.72215343e-01 1.07898462e+00 1.74348682e-01
-3.89929891e-01 5.70710480e-01 -5.90861261e-01 -8.10907662e-01
-1.24768209e+00 -6.12294436e-01 7.35353380e-02 -2.34967604e-01
1.77319814e-02 -7.29191661e-01 -4.59574103e-01 5.06207764e-01
-1.42236578e+00 -4.62518096e-01 3.42962712e-01 5.94217658e-01
1.71793967e-01 5.87068200e-01 3.87199938e-01 -8.70164812e-01
6.63285434e-01 -4.27154601e-01 -7.17255771e-01 -9.70764607e-02
-3.39991115e-02 -2.61558145e-01 1.08577085e+00 -1.07951653e+00
-6.54611409e-01 -2.62594283e-01 -3.22210044e-01 -8.45631659e-01
-5.02422631e-01 8.93021524e-02 -5.82801580e-01 -5.71960211e-01
1.12106740e+00 -1.68661729e-01 -7.51044527e-02 -1.33103490e-01
2.28145391e-01 5.14406979e-01 5.49624383e-01 -6.11962497e-01
1.63625109e+00 4.91532952e-01 1.12056918e-01 -5.53397536e-01
-3.18297207e-01 3.96279484e-01 -4.04457599e-01 -3.46579641e-01
5.18882632e-01 -5.92380047e-01 -5.55489719e-01 9.21863139e-01
-1.04784262e+00 -6.62197590e-01 -3.26853722e-01 3.35992217e-01
-5.92958152e-01 4.50234950e-01 -3.47302049e-01 -6.66589677e-01
8.32572347e-04 -1.45778823e+00 8.03510249e-01 5.00544123e-02
-2.50772424e-02 -9.36473727e-01 1.35636643e-01 7.99836367e-02
5.72796464e-01 1.01526368e+00 7.07960069e-01 -7.26053238e-01
-6.70724571e-01 -6.99015737e-01 2.93442696e-01 7.58116186e-01
2.02315688e-01 1.62307113e-01 -1.13205969e+00 -4.37222421e-01
2.83147871e-01 -4.33476627e-01 2.44955644e-01 -6.30237982e-02
1.23446167e+00 -4.36168730e-01 3.43006141e-02 7.03422487e-01
1.40221024e+00 4.92777407e-01 1.13133752e+00 6.09362066e-01
4.88822073e-01 2.96977639e-01 6.44129574e-01 6.48465008e-02
-4.11298215e-01 7.34045684e-01 1.15595746e+00 -2.15436146e-01
-1.92707717e-01 -3.41950178e-01 5.22173822e-01 -1.47753805e-01
2.66526401e-01 -5.15738547e-01 -8.66730332e-01 2.51578957e-01
-1.09277070e+00 -1.40163183e+00 4.31959718e-01 2.31356263e+00
7.19582736e-01 6.63971841e-01 -1.21744961e-01 3.08699995e-01
7.46956348e-01 2.87885308e-01 -5.46395838e-01 -6.72534823e-01
-1.12182111e-01 4.84953254e-01 6.48023665e-01 4.57759470e-01
-1.30167019e+00 9.30706024e-01 6.81913996e+00 6.26288533e-01
-1.39436591e+00 -1.32413521e-01 3.51418823e-01 -2.87892912e-02
-1.84452921e-01 -6.90199658e-02 -1.86086997e-01 4.54565644e-01
7.56416738e-01 -3.20698619e-01 4.38346237e-01 9.56597030e-01
1.67109147e-01 2.24068940e-01 -1.03572869e+00 7.71669805e-01
1.17065631e-01 -1.03629267e+00 1.51726976e-01 1.66670799e-01
5.81381917e-01 -1.88111350e-01 4.60365057e-01 3.15060169e-01
7.45205522e-01 -1.38897431e+00 7.42112398e-01 1.25315100e-01
6.29055858e-01 -9.18845952e-01 8.71363461e-01 5.74491918e-01
-4.55351025e-01 5.74367009e-02 -2.11377025e-01 -6.10309765e-02
-2.93742842e-03 1.55626491e-01 -1.21983302e+00 3.77308309e-01
7.33489394e-01 2.26991311e-01 -6.40933335e-01 1.06211364e+00
-4.14032787e-01 7.36658514e-01 -4.10930455e-01 1.90846100e-01
3.27380627e-01 2.33919516e-01 9.03207541e-01 9.34110284e-01
1.25815749e-01 -1.60485357e-01 1.62871450e-01 6.01991236e-01
-8.58644545e-02 -2.09251180e-01 -1.33495700e+00 2.46257499e-01
4.69368964e-01 7.93371975e-01 -4.24941450e-01 7.55369738e-02
-1.70848966e-01 9.54833269e-01 -8.57017785e-02 2.96995521e-01
-1.34131944e+00 -4.45838749e-01 1.03557193e+00 2.11181819e-01
1.29071489e-01 -2.54279822e-01 -6.50587156e-02 -8.40403616e-01
-3.97558175e-02 -1.58790267e+00 1.19208075e-01 -8.54435205e-01
-1.42732310e+00 7.21122026e-01 2.65194654e-01 -1.41447401e+00
-1.60735786e-01 -7.28123188e-01 -9.08004999e-01 8.04717898e-01
-9.47298825e-01 -1.13245058e+00 -2.90415794e-01 8.65756989e-01
2.33887240e-01 -4.30741906e-01 9.77339208e-01 3.04477841e-01
-2.60373116e-01 9.60088968e-01 -1.34313628e-01 4.00744706e-01
8.53426039e-01 -1.16620815e+00 8.42468619e-01 1.17898023e+00
1.21011563e-01 4.11500245e-01 1.07893348e+00 -6.69971228e-01
-1.09459412e+00 -1.14605844e+00 -2.88057506e-01 -7.53541529e-01
6.86160922e-01 -6.10255063e-01 -9.03511167e-01 6.24325931e-01
2.23299697e-01 1.80201486e-01 4.20540720e-01 -6.51235580e-01
-6.67257965e-01 -8.59431475e-02 -1.80415618e+00 1.02643943e+00
8.89142871e-01 -5.18260956e-01 -5.38492739e-01 2.00110644e-01
9.27473009e-01 -7.49011636e-01 -8.34413111e-01 4.59574610e-01
4.86080706e-01 -9.72536743e-01 1.29013681e+00 -6.20301545e-01
4.09240961e-01 -4.94937748e-01 -3.77935737e-01 -1.56220913e+00
1.94639519e-01 -7.23736048e-01 1.12752207e-01 1.01266336e+00
1.22053243e-01 -9.42382574e-01 6.52982354e-01 4.52134103e-01
2.88070012e-02 -4.04902309e-01 -9.94678855e-01 -9.88584697e-01
4.28573728e-01 -7.65558064e-01 8.59640896e-01 8.94249201e-01
-6.95505202e-01 -3.37814212e-01 -5.23945391e-01 6.43895507e-01
6.08087599e-01 -3.40995014e-01 1.42618001e+00 -5.87359309e-01
-5.50541520e-01 -2.02290088e-01 -1.27278042e+00 -3.33995044e-01
2.84398526e-01 -6.32743001e-01 -1.68541163e-01 -9.94392037e-01
-2.82658160e-01 -6.13537431e-01 1.64102390e-02 3.56074989e-01
-1.34443179e-01 5.56347013e-01 4.73751903e-01 -8.22543502e-02
7.49957338e-02 2.13558733e-01 1.27866757e+00 -4.81161743e-01
2.41513535e-01 8.51969272e-02 -5.10151625e-01 8.05186212e-01
1.01213551e+00 -7.51174271e-01 -5.91531098e-01 -4.31981146e-01
-3.61740738e-02 -5.41083872e-01 7.25598156e-01 -1.34078920e+00
-2.32506409e-01 -2.85744756e-01 4.92563426e-01 -1.29469842e-01
4.54331249e-01 -1.25945902e+00 3.81155729e-01 6.90616012e-01
-1.15165360e-01 2.38273088e-02 6.25983655e-01 3.32940698e-01
-1.74084865e-02 -3.18434179e-01 1.09039652e+00 -2.50829577e-01
-5.23305416e-01 2.14925453e-01 -1.24758221e-01 6.40339106e-02
1.67853773e+00 -3.48378837e-01 -5.88086009e-01 -7.28379428e-01
-4.63167936e-01 -2.22396210e-01 1.06485331e+00 5.24857163e-01
6.35535836e-01 -1.17827177e+00 -6.79931045e-01 3.63226563e-01
-2.18455493e-01 1.78409144e-02 5.61249107e-02 -1.37265027e-01
-1.10334229e+00 -3.29605669e-01 -5.47014356e-01 -3.33928853e-01
-1.41422105e+00 1.26109564e+00 7.38584816e-01 -1.88906357e-01
-5.51915586e-01 4.67107892e-01 2.88192689e-01 -4.51303691e-01
4.22502398e-01 2.50455767e-01 9.87665430e-02 -5.02706110e-01
5.13134241e-01 -1.57474607e-01 -6.59760013e-02 -5.16857326e-01
-2.02124685e-01 3.04434508e-01 6.68702424e-02 -1.39520228e-01
1.10369134e+00 3.84470671e-01 3.19981903e-01 2.29772888e-02
1.11876512e+00 3.50363702e-01 -1.26526785e+00 1.06862314e-01
-6.24194503e-01 -9.78482187e-01 -2.55306900e-01 -7.82844067e-01
-1.25100589e+00 6.79979682e-01 8.51962209e-01 4.40375924e-01
1.29832315e+00 -2.90781558e-01 5.86585641e-01 4.16805983e-01
6.84590399e-01 -5.63737214e-01 3.98564845e-01 3.07318717e-01
1.07906997e+00 -9.62356806e-01 1.19987736e-02 -3.14916581e-01
-5.74890733e-01 7.54704237e-01 7.82501996e-01 -5.97117901e-01
7.06903577e-01 6.55650795e-01 5.77212453e-01 1.61931589e-01
-2.83959955e-01 3.23801368e-01 -1.66888490e-01 1.14941549e+00
-2.75108159e-01 -1.37344182e-01 2.75105268e-01 3.62714753e-02
-6.25939429e-01 -4.88209516e-01 8.29533219e-01 1.08210242e+00
4.63763289e-02 -1.33205855e+00 -9.83907819e-01 -1.30575567e-01
-7.77436912e-01 1.81058839e-01 -6.15157485e-01 1.22886884e+00
3.69533420e-01 1.02989554e+00 -2.01736391e-01 -6.14449322e-01
5.96506476e-01 -3.54504257e-01 5.83717108e-01 -4.34569538e-01
-8.62677872e-01 -4.49774414e-01 -3.49562392e-02 -6.59453869e-01
-2.40821093e-01 -2.60261297e-01 -6.99363589e-01 -5.77655911e-01
-6.90363869e-02 -1.45566404e-01 7.72552133e-01 6.16114438e-01
8.56785476e-02 4.38736975e-01 1.01405120e+00 -1.09722161e+00
-6.66510999e-01 -6.58290863e-01 -3.56036186e-01 8.64690602e-01
3.15378845e-01 -6.08079910e-01 -6.08729720e-01 -2.24466138e-02] | [5.461292266845703, 7.849300861358643] |
0b242795-0217-4d13-86dc-89fdea44a444 | one-sided-box-filter-for-edge-preserving | 2108.05021 | null | https://arxiv.org/abs/2108.05021v1 | https://arxiv.org/pdf/2108.05021v1.pdf | One-Sided Box Filter for Edge Preserving Image Smoothing | Image smoothing is a fundamental task in signal processing. For such task, box filter is well-known. However, box filter can not keep some features of the signal, such as edges, corners and the jump in the step function. In this paper, we present a one-sided box filter that can smooth the signal but keep the discontinuous features in the signal. More specifically, we perform box filter on eight one-sided windows, leading to a one-sided box filter that can preserve corners and edges. Our filter inherits the constant $O(1)$ computational complexity of the original box filter with respect to the window size and also the linear $O(N)$ computational complexity with respect to the total number of samples. We performance several experiments to show the efficiency and effectiveness of this filter. We further compare our filter with other the-state-of-the-art edge preserving methods. Our filter can be deployed in a large range of applications where the classical box filter is adopted. | ['Yuanhao Gong'] | 2021-08-11 | null | null | null | null | ['image-smoothing'] | ['computer-vision'] | [ 1.58743218e-01 -1.41905412e-01 3.30486357e-01 -2.36256778e-01
-4.59442198e-01 -2.64389932e-01 1.37015963e-02 1.47069827e-01
-7.02907979e-01 5.12863517e-01 3.23381828e-04 -1.55306488e-01
1.80071384e-01 -7.90668309e-01 -6.30646348e-01 -6.62671149e-01
-4.75265056e-01 -6.83352411e-01 1.00174534e+00 -2.64375687e-01
2.64885724e-01 3.95284921e-01 -1.17256796e+00 5.80512658e-02
9.30070579e-01 1.15186024e+00 -3.44320163e-02 6.67368591e-01
5.66713735e-02 2.70235330e-01 -4.89586771e-01 -3.75976339e-02
3.88187557e-01 -4.51943189e-01 -1.09932631e-01 1.24545164e-01
3.13456059e-01 -1.74996689e-01 -3.82428169e-01 1.35024369e+00
5.55570960e-01 3.80560130e-01 3.57708484e-01 -9.26755607e-01
-2.86697507e-01 2.71460593e-01 -9.43828106e-01 2.97396213e-01
-5.87591417e-02 -1.66169584e-01 4.24791992e-01 -1.02355909e+00
5.43100357e-01 1.12263620e+00 1.03934884e+00 3.15031767e-01
-1.23766136e+00 -6.75633729e-01 2.68083990e-01 2.10435033e-01
-1.38231397e+00 -2.89356351e-01 8.12159002e-01 -1.81954727e-01
3.87940258e-01 4.69228983e-01 6.43209338e-01 9.19727311e-02
5.37796676e-01 5.37515640e-01 1.26061296e+00 -4.79084939e-01
2.80142397e-01 -9.29533020e-02 3.57510984e-01 7.65665472e-01
2.82593042e-01 -9.33066418e-04 -3.37586254e-01 -1.56878233e-01
9.42311585e-01 2.42284611e-01 -6.02011263e-01 -9.26425830e-02
-9.95782852e-01 6.50473893e-01 4.28552598e-01 3.13847750e-01
-1.19604290e-01 2.57388055e-01 4.20365185e-01 4.13759112e-01
7.63856411e-01 -1.16664290e-01 -2.30488911e-01 1.57769248e-01
-1.22906089e+00 2.51945674e-01 5.32204270e-01 8.22755516e-01
7.09856629e-01 2.01973826e-01 -3.54652852e-01 4.75558370e-01
1.03521466e-01 3.53456467e-01 2.70770222e-01 -7.47395933e-01
1.72929898e-01 4.26232606e-01 9.29285139e-02 -1.02007985e+00
-4.25683290e-01 -4.20841932e-01 -1.12778819e+00 7.31045485e-01
6.00895405e-01 -3.33594054e-01 -6.76427424e-01 1.43359733e+00
3.77494454e-01 4.47831303e-01 -3.83375525e-01 6.24733210e-01
6.13276124e-01 7.31487155e-01 -4.40610796e-01 -5.19052505e-01
1.79072714e+00 -8.83139014e-01 -1.06052375e+00 -1.87434152e-01
2.61331171e-01 -9.40399170e-01 8.94131660e-01 4.35088485e-01
-1.23092103e+00 -5.88478148e-01 -1.20911121e+00 -4.89205606e-02
-1.19636163e-01 1.51635230e-01 2.78977364e-01 7.44334161e-01
-7.80026197e-01 8.27591538e-01 -9.51513410e-01 -2.40454488e-02
3.75766128e-01 2.46538460e-01 -3.58066291e-01 1.54864565e-01
-8.59113216e-01 5.22247195e-01 -5.40454239e-02 2.43574083e-01
-1.87336206e-01 -8.64206493e-01 -8.13901603e-01 3.75396609e-01
5.85037053e-01 -2.14933157e-01 1.00528133e+00 -7.06134975e-01
-1.43052053e+00 4.78540480e-01 -5.77478230e-01 -5.76300204e-01
6.99110985e-01 -3.70423853e-01 -3.99850637e-01 1.61471888e-01
-2.71027714e-01 -4.21975516e-02 1.20423937e+00 -7.59483099e-01
-5.25591791e-01 -3.11136931e-01 -2.51834780e-01 -1.91129833e-01
-2.58921921e-01 9.40762609e-02 -3.19413453e-01 -1.16179907e+00
2.92311937e-01 -6.87398672e-01 -3.91997129e-01 6.63825452e-01
-1.33762911e-01 7.60505348e-02 9.97249544e-01 -7.79468417e-01
1.66149044e+00 -2.71036386e+00 -4.57356423e-01 3.53820026e-01
2.07000479e-01 2.88283974e-01 3.26089233e-01 4.19966936e-01
-6.22966550e-02 -1.07353844e-01 -5.08803427e-01 -1.96685195e-01
-2.90721476e-01 -3.08089137e-01 -2.32306898e-01 1.04445708e+00
9.53179672e-02 3.69987816e-01 -4.65227067e-01 -5.09964883e-01
6.91299886e-02 5.91065288e-01 -6.23725772e-01 -1.41015694e-01
4.01404709e-01 2.58938372e-01 -6.24660999e-02 2.35137239e-01
1.15577388e+00 1.16557017e-01 -1.71837494e-01 -3.55634063e-01
-6.12244666e-01 -1.11559138e-01 -1.81864977e+00 1.30603445e+00
-1.42725393e-01 8.99630129e-01 6.13951147e-01 -8.22384596e-01
1.10890424e+00 2.25885078e-01 2.39404306e-01 -3.58962089e-01
1.00634456e-01 3.78562927e-01 -7.63307884e-02 -2.21860245e-01
3.03416163e-01 -3.86125803e-01 2.35287085e-01 1.44113168e-01
-2.58512855e-01 3.22715864e-02 2.07879886e-01 -1.06214322e-01
1.03813374e+00 -2.48106688e-01 7.50957608e-01 -7.91395247e-01
6.97632670e-01 -3.61768246e-01 7.38148093e-01 4.55343068e-01
-4.45194423e-01 4.96678144e-01 4.91031617e-01 -3.96223336e-01
-7.96257734e-01 -7.32641399e-01 -1.70290917e-01 6.76674545e-01
2.54848868e-01 -2.61803418e-01 -9.97248530e-01 -2.73751557e-01
-1.28532186e-01 2.87677765e-01 -5.99375546e-01 -3.15566882e-02
-1.02277708e+00 -4.25688982e-01 2.70113200e-01 5.21719396e-01
8.04006696e-01 -6.98124886e-01 -1.09169436e+00 3.42506558e-01
3.92513156e-01 -7.67796814e-01 -1.19825232e+00 9.79538262e-02
-1.19217050e+00 -1.08641946e+00 -9.04005647e-01 -1.03734124e+00
8.55197370e-01 5.39558768e-01 5.78180671e-01 1.29101425e-01
-2.00877622e-01 -2.49971330e-01 -1.52844369e-01 -4.86900836e-01
1.38485849e-01 -5.35834134e-01 -2.23985463e-01 1.80435762e-01
-1.70755759e-02 -4.40141499e-01 -1.08929396e+00 3.69601935e-01
-9.44568217e-01 -2.86988050e-01 4.16999310e-01 8.26062858e-01
6.49929762e-01 6.85852766e-01 4.31203216e-01 -9.69423592e-01
6.15013778e-01 2.24523604e-01 -8.88576865e-01 -1.07321896e-01
-2.94831783e-01 -5.10931760e-02 8.15434456e-01 -5.47426820e-01
-1.04473042e+00 2.05842733e-01 -1.21839799e-01 -6.61323518e-02
3.52523983e-01 2.08160684e-01 1.28580108e-02 -5.24925470e-01
5.80346465e-01 1.95354491e-01 8.26617107e-02 -6.76230729e-01
3.77135336e-01 3.87874722e-01 7.70450056e-01 -5.49957417e-02
8.15662861e-01 1.11896181e+00 2.88481504e-01 -1.16882861e+00
-3.07967901e-01 -4.86975402e-01 -2.50169098e-01 -2.98700072e-02
4.94430542e-01 -6.61156058e-01 -9.84396935e-01 4.82465714e-01
-1.06299233e+00 -5.36854900e-02 -3.84651989e-01 5.20049810e-01
-3.32753748e-01 6.66385710e-01 -6.09878361e-01 -1.13699782e+00
-5.42353868e-01 -7.92049527e-01 6.80105984e-01 3.39206368e-01
-2.80961720e-03 -8.17629635e-01 -9.28494781e-02 -6.99316680e-01
4.65849787e-01 4.54815090e-01 6.45296097e-01 -1.81774363e-01
-3.28969270e-01 -3.62286985e-01 -1.46038130e-01 3.16936195e-01
1.74858402e-02 -8.16144869e-02 -7.43274033e-01 -4.61595178e-01
5.52757919e-01 5.04006684e-01 1.05666876e+00 8.24915409e-01
9.32515860e-01 -2.88126469e-01 -3.29796284e-01 6.34175301e-01
1.57077491e+00 4.28406417e-01 8.01950157e-01 1.72865346e-01
4.50587273e-03 4.23568428e-01 5.80691218e-01 3.95776689e-01
-2.99348593e-01 5.05061746e-01 3.83210741e-02 -5.21030962e-01
-3.77371535e-02 4.24757823e-02 2.74669826e-01 6.49951994e-01
4.35010344e-03 2.64626980e-01 -3.50479633e-01 4.04528052e-01
-1.69213986e+00 -1.00311136e+00 -4.02605116e-01 2.45275474e+00
7.35853314e-01 4.84535545e-01 1.77637711e-01 6.62334085e-01
8.86704504e-01 2.51165748e-01 -3.77685398e-01 -4.71063256e-01
5.18240966e-03 5.23674726e-01 8.32529902e-01 6.36500359e-01
-1.22072756e+00 5.12819767e-01 6.47419882e+00 9.90823030e-01
-1.08716881e+00 -9.71563999e-03 2.29838103e-01 1.10603109e-01
1.29275411e-01 -1.56746563e-02 -7.71057725e-01 4.52989250e-01
3.23441446e-01 -4.83993173e-01 1.42551780e-01 8.25484514e-01
4.96175826e-01 -3.07134748e-01 -8.18538547e-01 9.35310125e-01
-9.75302532e-02 -1.24063671e+00 -6.00407898e-01 -2.15043813e-01
4.46833014e-01 -7.33862817e-01 -2.41796017e-01 -9.53663737e-02
-3.25612038e-01 -7.46799648e-01 7.47226000e-01 3.01135451e-01
6.54006958e-01 -7.51704693e-01 5.50976038e-01 3.95590305e-01
-1.74264407e+00 6.24121800e-02 -4.02177542e-01 -1.63972348e-01
4.21219140e-01 1.09656906e+00 -2.19762266e-01 4.02884990e-01
8.38019788e-01 4.06121939e-01 -1.83104619e-01 1.53726673e+00
-3.21657285e-02 3.68985325e-01 -5.29967725e-01 -3.77583615e-02
-2.09621228e-02 -5.41095614e-01 6.66520774e-01 1.33221257e+00
5.60221195e-01 3.60336185e-01 1.11398019e-01 5.61974764e-01
7.92530105e-02 2.74358213e-01 -3.09196144e-01 3.37058276e-01
5.13911843e-01 8.92693400e-01 -8.96912158e-01 -4.85822588e-01
-7.49846697e-01 8.41335058e-01 -1.57513440e-01 2.28784382e-01
-6.33406401e-01 -1.22477365e+00 5.85329592e-01 3.78273815e-01
5.81195593e-01 -4.77344096e-01 -5.09299040e-01 -9.18625832e-01
2.81930208e-01 -6.91334486e-01 3.09503824e-01 -2.55728960e-01
-9.75948393e-01 6.35507226e-01 -1.37955651e-01 -1.33132732e+00
4.86530483e-01 -4.38374162e-01 -9.03689384e-01 9.01158333e-01
-1.54125977e+00 -4.03776288e-01 -3.86586577e-01 6.87983990e-01
6.61797404e-01 4.09754783e-01 4.51808602e-01 4.11120951e-01
-3.46090436e-01 4.43134636e-01 2.78846443e-01 8.91878977e-02
9.04516697e-01 -9.58009303e-01 6.48288488e-01 1.17949343e+00
-2.84532160e-01 8.50663662e-01 1.03483927e+00 -6.33159161e-01
-1.18050981e+00 -8.16899419e-01 7.88008630e-01 5.57361424e-01
5.12270391e-01 -4.77154136e-01 -1.17345536e+00 6.32537842e-01
1.91423327e-01 3.08472037e-01 4.20556456e-01 -4.04605091e-01
-5.44377565e-02 -3.89054418e-01 -1.32677329e+00 8.23491216e-01
8.82142425e-01 -1.72447592e-01 -5.75906634e-01 -1.13721356e-01
5.03183544e-01 -6.11010432e-01 -6.97065473e-01 3.37128282e-01
7.02309906e-01 -1.18680871e+00 9.90331054e-01 -1.19483113e-01
-1.24123320e-01 -6.42255247e-01 1.75026685e-01 -1.27510917e+00
-4.62834418e-01 -1.09293258e+00 -2.35557072e-02 9.67740119e-01
1.48178369e-01 -1.03049767e+00 7.46604264e-01 3.24848950e-01
-1.16423428e-01 -7.61507452e-01 -1.14665818e+00 -1.07160258e+00
-1.08078539e-01 -1.41915783e-01 3.34606290e-01 4.48194683e-01
1.21488556e-01 -1.30317718e-01 -4.73814428e-01 1.83781251e-01
8.34798276e-01 3.51431742e-02 6.17635131e-01 -1.23684680e+00
-3.01882267e-01 -2.73359060e-01 -4.10327345e-01 -1.27569759e+00
-4.51708704e-01 -2.97775805e-01 1.27434209e-01 -1.26526713e+00
-1.67064264e-01 -1.09334171e-01 -1.42109450e-02 6.30542636e-02
-4.42418635e-01 2.91588634e-01 2.46146560e-01 -1.60088435e-01
1.03546053e-01 3.98756742e-01 1.35705972e+00 1.07482478e-01
-5.52656889e-01 1.25853136e-01 -5.22723138e-01 1.04873288e+00
6.97454751e-01 -4.79786098e-01 -1.91483691e-01 -1.97674543e-01
-1.44798443e-01 -9.26732570e-02 1.19158283e-01 -1.16947746e+00
3.83419514e-01 9.21745300e-02 2.54949480e-01 -7.26591110e-01
3.93376142e-01 -1.03937829e+00 1.13713905e-01 8.89457643e-01
6.34030700e-02 4.02216613e-03 3.17178369e-01 6.13330543e-01
-3.15854549e-01 -3.49140584e-01 1.36133659e+00 -2.49662995e-02
-4.55420256e-01 1.41318068e-01 -2.15010136e-01 1.22193377e-02
1.17885554e+00 -5.02767086e-01 -3.59523803e-01 -3.16593647e-01
-6.09868348e-01 -6.33445196e-03 3.75503421e-01 -1.92574263e-01
6.87828183e-01 -1.23542762e+00 -6.06387734e-01 5.51684678e-01
-3.76468658e-01 -6.50881603e-02 4.80337262e-01 1.15343058e+00
-8.01871598e-01 1.30135417e-01 -1.38710782e-01 -2.77986139e-01
-1.68067122e+00 7.09221721e-01 1.90977484e-01 -1.08265907e-01
-1.25996768e+00 7.90854871e-01 3.69864017e-01 4.20252681e-01
2.34875262e-01 -7.30727792e-01 1.36203781e-01 -1.67206198e-01
1.05835462e+00 8.17149639e-01 -1.30505130e-01 -1.34699509e-01
-3.62915456e-01 1.00763500e+00 1.43605605e-01 -1.36557877e-01
1.24957633e+00 -2.47236174e-02 -1.65240094e-01 3.61644894e-01
1.14859235e+00 5.57321191e-01 -1.45926166e+00 -1.80761695e-01
-3.56876887e-02 -7.52295196e-01 1.10705115e-01 9.83198434e-02
-9.65724289e-01 8.93171430e-01 6.46582425e-01 3.81045878e-01
1.52244580e+00 -5.08465230e-01 1.03502178e+00 -1.62271202e-01
1.36095449e-01 -1.12275279e+00 -2.15499371e-01 9.31547955e-02
8.45015287e-01 -7.81168222e-01 3.56800050e-01 -1.15781200e+00
-2.27322698e-01 1.20347238e+00 7.79411420e-02 -8.31781805e-01
1.11884379e+00 5.00010610e-01 -4.80754822e-02 1.85090259e-01
-3.95368606e-01 -1.33905500e-01 2.60667413e-01 4.51521158e-01
5.00535190e-01 -2.03485359e-02 -9.96648014e-01 6.52397513e-01
1.04183428e-01 2.81902641e-01 4.73608494e-01 1.10918880e+00
-1.00687122e+00 -9.00040090e-01 -6.49716854e-01 2.51268744e-01
-6.56055808e-01 -9.28715914e-02 1.03069291e-01 7.32427776e-01
2.31588557e-02 1.02150524e+00 -5.25008999e-02 4.46837395e-02
8.14213634e-01 -3.06783477e-03 2.42339313e-01 -2.11434707e-01
-6.86081290e-01 5.13556480e-01 -2.66339332e-01 -6.51438236e-01
-2.37647876e-01 -5.08806050e-01 -1.36557329e+00 -3.05105031e-01
-4.52271104e-01 5.36797717e-02 5.09992838e-01 5.00282407e-01
1.73888937e-01 6.27174616e-01 4.62095290e-01 -6.32479787e-01
-6.38400793e-01 -9.04879391e-01 -1.01803362e+00 3.14094216e-01
6.77790523e-01 -5.23792267e-01 -4.85870153e-01 3.18743795e-01] | [11.068683624267578, -2.533808708190918] |
46a00611-851a-4535-9a82-518a3a2fc2c4 | cnn-based-synthesis-of-realistic-high | 1907.00787 | null | https://arxiv.org/abs/1907.00787v2 | https://arxiv.org/pdf/1907.00787v2.pdf | CNN-based synthesis of realistic high-resolution LiDAR data | This paper presents a novel CNN-based approach for synthesizing high-resolution LiDAR point cloud data. Our approach generates semantically and perceptually realistic results with guidance from specialized loss-functions. First, we utilize a modified per-point loss that addresses missing LiDAR point measurements. Second, we align the quality of our generated output with real-world sensor data by applying a perceptual loss. In large-scale experiments on real-world datasets, we evaluate both the geometric accuracy and semantic segmentation performance using our generated data vs. ground truth. In a mean opinion score testing we further assess the perceptual quality of our generated point clouds. Our results demonstrate a significant quantitative and qualitative improvement in both geometry and semantics over traditional non CNN-based up-sampling methods. | ['J. Marius Zöllner', 'Christoph B. Rist', 'Larissa T. Triess', 'David Peter', 'Markus Enzweiler'] | 2019-06-28 | null | null | null | null | ['depth-image-upsampling', 'point-set-upsampling', 'point-cloud-generation'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 3.57552677e-01 1.39609352e-01 9.29929838e-02 -6.61098480e-01
-1.40680373e+00 -6.29857242e-01 3.75742465e-01 5.10504901e-01
-3.37510318e-01 6.49180830e-01 -3.01369041e-01 -1.14204206e-01
1.23392045e-01 -1.27801347e+00 -1.22906685e+00 6.55671209e-02
-6.82423711e-02 8.79688978e-01 5.36112905e-01 -1.41712308e-01
2.69225061e-01 1.11850929e+00 -1.78823674e+00 7.90710673e-02
1.17752254e+00 1.30748141e+00 1.35104820e-01 5.30510306e-01
-3.23125839e-01 -2.46620476e-01 -6.17126584e-01 -4.59145516e-01
7.03161478e-01 3.11319858e-01 -5.41604996e-01 1.42572716e-01
1.28080738e+00 -3.21926296e-01 2.91526884e-01 1.14390016e+00
5.04444480e-01 3.54853235e-02 4.03926849e-01 -1.30916047e+00
-2.39191070e-01 1.05516665e-01 -5.10594904e-01 -3.50738049e-01
3.57874095e-01 4.06746149e-01 9.24889088e-01 -9.59028125e-01
5.61001718e-01 1.48410010e+00 1.03210390e+00 1.02947652e-01
-1.56429183e+00 -8.12969804e-01 -2.81750541e-02 -2.83358753e-01
-1.54451990e+00 -1.01668119e-01 8.80691946e-01 -4.16275650e-01
7.42651641e-01 1.27513364e-01 6.92147076e-01 6.02243185e-01
2.37660538e-02 3.14403743e-01 9.95913863e-01 -7.30170608e-02
4.66534942e-01 -9.63655263e-02 -3.93993199e-01 5.18795788e-01
3.99797976e-01 5.00431061e-01 -2.53940821e-01 -1.42771140e-01
9.90358293e-01 -3.44234020e-01 2.68590394e-02 -6.15883231e-01
-1.05777812e+00 7.90397286e-01 1.03684866e+00 -2.70501286e-01
-3.53872836e-01 5.63406110e-01 -1.36119118e-02 2.11307425e-02
7.49142051e-01 4.01088357e-01 -3.06804031e-01 1.27042919e-01
-1.33184087e+00 4.67452288e-01 3.93737853e-01 1.24122727e+00
1.09444845e+00 1.37532189e-01 2.31250837e-01 6.43194437e-01
2.95657724e-01 9.90282357e-01 -3.06673378e-01 -1.56626129e+00
4.43380088e-01 3.98418814e-01 2.93348491e-01 -1.05104327e+00
-1.23003699e-01 -4.49487358e-01 -3.79228354e-01 8.85719657e-01
1.11854643e-01 1.25141352e-01 -1.23973489e+00 1.25759089e+00
2.87367463e-01 1.49328783e-01 -1.40796393e-01 7.89292216e-01
7.15951681e-01 3.32722843e-01 2.31417328e-01 4.85536069e-01
8.85991812e-01 -5.13046384e-01 -1.56159565e-01 -9.37173069e-02
3.60711180e-02 -8.08925629e-01 1.28309822e+00 4.92334038e-01
-1.31418192e+00 -7.59214818e-01 -1.24062181e+00 -1.55189142e-01
-3.66979420e-01 -3.38112749e-02 3.40413988e-01 5.69396496e-01
-1.23936498e+00 1.09615922e+00 -7.24636018e-01 -2.29145736e-01
7.72371233e-01 3.22840244e-01 -1.41617984e-01 -9.35948342e-02
-6.27921283e-01 8.00024748e-01 3.28406185e-01 -2.65088141e-01
-7.00428367e-01 -9.98614192e-01 -9.85917270e-01 -4.90824655e-02
1.47268474e-01 -7.37657487e-01 1.39377153e+00 -5.21559060e-01
-1.16721106e+00 9.33070183e-01 -1.05396710e-01 -5.10162115e-01
8.70652735e-01 -1.68567792e-01 -4.33110669e-02 2.65838265e-01
3.87701541e-01 1.47948813e+00 5.43821216e-01 -2.12738776e+00
-8.35368931e-01 -3.84572327e-01 6.92287311e-02 1.53071627e-01
4.97725040e-01 -5.79851091e-01 -2.28683189e-01 -5.40468514e-01
3.09904605e-01 -5.81588924e-01 -5.35576463e-01 7.21535683e-01
-3.30329686e-01 1.15710102e-01 9.38662171e-01 -3.08134794e-01
1.21796221e-01 -2.11847973e+00 -5.99549234e-01 4.97510076e-01
9.42873657e-02 -4.31329161e-02 -2.24527821e-01 1.56169415e-01
1.51680604e-01 4.17720884e-01 -5.12883186e-01 -6.23169005e-01
-7.94024915e-02 3.17240089e-01 -5.75354517e-01 2.04097390e-01
5.37588358e-01 9.28210855e-01 -1.02043498e+00 -6.59440935e-01
8.34063470e-01 7.59271741e-01 -6.18002295e-01 -6.15335535e-03
-7.07590699e-01 5.02207577e-01 -2.22818241e-01 8.60379934e-01
1.08087313e+00 -5.97516559e-02 -4.30021733e-01 -3.64757746e-01
-5.13578616e-02 1.68436244e-01 -1.20699382e+00 1.98091316e+00
-7.23389804e-01 6.02653861e-01 8.98145586e-02 -1.28579199e-01
1.05300617e+00 -1.13337331e-01 5.95956743e-01 -7.26580143e-01
-8.45677312e-03 3.33139449e-01 -4.50405508e-01 2.71543533e-01
9.85625386e-01 -3.14289659e-01 1.54209763e-01 -1.12437546e-01
-1.58801749e-01 -1.26434076e+00 -2.99290895e-01 -3.98505256e-02
6.86562359e-01 3.30436647e-01 -1.81625530e-01 -2.23056570e-01
-5.57160154e-02 6.81813598e-01 1.90165043e-01 5.74016035e-01
-5.45134535e-03 1.30966783e+00 1.62838921e-02 -1.61387667e-01
-1.42771137e+00 -1.65739250e+00 -4.22102302e-01 3.44416499e-01
4.54252779e-01 -6.92024902e-02 -7.26825714e-01 -3.38307709e-01
4.56395030e-01 1.00775385e+00 -2.44932845e-01 2.40782320e-01
-5.35141706e-01 -9.54986513e-02 5.40347278e-01 8.56064141e-01
5.59855103e-01 -9.37384009e-01 -8.00805926e-01 1.30200669e-01
1.18495978e-01 -1.51843524e+00 -8.07801560e-02 8.13773461e-03
-1.24179721e+00 -9.62362707e-01 -2.23235279e-01 -4.34913993e-01
5.32993853e-01 1.72824100e-01 1.54827905e+00 6.80832639e-02
-1.74579442e-01 3.28454942e-01 -1.59510866e-01 -6.03379428e-01
-2.52945364e-01 -6.20480925e-02 -2.69959748e-01 -6.80973828e-01
4.36277762e-02 -7.75810778e-01 -5.91870368e-01 1.34002730e-01
-8.35171044e-01 -1.14441738e-01 4.98005480e-01 9.62352604e-02
1.10400760e+00 -4.34910841e-02 7.28604719e-02 -4.20464158e-01
5.24178326e-01 9.71647259e-03 -9.08373773e-01 -3.39000702e-01
-6.62004292e-01 -1.25129014e-01 3.91906887e-01 -5.35385311e-02
-6.26577139e-01 2.62324601e-01 -3.71216357e-01 -8.19107473e-01
-4.62821662e-01 9.44712162e-02 -1.64832786e-01 -2.90325046e-01
6.89095795e-01 -2.02437446e-01 -2.62147725e-01 -3.56542408e-01
7.40835965e-01 4.17786360e-01 8.30921113e-01 -9.58116472e-01
1.09606969e+00 9.74679887e-01 1.91823676e-01 -7.92659819e-01
-5.99808097e-01 -2.85586476e-01 -7.24521577e-01 -1.49865761e-01
8.62047195e-01 -9.13639724e-01 -6.15424275e-01 9.02397931e-02
-1.48255587e+00 -3.09177488e-01 -9.01719451e-01 2.77403027e-01
-8.15035522e-01 2.31353149e-01 -1.71709761e-01 -7.66225278e-01
-2.39517763e-01 -1.24932027e+00 1.97540259e+00 -6.51503056e-02
-3.83599773e-02 -7.40741968e-01 -8.42578057e-03 1.49026081e-01
1.76418737e-01 6.81731701e-01 6.32580459e-01 -1.21968642e-01
-1.12272537e+00 -6.33155778e-02 -7.12007523e-01 4.97627109e-01
-9.49562788e-02 2.77493149e-01 -1.16401303e+00 -1.42964974e-01
-5.20480514e-01 -2.35798821e-01 5.56568563e-01 3.97817135e-01
1.32685637e+00 2.19650701e-01 -1.96180835e-01 7.86252737e-01
1.79153013e+00 -4.74113040e-02 6.77016377e-01 7.81957060e-02
8.68921459e-01 5.59951246e-01 8.39029193e-01 4.06878032e-02
2.95298785e-01 6.11390114e-01 9.27299678e-01 -4.54747289e-01
-2.73083717e-01 -7.04774141e-01 -2.10046887e-01 2.32038632e-01
6.83771744e-02 -2.36060143e-01 -9.94208753e-01 6.56016588e-01
-1.53970206e+00 -5.67439616e-01 -3.04811090e-01 2.31946802e+00
5.51123857e-01 4.37719941e-01 1.61726272e-03 2.13899650e-02
5.84353805e-01 3.45714465e-02 -5.31328380e-01 -4.00504768e-01
-1.47878245e-01 8.28945339e-01 1.03683412e+00 7.99509585e-01
-8.83349001e-01 1.14105427e+00 6.95214939e+00 8.62622797e-01
-1.18875957e+00 7.16182739e-02 3.32980484e-01 3.04650627e-02
-6.08224034e-01 -6.80791959e-02 -4.98391926e-01 1.83716118e-01
7.25490928e-01 -8.81385803e-02 -5.09847999e-02 1.09126818e+00
3.51067126e-01 -1.54411435e-01 -1.10053623e+00 9.06482697e-01
-2.39758879e-01 -1.59094286e+00 2.81084388e-01 3.60884160e-01
6.95467114e-01 3.55249196e-01 1.16132684e-01 -2.29188383e-01
8.05406451e-01 -1.25642765e+00 1.17209268e+00 4.05347198e-01
1.28916895e+00 -8.23956966e-01 5.39758563e-01 1.40157565e-01
-1.40138435e+00 4.87546206e-01 -3.39175433e-01 1.49788961e-01
3.46531898e-01 9.03128207e-01 -9.94352639e-01 6.70574486e-01
7.56964505e-01 3.92093837e-01 -6.41791642e-01 1.18744338e+00
-1.55679822e-01 1.81882441e-01 -7.08166361e-01 4.21139181e-01
2.08863094e-01 -1.67352393e-01 5.27651548e-01 1.13554859e+00
2.91537315e-01 -2.23743469e-01 3.53436530e-01 1.45554841e+00
-1.47508293e-01 -7.63171837e-02 -8.67800653e-01 3.79273266e-01
7.38206863e-01 1.08835769e+00 -8.34138811e-01 -4.09953892e-01
1.37416571e-01 6.73847377e-01 2.56495252e-02 2.71089166e-01
-6.54614508e-01 -2.89646357e-01 8.30072701e-01 5.19355714e-01
3.32043260e-01 -5.21523535e-01 -1.07901227e+00 -7.20111847e-01
2.38809496e-01 -2.70786673e-01 -2.48266816e-01 -1.27315497e+00
-1.34230351e+00 6.27363920e-01 2.33539224e-01 -1.41484499e+00
-1.20525926e-01 -4.04428750e-01 -3.97743136e-01 1.04695499e+00
-1.65084231e+00 -1.34808445e+00 -7.05013394e-01 1.69821650e-01
5.37455559e-01 4.56423163e-01 4.87341881e-01 2.16171622e-01
2.51373500e-01 2.37153620e-01 -3.41550171e-01 -5.97472228e-02
2.74665862e-01 -1.45281875e+00 8.97821307e-01 7.00490177e-01
-2.88595613e-02 4.02689844e-01 7.81040192e-01 -8.56691301e-01
-5.87403357e-01 -1.50489211e+00 3.11245710e-01 -5.29858232e-01
1.81436852e-01 -2.17568561e-01 -8.50356340e-01 5.28845906e-01
-1.52638823e-01 1.05355941e-02 2.09552571e-02 -3.58285517e-01
-3.11109215e-01 -2.23232098e-02 -1.90011418e+00 4.24058229e-01
1.21050942e+00 -4.23226714e-01 -5.21332085e-01 8.87616873e-02
1.15621722e+00 -5.75419188e-01 -9.31686282e-01 9.60267246e-01
3.39639008e-01 -1.16257083e+00 1.31340325e+00 -1.47980303e-01
7.37120867e-01 -5.50831378e-01 -4.48878169e-01 -1.38297665e+00
6.77459165e-02 -6.31100535e-02 4.27683145e-01 7.66191065e-01
2.69838452e-01 -4.17219192e-01 1.12233067e+00 4.55432802e-01
-3.95376861e-01 -6.08837068e-01 -9.38329101e-01 -9.27123845e-01
3.50462645e-01 -1.02429581e+00 9.70702112e-01 6.72100782e-01
-8.47322345e-01 -4.68730293e-02 4.87927109e-01 4.11966711e-01
1.10877430e+00 2.03088701e-01 9.10536587e-01 -1.40192223e+00
1.90643772e-01 -4.81328130e-01 -7.36784458e-01 -9.37554538e-01
-2.86756884e-02 -6.60511434e-01 2.84701228e-01 -1.88817132e+00
-4.88201618e-01 -6.69012487e-01 1.60961479e-01 1.36737481e-01
3.27106804e-01 6.99970782e-01 2.98448443e-01 1.06379449e-01
-1.76770926e-01 5.63605428e-01 1.20157874e+00 -8.65645111e-02
-1.15862988e-01 -9.22853202e-02 -3.67323220e-01 8.12531412e-01
8.29010069e-01 -4.65493828e-01 -3.67878258e-01 -6.93838835e-01
1.25279605e-01 -2.32258961e-01 8.96120667e-01 -1.47313094e+00
-1.90618351e-01 -2.57423162e-01 4.47648704e-01 -1.19812202e+00
7.45863140e-01 -1.01943648e+00 1.26303464e-01 3.51356000e-01
-3.53245661e-02 1.70728728e-01 3.94258320e-01 3.71489614e-01
5.14573594e-06 1.40550137e-01 1.07450616e+00 -1.36028051e-01
-5.40363848e-01 3.58392984e-01 3.68962437e-01 -5.35981767e-02
8.50517809e-01 -7.31451869e-01 -1.52948707e-01 -4.88437116e-01
-6.58752739e-01 2.89157033e-01 1.08902884e+00 3.14211577e-01
1.10992873e+00 -1.31305563e+00 -6.17783427e-01 1.58508092e-01
3.33736390e-01 7.74155796e-01 -2.74293423e-01 1.24628298e-01
-1.02235925e+00 1.45491689e-01 -1.50123283e-01 -1.10588408e+00
-9.15022492e-01 1.95545822e-01 4.89928901e-01 2.41672590e-01
-7.21570492e-01 6.34586275e-01 -1.59566611e-01 -8.29209447e-01
2.54058242e-01 -9.05757368e-01 4.86536503e-01 -4.04167891e-01
-1.09543487e-01 2.71785200e-01 3.47080439e-01 -6.06500328e-01
-3.41289759e-01 8.43904018e-01 6.31372392e-01 -6.77400112e-01
1.11075401e+00 1.56495959e-01 2.31196344e-01 4.21108007e-01
1.11339307e+00 2.61829466e-01 -1.43765926e+00 4.52219062e-02
-2.20997483e-01 -8.35131228e-01 1.66158751e-01 -7.55604923e-01
-1.09769213e+00 1.01832950e+00 7.86691427e-01 6.31756037e-02
5.60298502e-01 2.34968551e-02 1.03817654e+00 1.68311819e-01
7.62172043e-01 -9.21797812e-01 -1.73601165e-01 3.13738793e-01
9.03548717e-01 -1.34252024e+00 2.34052483e-02 -8.76664937e-01
-3.14959407e-01 7.79179931e-01 7.09017396e-01 -5.61996341e-01
6.41387939e-01 4.33345765e-01 2.64232725e-01 -4.43069458e-01
-1.01140141e-01 -3.73850524e-01 3.30640525e-01 1.05323911e+00
6.54314235e-02 1.82756051e-01 9.93007571e-02 -9.31058750e-02
-9.83566463e-01 3.04342099e-02 3.51934552e-01 8.42926204e-01
-6.72787070e-01 -9.00806427e-01 -5.60901761e-01 3.20946962e-01
1.79243714e-01 1.54882455e-02 -4.50457633e-01 7.99605906e-01
2.99963504e-01 8.45526636e-01 6.04673624e-01 -4.86562163e-01
7.76446939e-01 -4.59641069e-01 5.15472651e-01 -8.28128636e-01
-2.95248300e-01 -8.65683407e-02 -7.41229625e-03 -9.00523007e-01
-2.92566776e-01 -3.63168418e-01 -1.55543435e+00 -4.97138113e-01
-9.11760628e-02 -8.90368298e-02 1.19249594e+00 4.80984122e-01
4.73983526e-01 5.23757935e-01 4.70294923e-01 -1.38443029e+00
-3.30231786e-01 -5.89864135e-01 -2.57664114e-01 4.66730595e-01
2.43923873e-01 -6.39021933e-01 -2.97905594e-01 -1.82327390e-01] | [8.332454681396484, -3.0319571495056152] |
b020dfe4-b6d6-475a-a000-ccfc78f3704f | a-dependable-hybrid-machine-learning-model | 2212.04546 | null | https://arxiv.org/abs/2212.04546v2 | https://arxiv.org/pdf/2212.04546v2.pdf | A Dependable Hybrid Machine Learning Model for Network Intrusion Detection | Network intrusion detection systems (NIDSs) play an important role in computer network security. There are several detection mechanisms where anomaly-based automated detection outperforms others significantly. Amid the sophistication and growing number of attacks, dealing with large amounts of data is a recognized issue in the development of anomaly-based NIDS. However, do current models meet the needs of today's networks in terms of required accuracy and dependability? In this research, we propose a new hybrid model that combines machine learning and deep learning to increase detection rates while securing dependability. Our proposed method ensures efficient pre-processing by combining SMOTE for data balancing and XGBoost for feature selection. We compared our developed method to various machine learning and deep learning algorithms to find a more efficient algorithm to implement in the pipeline. Furthermore, we chose the most effective model for network intrusion based on a set of benchmarked performance analysis criteria. Our method produces excellent results when tested on two datasets, KDDCUP'99 and CIC-MalMem-2022, with an accuracy of 99.99% and 100% for KDDCUP'99 and CIC-MalMem-2022, respectively, and no overfitting or Type-1 and Type-2 issues. | ['Mohammad Abu Yousuf', 'Mohammad Ali Moni', 'Fares Alharbi', 'Arnisha Akhter', 'Md Ashraf Uddin', 'Md. Manowarul Islam', 'Khondokar Fida Hasan', 'Md. Alamin Talukder'] | 2022-12-08 | null | null | null | null | ['network-intrusion-detection'] | ['miscellaneous'] | [-2.57846296e-01 -6.22017145e-01 1.48558587e-01 -4.55116123e-01
1.31723642e-01 -2.93128580e-01 5.58451116e-01 5.42845607e-01
-5.44390678e-01 5.67719042e-01 -6.68998361e-01 -6.45015419e-01
-6.35548890e-01 -1.00267684e+00 -8.59165117e-02 -4.44553584e-01
-2.88131833e-01 7.97903597e-01 4.20975447e-01 -2.67873138e-01
6.23102903e-01 1.12952542e+00 -1.55029023e+00 2.38284111e-01
6.34567678e-01 1.24201214e+00 -6.63475990e-01 7.26063132e-01
-3.63776356e-01 5.60513854e-01 -7.81040370e-01 -3.75791281e-01
4.82819527e-01 -2.55284727e-01 -6.09344602e-01 -4.83390987e-01
2.67351307e-02 -1.85783982e-01 -7.91829918e-03 8.44319642e-01
5.96309602e-01 -1.71706788e-02 5.88544130e-01 -1.65276098e+00
-2.34473720e-02 2.29988933e-01 -6.78494275e-01 7.15010822e-01
2.70223804e-02 3.03715497e-01 5.43849409e-01 -4.71719176e-01
3.49495083e-01 1.21296060e+00 7.24501133e-01 5.27505219e-01
-1.11637866e+00 -9.69526052e-01 -1.74405053e-01 5.90047896e-01
-1.23237193e+00 -3.17480445e-01 7.01505840e-01 -3.00819784e-01
1.13271797e+00 4.41770256e-01 4.22894150e-01 9.74202275e-01
3.38478386e-01 2.47747779e-01 9.33306038e-01 -4.44552511e-01
6.44999564e-01 4.01585788e-01 4.76753861e-01 4.86218423e-01
6.71419084e-01 3.11160147e-01 -1.64262086e-01 -4.40527409e-01
3.49034280e-01 2.93692350e-01 4.45435166e-01 2.40784883e-02
-4.73170280e-01 9.80433226e-01 1.10638827e-01 6.70912325e-01
-6.49317265e-01 -4.48578686e-01 7.67219603e-01 6.19865537e-01
2.57706404e-01 6.56789005e-01 -6.13913000e-01 -1.05550669e-01
-9.59937871e-01 1.02821089e-01 9.37074482e-01 1.97276816e-01
4.87065494e-01 4.17798519e-01 2.85106361e-01 7.09692955e-01
4.23856378e-01 1.72396228e-01 5.81366122e-01 -1.88825831e-01
-1.18857928e-01 1.07237875e+00 -2.14936063e-01 -1.20280600e+00
-7.63446093e-01 -6.84559882e-01 -9.81882155e-01 5.71076632e-01
3.59820396e-01 -1.90363452e-01 -1.01904058e+00 1.30810595e+00
3.72155100e-01 2.45717451e-01 -3.84015851e-02 4.43775743e-01
3.12860936e-01 5.07009089e-01 3.92598867e-01 -2.02159747e-01
9.95918870e-01 -2.31233805e-01 -5.20915329e-01 1.22665130e-02
5.82925022e-01 -6.30963385e-01 5.53311467e-01 5.81068695e-01
-5.12873769e-01 -3.65675986e-01 -1.30900693e+00 7.15601087e-01
-9.06624615e-01 -4.09723550e-01 6.99646354e-01 1.17906690e+00
-7.50247777e-01 7.00765312e-01 -7.65358090e-01 -7.19619334e-01
5.54648638e-01 7.13154376e-01 -4.06475127e-01 2.67562985e-01
-1.20972705e+00 9.40541685e-01 5.91176569e-01 -1.28768250e-01
-5.87122560e-01 -4.73924339e-01 -2.80678958e-01 3.65259754e-03
1.66993648e-01 -1.95016399e-01 6.76402926e-01 -7.65459418e-01
-1.25293314e+00 7.34111786e-01 3.18513185e-01 -7.75152326e-01
4.37116534e-01 3.42001766e-02 -1.00533760e+00 -5.50381746e-03
-2.63592780e-01 8.96959528e-02 6.85235441e-01 -9.74286377e-01
-8.78010035e-01 -6.40691280e-01 -1.01004444e-01 -3.68936121e-01
-6.05046391e-01 3.43478858e-01 1.96565419e-01 -2.19781250e-01
1.36063740e-01 -5.17544210e-01 -2.28275657e-01 -3.73884141e-01
-3.85346949e-01 -2.12081939e-01 1.30789876e+00 -4.12304640e-01
1.27603185e+00 -1.89589179e+00 -5.04113138e-01 8.41799915e-01
1.57012537e-01 1.09963977e+00 1.04597308e-01 4.78740603e-01
-3.21785241e-01 2.75720417e-01 -2.52303988e-01 5.98771982e-02
-9.17278901e-02 3.17854472e-02 -7.24953488e-02 3.22991908e-01
3.74058872e-01 2.55094856e-01 -5.97311258e-01 -2.33031243e-01
5.43054104e-01 2.71792859e-01 -4.75535363e-01 3.84451210e-01
-6.30678609e-02 1.81491137e-01 -4.37472671e-01 1.08891451e+00
7.24457622e-01 7.21259564e-02 1.94735602e-02 1.77625306e-02
-4.72578332e-02 -2.15325490e-01 -1.38499522e+00 8.30323875e-01
-1.89392447e-01 3.19739729e-01 -1.09081574e-01 -1.40449047e+00
1.32584488e+00 1.81027874e-01 7.02608764e-01 -7.80969799e-01
4.55062807e-01 4.51145977e-01 5.97137570e-01 -5.34300685e-01
-8.10690448e-02 -1.44622205e-02 2.70322293e-01 6.20250404e-01
1.55623809e-01 4.49511766e-01 3.69068533e-01 -1.83214918e-02
1.42586565e+00 -5.33184290e-01 4.31758434e-01 -2.29438052e-01
8.04696918e-01 1.43663540e-01 5.51152945e-01 8.44584286e-01
-6.01304293e-01 -1.12979440e-03 5.87796390e-01 -9.40577447e-01
-8.89655232e-01 -6.96774364e-01 -3.70142668e-01 9.57255065e-01
-3.31948906e-01 -8.04416165e-02 -5.97181380e-01 -8.02591443e-01
-7.21629895e-03 7.52130628e-01 -3.83939654e-01 -4.37365204e-01
-3.94171983e-01 -1.34850562e+00 8.66560400e-01 2.30805855e-02
6.70615077e-01 -1.26149333e+00 -5.58332503e-01 3.34388673e-01
5.73193252e-01 -7.24645913e-01 6.29198432e-01 3.77394229e-01
-7.18400955e-01 -1.60512006e+00 8.35322887e-02 -3.00175458e-01
3.52408111e-01 -1.70373291e-01 8.96015763e-01 4.31004256e-01
-7.78503656e-01 -1.30917072e-01 -4.73435670e-01 -7.92071581e-01
-5.04227459e-01 2.03734621e-01 3.68605852e-01 1.30031437e-01
9.74041402e-01 -6.74375951e-01 -4.14058924e-01 2.64609098e-01
-1.05734897e+00 -6.55073345e-01 7.30984688e-01 5.33674777e-01
2.33529508e-02 4.87811178e-01 9.69821095e-01 -1.08428347e+00
7.92176902e-01 -9.60698903e-01 -6.22020602e-01 -3.55392396e-02
-1.05958664e+00 -1.46752894e-01 8.23551238e-01 -3.07587922e-01
-6.34198844e-01 -2.59644419e-01 -5.75504661e-01 -2.39107594e-01
-6.44880176e-01 3.72706711e-01 -1.91142932e-01 -1.73726454e-01
9.50092852e-01 4.95308898e-02 1.87611744e-01 -4.43322062e-01
-3.43822718e-01 8.88804257e-01 1.36791512e-01 -3.34510356e-01
8.34527016e-01 3.31408203e-01 2.07875401e-01 -9.21247244e-01
-3.38114500e-01 -4.68217909e-01 -5.36706328e-01 -1.63488060e-01
4.31012988e-01 -1.47415444e-01 -8.68891358e-01 8.16977918e-01
-8.62326741e-01 2.87490606e-01 1.02732733e-01 2.66073585e-01
2.34878272e-01 1.80173859e-01 -2.97652215e-01 -1.09871161e+00
-7.51275182e-01 -1.04030180e+00 1.10938512e-01 3.93698573e-01
-2.98344910e-01 -8.93007338e-01 1.55545786e-01 -1.83876995e-02
1.00830126e+00 5.20559192e-01 9.49806511e-01 -1.83366203e+00
4.40565087e-02 -9.41842258e-01 -3.44722301e-01 4.82371449e-01
7.72988945e-02 4.52900797e-01 -9.28171217e-01 -2.06977740e-01
7.15002790e-02 -9.66520421e-03 4.56132889e-01 6.28656894e-02
1.13762474e+00 -1.47394404e-01 -1.52325541e-01 5.29839277e-01
1.47735989e+00 1.03903794e+00 6.17583036e-01 9.82102215e-01
3.86445552e-01 6.43741488e-01 3.33415776e-01 5.43189347e-01
-1.90059125e-01 3.31669271e-01 8.36192608e-01 4.43526171e-02
4.20470804e-01 3.25652480e-01 1.62980035e-02 2.20653981e-01
1.74134120e-01 -2.05562055e-01 -1.28152430e+00 1.84940964e-01
-1.58444083e+00 -1.02152812e+00 -3.10446173e-01 2.12885499e+00
3.24576706e-01 6.90669060e-01 4.62654322e-01 7.50258982e-01
7.02456295e-01 -1.71811759e-01 -7.23223805e-01 -9.37162399e-01
7.90259540e-02 6.22186244e-01 2.44885296e-01 1.24196872e-01
-1.23291314e+00 6.36433780e-01 5.54974747e+00 5.67226529e-01
-1.44350588e+00 -2.20091447e-01 7.17667162e-01 -8.42815563e-02
4.45867300e-01 -2.22018078e-01 -7.54821062e-01 6.56276405e-01
1.47203207e+00 -4.41892184e-02 6.49385229e-02 9.04037118e-01
1.88908204e-01 3.83836105e-02 -7.18875945e-01 6.74577773e-01
-7.92728439e-02 -1.21933568e+00 -1.30054746e-02 1.98104754e-02
2.70245939e-01 -9.67196375e-02 -1.31829277e-01 4.27616090e-01
2.19434395e-01 -9.24211860e-01 -1.60190370e-02 3.59929532e-01
1.97064146e-01 -1.09655404e+00 1.19586957e+00 1.74376562e-01
-5.76933742e-01 -3.52179915e-01 -2.21558034e-01 -9.44154188e-02
-2.76964188e-01 7.19413698e-01 -8.92551124e-01 6.11702979e-01
7.13670492e-01 3.30475092e-01 -6.41048729e-01 1.28762102e+00
2.61679322e-01 6.39429927e-01 -5.43873489e-01 -1.70934200e-01
3.46966684e-01 3.95318829e-02 5.49827337e-01 1.13680804e+00
1.06197663e-01 -2.02291921e-01 3.38139720e-02 4.38032240e-01
3.18883628e-01 3.04398626e-01 -4.51931745e-01 -1.53567702e-01
6.50529385e-01 1.37026334e+00 -8.67466807e-01 -2.50926614e-01
-1.45953476e-01 4.17047739e-01 -7.34958351e-02 -6.01649433e-02
-6.62528574e-01 -6.34416997e-01 7.75860727e-01 2.96527445e-01
-3.14026356e-01 1.02702148e-01 -4.89046723e-01 -7.63022780e-01
-3.94008428e-01 -1.13245916e+00 9.00222242e-01 8.62203762e-02
-1.42770660e+00 9.92854357e-01 -4.52101743e-03 -1.02464092e+00
-2.71964550e-01 -9.50244904e-01 -1.01184702e+00 7.27327764e-01
-1.11962068e+00 -1.02118862e+00 -2.34645620e-01 5.65991759e-01
2.68093795e-01 -8.30896497e-01 9.64131534e-01 5.77404618e-01
-1.15554619e+00 6.02372229e-01 1.13288909e-01 1.78451076e-01
5.14703929e-01 -1.03729928e+00 2.23850265e-01 1.09476829e+00
-1.77896559e-01 5.91226161e-01 7.53910244e-01 -5.50314844e-01
-9.61259961e-01 -9.56440210e-01 5.09353936e-01 -8.07412565e-02
5.73851049e-01 -1.45825475e-01 -1.13087881e+00 2.70919144e-01
-1.76635742e-01 -6.46308586e-02 1.04546642e+00 3.08113754e-01
-3.13602298e-01 -3.34477633e-01 -1.77853775e+00 3.10104221e-01
4.32683378e-01 5.99064454e-02 -3.77041668e-01 1.72580302e-01
5.58845745e-03 7.09358137e-03 -6.88812137e-01 5.54693520e-01
4.71896529e-01 -1.37089014e+00 8.13314199e-01 -8.95386755e-01
-1.66963600e-02 -1.69801906e-01 -3.81019041e-02 -8.86984527e-01
-2.02923134e-01 -2.96654999e-01 -1.57375216e-01 1.27788234e+00
3.39473426e-01 -8.78987014e-01 9.57297862e-01 7.14489818e-01
1.55288607e-01 -7.63065696e-01 -8.34413588e-01 -6.56916559e-01
-3.04511219e-01 -4.97212678e-01 7.16198444e-01 1.14869165e+00
-2.95770168e-01 -9.27713141e-02 -1.70828104e-01 2.55578697e-01
7.01363385e-01 -3.11708003e-01 8.16520989e-01 -1.78321946e+00
5.50431125e-02 -6.58123314e-01 -9.92721677e-01 4.61925566e-01
-1.03869103e-01 -5.47649324e-01 -7.38118112e-01 -1.10679269e+00
-1.52032539e-01 -6.05917037e-01 -7.99383402e-01 8.47872138e-01
1.19752206e-01 1.35965720e-01 -2.00410798e-01 1.51188016e-01
-2.55854100e-01 4.72706668e-02 2.47984409e-01 1.13021769e-01
-4.12964225e-01 1.85554907e-01 -6.79811835e-01 7.82901347e-01
1.25259507e+00 -5.58555007e-01 -2.34257296e-01 1.23768069e-01
-1.34491161e-01 -6.13104939e-01 4.80707772e-02 -1.43158710e+00
2.48926118e-01 -2.68803596e-01 6.77932322e-01 -3.93574387e-01
2.85319332e-02 -9.86938357e-01 1.38614863e-01 8.73388171e-01
-3.79630481e-03 2.79493839e-01 5.05629718e-01 2.63213545e-01
-8.01770762e-02 -3.16609234e-01 1.20392144e+00 6.65509403e-02
-8.85141671e-01 4.09931511e-01 -4.88293469e-01 -3.62985760e-01
1.37908149e+00 -3.30888867e-01 -3.41012120e-01 -1.09671699e-02
-5.64644396e-01 1.00739561e-01 1.62809536e-01 5.97261727e-01
6.29034400e-01 -8.98923695e-01 -6.61632121e-01 7.35982239e-01
4.33892086e-02 -4.80388850e-01 9.26390216e-02 5.79908252e-01
-9.56860363e-01 4.09314692e-01 -9.23404753e-01 -4.13968682e-01
-1.24414253e+00 4.02508259e-01 4.10002530e-01 -3.89956385e-01
-1.52419701e-01 6.97413445e-01 -6.41881943e-01 -3.70903224e-01
3.09451103e-01 4.39782202e-01 -4.03588325e-01 3.47219259e-02
7.38776207e-01 7.31186450e-01 5.10355890e-01 -3.95226419e-01
-6.79424226e-01 -9.82357189e-02 -6.55108154e-01 1.89216375e-01
1.47983873e+00 3.86457980e-01 -3.70604068e-01 1.19498678e-01
9.61002171e-01 -4.16529328e-01 -3.91355425e-01 8.06273893e-02
6.42817497e-01 -5.31568527e-01 8.54003429e-02 -1.07865345e+00
-1.06208909e+00 7.18739629e-01 1.16164744e+00 6.11610830e-01
1.20022368e+00 -7.34204948e-01 4.57461566e-01 4.48769957e-01
2.81753037e-02 -9.56419945e-01 -1.53684005e-01 4.92386073e-01
3.57058048e-01 -1.46452475e+00 -4.17195410e-02 1.36447534e-01
-1.85898036e-01 1.29840159e+00 1.06250560e+00 -1.95669264e-01
8.75233352e-01 1.74840391e-01 1.20411485e-01 -3.66790593e-01
-7.99525082e-01 -2.46416405e-02 -1.01428302e-02 8.08035314e-01
3.64506274e-01 -1.57682985e-01 -4.36413020e-01 4.51879680e-01
-1.08998036e-02 -2.42940813e-01 1.81462079e-01 1.13921750e+00
-6.75580859e-01 -1.25361490e+00 -5.06698787e-01 9.68961179e-01
-7.61410177e-01 2.47836381e-01 -6.52768254e-01 1.01708329e+00
2.05792993e-01 1.20199513e+00 6.95825070e-02 -7.39952147e-01
3.41844857e-01 3.00299048e-01 -2.61067480e-01 -3.69860291e-01
-8.25659215e-01 -3.07722181e-01 -3.92778553e-02 -4.69515920e-01
1.73370726e-02 -4.44208890e-01 -1.01274955e+00 -8.41180444e-01
-2.79072732e-01 3.88864964e-01 1.10199416e+00 1.12053418e+00
4.93397832e-01 5.17254233e-01 7.39938736e-01 -3.39915514e-01
-5.28481424e-01 -1.04264796e+00 -5.06664753e-01 4.64956820e-01
-7.68374233e-03 -7.11915672e-01 -4.88428324e-01 -5.54831982e-01] | [5.2505669593811035, 7.18936014175415] |
36d03122-77ca-4779-afb5-b3b6d642c7ed | a-representation-learning-framework-for-multi | null | null | https://www.aaai.org/ocs/index.php/AAAI/AAAI16/paper/view/12236 | https://www.aaai.org/ocs/index.php/AAAI/AAAI16/paper/view/12236/12016 | A Representation Learning Framework for Multi-Source Transfer Parsing | Cross-lingual model transfer has been a promising approach for inducing dependency parsers for low-resource languages where annotated treebanks are not available. The major obstacles for the model transfer approach are two-fold: 1. Lexical features are not directly transferable across languages; 2. Target language-specific syntactic structures are difficult to be recovered. To address these two challenges, we present a novel representation learning framework for multi-source transfer parsing. Our framework allows multi-source transfer parsing using full lexical features straightforwardly. By evaluating on the Google universal dependency treebanks (v2.0), our best models yield an absolute improvement of 6.53% in averaged labeled attachment score, as compared with delexicalized multi-source transfer models. We also significantly outperform the state-of-the-art transfer system proposed most recently. | ['Ting Liu', 'Haifeng Wang', 'David Yarowsky', 'Wanxiang Che', 'Jiang Guo'] | 2016-03-05 | null | null | null | null | ['cross-lingual-zero-shot-dependency-parsing'] | ['natural-language-processing'] | [-5.54481447e-02 3.60970587e-01 -4.23397064e-01 -5.70603132e-01
-1.76243854e+00 -8.67516279e-01 1.50968835e-01 1.08385742e-01
-5.34716070e-01 1.16444492e+00 4.16523695e-01 -4.21602815e-01
5.03391862e-01 -6.00885272e-01 -8.83855700e-01 -1.16258807e-01
6.47182465e-02 6.23610914e-01 2.62032747e-01 -4.58399534e-01
-7.52346218e-02 1.52201891e-01 -8.00335824e-01 6.33291602e-01
8.38035762e-01 3.67209136e-01 4.33670402e-01 3.80193740e-01
-5.45786321e-01 7.00092375e-01 -5.27298689e-01 -7.09191084e-01
-1.67834967e-01 -5.34889102e-01 -1.32982683e+00 -6.48835540e-01
2.53136843e-01 -1.73390940e-01 2.20021933e-01 8.94741476e-01
3.34625334e-01 -3.33789259e-01 6.44971967e-01 -6.57025993e-01
-1.08504272e+00 1.05304074e+00 -5.31114936e-01 3.07388097e-01
4.62362081e-01 -4.59280223e-01 1.35775971e+00 -1.00473034e+00
8.59201133e-01 1.49933302e+00 5.34924626e-01 9.42907095e-01
-1.17855287e+00 -7.80708373e-01 1.03537925e-01 1.56460050e-02
-1.12253678e+00 -5.74624717e-01 5.69458306e-01 -2.63712913e-01
1.36175060e+00 -3.09197456e-01 -1.29967541e-01 1.17143655e+00
2.54233330e-01 7.89947510e-01 1.15613461e+00 -1.02458298e+00
-2.06660375e-01 3.09602290e-01 1.28773794e-01 7.81651139e-01
2.24259689e-01 -1.85818613e-01 -6.59064710e-01 1.04269318e-01
7.26513147e-01 -8.22126091e-01 9.13705900e-02 1.94157317e-01
-9.43929613e-01 1.19568551e+00 3.47670525e-01 7.40235090e-01
-1.17475964e-01 -7.89421648e-02 6.76096022e-01 6.45489037e-01
8.47261369e-01 2.43502244e-01 -1.10696816e+00 -3.14513445e-02
-5.59933603e-01 -1.24753021e-01 8.48573923e-01 1.18051338e+00
8.97282958e-01 -1.32674323e-02 1.74587622e-01 9.88765776e-01
3.41101885e-01 4.19538289e-01 4.69326019e-01 -5.30511975e-01
1.22797585e+00 4.00253028e-01 -1.19967781e-01 -2.80369550e-01
-3.63391787e-01 1.89717278e-01 -4.27560508e-01 -8.94395858e-02
6.25272810e-01 -5.41032374e-01 -5.43867886e-01 1.98795795e+00
2.34929815e-01 -2.18336418e-01 5.20536363e-01 2.92674810e-01
9.55994010e-01 7.22433209e-01 8.27023149e-01 -1.39158428e-01
1.59458673e+00 -9.37338173e-01 -6.47588909e-01 -6.09479964e-01
1.34926832e+00 -9.82725084e-01 1.20216119e+00 -1.85590550e-01
-1.23333442e+00 -5.23427367e-01 -8.06108534e-01 -2.61234194e-01
-3.92050803e-01 1.80507347e-01 7.35312283e-01 8.24671805e-01
-1.28851962e+00 3.59811634e-01 -8.94873083e-01 -4.78263646e-01
2.83800393e-01 3.39243829e-01 -9.17454541e-01 -2.88011849e-01
-1.31533945e+00 1.22780132e+00 5.28447866e-01 -3.22826385e-01
-5.61749041e-01 -7.64415979e-01 -1.20739412e+00 -5.86442798e-02
-1.27881289e-01 -3.58968109e-01 1.37326968e+00 -8.76240194e-01
-1.60617185e+00 1.34085894e+00 -2.99635381e-01 -1.90165922e-01
4.22153138e-02 -5.79439223e-01 -2.87995636e-01 -8.61512870e-02
5.49246967e-01 6.21943831e-01 2.88908452e-01 -9.68445897e-01
-7.27351189e-01 -2.29764789e-01 2.33912356e-02 2.05643833e-01
-4.11781192e-01 9.57041204e-01 -1.77917987e-01 -4.73547637e-01
-3.11904103e-01 -8.76067162e-01 -1.73716232e-01 -5.98346651e-01
-8.53329226e-02 -6.40675664e-01 4.86130476e-01 -1.02257240e+00
9.45392191e-01 -1.69590318e+00 1.66289479e-01 -6.07279956e-01
-6.18308425e-01 5.29399633e-01 -4.61338580e-01 4.96240050e-01
-8.47604945e-02 4.83371019e-01 -1.51710972e-01 -5.13566315e-01
-3.03948343e-01 3.81430864e-01 -1.77583426e-01 5.88754425e-03
6.84778154e-01 9.59231198e-01 -8.38900566e-01 -6.85813725e-01
-1.71661139e-01 4.60816801e-01 -7.01344550e-01 5.38825274e-01
-6.48343340e-02 8.67561698e-01 -5.68815708e-01 6.90158665e-01
4.85869706e-01 2.25075573e-01 4.94018257e-01 2.45268364e-02
-2.18413994e-01 1.03510296e+00 -6.38310969e-01 2.21531701e+00
-1.05082321e+00 5.29481657e-02 1.46658450e-01 -7.87239015e-01
9.65359688e-01 7.34594107e-01 -1.61136165e-02 -4.40692276e-01
-5.94597636e-03 5.30789912e-01 7.56045282e-02 -1.83807507e-01
2.26149186e-01 -5.92835665e-01 -6.40354395e-01 5.14916182e-01
8.41763258e-01 -3.52174491e-02 1.19438834e-01 9.59548056e-02
1.12310064e+00 5.25473952e-01 5.67924678e-01 -5.84393382e-01
4.75127250e-01 9.95096471e-03 7.74190664e-01 2.14213014e-01
-7.47555718e-02 4.66457725e-01 4.17545974e-01 -2.26931676e-01
-5.29407740e-01 -1.12826526e+00 -2.23469324e-02 1.78844082e+00
-5.52914321e-01 -3.39425832e-01 -9.53113317e-01 -1.18297946e+00
-4.14208531e-01 7.64356494e-01 -5.62079608e-01 1.96549609e-01
-1.24691558e+00 -8.17389011e-01 8.24746609e-01 9.81894732e-01
5.72322123e-02 -1.47209322e+00 -7.63475597e-02 6.40923858e-01
-5.22845864e-01 -1.51669240e+00 -3.10976565e-01 3.06878388e-01
-9.39209163e-01 -7.23563910e-01 -5.05869687e-01 -1.28953218e+00
4.23655450e-01 -2.12063551e-01 1.55263186e+00 6.55537769e-02
3.40824723e-02 -2.64824219e-02 -6.32127702e-01 -4.80903804e-01
-7.28396833e-01 6.95584238e-01 -5.64835668e-02 -4.67266172e-01
5.65137923e-01 -5.05354583e-01 1.81659274e-02 -3.57898995e-02
-4.05294776e-01 -2.84075260e-01 6.67655945e-01 7.43394911e-01
4.71648365e-01 -8.53805542e-01 1.19144428e+00 -1.41469157e+00
3.92129004e-01 -6.88485503e-01 -4.05387968e-01 3.80703330e-01
-1.50579736e-01 1.25948340e-01 6.38689816e-01 -1.27306664e-02
-1.74131167e+00 1.64367050e-01 -6.13442957e-01 2.90371507e-01
-3.44395578e-01 4.87507850e-01 -4.99131531e-01 1.12569690e-01
4.75415349e-01 -2.68072098e-01 -4.72433001e-01 -9.56190586e-01
5.83480120e-01 6.11044347e-01 3.25087219e-01 -1.12430060e+00
5.86732447e-01 -1.03690349e-01 -4.21960264e-01 -6.87804878e-01
-1.32451856e+00 -3.46167833e-01 -1.52177060e+00 4.73590940e-01
1.26268613e+00 -1.24060893e+00 2.95597836e-02 2.23866791e-01
-1.80285728e+00 -5.45803964e-01 4.97291796e-02 4.40487713e-01
-4.41054493e-01 1.19685337e-01 -1.24769163e+00 -2.61289805e-01
-5.57830930e-01 -8.86694968e-01 1.05877447e+00 -7.59912096e-03
-2.93272108e-01 -1.46232927e+00 4.59034681e-01 3.50624114e-01
3.02596658e-01 1.38659626e-01 1.19182897e+00 -8.04790556e-01
-2.27260351e-01 1.45698175e-01 -2.00953260e-01 4.31587189e-01
4.55872118e-01 -2.68855840e-01 -8.88002634e-01 -3.78536761e-01
-3.36178303e-01 -8.47661912e-01 4.50860322e-01 1.16453312e-01
3.09878290e-01 4.40446706e-03 -3.21618736e-01 4.42771375e-01
1.37805438e+00 -1.08496808e-01 4.30288762e-01 3.15773398e-01
7.82376468e-01 7.97850311e-01 7.44169295e-01 -1.60214916e-01
8.26019824e-01 5.05703211e-01 -5.75763173e-03 -9.64405835e-02
-3.89515191e-01 -4.16597486e-01 8.72218668e-01 1.54389179e+00
-1.70494959e-01 -5.10159917e-02 -9.75533962e-01 8.87424529e-01
-1.47519767e+00 -3.69603723e-01 -2.15185165e-01 1.79413450e+00
1.25800133e+00 5.36524244e-02 -1.84557855e-01 -4.42021310e-01
7.31469512e-01 -3.86692025e-02 -1.18581848e-02 -8.75071347e-01
-4.84881513e-02 9.52616036e-01 2.94378906e-01 7.05488443e-01
-1.13236523e+00 1.92989540e+00 6.44652748e+00 4.02160645e-01
-8.03676844e-01 7.67791390e-01 2.67768741e-01 4.01354462e-01
-4.46225032e-02 2.59383321e-01 -1.47992587e+00 7.58716613e-02
1.55620730e+00 -2.12690458e-01 -1.23950891e-01 8.33411455e-01
-4.95187730e-01 3.89336169e-01 -1.26781070e+00 2.83564121e-01
-2.01865043e-02 -1.03417623e+00 4.60574478e-02 -6.66207597e-02
5.60577393e-01 6.09617293e-01 -3.18719238e-01 6.85297966e-01
8.84639323e-01 -9.43177700e-01 5.40463209e-01 -3.07332993e-01
1.22013283e+00 -8.67334068e-01 7.28160501e-01 2.83214033e-01
-1.43431556e+00 4.53092605e-01 -6.99380636e-01 -1.78600252e-01
6.14678919e-01 1.33088067e-01 -7.37043083e-01 7.01705515e-01
7.11483419e-01 5.31693995e-01 -4.84329224e-01 2.95480579e-01
-8.80216181e-01 1.05277586e+00 -1.79875776e-01 2.03355610e-01
2.78304160e-01 3.32251303e-02 7.70637691e-02 1.82615912e+00
3.48212153e-01 -2.93871183e-02 2.10148573e-01 3.99038434e-01
-4.00016159e-01 7.34359503e-01 -7.47000396e-01 2.65919179e-01
5.83143294e-01 1.34189105e+00 -5.34596562e-01 -1.08772077e-01
-9.54881787e-01 1.05917573e+00 1.17355335e+00 -1.36980280e-01
-4.62067813e-01 -2.59617209e-01 6.20708048e-01 -1.83923587e-01
3.65303606e-01 -2.77407587e-01 -6.87459782e-02 -1.33342242e+00
-2.21141547e-01 -6.83098912e-01 7.37944603e-01 -2.70899564e-01
-1.64589143e+00 9.92381632e-01 -7.86779542e-03 -7.78005838e-01
-5.40208697e-01 -9.22368228e-01 -6.31967783e-01 1.19952929e+00
-1.93711746e+00 -1.96112752e+00 4.51511383e-01 6.87855721e-01
7.33876884e-01 -3.34041893e-01 1.78876317e+00 4.16830987e-01
-3.95872235e-01 9.14583623e-01 -2.56950259e-01 5.65086424e-01
1.02756453e+00 -1.37938607e+00 9.07660186e-01 8.91789794e-01
3.16643596e-01 5.36360741e-01 1.17542289e-01 -4.20686364e-01
-9.60778952e-01 -1.31214988e+00 1.51720440e+00 -8.49978864e-01
9.46057796e-01 -6.03400469e-01 -1.25605309e+00 1.36614656e+00
5.88959336e-01 2.39129886e-01 1.11869419e+00 6.42868996e-01
-6.40983045e-01 2.27969855e-01 -1.11328864e+00 2.09528059e-01
1.05756807e+00 -4.78525639e-01 -1.00295722e+00 2.92607188e-01
8.25618505e-01 -2.59252250e-01 -1.21889853e+00 2.30729088e-01
1.47452965e-01 -5.74400365e-01 7.39078939e-01 -9.05418634e-01
5.21599233e-01 3.78311068e-01 -2.83025503e-01 -1.60195923e+00
-4.46235090e-01 -5.41767061e-01 2.91488320e-01 2.01777649e+00
8.68480265e-01 -6.73872948e-01 3.77234131e-01 3.12693834e-01
-2.65424013e-01 -1.82861805e-01 -1.23637402e+00 -7.84920275e-01
1.14560199e+00 -1.42669663e-01 3.58931810e-01 1.15650988e+00
2.46381670e-01 1.12137389e+00 -2.49551952e-01 1.24218717e-01
5.37042379e-01 -5.48156910e-02 4.43294227e-01 -1.36455834e+00
-3.36990744e-01 7.62408972e-02 3.32745686e-02 -7.49607027e-01
1.00706720e+00 -1.28754508e+00 9.25662816e-02 -1.43180585e+00
1.26333669e-01 -9.50563848e-01 -5.78363001e-01 9.45716143e-01
-3.36404890e-01 1.92849696e-01 3.11419725e-01 1.23679005e-01
-2.62865186e-01 2.93928355e-01 9.00232375e-01 3.47161889e-01
-5.11261038e-02 -2.05485970e-01 -1.00034487e+00 9.27864254e-01
1.06147838e+00 -1.01153409e+00 -9.25786197e-02 -1.09623575e+00
-1.40607372e-01 3.53194326e-01 -4.61221308e-01 -5.92458606e-01
-2.44744718e-01 -7.13260099e-02 -4.78816591e-02 -1.90974906e-01
1.73686743e-01 -2.14990973e-01 -6.00463927e-01 2.10696325e-01
-1.00994706e-01 2.77606070e-01 5.61783969e-01 1.30693346e-01
-3.61103535e-01 -5.57544351e-01 7.99674213e-01 -4.64793354e-01
-5.71878254e-01 8.48738998e-02 -4.39966649e-01 5.45596898e-01
7.11847723e-01 5.26978493e-01 -2.84857839e-01 9.35436934e-02
-4.81477439e-01 -1.66435093e-01 1.35396272e-01 7.64337957e-01
1.27037823e-01 -1.16255796e+00 -1.23084378e+00 4.07665595e-02
7.09436238e-02 -9.47220922e-02 -1.08297214e-01 2.24267110e-01
-3.29052597e-01 5.57225764e-01 -4.74941343e-01 -2.45475635e-01
-1.30129611e+00 2.13941634e-01 -1.25212997e-01 -8.98515046e-01
-4.73708391e-01 1.17242491e+00 3.87451470e-01 -1.01199353e+00
-3.79598528e-01 -1.74624160e-01 -3.62628013e-01 1.13471441e-01
6.21094525e-01 -8.28522295e-02 3.93999547e-01 -1.04150379e+00
-6.09910131e-01 5.88324666e-01 -4.99180824e-01 -1.61389172e-01
1.70377898e+00 7.86506850e-03 -2.17204288e-01 4.18149561e-01
1.28023350e+00 2.74446815e-01 -8.73657763e-01 -3.95540982e-01
5.67523599e-01 2.97181662e-02 -2.58673191e-01 -8.59890401e-01
-7.85539031e-01 1.18698800e+00 1.19067304e-01 -2.70572960e-01
7.58021712e-01 4.89781141e-01 1.03147042e+00 3.74998569e-01
7.43300676e-01 -8.83808076e-01 -1.07693963e-01 1.04746091e+00
8.12697291e-01 -1.26152837e+00 -3.98276538e-01 -8.02330554e-01
-7.40894675e-01 8.67039144e-01 8.06422949e-01 -2.20375970e-01
7.00508296e-01 6.41696870e-01 5.75949669e-01 4.58197147e-02
-7.59328067e-01 -2.25820273e-01 -6.28054217e-02 8.69665384e-01
1.36082268e+00 1.72527090e-01 -4.71625179e-01 9.83020365e-01
-2.37042874e-01 -3.05124670e-01 4.21354294e-01 1.05104256e+00
-2.60783941e-01 -1.95158851e+00 -8.08778703e-02 -2.01111421e-01
-1.11991727e+00 -5.81181526e-01 -1.97864622e-01 8.89255583e-01
-2.37011090e-01 1.00319815e+00 -7.90060535e-02 1.59821317e-01
4.28586483e-01 4.52244878e-01 7.22660303e-01 -1.46161532e+00
-7.28591919e-01 -9.54428613e-02 4.65467423e-01 -3.47230703e-01
-5.90777814e-01 -7.02016592e-01 -1.36182654e+00 2.29232326e-01
-3.54821563e-01 2.44296223e-01 5.92845321e-01 1.03112113e+00
2.67358571e-01 2.97794431e-01 3.03646207e-01 -9.65524375e-01
-3.46495926e-01 -1.23562288e+00 -3.03016305e-01 1.84618711e-01
-1.81249350e-01 -5.43363333e-01 -2.71672346e-02 4.82419997e-01] | [10.535527229309082, 9.817673683166504] |
d7c5ca83-1b27-4999-a0a8-eddc383be55e | double-sided-information-aided-temporal | 2205.07494 | null | https://arxiv.org/abs/2205.07494v1 | https://arxiv.org/pdf/2205.07494v1.pdf | Double-Sided Information Aided Temporal-Correlated Massive Access | This letter considers temporal-correlated massive access, where each device, once activated, is likely to transmit continuously over several consecutive frames. Motivated by that the device activity at each frame is correlated to not only its previous frame but also its next frame, we propose a double-sided information (DSI) aided joint activity detection and channel estimation algorithm based on the approximate message passing (AMP) framework. The DSI is extracted from the estimation results in a sliding window that contains the target detection frame and its previous and next frames. The proposed algorithm demonstrates superior performance over the state-of-the-art methods. | ['Yunfeng Guan', 'Meixia Tao', 'Weifeng Zhu'] | 2022-05-16 | null | null | null | null | ['activity-detection'] | ['computer-vision'] | [ 8.16195309e-01 -1.49136499e-01 -5.86551547e-01 1.13066159e-01
-7.19457150e-01 -2.22565889e-01 2.56966978e-01 1.23464905e-01
-4.02590722e-01 1.17957890e+00 1.30671725e-01 -3.49678099e-01
2.47664511e-01 -2.75791883e-01 -4.03508186e-01 -9.06938672e-01
-7.45548189e-01 -2.51699418e-01 4.32351142e-01 5.43080866e-01
-1.25401467e-01 1.57791227e-01 -6.66974962e-01 1.09683657e-02
1.38467073e-01 1.92599642e+00 5.07024944e-01 1.16184199e+00
4.53045398e-01 1.11146402e+00 -1.03350604e+00 5.40224239e-02
-1.76009685e-01 -7.15797782e-01 -1.93610013e-01 3.77645403e-01
-5.29464960e-01 -7.33816028e-01 -8.89717758e-01 5.02873778e-01
5.46714962e-01 -3.19020271e-01 3.52625191e-01 -1.00982285e+00
2.62484699e-01 4.05705631e-01 -8.23445082e-01 7.21747160e-01
6.16928995e-01 -2.44877756e-01 6.99299097e-01 -7.49204397e-01
4.32805121e-01 8.13319921e-01 2.73073047e-01 -3.31359431e-02
-7.20510721e-01 -7.38882303e-01 4.61965501e-02 2.30810001e-01
-1.85249543e+00 -5.47541261e-01 5.76635122e-01 1.19607009e-01
6.33886337e-01 2.60114908e-01 7.17461944e-01 9.08712149e-01
3.95123929e-01 1.10805011e+00 5.89115381e-01 -5.37524641e-01
5.77386141e-01 -3.57986689e-01 -4.55672473e-01 6.97810173e-01
2.51491457e-01 5.64752296e-02 -8.30742002e-01 -6.26077294e-01
6.56321764e-01 -2.20181793e-01 -5.34357369e-01 1.75288975e-01
-1.52342546e+00 2.75550872e-01 -5.08280456e-01 3.84122968e-01
-1.02302861e+00 4.89419192e-01 -9.06967074e-02 7.02477843e-02
1.63960442e-01 -4.87361580e-01 -9.40685943e-02 -6.07284725e-01
-1.24302781e+00 -2.63464361e-01 1.00920653e+00 1.17812431e+00
5.63064158e-01 1.78785697e-01 -6.53659046e-01 3.09840053e-01
6.23479962e-01 7.66042233e-01 -9.64855626e-02 -1.04106891e+00
6.29672945e-01 -3.40186894e-01 4.62713778e-01 -9.30437744e-01
-1.93507075e-01 -1.07409763e+00 -1.15950632e+00 -5.01937985e-01
2.97335029e-01 -7.50818372e-01 -1.34580255e-01 1.61980522e+00
2.74846911e-01 8.80225241e-01 1.60348386e-01 5.44901073e-01
-2.93402612e-01 9.95197892e-01 -2.11739913e-01 -1.34498012e+00
1.20863569e+00 -5.15541852e-01 -1.25344372e+00 -9.97155532e-02
7.20942095e-02 -7.53880382e-01 -2.45763242e-01 8.67073119e-01
-1.53491151e+00 -4.46247816e-01 -1.66417313e+00 9.14121091e-01
5.37755311e-01 3.81382555e-01 2.61604995e-01 9.30805564e-01
-7.14025021e-01 -4.88167740e-02 -8.18463147e-01 -5.63607253e-02
4.33758974e-01 4.57304120e-01 2.38064989e-01 8.37289356e-03
-1.03217971e+00 1.22733749e-01 2.62199402e-01 -1.47376731e-01
-9.98088121e-01 -3.11247736e-01 -5.15481174e-01 4.61482257e-02
5.30707419e-01 -5.80720484e-01 1.41160202e+00 -5.73776841e-01
-1.41677833e+00 -1.82685591e-02 -9.88609731e-01 -8.27740490e-01
1.89989507e-01 4.49414514e-02 -8.40464950e-01 4.79351550e-01
-2.85313845e-01 -2.84831971e-01 1.08792436e+00 -8.73605013e-01
-1.04382265e+00 -3.08965206e-01 -1.27829567e-01 2.06458300e-01
-3.35165054e-01 3.83833162e-02 -1.16928196e+00 -8.76535654e-01
1.17643662e-01 -8.57290149e-01 -2.79443830e-01 -2.04997838e-01
-5.21638811e-01 4.81838793e-01 9.28796887e-01 -4.86071467e-01
2.25960851e+00 -2.27934217e+00 8.23740289e-02 3.74206841e-01
9.29455385e-02 9.42119062e-02 3.87146235e-01 5.22456050e-01
6.08495295e-01 -3.50938588e-01 1.00164749e-01 -4.80096102e-01
-4.44331676e-01 2.62083650e-01 -2.48361290e-01 6.88532591e-01
-3.32182229e-01 2.76052803e-01 -9.55832124e-01 -3.51888984e-01
2.60280818e-01 5.05538881e-01 -5.52862473e-02 2.20097020e-01
8.65116641e-02 6.52204275e-01 -5.58288991e-01 5.79712391e-01
7.81299710e-01 -5.73220968e-01 6.36802554e-01 -3.48302305e-01
-1.76655740e-01 -9.27743909e-04 -1.33015335e+00 1.71243298e+00
-6.08481407e-01 5.98681867e-01 2.48147130e-01 -7.00221360e-01
1.76683381e-01 1.14253175e+00 7.40615964e-01 -8.28722656e-01
1.19876251e-01 2.75626838e-01 -1.20943218e-01 -2.93938667e-02
4.40359682e-01 1.55058295e-01 -2.30374262e-01 5.67862272e-01
-8.47559795e-02 7.02465534e-01 1.80567235e-01 5.12518406e-01
1.73068452e+00 -4.85693842e-01 1.16484058e+00 2.45460182e-01
6.63719714e-01 -8.03423464e-01 6.01659536e-01 1.21453047e+00
-2.71458000e-01 4.90506329e-02 1.07305035e-01 4.69141304e-02
-6.57153308e-01 -1.18361568e+00 4.58393209e-02 6.72764421e-01
6.28210604e-01 -6.80994272e-01 -5.17314732e-01 -3.33547622e-01
-3.19491267e-01 4.00408804e-01 -2.61726409e-01 2.32494082e-02
-3.78285140e-01 -5.31668961e-01 5.29102802e-01 2.94400513e-01
8.81120265e-01 -1.86672404e-01 -7.82161176e-01 9.34412599e-01
-5.22611916e-01 -1.49527860e+00 -5.81975996e-01 1.43304259e-01
-4.84914362e-01 -6.17969215e-01 -7.55335271e-01 -1.64381966e-01
3.69028628e-01 5.69365680e-01 9.93367314e-01 7.48629868e-03
1.61172226e-01 4.48645860e-01 -4.23643589e-01 -1.67236254e-01
-3.77616942e-01 -1.28396541e-01 -7.46974275e-02 7.56882906e-01
1.85700923e-01 -4.07277346e-01 -8.75424564e-01 5.01480937e-01
-6.52830362e-01 -1.18568406e-01 6.09364986e-01 5.51931262e-01
6.29989445e-01 4.38881457e-01 2.64968991e-01 -3.38873893e-01
1.91477269e-01 -6.75720811e-01 -4.36026722e-01 2.67556399e-01
-1.25316858e-01 -2.36464992e-01 3.80630553e-01 -2.83567309e-01
-1.12350857e+00 2.65970767e-01 -2.06315339e-01 3.55633870e-02
1.08232841e-01 3.42086554e-01 -2.81212956e-01 -8.72005150e-02
2.29054496e-01 5.27297497e-01 -5.54344237e-01 3.42495330e-02
4.71733697e-02 1.03047466e+00 6.23494864e-01 -1.55144140e-01
6.57476008e-01 8.40645015e-01 1.59019217e-01 -1.19739795e+00
-5.81301510e-01 -5.62974095e-01 -4.32716966e-01 -4.18935150e-01
4.02865797e-01 -1.49934494e+00 -8.03386450e-01 5.36566019e-01
-1.34863853e+00 -1.15022408e-02 2.10807323e-01 7.69336402e-01
-2.57756025e-01 6.60036147e-01 -6.06405318e-01 -1.43437672e+00
4.27172817e-02 -8.94025743e-01 1.18009615e+00 4.53625396e-02
-4.15008724e-01 -7.89188027e-01 -2.07414880e-01 -1.93991393e-01
6.05004907e-01 1.15144834e-01 2.92238623e-01 -6.35417476e-02
-8.92038107e-01 -5.22413611e-01 1.66306779e-01 -1.22492142e-01
3.42436641e-01 -4.73526776e-01 -7.22265363e-01 -6.02196932e-01
1.73431635e-01 4.08229440e-01 2.38777563e-01 8.20516884e-01
9.66439664e-01 -4.90780711e-01 -9.03724313e-01 2.79576361e-01
1.39717162e+00 7.68759966e-01 9.64577019e-01 -1.75564185e-01
1.58774294e-02 -6.27974629e-01 9.27603006e-01 1.40515196e+00
2.36813039e-01 1.26862919e+00 2.66854435e-01 2.27238219e-02
5.31465597e-02 1.02582552e-01 3.41839552e-01 6.15597069e-01
1.27056912e-01 -9.73318338e-01 -3.31777722e-01 4.07543749e-01
-1.81709468e+00 -1.32914519e+00 -1.42583102e-01 2.50711083e+00
5.24393797e-01 3.08061182e-01 1.70564830e-01 5.61290026e-01
6.60231352e-01 3.73991817e-01 -6.13134742e-01 2.38888487e-01
-4.31965142e-01 2.05853239e-01 9.08791780e-01 4.92599130e-01
-9.31661487e-01 2.07780197e-01 6.59019613e+00 1.38096821e+00
-6.87905967e-01 5.22824287e-01 7.88866460e-01 -3.00611943e-01
1.44845381e-01 -1.65675744e-01 -6.92988157e-01 8.32395971e-01
1.34210765e+00 -1.01254255e-01 2.24283949e-01 3.45202059e-01
7.66720355e-01 -1.15282094e+00 -8.82774115e-01 1.21095192e+00
-7.63192624e-02 -1.42596996e+00 -2.23249137e-01 4.73367989e-01
7.40931749e-01 -3.94770086e-01 2.85206689e-03 -2.14659482e-01
-1.44639611e-01 -3.25768918e-01 6.92743242e-01 4.62159663e-01
8.57528329e-01 -6.05834723e-01 5.97198069e-01 2.71685034e-01
-1.89147270e+00 -4.54980284e-01 2.66949415e-01 -2.58860588e-01
8.50423396e-01 1.20584500e+00 -6.32370710e-01 5.72083116e-01
2.31799737e-01 3.84130180e-01 2.14417249e-01 1.10800529e+00
-1.43210202e-01 1.16983891e+00 -5.17893136e-01 -4.93069142e-02
1.02981050e-02 1.59252748e-01 6.78173184e-01 1.13944876e+00
8.94365728e-01 4.25449699e-01 3.38309944e-01 -4.03921381e-02
5.98938130e-02 -2.33979091e-01 -4.09826815e-01 4.90625918e-01
1.22493994e+00 7.13010728e-01 -4.66348082e-01 -7.62111306e-01
-9.77318764e-01 1.34911084e+00 -5.79715788e-01 6.29463613e-01
-9.91692483e-01 -4.77226436e-01 4.15502042e-01 4.11924273e-02
5.64453363e-01 -6.25006616e-01 -2.38242537e-01 -8.42797875e-01
1.92610532e-01 -4.56558138e-01 3.22469056e-01 -4.11228389e-01
-4.12966818e-01 1.41705409e-01 -3.13998133e-01 -1.67011583e+00
-5.23543596e-01 2.07363218e-01 -4.37525183e-01 5.69461703e-01
-1.28164256e+00 -5.12344480e-01 -1.69754937e-01 8.68505239e-01
3.99168134e-01 -1.16459072e-01 8.13450456e-01 5.80485225e-01
-2.87138879e-01 6.67109609e-01 2.97241598e-01 3.71204577e-02
-1.74251050e-01 -7.33047783e-01 9.07273032e-03 1.04516137e+00
-4.77532744e-02 1.41384825e-01 7.13218629e-01 -5.44689238e-01
-1.53375769e+00 -6.96547747e-01 7.63311386e-01 2.05928341e-01
3.83993030e-01 -3.07514042e-01 -3.46212596e-01 2.67978877e-01
3.39996159e-01 1.40058137e-02 1.04749238e+00 -6.17228985e-01
3.68932903e-01 -2.09466308e-01 -1.01344192e+00 4.49700981e-01
7.80202985e-01 -5.80052853e-01 1.40772551e-01 -7.27138072e-02
1.36098161e-01 -3.71118337e-01 -6.21351600e-01 2.72491038e-01
7.65513897e-01 -9.61153388e-01 7.43974626e-01 4.66820866e-01
-5.72273076e-01 -6.33693457e-01 -6.88661158e-01 -6.68547034e-01
-1.17439680e-01 -1.38669252e+00 -1.24030101e+00 9.18570399e-01
4.72440630e-01 1.81250334e-01 9.46642339e-01 1.47833660e-01
4.71629828e-01 -3.88677448e-01 -1.60692418e+00 -8.00668955e-01
-1.08407879e+00 -1.00513184e+00 2.02886552e-01 6.70529157e-02
3.54123414e-01 4.99576420e-01 -1.12296021e+00 8.44789565e-01
7.69633889e-01 -4.04583573e-01 6.74950004e-01 -9.39271808e-01
-8.67275298e-01 3.53990197e-01 -4.76851672e-01 -2.01466990e+00
-4.32297856e-01 1.00172818e-01 3.62430997e-02 -8.58864248e-01
2.17959329e-01 -6.16219155e-02 -3.90433013e-01 -1.21233396e-01
-1.97695158e-02 3.63663018e-01 3.18266526e-02 1.51563361e-01
-1.19703174e+00 4.22998905e-01 5.88329852e-01 3.35690491e-02
-1.58814222e-01 5.94059825e-01 -1.47974029e-01 3.02222371e-01
6.64032519e-01 -3.38863611e-01 -4.52741206e-01 1.06089629e-01
8.12590355e-04 1.02462840e+00 -2.66055986e-02 -1.67503488e+00
4.07121032e-01 2.88654149e-01 3.80619496e-01 -8.95028889e-01
7.87851572e-01 -1.24228692e+00 5.32208979e-01 8.42738807e-01
-1.09243244e-01 -3.25067043e-01 -2.89664324e-02 1.29131591e+00
-1.45439915e-02 1.67217284e-01 5.90764046e-01 2.54532158e-01
-5.39327621e-01 4.29667234e-01 -1.27329659e+00 -4.36572373e-01
1.17895377e+00 -4.54434544e-01 1.98639378e-01 -1.03923309e+00
-7.91956306e-01 2.18563657e-02 -2.66675260e-02 -9.46204439e-02
6.32732749e-01 -1.23196077e+00 -3.41953218e-01 -1.48794139e-02
-1.80807292e-01 -5.98073840e-01 1.76635981e-01 1.10984814e+00
2.07262754e-01 7.72401989e-01 3.55898261e-01 -7.28221059e-01
-1.47852707e+00 4.35368240e-01 1.71511117e-02 -6.24362171e-01
-5.22380695e-02 5.35752952e-01 -1.29557490e-01 1.25859749e+00
5.73104382e-01 -3.59914592e-03 2.39093810e-01 -4.00533140e-01
1.14504921e+00 5.48088849e-01 -4.64679226e-02 -6.90449417e-01
-1.33293405e-01 2.80397922e-01 8.14211816e-02 -5.72778702e-01
3.98873270e-01 -1.00242341e+00 5.00893295e-01 3.85274529e-01
1.28535604e+00 3.01493764e-01 -1.38199830e+00 -6.56083226e-01
-1.31175011e-01 -8.84104133e-01 3.28114986e-01 -6.38734996e-01
-9.85198498e-01 3.56583685e-01 9.19489264e-01 2.61857897e-01
1.29201293e+00 -1.11258440e-01 9.58571374e-01 -6.52785525e-02
1.31113112e+00 -9.58939075e-01 1.13103800e-01 6.04167394e-02
1.79357708e-01 -5.48330247e-01 1.96976289e-02 -3.62307847e-01
-1.88243106e-01 9.46191788e-01 -1.64283931e-01 7.88043678e-01
9.58535552e-01 4.88823265e-01 -5.51540911e-01 8.74722302e-02
-9.20865655e-01 -2.77216673e-01 -3.07234973e-01 7.98544347e-01
2.59515494e-01 1.00315370e-01 -2.59316444e-01 2.15663731e-01
6.13517225e-01 4.36173856e-01 6.15685821e-01 1.28107929e+00
-7.20561028e-01 -1.13108885e+00 -7.30958760e-01 5.00163257e-01
-7.20517337e-01 -1.51762497e-02 2.29831338e-01 2.43293494e-01
2.84364279e-02 1.60000765e+00 1.16785698e-01 -3.80122423e-01
-1.01962034e-02 -5.25926411e-01 2.76712537e-01 -1.42103836e-01
2.17137516e-01 4.75527078e-01 3.13912719e-01 -7.17393577e-01
-5.42312145e-01 -8.43435526e-01 -1.25470316e+00 -3.78264844e-01
-2.43748292e-01 2.89936751e-01 4.02989447e-01 1.16969728e+00
4.00659740e-01 6.11514747e-01 1.23167026e+00 -9.19533074e-01
1.57248870e-01 -6.23693228e-01 -9.23824251e-01 -2.95851618e-01
5.59702039e-01 -4.24863100e-01 -1.43349677e-01 5.01697622e-02] | [6.2332563400268555, 1.4035987854003906] |
5ab16ada-2004-4dd4-8999-667189b08be9 | smap-single-shot-multi-person-absolute-3d | 2008.11469 | null | https://arxiv.org/abs/2008.11469v1 | https://arxiv.org/pdf/2008.11469v1.pdf | SMAP: Single-Shot Multi-Person Absolute 3D Pose Estimation | Recovering multi-person 3D poses with absolute scales from a single RGB image is a challenging problem due to the inherent depth and scale ambiguity from a single view. Addressing this ambiguity requires to aggregate various cues over the entire image, such as body sizes, scene layouts, and inter-person relationships. However, most previous methods adopt a top-down scheme that first performs 2D pose detection and then regresses the 3D pose and scale for each detected person individually, ignoring global contextual cues. In this paper, we propose a novel system that first regresses a set of 2.5D representations of body parts and then reconstructs the 3D absolute poses based on these 2.5D representations with a depth-aware part association algorithm. Such a single-shot bottom-up scheme allows the system to better learn and reason about the inter-person depth relationship, improving both 3D and 2D pose estimation. The experiments demonstrate that the proposed approach achieves the state-of-the-art performance on the CMU Panoptic and MuPoTS-3D datasets and is applicable to in-the-wild videos. | ['Wei Jiang', 'Jianan Zhen', 'Hujun Bao', 'Wentao Liu', 'Xiaowei Zhou', 'Qi Fang', 'Jiaming Sun'] | 2020-08-26 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/2395_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123600545.pdf | eccv-2020-8 | ['3d-multi-person-pose-estimation-absolute', '3d-depth-estimation', '3d-multi-person-pose-estimation-root-relative', '3d-multi-person-pose-estimation'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [ 2.66506672e-01 -1.70358688e-01 6.80454588e-03 -4.47513372e-01
-7.04498351e-01 -5.69679439e-01 2.60169357e-01 4.96118935e-03
-3.26221138e-01 1.96489364e-01 3.48284125e-01 4.56825465e-01
5.00061251e-02 -6.29082143e-01 -6.24500275e-01 -5.38724780e-01
1.65614888e-01 7.79164433e-01 6.13552213e-01 -1.71484008e-01
-2.13827342e-02 4.79621828e-01 -1.75680494e+00 1.89587604e-02
3.62031996e-01 1.03200829e+00 1.63744930e-02 7.94484437e-01
2.67844707e-01 2.70270556e-01 -4.32248980e-01 -2.91003078e-01
5.18975437e-01 -3.18959951e-01 -4.22491044e-01 6.27152383e-01
1.05489910e+00 -6.57740116e-01 -2.98670530e-01 8.50251198e-01
6.16055548e-01 -3.33485156e-02 3.75177026e-01 -9.91101921e-01
6.05640095e-03 -2.84023613e-01 -9.37335730e-01 -9.55132209e-03
1.03834128e+00 8.10926035e-02 7.52309322e-01 -6.96784079e-01
5.34256399e-01 1.59525633e+00 6.65109575e-01 3.64394128e-01
-1.10709751e+00 -5.62164307e-01 4.61127430e-01 -1.54510066e-01
-1.40034628e+00 -2.23621756e-01 8.75988126e-01 -5.68746209e-01
6.50866985e-01 1.16089940e-01 1.02238202e+00 9.16662395e-01
9.90276132e-03 6.09524369e-01 1.21231747e+00 -4.35057342e-01
7.86529109e-02 -2.38325030e-01 -1.13718957e-01 8.91642928e-01
5.48168540e-01 -2.00256810e-01 -8.40553761e-01 5.96810356e-02
1.08531761e+00 3.02987307e-01 -6.56367242e-02 -8.99652243e-01
-1.14204168e+00 4.44320410e-01 4.92946863e-01 -1.16278484e-01
-3.51972342e-01 4.54208553e-02 4.88070957e-02 -1.89122245e-01
5.32945991e-01 4.42156456e-02 -4.70959723e-01 -5.98887913e-02
-9.00225282e-01 5.17882168e-01 4.34359670e-01 8.58161807e-01
7.89262295e-01 -3.52063924e-01 1.47162348e-01 4.95538622e-01
4.26143110e-01 7.04990864e-01 9.16534737e-02 -9.58956957e-01
5.90506434e-01 8.47910702e-01 2.10823119e-01 -1.11070597e+00
-6.36852264e-01 -2.70563126e-01 -2.82858729e-01 1.91843450e-01
6.88253880e-01 -7.36028105e-02 -1.10886121e+00 1.65329397e+00
1.07441020e+00 -2.81762838e-01 -3.79744381e-01 1.33372796e+00
8.01880002e-01 1.29521132e-01 -2.08064869e-01 1.51554972e-01
1.67107725e+00 -7.64899611e-01 -4.80009586e-01 -8.83383214e-01
1.55249313e-01 -6.36593521e-01 6.67532623e-01 4.39019263e-01
-9.69813883e-01 -6.15364969e-01 -1.04110563e+00 -2.90036857e-01
-1.96978241e-01 2.96055585e-01 4.69884664e-01 7.07077801e-01
-6.08195901e-01 3.62737656e-01 -1.07890248e+00 -7.90986955e-01
3.01993806e-02 4.16982561e-01 -5.75352788e-01 -2.73105502e-01
-7.12192178e-01 7.64119208e-01 2.52125084e-01 2.41024271e-01
-4.88558590e-01 -3.54346484e-01 -1.11645186e+00 -3.44788253e-01
8.93228292e-01 -9.71197724e-01 1.14405096e+00 -5.82198739e-01
-1.33642566e+00 1.02839792e+00 -1.82535306e-01 6.01275973e-02
6.88811362e-01 -7.13025630e-01 1.80157144e-02 4.49818939e-01
8.26678500e-02 5.84648788e-01 8.22437465e-01 -1.27373719e+00
-6.45112276e-01 -1.22189999e+00 2.61786908e-01 7.30481088e-01
-7.17660934e-02 -2.36040846e-01 -1.00368762e+00 -3.68423700e-01
8.43872845e-01 -1.13647199e+00 -1.54120266e-01 4.44139779e-01
-3.75781953e-01 2.02363968e-01 5.11522532e-01 -8.49466681e-01
7.34255493e-01 -1.87037694e+00 6.77243531e-01 -4.52582203e-02
8.04962516e-02 -1.48200184e-01 2.15714604e-01 8.52211565e-02
2.75649607e-01 -5.34450412e-01 7.69290477e-02 -7.75059998e-01
-9.74580720e-02 1.54934525e-01 3.99931937e-01 8.13736677e-01
1.89214051e-02 5.08407056e-01 -8.79920244e-01 -6.38096988e-01
5.44435143e-01 6.48835182e-01 -6.53096259e-01 3.97785276e-01
-5.36929220e-02 5.79180002e-01 -4.17873323e-01 1.07956791e+00
7.59539843e-01 -1.35058286e-02 3.06691498e-01 -4.90186781e-01
-7.01585785e-02 5.84490784e-02 -1.62425673e+00 2.29269242e+00
-8.01197588e-02 2.04462916e-01 2.17329353e-01 -5.02914011e-01
6.71281934e-01 1.10605083e-01 6.63603544e-01 -4.21967894e-01
1.50649980e-01 1.56305477e-01 -4.18766320e-01 -4.09605950e-01
4.68639582e-01 -1.87403902e-01 -3.63788396e-01 5.29673025e-02
1.72586814e-01 -1.68804854e-01 -3.00210807e-03 -4.42098789e-02
7.85804093e-01 6.73717856e-01 4.71198589e-01 1.44750893e-01
2.91499883e-01 -1.93516701e-01 6.30163133e-01 4.59464967e-01
-1.96852833e-01 1.02814269e+00 2.30716765e-01 -5.26370287e-01
-8.03914011e-01 -1.25049376e+00 2.71223038e-01 7.84201264e-01
4.79745030e-01 -6.07662976e-01 -6.40220642e-01 -5.82357228e-01
1.42544299e-01 1.95019245e-02 -8.40961277e-01 3.34226787e-02
-4.43529606e-01 -5.30922830e-01 6.18555248e-02 5.63090265e-01
5.34912050e-01 -3.82090658e-01 -1.03729284e+00 -3.16010788e-02
-5.25815308e-01 -1.45306802e+00 -5.11301398e-01 -5.80428466e-02
-9.25088584e-01 -1.24381042e+00 -6.40361011e-01 -3.77584964e-01
6.65141523e-01 5.60176373e-01 9.34674263e-01 -2.00576752e-01
-5.01245141e-01 5.63933372e-01 -3.50407660e-01 -1.34717971e-01
2.99647778e-01 -1.91668674e-01 3.82619917e-01 1.13181263e-01
1.73271850e-01 -5.47054708e-01 -8.68188977e-01 3.75703305e-01
-3.49135071e-01 1.72811568e-01 6.51060283e-01 3.44434828e-01
8.49135816e-01 -8.19358304e-02 -2.24164918e-01 -5.40099323e-01
-2.86519289e-01 1.03620708e-01 -4.74217117e-01 1.81793392e-01
-5.30699752e-02 -1.78129286e-01 3.14748921e-02 -3.90697569e-01
-1.10946238e+00 6.83181882e-01 1.75943509e-01 -5.60191870e-01
-4.16454405e-01 -1.12092771e-01 -5.53386152e-01 1.80086214e-02
4.94592696e-01 5.74943125e-02 -7.78995827e-02 -7.23984838e-01
3.46630722e-01 4.06159967e-01 8.12242031e-01 -5.69075227e-01
7.46403813e-01 8.65532398e-01 2.04810187e-01 -8.84100318e-01
-1.22727883e+00 -8.37513328e-01 -1.40341914e+00 -5.22075415e-01
1.20782709e+00 -1.40462339e+00 -6.51907563e-01 6.39251232e-01
-1.04470551e+00 2.92640030e-02 -1.29984238e-03 6.15989029e-01
-4.05954003e-01 4.12722737e-01 -4.38075811e-01 -9.76945162e-01
-1.63008094e-01 -1.08181477e+00 1.75249600e+00 2.28307083e-01
-2.72808850e-01 -5.33246040e-01 2.46776342e-02 7.66107619e-01
-2.99534768e-01 6.90748394e-01 2.03517213e-01 7.45543372e-03
-5.38455486e-01 -4.29020703e-01 2.97734607e-03 -6.28195098e-03
9.85322297e-02 -3.99566263e-01 -1.00433433e+00 -2.57586986e-01
-7.05456957e-02 -2.45403126e-01 6.15912318e-01 4.30210859e-01
6.53284788e-01 5.74291013e-02 -1.96821272e-01 7.34554946e-01
1.30945623e+00 -3.95788699e-01 2.70571679e-01 3.73963147e-01
1.01702166e+00 8.03411663e-01 8.52115571e-01 6.66319489e-01
6.45283997e-01 1.03474760e+00 6.16333723e-01 -9.52291787e-02
-2.84396559e-01 -4.64767128e-01 3.29668969e-01 4.31185335e-01
-3.88900369e-01 2.05107033e-01 -6.91664100e-01 2.41915956e-01
-1.85527551e+00 -6.90715551e-01 -7.61423633e-02 2.33362889e+00
4.64754283e-01 2.24111915e-01 5.62796474e-01 2.62344368e-02
6.44488931e-01 1.14353113e-01 -6.26148164e-01 2.49673650e-01
8.92994925e-02 -2.15321496e-01 5.95649779e-01 3.34037989e-01
-1.10326695e+00 6.88926578e-01 5.72524929e+00 2.90076494e-01
-8.58470857e-01 -1.31162524e-01 1.14124432e-01 -3.74016821e-01
1.91745117e-01 -9.47086290e-02 -9.78161216e-01 1.22282058e-01
2.47819588e-01 5.80434561e-01 2.26508945e-01 7.94367731e-01
5.65263070e-03 -6.00021601e-01 -1.17620420e+00 1.28633738e+00
4.30967361e-01 -5.90721548e-01 -1.65454090e-01 2.65807301e-01
5.53327799e-01 -2.34519109e-01 -3.14455956e-01 -1.11130355e-02
-1.95536554e-01 -4.75580812e-01 1.13965786e+00 5.60758352e-01
7.01690257e-01 -6.73097014e-01 5.62275350e-01 6.06579602e-01
-1.44190621e+00 3.71703841e-02 -1.33490548e-01 -3.39929879e-01
2.15795115e-01 4.78248805e-01 -6.64214730e-01 7.12504983e-01
9.59838450e-01 6.38310254e-01 -8.16285491e-01 8.33607972e-01
-2.63536662e-01 -2.80822199e-02 -6.20103121e-01 2.90568650e-01
-3.05300206e-01 -1.28743440e-01 4.88953710e-01 1.06124532e+00
5.15377700e-01 2.81573355e-01 4.74715859e-01 4.15056944e-01
2.96037912e-01 -2.25253090e-01 -3.07457089e-01 3.34446430e-01
1.70886070e-01 1.25335431e+00 -8.85230780e-01 -2.56687790e-01
-4.10618186e-01 1.23720944e+00 1.95719704e-01 6.09560832e-02
-7.54037797e-01 7.23707154e-02 7.47011483e-01 4.68371034e-01
5.46349704e-01 -5.05615592e-01 -1.79094076e-01 -1.45790207e+00
2.37407640e-01 -7.23001420e-01 5.09637833e-01 -9.57255304e-01
-9.46816266e-01 3.57768774e-01 2.98677087e-01 -1.38037837e+00
-3.11386645e-01 -6.96106434e-01 -2.95458287e-02 5.59446871e-01
-1.16548288e+00 -1.36511183e+00 -6.30990744e-01 6.29579663e-01
6.52234793e-01 3.74276668e-01 7.66754925e-01 1.31618962e-01
-4.24467564e-01 2.85850406e-01 -4.76404786e-01 1.76546633e-01
8.43151808e-01 -1.28103018e+00 1.53332412e-01 9.29008901e-01
1.61028296e-01 4.29716349e-01 8.80698621e-01 -8.06698799e-01
-1.71489000e+00 -6.70553148e-01 7.23279834e-01 -8.10464084e-01
1.06665939e-01 -6.99957609e-01 -3.30010921e-01 5.94570100e-01
-3.75409305e-01 1.52903870e-01 4.84645367e-01 3.45368505e-01
-4.66489226e-01 -3.37088883e-01 -1.18291128e+00 3.40115994e-01
1.34895945e+00 -4.67574626e-01 -6.42061412e-01 2.13459954e-01
4.56894130e-01 -8.96596014e-01 -7.52131164e-01 3.19892883e-01
9.68932092e-01 -1.16787624e+00 1.34826612e+00 -9.16486010e-02
1.40572309e-01 -5.84534049e-01 -4.96436983e-01 -9.72615719e-01
-1.23734005e-01 -2.10822895e-01 -2.83398271e-01 9.89993989e-01
-2.21801743e-01 -1.22174852e-01 9.54409897e-01 6.28293753e-01
1.60613105e-01 -4.20229703e-01 -9.27991927e-01 -5.86611807e-01
