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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
98204de0-25d6-48df-bcb2-440aa9c3cf17 | code-attention-translating-code-to-comments | 1709.07642 | null | http://arxiv.org/abs/1709.07642v2 | http://arxiv.org/pdf/1709.07642v2.pdf | Code Attention: Translating Code to Comments by Exploiting Domain Features | Appropriate comments of code snippets provide insight for code functionality,
which are helpful for program comprehension. However, due to the great cost of
authoring with the comments, many code projects do not contain adequate
comments. Automatic comment generation techniques have been proposed to
generate comments from pieces of code in order to alleviate the human efforts
in annotating the code. Most existing approaches attempt to exploit certain
correlations (usually manually given) between code and generated comments,
which could be easily violated if the coding patterns change and hence the
performance of comment generation declines. In this paper, we first build
C2CGit, a large dataset from open projects in GitHub, which is more than
20$\times$ larger than existing datasets. Then we propose a new attention
module called Code Attention to translate code to comments, which is able to
utilize the domain features of code snippets, such as symbols and identifiers.
We make ablation studies to determine effects of different parts in Code
Attention. Experimental results demonstrate that the proposed module has better
performance over existing approaches in both BLEU and METEOR. | ['Hong-Yu Zhou', 'Wenhao Zheng', 'Ming Li', 'Jianxin Wu'] | 2017-09-22 | null | null | null | null | ['comment-generation'] | ['natural-language-processing'] | [-8.79984349e-02 1.20053999e-01 -8.01637024e-02 -4.09040660e-01
-5.42389691e-01 -6.66713119e-01 1.08394802e-01 3.36957455e-01
-5.45737259e-02 4.44558561e-01 2.93403059e-01 -4.42603827e-01
4.51688468e-01 -5.34265399e-01 -4.95958328e-01 -5.04914224e-02
2.19142810e-01 -2.29370415e-01 1.93297818e-01 -1.72560841e-01
7.75058866e-01 -2.80870169e-01 -1.38933885e+00 5.58542311e-01
1.45192027e+00 2.75321394e-01 5.65260291e-01 5.27215958e-01
-7.96498656e-01 8.11125994e-01 -7.51270771e-01 -7.89152443e-01
6.69240728e-02 -5.88805377e-01 -7.61674881e-01 -3.01010869e-02
1.90218776e-01 -2.70209581e-01 4.80611064e-02 1.35385060e+00
3.87308359e-01 -2.59305775e-01 5.17774642e-01 -1.21827638e+00
-1.32602370e+00 1.11621344e+00 -8.29280317e-01 1.54434279e-01
5.85326314e-01 1.39274567e-01 1.17729628e+00 -1.11486983e+00
4.66805667e-01 9.19660151e-01 8.06168497e-01 6.87989593e-01
-8.05471838e-01 -6.99560344e-01 1.09429762e-01 5.78305386e-02
-1.13418603e+00 -1.88219547e-01 6.81353211e-01 -9.31812763e-01
9.41623628e-01 1.28339589e-01 3.29088122e-01 7.20245481e-01
2.58020073e-01 6.70738041e-01 3.91217768e-01 -5.76715469e-01
-8.80720472e-06 4.35012609e-01 4.12684917e-01 6.36214972e-01
2.86728919e-01 -5.95525920e-01 -7.77521580e-02 -3.35622609e-01
4.01234776e-01 3.44760090e-01 -4.19130683e-01 -9.65569392e-02
-1.18717062e+00 8.11771095e-01 4.60693866e-01 2.45259225e-01
-4.10313681e-02 2.20203832e-01 6.73898458e-01 2.85001069e-01
3.50774288e-01 5.98165512e-01 -4.38790798e-01 -5.30939400e-01
-6.66407287e-01 2.76024580e-01 6.52766049e-01 1.52911174e+00
1.00363064e+00 -2.47041881e-01 -3.74215394e-01 9.69042599e-01
4.83649820e-01 3.10235083e-01 7.94537306e-01 -5.00174344e-01
1.00600016e+00 1.27115417e+00 1.47418767e-01 -1.08350241e+00
5.41900434e-02 -2.73015082e-01 -4.96176302e-01 -1.31165713e-01
-2.34456789e-02 -4.21508938e-01 -4.94951814e-01 1.37495399e+00
-1.92195177e-01 -2.91870475e-01 -1.93770945e-01 5.24625599e-01
1.05836356e+00 5.29945731e-01 -1.72760278e-01 1.20813727e-01
1.15710866e+00 -1.25701737e+00 -6.66042566e-01 -3.54253471e-01
1.05740869e+00 -1.07709014e+00 1.56244040e+00 -1.03474610e-01
-8.11739862e-01 -6.60836339e-01 -7.78547704e-01 -1.28292888e-01
-1.46997079e-01 5.25834978e-01 5.04273236e-01 7.06595540e-01
-8.00425887e-01 3.27684283e-01 -5.45675457e-01 -3.54596734e-01
4.72692311e-01 -7.04363659e-02 -7.13762492e-02 -5.24288565e-02
-8.51250589e-01 3.93331110e-01 2.41972014e-01 -2.81868819e-02
-5.88416576e-01 -7.43744016e-01 -9.75577354e-01 3.64151061e-01
1.87466219e-01 -1.70573190e-01 1.46880257e+00 -1.26704121e+00
-8.65894496e-01 5.50144255e-01 -2.59451777e-01 -7.86623731e-03
3.26168895e-01 -2.85523951e-01 -9.49873477e-02 -3.74028504e-01
3.48631889e-01 4.81969714e-01 4.65033233e-01 -1.01216114e+00
-4.84665424e-01 2.59477552e-02 4.46501285e-01 -1.18855521e-01
-6.95213139e-01 3.44197124e-01 -5.32871723e-01 -7.77738154e-01
-2.45271429e-01 -1.02312553e+00 -2.98565060e-01 4.86861169e-02
-4.06714439e-01 -3.12079638e-01 4.72296268e-01 -8.91049027e-01
1.79556930e+00 -2.32105875e+00 -1.24727972e-01 -5.74429668e-02
2.22913116e-01 2.68381327e-01 -2.98584789e-01 6.23388052e-01
-1.96071595e-01 5.92266738e-01 -4.47292656e-01 -3.28507572e-02
-1.58044070e-01 -1.61412373e-01 -3.06892842e-01 -3.24409339e-03
1.91499650e-01 8.87157083e-01 -9.03078854e-01 -3.59889925e-01
-5.29323101e-01 -1.83742121e-02 -1.15696824e+00 3.54709119e-01
-3.68894011e-01 -5.32416813e-03 -6.74628496e-01 4.81231481e-01
7.51919806e-01 -4.07947093e-01 -1.59863979e-01 4.01687205e-01
-1.40091881e-01 4.19068426e-01 -7.71972656e-01 1.85639834e+00
-7.47556150e-01 8.04114997e-01 -3.39976460e-01 -6.57226145e-01
1.05734646e+00 3.08047682e-01 -3.42893153e-02 -2.78224289e-01
-5.67495227e-02 1.26509354e-01 2.36693457e-01 -1.02386212e+00
5.27703404e-01 3.30478787e-01 -1.59403577e-01 8.07229698e-01
-3.85158211e-01 6.03574663e-02 5.40697873e-01 5.45990050e-01
1.41039586e+00 5.98853678e-02 2.62235761e-01 -3.01473647e-01
6.94724858e-01 1.90035719e-02 5.96713960e-01 5.82559705e-01
6.29976019e-02 7.84375012e-01 9.32452202e-01 -3.12271565e-01
-9.66820717e-01 -4.14527506e-01 1.58572242e-01 1.03032517e+00
-7.83413425e-02 -8.82339835e-01 -9.17941272e-01 -1.11071289e+00
-1.17647581e-01 6.35449231e-01 -7.00444162e-01 -2.07388937e-01
-4.72868294e-01 -5.81930041e-01 4.59359229e-01 7.52235234e-01
4.30802494e-01 -1.26275563e+00 -5.61735392e-01 2.45565981e-01
-3.83392960e-01 -7.13259697e-01 -1.02261913e+00 -2.46138364e-01
-6.58973873e-01 -1.16081572e+00 -7.53324986e-01 -8.22831213e-01
1.26803219e+00 5.28702438e-01 1.10920143e+00 9.97804165e-01
-1.93643913e-01 -8.40529948e-02 -9.63117659e-01 -5.55187047e-01
-6.42504454e-01 3.39112818e-01 -6.81654632e-01 -4.02131379e-01
6.49073660e-01 -3.26071441e-01 -4.23861891e-01 1.15327545e-01
-1.03903210e+00 1.19812749e-01 7.57332146e-01 8.72544408e-01
-2.24132523e-01 -2.87932783e-01 6.59472108e-01 -1.42401719e+00
9.67175126e-01 -7.80308664e-01 -3.55897099e-01 4.00912404e-01
-5.82033932e-01 1.90695971e-01 7.83136487e-01 -2.24467576e-01
-1.20712340e+00 -1.01084888e-01 -1.31360307e-01 2.07550734e-01
7.19442070e-02 1.02835572e+00 -1.90521795e-02 1.01398520e-01
7.68054962e-01 1.27093613e-01 -2.49770358e-01 -5.54082096e-01
7.96619207e-02 1.13918531e+00 7.47240335e-02 -6.16141438e-01
8.38169932e-01 -6.79598451e-02 -8.85686636e-01 -2.50973850e-01
-5.68513870e-01 -4.89144325e-01 -6.24958813e-01 1.75718628e-02
6.07946813e-01 -8.85949612e-01 -9.90457162e-02 2.67642111e-01
-1.57641137e+00 -4.53347266e-02 1.55103922e-01 1.78636372e-01
-6.88397288e-02 6.67103410e-01 -6.94362640e-01 -5.50304055e-01
-3.51979256e-01 -1.32420790e+00 7.23956108e-01 4.17739153e-01
-3.63889992e-01 -7.86123872e-01 1.06043734e-01 2.60031819e-01
5.16187191e-01 -6.05979487e-02 1.22799349e+00 -5.49447954e-01
-5.59515655e-01 -2.58617848e-01 -4.78534341e-01 4.05184478e-01
3.45917165e-01 4.34440225e-01 -5.88503420e-01 -3.26191574e-01
-2.23248184e-01 -5.51372953e-02 6.23100042e-01 -2.58112431e-01
1.36538005e+00 -4.93340254e-01 -2.36364722e-01 3.42554212e-01
1.42400312e+00 3.32917869e-01 8.05634737e-01 2.66386837e-01
7.37279892e-01 5.26410758e-01 6.13675475e-01 6.40917778e-01
4.87707168e-01 4.40412492e-01 4.83825833e-01 7.26047903e-02
1.06304519e-01 -2.47000858e-01 6.65217757e-01 1.30460203e+00
1.71963006e-01 -1.84033260e-01 -1.19856608e+00 9.28922474e-01
-1.71828175e+00 -7.20733881e-01 -4.95085835e-01 1.90335810e+00
1.08090413e+00 1.50954917e-01 -2.29257613e-01 -1.69148937e-01
8.10514808e-01 -2.68401474e-01 -3.71722996e-01 -4.93918389e-01
5.15695512e-01 -5.33553623e-02 1.61907166e-01 2.28474792e-02
-5.49221754e-01 5.38949311e-01 5.75154734e+00 4.58289564e-01
-8.59839678e-01 1.23874731e-01 3.05589586e-01 2.44744226e-01
-7.76608825e-01 3.31299394e-01 -6.69556439e-01 9.68554020e-01
6.77249908e-01 -5.67848861e-01 2.26811722e-01 1.27916670e+00
1.90308020e-01 7.22957850e-02 -1.25689352e+00 7.50442863e-01
2.63394117e-01 -1.08311498e+00 -1.80104487e-02 -2.79737085e-01
1.06037152e+00 -1.65413707e-01 -2.59459406e-01 5.73944032e-01
4.82647389e-01 -6.87470496e-01 6.08188450e-01 2.53941715e-01
6.93391025e-01 -5.82214832e-01 1.16012514e+00 4.54884380e-01
-1.11030698e+00 -2.26287097e-01 -7.25787878e-01 -2.51109362e-01
-2.36516744e-01 7.06337690e-01 -8.48380804e-01 3.03320199e-01
6.38320088e-01 9.97298300e-01 -1.20819545e+00 1.39977694e+00
-4.39556479e-01 5.55755854e-01 3.15922290e-01 -3.84602696e-01
3.86172086e-02 1.71274021e-02 1.35211468e-01 1.46534193e+00
7.21358716e-01 -1.71983957e-01 -1.77564900e-02 1.28189862e+00
-3.28106284e-01 4.00672644e-01 -5.96624970e-01 -1.88655689e-01
5.54174900e-01 1.28413129e+00 -6.43323600e-01 -3.93565208e-01
-9.76106405e-01 7.84207463e-01 3.64402175e-01 2.79168785e-01
-9.85196888e-01 -1.02385533e+00 5.66816151e-01 1.76604539e-01
2.90685236e-01 -4.50034533e-03 -3.17898214e-01 -1.38646305e+00
5.45901954e-01 -1.00896525e+00 9.17663276e-02 -9.45722401e-01
-1.05943060e+00 5.88497281e-01 -2.75430799e-01 -1.46365643e+00
-7.09180012e-02 -1.62339881e-01 -1.11015141e+00 9.50869381e-01
-1.20362663e+00 -7.19703555e-01 -6.96036398e-01 -1.19261090e-02
9.21430469e-01 -1.26822457e-01 5.09306848e-01 6.24503732e-01
-7.15458870e-01 8.26414764e-01 1.16054185e-01 5.21864116e-01
8.66215527e-01 -1.26676023e+00 9.31580007e-01 1.28607857e+00
1.66696236e-02 1.33511794e+00 4.92376477e-01 -7.51605451e-01
-9.57241476e-01 -1.09981656e+00 9.88628209e-01 -4.45510805e-01
5.38866580e-01 -4.72997993e-01 -1.11573124e+00 7.45708346e-01
3.88719052e-01 -2.28324115e-01 9.00891542e-01 -2.21513398e-02
-3.86636794e-01 2.31677830e-01 -7.28772640e-01 5.94641626e-01
8.75598967e-01 -4.90034133e-01 -6.25368297e-01 2.26468131e-01
8.30084026e-01 -3.12649310e-01 -5.14395177e-01 -3.58360494e-03
2.77703613e-01 -9.96874571e-01 3.17340910e-01 -5.26346564e-01
1.15622938e+00 -5.25490820e-01 3.18614870e-01 -1.37339878e+00
-2.66080260e-01 -4.13011760e-01 2.76009381e-01 1.70443177e+00
7.72980034e-01 -3.31565827e-01 5.75226307e-01 8.92668486e-01
-4.16933626e-01 -5.11439085e-01 -2.43769661e-01 -4.06697989e-01
1.06637768e-01 -1.35641187e-01 8.71047974e-01 1.05835998e+00
4.78292555e-01 2.68207818e-01 -9.82693955e-02 -2.01782510e-01
-5.77708967e-02 1.70662552e-01 9.26970303e-01 -1.13984704e+00
-3.83309782e-01 -3.32154155e-01 -2.16276318e-01 -1.26094067e+00
1.12838931e-01 -1.04967487e+00 3.24117482e-01 -1.61873531e+00
6.68447554e-01 -5.61695158e-01 7.37401024e-02 6.26129031e-01
-7.19577014e-01 -1.79084718e-01 3.26205306e-02 2.96602875e-01
-5.45268297e-01 3.64624023e-01 1.05977762e+00 -2.48360574e-01
-5.48757911e-02 1.02746170e-02 -1.08451712e+00 6.59654975e-01
6.30851626e-01 -7.36686051e-01 -5.01630008e-01 -9.49358702e-01
8.03788662e-01 2.32254900e-02 -2.31027063e-02 -9.03521836e-01
1.55931920e-01 -1.42111674e-01 -1.06211051e-01 -4.56710547e-01
-7.56095052e-01 -6.80217564e-01 -7.70885721e-02 4.09974009e-01
-7.32889056e-01 5.22303283e-01 7.57895634e-02 4.79218572e-01
-3.21279675e-01 -1.11120355e+00 5.00403464e-01 -3.84975016e-01
-5.14898777e-01 8.10664147e-02 -4.15855736e-01 1.72398031e-01
9.84126329e-01 -9.20060351e-02 -5.57715833e-01 -1.63005456e-01
-8.07334259e-02 3.30056280e-01 7.98168421e-01 8.35653126e-01
4.84508812e-01 -1.31905329e+00 -7.60253489e-01 2.01652259e-01
6.47444904e-01 7.87828192e-02 1.92047998e-01 5.20888150e-01
-8.94671321e-01 2.12769300e-01 -2.13072717e-01 -2.46593490e-01
-1.30067849e+00 6.01493001e-01 -6.75429180e-02 -1.25550166e-01
-4.50930536e-01 8.86304736e-01 1.42247796e-01 -5.90587854e-01
-2.28535421e-02 -7.05317914e-01 -3.67116660e-01 -1.64071575e-01
7.37772465e-01 4.52019833e-02 9.92885679e-02 -2.01715842e-01
-1.73939243e-01 4.62731570e-01 -4.22105998e-01 4.71443355e-01
1.31327045e+00 -1.44793883e-01 -4.31776762e-01 2.60716110e-01
1.12170458e+00 3.88314039e-01 -1.10389352e+00 -5.36252335e-02
2.58623481e-01 -7.96344936e-01 -5.42694509e-01 -5.82372665e-01
-1.18093634e+00 1.12584960e+00 1.77802682e-01 2.89662868e-01
7.44780362e-01 -1.93814367e-01 8.77418637e-01 4.14109379e-01
3.27990413e-01 -8.14161479e-01 2.53627330e-01 5.68183959e-01
8.40509117e-01 -1.38996685e+00 -9.44590941e-02 -5.55773556e-01
-5.70769608e-01 1.31456971e+00 1.12820160e+00 1.32487193e-01
3.72055799e-01 2.39997104e-01 1.57466292e-01 -1.14475168e-01
-8.94589841e-01 2.62821048e-01 2.52344795e-02 5.77129304e-01
1.07189131e+00 -3.60267371e-01 -6.35098875e-01 8.24251950e-01
-1.05427720e-01 -1.30297467e-01 1.25561476e+00 1.07204854e+00
-4.80905354e-01 -1.34792101e+00 -6.53101802e-02 6.64139152e-01
-5.70135534e-01 -4.86592025e-01 -4.62247431e-01 3.92641097e-01
2.01328695e-01 8.45538378e-01 -1.41637832e-01 -2.55977869e-01
2.73708522e-01 -2.25794478e-03 -2.36427218e-01 -1.32176816e+00
-8.55061591e-01 -4.13531333e-01 -1.02859363e-01 -1.21850044e-01
-2.95784116e-01 -6.47624433e-01 -1.39603496e+00 -2.39268079e-01
-6.76839113e-01 4.82316673e-01 3.91812027e-01 6.54582620e-01
5.69029927e-01 6.54865503e-01 4.77182209e-01 -2.33078316e-01
-4.15875494e-01 -1.21727836e+00 -1.38965055e-01 6.91463292e-01
1.19619995e-01 -3.91911447e-01 -3.17786843e-01 6.25988662e-01] | [7.650936126708984, 7.919658184051514] |
8244a763-ad86-4ac2-bf88-99c599990942 | edge-representation-learning-with-hypergraphs | 2106.15845 | null | https://arxiv.org/abs/2106.15845v2 | https://arxiv.org/pdf/2106.15845v2.pdf | Edge Representation Learning with Hypergraphs | Graph neural networks have recently achieved remarkable success in representing graph-structured data, with rapid progress in both the node embedding and graph pooling methods. Yet, they mostly focus on capturing information from the nodes considering their connectivity, and not much work has been done in representing the edges, which are essential components of a graph. However, for tasks such as graph reconstruction and generation, as well as graph classification tasks for which the edges are important for discrimination, accurately representing edges of a given graph is crucial to the success of the graph representation learning. To this end, we propose a novel edge representation learning framework based on Dual Hypergraph Transformation (DHT), which transforms the edges of a graph into the nodes of a hypergraph. This dual hypergraph construction allows us to apply message-passing techniques for node representations to edges. After obtaining edge representations from the hypergraphs, we then cluster or drop edges to obtain holistic graph-level edge representations. We validate our edge representation learning method with hypergraphs on diverse graph datasets for graph representation and generation performance, on which our method largely outperforms existing graph representation learning methods. Moreover, our edge representation learning and pooling method also largely outperforms state-of-the-art graph pooling methods on graph classification, not only because of its accurate edge representation learning, but also due to its lossless compression of the nodes and removal of irrelevant edges for effective message-passing. | ['Sung Ju Hwang', 'Minki Kang', 'DongKi Kim', 'Seul Lee', 'Jinheon Baek', 'Jaehyeong Jo'] | 2021-06-30 | null | http://proceedings.neurips.cc/paper/2021/hash/3def184ad8f4755ff269862ea77393dd-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/3def184ad8f4755ff269862ea77393dd-Paper.pdf | neurips-2021-12 | ['graph-reconstruction'] | ['graphs'] | [ 2.53954619e-01 2.35309482e-01 -3.87543172e-01 -9.42775421e-03
-3.64657551e-01 -5.43510854e-01 4.94412512e-01 6.58134639e-01
2.17297357e-02 3.85018855e-01 2.25695267e-01 -3.43971699e-01
-1.00902043e-01 -1.58644354e+00 -5.87360084e-01 -7.18940794e-01
-3.60475093e-01 3.04420620e-01 4.46670055e-02 -1.27771214e-01
3.20952124e-04 5.55357456e-01 -1.15737116e+00 1.63373008e-01
5.17131805e-01 7.77807057e-01 -3.91222313e-02 6.01762176e-01
-3.67705315e-01 8.27201605e-01 -4.58237886e-01 -4.76529598e-01
-1.83907093e-03 -4.22509938e-01 -6.86627924e-01 9.58946794e-02
4.28117096e-01 6.13319082e-03 -1.11151230e+00 1.16725278e+00
4.45037991e-01 2.80331910e-01 6.67924583e-01 -1.30460799e+00
-1.07641637e+00 9.13665652e-01 -6.42952919e-01 -5.45752905e-02
3.40553731e-01 -1.57953620e-01 1.69161808e+00 -5.63210845e-01
7.99917519e-01 1.11717641e+00 5.89905083e-01 3.39093298e-01
-1.28637302e+00 -6.13907218e-01 3.42184335e-01 3.18203896e-01
-1.44308257e+00 5.67009002e-02 1.11023557e+00 -2.50355005e-01
1.08176768e+00 3.07533085e-01 9.07077909e-01 7.75167525e-01
1.52133882e-01 7.50150442e-01 4.49906260e-01 -2.25374550e-01
-1.41930431e-02 -2.56974727e-01 4.26573426e-01 1.15235293e+00
5.47702253e-01 -3.41586292e-01 -1.91744357e-01 -1.08864503e-02
7.97294080e-01 3.05859387e-01 -5.23644090e-01 -3.47627789e-01
-9.30871844e-01 9.76849437e-01 1.25844026e+00 2.99568862e-01
-4.44786757e-01 6.01721585e-01 4.81072366e-01 3.57750773e-01
5.03952920e-01 2.46823132e-01 1.06281459e-01 4.32492226e-01
-3.72445226e-01 9.66128185e-02 7.28947699e-01 9.98906255e-01
9.43349600e-01 3.18767428e-02 -5.33646107e-01 6.68386519e-01
1.20191865e-01 9.24312994e-02 3.35026711e-01 -9.00377408e-02
7.13216305e-01 1.42662203e+00 -6.28835857e-01 -1.58088994e+00
-4.32108194e-01 -4.87726331e-01 -1.27446997e+00 -3.29933882e-01
9.84962210e-02 2.65380740e-01 -1.12379551e+00 1.75392199e+00
-6.89218240e-03 4.18354630e-01 -6.22663535e-02 5.04346430e-01
1.34262812e+00 6.77977204e-01 8.32129344e-02 2.16415405e-01
1.43352878e+00 -7.26515055e-01 -4.31861073e-01 -3.27889472e-01
8.31466377e-01 -9.55153778e-02 8.18663359e-01 -1.20989799e-01
-8.95875335e-01 -2.99146831e-01 -1.07035744e+00 -2.73074478e-01
-6.89548016e-01 -1.58173338e-01 9.95360672e-01 4.26497608e-01
-9.83904719e-01 6.33756697e-01 -5.95602572e-01 -3.08409721e-01
6.97198927e-01 2.35503927e-01 -8.30589294e-01 -5.58741391e-01
-1.09833944e+00 4.84366626e-01 4.91275221e-01 1.20664276e-01
-5.56207597e-01 -5.73024094e-01 -1.52614510e+00 6.54985309e-01
2.39806309e-01 -6.25691295e-01 5.03177226e-01 -5.16079724e-01
-8.24388146e-01 8.12235236e-01 -2.25950535e-02 -4.19243217e-01
-9.93581787e-02 4.09153968e-01 -3.06457311e-01 3.19011003e-01
-1.91034779e-01 4.37299669e-01 6.84615552e-01 -1.00758910e+00
-1.85673371e-01 -4.58801717e-01 2.81820774e-01 5.32890745e-02
-5.66038787e-01 -3.46157849e-01 -7.20987082e-01 -7.46043861e-01
4.68156219e-01 -8.28547657e-01 -2.16873333e-01 -9.56367850e-02
-6.32003903e-01 -2.72316843e-01 6.96400583e-01 -7.27827966e-01
1.47345257e+00 -2.21332598e+00 3.64583462e-01 4.62063074e-01
1.03100705e+00 1.65559337e-01 -5.22947311e-01 7.54714310e-01
-2.79538184e-01 3.45477134e-01 -3.31340969e-01 -1.71993703e-01
8.37629959e-02 1.92110598e-01 -3.63643318e-01 4.83210772e-01
3.75447154e-01 1.38187468e+00 -1.05756545e+00 -3.13802689e-01
1.40483573e-01 7.26813614e-01 -5.29286146e-01 1.81976184e-01
5.21332072e-03 -1.09247543e-01 -5.13533175e-01 4.94494468e-01
5.82602859e-01 -5.59422195e-01 5.34579813e-01 -2.96701670e-01
6.65120900e-01 3.37196022e-01 -9.77703273e-01 1.41830480e+00
-4.30469602e-01 6.13048375e-01 -1.83788672e-01 -1.33644760e+00
1.09309745e+00 1.45461902e-01 4.84038651e-01 -6.51179075e-01
6.43881559e-02 -1.24775760e-01 8.58173594e-02 -1.19157501e-01
4.53108072e-01 1.87029034e-01 -7.66398981e-02 5.06055534e-01
2.12244287e-01 -5.70536703e-02 4.87189889e-01 7.77226806e-01
1.55310500e+00 -3.40721577e-01 3.17416251e-01 7.79437870e-02
3.83383811e-01 -4.95249152e-01 3.62465829e-01 4.94716167e-01
1.03363611e-01 7.98221350e-01 9.68945622e-01 -4.58692819e-01
-5.99880397e-01 -1.00720060e+00 2.57949948e-01 8.34116638e-01
1.15727663e-01 -8.96366298e-01 -5.67687988e-01 -7.82357097e-01
2.01917872e-01 2.30275840e-01 -8.62510681e-01 -6.26012146e-01
-6.42198443e-01 -8.84127140e-01 3.86956096e-01 7.03795433e-01
3.59822601e-01 -1.10619259e+00 1.06125005e-01 2.71458298e-01
-6.26733750e-02 -1.13293588e+00 -6.13155961e-01 1.54398754e-01
-7.44352281e-01 -1.46640980e+00 -4.74831372e-01 -1.15430534e+00
1.18791163e+00 6.30853236e-01 1.26164424e+00 7.66333640e-01
-4.57399249e-01 2.70699114e-01 -4.52794284e-01 1.15385525e-01
-1.82226151e-01 2.53556371e-01 -4.74630237e-01 9.39952061e-02
1.91673040e-01 -7.73958266e-01 -3.94378662e-01 9.67261791e-02
-9.84107375e-01 1.43799171e-01 5.66522419e-01 8.57920885e-01
6.97005510e-01 2.58158058e-01 4.30792660e-01 -1.30197692e+00
9.93161440e-01 -5.81383944e-01 -4.00223553e-01 4.36651558e-01
-4.32963073e-01 1.94958404e-01 8.15308630e-01 -1.95839912e-01
-3.05665910e-01 -3.51723224e-01 1.38008343e-02 -3.33518505e-01
3.83423656e-01 9.11376178e-01 -2.07246467e-01 -3.74884069e-01
4.34500158e-01 2.27765605e-01 -1.39993764e-02 -1.97650105e-01
5.97442031e-01 2.04042301e-01 2.70264626e-01 -4.55107599e-01
8.04666996e-01 2.02598631e-01 4.46987361e-01 -8.13602090e-01
-3.53304237e-01 -3.59892130e-01 -4.19419020e-01 1.06635196e-02
5.69929779e-01 -6.07476234e-01 -6.26174212e-01 3.18663597e-01
-1.00600410e+00 -1.09250978e-01 -1.30687490e-01 1.78446010e-01
-1.11717828e-01 6.17789030e-01 -9.75818157e-01 -4.51072991e-01
-4.21075016e-01 -1.12796187e+00 9.59015012e-01 4.35579717e-02
7.46383891e-02 -1.27435625e+00 -7.12655112e-02 -3.31162997e-02
4.29349303e-01 5.31715989e-01 1.51575553e+00 -6.02143943e-01
-7.74295807e-01 -7.60096431e-01 -6.99053824e-01 1.59906015e-01
2.21274525e-01 -1.14910178e-01 -6.48450255e-01 -5.72350621e-01
-8.08373213e-01 -1.44667089e-01 1.29016542e+00 2.58908540e-01
1.25983989e+00 -2.62375355e-01 -5.21190822e-01 8.29225659e-01
1.36465013e+00 -3.68835002e-01 7.20030546e-01 -2.27485046e-01
1.12795293e+00 5.50663412e-01 -2.56241441e-01 6.54549375e-02
4.54473525e-01 4.71380472e-01 5.81707835e-01 -2.06370160e-01
-4.85664189e-01 -5.61801255e-01 2.26824358e-01 9.36368227e-01
-3.78033407e-02 -6.45094216e-01 -6.60385966e-01 4.29840118e-01
-1.86484623e+00 -7.43877351e-01 -2.56203651e-01 2.23636079e+00
2.93341249e-01 1.17564760e-01 -3.05200107e-02 1.93269432e-01
8.57190609e-01 6.16358817e-01 -3.99703652e-01 -2.66804516e-01
-3.95106450e-02 4.32586819e-01 4.13519382e-01 3.03625792e-01
-9.72141266e-01 9.32722449e-01 5.53203201e+00 5.40534735e-01
-8.77894580e-01 -3.12576890e-01 5.24606347e-01 4.04347539e-01
-6.64019704e-01 -9.25764814e-02 -2.93072373e-01 1.88788876e-01
5.15253425e-01 -4.04752582e-01 7.30512559e-01 6.99291289e-01
-4.41696018e-01 5.11036277e-01 -1.07012784e+00 1.09389377e+00
1.25056669e-01 -1.42038763e+00 4.74235892e-01 1.89458847e-01
5.56696773e-01 -1.11975171e-01 -1.76078901e-01 5.80695629e-01
3.42391759e-01 -1.29402149e+00 8.28846693e-02 1.65202588e-01
6.63392961e-01 -7.36381948e-01 7.36309528e-01 -8.74316022e-02
-1.95723546e+00 5.04496209e-02 -6.36914730e-01 6.24898337e-02
-7.38647394e-03 5.91298163e-01 -7.16199458e-01 9.42138255e-01
1.95628315e-01 8.94199550e-01 -6.61250770e-01 9.93191659e-01
-7.44572997e-01 5.32428682e-01 -5.28693683e-02 3.43164429e-02
3.09028864e-01 -4.60689873e-01 2.97241092e-01 1.16697574e+00
2.78818905e-01 6.47447957e-03 3.37392032e-01 1.06690145e+00
-8.18495452e-01 2.06059352e-01 -9.24995959e-01 -6.94512486e-01
4.71593559e-01 1.57168806e+00 -9.74135101e-01 -2.00745791e-01
-5.27925849e-01 7.75627375e-01 9.58637834e-01 5.56747258e-01
-5.75949252e-01 -7.93029308e-01 7.69547522e-01 3.68235186e-02
2.77414918e-01 -3.08502793e-01 1.00957230e-01 -1.06624317e+00
1.37573421e-01 -5.26861906e-01 6.93709850e-01 -4.84658509e-01
-1.30838335e+00 7.64174104e-01 -1.73890218e-01 -9.22737002e-01
4.58919269e-04 -7.85154641e-01 -1.02943897e+00 7.94331491e-01
-1.49185455e+00 -1.49493659e+00 -7.05238938e-01 4.59404141e-01
-1.33247286e-01 -3.66819315e-02 8.02239060e-01 3.69844466e-01
-6.16683483e-01 6.88237011e-01 -3.57455969e-01 5.31574488e-01
2.20409378e-01 -1.34573936e+00 9.28892970e-01 8.42382252e-01
5.87791443e-01 7.70933092e-01 2.49232557e-02 -6.63071394e-01
-1.92732692e+00 -1.29604948e+00 6.20714247e-01 -9.35500413e-02
6.65968955e-01 -7.18468487e-01 -1.22554016e+00 8.86765659e-01
-2.22370118e-01 5.75304568e-01 4.84675169e-01 2.50461400e-01
-6.39810503e-01 -3.97174284e-02 -8.85199308e-01 6.37836754e-01
1.27713633e+00 -7.68273056e-01 -1.07451059e-01 3.86746645e-01
8.03428233e-01 -2.63494134e-01 -8.53936434e-01 2.75181890e-01
4.66437340e-01 -5.39156199e-01 1.08073080e+00 -8.76109540e-01
4.04768437e-01 -1.16867661e-01 -4.67575388e-03 -1.56862915e+00
-7.26014078e-01 -4.15589780e-01 -2.24871174e-01 1.12262726e+00
3.56583506e-01 -8.97027731e-01 9.50614512e-01 2.54798919e-01
-1.45037189e-01 -7.58188903e-01 -6.17676973e-01 -4.53471452e-01
-2.07911447e-01 -2.59293407e-01 8.54898870e-01 1.01690102e+00
4.12429422e-02 4.02129799e-01 -1.46720126e-01 2.37706363e-01
5.17852962e-01 3.57401609e-01 7.47281134e-01 -1.25765634e+00
-1.67015016e-01 -8.37291360e-01 -1.07136762e+00 -7.48553157e-01
4.34971124e-01 -1.65893841e+00 -2.80428350e-01 -2.14616251e+00
2.77579099e-01 -1.93096206e-01 -5.26361406e-01 6.39488339e-01
-4.79414523e-01 3.90645295e-01 9.87447798e-02 -1.69112802e-01
-4.01933879e-01 5.86687088e-01 1.39421630e+00 -6.27974749e-01
-1.13603763e-01 -2.53571004e-01 -8.97354901e-01 2.79863417e-01
5.63794255e-01 -2.31616765e-01 -5.82021773e-01 -4.45755154e-01
4.22511607e-01 -1.13968275e-01 4.46251959e-01 -7.59487927e-01
1.09062865e-01 2.14057624e-01 8.03131312e-02 -9.13299620e-02
1.39446720e-01 -5.45038700e-01 1.22681424e-01 3.83874476e-01
-2.23371759e-01 1.25790060e-01 1.47034181e-02 7.53747761e-01
-3.53555471e-01 -5.97955771e-02 5.40142417e-01 -1.73782125e-01
-8.21420491e-01 8.78142893e-01 2.23882273e-01 1.20702803e-01
8.05702746e-01 -2.54173696e-01 -5.99685311e-01 -4.85916853e-01
-6.41546190e-01 2.30091572e-01 4.05173987e-01 3.81225705e-01
8.56669247e-01 -1.45863032e+00 -8.08398128e-01 4.89497602e-01
4.21995103e-01 -2.75825765e-02 2.17824504e-01 6.16716146e-01
-5.09234786e-01 9.03739780e-02 -1.13949083e-01 -1.49969652e-01
-1.11962104e+00 6.08377039e-01 1.66190878e-01 -4.61005479e-01
-1.07165897e+00 8.44576776e-01 5.38040102e-01 -3.18583161e-01
2.62622505e-01 -5.40095627e-01 -2.97871262e-01 5.22948010e-03
4.64720756e-01 1.99365571e-01 2.34716177e-01 -5.26026487e-01
-2.33992711e-01 5.36738098e-01 -1.96579114e-01 6.40827239e-01
1.26425481e+00 4.73775506e-01 -3.79815519e-01 1.06476480e-02
1.40421724e+00 -1.00097753e-01 -7.34388709e-01 -2.59382814e-01
-2.43748561e-01 -4.77017969e-01 1.43202513e-01 -2.22834408e-01
-1.47980845e+00 1.03209937e+00 -3.77765447e-02 5.97886205e-01
9.06577826e-01 1.63719311e-01 9.69335020e-01 3.71324182e-01
3.35275501e-01 -4.37725574e-01 6.15539812e-02 2.70398438e-01
9.05194998e-01 -1.14628816e+00 2.22657934e-01 -8.02270114e-01
-2.66623676e-01 1.10822105e+00 4.45011050e-01 -4.53027070e-01
5.63935518e-01 -6.59041330e-02 -5.91653407e-01 -5.11714578e-01
-5.51424921e-01 -2.73840964e-01 6.70074880e-01 6.62248850e-01
4.78227705e-01 4.29894269e-01 -1.14538878e-01 5.89385986e-01
1.15737934e-02 -5.20407379e-01 3.32027286e-01 6.30039811e-01
-2.45508969e-01 -1.13164651e+00 2.71556467e-01 7.85897553e-01
3.12462728e-03 -2.89914697e-01 -7.80243754e-01 9.02584374e-01
-4.41408187e-01 5.48020005e-01 6.20773137e-02 -6.10864699e-01
3.61951172e-01 -2.25575924e-01 5.12596071e-01 -9.08818424e-01
-4.88855332e-01 -5.38984418e-01 1.00716770e-01 -4.97980952e-01
1.44534260e-01 -1.35427555e-02 -1.39449167e+00 -4.62234229e-01
-3.36592108e-01 1.55861005e-01 4.13262486e-01 5.73206007e-01
5.43300450e-01 8.26487899e-01 5.17926037e-01 -9.37580585e-01
-1.87312245e-01 -8.76464248e-01 -8.78518999e-01 8.59803796e-01
5.24974838e-02 -6.82233274e-01 -3.57602626e-01 -5.62020183e-01] | [7.0993804931640625, 6.303421497344971] |
fd248292-2b1e-4a88-951e-c1935c3ce209 | unsupervised-multi-view-pedestrian-detection | 2305.12457 | null | https://arxiv.org/abs/2305.12457v1 | https://arxiv.org/pdf/2305.12457v1.pdf | Unsupervised Multi-view Pedestrian Detection | With the prosperity of the video surveillance, multiple visual sensors have been applied for an accurate localization of pedestrians in a specific area, which facilitate various applications like intelligent safety or new retailing. However, previous methods rely on the supervision from the human annotated pedestrian positions in every video frame and camera view, which is a heavy burden in addition to the necessary camera calibration and synchronization. Therefore, we propose in this paper an Unsupervised Multi-view Pedestrian Detection approach (UMPD) to eliminate the need of annotations to learn a multi-view pedestrian detector. 1) Firstly, Semantic-aware Iterative Segmentation (SIS) is proposed to extract discriminative visual representations of the input images from different camera views via an unsupervised pretrained model, then convert them into 2D segments of pedestrians, based on our proposed iterative Principal Component Analysis and the zero-shot semantic classes from the vision-language pretrained models. 2) Secondly, we propose Vertical-aware Differential Rendering (VDR) to not only learn the densities and colors of 3D voxels by the masks of SIS, images and camera poses, but also constraint the voxels to be vertical towards the ground plane, following the physical characteristics of pedestrians. 3) Thirdly, the densities of 3D voxels learned by VDR are projected onto Bird-Eyes-View as the final detection results. Extensive experiments on popular multi-view pedestrian detection benchmarks, i.e., Wildtrack and MultiviewX, show that our proposed UMPD approach, as the first unsupervised method to our best knowledge, performs competitively with the previous state-of-the-art supervised techniques. Code will be available. | ['Xu-Cheng Yin', 'Shiqi Ren', 'Chao Zhu', 'Mengyin Liu'] | 2023-05-21 | null | null | null | null | ['camera-calibration', 'pedestrian-detection'] | ['computer-vision', 'computer-vision'] | [-8.03340077e-02 -3.93657207e-01 1.10357277e-01 -2.68923581e-01
-3.40173572e-01 -4.90866393e-01 5.52217126e-01 -8.55548754e-02
-7.09396780e-01 3.60049903e-01 -2.10441381e-01 5.57707883e-02
5.36830783e-01 -7.79969275e-01 -7.38749444e-01 -8.34756613e-01
3.60964447e-01 4.67755169e-01 9.27774787e-01 -8.84342764e-04
6.28122315e-02 4.42466706e-01 -1.70157313e+00 7.08424225e-02
6.63581312e-01 8.78742993e-01 2.82940626e-01 6.03104770e-01
-1.88702624e-02 2.40042165e-01 -2.44519323e-01 -6.63987279e-01
3.56050372e-01 -7.93494135e-02 -3.04965466e-01 6.51710272e-01
3.45284551e-01 -5.13126612e-01 -3.31736684e-01 1.22093606e+00
2.89618224e-01 2.59722859e-01 7.11102188e-01 -1.32640743e+00
-3.30424756e-01 -8.05967376e-02 -1.10383642e+00 3.85708421e-01
4.61342216e-01 3.92251223e-01 6.39263511e-01 -1.04913712e+00
4.19478297e-01 1.32236004e+00 4.96409893e-01 4.85504091e-01
-9.82182741e-01 -5.18636286e-01 6.36467576e-01 3.35481316e-01
-1.60284996e+00 -2.81393081e-01 9.32187080e-01 -5.99425256e-01
5.62916815e-01 2.19542027e-01 1.03304124e+00 1.09082556e+00
1.63603853e-02 9.50777173e-01 9.99312580e-01 -1.94789052e-01
2.56014556e-01 3.77788454e-01 1.49216339e-01 8.47684860e-01
5.01111388e-01 3.42644826e-02 -1.57504514e-01 9.57124606e-02
8.06217790e-01 4.60231453e-01 -1.69044673e-01 -7.62161791e-01
-8.93665493e-01 5.86336672e-01 4.00020897e-01 -4.42860685e-02
-2.56452113e-01 -8.18961114e-02 3.32584679e-01 -4.79857475e-01
3.53390932e-01 -5.59397817e-01 -1.78689748e-01 3.07760715e-01
-8.54280949e-01 4.24745753e-02 4.17878211e-01 1.10782409e+00
7.92103827e-01 -4.92487960e-02 1.30681247e-01 5.29441774e-01
7.18829691e-01 7.90118337e-01 1.30698428e-01 -6.00680947e-01
6.33432269e-01 7.65750647e-01 3.07298332e-01 -1.07408953e+00
-2.33827665e-01 -3.60149920e-01 -8.20214033e-01 1.59659207e-01
6.11188471e-01 6.81642964e-02 -9.00844812e-01 1.15494490e+00
6.79719925e-01 2.51416713e-01 -9.99789592e-03 1.34846246e+00
8.50608230e-01 6.12568617e-01 1.83760732e-01 5.56013286e-02
1.54181027e+00 -8.79724145e-01 -1.37359872e-01 -5.07556856e-01
1.26694128e-01 -6.32614732e-01 8.19431484e-01 4.34158742e-01
-7.82733202e-01 -7.15556979e-01 -9.26596642e-01 7.61252865e-02
-3.71754110e-01 4.39237744e-01 4.85609710e-01 1.14673078e+00
-6.26572609e-01 -1.55303299e-01 -1.01987076e+00 -5.73059082e-01
4.99835610e-01 3.57577801e-02 -3.90481859e-01 -3.61178488e-01
-7.49168932e-01 5.96477091e-01 2.74078369e-01 3.42942655e-01
-1.22156942e+00 -4.05508906e-01 -8.97118151e-01 -4.59440164e-02
5.24123013e-01 -7.43535817e-01 6.00840151e-01 -8.22263241e-01
-1.16658473e+00 9.80519295e-01 -1.82710797e-01 -3.41105074e-01
7.41673291e-01 -2.58984327e-01 -1.44145921e-01 2.06176072e-01
2.94859737e-01 6.46718621e-01 7.23074675e-01 -1.64068663e+00
-9.24105346e-01 -5.97945452e-01 2.76851878e-02 2.92221844e-01
-1.52691320e-01 8.74026492e-02 -1.07094705e+00 -3.38322490e-01
1.29800364e-01 -8.52437556e-01 -5.04222095e-01 2.74240542e-02
-6.14388824e-01 -1.14839844e-01 9.03846562e-01 -8.71793747e-01
6.45702720e-01 -2.13731074e+00 2.43846372e-01 1.79798543e-01
1.50515720e-01 1.60576880e-01 2.73299903e-01 -2.28733078e-01
2.33150288e-01 -2.89423496e-01 -2.64120847e-01 -5.57085216e-01
-1.49741277e-01 1.35330603e-01 6.11548126e-02 8.22060943e-01
4.93862331e-02 5.79667747e-01 -9.04187799e-01 -7.94242084e-01
8.22578907e-01 5.62592566e-01 -4.47385848e-01 2.94148237e-01
-2.13858094e-02 5.61038673e-01 -4.85554725e-01 7.80665159e-01
1.01259041e+00 -1.14572514e-02 -1.50049359e-01 -3.73811066e-01
-3.24883610e-01 -4.41690832e-01 -1.47365546e+00 1.49550188e+00
-1.24829620e-01 1.36000291e-01 1.59997538e-01 -1.03720355e+00
7.48737097e-01 1.39861852e-02 4.91867572e-01 -4.59224105e-01
1.32014483e-01 -3.56607847e-02 -4.12354976e-01 -3.91598016e-01
3.19218546e-01 3.38457584e-01 -5.16920909e-02 -6.83431886e-03
8.15016311e-03 1.56850308e-01 3.16793442e-01 2.77307063e-01
3.75202477e-01 2.64901370e-01 1.98777303e-01 4.52330709e-02
9.66009319e-01 -1.01353057e-01 7.66274691e-01 4.52681035e-01
-3.81275177e-01 7.20571995e-01 2.57275909e-01 -5.31796455e-01
-9.46869552e-01 -1.33931828e+00 6.23111334e-03 7.09233105e-01
5.02275586e-01 -2.11842149e-01 -1.01058102e+00 -8.42749655e-01
-2.29615524e-01 5.51216900e-01 -2.31370375e-01 2.95363456e-01
-5.99767447e-01 -8.94345641e-01 2.42544681e-01 7.14598298e-01
7.50844002e-01 -4.73245025e-01 -9.52758431e-01 -1.60450786e-01
-1.66754454e-01 -1.57212138e+00 -5.42802989e-01 -4.22920845e-02
-5.10091424e-01 -1.28801095e+00 -1.03748083e+00 -6.68915451e-01
8.06672454e-01 7.07526565e-01 7.01445818e-01 -1.20131850e-01
-3.59404951e-01 7.64319301e-01 -1.19728483e-01 -7.29561271e-03
1.49269566e-01 -2.24245697e-01 1.89214751e-01 4.72029954e-01
5.64581633e-01 -4.25659001e-01 -7.88413346e-01 5.29302418e-01
-5.39773822e-01 1.97595745e-01 6.53650582e-01 5.15452623e-01
9.27129745e-01 7.69439712e-02 -1.24339446e-01 -6.31859183e-01
-1.85195759e-01 -2.93183446e-01 -1.00244808e+00 2.62692511e-01
-5.48398122e-02 -4.25226599e-01 5.85561216e-01 -1.88164130e-01
-1.12567544e+00 6.45585358e-01 -5.49401119e-02 -6.43001974e-01
-7.69433677e-01 -2.23683730e-01 -7.77548254e-01 -1.09189085e-03
2.99627334e-01 4.77194786e-01 -3.92137617e-01 -3.72977048e-01
5.48330009e-01 5.71813226e-01 5.81912398e-01 -4.24799800e-01
8.16114902e-01 8.31550002e-01 -7.48211294e-02 -1.07329035e+00
-4.92103517e-01 -8.87775958e-01 -1.05766213e+00 -5.83270371e-01
1.43018496e+00 -1.08571029e+00 -5.96168935e-01 5.09318113e-01
-1.40593123e+00 1.55398160e-01 1.01323225e-01 5.71772933e-01
-3.42585832e-01 7.96538413e-01 -4.00855482e-01 -1.11049747e+00
-1.60586223e-01 -1.44168115e+00 1.23079693e+00 5.32141984e-01
3.54550272e-01 -7.86257625e-01 -4.37167436e-01 6.09562933e-01
-2.25205302e-01 1.42621368e-01 3.67952138e-01 -3.73203486e-01
-9.71073747e-01 -7.88376480e-02 -4.16913539e-01 3.66681457e-01
-2.04376638e-01 -9.44903865e-02 -1.03057468e+00 -1.76313967e-01
-1.79507837e-01 1.80648401e-01 8.99694502e-01 6.37544274e-01
8.52529764e-01 1.51608810e-01 -6.88542247e-01 6.87479734e-01
1.33098042e+00 1.07129849e-01 3.55586350e-01 2.56316602e-01
1.03329134e+00 5.88123322e-01 5.78718007e-01 4.92176056e-01
6.47555053e-01 8.22620332e-01 5.37976682e-01 -1.48680642e-01
-6.49960339e-02 -3.24790806e-01 4.47693288e-01 5.60907304e-01
-3.71734232e-01 -2.55782772e-02 -8.31017435e-01 4.83841717e-01
-1.81270206e+00 -8.95701408e-01 -4.70303804e-01 2.42187953e+00
9.25090238e-02 3.03514838e-01 5.89033127e-01 -5.98554239e-02
9.36805129e-01 1.60588082e-02 -5.69112539e-01 2.28681713e-01
1.72314662e-02 -1.96848288e-01 7.55108476e-01 3.48556429e-01
-1.49860346e+00 9.05585408e-01 4.84553862e+00 6.37960494e-01
-7.39980340e-01 3.46126139e-01 7.74066091e-01 4.51957360e-02
1.73276708e-01 -1.12639099e-01 -1.26970649e+00 5.74621141e-01
3.33555788e-01 4.07401174e-01 2.27876857e-01 1.17882299e+00
2.48432726e-01 -2.59913683e-01 -8.91386747e-01 1.33468521e+00
3.68314117e-01 -8.37349415e-01 1.41229453e-02 -1.03727239e-03
5.25495887e-01 -2.35669091e-01 4.27104793e-02 2.31705187e-03
1.19190045e-01 -5.22184074e-01 9.17121649e-01 5.40450275e-01
3.76887470e-01 -8.89950931e-01 7.73347378e-01 4.76245344e-01
-1.71917319e+00 -4.71107988e-03 -5.74731290e-01 3.01680863e-01
4.34837073e-01 5.51888525e-01 -3.90063703e-01 6.95331097e-01
9.70334351e-01 8.29791009e-01 -8.22421253e-01 1.08014596e+00
-3.32530141e-01 3.12522173e-01 -4.05682564e-01 7.04569146e-02
5.78913614e-02 -4.19724941e-01 6.12637699e-01 1.29806828e+00
1.59861580e-01 4.61810678e-02 6.43575549e-01 9.81694460e-01
4.71635580e-01 4.23244834e-02 -4.34463084e-01 4.42440718e-01
1.36729807e-01 1.42775118e+00 -1.16664445e+00 -3.27743858e-01
-8.08276772e-01 1.31754935e+00 -5.57429828e-02 4.62274104e-01
-9.54918146e-01 -1.61869582e-02 4.87272888e-01 3.74357879e-01
5.79230070e-01 -4.70431864e-01 -1.55229762e-01 -1.38802612e+00
1.75717041e-01 -4.48355228e-01 3.80096704e-01 -5.57287157e-01
-1.21623993e+00 4.08693790e-01 1.67800620e-01 -1.30411148e+00
1.17174173e-02 -7.35501111e-01 -4.67232376e-01 7.83517063e-01
-1.45531261e+00 -1.36979342e+00 -5.12959719e-01 7.50551879e-01
7.13412225e-01 -1.53074235e-01 3.52783144e-01 5.49678624e-01
-7.78856933e-01 2.81808496e-01 -3.13479155e-01 3.94521028e-01
3.18833888e-01 -1.03288174e+00 4.17939186e-01 1.25152338e+00
7.73937851e-02 3.49215508e-01 3.89391154e-01 -6.40054643e-01
-1.30514574e+00 -1.24030375e+00 2.92225361e-01 -4.85953063e-01
1.86245903e-01 -5.13733625e-01 -5.49279571e-01 5.83595872e-01
3.45621035e-02 2.30094433e-01 4.74629939e-01 -3.19711000e-01
-1.38996214e-01 -1.80275470e-01 -9.74605978e-01 4.26578999e-01
1.03266263e+00 -3.03027242e-01 -4.00869995e-01 2.88008749e-01
3.58341694e-01 -4.90202934e-01 -3.70769590e-01 1.70930196e-02
3.91097248e-01 -1.12314570e+00 1.52053237e+00 -2.80180067e-01
2.87821423e-02 -8.38364720e-01 -3.42577279e-01 -9.21770096e-01
-8.92065316e-02 1.24364659e-01 3.22422236e-02 1.35190606e+00
-1.05672233e-01 -5.00286758e-01 8.98982942e-01 5.98441958e-01
-1.02369063e-01 -4.15438086e-01 -1.10554409e+00 -5.18496335e-01
-3.54841292e-01 -4.43073660e-01 1.86388120e-01 4.69981641e-01
-5.48131168e-01 3.04940730e-01 -2.63760120e-01 7.60666192e-01
1.15921783e+00 -5.74452728e-02 1.02865827e+00 -1.25661910e+00
-3.63174051e-01 -2.27082133e-01 -8.61307919e-01 -1.43482888e+00
-8.33371878e-02 -6.71783149e-01 -1.19603917e-01 -1.53069317e+00
5.81677556e-01 -2.15741023e-01 -7.09333494e-02 -1.15272157e-01
-2.89932936e-01 2.14261562e-01 3.20231676e-01 4.57463153e-02
-9.11003351e-01 5.21050334e-01 9.16983128e-01 -3.10787380e-01
-1.14737429e-01 2.22489655e-01 -3.10558408e-01 1.03546584e+00
4.19640392e-01 -2.13898391e-01 -2.46332496e-01 -2.88240612e-01
-2.94826537e-01 -2.09687293e-01 9.71901119e-01 -1.25289202e+00
3.82991165e-01 -1.52359277e-01 7.96086073e-01 -8.88176978e-01
5.99216998e-01 -9.28464413e-01 -1.22449156e-02 4.52026427e-01
3.15801203e-01 5.00797853e-02 -5.32323308e-02 9.37490940e-01
8.17258656e-02 -1.74988225e-01 9.18670774e-01 -4.98279780e-01
-1.02605486e+00 3.08145911e-01 -4.01396275e-01 -1.14648426e-02
1.33807456e+00 -4.14685518e-01 5.06660603e-02 1.96298100e-02
-6.69435978e-01 2.07345486e-01 5.92545569e-01 3.27397913e-01
8.68620455e-01 -1.13204694e+00 -4.92459655e-01 4.56298470e-01
-2.95931995e-02 2.92007685e-01 4.46023852e-01 8.64442408e-01
-4.77723002e-01 4.14130598e-01 -6.39885515e-02 -1.06675982e+00
-1.35818255e+00 8.48978758e-01 3.94232899e-01 5.44284470e-02
-8.28121841e-01 6.10339046e-01 5.37411869e-01 -2.37407610e-01
2.65044779e-01 -3.68349642e-01 -3.13374996e-01 -1.10379197e-01
6.11051321e-01 4.34671909e-01 -1.82949841e-01 -1.07197201e+00
-5.58691800e-01 8.56196940e-01 -2.34408258e-03 -1.75442994e-01
1.03561914e+00 -2.71686345e-01 2.07094908e-01 2.39601985e-01
8.81644011e-01 -5.87539449e-02 -1.51423156e+00 -7.13458732e-02
-2.69443095e-01 -6.87260926e-01 -9.11539346e-02 -2.05580190e-01
-1.24993825e+00 1.18147111e+00 9.19893086e-01 -2.75790244e-01
7.86473930e-01 4.73928172e-03 8.06403518e-01 7.16534778e-02
5.60693741e-01 -1.08033431e+00 4.03626077e-02 4.67719920e-02
2.44590223e-01 -1.23106933e+00 -8.60402267e-03 -7.94010222e-01
-7.68038571e-01 1.02427137e+00 8.49985719e-01 -5.87402061e-02
5.62594235e-01 5.91456778e-02 -1.32610515e-01 1.81342289e-01
-1.50826760e-02 -5.15827596e-01 1.34370551e-01 9.88762438e-01
6.40307814e-02 8.02436173e-02 6.35977313e-02 8.49105954e-01
3.64831656e-01 -2.87255138e-01 2.40710437e-01 5.65372765e-01
-4.51653898e-01 -7.78002977e-01 -8.71574581e-01 1.01486906e-01
4.91783060e-02 1.58010289e-01 -2.48391539e-01 6.17050231e-01
5.19984186e-01 1.05064094e+00 2.67157629e-02 -2.62239933e-01
4.71240729e-01 -1.42763719e-01 5.03387630e-01 -4.54012781e-01
-1.60075232e-01 1.85620695e-01 -2.14969605e-01 -4.23411936e-01
-5.38336098e-01 -9.48451281e-01 -1.18723977e+00 -4.81061973e-02
-2.49862954e-01 -1.62662730e-01 5.49634755e-01 9.74974573e-01
-1.00385640e-02 3.36683661e-01 4.41338450e-01 -1.25683391e+00
-1.31283253e-01 -5.25826514e-01 -4.43900079e-01 4.53480005e-01
-5.68080842e-02 -1.04167056e+00 -1.97396636e-01 2.49997228e-01] | [7.692292213439941, -0.8056875467300415] |
831cf9ce-2a2b-4a43-9a0f-24b17709d51b | a-comparison-of-strategies-for-source-free-1 | null | null | https://aclanthology.org/2022.acl-long.572 | https://aclanthology.org/2022.acl-long.572.pdf | A Comparison of Strategies for Source-Free Domain Adaptation | Data sharing restrictions are common in NLP, especially in the clinical domain, but there is limited research on adapting models to new domains without access to the original training data, a setting known as source-free domain adaptation. We take algorithms that traditionally assume access to the source-domain training data—active learning, self-training, and data augmentation—and adapt them for source free domain adaptation. Then we systematically compare these different strategies across multiple tasks and domains. We find that active learning yields consistent gains across all SemEval 2021 Task 10 tasks and domains, but though the shared task saw successful self-trained and data augmented models, our systematic comparison finds these strategies to be unreliable for source-free domain adaptation. | ['Steven Bethard', 'Yiyun Zhao', 'Xin Su'] | null | null | null | null | acl-2022-5 | ['source-free-domain-adaptation'] | ['computer-vision'] | [ 5.60842812e-01 7.70167887e-01 -9.88948882e-01 -6.40237570e-01
-1.23215008e+00 -6.69982672e-01 5.92095494e-01 5.15504599e-01
-9.20258701e-01 1.42309070e+00 5.01308382e-01 -3.51508230e-01
-2.29868278e-01 -4.46470350e-01 -6.53123677e-01 -4.74216998e-01
4.32587042e-02 1.03620410e+00 8.55912864e-02 -1.17437176e-01
-1.52948514e-01 2.37748265e-01 -7.19913125e-01 3.29872698e-01
1.14785814e+00 3.48448306e-01 -9.41586271e-02 4.17858750e-01
-4.27072495e-01 5.88977396e-01 -6.30641758e-01 -3.86966318e-01
3.73172224e-01 -3.02786112e-01 -1.13098609e+00 -1.67137936e-01
1.98969513e-01 -2.55472004e-01 -2.65717711e-02 5.37951648e-01
8.47011745e-01 2.24515185e-01 6.41364276e-01 -1.22034645e+00
-9.66696560e-01 4.64144409e-01 -1.24526091e-01 2.68736601e-01
3.19021791e-01 3.60320628e-01 4.49469417e-01 -8.41128767e-01
1.09365308e+00 9.33199763e-01 9.11038518e-01 1.13357091e+00
-1.60148680e+00 -8.54838252e-01 1.82203978e-01 -1.02158219e-01
-9.14747775e-01 -1.09550977e+00 4.54606354e-01 -5.08509696e-01
1.17757511e+00 -1.33580521e-01 3.28419209e-01 1.63789880e+00
-1.22818127e-01 6.67080045e-01 1.10146487e+00 -6.36282742e-01
5.80602407e-01 5.36046922e-01 2.61461496e-01 1.41244065e-02
4.36462194e-01 2.71272331e-01 -6.33754909e-01 -8.81187975e-01
5.71987391e-01 -1.15942396e-01 -1.27824396e-01 -6.56774282e-01
-1.13938415e+00 6.97464526e-01 1.61839291e-01 1.79328144e-01
-5.10389984e-01 -5.31120062e-01 6.47502720e-01 7.31158137e-01
9.37763870e-01 9.05583084e-01 -1.15046000e+00 -7.06547499e-03
-6.96844101e-01 2.75910109e-01 9.46145475e-01 1.08082080e+00
6.44769907e-01 -5.32383546e-02 7.81806372e-03 1.11059511e+00
1.72826536e-02 2.19500706e-01 7.07155108e-01 -8.70051146e-01
5.82748175e-01 6.56894088e-01 2.26596355e-01 -1.30114362e-01
-5.04758894e-01 -2.83971906e-01 -5.24226725e-01 2.03167304e-01
5.87231576e-01 -5.68024218e-01 -1.05902541e+00 1.90192068e+00
3.97104651e-01 -2.03744345e-03 4.25612658e-01 3.16640317e-01
9.07618880e-01 9.28247999e-03 8.15515816e-01 -5.73905110e-01
7.22928166e-01 -6.84634924e-01 -8.88930857e-01 -7.11583495e-01
9.40210223e-01 -4.18139845e-01 9.80289936e-01 3.02705854e-01
-1.09584188e+00 -4.14322019e-01 -8.53671789e-01 -1.02272399e-01
-5.57022810e-01 -5.77584863e-01 7.09582686e-01 4.94554818e-01
-9.67016220e-01 3.47096026e-01 -9.26187515e-01 -5.33081234e-01
8.15712631e-01 5.02691269e-01 -7.30260849e-01 -1.50240734e-01
-1.41268814e+00 1.22546446e+00 7.04707623e-01 -6.05057418e-01
-7.24340439e-01 -1.23373473e+00 -7.74759769e-01 -3.69818300e-01
4.87687975e-01 -7.72017896e-01 1.46380472e+00 -1.59324431e+00
-1.25676191e+00 1.05753851e+00 -5.68827391e-02 -5.85859656e-01
3.69869143e-01 -1.82478383e-01 -6.97635293e-01 -2.82136053e-01
-5.27597964e-02 7.52618372e-01 5.85524559e-01 -1.05377698e+00
-3.78266811e-01 -2.93848813e-01 -1.26143977e-01 5.10444880e-01
-4.72321540e-01 -5.92472777e-03 -2.59116311e-02 -5.11595368e-01
-2.10844800e-01 -7.24473536e-01 -4.58838373e-01 2.48209909e-01
-1.22511489e-02 -2.23623350e-01 5.52398086e-01 -6.38611138e-01
9.60635543e-01 -2.14119005e+00 -1.59976035e-01 -1.65687199e-03
3.60540807e-01 4.27718490e-01 -2.43824705e-01 2.11698100e-01
-5.51032662e-01 1.27548739e-01 -5.64089239e-01 -3.09228748e-01
-3.09834868e-01 4.62285548e-01 -2.41097853e-01 1.93102300e-01
4.16746348e-01 9.76512551e-01 -1.05461085e+00 -5.68768322e-01
-2.24356666e-01 7.56965503e-02 -6.94921076e-01 2.43287563e-01
-3.11238974e-01 7.03163266e-01 -3.50609571e-01 4.99427050e-01
5.70629776e-01 -3.33188266e-01 3.59037519e-01 4.49057132e-01
2.85958946e-01 5.00302553e-01 -7.28016019e-01 2.04948854e+00
-2.29539067e-01 2.02684715e-01 3.29393372e-02 -1.18476546e+00
8.51436555e-01 8.26738358e-01 7.12772965e-01 -8.87505710e-01
-5.20635068e-01 3.64134282e-01 7.51500726e-02 -5.35845816e-01
2.58810341e-01 -4.25911635e-01 -2.00058017e-02 3.14369142e-01
4.71617490e-01 1.59964293e-01 -1.37561828e-01 1.90053895e-01
1.31190324e+00 2.38953635e-01 9.37423587e-01 -1.73163995e-01
2.05641203e-02 5.68372130e-01 8.82801414e-01 9.10613298e-01
-3.54562461e-01 4.21096832e-01 2.29034111e-01 -3.13557029e-01
-1.11388803e+00 -1.23618710e+00 -4.22072738e-01 1.44736803e+00
-5.37049294e-01 -1.71778560e-01 -3.15136075e-01 -1.35824370e+00
1.87073827e-01 8.59205663e-01 -7.50465155e-01 -4.50517058e-01
-4.48069572e-01 -9.23536658e-01 4.69224811e-01 5.22736847e-01
5.61705567e-02 -1.13127220e+00 -2.85207089e-02 4.78137285e-01
9.24269408e-02 -6.63831651e-01 -1.68210506e-01 6.71235681e-01
-1.33306861e+00 -1.01862895e+00 -8.10228348e-01 -6.21512055e-01
6.54821396e-01 -4.63187963e-01 1.39187348e+00 -3.56747448e-01
-1.40172066e-02 4.97863501e-01 -4.26998287e-01 -8.71990979e-01
-8.16282928e-01 4.64320362e-01 6.51452839e-02 -6.46038294e-01
7.18671799e-01 -3.80791664e-01 -3.99634600e-01 1.40802145e-01
-7.82536924e-01 -3.43376994e-01 7.31869340e-01 8.35685313e-01
6.45515561e-01 -6.12716734e-01 1.17161810e+00 -1.79648161e+00
7.64637232e-01 -1.01229870e+00 -1.22693531e-01 3.05434287e-01
-1.11286080e+00 -2.09765241e-01 5.35491705e-01 -7.25469410e-01
-1.28110886e+00 3.42220992e-01 -1.08899206e-01 2.70141964e-03
-4.41547751e-01 4.42157894e-01 -1.77249044e-01 1.01547584e-01
1.55502212e+00 -9.41411108e-02 2.27045506e-01 -7.36257076e-01
2.83740371e-01 7.66934216e-01 3.77185613e-01 -5.29269159e-01
6.82108939e-01 1.56415939e-01 -6.35630548e-01 -5.15175521e-01
-8.31866026e-01 -4.13900375e-01 -8.35412562e-01 5.34220636e-01
6.03456259e-01 -1.01118970e+00 8.31331462e-02 1.57527015e-01
-7.94498622e-01 -8.40265095e-01 -1.04495251e+00 5.42333603e-01
-6.65096939e-01 6.77027628e-02 -1.19113475e-01 -3.37314934e-01
-4.58039224e-01 -7.61847198e-01 5.62125921e-01 6.93508685e-02
-8.01660478e-01 -1.48801637e+00 6.79113090e-01 1.11976974e-01
6.91075921e-01 1.29509494e-01 1.10558784e+00 -1.60972869e+00
6.29667146e-03 -1.53350487e-01 2.52668917e-01 3.96691024e-01
6.02763057e-01 -5.71266472e-01 -1.04462624e+00 -3.93595397e-01
-6.12246618e-02 -6.88643336e-01 5.90559840e-01 2.35798433e-01
9.91669536e-01 -4.20500249e-01 -6.07227623e-01 3.88820440e-01
1.09488130e+00 4.01136518e-01 5.18731415e-01 3.04442108e-01
2.21344098e-01 4.49005753e-01 5.58168590e-01 1.25630021e-01
1.96923912e-01 5.97814798e-01 -3.64658892e-01 -6.89567268e-01
-1.83535367e-01 -9.08121765e-02 9.62967426e-02 4.11630243e-01
2.53182650e-01 -1.32436618e-01 -1.33590567e+00 6.78727865e-01
-1.66607881e+00 -7.13075101e-01 3.24248314e-01 2.37782884e+00
1.61334896e+00 3.51787448e-01 1.51559994e-01 -2.16311887e-01
3.48889291e-01 -3.09569329e-01 -1.21858537e+00 -3.83473009e-01
-1.77854061e-01 5.36080956e-01 5.97320139e-01 2.97132254e-01
-1.02979374e+00 6.52752101e-01 8.17652798e+00 4.97818947e-01
-8.93330812e-01 6.33109391e-01 5.75316131e-01 -2.43422940e-01
-2.07134560e-01 -3.83477332e-03 -6.08695090e-01 3.16570640e-01
1.22707915e+00 -4.88020867e-01 -5.70169883e-04 1.02251291e+00
-7.23283589e-02 1.00505218e-01 -1.51272964e+00 5.29209495e-01
-1.57010362e-01 -1.22014582e+00 -1.85883015e-01 7.37064928e-02
8.02587390e-01 2.11005747e-01 -2.05691606e-01 5.95518231e-01
8.28194380e-01 -8.85483444e-01 8.16958025e-02 6.33368313e-01
1.10148656e+00 -3.27005416e-01 6.02743804e-01 4.37311769e-01
-2.48279914e-01 -1.39454126e-01 -1.93652645e-01 1.82043970e-01
9.06533375e-02 4.65556830e-01 -1.38531017e+00 3.45049620e-01
4.35753584e-01 7.00182378e-01 -6.04610324e-01 1.07503414e+00
2.02157497e-01 8.93170059e-01 -1.85497522e-01 5.47203243e-01
-3.53878379e-01 3.96643519e-01 4.46009636e-01 1.04706419e+00
-1.95846528e-01 5.25977090e-02 2.85588831e-01 5.57317257e-01
-1.64946318e-01 3.56982723e-02 -8.32594991e-01 -2.86825091e-01
7.03992367e-01 7.80816495e-01 -1.18398063e-01 -5.56054711e-01
-5.52032590e-01 8.72679591e-01 4.01123494e-01 5.18595576e-01
-2.35568792e-01 -1.08542912e-01 6.50972843e-01 4.67633188e-01
-4.19271648e-01 1.34462295e-02 -4.72569913e-01 -1.05261683e+00
-3.92881721e-01 -1.22853398e+00 9.13432181e-01 -6.02464020e-01
-1.83594501e+00 2.38515824e-01 3.65057170e-01 -1.12609279e+00
-5.50082922e-01 -4.38749194e-01 -2.04735741e-01 1.04386437e+00
-1.32309818e+00 -8.08435082e-01 1.31246541e-02 7.64861524e-01
4.50476348e-01 -5.79283118e-01 1.32811844e+00 4.28311050e-01
-1.54106647e-01 1.08084941e+00 4.08700615e-01 2.89034277e-01
1.51200557e+00 -1.28432167e+00 5.83582222e-01 2.18192041e-01
-2.65524238e-01 1.01291668e+00 4.31041747e-01 -1.01607788e+00
-8.68477941e-01 -9.57110286e-01 8.60117555e-01 -8.61125469e-01
2.85297871e-01 -4.22789603e-01 -1.39012969e+00 1.13749135e+00
2.02135772e-01 -3.54538555e-03 1.22101486e+00 5.37730575e-01
-2.64611334e-01 -3.80353630e-02 -1.62175035e+00 2.02272639e-01
1.25118566e+00 -3.78189236e-01 -9.76235688e-01 6.24293804e-01
9.29674029e-01 -4.03458893e-01 -1.01161945e+00 4.66800839e-01
1.58155710e-01 -2.51613498e-01 1.02594161e+00 -1.31510544e+00
1.39343619e-01 4.07384068e-01 2.07375944e-01 -1.45036125e+00
-4.90592211e-01 -4.52750951e-01 -3.84644642e-02 1.19324362e+00
8.90095353e-01 -1.11038160e+00 8.90639186e-01 1.08072567e+00
-3.20479155e-01 -3.07288796e-01 -9.58521783e-01 -7.35650659e-01
6.35110438e-01 -1.06399782e-01 5.49365938e-01 1.78431237e+00
1.84261486e-01 3.26755852e-01 6.58466816e-02 -4.09341782e-01
5.80227494e-01 -6.67031288e-01 6.72481298e-01 -1.40606630e+00
-4.59822834e-01 5.37417531e-02 9.57806595e-03 -5.22170126e-01
2.76081204e-01 -9.58258629e-01 -1.43333599e-01 -1.59578323e+00
2.24870861e-01 -8.51862252e-01 -4.31250423e-01 1.17173457e+00
-2.74772704e-01 -1.64232671e-01 -3.29160899e-01 3.13141376e-01
-4.98369187e-01 1.76390871e-01 9.89830315e-01 -1.29140792e-02
-5.21667838e-01 5.50813861e-02 -1.00784528e+00 5.25199950e-01
7.39015102e-01 -8.51547360e-01 -7.64962912e-01 -3.78464490e-01
6.57901689e-02 8.10483936e-03 2.57513057e-02 -6.02674663e-01
1.15592718e-01 -4.85337347e-01 7.03917325e-01 8.73424187e-02
1.50897145e-01 -8.87511969e-01 1.75284535e-01 3.82539362e-01
-8.37619126e-01 -1.72444165e-01 5.63720465e-01 5.33685446e-01
1.22762352e-01 -1.91892639e-01 8.90944064e-01 -3.54321480e-01
-8.59835029e-01 3.06870699e-01 -2.49042884e-01 7.42677331e-01
8.35100591e-01 -4.27742273e-01 -5.95772922e-01 -4.41330373e-02
-1.36842310e+00 1.50297105e-01 4.92259413e-01 5.00130117e-01
3.04918826e-01 -1.19749248e+00 -8.13314259e-01 3.50970268e-01
4.45537508e-01 2.56068438e-01 1.23780131e-01 7.41973221e-01
1.02833077e-01 2.80795544e-01 -2.63548553e-01 -3.92550379e-01
-1.02463412e+00 6.90728307e-01 5.01292825e-01 -3.01519364e-01
-4.10021335e-01 6.81141675e-01 2.49095231e-01 -1.03297675e+00
2.19661161e-01 1.26560867e-01 -7.28185326e-02 1.06921442e-01
3.43760490e-01 1.37673602e-01 3.37273002e-01 9.34670027e-03
-4.54381853e-01 -2.20052883e-01 -5.50731421e-01 -1.64765447e-01
1.31708241e+00 1.01855628e-01 3.88796449e-01 6.66757762e-01
9.45834637e-01 -2.59532601e-01 -1.13433242e+00 -8.83045137e-01
3.27932686e-01 -2.76467949e-01 -1.96399972e-01 -1.53487349e+00
-5.31776786e-01 4.02152300e-01 8.55913639e-01 -2.50110120e-01
1.00341570e+00 7.16099935e-03 2.47011468e-01 6.47386134e-01
3.08838487e-01 -1.27722192e+00 8.85047764e-03 3.64199013e-01
6.85060799e-01 -1.45067191e+00 2.77982384e-01 -2.77849454e-02
-9.60316479e-01 4.58247989e-01 8.03511500e-01 2.08356574e-01
7.51370668e-01 4.18236643e-01 2.12301329e-01 6.63739964e-02
-1.02062654e+00 1.49114400e-01 2.42851421e-01 1.10598671e+00
6.29202962e-01 -1.58421844e-01 -2.35624373e-01 7.00235426e-01
2.00750440e-01 6.11187160e-01 1.21108025e-01 1.21032584e+00
-1.21384479e-01 -1.51405287e+00 -1.63283542e-01 7.82534063e-01
-3.33875209e-01 -1.47352666e-01 -8.08222234e-01 1.12063813e+00
2.00529993e-01 6.02189600e-01 2.44486794e-01 1.38932243e-01
5.37800014e-01 7.78517067e-01 4.14277107e-01 -1.27232552e+00
-8.19100082e-01 -1.58155888e-01 3.81618619e-01 -3.40945899e-01
-3.44084501e-01 -7.31063604e-01 -1.30934453e+00 -3.08028087e-02
-7.94825703e-02 2.75930107e-01 3.60618591e-01 9.15075839e-01
6.63378596e-01 2.86969751e-01 1.57920361e-01 -2.01554392e-02
-6.60878003e-01 -1.05038977e+00 -3.24136198e-01 7.57872403e-01
4.72018450e-01 -5.47090888e-01 -2.11180001e-01 2.19608441e-01] | [10.678529739379883, 7.971100807189941] |
30c66ee3-b1ce-46e9-bed9-ece29a903661 | improving-text-matching-in-e-commerce-search | 2307.00370 | null | https://arxiv.org/abs/2307.00370v1 | https://arxiv.org/pdf/2307.00370v1.pdf | Improving Text Matching in E-Commerce Search with A Rationalizable, Intervenable and Fast Entity-Based Relevance Model | Discovering the intended items of user queries from a massive repository of items is one of the main goals of an e-commerce search system. Relevance prediction is essential to the search system since it helps improve performance. When online serving a relevance model, the model is required to perform fast and accurate inference. Currently, the widely used models such as Bi-encoder and Cross-encoder have their limitations in accuracy or inference speed respectively. In this work, we propose a novel model called the Entity-Based Relevance Model (EBRM). We identify the entities contained in an item and decompose the QI (query-item) relevance problem into multiple QE (query-entity) relevance problems; we then aggregate their results to form the QI prediction using a soft logic formulation. The decomposition allows us to use a Cross-encoder QE relevance module for high accuracy as well as cache QE predictions for fast online inference. Utilizing soft logic makes the prediction procedure interpretable and intervenable. We also show that pretraining the QE module with auto-generated QE data from user logs can further improve the overall performance. The proposed method is evaluated on labeled data from e-commerce websites. Empirical results show that it achieves promising improvements with computation efficiency. | ['Kewei Tu', 'Fei Huang', 'Pengjun Xie', 'Zhongqiang Huang', 'Tao Wang', 'Haihong Tang', 'Rong Xiao', 'Jianhui Ji', 'Ke Yu', 'Chenyue Jiang', 'Yue Zhang', 'Yong Jiang', 'Jiong Cai'] | 2023-07-01 | null | null | null | null | ['text-matching'] | ['natural-language-processing'] | [-1.41991228e-01 -4.49763574e-02 -6.09670937e-01 -5.17023027e-01
-1.02584517e+00 -4.43459302e-01 1.98873922e-01 2.70194620e-01
-1.62644163e-01 5.21692693e-01 -4.88683730e-02 -3.34892392e-01
-3.88085902e-01 -1.02847278e+00 -9.64287221e-01 7.19205886e-02
7.77932405e-02 8.26023519e-01 4.99737620e-01 -4.04702157e-01
3.91604066e-01 -1.12806655e-01 -1.62192988e+00 6.59608960e-01
1.17710638e+00 1.64899838e+00 4.63315487e-01 4.33123678e-01
-2.49433458e-01 9.07897651e-01 -2.55236417e-01 -8.00385833e-01
3.56959224e-01 -4.22945507e-02 -1.15039861e+00 -3.98531139e-01
-7.97882825e-02 -5.01849830e-01 1.68582767e-01 1.01298594e+00
1.72782645e-01 7.93215185e-02 4.32062060e-01 -8.75888288e-01
-8.39755893e-01 8.13868284e-01 -1.32596225e-01 3.15360397e-01
5.63366354e-01 -3.83525044e-01 1.75869298e+00 -1.11204290e+00
5.33219755e-01 1.09387267e+00 3.70811641e-01 -1.26733817e-02
-8.51076424e-01 -2.73678064e-01 1.56919658e-01 8.17067623e-01
-1.40419316e+00 -3.02307427e-01 6.29316628e-01 3.67332920e-02
1.32081592e+00 4.63554144e-01 4.47277784e-01 5.39239168e-01
3.66113454e-01 1.21815252e+00 6.22840405e-01 -5.42969882e-01
4.12094921e-01 6.54021084e-01 4.11871523e-01 5.85204959e-01
3.82277817e-01 -1.28035128e-01 -5.51782072e-01 -3.21529776e-01
4.45288599e-01 5.36119118e-02 8.70532766e-02 -3.40202227e-02
-5.88307977e-01 9.44914758e-01 4.53628629e-01 1.11339912e-01
-7.89207280e-01 -1.69496626e-01 2.27375537e-01 5.08556187e-01
3.88391465e-01 4.43257689e-01 -8.15020919e-01 1.33123949e-01
-6.81322575e-01 1.93747893e-01 1.19347739e+00 1.27266467e+00
8.13525140e-01 -5.75229466e-01 -2.75624335e-01 7.56731272e-01
6.91565335e-01 2.74877697e-01 5.74295819e-01 -5.72082400e-01
4.87715483e-01 8.10687721e-01 2.37501577e-01 -1.02657461e+00
-4.58084419e-02 -7.27103293e-01 -4.88118976e-01 -5.76507807e-01
-1.33279145e-01 2.30488136e-01 -3.61316681e-01 1.35174108e+00
3.83933634e-01 -1.44720167e-01 -3.06714252e-02 9.42710280e-01
6.63106561e-01 5.81744969e-01 3.13401014e-01 -1.84725642e-01
1.64627349e+00 -9.66503203e-01 -7.50513494e-01 -1.95894435e-01
5.74742079e-01 -7.78208733e-01 9.98891115e-01 5.29977679e-01
-1.24093986e+00 -6.27765477e-01 -9.11300540e-01 -3.07573229e-01
-3.32817078e-01 1.24742739e-01 8.73097897e-01 3.33474994e-01
-7.13597476e-01 4.07883167e-01 -5.30489683e-01 7.52222985e-02
1.01220431e-02 7.97905743e-01 1.66520223e-01 9.03276727e-02
-1.75588441e+00 8.31971109e-01 4.67028737e-01 1.25811519e-02
-2.87037075e-01 -4.05264318e-01 -6.03891671e-01 5.87137043e-01
6.56456590e-01 -8.77830744e-01 1.53538835e+00 -8.81716251e-01
-1.28916228e+00 3.12765658e-01 -3.09661508e-01 -3.48973691e-01
-1.14733272e-03 -3.92263800e-01 -7.10746408e-01 1.13800712e-01
1.11969084e-01 2.55641907e-01 6.02953732e-01 -1.01589799e+00
-1.21005094e+00 -4.49074835e-01 3.39485049e-01 2.39701301e-01
-3.53762954e-01 1.59033924e-01 -9.25160766e-01 -3.13880920e-01
1.90291896e-01 -5.40885150e-01 4.16105539e-02 -4.05079514e-01
-1.96148977e-01 -7.81799793e-01 2.58355856e-01 -9.53643799e-01
1.73595405e+00 -1.60999775e+00 -1.95973143e-01 4.84012216e-01
1.07298434e-01 8.37798044e-02 1.65129840e-01 3.11804175e-01
3.19527596e-01 1.60268009e-01 5.11373460e-01 2.02988219e-02
3.24627548e-01 1.33063018e-01 -3.32974404e-01 -3.02045643e-01
2.98557401e-01 1.39529538e+00 -6.76595390e-01 -7.16296613e-01
-2.31415033e-01 1.10502310e-01 -8.91642153e-01 3.68980408e-01
-5.54602265e-01 -2.01272115e-01 -8.93456876e-01 7.84239829e-01
3.61566752e-01 -9.19440448e-01 3.22305918e-01 -6.41877234e-01
4.19744909e-01 7.41787493e-01 -1.29152596e+00 1.14873838e+00
-6.24803245e-01 -2.21954420e-01 -5.27903764e-03 -8.15213323e-01
7.62269497e-01 2.43927076e-01 2.74664462e-01 -1.22811329e+00
-9.41398665e-02 4.18049276e-01 -2.65026271e-01 -6.75477028e-01
9.75840092e-01 1.10537345e-02 -9.94208157e-02 5.09733081e-01
3.31876501e-02 5.63373148e-01 2.48337641e-01 5.07109314e-02
8.43046844e-01 1.69354215e-01 4.58371460e-01 -1.69528395e-01
7.85556734e-01 1.91975385e-01 5.41169167e-01 6.04693353e-01
1.64136738e-01 -1.64414451e-01 1.51500732e-01 -2.86434293e-01
-8.21064949e-01 -7.13962197e-01 -2.05705330e-01 1.54345453e+00
5.22196889e-01 -5.29928446e-01 -4.44170147e-01 -8.42233777e-01
-5.87441400e-03 7.72044063e-01 -1.64386734e-01 -2.62821853e-01
-5.79374731e-01 -5.81296682e-01 -1.95478782e-01 6.38278544e-01
4.36829537e-01 -8.76248777e-01 -3.18310946e-01 4.95910645e-01
-6.82651341e-01 -8.05105507e-01 -3.49895567e-01 3.47068042e-01
-8.49067807e-01 -7.66242087e-01 -5.60018271e-02 -8.18806767e-01
4.30656821e-01 1.67421877e-01 1.48630261e+00 -3.27705359e-03
5.65916374e-02 2.08543509e-01 -5.73565722e-01 -1.46140277e-01
-2.37882197e-01 3.28344882e-01 -2.22886149e-02 -1.76987886e-01
9.83992517e-01 -7.62723833e-02 -7.93655574e-01 5.42282403e-01
-8.63518476e-01 -1.46548718e-01 9.93468285e-01 9.41409290e-01
1.00888252e+00 2.38283247e-01 9.88045096e-01 -8.85168612e-01
9.03628528e-01 -9.07783628e-01 -7.17489660e-01 9.27075326e-01
-1.31446660e+00 5.48467338e-01 7.85722613e-01 -2.14006647e-01
-1.16459060e+00 -2.76051790e-01 -3.26242656e-01 -6.64681345e-02
3.61725211e-01 8.86909127e-01 -1.32681474e-01 2.18186706e-01
4.32369292e-01 2.89731324e-01 -4.79966968e-01 -6.81933165e-01
2.51754194e-01 9.72095430e-01 1.83231667e-01 -6.64930940e-01
2.67157614e-01 -3.11032444e-01 -2.38070190e-01 -1.15547545e-01
-9.52003479e-01 -9.28685427e-01 -2.63048619e-01 1.51153654e-01
3.32260013e-01 -9.22465980e-01 -9.53288436e-01 -4.37828809e-01
-8.37742150e-01 3.39199513e-01 -1.48302481e-01 3.97200316e-01
-3.45434964e-01 3.99063736e-01 -8.21167290e-01 -8.23251784e-01
-9.42471325e-01 -9.72417593e-01 1.09898961e+00 1.53434068e-01
-7.50340000e-02 -1.03938353e+00 -8.18227082e-02 5.46754777e-01
4.98734355e-01 -8.46321583e-01 1.14837873e+00 -1.04794359e+00
-9.60000634e-01 -4.24814761e-01 -1.78569466e-01 3.22501659e-01
-2.21074447e-02 -5.98645806e-01 -7.27472246e-01 6.51675239e-02
1.20301999e-01 -2.52773851e-01 5.05381465e-01 -6.94329888e-02
1.15794849e+00 -7.72776127e-01 -2.37128943e-01 1.72246754e-01
1.63518798e+00 7.00387955e-02 4.31527406e-01 3.21449429e-01
8.12923685e-02 4.96362209e-01 1.20533180e+00 6.10677779e-01
8.33020270e-01 9.25854206e-01 2.98640907e-01 3.50523025e-01
2.96540320e-01 -4.55971450e-01 3.67321938e-01 1.05875409e+00
2.98092682e-02 -2.88150012e-01 -4.18901384e-01 3.79902065e-01
-1.98357940e+00 -9.50254798e-01 -8.37348327e-02 2.25751925e+00
1.17969215e+00 1.36136040e-01 -8.39233771e-02 -3.83211225e-02
4.26107824e-01 -6.44001901e-01 -5.72349370e-01 -3.92560989e-01
3.04107726e-01 4.23816383e-01 2.89441615e-01 4.53389347e-01
-9.56160784e-01 7.06605971e-01 5.96319866e+00 9.51475799e-01
-7.47580230e-01 1.51716009e-01 6.85424566e-01 1.81472197e-01
-6.26206875e-01 -1.00208737e-01 -1.44861913e+00 4.37400073e-01
1.09665513e+00 -3.51680696e-01 4.77372289e-01 1.27424717e+00
-1.94833249e-01 -6.92615733e-02 -1.38094878e+00 8.80173743e-01
6.70537800e-02 -1.03914452e+00 6.39693141e-02 3.33893523e-02
4.20774937e-01 -2.16260254e-01 -4.12729606e-02 8.26149583e-01
1.93407074e-01 -6.03625119e-01 5.68256080e-01 7.26244926e-01
5.32715142e-01 -6.15497768e-01 1.02092600e+00 6.26899540e-01
-1.34568191e+00 -3.20543140e-01 -7.08638728e-01 2.10891023e-01
1.25721067e-01 6.32993877e-01 -1.05441034e+00 8.67266715e-01
7.08876789e-01 4.25611109e-01 -6.40650153e-01 8.04665089e-01
-2.46872716e-02 4.24372822e-01 -4.54317987e-01 -3.04117590e-01
-2.04616785e-01 -2.30871484e-01 5.62770143e-02 9.02474582e-01
3.26003402e-01 8.18838105e-02 4.42593135e-02 1.04342806e+00
-1.16486497e-01 3.82162452e-01 -5.23539186e-02 7.06925020e-02
4.13075507e-01 1.39825201e+00 -2.17655957e-01 -6.77648306e-01
-5.29816389e-01 1.00139546e+00 4.63500232e-01 8.44586343e-02
-7.71431267e-01 -2.36820966e-01 2.70703256e-01 7.67973736e-02
6.08561277e-01 2.70536691e-01 -1.80057272e-01 -1.40008998e+00
4.06769186e-01 -8.29964936e-01 5.81320107e-01 -4.47174251e-01
-1.43512332e+00 5.27937710e-01 6.70376394e-05 -1.06560922e+00
-8.14739048e-01 -5.34415901e-01 -7.92063996e-02 1.13047731e+00
-1.87363148e+00 -9.65372324e-01 -1.04314066e-01 4.43820179e-01
5.00402331e-01 5.69358990e-02 8.66797149e-01 5.10059655e-01
-4.59068179e-01 8.56305838e-01 1.86614804e-02 -2.58975983e-01
5.30905306e-01 -1.16356575e+00 1.19411282e-01 5.20440698e-01
2.26814114e-02 1.08732080e+00 2.87667066e-01 -8.90896022e-01
-1.93261778e+00 -1.00597751e+00 1.49822879e+00 -4.96491164e-01
3.80540639e-01 -2.95470823e-02 -8.32425594e-01 6.88864291e-01
-3.87758389e-02 -2.01069564e-01 9.93018985e-01 3.69840086e-01
-1.41131163e-01 -3.83282334e-01 -1.10186696e+00 2.49905944e-01
7.81793535e-01 -5.29984891e-01 -9.16453123e-01 3.27826053e-01
7.61840343e-01 7.43993884e-03 -1.23555040e+00 3.52631271e-01
6.85083151e-01 -7.55694032e-01 1.06348848e+00 -5.89664817e-01
3.69600207e-01 -1.04936004e-01 -3.19058090e-01 -7.41985142e-01
-6.36277735e-01 -1.48585945e-01 -9.05362368e-01 9.40159380e-01
5.70599318e-01 -5.26820183e-01 6.89507127e-01 1.18956232e+00
1.20188355e-01 -9.43860769e-01 -4.93884921e-01 -6.20740354e-01
-5.60618341e-01 -3.69684011e-01 8.11650872e-01 4.15429294e-01
3.86558354e-01 7.83747375e-01 -9.01251361e-02 1.18794799e-01
5.94691277e-01 5.41674972e-01 3.51563990e-01 -1.33843744e+00
-9.42796767e-01 -2.78846145e-01 8.75914618e-02 -1.72009552e+00
-1.96383119e-01 -9.92277920e-01 -9.81957316e-02 -1.50740230e+00
4.26897883e-01 -7.47694790e-01 -4.88810241e-01 4.55877274e-01
-4.57072318e-01 -8.32725912e-02 -2.13233456e-01 4.75643158e-01
-1.21063387e+00 4.20224249e-01 1.04487967e+00 2.29187906e-02
-2.80636638e-01 4.06347662e-01 -1.02293885e+00 1.78053036e-01
6.55843616e-01 -2.89722413e-01 -5.29334247e-01 -3.76188844e-01
1.02549922e+00 3.27032745e-01 5.68634011e-02 -4.88033742e-01
5.75926542e-01 5.92599921e-02 4.35240060e-01 -7.80776858e-01
1.64700806e-01 -1.11443436e+00 4.42377776e-02 3.40653718e-01
-5.60370982e-01 5.90221062e-02 -1.64828807e-01 5.73761463e-01
-5.21594286e-01 -6.13265216e-01 1.22930728e-01 -1.22692183e-01
-8.64772737e-01 2.72217661e-01 1.28943279e-01 -7.89819285e-02
7.37621605e-01 -6.12586131e-03 -7.02542812e-02 -3.41469944e-01
-4.59931672e-01 3.76662582e-01 -1.18587680e-01 4.04663771e-01
6.11761272e-01 -1.38026750e+00 -4.22241271e-01 4.72311407e-01
2.49030873e-01 -1.11470036e-01 5.01515567e-02 7.66040742e-01
-1.05872251e-01 1.08885992e+00 5.67463003e-02 -3.76892596e-01
-9.75543857e-01 7.43177354e-01 -4.03829589e-02 -8.55210364e-01
-2.92976312e-02 7.35249400e-01 1.17291451e-01 -2.85470426e-01
1.79511756e-01 -3.70080084e-01 -6.15140200e-01 -1.57262757e-01
6.80463970e-01 2.50558048e-01 4.42699313e-01 -2.31652573e-01
-2.67765790e-01 2.75341034e-01 -3.99201810e-01 1.50483012e-01
1.19546616e+00 -4.58957642e-01 -2.04961315e-01 4.20825416e-03
1.09814262e+00 -4.33967449e-02 -5.58784008e-01 -7.20165312e-01
3.98438990e-01 -3.55047256e-01 2.43059590e-01 -1.08674169e+00
-6.57140672e-01 5.62031090e-01 3.99789065e-01 3.06134105e-01
1.40181541e+00 1.75696626e-01 1.02959085e+00 7.82373548e-01
7.87781954e-01 -1.33144057e+00 -1.16470359e-01 5.39161980e-01
6.77451611e-01 -1.40907609e+00 -1.32286653e-01 -5.57960570e-01
-5.92647791e-01 1.04584622e+00 5.49472749e-01 1.85061365e-01
7.68229306e-01 -9.18298662e-02 -5.09012759e-01 -1.97585776e-01
-1.25037265e+00 -2.60213047e-01 8.19301486e-01 5.72143234e-02
4.41222966e-01 4.24052402e-02 -5.31325817e-01 1.44779158e+00
-6.51695654e-02 3.77935261e-01 -2.02802002e-01 8.34673107e-01
-7.39939213e-01 -1.43141818e+00 -4.55279805e-04 7.89627969e-01
-7.17458010e-01 -3.94569278e-01 -1.56219965e-02 2.62358636e-01
-5.70288189e-02 1.00665593e+00 -1.05725363e-01 -6.75785482e-01
2.64692724e-01 4.02011126e-01 2.85217881e-01 -6.70274973e-01
-9.23610091e-01 1.61504462e-01 3.42503518e-01 -6.26925528e-01
-8.44534487e-02 -2.68359780e-01 -1.19523048e+00 -2.46691376e-01
-7.96787798e-01 4.64437217e-01 7.27487147e-01 7.77222574e-01
8.52876782e-01 2.50337631e-01 8.40567172e-01 9.78098065e-02
-1.35192752e+00 -1.02050424e+00 -4.80340779e-01 4.00282562e-01
-2.37701192e-01 -4.10310835e-01 -1.07806034e-01 -2.63386182e-02] | [10.961050033569336, 7.316137790679932] |
ccb602de-0428-4ccf-82ed-1a557b1323e3 | learning-to-infer-from-unlabeled-data-a-semi | 2211.02971 | null | https://arxiv.org/abs/2211.02971v1 | https://arxiv.org/pdf/2211.02971v1.pdf | Learning to Infer from Unlabeled Data: A Semi-supervised Learning Approach for Robust Natural Language Inference | Natural Language Inference (NLI) or Recognizing Textual Entailment (RTE) aims at predicting the relation between a pair of sentences (premise and hypothesis) as entailment, contradiction or semantic independence. Although deep learning models have shown promising performance for NLI in recent years, they rely on large scale expensive human-annotated datasets. Semi-supervised learning (SSL) is a popular technique for reducing the reliance on human annotation by leveraging unlabeled data for training. However, despite its substantial success on single sentence classification tasks where the challenge in making use of unlabeled data is to assign "good enough" pseudo-labels, for NLI tasks, the nature of unlabeled data is more complex: one of the sentences in the pair (usually the hypothesis) along with the class label are missing from the data and require human annotations, which makes SSL for NLI more challenging. In this paper, we propose a novel way to incorporate unlabeled data in SSL for NLI where we use a conditional language model, BART to generate the hypotheses for the unlabeled sentences (used as premises). Our experiments show that our SSL framework successfully exploits unlabeled data and substantially improves the performance of four NLI datasets in low-resource settings. We release our code at: https://github.com/msadat3/SSL_for_NLI. | ['Cornelia Caragea', 'Mobashir Sadat'] | 2022-11-05 | null | null | null | null | ['sentence-classification'] | ['natural-language-processing'] | [ 3.31263244e-01 2.89985418e-01 -3.82849276e-01 -9.06214595e-01
-8.05307508e-01 -7.28687048e-01 7.99565077e-01 1.61330283e-01
-3.51066887e-01 1.05694771e+00 2.51191467e-01 -7.80787468e-01
2.50900835e-01 -4.89243388e-01 -9.16996241e-01 -3.13600421e-01
3.99523675e-01 6.74682856e-01 -2.14752495e-01 -5.97666390e-02
-7.32904952e-03 4.73901816e-02 -1.27507436e+00 5.22166312e-01
9.20948803e-01 9.93939102e-01 4.19458887e-03 4.43651795e-01
-3.17468494e-01 1.21035361e+00 -5.48912585e-01 -5.25544286e-01
1.56434089e-01 -6.50033057e-01 -1.22032452e+00 3.76127958e-02
2.43856370e-01 -3.43868166e-01 -1.56126618e-01 8.89715493e-01
1.73730031e-01 -9.37789977e-02 8.27697217e-01 -1.41614211e+00
-4.39639568e-01 8.65052462e-01 -3.90823483e-01 -4.11300212e-02
5.26023626e-01 2.42709108e-02 1.36317980e+00 -1.12396884e+00
5.74758351e-01 1.10179627e+00 5.43769658e-01 4.49419886e-01
-1.07036304e+00 -5.39466858e-01 4.66279015e-02 4.06430781e-01
-1.37704015e+00 -5.85998952e-01 7.21315742e-01 -2.34310746e-01
1.17141283e+00 2.73068428e-01 1.62025541e-01 1.33598924e+00
-2.16733068e-02 1.22538733e+00 1.09691072e+00 -7.23058581e-01
2.58534610e-01 2.01266989e-01 3.82126600e-01 4.63911533e-01
5.15702330e-02 -1.77703097e-01 -3.64207059e-01 4.36800569e-02
1.48840785e-01 8.09535105e-03 -1.67960763e-01 1.96796954e-01
-1.26525366e+00 8.60177994e-01 2.17572525e-01 4.73520577e-01
-1.93499371e-01 -3.52932453e-01 3.39653224e-01 3.61452907e-01
5.26821136e-01 3.48920673e-01 -7.40544140e-01 6.11013658e-02
-1.03728080e+00 1.31453961e-01 1.00973284e+00 1.00894928e+00
8.68295968e-01 -3.28415573e-01 -1.74662694e-01 6.94881439e-01
2.81523168e-01 3.47888678e-01 5.53896606e-01 -6.94930732e-01
7.45462000e-01 9.13268030e-01 2.12900847e-01 -6.04462564e-01
-3.69416326e-01 -1.52211383e-01 -9.52318013e-01 -2.89911509e-01
5.99593997e-01 -2.46427000e-01 -7.74035037e-01 1.71886921e+00
2.08919242e-01 1.24607533e-01 4.36759740e-01 7.80332685e-01
1.06625247e+00 8.25866938e-01 -7.73562640e-02 -9.73034576e-02
1.14032471e+00 -8.67072165e-01 -6.92210615e-01 -6.00555837e-01
1.01802468e+00 -8.04278314e-01 1.10793126e+00 1.82595357e-01
-8.33733380e-01 -3.90966624e-01 -9.50987577e-01 -3.49684179e-01
-2.82352418e-01 3.64400029e-01 4.66360360e-01 3.92183252e-02
-7.30545461e-01 2.58257598e-01 -4.44948077e-01 -6.63484186e-02
3.53215992e-01 2.94020444e-01 -4.58633214e-01 -4.49730605e-01
-1.59989393e+00 9.69682932e-01 6.53645158e-01 3.05736154e-01
-6.75551295e-01 -4.62546229e-01 -1.17200065e+00 1.97549149e-01
7.61390686e-01 -3.51785153e-01 1.29695249e+00 -1.06063914e+00
-1.29276645e+00 1.02399206e+00 -5.04818678e-01 -4.62682307e-01
4.92089957e-01 -2.40796469e-02 -1.41997188e-01 -1.20273605e-01
2.71060079e-01 6.99631512e-01 7.05042362e-01 -1.07896018e+00
-5.38749099e-01 -3.49577695e-01 3.47433805e-01 1.08406201e-01
-1.22578055e-01 4.58589531e-02 -1.59052953e-01 -2.54181802e-01
3.17412503e-02 -1.00121009e+00 3.45480046e-03 -3.58706445e-01
-8.80491376e-01 -8.50394607e-01 7.61325121e-01 -5.80000281e-01
9.73805726e-01 -2.23195767e+00 -1.49925008e-01 -1.22184515e-01
1.56643122e-01 5.23728728e-01 -4.08823639e-02 3.99279565e-01
-1.89957589e-01 1.47672310e-01 -3.91803324e-01 -4.68689859e-01
6.90853000e-02 5.82164109e-01 -2.64458448e-01 1.51939645e-01
5.02821922e-01 1.04019439e+00 -9.37812924e-01 -6.90813124e-01
9.57355574e-02 7.74348751e-02 -3.82878840e-01 4.46297556e-01
-5.10375738e-01 7.14199901e-01 -7.29687884e-02 4.22550976e-01
5.64074814e-01 -6.06814623e-01 4.53264713e-01 -1.98249876e-01
1.18819147e-01 7.26423323e-01 -1.10909188e+00 1.45660961e+00
-6.21250331e-01 6.21802688e-01 -1.60676047e-01 -1.39970970e+00
7.91402698e-01 6.28725350e-01 -3.19146812e-02 -4.30480927e-01
2.36777738e-01 3.28500837e-01 3.27925608e-02 -7.64751732e-01
5.30410558e-02 -4.33149308e-01 -2.52153039e-01 6.37106419e-01
1.84582278e-01 -4.66349907e-02 5.21759748e-01 3.86296391e-01
8.23017478e-01 1.34554163e-01 5.35577953e-01 -4.19232473e-02
7.36718774e-01 4.76774983e-02 7.46011257e-01 7.56886542e-01
-3.81211080e-02 4.69636559e-01 6.18371487e-01 -4.01417851e-01
-9.85108078e-01 -7.85672247e-01 -5.40711470e-02 8.24340820e-01
-7.67087936e-02 -2.15771094e-01 -3.62621307e-01 -1.00299585e+00
-2.52015620e-01 1.11241055e+00 -3.32518429e-01 2.73958128e-02
-5.11110425e-01 -5.26268780e-01 4.31360722e-01 5.97457409e-01
4.58447546e-01 -1.19222140e+00 -8.33797827e-02 5.90813570e-02
-8.89957249e-01 -1.39778185e+00 -1.18587367e-01 3.60671997e-01
-4.78568345e-01 -1.08332670e+00 -2.65080273e-01 -8.75105023e-01
8.30190361e-01 1.90523807e-02 1.25275218e+00 1.00602150e-01
-1.06177352e-01 -1.08555891e-01 -4.41070378e-01 -5.36551118e-01
-7.11329877e-01 -2.17168871e-02 3.97840589e-02 5.63613251e-02
5.56395054e-01 -2.87802130e-01 -3.57768655e-01 1.65894225e-01
-1.09947503e+00 3.50048006e-01 4.34248388e-01 9.96573508e-01
4.53538418e-01 6.87534437e-02 6.84807241e-01 -1.16626799e+00
4.51251894e-01 -5.19484520e-01 -3.67215365e-01 5.05064368e-01
-3.60186666e-01 1.74550354e-01 9.04059529e-01 -2.56987095e-01
-1.27494717e+00 -1.41968518e-01 -3.50215048e-01 -1.51073903e-01
-4.98248398e-01 1.01215267e+00 -2.86496580e-01 6.14677548e-01
4.74450856e-01 5.79949245e-02 4.53950278e-02 -3.31529647e-01
2.26312011e-01 9.58901703e-01 3.82290810e-01 -5.24434626e-01
4.61086750e-01 1.66086063e-01 -1.45323291e-01 -6.04101539e-01
-1.73417854e+00 -3.39112669e-01 -8.73548388e-01 2.09547505e-02
6.45395994e-01 -8.62752020e-01 -5.30139148e-01 1.40107125e-01
-1.26582277e+00 -4.48163718e-01 -2.10429788e-01 6.15058839e-01
-2.10047096e-01 6.04528189e-01 -5.82601786e-01 -7.62253404e-01
-4.20477509e-01 -1.11335540e+00 7.62919426e-01 2.49859877e-02
-5.71061254e-01 -1.04177141e+00 -1.20894939e-01 8.00799847e-01
7.27055669e-02 -1.72719620e-02 1.07260048e+00 -1.20863032e+00
-2.25829571e-01 -6.14076912e-01 -2.84525394e-01 6.89629436e-01
2.03183547e-01 6.65387698e-03 -9.09291744e-01 -5.48953153e-02
2.21196353e-01 -9.89506602e-01 5.22308052e-01 1.71063066e-01
9.73850787e-01 -5.82365036e-01 -6.12371117e-02 -2.56732740e-02
1.16374028e+00 -1.29563287e-01 3.57297540e-01 -5.32271266e-02
6.56966448e-01 7.93508053e-01 7.45354533e-01 4.14092511e-01
6.03639483e-01 3.98245364e-01 1.25775889e-01 -1.35706559e-01
-2.30295360e-02 -2.04311997e-01 3.03418905e-01 7.41201520e-01
2.84300357e-01 -5.20308495e-01 -1.07467723e+00 5.55040419e-01
-1.99859858e+00 -8.62022579e-01 -2.33227849e-01 2.06691003e+00
1.30051005e+00 2.83756256e-01 -3.49335849e-01 4.29094285e-01
6.36051118e-01 -1.80640921e-01 -5.39624512e-01 -1.64783090e-01
-2.20443234e-01 5.10269366e-02 -1.17287114e-01 7.93669224e-01
-1.08610797e+00 9.41173673e-01 5.17114496e+00 7.90681243e-01
-1.07277763e+00 1.12549409e-01 7.72697628e-01 2.14304752e-03
-2.84036368e-01 1.52818605e-01 -9.57673132e-01 4.05928165e-01
1.00323868e+00 1.77239358e-01 9.94121507e-02 6.14426613e-01
3.25937659e-01 -2.52541065e-01 -1.52336228e+00 6.90515816e-01
5.14969081e-02 -1.31913733e+00 -3.50429155e-02 -3.06208223e-01
7.27536261e-01 1.60823241e-01 -2.35729873e-01 5.21995842e-01
1.93760961e-01 -9.67840135e-01 6.86074078e-01 9.46643949e-02
7.68711805e-01 -5.32186925e-01 1.15003669e+00 8.32937002e-01
-7.01655209e-01 1.33295402e-01 -2.13034689e-01 -5.04297316e-01
5.12337983e-02 8.10271919e-01 -1.14272857e+00 4.72746760e-01
3.53610843e-01 6.66214347e-01 -4.73743945e-01 4.55825210e-01
-9.47809815e-01 9.49895799e-01 -3.99463296e-01 -1.58338308e-01
4.59144413e-01 -8.04330185e-02 1.50330096e-01 1.13022256e+00
-1.36915162e-01 1.78719778e-03 3.40734243e-01 9.24670815e-01
-3.12910110e-01 4.63144630e-02 -7.23620772e-01 -1.80145562e-01
4.37419385e-01 1.17079675e+00 -6.30479515e-01 -6.53933644e-01
-6.80719972e-01 9.26223338e-01 5.99444449e-01 2.98853874e-01
-8.34248722e-01 -1.92130566e-01 9.15467665e-02 -9.42586064e-02
1.72049314e-01 -6.35363460e-02 -3.72035146e-01 -1.42721057e+00
2.39450380e-01 -8.20218146e-01 4.66615558e-01 -7.54373670e-01
-1.43134201e+00 4.47481960e-01 4.62647453e-02 -1.13341784e+00
-6.60698831e-01 -6.19480073e-01 -6.75814569e-01 6.91673934e-01