-6.30340993e-01 -4.81025577e-01 3.82166415e-01 4.24821407e-01
-3.56017977e-01 4.21554357e-01 -5.85655749e-01 6.04123712e-01
1.08190358e+00 4.28902239e-01 1.22691262e+00 -1.31527257e+00
-5.38481414e-01 1.60024121e-01 -8.41889203e-01 -1.26562643e+00
-4.21945393e-01 -2.60494143e-01 1.55616865e-01 -1.49760461e+00
3.24739903e-01 1.70669124e-01 1.23160616e-01 3.08982581e-01
-2.43634239e-01 5.30480683e-01 3.99475873e-01 1.41399354e-01
-8.41818273e-01 3.52196723e-01 1.20738947e+00 2.08987221e-01
3.16753192e-03 -7.88653716e-02 -5.33982456e-01 1.02280951e+00
3.33496690e-01 -3.71836811e-01 -2.19326958e-01 -5.66680253e-01
3.52256522e-02 1.81978986e-01 6.27309680e-01 -1.40878189e+00
1.37880996e-01 -1.25843287e-01 9.52140450e-01 -1.03878677e+00
9.73661005e-01 -7.70049393e-01 1.70994848e-01 4.49650556e-01
9.82757360e-02 -7.22506493e-02 3.67318578e-02 8.59133661e-01
1.02343224e-01 2.85452604e-01 5.86981595e-01 -6.10565305e-01
-7.54710197e-01 3.71161073e-01 2.49813750e-01 -2.11636573e-02
8.30597997e-01 -4.89933938e-01 1.37467563e-01 -3.16041708e-01
-7.89781690e-01 2.08043382e-01 8.13831747e-01 4.37656641e-01
6.36959016e-01 -1.30067205e+00 -5.25204539e-01 3.60412687e-01
2.43110523e-01 3.75337839e-01 5.01876235e-01 7.94210672e-01
-4.94560719e-01 2.56874949e-01 -2.60988146e-01 -1.13909197e+00
-1.68374503e+00 3.48865598e-01 3.89486492e-01 -6.92038313e-02
-6.78446829e-01 9.19341385e-01 3.75419945e-01 -5.23304522e-01
1.87766269e-01 -3.04805011e-01 5.55414855e-02 1.62946030e-01
5.24727345e-01 2.55964518e-01 3.27704437e-02 -1.05798888e+00
-6.22242451e-01 1.27138841e+00 1.89295650e-01 -2.63177484e-01
1.30550647e+00 -5.80549300e-01 1.22438952e-01 5.81677496e-01
1.09171128e+00 7.85128623e-02 -1.73284936e+00 -4.13285375e-01
-5.34032583e-01 -9.60853577e-01 -1.37550786e-01 -6.36783242e-01
-8.68234575e-01 9.50361192e-01 6.33296490e-01 -3.01163167e-01
1.08621609e+00 2.28896990e-01 6.47226989e-01 1.31208092e-01
6.11888826e-01 -1.25170088e+00 3.87182236e-01 1.72983631e-01
6.97568119e-01 -1.22596502e+00 6.17594957e-01 -6.73383594e-01
-4.61957753e-01 1.11877489e+00 8.22266638e-01 -4.30483557e-02
2.86444426e-01 -6.19211793e-02 -4.02233265e-02 -2.57387400e-01
-2.17119545e-01 -4.70441848e-01 6.04044616e-01 5.88712096e-01
2.55077720e-01 1.99193791e-01 1.19854525e-01 4.03868198e-01
-1.52649388e-01 -1.69594169e-01 2.52700269e-01 1.10847330e+00
-3.87535632e-01 -9.98322606e-01 -8.86577368e-01 -6.59712590e-04
-2.27198347e-01 5.36343396e-01 -5.55964410e-01 9.16153371e-01
5.66430092e-01 8.24894011e-01 1.82597861e-02 -5.32601178e-01
7.32547045e-01 8.72404724e-02 1.12351096e+00 -7.82564402e-01
-1.40311688e-01 6.24703109e-01 6.76189065e-02 -8.27463567e-01
-7.11990416e-01 -1.01914990e+00 -1.03822529e+00 -2.63465028e-02
-2.17344776e-01 -4.87855673e-01 7.94895291e-01 9.61933613e-01
1.44858524e-01 3.73820662e-01 2.82565296e-01 -1.44007528e+00
-3.60668659e-01 -8.16217899e-01 -7.44027793e-01 5.72863281e-01
4.79737729e-01 -1.00870085e+00 -1.80093527e-01 1.41521776e-02] | [7.077182769775391, -1.0500280857086182] |
1ae2f7c7-ea3c-430c-9a33-f3f3f1c50c4a | inno-at-semeval-2020-task-11-leveraging-pure | 2008.11584 | null | https://arxiv.org/abs/2008.11584v2 | https://arxiv.org/pdf/2008.11584v2.pdf | Inno at SemEval-2020 Task 11: Leveraging Pure Transformer for Multi-Class Propaganda Detection | The paper presents the solution of team "Inno" to a SEMEVAL 2020 task 11 "Detection of propaganda techniques in news articles". The goal of the second subtask is to classify textual segments that correspond to one of the 18 given propaganda techniques in news articles dataset. We tested a pure Transformer-based model with an optimized learning scheme on the ability to distinguish propaganda techniques between each other. Our model showed 0.6 and 0.58 overall F1 score on validation set and test set accordingly and non-zero F1 score on each class on both sets. | ['Vladimir Ivanov', 'Dmitry Grigorev'] | 2020-08-26 | null | null | null | null | ['propaganda-detection'] | ['natural-language-processing'] | [ 8.09833333e-02 4.14020531e-02 -3.90467912e-01 -1.16928697e-01
-8.56004179e-01 -7.28214502e-01 1.51648510e+00 2.33745128e-01
-4.88644928e-01 5.42597473e-01 6.14733756e-01 -6.51515603e-01
-2.03974858e-01 -6.08465910e-01 -5.55490911e-01 -3.47088009e-01
2.18069088e-02 5.20603657e-01 2.61905551e-01 -5.52818000e-01
1.04935884e+00 -1.03522941e-01 -1.19360495e+00 1.14613307e+00
7.03409970e-01 8.75718355e-01 -1.67374700e-01 9.21840966e-01
-7.86998570e-02 1.95322359e+00 -1.31107020e+00 -4.47087914e-01
2.30683342e-01 -5.14040053e-01 -1.29199708e+00 -3.53256255e-01
1.01975286e+00 9.07905027e-02 -3.56490880e-01 1.18784606e+00
2.08266139e-01 -1.60504580e-01 9.54856515e-01 -8.09203804e-01
-4.96282607e-01 1.30445802e+00 -4.73471075e-01 8.17716479e-01
8.31189096e-01 -2.89027840e-01 7.75486946e-01 -5.59248865e-01
1.10242712e+00 1.16182768e+00 7.70883739e-01 4.29119349e-01
-1.13797534e+00 -5.95585287e-01 -3.87313902e-01 3.79280657e-01
-8.16023529e-01 -2.63156146e-01 5.75319707e-01 -1.23443913e+00
1.09909987e+00 4.21557754e-01 6.22546494e-01 1.72354209e+00
1.00324512e+00 6.85694695e-01 1.56885421e+00 -4.24126267e-01
1.86844096e-02 4.96656001e-01 5.79036534e-01 8.97399604e-01
-3.78769264e-02 5.42068444e-02 -8.57511163e-01 -1.74472630e-01
7.55778477e-02 -6.82611108e-01 1.13808684e-01 5.82875431e-01
-1.25083518e+00 1.17812705e+00 1.62098557e-01 5.64020336e-01
-2.25450084e-01 -1.16913497e-01 1.03622270e+00 7.53628910e-01
8.56007934e-01 7.68288851e-01 -1.55396894e-01 -3.03909957e-01
-1.05495644e+00 6.66994691e-01 9.11815047e-01 5.35192072e-01
-5.75734675e-03 1.62299916e-01 -5.26671052e-01 6.17739677e-01
-3.12351763e-01 6.83368087e-01 5.46901762e-01 -4.40067828e-01
8.23211849e-01 5.05375326e-01 1.87696666e-02 -1.29362679e+00
-5.71619034e-01 -7.69540012e-01 -3.54993612e-01 -1.01661019e-01
4.54968095e-01 -2.92750031e-01 -1.04689741e+00 1.21054459e+00
-3.19333635e-02 -4.59098369e-01 -6.87560216e-02 4.82861906e-01
1.07166612e+00 8.23967755e-01 -2.01418534e-01 -3.39217365e-01
1.34445238e+00 -1.01860142e+00 -9.21130836e-01 -2.75050253e-01
9.42664504e-01 -1.09603000e+00 8.01590860e-01 8.07566106e-01
-5.57168782e-01 -3.65853578e-01 -9.74800944e-01 1.32102445e-01
-1.55915886e-01 2.06237167e-01 2.89342046e-01 5.99102497e-01
-1.87377080e-01 3.98840785e-01 -1.69845268e-01 -3.00937563e-01
1.33050591e-01 -3.37406874e-01 -1.75486952e-01 4.50713366e-01
-1.21486497e+00 1.35490263e+00 4.99151260e-01 -5.37800431e-01
-1.55142927e+00 -7.27482319e-01 -3.82973909e-01 -3.93800229e-01
4.89419818e-01 -3.52979675e-02 1.48693192e+00 -1.10594320e+00
-8.80478501e-01 1.21649444e+00 6.12222493e-01 -8.72688472e-01
7.39713609e-01 -6.42080367e-01 -8.15149724e-01 4.21003848e-02
6.01570547e-01 -1.73282117e-01 9.86159503e-01 -8.25423419e-01
-9.69297409e-01 -1.24655671e-01 3.50080431e-02 -2.72023201e-01
-2.25158438e-01 7.35495687e-01 7.61271596e-01 -8.26559961e-01
-2.36413211e-01 -8.15726757e-01 6.00515962e-01 -1.01237202e+00
-7.49949694e-01 -5.08000612e-01 8.65313470e-01 -1.08885527e+00
1.60377967e+00 -1.65431404e+00 1.02096193e-01 7.93220401e-02
3.42513174e-01 1.34330407e-01 1.51317090e-01 6.16485000e-01
5.85486442e-02 2.18344986e-01 5.37949860e-01 4.86316562e-01
-2.78363764e-01 -3.12969178e-01 -6.98736548e-01 8.42251837e-01
-2.96193093e-01 2.37643600e-01 -1.02225292e+00 -5.56406319e-01
-2.01273903e-01 -1.33485675e-01 -9.19014141e-02 6.98198006e-03
-1.98515266e-01 -1.94869551e-03 -5.09899020e-01 3.52398962e-01
1.09087124e-01 1.39740020e-01 1.83203742e-01 -4.29809578e-02
-4.80906725e-01 8.81916642e-01 -3.85142237e-01 1.12254381e+00
-1.94943354e-01 1.19006050e+00 -1.02546960e-01 -8.76150727e-01
8.86914849e-01 2.50762969e-01 -7.43316812e-03 -8.44394505e-01
6.14366174e-01 2.15263411e-01 3.36590201e-01 -7.82886565e-01
4.60904896e-01 -3.18563163e-01 -5.27541637e-01 4.64340210e-01
1.82267070e-01 1.65796265e-01 4.91950303e-01 5.49338698e-01
1.39157081e+00 -2.21117556e-01 4.35173601e-01 -8.41686785e-01
5.59904933e-01 6.61856294e-01 2.86455363e-01 1.11541450e+00
-6.07116446e-02 4.55414392e-02 7.06112683e-01 -8.90335202e-01
-8.92473221e-01 -7.03894258e-01 -1.45759508e-01 1.53937125e+00
-3.72444034e-01 -7.23860681e-01 -5.55348873e-01 -1.39214098e+00
-4.36771482e-01 1.25971258e+00 -1.15272760e+00 4.43073399e-02
-6.87087119e-01 -4.17754292e-01 1.07697213e+00 5.63933402e-02
4.79144365e-01 -8.24484408e-01 -9.06683147e-01 1.01630881e-01
-5.85189402e-01 -7.64756501e-01 -1.29519895e-01 3.77449751e-01
-3.47461849e-01 -1.33589816e+00 -2.85760611e-01 -8.79466772e-01
-1.74672455e-02 1.56784847e-01 1.00194943e+00 -1.84589878e-01
-1.00895554e-01 -9.30848867e-02 -8.00233424e-01 -6.44762039e-01
-1.00990367e+00 2.01511234e-01 9.06047970e-02 -3.18303645e-01
2.71965176e-01 1.39717907e-01 1.58036962e-01 1.04334131e-01
-2.60788888e-01 2.11790353e-01 1.59834698e-01 7.93464243e-01
-2.58569628e-01 1.42551094e-01 3.44040483e-01 -9.59830523e-01
1.00035739e+00 -6.20590091e-01 -2.75327325e-01 1.79483458e-01
-5.40927827e-01 -1.47485808e-01 7.85230875e-01 -5.65370321e-01
-9.30709183e-01 -3.42887968e-01 4.63175327e-02 1.26520529e-01
3.29400808e-01 6.83630228e-01 6.81395113e-01 3.68344575e-01
1.60578704e+00 2.55218089e-01 -2.95183361e-01 -6.47399545e-01
-1.74651772e-01 8.42362225e-01 5.87846756e-01 -4.56733227e-01
5.49347639e-01 1.23058841e-01 -5.48092365e-01 -7.12588131e-01
-1.53437948e+00 -5.11228383e-01 -2.47056738e-01 -5.71846128e-01
7.33417571e-01 -6.08689845e-01 -3.06665391e-01 6.42112017e-01
-1.38722193e+00 -8.72615352e-02 1.25938535e-01 4.48454171e-01
-3.90290409e-01 1.05023205e-01 -6.21549726e-01 -9.05277491e-01
-8.45303178e-01 -6.80163026e-01 6.25820339e-01 -1.97395101e-01
-4.97852474e-01 -7.88400412e-01 4.16660547e-01 6.24287009e-01
1.08803734e-01 2.92886645e-01 8.77093256e-01 -1.51239002e+00
6.81924880e-01 -4.94603157e-01 8.71533826e-02 2.52098471e-01
-1.38021946e-01 1.07495807e-01 -6.65444255e-01 -1.29995167e-01
5.21923363e-01 -6.71982825e-01 1.06255496e+00 1.13092519e-01
4.23647970e-01 -9.77647424e-01 -3.76146674e-01 -1.74363792e-01
9.91837025e-01 1.70716599e-01 3.86500388e-01 9.72151935e-01
5.22007704e-01 6.58818245e-01 7.54191518e-01 3.84218842e-01
-2.44654745e-01 7.24897325e-01 1.87197909e-01 6.99840844e-01
-2.03056619e-01 -4.50655818e-01 8.44719172e-01 5.52464545e-01
-1.30642414e-01 -4.00599450e-01 -1.42414892e+00 6.44171298e-01
-1.54640687e+00 -1.75376451e+00 -8.10244322e-01 1.67825508e+00
7.73984909e-01 5.43623626e-01 4.11080241e-01 3.85167480e-01
7.85539806e-01 3.88200849e-01 4.12118524e-01 -8.24147463e-01
-1.10177115e-01 1.10013142e-01 3.07342023e-01 5.11283219e-01
-1.62661016e+00 8.77209425e-01 7.14049911e+00 1.36621392e+00
-1.08696020e+00 3.77535045e-01 1.60328075e-01 -2.96416670e-01
1.92916337e-02 -1.56568378e-01 -8.92662704e-01 6.19933844e-01
1.28858435e+00 -3.11581761e-01 1.00446627e-01 9.48983669e-01
8.09911266e-02 -1.08759992e-01 -7.81870246e-01 3.76140803e-01
4.65993494e-01 -1.63727772e+00 2.13285416e-01 1.87289130e-04
8.47414613e-01 2.21767172e-01 -2.46314123e-01 6.62887931e-01
4.30597037e-01 -1.12398076e+00 1.37969160e+00 1.96519378e-03
4.12850261e-01 -5.34576774e-01 8.38146329e-01 1.06051075e+00
-5.27000070e-01 -3.17779034e-01 3.35931592e-03 -3.87077898e-01
1.54892862e-01 4.24792051e-01 -1.13483834e+00 3.18713099e-01
6.22952759e-01 6.29514456e-01 -7.58319438e-01 5.56828082e-01
-5.58523059e-01 1.18757784e+00 -8.04526731e-02 -3.10180336e-01
6.22542322e-01 4.77736741e-01 1.02059817e+00 1.65480435e+00
6.84726834e-02 -3.29881966e-01 3.52889419e-01 3.31481487e-01
3.29678416e-01 2.27152452e-01 -6.04106009e-01 -2.42762923e-01
4.33062881e-01 9.19723570e-01 -6.35301113e-01 -6.20368540e-01
2.30758488e-01 3.79359424e-01 9.86162797e-02 -2.79181689e-01
-1.09537399e+00 -3.04900050e-01 -1.68452084e-01 6.17960393e-01
-1.35842124e-02 4.85860258e-02 -3.16502601e-01 -8.92377555e-01
-4.18096930e-01 -1.32523990e+00 7.40113318e-01 -3.51450950e-01
-1.21220410e+00 8.87482882e-01 1.68808922e-01 -9.54459250e-01
-3.07306379e-01 -6.21180594e-01 -6.67536676e-01 3.28966945e-01
-4.21471089e-01 -1.27732384e+00 -6.53144270e-02 3.37873548e-01
1.09176075e+00 -8.17291379e-01 5.04545331e-01 6.89927116e-02
-3.08673084e-01 3.48455340e-01 -1.56701189e-02 4.99494255e-01
7.46820331e-01 -1.02611530e+00 1.26513764e-01 9.65973258e-01
2.00650543e-01 5.35324156e-01 1.47598290e+00 -1.05167580e+00
-1.02258229e+00 -9.27589834e-01 1.24495101e+00 -7.32240140e-01
1.23270392e+00 -7.70444497e-02 -5.10226846e-01 6.87950194e-01
4.15284961e-01 -8.82543743e-01 5.06155491e-01 5.30656278e-01
-1.01215792e+00 5.53663075e-01 -9.05502200e-01 7.31570199e-02
8.50239277e-01 -5.63830316e-01 -1.47343838e+00 9.69281077e-01
6.38115704e-02 -3.95535976e-01 -6.23361051e-01 2.13158950e-01
7.18870342e-01 -8.12158406e-01 5.58129013e-01 -1.15089047e+00
1.32674611e+00 9.75896493e-02 -2.60798365e-01 -1.07360995e+00
-3.68060708e-01 -3.99182379e-01 -1.14930980e-01 9.86028016e-01
5.57405889e-01 -1.46155044e-01 2.37207487e-01 -6.01929784e-01
-2.19281375e-01 -3.41626525e-01 -1.06034017e+00 -1.09033787e+00
2.08625883e-01 -4.78264868e-01 -4.94564891e-01 1.29943419e+00
4.39504653e-01 1.09983516e+00 -8.35652649e-01 -3.56473953e-01
3.91705960e-01 1.09314941e-01 7.38537014e-01 -1.12552750e+00
-3.10542434e-01 -6.19601309e-01 -3.34943354e-01 -6.41549468e-01
4.65912297e-02 -1.06901896e+00 1.14558458e-01 -1.50562108e+00
6.02212071e-01 1.17552869e-01 8.72864500e-02 4.02297884e-01
3.27773601e-01 2.61602253e-01 1.07662799e-02 3.98875684e-01
-4.45777237e-01 -4.79048081e-02 6.69547617e-01 -6.00491941e-01
3.85588370e-02 -1.18053257e-01 -5.04359961e-01 8.89242113e-01
8.76815915e-01 -9.72624898e-01 -2.80075759e-01 -4.76554781e-01
5.60145557e-01 -1.48705825e-01 3.56717885e-01 -8.94463181e-01
-3.69980931e-02 -3.96555424e-01 -9.48407948e-02 -7.31493175e-01
-1.53742447e-01 1.74817722e-02 1.42312899e-01 1.00386310e+00
-6.61367774e-01 -7.32745826e-02 2.77506579e-02 5.24965107e-01
4.87522297e-02 -5.32856286e-01 7.63267457e-01 -1.09477527e-01
-4.57298219e-01 -7.31385827e-01 -1.05851233e+00 3.22110772e-01
9.83503759e-01 1.78145897e-02 -1.09633720e+00 -2.74312124e-02
-5.79657853e-01 -1.94156140e-01 5.21109551e-02 6.10835671e-01
1.37172818e-01 -9.39797282e-01 -1.37060583e+00 -3.49033445e-01
1.12741500e-01 -9.73341644e-01 1.12396218e-02 1.08249176e+00
-8.20062578e-01 3.37343961e-01 -4.59411740e-01 -3.18940282e-01
-1.78490222e+00 4.77929622e-01 3.42992812e-01 -6.08030498e-01
-7.15466797e-01 1.04030943e+00 -1.86984167e-01 -6.13121956e-04
-7.71268979e-02 2.72593856e-01 -5.44296920e-01 4.82154787e-01
9.76736665e-01 8.32237959e-01 6.60627410e-02 -7.96882808e-01
-6.11377954e-01 1.03826625e-02 -7.36710489e-01 -1.09182619e-01
1.17668724e+00 5.57591200e-01 -1.38874531e-01 5.77073812e-01
9.16763127e-01 4.40218300e-01 -2.04847520e-03 -9.98167694e-03
2.52754331e-01 -2.95143366e-01 3.50882530e-01 -1.37888277e+00
-4.19783920e-01 5.01555920e-01 4.26163137e-01 8.56085658e-01
3.59509438e-01 1.78674355e-01 4.75088358e-01 4.76968884e-01
2.18337491e-01 -1.62781596e+00 2.10872233e-01 1.02071655e+00
1.32367730e+00 -7.88854420e-01 3.56325567e-01 -3.70684475e-01
-5.92751145e-01 1.01980460e+00 3.53260100e-01 -4.65029180e-01
1.49480894e-01 2.36550555e-01 1.09475128e-01 -9.03538465e-01
-9.69175577e-01 2.24173322e-01 5.46779275e-01 3.20837528e-01
6.75371408e-01 1.37786165e-01 -1.23253381e+00 4.29776460e-01
-6.00807965e-01 -3.40451628e-01 5.28598845e-01 9.66523707e-01
-9.22294438e-01 -2.70397782e-01 -5.33281446e-01 8.04089963e-01
-1.01423275e+00 -5.34011312e-02 -1.24369097e+00 8.66251051e-01
1.08586811e-01 1.21519005e+00 -2.04137817e-01 -1.19037712e+00
2.07444474e-01 2.53465548e-02 5.39978087e-01 -6.98683202e-01
-1.51103401e+00 3.25843208e-02 1.14543200e+00 -1.95178136e-01
-3.53839815e-01 -6.38041973e-01 -6.83033407e-01 -8.10505509e-01
-4.35551912e-01 5.02571464e-01 4.15311366e-01 1.04160106e+00
-3.05988342e-01 5.35845757e-01 4.46435332e-01 -3.29595387e-01
-1.09035408e+00 -1.54421425e+00 -2.43016168e-01 4.48835462e-01
1.80191323e-02 -6.88157380e-01 -5.08145869e-01 1.03294089e-01] | [8.476266860961914, 10.681117057800293] |
c1793c37-7002-4089-8964-681213052484 | global-optimality-and-finite-sample-analysis | 2111.02997 | null | https://arxiv.org/abs/2111.02997v3 | https://arxiv.org/pdf/2111.02997v3.pdf | Global Optimality and Finite Sample Analysis of Softmax Off-Policy Actor Critic under State Distribution Mismatch | In this paper, we establish the global optimality and convergence rate of an off-policy actor critic algorithm in the tabular setting without using density ratio to correct the discrepancy between the state distribution of the behavior policy and that of the target policy. Our work goes beyond existing works on the optimality of policy gradient methods in that existing works use the exact policy gradient for updating the policy parameters while we use an approximate and stochastic update step. Our update step is not a gradient update because we do not use a density ratio to correct the state distribution, which aligns well with what practitioners do. Our update is approximate because we use a learned critic instead of the true value function. Our update is stochastic because at each step the update is done for only the current state action pair. Moreover, we remove several restrictive assumptions from existing works in our analysis. Central to our work is the finite sample analysis of a generic stochastic approximation algorithm with time-inhomogeneous update operators on time-inhomogeneous Markov chains, based on its uniform contraction properties. | ['Romain Laroche', 'Remi Tachet', 'Shangtong Zhang'] | 2021-11-04 | null | null | null | null | ['policy-gradient-methods'] | ['methodology'] | [-2.44701907e-01 3.19675624e-01 -6.57399058e-01 1.24227814e-01
-7.01596439e-01 -6.87196255e-01 4.99611616e-01 1.16327535e-02
-7.61799395e-01 1.20611644e+00 2.64009923e-01 -6.58353150e-01
-8.55367035e-02 -4.07465577e-01 -7.91169882e-01 -8.84757698e-01
1.74261168e-01 6.90882683e-01 2.23739132e-01 -3.09730262e-01
1.60555989e-01 2.33762741e-01 -8.73840570e-01 -3.11217517e-01
8.55006278e-01 8.86903346e-01 -1.22800246e-01 9.83290374e-01
-4.57861386e-02 8.26100230e-01 -3.87936980e-01 -1.81403995e-01
6.18329287e-01 -9.49389756e-01 -5.92643321e-01 -1.53483957e-01
5.23869656e-02 -5.79979539e-01 -2.59512961e-01 1.37557554e+00
5.35568595e-01 4.66634333e-01 6.77384973e-01 -1.08006835e+00
-2.09315509e-01 8.55269611e-01 -5.05704820e-01 2.17347264e-01
3.96867171e-02 2.03382939e-01 8.60620379e-01 -1.93934828e-01
7.01257288e-01 1.32651031e+00 8.05372059e-01 8.31313193e-01
-1.14520454e+00 -3.15467149e-01 6.76873446e-01 2.47801282e-02
-7.72445202e-01 -2.05124989e-01 5.26423573e-01 -3.43560100e-01
6.75952733e-01 1.01905823e-01 9.69535530e-01 1.05404890e+00
3.06643873e-01 6.42940700e-01 1.37769449e+00 -4.66727227e-01
7.06980526e-01 9.31165740e-02 -4.75941263e-02 7.75271714e-01
3.12309444e-01 4.54245746e-01 1.05146155e-01 -3.63308668e-01
1.09819269e+00 -5.68767637e-03 -1.94557577e-01 -5.37378252e-01
-7.89531469e-01 1.01289725e+00 -1.14628166e-01 1.84781581e-01
-7.12745965e-01 7.81123757e-01 5.42903006e-01 4.50536847e-01
6.14129543e-01 2.52321959e-01 -4.68960255e-01 -6.70134604e-01
-8.68181407e-01 4.56699520e-01 1.05948567e+00 7.02920496e-01
4.35134590e-01 4.66134489e-01 -4.98786807e-01 3.19559723e-01
2.51516730e-01 7.59840429e-01 4.39559698e-01 -1.69517756e+00
1.79918021e-01 -1.20380513e-01 8.21574271e-01 -3.60616297e-01
-1.40644193e-01 -5.30245066e-01 -2.86620319e-01 4.82639313e-01
1.00775421e+00 -7.92682171e-01 -5.92788160e-01 2.07335091e+00
4.44494009e-01 1.48401126e-01 -7.71475658e-02 9.47266221e-01
-4.60934103e-01 6.33875430e-01 -2.05670282e-01 -8.44220042e-01
8.72348964e-01 -8.98434520e-01 -9.32004273e-01 9.41548944e-02
6.30395651e-01 -4.51383024e-01 1.12333024e+00 2.84364909e-01
-1.40130651e+00 -2.02692859e-02 -7.76662648e-01 4.16314811e-01
1.95208013e-01 -1.61804721e-01 3.98615003e-01 6.15305007e-01
-1.07831228e+00 8.37916553e-01 -1.02464306e+00 -2.50262469e-01
-4.03766073e-02 -3.76640633e-02 3.37221920e-01 5.20351946e-01
-1.07689202e+00 1.20808923e+00 2.28845179e-01 -1.58231169e-01
-1.10540426e+00 -6.87139392e-01 -4.72934991e-01 -4.75261547e-02
7.85399914e-01 -8.02178264e-01 2.05259085e+00 -1.38661122e+00
-2.35639000e+00 6.58738688e-02 -2.38321349e-01 -8.18456173e-01
1.17390776e+00 -5.58538958e-02 3.09748679e-01 7.00287372e-02
-5.62872328e-02 -7.02247620e-02 8.79804850e-01 -1.13761187e+00
-7.95271754e-01 -7.17447773e-02 2.87068754e-01 4.94281858e-01
-4.00316492e-02 -3.57071757e-01 -1.03563480e-01 -4.42132622e-01
-3.98804039e-01 -1.16067135e+00 -6.40194178e-01 -1.21779270e-01
-3.64200063e-02 -1.55067831e-01 5.08847773e-01 -4.85284209e-01
1.12116694e+00 -1.76791143e+00 2.44177908e-01 2.31200501e-01
-1.80122796e-02 1.63934141e-01 1.63093433e-01 5.75070858e-01
8.18905458e-02 4.97488827e-02 -2.71916002e-01 -3.22976679e-01
2.86659151e-01 2.16409534e-01 -8.28017414e-01 9.97756958e-01
-5.09244740e-01 6.60519421e-01 -1.34421360e+00 -3.35949779e-01
1.45474479e-01 9.92072523e-02 -7.15364933e-01 -7.77145624e-02
-5.35336018e-01 4.95107681e-01 -7.11301267e-01 -1.41405568e-01
3.26415718e-01 1.28628686e-01 4.31100070e-01 2.53039867e-01
-4.63575989e-01 2.01740682e-01 -1.24444151e+00 1.32841563e+00
-6.75604880e-01 1.89381436e-01 3.74897540e-01 -1.19878936e+00
4.23434764e-01 5.36239862e-01 7.98608065e-01 -5.81800818e-01
1.76084250e-01 3.72062087e-01 -7.91465715e-02 -2.12869003e-01
3.40145022e-01 -5.18319547e-01 1.89739943e-01 8.31279814e-01
-1.36401430e-01 -1.60810426e-01 3.66319984e-01 1.39950797e-01
9.01818514e-01 4.61014181e-01 3.22026134e-01 -6.54268920e-01
2.82427877e-01 -1.43422158e-02 7.34086335e-01 1.00954914e+00
-2.93716550e-01 -3.94820236e-02 1.05138874e+00 -2.25294530e-01
-1.43502176e+00 -8.06833327e-01 9.90589559e-02 9.28651571e-01
1.70021430e-02 -2.87449807e-01 -7.27376163e-01 -8.93233895e-01
8.58521909e-02 9.29000258e-01 -7.83572257e-01 -1.65150374e-01
-6.89677238e-01 -4.73672181e-01 2.77813971e-01 4.74465013e-01
2.95769304e-01 -6.20160758e-01 -8.03657770e-01 6.12655878e-01
1.33772314e-01 -6.64240360e-01 -9.66133416e-01 -3.87830427e-03
-1.05534375e+00 -1.06293654e+00 -8.68370235e-01 -1.09764174e-01
5.98950982e-01 -3.31724077e-01 6.22649491e-01 -4.55191731e-01
5.26721120e-01 9.28562522e-01 3.92004550e-02 -5.73009729e-01
-6.51778996e-01 1.97147275e-03 2.66549200e-01 -1.02862678e-02
-4.51500386e-01 -4.01816398e-01 -5.61122954e-01 1.28066659e-01
-6.00335836e-01 -1.67610928e-01 1.36697277e-01 9.58776653e-01
5.99633455e-01 -1.63373351e-01 3.95958334e-01 -8.59563351e-01
1.11770058e+00 -4.17917162e-01 -1.19816840e+00 1.35009930e-01
-9.50083613e-01 6.16098762e-01 9.23739433e-01 -7.52724588e-01
-1.18526471e+00 3.85138462e-03 7.67021179e-02 -4.51710135e-01
4.43018079e-01 2.73226023e-01 6.08685672e-01 2.18658224e-01
5.00446200e-01 1.67639524e-01 4.08078432e-01 -4.05180275e-01
5.13781309e-01 1.00423589e-01 1.19371004e-01 -1.09090900e+00
4.32449251e-01 6.88412607e-01 1.42124683e-01 -1.97006226e-01
-7.61383832e-01 -1.08610846e-01 -9.81034487e-02 -3.32224667e-01
4.20742303e-01 -4.51789558e-01 -1.21981752e+00 2.30692133e-01
-9.90898728e-01 -9.92360711e-01 -1.06888831e+00 7.38858819e-01
-1.10841489e+00 3.56482118e-01 -7.94326961e-01 -1.59069645e+00
-1.66971043e-01 -1.13310933e+00 7.27420866e-01 1.09318189e-01
7.15036616e-02 -1.42166996e+00 6.45604014e-01 -6.33628249e-01
5.98049521e-01 1.78221092e-01 5.62469244e-01 -3.12781215e-01
-3.79474573e-02 5.68300206e-03 2.46616885e-01 3.55925381e-01
-6.29928410e-02 -6.58216253e-02 -3.42328250e-01 -5.39242446e-01
3.80466610e-01 -8.12136903e-02 7.26005852e-01 7.89053977e-01
6.92685068e-01 -8.25468183e-01 -2.93084264e-01 4.01913822e-01
1.46444798e+00 3.77160937e-01 2.65762866e-01 4.41434383e-01
4.31119293e-01 2.89315641e-01 6.79375172e-01 7.93496072e-01
1.31306350e-01 5.34184694e-01 2.35010222e-01 2.29784906e-01
2.00432390e-01 -4.86517549e-01 8.49114537e-01 4.72672105e-01
-3.41945946e-01 -1.52735831e-02 -6.15755618e-01 5.05503476e-01
-2.29679370e+00 -1.13162446e+00 1.90216735e-01 2.56476092e+00
1.10798395e+00 6.66384250e-02 5.57132781e-01 -2.74798602e-01
6.88331664e-01 -1.04416549e-01 -8.92739713e-01 -7.42244899e-01
3.40305626e-01 1.50165096e-01 1.19489563e+00 1.17866969e+00
-5.48745275e-01 8.07108104e-01 7.32197475e+00 7.96634674e-01
-1.23047006e+00 4.52347547e-01 4.22252268e-01 -4.21597213e-01
-2.43330225e-01 3.62553537e-01 -7.59314001e-01 6.71594560e-01
1.20157051e+00 -4.47773963e-01 7.40553617e-01 9.91211116e-01
6.42230034e-01 -1.26267225e-01 -7.94491649e-01 4.48149025e-01
-5.11875212e-01 -1.03896642e+00 -5.11909783e-01 2.48286411e-01
8.63072395e-01 -1.80247098e-01 1.16077781e-01 3.32300156e-01
9.72335815e-01 -5.81080377e-01 1.15373588e+00 6.57617986e-01
6.07995749e-01 -8.68319035e-01 6.00866973e-01 2.93307751e-01
-9.42639053e-01 -2.34455705e-01 -9.86130312e-02 -2.44119987e-01
5.52541554e-01 4.47162449e-01 -5.31240225e-01 1.45434216e-01
1.87461108e-01 4.65665191e-01 2.83311486e-01 9.05586541e-01
-2.56381243e-01 9.18470621e-01 -7.06580341e-01 -2.22346112e-01
5.26801109e-01 -5.72886944e-01 1.02722979e+00 9.47156966e-01
2.03077987e-01 -2.85964519e-01 3.67819577e-01 6.09316647e-01
3.52865785e-01 1.01125903e-01 -4.39093709e-01 5.94390891e-02
1.77243873e-01 7.61087775e-01 -6.04646623e-01 -5.29396892e-01
-8.00283775e-02 7.03310490e-01 2.07446858e-01 7.25244582e-01
-9.67652261e-01 -2.12456569e-01 7.47258365e-01 -3.64832953e-03
3.97702157e-01 -4.96693701e-01 5.99618182e-02 -1.06058860e+00
8.47369060e-02 -7.53491580e-01 3.39654595e-01 -5.27705364e-02
-1.02243376e+00 8.52905735e-02 4.62490916e-01 -9.24050570e-01
-8.95452142e-01 -2.07805514e-01 -5.26428401e-01 7.92292595e-01
-1.26476049e+00 -3.78766596e-01 6.91244841e-01 4.33231801e-01
3.76904666e-01 1.33942395e-01 3.50801915e-01 -8.37919395e-03
-5.80317676e-01 3.37278336e-01 7.69759774e-01 -2.13167235e-01
5.21229863e-01 -1.49518299e+00 -7.72622898e-02 8.35759640e-01
-5.16387939e-01 5.99936664e-01 1.29209840e+00 -7.35695183e-01
-1.33198798e+00 -6.27906322e-01 2.37536281e-01 -2.42114350e-01
1.03482699e+00 -9.07259360e-02 -5.70261359e-01 9.71042991e-01
2.54187137e-01 7.53474161e-02 -1.82367742e-01 -1.43432602e-01
1.22605547e-01 -8.45367238e-02 -1.04383492e+00 7.22400725e-01
8.47449660e-01 -2.50413984e-01 -1.02658212e-01 3.00056577e-01
5.88803291e-01 -6.06035888e-01 -6.50374174e-01 1.21998694e-02
6.51143551e-01 -6.47378266e-01 4.36942369e-01 -9.27548110e-01
-1.87374979e-01 -2.72684693e-01 1.03346378e-01 -1.63566768e+00
-4.79133688e-02 -1.17702389e+00 -5.75045824e-01 8.42106998e-01
2.56484568e-01 -1.25306261e+00 6.16039753e-01 5.43006063e-01
-9.47386771e-02 -8.35094452e-01 -1.22258997e+00 -1.08795762e+00
4.88421261e-01 -3.49961400e-01 2.55634755e-01 6.37281954e-01
1.19253136e-01 1.04908012e-01 -5.26306689e-01 -2.12737724e-01
7.12242723e-01 3.79921403e-03 3.70267093e-01 -5.76630712e-01
-8.17726374e-01 -7.88378417e-01 2.80217260e-01 -1.15172708e+00
3.68464649e-01 -4.93042022e-01 2.00559318e-01 -1.39174485e+00
8.76482576e-02 -4.46334153e-01 -2.11691424e-01 1.93018556e-01
4.96840514e-02 -4.53292310e-01 2.35582277e-01 1.33001521e-01
-5.48700750e-01 8.66562188e-01 1.45924735e+00 1.11459076e-01
-4.81141418e-01 2.97538191e-01 -5.30944884e-01 6.30412400e-01
8.95659804e-01 -7.85764813e-01 -6.97007954e-01 -1.54615175e-02
2.90651798e-01 3.07603806e-01 1.00320712e-01 -5.65958798e-01
3.45447473e-02 -5.82624972e-01 -3.25078517e-01 4.06824835e-02
1.07459225e-01 -6.50037408e-01 9.90315061e-03 9.37218666e-01
-6.85401022e-01 2.79505551e-01 9.39267874e-02 8.53818476e-01
3.90074193e-01 -6.46596253e-01 9.28712606e-01 -1.94335327e-01
-1.46677226e-01 3.00272673e-01 -7.80149341e-01 3.30832303e-01
8.79492342e-01 3.14025909e-01 -4.93065231e-02 -9.49714661e-01
-7.07237184e-01 1.31819442e-01 6.76457942e-01 -2.97311217e-01
3.18417661e-02 -1.21524322e+00 -4.69150692e-01 -4.60835189e-01
-6.43344820e-01 -3.36906731e-01 -1.27432629e-01 1.09035492e+00
-2.62060583e-01 3.18498045e-01 6.20167963e-02 -1.27135843e-01
-5.71457386e-01 6.26052201e-01 7.64937460e-01 -5.64509094e-01
-6.51693583e-01 2.19204366e-01 -1.54778615e-01 -9.85851288e-02
2.11548299e-01 -4.77157742e-01 1.88721120e-01 -1.62977889e-01
4.00626540e-01 6.54530406e-01 -4.44723010e-01 -1.32516414e-01
6.03685081e-02 3.96460444e-01 8.64078701e-02 -1.08453810e+00
9.51267183e-01 -8.86107311e-02 1.90818951e-01 8.57768834e-01
8.41638088e-01 1.64614171e-01 -1.77213705e+00 -5.61743826e-02
-2.42514625e-01 -1.25154048e-01 1.57644317e-01 -4.78343487e-01
-1.08379340e+00 4.11258698e-01 6.08588934e-01 3.48317683e-01
6.29304171e-01 -3.04690599e-01 6.42472386e-01 3.53066236e-01
3.20937842e-01 -1.60809290e+00 -2.93823838e-01 7.40408003e-01
7.00210810e-01 -6.98734939e-01 9.95870829e-02 2.33879656e-01
-6.79071188e-01 8.55888844e-01 3.65826070e-01 -5.07908642e-01
6.08961582e-01 2.50325978e-01 -2.08806396e-01 3.22212607e-01
-8.13550293e-01 -3.67451340e-01 -2.05791131e-01 4.39013392e-01
2.72654593e-01 -9.02823173e-03 -1.07760262e+00 4.80613224e-02
7.12066665e-02 3.23645361e-02 5.84886372e-01 1.03429818e+00
-3.24228317e-01 -1.38667607e+00 -2.50771910e-01 2.37351209e-01
-6.31432414e-01 8.29619765e-02 4.03422453e-02 8.31547678e-01
-6.20721638e-01 5.79637349e-01 -3.73157323e-03 1.95486337e-01
2.64251798e-01 1.92815259e-01 7.49913096e-01 -2.45582461e-01
-3.71753603e-01 1.07889116e-01 -1.60487331e-02 -6.31312490e-01
-2.79643238e-01 -8.42550814e-01 -1.34486389e+00 -4.58929390e-01
-1.70371711e-01 6.52469099e-01 8.36477518e-01 9.90320385e-01
1.54893011e-01 3.91400397e-01 6.00666821e-01 -7.34139860e-01
-1.31522012e+00 -8.81413639e-01 -5.33316135e-01 2.53835410e-01
4.68143046e-01 -7.92721629e-01 -5.15867710e-01 -3.53838384e-01] | [4.270388603210449, 2.6068952083587646] |
903710d2-804e-4ec8-bfc4-3c4ca7c9d1ea | nlp_hz-at-semeval-2018-task-9-a-nearest | null | null | https://aclanthology.org/S18-1148 | https://aclanthology.org/S18-1148.pdf | NLP\_HZ at SemEval-2018 Task 9: a Nearest Neighbor Approach | Hypernym discovery aims to discover the hypernym word sets given a hyponym word and proper corpus. This paper proposes a simple but effective method for the discovery of hypernym sets based on word embedding, which can be used to measure the contextual similarities between words. Given a test hyponym word, we get its hypernym lists by computing the similarities between the hyponym word and words in the training data, and fill the test word{'}s hypernym lists with the hypernym list in the training set of the nearest similarity distance to the test word. In SemEval 2018 task9, our results, achieve 1st on Spanish, 2nd on Italian, 6th on English in the metric of MAP. | ['Wei Qiu', 'Mosha Chen', 'Luo Si', 'Linlin Li'] | 2018-06-01 | null | null | null | semeval-2018-6 | ['hypernym-discovery'] | ['natural-language-processing'] | [ 2.42911726e-02 2.75455296e-01 -3.07807863e-01 -8.54942352e-02
2.40862697e-01 -5.55427492e-01 6.85661435e-01 7.45983422e-01
-9.98025715e-01 4.94235963e-01 4.82580721e-01 -2.79284716e-01
-6.43136919e-01 -1.18611157e+00 -2.87548192e-02 -7.18324006e-01
-1.17192201e-01 7.97697484e-01 1.04272105e-01 -6.27179325e-01
2.92395562e-01 8.99688378e-02 -1.94182384e+00 2.98476989e-05
7.95515239e-01 5.49093723e-01 1.54440597e-01 1.19639687e-01
-6.81177616e-01 7.30822980e-02 -5.51555395e-01 -6.41325235e-01
3.55906337e-01 -2.16698080e-01 -8.65884602e-01 -4.70044494e-01
4.59417224e-01 4.01317805e-01 -5.21123946e-01 1.28940964e+00
3.56981039e-01 4.84757841e-01 6.17694795e-01 -1.38531864e+00
-6.11227810e-01 8.60220134e-01 -1.06568709e-01 2.28849247e-01
4.16834772e-01 -3.19586277e-01 1.63489795e+00 -1.06646264e+00
8.91402900e-01 1.18068957e+00 2.73939788e-01 5.27854800e-01
-6.54137969e-01 -8.85744333e-01 -4.53801006e-01 6.77580178e-01
-1.42990279e+00 2.87539691e-01 1.32557094e-01 -6.10759780e-02
1.49354208e+00 3.20835114e-01 9.50364172e-01 5.89185953e-01
-3.06806415e-01 9.94573440e-03 6.87955022e-01 -8.52832139e-01
9.37672257e-02 -1.42465532e-01 5.37233889e-01 5.88681519e-01
8.74640167e-01 -3.57769281e-02 -2.33663917e-01 -2.73454338e-01
1.30181070e-02 1.04930386e-01 -1.50890321e-01 -1.25895098e-01
-1.07799447e+00 9.34855282e-01 5.22865295e-01 7.31884539e-01
-2.21397787e-01 -1.93102583e-01 5.75329900e-01 2.45403022e-01
1.91020265e-01 1.10052490e+00 -4.28465217e-01 -1.44341821e-02
-5.52988112e-01 1.59100145e-01 8.23056519e-01 9.53444660e-01
1.00661695e+00 -6.63913965e-01 2.14955509e-02 9.02515650e-01
2.87329644e-01 4.57628548e-01 1.22271025e+00 -2.79356569e-01
4.51837301e-01 1.25493240e+00 -4.42740649e-01 -8.37321043e-01
-4.10535991e-01 1.56774685e-01 -3.42738748e-01 -2.09006965e-01
-2.32839525e-01 3.21307093e-01 -8.51088583e-01 1.40061736e+00
5.65294206e-01 5.13846755e-01 4.04462457e-01 5.41915476e-01
1.33170235e+00 4.52310324e-01 1.96654767e-01 -1.79145202e-01
1.63246047e+00 -6.77767873e-01 -7.37991512e-01 -7.98933133e-02
1.11586607e+00 -5.77289045e-01 1.21987152e+00 1.04644177e-02
-5.08379638e-01 -7.29962140e-02 -1.28431261e+00 -3.16025061e-03
-1.30861092e+00 -5.09643197e-01 4.74165678e-01 4.49599147e-01
-6.27037287e-01 7.47333646e-01 -6.26290683e-03 -9.30430233e-01
7.69847557e-02 1.80115595e-01 -6.72349870e-01 -3.63987565e-01
-2.00764513e+00 1.50974619e+00 1.62150753e+00 -5.82182288e-01
-2.88013518e-01 -7.28157163e-01 -1.19576979e+00 1.39949873e-01
2.93932617e-01 -3.59898806e-01 6.15539134e-01 -1.68520987e-01
-5.17077267e-01 1.31369472e+00 -4.71516885e-03 -6.38467014e-01
-3.92581224e-01 6.07060231e-02 -1.23061168e+00 -5.82375117e-02
4.58552688e-01 4.68082070e-01 3.23537022e-01 -8.71902764e-01
-7.79432118e-01 -1.84536204e-01 2.06793267e-02 4.30106670e-01
-9.34609175e-01 -9.67465267e-02 -1.52196184e-01 -2.59161144e-01
3.90697777e-01 -6.69466496e-01 -1.34580493e-01 -5.54670930e-01
-2.08130985e-01 -6.96761072e-01 7.74345636e-01 -2.30543450e-01
1.60829127e+00 -2.42555118e+00 -2.91753054e-01 6.48930311e-01
6.41562700e-01 6.38187289e-01 -4.26722854e-01 7.24173129e-01
-5.41303158e-01 2.69252270e-01 -2.22764790e-01 4.99623209e-01
5.65322489e-02 8.33344996e-01 -4.07286674e-01 1.25837117e-01
-2.83975452e-01 9.96600747e-01 -1.33712769e+00 -6.18949652e-01
3.43755007e-01 -6.31418750e-02 -2.69306242e-01 -6.03592359e-02
7.49143660e-02 -5.92400849e-01 -2.77282838e-02 7.12914169e-02
3.58989358e-01 5.87994978e-02 4.28044021e-01 -2.45682850e-01
1.13508485e-01 6.57499194e-01 -1.08470297e+00 1.29674697e+00
-5.94321966e-01 5.32541215e-01 -9.14365709e-01 -6.24005437e-01
1.08351958e+00 5.98475575e-01 5.77613950e-01 -8.51366997e-01
2.79647913e-02 6.44288301e-01 5.03589988e-01 -9.65519488e-01
6.82135403e-01 -2.57079631e-01 1.47932731e-02 4.54429269e-01
4.49794352e-01 -9.65186059e-02 7.83434272e-01 3.57144743e-01
1.36018395e+00 -4.73404527e-01 1.01047516e+00 -4.37158346e-01
4.18912321e-01 -2.01000959e-01 2.64832467e-01 2.99594343e-01
-6.14121323e-03 1.82319969e-01 2.31741771e-01 -6.60229087e-01
-1.16204369e+00 -1.31141913e+00 -4.34532762e-01 1.13085341e+00
3.66844028e-01 -1.39371598e+00 -2.32455000e-01 -7.15609848e-01
2.66095906e-01 1.06383932e+00 -6.21065795e-01 -4.30666327e-01
-5.33848643e-01 -5.12564480e-01 9.10716116e-01 8.76422971e-02
3.11133355e-01 -1.46841824e+00 -7.63010263e-01 1.00362701e-02
-1.90649852e-01 -1.02400196e+00 -8.87433961e-02 3.92526150e-01
-6.00885391e-01 -1.28583276e+00 2.22056154e-02 -1.15319347e+00
4.23702210e-01 6.26231283e-02 1.28540099e+00 3.63535911e-01
-5.89171171e-01 -2.03284472e-01 -8.22100878e-01 -4.61759597e-01
-9.26185921e-02 -2.69117318e-02 3.69919986e-01 -5.15248239e-01
1.25508809e+00 -6.49582744e-01 -2.64685392e-01 -9.70457569e-02
-1.19085062e+00 -4.10380542e-01 2.31511712e-01 7.15348661e-01
4.94265318e-01 1.36259109e-01 2.58900762e-01 -1.15296352e+00
7.64786541e-01 -6.63790464e-01 -1.62475973e-01 4.66836214e-01
-1.28848505e+00 3.51286411e-01 3.20704907e-01 -2.86628336e-01
-2.35877052e-01 -3.89839649e-01 -3.47193256e-02 -1.80654496e-01
-4.20923680e-02 6.70059621e-01 -3.46190274e-01 3.35913718e-01
9.22325671e-01 -7.14666629e-03 -4.12747681e-01 -3.16356391e-01
1.05738688e+00 5.40398061e-01 6.48006558e-01 -3.59751046e-01
8.38496327e-01 2.50464559e-01 2.49594659e-01 -7.99469709e-01
-7.67043948e-01 -1.13405252e+00 -8.82661104e-01 2.37815663e-01
8.53119969e-01 -5.48725069e-01 -5.79269230e-01 -6.30211532e-01
-1.14252138e+00 8.03592563e-01 -7.49855697e-01 6.07587755e-01
-9.90992635e-02 5.22898436e-01 8.89949948e-02 -3.64031971e-01
-6.13843918e-01 -3.44372898e-01 6.02806449e-01 3.95364612e-02
-8.12437117e-01 -1.04489052e+00 6.53273225e-01 -1.59661889e-01
-6.79868907e-02 1.24656605e-02 1.85019922e+00 -1.59045577e+00
2.50743806e-01 -5.53056657e-01 6.73635071e-03 1.41587019e-01
3.85993868e-01 -2.06406996e-01 -7.63625562e-01 -1.80300072e-01
-2.44585931e-01 -5.19596897e-02 8.86863172e-01 -2.03106720e-02
7.51311839e-01 -5.81882715e-01 -5.81370831e-01 5.63162684e-01
1.52657187e+00 5.43974154e-02 8.85547936e-01 6.24831080e-01
7.79039025e-01 6.33340955e-01 7.42019653e-01 3.05093348e-01
2.85947118e-02 3.87637824e-01 5.03752410e-01 2.66461521e-01
3.44837829e-02 -5.37210464e-01 7.20549077e-02 1.23945022e+00
1.72783077e-01 -1.22244038e-01 -9.79359269e-01 8.39234650e-01
-1.73141420e+00 -9.87775743e-01 -4.21034992e-01 2.36214685e+00
8.40975285e-01 -8.68413970e-02 -1.47498593e-01 3.67445379e-01
6.59498930e-01 2.58372068e-01 -1.68964982e-01 -5.28567135e-01
-4.56544608e-01 9.77311432e-01 4.56147134e-01 5.13152122e-01
-8.29174578e-01 1.45699775e+00 5.46306515e+00 9.27094102e-01
-4.13534552e-01 1.72224328e-01 -6.16386950e-01 -2.13621810e-01
-8.46762002e-01 5.58923364e-01 -6.61698580e-01 4.74408269e-01
7.24858522e-01 -7.05620170e-01 2.41453603e-01 5.39295733e-01
-3.25101048e-01 1.29589453e-01 -1.15048039e+00 1.23631001e+00
4.72725362e-01 -1.00454533e+00 4.81521517e-01 1.04468152e-01
6.35398269e-01 9.69964862e-02 -3.59251559e-01 4.68588799e-01
2.95740575e-01 -1.32616591e+00 -1.31363630e-01 -7.79533908e-02
8.24956954e-01 -7.40346432e-01 8.19695950e-01 4.57551405e-02
-1.34845364e+00 1.18862810e-02 -8.87178898e-01 -1.13246650e-01
1.19321786e-01 7.23942339e-01 -1.22819543e+00 3.97386938e-01
6.91703856e-01 8.02481294e-01 -6.87172771e-01 1.13572335e+00
-9.10290539e-01 3.48744512e-01 -2.10953444e-01 -3.86485279e-01
2.99885005e-01 -2.63253748e-01 6.74821794e-01 1.09794056e+00
3.97322565e-01 6.76604509e-02 2.27795988e-02 5.63312411e-01
-2.97354370e-01 6.52307153e-01 -1.10813415e+00 -3.00580710e-01
9.36297834e-01 1.57472146e+00 -4.29920405e-01 -5.65781236e-01
-1.70953438e-01 8.99500608e-01 3.42665434e-01 -1.23665007e-02
-5.47565699e-01 -9.84360635e-01 1.16053641e+00 -1.34978488e-01
-1.66790009e-01 6.00117221e-02 -2.44968042e-01 -8.34318280e-01
-9.68347639e-02 -4.94422227e-01 1.13966680e+00 -6.21643186e-01
-1.43649995e+00 6.15370333e-01 -1.05531970e-02 -1.26542580e+00
-4.17748779e-01 -1.03394699e+00 -6.51630521e-01 8.18916500e-01
-1.03710902e+00 -7.23033428e-01 -1.95845380e-01 4.98996466e-01
1.57429859e-01 -4.88776177e-01 1.32247674e+00 3.11969787e-01
1.23629861e-01 5.60137689e-01 3.80711854e-02 2.42257223e-01
6.38430357e-01 -1.19389927e+00 7.98965752e-01 5.59960127e-01
9.62362885e-01 8.28688502e-01 7.44619668e-01 -9.08053279e-01
-6.22262061e-01 -8.86200249e-01 1.92602384e+00 -6.82735026e-01
9.20814097e-01 -1.10724397e-01 -1.01628816e+00 3.71320993e-01
3.85448962e-01 -2.04348087e-01 9.97916937e-01 5.24215519e-01
-9.76279557e-01 2.70956606e-01 -9.78644967e-01 9.53961551e-01
1.20684814e+00 -6.13274515e-01 -1.41618586e+00 4.98536438e-01
1.26799595e+00 2.48172015e-01 -7.16605663e-01 6.48986936e-01
4.26842660e-01 -1.67224765e-01 9.66720700e-01 -1.29885018e+00
3.72022063e-01 -4.40075696e-01 -4.27517891e-01 -1.59833586e+00
-3.89392227e-01 -1.77832544e-01 -8.81482661e-02 9.68372345e-01
6.93260550e-01 -5.58764875e-01 4.78738993e-01 1.47441596e-01
7.74966553e-02 -5.71847379e-01 -1.26858795e+00 -1.13056684e+00
4.73031687e-04 -3.54857355e-01 1.03041780e+00 1.30164599e+00
8.39240491e-01 5.59416831e-01 1.23610543e-02 -1.49171248e-01
2.64866471e-01 -2.57885568e-02 9.60520506e-02 -1.52950764e+00
3.60380441e-01 -5.69825590e-01 -9.84559178e-01 -9.44576859e-02
5.63888133e-01 -1.74137342e+00 -1.08233348e-01 -1.55836511e+00
3.35782468e-01 1.04960166e-01 -6.58041418e-01 5.83734512e-01
-2.88782775e-01 8.50478485e-02 1.62425578e-01 1.18729398e-01
-5.83074808e-01 5.82328677e-01 8.11548710e-01 -1.59017086e-01
2.03777999e-02 -6.79067135e-01 -2.63743281e-01 6.45661414e-01
6.51750445e-01 -7.91015148e-01 -4.92540270e-01 -3.44209969e-01
6.92028999e-01 -1.01595187e+00 1.79365396e-01 -7.88577080e-01
2.42086127e-01 -8.73445347e-02 -1.88648820e-01 -1.99859187e-01
1.87439114e-01 -1.06829476e+00 -2.73146540e-01 7.71682858e-01
-6.09367669e-01 4.66405362e-01 -5.18376194e-02 2.50474662e-01
-2.10950196e-01 -8.86259139e-01 4.68157858e-01 -7.11375177e-02
-1.22888863e+00 3.07064503e-01 1.04249157e-01 5.94384491e-01
9.63646293e-01 -2.88062662e-01 -2.99834996e-01 6.10338524e-02
-4.02164519e-01 2.86290973e-01 2.36062944e-01 9.46471274e-01
9.25667465e-01 -1.76513946e+00 -6.22202992e-01 1.87941968e-01
1.01946783e+00 -5.70917606e-01 -2.27006599e-01 1.72765717e-01
-4.48420674e-01 2.21161306e-01 -2.19772145e-01 7.65642747e-02
-1.40795171e+00 7.64729679e-01 1.64137501e-02 -4.78895567e-02
-8.18322659e-01 7.12661445e-01 2.99344584e-02 -8.12937140e-01
4.50173579e-02 -1.75685480e-01 -8.04794431e-01 2.64193356e-01
5.84990501e-01 5.40119350e-01 1.31838277e-01 -7.90620744e-01
-6.58731937e-01 5.65781653e-01 2.12520227e-01 -4.38794196e-01
1.02446985e+00 3.51090789e-01 -8.70490670e-01 6.06945336e-01
1.54168487e+00 -2.83812851e-01 2.01989755e-01 -4.04392838e-01
7.17314601e-01 -6.48707807e-01 -5.89813888e-01 -6.31528080e-01
-5.86802542e-01 6.19164646e-01 7.15595722e-01 1.31323244e-02
6.88742042e-01 2.92634547e-01 9.18891251e-01 7.44857788e-01
1.76010922e-01 -1.32911694e+00 -2.03823686e-01 1.07384896e+00
4.64235067e-01 -9.30996776e-01 -1.04310952e-01 -4.57749784e-01
-6.06069505e-01 1.08101261e+00 6.96489811e-01 -1.24671482e-01
9.74327028e-01 -2.50776857e-01 4.17360626e-02 -7.58243561e-01
-5.35026312e-01 -7.98050463e-01 5.14926016e-01 5.88133514e-01
4.72814560e-01 2.73002356e-01 -1.15036142e+00 4.75347847e-01
-7.20538795e-01 -6.59728169e-01 1.12991743e-01 3.63121033e-01
-7.78070152e-01 -1.14095557e+00 2.68380493e-01 6.89932168e-01
1.51159748e-01 -8.63891602e-01 -5.67220569e-01 7.72673309e-01
6.81441486e-01 7.34898746e-01 1.73597693e-01 -6.79308176e-01
4.24751431e-01 4.87191081e-01 2.29908600e-01 -1.08656585e+00
-4.82979268e-01 -8.07812452e-01 9.84907895e-03 -3.72797668e-01
1.26272455e-01 -1.65485129e-01 -1.79285872e+00 -1.03536546e-01
-2.87293166e-01 4.70658541e-01 3.69693935e-01 1.35982728e+00
2.82717645e-02 4.17192243e-02 3.32061917e-01 4.37550217e-01
-3.34399074e-01 -1.05048704e+00 -8.71402860e-01 1.35794854e+00
-3.64205837e-01 -5.59949040e-01 -2.24944159e-01 -3.43810022e-01] | [9.874410629272461, 8.75112533569336] |
3e6faaa5-48b9-435e-80f5-f3d92953d977 | expansion-of-visual-hints-for-improved | 2211.00392 | null | https://arxiv.org/abs/2211.00392v1 | https://arxiv.org/pdf/2211.00392v1.pdf | Expansion of Visual Hints for Improved Generalization in Stereo Matching | We introduce visual hints expansion for guiding stereo matching to improve generalization. Our work is motivated by the robustness of Visual Inertial Odometry (VIO) in computer vision and robotics, where a sparse and unevenly distributed set of feature points characterizes a scene. To improve stereo matching, we propose to elevate 2D hints to 3D points. These sparse and unevenly distributed 3D visual hints are expanded using a 3D random geometric graph, which enhances the learning and inference process. We evaluate our proposal on multiple widely adopted benchmarks and show improved performance without access to additional sensors other than the image sequence. To highlight practical applicability and symbiosis with visual odometry, we demonstrate how our methods run on embedded hardware. | ['Juho Kannala', 'Arno Solin', 'Niki Loppi', 'Yuxin Hou', 'Andrea Pilzer'] | 2022-11-01 | null | null | null | null | ['stereo-matching-1'] | ['computer-vision'] | [-6.24668747e-02 7.35827610e-02 -3.59707624e-01 1.73399244e-02
-1.51245028e-01 -6.17560744e-01 7.33659863e-01 -4.01924038e-03
-3.88936341e-01 4.81245935e-01 1.61979407e-01 -3.91047597e-01
-4.59398776e-02 -5.39635599e-01 -9.02382135e-01 -3.96536440e-01
-1.78504631e-01 5.05399346e-01 5.91268241e-01 -2.26083428e-01
6.51317239e-01 7.67999351e-01 -1.93185067e+00 -2.89973438e-01
6.74145162e-01 7.12179601e-01 3.93962562e-01 6.92574322e-01
2.63810575e-01 5.64902544e-01 -2.48796582e-01 1.77914083e-01
6.78992748e-01 2.07102373e-01 -5.84324360e-01 3.04783195e-01
1.05574179e+00 -4.47011322e-01 -5.84369242e-01 1.21306729e+00
5.47260404e-01 1.49212390e-01 4.54568297e-01 -1.40614295e+00
-6.06333673e-01 1.73991714e-02 -4.98637319e-01 1.16788782e-01
7.51848638e-01 4.27478790e-01 8.69308591e-01 -9.45405304e-01
1.07465899e+00 1.21843612e+00 6.51948690e-01 3.21151227e-01
-1.03048050e+00 -4.07212436e-01 4.67147678e-02 3.75393033e-01
-1.30493176e+00 -2.64139146e-01 9.34254050e-01 -4.40376699e-01
1.06391776e+00 -4.06726450e-02 7.62730062e-01 9.56580400e-01
2.56313682e-01 7.11553752e-01 8.73555124e-01 -2.54387558e-01
3.90537769e-01 -8.67004972e-03 1.10310331e-01 1.11754429e+00
5.70118785e-01 5.68122268e-01 -6.07546687e-01 -1.29901111e-01
1.01831341e+00 4.47890490e-01 -2.42200360e-01 -1.38180494e+00
-1.17494416e+00 8.05135012e-01 1.08243835e+00 -3.95876437e-01
-2.71002024e-01 6.47158176e-02 1.23179965e-01 2.34229982e-01
-1.47926942e-01 4.26890701e-01 -1.13385379e-01 -3.32360864e-02
-3.82284075e-01 8.57337639e-02 7.14310050e-01 1.24388146e+00
1.35264552e+00 6.29630759e-02 2.71579862e-01 6.54974639e-01
3.75683457e-01 1.06133151e+00 4.32221323e-01 -1.36583900e+00
3.96304041e-01 7.29069114e-01 1.50717169e-01 -1.29651308e+00
-6.31019533e-01 -2.56831229e-01 -7.01631665e-01 7.00254261e-01
1.45444304e-01 2.36427069e-01 -1.10547888e+00 1.32188153e+00
4.00127381e-01 3.31161827e-01 -4.53880727e-02 1.32388866e+00
6.93891823e-01 1.04268566e-01 -4.53578115e-01 4.04203922e-01
9.25838470e-01 -1.03132474e+00 8.50737393e-02 -5.41549385e-01
5.36011994e-01 -5.48719406e-01 9.99608159e-01 2.03844473e-01
-6.70643330e-01 -6.36011839e-01 -1.30911553e+00 -2.92487681e-01
-2.18616083e-01 -3.29275697e-01 6.14912033e-01 4.22757596e-01
-1.20551920e+00 5.27762473e-01 -8.03180635e-01 -6.15427732e-01
1.07979931e-01 6.20865703e-01 -5.43823838e-01 -2.58244693e-01
-7.36404300e-01 9.28209424e-01 3.04662228e-01 -3.33743840e-01
-7.42848456e-01 -4.18558508e-01 -1.37014747e+00 -3.78330737e-01
2.01848358e-01 -1.21045291e+00 8.99020016e-01 -9.74242985e-02
-1.42303264e+00 1.09031093e+00 -2.94198096e-01 -5.63192070e-01
3.55075061e-01 -2.72224426e-01 1.72247201e-01 2.40756139e-01
1.24677829e-01 1.01739144e+00 8.10354471e-01 -1.44378769e+00
-7.30329275e-01 -5.96093297e-01 3.92897427e-01 5.47394514e-01
-1.20647270e-02 -8.72373402e-01 -6.02391839e-01 -6.25190586e-02
8.54608059e-01 -1.36256957e+00 -5.60918272e-01 4.24700007e-02
-2.94397742e-01 2.17481732e-01 9.29763675e-01 -7.29805082e-02
4.52386349e-01 -2.07730627e+00 9.12072435e-02 3.92224729e-01
4.88136888e-01 -7.13756904e-02 3.92452590e-02 4.23285998e-02
2.90761828e-01 -4.84958142e-01 2.52523512e-01 -4.11290944e-01
1.00349866e-01 6.19220614e-01 -3.79330248e-01 9.38599706e-01
-1.81862012e-01 7.30955780e-01 -1.33411515e+00 -3.56768042e-01
8.00613284e-01 3.15962940e-01 -9.21072125e-01 1.10529281e-01
1.21141843e-01 4.63482946e-01 -3.31129760e-01 7.98111916e-01
7.93971300e-01 -3.10658008e-01 -1.87767237e-01 -1.86086684e-01
-9.21792760e-02 6.47934198e-01 -1.29547930e+00 2.08045936e+00
-3.14981073e-01 5.46867609e-01 -2.74745077e-01 -7.92085588e-01
1.16866565e+00 -4.18517262e-01 3.79912913e-01 -8.75330925e-01
7.55219385e-02 2.23801881e-01 -2.81420678e-01 -1.32392541e-01
9.85150337e-01 3.65988284e-01 -9.38770697e-02 6.47789985e-02
4.52819541e-02 -6.31542742e-01 3.03937010e-02 1.36801764e-01
1.12080884e+00 3.25100064e-01 6.45610869e-01 -2.60541052e-01
3.76404613e-01 4.10100073e-01 4.10619229e-01 9.30749655e-01
-3.29903096e-01 6.65615737e-01 2.22105421e-02 -5.66537917e-01
-1.26491976e+00 -1.51244605e+00 -2.12026939e-01 6.35472775e-01
9.73279595e-01 -2.88570285e-01 -1.30732819e-01 -5.81770599e-01
5.14431059e-01 1.85920317e-02 -3.46034706e-01 -1.79253191e-01
-4.77460861e-01 -2.50200517e-02 1.72612116e-01 4.93733108e-01
3.92124206e-01 -7.08207726e-01 -1.03076398e+00 -7.93982893e-02
1.39181688e-01 -1.02295697e+00 -3.03500414e-01 2.51056165e-01
-1.03808868e+00 -1.27636969e+00 -5.31780839e-01 -7.73629010e-01
7.81606376e-01 9.76911724e-01 1.13058197e+00 -1.30620733e-01
-1.74481392e-01 7.24290669e-01 -1.92917570e-01 -8.06407928e-02
-5.73214814e-02 1.02166563e-01 4.63148415e-01 -6.13217235e-01
3.41866076e-01 -8.14075410e-01 -6.72955155e-01 4.56226915e-01
-4.34424937e-01 -1.18811075e-02 3.60716701e-01 9.33891237e-01
5.18022120e-01 -7.81276584e-01 -5.63984752e-01 -3.57100278e-01
-5.48781678e-02 -3.72999489e-01 -9.25154448e-01 -2.32779384e-01
-4.96111453e-01 4.24472034e-01 3.59132677e-01 -3.39191526e-01
-2.37294421e-01 3.99626374e-01 1.85897008e-01 -8.11636269e-01
-1.31362408e-01 1.11108884e-01 2.52279967e-01 -6.54881239e-01
9.31131124e-01 8.68490059e-03 5.54841533e-02 -1.81924716e-01
4.95816410e-01 4.48591322e-01 8.45183671e-01 -4.22980994e-01
9.74164188e-01 9.01785254e-01 4.88916993e-01 -8.35821271e-01
-5.02907336e-01 -8.56860936e-01 -5.67613125e-01 -5.97276501e-02
2.61435062e-01 -1.10350585e+00 -1.14165735e+00 1.25974968e-01
-1.01185834e+00 -2.14806482e-01 -3.00482869e-01 7.18817890e-01
-9.91439819e-01 6.37741923e-01 -4.19781178e-01 -5.03308892e-01
-1.15534961e-01 -1.30082214e+00 1.43880665e+00 2.24867433e-01
-2.04855293e-01 -1.01116836e+00 3.38968217e-01 -1.06487431e-01
-8.38630795e-02 9.28245187e-02 1.74793050e-01 -1.68860748e-01
-1.03198159e+00 8.19484368e-02 -2.71136224e-01 -2.69506872e-01
8.17875117e-02 -3.29591602e-01 -8.20819139e-01 -5.45036137e-01
-2.39314467e-01 -3.24055254e-01 8.39147270e-01 1.64782777e-01
5.38532317e-01 1.31277531e-01 -5.06208956e-01 1.08774316e+00
1.39906359e+00 -7.59419426e-02 4.29983914e-01 8.96745563e-01
8.94675851e-01 3.41724485e-01 7.02285230e-01 6.71701193e-01
7.09970653e-01 7.71177173e-01 1.07704461e+00 -1.34543777e-02
-1.70569867e-01 -6.09435618e-01 3.20044786e-01 5.64800143e-01
-1.29524797e-01 3.65057766e-01 -9.73478734e-01 5.39165318e-01
-1.79169559e+00 -7.21622407e-01 -4.16015089e-02 2.41497970e+00
3.61669540e-01 1.45055726e-01 -1.98546867e-03 9.36769843e-02
4.49599594e-01 2.80732065e-01 -7.19266832e-01 -7.90663064e-02
-1.27630532e-01 -3.31296325e-02 9.86918390e-01 9.01546240e-01
-9.71904397e-01 1.02418399e+00 6.42296267e+00 1.53111011e-01
-1.33524036e+00 -3.18214864e-01 -1.68194413e-01 4.24167179e-02
-3.58386219e-01 1.56546772e-01 -1.14928603e+00 1.69646680e-01
2.40366206e-01 -1.02806240e-01 4.71331030e-01 1.26126266e+00
-1.77888006e-01 -2.15569302e-01 -1.19296193e+00 1.33682227e+00
1.09758303e-01 -1.44381011e+00 -2.01629642e-02 2.55959839e-01
8.61920774e-01 6.46740258e-01 2.21630812e-01 4.83509488e-02
8.41872931e-01 -5.83096921e-01 6.65288568e-01 1.23137027e-01
4.71528232e-01 -4.68440443e-01 4.11578298e-01 3.79519671e-01
-1.26786959e+00 -1.97676912e-01 -8.07744205e-01 -4.12058085e-01
-4.43558581e-02 4.13916647e-01 -1.33833528e+00 5.73287725e-01
9.09185469e-01 1.17307997e+00 -6.72289431e-01 1.20717502e+00
-3.88944715e-01 -2.61204213e-01 -6.39841139e-01 -2.19505709e-02
2.98773855e-01 -1.81744799e-01 8.34599197e-01 8.24201405e-01
5.38765430e-01 -4.11583662e-01 5.16211748e-01 6.09264135e-01
2.58331895e-01 -3.37411255e-01 -1.40603590e+00 7.63195634e-01
5.95818579e-01 9.27984476e-01 -4.34167892e-01 -3.63779992e-01
-2.65964985e-01 1.04101253e+00 5.04368365e-01 2.87761569e-01
-5.42213202e-01 -1.31854668e-01 1.00619543e+00 -1.11464737e-02
2.83154041e-01 -6.86012328e-01 -4.51860964e-01 -1.30151308e+00
2.46673524e-02 -5.97163439e-01 3.10479105e-01 -9.47316527e-01
-9.46309566e-01 3.13404351e-01 -1.83164895e-01 -1.76624489e+00
-7.31436014e-01 -8.20239007e-01 -1.29008397e-01 7.78886199e-01
-1.60560763e+00 -7.93579400e-01 -7.45486259e-01 8.90368700e-01
2.43512824e-01 -8.97692610e-03 5.35523772e-01 2.23968159e-02
1.74884617e-01 1.79154053e-01 -7.10619539e-02 -1.16748624e-01
6.82077110e-01 -1.31860602e+00 7.19295502e-01 7.46645033e-01
4.06213552e-01 8.66355479e-01 9.46830690e-01 -4.90329713e-01
-1.85616577e+00 -7.67821848e-01 6.20999455e-01 -5.87554455e-01
6.62521243e-01 -2.88225174e-01 -5.14827371e-01 7.73533225e-01
-4.09635007e-02 2.93057740e-01 -7.18502253e-02 1.83093265e-01
-5.98195970e-01 -1.36556178e-01 -1.12040591e+00 9.84899759e-01
1.55296230e+00 -6.97499871e-01 -9.02045429e-01 1.99578464e-01
5.69796979e-01 -9.93339896e-01 -5.16341269e-01 5.96750319e-01
4.80362624e-01 -1.34044385e+00 1.35862279e+00 -2.38726854e-01
-1.95369739e-02 -5.95804572e-01 -4.16440070e-01 -1.36146629e+00
-4.00950760e-01 -5.78986764e-01 -1.44764960e-01 4.32600796e-01
-1.01490691e-01 -8.23321879e-01 1.21099627e+00 6.89859018e-02
-2.45186120e-01 -1.16850488e-01 -8.28812897e-01 -9.49161470e-01
-5.74989378e-01 -3.14377218e-01 3.69358152e-01 7.10674822e-01
3.67450207e-01 1.78051651e-01 -3.60828578e-01 6.59819305e-01
9.73654985e-01 3.59883934e-01 1.29282689e+00 -1.17739570e+00
-3.50071967e-01 -2.42672682e-01 -1.14975917e+00 -1.85057616e+00
-1.12594068e-02 -8.17211688e-01 1.11038208e-01 -1.04224598e+00
-4.31925151e-03 -3.61439019e-01 -1.32694945e-01 2.04171583e-01
-3.77300978e-02 6.05343878e-01 3.12890053e-01 3.41543972e-01
-7.73044050e-01 4.72857326e-01 1.14621496e+00 -2.08149314e-01
-2.37502992e-01 -1.29585132e-01 -1.41956568e-01 7.54799306e-01
4.56688315e-01 -2.48151898e-01 -3.32536668e-01 -4.58860427e-01
1.03306450e-01 -9.27871764e-02 7.25517333e-01 -1.30334115e+00
5.49957633e-01 1.01772532e-01 1.45657972e-01 -8.26735079e-01
5.50785422e-01 -9.06578302e-01 -1.03145860e-01 9.11131620e-01
2.24023864e-01 2.30696887e-01 8.57057795e-02 4.25814956e-01
-1.59027681e-01 1.29102506e-02 6.51704371e-01 -7.21515566e-02
-1.23138964e+00 3.95872205e-01 4.54084277e-02 -1.15771383e-01
8.00749540e-01 -5.90008140e-01 -5.64011395e-01 -3.95304799e-01
-4.53759938e-01 2.25207880e-01 1.17974317e+00 5.85252047e-01
7.52732694e-01 -1.41498888e+00 -8.35206136e-02 5.31769872e-01
4.91343260e-01 1.41155466e-01 2.51593366e-02 6.82313442e-01
-7.49469221e-01 5.41117311e-01 -6.11772239e-01 -1.33688712e+00
-1.17375743e+00 8.06026995e-01 1.93860173e-01 1.25115052e-01
-8.58432829e-01 4.69434500e-01 1.84402600e-01 -6.99127614e-01
4.65212733e-01 -4.76146460e-01 1.93944369e-02 -5.03220260e-01
2.28335291e-01 5.21537840e-01 -8.52804780e-02 -4.14461046e-01
-4.53425467e-01 9.89707053e-01 1.89006835e-01 -6.42712563e-02
9.62627292e-01 -4.51070815e-01 1.38477191e-01 3.98075193e-01
1.16915083e+00 1.31848499e-01 -1.89855874e+00 -4.31627274e-01
1.43565789e-01 -7.21746504e-01 -3.37743431e-01 8.11608434e-02
-6.16796196e-01 7.43938744e-01 6.03243649e-01 6.84066564e-02
6.63370728e-01 1.16497412e-01 4.63579297e-01 7.22605467e-01
1.00043941e+00 -7.12099493e-01 1.31909326e-01 1.05304313e+00
5.67291915e-01 -1.46300960e+00 1.46065608e-01 -4.26841885e-01
-3.87943089e-01 1.01960397e+00 4.98908937e-01 -6.45094991e-01
3.78054827e-01 2.81680495e-01 2.02948079e-01 -7.33255222e-02
-3.95626038e-01 -6.11913085e-01 3.52506697e-01 1.15435028e+00
-1.99542060e-01 -3.70645940e-01 1.14330374e-01 -3.73503566e-01
-4.23602074e-01 -1.33601487e-01 4.12021875e-01 1.03661668e+00
-7.85888612e-01 -9.12646949e-01 -5.63568413e-01 -2.49283135e-01
2.57265002e-01 2.23317407e-02 -1.76060706e-01 8.90981317e-01
-4.27401885e-02 6.37358665e-01 1.70828432e-01 -6.21147811e-01
3.38837683e-01 -4.93738025e-01 6.45650804e-01 -4.99288917e-01
-5.09405844e-02 -2.43953899e-01 -1.79581523e-01 -1.05375934e+00
-3.31627846e-01 -6.28372073e-01 -1.26423562e+00 -4.02104497e-01
2.57385314e-01 -6.97835982e-02 8.18770587e-01 7.32622862e-01
6.05300069e-01 5.62485643e-02 6.92791402e-01 -1.36964607e+00
-6.64857030e-01 -5.33466160e-01 -5.30565023e-01 4.30019408e-01
7.03281939e-01 -9.68554080e-01 -3.75099808e-01 -1.58843890e-01] | [7.8836669921875, -2.202723264694214] |
efb63804-d67e-41b5-855d-d5cdad56e965 | explicit-feature-interaction-aware-uplift | 2306.00315 | null | https://arxiv.org/abs/2306.00315v1 | https://arxiv.org/pdf/2306.00315v1.pdf | Explicit Feature Interaction-aware Uplift Network for Online Marketing | As a key component in online marketing, uplift modeling aims to accurately capture the degree to which different treatments motivate different users, such as coupons or discounts, also known as the estimation of individual treatment effect (ITE). In an actual business scenario, the options for treatment may be numerous and complex, and there may be correlations between different treatments. In addition, each marketing instance may also have rich user and contextual features. However, existing methods still fall short in both fully exploiting treatment information and mining features that are sensitive to a particular treatment. In this paper, we propose an explicit feature interaction-aware uplift network (EFIN) to address these two problems. Our EFIN includes four customized modules: 1) a feature encoding module encodes not only the user and contextual features, but also the treatment features; 2) a self-interaction module aims to accurately model the user's natural response with all but the treatment features; 3) a treatment-aware interaction module accurately models the degree to which a particular treatment motivates a user through interactions between the treatment features and other features, i.e., ITE; and 4) an intervention constraint module is used to balance the ITE distribution of users between the control and treatment groups so that the model would still achieve a accurate uplift ranking on data collected from a non-random intervention marketing scenario. We conduct extensive experiments on two public datasets and one product dataset to verify the effectiveness of our EFIN. In addition, our EFIN has been deployed in a credit card bill payment scenario of a large online financial platform with a significant improvement. | ['Xiuqiang He', 'Fuyuan Lyu', 'Han Gao', 'Xing Tang', 'Dugang Liu'] | 2023-06-01 | null | null | null | null | ['marketing'] | ['miscellaneous'] | [ 1.69538125e-01 -9.71063748e-02 -8.49362493e-01 -8.39984596e-01
-2.05041096e-01 -3.65041465e-01 4.18470591e-01 2.95413971e-01
-1.63556799e-01 3.19692105e-01 4.25325334e-01 -1.62727118e-01
-4.02425259e-01 -1.06220925e+00 -6.12906218e-01 -5.58518708e-01
-4.75796312e-02 4.03396785e-01 -1.57821909e-01 -2.99896002e-01
2.03413919e-01 1.36484593e-01 -1.51377726e+00 4.24368709e-01
1.07772601e+00 1.07131386e+00 -2.84775645e-02 -3.63903604e-02
-6.70252740e-02 4.07184392e-01 -1.74356475e-01 -4.68072265e-01
4.45327699e-01 -2.34169081e-01 -4.56464887e-01 4.22630459e-01
-2.98240073e-02 -5.61022103e-01 -1.68842629e-01 7.85561562e-01
3.56592715e-01 -6.79252446e-02 6.81647241e-01 -1.43206167e+00
-4.72152710e-01 9.52489793e-01 -7.71689951e-01 -2.76638567e-01
2.98951745e-01 2.89900750e-01 1.34438181e+00 -3.21129590e-01
3.18278283e-01 1.39052987e+00 2.07329363e-01 3.84013414e-01
-1.42818284e+00 -9.09653962e-01 5.15105128e-01 -2.10365243e-02
-8.39945734e-01 -1.67801782e-01 7.02422500e-01 -5.50844789e-01
2.73602933e-01 2.37062827e-01 7.39402413e-01 9.85333204e-01
1.24033436e-01 1.12553895e+00 9.82643843e-01 -5.00493273e-02
3.19951415e-01 4.75386471e-01 6.13708079e-01 5.98636419e-02
-7.88432488e-04 3.61134529e-01 -3.82081926e-01 -4.38065797e-01
6.12440705e-01 5.45693874e-01 -1.11579709e-01 -3.13269049e-01
-6.99465096e-01 1.28444088e+00 5.36743522e-01 1.71167180e-02
-6.14597678e-01 -2.72555560e-01 3.15016955e-01 2.65546530e-01
4.89258677e-01 2.44142219e-01 -8.57324660e-01 8.68434682e-02
-4.75185692e-01 6.26442134e-01 8.65745187e-01 7.38926113e-01
8.46079588e-01 -5.35202861e-01 -5.13827801e-01 9.44927156e-01
3.41258287e-01 1.97766691e-01 6.34129167e-01 -4.95348215e-01
4.81274635e-01 9.39286411e-01 2.12318450e-01 -1.07153571e+00
-5.07809579e-01 -4.19801503e-01 -7.06705272e-01 -3.33003938e-01
3.91415089e-01 -3.20208341e-01 -6.91108823e-01 2.12288189e+00
4.43804592e-01 1.69853449e-01 -4.06666726e-01 8.23623836e-01
6.82001889e-01 2.93572515e-01 3.40482444e-01 -4.74786133e-01
1.50246727e+00 -3.98710251e-01 -7.05165386e-01 -5.29462039e-01
8.79755259e-01 -6.22027993e-01 1.20062304e+00 2.61761278e-01
-7.67905951e-01 -1.91597059e-01 -6.96952224e-01 4.49044794e-01
-1.11692041e-01 3.73054668e-02 1.20562267e+00 6.84970796e-01
-7.07265809e-02 8.12338054e-01 -3.53998035e-01 -2.26308331e-01
5.26984155e-01 5.70244074e-01 -1.97663624e-02 -3.38116616e-01
-1.45599961e+00 3.48912537e-01 1.30795047e-01 -1.12485789e-01
-4.66476530e-01 -1.01006639e+00 -6.06576979e-01 3.81826788e-01
6.04659438e-01 -6.57515705e-01 1.22387075e+00 -1.10632193e+00
-1.26564920e+00 2.92112470e-01 -5.50741702e-02 -2.46303722e-01
3.85685503e-01 -5.20037003e-02 -3.86087358e-01 -5.79877555e-01
2.95582935e-02 4.42111671e-01 6.88280880e-01 -1.00321293e+00
-8.72131228e-01 -7.79916108e-01 3.52044702e-01 3.45404476e-01
-5.41400611e-01 -1.85188148e-02 -2.70318002e-01 -5.44641078e-01
9.06931683e-02 -8.99820864e-01 -3.23354065e-01 -3.39751542e-01
-5.53569317e-01 -3.43523920e-01 6.65795505e-01 -3.46566349e-01
1.40731144e+00 -2.29167724e+00 -9.60882306e-02 4.43429500e-01
3.15504581e-01 6.79398403e-02 -1.59062728e-01 4.77710128e-01
-2.13649496e-01 3.00261527e-01 1.24010667e-01 -3.70512307e-02
2.57856369e-01 2.15465426e-01 -2.24915221e-01 1.42713651e-01
1.10998377e-01 6.83589458e-01 -6.87398493e-01 -1.68446437e-01
6.38348330e-03 1.73598379e-01 -9.18401182e-01 1.86939389e-01
-2.44866744e-01 2.88690358e-01 -8.82735372e-01 6.16249979e-01
8.27172160e-01 -3.14293683e-01 3.39828730e-01 -2.61397421e-01
2.24378243e-01 2.12885559e-01 -1.39303148e+00 9.39638495e-01
-3.32546741e-01 -2.20138758e-01 8.12728703e-02 -9.51047719e-01
6.26472831e-01 1.09219342e-01 8.32792401e-01 -7.45098829e-01
2.94339210e-01 2.21346691e-02 2.06472501e-01 -3.68637890e-01
3.54709774e-01 -1.75559729e-01 -2.29188263e-01 5.25673032e-01
-2.49428749e-01 4.45717961e-01 2.31489152e-01 2.57711172e-01
1.06383657e+00 -4.25382614e-01 1.82031214e-01 -1.70759290e-01
8.29298273e-02 -5.01450837e-01 1.09322500e+00 8.05831432e-01
-2.85640478e-01 1.96832791e-01 8.84537995e-01 -2.57767618e-01
-4.80046391e-01 -6.48257613e-01 -3.09296876e-01 1.06814027e+00
1.55730039e-01 -4.53265458e-01 -6.14143968e-01 -8.88646305e-01
6.00086331e-01 7.84952521e-01 -6.86532378e-01 -2.60259151e-01
2.45999079e-02 -1.07908976e+00 -2.10202992e-01 4.50339854e-01
5.93662918e-01 -8.30214500e-01 6.36343732e-02 2.76794106e-01
-1.32235199e-01 -7.44600117e-01 -8.40972126e-01 4.52573672e-02
-7.29385197e-01 -1.12194157e+00 -1.12097174e-01 -3.18680614e-01
5.29506028e-01 3.13000262e-01 9.21558619e-01 -6.60510063e-02
3.35138030e-02 -1.10674366e-01 -4.00132209e-01 -4.85653788e-01
-2.55304594e-02 7.86715448e-02 -2.45662592e-02 4.67819959e-01
8.19881260e-01 -6.44892514e-01 -8.03557098e-01 7.67284632e-01
-1.09782207e+00 -9.45204347e-02 5.64827561e-01 8.86978269e-01
1.73322126e-01 2.40073025e-01 8.33941996e-01 -1.50159311e+00
7.86178887e-01 -7.92200863e-01 -3.95720303e-01 1.31969213e-01
-9.83283162e-01 -2.32638389e-01 2.77718306e-01 -8.69923294e-01
-9.79305089e-01 2.52751619e-01 7.09408000e-02 7.34464601e-02
-3.95165309e-02 8.76637876e-01 -6.15399182e-01 3.48650903e-01
4.97857690e-01 -2.98206598e-01 7.90738873e-03 -5.33923566e-01
2.55021363e-01 9.50278699e-01 -2.27234319e-01 -4.74356145e-01
7.34645844e-01 2.14782190e-02 -2.35322654e-01 -2.32420623e-01
-1.03369629e+00 -3.80417109e-01 -2.19423354e-01 4.07182314e-02
4.06852305e-01 -8.02674472e-01 -1.20865631e+00 4.31343585e-01
-5.55355906e-01 -1.05405085e-01 -1.05130762e-01 6.23150289e-01
-2.39804164e-01 1.15366243e-01 -7.58281052e-01 -6.70040905e-01
-8.00193027e-02 -1.24955308e+00 8.44593406e-01 4.39974576e-01
-2.40422264e-01 -9.57733333e-01 -4.56030339e-01 5.85013926e-01
2.06042960e-01 2.56422728e-01 1.20629680e+00 -7.97244132e-01
-4.86277938e-01 -5.73268533e-01 -2.80121595e-01 7.35928640e-02
3.45850348e-01 -5.09251021e-02 -7.45548487e-01 -3.18877161e-01
-1.60487574e-02 -1.29870743e-01 6.20793462e-01 8.21023881e-01
1.10725296e+00 -4.41376776e-01 -3.50751281e-01 2.33307719e-01
1.14141726e+00 3.21927696e-01 5.71009398e-01 -2.77576949e-02
5.27138770e-01 7.85075963e-01 9.30593848e-01 8.16188812e-01
3.87582541e-01 8.88909400e-01 4.31250721e-01 -3.29365611e-01
6.47660375e-01 -3.98494929e-01 3.50914806e-01 1.45227209e-01
4.09598202e-01 -2.67862797e-01 -3.58769149e-01 2.64622420e-01
-2.09463668e+00 -8.74866009e-01 -2.25964025e-01 2.46037030e+00
8.88272762e-01 2.51237065e-01 4.93925482e-01 1.17981836e-01
7.82995880e-01 -1.48954228e-01 -8.19898784e-01 -3.73746067e-01
1.25609741e-01 -1.01109453e-01 6.85340047e-01 2.23997027e-01
-1.06273973e+00 5.82430422e-01 5.53871775e+00 9.51969326e-01
-7.94368684e-01 -1.54998660e-01 8.81747425e-01 -1.09354645e-01
-5.24244368e-01 2.80183733e-01 -9.87129450e-01 7.30092645e-01
5.64319193e-01 -4.24002975e-01 4.51185614e-01 7.98314273e-01
6.11549497e-01 3.06300493e-03 -1.30058634e+00 4.94835705e-01
-4.59815234e-01 -9.21461403e-01 -6.45046160e-02 6.78644955e-01
4.60619658e-01 -4.53077465e-01 2.24884599e-01 6.57222271e-01
5.49147904e-01 -7.74148524e-01 3.22329849e-01 3.20948482e-01
3.59982610e-01 -9.16883767e-01 6.82153523e-01 6.54801071e-01
-8.93429399e-01 -6.28488183e-01 -4.45653312e-02 -1.71078503e-01
1.24878868e-01 8.57343793e-01 -6.55109346e-01 4.36412066e-01
3.41790527e-01 8.45260918e-01 -2.98151612e-01 8.20859611e-01
-4.75776120e-04 7.14908898e-01 -2.24595711e-01 9.36844125e-02
1.25419021e-01 -3.21131647e-01 5.79791218e-02 4.93394881e-01
1.15574025e-01 2.59026170e-01 4.87372398e-01 7.07313418e-01
-1.02679282e-01 5.33531249e-01 -2.59098202e-01 -3.11549783e-01
4.69308227e-01 1.35129571e+00 -2.24389821e-01 -7.73113295e-02
-5.68577945e-01 5.18583477e-01 -3.76386866e-02 3.46937269e-01
-7.76982903e-01 -1.56760320e-01 8.58844340e-01 5.00417173e-01
1.18645974e-01 5.27920485e-01 -2.78283834e-01 -1.16448534e+00
-1.13439783e-01 -9.21090662e-01 5.02219141e-01 -4.55923617e-01
-1.79524529e+00 -5.49952127e-02 -1.49600089e-01 -1.18095791e+00
-1.38776645e-01 -3.39248985e-01 -6.03150845e-01 9.86557603e-01
-1.23441255e+00 -1.24710703e+00 -2.08880603e-01 4.67342585e-01
2.44902998e-01 1.11531150e-02 7.12528765e-01 6.98327959e-01
-9.74296033e-01 8.34563851e-01 3.83199118e-02 -7.41726980e-02
8.21003616e-01 -1.07785869e+00 -1.39436513e-01 1.88635454e-01
-3.54371399e-01 7.60637641e-01 6.20562792e-01 -7.91230023e-01
-1.18631041e+00 -1.02620804e+00 9.26701248e-01 -3.26782055e-02
5.74750125e-01 -4.71345782e-01 -9.31289196e-01 8.20286155e-01
-2.97801346e-01 -3.54422897e-01 1.02217209e+00 8.02487373e-01
-2.07764477e-01 -3.19074690e-01 -1.38189673e+00 5.61649382e-01
7.64145494e-01 -1.00290097e-01 3.04326825e-02 6.01702273e-01
7.44200230e-01 -2.31459245e-01 -1.03895915e+00 4.00405258e-01
5.33258796e-01 -8.27378452e-01 7.16883659e-01 -8.57098699e-01
7.46102095e-01 1.74604386e-01 -1.77337125e-01 -1.33656347e+00
-5.78566670e-01 -3.90538573e-01 6.27738610e-02 1.65117764e+00
4.10823077e-01 -7.51552224e-01 8.78403723e-01 1.43622160e+00
3.55867714e-01 -9.18668449e-01 -3.86753947e-01 -3.88124555e-01
-1.01000130e-01 -2.45738178e-01 1.12275577e+00 1.20629680e+00
1.28969550e-01 5.38900614e-01 -5.80630541e-01 -5.61216585e-02
4.14131790e-01 2.11355984e-01 6.85359895e-01 -1.48117280e+00
-5.62479258e-01 -4.76849377e-01 -1.46141887e-01 -9.91009235e-01
1.30425140e-01 -7.18225777e-01 -4.35579240e-01 -1.15405345e+00
7.15452850e-01 -6.21023178e-01 -3.18426818e-01 5.15749931e-01
-3.65675718e-01 -4.54319179e-01 -2.29709572e-03 5.89158088e-02
-4.17874977e-02 4.80158746e-01 1.22619092e+00 -2.20574271e-02
-5.35945117e-01 6.33104980e-01 -1.20617723e+00 6.07839465e-01
6.39550686e-01 -4.33518887e-01 -6.44761443e-01 1.38661042e-01
5.30354232e-02 3.57355595e-01 1.98394731e-01 -7.52336532e-02
-7.02550039e-02 -6.57696724e-01 2.61072010e-01 -4.05706465e-01
7.80872777e-02 -8.53183448e-01 1.49966195e-01 2.81229496e-01
-5.81829011e-01 -3.22050512e-01 -1.03468850e-01 7.17057705e-01
-5.19539565e-02 -1.37383923e-01 5.08715689e-01 1.16220474e-01
-3.36658090e-01 7.36385524e-01 -1.24127967e-02 -1.42845407e-01
9.11559880e-01 5.91719374e-02 -3.23380560e-01 -5.74118555e-01
-6.95778787e-01 6.19712770e-01 2.81728029e-01 5.23108840e-01
2.14411914e-01 -1.51036096e+00 -6.05184734e-01 2.67274797e-01
3.32188874e-01 -3.51152539e-01 5.65826774e-01 7.06635833e-01
3.11429024e-01 8.55778679e-02 1.41045824e-01 -3.43785942e-01
-1.31815445e+00 6.11711025e-01 2.00352877e-01 -5.65235376e-01
-8.08147788e-02 4.81534690e-01 6.51226461e-01 -4.82604235e-01
1.68086112e-01 -1.07986843e-02 -5.05912900e-01 3.64797652e-01
4.31325793e-01 -8.12540483e-03 -1.29861921e-01 -4.48403955e-01
-8.86991546e-02 1.81112662e-01 -5.75598955e-01 2.25309432e-01
1.38260210e+00 -1.10431053e-01 1.60723761e-01 1.73948035e-01
1.12771547e+00 -1.76855519e-01 -1.20708108e+00 -6.98430896e-01
-8.50950778e-02 -8.64139259e-01 3.20698321e-01 -9.49608743e-01
-1.51644647e+00 5.00700474e-01 5.26113510e-01 3.72330934e-01
1.43055201e+00 -2.98834562e-01 8.28674912e-01 -1.78379282e-01
1.41797960e-01 -1.16278243e+00 -1.18219279e-01 1.31242499e-02
5.48947752e-01 -1.37235999e+00 -3.27779650e-04 -7.74582803e-01
-8.26403320e-01 6.19690657e-01 6.64903283e-01 1.37223586e-01
1.00483525e+00 -1.22430427e-02 -1.25076443e-01 -2.04195678e-01
-7.97713220e-01 -1.24042938e-02 2.86279678e-01 2.42088273e-01
5.96076369e-01 4.79676276e-01 -8.04111242e-01 1.24330091e+00
4.73469961e-03 1.73338667e-01 3.55936885e-01 7.15779960e-01
-1.31433919e-01 -1.50693142e+00 4.29368801e-02 9.83050823e-01
-3.70260537e-01 4.70393710e-02 -4.29050326e-01 7.41474032e-01
1.86607599e-01 1.14001787e+00 -1.40907794e-01 -8.10960293e-01
7.63869524e-01 -2.46505991e-01 -4.57534976e-02 -6.48450196e-01
-1.04431129e+00 3.43091846e-01 9.58709493e-02 -6.15226626e-01
-1.36680096e-01 -8.80173206e-01 -1.08773088e+00 -7.19191492e-01
-6.19711041e-01 7.20346943e-02 4.65487808e-01 9.55764592e-01
3.91344070e-01 5.24568856e-01 1.30375433e+00 -4.23111826e-01
-1.00949860e+00 -9.79696512e-01 -1.23319018e+00 9.17183042e-01
-1.27855599e-01 -8.11003506e-01 -3.89909208e-01 -3.01265866e-01] | [9.678936958312988, 5.400411128997803] |
f4c16ac0-1e8d-4bd4-bc1b-0f599c293577 | deep-adaptive-attention-for-joint-facial | 1803.05588 | null | http://arxiv.org/abs/1803.05588v2 | http://arxiv.org/pdf/1803.05588v2.pdf | Deep Adaptive Attention for Joint Facial Action Unit Detection and Face Alignment | Facial action unit (AU) detection and face alignment are two highly
correlated tasks since facial landmarks can provide precise AU locations to
facilitate the extraction of meaningful local features for AU detection. Most
existing AU detection works often treat face alignment as a preprocessing and
handle the two tasks independently. In this paper, we propose a novel
end-to-end deep learning framework for joint AU detection and face alignment,
which has not been explored before. In particular, multi-scale shared features
are learned firstly, and high-level features of face alignment are fed into AU
detection. Moreover, to extract precise local features, we propose an adaptive
attention learning module to refine the attention map of each AU adaptively.
Finally, the assembled local features are integrated with face alignment
features and global features for AU detection. Experiments on BP4D and DISFA
benchmarks demonstrate that our framework significantly outperforms the
state-of-the-art methods for AU detection. | ['Zhilei Liu', 'Jianfei Cai', 'Zhiwen Shao', 'Lizhuang Ma'] | 2018-03-15 | deep-adaptive-attention-for-joint-facial-1 | http://openaccess.thecvf.com/content_ECCV_2018/html/Zhiwen_Shao_Deep_Adaptive_Attention_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Zhiwen_Shao_Deep_Adaptive_Attention_ECCV_2018_paper.pdf | eccv-2018-9 | ['action-unit-detection', 'facial-action-unit-detection'] | ['computer-vision', 'computer-vision'] | [-9.08606648e-02 -6.85249045e-02 -1.07736140e-01 -4.85087216e-01
-1.01858056e+00 -2.50395626e-01 4.15168941e-01 -2.15226740e-01
-3.19127709e-01 -1.53650679e-02 3.13547790e-01 4.78723973e-01
4.36799169e-01 -7.54548550e-01 -5.16137719e-01 -9.01216745e-01
2.04648852e-01 8.09986219e-02 1.25269309e-01 -1.07328467e-01
1.13796946e-02 5.54483414e-01 -1.52954185e+00 1.65326193e-01
4.37187612e-01 1.52035189e+00 -1.33344531e-01 2.77919143e-01
1.93400323e-01 2.90256321e-01 -3.05757105e-01 -2.94138432e-01
4.22359496e-01 -4.70249802e-01 -4.45132375e-01 3.33535403e-01
6.94172978e-01 -8.89759183e-01 -2.66491115e-01 1.03786135e+00
8.10285985e-01 -4.40781303e-02 7.10975587e-01 -1.01844215e+00
-5.24204135e-01 1.41925499e-01 -1.34267032e+00 1.69223085e-01
4.31568861e-01 2.31506079e-01 1.08327818e+00 -1.36886656e+00
2.25098565e-01 1.34286082e+00 3.70074302e-01 6.12436473e-01
-9.84700382e-01 -1.05468547e+00 3.02048266e-01 1.53666466e-01
-1.75746083e+00 -8.31738353e-01 7.97068834e-01 -3.21178585e-01
7.64178991e-01 -3.19053717e-02 5.49843609e-01 7.00865388e-01
-2.02035442e-01 8.80639255e-01 4.06397551e-01 -2.80926377e-01
-2.64904797e-01 -1.92676395e-01 -2.79527575e-01 1.28240991e+00
-8.95306319e-02 -2.50141531e-01 -4.87600833e-01 -6.32974803e-02
9.58384573e-01 5.18755466e-02 -9.15816128e-02 2.70080771e-02
-5.61917543e-01 8.07808042e-01 7.08534420e-01 2.15554893e-01
-3.17507535e-01 2.27070406e-01 3.28764707e-01 -6.69451132e-02
6.68194175e-01 -2.94525027e-01 -1.75239980e-01 1.64527804e-01
-4.58731502e-01 1.03275441e-01 3.89314145e-02 7.89700210e-01
1.10472858e+00 -2.59985596e-01 -7.20400929e-01 8.91927540e-01
6.56146049e-01 3.79201442e-01 2.25918174e-01 -5.37489593e-01
4.57115471e-01 1.07583475e+00 7.67557770e-02 -1.25184214e+00
-4.33059365e-01 7.33348122e-03 -5.35745680e-01 1.52692288e-01
3.31640840e-01 -3.53558928e-01 -6.30870283e-01 1.71500611e+00
7.32976496e-01 5.41689098e-01 -1.47310063e-01 1.18399787e+00
1.08668077e+00 5.13662875e-01 -8.06467608e-02 -1.28585011e-01
1.55937171e+00 -1.14803612e+00 -4.96692508e-01 -3.23342860e-01
5.60717463e-01 -7.68662214e-01 8.58721077e-01 -1.74633209e-02
-1.10320616e+00 -6.39120579e-01 -7.37303138e-01 -3.46877545e-01
7.59989694e-02 8.33608031e-01 5.06515443e-01 4.43993151e-01
-9.30917621e-01 3.00556347e-02 -7.43526220e-01 -2.23955661e-01
8.14877748e-01 6.27480268e-01 -5.34267306e-01 6.49741814e-02
-9.50948775e-01 5.19535303e-01 -2.66516153e-02 5.30782998e-01
-1.11341178e+00 -1.05119795e-01 -1.10099113e+00 1.29945025e-01
4.28489387e-01 -3.02629709e-01 1.25929332e+00 -8.67184937e-01
-1.69511986e+00 1.05132771e+00 -3.43637258e-01 2.01769322e-02
2.92565554e-01 -2.72275567e-01 -3.45630832e-02 2.93135822e-01
1.22449555e-01 7.26385355e-01 1.21858037e+00 -8.06210399e-01
-8.28949451e-01 -7.16390431e-01 -6.06622137e-02 4.35386211e-01
-6.95419550e-01 5.74552178e-01 -8.77516210e-01 -3.26988339e-01
1.15905166e-01 -4.57391262e-01 -4.16829325e-02 4.11736488e-01
-1.28864333e-01 -7.68041313e-01 8.93139124e-01 -5.15606165e-01
1.16378689e+00 -2.18085337e+00 2.33967364e-01 1.89230472e-01
4.88576531e-01 1.83232099e-01 -4.23306525e-01 -1.15196452e-01
1.54247612e-01 -3.22294593e-01 1.02977931e-01 -8.91301394e-01
-1.01165853e-01 -2.01169074e-01 1.42310172e-01 8.60535920e-01
6.73294008e-01 1.00531614e+00 -7.57250130e-01 -5.94324172e-01
1.52117163e-01 3.49280119e-01 -5.74636281e-01 4.55661178e-01
-3.02254758e-03 4.55931365e-01 -6.52248025e-01 1.03812683e+00
6.70156181e-01 1.59651354e-01 -2.66314209e-01 -2.84018606e-01
-1.30438525e-02 -8.41114745e-02 -8.73010516e-01 1.59555399e+00
-3.65288973e-01 3.09645504e-01 3.61529827e-01 -6.52906418e-01
1.20309305e+00 3.50636840e-01 4.58432078e-01 -5.06069005e-01
7.29016423e-01 -5.50136864e-02 -8.15380439e-02 -4.01395142e-01
-4.24834862e-02 1.69001728e-01 1.00761794e-01 5.24319768e-01
9.23871621e-02 3.07508856e-01 -1.56762064e-01 -1.64890379e-01
8.12911153e-01 1.04783483e-01 3.60908210e-01 -7.15833977e-02
1.02902579e+00 -6.89560592e-01 7.95956671e-01 7.46048614e-02
-5.04340351e-01 7.48879731e-01 5.71655750e-01 -4.48044330e-01
-5.59594393e-01 -6.97823703e-01 -1.23333722e-01 1.39342201e+00
3.08706760e-01 -6.87970877e-01 -1.06624711e+00 -1.11404061e+00
9.51739214e-03 -1.87816560e-01 -1.02826869e+00 -2.53685117e-01
-4.49251980e-01 -5.80827951e-01 4.05969203e-01 7.52060413e-01
4.28247064e-01 -1.16860199e+00 -4.68014538e-01 1.45324722e-01
9.33552682e-02 -1.16289091e+00 -1.02140868e+00 -4.48047936e-01
-3.39721441e-01 -1.24223316e+00 -6.22235119e-01 -1.00012112e+00
9.91460681e-01 5.14140368e-01 5.51310003e-01 3.35482448e-01
-5.76262057e-01 2.99032684e-02 -2.84851789e-01 -3.97121072e-01
1.49091631e-01 -2.11774051e-01 1.65395007e-01 7.37082243e-01
8.27158928e-01 -2.94624180e-01 -9.49516594e-01 5.44643641e-01
-4.25264597e-01 -1.55067459e-01 7.67477393e-01 7.30889916e-01
5.67424297e-01 -4.59913254e-01 5.16427517e-01 -3.31430525e-01
4.00734484e-01 -2.00113326e-01 -6.21179998e-01 1.36425287e-01
1.20126836e-01 -2.87099808e-01 2.91160911e-01 -1.99187577e-01
-9.31078136e-01 5.44057012e-01 -3.50221664e-01 -7.37298608e-01
-2.04000071e-01 1.36865363e-01 -6.13649905e-01 -4.36620682e-01
4.43178415e-01 9.55009386e-02 2.77020842e-01 -3.69926155e-01
1.91418618e-01 9.44819987e-01 3.76584768e-01 -5.11924922e-01
7.96483219e-01 6.55221820e-01 -2.19964057e-01 -7.16344953e-01
-1.06141043e+00 -6.40093803e-01 -9.41888869e-01 -2.35539898e-01
1.02613628e+00 -1.23912346e+00 -8.05965602e-01 7.33913839e-01
-1.25218773e+00 -9.28194597e-02 2.01677814e-01 1.02019235e-01
-4.31299150e-01 3.98972422e-01 -6.46173239e-01 -8.05918574e-01
-7.20385909e-01 -1.31270564e+00 1.96043503e+00 4.89439756e-01
-1.52205238e-02 -3.96743000e-01 4.32423642e-03 2.07977384e-01
-3.51625942e-02 2.56051689e-01 1.75170407e-01 -1.58585310e-01
-3.06178212e-01 -5.02440333e-01 -7.10607111e-01 2.11179182e-01
5.33009350e-01 2.33506739e-01 -1.14674866e+00 -3.72816384e-01
-1.55643642e-01 -4.28632528e-01 8.25599313e-01 1.87100306e-01
1.27021241e+00 -2.79998958e-01 -4.01019037e-01 6.93754196e-01
8.52384508e-01 -1.25700891e-01 3.32419515e-01 -1.78934544e-01
8.74070942e-01 4.65999871e-01 9.76933479e-01 8.39784682e-01
1.84876144e-01 9.33040321e-01 6.77114189e-01 -2.13700026e-01
1.61532968e-01 -3.95796187e-02 5.94971061e-01 1.91115975e-01
-2.14868426e-01 3.89183402e-01 -5.35205245e-01 3.89955103e-01
-1.96206844e+00 -7.21841872e-01 2.09144309e-01 2.01263666e+00
7.22705066e-01 -1.56387389e-01 5.59043169e-01 -2.11247608e-01
8.52425218e-01 1.81664199e-01 -4.26583499e-01 -5.77577837e-02
1.86983630e-01 9.30480063e-02 1.26144988e-02 3.01957875e-01
-1.41719866e+00 1.29686117e+00 5.06330442e+00 8.14263344e-01
-1.05137098e+00 3.25932145e-01 7.75061846e-01 -1.91043198e-01
1.92274719e-01 -3.98285359e-01 -1.25462437e+00 2.75050402e-01
2.19142973e-01 3.16215791e-02 1.65464524e-02 1.05506504e+00
1.34661049e-01 1.74956188e-01 -1.07425880e+00 1.26481020e+00
3.57578248e-01 -8.81288946e-01 -9.88975316e-02 1.16617233e-02
6.93942845e-01 -1.55075192e-01 3.65836844e-02 5.93820326e-02
3.01943962e-02 -1.04416287e+00 4.82258648e-01 2.77316779e-01
9.86539781e-01 -1.11207461e+00 7.00327098e-01 -1.56706665e-03
-1.71326840e+00 -2.45931923e-01 -7.61967480e-01 -1.00430049e-01
-8.63626599e-02 -1.02789337e-02 -6.76074028e-01 1.50689900e-01
7.01727152e-01 9.91904199e-01 -6.31799757e-01 7.90476382e-01
-5.26574552e-01 1.70544297e-01 -4.73414689e-01 -4.31743637e-03
4.11775112e-01 -2.93565243e-01 3.46224636e-01 1.02504694e+00
3.17901433e-01 2.93501496e-01 2.85351008e-01 7.74031281e-01
-2.00024560e-01 5.34171760e-01 -3.94550472e-01 2.12686658e-01
1.59350321e-01 1.83274961e+00 -4.05163825e-01 6.46282583e-02
-8.64929676e-01 1.10615301e+00 7.63126016e-01 -1.23726025e-01
-9.00268197e-01 -1.84399232e-01 1.21510136e+00 -6.19418435e-02
2.36253560e-01 -5.49453534e-02 2.55122423e-01 -1.07605195e+00
6.52634352e-02 -6.14009440e-01 6.04888201e-01 -3.49980712e-01
-1.29483330e+00 6.78878188e-01 -4.83158350e-01 -1.27034807e+00
-3.99968363e-02 -6.42402947e-01 -1.03856051e+00 8.33041608e-01
-1.26058197e+00 -1.48794091e+00 -7.80973673e-01 8.08484852e-01
5.69208443e-01 -1.77760050e-01 6.68581903e-01 2.75319338e-01
-1.15119648e+00 1.11837673e+00 -2.79374540e-01 8.21196496e-01
7.42322922e-01 -8.19921613e-01 4.62387085e-01 9.54877377e-01
2.35773981e-01 3.39500189e-01 1.36281252e-01 -4.54284340e-01
-1.32006741e+00 -1.37194586e+00 3.47833693e-01 -4.99588907e-01
6.18910491e-01 -6.19145632e-01 -7.19132721e-01 6.93365574e-01
-1.47240743e-01 5.98939300e-01 4.27711487e-01 -7.08725899e-02
-2.79361874e-01 -3.96377891e-01 -9.03468132e-01 6.92317069e-01
1.25554121e+00 -4.74813193e-01 -1.26859814e-01 4.77713823e-01
2.84622610e-01 -5.68012774e-01 -6.90529227e-01 3.80453318e-01
5.94319165e-01 -8.94066393e-01 6.96952879e-01 -3.57179642e-01
2.04271600e-01 -3.18086237e-01 1.66840971e-01 -8.61917675e-01
-1.79671451e-01 -8.03692818e-01 -1.52563393e-01 1.42287767e+00
1.25509933e-01 -5.30540347e-01 9.50013876e-01 3.43546778e-01
-2.73119435e-02 -9.75949526e-01 -9.44947600e-01 -1.92847788e-01
-3.85168284e-01 -1.18107021e-01 7.61264145e-01 4.59834039e-01
8.65190849e-02 4.68368351e-01 -3.40831995e-01 2.79659778e-01
4.75390881e-01 4.67045397e-01 1.04313004e+00 -1.12871265e+00
-8.92878994e-02 -7.50316083e-01 -8.00479531e-01 -1.08433628e+00
3.47217858e-01 -5.73196650e-01 2.64261901e-01 -1.12634671e+00
3.61611605e-01 -1.90098241e-01 -1.81464940e-01 8.27979147e-01
-5.83325326e-01 7.12338805e-01 -7.41552562e-02 -8.54302868e-02
-6.25590146e-01 9.09965038e-01 1.15252888e+00 -5.72586283e-02
-2.89064944e-01 3.56973633e-02 -6.44865036e-01 8.15454423e-01
6.38826847e-01 -1.59048781e-01 1.17516716e-03 -3.04169923e-01
-3.19429934e-01 -2.61962771e-01 2.47750744e-01 -9.72003520e-01
1.62679106e-01 5.40972538e-02 6.38300419e-01 -4.62965280e-01
4.51608926e-01 -7.32106626e-01 -8.09158266e-01 3.34507704e-01
6.47309795e-02 -7.22334683e-02 1.96151882e-01 4.92503762e-01
-2.46614203e-01 3.51541452e-02 9.14212942e-01 9.31542590e-02
-7.63497412e-01 1.11486864e+00 -4.60299430e-03 -3.43891710e-01
1.48761272e+00 -9.73433908e-03 4.58469475e-03 -2.74651736e-01
-6.23038352e-01 5.27921915e-01 2.66832143e-01 7.25311756e-01
7.48077750e-01 -1.38142180e+00 -1.06171572e+00 5.64920843e-01
2.31166363e-01 3.54489625e-01 2.47698590e-01 1.00934279e+00
-4.05564249e-01 1.20785922e-01 -2.39702001e-01 -7.21455693e-01
-1.89373147e+00 5.08992255e-01 5.23780227e-01 3.10445637e-01
-4.46275234e-01 1.35502517e+00 6.58060372e-01 -1.00353263e-01
3.61092329e-01 6.02555610e-02 -4.59031314e-01 3.20285320e-01
1.08438456e+00 -9.51555073e-02 -9.56193916e-03 -1.10095334e+00
-3.96119237e-01 1.02624679e+00 -3.18670869e-01 3.86780113e-01
1.22281742e+00 -2.04199672e-01 -3.51787090e-01 -2.11364836e-01
1.35682547e+00 7.33769983e-02 -1.59246004e+00 -5.68549991e-01
-3.09861720e-01 -6.98464215e-01 -7.35658780e-02 1.62451584e-02
-1.42016804e+00 1.18375766e+00 5.47783852e-01 -4.21896040e-01
1.31769109e+00 1.84392273e-01 7.03929126e-01 6.12009056e-02
3.18927050e-01 -7.79027641e-01 4.07564521e-01 1.59389734e-01
1.06250536e+00 -1.36974418e+00 -1.78578749e-01 -6.93408191e-01
-5.23677945e-01 1.14425623e+00 1.46054435e+00 -2.28884533e-01
4.57247108e-01 2.41273746e-01 9.89620909e-02 -1.89536795e-01
-4.52433944e-01 -6.62710547e-01 5.86244106e-01 3.18387330e-01
3.27112406e-01 -3.94895166e-01 -7.83554390e-02 8.80618870e-01
2.24870339e-01 -3.76044780e-01 -1.67685226e-01 6.38001144e-01
-6.92593932e-01 -9.18948948e-01 -2.60089576e-01 3.37807804e-01
-4.41312730e-01 2.55047008e-02 -7.54191875e-01 5.66048563e-01
1.64205432e-01 7.27000415e-01 2.68144429e-01 -4.82269794e-01
1.01343319e-01 -1.98036715e-01 5.42427242e-01 -9.49299395e-01
-4.88269210e-01 3.90414029e-01 -3.05063009e-01 -8.71021271e-01
-9.88400131e-02 -7.67516613e-01 -1.26213384e+00 9.00028720e-02
-4.88134444e-01 -3.20483893e-02 1.67729110e-01 1.01278079e+00
3.88721377e-01 5.97836636e-02 9.52025294e-01 -1.31734633e+00
-3.27136934e-01 -1.15588975e+00 -3.89199048e-01 3.23735237e-01
3.27355474e-01 -9.07319784e-01 -1.60075799e-01 -4.82778728e-01] | [13.619840621948242, 1.467418909072876] |
411381c0-49a8-41be-b28b-e0aa4351a355 | affordance-learning-in-direct-perception-for | 1903.08746 | null | http://arxiv.org/abs/1903.08746v1 | http://arxiv.org/pdf/1903.08746v1.pdf | Affordance Learning In Direct Perception for Autonomous Driving | Recent development in autonomous driving involves high-level computer vision
and detailed road scene understanding. Today, most autonomous vehicles are
using mediated perception approach for path planning and control, which highly
rely on high-definition 3D maps and real time sensors. Recent research efforts
aim to substitute the massive HD maps with coarse road attributes. In this
paper, we follow the direct perception based method to train a deep neural
network for affordance learning in autonomous driving. Our goal in this work is
to develop the affordance learning model based on freely available Google
Street View panoramas and Open Street Map road vector attributes. Driving scene
understanding can be achieved by learning affordances from the images captured
by car-mounted cameras. Such scene understanding by learning affordances may be
useful for corroborating base maps such as HD maps so that the required data
storage space is minimized and available for processing in real time. We
compare capability in road attribute identification between human volunteers
and our model by experimental evaluation. Our results indicate that this method
could act as a cheaper way for training data collection in autonomous driving.
The cross validation results also indicate the effectiveness of our model. | ['Jean M. Uwabeza Vianney', 'Chen Sun', 'Dongpu Cao'] | 2019-03-20 | null | null | null | null | ['road-scene-understanding'] | ['computer-vision'] | [ 1.43101737e-01 3.75745296e-01 -5.86432330e-02 -8.62211287e-01
-1.93946674e-01 -2.39384085e-01 7.56486475e-01 4.84369975e-03
-5.14359713e-01 5.07442892e-01 -6.06203489e-02 -7.81101942e-01
-2.51189083e-01 -1.20353401e+00 -8.76977742e-01 -2.91046828e-01
2.67584771e-02 5.20657420e-01 4.26711619e-01 -7.99146116e-01
4.74580139e-01 8.18122506e-01 -2.32111192e+00 -8.71870071e-02
1.08956313e+00 8.49990964e-01 8.00567925e-01 7.93166697e-01
-4.65389341e-02 4.10248965e-01 1.34058475e-01 -1.74949188e-02
4.30719823e-01 3.53761733e-01 -2.97872007e-01 1.62577424e-02
6.17107511e-01 -6.58885539e-01 -4.52556551e-01 7.55305588e-01
3.39736402e-01 2.25726366e-01 5.73493421e-01 -1.50447965e+00
-3.62953901e-01 -3.23748915e-03 2.54366547e-02 3.76827605e-02
4.42972422e-01 3.59770209e-01 6.88880980e-01 -6.65917814e-01
3.88846070e-01 1.24477017e+00 5.10247171e-01 4.54920679e-01
-8.83343577e-01 -5.75808227e-01 3.18586789e-02 7.67874241e-01
-1.39852154e+00 -4.07552332e-01 8.46599340e-01 -4.79262114e-01
1.01057661e+00 3.15994471e-01 6.23792112e-01 7.29497850e-01
4.27326798e-01 5.28607488e-01 1.44366455e+00 -3.38921607e-01
7.26994202e-02 4.05631274e-01 1.45056948e-01 8.99617016e-01
3.64726096e-01 6.01982713e-01 -4.94832307e-01 4.27029908e-01
7.44475484e-01 1.01253414e-03 6.45082071e-02 -6.72356486e-01
-1.23456109e+00 9.10114110e-01 9.00609672e-01 -2.48239025e-01
-4.30526257e-01 -2.12194175e-02 2.53979210e-02 2.35342935e-01
1.22365378e-01 3.76310140e-01 -2.50393271e-01 1.30841332e-02
-3.37473571e-01 5.21688700e-01 4.84975398e-01 1.04630697e+00
1.30661082e+00 -8.27143341e-02 3.67959559e-01 5.52826285e-01
5.39891958e-01 9.25677001e-01 2.06697464e-01 -1.17257905e+00
3.27125251e-01 5.85263729e-01 1.38447881e-01 -1.04706311e+00
-6.81549430e-01 1.99483514e-01 -3.55601668e-01 8.34240317e-01
3.64831448e-01 2.94830054e-01 -1.17276239e+00 1.24297178e+00
3.01635265e-01 2.56524198e-02 1.59900084e-01 1.09226906e+00
8.03035676e-01 3.97862405e-01 1.38721824e-01 5.88804305e-01
1.28742480e+00 -6.41847312e-01 -6.46066606e-01 -1.92213535e-01
7.81600058e-01 -5.33450425e-01 1.43551147e+00 3.08346331e-01
-3.07003856e-01 -1.03161860e+00 -1.57373369e+00 -3.26228380e-01
-9.89724815e-01 7.59973079e-02 9.15370524e-01 8.49534452e-01
-9.97003794e-01 4.28540319e-01 -7.10936844e-01 -6.45918727e-01
5.25048316e-01 3.46862197e-01 -5.92228770e-01 -3.55166674e-01
-1.30710101e+00 1.40120113e+00 4.31325316e-01 6.59527481e-02
-1.04173601e+00 -3.42357874e-01 -1.14185095e+00 -4.19711113e-01
3.26345861e-01 -5.78617334e-01 9.35676992e-01 -1.46551862e-01
-1.30477738e+00 8.10172141e-01 -1.37878731e-01 -5.92605889e-01
4.87985283e-01 -3.70027989e-01 -6.13211989e-01 -1.22929916e-01
5.03968358e-01 1.10825455e+00 5.65251052e-01 -1.33908057e+00
-9.96670902e-01 -5.54417670e-01 2.98282266e-01 4.03500617e-01
2.63911247e-01 -7.43550658e-01 8.23512152e-02 3.51069838e-01
5.99092066e-01 -1.15252352e+00 -3.82810295e-01 1.08760595e-01
-3.65972668e-01 -3.66077632e-01 9.80829000e-01 -4.09083247e-01
4.32820290e-01 -1.85263550e+00 -5.89722097e-01 2.96302795e-01
3.77775848e-01 -1.36034973e-02 1.31805211e-01 1.58613294e-01
3.78408581e-01 -1.37595832e-01 -4.77512144e-02 2.75632322e-01
1.14623971e-01 5.15147090e-01 -2.17048302e-01 3.99547547e-01
1.38569951e-01 7.55526721e-01 -6.60185039e-01 -4.19039220e-01
8.59465599e-01 4.56113219e-01 -2.77589858e-01 2.72434294e-01
-7.07777739e-02 5.69549143e-01 -5.54405510e-01 5.02054811e-01
8.71087790e-01 5.98037302e-01 -4.49732006e-01 -1.05539791e-01
-5.00160754e-01 5.55605531e-01 -1.09015858e+00 1.83816469e+00
-5.49253762e-01 1.02485096e+00 -4.11329836e-01 -8.01653147e-01
1.32089448e+00 -1.68712139e-01 -2.46517919e-02 -1.13663960e+00
-4.73343320e-02 3.39007020e-01 8.86609256e-02 -8.42509568e-01
7.61531174e-01 4.18607751e-03 -4.46993262e-02 9.47056487e-02
-3.59263599e-01 -6.23058558e-01 -4.31780256e-02 -1.51335001e-01
4.58619028e-01 6.44077539e-01 9.34713259e-02 -3.86413366e-01
6.36573076e-01 4.87884998e-01 8.07602108e-02 6.08662724e-01
-3.82328600e-01 2.46608153e-01 -8.66247565e-02 -1.02580953e+00
-1.46823978e+00 -1.19757998e+00 -5.30156612e-01 9.78731632e-01
5.28314769e-01 9.03980136e-02 -7.02148020e-01 -1.17176354e-01
5.75002395e-02 9.56899464e-01 -4.27811623e-01 -1.14336880e-02
-5.66339910e-01 -3.46829236e-01 4.39934999e-01 5.05395412e-01
7.95855761e-01 -9.66206908e-01 -1.20747280e+00 1.61845863e-01
-8.02085996e-02 -1.21130097e+00 2.30678901e-01 -6.05261065e-02
-7.36181617e-01 -1.03696632e+00 -1.90433547e-01 -4.70274270e-01
3.46939385e-01 7.32150614e-01 5.72703958e-01 -9.44372043e-02
-2.61690110e-01 1.75763533e-01 -1.71018332e-01 -9.86794174e-01
-3.98005068e-01 1.77219421e-01 4.17448193e-01 -9.02231932e-02
8.65848184e-01 -5.35621166e-01 -6.23763978e-01 5.86437106e-01
-3.57208937e-01 4.17493492e-01 7.91969121e-01 3.65538388e-01
5.82928419e-01 -1.59077287e-01 5.60010254e-01 -5.59441090e-01
4.71002758e-01 -3.22132975e-01 -6.90762401e-01 -3.04526120e-01
-9.47746098e-01 1.95526168e-01 1.08122341e-01 1.58388063e-01
-1.10313821e+00 3.33234072e-01 -4.63339180e-01 -1.30735454e-03
-8.37973535e-01 1.24800861e-01 -4.01354283e-01 -1.96370140e-01
1.09142292e+00 1.65452972e-01 3.65301281e-01 -1.99960709e-01
6.04620993e-01 9.92953002e-01 6.30446017e-01 -2.74436951e-01
6.82269990e-01 7.42555797e-01 2.80577153e-01 -1.10888028e+00
-4.93892312e-01 -4.67849165e-01 -1.16425920e+00 -4.73158807e-01
1.13182354e+00 -9.93044674e-01 -8.70951533e-01 2.52557456e-01
-9.74273622e-01 -2.03486905e-01 5.96460402e-02 7.58359134e-01
-9.48910654e-01 5.35684861e-02 1.70603797e-01 -7.67780960e-01
2.92839050e-01 -1.24840176e+00 9.32951689e-01 2.84837812e-01
-5.97953536e-02 -6.10383213e-01 -4.35439982e-02 6.66590214e-01
3.71777505e-01 3.87345672e-01 6.69955075e-01 -1.81617066e-01
-1.10805750e+00 -3.49051565e-01 -4.51719314e-01 8.50068219e-03
9.94484723e-02 -4.29294139e-01 -1.45206714e+00 3.51403981e-01
-1.85507849e-01 -2.05958098e-01 8.08592796e-01 3.48068029e-01
9.26458061e-01 1.97470292e-01 -4.97355253e-01 5.69711447e-01
1.28409612e+00 2.71111459e-01 8.66173148e-01 7.83248901e-01
1.10324991e+00 1.17032099e+00 1.16311848e+00 -7.44075999e-02
9.61128533e-01 5.88497043e-01 7.51151741e-01 -1.14433691e-01
2.40063667e-02 -5.75368345e-01 -1.06305229e-02 2.78176993e-01
-4.13024694e-01 3.85873914e-01 -1.26435518e+00 5.36027312e-01
-1.80530739e+00 -7.86707759e-01 -6.62193000e-01 2.25612211e+00
1.10764638e-01 4.13206577e-01 -6.06229566e-02 3.96533459e-01
3.77353907e-01 -1.66177154e-01 -6.75791621e-01 -6.80264235e-01
-1.64976588e-03 -3.47539306e-01 8.97864759e-01 7.48304367e-01
-1.12247407e+00 9.71745312e-01 5.66725922e+00 3.79372478e-01
-1.00934446e+00 1.25588486e-02 3.45791340e-01 3.63851249e-01
-3.51231575e-01 -4.91964705e-02 -9.10774827e-01 1.08902920e-02
1.22445881e+00 1.52253509e-01 3.51401061e-01 1.06733286e+00
5.24866760e-01 -4.91186559e-01 -1.03065097e+00 9.68487740e-01
-1.08276889e-01 -1.32100856e+00 -6.99787661e-02 3.66003066e-01
4.10147518e-01 3.19381297e-01 -1.12223953e-01 2.42586166e-01
2.27915779e-01 -1.25321543e+00 5.72336912e-01 7.42207527e-01
6.87640190e-01 -8.05759788e-01 7.17815220e-01 6.93794131e-01
-1.16046178e+00 -1.18486337e-01 -7.26227999e-01 -5.85208118e-01
4.71777990e-02 1.68339029e-01 -1.38127685e+00 4.34878647e-01
7.26798952e-01 6.03757679e-01 -5.90383291e-01 7.92732954e-01
-9.90356207e-02 -3.94484028e-03 -4.21814501e-01 -3.40885252e-01
3.09144408e-01 -2.47888327e-01 4.06337261e-01 8.43811452e-01
2.27391973e-01 -2.96258507e-03 1.73297361e-01 7.94367015e-01
6.07089400e-01 1.41981954e-03 -1.44203579e+00 4.07796949e-01
3.87372375e-01 9.88962114e-01 -5.55671871e-01 -2.12816969e-01
-4.00594592e-01 4.22263414e-01 1.24348558e-01 1.53995484e-01
-6.97280765e-01 -5.79565883e-01 8.46902013e-01 6.78438365e-01
2.83151977e-02 -7.18051970e-01 -6.84997022e-01 -5.69174170e-01
-5.55442385e-02 -1.91393226e-01 -2.68476844e-01 -1.06815732e+00
-6.58768654e-01 7.20896006e-01 3.17637473e-01 -1.38387728e+00
-2.71801472e-01 -9.80021715e-01 -2.62752444e-01 1.02758718e+00
-1.94856811e+00 -1.30396008e+00 -8.62388551e-01 7.06627965e-01
5.33567727e-01 -2.98939466e-01 8.06595027e-01 4.42625433e-02
1.43478602e-01 1.26046032e-01 -3.22029531e-01 -2.82683074e-01
3.88156325e-01 -1.16752613e+00 5.73517025e-01 5.68923116e-01
2.39978936e-02 2.11748973e-01 9.44465220e-01 -5.42715013e-01
-1.31560576e+00 -9.16908145e-01 8.77237320e-01 -9.57876027e-01
2.45394379e-01 -4.96058166e-01 -7.89686799e-01 4.62404400e-01
-3.97614669e-04 -1.79581091e-01 4.08147335e-01 2.22570542e-02
-9.81357396e-02 -4.20414150e-01 -1.14540374e+00 5.62492013e-01
1.18116581e+00 -5.54290295e-01 -6.41464233e-01 1.69042826e-01
7.69112527e-01 -3.55672449e-01 -5.22402585e-01 4.58473891e-01
6.87581301e-01 -1.12495685e+00 9.17813778e-01 -3.21131170e-01
2.03213587e-01 -5.83522975e-01 -5.49180090e-01 -1.05824757e+00
-2.06562340e-01 -2.73183528e-02 4.18813586e-01 5.04279971e-01
5.47492385e-01 -7.76502073e-01 8.72866213e-01 5.96745849e-01
-5.32165349e-01 -3.97747248e-01 -8.85565698e-01 -6.06089115e-01
-1.65908381e-01 -1.06510746e+00 9.08168972e-01 5.77313185e-01
-2.13814586e-01 4.76779163e-01 -1.12575412e-01 4.03016210e-01
7.16951191e-01 -1.28996819e-01 1.14224422e+00 -1.68104112e+00
3.87222081e-01 -1.20941594e-01 -9.01202500e-01 -9.64371145e-01
1.46193355e-01 -6.93435550e-01 3.17204088e-01 -1.80612862e+00
-3.60404909e-01 -9.46859717e-01 1.26503864e-02 2.33921185e-01
1.23951703e-01 2.05372497e-01 -1.16276316e-01 1.91981733e-01
-6.35695085e-02 4.80107099e-01 1.54184389e+00 7.01677846e-03
-2.37974182e-01 1.05854325e-01 -4.72658813e-01 6.53200924e-01
1.10456038e+00 -1.91756591e-01 -8.18371534e-01 -2.33473480e-01
1.39511645e-01 -2.28560597e-01 6.88469291e-01 -1.31632376e+00
1.61121443e-01 -2.34925866e-01 1.91882670e-01 -9.65743065e-01
5.63490927e-01 -9.73572850e-01 -1.53279841e-01 4.31212813e-01
1.56050930e-02 -8.53270069e-02 2.58001417e-01 6.68961406e-01
5.04832901e-02 1.05473660e-01 4.40650403e-01 -2.18875811e-01
-1.65189004e+00 2.97506720e-01 -4.62720424e-01 -6.29324436e-01
1.15192282e+00 -8.95678639e-01 -1.81571662e-01 -4.27939773e-01
-7.17236817e-01 2.05967024e-01 3.22802663e-01 7.57958055e-01
9.18310642e-01 -1.39528739e+00 -7.25303769e-01 7.15379536e-01
5.95717013e-01 7.56988376e-02 1.89986199e-01 5.39073586e-01
-9.19214666e-01 9.47286546e-01 -9.56829369e-01 -9.64797437e-01
-8.77412796e-01 5.60583889e-01 2.55587190e-01 5.20372450e-01
-7.37648487e-01 4.44457412e-01 2.46133536e-01 -9.58557427e-01
-2.42116809e-01 -3.69136572e-01 -5.81363857e-01 -4.50983435e-01
3.73183787e-01 4.82143402e-01 3.18872631e-02 -1.00384152e+00
-2.85776585e-01 6.38461232e-01 1.53559983e-01 -3.24164510e-01
9.98226583e-01 -6.00490212e-01 4.09381300e-01 6.30364776e-01
8.43679190e-01 -4.42923963e-01 -1.63590407e+00 1.39209971e-01
4.67266850e-02 -6.12679005e-01 2.02001229e-01 -4.90594178e-01
-3.19207400e-01 1.25630474e+00 1.14144754e+00 1.59511954e-01
6.92083955e-01 2.94440370e-02 4.79333311e-01 8.46797168e-01
7.13667631e-01 -1.18310177e+00 -3.74864131e-01 3.22458982e-01
7.66045451e-01 -1.92879248e+00 -1.18245557e-01 -5.84255874e-01
-7.47490525e-01 9.64448392e-01 7.34064817e-01 -2.28416145e-01
6.83115840e-01 -8.15422311e-02 3.98816586e-01 -3.12964797e-01
-3.40076447e-01 -5.85249245e-01 3.78615648e-01 1.28085709e+00
5.88335171e-02 5.52266836e-01 1.07696258e-01 -2.60103568e-02
-7.84018934e-01 -9.93748158e-02 6.75625980e-01 6.85779512e-01
-1.08507001e+00 -8.73036861e-01 -3.84802490e-01 4.61077452e-01
3.53230506e-01 1.29977584e-01 9.98510793e-02 9.91531789e-01
4.17617738e-01 1.04621685e+00 1.87293351e-01 -5.93114018e-01
6.29976094e-01 9.48078707e-02 4.09020334e-01 -3.23978245e-01
5.10243416e-01 -5.05277574e-01 3.92167687e-01 -7.30163276e-01
-3.44499826e-01 -6.76092565e-01 -1.37958753e+00 -4.28767055e-01
-3.92218903e-02 -4.03207213e-01 1.33651090e+00 1.04562664e+00
2.57415265e-01 3.50641698e-01 5.74298561e-01 -1.06124043e+00
-6.57629520e-02 -8.53469491e-01 -4.39128309e-01 1.41859218e-01
5.92742026e-01 -1.07632637e+00 5.20859919e-02 5.85156642e-02] | [8.004969596862793, -1.9281450510025024] |
d7a532e2-30f6-4b81-a3d3-599b924530de | x-torch-differentiable-scientific-computing | 2010.01921 | null | https://arxiv.org/abs/2010.01921v1 | https://arxiv.org/pdf/2010.01921v1.pdf | $ξ$-torch: differentiable scientific computing library | Physics-informed learning has shown to have a better generalization than learning without physical priors. However, training physics-informed deep neural networks requires some aspect of physical simulations to be written in a differentiable manner. Unfortunately, some operations and functionals commonly used in physical simulations are scattered, hard to integrate, and lack higher order derivatives which are needed in physical simulations. In this work, we present $\xi$-torch, a library of differentiable functionals for scientific simulations. Example functionals are a root finder and an initial value problem solver, among others. The gradient of functionals in $\xi$-torch are written based on their analytical expression to improve numerical stability and reduce memory requirements. $\xi$-torch also provides second and higher order derivatives of the functionals which are rarely available in existing packages. We show two applications of this library in optimizing parameters in physics simulations. The library and all test cases in this work can be found at https://github.com/xitorch/xitorch/ and the documentation at https://xitorch.readthedocs.io. | ['Sam M. Vinko', 'Muhammad F. Kasim'] | 2020-10-05 | null | null | null | null | ['physical-simulations'] | ['miscellaneous'] | [-6.81997657e-01 -2.06852555e-01 1.60615176e-01 -5.10762453e-01
-5.87304354e-01 -4.51369375e-01 1.59271225e-01 -1.93399251e-01
-3.19804549e-01 1.38084102e+00 -5.74386954e-01 -5.80460846e-01
-2.63946295e-01 -7.85305262e-01 -1.06492567e+00 -1.06493914e+00
-3.39562684e-01 3.05849135e-01 9.77787003e-02 -2.22730786e-01
3.44942451e-01 7.47180462e-01 -1.26644111e+00 -2.98701197e-01
9.35712993e-01 8.91663909e-01 2.29838088e-01 8.07295501e-01
8.45673587e-03 3.79692912e-01 -4.28303152e-01 1.35236057e-02
2.96073347e-01 -5.72054923e-01 -4.91730988e-01 -8.43914270e-01
2.85968363e-01 -1.67554781e-01 -4.67682898e-01 1.11602366e+00
6.85641468e-01 6.31378412e-01 7.72721589e-01 -1.25633109e+00
-5.83163440e-01 5.81553042e-01 -1.93240076e-01 4.06994611e-01
-1.26250476e-01 5.81377029e-01 5.20814955e-01 -8.72480690e-01
2.57713497e-01 1.20126486e+00 9.11687434e-01 5.61124086e-01
-1.00127256e+00 -8.50571930e-01 -4.35316898e-02 1.51521340e-01
-1.35316086e+00 -2.25300342e-01 5.79165280e-01 -3.76911938e-01
1.24520969e+00 2.28389472e-01 7.32640803e-01 8.52097750e-01
6.79228663e-01 1.83339909e-01 9.24807608e-01 -1.72611654e-01
5.40396690e-01 1.12021424e-01 5.19044161e-01 9.60214674e-01
3.16142946e-01 5.72894871e-01 -4.86354768e-01 -1.97910130e-01
1.12739050e+00 -2.74141610e-01 -3.12877417e-01 -2.32775375e-01
-8.05802047e-01 8.82698178e-01 5.51504254e-01 -1.25374451e-01
-1.98611647e-01 7.38229930e-01 1.63869873e-01 1.52764753e-01
4.18091476e-01 5.81511319e-01 -5.67876399e-01 -2.34656677e-01
-6.21519506e-01 6.44751012e-01 1.09193885e+00 8.20400476e-01
9.79810834e-01 4.15102571e-01 2.33158693e-01 6.52376950e-01
2.99617171e-01 7.23820090e-01 1.62230745e-01 -1.56660342e+00
-1.43480316e-01 -6.41288832e-02 2.43202686e-01 -3.58301252e-01
-5.43123186e-01 -4.69646394e-01 -8.80783319e-01 6.28292680e-01
4.07110512e-01 -7.31351554e-01 -8.94358695e-01 1.39729238e+00
3.90454561e-01 4.48919952e-01 -7.09459037e-02 1.06439865e+00
1.15373969e+00 1.06813657e+00 -1.52094681e-02 -4.79338355e-02
1.04825985e+00 -8.40366840e-01 -2.61627525e-01 5.79305515e-02
5.71442187e-01 -6.50787354e-01 1.07691300e+00 2.62206912e-01
-1.40653837e+00 -4.39041346e-01 -1.19325721e+00 -1.40045539e-01
-4.92311835e-01 -4.73506264e-02 9.58338797e-01 3.88431728e-01
-1.18307865e+00 1.57215083e+00 -1.29614794e+00 -7.06788339e-03
3.27982128e-01 4.73869771e-01 2.19572708e-01 6.45222604e-01
-1.37621593e+00 1.05792356e+00 2.48430833e-01 -1.60759345e-01
-8.66188943e-01 -1.34264362e+00 -7.59109855e-01 1.13026932e-01
1.56249732e-01 -7.71942258e-01 1.67461324e+00 -4.97063369e-01
-2.13737178e+00 2.98208982e-01 -1.10866472e-01 -3.54890704e-01
4.25682157e-01 -2.09955201e-01 -1.49364844e-02 2.44161231e-03
-2.72295088e-01 5.29759884e-01 2.94373095e-01 -8.75199914e-01
1.40067533e-01 2.60443270e-01 1.04905888e-01 1.95256233e-01
9.93892029e-02 -2.24257514e-01 -3.85849386e-01 -2.71622926e-01
-3.56867999e-01 -1.00699162e+00 -3.92020881e-01 2.53875494e-01
-3.35082680e-01 -2.19689563e-01 7.16354012e-01 -5.08946896e-01
5.58653057e-01 -1.78422928e+00 -6.69984818e-02 4.62448001e-02
1.97038241e-02 3.18172932e-01 2.15568766e-01 5.28970838e-01
-3.05112958e-01 1.14009477e-01 -4.73484427e-01 -1.62208050e-01
3.14109355e-01 1.16885871e-01 -1.47431716e-01 5.45090556e-01
-8.01883172e-03 7.94222653e-01 -7.35929728e-01 -2.06211820e-01
3.99705470e-01 8.17132115e-01 -4.99008358e-01 -3.49606872e-02
-5.03294647e-01 6.44649565e-01 -6.99675381e-01 2.19190761e-01
7.82595277e-01 -4.58483934e-01 -3.16504270e-01 -6.92358315e-02
-4.54242289e-01 6.98499501e-01 -1.16206074e+00 1.49294472e+00
-5.07680297e-01 6.11994743e-01 2.66505390e-01 -1.14885092e+00
6.71276510e-01 1.22014627e-01 5.65700054e-01 -2.31737792e-01
2.71459311e-01 1.97638616e-01 1.38535574e-01 -1.33653939e-01
2.18252897e-01 -1.85173228e-01 3.61295968e-01 5.07531524e-01
-2.01666076e-02 -6.97617471e-01 4.34084386e-01 1.85869172e-01
9.32812214e-01 5.68468630e-01 1.55966952e-01 -8.09171796e-01
4.29394990e-01 2.81993151e-01 5.56951880e-01 8.05313170e-01
-2.69299597e-02 3.62998396e-01 4.49872881e-01 -4.03359354e-01
-1.17397869e+00 -1.01942253e+00 -5.72168589e-01 8.57932210e-01
4.89720814e-02 -2.68547863e-01 -8.70349228e-01 1.19661026e-01
2.89100140e-01 1.11807835e+00 -3.52775663e-01 -1.62672058e-01
-6.46520913e-01 -8.86240244e-01 2.80930638e-01 6.43005729e-01
5.17610371e-01 -1.28666961e+00 -7.02078879e-01 4.60560322e-02
5.43785334e-01 -5.25691628e-01 -2.52421260e-01 3.94429773e-01
-9.76062477e-01 -9.01271105e-01 -6.57778144e-01 -2.99209952e-01
3.85669023e-01 -2.30369270e-01 1.18259573e+00 4.15045470e-01
-4.63614851e-01 1.19231448e-01 1.21289015e-01 -5.70933402e-01
-2.79777914e-01 9.51975510e-02 1.56579867e-01 -9.68780935e-01
-8.97383913e-02 -8.87678266e-01 -7.68484056e-01 1.84924424e-01
-4.58367586e-01 1.26358435e-01 3.23403656e-01 7.30717778e-01
5.19232512e-01 5.81782386e-02 2.85719156e-01 -6.87401652e-01
4.01428729e-01 -3.49173188e-01 -1.20890164e+00 -1.42899528e-01
-4.77862418e-01 3.35070938e-01 8.66930902e-01 -4.81981248e-01
-1.01405811e+00 -8.02557915e-02 -6.50974572e-01 -4.74389225e-01
-9.18360949e-02 5.42500496e-01 4.33367267e-02 -4.32960540e-01
7.11214840e-01 -4.16188128e-02 -1.10791720e-01 -7.38547087e-01
8.56721550e-02 5.55318147e-02 2.09932938e-01 -9.61640596e-01
6.70734644e-01 -1.40581187e-02 1.93956599e-01 -9.03373897e-01
-7.88717091e-01 -6.57269657e-02 -3.49794358e-01 -6.78411722e-02
5.42518139e-01 -5.64146221e-01 -1.22530580e+00 6.84197068e-01
-1.00493598e+00 -1.05084097e+00 -3.77160877e-01 8.27274859e-01
-5.53804219e-01 8.78782049e-02 -7.99344778e-01 -7.60347068e-01
-4.40577686e-01 -1.45744050e+00 5.91757178e-01 7.98053086e-01
5.16137481e-03 -1.28024757e+00 5.31589612e-02 -3.37042987e-01
6.00937068e-01 1.03680193e-01 8.12236786e-01 -1.83036745e-01
-5.59536457e-01 8.09814706e-02 1.18167175e-03 1.70602322e-01
-3.57527763e-01 5.70875287e-01 -8.09727430e-01 -2.18281493e-01
8.55780840e-02 -1.60743743e-01 8.40926230e-01 9.20438826e-01
1.56076980e+00 -2.48298511e-01 -4.44779962e-01 1.11995196e+00
1.31866109e+00 3.61457199e-01 4.85623091e-01 1.18446134e-01
5.38132250e-01 1.53231010e-01 1.43214345e-01 4.37248528e-01
-2.69843405e-03 4.09773439e-01 1.16618894e-01 6.70776644e-04
8.28557685e-02 1.77086577e-01 4.41185653e-01 8.81119668e-01
-1.58703282e-01 -1.10782698e-01 -1.13246214e+00 -3.24084833e-02
-1.65460014e+00 -8.71480763e-01 -4.78545874e-01 2.03721070e+00
1.21016765e+00 1.87244534e-01 -1.55631170e-01 -2.55003959e-01
5.47102571e-01 -1.13385580e-01 -1.24894309e+00 -6.25888884e-01
2.99988717e-01 6.31041110e-01 9.42700505e-01 9.28283453e-01
-8.53683412e-01 8.44282448e-01 6.25385618e+00 9.38012481e-01
-1.41191185e+00 1.67299658e-01 5.61151206e-01 -2.85299778e-01
2.92521305e-02 2.79973090e-01 -1.02686524e+00 5.29230654e-01
1.31559718e+00 -5.29644251e-01 5.38674414e-01 1.08759141e+00
4.98302937e-01 -3.82080019e-01 -1.04621291e+00 7.18949080e-01
-6.65382504e-01 -1.52329969e+00 -4.91305709e-01 5.09392284e-02
5.76926231e-01 3.59546363e-01 -9.13346633e-02 4.33239937e-01
6.68778479e-01 -1.20444191e+00 4.79667008e-01 8.29597294e-01
6.62269235e-01 -8.96377325e-01 4.89641935e-01 2.80340314e-01
-9.36562777e-01 4.42726433e-01 -6.92467332e-01 -3.12924773e-01
2.63514936e-01 1.00534546e+00 -2.71729499e-01 3.40107828e-01
8.16458762e-01 8.30427051e-01 -1.37912124e-01 1.25463283e+00
-1.75059587e-01 7.47818768e-01 -6.41100883e-01 -5.19017696e-01
1.30817384e-01 -4.82400477e-01 5.93756080e-01 1.10364771e+00
3.49820495e-01 5.30894697e-01 -4.20949273e-02 1.32741976e+00
1.98236316e-01 -2.06554025e-01 -2.73150295e-01 -1.50851766e-02
3.96396101e-01 1.33355820e+00 -5.92736244e-01 -4.07439053e-01
-1.18457591e-02 3.32896173e-01 1.58325538e-01 6.07185006e-01
-1.30808711e+00 -5.97866118e-01 1.00998044e+00 1.29397679e-02
1.75042033e-01 -7.22821414e-01 -4.17264909e-01 -1.04551840e+00
-2.97152370e-01 -5.73115587e-01 -9.69780162e-02 -1.10130441e+00
-1.23416543e+00 3.16130012e-01 5.44132292e-01 -6.41536951e-01
-8.69939700e-02 -1.13932252e+00 -9.73905027e-01 1.23933268e+00
-1.12917495e+00 -2.24603266e-01 -3.05565357e-01 2.57021308e-01
1.97767973e-01 3.81332524e-02 7.43452847e-01 1.18467405e-01
-8.21850419e-01 5.07740676e-01 4.43725854e-01 -5.34997955e-02
3.11892211e-01 -1.24547029e+00 6.69169068e-01 4.77061450e-01
-5.44133306e-01 8.64328265e-01 1.16714561e+00 -7.22770333e-01
-1.62780809e+00 -7.58248091e-01 2.41537899e-01 1.10738553e-01
7.97957242e-01 -2.18587756e-01 -1.28142405e+00 6.09255731e-01
3.65479171e-01 -1.08684160e-01 2.31425345e-01 -3.24150883e-02
2.62762457e-01 3.12407445e-02 -1.08365774e+00 5.56766570e-01
9.58954096e-01 -8.58958289e-02 -1.91976279e-02 7.23345160e-01
4.42167193e-01 -8.88445199e-01 -1.06185234e+00 3.97395730e-01
4.42073494e-01 -9.39137459e-01 1.03193116e+00 -2.22279206e-01
3.21229428e-01 -1.49614230e-01 3.38790834e-01 -1.43784475e+00
-8.80517215e-02 -7.54980743e-01 -2.19050512e-01 7.62853980e-01
4.11449581e-01 -1.37650084e+00 7.28364825e-01 8.39366376e-01
-4.49098438e-01 -9.93121266e-01 -8.53149652e-01 -1.02133071e+00
8.58339429e-01 -3.48966241e-01 3.61692548e-01 6.86797738e-01
-7.69226700e-02 -5.40266261e-02 2.89311916e-01 1.88447520e-01
7.24813581e-01 -4.73129563e-02 4.68760371e-01 -8.83524954e-01
-5.46893239e-01 -7.60514677e-01 2.40455195e-03 -1.11957884e+00
3.52552861e-01 -7.68312275e-01 1.18418068e-01 -1.62655735e+00
-1.41480446e-01 -7.56882191e-01 1.59241389e-02 6.01113141e-01
7.38888755e-02 7.85191823e-03 -1.11775696e-01 1.28655076e-01
9.33906808e-03 8.99170816e-01 1.28080249e+00 2.19062805e-01
-1.63688675e-01 7.96012878e-02 -3.47973675e-01 8.59400809e-01
1.48664832e+00 -6.01740301e-01 -2.83116847e-01 -2.05899000e-01
-4.13892902e-02 1.74651161e-01 8.14332128e-01 -1.26048887e+00
2.00887471e-01 -4.85157520e-01 5.35258889e-01 -6.77516997e-01
8.30365181e-01 6.44409470e-03 2.61454552e-01 6.05966151e-01
-1.31858081e-01 2.72070110e-01 6.65315986e-01 -2.53606856e-01
2.71185666e-01 -7.09765792e-01 1.20524204e+00 -4.46368963e-01
-2.34679654e-01 3.92986685e-01 -4.47002769e-01 9.83395204e-02
6.68377757e-01 2.62975276e-01 -3.94640803e-01 -4.68540698e-01
-7.28867829e-01 1.79850712e-01 6.54769421e-01 -2.85456628e-01
2.70961642e-01 -1.02323544e+00 -4.22284514e-01 -1.65300295e-01
-7.67459512e-01 1.56561732e-01 3.47875208e-01 8.99443030e-01
-1.09928203e+00 3.91836703e-01 -2.02151030e-01 -3.72975439e-01
-6.78122699e-01 1.52387932e-01 1.11447871e+00 6.56513274e-02
-7.25669384e-01 1.03046322e+00 2.24558204e-01 -5.80057383e-01
1.59844071e-01 -4.52732801e-01 3.05047512e-01 -6.95610464e-01
3.21870834e-01 5.28739452e-01 -3.93842496e-02 -1.35944620e-01
-4.76515025e-01 6.14102423e-01 2.68477350e-01 -2.55470037e-01
1.55094111e+00 2.03707665e-01 -8.59917402e-02 4.47416961e-01
1.17448211e+00 -2.04062387e-01 -1.66067517e+00 1.24681383e-01
-4.27777350e-01 4.68576178e-02 2.39530534e-01 -8.47172916e-01
-1.44709051e+00 9.83367562e-01 3.98696899e-01 1.60825718e-03
7.05176353e-01 -3.33928138e-01 8.22082281e-01 6.76490724e-01
1.40792936e-01 -8.85772288e-01 -5.20335793e-01 9.10986304e-01
9.70479786e-01 -9.47908342e-01 5.01815736e-01 -3.59070063e-01
-3.24136913e-01 1.28311515e+00 7.74186134e-01 -5.59452116e-01
1.32622278e+00 6.99279249e-01 1.77102480e-02 -2.03225851e-01
-7.01774776e-01 1.00971080e-01 1.97773844e-01 2.72437215e-01
7.43504524e-01 -1.68323547e-01 -4.15784538e-01 3.88200521e-01
-6.63472831e-01 -7.49853924e-02 5.35114467e-01 9.17078495e-01
-3.03608865e-01 -1.12593794e+00 -1.41985893e-01 3.76435667e-01
-3.08348596e-01 -3.23864996e-01 2.15642639e-02 7.96289563e-01
-2.16139019e-01 5.91840386e-01 -8.79886895e-02 5.20386845e-02
6.17249049e-02 1.09011635e-01 6.01178586e-01 -4.67989713e-01
-6.21469676e-01 -2.34698430e-01 2.30055768e-02 -5.44968963e-01
7.07815820e-03 -6.04394794e-01 -2.00453210e+00 -8.89582217e-01
-8.21030291e-04 5.92417955e-01 1.06442046e+00 8.24496984e-01
4.29730415e-01 6.10075951e-01 2.72739753e-02 -1.55560088e+00
-5.28574467e-01 -8.44460607e-01 -5.51417410e-01 -2.41650388e-01
2.14688003e-01 -1.03264248e+00 -5.95098019e-01 -2.16702685e-01] | [6.4570746421813965, 3.4948158264160156] |
f2aeb634-6a8d-46c7-9d3a-fd39b83ef022 | star-net-action-recognition-using-spatio | 1902.10024 | null | http://arxiv.org/abs/1902.10024v1 | http://arxiv.org/pdf/1902.10024v1.pdf | STAR-Net: Action Recognition using Spatio-Temporal Activation Reprojection | While depth cameras and inertial sensors have been frequently leveraged for
human action recognition, these sensing modalities are impractical in many
scenarios where cost or environmental constraints prohibit their use. As such,
there has been recent interest on human action recognition using low-cost,
readily-available RGB cameras via deep convolutional neural networks. However,
many of the deep convolutional neural networks proposed for action recognition
thus far have relied heavily on learning global appearance cues directly from
imaging data, resulting in highly complex network architectures that are
computationally expensive and difficult to train. Motivated to reduce network
complexity and achieve higher performance, we introduce the concept of
spatio-temporal activation reprojection (STAR). More specifically, we reproject
the spatio-temporal activations generated by human pose estimation layers in
space and time using a stack of 3D convolutions. Experimental results on
UTD-MHAD and J-HMDB demonstrate that an end-to-end architecture based on the
proposed STAR framework (which we nickname STAR-Net) is proficient in
single-environment and small-scale applications. On UTD-MHAD, STAR-Net
outperforms several methods using richer data modalities such as depth and
inertial sensors. | ['William McNally', 'Alexander Wong', 'John McPhee'] | 2019-02-26 | null | null | null | null | ['multimodal-activity-recognition'] | ['computer-vision'] | [ 2.49124765e-01 -2.81754136e-01 -3.10647078e-02 -4.33503360e-01
-3.50133657e-01 -8.43228102e-02 3.87900651e-01 -4.73287642e-01
-6.87721431e-01 4.86745954e-01 4.87962067e-01 -7.01892599e-02
1.43874630e-01 -5.89296520e-01 -7.48011827e-01 -4.80745703e-01
6.32891506e-02 1.01898305e-01 -2.97933985e-02 -4.86154482e-03
-1.05095133e-01 5.90151966e-01 -1.55654883e+00 1.04689680e-01
4.11122561e-01 1.28849387e+00 -7.25226998e-02 7.28025377e-01
3.93498719e-01 1.07298267e+00 -2.72161841e-01 -3.47745985e-01
6.20020747e-01 -3.30175549e-01 -7.58366883e-01 4.23076749e-01
7.76076734e-01 -9.68517721e-01 -9.89191949e-01 7.44083345e-01
8.20079625e-01 4.12637025e-01 3.59005248e-03 -1.15909100e+00
-6.40954435e-01 -9.79513004e-02 -4.72667783e-01 3.44722301e-01
6.70501471e-01 5.43757141e-01 5.98504126e-01 -8.19541037e-01
3.09153914e-01 1.26035190e+00 7.87234008e-01 5.66519499e-01
-8.67785931e-01 -3.34300458e-01 2.08728686e-01 2.50894397e-01
-1.12182415e+00 -5.46970308e-01 8.21522474e-01 -2.70633459e-01
1.51025259e+00 -4.56614979e-02 8.49799156e-01 1.62312794e+00
1.01235181e-01 1.25195861e+00 8.67976069e-01 -1.15774155e-01
8.87157023e-02 -4.78864372e-01 -3.99188250e-01 7.45014787e-01
1.85084548e-02 2.09279522e-01 -8.18323970e-01 2.69921094e-01
1.39330900e+00 6.10154212e-01 -2.67586023e-01 -7.17187524e-01
-1.53274846e+00 4.77174014e-01 8.72606814e-01 1.96573474e-02
-7.28034198e-01 5.37382185e-01 5.34628272e-01 5.77307977e-02
4.85875070e-01 2.10514024e-01 -4.27330613e-01 -6.66692436e-01
-4.70106065e-01 3.42696272e-02 2.06347510e-01 8.02114069e-01
3.19479197e-01 1.56888202e-01 8.92580971e-02 6.59548879e-01
2.00881660e-01 5.34292579e-01 6.15793407e-01 -9.19913054e-01
6.91752136e-01 7.03539550e-01 3.16799313e-01 -1.15324223e+00
-6.71443641e-01 -3.17697495e-01 -1.04939520e+00 2.97220588e-01
4.24452662e-01 -1.51439562e-01 -1.02192247e+00 1.62013245e+00
4.50093597e-01 4.93562281e-01 3.79598960e-02 1.54324818e+00
5.50305426e-01 1.33640736e-01 1.50278062e-01 3.15501183e-01
1.15826499e+00 -1.18672347e+00 -5.63137770e-01 -5.66385686e-01
7.17848539e-01 -2.79009372e-01 1.13783658e+00 3.43793064e-01
-9.27973449e-01 -8.94075572e-01 -1.09002888e+00 -3.80368471e-01
-3.31180334e-01 3.38217765e-01 1.10517728e+00 5.22228777e-01
-8.62364769e-01 5.50863504e-01 -1.51621509e+00 -6.72604978e-01
4.71748143e-01 3.39864522e-01 -6.50674343e-01 -9.35195312e-02
-9.79517162e-01 8.28922570e-01 2.15292916e-01 6.23393953e-01
-8.62466455e-01 -1.73725426e-01 -1.09187770e+00 -2.63848752e-01
3.84121925e-01 -9.67230916e-01 1.32308328e+00 -9.23717439e-01
-1.82960629e+00 7.63579428e-01 2.31756404e-01 -6.76577985e-01
6.94227457e-01 -8.99153113e-01 -4.69376057e-01 3.45077246e-01
-1.73718035e-01 7.12565422e-01 7.29888499e-01 -5.02809227e-01
-5.27337313e-01 -7.48575449e-01 3.88639539e-01 4.99614477e-01
-4.91903156e-01 -6.49019480e-02 -4.80671376e-01 -6.33693099e-01
3.05544436e-01 -1.15849674e+00 -2.61607677e-01 3.39236587e-01
-2.62676001e-01 7.42884576e-02 9.51880097e-01 -6.99998617e-01
7.91479170e-01 -2.13669944e+00 2.82994211e-01 -2.26464614e-01
1.92472145e-01 4.13638979e-01 8.77697691e-02 5.74869029e-02
-5.09484112e-02 -5.03472567e-01 -3.36808525e-02 -5.58981657e-01
7.11233765e-02 3.88907731e-01 -6.21032380e-02 6.85632885e-01
5.91810606e-02 1.02900064e+00 -9.26972091e-01 -1.02155909e-01
8.81979823e-01 8.53287518e-01 -5.34528077e-01 4.31730747e-01
6.06064275e-02 7.48345017e-01 -4.32262659e-01 9.08399940e-01
2.68416375e-01 -4.12674546e-01 -1.87692776e-01 -3.44496965e-01
1.78199917e-01 3.26331973e-01 -1.11669350e+00 2.35219502e+00
-5.04323423e-01 5.36963642e-01 -8.21427032e-02 -8.90406609e-01
5.95815122e-01 3.23519826e-01 7.37796903e-01 -8.76426160e-01
2.63353735e-01 -4.84876707e-02 -4.08737659e-01 -6.42185211e-01
5.17730176e-01 1.08266041e-01 -1.49646819e-01 1.15551300e-01
6.30801767e-02 4.08138245e-01 -2.53288925e-01 -7.38627836e-02
1.32899225e+00 6.60345495e-01 1.88356757e-01 5.86698890e-01
2.45019495e-01 -1.67899281e-01 5.62812626e-01 4.67773408e-01
-5.74139595e-01 8.25948954e-01 -5.58210388e-02 -8.29926431e-01
-8.97914708e-01 -1.06760454e+00 2.31423497e-01 9.88518000e-01
1.53809786e-01 -1.67919040e-01 -5.59740663e-01 -6.39156222e-01
-5.12742847e-02 1.24931805e-01 -7.10736871e-01 -2.11880937e-01
-5.72167337e-01 -4.04436082e-01 6.64523840e-01 1.18140697e+00
1.21526897e+00 -1.00796402e+00 -1.22270441e+00 3.66879642e-01
-2.19704837e-01 -1.49308431e+00 -4.16640729e-01 5.53256879e-03
-1.00595105e+00 -1.03333139e+00 -1.00930667e+00 -2.14182422e-01
4.53761160e-01 4.83485579e-01 8.18305075e-01 -4.67402279e-01
-3.26942712e-01 7.51675069e-01 -4.85438198e-01 -4.16035987e-02
3.99909526e-01 -7.53826499e-02 4.54369873e-01 3.42859507e-01
5.57619512e-01 -7.00980246e-01 -9.30148900e-01 3.35828424e-01
-7.39193141e-01 1.45862386e-01 8.29527020e-01 5.95657647e-01
4.74292636e-01 -3.16532552e-01 8.86215642e-03 -2.90459067e-01
1.91819832e-01 -1.40769795e-01 -4.20030713e-01 5.65965520e-03
-2.24195510e-01 -2.10092247e-01 2.88247585e-01 -4.86579627e-01
-9.84939039e-01 4.96299624e-01 -2.51955867e-01 -9.69275117e-01
-5.10969222e-01 4.39944923e-01 -2.31590495e-01 -8.79958197e-02
6.44789577e-01 1.20888337e-01 -2.58262809e-02 -5.93874633e-01
2.83042967e-01 5.85736990e-01 9.61880386e-01 -1.96485206e-01
4.44568247e-01 7.09591687e-01 -9.14256796e-02 -7.02590525e-01
-8.93375576e-01 -5.13524890e-01 -9.44329798e-01 -3.89350504e-01
1.30829883e+00 -1.34792781e+00 -7.93664992e-01 9.76296127e-01
-9.82586503e-01 -3.45231920e-01 -1.70764580e-01 8.99834454e-01
-7.03908622e-01 3.20940346e-01 -5.50229192e-01 -7.33464241e-01
-9.23388451e-02 -1.01688480e+00 1.35864043e+00 2.83042192e-01
-1.51470631e-01 -8.05803835e-01 -1.19190775e-01 6.59760833e-01
3.26471180e-01 6.37320936e-01 8.49275738e-02 -1.28442749e-01
-6.14345610e-01 -4.93119836e-01 -1.87873095e-01 3.72506082e-01
1.83694690e-01 -6.00179970e-01 -9.94366407e-01 -2.99872190e-01
-1.49741188e-01 -7.12194562e-01 6.31780684e-01 4.38829213e-01
1.16468215e+00 -8.86202529e-02 -1.19247727e-01 8.52145672e-01
1.06490993e+00 -6.39752299e-02 8.10567617e-01 4.38876897e-01
1.18620253e+00 9.00286660e-02 4.79477465e-01 6.70484662e-01
4.98738766e-01 8.89895916e-01 5.55467069e-01 -3.60633790e-01
2.17136517e-02 -4.40051705e-01 5.32201946e-01 4.17693555e-01
-4.88994598e-01 3.17435414e-02 -8.69602323e-01 4.74477470e-01
-2.17871571e+00 -8.94404948e-01 1.85186073e-01 2.10363746e+00
4.43911791e-01 1.55816287e-01 3.37523133e-01 8.68951231e-02
2.90500551e-01 3.07756513e-01 -8.24616194e-01 5.34805842e-02
-1.37634069e-01 7.37192929e-02 4.80690002e-01 3.55793536e-02
-1.49753606e+00 8.10665190e-01 5.69307184e+00 1.05334811e-01
-1.25363433e+00 3.09769157e-02 3.85638714e-01 -3.45110744e-01
6.03407264e-01 -4.40584332e-01 -4.91938442e-01 1.93503156e-01
9.08096075e-01 4.89184022e-01 2.61471063e-01 1.15241563e+00
3.57733846e-01 -2.27002949e-01 -1.32273424e+00 1.47661185e+00
4.39727455e-02 -9.69173372e-01 -3.45915735e-01 1.36433095e-01
5.17046630e-01 3.50923777e-01 1.34992644e-01 3.49002391e-01
2.25861505e-01 -9.84663665e-01 6.20964587e-01 4.68894243e-01
6.82273626e-01 -5.37186205e-01 7.46796310e-01 2.09304318e-01
-1.20208859e+00 -1.83295295e-01 -3.03747147e-01 -5.13070583e-01
1.69059947e-01 3.02319378e-01 -5.34627199e-01 2.86731213e-01
9.32539403e-01 1.12173128e+00 -5.62060058e-01 9.32980120e-01
-3.18469673e-01 2.20272690e-01 -3.72182876e-01 2.32256427e-01
4.05554622e-01 1.07594617e-01 1.59950197e-01 8.71187866e-01
3.68305087e-01 3.39117795e-01 3.53525311e-01 4.71753657e-01
-1.67910438e-02 -4.70973641e-01 -7.35844254e-01 -1.67022884e-01
-9.80785340e-02 1.00914276e+00 -4.62932885e-01 -1.59147397e-01
-7.00322390e-01 1.65284669e+00 2.72793204e-01 3.66114020e-01
-9.29727972e-01 -7.26562440e-02 1.10229087e+00 -4.25329134e-02
3.01164180e-01 -7.17785895e-01 -1.78782403e-01 -1.38083863e+00
2.25590199e-01 -8.81201506e-01 4.02354300e-01 -1.01516783e+00
-1.09082139e+00 4.17761266e-01 -2.37509757e-01 -1.59703684e+00
-3.66793990e-01 -9.01887178e-01 -2.85285991e-02 7.04770088e-01
-1.20257926e+00 -1.28326523e+00 -6.51077390e-01 1.06637013e+00
5.14842212e-01 -4.01099883e-02 9.03988838e-01 4.29189712e-01
-7.00742424e-01 5.50662816e-01 -2.06921369e-01 5.90231657e-01
5.56901693e-01 -1.03797615e+00 5.62576234e-01 9.35933828e-01
2.71707326e-01 5.07524967e-01 3.23274881e-01 -3.23465765e-01
-1.68470490e+00 -1.09591234e+00 5.96229255e-01 -6.68246567e-01
3.44256848e-01 -3.57740164e-01 -5.11944711e-01 8.09869111e-01
-1.01725705e-01 5.58405519e-01 4.92926121e-01 6.04679063e-02
-3.35725516e-01 -1.42845348e-01 -9.64459956e-01 4.80638891e-01
1.43512297e+00 -8.64070237e-01 -4.95431036e-01 4.05005038e-01
4.24625278e-01 -7.52481818e-01 -8.30430448e-01 3.69337946e-01
6.48473322e-01 -1.09107971e+00 1.14788949e+00 -8.22713733e-01
4.77728158e-01 -4.63141024e-01 -3.51065934e-01 -1.07535875e+00
-3.38770926e-01 -3.97525012e-01 -4.11136210e-01 5.17803609e-01
-5.17428406e-02 -4.13100094e-01 1.13019907e+00 8.44533384e-01
-2.02778816e-01 -6.33147120e-01 -1.08092999e+00 -7.96507597e-01
-5.37323058e-01 -5.85379362e-01 3.75798821e-01 8.73086810e-01
-1.00493059e-01 2.51085192e-01 -7.95761526e-01 3.07866603e-01
4.75068122e-01 -9.52512696e-02 1.04067194e+00 -8.93679023e-01
-4.67822701e-01 -8.45526606e-02 -9.14349556e-01 -1.61402035e+00
-1.60489455e-01 -2.40171850e-01 4.10244465e-02 -1.45688224e+00
-1.69679567e-01 -4.57390165e-03 -4.64888453e-01 6.86021090e-01
-5.74141406e-02 5.79956174e-01 3.07414099e-03 2.22662777e-01
-7.36856639e-01 8.47915530e-01 1.18220186e+00 -5.80635220e-02
-5.43913990e-02 -1.42969355e-01 -3.02488446e-01 8.41486514e-01
5.32510459e-01 8.55924934e-02 -4.97753799e-01 -9.63447332e-01
9.33571458e-02 1.62359998e-01 8.89420748e-01 -1.53993487e+00
2.28482038e-01 -1.30347967e-01 8.12203467e-01 -4.58728373e-01
7.78712988e-01 -1.08588064e+00 9.28868428e-02 3.60339612e-01
-1.67364612e-01 7.92343840e-02 5.43630272e-02 6.90959752e-01
-3.31780583e-01 6.83618307e-01 5.42170525e-01 -2.85522908e-01
-1.15511978e+00 5.87180912e-01 -1.54759824e-01 -2.02775434e-01
1.03834760e+00 -6.53904915e-01 1.39646038e-01 -4.65638757e-01
-8.07951212e-01 -9.71038491e-02 3.63816023e-01 7.00827122e-01
8.26449752e-01 -1.68112314e+00 -2.77543873e-01 3.21963161e-01
2.80203402e-01 3.17513868e-02 3.56279075e-01 1.10626435e+00
-5.07444143e-01 6.31682038e-01 -5.07114708e-01 -7.93082952e-01
-1.06603789e+00 3.69162679e-01 5.26799738e-01 -9.80189368e-02
-9.55538869e-01 1.07821667e+00 2.29944009e-02 -4.70776796e-01
5.74651062e-01 -4.78061765e-01 2.07618356e-01 -3.18073183e-01
5.36344588e-01 2.09155872e-01 7.54062980e-02 -7.49965250e-01
-4.84935880e-01 3.63393754e-01 1.97203994e-01 -6.60767630e-02
1.43763900e+00 -2.61405051e-01 4.64385390e-01 2.76294440e-01
1.18054390e+00 -7.73699820e-01 -1.87061775e+00 -4.33614314e-01
-2.68410921e-01 -8.91362846e-01 3.04591537e-01 -6.31002247e-01
-1.23100102e+00 9.98557866e-01 9.26425993e-01 -2.84557402e-01
1.29239511e+00 -2.83077955e-01 1.12863553e+00 7.59679258e-01
6.54824853e-01 -1.31026411e+00 4.31062639e-01 5.55034578e-01
8.54815245e-01 -1.39984751e+00 -4.40781787e-02 -6.31329119e-02
-5.93632996e-01 8.81618679e-01 8.89539897e-01 -2.86431219e-02
3.83939624e-01 -3.07717174e-02 1.86599597e-01 -1.59070641e-01
-2.69701749e-01 -2.94004053e-01 1.26576334e-01 6.20289266e-01
3.51588607e-01 -6.71899840e-02 3.02354872e-01 3.11623722e-01
5.07369861e-02 3.41036856e-01 1.03260897e-01 1.34952581e+00
-4.93487641e-02 -6.63147688e-01 -3.38217646e-01 1.65248513e-01
-2.77229309e-01 2.27358684e-01 -4.38592672e-01 6.39072478e-01
1.63188517e-01 7.37239897e-01 6.30058944e-02 -7.64974475e-01
5.69797575e-01 9.23124552e-02 6.59166694e-01 -3.29371810e-01
-3.40558320e-01 -9.74883512e-02 1.89841330e-01 -1.15062714e+00
-8.47406030e-01 -7.24334717e-01 -1.01144528e+00 -3.88857365e-01
5.63295633e-02 -6.20939434e-01 5.39134026e-01 1.11269999e+00
5.76220036e-01 5.70641935e-01 2.98887730e-01 -1.23412979e+00
-4.75532800e-01 -1.06735539e+00 -5.51140606e-01 5.16984224e-01
3.40728849e-01 -9.42578852e-01 9.72516835e-02 1.78009257e-01] | [7.822884559631348, 0.400326669216156] |
1d92914d-fee5-4daa-9bab-23d84c0d879d | esg-valued-portfolio-optimization-and-dynamic | 2206.02854 | null | https://arxiv.org/abs/2206.02854v1 | https://arxiv.org/pdf/2206.02854v1.pdf | ESG-Valued Portfolio Optimization and Dynamic Asset Pricing | ESG ratings provide a quantitative measure for socially responsible investment. We present a unified framework for incorporating numeric ESG ratings into dynamic pricing theory. Specifically, we introduce an ESG-valued return that is a linearly constrained transformation of financial return and ESG score. This leads to a more complex portfolio optimization problem in a space governed by reward, risk and ESG score. The framework preserves the traditional risk aversion parameter and introduces an ESG affinity parameter. We apply this framework to develop ESG-valued: portfolio optimization; capital market line; risk measures; option pricing; and the computation of shadow riskless rates. | ['Svetlozar T. Rachev', 'Stefan Mittnik', 'W. Brent Lindquist', 'Davide Lauria'] | 2022-06-06 | null | null | null | null | ['portfolio-optimization'] | ['time-series'] | [-5.47528207e-01 4.56401259e-01 -3.52549374e-01 -4.46715891e-01
-3.79950732e-01 -8.20905507e-01 3.51855576e-01 -3.79137963e-01
-4.20156568e-01 7.46062338e-01 2.75721192e-01 -6.08127773e-01
-7.13898599e-01 -1.08742416e+00 3.74941863e-02 -3.18980336e-01
-3.25686902e-01 2.30318964e-01 -4.15891409e-02 -2.77433097e-01
6.05450332e-01 6.05949223e-01 -9.72811460e-01 -2.16035426e-01
7.03707993e-01 1.46677804e+00 -4.53484923e-01 5.11942029e-01
1.29796863e-01 8.30331087e-01 -4.55501527e-01 -1.21053672e+00
8.10316503e-01 -2.77526468e-01 -3.30931813e-01 -6.90896094e-01
-5.45020759e-01 -5.62881947e-01 7.25247115e-02 1.14248610e+00
3.59962523e-01 3.63182247e-01 1.08821535e+00 -1.37603128e+00
-8.07397604e-01 8.62802327e-01 -4.04236317e-01 2.38888189e-01
1.15184873e-01 -3.58058095e-01 1.73906100e+00 -8.49516749e-01
1.21003978e-01 1.23698747e+00 5.12843192e-01 7.65257716e-01
-1.03935647e+00 -2.30824798e-01 -5.72469607e-02 -4.21139866e-01
-5.78289092e-01 2.14964584e-01 4.10963982e-01 -6.06941402e-01
8.26198399e-01 6.22588217e-01 8.22166741e-01 5.01505077e-01
5.49927354e-01 4.56272870e-01 9.86308575e-01 -2.72414237e-01
5.85442424e-01 1.46417364e-01 -1.00646324e-01 -9.50855855e-03
6.69092238e-01 8.03826809e-01 -1.85942292e-01 -5.16727805e-01
1.09374833e+00 6.21230826e-02 1.24550387e-01 -4.96950924e-01
-5.70063889e-01 1.26045775e+00 1.52881801e-01 -2.45161712e-01
-5.19300342e-01 4.15984184e-01 -1.26799680e-02 6.85226262e-01
4.83131349e-01 7.00878084e-01 -1.43322930e-01 -1.47069231e-01
-5.69987297e-01 4.69557375e-01 1.10946190e+00 4.98351812e-01
2.24891324e-02 1.94983169e-01 -5.95774055e-01 5.54710448e-01
7.40259886e-01 6.20184302e-01 1.19447909e-01 -1.87533712e+00
6.22019291e-01 1.37840748e-01 8.33709240e-01 -9.18517411e-01
-2.51482338e-01 -5.93985140e-01 -2.19945431e-01 1.10729957e+00
3.92591298e-01 -4.03401315e-01 2.76806444e-01 1.83358836e+00
-7.07692280e-02 -3.02073658e-01 2.09386230e-01 5.02066970e-01
-3.26624990e-01 4.51364189e-01 -2.28505477e-01 -5.31560779e-01
6.75036132e-01 -7.92248189e-01 -7.16300726e-01 3.07090461e-01
3.03816468e-01 -4.11064744e-01 9.13918853e-01 5.19546986e-01
-1.52762079e+00 3.42947602e-01 -1.00259662e+00 6.27684176e-01
-4.68561724e-02 -4.87266660e-01 6.95925891e-01 1.38602197e+00
-1.46216381e+00 1.33308494e+00 -1.55667961e-01 2.99933314e-01
4.70442176e-01 1.66768253e-01 6.66665971e-01 7.89829850e-01
-1.10815167e+00 1.14203489e+00 -1.08526284e-02 -1.98020115e-01
-5.24968565e-01 -6.78176224e-01 -3.25562984e-01 3.25777888e-01
4.54021722e-01 -4.63306814e-01 1.40457153e+00 -6.34834886e-01
-2.00200081e+00 6.85267687e-01 1.09138870e+00 -7.95994759e-01
1.15471256e+00 -4.66645151e-01 -3.81599724e-01 6.29318804e-02
-9.68279243e-02 -9.85711664e-02 4.93405461e-01 -6.45340204e-01
-2.29272589e-01 -5.12403101e-02 3.44338506e-01 4.34428483e-01
-3.24542552e-01 6.43649042e-01 7.95157731e-01 -1.37737870e+00
-3.92099351e-01 -6.66615427e-01 -4.10006285e-01 2.69563466e-01
-2.15965416e-02 7.93558359e-02 -3.24386656e-01 -5.49443126e-01
1.40299392e+00 -1.83039594e+00 -1.18855491e-01 5.60185432e-01
1.19498419e-02 -1.86483324e-01 -1.18430816e-01 3.75430375e-01
-2.40716800e-01 5.78761816e-01 -1.29975021e-01 1.39184564e-01
8.22297871e-01 -3.35183531e-01 -7.21349955e-01 1.64013699e-01
-4.66435701e-02 1.12907636e+00 -8.01264405e-01 5.92936799e-02
-2.23868921e-01 -2.20130488e-01 -8.13901782e-01 3.88304323e-01
2.50674561e-02 -4.40793574e-01 -5.73217928e-01 5.69514990e-01
5.60467780e-01 1.71325617e-02 -2.26647794e-01 5.81440508e-01
-3.52084339e-01 1.24545872e-01 -1.16085672e+00 6.35175169e-01
-2.35743016e-01 -5.24729490e-03 -2.32066318e-01 -5.36029935e-01
1.13788164e+00 1.70833573e-01 5.94202757e-01 -2.59541363e-01
1.78505912e-01 5.40078521e-01 -1.67890966e-01 1.76881433e-01
6.00186110e-01 -8.68581176e-01 -3.48834813e-01 1.39370155e+00
-3.54710281e-01 -2.46551886e-01 -4.23669577e-01 1.63611546e-01
8.89961243e-01 3.16696763e-02 4.73340362e-01 -7.19894052e-01
7.56670907e-02 -7.17037380e-01 5.24138808e-01 5.32093167e-01
-5.46260595e-01 2.91124254e-01 1.14423454e+00 -2.55974196e-02
-9.13103163e-01 -1.59035814e+00 -5.62766977e-02 9.91213381e-01
-1.97107852e-01 1.98751256e-01 -6.39653444e-01 -7.18893647e-01
8.01658690e-01 1.11333251e+00 -7.84207404e-01 -2.65015632e-01
-1.21282265e-01 -1.05652297e+00 3.52429122e-01 6.28721654e-01
1.70595765e-01 -1.03439307e+00 -7.73749292e-01 1.96704343e-01
5.14538229e-01 -1.02507118e-02 -7.36390531e-01 9.62842330e-02
-7.87924170e-01 -7.79564202e-01 -1.23261952e+00 3.19828361e-01
4.59891967e-02 -2.12111786e-01 1.25697100e+00 -3.29923391e-01
-5.67246079e-02 7.15121210e-01 -1.55677855e-01 -6.26145780e-01
-1.08779326e-01 -8.52988660e-01 2.09708929e-01 1.76624700e-01
-3.83948684e-02 -4.01118457e-01 -9.06890035e-01 5.76173306e-01
-4.07931089e-01 -7.98103631e-01 -8.04640353e-02 6.60210848e-01
4.49454904e-01 -3.06845695e-01 1.27790117e+00 -7.18100488e-01
1.11117649e+00 -5.97639084e-01 -1.04600561e+00 4.63490784e-01
-1.27336717e+00 -1.71981920e-02 1.19401351e-01 -2.21279129e-01
-1.39403009e+00 -9.63414550e-01 3.16764712e-01 -1.70761108e-01
1.21347821e+00 3.53433728e-01 -2.97470510e-01 -2.98659116e-01
3.76487583e-01 -5.51539540e-01 -1.78642124e-02 -4.95019168e-01
6.80620134e-01 3.65072012e-01 2.94483632e-01 -6.22264385e-01
7.39244282e-01 1.64343685e-01 1.14316933e-01 3.60830933e-01
-8.18726122e-01 2.99421936e-01 -2.01199591e-01 -3.55316222e-01
7.23186672e-01 -4.16827261e-01 -1.01155198e+00 3.81791264e-01
-7.01889813e-01 -3.63664508e-01 -1.21882451e+00 6.75497472e-01
-1.24095309e+00 6.11609258e-02 -6.68241799e-01 -1.51277244e+00
-2.36883894e-01 -3.95781249e-01 1.08122781e-01 1.75618351e-01
-4.86854240e-02 -1.26073170e+00 4.28776056e-01 1.83552191e-01
8.03083122e-01 6.31908536e-01 5.76382220e-01 -5.53886890e-01
-5.56467652e-01 -2.12612435e-01 -2.30569348e-01 7.77006269e-01
-1.73689529e-01 6.10348433e-02 -6.52233005e-01 -1.92609787e-01
7.58297801e-01 -2.05272526e-01 6.58005357e-01 5.82604349e-01
7.83411026e-01 -6.24541759e-01 4.24109548e-01 8.11690569e-01
1.56136048e+00 5.89913964e-01 5.39070547e-01 6.67536259e-01
-3.03685199e-03 1.05356276e+00 1.10786903e+00 9.90540624e-01
-8.42945799e-02 4.30549413e-01 3.79532456e-01 3.94376308e-01
7.05822229e-01 -1.90231562e-01 5.67527950e-01 5.39883196e-01
-4.92592365e-01 -9.88517702e-02 -6.51122332e-01 -9.35602039e-02
-1.83764493e+00 -1.31440198e+00 4.92044210e-01 2.47222161e+00
5.48003495e-01 1.91208094e-01 5.92046082e-01 -8.99936557e-02
7.73890316e-01 1.10968754e-01 -9.53425646e-01 -8.81183326e-01
-2.59564281e-01 1.52696550e-01 9.36325490e-01 5.67964137e-01
-7.10689068e-01 3.49783391e-01 8.31495094e+00 7.19565928e-01
-4.70229983e-01 6.34675473e-02 1.09624028e+00 -4.98188227e-01
-1.26873100e+00 -4.22377512e-02 -5.07619739e-01 5.68984628e-01
7.93537199e-01 -1.41381598e+00 4.92855787e-01 1.02117229e+00
1.81359038e-01 3.52275401e-01 -9.19868171e-01 6.44463360e-01
-4.73339260e-01 -1.12178063e+00 -2.30089605e-01 5.61513543e-01
6.86924934e-01 -4.06042337e-01 5.89142323e-01 6.04273379e-02
9.06522930e-01 -1.19777501e+00 1.13052237e+00 1.26192129e+00
7.29337394e-01 -1.09589839e+00 5.91044188e-01 -1.89058483e-01
-7.49445140e-01 -4.49804783e-01 -5.65414310e-01 -1.40405089e-01
4.55353379e-01 5.77353537e-01 -6.98798075e-02 3.25476259e-01
2.79017955e-01 5.43359995e-01 -1.35820553e-01 1.09816599e+00
-1.98798314e-01 1.18359484e-01 -8.25880095e-02 -3.13048273e-01
1.08743973e-01 -8.96959245e-01 5.20033062e-01 6.61714971e-01
8.32060277e-01 3.44138980e-01 -6.16320074e-01 1.38299203e+00
5.36058955e-02 1.78444728e-01 -5.39239287e-01 -1.37719542e-01
3.85250658e-01 9.16887462e-01 -5.88778853e-01 1.15629815e-01
-2.87058234e-01 6.88600123e-01 -1.10816866e-01 3.05726737e-01
-8.47042859e-01 -5.10204911e-01 9.52404678e-01 4.77758944e-02
6.40258864e-02 1.50553599e-01 -6.37134194e-01 -1.27318883e+00
6.44787550e-02 -3.23430717e-01 7.60165393e-01 -5.18805742e-01
-1.59664798e+00 2.32631132e-01 5.92029504e-02 -1.26722431e+00
-4.67667371e-01 -8.37598145e-01 -1.06680775e+00 1.17878985e+00
-1.41059303e+00 -1.44062132e-01 5.93695700e-01 2.36617669e-01
-3.24716091e-01 -6.82525396e-01 5.15614390e-01 -6.78691491e-02
-4.62781161e-01 7.54125059e-01 6.24859810e-01 -4.35439140e-01
1.90546811e-01 -1.78497601e+00 7.92884767e-01 6.44042611e-01
-5.32141268e-01 5.37633419e-01 3.56686950e-01 -7.17711926e-01
-7.18422532e-01 -7.10454166e-01 6.34818137e-01 -7.51085579e-01
1.29121137e+00 1.93900153e-01 -4.02644455e-01 4.22673821e-01
-1.50459632e-01 -3.20592970e-01 9.80974495e-01 -2.55548418e-01
-2.44868457e-01 -2.58740276e-01 -1.62291932e+00 5.94202638e-01
1.13732624e+00 -4.65857089e-01 -6.34767652e-01 -4.64719087e-02
1.01546216e+00 3.83480936e-01 -1.24256301e+00 1.02546416e-01
7.00898588e-01 -1.07931912e+00 1.30711401e+00 -6.07925653e-01
1.51099131e-01 3.20983022e-01 -3.36393952e-01 -1.02292037e+00
-5.82622170e-01 -1.52570140e+00 -8.08826648e-03 1.14534819e+00
4.56455529e-01 -1.35666752e+00 6.58224702e-01 1.30706358e+00
1.42640710e-01 -9.46924627e-01 -1.26118398e+00 -1.28160465e+00
6.23779416e-01 -1.84110180e-01 9.01256859e-01 6.86667860e-01
4.07310188e-01 -5.64592719e-01 -5.61607540e-01 -7.33120143e-01
1.36459970e+00 5.69383949e-02 -2.29747251e-01 -1.19030297e+00
-6.19045734e-01 -1.09939337e+00 4.67786677e-02 -3.17659050e-01
-2.13466406e-01 -8.86295736e-01 -5.55798769e-01 -9.64908421e-01
7.34535977e-02 -4.43653375e-01 -9.93014574e-01 8.04170147e-02
1.69642270e-01 -1.66618273e-01 6.45972908e-01 1.49645850e-01
-2.15249568e-01 7.67051518e-01 1.09436619e+00 4.17461008e-01
1.94177032e-02 6.86146677e-01 -1.18536186e+00 8.93994570e-01
9.45748210e-01 -4.64659244e-01 -6.11095369e-01 3.10629815e-01
9.75705326e-01 4.96445656e-01 2.15384126e-01 -2.12460026e-01
-2.69629210e-01 -1.01859641e+00 -6.28719330e-02 -4.45259780e-01
1.44238621e-02 -3.25284034e-01 3.75690728e-01 6.71347976e-01
-7.17566311e-01 1.88808963e-01 -5.65202296e-01 5.77263832e-01
1.39749661e-01 -8.45450044e-01 8.77609253e-01 -1.80965871e-01
1.29226476e-01 2.67406791e-01 -1.60000280e-01 3.78305048e-01
1.19367397e+00 -2.64838099e-01 -4.93851870e-01 -5.22505343e-01
-6.39857173e-01 1.79666340e-01 6.94111586e-01 7.99552798e-02
7.23070502e-01 -1.76225364e+00 -5.56229115e-01 -4.14715081e-01
-3.98718387e-01 -8.90770853e-01 -2.32693523e-01 2.18487769e-01
-6.78801894e-01 1.51617855e-01 -5.92569947e-01 6.76427007e-01
-5.14113486e-01 3.91388476e-01 8.25837255e-01 -4.15351868e-01
-3.04696232e-01 1.00807154e+00 4.76254284e-01 -1.33775353e-01
1.22183315e-01 -2.68657625e-01 -5.08935809e-01 4.90401238e-01
7.73336112e-01 1.12287915e+00 -6.28600597e-01 -8.78578424e-02
-1.24909326e-01 5.01428366e-01 4.61356074e-01 -8.85217369e-01
1.55293751e+00 -2.68203974e-01 -6.76024556e-02 4.67136174e-01
7.55382657e-01 2.88116466e-03 -1.47425246e+00 2.84029037e-01
6.16063774e-01 -9.92490947e-01 -2.85209894e-01 -9.43390846e-01
-9.65629637e-01 6.68866575e-01 3.59083533e-01 5.11487246e-01
8.18565071e-01 -3.76722276e-01 4.08352733e-01 2.32886419e-01
5.77920556e-01 -1.52517509e+00 2.92996585e-01 2.83929497e-01
1.39320266e+00 -5.42570472e-01 3.51968892e-02 -2.10075349e-01
-9.32264626e-01 9.64661956e-01 8.15618262e-02 -5.97706556e-01
1.36196399e+00 1.82827905e-01 1.15806172e-02 5.44719100e-02
-6.51873291e-01 3.39320987e-01 2.95222223e-01 5.19107282e-01
1.24861054e-01 7.09091365e-01 -7.71820903e-01 1.40332711e+00
-5.19942164e-01 -1.83191463e-01 9.48738039e-01 7.15153575e-01
-5.66678882e-01 -1.10087943e+00 -8.71573612e-02 5.81593215e-01
-8.33735228e-01 -8.03379789e-02 -4.73185003e-01 2.31997192e-01
-6.77230179e-01 4.89463121e-01 3.44602726e-02 -1.12965144e-01
4.90821004e-01 -2.48823360e-01 2.09718958e-01 -6.67542100e-01
-6.99801624e-01 4.39222865e-02 2.57936791e-02 -7.28279889e-01
-2.56740510e-01 -1.03473973e+00 -8.47540021e-01 -3.49814624e-01
-2.28181034e-01 3.16804856e-01 3.38171452e-01 2.30928317e-01
-7.49417245e-02 1.06256895e-01 1.38270926e+00 -4.46239799e-01
-1.78734565e+00 -2.92580962e-01 -1.59220946e+00 5.50636649e-02
-1.40637770e-01 -7.47552633e-01 -8.70475471e-01 -6.71073794e-01] | [4.9518866539001465, 3.9434919357299805] |
62c28e48-1010-4de3-adbc-791f75436983 | signed-directed-graph-contrastive-learning | 2301.05163 | null | https://arxiv.org/abs/2301.05163v1 | https://arxiv.org/pdf/2301.05163v1.pdf | Signed Directed Graph Contrastive Learning with Laplacian Augmentation | Graph contrastive learning has become a powerful technique for several graph mining tasks. It learns discriminative representation from different perspectives of augmented graphs. Ubiquitous in our daily life, singed-directed graphs are the most complex and tricky to analyze among various graph types. That is why singed-directed graph contrastive learning has not been studied much yet, while there are many contrastive studies for unsigned and undirected. Thus, this paper proposes a novel signed-directed graph contrastive learning, SDGCL. It makes two different structurally perturbed graph views and gets node representations via magnetic Laplacian perturbation. We use a node-level contrastive loss to maximize the mutual information between the two graph views. The model is jointly learned with contrastive and supervised objectives. The graph encoder of SDGCL does not depend on social theories or predefined assumptions. Therefore it does not require finding triads or selecting neighbors to aggregate. It leverages only the edge signs and directions via magnetic Laplacian. To the best of our knowledge, it is the first to introduce magnetic Laplacian perturbation and signed spectral graph contrastive learning. The superiority of the proposed model is demonstrated through exhaustive experiments on four real-world datasets. SDGCL shows better performance than other state-of-the-art on four evaluation metrics. | ['Chong-Kwon Kim', 'Yoonhyuk Choi', 'Taewook Ko'] | 2023-01-12 | null | null | null | null | ['graph-mining'] | ['graphs'] | [ 1.54832467e-01 3.21457475e-01 -3.22633743e-01 -1.01624422e-01
-3.21467042e-01 -6.17874503e-01 7.72717178e-01 2.34042794e-01
5.46149537e-02 5.29857576e-01 -3.48578952e-02 -2.33368307e-01
-5.22845149e-01 -8.39878142e-01 -6.83668911e-01 -8.82908046e-01
-5.64232171e-01 3.98965716e-01 1.38520852e-01 -4.36265051e-01
6.72356635e-02 4.49454933e-01 -9.81981397e-01 -1.06301472e-01
8.66647065e-01 4.66286182e-01 1.01856247e-03 4.52191919e-01
1.24279775e-01 7.87104428e-01 -1.70345187e-01 -4.00293112e-01
3.46505851e-01 -4.86076653e-01 -5.31847477e-01 1.07437946e-01
7.64337480e-01 3.45170319e-01 -6.35643303e-01 1.24876308e+00
6.85559034e-01 9.66616347e-02 8.67685378e-01 -1.48584688e+00
-9.37573910e-01 8.50605011e-01 -1.12069845e+00 4.23023671e-01
5.03056824e-01 2.50808848e-03 1.45473206e+00 -8.23745728e-01
7.10007548e-01 1.29218316e+00 6.49227500e-01 1.72085553e-01
-1.44600427e+00 -7.49536932e-01 5.78026235e-01 2.99303323e-01
-1.11599612e+00 5.55024296e-02 1.64323199e+00 -4.35202032e-01
5.77686310e-01 2.94077009e-01 6.42223656e-01 1.21378386e+00
1.20768242e-01 5.69472075e-01 1.23423934e+00 -4.14148897e-01
4.29970073e-03 -4.16554064e-02 1.39800370e-01 1.31067812e+00
6.53145552e-01 -9.59813073e-02 -5.62725186e-01 -1.23920046e-01
3.95081758e-01 7.36273080e-02 -4.84868079e-01 -9.61346984e-01
-1.05723822e+00 8.44918311e-01 9.07374501e-01 4.17865306e-01
-2.40606442e-01 1.54590636e-01 2.69204438e-01 5.22956491e-01
6.47467315e-01 4.47839469e-01 -9.19106230e-03 4.57931519e-01
-5.22733450e-01 -1.29460618e-01 7.61742294e-01 7.42124081e-01
8.82716417e-01 1.68116897e-01 1.54180571e-01 6.89602256e-01
3.39969546e-01 6.30995512e-01 1.21776555e-02 -2.96172410e-01
5.38726985e-01 9.71861124e-01 -7.60480940e-01 -1.82148874e+00
-6.71197057e-01 -7.41390467e-01 -1.36421835e+00 -2.15821192e-02
1.76846460e-01 -1.79664969e-01 -7.70218790e-01 1.72444654e+00
2.52803952e-01 5.38859844e-01 -4.74746257e-01 6.68945074e-01
1.08657622e+00 4.18703467e-01 -1.55882224e-01 -2.24027932e-01
8.36046636e-01 -7.25106895e-01 -5.75330734e-01 -3.15328658e-01
6.47015810e-01 -3.17355841e-01 1.09535980e+00 2.63159633e-01
-7.13597715e-01 -2.19356537e-01 -1.28756952e+00 1.90604627e-01
-5.03148913e-01 -1.16818197e-01 7.82086492e-01 6.45322323e-01
-1.06744838e+00 6.12393975e-01 -6.87993824e-01 -4.22428161e-01
3.78285706e-01 2.53567368e-01 -5.81182778e-01 -1.78733617e-01
-1.16303432e+00 6.00321949e-01 1.10129483e-01 -1.06502570e-01
-3.69101703e-01 -6.10103488e-01 -1.03181958e+00 -3.04954257e-02
6.98038399e-01 -5.87078154e-01 3.90851259e-01 -8.38139474e-01
-1.01337862e+00 9.59341466e-01 2.02469677e-01 -1.74061790e-01
3.82165909e-01 3.87722462e-01 -4.96381551e-01 1.54372215e-01
2.00662568e-01 1.45938739e-01 1.08849657e+00 -1.53667700e+00
8.23147893e-02 -5.99121094e-01 5.74223287e-02 3.51028472e-01
-4.50596631e-01 -7.23291874e-01 -3.86178613e-01 -9.39800143e-01
5.07134020e-01 -1.16399944e+00 -2.70184148e-02 -7.01437071e-02
-6.88322663e-01 -2.16080457e-01 1.08147013e+00 -4.86866325e-01
1.29481089e+00 -2.06882358e+00 5.07387578e-01 5.83551347e-01
8.87334824e-01 1.05583988e-01 -2.34068155e-01 6.95770979e-01
-3.83281767e-01 1.97089061e-01 -2.62515754e-01 -5.46273328e-02
-1.21660288e-02 -8.92385468e-02 6.54238909e-02 8.45003247e-01
-2.10309327e-02 1.07236648e+00 -1.37712586e+00 -4.24194068e-01
2.30260387e-01 6.13338947e-01 -5.00534415e-01 -9.17040631e-02
2.09872529e-01 3.25227708e-01 -2.98575759e-01 5.53607047e-01
7.64747739e-01 -5.87755203e-01 6.25685215e-01 -5.58827162e-01
3.89763594e-01 -2.14213338e-02 -1.10057473e+00 1.55249763e+00
-1.50593728e-01 5.38834572e-01 -4.01211120e-02 -1.54546940e+00
8.89505863e-01 -1.93980843e-01 5.64138591e-01 -6.80911958e-01
-7.84861520e-02 8.55467767e-02 1.33371323e-01 -3.87341529e-01
-5.36049791e-02 6.32716902e-03 -2.01687604e-01 3.94793719e-01
1.79321200e-01 -3.85634974e-02 3.20046574e-01 8.07490528e-01
1.41829658e+00 -1.27334138e-02 5.31217694e-01 -4.46254164e-01
5.43015957e-01 -5.99061131e-01 3.75128925e-01 5.16262174e-01
-1.00081101e-01 3.87348562e-01 8.99670362e-01 -1.83062702e-01
-3.86225790e-01 -1.31869674e+00 4.70515698e-01 1.03503406e+00
3.24831218e-01 -5.62607110e-01 -3.63124251e-01 -1.05592322e+00
2.33392790e-01 3.73927206e-01 -8.00324798e-01 -5.42385697e-01
-5.54168820e-01 -6.86552942e-01 2.02017814e-01 1.04368314e-01
4.15327191e-01 -7.52047420e-01 4.16772127e-01 -2.50776768e-01
3.57954502e-02 -9.57445204e-01 -7.78829038e-01 1.12079889e-01
-6.84548676e-01 -1.48217881e+00 -3.89518082e-01 -1.22824597e+00
9.18474078e-01 6.65237784e-01 1.19611669e+00 3.79534550e-02
-2.24945992e-01 6.87482953e-01 -3.92161220e-01 -9.85151455e-02
-1.72112957e-01 3.43454719e-01 2.16684528e-02 3.23013961e-01
1.14041381e-02 -1.12567413e+00 -4.45967317e-01 6.26022520e-04
-6.16226673e-01 -5.05468361e-02 7.76348472e-01 7.87743986e-01
5.33728421e-01 2.01745227e-01 4.97505248e-01 -1.35419154e+00
8.55203211e-01 -5.78173995e-01 -2.66629010e-01 2.69097030e-01
-6.99418545e-01 1.80731267e-01 8.99964333e-01 -3.06152076e-01
-5.30642152e-01 -2.61555284e-01 4.58100975e-01 -5.16624987e-01
3.16555172e-01 7.87901819e-01 -1.84269384e-01 -4.14211243e-01
7.05392063e-01 7.10211173e-02 -1.54834380e-03 -2.66856939e-01
4.92420018e-01 4.82202657e-02 2.79052496e-01 -2.01994553e-01
1.26518500e+00 4.71453011e-01 5.64583123e-01 -1.20096576e+00
-5.58443964e-01 -4.08176094e-01 -6.28956378e-01 -5.52351952e-01
4.62538481e-01 -5.63207328e-01 -7.76114881e-01 4.13292468e-01
-5.62268794e-01 -6.44597858e-02 -2.40801685e-02 2.03578711e-01
-1.56693429e-01 7.81451285e-01 -3.09003443e-01 -5.59483171e-01
-3.48005861e-01 -7.48198688e-01 9.12722111e-01 -1.87647849e-01
2.18529895e-01 -1.39673817e+00 1.35523140e-01 7.27208033e-02
2.34188601e-01 7.68022954e-01 9.40260291e-01 -5.00597656e-01
-4.64469612e-01 -2.95094639e-01 -1.98112965e-01 8.09155703e-02
5.94167113e-01 -2.23105505e-01 -6.32985771e-01 -7.75399148e-01
-4.44467366e-01 -2.66748339e-01 1.11336017e+00 2.04872370e-01
1.18022752e+00 -3.36267859e-01 -5.91144562e-01 6.30877614e-01
1.27053452e+00 -2.50401646e-01 1.83134705e-01 -1.24282464e-01
1.41514492e+00 5.24109185e-01 1.34315476e-01 1.10722467e-01
4.95483071e-01 4.95233029e-01 6.13362253e-01 -7.82173797e-02
-3.05965871e-01 -5.11909664e-01 3.58838201e-01 1.26175439e+00
-7.00351819e-02 -3.66961718e-01 -9.45348322e-01 2.63585836e-01
-1.74976945e+00 -1.13663626e+00 -2.18769521e-01 2.00899625e+00
3.03859383e-01 2.01018140e-01 2.48859487e-02 2.67858922e-01
9.10404384e-01 8.72877479e-01 -7.33974397e-01 8.29836577e-02
-2.35511735e-01 2.04946056e-01 3.64972323e-01 6.12045228e-01
-1.26025558e+00 6.47578537e-01 4.84753036e+00 6.64449036e-01
-1.19640279e+00 -8.13127235e-02 5.59069216e-01 4.06136625e-02
-6.19095445e-01 -1.78093538e-02 -7.35291466e-02 3.37369353e-01
2.69782364e-01 -2.43931875e-01 6.42073929e-01 5.28017342e-01
-3.07564400e-02 2.53842562e-01 -9.98845518e-01 1.24265039e+00
4.58261877e-01 -1.17163801e+00 1.31108329e-01 1.00264512e-01
7.27649033e-01 1.58758804e-01 2.15472683e-01 3.27837586e-01
2.22757295e-01 -8.84194791e-01 9.53616500e-02 4.91885573e-01
7.20440745e-01 -6.55404389e-01 3.25248778e-01 2.02117190e-01
-1.53910434e+00 9.87980664e-02 -1.24019040e-02 1.36529788e-01
-8.24334472e-02 8.48901510e-01 -7.30740726e-01 8.60756338e-01
5.24916530e-01 1.36576331e+00 -8.41856182e-01 6.73742652e-01
-2.00915605e-01 8.49001348e-01 -1.32924438e-01 -1.24061644e-01
1.81496203e-01 -6.26296163e-01 9.61229563e-01 9.28082407e-01
1.63879693e-01 -2.38474756e-01 4.78797942e-01 6.39001667e-01
-4.22469050e-01 3.97217035e-01 -1.20850015e+00 -2.28504241e-01
4.21847880e-01 1.45908213e+00 -7.91159570e-01 9.23179612e-02
-4.65813220e-01 9.20657218e-01 7.26750433e-01 4.74306256e-01
-6.81260169e-01 -3.47325295e-01 9.98276100e-02 4.40905422e-01
1.51574060e-01 -2.07382709e-01 1.11341298e-01 -1.27631402e+00
2.47945517e-01 -9.03278530e-01 6.36523724e-01 -3.58411640e-01
-1.81368518e+00 4.46805149e-01 1.64814115e-01 -1.14501548e+00
1.26473650e-01 -6.84824169e-01 -7.24899232e-01 4.85030919e-01
-1.41620064e+00 -1.28499377e+00 -5.46872139e-01 6.16654456e-01
-9.14997887e-03 -3.03590238e-01 4.95230675e-01 3.27235848e-01
-6.43429995e-01 6.57097340e-01 -3.24003883e-02 9.30094719e-02
6.81451619e-01 -1.60016942e+00 4.12725449e-01 7.59296834e-01
3.51527482e-01 4.54161078e-01 6.02137268e-01 -7.40862131e-01
-1.77394569e+00 -1.14040899e+00 4.44522589e-01 -4.75028120e-02
8.19758415e-01 -6.06875777e-01 -8.65917683e-01 6.08304977e-01
2.29577884e-01 5.00377059e-01 4.59010124e-01 1.09975182e-01
-6.23208463e-01 -4.25591260e-01 -9.59261060e-01 7.47136295e-01
1.56639552e+00 -6.63926065e-01 -3.72709967e-02 6.18290365e-01
5.81279993e-01 -9.93169546e-02 -6.93786860e-01 5.31543970e-01
1.25941411e-01 -9.82756138e-01 8.90476763e-01 -4.38006073e-01
1.60258869e-03 -1.98569328e-01 1.71389058e-02 -1.79514182e+00
-6.35575175e-01 -8.52857769e-01 -3.16655338e-01 1.07603467e+00
4.95949388e-01 -1.06225479e+00 9.20484066e-01 -2.97509253e-01
9.46561247e-02 -8.41803849e-01 -6.35851562e-01 -7.68683314e-01
-1.02564998e-01 -6.35630032e-03 2.19063431e-01 1.52803826e+00
1.01683743e-01 9.11528528e-01 -4.40370768e-01 1.49412513e-01
9.85869706e-01 2.17826605e-01 7.92569220e-01 -1.46588683e+00
-3.63362819e-01 -5.64953148e-01 -7.40262389e-01 -6.97211802e-01
3.38292539e-01 -1.64225626e+00 -5.56571007e-01 -1.59167695e+00
2.36529797e-01 -2.33863652e-01 -1.87445849e-01 3.69819432e-01
-4.18401629e-01 1.79351240e-01 2.83359196e-02 -1.51140064e-01
-5.49367011e-01 6.70481503e-01 1.36006153e+00 -5.10180354e-01
-2.10644588e-01 -1.58561468e-01 -8.28512013e-01 5.95446706e-01
9.64584231e-01 -1.95717886e-01 -9.34450388e-01 4.01168726e-02
6.19062662e-01 -3.89594167e-01 2.82946557e-01 -8.46660852e-01
2.35676572e-01 1.00564368e-01 -5.16326493e-03 -3.24725658e-01
6.11378346e-03 -6.45704746e-01 1.01583138e-01 3.91583055e-01
-1.34546682e-01 1.35689899e-01 -1.78005874e-01 1.11758721e+00
-2.57233027e-02 2.98639923e-01 7.44018078e-01 -6.47777170e-02
-6.64984524e-01 5.92547953e-01 2.40069598e-01 4.46456820e-01
1.06081915e+00 -1.44156545e-01 -2.80266345e-01 -6.22040093e-01
-6.60710990e-01 1.81180909e-01 2.80488938e-01 4.49920565e-01
6.63829565e-01 -1.54045737e+00 -7.58371174e-01 1.66159555e-01
1.85379267e-01 -1.61217362e-01 2.18548596e-01 1.06231618e+00
-3.48302186e-01 -2.97183841e-02 1.45768300e-01 -6.42218411e-01
-1.26838601e+00 6.82227194e-01 3.56947303e-01 -4.98786807e-01
-8.29397082e-01 7.43406117e-01 2.35061854e-01 -7.99421906e-01
8.95033851e-02 1.38185890e-02 -2.87633568e-01 2.66996682e-01
-4.63501550e-02 6.76903605e-01 1.34339109e-01 -7.82836497e-01
-4.85136718e-01 7.54430354e-01 -1.39454708e-01 3.33010077e-01
1.39178932e+00 4.89198677e-02 -3.13917428e-01 5.20911872e-01
1.50856555e+00 2.64374286e-01 -8.48703325e-01 -3.84425521e-01
-9.58155990e-02 -3.72006208e-01 1.52469441e-01 -4.03161108e-01
-1.48703158e+00 6.19060218e-01 2.94457674e-01 7.02036679e-01
1.07598948e+00 8.26043189e-02 3.85662049e-01 4.79564935e-01
3.91340524e-01 -6.78181648e-01 5.76438904e-01 2.43348569e-01
1.12727153e+00 -1.39627850e+00 3.74423981e-01 -8.69712293e-01
-5.08890986e-01 6.13060474e-01 6.50999665e-01 -5.14525712e-01
1.07253122e+00 -5.13968468e-02 -2.80485511e-01 -7.52880692e-01
-4.84961480e-01 -2.46931389e-01 7.32727051e-01 7.53565133e-01
4.48801786e-01 2.67012507e-01 -3.65357816e-01 -8.90033320e-02
-1.50608078e-01 -8.08243275e-01 3.08070451e-01 7.30444908e-01
-1.67653307e-01 -9.93472219e-01 6.66348785e-02 8.96104217e-01
-1.54250324e-01 -3.14999074e-02 -9.71858978e-01 8.97155762e-01
-2.36129388e-01 8.46827745e-01 -4.07815337e-01 -6.23439133e-01
3.60594571e-01 -3.01321477e-01 6.53863966e-01 -6.20831966e-01
-2.21130446e-01 -2.51985252e-01 2.21154355e-02 -5.35850585e-01
-5.91186285e-01 -4.90747869e-01 -1.12661159e+00 -4.49157715e-01
-3.95146489e-01 4.71474715e-02 1.85156375e-01 4.77881134e-01
4.76944685e-01 5.76160252e-01 9.60338652e-01 -1.05712128e+00
-3.48012179e-01 -8.32845747e-01 -9.31603491e-01 5.91457129e-01
3.44877630e-01 -9.31723952e-01 -7.60985315e-01 -5.60031772e-01] | [7.241724967956543, 6.172756671905518] |
fd3a9fc8-5ada-4667-a722-b31b49d0c6f4 | eventhpe-event-based-3d-human-pose-and-shape | 2108.06819 | null | https://arxiv.org/abs/2108.06819v1 | https://arxiv.org/pdf/2108.06819v1.pdf | EventHPE: Event-based 3D Human Pose and Shape Estimation | Event camera is an emerging imaging sensor for capturing dynamics of moving objects as events, which motivates our work in estimating 3D human pose and shape from the event signals. Events, on the other hand, have their unique challenges: rather than capturing static body postures, the event signals are best at capturing local motions. This leads us to propose a two-stage deep learning approach, called EventHPE. The first-stage, FlowNet, is trained by unsupervised learning to infer optical flow from events. Both events and optical flow are closely related to human body dynamics, which are fed as input to the ShapeNet in the second stage, to estimate 3D human shapes. To mitigate the discrepancy between image-based flow (optical flow) and shape-based flow (vertices movement of human body shape), a novel flow coherence loss is introduced by exploiting the fact that both flows are originated from the identical human motion. An in-house event-based 3D human dataset is curated that comes with 3D pose and shape annotations, which is by far the largest one to our knowledge. Empirical evaluations on DHP19 dataset and our in-house dataset demonstrate the effectiveness of our approach. | ['Li Cheng', 'Minglun Gong', 'Shoushun Chen', 'Xiaoqin Hu', 'Pengyu Wang', 'Sen Wang', 'Xinxin Zuo', 'Chuan Guo', 'Shihao Zou'] | 2021-08-15 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Zou_EventHPE_Event-Based_3D_Human_Pose_and_Shape_Estimation_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Zou_EventHPE_Event-Based_3D_Human_Pose_and_Shape_Estimation_ICCV_2021_paper.pdf | iccv-2021-1 | ['3d-human-pose-and-shape-estimation'] | ['computer-vision'] | [-2.54658479e-02 -2.01746881e-01 1.48193344e-01 -1.89930081e-01
1.93552691e-02 -2.89340943e-01 4.37187642e-01 -1.82193145e-01
-3.66185129e-01 4.57628161e-01 6.26589179e-01 4.57641721e-01
2.42741376e-01 -6.71681166e-01 -6.07464612e-01 -3.54875118e-01
-2.94048488e-01 3.31229240e-01 4.01979804e-01 -8.28908011e-02
-1.04124807e-01 4.06411678e-01 -1.46354544e+00 -5.64536750e-02
4.69287217e-01 9.47308540e-01 -1.19478129e-01 9.28173661e-01
2.98174590e-01 1.09699094e+00 -2.78164804e-01 -3.59776437e-01
5.06750345e-01 -5.92805922e-01 -7.86614358e-01 4.23137248e-01
7.93450654e-01 -7.67865121e-01 -7.59800911e-01 6.97762370e-01
7.03819215e-01 3.23446572e-01 4.23619360e-01 -1.33320737e+00
-1.74120739e-01 -2.51334924e-02 -3.64210069e-01 4.24062163e-01
7.94687092e-01 6.66603267e-01 8.57933640e-01 -8.23021650e-01
1.01782584e+00 1.47340310e+00 7.06574321e-01 6.53469801e-01
-9.19353247e-01 -2.80876219e-01 2.51447596e-02 1.96431905e-01
-1.11253548e+00 -4.98789728e-01 9.82336223e-01 -8.04748893e-01
6.82121098e-01 -5.66906072e-02 1.34690797e+00 1.20661640e+00
9.31030288e-02 1.00860822e+00 5.98877847e-01 -5.28443381e-02
3.59504521e-02 -3.85374725e-01 -2.40772054e-01 9.41102207e-01
1.18250459e-01 3.12824100e-01 -8.52006733e-01 1.12187147e-01
1.18926096e+00 1.61410660e-01 -6.34817481e-01 -6.48778379e-01
-1.48907828e+00 6.98444545e-01 4.41337258e-01 -1.25466093e-01
-6.71607375e-01 4.24379766e-01 3.93661350e-01 5.22920825e-02
5.21142483e-01 -8.17825645e-02 -2.27354199e-01 -4.15073007e-01
-6.32928014e-01 5.65542758e-01 9.22138333e-01 6.76848471e-01
5.51833391e-01 -8.89283512e-03 -2.71094054e-01 2.84284860e-01
5.53723037e-01 6.85688972e-01 2.68574715e-01 -9.38255727e-01
5.02486110e-01 6.13019347e-01 3.18731993e-01 -1.39292240e+00
-4.95146096e-01 -1.11617774e-01 -7.47497797e-01 -8.36458951e-02
6.94640636e-01 -2.34004334e-01 -6.47728682e-01 1.76779926e+00
7.29851723e-01 6.77760363e-01 -2.73837537e-01 1.54271138e+00
7.80339777e-01 5.35411119e-01 1.10782962e-02 -8.29803869e-02
1.23350763e+00 -8.82279336e-01 -6.61514997e-01 -1.31467178e-01
2.24515393e-01 -5.84569693e-01 7.31512666e-01 1.96326956e-01
-1.27342665e+00 -8.53883862e-01 -8.02569449e-01 -1.45245016e-01
3.97330374e-02 -1.88619763e-01 4.69735116e-01 4.04188663e-01
-7.00738072e-01 5.49863815e-01 -1.29746139e+00 -5.14858663e-01
3.39296132e-01 1.55299213e-02 -2.89364070e-01 8.05189684e-02
-1.18960500e+00 5.74066818e-01 1.03369169e-01 4.27300453e-01
-8.45369458e-01 -8.29100668e-01 -1.12449563e+00 -2.79466808e-01
3.89577568e-01 -1.24034917e+00 9.63881314e-01 -7.55203843e-01
-1.56924391e+00 8.50060105e-01 -2.48743668e-01 -5.95803559e-01
9.65702951e-01 -6.28064811e-01 -1.51615113e-01 5.24185896e-01
-3.51583883e-02 6.51002288e-01 9.51911747e-01 -9.86883223e-01
-4.50580627e-01 -3.10302258e-01 -2.21255980e-02 2.38927081e-01
-1.61116764e-01 -1.41403735e-01 -6.21387005e-01 -6.06755197e-01
-4.41931821e-02 -1.00599730e+00 -6.56085461e-02 3.50876987e-01
-3.67518395e-01 -4.72740382e-02 7.93108582e-01 -6.93070352e-01
1.35814428e+00 -1.90828478e+00 4.49651927e-01 2.35257559e-02
4.09233063e-01 1.78526103e-01 8.33515003e-02 1.57404780e-01
2.07340717e-01 -3.78476590e-01 -2.72489488e-01 -5.42320073e-01
-1.35302141e-01 2.09679037e-01 -2.97685474e-01 8.46498847e-01
4.57788825e-01 1.20957923e+00 -1.23451161e+00 -6.48214400e-01
6.64107084e-01 6.98531508e-01 -6.07532144e-01 4.01040882e-01
-1.79846644e-01 9.98677015e-01 -3.97108972e-01 4.75627393e-01
5.44277787e-01 -3.90190661e-01 -1.04550578e-01 -4.92256105e-01
3.53124226e-03 6.54254928e-02 -1.34092319e+00 2.13218236e+00
-1.59124419e-01 3.75223339e-01 -2.25432798e-01 -8.53535056e-01
6.03587270e-01 4.79513228e-01 1.14215779e+00 -4.48904753e-01
2.54825056e-01 -9.44309384e-02 -1.46315798e-01 -8.15752447e-01
3.66741896e-01 6.42107576e-02 3.50976586e-02 3.54693383e-01
1.39959946e-01 5.98952323e-02 1.68175653e-01 1.20268077e-01
1.09725451e+00 6.46438241e-01 2.77334064e-01 3.42678726e-02
6.73379540e-01 -8.07502046e-02 7.46079683e-01 5.00887513e-01
-7.23080993e-01 9.04455364e-01 1.69516355e-01 -9.21509147e-01
-9.66267467e-01 -1.28476119e+00 3.25575262e-01 5.58280885e-01
4.76730734e-01 -4.12024021e-01 -5.28247237e-01 -6.68076754e-01
9.02289897e-02 -2.70078450e-01 -5.68203986e-01 -3.39324586e-02
-1.05080831e+00 -3.45573962e-01 3.76111180e-01 5.03099322e-01
6.48456931e-01 -1.13823509e+00 -1.32864559e+00 4.92489159e-01
-6.63338363e-01 -1.59807444e+00 -9.08427358e-01 -5.63519537e-01
-8.21575701e-01 -1.41039050e+00 -9.03409600e-01 -3.22257876e-01
4.85121936e-01 3.78290154e-02 1.26775920e+00 -2.07350571e-02
-5.52667677e-01 7.92458653e-01 -3.03529054e-01 -2.84480453e-01
-4.98357713e-02 -1.92365855e-01 8.46454799e-02 5.60461223e-01
2.20488682e-01 -6.89352632e-01 -1.13407815e+00 2.38159120e-01
-6.86474621e-01 2.99503077e-02 1.66379094e-01 6.39877617e-01
5.23865044e-01 -2.51230747e-01 1.04455771e-02 -4.48302329e-01
8.90500173e-02 -3.63467634e-01 -4.05941427e-01 -1.32088438e-01
1.60169989e-01 -2.14617640e-01 4.00906175e-01 -4.73095864e-01
-1.15066028e+00 5.41507006e-01 1.11597767e-02 -7.52425075e-01
-2.24770471e-01 1.13249458e-01 -5.03610484e-02 1.97626829e-01
5.18328369e-01 4.22803834e-02 1.80150568e-01 -2.59961933e-01
3.05204481e-01 1.37039786e-02 9.79809821e-01 -3.48448396e-01
8.33937585e-01 1.04307139e+00 2.92245477e-01 -9.89775717e-01
-9.97168481e-01 -8.00474942e-01 -9.20872033e-01 -6.84043467e-01
1.22748184e+00 -1.16939664e+00 -8.67881954e-01 9.82846737e-01
-1.26649165e+00 -2.97252357e-01 -5.08334398e-01 7.16451466e-01
-8.12737107e-01 5.05359173e-01 -7.42132545e-01 -8.53485584e-01
-3.48255545e-01 -8.29342663e-01 1.46484935e+00 1.28341898e-01
-2.99463481e-01 -1.16186619e+00 2.64439791e-01 2.67185628e-01
1.37680888e-01 9.43639338e-01 6.91252276e-02 -3.37079018e-02
-6.75647318e-01 -2.94822659e-02 -2.37065665e-02 1.55448750e-01
1.74244508e-01 -1.87862948e-01 -8.47974181e-01 -3.34782481e-01
1.12988584e-01 -2.48312384e-01 7.48017073e-01 6.17077172e-01
6.40119672e-01 -3.70425545e-02 -1.02463990e-01 8.60800624e-01
1.15930307e+00 -3.62734139e-01 5.33971846e-01 -7.03823790e-02
1.14149857e+00 6.92624152e-01 5.61144590e-01 8.89637351e-01
5.28286994e-01 9.14311647e-01 3.51194799e-01 -2.59624254e-02
-5.82634628e-01 -5.35170257e-01 5.97974539e-01 8.78918827e-01
-3.76007587e-01 -6.39587492e-02 -8.53358388e-01 5.02884865e-01
-2.01561880e+00 -1.01372766e+00 -3.32200825e-01 2.07790565e+00
6.56334043e-01 1.57068241e-02 4.42086279e-01 -2.95167435e-02
5.99585772e-01 3.04904550e-01 -6.09005809e-01 5.24503171e-01
-3.23293991e-02 1.93536788e-01 7.30035827e-02 5.14854074e-01
-1.29237342e+00 8.51403713e-01 5.21144342e+00 1.16382293e-01
-1.28545535e+00 -1.45810700e-04 1.32722527e-01 -2.84463942e-01
1.34853527e-01 -9.94902104e-02 -6.63699865e-01 5.73021770e-01
5.01283824e-01 9.42371562e-02 8.44068266e-03 3.17120373e-01
6.57618761e-01 1.30162593e-02 -1.14372015e+00 1.19208837e+00
1.37948290e-01 -1.02681637e+00 1.81987900e-02 5.14072925e-02
5.23354292e-01 -1.01541191e-01 -3.52662295e-01 -1.17409810e-01
1.32274434e-01 -6.33260250e-01 9.38454509e-01 8.84424508e-01
5.32032549e-01 -5.53616583e-01 4.05785203e-01 2.28924900e-01
-1.64558268e+00 2.32122570e-01 -1.06397130e-01 -1.89895540e-01
8.32701206e-01 7.46238291e-01 -3.13599080e-01 5.93016505e-01
7.82032728e-01 1.40086198e+00 -2.68557638e-01 1.03723586e+00
-2.68507957e-01 5.09033144e-01 -4.02502507e-01 4.43957239e-01
-1.41535819e-01 -1.06086671e-01 9.66167271e-01 1.06481898e+00
1.35046393e-01 1.19138412e-01 5.95415294e-01 1.03136098e+00
1.16014652e-01 -1.51732951e-01 -6.29211605e-01 2.04584002e-01
7.82287717e-02 1.16962850e+00 -5.99961758e-01 -3.11421961e-01
-4.72050160e-01 1.33995736e+00 2.86174566e-02 2.78907806e-01
-8.89472961e-01 -1.39061242e-01 7.63454378e-01 2.49990731e-01
2.77426422e-01 -4.75495964e-01 2.19898984e-01 -1.60961938e+00
1.74300373e-01 -4.32343364e-01 4.24649894e-01 -5.65989614e-01
-1.42120838e+00 1.65473655e-01 -6.58566207e-02 -1.49480069e+00
-3.68865967e-01 -3.63895416e-01 -5.26743710e-01 6.07547581e-01
-1.48755062e+00 -1.10694504e+00 -8.61399412e-01 1.07248652e+00
4.42674696e-01 3.15851003e-01 3.23721349e-01 4.47777361e-01
-4.14180726e-01 1.92619741e-01 -6.85871482e-01 4.76913214e-01
7.55147755e-01 -1.14543307e+00 6.89798892e-01 1.09127593e+00
2.37533838e-01 2.74377882e-01 4.90790695e-01 -7.73667634e-01
-1.69390082e+00 -1.17739511e+00 7.94988275e-01 -7.41643608e-01
4.41827744e-01 -3.94991845e-01 -6.50562644e-01 5.96671224e-01
-2.32305601e-01 6.24396563e-01 2.73126006e-01 -5.37296116e-01
-5.34880236e-02 8.06485713e-02 -8.48513424e-01 3.51181984e-01
1.65598488e+00 -5.28706431e-01 -6.34017289e-01 4.56290729e-02
6.51090145e-01 -7.17457950e-01 -9.78557289e-01 4.82872516e-01
6.88827395e-01 -1.07823753e+00 1.26836169e+00 -4.87771630e-01
5.33052206e-01 -4.27232474e-01 1.08626701e-01 -1.07607901e+00
-4.54850215e-03 -9.41955924e-01 -7.84266293e-01 8.81381094e-01
-3.49920422e-01 -3.08341265e-01 1.04731691e+00 3.57177526e-01
1.42550483e-01 -4.34145182e-01 -6.64774597e-01 -7.05349982e-01
-4.06796157e-01 -4.20262247e-01 3.27195525e-01 8.75191808e-01
-4.60009038e-01 1.91598713e-01 -7.18200684e-01 -8.23034253e-03
1.00363958e+00 1.36238158e-01 1.15970111e+00 -1.38012564e+00
-3.83573830e-01 -9.40255821e-02 -7.00468302e-01 -1.57670808e+00
7.02291429e-02 -6.11939251e-01 1.47364587e-01 -1.18962491e+00
-5.25518171e-02 9.08254683e-02 4.68877442e-02 1.13730282e-01
-3.97490382e-01 3.18102568e-01 4.49899286e-01 2.99624622e-01
-6.30115151e-01 6.82861745e-01 1.48657620e+00 9.72943604e-02
-4.29695159e-01 -1.87929608e-02 -2.99948975e-02 9.83125269e-01
2.75868177e-01 -1.11890540e-01 -4.15708780e-01 -4.13595229e-01
4.86499257e-02 2.33285353e-01 9.88810182e-01 -1.21206713e+00
3.84342521e-01 6.55957684e-02 4.56503212e-01 -6.54949367e-01
3.13416183e-01 -8.10199797e-01 2.07008287e-01 7.02728629e-01
-2.00460821e-01 1.36878074e-03 -1.68537647e-01 8.11253548e-01
-2.55833566e-01 4.31337833e-01 6.08197153e-01 -3.50651354e-01
-9.20386255e-01 8.89660895e-01 -9.71128419e-03 5.20558596e-01
7.88514197e-01 -4.33638781e-01 1.30083278e-01 -4.38792497e-01
-8.50931585e-01 2.51201332e-01 3.53983849e-01 6.27206206e-01
7.72287011e-01 -1.44135058e+00 -9.20585334e-01 3.19894880e-01
1.26501411e-01 2.89859712e-01 3.49069953e-01 1.00937593e+00
-6.06126904e-01 9.54339653e-02 -2.48385370e-01 -1.00617754e+00
-8.69414985e-01 2.43374020e-01 4.65351462e-01 -2.75049448e-01
-1.14071274e+00 7.07972944e-01 2.23426670e-01 -2.18170792e-01
2.43091494e-01 -2.63607681e-01 -4.08599712e-02 -2.55132951e-02
5.02155900e-01 6.90186203e-01 -1.95075184e-01 -9.49481905e-01
-3.24849516e-01 8.98584723e-01 4.20685649e-01 -1.44822836e-01
1.11331654e+00 -4.72636968e-01 9.32583064e-02 4.33403969e-01
1.29896891e+00 -1.88196495e-01 -1.79248190e+00 -3.16750109e-01
-1.73205033e-01 -6.78003311e-01 -2.48329267e-01 -3.23434711e-01
-1.38949740e+00 9.72735822e-01 6.09913468e-01 -2.81830505e-02
1.15511692e+00 -2.62304962e-01 1.42535865e+00 -9.19689983e-03
4.05278176e-01 -8.57759774e-01 4.75812644e-01 5.29431283e-01
7.11311936e-01 -1.17497933e+00 -1.01742372e-01 -6.28724456e-01
-3.93143147e-01 1.08425295e+00 6.76023901e-01 -3.56740147e-01
7.23509848e-01 8.73992741e-02 -1.16534539e-01 -2.71923989e-01
-5.40743291e-01 -4.56963331e-01 3.59662116e-01 6.78995550e-01
2.88092971e-01 -2.34783009e-01 -1.62079528e-01 1.69585958e-01
-2.61290878e-01 3.77905488e-01 3.19698811e-01 9.78676915e-01
-1.38303101e-01 -7.45902181e-01 -3.38716984e-01 -6.82278490e-03
-3.80564481e-01 3.05138886e-01 -2.38977820e-01 6.83664143e-01
3.02766770e-01 7.18962491e-01 2.32511893e-01 -2.19419122e-01
7.20009446e-01 -1.04854017e-01 7.11868346e-01 -3.36138010e-01
-4.79185849e-01 -1.03678569e-01 -1.50058135e-01 -1.24411857e+00
-1.02829432e+00 -6.81594372e-01 -1.26374280e+00 -2.93913066e-01
1.00203335e-01 -3.08830798e-01 1.40214428e-01 8.02247941e-01
3.53803039e-01 3.55193496e-01 5.11713028e-01 -1.18845117e+00
-2.66875803e-01 -5.56795299e-01 -4.56135899e-01 1.30218661e+00
5.82130075e-01 -8.10707748e-01 -3.37775320e-01 6.47724092e-01] | [7.336039066314697, -0.6248089671134949] |
af614ad4-f8ac-42e0-91d3-82a72d1f76ff | can-spoofing-countermeasure-and-speaker | 2303.07073 | null | https://arxiv.org/abs/2303.07073v3 | https://arxiv.org/pdf/2303.07073v3.pdf | Can spoofing countermeasure and speaker verification systems be jointly optimised? | Spoofing countermeasure (CM) and automatic speaker verification (ASV) sub-systems can be used in tandem with a backend classifier as a solution to the spoofing aware speaker verification (SASV) task. The two sub-systems are typically trained independently to solve different tasks. While our previous work demonstrated the potential of joint optimisation, it also showed a tendency to over-fit to speakers and a lack of sub-system complementarity. Using only a modest quantity of auxiliary data collected from new speakers, we show that joint optimisation degrades the performance of separate CM and ASV sub-systems, but that it nonetheless improves complementarity, thereby delivering superior SASV performance. Using standard SASV evaluation data and protocols, joint optimisation reduces the equal error rate by 27\% relative to performance obtained using fixed, independently-optimised sub-systems under like-for-like training conditions. | ['Nicholas Evans', 'Massimiliano Todisco', 'Hemlata Tak', 'Wanying Ge'] | 2023-03-13 | null | null | null | null | ['speaker-verification'] | ['speech'] | [ 3.60591710e-01 2.68797636e-01 8.28418136e-02 -5.14885604e-01
-1.26386726e+00 -7.95367658e-01 1.05090225e+00 1.56735796e-02
-4.50292587e-01 4.51364458e-01 2.24002391e-01 -9.02758837e-01
1.59999862e-01 -6.53863996e-02 -4.46812361e-01 -5.68243682e-01
-1.95973098e-01 3.75671208e-01 2.95191184e-02 -5.28757870e-01
5.85556254e-02 3.17724377e-01 -1.52088213e+00 2.57745892e-01
4.03511077e-01 5.39654016e-01 -3.09341520e-01 1.23618281e+00
3.77304815e-02 5.40403366e-01 -1.17984617e+00 -6.64110959e-01
2.81150430e-01 -2.13631377e-01 -9.76573825e-01 -2.98992276e-01
6.49480283e-01 -1.47309929e-01 -3.29361171e-01 9.71948624e-01
9.96237934e-01 -1.01556577e-01 1.32667273e-01 -1.53242636e+00
-2.44537711e-01 9.07463789e-01 -1.29008129e-01 2.95612305e-01
8.35201561e-01 1.44984826e-01 8.14607203e-01 -3.49778205e-01
3.84160936e-01 1.46646941e+00 1.00431025e+00 7.30903089e-01
-1.44393969e+00 -9.51928556e-01 -1.90555260e-01 -5.80936596e-02
-1.28894901e+00 -1.38683236e+00 6.15640879e-01 -8.82276595e-02
1.23049080e+00 6.90416873e-01 2.33287096e-01 1.30174148e+00
-3.24177653e-01 7.33616769e-01 1.08254659e+00 -3.71564567e-01
5.42705581e-02 7.34381258e-01 1.81675568e-01 4.23376679e-01
8.36817827e-03 5.82002163e-01 -6.09051704e-01 -6.63154066e-01
-5.71125783e-02 -7.47550845e-01 -3.10922772e-01 -5.75132407e-02
-1.19979310e+00 1.03021252e+00 1.84179947e-01 2.84648925e-01
-5.50992340e-02 -8.10946673e-02 6.33353293e-01 7.23618925e-01
2.42742598e-01 4.53965455e-01 -5.10179043e-01 -1.99902594e-01
-1.24214673e+00 4.03334826e-01 1.28098011e+00 7.04739630e-01
4.64041442e-01 3.01153928e-01 9.47339111e-04 6.34537101e-01
6.23232424e-01 6.21672690e-01 7.20706642e-01 -6.79380953e-01
7.32109785e-01 1.52851744e-02 2.81601418e-02 -8.12450349e-01
-2.17577145e-01 -4.77091670e-01 -2.93223113e-01 5.52055724e-02
4.76323962e-01 -3.31180245e-01 -8.28137279e-01 1.83795524e+00
4.19160545e-01 2.43968919e-01 2.02332467e-01 4.76575136e-01
7.85268366e-01 4.49235767e-01 -5.47938198e-02 -2.26874590e-01
1.36556208e+00 -8.00424397e-01 -5.77003658e-01 -4.65667546e-01
9.97299612e-01 -1.05282092e+00 5.23760140e-01 1.56070933e-01
-9.37269628e-01 -2.69522190e-01 -1.17458510e+00 4.03317958e-01
-4.20368850e-01 -5.64746320e-01 3.46447855e-01 1.78034639e+00
-1.63461173e+00 3.88510942e-01 -4.75750446e-01 -2.68309593e-01
3.44637245e-01 7.50291944e-01 -6.10759437e-01 2.87826568e-01
-1.40661621e+00 1.04745066e+00 1.77372724e-01 -1.55892551e-01
-1.15669668e+00 -4.58707988e-01 -9.86498415e-01 -9.93916243e-02
-1.10107191e-01 -3.06179702e-01 1.43795824e+00 -5.82659543e-01
-1.86983645e+00 9.20816481e-01 -3.71411145e-01 -7.58880615e-01
5.51678538e-01 7.52411336e-02 -9.52808499e-01 -7.31721073e-02
-1.32100031e-01 3.84478480e-01 1.28424597e+00 -1.26023042e+00
-4.34372842e-01 -3.15479815e-01 -3.16502810e-01 -1.30510956e-01
-2.97951430e-01 6.93070829e-01 3.79541814e-02 -3.08168262e-01
2.09940672e-02 -9.33048368e-01 3.25639509e-02 -7.59750843e-01
-5.29074371e-01 3.48057561e-02 1.03632438e+00 -1.09277391e+00
1.08493769e+00 -2.15013647e+00 -2.31670782e-01 4.86290097e-01
-2.74089482e-02 9.55149114e-01 -3.51150393e-01 3.69972318e-01
-4.01964307e-01 3.53844881e-01 -2.88324535e-01 -9.27562714e-01
1.51340753e-01 1.05623223e-01 -6.70195445e-02 7.72408783e-01
4.31432873e-02 6.25471592e-01 -8.27872694e-01 -3.73580754e-01
1.90752983e-01 6.46039188e-01 -4.68626946e-01 1.70452192e-01
3.68626028e-01 1.79019615e-01 2.08765045e-01 5.84808528e-01
8.90070438e-01 1.72534570e-01 1.38295650e-01 4.16762561e-01
2.33089298e-01 8.47991765e-01 -1.44704497e+00 1.04966152e+00
-3.15336764e-01 8.42665195e-01 9.22686994e-01 -1.07215977e+00
8.93053114e-01 8.54132235e-01 1.72303155e-01 -4.50017810e-01
3.22489798e-01 3.69921774e-01 1.61359444e-01 -4.11828846e-01
5.44028938e-01 -2.82452434e-01 6.46997895e-03 6.58273757e-01
1.79298773e-01 -1.71764001e-01 -4.05040771e-01 3.09697241e-01
1.19133472e+00 -6.42794847e-01 2.11480409e-01 -1.99672088e-01
1.07677329e+00 -3.34105730e-01 2.26718411e-01 8.94634962e-01
-8.60086322e-01 2.68792897e-01 -1.83862634e-02 1.55342981e-01
-8.54564369e-01 -6.77301884e-01 -4.40470397e-01 1.24419439e+00
-3.60851809e-02 -5.41007757e-01 -8.80683124e-01 -7.65891373e-01
2.78096527e-01 8.27815413e-01 -4.38166380e-01 -6.57873005e-02
-6.36622369e-01 -7.93510318e-01 1.47822464e+00 -5.81784956e-02
2.59333879e-01 -8.84915769e-01 -2.54585091e-02 3.99523824e-01
-2.72855222e-01 -1.26237166e+00 -3.19992393e-01 2.39703685e-01
-4.32062060e-01 -5.14616311e-01 -5.51354527e-01 -7.53158808e-01
1.70464531e-01 4.18090314e-01 9.31786120e-01 3.03846836e-01
1.26403481e-01 2.62427747e-01 -1.80649564e-01 -4.88384873e-01
-1.37716579e+00 1.60562217e-01 4.14134085e-01 -1.47675853e-02
4.31255877e-01 -3.08783263e-01 -5.46083972e-02 5.16551375e-01
-5.06936729e-01 -6.31617606e-01 1.77535027e-01 1.02186751e+00
-5.55515289e-01 -1.94386348e-01 4.67919588e-01 -5.22953629e-01
7.61309266e-01 -3.18791837e-01 -4.18572903e-01 -2.61085145e-02
-8.15304935e-01 2.49989718e-01 2.18347814e-02 -4.79444474e-01
-7.67388940e-01 -5.50310127e-02 -4.46101397e-01 -9.60828885e-02
-3.53297830e-01 4.43901867e-02 -2.43603408e-01 -7.29626894e-01
8.06679904e-01 3.47953171e-01 6.62048221e-01 -3.88007909e-01
2.88581312e-01 1.34315264e+00 5.27570844e-01 -5.89012429e-02
9.98616934e-01 1.88669786e-01 -4.85996068e-01 -9.57178295e-01
-1.51066408e-01 -7.85998166e-01 -2.28504792e-01 2.42385656e-01
4.71835613e-01 -9.59983528e-01 -9.74477708e-01 5.45569420e-01
-1.19666946e+00 -1.79797009e-01 1.29948571e-01 3.10750008e-01
-3.54624651e-02 8.87171984e-01 -4.80749130e-01 -1.19238317e+00
-4.86755282e-01 -1.31776047e+00 1.21833646e+00 -3.51248860e-01
-4.26839203e-01 -1.04594076e+00 2.39833832e-01 9.14438486e-01
6.96227252e-01 -2.33803049e-01 7.79506415e-02 -1.35949063e+00
1.07856117e-01 -5.54155111e-01 -1.12932464e-02 5.51839054e-01
-8.78289808e-03 -2.43790448e-01 -1.61355138e+00 -7.30140030e-01
1.86635345e-01 -7.98166618e-02 8.00529361e-01 -4.57217544e-03
3.88017684e-01 -6.73199296e-01 -3.05364788e-01 6.01745129e-01
8.96987498e-01 -1.07997559e-01 3.72158170e-01 5.46824932e-01
3.88403416e-01 8.07197213e-01 -5.23771644e-02 1.43444926e-01
3.94566208e-01 9.35341775e-01 3.58982831e-01 8.17645118e-02
-7.12328870e-03 -2.25695502e-02 9.85021412e-01 6.69861853e-01
3.87194216e-01 -2.76558548e-01 -1.01821983e+00 5.74857950e-01
-1.27992296e+00 -1.30365551e+00 -1.70120165e-01 2.28611803e+00
6.19460344e-01 1.00041546e-01 6.06331587e-01 6.33836329e-01
9.02249813e-01 3.83476734e-01 -1.71723403e-02 -8.90467703e-01
-2.11950079e-01 -1.73197761e-02 7.41972208e-01 1.10469353e+00
-1.07074618e+00 8.32517564e-01 7.68596554e+00 5.71836591e-01
-1.13862836e+00 3.70681077e-01 3.29954214e-02 7.64138103e-02
-1.86175033e-01 5.05517460e-02 -9.51471806e-01 4.71735150e-01
1.58278394e+00 1.02864437e-01 7.31885195e-01 5.22423327e-01
-1.36857331e-01 4.89280283e-01 -9.34530377e-01 1.10188079e+00
3.12115908e-01 -9.64560986e-01 -3.78589302e-01 4.91109341e-01
3.43587726e-01 5.11066139e-01 5.81356585e-02 3.80176932e-01
6.84713483e-01 -1.01684701e+00 7.94846833e-01 -5.02007186e-01
4.93092149e-01 -4.35524672e-01 1.01444542e+00 4.51273620e-01
-9.20935035e-01 -3.53674352e-01 1.27262339e-01 4.64363396e-02
4.55110252e-01 1.74615771e-01 -1.57530797e+00 4.86471832e-01
4.12052423e-01 -1.82580967e-02 -6.21371806e-01 6.72243714e-01
-1.09763011e-01 9.44369316e-01 -5.42530119e-01 -1.84161495e-02
1.19052231e-01 5.17951012e-01 1.04495740e+00 1.60816205e+00
-4.20454182e-02 -3.55930656e-01 -2.36499518e-01 2.11410463e-01
1.39845446e-01 -1.16183117e-01 -4.27364022e-01 -1.17974319e-01
7.09654272e-01 9.18002307e-01 7.13003576e-02 -4.90972698e-01
-1.12860035e-02 9.77051377e-01 1.21833161e-01 1.49394438e-01
-6.12430692e-01 -3.47256452e-01 9.49053884e-01 -3.03538088e-02
3.34511280e-01 -6.75580353e-02 -2.43816689e-01 -1.15743685e+00
-2.19405204e-01 -1.35470426e+00 3.74378264e-01 -1.45252600e-01
-9.17495966e-01 6.97694361e-01 -2.43400738e-01 -7.45508671e-01
-7.17012703e-01 -3.06499839e-01 -6.86100423e-01 1.03902805e+00
-1.46933174e+00 -9.89065289e-01 9.11191627e-02 7.00273991e-01
-5.63109815e-02 -5.24904311e-01 1.06715322e+00 2.61161923e-01
-6.30021036e-01 1.23441648e+00 1.69384316e-01 2.62193561e-01
6.86038792e-01 -1.14607286e+00 9.65542793e-01 9.76556778e-01
3.40279669e-01 8.00273418e-01 1.11493337e+00 -3.85641277e-01
-1.27706861e+00 -6.61013842e-01 1.38857210e+00 -8.30701768e-01
5.57829797e-01 -5.58440089e-01 -8.91562879e-01 5.46781361e-01
1.94183350e-01 -2.08260506e-01 7.82275140e-01 3.45430017e-01
-6.87445343e-01 1.86353907e-01 -1.56227815e+00 9.05621797e-02
8.30799580e-01 -1.10063469e+00 -7.14658439e-01 2.45400131e-01
7.10410714e-01 -2.46002957e-01 -7.98674583e-01 2.91802771e-02
5.79961300e-01 -1.05182862e+00 1.32945192e+00 -5.96336842e-01
-3.73639494e-01 -1.11583613e-01 -3.66063446e-01 -1.23393297e+00
-2.14088425e-01 -1.33634210e+00 -1.85017034e-01 1.64213336e+00
5.84812403e-01 -1.13387096e+00 8.64616930e-01 5.01497388e-01
1.27834365e-01 2.38680579e-02 -1.29026568e+00 -1.03469908e+00
5.65948896e-02 -8.02890062e-01 1.06194818e+00 1.18108809e+00
2.68530637e-01 2.31925339e-01 -4.03168052e-01 5.75076997e-01
8.07322800e-01 -5.67554533e-01 1.17554569e+00 -9.28306341e-01
-5.39419889e-01 -5.82560480e-01 -7.99025714e-01 -7.01577604e-01
2.49863431e-01 -1.06372952e+00 1.07409991e-01 -7.95487463e-01
-2.97702879e-01 -3.96423727e-01 -2.89877802e-01 4.59093809e-01
-3.66866052e-01 2.47896999e-01 3.65568221e-01 1.83567166e-01
-2.37474218e-01 1.70738891e-01 4.00647879e-01 -2.38363102e-01
-4.35523093e-01 1.99974939e-01 -8.53044033e-01 1.32002398e-01
7.47584581e-01 -6.60781622e-01 -5.00931256e-02 -3.15291345e-01
-2.61904597e-01 1.50708735e-01 4.77119833e-01 -9.39484000e-01
2.31933430e-01 4.50705081e-01 5.61208948e-02 -3.07356436e-02
4.47063208e-01 -5.17753065e-01 3.23322378e-02 7.20737159e-01
-4.32571232e-01 -7.83178508e-02 3.09456050e-01 3.24911326e-01
-1.08300559e-01 -1.49038270e-01 7.52278209e-01 1.97230577e-01
-3.43189061e-01 -1.03911735e-01 -4.72034901e-01 -1.29066095e-01
6.09701812e-01 -3.11351597e-01 -4.49504465e-01 -6.86074555e-01
-4.79554325e-01 1.47294402e-01 3.78902197e-01 3.79637063e-01
2.91469395e-01 -9.00590837e-01 -1.12714231e+00 6.17640793e-01
7.94453621e-02 -5.45829415e-01 -4.62824889e-02 7.77094364e-01
-2.76125669e-01 6.75110519e-01 1.45811811e-01 -7.62464225e-01
-1.94585943e+00 5.53879261e-01 3.77364010e-01 -1.23837255e-01
-2.57127076e-01 1.31334472e+00 -5.32447636e-01 -1.11701620e+00
3.34206849e-01 3.12751293e-01 1.19956611e-02 -9.18091163e-02
8.44556928e-01 3.80238652e-01 4.41269398e-01 -1.40847707e+00
-8.51795316e-01 -7.78621212e-02 -2.18943164e-01 -6.53018534e-01
1.02964425e+00 -1.98802531e-01 2.27446273e-01 -1.70806661e-01
1.60175908e+00 3.03739935e-01 -7.35212147e-01 -3.45952868e-01
3.50663774e-02 -4.90140885e-01 2.62194395e-01 -6.29563630e-01
-5.68710506e-01 9.05070603e-01 7.04175234e-01 6.24140978e-01
5.05058348e-01 2.44669337e-02 7.01405346e-01 1.76275790e-01
1.74248904e-01 -6.67944968e-01 -5.39839208e-01 2.86410809e-01
5.51582694e-01 -1.40943456e+00 -6.12700544e-02 -3.58150244e-01
-7.12110400e-01 6.07443511e-01 -7.29969591e-02 4.29256678e-01
6.46902859e-01 1.88391313e-01 4.98046547e-01 -6.94347546e-02
-5.30800223e-01 1.55466184e-01 -4.48923698e-03 1.03066564e+00
2.48759151e-01 2.47833550e-01 2.81656235e-01 3.42242308e-02
-9.35499251e-01 -4.65131700e-01 2.76233613e-01 1.02825642e+00
-1.45738497e-01 -1.29921341e+00 -8.41790020e-01 1.10436708e-01
-6.43685937e-01 -1.27176061e-01 -5.18816948e-01 5.15659273e-01
-4.00258154e-02 1.62071669e+00 -3.04857403e-01 -9.62606370e-01
-2.68241242e-02 3.25082600e-01 1.55666530e-01 -1.89845070e-01
-1.21549833e+00 -2.75209218e-01 5.93020558e-01 -4.62254226e-01
-3.37897509e-01 -1.01403368e+00 -5.09231746e-01 -7.51425743e-01
-9.18576241e-01 4.21933174e-01 1.27936673e+00 1.07893801e+00
4.94310588e-01 4.13377546e-02 1.02755904e+00 -8.11079979e-01
-1.28751636e+00 -1.05763757e+00 -3.08708161e-01 3.76925975e-01
1.03273082e+00 -2.36978978e-01 -9.73375499e-01 -2.31235489e-01] | [14.125195503234863, 5.933498859405518] |
c230f1a1-32bc-4633-853a-606ba594f312 | hero-roberta-and-longformer-hebrew-language | 2304.11077 | null | https://arxiv.org/abs/2304.11077v1 | https://arxiv.org/pdf/2304.11077v1.pdf | HeRo: RoBERTa and Longformer Hebrew Language Models | In this paper, we fill in an existing gap in resources available to the Hebrew NLP community by providing it with the largest so far pre-train dataset HeDC4, a state-of-the-art pre-trained language model HeRo for standard length inputs and an efficient transformer LongHeRo for long input sequences. The HeRo model was evaluated on the sentiment analysis, the named entity recognition, and the question answering tasks while the LongHeRo model was evaluated on the document classification task with a dataset composed of long documents. Both HeRo and LongHeRo presented state-of-the-art performance. The dataset and model checkpoints used in this work are publicly available. | ['Harel Haskey', 'Vitaly Shalumov'] | 2023-04-18 | null | null | null | null | ['document-classification'] | ['natural-language-processing'] | [-4.91231568e-02 2.28837449e-02 -2.47120395e-01 -4.94044811e-01
-1.03653908e+00 -7.99153745e-01 7.04222739e-01 2.99425781e-01
-8.63930285e-01 8.02772403e-01 3.69067311e-01 -4.83207375e-01
-8.89474228e-02 -5.36904156e-01 -4.34181720e-01 -1.64736465e-01
7.53298402e-02 9.81904387e-01 1.22918263e-01 -4.99535471e-01
3.25325906e-01 3.49726617e-01 -1.32984543e+00 9.25412834e-01
5.44361889e-01 1.04764712e+00 -2.70175971e-02 8.46952081e-01
-4.01176929e-01 1.07878828e+00 -3.64513069e-01 -7.99404919e-01
-9.26255435e-02 -2.58331269e-01 -1.17242956e+00 -3.72800112e-01
2.84296721e-01 1.46005511e-01 4.17986810e-02 5.18840373e-01
7.12719500e-01 -4.94825765e-02 3.92482221e-01 -1.02306008e+00
-6.80500269e-01 1.01420057e+00 2.70864367e-01 2.76076406e-01
4.43021238e-01 -1.24708287e-01 1.26484489e+00 -1.11459482e+00
1.06445885e+00 1.01344109e+00 6.97122395e-01 5.68571031e-01
-8.09066653e-01 -5.24576843e-01 -2.10252732e-01 2.24899426e-01
-1.28384089e+00 -4.68870699e-01 4.64681029e-01 -4.03869838e-01
1.77106798e+00 1.79931253e-01 2.98983932e-01 1.29041553e+00
2.35014156e-01 9.71445560e-01 1.20021474e+00 -6.40814424e-01
2.82810062e-01 3.83286744e-01 5.43063581e-01 6.70558989e-01
-3.90919268e-01 2.51498580e-01 -5.86209238e-01 -2.02476785e-01
-5.02100997e-02 -3.50018978e-01 1.03882179e-02 2.15486065e-01
-1.03450370e+00 7.35100389e-01 -1.31187811e-01 7.85599113e-01
-3.84012997e-01 -2.31817320e-01 6.79072380e-01 5.46745718e-01
6.83485031e-01 5.98027706e-01 -1.01667643e+00 -5.69287419e-01
-9.48960900e-01 2.42173269e-01 1.37382317e+00 9.79198992e-01
5.52807271e-01 -1.87220573e-01 -1.76102877e-01 7.89310873e-01
-2.75155324e-02 4.92447108e-01 7.61926472e-01 -5.99812686e-01
8.60544384e-01 6.75532043e-01 1.56119429e-02 -3.53236914e-01
-6.34719253e-01 -4.46595609e-01 -5.50188541e-01 -3.83295894e-01
4.10501212e-01 -1.22501671e-01 -8.81345272e-01 1.42018986e+00
-1.14567213e-01 -1.30405217e-01 4.61109430e-01 3.06407034e-01
9.58035946e-01 8.24915051e-01 1.43661126e-02 1.77599825e-02
1.35797751e+00 -1.30730307e+00 -8.75887871e-01 -2.65668184e-01
9.35373127e-01 -1.08100939e+00 9.70130146e-01 5.06923616e-01
-9.60289240e-01 -6.28965080e-01 -8.18088472e-01 -2.41820142e-01