-1.68947065e+00 -1.21572316e+00 3.24412435e-03 5.87632895e-01
6.72040820e-01 9.43982452e-02 8.50337446e-01 2.64148444e-01
-7.38598883e-01 5.76202154e-01 -5.37519827e-02 4.19677973e-01
8.28123510e-01 -1.05633342e+00 1.97046444e-01 8.35362315e-01
4.28674728e-01 6.78007126e-01 6.55401409e-01 -5.17685533e-01
-1.11477828e+00 -9.31518972e-01 1.75250483e+00 -3.20475519e-01
7.22338319e-01 -4.55388486e-01 -8.92866433e-01 9.94584739e-01
2.82505751e-01 1.03014251e-02 1.08253658e+00 2.77934223e-01
-3.16378176e-01 1.96153387e-01 -9.83811915e-01 5.66953778e-01
6.86206758e-01 -5.25267482e-01 -7.68828988e-01 6.67627335e-01
6.33836031e-01 -2.09761634e-01 -6.54809773e-01 4.69920874e-01
1.18608326e-01 -7.40491748e-01 5.73263049e-01 -6.99840724e-01
8.82169604e-01 -3.11087251e-01 -6.99471459e-02 -9.94114995e-01
9.43622664e-02 -2.08468080e-01 5.38785309e-02 1.27771568e+00
9.53872085e-01 -4.67053562e-01 6.66994631e-01 9.95165586e-01
3.41664478e-02 -7.88561583e-01 -8.67673814e-01 -6.01471066e-01
-7.57576749e-02 -5.57002962e-01 2.95776933e-01 1.09604645e+00
2.75827706e-01 1.04839313e+00 -4.10487741e-01 -5.15600368e-02
3.05292845e-01 6.13658488e-01 6.25567913e-01 -1.19236016e+00
-1.79158881e-01 -8.98145512e-02 -3.15212943e-02 -9.21897292e-01
6.84535265e-01 -1.26718235e+00 2.66106367e-01 -1.62471354e+00
3.56627405e-01 -3.11895877e-01 8.45290497e-02 9.53294218e-01
-4.18728471e-01 2.41387725e-01 -2.65824441e-02 3.04555684e-01
-8.16622794e-01 5.27000070e-01 9.69213545e-01 -3.39497447e-01
7.09305853e-02 1.22609973e-01 -5.48302591e-01 9.64192808e-01
9.18581188e-01 -3.58415812e-01 -4.54096377e-01 -5.36404788e-01
3.75604540e-01 8.92059654e-02 1.88884482e-01 -6.16446197e-01
1.81141004e-01 1.13940323e-02 2.28597268e-01 -6.82022572e-01
1.01918057e-01 -7.24192321e-01 -3.20684969e-01 2.61973053e-01
-6.37077630e-01 -3.75326425e-01 -4.30364572e-02 3.71050805e-01
-4.52703238e-01 -7.77242720e-01 5.02864778e-01 -2.52236128e-01
-6.10501766e-01 2.12619510e-02 -4.75579679e-01 4.34656054e-01
7.41378009e-01 2.16677114e-01 -7.01323897e-02 -5.15291452e-01
-7.72010565e-01 3.84592980e-01 8.87041986e-02 1.28641859e-01
6.74978673e-01 -9.65453744e-01 -9.36570048e-01 1.19089246e-01
2.44322032e-01 4.27925438e-01 8.14959407e-02 8.50940704e-01
-1.70874968e-01 6.09372020e-01 2.39182293e-01 -4.64741200e-01
-1.10601342e+00 5.40553153e-01 -6.82474002e-02 -6.18683219e-01
-4.86062557e-01 9.23107028e-01 1.30989060e-01 -8.08781743e-01
2.92078137e-01 -2.84798503e-01 -7.12908506e-02 -6.94248900e-02
3.42487156e-01 -1.66855752e-01 1.34518355e-01 -4.76241022e-01
-3.59387398e-01 -1.42856494e-01 -3.45826447e-01 4.11347076e-02
1.26881492e+00 -1.92909852e-01 -3.36275727e-01 6.93819821e-01
1.29294634e+00 -2.90565848e-01 -8.35264564e-01 -6.62441075e-01
2.20564455e-01 -6.78338334e-02 -1.04225978e-01 -8.28329086e-01
-5.18871069e-01 9.11272287e-01 -2.69435644e-01 4.86684404e-02
6.75377965e-01 2.13894382e-01 8.76613259e-01 7.50304818e-01
1.27422318e-01 -9.50709105e-01 1.42525556e-02 8.00765336e-01
8.17039371e-01 -1.87960780e+00 -6.11051023e-02 -4.10478324e-01
-8.43343079e-01 8.69244516e-01 6.47477806e-01 2.25452542e-01
5.50044596e-01 3.98968309e-01 1.35392845e-01 -1.11470841e-01
-7.09983885e-01 -4.55597118e-02 2.99829524e-02 1.51915193e-01
8.65550280e-01 5.60663417e-02 -3.94645602e-01 6.47163749e-01
-1.60755768e-01 -1.37940757e-02 4.22391534e-01 1.06434357e+00
-3.14835296e-03 -1.13442624e+00 -1.27989367e-01 5.17496049e-01
-4.66941476e-01 -3.38893265e-01 -5.30960321e-01 5.11373937e-01
5.30140214e-02 1.27484608e+00 -3.60347852e-02 -6.75005838e-02
-9.93948802e-02 4.62943435e-01 3.97333652e-01 -7.79724598e-01
-3.63108009e-01 -1.94441736e-01 4.14205104e-01 -1.48336828e-01
-5.82211196e-01 -4.23399478e-01 -1.37889743e+00 -1.95810318e-01
-3.31835657e-01 2.80640274e-01 4.75203425e-01 1.66317356e+00
1.94409311e-01 1.66664541e-01 6.74336135e-01 -6.28967702e-01
-6.85901105e-01 -1.16875291e+00 -2.91967839e-01 6.85133636e-01
1.77383363e-01 -3.52887243e-01 -4.67144221e-01 2.14393213e-01] | [10.229240417480469, 8.596946716308594] |
1f0aaa27-c201-450d-97c3-e9aac31ef393 | facial-makeup-transfer-combining-illumination | 1907.03398 | null | https://arxiv.org/abs/1907.03398v1 | https://arxiv.org/pdf/1907.03398v1.pdf | Facial Makeup Transfer Combining Illumination Transfer | To meet the women appearance needs, we present a novel virtual experience approach of facial makeup transfer, developed into windows platform application software. The makeup effects could present on the user's input image in real time, with an only single reference image. The input image and reference image are divided into three layers by facial feature points landmarked: facial structure layer, facial color layer, and facial detail layer. Except for the above layers are processed by different algorithms to generate output image, we also add illumination transfer, so that the illumination effect of the reference image is automatically transferred to the input image. Our approach has the following three advantages: (1) Black or dark and white facial makeup could be effectively transferred by introducing illumination transfer; (2) Efficiently transfer facial makeup within seconds compared to those methods based on deep learning frameworks; (3) Reference images with the air-bangs could transfer makeup perfectly. | ['Xiao-Dong Li', 'Xin Jin', 'Xiaokun Zhang', 'Rui Han', 'Ning Ning'] | 2019-07-08 | null | null | null | null | ['facial-makeup-transfer'] | ['computer-vision'] | [ 2.02689007e-01 2.86072046e-01 2.35324308e-01 -2.07658976e-01
-1.19766043e-02 -4.62354779e-01 4.07347172e-01 -5.93680739e-01
-1.92608945e-02 3.73318106e-01 -9.14637744e-02 -5.29748797e-02
5.06251574e-01 -9.86788929e-01 -7.40249336e-01 -4.55621392e-01
2.83458352e-01 -2.49274269e-01 2.28361249e-01 -2.96991915e-01
1.14717402e-01 8.19525599e-01 -2.11884785e+00 4.06288445e-01
7.28394449e-01 1.06941342e+00 8.40440542e-02 5.97088218e-01
-4.65745091e-01 4.70937431e-01 -5.80558002e-01 -5.76081276e-01
3.96226943e-01 -3.83303583e-01 -5.50571978e-01 3.81293029e-01
6.65960312e-01 -7.36258924e-01 4.96649295e-02 1.09615445e+00
6.88099742e-01 -6.89864904e-02 6.48516655e-01 -1.49887955e+00
-1.04648352e+00 -2.37585887e-01 -9.25248981e-01 -5.75730741e-01
6.14803314e-01 7.93413445e-02 7.79630765e-02 -1.11163855e+00
8.63655210e-01 1.48769617e+00 4.45779413e-01 1.03400731e+00
-1.06242704e+00 -1.25291073e+00 1.55921698e-01 -4.74718064e-02
-1.34212947e+00 -5.22371352e-01 8.42448354e-01 -3.35380942e-01
3.99247020e-01 3.95234168e-01 9.46268320e-01 8.63885164e-01
2.25775063e-01 5.42173982e-01 1.25961149e+00 -5.47015607e-01
-1.25520518e-02 5.04564285e-01 -4.35113966e-01 1.02823615e+00
-2.93694437e-01 2.95954704e-01 -1.40214562e-01 1.84863448e-01
1.27350664e+00 5.33999689e-02 -2.94535905e-01 -3.11600804e-01
-5.47083139e-01 1.76714361e-01 4.54990745e-01 4.66709174e-02
-6.32754192e-02 4.03822772e-02 8.90441760e-02 3.61929923e-01
5.78694701e-01 -2.22125337e-01 -3.48748922e-01 3.31376374e-01
-7.81421602e-01 -2.15065498e-02 4.04349416e-01 1.03989577e+00
1.15189159e+00 1.56121969e-01 9.12421569e-03 8.56078625e-01
4.77428526e-01 9.39517260e-01 1.91957161e-01 -8.44019830e-01
2.40012720e-01 5.65311968e-01 3.26253057e-01 -1.00845897e+00
-3.15506786e-01 4.20686930e-01 -6.49275899e-01 1.08454955e+00
2.65282601e-01 -5.19931912e-01 -1.10112929e+00 1.32545686e+00
8.82294357e-01 1.61919430e-01 -1.75169885e-01 8.06811810e-01
1.51832795e+00 6.20513082e-01 1.38573293e-02 -1.90595031e-01
1.36677742e+00 -9.60702360e-01 -1.09475136e+00 1.82549372e-01
2.06201196e-01 -1.14159715e+00 1.22950888e+00 3.94671202e-01
-1.27910018e+00 -9.45200861e-01 -1.03165686e+00 -2.08985165e-01
-3.72763962e-01 3.59075010e-01 6.03955150e-01 1.07605302e+00
-1.41607082e+00 4.36065674e-01 -3.31498593e-01 -2.61892349e-01
3.03664207e-01 4.57694322e-01 -7.27224886e-01 2.29099378e-01
-8.72046351e-01 7.47636020e-01 6.81126863e-02 3.34945112e-01
-7.14800894e-01 -8.44034910e-01 -9.60144460e-01 -8.36849436e-02
-9.52964202e-02 -7.10335314e-01 1.02173173e+00 -2.05879283e+00
-2.31675005e+00 1.30463493e+00 -3.16306293e-01 6.81160331e-01
6.17888927e-01 -1.63137257e-01 -5.19489586e-01 1.85755298e-01
-3.48094344e-01 7.22712040e-01 1.05720687e+00 -1.71736670e+00
-3.50957215e-01 -4.70022440e-01 7.45205814e-03 1.43516138e-01
-2.03314796e-01 3.90400171e-01 -6.80432618e-01 -3.54541361e-01
-1.41800106e-01 -7.32043266e-01 2.14504793e-01 7.14970887e-01
-1.56462803e-01 1.32692844e-01 1.13660014e+00 -7.96240389e-01
9.86115456e-01 -2.21820235e+00 -2.72770673e-01 2.98690498e-01
2.53916811e-02 6.74976707e-01 -4.56832647e-01 3.27113509e-01
-4.53832954e-01 7.88950920e-02 3.17406863e-01 -2.97439456e-01
-3.02715778e-01 -9.99618396e-02 -3.39139812e-02 3.46939504e-01
1.11769132e-01 9.16377127e-01 -7.23615944e-01 -6.88500226e-01
4.47765708e-01 9.36502576e-01 -4.11308914e-01 2.50450760e-01
2.03460619e-01 4.29216117e-01 -1.71290144e-01 7.87388384e-01
1.49384487e+00 4.46626067e-01 -5.09673636e-03 -2.51575798e-01
-9.41759795e-02 -6.29197836e-01 -1.19003463e+00 1.35250008e+00
-6.51991665e-01 3.94331336e-01 6.31457567e-01 -6.36980832e-02
1.14851773e+00 4.69837785e-01 2.31373712e-01 -9.04939950e-01
2.46546745e-01 -9.56861600e-02 -3.96323949e-01 -8.43967676e-01
3.08084279e-01 -3.73013079e-01 4.88580793e-01 6.02147877e-01
-5.80085926e-02 -2.64996737e-01 -4.27951097e-01 -2.20334888e-01
1.42951980e-01 8.41990769e-01 -6.71744645e-02 1.69625059e-02
5.46698034e-01 -6.67760909e-01 3.42348725e-01 -1.41739935e-01
-1.19404256e-01 8.10011446e-01 3.99394125e-01 -5.30685842e-01
-7.95783818e-01 -1.12922013e+00 -7.79602751e-02 1.28453779e+00
4.17609036e-01 -4.14196663e-02 -1.12069893e+00 -5.61296582e-01
-1.43785298e-01 1.98304370e-01 -9.17416990e-01 1.61023408e-01
-3.10394913e-01 -1.60548091e-01 2.22680584e-01 4.49165314e-01
6.05543613e-01 -1.42720592e+00 -4.53790337e-01 -2.95484751e-01
5.89693338e-02 -7.17997015e-01 -6.60197139e-01 -5.34949780e-01
-6.61756277e-01 -1.04512560e+00 -1.07053924e+00 -7.53396749e-01
1.23642492e+00 3.57255995e-01 7.82962263e-01 4.14527267e-01
-6.20137155e-01 4.53000486e-01 -2.57220924e-01 -6.69867158e-01
-1.99756637e-01 -7.49074817e-01 -1.22257628e-01 4.85395640e-01
-4.13566865e-02 -5.54644108e-01 -9.48790550e-01 2.28332758e-01
-8.72936726e-01 3.13846767e-01 2.85882026e-01 4.63102728e-01
2.36284062e-01 -3.40261817e-01 2.64183611e-01 -7.87961602e-01
3.61880124e-01 1.80063862e-02 -4.84016985e-01 3.78050059e-01
1.70660336e-02 -6.31589115e-01 4.21134740e-01 -3.85662347e-01
-1.62619901e+00 4.90202680e-02 -1.70371920e-01 -3.57368022e-01
-2.62651622e-01 -3.49505246e-01 -5.67669272e-01 -2.93482453e-01
5.34830213e-01 5.42948656e-02 4.33663726e-01 -1.96257651e-01
6.44800603e-01 7.91553199e-01 2.79207498e-01 -4.55479801e-01
7.90318847e-01 6.89909637e-01 -1.46849021e-01 -7.26930022e-01
-1.70479566e-01 4.59404439e-02 -1.17077279e+00 -7.93897688e-01
8.20875227e-01 -7.87545502e-01 -9.30461287e-01 7.28645861e-01
-1.24388266e+00 -4.22971189e-01 6.03130907e-02 6.30954355e-02
-3.78235936e-01 3.14307094e-01 -5.51886678e-01 -8.96709561e-01
-4.54838842e-01 -8.96443665e-01 1.17915404e+00 8.16275597e-01
6.07532971e-02 -1.07107532e+00 -1.74213141e-01 1.50032207e-01
3.34033728e-01 4.34823722e-01 8.25980604e-01 5.44407785e-01
-3.05544019e-01 -1.11006528e-01 -5.37407756e-01 3.93579870e-01
6.25054240e-01 7.13613093e-01 -1.43983579e+00 -7.06644356e-02
-4.94488806e-01 -1.09584115e-01 2.67599851e-01 2.33183116e-01
1.04226494e+00 -9.06803533e-02 -3.74275088e-01 7.57938385e-01
1.35141730e+00 4.63225424e-01 1.18085480e+00 6.78317100e-02
7.22564638e-01 8.97209048e-01 6.89907849e-01 1.96495831e-01
1.69236273e-01 6.21235192e-01 3.97896171e-01 -1.13964617e+00
-5.78023791e-01 -4.55656856e-01 5.17725587e-01 3.95240486e-01
-5.82804203e-01 2.80840039e-01 -3.46800894e-01 1.33962974e-01
-1.49267626e+00 -1.00250566e+00 -2.48534784e-01 2.27463078e+00
6.66072845e-01 -6.96012199e-01 -1.15574673e-02 -1.21097282e-01
7.08284080e-01 -2.72794843e-01 -2.35998243e-01 -9.30089295e-01
1.58244535e-01 4.62605923e-01 -2.43104789e-02 3.64538133e-01
-8.63844454e-01 1.02852023e+00 6.88427973e+00 5.78112066e-01
-1.74855602e+00 1.18562482e-01 7.27652669e-01 -2.01004203e-02
-5.04219651e-01 -2.92834163e-01 -5.10180235e-01 4.16079938e-01
1.76946834e-01 3.77569385e-02 2.54241526e-01 6.53250277e-01
3.77866954e-01 -1.06576510e-01 -8.92172396e-01 1.16662848e+00
1.99369565e-01 -9.21184659e-01 2.02064306e-01 3.06717083e-02
9.38639939e-01 -8.23579371e-01 3.70417804e-01 1.12017497e-01
-1.86689608e-02 -1.17397106e+00 6.61011934e-01 5.36748648e-01
1.50648654e+00 -8.11782777e-01 4.10122871e-01 -3.40143330e-02
-1.13999379e+00 2.24750355e-01 -4.49607402e-01 -6.62795603e-02
-1.37765869e-01 1.57145694e-01 -6.80156291e-01 6.52218521e-01
7.46751726e-01 3.11440587e-01 -4.66081828e-01 7.26740718e-01
-4.09768492e-01 -1.76845081e-02 -2.98189092e-02 1.53537631e-01
-1.61076769e-01 -4.74914700e-01 -1.57310218e-01 1.12605560e+00
4.13783848e-01 1.30645052e-01 -3.06100786e-01 7.94644773e-01
3.75025347e-02 7.03836262e-01 -7.11077452e-01 6.03359520e-01
1.48814172e-01 1.64748538e+00 -7.37186313e-01 -1.53420329e-01
-4.18272614e-01 1.49104548e+00 -6.60318658e-02 6.01628482e-01
-8.63149524e-01 -5.45591414e-01 4.70032066e-01 4.17733818e-01
1.67861596e-01 3.66217047e-01 -8.63276497e-02 -8.40099871e-01
1.59958359e-02 -5.56083381e-01 -1.25559866e-01 -1.33282125e+00
-8.44227910e-01 6.73218846e-01 -3.53256643e-01 -1.32333720e+00
2.73112893e-01 -6.96654141e-01 -1.06056559e+00 1.10224748e+00
-1.50545859e+00 -1.78070807e+00 -6.90570056e-01 8.24289083e-01
2.74201989e-01 1.58147104e-02 1.05598557e+00 3.98331910e-01
-4.13495719e-01 8.06001246e-01 -2.23507270e-01 7.72189274e-02
1.01130569e+00 -8.65967453e-01 2.12086458e-02 4.75509882e-01
-2.73077250e-01 5.30168772e-01 2.15170220e-01 -5.13710022e-01
-1.15889442e+00 -9.99479234e-01 5.56083918e-01 -4.42335933e-01
7.04727769e-02 -3.73469234e-01 -6.13406420e-01 6.94017351e-01
5.04057467e-01 2.29948476e-01 8.06199610e-01 4.54159081e-02
-4.04902160e-01 -4.27365810e-01 -1.51618326e+00 8.43140304e-01
8.99609089e-01 -2.82415986e-01 -1.55669615e-01 1.83356386e-02
1.13962837e-01 -5.83564222e-01 -6.38884664e-01 2.18331978e-01
1.18150616e+00 -1.18238068e+00 9.25271928e-01 -2.64611065e-01
6.91734433e-01 -5.26048601e-01 4.21782523e-01 -1.15233457e+00
-2.22100586e-01 -7.80684710e-01 4.15057749e-01 1.52950263e+00
3.53187472e-01 -6.30899549e-01 4.67744827e-01 9.53570545e-01
1.24574833e-01 -6.83660090e-01 -6.25160694e-01 -2.01388136e-01
5.71669675e-02 1.43601937e-04 9.70995247e-01 8.34261179e-01
-4.24523167e-02 -1.21247932e-01 -4.96773392e-01 8.20595622e-02
2.63879538e-01 1.42804503e-01 1.22957265e+00 -9.79988396e-01
5.49081974e-02 -3.72702301e-01 -1.85516909e-01 -5.94074607e-01
9.39348117e-02 -6.32106304e-01 -2.90291727e-01 -1.54057634e+00
2.02959195e-01 -4.67352957e-01 -6.03555236e-03 6.55440211e-01
-2.75671631e-01 6.85534418e-01 3.95023167e-01 -1.79178938e-01
3.03215478e-02 4.07131821e-01 1.88403416e+00 1.69582829e-01
-2.96728939e-01 -4.87276614e-02 -6.45352066e-01 9.83315349e-01
4.36549306e-01 1.51477143e-01 -5.73580503e-01 -4.19497490e-01
-3.38858142e-02 8.60951245e-02 3.25872540e-01 -6.04241848e-01
-1.17666237e-01 -3.23820442e-01 9.80049014e-01 -2.97665149e-01
2.83875555e-01 -9.19958949e-01 4.22012240e-01 2.76655614e-01
2.26102859e-01 -2.42942438e-01 4.98820901e-01 -1.21206408e-02
2.93767080e-02 -1.09973766e-01 9.55820680e-01 -1.59741700e-01
-7.33141661e-01 2.98901618e-01 -2.91159034e-01 -7.35293210e-01
1.51302207e+00 -7.92845190e-01 -2.51253963e-01 -4.40587670e-01
-8.69037509e-01 -1.48809254e-01 7.17405260e-01 6.83893383e-01
8.95823896e-01 -1.47687161e+00 -5.76381564e-01 7.39355266e-01
-7.60566592e-02 -3.30047965e-01 7.86542833e-01 6.12441957e-01
-9.65131044e-01 -1.49048850e-01 -6.56883180e-01 -3.85124236e-01
-1.81712103e+00 7.00265825e-01 5.30848861e-01 4.27272409e-01
-7.42675841e-01 8.64706993e-01 9.68991995e-01 -6.46968424e-01
4.99362908e-02 -1.06653154e-01 -3.82501870e-01 4.17549871e-02
1.02550817e+00 2.63995260e-01 -1.01259813e-01 -7.93613672e-01
-1.40651435e-01 1.28463352e+00 2.41026521e-01 1.13290949e-02
9.46998656e-01 -3.15827131e-01 -2.70539224e-01 1.15226589e-01
1.12920022e+00 4.34382975e-01 -1.38762999e+00 1.65402904e-01
-9.63366807e-01 -7.57836401e-01 -1.99585781e-01 -8.80076766e-01
-1.32572091e+00 1.04304886e+00 9.49234962e-01 -1.89520389e-01
1.52859843e+00 -2.95481980e-01 5.94668269e-01 -4.73939002e-01
5.01598597e-01 -1.09531558e+00 1.45292163e-01 9.63699743e-02
1.04580808e+00 -8.87947619e-01 -9.19309407e-02 -8.06560934e-01
-6.54252708e-01 1.37542021e+00 8.69666755e-01 5.30857369e-02
7.30558515e-01 4.52199042e-01 6.53706372e-01 -1.28800273e-01
-4.91026133e-01 -1.15172178e-01 3.75946581e-01 1.11204028e+00
4.49822992e-01 -5.18369414e-02 2.36177985e-02 4.33270216e-01
-1.73224449e-01 2.84398615e-01 3.34775656e-01 4.82049048e-01
-2.84496903e-01 -1.00875163e+00 -6.84064090e-01 -2.41702482e-01
-3.93048823e-01 1.10656433e-01 -3.27600896e-01 8.07800770e-01
6.80392206e-01 7.98827291e-01 1.83972239e-01 -2.40826920e-01
4.38142091e-01 -1.96229786e-01 8.43904734e-01 -5.53974271e-01
-6.37280524e-01 1.31402075e-01 -2.67847568e-01 -7.07254171e-01
-3.60960245e-01 -1.06714666e-01 -1.22053957e+00 -3.51497680e-01
-1.34459585e-01 -2.94416636e-01 7.15907812e-01 5.07266223e-01
3.57503384e-01 3.88729900e-01 9.51694727e-01 -1.35947037e+00
2.86186039e-01 -8.86856496e-01 -7.37019241e-01 5.25553465e-01
2.28848502e-01 -6.07192338e-01 -1.46310687e-01 5.17072856e-01] | [12.802413940429688, -0.13666968047618866] |
f4a2a5cf-dcbb-4a97-a61a-ebc9ab2c285b | flexible-portrait-image-editing-with-fine | 2204.01318 | null | https://arxiv.org/abs/2204.01318v1 | https://arxiv.org/pdf/2204.01318v1.pdf | Flexible Portrait Image Editing with Fine-Grained Control | We develop a new method for portrait image editing, which supports fine-grained editing of geometries, colors, lights and shadows using a single neural network model. We adopt a novel asymmetric conditional GAN architecture: the generators take the transformed conditional inputs, such as edge maps, color palette, sliders and masks, that can be directly edited by the user; the discriminators take the conditional inputs in the way that can guide controllable image generation more effectively. Taking color editing as an example, we feed color palettes (which can be edited easily) into the generator, and color maps (which contain positional information of colors) into the discriminator. We also design a region-weighted discriminator so that higher weights are assigned to more important regions, like eyes and skin. Using a color palette, the user can directly specify the desired colors of hair, skin, eyes, lip and background. Color sliders allow the user to blend colors in an intuitive manner. The user can also edit lights and shadows by modifying the corresponding masks. We demonstrate the effectiveness of our method by evaluating it on the CelebAMask-HQ dataset with a wide range of tasks, including geometry/color/shadow/light editing, hand-drawn sketch to image translation, and color transfer. We also present ablation studies to justify our design. | ['Ying He', 'Fei Hou', 'QiAn Fu', 'Linlin Liu'] | 2022-04-04 | null | null | null | null | ['sketch-to-image-translation'] | ['computer-vision'] | [ 5.82611859e-01 -9.45117697e-02 1.25325367e-01 -3.90866727e-01
-2.12867796e-01 -9.85668838e-01 5.51690936e-01 -4.89413768e-01
-1.29014552e-01 5.93397796e-01 -2.78562456e-02 -2.35622630e-01
3.93033206e-01 -1.09332466e+00 -7.37829149e-01 -5.79126239e-01
3.86705637e-01 1.47562027e-01 1.87188119e-01 -3.21512580e-01
1.79349288e-01 5.85658252e-01 -1.36810565e+00 2.92598069e-01
1.06189382e+00 8.88498127e-01 1.33499682e-01 8.41010332e-01
-1.73853368e-01 3.10765773e-01 -8.47405791e-01 -3.84560764e-01
3.84260446e-01 -5.41138709e-01 -1.51176751e-01 3.02311610e-02
4.98681575e-01 -5.66593707e-01 4.60455529e-02 8.26909542e-01
3.32052171e-01 -1.17640927e-01 7.74320960e-01 -1.43109310e+00
-1.36378276e+00 4.91365552e-01 -1.01466274e+00 -5.74945867e-01
3.36413205e-01 5.10900021e-01 8.54223669e-01 -7.65975296e-01
6.67629898e-01 1.24894965e+00 3.70163590e-01 7.67118454e-01
-1.43637216e+00 -1.03043187e+00 3.41489792e-01 -2.21598327e-01
-1.31284297e+00 -1.51096806e-01 9.27329123e-01 -2.68067747e-01
1.68691188e-01 8.52844715e-01 8.88321161e-01 1.09494114e+00
1.62652954e-02 5.73294759e-01 1.18021882e+00 -5.26325464e-01
1.46691263e-01 2.31779709e-01 -6.47597730e-01 8.05501223e-01
-1.60302073e-01 6.42844066e-02 -3.53155911e-01 -1.07617944e-03
1.34799576e+00 4.99414206e-02 -3.37948263e-01 -2.98114866e-01
-1.19035423e+00 6.74051523e-01 5.63859701e-01 -8.86974186e-02
-9.81206894e-02 3.81729811e-01 -2.00730890e-01 1.13393784e-01
3.04430574e-01 5.92464268e-01 -2.72696495e-01 2.17040673e-01
-6.51462555e-01 1.13274261e-01 4.08983767e-01 1.24357820e+00
7.90595293e-01 -6.54795319e-02 -8.96812439e-01 9.73308325e-01
2.17050046e-01 8.32453251e-01 -1.42333344e-01 -9.33933914e-01
3.59288096e-01 6.41583979e-01 3.19917262e-01 -7.85553515e-01
-5.24452366e-02 1.75034866e-01 -8.49933505e-01 7.30542898e-01
3.05351019e-01 -4.68176961e-01 -1.45362985e+00 1.93885517e+00
3.35668832e-01 -8.59549269e-02 -4.64492351e-01 8.85857105e-01
6.09278440e-01 7.15518832e-01 3.61234248e-02 2.38038346e-01
1.52787018e+00 -9.88934457e-01 -8.04234684e-01 -1.70818239e-01
-4.96638333e-03 -8.79566669e-01 1.72387421e+00 1.69277355e-01
-9.92314279e-01 -4.90248173e-01 -1.02970397e+00 -4.03521806e-01
-4.52838570e-01 6.27014160e-01 5.31258762e-01 5.61305225e-01
-1.13509190e+00 4.29387778e-01 -5.54167628e-01 -1.50073782e-01
3.78202826e-01 7.14853629e-02 2.58765928e-02 1.40961960e-01
-1.09374249e+00 6.77528441e-01 -1.11364096e-01 2.11464301e-01
-6.89456761e-01 -7.14201868e-01 -8.03669155e-01 4.09697294e-02
2.16648832e-01 -7.58521080e-01 1.01828575e+00 -1.26467419e+00
-2.16622353e+00 8.22126269e-01 9.61502567e-02 4.30815816e-01
7.42893934e-01 9.17794406e-02 -3.29229861e-01 -7.52276108e-02
-1.26046300e-01 1.02819479e+00 1.22013319e+00 -1.53916550e+00
-5.19176602e-01 -2.57112980e-02 3.68015468e-01 1.30133271e-01
-2.92742491e-01 -1.07957676e-01 -8.64949286e-01 -1.04420221e+00
-2.88799793e-01 -1.06710052e+00 -1.05868027e-01 6.95927143e-01
-9.85343039e-01 2.67022967e-01 1.06398082e+00 -6.26384318e-01
1.33223450e+00 -2.18530941e+00 2.64626324e-01 5.44410229e-01
1.34673595e-01 -4.18305174e-02 -3.25522691e-01 1.81143641e-01
-1.41144335e-01 2.40676850e-01 -4.18279856e-01 -2.81136751e-01
9.99297425e-02 1.26976714e-01 -2.84320951e-01 8.58886540e-03
4.54306901e-01 1.09136355e+00 -6.71037912e-01 -2.56503463e-01
2.23151669e-01 6.89012229e-01 -5.64383745e-01 2.61020929e-01
-4.15656209e-01 3.75881195e-01 -5.66539407e-01 5.54250658e-01
8.93066943e-01 -4.22314070e-02 7.64819011e-02 -4.62557197e-01
-2.03467593e-01 -1.51138470e-01 -1.24214482e+00 1.41016722e+00
-7.78779924e-01 5.56767583e-01 2.47012258e-01 -5.49463555e-02
9.53362048e-01 -1.13681301e-01 -1.38620034e-01 -5.53979218e-01
-1.02962535e-02 -3.06734502e-01 -2.69654363e-01 -1.21334270e-01
5.67468643e-01 6.97537288e-02 -2.95989662e-01 7.61806369e-01
-6.00717783e-01 -5.40287018e-01 -9.40204114e-02 1.85054749e-01
5.42355061e-01 3.29883844e-01 -7.23897815e-02 -1.25174478e-01
1.60312325e-01 -4.23335910e-01 2.50668854e-01 3.97180527e-01
4.88815576e-01 9.70011115e-01 7.95354366e-01 -2.42434949e-01
-1.19556630e+00 -1.20217967e+00 1.58757389e-01 1.45234168e+00
1.53535873e-01 -3.26522708e-01 -1.01808441e+00 -4.80841875e-01
-3.32533047e-02 8.37737143e-01 -1.03416348e+00 -1.60286337e-01
-4.74687338e-01 -3.06817979e-01 2.53434241e-01 6.08024657e-01
4.38720882e-01 -1.29197383e+00 -6.84133232e-01 -1.76169366e-01
2.39403367e-01 -6.70994580e-01 -1.34177625e+00 -9.27906409e-02
-4.87639487e-01 -8.50583911e-01 -1.02340424e+00 -8.37004185e-01
1.25466144e+00 -4.54908572e-02 9.31266248e-01 2.23670870e-01
-4.46753114e-01 1.74778298e-01 -2.48162493e-01 -3.41418117e-01
-3.04519594e-01 -8.78580734e-02 -4.02560115e-01 2.03435719e-01
-6.37788296e-01 -5.90383232e-01 -9.96462882e-01 6.14864349e-01
-1.19920814e+00 7.32267439e-01 3.67324054e-01 6.19558752e-01
5.88639498e-01 -2.48098493e-01 1.50325552e-01 -1.25736368e+00
8.27800751e-01 1.31857276e-01 -7.67087460e-01 5.24477839e-01
-1.75606027e-01 1.48223788e-01 8.76623988e-01 -7.02384710e-01
-1.19535422e+00 2.00828519e-02 1.79925729e-02 -4.03795868e-01
1.65141970e-02 -4.39436659e-02 -4.89590526e-01 -1.95340544e-01
5.30566752e-01 4.61143814e-03 -1.93045646e-01 -2.26929277e-01
1.07685483e+00 5.64842939e-01 5.52685678e-01 -7.92042136e-01
1.11620069e+00 4.47822839e-01 -2.36895755e-01 -4.10992742e-01
-2.21096933e-01 6.58017516e-01 -5.04679620e-01 -1.18378669e-01
8.09067428e-01 -4.40181583e-01 -7.35976517e-01 6.38376772e-01
-1.05801785e+00 -8.17772388e-01 -3.95567387e-01 -2.89509118e-01
-2.23693669e-01 -1.63838327e-01 -6.57307923e-01 -4.01484758e-01
-3.13141823e-01 -1.16774750e+00 1.29822052e+00 6.59019232e-01
-2.03781992e-01 -7.50048101e-01 -4.13213730e-01 -2.78515399e-01
6.20609581e-01 7.28056192e-01 1.23117304e+00 3.63471568e-01
-6.61855698e-01 -1.41800288e-02 -2.60881156e-01 2.09603086e-01
5.06897271e-01 7.23334312e-01 -7.81545281e-01 -1.65300786e-01
-8.73268247e-01 -7.42826937e-03 5.70306599e-01 2.56878674e-01
1.65201139e+00 -3.93859237e-01 -4.59803969e-01 9.67663169e-01
1.03547156e+00 4.05746669e-01 8.91390741e-01 -5.95290102e-02
1.03588557e+00 2.40296587e-01 1.51413172e-01 3.12821031e-01
2.42551193e-01 7.41254091e-01 2.01965392e-01 -7.57354259e-01
-3.20695311e-01 -5.96190393e-01 1.43180624e-01 2.49592692e-01
-2.00148374e-01 -2.87725508e-01 -3.07892174e-01 -3.38879563e-02
-1.33526719e+00 -6.83570445e-01 2.57202089e-01 2.08332086e+00
1.16486692e+00 4.48964536e-02 3.98204960e-02 -3.16457510e-01
9.37934637e-01 1.41816303e-01 -7.33554184e-01 -4.18887585e-01
7.17845559e-03 5.43284297e-01 4.31421459e-01 6.20019138e-01
-8.05466473e-01 1.12135863e+00 6.22135019e+00 8.02323937e-01
-1.52115536e+00 -2.41944402e-01 9.70751822e-01 -3.07294726e-01
-1.10741854e+00 -7.95961469e-02 -4.03731763e-01 6.39731586e-01
-7.50023723e-02 2.05500409e-01 1.05697620e+00 4.57735360e-01
2.40668207e-01 -5.56673780e-02 -1.10373414e+00 9.00254905e-01
-1.91781387e-01 -1.40089262e+00 3.64191234e-01 -1.65901527e-01
8.92016590e-01 -7.65570939e-01 4.69965637e-01 -2.33822912e-02
5.22162735e-01 -1.06085587e+00 1.13383281e+00 6.38281643e-01
1.92817378e+00 -6.40479147e-01 -2.20192581e-01 -1.67036191e-01
-1.17052865e+00 1.75063893e-01 6.05948307e-02 1.40591279e-01
1.48192748e-01 2.89968461e-01 -7.48532116e-01 1.34952784e-01
4.34242934e-01 3.18691134e-01 -6.05537176e-01 6.76860154e-01
-8.91848087e-01 2.74905980e-01 -4.20224816e-01 -1.56861797e-01
-1.66164979e-01 -5.10977209e-01 8.69736299e-02 1.14707065e+00
4.07867104e-01 1.52471676e-01 -1.60032868e-01 1.53470349e+00
-3.58906597e-01 -2.01975808e-01 -3.08388025e-01 7.34033138e-02
8.53528619e-01 1.59995854e+00 -8.45604062e-01 -2.10324928e-01
-4.39116471e-02 1.45812321e+00 1.16090044e-01 7.87112951e-01
-1.18258119e+00 -9.38700855e-01 6.99216366e-01 2.60532111e-01
3.32255602e-01 -3.87795269e-02 -5.17964661e-01 -7.63547003e-01
4.24476676e-02 -8.23744833e-01 2.62599550e-02 -1.37326407e+00
-1.10723186e+00 6.35044217e-01 -5.27218804e-02 -9.67896879e-01
1.62982345e-01 -6.46746755e-01 -1.09115553e+00 1.17808092e+00
-1.11345196e+00 -1.37255681e+00 -5.66196203e-01 6.84801877e-01
6.33526444e-02 2.37780511e-01 7.42850065e-01 2.47007251e-01
-5.30005038e-01 9.28295553e-01 -1.57660440e-01 3.17593753e-01
8.24185073e-01 -1.19089127e+00 7.74733245e-01 6.21206045e-01
2.96918359e-02 6.84547186e-01 4.33635950e-01 -5.43000817e-01
-1.30773187e+00 -1.02888489e+00 1.18556976e-01 -3.12440634e-01
1.26436651e-01 -8.95628333e-01 -5.16644359e-01 7.02278495e-01
3.40441108e-01 -1.12576008e-01 3.29196692e-01 -1.49072990e-01
-4.03933764e-01 -3.81887764e-01 -1.13451064e+00 1.12406874e+00
1.07303882e+00 -6.16678953e-01 3.37396301e-02 1.91957980e-01
6.85770929e-01 -7.26206064e-01 -3.72707129e-01 7.94819966e-02
8.19199085e-01 -7.10123420e-01 9.07949388e-01 -2.85553128e-01
5.02876461e-01 -6.91534698e-01 2.99300551e-01 -1.77743995e+00
-4.31742221e-01 -8.27090681e-01 4.34884012e-01 1.34015548e+00
7.48817921e-01 -4.82585460e-01 5.65098047e-01 7.61847973e-01
-1.02610841e-01 -6.63453281e-01 -4.89873201e-01 -4.35369276e-02
-1.71105981e-01 -2.40169168e-01 1.17535734e+00 8.31121624e-01
-2.38825738e-01 7.77692497e-02 -4.91483569e-01 -6.15274608e-02
2.61612952e-01 3.85010988e-01 7.05075383e-01 -8.08904648e-01
-2.32358709e-01 -5.91152191e-01 4.39247131e-01 -1.04843593e+00
-3.72375995e-01 -5.62799215e-01 -5.11068553e-02 -1.49915850e+00
8.17577243e-02 -8.54060590e-01 1.31690145e-01 9.63786066e-01
-3.48166794e-01 5.35084128e-01 5.94364405e-01 -2.55737543e-01
-7.47405887e-02 5.12605369e-01 2.07492256e+00 -3.23720574e-01
-3.38861734e-01 -2.18540996e-01 -1.14621794e+00 6.45244002e-01
6.34598792e-01 -1.31551012e-01 -4.43220705e-01 -7.55126476e-01
2.82257915e-01 -5.30265309e-02 4.68949169e-01 -5.62755227e-01
-3.38884555e-02 -4.15924579e-01 9.20742691e-01 -7.77767971e-02
2.32226878e-01 -7.13366747e-01 5.15091717e-01 2.41561830e-01
-5.02161920e-01 1.34527236e-01 2.45649055e-01 1.92531303e-01
3.29880983e-01 3.21770936e-01 8.52293491e-01 -7.72193000e-02
-3.26020747e-01 5.30627012e-01 -6.50284067e-02 -1.61012352e-01
1.12205374e+00 -3.44325095e-01 -3.12835902e-01 -7.06211448e-01
-6.08907938e-01 2.26177320e-01 8.33710968e-01 6.27330244e-01
6.04321778e-01 -1.70179617e+00 -4.41973507e-01 6.44026279e-01
9.72419828e-02 1.18284166e-01 1.78946778e-01 2.70294011e-01
-6.35410964e-01 -2.63020962e-01 -3.43194276e-01 -3.17325711e-01
-9.66108859e-01 3.74572784e-01 3.53699684e-01 2.17681527e-01
-5.27497172e-01 9.72628891e-01 6.10391915e-01 -3.06245983e-01
5.58042452e-02 -8.50222766e-01 1.05495326e-01 -1.75171494e-01
5.02902210e-01 4.62751091e-02 -3.63899171e-01 4.99624535e-02
-2.41385594e-01 8.58694613e-01 1.30013779e-01 -1.94536880e-01
1.12082577e+00 9.59374662e-03 -1.21710971e-01 9.32100117e-02
9.04809356e-01 5.99765897e-01 -1.91570055e+00 1.89351380e-01
-8.17416131e-01 -6.54465020e-01 -2.47926876e-01 -1.33189869e+00
-1.35347748e+00 7.20454812e-01 4.50258881e-01 2.83349276e-01
1.39276838e+00 -1.16841547e-01 6.34381294e-01 -3.38651478e-01
2.79347956e-01 -1.12540889e+00 2.23332271e-01 1.98926345e-01
1.22039092e+00 -6.07044220e-01 -2.35430241e-01 -5.36791623e-01
-9.92353678e-01 1.13383019e+00 7.45497048e-01 -2.09980458e-02
2.60634363e-01 7.86137879e-01 2.59632051e-01 4.34523895e-02
-3.25196415e-01 3.04493576e-01 4.14690435e-01 6.30462170e-01
3.71660054e-01 5.10441065e-01 4.20763008e-02 5.60764611e-01
-3.62741202e-01 -1.52485028e-01 3.47008109e-01 6.28294885e-01
-5.15799038e-02 -1.21690226e+00 -4.89213496e-01 4.50184375e-01
9.85561237e-02 -2.67316729e-01 -5.55141032e-01 6.72317147e-01
4.18798715e-01 4.91729021e-01 3.49304199e-01 -4.69222784e-01
3.96135569e-01 -2.29694739e-01 5.41380942e-01 -5.65297902e-01
-2.66111374e-01 5.04323132e-02 -6.39287457e-02 -5.82620561e-01
2.14525219e-02 -1.52929664e-01 -1.18490863e+00 -2.73407191e-01
-1.28912285e-01 -1.27468526e-01 4.37990665e-01 5.44370651e-01
4.66789812e-01 8.09688091e-01 7.00281620e-01 -9.68926609e-01
7.51142949e-02 -9.26129043e-01 -7.06760406e-01 4.79043245e-01
1.96874306e-01 -5.11823833e-01 -1.06933139e-01 2.90268511e-01] | [11.658341407775879, -0.6035923361778259] |
541ce0e6-9bef-4281-921d-52371fced18b | parameter-inference-in-a-computational-model | 2101.06266 | null | https://arxiv.org/abs/2101.06266v2 | https://arxiv.org/pdf/2101.06266v2.pdf | Parameter inference in a computational model of hemodynamics in pulmonary hypertension | Pulmonary hypertension (PH), defined by a mean pulmonary arterial pressure (mPAP) $>$ 20 mmHg, is characterized by increased pulmonary vascular resistance and decreased pulmonary arterial compliance. There are few measurable biomarkers of PH progression, but a conclusive diagnosis of the disease requires invasive right heart catheterization (RHC). Patient-specific computational models of the cardiovascular system are a potential noninvasive tool for determining additional indicators of disease severity. Using computational modeling, this study quantifies physiological parameters indicative of disease severity in nine PH patients. The model includes all four heart chambers and the pulmonary and systemic circulations. We consider two sets of calibration data: static (systolic \& diastolic values) RHC data and a combination of static and continuous, time-series waveform data. We determine a subset of identifiable parameters for model calibration using sensitivity analyses and multistart inference, and carry out uncertainty quantification post-inference. Results show that additional waveform data enables accurate calibration of the right atrial reservoir and pump function across the PH cohort. Model outcomes, including stroke work and pulmonary resistance-compliance relations, reflect typical right heart dynamics in PH phenotypes. Lastly, we show that estimated parameters agree with previous, non-modeling studies, supporting this type of analysis in translational PH research. | ['REU Program', 'Amanda L. Colunga', 'Mette S. Olufsen', 'Mitchel J. Colebank'] | 2021-01-15 | null | null | null | null | ['clinical-knowledge'] | ['miscellaneous'] | [ 1.06770433e-01 -4.49182279e-02 -1.25223622e-01 9.07704607e-02
-3.58404070e-01 -6.54534280e-01 1.56964853e-01 1.70410797e-01
-7.16193244e-02 7.61459172e-01 3.80844444e-01 -8.28331769e-01
-7.38506556e-01 -5.48823178e-01 -5.71659431e-02 -4.00334775e-01
-4.03984427e-01 9.86333966e-01 1.29373912e-02 5.17793179e-01
3.62945609e-02 7.27973461e-01 -4.16040570e-01 -2.51606345e-01
9.85353947e-01 6.56885266e-01 -1.54647186e-01 1.15046465e+00
3.00117522e-01 8.39245617e-01 -3.77810121e-01 4.73495945e-02
3.00087512e-01 -8.08897018e-01 -2.41882995e-01 -4.11241978e-01
2.98109293e-01 -5.85081637e-01 -4.49265540e-02 6.07110783e-02
1.18878555e+00 -2.03054652e-01 6.49766743e-01 -6.60420597e-01
-6.11178502e-02 6.75771534e-01 -8.88334364e-02 6.84579313e-01
-3.30089033e-02 7.82843113e-01 9.01937366e-01 -5.69490969e-01
1.93304390e-01 6.09485209e-01 1.24446952e+00 2.35013694e-01
-1.75896466e+00 -2.06334591e-01 -5.42933226e-01 -4.86092418e-01
-1.15992379e+00 -3.87250483e-01 4.05744940e-01 -9.37326431e-01
7.25891232e-01 4.32449222e-01 1.31217694e+00 3.32187623e-01
4.88047779e-01 -4.03278559e-01 1.33187628e+00 -1.91834375e-01
-5.45318797e-02 2.52269119e-01 1.30464599e-01 2.57340342e-01
7.03942180e-01 5.13441265e-01 -2.37188414e-01 -7.70449460e-01
1.06593645e+00 4.59969975e-02 -6.28950953e-01 -3.32625628e-01
-1.13357639e+00 3.85202825e-01 -4.61922526e-01 1.88375726e-01
-6.04350030e-01 3.38054925e-01 3.04043382e-01 1.20737851e-01
-1.13375500e-01 8.64937663e-01 -6.70827866e-01 -6.21972799e-01
-1.02723610e+00 3.24986756e-01 1.09134686e+00 4.37867910e-01
-8.00873712e-02 5.97188845e-02 -5.99717200e-01 9.62595701e-01
5.56532323e-01 1.07968545e+00 1.08445056e-01 -1.70503616e+00
2.07420811e-01 1.69764891e-01 4.48933035e-01 -7.93903470e-01
-6.52142465e-01 -7.85367131e-01 -8.19609940e-01 -3.30423936e-02
7.23838508e-01 -3.81240427e-01 -4.89908397e-01 1.33325851e+00
1.04045622e-01 4.29995582e-02 -2.28848264e-01 7.11509645e-01
2.76768029e-01 -1.13583267e-01 2.82469511e-01 -8.40870023e-01
1.23742342e+00 -3.78586203e-01 -2.69548088e-01 9.56756398e-02
4.25679982e-01 -4.56550360e-01 7.51719236e-01 2.13531956e-01
-1.42475736e+00 -3.00004482e-01 -4.97218192e-01 3.07618171e-01
2.96994060e-01 -6.36264160e-02 1.27306044e-01 8.51672113e-01
-9.22415316e-01 1.04442477e+00 -1.00150347e+00 -5.96479811e-02
3.85622770e-01 -9.86266658e-02 2.61172950e-01 3.40557367e-01
-1.05784941e+00 1.17027938e+00 -2.84873903e-01 2.15922028e-01
-6.61886811e-01 -1.78397512e+00 -2.85472304e-01 3.14233750e-01
-3.13919902e-01 -1.38206851e+00 8.25839221e-01 4.51855034e-01
-1.23429084e+00 5.57848752e-01 -9.63945612e-02 -3.66462708e-01
7.57962227e-01 -1.13097681e-02 -2.20133573e-01 5.71893394e-01
-1.79376125e-01 -3.47093999e-01 6.00471616e-01 -9.13948238e-01
-5.92617542e-02 -3.34992915e-01 -8.70753765e-01 1.07552245e-01
4.57278162e-01 9.42907110e-02 2.24113092e-01 -3.76615196e-01
3.03095758e-01 -7.91868925e-01 -4.70596164e-01 7.16277361e-02
-1.76445886e-01 7.47391820e-01 6.85460046e-02 -1.00780499e+00
1.64041066e+00 -1.65333164e+00 -2.76483357e-01 6.30473375e-01
8.31376493e-01 1.96107700e-01 7.11405277e-01 4.67782557e-01
-7.20920265e-02 6.86889410e-01 -5.64049244e-01 9.07610655e-02
1.14912555e-01 3.35515067e-02 -2.93758065e-01 6.58453286e-01
-4.04720530e-02 1.23849118e+00 -5.15175998e-01 -6.08224988e-01
4.78484839e-01 4.87719655e-01 -7.03801751e-01 2.51416147e-01
1.83281749e-01 9.68794286e-01 -1.52675271e-01 4.07449096e-01
4.31021541e-01 -6.16027832e-01 6.10996664e-01 1.17952146e-01
-4.13315058e-01 3.10535997e-01 -7.18503296e-01 7.06871331e-01
-2.52933323e-01 2.76239127e-01 1.11550316e-01 -6.41263247e-01
9.29328620e-01 6.35950267e-01 9.73288476e-01 -2.83252478e-01
5.11663370e-02 2.80278325e-01 8.02060485e-01 -6.77791119e-01
-4.73645538e-01 -8.44905913e-01 6.45633101e-01 6.84217215e-01
-4.10958767e-01 -5.77500105e-01 -1.34869903e-01 -2.90806163e-02
1.21303761e+00 -1.15025520e-01 3.56466085e-01 -6.90759301e-01
4.44438368e-01 -8.09775889e-02 7.16320693e-01 8.86391521e-01
-5.94160914e-01 7.31833994e-01 1.07361937e+00 -4.77617085e-02
-1.35922778e+00 -1.41204715e+00 -1.00959730e+00 -4.98716570e-02
-4.74202991e-01 -2.83623129e-01 1.75384302e-02 2.84195632e-01
9.11333144e-01 3.73822391e-01 -2.80870914e-01 6.01075739e-02
-8.16854775e-01 -8.73163462e-01 6.88790917e-01 6.05911374e-01
-2.42919773e-01 -6.24937594e-01 -8.87447178e-01 7.30145931e-01
7.00558797e-02 -7.15618432e-01 -9.20045301e-02 -2.15543345e-01
-1.37030613e+00 -1.41926873e+00 -8.93484592e-01 -1.72731593e-01
2.66864449e-01 -5.32490253e-01 1.27497387e+00 1.93977714e-01
-7.37906396e-01 6.46181643e-01 3.51303577e-01 -1.47579566e-01
-6.14333749e-01 -4.56658423e-01 1.82800978e-01 -6.03174388e-01
-1.99731693e-01 -1.11586869e+00 -1.16036272e+00 3.54773313e-01
2.24136204e-01 -4.15275723e-01 5.07718265e-01 6.69490576e-01
1.04574943e+00 -1.91496268e-01 8.17738116e-01 -1.04410946e+00
6.51481152e-01 -6.29664838e-01 -4.71026570e-01 1.58463083e-02
-1.45691180e+00 -6.32891834e-01 7.72448957e-01 -2.27544293e-01
-7.57277250e-01 -6.93524063e-01 1.99355781e-01 -4.80022639e-01
1.35122046e-01 5.77665627e-01 4.74525332e-01 2.88023025e-01
7.13553369e-01 1.84497505e-01 4.89852756e-01 -4.34502482e-01
1.15686424e-01 3.61656725e-01 5.55832922e-01 -9.56870556e-01
7.33358920e-01 2.61089683e-01 6.80979490e-01 -7.27165103e-01
-1.45698950e-01 -3.72666657e-01 -6.80222392e-01 -2.05744207e-01
5.31434715e-01 -6.26659036e-01 -1.07635152e+00 1.31538361e-01
-3.24513465e-01 -7.28803277e-01 -9.37043190e-01 9.70550954e-01
-8.60918045e-01 5.10830104e-01 -6.61160290e-01 -1.01866007e+00
-6.25347257e-01 -6.59596384e-01 2.26865426e-01 -1.65843785e-01
-7.47023046e-01 -1.50305498e+00 6.14346683e-01 2.72024870e-01
9.74907756e-01 7.01365113e-01 1.32063520e+00 -3.80376697e-01
-2.96422541e-01 -1.17035948e-01 2.72704498e-03 4.47969347e-01
8.50045029e-03 3.94962877e-01 -2.68339008e-01 -2.19319120e-01
3.40414971e-01 4.61938679e-01 2.81213820e-01 1.04761136e+00
7.73315310e-01 -2.56942987e-01 -2.34431699e-01 6.65607631e-01
1.28608489e+00 2.63578713e-01 4.69760776e-01 -3.60377997e-01
4.87026930e-01 3.88851851e-01 4.55915183e-03 9.20520008e-01
2.69991904e-01 9.32174474e-02 -3.81599337e-01 4.19209749e-02
-5.50175272e-02 -2.30170712e-01 -1.60021603e-01 8.00325811e-01
-3.58603835e-01 2.79352427e-01 -1.32630098e+00 5.48351586e-01
-1.12711155e+00 -9.61346924e-01 -6.46755576e-01 2.44416642e+00
9.94436324e-01 1.63077116e-01 3.89247894e-01 -2.93265432e-01
4.82155561e-01 -8.91993865e-02 -5.93759537e-01 -2.01643854e-01
4.46301661e-02 6.37441337e-01 5.06012797e-01 8.26213896e-01
-2.60718763e-01 -5.49406670e-02 8.39069653e+00 -6.86418772e-01
-1.00652039e+00 -1.26735702e-01 5.88187158e-01 -1.20249987e-01
-4.52420950e-01 4.25502449e-01 -4.61738944e-01 5.80835462e-01
1.40883648e+00 -7.17241943e-01 5.01611650e-01 3.33642095e-01
7.92962313e-01 -1.77342653e-01 -1.15175807e+00 7.11785793e-01
-4.67882752e-01 -1.35976279e+00 -8.08233082e-01 5.77831641e-02
1.63541257e-01 1.44201621e-01 -8.47476721e-02 1.21504106e-01
-1.43780202e-01 -1.14840913e+00 -9.74766389e-02 1.25538969e+00
1.39390361e+00 1.10043250e-01 4.25179571e-01 9.79876071e-02
-9.93583322e-01 -2.93425377e-02 -1.41469747e-01 9.39349309e-02
5.17815650e-01 1.04219186e+00 -1.05046546e+00 -2.44322997e-02
3.52623522e-01 6.17944300e-01 -2.88839251e-01 1.14410424e+00
5.55263124e-02 1.12582815e+00 -6.09860480e-01 5.73340237e-01
-1.01854980e+00 -3.39875638e-01 7.90533543e-01 6.83791220e-01
3.53681862e-01 4.41801220e-01 -2.65264034e-01 1.79684508e+00
3.59920472e-01 8.93837884e-02 -3.75980943e-01 -2.56095767e-01
8.85318458e-01 1.02481222e+00 -9.84386578e-02 -2.64273912e-01
-2.83859551e-01 -2.16865137e-01 -1.62973940e-01 3.64538580e-01
-5.89138091e-01 -3.87142599e-02 5.14584243e-01 7.48060524e-01
-2.99655348e-01 -1.58463225e-01 -1.03725243e+00 -8.91422331e-01
-2.96221767e-02 -5.64011335e-01 3.59324634e-01 -5.57513535e-01
-1.04930198e+00 -1.19814329e-01 -3.62547226e-02 -8.83859515e-01
-3.82154644e-01 -3.62633288e-01 -8.87200058e-01 1.89057064e+00
-1.48956466e+00 -4.80312347e-01 -1.38804808e-01 -4.88987342e-02
-5.80158055e-01 2.61684716e-01 8.70581210e-01 3.31357837e-01
-5.15006483e-01 2.59091586e-01 -1.75817162e-01 -1.03631601e-01
7.03361452e-01 -1.64730060e+00 -4.88342643e-02 5.31103611e-01
-1.15124381e+00 1.22027898e+00 6.02289319e-01 -1.11342072e+00
-1.42450547e+00 -7.94177830e-01 9.28296328e-01 -9.94346261e-01
4.67450678e-01 4.50452268e-01 -7.65483737e-01 5.05941153e-01
-4.23571795e-01 3.78565431e-01 1.07577193e+00 -1.38193518e-01
1.74018115e-01 -1.50617823e-01 -1.29520285e+00 2.19766468e-01
7.19764054e-01 -4.35386926e-01 -5.22903502e-01 8.66443142e-02
3.03745806e-01 -4.62248653e-01 -2.35821056e+00 7.46312737e-01
9.85153258e-01 -7.67769754e-01 9.47629094e-01 -7.00034559e-01
2.09856600e-01 -2.99432814e-01 1.50742456e-01 -1.04423285e+00
-5.65958738e-01 -1.01924205e+00 -3.77406418e-01 9.02437687e-01
3.36144149e-01 -1.12539601e+00 5.55896699e-01 1.32967091e+00
-3.73329937e-01 -8.56641412e-01 -7.16273069e-01 -4.76091236e-01
5.86917639e-01 -5.70914596e-02 2.02685088e-01 7.66988993e-01
3.86752009e-01 -5.60420426e-03 -1.31976698e-03 1.06502317e-01
7.48763978e-01 1.84195295e-01 4.44689214e-01 -1.48001444e+00
-7.42695153e-01 -6.12403572e-01 1.79366261e-01 -3.82649422e-01
-4.37231511e-01 -8.79373729e-01 -2.75289625e-01 -1.66550875e+00
8.38112086e-02 -1.00587034e+00 -1.98584497e-01 9.24163461e-02
-1.81383878e-01 -4.55403596e-01 -1.64816797e-01 4.38649714e-01
4.54971492e-01 8.86729509e-02 1.52199113e+00 6.16927624e-01
-6.36388183e-01 6.43746927e-02 -7.35659242e-01 1.09241977e-01
9.14231300e-01 -4.73601490e-01 -6.11386716e-01 4.93737906e-01
1.10227853e-01 9.90203679e-01 5.35648704e-01 -8.04852843e-01
-2.33371317e-01 -4.73280936e-01 7.24529088e-01 -4.00841564e-01
-1.02664158e-01 -6.96821272e-01 7.41837621e-01 9.56379414e-01
-2.89765388e-01 -6.65822998e-02 1.60139009e-01 2.10552216e-01
3.56620908e-01 1.66442767e-01 1.17909861e+00 -3.07631612e-01
5.40993810e-01 5.30690372e-01 -6.23745382e-01 9.37603116e-01
6.51339769e-01 -1.30629748e-01 -7.31682718e-01 -2.16784209e-01
-1.21705031e+00 3.14491749e-01 2.87318856e-01 -5.13032734e-01
4.01961774e-01 -7.45529234e-01 -1.04704475e+00 2.01889560e-01
-2.83640683e-01 -6.67565390e-02 5.29410601e-01 1.85306466e+00
-1.05420387e+00 5.71233809e-01 2.84155518e-01 -7.61998653e-01
-6.69421256e-01 3.10899109e-01 1.29044366e+00 2.63198856e-02
-1.06295633e+00 4.53977436e-01 -2.74464816e-01 -7.66111389e-02
-1.53464690e-01 -3.63738954e-01 1.67584807e-01 -3.56052309e-01
1.81704536e-01 7.47557759e-01 -5.57089746e-01 -1.25030771e-01
-2.60649472e-01 5.01374006e-01 6.48611367e-01 1.55125158e-02
9.11948860e-01 -4.05230582e-01 -3.60669523e-01 5.15075684e-01
6.73134983e-01 3.19735914e-01 -8.87869358e-01 2.96104215e-02
-3.88892442e-01 -3.50988269e-01 -1.57651156e-01 -1.13350308e+00
-7.66068876e-01 6.93150520e-01 5.44226289e-01 3.59677613e-01
7.32744873e-01 -1.80888072e-01 3.66257668e-01 -1.15758650e-01
-3.05160612e-01 -8.39344740e-01 -4.54615504e-01 -1.95467964e-01
9.22034025e-01 -6.23707891e-01 4.72410947e-01 -2.84093559e-01
-5.09236693e-01 8.08366418e-01 4.76172060e-01 6.11691456e-03
1.10983706e+00 4.18030858e-01 3.25535774e-01 -2.13390008e-01
-8.98057878e-01 3.88772041e-01 -3.93383875e-02 5.50618052e-01
6.53674245e-01 1.45368904e-01 -7.77984917e-01 8.68224204e-01
-2.21673608e-01 1.48008227e-01 5.90692163e-01 5.91524661e-01
-5.55909395e-01 -8.11756968e-01 -7.59177878e-02 1.08487868e+00
-6.91990554e-01 -2.77296484e-01 1.03697054e-01 7.86531508e-01
-1.97838351e-01 7.53530622e-01 -1.08716555e-01 3.13513935e-01
5.92814267e-01 5.67320049e-01 4.29839343e-01 -5.58631063e-01
-4.78262097e-01 2.76993036e-01 7.79581964e-02 -1.32262275e-01
-8.61568600e-02 -1.06464231e+00 -1.40062249e+00 -3.55638236e-01
-2.91157514e-01 7.47955218e-02 1.79406390e-01 3.39303643e-01
6.28438830e-01 7.05952764e-01 6.93031251e-01 -5.18810824e-02
-9.66338754e-01 -1.00043750e+00 -1.09478664e+00 4.01559509e-02
4.71599430e-01 -6.04512691e-01 -1.02927971e+00 -8.73771235e-02] | [14.103777885437012, 3.038763999938965] |
32bf654d-9c0e-4a5e-b9dd-02e27ec0c2da | fashionpedia-ontology-segmentation-and-an | 2004.12276 | null | https://arxiv.org/abs/2004.12276v2 | https://arxiv.org/pdf/2004.12276v2.pdf | Fashionpedia: Ontology, Segmentation, and an Attribute Localization Dataset | In this work we explore the task of instance segmentation with attribute localization, which unifies instance segmentation (detect and segment each object instance) and fine-grained visual attribute categorization (recognize one or multiple attributes). The proposed task requires both localizing an object and describing its properties. To illustrate the various aspects of this task, we focus on the domain of fashion and introduce Fashionpedia as a step toward mapping out the visual aspects of the fashion world. Fashionpedia consists of two parts: (1) an ontology built by fashion experts containing 27 main apparel categories, 19 apparel parts, 294 fine-grained attributes and their relationships; (2) a dataset with everyday and celebrity event fashion images annotated with segmentation masks and their associated per-mask fine-grained attributes, built upon the Fashionpedia ontology. In order to solve this challenging task, we propose a novel Attribute-Mask RCNN model to jointly perform instance segmentation and localized attribute recognition, and provide a novel evaluation metric for the task. We also demonstrate instance segmentation models pre-trained on Fashionpedia achieve better transfer learning performance on other fashion datasets than ImageNet pre-training. Fashionpedia is available at: https://fashionpedia.github.io/home/index.html. | ['Serge Belongie', 'Mengyun Shi', 'Menglin Jia', 'Mikhail Sirotenko', 'Yin Cui', 'Hartwig Adam', 'Claire Cardie', 'Bharath Hariharan'] | 2020-04-26 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/1203_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123460307.pdf | eccv-2020-8 | ['fine-grained-visual-recognition', 'fine-grained-visual-categorization'] | ['computer-vision', 'computer-vision'] | [ 1.50511339e-01 -8.22860003e-02 -3.82305443e-01 -7.95983672e-01
-7.07344770e-01 -9.41505373e-01 4.08526510e-01 2.99816102e-01
-2.13426799e-01 3.98582757e-01 4.61106300e-02 1.55703068e-01
-2.99378559e-02 -7.46620119e-01 -8.63680959e-01 -5.26554525e-01
5.17586805e-02 9.18563426e-01 -2.27940045e-02 8.93859118e-02
-1.12314127e-01 1.23549521e-01 -1.46201313e+00 7.26632774e-01
6.29093885e-01 1.56422973e+00 1.09902844e-01 5.07314622e-01
7.75209069e-02 6.31007135e-01 -5.65949202e-01 -5.78974247e-01
3.05691272e-01 -2.26858914e-01 -1.00067163e+00 5.47736406e-01
6.74405813e-01 2.00926438e-01 -4.10767458e-02 1.07366192e+00
1.52783841e-03 1.61107242e-01 9.86286640e-01 -1.44194841e+00
-1.38449979e+00 6.17809772e-01 -2.40850136e-01 -4.39670794e-02
3.52955125e-02 3.01226437e-01 1.13082600e+00 -6.30476058e-01
6.27840817e-01 1.18901575e+00 6.16981566e-01 3.15542847e-01
-1.52841425e+00 -4.45030987e-01 5.34265459e-01 1.91195056e-01
-1.74864852e+00 -1.33426219e-01 9.46363926e-01 -6.11511171e-01
5.08491576e-01 3.95459473e-01 9.85362589e-01 9.23763275e-01
-1.37168303e-01 1.08314073e+00 1.32369685e+00 -1.00441508e-01
1.90780118e-01 2.85829097e-01 3.62453490e-01 7.02597320e-01
-1.79973438e-01 -2.75873393e-01 -5.47725968e-02 3.48780841e-01
1.06936514e+00 -1.14243008e-01 2.64738888e-01 -4.86287534e-01
-1.30495834e+00 6.05455279e-01 7.31885791e-01 -4.43232320e-02
-5.11303186e-01 1.51669309e-01 5.24864256e-01 3.76899302e-01
6.26970291e-01 5.39319038e-01 -7.36379325e-01 3.09193909e-01
-7.58540452e-01 1.61791131e-01 7.09724188e-01 1.35332835e+00
8.84434760e-01 -2.75454462e-01 -4.08290327e-01 1.16438210e+00
2.01418817e-01 1.84828594e-01 -1.90973416e-01 -8.07631493e-01
1.97922528e-01 8.87543321e-01 -1.29681408e-01 -7.09514856e-01
-3.22045445e-01 -5.66838086e-01 -8.57464314e-01 -1.03139162e-01
6.81664348e-01 3.85711789e-01 -1.40991306e+00 1.60382676e+00
4.99297172e-01 1.80197023e-02 -2.70168811e-01 1.19825530e+00
1.34975946e+00 4.86584634e-01 9.58400369e-01 4.03729409e-01
2.15068460e+00 -1.07810724e+00 -4.50634390e-01 -3.66095930e-01
-1.67301387e-01 -7.98861206e-01 1.40790606e+00 1.61924854e-01
-9.25769746e-01 -9.98161912e-01 -1.01563954e+00 -2.12476537e-01
-8.86208534e-01 5.33876300e-01 7.58458078e-01 3.32209259e-01
-7.66987801e-01 2.43992701e-01 -5.17586291e-01 -4.90223497e-01
8.91764164e-01 3.75857592e-01 -3.56393486e-01 -7.27508888e-02
-1.06212330e+00 4.16709840e-01 6.24171436e-01 1.21373028e-01
-9.88657713e-01 -7.72806048e-01 -1.17711091e+00 -7.41833597e-02
4.96649057e-01 -9.66918707e-01 8.61831546e-01 -1.16594970e+00
-9.10578668e-01 1.66776443e+00 2.89370656e-01 -2.58372039e-01
2.13069797e-01 -1.11720031e-02 -5.58664441e-01 1.09075613e-01
4.16106254e-01 9.79931891e-01 8.76767218e-01 -1.50898480e+00
-5.87214410e-01 -3.99715960e-01 2.67476171e-01 2.80329064e-02
2.10506424e-01 -3.04792617e-02 -9.50903416e-01 -1.06859064e+00
3.60150971e-02 -8.95676017e-01 -2.42392719e-01 -2.32768849e-01
-9.56088066e-01 -4.69149381e-01 3.71977448e-01 -7.26926804e-01
7.41689324e-01 -2.25232410e+00 6.84454739e-02 2.18551546e-01
2.09545553e-01 -5.65305762e-02 -1.35618061e-01 -9.00047645e-02
5.27040549e-02 -3.76040488e-02 -3.18137705e-01 -2.06415877e-01
4.45358723e-01 3.26068938e-01 8.33412539e-03 2.72775352e-01
4.38359857e-01 1.33773839e+00 -6.70638859e-01 -8.05882037e-01
3.95779580e-01 4.25799370e-01 -2.78678477e-01 2.03216925e-01
-6.13691688e-01 8.59764695e-01 -5.08251309e-01 1.26679945e+00
6.83575988e-01 -3.71446878e-01 1.94153100e-01 -9.31151152e-01
2.11787298e-01 -2.82163978e-01 -1.16254079e+00 1.64502645e+00
-3.52185339e-01 2.31488720e-01 9.48366672e-02 -9.33385849e-01
1.09246993e+00 3.45300272e-04 4.36559528e-01 -7.95980573e-01
3.67068470e-01 -3.44411284e-02 -4.78885740e-01 -5.56219339e-01
5.69642007e-01 -9.09219831e-02 -9.15328205e-01 1.23684317e-01
4.56760168e-01 -6.17775694e-02 4.55578536e-01 3.71425375e-02
4.63840753e-01 4.60620731e-01 4.27305251e-01 -4.44340378e-01
4.77822930e-01 4.21448171e-01 5.82251668e-01 6.88901186e-01
-3.13992918e-01 4.61619973e-01 3.32411498e-01 -9.17542875e-01
-1.14304411e+00 -1.45642817e+00 -1.50241345e-01 1.54366028e+00
5.79785526e-01 -3.17546487e-01 -8.00190032e-01 -8.85410070e-01
1.65661231e-01 4.87578541e-01 -8.99266720e-01 7.29449466e-02
-5.11794209e-01 -6.98270082e-01 3.01921457e-01 8.84322286e-01
8.62083852e-01 -1.28201365e+00 -9.72500145e-02 4.84292395e-02
-4.26901013e-01 -1.44552672e+00 -7.21629798e-01 2.19594494e-01
-1.79656118e-01 -1.16087437e+00 -4.49970573e-01 -1.40159869e+00
7.33790219e-01 -3.79112035e-01 1.78781712e+00 -1.83069229e-01
-4.87829179e-01 6.19241416e-01 -1.20513007e-01 -2.87843257e-01
-3.43741953e-01 1.86280712e-01 -3.08524013e-01 2.63489485e-01
7.63601065e-01 -3.18552405e-01 -9.34816658e-01 6.64457262e-01
-5.83808005e-01 2.21205413e-01 8.21910262e-01 4.47091013e-01
1.44376385e+00 3.54534797e-02 4.90529865e-01 -1.07121003e+00
3.27064782e-01 -4.26839530e-01 -2.88941175e-01 3.89756233e-01
-2.05521151e-01 -3.40629280e-01 6.23927653e-01 -4.71283108e-01
-9.89198506e-01 4.83830631e-01 -2.50276148e-01 -2.87108094e-01
-9.91242588e-01 1.38032392e-01 -5.25941193e-01 3.41986835e-01
3.47277254e-01 1.83570579e-01 -3.50489825e-01 -7.08767176e-01
8.09565663e-01 4.95903045e-01 1.00596666e+00 -1.01719558e+00
4.91029054e-01 5.00053763e-01 -1.94767594e-01 -5.39628208e-01
-1.11941028e+00 -5.86744308e-01 -1.09236228e+00 -3.65235865e-01
1.51165247e+00 -1.12629807e+00 -8.89640808e-01 3.67803365e-01
-6.26875162e-01 -5.69972575e-01 -7.15045393e-01 1.38214111e-01
-9.43256855e-01 -2.35590622e-01 -9.29275811e-01 -1.90680966e-01
-1.60901740e-01 -1.10095179e+00 1.18588865e+00 1.17022581e-01
-2.66599327e-01 -7.75532484e-01 -3.72726291e-01 6.34544492e-01
6.95814267e-02 5.29313445e-01 1.16585326e+00 -7.57510304e-01
-6.04082465e-01 -3.85849401e-02 -6.62197232e-01 4.29882288e-01
1.46550998e-01 -3.84418815e-01 -9.55877185e-01 5.31383008e-02
-5.32127440e-01 -3.43743652e-01 1.09158766e+00 6.18845344e-01
1.58724785e+00 -2.93045968e-01 -2.75229245e-01 7.12254584e-01
1.55804431e+00 1.33772567e-01 3.31651062e-01 3.65579695e-01
9.63948727e-01 6.49127901e-01 9.97184575e-01 6.08488619e-02
7.38395870e-01 7.56287217e-01 4.42496359e-01 -6.02929533e-01
-6.72154546e-01 -3.85824591e-01 -3.03985238e-01 3.75134856e-01
-3.80541414e-01 1.66895594e-02 -4.43732083e-01 8.29145133e-01
-1.79851425e+00 -6.08158529e-01 -7.67494366e-02 1.76779687e+00
7.70527184e-01 8.84097349e-03 7.05917895e-01 -2.05401316e-01
8.98510396e-01 -5.21680079e-02 -6.83418095e-01 -4.44413275e-01
-7.37583414e-02 1.67459369e-01 3.63049209e-01 1.06429294e-01
-2.02349091e+00 1.11678886e+00 5.35124397e+00 9.28451240e-01
-5.40096283e-01 1.51223212e-01 1.13906693e+00 3.08186084e-01
1.08508818e-01 -4.24700946e-01 -5.89693785e-01 4.39197034e-01
3.31755012e-01 4.10055935e-01 3.96095604e-01 9.97222006e-01
-3.47537994e-01 3.76431614e-01 -1.24851072e+00 8.26674581e-01
-2.86621675e-02 -1.25245965e+00 9.90920439e-02 -2.89075404e-01
5.66777408e-01 -5.33545315e-01 2.27610216e-01 5.60210824e-01
5.83914459e-01 -1.17415476e+00 1.07957256e+00 3.91964257e-01
1.22152722e+00 -5.79990149e-01 4.46101874e-01 -3.27057958e-01
-1.94447923e+00 -5.41708097e-02 -1.76320225e-01 2.22005382e-01
3.67815346e-02 2.23219112e-01 -6.85915709e-01 5.54744482e-01
9.05825615e-01 8.01605582e-01 -5.29400170e-01 8.80743742e-01
-4.86275136e-01 4.81805712e-01 -7.39518777e-02 3.06763321e-01
2.57510751e-01 -3.10687959e-01 1.36826277e-01 1.44662559e+00
-2.32212365e-01 -1.16943037e-02 7.63491511e-01 1.15381277e+00
-2.00060636e-01 -3.49207595e-02 2.54302844e-02 1.95384026e-01
3.06742549e-01 1.60003293e+00 -1.31986380e+00 -4.77143496e-01
-3.56545120e-01 1.15307689e+00 1.87931359e-01 6.15190506e-01
-1.06387162e+00 -1.57911450e-01 1.05754018e+00 1.90246925e-01
6.28759325e-01 1.39329135e-02 -2.25124747e-01 -7.66675889e-01
-1.97723374e-01 -6.72878623e-01 9.55454707e-01 -5.84370077e-01
-1.75219703e+00 7.00903535e-01 8.45038816e-02 -1.07833874e+00
1.58792406e-01 -7.16192424e-01 -2.14401275e-01 4.82252955e-01
-1.32351148e+00 -2.15400457e+00 -5.02508938e-01 8.25793207e-01
9.47129488e-01 1.36875570e-01 7.36517191e-01 5.25062978e-01
-4.33787525e-01 5.08787632e-01 -4.96794283e-01 4.94114488e-01
5.38016617e-01 -1.52875197e+00 5.55876076e-01 1.81974396e-01
1.47483915e-01 4.32346553e-01 7.23550439e-01 -6.01014912e-01
-9.45058227e-01 -1.63562822e+00 4.46095228e-01 -6.44384325e-01
5.28040111e-01 -8.24395537e-01 -5.61543107e-01 1.08515298e+00
1.57062948e-01 1.23774059e-01 6.87948287e-01 3.00583035e-01
-4.18796122e-01 -1.94030896e-01 -1.35848725e+00 2.97134817e-01
1.28248465e+00 -4.45053667e-01 -5.11379957e-01 4.81910944e-01
6.61313474e-01 -1.74361035e-01 -1.44667304e+00 3.31057072e-01
5.02287328e-01 -4.21411753e-01 1.55723727e+00 -6.37193203e-01
2.95566559e-01 -4.63258624e-01 -4.11452889e-01 -1.29906070e+00
-7.23091364e-01 5.07590827e-03 1.20625563e-01 1.66230059e+00
2.57389039e-01 -2.88521230e-01 7.18439996e-01 4.71030593e-01
-3.99914712e-01 -8.88496459e-01 -6.99100912e-01 -8.20775747e-01
-9.59984958e-02 -4.31352615e-01 9.77601349e-01 8.80963802e-01
-5.99596798e-01 4.71573502e-01 -2.53838599e-01 3.19158018e-01
8.94910812e-01 7.02261388e-01 5.48402011e-01 -1.18747234e+00
-1.94925651e-01 -2.26650640e-01 -7.00317800e-01 -1.02500486e+00
1.14135727e-01 -1.12146795e+00 3.89241613e-02 -1.80421054e+00
3.80305797e-01 -8.94704223e-01 -6.05825782e-01 6.77381814e-01
1.66588482e-02 1.17006814e+00 1.74611822e-01 2.31674030e-01
-1.09407568e+00 2.44517088e-01 1.39519203e+00 -7.56943405e-01
-3.42328772e-02 -2.03916989e-03 -8.91480148e-01 7.56208777e-01
7.92140603e-01 -2.73186743e-01 -2.90886462e-01 -2.26212502e-01
-6.94732964e-02 -3.39431137e-01 9.58691657e-01 -8.77837777e-01
2.51246374e-02 -8.79320800e-02 8.51273119e-01 -6.77077234e-01
5.90217352e-01 -9.29439902e-01 2.97292292e-01 8.55757967e-02
-2.71189064e-01 -1.76726073e-01 1.71321765e-01 4.62983906e-01
-2.62383193e-01 3.24394345e-01 6.43676817e-01 -5.26773095e-01
-1.57831621e+00 4.58665013e-01 -2.30816409e-01 1.52659222e-01
1.21012712e+00 -2.12482348e-01 -3.63144428e-01 2.62223899e-01
-1.58317363e+00 2.04046935e-01 4.86290067e-01 5.31262100e-01
5.56750953e-01 -1.47467065e+00 -6.48292661e-01 3.69426906e-01
5.23700118e-01 -2.65369087e-01 5.26892424e-01 6.98710918e-01
-2.74622321e-01 2.33793586e-01 -5.03814876e-01 -7.37753212e-01
-1.33258224e+00 1.00863194e+00 6.32049739e-02 -1.69299409e-01
-5.82702518e-01 1.02073014e+00 8.26761842e-01 -6.91640973e-01
-2.08799336e-02 -2.91509569e-01 -3.44784409e-01 8.67552534e-02
6.18859194e-02 -5.82168885e-02 -1.33389324e-01 -8.71163666e-01
-4.56560075e-01 8.72497559e-01 -1.24490643e-02 5.70459783e-01
1.24329305e+00 -2.93606043e-01 -1.33265451e-01 3.66269261e-01
8.96921694e-01 -4.24889684e-01 -1.41463637e+00 -2.33892873e-01
-1.41170830e-01 -4.23497885e-01 -4.18891370e-01 -1.23809350e+00
-1.27044582e+00 4.11156237e-01 6.07242465e-01 3.57934862e-01
1.40815520e+00 8.64509225e-01 8.52662861e-01 -2.20472321e-01
4.05128896e-01 -1.15200949e+00 -4.26087044e-02 -7.74028003e-02
6.92171097e-01 -1.24551356e+00 -9.91010815e-02 -1.11937070e+00
-9.72168446e-01 6.36129439e-01 6.73809469e-01 -2.29280576e-01
5.86863995e-01 2.25175824e-02 1.74973339e-01 -6.82452023e-01
-1.82531655e-01 -5.74471295e-01 7.71590471e-01 7.30936289e-01
1.82191089e-01 6.51697516e-01 6.48153499e-02 1.05996072e+00
-1.10011227e-01 -1.18943751e-01 -2.48972252e-01 4.34854299e-01
-1.43694296e-01 -8.02918792e-01 -2.41330102e-01 6.12681270e-01
-2.82369703e-01 1.68894082e-02 -3.24396998e-01 9.98308837e-01
7.52142668e-01 9.27843451e-01 4.71427768e-01 -3.75443012e-01
6.33022428e-01 -5.78119755e-02 5.73801398e-01 -3.50827366e-01
-7.94215560e-01 2.11220980e-02 4.28767413e-01 -5.15583217e-01
-4.43833977e-01 -3.99471998e-01 -8.90185535e-01 -2.38227203e-01
2.82257378e-01 3.54508795e-02 4.42223132e-01 7.13532865e-01
1.96398169e-01 8.34800184e-01 1.52928278e-01 -6.56571746e-01
1.96691409e-01 -6.17667019e-01 -8.36955905e-01 9.05532777e-01
2.26695538e-02 -7.55548418e-01 2.36649185e-01 5.03604114e-01] | [9.759133338928223, 0.4427202045917511] |
008f1c48-bcc9-4e28-b4ed-43a9ee1c91eb | multi-spectral-vehicle-re-identification-with | 2208.00632 | null | https://arxiv.org/abs/2208.00632v2 | https://arxiv.org/pdf/2208.00632v2.pdf | Multi-spectral Vehicle Re-identification with Cross-directional Consistency Network and a High-quality Benchmark | To tackle the challenge of vehicle re-identification (Re-ID) in complex lighting environments and diverse scenes, multi-spectral sources like visible and infrared information are taken into consideration due to their excellent complementary advantages. However, multi-spectral vehicle Re-ID suffers cross-modality discrepancy caused by heterogeneous properties of different modalities as well as a big challenge of the diverse appearance with different views in each identity. Meanwhile, diverse environmental interference leads to heavy sample distributional discrepancy in each modality. In this work, we propose a novel cross-directional consistency network to simultaneously overcome the discrepancies from both modality and sample aspects. In particular, we design a new cross-directional center loss to pull the modality centers of each identity close to mitigate cross-modality discrepancy, while the sample centers of each identity close to alleviate the sample discrepancy. Such strategy can generate discriminative multi-spectral feature representations for vehicle Re-ID. In addition, we design an adaptive layer normalization unit to dynamically adjust individual feature distribution to handle distributional discrepancy of intra-modality features for robust learning. To provide a comprehensive evaluation platform, we create a high-quality RGB-NIR-TIR multi-spectral vehicle Re-ID benchmark (MSVR310), including 310 different vehicles from a broad range of viewpoints, time spans and environmental complexities. Comprehensive experiments on both created and public datasets demonstrate the effectiveness of the proposed approach comparing to the state-of-the-art methods. | ['Chenglong Li', 'Zhiqi Ma', 'Jixin Ma', 'Jin Tang', 'Xianpeng Zhu', 'Aihua Zheng'] | 2022-08-01 | null | null | null | null | ['vehicle-re-identification'] | ['computer-vision'] | [-9.57108513e-02 -8.52167666e-01 -3.67112160e-02 -4.77502555e-01
-8.04160237e-01 -5.45271873e-01 6.52220309e-01 -3.96043390e-01
-4.37681526e-01 3.93460989e-01 1.10263951e-01 1.13560095e-01
-6.43876716e-02 -4.94095057e-01 -7.51903534e-01 -1.01248455e+00
6.97162926e-01 1.70154795e-01 1.35506183e-01 -3.44471693e-01
-7.59957533e-04 6.20218813e-01 -1.99236810e+00 9.70065594e-02
9.91570950e-01 9.00854707e-01 3.88419479e-01 1.70826409e-02
-1.82131976e-01 5.64244352e-02 -2.33710930e-01 -2.65361547e-01
4.24972683e-01 -2.73465812e-02 -1.49473757e-01 2.11431697e-01
7.74257779e-01 -2.24389881e-01 -2.87726521e-01 1.29520941e+00
6.70313418e-01 3.42240721e-01 6.98417842e-01 -1.57738745e+00
-7.05356836e-01 -5.69966063e-02 -9.91172969e-01 1.15703143e-01
-8.79799053e-02 3.10289353e-01 5.29535055e-01 -1.03830695e+00
3.79815578e-01 1.32147932e+00 7.67449617e-01 5.55736601e-01
-1.14701486e+00 -1.17986381e+00 2.50734717e-01 6.07740402e-01
-1.48724794e+00 -6.71836972e-01 1.06979942e+00 -4.83635098e-01
4.96405542e-01 1.57612726e-01 2.43911818e-01 1.10529542e+00
-2.63986170e-01 4.98115003e-01 1.02404737e+00 1.39970690e-01
-2.94551432e-01 4.47652847e-01 -2.70683151e-02 2.29854748e-01
3.60123694e-01 1.66887730e-01 -3.12388539e-01 2.17891820e-02
1.43133625e-01 3.99291039e-01 -5.76602481e-02 -3.86573762e-01
-1.15497684e+00 4.48551536e-01 2.17540219e-01 -1.05051957e-01
-7.92414099e-02 -1.44013256e-01 6.58173025e-01 2.06357837e-01
3.21018875e-01 -3.65613043e-01 -4.30152416e-01 3.66651326e-01
-3.09495419e-01 2.09081203e-01 -6.19391613e-02 9.01158035e-01
1.05968964e+00 2.46363923e-01 -1.73329473e-01 1.24906552e+00
6.72647834e-01 1.20463896e+00 3.23654920e-01 -5.64638734e-01
8.95761549e-01 4.09150243e-01 1.55345932e-01 -1.20788622e+00
-4.26158249e-01 -5.00791848e-01 -1.13655937e+00 1.75698102e-01
1.20312668e-01 4.75284010e-02 -7.54718363e-01 1.99701977e+00
7.06249774e-01 3.92091751e-01 3.17887902e-01 1.19999886e+00
1.29052079e+00 6.35795653e-01 3.01393837e-01 -1.88730508e-01
1.40431166e+00 -8.26272070e-01 -5.62908888e-01 -1.53995678e-01
4.16807383e-01 -9.72911716e-01 7.65377104e-01 -1.13224439e-01
-3.74245971e-01 -9.53901708e-01 -1.08402026e+00 2.28879705e-01
-3.76076877e-01 3.92133713e-01 1.36153087e-01 5.47114968e-01
-6.30428493e-01 -2.01784242e-02 -1.47592232e-01 -2.73888350e-01
3.55979055e-01 7.67079089e-03 -7.52550483e-01 -3.64844412e-01
-1.27251446e+00 7.75505483e-01 2.61123538e-01 5.02783298e-01
-7.48050869e-01 -7.82972932e-01 -7.21417427e-01 -4.33854252e-01
3.24233145e-01 -4.05152678e-01 6.27937078e-01 -1.01617754e+00
-1.22743893e+00 8.33688080e-01 -2.58036196e-01 2.62880266e-01
4.87137914e-01 1.48466200e-01 -1.15992272e+00 -2.86338449e-01
1.50529012e-01 2.02290192e-01 8.49079490e-01 -1.73713994e+00
-6.90199316e-01 -6.57748640e-01 -2.80097902e-01 4.49755847e-01
-4.28895295e-01 5.14360182e-02 -4.79297459e-01 -3.77192706e-01
1.31893605e-01 -9.74542499e-01 2.40290254e-01 -2.47359931e-01
-2.63294697e-01 -2.77765244e-01 1.01356030e+00 -5.84617376e-01
7.93005824e-01 -2.43180919e+00 -1.91744164e-01 9.06874910e-02
1.60170689e-01 4.67870116e-01 -5.53625286e-01 1.77238837e-01
-3.52259547e-01 -1.56773418e-01 1.73998326e-01 -3.86293858e-01
-1.29052745e-02 6.39666319e-02 -1.45074978e-01 8.48756492e-01
2.73862574e-02 6.10795975e-01 -7.40851581e-01 -3.74523550e-01
4.55137670e-01 6.29931390e-01 -4.86907028e-02 8.93001556e-02
2.07808942e-01 5.45461893e-01 -5.00298023e-01 7.76575148e-01
1.50896347e+00 1.20447382e-01 -2.79509634e-01 -9.70024824e-01
-2.35629335e-01 -3.77285451e-01 -1.33067369e+00 1.31013668e+00
-4.09379482e-01 3.36709857e-01 1.60871729e-01 -9.94626999e-01
1.06913102e+00 1.52010262e-01 5.45719981e-01 -1.10677934e+00
1.19761750e-01 2.78093696e-01 -2.90233642e-01 -8.21209013e-01
6.42159343e-01 -1.88560754e-01 1.64984584e-01 1.08472995e-01
-2.64288694e-01 5.01094818e-01 -1.31773517e-01 -2.03502715e-01
3.26815903e-01 -3.68341245e-02 -2.44129211e-01 -1.06034264e-01
1.01722789e+00 -2.56720155e-01 9.46281910e-01 2.95678884e-01
-4.13856655e-01 6.21578336e-01 -1.42819524e-01 -3.95153105e-01
-1.12031460e+00 -9.45367575e-01 -2.31779590e-01 1.03107750e+00
8.84776473e-01 1.71697155e-01 -1.80715859e-01 -5.40324867e-01
2.31586114e-01 3.82122010e-01 -4.56940234e-01 -1.23312652e-01
-4.08334345e-01 -1.06123400e+00 5.21954358e-01 4.35580224e-01
7.82582998e-01 -5.45930207e-01 2.50185639e-01 -1.13519125e-01
-3.88169199e-01 -1.35841525e+00 -6.32484674e-01 -3.94752085e-01
-3.95063877e-01 -1.12087071e+00 -5.99054515e-01 -7.84425855e-01
5.59801400e-01 1.03272331e+00 6.94154918e-01 -1.17121927e-01
-1.41319841e-01 2.21816987e-01 -3.32607746e-01 -2.54603654e-01
-3.03450525e-01 -2.83223867e-01 3.46837729e-01 5.97393095e-01
5.58865964e-01 -3.08893472e-01 -6.19216383e-01 6.98459208e-01
-8.54532897e-01 -1.98519249e-02 2.89428145e-01 8.70418549e-01
7.44028151e-01 1.60764128e-01 6.42989457e-01 -2.82988012e-01
1.70750558e-01 -7.81532705e-01 -7.75292575e-01 3.75001699e-01
-5.65868020e-01 -2.72180498e-01 8.11727583e-01 -5.40431440e-01
-1.39041519e+00 -1.59519957e-03 -1.71038695e-02 -6.19193733e-01
-3.00547689e-01 2.46790215e-01 -5.88439465e-01 -1.62148058e-01
4.22370225e-01 3.84440184e-01 1.33335680e-01 -5.00684857e-01
2.19663575e-01 8.50585699e-01 5.81514716e-01 -4.45009738e-01
1.27874100e+00 5.73588312e-01 6.82608783e-02 -6.55008614e-01
-4.65929687e-01 -6.97212934e-01 -1.76084235e-01 -3.55257690e-01
7.99800396e-01 -1.49673927e+00 -7.26296902e-01 8.95632207e-01
-1.05662894e+00 1.99734032e-01 2.38275111e-01 5.98511994e-01
3.18362787e-02 4.59705651e-01 -1.25460505e-01 -6.29417062e-01
-2.16734439e-01 -1.22438920e+00 1.10433400e+00 5.89790702e-01
7.81384945e-01 -6.28102124e-01 7.79755637e-02 6.67939186e-01
3.88198197e-01 1.11813575e-01 5.51789701e-01 -4.25270140e-01
-4.21900183e-01 2.22149771e-02 -7.34601319e-01 4.73076344e-01
3.42273265e-01 4.82406765e-02 -1.11255610e+00 -3.63121301e-01
-4.11319196e-01 -1.61750644e-01 9.07368243e-01 -3.89361725e-04
1.00474453e+00 7.21847266e-02 -3.09114128e-01 9.43137884e-01
1.67212963e+00 9.27418694e-02 5.02202988e-01 5.48566818e-01
1.25896859e+00 7.52201200e-01 8.54127586e-01 5.01025081e-01
8.58088911e-01 7.82307446e-01 5.61068952e-01 -1.62102684e-01
-9.43458378e-02 -8.35015848e-02 3.12053800e-01 5.85011601e-01
-1.11558937e-01 -2.47719854e-01 -5.98610401e-01 5.11579692e-01
-1.79866099e+00 -1.19312525e+00 -4.22499210e-01 2.26381397e+00
4.57493722e-01 -4.59781557e-01 2.11074054e-01 -2.66156822e-01
1.13649988e+00 3.14183265e-01 -6.89271808e-01 1.01693705e-01
-6.65451348e-01 -8.24855566e-01 8.87933731e-01 3.83120999e-02
-1.16964447e+00 5.34655690e-01 5.08374929e+00 1.21508873e+00
-1.27203202e+00 2.93740749e-01 4.90329236e-01 8.04230347e-02
-6.99549139e-01 -2.91065931e-01 -1.00998902e+00 7.16590941e-01
4.38558310e-01 -2.36836169e-03 5.73374391e-01 5.65670311e-01
2.90237606e-01 1.89547792e-01 -6.49059057e-01 1.59583664e+00
3.60112071e-01 -1.04558718e+00 -1.58978388e-01 -6.14199154e-02
8.94963622e-01 4.86240208e-01 1.29523218e-01 6.45653233e-02
-6.42017573e-02 -7.99923003e-01 7.84865797e-01 6.13116562e-01
1.03252602e+00 -8.44636381e-01 7.83054709e-01 2.09399126e-02
-1.74735141e+00 -2.13000745e-01 -4.11623448e-01 5.64665377e-01
4.00123335e-02 4.78870422e-01 -1.81745246e-01 1.03742290e+00
8.58389199e-01 9.12602365e-01 -5.85351408e-01 7.84569740e-01
4.10453171e-01 6.48637936e-02 -2.84770668e-01 4.50077683e-01
-1.10040762e-01 -3.87611836e-01 5.67255080e-01 1.06959343e+00
4.18847471e-01 -1.69117257e-01 2.62000412e-01 7.35733747e-01
7.37231821e-02 1.49493916e-02 -4.49140966e-01 1.54083326e-01
7.68107474e-01 1.41010773e+00 -1.76199794e-01 -1.85794726e-01
-7.01316595e-01 6.24038279e-01 -1.78829022e-02 7.33892262e-01
-1.15018368e+00 -1.01726539e-01 1.14808798e+00 -8.41042995e-02
2.41768450e-01 -1.19381540e-01 -4.36192192e-02 -1.12579215e+00
2.82760352e-01 -7.22659886e-01 2.67474800e-01 -7.56664515e-01
-1.73468554e+00 5.01797259e-01 -2.00712122e-02 -1.84413970e+00
3.24673206e-01 -5.30104876e-01 -4.36155409e-01 1.08961225e+00
-2.14773726e+00 -1.54767692e+00 -8.09249103e-01 8.35647285e-01
2.96780109e-01 -5.76936781e-01 2.19779149e-01 8.74538958e-01
-9.07266676e-01 9.09511983e-01 4.94087487e-01 -1.17818817e-01
1.07918036e+00 -5.08994460e-01 -7.23439676e-04 6.74331784e-01
-5.55752337e-01 3.23559701e-01 5.55921078e-01 -5.08392394e-01
-1.77099228e+00 -1.38275909e+00 4.28224891e-01 -4.33319546e-02
5.32109380e-01 -4.58301194e-02 -9.28096771e-01 3.00785184e-01
-2.66117603e-01 3.83262575e-01 4.96153653e-01 6.28765440e-03
-8.49095464e-01 -9.21237111e-01 -1.08351588e+00 3.01382154e-01
9.50832248e-01 -7.31793582e-01 -1.30512044e-01 1.85169145e-01
4.87433255e-01 -2.87261218e-01 -7.54125535e-01 8.08340430e-01
6.15334272e-01 -9.44445014e-01 1.15384400e+00 -1.49373010e-01
1.31695578e-02 -9.17709470e-01 -3.60836476e-01 -1.21032822e+00
-4.03123647e-01 -1.35956347e-01 4.34779704e-01 1.75295544e+00
8.01901519e-02 -1.05541527e+00 3.55401248e-01 2.81052232e-01
-2.66151011e-01 -2.81639665e-01 -8.76372457e-01 -8.26083601e-01
-1.85604900e-01 -3.49049151e-01 1.05175269e+00 9.62254047e-01
-6.89607143e-01 6.52108192e-02 -6.89742506e-01 4.74649400e-01
8.81074607e-01 2.64744192e-01 9.82514679e-01 -1.25823891e+00
1.74200758e-01 -3.66800666e-01 -3.79627526e-01 -7.71200597e-01
4.25947696e-01 -7.97768414e-01 1.56871632e-01 -1.22107995e+00
6.67349756e-01 -8.64915133e-01 -5.99391460e-01 2.55778164e-01
-2.78528780e-01 4.73624170e-01 1.22776993e-01 3.49577099e-01
-5.93598187e-01 9.19532180e-01 1.28380883e+00 -4.37533259e-01
5.20374663e-02 -2.91172594e-01 -7.22972870e-01 5.20244777e-01
7.17550397e-01 -2.88346469e-01 -5.25594413e-01 -7.18363822e-01
2.47644752e-01 -1.70684278e-01 6.88422680e-01 -8.56207013e-01
2.35512763e-01 -5.52539229e-01 4.13194031e-01 -8.63246441e-01
2.68357754e-01 -1.06130743e+00 6.12712562e-01 -1.61070600e-01
4.32452485e-02 4.32364792e-02 2.71538645e-01 7.66591311e-01
-3.07521164e-01 1.81427136e-01 8.87459338e-01 2.12000415e-01
-1.10649788e+00 6.37710750e-01 1.32714897e-01 -1.59837380e-02
1.01932216e+00 -3.88635665e-01 -6.56005979e-01 7.20536858e-02
9.64161232e-02 4.75967258e-01 6.79268241e-01 9.14753735e-01
4.17695612e-01 -1.81719029e+00 -9.80252147e-01 4.20660257e-01
7.15548456e-01 -2.63820648e-01 1.25763583e+00 9.79394972e-01
-6.33287355e-02 2.02097185e-02 -3.04077387e-01 -7.14344025e-01
-1.44000268e+00 5.20442128e-01 5.16337693e-01 3.31213742e-01
-3.78503174e-01 6.13894403e-01 4.75203067e-01 -7.90741742e-01
-8.46710205e-02 3.83770168e-01 -4.79556203e-01 2.43987188e-01
6.69421315e-01 5.03720582e-01 6.71001747e-02 -1.57601285e+00
-6.64117038e-01 1.08717763e+00 1.93326160e-01 4.23079729e-01
1.07890928e+00 -7.12005317e-01 -9.77138206e-02 2.76596129e-01
1.49179184e+00 4.07520495e-02 -1.28209281e+00 -5.26013792e-01
-7.86064148e-01 -6.62176967e-01 -1.28524706e-01 -5.57933807e-01
-1.32221794e+00 5.24712265e-01 1.05226088e+00 -1.90456241e-01
1.20533657e+00 -1.51203945e-01 9.10748243e-01 1.61466777e-01
2.04542518e-01 -1.39585829e+00 -1.02099396e-01 4.84041274e-01
6.37275875e-01 -1.75490487e+00 -2.94922087e-02 -4.08771127e-01
-7.30288208e-01 9.95452225e-01 7.99275458e-01 1.81637317e-01
5.95424354e-01 -1.12771384e-01 3.44196200e-01 2.90538277e-02
-2.74282753e-01 -2.48465791e-01 4.37271625e-01 7.74844587e-01
2.06530541e-02 7.55105913e-02 -2.76009180e-03 4.16080952e-01
3.41033518e-01 -2.01495931e-01 -6.67460309e-03 3.09657782e-01
-1.63122162e-01 -9.04649019e-01 -6.84950054e-01 1.84363335e-01
-1.14310253e-02 1.84879795e-01 2.56296486e-01 4.86618429e-01
6.31062210e-01 1.18588662e+00 -1.57932471e-03 -1.00625002e+00
3.15031946e-01 -2.51058549e-01 2.42980450e-01 9.41885337e-02
-1.26215219e-01 3.57876942e-02 4.00792435e-02 -5.78993797e-01
-8.23451340e-01 -7.73733079e-01 -1.17433476e+00 -5.28597713e-01
-3.47398162e-01 -2.92022496e-01 8.15148830e-01 9.82013643e-01
4.99573618e-01 3.76130432e-01 1.22067010e+00 -9.82430220e-01
-6.62899971e-01 -7.15755582e-01 -6.94829464e-01 7.38453388e-01
6.14083230e-01 -9.48272169e-01 -5.82923293e-01 -8.90074298e-02] | [14.503548622131348, 0.8088288903236389] |
3ddb2599-5591-4727-8c6d-f394aaadd51f | long-context-language-decision-transformers | 2302.05507 | null | https://arxiv.org/abs/2302.05507v1 | https://arxiv.org/pdf/2302.05507v1.pdf | Long-Context Language Decision Transformers and Exponential Tilt for Interactive Text Environments | Text-based game environments are challenging because agents must deal with long sequences of text, execute compositional actions using text and learn from sparse rewards. We address these challenges by proposing Long-Context Language Decision Transformers (LLDTs), a framework that is based on long transformer language models and decision transformers (DTs). LLDTs extend DTs with 3 components: (1) exponential tilt to guide the agent towards high obtainable goals, (2) novel goal conditioning methods yielding significantly better results than the traditional return-to-go (sum of all future rewards), and (3) a model of future observations. Our ablation results show that predicting future observations improves agent performance. To the best of our knowledge, LLDTs are the first to address offline RL with DTs on these challenging games. Our experiments show that LLDTs achieve the highest scores among many different types of agents on some of the most challenging Jericho games, such as Enchanter. | ['Christopher Pal', 'David Vazquez', 'Issam Laradji', 'Pau Rodriguez', 'Nicolas Gontier'] | 2023-02-10 | null | null | null | null | ['offline-rl'] | ['playing-games'] | [ 4.71078828e-02 1.63682774e-01 -1.94739833e-01 -8.16112235e-02
-1.06553280e+00 -6.18840814e-01 1.07600188e+00 -9.87899825e-02
-6.71939552e-01 9.35974121e-01 4.97814804e-01 -7.37859964e-01
-3.29428539e-02 -7.76465416e-01 -4.52719390e-01 -3.04740936e-01
-4.70000833e-01 9.64194894e-01 3.45688909e-01 -7.69178331e-01
2.04498142e-01 -3.30494307e-02 -1.32196689e+00 1.67654961e-01
7.15299666e-01 7.55892336e-01 4.30367738e-01 9.74001229e-01
5.16207293e-02 1.82316196e+00 -2.98486561e-01 -3.64700317e-01
4.37962353e-01 -5.16776979e-01 -8.22737813e-01 -3.08380395e-01
-2.91518122e-01 -7.65089333e-01 -3.55713874e-01 7.83291638e-01
4.89588976e-01 4.45331514e-01 3.64393562e-01 -1.25886917e+00
-3.52256656e-01 1.29825640e+00 -3.49016935e-01 9.57895592e-02
5.41966796e-01 6.77899599e-01 1.37683308e+00 -5.09822965e-01
5.38936973e-01 1.44529319e+00 4.96386349e-01 6.99117303e-01
-1.00714040e+00 -5.59928596e-01 7.70429552e-01 2.14386135e-01
-6.56999767e-01 -5.10966957e-01 3.73536527e-01 -2.92079955e-01
1.37767279e+00 1.76899880e-03 8.66022766e-01 1.37516284e+00
1.76508769e-01 1.30279970e+00 1.33687484e+00 -2.76009917e-01
4.74460185e-01 -5.82983375e-01 -1.47371605e-01 7.91786790e-01
-3.63321900e-01 7.34460115e-01 -8.34809363e-01 -1.98850051e-01
4.86939311e-01 -9.73775461e-02 2.64505446e-01 -7.29193836e-02
-1.13141072e+00 1.03567576e+00 -9.02644545e-03 3.91053855e-02
-5.59113801e-01 6.16753101e-01 3.06113362e-01 5.84594011e-01
3.59323591e-01 6.31752908e-01 -4.11314696e-01 -7.67837882e-01
-5.21205544e-01 7.15418160e-01 9.86540496e-01 9.12056625e-01
4.06179011e-01 3.68169814e-01 -3.43080848e-01 3.75719398e-01
3.47247094e-01 6.02579176e-01 5.48648953e-01 -1.25666511e+00
7.49327600e-01 8.40116739e-02 3.65996540e-01 -2.78833002e-01
-7.15697765e-01 -5.49014866e-01 -1.07951559e-01 5.40941417e-01
5.91471970e-01 -7.16354847e-01 -7.87541449e-01 2.20780444e+00
-9.84331518e-02 8.67114365e-02 4.78890181e-01 7.08195269e-01
3.96354705e-01 5.78289092e-01 2.66375154e-01 -1.60872757e-01
8.31949115e-01 -1.22988391e+00 -4.83875215e-01 -1.11170197e+00
7.79956460e-01 -2.87674546e-01 1.26259530e+00 6.38859332e-01
-1.32037318e+00 4.13347930e-02 -8.48099113e-01 4.83470596e-02
-1.42694386e-02 -1.26933724e-01 1.24568272e+00 4.28136766e-01
-1.38582873e+00 5.47020018e-01 -1.15388215e+00 -2.12484986e-01
1.75878897e-01 3.89616162e-01 1.37980133e-01 1.09094210e-01
-1.18903947e+00 9.19557810e-01 2.95509309e-01 -1.61130413e-01
-1.71561170e+00 -3.15834075e-01 -9.97197926e-01 -1.48756113e-02
1.09458196e+00 -6.25928342e-01 2.11842680e+00 -9.28662479e-01
-2.13208318e+00 6.95143044e-01 -3.11043318e-02 -9.80358720e-01
7.10621297e-01 -4.23432857e-01 1.75516412e-01 -2.43763223e-01
3.13190669e-01 5.08201301e-01 7.86045551e-01 -6.63244903e-01
-9.57669437e-01 -2.39043310e-01 5.01281381e-01 7.06275523e-01
2.68687611e-03 7.30041787e-02 -1.62098035e-02 -4.11803901e-01
-4.32749152e-01 -9.88726437e-01 -6.80313349e-01 -5.65090418e-01
-3.18858236e-01 -5.34467995e-01 2.37812370e-01 -5.23193955e-01
9.48618829e-01 -1.66670632e+00 3.47316712e-01 -2.32821524e-01
5.02816796e-01 -1.58443768e-02 -5.04379094e-01 7.22443104e-01
3.05963248e-01 -1.62737593e-01 1.62588954e-01 -6.27242148e-01
4.11839128e-01 2.78743774e-01 -6.58346295e-01 2.67236102e-02
-9.07966271e-02 1.42082143e+00 -1.30694640e+00 -1.98675916e-01
1.66761234e-01 -3.40561539e-01 -7.35930502e-01 3.23207647e-01
-1.10083973e+00 2.66743004e-01 -7.74404526e-01 3.00748855e-01
-1.73367679e-01 -1.23648293e-01 4.31852013e-01 8.14180791e-01
-1.91880196e-01 9.04311955e-01 -6.84828579e-01 1.94331932e+00
-2.60555327e-01 3.11491996e-01 6.55609295e-02 -7.66295135e-01
4.95158523e-01 1.89105168e-01 4.00747150e-01 -9.44137037e-01
2.29761899e-01 1.33660913e-01 1.07677616e-01 -2.36310065e-01
7.95544505e-01 -4.04469669e-01 -5.01470506e-01 8.56647253e-01
5.22275604e-02 -2.40634203e-01 5.31691492e-01 7.22434819e-01
1.25137377e+00 5.41320741e-01 1.67348385e-01 -5.45318536e-02
-5.37745655e-02 1.63279548e-01 6.10024095e-01 1.49987113e+00
-2.85405785e-01 -7.52132162e-02 8.78823578e-01 -3.47587436e-01
-8.97321463e-01 -1.02564728e+00 1.02324891e+00 1.79399240e+00
-5.05036674e-02 -7.39520729e-01 -4.38819706e-01 -5.51240742e-01
-2.70043910e-01 1.03572810e+00 -5.84569931e-01 -1.18065193e-01