-1.04049969e+00 4.30496365e-01 4.17129636e-01 6.84579015e-01
-1.03113770e+00 4.90885913e-01 -7.60533929e-01 -6.37400031e-01
-1.38762044e-02 1.45708963e-01 -4.09900725e-01 -9.46903676e-02
-1.39602339e+00 1.13663960e+00 7.30058193e-01 -1.58741876e-01
-6.83614016e-01 -8.41879070e-01 -7.47148812e-01 5.14165312e-03
4.05352235e-01 -2.41226599e-01 1.47685301e+00 -6.71664536e-01
-1.65223134e+00 1.16375637e+00 -6.57974109e-02 -8.79412770e-01
3.45061451e-01 -4.88720059e-01 -8.73450160e-01 -6.94556460e-02
-1.20075308e-01 5.00124037e-01 2.16751501e-01 -6.53357685e-01
-5.16330421e-01 -3.24362189e-01 -3.48255426e-01 -1.12905964e-01
-1.93566203e-01 4.36615378e-01 -3.92199606e-01 -7.24241138e-01
-7.08632767e-01 -1.04210639e+00 9.81214922e-03 -6.60831153e-01
-2.50639826e-01 -4.74891990e-01 7.06117034e-01 -8.58541489e-01
1.46086490e+00 -2.08229947e+00 6.06907830e-02 -4.33717184e-02
-5.13969719e-01 5.37566662e-01 -5.05246103e-01 1.24049067e+00
-2.67250538e-01 3.74989770e-02 -8.12540203e-02 -7.53293574e-01
1.32220984e-01 3.64765972e-01 -6.23264968e-01 2.18518853e-01
1.47896865e-02 1.00557852e+00 -8.52951169e-01 -3.26084614e-01
1.59880847e-01 3.24282527e-01 -3.41462195e-01 3.73161048e-01
-2.73050368e-01 6.89104293e-03 -2.49227628e-01 5.31128883e-01
1.67041853e-01 -1.11109868e-01 -1.77181344e-02 1.11852817e-01
-3.59231472e-01 6.11907959e-01 -7.50829697e-01 2.01983547e+00
-6.45744026e-01 4.69064236e-01 -3.58682156e-01 -6.92825317e-01
1.01923394e+00 6.86843753e-01 2.77308047e-01 -6.95014298e-01
1.96232229e-01 5.39869010e-01 -2.56310374e-01 -4.83988732e-01
8.35019588e-01 -1.22979656e-01 -2.75546879e-01 4.04898822e-01
5.19331872e-01 -1.82413265e-01 6.08048499e-01 1.17407501e-01
1.04318976e+00 5.46276011e-02 5.22911787e-01 -3.55244935e-01
7.94019103e-01 1.83943838e-01 9.65051726e-02 5.30512035e-01
1.33640841e-01 5.23151815e-01 2.99787581e-01 -4.14578289e-01
-1.09272814e+00 -7.09269345e-01 -7.90423527e-02 1.34517920e+00
-5.73776722e-01 -9.13532495e-01 -7.91262627e-01 -8.64770174e-01
-2.91794211e-01 1.29858518e+00 -5.85839689e-01 3.14370126e-01
-4.30350780e-01 -3.09839636e-01 1.12524629e+00 4.52379823e-01
4.17891860e-01 -1.40154314e+00 -4.58256692e-01 4.31541920e-01
-2.37058640e-01 -1.29324019e+00 -4.73186105e-01 3.32181126e-01
-4.43769187e-01 -1.04099905e+00 -5.42183697e-01 -7.97458947e-01
8.14243257e-02 -6.47633016e-01 1.49662423e+00 -3.89707327e-01
-1.37275800e-01 3.69821966e-01 -7.28923321e-01 -6.35328650e-01
-7.03874111e-01 7.41251886e-01 -3.42252135e-01 -2.63359129e-01
5.26638269e-01 -2.32351959e-01 -6.27901927e-02 1.98106736e-01
-9.57597435e-01 -5.12890182e-02 3.77703935e-01 7.13340640e-01
4.03116375e-01 -2.40629360e-01 7.39836693e-01 -1.28901494e+00
6.15848958e-01 -4.13314909e-01 -5.53982258e-01 5.55045784e-01
-7.16212273e-01 2.15073496e-01 8.86444628e-01 -7.46009201e-02
-1.14896643e+00 -1.55643597e-01 -9.22360659e-01 -8.67079869e-02
-2.87409306e-01 7.74900079e-01 -8.27931166e-02 5.01756608e-01
5.35107970e-01 2.06957161e-01 -5.23966134e-01 -8.30751240e-01
5.02565861e-01 1.08861351e+00 5.92544854e-01 -5.24677515e-01
3.19178790e-01 2.47662783e-01 -4.64361310e-01 -7.25140750e-01
-1.30452836e+00 -8.80400300e-01 -6.27258182e-01 7.49942586e-02
9.44007754e-01 -7.62098908e-01 -1.53868124e-01 7.76012838e-01
-1.26359510e+00 -3.99479389e-01 -3.98293346e-01 2.08175033e-01
-3.47175181e-01 3.73359472e-02 -7.72444308e-01 -7.11589694e-01
-8.18678200e-01 -7.97336042e-01 1.20597506e+00 -3.21780354e-01
-4.17543560e-01 -1.25905621e+00 5.45341790e-01 4.31341171e-01
4.47072178e-01 -7.62112997e-03 9.26132619e-01 -1.25308645e+00
9.09329280e-02 -3.44452888e-01 2.90535271e-01 6.64381683e-01
-2.53070772e-01 -1.84507445e-01 -9.43460405e-01 -3.95503521e-01
-1.25944108e-01 -6.53930545e-01 8.06634307e-01 -3.06272447e-01
9.40536737e-01 -2.88645446e-01 -4.36441861e-02 3.30580682e-01
1.33671939e+00 3.10764581e-01 5.81961632e-01 4.98363107e-01
3.08904231e-01 5.69581032e-01 6.40769839e-01 4.09360081e-01
4.58876848e-01 5.72443664e-01 -8.42454806e-02 1.60133347e-01
-3.75453457e-02 -4.53162044e-01 3.89750421e-01 1.13327074e+00
2.92290777e-01 -6.09350324e-01 -1.31385350e+00 6.16256833e-01
-1.80491221e+00 -1.01103055e+00 2.93299947e-02 1.74728417e+00
9.93200004e-01 1.99908108e-01 -1.95922971e-01 1.69663534e-01
3.07658225e-01 2.16933578e-01 -9.32029355e-03 -6.57386482e-01
-2.82645702e-01 5.86326241e-01 4.22970802e-01 5.30109644e-01
-1.05833113e+00 1.10563433e+00 6.81844664e+00 9.92403328e-01
-1.11544013e+00 2.82373279e-01 3.47186118e-01 2.49849763e-02
-1.60188138e-01 2.75838822e-02 -1.30267489e+00 4.30889428e-01
1.60543275e+00 -8.49180669e-02 3.57135653e-01 9.09701347e-01
-9.67101082e-02 4.86817546e-02 -1.20043993e+00 8.69258940e-01
3.32432091e-01 -1.42392707e+00 -4.32010107e-02 -1.69758692e-01
6.53020799e-01 7.02347696e-01 -1.00324504e-01 7.58485198e-01
2.77529895e-01 -1.02951992e+00 7.66726494e-01 5.41607559e-01
6.25811338e-01 -7.25141823e-01 1.04909909e+00 5.99465966e-01
-1.03694677e+00 5.14627248e-02 -6.76293001e-02 6.34658411e-02
2.37926647e-01 5.66097260e-01 -7.03808129e-01 7.13025153e-01
5.39420545e-01 7.30549812e-01 -8.41740310e-01 6.10042214e-01
-4.57510591e-01 1.03139389e+00 -2.70242155e-01 -3.45560201e-02
5.71001589e-01 -4.83728014e-02 3.49667072e-01 1.84329200e+00
1.42405272e-01 1.00082867e-01 2.81243324e-01 3.51322383e-01
-1.74537629e-01 3.87979567e-01 -3.86499196e-01 -4.39531505e-01
3.15141171e-01 1.20634770e+00 -4.06304836e-01 -5.21061480e-01
-4.23038721e-01 9.00027931e-01 4.37102795e-01 2.02627838e-01
-6.56738460e-01 -6.03195786e-01 9.50973704e-02 -1.74016520e-01
4.69742030e-01 -8.34702551e-02 3.70274112e-02 -1.34648001e+00
-1.37321800e-01 -1.05990660e+00 8.23099971e-01 -7.65593827e-01
-1.37525415e+00 1.01628113e+00 9.61353909e-03 -7.49799430e-01
-5.74676514e-01 -8.17109108e-01 -2.60451108e-01 7.77986825e-01
-1.57980466e+00 -1.17267299e+00 1.18457593e-01 5.48887730e-01
5.99022985e-01 -4.51200038e-01 1.22314954e+00 5.58452368e-01
-4.58204538e-01 5.84668815e-01 3.79500538e-01 2.27186501e-01
8.10428023e-01 -1.10368061e+00 4.95288849e-01 5.49040437e-01
4.21034127e-01 6.07926309e-01 5.13011575e-01 -3.35477889e-01
-1.25150323e+00 -1.08811665e+00 1.78991091e+00 -6.02132916e-01
9.45667803e-01 -5.34966767e-01 -7.77771950e-01 9.36029255e-01
4.71686125e-01 -1.66459024e-01 9.18865085e-01 1.10295758e-01
-5.56840181e-01 1.09259099e-01 -1.01053941e+00 1.67346653e-02
6.99435532e-01 -1.00405443e+00 -1.09621036e+00 2.90764719e-01
6.24459922e-01 -6.40770555e-01 -1.03146064e+00 2.84069121e-01
5.24477303e-01 -8.99431229e-01 5.48650920e-01 -7.67707586e-01
4.63776946e-01 5.81340045e-02 -2.96486914e-01 -1.35227287e+00
1.47109881e-01 -5.35270214e-01 1.13064582e-02 1.59221685e+00
8.22473109e-01 -5.19293845e-01 4.12937969e-01 1.98805660e-01
-2.89964825e-01 -8.58538210e-01 -1.00012052e+00 -8.57416987e-01
1.72716618e-01 -9.23116386e-01 4.80428964e-01 8.74795616e-01
-1.85882533e-03 8.06371212e-01 -2.83279210e-01 -3.15213680e-01
1.49423465e-01 2.01908261e-01 5.70568621e-01 -8.03416789e-01
-4.67118621e-01 -1.59958482e-01 -8.66991207e-02 -7.71924198e-01
5.26903152e-01 -1.20141339e+00 -9.75102857e-02 -1.41994035e+00
-1.78151168e-02 -1.42979383e-01 -2.52494961e-01 8.97660196e-01
1.37856245e-01 2.48000864e-02 3.34002674e-01 -2.32547149e-02
-6.19808614e-01 5.07789850e-01 4.93419796e-01 -3.76678258e-02
-3.85267735e-02 -1.57268196e-01 -3.87423426e-01 5.82049668e-01
7.36951113e-01 -5.96451879e-01 -2.34183967e-01 -4.63116169e-01
5.93310118e-01 -4.03775163e-02 -3.70351262e-02 -9.18891132e-01
2.02740747e-02 3.06687236e-01 6.78999647e-02 -1.02272117e+00
2.50274509e-01 -6.48189783e-01 7.08184764e-03 3.78908306e-01
-6.20862007e-01 2.88951457e-01 3.62588763e-01 8.50803684e-03
-4.99041915e-01 -3.41249138e-01 7.36139476e-01 -6.73986152e-02
-8.65768433e-01 2.62201801e-02 -4.78439808e-01 4.10076946e-01
8.02243173e-01 4.40715015e-01 -5.34463286e-01 -2.57848591e-01
-6.17640138e-01 1.78443626e-01 3.21644656e-02 7.77721524e-01
3.17083210e-01 -9.82049227e-01 -8.04112911e-01 6.53570890e-02
3.89335513e-01 -6.50680304e-01 -8.79298374e-02 5.12191713e-01
-3.80089104e-01 1.04277122e+00 1.03451014e-02 -1.70790255e-01
-1.31781566e+00 6.54178739e-01 2.28900477e-01 -1.00930417e+00
-1.34574831e-01 6.69230521e-01 -3.67890656e-01 -9.83986437e-01
2.31508091e-01 -4.62444901e-01 -3.02840352e-01 2.46432006e-01
5.68902314e-01 4.73474979e-01 4.86321479e-01 -5.82915306e-01
-4.52530295e-01 2.69897312e-01 -6.97072670e-02 -2.40611866e-01
1.40517235e+00 9.31286514e-02 -2.01108396e-01 7.45314598e-01
1.58934641e+00 -7.20202550e-02 -3.58642310e-01 -1.91822037e-01
4.32582408e-01 1.52665854e-01 1.11626357e-01 -1.21557677e+00
-6.81176424e-01 7.99725831e-01 4.39510852e-01 3.92507426e-02
9.90267754e-01 1.12284891e-01 1.03381002e+00 9.10954475e-01
4.47067261e-01 -1.12904143e+00 -2.01648220e-01 1.29617822e+00
8.09299350e-01 -9.20794666e-01 -1.17809527e-01 -1.51948050e-01
-8.02021086e-01 1.10613573e+00 2.12218016e-01 6.79706037e-02
7.61226356e-01 4.94895995e-01 1.86892942e-01 -1.45645574e-01
-1.11901772e+00 -2.46639058e-01 3.70432943e-01 7.15875253e-02
7.14017332e-01 -1.17569104e-01 -5.61693907e-01 7.88714826e-01
-5.99020541e-01 3.54409456e-01 1.84475020e-01 1.07517660e+00
-9.55983177e-02 -1.49923718e+00 -7.91109633e-03 8.79421681e-02
-8.58809292e-01 -4.81260240e-01 -6.78711712e-01 8.64698827e-01
4.66100201e-02 9.99326766e-01 -1.04952417e-01 -2.15281606e-01
6.79830372e-01 4.87264901e-01 2.53595620e-01 -6.74090385e-01
-1.22757709e+00 -3.16976219e-01 7.06928253e-01 -6.51827037e-01
-3.74543101e-01 -6.09712839e-01 -1.23983705e+00 -1.17558040e-01
-8.73292834e-02 5.80125570e-01 8.57717216e-01 1.12474191e+00
3.63465816e-01 2.38657191e-01 3.18656445e-01 -3.58895361e-01
-6.98493421e-01 -1.20058215e+00 -5.17560780e-01 4.22756225e-01
8.54745880e-02 -6.56567048e-03 -2.03738213e-01 1.82892561e-01] | [10.32907485961914, 9.526021957397461] |
1f5727f0-13c1-45db-a15c-2be39886d02d | can-pre-trained-vision-and-language-models | 2302.11713 | null | https://arxiv.org/abs/2302.11713v2 | https://arxiv.org/pdf/2302.11713v2.pdf | Can Pre-trained Vision and Language Models Answer Visual Information-Seeking Questions? | Large language models have demonstrated an emergent capability in answering knowledge intensive questions. With recent progress on web-scale visual and language pre-training, do these models also understand how to answer visual information seeking questions? To answer this question, we present InfoSeek, a Visual Question Answering dataset that focuses on asking information-seeking questions, where the information can not be answered by common sense knowledge. We perform a multi-stage human annotation to collect a natural distribution of high-quality visual information seeking question-answer pairs. We also construct a large-scale, automatically collected dataset by combining existing visual entity recognition datasets and Wikidata, which provides over one million examples for model fine-tuning and validation. Based on InfoSeek, we analyzed various pre-trained Visual QA systems to gain insights into the characteristics of different pre-trained models. Our analysis shows that it is challenging for the state-of-the-art multi-modal pre-trained models to answer visual information seeking questions, but this capability is improved through fine-tuning on the automated InfoSeek dataset. We hope our analysis paves the way to understand and develop the next generation of multi-modal pre-training. | ['Ming-Wei Chang', 'Alan Ritter', 'Soravit Changpinyo', 'Haitian Sun', 'Yi Luan', 'Hexiang Hu', 'Yang Chen'] | 2023-02-23 | null | null | null | null | ['common-sense-reasoning'] | ['reasoning'] | [ 5.64863123e-02 4.02012169e-01 -1.45226941e-01 -2.43262857e-01
-1.04783618e+00 -1.19087946e+00 7.22785354e-01 2.92629540e-01
-3.38545829e-01 2.40935564e-01 3.61736655e-01 -8.61640215e-01
-1.18045621e-01 -4.01522458e-01 -6.58783555e-01 7.01075187e-03
3.71807277e-01 9.06364202e-01 5.53393364e-01 -3.99902701e-01
4.30102795e-01 2.59022534e-01 -1.61934173e+00 1.10325217e+00
8.03941071e-01 8.49245012e-01 2.97592849e-01 9.55081165e-01
-7.61366427e-01 1.27176392e+00 -6.02882743e-01 -5.07414639e-01
-2.78470758e-02 -3.88800621e-01 -1.56491542e+00 -8.76296684e-02
1.29208028e+00 -2.01855496e-01 -1.88804150e-01 6.29337072e-01
2.91426629e-01 5.37573956e-02 7.35376358e-01 -1.42097843e+00
-1.50937545e+00 7.97094181e-02 -2.24991411e-01 5.24348855e-01
6.92262411e-01 5.34464598e-01 1.35020268e+00 -9.75713491e-01
1.15328443e+00 1.41395450e+00 3.03737015e-01 6.53198600e-01
-1.04152930e+00 -2.91132152e-01 -3.85678373e-02 6.72385931e-01
-1.04054689e+00 -4.13466424e-01 6.68183625e-01 -6.97406530e-01
9.99642909e-01 3.80874366e-01 5.47044873e-01 9.36450064e-01
-3.09790552e-01 1.28097761e+00 1.31870031e+00 -7.69861341e-01
-5.07754181e-03 5.20016730e-01 3.01600516e-01 1.11092699e+00
-5.35196625e-02 -1.39665291e-01 -6.71647370e-01 -2.20294334e-02
4.82115328e-01 -3.58672887e-01 -3.84682029e-01 -6.10131979e-01
-1.31169045e+00 9.06528831e-01 9.79197502e-01 3.81996334e-01
-3.85985039e-02 8.12653638e-03 1.40589193e-01 4.21074092e-01
-1.63844213e-01 8.41946304e-01 -3.67275119e-01 1.45092458e-01
-7.06445694e-01 1.48750216e-01 8.91301453e-01 9.29778159e-01
1.10868275e+00 -3.56840193e-01 -7.21409440e-01 5.16527116e-01
5.27516544e-01 5.58137834e-01 2.33245373e-01 -1.15228832e+00
6.98902905e-01 1.23698473e+00 2.07512885e-01 -1.01446092e+00
-3.52804005e-01 3.90068106e-02 -2.74772316e-01 3.57919306e-01
9.58650649e-01 3.11351925e-01 -1.30033767e+00 1.16907775e+00
2.58531213e-01 -5.95338225e-01 3.99418548e-02 9.29837942e-01
1.51066995e+00 6.12949729e-01 4.65125769e-01 3.73460889e-01
1.80436587e+00 -1.16400743e+00 -6.03407562e-01 -5.35057127e-01
7.10534394e-01 -8.04264009e-01 1.94088507e+00 -6.86725527e-02
-8.35230410e-01 -7.48841465e-01 -6.10407233e-01 -5.97772062e-01
-9.47287381e-01 9.71616581e-02 5.55990219e-01 5.30656457e-01
-1.07106256e+00 -4.68512267e-01 -1.35508925e-01 -6.97534621e-01
6.69412971e-01 -3.42672855e-01 -5.07591963e-01 -3.26362282e-01
-9.79492426e-01 1.18304825e+00 1.46842718e-01 -4.17060852e-01
-1.11061120e+00 -9.98630285e-01 -8.26681495e-01 2.95818578e-02
6.88915253e-01 -8.23228419e-01 1.44244969e+00 -9.53115165e-01
-6.42091513e-01 1.42405617e+00 -3.34001988e-01 8.36526006e-02
6.97809532e-02 1.46677643e-01 -3.25972021e-01 7.31687307e-01
3.49708229e-01 9.78666544e-01 8.35085452e-01 -1.84361970e+00
-5.02723694e-01 -5.63229859e-01 5.81603229e-01 1.32039130e-01
-2.81705916e-01 -7.76188523e-02 -7.93704748e-01 -1.56702563e-01
-5.37445605e-01 -6.14195943e-01 1.53590530e-01 4.12276000e-01
-2.36683875e-01 -3.86158317e-01 8.96496713e-01 -8.94898295e-01
1.02730262e+00 -1.84016275e+00 2.59705111e-02 5.74390665e-02
5.86592913e-01 3.57656389e-01 -4.77688581e-01 5.52860022e-01
2.78634131e-01 1.83162719e-01 1.19030531e-02 2.36086007e-02
1.85153812e-01 1.23759024e-01 -2.84967363e-01 -1.49890751e-01
1.58435881e-01 1.58314896e+00 -9.17147636e-01 -9.94439662e-01
6.00859337e-02 3.64411056e-01 -3.79049182e-01 4.41103637e-01
-7.23091066e-01 8.65756422e-02 -4.64879125e-01 9.07935321e-01
2.75729358e-01 -1.11333776e+00 -6.92331493e-02 -4.53372449e-01
1.36012971e-01 -5.39776720e-02 -8.47874463e-01 1.72357428e+00
-3.84914905e-01 1.12155521e+00 1.14018790e-01 -6.78042471e-01
6.60162985e-01 -5.10194153e-02 -1.26698032e-01 -1.58565855e+00
-8.35804045e-02 -1.51811659e-01 -4.13190693e-01 -1.05951858e+00
7.14200854e-01 1.26507968e-01 -5.07377945e-02 6.12190545e-01
3.18402320e-01 -1.39714181e-01 5.68706632e-01 9.91944492e-01
9.34882700e-01 -1.08271331e-01 1.14847109e-01 -1.02621846e-01
4.16532964e-01 9.04944241e-01 -3.10996175e-01 9.83626187e-01
-3.82601857e-01 4.07585680e-01 2.95715988e-01 -4.29959923e-01
-9.08846200e-01 -1.15991557e+00 2.11931646e-01 1.67420340e+00
1.87112942e-01 -4.27988380e-01 -6.25256300e-01 -1.07706690e+00
2.10124567e-01 7.26082981e-01 -9.03767884e-01 2.57632881e-01
-1.43648237e-01 -1.22875996e-01 5.96677005e-01 4.85931426e-01
3.68197918e-01 -1.48837328e+00 -7.88535714e-01 -2.90342778e-01
-3.02105874e-01 -1.13180387e+00 -2.71521926e-01 -1.62218496e-01
-2.17565596e-01 -1.64408767e+00 -9.39996839e-01 -1.10010386e+00
5.66022098e-01 4.29491222e-01 1.89752233e+00 4.39335406e-01
-5.58418810e-01 1.29920399e+00 -3.98559421e-01 -4.45955575e-01
-4.77930635e-01 7.77222961e-02 -7.46744931e-01 -2.59980142e-01
7.29867816e-01 1.87786922e-01 -7.05501854e-01 3.04246128e-01
-1.12544358e+00 3.52922790e-02 6.48787558e-01 4.69067901e-01
4.29831058e-01 -6.23643458e-01 4.36901659e-01 -8.31784964e-01
1.04494250e+00 -4.15101647e-01 -4.27771896e-01 1.21058166e+00
-4.85224843e-01 3.37775737e-01 1.41869500e-01 -2.69528389e-01
-1.21662605e+00 -1.49417251e-01 2.06055120e-01 -2.86641538e-01
-3.18481892e-01 4.80263054e-01 1.56261459e-01 -2.53333122e-01
1.14967167e+00 8.15588161e-02 -2.33461529e-01 -4.96661961e-01
1.21948373e+00 6.04475558e-01 7.01261401e-01 -4.29169118e-01
8.08908224e-01 5.25903165e-01 -4.24809456e-01 -8.82291615e-01
-1.15485740e+00 -8.83670568e-01 -5.10029554e-01 -5.68384051e-01
1.31306005e+00 -6.96596682e-01 -1.03012168e+00 -1.16711922e-01
-1.06157100e+00 -5.42572260e-01 -2.65039444e-01 -4.19147044e-01
-3.00327808e-01 5.02574921e-01 -2.42212862e-01 -7.32234716e-01
-4.00941044e-01 -8.04096758e-01 1.03976083e+00 4.93022084e-01
-1.52589932e-01 -1.08560920e+00 4.44000572e-01 1.21849263e+00
5.59445560e-01 -2.08925217e-01 1.24547553e+00 -5.84894180e-01
-8.63984823e-01 2.79416293e-01 -1.00703287e+00 -1.94656089e-01
-3.02986473e-01 -5.66469543e-02 -8.80242169e-01 -9.62117538e-02
-6.59140289e-01 -1.08488512e+00 8.88086438e-01 -1.96508393e-01
9.15287256e-01 -2.08315596e-01 -3.30316484e-01 1.33996531e-01
1.35290921e+00 -9.11251754e-02 6.29342496e-01 5.18846869e-01
1.03310013e+00 7.86344230e-01 4.45216745e-01 -1.55977085e-01
1.05367482e+00 5.70456684e-01 6.17004395e-01 -3.27121735e-01
-5.74513018e-01 -5.62362790e-01 -2.04172403e-01 3.08354765e-01
2.56040305e-01 -1.80236146e-01 -1.47122061e+00 1.00354683e+00
-1.66556323e+00 -1.09890902e+00 -2.71163940e-01 1.54799902e+00
8.80843341e-01 -4.28817391e-01 1.75688326e-01 -3.41574818e-01
1.11608304e-01 8.76261517e-02 -5.06054223e-01 -1.99002504e-01
-1.30738243e-01 1.47040144e-01 1.19460583e-01 5.72917283e-01
-8.76089394e-01 1.11642885e+00 6.89152765e+00 7.66253352e-01
-7.13029861e-01 5.25151417e-02 3.32951933e-01 3.56934816e-01
-8.85224700e-01 1.85533375e-01 -7.42067456e-01 -1.54937571e-02
6.87040091e-01 1.63494870e-01 3.85166466e-01 8.78040969e-01
-3.59798312e-01 -2.55119830e-01 -8.86633515e-01 1.14787710e+00
6.28519356e-01 -1.86317647e+00 5.73172390e-01 -2.67345101e-01
6.65804803e-01 -3.58463936e-02 -7.37496987e-02 6.72768295e-01
4.31659877e-01 -1.30076838e+00 4.88314688e-01 8.29973578e-01
7.66213179e-01 -1.86147407e-01 1.75332785e-01 2.78763443e-01
-1.13051295e+00 -1.59933195e-01 -1.01420753e-01 1.85225904e-01
3.33062291e-01 6.32238463e-02 -1.02281225e+00 2.60853678e-01
1.01734805e+00 3.63237590e-01 -1.75959718e+00 1.03003907e+00
-2.36471176e-01 2.05854386e-01 4.83286604e-02 -3.23572367e-01
2.72194922e-01 3.50432545e-01 1.55198202e-01 1.17019033e+00
-1.78861588e-01 -7.68731683e-02 1.05829857e-01 9.55160379e-01
-1.82194099e-01 1.25531107e-01 -4.68655407e-01 -4.70076114e-01
1.69268429e-01 1.44766212e+00 -5.99613369e-01 -5.67876935e-01
-6.11979067e-01 8.53907943e-01 7.80579507e-01 6.57997787e-01
-3.52141052e-01 -4.20875192e-01 1.49904788e-01 3.58830333e-01
1.31322831e-01 -1.52843341e-01 -1.70880228e-01 -1.31541026e+00
-1.74985796e-01 -1.33316958e+00 9.67882276e-01 -1.62364268e+00
-1.37815583e+00 5.25191844e-01 -3.43051367e-02 -5.93372047e-01
-2.19508231e-01 -1.08125150e+00 -1.80982396e-01 5.98233044e-01
-1.69417417e+00 -1.68022537e+00 -9.08413768e-01 1.05247283e+00
4.37693059e-01 -1.62353575e-01 7.01472878e-01 9.33948532e-02
-1.13399187e-02 4.27197278e-01 -3.78522873e-01 2.73110479e-01
8.72190475e-01 -1.30401230e+00 1.42014816e-01 5.63475788e-01
7.60198712e-01 4.51481819e-01 4.97412384e-01 -5.31859815e-01
-1.55574751e+00 -6.36908591e-01 9.27079499e-01 -1.46355939e+00
7.56521106e-01 -3.27724785e-01 -1.25742006e+00 7.82692313e-01
7.87266672e-01 -2.31256887e-01 7.66007721e-01 2.44425014e-01
-9.14664686e-01 2.75290161e-01 -9.21630502e-01 5.31195045e-01
8.48091364e-01 -1.04018569e+00 -1.33570504e+00 3.81440371e-01
5.92919767e-01 -6.53311461e-02 -6.10041797e-01 2.52155364e-01
3.67347270e-01 -8.28256965e-01 1.23772538e+00 -1.02158523e+00
4.14710879e-01 -4.14488614e-01 -8.20250660e-02 -8.90949428e-01
-2.15169817e-01 -2.40991026e-01 -1.97554722e-01 1.11365128e+00
6.78291678e-01 3.69157419e-02 8.04826856e-01 6.81594849e-01
1.91568762e-01 -2.46549994e-01 -5.97157896e-01 -2.85352588e-01
-1.14207000e-01 -2.79937118e-01 1.48799807e-01 1.10068703e+00
1.57294780e-01 9.30937648e-01 -8.07916373e-02 -4.89361733e-02
4.90795463e-01 3.63489419e-01 1.19583893e+00 -1.17431819e+00
-3.44903767e-01 -2.58985370e-01 6.05490878e-02 -1.28416121e+00
-4.53222245e-02 -9.50778127e-01 -2.13420585e-01 -2.36500335e+00
5.18505454e-01 -9.51685682e-02 4.13928032e-02 8.44182968e-01
-4.18937594e-01 4.40939426e-01 3.69330406e-01 2.14970186e-01
-1.40202129e+00 2.71498352e-01 1.63017249e+00 -4.42742079e-01
-1.50756881e-01 -7.10987270e-01 -9.25894558e-01 5.00490963e-01
4.68671054e-01 -8.18275288e-02 -5.27526498e-01 -6.69779658e-01
7.53264606e-01 -8.79933834e-02 6.77593291e-01 -6.86835408e-01
3.39991510e-01 -3.13984871e-01 5.30935228e-01 -8.07942212e-01
3.13647687e-01 -6.83425546e-01 -3.92250478e-01 1.56212807e-01
-5.72307348e-01 1.95822909e-01 3.88497442e-01 5.39311588e-01
-2.08822027e-01 -1.14210568e-01 5.46523929e-01 -4.76464719e-01
-1.52222192e+00 -3.27463187e-02 -4.14701968e-01 7.60857761e-01
8.46535921e-01 1.32058673e-02 -1.10309088e+00 -6.89394116e-01
-6.00949109e-01 6.06766701e-01 4.66520369e-01 7.38082349e-01
6.71284318e-01 -1.24162519e+00 -5.18989861e-01 -1.53462738e-01
1.01178277e+00 -4.71318424e-01 4.27973568e-01 2.94009387e-01
-5.78559160e-01 7.53453553e-01 -3.54261547e-01 -5.80927849e-01
-1.35249388e+00 9.36879218e-01 1.83247283e-01 -2.84051329e-01
-1.68125704e-01 8.29529703e-01 1.97769582e-01 -6.62358522e-01
1.58550709e-01 -4.12767567e-02 -6.14729702e-01 3.57607186e-01
6.87338173e-01 1.30486473e-01 -1.91814885e-01 -7.40463257e-01
-4.80367303e-01 6.27783537e-01 1.14202239e-01 -1.72079697e-01
8.52297902e-01 -3.03588808e-01 7.67000094e-02 2.41039410e-01
1.10736322e+00 2.42530461e-02 -7.58082330e-01 -2.09070146e-01
8.81967396e-02 -4.16914433e-01 -2.09988758e-01 -1.30037510e+00
-7.56785750e-01 1.14761722e+00 4.65843618e-01 4.51402724e-01
8.76025617e-01 7.65752196e-01 3.44944328e-01 7.99628973e-01
1.36241354e-02 -1.04780877e+00 8.23179901e-01 5.91298699e-01
1.11993265e+00 -1.73656034e+00 -2.25254089e-01 -1.19228706e-01
-9.08621669e-01 8.11155260e-01 9.90653634e-01 3.71841818e-01
3.33775252e-01 -4.24915135e-01 7.78677404e-01 -9.47861969e-01
-6.94669545e-01 -7.32840002e-01 1.06811738e+00 8.88094068e-01
2.58197904e-01 -2.57164985e-01 3.42603058e-01 1.28519818e-01
-6.79396093e-02 -1.35498717e-01 1.65919498e-01 7.61148453e-01
-6.87546909e-01 -7.57164955e-01 -4.12413329e-01 3.19372535e-01
-1.70571413e-02 -2.29118258e-01 -9.39341605e-01 1.12452471e+00
-3.36062253e-01 1.20959854e+00 -8.95192251e-02 7.07849190e-02
4.18594807e-01 2.83666790e-01 4.57037151e-01 -5.14989078e-01
-5.20064592e-01 -6.85876846e-01 2.57080615e-01 -6.56993449e-01
-4.12687689e-01 3.39372754e-02 -1.06049430e+00 6.01151995e-02
1.97886810e-01 1.37647286e-01 5.47157288e-01 9.01708484e-01
6.19957864e-01 3.35700184e-01 -3.98701549e-01 -4.64474946e-01
-2.50806343e-02 -6.79847181e-01 6.15219735e-02 1.00926459e+00
4.30903107e-01 -3.96904469e-01 -3.89562011e-01 1.97996795e-01] | [10.945703506469727, 1.7703304290771484] |
24b492fb-cdc7-40e0-955e-f4420d2f36bf | meev-body-mesh-estimation-on-egocentric-video | 2210.14165 | null | https://arxiv.org/abs/2210.14165v1 | https://arxiv.org/pdf/2210.14165v1.pdf | MEEV: Body Mesh Estimation On Egocentric Video | This technical report introduces our solution, MEEV, proposed to the EgoBody Challenge at ECCV 2022. Captured from head-mounted devices, the dataset consists of human body shape and motion of interacting people. The EgoBody dataset has challenges such as occluded body or blurry image. In order to overcome the challenges, MEEV is designed to exploit multiscale features for rich spatial information. Besides, to overcome the limited size of dataset, the model is pre-trained with the dataset aggregated 2D and 3D pose estimation datasets. Achieving 82.30 for MPJPE and 92.93 for MPVPE, MEEV has won the EgoBody Challenge at ECCV 2022, which shows the effectiveness of the proposed method. The code is available at https://github.com/clovaai/meev | ['Dongyoon Wee', 'Nicolas Monet'] | 2022-10-21 | null | null | null | null | ['3d-pose-estimation', '3d-human-pose-estimation', '3d-human-pose-and-shape-estimation'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [-5.38642645e-01 9.86501947e-02 6.44430965e-02 -1.85577959e-01
-4.44116980e-01 -4.05903697e-01 4.66509789e-01 -7.30938494e-01
-2.38471285e-01 5.20921588e-01 7.25193620e-01 5.60353935e-01
3.53886813e-01 -3.17023426e-01 -3.57636362e-01 -4.06440884e-01
1.23027295e-01 2.54731804e-01 1.09995350e-01 -1.03384823e-01
-2.21307218e-01 1.24303317e-02 -1.40458798e+00 -3.26501094e-02
6.18499935e-01 1.05102909e+00 -1.62579551e-01 7.64251173e-01
6.30478919e-01 1.76495537e-01 -2.51624972e-01 -7.20807433e-01
5.30697048e-01 -5.32430857e-02 -5.12572050e-01 1.90341339e-01
1.19808674e+00 -6.50204718e-01 -1.01631832e+00 8.36509407e-01
1.22072709e+00 2.09057689e-01 4.74859029e-01 -1.52442348e+00
-1.82995886e-01 -1.69882566e-01 -6.68355703e-01 2.36358777e-01
7.38780081e-01 3.00213873e-01 5.08859038e-01 -1.25679588e+00
8.24063063e-01 1.44905734e+00 7.90467918e-01 9.06443954e-01
-7.98156917e-01 -5.89610100e-01 -1.81239039e-01 3.00421715e-01
-1.66369390e+00 -7.71819890e-01 6.79843426e-01 -5.93790412e-01
6.40227079e-01 2.02688426e-01 8.51667702e-01 1.24691427e+00
1.29700556e-01 1.06547427e+00 8.34836900e-01 8.11793059e-02
-8.84838700e-02 -8.22422653e-02 2.74112046e-01 6.53514087e-01
4.56605822e-01 7.49712735e-02 -8.09425950e-01 -2.18873516e-01
6.68822169e-01 1.74327623e-02 -5.60066044e-01 -6.88015640e-01
-1.13772142e+00 2.78789669e-01 4.19511884e-01 -1.30232975e-01
-2.55359709e-01 4.22828877e-03 3.80462021e-01 -1.18793689e-01
2.44814992e-01 -3.03094685e-01 -2.16862336e-01 -3.20183516e-01
-7.17986763e-01 5.71295440e-01 5.66345811e-01 1.28553641e+00
1.16855711e-01 -1.90902486e-01 -4.52396810e-01 6.83419287e-01
4.99657661e-01 9.90832806e-01 9.48396400e-02 -1.09904647e+00
8.09964716e-01 4.48644072e-01 5.26897728e-01 -1.23512673e+00
-7.05561519e-01 -1.81661755e-01 -9.24850583e-01 -1.57283023e-01
5.33435464e-01 -3.07645708e-01 -7.54509985e-01 1.73423183e+00
8.20446312e-01 -2.03040359e-03 -1.42044097e-01 1.37522149e+00
1.39811504e+00 4.11181599e-01 1.35289850e-02 2.15256646e-01
1.62083387e+00 -1.10043001e+00 -9.73507404e-01 -3.25818211e-01
4.04757529e-01 -5.04264534e-01 7.52622187e-01 3.93664479e-01
-1.20153189e+00 -7.49212563e-01 -8.15300763e-01 -2.87487715e-01
-5.01844799e-03 4.89874154e-01 3.30627203e-01 9.03727889e-01
-1.10609758e+00 1.78575888e-01 -6.80340827e-01 -6.46166980e-01
3.10106844e-01 5.11287332e-01 -7.97994137e-01 -1.71550125e-01
-1.11028671e+00 6.98942363e-01 -4.12389217e-03 2.66509861e-01
-5.08702457e-01 -3.51572096e-01 -1.02704632e+00 -2.53484190e-01
1.51762679e-01 -1.13268018e+00 1.02258062e+00 -9.26163122e-02
-1.40132129e+00 1.07629967e+00 -3.30534399e-01 -4.69365954e-01
1.05758786e+00 -8.64426672e-01 -3.89296979e-01 2.30422348e-01
-1.72922954e-01 7.61327267e-01 7.86019862e-01 -7.33557701e-01
-3.40351671e-01 -8.74448001e-01 -2.14782357e-01 5.17010093e-01
-2.84882039e-01 -1.89131107e-02 -1.12531698e+00 -4.38626766e-01
1.89387441e-01 -1.31165087e+00 2.20035642e-01 1.92090012e-02
-4.21307355e-01 -1.31638572e-01 1.09997964e+00 -9.72537458e-01
9.67411458e-01 -2.15001798e+00 3.28509837e-01 -1.05739981e-01
6.51610970e-01 2.92421609e-01 9.85318571e-02 6.63846657e-02
2.76022971e-01 -2.57060707e-01 3.02204698e-01 -8.72672737e-01
2.10061789e-01 -2.73343861e-01 3.17775384e-02 1.04913068e+00
-4.12024200e-01 1.21495891e+00 -4.42833483e-01 -6.13512993e-01
5.70819080e-01 7.86308348e-01 -4.42572504e-01 1.16308130e-01
5.74847221e-01 7.25286782e-01 -3.45060080e-01 6.45775855e-01
1.22310519e+00 -1.54282048e-01 -1.34959333e-02 -5.48840046e-01
2.37125829e-01 -1.21202663e-01 -1.22778118e+00 1.77868629e+00
1.53442547e-02 5.98572969e-01 2.36527845e-01 -2.63375878e-01
6.67528868e-01 4.20503527e-01 5.84012032e-01 -5.46743393e-01
4.84564662e-01 -2.11617872e-01 -3.95962864e-01 -4.99538660e-01
5.22749364e-01 3.67569119e-01 -2.36108318e-01 -1.91087782e-01
1.10591531e-01 -9.46336687e-02 -2.25348510e-02 1.69542462e-01
9.66788054e-01 3.89415711e-01 2.40447700e-01 -7.45014846e-02
6.15450740e-01 -2.83100188e-01 7.19572961e-01 5.82011044e-01
-9.07084167e-01 9.76429224e-01 -2.09889159e-01 -7.08833396e-01
-9.10947204e-01 -1.29681110e+00 -1.11853264e-01 7.19401777e-01
5.44983149e-01 -6.35766089e-01 -9.46674228e-01 -5.37563622e-01
1.17317699e-02 3.93170528e-02 -5.29415011e-01 -1.65628307e-02
-5.93458652e-01 -4.36105072e-01 5.40864408e-01 5.53914428e-01
1.14117122e+00 -5.96687615e-01 -6.16612852e-01 -1.80325389e-01
-7.67902792e-01 -1.53161824e+00 -9.36253369e-01 -1.02531064e+00
-7.66664505e-01 -1.09122574e+00 -1.19037366e+00 -2.68758297e-01
3.70719939e-01 4.10152286e-01 1.01058459e+00 -2.31527224e-01
-5.15900135e-01 8.28732252e-01 -1.63290247e-01 -1.97568685e-01
5.20676672e-01 -6.84266090e-02 6.04306936e-01 -3.74768488e-02
4.77899909e-01 -4.09866065e-01 -1.07144630e+00 6.30198538e-01
6.16254248e-02 6.47657737e-02 1.31646559e-01 5.76225162e-01
3.43506217e-01 -4.15207863e-01 -1.34939179e-01 -8.70273933e-02
4.88466099e-02 -1.71875432e-01 -4.56654340e-01 3.00909206e-03
-1.97862796e-02 -6.31040156e-01 -1.72656551e-01 -3.12728703e-01
-9.66271222e-01 2.81360507e-01 -4.01981622e-02 -5.73717356e-01
-1.91230640e-01 -3.89095873e-01 -5.01204967e-01 -1.58895805e-01
4.73094195e-01 1.41262129e-01 -1.39332563e-01 -4.98755544e-01
2.21965313e-01 7.87255108e-01 1.02504873e+00 -3.32321405e-01
7.33475983e-01 6.92178309e-01 7.14519471e-02 -9.19515193e-01
-5.87272048e-01 -6.58846736e-01 -8.39930356e-01 -4.70537305e-01
9.72044826e-01 -1.56441140e+00 -9.07775342e-01 7.87636459e-01
-1.04059291e+00 -1.19595304e-01 5.31172827e-02 6.84753180e-01
-6.11728668e-01 6.23982906e-01 -6.20710373e-01 -8.99509668e-01
-8.35729837e-01 -7.71375477e-01 1.31508672e+00 4.19510156e-01
-5.76581359e-01 -6.70479298e-01 2.68734861e-02 9.41743910e-01
1.52527928e-01 4.19023246e-01 -2.97206670e-01 -2.84855366e-01
-3.90141040e-01 -6.21715188e-01 -1.23373695e-01 -2.42815726e-02
1.62709672e-02 -4.60472792e-01 -1.07144165e+00 -7.49301672e-01
-7.06909671e-02 -3.45556527e-01 5.78312516e-01 6.30050004e-01
6.37074411e-01 -1.02253370e-01 -4.21613872e-01 7.68409669e-01
8.69347930e-01 -3.53568405e-01 7.81478882e-01 7.98241869e-02
8.86122942e-01 5.39012909e-01 6.67745829e-01 7.24556804e-01
8.85469496e-01 1.27598977e+00 2.79971004e-01 1.38874933e-01
-4.40022767e-01 -2.85800874e-01 4.06839311e-01 4.83518571e-01
-4.47278798e-01 -2.85028666e-01 -9.06226695e-01 4.88473743e-01
-1.89762890e+00 -9.61253226e-01 -4.95435804e-01 2.25391054e+00
1.21735953e-01 -2.77953088e-01 5.24720311e-01 -1.92148328e-01
7.87171483e-01 1.58739343e-01 -4.06452596e-01 2.74664164e-01
-2.05864444e-01 -4.57214803e-01 3.42286199e-01 3.41329157e-01
-1.46507919e+00 9.06431675e-01 5.88587523e+00 4.77621496e-01
-7.29986668e-01 3.13113332e-01 1.72916159e-01 -3.96173239e-01
4.65261191e-01 -5.32967329e-01 -1.11726844e+00 6.29513264e-01
5.72615981e-01 1.16694244e-02 5.43007925e-02 8.39018166e-01
1.49691328e-01 -5.80384843e-02 -8.13382924e-01 1.74084258e+00
3.15760642e-01 -8.45364571e-01 -4.01412100e-01 3.33761483e-01
6.17828786e-01 1.99184164e-01 1.49791166e-01 2.48514667e-01
-1.57670379e-01 -9.12974894e-01 6.30250633e-01 5.20214081e-01
7.02168405e-01 -4.68673527e-01 9.52174783e-01 2.42595941e-01
-1.52047658e+00 6.88072667e-02 -4.46653634e-01 -8.70901048e-02
3.73294026e-01 3.68380606e-01 -3.72266471e-01 5.34295142e-01
1.07900822e+00 7.27919996e-01 -6.64920628e-01 1.13517749e+00
-1.90595284e-01 2.27130234e-01 -5.91996849e-01 3.70598704e-01
-4.64365900e-01 7.41738230e-02 8.63653421e-01 1.06331325e+00
4.23723489e-01 3.51124585e-01 2.08565801e-01 3.56083602e-01
-9.87916514e-02 -3.98798883e-02 -7.51356006e-01 4.82179314e-01
3.02193582e-01 1.30193555e+00 -1.45334572e-01 -3.24123591e-01
-3.02743554e-01 1.40804660e+00 3.23041826e-02 2.62717426e-01
-9.85301733e-01 2.23479830e-02 8.17510307e-01 2.13552579e-01
2.08204418e-01 -2.98933774e-01 -2.09872141e-01 -1.55673194e+00
1.74968973e-01 -7.03021049e-01 4.43487674e-01 -8.72459829e-01
-9.72966313e-01 4.28854585e-01 1.35097522e-02 -1.39234734e+00
-1.73112839e-01 -5.00885189e-01 -4.95008618e-01 8.10011983e-01
-4.85724539e-01 -1.33048534e+00 -8.40791762e-01 8.27279091e-01
4.38250393e-01 -1.82309151e-01 4.55671221e-01 3.58441144e-01
-5.43152511e-01 8.66711497e-01 -2.90501326e-01 2.77382821e-01
9.63805318e-01 -1.03894639e+00 8.14143777e-01 7.43383110e-01
-1.84619933e-01 5.24020255e-01 8.76750588e-01 -7.78475523e-01
-1.75432551e+00 -8.97492290e-01 8.17989349e-01 -1.17706394e+00
-1.67058855e-02 -5.11417806e-01 -3.98654372e-01 6.59481585e-01
8.72301981e-02 5.85193753e-01 5.34666896e-01 -1.11013390e-01
-8.35826918e-02 5.79131916e-02 -1.40356338e+00 5.81494391e-01
1.47429430e+00 -2.65489578e-01 -6.90006077e-01 1.76562052e-02
2.93272495e-01 -9.94066417e-01 -9.33450580e-01 5.71449399e-01
9.14806128e-01 -9.16108608e-01 1.22917628e+00 -1.60950109e-01
3.48522216e-02 -4.68791366e-01 -4.01822954e-01 -9.28516448e-01
-2.84768790e-01 -6.72153115e-01 -7.76716232e-01 1.13753653e+00
-2.64231920e-01 -4.89879996e-01 1.15271544e+00 1.15222108e+00
2.78205544e-01 -4.07003552e-01 -1.06536448e+00 -7.75013506e-01
-2.45847255e-01 -1.07775517e-01 4.02317345e-01 5.50726056e-01
-9.84331593e-02 2.55731344e-01 -1.06089747e+00 2.53600895e-01
1.18086815e+00 -1.05075136e-01 1.63455796e+00 -1.14559519e+00
-6.56861961e-02 2.33299613e-01 -8.44897926e-01 -1.35747266e+00
-2.20099628e-01 -3.53953421e-01 -3.14492494e-01 -1.37020969e+00
4.07866716e-01 2.89314181e-01 1.37804434e-01 2.48660371e-01
-1.98467866e-01 7.86750734e-01 7.05471218e-01 1.72745645e-01
-8.12835515e-01 7.27627814e-01 1.13113928e+00 -2.54461855e-01
-1.95871010e-01 6.72409981e-02 -3.04715037e-01 9.09303010e-01
8.09012890e-01 -6.60427064e-02 -2.48810127e-01 -3.31286579e-01
-1.58990517e-01 1.01076327e-01 8.96981120e-01 -1.45004058e+00
4.12418604e-01 3.35398436e-01 9.92475629e-01 -1.05633795e+00
9.02250588e-01 -7.03324735e-01 4.68749732e-01 5.87954104e-01
3.67366999e-01 -7.64196590e-02 2.90312082e-01 4.31295395e-01
-6.88005835e-02 4.59523320e-01 5.99434495e-01 1.09702893e-01
-8.14102054e-01 7.00520992e-01 1.30866170e-01 2.56621867e-01
8.34544361e-01 -2.64573634e-01 -4.42906320e-01 -6.90606236e-01
-8.45796466e-01 6.59446299e-01 6.78193152e-01 6.52083099e-01
7.27985322e-01 -1.48033226e+00 -7.52731204e-01 -1.83798149e-02
1.74932733e-01 -2.05120832e-01 8.80193770e-01 1.10937226e+00
-2.79553473e-01 5.33278644e-01 -3.32232565e-01 -7.68897593e-01
-1.82754862e+00 3.81267309e-01 3.40482086e-01 -3.52962613e-02
-9.97231543e-01 8.19157004e-01 3.73726696e-01 -4.62220669e-01
1.60208493e-01 2.97875047e-01 -2.34212637e-01 -1.65565789e-01
8.35980833e-01 7.80638218e-01 -2.43797615e-01 -1.28577816e+00
-7.16253102e-01 1.04122210e+00 1.49849266e-01 -6.04146011e-02
8.77624512e-01 -8.69026661e-01 3.67517471e-01 -2.23830774e-01
1.10382748e+00 2.83703864e-01 -1.48008156e+00 -2.34892219e-01
-5.62976360e-01 -6.98565483e-01 -1.53507158e-01 -5.52567124e-01
-9.20326829e-01 9.90716636e-01 1.12291861e+00 -5.82684219e-01
9.12739754e-01 9.56112146e-03 9.92637753e-01 3.17114502e-01
7.79186726e-01 -9.62742805e-01 -1.37086406e-01 4.62809950e-01
1.06432009e+00 -1.44106638e+00 2.88786620e-01 -6.38313591e-01
-8.20740283e-01 6.73121333e-01 8.67132604e-01 -3.21339034e-02
5.29600263e-01 -6.78950846e-02 9.51129273e-02 -2.11991861e-01
-3.59540015e-01 -1.28881916e-01 7.23996878e-01 8.04976046e-01
1.65265545e-01 1.05467454e-01 -2.90115446e-01 8.14467490e-01
-4.90929961e-01 1.56444192e-01 8.25653896e-02 6.12745404e-01
-5.67997023e-02 -7.21492171e-01 -6.52890027e-01 -7.70251378e-02
-4.40644890e-01 3.05130094e-01 -5.12250543e-01 7.48952270e-01
2.17419863e-01 8.94173324e-01 -1.80366546e-01 -6.75356448e-01
4.30056155e-01 -9.16797221e-02 5.91276050e-01 -6.40681982e-02
-2.03619823e-01 -6.94828853e-02 2.19978645e-01 -1.00387704e+00
-2.90128082e-01 -5.94041109e-01 -9.18868542e-01 -6.18881881e-01
1.63145095e-01 -1.98576346e-01 4.18270677e-01 5.97414196e-01
6.45335793e-01 8.42652097e-02 2.17692852e-01 -1.36720324e+00
-2.84445763e-01 -1.00291467e+00 -4.02495801e-01 6.80831790e-01
2.77618557e-01 -9.42405224e-01 -1.10833526e-01 1.24455839e-01] | [6.991231918334961, -0.918549120426178] |
fe56b032-150a-400c-9e3d-017a5e0a906c | multiple-sequence-alignment-for-short | 1510.09037 | null | http://arxiv.org/abs/1510.09037v2 | http://arxiv.org/pdf/1510.09037v2.pdf | Multiple sequence alignment for short sequences | Multiple sequence alignment (MSA) has been one of the most important problems
in bioinformatics for more decades and it is still heavily examined by many
mathematicians and biologists. However, mostly because of the practical
motivation of this problem, the research on this topic is focused on aligning
long sequences. It is understandable, since the sequences that need to be
aligned (usually DNA or protein sequences) are generally quite long (e. g., at
least 30-40 characters). Nevertheless, it is a challenging question that
exactly where MSA starts to become a real hard problem (since it is known that
MSA is NP-complete [2]), and the key to answer this question is to examine
short sequences. If the optimal alignment for short sequences could be
determined in polynomial time, then these results may help to develop faster or
more accurate heuristic algorithms for aligning long sequences. In this work,
it is shown that for length-1 sequences using arbitrary metric, as well as for
length-2 sequences using unit metric, the optimum of the MSA problem can be
achieved by the trivial alignment. | [] | 2015-11-15 | null | null | null | null | ['multiple-sequence-alignment'] | ['medical'] | [ 7.20865011e-01 -1.01711787e-01 -1.03063636e-01 -2.83851892e-01
-5.35147607e-01 -9.43778872e-01 -5.75157255e-02 5.14445662e-01
-4.89488423e-01 1.09843016e+00 -2.92651236e-01 -5.91886163e-01
-9.37176794e-02 -7.22190738e-01 -4.42458898e-01 -1.08364677e+00
-2.59363949e-01 6.72448456e-01 4.80439305e-01 -3.88610572e-01
6.81099653e-01 8.14078987e-01 -1.45023751e+00 -1.49147376e-01
1.09070301e+00 4.44358379e-01 6.72946811e-01 8.88525128e-01
-4.16793197e-01 -9.98779852e-03 -5.64765334e-01 -5.26676476e-01
3.26922119e-01 -8.34609985e-01 -1.31829417e+00 6.88774586e-02
1.60063103e-01 1.74681917e-01 3.94292265e-01 1.37380922e+00
1.79597050e-01 -5.53131327e-02 3.00127923e-01 -1.28113866e+00
-1.13850832e-03 4.32516307e-01 -6.21835470e-01 1.65092856e-01
2.95563161e-01 -4.58455421e-02 1.36695242e+00 -3.12420934e-01
6.79749250e-01 9.54559147e-01 4.36895192e-01 3.04072917e-01
-1.22604477e+00 -3.52812946e-01 -1.70711368e-01 7.61642337e-01
-1.43063104e+00 9.28173959e-02 3.65698695e-01 -3.30413222e-01
6.79013431e-01 6.57028198e-01 6.88275993e-01 6.31441653e-01
3.95218074e-01 6.23999119e-01 1.19353557e+00 -6.49986088e-01
1.69351920e-01 -2.23976418e-01 3.36210847e-01 2.86379576e-01
4.49661762e-01 -3.79755944e-01 5.58185652e-02 -4.68007654e-01
4.51775670e-01 -1.20532028e-01 -3.95761788e-01 -3.00635874e-01
-1.30088007e+00 9.85257745e-01 -1.07195117e-01 7.51201272e-01
-2.18757510e-01 -2.80690551e-01 3.82712275e-01 4.09652025e-01
-2.30762009e-02 8.22580397e-01 -4.24790204e-01 -5.90385973e-01
-8.09285879e-01 5.04331887e-01 1.05406725e+00 7.70999312e-01
8.60490441e-01 -3.66182417e-01 7.03318596e-01 6.66829765e-01
-2.79802501e-01 3.85779023e-01 2.29177833e-01 -6.94495857e-01
2.13000119e-01 5.51393509e-01 3.74603838e-01 -1.01198411e+00
-4.63575870e-01 -1.79353997e-01 -9.12352264e-01 -4.67666201e-02
1.10202026e+00 9.23778564e-02 -2.97606736e-01 1.73472714e+00
3.61651957e-01 1.44875988e-01 4.55852970e-03 1.04427505e+00
3.65894586e-02 9.07549024e-01 -3.07194114e-01 -7.38115609e-01
1.53020275e+00 -9.24362063e-01 -5.63560903e-01 -1.83461621e-01
6.89352572e-01 -1.17626071e+00 8.52862179e-01 3.94075572e-01
-8.74351919e-01 -1.55727789e-01 -1.15541089e+00 1.29337721e-02
-1.08753860e-01 -4.65319097e-01 5.35202086e-01 5.51939607e-01
-8.67062032e-01 6.79991007e-01 -8.10379565e-01 -8.08813334e-01
-3.40895087e-01 3.79850775e-01 -3.95757675e-01 -1.78178132e-01
-1.10361814e+00 1.08514798e+00 4.34130758e-01 -2.91777905e-02
-1.65765792e-01 -2.62983348e-02 -4.12469000e-01 6.00444004e-02
6.34582043e-01 -4.31509674e-01 1.04379416e+00 -1.17480314e+00
-1.19375932e+00 8.89909983e-01 -5.71356237e-01 -4.93079305e-01
4.40432400e-01 2.35788524e-01 -1.03897348e-01 2.50927806e-01
-8.34222585e-02 1.07002258e-01 4.05645013e-01 -7.65815556e-01
-7.61576593e-01 -5.22208512e-01 2.45092705e-01 4.14825864e-02
-2.66433191e-02 3.70987743e-01 -5.25786616e-02 -4.65845823e-01
3.66457582e-01 -1.40213585e+00 -8.62360954e-01 -2.45239928e-01
-2.82930106e-01 -3.50976974e-01 2.43482903e-01 -7.59743452e-01
1.15466094e+00 -1.93342674e+00 5.56710660e-01 3.92162383e-01
8.53656679e-02 2.48517439e-01 -1.74323320e-01 8.46626580e-01
-2.35739425e-01 4.90070991e-02 -7.06339061e-01 4.40131694e-01
-4.02445883e-01 4.40656960e-01 -2.59159446e-01 7.10508585e-01
4.28813621e-02 5.21377683e-01 -1.01230097e+00 -4.63015527e-01
-1.28912523e-01 3.91542427e-02 -2.29880527e-01 1.07438296e-01
-1.84818476e-01 4.68447298e-01 -5.15056193e-01 3.92449498e-01
8.46174419e-01 -3.29001278e-01 6.29304707e-01 1.69720337e-01
-3.85961294e-01 1.99835628e-01 -1.23912120e+00 1.30531478e+00
-3.47624943e-02 6.35718524e-01 9.87760574e-02 -1.36163259e+00
9.08884764e-01 2.92223066e-01 7.69174874e-01 -1.33975416e-01
-7.64652342e-02 5.35224140e-01 5.18667161e-01 -5.13789773e-01
5.46168268e-01 -2.73827344e-01 -4.03382480e-02 5.03620088e-01
-7.18184829e-01 -1.19541205e-01 7.04442084e-01 -2.49447480e-01
1.24967623e+00 -3.88485610e-01 7.66817212e-01 -4.39132214e-01
1.01678240e+00 4.91620302e-01 8.51306021e-01 3.35909367e-01
-1.31206349e-01 4.83627766e-01 6.74844563e-01 -5.80360591e-01
-1.51466882e+00 -6.53948545e-01 -1.57977745e-01 6.94115102e-01
3.84697527e-01 -4.42867696e-01 -9.52634156e-01 -2.44801521e-01
-2.58377433e-01 3.82426083e-01 -2.00876907e-01 2.20831946e-01
-9.42885816e-01 -8.20619881e-01 4.02160585e-01 1.12222813e-01
1.83501706e-01 -8.70700777e-01 -7.97038555e-01 6.62971854e-01
-5.04646480e-01 -1.01645684e+00 -3.51002157e-01 3.21489632e-01
-1.09113038e+00 -1.22908473e+00 -1.00042057e+00 -6.68989539e-01
6.32151246e-01 6.07611060e-01 8.78493130e-01 3.81068587e-01
-3.22424233e-01 -2.51522422e-01 -5.44444442e-01 -2.67634094e-01
-7.19649732e-01 5.14601111e-01 -3.35047692e-02 -2.33247533e-01
4.69505072e-01 -7.27609277e-01 -3.43577027e-01 7.91926742e-01
-1.14035511e+00 7.89270550e-02 5.33802271e-01 7.40563035e-01
6.24637485e-01 5.13341231e-03 5.11138976e-01 -6.66475058e-01
4.70360786e-01 -2.12918475e-01 -8.92011464e-01 6.22987509e-01
-2.44001597e-01 5.74462041e-02 9.76731479e-01 -2.79436052e-01
-2.54154205e-01 2.07759619e-01 -4.42667693e-01 2.12251633e-01
-2.24571854e-01 6.11575365e-01 -3.22824419e-01 -2.29816616e-01
4.41945374e-01 5.11323214e-01 2.05728814e-01 -3.27889800e-01
-5.27931042e-02 8.30016792e-01 2.59825736e-01 -6.04181230e-01
6.67872488e-01 4.20349538e-01 4.13076639e-01 -1.25901878e+00
-5.78158021e-01 -7.88702965e-01 -7.72459507e-01 8.94491747e-02
6.98276103e-01 -8.92468467e-02 -8.56509268e-01 2.85004675e-01
-1.17842245e+00 9.05910507e-02 3.07118356e-01 2.67223746e-01
-7.96963274e-01 1.07326472e+00 -3.58791977e-01 -6.91197157e-01
-1.63515374e-01 -1.55910718e+00 7.31341064e-01 8.21009725e-02
-5.02627015e-01 -6.59844041e-01 1.42467618e-01 2.87162870e-01
1.58314392e-01 3.06966305e-01 1.26322794e+00 -7.14872181e-01
-7.00257242e-01 -3.68720144e-02 -8.77834409e-02 1.32042840e-01
3.31316561e-01 1.61560342e-01 -2.10999086e-01 -3.66869628e-01
1.82775453e-01 9.85256210e-02 2.90605634e-01 2.02688634e-01
9.56606627e-01 -2.09725603e-01 -3.00998509e-01 3.19411814e-01
1.48901284e+00 3.81984532e-01 6.99704528e-01 6.37277782e-01
1.89640358e-01 9.24717009e-01 1.03283262e+00 2.55818903e-01
7.92733431e-02 8.53841901e-01 4.00845885e-01 2.14134529e-01
6.29785001e-01 3.44148815e-01 8.27866048e-03 1.01010311e+00
-3.19120020e-01 -2.98394877e-02 -1.06726575e+00 5.45608282e-01
-1.86274648e+00 -1.16968548e+00 -5.96269846e-01 2.46494317e+00
1.02562642e+00 -1.92369595e-01 3.53098512e-01 4.99416530e-01
9.32577670e-01 -9.67269018e-02 -5.71344197e-01 -7.77685463e-01
-4.51568812e-01 3.83886322e-02 6.40905976e-01 4.29972827e-01
-8.33348513e-01 7.43735492e-01 6.42761803e+00 8.00403714e-01
-1.15234101e+00 -4.43858743e-01 4.16665405e-01 3.16518366e-01
-1.57976985e-01 2.56888121e-01 -7.35540211e-01 5.60625017e-01
9.23038900e-01 -6.90456867e-01 4.65664476e-01 7.87115216e-01
1.90540984e-01 -3.17872405e-01 -8.81549656e-01 8.54505301e-01
-1.99794739e-01 -1.08219385e+00 -2.08768070e-01 3.27551812e-01
6.30906165e-01 -3.02578062e-01 -3.01692903e-01 -3.50454986e-01
5.89866377e-02 -9.09424424e-01 1.97186664e-01 -1.49588197e-01
2.94820845e-01 -1.11710131e+00 9.03557599e-01 7.14133799e-01
-1.06694472e+00 2.06210271e-01 -8.75137329e-01 -9.39921886e-02
4.44310635e-01 7.38985360e-01 -9.15144205e-01 6.62615120e-01
3.62666309e-01 3.84250075e-01 -2.01103613e-01 1.50357449e+00
-7.84573108e-02 5.40829062e-01 -4.60866719e-01 -3.52003723e-01
4.77023453e-01 -7.89301693e-01 6.45627499e-01 1.09732473e+00
4.95088637e-01 3.78148526e-01 2.43960887e-01 1.26694322e-01
4.29879069e-01 5.28459191e-01 -4.53194857e-01 -3.45568627e-01
1.53404310e-01 1.08510208e+00 -1.06025279e+00 -2.48731479e-01
-3.22357208e-01 9.60503042e-01 2.39633903e-01 4.18245420e-02
-7.84201026e-01 -5.88511944e-01 1.09768999e+00 8.82882159e-03
2.67993838e-01 -6.41034901e-01 -3.90750736e-01 -8.09222460e-01
7.32191354e-02 -1.26070380e+00 5.12959778e-01 -2.79653817e-01
-1.11212969e+00 6.92602456e-01 -2.62746960e-01 -1.16265774e+00
-5.24039626e-01 -4.90668684e-01 -1.02891602e-01 8.65774691e-01
-1.38626468e+00 -4.72924322e-01 1.03458628e-01 1.91659018e-01
5.53052127e-01 3.71954769e-01 8.23876560e-01 4.69357073e-02
-3.09939563e-01 3.67584199e-01 5.54462016e-01 -2.65790313e-01
5.80278814e-01 -1.11066449e+00 3.60541612e-01 7.12601542e-01
5.36091812e-02 7.29123056e-01 1.24213016e+00 -4.73474473e-01
-1.68978190e+00 -5.15795887e-01 1.26719415e+00 5.78782894e-03
8.69973242e-01 -8.12580809e-02 -1.21810079e+00 3.53279889e-01
8.71845037e-02 -5.13979971e-01 9.09095347e-01 -1.92198768e-01
-7.21782520e-02 1.71328694e-01 -1.02834201e+00 8.35952878e-01
9.11202729e-01 -7.16477185e-02 -5.55027664e-01 3.67049307e-01
4.71545786e-01 -2.90680647e-01 -7.57247925e-01 2.61954844e-01
4.64918584e-01 -9.38144505e-01 7.97755003e-01 -6.09712899e-01
3.69040817e-01 -5.80756962e-01 -1.91786855e-01 -1.08399200e+00
-1.16436101e-01 -7.10845351e-01 4.61239278e-01 6.86170280e-01
3.19676429e-01 -7.05950141e-01 9.88607943e-01 1.05995804e-01
8.02567899e-02 -9.23232138e-01 -1.09988737e+00 -1.15529168e+00
8.94758254e-02 -7.38650486e-02 7.17768848e-01 1.07364643e+00
1.56558007e-01 1.59609243e-01 -3.79669249e-01 2.70595163e-01
4.80209529e-01 7.46636152e-01 8.70519161e-01 -1.32800233e+00
-2.79431850e-01 -4.65790659e-01 -6.44604743e-01 -1.22430420e+00
1.34420589e-01 -6.87437356e-01 7.96242803e-02 -1.43087673e+00
1.85966462e-01 -5.63626945e-01 1.85509026e-01 2.08678365e-01
-3.70745122e-01 -7.31147155e-02 1.03089504e-01 9.76962149e-02
-1.84189707e-01 6.30431399e-02 1.24771011e+00 7.07763582e-02
1.03520490e-01 3.26458156e-01 -3.24771076e-01 5.45560241e-01
8.06664228e-01 -4.64562148e-01 -4.60154191e-02 -1.00658692e-01
3.85197669e-01 3.10840547e-01 -2.60369003e-01 -5.93031406e-01
7.43347257e-02 -8.41173351e-01 -5.04662037e-01 -6.92781508e-01
2.68736631e-01 -1.09388936e+00 4.57870424e-01 6.48549318e-01
-8.52493346e-02 5.16631603e-01 -1.21200599e-01 3.13633442e-01
-2.87230313e-01 -9.25521433e-01 7.97251821e-01 -1.82779923e-01
-5.50719619e-01 9.37049687e-02 -5.95734179e-01 -1.68749094e-01
1.28134823e+00 -5.50575316e-01 -5.06503470e-02 -3.46879154e-01
-5.17071366e-01 2.34802425e-01 9.82169926e-01 1.03469811e-01
4.16078985e-01 -8.52111340e-01 -7.56622314e-01 -2.71452129e-01
1.58475563e-01 -9.28068310e-02 2.45245486e-01 9.48257327e-01
-1.12220526e+00 5.36205888e-01 -3.33813846e-01 -9.23003852e-01
-1.96022141e+00 8.36068213e-01 -1.51608109e-01 -3.24173659e-01
-4.55039084e-01 5.19070029e-01 -6.64335042e-02 -1.45773897e-02
6.11672457e-03 -3.65624502e-02 7.23662823e-02 -9.74688306e-02
7.90799081e-01 4.48095083e-01 -7.44599849e-02 -6.50576770e-01
-2.79130727e-01 7.87909091e-01 -1.41926259e-01 7.71518871e-02
1.48976195e+00 -1.23090602e-01 -7.88443685e-01 3.04393202e-01
1.13899386e+00 1.18878141e-01 -6.08952761e-01 -3.62684503e-02
6.69812262e-01 -6.89416111e-01 -8.67351472e-01 -1.63866922e-01
-6.26618147e-01 9.56028104e-01 1.11521773e-01 5.02213836e-01
1.14997995e+00 -2.33589694e-01 9.98483121e-01 6.87036276e-01
7.66820312e-01 -6.79868639e-01 -4.18317407e-01 6.70675755e-01
6.71419322e-01 -8.60578060e-01 -5.24818003e-02 -9.00821567e-01
-2.47774228e-01 1.40009236e+00 2.69183129e-01 1.06592380e-01
-6.72648251e-02 2.94263750e-01 -5.19324467e-02 3.40520591e-01
-4.94158715e-01 -1.36139959e-01 -2.07750961e-01 2.42576912e-01
5.34421980e-01 7.16346204e-02 -1.13184834e+00 4.17121015e-02
-5.12260318e-01 -1.85430959e-01 8.87124360e-01 1.03746605e+00
-8.20330560e-01 -1.91734648e+00 -6.42187655e-01 2.11818799e-01
-5.55209637e-01 2.35780571e-02 -4.61609095e-01 5.55012226e-01
-2.35422656e-01 7.71151721e-01 -1.85741290e-01 -7.87802488e-02
-6.09451383e-02 1.88618507e-02 5.92157960e-01 -4.25352693e-01
-5.31142056e-01 -4.83537875e-02 1.00598142e-01 -2.73698837e-01
-4.18669134e-01 -6.20538056e-01 -1.29863238e+00 -6.66917086e-01
-4.52558428e-01 7.41354644e-01 8.33648205e-01 1.08678293e+00
4.69525717e-03 -1.47960305e-01 6.95553422e-01 -2.89293855e-01
-8.27013493e-01 -6.35797262e-01 -6.72573566e-01 2.74588108e-01
1.15927577e-01 -2.95523077e-01 -2.64369696e-01 -6.52539283e-02] | [4.891394138336182, 5.142265319824219] |
62e09f35-275b-4f63-8504-14fa8ac3c991 | cloud-based-deep-learning-end-to-end-full | 2304.13506 | null | https://arxiv.org/abs/2304.13506v1 | https://arxiv.org/pdf/2304.13506v1.pdf | Cloud-Based Deep Learning: End-To-End Full-Stack Handwritten Digit Recognition | Herein, we present Stratus, an end-to-end full-stack deep learning application deployed on the cloud. The rise of productionized deep learning necessitates infrastructure in the cloud that can provide such service (IaaS). In this paper, we explore the use of modern cloud infrastructure and micro-services to deliver accurate and high-speed predictions to an end-user, using a Deep Neural Network (DNN) to predict handwritten digit input, interfaced via a full-stack application. We survey tooling from Spark ML, Apache Kafka, Chameleon Cloud, Ansible, Vagrant, Python Flask, Docker, and Kubernetes in order to realize this machine learning pipeline. Through our cloud-based approach, we are able to demonstrate benchmark performance on the MNIST dataset with a deep learning model. | ['Terry Luo', 'Ashwin Kumar', 'Aadarsh Jha', 'Ruida Zeng'] | 2023-02-01 | null | null | null | null | ['handwritten-digit-recognition'] | ['computer-vision'] | [-8.32070827e-01 -2.51523405e-01 5.73033750e-01 -7.33407140e-01
-2.98562765e-01 -7.39517570e-01 2.83097893e-01 -2.50685036e-01
-2.45776728e-01 2.93003380e-01 -3.94987077e-01 -8.05822551e-01
-1.33968070e-01 -8.85563076e-01 -6.53768361e-01 -6.08486712e-01
-8.75602812e-02 8.07993650e-01 1.62882246e-02 -3.08141019e-02
-7.19628930e-02 9.08859670e-01 -1.71648920e+00 7.27476299e-01
3.01883876e-01 1.25430286e+00 3.20457071e-01 1.08185077e+00
-2.42175564e-01 1.32365847e+00 -5.41857958e-01 -6.08434618e-01
5.51581562e-01 6.03688717e-01 -7.02210426e-01 -5.91607511e-01
3.12021643e-01 -4.10562426e-01 -1.44309983e-01 6.36997402e-01
9.82996583e-01 -5.67154169e-01 -4.95318696e-02 -1.36929142e+00
-2.19430178e-01 4.14240927e-01 3.05210315e-02 1.90748289e-01
-2.55228467e-02 7.17727005e-01 3.80531698e-01 -9.49717641e-01
5.65560818e-01 9.29565251e-01 9.87524569e-01 6.52242541e-01
-7.74790525e-01 -7.48456836e-01 -2.90309906e-01 1.72504708e-01
-8.82460296e-01 -4.34181929e-01 3.07278395e-01 -6.98110700e-01
1.56854546e+00 3.26881975e-01 5.47845542e-01 1.06845331e+00
6.00128233e-01 8.10870767e-01 1.05026472e+00 -2.30352193e-01
6.72389328e-01 3.43721956e-02 5.13752937e-01 5.76498866e-01
-1.50509834e-01 1.18312150e-01 -5.21556616e-01 -8.48436654e-02
3.99756193e-01 3.50694656e-01 4.42921847e-01 1.91879928e-01
-7.53351986e-01 4.01892871e-01 4.64337021e-01 4.24763560e-02
-9.18604136e-01 4.01133180e-01 8.28482211e-01 3.31480980e-01
6.89659774e-01 1.00435711e-01 -1.21724129e+00 -6.33579373e-01
-7.50651658e-01 6.55963659e-01 1.34562469e+00 7.73834229e-01
8.36092889e-01 1.31838083e-01 1.03857517e-01 3.21357906e-01
2.93347567e-01 -1.26858920e-01 8.79361451e-01 -9.39935207e-01
4.95554768e-02 6.88102067e-01 -2.27622107e-01 -2.23022550e-01
-4.43833619e-01 -6.51192129e-01 -5.48550189e-01 9.48718131e-01
2.47186467e-01 -5.66232562e-01 -1.09378719e+00 7.81574428e-01
5.56931138e-01 3.51898164e-01 3.43833536e-01 1.07947981e+00
1.03901958e+00 1.99666485e-01 2.41950482e-01 7.31941104e-01
1.23342860e+00 -1.20938516e+00 -2.02881590e-01 1.56271160e-02
7.39657998e-01 -8.39634717e-01 1.19406784e+00 8.95039737e-01
-1.08478737e+00 -5.95278144e-01 -8.48539650e-01 -3.35667074e-01
-7.09921718e-01 2.15647951e-01 1.19999647e+00 5.93695521e-01
-1.28025174e+00 9.19003785e-01 -1.52377236e+00 -1.18121773e-01
5.83655894e-01 5.15164375e-01 -3.00553918e-01 1.78286489e-02
-3.71523380e-01 6.86685920e-01 3.47550184e-01 -9.10916831e-03
-1.38312972e+00 -1.24706030e+00 -1.68260768e-01 2.20047042e-01
-4.15556192e-01 -7.50783324e-01 1.61555016e+00 -9.32915151e-01
-1.60768402e+00 1.01120138e+00 3.29633683e-01 -8.32867742e-01
8.44721973e-01 -8.97964761e-02 -1.14981636e-01 -2.25675225e-01
-3.58331919e-01 3.41587543e-01 1.43872008e-01 -6.46278977e-01
-7.91205764e-01 -8.75817120e-01 -3.96711491e-02 -3.55821699e-01
-1.43509224e-01 3.58888537e-01 4.64848429e-03 -1.09828621e-01
-3.19600403e-01 -5.89041293e-01 -3.29768419e-01 -1.81452900e-01
-3.15858960e-01 -4.41293001e-01 1.14753985e+00 -8.70824575e-01
2.84580350e-01 -2.14733672e+00 -4.17532682e-01 -5.16025126e-02
4.04768407e-01 3.72080177e-01 6.96960315e-02 5.75390518e-01
-2.49274462e-01 -2.63955235e-01 2.10630298e-01 -5.68330705e-01
1.76760539e-01 3.30973148e-01 -3.61837208e-01 6.43264130e-02
-1.93346456e-01 6.98655009e-01 -3.65803838e-01 -1.02782026e-02
2.54193962e-01 7.19215274e-01 -5.02763927e-01 5.36108077e-01
-4.81927425e-01 8.30162391e-02 -2.01167598e-01 9.27401841e-01
9.92112100e-01 -2.93586165e-01 3.88536835e-03 2.37541512e-01
-3.57322127e-01 9.67579558e-02 -8.19429696e-01 1.97112787e+00
-7.25124836e-01 5.46718419e-01 4.23229754e-01 -6.17739081e-01
1.19562018e+00 2.62349159e-01 4.87154484e-01 -3.47247303e-01
-7.40843117e-02 3.49947244e-01 -1.89903110e-01 -5.13581634e-01
1.50919914e-01 4.12965715e-01 4.99200314e-01 2.12463111e-01
4.31899011e-01 3.88053328e-01 3.16830017e-02 1.01117805e-01
1.59449744e+00 3.82299930e-01 -3.95794064e-01 -2.99946070e-01
6.77208006e-02 2.96583444e-01 3.34369451e-01 3.90396595e-01
-1.37653530e-01 1.89464763e-01 6.72303200e-01 -1.23345256e+00
-1.16449916e+00 -9.98030663e-01 -1.94746032e-01 1.59037054e+00
-9.67756331e-01 -3.91214013e-01 -1.20102525e+00 -4.04341072e-01
2.14064047e-01 5.26001453e-01 -2.34127462e-01 4.13987666e-01
-3.21060754e-02 -7.32333064e-01 7.57057786e-01 5.95535994e-01
5.98881006e-01 -1.12267661e+00 -1.00366282e+00 4.16339666e-01
7.62716234e-01 -1.10179555e+00 5.93368948e-01 5.16130865e-01
-7.10635483e-01 -1.10239649e+00 -4.56913382e-01 -6.26796782e-01
-2.18754634e-02 -3.18323135e-01 1.43376958e+00 -1.30054459e-01
-9.13904428e-01 1.44774899e-01 -1.10460669e-02 -9.73687232e-01
-1.34669289e-01 2.62703955e-01 -1.71127185e-01 -4.09522563e-01
8.83418024e-01 -1.02855933e+00 -8.36952090e-01 -3.28558028e-01
-7.10523784e-01 4.28218961e-01 3.50176334e-01 3.08576375e-01
3.79486263e-01 -3.45733643e-01 6.26207054e-01 -7.04384327e-01
5.39266169e-01 -8.78466547e-01 -1.18997788e+00 -5.39306253e-02
-8.90786052e-01 -4.09630477e-01 8.34570885e-01 1.30222380e-01
-8.74910891e-01 6.08070791e-01 -7.32492685e-01 -8.07158113e-01
-4.22926039e-01 5.86004019e-01 -9.00050998e-02 -1.29362047e-01
9.95465696e-01 3.24828207e-01 2.24750653e-01 -9.77342188e-01
1.27192184e-01 1.17072845e+00 6.95441246e-01 -6.09747767e-01
-2.42102563e-01 2.81061858e-01 -7.59260580e-02 -2.28076279e-01
-2.09904507e-01 -2.24963039e-01 -2.63745308e-01 -2.08179101e-01
5.37385583e-01 -1.28138649e+00 -1.58142793e+00 8.96622479e-01
-1.36815524e+00 -7.82575905e-01 -5.67966178e-02 9.28931013e-02
-4.63893414e-01 -2.58538634e-01 -1.12511361e+00 -7.35576808e-01
-1.13549328e+00 -1.15324187e+00 8.06842983e-01 2.89404005e-01
3.90031129e-01 -6.95871174e-01 2.28909925e-01 4.40319270e-01
8.43182147e-01 4.30735052e-01 3.04896504e-01 -1.08215654e+00
-7.52610147e-01 -5.27992547e-01 -3.32785249e-01 7.38251090e-01
-5.76429605e-01 4.35255826e-01 -1.54525793e+00 -1.01762891e-01
-3.36807221e-01 -4.88315910e-01 5.05109847e-01 2.99644530e-01
1.80980074e+00 -2.47338399e-01 -1.04080103e-01 1.07847321e+00
1.72188056e+00 -7.97760710e-02 5.53622186e-01 5.90561628e-01
6.14320636e-01 2.93991238e-01 8.23622383e-03 8.77298057e-01
5.25408983e-01 1.12787127e-01 9.92345929e-01 -2.83782613e-02
1.23172648e-01 3.33611101e-01 1.29420878e-02 3.49300414e-01
-3.15846562e-01 4.83920276e-01 -1.62868667e+00 1.86280206e-01
-1.89143217e+00 -4.15444613e-01 -4.72404569e-01 1.62516391e+00
8.37183952e-01 -7.96064064e-02 1.41470075e-01 -7.08320737e-02
1.27467841e-01 -6.43150091e-01 -8.45320523e-01 -5.26649475e-01
2.99903840e-01 6.38714492e-01 3.71652335e-01 -7.76475994e-03
-7.21217155e-01 1.16784620e+00 6.07892513e+00 5.69825888e-01
-1.53053379e+00 5.55456281e-01 9.26989853e-01 -3.29615027e-01
2.39072636e-01 -2.15599284e-01 -7.54212916e-01 7.60304153e-01
1.72466207e+00 -1.85862586e-01 5.99537134e-01 1.96516764e+00
1.14298835e-01 2.72177666e-01 -1.11189735e+00 7.73681104e-01
-7.51436472e-01 -1.93571866e+00 -7.32227981e-01 1.53420687e-01
2.04561293e-01 1.10870242e+00 -1.17679350e-01 5.07591009e-01
7.91779518e-01 -1.18204200e+00 5.53728998e-01 5.66111386e-01
5.52603245e-01 -1.12280476e+00 8.57272446e-01 4.93915021e-01
-5.62222898e-01 -9.32957977e-02 -5.08811355e-01 -4.78925943e-01
-5.70637286e-01 1.16435540e+00 -1.12738192e+00 3.58495206e-01
1.26576018e+00 4.34092432e-01 -5.52431345e-01 8.58383954e-01
2.72552997e-01 5.71063101e-01 -1.98124975e-01 4.61246632e-02
1.32731944e-01 1.25586212e-01 -2.79080451e-01 1.41618371e+00
1.62677348e-01 -1.72181085e-01 6.88965395e-02 7.65559375e-01
-8.11310336e-02 -8.98224488e-02 -3.58502746e-01 1.92496657e-01
2.15057597e-01 1.82456279e+00 -4.19686377e-01 -7.08530664e-01
-1.14125580e-01 9.36886668e-01 5.15862286e-01 1.38281986e-01
-4.98571128e-01 -4.25812125e-01 1.27176261e+00 -7.71828443e-02
-4.69287969e-02 -3.01713496e-01 -7.38453507e-01 -1.01519728e+00
1.81227714e-01 -1.04001176e+00 1.46324769e-01 -9.97958839e-01
-1.37287962e+00 8.74057949e-01 -6.15919709e-01 -4.85877991e-01
-4.39888716e-01 -1.22502613e+00 -7.96362102e-01 1.27425587e+00
-1.42456138e+00 -1.38512886e+00 -7.47395217e-01 7.46650457e-01
3.52838814e-01 -7.12505698e-01 1.14173710e+00 6.22799397e-01
-7.24159896e-01 1.81103289e-01 3.27973008e-01 2.28369385e-01
2.14811265e-01 -1.56535971e+00 7.06807852e-01 5.60365796e-01
-3.35786968e-01 3.99729609e-01 6.65267825e-01 -3.82760406e-01
-1.86678147e+00 -1.39724624e+00 4.22079086e-01 -4.15759742e-01
6.48357689e-01 -4.34435725e-01 -7.77468085e-01 1.02682483e+00
3.40460628e-01 6.44642293e-01 8.36199880e-01 1.39933661e-01
-5.01305640e-01 -5.45220494e-01 -1.44542241e+00 7.88822845e-02
5.21983922e-01 -2.82925427e-01 2.27051117e-02 9.15757179e-01
6.02443695e-01 -8.57240021e-01 -1.17438614e+00 9.35598463e-02
5.10620713e-01 -1.46401024e+00 8.34961832e-01 -9.31076288e-01
1.05781710e+00 1.42397985e-01 -2.54097551e-01 -1.24255812e+00
1.75067410e-01 -6.00987315e-01 -4.40987527e-01 1.02647650e+00
-2.36757984e-03 -8.12860012e-01 1.31148028e+00 1.16509485e+00
-5.68091333e-01 -9.93822932e-01 -1.08784866e+00 -6.11508608e-01
1.65579706e-01 -7.43163526e-01 1.28769207e+00 7.63782203e-01
-2.40047261e-01 -1.02545172e-01 1.84424058e-01 7.70367831e-02
7.10886419e-01 -1.35245442e-01 1.08430386e+00 -1.24851465e+00
-5.60975909e-01 -1.02902867e-01 -8.22613835e-01 -1.41583249e-01
1.61751151e-01 -8.86207283e-01 -3.91880989e-01 -1.36573553e+00
5.41315414e-02 -4.94787931e-01 -1.28259778e-01 9.09341514e-01
4.54243660e-01 -1.95406489e-02 4.27374780e-01 1.58699110e-01
-1.99581757e-01 -6.00050539e-02 4.82680559e-01 1.91356570e-01
2.05602631e-01 1.18113637e-01 -2.71343321e-01 6.53822243e-01
7.11259246e-01 -3.24773401e-01 1.75592914e-01 -7.07970619e-01
2.23997086e-01 -6.34717047e-02 8.14529240e-01 -1.38978446e+00
5.52454531e-01 3.27433646e-01 6.76461637e-01 -4.72698182e-01
2.35108703e-01 -9.14184749e-01 3.14460546e-01 7.13818252e-01
-1.61288381e-01 2.72422433e-01 3.34661245e-01 -4.21659164e-02
8.16399604e-03 -3.14100794e-02 5.98427415e-01 -3.26885968e-01
-7.35702693e-01 4.44259942e-01 -1.41285643e-01 -6.20736778e-01
1.17281878e+00 3.34916711e-01 -6.56779826e-01 2.47687504e-01
-8.88583601e-01 -3.49070169e-02 2.24440247e-01 1.66433901e-01
4.22852278e-01 -8.09892595e-01 -8.86604071e-01 3.18672895e-01
-8.60866606e-02 2.55360067e-01 3.35740805e-01 3.74633908e-01
-1.35286868e+00 2.83288509e-01 -6.19730413e-01 -6.49022639e-01
-1.27763975e+00 4.48173642e-01 7.65737534e-01 -1.50762662e-01
-7.36791730e-01 8.15804541e-01 -5.96534669e-01 -9.65516567e-01
5.13190567e-01 -4.03918803e-01 4.66801107e-01 -4.39670593e-01
8.63805354e-01 3.81316841e-01 1.00251961e+00 5.19882321e-01
-5.18145263e-01 -3.46747547e-01 9.35773253e-02 1.28389373e-01
1.82512546e+00 6.82751417e-01 -4.64328378e-01 7.32693598e-02
1.11759698e+00 -7.02841997e-01 -1.56697845e+00 3.75390023e-01
3.86941023e-02 -2.49089256e-01 4.48681355e-01 -1.28788304e+00
-1.39103270e+00 9.91296828e-01 1.15416718e+00 2.40472540e-01
1.13145840e+00 -2.10462630e-01 5.75898349e-01 6.92092836e-01
3.90408456e-01 -9.04520810e-01 -4.25251961e-01 5.43157578e-01
6.95624292e-01 -1.13011038e+00 -3.61175328e-01 8.43431503e-02
-3.92479926e-01 1.71090746e+00 8.20030510e-01 -4.30500984e-01
8.71387184e-01 1.08885872e+00 4.38365787e-01 -2.68863827e-01
-1.37141442e+00 2.43640393e-01 -5.62115371e-01 4.92463559e-01
6.76481962e-01 4.03137833e-01 2.26679787e-01 9.50518191e-01
-5.50280273e-01 9.41115201e-01 4.69420284e-01 9.21921730e-01
-1.44938789e-02 -7.14446247e-01 -2.44706988e-01 6.06232464e-01
-1.00720048e+00 -2.33280599e-01 -3.29944752e-02 4.39334720e-01
1.52634755e-01 5.08642912e-01 3.99664938e-01 -4.79928523e-01
2.53009170e-01 8.13059807e-02 6.06699986e-03 -2.87967533e-01
-1.39917862e+00 -4.04888868e-01 -3.63820121e-02 -8.24677944e-01
5.21812797e-01 -1.93478048e-01 -1.31596327e+00 -8.84688437e-01
4.17887837e-01 -2.02781260e-01 1.99011230e+00 7.65087128e-01
1.17285919e+00 6.92824006e-01 4.26958025e-01 -1.10779977e+00
-8.57444823e-01 -1.08704376e+00 -6.27355337e-01 -1.96348086e-01
-1.32314056e-01 2.25402921e-01 -1.60588339e-01 5.37056401e-02] | [8.522989273071289, 2.913975238800049] |
5499b59a-d5e9-437f-984a-acddf08144d8 | quantifying-quality-of-class-conditional | 2210.07617 | null | https://arxiv.org/abs/2210.07617v1 | https://arxiv.org/pdf/2210.07617v1.pdf | Quantifying Quality of Class-Conditional Generative Models in Time-Series Domain | Generative models are designed to address the data scarcity problem. Even with the exploding amount of data, due to computational advancements, some applications (e.g., health care, weather forecast, fault detection) still suffer from data insufficiency, especially in the time-series domain. Thus generative models are essential and powerful tools, but they still lack a consensual approach for quality assessment. Such deficiency hinders the confident application of modern implicit generative models on time-series data. Inspired by assessment methods on the image domain, we introduce the InceptionTime Score (ITS) and the Frechet InceptionTime Distance (FITD) to gauge the qualitative performance of class conditional generative models on the time-series domain. We conduct extensive experiments on 80 different datasets to study the discriminative capabilities of proposed metrics alongside two existing evaluation metrics: Train on Synthetic Test on Real (TSTR) and Train on Real Test on Synthetic (TRTS). Extensive evaluation reveals that the proposed assessment method, i.e., ITS and FITD in combination with TSTR, can accurately assess class-conditional generative model performance. | ['Sheraz Ahmed', 'Andreas Dengel', 'Peter Schichtel', 'Sankrutyayan Thota', 'Maria Walch', 'Alireza Koochali'] | 2022-10-14 | null | null | null | null | ['fault-detection'] | ['miscellaneous'] | [ 1.74533546e-01 -3.44113737e-01 1.94877550e-01 -4.02137727e-01
-8.87830973e-01 -4.91538554e-01 7.81922817e-01 -2.09791347e-01
-1.95690840e-01 6.46330535e-01 -9.12301540e-02 -2.62065947e-01
-5.04894614e-01 -7.63476133e-01 -2.78066486e-01 -9.73723888e-01
3.04115657e-02 3.21147949e-01 8.42092335e-02 -1.21955328e-01
1.82561100e-01 4.94887978e-02 -1.74000311e+00 -3.41226682e-02
1.26581907e+00 1.15871727e+00 -3.06622721e-02 3.59056383e-01
1.37385517e-01 5.45389473e-01 -8.70807827e-01 -3.72197270e-01
-1.40943518e-02 -6.33894742e-01 -2.30712742e-01 -3.89117114e-02
-1.32662565e-01 -1.64401103e-02 -1.71818271e-01 1.01700771e+00
5.44073105e-01 -5.77730015e-02 8.89974594e-01 -1.42861795e+00
-7.11013258e-01 1.32343933e-01 -4.61960912e-01 4.24189091e-01
3.17571014e-01 2.24214077e-01 6.68104827e-01 -9.90299046e-01
3.19330573e-01 1.05660307e+00 7.86622047e-01 9.03029367e-02
-9.29836750e-01 -4.88498867e-01 -1.93360940e-01 1.52283326e-01
-1.33007681e+00 -9.65308398e-02 8.29210162e-01 -5.30657172e-01
4.93132949e-01 1.99444726e-01 5.20842791e-01 1.18678188e+00
4.79598045e-01 6.30711854e-01 1.50917614e+00 -2.61733443e-01
2.94723302e-01 1.17061418e-02 -4.89115305e-02 3.25741947e-01
1.52415156e-01 2.91882753e-01 -5.13560951e-01 2.61462629e-02
6.56984568e-01 1.78015530e-01 -1.97570160e-01 4.07395177e-02
-1.14005923e+00 6.73506796e-01 1.62259892e-01 4.12151933e-01
-3.46396834e-01 -9.40313488e-02 3.03321123e-01 3.01266700e-01
6.22112393e-01 5.82970530e-02 -1.18830457e-01 -4.76379603e-01
-9.90808547e-01 1.99738979e-01 4.84036773e-01 9.02262270e-01
5.36345303e-01 4.40260887e-01 -3.60468447e-01 7.01956868e-01
2.79234827e-01 7.60034144e-01 6.27721548e-01 -3.32981050e-01
2.37631693e-01 6.27588212e-01 -8.90407786e-02 -1.04154956e+00
-2.14104220e-01 -9.19383407e-01 -1.01073158e+00 -7.57475197e-02
1.35736778e-01 -5.48692532e-02 -9.78035808e-01 1.66739583e+00
2.56873995e-01 2.50387311e-01 -8.81776363e-02 6.69307470e-01
6.43394828e-01 6.41705811e-01 7.30100051e-02 -3.84489864e-01
1.09758568e+00 -3.11805278e-01 -7.65861630e-01 4.71618399e-02
4.18368489e-01 -9.45822716e-01 1.05294085e+00 5.32785416e-01
-7.18595445e-01 -7.48186827e-01 -1.04270661e+00 5.70892334e-01
-2.93781400e-01 2.66937345e-01 5.16515493e-01 7.75155246e-01
-7.05461562e-01 4.58820224e-01 -1.03546536e+00 -2.44868472e-01
2.94131219e-01 -1.22783080e-01 -2.29953349e-01 -2.67961323e-01
-1.20887923e+00 6.73830152e-01 2.97553033e-01 1.84844926e-01
-1.02059436e+00 -5.96236527e-01 -6.69447184e-01 -1.37800992e-01
1.27123207e-01 -3.92358720e-01 9.31281924e-01 -4.86792743e-01
-1.07877934e+00 5.03335834e-01 3.22077870e-01 -4.06803250e-01
6.32555187e-01 -2.76062697e-01 -8.63013625e-01 -2.43645515e-02
2.55857706e-01 5.02830110e-02 8.21697235e-01 -1.20752728e+00
-6.17656529e-01 -4.00447279e-01 -3.01775694e-01 -1.33003384e-01
-3.90462607e-01 -2.83281535e-01 -1.45736307e-01 -8.86552989e-01
2.82133371e-01 -8.11445594e-01 1.17349803e-01 -3.11732262e-01
-4.22684997e-01 -3.73483688e-01 1.27466094e+00 -7.10290790e-01
1.46583962e+00 -2.18029332e+00 -3.44972700e-01 1.62180036e-01
5.97377531e-02 2.68812418e-01 1.56405583e-01 7.66314983e-01
-1.59529567e-01 -1.62531197e-01 -6.40664041e-01 -9.06002894e-02
1.62524968e-01 2.57098407e-01 -3.36986959e-01 4.43178296e-01
5.55845141e-01 7.30439782e-01 -9.63980794e-01 -5.23141265e-01
2.13719428e-01 5.78403533e-01 -2.52582848e-01 2.21493259e-01
2.27240361e-02 4.07295316e-01 -4.82705146e-01 7.52771854e-01
6.06779099e-01 -3.19899470e-01 -1.83653504e-01 -1.93024188e-01
-3.93848028e-03 1.47988359e-02 -9.20251906e-01 1.58056808e+00
-3.41317534e-01 5.11827171e-01 -8.35652351e-01 -1.05462444e+00
1.20046341e+00 4.44199473e-01 5.29072702e-01 -1.04964030e+00
-3.39807570e-02 3.74844849e-01 7.56155923e-02 -5.42023718e-01
3.30137283e-01 -3.83725554e-01 -2.09569588e-01 3.46332371e-01
1.79907292e-01 3.09399422e-02 2.23655969e-01 1.21380240e-01
1.06150162e+00 4.53565150e-01 1.88855827e-01 -1.59861997e-01
1.93692446e-01 -2.03301668e-01 4.81949061e-01 2.06402138e-01
-1.21878281e-01 8.67073476e-01 5.28505325e-01 -2.59813160e-01
-1.19812667e+00 -1.22043943e+00 -4.15995270e-01 6.29113019e-01
-1.34018315e-02 -2.34451964e-01 -6.21593952e-01 -7.22903252e-01
-1.46921262e-01 6.93960428e-01 -6.57230079e-01 -5.31100869e-01
-1.75437212e-01 -1.27262068e+00 6.95248425e-01 6.72874272e-01
5.45217812e-01 -8.19961488e-01 -8.48206401e-01 2.43435338e-01
-1.84464931e-01 -1.01232767e+00 -2.50991583e-01 1.36922002e-01
-8.00764561e-01 -9.22417581e-01 -8.35682213e-01 -3.75544518e-01
4.65583742e-01 9.61275697e-02 1.26130414e+00 -7.45225400e-02
-3.14064354e-01 4.23143357e-01 -6.83492064e-01 -3.95326436e-01
-3.34513277e-01 -2.69201458e-01 -8.41153786e-02 7.07632005e-02
3.33630800e-01 -8.42290640e-01 -8.02844226e-01 6.54608905e-01
-1.15816820e+00 -1.73514262e-01 7.93864191e-01 9.85623240e-01
7.35054731e-01 2.80788183e-01 1.05688834e+00 -6.72047615e-01
7.70685434e-01 -6.53353810e-01 -3.68027896e-01 3.22988778e-01
-1.04992747e+00 -7.27660954e-02 3.84776980e-01 -4.58138943e-01
-1.01053870e+00 -6.56325996e-01 -1.23122983e-01 -5.06995142e-01
8.64596590e-02 1.00360239e+00 -9.23519060e-02 3.76382828e-01
6.96658313e-01 5.78050852e-01 -2.07474396e-01 -3.04483414e-01
-9.69607905e-02 5.77544510e-01 7.67173707e-01 -4.48766023e-01
7.96555459e-01 3.57316196e-01 9.57853645e-02 -6.67511165e-01
-6.36195898e-01 -4.11511421e-01 -3.42902511e-01 -4.60288495e-01
6.72603250e-01 -7.93863297e-01 -3.63165230e-01 4.71940905e-01
-8.48468482e-01 2.21893191e-01 -3.26868355e-01 5.69444418e-01
-5.20989418e-01 3.11581850e-01 -8.54486302e-02 -1.13509393e+00
-3.40283871e-01 -7.99230635e-01 1.28347850e+00 1.10024624e-01
-2.09668372e-02 -1.11467707e+00 3.56382430e-01 4.94801626e-02
3.89575452e-01 8.60549867e-01 8.89231503e-01 -6.30222023e-01
-3.46415401e-01 -7.34513283e-01 -1.56625465e-01 6.26913846e-01
1.66843757e-01 1.44223392e-01 -1.08860457e+00 -2.08947062e-01
3.20475429e-01 -1.29208684e-01 5.71791530e-01 3.67499322e-01
9.42008257e-01 -1.33895874e-01 -6.50956109e-02 2.98658460e-01
1.50479245e+00 6.05282545e-01 1.07153416e+00 5.55275902e-02
4.64854449e-01 3.33204418e-01 7.96282589e-01 7.62652516e-01
1.90901130e-01 5.02340436e-01 4.43746090e-01 6.06240295e-02
-1.02840774e-01 -4.06043410e-01 2.22133920e-01 1.20504212e+00
-4.38179195e-01 -5.55114150e-01 -1.09652543e+00 5.71697295e-01
-1.74011636e+00 -1.03280842e+00 -2.70597458e-01 2.34343719e+00
5.90953648e-01 3.63333762e-01 -2.36975010e-02 7.31925368e-01
5.72621405e-01 -9.90974382e-02 -4.01818395e-01 1.74715385e-01
-3.03086072e-01 3.10283363e-01 -3.26071680e-02 -2.82882750e-01
-9.16179895e-01 2.04842746e-01 5.68836594e+00 1.16147149e+00
-1.10536313e+00 2.26555660e-01 7.72136271e-01 3.87911886e-01
-4.12080437e-01 -6.64974153e-02 -3.42331946e-01 8.28950584e-01
1.03743386e+00 -2.94044793e-01 -1.36709720e-01 7.82934844e-01
7.86866397e-02 -6.54162914e-02 -1.02788281e+00 1.05102074e+00
2.01936848e-02 -8.60846162e-01 1.23841958e-02 1.31398574e-01
7.97111988e-01 -3.04889917e-01 5.07925928e-01 3.98484290e-01
-1.29884645e-01 -1.10728133e+00 6.11009836e-01 7.77656972e-01
8.71059954e-01 -7.36809075e-01 1.10462224e+00 3.06027591e-01
-1.08320606e+00 2.76809722e-01 -1.97884262e-01 2.95361698e-01
1.42681673e-01 9.40900624e-01 -8.12953472e-01 1.09689581e+00
6.50275350e-01 8.01959276e-01 -6.13043129e-01 1.16650498e+00
6.49626106e-02 8.48319709e-01 -1.36850387e-01 1.14698365e-01
1.90027744e-01 -3.12476277e-01 4.05852705e-01 9.71981525e-01
8.44768107e-01 -1.10041007e-01 -1.22591212e-01 8.32368314e-01
3.36778879e-01 -5.12810275e-02 -4.27626491e-01 -2.39019945e-01
3.02739680e-01 1.17154944e+00 -7.76243508e-01 -1.09605409e-01
-2.01156378e-01 5.51863134e-01 -3.65136594e-01 3.74375969e-01
-1.09189546e+00 -3.15642267e-01 1.76803187e-01 1.00599959e-01
3.64653729e-02 -1.14579625e-01 -1.78725451e-01 -8.75844836e-01
2.18473837e-01 -9.15193558e-01 4.09887582e-01 -8.80296707e-01
-1.50117576e+00 8.73579264e-01 1.82905003e-01 -2.09006882e+00
-5.15405715e-01 -3.16858262e-01 -7.23433554e-01 8.54176998e-01
-1.15106165e+00 -1.22865915e+00 -5.56444466e-01 5.27395368e-01
5.27385354e-01 -2.07282886e-01 5.87529361e-01 6.23015642e-01
-3.86805058e-01 5.35284698e-01 1.24738127e-01 -3.03491745e-02
4.55688685e-01 -1.13163853e+00 1.41307905e-01 1.00327504e+00
1.99461102e-01 4.44711387e-01 9.52055812e-01 -4.56686139e-01
-1.23979962e+00 -1.12055063e+00 4.26920682e-01 -4.57635134e-01
4.85904098e-01 -5.21322004e-02 -9.99215126e-01 2.79907793e-01
2.27792896e-02 -2.59164311e-02 6.57530248e-01 -2.08763525e-01
-3.29466969e-01 -1.83776408e-01 -9.84671891e-01 1.31276175e-01
8.03258240e-01 -5.03160775e-01 -4.55695212e-01 2.87135452e-01
6.13770723e-01 -2.16399670e-01 -1.14356232e+00 6.64352715e-01
5.86342156e-01 -1.16076326e+00 7.57191539e-01 -2.96641469e-01
5.96043646e-01 -4.01014686e-01 -3.58211637e-01 -1.17758143e+00
-4.26080376e-02 -4.29903448e-01 -9.04426649e-02 1.35648859e+00
2.56566226e-01 -7.45661199e-01 4.09969777e-01 1.74694493e-01
-4.78799701e-01 -8.56206834e-01 -1.06117523e+00 -1.11925101e+00
-1.43786624e-01 -6.04790211e-01 7.11060226e-01 8.60267043e-01
-3.15227926e-01 1.68316379e-01 -4.07595664e-01 9.30155963e-02
7.03864336e-01 5.88592477e-02 5.31180382e-01 -1.33866620e+00
-2.28546456e-01 -2.59211093e-01 -8.27082574e-01 -4.43865806e-01
-4.17922527e-01 -4.91371870e-01 6.37917072e-02 -1.49943268e+00
1.41920373e-01 -5.19474089e-01 -5.61294675e-01 2.27303252e-01
-2.92021424e-01 3.76029015e-01 -2.04675913e-01 4.11840171e-01
-3.86331230e-01 9.20555234e-01 1.23708236e+00 -2.40660086e-02
2.21988797e-01 9.92547199e-02 -3.44030261e-01 6.13089502e-01
4.72050071e-01 -5.13983309e-01 -8.16661417e-01 -6.11953773e-02
3.61365169e-01 2.10405543e-01 5.73725641e-01 -1.35472167e+00
1.40233459e-02 5.57883009e-02 2.80705601e-01 -7.43823051e-01
1.11939721e-01 -6.45385087e-01 4.93827462e-01 5.08281767e-01
1.32062450e-01 4.54788268e-01 2.28830427e-03 7.38536358e-01
-7.37694800e-01 7.49525130e-02 5.36143720e-01 2.78898746e-01
-5.41983843e-01 3.40789795e-01 1.01393223e-01 1.65740624e-01
9.24620986e-01 -4.09871489e-01 -4.48598921e-01 -4.51250792e-01
-5.89134455e-01 -5.85821196e-02 2.50735462e-01 4.11287844e-01
8.30723763e-01 -1.57736671e+00 -8.03223073e-01 2.19948649e-01
4.73460943e-01 -8.91061723e-02 5.05291104e-01 1.11899126e+00
-2.72409379e-01 3.29570621e-01 -1.04649983e-01 -9.45339501e-01
-7.69504786e-01 4.24111724e-01 7.29240403e-02 -4.40791965e-01
-4.48494673e-01 5.86017847e-01 5.19901037e-01 -1.04986034e-01
-1.40747488e-01 -4.97634530e-01 -3.83117497e-02 6.76584318e-02
2.56613463e-01 5.00667393e-01 4.77766931e-01 -4.67927247e-01
-2.86487699e-01 3.63213211e-01 2.36405462e-01 -1.55944020e-01
1.35699487e+00 1.07228555e-01 2.03778073e-01 6.42202258e-01
9.71978009e-01 -4.14387584e-01 -1.18426371e+00 -4.73318845e-02
1.37786716e-02 -3.47211421e-01 -1.65650338e-01 -8.14186335e-01
-8.33369434e-01 9.13973987e-01 9.15580153e-01 6.93366349e-01
1.34802747e+00 -1.65439963e-01 6.45085394e-01 -1.05298333e-01
5.98719954e-01 -9.59629476e-01 4.13807809e-01 1.57971352e-01
8.35980773e-01 -1.08356845e+00 -1.23113893e-01 -2.43531942e-01
-6.30713105e-01 7.81158566e-01 2.63433099e-01 -8.73574391e-02
7.14387953e-01 1.01718888e-01 -1.38784781e-01 -2.48959139e-01
-7.40481615e-01 -1.09752305e-01 5.80814064e-01 6.96428061e-01
5.69361508e-01 4.40232754e-02 -2.49410599e-01 5.30233383e-01
-3.40574265e-01 1.32558504e-02 2.46674225e-01 9.51405048e-01
-2.02773437e-01 -8.65096033e-01 -3.64831597e-01 6.12528384e-01
-3.64614099e-01 1.66477695e-01 1.58114597e-01 9.08247232e-01
2.32476860e-01 9.36059296e-01 8.08454454e-02 -5.74257076e-01
4.58912253e-01 4.80562262e-02 3.61374229e-01 -3.19695920e-01
-2.06383660e-01 1.92052916e-01 -1.80279553e-01 -1.92903191e-01
-7.40240097e-01 -6.96702719e-01 -8.69285941e-01 -9.83570889e-02
-4.40132320e-01 1.66324332e-01 6.81543946e-01 1.01570964e+00
1.47286773e-01 8.56809556e-01 8.27024400e-01 -2.87431568e-01
-6.58355057e-01 -1.21434951e+00 -6.75957382e-01 5.07335067e-01
9.24535319e-02 -8.60099196e-01 -3.33180279e-01 2.69368351e-01] | [7.332810401916504, 2.5183444023132324] |
fbad6b40-6742-44d9-b52e-e000d6000b6f | automated-audio-captioning-with-recurrent | 1706.10006 | null | http://arxiv.org/abs/1706.10006v2 | http://arxiv.org/pdf/1706.10006v2.pdf | Automated Audio Captioning with Recurrent Neural Networks | We present the first approach to automated audio captioning. We employ an
encoder-decoder scheme with an alignment model in between. The input to the
encoder is a sequence of log mel-band energies calculated from an audio file,
while the output is a sequence of words, i.e. a caption. The encoder is a
multi-layered, bi-directional gated recurrent unit (GRU) and the decoder a
multi-layered GRU with a classification layer connected to the last GRU of the
decoder. The classification layer and the alignment model are fully connected
layers with shared weights between timesteps. The proposed method is evaluated
using data drawn from a commercial sound effects library, ProSound Effects. The
resulting captions were rated through metrics utilized in machine translation
and image captioning fields. Results from metrics show that the proposed method
can predict words appearing in the original caption, but not always correctly
ordered. | ['Sharath Adavanne', 'Tuomas Virtanen', 'Konstantinos Drossos'] | 2017-06-30 | null | null | null | null | ['audio-captioning'] | ['audio'] | [ 7.82040656e-01 2.76597738e-01 2.79558063e-01 -2.96862692e-01
-1.04703259e+00 -3.01630497e-01 3.38138074e-01 -7.19924364e-03
-1.66191474e-01 6.49673760e-01 7.25437462e-01 -7.25417361e-02
4.40752745e-01 -3.51798356e-01 -9.83491778e-01 -6.23864651e-01
-5.88922389e-02 3.01476985e-01 8.32951590e-02 -3.11727002e-02
3.26689094e-01 -2.14530006e-01 -1.66088533e+00 9.59566355e-01
2.88102627e-01 1.23482418e+00 7.02221751e-01 1.41130531e+00
3.96959819e-02 1.16693628e+00 -4.42963302e-01 -2.86398768e-01
-1.51434720e-01 -8.92634749e-01 -8.39283586e-01 -1.65369093e-01
1.94909170e-01 -5.78349009e-02 -3.03811193e-01 7.71538377e-01
8.34057570e-01 -1.71643179e-02 6.62231505e-01 -1.00966263e+00
-7.41416216e-01 1.01523387e+00 -6.19963668e-02 1.22266598e-01
6.60108566e-01 -2.23693684e-01 1.41385841e+00 -8.80843699e-01
4.48809832e-01 9.96947587e-01 6.06622040e-01 5.95901132e-01
-1.11457491e+00 -5.38502038e-01 -2.63617128e-01 3.06090772e-01
-1.20302379e+00 -5.84304333e-01 7.26690650e-01 -7.01870501e-01
1.08941901e+00 3.10749888e-01 5.28526366e-01 1.13314414e+00
3.32748294e-01 5.03068268e-01 8.27679217e-01 -6.09158278e-01
3.24942857e-01 1.07016571e-01 -5.72299510e-02 3.71421188e-01
-5.29918790e-01 5.17765470e-02 -8.16614628e-01 -2.00094491e-01
5.77210724e-01 -7.04493284e-01 -2.55752116e-01 8.35395530e-02
-1.08032835e+00 6.32464349e-01 3.18103641e-01 3.89696151e-01
-7.44897187e-01 6.21427774e-01 6.02245212e-01 3.07254225e-01
4.96529520e-01 2.45801777e-01 -2.45345622e-01 -3.73927087e-01
-1.09567332e+00 -2.14321576e-02 5.93200684e-01 6.90505385e-01
4.27620679e-01 8.93420726e-02 -3.61195117e-01 9.96905565e-01
4.24919873e-01 2.98835248e-01 7.85614371e-01 -4.91883755e-01
6.87922060e-01 -1.12436615e-01 1.43776044e-01 -8.92027915e-01
2.16481872e-02 -3.97763699e-01 -5.62252939e-01 -6.40441291e-03
-1.50929362e-01 -3.01966131e-01 -7.91913271e-01 1.82834911e+00
-3.82226914e-01 6.08986735e-01 2.02576175e-01 9.00915504e-01
7.92244554e-01 1.35265112e+00 1.83976084e-01 -2.56754667e-01
1.27132797e+00 -1.13020623e+00 -8.71154130e-01 -1.48838192e-01
4.61822838e-01 -1.13733733e+00 8.77524614e-01 3.45577925e-01
-1.43898284e+00 -8.19730699e-01 -1.05091989e+00 -7.21355900e-02
-2.30895191e-01 4.46092457e-01 3.32707986e-02 2.32350379e-01
-1.48726666e+00 5.77908695e-01 -3.53531122e-01 -3.06999460e-02
-1.32289410e-01 3.68414104e-01 -4.43052836e-02 6.01392329e-01
-1.45686245e+00 7.92609155e-01 5.52522957e-01 -9.30082873e-02
-1.06550825e+00 -3.17038149e-01 -8.02804947e-01 3.25088471e-01
-6.12200737e-01 -4.64554042e-01 1.73179185e+00 -1.53732479e+00
-1.68204784e+00 7.44935393e-01 -1.59903467e-01 -9.26453233e-01
2.00597420e-01 -1.75535589e-01 -5.08395672e-01 2.24554926e-01
-1.82858780e-01 1.02102959e+00 1.07963014e+00 -1.28090131e+00
-7.07540691e-01 2.46847197e-01 -3.49903464e-01 6.36063755e-01
-1.07042283e-01 1.35597080e-01 -6.61577284e-02 -8.13537836e-01
-1.78678736e-01 -8.29339445e-01 1.42894223e-01 -4.53541487e-01
-4.41179991e-01 1.19604520e-01 7.49293685e-01 -1.25436378e+00
1.57087088e+00 -2.25100541e+00 3.85607570e-01 -1.09131886e-02
-3.89080167e-01 6.67301789e-02 -3.08489442e-01 5.60797274e-01
-5.23273766e-01 -9.80199203e-02 -2.39522666e-01 -4.99618322e-01
-1.72865346e-01 -1.13838501e-01 -7.66001463e-01 3.96932662e-02
1.41038060e-01 6.79285824e-01 -9.80876207e-01 -3.18072915e-01
1.07266970e-01 7.37820864e-01 -5.28747976e-01 6.10348642e-01
-4.18140531e-01 1.49750635e-01 -1.03343666e-01 2.62074359e-02
1.06576212e-01 -1.29889399e-01 -2.17976749e-01 -1.10292509e-01
-3.37325454e-01 8.77894402e-01 -9.02054846e-01 1.79002392e+00
-8.21797669e-01 1.11327851e+00 -8.53500143e-02 -7.35323370e-01
1.05947018e+00 1.09348941e+00 1.61112174e-01 -5.57059824e-01
1.62262172e-01 3.88081342e-01 -1.39533579e-01 -6.10626221e-01
8.10029328e-01 -6.13551959e-02 3.24388370e-02 4.94874865e-01
2.52766609e-01 -1.24545783e-01 -9.68171731e-02 3.70184183e-02
9.22643006e-01 1.90546647e-01 1.84263229e-01 1.59829497e-01
5.43714941e-01 -2.58157969e-01 -7.72201270e-02 4.57535684e-01
3.65324438e-01 1.17804432e+00 3.68555754e-01 -2.09978089e-01
-1.63133752e+00 -7.81588495e-01 9.89251286e-02 1.33131289e+00
-4.32901204e-01 -3.50866735e-01 -1.05737340e+00 2.80970912e-02
-5.25430501e-01 8.79571795e-01 -5.75398147e-01 -1.09047480e-01
-3.16076636e-01 -3.73477727e-01 6.55222356e-01 3.09591323e-01
2.12425441e-01 -1.51296139e+00 -6.45409405e-01 6.47375524e-01
-5.58116972e-01 -9.31293309e-01 -6.11393571e-01 3.76206994e-01
-6.86715007e-01 -2.65065610e-01 -8.82191479e-01 -1.20134842e+00
3.93333077e-01 -2.74875522e-01 1.20973372e+00 -3.15517992e-01
1.36425365e-02 2.22196691e-02 -5.50892651e-01 -3.24065924e-01
-1.00904632e+00 1.26356080e-01 -8.13582242e-02 3.93764526e-01
-6.78719371e-04 -7.76469827e-01 -4.20235842e-01 4.18608412e-02
-7.03860223e-01 7.21588731e-01 7.39327610e-01 6.39472127e-01
6.07880652e-01 -6.07919753e-01 6.73222363e-01 -4.35163677e-01
7.50677645e-01 -5.83950877e-01 -3.41816574e-01 3.99179459e-02
-3.01951636e-02 1.91068009e-01 6.81410015e-01 -5.22453070e-01
-7.03708172e-01 3.97734433e-01 -3.52850914e-01 -2.28046998e-01
-5.46746813e-02 5.10739028e-01 -1.04582077e-02 3.00906569e-01
7.76291728e-01 3.42406631e-01 -3.90318096e-01 -4.91524637e-01
4.42838371e-01 1.36428559e+00 8.60682487e-01 -2.60466218e-01
2.85721242e-01 -1.89337581e-01 -5.97696304e-01 -6.87061489e-01
-6.01518035e-01 -2.92216033e-01 -4.40355152e-01 -5.85756540e-01
1.12106597e+00 -1.07996249e+00 -7.82677066e-03 3.98738295e-01
-1.63172996e+00 -1.56479046e-01 -2.94273704e-01 5.83841383e-01
-9.09414828e-01 -1.80613384e-01 -7.55009472e-01 -9.08274174e-01
-6.54520273e-01 -1.04221833e+00 1.18817008e+00 2.07977235e-01
-4.42408353e-01 -7.06094146e-01 4.22290146e-01 3.87698144e-01
3.63570631e-01 7.80154914e-02 9.86448944e-01 -8.18499029e-01
-2.29097143e-01 -1.17183449e-02 3.02551892e-02 6.10920668e-01
2.88860872e-02 -1.30444720e-01 -1.52490222e+00 2.61924490e-02
-9.39275250e-02 -5.68253219e-01 6.86347365e-01 4.66566920e-01
1.10015118e+00 -5.11215150e-01 7.64998794e-02 1.49814576e-01
1.19943571e+00 5.93621254e-01 8.89740586e-01 7.12263510e-02
5.14013350e-01 3.62768590e-01 1.79783866e-01 2.70800471e-01
4.44550104e-02 8.36516201e-01 3.62440914e-01 -1.22094430e-01
-4.17183608e-01 -5.83837092e-01 9.16388988e-01 1.58258760e+00
1.37452051e-01 -4.62562233e-01 -1.00315607e+00 7.13572741e-01
-1.66208184e+00 -1.06474793e+00 -1.42827615e-01 2.21375465e+00
1.10024250e+00 1.25864610e-01 -1.05333634e-01 4.18837965e-01
1.02239680e+00 5.67398891e-02 -2.62232900e-01 -1.15237772e+00
1.00180537e-01 1.93535954e-01 1.35856509e-01 8.09033811e-01
-8.83079708e-01 7.95916259e-01 6.29631901e+00 7.81803608e-01
-1.21098971e+00 2.05655232e-01 7.30158389e-01 -1.85271148e-02
-2.07993537e-01 -3.04489881e-01 -3.34117144e-01 7.21291721e-01
1.61752391e+00 -7.39668757e-02 6.12673700e-01 6.77205622e-01
4.99528736e-01 6.21008240e-02 -1.18595791e+00 1.08827460e+00
2.44282424e-01 -1.22294474e+00 2.84537166e-01 -2.85831392e-01
7.47384608e-01 8.25989842e-02 3.30601692e-01 3.00742220e-02
-7.29949102e-02 -1.21309900e+00 1.02741170e+00 5.57831109e-01
9.73524272e-01 -7.81717122e-01 6.87993824e-01 9.66099948e-02
-1.31385970e+00 1.71170905e-01 -2.58833855e-01 -2.76219070e-01
2.34394193e-01 2.92180806e-01 -1.29326010e+00 3.10512751e-01
4.34351534e-01 4.54883248e-01 -1.68253496e-01 1.15198660e+00
-1.55894995e-01 9.54170465e-01 -4.10189927e-02 -5.97904585e-02
3.87877017e-01 -1.93546545e-02 6.00138366e-01 1.67807400e+00
5.08739591e-01 -1.37759462e-01 -3.50082606e-01 4.58601952e-01
-3.05936962e-01 2.83372313e-01 -5.75829625e-01 -2.68266708e-01
3.46475989e-01 1.04232442e+00 -4.14728731e-01 -3.45112294e-01
-2.14401290e-01 1.15271306e+00 7.58942729e-03 2.84138560e-01
-9.45070267e-01 -4.85695332e-01 3.87250334e-01 2.46871542e-02
2.48620778e-01 1.17684826e-01 -2.17706457e-01 -6.65587723e-01
-9.68498737e-02 -7.99729407e-01 7.09635094e-02 -1.55833352e+00
-9.02754366e-01 1.09182394e+00 -2.60953665e-01 -1.40133202e+00
-7.02541530e-01 -3.47685039e-01 -5.55056095e-01 1.16587162e+00
-1.13827085e+00 -8.75122547e-01 4.12717499e-02 2.77430475e-01
7.03550756e-01 -3.22606325e-01 1.04491961e+00 4.98455256e-01
-7.50181824e-02 3.24873596e-01 5.76077141e-02 1.24251664e-01
4.74242091e-01 -1.09431756e+00 5.46310186e-01 5.86067200e-01
5.33993185e-01 3.90627049e-02 1.03689373e+00 -5.46952784e-01
-6.93731844e-01 -1.24297297e+00 1.49509847e+00 6.02235496e-02
6.58000231e-01 -6.52669191e-01 -9.22815800e-01 4.49927628e-01
6.97005212e-01 -5.14992237e-01 7.59098172e-01 -4.33344275e-01
-6.67970031e-02 1.72583833e-01 -7.62991488e-01 3.62058580e-01
5.10232449e-01 -1.00949299e+00 -9.13555741e-01 3.89186978e-01
9.73019660e-01 -6.20688617e-01 -4.34572369e-01 -1.39838398e-01
5.68644047e-01 -6.16995394e-01 6.81804299e-01 -4.13918853e-01
9.71038461e-01 -2.69506037e-01 -3.07032198e-01 -1.29563260e+00
-2.50356533e-02 -7.28442013e-01 -5.92535064e-02 1.17026162e+00
9.15604353e-01 1.52416185e-01 4.79726225e-01 -6.01449981e-02
-3.44817817e-01 -6.55446470e-01 -1.08706260e+00 -3.47805470e-01
-4.13766712e-01 -6.71502352e-01 1.98250845e-01 4.96056259e-01
-9.48761180e-02 8.22235942e-01 -8.88549805e-01 1.95562094e-01
2.18871206e-01 -3.44519585e-01 2.07632389e-02 -9.94565010e-01
-5.25984228e-01 -1.99453175e-01 -4.25940573e-01 -1.00744796e+00
-1.09826542e-01 -9.54249144e-01 4.98823851e-01 -1.62426317e+00
1.58404425e-01 -1.71405599e-02 -3.73001516e-01 4.47620183e-01
2.24274009e-01 4.11393970e-01 7.48245940e-02 1.42861083e-01
-2.37983584e-01 5.23156345e-01 6.32737219e-01 -8.56452212e-02
-3.40153724e-01 1.80289075e-02 -5.33592366e-02 5.97556651e-01
8.43576550e-01 -7.16879547e-01 -4.72070456e-01 -5.10543346e-01
3.94282192e-01 4.79821146e-01 4.04106647e-01 -1.36951840e+00
3.04258257e-01 4.89293039e-01 1.80979416e-01 -5.99605024e-01
4.82318580e-01 -6.72395468e-01 5.06039798e-01 2.93749899e-01
-1.01554644e+00 3.56577814e-01 1.85684919e-01 2.73060799e-01
-5.48167884e-01 -4.45181280e-01 7.36305416e-01 1.52350888e-01
-3.20496380e-01 -4.29033972e-02 -5.70425153e-01 -2.50589311e-01
6.07743561e-01 1.42624043e-02 1.48006484e-01 -8.32794785e-01
-9.13658202e-01 -2.87446380e-01 9.04812217e-02 6.97632968e-01
6.80779338e-01 -1.60964215e+00 -9.30589736e-01 6.03618100e-02
-1.85560316e-01 -3.77683222e-01 -2.84140538e-02 3.12566012e-01
-5.13727188e-01 4.35163021e-01 -3.41334701e-01 -4.13160563e-01
-1.31812024e+00 8.43178853e-02 5.21881998e-01 -2.06259042e-01
-3.70303929e-01 1.02122581e+00 3.82112265e-02 1.55995637e-01
4.75372344e-01 -4.09506142e-01 -5.30402482e-01 2.68623680e-01
6.70300722e-01 -6.37358204e-02 1.86930582e-01 -9.99745369e-01
2.34189183e-02 3.06444049e-01 1.44198492e-01 -9.65607285e-01
1.28915203e+00 -7.63105974e-02 -6.57214820e-02 1.02305651e+00
1.29816377e+00 -3.51259500e-01 -1.04437399e+00 4.10551727e-02
-3.25435735e-02 1.02157049e-01 1.12206079e-01 -9.77091789e-01
-7.09201455e-01 1.15730369e+00 8.39820862e-01 3.78683180e-01
1.24122572e+00 -1.65772706e-01 7.97296822e-01 1.19932286e-01
1.12509698e-01 -1.17718649e+00 1.47074148e-01 6.82613969e-01
1.17151761e+00 -4.91304636e-01 -6.31173670e-01 1.82167470e-01
-9.23741639e-01 1.01328266e+00 1.47933632e-01 -9.08159241e-02
2.27033794e-01 3.28674853e-01 1.38166592e-01 2.48484895e-01
-1.07796907e+00 1.55044213e-01 4.02775586e-01 4.16578978e-01
7.75020301e-01 -4.57649641e-02 -1.33829355e-01 6.34736776e-01
-6.42419457e-01 -4.46409099e-02 4.81314331e-01 4.32233006e-01
-6.16490602e-01 -9.35895860e-01 -3.91426593e-01 2.45151460e-01
-5.61220706e-01 -5.08995295e-01 -4.77892071e-01 5.61211584e-03
1.20217979e-01 9.30098891e-01 2.96232641e-01 -7.76656866e-01
2.25035027e-01 5.63819230e-01 -1.41081838e-02 -6.12424850e-01
-9.81092393e-01 4.82031666e-02 1.21733554e-01 -1.90158099e-01
-2.10086003e-01 -4.33939129e-01 -1.25554013e+00 4.18094009e-01
-1.88859403e-01 6.97543919e-01 1.08446825e+00 6.58617914e-01
3.28099817e-01 8.41968596e-01 9.83524382e-01 -1.02357733e+00
-2.19773576e-01 -1.30213761e+00 -1.60024732e-01 1.76130444e-01
5.79443097e-01 -1.71445291e-02 -3.10094357e-01 8.13862681e-01] | [15.281448364257812, 4.9098052978515625] |
700c7328-763f-4a86-986a-8959293647e8 | debiased-pseudo-labeling-in-self-training | 2202.07136 | null | https://arxiv.org/abs/2202.07136v5 | https://arxiv.org/pdf/2202.07136v5.pdf | Debiased Self-Training for Semi-Supervised Learning | Deep neural networks achieve remarkable performances on a wide range of tasks with the aid of large-scale labeled datasets. Yet these datasets are time-consuming and labor-exhaustive to obtain on realistic tasks. To mitigate the requirement for labeled data, self-training is widely used in semi-supervised learning by iteratively assigning pseudo labels to unlabeled samples. Despite its popularity, self-training is well-believed to be unreliable and often leads to training instability. Our experimental studies further reveal that the bias in semi-supervised learning arises from both the problem itself and the inappropriate training with potentially incorrect pseudo labels, which accumulates the error in the iterative self-training process. To reduce the above bias, we propose Debiased Self-Training (DST). First, the generation and utilization of pseudo labels are decoupled by two parameter-independent classifier heads to avoid direct error accumulation. Second, we estimate the worst case of self-training bias, where the pseudo labeling function is accurate on labeled samples, yet makes as many mistakes as possible on unlabeled samples. We then adversarially optimize the representations to improve the quality of pseudo labels by avoiding the worst case. Extensive experiments justify that DST achieves an average improvement of 6.3% against state-of-the-art methods on standard semi-supervised learning benchmark datasets and 18.9%$ against FixMatch on 13 diverse tasks. Furthermore, DST can be seamlessly adapted to other self-training methods and help stabilize their training and balance performance across classes in both cases of training from scratch and finetuning from pre-trained models. | ['Mingsheng Long', 'Jianmin Wang', 'Pengfei Wan', 'Ximei Wang', 'Junguang Jiang', 'Baixu Chen'] | 2022-02-15 | null | null | null | null | ['semi-supervised-image-classification', 'texture-classification'] | ['computer-vision', 'computer-vision'] | [ 4.49602813e-01 1.72065943e-01 -3.33957404e-01 -7.13163316e-01
-7.47690380e-01 -6.47047162e-01 3.52234930e-01 -1.90074280e-01
-6.20794594e-01 9.75877464e-01 -4.00678068e-01 -3.01295161e-01
2.65082508e-01 -6.19873047e-01 -9.95838165e-01 -8.21080923e-01
3.29535395e-01 5.75543880e-01 6.76612630e-02 -6.22990308e-03
-9.82849672e-02 2.21028149e-01 -1.51148760e+00 2.10898340e-01
1.20183420e+00 1.09490645e+00 -3.02154906e-02 1.31715983e-01
-1.26209378e-01 8.31756115e-01 -8.40162933e-01 -4.98712063e-01
4.24080372e-01 -3.52914244e-01 -7.22258985e-01 3.58798653e-01
6.02108896e-01 -2.88330466e-01 -1.10818431e-01 1.26783741e+00
5.22256255e-01 -4.32910509e-02 5.76138258e-01 -1.37151122e+00
-7.80701518e-01 7.23081648e-01 -5.74912012e-01 -9.07019600e-02
-4.24029648e-01 3.22788209e-01 6.05662823e-01 -9.36351001e-01
4.04492736e-01 8.96595776e-01 7.74017990e-01 9.16334808e-01
-1.33132970e+00 -1.13430536e+00 1.15790637e-02 -1.15982257e-01
-1.44535434e+00 -6.16327226e-01 8.03446233e-01 -3.67869288e-01
4.90181893e-01 2.31086500e-02 1.20216943e-01 1.47638953e+00
-1.10626593e-01 8.37833166e-01 1.39957702e+00 -3.84615809e-01
4.30270314e-01 5.66339672e-01 2.11321920e-01 6.13631487e-01
3.78974736e-01 3.03505808e-01 -2.82529145e-01 -6.12096600e-02
5.77768266e-01 3.19676548e-02 -1.11297369e-01 -4.76062804e-01
-1.11378992e+00 7.47683406e-01 5.72145939e-01 4.97108363e-02
-1.25765249e-01 -2.24016029e-02 5.26442707e-01 5.54739416e-01
5.83536983e-01 6.86475277e-01 -7.14244843e-01 1.67248979e-01
-1.06074893e+00 -4.97017093e-02 5.08339942e-01 1.08165634e+00
8.51689756e-01 4.01992977e-01 -1.69819579e-01 1.10223448e+00
1.35783281e-03 3.92268836e-01 8.12359214e-01 -7.66469240e-01
4.70083028e-01 7.44583428e-01 1.71171710e-01 -5.89970827e-01
-2.22196430e-01 -9.85613644e-01 -1.09485459e+00 2.59427071e-01
7.37266898e-01 -2.98610747e-01 -1.15918791e+00 1.94643307e+00
3.35175663e-01 1.63030282e-01 3.41488533e-02 8.15699756e-01
6.52930379e-01 2.80536741e-01 1.88175678e-01 -3.11615676e-01
8.79505634e-01 -1.38544679e+00 -6.25788629e-01 -6.01230621e-01
8.62067759e-01 -4.12270635e-01 1.23429155e+00 4.19935554e-01
-8.13574076e-01 -7.12738752e-01 -1.24398518e+00 3.04621041e-01
-3.67663085e-01 3.55719298e-01 5.02045512e-01 7.71334112e-01
-7.46580780e-01 7.96313941e-01 -6.91911399e-01 1.19027212e-01
8.16303372e-01 4.58881795e-01 -4.20489460e-01 -3.09492573e-02
-1.23813319e+00 7.07853496e-01 4.78266686e-01 9.53852609e-02
-9.92951572e-01 -8.50299776e-01 -7.10708916e-01 -1.32623345e-01
5.39692521e-01 -3.26564848e-01 1.29235327e+00 -1.52283835e+00
-1.37947011e+00 1.08407807e+00 1.89348999e-02 -5.26946545e-01
9.33112979e-01 -2.56395191e-01 -4.80751961e-01 -1.55801490e-01
2.06768468e-01 7.91590393e-01 1.14830363e+00 -1.52950490e+00
-4.20838654e-01 -2.10086465e-01 -2.24476650e-01 1.60771489e-01
-5.88841259e-01 -2.80753821e-01 -2.44118646e-01 -7.15672314e-01
1.27619132e-01 -1.14118636e+00 -2.49943286e-01 -1.25335425e-01
-5.98413348e-01 -1.40901417e-01 7.87353814e-01 -4.17957380e-02
1.10225868e+00 -2.22709179e+00 -3.71011138e-01 -2.63840351e-02
3.33571017e-01 7.43953228e-01 -1.12008125e-01 -1.35465756e-01
-3.57929528e-01 3.13948333e-01 -3.63557726e-01 -5.73741436e-01
-1.66579619e-01 3.40779185e-01 -5.44560850e-01 5.00947416e-01
2.90232331e-01 8.72705817e-01 -1.04913056e+00 -5.17190576e-01
-1.52963419e-02 1.92394614e-01 -2.59259552e-01 3.54250491e-01
-1.75469816e-01 5.67379057e-01 -2.61737168e-01 7.36732602e-01
6.98929906e-01 -6.04432881e-01 9.84191895e-03 -9.51907188e-02
4.03187156e-01 1.55664101e-01 -1.16622162e+00 1.44688690e+00
-4.79949087e-01 2.85133272e-01 -1.88697800e-01 -1.07956100e+00
9.29169178e-01 1.98101193e-01 1.64688185e-01 -4.82486844e-01
3.45412642e-01 3.45132262e-01 -1.42678156e-01 -2.06446588e-01
2.74290979e-01 -1.16834745e-01 1.27474874e-01 7.14879870e-01
2.11772799e-01 6.36174157e-02 8.33761841e-02 -4.60016094e-02
8.14637005e-01 1.20321453e-01 1.58467039e-01 -2.60515362e-01
3.26307893e-01 9.28717330e-02 7.86166370e-01 8.06814253e-01
-5.24435103e-01 5.84022582e-01 1.26784250e-01 -6.29002392e-01
-9.28983867e-01 -9.20066416e-01 -3.57017130e-01 1.15530992e+00
1.19651698e-01 5.05261384e-02 -8.62906039e-01 -1.36013901e+00
6.89518079e-02 8.18155944e-01 -7.63864934e-01 -5.82212448e-01
-3.80702883e-01 -8.64446580e-01 6.54966712e-01 6.51814222e-01
6.24872386e-01 -1.16386735e+00 -2.82116205e-01 1.84454694e-01
1.29720997e-02 -1.02542877e+00 -3.98059040e-01 7.32580721e-01
-9.57611561e-01 -9.67749417e-01 -5.45229852e-01 -7.61831403e-01
1.23313022e+00 3.48207474e-01 1.27712739e+00 1.80202305e-01
7.06197843e-02 -3.29586148e-01 -3.18342566e-01 -3.62521112e-01
-6.95401549e-01 2.00353265e-01 3.94960076e-01 1.68755859e-01
3.63764256e-01 -4.70702350e-01 -5.22503912e-01 8.37209284e-01
-9.24344420e-01 -4.06399481e-02 4.97219354e-01 1.33968544e+00
7.23076642e-01 1.90806642e-01 9.32864904e-01 -1.47844958e+00
2.48778835e-01 -6.18847609e-01 -5.58122039e-01 2.80161887e-01
-1.07370710e+00 1.67535782e-01 1.16554773e+00 -8.63749862e-01
-9.51412737e-01 1.94863364e-01 6.92382753e-02 -7.96986401e-01
-2.81572223e-01 4.81077135e-02 -2.27903262e-01 -1.46406755e-01
1.12083173e+00 2.10824519e-01 -9.75538418e-03 -2.40122050e-01
2.75979459e-01 7.39211500e-01 3.94954979e-01 -5.84206760e-01
9.43116307e-01 3.89224291e-01 -2.81069547e-01 -2.57648677e-01
-1.47780359e+00 -1.54215753e-01 -4.85759705e-01 -5.87126277e-02
2.41922870e-01 -1.02988636e+00 -2.10541949e-01 7.12046087e-01
-5.74306011e-01 -6.65197551e-01 -4.82746720e-01 9.20961499e-02
-2.64157504e-01 2.33993456e-01 -5.35186887e-01 -6.14636183e-01
-4.49736208e-01 -1.31080246e+00 8.04827034e-01 3.44446868e-01
-1.45839781e-01 -9.58491504e-01 -2.52020657e-01 5.09392083e-01
3.36962759e-01 1.29040122e-01 5.26297688e-01 -1.05901039e+00
-2.00618312e-01 -3.07947665e-01 -2.19511092e-01 8.45813870e-01
2.24964738e-01 -1.89181238e-01 -1.21271122e+00 -5.93784750e-01
-5.94317354e-02 -1.08321929e+00 7.78717160e-01 1.77067704e-02
1.34716678e+00 -2.48770446e-01 -2.96019167e-01 7.98895597e-01
1.29529095e+00 2.94644870e-02 3.52559328e-01 4.40919399e-01
7.52458096e-01 5.00905335e-01 7.95507967e-01 1.60098359e-01
1.28283679e-01 2.98581123e-01 5.02464652e-01 -1.34664536e-01
-9.97288153e-02 -3.49296510e-01 1.32279813e-01 5.57058036e-01
4.46058154e-01 -2.28594914e-01 -8.79318357e-01 4.43995535e-01
-1.69172335e+00 -6.28571332e-01 -1.04127759e-02 2.38606071e+00
1.30163348e+00 4.57972020e-01 -1.49258271e-01 3.69012833e-01
8.64440560e-01 -2.73438543e-02 -1.00674367e+00 -6.42146319e-02
-1.14814758e-01 1.32245362e-01 7.91435838e-01 5.45105189e-02
-1.34753478e+00 1.12980998e+00 6.29450130e+00 9.51771498e-01
-1.28312218e+00 2.61024833e-01 1.09874761e+00 -4.12407331e-02
3.60769331e-02 -2.53059089e-01 -9.88614917e-01 6.60746634e-01
8.55971992e-01 2.04709828e-01 3.27548891e-01 1.21798515e+00
-1.55019343e-01 9.82793644e-02 -1.16639030e+00 8.14586997e-01
-1.08813882e-01 -1.06684268e+00 -1.33418396e-01 -2.38341421e-01
1.12594104e+00 6.43643886e-02 1.71935692e-01 5.44415534e-01
6.71611845e-01 -9.34766829e-01 8.77781630e-01 -2.99724396e-02
1.20530057e+00 -7.04688430e-01 6.99555516e-01 6.52794242e-01
-6.68480217e-01 -1.45188510e-01 -3.39449704e-01 1.26014590e-01
-2.23079994e-01 8.53170693e-01 -8.65306437e-01 2.08337680e-01
5.08760929e-01 4.30956095e-01 -7.11083591e-01 6.32174611e-01
-5.43163776e-01 9.83872414e-01 -2.39067420e-01 1.63537756e-01
1.62758768e-01 8.15457776e-02 1.30533129e-01 9.64807272e-01
-1.14532709e-01 -3.05386811e-01 1.44641250e-01 9.15386081e-01
-3.68893027e-01 -1.22553565e-01 -4.34076071e-01 -1.09608592e-02
8.06853890e-01 1.20558405e+00 -6.55757904e-01 -4.38846409e-01
-9.80404317e-02 9.42055285e-01 6.92700505e-01 3.44347000e-01
-9.48742151e-01 -2.13873833e-01 3.34806681e-01 -1.76699657e-03
2.97234859e-02 1.50847286e-01 -6.13056660e-01 -1.15618432e+00
7.58230388e-02 -1.01906979e+00 2.60266483e-01 -4.72842127e-01
-1.49490905e+00 9.05007541e-01 -3.42844069e-01 -1.43517876e+00
-2.88201034e-01 -4.01202619e-01 -4.11041558e-01 6.45904481e-01
-1.66304636e+00 -8.32078934e-01 -2.90418178e-01 4.75991100e-01
5.39576769e-01 -3.56871933e-01 8.34450364e-01 3.17963660e-01
-9.17759955e-01 1.23901212e+00 2.45806187e-01 1.85678259e-01
1.15994203e+00 -1.35196054e+00 4.34515983e-01 8.13838661e-01
1.10884711e-01 6.00947261e-01 5.74358046e-01 -5.73849082e-01
-1.03869903e+00 -1.50124311e+00 5.95309556e-01 -4.14912492e-01
5.73275268e-01 -3.73134315e-01 -1.06854033e+00 6.04107440e-01
-2.05398440e-01 6.15054250e-01 7.76632130e-01 -9.28068608e-02
-7.24440575e-01 -2.89334327e-01 -1.27611208e+00 5.00355124e-01
1.05779326e+00 -3.86106461e-01 -4.58708942e-01 5.99484444e-01
7.23950446e-01 -5.22144437e-01 -6.36432648e-01 5.65918803e-01
2.01060265e-01 -9.76013005e-01 8.22344720e-01 -4.85976309e-01
4.71284300e-01 -2.56492376e-01 1.50735885e-01 -1.50822330e+00
-1.66216299e-01 -5.70306063e-01 -9.23783556e-02 1.36135423e+00
5.42718828e-01 -6.44163787e-01 1.15075731e+00 6.21609747e-01
-1.20948456e-01 -9.30049956e-01 -6.15288734e-01 -9.63775635e-01
9.02684405e-02 -2.47489065e-01 5.63516080e-01 1.41911697e+00
-3.42785716e-01 3.00841242e-01 -4.16437298e-01 -1.43011995e-02
8.90415192e-01 -4.11756802e-03 7.91834593e-01 -1.16420054e+00
-2.34493956e-01 -1.72964111e-01 1.11443512e-01 -9.32172477e-01
5.03265500e-01 -8.32708776e-01 2.69869626e-01 -8.66953135e-01
-2.71996465e-02 -1.02327085e+00 -4.93238300e-01 8.16062689e-01
-5.49421310e-01 4.28550869e-01 -1.68238014e-01 4.29125816e-01
-6.78263605e-01 4.06138450e-01 1.35483611e+00 -1.26367107e-01
-2.23094463e-01 6.27990887e-02 -7.90724158e-01 6.48543358e-01
9.35939550e-01 -7.83137262e-01 -6.55469120e-01 -4.37303931e-01
5.96534126e-02 -2.50864387e-01 6.83185086e-02 -9.74355459e-01
1.33529752e-01 -1.82108939e-01 3.78080398e-01 -2.40266353e-01
1.54544003e-02 -9.25347090e-01 -4.84947450e-02 3.42775047e-01
-5.15056312e-01 -2.97293842e-01 3.29992399e-02 5.80515802e-01
-1.71875075e-01 -4.18256074e-01 1.17630351e+00 -1.06659427e-01
-5.56840241e-01 3.04459572e-01 1.05050899e-01 3.27213496e-01
1.12834537e+00 -2.74789840e-01 -4.75823998e-01 -3.90321463e-02
-5.36990404e-01 3.83825958e-01 5.47385812e-01 3.54051322e-01
3.26337010e-01 -1.26356900e+00 -4.43154722e-01 4.15843785e-01
2.61202097e-01 4.20918643e-01 1.89563125e-01 4.64641631e-01
-2.40870833e-01 6.38655154e-03 -1.91605419e-01 -5.88852644e-01
-8.45479727e-01 7.73185790e-01 3.31762344e-01 -4.57862318e-01
-2.90270954e-01 1.08694971e+00 1.22808419e-01 -6.67949021e-01
4.86539721e-01 2.22343840e-02 5.90642653e-02 -2.66239028e-02
4.55780119e-01 2.08411649e-01 2.18570545e-01 -4.44891930e-01
-1.44379094e-01 1.55389756e-01 -4.00604725e-01 3.70491326e-01
1.07880950e+00 2.21475288e-02 2.84399897e-01 4.44803387e-01
1.23601258e+00 -2.79565245e-01 -1.63958418e+00 -5.24934649e-01
-1.27286866e-01 -2.27394342e-01 6.79762214e-02 -9.86679733e-01
-1.32974219e+00 7.32277930e-01 4.34149057e-01 9.02796760e-02
1.07652581e+00 -4.34607178e-01 8.24547827e-01 5.44151425e-01
4.35957372e-01 -1.16065454e+00 2.21650124e-01 3.26867074e-01
5.10887325e-01 -1.68271041e+00 -3.99471894e-02 -4.03707922e-01
-8.83808136e-01 8.03750992e-01 1.03485966e+00 -1.84018165e-01
4.43267047e-01 2.96637833e-01 4.16814804e-01 1.55666307e-01
-7.07033098e-01 2.73760140e-01 1.89427465e-01 3.81733418e-01
1.49232835e-01 3.35330255e-02 5.10090925e-02 7.50941813e-01
-7.64020309e-02 7.04018548e-02 2.74517894e-01 1.01306117e+00
-1.78260311e-01 -1.09200549e+00 -3.18712920e-01 5.51683605e-01
-4.13823336e-01 -1.19447961e-01 -2.27995872e-01 6.68456435e-01
2.22031817e-01 8.92249644e-01 -1.58978552e-01 -6.04906678e-01
1.99923202e-01 2.50610441e-01 1.16453469e-01 -6.44077539e-01
-7.32609808e-01 -2.08245948e-01 -2.10758727e-02 -4.05986190e-01
-3.22042644e-01 -3.42903197e-01 -1.12150502e+00 -5.53150699e-02
-6.30090535e-01 1.83870688e-01 5.64785540e-01 9.72700655e-01
4.05527651e-01 5.52230835e-01 9.96494114e-01 -6.56596243e-01
-1.27102709e+00 -9.94410694e-01 -4.77301151e-01 6.43602133e-01
2.20164075e-01 -8.94791305e-01 -6.95532262e-01 7.74950013e-02] | [9.460481643676758, 3.6106793880462646] |
266337a6-c385-452d-bd71-39eac3730042 | enhanced-meta-learning-for-cross-lingual | 1911.06161 | null | https://arxiv.org/abs/1911.06161v2 | https://arxiv.org/pdf/1911.06161v2.pdf | Enhanced Meta-Learning for Cross-lingual Named Entity Recognition with Minimal Resources | For languages with no annotated resources, transferring knowledge from rich-resource languages is an effective solution for named entity recognition (NER). While all existing methods directly transfer from source-learned model to a target language, in this paper, we propose to fine-tune the learned model with a few similar examples given a test case, which could benefit the prediction by leveraging the structural and semantic information conveyed in such similar examples. To this end, we present a meta-learning algorithm to find a good model parameter initialization that could fast adapt to the given test case and propose to construct multiple pseudo-NER tasks for meta-training by computing sentence similarities. To further improve the model's generalization ability across different languages, we introduce a masking scheme and augment the loss function with an additional maximum term during meta-training. We conduct extensive experiments on cross-lingual named entity recognition with minimal resources over five target languages. The results show that our approach significantly outperforms existing state-of-the-art methods across the board. | ['Chin-Yew Lin', 'Börje F. Karlsson', 'Qianhui Wu', 'Hui Chen', 'Zijia Lin', 'Guoxin Wang', 'Biqing Huang'] | 2019-11-14 | null | null | null | null | ['cross-lingual-ner'] | ['natural-language-processing'] | [ 3.04956716e-02 -1.51849777e-01 -2.95988828e-01 -8.72505426e-01
-1.08258522e+00 -5.88119864e-01 3.46786439e-01 1.05299696e-01
-8.05083990e-01 7.27248430e-01 2.70969361e-01 -1.18929371e-01
2.66063124e-01 -5.50429642e-01 -6.71292067e-01 -1.57417893e-01
8.12764317e-02 4.47172731e-01 6.23982251e-02 -2.79722691e-01
5.88099025e-02 3.37993056e-01 -9.88123059e-01 7.48326600e-01
1.28787005e+00 6.11138821e-01 4.36569422e-01 1.08914457e-01
-6.67143881e-01 6.69380546e-01 -6.96410716e-01 -8.68517399e-01
6.72965720e-02 -4.54044819e-01 -1.07324219e+00 -3.79873402e-02
2.44839832e-01 2.72441000e-01 9.47656706e-02 9.01338458e-01
6.38894618e-01 3.51267129e-01 5.00847995e-01 -7.29205728e-01
-7.75645554e-01 1.09268749e+00 -2.42787957e-01 1.61337987e-01
8.03215429e-02 -2.32198104e-01 8.13250780e-01 -1.07923925e+00
6.90955222e-01 1.15943968e+00 8.97823572e-01 8.87067139e-01
-9.97685432e-01 -7.64586985e-01 3.94466281e-01 1.39993191e-01
-1.42265022e+00 -6.08977020e-01 7.47747421e-01 1.73212990e-01
1.21022987e+00 -6.20150231e-02 -3.07283998e-02 1.08588624e+00
-3.58484447e-01 9.65363264e-01 1.06446731e+00 -6.94210172e-01
2.68439557e-02 5.40206254e-01 -1.33709861e-02 5.18564522e-01
9.91951674e-02 -2.73317397e-01 -3.49099606e-01 -1.88612267e-01
2.20719650e-01 -2.17054918e-01 -2.20982954e-01 -9.55127031e-02
-1.20502090e+00 5.75595975e-01 4.40534800e-01 7.06408799e-01
-2.90496647e-01 -2.38840938e-01 5.60907900e-01 3.99878263e-01
6.02515161e-01 7.10584402e-01 -1.13016868e+00 4.64902520e-02
-8.82582605e-01 -4.08733338e-01 1.03018427e+00 1.04367936e+00
9.09417093e-01 1.55284088e-02 -1.67387411e-01 1.42238784e+00
5.32869138e-02 4.10805047e-01 8.63609076e-01 -3.52760762e-01
1.01406813e+00 6.52120173e-01 -7.69180954e-02 -4.10295606e-01
-2.45237365e-01 -6.14767075e-01 -7.31404305e-01 -5.21405041e-01
9.79795903e-02 -4.23046917e-01 -9.34126914e-01 1.80900764e+00
4.05044854e-01 3.08202237e-01 5.18840432e-01 4.24990743e-01
7.31405735e-01 7.22514153e-01 4.40588802e-01 -7.53186122e-02
1.22341406e+00 -1.33477068e+00 -5.09032369e-01 -6.35917306e-01
1.15419877e+00 -7.13200033e-01 1.10950577e+00 -7.14248270e-02
-6.59670830e-01 -7.20798373e-01 -7.50426114e-01 7.69164190e-02
-6.10829055e-01 6.08783126e-01 4.16608274e-01 6.74875855e-01
-6.15911126e-01 5.58438838e-01 -7.22095728e-01 -5.50527275e-01
3.31507325e-02 1.64528280e-01 -4.88684207e-01 -3.01312894e-01
-1.43949473e+00 1.04563534e+00 9.89315748e-01 1.23631619e-01
-6.34704888e-01 -7.99418628e-01 -9.63792920e-01 5.21582551e-03
2.57424921e-01 -6.42424226e-01 1.17512918e+00 -1.26142311e+00
-1.49092996e+00 8.91862571e-01 -3.44008327e-01 -2.98598379e-01
1.64331451e-01 -2.27088287e-01 -7.96521544e-01 -3.43107343e-01
1.71592996e-01 6.69615686e-01 4.00122404e-01 -1.29562676e+00
-6.23694718e-01 -7.46348798e-02 -1.94121674e-02 2.55879104e-01
-6.40354216e-01 3.27289104e-01 -5.07146955e-01 -7.92945147e-01
-1.55043498e-01 -7.22389162e-01 -4.15042907e-01 -7.79697597e-01
-4.29074705e-01 -3.06531131e-01 1.97841436e-01 -6.35487080e-01
1.38468742e+00 -2.09186554e+00 -1.53701464e-02 1.48965400e-02
-4.69330490e-01 5.28520942e-01 -5.17117202e-01 5.72003722e-01
-1.31604373e-01 2.72967488e-01 -3.66763204e-01 -5.71479619e-01
-2.14554429e-01 1.99069753e-01 -2.73562640e-01 -7.03882724e-02
3.29098076e-01 8.10036421e-01 -9.95788217e-01 -3.94336373e-01
-1.96072236e-01 4.49664295e-01 -4.23392951e-01 5.15215516e-01
7.38411490e-03 4.85529959e-01 -5.71106851e-01 4.63763714e-01
5.45371771e-01 -2.16258079e-01 4.49685007e-01 -3.64810228e-01
-4.25553285e-02 6.56703293e-01 -1.20836592e+00 1.93058598e+00
-1.07840371e+00 -6.94052652e-02 -2.25246444e-01 -9.24539387e-01
1.24473572e+00 3.54385704e-01 9.50964987e-02 -5.69367170e-01
-1.14846654e-01 5.21209776e-01 -1.07699625e-01 -3.50091070e-01
4.26539481e-01 -3.08029324e-01 -2.49913231e-01 4.03930217e-01
4.43918914e-01 3.07989389e-01 9.37986448e-02 -2.50418186e-01
9.19721603e-01 2.29931518e-01 5.46717942e-01 3.99677865e-02
7.52839029e-01 8.55314732e-02 8.78259838e-01 6.89929724e-01
8.43305234e-03 2.36400053e-01 -2.46206298e-01 -3.57940257e-01
-8.53482127e-01 -7.09392369e-01 -1.24270335e-01 1.52686465e+00
-1.44165173e-01 -4.04681206e-01 -7.60618210e-01 -1.21263993e+00
-3.17765534e-01 1.09335518e+00 -3.70915711e-01 -1.48044765e-01
-8.18383217e-01 -8.92556190e-01 9.52312708e-01 6.63023889e-01
8.21198404e-01 -1.28703046e+00 9.11113322e-02 4.01813656e-01
-3.50589752e-01 -1.22594500e+00 -6.43013358e-01 4.15541679e-01
-1.03756034e+00 -5.49317598e-01 -7.87084460e-01 -1.22158515e+00
8.49537551e-01 4.08269837e-02 1.28243160e+00 -1.26976445e-01
1.42643794e-01 1.64799497e-01 -5.75483441e-01 -1.76134020e-01
-7.01570332e-01 6.64854407e-01 -1.35698067e-02 4.37724777e-03
5.27330577e-01 -2.70598561e-01 -5.47475368e-03 3.10128152e-01
-8.26067567e-01 -7.90238678e-02 8.45916212e-01 9.93508816e-01
7.09949553e-01 -2.26920322e-02 7.67741323e-01 -1.18274057e+00
4.90384728e-01 -6.72401011e-01 -2.79320627e-01 9.25943792e-01
-4.83396858e-01 4.78863209e-01 9.70556617e-01 -6.06138706e-01
-1.53241503e+00 1.65237248e-01 -2.62872219e-01 -1.97578907e-01
-3.56296271e-01 8.13090324e-01 -6.66397989e-01 3.93305458e-02
5.85829914e-01 2.99690038e-01 -8.27445447e-01 -1.01737320e+00
6.15143359e-01 8.21695864e-01 3.31547707e-01 -8.34314227e-01
5.70178151e-01 6.78977370e-02 -5.91683209e-01 -5.08607984e-01
-1.44338548e+00 -5.91795743e-01 -8.95733535e-01 2.57167965e-01
6.43058419e-01 -9.67711151e-01 -9.03382152e-02 3.05632442e-01
-1.41325927e+00 -2.48590469e-01 -5.19794449e-02 5.93542576e-01
-1.50635749e-01 1.43600032e-01 -6.05051279e-01 -6.00037575e-01
-4.67769831e-01 -8.16888928e-01 8.81521583e-01 1.90405950e-01
6.77578449e-02 -1.42413878e+00 4.86526698e-01 2.25094199e-01
4.75939184e-01 -3.43619645e-01 7.73638070e-01 -1.41494548e+00
-8.56297016e-02 -5.20212203e-02 7.05489563e-03 6.61745846e-01
3.07037652e-01 -4.24699664e-01 -9.66314018e-01 -3.15434515e-01
3.95207368e-02 -5.04862130e-01 8.16593707e-01 -2.14272901e-01
1.00701940e+00 -4.31826591e-01 -5.02667725e-01 8.25058341e-01
1.39422190e+00 4.78297286e-02 4.99603420e-01 7.02107608e-01
8.35785270e-01 6.20947778e-01 6.49955392e-01 7.67905414e-02
5.83532691e-01 6.33685827e-01 -2.28536814e-01 -1.27819970e-01
-1.55342609e-01 -5.46803355e-01 6.03690982e-01 1.34006333e+00
1.95197314e-01 -3.21184874e-01 -1.07297981e+00 6.48606300e-01
-1.53527093e+00 -8.82347047e-01 4.59220022e-01 2.18542814e+00
1.24500084e+00 -3.30895581e-03 -3.15222144e-01 -5.61516881e-01
1.02311170e+00 1.60802938e-02 -5.53655028e-01 -2.94234514e-01
-2.74966091e-01 2.56469995e-01 4.71867412e-01 2.19814330e-01
-1.19638383e+00 1.40463173e+00 6.05141306e+00 1.01539481e+00
-1.19731128e+00 3.60387623e-01 5.35154760e-01 4.62535739e-01
-3.68453473e-01 1.36736393e-01 -1.37798560e+00 2.93022424e-01
1.26995707e+00 -2.11964637e-01 1.10912412e-01 1.04973435e+00
-2.06647247e-01 6.23466849e-01 -1.15709734e+00 6.15068913e-01
2.55964607e-01 -9.80030298e-01 2.95245856e-01 -4.25624192e-01
9.16092277e-01 3.66450757e-01 -3.47900510e-01 7.34779000e-01
3.91442150e-01 -7.56608963e-01 3.96249890e-01 4.42417800e-01
6.22816741e-01 -7.14771986e-01 8.63753855e-01 5.47414780e-01
-1.35949290e+00 1.86192408e-01 -5.77911735e-01 5.89009166e-01
2.67525077e-01 3.81963968e-01 -1.02606213e+00 8.62893999e-01
4.87282455e-01 6.97635472e-01 -7.23681867e-01 1.13719249e+00
-4.25412446e-01 7.84086049e-01 -2.37231344e-01 4.16393876e-02
1.65222198e-01 9.27396044e-02 3.69813055e-01 1.75004137e+00
4.18086857e-01 -1.42185718e-01 3.55656832e-01 6.48592234e-01
-4.73450810e-01 6.90734863e-01 -5.04685938e-01 -1.01818584e-01
7.24215925e-01 1.25069451e+00 -3.38113874e-01 -4.37448978e-01
-6.58752441e-01 1.41009867e+00 9.02297974e-01 2.70555079e-01
-6.53790593e-01 -7.11235285e-01 2.99974561e-01 -2.39109874e-01
5.59752226e-01 -9.86636430e-02 -5.06136864e-02 -1.63396323e+00
7.37840906e-02 -8.70039165e-01 7.32948244e-01 -4.94478613e-01
-1.78119814e+00 1.09787858e+00 -2.19687149e-01 -1.29268503e+00
-2.74777353e-01 -5.14383614e-01 -6.43606365e-01 9.01445091e-01
-1.80929327e+00 -1.51344359e+00 1.25119984e-01 5.15852690e-01
6.11588955e-01 -4.22938406e-01 9.77161586e-01 5.16701341e-01
-5.20242453e-01 1.07196844e+00 1.64859802e-01 6.04356706e-01
1.14494610e+00 -1.04007733e+00 6.40697300e-01 9.08299446e-01
4.64759618e-01 9.49199021e-01 1.85228974e-01 -7.28798866e-01
-1.00442851e+00 -1.37205255e+00 1.37708485e+00 -4.15195495e-01
7.06699610e-01 -1.72527552e-01 -1.27361584e+00 8.42391729e-01
2.56161630e-01 9.94236022e-02 1.08165491e+00 4.30597216e-01
-7.22078741e-01 -2.34061003e-01 -1.06754553e+00 4.87168550e-01
1.04909348e+00 -6.67054772e-01 -1.02761364e+00 2.63943911e-01
7.86997795e-01 -2.90049911e-01 -1.08360124e+00 5.46307981e-01
1.42981768e-01 -4.24158961e-01 7.15497077e-01 -1.27652693e+00
-1.70949213e-02 -3.56045544e-01 -2.08928823e-01 -1.70033991e+00
-2.42408305e-01 -3.58981907e-01 2.31941998e-01 1.91762042e+00
9.46817338e-01 -5.09356201e-01 3.74490976e-01 5.07354498e-01
-3.88005078e-01 -4.69962031e-01 -8.28306913e-01 -1.04695117e+00
2.66499192e-01 -4.47583765e-01 6.77944362e-01 1.39334023e+00
6.38150871e-02 5.31437933e-01 -3.61511379e-01 3.64029855e-01
4.41223145e-01 2.02596597e-02 5.35789609e-01 -8.74603570e-01
-3.21224988e-01 -4.04377989e-02 -1.30999252e-01 -1.14697552e+00
6.84834480e-01 -1.37764466e+00 1.82740405e-01 -1.24607182e+00
3.47513258e-01 -7.94011652e-01 -6.89688325e-01 9.15846765e-01
-5.36407471e-01 -1.35003492e-01 1.73844978e-01 2.71605283e-01
-8.55571687e-01 6.22985721e-01 8.35377574e-01 -5.13707325e-02
-3.29555720e-01 -1.25483526e-02 -6.63669050e-01 5.34961998e-01
7.07021773e-01 -9.27319705e-01 -1.01047777e-01 -8.41029942e-01
1.15849279e-01 -1.84629723e-01 -2.12644026e-01 -8.79773319e-01
3.66040498e-01 -1.70074999e-01 3.27610940e-01 -2.57618517e-01
-7.27064535e-02 -7.45175540e-01 -6.58442732e-03 1.61582425e-01
-6.22039199e-01 1.22956775e-01 3.40995133e-01 4.00213152e-01
-4.89870310e-01 -5.60860813e-01 7.15307891e-01 -2.08419472e-01
-1.02343512e+00 2.21690968e-01 1.80859827e-02 4.53854531e-01
5.86973786e-01 1.81918278e-01 -2.53369629e-01 5.18985689e-02
-6.46422744e-01 2.42703557e-01 3.88952583e-01 6.60378933e-01
3.14670712e-01 -1.47235703e+00 -8.17824125e-01 3.13353874e-02
4.76676881e-01 -4.36443150e-01 1.89342096e-01 4.32517231e-01
3.57509293e-02 3.83029908e-01 -4.40283259e-03 -2.17066392e-01
-9.74001229e-01 5.41683555e-01 5.19614041e-01 -5.69812059e-01
-1.94858551e-01 8.05430591e-01 3.72182280e-02 -1.20975316e+00
6.66601211e-02 -1.37958854e-01 -2.97135413e-01 -1.19112410e-01
5.86176813e-01 1.33933112e-01 2.28373885e-01 -7.58449733e-01
-5.43200374e-01 4.73976463e-01 -3.41077924e-01 -3.05844937e-02
1.38912189e+00 -2.43929893e-01 5.98596074e-02 6.68696105e-01
1.33014715e+00 3.05702329e-01 -8.18685889e-01 -6.59596264e-01
5.65501094e-01 -7.32431039e-02 -2.28434935e-01 -1.00334477e+00
-1.06686270e+00 6.46437407e-01 3.25388312e-01 -3.28355998e-01
1.11154962e+00 1.69232488e-01 8.58738661e-01 1.01368105e+00
5.08542776e-01 -1.17275071e+00 -1.14504881e-01 9.28932071e-01
5.25638223e-01 -1.32659650e+00 -3.50107372e-01 -3.13814491e-01
-9.79036152e-01 1.20918298e+00 7.72323608e-01 1.63593188e-01
6.59170330e-01 1.35688797e-01 3.71554226e-01 1.95063099e-01
-7.59458065e-01 -2.44954303e-01 6.14453316e-01 4.72302347e-01
9.12487149e-01 -3.36944163e-02 -3.01043719e-01 9.99554098e-01
3.53676006e-02 -2.33880151e-02 1.01791143e-01 9.26730096e-01
-4.61789608e-01 -1.72003150e+00 -2.08370052e-02 2.04212189e-01
-5.46203911e-01 -5.64835191e-01 -3.44980776e-01 5.86745083e-01
1.46793067e-01 7.72332072e-01 -2.75218040e-01 -2.99876869e-01
6.94396555e-01 4.94755000e-01 2.20137119e-01 -1.00607085e+00
-9.34750617e-01 -2.09499449e-01 4.10379201e-01 -3.60730708e-01
-6.18912518e-01 -4.52365249e-01 -1.23064530e+00 1.65003508e-01
-4.51672614e-01 5.27100801e-01 6.56204462e-01 1.11223555e+00
6.17363036e-01 2.30031580e-01 7.24601150e-01 -4.81050611e-01
-7.17277884e-01 -1.12418377e+00 -3.77989531e-01 4.02567536e-01
-1.58015653e-01 -3.88222903e-01 -3.42472702e-01 2.07349196e-01] | [9.946247100830078, 9.640213012695312] |
4b354729-8e42-4a60-aec4-19e432aa0f29 | quick-algorithms-for-independent-vector | 1910.10242 | null | https://arxiv.org/abs/1910.10242v2 | https://arxiv.org/pdf/1910.10242v2.pdf | Algorithm for Independent Vector Extraction Based on Semi-Time-Variant Mixing Model | A new algorithm for dynamic independent vector extraction is proposed. It is based on the mixing model where mixing parameters related to the source-of-interest (SOI) are time-variant while the separating parameters are time-invariant. A contrast function based on the quasi-likelihood approach is optimized using the Newton-Raphson approach. The update is computed without imposing the orthogonal constraint, and the orthogonality is enforced afterward. This yields an algorithm that is significantly faster than gradient-based algorithms while different from fixed-point methods, which are even faster. We show advantageous properties of the proposed algorithm compared to the fixed-point methods in an on-line processing regime where stable convergence to the SOI is the important issue. The effectiveness of the method is demonstrated in a speech extraction experiment with a dense microphone array. | ['Jaroslav Čmejla', 'Tomáš Kounovský', 'Václav Kautský', 'Zbyněk Koldovský'] | 2019-10-22 | null | null | null | null | ['speech-extraction'] | ['speech'] | [ 1.14634730e-01 -3.95819336e-01 2.76055753e-01 -9.48077142e-02
-8.33928287e-01 -5.93741655e-01 4.70631242e-01 -2.03392923e-01
-5.35611808e-01 6.34215415e-01 1.52755529e-01 -2.29017466e-01
-5.98916948e-01 -1.47064850e-01 -2.38203481e-01 -1.19161832e+00
-6.17901325e-01 1.90873966e-01 -5.45196533e-02 -1.44865289e-01
2.99932331e-01 5.78086019e-01 -1.27742350e+00 -5.33046126e-01
6.98140621e-01 7.09480941e-01 1.83964804e-01 1.05550027e+00
5.29824570e-02 2.54788637e-01 -5.44167519e-01 2.41450384e-01
5.14534891e-01 -4.72019613e-01 -2.52536774e-01 1.90603182e-01
1.41622856e-01 3.03028636e-02 -1.06704526e-01 1.16770232e+00
6.92648113e-01 6.00398481e-01 5.81880689e-01 -9.70422208e-01
-8.62117186e-02 3.15090626e-01 -5.31264246e-01 3.09507132e-01
2.92515934e-01 -4.89583343e-01 8.85263622e-01 -1.28194237e+00
4.91308928e-01 1.10863304e+00 7.62064278e-01 1.37868762e-01
-1.24090171e+00 -5.72715938e-01 -9.96762216e-02 9.24610049e-02
-1.45478380e+00 -9.86496091e-01 9.88670468e-01 -2.72278696e-01
8.91912997e-01 4.98536527e-01 5.74136615e-01 6.18468046e-01
2.39129905e-02 5.51095426e-01 1.01915157e+00 -8.83669078e-01
3.49699140e-01 1.06159188e-01 2.96892107e-01 7.02475369e-01
1.23819433e-01 2.41676256e-01 -7.45065808e-01 -4.02256161e-01
6.28889441e-01 -5.07989705e-01 -5.55451334e-01 -5.39640009e-01
-1.28642881e+00 7.43665934e-01 -3.63065898e-01 5.64263523e-01
-3.40717971e-01 -3.26279625e-02 1.16759360e-01 2.77600586e-01
7.27549434e-01 3.18741262e-01 -4.40972477e-01 -2.13112935e-01
-1.44051623e+00 2.01671392e-01 1.21369183e+00 8.42400789e-01
4.36387867e-01 6.01324320e-01 2.20239177e-01 1.03710008e+00
7.21405387e-01 8.82525623e-01 6.25576794e-01 -8.91601384e-01
4.07629371e-01 -3.40820760e-01 2.35439137e-01 -7.41235673e-01
-4.39298421e-01 -7.99144506e-01 -6.84866369e-01 2.83419549e-01
7.30437398e-01 -6.55116737e-01 -7.00184882e-01 1.68464196e+00
5.84660649e-01 5.29649317e-01 1.12974279e-01 7.08311558e-01
4.25789148e-01 9.09482837e-01 -6.59451962e-01 -1.13171244e+00
9.48990762e-01 -8.38789403e-01 -1.38186347e+00 -4.39921319e-02
6.38585463e-02 -1.30762219e+00 3.27407956e-01 6.77959025e-01
-1.19357264e+00 -5.19011259e-01 -1.32168853e+00 3.42939079e-01
-7.61813074e-02 -3.08256992e-03 3.86081845e-01 8.94435108e-01
-1.14102268e+00 6.71711683e-01 -8.10776055e-01 1.25199882e-02
-5.04913032e-01 4.91246045e-01 -2.56632596e-01 6.72123849e-01
-8.54692638e-01 7.23145247e-01 1.36403531e-01 4.67786580e-01
-3.57683152e-01 -5.25942683e-01 -8.40055525e-01 -1.20387696e-01
1.41280338e-01 -2.81553745e-01 1.29571760e+00 -8.51355851e-01
-2.31042409e+00 2.34711617e-01 -6.60965323e-01 -1.41956836e-01
5.19250095e-01 -3.56661469e-01 -6.40770853e-01 2.81687140e-01
-2.02533044e-02 -1.42832309e-01 1.64899099e+00 -1.07293665e+00
-4.39404786e-01 2.95425784e-02 -6.49262905e-01 3.43276590e-01
-3.70665222e-01 1.42288178e-01 -5.17201066e-01 -7.03509748e-01
7.84750998e-01 -1.01050341e+00 -3.20098162e-01 -3.65155578e-01
-4.95342255e-01 2.78532505e-01 7.93425024e-01 -8.45988333e-01
1.47617495e+00 -2.27063012e+00 4.24032837e-01 6.01652086e-01
-8.32076892e-02 1.11693762e-01 3.54298241e-02 7.30163753e-01
-4.64307755e-01 -4.31716830e-01 -3.49221885e-01 -5.57040334e-01
-3.04586768e-01 -2.59317160e-01 -1.68992996e-01 1.01192319e+00
-2.32048333e-01 2.00861871e-01 -8.71031404e-01 -3.99139851e-01
3.43724340e-01 3.72615874e-01 -4.63279635e-01 2.01815665e-01
5.51113427e-01 5.99146485e-01 -6.84605166e-02 3.56958866e-01
6.39784098e-01 1.38761878e-01 1.71350390e-01 -3.19675624e-01
-6.97035134e-01 1.07739367e-01 -1.92582726e+00 1.48785508e+00
-6.07925296e-01 9.76409197e-01 7.26323247e-01 -1.02773523e+00
8.55841398e-01 8.50134790e-01 7.19430625e-01 1.42632453e-02
3.79870720e-02 5.14046013e-01 2.40882874e-01 -3.47995818e-01
2.30523393e-01 -3.38729918e-01 4.23792630e-01 2.28323668e-01
3.89195323e-01 -3.57936233e-01 3.26464742e-01 1.39978960e-01
7.36490726e-01 4.87911105e-02 7.22176194e-01 -7.80321121e-01
7.03193426e-01 -3.79030019e-01 5.95245898e-01 6.43136203e-01
-1.49194196e-01 5.18620610e-01 -4.79003899e-02 2.82481492e-01
-7.11794019e-01 -9.21372652e-01 -4.26247209e-01 3.61985296e-01
-5.39771505e-02 -4.24150676e-01 -6.95489824e-01 1.58921219e-02
-2.23295882e-01 5.82161129e-01 -1.72174841e-01 4.42950040e-01
-5.91597497e-01 -8.95776033e-01 1.79066118e-02 -6.87294304e-02
3.74772578e-01 -4.03778315e-01 -3.44666749e-01 7.75669336e-01
-2.10591465e-01 -1.00287104e+00 -5.17904282e-01 4.03167725e-01
-1.14741778e+00 -7.30485857e-01 -9.65050042e-01 -6.35289073e-01
5.60157716e-01 3.69821936e-01 5.49112558e-01 -4.97466296e-01
5.25879003e-02 6.89311504e-01 -4.50053029e-02 -5.15341759e-01
-4.18215454e-01 -3.21586370e-01 4.88015592e-01 5.54556787e-01
-1.35489544e-02 -8.96800101e-01 -3.46611857e-01 2.05937162e-01
-4.47427124e-01 -1.68390736e-01 -4.82628942e-02 8.84340882e-01
4.41864073e-01 3.90969068e-01 4.56339449e-01 -5.49593747e-01
6.68113708e-01 -2.67859817e-01 -8.80684555e-01 -1.08644761e-01
-6.63561702e-01 2.24834830e-02 5.88243484e-01 -7.63354301e-01
-1.34091353e+00 3.93308401e-01 -1.86081469e-01 -1.99279442e-01
2.04422280e-01 2.64010042e-01 -1.80127263e-01 -3.49103779e-01
3.69018704e-01 6.95794001e-02 -1.02480374e-01 -7.54690409e-01
3.88382316e-01 7.15881824e-01 5.74326038e-01 -1.79721639e-01
1.27387416e+00 5.41662395e-01 1.30901799e-01 -1.77361083e+00
-3.53064567e-01 -9.50073242e-01 -7.09735751e-01 -3.74725163e-01
4.43216562e-01 -6.53743565e-01 -3.90497506e-01 6.59564197e-01
-1.21770883e+00 6.92956671e-02 -3.20176840e-01 1.40049589e+00
-5.51077306e-01 6.41174018e-01 -5.14615595e-01 -1.34223628e+00
-2.40155712e-01 -9.36521530e-01 5.84609270e-01 2.29846746e-01
-3.21496665e-01 -1.22777975e+00 4.25448596e-01 -1.69041812e-01
3.71300817e-01 -2.68974543e-01 4.35888827e-01 -6.85097754e-01
-2.97888428e-01 -2.46547326e-01 4.21303153e-01 3.41251194e-01
3.85201097e-01 3.64478052e-01 -1.17543972e+00 -4.48137671e-01
8.22102487e-01 6.51954889e-01 3.80714327e-01 8.55866671e-01
2.23076507e-01 -2.17641205e-01 -3.87699813e-01 9.04400587e-01
1.49828756e+00 5.69187880e-01 2.27448449e-01 1.13364838e-01
4.62164402e-01 5.25369942e-01 4.32136357e-01 4.95845020e-01
-2.39928693e-01 7.64923155e-01 -7.19995499e-02 -2.38004908e-01
3.95159423e-02 1.59093261e-01 4.74662036e-01 1.58604729e+00
-7.73890987e-02 -2.60566473e-01 -5.73095202e-01 6.19021475e-01
-1.65814352e+00 -9.43860173e-01 -4.90810752e-01 2.47859979e+00
6.02297127e-01 -4.61178534e-02 -1.36104867e-01 6.53410733e-01
6.49492681e-01 2.36326978e-01 -2.93016672e-01 -3.13376427e-01
-2.29247250e-02 3.42991114e-01 7.16904581e-01 1.22833490e+00
-1.13611281e+00 5.55402815e-01 7.39147663e+00 7.92353511e-01
-1.10471594e+00 3.23459208e-01 -2.18298897e-01 -3.73010114e-02
4.13925573e-02 9.28968564e-02 -8.23022425e-01 3.07870150e-01
8.99115026e-01 -3.57581526e-01 5.33854306e-01 4.03255373e-01
4.62909043e-01 -2.26135969e-01 -8.47237527e-01 1.24314857e+00
7.09249079e-02 -7.94294417e-01 -7.15839803e-01 -4.79913726e-02
4.48248357e-01 -1.67466447e-01 -1.45322243e-02 -2.77428895e-01
-1.95123151e-01 -3.48170966e-01 9.35315251e-01 4.53664124e-01
3.39687288e-01 -5.60783923e-01 1.79503322e-01 3.49908084e-01
-1.39023447e+00 -8.74170940e-03 -8.41209292e-02 -1.30752861e-01
6.62086785e-01 9.83061194e-01 -6.43562317e-01 3.51179987e-01
3.73309851e-01 4.61532861e-01 -2.98970714e-02 1.35944188e+00
-1.81195542e-01 8.78320694e-01 -8.17789495e-01 -9.77586582e-03
8.19501579e-02 -9.16411638e-01 1.39065063e+00 1.50580943e+00
7.15058088e-01 1.48464397e-01 -3.31335455e-01 2.96837360e-01
5.89132488e-01 3.83164972e-01 -6.36112928e-01 8.47299322e-02
4.10689712e-01 1.33978927e+00 -9.15683448e-01 -2.59993196e-01
-4.85958964e-01 8.72961223e-01 -3.58586699e-01 8.21339071e-01
-4.03811455e-01 -7.40721524e-01 4.37688887e-01 -3.25655609e-01
5.99546432e-01 -7.17419624e-01 -3.26063745e-02 -1.16489065e+00
7.14603160e-03 -9.61340368e-01 4.59619835e-02 -3.43780160e-01
-8.40124488e-01 6.03431404e-01 3.49762410e-01 -1.56184042e+00
-7.64675558e-01 -5.66206872e-01 -5.91513097e-01 1.01840305e+00
-1.24513292e+00 -4.48256522e-01 4.67201799e-01 5.67587316e-01
5.84058583e-01 -3.91257793e-01 9.34195101e-01 2.96131551e-01
-5.51808298e-01 3.47897857e-01 7.63422072e-01 -2.58013725e-01
4.68329281e-01 -1.16867125e+00 1.38786048e-01 1.37910461e+00
6.01383507e-01 1.02650058e+00 1.21744049e+00 -4.17969584e-01
-1.57332790e+00 -2.49998048e-01 9.89253700e-01 5.32589890e-02
8.78922224e-01 -4.65482652e-01 -7.26731837e-01 3.98863077e-01
3.39162767e-01 -2.39323542e-01 8.89616072e-01 9.59454477e-02
1.39116362e-01 -2.20618472e-01 -8.73452783e-01 4.11392301e-01
7.62437940e-01 -3.60637188e-01 -5.92167556e-01 4.83941287e-01
2.52500743e-01 -4.21937793e-01 -6.35951936e-01 1.02190234e-01
5.42291403e-01 -9.66986895e-01 1.11099923e+00 -1.25660792e-01
-4.16051805e-01 -4.63300139e-01 -2.36999124e-01 -1.50847530e+00
-5.00994503e-01 -1.53322685e+00 -1.91823035e-01 1.17736840e+00
5.27765036e-01 -1.00171924e+00 3.46402168e-01 1.40525907e-01
2.12516133e-02 -2.18715027e-01 -1.32966554e+00 -9.81613874e-01
-5.08332670e-01 -5.07180631e-01 1.03459992e-01 7.47774065e-01
1.24263652e-01 5.14685869e-01 -5.11951506e-01 6.54107332e-01
1.07205164e+00 1.01740874e-01 4.66240436e-01 -1.10787249e+00
-7.90853381e-01 -2.29160994e-01 -8.94601643e-02 -1.22602785e+00
1.42829390e-02 -4.51611936e-01 1.82261929e-01 -1.09392822e+00
-4.87646580e-01 -3.05353314e-01 -2.07887620e-01 -1.18377268e-01
-6.70113415e-02 -1.81642119e-02 1.15751334e-01 1.77371338e-01
1.47325039e-01 3.45803261e-01 8.20275843e-01 -1.97139196e-02
-6.59279585e-01 4.56751078e-01 -3.14894877e-02 1.01779139e+00
5.91331720e-01 -4.15303051e-01 -6.40702963e-01 -2.07580104e-01
-1.70162112e-01 4.31077987e-01 -1.79356247e-01 -1.12778795e+00
4.48542774e-01 2.91755438e-01 1.09209396e-01 -6.41306102e-01
8.53252292e-01 -8.46344352e-01 4.61668551e-01 3.85814011e-01
9.48541611e-02 1.38560086e-02 1.91668078e-01 5.87527752e-01
-4.29439604e-01 -7.45843112e-01 7.49545753e-01 2.71031171e-01
-3.86699617e-01 -3.51013179e-04 -6.47612870e-01 -2.02670172e-01
5.18579662e-01 -1.81013778e-01 6.25807345e-01 -8.69463086e-01
-6.71101332e-01 -3.52905631e-01 -1.18326925e-01 -1.54110283e-01
4.89031285e-01 -1.21047199e+00 -7.16642618e-01 4.61001396e-01
-4.67841744e-01 -5.25716603e-01 -2.74158157e-02 9.85710979e-01
-3.81526887e-01 3.17764372e-01 2.98202962e-01 -5.96222579e-01
-1.49937820e+00 4.46588546e-01 3.78124386e-01 -1.59339383e-01
-5.11990309e-01 1.03870738e+00 -1.97496582e-02 -2.03650549e-01
2.99518585e-01 -1.83510035e-01 -1.95422605e-01 2.47982353e-01
6.26686275e-01 6.66910648e-01 1.28187433e-01 -8.23242188e-01
-3.56814235e-01 9.17548776e-01 3.36455047e-01 -1.00403774e+00
1.20753121e+00 -3.47893149e-01 -2.84710795e-01 7.86770582e-01
1.42435205e+00 9.51717317e-01 -9.43578064e-01 -1.88834548e-01
-3.24753001e-02 -5.63795805e-01 3.75671148e-01 -2.68001884e-01
-8.61727238e-01 7.54065990e-01 8.66061032e-01 2.86709338e-01
1.14837384e+00 -7.18508780e-01 4.51670080e-01 3.34046006e-01
2.80601263e-01 -1.10749078e+00 -4.52155620e-01 3.54891270e-01
7.96170771e-01 -8.98980737e-01 2.79657513e-01 -5.83810210e-01
1.50867641e-01 1.28602660e+00 -2.20955059e-01 3.49914879e-02
1.13124144e+00 4.75249380e-01 2.99173594e-01 2.00459599e-01
-2.17334375e-01 -8.60068202e-02 4.52986747e-01 5.78713238e-01
4.69814122e-01 -1.81770120e-02 -5.52345991e-01 1.31644130e-01
-1.35309368e-01 -3.28701198e-01 3.21074873e-01 7.26408005e-01
-2.71647274e-01 -1.34230161e+00 -9.16982234e-01 4.56743464e-02
-5.64433753e-01 -2.07689658e-01 2.25196198e-01 7.06400037e-01
-2.55563378e-01 1.48019862e+00 -2.95380145e-01 -1.08981825e-01
3.05472285e-01 2.76245415e-01 5.82560778e-01 -3.74657601e-01
-3.19860101e-01 9.00437117e-01 2.14364827e-01 -4.10027117e-01
-5.57394207e-01 -9.45698917e-01 -1.11726809e+00 1.55449033e-01
-8.55968058e-01 6.26335442e-01 1.06075549e+00 8.14076364e-01
-7.92801902e-02 1.98153213e-01 1.05065072e+00 -1.13157654e+00
-4.81363773e-01 -7.73809254e-01 -1.02449977e+00 -6.44252300e-02
6.52541637e-01 -5.23077428e-01 -8.61228108e-01 6.11305684e-02] | [15.150936126708984, 5.69337797164917] |
2508892c-dca6-481c-880f-f9946baca9dd | heterogeneous-branch-collaborative-learning | 2303.11621 | null | https://arxiv.org/abs/2303.11621v1 | https://arxiv.org/pdf/2303.11621v1.pdf | Heterogeneous-Branch Collaborative Learning for Dialogue Generation | With the development of deep learning, advanced dialogue generation methods usually require a greater amount of computational resources. One promising approach to obtaining a high-performance and lightweight model is knowledge distillation, which relies heavily on the pre-trained powerful teacher. Collaborative learning, also known as online knowledge distillation, is an effective way to conduct one-stage group distillation in the absence of a well-trained large teacher model. However, previous work has a severe branch homogeneity problem due to the same training objective and the independent identical training sets. To alleviate this problem, we consider the dialogue attributes in the training of network branches. Each branch learns the attribute-related features based on the selected subset. Furthermore, we propose a dual group-based knowledge distillation method, consisting of positive distillation and negative distillation, to further diversify the features of different branches in a steadily and interpretable way. The proposed approach significantly improves branch heterogeneity and outperforms state-of-the-art collaborative learning methods on two widely used open-domain dialogue datasets. | ['Kan Li', 'Bin Sun', 'Shaoxiong Feng', 'Yiwei Li'] | 2023-03-21 | null | null | null | null | ['dialogue-generation', 'dialogue-generation'] | ['natural-language-processing', 'speech'] | [ 1.30024433e-01 4.60393518e-01 -2.64021993e-01 -4.68509823e-01
-5.94734430e-01 -4.97008413e-01 7.28916585e-01 2.29168624e-01
-5.23821294e-01 1.30351305e+00 3.56262326e-01 -2.89035320e-01
-1.42555803e-01 -1.03755772e+00 -2.01341003e-01 -8.50366235e-01
2.88758755e-01 9.54025745e-01 2.41444632e-01 -6.23815358e-01
6.40185475e-02 -7.94575270e-03 -1.29785991e+00 2.42949486e-01
1.64344668e+00 8.83520722e-01 8.21537971e-02 3.31625074e-01
-5.81016421e-01 7.41417229e-01 -7.57818103e-01 -7.56923020e-01
1.55042941e-02 -8.94073904e-01 -1.09588051e+00 7.40317851e-02
-1.08955121e-02 -2.85531640e-01 -2.77416408e-02 7.60712266e-01
7.37838387e-01 4.90417421e-01 7.20509827e-01 -9.54644561e-01
-4.24780101e-01 1.24733663e+00 -4.04000491e-01 -1.17330596e-01
2.76057303e-01 1.29816309e-01 1.18774772e+00 -8.09664428e-01
4.51951891e-01 1.20943522e+00 3.43885064e-01 6.26862466e-01
-1.29855657e+00 -7.87399769e-01 1.77933499e-01 2.83319980e-01
-1.14376593e+00 -2.00820297e-01 1.11285698e+00 -2.97581822e-01
4.48951215e-01 2.18291972e-02 7.06373036e-01 9.71871853e-01
-5.38141787e-01 1.03076053e+00 1.08625996e+00 -4.70831513e-01
1.96931154e-01 4.69046086e-01 -2.73843613e-02 6.32702231e-01
-1.22229218e-01 -3.24609458e-01 -5.17672420e-01 -3.65715265e-01
3.74172658e-01 -2.62874842e-01 -3.24729919e-01 -5.51304519e-01
-1.07156432e+00 1.13159382e+00 3.08373868e-01 2.65504450e-01
-2.68578470e-01 -5.54139614e-01 4.81357902e-01 3.63606215e-01
6.72692895e-01 9.77485538e-01 -8.34602833e-01 -2.50246495e-01
-6.77097499e-01 2.10121900e-01 1.19085097e+00 7.54474282e-01
8.93503785e-01 -1.21149443e-01 -5.98054528e-01 1.16173339e+00
1.69536471e-01 -4.01664339e-02 6.15172505e-01 -6.60280883e-01
7.32204378e-01 9.44965243e-01 -8.80363435e-02 -6.59241319e-01
-2.05343738e-01 -5.91206431e-01 -9.88394678e-01 -5.99750131e-03
5.87772548e-01 -7.65530765e-01 -6.04507804e-01 1.69861054e+00
7.54252434e-01 8.95063728e-02 5.20823598e-01 5.55107951e-01
1.00822091e+00 6.18150294e-01 1.02745086e-01 -3.29131573e-01
1.01946568e+00 -1.29813397e+00 -6.94573700e-01 1.63863286e-01
8.39726925e-01 -5.09984732e-01 9.10669208e-01 4.12347913e-01
-9.52932239e-01 -3.70564222e-01 -8.34038734e-01 -9.11900401e-03
-3.20038080e-01 1.70359150e-01 7.76715636e-01 4.85789657e-01
-5.93331039e-01 5.76655209e-01 -1.06850080e-01 1.21705700e-02
5.25479972e-01 3.56948167e-01 -1.49171844e-01 6.06152862e-02
-1.40736437e+00 9.56431091e-01 7.92386651e-01 -1.60553697e-02
-8.03938746e-01 -6.85711086e-01 -5.77566326e-01 2.13157907e-01
7.99860358e-01 -8.07873309e-01 1.21756732e+00 -1.02649379e+00
-2.18037438e+00 4.51064616e-01 4.22394514e-01 -4.16211993e-01
8.02353799e-01 -2.91028142e-01 1.31736740e-01 -6.42601475e-02
-2.57558048e-01 5.63672841e-01 6.94258451e-01 -1.19664872e+00
-7.72240281e-01 -3.31505567e-01 3.03207010e-01 7.61191130e-01
-7.22660244e-01 -4.30712819e-01 -2.27128863e-01 -4.24664140e-01
-2.12110087e-01 -8.23166370e-01 -4.25001264e-01 -3.20816964e-01
-3.51930082e-01 -6.93405390e-01 6.38714910e-01 -5.59507549e-01
1.29090405e+00 -1.74241793e+00 6.00715041e-01 5.76368645e-02
4.39768970e-01 6.96983099e-01 -5.08157648e-02 3.26666325e-01
2.95367479e-01 -1.83879193e-02 -1.61869317e-01 -3.17509651e-01
-2.27204978e-01 1.78037688e-01 -5.63818291e-02 -1.59474805e-01
1.42175049e-01 5.06681204e-01 -1.38094628e+00 -6.70968652e-01
1.78941526e-02 9.23561305e-02 -7.09710836e-01 6.25823498e-01
-5.39574146e-01 8.65505934e-01 -5.62403142e-01 1.97709307e-01
4.40081656e-01 -8.25233832e-02 4.47485268e-01 -6.44489452e-02
3.15602161e-02 4.02726769e-01 -8.93879235e-01 1.85004282e+00
-7.64093161e-01 2.93732047e-01 -2.13174541e-02 -1.21900630e+00
1.34904933e+00 2.73441166e-01 3.85035694e-01 -1.87083468e-01
2.53963590e-01 2.72172332e-01 3.12712878e-01 -4.13184434e-01
6.15158439e-01 -1.64800547e-02 -4.39775512e-02 5.96352518e-01
3.76449168e-01 -2.99011201e-01 1.86980203e-01 2.52552509e-01
7.70726383e-01 1.57743722e-01 3.29367429e-01 -5.23825698e-02
6.95204556e-01 -1.48204416e-01 7.61904061e-01 6.72674477e-01
-6.19461164e-02 1.16173677e-01 5.61968982e-01 -2.89969981e-01
-7.35367656e-01 -6.03811443e-01 2.01197088e-01 1.55392659e+00
4.14211974e-02 -3.62046719e-01 -6.43751860e-01 -1.23707664e+00
-1.03369439e-02 7.56007969e-01 -5.21304309e-01 -3.63322705e-01
-4.95087743e-01 -7.60035455e-01 6.26549184e-01 3.79040599e-01
9.86525893e-01 -8.68931770e-01 -1.40929148e-01 2.24814773e-01
-3.48661810e-01 -7.61431098e-01 -2.28259638e-02 2.99008042e-01
-7.84618735e-01 -8.79622281e-01 -9.35104847e-01 -6.84845805e-01
6.64897025e-01 3.38899263e-04 1.13965583e+00 1.58761129e-01
2.25071713e-01 -1.27138078e-01 -6.46469891e-01 -2.82853961e-01
-5.96676290e-01 6.84821665e-01 -1.08849049e-01 1.31700769e-01
3.74306887e-01 -4.99579400e-01 -4.03118253e-01 2.84033209e-01
-4.65106726e-01 5.14031887e-01 6.62586629e-01 1.51275504e+00
2.03976333e-01 1.65415689e-01 9.97875690e-01 -1.10431230e+00
1.11869943e+00 -5.90995312e-01 -3.00123930e-01 5.41066945e-01
-7.25757837e-01 4.18298125e-01 9.65905786e-01 -6.90530062e-01
-1.54242361e+00 -5.73122427e-02 2.12024786e-02 -1.33244514e-01
-2.74921991e-02 7.17115104e-01 -1.58460587e-01 1.07024737e-01
7.60151267e-01 1.99522004e-01 2.82977074e-02 -4.42353815e-01
7.79831827e-01 7.75519490e-01 2.30820045e-01 -8.65884960e-01
6.66783392e-01 -6.20037131e-02 -3.03729743e-01 -5.95073164e-01
-9.90305960e-01 -2.60196209e-01 -8.56176853e-01 -3.52465883e-02
6.57645583e-01 -7.68208802e-01 -5.48285604e-01 6.61452830e-01
-1.10404348e+00 -4.71884727e-01 -2.57146209e-01 4.82188791e-01
-2.37424731e-01 3.12449247e-01 -2.95653194e-01 -7.53085852e-01
-4.97234464e-01 -9.53410327e-01 4.14259255e-01 6.53105557e-01
-1.47849079e-02 -1.06414318e+00 9.97295305e-02 7.03964829e-01
4.76664484e-01 -3.20430696e-02 1.03435922e+00 -1.41269386e+00
-1.84347540e-01 9.99798775e-02 -1.31036505e-01 4.14654911e-01
3.87550950e-01 -5.72415888e-02 -8.68929207e-01 -9.93077829e-02
-2.30366513e-01 -8.28431249e-01 8.52841675e-01 -1.03231981e-01
1.06717944e+00 -4.07330990e-01 -3.13812822e-01 1.90777391e-01
7.11582124e-01 2.35536829e-01 1.81606874e-01 1.06339023e-01
7.84537435e-01 8.41580689e-01 8.29668045e-01 6.86230600e-01
4.59732413e-01 6.11835539e-01 1.50569916e-01 -2.14523032e-01
1.92469973e-02 -3.43915075e-01 -2.29429733e-02 9.55992818e-01
-2.41866171e-01 -1.53171480e-01 -9.12327588e-01 4.78816390e-01
-1.97726583e+00 -6.66369081e-01 3.20212543e-01 2.18421197e+00
1.68940032e+00 -2.99764872e-02 2.02647686e-01 -4.08903277e-03
8.06929410e-01 8.22184514e-03 -5.91030836e-01 -3.55051666e-01
7.60460347e-02 8.22498649e-02 -9.34274495e-03 4.17614520e-01
-9.24263895e-01 1.11433947e+00 5.07178736e+00 1.01100612e+00
-9.66788471e-01 -9.70358402e-02 7.32108772e-01 1.99436963e-01
-3.69585335e-01 9.82019082e-02 -1.02020776e+00 4.53244209e-01
7.12742388e-01 -3.32129180e-01 2.74634272e-01 9.57056642e-01
-8.03047121e-02 -1.52715787e-01 -1.01635551e+00 6.48583710e-01
6.93225954e-03 -1.24057734e+00 2.26238683e-01 -1.47043362e-01
9.92139518e-01 -4.79241848e-01 -2.28373110e-01 8.66641104e-01
6.93718851e-01 -8.54389608e-01 -3.05682942e-02 3.05568755e-01
4.78547096e-01 -8.44805360e-01 8.60680878e-01 6.04479969e-01
-7.96084583e-01 -2.82259583e-01 -2.67513156e-01 -2.64849998e-02
-1.04346178e-01 4.89919543e-01 -1.48137164e+00 7.66015351e-01
3.42726618e-01 5.13389945e-01 -4.72083807e-01 9.56467748e-01
-5.80616236e-01 6.75129712e-01 -1.72763452e-01 -3.15414578e-01
2.45912433e-01 -4.00979280e-01 2.71843523e-01 9.42253649e-01
5.82428882e-04 2.39686742e-01 5.35799682e-01 6.99502289e-01
-1.81199402e-01 3.48807037e-01 -4.15754527e-01 -3.82994451e-02
7.94137359e-01 1.26036501e+00 -2.34750479e-01 -4.88308012e-01
-3.92818034e-01 7.87055433e-01 5.75571418e-01 1.40639320e-02
-5.88581383e-01 -6.16070390e-01 3.12466532e-01 -4.03780907e-01
2.99766392e-01 -1.18466886e-02 -1.31159261e-01 -1.16002202e+00
-3.93670857e-01 -1.14696050e+00 4.51947689e-01 -2.08379060e-01
-1.25553548e+00 5.39992988e-01 2.40038447e-02 -1.12060893e+00
-6.45626247e-01 -4.98161768e-04 -7.33737230e-01 9.32350218e-01
-1.50688863e+00 -1.22467935e+00 -5.34624696e-01 7.12579131e-01
7.08849370e-01 -4.89868581e-01 1.03237391e+00 2.06470024e-02
-7.09963322e-01 8.49316120e-01 2.39981607e-01 2.49465719e-01
1.01483512e+00 -1.43255913e+00 -1.82257757e-01 2.39406556e-01
-1.57957524e-01 3.80984008e-01 6.26019597e-01 -5.39627910e-01
-1.03670824e+00 -8.78544271e-01 6.99891329e-01 -4.33359109e-02
6.22703671e-01 -2.32406080e-01 -1.01152074e+00 2.89467007e-01
3.17082882e-01 -3.30193669e-01 9.93936956e-01 6.32221460e-01
-5.08878790e-02 -1.68419883e-01 -1.06358719e+00 5.99476933e-01
7.08100021e-01 -7.65698254e-02 -9.39359486e-01 2.68117577e-01
6.51345849e-01 -4.59328264e-01 -1.03304255e+00 3.06945026e-01
3.49664807e-01 -7.05767989e-01 6.14504695e-01 -7.67379105e-01
4.82804894e-01 1.08961791e-01 4.05079037e-01 -1.83858597e+00
-1.06956199e-01 -8.59225690e-01 -1.76616609e-01 1.69308424e+00
6.62559450e-01 -5.74921846e-01 8.67954016e-01 5.98222017e-01
-8.00492391e-02 -1.12774611e+00 -5.83974302e-01 -5.75187802e-01
3.42068702e-01 4.93281990e-01 6.52689219e-01 1.27898812e+00
3.37529570e-01 9.90563095e-01 -4.16937649e-01 -4.25107330e-01
2.16870412e-01 3.62361103e-01 1.20104980e+00 -1.56383324e+00
-4.96150941e-01 -3.23663414e-01 5.05906343e-02 -1.35333598e+00
2.67297208e-01 -7.42957175e-01 1.44773901e-01 -1.44563019e+00
2.36242879e-02 -9.53534365e-01 -2.70143915e-02 6.11889124e-01
-5.70634186e-01 -3.66596013e-01 2.40882542e-02 6.08091317e-02
-6.48054659e-01 1.15296245e+00 1.54748857e+00 -1.76882476e-01
-7.56304741e-01 2.70377696e-01 -7.32580304e-01 6.35174513e-01
9.05462205e-01 -3.32033157e-01 -9.04777348e-01 -2.01482132e-01
-1.01241600e-02 1.46372706e-01 -2.46349782e-01 -4.53608185e-01
2.99010158e-01 -3.40029836e-01 1.94074333e-01 -4.59010065e-01
4.83925819e-01 -4.28303778e-01 -4.78479296e-01 2.82794535e-01
-6.44501865e-01 -4.65718657e-01 -5.84736541e-02 4.76997435e-01
-3.01698655e-01 -4.45987284e-01 7.00568140e-01 -2.97372639e-01
-5.77660203e-01 1.41151547e-01 -1.65531278e-01 2.43301496e-01
1.04828751e+00 7.13994447e-03 -5.58068931e-01 -4.50653136e-01
-5.95308602e-01 6.15924358e-01 4.60650250e-02 3.34016085e-01
2.34067231e-01 -1.10427380e+00 -9.38452125e-01 -2.14612614e-02
-1.09372713e-01 5.90566337e-01 1.93772554e-01 7.69425690e-01
-1.04264624e-01 1.27281636e-01 -3.16456348e-01 -3.09513330e-01
-1.26746571e+00 2.23286122e-01 2.02833414e-01 -8.11948180e-01
-4.08283830e-01 1.07403409e+00 1.25753492e-01 -6.56633794e-01
2.77814984e-01 3.20354402e-02 -7.17584133e-01 5.23804784e-01
3.59877616e-01 3.55410814e-01 -1.29473552e-01 -1.93173468e-01
4.93771769e-02 4.69586588e-02 -4.89702433e-01 -3.49934548e-02
1.19102728e+00 -1.21375188e-01 1.58808436e-02 2.87050992e-01
6.52784765e-01 -1.25600368e-01 -1.25724185e+00 -5.34617066e-01
-1.48408994e-01 -2.21061990e-01 3.92613094e-03 -1.05089414e+00
-1.00570226e+00 7.90484071e-01 1.40977129e-02 2.06772342e-01
9.21939433e-01 -2.05953807e-01 5.93877256e-01 6.98726058e-01
2.46269375e-01 -1.34598446e+00 3.34372759e-01 7.54167914e-01
7.43609667e-01 -1.39960790e+00 1.96572989e-01 -5.24034858e-01
-8.14346731e-01 1.11521578e+00 1.18589222e+00 2.58461326e-01
3.75233948e-01 -2.25550666e-01 -9.61808115e-02 1.02489442e-01
-9.03213084e-01 -7.24079087e-02 1.33240283e-01 4.75591481e-01
5.72088540e-01 4.78226952e-02 -7.42060125e-01 8.55022907e-01
-1.93852961e-01 -2.04741254e-01 4.61105913e-01 8.24221134e-01
-3.43532234e-01 -1.38109303e+00 3.25085944e-03 4.09939349e-01
-8.79491195e-02 -1.56935468e-01 -6.83764517e-01 5.66633701e-01
7.19151497e-02 1.07564855e+00 -2.82393306e-01 -6.21502995e-01
-3.12478375e-02 2.58439332e-01 4.29767847e-01 -8.93681526e-01
-1.00079274e+00 -3.20325583e-01 5.48492193e-01 5.23665585e-02
-3.88595641e-01 -2.22770199e-01 -1.04427826e+00 -1.32678300e-01
-9.85206306e-01 6.76402390e-01 4.73606020e-01 1.12801576e+00
4.38580185e-01 4.97310579e-01 8.93846869e-01 -6.35854900e-01
-1.02959943e+00 -1.30296063e+00 -3.95633310e-01 3.09006453e-01
-6.59638345e-02 -9.29045618e-01 -3.89509201e-01 -2.44044408e-01] | [12.591045379638672, 8.100469589233398] |
8f38738d-d472-4b5c-95d7-9063cf36ab36 | genesis-generative-scene-inference-and | 1907.13052 | null | https://arxiv.org/abs/1907.13052v4 | https://arxiv.org/pdf/1907.13052v4.pdf | GENESIS: Generative Scene Inference and Sampling with Object-Centric Latent Representations | Generative latent-variable models are emerging as promising tools in robotics and reinforcement learning. Yet, even though tasks in these domains typically involve distinct objects, most state-of-the-art generative models do not explicitly capture the compositional nature of visual scenes. Two recent exceptions, MONet and IODINE, decompose scenes into objects in an unsupervised fashion. Their underlying generative processes, however, do not account for component interactions. Hence, neither of them allows for principled sampling of novel scenes. Here we present GENESIS, the first object-centric generative model of 3D visual scenes capable of both decomposing and generating scenes by capturing relationships between scene components. GENESIS parameterises a spatial GMM over images which is decoded from a set of object-centric latent variables that are either inferred sequentially in an amortised fashion or sampled from an autoregressive prior. We train GENESIS on several publicly available datasets and evaluate its performance on scene generation, decomposition, and semi-supervised learning. | ['Oiwi Parker Jones', 'Martin Engelcke', 'Ingmar Posner', 'Adam R. Kosiorek'] | 2019-07-30 | null | https://openreview.net/forum?id=BkxfaTVFwH | https://openreview.net/pdf?id=BkxfaTVFwH | iclr-2020-1 | ['scene-generation', 'unsupervised-object-segmentation'] | ['computer-vision', 'computer-vision'] | [ 4.54413086e-01 9.73883793e-02 1.26105130e-01 -2.77091593e-01
-5.00700712e-01 -7.84384727e-01 1.46089303e+00 -4.62048650e-01
-1.31293666e-02 5.09763539e-01 2.67856777e-01 -7.08603719e-03
-1.96937263e-01 -7.89155066e-01 -8.91679764e-01 -9.75227416e-01
1.40379369e-01 8.72540534e-01 3.58349383e-02 2.66827941e-01
5.02846502e-02 4.40017164e-01 -1.76416719e+00 2.81101435e-01
5.32279551e-01 5.25270283e-01 5.80727041e-01 8.79004717e-01
-5.10391295e-02 8.66899371e-01 -4.07548279e-01 -1.54429227e-01
2.48273939e-01 -6.98181331e-01 -6.94589555e-01 8.91349494e-01
3.15260023e-01 -1.66604713e-01 -3.60026211e-01 8.42655301e-01
1.24894537e-01 9.73097906e-02 1.13461804e+00 -1.27011979e+00
-8.65594506e-01 4.92339373e-01 -2.20338285e-01 -1.90013647e-01
2.43318349e-01 3.35350692e-01 1.01424778e+00 -7.61737168e-01
7.41652369e-01 1.48963416e+00 2.61294901e-01 6.03070080e-01
-2.08254671e+00 -1.40115321e-01 2.73967564e-01 -7.72091076e-02
-9.83396232e-01 -4.64660555e-01 1.16547883e+00 -8.03899050e-01
9.73840654e-01 1.44600242e-01 8.74626100e-01 1.47118700e+00
1.54795840e-01 9.88612711e-01 1.24486065e+00 -2.45302439e-01
4.60149676e-01 7.74808899e-02 -4.83077049e-01 3.98079127e-01
3.57449017e-02 4.88293953e-02 -5.74349046e-01 -8.51818994e-02
9.99428928e-01 1.17069185e-01 2.05624431e-01 -1.10253739e+00
-1.38729048e+00 9.17652130e-01 2.60791123e-01 -1.77764580e-01
-6.63222432e-01 4.05571759e-01 -7.87292346e-02 -2.21160382e-01
5.50295770e-01 4.18171495e-01 -1.10824250e-01 -6.53543649e-03
-8.61567855e-01 5.10536373e-01 6.04523420e-01 1.24384797e+00
9.58828509e-01 3.42071295e-01 -8.63860920e-02 7.22098351e-01
5.62868476e-01 4.27470118e-01 3.56545150e-01 -1.19842505e+00
-9.28603932e-02 5.20520747e-01 -1.99244618e-01 -6.28957748e-01
-6.73864931e-02 -2.86289662e-01 -6.91358387e-01 2.74949461e-01
2.13201623e-02 2.78012063e-02 -1.24252009e+00 1.70493448e+00
2.83130169e-01 1.73728496e-01 2.05702018e-02 5.92342019e-01
7.94995248e-01 5.16018391e-01 1.06349654e-01 -1.39384074e-02
1.02393508e+00 -9.63100910e-01 -3.91619951e-01 -5.14444530e-01
2.39806790e-02 -7.53308594e-01 7.55146205e-01 6.59134209e-01
-1.20974350e+00 -7.50857294e-01 -6.58289671e-01 -1.62616089e-01
-2.62285441e-01 8.17520320e-02 9.79307652e-01 4.79474425e-01
-1.10853052e+00 1.43109575e-01 -1.01729560e+00 -2.64604241e-01
5.60145617e-01 -4.63886932e-02 -3.65012705e-01 -8.04514159e-03
-3.77377003e-01 6.42536938e-01 3.80078644e-01 -3.88753936e-02
-1.97071028e+00 -5.10452688e-01 -1.14400387e+00 -2.38895357e-01
3.53613079e-01 -1.12729049e+00 1.23649490e+00 -8.41111064e-01
-1.78874779e+00 9.02543604e-01 -2.85893410e-01 -5.98475754e-01
4.83030438e-01 -2.77072877e-01 2.03289598e-01 8.75924826e-02
2.29706645e-01 1.03805256e+00 1.28577626e+00 -1.64350867e+00
-3.63847703e-01 -1.47907525e-01 1.53347060e-01 3.12766701e-01
3.09285253e-01 -2.72844702e-01 -3.95754725e-01 -5.13373256e-01
2.78008938e-01 -1.10578108e+00 -4.70167309e-01 -5.83073795e-02
-5.95546722e-01 -5.67850359e-02 7.91741014e-01 -1.95760399e-01
4.33790863e-01 -2.01442552e+00 8.43457103e-01 -2.25885093e-01
3.33896786e-01 -3.49944830e-01 -9.26007554e-02 4.32627589e-01
-8.37238804e-02 -7.90605024e-02 -4.42811638e-01 -9.14979100e-01
2.43715882e-01 6.64908826e-01 -6.13735735e-01 4.67837065e-01
5.07140398e-01 9.75968957e-01 -1.06656253e+00 -3.07744145e-01
8.09296370e-01 6.40606582e-01 -9.13339257e-01 3.46721828e-01
-8.12470138e-01 7.08313942e-01 -2.12693080e-01 3.05857599e-01
5.10541499e-01 -1.86554506e-01 2.08011180e-01 9.15295631e-02
-2.10595392e-02 5.15965462e-01 -1.11001444e+00 2.14517856e+00
-2.93510407e-01 6.54792070e-01 -3.79278809e-01 -1.05137372e+00
8.47451687e-01 1.74642950e-01 6.17912650e-01 -1.39591634e-01
-2.33729676e-04 -2.45796129e-01 -2.51796305e-01 -4.67564702e-01
5.43953538e-01 -4.12739992e-01 -2.40825474e-01 3.97673458e-01
6.89475715e-01 -7.96231806e-01 3.02838773e-01 3.81112754e-01
8.77439737e-01 7.81840563e-01 2.57568091e-01 -5.81856770e-03
8.84293392e-02 -6.61695153e-02 3.80489409e-01 8.36104333e-01
2.37734362e-01 9.39204752e-01 4.80205268e-01 -3.02544594e-01
-1.05612373e+00 -1.60816276e+00 -4.00408916e-02 8.50519121e-01
-4.42960598e-02 -3.54643136e-01 -4.89833623e-01 -4.05583471e-01
1.11814886e-02 9.93171751e-01 -7.52690375e-01 -1.85372606e-01
-3.74067165e-02 -7.02330828e-01 2.33164638e-01 3.88946056e-01
9.55853015e-02 -1.30852532e+00 -8.73570979e-01 3.06708038e-01
-3.98277044e-02 -1.23510289e+00 1.64191946e-01 3.37009519e-01
-8.23145330e-01 -9.08240557e-01 -4.66540337e-01 -4.91147190e-01
7.81310797e-01 4.40126598e-01 1.28579688e+00 -6.24129534e-01
-6.13757253e-01 7.62035966e-01 -2.24129900e-01 -4.52195495e-01
-4.88272578e-01 -1.84796825e-01 3.75487865e-03 1.66345507e-01
2.12215677e-01 -8.24672580e-01 -1.87044606e-01 -1.11611271e-02
-1.03435385e+00 5.36407530e-01 6.77002132e-01 6.70498490e-01
7.93119311e-01 1.79133168e-03 -9.04878508e-03 -1.00617504e+00
2.14404702e-01 -5.56790471e-01 -6.37527764e-01 -1.09899826e-01
-1.09921500e-01 4.09826547e-01 3.43941629e-01 -6.61714673e-01
-1.25023305e+00 2.45773941e-01 2.60865986e-01 -7.43414879e-01
-8.01605284e-01 2.87333041e-01 -2.80567139e-01 3.70243639e-01
6.36253417e-01 6.12085879e-01 -2.11143062e-01 -5.11688650e-01
7.78896093e-01 -2.16588136e-02 5.56236863e-01 -7.28352308e-01
9.84843254e-01 7.76296377e-01 1.33538306e-01 -9.44742978e-01
-6.27589285e-01 -3.53588372e-01 -8.92547369e-01 -1.82313904e-01
1.14029586e+00 -9.45314288e-01 -4.03647542e-01 5.74689806e-01
-1.13312554e+00 -4.73747522e-01 -6.78531051e-01 4.21598107e-01
-1.14163613e+00 4.82194945e-02 -2.72356063e-01 -1.00448918e+00
4.15213406e-01 -1.14474571e+00 1.43127537e+00 -2.80061066e-02
-2.61684537e-01 -9.59207177e-01 2.65196085e-01 1.88351721e-01
4.12716307e-02 5.43318510e-01 8.03844452e-01 -1.82963565e-01
-9.15514171e-01 2.94029620e-02 1.79241568e-01 2.58769631e-01
4.61505055e-01 2.06805199e-01 -1.17417216e+00 -2.58180927e-02
1.52109587e-03 -3.03097129e-01 9.45724666e-01 5.86635232e-01
1.02030754e+00 -2.17838034e-01 -2.04677999e-01 7.57434011e-01
1.40670645e+00 5.01122884e-02 5.99710882e-01 1.31499693e-01
8.26311290e-01 6.87658250e-01 4.78196852e-02 4.39959019e-01
4.30947840e-01 5.05708873e-01 7.56537497e-01 1.45825520e-01
-1.32425070e-01 -5.08323610e-01 3.10400665e-01 5.23862302e-01
-1.12155974e-01 -1.07667975e-01 -8.29753935e-01 9.13724005e-01
-1.77573586e+00 -1.09081173e+00 -4.19156179e-02 1.94521427e+00
7.23360658e-01 1.08049117e-01 1.14724264e-01 -8.01315606e-02
2.49789357e-01 2.87777245e-01 -6.20051742e-01 -6.38897941e-02
-2.72091597e-01 1.25045136e-01 1.30801171e-01 4.18503135e-01
-1.09899533e+00 1.09772623e+00 7.03128004e+00 2.68001884e-01
-8.19100261e-01 -8.33480805e-02 2.46181384e-01 -3.55805188e-01
-5.07953405e-01 4.36539024e-01 -8.05324614e-01 2.39750072e-01
5.74674010e-01 -1.62465665e-02 5.51263809e-01 1.07819223e+00
4.15081605e-02 -1.22286692e-01 -1.33466709e+00 9.31691825e-01
2.74516135e-01 -1.28486145e+00 3.85102779e-01 3.49617094e-01
1.02069783e+00 4.30622883e-02 3.01128536e-01 2.41549924e-01
9.40460026e-01 -1.01240623e+00 1.23034871e+00 7.79853404e-01
3.22271854e-01 -4.28895146e-01 -1.74716637e-02 5.76758862e-01
-8.71767461e-01 1.34110019e-01 -3.77703309e-01 -1.71926290e-01
3.23668092e-01 5.03570020e-01 -8.70335042e-01 3.97179067e-01
5.70686877e-01 9.71074522e-01 -6.87494814e-01 7.90578365e-01
-6.34540319e-01 6.27243459e-01 -1.94057643e-01 2.53054976e-01
2.09029704e-01 -2.93581337e-01 8.16186786e-01 1.05750394e+00
3.53125446e-02 -3.29063505e-01 2.73988843e-01 1.52589607e+00
1.85560077e-01 -5.68000317e-01 -8.24229240e-01 -1.70801058e-01
1.84116617e-01 1.28187978e+00 -7.79027224e-01 -3.30669940e-01
-2.48052940e-01 8.62054110e-01 1.58157617e-01 5.59304237e-01
-8.41129243e-01 2.65578330e-01 7.75104165e-01 -7.50917289e-03
6.43077254e-01 -7.01492846e-01 -2.75688887e-01 -1.24538338e+00
-2.30996475e-01 -8.28510702e-01 -1.33171096e-01 -1.01857948e+00
-1.21184695e+00 2.87219912e-01 4.86965090e-01 -1.06951678e+00
-8.12085927e-01 -5.24866343e-01 -2.93347210e-01 8.09221625e-01
-1.32944894e+00 -1.57521999e+00 -2.50995785e-01 5.30797780e-01
1.01271486e+00 -8.20814669e-02 8.60180676e-01 -3.28896165e-01
-2.67719924e-01 -1.45294622e-01 -8.64758641e-02 -1.95550561e-01
3.70779663e-01 -1.51556861e+00 6.04357362e-01 9.51943994e-01
7.04116404e-01 6.98087096e-01 9.83608544e-01 -5.68788409e-01
-1.45141792e+00 -1.13684726e+00 4.95939910e-01 -8.75309408e-01
3.63786131e-01 -9.02101934e-01 -6.66423142e-01 1.04811370e+00
1.77273840e-01 -1.12821855e-01 6.00699246e-01 1.12141781e-01
-4.75602686e-01 2.93970078e-01 -7.02711105e-01 6.55802369e-01
1.17452502e+00 -7.04084635e-01 -5.65410852e-01 1.73274115e-01
6.22060776e-01 -3.48846078e-01 -4.77142751e-01 1.25751913e-01
3.96149158e-01 -1.13662279e+00 1.22950423e+00 -6.57545269e-01
7.28051364e-01 -4.05795276e-01 -3.93434733e-01 -1.26204193e+00
-6.37183845e-01 -6.60224974e-01 -3.92354220e-01 1.08151984e+00
8.09047520e-02 -4.46525484e-01 7.28420079e-01 3.51201952e-01
-2.25907564e-01 -3.09849054e-01 -5.12693465e-01 -6.68324947e-01
-1.07176915e-01 -7.46309936e-01 4.61717963e-01 7.00503767e-01
-7.31156886e-01 6.03536844e-01 -4.22561795e-01 9.74854603e-02
1.02576351e+00 2.28246450e-01 1.36376941e+00 -1.21003389e+00
-7.56282091e-01 -4.80582356e-01 -6.15926087e-01 -1.20097840e+00
2.55905896e-01 -6.16429687e-01 5.03884196e-01 -1.70260167e+00
3.26636493e-01 -2.19133526e-01 1.90476209e-01 3.32644999e-01
-4.67345910e-03 1.37413979e-01 1.21633306e-01 2.22331002e-01
-5.09092629e-01 8.71164262e-01 1.03406024e+00 -1.11143678e-01
-9.66293886e-02 -1.41040504e-01 -6.00196660e-01 9.57696378e-01
5.43993831e-01 -4.31949943e-01 -7.56291926e-01 -5.82325459e-01
2.90731996e-01 -3.77310097e-01 9.49486971e-01 -8.92524421e-01
4.43926789e-02 -5.61437190e-01 5.49439490e-01 -6.96523488e-01
6.90390229e-01 -6.04625106e-01 6.49896443e-01 1.72658078e-02
-3.05228084e-01 -2.94125199e-01 1.19849995e-01 7.30072260e-01
-4.28057536e-02 5.25984820e-03 5.74751616e-01 -4.70744759e-01
-6.66659772e-01 2.83671230e-01 -6.41402960e-01 -1.95720181e-01
9.54068720e-01 -3.71486634e-01 5.18220514e-02 -2.75676161e-01
-7.32225835e-01 -3.42732131e-01 5.85578561e-01 6.38982832e-01
6.72657728e-01 -1.32938874e+00 -6.60905242e-01 4.44451720e-01
1.40840054e-01 5.76052725e-01 2.91446209e-01 4.54222471e-01
-2.25508183e-01 4.82665032e-01 -2.67991692e-01 -1.13262045e+00
-1.00338042e+00 7.26865113e-01 1.20500714e-01 -6.42258152e-02
-5.71111619e-01 9.62427318e-01 8.72005999e-01 -5.89459300e-01
-1.10170238e-01 -3.26350838e-01 -4.47449554e-03 -8.73553529e-02
1.14025511e-01 1.04526326e-01 -4.18583483e-01 -8.29414189e-01
-4.06850316e-02 3.90403390e-01 1.58493221e-01 -5.73951542e-01
1.57877135e+00 -8.63731503e-02 -1.68549523e-01 1.02994835e+00
8.55351090e-01 -3.03083032e-01 -1.78154576e+00 -3.44239146e-01
-3.17882240e-01 -5.50723732e-01 -8.06685239e-02 -3.29433650e-01
-5.66559494e-01 1.09979665e+00 1.54521734e-01 3.43944132e-01
9.93183136e-01 3.65841717e-01 1.34674743e-01 2.96962354e-02
4.99322325e-01 -7.26581752e-01 4.23920691e-01 4.92642611e-01
9.34106231e-01 -9.68832314e-01 1.04277544e-01 -2.69760251e-01
-6.43467963e-01 7.49364793e-01 4.15461153e-01 -4.40907478e-01
4.71880406e-01 2.04193860e-01 -4.04054523e-01 -2.50336796e-01
-1.13937879e+00 -3.81924272e-01 5.32226861e-01 9.55031335e-01
3.29648137e-01 1.86768815e-01 3.68790060e-01 1.15295999e-01
-5.00599623e-01 -2.88248658e-01 4.04508114e-01 9.92327273e-01
-2.84720331e-01 -9.25752282e-01 -2.80998319e-01 2.58627266e-01
-1.38251148e-02 -2.58789305e-02 -4.71656710e-01 4.85038489e-01
2.90994376e-01 6.50154650e-01 2.01379508e-01 -1.27469927e-01
2.51406934e-02 -4.00181971e-02 9.33399975e-01 -1.07008743e+00
-5.08635268e-02 4.86633778e-01 -2.53148079e-01 -4.44564939e-01
-6.53106153e-01 -1.11874819e+00 -7.71697998e-01 1.26316741e-01
-4.25994545e-02 -2.38572925e-01 7.71550953e-01 9.81716454e-01
2.18443513e-01 8.51847112e-01 3.95536125e-01 -1.41181612e+00
-1.62870139e-01 -9.92936254e-01 -4.48747486e-01 4.54022288e-01
3.40147525e-01 -9.34759080e-01 -3.27364057e-01 7.97812998e-01] | [10.06421184539795, 0.2712155878543854] |
0c3f9f0a-daaa-46de-a4f0-29711dc18e3e | cuni-system-for-wmt16-automatic-post-editing | 1606.07481 | null | http://arxiv.org/abs/1606.07481v1 | http://arxiv.org/pdf/1606.07481v1.pdf | CUNI System for WMT16 Automatic Post-Editing and Multimodal Translation Tasks | Neural sequence to sequence learning recently became a very promising
paradigm in machine translation, achieving competitive results with statistical
phrase-based systems. In this system description paper, we attempt to utilize
several recently published methods used for neural sequential learning in order
to build systems for WMT 2016 shared tasks of Automatic Post-Editing and
Multimodal Machine Translation. | ['Ondřej Bojar', 'Marek Tlustý', 'Jindřich Helcl', 'Jindřich Libovický', 'Pavel Pecina'] | 2016-06-23 | cuni-system-for-wmt16-automatic-post-editing-1 | https://aclanthology.org/W16-2361 | https://aclanthology.org/W16-2361.pdf | ws-2016-8 | ['multimodal-machine-translation'] | ['natural-language-processing'] | [ 7.87316978e-01 -2.76966602e-01 -5.36591828e-01 -2.96277493e-01
-1.28074586e+00 -4.64388251e-01 9.40023959e-01 6.13007061e-02
-7.44288862e-01 1.15986896e+00 4.38545793e-01 -8.64374697e-01
4.36936557e-01 1.12940006e-01 -8.65688622e-01 -2.09351599e-01
3.77159894e-01 9.75934029e-01 -2.24688947e-01 -6.24239326e-01
3.53417188e-01 1.31719425e-01 -9.92079556e-01 1.09626997e+00
7.48476803e-01 3.94704968e-01 2.46601135e-01 8.41874182e-01
-3.03616077e-01 5.18304527e-01 -6.05024099e-01 -8.42802584e-01
1.32636860e-01 -8.03213239e-01 -8.57660413e-01 -5.12732208e-01
8.27423751e-01 1.36484116e-01 -9.07591805e-02 8.41542959e-01
1.02321637e+00 1.02925375e-01 8.30308139e-01 -5.64899266e-01
-1.07634747e+00 1.22064292e+00 -3.94584894e-01 3.28086823e-01
7.90211201e-01 3.70542370e-02 1.14918542e+00 -1.39202440e+00
8.22497249e-01 1.32607329e+00 4.73942131e-01 1.01068592e+00
-1.23768520e+00 -5.01055241e-01 -4.00695890e-01 5.17740369e-01
-1.06019342e+00 -8.47633243e-01 3.67821097e-01 -1.09964274e-02
1.84365916e+00 4.72919077e-01 1.87072322e-01 1.59663773e+00
6.32483959e-01 1.13185215e+00 1.17863595e+00 -7.81273901e-01
-6.76948652e-02 7.93598220e-02 -2.73437262e-01 5.55494964e-01
-4.00584221e-01 8.87181312e-02 -9.44240034e-01 -2.72429049e-01
2.80162632e-01 -6.42135859e-01 8.46644193e-02 3.67339432e-01
-1.82337320e+00 7.52542853e-01 -3.54916081e-02 4.90309119e-01
-4.55441773e-01 1.17863119e-01 7.60120332e-01 8.72505784e-01
6.02065265e-01 8.36978972e-01 -7.43698597e-01 -4.21328872e-01
-1.00954449e+00 1.54863864e-01 7.26543307e-01 9.50239778e-01
2.31814578e-01 5.90576492e-02 -6.95516825e-01 1.05874813e+00
7.60916695e-02 8.19052637e-01 8.04426968e-01 -6.49121165e-01
9.67648089e-01 6.95337355e-02 -5.43729067e-02 -4.19567406e-01
-2.21688017e-01 -2.12290719e-01 -9.47545707e-01 -4.59962428e-01
1.58495277e-01 -2.68724412e-01 -1.07142138e+00 1.43986082e+00
-2.22045019e-01 -4.41310890e-02 2.03915387e-01 7.42544770e-01
8.66964102e-01 1.01613450e+00 1.21381372e-01 -5.13092518e-01
9.82280016e-01 -1.16021347e+00 -8.40479553e-01 -1.71400800e-01
7.03852057e-01 -1.39612436e+00 1.00032640e+00 2.09650129e-01
-1.38664448e+00 -4.55924183e-01 -7.20078290e-01 -7.17312694e-02
-3.44371229e-01 2.84739316e-01 4.54384625e-01 3.90537590e-01
-1.49494660e+00 6.87447309e-01 -5.21498680e-01 -7.47631729e-01
1.24885589e-01 6.31613016e-01 -4.45808113e-01 1.93951026e-01
-1.53879809e+00 1.66966045e+00 4.89297450e-01 1.40596986e-01
-6.50335491e-01 -3.43550920e-01 -6.17808878e-01 -2.48800054e-01
-2.42337093e-01 -1.01119673e+00 1.74595010e+00 -9.09919977e-01
-2.01685333e+00 1.10609770e+00 -7.30389237e-01 -8.58349681e-01
3.36710125e-01 -1.77891850e-01 -4.65403676e-01 -1.84204817e-01
-2.10349888e-01 8.54438663e-01 8.42104375e-01 -6.74843609e-01
-6.18360698e-01 -3.47734652e-02 -4.93540853e-01 3.14396530e-01
-2.70748019e-01 7.12798476e-01 -1.48221985e-01 -4.97747958e-01
-4.12579209e-01 -1.06086779e+00 -3.18191141e-01 -7.93155849e-01
-3.21869880e-01 -6.37613773e-01 2.42406368e-01 -1.10293853e+00
1.07384253e+00 -1.57979739e+00 9.07781124e-01 -2.00961351e-01
-3.35488945e-01 5.79589427e-01 -7.28329599e-01 7.47402728e-01
1.17881350e-01 1.54377490e-01 -4.09256399e-01 -7.30791092e-01
2.72459947e-02 1.42162696e-01 -4.73631471e-01 3.83239836e-02
3.92916709e-01 1.46538579e+00 -8.59633267e-01 -4.53535646e-01
9.32555795e-02 2.12943822e-01 -1.21062314e-02 2.21234649e-01
-4.54659522e-01 6.01556718e-01 -1.59446239e-01 6.55556798e-01
2.41857097e-01 2.33146638e-01 2.00783819e-01 1.27765909e-01
-1.32340446e-01 6.61272168e-01 -9.26740021e-02 2.21800756e+00
-4.72164780e-01 6.45566106e-01 -2.83303559e-01 -7.31182992e-01
7.19963193e-01 8.05702567e-01 1.60310805e-01 -1.01259923e+00
2.58487880e-01 6.92979753e-01 1.10709615e-01 -4.94941711e-01
7.17751205e-01 -2.90215969e-01 -1.18139826e-01 4.56625074e-01
4.66160566e-01 -3.54462236e-01 2.64513671e-01 1.14134680e-02
8.50476384e-01 3.72244477e-01 4.12727684e-01 -1.27040491e-01
6.21199369e-01 3.29070598e-01 1.50680155e-01 6.91261172e-01
1.08566927e-03 6.69494629e-01 1.51410280e-02 -5.05713344e-01
-1.44190812e+00 -9.10889506e-01 4.48331803e-01 1.47830784e+00
-5.65480590e-01 -2.49466971e-01 -7.01586723e-01 -6.52974904e-01
-2.55175501e-01 7.84882188e-01 -5.11636436e-01 -1.18727103e-01
-1.08368826e+00 -6.06130540e-01 8.24191689e-01 2.49530792e-01
-2.60595512e-02 -1.38046944e+00 2.97655314e-01 4.68412757e-01
-4.50910032e-01 -1.06805003e+00 -7.90643275e-01 2.94185758e-01
-9.18381035e-01 -2.17075631e-01 -1.19805431e+00 -1.25839472e+00
1.21599913e-01 1.96768478e-01 1.39103258e+00 -1.04225546e-01
1.73677564e-01 -1.45764858e-01 -4.71058279e-01 -6.23097181e-01
-9.39687788e-01 7.86612451e-01 4.10078764e-01 -2.48733357e-01
5.67297757e-01 -3.48135948e-01 -8.75790603e-04 4.71219718e-02
-6.32924795e-01 2.35322177e-01 9.62700248e-01 9.23427999e-01
3.28740746e-01 -1.17851686e+00 5.43876648e-01 -5.88931382e-01
1.15560377e+00 -3.22569847e-01 -2.50959665e-01 5.98858058e-01
-4.80628282e-01 1.36630923e-01 6.97372913e-01 -5.30822098e-01
-8.53797615e-01 4.65287603e-02 -4.62244272e-01 -1.75753862e-01
7.20152492e-03 8.69118214e-01 3.58364373e-01 -1.29950223e-02
6.32482052e-01 6.49275422e-01 -5.81128933e-02 -5.77042401e-01
5.12398303e-01 8.80719662e-01 3.57826352e-01 -4.87664223e-01
6.15245342e-01 -2.53753215e-01 -1.88430443e-01 -4.15375531e-01
-6.86012506e-01 -3.12045097e-01 -8.59942853e-01 1.80224583e-01
9.24889266e-01 -6.58237219e-01 -4.21388894e-01 2.54123986e-01
-1.70718288e+00 -2.97893792e-01 2.08835155e-01 4.59130824e-01
-5.67740321e-01 1.09437242e-01 -8.84268582e-01 -5.09043455e-01
-8.22445214e-01 -1.07042885e+00 1.07984400e+00 2.89370082e-02
-5.69467902e-01 -1.06031096e+00 5.24358034e-01 4.85246390e-01
8.05641294e-01 -3.60138386e-01 9.69551206e-01 -8.56728375e-01
-1.19361006e-01 1.17878161e-01 -1.12816803e-02 2.92783588e-01
1.74130239e-02 -1.57670975e-01 -4.82576966e-01 -3.56497377e-01
-3.82144421e-01 -4.18031037e-01 9.45900679e-01 2.62481242e-01
5.83586991e-01 -1.91895932e-01 -1.33068323e-01 4.07158673e-01
1.01556051e+00 8.93417820e-02 4.12308782e-01 2.78183669e-01
7.10780144e-01 5.72312832e-01 3.74378532e-01 -3.03637087e-02
2.31007367e-01 8.61611068e-01 -9.03738961e-02 -7.49051422e-02
-3.09909552e-01 -2.50966817e-01 8.56914461e-01 1.35729766e+00
-1.46698818e-01 -2.94415474e-01 -1.15965390e+00 4.95198727e-01
-1.93731165e+00 -8.35023701e-01 -1.75067097e-01 1.78238869e+00
1.21675444e+00 -1.10723838e-01 -3.92290801e-02 -7.03287601e-01
7.12065935e-01 -8.91868845e-02 -1.86924413e-01 -1.14893210e+00
-5.97225964e-01 7.19550550e-01 5.56241870e-01 6.03861511e-01
-7.83873260e-01 1.50063157e+00 7.65249872e+00 9.10116017e-01
-1.10556686e+00 6.00586057e-01 3.33520532e-01 -1.44934162e-01
-2.35891148e-01 -3.07728469e-01 -8.01863611e-01 2.98745602e-01
1.55808854e+00 -1.41497865e-01 7.27410853e-01 7.69777447e-02
2.18561798e-01 2.28575855e-01 -1.34331954e+00 9.08931077e-01
3.18441480e-01 -1.66450036e+00 5.74156582e-01 -3.75332564e-01
1.21514893e+00 7.86736608e-01 2.71445036e-01 6.20115042e-01
3.04291099e-01 -1.27665782e+00 4.51395005e-01 3.72011214e-01
9.22368884e-01 -6.27427578e-01 9.88136292e-01 3.08281690e-01
-5.08454025e-01 1.74592376e-01 -3.53832632e-01 -2.01082215e-01
2.87976503e-01 7.75542781e-02 -1.20166469e+00 6.06411755e-01
1.96283817e-01 7.31219411e-01 -4.31983888e-01 8.47439349e-01
-3.47092807e-01 8.49367976e-01 4.86569628e-02 -4.60652322e-01
4.64421958e-01 3.47130299e-02 7.10090041e-01 1.69585085e+00
4.12355006e-01 -3.93193513e-01 1.02102123e-01 2.73066968e-01
-6.08152330e-01 6.19666994e-01 -6.94939435e-01 -4.21845347e-01
9.28281993e-02 1.00706720e+00 -1.00141779e-01 -5.67818046e-01
-4.14182812e-01 1.54885519e+00 4.04480785e-01 4.59980965e-01
-7.26656199e-01 -1.21002071e-01 5.36516905e-01 -4.09715801e-01
9.51598212e-02 -6.11128271e-01 -4.30039823e-01 -1.49794412e+00
-1.55128598e-01 -1.35808086e+00 1.58517927e-01 -6.73242152e-01
-1.35781491e+00 9.10716057e-01 -2.42832467e-01 -1.08993959e+00
-5.67832351e-01 -6.30250156e-01 -4.96271789e-01 1.32575786e+00
-1.23808146e+00 -1.48212242e+00 7.79867351e-01 3.57848495e-01
9.88799572e-01 -8.57993424e-01 1.24676180e+00 2.07093433e-01
-2.70268440e-01 7.87398040e-01 5.60493767e-01 -6.03244975e-02
1.37678874e+00 -1.16525924e+00 1.12108648e+00 6.95186019e-01
6.22082233e-01 7.00557888e-01 7.93463171e-01 -6.68220818e-01
-1.87936592e+00 -8.77634287e-01 1.84054255e+00 -8.33633006e-01
8.20885897e-01 -6.17607594e-01 -6.53777838e-01 4.96250272e-01
1.05783916e+00 -7.69751608e-01 7.03994513e-01 2.02858821e-01
-7.16694966e-02 1.66856915e-01 -7.70191312e-01 7.57005215e-01
8.88530970e-01 -8.54281843e-01 -1.00117755e+00 4.95520055e-01
1.05371058e+00 -4.60473984e-01 -6.16522133e-01 5.84275544e-01
4.83133405e-01 -4.96151671e-02 6.78005457e-01 -1.02642941e+00
6.31062627e-01 -1.07023455e-01 -6.79003820e-02 -1.79623628e+00
-3.21463287e-01 -1.33200336e+00 -2.67021000e-01 8.98779571e-01
1.25170922e+00 -4.33636397e-01 5.89993477e-01 5.81620634e-02
-1.77197754e-01 -7.72914231e-01 -1.36725175e+00 -6.00205123e-01
5.58967412e-01 -1.05274066e-01 3.09013247e-01 9.48316276e-01
2.33241886e-01 9.88734663e-01 -8.58209074e-01 -5.07530332e-01
1.44088581e-01 -1.72377497e-01 5.55467725e-01 -7.15873659e-01
-4.72147882e-01 -8.17535400e-01 5.60752712e-02 -1.03346980e+00
3.89863282e-01 -1.41523373e+00 1.30059227e-01 -1.56137764e+00
4.69994098e-01 3.75130445e-01 -6.21178508e-01 5.19205213e-01
-3.00290704e-01 7.09341943e-01 1.94384426e-01 3.85090113e-01
-5.23204625e-01 5.11881113e-01 1.13449991e+00 -5.37625670e-01
-1.17009759e-01 6.06102571e-02 -4.55870450e-01 -4.88119274e-02
8.93355846e-01 -4.87517506e-01 2.79073179e-01 -1.11169839e+00
4.54659820e-01 8.53361711e-02 -1.25821650e-01 -4.38353688e-01
3.05039704e-01 -1.25621021e-01 1.97148636e-01 -6.09660864e-01
2.11663693e-01 -4.07629609e-01 -3.18491310e-01 5.01477122e-01
-8.23104799e-01 6.30372584e-01 3.98981124e-01 4.03065607e-02
-3.89010519e-01 -1.46109655e-01 3.36578190e-01 -3.35764498e-01
-5.45503080e-01 1.61078244e-01 -7.50749588e-01 -2.71888226e-01
4.12223220e-01 2.74354607e-01 -1.24701560e-01 -3.93026143e-01
-6.29186511e-01 2.63119727e-01 1.67901173e-01 9.11667645e-01
6.61873877e-01 -1.25052476e+00 -1.51138103e+00 -3.21304828e-01
1.92121476e-01 -9.85323071e-01 -2.55514681e-01 1.21687162e+00
-2.96555400e-01 1.03239679e+00 -4.42220867e-01 -5.64265013e-01
-1.43211508e+00 2.91057080e-01 5.93542047e-02 -4.05175894e-01
-2.93930713e-02 1.09371340e+00 -7.00214088e-01 -9.29846942e-01
1.04216747e-01 8.68659541e-02 -8.38541090e-02 -8.65076482e-02
5.62576771e-01 -5.26290499e-02 4.29726213e-01 -5.55352688e-01
-2.94965386e-01 1.38541058e-01 -4.82594460e-01 -7.05341995e-01
1.21286690e+00 -1.65083811e-01 -5.75041950e-01 7.75786817e-01
1.08528960e+00 -7.50277936e-02 -3.34753871e-01 -6.44956648e-01
4.46394324e-01 2.39390433e-02 -2.32816696e-01 -1.44805729e+00
-1.51434869e-01 1.06181407e+00 4.61607695e-01 -3.44288796e-01
9.61645901e-01 -7.60239735e-03 1.15302932e+00 8.66659582e-01
4.29358125e-01 -1.22444534e+00 -4.18527365e-01 1.39519715e+00
9.83967006e-01 -1.55073237e+00 -3.42635214e-01 2.78276175e-01
-6.54634714e-01 1.47275269e+00 2.63633490e-01 1.63672045e-01
-7.63243809e-02 5.42779341e-02 3.03579420e-01 2.67071009e-01
-1.10951734e+00 -5.27140126e-02 8.48012567e-01 4.14623410e-01
1.10457754e+00 3.27723771e-01 -8.93715560e-01 -2.23537758e-02
-2.54040092e-01 -1.60092369e-01 1.95147991e-01 6.49158180e-01
-2.45056942e-01 -1.86031902e+00 -8.06453750e-02 2.42544428e-01
-7.84749508e-01 -8.39421690e-01 -1.11212504e+00 9.97094288e-02
-2.42385402e-01 8.09142292e-01 -2.74056762e-01 -6.80385292e-01
6.19473867e-02 4.72363383e-01 9.37271535e-01 -1.01504695e+00
-1.16075408e+00 3.03192846e-02 5.26000977e-01 -4.04218584e-01
-5.01808226e-01 -8.66716862e-01 -8.14710081e-01 -1.92556635e-01
-1.51477590e-01 2.95531899e-01 1.14429390e+00 1.12010336e+00
3.69273156e-01 1.88098013e-01 5.44321001e-01 -8.90729308e-01
-6.00071490e-01 -1.67542386e+00 3.23545516e-01 4.75735515e-02
1.75228924e-01 1.42599106e-01 -8.36966652e-03 9.86811817e-02] | [11.60059642791748, 10.382320404052734] |
92aa6117-fc85-492f-867e-670bac0e66a4 | personalized-federated-learning-with-hidden | 2211.10684 | null | https://arxiv.org/abs/2211.10684v2 | https://arxiv.org/pdf/2211.10684v2.pdf | Personalized Federated Learning with Hidden Information on Personalized Prior | Federated learning (FL for simplification) is a distributed machine learning technique that utilizes global servers and collaborative clients to achieve privacy-preserving global model training without direct data sharing. However, heterogeneous data problem, as one of FL's main problems, makes it difficult for the global model to perform effectively on each client's local data. Thus, personalized federated learning (PFL for simplification) aims to improve the performance of the model on local data as much as possible. Bayesian learning, where the parameters of the model are seen as random variables with a prior assumption, is a feasible solution to the heterogeneous data problem due to the tendency that the more local data the model use, the more it focuses on the local data, otherwise focuses on the prior. When Bayesian learning is applied to PFL, the global model provides global knowledge as a prior to the local training process. In this paper, we employ Bayesian learning to model PFL by assuming a prior in the scaled exponential family, and therefore propose pFedBreD, a framework to solve the problem we model using Bregman divergence regularization. Empirically, our experiments show that, under the prior assumption of the spherical Gaussian and the first order strategy of mean selection, our proposal significantly outcompetes other PFL algorithms on multiple public benchmarks. | ['Jiancheng Lv', 'Qing Ye', 'Yuhao Zhou', 'Mingjia Shi'] | 2022-11-19 | null | null | null | null | ['personalized-federated-learning'] | ['methodology'] | [-7.20593095e-01 -1.73357710e-01 -4.14491296e-01 -4.86671627e-01
-8.88528228e-01 -4.49564159e-01 2.08254442e-01 2.41631083e-03
-3.47945035e-01 8.70770812e-01 1.08403914e-01 -3.21548194e-01
-2.44116843e-01 -8.55844021e-01 -7.80565083e-01 -1.19204521e+00
1.43457711e-01 5.83644271e-01 -4.41563129e-03 2.10335612e-01
8.31113756e-03 4.10135925e-01 -1.10339987e+00 2.97295839e-01
8.90236557e-01 1.13898933e+00 -2.53606550e-02 1.30425572e-01
-2.74058759e-01 9.75871980e-01 -4.55461591e-01 -6.80520833e-01
5.50315976e-01 -6.51908480e-03 -7.60331631e-01 -2.81735659e-01
2.61197031e-01 -6.12784147e-01 -1.76013947e-01 1.32206059e+00
3.69707942e-01 1.37304604e-01 5.30626893e-01 -1.39589834e+00
-4.01862741e-01 5.91954231e-01 -4.33436960e-01 -1.26082867e-01
-2.67189234e-01 3.23632844e-02 1.02525306e+00 -5.50846994e-01
5.01300991e-01 1.04303813e+00 6.91365719e-01 4.47277844e-01
-1.22446704e+00 -5.35661936e-01 2.95049846e-01 3.04677486e-01
-1.56915343e+00 -2.55766481e-01 7.95295835e-01 -2.54959524e-01
2.41953209e-01 3.93663675e-01 9.06722844e-02 9.64519024e-01
3.66119266e-01 8.98157239e-01 1.10365653e+00 -1.42666370e-01
7.67739117e-01 5.80546618e-01 2.03590572e-01 2.19112962e-01
1.30345389e-01 -3.09774838e-02 -5.84961236e-01 -9.89762843e-01
1.68112069e-01 4.89733428e-01 -3.52220774e-01 -7.89139867e-01
-2.86113173e-01 1.02837372e+00 2.10794151e-01 -9.83109977e-03
-4.95249569e-01 -8.87934566e-02 4.11132574e-01 3.35169822e-01
6.31137013e-01 -3.25389802e-01 -8.02691936e-01 1.50996119e-01
-7.93552637e-01 1.43462539e-01 1.09441566e+00 8.67721200e-01
1.07632089e+00 -4.59149629e-01 -1.27501875e-01 5.96018255e-01
5.53264856e-01 2.94540137e-01 4.35493231e-01 -1.07725143e+00
4.84317213e-01 4.90133107e-01 2.35410929e-01 -8.14162254e-01
1.72352612e-01 -4.82629061e-01 -9.92435932e-01 4.39057648e-02
5.61056852e-01 -2.52580494e-01 -2.16964737e-01 1.80293393e+00
7.60632932e-01 1.26246035e-01 1.11697264e-01 8.82525206e-01
1.93652779e-01 5.07639289e-01 4.05980535e-02 -4.29324627e-01
9.37939346e-01 -7.49668360e-01 -7.29915857e-01 3.45191777e-01
8.38336587e-01 -3.53381634e-01 7.40110993e-01 7.08861709e-01
-5.67531645e-01 -2.06441600e-02 -5.84790409e-01 1.55614376e-01
-2.67072380e-01 -3.42051148e-01 5.05418301e-01 8.20222974e-01
-9.32766557e-01 5.21483123e-01 -8.30786228e-01 -2.03262284e-01
6.78481698e-01 2.39686906e-01 -3.35219443e-01 -4.46712911e-01
-1.00973821e+00 5.04515409e-01 6.05243668e-02 -2.71127403e-01
-9.44237292e-01 -8.17976832e-01 -1.64575636e-01 3.06534111e-01
4.88035887e-01 -7.75785804e-01 1.15737402e+00 -6.41982913e-01
-1.38206410e+00 4.04822409e-01 -9.57560092e-02 -5.70079565e-01
9.29455757e-01 -9.67682898e-02 -4.11784798e-02 -2.34109666e-02
-4.22435760e-01 -1.97569668e-01 9.19808269e-01 -1.27317357e+00
-7.32870579e-01 -7.68973231e-01 -2.68683076e-01 -6.38809502e-02
-6.68495595e-01 4.27318402e-02 -4.21950936e-01 -3.28732133e-01
-1.77297607e-01 -7.23971725e-01 -2.61878759e-01 2.61736304e-01
-1.44871343e-02 -5.96541822e-01 1.23577428e+00 -7.59076238e-01
1.18491971e+00 -2.42540836e+00 -1.09497838e-01 5.74487567e-01
3.08215737e-01 1.30966306e-01 2.61787653e-01 4.63307738e-01
4.45633709e-01 1.82288349e-01 -3.38561609e-02 -6.12775624e-01
2.12747872e-01 3.76496494e-01 -2.88568050e-01 8.52547050e-01
-7.78839946e-01 3.08680892e-01 -5.53636909e-01 -5.48328876e-01
-1.42635405e-01 4.22379315e-01 -7.11633801e-01 5.19772410e-01
-3.67396712e-01 3.74735713e-01 -6.78428888e-01 3.23767304e-01
1.19484341e+00 -3.68752241e-01 5.19178689e-01 -1.15962468e-01
8.46325159e-02 -1.82446882e-01 -1.36944199e+00 1.51690292e+00
-4.37768728e-01 -1.99985743e-01 8.14032733e-01 -9.43270028e-01
1.01629210e+00 4.66623336e-01 8.08355451e-01 -1.99493155e-01
1.99222267e-02 1.92356423e-01 -4.61752415e-01 -4.49986190e-01
-2.85804868e-01 5.34519777e-02 1.82906911e-01 5.68106234e-01
-1.71602547e-01 5.13552129e-01 -4.02075648e-01 3.01548332e-01
1.10883427e+00 -1.41874135e-01 5.53329587e-02 -3.04975510e-01
4.73203450e-01 -4.14394766e-01 1.01277864e+00 8.25105309e-01
-3.79698634e-01 3.20566356e-01 4.47507590e-01 -6.80466056e-01
-5.59374750e-01 -8.88460159e-01 -1.53009832e-01 1.27126467e+00
1.53880149e-01 -4.13905531e-01 -6.96788371e-01 -1.13898087e+00
4.56746459e-01 8.31961989e-01 -3.83662462e-01 -6.19557165e-02
3.40769580e-03 -7.48527527e-01 1.98702931e-01 8.82727578e-02
6.06603682e-01 -3.88822585e-01 -2.56918073e-01 1.33013338e-01
-1.85393602e-01 -5.86597562e-01 -6.48790002e-01 3.02502960e-01
-7.70924449e-01 -1.01486397e+00 -3.08148086e-01 -2.48487294e-01
4.61642146e-01 2.20191389e-01 7.22693205e-01 -1.30705908e-01
-8.73387754e-02 4.45699155e-01 -2.31696025e-01 -1.65847242e-01
-3.20726156e-01 -5.57131246e-02 -2.05611140e-02 6.15334928e-01
5.92218637e-01 -6.44272506e-01 -6.97780371e-01 4.44414973e-01
-9.23424125e-01 -3.67604077e-01 1.95942193e-01 8.73528004e-01
4.86200631e-01 1.27095461e-01 6.06554627e-01 -1.10841537e+00
5.13582945e-01 -9.12650883e-01 -6.67276800e-01 7.28456020e-01
-9.87634242e-01 -5.62258922e-02 9.44901526e-01 -3.40008199e-01
-1.35257900e+00 -1.45973429e-01 2.29730308e-01 -9.56310093e-01
-1.76970735e-01 4.18936133e-01 -6.56932712e-01 -4.56122458e-02
5.94086170e-01 1.60630286e-01 3.50317895e-01 -1.01619959e+00
2.36902535e-01 1.09915662e+00 6.97460324e-02 -1.02137709e+00
3.74672413e-01 5.88945627e-01 -7.14987740e-02 -4.38746303e-01
-5.30214429e-01 -5.61173558e-01 -3.64128798e-01 5.17517254e-02
5.43442607e-01 -9.20545816e-01 -1.03403902e+00 4.91055399e-01
-9.37383771e-01 -9.75652188e-02 -2.48762310e-01 4.27138507e-01
-4.91874874e-01 5.37024379e-01 -4.33476031e-01 -1.12879980e+00
-5.78016043e-01 -9.26359892e-01 4.78804290e-01 8.07284564e-02
2.54865944e-01 -1.19643128e+00 1.60831183e-01 6.45181775e-01
7.09556103e-01 -5.43031394e-02 7.92084694e-01 -1.15987396e+00
-6.51108682e-01 -2.95949817e-01 2.63518225e-02 8.11619341e-01
1.48457468e-01 -2.12655753e-01 -1.00421000e+00 -6.68545842e-01
5.28748631e-01 -3.36833298e-01 4.77379560e-01 7.85418041e-03
1.71772242e+00 -8.16457868e-01 -1.86323956e-01 8.48271430e-01
1.51432562e+00 -1.59482956e-01 3.16844285e-01 2.54758354e-02
4.76074338e-01 5.63412905e-01 5.66562831e-01 1.11654258e+00
5.35399199e-01 4.68526989e-01 6.63445830e-01 2.29886979e-01
3.70147884e-01 -3.54261994e-01 4.44747746e-01 5.72450340e-01
3.79853040e-01 -2.13497624e-01 -7.05790401e-01 3.57383013e-01
-2.18432379e+00 -8.86577368e-01 2.21345201e-03 2.53129053e+00
9.24891591e-01 -6.22744322e-01 3.43108103e-02 -4.76968825e-01
7.87136912e-01 -1.55456811e-01 -8.11004579e-01 -2.51030594e-01
1.73514374e-02 -1.60541385e-01 6.79033279e-01 1.90849826e-01
-9.62997973e-01 5.05779862e-01 4.89163876e+00 1.14012313e+00
-1.02992344e+00 4.79254842e-01 8.01114798e-01 -1.89559922e-01
-4.51805294e-01 1.42762318e-01 -8.96017611e-01 8.88372242e-01
9.00170445e-01 -3.56235325e-01 6.87577963e-01 1.17700827e+00
1.60341948e-01 7.41206184e-02 -1.22469985e+00 1.03383970e+00
-3.09922159e-01 -1.04641259e+00 -5.87024763e-02 3.99565935e-01
7.87932575e-01 3.34623992e-01 7.56880045e-02 2.07409590e-01
7.63038397e-01 -6.27810001e-01 3.71986687e-01 8.12615871e-01
3.03361416e-01 -8.86010110e-01 7.90785313e-01 9.74173844e-01
-5.70684433e-01 -3.93953800e-01 -5.27344465e-01 4.33462203e-01
-2.79418230e-01 7.81831384e-01 -6.33480191e-01 7.85223663e-01
9.80786324e-01 2.97364146e-01 -2.59207517e-01 1.13172054e+00
2.24319071e-01 8.28277707e-01 -6.33525312e-01 3.04683417e-01
-1.71956390e-01 -5.63194871e-01 3.05192769e-01 7.87456512e-01
2.95184612e-01 -1.03537306e-01 5.44639707e-01 6.23998642e-01
-2.69152224e-01 6.73393250e-01 -4.22334015e-01 3.92284989e-01
6.48706198e-01 1.11311340e+00 2.64143527e-01 -1.30345657e-01
-5.51902354e-01 6.63220882e-01 6.60099506e-01 4.88576502e-01
-5.69186807e-01 -1.11579761e-01 8.13410521e-01 -4.17276062e-02
3.49138200e-01 1.30600631e-01 -1.79884762e-01 -1.35127831e+00
1.78538904e-01 -1.05055118e+00 1.03128481e+00 -3.69719863e-02
-1.97022724e+00 1.09272972e-01 -2.57795960e-01 -8.57575655e-01
-1.12950005e-01 -2.45185837e-01 -5.99533200e-01 1.02487171e+00
-1.45250964e+00 -1.22826207e+00 2.10721605e-02 1.24504757e+00
-1.63531780e-01 -1.44055754e-01 8.63223433e-01 3.48587275e-01
-6.08117580e-01 9.33739305e-01 9.58808124e-01 -7.78171346e-02
1.10924721e+00 -8.44006181e-01 -5.98911166e-01 4.90370810e-01
-1.33612901e-01 7.81252325e-01 5.33529520e-01 -5.47298789e-01
-1.61882830e+00 -1.17540622e+00 5.60082674e-01 -1.80403948e-01
5.01709342e-01 -2.18694240e-01 -1.18082404e+00 5.90105176e-01
-2.64596790e-02 5.24074137e-01 1.13511157e+00 2.97190279e-01
-5.65177321e-01 -7.84769416e-01 -1.88542080e+00 -1.20106656e-02
5.42900085e-01 -4.88518238e-01 -3.07718292e-02 5.30967236e-01
5.03301501e-01 1.07503757e-01 -1.23122239e+00 1.23729780e-01
2.66044080e-01 -1.03441501e+00 4.45228159e-01 -8.58455777e-01
-3.35060716e-01 -1.86612412e-01 -6.21917546e-01 -1.13623667e+00
-2.41615057e-01 -8.60641241e-01 -5.56394577e-01 1.54848242e+00
-5.80196977e-02 -1.05119693e+00 8.41807604e-01 1.13504648e+00
3.25749099e-01 -6.13813758e-01 -1.33557606e+00 -7.78841019e-01
3.23935449e-01 -2.47316822e-01 9.86801624e-01 1.04892492e+00
-1.82916075e-01 -4.66228545e-01 -5.25769591e-01 3.72462422e-01
1.07316554e+00 1.93626449e-01 9.95993555e-01 -1.31397438e+00
-4.99439120e-01 7.49436170e-02 2.86728498e-02 -9.43064332e-01
3.99579048e-01 -9.07331944e-01 -2.10500374e-01 -1.03947127e+00
3.63878936e-01 -7.99760342e-01 -6.53040886e-01 6.10717177e-01
1.35276407e-01 -5.36431730e-01 2.16451705e-01 4.85090971e-01
-7.58953929e-01 7.16671765e-01 9.04309988e-01 -1.39748588e-01
-5.20023890e-02 5.41936338e-01 -7.51508713e-01 3.79664719e-01
5.98850489e-01 -5.98925710e-01 -3.57647061e-01 -2.94403046e-01
-1.88416727e-02 1.09573729e-01 2.30106503e-01 -5.18149793e-01
7.69834518e-01 -5.31432092e-01 7.90345073e-02 -4.72775459e-01
2.47008484e-02 -1.40383887e+00 4.91058618e-01 2.41153002e-01
-2.24540830e-01 -5.83719432e-01 -4.01145518e-01 8.27936053e-01
-1.63479820e-01 -2.27066621e-01 8.52587819e-01 -8.96930695e-03
-9.86160561e-02 7.53266692e-01 -4.80117202e-02 -1.86364874e-02
1.09536231e+00 3.91493201e-01 -3.82546395e-01 -5.16908884e-01
-7.83171296e-01 5.56306124e-01 6.34312570e-01 3.73038761e-02
2.42148459e-01 -1.12916470e+00 -6.79911137e-01 1.57762602e-01
-1.31789804e-01 1.81343973e-01 4.38663960e-01 9.08733785e-01
-5.88711537e-02 1.93810359e-01 2.72048354e-01 -5.09514689e-01
-1.09330595e+00 8.26142967e-01 3.63334864e-01 -2.67495483e-01
-3.96447480e-01 5.79507053e-01 2.03388661e-01 -6.01297319e-01
6.83494985e-01 3.21820319e-01 2.83688605e-01 7.19684968e-03
4.23896909e-01 5.84244609e-01 2.79783368e-01 -2.00201213e-01
-2.34512433e-01 -1.15770720e-01 -3.75929922e-01 2.08419740e-01
1.42687511e+00 -2.97400475e-01 -5.08758128e-01 2.53156781e-01
1.49984705e+00 1.47323072e-01 -1.36745119e+00 -6.83552206e-01
-2.13117093e-01 -7.79431581e-01 3.14498961e-01 -7.90760279e-01
-1.21541667e+00 6.82030559e-01 6.91654563e-01 2.82064453e-02
9.87997532e-01 -1.68093741e-01 5.64024627e-01 4.23587352e-01
8.62934709e-01 -1.14225817e+00 -4.10408854e-01 3.39445770e-01
4.86678928e-01 -1.12684011e+00 -9.69975665e-02 -1.78554934e-02
-5.51095188e-01 9.42593813e-01 6.35271192e-01 1.62440538e-01
1.06298959e+00 1.23952031e-01 2.59723375e-03 2.30068862e-01
-1.04979980e+00 4.82254475e-01 -1.48521051e-01 4.50788289e-01
-1.87426940e-01 1.67500228e-04 -1.64968669e-01 1.02542067e+00
3.53025198e-01 1.41633719e-01 4.44193631e-02 8.47502828e-01
-2.30648249e-01 -1.34335256e+00 -5.92928529e-01 3.91194969e-01
-7.19456971e-01 3.47289383e-01 -8.96118283e-02 2.21932247e-01
2.00891078e-01 9.15714145e-01 -1.22686796e-01 -7.56357983e-02
-1.09048679e-01 3.67582202e-01 -9.72276405e-02 -2.85077542e-01
-6.33226097e-01 1.58965021e-01 -3.42869103e-01 -6.31409943e-01
1.44180432e-01 -8.30414832e-01 -1.00819886e+00 -6.08491421e-01
-3.43895823e-01 5.44752240e-01 1.01142454e+00 8.28862369e-01
7.70138323e-01 -2.66769707e-01 1.34815049e+00 -1.10256389e-01
-1.67714262e+00 -6.97479665e-01 -9.19602156e-01 5.59950233e-01
1.19015545e-01 -2.51687944e-01 -8.41327190e-01 -2.77779877e-01] | [5.837806224822998, 6.3294901847839355] |
ec0dfbf2-f602-41df-b011-e66b6efa35f2 | sentence-constituent-aware-aspect-category | 2010.01461 | null | https://arxiv.org/abs/2010.01461v1 | https://arxiv.org/pdf/2010.01461v1.pdf | Sentence Constituent-Aware Aspect-Category Sentiment Analysis with Graph Attention Networks | Aspect category sentiment analysis (ACSA) aims to predict the sentiment polarities of the aspect categories discussed in sentences. Since a sentence usually discusses one or more aspect categories and expresses different sentiments toward them, various attention-based methods have been developed to allocate the appropriate sentiment words for the given aspect category and obtain promising results. However, most of these methods directly use the given aspect category to find the aspect category-related sentiment words, which may cause mismatching between the sentiment words and the aspect categories when an unrelated sentiment word is semantically meaningful for the given aspect category. To mitigate this problem, we propose a Sentence Constituent-Aware Network (SCAN) for aspect-category sentiment analysis. SCAN contains two graph attention modules and an interactive loss function. The graph attention modules generate representations of the nodes in sentence constituency parse trees for the aspect category detection (ACD) task and the ACSA task, respectively. ACD aims to detect aspect categories discussed in sentences and is a auxiliary task. For a given aspect category, the interactive loss function helps the ACD task to find the nodes which can predict the aspect category but can't predict other aspect categories. The sentiment words in the nodes then are used to predict the sentiment polarity of the aspect category by the ACSA task. The experimental results on five public datasets demonstrate the effectiveness of SCAN. | ['Sheng-hua Zhong', 'Cunxiang Yin', 'Yuncong Li'] | 2020-10-04 | null | null | null | null | ['aspect-category-detection'] | ['natural-language-processing'] | [ 1.42451972e-01 3.82641047e-01 -2.86921173e-01 -7.31818140e-01
-4.94717747e-01 -5.16366303e-01 3.98882866e-01 3.65930945e-01
-3.97155248e-02 1.31560102e-01 5.28923929e-01 -4.90975708e-01
2.34608352e-01 -1.16608572e+00 -3.06938678e-01 -5.59089124e-01
4.24150288e-01 4.02975500e-01 2.90944517e-01 -6.20467842e-01
4.26237047e-01 -2.86716204e-02 -1.18317163e+00 7.31566608e-01
7.41158009e-01 1.18742478e+00 2.82207817e-01 3.94605249e-01
-8.76385450e-01 7.63774574e-01 -9.03877258e-01 -5.75035870e-01
-1.66041940e-01 -6.84073210e-01 -9.01363909e-01 2.40375876e-01
1.24117225e-01 4.36348230e-01 2.00707659e-01 1.22283292e+00
3.71390164e-01 4.23540287e-02 7.92037547e-01 -1.25383413e+00
-5.67374170e-01 9.53128278e-01 -6.63225830e-01 3.85860771e-01
2.13666916e-01 -1.99964345e-01 1.72689068e+00 -8.06220770e-01
3.17393899e-01 1.61468732e+00 5.75564682e-01 4.96878207e-01
-6.22962296e-01 -5.33270121e-01 8.95770192e-01 1.42270640e-01
-8.64677012e-01 -4.31952009e-04 1.20744681e+00 -1.39313892e-01
1.14412093e+00 2.96040595e-01 8.27820241e-01 6.29044950e-01
4.53810960e-01 1.00578725e+00 7.60509193e-01 -6.12027198e-02
1.88324362e-01 3.23832005e-01 7.50633776e-01 4.52775568e-01
1.44032747e-01 -7.29369879e-01 -2.19775826e-01 2.38525122e-03
-1.30570501e-01 -7.94943497e-02 -7.94263184e-02 1.70415025e-02
-8.01405132e-01 1.14444482e+00 7.18810380e-01 2.99867094e-01
-3.48652571e-01 -1.26714125e-01 7.90417016e-01 3.19947988e-01
8.12668800e-01 5.28406560e-01 -7.97820687e-01 3.25273573e-01
-1.54851794e-01 1.23898312e-01 9.11264181e-01 9.69869792e-01
8.96863282e-01 -1.10547096e-01 -4.96495485e-01 8.36534917e-01
6.12914622e-01 4.34873998e-01 5.83600640e-01 -3.08173411e-02
8.80563080e-01 1.43658793e+00 -6.32552683e-01 -1.59887111e+00
-6.07746899e-01 -5.08762836e-01 -6.97320044e-01 -3.24172020e-01
-9.12446678e-02 -1.71813622e-01 -8.91697288e-01 1.49201250e+00
4.94855255e-01 -4.35565054e-01 2.18239605e-01 9.59838629e-01
1.59917545e+00 8.22410464e-01 2.56903946e-01 -7.83803910e-02
1.86914515e+00 -1.13281393e+00 -6.93122089e-01 -8.60695004e-01
6.75282121e-01 -9.56116855e-01 1.37702465e+00 -1.49394423e-01
-6.97221160e-01 -4.30173695e-01 -9.42294180e-01 -6.82275668e-02
-5.71494341e-01 -5.47207892e-02 8.15458536e-01 3.36833775e-01
-8.97188663e-01 8.04145411e-02 -3.00683051e-01 -2.84481376e-01
5.08020699e-01 3.16795439e-01 1.30054206e-01 1.30262062e-01
-1.47805929e+00 5.30390978e-01 1.05945870e-01 6.30638599e-02
-4.67086613e-01 -3.69731545e-01 -1.26765943e+00 2.55978018e-01
2.56422311e-01 -6.04629517e-01 1.02290976e+00 -1.58719695e+00
-9.29421961e-01 8.98446977e-01 -4.07989740e-01 -1.45458700e-02
-2.95966446e-01 2.38942385e-01 -5.97327888e-01 8.06798339e-02
6.57379389e-01 5.50997257e-01 8.29763710e-01 -1.00703251e+00
-7.78926253e-01 -4.73210633e-01 4.29529220e-01 7.34679520e-01
-4.75060731e-01 -2.72396835e-03 -5.27965069e-01 -7.35569239e-01
3.38508129e-01 -6.75460517e-01 -2.93320060e-01 -5.27979314e-01
-9.35627103e-01 -7.71436691e-01 8.97061229e-01 -2.58485705e-01
1.14612377e+00 -2.05408716e+00 1.52372280e-02 1.23908013e-01
2.18016490e-01 -1.52309909e-01 -3.22551876e-01 4.00060602e-02
-2.44566292e-01 2.09540963e-01 -2.51405180e-01 -1.07855015e-01
-2.11549342e-01 1.24363609e-01 -4.85465199e-01 3.54016759e-02
4.54047084e-01 8.16072702e-01 -8.42344463e-01 -5.28293967e-01
-2.00938344e-01 2.15925947e-01 -5.81907928e-01 3.01861584e-01
-4.41451371e-01 1.68038294e-01 -8.15654993e-01 6.71612084e-01
5.85911572e-01 -2.81618267e-01 9.83236954e-02 -5.65541983e-01
1.89548627e-01 9.79633868e-01 -6.96734309e-01 8.72392595e-01
-5.24557292e-01 5.05176842e-01 -1.15594164e-01 -1.23528492e+00
1.13595057e+00 2.97023803e-02 3.92069928e-02 -6.07350886e-01
4.01690930e-01 5.40051013e-02 3.52704436e-01 -4.26399201e-01
6.95246518e-01 -5.22558093e-01 -5.57397068e-01 4.63746428e-01
-1.44870043e-01 -4.39004570e-01 1.91343099e-01 4.04489547e-01
8.68443012e-01 -4.19150591e-01 2.33964443e-01 -5.04352868e-01
9.75617111e-01 2.38055915e-01 7.62834907e-01 4.01519299e-01
-1.69579506e-01 5.47692358e-01 1.18236697e+00 -6.54164016e-01
-5.73264241e-01 -5.09093046e-01 1.10825539e-01 1.26101243e+00
6.09104753e-01 -4.62556720e-01 -6.12440050e-01 -1.30749846e+00
-3.06173444e-01 8.71117532e-01 -8.29622388e-01 -3.18335056e-01
-3.74887258e-01 -8.62970233e-01 -1.12373814e-01 5.28441370e-01
6.13830686e-01 -1.52219200e+00 2.78620366e-02 -8.02354813e-02
-3.54237437e-01 -8.49866688e-01 -7.26341486e-01 2.92499244e-01
-5.54513395e-01 -1.27994096e+00 -3.38447213e-01 -1.21217132e+00
1.06826162e+00 4.98731762e-01 1.38654387e+00 1.77347660e-01
3.40153217e-01 1.71091065e-01 -5.63005090e-01 -6.21920645e-01
-3.33354145e-01 3.42123955e-01 -2.68619925e-01 1.60285637e-01
9.07816887e-01 -1.89465716e-01 -6.58366203e-01 3.45676094e-01
-7.21697271e-01 -1.34494985e-02 4.46725518e-01 4.76648986e-01
7.13917077e-01 3.17243576e-01 6.92792833e-01 -1.47624373e+00
9.42970216e-01 -7.16230571e-01 -2.60544777e-01 -6.30726293e-02
-7.49845147e-01 -2.23681644e-01 9.74811494e-01 -2.16657802e-01
-8.35746825e-01 -3.27169001e-01 -4.42119092e-01 1.60583720e-01
8.77946913e-02 8.30517888e-01 -7.04440773e-01 5.98721206e-01
1.76234141e-01 2.09647849e-01 -2.79255152e-01 9.60705802e-02
-2.31683115e-03 7.51948535e-01 -1.87539026e-01 4.73949779e-03
4.22276616e-01 3.03151608e-01 -2.17378229e-01 -5.02408743e-01
-1.61561918e+00 -6.89896524e-01 -2.29398221e-01 -2.57814288e-01
1.04253113e+00 -8.64104569e-01 -3.33561450e-01 4.17407393e-01
-1.27772033e+00 1.13702118e-01 -2.69017607e-01 -3.43601443e-02
-1.25036508e-01 3.31661403e-01 -2.69687295e-01 -5.11837006e-01
-7.90425777e-01 -1.43286073e+00 1.32282615e+00 5.24834454e-01
-4.42636400e-01 -1.17017269e+00 -1.64504260e-01 3.88311535e-01
2.13023052e-01 -2.20682129e-01 1.39601493e+00 -1.10111046e+00
-1.17490977e-01 -3.03219289e-01 -2.42858723e-01 2.84041226e-01
3.73062909e-01 -2.93913007e-01 -8.76667976e-01 -1.08935609e-01
1.34280309e-01 -1.08555406e-02 9.25027668e-01 4.64295626e-01