-4.46048439e-01 -5.17818332e-01 6.45808995e-01 5.87793410e-01
3.24683726e-01 -1.25789154e+00 -8.50783050e-01 4.54400450e-01
-3.65389019e-01 -1.08366299e+00 -7.66159594e-01 4.75273967e-01
-6.07217610e-01 -9.70982850e-01 -3.50913823e-01 -5.78054130e-01
-7.52896518e-02 6.09289259e-02 1.30439746e+00 -2.19207749e-01
1.62435561e-01 5.40602803e-01 -5.13249695e-01 -5.30263186e-01
-4.71162200e-01 8.50055069e-02 1.52840689e-01 -5.13497472e-01
1.10565029e-01 -8.00586820e-01 -1.80873990e-01 -7.06927106e-02
-2.65958697e-01 5.42606652e-01 5.83225667e-01 9.64476168e-01
1.10738114e-01 -6.30971864e-02 3.18449467e-01 -8.49547982e-01
1.13231349e+00 -3.81775141e-01 -9.82682705e-01 2.88188402e-02
-4.92297679e-01 3.38448197e-01 7.71798074e-01 -5.17951787e-01
-1.11842632e+00 -2.19185457e-01 -1.72291294e-01 -7.03994408e-02
2.52731293e-01 5.71294665e-01 4.99667346e-01 2.63496816e-01
8.48252475e-01 5.59863448e-01 5.68293370e-02 7.34187886e-02
4.45326686e-01 1.68506086e-01 3.34524423e-01 -1.06567454e+00
7.00107574e-01 1.80622518e-01 -4.13419791e-02 -3.14722598e-01
-8.77749979e-01 -5.60657308e-02 2.67051578e-01 -3.64375293e-01
6.09271646e-01 -1.08324409e+00 -1.30060399e+00 6.69955432e-01
-7.10985124e-01 -1.46031094e+00 -4.36149597e-01 4.51992989e-01
-1.05307174e+00 1.47414833e-01 -1.01752520e+00 -1.34813130e+00
-3.85276675e-01 -1.19262230e+00 8.63529503e-01 2.44510233e-01
7.73488823e-03 -8.12586665e-01 1.33518770e-01 2.39293933e-01
4.33538616e-01 -1.53292015e-01 4.96760726e-01 -6.71922624e-01
-6.63736939e-01 2.69623309e-01 2.92943418e-01 -1.38523772e-01
-3.23198438e-01 -6.04938686e-01 -5.93283474e-01 -5.06936848e-01
-7.39910975e-02 -1.18757999e+00 7.25027442e-01 4.72906291e-01
5.47989607e-01 -3.38095278e-01 3.52332927e-02 4.27904606e-01
1.06485760e+00 3.90036821e-01 4.03845221e-01 5.48376083e-01
2.19357938e-01 3.40736538e-01 7.98013210e-01 8.17158759e-01
8.83271337e-01 5.31835556e-01 6.78909361e-01 3.09478074e-01
9.92594510e-02 -8.38364005e-01 1.06643212e+00 3.89995158e-01
-1.38758138e-01 -2.49070644e-01 -8.44814181e-01 5.17640114e-01
-2.26933527e+00 -1.12243748e+00 2.13089749e-01 1.77945268e+00
8.72332752e-01 4.82410520e-01 5.18774211e-01 -4.34405595e-01
1.77079113e-04 4.58366424e-01 -1.03523111e+00 -3.54506820e-01
4.52586040e-02 3.40063274e-01 4.48050529e-01 1.00179338e+00
-8.35905969e-01 1.69681847e+00 6.69416380e+00 9.62902427e-01
-6.93468869e-01 2.20900193e-01 4.40422624e-01 -5.19269407e-01
-1.77986458e-01 2.32292220e-01 -8.20122063e-01 1.36662468e-01
6.37617350e-01 -4.38622892e-01 1.23251474e+00 9.32238221e-01
2.53785968e-01 -1.60944074e-01 -9.82841134e-01 7.18220711e-01
-1.18588820e-01 -1.05510259e+00 -2.88000643e-01 1.21030450e-01
6.44591570e-01 6.07070029e-01 2.16541350e-01 1.00795162e+00
1.80887163e+00 -1.21529341e+00 1.18303204e+00 2.86908180e-01
6.09138787e-01 -5.61821103e-01 3.41245174e-01 7.40324497e-01
-1.07298195e+00 -5.22645593e-01 -1.86363578e-01 -7.07745731e-01
1.43073514e-01 -9.00068358e-02 -9.74630773e-01 2.61523873e-01
3.82580429e-01 9.32233214e-01 -1.47130325e-01 4.09143120e-01
-6.44271016e-01 6.99231029e-01 -4.75070924e-01 -5.18526614e-01
8.25084925e-01 -2.98778892e-01 6.92258120e-01 5.78876019e-01
1.30611107e-01 1.60740584e-01 7.18491495e-01 7.87462413e-01
9.06126127e-02 -2.30404437e-01 -5.15000880e-01 -4.15278137e-01
3.21843594e-01 8.84966075e-01 -2.35007077e-01 -2.48083889e-01
-1.70288309e-01 7.77462006e-01 7.59104967e-01 4.22859222e-01
-6.63587391e-01 6.63192049e-02 7.82422245e-01 -2.28660405e-01
1.33198664e-01 -4.27704245e-01 -9.20742452e-02 -1.33019769e+00
-1.10229954e-01 -1.27210557e+00 6.79902017e-01 -8.43831480e-01
-1.01851225e+00 5.71440637e-01 -2.75723934e-01 -8.59839559e-01
-9.14483488e-01 -4.05670166e-01 -5.32538772e-01 6.75232530e-01
-1.59624922e+00 -1.20780444e+00 1.77676678e-01 7.84816742e-01
9.18210447e-01 -4.67131585e-01 7.80044019e-01 -3.24097991e-01
-1.83610335e-01 3.00480902e-01 -1.68693084e-02 1.63683165e-02
2.54696459e-01 -1.54831302e+00 6.97304606e-01 1.00728619e+00
6.72914386e-02 3.26478392e-01 7.55139768e-01 -7.45337546e-01
-1.54309547e+00 -6.45947874e-01 5.78289270e-01 -4.65502739e-01
1.03275120e+00 -5.22336006e-01 -1.59786213e-02 1.40817583e+00
1.82315111e-01 -6.60454512e-01 3.10351878e-01 4.90445703e-01
-1.79942325e-01 1.78160697e-01 -5.81394315e-01 1.19882345e+00
1.34936130e+00 -4.54706788e-01 -7.06371725e-01 4.63227123e-01
8.42831731e-01 -7.08921909e-01 -3.40709269e-01 -2.21891299e-01
5.35069048e-01 -1.14222252e+00 7.45679677e-01 -9.27757204e-01
5.91790617e-01 -5.72425053e-02 -1.06566921e-01 -1.54529333e+00
-2.36170322e-01 -1.32772756e+00 -1.72990467e-02 5.14777720e-01
5.36416352e-01 -7.73758948e-01 9.19645071e-01 5.51353276e-01
-1.99998945e-01 -5.85004270e-01 -7.54631639e-01 -7.61646152e-01
2.46492013e-01 -7.95146108e-01 4.66214895e-01 4.90965843e-01
4.33891803e-01 5.58198869e-01 -9.93369222e-01 -1.41012162e-01
7.69479930e-01 1.82734817e-01 8.15392137e-01 -8.00856233e-01
-8.31710637e-01 -5.80713928e-01 2.46620223e-01 -1.62711263e+00
2.25889549e-01 -9.70457673e-01 2.94675440e-01 -1.49298346e+00
2.28519306e-01 -5.84756434e-01 1.90129608e-01 8.07062685e-01
-1.90420851e-01 -3.30345780e-01 6.10662162e-01 -6.91668913e-02
-1.34679794e+00 9.63043809e-01 1.20679355e+00 -2.45659240e-02
-2.73916990e-01 6.75231293e-02 -1.04487896e+00 7.68329740e-01
1.03260648e+00 -4.19589937e-01 -6.79099977e-01 -5.05998015e-01
5.91251433e-01 5.84141254e-01 4.70658578e-02 -8.49000871e-01
5.25176406e-01 -7.56547809e-01 -2.53278077e-01 -3.22287261e-01
5.38922846e-01 -2.03860879e-01 -2.64451087e-01 7.40081251e-01
-7.07564712e-01 4.43547256e-02 4.79484908e-02 5.57630718e-01
2.77347267e-01 -1.86781213e-02 3.70051473e-01 -4.50567842e-01
-8.40425134e-01 3.95615786e-01 -1.03086519e+00 5.25592029e-01
9.92286384e-01 1.66346103e-01 -4.46946949e-01 -1.08429277e+00
-8.07870924e-01 9.63526726e-01 1.23083830e-01 3.38908792e-01
5.62244594e-01 -9.43798900e-01 -8.30019414e-01 -1.96610793e-01
-8.65971670e-02 -1.04760118e-01 -1.01397701e-01 5.89840591e-01
-1.07332572e-01 5.10063350e-01 5.81819266e-02 -6.88349316e-03
-1.02749467e+00 4.06817406e-01 5.01577735e-01 -1.27788270e+00
-7.02939153e-01 9.00956333e-01 2.85049468e-01 -7.67724752e-01
2.83742964e-01 -2.23036572e-01 -1.40623763e-01 -3.97053331e-01
6.18607640e-01 5.88802733e-02 -4.29164290e-01 1.41699901e-02
-7.60858729e-02 -2.05806748e-04 -2.61107832e-01 -7.64884353e-01
1.47571588e+00 -1.21591464e-02 2.35578567e-01 3.14987361e-01
1.99640900e-01 -1.19767845e-01 -1.71514177e+00 -5.82802176e-01
-6.48709238e-02 -3.25284094e-01 2.96378452e-02 -1.12451911e+00
-6.09713018e-01 5.72766304e-01 2.11118758e-02 1.71925619e-01
9.42556262e-01 -1.40397018e-02 9.23385084e-01 6.99351847e-01
9.83109236e-01 -1.15156567e+00 5.16158998e-01 1.24597561e+00
7.14709878e-01 -8.66103649e-01 -3.91265154e-01 2.68636048e-01
-1.05304348e+00 7.92430699e-01 6.92265332e-01 -1.49086058e-01
3.53812836e-02 5.43135405e-01 -1.58541083e-01 -1.72866955e-01
-1.58582711e+00 -7.89307892e-01 -5.21602333e-01 8.50264192e-01
-1.58655390e-01 1.73595712e-01 -1.05497606e-01 8.18473220e-01
-6.44498110e-01 1.65423334e-01 5.16232431e-01 1.06755280e+00
-4.99755323e-01 -1.30278838e+00 -1.45767659e-01 6.52667701e-01
-4.23584670e-01 -5.11199117e-01 -3.07066768e-01 3.72836322e-01
-4.98344928e-01 1.28005552e+00 -2.78630465e-01 -4.82344389e-01
1.78552449e-01 -3.25373709e-02 6.35363460e-01 -5.49600065e-01
-7.18560159e-01 -1.02852039e-01 5.26884615e-01 -9.47326839e-01
3.65344249e-03 -7.99836338e-01 -1.24835861e+00 -6.51010513e-01
-2.71116961e-02 2.92896003e-01 2.69155085e-01 1.17643416e+00
9.93426442e-02 5.42726159e-01 4.78652060e-01 -8.64152312e-01
-1.01467240e+00 -9.22351480e-01 -6.24633908e-01 1.48629069e-01
2.66313344e-01 -6.19123161e-01 -1.30335167e-01 -4.70129043e-01] | [3.8423376083374023, 1.4546858072280884] |
4a357936-9366-4cc8-ba75-d6da432d89f8 | guiding-text-to-image-diffusion-model-towards | 2301.05221 | null | https://arxiv.org/abs/2301.05221v1 | https://arxiv.org/pdf/2301.05221v1.pdf | Guiding Text-to-Image Diffusion Model Towards Grounded Generation | The goal of this paper is to augment a pre-trained text-to-image diffusion model with the ability of open-vocabulary objects grounding, i.e., simultaneously generating images and segmentation masks for the corresponding visual entities described in the text prompt. We make the following contributions: (i) we insert a grounding module into the existing diffusion model, that can be trained to align the visual and textual embedding space of the diffusion model with only a small number of object categories; (ii) we propose an automatic pipeline for constructing a dataset, that consists of {image, segmentation mask, text prompt} triplets, to train the proposed grounding module; (iii) we evaluate the performance of open-vocabulary grounding on images generated from the text-to-image diffusion model and show that the module can well segment the objects of categories beyond seen ones at training time; (iv) we adopt the guided diffusion model to build a synthetic semantic segmentation dataset, and show that training a standard segmentation model on such dataset demonstrates competitive performance on zero-shot segmentation(ZS3) benchmark, which opens up new opportunities for adopting the powerful diffusion model for discriminative tasks. | ['Weidi Xie', 'Yanfeng Wang', 'Ya zhang', 'Xiaoyun Zhang', 'Qinye Zhou', 'Ziyi Li'] | 2023-01-12 | null | null | null | null | ['zero-shot-segmentation'] | ['computer-vision'] | [ 4.77764010e-01 6.28073573e-01 -1.70874611e-01 -4.44980592e-01
-8.41181219e-01 -7.61004210e-01 8.10342014e-01 7.54542351e-02
-3.66793483e-01 -6.11912459e-02 9.04006660e-02 -7.37671107e-02
1.79396003e-01 -8.60828817e-01 -7.79845059e-01 -4.07781333e-01
3.31182808e-01 9.09660399e-01 9.72145557e-01 -1.74465761e-01
1.46650389e-01 1.78907886e-01 -1.51020646e+00 1.55173540e-01
1.00797176e+00 1.23840487e+00 4.02343869e-01 5.40769756e-01
-4.64397699e-01 6.93792045e-01 -4.88713056e-01 -4.76008296e-01
4.39775944e-01 -5.09127080e-01 -1.08102643e+00 7.43396461e-01
7.65680730e-01 -4.11947668e-01 -4.14365172e-01 1.06991076e+00
1.24387309e-01 1.85999230e-01 1.11917150e+00 -1.34709227e+00
-1.27121162e+00 6.01030588e-01 -4.52794790e-01 3.31840664e-02
2.15593241e-02 4.03609008e-01 1.19194841e+00 -9.77199078e-01
1.25280154e+00 1.16056550e+00 3.77252966e-01 6.68248057e-01
-1.36125791e+00 -2.58670360e-01 3.59089732e-01 9.01239589e-02
-1.36038637e+00 -2.03265354e-01 6.65615857e-01 -7.73310661e-01
8.79247427e-01 1.87708158e-03 8.14752221e-01 1.32542360e+00
-3.63284975e-01 1.16699326e+00 9.30758536e-01 -3.45316887e-01
4.55469310e-01 3.56622636e-01 5.49761236e-01 9.29204762e-01
8.71751681e-02 -2.72087842e-01 -2.81850696e-01 3.71082753e-01
7.67178833e-01 -1.40126929e-01 5.47632538e-02 -8.56189191e-01
-1.37604368e+00 9.78870392e-01 6.71562076e-01 3.92133683e-01
-1.85911953e-01 5.05227037e-02 3.09907734e-01 2.69190148e-02
6.97993994e-01 3.78815651e-01 -5.69325387e-02 2.54631132e-01
-1.24681485e+00 2.71275997e-01 7.90638506e-01 1.40877259e+00
1.08290887e+00 -3.60966362e-02 -6.03445172e-01 8.47882271e-01
2.68493474e-01 7.09958255e-01 4.78371710e-01 -8.65858138e-01
3.92347127e-01 7.08176017e-01 -9.33508798e-02 -6.95878088e-01
-1.17570467e-01 -7.11054504e-02 -5.03807843e-01 -2.78474689e-01
3.10070693e-01 -5.90537442e-03 -1.47357547e+00 1.51224697e+00
4.54115450e-01 5.13721369e-02 1.86559767e-01 7.31226802e-01
1.04675484e+00 5.72160602e-01 2.49209553e-01 1.81133091e-01
1.31393242e+00 -1.55131590e+00 -7.08248079e-01 -4.06191766e-01
7.51473486e-01 -4.88634855e-01 1.36711872e+00 -6.45886883e-02
-1.03884697e+00 -7.01166093e-01 -1.15181589e+00 -5.00659287e-01
-6.51544392e-01 -6.62881956e-02 5.58571279e-01 4.09689367e-01
-1.21273339e+00 2.99141109e-01 -8.01633298e-01 -7.94989467e-01
6.74154162e-01 -9.00212005e-02 -6.96331933e-02 -1.28777564e-01
-9.34105933e-01 7.10948825e-01 6.29948378e-01 -4.18106616e-01
-1.31822336e+00 -5.52039444e-01 -1.15480375e+00 -5.13510741e-02
5.11510909e-01 -7.26512909e-01 9.89284933e-01 -1.08140600e+00
-1.09642768e+00 1.27727091e+00 1.19520515e-01 -4.22145933e-01
5.17192006e-01 8.68163034e-02 -2.59876046e-02 5.37234724e-01
5.54111719e-01 1.47127819e+00 1.04391944e+00 -1.53661728e+00
-5.65205753e-01 -3.61098409e-01 5.82248941e-02 1.12044260e-01
-5.26740730e-01 -2.50752360e-01 -1.08186126e+00 -8.89361262e-01
3.75978239e-02 -9.52760458e-01 -2.80794829e-01 1.18516691e-01
-8.76454055e-01 -2.39705160e-01 1.07017529e+00 -5.70456266e-01
9.91937637e-01 -2.16514802e+00 5.10374248e-01 1.01523593e-01
2.08717406e-01 5.39519340e-02 -3.08065891e-01 3.65979344e-01
1.88302651e-01 1.08014025e-01 -7.01784611e-01 -7.67041862e-01
3.42636436e-01 4.61329341e-01 -4.62362349e-01 1.79639697e-01
2.88828731e-01 1.48293412e+00 -8.07240367e-01 -8.69576693e-01
2.58169979e-01 1.11439817e-01 -5.19888520e-01 4.80574846e-01
-7.13434815e-01 3.44871759e-01 -4.36649829e-01 6.15241408e-01
5.27141333e-01 -5.57564974e-01 -1.97497532e-01 -1.80996522e-01
1.92034960e-01 -3.51319909e-01 -1.01854944e+00 1.97921968e+00
-9.21593681e-02 4.99718934e-01 -2.23900020e-01 -8.45998526e-01
7.11214542e-01 -3.71480957e-02 4.30617332e-01 -6.27302587e-01
1.57505080e-01 8.83153453e-02 -4.44078475e-01 -6.21968150e-01
6.01533473e-01 5.91791198e-02 -3.00239623e-01 6.26847208e-01
6.32257402e-01 -8.42957556e-01 4.96927053e-01 8.00894260e-01
8.38438392e-01 1.35270551e-01 -2.30310589e-01 -1.90156043e-01
1.75191417e-01 3.12584370e-01 -5.67006180e-03 7.75254726e-01
-2.57561535e-01 8.08631361e-01 3.48428518e-01 -9.44192111e-02
-1.29968250e+00 -1.20487738e+00 -1.78194419e-01 1.20895898e+00
6.17965937e-01 -2.93594003e-01 -1.22621417e+00 -1.16531670e+00
-7.53640011e-02 7.86099613e-01 -9.02873516e-01 1.07479103e-01
-9.01286304e-02 -3.76612782e-01 5.35890281e-01 6.81556940e-01
6.01035833e-01 -1.00770020e+00 -2.88837373e-01 -1.80712461e-01
-2.56224036e-01 -1.46241033e+00 -7.55684614e-01 9.92267802e-02
-5.85878670e-01 -1.12618518e+00 -9.75717485e-01 -1.21973586e+00
8.92937362e-01 3.54033858e-01 9.48174596e-01 2.32550632e-02
-2.50943214e-01 9.55185592e-01 -4.06241834e-01 -2.67497838e-01
-4.43454325e-01 1.00048915e-01 -4.87487435e-01 2.23493963e-01
2.49144912e-01 -1.65659264e-01 -7.87714303e-01 3.76616329e-01
-1.41383731e+00 3.33130598e-01 3.85720640e-01 4.46239501e-01
7.91858971e-01 -3.75137389e-01 3.29476833e-01 -1.04496837e+00
3.98743302e-01 -3.90932500e-01 -4.71787870e-01 3.04036230e-01
-5.50484180e-01 7.23213404e-02 1.13948353e-01 -4.79016662e-01
-8.96642923e-01 1.43457398e-01 -2.77556837e-01 -5.74151278e-01
-1.97571933e-01 8.57755616e-02 -7.54555166e-02 1.39944057e-03
6.44040883e-01 2.92424709e-01 -1.28715828e-01 -3.88042480e-01
1.21759880e+00 7.08240747e-01 5.71672618e-01 -4.95623380e-01
8.28639567e-01 8.39585721e-01 -3.09438467e-01 -9.05704796e-01
-1.17430818e+00 -7.83178031e-01 -9.45193887e-01 -1.38328403e-01
1.53208423e+00 -8.64736557e-01 2.97414750e-01 4.83265579e-01
-9.85493958e-01 -9.01676893e-01 -6.89471126e-01 -1.69540197e-02
-8.14346492e-01 7.50735104e-01 -6.15320504e-01 -2.36187741e-01
-2.40152001e-01 -1.30786908e+00 1.61617863e+00 3.06977220e-02
-1.53477445e-01 -1.33231723e+00 -8.59862715e-02 5.39107978e-01
8.93382132e-02 1.45065665e-01 8.86705697e-01 -9.46238101e-01
-8.26967001e-01 5.77800684e-02 -4.91371512e-01 6.66477144e-01
-1.46001950e-01 6.04978129e-02 -1.01399803e+00 -1.67834893e-01
-8.63577127e-02 -7.73407102e-01 1.16988766e+00 2.63159364e-01
1.09852850e+00 -6.98900148e-02 -5.41038752e-01 7.73624718e-01
1.29575348e+00 -2.02896759e-01 6.51680827e-01 2.22910225e-01
1.09313095e+00 7.66699433e-01 6.09510720e-01 -8.47547725e-02
5.85952163e-01 6.59084260e-01 3.73744398e-01 -4.90410477e-01
-6.32379949e-01 -4.57773089e-01 1.52875617e-01 9.17832017e-01
4.89835262e-01 -3.24241996e-01 -7.97120929e-01 1.01175010e+00
-1.87963641e+00 -6.41823530e-01 -1.36773452e-01 1.73121917e+00
6.50236249e-01 6.75252974e-02 1.63391247e-01 -3.07885706e-01
5.84198356e-01 4.46610063e-01 -5.40508270e-01 -9.08537433e-02
-1.71642318e-01 3.35407965e-02 3.98816317e-01 5.12022078e-01
-1.19846368e+00 1.59475029e+00 6.10144472e+00 1.13104784e+00
-1.01602101e+00 4.91459906e-01 5.72972119e-01 2.88332969e-01
-6.21549010e-01 1.69515498e-02 -7.45065451e-01 2.87040889e-01
5.46698689e-01 -9.57691148e-02 2.16312811e-01 1.00009692e+00
-2.26042986e-01 2.44227843e-03 -1.29549897e+00 7.26981580e-01
3.84216040e-01 -1.30792177e+00 5.59502482e-01 1.22259371e-01
1.23926520e+00 1.15292333e-01 2.31768876e-01 4.73832518e-01
4.98272449e-01 -7.37134457e-01 1.07868493e+00 2.09325731e-01
8.10449541e-01 -1.47670910e-01 2.76486605e-01 3.35389107e-01
-9.78168964e-01 1.23945080e-01 -4.18855250e-01 6.06370687e-01
3.24599385e-01 2.38242552e-01 -8.94312084e-01 4.72485304e-01
4.09701645e-01 8.26750100e-01 -9.63589072e-01 7.31589258e-01
-3.82729828e-01 6.12362504e-01 -5.20472340e-02 1.47034824e-01
7.02146888e-01 -3.69261265e-01 3.81428063e-01 1.23133874e+00
1.47147253e-01 -2.19475716e-01 4.05486345e-01 1.22486103e+00
-2.38085657e-01 2.23208889e-01 -6.99852288e-01 -2.56295860e-01
1.18599921e-01 1.30205548e+00 -1.31981671e+00 -6.62367404e-01
-4.06410396e-01 1.44271827e+00 3.66027266e-01 5.83427787e-01
-9.35735524e-01 -2.88234174e-01 1.52104363e-01 4.50759567e-02
8.14639986e-01 -1.90423697e-01 -2.95173794e-01 -1.21993232e+00
-1.77122504e-01 -6.30523920e-01 3.51209909e-01 -1.09579682e+00
-1.28758919e+00 7.35431671e-01 4.09379713e-02 -8.10840547e-01
-7.63031607e-03 -4.99082208e-01 -3.72875869e-01 6.50263071e-01
-1.33981228e+00 -1.69954216e+00 -4.54906374e-01 7.40066290e-01
1.02229297e+00 1.43613756e-01 5.87873101e-01 8.80318973e-03
-3.69036257e-01 3.17374200e-01 -1.72851048e-02 2.99282759e-01
5.46490133e-01 -1.50164425e+00 7.60419428e-01 9.23943996e-01
5.28359592e-01 4.06797796e-01 4.94267881e-01 -8.35142612e-01
-1.02410293e+00 -1.60109520e+00 4.96299922e-01 -9.29607034e-01
9.99953151e-01 -7.65619218e-01 -9.50475335e-01 9.32704389e-01
4.04304177e-01 -6.23370782e-02 4.45851266e-01 -2.96153724e-01
-3.44513267e-01 3.19573909e-01 -9.87540483e-01 7.12841034e-01
1.17738974e+00 -7.57461369e-01 -8.55017185e-01 6.81532145e-01
1.20550931e+00 -4.29533958e-01 -7.67247975e-01 9.88752618e-02
1.00794047e-01 -6.63572729e-01 1.01665521e+00 -4.53177482e-01
5.72121799e-01 -8.07250477e-03 -1.65029272e-01 -1.12943399e+00
-3.69270556e-02 -4.70302254e-01 -1.12217534e-02 1.48888052e+00
4.48236525e-01 -2.72298574e-01 5.97319186e-01 4.87085074e-01
-2.21267477e-01 -7.78044879e-01 -7.66998529e-01 -7.09843695e-01
9.34931710e-02 -5.32998919e-01 2.39551276e-01 9.33427155e-01
-3.16359431e-01 6.08929992e-01 -1.64915752e-02 -6.35491684e-02
5.57573378e-01 1.12897605e-01 8.64753962e-01 -1.00018382e+00
-2.81404763e-01 -3.33396107e-01 -2.17736557e-01 -1.59491408e+00
1.80601895e-01 -1.24437296e+00 1.54238135e-01 -2.03875184e+00
4.44885731e-01 -5.14703989e-01 -3.66918072e-02 3.72024864e-01
-2.22111061e-01 5.27573347e-01 2.59546459e-01 4.69385922e-01
-1.12907827e+00 7.49338031e-01 1.74364793e+00 -6.73953176e-01
-4.86754254e-02 -4.55165625e-01 -7.52589464e-01 6.91092253e-01
1.02869026e-01 -3.68085355e-01 -6.64199054e-01 -5.50109088e-01
-5.74707016e-02 -2.95780480e-01 3.60146433e-01 -7.66102612e-01
1.40431046e-01 1.92169603e-02 -4.73188981e-02 -4.61379290e-01
2.36583039e-01 -7.25738287e-01 -3.85535806e-01 4.61074002e-02
-4.85560656e-01 -4.86331850e-01 1.09865582e-02 9.19236302e-01
-1.03794776e-01 -3.63338947e-01 7.42736876e-01 -7.15394020e-02
-1.24015760e+00 4.83242750e-01 -1.03324905e-01 5.37945628e-01
1.37374723e+00 -3.99235904e-01 -4.67699856e-01 -2.25918666e-01
-8.46433342e-01 4.16935533e-01 8.97597432e-01 5.85409045e-01
4.06605244e-01 -1.26049733e+00 -2.73227334e-01 1.86470583e-01
5.78496635e-01 2.34539494e-01 3.29742819e-01 8.09631288e-01
-4.87816840e-01 3.28135163e-01 6.09490834e-02 -9.98993874e-01
-9.60161805e-01 9.64552224e-01 1.73680559e-01 -9.54923555e-02
-8.43463302e-01 1.17214465e+00 8.06164145e-01 -3.36060286e-01
2.99060732e-01 -5.62288284e-01 -4.38370034e-02 1.94594160e-01
2.04340205e-01 -1.06159560e-02 -1.16502568e-01 -8.79661977e-01
-2.93342788e-02 7.57141113e-01 -1.76553801e-01 -4.18848038e-01
1.15741515e+00 -2.33148649e-01 5.41560864e-03 4.72904325e-01
1.22678506e+00 -1.36084452e-01 -1.44892871e+00 -2.56285518e-01
1.84107795e-02 -2.19824582e-01 2.77434886e-02 -6.35819912e-01
-8.85749519e-01 1.09477437e+00 5.95708966e-01 5.35378575e-01
7.49604762e-01 5.18796682e-01 1.04217112e+00 1.88933492e-01
2.74838507e-01 -1.27384233e+00 6.22192025e-01 4.46488529e-01
8.48604620e-01 -1.15807819e+00 -3.96815419e-01 -6.31666124e-01
-1.09163642e+00 6.80100560e-01 4.99445975e-01 -1.14211842e-01
4.19446111e-01 -1.36834159e-01 2.95404822e-01 -6.80611253e-01
-2.92399019e-01 -7.41930246e-01 4.97047514e-01 8.54369283e-01
-2.36905292e-01 -2.37406924e-01 -2.04118248e-02 5.20403445e-01
7.95545131e-02 -2.94961065e-01 3.39580566e-01 6.87787354e-01
-5.76446295e-01 -9.60170984e-01 3.10407616e-02 2.26293311e-01
4.50611040e-02 -1.97177336e-01 -8.31148207e-01 8.43921363e-01
1.73125386e-01 8.40053022e-01 3.09005290e-01 -1.33476749e-01
2.60767817e-01 1.72026619e-01 3.84781748e-01 -1.05089045e+00
-3.06976944e-01 4.17052656e-02 -8.38999748e-02 -5.73433042e-01
-3.29105109e-01 -4.11025017e-01 -1.17942345e+00 3.89193803e-01
-2.42807046e-01 5.29556088e-02 5.51907539e-01 1.03163826e+00
2.18257338e-01 6.10074401e-01 2.60660887e-01 -9.60912406e-01
-3.71556431e-01 -9.67707753e-01 -7.79617608e-01 1.07379723e+00
3.19601148e-02 -7.10653365e-01 -2.84829438e-01 4.57600325e-01] | [9.72454833984375, 0.8254665732383728] |
17c0d985-8a79-49f3-9c58-67a0d466eda0 | mutual-and-self-prototype-alignment-for-semi | 2206.01739 | null | https://arxiv.org/abs/2206.01739v1 | https://arxiv.org/pdf/2206.01739v1.pdf | Mutual- and Self- Prototype Alignment for Semi-supervised Medical Image Segmentation | Semi-supervised learning methods have been explored in medical image segmentation tasks due to the scarcity of pixel-level annotation in the real scenario. Proto-type alignment based consistency constraint is an intuitional and plausible solu-tion to explore the useful information in the unlabeled data. In this paper, we propose a mutual- and self- prototype alignment (MSPA) framework to better utilize the unlabeled data. In specific, mutual-prototype alignment enhances the information interaction between labeled and unlabeled data. The mutual-prototype alignment imposes two consistency constraints in reverse directions between the unlabeled and labeled data, which enables the consistent embedding and model discriminability on unlabeled data. The proposed self-prototype alignment learns more stable region-wise features within unlabeled images, which optimizes the classification margin in semi-supervised segmentation by boosting the intra-class compactness and inter-class separation on the feature space. Extensive experimental results on three medical datasets demonstrate that with a small amount of labeled data, MSPA achieves large improvements by leveraging the unlabeled data. Our method also outperforms seven state-of-the-art semi-supervised segmentation methods on all three datasets. | ['Zhicheng Jiao', 'Chunna Tian', 'Zhenxi Zhang'] | 2022-06-03 | null | null | null | null | ['semi-supervised-medical-image-segmentation'] | ['computer-vision'] | [ 2.04932421e-01 5.32003403e-01 -7.97191858e-01 -7.05819130e-01
-8.05899084e-01 -3.57234716e-01 1.93607315e-01 3.20718698e-02
-4.19915825e-01 4.67575073e-01 9.10570696e-02 -2.02326011e-02
-1.21371165e-01 -4.21845645e-01 -2.28428811e-01 -1.06208980e+00
1.76134989e-01 4.18593228e-01 3.05767238e-01 2.39247233e-01
-5.69491014e-02 2.07543954e-01 -1.23970330e+00 1.72925517e-01
1.14424002e+00 9.97550786e-01 4.63255167e-01 5.63750677e-02
-1.94304749e-01 3.73566121e-01 1.80724457e-01 -3.38174626e-02
4.69384074e-01 -5.97614467e-01 -9.90307808e-01 6.87758923e-01
3.43563169e-01 -6.66401759e-02 -2.23078325e-01 1.23454309e+00
3.71379912e-01 -1.75676242e-01 8.02927792e-01 -1.24339533e+00
-6.04441881e-01 6.20894492e-01 -1.01632667e+00 1.36025771e-01
-3.04393947e-01 1.34085581e-01 9.89643633e-01 -8.43940258e-01
7.07509398e-01 8.66866529e-01 5.70824981e-01 5.12804151e-01
-1.18669689e+00 -4.98583674e-01 6.27918988e-02 8.35162252e-02
-1.36292493e+00 -5.49098626e-02 9.63166475e-01 -4.25762981e-01
2.70389527e-01 3.08359236e-01 7.03759551e-01 4.09357548e-01
-3.73202600e-02 1.23423111e+00 1.27754772e+00 -5.71341753e-01
3.16849232e-01 4.39729333e-01 4.50283885e-01 9.35473084e-01
1.90737486e-01 -8.48103873e-03 -3.82482052e-01 -1.59287844e-02
8.81259382e-01 2.55817641e-02 -2.06085160e-01 -8.36034179e-01
-1.28697467e+00 7.89430737e-01 6.04130149e-01 4.18570578e-01
-1.46344706e-01 -5.80465436e-01 4.03677613e-01 -1.15695652e-02
5.17383516e-01 3.05893093e-01 -5.47479451e-01 3.39377373e-01
-1.06941152e+00 -3.51576626e-01 3.76365751e-01 1.01543236e+00
1.01763058e+00 -2.01065779e-01 -8.04093853e-02 8.97364199e-01
5.85805237e-01 2.95027941e-01 9.41463947e-01 -8.16037774e-01
4.68900800e-01 9.66974258e-01 -3.58782381e-01 -6.82178140e-01
-4.36433494e-01 -5.10348141e-01 -1.00000894e+00 -7.68247768e-02
5.08729994e-01 -4.67414781e-02 -1.05792272e+00 1.46684444e+00
8.34847629e-01 2.18630508e-01 3.94558869e-02 9.93060887e-01
8.08111370e-01 2.80827671e-01 -5.45102134e-02 -5.56886137e-01
1.29328001e+00 -1.39139068e+00 -7.66793966e-01 -2.65719503e-01
1.01934385e+00 -6.40541255e-01 1.12528813e+00 -1.38595784e-02
-7.64223933e-01 -6.57832325e-01 -1.12763000e+00 9.86950025e-02
-9.70211178e-02 3.62146050e-01 5.87458193e-01 6.16743743e-01
-6.39990926e-01 4.75812793e-01 -1.02983522e+00 -1.84757918e-01
7.41332710e-01 4.10488248e-01 -6.28570139e-01 -1.82881206e-02
-8.56979728e-01 5.19615471e-01 7.19490469e-01 1.62808895e-01
-1.73987001e-01 -5.82683802e-01 -1.06405938e+00 -2.52760291e-01
3.16282630e-01 -2.53885478e-01 7.44504571e-01 -1.08686376e+00
-1.18829238e+00 1.23672140e+00 -5.09694070e-02 -1.25710204e-01
5.43004870e-01 1.03315197e-01 -1.81792498e-01 3.10843170e-01
2.77900964e-01 1.08682954e+00 7.21279860e-01 -1.28297162e+00
-5.04662693e-01 -7.22599506e-01 -5.47875404e-01 4.64984238e-01
-4.70608264e-01 -5.45420170e-01 -5.28098404e-01 -7.70259500e-01
8.74144256e-01 -1.14221179e+00 -4.69694376e-01 3.50449681e-01
-6.68897808e-01 -9.63597223e-02 1.07744706e+00 -3.20199579e-01
1.17112720e+00 -2.32307386e+00 -1.42596429e-02 3.23200047e-01
1.83160052e-01 2.64975995e-01 -1.05977952e-01 -7.98004270e-02
-1.26120761e-01 -5.43119991e-03 -6.20559335e-01 -1.47089109e-01
-3.41749817e-01 3.90575051e-01 5.70791513e-02 7.44373918e-01
1.95550248e-01 1.04903162e+00 -9.33264792e-01 -1.40484703e+00
2.83668995e-01 1.40742630e-01 -2.86088556e-01 2.64990330e-01
2.09322542e-01 7.95559406e-01 -7.68255770e-01 7.77394831e-01
8.38619709e-01 -5.77612102e-01 2.57101178e-01 -3.01328719e-01
3.02992493e-01 -2.09562421e-01 -1.16213584e+00 1.99847472e+00
-4.73810174e-02 3.42430562e-01 -1.18682258e-01 -1.25286162e+00
8.33008647e-01 8.37759450e-02 7.17486024e-01 -5.65713465e-01
5.99574260e-02 3.43780220e-01 1.03455577e-02 -6.36494398e-01
-5.32125065e-04 -9.07032117e-02 1.28674746e-01 5.12114704e-01
1.68103129e-01 -1.76735699e-01 1.09135315e-01 1.65250927e-01
4.89980489e-01 1.91082388e-01 6.81286931e-01 -7.01444387e-01
4.97572780e-01 -6.38010129e-02 8.47046852e-01 4.66346502e-01
-5.92944920e-01 7.95836031e-01 4.26272750e-02 -4.27803278e-01
-8.17298234e-01 -9.80655730e-01 -5.49744248e-01 6.72829688e-01
7.18348503e-01 1.25516709e-02 -9.32664514e-01 -1.35860062e+00
-1.38016626e-01 3.22182983e-01 -7.68602908e-01 -2.72915382e-02
-4.25553054e-01 -8.18418264e-01 1.79664791e-01 7.57029116e-01
6.35299325e-01 -8.99870992e-01 -5.87837696e-01 -7.38617927e-02
-1.69312060e-01 -9.39029396e-01 -8.31958234e-01 4.21643376e-01
-1.30899680e+00 -1.01619494e+00 -7.47325003e-01 -1.24643362e+00
1.36068130e+00 3.28516215e-01 6.67821527e-01 1.21737517e-01
-5.16654313e-01 -1.54852197e-02 -4.05365527e-01 -8.08567628e-02
-2.63927549e-01 6.89434931e-02 -1.11237578e-01 -1.18331667e-02
3.59679222e-01 -2.57014126e-01 -7.18068957e-01 6.78683937e-01
-1.01265550e+00 4.16775823e-01 7.61218071e-01 1.34794557e+00
1.10260630e+00 -1.82552487e-01 4.72956330e-01 -1.07896018e+00
-5.07775284e-02 -3.71867985e-01 -2.51560569e-01 5.39916277e-01
-9.63482738e-01 3.83819789e-01 3.21917295e-01 -5.77853918e-01
-1.05394745e+00 5.20167291e-01 1.27892673e-01 -2.60883838e-01
-1.86037615e-01 4.41001892e-01 -3.62616181e-01 -2.91211251e-02
5.32058120e-01 3.13469499e-01 2.05325767e-01 -2.27723822e-01
4.82200533e-01 8.29473197e-01 4.22324061e-01 -4.28445727e-01
6.32711411e-01 8.34735096e-01 -1.40905511e-02 -6.61796689e-01
-9.78139579e-01 -9.51525629e-01 -1.28089547e+00 -4.19620313e-02
8.84260893e-01 -7.66800940e-01 6.36546612e-02 4.23034132e-01
-5.17154753e-01 -2.07149416e-01 -7.03482330e-01 7.31904685e-01
-5.79223812e-01 8.24547410e-01 -4.82032806e-01 -4.38598305e-01
-3.90155345e-01 -1.48625195e+00 1.24078918e+00 5.30412316e-01
-1.30085945e-01 -1.06061900e+00 3.80486548e-02 6.08304977e-01
-1.69145793e-01 1.01186909e-01 7.58733690e-01 -1.05950940e+00
-2.91819304e-01 -1.43886194e-01 -3.34681600e-01 3.84970367e-01
5.65649569e-01 -2.92509228e-01 -8.66446495e-01 -3.01726162e-01
1.04592279e-01 -6.08660877e-01 9.58733916e-01 5.97790539e-01
1.24238598e+00 -2.14467660e-01 -5.04979730e-01 6.61574185e-01
1.16362619e+00 -3.32478806e-02 3.08331609e-01 1.20027788e-01
8.31028819e-01 8.22374642e-01 1.14578903e+00 2.53662467e-01
2.83240378e-01 4.49367315e-01 2.28389651e-01 -5.20872235e-01
-1.62770376e-02 -2.21193627e-01 -2.27483243e-01 1.09577131e+00
3.85234475e-01 2.75503457e-01 -7.02749014e-01 6.16638660e-01
-1.94750834e+00 -5.47570407e-01 -1.10642605e-01 1.96776783e+00
1.28014922e+00 -4.09975275e-02 -7.23320991e-03 1.35291636e-01
8.55978549e-01 3.56172696e-02 -9.35253620e-01 2.05048025e-01
-1.39882907e-01 -1.65196791e-01 6.45976841e-01 2.83308446e-01
-1.46355891e+00 7.13601470e-01 6.01354408e+00 1.14424622e+00
-1.08448505e+00 9.28084925e-02 1.17762315e+00 1.89441368e-01
-2.69545168e-01 -1.33060277e-01 -7.34649301e-01 4.97520208e-01
2.96952814e-01 7.61305913e-02 -2.14113981e-01 1.11918521e+00
-4.43552323e-02 -2.15955570e-01 -1.16136014e+00 1.02031660e+00
1.34975063e-02 -1.35308027e+00 -5.97547106e-02 9.10582989e-02
1.10805833e+00 -1.59936070e-01 1.11366913e-01 -9.05553997e-02
-3.92338522e-02 -7.73242950e-01 4.51852173e-01 3.55936140e-02
8.92083406e-01 -5.17113149e-01 8.69905829e-01 4.71780300e-01
-1.20993745e+00 1.18292466e-01 -2.59020418e-01 4.54892963e-01
7.31156543e-02 6.77560270e-01 -6.93161368e-01 5.22357881e-01
5.37301958e-01 8.53062212e-01 -6.54244304e-01 8.27404320e-01
-1.77626356e-01 6.07553005e-01 -2.48575404e-01 2.64777094e-01
3.57194930e-01 -5.48716009e-01 2.17917264e-01 1.10703933e+00
-2.11154819e-01 1.25157356e-01 4.81197208e-01 6.73961878e-01
1.16292788e-02 4.48618680e-01 -2.24036291e-01 1.00164406e-01
4.60002780e-01 1.40544260e+00 -1.24873650e+00 -3.85442108e-01
-2.46510983e-01 1.02494764e+00 2.22680092e-01 1.76915169e-01
-5.86138904e-01 -5.01788370e-02 -7.32478569e-04 -2.99891327e-02
8.97979736e-02 1.88792005e-01 -7.51610816e-01 -1.10169530e+00
1.03133477e-01 -5.71807981e-01 6.81853592e-01 -3.49842489e-01
-1.31574118e+00 4.40552294e-01 8.95959958e-02 -1.49092698e+00
2.61756703e-02 -3.71881843e-01 -4.96710926e-01 5.73103309e-01
-1.44540203e+00 -1.46541381e+00 -2.29294121e-01 3.78381312e-01
7.46170640e-01 -1.78390071e-01 5.94674349e-01 -4.89914343e-02
-6.19692147e-01 7.58862615e-01 2.83144236e-01 3.07915181e-01
8.53204668e-01 -1.28263843e+00 -8.28896686e-02 6.69527292e-01
4.21386927e-01 5.91588855e-01 3.13525438e-01 -5.93193233e-01
-1.03142989e+00 -1.09415030e+00 5.43772638e-01 -4.12947200e-02
4.75477040e-01 5.77801801e-02 -1.07614255e+00 4.99569386e-01
1.09829591e-03 6.41850650e-01 1.06449854e+00 -1.95138708e-01
-3.05218756e-01 -3.06865051e-02 -1.42652011e+00 3.20464879e-01
7.89270401e-01 -3.79485369e-01 -6.35298431e-01 4.26787466e-01
6.65875673e-01 -3.70259792e-01 -8.60326231e-01 6.53258324e-01
5.76524377e-01 -8.06215405e-01 8.10265481e-01 -3.96410286e-01
3.31068814e-01 -3.78687263e-01 3.81180197e-02 -1.02441919e+00
-2.02811822e-01 -2.63520420e-01 1.01099923e-01 1.08318365e+00
4.48522836e-01 -5.10284424e-01 1.11852074e+00 7.38553643e-01
-7.62630254e-02 -1.27528977e+00 -1.02535629e+00 -7.81924248e-01
1.86999395e-01 2.80236662e-03 2.42463604e-01 1.25917065e+00
2.87505597e-01 -4.07064930e-02 -8.32617730e-02 -8.81545059e-03
7.14622855e-01 4.43156272e-01 3.52399647e-01 -1.07452142e+00
-3.17003369e-01 -2.43757263e-01 -5.24755597e-01 -1.02286649e+00
2.24095359e-01 -1.04565191e+00 1.49834305e-01 -1.33241761e+00
7.12386191e-01 -7.06512868e-01 -4.29665029e-01 6.20505452e-01
-6.07760608e-01 3.72442842e-01 -1.66727543e-01 5.93186140e-01
-6.97374701e-01 5.66880167e-01 1.56443989e+00 -3.29288781e-01
-4.11525160e-01 -1.58090025e-01 -5.78181088e-01 8.64687502e-01
8.11149955e-01 -4.95540321e-01 -6.83167696e-01 -2.00342774e-01
-4.14538354e-01 -1.61661059e-01 -1.31393224e-01 -6.55358791e-01
1.94649577e-01 -2.95027077e-01 3.77693355e-01 -6.63717926e-01
-1.04595125e-01 -8.05598855e-01 -2.18661875e-01 5.07054865e-01
-6.23374462e-01 -5.32122076e-01 -3.05498820e-02 6.56659782e-01
-3.70275468e-01 -2.64605045e-01 1.18572330e+00 -1.12012876e-02
-7.39488482e-01 4.58088130e-01 2.95083463e-01 2.71897227e-01
1.31088972e+00 -6.61500633e-01 3.40444259e-02 1.30282000e-01
-7.68242240e-01 4.63372320e-01 5.26579618e-01 1.82460874e-01
6.36145890e-01 -1.41249919e+00 -6.03548646e-01 3.60368460e-01
4.50843364e-01 3.22770387e-01 3.74416530e-01 1.04000115e+00
-4.70771849e-01 2.76092380e-01 -3.07709336e-01 -1.00878143e+00
-1.39552164e+00 5.55969417e-01 1.44960850e-01 -4.20551807e-01
-5.35453677e-01 9.94593561e-01 5.35222709e-01 -5.87854087e-01
8.99013057e-02 -3.22898895e-01 -1.27522230e-01 -1.64753255e-02
4.09742922e-01 3.01799297e-01 -2.35860676e-01 -8.69454563e-01
-3.98983628e-01 7.32546926e-01 -3.36879879e-01 4.40298729e-02
1.03501332e+00 -3.31150264e-01 2.26450823e-02 4.68910843e-01
1.46314466e+00 -3.30759525e-01 -1.42497599e+00 -6.02047563e-01
1.29711181e-02 -5.30656636e-01 5.05480841e-02 -5.16542256e-01
-1.36093545e+00 8.57695401e-01 8.10737312e-01 -2.43132800e-01
1.12766731e+00 2.55923539e-01 6.52562082e-01 2.22225040e-01
2.35870153e-01 -1.30227113e+00 1.46087468e-01 -1.85947999e-01
4.13474768e-01 -1.79503751e+00 2.74544477e-01 -8.08206677e-01
-1.09646988e+00 9.51555431e-01 7.55489290e-01 5.86109422e-02
8.93272221e-01 1.18866771e-01 3.17137867e-01 -1.11177966e-01
-1.98719591e-01 -1.19030967e-01 6.33532941e-01 5.82208335e-01
3.67105097e-01 1.31320387e-01 -5.13480842e-01 5.01111567e-01
1.83553651e-01 -2.27414772e-01 5.39863855e-02 9.10631835e-01
-3.60157847e-01 -1.11309171e+00 -1.97380096e-01 6.41728401e-01
-1.97950855e-01 5.23627438e-02 -3.05902690e-01 7.38390744e-01
2.03344271e-01 8.07560861e-01 -1.83937862e-03 -1.40433848e-01
-2.32758075e-01 -1.68348402e-01 4.18566018e-01 -7.16595292e-01
-1.56587750e-01 4.02355105e-01 -4.25149769e-01 -4.05942529e-01
-6.59356296e-01 -6.67688489e-01 -1.80726898e+00 5.02909124e-01
-9.04003084e-01 1.41806319e-01 4.49818224e-01 1.05500031e+00
2.57218510e-01 8.95152986e-02 8.10606539e-01 -5.02769172e-01
-7.14169025e-01 -9.00991082e-01 -8.79271805e-01 6.11985743e-01
9.92258191e-02 -4.80517536e-01 -2.41847754e-01 3.06539655e-01] | [14.748893737792969, -2.080825090408325] |
f7ec971f-5479-4c1d-9dfd-37c48c0ceded | part-stacked-cnn-for-fine-grained-visual | 1512.08086 | null | http://arxiv.org/abs/1512.08086v1 | http://arxiv.org/pdf/1512.08086v1.pdf | Part-Stacked CNN for Fine-Grained Visual Categorization | In the context of fine-grained visual categorization, the ability to
interpret models as human-understandable visual manuals is sometimes as
important as achieving high classification accuracy. In this paper, we propose
a novel Part-Stacked CNN architecture that explicitly explains the fine-grained
recognition process by modeling subtle differences from object parts. Based on
manually-labeled strong part annotations, the proposed architecture consists of
a fully convolutional network to locate multiple object parts and a two-stream
classification network that en- codes object-level and part-level cues
simultaneously. By adopting a set of sharing strategies between the computation
of multiple object parts, the proposed architecture is very efficient running
at 20 frames/sec during inference. Experimental results on the CUB-200-2011
dataset reveal the effectiveness of the proposed architecture, from both the
perspective of classification accuracy and model interpretability. | ['Ya zhang', 'DaCheng Tao', 'Shaoli Huang', 'Zhe Xu'] | 2015-12-26 | part-stacked-cnn-for-fine-grained-visual-1 | http://openaccess.thecvf.com/content_cvpr_2016/html/Huang_Part-Stacked_CNN_for_CVPR_2016_paper.html | http://openaccess.thecvf.com/content_cvpr_2016/papers/Huang_Part-Stacked_CNN_for_CVPR_2016_paper.pdf | cvpr-2016-6 | ['fine-grained-visual-categorization'] | ['computer-vision'] | [ 2.39895552e-01 8.42111558e-02 -2.46062115e-01 -6.26002371e-01
-3.25743198e-01 -5.18127799e-01 6.22703791e-01 2.58401304e-01
-1.82095379e-01 2.08459198e-01 -1.70512825e-01 -4.15378422e-01
-1.66927576e-01 -5.84341764e-01 -1.03720605e+00 -3.98556620e-01
2.34958738e-01 4.88100648e-01 3.18603814e-01 2.12529659e-01
4.35533136e-01 8.74761820e-01 -1.90502572e+00 7.97049344e-01
6.62222803e-01 1.68010509e+00 9.49558988e-02 6.89902127e-01
-3.66172105e-01 1.13719499e+00 -4.51817840e-01 -4.18776810e-01
3.23676199e-01 1.12962544e-01 -8.48205984e-01 7.42328942e-01
8.89054596e-01 -4.26947176e-01 -2.51727737e-02 8.50243628e-01
-7.14333653e-02 -2.16525853e-01 5.15674055e-01 -1.33665335e+00
-7.57843435e-01 4.87770170e-01 -4.62726265e-01 1.89488843e-01
-1.29878387e-01 3.44719380e-01 1.04045832e+00 -1.00732851e+00
1.42525211e-01 1.42175055e+00 5.68917155e-01 3.52888793e-01
-1.29479253e+00 -4.72187549e-01 7.01583922e-01 4.19677347e-01
-1.53004265e+00 -4.26283449e-01 5.93038619e-01 -6.61397457e-01
8.92600894e-01 3.06469470e-01 4.76442516e-01 7.30541945e-01
-8.27450305e-02 7.50887573e-01 1.05320168e+00 -2.33254686e-01
2.98145235e-01 3.96198481e-02 6.56660199e-01 8.73311877e-01
6.74988568e-01 -1.39057994e-01 -4.83507305e-01 1.85221490e-02
7.08810091e-01 3.95436972e-01 -1.21903837e-01 -6.04857802e-01
-1.06253779e+00 4.28737700e-01 8.32816839e-01 1.71871409e-01
-4.04263288e-01 5.13684571e-01 4.11171168e-01 -1.41792253e-01
4.74873513e-01 1.27485976e-01 -5.81350148e-01 2.61047632e-01
-9.93976533e-01 -1.44692964e-03 6.31018281e-01 1.07808220e+00
8.40930402e-01 1.05099954e-01 -3.05771858e-01 4.95707333e-01
4.27099168e-01 1.66241646e-01 2.50229299e-01 -9.14510608e-01
3.23220581e-01 1.12879145e+00 1.27147034e-01 -8.08888733e-01
-2.32834637e-01 -6.10973358e-01 -1.12779903e+00 3.24434251e-01
4.47967887e-01 5.58440149e-01 -1.17413616e+00 1.36679435e+00
3.15866888e-01 5.10989949e-02 -3.02973747e-01 1.13361537e+00
6.83151186e-01 2.10868508e-01 3.31561327e-01 3.16772997e-01
1.81313038e+00 -1.15489137e+00 -4.35041338e-01 -3.64457279e-01
6.89903274e-02 -4.36809868e-01 1.01117873e+00 2.29927152e-01
-1.00874984e+00 -1.07574320e+00 -1.18626022e+00 -4.52160656e-01
-5.30266345e-01 5.91850221e-01 6.99439406e-01 3.38081419e-01
-8.63417864e-01 5.55024445e-01 -7.79964268e-01 -7.46569708e-02
9.35633540e-01 3.05975676e-01 -3.70208472e-01 -3.94248851e-02
-3.87378633e-01 6.78189397e-01 6.46915793e-01 4.73793238e-01
-9.92231250e-01 -7.14475155e-01 -8.02106917e-01 7.80986309e-01
3.01106811e-01 -8.43059540e-01 1.21020555e+00 -1.26344657e+00
-1.00791550e+00 7.53057897e-01 -4.27400500e-01 -6.31049097e-01
5.16616046e-01 -1.67611241e-01 -7.88907260e-02 3.44680518e-01
-8.86800885e-02 7.64438748e-01 1.12467849e+00 -1.51054013e+00
-7.73492813e-01 -5.48896492e-01 3.17017078e-01 -1.87978223e-01
-8.38646442e-02 -2.48027265e-01 -4.09560859e-01 -5.68035066e-01
2.99136430e-01 -7.37137616e-01 -3.38512696e-02 6.73176527e-01
-4.59045857e-01 -3.16120774e-01 7.76346505e-01 -3.68383318e-01
8.78087461e-01 -2.17739201e+00 -1.40238717e-01 -1.82651393e-02
7.11057186e-01 2.87546754e-01 2.78691240e-02 2.28304677e-02
-1.15282819e-01 2.15797797e-01 -1.73328981e-01 -4.49303061e-01
9.88570750e-02 1.59618646e-01 -4.15091097e-01 3.46358687e-01
6.83662415e-01 1.08310819e+00 -5.26111782e-01 -3.96888614e-01
3.75605762e-01 3.99793118e-01 -4.35799003e-01 2.39845455e-01
-3.18359673e-01 -7.94009399e-03 -3.61762732e-01 8.56723547e-01
9.03847694e-01 -8.48838925e-01 1.64977640e-01 -5.47396541e-01
-2.54793167e-02 -4.92336880e-03 -1.04594064e+00 1.34922910e+00
-3.94772261e-01 6.38196528e-01 7.07591176e-02 -8.81558239e-01
5.99246383e-01 8.33222046e-02 -1.34489655e-01 -6.20544016e-01
1.97256371e-01 -7.30363186e-03 1.68048255e-02 -1.65688813e-01
4.23995256e-01 1.22609861e-01 8.98800194e-02 3.48040700e-01
1.37609422e-01 2.43965104e-01 4.87464927e-02 1.38600826e-01
6.86281800e-01 1.72120258e-02 4.72285002e-01 -2.67615527e-01
5.10200739e-01 1.73870828e-02 3.67172331e-01 8.50995898e-01
-3.14800113e-01 5.04274845e-01 3.17963660e-01 -1.01302159e+00
-1.03622448e+00 -8.74913335e-01 2.72678323e-02 1.20990098e+00
4.07781154e-01 -3.06168616e-01 -7.81885326e-01 -6.87593043e-01
2.61574805e-01 4.34092730e-01 -1.00075376e+00 -7.59192631e-02
-5.30630767e-01 -1.22210577e-01 5.09344459e-01 9.92005885e-01
5.60256422e-01 -9.24252152e-01 -9.14344013e-01 -8.46660361e-02
-9.06329006e-02 -1.24104369e+00 -3.89900416e-01 3.71169358e-01
-8.04071546e-01 -1.38607669e+00 -3.51286948e-01 -5.53165078e-01
9.70267117e-01 6.92096770e-01 1.16669071e+00 6.31488621e-01
-4.74881232e-01 3.27656120e-02 -1.72126710e-01 -3.65938097e-01
-9.63966399e-02 -1.13111295e-01 -1.27438888e-01 4.59407151e-01
2.64629155e-01 -2.12173194e-01 -5.66649318e-01 2.27161914e-01
-9.05379474e-01 3.93655509e-01 9.13565814e-01 8.37051809e-01
6.84510946e-01 -2.72863656e-01 2.03635991e-01 -7.04357624e-01
1.33249357e-01 -8.10450539e-02 -7.74292707e-01 5.19151270e-01
-5.31401098e-01 2.83327967e-01 7.42689312e-01 -3.98679823e-01
-1.04921961e+00 2.15856448e-01 2.46744096e-01 -7.86715329e-01
-6.09111428e-01 3.31854336e-02 -1.94403559e-01 -2.97638953e-01
2.29398176e-01 3.91703159e-01 -2.55581111e-01 -7.11607575e-01
6.01357341e-01 6.10640287e-01 5.27338028e-01 -5.80999732e-01
7.52540290e-01 5.91251194e-01 2.79404372e-02 -5.74041307e-01
-1.10747933e+00 -5.11385739e-01 -8.78227472e-01 -2.39407830e-02
9.24306333e-01 -1.04361999e+00 -1.35062647e+00 4.09956366e-01
-1.35563028e+00 -8.34889486e-02 -3.02739829e-01 -2.04819087e-02
-4.19373274e-01 2.47592643e-01 -5.06134450e-01 -6.15546405e-01
-3.96133542e-01 -1.18703866e+00 1.45315993e+00 2.78051883e-01
1.05076227e-02 -5.53264141e-01 -6.76253319e-01 5.93198836e-01
2.86219716e-01 -6.39656335e-02 1.03032243e+00 -6.02755368e-01
-1.19523692e+00 -6.97549656e-02 -9.49182808e-01 2.49673009e-01
-9.15474892e-02 -1.16524217e-03 -1.21717787e+00 -1.07184112e-01
-3.18608433e-01 -3.08995754e-01 9.97672975e-01 1.68802366e-01
1.87388349e+00 -5.11448622e-01 -2.98989296e-01 6.20608270e-01
1.40115190e+00 -7.78607503e-02 2.79851764e-01 1.18154280e-01
9.25827920e-01 4.55675840e-01 3.61047298e-01 5.74529886e-01
4.52073485e-01 5.28072476e-01 7.83945203e-01 -1.04288302e-01
-3.25328857e-01 -9.75144506e-02 -3.20516616e-01 2.74005234e-01
-2.99065076e-02 -1.51999295e-01 -7.69150674e-01 5.40140688e-01
-1.82215643e+00 -1.01659834e+00 -2.00844899e-01 1.74906194e+00
4.28551286e-01 2.55297512e-01 -1.63289711e-01 1.31164446e-01
7.03278720e-01 1.62817910e-02 -6.31332040e-01 -3.64010006e-01
1.89692955e-02 1.15983486e-02 4.88414019e-01 3.17572981e-01
-1.12001324e+00 8.68635952e-01 5.85345888e+00 6.85351908e-01
-1.14592957e+00 1.78996380e-02 8.22940946e-01 1.38403043e-01
1.08410083e-01 -1.55929342e-01 -8.44217241e-01 2.63510287e-01
5.73222041e-01 1.43672690e-01 3.59223336e-01 1.07023644e+00
5.22813536e-02 1.13004632e-02 -1.47414756e+00 9.53834116e-01
1.30665511e-01 -1.67691660e+00 5.53689241e-01 -8.61953050e-02
4.18345451e-01 -3.27013314e-01 -3.11426958e-03 6.51835799e-02
-1.91295668e-02 -1.06538272e+00 1.45123208e+00 5.99513292e-01
7.44975150e-01 -4.88928109e-01 4.83970731e-01 4.86192763e-01
-1.58291519e+00 -2.98509181e-01 -2.31450841e-01 -1.97067216e-01
-1.94959521e-01 3.49612087e-01 -9.11375523e-01 5.10037065e-01
8.32341790e-01 5.37930608e-01 -9.43933129e-01 9.26984370e-01
-8.93637463e-02 4.95746493e-01 -1.46309827e-02 1.58789769e-01
2.71662563e-01 3.79822075e-01 2.96999495e-02 1.31436741e+00
-1.15513295e-01 1.49206012e-01 1.64886400e-01 1.04566717e+00
-1.91443056e-01 -6.46167159e-01 -1.38872057e-01 8.50064401e-03
3.42728674e-01 1.49564612e+00 -9.21054661e-01 -7.11210549e-01
-2.53639609e-01 9.11897898e-01 6.99028254e-01 2.25950047e-01
-8.60078156e-01 -2.58484095e-01 8.91521513e-01 9.30386558e-02
8.44343066e-01 -2.40860298e-01 -7.39695728e-01 -1.11908746e+00
1.57956421e-01 -7.27675617e-01 2.38010779e-01 -1.07609022e+00
-1.22385430e+00 8.20002139e-01 -2.09187686e-01 -1.21548474e+00
7.95770362e-02 -9.84712660e-01 -2.95905352e-01 7.72039771e-01
-1.57900810e+00 -1.59931743e+00 -7.93334067e-01 4.90564793e-01
7.77059793e-01 2.81992495e-01 6.52706683e-01 1.52551636e-01
-4.39590365e-01 5.71017981e-01 -4.10822421e-01 1.89095840e-01
1.80961370e-01 -1.27284753e+00 5.33194423e-01 7.70492792e-01
2.44922265e-01 8.92055929e-01 4.66475904e-01 -2.72212356e-01
-1.32634640e+00 -1.44005227e+00 6.56970143e-01 -4.56412971e-01
4.23832774e-01 -6.19742036e-01 -1.09016514e+00 6.72201574e-01
9.44256876e-03 4.33457673e-01 4.41803336e-01 -6.33698478e-02
-9.63960886e-01 -3.30325931e-01 -1.08224416e+00 3.42880070e-01
1.00426221e+00 -8.24628055e-01 -7.02680111e-01 1.51693583e-01
7.17982411e-01 -2.28756458e-01 -5.35143018e-01 4.07260239e-01
7.54579961e-01 -9.76248562e-01 1.12905693e+00 -7.55802989e-01
3.38754207e-01 -6.59045041e-01 -2.46771023e-01 -7.96047211e-01
-5.54251313e-01 2.06148610e-01 -4.07977670e-01 9.01094377e-01
3.14582139e-02 -3.95203352e-01 6.05600655e-01 5.02219558e-01
-1.58256292e-01 -7.73698390e-01 -8.05102229e-01 -6.79943919e-01
-5.41752577e-01 -5.76663613e-01 6.33589566e-01 5.21211326e-01
-6.21502817e-01 3.61498594e-01 -2.24853367e-01 5.37924290e-01
7.75329053e-01 6.22864723e-01 6.40398145e-01 -1.40709794e+00
-3.08904111e-01 -5.67444026e-01 -5.97221434e-01 -1.10528266e+00
1.85098439e-01 -6.50556505e-01 2.69851208e-01 -1.47538066e+00
4.70994383e-01 -2.54256278e-01 -4.29249734e-01 8.11071396e-01
-1.43225089e-01 6.04466617e-01 4.95479733e-01 2.87775815e-01
-9.47292387e-01 3.24069649e-01 1.10058045e+00 -4.24674273e-01
4.64422882e-01 -2.14264944e-01 -7.76030183e-01 9.15159702e-01
3.68575126e-01 -2.01736227e-01 -1.05172925e-01 -7.04979718e-01
-2.43083328e-01 -1.71219990e-01 9.86450553e-01 -9.84291136e-01
2.59123296e-01 -4.80297282e-02 7.89357126e-01 -8.29269826e-01
5.00317335e-01 -1.18756616e+00 1.02865890e-01 5.77649295e-01
-4.27486658e-01 -1.26716673e-01 3.55783820e-01 7.18344212e-01
-2.93949574e-01 8.44041780e-02 8.43578696e-01 -8.40230510e-02
-1.03344893e+00 4.22743767e-01 -1.67465836e-01 -2.87996560e-01
1.11202347e+00 -3.02499413e-01 -4.97458220e-01 1.34644598e-01
-6.60277069e-01 -1.26040382e-02 3.79302949e-01 4.87254798e-01
6.48059130e-01 -1.17123628e+00 -2.97612071e-01 3.37430447e-01
5.32959938e-01 2.39709690e-02 4.76682872e-01 4.67456669e-01
-4.70259994e-01 8.29014301e-01 -2.85649091e-01 -8.43341410e-01
-1.25537729e+00 7.26621926e-01 4.04054463e-01 -1.65148973e-02
-4.01403993e-01 8.63346279e-01 6.83786333e-01 -1.82453886e-01
2.73966819e-01 -8.07320058e-01 6.18616864e-02 -1.52804807e-01
7.79458642e-01 2.21931994e-01 1.26474619e-01 -6.77133501e-01
-6.22830153e-01 5.40849447e-01 -1.20655708e-01 3.98153841e-01
1.11920881e+00 -1.71729177e-01 -1.38716802e-01 3.90365332e-01
1.00291657e+00 -4.65678960e-01 -1.60917318e+00 -2.43468791e-01
1.11562803e-01 -5.43993652e-01 -7.06615970e-02 -9.93932366e-01
-9.27674592e-01 1.19476581e+00 5.98862112e-01 1.27974197e-01
1.16023707e+00 1.70334265e-01 2.44519189e-01 4.90601301e-01
3.28869164e-01 -4.89762694e-01 7.28080124e-02 5.81661224e-01
8.65222156e-01 -1.40000188e+00 -1.78197205e-01 -5.20338535e-01
-4.09134984e-01 1.14763403e+00 9.33463991e-01 -7.44863972e-02
4.44908261e-01 6.21480346e-02 -7.76352882e-02 -1.66355237e-01
-9.62520123e-01 -1.98607579e-01 6.99804664e-01 2.89315790e-01
2.31591165e-01 1.67833343e-01 2.12275386e-01 8.18126738e-01
2.99603462e-01 1.08457446e-01 1.76628858e-01 7.50991404e-01
-5.96527934e-01 -4.75617170e-01 -2.19576970e-01 4.26667273e-01
-2.73066550e-01 -1.87989220e-01 -4.16902781e-01 6.83458269e-01
4.56873000e-01 8.10727119e-01 4.79542643e-01 -1.84367806e-01
1.96668923e-01 9.07993615e-02 3.95013630e-01 -4.76927012e-01
-6.43635213e-01 -2.61524230e-01 -1.05398960e-01 -8.05997193e-01
-4.09772009e-01 -1.31773099e-01 -1.13388169e+00 -3.21810544e-01
-2.62936473e-01 -5.13143800e-02 6.93374038e-01 1.24143112e+00
6.68013573e-01 8.44217598e-01 3.64266068e-01 -1.10861063e+00
-5.73777556e-01 -9.68460023e-01 -4.18096185e-01 3.17817152e-01
6.45284474e-01 -6.71337306e-01 -6.98894784e-02 2.70359010e-01] | [9.564166069030762, 1.5992463827133179] |
46655ffb-cdf9-461b-8286-3e814c02ce48 | better-exploiting-motion-for-better-action | null | null | http://openaccess.thecvf.com/content_cvpr_2013/html/Jain_Better_Exploiting_Motion_2013_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2013/papers/Jain_Better_Exploiting_Motion_2013_CVPR_paper.pdf | Better Exploiting Motion for Better Action Recognition | Several recent works on action recognition have attested the importance of explicitly integrating motion characteristics in the video description. This paper establishes that adequately decomposing visual motion into dominant and residual motions, both in the extraction of the space-time trajectories and for the computation of descriptors, significantly improves action recognition algorithms. Then, we design a new motion descriptor, the DCS descriptor, based on differential motion scalar quantities, divergence, curl and shear features. It captures additional information on the local motion patterns enhancing results. Finally, applying the recent VLAD coding technique proposed in image retrieval provides a substantial improvement for action recognition. Our three contributions are complementary and lead to outperform all reported results by a significant margin on three challenging datasets, namely Hollywood 2, HMDB51 and Olympic Sports. | ['Herve Jegou', 'Patrick Bouthemy', 'Mihir Jain'] | 2013-06-01 | null | null | null | cvpr-2013-6 | ['video-description'] | ['computer-vision'] | [ 2.59497285e-01 -6.83382392e-01 -6.66549861e-01 -3.40791419e-02
-6.26998782e-01 -6.78326070e-01 9.73539531e-01 7.09115155e-03
-4.52560812e-01 5.46349525e-01 6.88245714e-01 2.83252656e-01
-3.83087844e-01 -2.60227233e-01 -6.32638857e-02 -9.72748399e-01
-4.18740720e-01 -1.37241870e-01 4.18747097e-01 -7.83400685e-02
6.26459837e-01 7.78257430e-01 -1.69011986e+00 4.47910666e-01
1.92844033e-01 1.13358700e+00 -2.60326982e-01 1.03584671e+00
2.81947106e-01 1.41208541e+00 -3.09502155e-01 -1.35438621e-01
3.29954982e-01 -5.89320719e-01 -1.04890108e+00 4.25090909e-01
6.58876717e-01 -4.83110458e-01 -7.22247481e-01 5.53074062e-01
3.74505311e-01 4.84700292e-01 7.68888950e-01 -1.15751469e+00
-4.70015943e-01 -1.52420849e-01 -4.00587857e-01 7.75191247e-01
8.43485713e-01 6.43399730e-02 1.20446467e+00 -6.93687081e-01
9.93732631e-01 1.17104924e+00 5.60063064e-01 3.21784347e-01
-9.55565751e-01 5.48793599e-02 -1.12269558e-02 8.99530470e-01
-1.30142021e+00 -5.54399729e-01 6.90373242e-01 -7.48179853e-01
1.10348105e+00 5.08028388e-01 8.61035466e-01 9.99601543e-01
2.42890134e-01 1.20422018e+00 8.11255753e-01 -4.50827330e-01
1.78215250e-01 -6.80987298e-01 1.28899261e-01 4.75151986e-01
-7.90962651e-02 1.96759135e-01 -8.27755868e-01 -7.90258124e-02
7.36618459e-01 -1.26216114e-01 -2.90589750e-01 -7.80054808e-01
-1.32918239e+00 6.59617305e-01 -1.63925529e-01 3.72290403e-01
-4.60723728e-01 3.61396015e-01 7.08439291e-01 1.88417092e-01
2.54156083e-01 -1.08774006e-01 2.93773636e-02 -1.08491743e+00
-9.07173872e-01 3.70924711e-01 5.06125987e-01 6.02738082e-01
2.10072845e-01 5.38500100e-02 -4.64434206e-01 5.86324096e-01
1.63004324e-01 5.02488256e-01 5.83602369e-01 -1.43780363e+00
3.36902261e-01 5.33537745e-01 2.16649175e-01 -1.44972730e+00
-2.33927861e-01 2.27679744e-01 -6.94916189e-01 1.93593457e-01
3.39072704e-01 4.41466957e-01 -6.61421657e-01 1.41736484e+00
3.60690087e-01 3.73034239e-01 9.21066999e-02 9.96525645e-01
4.79209989e-01 2.41009519e-01 6.90258071e-02 -1.56678334e-01
1.18581200e+00 -9.46521759e-01 -8.61072898e-01 2.96477884e-01
7.63592839e-01 -7.57562041e-01 5.16379714e-01 4.98664320e-01
-1.15747046e+00 -7.55296767e-01 -1.14633358e+00 -4.22461256e-02
-1.86202541e-01 4.48710829e-01 6.68156743e-01 5.29262483e-01
-9.26185071e-01 8.68571103e-01 -1.19821155e+00 -3.90139133e-01
2.93402195e-01 3.05584162e-01 -7.10759938e-01 8.44655372e-03
-8.80270123e-01 7.23284245e-01 1.71272621e-01 -1.68561079e-02
-6.74874306e-01 -2.29259655e-01 -9.07344043e-01 -2.63191521e-01
7.88207725e-02 -3.06014717e-01 9.60392773e-01 -8.36945176e-01
-1.48265982e+00 7.54540682e-01 -1.95796043e-01 -5.66784501e-01
6.26206040e-01 -2.98646420e-01 -5.70236623e-01 8.74971151e-01
-8.22267607e-02 5.60323894e-01 8.53603303e-01 -5.56814909e-01
-7.05421865e-01 -3.44903797e-01 1.17915258e-01 3.32439572e-01
-2.24444538e-01 -9.82609391e-02 -4.29912239e-01 -9.20515120e-01
5.72312474e-02 -1.05879664e+00 3.70322727e-02 5.62533066e-02
6.17032610e-02 -1.92280501e-01 8.99709702e-01 -6.63142860e-01
1.51301777e+00 -2.39284253e+00 6.29751921e-01 1.60414036e-02
-3.97339538e-02 4.69019592e-01 -1.81973264e-01 5.89368463e-01
-1.47202641e-01 -2.81423807e-01 1.02394119e-01 -6.79290295e-02
1.87582001e-02 4.15150821e-01 -2.62174189e-01 8.17123532e-01
2.32578635e-01 8.85251939e-01 -7.80227542e-01 -5.50440669e-01
6.02988183e-01 4.64187473e-01 -4.77115691e-01 -1.21989451e-01
4.55534518e-01 5.40036798e-01 -5.02023220e-01 5.98678648e-01
3.84582728e-01 -2.60381331e-03 1.07629910e-01 -1.76504955e-01
-6.42716661e-02 7.89506733e-02 -1.46903729e+00 1.81477249e+00
1.48198947e-01 8.82878423e-01 -3.33799362e-01 -1.02037692e+00
5.92886984e-01 2.77112961e-01 1.17142963e+00 -7.65823662e-01
1.29783778e-02 -1.50777295e-01 -3.54332507e-01 -7.12863624e-01
5.98813355e-01 1.85016364e-01 2.81544346e-02 1.96348816e-01
3.40495855e-02 4.94301170e-01 5.59075236e-01 7.26735070e-02
1.34814739e+00 4.59892422e-01 4.94807750e-01 -1.30031556e-01
1.10311747e+00 5.20267375e-02 4.24932778e-01 4.56749678e-01
-7.63962686e-01 7.09389746e-01 4.96078104e-01 -5.55133224e-01
-8.21605146e-01 -9.88253832e-01 -3.58597226e-02 8.22036266e-01
1.60912737e-01 -7.81672359e-01 -4.59320456e-01 -5.32990396e-01
2.26329602e-02 8.26436654e-02 -7.36406088e-01 -2.69594222e-01
-7.92270362e-01 -3.66816908e-01 6.22722864e-01 9.70826089e-01
5.93629003e-01 -8.37678373e-01 -8.02047133e-01 4.92732376e-02
-3.17589790e-01 -1.30051804e+00 -7.12674618e-01 -3.42320442e-01
-9.43354726e-01 -1.24752927e+00 -7.94205368e-01 -4.65880215e-01
1.18947901e-01 5.09166658e-01 6.87066615e-01 -1.04334265e-01
-6.67771637e-01 9.63425159e-01 -6.57319486e-01 4.31314200e-01
-6.46172985e-02 -3.78066719e-01 5.49075603e-02 2.67333746e-01
6.08334839e-01 -4.13820118e-01 -6.81749880e-01 3.65086794e-01
-9.02182043e-01 -3.29662502e-01 3.75030756e-01 7.35021114e-01
5.48715651e-01 -2.84791827e-01 -1.58075422e-01 -2.41753459e-02
1.90535605e-01 -1.64272532e-01 -1.21867426e-01 1.12965748e-01
-2.04832286e-01 1.61643669e-01 8.41655582e-02 -4.93521333e-01
-7.78433025e-01 2.61836231e-01 6.57177716e-03 -6.83988750e-01
-2.37409949e-01 1.27048001e-01 1.61242858e-01 -2.47872949e-01
3.05984527e-01 5.76476753e-01 -1.11726252e-02 -4.88968909e-01
3.83074760e-01 4.32205200e-01 5.45063853e-01 -2.37896934e-01
4.69794750e-01 9.98719990e-01 4.20397371e-01 -1.21953917e+00
-3.21252823e-01 -1.15833259e+00 -1.32515872e+00 -2.92201400e-01
1.27181876e+00 -8.39922965e-01 -9.88708913e-01 6.46257222e-01
-9.07250464e-01 8.72011483e-02 -5.05023360e-01 8.47607672e-01
-1.10908639e+00 1.01874232e+00 -6.02564335e-01 -7.45129108e-01
9.84244347e-02 -9.67479527e-01 1.25273848e+00 -2.31002703e-01
-4.11534250e-01 -1.14644945e+00 5.98183334e-01 4.26661879e-01
1.07445754e-02 4.82328147e-01 5.52391887e-01 -3.97109419e-01
-6.29054904e-01 -2.35505030e-01 9.75062922e-02 5.08603275e-01
1.23238862e-01 1.86555177e-01 -5.60053229e-01 -2.03066394e-01
-1.85803801e-01 -2.14193746e-01 1.06184387e+00 4.46340770e-01
7.00761080e-01 3.14414650e-02 -9.56543460e-02 6.35568202e-01
1.23953581e+00 4.12862211e-01 9.41703677e-01 3.83152932e-01
6.76446021e-01 3.34929436e-01 7.75291562e-01 8.08667541e-01
8.40298533e-02 1.18260312e+00 5.34716696e-02 3.33938718e-01
-3.37641478e-01 1.09180529e-03 7.73364007e-01 8.34320247e-01
-8.04818094e-01 7.08994791e-02 -6.73584402e-01 5.27361631e-01
-2.05927563e+00 -1.48924589e+00 -3.17061007e-01 2.08498621e+00
2.96851873e-01 -1.67368159e-01 4.24970776e-01 4.03025210e-01
2.90377408e-01 6.21807098e-01 -2.14527279e-01 -3.33515972e-01
-2.66975194e-01 1.61135614e-01 4.84555572e-01 4.70570654e-01
-1.59881806e+00 7.41189420e-01 7.17847538e+00 8.87782812e-01
-8.42072904e-01 -6.85758591e-02 -2.24575121e-03 -1.05091266e-01
3.76261532e-01 -9.05075073e-02 -6.33101463e-01 2.33944982e-01
7.79218316e-01 -1.04320683e-01 4.25262339e-02 7.57790148e-01
3.04751009e-01 -2.53703088e-01 -1.11446631e+00 1.25457025e+00
4.35328305e-01 -1.37001622e+00 2.31813133e-01 2.03071237e-01
4.55419213e-01 -3.53769630e-01 7.92788342e-02 -7.90449828e-02
-1.64808288e-01 -8.96723747e-01 7.30272174e-01 8.06465745e-01
3.65000486e-01 -6.64376020e-01 5.88856280e-01 -1.35497078e-01
-1.65453291e+00 -1.10745862e-01 -2.49560550e-01 -2.17689350e-01
1.78943172e-01 -1.26774102e-01 -2.01425597e-01 6.80690765e-01
3.75419974e-01 1.46753430e+00 -7.10232496e-01 9.65390563e-01
1.06221899e-01 3.48193944e-01 2.00060636e-01 2.91456699e-01
4.92301494e-01 -3.09826553e-01 7.32311010e-01 1.25981927e+00
4.04056758e-02 1.97988883e-01 1.99346289e-01 1.43254846e-01
7.30785787e-01 1.87687442e-01 -6.45356417e-01 -3.22127879e-01
-3.14026296e-01 9.70833957e-01 -6.56431139e-01 -2.93008447e-01
-5.63526511e-01 1.32930231e+00 -5.46370447e-02 3.42227608e-01
-7.35277891e-01 -7.95847774e-02 1.21331882e+00 -1.56578213e-01
7.19708025e-01 -8.33115101e-01 2.28550926e-01 -1.27921796e+00
2.44954243e-01 -8.38083625e-01 5.58979154e-01 -4.56697494e-01
-8.78795266e-01 1.70734406e-01 1.92438722e-01 -1.77601242e+00
-5.18585622e-01 -8.80501926e-01 -2.29574889e-01 3.41956854e-01
-1.10835397e+00 -9.72081661e-01 -2.34339193e-01 8.44205797e-01
6.15839481e-01 -2.25452393e-01 7.04650819e-01 5.54680705e-01
-3.99095923e-01 2.41029724e-01 2.76191950e-01 3.11280936e-01
5.15014648e-01 -1.10148752e+00 1.15565121e-01 6.81647122e-01
4.97653097e-01 2.20103592e-01 4.25772369e-01 -4.04724568e-01
-1.57048833e+00 -6.34283960e-01 7.72431493e-01 -8.45028996e-01
7.89781988e-01 2.17264965e-02 -5.95140398e-01 5.02666116e-01
4.58293520e-02 5.02209999e-02 6.72464430e-01 -4.36110288e-01
-5.12311421e-03 1.73251498e-02 -6.84864819e-01 4.07367468e-01
1.23352289e+00 -6.70047224e-01 -5.96245944e-01 1.81322396e-01
3.60179693e-02 -2.00410172e-01 -1.06602573e+00 2.72162348e-01
8.28912616e-01 -1.15936852e+00 1.36753559e+00 -1.12290776e+00
3.75275284e-01 -2.03581482e-01 -4.57463652e-01 -6.88069105e-01
-5.04945993e-01 -5.75102031e-01 -3.93206358e-01 8.95906687e-01
-3.19578916e-01 -1.64526131e-03 8.24345589e-01 3.10746700e-01
1.24305934e-01 -5.91039777e-01 -1.30863655e+00 -9.92826402e-01
-1.86188102e-01 -5.01919568e-01 -6.34336546e-02 7.17356384e-01
1.63450420e-01 -9.12209079e-02 -6.38330579e-01 -3.02388608e-01
4.18302000e-01 -1.68685615e-02 6.90033078e-01 -9.11034286e-01
-2.53784239e-01 -4.30170447e-01 -1.59062040e+00 -1.38461483e+00
1.32798105e-01 -6.59533918e-01 -3.00666481e-01 -1.17854214e+00
1.78256184e-01 3.34460288e-01 -3.19468707e-01 2.01558709e-01
1.37945816e-01 4.28646505e-01 3.43567610e-01 4.21013504e-01
-8.79185557e-01 6.97799563e-01 1.33293653e+00 -1.84708923e-01
1.22558251e-01 -5.03268130e-02 1.00405931e-01 9.00754213e-01
2.57199705e-01 -5.27336597e-02 -3.56184810e-01 -9.33913738e-02
-2.33379826e-01 1.09384574e-01 5.79675913e-01 -1.26857662e+00
9.00427103e-02 -2.65750706e-01 4.12875772e-01 -5.77540815e-01
6.34018838e-01 -6.37252748e-01 -5.93799986e-02 6.32575095e-01
-5.06051660e-01 3.10421556e-01 -1.70233231e-02 8.54447961e-01
-5.75722694e-01 -6.90240338e-02 6.20512486e-01 5.95150664e-02
-1.49652946e+00 2.07011133e-01 -6.33655548e-01 1.54035896e-01
1.32590270e+00 -5.91783464e-01 -2.38201227e-02 -4.31451917e-01
-1.14677548e+00 -2.53763676e-01 3.16840112e-01 6.37293994e-01
6.32174313e-01 -1.90810728e+00 -5.00834167e-01 2.55464405e-01
1.01833478e-01 -1.07092810e+00 4.30507243e-01 1.38911569e+00
-7.35885799e-01 8.04069936e-01 -5.65080345e-01 -7.68122196e-01
-1.81420088e+00 4.79372829e-01 1.55078083e-01 -2.84839958e-01
-7.68207252e-01 4.41042662e-01 -6.60361648e-02 4.53890175e-01
3.39661121e-01 -4.28270370e-01 -4.55807924e-01 3.47235918e-01
7.55056858e-01 8.62542391e-01 -8.18629861e-02 -1.29100943e+00
-7.22064555e-01 9.80504513e-01 2.42310300e-01 -1.48460597e-01
1.10710955e+00 -2.69332021e-01 1.91160873e-01 5.11096478e-01
1.42895007e+00 -1.59193724e-01 -1.30024314e+00 4.30637691e-03
3.53405923e-01 -8.92663062e-01 -9.64445174e-02 -2.21719638e-01
-9.27026451e-01 9.26242054e-01 8.39311540e-01 1.56996220e-01
1.18977642e+00 -1.95002019e-01 6.71501040e-01 3.71700346e-01
2.54343510e-01 -1.25748301e+00 5.62867284e-01 5.58210194e-01
7.74937034e-01 -1.05547750e+00 2.58708209e-01 -3.23136419e-01
-9.54813778e-01 1.28319144e+00 1.60187736e-01 -3.25339526e-01
6.21059954e-01 2.25805901e-02 -3.38106193e-02 -1.26568258e-01
-6.94134295e-01 -4.04811144e-01 7.36324549e-01 6.27326667e-01
4.37015980e-01 -2.19289452e-01 -6.19246542e-01 8.98252651e-02
3.59655619e-01 4.63282317e-02 2.53820032e-01 1.29956090e+00
-3.22379291e-01 -1.23814082e+00 -1.58254460e-01 -1.41086906e-01
-4.78412420e-01 2.45048180e-01 -6.11898065e-01 9.17692363e-01
4.44334112e-02 7.04317570e-01 1.35540053e-01 -4.23066020e-01
3.24121177e-01 1.19004704e-01 7.67471790e-01 -1.11887231e-01
-1.90583795e-01 -1.29997255e-02 2.21580043e-01 -1.21612358e+00
-1.13561511e+00 -1.00083351e+00 -1.06734169e+00 -2.22837195e-01
2.27964655e-01 -5.87178618e-02 1.98984742e-01 1.02980912e+00
4.18447167e-01 2.09132254e-01 5.09156585e-01 -9.01365638e-01
-6.42855823e-01 -6.68458581e-01 -7.89281428e-01 9.60261524e-01
3.95901054e-01 -9.87273395e-01 -2.75556684e-01 5.44248939e-01] | [8.040848731994629, 0.3208381235599518] |
1c1d6da4-4caf-4f40-9557-278c70b73cee | star-ris-enabled-simultaneous-indoor-and | 2302.03342 | null | https://arxiv.org/abs/2302.03342v1 | https://arxiv.org/pdf/2302.03342v1.pdf | STAR-RIS-Enabled Simultaneous Indoor and Outdoor 3D Localization: Theoretical Analysis and Algorithmic Design | Recent research and development interests deal with metasurfaces for wireless systems beyond their consideration as intelligent tunable reflectors. Among the latest proposals is the simultaneously transmitting (a.k.a. refracting) and reflecting reconfigurable intelligent surface (STAR-RIS) which intends to enable bidirectional indoor-to-outdoor, and vice versa, communications thanks to its additional refraction capability. This double functionality provides increased flexibility in concurrently satisfying the quality-of-service requirements of users located at both sides of the metasurfaces, for example, the achievable data rate and localization accuracy. In this paper, we focus on STAR-RIS-empowered simultaneous indoor and outdoor three-dimensional (3D) localization, and study the fundamental performance limits via Fisher information analyses and Cram\'er Rao lower bounds (CRLBs). We also devise an efficient localization algorithm based on an off-grid compressive sensing (CS) technique relying on atomic norm minimization (ANM). The impact of the training overhead, the power splitting at the STAR-RIS, the power allocation between the users, the STAR-RIS size, the imperfections of the STAR-RIS-to-BS channel, as well as the role of the multi-path components on the positioning performance are assessed via extensive computer simulations. It is theoretically showcased that high-accuracy, up to centimeter level, 3D localization can be simultaneously achieved for indoor and outdoor users, which is also accomplished via the proposed ANM-based estimation algorithm. | ['George C. Alexandropoulos', 'Aymen Fakhreddine', 'Jiguang He'] | 2023-02-07 | null | null | null | null | ['compressive-sensing'] | ['computer-vision'] | [ 4.22545135e-01 2.84421682e-01 4.37115014e-01 1.52820900e-01
-8.79421353e-01 -4.19693381e-01 2.71861643e-01 -1.63342357e-01
-8.66950527e-02 6.04545176e-01 5.83621711e-02 -4.08149987e-01
-7.90862918e-01 -8.97187591e-01 -5.89748919e-01 -1.24523151e+00
-4.74453688e-01 8.77538696e-02 -1.75418735e-01 -2.93496907e-01
1.27206475e-01 6.40570402e-01 -1.55311263e+00 -3.85572344e-01
1.04626203e+00 1.24935722e+00 2.78313428e-01 4.46965456e-01
3.07030886e-01 -1.51375145e-01 -4.19745535e-01 8.84137452e-02
1.68551236e-01 1.53311649e-02 4.17667091e-01 -1.35322556e-01
2.16179364e-03 -2.06377968e-01 -1.68575555e-01 6.16406202e-01
6.51094258e-01 -4.60456088e-02 7.41379142e-01 -8.12521398e-01
-2.44917348e-01 3.21997017e-01 -6.23414814e-01 -2.40094587e-01
3.83032650e-01 -4.49693918e-01 3.67614537e-01 -8.51483762e-01
3.33582312e-02 6.05093718e-01 7.42268622e-01 -6.35738205e-03
-7.13483691e-01 -6.12411022e-01 -5.86051583e-01 -3.71523857e-01
-1.78678000e+00 -7.57774711e-01 6.91390634e-01 -2.07063794e-01
2.91566998e-01 7.11283803e-01 5.22836268e-01 5.11439860e-01
2.34783590e-01 -1.64843649e-01 9.73383665e-01 -8.53601456e-01
4.41146612e-01 1.43905342e-01 -4.68350857e-01 7.91608393e-01
8.96082640e-01 8.95148665e-02 -4.81571674e-01 -2.28438407e-01
7.92270124e-01 -3.60003799e-01 -5.06501913e-01 -3.70771408e-01
-9.80397522e-01 2.18571350e-01 4.88880485e-01 8.06126714e-01
-4.31635559e-01 2.62658805e-01 -5.95166087e-01 1.97913453e-01
3.73522162e-01 3.04158539e-01 -7.33981729e-02 2.93302685e-01
-1.04485703e+00 -2.88221657e-01 6.85642660e-01 1.12242377e+00
3.74805033e-01 1.65324077e-01 -7.26583302e-02 4.50385273e-01
9.13998246e-01 1.32210672e+00 -7.91961923e-02 -6.43233180e-01
5.57675421e-01 9.51070562e-02 6.27216697e-01 -1.07965684e+00
-4.98383284e-01 -1.34043372e+00 -9.51433897e-01 -2.10175708e-01
6.68949783e-02 -5.25366247e-01 -4.29112971e-01 1.49841428e+00
5.02430677e-01 1.90414310e-01 3.89673501e-01 6.91171348e-01
5.20088792e-01 6.83084786e-01 -6.67235017e-01 -5.57126701e-01
1.11637402e+00 -1.43962666e-01 -3.82375121e-01 -4.94461544e-02
5.80104232e-01 -5.48372984e-01 3.85645092e-01 3.96189630e-01
-1.13023078e+00 -3.62707339e-02 -1.19417393e+00 5.69773734e-01
9.92634073e-02 5.90781808e-01 1.70563698e-01 1.31230450e+00
-1.12516379e+00 9.88055468e-02 -7.24871039e-01 -8.12172070e-02
1.77477449e-01 3.47287476e-01 -7.11006969e-02 -9.22512561e-02
-7.70219564e-01 4.37612355e-01 -6.09268486e-01 4.53860939e-01
-3.20148379e-01 -7.66850710e-01 -3.82577479e-01 1.42999977e-01
1.36127517e-01 -7.33361959e-01 5.33602417e-01 -4.58774000e-01
-1.78539956e+00 2.41833165e-01 -3.91961008e-01 -2.25852489e-01
1.80577219e-01 1.70544237e-02 -6.40553415e-01 2.90703624e-01
5.26376702e-02 -4.99373525e-01 4.27550226e-01 -1.43808222e+00
-4.53334928e-01 -6.59472942e-01 -3.22053194e-01 2.40649834e-01
-3.32311988e-01 -4.57041353e-01 3.76513302e-02 -3.56895775e-01
8.01519096e-01 -1.13974094e+00 -3.23336750e-01 -1.61621392e-01
-4.52116311e-01 2.78069526e-01 5.83993256e-01 -1.68324962e-01
8.54170084e-01 -2.36962938e+00 -9.98822972e-02 8.42450380e-01
-1.76532671e-01 -2.32269596e-02 -8.48252401e-02 7.42714226e-01
2.60609925e-01 -2.81657428e-01 -8.15959647e-02 -3.16345990e-01
-2.33758390e-01 -2.92841047e-01 3.99538130e-02 1.10083485e+00
-6.11212790e-01 2.97226578e-01 -6.65518880e-01 1.07338928e-01
1.21653423e-01 6.48193657e-01 -5.46701789e-01 2.42676549e-02
3.68649125e-01 8.66073966e-01 -9.03414786e-01 6.82272315e-01
1.00013494e+00 2.48803254e-02 1.24005631e-01 -2.01791510e-01
-9.19657946e-01 -1.44446254e-01 -1.30138695e+00 1.46362126e+00
-9.16922867e-01 2.90398806e-01 4.84522223e-01 -8.51135433e-01
1.13001287e+00 3.76694381e-01 6.70176148e-01 -9.87641811e-01
-1.40076414e-01 8.38935971e-01 -3.13584238e-01 -3.61810207e-01
4.22569662e-01 -5.08895814e-02 -1.08931325e-02 1.46189854e-01
-3.08440030e-01 1.05736278e-01 -4.06524271e-01 -5.62634580e-02
1.03829050e+00 -1.31782725e-01 8.20137635e-02 -9.49934185e-01
7.84969449e-01 -4.04542863e-01 9.09122154e-02 7.23790824e-01
6.68614209e-01 1.92824081e-01 -2.00634345e-01 3.18488657e-01
-4.87093776e-01 -1.02712464e+00 -3.90078336e-01 3.82071704e-01
7.77949929e-01 1.06040753e-01 -3.31886828e-01 2.87569851e-01
1.91298082e-01 8.96970451e-01 -2.12487191e-01 -5.65928742e-02
-2.70861626e-01 -7.39493966e-01 4.62694913e-01 -3.36512387e-01
5.66051662e-01 1.96109742e-01 -8.89754891e-01 7.49579892e-02
-4.77927141e-02 -1.06061852e+00 3.29056978e-01 -9.74727646e-02
-6.63625062e-01 -6.79841280e-01 -7.53734827e-01 -3.17899674e-01
6.75977767e-01 8.48178744e-01 4.26076770e-01 -1.30737051e-01
5.84374964e-02 9.75750029e-01 -3.49566281e-01 -2.30063871e-01
3.61528508e-02 -1.28189743e-01 3.09052169e-01 3.10328901e-01
-6.38646483e-01 -9.52917278e-01 -9.78087604e-01 7.61916339e-01
-2.31332541e-01 -3.84851731e-02 6.69797480e-01 1.29272759e-01
4.96362090e-01 1.08787261e-01 6.64234042e-01 -2.23885238e-01
1.32872611e-01 -8.62587869e-01 -8.47354650e-01 2.35710740e-01
-4.10586566e-01 -2.12436393e-01 4.31957006e-01 1.40919626e-01
-1.09898031e+00 -3.28049004e-01 -2.34502167e-01 3.25395465e-01
1.62425682e-01 5.36323845e-01 -3.63759249e-01 -8.65504742e-01
5.08128703e-01 3.20146948e-01 -3.04372519e-01 -1.23373315e-01
2.38475710e-01 9.22284365e-01 1.54427394e-01 -3.33829850e-01
1.19579303e+00 8.41853917e-01 7.37815499e-01 -1.66201365e+00
-5.58609247e-01 -4.34169382e-01 5.03083318e-02 -3.59393269e-01
3.07485551e-01 -1.06596148e+00 -7.89746821e-01 1.79753736e-01
-7.61608362e-01 -1.70634612e-01 -1.82803255e-02 8.60751390e-01
-3.55781823e-01 1.75585791e-01 1.84489459e-01 -1.43500364e+00
-4.29443151e-01 -6.54595912e-01 1.05666602e+00 3.29435974e-01
3.05120587e-01 -7.49692202e-01 -2.68769283e-02 3.75213027e-01
1.10198665e+00 5.51330626e-01 5.02836227e-01 4.83810604e-02
-7.49304473e-01 -3.47463071e-01 -4.28511165e-02 -4.46832627e-01
1.03626348e-01 -6.48578584e-01 -7.32426584e-01 -4.97355312e-01
1.89388707e-01 4.63126808e-01 3.11973155e-01 7.21752405e-01
8.06224525e-01 -4.28099096e-01 -7.00469375e-01 8.06728601e-01
1.66760647e+00 -1.38408199e-01 7.64433086e-01 -2.03717932e-01
1.94711238e-01 3.38716149e-01 7.22640514e-01 7.83064306e-01
4.02560085e-01 8.42184484e-01 7.65705228e-01 5.64910062e-02
1.36053115e-01 1.98545128e-01 2.92739719e-01 7.44772613e-01
-2.97508061e-01 -8.35213542e-01 -3.06589037e-01 2.93462843e-01
-1.22872436e+00 -6.40336931e-01 -5.82418025e-01 2.64722490e+00
2.72722095e-02 -4.78199691e-01 -4.11106586e-01 5.82409464e-02
5.29957056e-01 1.87267493e-02 -1.74198687e-01 1.82404190e-01
-1.44469082e-01 1.50012672e-01 1.02948642e+00 5.10308921e-01
-4.18442458e-01 -5.15481159e-02 4.48975563e+00 8.06860685e-01
-1.00667238e+00 2.74916768e-01 -2.67829280e-03 -7.69719407e-02
-1.05775952e+00 -1.05716176e-01 -6.85711443e-01 3.29364836e-01
8.64307940e-01 1.27025947e-01 2.68376291e-01 2.36318499e-01
5.00320613e-01 -5.85988224e-01 -3.44092429e-01 9.26344395e-01
6.21503741e-02 -1.17900407e+00 -2.46720314e-01 5.16134918e-01
7.37111390e-01 7.88923949e-02 2.07189709e-01 -3.33589137e-01
-4.83493239e-01 -3.47579569e-01 7.55770504e-01 7.03786433e-01
1.05046391e+00 -5.70315778e-01 4.68833238e-01 5.29408514e-01
-1.30244040e+00 -3.27865481e-01 -3.54413316e-02 2.61268228e-01
2.28943795e-01 1.31858528e+00 -4.23314422e-01 1.15884328e+00
3.44524950e-01 7.57290050e-02 9.51295793e-02 9.84652877e-01
3.12317461e-02 4.77899671e-01 -7.82005548e-01 -4.23896670e-01
1.04355820e-01 -5.19135296e-01 1.00789213e+00 7.45576978e-01
1.51219845e+00 5.54429352e-01 -4.49444085e-01 7.11785793e-01
1.54378012e-01 4.35216315e-02 -5.42674601e-01 4.06241387e-01
8.11735332e-01 1.30663776e+00 -5.49013138e-01 3.10180962e-01
-1.94136590e-01 4.82945740e-01 -5.64055979e-01 5.55187583e-01
-6.76393688e-01 -4.60995734e-01 4.97065574e-01 7.71841168e-01
3.76781166e-01 -8.83674681e-01 -2.68484086e-01 -8.22129071e-01
6.34486750e-02 -7.76356682e-02 -1.83344007e-01 -4.81892973e-01
-5.92269361e-01 -4.26695310e-02 -2.88187116e-01 -1.31361639e+00
3.39058936e-01 -1.08981661e-01 -3.16001326e-01 7.13232219e-01
-1.74430132e+00 -1.04867601e+00 -4.59776819e-01 4.71279860e-01
-2.58794725e-01 1.29238404e-02 6.70898616e-01 7.25115061e-01
-2.11004496e-01 4.58764762e-01 1.04944134e+00 -5.14828682e-01
8.98085088e-02 -5.22882462e-01 -3.48247796e-01 7.76123822e-01
4.63383310e-02 7.08894670e-01 7.32839227e-01 -4.17512149e-01
-2.16218853e+00 -9.37092185e-01 2.91860372e-01 1.40698150e-01
2.48911798e-01 -3.76016349e-01 7.06323832e-02 2.29472108e-02
-2.27565188e-02 -1.01209268e-01 9.45448995e-01 -2.01253250e-01
1.59132168e-01 -4.08340931e-01 -1.42686915e+00 5.39143860e-01
1.13897383e+00 -1.18044680e-02 2.72823036e-01 4.15675014e-01
2.00226784e-01 -5.50375223e-01 -7.86648691e-01 6.42906308e-01
7.61632025e-01 -8.96171570e-01 1.02736378e+00 4.85077202e-01
-1.11414909e-01 -4.19107765e-01 -7.24685609e-01 -1.08501852e+00
-2.04139709e-01 -7.23035455e-01 5.74753694e-02 1.16276717e+00
3.92652541e-01 -1.05279946e+00 6.74636662e-01 1.16570756e-01
-3.33177447e-01 -5.24526834e-01 -1.50401747e+00 -8.67525101e-01
-4.56208736e-01 -2.23418593e-01 4.43523824e-01 5.95380545e-01
-2.10117579e-01 1.18700422e-01 -2.30170533e-01 1.17134619e+00
9.73219693e-01 2.13020504e-01 6.70054734e-01 -1.32064402e+00
-3.46274018e-01 2.54536960e-02 -8.87726173e-02 -1.24717557e+00
-4.89339650e-01 -6.98459744e-01 2.83573847e-02 -1.65114534e+00
-5.55190802e-01 -1.26899290e+00 -3.31455059e-02 -1.64129466e-01
6.95091069e-01 3.30287755e-01 -3.45234007e-01 -1.72073301e-02
-4.93923157e-01 7.61921823e-01 9.67251718e-01 3.26892167e-01
-3.68360311e-01 6.70486987e-01 -4.80421126e-01 3.29430491e-01
6.30155981e-01 -3.71323019e-01 -2.85479933e-01 -6.02115095e-01
5.58609307e-01 5.63783348e-01 5.14376879e-01 -1.59770787e+00
3.31196606e-01 2.33733300e-02 1.60473600e-01 -3.83041739e-01
5.05653083e-01 -1.21701086e+00 7.40971923e-01 6.14346087e-01
1.71682507e-01 -7.06643462e-01 -1.42262474e-01 9.57510948e-01
4.37003911e-01 -7.96068907e-02 7.45641232e-01 2.26017892e-01
1.33080006e-01 -3.76272947e-02 -4.93401855e-01 -3.72561932e-01
1.16108704e+00 -2.94151813e-01 -3.21442366e-01 -6.33484066e-01
-1.55553132e-01 -2.27859281e-02 2.05694884e-01 -2.67724901e-01
3.31059664e-01 -9.73522544e-01 -5.32813191e-01 1.86270595e-01
4.09554392e-02 -3.48084390e-01 5.13250530e-01 1.20709288e+00
-2.96074867e-01 5.53925633e-01 1.92169309e-01 -6.09976947e-01
-1.05173242e+00 -3.01488042e-01 4.12396133e-01 2.18903765e-01
-2.73182958e-01 7.00700521e-01 -2.92417854e-01 -1.83624640e-01
-8.44019055e-02 -1.88297965e-02 4.42848727e-02 -2.26575911e-01
5.19351922e-02 7.16533244e-01 3.63293469e-01 -5.00993907e-01
-4.47894275e-01 1.03312683e+00 9.91530299e-01 -1.71754837e-01
1.31731713e+00 -6.07354820e-01 -1.63119316e-01 9.22720879e-02
7.07258105e-01 1.07756996e+00 -5.66472590e-01 -1.36567816e-01
-4.23365295e-01 -5.12685299e-01 2.30401039e-01 -5.29756129e-01
-8.61009955e-01 4.57319468e-01 8.93838465e-01 4.40600544e-01
8.66979599e-01 1.65049825e-02 5.02792418e-01 5.28256297e-01
1.23525274e+00 -6.40055180e-01 -2.21121043e-01 3.59977596e-02
6.19352341e-01 -5.03623188e-01 1.63148776e-01 -5.23839951e-01
2.84819484e-01 9.31621492e-01 -1.04394116e-01 1.27336942e-03
9.18694675e-01 2.17561871e-01 -3.72587055e-01 -1.02934502e-01
3.67371105e-02 4.03282680e-02 2.14109197e-02 4.34452057e-01
3.33991826e-01 2.09958643e-01 -2.91379899e-01 6.18453681e-01
2.81222071e-02 -2.49923483e-01 4.68173623e-01 7.44766653e-01
-8.40019822e-01 -8.68915021e-01 -6.33378506e-01 5.03172398e-01
-2.22503006e-01 8.33130628e-02 2.76945382e-01 4.62034434e-01
1.03430025e-01 1.23752737e+00 -2.05754936e-01 -7.89275393e-02
5.14228940e-01 -7.86884367e-01 4.56671238e-01 -3.22872221e-01
1.44947302e-02 5.24208806e-02 1.80569008e-01 -3.49373877e-01
-4.36187208e-01 -6.74605191e-01 -1.10918093e+00 1.15098106e-02
-6.85202003e-01 6.81980610e-01 1.07481825e+00 1.02867317e+00
8.63229632e-01 3.20176244e-01 1.10482883e+00 -8.22128892e-01
-3.56919169e-01 -4.86579657e-01 -1.04390025e+00 -8.18762660e-01
3.06143224e-01 -6.94751382e-01 -6.05133355e-01 -8.94661129e-01] | [6.27695369720459, 1.2458192110061646] |
5d8b2dc8-0a56-478c-89da-01c732403812 | towards-high-performance-one-stage-human-pose | 2301.04842 | null | https://arxiv.org/abs/2301.04842v1 | https://arxiv.org/pdf/2301.04842v1.pdf | Towards High Performance One-Stage Human Pose Estimation | Making top-down human pose estimation method present both good performance and high efficiency is appealing. Mask RCNN can largely improve the efficiency by conducting person detection and pose estimation in a single framework, as the features provided by the backbone are able to be shared by the two tasks. However, the performance is not as good as traditional two-stage methods. In this paper, we aim to largely advance the human pose estimation results of Mask-RCNN and still keep the efficiency. Specifically, we make improvements on the whole process of pose estimation, which contains feature extraction and keypoint detection. The part of feature extraction is ensured to get enough and valuable information of pose. Then, we introduce a Global Context Module into the keypoints detection branch to enlarge the receptive field, as it is crucial to successful human pose estimation. On the COCO val2017 set, our model using the ResNet-50 backbone achieves an AP of 68.1, which is 2.6 higher than Mask RCNN (AP of 65.5). Compared to the classic two-stage top-down method SimpleBaseline, our model largely narrows the performance gap (68.1 AP vs. 68.9 AP) with a much faster inference speed (77 ms vs. 168 ms), demonstrating the effectiveness of the proposed method. Code is available at: https://github.com/lingl_space/maskrcnn_keypoint_refined. | ['Jie Xu', 'Linhao Xu', 'Lin Zhao', 'Ling Li'] | 2023-01-12 | null | null | null | null | ['keypoint-detection', 'human-detection'] | ['computer-vision', 'computer-vision'] | [-1.90357387e-01 -9.87714082e-02 1.50008827e-01 -1.19831063e-01
-7.49318480e-01 -3.52490038e-01 3.86904001e-01 -1.76705986e-01
-9.60840821e-01 6.03247762e-01 3.61962557e-01 2.43652806e-01
1.49266317e-01 -6.71982825e-01 -6.12098694e-01 -5.80729127e-01
-5.45855202e-02 3.10466409e-01 4.35569137e-01 -2.85398483e-01
-1.28089055e-01 4.83557612e-01 -1.64190996e+00 -2.88800290e-03
4.69649136e-01 9.69485402e-01 3.07484537e-01 6.67793691e-01
5.25819123e-01 2.44276643e-01 -8.03623319e-01 -3.82295907e-01
2.73517132e-01 2.53069215e-02 -5.92537582e-01 -3.51314068e-01
5.93929470e-01 -6.02019310e-01 -5.44186950e-01 8.29869747e-01
1.07103062e+00 3.68006140e-01 2.92582154e-01 -1.06676280e+00
-1.99117988e-01 4.17739362e-01 -6.64076626e-01 2.27558702e-01
5.72525144e-01 7.22278655e-02 7.96228349e-01 -1.05814517e+00
3.85953695e-01 1.20829737e+00 8.11839104e-01 6.43633664e-01
-8.02330434e-01 -7.56562948e-01 4.28798467e-01 2.75099158e-01