1.03676999e+00 -5.77304780e-01 -2.81227320e-01 2.89197892e-01
1.06671619e+00 3.07089955e-01 5.13341784e-01 4.07985598e-01
9.94527876e-01 7.28676140e-01 7.82381117e-01 -1.12502545e-01
6.64404988e-01 3.74053478e-01 6.29515052e-01 -1.92939639e-01
-2.10555978e-02 -1.91200301e-01 5.71172476e-01 1.17570245e+00
6.39722884e-01 -5.18798709e-01 -6.53537869e-01 9.04379785e-01
-1.55621839e+00 -5.59816122e-01 -5.53763628e-01 1.55719328e+00
7.10440576e-01 4.65440989e-01 -6.30866736e-02 1.77433044e-01
9.21651781e-01 6.88237906e-01 -7.48457134e-01 -9.06238973e-01
-2.95803189e-01 -2.39587054e-01 -8.03864375e-02 4.02080059e-01
-1.26722527e+00 1.14094210e+00 4.91085911e+00 7.89941609e-01
-1.00642467e+00 -1.22279199e-02 1.07089412e+00 3.52873713e-01
-9.95685399e-01 2.12626964e-01 -9.77090776e-01 4.19190735e-01
3.82957608e-01 -1.50512323e-01 -2.46951759e-01 1.22994447e+00
1.58562973e-01 2.67597735e-02 -6.89022362e-01 3.79403085e-01
3.04601282e-01 -1.02175713e+00 5.31201780e-01 -3.31131876e-01
5.19960046e-01 -1.32896975e-01 2.40384247e-02 6.10514998e-01
-7.54921436e-02 -6.61549211e-01 4.49327826e-01 3.68718766e-02
3.75026643e-01 -1.00972366e+00 1.16424096e+00 5.02835736e-02
-1.44480169e+00 7.49934539e-02 -8.21170032e-01 -1.78634897e-01
-6.15205290e-03 8.51579964e-01 -7.23496318e-01 8.52608904e-02
8.15649569e-01 1.01037705e+00 -6.14573359e-01 4.16115731e-01
-6.54161513e-01 7.73543417e-01 1.95674926e-01 -6.53683424e-01
5.69242358e-01 -3.67823094e-01 8.69374514e-01 1.04627657e+00
-7.12119341e-02 -5.30218072e-02 6.54389635e-02 7.18154788e-01
-2.73386985e-01 6.21670842e-01 -5.74450195e-01 -1.01522297e-01
4.08951566e-03 1.66934371e+00 -1.05635285e+00 -3.53237361e-01
-5.07855415e-01 7.31740475e-01 2.89317667e-01 9.48912725e-02
-4.34873074e-01 -7.42820740e-01 7.92674959e-01 5.90509921e-02
4.18250740e-01 5.98450243e-01 -7.54080474e-01 -1.00672984e+00
6.06396385e-02 -7.58139729e-01 6.88954055e-01 -8.73810947e-01
-1.50875604e+00 1.04533505e+00 -5.10978997e-01 -1.29428673e+00
1.55627847e-01 -5.28229237e-01 -1.33582222e+00 8.93313825e-01
-1.76195180e+00 -1.10259807e+00 -1.33312508e-01 3.81476372e-01
1.03124237e+00 -4.03932869e-01 5.63265741e-01 -1.49495676e-01
-5.16037285e-01 5.19797802e-01 -5.92587769e-01 2.79874861e-01
1.79922298e-01 -1.45857239e+00 5.41566610e-01 7.52186358e-01
-2.76251197e-01 5.13916910e-01 7.02607751e-01 -8.33956361e-01
-1.03283310e+00 -1.43393457e+00 1.49228764e+00 -2.13549927e-01
6.58728004e-01 -3.00801247e-01 -8.89052391e-01 4.65107858e-01
3.27496409e-01 -3.76397550e-01 8.36399853e-01 3.45189005e-01
-2.94211894e-01 -1.25870302e-01 -8.81832302e-01 6.99887216e-01
6.62944853e-01 -3.66153479e-01 -8.47234726e-01 5.24242342e-01
1.21651709e+00 -4.43531759e-02 -3.19671899e-01 4.20040846e-01
-8.39642063e-02 -7.74095953e-01 6.51732683e-01 -7.96366990e-01
9.69920576e-01 -4.68722790e-01 -6.83526918e-02 -1.57593000e+00
-2.83017397e-01 3.08250971e-02 3.20162892e-01 1.66971564e+00
9.32971716e-01 -5.71458697e-01 6.54570520e-01 4.84980166e-01
-3.53922307e-01 -9.62087274e-01 -5.82963109e-01 -1.16162024e-01
5.16330600e-02 -5.13405442e-01 8.26426625e-01 8.41620386e-01
1.62041813e-01 1.29475403e+00 2.12320313e-01 1.69051930e-01
9.71084535e-02 8.65632534e-01 4.47054297e-01 -9.95223045e-01
-5.95619939e-02 -6.96865976e-01 -2.70285547e-01 -9.52437818e-01
5.26575506e-01 -1.21049035e+00 1.75038636e-01 -1.91657102e+00
3.92125487e-01 -4.86606568e-01 -1.71976328e-01 5.56647599e-01
-7.47720838e-01 9.17711407e-02 1.28085678e-03 6.43142462e-02
-8.31137598e-01 7.08955407e-01 1.50587738e+00 -6.84686720e-01
-1.71840668e-01 4.75397378e-01 -1.57505310e+00 1.02747953e+00
8.67096066e-01 -5.28746367e-01 -4.94225204e-01 -2.04448551e-01
8.14023852e-01 -2.77206600e-01 -2.27662876e-01 -3.99544060e-01
2.80120708e-02 -5.34075089e-02 2.02725857e-01 -1.03734136e+00
-8.32178220e-02 -7.50696063e-01 -7.07885265e-01 4.29724663e-01
-5.40179193e-01 2.06620827e-01 1.45034730e-01 5.59628665e-01
-5.23140788e-01 -2.89129108e-01 6.07965708e-01 -1.30847409e-01
-7.97533154e-01 2.64458358e-01 -5.27056873e-01 3.65626782e-01
7.15641499e-01 -9.31926146e-02 -4.46472347e-01 -6.56105518e-01
-4.11423802e-01 5.46420038e-01 1.12775080e-01 8.11918139e-01
7.85757124e-01 -1.21201289e+00 -5.72785258e-01 1.89102903e-01
4.35997993e-01 4.29918587e-01 2.30219230e-01 6.02195382e-01
-1.76709756e-01 2.04367206e-01 3.45142037e-01 -4.80353653e-01
-1.22626591e+00 5.62270820e-01 4.16473001e-01 -4.89963591e-01
-2.00261280e-01 9.77550328e-01 7.85747588e-01 -8.90390337e-01
-2.05181450e-01 -1.92777425e-01 -1.18308878e+00 3.41669112e-01
5.00363290e-01 -1.77217886e-01 3.40326247e-03 -1.08756423e+00
-5.45106411e-01 7.15917766e-01 -2.02333942e-01 4.08401877e-01
1.22913706e+00 -2.91318536e-01 -5.87505937e-01 4.05829012e-01
1.42118180e+00 6.29803091e-02 -6.95994616e-01 -1.54663632e-02
-1.06051177e-01 3.98665257e-02 6.75128177e-02 -7.36639023e-01
-1.55828452e+00 9.35265124e-01 -1.82908121e-02 5.85079253e-01
1.33139205e+00 4.10886914e-01 8.00819516e-01 4.17625979e-02
-1.20769501e-01 -9.73315239e-01 4.85339984e-02 9.33936179e-01
9.73056555e-01 -1.19988084e+00 -4.50474359e-02 -8.17376077e-01
-1.16347420e+00 9.77752149e-01 1.01479375e+00 -1.34326652e-01
8.27165902e-01 -2.33364142e-02 4.20011938e-01 -6.82059050e-01
-6.19711339e-01 -2.49577820e-01 5.46295702e-01 4.65963483e-01
4.98461008e-01 3.64894792e-02 -5.86708903e-01 1.12747633e+00
-4.48469877e-01 -1.01006198e+00 4.21103895e-01 4.76633608e-01
-5.24571240e-01 -7.87144303e-01 -4.41817306e-02 8.18707705e-01
-6.32113636e-01 -5.07871985e-01 -8.67025971e-01 3.75726283e-01
-1.11473814e-01 1.28049481e+00 3.28369647e-01 -5.45754254e-01
5.55447221e-01 -1.45724207e-01 -4.59138304e-01 -9.16885555e-01
-1.10870409e+00 1.05294019e-01 2.68017262e-01 -3.12565178e-01
-4.16164100e-01 -5.07625878e-01 -1.55198026e+00 7.15575144e-02
-4.28655386e-01 4.74656731e-01 4.46093112e-01 1.13158500e+00
3.70446026e-01 8.44121516e-01 1.09083319e+00 -1.91092238e-01
1.74808711e-01 -9.86291766e-01 -5.68437278e-01 6.13618255e-01
6.23752847e-02 -3.02108258e-01 -4.88543630e-01 -3.59515280e-01] | [11.483041763305664, 6.626908302307129] |
37d36f81-99ba-43b9-b61a-68c8f80c925b | ls-net-fast-single-shot-line-segment-detector | 1912.09532 | null | https://arxiv.org/abs/1912.09532v2 | https://arxiv.org/pdf/1912.09532v2.pdf | LS-Net: Fast Single-Shot Line-Segment Detector | In low-altitude Unmanned Aerial Vehicle (UAV) flights, power lines are considered as one of the most threatening hazards and one of the most difficult obstacles to avoid. In recent years, many vision-based techniques have been proposed to detect power lines to facilitate self-driving UAVs and automatic obstacle avoidance. However, most of the proposed methods are typically based on a common three-step approach: (i) edge detection, (ii) the Hough transform, and (iii) spurious line elimination based on power line constrains. These approaches not only are slow and inaccurate but also require a huge amount of effort in post-processing to distinguish between power lines and spurious lines. In this paper, we introduce LS-Net, a fast single-shot line-segment detector, and apply it to power line detection. The LS-Net is by design fully convolutional and consists of three modules: (i) a fully convolutional feature extractor, (ii) a classifier, and (iii) a line segment regressor. Due to the unavailability of large datasets with annotations of power lines, we render synthetic images of power lines using the Physically Based Rendering (PBR) approach and propose a series of effective data augmentation techniques to generate more training data. With a customized version of the VGG-16 network as the backbone, the proposed approach outperforms existing state-of-the-art approaches. In addition, the LS-Net can detect power lines in near real-time (20.4 FPS). This suggests that our proposed approach has a promising role in automatic obstacle avoidance and as a valuable component of self-driving UAVs, especially for automatic autonomous power line inspection. | ['Davide Roverso', 'Van Nhan Nguyen', 'Robert Jenssen'] | 2019-12-19 | null | null | null | null | ['line-detection'] | ['computer-vision'] | [ 1.05713224e-02 -3.87758374e-01 2.59472191e-01 1.22334838e-01
-1.60795912e-01 -8.28168452e-01 3.30548167e-01 7.92032480e-02
-2.97597855e-01 5.92648029e-01 -7.18492150e-01 -6.80885911e-01
-1.33136109e-01 -1.03533781e+00 -4.05608475e-01 -4.94911641e-01
-2.02295497e-01 5.22487098e-03 6.22659743e-01 -6.17305458e-01
9.68736410e-02 1.25448585e+00 -1.44443500e+00 -3.41854304e-01
1.21296799e+00 1.08998144e+00 -5.99070303e-02 7.15294421e-01
1.36427075e-01 4.95059878e-01 -6.50042295e-01 1.37326390e-01
5.76375484e-01 -4.28322464e-01 -3.11028272e-01 2.69420624e-01
1.60452768e-01 -6.06609285e-01 -4.64177728e-01 9.68120813e-01
5.30563295e-01 1.30699828e-01 4.96431559e-01 -1.63255382e+00
5.53888939e-02 2.79913731e-02 -1.08811772e+00 2.63350427e-01
3.79455745e-01 3.76295298e-01 4.66530234e-01 -5.30170977e-01
2.03195252e-02 5.88159502e-01 1.06073070e+00 5.56449406e-02
-7.98931181e-01 -6.15959346e-01 -2.41824642e-01 1.16975874e-01
-1.50800145e+00 -1.82444200e-01 1.02778542e+00 -5.20696819e-01
8.01485837e-01 3.16784471e-01 1.05656445e+00 4.15058136e-01
2.26978481e-01 3.27784032e-01 7.68373191e-01 -4.84690905e-01
1.86955437e-01 1.19458705e-01 -2.00954974e-01 1.18684852e+00
3.50366384e-01 1.15887135e-01 -4.57828343e-02 -1.64566725e-01
1.00544786e+00 -2.75558740e-01 -6.69191480e-01 -3.99796426e-01
-9.13710117e-01 5.80009520e-01 6.67289734e-01 1.71499103e-01
-1.73670724e-01 -1.03985995e-01 3.03405195e-01 -4.83108684e-03
1.28833830e-01 2.89613485e-01 -2.41480291e-01 5.11752740e-02
-1.16924620e+00 2.27328151e-01 7.37885892e-01 8.90105665e-01
6.21972084e-01 5.11225164e-01 4.59990986e-02 5.67837715e-01
2.71667033e-01 4.02889878e-01 2.81292945e-01 -2.78729051e-01
4.66043085e-01 6.58541977e-01 2.48782143e-01 -1.26632726e+00
-8.43829691e-01 -6.51276588e-01 -8.74694824e-01 7.51489878e-01
1.61668062e-01 -7.87304878e-01 -5.53390920e-01 1.08469033e+00
3.72028351e-01 1.99082315e-01 -5.75157553e-02 1.04413319e+00
6.16925716e-01 8.49315047e-01 -2.59325594e-01 -1.07010134e-01
1.26736116e+00 -9.60066259e-01 -5.97223222e-01 -1.55493587e-01
8.39548588e-01 -6.64336979e-01 8.39580476e-01 4.07288671e-01
-7.00425267e-01 -6.93990529e-01 -1.61223376e+00 2.46971369e-01
-4.01999801e-01 8.04926038e-01 6.31111503e-01 5.81644595e-01
-7.79045284e-01 6.73326075e-01 -5.82565784e-01 -4.37648743e-01
4.20175254e-01 1.76516861e-01 -3.79351854e-01 4.81796443e-01
-9.59113896e-01 5.13284862e-01 4.25127596e-01 4.65962261e-01
-6.74143255e-01 -6.13367140e-01 -8.04847479e-01 1.80607647e-01
6.02522254e-01 -5.59699416e-01 6.26917481e-01 -7.61990130e-01
-1.42150354e+00 3.98116171e-01 4.65645105e-01 -6.30352080e-01
7.26332605e-01 -2.86329746e-01 -4.60093647e-01 3.83143544e-01
-1.28286192e-02 3.99020135e-01 8.77656817e-01 -1.11499286e+00
-1.11805725e+00 -4.41150879e-03 9.24472734e-02 2.50411183e-01
-3.22452933e-01 -3.17970246e-01 -1.43449828e-01 -5.54084182e-01
-1.31180093e-01 -8.30269217e-01 -2.55128331e-02 3.65054727e-01
-6.93804324e-01 -9.67394784e-02 1.27016044e+00 -6.58612251e-01
1.17320943e+00 -1.98356748e+00 -3.80254626e-01 1.15176082e-01
2.96926708e-03 7.98027575e-01 9.63179320e-02 4.77795541e-01
-1.57406092e-01 -7.01851398e-02 -3.59833509e-01 2.50716180e-01
-3.65712434e-01 -4.37588185e-01 -2.99629271e-01 6.47223115e-01
2.95632750e-01 4.02771801e-01 -6.33867502e-01 -3.03481698e-01
6.81441486e-01 3.67368728e-01 -1.22747488e-01 1.53045371e-01
1.24963097e-01 4.57343519e-01 -3.67270947e-01 5.38802683e-01
9.61627722e-01 4.20498729e-01 -1.93079710e-01 -6.61788046e-01
-7.55848646e-01 -5.13247848e-01 -1.16329229e+00 1.46164417e+00
-1.35242313e-01 7.63365209e-01 1.09623298e-02 -7.76347339e-01
1.30788124e+00 2.31404454e-01 6.11598194e-01 -2.53058583e-01
3.24100792e-01 6.34211972e-02 7.41452947e-02 -5.28267086e-01
4.00871187e-01 1.84013367e-01 6.30254671e-02 -1.57305181e-01
-3.05472855e-02 -4.72355992e-01 4.14144784e-01 1.40673807e-02
1.06611526e+00 4.39045936e-01 2.65154958e-01 -3.53196204e-01
9.19709504e-01 3.65576386e-01 7.85333872e-01 3.38554293e-01
-1.65575355e-01 3.63180995e-01 5.04147887e-01 -5.78873932e-01
-9.68021631e-01 -7.37406671e-01 -4.18963283e-02 6.41359612e-02
4.27237004e-01 -3.27071398e-01 -9.20174956e-01 -6.74906433e-01
-7.83665404e-02 6.83273911e-01 2.91891093e-03 -1.80407047e-01
-2.88331509e-01 -6.38839483e-01 8.04236472e-01 5.78925848e-01
1.15524149e+00 -6.41454279e-01 -1.04701102e+00 2.40870371e-01
2.22740591e-01 -1.41138744e+00 -1.31020054e-01 1.40267774e-01
-6.53285384e-01 -1.45474386e+00 -6.25406146e-01 -7.32752442e-01
7.44440734e-01 5.87330818e-01 5.28294146e-01 3.91410559e-01
-6.30728006e-01 2.37742186e-01 -3.86196792e-01 -5.56771755e-01
8.66661500e-03 -7.24679753e-02 -3.42930257e-02 6.40538260e-02
-7.41778389e-02 -3.37193638e-01 -6.37205005e-01 4.00874168e-01
-6.11278594e-01 2.46442512e-01 4.90320235e-01 6.08841419e-01
2.43542001e-01 5.95638871e-01 2.55203992e-01 -5.46492994e-01
6.72702253e-01 3.37107922e-03 -1.41942120e+00 1.33249044e-01
-2.38150328e-01 -6.80090368e-01 1.29661071e+00 -3.59563753e-02
-8.70986998e-01 3.39705974e-01 -2.38419548e-01 -5.39459705e-01
-3.77660364e-01 3.07422668e-01 -2.36346424e-02 -6.56196654e-01
4.70990241e-01 1.51086196e-01 -1.68475494e-01 -8.10826346e-02
1.14777256e-02 5.14944553e-01 5.09803832e-01 -2.51703542e-02
1.49986410e+00 5.26151240e-01 6.15681529e-01 -1.54492080e+00
-5.21018982e-01 -2.18547374e-01 -9.65805590e-01 -5.29964685e-01
7.75494516e-01 -7.76799560e-01 -6.89424038e-01 7.02047825e-01
-1.09826481e+00 -3.96095306e-01 -2.47338384e-01 3.37589175e-01
-2.70115435e-01 7.62847126e-01 -3.37809324e-01 -9.02994215e-01
-4.78181362e-01 -1.00278425e+00 6.97994590e-01 8.28388810e-01
2.86983520e-01 -8.94105434e-01 -2.05189377e-01 1.44069299e-01
2.29668617e-01 6.95651948e-01 7.39196837e-01 -2.14631617e-01
-3.16257596e-01 -5.88859558e-01 -3.07904959e-01 3.92612368e-01
2.40596816e-01 5.42078555e-01 -7.35830963e-01 -5.64307094e-01
-3.13622534e-01 -1.55021593e-01 4.09043342e-01 1.97155103e-01
9.25819755e-01 2.21911930e-02 -4.43300426e-01 9.21329916e-01
1.60348904e+00 6.01922870e-01 5.59021711e-01 3.67116570e-01
7.19184399e-01 3.04547608e-01 8.63796175e-01 7.13976562e-01
6.66676462e-02 5.81621051e-01 6.28323555e-01 -8.03070486e-01
1.37572736e-01 -1.74942821e-01 9.18849856e-02 4.14484859e-01
-1.73042446e-01 -3.24754596e-01 -8.28097522e-01 2.52115428e-01
-1.64441240e+00 -7.76556849e-01 -5.87090433e-01 2.29155564e+00
1.70448422e-01 2.92474896e-01 2.12214366e-01 5.21271348e-01
5.06356895e-01 5.00038601e-02 -4.01419789e-01 -2.94880360e-01
-3.12679224e-02 -1.44207597e-01 7.65938520e-01 2.13631645e-01
-1.43134308e+00 8.28682661e-01 4.82740307e+00 7.23951399e-01
-1.25270545e+00 -4.39642161e-01 2.42268756e-01 6.46031141e-01
2.77636737e-01 1.66501656e-01 -6.67457640e-01 3.00726682e-01
1.60880789e-01 -1.31123206e-02 2.62662530e-01 7.54573286e-01
3.39766800e-01 -4.42101508e-01 -5.15540540e-01 9.76426065e-01
6.90878704e-02 -9.00877953e-01 -1.11918293e-01 -2.51314729e-01
4.22324121e-01 -3.91025811e-01 -4.76755172e-01 -8.40118974e-02
-3.17984432e-01 -5.67993224e-01 5.46795666e-01 4.39471006e-01
8.12132895e-01 -9.39665675e-01 8.29833746e-01 4.86017853e-01
-1.63071454e+00 -1.54371589e-01 -5.09124219e-01 1.46280620e-02
2.21111715e-01 5.53345144e-01 -6.51908278e-01 1.01743066e+00
6.80674911e-01 5.98155200e-01 -4.73647892e-01 1.47609091e+00
-5.06985545e-01 4.84915704e-01 -4.81989920e-01 1.03698246e-01
4.01043862e-01 -5.16976118e-01 7.21959949e-01 9.02101576e-01
5.09908795e-01 -5.68802580e-02 3.70472133e-01 7.72997677e-01
3.82201254e-01 7.35334530e-02 -8.65053415e-01 9.27305669e-02
2.95516312e-01 1.87123191e+00 -1.03915143e+00 -3.78442705e-02
-6.44916236e-01 9.44788933e-01 -3.61300632e-02 2.36450866e-01
-1.13508284e+00 -1.26195276e+00 3.64807963e-01 2.79939771e-01
1.09283991e-01 -5.58746517e-01 1.11535430e-01 -8.83789062e-01
-2.20498100e-01 -4.18369085e-01 9.58846435e-02 -1.04607856e+00
-7.22008467e-01 8.75402391e-01 -1.66022524e-01 -1.73856902e+00
3.37387249e-02 -6.84348226e-01 -8.95408094e-01 9.19409215e-01
-1.80922616e+00 -1.18951094e+00 -1.05534470e+00 4.66755718e-01
7.06275702e-01 -3.29932094e-01 6.83356464e-01 1.60035953e-01
-9.16686833e-01 1.89136788e-01 -2.31357679e-01 4.40081000e-01
3.22598398e-01 -1.03176272e+00 2.76159436e-01 1.18971705e+00
-9.51668471e-02 2.10750271e-02 6.69019520e-01 -7.82517195e-01
-1.42205417e+00 -1.10337543e+00 -1.05346376e-02 3.91226113e-01
6.01252675e-01 -1.50556132e-01 -7.83270359e-01 4.32162613e-01
2.71420807e-01 1.54944226e-01 3.50109369e-01 -6.70926630e-01
5.44086218e-01 -4.37142432e-01 -1.21435881e+00 6.88697577e-01
5.80116689e-01 6.66118562e-02 -1.35585934e-01 3.68462533e-01
1.30535215e-01 -3.56897056e-01 -6.36052847e-01 5.75182736e-01
3.91104460e-01 -1.31376374e+00 6.79601490e-01 -6.82884315e-03
-4.92496714e-02 -8.59513760e-01 6.37903869e-01 -1.39732289e+00
-4.73000407e-01 -6.58784568e-01 3.37387472e-01 1.33653867e+00
9.53658670e-02 -7.56108403e-01 7.22313106e-01 6.73759729e-02
-4.15600926e-01 -7.64375567e-01 -6.70640171e-01 -6.57632530e-01
-5.08328080e-01 -6.12918325e-02 4.55202311e-01 8.26653123e-01
-2.18883321e-01 2.80173928e-01 -2.15041712e-01 7.13209569e-01
6.98122144e-01 -6.68715909e-02 8.82727623e-01 -1.27828062e+00
6.10234402e-02 -1.38672218e-01 -5.72694898e-01 -8.78998876e-01
-9.97415781e-02 -4.30893034e-01 -7.30729476e-02 -1.72158384e+00
-5.54124892e-01 -4.32734370e-01 2.56135315e-01 4.85660374e-01
1.51666239e-01 1.96049929e-01 6.54645264e-02 6.53117076e-02
-6.86549842e-02 4.80313510e-01 9.99560356e-01 -1.48212137e-02
-2.58799493e-01 1.24048613e-01 -9.42516625e-02 1.20052457e+00
1.08691871e+00 -1.40851320e-04 -5.11051238e-01 -2.53156215e-01
4.81120199e-02 2.28358552e-01 5.01244128e-01 -1.77789807e+00
4.84778076e-01 8.94387290e-02 7.35687256e-01 -7.88521528e-01
2.50153452e-01 -1.23659003e+00 -1.82941481e-01 7.40108609e-01
4.45603788e-01 1.88314617e-01 6.71822429e-01 2.38493159e-01
-1.14012904e-01 -5.31056702e-01 8.51426780e-01 7.77822882e-02
-9.62715685e-01 2.51758456e-01 -6.32262588e-01 -3.74772668e-01
1.55166841e+00 -4.75983292e-01 -3.99679303e-01 -1.98321402e-01
-1.09179839e-01 4.18211639e-01 4.45042670e-01 1.64571956e-01
7.99502015e-01 -1.02341247e+00 -5.10212660e-01 6.17876351e-01
-1.86182469e-01 8.77507702e-02 4.22589988e-01 8.66023600e-01
-1.35192108e+00 3.98028225e-01 -4.52960491e-01 -4.81500149e-01
-1.16035128e+00 3.97664756e-01 5.98427534e-01 1.59987826e-02
-8.34323823e-01 4.33409095e-01 -6.79180473e-02 5.06326370e-02
-4.94627059e-02 -2.19942808e-01 -5.78279078e-01 6.72714319e-03
2.73602188e-01 5.89547038e-01 9.58948582e-02 -6.44305050e-01
-2.86236912e-01 1.13537169e+00 3.97269994e-01 2.50002474e-01
1.05669868e+00 7.19462857e-02 1.31342977e-01 1.50380880e-01
5.04536808e-01 8.33066478e-02 -1.37433016e+00 4.18306351e-01
-3.15972477e-01 -4.26940382e-01 3.92165422e-01 -6.44139588e-01
-1.18134415e+00 8.81025076e-01 6.84637249e-01 4.45289165e-01
1.44399130e+00 -8.42514932e-01 7.75653124e-01 2.63389736e-01
5.75519383e-01 -1.04421854e+00 -3.83578986e-01 9.55922976e-02
7.84530461e-01 -9.27656114e-01 2.39283234e-01 -8.76358569e-01
-4.44765359e-01 1.75651765e+00 8.45389187e-01 -3.06590199e-01
4.35384125e-01 6.50694668e-01 1.15344360e-01 3.90553959e-02
-2.77531266e-01 -2.41506264e-01 2.36451834e-01 6.41798258e-01
2.10558429e-01 -4.19349551e-01 -1.71322092e-01 4.14341688e-01
2.00006608e-02 9.51469094e-02 6.34237707e-01 1.10633636e+00
-5.24416625e-01 -5.87606549e-01 -4.72771794e-01 5.01299679e-01
-6.56269789e-02 1.71021655e-01 -1.14497848e-01 1.06753290e+00
1.93430439e-01 1.05883110e+00 2.50016730e-02 -4.86241013e-01
8.90200973e-01 -4.47614372e-01 2.42646173e-01 -2.03781798e-01
-5.95580816e-01 -2.61865333e-02 -5.61573841e-02 -3.87677044e-01
-6.36082441e-02 -1.98093131e-01 -1.39208448e+00 -9.40527171e-02
-6.42207146e-01 2.10182473e-01 5.88237584e-01 6.57916129e-01
1.77296653e-01 8.84297013e-01 9.26119983e-01 -9.47468758e-01
-3.71678561e-01 -7.32182205e-01 -8.94546151e-01 -1.41398488e-02
2.82889549e-02 -7.31905103e-01 -3.87058049e-01 -2.35637605e-01] | [8.791184425354004, -0.9968221187591553] |
a38c65d4-a1c7-4b68-a7fb-0778e61dced5 | multi-task-learning-improves-performance-in | 2307.01401 | null | https://arxiv.org/abs/2307.01401v1 | https://arxiv.org/pdf/2307.01401v1.pdf | Multi-Task Learning Improves Performance In Deep Argument Mining Models | The successful analysis of argumentative techniques from user-generated text is central to many downstream tasks such as political and market analysis. Recent argument mining tools use state-of-the-art deep learning methods to extract and annotate argumentative techniques from various online text corpora, however each task is treated as separate and different bespoke models are fine-tuned for each dataset. We show that different argument mining tasks share common semantic and logical structure by implementing a multi-task approach to argument mining that achieves better performance than state-of-the-art methods for the same problems. Our model builds a shared representation of the input text that is common to all tasks and exploits similarities between tasks in order to further boost performance via parameter-sharing. Our results are important for argument mining as they show that different tasks share substantial similarities and suggest a holistic approach to the extraction of argumentative techniques from text. | ['Marco Morucci', 'Isaac Mehlhaff', 'Shashank Shekhar', 'Amirhossein Farzam'] | 2023-07-03 | null | null | null | null | ['multi-task-learning', 'argument-mining'] | ['methodology', 'natural-language-processing'] | [ 1.05287530e-01 6.59001827e-01 -7.48686373e-01 -4.87066865e-01
-1.26917183e+00 -9.79471207e-01 1.15557325e+00 7.89594293e-01
-5.68066001e-01 7.86486149e-01 7.88292646e-01 -1.12576449e+00
-4.30053115e-01 -7.23332345e-01 -8.09034646e-01 -8.00148621e-02
1.76538914e-01 9.39461708e-01 4.41561863e-02 -4.63191092e-01
5.96263409e-01 -6.64421618e-02 -1.30344880e+00 1.18486750e+00
6.26817405e-01 6.69617653e-01 -5.11379123e-01 6.40403271e-01
-7.54083455e-01 1.12809718e+00 -5.82661510e-01 -8.91161501e-01
9.90866125e-03 -2.01867998e-01 -1.71700728e+00 -7.60467887e-01
3.34257543e-01 2.21605152e-02 3.14282417e-01 6.31283164e-01
4.43796903e-01 -2.16945991e-01 8.50421071e-01 -1.08855462e+00
-4.34869677e-01 1.40527320e+00 -5.11926234e-01 3.95794183e-01
1.38147265e-01 -5.00594795e-01 1.62322628e+00 -6.46697640e-01
9.50261176e-01 1.45291054e+00 7.00265527e-01 2.88074940e-01
-1.21193159e+00 -6.07006848e-01 3.26888055e-01 2.21873391e-02
-1.95042178e-01 -4.04352605e-01 9.22720075e-01 -6.13866806e-01
1.23874474e+00 1.05048336e-01 5.74948967e-01 1.33042407e+00
-7.13100359e-02 1.22777879e+00 1.14054394e+00 -5.05557358e-01
1.54142063e-02 -1.84821784e-01 4.69775617e-01 5.20607054e-01
5.20102739e-01 -4.01781291e-01 -5.89513481e-01 -6.95632517e-01
6.86462224e-02 -4.48680699e-01 4.40142423e-01 -2.62364656e-01
-1.18896985e+00 1.60327446e+00 2.30943435e-03 3.86607140e-01
-1.88439041e-01 2.26243228e-01 1.26522470e+00 4.87923592e-01
9.25205350e-01 7.93148160e-01 -1.04930711e+00 -2.49994352e-01
-7.65585542e-01 9.55081642e-01 1.27870810e+00 4.72433269e-01
4.23340350e-01 -5.61269045e-01 -2.13419914e-01 9.32923555e-01
1.94947451e-01 2.21451551e-01 4.47714537e-01 -1.05378854e+00
1.01317978e+00 9.02389467e-01 3.02649923e-02 -6.61234498e-01
-5.32045186e-01 -4.04248148e-01 -7.13864015e-03 2.28606418e-01
8.12551498e-01 -4.44004744e-01 -4.80481565e-01 1.59578824e+00
4.60984111e-01 -8.28769982e-01 2.33712375e-01 5.12749910e-01
1.13338900e+00 2.08435476e-01 6.10256612e-01 2.85379410e-01
1.69975221e+00 -8.50267947e-01 -4.45878476e-01 -5.81238031e-01
1.21097541e+00 -1.02528942e+00 9.98067617e-01 2.45904431e-01
-1.11095500e+00 5.76837063e-02 -9.52134430e-01 -6.33629203e-01
-5.82164109e-01 5.74600231e-03 9.97164249e-01 3.99962872e-01
-2.91464210e-01 6.66957200e-01 -2.18969166e-01 5.46702705e-02
9.59799945e-01 1.17404714e-01 -1.66609675e-01 5.48514605e-01
-1.21916091e+00 8.97676945e-01 4.40633774e-01 -5.31396449e-01
-3.79090577e-01 -7.57870138e-01 -8.15109193e-01 1.03313491e-01
7.13838518e-01 -8.51689517e-01 1.67691255e+00 -1.13522840e+00
-1.04521108e+00 1.56805301e+00 -5.65270595e-02 -8.11615467e-01
6.64579690e-01 -8.03023338e-01 3.22623849e-02 -1.61878258e-01
7.04480827e-01 2.27617025e-01 8.36083949e-01 -8.84033024e-01
-8.11017931e-01 -1.98953658e-01 1.73002258e-01 1.45077720e-01
-9.66387913e-02 6.90933645e-01 3.78453702e-01 -6.47780299e-01
-1.89027146e-01 -5.86108863e-01 -1.37154646e-02 -3.36422116e-01
-5.11634529e-01 -9.16836202e-01 9.21958208e-01 -3.84502947e-01
1.06670058e+00 -1.67540395e+00 7.82694624e-05 7.20698908e-02
5.22604406e-01 1.31290957e-01 4.97463867e-02 5.17426372e-01
-2.31234193e-01 6.29890382e-01 1.30007779e-02 -2.52044976e-01
2.33065605e-01 1.44969821e-01 -4.25426513e-01 3.01780581e-01
1.96308643e-01 1.12820816e+00 -1.08017850e+00 -7.90626466e-01
-1.28431603e-01 -1.47033766e-01 -5.30353189e-01 -2.03492954e-01
-5.83188057e-01 -5.70770279e-02 -1.03170455e+00 3.91724050e-01
1.75337017e-01 -5.47087610e-01 5.67249596e-01 2.38421746e-03
-1.40944228e-01 1.40333414e+00 -7.92675316e-01 1.68246651e+00
-4.60128278e-01 7.79274464e-01 2.59859622e-01 -1.54639328e+00
5.35931349e-01 3.27234417e-01 3.65474075e-01 -7.71235704e-01
4.50522393e-01 8.13193977e-01 1.93603814e-01 -3.70136321e-01
2.79035419e-01 -8.97011086e-02 -3.92388940e-01 1.38786328e+00
-4.97067124e-02 -2.39991825e-02 4.95930552e-01 4.50073838e-01
1.16123378e+00 1.94475755e-01 7.28451729e-01 -7.70698667e-01
2.93851137e-01 5.24256170e-01 4.50979114e-01 8.37351739e-01
2.32604727e-01 -1.98472738e-02 8.27058733e-01 -1.06229544e+00
-1.42862928e+00 -5.50489306e-01 -1.79090276e-01 1.54348540e+00
-4.56104338e-01 -5.70024908e-01 -5.23752630e-01 -1.35893178e+00
3.94745767e-01 4.22447741e-01 -8.51483583e-01 5.92376649e-01
-1.10408235e+00 -8.29518616e-01 6.72020912e-01 5.49764931e-01
3.41376215e-01 -1.29126763e+00 -1.15365994e+00 5.35159945e-01
-4.28238422e-01 -1.00685823e+00 3.41152012e-01 7.01324403e-01
-6.49150550e-01 -1.68010461e+00 -2.57259190e-01 -6.46023512e-01
1.96959041e-02 -3.57051753e-02 1.74027753e+00 3.13422084e-01
-7.07356483e-02 -1.29214793e-01 -3.43225688e-01 -1.12604666e+00
-9.30363953e-01 5.90991676e-01 -4.69357282e-01 -6.15832567e-01
5.46115279e-01 -3.73218149e-01 -3.38771909e-01 -2.09537745e-01
-4.66107786e-01 2.50587940e-01 2.21533373e-01 9.49295461e-01
1.81523964e-01 -4.83949482e-01 6.79879189e-01 -1.52485502e+00
1.33719850e+00 -6.20843947e-01 -3.44137013e-01 -3.20192077e-03
-7.50103533e-01 5.35359383e-01 4.59321260e-01 -2.69055456e-01
-1.19223702e+00 -6.07089400e-01 -3.14995386e-02 4.43455011e-01
6.32541627e-02 6.34633243e-01 3.70070875e-01 5.41647196e-01
9.77453411e-01 -7.84688532e-01 1.81850523e-01 -5.03016174e-01
7.71701932e-01 5.11366248e-01 2.33769283e-01 -1.21938109e+00
6.31406307e-01 7.53682792e-01 5.19697405e-02 -3.62158835e-01
-1.60161746e+00 -3.16738963e-01 -2.10557669e-01 3.47864091e-01
6.85753644e-01 -8.02862167e-01 -8.77456427e-01 1.05917707e-01
-1.42115140e+00 -6.07687771e-01 -4.12500948e-01 1.25507846e-01
-5.39877474e-01 2.23448649e-01 -7.63311565e-01 -7.00577378e-01
-9.04955387e-01 -7.12805629e-01 1.15201890e+00 -7.08449632e-02
-9.29862678e-01 -1.31109345e+00 2.51834244e-01 6.54692292e-01
4.88848612e-02 1.71607673e-01 1.44681144e+00 -1.39137077e+00
1.13772966e-01 2.90852249e-01 -5.41275978e-01 6.42060786e-02
-2.01803476e-01 -1.59175605e-01 -8.77288043e-01 2.91403025e-01
-1.90518603e-01 -7.36336231e-01 1.18514562e+00 5.59640288e-01
1.01156914e+00 -5.76129317e-01 -6.05979323e-01 1.79457948e-01
8.00917506e-01 -3.24352026e-01 2.29255006e-01 1.36945796e+00
4.62914735e-01 1.06526661e+00 6.50397599e-01 1.66887760e-01
2.16334239e-01 2.32967526e-01 2.17353567e-01 -3.94853383e-01
8.09561685e-02 -1.46582574e-01 -1.46501837e-02 -8.68789107e-02
4.64045070e-02 -1.33177340e-01 -1.15359414e+00 8.66439879e-01
-2.38091516e+00 -1.29535139e+00 -6.43868744e-01 1.47248328e+00
1.06857896e+00 5.64765871e-01 3.14427733e-01 4.58441585e-01
2.82543719e-01 2.16585934e-01 -1.61692798e-01 -1.05522919e+00
-2.59027094e-01 7.37447917e-01 3.82076025e-01 3.76248688e-01
-1.47587597e+00 1.06509316e+00 6.85134935e+00 8.37258101e-01
-6.96700454e-01 1.73188075e-01 8.13735366e-01 -1.30156711e-01
-6.33009613e-01 3.17760050e-01 -6.77005053e-01 2.24013731e-01
6.97720408e-01 -2.52739459e-01 -3.84475291e-02 1.03789175e+00
7.52954334e-02 -5.62743917e-02 -1.18811321e+00 4.08930033e-01
-2.02435285e-01 -2.06356049e+00 -6.97095096e-02 3.24097484e-01
5.38939357e-01 3.16581249e-01 -1.60169587e-01 2.48924866e-01
1.09862733e+00 -1.05326867e+00 1.07736504e+00 -2.97215432e-01
4.08908665e-01 -4.94235456e-01 8.26755106e-01 2.99928278e-01
-8.11833262e-01 -4.84088719e-01 2.04403475e-01 -5.49382925e-01
1.96866095e-01 7.35485911e-01 -7.91057706e-01 4.28534776e-01
6.66353822e-01 5.13655543e-01 -4.37420636e-01 1.65192276e-01
-5.70448101e-01 7.12782621e-01 -3.73370886e-01 -1.75519601e-01
5.85839272e-01 1.84676751e-01 5.41660190e-01 1.32056713e+00
-1.32692531e-01 -9.00525376e-02 1.33790269e-01 9.49418366e-01
-4.72339481e-01 4.79887009e-01 -7.56333470e-01 -2.77032495e-01
4.82719392e-01 1.29400516e+00 -6.44450903e-01 -6.18221998e-01
-4.41505402e-01 2.61111438e-01 7.08083510e-01 -3.57644297e-02
-5.79581559e-01 4.46823649e-02 5.85505366e-01 1.96952149e-01
2.16234535e-01 -4.24287245e-02 -8.64789903e-01 -9.21991527e-01
-3.45224887e-02 -1.15627217e+00 1.09148490e+00 -2.85696089e-01
-1.52366710e+00 1.21453702e-01 1.97748452e-01 -6.18025661e-01
-7.76542723e-01 -8.75151098e-01 -7.74483860e-01 7.62039423e-01
-1.50720251e+00 -1.50515866e+00 2.52888709e-01 2.73221552e-01
6.63112760e-01 -3.85844350e-01 8.25744629e-01 -1.24044858e-01
-1.28478371e-02 1.95221156e-01 -1.04813658e-01 5.44051409e-01
7.63305902e-01 -1.36243808e+00 8.10826242e-01 4.51601177e-01
4.21401381e-01 7.02312469e-01 5.42429805e-01 -7.08224475e-01
-9.70235050e-01 -5.48893154e-01 1.33498585e+00 -9.08506811e-01
1.14937043e+00 -1.17495142e-01 -6.78746402e-01 4.98697132e-01
4.41350430e-01 -4.09159839e-01 9.45371866e-01 1.29794705e+00
-8.07027638e-01 5.23033023e-01 -6.94802046e-01 5.18080890e-01
1.21102893e+00 -7.36095428e-01 -1.24310541e+00 4.04092908e-01
3.51863921e-01 -3.30970585e-01 -3.39469522e-01 1.67764708e-01
7.97120988e-01 -8.03293288e-01 9.91969228e-01 -1.22397184e+00
1.17741239e+00 2.99996287e-01 1.62860006e-01 -1.07796395e+00
-3.13292742e-02 -5.66194832e-01 -2.14890704e-01 1.13723600e+00
9.94505167e-01 -5.81512809e-01 6.59554601e-01 4.14400041e-01
-1.43503565e-02 -9.15596783e-01 -6.73882544e-01 -4.18668926e-01
7.11293519e-01 -5.72230697e-01 5.26503742e-01 1.22473991e+00
1.30604059e-01 1.13797772e+00 2.30464831e-01 -7.36100316e-01
6.67600513e-01 7.72993565e-01 1.01340759e+00 -1.92591763e+00
-2.12941289e-01 -8.73227954e-01 5.14584422e-01 -5.19428134e-01
7.41013229e-01 -1.14965093e+00 -5.33698499e-01 -1.78003705e+00
4.26206648e-01 -5.93439400e-01 6.27256781e-02 6.41097009e-01
-7.38795921e-02 -1.68778494e-01 -1.13689363e-01 2.50252664e-01
-4.74703282e-01 7.96941072e-02 7.74621129e-01 -2.85938650e-01
-1.76479191e-01 -2.62932539e-01 -1.20033109e+00 1.17426157e+00
7.22246051e-01 -7.97174335e-01 -2.76957482e-01 -6.57692134e-01
1.12705350e+00 -5.25715590e-01 4.27494287e-01 -1.27676114e-01
-1.49950042e-01 -1.86106473e-01 2.33262494e-01 -5.35110474e-01
-1.19533867e-01 -9.33548510e-02 -5.42515814e-01 2.41415873e-01
-6.02933347e-01 1.19335335e-02 1.29315943e-01 4.61378992e-01
-6.59490228e-02 -3.98367196e-01 3.36608946e-01 -4.88264233e-01
-2.65394807e-01 -3.62918675e-01 -2.83018678e-01 7.75012434e-01
4.81787950e-01 -7.60884956e-03 -8.73589635e-01 -1.30814746e-01
-4.09590334e-01 2.08935648e-01 1.24701105e-01 2.61680722e-01
1.77181680e-02 -8.25063288e-01 -1.23478758e+00 -4.07199591e-01
1.56060476e-02 9.37050208e-02 -3.47677737e-01 6.07991755e-01
-3.25648218e-01 3.57728302e-01 -2.74235215e-02 -2.30680138e-01
-1.54115558e+00 3.13479751e-01 -1.48912687e-02 -1.01053905e+00
-8.04915130e-01 7.78880239e-01 6.33173138e-02 -4.25009191e-01
-2.72130281e-01 -5.55429682e-02 -3.21938038e-01 5.16497850e-01
2.07855567e-01 3.76159288e-02 1.14258222e-01 -2.91599967e-02
-3.63124609e-01 1.42898411e-01 -4.26942348e-01 -3.90294373e-01
1.64324749e+00 2.66111672e-01 -1.85260817e-01 1.76917911e-01
8.23045850e-01 3.58549416e-01 -5.72929680e-01 -2.62724906e-01
5.11841178e-01 -8.87911916e-02 -5.27574010e-02 -8.87905657e-01
-4.34743553e-01 6.94720626e-01 -3.65471780e-01 4.68787342e-01
2.97167689e-01 5.64616442e-01 6.78286195e-01 6.65642440e-01
-5.43587208e-02 -1.74555874e+00 7.15452284e-02 8.22952509e-01
7.06488311e-01 -1.30524969e+00 5.00467479e-01 -5.88700771e-01
-4.63841051e-01 1.15694904e+00 1.18312053e-01 -9.91584510e-02
3.95088911e-01 5.76336503e-01 -3.17405090e-02 -9.55489278e-01
-8.70922089e-01 -2.04728127e-01 1.96040615e-01 1.94690704e-01
1.06878042e+00 4.65327427e-02 -9.80699599e-01 8.10088634e-01
-4.61959958e-01 -4.50608701e-01 7.63205290e-02 1.12507641e+00
-1.69786885e-01 -1.59253573e+00 -1.44168377e-01 6.23672009e-01
-1.00420177e+00 -3.28640938e-01 -9.90009725e-01 1.14534342e+00
-1.34330943e-01 9.21713293e-01 -7.66430721e-02 2.62437433e-01
-1.53242379e-01 4.83352900e-01 4.48367804e-01 -7.72891045e-01
-1.19838774e+00 1.78377461e-02 1.36279702e+00 -4.93082076e-01
-6.11825705e-01 -7.34575331e-01 -1.41159129e+00 -4.30861413e-01
-1.32132426e-01 5.11345863e-01 7.77630031e-01 1.38168776e+00
4.11205977e-01 4.09462512e-01 -4.28051092e-02 -3.85768265e-01
-5.72851539e-01 -8.02777350e-01 5.95384128e-02 9.31146920e-01
-8.69978499e-03 -8.21844101e-01 -2.34701708e-01 -1.12243861e-01] | [9.586353302001953, 9.613236427307129] |
b93e8a89-8926-4e17-b2e0-8f7f2d99e65a | paddlespeech-an-easy-to-use-all-in-one-speech-1 | 2205.12007 | null | https://arxiv.org/abs/2205.12007v1 | https://arxiv.org/pdf/2205.12007v1.pdf | PaddleSpeech: An Easy-to-Use All-in-One Speech Toolkit | PaddleSpeech is an open-source all-in-one speech toolkit. It aims at facilitating the development and research of speech processing technologies by providing an easy-to-use command-line interface and a simple code structure. This paper describes the design philosophy and core architecture of PaddleSpeech to support several essential speech-to-text and text-to-speech tasks. PaddleSpeech achieves competitive or state-of-the-art performance on various speech datasets and implements the most popular methods. It also provides recipes and pretrained models to quickly reproduce the experimental results in this paper. PaddleSpeech is publicly avaiable at https://github.com/PaddlePaddle/PaddleSpeech. | ['Liang Huang', 'Yanjun Ma', 'dianhai yu', 'Xiaoguang Hu', 'Zeyu Chen', 'Enlei Gong', 'Xiaojie Chen', 'Yuxin Huang', 'Renjie Zheng', 'Xintong Li', 'Junkun Chen', 'Tian Yuan', 'HUI ZHANG'] | 2022-05-20 | paddlespeech-an-easy-to-use-all-in-one-speech | https://aclanthology.org/2022.naacl-demo.12 | https://aclanthology.org/2022.naacl-demo.12.pdf | naacl-acl-2022-7 | ['environmental-sound-classification', 'speech-to-text-translation', 'keyword-spotting', 'speaker-identification'] | ['audio', 'natural-language-processing', 'speech', 'speech'] | [-2.97641009e-01 -1.71158053e-02 -1.01712465e-01 -6.24382854e-01
-1.10157466e+00 -5.50916493e-01 7.68170774e-01 -1.51578948e-01
-2.16917500e-01 2.89098889e-01 4.37001914e-01 -9.02910173e-01
3.89197528e-01 -2.66196281e-01 -1.74480975e-01 -5.79129398e-01
1.60119981e-01 6.13421023e-01 2.80156642e-01 -3.42697918e-01
-7.87540302e-02 1.94449693e-01 -1.32311833e+00 7.28215575e-01
4.24318582e-01 6.16231084e-01 7.66588211e-01 1.13022745e+00
-4.13436741e-01 8.11400175e-01 -5.24867833e-01 -4.36679691e-01
-1.10590957e-01 -2.28977069e-01 -9.19633150e-01 -1.97425947e-01
-7.64339371e-03 -2.47300237e-01 -4.20312375e-01 6.88564360e-01
8.16109002e-01 2.14282691e-01 2.25375265e-01 -1.04650414e+00
-5.44559658e-01 7.79366851e-01 5.27128354e-02 6.04011476e-01
6.01741254e-01 1.14395842e-01 9.49996948e-01 -1.38994670e+00
1.99589342e-01 1.36698449e+00 3.12866062e-01 5.94047129e-01
-7.70369887e-01 -5.14688134e-01 -3.55498046e-02 -4.84483577e-02
-1.33652008e+00 -1.27763891e+00 3.59342754e-01 -1.61825180e-01
1.71809733e+00 6.56096101e-01 2.30471626e-01 1.41895306e+00
8.43876880e-03 1.15180767e+00 7.72098362e-01 -5.31431794e-01
1.76059399e-02 -5.63809201e-02 1.90320238e-01 5.46153247e-01
-5.45397758e-01 -1.30333275e-01 -6.94981217e-01 -1.45744935e-01
6.61686599e-01 -2.75166482e-01 -1.97584018e-01 4.03213114e-01
-1.05286801e+00 5.03101408e-01 -2.52828181e-01 5.60269773e-01
-3.18327904e-01 -1.22326523e-01 5.82040668e-01 4.03048903e-01
5.37018120e-01 -8.19681808e-02 -7.49466598e-01 -9.06967282e-01
-7.58565128e-01 5.80223575e-02 9.22042012e-01 1.26168418e+00
2.88790494e-01 4.99557644e-01 6.87114075e-02 1.44553101e+00
4.17758435e-01 6.15552902e-01 7.09369600e-01 -7.71745503e-01
9.14416730e-01 -6.68732226e-02 -2.63171226e-01 -3.68365884e-01
-3.15930307e-01 -1.20574050e-01 -5.55852473e-01 -1.24238558e-01
5.52031212e-02 -1.85420409e-01 -9.35605466e-01 1.03573716e+00
2.18626812e-01 -1.19126357e-01 1.41773045e-01 6.88856423e-01
1.30855787e+00 1.31081617e+00 6.35779351e-02 -1.89836651e-01
1.39713216e+00 -1.40644884e+00 -9.53458548e-01 -6.57608211e-01
5.93433201e-01 -1.19054234e+00 1.51444137e+00 2.81708211e-01
-1.40919268e+00 -4.67682689e-01 -5.68033636e-01 -2.93821722e-01
-5.33901274e-01 3.31162095e-01 3.75430197e-01 8.07248890e-01
-1.25571227e+00 2.61200905e-01 -1.12568724e+00 -4.98204648e-01
-5.80335222e-02 8.84688273e-02 -3.46267909e-01 2.29118645e-01
-9.86568809e-01 7.58884728e-01 3.96937728e-01 -2.23262891e-01
-6.18439734e-01 -5.43040454e-01 -8.12437296e-01 3.98200154e-02
5.39052606e-01 -3.40671927e-01 2.17267227e+00 -5.38968623e-01
-2.13540363e+00 7.07867503e-01 -5.36629736e-01 -3.39868963e-01
1.57833844e-01 -4.31097776e-01 -9.02332246e-01 5.52239865e-02
-1.04035012e-01 3.38222474e-01 8.28217387e-01 -5.97133279e-01
-4.40794945e-01 -6.62916899e-02 -6.21636987e-01 3.70652020e-01
-2.16391608e-01 8.49503756e-01 -9.71593618e-01 -9.67979968e-01
-4.53599915e-02 -6.94757044e-01 2.63469577e-01 -4.22293603e-01
-4.32259262e-01 -3.75732780e-01 1.04087853e+00 -7.90634692e-01
1.52887654e+00 -2.31729007e+00 -2.44249746e-01 -3.28255355e-01
-1.20355681e-01 8.42475653e-01 -1.97144881e-01 1.20711565e+00
-1.86687976e-01 1.68601528e-01 4.44799736e-02 -9.37682927e-01
2.20098272e-01 3.76667529e-01 -3.32174122e-01 7.67867789e-02
-1.92181580e-02 7.91417301e-01 -7.21795619e-01 -2.57078767e-01
8.36067259e-01 4.69386697e-01 -1.65465951e-01 4.23731863e-01
6.12062886e-02 1.71545923e-01 -2.25665689e-01 5.15100241e-01
4.76908565e-01 4.52205576e-02 1.61796167e-01 3.04683805e-01
-4.10202593e-01 1.34607804e+00 -7.29017973e-01 1.60011113e+00
-6.08161509e-01 6.77023292e-01 4.66036171e-01 -5.89303315e-01
7.36412942e-01 7.70751119e-01 3.34802619e-03 -4.36835229e-01
2.92023838e-01 3.76722753e-01 -8.96722823e-02 -4.21180367e-01
5.83358884e-01 1.37193024e-01 1.55978903e-01 4.31445777e-01
3.25588584e-01 -4.95990694e-01 1.87908396e-01 3.13220203e-01
9.03756082e-01 -3.71602833e-01 7.31233299e-01 -1.59290209e-01
5.19741118e-01 -4.84218746e-01 1.81475505e-01 5.54022253e-01
-1.52840555e-01 4.97805685e-01 6.63238540e-02 -6.50859624e-02
-1.03289509e+00 -1.17118597e+00 -3.18827317e-03 1.50395918e+00
-7.35329926e-01 -9.80090022e-01 -8.45996201e-01 -3.63675326e-01
-3.69520366e-01 9.73131120e-01 1.81881785e-02 4.50498044e-01
-3.11465323e-01 -9.91294160e-02 8.53782237e-01 4.67567861e-01
4.93869841e-01 -1.31035650e+00 1.64226398e-01 2.47864068e-01
-3.97540748e-01 -1.51790440e+00 -8.51651192e-01 1.97951362e-01
-3.85936171e-01 -4.90410924e-01 -6.28807843e-01 -9.86558020e-01
8.42067227e-02 5.81045389e-01 8.75142336e-01 -9.57920998e-02
2.30993018e-01 9.26183537e-02 -6.63924158e-01 -4.64846909e-01
-8.71853530e-01 2.54948944e-01 1.29550233e-01 -3.08451831e-01
3.77524465e-01 -3.96594584e-01 -1.11107461e-01 3.13814729e-01
-5.58695853e-01 1.82639256e-01 3.27798963e-01 4.55836743e-01
2.76991993e-01 3.48552354e-02 3.92548531e-01 -3.91698420e-01
9.55147326e-01 -2.21810609e-01 -4.00100261e-01 4.27271239e-03
-2.80213773e-01 -3.33216608e-01 6.83079898e-01 -1.17265806e-01
-9.93401587e-01 -9.52717364e-02 -1.31357443e+00 -2.93619841e-01
-5.09665608e-01 5.39301038e-01 -2.62868553e-01 3.09141368e-01
8.67896751e-02 5.67693353e-01 -5.15968055e-02 -7.43290603e-01
4.94177550e-01 1.49697864e+00 5.67048728e-01 -4.12765622e-01
3.73000830e-01 -1.95972323e-02 -9.62955236e-01 -1.44923031e+00
-5.60062945e-01 -8.02733779e-01 -3.77971232e-01 -5.74175753e-02
6.87096477e-01 -8.45977366e-01 -4.49737042e-01 8.41730654e-01
-1.13986313e+00 -7.99553394e-01 -1.85208321e-02 2.96014041e-01
-5.59366465e-01 4.21487719e-01 -9.97734606e-01 -9.97289717e-01
-7.15164721e-01 -1.23223925e+00 1.27749848e+00 -8.69360194e-02
-4.13474768e-01 -9.92688179e-01 -1.26094177e-01 4.04601455e-01
2.92047024e-01 -7.59747088e-01 3.59001368e-01 -1.03341997e+00
-4.05751960e-03 1.19292907e-01 8.91027302e-02 6.21244133e-01
2.66037047e-01 3.13652188e-01 -1.15050840e+00 -2.78389752e-01
-3.68438326e-02 -2.11012140e-01 4.85112607e-01 3.52257937e-01
1.14191425e+00 -6.21028006e-01 -1.25845537e-01 4.52803761e-01
6.38285816e-01 5.81909001e-01 6.70156360e-01 6.73494339e-02
3.76672298e-01 4.94414568e-01 4.11631137e-01 4.01263237e-01
5.67417800e-01 8.28813910e-01 -8.86817798e-02 8.35593417e-02
-1.99542344e-01 -4.29986775e-01 6.87464595e-01 1.76703167e+00
5.38053691e-01 -5.47263920e-01 -1.27957845e+00 5.56935132e-01
-1.72761595e+00 -1.03531837e+00 -3.82372260e-01 1.94417179e+00
8.99310291e-01 1.82870612e-01 3.94715995e-01 1.68643340e-01
5.81592739e-01 3.86497140e-01 4.33120802e-02 -7.92838037e-01
-2.61155218e-02 3.48169029e-01 4.74113114e-02 8.72588098e-01
-1.09622157e+00 1.49232864e+00 6.87265539e+00 1.12520778e+00
-1.20770812e+00 3.77566278e-01 3.82226408e-01 4.72056121e-02
1.12225441e-02 -2.86798030e-01 -8.81081820e-01 4.61895287e-01
1.54416645e+00 -2.78912514e-01 5.68746209e-01 8.77130210e-01
7.14806616e-01 1.46553293e-01 -6.04518950e-01 1.10337722e+00
-2.21665755e-01 -1.43294394e+00 1.92371011e-02 -7.05867410e-02
3.48371267e-02 8.13214719e-01 -6.16623126e-02 3.02355587e-01
4.66442049e-01 -6.40492201e-01 7.98731685e-01 -2.83547103e-01
9.27466512e-01 -5.52641034e-01 5.42338371e-01 4.88135427e-01
-1.43072355e+00 4.83938232e-02 -1.73578456e-01 -1.55901343e-01
4.88420397e-01 3.43893021e-01 -1.16063762e+00 5.02767742e-01
8.10059965e-01 5.42765319e-01 -1.60220906e-01 6.93898916e-01
-4.33268368e-01 1.06312382e+00 -3.81012827e-01 -1.72622621e-01
3.93751591e-01 -3.81511748e-02 4.93949592e-01 1.85511315e+00
3.41887832e-01 4.89209175e-01 2.23244861e-01 2.67082900e-01
1.21211573e-01 4.13670361e-01 -5.95298767e-01 -5.07867813e-01
1.07321727e+00 1.14150047e+00 -4.68356311e-01 -4.85208541e-01
-6.25403404e-01 9.19241965e-01 1.97264329e-02 2.35386252e-01
-5.70568264e-01 -5.86271167e-01 1.03865016e+00 1.36762187e-01
3.62437487e-01 -7.98326075e-01 -1.19660668e-01 -9.85762358e-01
-6.68634251e-02 -1.25482559e+00 1.91299990e-01 -1.16826928e+00
-1.04906380e+00 9.69623208e-01 -8.52920935e-02 -7.44002402e-01
-6.20086610e-01 -7.45425820e-01 -8.97174060e-01 9.80741382e-01
-1.09279394e+00 -9.85000193e-01 -9.63682458e-02 7.00116634e-01
1.32560170e+00 -4.68511045e-01 1.04180598e+00 2.29454294e-01
-7.61450648e-01 4.68177199e-01 9.88096222e-02 3.45716923e-01
4.79558915e-01 -1.08245099e+00 1.60392165e+00 9.60826814e-01
2.36133322e-01 7.94037879e-01 6.95422947e-01 -5.05023956e-01
-1.44544733e+00 -8.52635503e-01 1.28105366e+00 -3.03596258e-01
1.13764071e+00 -6.71957910e-01 -8.39536667e-01 8.02075505e-01
5.65451980e-01 -4.64435637e-01 6.89396441e-01 6.16575740e-02
-2.81236976e-01 9.12230462e-02 -7.17943907e-01 8.08332920e-01
8.87029409e-01 -1.00082016e+00 -6.75684512e-01 2.67655849e-01
9.97468352e-01 -6.03801370e-01 -7.02939451e-01 -2.03764871e-01
3.13293129e-01 -7.64994323e-01 6.57337368e-01 -2.13538721e-01
2.50049680e-02 -1.18405789e-01 -3.24208081e-01 -1.46658754e+00
-1.44287661e-01 -1.35202324e+00 -1.72548085e-01 1.34740973e+00
6.31239355e-01 -1.00970113e+00 3.80920410e-01 1.46477506e-01
-8.58106911e-01 -5.67828417e-01 -1.10776830e+00 -1.04132581e+00
1.26859188e-01 -1.13159752e+00 7.27661729e-01 6.93621933e-01
4.01371658e-01 4.27056491e-01 -2.26170704e-01 2.31650084e-01
9.77347791e-02 -6.06462479e-01 8.95645022e-01 -5.45298338e-01
-3.49677086e-01 -4.50660288e-01 -2.29143947e-02 -1.63447857e+00
9.46044698e-02 -7.75100470e-01 5.51840141e-02 -1.63075376e+00
-3.84692430e-01 -1.31255105e-01 2.24026293e-01 9.26614881e-01
8.63736346e-02 -2.33739883e-01 3.91162604e-01 6.79570884e-02
-1.60101548e-01 5.55127323e-01 9.07695115e-01 8.19336325e-02
-2.12094530e-01 1.33213520e-01 -5.59043109e-01 4.52728331e-01
1.37356007e+00 -3.27988952e-01 -5.96625090e-01 -6.04680002e-01
-4.52200413e-01 5.41613698e-02 7.74305761e-02 -8.53514314e-01
2.40230769e-01 -8.41980651e-02 -3.73419046e-01 -8.38818073e-01
8.72781456e-01 -4.40725654e-01 -1.48690492e-01 1.60837546e-02
-5.35733923e-02 3.53945613e-01 4.49024260e-01 -1.33731201e-01
-3.34339887e-01 -1.11077830e-01 6.65857196e-01 -2.25945003e-02
-7.39200532e-01 1.42196521e-01 -8.55926514e-01 4.14203145e-02
5.60807049e-01 -3.33408527e-02 -6.40472233e-01 -7.62607217e-01
-6.85536981e-01 6.61124736e-02 1.67396385e-02 1.00975323e+00
6.66531503e-01 -9.69646811e-01 -7.01039910e-01 7.01299131e-01
1.15061566e-01 -4.57424611e-01 4.06442657e-02 6.53856933e-01
-4.97633874e-01 9.68582869e-01 3.48153412e-01 -4.41322774e-01
-1.76551509e+00 2.88944125e-01 4.34132546e-01 1.32054985e-01
-6.65478230e-01 9.02643442e-01 1.53134856e-02 -8.36709261e-01
2.85502344e-01 -5.16976535e-01 2.81872660e-01 -4.42155659e-01
9.90186989e-01 3.26860189e-01 4.74588990e-01 -6.92823887e-01
-5.99595129e-01 -1.61210880e-01 -2.36821339e-01 -5.83426893e-01
1.24572408e+00 -3.06201160e-01 4.73899096e-02 7.04593122e-01
1.07141864e+00 5.64074405e-02 -8.42185915e-01 -4.28685173e-03
-4.97056320e-02 -3.00868124e-01 2.43670195e-01 -9.18643415e-01
-7.45653152e-01 1.02286482e+00 1.31214708e-01 4.99816895e-01
8.99963677e-01 1.55396312e-01 1.02275002e+00 4.13632214e-01
4.15829062e-01 -1.19343257e+00 -1.75926596e-01 1.10352170e+00
1.17083275e+00 -9.52380598e-01 -3.90489310e-01 -7.36062646e-01
-8.52397799e-01 9.84109104e-01 2.64321029e-01 5.71314216e-01
8.89024854e-01 7.68193185e-01 5.93398511e-01 1.43845543e-01
-1.17071986e+00 -3.31181407e-01 9.55240428e-02 6.84648097e-01
8.59451115e-01 2.26532742e-01 -1.04332596e-01 5.33239007e-01
-6.21004641e-01 -1.83501720e-01 2.60457486e-01 8.92018437e-01
-6.00437224e-01 -1.49371028e+00 -3.55409712e-01 9.27442014e-02
-5.53175211e-01 -6.34912610e-01 -6.27515435e-01 6.78890049e-01
-6.24805629e-01 1.50120199e+00 2.76661143e-02 -4.70665574e-01
3.29143316e-01 2.69190401e-01 1.48078175e-02 -8.88885915e-01
-6.96269155e-01 6.25062943e-01 7.25323915e-01 -5.16276062e-01
2.12111950e-01 -6.94363892e-01 -1.44714415e+00 -7.39165664e-01
-1.15631692e-01 4.03522968e-01 9.20174539e-01 8.45085442e-01
6.26560032e-01 4.42992151e-01 5.15681684e-01 -8.33238244e-01
-4.14968431e-01 -1.21397960e+00 -4.69110221e-01 -2.77894735e-01
3.48032594e-01 -2.02129677e-01 -2.22053736e-01 -1.01879463e-01] | [14.400145530700684, 6.890883922576904] |
bd739298-2658-45d1-932c-3d346f160615 | sleep-apnea-detection-from-single-lead-ecg-a | null | null | https://ieeexplore.ieee.org/abstract/document/9714370 | https://ieeexplore.ieee.org/abstract/document/9714370 | Sleep Apnea Detection From Single-Lead ECG: A Comprehensive Analysis of Machine Learning and Deep Learning Algorithms | https://ieeexplore.ieee.org/abstract/document/9714370 | ['Mohamad', 'Mahsa ; Forouzanfar', 'Bahrami'] | 2022-02-15 | null | null | null | ieee-transactions-on-instrumentation-and-1 | ['sleep-apnea-detection'] | ['medical'] | [-3.13157916e-01 -2.12861970e-02 -3.29496741e-01 -8.66282284e-02
-6.42045856e-01 1.75915539e-01 -2.73249522e-02 8.82882550e-02
-3.50012392e-01 1.08797061e+00 3.42663795e-01 -2.71520108e-01
-7.25315139e-02 -7.58162916e-01 -1.54720386e-02 -8.51412416e-01
-1.62162296e-02 2.78586954e-01 -1.84649095e-01 1.31346747e-01
-1.51133463e-01 2.36557782e-01 -1.30721200e+00 1.85864687e-01
1.22698009e+00 1.31505632e+00 1.94594618e-02 5.61951399e-01
2.93717571e-02 -1.64390523e-02 -9.02085662e-01 8.46068412e-02
3.11325043e-01 -4.23574179e-01 -3.25371385e-01 -1.62073046e-01
3.76922153e-02 -4.76793796e-01 -3.08701962e-01 1.45377493e+00
2.39197165e-01 -5.23969643e-02 9.23341036e-01 -1.21077168e+00
-2.73086578e-01 5.95081568e-01 -6.47712290e-01 4.50078934e-01
7.56109953e-01 3.13248951e-03 5.80859780e-01 -1.27550519e+00
-3.69196595e-03 3.07951182e-01 -2.31310558e-02 -2.43887708e-01
-9.63573873e-01 -1.29797435e+00 2.11304054e-02 5.77565789e-01
-1.89953220e+00 -6.24728143e-01 1.25720930e+00 -5.62946379e-01
9.14828837e-01 6.65222108e-01 2.74278283e-01 9.23041582e-01
1.32083774e-01 8.46824706e-01 9.38871324e-01 -5.10305762e-01
1.98661894e-01 -2.53157586e-01 7.19423473e-01 4.26357239e-01
7.50600755e-01 -1.03785723e-01 -6.96889609e-02 -2.28507847e-01
6.41437113e-01 2.99727861e-02 -8.38520586e-01 6.71492964e-02
-4.49614793e-01 1.00520337e+00 3.81963074e-01 5.44992864e-01
-5.71101665e-01 -4.98921759e-02 4.82970268e-01 6.03702784e-01
2.03268111e-01 -2.69788772e-01 6.27659917e-01 -4.43862140e-01
-6.39973223e-01 4.62161511e-01 7.34855592e-01 1.32278454e+00
5.20547807e-01 -2.52562732e-01 1.28982806e+00 1.08801806e+00
5.65329373e-01 6.21778309e-01 -2.68124372e-01 -8.58454108e-01
6.40448987e-01 8.59574899e-02 6.72167718e-01 -6.24067008e-01
-3.32360774e-01 3.54894884e-02 -9.86862600e-01 2.00005189e-01
5.14235079e-01 -6.11887991e-01 -4.76991475e-01 9.78706181e-01
-3.23736787e-01 -3.95253628e-01 -1.77389652e-01 7.68545389e-01
2.15296701e-01 1.23550308e+00 7.41389543e-02 -8.03551003e-02
6.93291306e-01 -7.02807605e-02 -1.15126312e+00 5.72141036e-02
-1.13696251e-02 -9.46712077e-01 9.61952150e-01 6.45871878e-01
-9.83976901e-01 -1.81921899e-01 -1.34222639e+00 6.46811545e-01
-5.01705885e-01 -8.75616074e-03 3.94707799e-01 5.81170142e-01
-6.77304983e-01 -1.05053119e-01 -1.12473309e+00 2.08416674e-02
3.53738755e-01 1.43419489e-01 4.39434350e-01 -2.97192216e-01
-1.68031967e+00 1.55383110e-01 3.24028939e-01 2.36990020e-01
9.84311104e-02 -1.71631217e-01 -2.86647052e-01 -4.94666010e-01
4.56299752e-01 2.31668517e-01 1.93611193e+00 -3.22761416e-01
-7.22955167e-01 2.72386819e-01 1.10876389e-01 8.61178115e-02
1.49037257e-01 -3.10439318e-01 -7.04717398e-01 5.59696853e-01
-4.01137352e-01 -4.25483257e-01 3.91552269e-01 -1.07379889e+00
-8.67649972e-01 -2.12220728e-01 6.19966760e-02 1.62940487e-01
-8.93493831e-01 5.24550259e-01 -2.57285754e-03 -1.40217273e-02
-1.08566908e-02 -1.09126389e+00 5.08354902e-02 -1.94400445e-01
-4.64321434e-01 3.04754265e-02 7.86631823e-01 -3.35373670e-01
1.18007541e+00 -1.67914379e+00 -6.68303072e-01 4.43550676e-01
-4.34417725e-01 9.36596617e-02 1.61425948e-01 1.11769652e+00
3.05049658e-01 3.72719944e-01 -2.99418181e-01 2.94808358e-01
-2.01104749e-02 -5.02632678e-01 2.43438289e-01 9.28547263e-01
-5.52184463e-01 1.36616841e-01 -5.47622979e-01 1.03986427e-01
3.16201717e-01 1.02521193e+00 5.78054607e-01 -1.92712009e-01
6.41691208e-01 -2.02230826e-01 -1.15247285e+00 1.00783205e+00
9.66147244e-01 2.98258334e-01 -2.05866992e-01 -3.33941132e-01
-1.05738193e-01 -1.02543525e-01 -1.96901786e+00 1.06322646e+00
-1.04359649e-01 4.21198189e-01 5.79706669e-01 -6.06299818e-01
3.23088437e-01 8.89353514e-01 3.86820674e-01 -5.67861974e-01
4.97975349e-01 5.21367669e-01 1.59860834e-01 -1.42829388e-01
-2.33037323e-01 -3.61642629e-01 -4.98300999e-01 2.97929883e-01
-2.08949625e-01 -2.12989524e-01 1.10360897e+00 3.75548065e-01
9.50172246e-01 2.49304473e-01 9.03187692e-01 -1.79883793e-01
1.21395374e-02 -3.88690345e-02 4.51024532e-01 3.23152572e-01
-2.93427259e-01 -1.63509011e-01 1.93690389e-01 7.30211437e-02
-1.23736882e+00 -9.65494514e-01 -8.54369402e-01 4.01035428e-01
1.82574525e-01 -6.14688993e-01 -6.14162445e-01 -2.97231395e-02
-6.78442866e-02 1.53906131e+00 3.94721895e-01 -8.32924619e-02
-2.50904590e-01 -6.65590644e-01 3.52804571e-01 7.08442152e-01
6.12613186e-02 -1.09266925e+00 -8.26739848e-01 1.28591493e-01
1.26839727e-01 -5.26408553e-01 -1.90314949e-01 4.79881197e-01
-1.20135295e+00 -5.86827517e-01 -4.74360764e-01 1.56681880e-01
1.26316774e+00 2.97239780e-01 7.18676150e-01 -1.45170360e-03
-4.96127009e-01 7.04195559e-01 -7.30810940e-01 -5.47953129e-01
-2.48511165e-01 -4.83309001e-01 3.99014086e-01 -8.29663515e-01
5.95733523e-01 -6.94993436e-01 -7.53717661e-01 -2.08976921e-02
-8.41752410e-01 -3.07617635e-01 1.02151477e+00 7.07792997e-01
2.49427095e-01 4.85802174e-01 2.53396004e-01 -6.02057755e-01
6.82828963e-01 1.20911926e-01 -1.16800320e+00 -1.90426096e-01
-1.30687380e+00 -6.05011761e-01 5.05659759e-01 6.66078134e-03
-1.46068847e+00 -2.62606382e-01 4.36158061e-01 2.12889284e-01
-1.04367936e+00 4.04902428e-01 -5.31443894e-01 3.98336142e-01
3.61210465e-01 4.28058118e-01 -2.14635193e-01 -6.28579199e-01
-2.51437455e-01 5.52251518e-01 2.82682151e-01 -9.78034258e-01
6.99290812e-01 3.08323175e-01 -6.92645192e-01 -1.49852872e+00
-1.18777841e-01 -1.55394599e-01 -1.02741599e-01 -2.74693966e-01
5.16444921e-01 -1.13256633e+00 -1.30967677e-01 2.59760946e-01
-2.60730654e-01 -2.90972292e-01 -1.85329154e-01 9.04071391e-01
-2.92654306e-01 -2.22092330e-01 -7.41699219e-01 -4.41715062e-01
-1.69610187e-01 -4.12899882e-01 -8.33045989e-02 5.78769922e-01
-8.79302204e-01 -4.91356581e-01 -1.58166602e-01 7.18303382e-01
2.79917419e-01 -4.13260944e-02 -3.06348354e-01 -1.15898204e+00
-8.55222940e-01 -9.46097970e-01 1.80878658e-02 5.05305529e-01
4.32539761e-01 6.06758177e-01 -8.43216658e-01 -5.91168344e-01
3.20390880e-01 -3.33811194e-01 3.54978591e-01 8.99111807e-01
3.11305404e-01 -6.99910045e-01 -2.61049122e-01 1.63839668e-01
1.62442100e+00 9.27075624e-01 5.21841049e-01 2.95661807e-01
2.81861335e-01 3.07093957e-03 6.63744688e-01 2.53535479e-01
-3.37416232e-01 5.96508384e-01 4.75245938e-02 2.75502950e-01
4.18420404e-01 3.54528986e-02 -6.26764493e-04 -1.63668454e-01
-5.53632915e-01 -6.57580137e-01 -1.59452021e+00 3.74477357e-01
-9.79563117e-01 -6.38954759e-01 2.45460838e-01 1.90980840e+00
8.40537071e-01 3.75925750e-01 2.31645219e-02 5.20660579e-01
-2.90921610e-03 4.79912490e-01 -1.30054310e-01 -7.45505393e-01
3.62446338e-01 -4.41568047e-02 6.18167162e-01 5.51652253e-01
-1.03060293e+00 2.11159661e-01 9.27254736e-01 6.99198723e-01
-8.39303255e-01 -1.92453042e-01 -8.80986303e-02 -1.75237611e-01
-8.97683129e-02 2.39022315e-01 -6.43390834e-01 2.34689012e-01
1.48778355e+00 -1.21922338e+00 1.00150175e-01 7.28927433e-01
3.09345275e-01 -3.27295303e-01 9.61699262e-02 7.62866020e-01
-3.53087783e-01 -7.26186812e-01 -8.73296559e-01 3.41098011e-01
3.35606277e-01 3.34646970e-01 2.53479809e-01 -3.40323001e-01
1.89205825e-01 -4.03051287e-01 6.50389254e-01 -1.71408872e-03
5.56814432e-01 -8.06278944e-01 1.08960462e+00 -2.42960349e-01
-7.50623107e-01 -5.65577984e-01 -4.21950132e-01 -2.14357674e-01
8.00856948e-01 8.25273931e-01 -6.90640211e-01 5.28829992e-01
3.83653134e-01 3.73764336e-01 -1.67899385e-01 1.24096477e+00
-9.91959929e-01 9.93310928e-01 -4.81261611e-01 -2.16351703e-01
-2.87079424e-01 -3.93986344e-01 5.02943397e-01 1.17275250e+00
2.13305697e-01 8.43697131e-01 6.23358451e-02 1.88848630e-01
-2.11792275e-01 4.14113879e-01 -4.82784629e-01 -1.40840054e-01
1.24521410e+00 7.99422681e-01 -1.04173613e+00 -2.20893726e-01
-1.14098465e+00 5.00128746e-01 -3.76049817e-01 8.09444606e-01
-6.63929939e-01 -6.53913140e-01 2.02119574e-01 3.74089599e-01
4.01742160e-01 2.03085512e-01 -2.72227898e-02 -9.55112219e-01
4.39677596e-01 -5.47729731e-01 5.08862734e-01 -4.03229922e-01
-2.75402486e-01 9.18866932e-01 4.54360187e-01 -1.67723131e+00
-2.22759068e-01 -7.39304602e-01 -5.30268908e-01 1.14569545e+00
-4.60076123e-01 -8.43111992e-01 1.65177494e-01 -2.98839267e-02
-9.24948528e-02 -1.03213573e-02 6.37785971e-01 3.02253753e-01
-3.58124703e-01 9.50631127e-02 5.09639323e-01 -6.66503161e-02
5.38093984e-01 -1.01360881e+00 -1.91594511e-01 6.74468040e-01
-5.50205290e-01 8.24591637e-01 8.68237853e-01 -3.95829111e-01
-1.52573812e+00 -5.26964784e-01 6.51731908e-01 2.82035202e-01
8.90369236e-01 -3.43343288e-01 -1.04596663e+00 8.54620695e-01
4.56080705e-01 -1.45352900e-01 3.77538383e-01 1.13612413e-01
2.78211664e-02 1.67270809e-01 -9.76218164e-01 4.98861194e-01
4.60784256e-01 -3.22852224e-01 -3.00543278e-01 1.46964684e-01
-8.87571275e-02 -6.27200663e-01 -1.44407415e+00 6.29307032e-02
6.80593550e-01 -5.92573762e-01 6.70526683e-01 2.86902599e-02
6.60063326e-02 1.23207070e-01 -4.85517144e-01 -6.74405038e-01
-4.57862347e-01 -1.44722790e-01 1.54536605e-01 1.20253146e+00
8.51592064e-01 -1.38401794e+00 6.42705917e-01 2.96346098e-01
-4.34210002e-01 -9.40317929e-01 -9.68098402e-01 -8.82220089e-01
-1.51975349e-01 -1.67755067e-01 1.46399900e-01 7.79523492e-01
7.98456907e-01 1.08820014e-02 -3.39024007e-01 4.19446379e-01
7.80095637e-01 3.38140786e-01 -9.69907790e-02 -1.37099311e-01
4.33011234e-01 -7.66802192e-01 -3.17697138e-01 -3.38325441e-01
3.50594148e-02 9.00913775e-02 -3.17132503e-01 -1.65493059e+00
-1.71668589e-01 1.83039814e-01 -2.48459995e-01 3.29176396e-01
2.57424653e-01 -4.34825808e-01 2.58620054e-01 -1.82042167e-01
-7.96050429e-01 3.34336460e-01 8.51112366e-01 1.48208234e-02
-1.58930048e-01 7.69947112e-01 -7.13383377e-01 6.77910984e-01
3.16191840e+00 -8.56836915e-01 -4.68063533e-01 -3.44601184e-01
-2.27324262e-01 5.14435768e-02 9.65102240e-02 -8.83337915e-01
2.51225442e-01 -6.78625762e-01 3.09211642e-01 5.49018160e-02
5.09315193e-01 -8.38251710e-01 4.79049683e-01 1.21085596e+00
-2.43927136e-01 -2.09367320e-01 5.63204348e-01 -8.62295702e-02
-2.40408719e-01 -4.53865826e-01 1.04734349e+00 3.08654338e-01
-1.03145397e+00 -3.93602252e-02 -5.10916650e-01 1.34389540e-02
5.85044146e-01 -6.32575512e-01 -4.13145036e-01 1.20221183e-01
-1.44955739e-01 4.32450742e-01 6.58731282e-01 7.09189922e-02
7.07508266e-01 -8.77919436e-01 -8.87721956e-01 -1.94608599e-01
6.60384417e-01 -7.27885962e-01 5.40633440e-01 8.11666727e-01
-3.18703294e-01 8.57726336e-01 -3.66210580e-01 2.90210601e-02
-1.36311531e+00 3.59084964e-01 -8.09582621e-02 6.06520116e-01
-1.00131643e+00 7.51010180e-01 -5.61197817e-01 3.57567549e-01
1.56762615e-01 -2.14898109e-01 -4.03725989e-02 8.06145370e-02
4.49962825e-01 3.96336854e-01 -5.44282377e-01 -8.02996039e-01
-8.05698216e-01 5.46134710e-01 -3.51276457e-01 1.06963459e-02
1.46261585e+00 -1.20918080e-01 -2.94584278e-02 1.56401340e-02
1.12934256e+00 3.86198342e-01 -8.61152053e-01 5.94842285e-02
2.22118661e-01 -2.97739476e-01 -1.86308816e-01 -4.59890813e-01
-8.51532757e-01 4.66916114e-01 1.10527718e+00 -2.89402932e-01
1.55692518e+00 7.56456852e-02 5.42426825e-01 1.23807257e-02
2.73922324e-01 -1.15339959e+00 -7.63172388e-01 2.34406650e-01
1.09238672e+00 -1.18385947e+00 1.04477870e+00 -4.75850463e-01
-5.20367801e-01 1.08975267e+00 5.44038452e-02 1.79435149e-01
9.14924264e-01 7.64937639e-01 3.31670761e-01 -2.80945092e-01
-3.07123750e-01 3.57236952e-01 8.92145783e-02 3.72082710e-01
4.80112523e-01 4.83792305e-01 -7.02895522e-01 3.38601708e-01
1.16254814e-01 8.35307896e-01 8.52750063e-01 1.47577405e+00
-1.03506967e-01 -1.02934825e+00 -5.46081305e-01 4.40902263e-01
-6.72953844e-01 3.79036218e-02 -3.97510231e-01 1.43106544e+00
-1.45765087e-02 8.92657697e-01 2.99495667e-01 -4.42087464e-02
7.47049749e-01 5.73288381e-01 7.31186271e-01 -5.40904284e-01
3.68503481e-01 7.16078639e-01 6.55866802e-01 -7.11810529e-01
-7.29970858e-02 -1.29796147e+00 -1.73024523e+00 -4.98861700e-01
8.33401158e-02 5.94135702e-01 6.30989969e-01 -8.90296221e-01
1.40033156e-01 5.68294644e-01 4.09986615e-01 -2.15595424e-01
-2.84747392e-01 -1.03248787e+00 -1.20038426e+00 3.22784871e-01
2.40293398e-01 -7.75891840e-01 -6.17447793e-01 3.76100093e-01] | [6.396810531616211, 2.775449275970459] |
3865fa1e-38bb-4dc4-8e23-bcc128259b7e | proceedings-first-workshop-on-causal | 1608.07398 | null | http://arxiv.org/abs/1608.07398v1 | http://arxiv.org/pdf/1608.07398v1.pdf | Proceedings First Workshop on Causal Reasoning for Embedded and safety-critical Systems Technologies | Formal approaches for automated causality analysis, fault localization,
explanation of events, accountability and blaming have been proposed
independently by several communities --- in particular, AI, concurrency,
model-based diagnosis, formal methods. Work on these topics has significantly
gained speed during the last years. The goals of CREST are to bring together
and foster exchange between researchers from the different communities, and to
present and discuss recent advances and new ideas in the field.
The workshop program consisted of a set of invited and contributed
presentations that illustrate different techniques for, and applications of,
causality analysis and fault localization.
The program was anchored by two keynote talks. The keynote by Hana Chockler
(King's College) provided a broad perspective on the application of causal
reasoning based on Halpern and Pearl's definitions of actual causality to a
variety of application domains ranging from formal verification to legal
reasoning. The keynote by Chao Wang (Virginia Tech) concentrated on
constraint-based analysis techniques for debugging and verifying concurrent
programs.
Workshop papers deal with compositional causality analysis and a wide
spectrum of application for causal reasoning, such as debugging of
probabilistic models, accountability and responsibility, hazard analysis in
practice based on Lewis' counterfactuals, and fault localization and repair. | ['Gregor Gössler', 'Oleg Sokolsky'] | 2016-08-26 | null | null | null | null | ['fault-localization'] | ['computer-code'] | [-6.86806813e-02 3.56170624e-01 -2.35054776e-01 -2.80337423e-01
-5.09461582e-01 -5.13382256e-01 7.22599685e-01 5.49454391e-01
2.91344553e-01 7.91346490e-01 2.21177295e-01 -7.93887198e-01
-5.22199214e-01 -4.52449262e-01 -4.46610630e-01 -1.51232436e-01
-8.59758556e-01 5.67676425e-01 7.24516273e-01 -6.60872087e-02
5.22948980e-01 5.05216002e-01 -1.58865440e+00 1.46664768e-01
4.68927622e-01 3.37898612e-01 -5.60392022e-01 8.91201198e-01
1.09408110e-01 1.76755393e+00 -6.24387980e-01 -1.20324381e-01
-5.58634400e-01 -5.23036182e-01 -1.15692794e+00 -4.59047526e-01
-3.69853765e-01 -1.62414506e-01 4.26358636e-03 8.90907526e-01
-4.01525348e-02 -3.67991090e-01 1.78303868e-01 -2.49843287e+00
-1.32077545e-01 1.27923489e+00 -6.85015917e-01 1.27103463e-01
8.64656806e-01 -1.40779376e-01 1.00888288e+00 -3.71950775e-01
4.64802325e-01 1.47774434e+00 8.11080396e-01 5.81722677e-01
-1.19237006e+00 -6.00352287e-01 3.37859064e-01 5.27374506e-01
-1.34859967e+00 -1.52746826e-01 5.01191556e-01 -6.04927242e-01
1.29778790e+00 6.80969298e-01 3.86863023e-01 7.64649689e-01
3.64784211e-01 8.17003548e-01 1.16378593e+00 -8.36920500e-01
3.23252618e-01 -1.35441661e-01 5.21773875e-01 6.79131985e-01
5.44363499e-01 4.38065141e-01 -5.06991804e-01 -1.05703449e+00
5.31889200e-01 -1.17759779e-01 -8.05059522e-02 -4.29358423e-01
-1.39902067e+00 7.90564537e-01 -3.96863520e-01 4.74897772e-01
6.52725920e-02 7.68799603e-01 6.97988331e-01 3.07764590e-01
-1.43212900e-01 -4.12132964e-02 -4.07751560e-01 -2.85016090e-01
-5.07277429e-01 8.13373446e-01 1.28875065e+00 6.74971402e-01
-7.00962022e-02 1.98927417e-01 5.83120063e-02 1.26698008e-02
1.10150695e+00 3.22055012e-01 -2.63273209e-01 -1.36353433e+00
-6.75978959e-02 4.65815127e-01 4.55479890e-01 -7.39957333e-01
-3.08790743e-01 3.17884535e-01 -4.06638205e-01 7.94812799e-01
2.89191812e-01 -1.98012516e-01 -4.03432786e-01 1.64496958e+00
2.05501005e-01 6.76238239e-02 -1.76765889e-01 6.12424970e-01
1.07847862e-01 4.93926227e-01 3.89602333e-02 -5.46927154e-01
1.08542943e+00 -4.17885572e-01 -8.12849522e-01 3.15377891e-01
7.64988601e-01 -5.49202919e-01 2.35938162e-01 8.49263847e-01
-1.26973963e+00 3.71596396e-01 -1.02436757e+00 6.58562720e-01
2.12051898e-01 -7.48238921e-01 8.59369218e-01 6.44856572e-01
-1.17726707e+00 2.88043052e-01 -1.23928678e+00 -4.14152324e-01
-1.14247538e-01 3.45220596e-01 -6.89799562e-02 -1.16629243e-01
-1.30894923e+00 9.63141620e-01 2.76568420e-02 -7.08250552e-02
-1.20819879e+00 -5.02509534e-01 -7.99041390e-01 -2.01202840e-01
4.90705997e-01 -4.11583841e-01 1.59076023e+00 -6.87189400e-01
-1.16962850e+00 5.97909093e-01 -1.02616675e-01 -4.78398323e-01
5.71347594e-01 9.52439159e-02 -9.34578538e-01 -2.90946037e-01
3.12650055e-01 -1.86772466e-01 4.16753888e-02 -1.11913919e+00
-7.93000758e-01 -3.41722757e-01 1.70674115e-01 -5.04318833e-01
6.00264013e-01 1.10069716e+00 3.64367902e-01 -3.33256841e-01
-8.43190476e-02 -8.64138663e-01 -3.80314797e-01 -5.00694692e-01
-4.84974951e-01 -4.40222263e-01 7.58628011e-01 -3.00217807e-01
1.28201604e+00 -1.90707099e+00 7.40105882e-02 4.30710495e-01
1.08337060e-01 -4.26858723e-01 4.52971309e-01 1.18421686e+00
-6.04331613e-01 5.08454323e-01 -2.33537599e-01 3.74520779e-01
5.27643025e-01 1.39045894e-01 -6.62826598e-01 6.69578016e-01
7.43409544e-02 3.80961537e-01 -9.90461230e-01 -5.36946774e-01
1.02407038e-01 -9.22490805e-02 -2.26968184e-01 1.24572344e-01
-3.40204448e-01 1.39269233e-01 -1.73623547e-01 7.91224062e-01
3.50420594e-01 -4.02725302e-02 7.16913521e-01 6.06226444e-01
-5.89106679e-01 2.98262686e-01 -1.57659352e+00 1.17264366e+00
-9.50811207e-02 5.43669224e-01 1.66487306e-01 -6.32712841e-01
5.82447052e-01 1.04886508e+00 3.05844694e-01 -2.97333896e-01
-9.32046026e-02 3.55537981e-01 9.57793146e-02 -4.90104705e-01
1.54678062e-01 -1.75346881e-01 -5.52032173e-01 1.16140807e+00
-4.72970277e-01 -7.21302852e-02 3.17473531e-01 4.14769679e-01
1.62945104e+00 2.80564964e-01 6.45211935e-01 -3.43017399e-01
4.05298650e-01 5.09082675e-01 7.65674174e-01 6.70376360e-01
-3.55140537e-01 -7.31876642e-02 1.28243148e+00 -4.96639460e-01
-8.78793716e-01 -8.70252192e-01 1.17941678e-01 8.16240847e-01
-8.36872384e-02 -8.28075409e-01 -4.55172151e-01 -7.46759832e-01
-1.75480191e-02 9.83885586e-01 -6.62101448e-01 -3.57216373e-02
-7.42625594e-01 -4.06128377e-01 9.45312083e-01 7.52395570e-01
-1.07722871e-01 -8.27813625e-01 -7.90578246e-01 1.97667569e-01
-7.65703470e-02 -5.74043870e-01 2.04587549e-01 2.80634046e-01
-8.85880768e-01 -1.52986407e+00 7.13458955e-02 -3.78821820e-01
3.55563909e-01 -2.30740026e-01 1.03022206e+00 3.99405867e-01
-5.01948297e-01 5.77739000e-01 -2.27939799e-01 -6.47800803e-01
-8.03454995e-01 -8.86747181e-01 -1.87697839e-02 -6.85774446e-01
5.18033765e-02 -4.64037567e-01 2.08315626e-01 6.85293734e-01
-8.38506520e-01 -5.43143570e-01 2.88236476e-02 8.43688965e-01
-9.11410972e-02 2.76860476e-01 2.88699627e-01 -9.61736262e-01
6.24640763e-01 -5.33342898e-01 -9.70614791e-01 4.83008027e-01
-8.72244239e-01 5.38972281e-02 5.12892492e-02 -3.80201191e-02
-1.03636944e+00 -4.20112610e-01 3.28746498e-01 -8.73902813e-02
-1.04609065e-01 7.42678165e-01 -2.76686966e-01 3.72621976e-02
6.97020769e-01 -4.64778304e-01 -1.70603037e-01 2.88770795e-01
1.97305888e-01 9.78598297e-02 3.78799319e-01 -1.11554050e+00
5.57145715e-01 3.72919202e-01 1.84167638e-01 -1.79918520e-02
1.54650211e-01 -3.54152657e-02 -4.66431640e-02 -2.57351726e-01
3.72060120e-01 -5.61797976e-01 -1.25142312e+00 4.96465385e-01
-1.45589209e+00 -6.04795635e-01 -2.16545671e-01 4.00391817e-01
-4.21522349e-01 3.42439383e-01 -7.80014575e-01 -1.49786663e+00
1.99077621e-01 -9.79316175e-01 6.30885422e-01 -1.31644920e-01
-1.04007840e+00 -1.21850634e+00 6.54947042e-01 -7.07835928e-02
8.81948601e-03 7.00992882e-01 1.18624735e+00 -5.55108249e-01
-5.65870464e-01 -5.56039572e-01 4.02287953e-02 -2.31377944e-01
-3.98663223e-01 1.00292969e+00 -6.89157605e-01 5.77124022e-02
-1.78723693e-01 -2.96783745e-01 -1.73825815e-01 2.47383133e-01
5.62535882e-01 -3.04830253e-01 -6.37081444e-01 -5.43506205e-01
1.32163119e+00 5.39221108e-01 7.59549201e-01 4.54304814e-01
4.00556207e-01 7.34868586e-01 7.28612006e-01 6.59151375e-01
4.56679523e-01 5.69215536e-01 6.21056318e-01 4.82689232e-01
3.41022551e-01 1.13891318e-01 5.03808856e-01 3.61061990e-01
-4.78785485e-01 2.11555827e-02 -1.54927492e+00 7.44574785e-01
-2.41084170e+00 -1.36081576e+00 -1.02951062e+00 2.07126188e+00
7.41531014e-01 2.36101627e-01 2.94957459e-01 6.34756804e-01
1.06775081e+00 -1.22399449e-01 1.36803523e-01 -7.87431180e-01
3.46999109e-01 -7.81555027e-02 4.23163623e-01 8.07835460e-01
-5.18904626e-01 2.78783470e-01 7.53216934e+00 2.14132920e-01
-6.14992917e-01 3.07184428e-01 6.33539930e-02 3.37103188e-01
-8.53295505e-01 7.90869296e-01 -3.33607316e-01 1.14010880e-02
1.48670721e+00 -6.36544108e-01 2.17783719e-01 7.61940181e-01
2.46807292e-01 -4.45518225e-01 -1.31716621e+00 3.20177883e-01
-1.00539513e-01 -1.29072762e+00 -6.93914950e-01 -2.61137217e-01
8.53650987e-01 -1.63700029e-01 -5.89006841e-01 6.48532137e-02
1.27623451e+00 -9.84725475e-01 1.28774786e+00 4.41990674e-01
2.01073214e-01 -9.30727899e-01 8.47810268e-01 1.49600878e-01
-8.68731141e-01 -3.57932061e-01 5.09425819e-01 -6.60360277e-01
2.93491125e-01 8.65940273e-01 -6.63691223e-01 7.68217921e-01
7.64940441e-01 1.35273740e-01 2.13116065e-01 1.15435076e+00
-6.25921071e-01 6.37247801e-01 -3.75829190e-02 -1.07590124e-01
-2.43380427e-01 5.42911053e-01 5.40996790e-01 1.14578331e+00
-1.95223778e-01 1.83694363e-01 1.57726277e-02 8.04193556e-01
7.67624915e-01 -8.29323590e-01 -3.93631190e-01 7.86279216e-02
6.78230405e-01 7.07915485e-01 -8.51755023e-01 -2.04632327e-01
-2.07632005e-01 1.30928457e-01 -4.49610651e-01 2.16042325e-01
-1.22376251e+00 -3.86067450e-01 2.60964394e-01 2.58853108e-01
-2.74400711e-01 -3.82613719e-01 -5.02290189e-01 -6.22632563e-01
-8.88496190e-02 -1.11121738e+00 7.81508923e-01 -8.64853799e-01
-9.02299702e-01 1.30897880e-01 7.88016915e-01 -7.59164691e-01
-6.31912649e-01 -1.26708895e-01 -1.03294384e+00 9.83903050e-01
-8.28884900e-01 -9.72691596e-01 9.36445780e-03 4.95764136e-01
-2.17566773e-01 3.31008017e-01 9.56289828e-01 1.80219725e-01
-6.09951496e-01 5.14291003e-02 -5.43654919e-01 -1.06159389e-01
7.90752649e-01 -1.34639430e+00 3.48512441e-01 1.15570474e+00
-3.09139013e-01 7.90941238e-01 1.17887592e+00 -9.03782248e-01
-1.53587770e+00 -7.20582008e-01 1.34683824e+00 -5.93228936e-01
1.36101604e+00 -9.12254006e-02 -6.35892391e-01 1.26932466e+00
4.11081851e-01 -6.91921338e-02 3.20216000e-01 2.15189815e-01
-4.61509794e-01 -1.59525931e-01 -1.01646435e+00 4.96438801e-01
6.45610034e-01 -7.18256831e-01 -6.02456629e-01 3.81780893e-01
6.19803548e-01 -3.28307450e-01 -6.80772185e-01 1.96727335e-01
4.93258208e-01 -1.20497072e+00 2.98886716e-01 -4.70965743e-01
2.83305109e-01 -8.00398171e-01 -4.12285179e-02 -5.73932409e-01
-2.48502254e-01 -1.06546223e+00 4.76549007e-02 1.42446351e+00
2.47714087e-01 -8.14275503e-01 2.80892402e-01 9.32910204e-01
-3.99442881e-01 -6.21778071e-02 -8.09976757e-01 -6.18807971e-01
-3.31956208e-01 -8.73443007e-01 4.06227797e-01 1.28756583e+00
1.00913608e+00 -1.37883902e-01 -5.78766540e-02 3.48864377e-01
7.66630292e-01 2.59734318e-02 5.79058647e-01 -1.46884620e+00
-3.17029029e-01 -4.37759101e-01 -6.62229300e-01 3.32403421e-01
1.48043290e-01 -4.23276871e-01 2.07599372e-01 -1.66990328e+00
2.15929240e-01 -4.06008661e-01 2.07417943e-02 1.01768208e+00
2.67193824e-01 -3.30046386e-01 -2.41449788e-01 3.32196921e-01
-7.28726864e-01 -3.51384044e-01 4.34438705e-01 2.97446968e-03
1.05432607e-01 1.37654155e-01 -5.53428471e-01 6.09752834e-01
4.70647693e-01 -9.35172200e-01 -4.25786257e-01 -1.26492158e-01
9.91657376e-01 9.37516630e-01 1.05064154e+00 -7.45150626e-01
5.66671610e-01 -7.36708105e-01 -4.40689951e-01 -2.94485211e-01
-5.04084468e-01 -8.88127744e-01 9.68292475e-01 9.76914585e-01
-3.34951311e-01 5.19516945e-01 3.23509753e-01 5.72416365e-01
-4.40603018e-01 -3.22335869e-01 2.34884337e-01 -7.28235394e-02
-7.57392406e-01 -3.57974440e-01 -7.35568762e-01 -1.49552315e-01
1.60904205e+00 1.89424053e-01 -6.54738009e-01 -3.21974814e-01
-7.23834515e-01 3.51145983e-01 4.40162718e-01 1.93456933e-01
3.40239227e-01 -1.08519304e+00 -6.36441946e-01 -3.84244800e-01
-5.42134531e-02 -2.57267803e-01 2.07420379e-01 1.44427776e+00
-4.81090367e-01 5.71429610e-01 -2.26550475e-01 -5.83653033e-01
-1.29420662e+00 3.97895247e-01 1.05956025e-01 -3.43254298e-01
-2.22865954e-01 6.89334333e-01 -2.67313391e-01 1.15576840e-03
1.73034936e-01 -4.11100864e-01 2.95307249e-01 -3.16273838e-01
8.31834733e-01 7.48339653e-01 1.11278504e-01 1.47527263e-01
-1.18012965e+00 5.48004024e-02 1.98957458e-01 -7.26620257e-01
1.22271800e+00 1.07281491e-01 -9.96549428e-01 1.21985900e+00
2.44171828e-01 1.37424752e-01 -7.74462044e-01 3.83723348e-01
5.80707312e-01 -3.78589630e-02 -3.56949687e-01 -1.06472576e+00
-3.62420917e-01 5.45525372e-01 2.18044788e-01 7.42039442e-01
7.00777531e-01 1.26540124e-01 -1.34773016e-01 -1.38650388e-01
8.38220477e-01 -5.94410062e-01 -2.59893984e-01 4.34152424e-01
9.27755475e-01 -6.30816817e-01 9.04693007e-02 -2.94490963e-01
-4.48605418e-01 1.29617345e+00 2.96044558e-01 -5.65591715e-02
4.41960037e-01 9.92814362e-01 -1.73586935e-01 -3.54262948e-01
-1.11769283e+00 4.19428587e-01 -4.96251643e-01 7.90364742e-01
5.28429329e-01 3.27338815e-01 -2.92066425e-01 7.41987050e-01
4.24301863e-01 4.69537526e-01 1.17617989e+00 1.60465002e+00
-1.82200134e-01 -1.43208897e+00 -1.15926182e+00 -1.83697164e-01
-5.09789169e-01 7.95586333e-02 -5.69017351e-01 1.36490774e+00
-1.39043435e-01 1.25302696e+00 -1.56665176e-01 -3.11577052e-01
3.68129998e-01 2.47558728e-02 6.10709190e-01 -5.32796085e-01
-7.33635366e-01 -1.03957288e-01 7.63646185e-01 -6.40210569e-01
-7.05118954e-01 -1.06394720e+00 -1.67435229e+00 -8.21905315e-01
-2.62433082e-01 7.70388246e-01 5.31942308e-01 9.92054939e-01
-2.77351558e-01 7.87303090e-01 3.70206863e-01 -2.39699423e-01
-4.15635794e-01 -5.20522177e-01 -8.38004291e-01 -4.77723300e-01
7.69736245e-02 -5.49015343e-01 -5.25579870e-01 9.74067226e-02] | [8.319266319274902, 6.107024192810059] |
f9b132c1-8c91-4001-b3ad-20dacf6eff72 | fast-and-interpretable-nonlocal-neural | 2306.01950 | null | https://arxiv.org/abs/2306.01950v1 | https://arxiv.org/pdf/2306.01950v1.pdf | Fast and Interpretable Nonlocal Neural Networks for Image Denoising via Group-Sparse Convolutional Dictionary Learning | Nonlocal self-similarity within natural images has become an increasingly popular prior in deep-learning models. Despite their successful image restoration performance, such models remain largely uninterpretable due to their black-box construction. Our previous studies have shown that interpretable construction of a fully convolutional denoiser (CDLNet), with performance on par with state-of-the-art black-box counterparts, is achievable by unrolling a dictionary learning algorithm. In this manuscript, we seek an interpretable construction of a convolutional network with a nonlocal self-similarity prior that performs on par with black-box nonlocal models. We show that such an architecture can be effectively achieved by upgrading the $\ell 1$ sparsity prior of CDLNet to a weighted group-sparsity prior. From this formulation, we propose a novel sliding-window nonlocal operation, enabled by sparse array arithmetic. In addition to competitive performance with black-box nonlocal DNNs, we demonstrate the proposed sliding-window sparse attention enables inference speeds greater than an order of magnitude faster than its competitors. | ['Yao Wang', 'Adeen Flinker', 'Amirhossein Khalilian-Gourtani', 'Nikola Janjušević'] | 2023-06-02 | null | null | null | null | ['image-restoration', 'grayscale-image-denoising', 'dictionary-learning'] | ['computer-vision', 'computer-vision', 'methodology'] | [ 5.08565187e-01 3.96110922e-01 -7.83639774e-03 -5.57478428e-01
-7.22527921e-01 -2.93080956e-01 7.17093468e-01 -1.71443716e-01
-2.35543162e-01 5.68178892e-01 4.90365654e-01 -3.69757235e-01
-2.01676011e-01 -6.57442987e-01 -1.24971783e+00 -8.61606240e-01
-8.24899226e-03 1.23724878e-01 -2.31804177e-01 -2.25869775e-01
6.06939569e-02 4.12377089e-01 -1.57790208e+00 3.91504258e-01
7.88049579e-01 1.15739548e+00 3.33716989e-01 5.57886958e-01
-3.22710313e-02 9.89301980e-01 -3.95825922e-01 -2.25394547e-01
5.07862270e-01 -4.19736058e-01 -4.19612914e-01 1.26246288e-01
1.13889956e+00 -5.02038360e-01 -5.42604029e-01 1.22319388e+00
2.43954599e-01 1.15519792e-01 2.53973007e-01 -8.79094660e-01
-1.14530194e+00 7.30939329e-01 -3.56419951e-01 1.70835212e-01
-1.99538276e-01 1.45089939e-01 1.25705314e+00 -1.09495413e+00
5.62534213e-01 1.45135164e+00 9.29034352e-01 3.59296799e-01
-1.58957517e+00 -5.18146336e-01 2.60415316e-01 3.57428640e-01
-1.45531893e+00 -7.18154848e-01 7.88521290e-01 -8.75698775e-02
1.10045731e+00 3.02991420e-01 5.06494224e-01 9.47562099e-01
2.85507262e-01 5.67170560e-01 1.02422428e+00 -5.54977059e-01
4.12003100e-01 -4.31378305e-01 -1.21093139e-01 8.98973286e-01
3.79307151e-01 1.54352099e-01 -8.37005079e-01 1.83207124e-01
1.17551887e+00 1.50562704e-01 -3.09317797e-01 -3.63655329e-01
-1.21309245e+00 7.23239541e-01 8.62915814e-01 2.78540790e-01
-4.86914724e-01 8.83177221e-01 1.40126199e-01 2.04173952e-01
5.19175231e-01 3.78952801e-01 -1.73889160e-01 2.15542972e-01
-1.31244063e+00 -3.15716378e-02 5.77922165e-01 8.40117157e-01
1.00938988e+00 6.35102689e-01 3.87947820e-02 4.59812641e-01
1.41802058e-01 3.43539417e-01 3.62014592e-01 -1.26481092e+00
-1.16212619e-02 4.15743768e-01 -1.88332543e-01 -1.14889133e+00
-2.14951530e-01 -1.00020540e+00 -1.55355346e+00 4.96796280e-01
2.05022797e-01 1.62033156e-01 -1.06735754e+00 1.93503821e+00
7.92649016e-03 4.57407624e-01 4.69661504e-02 8.61936748e-01
6.39362574e-01 7.04533100e-01 -5.10481885e-03 -1.02774508e-01
1.41000915e+00 -1.10353446e+00 -8.68042946e-01 -3.14395189e-01
2.60339737e-01 -5.10910273e-01 9.89219427e-01 7.53747225e-01
-1.11431563e+00 -8.32568526e-01 -1.34010291e+00 -7.18977988e-01
8.25013369e-02 6.25887588e-02 7.30341434e-01 4.33479100e-01
-1.50275743e+00 7.31313646e-01 -1.12244880e+00 -3.10794741e-01
6.63503289e-01 2.85218328e-01 -5.53222656e-01 -7.33883977e-02
-7.60342121e-01 9.47065771e-01 3.30908328e-01 4.92790341e-01
-1.14848566e+00 -8.86733174e-01 -9.52358365e-01 3.64208728e-01
3.23026478e-01 -9.70279515e-01 9.03233588e-01 -1.10728693e+00
-1.40891922e+00 7.70235360e-01 -4.91305083e-01 -9.99563158e-01
1.46342337e-01 -4.64329213e-01 -2.77074516e-01 2.67927408e-01
-8.62158090e-02 7.09068954e-01 1.36826909e+00 -1.33255088e+00
-2.03873977e-01 -3.01293224e-01 2.15837806e-01 -6.24632537e-02
-3.86719525e-01 -3.17522585e-01 -2.07147121e-01 -1.19236386e+00
3.67298871e-01 -6.87468231e-01 -4.37732726e-01 6.38839364e-01
-3.66054654e-01 1.36383101e-01 8.42963159e-01 -6.17740154e-01
1.28664124e+00 -2.12261868e+00 3.20960969e-01 1.09749541e-01
7.25488544e-01 2.11064249e-01 -2.72303462e-01 2.31890485e-01
-3.96567017e-01 1.96398944e-02 -5.24612486e-01 -6.44788146e-01
7.48081803e-02 6.17448032e-01 -5.28977215e-01 6.23330474e-01
3.75536978e-01 1.00343943e+00 -7.07254231e-01 -7.54818320e-02
2.61087239e-01 7.42106616e-01 -8.11432838e-01 -1.45475268e-01
-1.55631363e-01 2.84528852e-01 -2.67932490e-02 5.29598653e-01
7.16874063e-01 -4.65874583e-01 3.06933254e-01 -7.26963043e-01
-7.02888668e-02 6.87293783e-02 -1.18081284e+00 2.06876493e+00
-6.46172225e-01 8.98221254e-01 4.84298557e-01 -1.34436786e+00
9.99957383e-01 1.68814570e-01 4.19290401e-02 -6.99814200e-01
-9.94149595e-02 2.18346804e-01 -8.46376196e-02 -7.74771869e-02
5.49765825e-01 -3.52441519e-01 2.40699798e-01 2.20933944e-01
2.72940218e-01 1.41724378e-01 -7.40767270e-02 2.38116190e-01
1.33650994e+00 1.04715213e-01 4.27761585e-01 -7.59366691e-01
4.77510363e-01 -1.96588546e-01 5.48102975e-01 9.56837058e-01
1.12143070e-01 7.41667747e-01 3.78829330e-01 -7.26919234e-01
-1.11809731e+00 -1.08115196e+00 -2.18188182e-01 1.07342291e+00
-3.60015444e-02 -5.30346453e-01 -8.13618004e-01 5.25621092e-03
-2.15950087e-01 5.86792350e-01 -7.03437448e-01 -3.14623825e-02
-8.70339215e-01 -5.92908859e-01 6.90845072e-01 6.72891438e-01
7.74807155e-01 -7.04881907e-01 -5.26812673e-01 2.48028383e-01
1.16374061e-01 -1.14142907e+00 -2.67642826e-01 5.63827991e-01
-9.88601983e-01 -5.95122159e-01 -4.89238709e-01 -8.74454141e-01
9.67366517e-01 2.19786227e-01 1.28101230e+00 1.32238522e-01
-7.61588290e-02 2.81906221e-02 -7.75162280e-02 4.75442894e-02
-2.80916661e-01 -1.22313701e-01 2.96470691e-02 1.56399250e-01
-8.88047833e-03 -1.15579176e+00 -8.74036193e-01 -1.08135931e-01
-1.28320420e+00 3.62122566e-01 8.86992693e-01 1.06764615e+00
8.84426475e-01 1.06124565e-01 5.56606293e-01 -8.97758365e-01
2.06188634e-01 -3.45735550e-01 -4.72408235e-01 2.43226215e-02
-4.73369569e-01 5.55279851e-01 8.70646179e-01 -1.83850765e-01
-1.08263576e+00 9.72795710e-02 -2.26361945e-01 -6.20351136e-01
-3.04822791e-02 6.68925822e-01 -2.98906118e-01 -1.60906956e-01
8.42584193e-01 1.56689197e-01 6.33688048e-02 -6.72813773e-01
7.38296926e-01 2.49668971e-01 1.03300762e+00 -6.82338238e-01
9.64184225e-01 8.17749977e-01 2.80635178e-01 -8.62578392e-01
-9.03966963e-01 -1.27056107e-01 -4.24687445e-01 2.63691694e-01
9.53982115e-01 -1.33159137e+00 -5.20084560e-01 5.08782983e-01
-1.19611859e+00 -3.89196873e-01 -4.87211883e-01 -2.02831104e-02
-4.49752122e-01 3.42482269e-01 -4.97732699e-01 -4.52984184e-01
-3.69957507e-01 -9.71185803e-01 1.05437613e+00 -1.56480014e-01
-2.24505767e-01 -1.09846389e+00 -4.42434967e-01 2.88028061e-01
7.03437984e-01 3.02738667e-01 1.01989079e+00 -3.73646259e-01
-9.71833110e-01 6.93674460e-02 -3.87084723e-01 6.93900764e-01
-1.10857524e-01 -4.55960006e-01 -1.19010556e+00 -3.09874505e-01
5.03874958e-01 -2.42842678e-02 1.22568142e+00 5.70986509e-01
1.48531115e+00 -5.82285881e-01 8.13754424e-02 1.13733232e+00
1.68110609e+00 -2.53647357e-01 8.48307133e-01 2.03424394e-01
9.19232368e-01 8.82141218e-02 -1.19648553e-01 4.83143836e-01
3.12446326e-01 4.24925506e-01 8.40055585e-01 -2.90965170e-01
-5.84253192e-01 -1.93213597e-01 3.09111446e-01 8.67639840e-01
-1.08394630e-01 -2.36300811e-01 -7.58574724e-01 6.38186872e-01
-1.84764719e+00 -9.13472056e-01 -7.54348785e-02 1.85012484e+00
8.72583687e-01 9.10089538e-02 -6.16469860e-01 9.57815498e-02
4.88498360e-01 6.03410542e-01 -6.19614959e-01 -3.83466035e-01
-5.87656796e-01 8.20110381e-01 5.80163002e-01 7.28576362e-01
-1.07279599e+00 7.17234850e-01 5.98081207e+00 1.01987433e+00
-1.02098823e+00 3.67842346e-01 7.91545212e-01 5.40667167e-03
-6.47683620e-01 9.32897255e-02 -6.37276590e-01 5.65702915e-02
9.06814039e-01 6.44666106e-02 6.96080685e-01 6.95388734e-01
2.28923291e-01 6.38790503e-02 -1.49311817e+00 1.13337886e+00
2.24861249e-01 -1.78555644e+00 3.48681659e-01 -3.69519852e-02
9.81288731e-01 -7.76006654e-02 1.59307882e-01 9.49585810e-02
2.61643678e-01 -1.36395538e+00 9.37922657e-01 4.85394418e-01
7.83614874e-01 -4.48369026e-01 4.29260463e-01 2.90753782e-01
-1.09703004e+00 -9.54235196e-02 -3.65838349e-01 -3.73045206e-01
8.59276950e-02 8.49184036e-01 -2.27777004e-01 3.51372004e-01
6.81316495e-01 9.14646208e-01 -3.79730910e-01 6.46217823e-01
-3.52853924e-01 6.34287775e-01 -3.73053700e-01 5.11878014e-01
2.49872565e-01 -6.59754649e-02 5.35249352e-01 1.05857873e+00
5.61969221e-01 1.72894374e-01 -2.08549723e-01 1.19853175e+00
-4.35604244e-01 -4.93235946e-01 -3.46047610e-01 1.81985557e-01
3.89850348e-01 1.09101009e+00 -4.10804689e-01 -5.60475111e-01
-3.07766467e-01 1.11028838e+00 3.22893322e-01 4.57134187e-01
-7.54074574e-01 -5.91597520e-02 9.02049601e-01 1.50965169e-01
7.26681948e-01 -4.34196085e-01 -8.21736395e-01 -1.29622900e+00
2.88578197e-02 -1.03072870e+00 -7.11726770e-02 -8.77253592e-01
-1.28878641e+00 6.22106075e-01 -1.80745095e-01 -1.08259022e+00
-1.53009901e-02 -6.87202513e-01 -4.62869138e-01 8.23084176e-01
-1.70871198e+00 -1.21381509e+00 -3.77622068e-01 6.39995158e-01
4.83618468e-01 4.10427116e-02 1.01000214e+00 3.72925282e-01
-3.61456662e-01 5.56200206e-01 3.03448051e-01 -2.11455375e-02
3.42691839e-01 -1.26263511e+00 5.53598046e-01 1.33355606e+00
3.23416233e-01 9.97385323e-01 7.71364689e-01 -3.88560116e-01
-1.75491428e+00 -1.18294132e+00 5.70669949e-01 -1.18194669e-01
7.88169086e-01 -3.94037843e-01 -9.90396976e-01 9.00961399e-01
5.47262490e-01 4.50378358e-01 3.38362813e-01 5.30629233e-02
-6.04549885e-01 -5.30821264e-01 -9.43398833e-01 6.70856357e-01
1.32795477e+00 -7.21255243e-01 -6.65211380e-01 2.10022107e-01
6.84283316e-01 -4.79529589e-01 -7.30079830e-01 3.06680202e-01
4.67304468e-01 -1.01371920e+00 1.23580599e+00 -2.55554080e-01
7.22015381e-01 -4.97983605e-01 -4.67064381e-01 -1.14317763e+00
-4.42916036e-01 -9.59852815e-01 -3.33895683e-01 7.68654644e-01
-1.30336031e-01 -4.37153786e-01 8.10529828e-01 4.53605890e-01
-6.91839635e-01 -5.80952108e-01 -1.22857809e+00 -6.49197817e-01
2.16653645e-02 -5.73352456e-01 2.96215117e-01 6.92205906e-01
-5.03506780e-01 2.63107419e-01 -6.00553632e-01 4.11618650e-01
9.38459814e-01 -4.66601690e-03 5.12103498e-01 -1.04067230e+00
-4.31637198e-01 -4.47560459e-01 -5.09404480e-01 -1.72343099e+00
2.37187147e-01 -9.33059633e-01 1.38713658e-01 -1.43756711e+00
-2.15289257e-02 -3.05649489e-01 -3.95207018e-01 6.69833541e-01
1.41527101e-01 7.23366916e-01 1.48335725e-01 4.93675545e-02
-3.60226750e-01 6.76007807e-01 1.26666927e+00 -1.76109031e-01
2.73671985e-01 -6.78494930e-01 -1.00667942e+00 8.98143351e-01
5.17544448e-01 -2.63711274e-01 -3.92519683e-01 -1.05599892e+00
4.33172911e-01 -1.50875285e-01 9.19870913e-01 -1.22191525e+00
4.23343688e-01 2.91463345e-01 2.20418677e-01 -3.34343284e-01
4.98551309e-01 -8.84533763e-01 3.99880260e-01 3.09194893e-01
-3.79803270e-01 -5.92672080e-02 3.01217973e-01 7.67348886e-01
-4.17146236e-01 -1.06006972e-01 6.75691903e-01 -2.22976461e-01
-8.77200067e-01 1.97324082e-02 -2.72093266e-01 -1.38256595e-01
3.96279395e-01 -4.42025959e-01 -4.69965994e-01 -2.63086319e-01
-8.72271538e-01 -3.60263437e-01 2.82587081e-01 3.50634567e-02
7.28261888e-01 -1.33247578e+00 -6.76627994e-01 4.26587880e-01
-3.74581397e-01 2.96496868e-01 4.20213103e-01 7.00859427e-01
-8.72189760e-01 2.92565823e-01 -3.29868644e-01 -6.15285039e-01
-8.52713227e-01 3.33065897e-01 2.65617132e-01 -1.83222935e-01
-7.40523517e-01 1.06980920e+00 5.90741456e-01 -1.56888053e-01
2.07193807e-01 -7.79891312e-01 4.78011608e-01 -2.91003406e-01
4.82617944e-01 1.62742749e-01 2.09250778e-01 -5.02501547e-01
-2.12300658e-01 4.53976363e-01 2.26029813e-01 2.48062685e-01
1.60894024e+00 -1.49044767e-01 -5.35907149e-01 1.49323121e-01
1.08345807e+00 -8.09276849e-02 -1.45747137e+00 -4.61764902e-01
-3.46981436e-01 -3.70925695e-01 5.96041262e-01 -5.54756105e-01
-1.24547565e+00 9.61497962e-01 5.43157876e-01 -3.72391529e-02
1.38548625e+00 -1.82990402e-01 9.38459098e-01 5.65807939e-01
2.23358050e-01 -6.17017746e-01 8.93676654e-02 4.22534555e-01
1.25325191e+00 -1.15291131e+00 2.07399085e-01 -2.83940554e-01
-4.62245755e-02 1.08232319e+00 2.42689371e-01 -5.23655295e-01
8.10375810e-01 3.83034080e-01 -1.84001192e-01 -1.84161544e-01
-7.79611826e-01 2.62735337e-02 2.72414505e-01 3.80807877e-01
1.67944714e-01 -8.96538496e-02 8.98512602e-02 4.68253702e-01
-2.60063767e-01 -9.74974334e-02 4.43239212e-01 6.74186587e-01
-4.67983127e-01 -9.54261243e-01 -3.63205284e-01 2.20606446e-01
-4.64740425e-01 -6.79840505e-01 3.17890525e-01 6.17971480e-01
3.73805434e-01 8.05612803e-01 3.67161661e-01 -1.48434296e-01
-2.41812714e-03 -2.51526564e-01 3.40419650e-01 -5.56144416e-01
-4.78656679e-01 8.95256847e-02 1.21758934e-02 -6.89847052e-01
-7.00233638e-01 -2.89271772e-01 -1.15909445e+00 -5.95400035e-01
3.70218866e-02 -3.72147799e-01 4.85763133e-01 1.07728028e+00
5.67725003e-01 7.25175858e-01 -5.18280342e-02 -1.04263151e+00
-5.54041445e-01 -7.29978681e-01 -4.67417330e-01 2.65088141e-01
6.90343082e-01 -3.71620923e-01 -2.99115300e-01 4.66599911e-01] | [11.395543098449707, -2.096381902694702] |
4a9c1933-5b33-4005-b1a1-7038126d9093 | stochastic-transformer-networks-with-linear | 2109.13318 | null | https://arxiv.org/abs/2109.13318v2 | https://arxiv.org/pdf/2109.13318v2.pdf | Stochastic Transformer Networks with Linear Competing Units: Application to end-to-end SL Translation | Automating sign language translation (SLT) is a challenging real world application. Despite its societal importance, though, research progress in the field remains rather poor. Crucially, existing methods that yield viable performance necessitate the availability of laborious to obtain gloss sequence groundtruth. In this paper, we attenuate this need, by introducing an end-to-end SLT model that does not entail explicit use of glosses; the model only needs text groundtruth. This is in stark contrast to existing end-to-end models that use gloss sequence groundtruth, either in the form of a modality that is recognized at an intermediate model stage, or in the form of a parallel output process, jointly trained with the SLT model. Our approach constitutes a Transformer network with a novel type of layers that combines: (i) local winner-takes-all (LWTA) layers with stochastic winner sampling, instead of conventional ReLU layers, (ii) stochastic weights with posterior distributions estimated via variational inference, and (iii) a weight compression technique at inference time that exploits estimated posterior variance to perform massive, almost lossless compression. We demonstrate that our approach can reach the currently best reported BLEU-4 score on the PHOENIX 2014T benchmark, but without making use of glosses for model training, and with a memory footprint reduced by more than 70%. | ['Sotirios Chatzis', 'Dimitris N. Metaxas', 'Dimitrios Kosmopoulos', 'Konstantinos P. Panousis', 'Andreas Voskou'] | 2021-09-01 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Voskou_Stochastic_Transformer_Networks_With_Linear_Competing_Units_Application_To_End-to-End_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Voskou_Stochastic_Transformer_Networks_With_Linear_Competing_Units_Application_To_End-to-End_ICCV_2021_paper.pdf | iccv-2021-1 | ['sign-language-translation'] | ['computer-vision'] | [ 7.38827169e-01 1.15827799e-01 -1.83225974e-01 -2.25913629e-01
-1.23785186e+00 -4.26687866e-01 7.10147679e-01 -1.45745739e-01
-8.82894933e-01 5.72200954e-01 3.17821175e-01 -4.83411670e-01
1.70321926e-01 -5.01998603e-01 -1.00298059e+00 -7.37819135e-01
4.42773700e-01 8.50863039e-01 3.14644754e-01 -6.24020286e-02
-5.91943460e-03 -1.35024756e-01 -1.56693220e+00 5.02446651e-01
8.80412340e-01 1.17393601e+00 3.34093451e-01 8.12254190e-01
-3.72901894e-02 8.26023459e-01 -5.79083860e-01 -8.82383287e-01
2.38572583e-01 -5.77267110e-01 -5.74792266e-01 -1.35887444e-01
6.82232141e-01 -6.46775663e-01 -1.49603829e-01 8.32215369e-01
7.13972986e-01 -2.02800348e-01 6.57603145e-01 -8.96926999e-01
-5.10377288e-01 7.22619236e-01 -3.80394042e-01 -3.57165128e-01
6.61851242e-02 4.87412333e-01 1.27559364e+00 -7.53085136e-01
8.44299674e-01 1.05951738e+00 6.91533983e-01 7.44741797e-01
-1.21060777e+00 -3.85806382e-01 -2.37529003e-03 3.68005186e-01
-1.18454206e+00 -5.11392653e-01 3.77948344e-01 7.15795755e-02
1.30250537e+00 2.19881028e-01 8.44353437e-01 1.54336536e+00
1.00781322e-01 1.19821060e+00 9.62354541e-01 -6.13206923e-01
2.77483046e-01 -2.67073274e-01 -1.84702918e-01 6.69510901e-01
1.87133297e-01 -5.59493564e-02 -7.15381563e-01 8.62296745e-02
2.11940512e-01 -2.72137672e-01 -1.50718451e-01 -2.48553291e-01
-1.18350661e+00 4.79035139e-01 1.65126324e-01 7.52190426e-02
-5.38582385e-01 6.58228934e-01 6.43551648e-01 3.87610555e-01
2.22773135e-01 -2.54701316e-01 -3.67941678e-01 -4.78798598e-01
-1.45972335e+00 2.95598239e-01 7.70500839e-01 9.34202135e-01
2.66181856e-01 1.04895070e-01 -3.49204481e-01 7.78549790e-01
4.73650694e-01 8.03084850e-01 5.99541008e-01 -7.46380687e-01
7.80012608e-01 1.66500479e-01 -1.82685524e-01 -1.22188307e-01
1.01346597e-01 -4.39765751e-01 -6.09628737e-01 2.98802316e-01
6.20965362e-01 4.32445994e-03 -1.32418120e+00 1.92232394e+00
1.47202447e-01 5.33540659e-02 -3.20953317e-02 8.57548714e-01
2.63571084e-01 4.76053268e-01 2.35162243e-01 6.77916184e-02
1.49688315e+00 -1.09794998e+00 -8.10501993e-01 -2.72921264e-01
6.00522161e-01 -8.94096017e-01 1.27852750e+00 6.72567308e-01
-1.45240045e+00 -5.38634844e-02 -1.10372591e+00 -5.68434000e-01
-2.47357681e-01 1.75987154e-01 3.82771194e-01 5.91576219e-01
-1.17688036e+00 5.13551950e-01 -9.01239693e-01 -3.64779621e-01
5.11384428e-01 3.21057558e-01 -1.71460718e-01 -2.40884513e-01
-9.39614356e-01 1.23573840e+00 2.39485294e-01 3.48150164e-01
-7.40716815e-01 -4.69660938e-01 -7.29864597e-01 -3.85328865e-04
4.29978907e-01 -1.03489423e+00 1.52562213e+00 -8.81135702e-01
-1.99286699e+00 6.54248476e-01 -3.02243471e-01 -8.93906653e-01
1.07792687e+00 -4.04148430e-01 1.73172001e-02 7.68287107e-02
-5.02842247e-01 6.30892634e-01 1.03307986e+00 -8.70862186e-01
-6.54498935e-01 -9.85651240e-02 -2.23947421e-01 1.43843740e-01
-1.96439713e-01 2.50281598e-02 -6.81065321e-01 -6.41421735e-01
-8.28282759e-02 -1.07438099e+00 1.03667937e-02 2.39465773e-01
-3.45923156e-01 -1.47699416e-01 6.16076291e-01 -1.26005805e+00
1.26534951e+00 -1.80379033e+00 3.79076451e-01 2.49878645e-01
-7.24961236e-02 5.72879553e-01 -2.21695632e-01 5.45747817e-01
3.90745878e-01 -1.14418983e-01 -5.98997712e-01 -8.75477970e-01
4.84949082e-01 3.50424945e-01 -4.41697031e-01 1.48505211e-01
-6.38900250e-02 1.26714563e+00 -7.44279623e-01 -6.02898657e-01
3.19636256e-01 6.94569767e-01 -7.01865256e-01 -2.20956042e-01
-6.27442241e-01 2.66901180e-02 -5.01968060e-03 4.98557031e-01
3.38586628e-01 -1.86509602e-02 4.36308712e-01 -1.63161367e-01
8.45284294e-03 7.25419223e-01 -7.93129981e-01 2.20301700e+00
-6.25872433e-01 6.51405752e-01 -8.15159678e-02 -7.30719566e-01
3.20694983e-01 4.52182889e-01 1.83463618e-01 -7.84515619e-01
3.54034871e-01 6.88596249e-01 -2.07800448e-01 -3.68765503e-01
5.24273336e-01 -1.44761220e-01 -3.31197791e-02 6.82024419e-01
2.88963288e-01 -1.72244489e-01 2.78900743e-01 6.68387115e-02
1.14060903e+00 5.55259109e-01 -1.40781254e-02 3.44834983e-01
8.16214904e-02 -1.64742097e-01 2.31531501e-01 5.76756001e-01
-2.59979814e-02 7.08239973e-01 1.85628548e-01 -2.24782705e-01
-1.34717274e+00 -1.17703712e+00 2.52292961e-01 8.84587467e-01
-3.02201480e-01 -4.60889757e-01 -1.09866929e+00 -6.56155169e-01
-2.72158116e-01 1.11086094e+00 -3.61344427e-01 -2.23430961e-01
-8.15891504e-01 -5.39214075e-01 9.28045034e-01 3.69321913e-01
5.53746819e-01 -1.10257137e+00 -9.91715193e-01 2.33543366e-01
-4.77550000e-01 -1.08653390e+00 -7.01204896e-01 2.29942128e-01
-9.86349106e-01 -5.82208753e-01 -8.76683295e-01 -4.90125865e-01
2.73866564e-01 -1.87550142e-01 1.08989179e+00 -1.35954380e-01
-1.96534142e-01 1.01320803e-01 -3.54636967e-01 -4.71093059e-01
-5.23572147e-01 3.38497758e-01 -2.85272598e-01 -1.14367478e-01
4.05413955e-01 -8.04421782e-01 -5.74660420e-01 -1.62539423e-01
-1.18272829e+00 4.32838291e-01 1.12956619e+00 9.13076997e-01
6.28452539e-01 -5.71932554e-01 2.60545313e-01 -5.52250624e-01
4.39558834e-01 -5.54572493e-02 -5.40877759e-01 5.79175293e-01
-7.06963897e-01 3.80212843e-01 4.20139015e-01 -4.12563980e-01
-8.98147225e-01 1.55744739e-02 -5.49193680e-01 -4.00773138e-01
1.65085033e-01 3.78427148e-01 -8.28510970e-02 2.68889517e-01
4.53639984e-01 7.86130190e-01 2.34917209e-01 -5.71938097e-01
5.85981905e-01 8.18729639e-01 4.71948206e-01 -3.41281503e-01
8.39732707e-01 3.63449425e-01 -2.59695113e-01 -4.78555322e-01
-6.86114371e-01 1.19124383e-01 -3.45734179e-01 -2.49209389e-01
7.34355271e-01 -7.00969160e-01 -7.11365819e-01 5.41189671e-01
-9.43774164e-01 -7.33103037e-01 -6.15871727e-01 5.25514364e-01
-7.84283817e-01 3.68640810e-01 -7.52067745e-01 -7.93347239e-01
-8.87775421e-01 -9.42084610e-01 1.19144583e+00 -2.34996065e-01
-2.73490071e-01 -6.08980119e-01 -2.24229842e-01 7.46864736e-01
6.16127014e-01 9.48235765e-03 8.40564489e-01 -2.36133158e-01
-8.11192691e-01 -4.06356156e-01 -5.99956438e-02 7.16505051e-01
-2.20491633e-01 -2.41895184e-01 -1.08779454e+00 -2.38179967e-01
-3.14385414e-01 -5.64185798e-01 1.03426421e+00 2.72475094e-01
8.72147560e-01 -4.75017548e-01 7.91720971e-02 5.07332146e-01
1.34231317e+00 -2.02114522e-01 8.64820004e-01 1.51721746e-01
4.68450785e-01 3.11137259e-01 2.14643463e-01 2.02683747e-01
6.56901240e-01 8.34130526e-01 3.14295381e-01 1.54462308e-01
-8.43224764e-01 -5.41023195e-01 7.35342979e-01 1.15861666e+00
-3.33897024e-01 -3.89079303e-01 -5.38047135e-01 6.20205998e-01
-1.84983599e+00 -1.03195381e+00 2.96599239e-01 2.40795588e+00
1.14332390e+00 2.42633894e-01 4.80752029e-02 3.94472480e-01
7.79577568e-02 1.22682691e-01 -5.29599190e-01 -5.73734760e-01
-1.08209558e-01 4.33068067e-01 6.55094564e-01 5.53377807e-01
-6.22608006e-01 1.03705776e+00 5.66170168e+00 1.10479724e+00
-1.03918540e+00 3.07916611e-01 5.97762428e-02 -5.59463859e-01
-3.34534705e-01 -7.63815492e-02 -5.12845218e-01 6.44500315e-01
1.18945515e+00 3.69820476e-01 8.36741686e-01 5.26983440e-01
9.22945365e-02 -8.78789052e-02 -9.52495277e-01 7.05978572e-01
6.50342479e-02 -1.05399203e+00 3.36416334e-01 2.45440528e-01
2.92493314e-01 3.63524199e-01 2.13900223e-01 2.97553122e-01
4.10866618e-01 -7.63996542e-01 1.35856104e+00 5.71804047e-01
9.42595005e-01 -5.60484111e-01 6.52443290e-01 3.74081254e-01
-9.69478071e-01 8.07417855e-02 -5.14129773e-02 1.95307583e-01
5.02775133e-01 4.87742722e-01 -8.40699613e-01 6.05876923e-01
4.16683614e-01 2.10990176e-01 -1.47221550e-01 9.88929689e-01
-6.85905993e-01 1.05203998e+00 -7.56470799e-01 -1.67319164e-01
3.38084549e-01 3.14639322e-02 5.51247776e-01 1.36349392e+00