-1.85874569e+00 -4.60059822e-01 3.76361012e-01 -1.21672653e-01
7.65624940e-01 4.89621043e-01 9.02110100e-01 1.31356406e+00
-6.01587482e-02 9.33227122e-01 8.10837328e-01 -1.37484446e-01
-1.91526469e-02 -1.29386678e-01 -2.35954709e-02 5.52924395e-01
2.69079626e-01 1.56572193e-01 -5.75259030e-01 1.45841137e-01
9.68856394e-01 2.39533663e-01 -3.61739069e-01 -1.02763198e-01
-1.33820879e+00 5.39138079e-01 8.27647209e-01 2.31459975e-01
-3.96228731e-01 2.32772201e-01 3.65810513e-01 -9.39512774e-02
1.67970285e-01 4.30185318e-01 -4.19869721e-01 -2.20005348e-01
-1.03403676e+00 6.02898180e-01 4.57893997e-01 1.07098925e+00
2.18808100e-01 -2.25073248e-01 -6.74166918e-01 6.99784875e-01
1.45319983e-01 7.39163041e-01 2.69895673e-01 -7.94758797e-01
4.03088480e-01 4.47775096e-01 3.26683760e-01 -1.15940416e+00
-7.30470836e-01 -8.25723827e-01 -8.44667017e-01 -1.87890232e-01
6.84989929e-01 -2.95807093e-01 -8.57009768e-01 1.78910291e+00
4.22736973e-01 -2.04326898e-01 -4.96589035e-01 1.23325193e+00
9.43820238e-01 3.98188055e-01 6.03445023e-02 1.97408438e-01
1.81901896e+00 -1.07086372e+00 -6.64053440e-01 -3.52239311e-01
4.69910711e-01 -5.44774830e-01 8.60754192e-01 3.78066838e-01
-1.03775394e+00 -9.41444576e-01 -9.47465837e-01 -2.21638545e-01
-4.03731823e-01 5.80395937e-01 6.50879204e-01 4.48317617e-01
-9.60828543e-01 4.01813924e-01 -7.10159540e-01 -4.28120732e-01
3.48898977e-01 4.56909239e-01 -4.67229068e-01 -4.53376299e-04
-1.41637981e+00 8.28080952e-01 3.27550530e-01 4.42517608e-01
-6.87536120e-01 -4.59159374e-01 -8.62259448e-01 5.47317090e-03
7.14244604e-01 -7.67652869e-01 1.17312050e+00 -2.13221401e-01
-1.28529274e+00 4.36626583e-01 -1.90490872e-01 -4.27232593e-01
9.57270443e-01 -8.46457183e-01 -2.06287205e-01 1.99556381e-01
2.45131746e-01 1.05520797e+00 6.72790706e-01 -8.50162923e-01
-8.23507905e-01 -5.17470360e-01 1.22324698e-01 3.95365179e-01
-3.08728397e-01 3.04590212e-03 -1.12107503e+00 -7.96983123e-01
1.35933906e-01 -1.05705953e+00 -3.82480830e-01 6.72822744e-02
-4.96172816e-01 -2.40691513e-01 2.94647306e-01 -9.10918355e-01
1.34244287e+00 -1.93631601e+00 1.22338273e-01 2.15227544e-01
3.75993073e-01 2.78519958e-01 7.44599029e-02 2.96749413e-01
4.57413085e-02 -3.59348506e-01 8.50548744e-02 -5.62497199e-01
-5.74701168e-02 -3.38230640e-01 9.41455141e-02 6.42693460e-01
-1.39862806e-01 1.16875160e+00 -5.41950345e-01 -3.54274660e-01
3.85236084e-01 6.93389356e-01 -6.12331390e-01 6.65701255e-02
3.33661467e-01 3.88698816e-01 -3.40218842e-01 6.59261525e-01
7.51991332e-01 -1.74704328e-01 6.68457244e-03 -3.50541204e-01
-1.81621298e-01 -3.17194723e-02 -1.51222694e+00 1.88115501e+00
-1.40087306e-01 3.97948593e-01 -1.73682123e-02 -4.72206622e-01
6.80436313e-01 1.72239944e-01 4.00172621e-01 -6.71238124e-01
3.97205204e-01 -2.53903300e-01 -1.08498700e-01 -8.22202340e-02
5.92993677e-01 3.57406974e-01 -2.65846401e-01 4.95495275e-02
-9.31484997e-02 5.03209889e-01 2.18927816e-01 1.49540126e-01
8.61706734e-01 2.93294966e-01 2.64779508e-01 -1.60503790e-01
5.70311546e-01 -2.92919606e-01 4.51578081e-01 8.39339435e-01
-4.08238322e-01 7.12573707e-01 1.37789160e-01 -4.11169350e-01
-7.86622167e-01 -9.92537022e-01 -1.51872765e-02 1.29071391e+00
3.20309401e-01 -9.23653781e-01 -1.01143670e+00 -6.24245822e-01
-2.55406667e-02 1.66560516e-01 -8.58721375e-01 -6.43957928e-02
-7.83068597e-01 -6.39483690e-01 7.24618137e-01 9.77016449e-01
8.05213630e-01 -9.86503422e-01 -7.26246893e-01 5.53299785e-02
-5.46114504e-01 -1.27250624e+00 -6.57275558e-01 -2.05946892e-01
-3.92170936e-01 -8.83553684e-01 -1.10227656e+00 -5.18278301e-01
6.86571836e-01 3.58604133e-01 7.51890004e-01 4.05136235e-02
-4.37205970e-01 1.35514662e-01 -2.82943279e-01 -3.23508531e-01
4.70913857e-01 3.10245991e-01 4.10143405e-01 -2.40847692e-01
3.72018099e-01 -2.30091020e-01 -1.07867742e+00 5.36115944e-01
-2.44805977e-01 8.42129663e-02 8.26157212e-01 4.40437227e-01
4.98667121e-01 -8.28333348e-02 2.75964320e-01 -3.72346520e-01
2.62933820e-01 5.67530170e-02 -3.77004713e-01 -2.88484190e-02
-2.78939635e-01 -2.56945580e-01 4.66775417e-01 -4.62154806e-01
-8.52719665e-01 2.54219532e-01 -4.69410926e-01 -1.95239380e-01
-2.19008163e-01 -2.45615430e-02 -2.09959880e-01 5.77667020e-02
6.51991785e-01 1.97079659e-01 2.75777448e-02 -5.85326433e-01
3.50582540e-01 4.92858678e-01 6.30473673e-01 -3.39802176e-01
8.85341525e-01 5.83207846e-01 -2.35132635e-01 -6.75205827e-01
-8.63333821e-01 -7.57546842e-01 -7.11306155e-01 -3.00050318e-01
9.16155338e-01 -1.23268855e+00 -1.18487048e+00 4.92187709e-01
-9.75133717e-01 -5.80672771e-02 -1.14216872e-01 5.91815352e-01
-3.73009086e-01 2.60491163e-01 -7.12759078e-01 -7.28288472e-01
-6.28697157e-01 -9.00537968e-01 1.27208543e+00 1.85132667e-01
-3.95785660e-01 -4.01194453e-01 -2.25668907e-01 3.66901428e-01
3.09876144e-01 6.75884709e-02 -1.28718525e-01 -4.96818483e-01
-2.71880746e-01 -4.18392956e-01 -3.80866140e-01 3.67402099e-02
-1.03160433e-01 -5.19654512e-01 -1.13966763e+00 -6.79004252e-01
-1.87904507e-01 -1.17525123e-01 9.08891618e-01 6.28783047e-01
1.27659893e+00 -7.17888921e-02 -5.83616316e-01 6.40710890e-01
8.94647360e-01 -2.13146642e-01 6.90052032e-01 4.32691723e-01
8.71805131e-01 5.99279761e-01 7.47960806e-01 5.37656963e-01
4.56122011e-01 9.81444001e-01 2.79351652e-01 -3.37299943e-01
-2.21080571e-01 -4.67147619e-01 3.48165721e-01 4.04646218e-01
-5.61811090e-01 3.61059420e-02 -7.52481639e-01 1.95788309e-01
-1.97528636e+00 -9.43726122e-01 6.70157894e-02 2.26900244e+00
6.18930876e-01 3.04307938e-01 5.59708953e-01 5.84147684e-02
8.63590956e-01 1.20789550e-01 -2.77283996e-01 3.11839640e-01
2.19244823e-01 -1.41902128e-02 6.17130816e-01 3.95018041e-01
-1.38856637e+00 1.02210438e+00 5.98827124e+00 1.02471137e+00
-7.13245451e-01 -6.62901700e-02 1.87140405e-01 -4.34316069e-01
4.98952806e-01 -5.03099680e-01 -1.32284808e+00 5.36767423e-01
5.98913550e-01 2.92323619e-01 3.72170866e-01 9.11925435e-01
1.04673117e-01 -1.31417349e-01 -1.06157398e+00 1.34432912e+00
4.89600413e-02 -9.17048573e-01 -1.33345068e-01 2.24544391e-01
1.30613714e-01 -1.94046140e-01 -6.20326810e-02 5.49371779e-01
-2.37478301e-01 -1.08302712e+00 7.98465014e-01 4.65728253e-01
7.80632436e-01 -1.05167556e+00 9.14887190e-01 5.16891956e-01
-1.66599381e+00 -1.46894991e-01 -4.40437376e-01 -1.77855745e-01
2.34001800e-01 4.72910345e-01 -4.89739895e-01 5.36213458e-01
1.13295650e+00 5.45949280e-01 -9.47998464e-01 1.00007260e+00
-4.54637170e-01 8.32916871e-02 -4.40127552e-01 -8.78477097e-02
-6.28951266e-02 2.47974291e-01 5.58759928e-01 1.34162986e+00
8.60392153e-02 2.38118563e-02 3.21030021e-01 5.25004029e-01
1.87188070e-02 9.81294364e-02 -3.24140072e-01 5.88022530e-01
5.85535228e-01 1.49040985e+00 -7.64332771e-01 -3.39988649e-01
-1.39839321e-01 1.11061668e+00 4.19601738e-01 1.97944537e-01
-1.28644001e+00 -5.00920057e-01 5.20665109e-01 2.62563378e-01
4.61296916e-01 -2.30406433e-01 9.05066729e-02 -1.02907670e+00
2.51927137e-01 -8.92854631e-01 4.14926380e-01 -4.85869110e-01
-9.28201914e-01 6.75757706e-01 1.43807799e-01 -1.20564771e+00
-9.25971270e-02 -6.98624253e-01 -2.59623110e-01 6.99829996e-01
-1.04273379e+00 -1.30347538e+00 -5.68468869e-01 6.83724344e-01
5.29995263e-01 1.65632784e-01 6.99366212e-01 5.43007314e-01
-6.84631646e-01 1.06216240e+00 -4.02874112e-01 4.11493480e-01
8.72299075e-01 -1.06700015e+00 6.57073021e-01 9.30836439e-01
-1.42892271e-01 1.02722967e+00 5.85804939e-01 -6.83505058e-01
-1.13705361e+00 -9.58410323e-01 6.51470900e-01 -7.02595592e-01
2.62434632e-01 -7.97973216e-01 -4.33099329e-01 5.03595591e-01
-2.73381591e-01 5.73778525e-02 4.36028242e-01 3.83001357e-01
-1.78552523e-01 -3.44092585e-02 -9.40874696e-01 9.36042964e-01
1.44575167e+00 -1.92488536e-01 -4.94771659e-01 1.39947012e-01
6.67128384e-01 -5.10881782e-01 -7.69149065e-01 4.95066881e-01
1.00996816e+00 -8.46986711e-01 1.41308331e+00 -1.09357007e-01
-8.36514756e-02 -3.34345192e-01 -1.33906886e-01 -8.54047954e-01
-6.85141206e-01 -5.41660130e-01 -4.60397661e-01 9.05865431e-01
1.58221841e-01 -5.71018934e-01 7.90217161e-01 3.73872757e-01
1.64730951e-01 -9.24886942e-01 -8.23670030e-01 -7.33991206e-01
-2.32399404e-01 -4.88444984e-01 6.52717710e-01 3.40223610e-01
-1.94465399e-01 4.05697197e-01 -5.69971740e-01 1.70840859e-01
6.51375413e-01 -1.44603610e-01 9.01509464e-01 -1.16351497e+00
-2.90884972e-01 -3.61927330e-01 -2.73925364e-01 -1.49806023e+00
-3.27818334e-01 -4.64786053e-01 4.88487147e-02 -1.47215390e+00
2.80881107e-01 -2.32652768e-01 -2.93927729e-01 5.08592427e-01
-3.89463365e-01 5.29883504e-01 4.07478243e-01 1.92389309e-01
-7.72052109e-01 5.86394310e-01 1.33979702e+00 1.32072031e-01
-1.51591361e-01 2.47028098e-01 -8.04459929e-01 9.21698570e-01
9.05057311e-01 -1.86576515e-01 -2.31569603e-01 -1.97130591e-01
7.65484124e-02 -2.40061164e-01 6.68013096e-01 -1.42866457e+00
4.82509524e-01 3.49361330e-01 1.12119579e+00 -8.85924160e-01
6.87138200e-01 -6.55510485e-01 -7.93078691e-02 7.18258679e-01
-1.27854660e-01 9.65268016e-02 1.84720024e-01 4.22224134e-01
2.07860827e-01 1.97858557e-01 6.04409814e-01 -1.95038825e-01
-7.77578413e-01 5.37259459e-01 -3.87340710e-02 -6.32803217e-02
9.60931003e-01 -3.56438965e-01 -1.74182355e-01 -2.74986327e-01
-8.81747663e-01 3.03119600e-01 1.99745372e-01 5.84365726e-01
6.26085699e-01 -1.34986424e+00 -6.41303122e-01 2.11343139e-01
1.36635423e-01 7.16764852e-02 4.65814948e-01 9.96679783e-01
-2.85163015e-01 6.16100132e-01 -1.61646038e-01 -5.66390693e-01
-1.27368832e+00 4.73299503e-01 2.20355079e-01 -1.95574865e-01
-8.87401760e-01 1.04588890e+00 3.17629784e-01 -4.27323639e-01
6.10181153e-01 -8.94843116e-02 -5.22869945e-01 2.10426956e-01
9.39661264e-01 6.93789124e-01 -8.12590718e-02 -6.94149554e-01
-7.64230490e-01 7.04194427e-01 -2.36948833e-01 -1.44500062e-01
1.14580166e+00 -1.27126500e-01 1.68670639e-01 -1.28612652e-01
9.23259616e-01 1.92355976e-01 -1.40483034e+00 -9.38917249e-02
-4.85568643e-01 -4.24674690e-01 -7.74051175e-02 -8.30923140e-01
-8.88420403e-01 6.62226200e-01 7.59134114e-01 -1.89376369e-01
9.30303812e-01 9.99604911e-02 8.52324486e-01 3.89233112e-01
6.12569571e-01 -1.32896829e+00 1.72840096e-02 3.76539350e-01
9.92591858e-01 -1.08273876e+00 3.14215481e-01 -5.71001112e-01
-5.54909885e-01 6.84891760e-01 1.01782596e+00 -1.86044484e-01
3.37037176e-01 2.08606154e-01 -2.03779668e-01 -2.59014964e-01
-2.56470084e-01 -4.87875640e-01 6.22095048e-01 5.08176863e-01
3.34559560e-01 6.54418245e-02 -1.80696055e-01 8.58969152e-01
-5.84157765e-01 -2.18583465e-01 -1.15462795e-01 8.14485312e-01
-4.67956126e-01 -6.41708195e-01 -4.60459381e-01 2.17282265e-01
-5.55464029e-01 -3.11986823e-02 -1.69884250e-01 9.77316499e-01
1.86510652e-01 8.44915688e-01 -6.39449731e-02 -5.77984214e-01
6.17589772e-01 -2.40869626e-01 4.88807768e-01 -4.72628415e-01
-5.47339201e-01 1.99364692e-01 6.09202571e-02 -1.00696480e+00
-2.17886329e-01 -4.42718655e-01 -1.13937783e+00 -4.94729429e-01
-4.11713302e-01 2.71325819e-02 4.04161513e-01 7.86699951e-01
3.58071238e-01 7.91993737e-01 2.10348099e-01 -1.16667700e+00
-4.94436920e-01 -1.05708659e+00 -2.98116386e-01 1.63883761e-01
9.56095457e-02 -8.91268551e-01 -9.42421779e-02 -3.53906333e-01] | [7.141965866088867, -0.7796465158462524] |
5793d2c7-6105-43f0-9617-95c3439d3583 | utterance-to-utterance-interactive-matching | 1911.06940 | null | https://arxiv.org/abs/1911.06940v1 | https://arxiv.org/pdf/1911.06940v1.pdf | Utterance-to-Utterance Interactive Matching Network for Multi-Turn Response Selection in Retrieval-Based Chatbots | This paper proposes an utterance-to-utterance interactive matching network (U2U-IMN) for multi-turn response selection in retrieval-based chatbots. Different from previous methods following context-to-response matching or utterance-to-response matching frameworks, this model treats both contexts and responses as sequences of utterances when calculating the matching degrees between them. For a context-response pair, the U2U-IMN model first encodes each utterance separately using recurrent and self-attention layers. Then, a global and bidirectional interaction between the context and the response is conducted using the attention mechanism to collect the matching information between them. The distances between context and response utterances are employed as a prior component when calculating the attention weights. Finally, sentence-level aggregation and context-response-level aggregation are executed in turn to obtain the feature vector for matching degree prediction. Experiments on four public datasets showed that our proposed method outperformed baseline methods on all metrics, achieving a new state-of-the-art performance and demonstrating compatibility across domains for multi-turn response selection. | ['Zhen-Hua Ling', 'Jia-Chen Gu', 'Quan Liu'] | 2019-11-16 | null | null | null | null | ['conversational-response-selection'] | ['natural-language-processing'] | [ 3.01692307e-01 -1.65666789e-02 -3.04505140e-01 -9.53624666e-01
-1.34099483e+00 -1.94027156e-01 6.08415306e-01 1.30666465e-01
-4.10544306e-01 2.66553491e-01 6.81071043e-01 -1.67556480e-01
-7.78106824e-02 -5.46626031e-01 -1.16033517e-01 -3.83709490e-01
2.34532073e-01 7.57686853e-01 1.65264353e-01 -6.29028320e-01
7.10436165e-01 -1.54371828e-01 -1.20323861e+00 8.41788173e-01
8.25332403e-01 9.90138233e-01 4.78733480e-01 8.71452630e-01
-5.22659719e-01 1.28472698e+00 -4.85390335e-01 -3.42437118e-01
-1.43805444e-01 -7.80735552e-01 -1.50792587e+00 -1.12395152e-01
1.30923346e-01 -3.15879703e-01 -3.89199406e-01 6.02176487e-01
7.08927393e-01 8.23977888e-01 4.71775949e-01 -7.74820626e-01
-1.07031000e+00 8.14290106e-01 -2.83407897e-01 1.30280703e-01
1.06516540e+00 -1.53462403e-02 1.60135043e+00 -1.20977676e+00
7.22472429e-01 1.57919014e+00 1.74104899e-01 6.62825763e-01
-9.59654510e-01 -3.75703305e-01 3.00048947e-01 7.46450126e-01
-8.95974338e-01 -3.72168332e-01 8.64337385e-01 -2.26976812e-01
1.36737943e+00 4.87209052e-01 2.48280257e-01 7.42315888e-01
-6.45284355e-02 1.37756598e+00 5.73175788e-01 -5.42089343e-01
-2.57897943e-01 -7.84702599e-03 6.47293806e-01 1.06382377e-01
-1.30561852e+00 -4.90464658e-01 -5.01807868e-01 -2.91583091e-01
2.69624382e-01 5.85032590e-02 3.50414515e-02 4.94400293e-01
-8.97852480e-01 1.09581375e+00 5.65091968e-01 3.55726331e-01
-5.72404861e-01 -1.35378450e-01 7.17361808e-01 9.34344709e-01
5.90449810e-01 4.14314151e-01 -1.72609001e-01 -2.29688689e-01
-2.99417317e-01 4.02406365e-01 6.67411149e-01 1.05052745e+00
8.75072241e-01 -4.33814019e-01 -9.79899883e-01 1.49864471e+00
2.10143819e-01 5.09845354e-02 4.44843143e-01 -7.52088726e-01
7.74883330e-01 8.12793434e-01 1.71368077e-01 -1.03105068e+00
-3.80643636e-01 1.11261517e-01 -6.11337662e-01 -7.31189370e-01
1.81275383e-01 6.38668537e-02 -3.70277673e-01 1.62515557e+00
4.66718048e-01 -1.61392525e-01 -7.73060694e-02 1.30859435e+00
1.35284591e+00 6.47157431e-01 2.01491117e-01 -2.46585339e-01
1.38777411e+00 -1.36823857e+00 -9.72597480e-01 -1.34086549e-01
8.63502562e-01 -9.75169659e-01 1.41496837e+00 -1.72115207e-01
-1.24921536e+00 -4.76329952e-01 -4.00191873e-01 -4.77827787e-01
-8.01468920e-03 1.25187442e-01 3.06274265e-01 -1.42692164e-01
-9.91681457e-01 2.32523277e-01 -3.02857086e-02 -3.27187181e-01
-3.43606979e-01 3.36621553e-01 -5.72513454e-02 -1.54826045e-01
-1.96778226e+00 1.16214418e+00 -2.00228572e-01 2.56637663e-01
-7.06639886e-01 -3.31795484e-01 -7.12316275e-01 1.38833970e-01
1.79465726e-01 -3.89979273e-01 1.71985102e+00 -9.84610856e-01
-1.82623732e+00 7.59182215e-01 -5.18340766e-01 -1.84679151e-01
1.53704733e-01 -1.10384822e-01 -2.36406669e-01 3.19604352e-02
8.36374834e-02 6.12719297e-01 4.71303344e-01 -6.41119480e-01
-7.20580637e-01 -2.31082737e-01 4.52093095e-01 7.90322363e-01
-1.10407837e-01 6.67775929e-01 -6.68428361e-01 -1.90996975e-01
3.41145694e-01 -8.08727920e-01 -1.07430421e-01 -1.04318619e+00
-3.24969947e-01 -1.32231390e+00 8.17988634e-01 -7.67385304e-01
1.46475697e+00 -1.76707184e+00 4.32736307e-01 -1.70612916e-01
1.53868467e-01 -4.74657826e-02 -6.52999341e-01 1.06060970e+00
1.71908438e-02 -1.61064550e-01 2.59746760e-01 -6.02239370e-01
2.42261905e-02 -2.83071995e-01 -1.74231961e-01 2.82784462e-01
2.61749893e-01 9.73703146e-01 -1.10469234e+00 -4.94448751e-01
9.42807794e-02 1.45166561e-01 -5.79426348e-01 9.73363996e-01
-2.15318635e-01 5.41577756e-01 -6.36468828e-01 5.86403191e-01
2.14654073e-01 -7.27906004e-02 3.94841582e-01 2.17177719e-01
1.73376411e-01 1.02778184e+00 -3.81830037e-01 1.59348464e+00
-9.57385659e-01 5.84485471e-01 1.00584671e-01 -9.46587205e-01
1.26947117e+00 4.95817453e-01 4.79447633e-01 -1.22502589e+00
1.72165900e-01 5.80652580e-02 4.90227453e-02 -7.98130393e-01
9.88976955e-01 1.81003794e-01 -7.09244430e-01 9.12231863e-01
1.35667948e-02 -1.18104450e-01 8.85027498e-02 5.81194580e-01
1.00734746e+00 -1.61673024e-01 5.98033294e-02 -7.20952824e-02
8.00979137e-01 -2.82115757e-01 4.49183106e-01 7.26844907e-01
-2.22229257e-01 4.74186003e-01 5.81655443e-01 -5.92978060e-01
-6.60033166e-01 -3.42273057e-01 2.83166111e-01 2.04760337e+00
2.81032383e-01 -2.27691337e-01 -7.08626151e-01 -7.34931231e-01
-3.87751400e-01 7.67994642e-01 -7.70694435e-01 -1.78509846e-01
-6.38053417e-01 -1.90967828e-01 3.60183865e-01 3.51210296e-01
3.49187851e-01 -1.55640972e+00 -1.25750721e-01 5.39120317e-01
-9.46824849e-01 -9.58034694e-01 -1.03617454e+00 2.17860639e-02
-3.97709787e-01 -7.98799694e-01 -4.78801548e-01 -1.02075171e+00
3.12432587e-01 5.07996798e-01 1.27680957e+00 4.40758139e-01
1.38370410e-01 1.97391257e-01 -9.05943811e-01 2.28460990e-02
-4.04572099e-01 2.52207190e-01 -1.36705607e-01 9.37577337e-02
6.88025832e-01 -4.62329417e-01 -6.19557142e-01 6.51660204e-01
-5.90713561e-01 -1.21525183e-01 3.00756007e-01 1.21595478e+00
2.05300659e-01 -8.93066347e-01 1.06869745e+00 -8.10440421e-01
1.36333156e+00 -7.58104503e-01 -7.65159205e-02 6.25270128e-01
-2.34410197e-01 -2.10650191e-01 5.64271867e-01 -5.91286540e-01
-1.23421490e+00 -1.24666311e-01 -2.74403632e-01 -9.49823931e-02
7.16944486e-02 5.77913821e-01 -1.36062697e-01 2.78230190e-01
3.90916318e-01 2.24968657e-01 -2.54738867e-01 -3.58820945e-01
4.97709215e-01 1.32069016e+00 1.67662561e-01 -5.33804119e-01
-1.60104960e-01 -2.48250157e-01 -8.96921337e-01 -5.76742828e-01
-7.45239794e-01 -1.06216931e+00 -5.65954208e-01 -5.53639293e-01
1.00970232e+00 -4.83123779e-01 -1.11767805e+00 3.43480289e-01
-1.55697680e+00 -3.13847840e-01 3.60546201e-01 3.55482668e-01
-4.70406890e-01 2.48263136e-01 -1.10783994e+00 -1.17714357e+00
-7.00695395e-01 -1.39131153e+00 1.03533673e+00 2.08360836e-01
-4.47122037e-01 -8.92349780e-01 3.31307240e-02 7.27686048e-01
5.33658683e-01 -6.12230301e-01 9.03781652e-01 -1.15283203e+00
-3.20812315e-01 -4.71041143e-01 -2.82842636e-01 -2.01009184e-01
-2.98277382e-02 -2.07128778e-01 -8.24634790e-01 2.27829292e-02
1.36476249e-01 -7.61249185e-01 7.66620994e-01 2.18993291e-01
8.96583796e-01 -4.81649786e-01 -9.22695268e-03 -4.46423404e-02
8.30051422e-01 4.42376047e-01 7.11186290e-01 1.44538105e-01
5.18610060e-01 1.02517152e+00 1.05966032e+00 3.83363366e-01
8.41574967e-01 1.05481744e+00 4.37049717e-01 2.34007426e-02
2.41057083e-01 -1.69717640e-01 1.63201243e-01 1.16730130e+00
2.83145189e-01 -3.92773092e-01 -6.11558080e-01 7.41829991e-01
-2.16319203e+00 -1.18068159e+00 -2.85993308e-01 2.09052396e+00
8.53795767e-01 -2.65029520e-01 3.28097075e-01 -4.53513145e-01
7.78541863e-01 2.37197310e-01 -5.23580849e-01 -9.81754720e-01
1.37812018e-01 -2.82568872e-01 -1.99377671e-01 9.08735633e-01
-8.44218612e-01 1.19910586e+00 6.06203938e+00 8.89699936e-01
-8.80012453e-01 2.91016579e-01 8.25288355e-01 2.50151455e-02
-5.50197601e-01 -5.88888943e-04 -6.48783028e-01 2.80102581e-01
7.72845209e-01 8.24899301e-02 5.23029447e-01 7.60129571e-01
3.00849676e-01 -5.66486046e-02 -9.35906529e-01 4.83082324e-01
2.16018595e-02 -1.18233991e+00 -2.34975159e-01 -4.83838350e-01
4.11154002e-01 -7.77584165e-02 -1.45332143e-01 8.46461892e-01
5.43057203e-01 -7.91088462e-01 3.10390353e-01 5.27372777e-01
5.90872645e-01 -7.45051026e-01 9.45992887e-01 5.23143291e-01
-1.24218023e+00 -2.03000739e-01 -4.56078976e-01 -1.79549694e-01
2.81280071e-01 1.11466862e-01 -1.10242569e+00 6.95788920e-01
4.82947290e-01 4.93757516e-01 -1.34013027e-01 4.14572507e-01
-1.59174517e-01 4.22544271e-01 7.45562417e-03 -5.60440958e-01
5.64791560e-01 -3.42040241e-01 3.70262325e-01 1.47625113e+00
-1.87284559e-01 3.57548088e-01 4.67859119e-01 5.92126548e-01
-2.43596122e-01 5.41870236e-01 -2.55032539e-01 3.01675647e-01
8.82755995e-01 1.27421844e+00 -2.26575017e-01 -4.76676285e-01
-4.15325820e-01 8.70481431e-01 8.39374304e-01 4.25520986e-01
-7.14796722e-01 -5.94099343e-01 6.50713384e-01 -3.07142943e-01
-7.87301734e-02 3.54295552e-01 -1.70532450e-01 -7.38478124e-01
-7.01434314e-02 -8.60211730e-01 5.47831118e-01 -5.66737890e-01
-1.41495669e+00 9.34619248e-01 -1.74791679e-01 -1.30554795e+00
-5.82551718e-01 1.91486329e-01 -1.02691483e+00 1.28846693e+00
-1.12869084e+00 -1.16257524e+00 2.95728445e-02 5.02944708e-01
1.15746450e+00 -1.24301188e-01 9.31943834e-01 3.15477073e-01
-4.80988950e-01 9.65896606e-01 -1.51815325e-01 2.29222700e-01
7.56717741e-01 -9.17951584e-01 3.78722012e-01 5.09929836e-01
-2.29438215e-01 8.00900280e-01 4.32240933e-01 -5.35912454e-01
-1.37340260e+00 -6.96008146e-01 1.54649854e+00 -2.71703333e-01
5.43127239e-01 -3.56562734e-01 -1.05651331e+00 4.21837091e-01
7.06438363e-01 -5.53594649e-01 7.54226267e-01 4.92378831e-01
-2.69147575e-01 1.26620516e-01 -8.23048830e-01 6.16306007e-01
7.71870434e-01 -1.00844967e+00 -6.31749630e-01 5.79173148e-01
1.21565616e+00 -6.21438444e-01 -8.64971459e-01 4.21248227e-02
6.04225814e-01 -8.12617064e-01 7.06301332e-01 -8.94635201e-01
6.77373767e-01 1.93537176e-01 -2.93348879e-01 -1.19908321e+00
-5.74119210e-01 -7.98902392e-01 2.03320146e-01 1.30344272e+00
4.87413466e-01 -3.10697824e-01 4.18714523e-01 8.00191224e-01
-5.36532283e-01 -1.01486504e+00 -9.13371027e-01 -1.40897587e-01
-2.94404849e-02 -3.62825692e-01 7.62755215e-01 1.01812816e+00
7.06777394e-01 9.57473636e-01 -7.59706557e-01 -4.03048456e-01
-2.08915770e-02 4.56475794e-01 6.97173655e-01 -6.18114769e-01
-1.41976908e-01 -6.96454942e-01 7.19141513e-02 -1.70810151e+00
5.27545393e-01 -8.74350131e-01 4.20251042e-01 -1.65472186e+00
2.55777568e-01 -4.10941064e-01 -1.41763955e-01 2.77842075e-01
-7.04276383e-01 -1.23639017e-01 2.70010024e-01 4.50415730e-01
-1.02569902e+00 7.41381526e-01 1.29797518e+00 -2.25046441e-01
-4.37115610e-01 3.18090469e-01 -4.90425229e-01 1.56519637e-01
6.55101597e-01 -4.53937382e-01 -3.50323588e-01 -6.14601791e-01
-5.32444706e-03 8.55055034e-01 -2.77432203e-01 -1.57316431e-01
5.02371728e-01 -3.99855375e-01 -3.46111268e-01 -8.60637665e-01
4.72132742e-01 -3.38403136e-01 -4.25179124e-01 -2.31892411e-02
-1.33891726e+00 1.94846436e-01 -4.14356947e-01 5.17969072e-01
-4.60262746e-01 -2.43352517e-01 3.61484230e-01 -1.37197524e-01
-6.19157612e-01 2.33760282e-01 -5.94355822e-01 1.80841261e-03
5.70506275e-01 -4.49971408e-02 -2.52344131e-01 -1.15052986e+00
-6.81723595e-01 8.39079559e-01 -7.95817897e-02 8.03243637e-01
8.34989488e-01 -1.50466621e+00 -1.03315091e+00 -1.17695071e-01
3.96163672e-01 -1.53570399e-01 8.74433219e-01 1.01724136e+00
2.55570203e-01 5.59941590e-01 1.93948030e-01 -5.71945548e-01
-1.55792236e+00 2.87709922e-01 3.32462847e-01 -7.05189109e-01
-1.62784845e-01 1.32585907e+00 1.52653292e-01 -1.05440140e+00
3.20002258e-01 1.19859159e-01 -6.37057662e-01 2.75027424e-01
7.53788948e-01 7.27362335e-02 1.26232699e-01 -7.94013143e-01
-4.06127363e-01 1.17662124e-01 -4.20266777e-01 -1.60540178e-01
1.13636041e+00 -4.86893684e-01 -3.86898398e-01 4.20541495e-01
1.51099014e+00 -4.13032919e-01 -8.13843668e-01 -7.83323348e-01
2.58648098e-01 -5.54867625e-01 -4.20548081e-01 -6.28475189e-01
-8.22657704e-01 9.20856535e-01 -5.87766841e-02 4.26579237e-01
1.12000263e+00 7.82955065e-02 1.02776945e+00 5.87262988e-01
6.02401346e-02 -1.37309778e+00 4.75973934e-01 1.12166393e+00
1.14394963e+00 -1.31887746e+00 -4.21588600e-01 -1.50131971e-01
-1.24772060e+00 9.62542713e-01 1.01074696e+00 -9.39266570e-03
1.95518911e-01 -2.27278471e-01 4.41528291e-01 -1.29216835e-01
-1.34744895e+00 -2.01895699e-01 4.05343324e-01 3.85020018e-01
1.06526649e+00 7.58681744e-02 -6.82724476e-01 4.34590280e-01
5.23664318e-02 -4.59154159e-01 3.01398724e-01 6.94991648e-01
-4.63970274e-01 -1.29198611e+00 -6.79965178e-03 3.91168445e-01
-2.49796972e-01 -2.31677234e-01 -8.85689139e-01 2.41057500e-01
-7.14847982e-01 1.56247866e+00 1.69584081e-02 -1.06239259e+00
4.67662573e-01 1.86695993e-01 -1.57055795e-01 -8.17517519e-01
-1.51154184e+00 9.42770764e-02 5.44063389e-01 -6.40278220e-01
-2.96498090e-01 -4.25342977e-01 -1.26985836e+00 -2.36607730e-01
-6.68814480e-01 4.44582820e-01 4.41012532e-01 1.08265305e+00
4.00008202e-01 5.75317800e-01 1.29885066e+00 -8.21155965e-01
-7.56631196e-01 -1.65827370e+00 -1.17671348e-01 8.30317914e-01
8.53236839e-02 -3.77339691e-01 4.56937626e-02 -6.07300937e-01] | [12.500687599182129, 7.829361438751221] |
4e4b800a-e147-4dd6-9dab-136cd9c7405f | the-combination-of-context-information-to | 1810.04000 | null | http://arxiv.org/abs/1810.04000v1 | http://arxiv.org/pdf/1810.04000v1.pdf | The combination of context information to enhance simple question answering | With the rapid development of knowledge base,question answering based on
knowledge base has been a hot research issue. In this paper, we focus on
answering singlerelation factoid questions based on knowledge base. We build a
question answering system and study the effect of context information on fact
selection, such as entity's notable type,outdegree. Experimental results show
that context information can improve the result of simple question answering. | ['Lin Li', 'Zhaohui Chao'] | 2018-10-09 | null | null | null | null | ['knowledge-base-question-answering', 'fact-selection'] | ['natural-language-processing', 'natural-language-processing'] | [-7.84411669e-01 2.64699787e-01 -3.14643502e-01 -3.34876120e-01
-4.64277864e-01 -4.29946274e-01 3.46165895e-01 4.03256923e-01
-4.42469180e-01 1.24821258e+00 4.50843811e-01 -4.35647786e-01
-4.30395752e-01 -1.53946102e+00 -3.20683390e-01 1.78132132e-01
2.04621524e-01 3.77597004e-01 1.17086697e+00 -8.40550840e-01
4.78275537e-01 -1.17482208e-01 -1.54075730e+00 5.50078869e-01
1.31308222e+00 7.89470196e-01 5.36658950e-02 3.38100761e-01
-9.81456935e-01 1.40895188e+00 -1.06994891e+00 -6.31593227e-01
-2.40056425e-01 -4.95097607e-01 -1.38177323e+00 -6.15970612e-01
8.77661034e-02 1.46471960e-02 -4.20515656e-01 8.66581142e-01
4.64442015e-01 2.76707798e-01 3.03197712e-01 -1.14284348e+00
-7.86768615e-01 5.89920342e-01 6.85669035e-02 9.62476611e-01
1.08078003e+00 -5.71122110e-01 9.84340012e-01 -6.18574500e-01
7.67019749e-01 1.20995593e+00 6.08657062e-01 1.35462359e-01
-2.30468437e-01 -3.27980101e-01 1.96524441e-01 1.02535343e+00
-1.59503877e+00 7.44225532e-02 5.87136328e-01 -7.51115680e-02
1.03188050e+00 4.70854819e-01 4.46637779e-01 1.56873301e-01
-6.07665218e-02 5.01649439e-01 1.16508973e+00 -7.16917932e-01
1.46508485e-01 2.97602087e-01 7.41987109e-01 6.98549986e-01
6.10080719e-01 -5.36983967e-01 -2.44136840e-01 -3.44905049e-01
6.31711602e-01 -1.65178567e-01 -4.06464487e-01 4.87787962e-01
-6.77710414e-01 9.13079023e-01 5.50236583e-01 5.48549294e-01
-3.59122992e-01 -2.07340404e-01 5.69681227e-02 5.62431335e-01
1.10911690e-01 7.52998292e-01 -8.84747624e-01 1.13632567e-02
-2.03836486e-01 5.99854171e-01 1.56765270e+00 8.66549969e-01
8.86698961e-01 -6.90954506e-01 -3.68959814e-01 7.59564996e-01
1.25959218e-01 4.10760731e-01 4.79554951e-01 -7.05406666e-01
4.59839672e-01 1.41375101e+00 3.56093824e-01 -1.20890105e+00
-4.53396380e-01 -2.28361964e-01 -1.23895943e-01 -7.59544790e-01
3.72498155e-01 -4.82180446e-01 -6.39013112e-01 1.15954900e+00
6.46184742e-01 -7.81409889e-02 2.34035358e-01 7.32922614e-01
1.61193848e+00 6.60313010e-01 1.81895524e-01 -3.94928664e-01
1.87255478e+00 -6.76189423e-01 -1.23082078e+00 2.31090546e-01
6.43417060e-01 -8.46073508e-01 7.49613047e-01 2.38279253e-02
-4.68333334e-01 -3.57763946e-01 -5.14141142e-01 -2.44985029e-01
-9.93035793e-01 -1.92897111e-01 8.40836644e-01 8.34606707e-01
-6.07369006e-01 -1.23230718e-01 -1.79703072e-01 -5.69435537e-01
1.28832951e-01 1.08674005e-01 -1.14020713e-01 -3.65158677e-01
-1.99371362e+00 1.15151763e+00 9.62251186e-01 -3.54894489e-01
-3.91981937e-02 -6.01220250e-01 -2.70777792e-01 1.81828201e-01
1.08709180e+00 -9.52677786e-01 1.16854262e+00 -3.43404144e-01
-1.03640449e+00 4.39204156e-01 -2.72948086e-01 -3.32321584e-01
-2.68618733e-01 -2.19658151e-01 -1.19001663e+00 4.18237269e-01
2.17775524e-01 -1.10252514e-01 -5.28045259e-02 -1.06221879e+00
-1.02891791e+00 -3.12503278e-01 8.69304836e-01 4.42490071e-01
-2.10729092e-02 2.56334931e-01 -4.35499996e-01 -9.48401615e-02
4.32216562e-02 -3.51862967e-01 -8.80728289e-02 -5.89045882e-01
-1.04186036e-01 -8.63149941e-01 8.22343469e-01 -7.90137529e-01
1.82556915e+00 -1.58148324e+00 -6.43979907e-01 2.58028626e-01
5.94055988e-02 3.97341400e-01 5.31832457e-01 7.94839561e-01
4.10977423e-01 3.86289150e-01 2.35558197e-01 1.19989681e+00
-1.80932850e-01 4.67834890e-01 -3.67375553e-01 -5.18353939e-01
-5.86692728e-02 1.02580202e+00 -8.95306706e-01 -1.05447817e+00
-3.20239544e-01 -1.25208333e-01 -5.44767141e-01 2.42610231e-01
-6.91132247e-01 -1.25750139e-01 -1.23716021e+00 9.07508850e-01
4.75456148e-01 -4.88950402e-01 7.52509683e-02 -7.48509690e-02
1.23392038e-01 3.51906270e-01 -1.01181161e+00 1.17825484e+00
-4.67179924e-01 4.87030968e-02 -4.04078424e-01 -6.82686687e-01
1.07014513e+00 4.96998459e-01 -8.41456577e-02 -7.89692998e-01
2.68049277e-02 2.90814549e-01 2.73689032e-01 -1.35399663e+00
4.05809820e-01 -2.94267625e-01 2.78357230e-02 1.52639747e-01
-1.69674233e-01 -1.59526750e-01 5.47564685e-01 5.03298640e-01
1.16153419e+00 -2.11780727e-01 5.96525967e-01 -4.01952028e-01
6.92288518e-01 5.28055191e-01 2.71027148e-01 5.83170235e-01
-2.60936785e-02 -9.76475403e-02 4.64248568e-01 -4.01788533e-01
-3.02111328e-01 -8.58887434e-01 -2.50342399e-01 1.00065768e+00
4.15495366e-01 -6.14868999e-01 -5.42032719e-01 -9.55275774e-01
1.73675358e-01 7.75576353e-01 -5.31373799e-01 5.44147044e-02
-5.57326317e-01 -7.94114769e-01 3.68141413e-01 2.20880330e-01
9.86459434e-01 -1.14104080e+00 -2.44481534e-01 2.79614598e-01
-7.26473749e-01 -1.02324986e+00 2.03914508e-01 -3.54593575e-01
-8.94096017e-01 -1.50170100e+00 -1.89583033e-01 -8.69085610e-01
3.87662441e-01 3.40788662e-01 1.52942038e+00 6.58189774e-01
2.08664298e-01 5.45170605e-01 -9.83338237e-01 -7.13596463e-01
2.18549907e-01 4.72556919e-01 -4.25003856e-01 -3.98217052e-01
7.38683701e-01 -4.78095829e-01 -7.32183099e-01 5.68341494e-01
-1.27758718e+00 -7.43681908e-01 4.59593326e-01 3.90555322e-01
1.79292366e-01 4.70382005e-01 1.27392328e+00 -1.24848032e+00
1.16094005e+00 -1.00339186e+00 -2.63426065e-01 9.55186963e-01
-6.39375150e-01 1.39244989e-01 4.66377079e-01 9.59608704e-02
-1.61467528e+00 -7.34513044e-01 -3.07256252e-01 5.13476491e-01
-4.33396772e-02 1.18991721e+00 -3.66981089e-01 -1.37523204e-01
9.17847574e-01 -1.16699524e-01 -4.85041082e-01 -4.83856916e-01
3.15225929e-01 6.06779933e-01 1.71500072e-01 -7.50583291e-01
4.84292448e-01 1.78124160e-01 -2.62227118e-01 -6.63844585e-01
-1.31697536e+00 -7.27370739e-01 -1.32194042e-01 -1.51893839e-01
6.21313453e-01 -7.12296426e-01 -8.19946647e-01 -1.77484736e-01
-1.04259503e+00 5.09835482e-01 -8.45279992e-02 5.27259707e-01
1.84255466e-01 1.66370615e-01 -2.14512274e-01 -7.24969625e-01
-9.80367810e-02 -1.43426865e-01 2.71859348e-01 6.64266706e-01
5.23938127e-02 -1.19520891e+00 4.59598631e-01 7.45904088e-01
4.80988234e-01 7.66182989e-02 1.09663093e+00 -1.01306713e+00
-1.04033387e+00 -4.24472630e-01 -2.15122133e-01 -2.25987136e-01
1.04654446e-01 -3.21190029e-01 -5.98540783e-01 6.01661980e-01
1.16472498e-01 -1.71804979e-01 7.30575860e-01 6.30562156e-02
8.51923108e-01 -5.71835756e-01 -4.83071685e-01 1.19845524e-01
1.61623383e+00 2.97964245e-01 1.00196135e+00 5.49127996e-01
3.61840129e-01 7.38733113e-01 1.00806046e+00 1.07859805e-01
9.64440107e-01 2.30507448e-01 3.64874229e-02 6.28129601e-01
4.57446501e-02 -5.06202042e-01 -4.32147801e-01 8.43283176e-01
-5.12427092e-01 -1.71504810e-01 -8.82803798e-01 6.55373335e-01
-1.76325548e+00 -1.11726439e+00 -4.03521717e-01 1.66380787e+00
1.00628388e+00 -8.84894002e-03 2.30682176e-03 1.25938147e-01
6.69841409e-01 -5.75319342e-02 -1.47398070e-01 -2.63184309e-03
-2.94549316e-01 2.74286807e-01 6.00883663e-02 6.19792044e-01
-5.76542258e-01 1.03006542e+00 6.84040022e+00 1.00461292e+00
-4.63962317e-01 7.17845410e-02 2.12727457e-01 4.97453332e-01
-7.03574240e-01 4.53084052e-01 -1.12191832e+00 4.08245265e-01
6.50612712e-01 -7.28286624e-01 1.36971790e-02 6.78861201e-01
-3.38075042e-01 -7.29555309e-01 -3.36614192e-01 5.15416622e-01
7.83282891e-02 -1.28987169e+00 2.73912489e-01 -2.39272565e-01
7.05470920e-01 -5.48204839e-01 -6.69536471e-01 7.25217342e-01
4.49351460e-01 -7.76259124e-01 -1.79492682e-01 8.29605281e-01
1.20096095e-01 -7.36238956e-01 1.11671793e+00 5.69880247e-01
-1.06537807e+00 -5.03780097e-02 -4.75652695e-01 -3.32209766e-01
2.07180426e-01 9.37824368e-01 -8.44365060e-01 1.14798212e+00
6.47539675e-01 -1.05289616e-01 -6.74486816e-01 1.17956400e+00
-6.11668646e-01 8.08373153e-01 -3.84851277e-01 -5.78474343e-01
-2.46621251e-01 3.46024930e-02 1.21256396e-01 9.38479304e-01
1.47838473e-01 1.17277241e+00 1.78766102e-01 2.67735451e-01
-1.64793819e-01 6.53801322e-01 -4.44420397e-01 2.47793019e-01
6.91904008e-01 9.44378912e-01 -7.14672267e-01 -6.21831894e-01
-5.90101004e-01 2.74548352e-01 3.97158086e-01 3.03755522e-01
-5.57684243e-01 -8.83653164e-01 -8.69932249e-02 2.46330574e-01
3.60255957e-01 7.09685683e-02 -9.91792902e-02 -1.22187567e+00
3.24033797e-01 -6.13823414e-01 1.06438291e+00 -7.88673699e-01
-1.17087901e+00 5.00617146e-01 2.50915855e-01 -5.69027066e-01
6.36808872e-02 -3.13027024e-01 -6.81296825e-01 7.40885317e-01
-1.77469420e+00 -7.38855243e-01 -3.07372451e-01 7.34453380e-01
-2.15812370e-01 3.60560082e-02 8.03002775e-01 4.59639311e-01
-3.00596595e-01 2.92425483e-01 -3.87924731e-01 1.27974659e-01
4.95095938e-01 -1.15914571e+00 -2.04206973e-01 5.11343479e-01
1.58183143e-01 1.03599262e+00 6.74557269e-01 -8.49703491e-01
-1.09762919e+00 -6.51928782e-01 1.42521405e+00 -7.94943571e-01
5.35106957e-01 4.35732901e-01 -1.21995270e+00 5.30312955e-01
3.77735734e-01 -9.29932743e-02 9.54151034e-01 7.27890670e-01
-4.54852045e-01 -3.06653082e-01 -1.35803914e+00 2.95339525e-01
1.01919031e+00 -4.61432636e-01 -1.45251119e+00 3.75679165e-01
1.00534439e+00 -1.70835897e-01 -1.09053373e+00 5.28348088e-01
1.93606928e-01 -7.40020871e-01 9.42213416e-01 -7.46569335e-01
1.79803461e-01 -6.50415957e-01 -7.67509192e-02 -8.90107632e-01
-2.85266638e-01 -2.41088897e-01 -4.77668405e-01 1.05071199e+00
7.95972168e-01 -9.18080986e-01 8.21967542e-01 7.73575842e-01
2.50459224e-01 -7.79430211e-01 -8.72776926e-01 -6.65351987e-01
-1.37664929e-01 2.96452940e-01 8.89023721e-01 1.21424830e+00
4.47340608e-01 8.78344417e-01 2.40240872e-01 3.54652643e-01
-3.04717094e-01 6.16895139e-01 4.88609731e-01 -1.38848805e+00
-1.19219556e-01 -1.19176526e-02 -2.71395743e-01 -1.08662093e+00
-2.63448089e-01 -6.14185452e-01 -4.39252406e-01 -2.13893557e+00
8.08146670e-02 -3.11652869e-01 -2.25436389e-01 3.02962899e-01
-1.06788218e+00 -3.96991372e-01 -1.95462227e-01 -2.36546788e-02
-1.04426575e+00 4.66515213e-01 1.57904077e+00 3.72988760e-01
-2.11313833e-02 -1.84839126e-02 -1.00738347e+00 6.38019860e-01
9.45516765e-01 -4.22032028e-01 -5.45514882e-01 -2.57760018e-01
8.78310442e-01 4.36076313e-01 -6.87868968e-02 -8.13709557e-01
5.91710031e-01 -4.84601617e-01 -9.62642860e-03 -5.93884110e-01
1.03647508e-01 -7.45473742e-01 -2.90598780e-01 3.99472684e-01
-1.10953219e-01 2.45301813e-01 1.88326016e-02 5.54224193e-01
-8.28091741e-01 -5.84685564e-01 -2.66677532e-02 -3.88499677e-01
-9.21891809e-01 2.54248470e-01 -3.98053341e-02 8.40124965e-01
8.18522692e-01 5.33695109e-02 -6.67360127e-01 -4.28648353e-01
-4.13856685e-01 5.13556957e-01 -3.23402062e-02 2.80333340e-01
4.51732129e-01 -1.18633556e+00 -7.19420493e-01 -5.32500029e-01
3.35302293e-01 -3.82438116e-02 2.94822156e-01 2.94048816e-01
-7.20112503e-01 7.88783371e-01 1.02948688e-01 1.75154656e-01
-1.01326358e+00 4.87384200e-01 2.47928530e-01 -4.62316245e-01
1.04703195e-02 8.99094164e-01 -3.05367380e-01 -5.46185970e-01
-1.48994505e-01 -2.43050218e-01 -1.01031840e+00 4.43552807e-02
6.19530261e-01 3.94831151e-01 2.24892348e-02 2.70853247e-02
-3.00740898e-01 3.62715334e-01 -1.01335421e-01 -2.45524310e-02
7.50251055e-01 -7.89605305e-02 -5.18616617e-01 3.29978585e-01
7.66257524e-01 4.00138408e-01 -1.10652059e-01 -3.26965392e-01
3.73650849e-01 -7.35681415e-01 -1.64891899e-01 -1.27605271e+00
-6.61655903e-01 2.18213305e-01 1.17315792e-01 1.05215514e+00
9.68008518e-01 2.99735755e-01 9.55975652e-01 9.11130011e-01
6.09083235e-01 -1.07577038e+00 -2.03802928e-01 8.72687399e-01
8.10868502e-01 -1.31582427e+00 1.92497984e-01 -1.02714682e+00
-4.01831776e-01 7.68238425e-01 9.09984589e-01 -5.42755239e-02
1.02455795e+00 -2.97440439e-01 3.31253290e-01 -8.00523877e-01
-8.87405753e-01 -7.59713829e-01 3.13241869e-01 4.07062799e-01
5.73948205e-01 -1.25247343e-02 -1.07610047e+00 8.23788941e-01
-5.36591947e-01 4.30950612e-01 3.69085580e-01 1.15000308e+00
-1.13139021e+00 -1.20879352e+00 -5.37479818e-01 8.83993626e-01
-8.12679291e-01 -2.58110374e-01 -5.00329077e-01 8.37110996e-01
3.13509345e-01 1.31947315e+00 -2.97828406e-01 -9.05813947e-02
6.76163256e-01 3.43225896e-01 5.30389190e-01 -7.81444013e-01
-5.96344829e-01 -9.34864521e-01 6.21959448e-01 3.27007473e-02
-6.30828381e-01 -2.64867276e-01 -1.32485175e+00 -2.71541595e-01
-9.65994000e-01 1.01644230e+00 2.98474193e-01 1.17928028e+00
4.47917163e-01 2.08234668e-01 4.97185349e-01 9.06108320e-01
-1.42269567e-01 -1.12113285e+00 -4.75688696e-01 2.60414481e-01
-5.42539693e-02 -6.98194206e-01 -2.84970611e-01 -1.32082775e-01] | [10.539799690246582, 7.932443618774414] |
4575156f-7386-4c0d-a207-bd4eca7c45c4 | social-honeypot-for-humans-luring-people | 2303.17946 | null | https://arxiv.org/abs/2303.17946v1 | https://arxiv.org/pdf/2303.17946v1.pdf | Social Honeypot for Humans: Luring People through Self-managed Instagram Pages | Social Honeypots are tools deployed in Online Social Networks (OSN) to attract malevolent activities performed by spammers and bots. To this end, their content is designed to be of maximum interest to malicious users. However, by choosing an appropriate content topic, this attractive mechanism could be extended to any OSN users, rather than only luring malicious actors. As a result, honeypots can be used to attract individuals interested in a wide range of topics, from sports and hobbies to more sensitive subjects like political views and conspiracies. With all these individuals gathered in one place, honeypot owners can conduct many analyses, from social to marketing studies. In this work, we introduce a novel concept of social honeypot for attracting OSN users interested in a generic target topic. We propose a framework based on fully-automated content generation strategies and engagement plans to mimic legit Instagram pages. To validate our framework, we created 21 self-managed social honeypots (i.e., pages) on Instagram, covering three topics, four content generation strategies, and three engaging plans. In nine weeks, our honeypots gathered a total of 753 followers, 5387 comments, and 15739 likes. These results demonstrate the validity of our approach, and through statistical analysis, we examine the characteristics of effective social honeypots. | ['Pier Paolo Tricomi', 'Luca Pajola', 'Mauro Conti', 'Sara Bardi'] | 2023-03-31 | null | null | null | null | ['marketing'] | ['miscellaneous'] | [-2.60940760e-01 1.00615434e-01 -4.14333373e-01 9.05724522e-03
-5.70049621e-02 -8.63689542e-01 1.08203065e+00 1.37569413e-01
-4.31519210e-01 8.45534027e-01 1.48577066e-02 -1.15672266e-02
-1.04932643e-01 -1.21738398e+00 1.52569547e-01 -2.56443024e-01
-1.09834045e-01 5.57451844e-01 6.20739579e-01 -4.25849050e-01
6.02898538e-01 4.96422112e-01 -1.48802412e+00 -4.22952278e-03
8.21949780e-01 5.86293995e-01 -2.34787017e-01 4.37259406e-01
-3.25442374e-01 7.07751930e-01 -1.16477835e+00 -5.90239823e-01
1.05719730e-01 -4.60869461e-01 -7.69649386e-01 -3.14017013e-03
-1.37813494e-01 -3.95651519e-01 -1.42374068e-01 9.37411249e-01
4.27934051e-01 1.60749912e-01 3.82145256e-01 -1.60624301e+00
-2.47538641e-01 1.02214682e+00 -8.40648234e-01 2.29497388e-01
4.32099253e-01 4.36874121e-01 1.18998039e+00 -4.08677965e-01
7.52489626e-01 1.37492132e+00 3.38370055e-01 4.05933797e-01
-9.69789207e-01 -8.05587649e-01 -2.20798507e-01 -3.09014738e-01
-8.93438518e-01 -2.95726031e-01 8.93519104e-01 -2.28797942e-01
2.05630600e-01 4.94119555e-01 1.11939168e+00 1.45214391e+00
-2.00799435e-01 6.60909414e-01 1.45885181e+00 5.65250739e-02
5.26103079e-01 7.02354968e-01 3.84487718e-01 3.43882531e-01
7.08060980e-01 -3.27355921e-01 -6.46638215e-01 -9.98342216e-01
3.09831649e-01 6.23092130e-02 2.29303688e-01 -1.04695326e-02
-7.78974831e-01 1.40887642e+00 4.17082071e-01 5.91482282e-01
-3.48756015e-01 -1.73522949e-01 3.50454539e-01 1.88938558e-01
6.44962728e-01 8.61023188e-01 -1.42569933e-02 -5.37391424e-01
-6.20429575e-01 2.99645841e-01 1.40896273e+00 2.81144917e-01
9.43841279e-01 -3.64340574e-01 -1.47385314e-01 1.01258361e+00
6.35339499e-01 5.91484070e-01 5.75113714e-01 -9.49018836e-01
7.36781806e-02 1.18372452e+00 1.44200802e-01 -1.52189350e+00
-2.99641907e-01 -3.52047831e-01 -4.25130039e-01 -7.56829903e-02
4.37000275e-01 -4.58438396e-01 -1.29623234e-01 1.40885365e+00
6.72519207e-01 -1.93709403e-01 -5.11354923e-01 6.38781726e-01
7.36515701e-01 5.05558133e-01 -7.56758079e-03 -2.74700433e-01
1.65675557e+00 -6.00589871e-01 -5.04023254e-01 -1.51053876e-01
5.52282453e-01 -6.56196535e-01 1.24331188e+00 1.23267442e-01
-9.15456176e-01 9.52019915e-02 -5.60232401e-01 5.91764092e-01
-7.71525502e-01 -4.17229474e-01 6.09925508e-01 1.03537834e+00
-8.73211801e-01 7.71790087e-01 -1.79482803e-01 -9.15215254e-01
4.63374108e-01 2.39236385e-01 1.87596813e-01 4.86189276e-01
-1.26321578e+00 4.57046121e-01 -4.27219197e-02 -4.57075775e-01
-4.28641826e-01 -2.85987705e-01 -1.75920203e-01 -1.95182767e-02
6.32422805e-01 -2.83194065e-01 1.10181987e+00 -6.39439940e-01
-1.52713883e+00 9.40747797e-01 3.63091767e-01 -4.87881064e-01
6.34520650e-01 -8.18107799e-02 -2.66953230e-01 5.42091072e-01
6.11689687e-01 4.01904106e-01 1.01532888e+00 -1.23650002e+00
-4.01407421e-01 -1.54420614e-01 3.08440685e-01 -1.09923080e-01
-9.64459002e-01 5.53798795e-01 2.19406098e-01 -4.70506102e-01
-2.59001404e-01 -9.21842635e-01 -1.38961464e-01 -3.50677282e-01
-8.10360193e-01 -3.34236532e-01 1.40817595e+00 -3.11044693e-01
1.18901467e+00 -2.00699306e+00 -3.99896801e-01 5.83542049e-01
6.17002904e-01 6.34722471e-01 7.31175318e-02 7.99857080e-01
7.03996956e-01 8.35687995e-01 1.13803633e-01 -1.46124586e-01
1.92818448e-01 -1.67981118e-01 -2.16742769e-01 1.35478050e-01
-1.07001081e-01 8.76192033e-01 -1.20216823e+00 -8.64663899e-01
-4.17789966e-02 1.36645228e-01 -3.65567237e-01 1.91029236e-01
-2.37768054e-01 2.12509081e-01 -9.91994500e-01 6.67795122e-01
4.75367665e-01 -5.49495220e-01 2.39501178e-01 6.14066064e-01
-9.41681042e-02 3.01801801e-01 -8.28305244e-01 5.62358320e-01
-1.98143139e-01 6.02081478e-01 2.20805064e-01 -6.26691729e-02
1.12885058e+00 -2.33820394e-01 7.06609309e-01 -3.11464649e-02
5.70348799e-01 3.38688970e-01 -1.15846112e-01 -1.45132244e-01
7.39343405e-01 3.53217900e-01 -2.27066845e-01 1.58333540e+00
-2.71803737e-01 -4.34211008e-02 4.76917773e-01 6.32833719e-01
1.31697655e+00 -4.69654411e-01 3.85239393e-01 -9.88617837e-02
3.34849656e-01 -4.36489098e-02 -4.06218991e-02 1.09186435e+00
-5.59617996e-01 -1.61641352e-02 9.40587997e-01 -2.98298597e-01
-7.15827048e-01 -8.73131514e-01 2.49859348e-01 1.23949528e+00
2.88779706e-01 -5.53026438e-01 -1.02484322e+00 -1.11417997e+00
-1.72899738e-01 3.27879578e-01 -3.80725980e-01 4.31847237e-02
-5.20272613e-01 -8.98453772e-01 6.09330118e-01 -5.76649010e-01
1.16752255e+00 -1.66015148e+00 -5.96318305e-01 3.44764680e-01
-3.23097646e-01 -9.38050628e-01 -4.09881711e-01 -5.42168736e-01
-4.14661229e-01 -1.43122017e+00 -7.77988851e-01 -2.45455414e-01
4.04822320e-01 8.08390796e-01 9.02657688e-01 5.52423418e-01
-2.66066790e-01 4.16222602e-01 -7.67828286e-01 -2.77095228e-01
-6.69973254e-01 6.21809125e-01 3.12976241e-01 5.14975727e-01
4.80023235e-01 -8.70682657e-01 -5.21466553e-01 7.97013462e-01
-8.69274974e-01 -5.89134395e-01 3.50967467e-01 4.50306803e-01
-5.92190862e-01 -9.31063220e-02 7.30670869e-01 -1.10858369e+00
1.20717502e+00 -8.77904475e-01 -2.16495723e-01 -3.91622074e-02
-3.90909702e-01 -5.56713283e-01 5.36419213e-01 -9.58861470e-01
-7.90342271e-01 -4.92950767e-01 2.79621392e-01 3.00222516e-01
2.44788993e-02 6.25546575e-02 1.42763004e-01 -2.93228239e-01
1.01168311e+00 -5.88564314e-02 3.51031333e-01 -3.43739390e-01
3.53869915e-01 1.08639622e+00 -5.72343051e-01 -4.21764195e-01
1.30415523e+00 5.11633575e-01 -3.89097452e-01 -1.43752289e+00
-6.24858081e-01 -4.61330414e-01 6.03051148e-02 -6.03665590e-01
4.08331811e-01 -9.74303037e-02 -1.24061394e+00 9.15956557e-01
-1.11340117e+00 -1.43873930e-01 -4.44603324e-01 6.76342249e-02
-6.37980774e-02 5.60438633e-01 -8.26078594e-01 -9.11233723e-01
-4.67173219e-01 -5.56096494e-01 6.13490164e-01 6.81748211e-01
-7.24964678e-01 -7.85628259e-01 2.44848952e-01 5.34574449e-01
9.11178291e-01 2.43605331e-01 1.08337134e-01 -1.24738479e+00
-6.37214959e-01 -4.55440283e-01 -5.70178926e-01 -1.61392868e-01
2.71325648e-01 2.37818152e-01 -7.56300628e-01 -1.02664694e-01
-2.24962104e-02 -5.73935628e-01 3.57803851e-01 -4.04666997e-02
4.81572568e-01 -7.83721089e-01 -5.22288442e-01 -1.90758392e-01
8.11627746e-01 -6.71367049e-02 5.25795639e-01 4.13885504e-01
3.28632474e-01 1.08787763e+00 5.60459614e-01 8.98724973e-01
1.81848660e-01 4.99432445e-01 5.72674811e-01 -5.47023490e-02
3.51062089e-01 -5.11371493e-01 4.82028157e-01 2.33627751e-01
-1.39148861e-01 -3.13141644e-01 -5.47547221e-01 2.98099816e-01
-1.49024212e+00 -1.01394856e+00 -1.53979376e-01 1.70755446e+00
6.55316234e-01 2.74983793e-01 8.70654821e-01 -4.11489829e-02
1.10507143e+00 6.96215332e-01 -3.98193568e-01 -3.33996952e-01
1.04965560e-01 1.08195484e-01 2.56894559e-01 4.33507264e-01
-7.66500711e-01 1.02886581e+00 5.63295317e+00 1.03041685e+00
-7.04451919e-01 1.79402247e-01 6.68560147e-01 6.54305965e-02
-6.07658215e-02 2.22101808e-01 -9.93089736e-01 8.12853515e-01
8.90278876e-01 -3.31241965e-01 6.12389147e-01 9.37940300e-01
4.93990511e-01 -2.78527737e-01 -2.53534257e-01 5.39038539e-01
-1.27516594e-02 -1.29934609e+00 -2.13081285e-01 6.81765795e-01
4.79465008e-01 -1.96795419e-01 -1.23262014e-02 1.57210365e-01
7.47886539e-01 -4.81436998e-01 1.13854101e-02 1.04418479e-01
1.53733358e-01 -4.45926666e-01 4.57635134e-01 4.96168375e-01
-8.29097688e-01 -3.03605855e-01 -8.67359415e-02 -1.14877410e-01
3.93019199e-01 7.97755122e-01 -8.56364965e-01 -6.62356839e-02
6.03079021e-01 5.04179716e-01 -8.72110903e-01 8.98209214e-01
-3.71537685e-01 7.62684941e-01 -4.92972642e-01 -1.09028709e+00
3.19500089e-01 -2.50225514e-01 9.78158474e-01 7.53740966e-01
1.21831968e-01 3.29004042e-02 -1.63478572e-02 1.00978673e+00
-4.48482633e-02 4.30874586e-01 -9.70318139e-01 -5.61510265e-01
7.51157284e-01 1.83467972e+00 -1.21653318e+00 -2.33392507e-01
3.21307600e-01 6.63837671e-01 1.21297993e-01 2.23408267e-01
-7.77859926e-01 -7.91207790e-01 6.06471002e-01 7.52721429e-01
-3.60576242e-01 -5.34801520e-02 -4.72892029e-03 -9.61554348e-01
-2.62259454e-01 -1.14210629e+00 1.85582981e-01 -4.15226966e-01
-1.58721066e+00 6.11540735e-01 -6.46009445e-02 -1.08331394e+00
-1.95614874e-01 -1.77252516e-01 -1.07203221e+00 2.83478677e-01
-8.29647005e-01 -1.18476951e+00 -3.55696231e-01 1.87952563e-01
6.06490225e-02 -3.48124266e-01 2.23200113e-01 -7.75873959e-02
-3.68905574e-01 2.96536922e-01 -3.43418390e-01 2.53946751e-01
4.74474251e-01 -7.55992770e-01 3.21335435e-01 4.86372054e-01
9.07865465e-02 9.15994287e-01 6.04908764e-01 -8.71664047e-01
-8.57078791e-01 -7.35536873e-01 1.00473440e+00 -4.91679698e-01
1.14901137e+00 -4.62210298e-01 -5.13054848e-01 2.27215767e-01
2.06228256e-01 -5.37826061e-01 6.95958555e-01 2.52905097e-02
-1.89921856e-02 2.08056629e-01 -1.50533628e+00 1.01318431e+00
9.60399628e-01 -2.51825988e-01 -5.23892678e-02 7.49265194e-01
7.14805782e-01 2.29392201e-01 -3.99616241e-01 -3.51490170e-01
4.75509793e-01 -1.31249547e+00 7.23486602e-01 -4.52141881e-01
4.74757701e-01 9.78343636e-02 4.13429171e-01 -1.00146139e+00
1.34208888e-01 -1.40995419e+00 -2.69904852e-01 1.57394850e+00
2.74259359e-01 -1.21796477e+00 1.12099707e+00 6.22225404e-01
5.36030710e-01 -4.61608768e-01 -5.17578542e-01 -6.73877299e-01
-4.06201929e-01 1.42301634e-01 6.94546402e-01 1.11786640e+00
1.54714420e-01 5.00389159e-01 -5.41467786e-01 -4.58119422e-01
8.42969060e-01 -2.46461064e-01 1.18533802e+00 -1.55302370e+00
-3.82297426e-01 -5.34859002e-01 1.40370578e-02 -8.58899891e-01
-3.32902633e-02 -4.26894277e-01 -4.95341808e-01 -1.09331608e+00
1.42114326e-01 -7.04111040e-01 4.46732432e-01 3.84654552e-01
1.09528869e-01 6.66196048e-01 3.22518557e-01 3.91764760e-01
-7.01164544e-01 8.47113654e-02 1.25685537e+00 -8.32905322e-02
-7.92227566e-01 4.71604735e-01 -1.00786746e+00 8.50692391e-01
1.07800698e+00 -6.27205431e-01 -1.49712950e-01 6.97321475e-01
4.53199178e-01 -3.53599221e-01 3.82155716e-01 -6.66695952e-01
-5.22995815e-02 -3.55358839e-01 -2.96359271e-01 -2.97986209e-01
3.57903212e-01 -3.73460025e-01 -1.06954418e-01 7.55795777e-01
-2.36481145e-01 -4.16805178e-01 -5.65615892e-01 5.11583388e-01
2.12386280e-01 -5.49102247e-01 8.34692895e-01 -3.88103187e-01
1.05839789e-01 2.54459381e-01 -7.56469488e-01 4.61009055e-01
1.33102369e+00 -3.52418691e-01 -9.53866720e-01 -8.19481611e-01
-5.24122596e-01 1.75002158e-01 6.90726101e-01 3.37040603e-01
4.05653059e-01 -1.06620824e+00 -5.83850145e-01 4.55275774e-02
-8.30070600e-02 -7.33489811e-01 -1.18780456e-01 6.08453631e-01
-4.53856021e-01 -1.17384069e-01 -2.54940301e-01 -3.30439955e-01
-1.20018685e+00 1.52551278e-01 -6.10412136e-02 -5.92185318e-01
-3.11697811e-01 4.13805962e-01 -4.14494395e-01 -5.49014270e-01
2.03863159e-02 4.65575635e-01 -6.47119880e-01 6.39670432e-01
7.64451504e-01 8.33053648e-01 -5.85362315e-01 -6.08752549e-01
-1.26754537e-01 9.95632336e-02 -1.05044790e-01 -2.88412601e-01
1.01960111e+00 2.06494108e-02 -5.60154259e-01 1.20470338e-01
9.46882606e-01 4.01914805e-01 -4.78471190e-01 -4.24265265e-02
1.46227300e-01 -8.24556530e-01 -7.61432648e-01 -3.28688830e-01
-9.19828475e-01 4.46286261e-01 3.69927138e-02 1.58972228e+00
4.54610974e-01 1.46213606e-01 1.10372806e+00 2.95768827e-01
6.68688536e-01 -1.09705448e+00 8.74897122e-01 3.99042577e-01
4.98487830e-01 -1.11791837e+00 -1.09266266e-01 -5.81864417e-01
-6.22952223e-01 7.11325049e-01 7.20964432e-01 -1.65308818e-01
7.21499681e-01 -2.86340684e-01 -3.98396589e-02 -4.87312883e-01
-4.51153994e-01 -1.05969325e-01 -3.54005992e-01 7.01673210e-01
-4.57372032e-02 1.21290865e-03 -8.50320876e-01 3.45947027e-01
-2.89656311e-01 -4.25165713e-01 1.08841372e+00 8.89793873e-01
-8.80964220e-01 -1.32321870e+00 -3.77672583e-01 8.06113720e-01
-5.55926204e-01 1.83508098e-01 -1.31506300e+00 7.57057190e-01
-2.88737983e-01 1.34015799e+00 -5.25482476e-01 -5.55100918e-01
-2.42167804e-02 -3.64082277e-01 -4.83200669e-01 -7.85922408e-01
-1.15277386e+00 -2.95301706e-01 4.45175350e-01 -3.78006130e-01
-4.48145509e-01 -4.73729312e-01 -6.71244860e-01 -9.03018355e-01
-4.40432042e-01 5.72340488e-01 5.79410076e-01 6.25982583e-01
2.71280289e-01 -3.40593576e-01 1.32840049e+00 -8.32303226e-01
-6.82006001e-01 -1.08167648e+00 -8.57568860e-01 3.41417134e-01
-2.81708956e-01 -5.85380852e-01 -9.88753378e-01 -6.27098203e-01] | [8.122415542602539, 10.106975555419922] |
bbb150b1-1000-4592-a49c-b62e47147daa | real-time-simultaneous-multi-object-3d-shape | 2305.09510 | null | https://arxiv.org/abs/2305.09510v1 | https://arxiv.org/pdf/2305.09510v1.pdf | Real-time Simultaneous Multi-Object 3D Shape Reconstruction, 6DoF Pose Estimation and Dense Grasp Prediction | Robotic manipulation systems operating in complex environments rely on perception systems that provide information about the geometry (pose and 3D shape) of the objects in the scene along with other semantic information such as object labels. This information is then used for choosing the feasible grasps on relevant objects. In this paper, we present a novel method to provide this geometric and semantic information of all objects in the scene as well as feasible grasps on those objects simultaneously. The main advantage of our method is its speed as it avoids sequential perception and grasp planning steps. With detailed quantitative analysis, we show that our method delivers competitive performance compared to the state-of-the-art dedicated methods for object shape, pose, and grasp predictions while providing fast inference at 30 frames per second speed. | ['Volkan Isler', 'Jinwook Huh', 'Selim Engin', 'Isaac Kasahara', 'Nikhil Chavan-Dafle', 'Shubham Agrawal'] | 2023-05-16 | null | null | null | null | ['3d-shape-reconstruction'] | ['computer-vision'] | [ 4.08436209e-02 -1.69773951e-01 -1.10115223e-01 -3.28641772e-01
-4.29652959e-01 -8.51200044e-01 2.53434777e-01 7.17267334e-01
-2.75295913e-01 2.53323168e-01 -1.45013139e-01 1.81970045e-01
-3.76665533e-01 -7.77404964e-01 -8.79990935e-01 -4.87498999e-01
-2.83711791e-01 1.17515147e+00 7.08527625e-01 -1.38409331e-01
6.47121847e-01 1.04358780e+00 -1.82226884e+00 1.00931242e-01
5.29301405e-01 1.30791318e+00 1.10701895e+00 6.76338613e-01
1.06329203e-01 3.07904392e-01 -1.92093149e-01 -5.23751490e-02
3.11622202e-01 4.99516577e-01 -8.73440802e-01 2.28106484e-01
2.11385757e-01 -7.03577816e-01 -2.43834749e-01 9.43211317e-01
1.03102475e-01 3.02995652e-01 6.66117847e-01 -1.10133779e+00
-1.42031699e-01 4.04384375e-01 -2.74525046e-01 -3.28519493e-01
5.68848491e-01 2.44717345e-01 9.13906932e-01 -9.52016056e-01
7.36798584e-01 1.55822778e+00 1.05751708e-01 2.09533945e-01
-1.07937014e+00 -7.40012378e-02 3.57206821e-01 2.87996233e-01
-1.04899395e+00 -2.12630510e-01 6.60795748e-01 -4.62786078e-01
8.48968625e-01 1.40897483e-01 5.39323390e-01 6.11095667e-01
2.49934375e-01 8.28022838e-01 7.94675231e-01 -3.79078865e-01
5.59002042e-01 -2.16536492e-01 9.06934515e-02 8.49809706e-01
1.75755039e-01 -4.68250029e-02 -2.98041523e-01 -2.17519905e-02
1.12309349e+00 2.86034793e-01 1.50475591e-01 -9.55672801e-01
-1.50403512e+00 3.42644662e-01 7.53408074e-01 -2.01492488e-01
-7.85254359e-01 4.10187900e-01 2.83831567e-01 -1.43391937e-01
-1.71468575e-02 4.02800739e-01 -4.77687985e-01 -9.80167240e-02
-1.57020524e-01 6.39420986e-01 8.93336415e-01 1.57775259e+00
8.87339234e-01 -6.29059553e-01 -1.22986719e-01 5.43498516e-01
3.38757604e-01 6.63571060e-01 -5.19519985e-01 -1.51182425e+00
6.45683885e-01 5.49523294e-01 6.64873481e-01 -1.07873797e+00
-4.99004215e-01 3.03388089e-01 -1.85562178e-01 4.64513004e-01
4.35803205e-01 3.81828815e-01 -1.12069499e+00 1.21366620e+00
4.65432853e-01 -4.91465569e-01 -3.09919566e-01 1.12345290e+00
4.12189543e-01 3.91470134e-01 2.58446522e-02 2.61125892e-01
1.40198922e+00 -9.60395336e-01 -4.48609442e-01 -1.49278164e-01
-8.21911842e-02 -9.28120077e-01 7.19726264e-01 4.95519966e-01
-1.00135720e+00 -4.70481515e-01 -6.23536110e-01 -2.19363868e-01
-3.38130742e-01 3.86806041e-01 9.61307108e-01 5.93162514e-02
-7.67815828e-01 8.37142527e-01 -1.15295017e+00 -4.47978526e-01
3.43937188e-01 6.71013951e-01 -2.52771348e-01 -4.16616470e-01
-1.84574232e-01 1.36263180e+00 6.41372800e-01 2.60260195e-01
-1.22916007e+00 -1.90783173e-01 -8.92291009e-01 1.34194627e-01
9.24898446e-01 -2.50892937e-01 1.42132533e+00 -1.35565564e-01
-1.58319461e+00 6.23111546e-01 -2.07668483e-01 -5.27218096e-02
4.57108885e-01 -5.34632504e-01 5.64434290e-01 4.41155076e-01
-4.58459631e-02 9.22137678e-01 6.32149816e-01 -1.71050513e+00
-6.01016760e-01 -7.03115106e-01 5.84332407e-01 3.50843370e-01
4.15951759e-01 -7.22416267e-02 -6.92666709e-01 -1.45366207e-01
7.56148815e-01 -1.00128758e+00 -4.05792087e-01 6.25530720e-01
-5.52330732e-01 -5.46496630e-01 7.93499231e-01 -6.06001377e-01
2.07681686e-01 -1.78919327e+00 4.40634727e-01 3.03782642e-01
-7.23324716e-02 -2.11151857e-02 -1.05672270e-01 7.59009004e-01
6.07073426e-01 -5.38369060e-01 1.19996458e-01 -2.29146957e-01
2.67798156e-01 4.64262813e-01 -4.40981388e-01 4.57154304e-01
1.28230348e-01 8.70766401e-01 -1.26094043e+00 -3.58582944e-01
8.68052185e-01 2.36747935e-01 -5.14728665e-01 4.41924363e-01
-7.71107435e-01 6.91175580e-01 -8.54056418e-01 8.71634424e-01
6.90626323e-01 9.42495242e-02 2.02469394e-01 -4.17833686e-01
-2.53558189e-01 4.99824136e-01 -1.24536610e+00 1.92319322e+00
-3.60276371e-01 -3.04502230e-02 3.88711005e-01 -7.72567332e-01
1.05791259e+00 8.52300748e-02 4.80884761e-01 -8.70296583e-02
1.83021948e-01 3.17610919e-01 -2.91100621e-01 -6.11826301e-01
4.64658409e-01 3.05183828e-01 -4.44384366e-02 1.14073418e-01
1.27551835e-02 -8.06335568e-01 2.19947964e-01 -1.63594350e-01
7.79475152e-01 6.43544495e-01 1.38294563e-01 -3.12050670e-01
7.96684027e-02 2.04069853e-01 4.52226810e-02 6.92041099e-01
7.89713338e-02 3.32149833e-01 2.12699786e-01 -3.70643258e-01
-1.19300544e+00 -1.40819538e+00 1.81336328e-02 9.14196253e-01
9.09181654e-01 -1.50408089e-01 -5.88248014e-01 -2.99726814e-01
2.55112976e-01 4.84027207e-01 -3.68075937e-01 2.39864022e-01
-6.43783867e-01 1.54630557e-01 -4.38408881e-01 8.87900233e-01
1.38429865e-01 -1.31742299e+00 -1.16179788e+00 1.43680602e-01
1.39445420e-02 -1.26225865e+00 -1.45494733e-02 1.93843752e-01
-1.04706562e+00 -1.35347247e+00 -3.59733343e-01 -8.33547711e-01
1.18717539e+00 3.87662053e-01 8.05290103e-01 -2.28496864e-02
-4.75385457e-01 4.53980535e-01 -4.90713090e-01 -4.15526360e-01
-2.31169805e-01 -1.85078576e-01 2.62013748e-02 -4.61145282e-01
-7.78991207e-02 -2.98069626e-01 -4.76161063e-01 2.91930497e-01
-4.38046157e-01 2.76419789e-01 5.55819809e-01 3.93448293e-01
8.25835764e-01 1.37997642e-01 -3.74222472e-02 -3.14296514e-01
9.47314650e-02 -1.06185861e-01 -9.95342493e-01 3.53460521e-01
-1.30791873e-01 7.31633902e-02 2.69605279e-01 -1.05078600e-01
-1.02680993e+00 6.57398105e-01 2.63936788e-01 -4.82617557e-01
-4.82475132e-01 2.84031868e-01 -1.57332778e-01 -5.71183898e-02
7.65724927e-02 -9.83101279e-02 -3.56784239e-02 -8.78601491e-01
4.10021484e-01 5.19360602e-01 6.45868480e-01 -1.01368415e+00
4.69008416e-01 5.31742811e-01 4.31494325e-01 -5.52618682e-01
-7.82665014e-01 -6.50833428e-01 -1.40768468e+00 -3.38100046e-01
7.18183756e-01 -5.33422112e-01 -1.39724815e+00 3.92515659e-01
-1.49687767e+00 -1.67631269e-01 -4.90320772e-02 5.96872807e-01
-1.11374140e+00 1.67168736e-01 -5.73503792e-01 -1.24117839e+00
-7.22568035e-02 -1.48692262e+00 1.58136761e+00 1.24982223e-02
-1.31862583e-02 -3.66577864e-01 -7.42622793e-01 2.68247873e-01
3.44323330e-02 3.25439453e-01 9.27745104e-01 -3.24672490e-01
-1.43128097e+00 -2.10846886e-01 -5.15464783e-01 -1.20356463e-01
4.26295698e-01 -5.98854907e-02 -4.83283162e-01 -3.04526180e-01
-3.86040241e-01 -2.95303047e-01 6.56708300e-01 4.05737698e-01
1.30711007e+00 -1.07673943e-01 -6.06000602e-01 5.02138920e-02
1.45653009e+00 1.75806046e-01 3.01922143e-01 -3.05116805e-03
6.19761407e-01 9.82259750e-01 1.36127067e+00 4.87606406e-01
3.70177835e-01 9.70527351e-01 1.12756300e+00 4.78361338e-01
1.01650335e-01 -2.26465181e-01 -2.27219332e-02 3.34685296e-01
-3.41661960e-01 -1.57159880e-01 -8.67981553e-01 6.38459921e-01
-2.04916215e+00 -5.23797929e-01 -2.93507054e-02 2.33263993e+00
5.27733862e-01 6.07809089e-02 4.92003299e-02 -9.92000923e-02
7.77873814e-01 -3.10011774e-01 -6.56288743e-01 -1.28961369e-01
5.49278140e-01 3.94180566e-02 7.14871228e-01 5.26529789e-01
-1.15751886e+00 1.06260872e+00 6.67057991e+00 4.80842203e-01
-6.10331953e-01 -3.07515621e-01 -7.22740442e-02 1.65669724e-01
9.53788608e-02 3.94350588e-02 -7.61622131e-01 7.73230866e-02
1.76757812e-01 3.24786663e-01 8.05660367e-01 1.06622636e+00
-4.11908664e-02 -5.84372640e-01 -1.52911830e+00 7.44364440e-01
-1.86765894e-01 -1.14706063e+00 -7.50977024e-02 -1.97840586e-01
2.78013647e-01 4.38077003e-02 -3.84744138e-01 -2.31746078e-01
4.33313638e-01 -6.40605569e-01 1.16456580e+00 5.60827732e-01
4.85828847e-01 -6.17205441e-01 5.86052954e-01 5.29789686e-01
-1.11270964e+00 -2.66980708e-01 -5.73538899e-01 -2.96873868e-01
2.08048314e-01 2.73019671e-01 -1.12683153e+00 5.57540596e-01
6.34465814e-01 5.15120745e-01 -1.45230461e-02 1.21061599e+00
-4.68397945e-01 -1.06538527e-01 -6.30539358e-01 -1.92249537e-01
6.77496195e-02 -1.25397712e-01 6.33516312e-01 8.57632637e-01
2.61468083e-01 3.10066134e-01 8.89310062e-01 1.02303696e+00
4.30902958e-01 -3.63775611e-01 -3.88669342e-01 -2.79426798e-02
8.09055924e-01 9.61643755e-01 -1.17442226e+00 -2.88719118e-01
7.93219209e-02 7.58276999e-01 3.68889511e-01 6.24694712e-02
-4.19889569e-01 -1.00822411e-01 6.46182239e-01 4.58233766e-02
5.94045579e-01 -9.31396365e-01 -2.86160052e-01 -6.34384036e-01
4.51473385e-01 -1.93775773e-01 -5.22705168e-02 -9.16727841e-01
-1.03359628e+00 2.48022571e-01 3.47446710e-01 -1.02928662e+00
4.29917946e-02 -1.27750289e+00 -8.72047395e-02 7.13770390e-01
-1.35730243e+00 -1.32966948e+00 -4.85146970e-01 1.99688435e-01
7.88897336e-01 2.57017523e-01 8.77302289e-01 -3.55336607e-01
8.97035450e-02 -3.64674479e-01 -1.54424414e-01 -6.71842545e-02
2.36053810e-01 -1.21016002e+00 3.42491388e-01 2.86208153e-01
-1.41854540e-01 6.36334479e-01 8.26020181e-01 -7.55009115e-01
-2.05974102e+00 -6.98062241e-01 4.83438462e-01 -5.78479648e-01
3.76725852e-01 -3.59572113e-01 -5.59780538e-01 7.08370864e-01
-3.77946347e-01 -1.22478314e-01 -2.75142759e-01 1.27908587e-01
-1.79098561e-01 9.31844935e-02 -1.21687472e+00 4.55136895e-01
1.08425426e+00 -1.06341042e-01 -7.00969636e-01 6.21408641e-01
7.14487076e-01 -1.09425449e+00 -8.73411417e-01 6.48871839e-01
7.34666467e-01 -7.38934696e-01 1.28400254e+00 -3.84992003e-01
3.42997342e-01 -3.06316167e-01 -4.44751143e-01 -9.49648976e-01
-2.71153510e-01 -3.44277591e-01 -2.06412494e-01 6.98632181e-01
-5.73001392e-02 -3.31223130e-01 6.43934846e-01 8.33472610e-01
-2.98855871e-01 -7.74079382e-01 -8.22474360e-01 -8.19588304e-01
-5.41105866e-01 -3.84458691e-01 6.16801083e-01 2.53034085e-01
1.50465351e-02 -2.35994607e-01 2.22855136e-02 6.12908900e-01
7.88472891e-01 8.09273124e-01 7.16304064e-01 -1.34406710e+00
4.00261804e-02 -2.33927935e-01 -5.17625570e-01 -1.26095605e+00
1.25451311e-01 -6.52386069e-01 7.14100301e-01 -1.88466096e+00
2.98873395e-01 -1.00102937e+00 -9.40916985e-02 7.37834692e-01
4.93568778e-02 -1.39653593e-01 5.75617492e-01 2.45259970e-01
-6.47932053e-01 4.60477650e-01 1.49287009e+00 -1.08546531e-02
-3.88627574e-02 6.97600022e-02 6.05444238e-02 7.95382440e-01
5.73220074e-01 -2.55522847e-01 -5.29485047e-02 -7.39488661e-01
-2.95295417e-01 2.98662007e-01 7.88748145e-01 -1.01517880e+00
1.15390308e-01 -6.02546811e-01 2.87493765e-01 -1.05949664e+00
8.83428216e-01 -1.17051125e+00 -1.68385059e-01 6.51123762e-01
-3.39111805e-01 -1.77040353e-01 1.34204313e-01 7.53499031e-01
2.71835685e-01 -5.13754725e-01 4.07614529e-01 -3.44613492e-01
-9.33390915e-01 3.86013865e-01 -1.01991072e-01 -7.39014208e-01
1.15373075e+00 -3.89382467e-02 -7.67326355e-02 -2.50681430e-01
-9.43872750e-01 4.59603846e-01 7.24750340e-01 7.78260887e-01
7.49775946e-01 -1.01900661e+00 -3.95646870e-01 5.20504862e-02
1.19746342e-01 5.42218804e-01 1.00172050e-02 5.03625154e-01
-7.04263687e-01 5.82210124e-01 -4.88382071e-01 -8.20451498e-01
-1.17358816e+00 7.92761445e-01 -2.63313979e-01 2.54006207e-01
-7.76386678e-01 7.12635994e-01 1.61980733e-01 -5.12883544e-01
4.79688466e-01 -5.91204226e-01 1.19215719e-01 -5.38941741e-01
2.09138826e-01 5.02549350e-01 -4.33560796e-02 -4.44537133e-01
-5.38823485e-01 7.05556393e-01 2.22671926e-01 3.65297198e-02
1.36733377e+00 -1.36593997e-01 -5.29790401e-01 4.50906932e-01
6.50048435e-01 -2.63551563e-01 -1.74688625e+00 -1.84974775e-01
9.98257101e-02 -9.01722908e-01 -1.75447851e-01 -9.72758114e-01
-6.17588758e-01 8.67996395e-01 1.53333917e-01 -7.17638955e-02
7.86885858e-01 5.16791165e-01 6.68306708e-01 8.38409424e-01
1.40997338e+00 -1.11912930e+00 3.31471041e-02 5.71593523e-01
1.33779728e+00 -1.25364089e+00 1.56315744e-01 -1.18592942e+00
-3.52910072e-01 1.31629741e+00 6.28491580e-01 -3.91059220e-01
3.81660759e-01 2.16562763e-01 -2.18116269e-01 -4.14399743e-01
-4.85007584e-01 -2.66635686e-01 3.56877714e-01 7.19559193e-01
-1.37675762e-01 2.30268478e-01 1.18977979e-01 3.15445550e-02
-1.05483390e-01 -1.24107175e-01 2.00925797e-01 1.36421037e+00
-8.52977991e-01 -9.78887558e-01 -3.60064685e-01 2.65514314e-01
1.06688462e-01 4.12982106e-01 -1.79063290e-01 3.93932045e-01
-1.95456427e-02 8.73715103e-01 1.07328527e-01 -9.99004319e-02
4.28681225e-01 -2.43027449e-01 1.21082175e+00 -9.72605109e-01
-1.91953361e-01 -7.77439922e-02 2.68962011e-02 -9.67038631e-01
-3.93019944e-01 -7.21073091e-01 -1.71655929e+00 9.41260085e-02
-4.44609016e-01 -2.28049442e-01 1.19075608e+00 1.05547416e+00
4.15284455e-01 3.77963006e-01 3.75151753e-01 -1.84958673e+00
-8.21592391e-01 -6.92996621e-01 -4.52469856e-01 4.38964605e-01
2.39962041e-01 -1.23499775e+00 3.69966835e-01 2.20901184e-02] | [5.802702903747559, -0.8396283984184265] |
23e60ffa-6e8f-4554-9736-d5660658578d | dual-side-feature-fusion-3d-pose-transfer | 2305.14951 | null | https://arxiv.org/abs/2305.14951v1 | https://arxiv.org/pdf/2305.14951v1.pdf | Dual-Side Feature Fusion 3D Pose Transfer | 3D pose transfer solves the problem of additional input and correspondence of traditional deformation transfer, only the source and target meshes need to be input, and the pose of the source mesh can be transferred to the target mesh. Some lightweight methods proposed in recent years consume less memory but cause spikes and distortions for some unseen poses, while others are costly in training due to the inclusion of large matrix multiplication and adversarial networks. In addition, the meshes with different numbers of vertices also increase the difficulty of pose transfer. In this work, we propose a Dual-Side Feature Fusion Pose Transfer Network to improve the pose transfer accuracy of the lightweight method. Our method takes the pose features as one of the side inputs to the decoding network and fuses them into the target mesh layer by layer at multiple scales. Our proposed Feature Fusion Adaptive Instance Normalization has the characteristic of having two side input channels that fuse pose features and identity features as denormalization parameters, thus enhancing the pose transfer capability of the network. Extensive experimental results show that our proposed method has stronger pose transfer capability than state-of-the-art methods while maintaining a lightweight network structure, and can converge faster. | ['Feipeng Da', 'Jue Liu'] | 2023-05-24 | null | null | null | null | ['pose-transfer'] | ['computer-vision'] | [ 2.27883771e-01 4.67708372e-02 3.43522951e-02 -3.56525660e-01
-7.45554030e-01 -3.99122834e-01 1.60906240e-01 -3.01246583e-01
-4.30446386e-01 6.20592058e-01 -7.33875623e-03 3.56139660e-01
1.59414336e-01 -1.06803882e+00 -1.27036452e+00 -7.57965505e-01
2.07089618e-01 6.04354084e-01 4.16567028e-01 -4.06756878e-01
-1.48144871e-01 6.20672524e-01 -1.34632277e+00 1.88403293e-01
6.91928983e-01 1.41425943e+00 -7.01564401e-02 2.35384300e-01
2.25490741e-02 1.40368426e-02 -5.07287681e-01 -4.59352881e-01
4.90271956e-01 -3.73950377e-02 -5.73345006e-01 -2.51008838e-01
9.44924235e-01 -5.85346997e-01 -4.83401448e-01 1.11724937e+00
9.71444488e-01 1.55230621e-02 4.55361515e-01 -1.18345034e+00
-6.59601390e-01 5.18549323e-01 -7.29174733e-01 -2.33627439e-01
1.85213670e-01 -7.44947642e-02 2.71470755e-01 -1.19663155e+00
7.03521073e-01 1.47299087e+00 1.01512265e+00 7.71857440e-01
-1.11258566e+00 -1.31528413e+00 1.60491187e-02 -7.92198330e-02
-1.53559232e+00 -3.27377856e-01 9.48169947e-01 -1.74891695e-01
4.26945448e-01 2.85006791e-01 6.54464126e-01 1.21885824e+00
2.92335391e-01 3.87423486e-01 6.75609231e-01 1.78679571e-01
-7.23633096e-02 -3.85569006e-01 -5.11714399e-01 6.09952033e-01
3.11448663e-01 -7.18506565e-03 -5.37860274e-01 -1.28199667e-01
1.37928343e+00 1.97925493e-01 -2.02023223e-01 -4.62116718e-01
-1.28050876e+00 5.89409411e-01 9.83502090e-01 -9.24742520e-02
-2.55780309e-01 5.96980691e-01 4.21256751e-01 3.02884907e-01
7.96144426e-01 -1.27981277e-02 -5.22533834e-01 1.17387854e-01
-7.91474819e-01 3.29924494e-01 6.40345752e-01 9.20941412e-01
9.62757766e-01 -5.91704212e-02 -5.47474883e-02 5.53118825e-01
3.11905473e-01 9.05574262e-01 3.06570947e-01 -6.43798590e-01
9.21867430e-01 5.77743292e-01 -2.90960759e-01 -1.37787044e+00
-4.52223092e-01 -5.01544595e-01 -1.27409804e+00 9.89070013e-02
2.54261404e-01 -2.52877921e-01 -1.10080087e+00 1.88317978e+00
5.90276062e-01 4.17944938e-01 -2.31274679e-01 1.03344059e+00
1.00796223e+00 5.67552209e-01 -9.67681929e-02 1.08474456e-01
1.18666685e+00 -7.57745862e-01 -6.29030645e-01 -2.94753611e-01
2.71141976e-01 -1.13055050e+00 7.48311877e-01 -5.92144206e-02
-1.15314448e+00 -7.19582915e-01 -1.09909284e+00 -3.14436972e-01
-1.08494200e-01 8.12393054e-02 6.75226927e-01 2.46033654e-01
-7.73977876e-01 9.90419447e-01 -9.91140664e-01 -3.13780196e-02
5.90405345e-01 8.37316573e-01 -5.49471974e-01 6.68575689e-02
-1.28137898e+00 6.74137354e-01 2.99679339e-01 4.68742400e-01
-5.52301109e-01 -8.55325758e-01 -9.23931539e-01 -3.99371237e-02
2.49840930e-01 -8.18426907e-01 8.54586244e-01 -8.81883919e-01
-1.54613531e+00 5.03928244e-01 1.92428529e-01 2.99946457e-01
8.14527750e-01 -3.98521423e-01 -2.99624294e-01 -9.16499123e-02
9.71086472e-02 7.48214841e-01 1.02429831e+00 -1.10465908e+00
-3.33208144e-01 -5.70859075e-01 -4.60582711e-02 2.67041892e-01
-6.14241779e-01 -4.50359493e-01 -7.58491099e-01 -9.66897488e-01
4.47531998e-01 -1.03730583e+00 -3.27609144e-02 5.58481455e-01
-9.12697464e-02 6.02565929e-02 1.06457043e+00 -7.64694989e-01
9.83540356e-01 -2.37363482e+00 5.21676362e-01 4.57664877e-01
2.42052287e-01 1.34548932e-01 -2.37234458e-01 2.17426956e-01
-1.54729173e-01 -3.15578021e-02 -2.46818084e-02 -2.96633929e-01
-1.34691402e-01 3.44718009e-01 -1.67572647e-01 6.40423238e-01
2.66472280e-01 1.04298151e+00 -7.05953479e-01 -5.53150773e-01
6.47246987e-02 8.89961839e-01 -7.94203222e-01 7.68711120e-02
3.46400589e-02 5.42071998e-01 -5.70934832e-01 5.02945900e-01
1.11987078e+00 -1.05237551e-01 -1.59299389e-01 -8.05486679e-01
3.17268997e-01 1.40580926e-02 -1.60045159e+00 2.13173008e+00
-3.20229292e-01 -9.11641866e-02 2.00564757e-01 -5.56769550e-01
8.79954159e-01 3.16772461e-01 5.02451718e-01 -3.99926335e-01
5.55928946e-01 3.48113894e-01 -1.49581254e-01 -6.24162965e-02
2.62852103e-01 -5.06506152e-02 -3.14502388e-01 2.40467757e-01
8.94230083e-02 -3.20430771e-02 -3.03057015e-01 -1.76249817e-01
6.77470922e-01 3.81309479e-01 -2.62103289e-01 -1.12883121e-01
5.65105736e-01 -4.63252723e-01 8.21661770e-01 1.28499484e-02
2.96151608e-01 8.39024723e-01 1.16248071e-01 -4.93763864e-01
-1.02789938e+00 -1.27734220e+00 -1.42187318e-02 1.01957250e+00
5.14510572e-01 -2.05453768e-01 -8.57451320e-01 -5.51883876e-01
2.91001469e-01 -1.65143669e-01 -7.36113966e-01 -5.04924834e-01
-1.12899125e+00 -4.54041958e-01 7.51226187e-01 9.81684387e-01
6.53230369e-01 -6.52626872e-01 -9.67682898e-02 7.65405968e-02
-1.78168714e-01 -8.44681561e-01 -9.33462262e-01 -2.29887828e-01
-1.13824344e+00 -7.85909534e-01 -7.99687207e-01 -1.11382532e+00
1.11488867e+00 -1.66172773e-01 7.25910008e-01 2.12766200e-01
-3.02925371e-02 -2.79543817e-01 -4.51064892e-02 -2.65696615e-01
-1.90781519e-01 4.36851501e-01 3.57335836e-01 6.25240207e-02
-3.24906081e-01 -9.23102796e-01 -7.83031225e-01 5.07008910e-01
-1.09363365e+00 1.34326100e-01 7.47449934e-01 7.61448264e-01
7.01351464e-01 -3.91684651e-01 4.32341099e-01 -7.30966568e-01
3.40056002e-01 -1.94827154e-01 -3.03151906e-01 -5.10427020e-02
-2.66627491e-01 7.33487755e-02 6.38025463e-01 -7.67313898e-01
-8.00617993e-01 3.34358335e-01 -2.63044059e-01 -8.76467526e-01
3.62674683e-01 1.41752064e-01 -5.13324201e-01 -5.80688059e-01
6.04589403e-01 -5.13732471e-02 4.33547437e-01 -7.73030281e-01
3.83536160e-01 3.45305353e-01 6.29241347e-01 -8.03377092e-01
1.42577624e+00 2.67646044e-01 2.14871168e-01 -1.43255830e-01
-5.36653101e-01 7.85683170e-02 -7.07265198e-01 -2.30858162e-01
6.13133430e-01 -1.05564976e+00 -8.26477826e-01 7.72875607e-01
-1.26767373e+00 2.42174834e-01 -2.19560474e-01 2.82644212e-01
-3.69560361e-01 1.91014796e-01 -8.52027714e-01 8.61959159e-02
-7.06897914e-01 -1.19227242e+00 1.35930061e+00 2.41283938e-01
-1.29862810e-02 -6.61686420e-01 -3.62787426e-01 1.29152790e-01
6.07986569e-01 7.26872742e-01 6.31052911e-01 -1.90778285e-01
-5.12337208e-01 -5.16999543e-01 -1.89783469e-01 3.70173097e-01
2.34084427e-01 -3.40781689e-01 -6.89557076e-01 -6.78076446e-01
-1.61206126e-02 -2.83446819e-01 4.87762988e-01 5.85280650e-04
1.24206877e+00 -3.43932271e-01 -3.24538589e-01 9.79435205e-01
1.30767727e+00 -3.59165102e-01 5.53070486e-01 2.70797256e-02
1.32633328e+00 1.24576382e-01 4.86434489e-01 5.85555173e-02
2.54756957e-01 5.80064535e-01 5.57213306e-01 -4.77934629e-01
-3.49384338e-01 -3.31676811e-01 3.52852941e-01 1.25291061e+00
-4.15101916e-01 2.13541791e-01 -6.90723360e-01 1.31364256e-01
-1.85520196e+00 -6.58476710e-01 1.97990209e-01 2.18219638e+00
1.20283389e+00 -6.53574616e-03 -2.44138017e-01 -3.56597602e-02
8.28104258e-01 1.02452710e-01 -6.50431633e-01 -1.76791847e-02
2.10479468e-01 4.46453989e-01 7.09027290e-01 3.14021766e-01
-9.74604607e-01 7.44713783e-01 5.38869190e+00 1.00011551e+00
-1.47224164e+00 1.58074543e-01 2.80019939e-01 -3.26664537e-01
-7.29987174e-02 -4.08227503e-01 -6.67658567e-01 3.70945305e-01
4.02971059e-01 -9.20544490e-02 4.16433513e-01 8.49272966e-01
-3.14886183e-01 4.83682424e-01 -1.37843156e+00 9.93439972e-01
1.15752339e-01 -1.23323381e+00 4.02007490e-01 2.83610877e-02
6.69099331e-01 -1.10530086e-01 1.05780497e-01 7.36945197e-02
-1.74001321e-01 -1.00569236e+00 8.37685883e-01 5.09432912e-01
1.35411763e+00 -1.08418107e+00 9.29119229e-01 1.31967008e-01
-1.47204828e+00 2.71772504e-01 -3.59850228e-01 -1.56966433e-01
3.88977118e-02 4.59374607e-01 -3.65345597e-01 8.93236458e-01
7.62863338e-01 6.22184992e-01 -4.87586439e-01 6.73450947e-01
-4.91549447e-02 6.10273816e-02 -5.73299468e-01 2.93536752e-01
-1.61251098e-01 1.77483395e-01 4.24640566e-01 5.75928450e-01
4.41656560e-01 -9.84893367e-02 4.48179424e-01 7.14704812e-01
-5.50796092e-01 9.60121006e-02 -2.14731500e-01 2.84610689e-01
5.36572099e-01 1.21841300e+00 -6.24813735e-01 -1.88923106e-01
-1.22186415e-01 1.12627101e+00 2.24767506e-01 -2.67272238e-02
-1.04476237e+00 -5.15370846e-01 8.39798510e-01 2.76692390e-01
2.86309332e-01 -1.41074672e-01 -3.65376860e-01 -1.12454271e+00
3.98022592e-01 -7.29304314e-01 -3.45358290e-02 -3.14816773e-01
-1.39199424e+00 4.44186628e-01 -2.01097414e-01 -1.45453608e+00
1.66341603e-01 -4.36527282e-01 -3.16306233e-01 1.03955626e+00
-1.00006974e+00 -1.55922174e+00 -6.16815448e-01 6.95634902e-01
1.65898055e-01 2.17417985e-01 7.59393752e-01 9.42175984e-01
-4.81443495e-01 1.01146591e+00 -1.30832791e-01 4.90846097e-01
1.01923823e+00 -7.82560706e-01 6.74984813e-01 5.68354905e-01
-3.57233256e-01 6.64257109e-01 3.17162037e-01 -9.20302570e-01
-1.67841387e+00 -1.41279018e+00 5.13397396e-01 -3.14802498e-01
2.84967124e-01 -6.30566239e-01 -8.27322125e-01 6.29378021e-01
-4.70961988e-01 3.15427661e-01 2.88959980e-01 -1.93303764e-01
-2.52115607e-01 -3.79653871e-01 -1.21489632e+00 5.02318919e-01
1.33219755e+00 -4.26051319e-01 -4.55904037e-01 1.67138517e-01
9.15436029e-01 -1.24228740e+00 -1.27357805e+00 7.70154595e-01
6.42881215e-01 -3.12948138e-01 1.24129391e+00 -2.65929371e-01
6.78806841e-01 -5.29211283e-01 -3.24984752e-02 -1.12395322e+00
-4.88007873e-01 -3.51772010e-01 -4.98279147e-02 1.33697379e+00
2.10165083e-01 -6.98550880e-01 7.83299565e-01 4.16905642e-01
-9.17531401e-02 -1.04494250e+00 -1.21276379e+00 -6.59946918e-01
1.97458401e-01 -4.20153067e-02 1.03652692e+00 8.40933144e-01
-5.89131832e-01 3.39765966e-01 -4.04141843e-01 2.86614269e-01
5.21047711e-01 2.46272888e-02 8.62626195e-01 -1.30734932e+00
1.42790601e-02 -6.66199401e-02 -6.42432034e-01 -9.23204362e-01
-1.65403649e-01 -1.10966349e+00 9.88384336e-02 -1.32336187e+00
-8.38666782e-02 -5.50770700e-01 -1.38233155e-01 8.09722185e-01
-1.19786449e-01 7.32814848e-01 2.63745487e-01 2.22379789e-01
-9.15452689e-02 6.88896894e-01 1.80046558e+00 -1.12978660e-01
4.81241196e-03 -2.64064729e-01 -4.74343926e-01 8.23963940e-01
4.56879705e-01 -5.68331480e-01 -3.32620203e-01 -8.87630820e-01
3.06014806e-01 -1.59427300e-01 4.42010492e-01 -1.21598184e+00
2.90863007e-01 7.97620639e-02 1.06943297e+00 -4.33190495e-01
4.64275122e-01 -1.18968999e+00 5.30738711e-01 5.84818661e-01
1.99443012e-01 2.09485412e-01 3.50084424e-01 4.95372176e-01
-1.85576770e-02 3.42393845e-01 7.88358510e-01 1.67842478e-01
-2.08865255e-01 8.65200996e-01 5.00214636e-01 -5.93243651e-02
1.04266584e+00 -2.42298022e-01 -6.52885363e-02 6.41427003e-04
-5.95908880e-01 1.62637785e-01 6.71769559e-01 8.19768429e-01
5.96481085e-01 -1.93622637e+00 -6.07282996e-01 4.27921981e-01
-2.13393703e-01 7.79274940e-01 3.02516341e-01 7.43718088e-01
-7.57270038e-01 -3.99335951e-01 -5.41709900e-01 -6.23093426e-01
-1.15188229e+00 2.62383997e-01 2.08951935e-01 2.88864765e-02
-6.89309776e-01 1.01338196e+00 2.18137056e-01 -6.97133183e-01
2.72387266e-01 -2.01162383e-01 2.10621595e-01 4.01217639e-02
2.46529251e-01 4.54417378e-01 1.24096632e-01 -7.89237022e-01
-5.09069681e-01 1.02521598e+00 -1.22598059e-01 2.57317185e-01
1.43901777e+00 1.83256000e-01 -4.17042524e-01 1.59472510e-01
1.48876894e+00 -6.75347745e-02 -1.16065371e+00 -2.19107598e-01
-7.82100081e-01 -5.37940502e-01 -4.82027903e-02 -5.06072164e-01
-1.67060030e+00 6.63838565e-01 6.88394129e-01 -5.41083097e-01
1.11461985e+00 -1.31341860e-01 1.42909563e+00 6.34033978e-02
5.64405024e-01 -9.65402842e-01 -7.74022788e-02 4.63772476e-01
1.06507623e+00 -7.34810829e-01 1.23602882e-01 -9.33332086e-01
8.98425132e-02 1.09701002e+00 9.30660307e-01 -4.55921113e-01
6.36555195e-01 5.18093944e-01 8.03895965e-02 -1.38959423e-01
-7.63888434e-02 4.85352159e-01 5.86267471e-01 3.35202307e-01
2.03048408e-01 -6.29447028e-02 -3.72024864e-01 6.96786344e-01
-4.27044064e-01 3.63990888e-02 -1.86767086e-01 1.03746045e+00
-1.04964428e-01 -1.27750635e+00 -4.84809220e-01 4.34043676e-01
-3.44952643e-01 7.00832605e-02 -2.27902964e-01 5.80389559e-01
5.52043140e-01 2.15466261e-01 1.27091065e-01 -7.56894112e-01
6.66360080e-01 -1.03543647e-01 6.27512634e-01 -4.94796962e-01
-8.81816268e-01 2.84129493e-02 -2.74509132e-01 -8.50448132e-01
-9.72144082e-02 -2.23972216e-01 -1.50020254e+00 -6.53292179e-01
-3.91675502e-01 -1.70407653e-01 5.61230361e-01 7.83546507e-01
5.74220836e-01 7.52611160e-01 5.44993520e-01 -1.51271021e+00
-7.07168341e-01 -9.23121154e-01 -2.59986818e-01 7.44577050e-01
1.55533195e-01 -8.74450982e-01 -7.94975013e-02 7.86116645e-02] | [7.235569000244141, -1.4395400285720825] |
1c905f8f-d743-4722-9dd5-06b26475f556 | neural-character-based-composition-models-for | 1809.00378 | null | http://arxiv.org/abs/1809.00378v1 | http://arxiv.org/pdf/1809.00378v1.pdf | Neural Character-based Composition Models for Abuse Detection | The advent of social media in recent years has fed into some highly
undesirable phenomena such as proliferation of offensive language, hate speech,
sexist remarks, etc. on the Internet. In light of this, there have been several
efforts to automate the detection and moderation of such abusive content.
However, deliberate obfuscation of words by users to evade detection poses a
serious challenge to the effectiveness of these efforts. The current state of
the art approaches to abusive language detection, based on recurrent neural
networks, do not explicitly address this problem and resort to a generic OOV
(out of vocabulary) embedding for unseen words. However, in using a single
embedding for all unseen words we lose the ability to distinguish between
obfuscated and non-obfuscated or rare words. In this paper, we address this
problem by designing a model that can compose embeddings for unseen words. We
experimentally demonstrate that our approach significantly advances the current
state of the art in abuse detection on datasets from two different domains,
namely Twitter and Wikipedia talk page. | ['Pushkar Mishra', 'Ekaterina Shutova', 'Helen Yannakoudakis'] | 2018-09-02 | neural-character-based-composition-models-for-1 | https://aclanthology.org/W18-5101 | https://aclanthology.org/W18-5101.pdf | ws-2018-10 | ['abuse-detection'] | ['natural-language-processing'] | [-6.00028858e-02 -9.48704332e-02 -1.80424377e-01 -1.31832911e-02
-4.92387652e-01 -7.81719208e-01 1.05728018e+00 5.86715698e-01
-5.29927552e-01 7.46613324e-01 3.60479563e-01 -5.25418162e-01
4.10636902e-01 -7.83270419e-01 -2.53316641e-01 -3.62188429e-01
8.46318528e-02 -9.28020924e-02 1.66618526e-01 -4.94843155e-01
4.62179273e-01 5.28318763e-01 -1.13664174e+00 9.83023122e-02
6.56594753e-01 4.38130081e-01 -5.18941641e-01 5.58460057e-01
-4.49436545e-01 7.86180913e-01 -9.92934763e-01 -8.39115798e-01
-1.12412475e-01 -3.47677201e-01 -6.35330558e-01 -1.81329712e-01
4.22931761e-01 -5.69399893e-01 -6.91659570e-01 1.45046592e+00
5.12744009e-01 -1.71100736e-01 6.29899561e-01 -1.06132627e+00
-9.00459528e-01 8.34612012e-01 -6.24985218e-01 8.08988273e-01
3.09777111e-01 7.20136017e-02 9.56139147e-01 -7.51795948e-01
6.28552258e-01 1.28952289e+00 5.25163352e-01 5.84595025e-01
-1.12063086e+00 -8.33702028e-01 -1.27617523e-01 -3.91925387e-02
-1.39918077e+00 -3.73570293e-01 9.23139691e-01 -6.19319260e-01
7.14790821e-01 3.81115943e-01 6.18789911e-01 1.86767662e+00
-1.06902048e-01 7.12030947e-01 8.83124650e-01 -5.70027173e-01
1.63159445e-02 5.04744351e-01 5.44251323e-01 6.02268875e-01
7.20913291e-01 -2.14598238e-01 -2.92127460e-01 -7.42238343e-01
1.54991671e-01 8.03330764e-02 -3.04162681e-01 5.79728261e-02
-4.95668173e-01 1.59313560e+00 1.89542398e-01 8.38687003e-01
-5.33935875e-02 7.40248198e-03 8.83796453e-01 1.43816173e-01
7.90715337e-01 7.81407118e-01 2.24901646e-01 -2.72160172e-01
-7.96437800e-01 2.43663043e-01 9.75807488e-01 3.60293031e-01
4.91807967e-01 -6.54571503e-02 1.07163511e-01 1.04387128e+00
2.25372136e-01 3.62234980e-01 8.07228029e-01 -1.47482887e-01
4.73378628e-01 5.76535165e-01 3.21432669e-03 -1.64203036e+00
-1.02172472e-01 -3.43897045e-01 -4.77889895e-01 -2.44198903e-01
3.18353653e-01 -2.71396309e-01 -1.01570404e+00 1.59223807e+00
3.22062761e-01 -1.96394518e-01 -3.04230839e-01 7.44188905e-01
4.65873837e-01 8.01737607e-01 1.33832887e-01 -5.46763428e-02
1.35802054e+00 -6.67826056e-01 -1.13566613e+00 -2.37527356e-01
1.03162944e+00 -7.01995850e-01 9.92144465e-01 2.91508406e-01
-7.28813469e-01 1.78351492e-01 -1.29871893e+00 -1.94621220e-01
-1.04302251e+00 -5.95334411e-01 4.48480487e-01 1.15792823e+00
-4.41047996e-01 5.76618493e-01 -4.48823750e-01 -4.46052432e-01
5.32585084e-01 8.06261152e-02 -4.06794220e-01 1.06370635e-01
-1.62746954e+00 1.20656168e+00 2.33748943e-01 -1.20137986e-02
-6.98967218e-01 -3.35174084e-01 -8.09015095e-01 -2.62165546e-01
4.90410686e-01 -9.23212525e-03 1.10565293e+00 -1.09180570e+00
-1.04211259e+00 1.14444923e+00 2.49097601e-01 -4.95783448e-01
4.45874840e-01 -5.97195446e-01 -7.78804362e-01 -7.11561367e-02
-4.08143289e-02 -3.63450940e-03 1.15309930e+00 -9.00241375e-01
-7.85135776e-02 -4.80735838e-01 1.77260175e-01 -2.21056551e-01
-1.25028622e+00 5.02863944e-01 2.64326960e-01 -7.82790244e-01
-4.32786226e-01 -7.95418262e-01 1.22804791e-01 -1.12730317e-01
-7.95151174e-01 -3.77351552e-01 1.22756553e+00 -7.18543649e-01
1.83254910e+00 -2.38342547e+00 1.27124488e-01 4.85183597e-02
5.88958681e-01 8.74055922e-01 2.27887928e-01 7.94439912e-01
1.65050790e-01 8.34573805e-01 -3.02319795e-01 -5.96129835e-01
-7.80196711e-02 4.55741763e-01 -8.13384295e-01 8.37091088e-01
2.82242876e-02 4.92288142e-01 -1.09699667e+00 -3.94283831e-01
-3.79837155e-02 5.76314986e-01 -4.20126289e-01 2.70226777e-01
4.69955318e-02 -8.51746574e-02 -6.78492427e-01 5.12949347e-01
4.78574216e-01 3.40339169e-02 3.20592336e-02 3.27531129e-01
-2.04651847e-01 7.11328983e-01 -6.21469855e-01 9.92355585e-01
-2.31037557e-01 9.15798783e-01 5.16839400e-02 -7.20504999e-01
6.90362096e-01 1.77940890e-01 -9.06706080e-02 -2.64954448e-01
6.15829110e-01 4.37423587e-01 1.41386334e-02 -6.47145331e-01
7.40695477e-01 -2.73072928e-01 -2.29966462e-01 6.46723866e-01
5.53249046e-02 2.47623503e-01 -4.47225273e-02 4.13031876e-01
1.20058107e+00 -6.10195994e-01 4.49178696e-01 1.68759488e-02
4.80479956e-01 -2.39912674e-01 5.97110391e-02 6.53396666e-01
-5.03741980e-01 3.49302799e-01 4.74485159e-01 -5.08731723e-01
-1.17792022e+00 -8.32605898e-01 -8.15338269e-02 1.10988522e+00
-1.24715254e-01 -6.16773784e-01 -9.53947723e-01 -1.01534069e+00
2.31293634e-01 8.36217940e-01 -6.55808806e-01 -4.19785380e-01
-6.91992104e-01 -8.11523855e-01 1.09203112e+00 4.04616222e-02
2.46551633e-01 -1.05699313e+00 -3.97417068e-01 3.54983807e-01
-1.11353956e-01 -1.14032161e+00 -4.55365330e-01 -4.08887155e-02
-3.98249358e-01 -8.02651346e-01 -5.82193792e-01 -2.68525273e-01
3.81855369e-01 3.11810315e-01 6.06322408e-01 4.13271517e-01
-5.43804765e-01 -5.13702892e-02 -6.84849620e-01 -5.56219697e-01
-8.12494218e-01 4.29894716e-01 2.54040480e-01 2.86300421e-01
6.62105739e-01 -6.49380326e-01 -2.82002419e-01 -1.68825686e-01
-1.39316404e+00 -6.20955646e-01 -2.01241560e-02 5.22034168e-01
-3.59427571e-01 -3.69934857e-01 4.64796990e-01 -1.18030906e+00
1.13986862e+00 -8.81227553e-01 -2.03182280e-01 -2.34228164e-01
-3.51093620e-01 4.81138378e-02 6.59549177e-01 -9.25715148e-01
-4.71060067e-01 -6.60782754e-01 -4.01508451e-01 -3.38336229e-01
-1.22377083e-01 3.51763695e-01 3.28940690e-01 -9.18007568e-02
7.68723190e-01 2.23612472e-01 -6.33418486e-02 -7.16429949e-01
3.78385633e-01 1.24707127e+00 7.24204332e-02 -2.72097085e-02
9.96267378e-01 5.43826878e-01 -6.25961959e-01 -1.41162670e+00
-1.09955037e+00 -6.58318639e-01 -1.97929665e-01 -6.28201887e-02
8.35206211e-01 -5.07774055e-01 -2.43445829e-01 5.27744710e-01
-1.31774271e+00 3.33177954e-01 2.06298247e-01 8.29230323e-02
-8.11985973e-03 9.40723419e-01 -8.66077721e-01 -1.17444837e+00
-5.02336502e-01 -7.54653156e-01 5.52854717e-01 -1.43256649e-01
-7.74895072e-01 -9.62286770e-01 4.43337381e-01 2.90726662e-01
7.27614701e-01 4.91594434e-01 9.58244979e-01 -1.19469202e+00
4.29766402e-02 -6.05285585e-01 2.45393291e-02 5.37105680e-01
2.52175689e-01 1.65005669e-01 -1.06859696e+00 -2.90813148e-01
2.30863586e-01 -6.19292498e-01 7.12825418e-01 -4.54325497e-01
8.26564431e-01 -8.02938998e-01 -3.84268761e-01 2.12093964e-01
1.25570929e+00 -7.51833171e-02 5.97201288e-01 5.17752945e-01
7.42387056e-01 5.51434994e-01 -8.68810043e-02 5.56579232e-01
-2.19917297e-01 5.74251890e-01 6.67812288e-01 3.13725501e-01
2.00989187e-01 -5.78544974e-01 4.43361402e-01 7.76082456e-01
2.58794606e-01 -5.67311943e-01 -9.10043359e-01 8.82422447e-01
-1.47366631e+00 -1.02521050e+00 -1.89034447e-01 1.97965944e+00
8.42280626e-01 2.76908636e-01 2.40719214e-01 4.44344580e-01
8.62786770e-01 6.97157681e-01 -1.81021631e-01 -1.12169492e+00
3.66622023e-02 4.55925390e-02 5.19903660e-01 5.45798361e-01
-1.26852548e+00 9.59830105e-01 5.93985939e+00 8.08465004e-01
-1.22443867e+00 3.92448604e-01 3.06458950e-01 -2.93522738e-02
-2.31920809e-01 -3.45502287e-01 -9.02851820e-01 7.64405847e-01
1.08410513e+00 -3.76633927e-02 3.77112448e-01 9.52457666e-01
-2.87599266e-02 2.19424769e-01 -6.81222022e-01 8.81886721e-01
6.06975794e-01 -1.08092117e+00 1.75910220e-01 3.71128619e-01
2.57259071e-01 5.95675036e-02 2.62286872e-01 3.50135207e-01
1.54973909e-01 -1.13302147e+00 5.62871873e-01 -2.31520772e-01
3.13449532e-01 -7.95958519e-01 7.04062045e-01 6.17947638e-01
-2.52257079e-01 -1.04730159e-01 -3.34482759e-01 -2.04897985e-01
3.24624151e-01 7.02218115e-01 -7.67564535e-01 -1.01889275e-01
2.40837753e-01 4.52234417e-01 -4.42623645e-01 5.07535040e-01
-3.90603215e-01 8.40893209e-01 -4.41421241e-01 -5.63519061e-01
7.02854335e-01 1.16273373e-01 9.27198708e-01 1.47479796e+00
3.07819620e-02 -4.69722748e-02 -2.20845297e-01 7.34778821e-01
-3.39549035e-01 2.97912657e-01 -1.23933971e+00 -8.30816627e-01
2.80423194e-01 1.23052084e+00 -4.11812484e-01 -2.28591397e-01
-4.46854740e-01 1.14085436e+00 7.83330381e-01 -9.63526219e-02
-8.48308921e-01 -4.60806847e-01 9.19071853e-01 4.51052040e-01
2.18996286e-01 -4.50758904e-01 1.05907790e-01 -1.31592071e+00
-7.14259595e-02 -9.70391452e-01 3.20857912e-01 -2.55685091e-01
-1.60762668e+00 5.61477542e-01 -1.50383547e-01 -6.45499587e-01
-2.36685067e-01 -4.12183434e-01 -3.57936352e-01 6.29438818e-01
-1.41668034e+00 -7.73285329e-01 3.06446612e-01 3.61819923e-01
4.22875464e-01 9.91405472e-02 8.05939436e-01 5.23358226e-01
-6.65214837e-01 6.35061204e-01 -5.24780601e-02 6.20569229e-01
5.65998435e-01 -8.17896903e-01 3.80045980e-01 7.49658525e-01
1.74549714e-01 7.74207115e-01 1.07715249e+00 -6.74134016e-01
-1.18526316e+00 -7.76742578e-01 1.30612922e+00 -7.29703307e-01
1.43169665e+00 -7.92481840e-01 -1.13724482e+00 6.77836955e-01
2.71267027e-01 -2.48893835e-02 8.93950880e-01 1.00812256e-01
-8.18134725e-01 5.19788265e-01 -1.13538969e+00 8.04530382e-01
9.72070158e-01 -8.10097098e-01 -8.47507536e-01 4.68884975e-01
6.63484752e-01 8.78230929e-02 -2.96788424e-01 -1.03940025e-01
4.06704903e-01 -7.45439172e-01 7.21011937e-01 -1.17715025e+00
5.91444790e-01 1.51129380e-01 -6.70219138e-02 -1.07862020e+00
-5.44766597e-02 -7.01718390e-01 -4.97493148e-01 1.31962740e+00
2.81899869e-01 -7.49569833e-01 4.75435585e-01 3.78538787e-01
9.62876976e-02 -4.75707591e-01 -1.06471241e+00 -6.82733774e-01
3.57183307e-01 -1.80680022e-01 6.44101277e-02 1.47550607e+00
2.71182984e-01 5.08891046e-01 -7.94822216e-01 1.37560353e-01
5.00505447e-01 -3.13427985e-01 6.45820856e-01 -1.00821102e+00
-5.54276221e-02 -3.86965424e-01 -5.43409944e-01 -7.61455774e-01
8.94993618e-02 -7.00347543e-01 -3.63887340e-01 -7.46540666e-01
2.85769820e-01 -4.09563966e-02 -2.32206974e-02 2.18939409e-01
-7.29015321e-02 5.57817459e-01 2.73309380e-01 1.89798892e-01
-3.33858103e-01 3.59108150e-01 7.33850479e-01 -1.73300520e-01
-4.05194238e-02 -3.96004498e-01 -7.36699343e-01 9.00189817e-01
7.59106159e-01 -9.39048707e-01 8.19298252e-03 -3.02383840e-01
5.89447141e-01 -5.04356265e-01 2.43437544e-01 -5.92143297e-01
-2.79898643e-01 8.56472924e-02 -2.75734335e-01 -1.42309964e-01
4.10594076e-01 -5.89222729e-01 -4.18461859e-01 6.01779699e-01
-3.59647393e-01 1.44188881e-01 5.10387607e-02 7.30668247e-01
-1.25148192e-01 -7.05692112e-01 9.37261045e-01 -7.25782886e-02
-2.12464333e-01 4.38857824e-02 -8.32521141e-01 2.25692824e-01
1.05071843e+00 -8.33473206e-02 -4.94999379e-01 -4.98293936e-01
-6.35417104e-01 -2.67761588e-01 2.73297936e-01 6.49399579e-01
4.74615991e-01 -1.03440094e+00 -6.55265689e-01 8.99533853e-02
1.50180891e-01 -7.30524361e-01 -8.84662271e-02 4.78806585e-01
-4.69656736e-01 3.04038227e-01 8.14674944e-02 -5.39625362e-02
-1.46239257e+00 8.71387243e-01 1.86847389e-01 -3.35480213e-01
-5.32957852e-01 6.49352491e-01 -1.77019581e-01 -1.06016204e-01
7.88687319e-02 1.02337159e-01 -3.79732221e-01 3.75525177e-01
8.59973788e-01 3.32625389e-01 -1.03621267e-01 -9.20248806e-01
-2.87169904e-01 8.32031220e-02 -6.43286943e-01 5.17768934e-02
1.12693620e+00 -7.85028413e-02 -1.80655941e-01 7.06435800e-01
1.71341431e+00 4.83507037e-01 -2.76400089e-01 -1.05745234e-01
1.44785792e-01 -7.32399106e-01 8.91528353e-02 -4.36984360e-01
-5.62617362e-01 9.84214187e-01 1.58980772e-01 1.10150552e+00
3.85482013e-01 -6.99224100e-02 1.30837774e+00 2.96174258e-01
2.51015514e-01 -9.15156245e-01 1.90760091e-01 6.41053021e-01
7.33266294e-01 -1.14337313e+00 -3.68336923e-02 -5.33323526e-01
-3.47140044e-01 1.02293289e+00 2.63233840e-01 -3.91680658e-01
6.95490718e-01 9.19574425e-02 4.32303660e-02 -3.19093883e-01
-3.06979418e-01 1.44562513e-01 -1.49256110e-01 1.73458397e-01
3.44603598e-01 3.12114023e-02 -9.43777859e-01 1.48761928e-01
-1.05499841e-01 -4.57832009e-01 9.29756403e-01 1.22886860e+00
-6.59004450e-01 -9.82281983e-01 -3.34101170e-01 2.94558793e-01
-1.22580528e+00 -1.72164515e-01 -9.76779997e-01 9.29849327e-01
-1.19436689e-01 8.33472431e-01 -4.82914858e-02 -3.71684343e-01
-1.48440646e-02 2.92684376e-01 1.24231599e-01 -7.74908423e-01
-8.28158140e-01 -3.85261267e-01 4.93850172e-01 -1.31376103e-01
-1.12529635e-01 -4.11675304e-01 -6.49059296e-01 -7.45219886e-01
-6.20581448e-01 1.55262262e-01 8.73346329e-01 9.22940314e-01
2.55435109e-01 -1.48101062e-01 5.66378593e-01 -4.48707849e-01
-9.91615772e-01 -1.16675496e+00 -5.85152388e-01 9.00953114e-01
6.86929762e-01 -7.67179608e-01 -9.57633793e-01 -5.08750975e-01] | [8.673258781433105, 10.441415786743164] |
d7d00c71-cdf6-4b5a-aef8-a6a47b882f4c | membership-inference-attack-on-graph-neural | 2101.06570 | null | https://arxiv.org/abs/2101.06570v3 | https://arxiv.org/pdf/2101.06570v3.pdf | Membership Inference Attack on Graph Neural Networks | Graph Neural Networks (GNNs), which generalize traditional deep neural networks on graph data, have achieved state-of-the-art performance on several graph analytical tasks. We focus on how trained GNN models could leak information about the \emph{member} nodes that they were trained on. We introduce two realistic settings for performing a membership inference (MI) attack on GNNs. While choosing the simplest possible attack model that utilizes the posteriors of the trained model (black-box access), we thoroughly analyze the properties of GNNs and the datasets which dictate the differences in their robustness towards MI attack. While in traditional machine learning models, overfitting is considered the main cause of such leakage, we show that in GNNs the additional structural information is the major contributing factor. We support our findings by extensive experiments on four representative GNN models. To prevent MI attacks on GNN, we propose two effective defenses that significantly decreases the attacker's inference by up to 60% without degradation to the target model's performance. Our code is available at https://github.com/iyempissy/rebMIGraph. | ['Megha Khosla', 'Wolfgang Nejdl', 'Iyiola E. Olatunji'] | 2021-01-17 | null | null | null | null | ['membership-inference-attack'] | ['computer-vision'] | [ 1.5290016e-01 5.8838421e-01 -3.4060284e-01 1.2030529e-01
-2.0261668e-01 -9.3640667e-01 6.3089454e-01 2.9664722e-01
-2.6338547e-01 5.0099993e-01 -1.0090885e-01 -1.2302306e+00
-9.1287173e-02 -1.1935782e+00 -1.0021166e+00 -4.7076058e-01
-4.6428928e-01 3.7774333e-01 2.1297553e-01 -1.9678906e-01
2.0999315e-01 6.3593161e-01 -5.3088266e-01 -1.3814053e-01
4.2833266e-01 6.1568010e-01 -6.2868071e-01 9.1278923e-01
2.7577373e-01 9.1423166e-01 -9.6496141e-01 -8.6914635e-01
4.8759714e-01 -1.0130049e-01 -8.9963651e-01 -4.7960740e-01
4.6952862e-01 -5.4091966e-01 -1.1243495e+00 1.4808912e+00
2.4423170e-01 -2.2515650e-01 6.2260693e-01 -1.4459593e+00
-7.3610884e-01 1.3052725e+00 -5.6670260e-01 3.3746970e-01
9.1674160e-03 1.3590167e-01 1.2087927e+00 -1.0341041e-01
4.1695794e-01 1.2398071e+00 8.0461806e-01 6.0412884e-01
-1.4113665e+00 -9.1022432e-01 5.2818600e-02 -1.2153392e-02
-1.2475613e+00 -3.8368410e-01 8.0325651e-01 2.6617115e-02
8.2022738e-01 2.7964991e-01 2.5652248e-01 1.4435571e+00
2.7444401e-01 5.2016580e-01 5.5517226e-01 -3.4803063e-01
8.9185894e-02 -6.0508832e-02 3.2131574e-01 8.6065233e-01
9.3716621e-01 2.7194285e-01 -3.7942439e-02 -7.0150179e-01
6.5214205e-01 -1.6329734e-01 -4.2865491e-01 -3.2843858e-01
-4.9606448e-01 1.2808501e+00 7.7394742e-01 1.2027357e-01
-4.7525141e-02 9.3308449e-01 5.1602018e-01 2.9231179e-01
2.1177556e-01 4.7492471e-01 -4.5958963e-01 2.6500985e-01
-5.3743786e-01 -1.8510418e-01 1.3181925e+00 5.6433529e-01
5.1414382e-01 2.8506681e-01 2.7845791e-01 -5.6354672e-02
1.8996610e-01 4.5661828e-01 -2.2096823e-01 -6.4811444e-01
3.1050974e-01 3.2363066e-01 -4.4255608e-01 -1.3180046e+00
-3.0608943e-01 -8.1823665e-01 -7.7002984e-01 -6.7851380e-02
6.7594099e-01 -5.4455036e-01 -9.5070934e-01 2.0338221e+00
-1.8994989e-02 3.2004753e-01 -1.2232614e-01 3.0884990e-01
6.9559538e-01 4.6454844e-01 5.6267433e-02 3.4254083e-01
1.0438063e+00 -6.0143226e-01 -2.4960169e-01 -3.9594704e-01
9.5022672e-01 -2.2389580e-01 5.3951383e-01 2.8065053e-01
-7.9356509e-01 1.3982995e-01 -1.1449655e+00 2.2709413e-01
-5.6650281e-01 -4.2590302e-01 8.6792970e-01 1.3070788e+00
-1.2415190e+00 7.9505312e-01 -9.0222353e-01 -3.1716096e-01
6.8487281e-01 6.3907111e-01 -3.1567180e-01 2.1517551e-01
-1.5503608e+00 7.8692210e-01 5.1072675e-01 -5.6024946e-02
-1.4096154e+00 -4.7888359e-01 -8.1349373e-01 4.5290908e-01
6.6635078e-01 -4.4644001e-01 1.0995244e+00 -5.5462062e-01
-1.0081195e+00 4.5818558e-01 2.9895398e-01 -9.4065714e-01
1.7759237e-01 5.1027924e-02 -2.6487571e-01 3.1989211e-01
-3.9519188e-01 1.5918222e-01 7.4671036e-01 -1.3700511e+00
4.9145907e-02 -6.6489778e-02 6.2053931e-01 -4.3031564e-01
-5.6387496e-01 2.6366116e-02 -1.5682602e-01 -4.0661490e-01
-2.2483027e-01 -9.0832371e-01 -2.9045078e-01 -2.3316577e-01
-1.1374890e+00 -5.0221343e-02 9.9429655e-01 -4.3932089e-01
1.3548652e+00 -1.7976381e+00 -2.4541885e-01 8.4111238e-01
8.0412751e-01 7.5194800e-01 -1.9798505e-01 8.5523307e-01
-8.2187183e-02 6.7464650e-01 -1.6861100e-02 -2.2578514e-01
1.0334885e-01 1.8656820e-01 -6.1795461e-01 7.6540297e-01
-1.6157684e-01 1.0959804e+00 -6.7717725e-01 4.8259668e-02
-4.2404406e-02 4.6642813e-01 -5.7746720e-01 -4.9574144e-02
-3.6981514e-01 8.9818887e-02 -4.3110910e-01 4.4337842e-01
7.0933992e-01 -7.2080952e-01 6.5948492e-01 -2.3370590e-02
7.8997809e-01 5.8514327e-01 -7.5039840e-01 8.1596142e-01
-3.0009227e-02 6.1630964e-01 2.9764813e-01 -1.0172673e+00
6.6338468e-01 1.4096449e-01 -3.0045943e-02 -1.6820523e-01
4.9070662e-01 -6.3286513e-02 5.6085992e-01 1.4069216e-01
9.9487081e-02 9.2703730e-02 -8.8653728e-02 9.0364414e-01
9.0297312e-03 3.7720314e-01 7.2366660e-03 8.2480901e-01
1.6843429e+00 -6.5665656e-01 2.2185245e-01 -4.0428221e-01
2.6085794e-01 -2.9457060e-01 2.8050882e-01 1.3725570e+00
-2.8228918e-01 -6.8989635e-02 1.2703856e+00 -4.6793547e-01
-9.1160089e-01 -1.1935968e+00 2.1415856e-01 9.0842807e-01
1.1679580e-01 -8.7812376e-01 -8.7053841e-01 -1.1876873e+00
9.8471545e-02 8.4332681e-01 -7.6887625e-01 -5.6410313e-01
-4.4147003e-01 -9.5032507e-01 1.3933984e+00 2.6951814e-01
4.2703444e-01 -8.3687454e-01 -5.0584979e-02 -9.9177659e-02
6.9717117e-02 -1.0735468e+00 -2.5459275e-01 3.4087124e-01
-5.5893481e-01 -1.3755732e+00 6.6583261e-02 -3.9652890e-01
8.5680306e-01 1.0668683e-01 1.1603602e+00 7.0205331e-01
-1.1450964e-01 3.8273314e-01 -3.5938971e-02 -2.9017064e-01
-8.0859405e-01 4.5509416e-01 7.2592899e-02 -3.0772570e-01
3.5867840e-01 -9.7404426e-01 -4.4181675e-01 -4.1095264e-02
-9.8162502e-01 -4.4964224e-01 3.3586305e-01 3.9293224e-01
-2.0823888e-01 5.0254560e-01 2.6898855e-01 -1.4686705e+00
8.9267647e-01 -5.7320553e-01 -7.3892581e-01 9.1635607e-02
-6.1307961e-01 1.3901502e-01 8.9511102e-01 -5.1960349e-01
-3.0440271e-01 -5.1317418e-01 -1.8264653e-01 -4.8518229e-01
-1.1341977e-02 6.3459611e-01 -4.5689419e-02 -6.5594947e-01
8.4731472e-01 -6.3081414e-02 1.5537257e-03 -1.4824781e-01
3.6526725e-01 2.6510251e-03 2.8747123e-01 -6.9623560e-01
1.3306428e+00 3.4683618e-01 4.4475907e-01 -6.9025743e-01
-6.6591340e-01 3.4335732e-01 -1.4606293e-01 5.5952813e-02
4.4695154e-01 -4.4420323e-01 -1.1988826e+00 4.7046357e-01
-1.1664579e+00 -6.9348669e-01 1.4495672e-01 3.6442511e-02
7.0607729e-02 7.2846717e-01 -1.2169403e+00 -7.7343738e-01
-5.4752487e-01 -9.4427919e-01 2.2934163e-01 1.1743256e-01
-9.3530588e-02 -1.5111946e+00 -1.6943245e-01 4.7106821e-02
5.3846657e-01 3.3783644e-01 1.2642478e+00 -1.3712995e+00
-7.0641255e-01 -4.1431192e-01 -4.1801766e-01 2.6407626e-01
-1.0819505e-01 1.8047281e-01 -8.4885073e-01 -6.2776840e-01
-1.3937123e-01 -3.0470604e-01 9.4534117e-01 3.5958797e-01
1.3929964e+00 -7.4810326e-01 -5.3954524e-01 8.6542147e-01
1.5369906e+00 -6.7606375e-02 7.2847700e-01 1.1026739e-01
8.9905709e-01 1.7343260e-01 -3.9809093e-01 7.9264671e-02
1.4774501e-01 1.2999737e-01 1.0712492e+00 -8.1231751e-02
1.5343256e-01 -7.4433851e-01 4.6517611e-01 3.1760514e-01
2.3913904e-01 -9.5938385e-01 -8.9593399e-01 1.6132723e-01
-1.5344441e+00 -8.5932797e-01 -8.7154016e-02 2.0579181e+00
4.2897266e-01 7.7153552e-01 1.1629773e-01 6.2700294e-02
7.9579288e-01 5.2620173e-01 -5.3880203e-01 -5.7195640e-01
2.8453036e-03 5.4531288e-01 1.1829839e+00 9.3952501e-01
-1.1184428e+00 9.6976531e-01 6.5506058e+00 9.2389059e-01
-9.0653276e-01 -9.2202805e-02 7.2874707e-01 2.9700782e-02
-4.6317706e-01 3.2659012e-01 -7.6311368e-01 3.0890310e-01
1.3681968e+00 -2.8651011e-01 7.7420592e-01 7.3761082e-01
-3.8342944e-01 3.0657881e-01 -9.4237119e-01 2.8846473e-01
-4.9988300e-02 -1.4685456e+00 2.3578131e-01 7.5518245e-01
4.0648809e-01 1.8360007e-01 2.9465222e-01 3.8200387e-01
1.1195661e+00 -1.3191721e+00 2.8419074e-01 2.3606554e-02
6.4208251e-01 -1.0158197e+00 5.7500201e-01 3.9213234e-01
-8.9143485e-01 -2.3844118e-01 -3.8168269e-01 -7.4966960e-02
-1.8425824e-02 5.9684455e-01 -9.3622959e-01 5.2071744e-01
3.4211847e-01 2.5027460e-01 -7.8474194e-01 4.2117900e-01
-6.4002180e-01 1.2214227e+00 -6.3386726e-01 -3.3094525e-02
4.3915460e-01 -4.8812915e-02 7.9120201e-01 9.7671324e-01
-1.8482815e-02 -6.1041489e-02 1.6203588e-01 1.1524261e+00
-5.1453030e-01 -5.1210576e-01 -9.1003126e-01 -4.7008869e-01
6.6149634e-01 1.1428353e+00 -7.7845985e-01 -1.1158265e-01
-1.4196689e-01 5.3626603e-01 5.9353095e-01 4.6163896e-01
-9.4564766e-01 -6.6984975e-01 5.1284164e-01 2.1181005e-01
4.1102099e-01 -2.5743568e-01 -1.3312040e-01 -1.0128777e+00
-3.4342158e-01 -1.0841707e+00 6.7764795e-01 -2.7245554e-01
-1.2886984e+00 5.7947332e-01 -1.2688296e-01 -3.7988549e-01
-8.1984013e-02 -8.1768018e-01 -1.1078261e+00 6.5723675e-01
-1.1237067e+00 -1.1408178e+00 3.2043308e-01 5.9343153e-01
-4.3803269e-01 -1.2991813e-01 8.0073065e-01 1.2798551e-01
-7.9141533e-01 1.1319155e+00 -1.2114404e-01 9.7533786e-01
2.2440866e-01 -1.1558721e+00 9.5796728e-01 1.4218861e+00
3.7264135e-01 1.0568579e+00 8.0163741e-01 -9.2852968e-01
-1.4208257e+00 -8.2805902e-01 5.3622842e-01 -4.7225848e-01
9.6513873e-01 -6.1439973e-01 -7.2369230e-01 1.3011918e+00
4.8627410e-02 2.5141221e-02 5.7910615e-01 2.0006043e-01
-8.3459908e-01 1.6376272e-01 -1.1982362e+00 9.0748250e-01
1.0938413e+00 -6.4918572e-01 6.4834863e-02 2.1493597e-01
9.0894121e-01 -2.2328505e-01 -6.4676309e-01 4.0517163e-01
2.6127362e-01 -1.0416268e+00 9.7979784e-01 -9.9875218e-01
1.4290014e-01 -7.5024897e-03 -5.6935355e-02 -1.2105603e+00
-3.2438037e-01 -1.0001674e+00 -7.5671244e-01 8.1098813e-01
5.6765014e-01 -1.0936260e+00 1.2265629e+00 4.9682117e-01
4.6129400e-01 -6.5896738e-01 -7.2237021e-01 -6.9725162e-01
4.1501033e-01 -2.5665188e-01 5.8048868e-01 7.9882050e-01
4.0443745e-03 3.1331581e-01 -3.9246577e-01 6.7577302e-01
9.2405432e-01 -2.5249413e-01 9.5038521e-01 -1.1972340e+00
-4.7557971e-01 -4.7436157e-01 -4.2572296e-01 -9.2159408e-01
4.3194896e-01 -1.1760160e+00 -5.9952188e-01 -1.2556041e+00
1.0208112e-01 -1.7608649e-01 -3.0099583e-01 6.1623645e-01
7.3326766e-02 2.2943702e-01 2.2333173e-01 -1.9310300e-01
-3.9291337e-01 1.0722444e-01 7.1266472e-01 -1.4413935e-01
1.2885438e-01 3.0646917e-01 -1.2424068e+00 6.3246369e-01
1.1646650e+00 -6.8439907e-01 -6.4559084e-01 -3.2999718e-01
4.3713927e-01 -5.9272554e-02 5.3117502e-01 -8.1206220e-01
2.5252840e-01 1.9524568e-01 2.3241186e-01 -2.7667779e-01
1.9959439e-01 -5.9899902e-01 1.2905183e-01 9.4080013e-01
-4.0336499e-01 -4.3948520e-02 1.7813495e-01 7.8536242e-01
4.8195258e-01 -2.4579218e-01 8.7663990e-01 -1.6187397e-01
6.9700092e-02 7.8164136e-01 -4.4845119e-01 1.3702521e-01
6.5811801e-01 1.2259329e-01 -1.0508331e+00 -1.0606034e+00
-6.6651285e-01 7.1256198e-02 5.6367844e-01 -8.4009804e-02
4.0030402e-01 -8.6935866e-01 -5.0387275e-01 1.9762349e-01
-3.5570225e-01 -3.4580755e-01 -1.7108640e-01 6.2111825e-01
-5.8713502e-01 3.8540089e-01 -2.1186002e-02 -3.0038781e-02
-1.1830487e+00 9.2511898e-01 6.2116444e-01 -6.3323659e-01
-4.9423563e-01 9.0049481e-01 1.5156457e-01 -4.9306488e-01
4.1558251e-01 1.3635305e-01 2.0057689e-01 -6.4533442e-01
2.9308280e-01 3.4995130e-01 -1.7658421e-01 -2.3218325e-01
-2.8595179e-01 -1.3104740e-01 -5.7970113e-01 1.7134540e-01
1.0066723e+00 1.6046935e-01 -2.8382078e-01 -2.3710686e-01
1.1709230e+00 3.2725081e-01 -8.0562633e-01 -3.6877963e-01
-7.5965650e-02 -2.9493243e-01 -1.5540766e-02 -5.5487847e-01
-1.3276688e+00 8.2981795e-01 -2.5372946e-01 7.9235995e-01
7.7846271e-01 6.1484920e-03 9.3203253e-01 5.9214222e-01
3.8094413e-01 -3.9068550e-01 -9.2034834e-03 5.5580515e-01
2.3052633e-01 -6.9284642e-01 2.6054132e-01 -5.3844911e-01
-6.7169055e-02 1.0537355e+00 7.1166122e-01 -4.7472492e-01
1.0352015e+00 3.1229132e-01 -2.8485218e-01 -4.8453912e-01
-7.5832921e-01 1.5721753e-01 -1.6928223e-01 6.6132116e-01
-3.7372876e-02 7.5860344e-02 1.2521639e-01 3.6242270e-01
-3.2967725e-01 -6.2243545e-01 1.0355033e+00 6.7397785e-01
-3.7996870e-01 -1.1928850e+00 -6.5929703e-02 6.1091506e-01
-9.8387623e-01 -3.4021580e-01 -7.6480323e-01 7.9034650e-01
-4.7669405e-01 1.0276347e+00 -2.2537974e-01 -7.2552925e-01
-2.5250769e-01 -2.7914780e-01 3.9159799e-01 -5.8279765e-01
-6.0086316e-01 -3.6488748e-01 4.4457597e-01 -4.4936794e-01
3.2011217e-01 -5.8369521e-02 -9.2358583e-01 -1.3239548e+00
-5.2678806e-01 2.7176005e-01 2.9879653e-01 5.0591421e-01
3.9464912e-01 2.6488447e-01 6.0655999e-01 -3.8071823e-01
-8.5634047e-01 -8.1280756e-01 -6.9456393e-01 1.8605942e-01
4.1776851e-01 -4.0852916e-01 -1.0351235e+00 -6.6202563e-01] | [6.058763027191162, 7.307633876800537] |
1f98599d-98de-43f0-aa58-1cc3cd3aab44 | fsitm-a-feature-similarity-index-for-tone | 1704.05624 | null | http://arxiv.org/abs/1704.05624v1 | http://arxiv.org/pdf/1704.05624v1.pdf | FSITM: A Feature Similarity Index For Tone-Mapped Images | In this work, based on the local phase information of images, an objective
index, called the feature similarity index for tone-mapped images (FSITM), is
proposed. To evaluate a tone mapping operator (TMO), the proposed index
compares the locally weighted mean phase angle map of an original high dynamic
range (HDR) to that of its associated tone-mapped image calculated using the
output of the TMO method. In experiments on two standard databases, it is shown
that the proposed FSITM method outperforms the state-of-the-art index, the tone
mapped quality index (TMQI). In addition, a higher performance is obtained by
combining the FSITM and TMQI indices. The MATLAB source code of the proposed
metric(s) is available at
https://www.mathworks.com/matlabcentral/fileexchange/59814. | ['Atena Shahkolaei', 'Hossein Ziaei Nafchi', 'Reza Farrahi Moghaddam', 'Mohamed Cheriet'] | 2017-04-19 | null | null | null | null | ['tone-mapping'] | ['computer-vision'] | [ 5.35640836e-01 -4.19123143e-01 -4.82945852e-02 -1.09989718e-01
-7.50979602e-01 -1.71872973e-01 3.47406447e-01 -2.24601496e-02
-1.59758031e-01 4.58094358e-01 -2.43983399e-02 5.52038923e-02
-6.17208242e-01 -8.43466759e-01 -2.47048020e-01 -7.26148546e-01
-4.61190268e-02 8.83766823e-03 5.89280307e-01 -1.65883422e-01
7.41482913e-01 3.30961496e-01 -1.65177822e+00 -2.54893880e-02
9.82609510e-01 1.46066272e+00 5.18507957e-01 4.73937064e-01
1.79369301e-01 5.16439080e-01 -6.69140995e-01 -6.10507205e-02
3.42372596e-01 -7.03500450e-01 -5.77843904e-01 -3.33110988e-01
3.18102539e-01 -9.13902372e-03 -3.79208297e-01 1.33599770e+00
5.18312514e-01 1.51964962e-01 4.90819186e-01 -1.19043100e+00
-6.15648985e-01 1.35658056e-01 -7.02970803e-01 6.24439955e-01
3.81095171e-01 -7.49024153e-02 6.63060963e-01 -9.55689669e-01
6.97334826e-01 7.96480536e-01 2.23944098e-01 -2.96596348e-01
-1.14769983e+00 -8.43348563e-01 -6.27570450e-01 6.85913146e-01
-1.85898876e+00 -1.04786374e-01 9.51057136e-01 -2.37896115e-01
6.57822669e-01 5.34066856e-01 5.31095624e-01 1.54262051e-01
4.96733904e-01 3.42799962e-01 1.85766232e+00 -5.06907165e-01
-6.28783330e-02 2.24327426e-02 -1.64038852e-01 6.10011637e-01
1.53084144e-01 4.20438826e-01 -7.14579403e-01 -6.38728635e-03
7.15907574e-01 -5.99285305e-01 -5.76646984e-01 -2.01666474e-01
-1.13830900e+00 4.08684671e-01 3.39892447e-01 5.55543184e-01
-2.74636209e-01 -2.01923355e-01 -2.41471864e-02 6.33777022e-01
4.60505247e-01 2.90145040e-01 1.45525306e-01 -5.11673987e-01
-1.03208697e+00 -1.89224795e-01 4.48778659e-01 5.12021422e-01
8.60934615e-01 5.41960374e-02 3.74613479e-02 8.60662878e-01
1.98176771e-01 8.59466195e-01 4.00194407e-01 -8.80094051e-01
9.87632200e-02 4.73667473e-01 -2.48005167e-02 -1.35773480e+00
-2.84761637e-01 -2.67149121e-01 -6.76950395e-01 4.60882455e-01
1.50444046e-01 3.80538136e-01 -6.18413091e-01 1.33702672e+00
1.73781529e-01 1.33331552e-01 7.32646836e-03 9.60921824e-01
8.06565404e-01 9.50905502e-01 -3.96330535e-01 -3.64929289e-01
1.11761737e+00 -4.84417439e-01 -9.55071867e-01 9.83505324e-02
-2.14008778e-01 -1.37692738e+00 9.84016538e-01 6.40183151e-01
-1.06276870e+00 -9.38292801e-01 -1.47212172e+00 3.05716604e-01
-2.43462056e-01 -1.01036474e-01 -3.22803333e-02 5.23694694e-01
-1.03519535e+00 6.19281769e-01 -5.15643775e-01 -4.49332952e-01
1.00886963e-01 1.12375334e-01 -2.37328336e-01 -4.18299548e-02
-1.26921844e+00 1.10671043e+00 2.48284653e-01 -1.04939930e-01
-3.53837311e-01 -6.70634151e-01 -5.86853802e-01 -1.27108470e-01
2.20697567e-01 3.70930508e-02 8.28943908e-01 -6.62687182e-01
-1.64095128e+00 6.72300041e-01 1.66596130e-01 7.31688887e-02
2.26260990e-01 2.09176898e-01 -6.60385132e-01 7.96897948e-01
-5.66307418e-02 3.45953554e-01 8.33532691e-01 -1.23640096e+00
-5.23954093e-01 -1.13181762e-01 -2.51328915e-01 1.80047452e-01
-5.25889322e-02 1.74356669e-01 -3.88817936e-01 -7.76207089e-01
2.53112048e-01 -6.39892817e-01 4.16347951e-01 1.03995562e-01
-1.82147205e-01 1.98652908e-01 1.00496209e+00 -8.08398664e-01
1.58679104e+00 -2.35585022e+00 -1.68214962e-01 6.44838452e-01
-2.86770135e-01 3.08468521e-01 -9.86834764e-02 4.35964376e-01
-1.75270379e-01 -2.34919325e-01 -3.36576313e-01 4.25074965e-01
3.11989971e-02 -2.42083937e-01 7.72393942e-02 5.81147373e-01
3.44693996e-02 5.54679573e-01 -6.89049840e-01 -6.06518924e-01
5.06437719e-01 4.95306373e-01 2.16904819e-01 -6.41397610e-02
3.67765725e-01 4.57768649e-01 -7.95455277e-02 6.18482292e-01
1.10111320e+00 -1.16869099e-01 1.82923168e-01 -7.07895279e-01
-3.37571949e-01 -7.18262196e-02 -1.55957770e+00 1.27187502e+00
-3.10081929e-01 9.83380079e-01 -3.60832691e-01 -5.67162693e-01
1.44248033e+00 3.15982759e-01 7.71211326e-01 -1.66743326e+00
1.82306364e-01 3.50143164e-01 -3.65791768e-02 -4.58826482e-01
5.61949193e-01 1.50435179e-01 9.64570642e-02 5.86094379e-01
-2.93017589e-02 -2.39262342e-01 4.64360327e-01 -3.54223885e-02
8.03645849e-01 1.35386154e-01 6.04699135e-01 -5.09801030e-01
7.90187955e-01 -1.17366001e-01 3.18560272e-01 3.51137072e-01
-3.17301542e-01 6.32430017e-01 2.38220632e-01 -9.04856622e-02
-1.19603074e+00 -1.40759552e+00 -6.95061207e-01 4.27805096e-01
7.73743153e-01 -2.36479387e-01 -6.18900597e-01 1.63423240e-01
-1.95371076e-01 3.08715254e-01 -2.89956391e-01 -1.08572796e-01
-4.62136805e-01 -7.37887919e-01 3.21001977e-01 1.43207431e-01
8.21066499e-01 -8.99288595e-01 -8.49652052e-01 -1.13799572e-02
-5.50523281e-01 -9.05723453e-01 -5.90195954e-01 -2.00181216e-01
-5.89405775e-01 -8.28016818e-01 -6.96567416e-01 -5.88184714e-01
4.18921381e-01 3.58322561e-01 8.99115026e-01 3.52130160e-02
-3.71576130e-01 2.32138067e-01 -7.72405326e-01 -2.17056833e-02
-3.06869894e-01 -3.53529334e-01 -1.68956339e-01 2.95955181e-01
2.07851604e-01 -5.31841815e-01 -7.22699940e-01 7.54127741e-01
-1.12185168e+00 1.35798946e-01 7.56980598e-01 5.43642759e-01
9.48610902e-01 5.25253296e-01 3.04310620e-01 -2.99962699e-01
4.71054316e-01 -1.23052280e-02 -9.56235170e-01 2.49549463e-01
-1.04080701e+00 -3.65215242e-02 2.66368747e-01 -2.87064999e-01
-9.52911615e-01 -3.33142489e-01 1.41377971e-01 -2.28521228e-01
1.66590109e-01 4.36259568e-01 -9.31425169e-02 -5.16060591e-01
5.44241071e-01 3.49004924e-01 1.70154899e-01 -2.46717170e-01
-2.67135240e-02 8.60524058e-01 8.25504601e-01 -2.04169989e-01
9.94041145e-01 4.24922973e-01 1.55727938e-01 -7.49022782e-01
-2.83104390e-01 -4.53575432e-01 -2.48865455e-01 -6.47092462e-01
7.69462943e-01 -5.34126341e-01 -5.58384240e-01 6.38166130e-01
-4.70092088e-01 -3.29716057e-01 -3.76501828e-02 6.59697473e-01
-6.99016333e-01 4.04531449e-01 -2.77478874e-01 -4.57189977e-01
-4.01274115e-01 -1.02592421e+00 7.65748680e-01 6.60898268e-01
-6.02238066e-02 -7.00073838e-01 2.49905705e-01 1.52205035e-01
6.92969739e-01 3.36656958e-01 6.66465640e-01 -1.92421023e-02
-6.79277182e-01 -7.23566636e-02 -2.87320763e-01 2.45342746e-01
2.14454949e-01 2.09620297e-01 -6.73538804e-01 -1.81909084e-01
7.17758983e-02 1.12499982e-01 4.94984061e-01 5.56499481e-01
8.61299813e-01 1.72768325e-01 1.20732188e-01 6.51186824e-01
2.00001478e+00 6.85252249e-01 1.17622101e+00 7.38239110e-01
5.81640899e-02 2.62139171e-01 1.36927831e+00 4.68477786e-01
6.66455105e-02 1.15904558e+00 1.24924868e-01 -2.29421034e-01
-2.93985367e-01 1.36305407e-01 3.31652910e-01 7.63352811e-01
-1.34811997e-01 -2.27776527e-01 -7.07802951e-01 1.02208257e-01
-1.47519779e+00 -1.05582774e+00 -1.01970315e-01 2.28437400e+00
6.95944786e-01 2.45454654e-01 -1.00425869e-01 6.39424980e-01
7.70607769e-01 2.68423766e-01 -1.98234037e-01 -6.70528591e-01
-1.56515390e-01 7.02853203e-01 5.26764989e-01 6.94280863e-01
-9.75988567e-01 2.71720678e-01 5.26354456e+00 1.06908977e+00
-1.26677442e+00 6.70292154e-02 2.43801683e-01 2.52637297e-01
-1.02449365e-01 -1.38802439e-01 -1.25234038e-01 5.72641313e-01
9.99066114e-01 -7.28956699e-01 6.55548334e-01 3.51681322e-01
3.59715253e-01 -6.51524007e-01 -2.52402961e-01 1.04124010e+00
1.91813648e-01 -8.57533097e-01 -2.35808820e-01 1.17135063e-01
4.75313306e-01 -2.48843953e-01 5.32098114e-01 -4.73695099e-01
-4.76000190e-01 -5.41064203e-01 6.29816115e-01 8.69588435e-01
1.23048663e+00 -9.18744981e-01 8.27725589e-01 -3.22384655e-01
-1.55125439e+00 7.08068907e-02 -2.21804351e-01 4.40664947e-01
2.38206554e-02 6.54183447e-01 -3.97055596e-01 8.94195199e-01
9.49739516e-01 5.06670177e-01 -8.46630931e-01 1.34052348e+00
9.61373653e-03 3.85406286e-01 -1.74008280e-01 3.13451797e-01
-1.51330769e-01 -4.00257766e-01 5.70005357e-01 8.46751869e-01
6.35685980e-01 3.40041965e-02 -2.25619569e-01 6.71247900e-01
4.16445136e-01 1.58418417e-01 -2.20309943e-01 2.76905037e-02
7.33731091e-01 1.45198452e+00 -8.80698204e-01 -1.61440685e-01
-4.36451524e-01 8.22847962e-01 -6.44860685e-01 1.87614903e-01
-8.22842419e-01 -1.00784433e+00 2.62842298e-01 3.96891683e-02
2.93815643e-01 -1.71689913e-02 -1.02136709e-01 -5.51710188e-01
2.02260748e-01 -7.63096988e-01 2.04354137e-01 -1.26037610e+00
-8.48976791e-01 8.63303483e-01 1.69914588e-01 -2.16974044e+00
-8.94491300e-02 -4.69028533e-01 -3.84145409e-01 1.05069113e+00
-1.55026007e+00 -6.68328583e-01 -7.55188048e-01 5.04739821e-01
6.49623647e-02 -2.83428561e-02 5.26359200e-01 6.45842016e-01
-1.67704910e-01 5.99263608e-01 2.59655744e-01 -5.71659580e-02
8.76743197e-01 -1.08678567e+00 1.05978698e-01 9.64926958e-01
-9.93381068e-02 1.10895187e-01 7.94714570e-01 -5.64930975e-01
-1.04349923e+00 -9.02811944e-01 6.76037252e-01 1.08384877e-01
4.93028581e-01 1.57449648e-01 -9.42615747e-01 1.90257967e-01
1.93031564e-01 -6.54108077e-02 6.20150089e-01 -8.08715641e-01
-6.06647320e-02 -5.81547379e-01 -1.32785773e+00 3.18128839e-02
5.06180584e-01 -5.25214493e-01 -1.51096061e-01 -1.19581841e-01
3.78178030e-01 -5.38351893e-01 -1.55946863e+00 4.61942345e-01
7.68379331e-01 -1.19167519e+00 9.39669251e-01 7.53721178e-01
2.50062138e-01 -8.73576999e-01 -4.34801459e-01 -1.09658849e+00
-5.29738009e-01 -4.76172924e-01 1.75710365e-01 9.41930354e-01
1.83206931e-01 -6.79802775e-01 2.88084835e-01 -2.22164318e-01
-2.38922331e-02 -7.24800050e-01 -1.05165887e+00 -9.70258176e-01
-4.67027724e-01 -3.73739488e-02 5.87761402e-01 6.61713064e-01
-3.42165343e-02 -1.66133925e-01 -2.65592247e-01 3.18795651e-01
7.78908074e-01 2.62465507e-01 2.41338044e-01 -9.98429418e-01
-1.22059666e-01 -3.75125349e-01 -9.86139715e-01 -3.53096694e-01
-5.45773268e-01 -5.76938748e-01 1.39903389e-02 -1.39847243e+00
1.63104773e-01 -3.63941163e-01 -6.39445841e-01 6.10249974e-02
-6.95380121e-02 1.02191973e+00 6.25255644e-01 3.10594767e-01
-2.81888932e-01 2.35697627e-01 1.55621779e+00 -1.08058006e-01
-1.56908050e-01 -2.28572533e-01 -5.02896488e-01 3.51802796e-01
9.45319057e-01 -5.12694180e-01 -3.29903185e-01 3.21334116e-02
4.68698004e-03 2.97488779e-01 2.07858130e-01 -1.48192596e+00
1.80384606e-01 -1.55042619e-01 3.26674283e-01 -7.81734228e-01
3.85976493e-01 -6.92509532e-01 7.21830428e-01 7.42277682e-01
-1.66252330e-02 2.38754183e-01 -5.10324389e-02 -1.67714674e-02
-5.94704032e-01 -1.59089237e-01 1.15269864e+00 3.34094882e-01
-9.50570941e-01 5.25084138e-02 -2.90089011e-01 -2.68244654e-01
1.04075897e+00 -6.39212310e-01 -6.61476791e-01 -3.09369177e-01
-3.40812445e-01 -2.96414047e-01 6.59274161e-01 2.32011512e-01
8.45711529e-01 -1.72269034e+00 -6.46563768e-01 3.05893630e-01
3.34056884e-01 -7.25454569e-01 3.72118652e-01 1.20754015e+00
-8.61247420e-01 1.44511223e-01 -5.95599532e-01 -5.53527772e-01
-1.36542857e+00 3.28232229e-01 4.08425421e-01 -9.67300609e-02
-4.40904409e-01 1.87501267e-01 -2.61030167e-01 -3.51990461e-02
-2.04548150e-01 -1.76537305e-01 -3.17802727e-01 -1.43529147e-01
7.17783213e-01 7.08133399e-01 1.86617374e-01 -1.07225513e+00
-4.29797500e-01 1.17238224e+00 4.31273907e-01 -3.95954698e-01
1.09851146e+00 -3.88354361e-01 -1.73763111e-01 1.41956344e-01
1.31536853e+00 8.05671141e-02 -7.91670799e-01 -3.58808815e-01
-1.38739318e-01 -1.00868309e+00 1.17942713e-01 -8.85600209e-01
-1.27968061e+00 4.67795104e-01 1.40751827e+00 2.61954755e-01
1.94427049e+00 -1.72131762e-01 8.35609555e-01 -5.94530851e-02
3.63785684e-01 -1.20198166e+00 1.75913244e-01 1.43779129e-01
8.37101579e-01 -9.04677570e-01 2.35450283e-01 -4.87106562e-01
-6.35883629e-01 1.18613231e+00 3.91871423e-01 -5.26619367e-02
8.32701862e-01 3.64257574e-01 3.84316832e-01 7.49865025e-02
-4.06158954e-01 -2.65347987e-01 3.65020126e-01 6.50594831e-01
1.71589434e-01 -7.05388337e-02 -5.52575886e-01 -6.07393160e-02
-3.24741125e-01 2.30082184e-01 6.66632652e-01 8.61062765e-01
-6.22611403e-01 -9.89675879e-01 -7.37762868e-01 4.28858221e-01
-3.89359623e-01 1.18733212e-01 -1.61037396e-03 6.80465460e-01
1.46581054e-01 1.08979201e+00 3.66920680e-01 -8.45806003e-01
4.00485963e-01 -4.28902984e-01 6.75347269e-01 1.35298222e-01
-1.09862536e-01 1.78095207e-01 -1.18834987e-01 -7.56388366e-01
-8.74674439e-01 -4.06719148e-01 -1.25564873e+00 -4.68804508e-01
-2.14802802e-01 1.29183093e-02 6.25484049e-01 3.71929914e-01
1.07672088e-01 4.20074552e-01 1.10393381e+00 -5.37601769e-01
-6.07600994e-02 -8.64576638e-01 -9.29062724e-01 4.57153738e-01
1.84236765e-01 -7.99071610e-01 -2.84937501e-01 8.56893808e-02] | [11.67646598815918, -1.951554775238037] |
b0077d2a-9d21-48a4-89b3-eda63335f098 | hyperspherical-embedding-for-point-cloud | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Zhang_Hyperspherical_Embedding_for_Point_Cloud_Completion_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Zhang_Hyperspherical_Embedding_for_Point_Cloud_Completion_CVPR_2023_paper.pdf | Hyperspherical Embedding for Point Cloud Completion | Most real-world 3D measurements from depth sensors are incomplete, and to address this issue the point cloud completion task aims to predict the complete shapes of objects from partial observations. Previous works often adapt an encoder-decoder architecture, where the encoder is trained to extract embeddings that are used as inputs to generate predictions from the decoder. However, the learned embeddings have sparse distribution in the feature space, which leads to worse generalization results during testing. To address these problems, this paper proposes a hyperspherical module, which transforms and normalizes embeddings from the encoder to be on a unit hypersphere. With the proposed module, the magnitude and direction of the output hyperspherical embedding are decoupled and only the directional information is optimized. We theoretically analyze the hyperspherical embedding and show that it enables more stable training with a wider range of learning rates and more compact embedding distributions. Experiment results show consistent improvement of point cloud completion in both single-task and multi-task learning, which demonstrates the effectiveness of the proposed method. | ['Matthew Johnson-Roberson', 'Ram Vasudevan', 'Haomeng Zhang', 'Junming Zhang'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['point-cloud-completion'] | ['computer-vision'] | [ 3.28074954e-02 1.84944078e-01 -7.94525295e-02 -5.98907113e-01
-2.25488946e-01 -3.20957899e-01 5.55708289e-01 -1.60068050e-01
-1.75161675e-01 5.10261118e-01 2.64747977e-01 -2.43139379e-02
-3.42198201e-02 -8.19873273e-01 -9.66873229e-01 -7.49603629e-01
1.01521805e-01 5.14642596e-01 6.82484061e-02 1.37723163e-01
2.97775239e-01 5.12522578e-01 -1.49986041e+00 8.57018530e-02
8.48131180e-01 1.09847260e+00 7.10731626e-01 4.93903190e-01
-2.16719389e-01 3.26502651e-01 -4.92367268e-01 -1.63888469e-01
4.69265610e-01 8.25346038e-02 -1.74778491e-01 2.53922064e-02
1.80522561e-01 -5.50368845e-01 -6.96444452e-01 9.87722397e-01
4.78664041e-01 6.25258386e-02 9.34606314e-01 -1.25565660e+00
-9.45520401e-01 2.09763348e-01 -3.79657209e-01 -1.20746851e-01
-6.92475215e-02 -2.05751564e-02 9.81466353e-01 -1.14475834e+00
2.58312523e-01 1.16958499e+00 3.26472372e-01 5.66734791e-01
-7.88838446e-01 -8.27674806e-01 6.77031800e-02 2.08122328e-01
-1.54928899e+00 -2.57209808e-01 9.62285459e-01 -4.65762019e-01
1.10185242e+00 -1.63856730e-01 6.67283356e-01 8.81638288e-01
2.74640650e-01 7.07556427e-01 5.61916173e-01 -1.00851439e-01
2.22720325e-01 4.87290382e-01 -4.25035119e-01 4.46975976e-01
4.20772940e-01 2.25919142e-01 -3.93352866e-01 1.11393906e-01
9.80781496e-01 3.66819739e-01 -3.48683655e-01 -6.99664593e-01
-1.29654300e+00 8.79995346e-01 6.50054097e-01 -1.12557836e-01
-2.74687469e-01 -5.79217821e-02 4.38450053e-02 1.87212825e-01
3.98238361e-01 2.43628636e-01 -2.65184850e-01 -1.44636065e-01
-5.19261420e-01 2.32203543e-01 5.96590161e-01 1.35786939e+00
8.82443368e-01 1.11083768e-01 1.83205858e-01 8.75597715e-01
6.39584005e-01 8.93402815e-01 5.78294575e-01 -4.87575471e-01
8.56848896e-01 7.24414289e-01 2.58089807e-02 -9.46244776e-01
-2.52279460e-01 -2.57948279e-01 -7.24749625e-01 4.32925642e-01
-1.03013501e-01 -1.98420495e-01 -8.59549642e-01 1.56629670e+00
3.97301793e-01 3.38977545e-01 3.10952812e-01 1.24485612e+00
7.21528828e-01 9.07162845e-01 -3.59375864e-01 3.39045763e-01
9.08287108e-01 -8.18392515e-01 -4.72460121e-01 -4.03422982e-01
3.72395545e-01 -5.95063210e-01 1.01262164e+00 2.42605850e-01
-6.35267317e-01 -5.89114547e-01 -1.51265419e+00 -3.25953424e-01
-3.40052158e-01 3.98906082e-01 3.88245583e-01 2.34947830e-01
-6.06809080e-01 6.62150025e-01 -9.73318398e-01 5.99031113e-02
3.56146604e-01 4.73262131e-01 -4.67508376e-01 -1.87196583e-02
-9.20971692e-01 8.43894541e-01 5.65865457e-01 -4.84613180e-02
-7.86486506e-01 -5.17810822e-01 -1.03765130e+00 1.91085711e-01
-2.59151220e-01 -4.45947260e-01 8.56888473e-01 -3.77881110e-01
-1.55965340e+00 4.59212601e-01 -9.62898880e-02 -2.21132666e-01
1.36380434e-01 -2.86000341e-01 -2.25212485e-01 6.85336068e-02
2.37476863e-02 1.00997829e+00 1.01161325e+00 -1.11480856e+00
-5.73727190e-01 -6.38536334e-01 -6.86931163e-02 6.45110369e-01
-7.19316363e-01 -7.53155231e-01 -4.79255050e-01 -4.00676668e-01
3.40306193e-01 -9.83453214e-01 1.23781905e-01 2.07774669e-01
-2.44497016e-01 -1.20806046e-01 9.68211055e-01 -4.82329369e-01
7.62462914e-01 -2.37669969e+00 3.89418006e-01 -4.13144045e-02
2.05897555e-01 -4.72130664e-02 -9.50317010e-02 2.58688152e-01
-2.01246426e-01 -2.10483804e-01 -1.85502917e-01 -5.72309077e-01
1.05349965e-01 4.22699779e-01 -6.40064836e-01 5.62086880e-01
3.74957979e-01 6.78805768e-01 -6.40810728e-01 -2.74602890e-01
5.31915963e-01 6.90858126e-01 -5.37890196e-01 5.06844640e-01
-1.45493478e-01 3.88384819e-01 -7.01317012e-01 5.58538795e-01
9.83713806e-01 -2.07966968e-01 -3.53731960e-01 -1.32489624e-02
2.41171774e-02 2.79922366e-01 -1.08485246e+00 1.86544371e+00
-5.35355747e-01 6.87736273e-01 -7.30038509e-02 -9.85618651e-01
1.36098099e+00 1.11494079e-01 4.93574053e-01 -5.39118648e-01
5.21466136e-02 1.26681477e-01 -1.33428097e-01 -6.41022563e-01
3.26141298e-01 -4.32895571e-01 2.86560923e-01 2.00436383e-01
7.98012093e-02 -3.43340695e-01 -3.79868329e-01 -2.21418113e-01
8.27430069e-01 2.28866175e-01 -3.54994903e-03 6.40298203e-02
4.23172176e-01 -1.80036709e-01 5.65834284e-01 1.13498546e-01
1.40419543e-01 8.81759286e-01 2.40787521e-01 -4.33310091e-01
-1.30266619e+00 -1.20464408e+00 -4.24959660e-01 5.39846003e-01
4.87952620e-01 -2.21744612e-01 -3.28394830e-01 -5.29934704e-01
4.04767364e-01 7.30134666e-01 -5.18173873e-01 -4.12563860e-01
-3.54437113e-01 -5.68935156e-01 2.19429865e-01 6.22856975e-01
3.41725469e-01 -9.48607683e-01 -5.76980650e-01 -2.34343503e-02
-6.21141009e-02 -1.27443576e+00 -2.18595818e-01 8.21292028e-02
-1.24128234e+00 -9.04495299e-01 -7.27015793e-01 -8.21875870e-01
8.44459653e-01 2.94926316e-01 4.94797349e-01 -2.24220425e-01
-5.44531643e-02 -1.36764245e-02 -3.94047648e-01 -6.74943566e-01
-1.25041567e-02 7.93477967e-02 2.26793349e-01 4.01216522e-02
6.93574190e-01 -7.79693842e-01 -5.58036268e-01 2.48229757e-01
-9.33568358e-01 1.31937981e-01 8.58212352e-01 6.45992160e-01
5.95703363e-01 -1.19768299e-01 6.04474127e-01 -4.69565183e-01
4.95892942e-01 -5.69683790e-01 -7.02709675e-01 -4.80927490e-02
-6.87588751e-01 4.27600324e-01 8.73811305e-01 -3.17404628e-01
-7.70574808e-01 2.22538903e-01 -1.50750250e-01 -9.55924451e-01
1.69313245e-03 1.81987077e-01 -2.88033694e-01 -1.16058528e-01
4.91110981e-01 4.84760642e-01 1.37506515e-01 -5.80838978e-01
2.65314460e-01 9.95028973e-01 3.04388911e-01 -4.64178085e-01
9.62440014e-01 5.16905129e-01 -5.34769930e-02 -9.63516891e-01
-3.68873656e-01 -4.73056555e-01 -6.97448790e-01 1.36074722e-02
7.50893176e-01 -1.21037841e+00 -6.08346641e-01 7.01982006e-02
-1.21901345e+00 -5.55837061e-03 -3.85555476e-01 9.43590999e-01
-6.17180586e-01 4.11009341e-01 -1.68100834e-01 -6.40548348e-01
-3.08512717e-01 -1.20427155e+00 1.39827430e+00 3.03030252e-01
3.32594961e-01 -9.27923858e-01 1.28586292e-01 -4.23842259e-02
1.20682664e-01 1.27271831e-01 1.02366734e+00 -5.40698051e-01
-8.00098419e-01 -2.76815355e-01 -4.12186116e-01 5.46360493e-01
1.89477965e-01 -2.38597780e-01 -1.02554488e+00 -3.59409243e-01
4.01217729e-01 -2.60926664e-01 8.76868725e-01 2.14673415e-01
1.20593691e+00 -9.98865440e-02 -4.68248010e-01 1.08934176e+00
1.54532444e+00 7.70778731e-02 5.63609302e-01 3.50629717e-01
7.10732937e-01 4.28719282e-01 5.41647911e-01 6.66722238e-01
4.11394209e-01 3.64450544e-01 8.97165298e-01 3.11652064e-01
2.58105993e-02 -4.88040000e-01 3.93159807e-01 1.02957368e+00
1.68963403e-01 -1.74381927e-01 -8.63410771e-01 5.70536017e-01
-1.68948054e+00 -3.89264435e-01 4.17364687e-01 2.36020422e+00
5.46668828e-01 1.26695499e-01 -4.89982307e-01 -1.43834785e-01
5.61998785e-01 2.52378225e-01 -1.01779354e+00 -1.43490210e-01
-5.61497584e-02 -3.78322117e-02 5.25327265e-01 4.42338556e-01
-8.46348405e-01 8.17523658e-01 6.06127834e+00 4.59930152e-01
-1.44824457e+00 3.12722474e-03 -3.59962732e-02 -3.54288131e-01
-4.29609865e-01 -8.80367458e-02 -9.20296550e-01 5.53483367e-01
6.21327698e-01 -2.05936074e-01 4.44732368e-01 1.10782695e+00
2.94732880e-02 4.71806586e-01 -1.37402844e+00 1.34833980e+00
3.08491468e-01 -1.17602789e+00 2.54558921e-01 2.65542179e-01
5.67291737e-01 5.47300994e-01 1.86044071e-02 4.45376396e-01
-2.42940202e-01 -9.65027153e-01 6.36436224e-01 5.57666898e-01
9.68692660e-01 -5.96621156e-01 6.50530100e-01 6.88953876e-01
-1.12574708e+00 -2.66946554e-01 -1.19337845e+00 -1.02054395e-01
4.18537594e-02 5.02582669e-01 -1.31919873e+00 3.85440737e-01
5.16111791e-01 1.05092835e+00 -4.60952103e-01 1.10791636e+00
-2.20242903e-01 2.03198612e-01 -5.95742404e-01 -2.00599849e-01
-9.03469743e-04 -5.43692112e-01 7.41080225e-01 8.71494353e-01
6.87117338e-01 7.65209422e-02 4.29510623e-02 1.09441531e+00
-1.30647004e-01 5.15224449e-02 -9.70690608e-01 -1.80797443e-01
4.44171369e-01 1.12781823e+00 -2.04442725e-01 -1.07845694e-01
-5.47949553e-01 8.03867757e-01 5.62662542e-01 4.18683320e-01
-8.13495755e-01 -6.03547454e-01 8.20866048e-01 1.28454998e-01
5.76651692e-01 -5.14125824e-01 -5.23139834e-01 -1.34577215e+00
2.94961333e-01 -2.44731843e-01 -1.37341529e-01 -9.23497796e-01
-1.11846280e+00 4.89295244e-01 -1.55711537e-02 -1.72597706e+00
-1.90979183e-01 -9.61311698e-01 -6.84694767e-01 8.89762700e-01
-1.77023423e+00 -9.22010601e-01 -4.70922530e-01 5.41091919e-01
4.63930964e-01 -4.88823414e-01 6.35561466e-01 2.45050594e-01
-2.98275262e-01 5.05473375e-01 4.48730171e-01 1.14123054e-01
5.44914544e-01 -1.21942282e+00 3.33751261e-01 4.37221110e-01
1.51822865e-01 4.53673810e-01 5.18826544e-01 -6.69114470e-01
-1.64307594e+00 -1.32487285e+00 5.35239458e-01 -3.78570497e-01
3.33733082e-01 -6.04536772e-01 -9.19505656e-01 6.24768198e-01
-1.37804955e-01 3.75258550e-02 5.66310942e-01 -5.11629023e-02
-2.07873523e-01 -2.39789754e-01 -8.92544270e-01 3.63875717e-01
8.06991220e-01 -6.02449298e-01 -7.86917269e-01 5.58627307e-01
8.27838182e-01 -5.96920609e-01 -9.75430727e-01 5.02551019e-01
4.36657041e-01 -7.10652471e-01 8.87529314e-01 -4.75545883e-01
6.66058660e-01 -2.79317498e-01 -4.63765442e-01 -1.39053094e+00
-2.32323840e-01 3.50055061e-02 -3.62227201e-01 7.54775643e-01
4.50904369e-01 -7.38954484e-01 1.05859923e+00 2.73410022e-01
-3.98628086e-01 -1.00207448e+00 -1.07726502e+00 -7.23996699e-01
2.22985163e-01 -2.69710720e-01 8.98582518e-01 6.79627061e-01
-2.07575820e-02 4.92990553e-01 -2.16834992e-01 5.75420380e-01
5.93147814e-01 3.80256474e-01 8.25047612e-01 -1.26146090e+00
-1.13962486e-01 -1.43249795e-01 -7.03482866e-01 -1.56240535e+00
2.17052072e-01 -1.02819729e+00 1.56627446e-01 -1.49965763e+00
-5.38881645e-02 -5.25929630e-01 -1.84916899e-01 2.87140697e-01
8.11425149e-02 -1.80231750e-01 9.35901254e-02 4.25661027e-01
-1.91335663e-01 1.38500667e+00 1.37591147e+00 -2.48955891e-01
-1.34685725e-01 2.79635806e-02 -4.63419139e-01 6.00917220e-01
6.84142113e-01 -5.33648431e-01 -7.60383964e-01 -1.03717065e+00
5.13056703e-02 -2.95495391e-02 1.33086503e-01 -1.13800395e+00
2.90587932e-01 -7.69088194e-02 7.07371533e-01 -7.19769657e-01
8.37123990e-01 -1.01639521e+00 -2.17852786e-01 9.59320553e-03
-7.38434047e-02 -1.02945149e-01 1.62722394e-01 8.49682689e-01
-3.09731632e-01 -2.91509032e-01 4.95463789e-01 1.64599344e-01
-4.49368596e-01 7.03967154e-01 3.68607759e-01 -2.14252308e-01
1.03843594e+00 -4.79693830e-01 -6.52029514e-02 -3.03265721e-01
-5.38578510e-01 3.87863755e-01 5.96308291e-01 8.35269094e-01
1.16437507e+00 -1.56013560e+00 -7.43034720e-01 7.23579049e-01
3.30382258e-01 5.42824388e-01 1.28861338e-01 3.94151717e-01
-5.61096191e-01 4.59408492e-01 -3.26376855e-01 -9.70129371e-01
-5.97979784e-01 5.12766838e-01 3.42778027e-01 3.55097562e-01
-1.02870810e+00 7.52366006e-01 4.83835071e-01 -8.24791312e-01
1.69347450e-01 -4.81774241e-01 -2.17645407e-01 -3.00593823e-01
4.33675975e-01 3.21966782e-02 5.46016060e-02 -5.94561040e-01
-1.81956366e-01 7.06889987e-01 -3.59401666e-02 -9.30450391e-03
1.52822649e+00 -5.12149967e-02 1.55146047e-01 4.86453295e-01
1.42856359e+00 -2.70147264e-01 -1.63021588e+00 -3.78292590e-01
-5.11896551e-01 -7.78349578e-01 1.05406508e-01 -1.09164767e-01
-9.73167598e-01 1.34258497e+00 6.31895959e-01 -9.23028812e-02
8.27422500e-01 -1.04723617e-01 8.22293460e-01 6.06068969e-01
2.48992518e-01 -9.35192108e-01 -9.26037319e-03 5.37262261e-01
1.01434886e+00 -1.32463801e+00 3.69011946e-02 -2.27599770e-01
-6.82136655e-01 1.33923686e+00 7.80701935e-01 -2.95849800e-01
9.06235278e-01 1.15898907e-01 -2.51159906e-01 -2.53556341e-01
-7.88106561e-01 1.92418560e-01 1.77751079e-01 6.46004200e-01
8.68902914e-03 6.56831264e-02 3.82817864e-01 6.20509684e-01
-5.82024872e-01 -2.24871710e-01 3.38922650e-01 6.78021610e-01
-7.33908594e-01 -7.99932778e-01 -2.06646368e-01 4.84995335e-01
1.56208023e-01 7.97202662e-02 -2.20882162e-01 5.74326217e-01
9.69766602e-02 5.34010291e-01 2.38496065e-01 -6.40103996e-01
4.43623811e-01 -1.21330451e-02 4.31644082e-01 -7.79436827e-01
2.07735866e-01 -1.65623546e-01 -5.05824029e-01 -3.00234884e-01
7.41973147e-02 -4.60187167e-01 -1.56652892e+00 2.30492540e-02
-4.31988239e-01 3.05712491e-01 1.05686665e+00 8.94064546e-01
6.30910039e-01 3.53873312e-01 8.24223757e-01 -9.38582242e-01
-8.89944851e-01 -1.12198579e+00 -7.68785119e-01 4.31841791e-01
5.41834772e-01 -9.32828128e-01 -4.38472539e-01 -1.70308888e-01] | [8.221424102783203, -3.4285433292388916] |
82243d90-af11-4415-a76e-c56f6457cd71 | ai-security-for-geoscience-and-remote-sensing | 2212.09360 | null | https://arxiv.org/abs/2212.09360v2 | https://arxiv.org/pdf/2212.09360v2.pdf | AI Security for Geoscience and Remote Sensing: Challenges and Future Trends | Recent advances in artificial intelligence (AI) have significantly intensified research in the geoscience and remote sensing (RS) field. AI algorithms, especially deep learning-based ones, have been developed and applied widely to RS data analysis. The successful application of AI covers almost all aspects of Earth observation (EO) missions, from low-level vision tasks like super-resolution, denoising and inpainting, to high-level vision tasks like scene classification, object detection and semantic segmentation. While AI techniques enable researchers to observe and understand the Earth more accurately, the vulnerability and uncertainty of AI models deserve further attention, considering that many geoscience and RS tasks are highly safety-critical. This paper reviews the current development of AI security in the geoscience and RS field, covering the following five important aspects: adversarial attack, backdoor attack, federated learning, uncertainty and explainability. Moreover, the potential opportunities and trends are discussed to provide insights for future research. To the best of the authors' knowledge, this paper is the first attempt to provide a systematic review of AI security-related research in the geoscience and RS community. Available code and datasets are also listed in the paper to move this vibrant field of research forward. | ['Pedram Ghamisi', 'Peter M. Atkinson', 'Shizhen Chang', 'Weikang Yu', 'Tao Bai', 'Yonghao Xu'] | 2022-12-19 | null | null | null | null | ['scene-classification'] | ['computer-vision'] | [ 4.12806779e-01 6.09569177e-02 2.92066205e-03 -1.09818161e-01
-2.50621080e-01 -6.00380540e-01 6.46405935e-01 3.02637845e-01
-3.10061812e-01 5.31197906e-01 -3.19874316e-01 -5.64023316e-01
-2.78022438e-01 -1.16917753e+00 -5.65092623e-01 -9.69400644e-01
-5.33220172e-01 8.01621303e-02 -1.28510743e-01 -3.89042258e-01
2.03427374e-01 8.20416749e-01 -1.59806716e+00 -9.81886685e-02
1.14435923e+00 1.31630671e+00 -1.10415816e-01 6.42419279e-01
2.12880909e-01 7.92979658e-01 -3.99545074e-01 -3.96818221e-01
4.58128214e-01 -2.51409300e-02 -6.08912647e-01 -2.57965714e-01
7.86445215e-02 -3.03900808e-01 -2.22562566e-01 1.51476789e+00
5.48708141e-01 1.02978230e-01 4.73623067e-01 -1.23739672e+00
-6.96353495e-01 6.05062485e-01 -6.56025946e-01 4.34512734e-01
-1.51823051e-02 1.72850087e-01 7.25216687e-01 -5.98620296e-01
-1.63557567e-02 9.79273438e-01 5.49732387e-01 2.38440588e-01
-5.74714899e-01 -9.34630990e-01 2.90108800e-01 4.28089887e-01
-1.32499409e+00 -1.29529804e-01 6.36842787e-01 -5.48793435e-01
9.10569727e-01 4.09827888e-01 6.18735135e-01 6.28116131e-01
3.35523158e-01 8.25240672e-01 1.22125649e+00 -3.26801121e-01
4.60735559e-01 -3.96054983e-02 -1.13417141e-01 3.43235344e-01
4.62680489e-01 7.62030065e-01 -1.65022507e-01 -1.70161724e-01
3.51628363e-01 3.48815918e-02 -1.69695079e-01 6.43489957e-02
-1.01678431e+00 1.15550017e+00 8.65908206e-01 3.09953600e-01
-6.49501622e-01 8.89353007e-02 2.36550719e-01 3.30606669e-01
6.77256405e-01 5.82346201e-01 -3.14431787e-01 5.49718320e-01
-1.02146316e+00 2.47911870e-01 6.95589364e-01 5.66567063e-01
5.87188721e-01 9.51676190e-01 3.69808495e-01 2.74811566e-01
5.28450608e-01 9.23998833e-01 3.65076475e-02 -7.99078107e-01
1.08635820e-01 1.14247635e-01 5.13227619e-02 -1.44674385e+00
-3.23912829e-01 -5.43050230e-01 -1.36723876e+00 4.57782120e-01
-2.56148458e-01 -3.72446686e-01 -9.19521034e-01 1.30875504e+00
3.66468310e-01 4.78553563e-01 4.24687862e-01 1.01297414e+00
9.33290899e-01 1.12367141e+00 1.95746005e-01 9.47789699e-02
1.28495717e+00 -6.05744958e-01 -5.87008238e-01 -7.80412614e-01
4.96683381e-02 -4.73025411e-01 2.38513276e-01 3.55530947e-01
-6.39621973e-01 -3.23678762e-01 -1.08841181e+00 5.47955751e-01
-7.70576894e-01 -2.36540064e-01 9.73209381e-01 6.89872921e-01
-7.59593844e-01 4.47140187e-01 -8.26792300e-01 -1.73837140e-01
6.94542110e-01 2.59854585e-01 -1.21645793e-01 -3.53437252e-02
-1.94218326e+00 1.24639201e+00 4.23492700e-01 2.92405784e-01
-1.22274292e+00 -7.70787120e-01 -9.74737585e-01 -4.81611444e-03
4.71239418e-01 -3.90742958e-01 8.24692428e-01 -8.88619065e-01
-1.25153983e+00 8.15600395e-01 4.55246449e-01 -1.16172087e+00
3.33925426e-01 -4.52415645e-01 -6.34630322e-01 1.81924611e-01
-7.04071373e-02 4.20758247e-01 1.12115431e+00 -1.26067126e+00
-6.11501455e-01 -5.63222408e-01 -2.72935443e-02 -9.07812566e-02
-1.29359350e-01 2.15357736e-01 3.75275284e-01 -8.95864367e-01
-6.99317157e-02 -1.08620393e+00 -5.78918219e-01 1.28098994e-01
-2.89857775e-01 1.78852931e-01 8.37601602e-01 -7.89771914e-01
1.06200278e+00 -1.96609533e+00 1.53063565e-01 2.36318842e-01
1.26785338e-01 6.72725320e-01 7.02895597e-02 4.93805796e-01
-5.05237542e-02 2.93365985e-01 -8.52019846e-01 1.99995413e-01
-3.51596892e-01 8.66953209e-02 -9.21644986e-01 7.46632338e-01
2.27688998e-01 7.64165819e-01 -8.96399379e-01 2.06846185e-02
6.31613970e-01 4.78224605e-01 2.06745826e-02 1.26798928e-01
-3.78353626e-01 7.73665190e-01 -8.83056819e-01 9.70694661e-01
8.74919832e-01 1.17205732e-01 -3.71030986e-01 7.56696332e-03
-2.90432006e-01 -2.77377337e-01 -1.05165982e+00 1.13014436e+00
-3.73390436e-01 6.18677199e-01 3.51266146e-01 -1.32960999e+00
9.90831912e-01 3.00719827e-01 4.71987307e-01 -5.86492479e-01
1.04205571e-02 2.20037192e-01 5.48108406e-02 -4.74208653e-01
4.93777901e-01 -9.05889198e-02 -1.09890044e-01 4.65946376e-01
-5.05417347e-01 -8.26732397e-01 -2.93659717e-01 1.76242031e-02
4.66896087e-01 -8.99630934e-02 4.89636451e-01 -1.55337587e-01
6.86278462e-01 2.83704758e-01 2.46449813e-01 7.62955368e-01
-3.06712806e-01 8.76349807e-02 -3.17683220e-01 -8.70209098e-01
-8.07514012e-01 -7.64855444e-01 -3.41041476e-01 8.02058816e-01
2.19816521e-01 2.06635550e-01 -5.90158224e-01 -3.85873973e-01
2.06234187e-01 8.34226310e-01 -5.11485040e-01 -3.26429009e-01
-1.79164633e-01 -1.22354984e+00 8.37952912e-01 2.53258348e-01
9.61871088e-01 -1.15479112e+00 -9.01923239e-01 9.48030874e-02
-1.31473646e-01 -1.07443297e+00 5.70666552e-01 -1.29373252e-01
-8.56919229e-01 -9.56559122e-01 -6.49172962e-01 -1.32259890e-01
3.80495101e-01 5.48117340e-01 9.26953793e-01 5.74225513e-03
-4.66442466e-01 4.45349336e-01 -4.97895747e-01 -1.09764636e+00
-4.45545316e-01 -3.65605116e-01 2.98979223e-01 -1.51963755e-02
3.85021031e-01 -4.14037675e-01 -6.25064492e-01 8.99921879e-02
-1.18928230e+00 -4.31998014e-01 6.32868171e-01 7.29704499e-01
3.95131230e-01 6.65218413e-01 4.43172842e-01 -5.39289951e-01
1.53698057e-01 -9.37053204e-01 -8.75726521e-01 3.22592378e-01
-7.55790532e-01 -2.08935231e-01 5.41005671e-01 -7.50907809e-02
-1.00057888e+00 5.82504738e-03 -1.06313236e-01 -2.85800844e-01
-4.31766242e-01 9.59315121e-01 6.75854832e-02 -5.74392915e-01
8.40909123e-01 3.44821244e-01 -1.23392127e-01 -2.34369293e-01
3.87764871e-01 7.90348530e-01 5.09511530e-01 -2.93459147e-01
1.25640988e+00 8.60968769e-01 1.05798826e-01 -1.21105671e+00
-9.94745374e-01 -1.96815118e-01 -9.61559117e-02 -3.23221624e-01
8.81092370e-01 -1.07456505e+00 -3.30334008e-01 7.86676407e-01
-1.00975370e+00 -9.23425630e-02 -2.44616136e-01 6.71537757e-01
-6.08961210e-02 5.73244452e-01 -2.61847317e-01 -1.14581680e+00
-8.54457736e-01 -1.08306563e+00 8.34370792e-01 4.32607681e-01
2.10570157e-01 -9.75219250e-01 1.23075629e-02 3.10272783e-01
7.61058271e-01 6.19405031e-01 4.76923704e-01 -4.92038935e-01
-8.20373058e-01 -4.00991410e-01 -7.61364400e-02 4.06569719e-01
-1.00203209e-01 1.65770322e-01 -1.19691610e+00 -4.19250250e-01
2.30575249e-01 -3.27467829e-01 1.25410044e+00 6.78142250e-01
1.09155107e+00 -4.40479457e-01 -1.28088608e-01 8.56845975e-01
1.35387516e+00 3.21904063e-01 7.57271409e-01 7.50571012e-01
5.65642178e-01 9.44944859e-01 8.52935731e-01 7.20006287e-01
1.87132642e-01 3.97477262e-02 1.25491893e+00 -2.80224115e-01
4.69748676e-01 2.33684152e-01 2.19296098e-01 1.00897826e-01
-3.66830677e-01 -4.57480669e-01 -1.22616613e+00 4.11431462e-01
-1.46117568e+00 -1.25130975e+00 4.02956232e-02 2.10624957e+00
3.18179399e-01 -3.15187573e-01 -3.43590796e-01 3.08413029e-01
7.78231561e-01 6.78233087e-01 -7.80233324e-01 -1.10407971e-01
-2.66842812e-01 1.36636570e-01 8.42347980e-01 3.15096110e-01
-1.69569683e+00 1.27024174e+00 5.80052233e+00 7.47930169e-01
-1.49662447e+00 -2.00964168e-01 6.96199358e-01 5.92126191e-01
-3.99042398e-01 -5.86592266e-03 -5.19576013e-01 2.45860070e-01
7.94150889e-01 -1.85311928e-01 5.89619935e-01 1.02534485e+00
6.46058246e-02 -1.72585696e-01 -4.87418741e-01 6.45983994e-01
-6.80865645e-02 -1.40308797e+00 1.38367638e-01 -8.68609995e-02
8.03879023e-01 5.49403071e-01 5.56978762e-01 -1.33847445e-01
4.72210050e-01 -1.21078598e+00 5.38732290e-01 2.73325920e-01
5.43887317e-01 -9.54446852e-01 7.70704508e-01 3.10427099e-01
-1.22849846e+00 -3.19668740e-01 -4.64241862e-01 -9.82198864e-02
-8.56203064e-02 8.27779233e-01 -2.47030824e-01 8.76035035e-01
1.19061816e+00 9.53985989e-01 -1.28129125e-01 1.05416048e+00
-4.64496166e-01 6.89760864e-01 -3.97420675e-01 1.39798313e-01
4.36495572e-01 -3.08094084e-01 9.95924652e-01 9.42900360e-01
3.89349043e-01 5.66398263e-01 1.17542237e-01 8.69065106e-01
2.72968411e-01 -3.27208728e-01 -9.22588587e-01 -4.58710372e-01
5.51997900e-01 1.06100726e+00 -3.92110527e-01 2.20823362e-02
-3.39536041e-01 4.92884248e-01 -4.98102069e-01 3.24051768e-01
-6.29165232e-01 -3.51481974e-01 7.64389515e-01 -3.28195482e-01
-5.28014218e-03 -3.71017635e-01 -4.83502060e-01 -1.15309870e+00
-4.17465419e-01 -1.08031952e+00 6.13582015e-01 -4.84161437e-01
-1.25573361e+00 7.89913833e-01 1.17844880e-01 -1.32216930e+00
-9.39056650e-02 -4.44615930e-01 -8.10876608e-01 5.99083662e-01
-2.18579769e+00 -1.33011818e+00 -4.38293546e-01 4.58317041e-01
2.66091287e-01 -6.75309896e-01 9.58818793e-01 -7.30540901e-02
-4.88097399e-01 5.46651147e-02 2.00313523e-01 6.64896071e-02
1.42491266e-01 -7.04905808e-01 7.21139669e-01 1.36106467e+00
6.60140663e-02 3.14523727e-01 9.16297257e-01 -6.79378867e-01
-1.54390442e+00 -1.32517326e+00 4.46077377e-01 6.37592524e-02
8.30329001e-01 2.47663468e-01 -7.90317237e-01 5.84568620e-01
2.71148272e-02 1.12895325e-01 6.00849807e-01 -5.84225178e-01
-2.91860908e-01 -2.00992167e-01 -1.38268924e+00 4.28621858e-01
2.66297966e-01 -4.86109734e-01 -3.76771688e-01 4.61903781e-01
5.86167693e-01 -2.80295730e-01 -8.38233829e-01 6.88297331e-01
2.16581360e-01 -9.14129436e-01 1.40308261e+00 -4.06997204e-01
3.42555165e-01 -3.44460011e-01 -1.86116695e-01 -1.24818289e+00
-1.36057019e-01 -5.62294483e-01 -2.45117486e-01 7.90887177e-01
4.47954014e-02 -8.81145895e-01 5.46646714e-01 4.87935662e-01
-1.89108506e-01 -4.18374985e-01 -8.84270430e-01 -6.58757031e-01
2.09801257e-01 -5.37667513e-01 7.20646501e-01 1.17355382e+00
-4.99995232e-01 -2.55548447e-01 -7.34687507e-01 9.49129164e-01
1.16218805e+00 2.46162415e-01 4.70402420e-01 -1.44473946e+00
-3.31752039e-02 -3.28547776e-01 -4.00749385e-01 -3.81476641e-01
1.63853973e-01 -6.06388330e-01 -1.47051677e-01 -1.26390159e+00
-4.19321954e-01 -5.63122153e-01 -2.10913613e-01 4.83022362e-01
-1.67452350e-01 1.87873363e-01 1.86104119e-01 3.56399655e-01
-5.37271164e-02 5.56224763e-01 8.04613590e-01 -5.41552007e-01
-2.29961425e-02 5.54125071e-01 -7.14121222e-01 8.19228709e-01
1.27769327e+00 -5.12179434e-01 -2.21162334e-01 -7.99043953e-01
4.45442170e-01 8.37229043e-02 6.39332473e-01 -9.36526895e-01
2.96523541e-01 -5.48992634e-01 2.32594743e-01 -4.85400707e-01
4.96363901e-02 -1.17875648e+00 1.72264546e-01 9.12973046e-01
-1.47106886e-01 -2.33081296e-01 4.65831637e-01 6.48878992e-01
-4.93877620e-01 -5.08350283e-02 1.13636386e+00 -3.21099222e-01
-1.09959662e+00 7.56197870e-01 -6.58293545e-01 -1.72680497e-01
1.42315388e+00 9.32349414e-02 -3.04222614e-01 -5.40528059e-01
-5.69954455e-01 4.39234167e-01 7.87876546e-02 3.25944334e-01
8.64328265e-01 -7.00033426e-01 -1.16706884e+00 1.75598770e-01
8.45346227e-02 2.73643583e-01 5.95437765e-01 4.74359214e-01
-6.97529376e-01 3.65412533e-01 -2.17555955e-01 -4.30028498e-01
-7.81563282e-01 7.47831106e-01 4.45873678e-01 -1.23608083e-01
-4.88322198e-01 1.06242645e+00 1.79881573e-01 -3.51651609e-01
2.90107548e-01 1.29422352e-01 -4.53810900e-01 -1.20912500e-01
8.59702289e-01 5.95168710e-01 -1.45937204e-01 -6.63922429e-01
-5.20041883e-01 6.72250032e-01 1.84112072e-01 -1.78347912e-03
1.58502436e+00 -1.72224924e-01 -2.30402529e-01 1.27922716e-02
4.89866197e-01 -2.95901448e-01 -1.03491211e+00 -3.30308318e-01
7.03729242e-02 -4.00225043e-01 6.56948924e-01 -8.30608547e-01
-1.48106074e+00 1.31617439e+00 6.82252109e-01 3.80870908e-01
1.28553724e+00 -3.33665848e-01 8.55432689e-01 5.05577385e-01
4.39854145e-01 -9.79951262e-01 -9.92880687e-02 5.00686765e-01
1.09310853e+00 -1.57037389e+00 4.37881172e-01 -1.21826187e-01
-8.83380651e-01 1.05862808e+00 3.15969259e-01 -1.71113342e-01
9.37167227e-01 3.64996552e-01 8.52825269e-02 -1.59555763e-01
-2.67343465e-02 -3.18691134e-01 3.32303166e-01 9.17993784e-01
-8.72441083e-02 8.46063569e-02 3.65408838e-01 1.85210124e-01
6.75245970e-02 -1.75998613e-01 3.43163043e-01 9.38323975e-01
-7.21861064e-01 -7.75158644e-01 -7.35493243e-01 2.00578377e-01
-7.66890347e-01 -4.62914497e-01 -3.70254397e-01 4.11897421e-01
-2.07054108e-01 1.17019081e+00 -2.05315799e-01 -3.64521414e-01
-2.46732891e-01 -6.05194807e-01 -1.69465899e-01 -3.16486418e-01
-4.90641952e-01 -2.60761857e-01 -1.26487195e-01 -3.94868463e-01
-7.19038725e-01 -5.18779933e-01 -1.16885114e+00 -5.12391925e-01
-3.97046119e-01 2.78562993e-01 1.05247188e+00 9.19304192e-01
2.21519783e-01 2.32791528e-01 9.01890993e-01 -1.07752621e+00
-5.44791818e-01 -5.97114265e-01 -6.36033118e-01 -4.02098030e-01
5.74387431e-01 -4.18010890e-01 -5.68507612e-01 -3.27295810e-01] | [5.544480800628662, 7.858702182769775] |
8ca331f3-1268-403f-b126-3ec200a2f074 | progressive-learning-with-cross-window | 2211.12425 | null | https://arxiv.org/abs/2211.12425v2 | https://arxiv.org/pdf/2211.12425v2.pdf | Progressive Learning with Cross-Window Consistency for Semi-Supervised Semantic Segmentation | Semi-supervised semantic segmentation focuses on the exploration of a small amount of labeled data and a large amount of unlabeled data, which is more in line with the demands of real-world image understanding applications. However, it is still hindered by the inability to fully and effectively leverage unlabeled images. In this paper, we reveal that cross-window consistency (CWC) is helpful in comprehensively extracting auxiliary supervision from unlabeled data. Additionally, we propose a novel CWC-driven progressive learning framework to optimize the deep network by mining weak-to-strong constraints from massive unlabeled data. More specifically, this paper presents a biased cross-window consistency (BCC) loss with an importance factor, which helps the deep network explicitly constrain confidence maps from overlapping regions in different windows to maintain semantic consistency with larger contexts. In addition, we propose a dynamic pseudo-label memory bank (DPM) to provide high-consistency and high-reliability pseudo-labels to further optimize the network. Extensive experiments on three representative datasets of urban views, medical scenarios, and satellite scenes demonstrate our framework consistently outperforms the state-of-the-art methods with a large margin. Code will be available publicly. | ['Jiayi Ma', 'Yongjun Zhang', 'Yansheng Li', 'Bo Dang'] | 2022-11-22 | null | null | null | null | ['semi-supervised-semantic-segmentation'] | ['computer-vision'] | [ 3.79828632e-01 2.74658591e-01 -6.05483472e-01 -9.90795016e-01
-8.85673702e-01 -4.36086535e-01 3.31450552e-02 -1.46305025e-01
-6.47445142e-01 6.50251329e-01 -9.75859687e-02 -3.16918194e-01
-2.79711671e-02 -3.32237363e-01 -7.26877511e-01 -7.65344501e-01
1.47209555e-01 4.79404658e-01 4.42083299e-01 1.55721635e-01
-2.93100886e-02 2.28519991e-01 -1.26860201e+00 1.42792851e-01
1.11192083e+00 1.06948686e+00 5.75671196e-01 -6.57423064e-02
-1.39330756e-02 6.40796661e-01 -1.02806896e-01 -1.75690666e-01
3.18548918e-01 -1.36390746e-01 -8.18084300e-01 5.31156421e-01
2.56902009e-01 -4.32917714e-01 1.30967721e-01 1.26939404e+00
2.77558535e-01 9.03969109e-02 1.70087993e-01 -1.12946951e+00
-4.10732478e-01 5.86515903e-01 -9.55237150e-01 1.54170930e-01
-4.36311930e-01 2.07731441e-01 1.06603837e+00 -8.80771101e-01
7.60195971e-01 9.41513777e-01 3.64057869e-01 4.68731195e-01
-1.10600042e+00 -8.35772693e-01 6.77933633e-01 1.78729594e-01
-1.33454454e+00 -1.76005632e-01 8.92148435e-01 -1.58167183e-01
5.53923845e-01 1.20293885e-01 5.05664527e-01 8.84520948e-01
-3.08038265e-01 8.88411283e-01 9.61378992e-01 -3.31243485e-01
2.57005006e-01 2.20008090e-01 3.50968093e-01 9.25488234e-01
1.52342737e-01 8.02237261e-03 -5.54595530e-01 1.29696295e-01
5.44223905e-01 1.74163297e-01 -3.45029563e-01 -5.23666978e-01
-1.13172364e+00 7.59549141e-01 4.88369107e-01 1.17618449e-01
-1.27273619e-01 -3.23353857e-02 2.62902617e-01 -1.07251734e-01
5.78593969e-01 2.23984480e-01 -8.36694956e-01 1.38982937e-01
-9.31432903e-01 -2.16723680e-01 3.72797281e-01 1.06443644e+00
9.20046449e-01 -1.85151860e-01 -7.00877905e-02 9.34135675e-01
4.71919656e-01 4.05953914e-01 2.12977812e-01 -1.15955472e+00
6.60874188e-01 7.68777907e-01 -1.35753483e-01 -7.18234956e-01
-5.08771777e-01 -6.45551920e-01 -9.03991222e-01 9.16057080e-03
2.57062525e-01 -1.18432626e-01 -1.18058729e+00 1.96433389e+00
7.19704926e-01 3.90995860e-01 -2.63967514e-01 1.19817793e+00
5.02084553e-01 3.99493784e-01 2.96469890e-02 -2.76344031e-01
1.47645164e+00 -1.34479094e+00 -5.99409699e-01 -6.08292639e-01
5.49023986e-01 -3.37033600e-01 1.19421637e+00 3.04753333e-01
-8.01719248e-01 -4.46995616e-01 -1.06497335e+00 -2.14900151e-01
-1.90831088e-02 3.62865567e-01 7.14976192e-01 5.31667769e-01
-7.36433268e-01 3.32240909e-01 -1.14931786e+00 -2.68238056e-02
7.70828247e-01 3.54007393e-01 -3.22410941e-01 -3.12324971e-01
-1.05983114e+00 4.87151116e-01 5.90129972e-01 4.33834761e-01
-6.59510970e-01 -6.21553302e-01 -9.31250036e-01 -8.11196677e-03
9.93460417e-01 -3.27774704e-01 1.01990104e+00 -9.81664896e-01
-1.17723095e+00 9.89121258e-01 -1.50275797e-01 -3.13305557e-01
5.32750010e-01 -2.30388969e-01 -1.98766634e-01 4.24178183e-01
2.03604296e-01 1.11685777e+00 6.63051128e-01 -1.41715586e+00
-6.89612806e-01 -5.26950955e-01 -4.25725952e-02 2.48343095e-01
-4.78527337e-01 -3.07735234e-01 -1.14619923e+00 -6.80038154e-01
5.47995746e-01 -1.11087632e+00 -5.00361323e-01 2.92562723e-01
-6.04315639e-01 -5.99124543e-02 7.25557983e-01 -5.47287226e-01
1.12977576e+00 -2.10927844e+00 1.27494903e-02 3.76990497e-01
3.82206798e-01 2.61803418e-01 -1.66337669e-01 -2.28132561e-01
4.00393084e-02 8.39192569e-02 -6.56878591e-01 -6.14293218e-01
-1.61752641e-01 5.94267607e-01 -1.62460163e-01 4.46610570e-01
4.01519775e-01 9.60796475e-01 -8.83160055e-01 -9.40690398e-01
9.58837941e-02 2.00861692e-01 -5.99669039e-01 1.02714054e-01
-3.08588833e-01 7.11107314e-01 -7.02602684e-01 9.78947341e-01
8.12948823e-01 -9.13445175e-01 4.26517218e-01 -2.15024695e-01
2.40584999e-01 4.49773436e-03 -1.08085907e+00 2.00442290e+00
-1.71331346e-01 1.91128105e-01 2.09653437e-01 -1.17370188e+00
7.05519855e-01 8.56734533e-03 4.79573995e-01 -7.51327693e-01
9.81870592e-02 1.84995294e-01 -1.62973523e-01 -4.72913414e-01
3.38629603e-01 5.00561036e-02 1.13611378e-01 5.48915148e-01
-3.42840292e-02 1.80139095e-01 3.38877052e-01 3.17947865e-01
5.88442802e-01 1.99321166e-01 9.03669298e-02 -3.65668207e-01
4.59016055e-01 -8.57706591e-02 1.25229847e+00 6.61305189e-01
-4.48166102e-01 7.75365531e-01 5.32714307e-01 -3.08944315e-01
-9.02845979e-01 -8.25540185e-01 -1.94431469e-01 9.14497316e-01
5.27704298e-01 -1.34299114e-01 -7.96127617e-01 -1.09942520e+00
-2.39957079e-01 4.53760952e-01 -6.05509639e-01 -1.77925527e-02
-6.37105882e-01 -9.34506238e-01 1.99377224e-01 7.04231262e-01
5.49329698e-01 -9.43076968e-01 -4.89213586e-01 1.84353273e-02
-4.00593787e-01 -1.32831573e+00 -6.90637052e-01 5.11215389e-01
-9.37542200e-01 -1.26747978e+00 -4.43837702e-01 -8.86512458e-01
1.16648841e+00 4.83802944e-01 1.07297707e+00 3.19617510e-01
-2.19041675e-01 -1.91621929e-01 -3.64001632e-01 -7.92596787e-02
2.26399489e-02 3.37631941e-01 -3.01991433e-01 -1.21530093e-01
3.00371051e-01 -3.40916127e-01 -5.45053661e-01 5.11619627e-01
-1.08554626e+00 4.58949059e-01 5.81421793e-01 1.09142983e+00
1.00786710e+00 -1.15837649e-01 5.15323043e-01 -1.31786036e+00
-6.29873276e-02 -4.69272107e-01 -7.55320013e-01 4.64580238e-01
-1.00616753e+00 1.84383214e-01 3.81872088e-01 -4.63327080e-01
-1.08820510e+00 2.36669049e-01 -1.89118654e-01 -5.65494239e-01
-1.34929763e-02 4.61233228e-01 -3.24944824e-01 2.28585318e-01
1.48824722e-01 -1.21596549e-02 -6.56572208e-02 -4.53720987e-01
4.48767334e-01 4.29618835e-01 5.11681497e-01 -6.70787096e-01
5.80090582e-01 8.70275676e-01 -2.04320371e-01 -1.31701097e-01
-1.38482809e+00 -5.94862044e-01 -6.55475378e-01 -9.59241241e-02
8.25586021e-01 -9.97913837e-01 -4.26011682e-01 4.71516550e-01
-7.35033453e-01 -5.15397549e-01 -2.71231323e-01 4.00710016e-01
-2.07098648e-01 5.23362219e-01 -7.18287170e-01 -4.60378617e-01
-3.30944955e-01 -1.41901863e+00 1.17236054e+00 3.68874401e-01
2.29276329e-01 -7.48308659e-01 -3.06883663e-01 7.90010095e-01
2.10860688e-02 7.37885991e-03 7.84844279e-01 -6.55098259e-01
-8.20555806e-01 3.02365541e-01 -5.59410572e-01 5.04500985e-01
1.33237630e-01 -2.42305517e-01 -8.26090574e-01 -3.44314575e-01
8.55258405e-02 -6.59820974e-01 1.22417438e+00 4.53689575e-01
1.61568058e+00 -2.81447228e-02 -3.92551631e-01 9.36818421e-01
1.21967304e+00 -7.00096861e-02 2.43079945e-01 2.63612628e-01
8.80942643e-01 7.40611196e-01 1.02178967e+00 4.41189528e-01
6.02378011e-01 3.73199463e-01 6.27157748e-01 -4.16751742e-01
1.18085571e-01 -5.92536256e-02 -2.84643590e-01 7.34290421e-01
3.43869239e-01 -2.19072461e-01 -7.48493850e-01 3.58550817e-01
-2.02984214e+00 -5.05824208e-01 9.75102559e-02 1.80472481e+00
1.08417547e+00 4.13973778e-01 -2.74210364e-01 -1.10275947e-01
8.86689544e-01 3.36931705e-01 -1.11708164e+00 4.15814638e-01
-1.22236557e-01 -1.41009465e-01 6.64863825e-01 4.70076889e-01
-1.26307070e+00 1.04939079e+00 5.36549091e+00 9.50276375e-01
-1.05708086e+00 2.88804591e-01 1.24979901e+00 -1.70198768e-01
-4.77574885e-01 -1.07422564e-02 -7.77506888e-01 6.43625438e-01
3.65243763e-01 3.93848419e-01 2.56482750e-01 9.72879052e-01
1.32451847e-01 -2.63269901e-01 -9.71929431e-01 9.02925909e-01
-3.60054686e-03 -1.33086514e+00 -2.16429904e-01 5.30434698e-02
9.62901235e-01 2.33930975e-01 3.01620103e-02 7.56360497e-03
3.16372722e-01 -6.74833894e-01 8.18470180e-01 1.47224069e-01
1.00009334e+00 -7.16694176e-01 7.58051336e-01 4.33318049e-01
-1.19026148e+00 -1.43693000e-01 -3.24802011e-01 4.92916733e-01
3.62142891e-01 9.90317285e-01 -4.99547243e-01 4.23341662e-01
7.59676278e-01 8.16996634e-01 -5.49627483e-01 5.34907401e-01
-4.55889761e-01 6.48136079e-01 -3.15893143e-01 3.09330970e-01
3.99161369e-01 -3.76640975e-01 -2.50966251e-02 1.03693604e+00
-2.00044848e-02 3.06911528e-01 4.17287350e-01 9.05580282e-01
-2.42850572e-01 -2.17541963e-01 1.03208072e-01 7.92432576e-02
6.62954152e-01 1.50022328e+00 -1.18949997e+00 -4.11805093e-01
-3.60754251e-01 8.02906513e-01 6.14107490e-01 3.50938320e-01
-9.98535454e-01 1.87030435e-01 3.91328722e-01 -2.16471404e-01
3.48363608e-01 -1.71339095e-01 -4.74306911e-01 -1.25548816e+00
2.54541993e-01 -7.43798614e-01 4.68312114e-01 -5.49813390e-01
-1.32983994e+00 5.35790801e-01 -1.43807843e-01 -1.19401658e+00
1.09565817e-01 -3.87182087e-01 -2.94382930e-01 6.68936610e-01
-1.99021733e+00 -1.21510231e+00 -3.95866424e-01 5.14372170e-01
6.38762593e-01 8.85621980e-02 3.25621456e-01 4.99894559e-01
-7.96555042e-01 4.78125930e-01 -6.48823008e-03 2.22770289e-01
7.93300450e-01 -1.03967011e+00 3.13931525e-01 9.98187423e-01
1.26615673e-01 5.29097974e-01 2.37661362e-01 -6.16286516e-01
-9.64640379e-01 -1.19125915e+00 5.11001170e-01 -6.36590570e-02
4.14821893e-01 -2.88411766e-01 -1.13215768e+00 5.02819657e-01
-1.99793249e-01 4.73487079e-01 6.89786494e-01 2.63678897e-02
-4.49278891e-01 -2.33283564e-01 -1.04696381e+00 3.60318005e-01
1.18951440e+00 -3.38345379e-01 -3.00540656e-01 4.67064172e-01
1.14391720e+00 -5.11883497e-01 -4.12569255e-01 7.56455004e-01
3.86513203e-01 -1.04932976e+00 8.69164526e-01 -3.65851372e-01