4.79148060e-01 -5.37730694e-01 -4.14316148e-01 5.49333453e-01
-1.99806586e-01 5.58564924e-02 -2.48599827e-01 1.41585395e-01
2.61777639e-01 5.60414195e-01 -4.31289345e-01 -4.70480323e-01
-3.18238467e-01 1.55330539e+00 2.06799358e-01 3.60914052e-01
-8.77972066e-01 -2.02209964e-01 5.42519808e-01 -3.68468240e-02
6.93353951e-01 -2.57460296e-01 -3.03641587e-01 -1.19976652e+00
4.94005442e-01 -1.04432809e+00 1.90159827e-01 -6.40968621e-01
-9.55690801e-01 3.17108035e-01 -1.56897947e-01 -1.18003416e+00
-6.02144003e-01 -5.00801146e-01 -1.34718180e-01 9.12260950e-01
-1.79882741e+00 -1.60206258e+00 6.30947500e-02 4.72567588e-01
6.02828503e-01 2.64425963e-01 9.48249638e-01 5.70145130e-01
-3.26734602e-01 1.16482270e+00 1.54494494e-01 -3.79426815e-02
5.30560195e-01 -1.16999626e+00 6.73384249e-01 8.69242549e-01
3.70032161e-01 5.42579651e-01 8.34004402e-01 -4.84914690e-01
-1.52901316e+00 -9.93423939e-01 1.56344092e+00 -3.84780437e-01
4.68096763e-01 -6.27073467e-01 -4.84252542e-01 6.89203978e-01
3.35620433e-01 -2.20226794e-01 3.99608523e-01 -9.48658884e-02
-5.89856386e-01 -1.08466677e-01 -1.06731343e+00 9.32542503e-01
1.29282522e+00 -5.71493268e-01 -6.24857008e-01 1.75757706e-01
6.07252061e-01 -4.74028111e-01 -5.75295925e-01 7.47554451e-02
1.22976255e+00 -5.98731816e-01 8.51500452e-01 -2.82578349e-01
4.14205730e-01 -3.38574111e-01 -3.26674581e-01 -1.02228379e+00
3.45283031e-01 -8.13936889e-01 -5.02606452e-01 8.60667348e-01
6.33726358e-01 -4.51276511e-01 8.09008718e-01 4.67589140e-01
-3.30946714e-01 -7.51077473e-01 -1.31812084e+00 -8.61793101e-01
4.46675122e-02 -8.92265081e-01 4.13286477e-01 2.85040259e-01
-1.85448050e-01 4.35929269e-01 -7.38841593e-01 -2.04749808e-01
7.93122530e-01 -1.32728994e-01 7.26270318e-01 -8.42745602e-01
-5.01390517e-01 -5.49276352e-01 -1.88819140e-01 -1.20320737e+00
-2.11345717e-01 -9.70366418e-01 4.15424883e-01 -1.87266707e+00
-2.45977044e-02 -7.28079751e-02 -1.54610723e-01 7.07531095e-01
8.82096402e-03 4.60796893e-01 4.28507000e-01 1.65478379e-01
-4.74204063e-01 7.39786565e-01 1.08124948e+00 -1.50608733e-01
1.75399259e-01 -1.10274218e-01 -3.66051406e-01 6.13602400e-01
5.87115467e-01 -4.61304873e-01 -4.11802560e-01 -9.47753906e-01
3.15941304e-01 -2.20345959e-01 4.34523702e-01 -1.14214361e+00
3.35729152e-01 3.70967686e-01 -1.29718885e-01 -5.16851962e-01
6.43571615e-01 -1.01033831e+00 4.19625714e-02 5.21859348e-01
-2.99210578e-01 -2.47146383e-01 6.40655980e-02 4.53887641e-01
-2.45668087e-02 -1.82407409e-01 6.41802371e-01 1.12048388e-01
-3.30572903e-01 1.76244359e-02 -4.96896386e-01 -2.78552592e-01
5.31649709e-01 -2.92962730e-01 -9.76493582e-02 -3.60148668e-01
-4.44905460e-01 -5.34659773e-02 3.88089538e-01 2.45617881e-01
5.12929499e-01 -1.11206126e+00 -8.25330079e-01 4.27409038e-02
4.84560877e-02 -1.41819805e-01 1.77441016e-01 9.77613688e-01
-4.87550139e-01 2.28263587e-01 5.74440211e-02 -3.51799309e-01
-1.21078193e+00 1.07702345e-01 1.48990691e-01 -6.65594339e-01
-9.21152294e-01 8.29651594e-01 -4.45589840e-01 -4.66106117e-01
4.42266434e-01 -6.42859995e-01 4.11122024e-01 4.74989861e-02
5.07920802e-01 1.53853402e-01 2.59053558e-01 -4.24876928e-01
-4.75261025e-02 5.14833510e-01 -6.54498562e-02 -8.47818434e-01
1.33771741e+00 -3.38749364e-02 1.96334705e-01 3.73767585e-01
9.43558693e-01 -1.93689436e-01 -1.32400680e+00 -4.96902198e-01
8.76278505e-02 -5.60730845e-02 1.07274331e-01 -1.40731525e+00
-8.85204911e-01 1.06708109e+00 5.70797324e-01 -4.74710226e-01
1.26568210e+00 -4.26021129e-01 1.57060122e+00 6.27733588e-01
5.42241693e-01 -1.25258088e+00 -2.91778386e-01 5.78476012e-01
8.34337056e-01 -8.58225644e-01 -2.41931662e-01 9.91575345e-02
-5.06932914e-01 8.03068817e-01 -1.74131185e-01 -2.46474221e-02
2.67023206e-01 3.97837102e-01 9.29456428e-02 1.97880238e-01
-9.52856362e-01 -2.54260749e-01 2.35528544e-01 3.88874769e-01
2.05039948e-01 3.63602899e-02 -5.70353210e-01 3.80591929e-01
-2.90335983e-01 4.28927809e-01 -1.08101629e-01 1.07587218e+00
-2.97519386e-01 -1.36876428e+00 4.41265628e-02 4.80827361e-01
-4.64066595e-01 -5.58622420e-01 -2.07429513e-01 6.43911302e-01
9.69592482e-02 6.98381543e-01 -2.30259195e-01 -4.40815747e-01
3.90187204e-01 6.53797448e-01 7.40448117e-01 -1.12145744e-01
-6.77646995e-01 -4.05992083e-02 1.93077594e-01 -5.80679536e-01
-1.36465073e-01 -8.43590081e-01 -1.00757456e+00 -3.83988261e-01
-1.51799217e-01 -2.95882851e-01 1.04827750e+00 1.23649955e+00
2.31085122e-01 4.77128327e-01 -1.22669920e-01 -7.98603833e-01
-6.85461104e-01 -1.41834497e+00 -1.34187073e-01 1.28291771e-01
3.20075065e-01 -2.71651894e-01 -1.18470021e-01 2.57515758e-01] | [9.209833145141602, -6.530070781707764] |
d3c9f3d1-923c-4f1e-9faa-831b66dfe984 | stylized-data-to-text-generation-a-case-study | 2305.03256 | null | https://arxiv.org/abs/2305.03256v1 | https://arxiv.org/pdf/2305.03256v1.pdf | Stylized Data-to-Text Generation: A Case Study in the E-Commerce Domain | Existing data-to-text generation efforts mainly focus on generating a coherent text from non-linguistic input data, such as tables and attribute-value pairs, but overlook that different application scenarios may require texts of different styles. Inspired by this, we define a new task, namely stylized data-to-text generation, whose aim is to generate coherent text for the given non-linguistic data according to a specific style. This task is non-trivial, due to three challenges: the logic of the generated text, unstructured style reference, and biased training samples. To address these challenges, we propose a novel stylized data-to-text generation model, named StyleD2T, comprising three components: logic planning-enhanced data embedding, mask-based style embedding, and unbiased stylized text generation. In the first component, we introduce a graph-guided logic planner for attribute organization to ensure the logic of generated text. In the second component, we devise feature-level mask-based style embedding to extract the essential style signal from the given unstructured style reference. In the last one, pseudo triplet augmentation is utilized to achieve unbiased text generation, and a multi-condition based confidence assignment function is designed to ensure the quality of pseudo samples. Extensive experiments on a newly collected dataset from Taobao have been conducted, and the results show the superiority of our model over existing methods. | ['Liqiang Nie', 'Wei Zhou', 'Zhongzhou Zhao', 'Xuming Lin', 'Xuemeng Song', 'Liqiang Jing'] | 2023-05-05 | null | null | null | null | ['data-to-text-generation'] | ['natural-language-processing'] | [ 4.16762918e-01 1.66876465e-01 -1.30209982e-01 -4.97462034e-01
-7.63355076e-01 -5.59574366e-01 6.08988822e-01 -1.96919575e-01
2.23585650e-01 8.60225797e-01 4.98531342e-01 -6.28087372e-02
8.50108042e-02 -1.05868506e+00 -5.47048509e-01 -4.33142036e-01
6.53496981e-01 5.60373008e-01 -1.95218742e-01 -4.44227219e-01
2.97197819e-01 1.31358683e-01 -1.16997445e+00 2.55675733e-01
1.31721246e+00 1.06826329e+00 5.67576587e-02 2.62489110e-01
-6.39625072e-01 4.69678491e-01 -7.74583220e-01 -4.80841070e-01
1.29474923e-01 -8.22503448e-01 -6.19579017e-01 4.64526653e-01
1.41731381e-01 -3.40571582e-01 -1.69107560e-02 9.14721251e-01
6.82850063e-01 4.08798978e-02 9.23765540e-01 -1.24633133e+00
-9.13264930e-01 9.77714300e-01 -3.10583174e-01 -5.83371162e-01
4.98117834e-01 2.73616523e-01 1.22094941e+00 -9.76927817e-01
6.51384830e-01 1.58154702e+00 2.77415186e-01 8.77664745e-01
-1.41290224e+00 -5.93910456e-01 2.75619447e-01 -2.26945847e-01
-1.29526675e+00 -4.83407259e-01 1.19174588e+00 -3.45832944e-01
3.18381816e-01 2.00896889e-01 6.21716201e-01 1.26617074e+00
4.32999134e-02 1.02328503e+00 1.11666429e+00 -4.48030055e-01
2.83230096e-01 4.52614754e-01 -2.40828052e-01 7.09948957e-01
3.14005136e-01 -7.62862638e-02 -5.94613433e-01 8.75545889e-02
6.42944336e-01 -3.76387984e-01 -2.79748529e-01 -2.72794604e-01
-1.42098677e+00 8.53565514e-01 2.29689896e-01 1.88530728e-01
-1.89396799e-01 -1.59715235e-01 3.74020219e-01 1.58450559e-01
4.12938952e-01 5.26504457e-01 -2.56368548e-01 -4.92945407e-03
-7.50504851e-01 4.79577094e-01 6.64005697e-01 1.61151063e+00
7.41931498e-01 3.76442552e-01 -6.89581871e-01 7.66396105e-01
4.65039283e-01 6.88858211e-01 5.00573337e-01 -2.81825900e-01
9.99478698e-01 1.01842821e+00 4.83546183e-02 -1.10835218e+00
-1.00935891e-01 -1.34070262e-01 -1.02420700e+00 -1.58474833e-01
9.91728157e-02 -3.98946613e-01 -7.83161044e-01 1.68565512e+00
4.19172436e-01 -4.40550536e-01 2.42383271e-01 7.29752302e-01
9.37708497e-01 7.37720788e-01 -3.72632533e-01 -1.61178812e-01
1.33779347e+00 -7.59236991e-01 -9.78910506e-01 -2.40888000e-01
5.50184071e-01 -6.97730839e-01 1.67712176e+00 1.09792434e-01
-9.76172507e-01 -6.86942637e-01 -1.13756156e+00 -2.87730366e-01
-3.10611874e-01 5.42218864e-01 2.79689580e-01 7.07986891e-01
-5.96460044e-01 2.06574112e-01 -1.84968829e-01 -1.44891664e-01
4.82950270e-01 7.38184229e-02 6.31338870e-03 1.54083282e-01
-1.34071064e+00 4.11627650e-01 6.05625689e-01 2.54154652e-01
-5.92581868e-01 -6.03610575e-01 -9.88534868e-01 -1.08882718e-01
5.51626265e-01 -7.79970229e-01 9.57691073e-01 -8.13639343e-01
-1.76364255e+00 4.57635015e-01 -2.76910633e-01 1.21600352e-01
6.54136539e-01 -3.16637754e-02 -4.63520378e-01 -2.58853436e-01
3.56933564e-01 6.58128798e-01 9.50591624e-01 -1.65528464e+00
-5.71140826e-01 -2.03151047e-01 -1.47504121e-01 3.87397647e-01
-4.93298233e-01 -4.18370157e-01 -5.32686532e-01 -1.19159377e+00
9.62964445e-02 -6.44607008e-01 -4.07785103e-02 -2.36684695e-01
-1.05589879e+00 -2.91651070e-01 6.69584572e-01 -4.87226546e-01
1.44794595e+00 -2.05215096e+00 3.62460941e-01 3.34341198e-01
9.06565562e-02 -1.51402339e-01 -8.78026560e-02 4.24589425e-01
2.31996968e-01 3.69395941e-01 -4.21170473e-01 -3.81325990e-01
3.66047382e-01 -3.39504331e-02 -6.74563944e-01 -2.47808233e-01
6.84979022e-01 8.90034497e-01 -8.73598456e-01 -8.98971319e-01
5.20198271e-02 4.64165807e-02 -4.28784490e-01 4.44109976e-01
-6.36487842e-01 3.95661891e-01 -7.77263403e-01 6.92830563e-01
5.20906925e-01 3.93118449e-02 1.93933949e-01 -5.33320189e-01
1.05234951e-01 1.57470182e-01 -1.47090256e+00 1.69572449e+00
-5.33497930e-01 1.69805154e-01 -2.83672392e-01 -4.35500324e-01
1.30305552e+00 1.98572040e-01 -2.15215310e-02 -5.14560759e-01
3.67014080e-01 2.53868163e-01 -1.83833390e-01 -4.08019423e-01
8.89859974e-01 -2.36596927e-01 -4.53005284e-01 6.19428933e-01
-2.39240259e-01 -8.65392983e-01 2.94857562e-01 2.31776431e-01
6.52646065e-01 3.97303432e-01 -4.02774625e-02 -2.21559256e-01
7.44825065e-01 -8.65362585e-02 6.18287563e-01 3.07993919e-01
2.47081831e-01 7.02926457e-01 7.67026544e-01 -1.19520597e-01
-1.02675986e+00 -7.70376801e-01 1.60690069e-01 5.61517417e-01
2.74270535e-01 -4.72058147e-01 -9.20722663e-01 -1.07289743e+00
-1.39473379e-01 1.06195641e+00 -7.08432376e-01 -2.11002782e-01
-4.90281343e-01 -5.46805441e-01 3.55697483e-01 4.77645725e-01
6.34009778e-01 -1.12536192e+00 -8.81965235e-02 2.77230084e-01
-4.04631704e-01 -1.08207083e+00 -1.04407215e+00 -1.38879135e-01
-6.22713268e-01 -7.32266128e-01 -5.77748954e-01 -8.81094277e-01
9.92424905e-01 -1.32978708e-01 9.77580726e-01 -1.65160298e-01
1.22000180e-01 -4.01363410e-02 -4.32647586e-01 -4.93231148e-01
-5.90447485e-01 2.90315092e-01 -3.13505568e-02 4.91537362e-01
8.18939060e-02 -2.69890159e-01 -2.42555544e-01 3.11126381e-01
-1.21661067e+00 4.18886334e-01 7.61022747e-01 1.02822268e+00
5.91109991e-01 1.29051536e-01 7.52891958e-01 -1.04108822e+00
1.20245886e+00 -1.68415949e-01 -4.34456587e-01 4.18808520e-01
-7.98399925e-01 4.74931836e-01 1.19662368e+00 -5.03715217e-01
-1.19768834e+00 -1.64760109e-02 1.51556894e-01 -2.39080533e-01
1.31178796e-01 4.93505418e-01 -9.83393550e-01 4.56917524e-01
5.65213799e-01 5.63246667e-01 -3.93375754e-02 -1.90895900e-01
7.48965144e-01 8.41027856e-01 2.35076144e-01 -9.86420631e-01
1.19548476e+00 2.97056079e-01 6.32308647e-02 -5.59596002e-01
-6.82741106e-01 3.29635590e-01 -5.27597845e-01 -6.77557886e-02
7.58897603e-01 -5.56272507e-01 -4.26356405e-01 4.30312812e-01
-1.19420707e+00 -1.80778310e-01 -5.45752406e-01 1.27476767e-01
-6.75868928e-01 2.69953966e-01 -1.41247541e-01 -8.32523823e-01
-6.23113453e-01 -1.18044579e+00 1.33451021e+00 1.88304633e-01
-2.68796772e-01 -9.19550657e-01 -1.61424458e-01 2.65474141e-01
1.54167101e-01 4.47484642e-01 1.32210755e+00 -4.81381893e-01
-6.25175774e-01 -4.23084758e-02 -3.12626749e-01 4.50020760e-01
7.30388761e-01 7.10176826e-02 -6.74139917e-01 1.19834347e-02
-9.14200321e-02 -4.88034427e-01 3.64724576e-01 -9.11405981e-02
1.00253356e+00 -5.18909037e-01 -1.79345861e-01 5.78007936e-01
1.17016029e+00 3.40514153e-01 4.17012453e-01 5.23341633e-02
1.00933444e+00 8.07283640e-01 8.60548973e-01 4.80720729e-01
5.42699337e-01 6.17219746e-01 -6.92933723e-02 -9.18221381e-03
-1.71277896e-01 -6.85019910e-01 3.60131770e-01 8.29089463e-01
3.96232426e-01 -4.09189224e-01 -3.93419683e-01 4.62158084e-01
-1.61597252e+00 -7.33301878e-01 9.63116363e-02 1.95715141e+00
1.27174997e+00 4.39162999e-01 1.45917132e-01 4.04335082e-01
7.22289741e-01 1.94123134e-01 -6.59192860e-01 -2.25345060e-01
-2.11893573e-01 -4.90981042e-02 -1.62654608e-01 4.34725285e-01
-7.24834442e-01 1.07266772e+00 5.01210117e+00 1.04655218e+00
-1.13368940e+00 -4.35045898e-01 7.27442086e-01 1.15464889e-01
-1.04100740e+00 -1.02309771e-02 -8.67470860e-01 7.27070749e-01
3.54398072e-01 -4.20810342e-01 5.03424406e-01 6.86672926e-01
4.00902331e-01 2.04738155e-01 -1.26274157e+00 9.62653220e-01
2.44513750e-01 -1.24821258e+00 6.98539376e-01 -1.97645545e-01
6.89554214e-01 -1.03921187e+00 5.23881912e-02 2.70667553e-01
1.29215956e-01 -8.31931949e-01 1.17524970e+00 5.55709362e-01
1.29114842e+00 -8.19083393e-01 3.68581235e-01 1.44964457e-01
-1.34122729e+00 2.10939646e-01 -4.34112623e-02 3.57784003e-01
1.56521291e-01 8.01531017e-01 -9.21412885e-01 9.84851599e-01
1.24534063e-01 6.84422493e-01 -6.56403482e-01 2.38756076e-01
-4.27253604e-01 3.55270535e-01 5.44933267e-02 -5.70365667e-01
7.47006983e-02 -3.45083892e-01 5.10426581e-01 8.37585986e-01
3.31917912e-01 -1.12971246e-01 7.54107833e-02 1.45093846e+00
-1.66996062e-01 3.48711908e-01 -6.33151054e-01 -2.25675493e-01
6.78519547e-01 1.31187141e+00 -4.81987804e-01 -3.78690511e-01
-9.52470079e-02 1.06293428e+00 1.89952835e-01 3.24120343e-01
-6.99444294e-01 -9.80889440e-01 2.05487996e-01 -1.19368635e-01
4.96661812e-02 -1.64284706e-02 -8.44644368e-01 -1.27515948e+00
4.71569508e-01 -1.25288415e+00 -5.26598990e-02 -7.02838123e-01
-1.31364965e+00 7.80480087e-01 -3.88724580e-02 -1.41032016e+00
-2.68090338e-01 -2.78781921e-01 -7.09697604e-01 1.23930109e+00
-1.14930212e+00 -1.54608393e+00 -4.28792596e-01 4.02489901e-01
7.13199019e-01 -4.74177063e-01 4.82475072e-01 -1.75521243e-02
-8.11140120e-01 8.84963691e-01 -1.98528066e-01 1.69672951e-01
5.68077803e-01 -1.36177003e+00 6.89434707e-01 8.72992158e-01
-1.32803649e-01 6.50646508e-01 6.29462242e-01 -1.02085066e+00
-1.60292649e+00 -1.40785170e+00 8.63619387e-01 -4.55369145e-01
4.92677033e-01 -8.14303815e-01 -6.15586936e-01 5.27104497e-01
4.11198288e-03 -5.25592506e-01 4.24580753e-01 -2.14760080e-01
-1.05114207e-01 -3.58732104e-01 -1.12768734e+00 1.07989550e+00
1.02613938e+00 -2.78484374e-01 -5.59952974e-01 1.46293670e-01
1.03647220e+00 -4.80739474e-01 -7.15288877e-01 3.87878060e-01
3.97118866e-01 -5.25246024e-01 4.61727262e-01 -4.96432722e-01
6.37538016e-01 -5.69004357e-01 -1.42290935e-01 -1.51186180e+00
-1.60817936e-01 -9.93033946e-01 6.82044327e-02 1.99010015e+00
6.10068798e-01 -4.10527945e-01 8.36442530e-01 4.79761839e-01
-2.64022827e-01 -9.73028362e-01 -4.65528637e-01 -8.20066154e-01
-5.99901527e-02 -2.85203785e-01 1.24815834e+00 7.13684738e-01
-1.23953506e-01 8.72962534e-01 -5.10120630e-01 -1.62281185e-01
4.18262392e-01 3.53838563e-01 1.01020467e+00 -8.64311516e-01
-1.14259347e-01 -4.77718681e-01 1.18104294e-01 -1.03520119e+00
9.72657204e-02 -8.67249429e-01 2.88033664e-01 -1.71519911e+00
-1.22411318e-01 -7.37855077e-01 4.52046126e-01 2.55961657e-01
-5.31138957e-01 -3.09903294e-01 1.12206049e-01 -6.24958314e-02
-3.94726321e-02 1.09449422e+00 1.62249565e+00 -3.88938367e-01
-3.97868723e-01 -9.10622999e-02 -1.17467690e+00 2.72971511e-01
6.25953138e-01 -1.88247949e-01 -8.43363941e-01 -3.35141987e-01
1.41816795e-01 -4.32587005e-02 -7.61610195e-02 -7.75179088e-01
-2.29341805e-01 -5.11493325e-01 2.76561022e-01 -6.94412231e-01
1.56473324e-01 -7.49467552e-01 -1.25112176e-01 2.24211827e-01
-5.34768462e-01 1.43384442e-01 -5.06351404e-02 3.82436901e-01
-1.81981549e-02 -1.65173069e-01 5.20593166e-01 7.55932704e-02
-2.69048780e-01 3.99703056e-01 6.35319129e-02 4.70855385e-01
8.81652415e-01 -1.57495260e-01 -3.07156146e-01 -1.91610143e-01
-9.29713771e-02 3.89864445e-01 4.50554162e-01 5.11500180e-01
8.13742161e-01 -1.93015075e+00 -8.88372183e-01 5.05700767e-01
3.89337957e-01 3.39792252e-01 -4.56804186e-02 2.50810742e-01
-2.63244927e-01 2.94205099e-01 6.27421588e-02 -2.52951235e-01
-8.02238047e-01 5.11940897e-01 -1.20284305e-04 -2.88174748e-01
-2.55203962e-01 6.14133060e-01 1.50474146e-01 -5.38426816e-01
8.64707455e-02 -6.40061975e-01 -8.76041576e-02 1.41549215e-01
3.45928967e-01 1.95175797e-01 -4.06761952e-02 -3.61375391e-01
-4.32310998e-02 5.45380414e-01 -6.01928160e-02 -5.32344401e-01
8.35683882e-01 -1.01587079e-01 -1.66041136e-01 6.76602304e-01
8.17990243e-01 2.65809506e-01 -1.10476410e+00 -1.99697435e-01
-9.01006535e-02 -4.05838817e-01 -2.96552122e-01 -7.86960602e-01
-7.78655231e-01 9.40344810e-01 9.53259096e-02 2.33395934e-01
1.18305993e+00 -4.13843960e-01 1.06392789e+00 2.21944541e-01
1.74735501e-01 -1.20761287e+00 4.16762978e-01 4.69017416e-01
1.17859197e+00 -1.04868531e+00 -2.29561090e-01 -5.30803144e-01
-9.99973834e-01 1.06918001e+00 1.00639904e+00 3.41277301e-01
2.52968699e-01 3.30013558e-02 7.89150819e-02 7.38036484e-02
-7.11689413e-01 -8.23350623e-02 5.15470624e-01 7.18895376e-01
3.62466067e-01 -1.37606310e-02 -1.95836619e-01 8.59700143e-01
-5.18656909e-01 -2.25650016e-02 4.46854681e-01 8.26365888e-01
-2.03187510e-01 -1.34576583e+00 -3.53502542e-01 6.31879508e-01
3.57634947e-02 -2.48333693e-01 -6.51613176e-01 4.74005610e-01
1.13256060e-01 1.01401579e+00 -2.44123131e-01 -8.51433575e-01
6.38336182e-01 6.27184883e-02 2.94883937e-01 -7.93538332e-01
-5.08895576e-01 1.44057587e-01 2.30842263e-01 -9.17082205e-02
1.45293185e-02 -5.58281064e-01 -1.27348638e+00 -2.67111570e-01
-3.15877855e-01 2.49630034e-01 3.79134148e-01 8.98254573e-01
3.98378432e-01 6.96638286e-01 1.08431375e+00 -5.98991632e-01
-5.44595897e-01 -1.00886083e+00 -5.67465842e-01 5.43439269e-01
1.41477287e-01 -4.48663622e-01 -2.43896708e-01 9.55664143e-02] | [11.713399887084961, 8.956891059875488] |
f97efb85-7a4a-4bee-8ba5-826bb8ddd87c | m-3vsnet-unsupervised-multi-metric-multi-view | 2005.00363 | null | https://arxiv.org/abs/2005.00363v2 | https://arxiv.org/pdf/2005.00363v2.pdf | M^3VSNet: Unsupervised Multi-metric Multi-view Stereo Network | The present Multi-view stereo (MVS) methods with supervised learning-based networks have an impressive performance comparing with traditional MVS methods. However, the ground-truth depth maps for training are hard to be obtained and are within limited kinds of scenarios. In this paper, we propose a novel unsupervised multi-metric MVS network, named M^3VSNet, for dense point cloud reconstruction without any supervision. To improve the robustness and completeness of point cloud reconstruction, we propose a novel multi-metric loss function that combines pixel-wise and feature-wise loss function to learn the inherent constraints from different perspectives of matching correspondences. Besides, we also incorporate the normal-depth consistency in the 3D point cloud format to improve the accuracy and continuity of the estimated depth maps. Experimental results show that M3VSNet establishes the state-of-the-arts unsupervised method and achieves comparable performance with previous supervised MVSNet on the DTU dataset and demonstrates the powerful generalization ability on the Tanks and Temples benchmark with effective improvement. Our code is available at https://github.com/whubaichuan/M3VSNet | ['Xiao Liu', 'Yijia He', 'Hongwei Yi', 'Can Huang', 'Jingbin Liu', 'Baichuan Huang'] | 2020-04-30 | null | null | null | null | ['point-cloud-reconstruction'] | ['computer-vision'] | [-1.55774549e-01 -2.65197188e-01 -1.25572652e-01 -6.44402206e-01
-8.34583342e-01 -3.31923991e-01 3.88646036e-01 -2.99003869e-01
-1.61136135e-01 6.01631343e-01 1.18095227e-01 7.38748815e-03
-3.07711363e-01 -8.47363532e-01 -9.62012351e-01 -6.78719461e-01
4.29929465e-01 5.43747008e-01 4.17875081e-01 -2.41875023e-01
3.38529795e-01 5.36499977e-01 -1.54653752e+00 1.44900098e-01
9.63482440e-01 1.14369988e+00 4.55000550e-01 6.13703765e-02
1.45561425e-02 6.28080785e-01 1.89028308e-01 -7.87326992e-02
7.26873457e-01 1.30078107e-01 -4.91066217e-01 -8.73229206e-02
9.78139281e-01 -6.47608519e-01 -7.96893418e-01 1.02495658e+00
5.56251943e-01 3.67688164e-02 4.47032183e-01 -1.33033884e+00
-4.09250051e-01 2.68263333e-02 -6.14070237e-01 -1.49168447e-01
2.37810656e-01 1.75908938e-01 9.29000080e-01 -1.03162992e+00
6.81750596e-01 1.05962753e+00 7.29166269e-01 3.46912920e-01
-9.36108351e-01 -8.09870601e-01 6.52795509e-02 4.27230239e-01
-1.45354807e+00 -2.28313193e-01 1.01398826e+00 -4.26253974e-01
8.05256963e-01 3.45639102e-02 5.60492396e-01 8.76153052e-01
-9.67900679e-02 7.92297781e-01 1.05529571e+00 -4.99713346e-02
-3.01435892e-03 -1.74659625e-01 -2.61132836e-01 7.12198675e-01
-7.43230106e-03 3.60670477e-01 -6.71810806e-01 8.75985622e-03
1.13987839e+00 3.86054128e-01 -4.58457083e-01 -9.96060848e-01
-1.42314601e+00 8.12722385e-01 7.84170508e-01 -6.85784325e-04
-4.41333875e-02 1.83217362e-01 3.07198256e-01 1.88691810e-01
6.68707728e-01 9.04041156e-03 -5.36152720e-01 5.63371591e-02
-7.83981800e-01 8.64061415e-02 2.43643701e-01 1.25113499e+00
1.08841360e+00 1.58675015e-01 2.96708852e-01 7.63647795e-01
4.07797784e-01 8.63346934e-01 9.53545421e-02 -1.37295377e+00
8.48136008e-01 8.21895480e-01 7.86855891e-02 -1.09142005e+00
-3.59566569e-01 -3.76898259e-01 -1.06059325e+00 4.49210346e-01
2.14680389e-01 1.85685515e-01 -6.58857524e-01 1.39510691e+00
4.63902235e-01 4.25240099e-01 -4.43540653e-03 1.03760219e+00
1.10089624e+00 5.79741120e-01 -6.05245113e-01 4.61547524e-02
5.01028478e-01 -8.94307852e-01 -3.05498481e-01 -1.96128815e-01
4.09782648e-01 -6.24430716e-01 9.80716884e-01 3.00245076e-01
-9.94257331e-01 -5.84631443e-01 -1.15908265e+00 -2.55022615e-01
-3.34721170e-02 -3.35260183e-02 6.68138027e-01 -8.37633088e-02
-9.71898973e-01 8.82881582e-01 -8.75190973e-01 -1.75865158e-01
4.68564808e-01 2.54213631e-01 -6.06046081e-01 -4.52768654e-01
-8.84172678e-01 6.15106404e-01 2.41252542e-01 1.70635760e-01
-7.82766342e-01 -1.00622332e+00 -1.07828438e+00 -3.48740757e-01
2.28344396e-01 -9.79849935e-01 8.70190442e-01 -4.42140907e-01
-1.19683683e+00 9.30207074e-01 -1.94549620e-01 5.41107431e-02
7.79840946e-01 -1.08189680e-01 -8.48292261e-02 1.55385315e-01
3.64847183e-01 6.82467878e-01 4.06645596e-01 -1.62373984e+00
-7.05547571e-01 -6.48504913e-01 9.85694453e-02 3.51352394e-01
-7.37066567e-02 -5.08059263e-01 -4.99758184e-01 -3.42294097e-01
8.66015077e-01 -6.38617694e-01 -1.62518799e-01 5.00769436e-01
-3.36357743e-01 -9.09796208e-02 7.77637482e-01 -5.54219723e-01
6.74727201e-01 -2.20182395e+00 2.24102154e-01 2.01586157e-01
2.69779027e-01 -1.08976707e-01 2.09681727e-02 2.99508095e-01
1.13152325e-01 -3.98759723e-01 -4.02565032e-01 -4.65042055e-01
-1.14413798e-01 3.33846718e-01 5.24194986e-02 8.19092274e-01
-2.57562369e-01 7.41274416e-01 -9.46288824e-01 -5.34863591e-01
6.75566196e-01 5.03405392e-01 -4.18006629e-01 1.87127545e-01
-4.96436134e-02 9.25781906e-01 -5.31835139e-01 7.92132437e-01
1.02384698e+00 -3.08500022e-01 -4.09917563e-01 -5.10346770e-01
-2.49373794e-01 -3.79737839e-02 -1.28821981e+00 2.47166181e+00
-5.48456311e-01 4.65819001e-01 -1.55961874e-03 -9.21181142e-01
1.02246809e+00 -5.73193934e-03 7.11058915e-01 -7.79772401e-01
8.68713111e-02 4.19933975e-01 -4.33199525e-01 -3.74959499e-01
2.48975605e-01 -9.05834511e-02 2.57631302e-01 -9.51752160e-03
4.93796840e-02 -4.69782740e-01 -2.25940317e-01 6.41953051e-02
7.38100111e-01 5.23797631e-01 5.37539870e-02 -1.40173256e-01
6.51362717e-01 6.48463191e-03 9.62168753e-01 2.92049646e-01
-6.57239631e-02 1.15929818e+00 -9.13775712e-02 -5.02208889e-01
-1.27175093e+00 -1.28196549e+00 -3.94857287e-01 3.11755717e-01
6.30127311e-01 -9.78813320e-02 -2.56702721e-01 -5.07222474e-01
2.87587821e-01 4.98506337e-01 -2.40923077e-01 8.07741582e-02
-5.41544139e-01 -2.58899271e-01 9.10249427e-02 8.19191515e-01
9.03662562e-01 -6.67597473e-01 -1.41168967e-01 -8.47184435e-02
-5.04948556e-01 -1.32567000e+00 -4.34477210e-01 -2.34085724e-01
-1.27347481e+00 -1.14945507e+00 -5.55243015e-01 -7.82778919e-01
7.52466023e-01 4.94537294e-01 1.00662613e+00 -1.21377513e-01
1.02416918e-01 2.69516051e-01 -2.79123813e-01 -1.78121150e-01
-6.67752465e-03 -1.41826525e-01 1.71543688e-01 -1.27888978e-01
3.59725952e-01 -1.19366455e+00 -9.30227041e-01 5.41659534e-01
-6.40146434e-01 2.50237584e-01 3.79683256e-01 7.43501842e-01
1.04574096e+00 -2.09566504e-01 6.85699284e-02 -5.03476381e-01
-3.76412511e-01 -3.18398774e-01 -8.49240243e-01 -4.21751626e-02
-5.97821295e-01 -2.14734927e-01 4.89275396e-01 7.02627301e-02
-7.80587316e-01 3.48482579e-01 -3.56063426e-01 -9.61973429e-01
-1.67921573e-01 1.48200557e-01 -3.39373022e-01 -5.55953562e-01
4.50065225e-01 4.88319457e-01 4.49352190e-02 -5.48616648e-01
1.58560634e-01 5.01755297e-01 5.95481336e-01 -3.00271124e-01
1.20226896e+00 1.00448060e+00 2.63102174e-01 -5.21976650e-01
-9.81662869e-01 -8.69599879e-01 -9.31990147e-01 -3.80577147e-01
7.35977054e-01 -1.37733841e+00 -5.64908743e-01 5.47967851e-01
-1.18458664e+00 -2.16235548e-01 -1.38200417e-01 6.36993408e-01
-9.16897058e-01 6.82957768e-01 -5.54880738e-01 -3.94084185e-01
-3.29952121e-01 -1.01710677e+00 1.27365589e+00 -7.43035600e-03
3.47750634e-01 -9.91170824e-01 -1.07325204e-02 6.39537752e-01
7.98766315e-02 5.82423925e-01 4.59268212e-01 3.45562845e-02
-1.07662785e+00 7.30589926e-02 -3.26057106e-01 5.65026879e-01
1.46844983e-01 -2.41868809e-01 -9.48468983e-01 -4.36298192e-01
2.09182471e-01 -3.69297147e-01 8.55793417e-01 4.89511043e-01
1.24104679e+00 1.44762009e-01 -1.65526703e-01 1.38117361e+00
1.91043043e+00 -1.29416332e-01 6.97759151e-01 7.18635440e-01
1.19814992e+00 4.52729523e-01 8.20972443e-01 5.32685041e-01
7.08048582e-01 6.45106375e-01 1.00026464e+00 -1.43028289e-01
7.96035156e-02 -4.63047653e-01 6.56017736e-02 1.11860967e+00
-2.59684473e-01 1.47660479e-01 -9.48640347e-01 5.14497638e-01
-2.12804413e+00 -7.70133138e-01 -4.13865596e-01 2.20654058e+00
3.43244314e-01 -2.40711570e-02 -2.99327165e-01 1.51538132e-02
6.71130538e-01 3.85886252e-01 -8.18749249e-01 3.30183625e-01
-3.49640399e-01 -2.28796467e-01 7.79610693e-01 4.84966725e-01
-1.09732807e+00 8.06012034e-01 5.14827871e+00 7.98507869e-01
-9.46422100e-01 2.83102304e-01 3.26006263e-01 -1.86888829e-01
-4.67111528e-01 -2.51256302e-02 -7.63519943e-01 3.62212390e-01
2.34149441e-01 1.32748351e-01 2.76595414e-01 8.83914292e-01
2.89940417e-01 3.73124518e-02 -1.14703906e+00 1.49255824e+00
1.62924126e-01 -1.54541743e+00 1.71163771e-02 -1.51697267e-02
1.02517164e+00 6.68056726e-01 -9.04131457e-02 -1.49951905e-01
1.64464340e-01 -6.43417120e-01 7.00235307e-01 6.26998782e-01
9.80121493e-01 -7.50072479e-01 7.24847078e-01 5.37061095e-01
-1.39072204e+00 8.91820900e-03 -6.87893093e-01 3.78217362e-02
4.12987620e-01 5.59364974e-01 -1.78481475e-01 1.14038348e+00
8.52973700e-01 1.36401165e+00 -4.04090255e-01 1.14618027e+00
-2.55210280e-01 5.80703765e-02 -4.13280308e-01 4.21808630e-01
3.12730074e-01 -5.32096207e-01 6.34451687e-01 6.40262604e-01
5.09893417e-01 6.45815283e-02 2.21212611e-01 8.80902648e-01
9.52749178e-02 1.23088267e-02 -8.05742264e-01 6.97698653e-01
3.50139052e-01 1.20773363e+00 -2.38567784e-01 -5.59923500e-02
-6.74620807e-01 8.33051801e-01 3.61214131e-01 2.45108873e-01
-6.34019136e-01 -1.51667511e-02 6.48341537e-01 2.74952769e-01
2.67796427e-01 -3.76264840e-01 -5.20158291e-01 -1.41451108e+00
4.70573545e-01 -4.98602122e-01 2.79968470e-01 -1.05613959e+00
-1.52969813e+00 3.83877486e-01 -1.07961977e-02 -1.97795558e+00
9.09946561e-02 -5.50727189e-01 -5.89967668e-01 7.63357759e-01
-1.90953183e+00 -1.37753987e+00 -8.64723980e-01 7.98245132e-01
4.83174264e-01 -2.65036881e-01 3.43877584e-01 3.67138237e-01
-1.35319099e-01 2.82590151e-01 4.90686029e-01 1.05632000e-01
7.41866171e-01 -1.04596686e+00 4.24282193e-01 6.38492465e-01
-1.47514030e-01 2.29948163e-01 3.59358341e-01 -5.60067952e-01
-1.48160708e+00 -1.25178075e+00 4.95133221e-01 -4.61664528e-01
4.77014303e-01 -2.85982788e-01 -9.13426757e-01 6.72711790e-01
-1.22971192e-01 3.82397950e-01 2.92125344e-01 -1.82659969e-01
-3.22573751e-01 -4.53249842e-01 -1.27555752e+00 2.31727630e-01
1.50941861e+00 -5.91266751e-01 -4.62082207e-01 5.56661427e-01
8.16725552e-01 -8.07437897e-01 -9.67775822e-01 9.12845492e-01
3.65042388e-01 -1.28536963e+00 1.19409764e+00 -1.63440093e-01
6.66206539e-01 -6.27901018e-01 -6.14049435e-01 -1.24153054e+00
-3.36439461e-01 -1.00609720e-01 -1.04646310e-02 1.05094826e+00
1.66470259e-01 -7.93080807e-01 1.07615101e+00 1.80300027e-01
-5.26257098e-01 -8.99023890e-01 -1.18972981e+00 -9.92352605e-01
7.60057196e-02 -4.47717756e-01 5.99735141e-01 1.07474029e+00
-3.88999164e-01 1.58237293e-02 -3.66608799e-01 5.48019230e-01
1.18780732e+00 3.29193443e-01 8.66916895e-01 -1.20745111e+00
-1.86664447e-01 -1.17301911e-01 -7.57490575e-01 -1.32332897e+00
1.80826783e-01 -1.03992426e+00 -5.15613630e-02 -1.93506491e+00
1.37452334e-01 -4.66962397e-01 -2.49126241e-01 1.25943467e-01
1.57495975e-01 2.03900203e-01 -7.90263992e-03 4.62547690e-01
-5.47400534e-01 9.57363486e-01 1.54100072e+00 -1.76664323e-01
1.00225993e-01 -2.44498495e-02 -1.78759053e-01 8.83972824e-01
7.20140874e-01 -2.35611424e-01 -4.83151734e-01 -8.63738775e-01
1.20981745e-01 1.41909331e-01 5.35637915e-01 -1.24661648e+00
4.78829682e-01 -3.25536013e-01 3.53819460e-01 -1.20129788e+00
6.22300267e-01 -9.34235036e-01 2.89905995e-01 1.68640241e-01
1.70379326e-01 2.07598670e-03 -1.14344262e-01 6.54150844e-01
-4.33961421e-01 1.99129339e-02 7.96004832e-01 -3.33895057e-01
-1.04132700e+00 1.01375973e+00 6.22587264e-01 1.08963981e-01
9.08650517e-01 -4.45665628e-01 -2.35645652e-01 -2.75604069e-01
-3.84015292e-01 5.88270187e-01 8.91156614e-01 3.40148151e-01
1.08434129e+00 -1.62394381e+00 -7.22448409e-01 1.79509029e-01
4.74192291e-01 8.21506500e-01 5.17683625e-01 7.47549415e-01
-8.19488645e-01 2.26162285e-01 -2.79940248e-01 -1.12524021e+00
-9.91342127e-01 3.41426730e-01 5.09629726e-01 1.68843657e-01
-1.01089799e+00 7.47641861e-01 4.33439076e-01 -9.72295403e-01
2.24190339e-01 -1.55543774e-01 -2.28126179e-02 -4.24647957e-01
2.36236706e-01 4.49051231e-01 1.40878018e-02 -7.35354483e-01
-2.73658931e-01 1.11705887e+00 3.01811397e-01 2.16195703e-01
1.66929042e+00 -1.95013106e-01 -1.64261967e-01 6.10644162e-01
1.34559453e+00 -3.53737384e-01 -1.67199326e+00 -5.14102995e-01
-3.73046249e-01 -9.75847244e-01 2.50931859e-01 -2.45929778e-01
-1.41423368e+00 8.98903012e-01 6.75536871e-01 -5.70772350e-01
9.06544447e-01 3.97561640e-02 1.07853150e+00 3.51471782e-01
8.54448915e-01 -1.07141304e+00 -1.32769570e-02 4.78093743e-01
9.59795713e-01 -1.65742159e+00 2.02581599e-01 -6.34774268e-01
-3.93351465e-01 1.13954186e+00 9.46359575e-01 -2.67055035e-01
7.18247235e-01 1.06975455e-02 1.74980789e-01 -3.61946374e-01
-2.77162820e-01 -4.48261574e-02 2.54550189e-01 7.12889969e-01
-2.84167100e-02 -2.51273721e-01 5.85048757e-02 -3.63684595e-02
-1.46268457e-01 -8.07822198e-02 1.76740527e-01 9.25333381e-01
-4.07782316e-01 -8.99791598e-01 -1.93335325e-01 3.90108049e-01
1.08552985e-01 -2.60527451e-02 2.16704775e-02 7.17776597e-01
1.96416885e-01 6.45309865e-01 8.36951211e-02 -6.34229541e-01
6.65773749e-01 -3.44953150e-01 6.42826676e-01 -4.58529890e-01
1.61911715e-02 -1.59518391e-01 -1.48056298e-01 -9.02474642e-01
-7.60903180e-01 -8.36745441e-01 -1.14273906e+00 -5.54294944e-01
-1.58858567e-01 -2.24987447e-01 6.73247874e-01 9.54008281e-01
2.60106415e-01 8.65695551e-02 1.05729723e+00 -1.11670196e+00
-4.27999407e-01 -7.54612505e-01 -6.70562387e-01 4.32344139e-01
3.74282628e-01 -8.09760928e-01 -6.15115047e-01 -2.73782045e-01] | [8.623281478881836, -2.6853835582733154] |
7ebb9b2a-a308-49af-b2a1-b6a0bac4f866 | panacea-an-automated-misinformation-detection | 2303.01241 | null | https://arxiv.org/abs/2303.01241v1 | https://arxiv.org/pdf/2303.01241v1.pdf | PANACEA: An Automated Misinformation Detection System on COVID-19 | In this demo, we introduce a web-based misinformation detection system PANACEA on COVID-19 related claims, which has two modules, fact-checking and rumour detection. Our fact-checking module, which is supported by novel natural language inference methods with a self-attention network, outperforms state-of-the-art approaches. It is also able to give automated veracity assessment and ranked supporting evidence with the stance towards the claim to be checked. In addition, PANACEA adapts the bi-directional graph convolutional networks model, which is able to detect rumours based on comment networks of related tweets, instead of relying on the knowledge base. This rumour detection module assists by warning the users in the early stages when a knowledge base may not be available. | ['Yulan He', 'Maria Liakata', 'Rob Procter', 'Arkaitz Zubiaga', 'Lin Gui', 'Elena Kochkina', 'Lixing Zhu', 'Miguel Arana-Catania', 'Runcong Zhao'] | 2023-02-28 | null | null | null | null | ['rumour-detection'] | ['natural-language-processing'] | [-4.06383395e-01 6.43098056e-01 -5.26561022e-01 1.96105242e-01
-3.45375687e-01 -5.16659498e-01 1.04855573e+00 1.11756039e+00
-2.29843959e-01 6.31052971e-01 6.01341963e-01 -7.25445449e-01
2.17608176e-02 -1.28696084e+00 -3.89247686e-01 1.50542423e-01
-2.12643921e-01 6.28243685e-01 5.94271302e-01 -9.71700430e-01
7.77411163e-01 3.17874104e-01 -9.45362985e-01 6.05865419e-01
7.39728510e-01 1.02037990e+00 -6.03662312e-01 5.78547895e-01
-1.88932508e-01 1.80625057e+00 -6.30962014e-01 -1.00122392e+00
-2.16020480e-01 -1.59614593e-01 -1.00187755e+00 -2.88516402e-01
5.27966619e-01 -4.38895255e-01 -5.99066675e-01 1.35464346e+00
1.01762168e-01 -4.26741868e-01 4.73892599e-01 -1.14271569e+00
-1.08249128e+00 1.13018703e+00 -3.81475002e-01 7.74507701e-01
5.61216116e-01 1.71994209e-01 1.09086430e+00 -7.44104326e-01
9.99386370e-01 9.91552770e-01 1.11108708e+00 1.58107713e-01
-8.19363177e-01 -4.88742262e-01 -3.06992859e-01 6.43088102e-01
-9.30112898e-01 -1.89993739e-01 1.03790259e+00 -8.77071738e-01
6.90277398e-01 2.03850895e-01 5.00201702e-01 1.23835111e+00
3.10004473e-01 3.94852608e-01 1.09839690e+00 -3.39242034e-02
2.57482141e-01 2.58060783e-01 4.51629817e-01 1.12595749e+00
6.17722929e-01 -1.95250943e-01 -6.32911026e-01 -7.11235404e-01
4.05000985e-01 -6.47791773e-02 -2.10770056e-01 1.19279198e-01
-1.02198863e+00 1.38982928e+00 1.04466963e+00 6.75937891e-01
-1.04235172e+00 -2.30953678e-01 7.13569045e-01 4.94592339e-01
1.03590119e+00 5.79320550e-01 1.29983202e-01 3.37172866e-01
-1.41291714e+00 3.36621016e-01 1.42989373e+00 1.99000701e-01
4.97522086e-01 -3.93136255e-02 -5.32968700e-01 4.03251350e-01
2.57631689e-01 8.41409937e-02 1.73485622e-01 -5.00807106e-01
5.87497711e-01 1.11268091e+00 1.97803900e-01 -1.90628266e+00
-7.22049475e-01 -6.72286272e-01 -8.56982529e-01 2.50139832e-01
2.09470347e-01 -2.17772618e-01 -4.23061877e-01 9.76709247e-01
2.32578561e-01 2.24698141e-01 -1.02577426e-01 1.02907920e+00
1.25840914e+00 3.28923792e-01 -2.28108630e-01 -2.10834861e-01
1.53196645e+00 -7.79603422e-01 -9.30517972e-01 7.06741139e-02
3.41928810e-01 -5.47787070e-01 3.84648502e-01 2.64041841e-01
-9.73697424e-01 1.18409609e-03 -1.31675673e+00 4.98749577e-02
-7.75361180e-01 -5.29258788e-01 4.02741522e-01 3.92653912e-01
-1.16387630e+00 8.32848907e-01 -1.29798546e-01 -2.41718888e-01
7.96495557e-01 -4.39511389e-01 -2.36995965e-01 5.04000485e-02
-1.83864379e+00 1.20115626e+00 5.79509974e-01 -1.31022468e-01
-1.05488849e+00 -4.29407477e-01 -7.56719589e-01 8.23425055e-02
6.31405592e-01 -6.37973607e-01 8.95740271e-01 -6.71478748e-01
-1.26184464e+00 9.71096218e-01 5.49881101e-01 -1.13327873e+00
9.84415293e-01 1.74163416e-01 -9.10337031e-01 5.61226249e-01
3.77152562e-01 -2.25160390e-01 9.55571771e-01 -1.06716847e+00
-2.93636143e-01 -1.45927951e-01 2.80263752e-01 -4.33652192e-01
-2.12965578e-01 3.26489180e-01 2.19035715e-01 -4.04361665e-01
-1.29761443e-01 -4.78354305e-01 -8.21064413e-03 -2.87300169e-01
-8.77897322e-01 -3.52216125e-01 7.94059753e-01 -1.01871932e+00
1.59064829e+00 -1.54518580e+00 -5.63743532e-01 2.94078529e-01
9.97688890e-01 5.43683887e-01 1.41690373e-01 8.90958369e-01
2.60192901e-01 4.17858005e-01 3.45391929e-02 1.57378599e-01
-8.45086575e-02 -2.88183928e-01 -3.62198740e-01 6.17027164e-01
2.11233705e-01 9.00551438e-01 -1.20788407e+00 -4.94330168e-01
-3.28022897e-01 1.54586375e-01 -3.46228749e-01 -7.29523748e-02
-5.39857626e-01 -2.35743169e-02 -3.90836716e-01 4.50135231e-01
6.86115563e-01 -7.15492606e-01 2.51914501e-01 -3.95367928e-02
-1.76988572e-01 7.66287982e-01 -5.55534959e-01 8.17812741e-01
-6.45425767e-02 7.05991328e-01 2.29898319e-01 -6.95072889e-01
8.78905892e-01 4.23439890e-01 1.69192985e-01 -3.79159302e-01
4.71465111e-01 2.24517301e-01 -4.06841934e-01 -7.84053206e-01
7.13406146e-01 -5.89527525e-02 6.40533864e-02 6.01784348e-01
-1.09504007e-01 2.51704663e-01 1.86472774e-01 9.31378841e-01
1.37248063e+00 -6.11649930e-01 4.75879014e-01 -2.22863391e-01
5.84291458e-01 4.56809849e-01 9.95835513e-02 8.48612249e-01
-2.04605460e-01 8.36648196e-02 9.61440265e-01 -6.42257690e-01
-9.09713566e-01 -6.09054744e-01 1.96445793e-01 8.88028920e-01
-1.88951671e-01 -4.73325729e-01 -4.04462993e-01 -1.17280018e+00
3.09976310e-01 1.03923774e+00 -1.13419163e+00 2.22453430e-01
-4.85295542e-02 -3.26990813e-01 7.02836633e-01 -1.13153808e-01
6.78518355e-01 -1.11535704e+00 -2.48215586e-01 4.65741724e-01
-4.47325945e-01 -1.04058766e+00 -1.01890950e-03 -3.74859005e-01
-4.46517944e-01 -1.58143723e+00 -4.23178762e-01 -3.22313815e-01
3.85783724e-02 1.34719685e-01 1.22437584e+00 6.84647202e-01
4.12183166e-01 -2.41445713e-02 -3.11348528e-01 -4.25894737e-01
-1.19064140e+00 4.97423783e-02 -2.07887709e-01 2.86191463e-01
2.31133327e-01 -5.54075241e-01 -5.13886929e-01 -8.15484002e-02
-1.00533891e+00 -1.57525048e-01 2.88881719e-01 5.06802082e-01
-2.96433479e-01 -1.75351113e-01 8.11105967e-01 -1.23348677e+00
1.30241418e+00 -1.37773800e+00 -4.42115158e-01 -5.05198427e-02
-9.88554180e-01 -1.84885040e-01 2.98442811e-01 -9.44904685e-02
-5.60965002e-01 -9.25478101e-01 -1.44171104e-01 -2.77846366e-01
1.32123500e-01 1.25579274e+00 7.22410262e-01 -3.50798294e-03
1.34081662e+00 -2.03465462e-01 1.51245594e-01 -3.71085882e-01
4.97399390e-01 9.89033461e-01 4.74935085e-01 3.40332329e-01
7.25973725e-01 5.44205844e-01 -4.17420119e-02 -5.58520019e-01
-1.45448148e+00 -6.45733833e-01 -4.73304689e-01 -5.98698139e-01
5.98302662e-01 -5.50649226e-01 -1.04805291e+00 3.08363616e-01
-1.49995458e+00 1.89589635e-01 2.48589948e-01 3.94826308e-02
-1.32270113e-01 5.57118058e-01 -1.21023595e+00 -8.45720112e-01
-6.12359524e-01 -1.62138432e-01 1.65814474e-01 -1.89489782e-01
-1.39110982e-01 -1.05679202e+00 2.32400134e-01 7.68918276e-01
8.05216253e-01 6.63797021e-01 5.54588854e-01 -1.36653590e+00
-2.09800363e-01 -5.40850699e-01 -6.49655402e-01 9.05993879e-02
-3.27817053e-01 -2.78618246e-01 -6.94002926e-01 -3.81920934e-02
-1.41331360e-01 -3.88061106e-01 1.14859509e+00 -1.38626128e-01
2.06293315e-01 -1.35104442e+00 -8.63703489e-02 -2.77365208e-01
1.33892858e+00 -8.04674566e-01 6.13289475e-01 7.11625516e-01
6.15418375e-01 5.93620181e-01 2.95829680e-02 5.95231533e-01
8.75311136e-01 3.38427812e-01 1.11497271e+00 1.33716077e-01
-8.33588168e-02 -4.74649310e-01 2.73545653e-01 6.57332420e-01
-2.66854674e-01 -2.62489378e-01 -1.14268613e+00 6.21979594e-01
-1.96707666e+00 -1.42705166e+00 -9.30835664e-01 1.57796168e+00
7.85579681e-01 6.41966820e-01 5.12215078e-01 1.56232081e-02
9.59697604e-01 5.78827858e-01 -8.19676965e-02 -6.11527741e-01
-1.37113467e-01 -3.48156005e-01 5.85947096e-01 6.87475562e-01
-9.17712212e-01 8.22030723e-01 6.08104944e+00 5.67147315e-01
-9.06833291e-01 5.99639297e-01 1.20829351e-01 3.00177485e-01
-4.42335129e-01 7.56327659e-02 -2.63296455e-01 4.95388091e-01
1.01322782e+00 -7.95795582e-03 2.51935452e-01 6.51482761e-01
1.87895328e-01 -1.75549164e-02 -5.35680234e-01 3.69069517e-01
5.53567827e-01 -2.34883356e+00 1.21162981e-01 -5.77507392e-02
5.98836660e-01 6.85483277e-01 -3.23136419e-01 3.13300520e-01
7.34114349e-01 -6.70816779e-01 8.19164336e-01 8.29207420e-01
4.34925288e-01 -5.52941501e-01 1.20269930e+00 7.47886300e-01
-2.86937892e-01 -1.66373879e-01 -1.50356248e-01 -2.45743081e-01
2.08854198e-01 1.05216074e+00 -1.41422272e+00 7.39586651e-01
3.02140445e-01 5.90873837e-01 -6.64059639e-01 9.32971954e-01
-6.96913302e-01 1.01280642e+00 1.16690606e-01 -2.98357010e-01
6.00572169e-01 3.23434979e-01 9.89233613e-01 1.50619483e+00
-1.06095642e-01 -9.36770439e-02 3.06609690e-01 1.15865266e+00
-4.87635881e-01 8.15733448e-02 -6.04131520e-01 -8.14976990e-02
2.80486137e-01 1.24884641e+00 -4.12913531e-01 -6.83908641e-01
-6.52207881e-02 8.37901354e-01 6.77864194e-01 -5.36410324e-02
-4.76774305e-01 -1.31603688e-01 -1.65483892e-01 7.31649458e-01
3.34344327e-01 -3.47013623e-02 1.13510666e-02 -1.18121707e+00
-5.14460921e-01 -7.15767801e-01 9.02808428e-01 -9.22510803e-01
-1.72965729e+00 7.58733928e-01 -4.95782822e-01 -8.61875117e-01
-1.38845965e-01 -1.95387527e-01 -9.67181921e-01 6.49309337e-01
-1.85593569e+00 -1.24688351e+00 -9.89575237e-02 6.58534467e-01
-1.29996940e-01 -2.26317644e-01 5.57962179e-01 8.12448487e-02
-3.51203591e-01 1.57434598e-01 -3.44532102e-01 6.51426554e-01
3.72334749e-01 -9.26037848e-01 3.21611434e-01 6.01345301e-01
2.51378655e-03 2.61300951e-01 1.05479789e+00 -1.22788298e+00
-8.01909208e-01 -9.92513716e-01 1.47232962e+00 -5.71890771e-01
1.73270226e+00 5.30981496e-02 -1.34067535e+00 4.13189203e-01
5.16957819e-01 -9.67647359e-02 4.40237075e-01 2.98645943e-01
-9.02491033e-01 4.18988675e-01 -1.29431295e+00 2.66988754e-01
6.25647604e-01 -6.38996065e-01 -1.00655401e+00 8.13305736e-01
5.10435164e-01 -6.43597618e-02 -7.96977401e-01 -3.24869826e-02
6.21757424e-03 -1.15300703e+00 4.41561371e-01 -1.01551175e+00
8.94082308e-01 -2.91215219e-02 4.36564028e-01 -1.18958294e+00
-7.31339037e-01 -7.83368826e-01 -5.80101609e-01 1.12988627e+00
7.25426912e-01 -6.84964657e-01 5.50160110e-01 -1.63745418e-01
3.49830806e-01 -3.53498846e-01 -9.19166028e-01 -5.56730568e-01
-1.51086912e-01 -4.40234840e-01 4.48575199e-01 1.56811368e+00
6.57236576e-01 7.39949763e-01 -5.32370210e-01 2.94547409e-01
5.54318368e-01 2.29912266e-01 3.76312345e-01 -1.65692484e+00
1.18887946e-01 -7.06830680e-01 -5.55292010e-01 -2.02653244e-01
-4.65006158e-02 -1.19108462e+00 -4.10127848e-01 -1.85666406e+00
2.05587044e-01 8.25484321e-02 -3.81721705e-01 4.58473265e-01
6.99352324e-02 5.65375686e-01 1.00992694e-01 5.79970956e-01
-9.64651823e-01 9.63768810e-02 1.01115906e+00 -2.79345095e-01
1.52133763e-01 9.65630412e-02 -7.66038477e-01 1.04488695e+00
9.52906847e-01 -8.72444391e-01 4.52394217e-01 2.45225847e-01
1.14945936e+00 4.24692839e-01 9.15866315e-01 -8.37959051e-01
5.66295028e-01 7.11910278e-02 -5.63165508e-02 -8.23591888e-01
-1.82621449e-01 -4.05878693e-01 -3.45960557e-01 6.90359294e-01
-4.23779696e-01 2.79926043e-03 -7.04361871e-02 1.01486969e+00
-8.91909450e-02 -2.57666647e-01 6.06122375e-01 -3.54158133e-01
-3.31191391e-01 1.58321396e-01 -6.69258893e-01 2.14927122e-01
5.46351612e-01 2.26298466e-01 -1.04565656e+00 -9.00390744e-01
-6.86114252e-01 1.33311793e-01 1.80081829e-01 3.55915248e-01
4.19602305e-01 -1.20110083e+00 -1.59106672e+00 -4.02086943e-01
3.35258722e-01 -9.44160223e-01 1.61831826e-01 1.19970679e+00
-5.69553256e-01 1.49127558e-01 -2.10252419e-01 1.92533582e-02
-7.61616826e-01 1.27556705e+00 7.59167075e-02 -6.60343289e-01
-6.00017726e-01 3.51191998e-01 -9.51185584e-01 -7.83592537e-02
-1.30628556e-01 1.33841783e-01 -1.00276458e+00 5.11719108e-01
7.62400210e-01 7.02519000e-01 1.99266255e-01 -8.19423854e-01
-3.48668456e-01 -4.43908989e-01 -5.93783557e-02 -9.26319323e-03
1.45962596e+00 -8.43439400e-02 -5.96808970e-01 3.96211863e-01
7.08551645e-01 5.38785279e-01 -5.27052879e-01 -6.76044762e-01
4.14513677e-01 -2.41508201e-01 4.36234057e-01 -1.26797748e+00
-8.62457633e-01 4.58079487e-01 -2.64416486e-01 1.48819208e+00
1.99467629e-01 1.78461269e-01 7.02548742e-01 2.68879354e-01
9.58722159e-02 -9.97496009e-01 6.98104799e-02 8.86644304e-01
1.43455577e+00 -1.33480132e+00 2.41867736e-01 -4.29104507e-01
-7.22421110e-01 1.28660309e+00 1.11239478e-01 -2.96212316e-01
7.94272125e-01 -9.52644199e-02 7.75400102e-02 -1.16789913e+00
-5.88483632e-01 -1.45581856e-01 6.00282371e-01 2.77266741e-01
2.07818914e-02 1.95920363e-01 -5.37365973e-01 5.63193440e-01
-1.77468002e-01 1.05363712e-01 9.60266769e-01 2.60175169e-01
-8.38528395e-01 -2.85074592e-01 -4.24640596e-01 5.18066287e-01
-9.61971998e-01 -4.49262112e-01 -8.99359763e-01 5.83384216e-01
-5.47341220e-02 1.38279021e+00 -3.81354868e-01 -6.10386193e-01
1.36871159e-01 -3.35316435e-02 -2.67040282e-01 -4.60228890e-01
-1.16625810e+00 -5.91329336e-01 9.37962353e-01 -5.09912848e-01
-3.69784892e-01 -2.19492257e-01 -8.14012468e-01 -1.06868947e+00
-5.77223420e-01 3.18919688e-01 5.75851500e-01 9.43785548e-01
5.11124134e-01 2.54828721e-01 4.95626569e-01 -4.56033081e-01
-5.81974030e-01 -1.30356157e+00 -5.49710751e-01 4.43847239e-01
6.52671576e-01 -3.70623648e-01 -4.69226480e-01 -5.00524044e-01] | [8.24092960357666, 10.07502269744873] |
4b4dbfda-8f69-452a-bc49-af7e4cfe916f | towards-automated-covid-19-presence-and | 2305.08660 | null | https://arxiv.org/abs/2305.08660v1 | https://arxiv.org/pdf/2305.08660v1.pdf | Towards Automated COVID-19 Presence and Severity Classification | COVID-19 presence classification and severity prediction via (3D) thorax computed tomography scans have become important tasks in recent times. Especially for capacity planning of intensive care units, predicting the future severity of a COVID-19 patient is crucial. The presented approach follows state-of-theart techniques to aid medical professionals in these situations. It comprises an ensemble learning strategy via 5-fold cross-validation that includes transfer learning and combines pre-trained 3D-versions of ResNet34 and DenseNet121 for COVID19 classification and severity prediction respectively. Further, domain-specific preprocessing was applied to optimize model performance. In addition, medical information like the infection-lung-ratio, patient age, and sex were included. The presented model achieves an AUC of 79.0% to predict COVID-19 severity, and 83.7% AUC to classify the presence of an infection, which is comparable with other currently popular methods. This approach is implemented using the AUCMEDI framework and relies on well-known network architectures to ensure robustness and reproducibility. | ['Frank Kramer', 'Elisabeth André', 'Bernhard Bauer', 'Wolfgang Reif', 'Miriam Elia', 'Fabio Hellmann', 'Silvan Mertes', 'Niklas Schröter', 'Dominik Müller'] | 2023-05-15 | null | null | null | null | ['severity-prediction'] | ['computer-vision'] | [ 2.05654472e-01 8.57515819e-03 7.41700754e-02 -3.14836830e-01
-3.88747960e-01 -1.77542523e-01 2.58995861e-01 6.16333306e-01
-7.61725366e-01 8.50284219e-01 1.27350643e-01 -4.40912515e-01
-6.22631311e-01 -7.13187397e-01 -1.59264684e-01 -6.48883820e-01
-3.88249129e-01 1.10854232e+00 1.68487340e-01 6.68166056e-02
-8.18048194e-02 7.93771505e-01 -1.14407647e+00 3.70810419e-01
6.10960364e-01 1.13947535e+00 1.06627218e-01 9.01975453e-01
2.82363117e-01 8.92415643e-01 -4.43492442e-01 -1.60427600e-01
1.95754573e-01 -1.85937271e-01 -5.65558791e-01 -5.21161616e-01
-1.45173315e-02 -2.97468275e-01 -3.99698578e-02 2.12314516e-01
7.36890435e-01 -1.08559586e-01 1.04970205e+00 -7.65336752e-01
1.43395290e-01 4.00686920e-01 -1.58249259e-01 6.81688964e-01
3.01728025e-02 2.31781393e-01 6.03035212e-01 -5.67393243e-01
3.11910868e-01 6.47100747e-01 1.16408956e+00 3.43117595e-01
-8.15539062e-01 -7.36011326e-01 -4.05247182e-01 2.11134538e-01
-1.20084047e+00 1.94882676e-01 2.87417859e-01 -9.11495745e-01
1.02416301e+00 1.88509971e-01 7.67441034e-01 1.14319253e+00
5.85990608e-01 -1.55754253e-01 1.01266611e+00 -8.71812180e-02
8.09091032e-02 3.35111097e-02 7.18302354e-02 5.64189076e-01
4.90557224e-01 2.33122148e-02 -8.97678453e-03 -3.93123150e-01
7.45259762e-01 5.21121264e-01 -2.13431180e-01 -3.24866861e-01
-1.31173086e+00 7.70619094e-01 4.65644240e-01 5.03203511e-01
-9.13846314e-01 -2.05409661e-01 7.68919468e-01 -1.83064807e-02
4.33103532e-01 4.92334187e-01 -8.33076417e-01 -1.35687441e-02
-8.26800108e-01 7.23033249e-02 6.56305611e-01 3.05821806e-01
1.11974008e-01 3.91687378e-02 -3.89446288e-01 8.16620469e-01
2.02857792e-01 4.88460571e-01 5.61436415e-01 -5.00005960e-01
5.20644188e-01 7.10763693e-01 -2.56890766e-02 -6.44923329e-01
-9.98458445e-01 -7.78104961e-01 -1.40102887e+00 -7.24438131e-02
2.30607614e-01 -4.89678532e-01 -9.00027573e-01 1.47707629e+00
1.52692199e-01 4.15392697e-01 1.30723072e-02 7.75460660e-01
7.85416067e-01 1.98799774e-01 4.35268074e-01 -2.76002496e-01
1.34284508e+00 -4.70957011e-01 -3.36402655e-01 1.70831665e-01
5.71868896e-01 -4.69691426e-01 3.53269488e-01 2.80112565e-01
-9.28402781e-01 -3.73687804e-01 -7.94074714e-01 7.15658844e-01
-2.70246267e-01 7.12371767e-02 1.03310905e-01 7.37010121e-01
-9.48217392e-01 9.06206489e-01 -7.78862417e-01 -4.99454558e-01
3.17833275e-01 6.38830960e-01 -3.61925870e-01 -6.32394403e-02
-1.35047579e+00 1.35344040e+00 4.99298602e-01 5.67893609e-02
-7.68798530e-01 -9.65246737e-01 -6.11988425e-01 2.23052539e-02
-1.58210788e-02 -1.48453760e+00 9.62666273e-01 -3.76602054e-01
-1.13131499e+00 9.22932923e-01 2.75694311e-01 -6.48130596e-01
8.78231704e-01 -2.31108814e-01 -3.21975410e-01 3.49076748e-01
-2.78233618e-01 3.45653519e-02 7.60209858e-01 -7.95884907e-01
-4.47450459e-01 -5.77220082e-01 -2.58820325e-01 2.50390589e-01
-1.80643216e-01 6.33999407e-02 3.01583335e-02 -6.38654888e-01
-8.81701186e-02 -7.74491489e-01 -5.48412323e-01 -3.68339837e-01
-1.76913157e-01 -6.72648177e-02 5.66944242e-01 -8.19206119e-01
1.08102357e+00 -1.81991911e+00 -5.84444292e-02 4.24668401e-01
6.61637664e-01 6.40953660e-01 1.81570783e-01 4.11225557e-01
-3.05110186e-01 5.10624563e-03 -3.53357822e-01 -2.00516000e-01
-4.47985858e-01 -2.85520870e-02 3.96372736e-01 5.74863374e-01
2.33913690e-01 7.13397086e-01 -5.82408428e-01 -6.16671979e-01
7.80695021e-01 7.88832963e-01 -5.56316674e-01 6.25677109e-01
1.44836962e-01 8.30684960e-01 -4.58251625e-01 3.36362839e-01
5.60175300e-01 -4.98940766e-01 2.54223526e-01 -5.07779457e-02
1.01800874e-01 -1.57034218e-01 -5.58902800e-01 1.23129630e+00
-7.32902586e-01 7.29067028e-02 -2.61083688e-03 -1.14187396e+00
8.51849675e-01 6.93864822e-01 9.39192772e-01 -2.78604627e-01
7.26976633e-01 2.02526748e-01 1.42316118e-01 -7.78248250e-01
-2.24528819e-01 -4.53336477e-01 1.78925633e-01 3.48691314e-01
-1.19099461e-01 -4.27660644e-02 -4.85808067e-02 -2.27696985e-01
1.29104280e+00 -2.93492734e-01 5.56567490e-01 -3.00312340e-01
7.31807530e-01 -1.94414377e-01 2.74969846e-01 7.21873581e-01
-2.60530323e-01 7.34355032e-01 1.73318490e-01 -6.30169094e-01
-9.50144172e-01 -9.37643409e-01 -4.00766104e-01 6.46559656e-01
-3.95602912e-01 1.50847420e-01 -6.92157865e-01 -7.15439081e-01
-1.90359838e-02 5.83675563e-01 -7.89373100e-01 -2.14797724e-02
-7.27730095e-01 -9.82487381e-01 5.01353025e-01 6.82106316e-01
1.37009323e-01 -1.23035026e+00 -9.36651587e-01 3.51691961e-01
-1.65246889e-01 -9.36857283e-01 2.51257837e-01 3.68603408e-01
-1.02646220e+00 -1.61012292e+00 -1.13370717e+00 -6.46513522e-01
3.24781150e-01 -3.16268563e-01 1.13944209e+00 3.73274982e-01
-4.66462821e-01 1.98782250e-01 -3.51396501e-01 -5.28297722e-01
-5.04989505e-01 3.29610765e-01 8.22293460e-02 -1.28145784e-01
2.07406312e-01 -7.44755566e-01 -9.80193377e-01 2.39226907e-01
-7.27436364e-01 -1.63597405e-01 7.82224357e-01 7.33332157e-01
5.15729904e-01 -1.58127576e-01 6.59173250e-01 -9.53081787e-01
6.00342333e-01 -9.18531656e-01 -1.39217556e-01 1.19474322e-01
-8.37122977e-01 -3.91478479e-01 8.37160826e-01 -5.61016425e-02
-6.11840189e-01 -2.94950660e-02 -4.26749974e-01 -6.39121592e-01
-5.59808671e-01 3.28275234e-01 5.52170932e-01 2.77473450e-01
6.30363643e-01 3.00848205e-02 1.59272835e-01 -4.86910790e-01
-4.17328060e-01 7.61075437e-01 2.75735706e-01 -9.48420689e-02
5.32610834e-01 1.40899628e-01 3.73470426e-01 -5.74730873e-01
-8.25218379e-01 -7.42764473e-01 -9.62690234e-01 -1.85373887e-01
1.35501623e+00 -9.59920526e-01 -7.89670587e-01 5.81098258e-01
-1.06142128e+00 -8.77998173e-02 -1.59785509e-01 7.82349467e-01
-3.65588844e-01 1.22957237e-01 -5.06568193e-01 -5.88749230e-01
-9.26312566e-01 -1.07499385e+00 7.71805525e-01 -1.42187625e-01
-3.51995081e-01 -1.18638229e+00 4.20464456e-01 2.92136192e-01
9.31630731e-01 6.94486499e-01 1.14229548e+00 -1.19360185e+00
2.98162624e-02 -5.23647904e-01 -2.78077096e-01 5.48354805e-01
-1.78105030e-02 -1.99671388e-01 -8.14001501e-01 -2.48213574e-01
1.29416063e-02 -1.14434622e-01 6.98039472e-01 6.72820330e-01
1.38683295e+00 -9.57323909e-02 -2.41620705e-01 4.97425050e-01
1.39926648e+00 3.01896721e-01 3.16553295e-01 1.35240123e-01
5.86783230e-01 4.06492651e-01 1.75308764e-01 7.83862114e-01
4.13233370e-01 3.42876673e-01 7.73316741e-01 -3.06041747e-01
3.88947427e-02 2.60510623e-01 -4.32231337e-01 7.03814268e-01
-5.70186675e-01 -3.33577842e-01 -1.50815415e+00 3.15973371e-01
-1.44563890e+00 -8.42250049e-01 -5.20400226e-01 2.04598236e+00
3.23933095e-01 1.60796463e-01 2.41491526e-01 3.36887836e-01
8.75091910e-01 -1.10585801e-01 -4.32502002e-01 -4.31058943e-01
3.01721960e-01 6.17743850e-01 3.41081142e-01 -6.63471455e-03
-1.08904552e+00 1.58243775e-01 6.78103685e+00 3.17107439e-01
-1.15816474e+00 2.41254777e-01 7.80928612e-01 9.97077897e-02
3.01846743e-01 -6.23710454e-01 -3.49279255e-01 4.18065518e-01
1.11636233e+00 1.59080356e-01 5.04231788e-02 7.14171648e-01
2.72664011e-01 9.12496522e-02 -8.64508688e-01 7.42572069e-01
1.34325102e-01 -1.12103045e+00 -2.94779450e-01 -2.41418630e-01
4.74399835e-01 3.27395946e-01 -5.84772713e-02 3.30106527e-01
-2.14131251e-02 -1.11634219e+00 8.61681402e-02 6.09105647e-01
1.20625639e+00 -7.04815805e-01 1.49755156e+00 4.80623484e-01
-1.09766328e+00 -2.57427692e-01 6.78400919e-02 3.79942581e-02
3.73971164e-01 7.84516096e-01 -1.46686161e+00 6.13100708e-01
8.01245868e-01 5.54893196e-01 -4.48125929e-01 1.17265773e+00
3.93171376e-03 5.90637684e-01 -2.84283996e-01 -1.38265640e-02
5.91324409e-03 9.87972841e-02 3.44283134e-01 1.33865201e+00
4.34394807e-01 1.78490594e-01 1.62326247e-02 3.93970042e-01
1.36132345e-01 4.62492049e-01 -7.74194419e-01 3.67226690e-01
7.14014024e-02 1.13367498e+00 -5.45000911e-01 -4.82497603e-01
-1.28267318e-01 4.74702775e-01 6.09683758e-03 -1.17920242e-01
-9.78335977e-01 -9.96089429e-02 3.59327972e-01 5.41035533e-01
2.98558056e-01 3.14996004e-01 -4.36498463e-01 -7.75564253e-01
-4.06540334e-01 -6.61576152e-01 7.17047691e-01 -5.97090065e-01
-1.42480671e+00 1.03211856e+00 2.78677531e-02 -1.31745183e+00
-4.19406772e-01 -6.48010135e-01 -7.79581010e-01 8.53711843e-01
-1.59092093e+00 -9.20178056e-01 -6.15760922e-01 5.21071136e-01
3.12967479e-01 -2.39159092e-01 1.27947330e+00 3.24451059e-01
-6.17909610e-01 3.38140815e-01 1.80557705e-02 4.26079221e-02
4.59070235e-01 -1.04595459e+00 -7.73888230e-02 3.49294275e-01
-8.74869645e-01 4.16381985e-01 2.91069537e-01 -7.60411322e-01
-6.80855691e-01 -1.37395608e+00 1.01553047e+00 -5.12940884e-01
5.18116117e-01 1.64330661e-01 -7.18188643e-01 3.39087039e-01
1.14154279e-01 1.00283168e-01 1.02464378e+00 -1.24972798e-01
-2.15881065e-01 9.96823385e-02 -1.43479931e+00 1.57262720e-02
8.41278553e-01 -8.38822648e-02 -7.27734268e-01 3.89484167e-01
3.56967062e-01 -4.35723096e-01 -1.37064564e+00 1.12005329e+00
6.09764218e-01 -1.20177436e+00 1.03721201e+00 -7.51891613e-01
5.25974274e-01 1.27679810e-01 1.11737281e-01 -1.16561759e+00
-4.84775335e-01 4.83552366e-02 -4.40675765e-02 3.79986197e-01
4.35277045e-01 -7.32384682e-01 6.60352170e-01 3.32232267e-01
-9.45607722e-02 -1.11637127e+00 -8.27976286e-01 -3.92451465e-01
8.09762999e-02 -3.34205389e-01 6.27484560e-01 8.98278296e-01
-3.50022614e-01 2.77153909e-01 -2.58896083e-01 1.76450491e-01
4.22007203e-01 -2.34315634e-01 4.15299892e-01 -1.73540771e+00
-2.20287561e-01 -4.47879136e-01 -4.55242783e-01 1.38240531e-02
-4.21882868e-02 -9.24481332e-01 -3.01706403e-01 -1.79949498e+00
3.77861589e-01 -6.69546545e-01 -9.39159632e-01 2.91376948e-01
-2.30571225e-01 4.84179497e-01 -5.03883464e-03 1.33170635e-01
-3.69634360e-01 1.63693920e-01 1.17556059e+00 2.55956620e-01
-2.09584236e-01 5.11819243e-01 -1.25659868e-01 7.46816456e-01
1.17215312e+00 -6.72257662e-01 -3.09006661e-01 -1.11183248e-01
-7.66520277e-02 5.84065735e-01 3.75391245e-01 -1.24874151e+00
-2.02817485e-01 -1.42441377e-01 6.84571981e-01 -7.11890876e-01
3.11317742e-01 -1.10131323e+00 4.11364645e-01 1.04335332e+00
-1.56617612e-01 4.68107909e-01 1.79863244e-01 3.91839355e-01
-5.08149527e-02 -1.79874972e-01 9.66048121e-01 -2.41342559e-01
-1.32818565e-01 5.51246226e-01 -6.15872443e-01 1.29798159e-01
1.44215858e+00 -8.34491178e-02 -1.05042376e-01 -9.28543359e-02
-8.28340709e-01 -3.18993889e-02 8.89054090e-02 1.81493357e-01
5.40140212e-01 -7.55693197e-01 -9.52999651e-01 1.92396209e-01
1.54179528e-01 9.10685062e-02 6.50663972e-01 1.38241518e+00
-9.75150883e-01 3.91385466e-01 -3.59966159e-01 -7.35365629e-01
-1.43073547e+00 3.57866883e-01 5.44583440e-01 -7.61570752e-01
-6.55813158e-01 6.59064651e-01 2.47881580e-02 -3.21434200e-01
1.94115803e-01 -1.54770106e-01 -7.03332007e-01 -4.60277591e-03
4.37181801e-01 4.84595418e-01 4.40148324e-01 -5.29634595e-01
-5.85641623e-01 5.28168082e-01 4.95281629e-02 3.98820966e-01
1.64049113e+00 2.69223630e-01 -5.90128489e-02 2.12490335e-01
1.07400787e+00 -3.97924364e-01 -6.01114035e-01 -2.36374494e-02
-2.73551375e-01 -2.92170309e-02 -9.37539414e-02 -1.13813972e+00
-1.00926816e+00 1.04078579e+00 9.65282738e-01 5.62045500e-02
1.14136362e+00 -1.64280817e-01 6.06305003e-01 2.36224890e-01
9.90760699e-02 -5.33035159e-01 3.06426343e-02 5.84524035e-01
7.26248801e-01 -1.11122298e+00 -1.17351070e-01 -2.01307088e-01
-6.18732452e-01 9.64210391e-01 4.61508870e-01 -1.48866996e-01
9.81949151e-01 1.24351874e-01 2.57166773e-01 -2.84592718e-01
-5.90045512e-01 8.15614760e-02 2.89439440e-01 6.49084985e-01
4.49319035e-01 2.80261666e-01 -2.58090675e-01 6.89686537e-01
-6.91241622e-02 7.30545670e-02 9.34781507e-02 7.33225048e-01
-2.34268963e-01 -8.15148890e-01 -3.91112000e-01 1.04617786e+00
-9.71513927e-01 -1.70956731e-01 -8.97992253e-02 8.11217844e-01
4.09402788e-01 6.31112516e-01 -1.19481280e-01 -4.73604888e-01
4.91177708e-01 2.94054657e-01 3.02886933e-01 -6.90495372e-01
-1.14868701e+00 -3.79780501e-01 -7.13248178e-02 -3.20351154e-01
-4.27291304e-01 -4.50203419e-01 -1.07573271e+00 -1.12232380e-01
-2.01958209e-01 1.93335265e-02 8.28635514e-01 8.43467295e-01
3.77267122e-01 9.84778166e-01 6.47605360e-01 -6.60925984e-01
-5.86345494e-01 -1.21838152e+00 -4.22497600e-01 2.84169912e-01
4.52493578e-01 -6.89787388e-01 -4.33279127e-01 -1.78654775e-01] | [15.449665069580078, -1.8050180673599243] |
72a2bb61-7ab4-439a-9f7d-0cd47e76ac2f | entity-resolution-with-hierarchical-graph | null | null | https://dl.acm.org/doi/10.1145/3514221.3517872 | https://dl.acm.org/doi/pdf/10.1145/3514221.3517872 | Entity Resolution with Hierarchical Graph Attention Networks | Entity Resolution (ER) links entities that refer to the same real-world entity from different sources. Existing work usually takes pairs of entities as input and judges those pairs independently. However, there is often interdependence between different pairs of ER decisions, e.g., the entities from the same data source are usually semantically related to each other. Furthermore, current ER approaches are mainly based on attribute similarity comparison, but ignore interdependence between attributes. To address the limits of existing methods, we propose HierGAT, a new method for ER based on a Hierarchical Graph Attention Transformer Network, which can model and exploit the interdependence between different ER decisions. The benefit of our method comes from: 1) The graph attention network model for joint ER decisions; 2) The graph-attention capability to identify the discriminative words from attributes and find the most discriminative attributes. Furthermore, we propose to learn contextual embeddings to enrich word embeddings for better performance. The experimental results on publicly available benchmark datasets show that HierGAT outperforms DeepMatcher by up to 32.5% of F1 score and up to 8.7% of F1 score compared with Ditto. | ['Xinqiao Lv', 'Hai Jin', 'Gao Cong', 'Yuhong Gu', 'Dezhong Yao'] | 2022-06-01 | null | null | null | sigmod-pods-2022-6 | ['entity-resolution'] | ['natural-language-processing'] | [-3.89262199e-01 2.13245109e-01 -2.88669854e-01 -3.10033083e-01
-5.15363216e-01 -3.24945688e-01 4.59624439e-01 6.92596972e-01
-4.61759031e-01 5.20529091e-01 5.25986612e-01 7.80872777e-02
-1.52596742e-01 -1.05688608e+00 -5.42547464e-01 -4.29500431e-01
-2.64493134e-02 6.14568651e-01 3.61486524e-01 -3.05201948e-01
-1.65519416e-01 3.11427534e-01 -1.06717777e+00 1.71595216e-02
1.07867718e+00 9.75801229e-01 1.03453815e-01 -2.28224732e-02
-5.29839993e-01 6.46493077e-01 -4.82993662e-01 -8.09858561e-01
-1.19832065e-02 -1.27483711e-01 -6.58911645e-01 -4.42727476e-01
1.91452622e-01 -6.30341768e-02 -5.39657474e-01 1.16681409e+00
4.66016859e-01 1.68288261e-01 6.46892667e-01 -1.49173415e+00
-1.35007298e+00 8.62163067e-01 -6.27771974e-01 1.71823859e-01
1.59852296e-01 -2.90683895e-01 1.56625783e+00 -1.05223095e+00
5.72635651e-01 1.50023782e+00 5.26018798e-01 1.57968253e-01
-1.17691338e+00 -8.82681012e-01 5.53031564e-01 4.99016792e-01
-1.73523200e+00 -1.88668415e-01 6.89783812e-01 -1.76010117e-01
9.51256156e-01 -5.69041297e-02 5.34787953e-01 9.09633815e-01
-8.44747722e-02 7.21630812e-01 5.15972137e-01 -1.11126393e-01
-1.26643807e-01 1.58117205e-01 2.13605940e-01 5.78697383e-01
6.33683264e-01 -3.35292697e-01 -2.78032184e-01 -1.76598787e-01
5.35362363e-01 1.85966894e-01 -2.74479300e-01 -4.41226453e-01
-1.15748537e+00 9.08824325e-01 9.91926193e-01 5.12005448e-01
-2.80582637e-01 1.82961896e-01 3.74019593e-01 8.98678899e-02
3.49558383e-01 4.75102663e-01 -3.83497030e-01 4.12512600e-01
-2.48662382e-01 3.55084315e-02 5.73303938e-01 1.38181961e+00
9.69026923e-01 -2.19042242e-01 -3.44253778e-01 8.17920268e-01
5.76904774e-01 3.73880625e-01 3.70850444e-01 -1.90744936e-01
6.66315317e-01 1.11058843e+00 8.45982134e-02 -1.50205910e+00
-3.51425886e-01 -4.89265770e-01 -8.73719871e-01 -5.08124769e-01
-1.38009787e-01 -5.75874597e-02 -8.50001156e-01 1.89998949e+00
3.84785831e-01 3.29636514e-01 1.57111660e-01 8.50266635e-01
1.38745284e+00 6.37785673e-01 5.73605597e-01 1.33317903e-01
1.43036473e+00 -8.79384160e-01 -1.02016079e+00 -3.05219293e-01
8.85857046e-01 -5.04283667e-01 8.92184317e-01 -2.81492352e-01
-5.88002205e-01 -4.47058797e-01 -9.63044167e-01 -3.27581704e-01
-7.75973916e-01 2.10106209e-01 5.26175678e-01 1.22148059e-01
-7.45008886e-01 3.19032788e-01 -5.07908165e-01 -4.48596239e-01
3.90082777e-01 3.71816009e-01 -4.89190906e-01 -1.29916459e-01
-1.75710440e+00 8.55746567e-01 6.88006520e-01 -4.25456576e-02
-2.59176552e-01 -6.47513032e-01 -1.10071087e+00 4.82754737e-01
5.84147692e-01 -5.90612352e-01 6.31910563e-01 -8.26746762e-01
-7.19747424e-01 6.41391754e-01 -2.53849417e-01 -3.61970782e-01
1.34160221e-01 -3.78307581e-01 -9.08203781e-01 -1.38001993e-01
4.15232599e-01 5.36309063e-01 1.06340833e-01 -1.17521763e+00
-8.35627437e-01 -3.82207572e-01 1.13792792e-01 2.48963833e-01
-6.85564458e-01 -1.12122722e-01 -7.81098187e-01 -6.97747231e-01
-9.26647410e-02 -7.02834487e-01 -5.33999987e-02 -2.51459986e-01
-6.28937006e-01 -7.10618079e-01 7.82654583e-01 -5.70469022e-01
1.48583829e+00 -2.18748713e+00 2.74809897e-01 6.91188499e-02
6.37101829e-01 3.78495306e-01 -1.40548974e-01 5.63598216e-01
-1.83103830e-01 4.23984677e-01 6.08966872e-02 -2.11702392e-01
1.02546573e-01 4.51275371e-02 -2.11842462e-01 1.28690898e-01
3.70699942e-01 1.08687854e+00 -1.07640862e+00 -5.90197027e-01
-1.47755474e-01 4.93977070e-01 -1.07331440e-01 2.62274384e-01
5.62721193e-02 7.77347088e-02 -7.38741577e-01 5.02856910e-01
6.22982085e-01 -3.80317062e-01 3.90026718e-01 -6.80750251e-01
2.69387603e-01 3.29074144e-01 -1.18467975e+00 1.41841054e+00
-4.29006130e-01 4.54521418e-01 -3.19794416e-01 -8.48348737e-01
1.06701171e+00 3.15584719e-01 3.32482338e-01 -6.76151693e-01
4.44556959e-02 2.30293974e-01 -1.21386215e-01 -5.73635876e-01
5.45633197e-01 9.37763900e-02 -2.58621305e-01 -2.98150294e-02
1.94090009e-01 5.68677902e-01 2.85016298e-01 5.01218319e-01
8.81339729e-01 -1.47160485e-01 4.44845408e-01 -1.57132164e-01
4.25760776e-01 -2.97168255e-01 1.10298777e+00 3.46341759e-01
-3.49000990e-02 2.61641711e-01 6.41623735e-01 -3.86302769e-01
-7.01656699e-01 -9.60175157e-01 3.71996053e-02 1.01751614e+00
7.64102936e-01 -6.38545573e-01 -1.75354853e-01 -1.10374105e+00
2.42560521e-01 8.32336187e-01 -7.75286674e-01 -2.79176891e-01
-4.44075286e-01 -6.53890133e-01 3.73771369e-01 9.27833080e-01
4.94950026e-01 -8.26641917e-01 1.16685785e-01 3.65843415e-01
-3.29594165e-01 -1.37075043e+00 -5.12742519e-01 -1.65443525e-01
-4.35927123e-01 -1.12015581e+00 -4.66824353e-01 -8.79912257e-01
5.48523784e-01 3.36709291e-01 1.23085034e+00 3.98599565e-01
1.20556280e-01 2.53748205e-02 -5.10576904e-01 -2.79655963e-01
1.55172557e-01 3.99492800e-01 -1.99705269e-02 2.98350602e-01
8.99570346e-01 -4.96380299e-01 -4.92677361e-01 3.72450322e-01
-7.44893074e-01 1.59130301e-02 7.72094190e-01 8.59978378e-01
6.15459740e-01 -5.33162244e-02 8.34593773e-01 -1.18000507e+00
4.80613619e-01 -1.01975882e+00 -4.25991774e-01 5.76106787e-01
-8.21811140e-01 2.95165300e-01 6.78265572e-01 -4.42859143e-01
-9.61186409e-01 -2.94892937e-01 7.58515894e-02 -4.90308881e-01
1.65610071e-02 6.21168196e-01 -7.08455384e-01 4.06173110e-01
1.18132204e-01 -1.83760211e-01 -6.57885849e-01 -3.67451102e-01
4.58909333e-01 5.17674148e-01 1.67747170e-01 -4.34235632e-01
7.19216108e-01 2.71866500e-01 -1.78542852e-01 -3.86801213e-01
-8.53780985e-01 -4.85962182e-01 -4.87594694e-01 1.95817292e-01
9.77778256e-01 -1.12445772e+00 -4.45524395e-01 1.15778029e-01
-1.10191345e+00 1.08295046e-01 1.89990504e-03 4.94220942e-01
2.02228338e-01 3.46507341e-01 -5.91267526e-01 -3.73726785e-01
-3.54814500e-01 -9.57655132e-01 1.02781034e+00 5.51866233e-01
-5.91981746e-02 -1.02674019e+00 -1.09743357e-01 7.59337768e-02
3.18980187e-01 2.29855984e-01 1.23926568e+00 -1.17734969e+00
-6.08629823e-01 -1.79146469e-01 -6.97784662e-01 -1.70110211e-01
3.16991150e-01 -4.01720731e-03 -5.82176447e-01 -2.91805476e-01
-7.18602240e-01 2.12507453e-02 9.67754006e-01 -1.65397540e-01
9.75568771e-01 -3.10001254e-01 -8.67348015e-01 3.69898617e-01
1.49602377e+00 1.18230484e-01 5.70643306e-01 2.19696954e-01
1.29383957e+00 7.28202999e-01 6.89206660e-01 9.78248492e-02
9.26532507e-01 8.82274568e-01 4.40199256e-01 -2.64264941e-01
-1.80600286e-01 -6.24504626e-01 3.74288633e-02 9.47257698e-01
1.33601353e-01 -6.61970675e-01 -8.98737371e-01 1.00569916e+00
-2.05398512e+00 -5.90275645e-01 -2.74848908e-01 2.10293555e+00
4.56003934e-01 -2.96106166e-03 -5.73913679e-02 -3.73184264e-01
1.13931477e+00 7.85869062e-02 -6.20039523e-01 -6.07468747e-02
-1.54782295e-01 -1.53128356e-01 5.38405299e-01 2.85929948e-01
-1.11027443e+00 1.13728583e+00 4.30901098e+00 8.99604976e-01
-7.98134804e-01 8.47992972e-02 4.32200193e-01 1.44452780e-01
-6.58501804e-01 2.02512428e-01 -9.95617449e-01 6.37943089e-01
5.10417223e-01 -4.47864652e-01 -5.69975972e-02 7.08291352e-01
-4.07146007e-01 3.56025308e-01 -1.02122176e+00 8.63103747e-01
7.29094893e-02 -9.30629432e-01 2.53672749e-01 1.42625690e-01
5.52866340e-01 -1.07539974e-01 -2.19073370e-01 6.49270177e-01
6.51220441e-01 -1.04910147e+00 3.42652172e-01 5.06582201e-01
5.90141773e-01 -1.01299083e+00 9.98170197e-01 -4.42382433e-02
-1.84684885e+00 2.30342932e-02 -4.79832351e-01 4.78073329e-01
2.91602183e-02 6.99911296e-01 -6.31879687e-01 1.02132821e+00
7.99452305e-01 1.04732156e+00 -6.17974341e-01 9.11010385e-01
-6.48475528e-01 3.90378773e-01 -1.20326839e-01 -1.70703217e-01
1.07418656e-01 -2.48895139e-01 5.09304583e-01 1.32603681e+00
4.43240941e-01 6.83346540e-02 2.17934355e-01 8.21694970e-01
-5.67322135e-01 4.28121656e-01 -5.72259545e-01 -1.62823930e-01
1.01061893e+00 1.34205818e+00 -5.20318329e-01 -3.36017430e-01
-7.18768120e-01 9.07360077e-01 9.10606146e-01 4.67087299e-01
-8.81329596e-01 -8.68108392e-01 8.41646731e-01 2.22813785e-02
6.26008749e-01 -8.27279091e-02 8.38978067e-02 -1.21177924e+00
1.47137612e-01 -3.92223209e-01 8.03963244e-01 -6.78911746e-01
-1.67143404e+00 6.21500671e-01 -1.11411653e-01 -1.04147232e+00
2.19857574e-01 -2.38164559e-01 -6.53622329e-01 8.89823079e-01
-1.70526814e+00 -1.32877707e+00 -3.23340416e-01 3.52841049e-01
2.32265443e-01 -1.30250588e-01 7.83519149e-01 5.08127332e-01
-7.98242688e-01 6.73014700e-01 5.48243187e-02 6.68549895e-01
8.85227561e-01 -1.48515534e+00 6.38463259e-01 7.83344984e-01
1.74902409e-01 7.19970942e-01 3.07902545e-01 -7.47140765e-01
-1.17765498e+00 -1.41555142e+00 1.40111053e+00 -2.46318400e-01
7.35517740e-01 -3.59178752e-01 -1.28970706e+00 9.26684916e-01
3.13147277e-01 2.79387832e-01 8.85458529e-01 4.71271008e-01
-5.99172235e-01 -3.41229975e-01 -9.31893229e-01 6.62341714e-01
1.21127963e+00 -4.06866431e-01 -7.93436944e-01 1.19801149e-01
9.75298464e-01 -1.00762159e-01 -1.20373333e+00 5.13342738e-01
1.75831690e-01 -5.33443928e-01 9.46052969e-01 -7.40037680e-01
2.93870270e-01 -5.55241823e-01 -9.34430584e-02 -1.56527174e+00
-6.35452688e-01 -1.52835041e-01 -3.22843820e-01 1.97792113e+00
5.07760108e-01 -8.13255966e-01 1.81925014e-01 6.44670546e-01
1.21171169e-01 -7.87208974e-01 -7.88218975e-01 -7.75654435e-01
-4.30854484e-02 1.32798776e-01 1.10111499e+00 1.35557282e+00
-2.26582244e-01 8.38990390e-01 -2.07733721e-01 5.64555883e-01
4.47100520e-01 3.92722517e-01 5.49437940e-01 -1.50307965e+00
7.13018626e-02 -4.49637622e-01 -6.38708532e-01 -1.00231397e+00
3.80744517e-01 -1.04434943e+00 -1.79475218e-01 -1.93018508e+00
2.76480436e-01 -8.13251436e-01 -7.34334826e-01 6.34154916e-01
-7.38393545e-01 -1.49189934e-01 1.22302547e-01 1.15455709e-01
-7.68184543e-01 8.71448755e-01 8.14866781e-01 -1.93747208e-01
-2.03782605e-04 -5.55396378e-01 -1.00717902e+00 4.90942687e-01
6.11916959e-01 -7.26469338e-01 -3.90406936e-01 -7.16244161e-01
3.94664586e-01 -1.59838691e-01 1.91402093e-01 -6.69485927e-01
2.68955052e-01 -9.64025185e-02 2.90664077e-01 -3.38329166e-01
7.13163018e-02 -9.48275328e-01 6.88118264e-02 2.55087931e-02
-3.03020149e-01 2.32614577e-01 -8.23839847e-03 1.02360785e+00
-3.80353719e-01 -3.71743664e-02 3.51560444e-01 1.82500929e-01
-1.00963843e+00 5.69643915e-01 1.84104294e-01 4.54196751e-01
1.00329256e+00 1.80613458e-01 -3.79764944e-01 -1.85718030e-01
-5.78669190e-01 7.22190797e-01 3.30584705e-01 8.12906325e-01
4.24650520e-01 -1.79203522e+00 -8.48292649e-01 -1.61802813e-01
5.47182977e-01 1.36331528e-01 8.17308798e-02 6.55460358e-01
4.93588597e-02 1.80867836e-01 6.54283389e-02 -2.80474454e-01
-1.20836914e+00 7.67784894e-01 2.68390387e-01 -4.89261866e-01
-5.60258627e-01 8.42599928e-01 5.78998864e-01 -3.67307991e-01
5.20658866e-02 -1.22440636e-01 -6.21162295e-01 2.58809447e-01
4.10176724e-01 2.13632420e-01 1.84611939e-02 -8.67907584e-01
-5.32487094e-01 5.67057014e-01 -3.43588889e-01 3.41582447e-01
1.22820461e+00 -1.71369940e-01 -6.08006753e-02 3.89467478e-01
1.17737687e+00 2.02509373e-01 -8.38688254e-01 -7.73604214e-01
2.58313328e-01 -5.31000018e-01 9.03672352e-02 -5.47351837e-01
-1.41698027e+00 7.63734221e-01 3.11088681e-01 3.41693223e-01
1.03575099e+00 2.13800207e-01 1.01298785e+00 2.70910949e-01
3.02800119e-01 -8.75763178e-01 -1.46925062e-01 2.55022079e-01
7.47028291e-01 -1.16732168e+00 -5.89559115e-02 -6.81029856e-01
-8.06366026e-01 9.75366890e-01 8.94762993e-01 -4.44354527e-02
5.69725454e-01 -6.08248897e-02 -2.07690552e-01 -3.44860315e-01
-5.47615230e-01 -4.74461287e-01 4.91803646e-01 5.03616869e-01
5.23666859e-01 1.80092141e-01 -3.83178771e-01 9.24791634e-01
3.22944134e-01 -3.42083424e-01 1.55694276e-01 5.07292986e-01
-2.06025854e-01 -1.15006399e+00 1.85806438e-01 3.25666279e-01
-3.69992077e-01 -3.74891400e-01 -5.29103756e-01 8.72298121e-01
1.57986000e-01 8.76147628e-01 6.06515333e-02 -4.02641863e-01
4.70515966e-01 -8.11482817e-02 -1.60576478e-01 -5.84051549e-01
-4.87761229e-01 -2.57917881e-01 1.62426323e-01 -3.26222420e-01
-2.95980453e-01 -4.44510520e-01 -1.38942862e+00 -2.57462919e-01
-5.32351494e-01 2.78375030e-01 2.54565358e-01 8.54836464e-01
8.82804155e-01 6.86161697e-01 5.21012366e-01 -7.95351490e-02
-3.80561166e-02 -9.08652246e-01 -5.23548424e-01 6.67692900e-01
6.58265352e-02 -9.82027173e-01 -1.64697066e-01 -3.33801627e-01] | [8.889159202575684, 8.148289680480957] |
9b6cec66-28ee-4cf4-b346-45af79bd9d17 | on-the-local-cache-update-rules-in-streaming | 2303.16340 | null | https://arxiv.org/abs/2303.16340v1 | https://arxiv.org/pdf/2303.16340v1.pdf | On the Local Cache Update Rules in Streaming Federated Learning | In this study, we address the emerging field of Streaming Federated Learning (SFL) and propose local cache update rules to manage dynamic data distributions and limited cache capacity. Traditional federated learning relies on fixed data sets, whereas in SFL, data is streamed, and its distribution changes over time, leading to discrepancies between the local training dataset and long-term distribution. To mitigate this problem, we propose three local cache update rules - First-In-First-Out (FIFO), Static Ratio Selective Replacement (SRSR), and Dynamic Ratio Selective Replacement (DRSR) - that update the local cache of each client while considering the limited cache capacity. Furthermore, we derive a convergence bound for our proposed SFL algorithm as a function of the distribution discrepancy between the long-term data distribution and the client's local training dataset. We then evaluate our proposed algorithm on two datasets: a network traffic classification dataset and an image classification dataset. Our experimental results demonstrate that our proposed local cache update rules significantly reduce the distribution discrepancy and outperform the baseline methods. Our study advances the field of SFL and provides practical cache management solutions in federated learning. | ['Jie Xu', 'Jieming Bian', 'Heqiang Wang'] | 2023-03-28 | null | null | null | null | ['traffic-classification'] | ['miscellaneous'] | [-3.27748597e-01 -6.90418482e-01 -6.88142180e-01 -5.73202550e-01
-7.59088695e-01 -3.37110668e-01 2.47004390e-01 4.34028178e-01
-4.45350051e-01 8.86492491e-01 1.74086824e-01 -3.21252227e-01
-4.32944715e-01 -9.77191269e-01 -7.84087718e-01 -7.88684666e-01
-2.25268334e-01 4.10795301e-01 8.37058187e-01 2.65363846e-02
2.75228941e-03 3.94380361e-01 -1.95611882e+00 5.12028873e-01
8.48632574e-01 1.29320288e+00 9.29448158e-02 4.86045450e-01
-7.79220164e-01 1.22903061e+00 -7.40429759e-01 -1.29196793e-01
1.49318680e-01 -2.78214306e-01 -7.33133674e-01 -1.95891708e-01
5.94517112e-01 -6.50179863e-01 -5.87589324e-01 8.50854993e-01
4.14980322e-01 1.53352723e-01 1.57530680e-01 -1.47172785e+00
-1.98352933e-01 8.83970499e-01 -5.51483393e-01 7.83372581e-01
1.89107703e-03 7.85337389e-03 5.94455063e-01 -4.34198618e-01
4.73124772e-01 9.78206694e-01 4.40671206e-01 5.31249344e-01
-7.92633891e-01 -8.76013577e-01 4.51708108e-01 4.01158661e-01
-9.86087143e-01 -5.32091618e-01 5.17619133e-01 -1.40220582e-01
4.88109529e-01 3.38088244e-01 3.48416835e-01 6.23803020e-01
1.70361415e-01 8.46092165e-01 8.39017689e-01 -2.69779205e-01
7.61212468e-01 -2.77552437e-02 4.34249520e-01 3.34105015e-01
3.39743644e-01 -3.46524902e-02 -1.17240942e+00 -7.54116893e-01
5.50068259e-01 2.53599763e-01 -3.90484661e-01 -3.48287195e-01
-9.00659561e-01 7.33588040e-01 2.92678364e-02 8.91422182e-02
-3.55303198e-01 1.66903302e-01 8.07589889e-01 6.37769401e-01
6.30551159e-01 -7.45013714e-01 -5.32297492e-01 -2.08559975e-01
-1.14408183e+00 2.11169541e-01 1.04346883e+00 1.08257592e+00
7.34213710e-01 8.67151748e-03 -1.54135317e-01 8.65083814e-01
7.18142539e-02 3.77718866e-01 9.21299875e-01 -1.00647986e+00
6.16002023e-01 4.24626470e-01 3.37615423e-02 -6.09992564e-01
-6.29313290e-02 -5.99451840e-01 -6.29045427e-01 -1.81395426e-01
4.39480424e-01 2.26026382e-02 -1.75430961e-02 1.64274061e+00
8.33482921e-01 5.69939137e-01 7.24715833e-03 9.19773042e-01
4.79818583e-01 5.34063637e-01 -9.61827405e-04 -7.01291323e-01
8.70170414e-01 -9.41931784e-01 -8.11459541e-01 5.53937733e-01
9.93301809e-01 -4.44428027e-01 8.73499572e-01 4.13023084e-01
-1.19060278e+00 -2.86533833e-01 -7.03140795e-01 3.44567657e-01
3.81555222e-02 -3.97657603e-01 4.95724678e-01 5.71360469e-01
-8.73579502e-01 4.75791514e-01 -7.99978018e-01 -3.43857408e-01
5.09576380e-01 2.33812258e-01 -9.07473490e-02 -1.31780311e-01
-9.00170982e-01 -2.12991133e-01 2.40734160e-01 -5.44527113e-01
-7.80820787e-01 -1.12813187e+00 -1.66797101e-01 1.85691833e-01
3.70101094e-01 -6.88716650e-01 1.53423738e+00 -9.13949132e-01
-1.11016631e+00 2.92569488e-01 -1.13480970e-01 -6.22671902e-01
4.51808453e-01 -4.53418493e-02 -5.20009100e-01 1.55762717e-01
-2.06512094e-01 -3.04896366e-02 5.63090265e-01 -1.08248293e+00
-1.20098829e+00 -5.01624286e-01 -2.78062403e-01 4.81650308e-02
-9.69356418e-01 -7.05085024e-02 -1.87680155e-01 -6.06908619e-01
-1.81655526e-01 -2.25408792e-01 1.42208571e-02 2.74092108e-01
4.15983975e-01 -4.50383753e-01 1.30163109e+00 4.47015725e-02
1.67501426e+00 -2.35355353e+00 -4.10221100e-01 7.38417357e-02
1.26936674e-01 2.25634381e-01 -2.22127467e-01 3.08473974e-01
5.06810308e-01 -2.16809526e-01 1.02188244e-01 -8.55131894e-02
-2.51604080e-01 5.85680604e-01 -6.46305442e-01 4.46933657e-01
-6.27883732e-01 2.16811344e-01 -1.18798602e+00 -7.17262983e-01
-2.88151741e-01 2.33398601e-01 -7.04812706e-01 4.26690847e-01
-4.65431750e-01 1.35963887e-01 -5.84745526e-01 4.69803214e-01
1.02351332e+00 -4.26201135e-01 2.68700182e-01 -1.25590146e-01
-1.76587746e-01 2.28097767e-01 -1.26935804e+00 1.51658046e+00
-4.95253265e-01 -6.81794388e-03 4.84217137e-01 -8.14832509e-01
1.12793612e+00 3.55879158e-01 9.43334222e-01 -8.22392523e-01
-2.34262589e-02 5.26981592e-01 -6.48733556e-01 -3.83005530e-01
4.32499915e-01 7.77387694e-02 3.30273092e-01 1.07450843e+00
5.18117892e-03 8.65099132e-01 1.02425754e-01 4.44246322e-01
1.36605549e+00 -2.42223889e-01 -4.96222218e-03 -3.92076463e-01
7.00147867e-01 -3.10692936e-03 8.14158499e-01 8.48842800e-01
-5.75702608e-01 5.92282368e-03 4.28271174e-01 -8.30729008e-01
-7.49734461e-01 -9.62743998e-01 -4.09728400e-02 1.55830240e+00
3.52389008e-01 -5.71127415e-01 -6.43054545e-01 -1.02502394e+00
3.97583306e-01 5.07441998e-01 -1.28037050e-01 -4.95567843e-02
-6.63205802e-01 -3.81274730e-01 3.67947191e-01 3.65438282e-01
6.81628764e-01 -6.52945220e-01 -7.87010014e-01 3.92398536e-01
-9.62957144e-02 -8.98986757e-01 -9.23863471e-01 -2.13456899e-02
-1.30607879e+00 -1.09495783e+00 -9.87286717e-02 -5.12388051e-01
1.97435364e-01 7.82519042e-01 1.05369174e+00 1.60039842e-01
-2.25572139e-02 6.10371768e-01 -4.22008872e-01 5.64890029e-03
-2.94562578e-01 1.62844867e-01 5.67472801e-02 2.81477213e-01
4.29054797e-01 -6.64282918e-01 -7.51848698e-01 5.75550199e-01
-1.37840831e+00 -3.37494493e-01 -9.27227512e-02 8.44250679e-01
7.20954478e-01 1.03191864e-02 1.05362344e+00 -1.18532157e+00
5.09822369e-01 -1.17850947e+00 -4.71881449e-01 3.67199361e-01
-8.28000307e-01 -9.66497511e-02 9.38612103e-01 -6.17610574e-01
-1.09784198e+00 -1.54122651e-01 2.29388222e-01 -7.77988970e-01
-1.03807293e-01 1.60767540e-01 -1.55386046e-01 1.42521307e-01
5.37779152e-01 3.56850713e-01 2.17804253e-01 -8.31194401e-01
2.22443447e-01 1.02754295e+00 6.93475723e-01 -9.13796186e-01
2.36489385e-01 6.73830032e-01 -2.40140080e-01 -4.94178623e-01
-7.26703584e-01 -5.74362159e-01 -3.73751581e-01 -3.42889905e-01
-1.02682993e-01 -1.12416351e+00 -7.60420203e-01 4.68661785e-01
-7.67903805e-01 -3.37822080e-01 -6.74215555e-01 3.61434102e-01
-5.78306913e-01 3.28959614e-01 -6.36276066e-01 -6.94594026e-01
-6.85506821e-01 -7.83452749e-01 6.36526227e-01 1.65751696e-01
3.21068913e-01 -9.92020845e-01 1.69345334e-01 1.66209176e-01
8.65801811e-01 1.64924283e-02 8.91621649e-01 -8.74636114e-01
-6.53802216e-01 1.85371131e-01 -1.26059860e-01 6.41493797e-02
2.57590920e-01 -3.35638195e-01 -6.48267567e-01 -8.48593831e-01
2.06976742e-01 -3.72033149e-01 5.97243488e-01 -8.62430334e-02
1.45090520e+00 -6.73788726e-01 -2.99410701e-01 8.17193866e-01
1.54558504e+00 -1.23142779e-01 1.17052227e-01 2.96053737e-01
3.39905024e-01 2.13174477e-01 6.77716494e-01 1.28056014e+00
4.30430591e-01 4.25824404e-01 5.16189873e-01 5.64371586e-01
-2.54053622e-01 -2.35921487e-01 5.76945841e-01 1.17259693e+00
6.12875760e-01 -3.72314125e-01 -4.32189345e-01 6.39099360e-01
-1.98207676e+00 -9.26066995e-01 -2.27687314e-01 2.55359221e+00
8.93939435e-01 -1.00611895e-01 4.66154516e-01 1.63361877e-01
7.31318414e-01 1.30557433e-01 -9.29042637e-01 -4.44106936e-01
-1.50348589e-01 6.15765043e-02 6.48144543e-01 7.47001842e-02
-6.09999299e-01 4.51922327e-01 5.91662741e+00 1.03378427e+00
-1.39923465e+00 4.58929151e-01 3.91079277e-01 -7.26166487e-01
-6.04112387e-01 -2.23425642e-01 -8.10444117e-01 9.84640539e-01
1.41173434e+00 -5.56845725e-01 3.82606596e-01 9.35010254e-01
1.39823005e-01 -7.22352788e-02 -9.45127487e-01 1.01074696e+00
-1.66109070e-01 -1.47741127e+00 3.91233325e-01 3.95957604e-02
6.03104532e-01 2.63696492e-01 -1.08823419e-01 8.15487280e-02
1.89838454e-01 -2.66510576e-01 6.02770925e-01 6.46273911e-01
6.92590237e-01 -9.83771145e-01 4.70229387e-01 6.38584375e-01
-1.35407889e+00 -5.26206434e-01 -4.08608139e-01 2.65651613e-01
-1.40466705e-01 8.90701115e-01 -7.94789195e-02 4.06426251e-01
8.79745960e-01 4.21267331e-01 -4.53928202e-01 1.11325240e+00
8.48519683e-01 1.06213892e+00 -4.15942073e-01 1.46388620e-01
-1.30400032e-01 8.20370540e-02 5.29827654e-01 9.79319751e-01
3.36105108e-01 1.90816894e-01 2.90168315e-01 3.84410232e-01
-3.44587356e-01 4.55346882e-01 -2.21062914e-01 5.78159213e-01
1.13719785e+00 9.08749819e-01 -2.17619702e-01 -3.47285748e-01
-6.71152472e-01 4.91972387e-01 4.58405465e-01 1.12825297e-01
-6.48172498e-01 -2.32103229e-01 9.16073799e-01 4.60215509e-01
5.00661552e-01 -8.69858116e-02 4.69585396e-02 -1.15093029e+00
7.24811554e-02 -8.91871214e-01 1.01606774e+00 6.64807856e-02
-1.59778821e+00 5.29913306e-01 -8.83838683e-02 -1.14018250e+00
-1.36418650e-02 1.84863850e-01 -6.47948027e-01 3.04325014e-01
-1.52682400e+00 -8.03739190e-01 -4.45150882e-01 1.12679768e+00
5.58193564e-01 -3.62026006e-01 4.53267276e-01 7.82300651e-01
-5.18981993e-01 1.08045220e+00 7.05460370e-01 -3.93392473e-01
8.26514721e-01 -6.16801858e-01 -1.65978760e-01 4.10642177e-01
-1.89572364e-01 3.65899116e-01 3.44499886e-01 -5.32344401e-01
-1.67408669e+00 -1.36779034e+00 5.57693064e-01 2.91521728e-01
5.86847186e-01 -4.07636538e-02 -1.43648160e+00 3.28187406e-01
-1.59415439e-01 8.50858748e-01 8.92944872e-01 -3.67967337e-01
-5.57864070e-01 -8.56587052e-01 -1.53031218e+00 -1.37637928e-01
9.85320866e-01 -2.25790471e-01 1.75958559e-01 2.83543706e-01
8.10100198e-01 -1.49791494e-01 -1.16155064e+00 2.31941387e-01
4.85764444e-01 -1.27215755e+00 4.35648382e-01 -6.82127595e-01
-1.54909074e-01 -1.48068622e-01 -5.11028886e-01 -8.35510910e-01
-1.81661487e-01 -6.59275353e-01 -8.77984226e-01 1.46377409e+00
-2.67017603e-01 -6.66228592e-01 1.08162785e+00 4.22761053e-01
7.80813396e-02 -1.06858790e+00 -1.28944969e+00 -1.02125740e+00
1.13333106e-01 -1.42681837e-01 1.19847775e+00 6.92007065e-01
-6.19084053e-02 -3.35200459e-01 -1.79429531e-01 7.28655532e-02
9.82394397e-01 2.48985812e-01 8.92099619e-01 -1.08362317e+00
-3.67174923e-01 -6.58991188e-02 -2.24088561e-02 -1.27730930e+00
1.95450380e-01 -8.98092389e-01 -5.20217836e-01 -9.54419374e-01
3.82994771e-01 -9.91835475e-01 -5.78660667e-01 2.87209332e-01
1.15201242e-01 -3.64224970e-01 2.74472445e-01 6.94565892e-01
-1.12322974e+00 6.25060260e-01 8.92594814e-01 1.03139564e-01
-1.85390964e-01 4.82759066e-02 -2.07776487e-01 1.53442964e-01
6.14936531e-01 -5.76322675e-01 -5.93138576e-01 -5.37813485e-01
-2.00685024e-01 3.76500189e-01 1.39600992e-01 -1.13696110e+00
7.63038337e-01 -4.81246680e-01 -3.24952081e-02 -8.06091189e-01
-2.96906829e-01 -9.56457198e-01 1.96710244e-01 4.99989480e-01
-4.80860054e-01 2.02694908e-01 -1.36985984e-02 8.43814373e-01
-1.74964771e-01 7.94667974e-02 8.72527659e-01 2.39421278e-01
-4.18212116e-01 6.44056678e-01 -3.34587365e-01 3.49131435e-01
1.13743448e+00 4.14305590e-02 -6.68260336e-01 -3.33155423e-01
-2.76640207e-01 5.81445277e-01 5.32746255e-01 3.59712392e-01
6.51678383e-01 -1.50507188e+00 -6.28349483e-01 4.85382229e-01
-1.41208479e-02 9.34028476e-02 4.62517172e-01 6.66135192e-01
-3.86263102e-01 1.42188609e-01 -6.74603879e-02 -7.14728177e-01
-1.26227045e+00 6.67268813e-01 3.13666046e-01 -4.28640433e-02
-6.58578932e-01 4.76260990e-01 -2.74653465e-01 -2.41865054e-01
7.45255411e-01 7.51915127e-02 2.34079346e-01 -2.08577111e-01
8.40611339e-01 8.85556817e-01 1.75052419e-01 -3.67069453e-01
-1.66181266e-01 1.30014285e-01 -2.07263127e-01 3.30475897e-01
1.27595115e+00 -4.36188579e-01 -2.89800733e-01 5.68412542e-01
1.36611199e+00 -1.84521258e-01 -1.19519961e+00 -9.31674838e-01
-6.21706955e-02 -9.35107231e-01 2.35088974e-01 -3.02324086e-01
-1.59720898e+00 3.93040150e-01 8.60733867e-01 2.71288399e-02
1.35195422e+00 -1.19815432e-01 1.60512471e+00 7.12815672e-02
8.65284920e-01 -9.89188015e-01 4.18660156e-02 4.12977695e-01
1.07358262e-01 -7.88627744e-01 -6.49456456e-02 -9.47176814e-02
-2.99412102e-01 1.26354647e+00 9.12910879e-01 1.76186621e-01
1.03441501e+00 5.42653739e-01 -2.33238167e-03 1.25854313e-01
-1.66680181e+00 3.36981565e-01 -3.24638128e-01 3.45816463e-01
2.97298521e-01 -1.46215424e-01 -5.36351621e-01 6.32976949e-01
1.60446197e-01 4.02996302e-01 4.61150885e-01 1.23259974e+00
-7.48638570e-01 -1.24754238e+00 -4.56629097e-01 6.80016041e-01
-4.09087479e-01 5.92475653e-01 2.72487521e-01 2.19944313e-01
1.34304389e-01 9.56624210e-01 6.45078003e-01 -2.20935240e-01
2.30790526e-02 7.00744241e-02 2.55375803e-01 -9.22776014e-02
-7.88051248e-01 7.37558827e-02 -4.26068366e-01 -6.44900858e-01
-1.86249986e-01 -5.47897875e-01 -1.44941688e+00 -8.05563271e-01
-3.79233927e-01 5.70274889e-01 6.81750953e-01 5.59867382e-01
5.78531206e-01 2.35597610e-01 1.20120621e+00 -4.87353243e-02
-1.19485068e+00 -4.38657045e-01 -8.71735811e-01 4.73700851e-01
4.97471750e-01 -3.39010507e-01 -4.68337297e-01 -9.80011672e-02] | [5.845589637756348, 6.2722249031066895] |
a1efd442-d002-4256-952c-8afca91c229c | share-with-thy-neighbors-single-view | 2204.10310 | null | https://arxiv.org/abs/2204.10310v3 | https://arxiv.org/pdf/2204.10310v3.pdf | Share With Thy Neighbors: Single-View Reconstruction by Cross-Instance Consistency | Approaches for single-view reconstruction typically rely on viewpoint annotations, silhouettes, the absence of background, multiple views of the same instance, a template shape, or symmetry. We avoid all such supervision and assumptions by explicitly leveraging the consistency between images of different object instances. As a result, our method can learn from large collections of unlabelled images depicting the same object category. Our main contributions are two ways for leveraging cross-instance consistency: (i) progressive conditioning, a training strategy to gradually specialize the model from category to instances in a curriculum learning fashion; and (ii) neighbor reconstruction, a loss enforcing consistency between instances having similar shape or texture. Also critical to the success of our method are: our structured autoencoding architecture decomposing an image into explicit shape, texture, pose, and background; an adapted formulation of differential rendering; and a new optimization scheme alternating between 3D and pose learning. We compare our approach, UNICORN, both on the diverse synthetic ShapeNet dataset - the classical benchmark for methods requiring multiple views as supervision - and on standard real-image benchmarks (Pascal3D+ Car, CUB) for which most methods require known templates and silhouette annotations. We also showcase applicability to more challenging real-world collections (CompCars, LSUN), where silhouettes are not available and images are not cropped around the object. | ['Mathieu Aubry', 'Alexei A. Efros', 'Matthew Fisher', 'Tom Monnier'] | 2022-04-21 | null | null | null | null | ['single-view-3d-reconstruction', '3d-object-reconstruction', '3d-object-reconstruction-from-a-single-image'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 3.24016750e-01 1.28977790e-01 4.83337753e-02 -5.01654327e-01
-9.68034565e-01 -7.92667747e-01 9.10515606e-01 -8.10600668e-02
-2.15286195e-01 5.61084807e-01 -2.13711902e-01 7.99975172e-02
1.61302656e-01 -5.05462348e-01 -1.25328755e+00 -6.98816299e-01
1.57038346e-01 1.00700307e+00 5.94106197e-01 -1.31795928e-01
1.69742063e-01 8.21881771e-01 -1.70513463e+00 3.85347754e-01
3.46799016e-01 1.17833877e+00 4.24623825e-02 5.00391543e-01
2.10361272e-01 6.72559738e-01 -1.85484305e-01 -3.64969075e-01
7.00873017e-01 -2.38772184e-01 -8.01159859e-01 8.06277454e-01
1.26006591e+00 -2.71883994e-01 5.16689662e-03 7.62307167e-01
3.27761292e-01 8.92503262e-02 7.98767984e-01 -1.15730309e+00
-3.65052342e-01 -1.41106129e-01 -6.63262486e-01 -8.85684043e-02
3.31846058e-01 2.66066402e-01 9.28315699e-01 -1.02961326e+00
1.03511107e+00 1.17608345e+00 7.24546790e-01 4.94803578e-01
-1.77088249e+00 -2.23241791e-01 3.88469726e-01 -2.62970515e-02
-1.23857570e+00 -4.20173585e-01 8.04645360e-01 -6.82744861e-01
8.33562851e-01 1.56541154e-01 6.96868002e-01 1.24425972e+00
-7.43438005e-02 6.33694589e-01 1.38199222e+00 -1.48364782e-01
2.57819384e-01 3.89557838e-01 -1.34153634e-01 7.36667633e-01
7.60630071e-02 4.05141383e-01 -2.85996556e-01 -9.16833207e-02
8.85717273e-01 4.42458056e-02 -3.88668925e-01 -1.50888276e+00
-1.11471546e+00 4.76731807e-01 2.60088801e-01 -2.39656374e-01
-1.51798353e-01 1.61368288e-02 3.33245724e-01 2.79946834e-01
3.74420702e-01 2.13870823e-01 -7.64475882e-01 3.49680364e-01
-6.96766376e-01 3.31811279e-01 8.65007341e-01 1.41556513e+00
9.95782137e-01 1.07073151e-01 2.37714261e-01 5.80722272e-01
5.18391505e-02 6.74351156e-01 7.21955746e-02 -1.07579923e+00
3.94868135e-01 3.41728270e-01 7.48461634e-02 -7.39094973e-01
-6.37937635e-02 -2.84073740e-01 -5.15038073e-01 5.38366199e-01
4.73515153e-01 3.01960319e-01 -1.11078084e+00 1.73056543e+00
5.54516852e-01 3.59971821e-01 5.10160960e-02 7.70759940e-01
6.87473178e-01 2.09921241e-01 -3.56247127e-01 3.96451168e-02
1.12325561e+00 -9.43823814e-01 -4.79280315e-02 -2.60068327e-01
2.68321246e-01 -7.38642991e-01 9.31231856e-01 6.16891444e-01
-1.25699496e+00 -6.05108321e-01 -8.62268448e-01 -1.63116664e-01
-2.68220276e-01 -6.87395111e-02 3.30765992e-01 4.36313033e-01
-9.74145293e-01 7.34470189e-01 -7.97280550e-01 -1.21384285e-01
5.01646638e-01 2.65029490e-01 -7.60583758e-01 -3.35310996e-01
-3.74052137e-01 8.56730998e-01 2.80782342e-01 -1.03682630e-01
-1.22141194e+00 -1.00967908e+00 -1.21587157e+00 -2.43393153e-01
4.55707550e-01 -7.23025799e-01 9.07189608e-01 -1.30441582e+00
-1.38893104e+00 1.45597863e+00 2.04119295e-01 -2.59849101e-01
7.55969167e-01 1.08044006e-01 7.25759640e-02 1.00249074e-01
-9.42411795e-02 7.98694968e-01 1.16574895e+00 -1.86713445e+00
-3.65948409e-01 -6.21786416e-01 1.66716650e-01 2.64936477e-01
3.83738607e-01 -4.41180497e-01 -4.86371964e-01 -4.52130169e-01
1.69564784e-01 -1.11012030e+00 -1.08281642e-01 2.58452594e-01
-2.94099897e-01 9.65897217e-02 9.06401277e-01 -6.51284575e-01
1.60110816e-01 -2.13149476e+00 6.06535554e-01 1.76714569e-01
-6.39589503e-02 1.44732296e-02 -2.71508753e-01 5.07584214e-03
-3.16093564e-01 -2.78428078e-01 -3.74918997e-01 -6.28380954e-01
-1.14657983e-01 6.50628328e-01 -3.12148571e-01 8.41787279e-01
4.01830614e-01 6.48171246e-01 -7.92066097e-01 -3.98213893e-01
4.29155052e-01 5.51195920e-01 -7.36140549e-01 3.48733455e-01
-5.29242635e-01 7.60410964e-01 -3.85375097e-02 5.05025625e-01
7.10074067e-01 -4.04227316e-01 1.19147405e-01 -3.49813700e-01
8.71701017e-02 2.42213324e-01 -1.63376617e+00 1.83735335e+00
-5.12789071e-01 2.91603774e-01 3.62733483e-01 -1.11756408e+00
7.14932501e-01 2.19461560e-01 5.45323551e-01 -4.21568185e-01
3.87204103e-02 1.12444974e-01 -3.02087188e-01 -2.40773112e-01
5.95401302e-02 -1.25152558e-01 2.71570504e-01 3.30377460e-01
4.39351201e-01 -9.02793229e-01 4.11988795e-02 1.30712390e-01
7.05773771e-01 5.84728599e-01 1.01547189e-01 -2.99878895e-01
4.10500497e-01 -2.15087160e-01 5.32210290e-01 3.57057661e-01
1.75681964e-01 1.22207153e+00 4.41290587e-01 -5.85585415e-01
-1.45390320e+00 -1.24688029e+00 -4.80418682e-01 5.70312083e-01
1.62666753e-01 -4.45312485e-02 -4.59538251e-01 -9.56786692e-01
1.29290909e-01 4.92953986e-01 -8.58438730e-01 2.88858414e-01
-6.58429861e-01 -3.42323005e-01 5.08440286e-02 5.70695341e-01
1.14333890e-01 -8.58989358e-01 -5.88894486e-01 -1.76106796e-01
1.74404651e-01 -1.40019393e+00 -5.31956732e-01 4.69558835e-01
-8.69759381e-01 -1.41640723e+00 -5.70544720e-01 -7.02097952e-01
9.64220941e-01 1.41648084e-01 1.54827094e+00 1.32864073e-01
-3.53088826e-01 9.35366154e-01 8.61515477e-02 -7.77457133e-02
-3.70685786e-01 -3.23542595e-01 -2.14983895e-01 8.09878185e-02
-3.02581638e-01 -8.17980468e-01 -3.35088313e-01 4.33980703e-01
-8.82253706e-01 1.95975274e-01 3.88287127e-01 1.04681563e+00
1.03798962e+00 -4.59191859e-01 -4.52253297e-02 -1.20434868e+00
-4.73668277e-01 -3.27706844e-01 -9.31636333e-01 2.43260399e-01
-4.60541248e-01 -4.99697328e-02 5.46170533e-01 -4.37264025e-01
-8.97956371e-01 5.77826381e-01 5.91067933e-02 -9.33969438e-01
-4.17287380e-01 -1.58743575e-01 -4.34457421e-01 -2.90391207e-01
6.97784066e-01 2.72032887e-01 2.90037513e-01 -4.25373405e-01
4.31139141e-01 -9.80741978e-02 5.72963238e-01 -9.64098215e-01
8.75742495e-01 8.58771503e-01 1.64870441e-01 -7.25285709e-01
-8.04027617e-01 -5.53282738e-01 -1.04004025e+00 1.07447140e-01
8.58111858e-01 -1.09127676e+00 -4.88041073e-01 3.17041516e-01
-1.11532056e+00 -5.24492264e-01 -5.21135151e-01 2.71541446e-01
-1.09641969e+00 3.73773426e-01 -4.12901640e-01 -4.18931782e-01
1.46391466e-01 -1.27446592e+00 1.44241679e+00 -2.42598787e-01
2.17802703e-01 -1.07941210e+00 -7.27568790e-02 5.77980638e-01
1.27007172e-01 4.73630995e-01 7.49735534e-01 -6.05992913e-01
-9.46016788e-01 8.61170068e-02 -1.94726497e-01 6.39967799e-01
1.01106456e-02 -1.69237405e-02 -1.09645700e+00 -5.08614540e-01
4.59767580e-02 -7.83389211e-01 7.58966684e-01 5.10648340e-02
1.25242078e+00 -1.01233788e-01 -2.05596119e-01 9.40535665e-01
1.52027798e+00 -1.49454594e-01 4.45842326e-01 -9.49109904e-03
9.97270167e-01 7.26409793e-01 3.85693550e-01 2.07297951e-01
3.54370922e-01 8.87825429e-01 7.56565571e-01 -2.34530538e-01
-3.89114648e-01 -1.64461240e-01 1.39388472e-01 6.84303164e-01
-1.54211551e-01 8.22166633e-03 -8.22017729e-01 6.07939422e-01
-1.44440663e+00 -8.94158781e-01 4.25118953e-02 2.46712422e+00
8.22785795e-01 6.77669793e-02 1.47864178e-01 -3.75080854e-02
2.88150370e-01 7.62933269e-02 -7.35223889e-01 4.49339040e-02
-1.56884119e-01 3.28487247e-01 4.44597989e-01 7.14024603e-01
-1.10466945e+00 6.59010351e-01 5.63625097e+00 4.46636587e-01
-1.16708231e+00 1.57767273e-02 7.52051234e-01 1.27669990e-01
-4.86162543e-01 9.71745793e-03 -8.27027082e-01 1.45246536e-01
3.58211637e-01 3.94646496e-01 4.92724240e-01 8.76513004e-01
-4.05516446e-01 -2.36234860e-03 -1.59232152e+00 7.24429488e-01
4.68455523e-01 -1.24942160e+00 6.40128031e-02 1.53342858e-02
9.40841258e-01 7.48875961e-02 3.47134843e-02 9.75594819e-02
4.94973630e-01 -9.54253316e-01 9.27953184e-01 3.68557543e-01
9.21427667e-01 -4.48881269e-01 3.68232876e-01 3.99944514e-01
-9.98670757e-01 2.70847470e-01 -3.16751122e-01 4.08706337e-01
-3.44205573e-02 3.47212821e-01 -6.77323341e-01 8.11676800e-01
6.20362818e-01 9.97482598e-01 -4.84924197e-01 6.17848575e-01
-2.41030902e-01 2.06503615e-01 -4.66940761e-01 5.95143497e-01
-2.29145922e-02 -2.95136213e-01 5.09191871e-01 9.59371328e-01
-8.10902119e-02 5.92760853e-02 5.18779159e-01 9.13612485e-01
2.23865643e-01 -1.94041625e-01 -7.09303379e-01 5.35104513e-01
1.40872538e-01 1.10328507e+00 -6.03957295e-01 -1.57515451e-01
-6.43188238e-01 8.79096687e-01 4.34714705e-01 3.24743211e-01
-6.33336008e-01 4.17977750e-01 5.47890544e-01 3.94135505e-01
8.26804042e-01 -1.43557012e-01 -1.46281883e-01 -1.36477947e+00
1.72690630e-01 -1.02787471e+00 2.81739473e-01 -8.31120193e-01
-1.47769380e+00 5.67941785e-01 2.37705126e-01 -1.26208603e+00
-3.29138577e-01 -8.22895765e-01 -2.60387570e-01 8.37012470e-01
-1.56633854e+00 -1.39458501e+00 -3.01490307e-01 7.34859526e-01
5.66602707e-01 -8.51086602e-02 7.93060005e-01 2.42811769e-01
-1.91874847e-01 3.20400089e-01 -2.07745090e-01 -3.81036475e-02
7.26053059e-01 -1.49556983e+00 2.30981857e-01 3.53553921e-01
4.41998690e-01 3.27782869e-01 4.45804924e-01 -3.46306503e-01
-1.52736568e+00 -1.02453399e+00 2.95970857e-01 -9.66048241e-01
4.00679082e-01 -7.10801244e-01 -9.91214812e-01 9.74876702e-01
6.09936900e-02 6.87256932e-01 1.98534891e-01 -1.84451155e-02
-6.26410067e-01 -1.28176197e-01 -1.12884033e+00 2.79791474e-01
1.08434284e+00 -4.63652879e-01 -4.79052156e-01 5.25867462e-01
4.74262178e-01 -9.45383191e-01 -8.79464507e-01 4.49092716e-01
5.17094553e-01 -1.19764328e+00 1.28759825e+00 -6.76713586e-01
5.53013504e-01 -3.63867849e-01 -4.70296621e-01 -1.28330004e+00
1.18542977e-01 -2.62370199e-01 4.17146683e-02 1.03977799e+00
2.37089768e-01 -3.70193273e-01 7.69418180e-01 6.11554027e-01
-2.06883639e-01 -1.07522488e+00 -8.13094258e-01 -7.92206943e-01
1.62472308e-01 -4.08486634e-01 4.04029578e-01 1.03570580e+00
-9.33697104e-01 3.67305696e-01 -2.83823997e-01 2.71741092e-01
8.28633368e-01 2.82958627e-01 1.23165047e+00 -1.19020200e+00
-7.58127868e-01 -1.89241186e-01 -6.26476824e-01 -1.04057813e+00
3.53862494e-01 -8.44103932e-01 -1.60106830e-02 -8.51830900e-01
2.69737512e-01 -6.64076388e-01 2.07777739e-01 4.38557088e-01
1.61467910e-01 2.53451765e-01 3.73979285e-02 4.21268381e-02
-5.08497596e-01 5.31635523e-01 1.45871127e+00 -1.26316950e-01
1.93894356e-01 8.61822069e-02 -2.44726047e-01 8.45733881e-01
2.38669753e-01 -4.20298010e-01 -6.45730674e-01 -4.55204606e-01
1.08466752e-01 2.40452886e-01 9.09489810e-01 -7.15803325e-01
1.48227839e-02 -2.38599479e-01 5.33281922e-01 -6.27888978e-01
7.39515305e-01 -1.16146278e+00 3.37243766e-01 9.60534289e-02
-2.02758178e-01 1.17706969e-01 2.12911755e-01 8.39853823e-01
-2.94376500e-02 -1.22740164e-01 1.09230578e+00 -3.31330001e-01
-5.71939051e-01 5.72767437e-01 3.77565205e-01 5.16829252e-01
1.00996995e+00 -3.62683207e-01 -1.00074150e-02 -1.14784092e-01
-9.33910012e-01 1.17002539e-01 1.10969782e+00 2.73957253e-01
5.14957666e-01 -1.22828674e+00 -7.56350756e-01 6.44852221e-01
1.20337419e-01 5.60044944e-01 1.22167669e-01 7.51004159e-01
-3.73058259e-01 1.30844973e-02 -1.97101891e-01 -1.15427518e+00
-1.28223217e+00 6.63250983e-01 6.91907823e-01 -8.97629708e-02
-7.91554034e-01 8.05978060e-01 6.39578819e-01 -8.84602726e-01
4.43676740e-01 -1.32020757e-01 2.28983760e-01 -1.83398172e-01
1.34143159e-01 -7.69047290e-02 3.81664544e-01 -7.24918723e-01
-2.06588283e-01 7.17607081e-01 -8.84478092e-02 -1.43035814e-01
1.46847653e+00 -3.82117322e-03 3.99561599e-02 6.38934851e-01
1.23071158e+00 4.48886044e-02 -1.80110335e+00 -4.24968958e-01
-2.76379496e-01 -6.30520880e-01 -2.32751742e-01 -6.32322967e-01
-1.25827730e+00 8.54021072e-01 5.41538537e-01 -1.41599819e-01
7.82462776e-01 1.82282284e-01 3.80264372e-01 3.27819645e-01
6.06600583e-01 -7.56648183e-01 3.13423276e-01 5.05895913e-01
1.14779150e+00 -1.46691918e+00 1.73018530e-01 -4.46198702e-01
-5.39598405e-01 9.43841517e-01 6.43770039e-01 -4.08703774e-01
6.96280897e-01 3.74309748e-01 3.55591241e-04 -2.67306536e-01
-8.74670446e-01 4.22610119e-02 5.43838143e-01 7.40042150e-01
2.29757741e-01 -2.16231346e-01 3.47865045e-01 2.11983889e-01
4.38391529e-02 -4.97027665e-01 3.00763309e-01 8.16109121e-01
5.72422445e-02 -1.06289041e+00 -2.96670556e-01 3.03108335e-01
-2.62101352e-01 9.99496356e-02 -3.67701352e-01 1.10904634e+00
2.15165481e-01 2.26344362e-01 2.31905729e-01 1.13818131e-01
5.40928066e-01 -9.93565693e-02 1.00438964e+00 -9.78348434e-01
-5.11026204e-01 1.19585320e-01 -2.77682189e-02 -6.84667051e-01
-6.25229836e-01 -1.03570724e+00 -9.31445777e-01 -1.63938612e-01
-2.69875765e-01 -1.51280686e-01 5.17027915e-01 9.53625321e-01
2.00047955e-01 2.63791740e-01 6.26484513e-01 -1.30429590e+00
-6.89727426e-01 -5.23736298e-01 -5.23926914e-01 7.91821003e-01
4.88189489e-01 -8.94268990e-01 -5.09063601e-01 3.61705929e-01] | [8.44853401184082, -2.9495432376861572] |
4b2563ba-f06b-40f1-a6ca-9b1011d0ae0e | co-attention-hierarchical-network-generating | 1911.08648 | null | https://arxiv.org/abs/1911.08648v1 | https://arxiv.org/pdf/1911.08648v1.pdf | Co-Attention Hierarchical Network: Generating Coherent Long Distractors for Reading Comprehension | In reading comprehension, generating sentence-level distractors is a significant task, which requires a deep understanding of the article and question. The traditional entity-centered methods can only generate word-level or phrase-level distractors. Although recently proposed neural-based methods like sequence-to-sequence (Seq2Seq) model show great potential in generating creative text, the previous neural methods for distractor generation ignore two important aspects. First, they didn't model the interactions between the article and question, making the generated distractors tend to be too general or not relevant to question context. Second, they didn't emphasize the relationship between the distractor and article, making the generated distractors not semantically relevant to the article and thus fail to form a set of meaningful options. To solve the first problem, we propose a co-attention enhanced hierarchical architecture to better capture the interactions between the article and question, thus guide the decoder to generate more coherent distractors. To alleviate the second problem, we add an additional semantic similarity loss to push the generated distractors more relevant to the article. Experimental results show that our model outperforms several strong baselines on automatic metrics, achieving state-of-the-art performance. Further human evaluation indicates that our generated distractors are more coherent and more educative compared with those distractors generated by baselines. | ['Yunfang Wu', 'Senlin Luo', 'Xiaorui Zhou'] | 2019-11-20 | null | null | null | null | ['distractor-generation'] | ['natural-language-processing'] | [ 1.08078532e-02 2.27369666e-01 1.91372499e-01 -5.69453202e-02
-8.03107679e-01 -4.32506651e-01 6.33561790e-01 2.78668310e-02
-3.73039961e-01 7.71140873e-01 8.99708629e-01 -1.78029969e-01
2.91983426e-01 -7.57025123e-01 -7.55900085e-01 -3.64395827e-01
6.92514956e-01 3.85258079e-01 4.17927951e-01 -6.60607994e-01
6.18713915e-01 -1.31207824e-01 -1.41508019e+00 4.94391084e-01
1.51188123e+00 4.24393892e-01 8.42734337e-01 3.35213840e-01
-6.39458537e-01 7.61418641e-01 -1.20748627e+00 -2.50807315e-01
-2.27432549e-01 -1.09406376e+00 -7.47731328e-01 -3.80985290e-01
6.05932176e-01 -4.37792450e-01 -3.37480962e-01 1.07385933e+00
7.76101351e-01 3.52903575e-01 8.71867418e-01 -6.10685766e-01
-1.44960725e+00 9.20688272e-01 -3.49390358e-01 5.28873086e-01
5.98421931e-01 3.66977423e-01 9.99799490e-01 -1.10327470e+00
4.64249641e-01 1.42847919e+00 1.39366969e-01 8.92983377e-01
-1.09360552e+00 -6.23612881e-01 4.37935978e-01 4.68441129e-01
-1.25369966e+00 -3.41802120e-01 7.44546533e-01 -3.15432012e-01
1.00126839e+00 2.92394668e-01 5.48004389e-01 1.67550015e+00
2.14640632e-01 8.88108909e-01 8.95948946e-01 -4.54173356e-01
1.58892706e-01 -4.47580562e-04 6.81255907e-02 3.43677491e-01
2.28524864e-01 2.63496023e-02 -5.30117035e-01 3.46141458e-01
7.13584542e-01 -4.28718299e-01 -5.75654089e-01 3.07090521e-01
-1.27082086e+00 8.96165371e-01 4.88019466e-01 5.28262436e-01
-4.65667039e-01 3.80619504e-02 1.63809180e-01 -3.04890350e-02
8.28392431e-02 9.98603523e-01 -1.47693440e-01 -1.99029177e-01
-9.27480996e-01 4.48394358e-01 5.09562075e-01 1.06108963e+00
3.59294355e-01 3.98812264e-01 -1.12406576e+00 8.23291540e-01
1.80406645e-01 5.11440158e-01 8.16182494e-01 -7.68194914e-01
6.62626684e-01 4.91451561e-01 1.46871701e-01 -1.04333520e+00
-2.14547426e-01 -5.97100198e-01 -6.81507051e-01 -1.68818921e-01
5.08453608e-01 -1.43878028e-01 -7.26151466e-01 1.88633931e+00
-1.84668213e-01 -2.00425103e-01 -2.54966319e-02 1.05463302e+00
1.28120637e+00 8.44385982e-01 9.16972831e-02 -7.92706460e-02
1.60516310e+00 -1.39659655e+00 -1.16911006e+00 -5.92902958e-01
4.56663787e-01 -8.18946123e-01 1.73886693e+00 1.80718690e-01
-1.34347785e+00 -8.61962199e-01 -1.02060699e+00 -4.50672030e-01
-1.66810840e-01 6.94614425e-02 -7.75838494e-02 5.15791646e-04
-7.99615264e-01 2.38628611e-01 -1.80613995e-01 -1.23865031e-01
3.15394431e-01 -4.10279572e-01 2.55155236e-01 1.68895368e-02
-1.64914703e+00 1.25937939e+00 4.78918582e-01 -2.44043525e-02
-8.09156001e-01 -6.65448904e-01 -8.14911664e-01 5.19975662e-01
5.74580312e-01 -9.66208696e-01 1.31481183e+00 -1.00078166e+00
-1.49211454e+00 3.96824002e-01 -3.66981924e-01 -8.29569027e-02
1.92326769e-01 -4.98887807e-01 -3.24954063e-01 1.54316977e-01
2.59242207e-01 8.35074127e-01 8.27369094e-01 -1.17510283e+00
-3.18989486e-01 1.05512656e-01 -5.14308847e-02 5.28347552e-01
-2.01610476e-01 -2.56716907e-01 -2.83957481e-01 -1.14557183e+00
-7.77852535e-02 -6.05104804e-01 -3.17825899e-02 -4.20396417e-01
-5.87486744e-01 -5.60711026e-01 3.26490462e-01 -8.02549481e-01
1.44156742e+00 -1.96801305e+00 2.73195237e-01 -4.38356131e-01
1.79099187e-01 4.59922940e-01 -5.56394100e-01 5.29704392e-01
-6.09756336e-02 3.32232207e-01 4.71144589e-03 -3.90494801e-02
-9.06802937e-02 -8.31683576e-02 -3.54038268e-01 -3.43342930e-01
4.06480759e-01 1.24815691e+00 -1.22721493e+00 -3.31985652e-01
-2.63204455e-01 1.57677010e-01 -6.15543067e-01 4.11833584e-01
-7.57257462e-01 4.07594472e-01 -4.21865791e-01 -3.55239958e-02
4.52694237e-01 -2.77492046e-01 -2.66788125e-01 -6.65702149e-02
4.15970199e-02 1.03807831e+00 -7.28128791e-01 1.84572208e+00
-4.02911097e-01 5.01581669e-01 -4.21993345e-01 -5.88623941e-01
8.49116981e-01 1.88975587e-01 -3.54516774e-01 -1.13542914e+00
8.70670527e-02 9.32961404e-02 3.94935042e-01 -9.37530577e-01
7.30173051e-01 -3.34005296e-01 -7.29363039e-02 4.50567484e-01
-6.34623915e-02 -1.86428174e-01 3.04693431e-01 4.76898462e-01
8.43733132e-01 3.60488057e-01 1.89221874e-01 -3.00788075e-01
5.22810876e-01 -1.78242758e-01 3.71348530e-01 8.14070165e-01
-7.13293850e-02 9.82157707e-01 6.35681391e-01 1.84777930e-01
-8.09036195e-01 -1.05071509e+00 4.28968638e-01 1.02875829e+00
1.72024593e-01 -5.07295489e-01 -9.46514308e-01 -7.58744299e-01
-4.45137531e-01 1.60757732e+00 -5.86758852e-01 -4.93632615e-01
-7.60330677e-01 -1.95044115e-01 5.56688309e-01 6.38151050e-01
4.28587854e-01 -1.33998847e+00 -4.13410723e-01 4.05823052e-01
-8.87619138e-01 -9.52089667e-01 -1.13533330e+00 -3.16444904e-01
-5.82982719e-01 -8.03213060e-01 -8.54383469e-01 -8.27565968e-01
6.22119308e-01 3.88424397e-01 1.37571406e+00 1.46423280e-01
1.52120188e-01 -5.75582795e-02 -3.90779734e-01 -4.23495620e-01
-4.37243402e-01 1.15035698e-01 -3.89012784e-01 -4.02178913e-01
4.84494537e-01 -3.24645698e-01 -6.85372829e-01 -2.84985714e-02
-9.33249116e-01 2.15702951e-01 7.72138894e-01 9.27689135e-01
2.77604520e-01 -3.60386610e-01 1.01320803e+00 -6.20630443e-01
1.30644286e+00 -6.09411716e-01 -1.38471156e-01 2.03610972e-01
-3.11477929e-01 4.23745632e-01 1.02967823e+00 -5.91089487e-01
-1.16679668e+00 -5.55146754e-01 -2.95277864e-01 -2.84890592e-01
-1.53108418e-01 5.98201632e-01 -4.23137248e-01 8.26267064e-01
8.32579136e-01 6.64641678e-01 3.74174118e-02 -3.53466898e-01
5.63211858e-01 2.59970486e-01 3.67207557e-01 -6.28052354e-01
6.48584068e-01 -2.68435776e-01 -2.80792356e-01 -6.85808063e-01
-1.16652215e+00 -1.62329689e-01 -1.70945436e-01 -1.16058234e-02
1.10925269e+00 -8.02040160e-01 -4.46110576e-01 1.97715431e-01
-1.67631316e+00 -2.26580333e-02 -4.52211708e-01 2.95137197e-01
-2.67340481e-01 5.66438258e-01 -5.60260057e-01 -5.17810762e-01
-3.08148772e-01 -1.18954372e+00 8.24792922e-01 4.30443913e-01
-5.70817888e-01 -7.74281740e-01 -1.01214297e-01 5.51820099e-01
6.76026702e-01 -1.57100081e-01 1.26243734e+00 -8.48149478e-01
-6.75211608e-01 2.77989328e-01 -3.15120101e-01 4.21986043e-01
7.60879666e-02 -1.90827623e-01 -5.83166003e-01 2.39199430e-01
2.09295511e-01 -3.41053098e-01 8.86358857e-01 5.57108931e-02
1.12308061e+00 -7.93231189e-01 1.05425867e-03 8.37518498e-02
9.70862269e-01 1.22866370e-01 1.01287949e+00 2.52344441e-02
7.30578899e-01 8.24754596e-01 4.67498153e-01 6.80918545e-02
8.06954861e-01 6.41056418e-01 2.17094824e-01 1.03152886e-01
-6.63001001e-01 -6.38920546e-01 5.16991615e-01 1.28429568e+00
3.27886462e-01 -4.56350654e-01 -7.02573597e-01 6.67638481e-01
-1.59414017e+00 -1.26592302e+00 -7.22477317e-01 2.01127458e+00
1.17100179e+00 1.89952895e-01 -9.94499698e-02 -3.68051022e-01
5.98971307e-01 1.02440119e-01 -4.35549736e-01 -4.41993475e-01
-2.52136230e-01 3.52629781e-01 -2.92404413e-01 5.08178294e-01
-2.42675275e-01 1.23303735e+00 5.77410364e+00 9.80163813e-01
-9.68600154e-01 8.36523101e-02 3.25239450e-01 -3.58725339e-01
-8.09981525e-01 -8.95069540e-02 -9.08881783e-01 7.97678053e-01
8.83931935e-01 -3.51659358e-01 3.38186353e-01 4.45561469e-01
5.91314018e-01 -2.14904785e-01 -1.12461662e+00 6.26541793e-01
4.37005699e-01 -1.13977790e+00 5.68993449e-01 -2.79230326e-01
7.29369223e-01 -4.31703687e-01 1.01982057e-02 7.44394958e-01
1.76280402e-02 -1.27264547e+00 1.01124632e+00 8.60450506e-01
4.07604545e-01 -6.72131717e-01 4.43754733e-01 6.92917645e-01
-6.87324107e-01 1.87423930e-01 -4.93742019e-01 -1.60359472e-01
1.66748330e-01 5.59326291e-01 -6.02636397e-01 3.28207731e-01
3.12040478e-01 2.76505023e-01 -7.81431019e-01 7.92252541e-01
-8.20776045e-01 7.07834005e-01 1.15207791e-01 -6.18924797e-01
3.56211841e-01 -8.50011334e-02 6.49320722e-01 9.54357684e-01
4.96809721e-01 2.93744802e-01 -1.17851332e-01 1.40314603e+00
-2.15802193e-01 3.51362288e-01 -4.57839519e-01 -1.07964180e-01
6.21197343e-01 8.38116884e-01 -1.56436518e-01 -4.17760551e-01
-4.32989955e-01 1.21936333e+00 4.70585316e-01 5.13887584e-01
-9.12836611e-01 -6.12205565e-01 3.44408363e-01 3.45423043e-01
3.53109360e-01 -1.55567706e-01 -3.62650841e-01 -1.31059635e+00
2.20591649e-01 -1.25435972e+00 3.96048836e-02 -1.07806897e+00
-1.38389826e+00 6.33729517e-01 -2.47990251e-01 -1.04762805e+00
-1.92591459e-01 -2.11511657e-01 -8.19705129e-01 1.33380389e+00
-1.37694204e+00 -8.39765549e-01 -1.96035266e-01 2.00423792e-01
9.89394903e-01 -8.33738875e-03 6.58280075e-01 -1.00083733e-02
-4.12345409e-01 6.46553040e-01 -2.46771812e-01 -2.07294766e-02
8.42462420e-01 -1.28046632e+00 3.52532953e-01 1.05356061e+00
-5.41645810e-02 1.02317619e+00 7.84748197e-01 -8.49191546e-01
-9.99626696e-01 -9.09833968e-01 1.34432006e+00 -6.62815869e-01
3.81267756e-01 -4.22049284e-01 -1.43185556e+00 2.21728355e-01
7.94893563e-01 -6.62685394e-01 7.27916420e-01 -6.12634176e-04
-4.52561051e-01 2.11258262e-01 -6.84065104e-01 1.11368406e+00
1.01300144e+00 -2.86657840e-01 -1.31491899e+00 1.94851324e-01
9.93508577e-01 -1.82260185e-01 -1.65684938e-01 1.00552015e-01
7.70174712e-02 -1.01483154e+00 6.78600729e-01 -6.38608813e-01
1.09722674e+00 -3.52694511e-01 3.20915580e-01 -1.72395670e+00
-7.16335177e-01 -4.09469455e-01 -3.28940272e-01 1.37979424e+00
5.27925193e-01 -1.96269915e-01 2.80538142e-01 2.43296534e-01
-5.28919160e-01 -7.29665577e-01 -7.43562877e-01 -8.31386685e-01
7.60849535e-01 4.50484566e-02 5.55252969e-01 7.57214308e-01
1.87804863e-01 1.34112227e+00 -1.97599962e-01 -3.69286954e-01
1.45613253e-01 -7.09730759e-02 3.43379408e-01 -8.05032194e-01
-2.51961917e-01 -9.82172430e-01 4.98307347e-01 -1.56186724e+00
1.96955726e-01 -8.95945430e-01 4.44215417e-01 -1.97347915e+00
1.60995990e-01 -2.09272355e-02 9.04935151e-02 1.67703226e-01
-1.04796255e+00 -3.05360287e-01 1.94014326e-01 4.20351252e-02
-4.61507738e-01 1.10132134e+00 1.89639378e+00 -6.00333139e-03
-1.23373553e-01 -3.36511135e-01 -1.30661964e+00 3.84494305e-01
5.85546792e-01 -2.87529051e-01 -5.78398168e-01 -7.55849063e-01
3.07649970e-01 1.29880220e-01 1.45539775e-01 -7.53879786e-01
1.51513278e-01 -3.19494694e-01 4.08877283e-01 -5.86027026e-01
1.30322546e-01 -3.46216381e-01 -5.34663379e-01 4.61029708e-01
-7.09887266e-01 2.66786695e-01 2.54676282e-01 4.51461673e-01
-1.60848007e-01 -6.58605635e-01 7.42570817e-01 -3.54612947e-01
-2.67082959e-01 -1.79929480e-01 -7.93594599e-01 6.84050739e-01
7.95465469e-01 -3.03427726e-02 -5.56253791e-01 -6.84653938e-01
-3.28198880e-01 3.46529573e-01 2.45187968e-01 7.46747613e-01
8.06412816e-01 -1.26186705e+00 -1.15312314e+00 -1.58614382e-01
1.83200389e-01 9.75495800e-02 2.43667319e-01 4.97239619e-01
-3.05328935e-01 6.78784549e-01 2.84396224e-02 -8.74483436e-02
-6.73853576e-01 8.91517460e-01 2.68646508e-01 -2.30132326e-01
-4.68759120e-01 1.06062376e+00 5.29772282e-01 -1.30069047e-01
1.95530459e-01 -2.83303112e-01 -4.96551007e-01 2.80303270e-01
7.14359879e-01 2.07968101e-01 -1.02949724e-01 -3.30096930e-01
-8.02162215e-02 1.19028740e-01 -2.70261139e-01 -1.32957965e-01
8.44792187e-01 -1.89457759e-01 3.65232415e-02 4.32800025e-01
7.40974367e-01 3.15588951e-01 -9.35374081e-01 -1.00391075e-01
1.00165427e-01 -2.42485777e-01 -2.41424844e-01 -1.20513105e+00
-7.48495460e-01 1.17479897e+00 -1.07219912e-01 7.37737790e-02
9.56105530e-01 -1.09962955e-01 1.15693045e+00 2.63196707e-01
-1.47337258e-01 -1.12744498e+00 8.17560256e-01 1.04589283e+00
1.43586063e+00 -8.35219145e-01 -3.74685138e-01 -4.63732898e-01
-9.09198940e-01 1.02569473e+00 1.39392340e+00 -2.60134339e-02
-9.78094116e-02 -1.24856852e-01 4.44509275e-02 1.95117220e-02
-9.94952857e-01 -2.44810596e-01 5.02818763e-01 4.41308051e-01
8.08389843e-01 -3.76619786e-01 -8.13752890e-01 9.90212202e-01
-4.17630583e-01 -2.37208128e-01 8.12309623e-01 5.53625047e-01
-5.57234466e-01 -1.01947749e+00 -3.12907100e-01 5.43644845e-01
-3.21762979e-01 -7.82527983e-01 -5.32344222e-01 3.51471484e-01
7.20829517e-02 9.00319636e-01 1.78978760e-02 4.05807905e-02
5.70225000e-01 3.73702735e-01 5.51847219e-01 -1.02631140e+00
-8.29934359e-01 1.14018090e-01 1.21001773e-01 -2.26214200e-01
1.96665198e-01 -5.40906429e-01 -1.38728905e+00 -6.18294887e-02
-4.68846291e-01 1.88457713e-01 2.97768056e-01 1.03806484e+00
5.78431487e-01 1.02950263e+00 3.22693706e-01 -2.15832964e-01
-8.90684783e-01 -1.39712477e+00 -2.71435045e-02 4.86891627e-01
1.03187449e-01 -4.45885748e-01 -2.66164631e-01 -1.56793952e-01] | [11.654434204101562, 8.392521858215332] |
d4fe6635-ec38-40ee-9f26-64cf125f6454 | deep-learning-for-cancer-prognosis-prediction | 2306.14596 | null | https://arxiv.org/abs/2306.14596v2 | https://arxiv.org/pdf/2306.14596v2.pdf | Deep Learning for Cancer Prognosis Prediction Using Portrait Photos by StyleGAN Embedding | Survival prediction for cancer patients is critical for optimal treatment selection and patient management. Current patient survival prediction methods typically extract survival information from patients' clinical record data or biological and imaging data. In practice, experienced clinicians can have a preliminary assessment of patients' health status based on patients' observable physical appearances, which are mainly facial features. However, such assessment is highly subjective. In this work, the efficacy of objectively capturing and using prognostic information contained in conventional portrait photographs using deep learning for survival predication purposes is investigated for the first time. A pre-trained StyleGAN2 model is fine-tuned on a custom dataset of our cancer patients' photos to empower its generator with generative ability suitable for patients' photos. The StyleGAN2 is then used to embed the photographs to its highly expressive latent space. Utilizing the state-of-the-art survival analysis models and based on StyleGAN's latent space photo embeddings, this approach achieved a C-index of 0.677, which is notably higher than chance and evidencing the prognostic value embedded in simple 2D facial images. In addition, thanks to StyleGAN's interpretable latent space, our survival prediction model can be validated for relying on essential facial features, eliminating any biases from extraneous information like clothing or background. Moreover, a health attribute is obtained from regression coefficients, which has important potential value for patient care. | ['Yixing Huang', 'Florian Putz', 'Christoph Bert', 'Rainer Fietkau', 'Andreas Maier', 'Dominik Kornek', 'Ahmed Gomaa', 'Amr Hagag'] | 2023-06-26 | null | null | null | null | ['survival-analysis', 'management'] | ['miscellaneous', 'miscellaneous'] | [ 4.20017183e-01 4.15542543e-01 -3.73632133e-01 -3.56061310e-01
-8.33959401e-01 -2.27576882e-01 5.29127598e-01 9.35314521e-02
-2.81855613e-01 8.38509202e-01 5.50074458e-01 -1.12591453e-01
5.32895140e-03 -9.32332337e-01 -8.62387121e-02 -1.24956238e+00
-1.15673445e-01 2.10253671e-01 -6.27200425e-01 -3.02603059e-02
-4.15190995e-01 6.60904586e-01 -1.33952522e+00 2.03707650e-01
5.38328707e-01 1.12461770e+00 -2.10923940e-01 6.79341316e-01
1.71794549e-01 4.82571483e-01 -4.68709737e-01 -6.19072258e-01
-1.77283436e-01 -3.50927532e-01 -4.06150639e-01 -2.41360860e-03
2.40787908e-01 -5.00981212e-01 -5.54397941e-01 7.39691019e-01
5.70616901e-01 -4.33915406e-01 8.28646064e-01 -1.10681176e+00
-8.07659864e-01 2.37352818e-01 -1.44516215e-01 -4.55369562e-01
2.19466135e-01 3.71555179e-01 9.24041212e-01 -5.50125480e-01
5.89278460e-01 8.87028933e-01 7.22452164e-01 9.10128772e-01
-1.34068501e+00 -5.28255224e-01 -4.18708056e-01 -9.97852236e-02
-1.49577749e+00 -4.71806794e-01 9.13641095e-01 -4.28228676e-01
2.99708307e-01 6.40392244e-01 1.01902568e+00 1.55928206e+00
4.31098193e-01 6.29395366e-01 1.07292640e+00 -2.60006398e-01
1.15156427e-01 2.80302256e-01 -4.92858022e-01 1.01453269e+00
-1.35258377e-01 4.80528265e-01 -2.86476791e-01 -1.71061680e-01
5.89607894e-01 3.87868613e-01 -5.75379193e-01 -4.36329246e-02
-1.11722255e+00 8.61552000e-01 7.11744189e-01 4.38358188e-01
-2.54150391e-01 2.48013869e-01 2.52445966e-01 1.32479936e-01
6.26135707e-01 1.01576261e-02 -1.72678366e-01 6.84697032e-02
-9.57021713e-01 -5.38430274e-01 2.92702705e-01 4.62131411e-01
2.09170029e-01 3.95517461e-02 -3.82785231e-01 6.02914512e-01
3.99372965e-01 6.35637105e-01 7.20410705e-01 -4.68905181e-01
-2.61237174e-01 9.36393678e-01 -3.24733019e-01 -1.16061282e+00
-6.46475017e-01 -5.78514814e-01 -1.29335582e+00 2.42531061e-01
1.79438084e-01 2.52565652e-01 -8.28485847e-01 1.84491491e+00
2.36622646e-01 -1.14751436e-01 4.04183716e-01 7.54250944e-01
1.03626096e+00 2.65300423e-01 3.93540174e-01 -2.57084191e-01
1.51401043e+00 -6.81009054e-01 -5.73888063e-01 6.23505861e-02
1.05474091e+00 -4.21744823e-01 1.00456941e+00 3.41798514e-02
-4.86779928e-01 -1.94311365e-01 -8.16340029e-01 -5.02314828e-02
-2.07104295e-01 6.96905375e-01 7.83880234e-01 9.24869418e-01
-1.18172157e+00 6.46675169e-01 -5.85542262e-01 -5.61540365e-01
7.95645177e-01 5.08708417e-01 -9.27404284e-01 -1.49542049e-01
-1.15158379e+00 6.04964793e-01 1.00345306e-01 6.87541664e-02
-7.66576827e-01 -7.64683962e-01 -9.39799845e-01 3.59283760e-02
-1.64166003e-01 -1.09272790e+00 6.82525635e-01 -1.19876611e+00
-1.69089007e+00 1.14704645e+00 8.22468698e-02 -1.60266802e-01
6.17514968e-01 3.45940441e-01 -6.74570918e-01 4.75319713e-01
-3.84827793e-01 6.85577929e-01 1.02386439e+00 -8.44819188e-01
-1.54158592e-01 -5.33500373e-01 -2.48700142e-01 -1.76148564e-01
-8.91364157e-01 -3.09081435e-01 -2.49202803e-01 -8.67985725e-01
-3.18715185e-01 -1.08347058e+00 -5.39624505e-02 8.38196754e-01
-3.01166207e-01 1.90840051e-01 8.02123785e-01 -8.52403879e-01
1.11352134e+00 -2.35701585e+00 1.93435311e-01 1.22164004e-01
4.54602569e-01 9.26545411e-02 -2.74090737e-01 3.01059902e-01
-1.36773258e-01 2.42073625e-01 -3.40942383e-01 -6.73971713e-01
-3.11689824e-02 -2.83890236e-02 8.63257870e-02 8.90439868e-01
1.49060339e-01 1.34490120e+00 -8.09339046e-01 -6.75759435e-01
2.92070478e-01 9.37764287e-01 -2.49452040e-01 2.72051066e-01
1.11703947e-01 6.88977480e-01 -2.43146047e-01 9.12379920e-01
2.72711694e-01 -2.43201450e-01 3.05303305e-01 -2.82388180e-01
4.41298902e-01 -4.02753234e-01 -9.41174552e-02 1.66447580e+00
-5.29350817e-01 7.07487285e-01 -7.51275644e-02 -6.08078778e-01
8.78901243e-01 5.00843763e-01 7.13411748e-01 -4.50671852e-01
4.89313126e-01 -8.94134305e-03 -3.65842551e-01 -3.30239743e-01
-4.21234965e-02 -6.18144453e-01 -1.91453457e-01 2.08602980e-01
-6.33592233e-02 1.20891258e-01 -6.01588249e-01 9.60382819e-02
1.03800797e+00 -8.47345293e-02 3.20607007e-01 -3.68120521e-02
5.41712642e-01 -2.38833979e-01 1.63543761e-01 -2.70338114e-02
-2.32132494e-01 7.05392838e-01 5.59799314e-01 -4.29644972e-01
-9.14949238e-01 -7.71109343e-01 -3.57231975e-01 4.27083075e-01
-4.31937426e-01 -3.73691112e-01 -5.27176857e-01 -9.51176405e-01
-4.20139357e-02 5.24751425e-01 -1.25440502e+00 -6.67093992e-01
5.74100949e-03 -8.55325043e-01 6.92551315e-01 5.51044524e-01
7.27139190e-02 -7.71149457e-01 -3.27521741e-01 7.93510899e-02
1.05416246e-01 -7.86107838e-01 -4.14803088e-01 -2.00663537e-01
-7.95081317e-01 -1.06996810e+00 -1.04798806e+00 -4.38051790e-01
1.05366290e+00 -2.98884243e-01 8.61443698e-01 4.14294600e-01
-6.57357156e-01 3.76931846e-01 -3.07702214e-01 1.34964073e-02
-6.89479291e-01 -1.30570769e-01 4.47816998e-02 3.37098181e-01
1.08582087e-01 -5.11902273e-01 -1.05102432e+00 1.24211006e-01
-8.61346543e-01 2.61084497e-01 8.95297229e-01 1.18624425e+00
5.41956961e-01 1.09879104e-02 3.17306101e-01 -7.99632132e-01
3.15781645e-02 -4.95115578e-01 -2.09082082e-01 1.34109169e-01
-5.95197976e-01 -1.81952044e-01 6.72226012e-01 -2.96773970e-01
-7.97445834e-01 1.09153077e-01 -3.16910654e-01 -4.81515646e-01
-2.35165104e-01 3.87011707e-01 -2.31183022e-01 -2.35912681e-01
3.19600642e-01 1.35165364e-01 3.36241663e-01 -1.63557917e-01
1.02316156e-01 7.02583373e-01 3.97946417e-01 2.54592393e-02
6.12181962e-01 7.51145184e-01 4.83510345e-01 -7.97886491e-01
-6.90776408e-01 -1.21032298e-01 -6.02323592e-01 -2.85004169e-01
9.65592206e-01 -8.07585359e-01 -1.02738464e+00 5.00405610e-01
-6.28156543e-01 -4.12796974e-01 -2.30026007e-01 4.21945751e-01
-3.87191653e-01 2.40757048e-01 -6.99320555e-01 -6.35004580e-01
-5.73769331e-01 -9.09944654e-01 1.42174292e+00 -1.46144768e-02
-3.38584632e-01 -1.39220667e+00 4.54385802e-02 4.10708159e-01
3.64763170e-01 8.76301050e-01 1.16060758e+00 -2.32282221e-01
-1.92279160e-01 -7.63504684e-01 -1.49166822e-01 2.43182912e-01
3.95937860e-01 1.62587881e-01 -1.37384069e+00 -4.95871514e-01
-2.89974302e-01 -3.15648377e-01 8.57948244e-01 2.31750250e-01
1.35843432e+00 -4.50194240e-01 -5.72074831e-01 9.13651288e-01
1.44113576e+00 -1.86456829e-01 7.34039843e-01 -3.61788869e-02
7.79842496e-01 7.05129921e-01 1.16684206e-01 4.30988729e-01
2.59367555e-01 6.53243780e-01 5.82699895e-01 -4.74825203e-01
-3.07393193e-01 -5.07019639e-01 3.88469636e-01 6.47200167e-01
1.21990018e-01 -5.13053834e-01 -8.30601513e-01 3.32995296e-01
-1.26587701e+00 -7.55004108e-01 -2.66486127e-02 1.97168505e+00
8.16095769e-01 -8.51636678e-02 -1.85202911e-01 2.31774002e-01
4.08537835e-01 3.69149186e-02 -3.24204028e-01 -1.43959999e-01
-4.28009391e-01 -1.23889989e-03 3.87588024e-01 2.22011179e-01
-9.76291955e-01 5.23771524e-01 6.06490612e+00 8.09428453e-01
-1.47197580e+00 1.62030682e-02 1.13189089e+00 -2.70756543e-01
-6.51949704e-01 -2.09616125e-01 -3.10417384e-01 5.01910627e-01
1.14318609e+00 1.71969116e-01 2.04327498e-02 6.33617640e-01
9.07890573e-02 1.07483581e-01 -1.12791097e+00 9.77732658e-01
1.23740606e-01 -1.24138176e+00 1.08223908e-01 5.16941786e-01
4.53353137e-01 -6.07504129e-01 5.05591989e-01 1.18112400e-01
-7.93476403e-02 -1.56134629e+00 3.80056858e-01 9.03026938e-01
1.76089740e+00 -7.64234602e-01 9.24580693e-01 1.50278434e-01
-7.40062714e-01 -2.36743852e-01 -6.02120487e-03 3.91001910e-01
1.43254220e-01 6.91809773e-01 -1.07198882e+00 4.30290282e-01
2.15338126e-01 7.84257233e-01 -6.28684282e-01 3.39565456e-01
-2.22573116e-01 6.14504457e-01 -1.01005472e-01 1.84119672e-01
-2.56199911e-02 4.24346924e-02 2.60791421e-01 1.00280333e+00
7.02825963e-01 3.54452580e-01 -2.87564635e-01 5.99208474e-01
-6.09530173e-02 1.33269563e-01 -5.36361158e-01 -4.02274519e-01
-6.28696755e-02 1.51862037e+00 -5.47073483e-01 -8.47992077e-02
-3.55062336e-01 1.16576779e+00 1.45697519e-01 -2.39767618e-02
-6.74160779e-01 2.33424410e-01 6.35383606e-01 3.11296761e-01
4.10076119e-02 1.54006287e-01 -1.52209252e-01 -1.07164788e+00
-3.98058206e-01 -6.01412535e-01 3.45341921e-01 -7.58500218e-01
-1.22650313e+00 8.73869538e-01 -5.35186112e-01 -1.33254457e+00
8.90026316e-02 -6.58984661e-01 -5.64671814e-01 7.06124067e-01
-1.42077935e+00 -1.94377756e+00 -4.62779641e-01 5.10492027e-01
-4.85748872e-02 -2.77091414e-01 1.46770716e+00 1.30337775e-01
-7.69976079e-01 1.10506487e+00 1.72877640e-01 1.83245942e-01
8.01616728e-01 -1.13233352e+00 -1.47776023e-01 3.31965983e-01
-2.14238644e-01 1.92515224e-01 4.36036080e-01 -5.73483109e-01
-1.39725256e+00 -1.15377927e+00 8.08770359e-01 -4.92742926e-01
5.54068148e-01 -2.31171146e-01 -6.60812795e-01 3.89437884e-01
-1.41090751e-01 2.70166099e-01 1.39733553e+00 -2.39371628e-01
-1.23923212e-01 -1.83596700e-01 -1.27156985e+00 6.02486551e-01
8.87732446e-01 -6.27913773e-01 9.49281603e-02 2.69990027e-01
4.37321723e-01 -9.94948968e-02 -1.40718699e+00 5.01416087e-01
8.80860984e-01 -7.51722515e-01 8.38815749e-01 -4.66092378e-01
7.30487525e-01 1.31787313e-02 -2.11481571e-01 -1.27341974e+00
-4.34499562e-01 -3.65974277e-01 5.40348813e-02 1.28632939e+00
2.22435623e-01 -4.64717895e-01 9.42758679e-01 7.56976426e-01
1.15032710e-01 -9.48469460e-01 -9.02504265e-01 -2.97339678e-01
-3.73638459e-02 -1.43923208e-01 6.78054035e-01 8.57310414e-01
-1.14484124e-01 1.10782357e-02 -5.43989062e-01 4.42178100e-02
6.47150159e-01 6.62679523e-02 5.38063407e-01 -1.23718190e+00
-2.05105916e-01 -4.34367508e-01 -8.32984746e-01 -1.96947873e-01
4.20996159e-01 -1.05364680e+00 -4.67061371e-01 -1.12659204e+00
3.53655636e-01 -6.68201745e-01 -5.04420698e-01 6.68842196e-01
-1.17361166e-01 7.40094721e-01 -7.29048476e-02 2.26186782e-01
-4.97571006e-02 9.19617236e-01 1.42751074e+00 -3.62799853e-01
2.08297238e-01 -2.23216027e-01 -7.25830197e-01 5.59654415e-01
7.59501159e-01 -4.73582715e-01 -2.24076658e-01 -7.13306963e-02
5.60965203e-02 5.31745613e-01 7.45673716e-01 -7.63978660e-01
-9.81755629e-02 -1.16574332e-01 7.20081329e-01 -1.20696370e-02
6.11525953e-01 -1.04269505e+00 5.70649087e-01 6.65367365e-01
-2.30706155e-01 -4.09361601e-01 1.31436065e-01 6.95387065e-01
-2.33979195e-01 1.31582394e-01 7.22101927e-01 1.94036871e-01
-4.32978839e-01 6.18561864e-01 -2.53310502e-01 -5.64506590e-01
1.24106312e+00 -3.13267708e-01 -2.70976573e-01 -4.54822093e-01
-8.50319445e-01 -2.25058779e-01 9.23619270e-01 6.49276972e-02
7.26766169e-01 -1.65353036e+00 -8.49924982e-01 4.72999454e-01
4.73400027e-01 -3.55925798e-01 7.28335559e-01 9.67394888e-01
-4.53236431e-01 3.25425804e-01 -3.24003309e-01 -5.23275495e-01
-1.48997772e+00 6.62485003e-01 1.28407121e-01 -2.12997794e-01
-7.52042651e-01 7.56865919e-01 3.49119186e-01 2.50755269e-02
-1.26355648e-01 -2.55569033e-02 -3.97188962e-01 2.62960225e-01
5.09499729e-01 -5.79809174e-02 -2.44989321e-02 -9.72941160e-01
-3.39400470e-01 5.07285595e-01 2.17460454e-01 4.08265293e-02
1.41109514e+00 -2.86908243e-02 2.05208752e-02 1.12911113e-01
1.65167129e+00 1.02329567e-01 -1.28799713e+00 1.29443388e-02
-5.02408266e-01 -5.68669558e-01 3.66134524e-01 -9.19265866e-01
-1.41683841e+00 8.58920038e-01 8.08110833e-01 -2.38460615e-01
1.47528875e+00 7.42440894e-02 9.07482207e-01 -3.03593874e-01
2.10888013e-01 -4.29489344e-01 7.18550310e-02 -3.69296283e-01
1.03145862e+00 -1.30334461e+00 -1.51577871e-02 -2.28725359e-01
-5.61384916e-01 1.16628921e+00 4.49100174e-02 3.51453215e-01
6.12689137e-01 1.53344288e-01 3.06961507e-01 -1.29424512e-01
-7.40758896e-01 1.83507606e-01 5.43762207e-01 5.18775165e-01
3.03587019e-01 4.99944389e-01 -6.50001736e-03 7.73888111e-01
-4.09343570e-01 1.59275178e-02 2.38797486e-01 5.31022727e-01
7.19079152e-02 -1.10386539e+00 -1.39822185e-01 4.33780998e-01
-5.02595305e-01 3.23310569e-02 -3.27102095e-01 5.26384473e-01
6.29383549e-02 4.70066160e-01 -2.90911384e-02 -4.47990626e-01
-3.45851891e-02 -3.24239110e-04 3.69241983e-01 -3.88515562e-01
-2.85235345e-01 -1.94242522e-01 -1.69364400e-02 -4.20871973e-01
-4.01874512e-01 -5.07046938e-01 -8.20166290e-01 -4.64159757e-01
-2.27195233e-01 -2.45639443e-01 6.97103918e-01 6.60928845e-01
2.29243726e-01 6.03134632e-01 6.81401670e-01 -6.75822616e-01
-2.66546130e-01 -6.28854632e-01 -6.18156970e-01 4.39075291e-01
4.78044987e-01 -5.12569427e-01 -3.52878362e-01 2.35359415e-01] | [15.278616905212402, -2.811782121658325] |
29ca0700-eb96-47b4-ac01-b211b57edcf7 | investigating-neighborhood-modeling-and | 2112.11734 | null | https://arxiv.org/abs/2112.11734v2 | https://arxiv.org/pdf/2112.11734v2.pdf | D-HYPR: Harnessing Neighborhood Modeling and Asymmetry Preservation for Digraph Representation Learning | Digraph Representation Learning (DRL) aims to learn representations for directed homogeneous graphs (digraphs). Prior work in DRL is largely constrained (e.g., limited to directed acyclic graphs), or has poor generalizability across tasks (e.g., evaluated solely on one task). Most Graph Neural Networks (GNNs) exhibit poor performance on digraphs due to the neglect of modeling neighborhoods and preserving asymmetry. In this paper, we address these notable challenges by leveraging hyperbolic collaborative learning from multi-ordered and partitioned neighborhoods, and regularizers inspired by socio-psychological factors. Our resulting formalism, Digraph Hyperbolic Networks (D-HYPR) - albeit conceptually simple - generalizes to digraphs where cycles and non-transitive relations are common, and is applicable to multiple downstream tasks including node classification, link presence prediction, and link property prediction. In order to assess the effectiveness of D-HYPR, extensive evaluations were performed across 8 real-world digraph datasets involving 21 prior techniques. D-HYPR statistically significantly outperforms the current state of the art. We release our code at https://github.com/hongluzhou/dhypr | ['Mubbasir Kapadia', 'Gerard de Melo', 'Zuohui Fu', 'Samuel S. Sohn', 'Advith Chegu', 'Honglu Zhou'] | 2021-12-22 | null | null | null | null | ['link-property-prediction'] | ['graphs'] | [ 1.74910761e-02 4.67756897e-01 -5.06009877e-01 -1.16344266e-01
-7.84609541e-02 -5.92177749e-01 7.31784284e-01 4.30888325e-01
2.03294784e-01 6.04737043e-01 4.69531566e-01 -7.11889446e-01
-6.36538386e-01 -1.34722650e+00 -3.97291273e-01 -3.28046441e-01
-6.68942869e-01 7.60483563e-01 1.59354091e-01 -2.43070647e-01
-8.62852857e-02 6.26660943e-01 -7.81095028e-01 -6.86515421e-02
6.58761799e-01 3.31049323e-01 1.66908838e-03 5.77477694e-01
2.77856857e-01 1.09104896e+00 -8.77945125e-02 -6.69194579e-01
2.20434576e-01 -3.42934936e-01 -8.80271196e-01 -2.07368791e-01
3.22228849e-01 -2.21058160e-01 -1.30813360e+00 8.31644893e-01
5.02680123e-01 2.88518757e-01 9.48078752e-01 -1.58617795e+00
-1.38774621e+00 1.09235179e+00 -6.36174798e-01 3.06873769e-01
1.13789566e-01 -1.00178592e-01 1.84183788e+00 -6.21422172e-01
8.08158338e-01 1.39962542e+00 8.88623714e-01 3.30247760e-01
-1.54762518e+00 -5.25575578e-01 3.61614227e-01 2.38131031e-01
-1.38938427e+00 -9.51284096e-02 8.86380255e-01 -3.95000637e-01
9.72369552e-01 1.52182251e-01 5.14733672e-01 1.28263652e+00
7.25079998e-02 5.88338733e-01 7.86117077e-01 -1.25373214e-01
1.95245463e-02 -3.59868348e-01 4.23324943e-01 8.73252630e-01
6.64455652e-01 4.64869477e-02 -3.19350183e-01 -1.09228611e-01
1.00307989e+00 9.79698822e-02 -1.06071286e-01 -6.11113012e-01
-8.24056804e-01 1.09221089e+00 1.01832938e+00 2.40932927e-01
-2.05575481e-01 1.71773449e-01 3.25492263e-01 4.95739609e-01
5.97532332e-01 4.34647501e-01 -1.13370540e-02 3.46801370e-01
-4.64460313e-01 4.90037613e-02 8.95685673e-01 1.03558242e+00
8.27546418e-01 7.74663016e-02 -1.27292112e-01 9.47071970e-01
2.48131871e-01 -6.57437090e-03 -3.02837398e-02 -6.92693472e-01
4.83328581e-01 6.87343538e-01 -5.94928026e-01 -1.54177904e+00
-8.18167269e-01 -6.55217290e-01 -1.23016787e+00 -3.65699530e-01
2.38973394e-01 -1.88891500e-01 -6.82373583e-01 1.85912585e+00
-6.09727465e-02 1.90576777e-01 -1.42693788e-01 6.28461123e-01
1.32126391e+00 5.86652935e-01 -1.14862528e-02 1.79180250e-01
8.70587587e-01 -8.81189823e-01 -3.19766819e-01 -3.15983534e-01
9.13021922e-01 -1.06215462e-01 8.64180386e-01 -2.59262659e-02
-9.67467666e-01 -1.42081156e-01 -8.84235263e-01 -3.68190855e-01
-5.11825204e-01 -1.98400825e-01 9.59930539e-01 4.30953801e-01
-1.61786914e+00 6.72853410e-01 -6.78017557e-01 -7.30905890e-01
4.30226803e-01 2.38044694e-01 -4.03391033e-01 -4.94387805e-01
-1.36259067e+00 6.67895436e-01 3.36497247e-01 -9.36118290e-02
-8.05179298e-01 -6.98967874e-01 -1.04520273e+00 3.48263651e-01
3.19201171e-01 -7.20862031e-01 6.31441832e-01 -3.91792923e-01
-9.34364200e-01 7.98948705e-01 2.33268037e-01 -5.76466918e-01
2.43498772e-01 2.32658952e-01 -4.47255999e-01 1.36769578e-01
5.65677471e-02 6.19295657e-01 2.76024371e-01 -1.05063522e+00
3.88411917e-02 -3.43352258e-01 4.92981404e-01 2.55631089e-01
-6.18479848e-01 -2.47577325e-01 -4.24620658e-01 -6.37134433e-01
2.40542054e-01 -1.18728542e+00 -2.00680345e-01 -1.94432706e-01
-7.12595642e-01 -5.56242168e-01 5.55961728e-01 -4.65834051e-01
1.36069381e+00 -1.84002602e+00 2.44490191e-01 5.57220817e-01
9.31916714e-01 -6.92901853e-03 -4.39884990e-01 1.00424397e+00
-1.56626597e-01 1.91922203e-01 3.50018069e-02 -3.34093750e-01
1.24403574e-01 2.83766359e-01 -1.43189773e-01 6.50504947e-01
3.19061205e-02 1.10260987e+00 -1.13255870e+00 -5.29425263e-01
-6.19506910e-02 4.43044156e-01 -7.39182830e-01 3.03980913e-02
-7.84345567e-02 1.13218948e-02 -2.86891669e-01 7.40491390e-01
5.51203966e-01 -9.20646250e-01 8.35029006e-01 5.89335859e-02
2.83558547e-01 4.64830607e-01 -8.30651581e-01 1.30577171e+00
-3.83430004e-01 9.51881051e-01 -2.71597654e-01 -1.06914735e+00
1.09613836e+00 -7.40667060e-02 3.91437382e-01 -5.30231535e-01
-7.40303248e-02 -7.29881674e-02 4.11459863e-01 -1.55829087e-01
4.00583267e-01 4.50595856e-01 2.98347652e-01 6.34401798e-01
2.20493466e-01 3.44528288e-01 3.62149298e-01 1.06193841e+00
1.66151345e+00 -2.51052022e-01 5.08987129e-01 -3.90057057e-01
-9.55891907e-02 -3.72657508e-01 4.74153548e-01 9.19648588e-01
-1.21826611e-01 6.85931087e-01 1.04470325e+00 -2.85446674e-01
-8.39870870e-01 -1.32614815e+00 8.39458033e-02 1.47198653e+00
1.58923835e-01 -7.82655358e-01 2.76825135e-03 -6.18780911e-01
2.11086616e-01 4.91168588e-01 -7.65603006e-01 -2.93332011e-01
-5.34526408e-01 -9.04605389e-01 5.13302922e-01 7.58848071e-01
1.24020033e-01 -9.36776817e-01 4.01364118e-01 1.53381094e-01
9.05411169e-02 -1.11817217e+00 -3.67908031e-01 1.05175804e-02
-8.48568678e-01 -1.16049755e+00 -3.73154402e-01 -9.41789925e-01
6.15799665e-01 5.70358872e-01 1.56072176e+00 3.34668010e-01
-1.60644144e-01 3.37319255e-01 -3.17841947e-01 1.93411186e-01
-1.35652021e-01 5.70211291e-01 -1.04791500e-01 -2.84366161e-01
1.71103179e-01 -1.16654873e+00 -6.96462691e-01 3.64335239e-01
-5.13644695e-01 3.94452326e-02 7.18500555e-01 6.74580216e-01
1.92158848e-01 5.80887310e-03 8.59653950e-01 -1.51657343e+00
9.71340358e-01 -1.08361220e+00 -2.80792713e-01 2.58615732e-01
-8.12766433e-01 -2.49498770e-01 5.61683774e-01 -2.80084550e-01
-6.28752530e-01 -6.05902135e-01 1.66560903e-01 -3.20213497e-01
2.93044955e-01 9.25703168e-01 -1.75640341e-02 -2.85464060e-02
7.95143723e-01 -9.24526453e-02 -6.85343593e-02 -2.54813313e-01
5.16611397e-01 2.62919366e-01 2.75946587e-01 -6.07210398e-01
9.07043576e-01 2.42733866e-01 3.76545072e-01 -9.05851781e-01
-3.93078238e-01 -3.07485819e-01 -5.51609576e-01 -1.82629392e-01
4.55656797e-01 -9.38400269e-01 -5.71114480e-01 1.10501647e-01
-7.61114895e-01 -5.84395349e-01 1.56127810e-01 3.38448703e-01
-1.49368793e-01 3.87639016e-01 -9.99939084e-01 -5.89137495e-01
-2.50531793e-01 -4.86942172e-01 5.00487506e-01 -1.18406512e-01
-3.24142337e-01 -1.59077442e+00 2.33124912e-01 1.54890448e-01
4.09020990e-01 2.52803177e-01 1.46573961e+00 -8.90336335e-01
-6.41781151e-01 5.99714741e-02 -6.68237031e-01 -1.30663827e-01
1.85632452e-01 7.28751719e-02 -4.98189151e-01 -4.92510736e-01
-1.04405689e+00 -5.51996708e-01 1.22533357e+00 3.94519866e-01
1.12981331e+00 -3.79080236e-01 -6.12348855e-01 4.92688954e-01
1.23298764e+00 -3.03483695e-01 6.02755547e-01 -1.81288626e-02
1.01321065e+00 6.69265509e-01 5.96500039e-02 3.58995467e-01
9.24079895e-01 4.15198505e-01 6.16940141e-01 -1.08264551e-01
-5.35069883e-01 -6.05983913e-01 5.75031154e-02 8.56068790e-01
-2.31226653e-01 -8.88658345e-01 -1.08038914e+00 7.18460798e-01
-1.95452452e+00 -8.42579782e-01 -2.86820322e-01 1.73042774e+00
4.53919530e-01 1.52579203e-01 4.36433643e-01 -2.20424682e-01
9.75179076e-01 6.02709174e-01 -6.34977221e-01 -1.33328468e-01
-1.78161979e-01 7.95928910e-02 4.32153791e-01 5.08284509e-01
-9.74112213e-01 9.88527358e-01 5.60994244e+00 6.16684914e-01
-5.38596630e-01 -6.86312243e-02 5.09775102e-01 -3.31169032e-02
-6.21550560e-01 9.22653005e-02 -4.26826328e-01 7.18626454e-02
7.18863666e-01 -3.50447863e-01 7.26917565e-01 6.94686115e-01
-1.69654965e-01 5.46015561e-01 -1.32239258e+00 8.72818530e-01
9.05075483e-03 -1.38719583e+00 8.17588419e-02 2.49236301e-01
9.82230365e-01 4.74069625e-01 3.57545674e-01 5.29434264e-01
1.15856516e+00 -1.21040523e+00 -2.19490826e-02 3.28759462e-01
6.17756307e-01 -5.42015672e-01 4.14033175e-01 3.36430185e-02
-1.40861917e+00 -1.11511417e-01 -5.50836384e-01 -1.94539338e-01
-1.63814127e-01 4.30890620e-01 -1.00854266e+00 6.98528469e-01
4.23917413e-01 1.31059289e+00 -8.53799403e-01 8.29646111e-01
-2.16781244e-01 8.27381849e-01 -2.19919398e-01 -1.53555930e-01
3.31988603e-01 -3.25245827e-01 5.48089564e-01 1.15504158e+00
1.09980658e-01 -2.73949392e-02 2.41144001e-01 9.09356415e-01
-6.58130586e-01 1.58096284e-01 -1.08089387e+00 -3.20573688e-01
8.39416623e-01 1.34057152e+00 -8.62480402e-01 2.58509606e-01
-5.23927093e-01 4.10299718e-01 1.03140807e+00 5.81723928e-01
-6.65970862e-01 -3.05892289e-01 5.20416737e-01 4.80463773e-01
1.27257556e-01 -3.67382765e-01 -1.12597965e-01 -9.94451702e-01
-3.58870536e-01 -5.16102195e-01 8.47293675e-01 -6.37470067e-01
-1.83187044e+00 6.06428742e-01 5.46206348e-02 -7.80172408e-01
-2.09585100e-01 -5.82118928e-01 -9.72160697e-01 5.10477960e-01
-1.31896698e+00 -1.47178805e+00 -2.97021687e-01 6.93667471e-01
-5.32810986e-02 -2.76619971e-01 7.04385221e-01 2.52868235e-01
-6.20905042e-01 6.89742923e-01 9.44547802e-02 3.81831437e-01
4.88911122e-01 -1.47038221e+00 6.78182423e-01 7.59732604e-01
2.93673575e-01 7.39629924e-01 4.02577817e-01 -7.77487338e-01
-1.28804016e+00 -1.43884552e+00 7.81045258e-01 -2.21469790e-01
1.08368587e+00 -6.71033740e-01 -1.00044823e+00 1.22898817e+00
9.64803994e-02 2.90546596e-01 7.10021198e-01 8.95097196e-01
-6.51693881e-01 -8.84213224e-02 -8.82612765e-01 9.04830992e-01
1.85320091e+00 -6.48856521e-01 -7.84290656e-02 6.61478758e-01
6.96710646e-01 -2.25747060e-02 -1.01766312e+00 3.57287139e-01
3.42981875e-01 -8.75500023e-01 1.06436861e+00 -8.72112870e-01
5.19350111e-01 1.74351364e-01 -2.25050315e-01 -1.49743724e+00
-8.56615961e-01 -6.09391391e-01 -3.66570652e-01 1.24353492e+00
4.81564254e-01 -9.91802454e-01 9.74398971e-01 2.39223406e-01
-1.22187898e-01 -7.77698576e-01 -5.93144298e-01 -7.03485668e-01
2.20150396e-01 6.45485520e-02 5.38792849e-01 1.28675818e+00
4.81432304e-02 7.23023534e-01 -4.96862769e-01 3.48078132e-01
6.95459485e-01 3.30113828e-01 7.39511788e-01 -1.54896128e+00
-4.14173812e-01 -5.06075382e-01 -6.18847311e-01 -1.11103201e+00
5.41978359e-01 -1.57484937e+00 -5.49239874e-01 -1.98686743e+00
3.22028100e-01 -6.09629512e-01 -3.20746839e-01 5.20315349e-01
3.58999893e-02 2.65548646e-01 5.35737239e-02 2.67432153e-01
-7.64223218e-01 5.80104887e-01 1.12561262e+00 -2.81536102e-01
-2.06256747e-01 9.87613276e-02 -9.47052479e-01 4.48576897e-01
1.06583285e+00 -2.67951429e-01 -9.26933467e-01 -4.62434590e-01
5.55213213e-01 7.82008395e-02 5.12186408e-01 -7.09290028e-01
4.29492950e-01 1.01081364e-01 1.98219329e-01 -4.46837366e-01
3.01346272e-01 -3.67148042e-01 6.58847019e-02 1.95452020e-01
-6.34627044e-01 2.15890348e-01 -3.98715883e-01 8.98423254e-01
1.51061594e-01 1.36215642e-01 4.40084606e-01 -3.78382914e-02
-5.69661915e-01 5.94868720e-01 -5.06991036e-02 4.30799872e-01
7.18609869e-01 -3.35458592e-02 -1.04307497e+00 -7.79821575e-01
-7.86027253e-01 4.00823802e-01 2.06148863e-01 4.55271244e-01
7.11006939e-01 -1.34704399e+00 -9.18540835e-01 -5.40988632e-02
1.60517111e-01 -1.68109477e-01 2.65301675e-01 7.14539289e-01
-4.70157355e-01 2.48775572e-01 -8.66599232e-02 -2.47729048e-01
-1.20033598e+00 5.63028812e-01 1.98839635e-01 -5.70478261e-01
-8.91222596e-01 6.34492755e-01 4.00290698e-01 -6.52568758e-01
1.97263852e-01 -6.28517196e-02 -3.15022528e-01 -5.99329360e-02
-1.05877109e-01 7.17666805e-01 -5.13314195e-02 -4.71553981e-01
-3.07487398e-01 3.88268381e-02 -4.39902633e-01 3.89781952e-01
1.48644078e+00 2.32031289e-02 -3.06877196e-01 1.85427696e-01
1.33303642e+00 -1.76325500e-01 -1.01980329e+00 -5.33737600e-01
-2.34401468e-02 -2.32126817e-01 -1.34593561e-01 -4.40757304e-01
-1.15673006e+00 6.24290705e-01 -1.33601367e-01 6.58333182e-01
7.34002054e-01 4.48033750e-01 3.86665344e-01 6.29653513e-01
2.14289606e-01 -5.65848172e-01 2.58294284e-01 6.16111517e-01
8.97192478e-01 -1.10727298e+00 2.29888260e-01 -6.78702295e-01
-5.17565072e-01 9.14543152e-01 8.10932100e-01 -3.55772525e-01
9.97256517e-01 -1.35999620e-01 -6.22061074e-01 -4.04039949e-01
-9.65221822e-01 -2.51820445e-01 2.68290371e-01 8.59793127e-01
4.83270526e-01 3.04007977e-01 6.92470595e-02 3.19251150e-01
-2.86284983e-01 -5.07052779e-01 6.45609617e-01 6.04744554e-01
-7.64341727e-02 -8.34198773e-01 3.44523251e-01 7.47624278e-01
-1.51056692e-01 -2.86135882e-01 -8.29445004e-01 1.04355156e+00
-3.70204926e-01 9.36047316e-01 2.27355585e-01 -4.55482990e-01
-4.69886810e-02 -5.36237419e-01 3.48271042e-01 -7.08726227e-01
-2.86552787e-01 -3.81086826e-01 4.07019407e-01 -3.04979533e-01
-1.55841693e-01 -5.80409586e-01 -9.65650439e-01 -8.74782860e-01
3.85638350e-03 -3.04165259e-02 -3.87747735e-02 2.90657133e-01
4.97506231e-01 5.79702139e-01 5.44864118e-01 -5.09501278e-01
-3.20531398e-01 -1.07159126e+00 -9.21451151e-01 2.15760693e-01
1.32569447e-01 -7.23092794e-01 -2.94389457e-01 -6.98721111e-01] | [7.018763065338135, 6.224867343902588] |
d9ce3be7-aa76-444b-9a25-e903fc35e5cb | district-dialogue-state-tracking-with | 2212.02851 | null | https://arxiv.org/abs/2212.02851v1 | https://arxiv.org/pdf/2212.02851v1.pdf | DiSTRICT: Dialogue State Tracking with Retriever Driven In-Context Tuning | Dialogue State Tracking (DST), a key component of task-oriented conversation systems, represents user intentions by determining the values of pre-defined slots in an ongoing dialogue. Existing approaches use hand-crafted templates and additional slot information to fine-tune and prompt large pre-trained language models and elicit slot values from the dialogue context. Significant manual effort and domain knowledge is required to design effective prompts, limiting the generalizability of these approaches to new domains and tasks. In this work, we propose DiSTRICT, a generalizable in-context tuning approach for DST that retrieves highly relevant training examples for a given dialogue to fine-tune the model without any hand-crafted templates. Experiments with the MultiWOZ benchmark datasets show that DiSTRICT outperforms existing approaches in various zero-shot and few-shot settings using a much smaller model, thereby providing an important advantage for real-world deployments that often have limited resource availability. | ['Vatche Isahagian', 'Evelyn Duesterwald', 'Praveen Venkateswaran'] | 2022-12-06 | null | null | null | null | ['dialogue-state-tracking'] | ['natural-language-processing'] | [ 3.01564962e-01 3.36209536e-01 -3.97777945e-01 -5.97319901e-01
-7.88673520e-01 -8.07044089e-01 9.80480433e-01 1.41976118e-01
-4.96752113e-01 9.38113689e-01 6.25125945e-01 -2.91369110e-01
5.00169657e-02 -4.79086101e-01 2.50688076e-01 -1.12336963e-01
1.90023437e-01 1.13640749e+00 6.60307705e-01 -9.17373657e-01
5.51292539e-01 2.99866069e-02 -1.34722936e+00 5.29588461e-01
5.00570714e-01 5.93146980e-01 3.71317357e-01 9.29126263e-01
-5.74496150e-01 8.80362153e-01 -9.42838192e-01 -2.44800285e-01
8.39760676e-02 -6.24094129e-01 -1.26079631e+00 2.45313361e-01
-1.23155192e-02 -4.88173962e-01 -3.03827882e-01 4.88069952e-01
6.43654168e-01 6.78560674e-01 2.49840587e-01 -9.58822191e-01
1.76377624e-01 7.45001733e-01 -5.57328314e-02 3.67117912e-01
8.30705643e-01 2.62783378e-01 1.09397578e+00 -5.53144157e-01
6.17964089e-01 1.54025161e+00 5.55544913e-01 9.60158288e-01
-1.16999793e+00 -1.63429633e-01 1.98633790e-01 -7.34325796e-02
-8.86278093e-01 -8.78848314e-01 5.17513514e-01 -2.13087305e-01
1.43176115e+00 5.90113461e-01 3.71693105e-01 1.06988418e+00
-2.77463347e-01 7.27815986e-01 1.05850804e+00 -6.81682467e-01
1.90606922e-01 2.65989184e-01 2.28195667e-01 4.71277356e-01
-5.23600459e-01 -3.90014797e-01 -8.86247754e-01 -6.89728737e-01
6.76858008e-01 -1.92242786e-01 -1.70911655e-01 -2.32934043e-01
-1.10725236e+00 9.65861142e-01 -2.64033169e-01 3.33623916e-01
-2.08096966e-01 -3.42776060e-01 7.79636383e-01 4.26035255e-01
3.31429273e-01 9.02977049e-01 -6.67785227e-01 -1.12302077e+00
-6.18479371e-01 3.96195412e-01 1.39272034e+00 1.03061914e+00
6.95105970e-01 -3.07569236e-01 -9.27004814e-01 1.37282026e+00
-8.56798217e-02 4.85002622e-02 6.67403877e-01 -1.18075860e+00
5.94943404e-01 6.98578179e-01 5.77709198e-01 -4.49328452e-01
-3.73163998e-01 1.50161982e-01 -8.25725496e-02 -3.74038368e-01
6.38231099e-01 -3.28739583e-01 -5.61356187e-01 1.68585145e+00
5.67176163e-01 -1.68302283e-01 2.15948835e-01 8.47670317e-01
6.73502505e-01 6.04849160e-01 1.23716563e-01 -3.39304954e-01
1.62126076e+00 -1.06421602e+00 -7.97227323e-01 -6.45838976e-01
5.48244357e-01 -9.03373659e-01 1.54505348e+00 1.80735305e-01
-9.10624385e-01 -2.62096435e-01 -8.09450388e-01 -6.52082637e-02
-8.57979506e-02 -1.76805899e-01 7.09879041e-01 6.47639453e-01
-7.93022275e-01 4.41694647e-01 -4.55305159e-01 -7.03040302e-01
-1.58085033e-01 2.17144728e-01 5.32909594e-02 1.15371227e-01
-1.51086426e+00 1.15589845e+00 2.00865135e-01 -2.14157805e-01
-6.20367587e-01 -3.47033143e-01 -7.44073570e-01 1.24206804e-01
8.89922619e-01 -3.14729869e-01 2.31364632e+00 -5.04799426e-01
-2.17602992e+00 6.87232137e-01 -3.48536938e-01 -4.25743371e-01
3.66204500e-01 -1.03134334e-01 -4.95118797e-02 -2.55683083e-02
-6.39030710e-02 3.25603634e-01 6.91705287e-01 -7.84940362e-01
-8.50094795e-01 -2.46968074e-03 5.55314183e-01 6.70853138e-01
-3.67329389e-01 2.07819119e-01 -7.59744704e-01 -1.32308573e-01
-2.62508661e-01 -9.18683290e-01 -4.37735677e-01 -5.17157555e-01
-2.93400407e-01 -6.88086569e-01 8.11775386e-01 -3.07709992e-01
1.61303997e+00 -1.69670284e+00 -1.71406165e-01 -9.69027802e-02
1.24825604e-01 5.84396660e-01 -1.65513098e-01 8.77580106e-01
6.81900442e-01 -2.32618049e-01 3.29239927e-02 -2.10895792e-01
2.99594969e-01 3.69453043e-01 -8.80615488e-02 -2.01822892e-01
-1.11509740e-01 7.38249481e-01 -1.13267672e+00 -5.62504888e-01
4.96520370e-01 2.07475536e-02 -5.40624499e-01 7.68056035e-01
-7.78081954e-01 2.49765664e-01 -6.01048112e-01 8.64967629e-02
-2.25855336e-02 -3.33049059e-01 6.63294196e-01 1.71940327e-01
-9.51525848e-03 1.06719255e+00 -8.33924234e-01 1.71951663e+00
-6.96475267e-01 5.50170898e-01 -9.70383582e-04 -6.21171951e-01
9.59623218e-01 4.76718813e-01 2.98394799e-01 -6.57010674e-01
2.22435072e-01 -1.35399461e-01 -8.72880593e-02 -6.18579805e-01
8.46880615e-01 -8.25095922e-02 -5.87982118e-01 8.94376516e-01
1.45500749e-01 -1.42237365e-01 3.68834764e-01 4.32765305e-01
1.31230354e+00 -3.64022225e-01 6.90757692e-01 8.25546905e-02
5.75494349e-01 2.83167601e-01 3.54706675e-01 1.02418160e+00
-3.36303383e-01 2.15270191e-01 6.22119069e-01 -3.95909965e-01
-9.72168028e-01 -4.42909718e-01 1.04420744e-01 2.01413488e+00
-5.51738143e-02 -7.21619308e-01 -9.25274253e-01 -9.06753480e-01
-5.54791503e-02 9.34145927e-01 -3.97665918e-01 -7.89779052e-02
-5.78689158e-01 -3.97910863e-01 4.28715199e-01 1.75783426e-01
4.95611936e-01 -1.27222645e+00 -7.58593261e-01 5.80788970e-01
-5.83620667e-01 -1.14705610e+00 -8.78368020e-01 1.19796775e-01
-5.19247472e-01 -1.00233495e+00 -5.67556322e-01 -5.53793311e-01
2.98388183e-01 3.88949722e-01 1.35214198e+00 1.37048468e-01
-6.41172901e-02 5.33405542e-01 -4.28816497e-01 -1.43401772e-01
-7.64758587e-01 3.88778299e-01 -1.14477858e-01 -1.89356536e-01
8.10106993e-01 -2.07896814e-01 -3.79991978e-01 7.32135415e-01
-4.41105396e-01 1.99460849e-01 2.17327058e-01 1.17339969e+00
-1.45028442e-01 -3.44134748e-01 7.36088753e-01 -1.26800096e+00
1.47775459e+00 -1.91709623e-01 -4.17492360e-01 4.56310362e-01
-7.86746204e-01 2.40096286e-01 2.60492444e-01 -5.49382567e-01
-1.41755545e+00 -1.41897753e-01 -9.75707024e-02 1.00162573e-01
-1.92134514e-01 3.75484437e-01 2.36605499e-02 2.40979344e-01
9.29202974e-01 1.44763842e-01 6.31098002e-02 -4.80263919e-01
3.85029852e-01 1.13525236e+00 4.19392288e-01 -8.89240384e-01
3.68855506e-01 -1.14841282e-01 -7.46325254e-01 -7.05027640e-01
-1.05681765e+00 -9.48009551e-01 -5.29202223e-01 -2.19078660e-01
3.48476797e-01 -6.10183895e-01 -8.92819226e-01 8.45723320e-03
-1.06697702e+00 -9.02073264e-01 -1.34864688e-01 8.23288560e-02
-6.06879711e-01 3.39662880e-01 -6.93903685e-01 -1.12247539e+00
-5.90651929e-01 -8.99896920e-01 1.05400169e+00 4.31394398e-01
-9.11712945e-01 -9.60700870e-01 1.92919314e-01 6.96413279e-01
6.59602225e-01 -5.06086409e-01 6.83585107e-01 -1.20993888e+00
-1.96590737e-01 -2.34663054e-01 4.01210301e-02 -4.00579758e-02
2.98358530e-01 -3.91008466e-01 -9.88132179e-01 -2.58506417e-01
-2.19430536e-01 -6.69074953e-01 3.16100866e-01 -1.49953887e-01
6.34432197e-01 -5.62643349e-01 -1.83964401e-01 -2.47683540e-01
5.87035716e-01 3.76788318e-01 1.89976558e-01 4.10266072e-01
6.26014695e-02 6.76689923e-01 1.12742424e+00 1.00621188e+00
5.46798170e-01 1.01830876e+00 -3.14510792e-01 1.91645712e-01
6.10769577e-02 -3.30282629e-01 1.88718632e-01 5.17457783e-01
2.31521472e-01 -2.80433834e-01 -9.18124676e-01 6.05689287e-01
-2.08864284e+00 -1.05200648e+00 5.42694569e-01 2.05795622e+00
1.54115188e+00 3.81549656e-01 3.19744498e-01 -1.94462001e-01
7.40390420e-01 2.87347615e-01 -6.24495149e-01 -6.12117648e-01
4.29543465e-01 2.40553617e-01 5.35566621e-02 7.68475294e-01
-8.54386270e-01 1.38845694e+00 6.67534399e+00 6.94600046e-01
-9.40306365e-01 1.31017715e-01 3.40562433e-01 -1.54298231e-01
-7.72181824e-02 3.84859145e-02 -9.81464028e-01 2.03515723e-01
1.14338112e+00 -5.27358711e-01 7.10304618e-01 9.32425916e-01
3.78533930e-01 -2.51132697e-01 -1.04345953e+00 7.30856001e-01
-4.70151342e-02 -1.31258106e+00 -3.92883092e-01 -2.42600322e-01
3.63006145e-01 -7.42296726e-02 -4.11391079e-01 7.95547903e-01
7.63356864e-01 -5.77114701e-01 1.41679928e-01 2.12448433e-01
7.21105278e-01 -4.97943729e-01 5.28980136e-01 5.94013691e-01
-7.57340193e-01 4.02995124e-02 -3.25823426e-01 -1.98839813e-01
2.49224782e-01 3.40872481e-02 -1.97158706e+00 -1.82620585e-01
2.45553106e-01 1.15910836e-01 -2.31494129e-01 6.25479579e-01
-1.13332935e-01 6.89150214e-01 -3.15658122e-01 -4.46933776e-01
2.56625265e-01 2.34524786e-01 5.27026236e-01 1.39248550e+00
-1.61536187e-01 3.37326646e-01 5.03618300e-01 3.75242174e-01
-9.08745080e-02 8.12083259e-02 -2.72808582e-01 -1.49693683e-01
1.07421815e+00 1.42955899e+00 -4.89281029e-01 -6.42796874e-01
-3.38904828e-01 9.00777459e-01 2.36827508e-01 1.76612392e-01
-4.47510898e-01 -4.29220140e-01 8.38446498e-01 -6.24302551e-02
1.08006582e-01 -1.69912681e-01 -7.92672485e-02 -1.03802454e+00
-3.73220205e-01 -1.30463648e+00 5.15183449e-01 -5.02060235e-01
-1.09017980e+00 7.16441512e-01 1.32952750e-01 -9.05185163e-01
-1.01196313e+00 -2.29932159e-01 -6.41736329e-01 9.25634325e-01
-1.22315156e+00 -6.85499072e-01 -3.19743037e-01 5.92658341e-01
1.21321726e+00 -2.62005746e-01 1.25451207e+00 -2.08088055e-01
-5.65394819e-01 6.05841279e-01 -2.73148060e-01 2.03011557e-01
1.10982752e+00 -1.32341516e+00 7.35163093e-01 3.36926430e-01
-1.22932926e-01 8.53096426e-01 1.09644246e+00 -5.97840726e-01
-1.22966552e+00 -5.45461774e-01 1.13724756e+00 -6.44512892e-01
5.59434712e-01 -5.12860596e-01 -1.04560554e+00 4.98977572e-01
4.67584103e-01 -4.97488856e-01 8.72230947e-01 7.28539586e-01
-1.47978172e-01 2.02764779e-01 -1.13294876e+00 6.42821550e-01
8.42282414e-01 -7.59645641e-01 -9.59897637e-01 3.83052766e-01
6.66622162e-01 -8.63696337e-01 -8.96579862e-01 -1.84903089e-02
5.94281197e-01 -7.98238933e-01 7.55001843e-01 -7.84773648e-01
-1.99510112e-01 1.07599162e-01 2.12662175e-01 -1.34281766e+00
-3.74992490e-02 -1.25534105e+00 -1.91092804e-01 1.15632856e+00
5.87255061e-01 -4.65101153e-01 7.89473593e-01 1.35675621e+00
-4.77665402e-02 -5.25090158e-01 -7.60240495e-01 -3.33297282e-01
-4.77559686e-01 -1.35654852e-01 7.88346529e-01 9.06821847e-01
7.91786134e-01 1.05492592e+00 -6.34059310e-01 -2.65904099e-01
1.13178492e-01 1.06535904e-01 1.19734824e+00 -1.16365111e+00
-3.55030388e-01 -2.38785028e-01 2.69520342e-01 -1.40522337e+00
4.52747792e-02 -1.70564964e-01 5.17532825e-01 -1.27490258e+00
-2.63589956e-02 -6.17569447e-01 -3.46705168e-02 7.33117223e-01
-4.08081681e-01 -4.02238131e-01 1.02925248e-01 1.77069724e-01
-1.08960605e+00 4.42035228e-01 1.14134836e+00 -4.13849615e-02
-6.16007447e-01 4.12381083e-01 -7.66532362e-01 3.46776366e-01
9.54324603e-01 -2.54875898e-01 -7.74778664e-01 3.43995057e-02
-2.46928915e-01 4.82541114e-01 -3.01061034e-01 -6.70650721e-01
4.89854902e-01 -6.62025988e-01 -1.92023426e-01 -2.71972477e-01
6.60036683e-01 -3.77170771e-01 -4.31875259e-01 -5.59454821e-02
-9.92779016e-01 -1.45841986e-01 2.32201025e-01 4.97233450e-01
3.13832685e-02 -4.37893122e-01 5.70697308e-01 -3.90708476e-01
-9.79982197e-01 -2.34241318e-02 -7.19645023e-01 2.79056907e-01
8.13528001e-01 -8.90220702e-02 -3.91902506e-01 -7.86215246e-01
-5.06394148e-01 4.01394099e-01 2.17034519e-01 6.91446364e-01
3.31812143e-01 -9.23084974e-01 -3.21738154e-01 7.26680458e-02
4.29239959e-01 -1.78347886e-01 1.78103566e-01 3.67718667e-01
4.22603115e-02 8.02389443e-01 -1.00719951e-01 -4.52574670e-01
-1.59366083e+00 2.01165646e-01 3.19489181e-01 -5.93091071e-01
-5.36782324e-01 7.06230521e-01 -2.94360906e-01 -6.45016909e-01
5.15127480e-01 -9.85280573e-02 -3.65386218e-01 1.05228215e-01
7.53367782e-01 -3.09558194e-02 1.14652567e-01 -3.38700533e-01
-6.02123588e-02 -2.99722463e-01 -3.25822800e-01 -4.52746570e-01
9.78911579e-01 -4.11949158e-01 1.56950265e-01 6.02727354e-01
6.64547205e-01 -1.53608710e-01 -1.28242600e+00 -9.49024558e-01
4.40349877e-01 -5.67162156e-01 -2.12843176e-02 -1.11374736e+00
-2.00398251e-01 5.46555161e-01 7.23697767e-02 4.48122501e-01
7.83691406e-01 5.25810104e-03 1.00298071e+00 9.43719029e-01
6.02284729e-01 -1.46233273e+00 3.93311590e-01 9.35746431e-01
7.02171683e-01 -1.17904699e+00 -3.11141014e-01 -1.84237778e-01
-1.24974489e+00 1.08593845e+00 9.93081629e-01 5.93746066e-01
1.41826883e-01 1.70254886e-01 4.79934454e-01 -1.57093793e-01
-1.39712954e+00 -4.02996451e-01 9.21329260e-02 4.02368426e-01
6.54937208e-01 -5.32612111e-03 -3.38425368e-01 5.88120639e-01
-1.92957699e-01 -1.02335423e-01 4.40038472e-01 1.21552968e+00
-8.51958275e-01 -1.54918087e+00 -2.60438025e-01 6.60484433e-01
-1.33661479e-01 -4.92525436e-02 -7.22743869e-01 4.89726573e-01
-6.75402284e-01 1.27042079e+00 -4.32673506e-02 -3.88525307e-01
6.10261738e-01 5.52865088e-01 1.75370887e-01 -1.13030159e+00
-9.98762369e-01 1.19258419e-01 9.63163555e-01 -4.59784061e-01
-1.92404807e-01 -6.27420723e-01 -1.07142055e+00 -3.97789598e-01
-4.09863979e-01 6.39701664e-01 2.23980770e-01 9.72602248e-01
4.71063137e-01 2.73292333e-01 6.44939423e-01 -7.21586525e-01
-9.61923659e-01 -1.50910699e+00 -1.26173973e-01 4.37719047e-01
2.10334599e-01 -7.47289181e-01 -5.69335818e-02 -2.55503535e-01] | [12.896936416625977, 7.853987693786621] |
54cbb521-0bab-4c15-9f5c-174dbb0ee8c9 | single-domain-generalization-for-lidar | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Kim_Single_Domain_Generalization_for_LiDAR_Semantic_Segmentation_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Kim_Single_Domain_Generalization_for_LiDAR_Semantic_Segmentation_CVPR_2023_paper.pdf | Single Domain Generalization for LiDAR Semantic Segmentation | With the success of the 3D deep learning models, various perception technologies for autonomous driving have been developed in the LiDAR domain. While these models perform well in the trained source domain, they struggle in unseen domains with a domain gap. In this paper, we propose a single domain generalization method for LiDAR semantic segmentation (DGLSS) that aims to ensure good performance not only in the source domain but also in the unseen domain by learning only on the source domain. We mainly focus on generalizing from a dense source domain and target the domain shift from different LiDAR sensor configurations and scene distributions. To this end, we augment the domain to simulate the unseen domains by randomly subsampling the LiDAR scans. With the augmented domain, we introduce two constraints for generalizable representation learning: sparsity invariant feature consistency (SIFC) and semantic correlation consistency (SCC). The SIFC aligns sparse internal features of the source domain with the augmented domain based on the feature affinity. For SCC, we constrain the correlation between class prototypes to be similar for every LiDAR scan. We also establish a standardized training and evaluation setting for DGLSS. With the proposed evaluation setting, our method showed improved performance in the unseen domains compared to other baselines. Even without access to the target domain, our method performed better than the domain adaptation method. The code is available at https://github.com/gzgzys9887/DGLSS. | ['Kuk-Jin Yoon', 'Changgyoon Oh', 'Yoonsu Kang', 'Hyeonseong Kim'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['lidar-semantic-segmentation'] | ['computer-vision'] | [ 2.76883751e-01 4.66919206e-02 -1.67020395e-01 -7.20457137e-01
-5.60596943e-01 -5.51617324e-01 4.79693919e-01 -1.63205668e-01
-2.45304585e-01 5.96366227e-01 -1.25777468e-01 3.75254378e-02
-5.83933927e-02 -7.95454144e-01 -8.93561721e-01 -5.43545365e-01
3.50555718e-01 7.34102190e-01 6.80401087e-01 -1.52130798e-01
-1.52187413e-02 5.36827505e-01 -1.81046081e+00 1.38081416e-01
1.05678093e+00 1.01879489e+00 5.64557731e-01 7.96810463e-02
-1.83104649e-01 2.81683449e-02 -4.18279886e-01 -4.21596803e-02
7.62029707e-01 -8.61419216e-02 -5.08801281e-01 1.39112338e-01
6.98239267e-01 -2.93421954e-01 -3.45910370e-01 1.17776084e+00
5.20407438e-01 2.42817372e-01 5.67103386e-01 -1.55863869e+00
-4.61263150e-01 1.05446547e-01 -5.20228922e-01 7.48120546e-02
-1.72110461e-02 1.15827605e-01 7.32490957e-01 -8.86120915e-01
7.57156193e-01 1.25427055e+00 5.83544910e-01 5.94876111e-01
-1.21360922e+00 -1.11295664e+00 4.31854546e-01 3.21702808e-01
-1.35634482e+00 -3.28748584e-01 9.08551633e-01 -4.82362598e-01
5.37395597e-01 -2.93116242e-01 3.85277122e-01 1.34036374e+00
-9.38838646e-02 7.91908681e-01 1.07100773e+00 1.07577574e-02
6.46792769e-01 2.04822987e-01 2.68762290e-01 2.68402249e-01
5.51789582e-01 3.08294237e-01 -6.59448326e-01 1.38729870e-01
5.48532605e-01 -1.33631796e-01 -2.74680834e-02 -9.17334616e-01
-8.83234441e-01 9.47972000e-01 4.56878752e-01 -2.54811114e-03
-9.14620683e-02 -1.59197345e-01 3.52775544e-01 9.26319361e-02
3.41328144e-01 1.95186347e-01 -5.12132466e-01 1.29661098e-01
-8.07127118e-01 3.32918704e-01 5.98900855e-01 1.37296808e+00
1.18219757e+00 -2.84838881e-02 9.09724981e-02 8.94456565e-01
8.42471048e-02 8.59915733e-01 4.44167107e-01 -1.10563016e+00
5.85396767e-01 5.61242998e-01 3.81034575e-02 -6.95222855e-01
-3.58661056e-01 -5.23803413e-01 -6.12536430e-01 1.87833235e-01
3.81973535e-01 -4.01204303e-02 -1.28563821e+00 1.99203193e+00
4.36288387e-01 4.26836222e-01 2.37870499e-01 8.64983201e-01
9.20931458e-01 4.06141460e-01 -2.16215365e-02 3.32463205e-01
9.13388133e-01 -6.92383826e-01 -3.13494563e-01 -7.88677514e-01
5.85793912e-01 -4.07097369e-01 1.16096234e+00 2.06397697e-01
-5.77979147e-01 -9.56206322e-01 -1.21778417e+00 -7.06485435e-02
-3.84254575e-01 5.64791225e-02 3.30624700e-01 3.80629092e-01
-8.69077504e-01 3.09409082e-01 -8.78677607e-01 -6.49879038e-01
5.81913233e-01 2.22323924e-01 -4.56356257e-01 -4.80725467e-01
-1.14653170e+00 7.64363945e-01 6.00867331e-01 -2.72514224e-01
-9.99504507e-01 -8.39868128e-01 -1.14176643e+00 -2.08182514e-01
3.10332477e-01 -5.59708297e-01 1.13302374e+00 -7.01352060e-01
-1.18598998e+00 9.77914572e-01 -3.72904748e-01 -6.36575520e-01
3.73700798e-01 -2.97082573e-01 -3.43854755e-01 1.66499093e-02
4.98901308e-01 1.15234160e+00 8.82203996e-01 -1.61488199e+00
-7.41622746e-01 -4.35420156e-01 -2.06558257e-01 2.73674846e-01
-5.53365983e-02 -7.78130114e-01 -3.29657972e-01 -3.45248669e-01
4.53141749e-01 -1.02381170e+00 -5.84111288e-02 4.99303192e-02
-1.43095076e-01 -1.10403910e-01 1.23522711e+00 -2.55927801e-01
5.56628704e-01 -2.59697628e+00 -4.58651073e-02 9.95181873e-02
3.74927260e-02 3.17387402e-01 -3.84288162e-01 7.91507661e-02
-3.46289054e-02 -2.12755695e-01 -5.42587399e-01 -3.73857617e-01
-2.01977957e-02 7.81396627e-01 -5.23766041e-01 3.52990150e-01
3.88300419e-01 7.46624887e-01 -1.00271797e+00 -3.61392617e-01
3.59966099e-01 1.97350591e-01 -6.89419270e-01 1.46780908e-01
-3.06594789e-01 8.02460253e-01 -5.24552822e-01 5.80627441e-01
1.21118522e+00 -7.12788478e-03 -3.68630327e-02 -2.54428655e-01
1.00572646e-01 1.13395847e-01 -1.28116941e+00 2.08119941e+00
-3.50265861e-01 4.27690446e-01 1.42452583e-01 -1.23028541e+00
1.26576841e+00 -2.39998892e-01 5.18040240e-01 -8.72450590e-01
-8.42386037e-02 4.33353901e-01 -3.40509675e-02 -3.15107197e-01
3.84408325e-01 -3.11889917e-01 -1.64463773e-01 3.41494866e-02
3.19261014e-01 -6.93256676e-01 -1.52187021e-02 1.44335479e-01
8.30946028e-01 2.30995879e-01 8.44259635e-02 -1.91479415e-01
3.21870774e-01 2.06458151e-01 7.99459398e-01 6.30706787e-01
-3.88417482e-01 7.05165029e-01 1.27403498e-01 -2.24841878e-01
-9.10509348e-01 -1.48412585e+00 -4.89526272e-01 7.85946667e-01
8.35712314e-01 -6.61414191e-02 -5.08268535e-01 -8.56561601e-01
4.32223886e-01 9.58438158e-01 -3.99316430e-01 -4.95206565e-01
-4.48147178e-01 -2.45978042e-01 3.81926239e-01 6.38110876e-01
1.11018693e+00 -6.69165134e-01 -5.49590230e-01 -5.21165729e-02
-7.57495835e-02 -1.62392521e+00 -1.30912215e-01 4.07842249e-01
-8.26283157e-01 -9.45841551e-01 -4.58095938e-01 -7.17177272e-01
4.68688667e-01 4.18498665e-01 9.10521865e-01 -5.03663123e-01
-8.56227800e-02 4.54194188e-01 -3.69117439e-01 -5.49645126e-01
-2.02707976e-01 2.32783973e-01 1.78604826e-01 -1.75595433e-01
6.70066476e-01 -6.84168220e-01 -4.15857702e-01 5.50785422e-01
-8.99006903e-01 -9.51032341e-02 5.13844728e-01 6.66633248e-01
6.64263189e-01 2.92909145e-02 6.23545945e-01 -7.02995062e-01
1.46612599e-01 -7.08894908e-01 -5.58725119e-01 -2.68968910e-01
-5.46231985e-01 1.44146502e-01 4.63657379e-01 -3.81046772e-01
-1.00798488e+00 3.02768201e-01 -2.00720593e-01 -5.57040155e-01
-7.04298019e-01 2.26331607e-01 -5.35924554e-01 1.01934947e-01
7.82653093e-01 9.73428115e-02 6.16212226e-02 -5.61146677e-01
3.21095437e-01 6.29528165e-01 6.06289625e-01 -8.59756112e-01
1.05479193e+00 8.26626718e-01 2.97139995e-02 -9.08524930e-01
-9.62142050e-01 -6.04013860e-01 -8.44671726e-01 8.65888447e-02
7.76753604e-01 -1.19100726e+00 9.17667076e-02 4.79723871e-01
-9.14526165e-01 -4.97765899e-01 -6.14641786e-01 5.06253004e-01
-4.58338439e-01 4.40634102e-01 1.47646591e-01 -4.88389254e-01
3.02396536e-01 -1.11231863e+00 1.40326953e+00 2.17978939e-01
-1.11839034e-01 -8.45032692e-01 -9.07830428e-03 2.05587745e-01
2.36219481e-01 3.70929360e-01 7.59566605e-01 -8.61501634e-01
-4.81888801e-01 1.47097126e-01 -3.65719348e-01 7.72512913e-01
2.00800747e-01 -5.36041200e-01 -1.14923263e+00 -3.53303850e-01
9.27473232e-02 -3.82331967e-01 1.12717283e+00 3.28528047e-01
1.07402325e+00 3.83700341e-01 -4.86312717e-01 8.70966494e-01
1.18647122e+00 3.05118978e-01 4.33639228e-01 3.95550549e-01
5.74920356e-01 6.68886662e-01 8.69335651e-01 2.41599292e-01
4.74717259e-01 5.64085901e-01 6.77477896e-01 -5.13305664e-02
-3.10896993e-01 -4.87974405e-01 3.70093495e-01 2.22161666e-01
5.03992140e-01 -1.53585210e-01 -9.90578055e-01 8.23621094e-01
-1.76229024e+00 -5.56320965e-01 4.94123623e-02 2.17639375e+00
5.23754776e-01 3.50511909e-01 -5.13920886e-03 -2.24270970e-02
6.94862902e-01 1.20887719e-01 -1.12136626e+00 -2.01984107e-01
-2.40254134e-01 4.90270078e-01 6.39140844e-01 4.57628787e-01
-1.04159164e+00 1.19084120e+00 5.06580353e+00 6.65577412e-01
-1.26276064e+00 1.42341405e-01 6.41353056e-02 -1.42742256e-02
-3.23661983e-01 -4.38603088e-02 -1.14995420e+00 4.77239877e-01
6.20939612e-01 -5.08807413e-02 1.21793844e-01 1.11191034e+00
1.01253903e-03 -1.76816449e-01 -1.21393943e+00 9.65321779e-01
-1.25136182e-01 -9.54382300e-01 4.31201681e-02 6.39992133e-02
7.60898530e-01 3.58489275e-01 1.84244290e-01 4.63370591e-01
2.92119086e-01 -8.89272690e-01 6.76349938e-01 2.24634558e-02
9.99885082e-01 -4.26943570e-01 6.54627442e-01 4.71150398e-01
-1.18065679e+00 -9.78214294e-02 -5.92606306e-01 5.93945384e-02
2.41944734e-02 6.77357733e-01 -1.00295818e+00 6.64810896e-01
8.26298177e-01 1.08865201e+00 -5.42663097e-01 8.20807517e-01
-3.06926191e-01 5.69933534e-01 -5.25042295e-01 5.78111053e-01
2.13320255e-01 -2.43384928e-01 6.80237830e-01 1.00067961e+00
5.03246069e-01 -1.45478204e-01 3.18768471e-01 1.02852654e+00
7.43211508e-02 -3.52006018e-01 -8.34800482e-01 2.53161818e-01
7.86251903e-01 9.30053532e-01 -3.83080125e-01 -1.96518824e-01
-2.96746671e-01 8.45029950e-01 6.07837215e-02 5.39292872e-01
-9.16513443e-01 -2.20344186e-01 1.02147961e+00 3.09723526e-01
5.53942442e-01 -4.28130060e-01 -5.37314117e-01 -1.13751054e+00
1.77102476e-01 -5.31001389e-01 3.06269616e-01 -6.76842630e-01
-1.43525457e+00 4.39059347e-01 4.93830383e-01 -1.30290020e+00
-1.14230938e-01 -6.66358948e-01 -3.22926342e-01 8.72136414e-01
-1.92977333e+00 -1.10827327e+00 -7.03609169e-01 8.05046678e-01
5.80009043e-01 -1.62046075e-01 4.78914946e-01 2.79644012e-01
-2.64941245e-01 4.67872083e-01 4.12808247e-02 -1.11628182e-01
9.45083439e-01 -1.01943910e+00 4.93538022e-01 7.35323966e-01
-1.58699006e-01 4.28516805e-01 6.23300970e-01 -7.25398660e-01
-8.49198878e-01 -1.53574109e+00 3.97320151e-01 -3.25963438e-01
3.72882247e-01 -6.45988107e-01 -1.11498630e+00 6.99739575e-01
-2.18602344e-01 2.03941509e-01 3.19879562e-01 -1.03125341e-01
-4.67685312e-01 -3.54211777e-01 -1.37936759e+00 1.83488488e-01
1.43350863e+00 -3.82308364e-01 -7.04228818e-01 1.99649751e-01
7.05081463e-01 -6.11860693e-01 -5.47895312e-01 7.77052939e-01
2.26130441e-01 -9.77653086e-01 1.11018038e+00 -3.68171692e-01
3.17092448e-01 -5.40887654e-01 -5.97220719e-01 -1.35545242e+00
-2.44450584e-01 1.05586015e-01 1.32654294e-01 1.09213221e+00
1.98819697e-01 -9.12014842e-01 9.56320941e-01 3.00944895e-01
-6.02344632e-01 -4.07156110e-01 -1.23998046e+00 -1.27341831e+00
4.12387252e-01 -5.53960562e-01 7.39500225e-01 7.69435763e-01
-7.00407982e-01 3.78963351e-01 2.56273299e-02 6.66322768e-01
7.31070280e-01 2.70034701e-01 9.11384583e-01 -1.54508209e+00
-6.18417822e-02 -7.38241002e-02 -5.75974107e-01 -1.32518756e+00
5.57322979e-01 -1.03774691e+00 1.03383631e-01 -1.55729043e+00
-3.77472192e-02 -6.51430488e-01 -5.45537286e-02 4.55160379e-01
1.03627317e-01 -1.49993720e-02 8.85395259e-02 1.79968163e-01
-3.75103772e-01 7.26153433e-01 1.18506134e+00 -1.60655051e-01
-1.91425607e-01 2.05613114e-03 -6.70856118e-01 6.30170166e-01
1.08819377e+00 -4.17833030e-01 -7.93787420e-01 -5.99891901e-01
-1.97753489e-01 -6.05762005e-01 6.19778574e-01 -1.35541284e+00
1.02721654e-01 -1.72888771e-01 2.20941752e-01 -8.07598770e-01
5.69435358e-01 -9.10708249e-01 1.76298898e-02 2.72316098e-01
1.21656042e-02 -4.95747715e-01 5.23138046e-01 7.03185380e-01
-2.95336336e-01 -1.30165666e-01 1.18222845e+00 7.96904713e-02
-1.26887727e+00 2.44711563e-01 2.80081276e-02 4.05904621e-01
1.09184301e+00 -6.35692716e-01 -2.71559060e-01 -2.49559611e-01
-7.43053555e-01 4.43472862e-01 7.41182566e-01 5.83479404e-01
6.26522362e-01 -1.21375680e+00 -5.17138481e-01 5.86590767e-01
4.02908027e-01 7.40251124e-01 2.26723447e-01 4.68501627e-01
-4.28939648e-02 3.52506995e-01 -3.06263089e-01 -1.14764440e+00
-8.01791310e-01 4.41603839e-01 3.04330349e-01 7.34023303e-02
-5.92042983e-01 9.11503494e-01 6.36123896e-01 -7.76374698e-01
2.65571684e-01 -6.67782545e-01 1.13803551e-01 -1.20967492e-01
7.63632311e-03 1.29569978e-01 1.77599937e-01 -6.55427992e-01
-5.02111018e-01 7.54403949e-01 9.84872356e-02 7.91743398e-02
1.22282076e+00 -1.76038176e-01 4.40598369e-01 3.62411708e-01
1.20708406e+00 -1.76366135e-01 -1.59379911e+00 -3.77983421e-01
1.84922721e-02 -5.30276656e-01 3.36119942e-02 -7.02667117e-01
-1.01298761e+00 1.03576469e+00 8.24468791e-01 -2.41069466e-01
1.01900184e+00 3.14150125e-01 9.48224008e-01 2.21181557e-01
5.03410518e-01 -1.05985212e+00 3.72326784e-02 7.98991621e-01
7.46137977e-01 -1.51553524e+00 -1.56893462e-01 -5.28177321e-01
-8.86754632e-01 8.30852628e-01 7.97393203e-01 -2.00533852e-01
7.59916306e-01 1.95397764e-01 -3.03089209e-02 -2.16487333e-01
-2.45073318e-01 -4.20759469e-01 1.96370691e-01 1.16382456e+00
-9.34825093e-02 -5.14467061e-02 1.90408722e-01 6.51590765e-01
-3.54247928e-01 5.68189248e-02 2.13245243e-01 8.07061255e-01
-6.36866271e-01 -1.10790706e+00 -3.65788341e-01 2.87312925e-01
3.74934554e-01 1.46615669e-01 -4.38286424e-01 1.01798522e+00
6.64874911e-01 8.87294769e-01 1.72948420e-01 -5.24541080e-01
7.21169174e-01 1.56832695e-01 3.34577978e-01 -1.00775111e+00
1.24673255e-01 -1.71591133e-01 -9.05339420e-02 -6.43425524e-01
-2.68045574e-01 -9.02500808e-01 -1.43990254e+00 5.29707409e-02
1.21071814e-02 6.54724538e-02 6.01304471e-01 7.68160760e-01
6.24470055e-01 3.18710178e-01 4.51383829e-01 -8.28093529e-01
-5.74176669e-01 -7.48103023e-01 -7.45595217e-01 4.62021500e-01
3.39974940e-01 -1.25156534e+00 -5.61292768e-01 -1.28454983e-01] | [8.201592445373535, -2.5778722763061523] |
35e3c1e2-2d2a-46a7-8b00-7b7de23ba737 | bert2code-can-pretrained-language-models-be | 2104.08017 | null | https://arxiv.org/abs/2104.08017v1 | https://arxiv.org/pdf/2104.08017v1.pdf | BERT2Code: Can Pretrained Language Models be Leveraged for Code Search? | Millions of repetitive code snippets are submitted to code repositories every day. To search from these large codebases using simple natural language queries would allow programmers to ideate, prototype, and develop easier and faster. Although the existing methods have shown good performance in searching codes when the natural language description contains keywords from the code, they are still far behind in searching codes based on the semantic meaning of the natural language query and semantic structure of the code. In recent years, both natural language and programming language research communities have created techniques to embed them in vector spaces. In this work, we leverage the efficacy of these embedding models using a simple, lightweight 2-layer neural network in the task of semantic code search. We show that our model learns the inherent relationship between the embedding spaces and further probes into the scope of improvement by empirically analyzing the embedding methods. In this analysis, we show that the quality of the code embedding model is the bottleneck for our model's performance, and discuss future directions of study in this area. | ['Rifat Shahriyar', 'Anindya Iqbal', 'Tahmid Hasan', 'Tanveer Muttaqueen', 'Kazi Sajeed Mehrab', 'Md. Mahim Anjum Haque', 'Masum Hasan', 'Abdullah Al Ishtiaq'] | 2021-04-16 | null | null | null | null | ['code-search', 'code-search'] | ['computer-code', 'computer-vision'] | [-2.21965566e-01 -5.31090572e-02 -5.75481594e-01 -2.26079315e-01
-5.26408434e-01 -7.66577840e-01 3.39494944e-01 5.42210042e-01
-2.52814829e-01 -1.14394732e-01 5.17642021e-01 -7.50734448e-01
-1.86841339e-02 -6.96806252e-01 -6.18088722e-01 6.59585893e-02
-2.16625735e-01 6.41479343e-02 2.42052913e-01 -2.65750289e-01
5.11838377e-01 -7.42842555e-02 -1.49195635e+00 1.78403467e-01
8.76655340e-01 5.97572207e-01 4.89817470e-01 6.45309448e-01
-5.91333568e-01 9.69456971e-01 -3.18533629e-01 -3.82512659e-01
1.09342299e-01 -2.18086272e-01 -9.86775696e-01 -6.01518691e-01
3.14459264e-01 -2.64685482e-01 -4.56243038e-01 1.28827500e+00
3.14895473e-02 -2.17603326e-01 4.07409370e-01 -1.18450570e+00
-1.03570652e+00 8.76821816e-01 -3.88063371e-01 2.37670258e-01
5.57075381e-01 1.38960997e-04 1.55877149e+00 -8.49485397e-01
6.51798129e-01 9.19785500e-01 8.92440081e-01 4.75483358e-01
-1.24398422e+00 -4.92296845e-01 -5.13961762e-02 9.50334445e-02
-1.22608602e+00 -3.38467062e-01 6.98180735e-01 -8.66715491e-01
1.29285097e+00 1.25361949e-01 4.91877794e-01 6.53347433e-01
1.33765593e-01 7.53384769e-01 2.65169770e-01 -5.40296257e-01
7.93816075e-02 4.42998528e-01 3.77250522e-01 1.06750047e+00
3.83043319e-01 -1.63915321e-01 -2.96401650e-01 -6.52658105e-01
4.09898490e-01 1.86016187e-01 -2.86227107e-01 -8.13698947e-01
-9.94725287e-01 1.21089244e+00 5.94953358e-01 5.89471996e-01
1.05832048e-01 6.54908836e-01 6.60871029e-01 4.14658487e-01
2.09038287e-01 9.29418147e-01 -4.23642546e-01 -4.96425927e-01
-8.35611522e-01 2.86046505e-01 1.08399093e+00 1.05839312e+00
8.17260742e-01 -4.19700116e-01 3.42036486e-01 9.24891949e-01
5.73193014e-01 9.33539271e-02 7.15943098e-01 -8.81966054e-01
5.40541410e-01 9.57967460e-01 -2.04166006e-02 -1.20831180e+00
-1.13301516e-01 -2.05046535e-01 9.14905891e-02 -1.40319973e-01
1.65636852e-01 2.10648984e-01 -2.69870967e-01 1.57165408e+00
-1.15906440e-01 -3.39559674e-01 -1.59468845e-01 6.82523429e-01
3.65369320e-01 5.05469561e-01 -1.39437601e-01 5.51887095e-01
1.36020279e+00 -1.19068444e+00 -2.83188492e-01 -5.18218279e-01
1.23175108e+00 -6.88162267e-01 1.32522774e+00 -1.28891125e-01
-8.97833765e-01 -2.27981091e-01 -1.08285236e+00 -3.55955005e-01
-3.77801687e-01 5.69542907e-02 7.41082788e-01 5.70306361e-01
-1.23685718e+00 4.49936211e-01 -9.57205176e-01 -6.10037744e-01
2.28340670e-01 -5.30187599e-03 -8.42436850e-02 -1.69420302e-01
-1.05289447e+00 7.43848681e-01 1.89224839e-01 -5.19998610e-01
-5.28334379e-01 -7.87690043e-01 -9.62007523e-01 5.18485963e-01
2.29597241e-01 -5.42669296e-01 1.30672717e+00 -8.71259153e-01
-6.62344754e-01 6.73352718e-01 -1.45363912e-01 -3.22579205e-01
2.46512387e-02 -1.52612954e-01 -1.59241319e-01 4.85457182e-02
2.80923218e-01 3.93931180e-01 3.72184455e-01 -8.44500422e-01
-4.78129983e-01 -2.16588691e-01 4.90646005e-01 -2.70449489e-01
-8.04184973e-01 2.70749301e-01 -6.02287531e-01 -6.43628776e-01
-1.44526526e-01 -9.40642118e-01 -6.02689423e-02 4.78683323e-01
6.03447594e-02 -4.69830781e-01 4.83698398e-01 -5.67067027e-01
1.62662256e+00 -2.65230107e+00 -2.09967303e-03 1.47726908e-01
5.56354105e-01 -4.23073508e-02 -2.04854935e-01 8.49135160e-01
-2.54356582e-03 5.22695124e-01 -1.41583622e-01 1.24940500e-01
3.27102721e-01 -4.30660695e-02 -5.59403598e-01 3.07669729e-01
-1.14427004e-02 1.05514455e+00 -1.18502986e+00 -4.46515799e-01
-3.62018019e-01 3.21378738e-01 -1.04116666e+00 1.07529841e-01
-3.42645615e-01 -5.80757916e-01 -7.08512485e-01 5.15941918e-01
5.19396402e-02 -6.72497571e-01 1.11213520e-01 -3.11570968e-02
-6.77345023e-02 7.05777943e-01 -6.35367811e-01 2.00745153e+00
-8.34724784e-01 1.06902325e+00 3.35498457e-03 -8.91391337e-01
5.41084230e-01 1.58976436e-01 4.23530787e-01 -6.16642833e-01
-3.76308858e-01 3.43175650e-01 -1.08359970e-01 -7.60779321e-01
5.62481463e-01 2.36829296e-01 -2.36423641e-01 1.00749707e+00
-3.17351848e-01 4.50291187e-02 1.13898270e-01 4.57562417e-01
1.54158616e+00 -8.02729353e-02 1.29248142e-01 -3.45522106e-01
2.06147358e-01 3.21735352e-01 9.07924026e-02 8.73720944e-01
-1.19399704e-01 1.53261080e-01 6.88688278e-01 -6.70757771e-01
-1.19645429e+00 -7.76084363e-01 -1.13448672e-01 1.23564470e+00
1.60429507e-01 -9.03529763e-01 -6.10594690e-01 -7.11936414e-01
3.38897198e-01 5.83157539e-01 -5.13416708e-01 -3.49284172e-01
-4.64561433e-01 -1.67856857e-01 7.70583212e-01 7.84542620e-01
5.56685813e-02 -6.01540506e-01 -8.55103135e-01 1.40320718e-01
-7.01470971e-02 -8.37126374e-01 -8.09622884e-01 8.47226307e-02
-7.46821284e-01 -1.14795768e+00 -2.97540963e-01 -8.61864567e-01
6.86941862e-01 4.06973273e-01 1.35361218e+00 7.16725349e-01
-5.10150254e-01 5.92405319e-01 -5.14910638e-01 -3.01570054e-02
-6.05709374e-01 3.36346447e-01 -3.41755688e-01 -7.54734099e-01
6.61938965e-01 -4.37420785e-01 -5.72726071e-01 1.19964853e-01
-1.16610169e+00 -2.47553848e-02 4.45583165e-01 8.77060771e-01
-2.29735479e-01 -5.95464520e-02 5.06203398e-02 -7.80076981e-01
1.01585031e+00 -8.33292663e-01 -8.16213191e-01 3.51006925e-01
-1.00260425e+00 6.97420239e-01 5.18010914e-01 -2.04990253e-01
-4.26926523e-01 -1.49574816e-01 -1.64205015e-01 -2.14699760e-01
3.08564186e-01 9.68283772e-01 5.13002813e-01 -3.33548725e-01
8.77060533e-01 2.21186519e-01 -2.61360239e-02 -5.95064461e-01
5.48094153e-01 6.55717134e-01 2.41756439e-01 -8.63987327e-01
8.68923128e-01 3.22678983e-01 -6.95844233e-01 -3.71647447e-01
-4.09294128e-01 -7.00688124e-01 -1.33531198e-01 3.05967689e-01
7.56142318e-01 -8.93208027e-01 -4.42891091e-01 -3.42561513e-01
-1.22536731e+00 -8.04130435e-02 -9.93190482e-02 3.11760902e-01
-3.20160121e-01 5.31497121e-01 -7.06703126e-01 -4.11957562e-01
-1.28492892e-01 -1.45937264e+00 9.98242676e-01 -1.15640990e-01
-5.21425724e-01 -1.12146783e+00 4.66010600e-01 1.18006751e-01
7.81524956e-01 -3.83123040e-01 1.60490406e+00 -6.79000854e-01
-8.12902033e-01 -3.57879668e-01 -4.69744116e-01 3.15782689e-02
4.99440059e-02 -6.49445206e-02 -6.08975768e-01 -2.94951260e-01
-1.64283067e-01 -2.82106042e-01 7.88593352e-01 -1.92446068e-01
1.19710445e+00 -5.19014955e-01 -5.43570757e-01 7.06674159e-01
1.79628778e+00 4.57824692e-02 2.73170739e-01 5.77009857e-01
5.78790486e-01 3.79364312e-01 -3.02651878e-02 3.44329685e-01
5.80904961e-01 6.48012042e-01 2.53371179e-01 2.27709562e-01
1.29475176e-01 -4.95755047e-01 2.82931089e-01 1.07972229e+00
5.24371088e-01 1.89652756e-01 -1.45420671e+00 8.74748766e-01
-1.63672042e+00 -7.82146990e-01 4.52925533e-01 1.94436347e+00
9.65527356e-01 7.88477585e-02 -4.07590084e-02 -3.29680800e-01
2.70019263e-01 1.28056273e-01 -5.69256365e-01 -5.53586662e-01
5.33505142e-01 -7.48411193e-02 6.09578907e-01 4.34116721e-01
-6.98879480e-01 6.42034352e-01 6.61123037e+00 4.46918607e-01
-9.95215774e-01 7.08278269e-02 5.06285243e-02 1.90622240e-01
-8.72413695e-01 3.85642827e-01 -4.76882786e-01 5.57010531e-01
9.04469490e-01 -6.24205530e-01 9.05220330e-01 1.40874958e+00
-2.69763559e-01 1.29760444e-01 -1.58236754e+00 9.18799818e-01
1.48249820e-01 -1.52268517e+00 -1.28833175e-01 6.17052838e-02
5.29942751e-01 4.20571625e-01 -1.18234595e-02 4.92579132e-01
5.19447803e-01 -9.80886221e-01 6.71128213e-01 1.28328919e-01
5.54134607e-01 -2.89108574e-01 4.17730004e-01 3.14078003e-01
-1.26100111e+00 -5.72631478e-01 -3.35984707e-01 -1.18934020e-01
-4.40897077e-01 3.32794487e-01 -7.45333910e-01 -4.68801744e-02
5.94783902e-01 7.07572877e-01 -8.84008765e-01 9.46892798e-01
4.37010005e-02 4.99906212e-01 -1.23191960e-01 -3.51094574e-01
2.79648095e-01 1.85807705e-01 2.93481857e-01 1.38594663e+00
3.99539232e-01 -3.19858730e-01 1.43196970e-01 1.53359604e+00
-3.05963874e-01 5.04560098e-02 -8.06672931e-01 -9.20789242e-01
5.54952502e-01 8.58738065e-01 -6.16240382e-01 -1.33803084e-01
-9.85121191e-01 6.63912117e-01 5.13238311e-01 3.92811924e-01
-7.00394690e-01 -8.46546590e-01 9.44856167e-01 1.12288699e-01
3.32367152e-01 -4.38626558e-01 -1.05391122e-01 -1.40312541e+00
4.53151643e-01 -9.48643088e-01 2.35551402e-01 -8.27456892e-01
-1.09430540e+00 2.98739165e-01 -1.54226292e-02 -8.76263797e-01
-3.94001067e-01 -6.06550038e-01 -4.73714292e-01 7.74558723e-01
-1.37803340e+00 -5.21887243e-01 -1.70498770e-02 1.17409760e-02
5.60885251e-01 -1.36859909e-01 8.17194402e-01 4.30735797e-01
-1.43773243e-01 6.83260560e-01 4.03323412e-01 6.20643556e-01
4.41590250e-01 -1.03055310e+00 8.48316550e-01 6.74508631e-01
4.72377658e-01 1.33451426e+00 6.59950316e-01 -4.01038378e-01
-1.98193789e+00 -5.94753683e-01 1.04636741e+00 -6.78816855e-01
1.25528133e+00 -5.92436731e-01 -9.17197287e-01 6.31679714e-01
1.20158933e-01 -1.25508774e-02 7.68930852e-01 3.44191760e-01
-1.01243782e+00 1.09071612e-01 -6.22437119e-01 4.85916734e-01
8.87095928e-01 -1.28830433e+00 -6.62837744e-01 2.45749757e-01
1.10006225e+00 -5.08456863e-02 -8.45184922e-01 -1.55150771e-01
7.27673233e-01 -7.14216828e-01 9.21381652e-01 -6.05938017e-01
9.06073987e-01 -1.23825908e-01 -2.16243073e-01 -1.04972959e+00
-1.99679464e-01 -2.67842829e-01 -3.23971175e-02 7.57533073e-01
5.63706100e-01 -4.56740975e-01 7.35685945e-01 9.40795183e-01
1.46363437e-01 -8.05172682e-01 -5.49686611e-01 -7.28184819e-01
2.18964532e-01 -4.48369801e-01 7.14022577e-01 1.01160467e+00
6.47009969e-01 1.40148476e-01 1.98428646e-01 -1.69938356e-02
2.89577454e-01 4.51132357e-01 5.59868574e-01 -1.23075914e+00
-5.93824923e-01 -7.96707153e-01 -5.94189525e-01 -1.24947286e+00
3.23867947e-01 -1.17642570e+00 -3.57558690e-02 -1.45324016e+00
6.17538393e-01 -5.40847063e-01 -2.68659979e-01 4.93075669e-01
-1.04283229e-01 -3.19584459e-01 8.40954259e-02 5.20350277e-01
-6.27019107e-01 1.42751664e-01 4.24448639e-01 -4.54488367e-01
2.07620338e-01 -4.58710819e-01 -9.63544071e-01 4.16159630e-01
4.69965845e-01 -7.04734385e-01 -6.39761388e-01 -1.04212320e+00
1.13639092e+00 6.83061592e-03 2.84624815e-01 -7.55948782e-01
5.20352721e-01 5.03566600e-02 -3.00987393e-01 3.62215042e-02
-1.68311745e-01 -1.03168595e+00 -1.97100848e-01 5.38624823e-01
-8.58532965e-01 4.73783970e-01 1.62281141e-01 6.13964915e-01
-2.33884051e-01 -6.05136335e-01 3.83693278e-01 -2.18073353e-01
-7.14601398e-01 2.60403045e-02 -3.56534421e-01 4.98418272e-01
6.09624267e-01 -1.17556691e-01 -3.92264813e-01 -1.94843397e-01
-2.10572127e-02 1.93061754e-01 1.08057606e+00 7.17146635e-01
5.17638803e-01 -1.22559893e+00 -2.15994701e-01 1.57947928e-01
6.17497623e-01 -5.48343539e-01 -3.81279469e-01 5.03175378e-01
-7.06146240e-01 6.38515234e-01 9.06784385e-02 -3.14095467e-01
-9.08651412e-01 8.28696370e-01 1.90203071e-01 -4.81595434e-02
-5.66539824e-01 7.26243973e-01 1.71047881e-01 -4.44809109e-01
3.24335486e-01 -5.77731669e-01 9.03804898e-02 -3.76710773e-01
4.47306812e-01 3.74334082e-02 -1.42115086e-01 -2.24896297e-01
-3.88360858e-01 3.95092428e-01 -1.56410992e-01 8.67816210e-02
1.36416125e+00 -2.74480376e-02 -4.56703216e-01 4.16909903e-01
1.69486189e+00 2.30236724e-01 -7.14747906e-01 -3.21318328e-01
5.69837391e-01 -6.97785139e-01 6.75015524e-03 -4.09043282e-01
-8.36471677e-01 8.52767825e-01 4.11522269e-01 3.87456477e-01
7.21662998e-01 4.25833404e-01 8.42931092e-01 7.85643399e-01
4.56058085e-01 -9.59863245e-01 1.18743584e-01 5.11951387e-01
3.57674152e-01 -1.08189535e+00 -6.34960607e-02 -7.97605608e-03
-7.84819126e-02 1.26394570e+00 3.03149372e-01 5.54180965e-02
6.30669892e-01 4.39852029e-01 -9.58097875e-02 -6.27766371e-01
-8.57419372e-01 2.67818093e-01 9.34482142e-02 2.05775693e-01
7.34423101e-01 -2.30546370e-01 -1.40054852e-01 3.19325924e-01
-7.77190849e-02 -1.36127487e-01 4.45664942e-01 1.14841783e+00
-5.60207009e-01 -1.18346083e+00 -5.89128351e-03 5.11956573e-01
-3.26782167e-01 -5.22423565e-01 -3.41768891e-01 4.95937526e-01
-2.76001155e-01 6.25397086e-01 -5.07388040e-02 -3.80979955e-01
6.16152398e-02 2.36902252e-01 1.53980762e-01 -7.41409302e-01
-5.34712791e-01 -6.29945457e-01 -1.60217211e-01 -8.13568354e-01
8.43851492e-02 -4.52289551e-01 -1.19102013e+00 -2.09032491e-01
-3.78366828e-01 5.51432312e-01 1.06658924e+00 7.00964153e-01
6.39626563e-01 5.87620884e-02 4.65318501e-01 -1.87752888e-01
-8.38490665e-01 -4.14668560e-01 -1.94413856e-01 4.20260102e-01
4.16537821e-01 -5.03718674e-01 -4.61503416e-01 2.22996131e-01] | [7.55244779586792, 8.068450927734375] |
ac5936c7-54c6-4834-bfac-12c784eac5b4 | graph-enhanced-dual-attention-network-for | null | null | https://aclanthology.org/2020.coling-main.136 | https://aclanthology.org/2020.coling-main.136.pdf | Graph Enhanced Dual Attention Network for Document-Level Relation Extraction | Document-level relation extraction requires inter-sentence reasoning capabilities to capture local and global contextual information for multiple relational facts. To improve inter-sentence reasoning, we propose to characterize the complex interaction between sentences and potential relation instances via a Graph Enhanced Dual Attention network (GEDA). In GEDA, sentence representation generated by the sentence-to-relation (S2R) attention is refined and synthesized by a Heterogeneous Graph Convolutional Network before being fed into the relation-to-sentence (R2S) attention . We further design a simple yet effective regularizer based on the natural duality of the S2R and R2S attention, whose weights are also supervised by the supporting evidence of relation instances during training. An extensive set of experiments on an existing large-scale dataset show that our model achieve competitive performance, especially for the inter-sentence relation extraction, while the neural predictions can also be interpretable and easily observed. | ['Shikun Zhang', 'Xiangyu Xi', 'Rui Xie', 'Zhonghao Sheng', 'Wei Ye', 'Bo Li'] | 2020-12-01 | null | null | null | coling-2020-8 | ['document-level-relation-extraction'] | ['natural-language-processing'] | [ 0.34779376 1.0581323 -0.20508957 -0.55965227 -0.76069313 -0.48929685
0.6950808 0.5085981 0.06197954 0.73289686 0.67418957 -0.6260191
-0.23261647 -1.2833298 -0.87063223 -0.0651768 -0.03494391 0.47755867
0.13653105 -0.6422836 -0.22699055 0.22620451 -0.88766325 0.6061486
0.8993934 0.8968155 0.00796616 0.8920471 -0.32099643 1.4527168
-0.61416435 -0.99736476 0.02129027 -0.4692323 -1.4099412 0.06783313
-0.02872911 -0.12205819 -0.6241222 1.1196502 0.09595835 0.11179262
0.26066807 -1.0108355 -1.2671082 1.3114891 -0.5169346 0.42620918
0.84596515 -0.03680716 1.7988751 -0.81151235 0.7233165 1.4707078
0.4940251 0.16287668 -0.94501424 -0.31780395 0.67356724 0.334162
-1.2303274 -0.2763471 0.9449773 -0.01373862 1.7971992 0.45394552
0.6236103 1.0989647 0.2654683 0.87681973 0.471816 -0.26119927
-0.35685566 -0.10296585 0.6045126 0.9225726 0.4916236 -0.46720746
-0.36983773 0.07059917 0.47932845 -0.29160777 -0.35045016 0.54616934
-0.99507076 0.7432342 0.70231974 0.3538961 -0.45388487 -0.03827733
0.4034615 0.25114614 0.6359139 0.67502296 -0.6872305 0.34244102
-0.2732311 0.05611047 0.6963704 1.1944413 0.58546495 -0.30100924
-0.6749304 0.5237825 0.5265388 0.22946176 0.22349194 -0.5135115
1.1702845 1.2314335 -0.45815408 -1.3740314 -0.49925426 -0.70993024
-1.2316134 -0.50698763 -0.17147237 -0.21941783 -0.5634923 1.6158575
0.3215888 0.0727468 0.6226485 0.74340034 1.6436572 0.6218889
0.1147141 -0.25142574 1.6629703 -1.1720659 -1.0889133 -0.67741376
0.8056857 -0.08811776 0.7725126 -0.19683984 -1.1687937 -0.57734394
-1.1714463 -0.8035252 -0.39149123 0.05737108 0.8133384 -0.28253022
-0.813872 0.44696635 -0.4358359 -0.11479659 0.6593599 0.43565643
-0.60528964 0.09074285 -1.9919485 1.1314647 0.68979794 0.59548396
-0.38781583 -0.40959027 -1.4244988 0.50682694 0.83340174 -1.1057565
0.862697 -0.4545686 -1.3726537 0.9948161 -0.31821862 -0.5573994
0.02328216 -0.19982336 -0.5616263 0.23250663 0.3213977 0.15309313
0.21716566 -1.1564684 -0.09519228 -0.34952438 0.7114035 0.39426738
0.023749 0.06880138 -0.39057934 -0.52281046 0.12446465 -0.29294428
-0.24547507 -0.6833677 -1.1804162 -0.6515714 0.47859663 -0.8491623
1.4578588 -1.644183 0.41327295 -0.04029907 0.69416726 0.30655232
-0.33320644 0.55498105 -0.5482472 0.45176747 -0.24875386 -0.40380192
-0.23398024 0.33048728 -0.4516469 -0.19992098 1.0041281 1.5388886
-1.2428353 -0.6893793 -0.22628747 0.1413656 -0.17960367 0.51092196
-0.3146218 0.17112458 -0.7158194 0.33092502 0.41897672 -0.92112184
0.54281026 -0.32956666 0.6265584 0.86611444 -0.7433236 1.4467714
-0.52855265 0.41006145 -0.12102065 -1.0368086 1.1237742 0.32756633
0.03730386 -0.5895635 0.11489863 -0.34075162 0.17229262 -0.82933474
0.59684885 -0.03138856 -0.10454737 0.31988373 0.32221055 -0.06087712
0.30588698 0.91935754 1.252164 0.01641644 0.7584924 -0.06862082
0.94843745 -0.28098089 0.6138064 0.57561386 0.20836133 0.36689004
1.2113104 -0.44215253 -0.56167626 -0.75374955 0.2553836 0.72944313
0.21310857 -0.7927097 -0.37631124 -1.2956753 -0.19340008 0.830402
-0.84345984 -0.3948678 -0.58139145 -0.6210086 0.3790798 0.876353
0.53632563 -1.2383175 0.25936037 0.32103613 -0.43595797 -1.6371529
-0.3037376 0.25840312 -0.4404321 -1.2581499 0.18837203 -0.6658377
0.7207228 0.08032929 1.5052792 0.42853972 0.26327753 -0.05510183
-0.51244086 -0.20731437 -0.36547238 0.17625485 -0.38656148 0.094297
0.3897641 -0.6397047 -0.1123801 -0.35316885 -0.73345536 0.4698724
0.56171685 0.88827074 0.45008183 0.02107225 0.7056179 -1.4594553
1.0284964 -0.66764563 -0.143591 0.68919414 -0.52669203 0.26815465
0.7871022 0.04357828 -1.1322818 -0.26975024 -0.2555832 -0.02405461
0.01423822 1.0725273 -0.63874817 0.52779037 0.3404236 -0.02051537
-0.39292017 0.11803872 0.56899744 0.43979502 0.48746696 -0.5426943
0.75822216 0.02404815 0.2629733 -0.38602874 -1.7126272 -0.01824908
-0.6609929 0.1303787 1.0704979 -0.8794935 -0.8962333 -0.10639754
-1.8253112 -0.2320931 -0.2975439 0.07632241 -0.15245971 0.39648438
-0.8616969 -0.8026799 -0.743726 -1.008621 1.253976 0.15704036
-0.27279866 -1.2217841 -0.1704814 0.6507072 -0.11132941 0.31413412
1.1200342 -1.0464076 -0.6039745 -0.22307585 -0.7337574 0.06727776
0.20686224 0.04985345 -0.7515195 0.21445622 -0.19053115 -0.3999483
0.918575 -0.02019454 1.1922805 -0.58018017 -0.22416784 0.30626294
1.101665 -0.19819576 0.7149467 -0.04621237 1.1677858 0.46762016
0.35468367 -0.00681163 1.0869502 0.26446337 0.38567343 -0.2239455
-0.34166202 -0.50949764 0.17180361 1.1057949 -0.17656712 -0.55570644
-0.81705385 0.38094 -2.0153875 -1.0260816 -0.5847201 1.3226625
1.009372 0.50849414 -0.38908055 0.12306794 0.64687383 0.41936156
-0.24183597 -0.47684115 -0.5066486 0.16742092 0.01341913 1.043354
-0.9258408 1.1398176 5.8202024 0.6547894 -0.505609 -0.05788687
0.828242 0.35257953 -0.8628234 0.16005543 -0.77080333 -0.06099544
0.7639469 -0.2317253 0.21626359 0.49872822 -0.18012309 0.14765395
-1.1802076 0.47503892 0.1445757 -1.6615739 0.30505425 -0.25509363
0.49335074 -0.33838 -0.37520155 0.62520564 0.5150414 -1.0707207
0.32088688 0.5450747 0.69288486 -0.42185694 1.032364 0.17661154
-1.5607425 0.04839024 -0.12560353 -0.34850702 0.14782263 0.6693486
-0.7953163 1.4141507 0.23184805 1.1340224 -0.80640894 -0.06852761
-0.9955182 0.38711038 0.0686143 -0.22560088 0.1169142 -0.23271812
0.47691724 1.2580392 -0.2810863 0.5566461 -0.04447556 1.0787102
-0.5171604 0.00799347 -0.77288115 -0.25087902 0.21229902 1.5568186
-0.5985051 -0.68182105 -0.5681536 0.77533555 1.1561062 0.52044815
-0.77473295 -0.47543544 0.28196973 -0.29587913 0.23362875 -0.02658975
-0.60069513 -1.3954238 0.3165864 -0.57766676 0.7204941 -1.059597
-1.5899055 1.1608776 -0.09566975 -0.7020884 -0.10552885 -0.53345096
-0.8526749 1.0732094 -1.4834572 -1.4404156 -0.12425368 0.46441594
0.28343913 0.0697595 0.96349835 -0.04515231 -0.8617663 0.5865564
-0.9641527 0.72139585 -0.0182346 -1.4368854 0.83384514 0.96437186
0.34191978 0.8243093 0.38658753 -0.9355316 -1.3309078 -1.2783598
1.389395 -0.54316574 1.0603185 -0.45479202 -1.1278524 1.2223636
0.5076699 0.18993427 0.6669195 0.6090216 -0.53963494 -0.08595946
-0.76181114 0.5797908 1.3029529 -0.9454121 -1.0980805 0.6507274
1.497872 -0.6199535 -1.0618477 0.6660111 -0.20091785 -0.48601997
0.81109554 -1.2292519 1.1774379 -0.11771943 0.07811839 -1.0181545
-0.6787495 -0.5993002 -0.7750177 1.4305931 1.0589762 -0.47260273
0.44875404 0.7053014 -0.35528335 -1.3068057 -0.7094306 -0.30923963
-0.21137173 -0.38143837 0.76507586 1.0556548 0.4477549 1.3565414
-0.1789324 0.49025497 0.19208038 0.5723266 0.3681593 -1.0825654
-0.6002727 -0.42074725 -0.21097963 -1.0463904 0.6843857 -1.3008692
-0.1508792 -2.1699524 0.3615656 0.11618938 -0.17734052 0.567211
-1.0039084 -0.28636596 -0.03087142 -0.3602701 -0.98917085 0.73644215
1.8794914 -0.32862744 0.07014149 -0.19894697 -1.1079012 0.52662665
0.44296682 -0.18594576 -0.5056628 -0.52184665 0.7102351 0.49382642
0.17792565 -0.27252725 0.22284742 -0.215898 0.14173242 -0.5061591
0.25707033 -0.49948946 -0.2875595 0.03958128 -0.75605136 -0.06893181
0.01403678 0.7102886 -0.32206196 0.07083861 0.0675291 -0.12874697
-0.15015966 0.49168715 0.05013098 0.3072028 0.8361167 0.28802323
-0.8015013 -0.5364712 -0.8523974 0.60218334 -0.42323753 0.45622024
0.8740799 -1.4342347 -0.81243634 0.07002932 0.16252095 0.63245916
0.22256158 0.6826981 -0.17475386 0.33602303 0.32560956 -0.03139839
-1.0942373 1.0027844 0.28335887 -0.98231095 -0.72484857 1.1175891
0.02757257 -0.30250672 -0.20703644 -0.65069324 -0.5580455 -0.19300717
0.4153702 -0.3033476 -0.01110005 -0.64130825 -0.5675261 0.28556517
-0.12220317 0.3197888 1.4169995 -0.07511855 -0.64870656 0.30319256
1.0940866 0.03490089 -0.7063561 -0.4196375 0.0285594 0.03563729
-0.11862478 -0.5548032 -0.9794323 0.6922993 -0.7862038 0.8215667
1.0053799 0.39777222 0.86089987 0.4834036 -0.06626645 -0.5695014
0.03900154 0.8059407 1.4343581 -1.4011587 0.20758663 -1.0673888
-0.9663177 1.2934296 0.93313396 -0.24233714 0.5567939 0.43640825
-0.24719222 -0.59526825 -1.0212314 -0.3684483 0.61831117 0.3486773
0.59367114 0.10528272 -0.18479517 1.2347041 -0.17494975 -0.52444214
0.43280292 0.38960287 0.07546225 -1.0472642 0.28624004 0.45602545
-0.254749 -0.42229202 -0.73567903 0.4923969 -0.11932234 1.2317284
-0.07861766 -0.6301618 0.3221631 -0.11233005 0.38041118 -0.9102693
-0.819407 -0.43412697 0.85274565 -0.54651225 -0.36860722 -0.1563471
-1.625641 -0.07246952 -0.3642382 0.13489138 -0.08708227 1.4101845
0.57532346 1.3293284 0.539656 -0.03079312 -0.12310796 -1.1203663
-0.32894164 0.70532125 0.31759688 -0.3724065 0.02055954 -0.26913127] | [9.269532203674316, 8.582245826721191] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.