3.50979537e-01 -3.89791638e-01 -2.31920672e-03 -7.99735606e-01
-4.76686843e-02 -2.19303265e-01 -1.61527097e-01 1.15833318e+00
4.40319866e-01 -5.33087134e-01 1.16179359e+00 8.34652007e-01
-3.20255667e-01 -1.17534542e+00 -8.44605684e-01 -5.08840263e-01
-2.85980612e-01 -5.22128999e-01 4.35707957e-01 1.09624076e+00
-3.61884326e-01 8.71878937e-02 -4.39460307e-01 3.94691080e-01
6.69087827e-01 3.43052655e-01 2.94388324e-01 -1.09649861e+00
-3.99653882e-01 -1.84519351e-01 1.83685794e-01 -1.24571550e+00
2.53695995e-01 -8.25516939e-01 2.69866735e-01 -1.39136350e+00
5.71811199e-01 -8.84798229e-01 -4.29868668e-01 6.93747818e-01
-6.32780194e-01 2.92630076e-01 1.42380586e-02 3.65873188e-01
-1.02924764e+00 5.78462243e-01 1.31294203e+00 -1.74339950e-01
-6.17937893e-02 -1.31301999e-01 -8.01558793e-01 8.07991028e-01
8.20211649e-01 -7.14222550e-01 -6.74156189e-01 -6.32678509e-01
2.40766093e-01 -1.26755387e-01 1.31046325e-01 -5.50055206e-01
2.54143804e-01 -3.73087019e-01 2.14811087e-01 -8.49541664e-01
1.33533105e-01 -8.80684674e-01 -2.16013119e-01 3.97443682e-01
-3.45432937e-01 -2.03238830e-01 -9.44448858e-02 7.07085669e-01
-2.38067523e-01 -2.53482133e-01 9.43445265e-01 -1.50192752e-01
-9.72071171e-01 5.62210739e-01 3.23474824e-01 3.06052268e-01
9.57444787e-01 -1.09155819e-01 -2.86926568e-01 4.00704518e-02
-5.81431866e-01 9.03690398e-01 5.46505272e-01 3.01209569e-01
5.46793163e-01 -1.02214038e+00 -3.95256877e-01 3.71284783e-01
1.91998363e-01 6.65232360e-01 5.67408800e-01 8.24437141e-01
-4.04327005e-01 2.32337505e-01 7.44398385e-02 -8.99924755e-01
-1.12622643e+00 5.82593381e-01 1.32582620e-01 -5.36639512e-01
-6.75273895e-01 1.02868295e+00 3.61960143e-01 -5.11539996e-01
4.59797919e-01 -4.13541049e-01 4.21459079e-02 -1.30015269e-01
4.32277799e-01 4.91273925e-02 -5.53607307e-02 -2.64414907e-01
-4.62618589e-01 4.75487649e-01 -2.60321021e-01 -1.14270337e-01
1.29096246e+00 -4.12015468e-01 -4.48717177e-02 2.28675470e-01
1.11307967e+00 -2.76361734e-01 -1.96747315e+00 -6.20597541e-01
1.02246501e-01 -3.61305624e-01 4.40850668e-02 -7.67252386e-01
-1.64312696e+00 7.17338681e-01 4.62239593e-01 -2.87784129e-01
1.34921837e+00 1.46011874e-01 8.69285166e-01 5.50448596e-01
3.85489792e-01 -1.55152762e+00 2.64777213e-01 2.23509341e-01
3.06922913e-01 -1.72215629e+00 9.60583314e-02 -6.26332521e-01
-9.58480597e-01 7.52714276e-01 8.63285542e-01 1.93885356e-01
7.48296440e-01 4.09310132e-01 1.68921158e-01 -1.27479136e-01
-7.52716482e-01 -2.10281223e-01 2.50546485e-01 2.66739368e-01
9.21003520e-02 -1.34674028e-01 -3.04851860e-01 6.59816146e-01
3.64251733e-01 -4.23830971e-02 1.70709804e-01 9.78486598e-01
-4.73288894e-01 -1.08159256e+00 -8.01125765e-02 4.29451346e-01
-4.17720437e-01 -1.61136463e-01 1.28632054e-01 6.72937989e-01
3.95776898e-01 9.62659538e-01 -2.30647177e-02 -1.35830745e-01
-1.71133652e-01 -2.14349210e-01 9.60813165e-02 -7.24974215e-01
-2.52155542e-01 4.00676101e-01 -9.97520089e-02 -6.59407079e-01
-7.14166462e-01 -4.88348126e-01 -1.61961782e+00 1.87824473e-01
-5.41755199e-01 -2.97206192e-04 6.37461662e-01 1.10680687e+00
3.54576945e-01 6.09997988e-01 6.03893459e-01 -5.71213365e-01
-5.24385512e-01 -8.39792550e-01 -7.07639635e-01 4.63836074e-01
2.54087269e-01 -6.03115380e-01 -2.40320474e-01 1.83916509e-01] | [9.519834518432617, 1.0385048389434814] |
4026c0e4-3735-4acb-a36d-b7a0fade09b9 | exploiting-more-information-in-sparse-point | 2210.00519 | null | https://arxiv.org/abs/2210.00519v1 | https://arxiv.org/pdf/2210.00519v1.pdf | Exploiting More Information in Sparse Point Cloud for 3D Single Object Tracking | 3D single object tracking is a key task in 3D computer vision. However, the sparsity of point clouds makes it difficult to compute the similarity and locate the object, posing big challenges to the 3D tracker. Previous works tried to solve the problem and improved the tracking performance in some common scenarios, but they usually failed in some extreme sparse scenarios, such as for tracking objects at long distances or partially occluded. To address the above problems, in this letter, we propose a sparse-to-dense and transformer-based framework for 3D single object tracking. First, we transform the 3D sparse points into 3D pillars and then compress them into 2D BEV features to have a dense representation. Then, we propose an attention-based encoder to achieve global similarity computation between template and search branches, which could alleviate the influence of sparsity. Meanwhile, the encoder applies the attention on multi-scale features to compensate for the lack of information caused by the sparsity of point cloud and the single scale of features. Finally, we use set-prediction to track the object through a two-stage decoder which also utilizes attention. Extensive experiments show that our method achieves very promising results on the KITTI and NuScenes datasets. | ['Zheng Fang', 'Zhiheng Li', 'Zuoxu Gu', 'Jiayao Shan', 'Yubo Cui'] | 2022-10-02 | null | null | null | null | ['3d-single-object-tracking'] | ['computer-vision'] | [-1.92415059e-01 -5.11426210e-01 -1.36593446e-01 -1.25651091e-01
-5.08754134e-01 -4.21305567e-01 4.83917505e-01 -2.53770381e-01
-1.10324211e-01 1.75397575e-01 1.28696501e-01 4.51578246e-03
-6.77379742e-02 -4.06386346e-01 -7.75681674e-01 -6.86114192e-01
1.62403226e-01 3.37410033e-01 5.19513905e-01 1.26580656e-01
1.59581006e-01 7.60541618e-01 -1.56349647e+00 -1.01735026e-01
6.96890295e-01 1.11200202e+00 6.17488146e-01 5.27575724e-02
-3.26972604e-01 3.47443253e-01 -4.90896732e-01 -1.98960766e-01
6.51220560e-01 -4.78188656e-02 9.83465090e-02 4.12886292e-01
6.98656499e-01 -4.64998990e-01 -6.36261761e-01 1.34874129e+00
6.06380582e-01 -7.25924894e-02 3.31912786e-01 -1.25964284e+00
-5.89923620e-01 1.62582561e-01 -9.64694619e-01 3.24442208e-01
1.75981864e-01 2.82541335e-01 6.83993876e-01 -9.87593114e-01
4.34371114e-01 1.54701519e+00 7.29060411e-01 4.30489182e-01
-6.29914522e-01 -8.46131206e-01 3.88588995e-01 1.62888467e-01
-1.46378398e+00 -3.09839904e-01 8.12289476e-01 -4.76648688e-01
6.85168922e-01 3.95310186e-02 8.52112949e-01 7.87037909e-01
4.06436473e-02 8.73044550e-01 7.40691960e-01 -4.04036837e-03
-2.46036440e-01 -1.42159671e-01 -1.12248659e-01 5.69354177e-01
5.46968997e-01 4.20360655e-01 -3.68679911e-01 -1.15607204e-02
9.25606430e-01 5.46816528e-01 -2.42043838e-01 -5.51833391e-01
-1.43085301e+00 5.87231576e-01 7.86951959e-01 3.22583735e-01
-3.46278250e-01 2.39406917e-02 2.34740794e-01 3.09536178e-02
4.05599833e-01 -1.42229140e-01 -2.55276620e-01 8.35735127e-02
-7.81276584e-01 3.06835175e-01 3.11234325e-01 1.40999854e+00
5.83095789e-01 1.60926059e-01 -3.40397298e-01 4.88500774e-01
7.12792993e-01 9.16709185e-01 2.72121131e-01 -6.18452370e-01
5.38275778e-01 7.26912737e-01 2.13165134e-01 -1.21914005e+00
-2.38571689e-01 -8.39437544e-01 -8.00287187e-01 1.67858616e-01
1.16803586e-01 1.76267371e-01 -9.45813000e-01 1.51526213e+00
6.67856634e-01 4.92005289e-01 -3.41876477e-01 1.51235795e+00
9.00979996e-01 6.74133718e-01 -1.42004341e-01 -2.36006439e-01
1.29758751e+00 -8.50468814e-01 -5.91403365e-01 -1.71524391e-01
4.17448729e-01 -8.62255573e-01 4.68272179e-01 -1.38772994e-01
-9.16628718e-01 -8.43185425e-01 -7.83876359e-01 -4.49263006e-02
9.55016091e-02 1.83538929e-01 4.41166729e-01 3.18764567e-01
-5.02084911e-01 1.69132903e-01 -9.74004805e-01 -1.87656611e-01
5.80316603e-01 3.71792793e-01 -1.23175129e-01 -2.37650424e-01
-7.67423749e-01 8.33021104e-01 3.14339668e-01 3.51891786e-01
-7.77543366e-01 -6.12294018e-01 -8.03341389e-01 1.43086776e-01
4.19898272e-01 -7.60537088e-01 9.50556099e-01 -2.52417147e-01
-1.08437788e+00 7.76420236e-01 -2.66655803e-01 -2.93522626e-01
3.84344608e-01 -4.06812370e-01 -1.91063404e-01 -3.50427717e-01
2.84359336e-01 5.92176259e-01 8.31426084e-01 -9.84975100e-01
-8.10416043e-01 -7.27024734e-01 -1.11162566e-01 3.91828179e-01
-1.68053105e-01 6.65312111e-02 -8.81917179e-01 -5.56610346e-01
5.55299938e-01 -9.53900218e-01 -2.78288871e-01 3.78875077e-01
-2.48191819e-01 -3.34472537e-01 1.09473443e+00 -3.71854246e-01
7.48136759e-01 -2.34996104e+00 2.39306360e-01 -7.12447613e-02
2.22759068e-01 3.34012240e-01 -5.58786392e-02 -9.46919546e-02
2.29508370e-01 -4.42958027e-01 1.58236384e-01 -3.69775981e-01
3.49507318e-03 1.21306963e-01 -3.11334878e-01 6.72834694e-01
1.81201354e-01 8.78987968e-01 -8.11316669e-01 -6.14062488e-01
4.23712403e-01 6.39653862e-01 -5.88431180e-01 2.03510195e-01
-3.01452458e-01 6.50328100e-01 -9.18043077e-01 8.15705359e-01
1.03843522e+00 -2.48543456e-01 -3.56053382e-01 -4.66208875e-01
-5.59859753e-01 1.04075082e-01 -1.22121549e+00 1.94856548e+00
-6.01673126e-02 3.70411307e-01 9.30549577e-02 -7.24961758e-01
1.21386778e+00 5.93316108e-02 7.20900118e-01 -5.67914188e-01
3.50868404e-01 2.52751201e-01 -1.01708658e-02 -2.16537878e-01
3.50574732e-01 -1.27194956e-01 3.95384356e-02 -8.46827403e-03
-1.34590060e-01 -1.65058836e-01 -2.04750195e-01 -5.14922179e-02
9.41291332e-01 2.27972493e-01 -7.61573948e-03 1.56734437e-01
6.35773361e-01 1.62502497e-01 9.14598882e-01 4.67848688e-01
-2.32187584e-01 5.40308535e-01 -7.68999085e-02 -3.90602052e-01
-9.05912697e-01 -7.76994050e-01 -5.63123636e-02 3.49802703e-01
6.13398731e-01 -3.36171836e-01 -1.90116778e-01 -5.33271551e-01
3.41270506e-01 1.50200874e-01 -8.52465034e-02 -3.59716117e-01
-6.14677966e-01 -3.12958330e-01 6.42084479e-02 4.36157733e-01
5.26638150e-01 -6.58333242e-01 -5.73893964e-01 1.62372261e-01
-1.70609821e-02 -1.35247207e+00 -7.26209164e-01 -4.88372752e-03
-1.03217638e+00 -8.80456030e-01 -8.16055655e-01 -8.65076721e-01
6.08187199e-01 9.80288208e-01 5.98767221e-01 2.10008696e-02
-3.94732468e-02 3.87443751e-02 -2.69529104e-01 -4.90904838e-01
2.04745129e-01 -1.36322200e-01 1.78308740e-01 -6.17375821e-02
6.03576303e-01 -2.59644359e-01 -4.79168326e-01 5.02655268e-01
-4.19477761e-01 -8.93718377e-02 6.69461668e-01 6.24687433e-01
8.26168716e-01 -4.44983272e-03 9.79752168e-02 -1.69412002e-01
6.09383844e-02 -3.27851415e-01 -1.12310469e+00 -6.30483925e-02
-8.89515579e-02 -1.45270392e-01 3.29470903e-01 -5.39606869e-01
-7.22074687e-01 5.59816301e-01 -1.79718331e-01 -1.41544080e+00
1.16458267e-01 1.96807578e-01 -3.51687819e-01 -4.72942859e-01
1.03237495e-01 3.61875117e-01 -1.53226489e-02 -7.67655075e-01
9.56780836e-02 5.40589273e-01 4.61688638e-01 -2.30727479e-01
1.39553285e+00 4.96525109e-01 1.27241254e-01 -5.32063127e-01
-1.13151991e+00 -5.37108362e-01 -4.75234807e-01 -2.03962997e-01
8.15554559e-01 -1.32447994e+00 -8.57415974e-01 3.28434289e-01
-1.30818605e+00 3.23164165e-01 -2.96801448e-01 8.07504833e-01
-3.36360306e-01 4.93313104e-01 -3.63514692e-01 -7.33586192e-01
-2.97591597e-01 -1.13550627e+00 1.37639832e+00 4.94337648e-01
4.31897581e-01 -4.86836076e-01 3.07603441e-02 9.92855057e-02
2.93538392e-01 2.63688881e-02 3.21304888e-01 -2.58352995e-01
-1.20674860e+00 -3.02789450e-01 -4.76777613e-01 -2.08840910e-02
9.31615382e-02 -4.21395868e-01 -5.87910295e-01 -3.91734123e-01
3.21447641e-01 -1.20116239e-02 6.90870404e-01 4.61141169e-01
1.06776631e+00 5.75059019e-02 -6.51990712e-01 9.07768369e-01
1.23129308e+00 1.63032845e-01 1.73291549e-01 1.68182001e-01
9.59474146e-01 6.29018471e-02 1.10360599e+00 5.68079293e-01
3.07725489e-01 9.97363210e-01 6.77255690e-01 1.50160536e-01
-2.81688303e-01 -4.07177955e-01 2.41077542e-01 1.00567210e+00
1.41029814e-02 1.02230899e-01 -6.33430839e-01 4.75210071e-01
-1.87389195e+00 -8.65411162e-01 -2.39052743e-01 2.08100104e+00
4.40644503e-01 3.50595891e-01 -6.22227415e-02 -2.44850799e-01
9.20325220e-01 2.36228690e-01 -6.49463713e-01 5.16071260e-01
-1.40619248e-01 -2.50711173e-01 4.52853650e-01 5.25157601e-02
-9.93627012e-01 8.95351052e-01 5.28057289e+00 7.38241792e-01
-1.22454536e+00 1.80079594e-01 -1.84638500e-01 -3.38883728e-01
-1.49114549e-01 4.48230170e-02 -1.41802943e+00 8.36879790e-01
3.15081626e-01 -2.09972695e-01 -1.66685004e-02 9.49956059e-01
7.35327899e-02 3.61592561e-01 -8.75577986e-01 1.35847604e+00
9.08816978e-02 -1.16709566e+00 -4.07968946e-02 1.27080068e-01
5.79982519e-01 3.54431570e-01 -2.11624173e-03 2.57821470e-01
-4.99578975e-02 -3.99979055e-01 8.85367095e-01 3.92701566e-01
5.09206891e-01 -4.69818741e-01 5.55869579e-01 7.22622275e-01
-1.50596011e+00 -4.00047116e-02 -9.31134105e-01 9.95624438e-02
2.95283467e-01 7.30938017e-01 -5.07066488e-01 5.89690804e-01
8.05088758e-01 1.13632810e+00 -3.08798134e-01 1.77330899e+00
-5.21617336e-03 1.26920208e-01 -6.99959517e-01 -7.09944069e-02
3.07640731e-01 -1.06092870e-01 9.57187772e-01 6.98419452e-01
7.90621102e-01 2.42707700e-01 4.91685569e-01 8.68810117e-01
2.18558554e-02 -2.34935790e-01 -7.30443656e-01 2.10865662e-01
5.94154000e-01 1.11453903e+00 -3.99454117e-01 -1.39115512e-01
-6.89678550e-01 6.79404080e-01 1.46888852e-01 -1.69976486e-03
-9.06888783e-01 -1.10461451e-01 6.60592973e-01 1.11500107e-01
8.18804204e-01 -5.59448421e-01 1.49261568e-05 -1.42441428e+00
2.48272866e-01 -6.10064983e-01 1.96377695e-01 -8.44127774e-01
-1.19987583e+00 4.43487197e-01 -2.45974883e-01 -1.82842171e+00
2.43219048e-01 -3.85276496e-01 -4.62473601e-01 7.98186421e-01
-1.67799747e+00 -1.11894608e+00 -5.78503311e-01 7.39128828e-01
5.55348635e-01 -1.81409284e-01 1.40105695e-01 7.21739888e-01
-3.71568054e-01 3.03393096e-01 -8.13783333e-03 -9.64789465e-03
5.98905861e-01 -6.22601449e-01 3.87367547e-01 6.94853365e-01
2.26505086e-01 4.88533825e-01 4.59125459e-01 -9.15451765e-01
-1.96519184e+00 -1.21204436e+00 8.32077384e-01 -3.51784050e-01
3.97965491e-01 -3.07734400e-01 -9.52075779e-01 6.37828112e-01
-2.52153695e-01 3.46554965e-01 1.15211904e-01 -6.20545112e-02
-1.22195721e-01 -1.11679807e-01 -8.63273263e-01 3.68266314e-01
1.54778123e+00 -1.85511187e-01 -7.97235727e-01 5.31000614e-01
9.01313365e-01 -9.86536920e-01 -7.44694769e-01 4.82101172e-01
2.43439242e-01 -6.53166890e-01 1.15932024e+00 -2.70724356e-01
3.04148649e-03 -8.95983219e-01 -1.44298419e-01 -1.04064083e+00
-6.45453811e-01 -3.20588142e-01 -4.26262528e-01 1.01299882e+00
-2.23304167e-01 -3.11290115e-01 9.58999574e-01 9.12992135e-02
-5.31817794e-01 -4.55783486e-01 -1.04221249e+00 -9.05143678e-01
-4.30829763e-01 -1.61160186e-01 7.56157517e-01 7.09052861e-01
-5.44525146e-01 3.62904400e-01 -5.19354939e-01 4.70361710e-01
1.00276244e+00 7.66833007e-01 9.17265654e-01 -1.39661038e+00
1.34204971e-02 -2.29469448e-01 -7.86908388e-01 -1.84778678e+00
5.52555807e-02 -9.69405890e-01 1.15457185e-01 -1.20815408e+00
2.27940008e-01 -7.70497978e-01 -1.32235005e-01 2.26568043e-01
-2.65348405e-01 1.39301911e-01 5.59251726e-01 5.51406205e-01
-8.27219784e-01 9.16564941e-01 1.59777582e+00 -2.77698755e-01
5.12713380e-02 3.35102677e-01 -5.11846721e-01 4.48384941e-01
2.17492133e-01 -6.52566969e-01 -4.13331836e-02 -9.47181463e-01
-1.50648788e-01 1.49459496e-01 6.39798522e-01 -1.22983086e+00
6.38028204e-01 -3.60949188e-02 6.19208038e-01 -1.45468438e+00
5.50927699e-01 -1.35975873e+00 2.10656196e-01 5.86017966e-01
1.79549992e-01 -8.56703222e-02 1.39832839e-01 7.33986259e-01
-3.29557240e-01 -9.93940830e-02 5.87670863e-01 -3.89368795e-02
-8.12555075e-01 8.90764654e-01 2.68556952e-01 -2.02330843e-01
1.08966577e+00 -3.06440532e-01 -1.00079909e-01 2.22778320e-02
-3.99769217e-01 7.05873549e-01 5.88340163e-01 6.62779272e-01
6.85524344e-01 -1.75887513e+00 -7.10715115e-01 3.72121751e-01
-4.51973118e-02 4.10556614e-01 3.63976508e-01 8.98974240e-01
-4.83691022e-02 5.94886780e-01 -3.98592740e-01 -1.22656584e+00
-1.17099023e+00 8.06434333e-01 2.46520728e-01 1.35406822e-01
-9.97640431e-01 6.93645120e-01 2.61751533e-01 -1.89396754e-01
4.86766070e-01 -2.89959460e-01 -1.62267182e-02 -1.96508780e-01
4.32921499e-01 -1.31605402e-01 -8.77575502e-02 -7.87245512e-01
-5.74100137e-01 1.22323358e+00 -6.56655431e-02 5.08086324e-01
1.31710255e+00 -1.60652831e-01 2.14694634e-01 1.19570881e-01
1.06179190e+00 -8.80020484e-02 -1.51857233e+00 -6.54874504e-01
-2.36794859e-01 -1.02407193e+00 7.59813339e-02 -6.69163242e-02
-1.35763621e+00 9.78245795e-01 6.44713283e-01 -8.02709833e-02
8.97410452e-01 9.89820659e-02 9.97341037e-01 3.17877859e-01
5.25500894e-01 -4.81909156e-01 -7.29331076e-02 6.87706590e-01
6.21982098e-01 -1.31884074e+00 1.66733071e-01 -6.49982452e-01
-3.76737028e-01 8.05759370e-01 7.43890822e-01 -2.35312492e-01
6.74137652e-01 1.64153531e-01 -2.36157104e-01 -2.66433626e-01
-6.43735409e-01 -4.80049670e-01 3.87511462e-01 5.96462071e-01
-8.52967612e-03 -3.78001958e-01 -3.24335019e-03 4.80154634e-01
8.32740366e-02 1.39129326e-01 -9.88745391e-02 9.66370463e-01
-6.47878408e-01 -7.99501121e-01 -5.64145863e-01 3.48838419e-01
-1.42533362e-01 2.15152070e-01 -8.06926042e-02 5.66948175e-01
2.62682140e-01 4.58375514e-01 1.15550853e-01 -4.73715752e-01
6.44163251e-01 -3.54409575e-01 6.10035360e-01 -5.83267212e-01
-5.01286685e-01 4.32906806e-01 -3.84573609e-01 -4.94797826e-01
-5.28410673e-01 -8.76979709e-01 -1.10578847e+00 -1.43843278e-01
-7.07240760e-01 6.37807399e-02 6.76291823e-01 7.68449962e-01
5.67115009e-01 4.28583384e-01 6.74761355e-01 -1.00123489e+00
-7.76210427e-01 -7.28892624e-01 -4.56015557e-01 3.20055544e-01
3.45063061e-01 -1.10191739e+00 -1.70020968e-01 -2.69817978e-01] | [6.626049041748047, -2.375286102294922] |
c239c50c-9cf7-4011-bda0-5499d0d925b1 | human-guided-ground-truth-generation-for | 2303.13069 | null | https://arxiv.org/abs/2303.13069v1 | https://arxiv.org/pdf/2303.13069v1.pdf | Human Guided Ground-truth Generation for Realistic Image Super-resolution | How to generate the ground-truth (GT) image is a critical issue for training realistic image super-resolution (Real-ISR) models. Existing methods mostly take a set of high-resolution (HR) images as GTs and apply various degradations to simulate their low-resolution (LR) counterparts. Though great progress has been achieved, such an LR-HR pair generation scheme has several limitations. First, the perceptual quality of HR images may not be high enough, limiting the quality of Real-ISR outputs. Second, existing schemes do not consider much human perception in GT generation, and the trained models tend to produce over-smoothed results or unpleasant artifacts. With the above considerations, we propose a human guided GT generation scheme. We first elaborately train multiple image enhancement models to improve the perceptual quality of HR images, and enable one LR image having multiple HR counterparts. Human subjects are then involved to annotate the high quality regions among the enhanced HR images as GTs, and label the regions with unpleasant artifacts as negative samples. A human guided GT image dataset with both positive and negative samples is then constructed, and a loss function is proposed to train the Real-ISR models. Experiments show that the Real-ISR models trained on our dataset can produce perceptually more realistic results with less artifacts. Dataset and codes can be found at https://github.com/ChrisDud0257/HGGT | ['Lei Zhang', 'Hui Zeng', 'Ming Liu', 'Xindong Zhang', 'Jie Liang', 'Du Chen'] | 2023-03-23 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Chen_Human_Guided_Ground-Truth_Generation_for_Realistic_Image_Super-Resolution_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Chen_Human_Guided_Ground-Truth_Generation_for_Realistic_Image_Super-Resolution_CVPR_2023_paper.pdf | cvpr-2023-1 | ['image-super-resolution', 'image-enhancement'] | ['computer-vision', 'computer-vision'] | [ 4.60705161e-01 1.49721906e-01 -1.93858948e-02 -1.84720531e-01
-1.06212795e+00 -9.48109031e-02 1.60512865e-01 -5.30165136e-01
-5.29134944e-02 7.99164712e-01 1.39694169e-01 -9.24568344e-03
3.21824461e-01 -8.37183654e-01 -5.46692550e-01 -8.23212206e-01
3.00391972e-01 -2.59243697e-01 2.30932489e-01 -4.00281101e-01
1.50498003e-01 2.32253194e-01 -1.58371127e+00 4.34729844e-01
1.23314977e+00 9.46148694e-01 6.06122136e-01 6.78965390e-01
3.04759592e-01 7.76012063e-01 -8.59929323e-01 -2.26688713e-01
4.13132548e-01 -7.17382073e-01 -5.62484682e-01 2.42117673e-01
3.16581398e-01 -5.98387480e-01 -3.57639074e-01 1.28637302e+00
8.20577562e-01 3.25487584e-01 3.28454077e-01 -1.15649736e+00
-1.35438168e+00 1.87680379e-01 -9.63763297e-01 1.26000851e-01
5.08676469e-01 3.25091392e-01 4.95355844e-01 -1.01664102e+00
4.55905259e-01 1.38513708e+00 2.94393212e-01 8.21405232e-01
-1.39566386e+00 -7.97927678e-01 -6.63199052e-02 1.17189541e-01
-1.46652710e+00 -4.23795909e-01 7.17170715e-01 6.85260221e-02
3.21885288e-01 4.57014710e-01 3.58202666e-01 1.28513765e+00
6.42199516e-02 4.71134871e-01 1.53565884e+00 -3.17440152e-01
1.11944318e-01 3.61913174e-01 -3.58349770e-01 3.74463111e-01
2.89402716e-02 5.01735866e-01 -3.19366664e-01 1.92735448e-01
1.21586025e+00 -1.25993177e-01 -7.65752852e-01 2.62234300e-01
-1.02236676e+00 4.58604187e-01 7.15148568e-01 3.49415570e-01
-2.67108411e-01 -2.70676702e-01 3.15909944e-02 2.88766056e-01
4.65974599e-01 4.22127604e-01 -1.77508295e-02 3.15795541e-01
-7.69984901e-01 5.37318923e-02 -2.20933594e-02 9.41054821e-01
7.96100557e-01 2.35861003e-01 -5.12091577e-01 1.23278701e+00
-1.29363060e-01 6.50816143e-01 4.94164109e-01 -1.15510559e+00
3.32778633e-01 2.43510649e-01 5.36571860e-01 -9.80088651e-01
3.53408158e-02 -4.01017278e-01 -1.19980741e+00 4.80215102e-01
6.74023032e-02 1.27321929e-01 -1.16932762e+00 1.57741725e+00
1.45431340e-01 -5.36558554e-02 7.91899338e-02 1.35329652e+00
9.96030927e-01 8.69156957e-01 7.83523321e-02 -4.45055336e-01
1.21742558e+00 -9.31377232e-01 -1.07286191e+00 -2.09771782e-01
-3.04749664e-02 -9.26996171e-01 1.61159933e+00 6.35715008e-01
-1.41953051e+00 -1.15932643e+00 -1.03141713e+00 -1.65748447e-01
-8.94124210e-02 3.91728759e-01 6.20307028e-02 5.11983037e-01
-1.25258577e+00 5.90365767e-01 -3.28248769e-01 -5.71736731e-02
2.70720959e-01 -1.99314952e-01 -1.80129841e-01 -4.00739044e-01
-1.42333198e+00 9.45278466e-01 3.14432740e-01 3.44940484e-01
-9.35487628e-01 -4.73474443e-01 -8.98092091e-01 -3.03971171e-01
2.36618251e-01 -4.81330067e-01 1.16604221e+00 -1.11758542e+00
-1.36942744e+00 9.30205882e-01 -1.19849123e-01 5.54615520e-02
6.02266133e-01 5.22185713e-02 -1.02317739e+00 2.10954472e-01
7.26630166e-02 6.89675331e-01 1.19660401e+00 -1.90203214e+00
-4.97757852e-01 -4.79570851e-02 -4.72157411e-02 3.98931026e-01
-1.73763946e-01 1.79540113e-01 -4.26043272e-01 -8.51034820e-01
1.59819126e-01 -5.73018312e-01 -3.27420980e-01 -2.73257717e-02
-5.40294290e-01 2.83881247e-01 6.40498877e-01 -9.39118266e-01
1.15904880e+00 -2.24677205e+00 -2.62439013e-01 -1.83386669e-01
1.27837569e-01 3.17261457e-01 -3.64625424e-01 -2.09728964e-02
-3.05795610e-01 2.65581459e-01 -1.13170207e-01 -2.47327805e-01
-2.67008722e-01 5.23054041e-02 -4.18347061e-01 1.30319417e-01
2.08194077e-01 6.91485524e-01 -9.58343387e-01 -5.53454816e-01
3.54020506e-01 7.95456529e-01 1.68980416e-02 6.13664865e-01
9.22161154e-03 7.77201474e-01 -3.78330618e-01 5.18896580e-01
1.04919004e+00 -1.86663792e-01 -1.93344668e-01 -7.18011320e-01
-1.90974697e-02 -2.06651464e-01 -1.14838457e+00 1.18181241e+00
-5.20305872e-01 2.69212872e-01 -1.00816943e-01 -4.06671047e-01
1.26235688e+00 3.38445991e-01 1.79477215e-01 -1.30259371e+00
7.75953382e-03 1.35897100e-01 -3.06539983e-01 -6.14826202e-01
7.94422209e-01 -3.69588524e-01 2.84826308e-01 2.42734224e-01
-3.90059084e-01 -1.91954687e-01 -4.69435751e-02 8.05151835e-02
6.08565390e-01 1.69505462e-01 7.70302564e-02 3.28691393e-01
5.02039731e-01 -3.36674005e-01 7.87601769e-01 5.21536529e-01
-2.08667114e-01 1.38271320e+00 -1.71886962e-02 -2.02591345e-01
-1.41461945e+00 -1.29639530e+00 -1.46626055e-01 7.79896557e-01
5.32550156e-01 4.36719954e-02 -7.01126754e-01 -3.26670825e-01
-6.69775724e-01 9.24521625e-01 -5.99565744e-01 -3.01300406e-01
-4.57238615e-01 -8.64409208e-01 2.68875301e-01 3.43579113e-01
9.09545302e-01 -1.43638456e+00 -2.82698959e-01 -6.19490556e-02
-7.61088252e-01 -1.11294901e+00 -5.63321412e-01 -4.08034861e-01
-6.71450198e-01 -8.37907910e-01 -1.12556982e+00 -6.16903007e-01
9.45811331e-01 5.32365859e-01 1.11078620e+00 1.45490631e-01
-3.46988797e-01 -1.57674089e-01 -5.19514680e-01 2.10008714e-02
-7.02603340e-01 -6.87609434e-01 -7.62872174e-02 1.55293986e-01
-2.01381683e-01 -2.46560872e-01 -9.94345427e-01 7.17018902e-01
-1.10959613e+00 4.39910978e-01 6.69396162e-01 7.83879936e-01
9.40568328e-01 5.96621215e-01 6.60924613e-01 -5.51019013e-01
5.76270282e-01 -6.43865541e-02 -3.45249683e-01 2.43595555e-01
-5.79311788e-01 -2.86482424e-01 8.96107733e-01 -6.18917346e-01
-1.65356660e+00 -3.58371258e-01 -1.85116604e-01 -6.81342661e-01
-3.37550342e-01 -6.00117780e-02 -3.10278356e-01 -8.13454464e-02
8.72590661e-01 4.18513149e-01 -2.77900457e-01 -4.16919231e-01
4.37300950e-01 7.13423193e-01 7.58471787e-01 -4.99875814e-01
9.57451165e-01 3.90008122e-01 -3.84724647e-01 -6.77327812e-01
-1.00638437e+00 -6.88015148e-02 -8.41624960e-02 -4.46910799e-01
8.30870211e-01 -1.01428604e+00 -1.65228069e-01 5.14569521e-01
-7.55096376e-01 -6.73014224e-01 -3.38980556e-01 2.94819385e-01
-6.47514403e-01 4.35821325e-01 -8.03680956e-01 -1.03420949e+00
-3.52394491e-01 -1.07053661e+00 1.09737051e+00 6.42010868e-01
2.17019215e-01 -4.77098793e-01 -2.83604503e-01 5.64815521e-01
6.06749952e-01 3.42034370e-01 6.05831146e-01 3.07442039e-01
-5.42033553e-01 6.16385899e-02 -5.21912336e-01 6.67600811e-01
3.26941907e-01 -1.42879337e-01 -1.08971059e+00 -4.50360507e-01
2.38904685e-01 -5.02687693e-01 5.47661602e-01 3.85334998e-01
1.53104973e+00 -3.24230641e-01 1.61597859e-02 5.59575975e-01
1.56170380e+00 2.19131574e-01 1.42342544e+00 2.31914714e-01
5.23223698e-01 5.56120038e-01 1.26935852e+00 1.61107406e-01
1.66508600e-01 7.75953829e-01 2.75459617e-01 -8.41722250e-01
-5.22380412e-01 -4.27085936e-01 5.38482070e-01 3.78216624e-01
-3.85372460e-01 -3.90361369e-01 -2.04695553e-01 4.21751231e-01
-1.47283661e+00 -1.07614458e+00 -2.26208657e-01 2.49393582e+00
1.00948048e+00 -9.26390514e-02 1.15990720e-03 1.86237976e-01
1.09416032e+00 1.40138626e-01 -5.60802519e-01 -2.37478405e-01
-4.20908630e-01 1.39994889e-01 2.64154226e-01 2.83809900e-01
-7.56263137e-01 7.78777897e-01 5.73236132e+00 9.74571288e-01
-1.11031640e+00 1.28490105e-01 1.21629977e+00 -1.28718438e-02
-2.94727921e-01 -2.93570042e-01 -4.51443315e-01 5.96198261e-01
7.12865055e-01 -1.14336379e-01 5.67370415e-01 5.57461202e-01
8.93806040e-01 -9.06330794e-02 -5.43063521e-01 1.13248348e+00
2.06462443e-02 -6.96876943e-01 6.55846149e-02 -1.39343679e-01
9.15535212e-01 -3.99052918e-01 5.39623320e-01 8.74675810e-02
3.42307240e-01 -1.21402693e+00 5.69617033e-01 5.62366843e-01
1.43358088e+00 -6.26828969e-01 5.93754172e-01 1.60902172e-01
-1.14015496e+00 -7.02359080e-02 -6.87138855e-01 4.71592069e-01
3.58992755e-01 6.93847954e-01 -2.72046685e-01 8.30723882e-01
1.05051506e+00 4.36535597e-01 -7.68692076e-01 8.80116165e-01
-4.91473496e-01 3.69877756e-01 1.80762112e-01 7.41243124e-01
-3.68567288e-01 -3.20460409e-01 4.79487956e-01 8.57060611e-01
6.03639781e-01 4.78615135e-01 1.04934812e-01 1.23470628e+00
1.28880098e-01 5.77762462e-02 -3.30638289e-01 3.01678985e-01
3.63555640e-01 1.60172272e+00 -4.71723229e-01 -4.00219828e-01
-2.92137027e-01 1.35408509e+00 -1.09865427e-01 7.95157611e-01
-1.03005862e+00 -3.74594867e-01 1.96047127e-01 3.17703605e-01
4.14371025e-03 2.92917103e-01 -1.90277234e-01 -1.22664261e+00
1.77191403e-02 -1.24862576e+00 4.19454187e-01 -1.70721555e+00
-1.37430024e+00 1.03664589e+00 -2.94886619e-01 -1.75109494e+00
8.58779475e-02 -4.82523777e-02 -6.19463086e-01 1.16326308e+00
-1.52272952e+00 -1.01564610e+00 -8.85907590e-01 7.28137851e-01
5.32723844e-01 2.95335382e-01 4.24423128e-01 3.53787810e-01
-5.60033381e-01 6.78533018e-01 -2.70672679e-01 -6.79767728e-02
9.24805701e-01 -1.12501001e+00 1.29752159e-01 1.10183513e+00
-4.04195070e-01 4.77808297e-01 9.48513627e-01 -6.05205119e-01
-7.87944257e-01 -1.48750067e+00 3.90997529e-01 -7.97415674e-02
8.66376981e-02 -7.54926503e-02 -1.28443801e+00 3.61930817e-01
1.28193662e-01 1.46397442e-01 6.65673688e-02 -4.89143342e-01
-1.89701021e-01 -1.52194783e-01 -1.55915141e+00 8.26781809e-01
9.83246982e-01 -3.38680983e-01 -3.67545187e-01 8.95497948e-02
8.90363395e-01 -4.57688302e-01 -1.07480490e+00 6.05269790e-01
2.03022927e-01 -1.19494963e+00 1.17299271e+00 -1.31454796e-03
6.82355464e-01 -8.45383286e-01 5.02428785e-02 -1.44497371e+00
-3.69935870e-01 -4.39676583e-01 1.58627287e-01 1.26895726e+00
2.54765421e-01 -5.08138299e-01 4.63234186e-01 5.58731377e-01
-6.75004199e-02 -4.95904326e-01 -2.49926299e-01 -9.38275039e-01
-2.56799072e-01 -6.71417452e-03 5.34901977e-01 9.71901715e-01
-3.56970251e-01 1.55524284e-01 -6.63505971e-01 4.42291051e-01
9.49092627e-01 2.18172520e-01 6.71893299e-01 -5.85807800e-01
-2.56223261e-01 -1.49266198e-01 4.99378145e-02 -7.39500284e-01
-4.23898846e-01 -3.82284999e-01 2.84336567e-01 -1.57917798e+00
3.34533870e-01 -4.87500519e-01 -4.02788669e-01 4.18449700e-01
-5.12186229e-01 7.48460174e-01 -3.91862877e-02 2.09131896e-01
-4.52752590e-01 7.74945855e-01 1.92749023e+00 1.99866906e-01
-1.75568447e-01 -1.51906282e-01 -7.69291222e-01 4.16244864e-01
7.87434697e-01 -7.73754418e-02 -5.27667761e-01 -8.63574892e-02
-8.27951953e-02 3.87816429e-01 6.41301811e-01 -9.22762096e-01
-2.76337028e-01 -3.34917903e-01 7.71431923e-01 -5.07477522e-01
4.32410359e-01 -5.25653064e-01 5.36522090e-01 2.19908684e-01
-4.40568089e-01 -1.94210872e-01 -6.51470348e-02 2.67507821e-01
-2.91089386e-01 7.60178939e-02 1.35905910e+00 -3.18803430e-01
-6.21769905e-01 4.99230713e-01 1.22725718e-01 -8.53689164e-02
6.81478202e-01 -4.54748958e-01 -5.95961630e-01 -5.90619445e-01
-6.74549878e-01 4.73762900e-02 8.42421889e-01 5.22648633e-01
1.04810750e+00 -1.38591921e+00 -8.76725137e-01 1.20157704e-01
2.19647005e-01 -2.32557617e-02 9.23544288e-01 6.56316817e-01
-1.88812047e-01 -2.04036728e-01 -3.55050713e-01 -2.42527157e-01
-1.26252353e+00 9.13141191e-01 4.94522780e-01 -1.37319401e-01
-8.57920587e-01 5.08955836e-01 6.42383754e-01 -1.97487175e-01
-3.31656216e-03 5.97220324e-02 -3.02320272e-01 -5.20993888e-01
8.92576456e-01 3.57689470e-01 -1.26464620e-01 -6.71226919e-01
1.32321149e-01 5.07926166e-01 -2.48574093e-02 -1.99502125e-01
1.16464996e+00 -5.71411133e-01 1.29564181e-01 5.97132929e-02
9.31932330e-01 -8.50563571e-02 -1.37475789e+00 -2.29571894e-01
-6.62547886e-01 -9.94872510e-01 1.01599716e-01 -9.33028102e-01
-1.41626072e+00 7.26959825e-01 9.58063245e-01 2.31396362e-01
1.83976197e+00 -1.51520506e-01 8.63271892e-01 -3.33073378e-01
4.35251206e-01 -1.06784427e+00 4.72680807e-01 -1.82013184e-01
1.22360861e+00 -1.28826833e+00 -1.15557514e-01 -5.96423745e-01
-9.83972073e-01 7.95791864e-01 9.76766706e-01 4.30055410e-02
-1.19097255e-01 1.28374085e-01 3.38250160e-01 5.68880886e-02
-5.44494808e-01 -3.45560938e-01 1.00287043e-01 8.59868467e-01
2.92302310e-01 -8.95610973e-02 -2.38600850e-01 3.75883043e-01
-1.87880203e-01 1.61236584e-01 7.64057517e-01 2.45520324e-01
-3.49762827e-01 -9.23232973e-01 -7.37739921e-01 2.39806682e-01
-5.36551356e-01 -6.77185580e-02 1.09744497e-01 4.67892557e-01
1.64184600e-01 1.28130412e+00 -2.43584245e-01 -5.16740620e-01
4.77786630e-01 -5.91801763e-01 4.22178864e-01 -3.40879589e-01
-1.50751114e-01 1.62305817e-01 -7.57220536e-02 -6.80567980e-01
-2.30329514e-01 -2.23975107e-01 -1.09743595e+00 -2.83200085e-01
-2.74055660e-01 -7.80049041e-02 2.67850369e-01 3.78836393e-01
1.66979596e-01 6.00163639e-01 1.04257071e+00 -8.63437235e-01
-2.95048922e-01 -9.48990107e-01 -7.36549139e-01 8.17600548e-01
3.39235663e-01 -4.60959852e-01 -5.63457727e-01 2.61900276e-01] | [11.12280559539795, -2.0237152576446533] |
23f2f05c-a33a-48b3-b4bd-8b99f34912b3 | knowledge-acquisition-strategies-for-goal | null | null | https://aclanthology.org/W14-4326 | https://aclanthology.org/W14-4326.pdf | Knowledge Acquisition Strategies for Goal-Oriented Dialog Systems | null | ['er', 'Aasish Pappu', 'Alex Rudnicky'] | 2014-06-01 | null | null | null | ws-2014-6 | ['goal-oriented-dialog'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.505391597747803, 3.6524617671966553] |
57a63624-9f68-40c7-8462-5b79aefb36f7 | global-local-context-network-for-person | 2112.02500 | null | https://arxiv.org/abs/2112.02500v4 | https://arxiv.org/pdf/2112.02500v4.pdf | MovieNet-PS: A Large-Scale Person Search Dataset in the Wild | Person search aims to jointly localize and identify a query person from natural, uncropped images, which has been actively studied over the past few years. In this paper, we delve into the rich context information globally and locally surrounding the target person, which we refer to as scene and group context, respectively. Unlike previous works that treat the two types of context individually, we exploit them in a unified global-local context network (GLCNet) with the intuitive aim of feature enhancement. Specifically, re-ID embeddings and context features are simultaneously learned in a multi-stage fashion, ultimately leading to enhanced, discriminative features for person search. We conduct the experiments on two person search benchmarks (i.e., CUHK-SYSU and PRW) as well as extend our approach to a more challenging setting (i.e., character search on MovieNet). Extensive experimental results demonstrate the consistent improvement of the proposed GLCNet over the state-of-the-art methods on all three datasets. Our source codes, pre-trained models, and the new dataset are publicly available at: https://github.com/ZhengPeng7/GLCNet. | ['Bingbing Ni', 'Rong Quan', 'Peng Zheng', 'Jie Qin', 'Xiaogang Cheng', 'Yichao Yan'] | 2021-12-05 | null | null | null | null | ['person-search'] | ['computer-vision'] | [-3.75318862e-02 -7.05033422e-01 -9.49037820e-02 -4.08252686e-01
-5.66224277e-01 -4.00374621e-01 7.54281104e-01 4.26396020e-02
-8.05826664e-01 4.77340579e-01 6.24382973e-01 2.89242774e-01
-1.69351101e-01 -4.55030650e-01 -2.30923027e-01 -5.26877284e-01
1.87103912e-01 4.11114603e-01 6.93695024e-02 4.66027968e-02
1.86331928e-01 9.17157754e-02 -1.37695062e+00 -2.48803303e-01
7.57739902e-01 8.75765324e-01 1.65206715e-01 4.45070058e-01
4.37822580e-01 4.18647259e-01 -3.70973408e-01 -6.23484433e-01
2.00640842e-01 -3.07450891e-01 -8.20809722e-01 2.22016513e-01
6.99962974e-01 -4.14390653e-01 -9.36026216e-01 1.08484793e+00
1.00349844e+00 4.82681215e-01 4.09422040e-01 -1.01694536e+00
-1.11014187e+00 9.72336158e-02 -6.59462690e-01 5.06166101e-01
7.31692970e-01 4.48652625e-01 1.26205409e+00 -1.04710472e+00
4.38993275e-01 1.43764246e+00 2.99202591e-01 5.65926313e-01
-1.15778184e+00 -6.14734054e-01 4.48201239e-01 6.15067720e-01
-1.77258778e+00 -3.11195433e-01 6.78222835e-01 -2.70029396e-01
5.21218479e-01 2.78735429e-01 6.48426056e-01 1.36795151e+00
-3.71067911e-01 1.17408907e+00 1.02643609e+00 -4.39135522e-01
-1.20272964e-01 4.03122567e-02 3.62992078e-01 6.42956913e-01
1.88685372e-01 2.32218534e-01 -7.10554063e-01 -2.62740999e-01
8.23844790e-01 2.57947206e-01 -5.04225791e-01 -2.15399787e-01
-1.34263170e+00 7.91750193e-01 5.77485561e-01 1.53376892e-01
-3.44410956e-01 2.20053136e-01 4.46866065e-01 -3.53997238e-02
2.66829878e-01 7.01415092e-02 1.00921337e-02 -1.79036632e-01
-6.12488568e-01 5.80942929e-01 6.08923376e-01 1.12124312e+00
5.89016497e-01 -5.01206756e-01 -6.76745772e-01 1.12368917e+00
2.48341799e-01 4.24800307e-01 4.51530546e-01 -5.66764534e-01
5.19208372e-01 4.96249527e-01 3.00924242e-01 -1.08390701e+00
-7.04759583e-02 -7.29652762e-01 -6.67201638e-01 -6.24075234e-01
3.93611521e-01 2.78187543e-02 -9.07550871e-01 1.90495110e+00
3.97142082e-01 4.53438282e-01 -1.92986310e-01 1.46811652e+00
9.85985100e-01 4.51413929e-01 1.23696908e-01 5.18738031e-01
1.68926787e+00 -1.63533199e+00 -4.18752462e-01 -4.29364234e-01
1.77381076e-02 -6.53153539e-01 1.29058218e+00 8.53311792e-02
-9.53022778e-01 -8.15727353e-01 -7.23211288e-01 -3.27405900e-01
-4.30404156e-01 3.62272769e-01 6.31135643e-01 3.40502262e-01
-9.65853989e-01 2.30934232e-01 -5.29437661e-01 -7.66628802e-01
4.64854866e-01 2.19135880e-01 -3.40874195e-01 -5.42963147e-01
-1.33497083e+00 4.85421002e-01 1.01027735e-01 2.40345329e-01
-8.10049593e-01 -2.50723273e-01 -6.31353498e-01 1.43700555e-01
6.43811882e-01 -8.45426977e-01 1.09723353e+00 -4.07726824e-01
-9.99926507e-01 9.60798979e-01 -4.96313632e-01 -6.91562667e-02
4.62006301e-01 -5.79179466e-01 -3.85398597e-01 2.45113030e-01
3.99804741e-01 6.25327885e-01 5.26553571e-01 -1.07689083e+00
-7.33780980e-01 -3.95537972e-01 2.11303309e-01 3.18474978e-01
-6.67207122e-01 4.09182727e-01 -1.52191770e+00 -8.55024278e-01
-3.17785561e-01 -9.66185510e-01 -1.95010141e-01 -1.20494477e-02
-5.20819962e-01 -6.73573256e-01 4.53542173e-01 -7.31195450e-01
1.40253365e+00 -1.98505521e+00 3.88236463e-01 9.87432972e-02
2.44814590e-01 2.68422961e-01 -1.86580122e-01 4.99940604e-01
1.88949913e-01 1.39136368e-03 -7.72552192e-02 -7.50882864e-01
3.25794935e-01 -2.91708224e-02 1.74984619e-01 5.85048676e-01
2.04059295e-02 1.18164718e+00 -1.08740079e+00 -5.87633729e-01
6.13387078e-02 4.75032687e-01 -3.97779435e-01 7.03512952e-02
1.64458752e-01 6.33895934e-01 -7.08553553e-01 9.64121222e-01
4.15172666e-01 -4.89054978e-01 -1.85791016e-01 -1.61822885e-02
9.60170999e-02 -3.35510224e-02 -1.09927332e+00 1.90121651e+00
-1.67675808e-01 4.69987452e-01 -1.53457850e-01 -6.54456139e-01
6.95057452e-01 1.51325643e-01 1.23890139e-01 -8.34508955e-01
1.04193082e-02 5.41263707e-02 -2.80903876e-01 -4.62557197e-01
6.91828072e-01 5.00346780e-01 -3.15562576e-01 3.87592673e-01
7.82156587e-02 7.93686271e-01 3.03365082e-01 3.24372083e-01
8.35181832e-01 -2.21656952e-02 2.81286240e-01 -2.22630069e-01
8.65805268e-01 -2.94887781e-01 6.58900797e-01 9.55372870e-01
-5.68174541e-01 5.71664214e-01 1.31625116e-01 -3.44008833e-01
-8.31407130e-01 -9.72874403e-01 8.89271125e-03 1.29640567e+00
7.27965057e-01 -7.23286688e-01 -5.45284212e-01 -6.03954792e-01
1.18453719e-01 1.67775899e-01 -6.64941669e-01 3.50860395e-02
-6.14933491e-01 -5.04823267e-01 4.09486115e-01 5.72227657e-01
8.95278811e-01 -1.02034020e+00 -3.28977376e-01 5.95432054e-03
-1.58723608e-01 -1.23541081e+00 -1.30731308e+00 -5.13850093e-01
-2.95052230e-01 -9.77142096e-01 -1.11438191e+00 -1.09910142e+00
5.05621552e-01 3.68498027e-01 1.07601821e+00 2.93754131e-01
-5.72628081e-01 8.51928174e-01 -4.91513103e-01 2.49189418e-03
6.31396472e-01 1.08209521e-01 1.47864431e-01 3.06950122e-01
8.06242228e-01 -3.78965199e-01 -1.17037630e+00 6.00875199e-01
-6.32105112e-01 -1.11209452e-01 5.33187211e-01 1.04355597e+00
7.65772164e-01 -1.04609497e-01 4.33821678e-01 -4.82806623e-01
7.89557040e-01 -5.82873702e-01 -3.25711608e-01 5.34785807e-01
-4.85847354e-01 -2.29468390e-01 2.90236831e-01 -4.49371725e-01
-9.37253952e-01 -1.50857940e-01 5.68793677e-02 -4.16499317e-01
-2.62981862e-01 4.81984138e-01 -2.65462786e-01 -1.24828324e-01
2.42463648e-01 5.32441795e-01 -5.84843338e-01 -6.99626803e-01
3.19028169e-01 6.52513027e-01 7.67239273e-01 -7.59393156e-01
8.67564142e-01 4.14287359e-01 -3.43062341e-01 -6.36070609e-01
-7.27945805e-01 -9.40069079e-01 -4.70831484e-01 -1.66223571e-01
8.75212610e-01 -1.08340561e+00 -6.65155292e-01 5.72375000e-01
-9.38589215e-01 -2.69730259e-02 1.08236335e-01 4.55962718e-01
-1.23374894e-01 6.36034548e-01 -5.30312181e-01 -7.30409086e-01
-5.13656914e-01 -1.14419520e+00 1.20882392e+00 8.17566335e-01
-7.36791492e-02 -9.57218826e-01 9.30779241e-03 4.61812466e-01
2.43994698e-01 1.43404365e-01 4.44932640e-01 -7.01804578e-01
-8.14019978e-01 -4.39091504e-01 -5.11635184e-01 3.54267433e-02
1.78166047e-01 -6.73997998e-01 -8.05406213e-01 -5.72390318e-01
-3.68719220e-01 -4.60392654e-01 1.20427287e+00 6.30754083e-02
1.24810767e+00 -1.41428590e-01 -6.44200683e-01 5.99052668e-01
1.32792950e+00 -1.91791520e-01 3.59565467e-01 1.94593295e-01
7.02997446e-01 5.12444258e-01 7.41710007e-01 4.77299184e-01
6.03988588e-01 1.15520489e+00 4.19069966e-03 9.34030116e-03
-2.21205667e-01 -4.84858543e-01 -1.48038253e-01 3.44356418e-01
-2.23311886e-01 -4.59602565e-01 -8.73026252e-01 8.52753818e-01
-2.18676233e+00 -9.36805785e-01 3.37879866e-01 1.96989882e+00
7.87226319e-01 -1.66728362e-01 3.39708894e-01 -3.37353289e-01
1.07972538e+00 3.68881762e-01 -6.54516876e-01 2.67233193e-01
-1.11618780e-01 9.63975117e-02 2.75327176e-01 4.29543108e-01
-1.41294646e+00 1.12188315e+00 4.88520861e+00 1.01580954e+00
-6.75168753e-01 3.40361595e-01 4.73168612e-01 -3.18733066e-01
5.71172386e-02 -5.41177904e-03 -1.04508603e+00 7.56181121e-01
2.09582239e-01 -2.14444652e-01 5.68421483e-01 8.45361531e-01
1.39392197e-01 -2.83746496e-02 -1.21438885e+00 1.48190904e+00
3.54937464e-01 -9.66844201e-01 -1.60269871e-01 7.40662143e-02
6.71203494e-01 -8.36069509e-02 1.72352821e-01 3.86651933e-01
1.32691473e-01 -1.08623588e+00 6.21038854e-01 7.47485280e-01
7.37321973e-01 -6.83405101e-01 6.03680074e-01 1.51502952e-01
-1.62877166e+00 -1.77675381e-01 -2.97713161e-01 2.22862571e-01
3.67713779e-01 2.36759648e-01 -1.24904223e-01 5.84546804e-01
1.01924551e+00 9.60203469e-01 -1.01725125e+00 1.41124940e+00
-3.41623366e-01 4.46984380e-01 -3.20079774e-01 -2.00369686e-01
2.46389821e-01 -1.70459673e-01 6.88803673e-01 1.28173065e+00
-6.04289770e-02 8.18542838e-02 5.34159362e-01 1.03396904e+00
-1.40308842e-01 6.36905655e-02 -4.21618782e-02 7.83751830e-02
5.86910367e-01 1.28609693e+00 -3.01423907e-01 -3.24647188e-01
-6.41162932e-01 1.42029512e+00 4.22895789e-01 6.33796215e-01
-7.80239522e-01 -3.25498372e-01 7.14966655e-01 -2.06894681e-01
4.25618857e-01 -1.66776121e-01 2.84536421e-01 -1.49382317e+00
3.85376126e-01 -7.86092103e-01 5.63671768e-01 -4.77132857e-01
-1.76483333e+00 5.74310422e-01 8.23293254e-02 -1.09944320e+00
-1.21712657e-02 -3.49625379e-01 -9.17766035e-01 1.25152659e+00
-1.70469439e+00 -1.54304993e+00 -6.63811982e-01 8.65389109e-01
6.28516376e-01 -3.74084592e-01 5.47480345e-01 5.32601297e-01
-7.72144318e-01 1.08566856e+00 -6.59611002e-02 4.73749489e-01
9.35612440e-01 -1.04967332e+00 5.02462983e-01 8.95643055e-01
1.45789295e-01 1.03274250e+00 3.42598110e-01 -6.42775834e-01
-1.28391433e+00 -1.01010299e+00 1.02232492e+00 -4.77663130e-01
5.58829725e-01 -5.71084738e-01 -5.30942798e-01 5.46170712e-01
3.29257846e-01 1.47029936e-01 7.26437151e-01 2.82943785e-01
-2.76031256e-01 9.45708379e-02 -8.31915557e-01 8.64837885e-01
1.57915306e+00 -5.85618615e-01 -4.29446638e-01 3.90572369e-01
6.00756645e-01 -2.89002031e-01 -6.93524122e-01 1.05006918e-01
6.34900630e-01 -8.39465618e-01 1.23636496e+00 -5.52240312e-01
1.85536623e-01 -3.09349656e-01 -2.76560396e-01 -9.76857424e-01
-5.81845760e-01 -5.43904960e-01 -1.63637906e-01 1.39790130e+00
5.69564328e-02 -6.45977855e-01 7.80863047e-01 5.48360884e-01
1.94416657e-01 -1.04529154e+00 -7.49359906e-01 -8.60250831e-01
-1.78538069e-01 2.02262234e-02 6.73491776e-01 5.80427408e-01
-3.19208145e-01 3.48774880e-01 -7.17747450e-01 2.28891417e-01
7.79246509e-01 3.60658169e-01 7.97991097e-01 -1.07011974e+00
-5.84869027e-01 -4.03221607e-01 -4.88170534e-01 -1.71607924e+00
1.37978066e-02 -8.15785348e-01 -1.11273855e-01 -1.44389439e+00
7.24977732e-01 -4.49141294e-01 -5.14657199e-01 2.37660527e-01
-7.63070285e-01 2.17622906e-01 2.60699570e-01 5.26710749e-01
-9.88711953e-01 8.19517791e-01 1.04272485e+00 -1.97262451e-01
-7.97218550e-03 -1.70867573e-02 -9.39151347e-01 3.95223200e-01
6.36709750e-01 -9.83312055e-02 -3.16100746e-01 -7.13259995e-01
-2.38753229e-01 -3.51169109e-01 9.72164273e-01 -9.43652391e-01
6.11083508e-01 3.77473980e-02 5.75968862e-01 -5.32855392e-01
6.43783092e-01 -5.58301806e-01 -3.24211478e-01 7.31363147e-02
-4.01399881e-01 9.01184306e-02 -1.36650831e-01 1.03864181e+00
-4.04594421e-01 -1.36464179e-01 3.51448745e-01 -1.74699858e-01
-1.09564662e+00 7.58055687e-01 3.33579212e-01 7.74973780e-02
8.09608400e-01 -6.94781318e-02 -4.15108949e-01 -6.04854107e-01
-6.97747290e-01 8.23981643e-01 4.83298093e-01 6.04100049e-01
7.10331321e-01 -1.60200346e+00 -7.23407149e-01 -9.79778320e-02
4.45180923e-01 -1.75079629e-01 4.50684011e-01 7.43801653e-01
-7.29846284e-02 7.21287251e-01 2.33410954e-01 -5.03577054e-01
-1.36780751e+00 6.37411594e-01 2.33593121e-01 -3.17791104e-01
-5.18415630e-01 1.12383509e+00 3.26524258e-01 -2.21094102e-01
5.78261554e-01 2.12920472e-01 -2.93387771e-01 -2.15739504e-01
6.21526003e-01 1.51125908e-01 -6.79277420e-01 -8.26446414e-01
-5.52424312e-01 7.94084013e-01 -3.15769583e-01 -1.48143424e-02
1.16632628e+00 -4.43764657e-01 7.86431208e-02 -9.91853625e-02
1.18673944e+00 5.32798767e-02 -1.20247757e+00 -8.93612921e-01
-3.45219858e-02 -9.66288805e-01 -1.54117391e-01 -7.29773343e-01
-1.09283340e+00 6.38280153e-01 6.59795702e-01 -2.43892312e-01
1.13300538e+00 5.11977971e-01 9.73568082e-01 3.04700464e-01
3.32794219e-01 -1.11111760e+00 3.13364059e-01 2.08147749e-01
9.73891377e-01 -1.41084015e+00 1.57209486e-01 -3.94610316e-01
-6.14558995e-01 6.66338563e-01 8.30521405e-01 -1.64514288e-01
4.96965140e-01 -5.60244441e-01 -4.91182685e-01 -1.72657400e-01
-5.38464367e-01 -6.24045491e-01 4.06786263e-01 4.52876091e-01
8.64075348e-02 -1.47456275e-02 -2.32691020e-01 1.03109264e+00
7.40687456e-03 4.38644253e-02 -8.58128667e-02 1.06708312e+00
-1.29067734e-01 -1.26497519e+00 -4.08326477e-01 3.14887613e-01
-2.45672658e-01 -2.22157687e-01 -6.11687124e-01 7.21996903e-01
2.56928325e-01 1.06968117e+00 -1.72993615e-01 -4.96125132e-01
3.74109834e-01 -1.03724502e-01 3.58743727e-01 -6.16906583e-01
-3.95466238e-01 2.19172016e-02 1.64536424e-02 -5.11393249e-01
-3.95631641e-01 -8.44065845e-01 -5.82297504e-01 -2.48860925e-01
-2.41092876e-01 5.68351336e-02 1.78995803e-01 7.75722623e-01
4.26958114e-01 1.12512037e-01 5.02129495e-01 -7.02465832e-01
-3.80557865e-01 -1.09505510e+00 -3.73221815e-01 6.16476178e-01
1.70381084e-01 -8.39263260e-01 -1.47807240e-01 -1.14197299e-01] | [14.76541805267334, 0.8342705368995667] |
9cc2e657-2cd3-47f2-827b-e8d1266c2482 | uncertainty-guided-label-denoising-for | 2305.11029 | null | https://arxiv.org/abs/2305.11029v2 | https://arxiv.org/pdf/2305.11029v2.pdf | Uncertainty Guided Label Denoising for Document-level Distant Relation Extraction | Document-level relation extraction (DocRE) aims to infer complex semantic relations among entities in a document. Distant supervision (DS) is able to generate massive auto-labeled data, which can improve DocRE performance. Recent works leverage pseudo labels generated by the pre-denoising model to reduce noise in DS data. However, unreliable pseudo labels bring new noise, e.g., adding false pseudo labels and losing correct DS labels. Therefore, how to select effective pseudo labels to denoise DS data is still a challenge in document-level distant relation extraction. To tackle this issue, we introduce uncertainty estimation technology to determine whether pseudo labels can be trusted. In this work, we propose a Document-level distant Relation Extraction framework with Uncertainty Guided label denoising, UGDRE. Specifically, we propose a novel instance-level uncertainty estimation method, which measures the reliability of the pseudo labels with overlapping relations. By further considering the long-tail problem, we design dynamic uncertainty thresholds for different types of relations to filter high-uncertainty pseudo labels. We conduct experiments on two public datasets. Our framework outperforms strong baselines by 1.91 F1 and 2.28 Ign F1 on the RE-DocRED dataset. | ['Soujanya Poria', 'Kun Zhang', 'Pengfei Hong', 'Xiaocui Yang', 'Kun Huang', 'Qi Sun'] | 2023-05-18 | null | null | null | null | ['document-level-relation-extraction', 'relation-extraction'] | ['natural-language-processing', 'natural-language-processing'] | [-1.20171243e-02 4.00059968e-01 -2.53341466e-01 -6.92173898e-01
-1.27213717e+00 -7.66415238e-01 6.89137280e-01 3.87733668e-01
-2.64359534e-01 1.02244437e+00 3.24394017e-01 -3.50837968e-02
-3.72052103e-01 -9.14556146e-01 -8.44425917e-01 -6.11723304e-01
1.68947950e-01 7.44302392e-01 2.19171152e-01 -2.33687591e-02
-1.67656928e-01 1.30195737e-01 -1.13327670e+00 4.67008621e-01
1.22930872e+00 1.18933344e+00 -3.16022217e-01 1.56647503e-01
-2.31442943e-01 9.25677657e-01 -7.88266301e-01 -9.08727109e-01
3.65463234e-02 -1.10756032e-01 -1.04413927e+00 -6.51487336e-02
-1.76126026e-02 -8.11732262e-02 -3.47881317e-02 1.32887125e+00
3.38031292e-01 8.76575708e-02 9.85657513e-01 -1.22219002e+00
-5.70215940e-01 1.37739825e+00 -6.37742162e-01 1.67573079e-01
1.62227362e-01 -5.74930012e-01 1.28728282e+00 -9.96307015e-01
7.25600481e-01 1.47580552e+00 5.43799520e-01 3.00719827e-01
-1.29566038e+00 -7.09253132e-01 2.89078891e-01 2.04866484e-01
-1.70232677e+00 -2.97727585e-01 5.58794498e-01 -2.14079335e-01
5.77580452e-01 2.91590631e-01 -3.45870346e-01 1.12637007e+00
2.52126455e-02 6.92031503e-01 1.08672249e+00 -2.90431023e-01
3.05903286e-01 1.12296075e-01 3.54826033e-01 2.46789068e-01
4.59254205e-01 -2.09121078e-01 -6.36183739e-01 -1.67183042e-01
2.45230526e-01 -3.24631929e-01 -3.33645284e-01 2.20032156e-01
-8.81107509e-01 4.75356251e-01 5.28217256e-01 1.93104208e-01
-1.48879915e-01 -1.32049963e-01 3.30427825e-01 5.03112040e-02
9.16033089e-01 4.29400474e-01 -9.55460668e-01 2.72555239e-02
-4.97339219e-01 6.41932432e-03 6.51174307e-01 1.12353551e+00
5.55567324e-01 -6.89824998e-01 -5.69632649e-01 9.46204662e-01
5.23338974e-01 2.23688364e-01 -9.72040147e-02 -9.80223000e-01
7.03788698e-01 6.03507638e-01 6.88362047e-02 -1.05589247e+00
-2.64743358e-01 -7.49236286e-01 -9.82282639e-01 -4.14091885e-01
1.80579007e-01 -3.51652980e-01 -9.04234767e-01 1.56841028e+00
6.25683367e-01 3.03965062e-01 2.17544466e-01 8.21638823e-01
9.23921049e-01 6.97049677e-01 7.38555863e-02 -3.96798790e-01
1.63240945e+00 -6.83230162e-01 -1.17626739e+00 -1.17343485e-01
7.45356202e-01 -5.66299617e-01 5.98671854e-01 5.93352377e-01
-5.52392364e-01 -1.15613818e-01 -9.24779058e-01 -1.49935216e-01
-2.41176546e-01 1.02368608e-01 4.04820263e-01 3.05027068e-01
-2.76961863e-01 5.92450678e-01 -7.61550367e-01 3.95933032e-01
4.23130214e-01 -1.02162976e-02 -3.63956600e-01 -2.81758368e-01
-1.87239814e+00 7.09808469e-01 8.70380521e-01 2.66678631e-01
-5.14845848e-01 -6.20312154e-01 -8.48889112e-01 2.36844033e-01
1.16614509e+00 -3.46334368e-01 1.09517169e+00 -7.69434795e-02
-9.29682374e-01 4.71711516e-01 -1.99668825e-01 -3.71888220e-01
3.59176189e-01 -4.45831090e-01 -7.05627501e-01 -5.66824675e-02
3.64456713e-01 1.83446214e-01 6.16570234e-01 -1.65533400e+00
-6.74683869e-01 -5.09931505e-01 -8.78272206e-02 1.47774875e-01
-8.00798833e-02 -8.50252658e-02 -6.49325013e-01 -8.63946080e-01
5.14921725e-01 -7.13531613e-01 -2.24325005e-02 -7.97540426e-01
-7.72475719e-01 -6.83017552e-01 4.40258920e-01 -7.16772676e-01
1.51398885e+00 -2.06759763e+00 5.18089123e-02 3.81083727e-01
5.07731795e-01 1.53649002e-01 2.17805654e-01 -2.54835226e-02
6.14798330e-02 4.07796830e-01 -3.28191251e-01 -3.70243639e-01
2.02079967e-01 4.28215146e-01 -3.53428453e-01 -7.85628334e-02
4.12959903e-01 8.79145801e-01 -1.04006648e+00 -6.71752155e-01
-3.83971840e-01 3.52002174e-01 -2.70462316e-02 6.23922981e-02
-4.95073140e-01 2.78697193e-01 -4.45821255e-01 7.30764627e-01
1.00485361e+00 -3.60180557e-01 2.36571476e-01 -5.55444539e-01
3.89820993e-01 4.17730302e-01 -1.39943504e+00 1.33730674e+00
-4.06758577e-01 1.56980157e-01 -1.08969361e-02 -6.32574856e-01
1.10682404e+00 2.80586779e-01 1.88813120e-01 -1.74900845e-01
3.14626575e-01 3.10724497e-01 -4.38098788e-01 -2.40488619e-01
4.22530442e-01 -1.37126118e-01 -2.77116865e-01 1.02961034e-01
1.05317786e-01 -1.08093843e-01 2.42797509e-01 6.74868584e-01
1.23416436e+00 -4.60809767e-02 7.82374144e-02 1.06185963e-02
5.25109410e-01 -2.94901729e-01 1.09453368e+00 5.21475494e-01
-4.57709143e-03 5.44812679e-01 1.00836015e+00 1.16697669e-01
-3.30013573e-01 -1.02834666e+00 -5.49624205e-01 8.01730752e-01
2.70863861e-01 -8.80286574e-01 -5.19426286e-01 -1.30619967e+00
3.43510583e-02 1.02923059e+00 -4.63205725e-01 -2.57675648e-01
-1.07555471e-01 -9.77703571e-01 5.02248466e-01 3.67095828e-01
5.40862143e-01 -5.31093776e-01 6.10420048e-01 2.48032615e-01
-7.57062614e-01 -1.51222622e+00 -3.61697316e-01 3.84052247e-01
-4.50358272e-01 -9.98454034e-01 -1.88501164e-01 -3.41393262e-01
6.40829206e-01 -1.88265696e-01 1.31134593e+00 -4.20711935e-01
3.60100776e-01 -2.82253623e-01 -6.10138714e-01 -1.15742944e-01
-3.20865005e-01 2.15509832e-01 1.11318044e-01 -3.58361788e-02
8.08888614e-01 -3.76264244e-01 -2.03925267e-01 4.69348162e-01
-9.34179187e-01 -3.50481898e-01 6.64927840e-01 9.33692455e-01
8.58610630e-01 7.57176757e-01 7.63197601e-01 -1.36475277e+00
9.63364840e-01 -6.85613275e-01 -4.48275834e-01 5.09497702e-01
-1.09175205e+00 5.25131464e-01 4.72271025e-01 -1.69944167e-01
-1.52849495e+00 -1.35629028e-01 -9.01362021e-03 -6.14131391e-02
-6.00418523e-02 7.88707018e-01 -6.32487357e-01 5.95137358e-01
7.96429515e-01 -4.12073910e-01 -7.08213508e-01 -4.41552937e-01
5.93125045e-01 1.00640213e+00 7.66418457e-01 -8.16335142e-01
6.74837649e-01 1.58533230e-01 3.53877172e-02 9.14987922e-02
-1.75765491e+00 -4.04845506e-01 -3.29152942e-01 1.15898490e-01
6.11178100e-01 -1.13840938e+00 -4.06610608e-01 1.72375306e-01
-1.19898820e+00 3.64474148e-01 -2.32009456e-01 3.70438159e-01
1.98418528e-01 3.37339044e-01 -7.24617660e-01 -7.86457241e-01
-3.50492120e-01 -8.91839981e-01 1.30108118e+00 1.82750568e-01
-6.71910718e-02 -6.62111104e-01 -6.97226077e-02 6.19567394e-01
-1.57634377e-01 2.75760800e-01 6.09056115e-01 -6.98258162e-01
-4.92735952e-01 -1.49872139e-01 -5.99146962e-01 6.07707083e-01
2.72534996e-01 -1.32133499e-01 -9.23187435e-01 2.66099304e-01
-1.44164398e-01 -3.59964371e-01 1.18788052e+00 9.39227566e-02
1.04494691e+00 -2.41715550e-01 -5.42545497e-01 2.04939082e-01
9.54927087e-01 -1.71731517e-01 4.18892056e-01 8.66695642e-02
8.23755443e-01 6.42599344e-01 1.10949361e+00 5.02319455e-01
6.93915188e-01 5.83099842e-01 1.47233233e-01 3.78128618e-01
9.72513482e-02 -3.89151692e-01 2.25191623e-01 6.94738925e-01
6.40309602e-02 -4.66144592e-01 -9.29581404e-01 3.17098916e-01
-1.91717470e+00 -4.96536940e-01 -4.46481973e-01 1.72901630e+00
1.45840216e+00 5.37093759e-01 -4.64857250e-01 2.21769542e-01
9.77808297e-01 -1.16916098e-01 -3.67872059e-01 1.28955528e-01
-2.20012799e-01 -8.74772072e-02 3.70350540e-01 5.91947854e-01
-1.23644888e+00 9.24475193e-01 4.36022282e+00 1.25430012e+00
-2.54635304e-01 2.54330575e-01 8.23846817e-01 2.82860249e-01
-6.63614929e-01 1.49728119e-01 -1.20269287e+00 5.61776042e-01
8.46484661e-01 -3.80478986e-02 5.03787678e-03 7.64017761e-01
-6.24915473e-02 -1.32451504e-01 -9.55797493e-01 9.21341479e-01
-6.07212894e-02 -9.45123017e-01 -1.80792123e-01 -2.45594811e-02
1.02390671e+00 -2.87233770e-01 -3.43213171e-01 5.29428840e-01
7.61095941e-01 -8.65728915e-01 3.42600703e-01 5.86712122e-01
6.98503911e-01 -1.08238792e+00 1.31685126e+00 3.90388846e-01
-1.11803055e+00 1.01108685e-01 -3.16214532e-01 3.86512548e-01
3.18955958e-01 1.75213647e+00 -7.87999749e-01 1.00455153e+00
7.10901797e-01 6.16668165e-01 -5.20913601e-01 4.68821555e-01
-8.07184041e-01 6.08671725e-01 -4.66977626e-01 2.37696618e-01
-1.90311730e-01 -4.76666875e-02 2.92190969e-01 1.02986407e+00
1.71395943e-01 4.31774527e-01 -6.15630597e-02 9.47404146e-01
-6.17980182e-01 -1.72852322e-01 -1.15906939e-01 3.89164351e-02
9.30517375e-01 1.48056269e+00 -6.48616195e-01 -2.37159222e-01
6.34298101e-02 9.10874546e-01 4.41922218e-01 2.25257710e-01
-7.83415377e-01 -4.36420649e-01 2.82828957e-01 -2.49265075e-01
-3.25297453e-02 -2.90639941e-02 -4.03396100e-01 -1.42074776e+00
2.29560882e-01 -5.52415371e-01 6.71923578e-01 -5.02322197e-01
-1.63434565e+00 5.97523391e-01 9.69633237e-02 -9.35098886e-01
-1.93775252e-01 -2.99951583e-01 -4.37799543e-02 6.52135551e-01
-1.36580682e+00 -8.90366018e-01 -2.47425094e-01 -4.35388181e-03
1.45250767e-01 1.89353183e-01 4.98767376e-01 4.31899816e-01
-7.07682848e-01 6.50399685e-01 1.01887897e-01 2.02677429e-01
1.02334809e+00 -1.54698300e+00 3.25766593e-01 8.99868190e-01
2.36010671e-01 5.52141130e-01 6.07776880e-01 -1.03230727e+00
-6.72894776e-01 -1.24446988e+00 1.18282712e+00 -7.95105755e-01
6.03135824e-01 -5.11200905e-01 -1.03773952e+00 6.20675445e-01
-3.92770886e-01 3.75737727e-01 5.18168449e-01 4.88057822e-01
-5.07725596e-01 -3.58378530e-01 -1.29890776e+00 3.12687576e-01
1.11435497e+00 -3.56706917e-01 -6.39330208e-01 3.45993966e-01
1.19626307e+00 -5.21264851e-01 -1.21758461e+00 9.18175340e-01
1.29556537e-01 -5.73182464e-01 8.20504546e-01 -1.44927979e-01
4.89293486e-01 -5.84226012e-01 -6.20381013e-02 -1.46371531e+00
-1.10742472e-01 -3.14133912e-01 -6.70978963e-01 2.14516783e+00
7.80324459e-01 -5.23082018e-01 5.17153680e-01 8.35010588e-01
4.54486758e-02 -6.39302194e-01 -8.15521002e-01 -7.73664117e-01
-2.04636782e-01 -6.23328924e-01 7.85847247e-01 9.35366333e-01
-2.52782013e-02 7.09456563e-01 -2.87842751e-01 6.06554806e-01
7.17515647e-01 -9.11996961e-02 3.07363927e-01 -1.44514740e+00
-1.82876959e-01 1.34062946e-01 -1.66460603e-01 -9.33872044e-01
4.08407956e-01 -8.64800394e-01 2.59621143e-01 -1.48178995e+00
3.55005525e-02 -8.39663148e-01 -3.63406688e-01 5.19438922e-01
-6.53452635e-01 -8.06064606e-02 -3.01858068e-01 7.42266849e-02
-9.02677834e-01 7.94183135e-01 1.07873225e+00 -1.57839358e-01
-2.35932190e-02 5.40951528e-02 -9.27210331e-01 6.03217959e-01
7.38644958e-01 -8.18789899e-01 -5.10291636e-01 -2.01422572e-01
7.18341768e-01 -1.59687266e-01 2.42457651e-02 -6.17003143e-01
1.38533980e-01 1.54609624e-02 2.11028829e-01 -7.55942345e-01
-2.22446453e-02 -9.13595319e-01 -1.15439920e-02 -1.57548368e-01
-4.28958803e-01 -5.03757238e-01 -1.93341956e-01 8.98311555e-01
-4.25969154e-01 -3.65452349e-01 3.58044416e-01 1.10530697e-01
-4.25835490e-01 1.07633300e-01 1.36188000e-01 4.18225020e-01
8.60211611e-01 7.17733979e-01 -5.45490205e-01 -1.67532712e-01
-9.01517868e-01 6.62072480e-01 -1.37709662e-01 3.47708553e-01
3.54760557e-01 -1.48324370e+00 -9.08164263e-01 -2.45034710e-01
3.24506491e-01 7.76260495e-01 1.19474202e-01 4.79319870e-01
-2.14473978e-02 2.72905171e-01 7.21833527e-01 -4.38160986e-01
-1.06590796e+00 4.01058316e-01 -8.36697128e-03 -6.40097797e-01
-1.87776640e-01 1.25726366e+00 -2.61041552e-01 -6.26260459e-01
1.12034172e-01 -4.05995548e-01 -3.55536729e-01 2.88609266e-01
4.59200948e-01 4.15294021e-01 4.98038083e-01 -3.50205779e-01
-5.04214585e-01 4.69848365e-02 -3.78005385e-01 -1.11812547e-01
1.14089298e+00 -4.14093941e-01 -5.17333269e-01 4.10364032e-01
8.20701003e-01 1.83456644e-01 -1.04781795e+00 -6.00140333e-01
5.70496678e-01 -3.93806934e-01 4.10986632e-01 -1.13175380e+00
-9.14087474e-01 3.81657600e-01 1.95432097e-01 3.06166559e-01
1.02853417e+00 4.96042192e-01 7.14851558e-01 2.48494029e-01
3.52256596e-01 -1.22793865e+00 -9.92217138e-02 5.14945924e-01
8.37009907e-01 -1.48327172e+00 1.25983819e-01 -1.09891021e+00
-7.44181454e-01 7.36510694e-01 7.46948361e-01 5.39685071e-01
6.77495420e-01 4.91517752e-01 -1.51164858e-02 -2.89212316e-01
-8.04600596e-01 -1.88893080e-01 6.01970553e-01 2.91834086e-01
3.91493797e-01 2.81901330e-01 -5.26487410e-01 1.19732237e+00
-7.62980357e-02 -1.81287095e-01 3.35914582e-01 6.71141088e-01
-1.86853379e-01 -1.30174279e+00 -4.23641324e-01 6.43709660e-01
-5.31558037e-01 -1.78603560e-01 -3.94628286e-01 4.76647913e-02
4.41478401e-01 1.29558802e+00 -4.25333828e-01 -5.97723305e-01
2.28379533e-01 8.16806778e-02 4.97037135e-02 -6.54522717e-01
-7.27549344e-02 -4.18996774e-02 5.40516376e-01 -1.91017434e-01
-4.10845041e-01 -4.96118456e-01 -1.54485178e+00 -1.41269445e-01
-9.60839808e-01 5.04400313e-01 4.44395781e-01 1.12804699e+00
6.04850888e-01 7.56079555e-01 6.05161667e-01 2.29745245e-04
-6.41895294e-01 -1.13606644e+00 -7.71306634e-01 3.35411876e-01
1.41707838e-01 -9.43266392e-01 -5.74065387e-01 -1.57308102e-01] | [9.272506713867188, 8.604736328125] |
6b32e621-dd5f-4b3e-8f23-fe5b37783f77 | learnable-theory-vs-applications | 1807.10681 | null | http://arxiv.org/abs/1807.10681v1 | http://arxiv.org/pdf/1807.10681v1.pdf | Learnable: Theory vs Applications | Two different views on machine learning problem: Applied learning (machine
learning with business applications) and Agnostic PAC learning are formalized
and compared here. I show that, under some conditions, the theory of PAC
Learnable provides a way to solve the Applied learning problem. However, the
theory requires to have the training sets so large, that it would make the
learning practically useless. I suggest shedding some theoretical
misconceptions about learning to make the theory more aligned with the needs
and experience of practitioners. | ['Marina Sapir'] | 2018-07-27 | null | null | null | null | ['misconceptions'] | ['miscellaneous'] | [-3.50143877e-03 6.05505466e-01 -6.54687345e-01 -3.82378578e-01
-6.47690058e-01 -8.40203822e-01 5.81661046e-01 -2.31727719e-01
-3.94581288e-01 8.37633491e-01 9.75247622e-02 -6.89537287e-01
-4.47009325e-01 -6.38433039e-01 -7.81052411e-01 -1.08643866e+00
5.25303185e-02 5.95454156e-01 1.06075957e-01 -1.54363051e-01
3.90124023e-01 1.33547798e-01 -1.34251904e+00 4.23855126e-01
7.02301800e-01 7.93847501e-01 -1.79778531e-01 5.78617871e-01
-1.04367144e-01 1.16297281e+00 -5.82005978e-01 -8.93954039e-01
7.64492810e-01 -4.73703474e-01 -9.95129287e-01 -2.56435484e-01
6.01825230e-02 -1.93771809e-01 -2.76891053e-01 1.29906809e+00
1.27354562e-01 -6.48990571e-02 7.17750013e-01 -1.64109850e+00
-4.79701042e-01 1.13594782e+00 -1.41634136e-01 3.39810699e-01
-1.60203308e-01 4.92369056e-01 1.41594911e+00 -1.29198879e-01
2.00074524e-01 1.06008816e+00 5.29472113e-01 7.80572772e-01
-8.65549207e-01 -6.51629806e-01 1.47327539e-02 5.56795299e-02
-7.93298721e-01 -2.87145644e-01 2.35027730e-01 -3.86389703e-01
5.95888436e-01 5.62086105e-01 8.50725532e-01 1.01135135e+00
1.42336324e-01 8.84276569e-01 1.28642142e+00 -5.38577259e-01
6.58983886e-01 7.95367777e-01 -1.71256810e-02 1.50763160e-02
9.75912273e-01 2.68437177e-01 -4.81277019e-01 -3.63494962e-01
4.19139981e-01 -2.16112770e-02 -3.85564208e-01 -4.13165033e-01
-6.77717924e-01 1.03975344e+00 1.35744378e-01 4.56792533e-01
-2.48797596e-01 2.38091633e-01 5.63222349e-01 6.89020813e-01
4.34473097e-01 8.98275375e-01 -7.25195885e-01 -4.70228791e-01
-3.53498489e-01 1.26959279e-01 1.55909336e+00 1.04484200e+00
2.69151807e-01 1.31020963e-01 4.08746332e-01 2.66939938e-01
3.74715149e-01 3.14242214e-01 5.93999386e-01 -1.30853045e+00
4.31314528e-01 4.15754430e-02 3.21883768e-01 -6.45423293e-01
7.63578638e-02 -2.54621834e-01 -1.64223656e-01 1.60310268e-01
6.95904553e-01 -3.04617167e-01 -5.15593708e-01 1.40809560e+00
-2.31172621e-01 2.00761572e-01 6.37974262e-01 6.95824265e-01
2.91916788e-01 8.09304357e-01 1.40625149e-01 -6.58604026e-01
6.69385314e-01 -6.46376908e-01 -7.40254760e-01 -3.78539383e-01
7.72804797e-01 -2.81189710e-01 1.08560014e+00 6.08256698e-01
-9.93073702e-01 3.30225199e-01 -1.30223358e+00 6.70433044e-01
-3.35545808e-01 -9.73590970e-01 9.35464084e-01 1.22238886e+00
-5.52408397e-01 4.51442540e-01 -6.55933499e-01 -2.40489587e-01
5.06764948e-01 5.16548812e-01 3.12578976e-02 -1.03950478e-01
-8.91825616e-01 1.11873806e+00 9.82449949e-01 -7.76809528e-02
-7.06982315e-01 -3.84471565e-01 -4.97546494e-01 2.71774679e-01
6.96020365e-01 -4.73671108e-01 1.61928689e+00 -1.53089893e+00
-1.53147197e+00 8.53580296e-01 2.74652421e-01 -7.88251340e-01
7.70584047e-01 -4.63344693e-01 -3.88389200e-01 8.74576196e-02
-2.37421155e-01 -2.27293074e-01 8.70108128e-01 -1.55996644e+00
-1.15955532e+00 -3.01624566e-01 1.15229018e-01 3.63966674e-01
-3.44447255e-01 -3.47526260e-02 1.39680821e-02 -3.34884197e-01
-1.80216122e-03 -8.17652762e-01 -2.90511698e-01 -7.90344834e-01
1.75467759e-01 -3.10741037e-01 8.12895238e-01 -8.35091248e-02
6.85022414e-01 -2.45437312e+00 -2.79504865e-01 6.31971657e-01
1.89595848e-01 9.16674212e-02 -9.70117822e-02 3.30214918e-01
3.61400433e-02 5.80931604e-01 2.39463687e-01 5.76666713e-01
9.22965258e-02 6.66708946e-01 -6.79591537e-01 7.18952119e-01
-1.68505326e-01 8.67089152e-01 -9.64795113e-01 -2.15759620e-01
-9.79692712e-02 -1.58781111e-01 -8.22993994e-01 3.61964613e-01
-3.28606397e-01 9.99636129e-02 -6.25656605e-01 4.36178237e-01
4.03464675e-01 -8.51420984e-02 6.90798342e-01 5.25262177e-01
3.59029263e-01 6.74712360e-01 -1.21098995e+00 7.31094897e-01
-4.58524488e-02 4.84848887e-01 5.10153733e-02 -1.38321018e+00
5.01791120e-01 5.96906245e-01 5.39953053e-01 -3.24440092e-01
1.28944978e-01 5.40475726e-01 6.26121342e-01 -7.26736784e-01
-3.69981453e-02 -6.80221260e-01 4.97368053e-02 7.28085518e-01
1.30831361e-01 -5.97118437e-01 -2.66918421e-01 -5.34598790e-02
7.25933492e-01 -5.18701449e-02 1.72572181e-01 -6.73452556e-01
8.76091868e-02 1.71170667e-01 7.61485934e-01 1.19582868e+00
-1.48478135e-01 5.43899424e-02 5.96419752e-01 -6.38218164e-01
-8.66429627e-01 -9.45153177e-01 1.12379909e-01 1.27176058e+00
1.86436638e-01 -4.74948019e-01 -8.24233949e-01 -9.73144889e-01
-4.67091687e-02 1.08036935e+00 -3.32325697e-01 -3.01721036e-01
-5.96977592e-01 -9.10609007e-01 1.64009720e-01 4.71586555e-01
3.41506720e-01 -9.29663122e-01 -6.01444960e-01 2.08162609e-02
1.76537082e-01 -7.97689259e-01 -1.30480155e-01 5.97534776e-01
-9.62614298e-01 -1.23866618e+00 -1.35081679e-01 -6.32811606e-01
3.97942245e-01 5.16723506e-02 1.00027370e+00 4.97027189e-02
2.36711904e-01 1.01656592e+00 -2.77742088e-01 -1.21987331e+00
-6.55404091e-01 -8.71201605e-02 1.80680752e-01 -2.97410518e-01
1.00273764e+00 -5.19805372e-01 -2.07174987e-01 -8.09452012e-02
-7.50938058e-01 -4.35260326e-01 6.70966446e-01 1.11380959e+00
5.67948222e-02 5.90109110e-01 6.37287199e-01 -1.41060233e+00
5.14732182e-01 -5.66095650e-01 -6.55581951e-01 3.72312933e-01
-8.99969280e-01 -1.05267555e-01 4.26865637e-01 -7.46310472e-01
-1.17273855e+00 -3.52901429e-01 2.90414512e-01 -1.82591435e-02
1.19650900e-01 5.90264857e-01 -4.25469428e-01 -2.62340512e-02
9.48305428e-01 1.19793952e-01 1.05368547e-01 -1.99007094e-01
3.11508149e-01 8.27760756e-01 2.51771599e-01 -8.82302403e-01
9.25999045e-01 9.82184634e-02 -2.44023904e-01 -6.39910638e-01
-1.10184157e+00 -1.19597293e-01 -4.37629730e-01 1.27454653e-01
3.58197063e-01 -9.53977704e-01 -7.59078205e-01 -8.13493282e-02
-4.98816311e-01 -5.83754897e-01 -7.40076721e-01 6.42204940e-01
-1.08080876e+00 -1.73252597e-01 -6.05099320e-01 -1.07895625e+00
-1.71916172e-01 -9.59208965e-01 -7.82006606e-02 3.17450255e-01
-2.94412911e-01 -1.08640659e+00 -3.02885175e-01 4.59154874e-01
3.99445176e-01 5.61933853e-02 1.06450593e+00 -1.56984627e+00
-6.19147062e-01 -3.60097378e-01 5.04436195e-01 6.26518965e-01
5.53893335e-02 -4.70771678e-02 -1.20494914e+00 -1.99296221e-01
7.24323988e-01 -4.00518805e-01 2.95914680e-01 2.73071676e-01
1.41792166e+00 -1.39319015e+00 -1.62548989e-01 3.37950677e-01
1.53368986e+00 6.79236233e-01 5.32579720e-01 6.66249812e-01
2.24606350e-01 6.94626510e-01 6.30896330e-01 3.49665135e-01
9.55702513e-02 -1.87950581e-02 2.25624979e-01 6.07467055e-01
9.40591931e-01 -3.88334036e-01 2.84959346e-01 7.85185337e-01
-2.25568056e-01 -5.24460822e-02 -9.82968271e-01 4.09000874e-01
-2.10471916e+00 -1.17782462e+00 1.21939547e-01 2.30040336e+00
1.08845961e+00 6.37744546e-01 2.11924434e-01 3.78753915e-02
3.84143919e-01 -1.49145439e-01 -3.21638018e-01 -6.88032627e-01
-1.06309116e-01 3.03690195e-01 7.07165480e-01 2.82215953e-01
-9.73794341e-01 6.51483715e-01 8.27650070e+00 7.73352265e-01
-1.00021231e+00 2.22324029e-01 9.05064225e-01 7.34347701e-02
-5.94993055e-01 3.04234415e-01 -3.76842767e-01 3.97378683e-01
1.33174980e+00 -7.94875622e-01 5.68059087e-01 1.41580975e+00
-4.68800008e-01 -2.74815205e-02 -1.55924439e+00 9.93689954e-01
-3.31932902e-01 -1.46122050e+00 4.12435830e-02 7.53631582e-03
7.54175901e-01 -1.50267884e-01 3.93207222e-01 5.35639107e-01
7.97462225e-01 -1.37388325e+00 4.30753797e-01 -6.47365674e-02
3.52068841e-01 -8.81266236e-01 1.16810882e+00 7.82725215e-01
-2.20494181e-01 -7.81034887e-01 -3.20808500e-01 -3.64703894e-01
-5.11765599e-01 2.76947785e-02 -6.76064312e-01 1.52290106e-01
8.23318005e-01 4.85697091e-01 -1.91558152e-01 1.01320624e+00
-3.35332230e-02 1.02544773e+00 -3.80139202e-01 -6.15435652e-02
2.95901865e-01 -3.03482622e-01 2.67522901e-01 9.42429483e-01
9.80651230e-02 5.16288936e-01 -2.00222373e-01 6.09556854e-01
1.86675534e-01 1.30676776e-02 -9.86325622e-01 -4.80897129e-01
6.43710017e-01 8.71940255e-01 -4.03557956e-01 -3.69769037e-01
-5.52260876e-01 2.15517774e-01 -2.16930196e-01 3.68272960e-01
-2.67755210e-01 1.31325185e-01 5.21756113e-01 2.44543687e-01
-9.25100148e-02 1.95821330e-01 -6.61528468e-01 -1.10565579e+00
-2.87120759e-01 -1.32092559e+00 8.37909818e-01 -8.11687335e-02
-1.59249103e+00 -1.90191895e-01 2.17626974e-01 -9.82831001e-01
-4.92957234e-01 -8.86186182e-01 -8.72727036e-01 3.79223049e-01
-7.68621325e-01 -7.13330925e-01 2.70628870e-01 3.70243400e-01
2.32247651e-01 -5.89542150e-01 8.02941024e-01 -4.03478533e-01
-2.94683948e-02 6.12906516e-01 2.31423438e-01 5.24284765e-02
4.58438545e-01 -1.37394273e+00 -1.10434353e-01 4.19628114e-01
2.52288610e-01 5.55517495e-01 9.31442440e-01 -9.50725600e-02
-1.75817811e+00 -6.74050212e-01 5.43677211e-01 -8.45934808e-01
5.02824306e-01 -6.32633567e-02 -6.37621641e-01 1.31958055e+00
2.13249147e-01 -3.37689012e-01 9.51409459e-01 1.92124531e-01
-4.61647063e-01 -3.64331543e-01 -1.44052935e+00 7.29131103e-01
5.14348686e-01 -4.06341404e-01 -9.15068746e-01 6.78819180e-01
7.91750789e-01 -3.46904904e-01 -7.93332100e-01 4.43696529e-01
4.24700379e-01 -8.53615046e-01 6.25975311e-01 -9.65915024e-01
7.19671622e-02 1.85928389e-01 -2.56058007e-01 -1.07120216e+00
-8.51571411e-02 -1.09775007e+00 -3.46685886e-01 7.62248933e-01
4.97409999e-01 -1.38390434e+00 8.29966426e-01 1.27831089e+00
1.67625800e-01 -5.51823437e-01 -9.22330797e-01 -1.02568436e+00
6.09256864e-01 -5.54866850e-01 4.51053798e-01 1.15043485e+00
6.12543285e-01 1.81156576e-01 -7.86939785e-02 6.46860749e-02
3.32242370e-01 -2.48611569e-01 8.42270195e-01 -1.21788871e+00
-4.84963298e-01 -5.31084180e-01 -3.54749918e-01 -5.42980492e-01
1.51441693e-01 -5.45889914e-01 1.65506423e-01 -8.17167103e-01
2.64873624e-01 -8.94000649e-01 -6.40233815e-01 4.69340265e-01
2.93498598e-02 -2.50689119e-01 4.13623244e-01 2.89480954e-01
-2.77192742e-01 -3.85191068e-02 7.69878745e-01 -1.65507331e-01
-3.16977322e-01 5.07323325e-01 -1.40705287e+00 9.70010579e-01
1.20104146e+00 -5.40080428e-01 -6.11732781e-01 4.36641313e-02
3.38114113e-01 -2.67566234e-01 -1.35372175e-04 -6.41557395e-01
1.06565416e-01 -1.04940498e+00 2.72580743e-01 2.08183423e-01
2.40980208e-01 -1.23043549e+00 1.24744095e-01 8.13848078e-01
-4.94827539e-01 -2.43776903e-01 2.23727226e-02 3.61576319e-01
-1.52222235e-02 -7.68735230e-01 8.20197999e-01 -4.28260773e-01
-4.22057807e-01 2.08121106e-01 -3.14347923e-01 4.72114831e-01
1.44000983e+00 4.17281911e-02 -3.96020353e-01 -8.40189993e-01
-6.02887094e-01 7.67721608e-02 3.88239950e-01 1.20997474e-01
3.85908157e-01 -1.01283932e+00 -3.65014553e-01 6.31247880e-03
-4.24822494e-02 4.01487350e-02 -3.36417764e-01 6.69746578e-01
-5.10725021e-01 3.78529757e-01 -1.18386194e-01 -3.48055631e-01
-7.60671079e-01 7.51573920e-01 6.43876910e-01 3.71177197e-01
-8.18025172e-01 7.94835925e-01 1.65447503e-01 -1.43893987e-01
5.77027977e-01 3.03222779e-02 1.74391761e-01 -1.98376149e-01
6.69309258e-01 6.28185645e-02 -5.02254546e-01 2.77811587e-02
-6.11894801e-02 -5.36612533e-02 -1.26568273e-01 -2.53509521e-01
1.59386945e+00 3.25286686e-01 -2.79562563e-01 7.83300102e-01
6.93515837e-01 -8.76896828e-03 -1.03735971e+00 -4.03382719e-01
6.92744434e-01 -7.18863666e-01 -4.78688449e-01 -5.96745968e-01
-6.43675089e-01 6.88938379e-01 4.24905330e-01 4.85523105e-01
9.05213892e-01 6.69882223e-02 2.31417477e-01 1.01783562e+00
4.46948677e-01 -1.62822735e+00 1.54623315e-01 2.70368189e-01
5.67451596e-01 -1.28101504e+00 1.54581964e-01 -3.37563962e-01
-1.04800940e+00 1.20053887e+00 8.07432711e-01 -1.21723741e-01
1.02364230e+00 3.52859259e-01 1.21748470e-01 -8.65974056e-04
-1.05370486e+00 2.15985849e-01 7.24499067e-03 8.79540443e-01
1.10631488e-01 2.27500871e-01 -4.16619152e-01 9.51434433e-01
-5.37259221e-01 -8.89500752e-02 6.65357411e-01 1.10000277e+00
-6.35369956e-01 -8.10299873e-01 -2.88854271e-01 7.27228761e-01
-1.05204535e+00 1.24034785e-01 -2.24020645e-01 1.07818544e+00
9.16602239e-02 8.71200562e-01 4.35761847e-02 -3.57080221e-01
-2.08382905e-01 2.61942357e-01 3.79551411e-01 -8.18684042e-01
-4.98724520e-01 4.59404886e-02 2.69100405e-02 -1.78035244e-01
-3.88667613e-01 -2.23136067e-01 -9.57708180e-01 -5.47832251e-01
-2.87303448e-01 4.79704410e-01 2.16786146e-01 1.08123362e+00
-1.61545947e-01 -4.82277244e-01 5.90042949e-01 -3.85572970e-01
-1.35497892e+00 -7.82626748e-01 -9.12124455e-01 2.87391186e-01
1.49898052e-01 -1.38300747e-01 -6.59895957e-01 -5.45363780e-03] | [8.356961250305176, 4.754675388336182] |
c50d2252-bdad-488e-866e-a955bfe5c186 | paramnet-a-parameter-variable-network-for | 2305.06511 | null | https://arxiv.org/abs/2305.06511v1 | https://arxiv.org/pdf/2305.06511v1.pdf | ParamNet: A Parameter-variable Network for Fast Stain Normalization | In practice, digital pathology images are often affected by various factors, resulting in very large differences in color and brightness. Stain normalization can effectively reduce the differences in color and brightness of digital pathology images, thus improving the performance of computer-aided diagnostic systems. Conventional stain normalization methods rely on one or several reference images, but one or several images are difficult to represent the entire dataset. Although learning-based stain normalization methods are a general approach, they use complex deep networks, which not only greatly reduce computational efficiency, but also risk introducing artifacts. StainNet is a fast and robust stain normalization network, but it has not a sufficient capability for complex stain normalization due to its too simple network structure. In this study, we proposed a parameter-variable stain normalization network, ParamNet. ParamNet contains a parameter prediction sub-network and a color mapping sub-network, where the parameter prediction sub-network can automatically determine the appropriate parameters for the color mapping sub-network according to each input image. The feature of parameter variable ensures that our network has a sufficient capability for various stain normalization tasks. The color mapping sub-network is a fully 1x1 convolutional network with a total of 59 variable parameters, which allows our network to be extremely computationally efficient and does not introduce artifacts. The results on cytopathology and histopathology datasets show that our ParamNet outperforms state-of-the-art methods and can effectively improve the generalization of classifiers on pathology diagnosis tasks. The code has been available at https://github.com/khtao/ParamNet. | ['Xiuli Liu', 'Shaoqun Zeng', 'Tingwei Quan', 'Shenghua Cheng', 'Junbo Hu', 'Li Chen', 'Die Luo', 'Hongtao Kang'] | 2023-05-11 | null | null | null | null | ['parameter-prediction'] | ['miscellaneous'] | [ 1.27581075e-01 -3.61180365e-01 -5.43792136e-02 -3.05087745e-01
-4.27459568e-01 -4.72140461e-01 7.95220435e-02 9.93086919e-02
-6.12933517e-01 5.10097623e-01 -4.73000020e-01 -5.21486104e-01
1.74094662e-01 -1.02590406e+00 -2.39082292e-01 -1.18232000e+00
2.48509809e-01 1.55969784e-01 5.92587531e-01 4.34448896e-03
-7.26501048e-02 8.33859384e-01 -1.15899444e+00 1.57275006e-01
7.74196267e-01 9.36428964e-01 2.07608953e-01 7.82429218e-01
-4.71317887e-01 6.36984050e-01 -5.79344153e-01 -2.13787004e-01
9.06642973e-02 -5.57727993e-01 -6.36844218e-01 -2.92160094e-01
1.16922453e-01 -3.17549199e-01 -2.30854109e-01 1.30625570e+00
7.01859772e-01 -1.74511775e-01 7.45987058e-01 -1.31564796e+00
-5.72580695e-01 2.65862823e-01 -5.69457352e-01 4.11070526e-01
-3.54188055e-01 3.97056520e-01 7.24501312e-01 -4.54514414e-01
5.56656361e-01 1.12745690e+00 8.00854743e-01 6.30356491e-01
-1.09429252e+00 -1.08012998e+00 -2.05787167e-01 2.19864309e-01
-1.53990817e+00 1.21786602e-01 6.68413103e-01 -3.74377996e-01
6.25661016e-01 3.15598458e-01 8.90502930e-01 5.02393305e-01
6.14744663e-01 4.69909281e-01 1.12501395e+00 -1.25019178e-01
-2.84573957e-02 -7.12014660e-02 -4.38774005e-02 9.88572836e-01
3.73775840e-01 -2.86764801e-01 3.45270596e-02 -8.82803574e-02
1.07594705e+00 4.55411643e-01 -4.38683867e-01 -2.00295553e-01
-1.07596743e+00 7.67284632e-01 9.05691087e-01 5.34886241e-01
-6.65008426e-02 1.33590385e-01 4.11423743e-01 1.71118498e-01
2.55348027e-01 3.59244436e-01 -1.42834038e-01 2.33644083e-01
-5.71696699e-01 -1.84202358e-01 4.02401924e-01 3.23591292e-01
8.45584750e-01 -2.94205040e-01 -1.28765672e-01 9.99469638e-01
1.70356095e-01 5.17284513e-01 5.56911230e-01 -4.59543467e-01
1.31189167e-01 1.08231962e+00 -3.02554905e-01 -1.04813731e+00
-1.03929830e+00 -3.49457175e-01 -1.22622585e+00 5.03322065e-01
5.23355305e-01 -2.54780501e-02 -1.30364215e+00 1.59059405e+00
3.59249979e-01 -4.17055711e-02 -2.12431967e-01 1.02886474e+00
9.49452877e-01 4.53433245e-01 2.03833997e-01 -8.63357857e-02
1.50948000e+00 -8.24082315e-01 -8.60105693e-01 -2.70306412e-02
8.63693893e-01 -8.05913210e-01 1.32934487e+00 2.13451892e-01
-7.20167220e-01 -2.25949213e-01 -1.42581868e+00 -2.16185987e-01
-6.28273189e-01 1.51167482e-01 7.81620383e-01 5.21484494e-01
-1.02209508e+00 4.41080362e-01 -9.82733667e-01 -4.04242337e-01
7.04434514e-01 5.85366368e-01 -5.08285522e-01 -2.19739020e-01
-9.61350739e-01 8.11991632e-01 3.39941680e-01 3.35430741e-01
-3.72374624e-01 -7.26151586e-01 -6.75670624e-01 5.92658529e-03
3.28950994e-02 -4.84406590e-01 1.00925660e+00 -9.61334527e-01
-1.25510633e+00 8.85595143e-01 1.33641839e-01 6.77288845e-02
5.04041076e-01 5.76624870e-01 -4.55754668e-01 3.14435035e-01
-2.46706605e-01 6.69882834e-01 2.52834439e-01 -9.20526803e-01
-6.15898490e-01 -1.84626490e-01 -1.48614338e-02 6.70729578e-02
-5.35634041e-01 -9.89785865e-02 -7.30576217e-01 -5.60485959e-01
6.90461919e-02 -8.10161293e-01 -1.86803564e-01 5.99287689e-01
-2.65329212e-01 -4.60942611e-02 8.94919634e-01 -4.75741893e-01
1.02166784e+00 -2.37254810e+00 -3.54088217e-01 3.61970067e-01
4.03460354e-01 3.99799317e-01 -2.75258839e-01 -4.29129787e-02
-2.07028046e-01 3.76045018e-01 -2.28780806e-01 1.10500650e-02
-2.30054095e-01 2.08633304e-01 3.84545982e-01 7.94065416e-01
2.81822145e-01 7.73105025e-01 -8.16164851e-01 -8.11722457e-01
3.02136809e-01 8.32821012e-01 -2.36861452e-01 2.19845086e-01
7.96169713e-02 1.14675295e-02 -3.21013510e-01 7.30482936e-01
1.01808465e+00 -6.45846009e-01 2.30451882e-01 -5.95318198e-01
1.99276701e-01 -1.82641804e-01 -1.07715750e+00 1.18395972e+00
-2.91268170e-01 6.54835820e-01 1.58321157e-01 -5.70363164e-01
9.56504464e-01 9.10987407e-02 4.31801289e-01 -6.19301677e-01
6.61639392e-01 2.89738506e-01 3.89228225e-01 -4.92465079e-01
-3.89303528e-02 -3.87443691e-01 3.41621399e-01 5.17834246e-01
-1.40871376e-01 -4.09544855e-02 4.65238005e-01 1.84337720e-02
1.23153710e+00 -5.86193860e-01 2.64165342e-01 -1.26180589e-01
7.02323198e-01 -2.18951419e-01 7.51661360e-01 2.40933269e-01
-3.51390690e-01 6.31863654e-01 8.34309936e-01 -6.81156397e-01
-9.04743671e-01 -8.70411873e-01 -3.10856074e-01 6.95645273e-01
4.04562861e-01 -7.67607167e-02 -6.67627811e-01 -6.37477875e-01
-8.56672749e-02 -6.21420853e-02 -9.97928858e-01 -1.87150672e-01
-4.35768366e-01 -1.18456364e+00 5.02140105e-01 6.13984048e-01
6.18651748e-01 -1.07774699e+00 -3.14599782e-01 7.10117444e-03
-8.15809071e-02 -6.73301280e-01 -6.24235988e-01 3.00237536e-01
-7.93578684e-01 -1.47596133e+00 -5.36653876e-01 -9.28938746e-01
1.42649829e+00 2.96926469e-01 7.76072204e-01 9.27538753e-01
-6.18665755e-01 -3.57595831e-01 -2.10764647e-01 -4.76626068e-01
-4.42894429e-01 2.95883030e-01 -3.34933937e-01 -1.16379326e-02
5.62519431e-01 -3.25424135e-01 -8.59721184e-01 3.74459088e-01
-1.39535809e+00 2.59920239e-01 8.63305151e-01 1.13030982e+00
8.61291766e-01 -2.26364303e-02 1.82932988e-01 -1.12060511e+00
5.54442704e-01 -3.22926715e-02 -5.95914066e-01 2.33910292e-01
-5.36277533e-01 -1.32851422e-01 9.04235840e-01 -4.81709957e-01
-7.19376504e-01 1.31556183e-01 -2.00577542e-01 -2.20473617e-01
1.49425358e-01 4.33877051e-01 -1.16463110e-01 -4.76348102e-01
5.89797735e-01 -2.26652890e-01 5.52030861e-01 -1.02744013e-01
-2.10638985e-01 5.08956254e-01 3.99734765e-01 3.37185785e-02
8.80932868e-01 5.33722043e-01 2.81717032e-01 -2.47862801e-01
-5.38870335e-01 -4.10620660e-01 -5.46577096e-01 -2.29470074e-01
8.47391188e-01 -5.19731760e-01 -9.16280150e-01 7.38462210e-01
-7.67779529e-01 -5.37658036e-01 2.67159939e-01 3.81634474e-01
-1.79605633e-02 2.57397920e-01 -1.02974021e+00 -1.86322987e-01
-5.54839194e-01 -1.31834674e+00 4.86067414e-01 7.03411639e-01
6.09839670e-02 -1.13387537e+00 -1.24811076e-01 8.52030814e-02
5.19841015e-01 4.75027472e-01 1.13326943e+00 -3.31269473e-01
-4.18136895e-01 -3.49312276e-01 -6.30955219e-01 3.01874340e-01
5.81064582e-01 3.38702172e-01 -8.41387689e-01 -4.17875886e-01
-4.39271629e-01 -1.46631375e-01 8.64559591e-01 3.70039523e-01
1.53048706e+00 5.77118397e-02 -4.90059435e-01 1.07549334e+00
1.73123133e+00 2.95169830e-01 7.44803190e-01 5.67536235e-01
6.42427981e-01 2.65412390e-01 4.50187147e-01 -2.08946913e-02
1.89446718e-01 2.25824580e-01 5.17801285e-01 -1.00893092e+00
-3.04508656e-01 1.36029094e-01 -2.50756383e-01 5.93117714e-01
6.56004180e-04 -1.69144481e-01 -9.37549949e-01 4.72425461e-01
-1.53098881e+00 -5.90411484e-01 -6.23540096e-02 1.73630476e+00
1.08912671e+00 -4.57684994e-02 -1.53256515e-02 1.45629987e-01
6.84795856e-01 -1.68902785e-01 -7.63998091e-01 -2.30124071e-01
-9.89094526e-02 1.86832681e-01 7.01435268e-01 2.59134024e-01
-1.02529478e+00 5.20030320e-01 5.91442585e+00 8.73620272e-01
-1.59816885e+00 9.29061510e-03 8.60542238e-01 -1.56483337e-01
-8.50781575e-02 -6.11030638e-01 -5.49987912e-01 5.12549043e-01
4.55760926e-01 2.74024196e-02 1.37244806e-01 5.57042301e-01
9.94743332e-02 -1.35358706e-01 -8.50366890e-01 9.59415913e-01
3.06407232e-02 -1.41610110e+00 1.04705006e-01 1.46583691e-01
3.40786308e-01 -7.05676451e-02 3.64708491e-02 -5.83313480e-02
1.50751725e-01 -1.17938697e+00 8.66078809e-02 2.41230160e-01
1.13406122e+00 -9.05986607e-01 1.48273265e+00 -1.95483491e-01
-1.11502278e+00 3.53915207e-02 -4.54538912e-01 4.01769966e-01
-1.89079270e-01 5.04605770e-01 -9.56950009e-01 1.23808891e-01
7.37856150e-01 5.43427289e-01 -8.03426087e-01 1.31791735e+00
-2.68211365e-01 4.25962448e-01 -2.18835965e-01 -3.88562739e-01
2.15482607e-01 -6.90051690e-02 -1.99163347e-01 1.49545860e+00
2.20759228e-01 9.96028334e-02 5.50443567e-02 5.74513137e-01
-2.96456069e-01 3.95195991e-01 -2.71338969e-02 6.74189329e-02
4.32245433e-01 1.83987761e+00 -1.48233712e+00 -1.11862101e-01
-2.90286928e-01 5.20427227e-01 3.00912350e-01 2.87602037e-01
-9.00965512e-01 -8.60421538e-01 7.29554951e-01 3.29485610e-02
1.74635261e-01 1.76265299e-01 -3.53711903e-01 -7.26319551e-01
-1.76330432e-01 -7.63533652e-01 5.78116953e-01 -6.01200402e-01
-1.33410561e+00 7.10739255e-01 -5.18012941e-01 -1.40145051e+00
3.67526144e-01 -1.06167197e+00 -6.26225829e-01 8.40107799e-01
-1.77930892e+00 -1.18665075e+00 -8.39094102e-01 6.45329714e-01
-8.32608715e-02 -3.11517175e-02 8.59050393e-01 2.95104891e-01
-8.17458987e-01 8.55174065e-01 1.02971025e-01 5.09906948e-01
1.02931237e+00 -1.25844812e+00 -1.85040802e-01 7.82430649e-01
-6.26399994e-01 5.93577683e-01 3.86182368e-01 -3.37517798e-01
-1.18632865e+00 -1.10513926e+00 3.57951760e-01 7.70550892e-02
6.68280721e-01 -3.76873881e-01 -1.13191199e+00 1.89549416e-01
-2.85786558e-02 4.70166832e-01 1.10482812e+00 -1.61728069e-01
-2.92878896e-01 -4.25860465e-01 -1.42317033e+00 6.63373172e-01
5.30663073e-01 -3.50773990e-01 7.75724277e-02 4.19734389e-01
4.41544056e-01 -6.47208214e-01 -9.23075974e-01 2.03412518e-01
5.53610623e-01 -7.76974857e-01 7.82434046e-01 -6.02149889e-02
3.37991416e-01 -7.37990677e-01 4.99617934e-01 -1.15981209e+00
-6.47177935e-01 -2.02368870e-02 3.53709430e-01 1.00802481e+00
5.26947618e-01 -1.02088678e+00 7.74012089e-01 4.29915398e-01
1.10756597e-02 -1.10305703e+00 -8.18823576e-01 -5.41111469e-01
1.42431438e-01 1.01099322e-02 9.09002423e-01 9.19063687e-01
-1.23127662e-01 5.56544997e-02 1.57973558e-01 1.38436137e-02
3.90131563e-01 -7.62329921e-02 5.68386376e-01 -1.20355678e+00
2.47923583e-02 -7.93202877e-01 -6.53644621e-01 -2.37518042e-01
-4.06859294e-02 -9.27086890e-01 1.14251906e-02 -1.67976606e+00
5.28863370e-01 -5.74933410e-01 -7.30881512e-01 9.96379077e-01
-4.33470130e-01 8.97421181e-01 5.78778461e-02 1.66636452e-01
-3.57612520e-01 -3.09390621e-03 1.68263173e+00 -4.42475528e-01
-1.72940403e-01 -3.43411356e-01 -8.16741467e-01 5.77313066e-01
1.02907610e+00 -4.21657383e-01 -3.04175347e-01 -1.86827406e-01
2.34663174e-01 -5.38152337e-01 2.04986379e-01 -8.61383796e-01
3.55144173e-01 -2.74482638e-01 9.05606210e-01 -5.17859519e-01
4.35334966e-02 -8.66755605e-01 2.84410864e-01 7.46794283e-01
3.29638980e-02 -3.49844769e-02 3.82749200e-01 9.11220834e-02
-2.13757381e-01 -1.29848868e-01 1.20217884e+00 -1.51277648e-03
-3.48372757e-01 5.88186324e-01 -3.22269648e-01 -4.39453363e-01
9.81885016e-01 -4.59430873e-01 -6.71283722e-01 1.94879193e-02
-1.72090471e-01 3.00976604e-01 7.22324610e-01 9.15718600e-02
4.62521106e-01 -1.36137617e+00 -3.15806448e-01 1.61835641e-01
1.62151441e-01 2.19704062e-01 5.13502359e-01 9.92505968e-01
-1.24000633e+00 6.31007627e-02 -6.18252695e-01 -6.11371636e-01
-1.59533381e+00 3.20694953e-01 7.20479369e-01 -4.64500815e-01
-5.27266383e-01 8.70867908e-01 3.75900090e-01 -4.16362733e-01
5.18622808e-02 -5.27839184e-01 -3.98120373e-01 -1.62684426e-01
9.21624422e-01 1.13072403e-01 2.19785601e-01 -3.06976676e-01
-3.84693086e-01 6.46088064e-01 -3.99421901e-01 5.03306210e-01
1.42856896e+00 2.31229454e-01 -4.88002360e-01 1.67627022e-01
1.29603136e+00 -3.88562530e-01 -1.13133597e+00 -6.49240315e-02
-6.00206196e-01 -4.82122332e-01 3.26977670e-01 -9.52570021e-01
-1.57755697e+00 8.12771082e-01 9.07696187e-01 1.74055740e-01
1.55670989e+00 -1.97400048e-01 7.24138677e-01 1.38273060e-01
-3.87301594e-02 -1.06603634e+00 -5.01366612e-03 2.72929728e-01
5.23154199e-01 -1.12364852e+00 1.64607063e-01 -5.76653838e-01
-2.24488586e-01 1.43021274e+00 9.19598639e-01 8.58212188e-02
4.89281476e-01 6.92999125e-01 6.52161598e-01 -2.28333741e-01
-4.67000008e-01 1.00414127e-01 1.82006061e-01 5.90049446e-01
6.12575233e-01 3.83983552e-02 -3.85544419e-01 5.01892269e-01
-1.50002450e-01 4.64910381e-02 5.24057627e-01 7.74348259e-01
-2.26834431e-01 -1.04262316e+00 -4.68700826e-01 7.18468308e-01
-5.23263812e-01 1.83330506e-01 -2.66836703e-01 9.69409168e-01
2.49562308e-01 6.52456522e-01 1.93260163e-01 -3.58251929e-01
2.44010240e-01 -2.62868851e-01 3.12728345e-01 -2.89477021e-01
-6.75245225e-01 8.16781521e-02 -3.83151799e-01 -4.40818548e-01
-4.51220602e-01 -3.52320880e-01 -1.59550917e+00 -5.92134833e-01
-3.93479317e-01 -8.21738467e-02 4.58045363e-01 6.49683475e-01
2.91506443e-02 8.48933816e-01 4.77145731e-01 -4.73328233e-01
8.13223422e-03 -1.03239930e+00 -7.80099869e-01 3.73658746e-01
3.58761042e-01 -7.26646304e-01 -3.45963120e-01 1.12451978e-01] | [14.99111557006836, -2.9552969932556152] |
6e1b50f6-7446-423a-9c9e-c6af5e675165 | idtrackerai-tracking-all-individuals-in-large | 1803.04351 | null | http://arxiv.org/abs/1803.04351v1 | http://arxiv.org/pdf/1803.04351v1.pdf | idtracker.ai: Tracking all individuals in large collectives of unmarked animals | Our understanding of collective animal behavior is limited by our ability to
track each of the individuals. We describe an algorithm and software,
idtracker.ai, that extracts from video all trajectories with correct identities
at a high accuracy for collectives of up to 100 individuals. It uses two deep
networks, one detecting when animals touch or cross and another one for animal
identification, trained adaptively to conditions and difficulty of the video. | ['Francisco Romero-Ferrero', 'Gonzalo G. de Polavieja', 'Francisco J. H. Heras', 'Robert Hinz', 'Mattia G. Bergomi'] | 2018-03-12 | null | null | null | null | ['multi-animal-tracking-with-identification'] | ['computer-vision'] | [-2.52415210e-01 -4.72934574e-01 1.35156661e-01 -1.36621222e-01
5.54455854e-02 -1.05446529e+00 4.22993213e-01 2.18418390e-01
-8.01232755e-01 6.78877056e-01 -2.32512623e-01 4.27698195e-02
-1.35678351e-01 -7.00601459e-01 -5.79745054e-01 -5.20728886e-01
-8.81656170e-01 7.98604608e-01 6.03807569e-01 1.41607448e-02
6.00919835e-02 5.72440803e-01 -1.47877610e+00 1.09077945e-01
7.69684166e-02 7.42890060e-01 -2.72349447e-01 1.53605652e+00
3.24122429e-01 1.04333866e+00 -7.85565257e-01 -4.62419927e-01
4.49868143e-01 -3.38444829e-01 -6.56457901e-01 -4.55966622e-01
6.22411132e-01 -7.24014759e-01 -6.16519094e-01 8.75002801e-01
1.37487024e-01 -1.32665128e-01 6.32962048e-01 -1.39913630e+00
-5.89155018e-01 1.01870894e+00 -2.90600985e-01 1.05441844e+00
2.76603460e-01 6.27720714e-01 5.82213223e-01 -1.62416503e-01
9.19193685e-01 1.06374860e+00 1.39300084e+00 7.47318506e-01
-1.23835766e+00 -7.35885143e-01 -2.61943907e-01 3.34274359e-02
-1.55454457e+00 -6.68343246e-01 1.24283805e-01 -8.20237458e-01
8.53740573e-01 -6.75825253e-02 8.38970304e-01 1.02309203e+00
1.15274765e-01 2.86547899e-01 5.04062772e-01 2.43059069e-01
1.75736219e-01 -4.90113705e-01 2.01366261e-01 7.59658992e-01
4.02865410e-01 3.48645389e-01 -5.15531898e-01 -6.05212748e-01
9.03794765e-01 -4.54600789e-02 9.31398869e-02 -7.22649470e-02
-1.43942022e+00 5.27473390e-01 3.94745946e-01 2.49235079e-01
-4.39214110e-01 8.16076756e-01 2.95129180e-01 6.67311788e-01
1.56589031e-01 6.08332932e-01 -3.21812183e-01 -4.13054198e-01
-9.13216889e-01 7.28101909e-01 1.07978141e+00 8.26708257e-01
4.92914438e-01 1.61622129e-02 1.07767068e-01 1.89090654e-01
3.72535996e-02 7.04557776e-01 1.73965558e-01 -1.26248503e+00
-1.72036499e-01 6.77454889e-01 2.59885818e-01 -1.22279251e+00
-6.40297949e-01 -1.54440269e-01 -5.87446868e-01 1.43422559e-01
5.48216760e-01 -6.90084040e-01 -7.37133503e-01 1.75222099e+00
3.93297791e-01 4.55918700e-01 -1.80492308e-02 7.13949382e-01
9.19713557e-01 3.32361639e-01 1.88814521e-01 2.36390993e-01
1.04269755e+00 -5.67682147e-01 -4.59637105e-01 -2.66360164e-01
6.15388334e-01 -2.34616846e-01 -7.66217709e-02 -1.09685048e-01
-9.63570535e-01 -2.63291806e-01 -7.90193260e-01 4.52173203e-01
-6.63535058e-01 1.29342198e-01 9.03755724e-01 3.82955164e-01
-1.47527874e+00 1.03667438e+00 -1.12564027e+00 -7.20641553e-01
7.92619944e-01 9.68768299e-01 -5.62742054e-01 7.83408403e-01
-8.41512263e-01 5.69431961e-01 5.53879933e-03 7.73717090e-03
-1.12850952e+00 -4.61875200e-01 -4.98828977e-01 -1.87329680e-01
3.46994810e-02 -6.63895130e-01 1.49267924e+00 -1.33616138e+00
-1.25049722e+00 1.19620609e+00 2.65006959e-01 -1.09919965e+00
6.21395171e-01 -5.21521643e-02 -4.18476194e-01 4.56658095e-01
2.50082433e-01 8.65437925e-01 7.52725840e-01 -1.00466549e+00
-9.51380849e-01 -3.25090110e-01 -3.10258586e-02 -4.44087163e-02
-3.27607617e-02 2.52890229e-01 -4.54525426e-02 -5.75262487e-01
-4.34229463e-01 -1.12180603e+00 -1.97466701e-01 4.32842761e-01
3.96908745e-02 -1.82032045e-02 9.05183852e-01 -3.87067378e-01
1.01688945e+00 -1.92906463e+00 3.48648340e-01 1.16470918e-01
5.82930267e-01 3.81926090e-01 -4.16392863e-01 7.90425837e-01
5.29976428e-01 2.02734649e-01 -2.40924671e-01 -2.54914224e-01
-3.04429084e-01 1.25265807e-01 -4.09791395e-02 9.54966307e-01
4.98445854e-02 6.95191622e-01 -1.13806784e+00 -3.79911363e-01
-1.42333552e-01 1.36895999e-01 -4.56375033e-01 4.35799330e-01
6.43130094e-02 3.14261109e-01 -3.43926698e-01 8.48079562e-01
6.55009568e-01 -1.55120015e-01 2.16628000e-01 4.18343991e-01
-4.87496585e-01 -2.41489679e-01 -1.00480855e+00 9.62024033e-01
3.45853239e-01 9.47915673e-01 6.68144464e-01 -3.80653799e-01
5.27073383e-01 1.25818834e-01 7.81687558e-01 2.18061104e-01
3.87037426e-01 3.58274840e-02 2.77483791e-01 -5.14845014e-01
2.24855214e-01 4.43601668e-01 -3.38922888e-01 7.99743295e-01
3.33944112e-01 2.70172715e-01 5.23893416e-01 2.72600532e-01
1.66687584e+00 -2.77455151e-01 2.63807327e-01 -2.85122335e-01
2.43076891e-01 3.36622536e-01 3.86281639e-01 1.28064799e+00
-6.73113108e-01 1.35939345e-01 2.74030536e-01 -1.29806185e+00
-9.11418796e-01 -1.03301728e+00 2.64254689e-01 1.42222977e+00
4.98134941e-01 -2.51258671e-01 -1.01251781e+00 -5.05514681e-01
4.25955862e-01 -1.35482296e-01 -1.02770007e+00 -2.56616384e-01
-6.57255828e-01 -4.66381758e-01 1.11948597e+00 5.57095289e-01
5.58349490e-01 -1.47733724e+00 -1.01171899e+00 2.10812584e-01
1.54908031e-01 -9.02463853e-01 -5.66427350e-01 -1.19672315e-02
-3.98035914e-01 -1.21261179e+00 -1.12047352e-01 -7.89895535e-01
6.70065641e-01 4.27113384e-01 1.21398437e+00 9.49455142e-01
-4.95280713e-01 4.68084216e-01 -3.87961179e-01 -4.28605556e-01
-5.60398579e-01 3.44011188e-01 6.16044462e-01 -1.60076439e-01
6.90529644e-01 -7.25556493e-01 -4.46576208e-01 4.64328319e-01
-5.20599008e-01 -4.29977179e-01 1.95526242e-01 2.52174288e-01
-5.63059747e-02 -4.38134260e-02 2.05814987e-01 -2.94360489e-01
5.57558775e-01 -8.38531971e-01 -7.19361603e-01 3.19277383e-02
5.31151175e-01 -3.49310756e-01 7.33259320e-01 -7.53448009e-01
-1.36235893e-01 5.99080980e-01 2.33392175e-02 -3.92720312e-01
-5.67828059e-01 -6.57622293e-02 6.63517118e-01 -7.00831175e-01
6.62080824e-01 1.00170337e-01 1.54358849e-01 -1.32886499e-01
-8.17405954e-02 4.67853993e-01 9.22356129e-01 -1.71259835e-01
1.02329051e+00 7.39439845e-01 -1.01443864e-01 -1.09254110e+00
-5.28924406e-01 -4.01331604e-01 -1.21254027e+00 -7.91003585e-01
1.00423193e+00 -7.08293259e-01 -1.57412696e+00 1.16378677e+00
-1.05028379e+00 -6.17427349e-01 -1.49249017e-01 4.67618555e-01
-3.44589263e-01 -4.98119593e-02 -8.82027864e-01 -6.57159805e-01
-4.39482927e-01 -5.66154659e-01 9.59398329e-01 5.41973352e-01
-4.16069806e-01 -7.64662266e-01 6.64424896e-01 -4.59940016e-01
5.42499244e-01 2.64348388e-01 -7.38442391e-02 -1.03835714e+00
-6.44914091e-01 -4.02645707e-01 1.25630692e-01 -4.35212970e-01
8.37121829e-02 7.54763365e-01 -4.72459406e-01 -3.98987800e-01
-6.07786000e-01 -3.51593584e-01 7.33058453e-01 4.99239177e-01
7.40505457e-01 -5.27391613e-01 -8.21830273e-01 8.30750704e-01
1.08235991e+00 2.71322370e-01 5.40930824e-03 3.73386234e-01
5.97113967e-01 5.55794418e-01 5.97879514e-02 4.20587063e-01
4.38043833e-01 3.45728308e-01 5.32704055e-01 2.65236348e-01
1.83407888e-01 -1.42142415e-01 2.04876840e-01 1.85061932e-01
-4.96376485e-01 -4.90328133e-01 -1.17835939e+00 6.67263865e-01
-1.95406294e+00 -1.63785398e+00 -1.51582927e-01 1.89342558e+00
1.99288577e-01 -4.05049846e-02 8.84101152e-01 -4.34347898e-01
9.36780393e-01 -7.39219859e-02 -6.35123014e-01 -6.58676699e-02
1.08727813e-01 -4.71564353e-01 9.72121179e-01 2.87129164e-01
-1.60784411e+00 1.03446317e+00 8.72049618e+00 8.93279389e-02
-8.17209661e-01 -3.94736201e-01 1.13552988e-01 -1.59743831e-01
5.92877984e-01 -2.08140299e-01 -8.35084438e-01 5.27462661e-01
8.29456985e-01 -2.28228986e-01 9.92328584e-01 7.30671346e-01
-1.76526591e-01 -5.47264051e-03 -9.71961081e-01 6.27873719e-01
-6.36385381e-02 -1.46225393e+00 -1.62000269e-01 1.64290279e-01
4.52499241e-01 2.80895919e-01 -1.93668753e-01 2.82447934e-02
1.20279479e+00 -8.07312369e-01 9.73063648e-01 6.80891871e-01
3.23333710e-01 -6.23168766e-01 8.95557284e-01 5.04277468e-01
-1.41924310e+00 -3.77393425e-01 -1.97584957e-01 -5.01907051e-01
1.91201150e-01 -4.34733480e-01 -6.91620886e-01 -2.29439631e-01
1.16230059e+00 8.88941050e-01 -8.59273374e-01 1.18766844e+00
3.13008964e-01 5.88678658e-01 -8.65301847e-01 -4.27065730e-01
4.51299548e-01 -1.18214734e-01 9.71433818e-01 1.16148138e+00
3.56405050e-01 5.63208640e-01 2.88865000e-01 5.75203955e-01
-2.03351900e-01 -4.66659635e-01 -1.01071870e+00 -2.31544137e-01
9.36364055e-01 1.27906978e+00 -1.09097123e+00 -3.63564432e-01
7.15667680e-02 8.15892756e-01 5.80751121e-01 -1.76224336e-01
-8.38179648e-01 -7.70530626e-02 1.18903518e+00 1.47518709e-01
5.03080368e-01 -5.89589775e-01 3.22339177e-01 -9.11737502e-01
-5.40512562e-01 -6.61999762e-01 7.13033676e-01 -3.15560907e-01
-1.55878437e+00 7.10273743e-01 -4.46756072e-02 -1.22260880e+00
-7.91656300e-02 -5.03866255e-01 -8.08940589e-01 7.54108280e-02
-5.07485509e-01 -1.23824680e+00 -3.15396696e-01 3.05052996e-01
1.09689116e-01 -3.91274214e-01 5.19181311e-01 1.48424938e-01
-7.54343808e-01 3.92453045e-01 3.76822390e-02 5.37732840e-01
3.58383030e-01 -1.16692877e+00 1.05190432e+00 6.07574821e-01
3.14744599e-02 6.42104626e-01 8.67797673e-01 -8.68593931e-01
-1.31921268e+00 -9.11433518e-01 5.72582364e-01 -9.05591190e-01
1.00450134e+00 -5.17801523e-01 -7.31206000e-01 1.09612215e+00
1.42570287e-01 1.65376037e-01 5.95586300e-01 -2.74978280e-01
1.26327440e-01 4.70135473e-02 -1.16438770e+00 6.05748534e-01
1.52285898e+00 -2.06876084e-01 -5.00424445e-01 2.90682316e-01
3.56940478e-01 -4.31375653e-01 -8.18601727e-01 3.11325360e-02
9.43341434e-01 -9.34119880e-01 1.11280227e+00 -7.10563004e-01
7.88742825e-02 -4.95367885e-01 2.83855826e-01 -1.14439976e+00
-5.85597694e-01 -8.67866516e-01 3.58005799e-02 1.01526916e+00
1.09293848e-01 -4.94537711e-01 5.95837653e-01 2.69839674e-01
2.16217533e-01 -1.92293245e-02 -1.05162263e+00 -8.13504934e-01
-2.41043001e-01 2.84371316e-01 8.58728647e-01 7.62185037e-01
-1.14823848e-01 -4.24828194e-02 -6.57217383e-01 4.39121753e-01
7.70054698e-01 -4.97805141e-02 1.16028786e+00 -1.42384744e+00
1.32470429e-01 -6.43152773e-01 -9.36160266e-01 -7.06730366e-01
1.16867356e-01 -2.94757724e-01 2.31864393e-01 -1.02166820e+00
-2.10110601e-02 -2.98164841e-02 2.49139339e-01 5.46332836e-01
2.02751562e-01 5.79013109e-01 -9.30183101e-03 4.16061997e-01
-1.02601302e+00 1.62180141e-02 3.65518898e-01 -8.53747874e-02
-2.11437494e-01 1.60104141e-01 -9.26936790e-02 1.15045404e+00
6.50084853e-01 -9.65983391e-01 3.95254970e-01 -6.34074509e-01
5.04701547e-02 -5.50969504e-02 7.58585513e-01 -1.69172978e+00
6.94571078e-01 -6.67063072e-02 4.14789319e-01 -5.66647232e-01
1.10769674e-01 -8.98206830e-01 5.64963937e-01 8.48194778e-01
-6.41492844e-01 5.90899825e-01 1.72666416e-01 7.43633389e-01
1.89334244e-01 -1.64832458e-01 8.58350158e-01 -4.75419074e-01
-8.17991197e-01 5.88877261e-01 -1.08393288e+00 -8.89833868e-02
1.44687593e+00 -1.23834141e-01 -5.40815175e-01 -3.86018187e-01
-5.21966815e-01 5.06060541e-01 9.25559342e-01 3.33384156e-01
1.94692403e-01 -1.28776574e+00 -7.38343954e-01 2.66838133e-01
8.89182929e-03 -6.24149382e-01 2.90782273e-01 5.45709252e-01
-1.22856760e+00 -9.94116068e-02 -6.71203911e-01 -5.66437960e-01
-1.65156484e+00 5.80663145e-01 7.94714332e-01 1.51337326e-01
-5.25820136e-01 1.06238461e+00 -3.28632504e-01 -4.28080618e-01
7.59542212e-02 7.82205611e-02 -4.00262535e-01 -1.21564150e-01
8.35074902e-01 5.78617156e-01 -6.06931686e-01 -1.28862000e+00
-7.75020838e-01 7.70457923e-01 -4.85395528e-02 1.84110388e-01
1.38601685e+00 3.86069249e-03 -1.72363833e-01 2.63556451e-01
6.66362703e-01 -2.29546756e-01 -1.45145142e+00 2.12168530e-01
-1.18327916e-01 -4.33036536e-01 -3.52165818e-01 -3.08348984e-01
-8.90874267e-01 3.47055107e-01 5.39751172e-01 8.80500257e-01
5.59980690e-01 -6.94866702e-02 8.33296657e-01 7.68045485e-01
5.64169407e-01 -9.49615955e-01 -9.83437672e-02 5.73399425e-01
3.89131308e-01 -1.11904442e+00 4.23935987e-02 1.44397065e-01
-4.50214714e-01 1.15348709e+00 7.82245398e-01 -5.58611453e-01
6.11674130e-01 7.74115324e-01 2.37077549e-01 -5.95956385e-01
-8.05900931e-01 -2.53595889e-01 -2.17083111e-01 9.55919981e-01
-2.59969115e-01 2.26613949e-03 2.09068745e-01 1.92565218e-01
-2.40596280e-01 -4.39188043e-05 5.02548099e-01 1.00988472e+00
-6.68747783e-01 -3.98094505e-01 -3.27293217e-01 4.53565538e-01
-5.27584493e-01 2.94627428e-01 -1.30041540e+00 7.58021295e-01
3.13733995e-01 8.48978937e-01 4.71889079e-01 -6.93844974e-01
3.39524627e-01 -5.61273992e-01 2.07691327e-01 -3.36949468e-01
-1.18679595e+00 -5.73178113e-01 1.66064188e-01 -4.49693203e-01
-8.33191931e-01 -8.68678749e-01 -9.90509033e-01 -9.81793761e-01
-5.17452471e-02 1.48199379e-01 5.61334677e-02 7.07203567e-01
3.21969539e-01 -1.31886840e-01 5.63749075e-01 -1.33970928e+00
-2.01644450e-01 -7.40143776e-01 -6.00104332e-01 1.44412458e-01
6.83917344e-01 -6.31954968e-01 -6.77163780e-01 2.14155942e-01] | [7.856321811676025, -0.919897735118866] |
e5007608-a31f-47e2-ad03-2812b82328bc | framing-the-news-from-human-perception-to | 2304.14456 | null | https://arxiv.org/abs/2304.14456v1 | https://arxiv.org/pdf/2304.14456v1.pdf | Framing the News:From Human Perception to Large Language Model Inferences | Identifying the frames of news is important to understand the articles' vision, intention, message to be conveyed, and which aspects of the news are emphasized. Framing is a widely studied concept in journalism, and has emerged as a new topic in computing, with the potential to automate processes and facilitate the work of journalism professionals. In this paper, we study this issue with articles related to the Covid-19 anti-vaccine movement. First, to understand the perspectives used to treat this theme, we developed a protocol for human labeling of frames for 1786 headlines of No-Vax movement articles of European newspapers from 5 countries. Headlines are key units in the written press, and worth of analysis as many people only read headlines (or use them to guide their decision for further reading.) Second, considering advances in Natural Language Processing (NLP) with large language models, we investigated two approaches for frame inference of news headlines: first with a GPT-3.5 fine-tuning approach, and second with GPT-3.5 prompt-engineering. Our work contributes to the study and analysis of the performance that these models have to facilitate journalistic tasks like classification of frames, while understanding whether the models are able to replicate human perception in the identification of these frames. | ['Daniel Gatica-Perez', 'David Alonso del Barrio'] | 2023-04-27 | null | null | null | null | ['prompt-engineering'] | ['natural-language-processing'] | [ 3.03113014e-01 3.10347140e-01 -5.80002487e-01 -1.63949266e-01
-6.27693474e-01 -8.89910758e-01 1.16307902e+00 7.45551109e-01
-5.57586432e-01 6.60900831e-01 1.04311323e+00 -7.84533441e-01
-1.17603995e-01 -7.29409814e-01 -8.19552898e-01 -1.85094282e-01
2.75967866e-01 5.62400520e-01 2.59714663e-01 -4.17395324e-01
6.15503609e-01 1.76633790e-01 -1.25954056e+00 8.57446432e-01
2.61998951e-01 7.97027290e-01 2.38691166e-01 6.47286236e-01
-4.40988630e-01 1.40413451e+00 -7.34063864e-01 -4.80697066e-01
-2.15206906e-01 -4.33344781e-01 -9.80434299e-01 -4.26167324e-02
1.90704301e-01 1.17263801e-01 1.20206863e-01 7.65745044e-01
2.98380792e-01 -2.61316925e-01 4.86657262e-01 -5.73083937e-01
-3.20669532e-01 1.22690475e+00 -3.27714235e-01 8.02381456e-01
8.79558384e-01 -1.73561320e-01 1.04265487e+00 -5.78435242e-01
1.39772463e+00 1.65094578e+00 5.64436138e-01 -9.25473794e-02
-8.99593771e-01 -1.42901137e-01 6.34341538e-02 2.41551518e-01
-9.50997949e-01 -4.99785632e-01 4.89558965e-01 -1.04213488e+00
7.88317382e-01 5.83021522e-01 7.73101449e-01 1.36175931e+00
6.39843881e-01 4.04813915e-01 1.40916121e+00 -7.94176340e-01
1.47261456e-01 1.76863521e-01 4.72849548e-01 3.63968015e-01
8.09965730e-02 -1.98756546e-01 -6.69500887e-01 -2.97553539e-01
3.30489695e-01 -3.49109411e-01 -2.94299480e-02 6.21913731e-01
-1.53132260e+00 1.35383415e+00 -1.29087448e-01 6.72998190e-01
-6.15698099e-01 -1.09427646e-01 7.48764992e-01 2.14011341e-01
9.01533723e-01 4.77384895e-01 -1.80445254e-01 -4.61314291e-01
-8.34171534e-01 6.10105157e-01 1.28077209e+00 5.99821389e-01
1.00148775e-01 -4.70169634e-01 -5.50398707e-01 4.87658679e-01
3.44935060e-01 4.92416322e-01 1.19050905e-01 -7.15867341e-01
6.17827356e-01 4.69300002e-01 2.75507897e-01 -1.60764277e+00
-6.46823764e-01 -3.00924987e-01 -1.66993633e-01 -4.75598186e-01
5.40314734e-01 -3.46499801e-01 -4.18687403e-01 1.22069097e+00
2.15976477e-01 -4.38496202e-01 -2.66329229e-01 6.52408481e-01
7.00609267e-01 1.04287553e+00 2.45882407e-01 -5.41381955e-01
2.43203044e+00 -3.97923350e-01 -1.16152644e+00 -3.48475426e-01
5.41440308e-01 -1.60418332e+00 6.93794787e-01 2.86614299e-01
-1.01652050e+00 -4.18050408e-01 -8.11855078e-01 -2.31752738e-01
-4.40717489e-01 -4.13088910e-02 3.55526775e-01 4.33545172e-01
-5.68727434e-01 2.31933892e-01 -4.34063613e-01 -6.49902344e-01
3.63215268e-01 -3.20607364e-01 2.47271866e-01 2.15779856e-01
-1.51748645e+00 1.34085941e+00 3.13323587e-01 -2.61918962e-01
-4.10509795e-01 -6.47419691e-01 -6.50627792e-01 -1.85338899e-01
6.79457426e-01 -3.85901809e-01 1.43181181e+00 -9.05363262e-01
-1.22090805e+00 1.14993441e+00 -2.09881231e-01 -8.94904613e-01
5.35900056e-01 -4.17815298e-01 -7.25798428e-01 1.81836665e-01
5.12897730e-01 2.24001348e-01 8.62818658e-01 -5.98769546e-01
-7.81363249e-01 -1.82283390e-02 3.66804749e-01 -1.57338157e-01
2.16860652e-01 1.00911164e+00 3.14126611e-02 -8.55120957e-01
-3.07570964e-01 -9.10813630e-01 -2.68992223e-02 -5.85037649e-01
-3.52286696e-01 -2.96008021e-01 5.84775567e-01 -1.06031120e+00
1.73323953e+00 -1.98470342e+00 -1.19174987e-01 1.12191573e-01
3.05844456e-01 -1.40951753e-01 4.56162900e-01 8.41831505e-01
3.14010173e-01 2.22386077e-01 3.41189653e-01 3.82467419e-01
1.49868086e-01 6.76651448e-02 -7.81522930e-01 5.52721620e-01
-1.03471376e-01 5.65837204e-01 -8.74763310e-01 -5.16318798e-01
-3.12603787e-02 4.47332084e-01 -3.86780590e-01 -3.30267847e-01
-6.57436311e-01 2.79447228e-01 -5.84904134e-01 2.45276511e-01
9.87654105e-02 -4.25488502e-01 4.23868597e-01 -4.51515615e-01
-7.98431337e-01 7.52314746e-01 -8.99238884e-01 9.07762289e-01
-1.55399516e-01 1.13755238e+00 -1.39513373e-01 -6.25714660e-01
6.14993930e-01 3.95607263e-01 2.41766363e-01 -6.76310003e-01
5.20245135e-01 1.82299949e-02 -2.19748188e-02 -6.27380967e-01
7.50894964e-01 -1.08324170e-01 -3.47169250e-01 6.42709732e-01
-3.62866879e-01 2.39939913e-02 6.60009265e-01 2.05978096e-01
6.04041934e-01 -6.36133030e-02 6.27427101e-01 -6.00970209e-01
5.66962302e-01 4.76779073e-01 2.33613551e-01 7.72052705e-01
1.55936182e-01 3.39883029e-01 9.83140707e-01 -8.09072495e-01
-1.09754574e+00 -9.02562365e-02 -2.02612653e-01 1.25084007e+00
-2.91598290e-01 -7.09914327e-01 -8.87395203e-01 -2.84074157e-01
-3.70708257e-01 1.20075452e+00 -6.70718431e-01 5.78078628e-01
-7.49053180e-01 -7.42278695e-01 3.45961064e-01 -9.77307707e-02
-1.74084306e-02 -8.36357892e-01 -1.14857149e+00 7.23423898e-01
-6.08946919e-01 -1.52683640e+00 -3.43280941e-01 2.46885810e-02
-1.28299505e-01 -1.29025674e+00 -3.75910014e-01 -4.55772132e-01
2.99169749e-01 1.81468669e-02 1.25097859e+00 -7.02838525e-02
1.00915037e-01 3.30676734e-01 -6.52906060e-01 -1.10931194e+00
-1.14383888e+00 -6.23152591e-03 -2.08104253e-01 -3.93472575e-02
5.51678836e-01 1.41225502e-01 -1.59002289e-01 2.16760635e-01
-9.84117389e-01 2.99020529e-01 -1.69479265e-03 3.81997734e-01
1.62250131e-01 -1.04185373e-01 1.22141436e-01 -1.29216862e+00
9.86839235e-01 -5.43304145e-01 -6.18133307e-01 1.93638787e-01
-2.76641667e-01 -2.06939027e-01 2.87132084e-01 -1.27406970e-01
-9.26833272e-01 -6.01978600e-01 -2.05773234e-01 5.25263727e-01
6.15998618e-02 8.51141870e-01 5.06931305e-01 4.03177798e-01
9.65275228e-01 -3.46526235e-01 1.82453617e-02 -3.03531796e-01
4.43756163e-01 5.48086584e-01 8.30463693e-02 -4.88537669e-01
3.84705633e-01 5.64646244e-01 -3.47410887e-01 -9.80433047e-01
-1.31158698e+00 -4.63318110e-01 -1.87110037e-01 -5.85147798e-01
1.12233031e+00 -9.62981582e-01 -4.65409487e-01 3.33743952e-02
-1.54312611e+00 1.18852213e-01 -1.11234903e-01 7.55200446e-01
-3.75819594e-01 2.08977669e-01 -6.71859682e-01 -6.28912270e-01
-1.77949622e-01 -1.07071519e+00 6.71621203e-01 -7.94526190e-02
-9.17277336e-01 -1.06810260e+00 7.63005763e-02 4.65600342e-01
4.07036036e-01 4.61577654e-01 9.40673769e-01 -9.46191669e-01
-6.04429990e-02 -2.33041756e-02 -3.05247754e-02 -4.08275165e-02
-4.95673977e-02 2.57440776e-01 -6.88575745e-01 1.00368313e-01
4.52386618e-01 1.14210561e-01 3.62560034e-01 6.91311836e-01
4.36585158e-01 -8.32978129e-01 -3.41244310e-01 -2.88006663e-01
1.08932233e+00 2.93864608e-01 6.12875342e-01 8.47320855e-01
1.67557046e-01 8.58242035e-01 6.53854549e-01 7.31078744e-01
3.72596085e-01 7.83041120e-01 -2.11083591e-01 1.52818039e-01
-6.06134236e-02 -2.16853991e-01 6.04650974e-01 1.00309873e+00
-3.25701147e-01 -2.23975256e-01 -1.04683411e+00 2.22267061e-01
-1.67891765e+00 -1.35893500e+00 -5.43002248e-01 1.54004240e+00
7.96450496e-01 7.92922676e-01 2.21726701e-01 2.74043866e-02
9.80577707e-01 5.25891602e-01 5.38002193e-01 -6.28588855e-01
-1.86352998e-01 -2.10369546e-02 6.40469372e-01 6.38251364e-01
-1.09409320e+00 7.18046069e-01 6.60896397e+00 8.98875594e-01
-1.01368213e+00 2.99409747e-01 6.66455865e-01 3.63677800e-01
-3.61015469e-01 3.52899760e-01 -1.13180971e+00 6.53265536e-01
1.28495049e+00 -2.94699252e-01 1.19189993e-01 5.41317463e-01
9.52929854e-01 -3.38046372e-01 -9.21206176e-01 4.24044967e-01
2.17267111e-01 -2.05268931e+00 -2.65956614e-02 -1.21704172e-02
6.03463829e-01 -2.88777858e-01 -2.78686702e-01 2.63452232e-02
1.36939943e-01 -5.98192811e-01 1.32185960e+00 6.38968587e-01
1.95645571e-01 -3.80484045e-01 8.65949512e-01 2.66439945e-01
-6.48784101e-01 5.69518544e-02 -2.05632001e-01 -5.30059993e-01
7.19073713e-01 9.79526043e-01 -1.03576529e+00 3.97291541e-01
1.94824159e-01 4.68868524e-01 -4.75557566e-01 5.40039897e-01
-1.96941242e-01 9.16962862e-01 -1.80832312e-01 -4.03265327e-01
4.84269470e-01 3.03092450e-02 7.50996113e-01 1.54938507e+00
2.46772766e-01 2.28477865e-02 2.30847865e-01 4.84582514e-01
2.36942649e-01 2.79775709e-01 -3.22656900e-01 -4.58380997e-01
3.78963321e-01 9.52507675e-01 -1.32208025e+00 -5.34598768e-01
-4.41539973e-01 1.32562056e-01 -2.09318370e-01 2.76859347e-02
-1.00892997e+00 1.18237965e-01 6.18630759e-02 7.73158789e-01
2.29742289e-01 -1.35427043e-01 -9.81116965e-02 -7.03067243e-01
-4.32298928e-01 -1.45238614e+00 4.46037561e-01 -6.65581405e-01
-9.52420056e-01 6.19291425e-01 4.41426158e-01 -9.52848017e-01
-1.46839187e-01 -5.44875264e-01 -2.21805230e-01 4.97776031e-01
-1.08236325e+00 -9.67650771e-01 2.81834036e-01 2.65742466e-02
7.66140342e-01 1.19991161e-01 4.32768285e-01 1.96991608e-01
-8.08998495e-02 -3.53455216e-01 -1.46405056e-01 8.35156739e-02
6.62666917e-01 -6.62658691e-01 5.18971801e-01 6.79427445e-01
2.07379073e-01 5.76499879e-01 1.36631739e+00 -8.38990748e-01
-1.28525972e+00 -5.44736207e-01 1.68591726e+00 -6.00943565e-01
1.26015186e+00 -1.82052419e-01 -4.75095183e-01 5.77579319e-01
6.08740270e-01 -9.29847360e-01 7.17837214e-01 1.21070027e-01
-1.15185864e-01 3.44965518e-01 -7.35885918e-01 5.52738845e-01
4.84139949e-01 -4.82278824e-01 -1.27552843e+00 8.61043036e-01
8.51650596e-01 -6.53802872e-01 -7.67993152e-01 -1.44105107e-01
4.63911682e-01 -6.90244496e-01 7.95788407e-01 -5.29585004e-01
6.00271821e-01 -1.99604049e-01 -1.14898138e-01 -8.90719712e-01
-3.92601967e-01 -8.01172554e-01 1.58852890e-01 1.15337837e+00
5.07354081e-01 -6.05417430e-01 2.88780332e-02 2.79132813e-01
3.21298908e-03 -2.16562718e-01 -8.76037717e-01 -2.09560886e-01
-4.22512025e-01 -6.95581436e-01 2.33273536e-01 9.50312734e-01
3.46956141e-02 6.27600014e-01 -4.00722861e-01 -2.22558454e-01
5.32703847e-02 -3.61331813e-02 5.89505672e-01 -1.20934582e+00
-1.50124311e-01 -4.83336836e-01 -1.25423014e-01 -7.44694412e-01
-2.36546189e-01 -6.55806720e-01 -4.17987704e-01 -1.57877254e+00
1.15776202e-02 9.53037366e-02 3.41529787e-01 -9.85614285e-02
2.02261031e-01 -2.01138124e-01 4.54336613e-01 3.59619290e-01
-4.38866794e-01 -3.09302360e-01 1.06888330e+00 -9.24662352e-02
-5.88139221e-02 -8.07919260e-03 -8.09670269e-01 1.09575152e+00
4.82398212e-01 -7.56093681e-01 -4.89104092e-02 -1.22676738e-01
1.06464148e+00 -1.78460721e-02 3.66321564e-01 -7.21683681e-01
2.29766265e-01 -2.96976745e-01 2.04979807e-01 -6.90287173e-01
-2.46101409e-01 -7.18971789e-01 6.08292282e-01 7.17238963e-01
-6.35136783e-01 4.25434798e-01 4.54178676e-02 3.66903931e-01
-1.72466069e-01 -4.89203572e-01 4.35897887e-01 -5.04191101e-01
-5.08519173e-01 -4.78848040e-01 -1.06678033e+00 4.55918759e-01
8.85080457e-01 6.78561181e-02 -7.91329682e-01 -4.26223904e-01
-5.47115028e-01 -2.43447930e-01 1.04665995e-01 5.29129982e-01
6.41389489e-02 -9.33183491e-01 -1.07681048e+00 -3.08257490e-01
-2.34855771e-01 -6.37947559e-01 3.92313600e-02 1.18104410e+00
-1.07970178e+00 7.56874621e-01 1.59674268e-02 -4.33442086e-01
-1.03937876e+00 6.00430787e-01 -3.04490685e-01 -5.50643504e-01
-6.90501988e-01 3.85898292e-01 -1.19604841e-01 3.13580692e-01
-2.24344321e-02 -5.90459287e-01 -8.32200646e-01 9.54741180e-01
9.83115017e-01 4.91873235e-01 -1.32390112e-02 -1.00178087e+00
-3.48436862e-01 3.16922247e-01 -1.82428777e-01 -2.35186011e-01
1.29822493e+00 -2.83256680e-01 -3.29073995e-01 7.86057591e-01
8.16399693e-01 4.76046413e-01 -4.96630758e-01 -2.03674093e-01
4.08307314e-01 -2.96065569e-01 8.06849822e-03 -7.02649355e-01
-1.91733241e-01 1.88501209e-01 7.20135346e-02 1.04173768e+00
3.92506272e-01 1.63920134e-01 5.51626444e-01 2.21910942e-02
1.78527027e-01 -1.50155652e+00 -3.51705313e-01 7.54580796e-01
1.10905945e+00 -8.01701069e-01 5.50951362e-01 -4.59490061e-01
-6.39452636e-01 1.44245195e+00 -2.34403789e-01 3.50590199e-01
6.33443475e-01 1.56679392e-01 1.69919863e-01 -7.74565756e-01
-6.48734152e-01 1.17490299e-01 3.84777308e-01 -4.75872234e-02
7.27464139e-01 2.64562786e-01 -1.23941576e+00 4.51307803e-01
-4.19540077e-01 7.95369968e-03 7.76749432e-01 1.06270754e+00
-7.45751739e-01 -8.08714151e-01 -8.91343534e-01 3.58567685e-01
-1.10775816e+00 -2.05767676e-01 -2.97523022e-01 7.84691632e-01
1.37748361e-01 1.03922427e+00 1.69261582e-02 -4.98585626e-02
3.47896934e-01 1.06580034e-01 1.81082010e-01 -5.36493838e-01
-8.77657652e-01 4.16850984e-01 9.04702365e-01 -1.64520741e-01
-1.04849279e+00 -7.95599222e-01 -7.97280848e-01 -4.64733362e-01
-1.69895645e-02 4.58500236e-01 7.01384127e-01 1.28644812e+00
1.46948531e-01 5.92876315e-01 3.23822886e-01 -5.02528846e-01
-2.87899494e-01 -9.08994555e-01 -1.19605571e-01 2.17701912e-01
7.42744356e-02 -5.13215244e-01 -3.63352180e-01 4.74511653e-01] | [8.967133522033691, 9.768510818481445] |
724eaa2a-072b-418b-9cf3-849cdece2f80 | segflow-joint-learning-for-video-object | 1709.06750 | null | http://arxiv.org/abs/1709.06750v1 | http://arxiv.org/pdf/1709.06750v1.pdf | SegFlow: Joint Learning for Video Object Segmentation and Optical Flow | This paper proposes an end-to-end trainable network, SegFlow, for
simultaneously predicting pixel-wise object segmentation and optical flow in
videos. The proposed SegFlow has two branches where useful information of
object segmentation and optical flow is propagated bidirectionally in a unified
framework. The segmentation branch is based on a fully convolutional network,
which has been proved effective in image segmentation task, and the optical
flow branch takes advantage of the FlowNet model. The unified framework is
trained iteratively offline to learn a generic notion, and fine-tuned online
for specific objects. Extensive experiments on both the video object
segmentation and optical flow datasets demonstrate that introducing optical
flow improves the performance of segmentation and vice versa, against the
state-of-the-art algorithms. | ['Ming-Hsuan Yang', 'Yi-Hsuan Tsai', 'Shengjin Wang', 'Jingchun Cheng'] | 2017-09-20 | segflow-joint-learning-for-video-object-1 | http://openaccess.thecvf.com/content_iccv_2017/html/Cheng_SegFlow_Joint_Learning_ICCV_2017_paper.html | http://openaccess.thecvf.com/content_ICCV_2017/papers/Cheng_SegFlow_Joint_Learning_ICCV_2017_paper.pdf | iccv-2017-10 | ['unsupervised-video-object-segmentation'] | ['computer-vision'] | [ 3.17573957e-02 -2.35219523e-01 -4.79570508e-01 -3.68627548e-01
-1.86300501e-02 -5.07662416e-01 2.13931680e-01 -4.94817615e-01
-4.09130573e-01 6.80439472e-01 1.31383482e-02 -1.80611417e-01
4.90678139e-02 -7.67522454e-01 -7.06736147e-01 -5.33182144e-01
-2.54638016e-01 2.69605458e-01 5.87115347e-01 1.52602181e-01
3.32063407e-01 5.59961319e-01 -1.16614902e+00 -7.45693548e-03
9.89985526e-01 1.22271836e+00 1.54616743e-01 9.89625216e-01
-5.21235049e-01 1.27876019e+00 -1.93655238e-01 -2.76867479e-01
5.97229898e-01 -3.83086711e-01 -1.36013615e+00 4.16171104e-01
8.03182006e-01 -7.70161927e-01 -6.54674768e-01 1.05322647e+00
5.63905649e-02 1.85623214e-01 2.44266257e-01 -1.42386246e+00
-6.79232061e-01 4.73806262e-01 -3.84573728e-01 6.40349209e-01
9.84894559e-02 5.12118757e-01 8.79394114e-01 -4.73646611e-01
7.26664245e-01 1.31144953e+00 3.34698230e-01 5.24459541e-01
-9.49572265e-01 -4.39257562e-01 3.77310514e-01 2.06503987e-01
-8.93693149e-01 -1.38946608e-01 7.00260699e-01 -7.33568132e-01
6.11170530e-01 -1.08782962e-01 8.93302321e-01 5.31137228e-01
8.27260688e-02 1.18116510e+00 6.69367075e-01 1.09624662e-01
-5.83695397e-02 -3.02171379e-01 3.95449281e-01 9.84451592e-01
1.37381107e-01 2.29477003e-01 -3.03880602e-01 5.02133787e-01
1.26790440e+00 -3.36971506e-02 -5.16341567e-01 -2.88918465e-01
-1.23873663e+00 5.32809436e-01 8.07096541e-01 1.44429475e-01
-2.51132488e-01 5.04911482e-01 4.65051264e-01 1.58071239e-02
4.54314888e-01 1.41970411e-01 -5.95626771e-01 -2.08215594e-01
-1.15549147e+00 1.56434506e-01 8.25606406e-01 9.75317717e-01
1.06200302e+00 1.41724184e-01 -4.10821140e-01 4.76842612e-01
4.99452353e-01 2.43298262e-01 3.08101296e-01 -1.40592587e+00
4.13937598e-01 5.76179743e-01 1.38124824e-01 -8.90256166e-01
-3.61874700e-01 -3.80670637e-01 -5.96602380e-01 5.94400987e-02
8.27254891e-01 -2.75431186e-01 -1.23716474e+00 1.49791968e+00
3.89619142e-01 8.37710917e-01 -1.80570886e-01 1.31711471e+00
8.83395195e-01 7.18264759e-01 2.44580150e-01 -1.72809824e-01
9.08356130e-01 -1.73422039e+00 -7.35611975e-01 -1.60596281e-01
3.79496068e-01 -6.40722454e-01 5.48299909e-01 1.82213590e-01
-1.23260796e+00 -9.42521751e-01 -8.64648461e-01 -4.18997943e-01
-1.19936921e-01 -1.44693941e-01 8.28690469e-01 5.81195176e-01
-1.17863345e+00 6.30153835e-01 -9.92599487e-01 -7.84354806e-02
1.04756916e+00 3.89286786e-01 -1.71097994e-01 -7.80438725e-03
-9.18271780e-01 4.73699033e-01 7.06015885e-01 4.71754670e-01
-1.18756533e+00 -8.43953907e-01 -8.15167069e-01 4.28694254e-03
3.01933497e-01 -1.15671003e+00 1.17270839e+00 -1.24978328e+00
-1.79697573e+00 6.63512886e-01 -2.60696054e-01 -5.78679800e-01
1.01208413e+00 -4.18061227e-01 3.33601348e-02 6.46723568e-01
1.41621783e-01 1.10069907e+00 7.15817511e-01 -1.05544043e+00
-1.02048802e+00 -1.07743531e-01 3.72155070e-01 -5.36727794e-02
-1.83507815e-01 -1.19534492e-01 -8.10461044e-01 -5.54259241e-01
-6.48171604e-02 -6.38133645e-01 -2.89814919e-01 3.50146770e-01
-4.95562285e-01 -2.25639626e-01 1.29950452e+00 -6.72628224e-01
1.15111160e+00 -1.66935062e+00 1.07226007e-01 -1.37104005e-01
2.35200986e-01 5.62723577e-01 -2.40144446e-01 -1.01862982e-01
1.29279882e-01 1.11751445e-01 -3.88366282e-01 -1.77566126e-01
-3.99257064e-01 3.25602531e-01 -8.47716630e-02 5.10691106e-01
3.91035825e-01 1.30563807e+00 -1.19093382e+00 -7.61177838e-01
5.31414926e-01 3.58456165e-01 -7.24543035e-01 4.00015086e-01
-3.30109775e-01 9.02986884e-01 -5.65271616e-01 7.17495501e-01
9.51563716e-01 -3.87178391e-01 -2.03346550e-01 -3.56125712e-01
-2.98229992e-01 3.79509814e-02 -1.07191324e+00 1.89378536e+00
-1.13603592e-01 6.34333849e-01 1.14061303e-01 -1.14559746e+00
6.38602674e-01 7.20561817e-02 1.01057184e+00 -3.94724786e-01
4.02482897e-01 1.91204578e-01 -7.04970658e-02 -8.15926492e-01
4.51051176e-01 2.31795087e-01 5.43522656e-01 2.72295475e-01
4.92675573e-01 1.53773889e-01 7.49825299e-01 7.46527240e-02
6.11312747e-01 6.37994885e-01 -1.44105256e-01 -3.78810227e-01
1.09053254e+00 5.77806421e-02 7.55329072e-01 6.53365314e-01
-7.83570826e-01 5.37064493e-01 3.62464458e-01 -7.33483553e-01
-6.85821772e-01 -1.03142285e+00 -1.80252299e-01 9.07782793e-01
8.80192339e-01 -5.00964113e-02 -8.72260153e-01 -9.82652783e-01
1.03295021e-01 5.64549714e-02 -4.05283153e-01 1.65051579e-01
-9.18737888e-01 -4.26909178e-01 2.26430610e-01 8.01531494e-01
1.14389360e+00 -1.14023054e+00 -7.22924590e-01 3.74466389e-01
-4.16487008e-01 -1.62506568e+00 -7.08552361e-01 -4.12715912e-01
-1.09460437e+00 -1.34880447e+00 -8.07013333e-01 -9.00358260e-01
5.77000022e-01 4.22396988e-01 1.02725041e+00 2.75773674e-01
-3.05965602e-01 3.92895252e-01 -1.22976914e-01 -5.50158834e-03
-2.05623090e-01 7.79505894e-02 -6.18241072e-01 4.00969028e-01
5.50290048e-02 -4.33060706e-01 -1.05930901e+00 3.14322293e-01
-9.83756185e-01 3.57104093e-02 2.07618892e-01 5.96526921e-01
4.55139816e-01 -4.50492799e-01 3.20485175e-01 -8.88905644e-01
8.80638286e-02 -2.82907546e-01 -7.62036264e-01 1.73453867e-01
-2.94903159e-01 -1.49373353e-01 5.07178545e-01 2.30622999e-02
-1.07227802e+00 2.94054300e-01 -6.09521605e-02 -7.36748159e-01
-1.26426652e-01 1.79708943e-01 7.18951449e-02 -4.54196513e-01
1.03997640e-01 1.56094000e-01 5.06947599e-02 -2.42846012e-01
5.92364788e-01 4.57106620e-01 8.43055189e-01 -4.73668456e-01
6.69302583e-01 7.98269033e-01 2.49800626e-02 -5.65420508e-01
-9.34987068e-01 -7.77032673e-01 -1.05319595e+00 -6.06364429e-01
1.20683026e+00 -7.69504070e-01 -8.27006221e-01 9.43891585e-01
-1.37123132e+00 -4.70347196e-01 -2.71222115e-01 5.46118379e-01
-7.50940561e-01 4.26757455e-01 -8.30980182e-01 -5.03570378e-01
-3.34383190e-01 -1.39096904e+00 9.09103274e-01 8.03990483e-01
4.49980885e-01 -1.34176576e+00 -1.65120721e-01 6.00990295e-01
2.73117572e-01 3.34981024e-01 3.61639738e-01 -2.77583063e-01
-1.37077558e+00 1.70638427e-01 -6.83564425e-01 6.35622263e-01
1.93856418e-01 3.63950133e-01 -9.86690700e-01 -1.60034344e-01
-2.79827774e-01 -2.51344860e-01 1.14850605e+00 7.07625151e-01
1.38345563e+00 -1.28750011e-01 -3.05213839e-01 1.04196310e+00
1.58930159e+00 2.16743320e-01 6.47810578e-01 1.37313664e-01
1.10362458e+00 3.48491371e-01 3.71429682e-01 1.43102124e-01
5.60832202e-01 3.61055017e-01 7.08967507e-01 -2.23290384e-01
-4.05458480e-01 -9.46760327e-02 1.07266642e-01 6.31673396e-01
-2.47325599e-01 -2.41042018e-01 -7.09482014e-01 6.78486228e-01
-2.00208330e+00 -9.99084890e-01 -2.50304163e-01 1.71733963e+00
5.36141574e-01 1.66743323e-01 3.26445878e-01 -1.49657667e-01
8.11212242e-01 3.74287426e-01 -6.22009456e-01 -2.44689763e-01
1.06320344e-01 1.49899870e-01 6.71415091e-01 7.00845003e-01
-1.40449655e+00 1.20404708e+00 6.57519054e+00 3.21005791e-01
-1.35982299e+00 4.67886850e-02 7.48345554e-01 3.03566933e-01
6.11343570e-02 9.63742957e-02 -6.55052841e-01 4.61626798e-01
3.75878155e-01 -6.57415614e-02 3.92844200e-01 6.03991687e-01
4.10265744e-01 -1.04425229e-01 -9.71890450e-01 7.91006148e-01
-1.98281333e-01 -1.63746750e+00 1.02412388e-01 -1.51118562e-01
9.60330188e-01 1.69995531e-01 -1.82011977e-01 -1.03751253e-02
1.03952661e-01 -6.48845792e-01 8.16684186e-01 5.54710209e-01
3.15704077e-01 -4.34864759e-01 6.38733208e-01 1.34829432e-01
-1.33867383e+00 -1.73518881e-01 -8.33475813e-02 -1.82459466e-02
6.39029026e-01 4.35473233e-01 -3.88356119e-01 6.67018831e-01
6.81389749e-01 1.43534327e+00 -4.60378170e-01 1.59908187e+00
-2.44859055e-01 6.29355669e-01 -8.01570714e-03 3.70028973e-01
7.78210759e-01 -6.39841497e-01 5.55614233e-01 1.30848575e+00
-2.11685225e-01 -1.25802025e-01 6.53381824e-01 1.20239437e+00
-1.88596800e-01 -1.36283293e-01 -6.43418357e-02 -2.58914921e-02
1.16043344e-01 1.53044069e+00 -9.92717206e-01 -5.22429228e-01
-4.24806148e-01 8.13118875e-01 3.17340977e-02 6.61046565e-01
-9.50280130e-01 -3.18882048e-01 6.89187944e-01 -1.85099691e-01
5.79609036e-01 -2.87204593e-01 -2.85099804e-01 -1.22368288e+00
-1.54333487e-01 -1.06703691e-01 2.33985260e-01 -6.18248165e-01
-1.18575418e+00 3.85805815e-01 -1.62011206e-01 -1.12573135e+00
-8.86924639e-02 -7.72544265e-01 -7.30523229e-01 7.26438820e-01
-2.13781953e+00 -1.09576237e+00 -6.22977316e-01 6.56397581e-01
6.98850632e-01 1.99170664e-01 8.01661387e-02 5.43016195e-01
-8.45735312e-01 2.62808412e-01 -1.23689637e-01 6.39356792e-01
3.28869879e-01 -1.21941328e+00 2.33246490e-01 1.17619693e+00
3.00702509e-02 2.74168402e-01 2.10711569e-01 -4.70942855e-01
-1.11710715e+00 -1.47591078e+00 4.22709852e-01 -4.48154844e-02
5.89981318e-01 8.49059448e-02 -8.11014533e-01 6.55488849e-01
4.42338258e-01 7.95668006e-01 1.62146032e-01 -5.73695898e-01
-1.04905374e-01 -2.13973492e-01 -1.07850659e+00 2.28522032e-01
1.23506248e+00 -2.72874296e-01 -3.11192840e-01 3.93257082e-01
8.12375307e-01 -6.55507207e-01 -8.31585824e-01 3.71691346e-01
4.92275149e-01 -1.13197613e+00 1.06660366e+00 -8.10301483e-01
6.73670709e-01 -5.62635481e-01 3.07990223e-01 -9.80888247e-01
-2.55202085e-01 -8.63259554e-01 -3.16588491e-01 1.03804481e+00
2.74264868e-02 -5.86887300e-01 8.95385861e-01 3.65732104e-01
-3.31909865e-01 -7.02603281e-01 -5.77049792e-01 -6.61214948e-01
-3.10066827e-02 -2.95489818e-01 3.55173647e-01 7.99347520e-01
-6.07987463e-01 3.57951038e-02 -5.66688851e-02 3.58655974e-02
6.18127167e-01 2.99388289e-01 7.01998770e-01 -1.04296708e+00
-2.70720664e-02 -7.56428659e-01 -7.15799570e-01 -1.70718241e+00
3.90298665e-01 -9.44301248e-01 2.67729759e-02 -1.70398390e+00
2.01501921e-02 -2.19497457e-01 -4.18330520e-01 2.32880607e-01
-4.23117131e-01 3.02174687e-01 6.59902334e-01 1.84922457e-01
-7.27235556e-01 3.50150257e-01 2.10091019e+00 -3.47175181e-01
-3.43729496e-01 9.73740518e-02 -3.20038199e-01 7.20454693e-01
5.20062983e-01 -1.44642740e-01 -4.45305228e-01 -6.39985263e-01
-5.01653314e-01 1.57532245e-01 5.01370668e-01 -1.01475465e+00
2.98479259e-01 -3.02549094e-01 1.94752738e-01 -6.45194173e-01
-1.07266665e-01 -6.85137570e-01 -4.83189434e-01 6.57187402e-01
-3.24397027e-01 -2.78172642e-01 1.01279445e-01 5.72493315e-01
-4.93536770e-01 -1.52773663e-01 9.13356245e-01 -2.46960551e-01
-1.05566895e+00 8.85465682e-01 -1.66572724e-02 4.70613867e-01
1.06107485e+00 -5.13199985e-01 -3.05946738e-01 -8.93667638e-02
-6.22136116e-01 5.21987617e-01 1.70292351e-02 5.79529941e-01
4.20760334e-01 -1.10287702e+00 -5.92336595e-01 1.71303257e-01
-3.63258749e-01 2.75426626e-01 1.11638717e-01 1.04833639e+00
-1.10107684e+00 5.16938448e-01 -5.73758960e-01 -1.00018001e+00
-5.64011753e-01 3.91007334e-01 7.56072402e-01 -5.51147982e-02
-4.26157236e-01 1.09728396e+00 3.55950415e-01 -8.20928290e-02
2.61168510e-01 -5.91114998e-01 -2.59622931e-01 -1.64002433e-01
4.51912194e-01 7.83401251e-01 -3.67435098e-01 -8.28468382e-01
-2.37518281e-01 8.84717822e-01 1.81532074e-02 2.56341398e-01
1.02110255e+00 -4.10781145e-01 -3.29566896e-01 1.85782626e-01
1.39953983e+00 -6.13746703e-01 -1.90728188e+00 -1.58813633e-02
-1.01875767e-01 -7.62860060e-01 2.31366634e-01 -6.37930930e-01
-1.94343567e+00 1.01047623e+00 3.91482145e-01 3.15311491e-01
1.19435453e+00 -4.31862146e-01 1.32309186e+00 6.43521966e-03
2.25475639e-01 -8.49704027e-01 6.52781799e-02 6.81264520e-01
3.46095592e-01 -1.28770423e+00 -1.99931562e-01 -7.69641876e-01
-4.02383417e-01 1.52213788e+00 9.09649253e-01 -4.55053568e-01
7.70291865e-01 1.17359571e-02 8.49016979e-02 1.82263941e-01
-3.30314457e-01 -2.96612769e-01 3.63997459e-01 3.49025279e-01
3.08371514e-01 -2.57089466e-01 -3.37287188e-01 2.01575160e-02
2.69986331e-01 7.11335063e-01 4.68249500e-01 7.27016509e-01
-4.43970382e-01 -8.56908679e-01 -4.14223559e-02 2.52392441e-01
-4.74395961e-01 2.91504115e-01 -5.10072447e-02 7.39182353e-01
3.84511411e-01 8.12181115e-01 3.36551338e-01 -4.32848968e-02
1.37969837e-01 -2.07718030e-01 6.00911677e-01 -2.19470650e-01
-6.93988740e-01 3.58899869e-02 -2.92687535e-01 -9.85648274e-01
-1.11378586e+00 -3.63339782e-01 -1.53130543e+00 -1.12586223e-01
-2.66491890e-01 -1.97433606e-01 3.74237478e-01 1.12900066e+00
2.21894354e-01 6.84244156e-01 7.55094409e-01 -9.68456030e-01
-1.29258573e-01 -6.32945657e-01 -3.82481635e-01 5.70784271e-01
6.75346076e-01 -5.05774558e-01 -1.32610694e-01 5.98629892e-01] | [9.137460708618164, -0.24043810367584229] |
e68cfc4d-09ee-456d-abb2-e8aefd716783 | improved-neural-relation-detection-for | 1704.06194 | null | http://arxiv.org/abs/1704.06194v2 | http://arxiv.org/pdf/1704.06194v2.pdf | Improved Neural Relation Detection for Knowledge Base Question Answering | Relation detection is a core component for many NLP applications including
Knowledge Base Question Answering (KBQA). In this paper, we propose a
hierarchical recurrent neural network enhanced by residual learning that
detects KB relations given an input question. Our method uses deep residual
bidirectional LSTMs to compare questions and relation names via different
hierarchies of abstraction. Additionally, we propose a simple KBQA system that
integrates entity linking and our proposed relation detector to enable one
enhance another. Experimental results evidence that our approach achieves not
only outstanding relation detection performance, but more importantly, it helps
our KBQA system to achieve state-of-the-art accuracy for both single-relation
(SimpleQuestions) and multi-relation (WebQSP) QA benchmarks. | ['Bo-Wen Zhou', 'Bing Xiang', 'Mo Yu', 'Cicero dos Santos', 'Wenpeng Yin', 'Kazi Saidul Hasan'] | 2017-04-20 | improved-neural-relation-detection-for-1 | https://aclanthology.org/P17-1053 | https://aclanthology.org/P17-1053.pdf | acl-2017-7 | ['knowledge-base-question-answering'] | ['natural-language-processing'] | [-1.33041620e-01 6.40592515e-01 -2.81219631e-01 -2.36046761e-01
-1.18485248e+00 -3.68090749e-01 4.57457066e-01 2.98801929e-01
-3.54333103e-01 1.14886129e+00 2.30316967e-01 -7.99003899e-01
-3.15729558e-01 -1.33088708e+00 -8.86500597e-01 -1.74070641e-01
1.75041720e-01 8.99888754e-01 8.12269986e-01 -8.25798631e-01
-3.61100286e-01 4.30264324e-01 -1.00104582e+00 7.94131815e-01
1.10911953e+00 1.17826080e+00 -3.59752566e-01 7.60145426e-01
-4.51936245e-01 1.80183303e+00 -6.95580959e-01 -1.11951244e+00
-3.06060165e-01 -1.85922399e-01 -1.92348003e+00 -1.02847373e+00
6.09860420e-01 -1.15399010e-01 -7.09401488e-01 8.74411523e-01
4.88243520e-01 2.58964330e-01 2.54530549e-01 -1.01849079e+00
-1.08703756e+00 9.42942142e-01 -8.48834664e-02 6.11433387e-01
7.41087317e-01 -2.81393707e-01 1.69905555e+00 -5.12987316e-01
7.18282104e-01 1.30876112e+00 5.79400659e-01 4.02709305e-01
-1.01510286e+00 -5.54011345e-01 -1.27024963e-01 9.25162435e-01
-1.26409435e+00 -3.58560592e-01 3.54163557e-01 -8.36396497e-03
1.85758042e+00 4.58867460e-01 2.10674152e-01 8.17527950e-01
1.66703269e-01 6.90720737e-01 8.01440239e-01 -3.79516095e-01
-1.65927708e-01 7.62265101e-02 8.88949752e-01 9.60588515e-01
1.95944503e-01 -4.15770143e-01 -5.88585079e-01 -3.19781393e-01
5.39685845e-01 -6.21844709e-01 -5.11448324e-01 2.49371380e-02
-1.04402351e+00 7.87627995e-01 9.48299527e-01 3.79800797e-01
-4.54907537e-01 1.59293354e-01 3.72444719e-01 5.97639978e-01
2.30257034e-01 6.92629516e-01 -8.56528282e-01 1.88299641e-01
-2.05232278e-01 1.54041484e-01 1.26073372e+00 9.31627452e-01
7.60749817e-01 -3.78744423e-01 -6.00384176e-01 8.49266648e-01
7.81753287e-02 2.52065569e-01 3.66730601e-01 -7.88699389e-01
9.28876400e-01 1.10819125e+00 -1.70422032e-01 -1.13385880e+00
-6.00941718e-01 -3.57407898e-01 -8.67894948e-01 -5.60625732e-01
1.17654592e-01 1.60383850e-01 -8.26425135e-01 1.51810002e+00
4.66659069e-01 -8.73817969e-03 7.13429034e-01 7.00765967e-01
1.94378316e+00 6.55441701e-01 3.01936895e-01 -2.08212249e-02
1.83195245e+00 -1.26775706e+00 -1.03276062e+00 -3.22454572e-02
8.35854113e-01 -2.23433167e-01 9.03315902e-01 -2.98420250e-01
-1.07824588e+00 -3.98968369e-01 -7.76033044e-01 -8.99122417e-01
-8.02956879e-01 -1.17670201e-01 7.60365605e-01 2.87223041e-01
-1.16056490e+00 1.46469221e-01 -4.22695100e-01 -5.23996651e-01
2.53796011e-01 4.79969323e-01 -4.49666500e-01 2.48379916e-01
-2.23691535e+00 1.47633278e+00 7.30045140e-01 3.05326670e-01
-4.60793585e-01 -5.55704296e-01 -9.38579440e-01 4.41526204e-01
6.57733917e-01 -1.25742030e+00 1.39485276e+00 2.22758595e-02
-1.44781697e+00 7.70259321e-01 -5.04959822e-01 -9.45589423e-01
-1.00993708e-01 -5.71701169e-01 -7.16126442e-01 2.46241599e-01
1.02852963e-01 4.28126782e-01 3.60617191e-02 -1.08504713e+00
-6.03141785e-01 -2.04582572e-01 4.92893159e-01 3.71302933e-01
-2.79188808e-02 6.51041865e-02 -3.87712598e-01 -2.16322392e-02
9.71602574e-02 -3.46175581e-01 3.71003263e-02 -6.95379317e-01
-8.30105245e-01 -9.69368100e-01 5.60335159e-01 -1.14138484e+00
1.24545765e+00 -1.38972163e+00 4.41179611e-02 1.95884109e-01
4.81575787e-01 4.99644309e-01 -6.28656000e-02 4.60435390e-01
1.67416688e-02 2.54285019e-02 -1.23607457e-01 2.51283973e-01
-9.78011414e-02 3.49615604e-01 -6.63825870e-01 -3.34233999e-01
6.59810603e-01 1.66839480e+00 -8.51280510e-01 -8.06766033e-01
-3.77854735e-01 4.50770259e-01 -4.60726991e-02 3.64546984e-01
-4.03089941e-01 -1.53330213e-03 -3.72794122e-01 6.62689567e-01
3.04271579e-01 -7.53584564e-01 2.16572523e-01 -5.21094263e-01
4.88848865e-01 1.19654036e+00 -8.28103840e-01 1.37380373e+00
-4.40966725e-01 5.93292296e-01 -3.66317272e-01 -7.67653406e-01
1.04966938e+00 6.11984074e-01 -2.22831547e-01 -9.61458445e-01
-1.20026998e-01 2.59737104e-01 1.66157290e-01 -6.60245299e-01
8.29991102e-01 1.02822758e-01 2.80734301e-01 9.46777239e-02
5.30693412e-01 1.29163474e-01 2.14040235e-01 6.45686269e-01
1.47485995e+00 -7.93621615e-02 5.36797106e-01 4.36348543e-02
9.41043139e-01 3.55840698e-02 3.22212845e-01 7.37560987e-01
-3.89083549e-02 1.05731713e-03 6.93849087e-01 -3.36825281e-01
-3.34240496e-01 -1.14924300e+00 4.46812510e-02 1.27449238e+00
8.48063752e-02 -1.90597102e-01 -3.35326254e-01 -1.07978344e+00
2.91532725e-02 6.63655043e-01 -5.78327000e-01 -2.89424598e-01
-8.43361735e-01 -4.48832422e-01 1.05324018e+00 6.63261712e-01
8.01285386e-01 -1.56873131e+00 -1.83356717e-01 3.39777619e-01
-7.36855328e-01 -1.51029146e+00 1.32319838e-01 3.14490348e-01
-6.15186572e-01 -1.23140752e+00 -2.85576075e-01 -1.17463934e+00
3.67571227e-02 -4.59207892e-02 1.76479483e+00 1.23291872e-02
1.71709329e-01 2.37198934e-01 -2.33674824e-01 5.31560779e-02
-2.60958374e-01 7.05420613e-01 -3.31142724e-01 -4.28696960e-01
7.77025044e-01 -3.37363631e-01 -3.12787861e-01 3.36540788e-02
-5.43624461e-01 -2.44025365e-01 7.43304789e-01 7.84617424e-01
3.64728719e-01 -2.63340652e-01 1.08020234e+00 -9.70040619e-01
1.12617075e+00 -4.76778179e-01 -3.78713191e-01 1.04366469e+00
-4.50440437e-01 2.32796341e-01 3.26914102e-01 -7.26726875e-02
-1.45121729e+00 -5.59626579e-01 -3.71260345e-01 1.85039014e-01
-1.74703579e-02 7.68039823e-01 -2.19543815e-01 -1.45015597e-01
7.49316037e-01 2.68645119e-02 -6.55992985e-01 -3.76004785e-01
8.87869239e-01 5.25084198e-01 8.76257598e-01 -5.82780063e-01
5.30274034e-01 8.75987038e-02 -1.01803787e-01 -3.43381464e-01
-1.21869206e+00 -5.74577510e-01 -6.53768778e-01 2.83662647e-01
9.49037075e-01 -8.38949382e-01 -1.53337383e+00 -7.22459331e-02
-1.67699552e+00 -2.04187348e-01 -2.93124229e-01 1.19558826e-01
-7.34008700e-02 1.18014567e-01 -1.30163944e+00 -6.68868482e-01
-1.07280970e+00 -7.09814310e-01 8.44187081e-01 4.17924821e-01
-1.33973792e-01 -9.51507747e-01 4.44616586e-01 9.33659554e-01
4.17170852e-01 -1.94743305e-01 1.23515654e+00 -1.02985525e+00
-7.73858070e-01 1.00117274e-01 -8.01767170e-01 6.62867026e-03
1.39616057e-01 -4.33919072e-01 -1.00336957e+00 2.47645348e-01
-4.35852826e-01 -7.63112962e-01 1.05870855e+00 -1.86667770e-01
6.01435065e-01 -5.07771552e-01 -3.22456241e-01 2.77581245e-01
1.09945416e+00 2.09617987e-01 7.86101460e-01 6.35038912e-01
1.02657378e+00 8.20657015e-01 2.78292686e-01 -5.27218878e-01
1.11827886e+00 5.23211002e-01 1.51530355e-01 -2.06698164e-01
-3.23949069e-01 -2.24500969e-01 -9.19524580e-02 1.06530106e+00
3.99084724e-02 -1.49578691e-01 -1.25445902e+00 7.78070807e-01
-2.00560379e+00 -7.76358962e-01 -5.40103555e-01 1.49997056e+00
1.33986378e+00 -5.20832986e-02 -2.58639663e-01 -1.30227327e-01
6.49684906e-01 -6.49747998e-02 -4.23760891e-01 -6.57656550e-01
-4.31495488e-01 7.70568848e-01 2.50662833e-01 7.58235633e-01
-1.01032472e+00 1.48382103e+00 5.94731188e+00 6.54708326e-01
-5.51224172e-01 1.87431902e-01 5.73634326e-01 5.27775347e-01
-2.82521665e-01 -2.34901115e-01 -1.03618586e+00 -2.14525759e-01
1.02423322e+00 2.10571866e-02 1.05584122e-01 5.13971269e-01
-6.85781181e-01 8.07088539e-02 -9.82675612e-01 5.32591581e-01
-4.90636751e-03 -1.52760231e+00 4.87336338e-01 -4.71832931e-01
3.90138000e-01 8.74937102e-02 -2.70423919e-01 8.31783891e-01
8.38648021e-01 -1.13319516e+00 -2.03626886e-01 6.24722600e-01
3.95590276e-01 -6.72208130e-01 1.22054374e+00 3.39176543e-02
-1.32612014e+00 1.87954098e-01 -4.27810609e-01 -1.63859632e-02
2.15042397e-01 4.37150300e-01 -1.19026160e+00 1.15175903e+00
9.53158379e-01 1.09114051e-01 -6.49683654e-01 6.93612993e-01
-8.19644332e-01 6.55897915e-01 -1.21351138e-01 -1.68170482e-01
1.37143850e-01 4.33015190e-02 1.20991684e-01 1.30524862e+00
-2.14093670e-01 4.52284127e-01 -2.52676010e-01 8.62635016e-01
-7.16727197e-01 2.91021556e-01 -3.32030892e-01 1.11854888e-01
6.45050764e-01 1.30135655e+00 -1.70281947e-01 -7.36077785e-01
-3.12741399e-01 8.17565978e-01 1.20276415e+00 6.29284739e-01
-5.54833412e-01 -7.41048217e-01 5.65598845e-01 -6.93716228e-01
2.45593458e-01 2.28780597e-01 -9.71648693e-02 -1.22603810e+00
1.65712029e-01 -7.51100957e-01 1.06714964e+00 -8.10792685e-01
-1.49148583e+00 1.05704331e+00 -4.11938012e-01 -2.91547090e-01
-3.16495121e-01 -4.21308666e-01 -3.35503906e-01 9.84555304e-01
-2.04514003e+00 -1.42685914e+00 -2.88991839e-01 7.83137500e-01
-8.71417001e-02 2.07889695e-02 1.06282651e+00 4.49534386e-01
-6.86735332e-01 5.97542048e-01 -4.43782389e-01 6.82518065e-01
7.07723022e-01 -1.38930750e+00 7.39904225e-01 5.33916116e-01
5.94811618e-01 9.79830563e-01 3.55645031e-01 -6.98120296e-01
-1.07796681e+00 -1.03347981e+00 1.42638671e+00 -6.54825509e-01
9.53157604e-01 -6.41746894e-02 -1.51681685e+00 1.06031263e+00
5.87669969e-01 -3.11367959e-02 6.97274566e-01 6.09858036e-01
-7.90301681e-01 -2.33138129e-01 -1.03552926e+00 4.56757367e-01
9.48730171e-01 -1.02458644e+00 -1.28012705e+00 3.80184829e-01
1.59422076e+00 -5.81441581e-01 -1.34645474e+00 9.50476587e-01
2.68648475e-01 -6.13473177e-01 1.07176900e+00 -1.20709777e+00
3.19886416e-01 -2.45359272e-01 2.99832989e-02 -1.04719198e+00
-4.08661544e-01 -2.22268209e-01 -9.67284977e-01 1.41429079e+00
8.84793282e-01 -1.01715243e+00 8.22108924e-01 5.49845636e-01
5.51975556e-02 -9.46269691e-01 -1.13493240e+00 -5.69586694e-01
4.92295772e-02 -1.24161720e-01 6.73127651e-01 1.20778155e+00
2.43171304e-01 1.22922170e+00 2.61221603e-02 5.70611775e-01
1.83415741e-01 3.43898475e-01 5.52409410e-01 -1.10329688e+00
-2.72275269e-01 -3.53945822e-01 -2.80174822e-01 -1.16519415e+00
3.83783996e-01 -8.48340392e-01 -1.44577935e-01 -2.24234676e+00
1.15401305e-01 -1.96421310e-01 -5.93238950e-01 6.99271739e-01
-5.56821048e-01 1.47641420e-01 -1.79283470e-01 -7.86838215e-03
-9.75942016e-01 5.45284450e-01 9.43320692e-01 -3.52743149e-01
-2.70821571e-01 -2.72442847e-01 -6.62720561e-01 3.42956811e-01
6.63545430e-01 -2.88813204e-01 -1.83089122e-01 -5.38107574e-01
6.84479475e-01 1.92492440e-01 3.00501466e-01 -7.18258619e-01
6.25894070e-01 3.24434072e-01 9.83024538e-02 -6.70057774e-01
3.86874169e-01 -3.94606769e-01 -4.37185794e-01 4.03633863e-01
-6.64171994e-01 2.86322739e-02 2.41526455e-01 3.40924561e-01
-6.01815701e-01 1.23600006e-01 1.65270090e-01 -1.51976589e-02
-6.94989085e-01 1.81902163e-02 9.41418950e-03 4.50484157e-01
4.37400401e-01 5.35506189e-01 -1.30532861e+00 -5.21374881e-01
-6.63077593e-01 7.01587677e-01 -4.92211938e-01 3.50733459e-01
6.83636963e-01 -1.26028264e+00 -7.23241031e-01 -3.85978967e-01
2.18127221e-01 3.08172166e-01 4.42730412e-02 7.24555016e-01
-5.94197452e-01 1.09511030e+00 1.54725388e-01 -1.89484850e-01
-1.35100961e+00 4.92862493e-01 4.75510567e-01 -9.60628092e-01
-3.52767915e-01 1.46203780e+00 -1.42209735e-02 -9.62860525e-01
2.92896181e-01 -5.53850234e-01 -7.11746752e-01 9.56030861e-02
4.02230203e-01 3.24390233e-01 4.26473260e-01 -5.21791279e-01
-4.77340221e-01 1.33250564e-01 -3.65312487e-01 2.01168016e-01
9.28660214e-01 1.00327417e-01 -6.57338917e-01 3.29391420e-01
1.02261353e+00 -2.04728588e-01 -1.06603138e-01 -8.02478492e-01
6.18596077e-01 2.25181267e-01 -3.36093813e-01 -1.29285085e+00
-6.93107605e-01 8.49449754e-01 2.78531760e-01 4.79049176e-01
7.96261370e-01 4.72678721e-01 1.44795799e+00 1.26928008e+00
1.96375325e-01 -5.60817122e-01 -1.12885088e-01 8.82802367e-01
8.71751606e-01 -1.24392903e+00 -1.88124970e-01 -5.73362589e-01
-5.23361921e-01 8.52536798e-01 1.00677717e+00 1.62280545e-01
2.26974621e-01 -1.16517521e-01 3.28329116e-01 -6.57604396e-01
-1.02109051e+00 -7.92567194e-01 7.73255050e-01 3.75334352e-01
5.65357924e-01 -7.27635697e-02 -2.50324309e-01 4.78879780e-01
-3.73408467e-01 -1.75144479e-01 7.79625028e-02 6.11876905e-01
-4.20550734e-01 -8.40794146e-01 -7.75473714e-02 5.42369664e-01
-4.67584044e-01 -6.58443987e-01 -8.45634997e-01 7.82257795e-01
-2.28917509e-01 9.81649876e-01 -1.38749585e-01 -4.38226223e-01
5.98326445e-01 4.72794414e-01 2.56407529e-01 -5.25793850e-01
-9.90470886e-01 -7.88676500e-01 7.44662642e-01 -6.45918906e-01
-4.24425781e-01 9.96661484e-02 -1.38357508e+00 -2.45241541e-02
-6.87310398e-01 5.90446770e-01 1.07692741e-01 1.13480878e+00
5.78512371e-01 9.13269639e-01 -1.35502562e-01 4.37289119e-01
-9.16658863e-02 -1.28787458e+00 -1.02975339e-01 2.51007169e-01
4.13050920e-01 -4.75168318e-01 9.88363251e-02 -3.60837668e-01] | [10.544983863830566, 7.958098411560059] |
a09a0a66-2463-44c1-ac7d-0ba786612e38 | hybrid-rnn-at-semeval-2019-task-9-blending | null | null | https://aclanthology.org/S19-2210 | https://aclanthology.org/S19-2210.pdf | Hybrid RNN at SemEval-2019 Task 9: Blending Information Sources for Domain-Independent Suggestion Mining | Social media has an increasing amount of information that both customers and companies can benefit from. These social media posts can include Tweets or be in the form of vocalization of complements and complaints (e.g., reviews) of a product or service. Researchers have been actively mining this invaluable information source to automatically generate insights. Mining sentiments of customer reviews is an example that has gained momentum due to its potential to gather information that customers are not happy about. Instead of reading millions of reviews, companies prefer sentiment analysis to obtain feedback and to improve their products or services. In this work, we aim to identify information that companies can act on, or other customers can utilize for making their own experience better. This is different from identifying if reviews of a product or service is negative, positive, or neutral. To that end, we classify sentences of a given review as suggestion or not suggestion so that readers of the reviews do not have to go through thousands of reviews but instead can focus on actionable items and applicable suggestions. To identify suggestions within reviews, we employ a hybrid approach that utilizes a recurrent neural network (RNN) along with rule-based features to build a domain-independent suggestion mining model. In this way, a model trained on electronics reviews is used to extract suggestions from hotel reviews. | ['Aysu Ezen-Can', 'Ethem F. Can'] | 2019-06-01 | null | null | null | semeval-2019-6 | ['suggestion-mining'] | ['natural-language-processing'] | [ 5.07569164e-02 3.48875731e-01 -4.81942832e-01 -7.45280325e-01
-5.60609460e-01 -5.13072908e-01 5.23991346e-01 6.04983807e-01
-4.10237968e-01 6.53431237e-01 5.36110580e-01 -6.72631800e-01
3.27565938e-01 -1.02413237e+00 -2.34278128e-01 -3.95580947e-01
4.81853962e-01 -4.02292609e-03 -2.31985509e-01 -7.62865663e-01
8.12802970e-01 2.53341943e-01 -1.44706452e+00 5.67266107e-01
6.51861191e-01 9.59201813e-01 3.67063344e-01 3.53214651e-01
-7.16954887e-01 9.08141553e-01 -6.74368203e-01 -5.25992334e-01
-1.68344416e-02 -3.65922540e-01 -5.22294641e-01 3.27597022e-01
-3.40766221e-01 -2.74247713e-02 6.27929211e-01 9.42050040e-01
1.33188561e-01 1.03042386e-01 3.07260633e-01 -6.96955264e-01
-6.67194963e-01 9.43012893e-01 -5.52801132e-01 1.22269571e-01
4.10923451e-01 -2.04306930e-01 1.33063495e+00 -1.20242071e+00
5.10674715e-01 8.02900791e-01 8.42747465e-02 2.46970922e-01
-6.61944032e-01 -5.72782516e-01 4.59901839e-01 -5.83595298e-02
-7.83677697e-01 -3.48699003e-01 1.02282512e+00 -2.57837415e-01
1.00182474e+00 3.49432826e-01 7.27097750e-01 8.53997827e-01
4.25636292e-01 5.70978582e-01 8.19634557e-01 -4.63534266e-01
3.79160076e-01 1.17394972e+00 1.98295012e-01 2.54059881e-02
1.44528776e-01 -7.76756048e-01 -5.03307283e-01 -7.75516927e-02
-6.53447434e-02 4.59330529e-01 -5.01980167e-03 4.33984011e-01
-7.05145359e-01 1.43852496e+00 4.00677383e-01 4.26640421e-01
-9.10785556e-01 -5.58810592e-01 5.22552729e-01 3.85720164e-01
8.80925477e-01 7.25267291e-01 -6.71653986e-01 -1.63690537e-01
-4.50159520e-01 3.18821929e-02 9.68806148e-01 4.07799751e-01
9.90457594e-01 -7.35582784e-03 3.61011535e-01 1.09213674e+00
4.95929003e-01 3.20699126e-01 9.29723024e-01 -1.69304296e-01
3.91566545e-01 1.10615551e+00 1.95906729e-01 -1.54713976e+00
-3.39675516e-01 -5.92141211e-01 -5.89746594e-01 -1.68959409e-01
-4.51780736e-01 -5.54117560e-01 -5.11699259e-01 1.05690253e+00
2.11994246e-01 -5.77801585e-01 1.96992040e-01 8.42380106e-01
1.00898659e+00 9.59512174e-01 -2.69951165e-01 -5.59063673e-01
1.23725617e+00 -8.77193272e-01 -8.05579364e-01 -5.49138248e-01
7.35887587e-01 -1.00117683e+00 9.23903465e-01 8.06893110e-01
-7.90158272e-01 -5.82327545e-01 -9.65393782e-01 3.17879677e-01
-7.31947422e-01 2.17977110e-02 7.40168631e-01 4.46274340e-01
-7.17193246e-01 5.76684654e-01 -3.22391599e-01 -3.52475733e-01
8.57752487e-02 3.24564487e-01 -8.55248645e-02 2.08950758e-01
-1.22267473e+00 8.25279057e-01 -1.23221621e-01 1.78152919e-01
2.42420897e-01 -1.98969632e-01 -8.92172754e-01 -2.09502995e-01
3.26539308e-01 -2.38065228e-01 1.24665856e+00 -1.51579988e+00
-1.44283068e+00 4.30243492e-01 -4.01604712e-01 -4.43346173e-01
-2.22969636e-01 -2.78763890e-01 -8.53461921e-01 -6.30925968e-02
3.45831871e-01 3.35982621e-01 1.00804925e+00 -8.01566184e-01
-9.71973956e-01 -3.88858289e-01 3.57928649e-02 1.83532625e-01
-6.05210364e-01 5.50588608e-01 -2.28734557e-02 -3.43381792e-01
-1.81886251e-03 -9.22520757e-01 -4.85453784e-01 -9.39704895e-01
-6.90553069e-01 -4.78925377e-01 8.34631503e-01 -5.02545595e-01
1.23437893e+00 -1.95365393e+00 -4.62663025e-01 4.96892363e-01
1.09950518e-02 1.89930052e-01 9.48318988e-02 7.80508339e-01
6.15235344e-02 6.74814641e-01 2.63668448e-01 -1.77279249e-01
-2.80367225e-01 4.68998551e-02 -5.96100748e-01 5.87295629e-02
5.99765658e-01 5.77564359e-01 -8.32810819e-01 -8.34069997e-02
3.90445925e-02 4.18411285e-01 -4.17084903e-01 1.19156085e-01
-2.19097324e-02 4.51469630e-01 -7.83085406e-01 3.97494972e-01
3.45677704e-01 -4.37325478e-01 -9.45808738e-03 -2.05847491e-02
-5.40927470e-01 1.12533534e+00 -7.34851241e-01 7.93648183e-01
-8.94430220e-01 5.89095771e-01 -1.88991591e-01 -1.06070614e+00
1.43005145e+00 2.01664209e-01 3.67917746e-01 -6.26664758e-01
3.70711476e-01 1.41674027e-01 -8.20251225e-05 -3.25758070e-01
9.84955072e-01 -2.23791927e-01 -1.69635981e-01 8.50943029e-01
-6.13946974e-01 -2.38067761e-01 3.20951909e-01 3.79471540e-01
9.53299582e-01 -4.25241768e-01 5.20890415e-01 2.56398231e-01
6.25603080e-01 8.08816683e-03 2.80846477e-01 3.22831124e-01
3.73955548e-01 3.74262393e-01 3.66621345e-01 -2.89356738e-01
-7.41858959e-01 -2.43704438e-01 9.82587561e-02 1.12441325e+00
-2.40628883e-01 -6.63368404e-01 -6.10634312e-02 -5.38171768e-01
-2.55526781e-01 8.09091866e-01 -5.68462968e-01 -4.34458405e-02
-1.29933089e-01 -5.53933263e-01 -5.35318911e-01 2.18625754e-01
1.18066199e-01 -1.30758178e+00 -4.11663651e-01 4.19856191e-01
-1.38100371e-01 -7.48008668e-01 -4.05597955e-01 4.44218546e-01
-6.42137408e-01 -6.16560876e-01 -3.13161492e-01 -6.11240804e-01
8.31057131e-01 6.47016287e-01 9.48947847e-01 -7.58506656e-02
2.97611266e-01 1.34743631e-01 -1.09821796e+00 -9.02846694e-01
-3.95779788e-01 2.31355458e-01 1.74676880e-01 4.66784477e-01
7.39660978e-01 -4.98335689e-01 -4.81993824e-01 4.15250510e-01
-7.73251116e-01 -1.17883608e-01 6.56686902e-01 5.48658609e-01
3.19174945e-01 6.74230754e-02 1.40833533e+00 -1.35462058e+00
1.28326917e+00 -9.33885217e-01 -7.72416145e-02 -3.99885297e-01
-8.70135248e-01 -2.47500777e-01 8.55052948e-01 -3.36705893e-01
-8.74991059e-01 3.34983170e-02 -4.10833716e-01 2.40460515e-01
-4.25518528e-02 1.27316248e+00 3.63326550e-01 3.76620322e-01
6.68994427e-01 -2.54734177e-02 6.84798742e-03 -1.63647622e-01
1.46658972e-01 1.21086872e+00 -1.32230535e-01 2.60963976e-01
5.51241994e-01 2.64055043e-01 -6.56364083e-01 -7.72596717e-01
-1.18402886e+00 -9.37823415e-01 -2.28924945e-01 -4.30599213e-01
4.94046569e-01 -6.82426691e-01 -5.46353579e-01 -2.55245626e-01
-1.00893199e+00 3.20224971e-01 -3.17096114e-01 5.49634159e-01
1.40399978e-01 -3.53366509e-02 -4.48418587e-01 -1.07392955e+00
-5.95136762e-01 -9.14785564e-01 4.73914057e-01 5.40897906e-01
-7.24109590e-01 -8.05799723e-01 -8.92193615e-02 5.21508932e-01
6.82222247e-01 -8.34193751e-02 4.22607362e-01 -1.17368364e+00
-7.77710089e-03 -8.58926237e-01 -4.20530997e-02 7.35452950e-01
6.79287136e-01 7.41740614e-02 -7.47921526e-01 1.18021078e-01
3.96543056e-01 -3.79168987e-01 5.33996642e-01 3.77973557e-01
7.65817881e-01 -7.48180389e-01 2.78683249e-02 -3.50178391e-01
9.33099449e-01 3.76161247e-01 6.07951224e-01 5.53254843e-01
3.04213941e-01 9.79521155e-01 1.06352699e+00 8.37686718e-01
3.99701416e-01 1.87753350e-01 5.69365740e-01 -9.26967487e-02
8.85246933e-01 -5.58320470e-02 7.90355146e-01 1.15377402e+00
9.32614133e-02 -1.49562120e-01 -4.37264532e-01 5.56490183e-01
-1.58976293e+00 -8.45651448e-01 -3.65410209e-01 1.90520561e+00
7.34272778e-01 4.88177091e-01 1.36753917e-01 2.36992821e-01
6.67882681e-01 1.35708630e-01 -5.18388510e-01 -1.35699582e+00
9.16472375e-02 1.93444088e-01 3.23529989e-01 1.19533487e-01
-6.66322112e-01 7.49727368e-01 4.60457563e+00 4.43968564e-01
-1.48884511e+00 -2.43098974e-01 1.07420850e+00 2.19063926e-03
-6.15464032e-01 -1.40527382e-01 -1.18065608e+00 2.74398416e-01
1.15990019e+00 -1.80385455e-01 7.74919763e-02 1.15756404e+00
8.88372004e-01 -3.30522120e-01 -7.14373708e-01 5.82901120e-01
2.35266283e-01 -1.37376010e+00 -9.54288766e-02 1.21597216e-01
7.78465867e-01 4.87682074e-02 2.60458767e-01 2.49497429e-01
1.29189461e-01 -7.42581308e-01 4.72350776e-01 3.48946214e-01
9.80251357e-02 -9.63489950e-01 1.24952149e+00 5.93547165e-01
-9.63988781e-01 -5.62956482e-02 -3.92960370e-01 -7.12812424e-01
4.48547781e-01 9.37374532e-01 -1.34345353e+00 3.82031024e-01
5.40789783e-01 1.19885373e+00 -4.32962924e-01 5.22357941e-01
-4.05248940e-01 7.53381610e-01 -8.62006024e-02 -7.75990963e-01
3.50993544e-01 -4.65254813e-01 1.08322157e-02 1.13494945e+00
3.91586125e-01 7.13076368e-02 1.36172429e-01 5.47490001e-01
-1.84505016e-01 8.94237518e-01 -8.56163621e-01 -5.88996232e-01
1.91450074e-01 1.78647709e+00 -8.62072706e-01 -1.63315222e-01
-6.43568277e-01 6.42240107e-01 2.68164184e-02 8.28555897e-02
-3.63277495e-01 -5.09840429e-01 3.50903034e-01 3.01497072e-01
3.46319199e-01 1.56892866e-01 -2.63107985e-01 -7.71990299e-01
1.27937436e-01 -7.84119368e-01 -1.43575370e-01 -7.97932208e-01
-1.33847070e+00 8.16134632e-01 -7.76323199e-01 -1.34549046e+00
-4.04919565e-01 -2.66384095e-01 -1.00993776e+00 7.95216024e-01
-1.58562112e+00 -7.47605741e-01 7.56551996e-02 1.41851544e-01
8.74798775e-01 -9.61292759e-02 5.26691496e-01 1.31097049e-01
-2.94009387e-01 1.37925327e-01 -2.88297713e-01 -1.15595937e-01
8.76473486e-01 -8.87870371e-01 1.95544511e-01 4.69550937e-01
1.67807974e-02 9.50348973e-01 7.96154022e-01 -7.14889586e-01
-1.25776982e+00 -1.05975723e+00 1.37776506e+00 -2.42128253e-01
8.74974966e-01 1.84683148e-02 -6.09449923e-01 3.83812279e-01
2.68225610e-01 -5.82952976e-01 1.20665264e+00 3.17375958e-01
1.68530256e-01 -2.39103213e-01 -8.54182482e-01 5.48934281e-01
2.28104278e-01 -3.20772707e-01 -4.64590967e-01 4.09843951e-01
8.17193747e-01 1.80696189e-01 -5.57058454e-01 8.62644799e-03
3.99072707e-01 -1.04777718e+00 4.01157200e-01 -4.51374382e-01
8.88183475e-01 -1.46312505e-01 1.58049781e-02 -1.63047910e+00
8.06198865e-02 -5.19500375e-01 4.83591586e-01 1.46780515e+00
1.11951292e+00 -8.52800071e-01 7.42974699e-01 7.12324679e-01
-3.80322367e-01 -1.15044630e+00 -6.05772845e-02 -6.18231036e-02
-5.14217079e-01 -8.22335660e-01 4.13974792e-01 8.44251692e-01
6.04586422e-01 8.83338571e-01 -4.86418098e-01 -2.62641639e-01
-3.05276036e-01 2.06767187e-01 1.00131750e+00 -1.09344411e+00
-1.69327408e-02 -3.74610722e-01 6.72437474e-02 -1.10875106e+00
-6.32865354e-02 -6.77049637e-01 3.63699906e-02 -1.71249521e+00
-8.93684775e-02 -3.30019861e-01 -2.07071453e-01 2.48010918e-01
9.73542929e-02 2.80810028e-01 -9.55015123e-02 2.06438184e-01
-5.17776310e-01 2.06057459e-01 1.16644454e+00 -1.07267268e-01
-7.16786087e-01 5.89259684e-01 -1.67986441e+00 8.16723466e-01
1.08932722e+00 -4.04217511e-01 -3.32895219e-01 3.42783988e-01
1.35652876e+00 -1.23421006e-01 -3.08569223e-01 -2.49476671e-01
1.86423138e-01 -2.16678292e-01 1.14121124e-01 -8.98706079e-01
1.74293026e-01 -5.59364200e-01 -2.51219004e-01 1.61837563e-01
-5.25713027e-01 2.93938011e-01 -1.27001271e-01 4.70778942e-01
-5.87524533e-01 -4.58274692e-01 2.44519800e-01 -1.98301747e-01
-2.66748726e-01 -7.49063417e-02 -1.00173855e+00 -5.94370663e-01
6.57751143e-01 -2.21478716e-01 -1.75290138e-01 -1.08152795e+00
-4.73498225e-01 1.60779327e-01 1.71808571e-01 6.64956868e-01
9.06150520e-01 -9.45956528e-01 -6.34632468e-01 5.93041023e-03
1.96771726e-01 -1.97395757e-01 -4.33428362e-02 8.27036858e-01
1.09112514e-02 5.44762492e-01 2.51386732e-01 -1.52784064e-01
-1.09189200e+00 3.82722437e-01 -4.17751908e-01 -4.01159644e-01
-1.42974928e-01 8.65111828e-01 -1.34795129e-01 -3.56614977e-01
-1.25752121e-01 -4.14948434e-01 -1.13824224e+00 6.43760622e-01
8.60152900e-01 -5.32696806e-02 2.28224665e-01 -6.03622139e-01
-1.76598862e-01 2.20652342e-01 -5.05744576e-01 -9.54700727e-03
1.61505640e+00 -2.22061485e-01 -3.08311403e-01 8.41458738e-01
1.12186193e+00 5.53926647e-01 -4.14033264e-01 2.99575794e-02
-9.55365133e-03 -2.56890267e-01 1.36273280e-02 -6.36348724e-01
-1.12403965e+00 5.84688663e-01 2.29808427e-02 1.09547436e+00
9.71277058e-01 2.77013838e-01 9.90137100e-01 5.48518956e-01
1.05933249e-01 -1.37191522e+00 2.59414196e-01 4.86613929e-01
9.73455489e-01 -1.66900492e+00 2.49751564e-02 -2.08678111e-01
-1.10782456e+00 1.23803639e+00 3.50215971e-01 -1.95677027e-01
8.22600722e-01 -1.22061453e-03 1.57499701e-01 -3.45469832e-01
-9.18661535e-01 -9.70552117e-02 1.68842703e-01 9.29083824e-02
8.33926082e-01 -4.30456959e-02 -7.60919213e-01 9.14282203e-01
-3.86329800e-01 -3.24689537e-01 9.30996001e-01 8.97084892e-01
-8.40565979e-01 -1.23163283e+00 -1.66979268e-01 1.11948764e+00
-7.60149181e-01 -2.75828034e-01 -7.40253091e-01 1.33858055e-01
-1.55308023e-01 1.64782274e+00 -2.10588828e-01 -7.49976516e-01
1.18927754e-01 -1.07769489e-01 -6.47735953e-01 -1.26831615e+00
-9.66134846e-01 2.43522733e-01 5.98679960e-01 -1.04813308e-01
-6.89250469e-01 -7.54432082e-01 -1.30395293e+00 -1.75559089e-01
-8.72303247e-01 7.03802109e-01 1.30050743e+00 1.12412071e+00
3.17274839e-01 5.16954780e-01 1.30513608e+00 -7.60522604e-01
-6.74970523e-02 -1.08539665e+00 -4.54045892e-01 1.50917515e-01
1.71811417e-01 -1.71684667e-01 -5.70407569e-01 -5.63304611e-02] | [11.137450218200684, 6.793140411376953] |
86f9cddd-612f-4102-8103-b805d04c5162 | punctuation-restoration-in-spanish-customer | 2205.13961 | null | https://arxiv.org/abs/2205.13961v1 | https://arxiv.org/pdf/2205.13961v1.pdf | Punctuation Restoration in Spanish Customer Support Transcripts using Transfer Learning | Automatic Speech Recognition (ASR) systems typically produce unpunctuated transcripts that have poor readability. In addition, building a punctuation restoration system is challenging for low-resource languages, especially for domain-specific applications. In this paper, we propose a Spanish punctuation restoration system designed for a real-time customer support transcription service. To address the data sparsity of Spanish transcripts in the customer support domain, we introduce two transfer-learning-based strategies: 1) domain adaptation using out-of-domain Spanish text data; 2) cross-lingual transfer learning leveraging in-domain English transcript data. Our experiment results show that these strategies improve the accuracy of the Spanish punctuation restoration system. | ['Simon Corston-Oliver', 'Tere Roldán', 'David Rossouw', 'Shayna Gardiner', 'Xiliang Zhu'] | 2022-05-27 | null | https://aclanthology.org/2022.deeplo-1.9 | https://aclanthology.org/2022.deeplo-1.9.pdf | deeplo-2022-7 | ['punctuation-restoration'] | ['natural-language-processing'] | [ 3.34760517e-01 -5.80354854e-02 -1.80750847e-01 -6.81684911e-01
-1.60010171e+00 -4.58237231e-01 1.65371865e-01 -1.36393681e-01
-3.09753180e-01 7.26212025e-01 5.53896606e-01 -5.99701226e-01
3.79729807e-01 -3.03016990e-01 -6.34894550e-01 -4.15842652e-01
5.82622826e-01 5.83724439e-01 -2.22213447e-01 -3.84798437e-01
-3.88093814e-02 3.28132510e-01 -8.67835224e-01 8.06090474e-01
1.19713449e+00 7.76741207e-01 5.79367578e-01 4.34257627e-01
-4.46545660e-01 8.30160201e-01 -8.77796531e-01 -2.51231104e-01
1.29422620e-01 -5.31001270e-01 -7.79668212e-01 5.10325372e-01
1.05644748e-01 -1.11667201e-01 -4.21111494e-01 8.94994438e-01
5.35476446e-01 2.18370929e-01 5.11246383e-01 -7.55980372e-01
-9.28113282e-01 9.18239355e-01 -2.86124617e-01 1.68247089e-01
3.14574778e-01 -1.44192949e-01 6.80573583e-01 -1.25292683e+00
4.72306460e-01 1.23399794e+00 5.52761495e-01 7.65313864e-01
-1.06354129e+00 -7.33034968e-01 1.44507915e-01 3.61307800e-01
-1.49596202e+00 -7.82315433e-01 8.86393607e-01 -8.55468959e-02
1.19800782e+00 -1.60401344e-01 -1.58661857e-01 1.43857110e+00
-1.96344733e-01 1.15451479e+00 1.05596852e+00 -6.82669699e-01
2.46996894e-01 1.91576988e-01 -2.21101105e-01 1.34294212e-01
-4.22862589e-01 -2.41716236e-01 -8.20666373e-01 9.27346200e-02
6.07481420e-01 -1.25603333e-01 -1.57905579e-01 1.80440843e-01
-8.57335746e-01 7.59812474e-01 -1.32950485e-01 4.41208482e-01
-4.46098208e-01 -6.32695735e-01 6.45674407e-01 5.95658123e-01
7.03898847e-01 -4.46489491e-02 -9.14719045e-01 -4.42742616e-01
-1.09794784e+00 -3.21103036e-01 7.16252029e-01 1.36367738e+00
8.67672384e-01 4.42301780e-01 3.45148295e-02 1.60989797e+00
4.89524677e-02 6.64129555e-01 1.11631930e+00 -3.71246934e-01
1.04917490e+00 2.12506041e-01 1.79664826e-03 -2.01547101e-01
-2.13386659e-02 -2.25700110e-01 -6.79443777e-01 -5.36896288e-01
1.87286854e-01 -3.49887013e-01 -9.77175832e-01 1.39975274e+00
-6.00492097e-02 8.40674415e-02 6.20504022e-01 7.52534568e-01
5.93834162e-01 9.29609835e-01 2.16329172e-02 -2.49911442e-01
1.08064330e+00 -1.23252189e+00 -1.05394161e+00 -5.08682311e-01
6.96976244e-01 -1.08692276e+00 1.42119908e+00 5.02691925e-01
-8.57952476e-01 -5.09858608e-01 -8.53618681e-01 -1.89047530e-01
-2.21763045e-01 6.54133260e-01 -1.77014276e-01 5.81689179e-01
-8.34587455e-01 4.10137624e-01 -7.65526712e-01 -5.09122014e-01
-1.31372642e-02 9.64293182e-02 -4.10096824e-01 -4.68484312e-01
-1.04205966e+00 7.60170221e-01 2.60888845e-01 -7.25228339e-02
-7.11649060e-01 -4.59334850e-01 -9.28049028e-01 2.39411891e-01
3.38823557e-01 1.08964428e-01 1.77170038e+00 -1.34110510e+00
-1.81947637e+00 9.02287602e-01 -5.54555655e-01 -3.76375467e-01
2.26343274e-01 -3.04095775e-01 -9.98973608e-01 -5.76425828e-02
4.78980727e-02 7.99937323e-02 8.08484256e-01 -1.11661184e+00
-7.56202459e-01 -5.55592775e-01 -8.14070702e-01 2.32920110e-01
-5.09459376e-01 3.74926209e-01 -3.82739127e-01 -1.04179633e+00
-4.26719971e-02 -7.24282563e-01 7.06606656e-02 -8.48512530e-01
-4.54303287e-02 -4.46597546e-01 8.39746356e-01 -1.32210815e+00
1.40421283e+00 -2.44828320e+00 -2.38880187e-01 -3.34714586e-03
-6.31929636e-01 6.98434174e-01 -5.58557689e-01 5.87206423e-01
-1.78383082e-01 -3.97708267e-03 -2.62859017e-01 -5.76779723e-01
-7.89005831e-02 4.70651329e-01 -4.40322757e-01 5.45438286e-03
4.78707582e-01 6.25228941e-01 -8.22113872e-01 -3.81580740e-01
-1.04010412e-02 1.91741660e-01 -3.01403135e-01 5.22245526e-01
1.31712601e-01 2.86210299e-01 -2.54437894e-01 8.99918437e-01
6.13276899e-01 2.59021252e-01 5.24678648e-01 2.65814185e-01
-5.89335896e-03 1.06363940e+00 -8.52382779e-01 1.69033349e+00
-8.99687469e-01 2.11648524e-01 4.76422191e-01 -1.11220455e+00
1.53281617e+00 5.71385324e-01 2.60699660e-01 -8.15757275e-01
-1.07693970e-01 6.48979425e-01 -4.36865538e-01 -3.70825589e-01
7.94961095e-01 -5.97608984e-01 -3.30852121e-01 2.87017494e-01
2.89497316e-01 -2.16067731e-01 -1.10892422e-01 -5.42342775e-02
1.18533945e+00 -1.48521632e-01 3.73836190e-01 1.17357492e-01
4.76207078e-01 -6.57620654e-02 1.01312065e+00 4.04134214e-01
-4.75694835e-01 9.40979838e-01 1.08812891e-01 1.04549147e-01
-1.07087660e+00 -1.10188389e+00 1.47839934e-01 1.12168765e+00
-6.40219331e-01 -3.08832169e-01 -1.03006053e+00 -9.18318748e-01
-1.56340882e-01 8.63548279e-01 4.08289488e-03 -1.51142657e-01
-9.07528520e-01 -2.03423530e-01 7.73525298e-01 6.32838428e-01
1.85238898e-01 -1.41136825e+00 6.99185908e-01 6.14895940e-01
-7.96173871e-01 -1.54451358e+00 -1.00176239e+00 5.99273980e-01
-8.38634491e-01 -6.42497003e-01 -7.99239814e-01 -1.42540681e+00
3.21736008e-01 4.33198035e-01 7.51088858e-01 -3.15810382e-01
3.25490683e-01 2.99509823e-01 -9.24971759e-01 -2.50442773e-01
-1.02083051e+00 3.04947287e-01 3.94192964e-01 9.44186002e-02
1.16254747e+00 -4.09142166e-01 2.48999164e-01 3.74126315e-01
-8.78826141e-01 -6.02857232e-01 7.74126947e-01 1.10756147e+00
6.31915569e-01 -1.71261445e-01 1.23871255e+00 -6.92233264e-01
7.71400511e-01 -4.95672166e-01 -2.72767514e-01 4.80304092e-01
-3.65424037e-01 -1.60429388e-01 1.03911030e+00 -4.33623463e-01
-1.53049648e+00 3.55866104e-01 -3.94534826e-01 -4.87672180e-01
-3.13412249e-01 6.58955157e-01 -6.41374469e-01 5.17978191e-01
6.36966646e-01 5.52043378e-01 -5.25012128e-02 -1.05335820e+00
1.11849464e-01 1.75855923e+00 8.54570687e-01 -6.43906951e-01
4.39468563e-01 -1.86081573e-01 -1.09095275e+00 -1.14987242e+00
-7.41523206e-01 -9.40861642e-01 -7.06611693e-01 1.95108026e-01
5.92997193e-01 -1.23137259e+00 -5.03081679e-02 4.38127249e-01
-1.11232865e+00 -6.04748547e-01 -2.12805882e-01 4.73503441e-01
-7.32784390e-01 6.85737967e-01 -9.02573168e-01 -6.16793633e-01
-4.83907640e-01 -9.66376483e-01 1.04008532e+00 -1.53967533e-02
-1.42644823e-01 -7.01709449e-01 2.64315866e-02 7.46138752e-01
2.43229851e-01 -7.78466403e-01 7.89771020e-01 -1.06162691e+00
-9.95154232e-02 -1.36509582e-01 -1.13076948e-01 9.92086530e-01
4.60629225e-01 -3.84067625e-01 -8.78803849e-01 -2.42600873e-01
5.86032197e-02 -5.30320287e-01 5.21920919e-01 1.02586284e-01
1.10283768e+00 -4.18963939e-01 2.78903302e-02 2.49987707e-01
8.28204691e-01 2.42595106e-01 6.81932271e-01 3.29384744e-01
4.24434841e-01 5.44294775e-01 1.05200601e+00 5.72505593e-01
2.89430678e-01 7.67707586e-01 -3.85618627e-01 9.55029875e-02
-4.03245360e-01 -8.20789695e-01 9.93374825e-01 1.63710678e+00
6.70449495e-01 -2.91746557e-01 -1.00128126e+00 9.38434720e-01
-1.85297394e+00 -6.72525585e-01 1.34812921e-01 2.11613107e+00
1.33726704e+00 -2.07944676e-01 -5.02417721e-02 -4.73801084e-02
9.57653224e-01 -2.99297124e-01 -3.10712844e-01 -8.49324524e-01
-4.49587464e-01 3.89121741e-01 3.60712230e-01 3.33646387e-01
-1.03297031e+00 1.50336146e+00 6.22257710e+00 1.34119630e+00
-1.02393878e+00 3.81466150e-01 3.00533682e-01 1.74686924e-01
-2.09799454e-01 -2.24410802e-01 -1.04538047e+00 6.39466226e-01
1.36423194e+00 -2.15910092e-01 4.89906698e-01 1.08591330e+00
5.59907556e-01 3.64053905e-01 -9.79521930e-01 1.10378492e+00
3.01333845e-01 -9.11832154e-01 -9.39476341e-02 -2.48340815e-01
7.43853688e-01 3.49523544e-01 -2.23387480e-01 7.59215534e-01
4.52950746e-01 -8.50337625e-01 6.48861527e-01 -8.82395878e-02
9.18701589e-01 -9.16591644e-01 6.47751808e-01 5.82016110e-01
-8.17815840e-01 2.18396261e-03 -4.20654178e-01 3.03351581e-01
4.15649295e-01 6.10180557e-01 -1.44645941e+00 4.14152443e-01
6.77005768e-01 6.21381283e-01 -2.88427383e-01 6.53895020e-01
-4.24348444e-01 1.24527466e+00 -2.59444445e-01 5.90921417e-02
2.13642254e-01 -2.89559752e-01 4.15687054e-01 1.68027222e+00
5.16050756e-01 5.49064055e-02 4.19095725e-01 3.55895281e-01
-3.60273451e-01 4.39369172e-01 -4.45686966e-01 -2.15258807e-01
6.71894252e-01 9.47235525e-01 -9.67324898e-02 -3.84725332e-01
-4.60662723e-01 1.36556280e+00 5.04664302e-01 4.66596365e-01
-3.75202060e-01 -3.62078071e-01 8.56649935e-01 3.37289609e-02
6.95378184e-01 -5.14437973e-01 -5.71901202e-01 -1.36083794e+00
3.27422708e-01 -1.41323161e+00 2.66789466e-01 -6.98295832e-01
-1.63642883e+00 6.62904978e-01 -6.36797309e-01 -1.33626401e+00
-4.29056019e-01 -5.13752460e-01 -3.36495042e-01 9.24911797e-01
-1.88534820e+00 -1.24166834e+00 1.72945067e-01 7.53014863e-01
1.11676073e+00 -6.56511724e-01 9.49210644e-01 7.74161816e-01
-6.16698146e-01 1.05342603e+00 7.01682806e-01 3.83314401e-01
1.16254449e+00 -1.02200127e+00 5.33606350e-01 9.72517729e-01
1.95274711e-01 3.69030952e-01 2.45876446e-01 -6.72715843e-01
-1.09864080e+00 -1.35319853e+00 1.51116109e+00 -1.31519139e-01
7.46123314e-01 -4.69824106e-01 -1.18951976e+00 9.58686888e-01
1.64403692e-01 -2.90395021e-01 9.49150443e-01 2.90181518e-01
-4.14851516e-01 -1.99514627e-01 -1.03312230e+00 5.32101691e-01
6.46926522e-01 -9.37942803e-01 -9.30608809e-01 2.27336109e-01
6.48886144e-01 -1.64764464e-01 -7.05356956e-01 1.75130144e-01
-5.28753847e-02 -1.31333768e-01 4.75005686e-01 -7.60119259e-01
2.81473637e-01 -2.10096151e-01 -4.48128968e-01 -1.64580643e+00
-2.34676793e-01 -8.34244013e-01 2.46437445e-01 1.79910219e+00
5.89264989e-01 -2.42104962e-01 7.44554877e-01 2.95261145e-01
-6.68405950e-01 1.38365120e-01 -1.20534718e+00 -1.28995109e+00
2.77168483e-01 -4.49559450e-01 4.29371655e-01 1.05918431e+00
5.00977933e-01 5.67730308e-01 -3.19889456e-01 7.71363974e-02
1.21896416e-02 -2.57649094e-01 6.30575776e-01 -5.58153629e-01
-1.44547254e-01 5.59023535e-03 -4.38910201e-02 -1.33733153e+00
6.24122858e-01 -1.06992948e+00 5.66645801e-01 -1.06370807e+00
-8.23502392e-02 -5.81683576e-01 -2.33159840e-01 8.58044147e-01
-5.78180254e-02 -2.05505922e-01 1.07755467e-01 2.66548624e-04
-5.01213312e-01 8.18811536e-01 5.15061677e-01 -1.09220378e-01
-3.71548355e-01 1.29574448e-01 -5.42165875e-01 4.27767962e-01
1.07501221e+00 -6.45926714e-01 8.29278454e-02 -5.76414883e-01
-6.07160985e-01 1.72998473e-01 -3.27991247e-01 -6.29049301e-01
1.73405372e-02 -2.37024248e-01 -2.10458990e-02 -7.05281317e-01
2.17627972e-01 -6.81789935e-01 -4.43421036e-01 -1.50323808e-01
-2.12514624e-01 -7.66707212e-02 3.15937042e-01 3.24673474e-01
-6.64790571e-01 -4.26829487e-01 8.08632493e-01 3.07684615e-02
-6.85396850e-01 -1.13565542e-01 -1.04680288e+00 1.38951525e-01
4.64650184e-01 -2.14732904e-02 2.87143490e-03 -4.83957022e-01
-6.19524956e-01 8.25800896e-02 3.71170372e-01 8.72924149e-01
5.64966679e-01 -1.37985992e+00 -9.69938278e-01 3.01759511e-01
3.76437694e-01 -5.77882417e-02 2.95125037e-01 4.11650538e-01
-1.04190029e-01 5.10907650e-01 8.44692439e-02 -3.31241369e-01
-1.37267196e+00 6.17300689e-01 1.68894842e-01 -3.36859405e-01
-3.27780813e-01 5.81012189e-01 2.50580604e-03 -1.03591025e+00
3.26862544e-01 -2.59929836e-01 8.96133259e-02 -1.36570483e-01
5.80321729e-01 2.42089972e-01 7.89564252e-01 -6.55429125e-01
-3.95531088e-01 -3.18469554e-02 -3.43957424e-01 -2.11367816e-01
1.41762662e+00 -4.07436341e-01 2.89987832e-01 2.03096509e-01
1.30906260e+00 2.13427126e-01 -1.16963744e+00 -7.76926517e-01
4.25626844e-01 -2.36210689e-01 -7.75970444e-02 -1.02520859e+00
-6.47973239e-01 9.27632153e-01 1.29909858e-01 -1.69240922e-01
1.19980860e+00 -6.72251061e-02 1.28137362e+00 5.87265551e-01
2.04412237e-01 -1.83955073e+00 1.18731238e-01 1.05041230e+00
1.03748488e+00 -1.23282111e+00 -7.34476566e-01 -4.72719908e-01
-1.17169201e+00 1.09851646e+00 5.00202537e-01 2.45446846e-01
5.18296361e-01 3.62198353e-01 5.74855626e-01 5.35530031e-01
-6.53145134e-01 -2.12925002e-01 -9.56727043e-02 9.77474213e-01
5.24569333e-01 3.44536543e-01 -2.78859228e-01 1.26436234e+00
-3.54167581e-01 6.19431995e-02 7.32187331e-01 9.24961209e-01
-5.32915413e-01 -1.55251920e+00 -6.28381014e-01 -3.87630761e-02
-5.48751533e-01 -4.44732398e-01 -6.84817553e-01 1.08760342e-01
-3.10448498e-01 1.41315937e+00 -5.92361810e-03 -1.94849610e-01
6.75238788e-01 5.66592872e-01 1.48568586e-01 -9.44427907e-01
-5.49443543e-01 4.95257556e-01 3.55548739e-01 -1.25896826e-01
5.65649308e-02 -1.00398564e+00 -1.34086239e+00 -1.02188908e-01
-1.73799157e-01 2.79684037e-01 8.26865852e-01 9.25892889e-01
6.80974185e-01 2.42782354e-01 1.04751027e+00 -4.29987878e-01
-1.01933444e+00 -1.35941410e+00 -8.30324709e-01 4.24877733e-01
1.54050604e-01 -9.29001793e-02 -1.80293739e-01 4.00131553e-01] | [14.384541511535645, 6.900758266448975] |
893f2551-ca76-489a-9397-005f9fb2f798 | kiu-net-overcomplete-convolutional | 2010.01663 | null | https://arxiv.org/abs/2010.01663v2 | https://arxiv.org/pdf/2010.01663v2.pdf | KiU-Net: Overcomplete Convolutional Architectures for Biomedical Image and Volumetric Segmentation | Most methods for medical image segmentation use U-Net or its variants as they have been successful in most of the applications. After a detailed analysis of these "traditional" encoder-decoder based approaches, we observed that they perform poorly in detecting smaller structures and are unable to segment boundary regions precisely. This issue can be attributed to the increase in receptive field size as we go deeper into the encoder. The extra focus on learning high level features causes the U-Net based approaches to learn less information about low-level features which are crucial for detecting small structures. To overcome this issue, we propose using an overcomplete convolutional architecture where we project our input image into a higher dimension such that we constrain the receptive field from increasing in the deep layers of the network. We design a new architecture for image segmentation- KiU-Net which has two branches: (1) an overcomplete convolutional network Kite-Net which learns to capture fine details and accurate edges of the input, and (2) U-Net which learns high level features. Furthermore, we also propose KiU-Net 3D which is a 3D convolutional architecture for volumetric segmentation. We perform a detailed study of KiU-Net by performing experiments on five different datasets covering various image modalities like ultrasound (US), magnetic resonance imaging (MRI), computed tomography (CT), microscopic and fundus images. The proposed method achieves a better performance as compared to all the recent methods with an additional benefit of fewer parameters and faster convergence. Additionally, we also demonstrate that the extensions of KiU-Net based on residual blocks and dense blocks result in further performance improvements. The implementation of KiU-Net can be found here: https://github.com/jeya-maria-jose/KiU-Net-pytorch | ['Vishal M. Patel', 'Ilker Hacihaliloglu', 'Vishwanath A. Sindagi', 'Jeya Maria Jose Valanarasu'] | 2020-10-04 | null | null | null | null | ['3d-medical-imaging-segmentation', 'volumetric-medical-image-segmentation', 'liver-segmentation', 'ultrasound'] | ['medical', 'medical', 'medical', 'medical'] | [ 8.07596222e-02 3.27823073e-01 -1.36593699e-01 -3.07588130e-01
-5.91379225e-01 -2.31663480e-01 8.53227526e-02 5.97857928e-04
-3.96383226e-01 6.63091660e-01 2.15185449e-01 -3.99055809e-01
2.26854578e-01 -9.72314775e-01 -9.90402460e-01 -4.83884275e-01
-3.42358589e-01 9.84118730e-02 4.97152030e-01 1.07545787e-02
1.55913860e-01 3.40519726e-01 -1.03217256e+00 3.11927110e-01
6.24413252e-01 1.04817939e+00 1.60301223e-01 7.64812887e-01
-4.11083959e-02 7.72444606e-01 -7.43315667e-02 -9.56042483e-03
3.97176653e-01 -3.94854426e-01 -1.13728499e+00 1.81328356e-02
2.94365317e-01 -6.24113500e-01 -4.30720985e-01 1.03070807e+00
5.65726459e-01 -8.98737609e-02 5.72950542e-01 -6.50027514e-01
-6.73782051e-01 5.51455557e-01 -8.27977180e-01 5.23999453e-01
5.46972007e-02 -4.08749580e-02 7.60013282e-01 -5.54918587e-01
7.38436699e-01 1.11283815e+00 8.90410423e-01 5.90502322e-01
-1.09072697e+00 -4.96034235e-01 4.16551717e-02 -2.54479293e-02
-1.31684875e+00 -5.28785996e-02 5.83953798e-01 -4.64748830e-01
8.46166551e-01 4.76740338e-02 5.31206846e-01 7.32899368e-01
4.30728555e-01 9.19564247e-01 1.10985148e+00 -3.24831337e-01
-1.88038603e-03 -1.66158050e-01 2.72843838e-01 1.10324979e+00
2.04506621e-01 3.90835665e-02 1.26768142e-01 5.35540320e-02
1.41066337e+00 1.34268090e-01 -4.78038073e-01 -5.92999421e-02
-9.91784811e-01 8.43647778e-01 8.75932932e-01 5.19965529e-01
-3.57570767e-01 3.55179727e-01 5.12811542e-01 1.39139578e-01
3.04164290e-01 1.59050822e-01 -4.84305471e-01 1.67184547e-01
-9.06257987e-01 -5.21199554e-02 4.37996179e-01 7.59468615e-01
8.22915316e-01 -1.08282760e-01 -2.92576939e-01 7.60364294e-01
2.31294468e-01 -6.58087209e-02 7.65531242e-01 -1.04011559e+00
2.80734807e-01 7.04984367e-01 -2.18930662e-01 -6.83336139e-01
-6.43076718e-01 -5.52624345e-01 -1.08494747e+00 3.11307698e-01
3.11749607e-01 -4.25952703e-01 -1.32846713e+00 1.57401562e+00
1.83828220e-01 3.83274853e-01 -6.83525056e-02 1.01019084e+00
1.16387010e+00 6.39076591e-01 -1.03166744e-01 1.79107875e-01
1.51900196e+00 -1.05199242e+00 -4.93110090e-01 -4.25486043e-02
7.27999091e-01 -6.12160742e-01 8.79700780e-01 1.40922591e-01
-1.15465271e+00 -5.91761410e-01 -1.06984222e+00 -4.12233293e-01
-4.62450862e-01 3.59090686e-01 8.43264222e-01 4.21844304e-01
-1.34066296e+00 7.10758746e-01 -9.61649477e-01 -3.14593464e-01
8.25568318e-01 6.61925018e-01 -3.29488218e-01 1.03969775e-01
-1.09738088e+00 6.26156807e-01 3.33507240e-01 1.38816491e-01
-7.17348397e-01 -6.20527864e-01 -9.27375376e-01 2.10085362e-01
2.01700568e-01 -7.18493164e-01 1.08095109e+00 -9.41687405e-01
-1.29050028e+00 8.84417713e-01 -3.82033549e-02 -5.72867155e-01
4.06736046e-01 5.65651208e-02 1.33528844e-01 4.46945906e-01
-7.59713054e-02 1.16867578e+00 4.02056187e-01 -1.10172522e+00
-5.69153428e-01 -3.79686922e-01 3.44304651e-01 -7.31566250e-02
-8.46175924e-02 -2.59497136e-01 -5.51735342e-01 -7.60385633e-01
1.25036001e-01 -8.07798445e-01 -4.95825469e-01 2.26241872e-02
-4.45569783e-01 -4.30695340e-02 7.00992227e-01 -7.02982306e-01
1.18522835e+00 -1.98700237e+00 -5.04440395e-03 6.26167431e-02
4.84255105e-01 4.70611751e-01 -1.43628329e-01 1.00467861e-01
-1.23267315e-01 1.64016917e-01 -3.67922992e-01 -1.14500724e-01
-4.72029746e-01 3.26134682e-01 2.42977560e-01 4.26058859e-01
1.88184991e-01 1.07940912e+00 -7.60521054e-01 -5.55660427e-01
2.90281743e-01 5.80278873e-01 -7.23168910e-01 4.65367623e-02
6.92937151e-02 6.59768164e-01 -5.74976027e-01 6.83379292e-01
8.03595662e-01 -5.08114576e-01 -7.66798630e-02 -2.31157184e-01
-1.16157025e-01 4.03923616e-02 -1.04808915e+00 1.73101103e+00
-3.32381874e-01 6.40580416e-01 2.86941141e-01 -1.28919017e+00
6.70348346e-01 4.41240370e-01 6.53859496e-01 -6.90811276e-01
6.69465363e-01 2.12322220e-01 1.04769439e-01 -5.29719591e-01
3.44981365e-02 -1.87886283e-02 3.06567490e-01 2.03368306e-01
2.32527405e-01 3.20027202e-01 2.31690213e-01 7.37207308e-02
1.09782171e+00 4.20454890e-02 2.92086661e-01 -3.63194913e-01
7.07286417e-01 -1.78901911e-01 6.68456376e-01 5.83783269e-01
-3.43777210e-01 7.53807366e-01 6.98682904e-01 -6.40772760e-01
-9.75413263e-01 -8.18494618e-01 -4.36295986e-01 4.97274280e-01
1.19234554e-01 -3.34634990e-01 -9.72385526e-01 -6.19094193e-01
-2.21667826e-01 3.55537683e-02 -9.26674008e-01 1.34947360e-01
-6.13717854e-01 -8.01598251e-01 5.12953341e-01 8.88871968e-01
7.56827354e-01 -1.02391315e+00 -6.93450272e-01 1.94081694e-01
-4.83678132e-02 -1.14211893e+00 -4.00993884e-01 1.79840922e-01
-1.26871741e+00 -1.18171024e+00 -1.07405829e+00 -1.07410336e+00
1.08177757e+00 -1.37895510e-01 9.52863038e-01 2.29328170e-01
-5.74003041e-01 1.18074767e-01 -4.28015381e-01 -3.10098648e-01
-1.32730424e-01 1.97670653e-01 -4.33754116e-01 -2.07923070e-01
3.43484044e-01 -6.02742076e-01 -1.02369726e+00 9.59671885e-02
-1.06521499e+00 1.45155311e-01 7.70492911e-01 9.02179956e-01
6.98034346e-01 -8.38004947e-02 4.93608117e-01 -1.22215986e+00
4.89340693e-01 -3.96089196e-01 -5.42102635e-01 -1.08213753e-01
-2.97676951e-01 -1.88431144e-02 6.42475069e-01 -1.78979263e-01
-7.13830888e-01 1.10256098e-01 -6.49638772e-01 -5.19031882e-01
-5.76145016e-02 3.67078722e-01 4.17662024e-01 -3.58495414e-01
4.94255126e-01 9.92143378e-02 -2.49542315e-02 -5.43309152e-01
-1.46792680e-02 6.61777973e-01 3.53387505e-01 -3.06717455e-01
3.75315487e-01 6.30223632e-01 8.21452588e-02 -6.91930354e-01
-7.04339981e-01 -4.81583863e-01 -6.85393810e-01 -2.71626096e-02
1.08672798e+00 -8.13800395e-01 -8.01315784e-01 3.39322686e-01
-1.05023289e+00 -4.71855640e-01 -9.88204926e-02 5.28471947e-01
-4.50579524e-01 4.40943450e-01 -1.13033295e+00 -3.67054403e-01
-5.63083768e-01 -1.50789845e+00 9.65628445e-01 4.91395205e-01
1.34679273e-01 -1.15110874e+00 -7.42997415e-03 2.46653348e-01
4.36891317e-01 4.06403005e-01 1.02425694e+00 -2.63624668e-01
-5.54640591e-01 8.62761959e-03 -5.22068501e-01 5.75720489e-01
1.04513541e-01 -2.70810753e-01 -8.30564082e-01 -3.41195017e-01
-1.86129078e-01 -3.03793728e-01 1.07915604e+00 9.92329419e-01
1.55660832e+00 -1.93425983e-01 -3.72478902e-01 1.02304864e+00
1.81234229e+00 7.19609782e-02 7.70724952e-01 1.45701066e-01
6.76841736e-01 3.47256929e-01 1.71120748e-01 2.41524577e-01
3.52355719e-01 2.85934627e-01 5.51925421e-01 -7.65289068e-01
-3.18695992e-01 -2.47427840e-02 -6.50826748e-03 6.94633305e-01
-3.54070157e-01 1.13827333e-01 -7.01052010e-01 7.41232216e-01
-1.77777350e+00 -5.83070636e-01 -1.08937159e-01 1.92210472e+00
8.02693367e-01 1.10987440e-01 1.30166993e-01 1.04753822e-02
5.81156492e-01 -6.68861419e-02 -4.81684059e-01 -6.03022397e-01
2.96049565e-01 4.58998322e-01 8.58065665e-01 5.45020938e-01
-1.19199359e+00 8.80795121e-01 6.30872202e+00 6.51404023e-01
-1.39776707e+00 1.65463254e-01 1.06542599e+00 1.52261019e-01
-6.76541999e-02 -2.51065850e-01 -5.97665489e-01 4.08619851e-01
6.11572027e-01 4.46482837e-01 4.97271642e-02 6.31496310e-01
1.47453040e-01 -2.54966021e-01 -8.89041603e-01 8.91120255e-01
-2.80126125e-01 -1.66499698e+00 -9.17406827e-02 4.59729768e-02
7.56930292e-01 4.04457867e-01 -1.03276238e-01 2.49386743e-01
1.01779930e-01 -1.17068624e+00 1.82752281e-01 2.56871015e-01
1.00722682e+00 -5.58713734e-01 9.69772100e-01 2.27419153e-01
-1.07078421e+00 7.06714764e-02 -7.09355474e-01 -1.65720284e-02
-3.56930234e-02 5.04194260e-01 -7.54611611e-01 4.22911435e-01
7.97594786e-01 6.71382785e-01 -3.39824468e-01 1.11999655e+00
5.90231195e-02 6.33966506e-01 -1.83125243e-01 2.07778424e-01
7.40805864e-01 -2.47825935e-01 3.17962915e-01 1.37982130e+00
2.69676059e-01 3.76596063e-01 1.96192756e-01 9.63640451e-01
-3.31390381e-01 1.28443658e-01 -5.14410794e-01 2.72108763e-01
-1.32919163e-01 1.24405456e+00 -1.06396925e+00 -3.75192225e-01
-7.29638577e-01 9.17941988e-01 2.86713123e-01 3.73726755e-01
-6.77542686e-01 -4.70993519e-01 3.90558928e-01 2.30388880e-01
5.52602947e-01 -2.92837266e-02 -2.78533757e-01 -1.08244896e+00
-2.27771372e-01 -4.78156388e-01 3.70819598e-01 -5.11605501e-01
-8.39700162e-01 8.40442598e-01 -1.87496379e-01 -1.05578935e+00
9.19724479e-02 -8.19879532e-01 -5.37940741e-01 7.85278320e-01
-1.97126770e+00 -1.06938267e+00 -2.51515418e-01 6.09161556e-01
5.62569618e-01 8.78515244e-02 6.60177350e-01 5.96221149e-01
-5.36307871e-01 4.80120361e-01 8.46892074e-02 6.53950930e-01
4.15726811e-01 -1.29833126e+00 1.61736354e-01 6.15847826e-01
-2.03456953e-01 5.87445498e-01 2.82028884e-01 -5.61427355e-01
-9.21163619e-01 -1.06557167e+00 5.79815507e-01 -9.99102220e-02
3.07838559e-01 -2.48130605e-01 -8.00328791e-01 8.44059408e-01
3.19521308e-01 4.60689187e-01 6.20234311e-01 -2.10810214e-01
2.83204280e-02 -1.44337201e-02 -1.28555262e+00 3.33851129e-01
6.71716630e-01 -4.62071747e-02 -4.54607338e-01 2.42860943e-01
5.34230232e-01 -6.44065619e-01 -1.12418032e+00 4.05526578e-01
5.06478846e-01 -1.15333676e+00 8.97180557e-01 -3.35124433e-01
7.49059618e-01 -1.94700390e-01 1.48267061e-01 -9.68506932e-01
-4.90218788e-01 -3.23515624e-01 2.18874067e-02 7.93746233e-01
3.61457467e-01 -6.29516244e-01 9.52249348e-01 1.92267802e-02
-2.31422693e-01 -1.42158246e+00 -7.96853065e-01 -3.28699023e-01
2.80321509e-01 -1.17823608e-01 3.26082438e-01 7.34875381e-01
-2.62775093e-01 2.35317349e-01 -2.17229307e-01 1.81946941e-02
4.87807512e-01 6.82670772e-02 2.45517641e-01 -9.69081640e-01
-2.01912954e-01 -3.74265552e-01 -5.55275679e-01 -1.19741011e+00
-2.51937896e-01 -9.54737782e-01 -2.86536574e-01 -1.77600098e+00
3.94360214e-01 -4.64349508e-01 -3.02336961e-01 5.74433804e-01
-4.35734575e-04 6.05219662e-01 -2.28188876e-02 1.90044686e-01
-3.55727345e-01 1.66064426e-01 1.88715136e+00 2.01812256e-02
-1.83318794e-01 1.81900524e-02 -6.06006742e-01 8.22432995e-01
7.14289665e-01 -1.95057750e-01 -2.99204290e-01 -5.72616935e-01
-9.13213119e-02 1.78425729e-01 2.38973990e-01 -9.51308906e-01
6.27809390e-02 4.00787324e-01 6.74674153e-01 -5.69255173e-01
1.75506454e-02 -8.43670785e-01 -4.42103684e-01 7.08497882e-01
-2.07722440e-01 -8.56938586e-02 1.99808329e-01 2.41603479e-01
-4.04168636e-01 -2.67268032e-01 9.87562358e-01 -5.93451202e-01
-7.10891008e-01 5.59183955e-01 -3.69478673e-01 -1.96205348e-01
9.91325378e-01 -3.91183555e-01 -9.46292058e-02 -1.58029824e-01
-8.47878218e-01 3.77713203e-01 3.34119737e-01 7.34916851e-02
5.75849891e-01 -1.02490139e+00 -5.99498808e-01 2.53447324e-01
-2.85496861e-01 3.34516406e-01 3.52419466e-01 1.19390798e+00
-1.07859170e+00 7.06556082e-01 -5.20286977e-01 -6.29849374e-01
-9.95498002e-01 3.48381788e-01 5.39001286e-01 -4.15909439e-01
-9.06371474e-01 9.24683452e-01 5.57353556e-01 -3.79281044e-01
2.91552097e-01 -7.71836996e-01 -5.33246160e-01 -2.57101268e-01
3.76520127e-01 2.30150580e-01 -2.19849944e-02 -5.42782962e-01
-3.36872041e-01 7.79353678e-01 -1.75440997e-01 4.40076083e-01
1.39434624e+00 -1.49863437e-01 -2.38971129e-01 -1.40664242e-02
1.45017576e+00 -3.91300291e-01 -1.32190132e+00 -1.55784100e-01
-2.36389160e-01 -1.43700808e-01 4.54841435e-01 -6.25649273e-01
-1.55252171e+00 1.03383040e+00 8.24184954e-01 -1.98385271e-04
1.32518399e+00 1.80145223e-02 1.20933545e+00 -9.05823633e-02
1.26168743e-01 -9.38094020e-01 -8.77349451e-02 4.32864606e-01
4.67308730e-01 -1.37075853e+00 -1.12474561e-01 -5.48393250e-01
-4.01274592e-01 1.23529840e+00 6.43931925e-01 -5.22610605e-01
8.20795894e-01 4.17101264e-01 1.79900676e-01 -3.97674173e-01
-4.16815013e-01 -4.38332647e-01 3.71273160e-01 3.78052980e-01
7.93001771e-01 -1.15452493e-02 -6.25610590e-01 5.88081360e-01
1.51226163e-01 3.74776274e-01 6.13256574e-01 8.67241561e-01
-3.86439621e-01 -1.05694675e+00 -2.70372927e-01 6.91511512e-01
-9.55757737e-01 -2.87477896e-02 -3.64645161e-02 7.59734690e-01
3.25400144e-01 4.46311772e-01 1.28744841e-01 -1.11672327e-01
3.67116220e-02 -3.14837396e-01 6.15087390e-01 -7.06320345e-01
-6.24819517e-01 3.82925838e-01 -2.56369025e-01 -7.27731526e-01
-5.08042991e-01 -3.00507426e-01 -1.59311044e+00 -1.07904583e-01
-1.25609592e-01 -3.53809595e-02 4.32323188e-01 7.90874481e-01
1.96687683e-01 8.91209543e-01 3.59930456e-01 -8.27163458e-01
-1.74479321e-01 -9.22604501e-01 -5.84992290e-01 2.28862226e-01
5.53904951e-01 -5.42090774e-01 -1.70781873e-02 5.48228100e-02] | [14.633065223693848, -2.6099634170532227] |
45ce20d3-3602-46e6-bfef-e0062bb9bd1f | appearance-based-gaze-estimation-using | 1903.07296 | null | http://arxiv.org/abs/1903.07296v1 | http://arxiv.org/pdf/1903.07296v1.pdf | Appearance-Based Gaze Estimation Using Dilated-Convolutions | Appearance-based gaze estimation has attracted more and more attention
because of its wide range of applications. The use of deep convolutional neural
networks has improved the accuracy significantly. In order to improve the
estimation accuracy further, we focus on extracting better features from eye
images. Relatively large changes in gaze angles may result in relatively small
changes in eye appearance. We argue that current architectures for gaze
estimation may not be able to capture such small changes, as they apply
multiple pooling layers or other downsampling layers so that the spatial
resolution of the high-level layers is reduced significantly. To evaluate
whether the use of features extracted at high resolution can benefit gaze
estimation, we adopt dilated-convolutions to extract high-level features
without reducing spatial resolution. In cross-subject experiments on the
Columbia Gaze dataset for eye contact detection and the MPIIGaze dataset for 3D
gaze vector regression, the resulting Dilated-Nets achieve significant (up to
20.8%) gains when compared to similar networks without dilated-convolutions.
Our proposed Dilated-Net achieves state-of-the-art results on both the Columbia
Gaze and the MPIIGaze datasets. | ['Zhaokang Chen', 'Bertram E. Shi'] | 2019-03-18 | null | null | null | null | ['contact-detection'] | ['robots'] | [ 2.07059950e-01 -6.95944354e-02 7.29694888e-02 -7.49446213e-01
-6.31429479e-02 -2.00045228e-01 2.39209980e-01 -3.46504360e-01
-5.58116674e-01 4.57941800e-01 -5.03558926e-02 -1.02489263e-01
1.40777349e-01 -2.52653509e-01 -6.32936716e-01 -5.56930482e-01
-3.96537706e-02 -5.65089524e-01 1.98608398e-01 -1.24379024e-01
4.43017393e-01 4.28705812e-01 -2.15064955e+00 1.43838391e-01
8.89845729e-01 1.18040967e+00 2.91376635e-02 5.50422311e-01
1.83835253e-01 4.93170947e-01 -4.34441805e-01 -2.82687306e-01
2.26671889e-01 -2.66609609e-01 -5.72623491e-01 -1.96107447e-01
1.32421052e+00 -7.29769588e-01 -6.20162301e-02 1.00877190e+00
5.02632737e-01 7.94169232e-02 4.71504569e-01 -1.31955576e+00
-6.48464382e-01 -1.22358046e-01 -1.14523327e+00 4.87025946e-01
3.39120120e-01 2.68657714e-01 7.47911572e-01 -7.87243545e-01
3.91588181e-01 1.16771221e+00 6.12972319e-01 6.79502070e-01
-1.25650966e+00 -1.06115413e+00 3.32760960e-01 2.50079006e-01
-1.34227097e+00 -8.29669118e-01 5.85878789e-01 -3.27436268e-01
1.15301812e+00 3.10910225e-01 5.07369101e-01 9.43891406e-01
2.53499478e-01 7.26708889e-01 1.49447203e+00 -5.27530611e-01
-4.33875740e-01 1.08591452e-01 2.68817544e-01 7.17653334e-01
3.20633769e-01 9.86709744e-02 -7.05552220e-01 2.97623485e-01
7.03645110e-01 1.14096023e-01 -5.54803371e-01 -5.60476705e-02
-8.58685136e-01 5.66819608e-01 9.79291856e-01 -4.26892489e-02
-3.66334558e-01 2.39505947e-01 3.12191229e-02 2.17077971e-01
8.20180416e-01 6.18611693e-01 -4.19194937e-01 -1.65743500e-01
-1.06721151e+00 1.98490381e-01 3.27392370e-01 7.65112400e-01
8.61799419e-01 -1.45239234e-01 -1.99411780e-01 7.68035173e-01
4.82678115e-01 5.02772391e-01 2.77860492e-01 -8.94594431e-01
3.13641369e-01 6.82008862e-01 1.43238798e-01 -9.53825176e-01
-6.74584746e-01 -1.92588754e-02 -3.81534666e-01 9.10765111e-01
7.35292554e-01 -2.99203038e-01 -1.11945081e+00 1.88582289e+00
2.53608614e-01 -5.66285215e-02 -5.18824816e-01 1.15292525e+00
7.58866191e-01 2.56610304e-01 7.36677945e-02 -5.74564971e-02
1.52082229e+00 -7.79015481e-01 -7.64531493e-01 -3.90642494e-01
4.43006754e-01 -8.36588919e-01 1.23523271e+00 2.04916120e-01
-1.12613904e+00 -6.25293970e-01 -1.10520911e+00 -3.96868438e-01
-3.46044689e-01 2.41621032e-01 7.29784966e-01 7.71095097e-01
-1.32817626e+00 4.11416113e-01 -8.01480174e-01 -4.40349996e-01
8.34748983e-01 8.37509573e-01 -3.27960253e-01 1.19282618e-01
-7.72753954e-01 8.89808238e-01 -3.91444981e-01 3.99622023e-01
-9.66592953e-02 -6.78787947e-01 -9.32669282e-01 2.44933501e-01
7.11761937e-02 -4.82791722e-01 1.16657591e+00 -1.29656088e+00
-1.43655109e+00 9.72339213e-01 -8.15244913e-01 -1.26907110e-01
1.77315742e-01 -4.40565586e-01 -3.69972318e-01 -2.30944417e-02
-3.17824066e-01 1.27537155e+00 1.25577486e+00 -6.82745576e-01
-8.32749426e-01 -6.35547340e-01 3.40116471e-01 1.32162198e-01
-4.90261674e-01 3.97697926e-01 -3.17368567e-01 -3.18700582e-01
-1.33982107e-01 -1.11080325e+00 4.07558471e-01 3.60694677e-01
-1.97850570e-01 -4.55654174e-01 9.52204704e-01 -4.64676708e-01
1.29310918e+00 -1.98475826e+00 -2.15182409e-01 -7.94767141e-02
7.82858372e-01 3.57259840e-01 -1.41989216e-01 -2.77460396e-01
-2.44343653e-01 6.83240071e-02 3.19539577e-01 -5.90089381e-01
-2.68511176e-01 -3.68092597e-01 -1.03759551e-02 5.61330855e-01
6.17798388e-01 9.63545263e-01 -5.05702376e-01 -1.88291192e-01
1.44689277e-01 7.53508329e-01 -5.93901753e-01 -1.09959781e-01
7.86785483e-02 2.17793196e-01 -1.72494158e-01 6.09539509e-01
9.47223425e-01 -4.97672826e-01 -3.22703481e-01 -3.12610805e-01
-3.13785613e-01 3.13974828e-01 -6.00934744e-01 1.37923849e+00
-4.51133132e-01 1.71783674e+00 -1.70862600e-01 -1.21101819e-01
5.82604110e-01 -6.10701293e-02 7.34991580e-02 -1.06419826e+00
5.38588881e-01 -7.59127662e-02 4.77259129e-01 -4.24917698e-01
7.05406249e-01 3.09487820e-01 3.44891757e-01 5.39974332e-01
-1.69323698e-01 5.60616732e-01 -1.14690766e-01 -2.57357389e-01
6.81906939e-01 3.20591241e-01 -3.55570689e-02 -3.05955529e-01
3.96534503e-01 -3.84742707e-01 1.17103055e-01 1.66993558e-01
-4.78134960e-01 6.40625834e-01 6.33339405e-01 -4.15180117e-01
-7.16829717e-01 -4.76109713e-01 -2.93822378e-01 1.38971269e+00
2.15878963e-01 -2.93046653e-01 -1.06282413e+00 -6.27458394e-01
-3.25461105e-02 3.38163853e-01 -1.16227376e+00 -1.04905240e-01
-4.73149776e-01 -6.17139280e-01 3.63337755e-01 5.36210179e-01
5.83124220e-01 -1.09257638e+00 -1.07847297e+00 -3.86550963e-01
1.38333783e-01 -9.89145517e-01 -6.43117964e-01 -2.89989024e-01
-6.20997488e-01 -1.22371817e+00 -8.64035249e-01 -5.74992239e-01
9.23834205e-01 5.82567453e-01 8.42703223e-01 1.15804344e-01
-1.23530008e-01 -3.96876261e-02 -3.27071734e-02 -7.95787156e-01
2.23799542e-01 3.67232174e-01 1.07357092e-01 1.42867699e-01
1.12109315e+00 -2.40810469e-01 -1.00448120e+00 3.75193089e-01
-4.24035579e-01 -5.03864976e-05 4.49165106e-01 7.19279170e-01
-5.16929217e-02 -6.51160896e-01 3.69182438e-01 -7.04593003e-01
7.66645432e-01 9.15560350e-02 -8.45609665e-01 -1.43561102e-02
-8.94127786e-01 1.03408352e-01 -1.43546192e-02 -5.21559298e-01
-1.06491554e+00 -2.95228273e-01 1.54665291e-01 -5.70548832e-01
-2.06297204e-01 8.83199573e-02 3.15991014e-01 -4.87745613e-01
8.63222122e-01 -3.93811703e-01 2.82520592e-01 -3.76872241e-01
-1.51231671e-02 9.08570051e-01 -1.56363007e-02 1.71008214e-01
4.50484395e-01 4.06427175e-01 1.56636611e-01 -7.03345895e-01
-8.46144140e-01 -2.76860505e-01 -6.17910624e-01 -8.30973312e-02
9.48932350e-01 -9.57782865e-01 -1.36566257e+00 9.15556252e-01
-9.73810375e-01 -3.99975866e-01 2.79640853e-01 4.71250832e-01
-1.23954356e-01 -1.59394741e-01 -2.85174489e-01 -6.79860353e-01
-4.58731532e-01 -1.39997339e+00 1.09051204e+00 7.20979869e-01
-4.53021556e-01 -6.22489035e-01 -2.17285052e-01 7.63249844e-02
6.72766089e-01 6.82417527e-02 4.73401308e-01 -1.00220926e-01
-5.61083555e-01 -1.60922363e-01 -7.23162651e-01 2.26595655e-01
3.44108164e-01 2.69739658e-01 -1.60074580e+00 -3.47102255e-01
-4.48081553e-01 -2.82399386e-01 8.54386568e-01 8.06250632e-01
1.16948855e+00 1.05713140e-02 -4.58890110e-01 8.86565328e-01
1.02963400e+00 -1.89806849e-01 8.40945244e-01 3.79852235e-01
9.39498842e-01 7.52293408e-01 5.07218480e-01 1.37414545e-01
4.32277709e-01 7.77982593e-01 4.71872568e-01 -3.42425495e-01
-3.26246470e-01 1.37805864e-01 1.08329087e-01 1.03185333e-01
-4.81216252e-01 -9.76341888e-02 -7.78666258e-01 3.72360706e-01
-1.49969268e+00 -8.45885158e-01 -2.56554157e-01 2.25663519e+00
7.62454033e-01 3.15448940e-02 2.38447785e-01 -1.92735597e-01
6.32159770e-01 1.66860804e-01 -7.79103398e-01 -5.05241692e-01
2.38758698e-01 3.59220177e-01 5.10913074e-01 2.50150323e-01
-1.01442015e+00 7.87407517e-01 6.56442118e+00 2.34093368e-01
-1.86001205e+00 -1.31263092e-01 6.19571209e-01 -6.59849405e-01
2.73066878e-01 -4.00449932e-01 -1.08447170e+00 6.42391443e-01
9.67524827e-01 3.67933750e-01 4.06405836e-01 5.88753879e-01
2.09480897e-01 -3.35764050e-01 -1.10331738e+00 1.21095717e+00
2.11062744e-01 -1.13506234e+00 -5.31160474e-01 4.04258579e-01
5.98065138e-01 3.37958217e-01 5.37784755e-01 7.69379288e-02
-3.97251368e-01 -1.31331611e+00 4.75225061e-01 6.58654630e-01
1.35275877e+00 -7.04020023e-01 6.11092150e-01 -1.17655575e-01
-7.91577101e-01 -1.27090901e-01 -3.45333874e-01 -3.54397416e-01
-2.52374470e-01 -1.43533707e-01 -7.98835933e-01 -3.24605227e-01
1.20348406e+00 7.41917968e-01 -1.00461721e+00 1.08445001e+00
-1.24033622e-01 4.85797405e-01 -4.25225466e-01 -2.70331562e-01
-3.92077379e-02 1.63587615e-01 3.24357301e-01 8.11517656e-01
4.77325693e-02 -2.13261828e-01 -7.65978515e-01 1.03615451e+00
-3.63088191e-01 -2.78523743e-01 -3.53865772e-01 2.77929753e-01
3.45001459e-01 1.34542048e+00 -2.55607367e-01 -1.76392104e-02
-6.27632678e-01 8.58591557e-01 4.62856382e-01 6.44380271e-01
-7.40814328e-01 -7.86234498e-01 1.37511218e+00 4.94834632e-01
3.93681675e-01 -6.03341199e-02 -4.25769120e-01 -9.94954586e-01
-6.20357059e-02 -7.78654933e-01 -1.69415757e-01 -1.08575785e+00
-9.30135548e-01 7.45929658e-01 -1.15886651e-01 -1.16130185e+00
-3.70914012e-01 -8.03418040e-01 -4.90988404e-01 1.57065570e+00
-1.80067742e+00 -1.20735049e+00 -9.38093066e-01 6.17032349e-01
3.41962278e-01 -2.92426106e-02 7.17668235e-01 1.51047558e-01
-7.94023275e-01 1.19677103e+00 -2.03242972e-01 8.20031017e-02
1.14690208e+00 -1.07972777e+00 5.74489415e-01 9.15537119e-01
-1.78025648e-01 1.03033626e+00 5.86756945e-01 -9.68846157e-02
-1.04582536e+00 -6.62347734e-01 8.88862073e-01 -8.74571204e-01
1.97790951e-01 -5.52479565e-01 -1.01367426e+00 7.18088329e-01
4.57290858e-01 2.50461012e-01 4.74514037e-01 4.93859619e-01
-4.60044056e-01 -2.39746317e-01 -1.08614230e+00 7.31083632e-01
9.48124290e-01 -5.53387403e-01 -2.94297934e-01 -2.66521066e-01
3.42874914e-01 -5.30755162e-01 -6.07242644e-01 3.34684312e-01
1.00622725e+00 -1.14800513e+00 7.33382821e-01 -5.31292081e-01
3.58099282e-01 -1.51806518e-01 4.52944279e-01 -1.11442494e+00
-2.76212633e-01 -6.33317769e-01 -2.78926790e-01 8.68620098e-01
3.58258158e-01 -9.46027219e-01 7.11912155e-01 9.59998369e-01
2.30262011e-01 -8.62173796e-01 -4.54569668e-01 -1.19967550e-01
-1.46828443e-01 -1.65626809e-01 7.50915110e-01 6.72824025e-01
-8.66842121e-02 5.39453268e-01 -2.22192138e-01 -2.35308167e-02
4.40953076e-01 2.65593734e-02 1.00302815e+00 -1.43676293e+00
3.48217785e-01 -5.83071351e-01 -6.43016219e-01 -1.20242178e+00
2.69017547e-01 -1.56429127e-01 -6.47697449e-02 -1.11821985e+00
-1.01035126e-01 -3.06848913e-01 -3.82423669e-01 5.65537572e-01
-5.45084834e-01 8.27922046e-01 9.47552249e-02 1.54484451e-01
-3.31237316e-01 2.03014791e-01 1.32773614e+00 1.70564219e-01
-4.60126668e-01 1.09150842e-01 -8.60722482e-01 6.57620430e-01
5.45901597e-01 -1.42775476e-01 -4.11964506e-01 -5.79238296e-01
4.40820664e-01 -6.46555185e-01 4.82525349e-01 -7.56060600e-01
2.70856231e-01 2.20246345e-01 8.61921191e-01 -4.08922583e-01
4.72867250e-01 -5.87317109e-01 -4.80132073e-01 1.20162303e-02
-1.54138654e-01 -6.05359953e-03 7.43718684e-01 3.42362016e-01
-1.06486157e-01 9.72424746e-02 7.84564376e-01 4.43305552e-01
-7.90966868e-01 2.92844653e-01 -4.91906032e-02 -3.08185816e-01
7.46897161e-01 -7.84996450e-01 -6.01006031e-01 -3.92305553e-01
-3.76964808e-01 1.66838858e-02 8.19505274e-01 6.87249720e-01
4.54321533e-01 -9.15491819e-01 -3.70699733e-01 6.79702759e-01
2.43352920e-01 -1.57453075e-01 1.47487059e-01 1.16791821e+00
-3.60734850e-01 5.21838248e-01 -6.32446587e-01 -8.56732905e-01
-1.64849877e+00 2.64221698e-01 5.05553544e-01 4.53012586e-01
-3.18038642e-01 1.13479030e+00 5.92870772e-01 1.91604719e-01
1.71192303e-01 -7.11335242e-01 -5.31556070e-01 1.61405638e-01
8.69659722e-01 3.73578191e-01 1.17244445e-01 -7.94720411e-01
-4.46081549e-01 8.10603857e-01 -3.68110329e-01 4.04069185e-01
1.35424948e+00 -5.51187336e-01 5.11742234e-02 1.54291525e-01
1.30208910e+00 1.12198532e-01 -1.66873956e+00 -1.05449587e-01
-3.71722132e-01 -8.37711632e-01 4.33092564e-01 -7.28322148e-01
-1.07230747e+00 1.14345598e+00 1.10014725e+00 1.00871123e-01
1.43407261e+00 -1.43629327e-01 5.64279497e-01 2.46185854e-01
5.50546907e-02 -5.93622863e-01 -2.78098196e-01 3.23562682e-01
7.01480806e-01 -1.77433455e+00 1.73291173e-02 -2.90214121e-01
-4.35276747e-01 1.11476791e+00 9.26955104e-01 -1.60901919e-01
6.71726167e-01 -1.46435589e-01 2.15053186e-01 -5.21936417e-01
-5.23685336e-01 -4.95597363e-01 1.00378883e+00 6.37153924e-01
7.50101089e-01 -2.15223148e-01 2.03355655e-01 5.95897138e-02
-1.81009397e-01 7.99692944e-02 3.96970898e-01 5.13008833e-01
-2.95947671e-01 -6.52431786e-01 -2.30647936e-01 8.38847220e-01
-7.31541932e-01 -2.70935923e-01 -2.48402834e-01 1.00106585e+00
-7.61621445e-02 8.65824997e-01 6.31790340e-01 -3.50816905e-01
3.24839950e-01 -2.72126961e-03 7.83362806e-01 -4.10124511e-01
-4.83422756e-01 -6.02044426e-02 -8.99120197e-02 -8.58463883e-01
-7.00244486e-01 -6.25879884e-01 -8.09732914e-01 -6.96078837e-01
-6.11196756e-01 -6.78341866e-01 6.38773024e-01 7.00160682e-01
6.84634864e-01 6.31455839e-01 3.32023531e-01 -1.17779636e+00
-1.24078140e-01 -1.46618736e+00 -3.65296394e-01 2.55535424e-01
8.61432612e-01 -1.01695824e+00 -2.85297424e-01 4.29467224e-02] | [14.124470710754395, 0.06657512485980988] |
19cd41c1-6e46-428c-b8da-922110d06872 | dgc-net-dense-geometric-correspondence | 1810.08393 | null | http://arxiv.org/abs/1810.08393v2 | http://arxiv.org/pdf/1810.08393v2.pdf | DGC-Net: Dense Geometric Correspondence Network | This paper addresses the challenge of dense pixel correspondence estimation
between two images. This problem is closely related to optical flow estimation
task where ConvNets (CNNs) have recently achieved significant progress. While
optical flow methods produce very accurate results for the small pixel
translation and limited appearance variation scenarios, they hardly deal with
the strong geometric transformations that we consider in this work. In this
paper, we propose a coarse-to-fine CNN-based framework that can leverage the
advantages of optical flow approaches and extend them to the case of large
transformations providing dense and subpixel accurate estimates. It is trained
on synthetic transformations and demonstrates very good performance to unseen,
realistic, data. Further, we apply our method to the problem of relative camera
pose estimation and demonstrate that the model outperforms existing dense
approaches. | ['Esa Rahtu', 'Marc Pollefeys', 'Torsten Sattler', 'Juho Kannala', 'Iaroslav Melekhov', 'Aleksei Tiulpin'] | 2018-10-19 | null | null | null | null | ['dense-pixel-correspondence-estimation'] | ['computer-vision'] | [ 1.27716571e-01 -2.61105806e-01 -1.83380499e-01 -1.38505742e-01
-4.55238581e-01 -5.74038565e-01 5.91917217e-01 -4.24586236e-01
-4.91080284e-01 8.85271192e-01 3.08321357e-01 -1.28212675e-01
1.78270251e-01 -6.02905810e-01 -7.21776247e-01 -2.10937262e-01
3.11519891e-01 3.52924109e-01 3.20766509e-01 -2.08097428e-01
3.05729866e-01 6.73726857e-01 -1.13060105e+00 -2.12425470e-01
7.02593386e-01 7.21273959e-01 -1.10982835e-01 7.88936257e-01
1.96135208e-01 1.05275750e+00 -2.21395716e-01 -3.97299826e-01
6.26728892e-01 -2.84516007e-01 -1.06380367e+00 3.31602156e-01
1.46627152e+00 -9.35079634e-01 -7.30777919e-01 9.03703988e-01
2.79658347e-01 2.77935654e-01 3.87240857e-01 -1.26497817e+00
-5.68055153e-01 -1.07599139e-01 -8.50955069e-01 6.42669022e-01
5.41480482e-01 3.06889057e-01 8.90279174e-01 -9.66446102e-01
9.27709162e-01 1.45830381e+00 8.58709693e-01 3.33314508e-01
-1.41650569e+00 -3.11889410e-01 1.23733759e-01 1.35772079e-01
-1.06861663e+00 -5.59348881e-01 5.75305045e-01 -5.30406535e-01
9.55217183e-01 -1.13359377e-01 5.98754704e-01 9.95508730e-01
3.56796607e-02 7.56469607e-01 9.55940723e-01 -2.52248496e-01
-1.00767389e-01 -3.42282593e-01 -1.62622437e-01 7.19594300e-01
2.85326362e-01 3.50656390e-01 -3.90143901e-01 1.92320153e-01
1.28472733e+00 9.90353748e-02 -6.59923017e-01 -4.78817701e-01
-1.53756690e+00 7.83305109e-01 7.63619840e-01 1.86761349e-01
-1.86250091e-01 7.19973862e-01 -1.30932722e-02 7.56592304e-02
4.85463679e-01 5.26315629e-01 -2.52362907e-01 -2.73116738e-01
-1.19094419e+00 3.65913779e-01 7.98300862e-01 1.20817077e+00
9.66690421e-01 2.77444512e-01 -6.75910786e-02 4.33417261e-01
-1.01718428e-02 3.40340436e-01 2.11671904e-01 -1.58698499e+00
5.03257453e-01 1.23956606e-01 3.89232129e-01 -1.21010983e+00
-3.70107114e-01 -4.33875561e-01 -6.92785919e-01 2.42180735e-01
8.98872256e-01 -1.61320176e-02 -7.82684207e-01 1.53823352e+00
3.35824072e-01 4.99898344e-01 -1.71884075e-01 1.00925696e+00
4.89563286e-01 4.84767079e-01 -1.94116980e-01 -5.85130118e-02
8.66337180e-01 -1.43319428e+00 -7.11252749e-01 -5.22878170e-01
3.14725041e-01 -1.05687535e+00 7.18681872e-01 1.17482170e-01
-1.42878366e+00 -5.75758040e-01 -8.43150258e-01 -5.76300919e-01
-6.67340830e-02 -1.03354901e-01 7.75333285e-01 4.06647295e-01
-1.40207338e+00 7.67855227e-01 -9.08933222e-01 -6.97684169e-01
5.90262055e-01 3.82897645e-01 -4.82710242e-01 -3.87133151e-01
-8.50631893e-01 9.89903510e-01 8.31075311e-02 3.28677624e-01
-7.43606389e-01 -8.78458917e-01 -1.03794301e+00 -3.51308361e-02
2.88580507e-01 -1.09409010e+00 1.28518212e+00 -9.66915250e-01
-1.40176976e+00 7.17050850e-01 -3.56252313e-01 -5.06567001e-01
9.94983792e-01 -4.51075733e-01 2.11406663e-01 2.75556743e-01
6.06296398e-02 1.13158774e+00 8.04174602e-01 -9.48363662e-01
-5.96847713e-01 -1.06553987e-01 4.11614567e-01 8.03904608e-02
-2.18195751e-01 -7.68656656e-02 -4.00468767e-01 -6.78139567e-01
-1.24164216e-01 -1.07477081e+00 -4.10852164e-01 6.09111845e-01
-3.25852007e-01 3.21017772e-01 9.49812114e-01 -4.06191289e-01
6.23094976e-01 -1.73823953e+00 2.13083625e-02 -2.60725617e-01
4.02274013e-01 3.80265981e-01 -1.43440157e-01 2.42244869e-01
1.49638191e-01 -2.33794227e-01 -3.26123685e-01 -5.46203136e-01
-1.28098786e-01 2.54109651e-01 -2.34405324e-01 7.58009791e-01
3.84089798e-01 1.28243899e+00 -1.13867736e+00 -5.66481292e-01
6.87522888e-01 6.96417391e-01 -7.51363039e-01 1.48657903e-01
8.55668709e-02 6.76101029e-01 -2.63711870e-01 4.02719736e-01
9.44665551e-01 -3.88315737e-01 -1.11752048e-01 -5.31414628e-01
-1.54427469e-01 -2.01441973e-01 -1.01141691e+00 1.88232553e+00
-5.81337094e-01 1.29472995e+00 5.41419499e-02 -7.03457594e-01
7.33569741e-01 -1.88445076e-02 6.92878604e-01 -3.65667701e-01
2.14923829e-01 1.04680255e-01 -3.16588171e-02 -3.40104878e-01
7.13561237e-01 -9.40226242e-02 4.57125396e-01 3.52839440e-01
9.03710201e-02 -5.14842868e-01 3.11853349e-01 2.18021005e-01
1.04788935e+00 5.46656668e-01 3.71414632e-01 -2.54588366e-01
6.51461422e-01 2.93066911e-03 4.82360154e-01 6.15531564e-01
-5.72222531e-01 1.12241125e+00 2.00022474e-01 -8.40702116e-01
-1.25622308e+00 -9.52224076e-01 -7.77284056e-02 4.57800895e-01
5.12136996e-01 -2.18318045e-01 -7.36817539e-01 -5.74306667e-01
1.61426798e-01 1.71590611e-01 -6.63077295e-01 2.24187508e-01
-7.64285505e-01 -4.79464561e-01 3.77698481e-01 8.09200943e-01
8.89913797e-01 -6.81817472e-01 -6.29380643e-01 2.94994265e-01
-4.21094149e-01 -1.88700771e+00 -9.18121219e-01 -4.74830568e-01
-8.29049706e-01 -1.21724617e+00 -8.46149027e-01 -8.41845155e-01
6.07090652e-01 4.48664606e-01 1.19968951e+00 -1.19209886e-02
-4.51880336e-01 5.27100503e-01 3.60125601e-02 3.77108268e-02
-1.42408654e-01 2.69350819e-02 -1.89978018e-01 1.67086348e-01
2.32035935e-01 -6.61965311e-01 -9.80122149e-01 4.66643125e-01
-8.76716316e-01 1.05807774e-01 2.98023671e-01 8.12182844e-01
2.42756426e-01 -6.36641383e-01 5.61847761e-02 -7.17633903e-01
2.61148870e-01 -4.67079096e-02 -6.58914506e-01 -2.40127370e-02
-4.16757643e-01 4.95513603e-02 6.07477903e-01 -2.89562374e-01
-1.04496932e+00 4.17209864e-01 -2.90581398e-02 -7.20534325e-01
-1.47787556e-01 -1.10288829e-01 2.51542091e-01 -7.30443656e-01
6.03387177e-01 -8.48590881e-02 3.19765857e-03 -1.10337824e-01
4.37375218e-01 2.18472824e-01 8.12222660e-01 -3.42934340e-01
1.16796982e+00 1.08637309e+00 4.13393766e-01 -6.57653928e-01
-9.78479147e-01 -4.43288684e-01 -1.16986370e+00 -2.37669647e-01
9.92225528e-01 -9.28479910e-01 -6.38557553e-01 3.59353960e-01
-1.41355205e+00 -4.50088114e-01 -3.43195200e-01 6.84585929e-01
-9.93060827e-01 5.99418283e-01 -7.40975320e-01 -4.11309540e-01
-6.16129264e-02 -1.31996846e+00 1.23506820e+00 2.03370005e-01
-1.97296962e-01 -1.49355257e+00 1.57493025e-01 2.16374367e-01
7.27918983e-01 6.08862638e-01 8.89105797e-02 1.88205868e-01
-1.09922969e+00 -1.84602961e-02 -5.00925362e-01 2.39239633e-01
1.79547340e-01 1.69115677e-01 -1.11311972e+00 -3.46661508e-01
-2.35154167e-01 -4.27888751e-01 9.81489062e-01 6.29798412e-01
8.90011251e-01 3.94525414e-04 -1.91536933e-01 1.19074667e+00
1.42649746e+00 -2.69523472e-01 8.02475274e-01 3.56086791e-01
1.11678183e+00 4.61557180e-01 5.26242971e-01 8.99312794e-02
4.75959003e-01 8.92401755e-01 6.86719954e-01 -5.34461677e-01
-5.84600806e-01 -2.69315749e-01 1.05215751e-01 3.28845412e-01
-2.90362298e-01 -2.07252219e-01 -7.18646169e-01 6.34412110e-01
-1.93501770e+00 -8.95524979e-01 -2.64850229e-01 1.90006518e+00
4.24351454e-01 -8.32183883e-02 -6.27810135e-02 -1.39328822e-01
7.21014261e-01 4.59562033e-01 -4.21459675e-01 -3.48031342e-01
-1.51714340e-01 2.08454356e-01 6.22459173e-01 8.53056788e-01
-1.19548690e+00 1.23906755e+00 7.02864933e+00 3.30717057e-01
-9.71830726e-01 5.82182966e-02 6.72385871e-01 -5.39217368e-02
-1.46666273e-01 1.91406831e-01 -6.62542105e-01 1.61003828e-01
3.74492258e-01 -2.40842216e-02 3.28281790e-01 5.25111973e-01
2.53434330e-01 -1.95834994e-01 -1.35584390e+00 1.28076959e+00
1.46314651e-01 -1.51156032e+00 -7.32944086e-02 2.38728121e-01
1.15426970e+00 1.60223171e-01 1.83089599e-01 -3.21456522e-01
4.42694604e-01 -1.06117988e+00 5.48552394e-01 4.09795642e-01
7.27660239e-01 -3.75138551e-01 6.01050198e-01 2.31852755e-02
-1.10598290e+00 -3.72179672e-02 -5.68780959e-01 -2.40902841e-01
5.86479545e-01 4.84215707e-01 -5.42884767e-01 3.17559391e-01
6.30240977e-01 1.47283745e+00 -6.09471858e-01 1.21522772e+00
-1.12563409e-01 1.39300629e-01 -2.87584126e-01 3.53634685e-01
5.47833145e-01 -3.73215050e-01 4.50097680e-01 1.21951509e+00
2.94689149e-01 -1.96081355e-01 1.30253315e-01 1.06447470e+00
-1.63159505e-01 -8.75363797e-02 -7.79832661e-01 4.45549250e-01
7.89155811e-02 1.39870965e+00 -6.73926651e-01 -1.97410077e-01
-6.38659537e-01 1.15825319e+00 5.11206627e-01 6.29797697e-01
-6.30728781e-01 -9.81790051e-02 9.71757412e-01 1.88750014e-01
4.47074205e-01 -4.75744516e-01 1.09132461e-01 -1.49972427e+00
4.01553325e-02 -3.69273394e-01 6.26327917e-02 -9.53284264e-01
-1.15067816e+00 5.30409396e-01 -1.60392091e-01 -1.45510960e+00
-5.08783996e-01 -7.84054697e-01 -6.98875189e-01 8.09283197e-01
-2.02335334e+00 -1.27444959e+00 -7.62920260e-01 7.59721339e-01
6.21262789e-01 3.21713716e-01 3.86070848e-01 3.13167155e-01
-3.50027949e-01 3.80245030e-01 1.19913213e-01 4.04573500e-01
8.77038300e-01 -1.27457345e+00 7.37734377e-01 1.25119448e+00
1.72544926e-01 4.60893363e-01 6.16044044e-01 -1.92172468e-01
-1.27680039e+00 -1.32791197e+00 8.50649297e-01 -7.08388209e-01
6.28466547e-01 -2.22445577e-01 -5.21828234e-01 9.96970236e-01
3.64291221e-01 8.31779003e-01 2.72205025e-02 -3.83212268e-01
-3.16941559e-01 -4.08367217e-02 -9.83897150e-01 4.49982643e-01
1.26170814e+00 -4.57553029e-01 -2.14521185e-01 3.88865352e-01
4.23066020e-01 -7.66476333e-01 -8.63939881e-01 1.68037593e-01
5.00244141e-01 -1.20649648e+00 1.21074247e+00 -5.37709892e-01
6.78781033e-01 -3.30951989e-01 4.16566990e-02 -1.34142315e+00
-3.30933392e-01 -9.71220970e-01 -1.19345702e-01 6.95498407e-01
1.35112092e-01 -4.87543315e-01 8.13717008e-01 6.36158824e-01
-2.16963571e-02 -2.97662556e-01 -7.84392655e-01 -7.54091382e-01
1.34332389e-01 -2.83915341e-01 4.53706682e-01 9.48966384e-01
-3.75800520e-01 1.71404183e-02 -6.71194613e-01 -1.71317294e-01
8.27329636e-01 1.54926404e-01 1.10647678e+00 -9.95643735e-01
-2.97230333e-01 -4.36235249e-01 -9.51639116e-01 -1.54431772e+00
4.19281721e-01 -6.26972616e-01 5.42621836e-02 -1.52219188e+00
-5.92409540e-03 -9.97658372e-02 1.90687910e-01 1.96251228e-01
-3.49632651e-01 8.49455059e-01 3.49394888e-01 1.92236021e-01
-5.38697422e-01 4.28325415e-01 1.69447482e+00 -1.63709372e-01
9.62861255e-02 -1.07334644e-01 -4.72309440e-01 6.17688239e-01
5.02010643e-01 -3.30128819e-02 -3.79054457e-01 -7.97321379e-01
-8.86842608e-02 -1.56576693e-01 6.26115382e-01 -1.07280576e+00
2.52602786e-01 -1.47479191e-01 6.03827834e-01 -2.59510607e-01
3.92779589e-01 -8.73772800e-01 -9.09041539e-02 4.66003418e-01
-3.45012575e-01 3.47939491e-01 1.06886752e-01 5.89052677e-01
-3.76564026e-01 1.80465087e-01 9.46248710e-01 -2.08749220e-01
-7.26930857e-01 8.46023679e-01 2.45871209e-02 4.14817184e-01
1.03950179e+00 -4.67661977e-01 -3.82219136e-01 -8.03755641e-01
-6.75572991e-01 9.70317721e-02 7.48751819e-01 4.97920364e-01
6.21819615e-01 -1.40589511e+00 -7.52808452e-01 1.29131600e-01
8.19960143e-03 1.01754814e-01 -1.02970682e-01 1.10728633e+00
-1.10214210e+00 4.53285396e-01 -4.08322841e-01 -8.72613430e-01
-8.80959451e-01 6.36500418e-01 6.93699956e-01 -4.12546583e-02
-7.45069742e-01 6.32782280e-01 5.70297122e-01 -1.12461455e-01
5.45435175e-02 -4.20379460e-01 -1.39951492e-02 -2.96835363e-01
5.74608922e-01 4.68698084e-01 -1.84767283e-02 -1.01815701e+00
-2.25764558e-01 1.10599899e+00 -4.23243232e-02 2.43161083e-03
1.26267064e+00 -4.06860650e-01 6.08323142e-02 3.84340994e-02
1.53129756e+00 -8.99881423e-02 -2.01761627e+00 -2.83219129e-01
-3.15103233e-01 -1.09705031e+00 1.12043016e-01 -1.45547822e-01
-1.60681689e+00 1.21726000e+00 3.85005772e-01 -1.13212697e-01
8.82034302e-01 -1.77608833e-01 9.45630789e-01 1.12182990e-01
1.32210985e-01 -7.59796083e-01 1.38868794e-01 5.95337510e-01
5.01820266e-01 -1.42739511e+00 6.10413365e-02 -5.68295240e-01
-2.17273623e-01 1.49781799e+00 7.81012595e-01 -4.41813797e-01
5.68918765e-01 3.11640978e-01 1.06937431e-01 7.15492591e-02
-4.88689274e-01 -4.43510175e-01 2.49477416e-01 9.80121672e-01
4.21005160e-01 -4.38176662e-01 5.20894788e-02 -8.71638417e-01
-8.72748196e-02 1.65218860e-01 8.90569448e-01 7.29447484e-01
-9.80465412e-02 -9.98171449e-01 -3.13259751e-01 6.83159381e-02
-2.84576356e-01 -1.22606531e-01 -3.33724409e-01 1.05225515e+00
-2.16611400e-01 7.75892615e-01 4.41781253e-01 2.19026040e-02
3.68279755e-01 -4.84082103e-01 8.94615054e-01 -2.11547568e-01
-4.38544512e-01 3.89312953e-03 -1.29075035e-01 -1.09299695e+00
-9.13332224e-01 -7.64968216e-01 -7.53881216e-01 -5.69605410e-01
-1.27715781e-01 -2.72358090e-01 2.78133094e-01 8.55567455e-01
1.08553201e-01 2.42977872e-01 6.41418934e-01 -1.27879775e+00
-2.64559150e-01 -7.09660232e-01 -3.66119415e-01 7.90375531e-01
8.48805130e-01 -5.51097274e-01 -4.89899158e-01 2.61923283e-01] | [8.728808403015137, -1.9220070838928223] |
529db013-a0e6-4aba-890a-145abd3d3834 | signal-processing-with-optical-quadratic | 2212.00660 | null | https://arxiv.org/abs/2212.00660v2 | https://arxiv.org/pdf/2212.00660v2.pdf | Signal processing with optical quadratic random sketches | Random data sketching (or projection) is now a classical technique enabling, for instance, approximate numerical linear algebra and machine learning algorithms with reduced computational complexity and memory. In this context, the possibility of performing data processing (such as pattern detection or classification) directly in the sketched domain without accessing the original data was previously achieved for linear random sketching methods and compressive sensing. In this work, we show how to estimate simple signal processing tasks (such as deducing local variations in a image) directly using random quadratic projections achieved by an optical processing unit. The same approach allows for naive data classification methods directly operated in the sketched domain. We report several experiments confirming the power of our approach. | ['Laurent Jacques', 'Laurent Daudet', 'Vincent Schellekens', 'Rémi Delogne'] | 2022-12-01 | null | null | null | null | ['compressive-sensing'] | ['computer-vision'] | [ 6.99913144e-01 1.09505311e-01 2.60622978e-01 -5.02152974e-03
-2.21888274e-01 -6.70520067e-01 6.86163604e-01 -2.52920449e-01
-5.62324226e-01 7.01907218e-01 -1.20255873e-01 -3.74170184e-01
-2.65453368e-01 -8.74810159e-01 -5.73165357e-01 -7.99410462e-01
-1.48989350e-01 4.58460748e-01 -1.55349776e-01 1.68520302e-01
4.52454060e-01 1.12017715e+00 -1.59380400e+00 3.35462630e-01
3.22723359e-01 1.27786088e+00 1.95250764e-01 8.52413416e-01
-1.22990102e-01 4.78621781e-01 -2.42672667e-01 -2.80816793e-01
6.50996983e-01 -3.27274948e-01 1.36507265e-02 2.32608363e-01
5.45888186e-01 -5.24002194e-01 -4.99518096e-01 8.42668056e-01
2.34396532e-01 -8.85468051e-02 6.97537661e-01 -7.99662709e-01
-4.66874480e-01 3.54426235e-01 -1.71858847e-01 -4.64411050e-01
6.29098773e-01 2.67814603e-02 5.54627180e-01 -1.38607788e+00
8.39456379e-01 9.20043051e-01 5.10747612e-01 2.21659079e-01
-1.61054146e+00 -3.09171975e-01 -5.99615633e-01 1.43709823e-01
-1.72189915e+00 -7.07482934e-01 1.11839485e+00 -3.21934104e-01
4.17651027e-01 3.73654276e-01 5.59088886e-01 6.38577878e-01
-1.72185265e-02 4.86051470e-01 1.28121662e+00 -7.63369501e-01
6.19117022e-01 3.14492613e-01 -1.83709145e-01 8.88961852e-01
3.19216132e-01 9.99017656e-02 -5.73886216e-01 -1.90262377e-01
7.86599576e-01 4.44710314e-01 -5.62780559e-01 -6.02529287e-01
-1.45957184e+00 5.86738706e-01 1.14098653e-01 3.95811975e-01
-6.19049847e-01 6.21414594e-02 8.03878456e-02 6.19101346e-01
1.46176487e-01 5.39538145e-01 6.46022782e-02 2.14961156e-01
-1.35322690e+00 8.69792402e-02 1.14983666e+00 9.03819859e-01
8.51835072e-01 7.41352588e-02 -3.35045010e-02 2.31519058e-01
-1.12906940e-01 1.02095652e+00 4.17240560e-01 -9.23301518e-01
3.30123842e-01 1.47767067e-01 2.18802392e-01 -1.28210056e+00
-1.81315705e-01 -1.37451291e-02 -1.18662333e+00 6.78275526e-01
6.48961723e-01 -1.55920476e-01 -4.43436056e-01 1.24212956e+00
-9.59634501e-03 1.43123418e-01 7.35165104e-02 6.83582664e-01
1.66621894e-01 7.39712358e-01 -6.78229272e-01 -5.35723329e-01
1.03089845e+00 -2.84239680e-01 -6.85155809e-01 3.04538369e-01
3.17982286e-02 -1.00792658e+00 8.26110184e-01 1.08914840e+00
-9.07033086e-01 -4.03436065e-01 -1.13276851e+00 -1.67940453e-01
-2.17188999e-01 4.59450245e-01 4.62561965e-01 6.19684815e-01
-7.06598341e-01 9.63275909e-01 -5.65221906e-01 -2.00953722e-01
3.44872385e-01 2.22597778e-01 -7.07542598e-01 -3.87612194e-01
-4.26059812e-01 5.05712271e-01 8.62819701e-02 1.29300117e-01
-2.69285172e-01 -6.61019266e-01 -4.60238189e-01 2.64617771e-01
3.86952281e-01 -5.26175797e-01 5.19343734e-01 -5.99232435e-01
-2.02273345e+00 8.11147213e-01 -2.22506344e-01 -6.56480074e-01
4.94515985e-01 -1.80551603e-01 -8.38909745e-02 4.88026232e-01
-5.52276909e-01 1.68417647e-01 1.74687839e+00 -7.77597845e-01
2.58766234e-01 -3.95115912e-01 -3.28906745e-01 -2.29644239e-01
-3.04654419e-01 -4.37484860e-01 1.93316877e-01 -6.32864773e-01
7.57022262e-01 -8.70036542e-01 -2.42667779e-01 8.99524093e-01
-2.94242918e-01 3.39014083e-01 1.04518962e+00 -6.52720988e-01
8.70363057e-01 -2.27390337e+00 2.94438213e-01 5.38661420e-01
2.81827629e-01 2.65134901e-01 1.18628979e-01 8.14192832e-01
-8.35953429e-02 -5.04543841e-01 -3.53983730e-01 -4.43619072e-01
-4.07875441e-02 5.69422171e-02 -9.47177529e-01 8.00266504e-01
-3.56762460e-03 8.32525551e-01 -5.99796891e-01 -3.84853542e-01
5.43409646e-01 3.27067107e-01 -6.99236572e-01 2.58242577e-01
-3.12034395e-02 5.29144049e-01 -1.81599766e-01 4.54570204e-01
8.70355248e-01 1.55752510e-01 1.61947682e-01 -5.10297060e-01
-2.85364091e-01 -3.84246185e-02 -1.75861609e+00 1.77949858e+00
-8.24513078e-01 8.32316399e-01 5.04463375e-01 -1.30218697e+00
1.09774852e+00 4.25607771e-01 4.49308127e-01 -4.17625636e-01
-3.97072703e-01 3.46046507e-01 -4.50984508e-01 -4.04506773e-01
4.28782880e-01 -4.64620709e-01 6.44158050e-02 6.81856871e-01
-6.58150837e-02 -4.81844217e-01 -3.09242845e-01 -7.64778107e-02
1.12872386e+00 -8.57611001e-03 6.33004546e-01 -3.39567661e-01
5.16764641e-01 -2.20982134e-01 -1.36226431e-01 7.24335372e-01
4.08682257e-01 6.76885545e-01 2.49624208e-01 -5.67773998e-01
-1.16699028e+00 -1.16077900e+00 -1.03176065e-01 2.30736539e-01
-1.50172919e-01 -4.06533182e-01 -3.41527253e-01 -9.20070931e-02
8.92753974e-02 4.68136847e-01 -1.82055071e-01 2.00889513e-01
-7.02975273e-01 9.61261801e-03 3.36808264e-01 -2.03030586e-01
4.69553083e-01 -6.74520671e-01 -7.60394037e-01 9.86967310e-02
2.31117085e-01 -1.13330019e+00 2.96212807e-02 -3.15247104e-02
-1.11773443e+00 -6.58146381e-01 -8.57324123e-01 -2.43087307e-01
7.44351327e-01 5.15273154e-01 6.41701221e-01 -1.74523890e-01
-5.60947299e-01 4.95522678e-01 9.29256808e-03 -1.04536012e-01
-1.94069922e-01 -3.94948363e-01 3.53764892e-01 5.07271409e-01
-1.34097546e-01 -1.08057678e+00 -5.39676964e-01 1.28294468e-01
-1.10872018e+00 -4.91896905e-02 8.60802352e-01 8.42287540e-01
5.23889542e-01 -1.71406165e-01 9.03037488e-02 -6.93420291e-01
5.71810484e-01 -2.28393257e-01 -7.49531746e-01 1.02297492e-01
-3.46469849e-01 4.32713300e-01 1.24018109e+00 -5.93473613e-01
-6.93816900e-01 4.26597416e-01 -1.23468414e-01 -6.16980612e-01
-3.31286073e-01 2.13024363e-01 2.63182912e-02 -5.30038297e-01
7.57658541e-01 6.13599002e-01 3.54485959e-01 -6.61384106e-01
5.68639874e-01 6.29577339e-01 4.86967921e-01 -4.03451025e-01
1.18161535e+00 9.84866500e-01 7.98594236e-01 -1.49172843e+00
-1.66951418e-01 -3.12263817e-01 -6.77353799e-01 -2.12079100e-02
4.13406700e-01 -4.60686982e-01 -6.59172773e-01 2.92557418e-01
-1.17862535e+00 8.77409708e-03 -8.15203190e-01 6.14558756e-01
-6.34226143e-01 6.28325522e-01 -2.82956719e-01 -9.30299163e-01
-1.39614046e-01 -6.49399400e-01 9.95475829e-01 -3.16634476e-01
-6.46885438e-03 -7.08806872e-01 -1.14071026e-01 -2.38123089e-01
6.89726353e-01 9.52813774e-02 5.46575665e-01 -1.63637444e-01
-9.25109208e-01 -5.34584761e-01 -1.30282044e-01 2.49436378e-01
1.63239222e-02 -3.60052854e-01 -1.03157353e+00 -2.51031786e-01
3.83425236e-01 5.17524891e-02 7.61130035e-01 1.09531604e-01
1.37838507e+00 -4.36019003e-01 -2.86223382e-01 7.50939965e-01
1.55839717e+00 -4.11240011e-01 7.60027528e-01 -6.21869266e-01
2.19995126e-01 5.00204861e-01 3.38039041e-01 6.65872455e-01
-6.71188056e-01 1.06783497e+00 6.98921829e-02 2.74627715e-01
-1.38461247e-01 -3.61031651e-01 2.25335404e-01 4.20496881e-01
-1.02934860e-01 7.39527419e-02 -3.55358303e-01 6.30311221e-02
-1.26955914e+00 -1.37919080e+00 -9.14656296e-02 2.49358559e+00
2.78063327e-01 -1.57832339e-01 -3.01965117e-01 6.30115807e-01
3.30651164e-01 8.22569281e-02 -1.61941394e-01 -3.14592063e-01
-8.32787082e-02 7.26671338e-01 5.67730784e-01 7.35399067e-01
-5.55069506e-01 3.99888992e-01 5.88519526e+00 8.28139186e-01
-1.49416852e+00 -1.26336142e-01 4.82211635e-02 -1.21822819e-01
-4.00975645e-01 1.26147270e-01 -3.12750369e-01 3.46784800e-01
5.38237870e-01 -6.34367540e-02 9.74385142e-01 5.19753158e-01
-4.10218276e-02 -2.71544337e-01 -1.27203953e+00 1.39974391e+00
4.65892628e-02 -1.55666232e+00 1.98333308e-01 2.28855327e-01
2.36355633e-01 -5.39386809e-01 1.83638573e-01 -8.19787756e-02
-6.33546650e-01 -8.22265208e-01 5.92673659e-01 1.07199287e+00
1.05244005e+00 -1.28054366e-01 1.77377954e-01 6.60281301e-01
-8.97726238e-01 -8.95057172e-02 -3.17191064e-01 -4.32665944e-01
3.02823275e-01 1.25050151e+00 -9.37108278e-01 3.56735170e-01
-6.93273991e-02 3.86285782e-01 2.55090203e-02 7.90112078e-01
3.39435749e-02 6.48229837e-01 -9.47376132e-01 -1.20539702e-01
-9.04494151e-02 -8.99271429e-01 9.96762991e-01 1.08574140e+00
8.54249835e-01 3.62174660e-01 -1.48223221e-01 1.10321403e+00
1.40453279e-01 9.37010422e-02 -1.22627032e+00 2.34310329e-03
2.16194376e-01 1.14905083e+00 -6.50230944e-01 -2.08966196e-01
-3.06993991e-01 1.29991770e+00 -8.97311792e-02 4.46777046e-01
-1.34580448e-01 -6.57734334e-01 3.61083984e-01 4.81576025e-01
5.25506616e-01 -9.51981544e-01 -3.77382815e-01 -1.30252361e+00
3.47565174e-01 -4.09786642e-01 -2.00280845e-01 -6.81595802e-01
-8.49929988e-01 1.60474077e-01 -1.10490113e-01 -1.66895247e+00
-3.53363484e-01 -9.25345361e-01 -4.75586951e-01 8.75853598e-01
-1.14577031e+00 -7.21345723e-01 -1.47628620e-01 8.26756895e-01
2.20953718e-01 -2.02694342e-01 9.89786804e-01 1.67433381e-01
1.14287704e-01 2.89927393e-01 2.56055206e-01 -1.80252180e-01
4.95151490e-01 -1.02880633e+00 -1.64723903e-01 9.68284905e-01
4.68846112e-01 8.00433993e-01 9.01603937e-01 -2.55312383e-01
-2.12380481e+00 -4.77075785e-01 6.76176310e-01 2.07683310e-01
6.72346711e-01 -5.43928742e-01 -6.66648567e-01 4.09641951e-01
5.06065264e-02 3.22672099e-01 4.17785853e-01 -3.99805665e-01
-1.35442644e-01 -3.53794873e-01 -1.50752068e+00 4.36673850e-01
8.33112061e-01 -7.90606260e-01 -4.06852871e-01 4.69071299e-01
1.36315390e-01 -1.95912138e-01 -7.20235467e-01 -1.23845905e-01
8.72317314e-01 -9.88994420e-01 1.08774865e+00 -2.23139063e-01
3.97486031e-01 -2.71186233e-01 -4.96612221e-01 -1.11119306e+00
2.09330112e-01 -1.20784521e+00 -3.13939720e-01 5.59388936e-01
3.85774598e-02 -8.78429890e-01 8.38666141e-01 2.03079790e-01
3.00845742e-01 -5.00099003e-01 -1.10595095e+00 -6.48125529e-01
-3.25294346e-01 -5.86580753e-01 3.34570497e-01 7.43845940e-01
1.62501857e-01 -8.77959579e-02 -5.24779320e-01 3.91325384e-01
1.00749731e+00 6.01803958e-01 8.30172479e-01 -1.07844961e+00
-7.58399487e-01 -1.28275990e-01 -8.10196579e-01 -1.57199013e+00
1.98400505e-02 -9.05554652e-01 -2.14825153e-01 -9.21151876e-01
-4.73866254e-01 -5.85898161e-01 1.75220430e-01 -1.81305423e-01
5.96828699e-01 4.60237443e-01 2.52129048e-01 2.55850345e-01
-6.53066710e-02 2.23557830e-01 1.02453768e+00 8.59808177e-02
-1.75457988e-02 3.37862283e-01 -6.98353946e-02 5.68224072e-01
4.91168886e-01 -2.99892098e-01 -1.76539317e-01 -1.91074669e-01
5.20223320e-01 4.50548321e-01 5.60766995e-01 -1.38465238e+00
6.36942983e-01 -7.72219431e-03 2.89395332e-01 -4.00235474e-01
6.14218116e-01 -1.25108719e+00 2.80829579e-01 5.79624057e-01
-1.83295056e-01 -3.99373025e-01 -1.42923936e-01 5.75215459e-01
-3.29903901e-01 -5.65978289e-01 5.77690065e-01 -1.68155178e-01
-3.99542451e-01 1.19033456e-02 -4.58952665e-01 -5.57656169e-01
9.57101405e-01 -1.93091974e-01 1.15664378e-01 -7.38950670e-01
-1.00006986e+00 -6.47157788e-01 4.98956025e-01 -4.64255691e-01
1.16701150e+00 -1.27407515e+00 -6.37100875e-01 8.62918437e-01
-5.32486774e-02 -2.41063118e-01 4.38155346e-02 8.56188536e-01
-8.57399940e-01 6.07834637e-01 -2.84947693e-01 -4.96999204e-01
-1.17177391e+00 8.82253826e-01 2.42760658e-01 -1.49238825e-01
-7.89077699e-01 3.40878844e-01 -3.27119648e-01 -1.51943281e-01
-1.54470161e-01 -3.98996562e-01 4.94417667e-01 -9.62556750e-02
7.43108273e-01 6.92773938e-01 2.11702853e-01 -1.17733069e-01
3.84798972e-03 8.03252637e-01 4.02911723e-01 -4.58415806e-01
1.21939266e+00 -3.57471369e-02 -3.58862042e-01 3.36095989e-01
1.43891978e+00 5.68803847e-01 -1.03720331e+00 -4.85638559e-01
-2.04029128e-01 -5.96809566e-01 2.06969649e-01 -1.56633720e-01
-6.19308114e-01 1.23956823e+00 5.33008635e-01 4.87271905e-01
1.22049081e+00 -1.28766224e-01 5.39428592e-01 1.10998356e+00
9.32752132e-01 -6.71381176e-01 -2.86858946e-01 -6.42318577e-02
1.16449082e+00 -8.82323325e-01 4.17575657e-01 -5.10626853e-01
1.06407180e-01 1.72613502e+00 -5.87099791e-01 -4.22765732e-01
1.00138605e+00 5.10846436e-01 -4.24898148e-01 -4.43636589e-02
-5.14183581e-01 -7.38873333e-02 2.39418596e-01 3.20969194e-01
-5.28729148e-03 2.53781468e-01 -3.66950035e-01 -1.53307572e-01
-1.05642423e-03 4.10407931e-01 7.09621370e-01 8.65672529e-01
-3.53879571e-01 -1.08539188e+00 -6.54525101e-01 4.98379499e-01
-1.27189308e-01 -1.66624874e-01 -7.77946562e-02 3.48554432e-01
-1.06484359e-02 4.79361951e-01 3.53849269e-02 -5.80487475e-02
2.45017782e-01 3.18033516e-01 8.21236789e-01 -4.28004742e-01
2.63955295e-01 -4.32735860e-01 -2.34114602e-01 -8.68541420e-01
-3.54658842e-01 -7.79832602e-01 -8.40759516e-01 -1.40423611e-01
9.27416142e-03 -2.39394773e-02 1.25464082e+00 7.15464473e-01
2.68923759e-01 -2.42273256e-01 7.79781103e-01 -1.25117123e+00
-8.53202879e-01 -6.93619192e-01 -9.17115331e-01 2.38323152e-01
6.36288047e-01 -3.63935441e-01 -5.55554450e-01 1.26797289e-01] | [11.491315841674805, -2.284080743789673] |
8d7a2440-8ab6-41d4-a162-96075b94df6e | fast-refacing-of-mr-images-with-a-generative | 2305.16922 | null | https://arxiv.org/abs/2305.16922v1 | https://arxiv.org/pdf/2305.16922v1.pdf | Fast refacing of MR images with a generative neural network lowers re-identification risk and preserves volumetric consistency | With the rise of open data, identifiability of individuals based on 3D renderings obtained from routine structural magnetic resonance imaging (MRI) scans of the head has become a growing privacy concern. To protect subject privacy, several algorithms have been developed to de-identify imaging data using blurring, defacing or refacing. Completely removing facial structures provides the best re-identification protection but can significantly impact post-processing steps, like brain morphometry. As an alternative, refacing methods that replace individual facial structures with generic templates have a lower effect on the geometry and intensity distribution of original scans, and are able to provide more consistent post-processing results by the price of higher re-identification risk and computational complexity. In the current study, we propose a novel method for anonymised face generation for defaced 3D T1-weighted scans based on a 3D conditional generative adversarial network. To evaluate the performance of the proposed de-identification tool, a comparative study was conducted between several existing defacing and refacing tools, with two different segmentation algorithms (FAST and Morphobox). The aim was to evaluate (i) impact on brain morphometry reproducibility, (ii) re-identification risk, (iii) balance between (i) and (ii), and (iv) the processing time. The proposed method takes 9 seconds for face generation and is suitable for recovering consistent post-processing results after defacing. | ['Jonas Richiardi', 'Till Huelnhagen', 'Tobias Kober', 'Jean-Philippe Thiran', 'Bénédicte Maréchal', 'Nataliia Molchanova'] | 2023-05-26 | null | null | null | null | ['face-generation', 'brain-morphometry', 'de-identification'] | ['computer-vision', 'medical', 'natural-language-processing'] | [ 3.86063427e-01 3.10261279e-01 6.33111060e-01 -4.74332154e-01
-5.41815579e-01 -5.53278506e-01 4.81377482e-01 2.09054202e-01
-7.13127851e-01 5.97013652e-01 2.04726998e-02 1.57857407e-02
-3.29429328e-01 -6.30260110e-01 -5.01090646e-01 -6.50898695e-01
-1.48527145e-01 6.34616137e-01 6.41952753e-02 2.85403103e-01
1.03329949e-01 1.05131292e+00 -1.47386062e+00 -2.88615078e-02
9.70406890e-01 8.69590759e-01 -1.73850119e-01 2.90458947e-01
-2.61429306e-02 -5.69050610e-02 -6.72038615e-01 -7.55408049e-01
4.52445477e-01 -2.82791018e-01 -7.69538701e-01 2.65082195e-02
8.10039639e-01 -2.76159108e-01 -3.99962924e-02 1.31359601e+00
7.77602494e-01 6.99974671e-02 7.45096624e-01 -1.14759254e+00
-3.85832399e-01 4.44394529e-01 -6.25079632e-01 1.92661434e-01
2.18463168e-01 3.51305045e-02 1.32214621e-01 -4.98369902e-01
6.57662094e-01 1.15725458e+00 7.83063114e-01 7.31386721e-01
-1.57107389e+00 -9.37296569e-01 -3.10176104e-01 8.18533301e-02
-1.64834595e+00 -4.66455162e-01 6.34414017e-01 -6.34052575e-01
2.97113240e-01 4.88127708e-01 6.01901412e-01 9.89930391e-01
3.73316914e-01 -8.75568092e-02 1.60919917e+00 -1.28577799e-01
2.24030837e-01 1.48725122e-01 -9.98618975e-02 5.54426849e-01
4.58957046e-01 -7.25471452e-02 -1.10135451e-01 -4.40279931e-01
7.92723417e-01 -1.91743031e-01 -4.54653740e-01 -2.85658300e-01
-7.87975371e-01 6.36021018e-01 2.41982654e-01 5.10341525e-01
-3.37148786e-01 -1.81717962e-01 2.82219797e-01 -5.02203852e-02
5.54420829e-01 3.05175930e-01 1.80599257e-01 3.46975982e-01
-1.27026045e+00 3.27488065e-01 4.39072818e-01 3.85494262e-01
4.54694331e-01 -3.61473784e-02 -2.32653677e-01 6.65401459e-01
2.01215595e-01 4.71284240e-01 7.89007187e-01 -4.34099346e-01
2.21441925e-01 2.68752486e-01 -2.55417496e-01 -1.20552278e+00
-6.42728448e-01 -2.70915121e-01 -8.64569068e-01 5.04882097e-01
5.24273813e-01 -2.43100420e-01 -1.05247784e+00 1.75656033e+00
4.24063504e-01 -2.40745936e-02 -3.34162652e-01 9.39139664e-01
6.64774358e-01 1.72332585e-01 2.07160711e-01 -1.79825574e-01
1.40293169e+00 -1.94024816e-01 -5.81430495e-01 7.70389810e-02
2.31523547e-04 -8.13296139e-01 5.55515826e-01 1.96428150e-01
-9.48954105e-01 -3.68331909e-01 -9.42997336e-01 2.50930160e-01
-2.76116729e-01 -2.07883492e-01 1.48193702e-01 1.33328819e+00
-1.18002069e+00 7.16121137e-01 -8.60121310e-01 -2.08129719e-01
5.94425559e-01 6.23775899e-01 -7.65871286e-01 1.64785817e-01
-9.64388430e-01 9.99723971e-01 2.13661030e-01 4.67189681e-03
-6.16384387e-01 -1.01543796e+00 -6.30735159e-01 -1.32501975e-01
1.35863885e-01 -5.39943993e-01 4.79077637e-01 -8.98846447e-01
-1.34619629e+00 1.30704808e+00 3.71237285e-02 -4.25939620e-01
1.03238547e+00 3.32506709e-02 -5.99195719e-01 3.72504830e-01
2.30976269e-02 6.77030265e-01 1.20651746e+00 -1.23842645e+00
7.99425095e-02 -1.01528287e+00 -4.25297558e-01 -8.71525481e-02
-6.75539998e-03 2.15872973e-01 1.67922795e-01 -9.18098390e-01
1.95706949e-01 -9.90758359e-01 -4.53499332e-02 1.35588363e-01
-4.01920319e-01 3.42700660e-01 7.84363985e-01 -1.25318718e+00
7.80962765e-01 -2.15622544e+00 -1.04582265e-01 5.54381371e-01
3.67753446e-01 5.04490614e-01 8.58548656e-02 -1.57510877e-01
-5.09918034e-01 3.03775966e-01 -5.51894426e-01 -4.03720766e-01
-2.94379085e-01 -1.84131637e-01 1.01486869e-01 9.45602238e-01
-2.52796970e-02 6.57241762e-01 -3.72816294e-01 -5.22141218e-01
2.25135401e-01 7.61492550e-01 -4.48030472e-01 -6.02012733e-04
2.93948263e-01 8.02745044e-01 -6.78564981e-02 2.88003623e-01
1.16474473e+00 5.15743375e-01 -3.60203907e-02 -4.18142766e-01
-5.49993813e-02 -3.70993078e-01 -1.11864495e+00 1.40321052e+00
-2.13733181e-01 3.02210569e-01 3.67684901e-01 -7.96955347e-01
1.06597030e+00 4.54670191e-01 6.28650546e-01 -5.03645957e-01
4.08818424e-01 1.84757665e-01 1.20934501e-01 -3.18070322e-01
1.78975016e-01 -4.95276302e-01 2.73401201e-01 5.59277773e-01
-2.36479659e-02 -4.53858264e-02 -3.18487614e-01 -1.61135226e-01
6.22622013e-01 -9.24921706e-02 2.04863027e-02 -6.01308525e-01
7.01729894e-01 -4.81659800e-01 3.28313708e-01 2.87873805e-01
-2.33302683e-01 9.33557153e-01 3.28937382e-01 -3.25723082e-01
-9.61468875e-01 -1.13987041e+00 -3.76637697e-01 1.70474097e-01
-1.43100902e-01 1.94484383e-01 -1.44795978e+00 -5.17430425e-01
-2.65567992e-02 6.90246284e-01 -7.46848583e-01 -1.88356757e-01
-5.15574932e-01 -9.24493134e-01 9.10317361e-01 -1.34623423e-03
5.54526150e-01 -7.30897486e-01 -8.28474104e-01 2.03037355e-02
-2.15951473e-01 -9.76547182e-01 -6.84679806e-01 -2.11362645e-01
-9.27658439e-01 -1.09250760e+00 -1.02518523e+00 -3.99059862e-01
9.67686236e-01 -1.71987534e-01 6.66331410e-01 1.00112282e-01
-5.77813685e-01 2.63524979e-01 -1.98481813e-01 -4.09785807e-01
-5.28210163e-01 -1.83137104e-01 2.07174182e-01 4.88521248e-01
-9.37569812e-02 -7.80563295e-01 -6.75217211e-01 4.24132109e-01
-1.14716613e+00 -2.61009187e-01 3.59819144e-01 4.01093394e-01
5.74980199e-01 5.47962114e-02 2.90256590e-01 -9.90717232e-01
7.00951159e-01 -1.56215191e-01 -6.14231348e-01 2.41214335e-01
-8.14632058e-01 1.14293702e-01 4.16996926e-01 -3.98865938e-01
-1.16976583e+00 -9.36909616e-02 -2.00919107e-01 -5.72320580e-01
-3.82941067e-01 3.59627977e-02 -2.39214301e-01 -5.16429722e-01
7.06946373e-01 8.62032026e-02 5.41715920e-01 -7.05119371e-01
1.01296403e-01 5.45331597e-01 7.26837814e-01 -3.63406241e-01
8.17895353e-01 7.30219066e-01 2.06142709e-01 -7.51180828e-01
-8.15695152e-02 -5.81549481e-02 -8.72824490e-01 -2.53993541e-01
1.10086560e+00 -3.78428102e-01 -4.46113497e-01 6.58087373e-01
-9.89047706e-01 1.41562298e-01 -1.82347894e-02 3.84192884e-01
-2.93553889e-01 6.71813905e-01 -1.50123566e-01 -6.35209382e-01
-8.04231226e-01 -1.35478592e+00 6.29649699e-01 1.82749540e-01
-4.29633528e-01 -7.03504980e-01 -1.31909177e-01 4.97937948e-01
5.53792834e-01 7.59819508e-01 1.01840365e+00 -5.89409232e-01
-6.07601777e-02 -2.90597469e-01 -1.78453192e-01 3.67057770e-01
1.13884427e-01 -2.35596716e-01 -1.00904667e+00 -3.08829129e-01
3.03895473e-01 2.50866324e-01 2.39753544e-01 3.57110322e-01
1.03403485e+00 -3.78071636e-01 -8.81079882e-02 6.96366668e-01
1.23750103e+00 3.55587631e-01 9.32773829e-01 1.43786326e-01
5.56382954e-01 9.52887356e-01 1.53265655e-01 2.95424998e-01
-7.21994191e-02 7.41963685e-01 3.61549884e-01 -1.24132589e-01
-3.11331600e-01 -6.33921772e-02 2.15048388e-01 2.02878773e-01
-1.96701348e-01 2.07259089e-01 -8.79591465e-01 3.83929729e-01
-1.13906217e+00 -9.26269412e-01 -1.30533725e-01 2.53127480e+00
5.34826219e-01 -2.02405840e-01 3.35957080e-01 1.47045776e-01
1.21922302e+00 -5.74890822e-02 -4.18315142e-01 -4.74295169e-01
1.49270454e-02 7.50200331e-01 6.01962090e-01 3.82893741e-01
-8.05588365e-01 3.23309273e-01 5.60226536e+00 7.22163498e-01
-1.47793627e+00 5.33513367e-01 7.49313593e-01 -2.30839908e-01
-3.03855896e-01 -2.72165388e-01 -3.36498290e-01 6.10801339e-01
7.94879556e-01 -1.53524324e-01 4.14742082e-01 4.83377069e-01
1.67657897e-01 -1.51563376e-01 -8.55722427e-01 1.15975308e+00
4.06095088e-01 -9.58157718e-01 1.61674216e-01 2.90634066e-01
2.92457819e-01 -3.11035633e-01 8.41590837e-02 -3.48675609e-01
-4.36996818e-01 -1.28877497e+00 9.76810694e-01 6.41787350e-01
1.08524060e+00 -8.91620994e-01 8.18592429e-01 -5.01255728e-02
-7.41614759e-01 3.06226254e-01 -1.77960526e-02 5.15822113e-01
3.24996531e-01 5.45625031e-01 -7.44128048e-01 6.21716976e-01
6.34273648e-01 -2.43097439e-01 -5.58895826e-01 1.00146437e+00
4.86400239e-02 2.13053241e-01 -3.24406683e-01 3.73576790e-01
-1.32757127e-01 -4.67492282e-01 8.73546124e-01 9.34602022e-01
3.93822193e-01 1.21383019e-01 -4.26576316e-01 1.14068961e+00
1.52437761e-01 2.16893330e-01 -2.74507642e-01 1.64555877e-01
1.86404079e-01 1.15425789e+00 -1.06296456e+00 2.40135342e-02
-1.31817013e-01 9.76013422e-01 -9.53397751e-02 -2.57629137e-02
-8.05535078e-01 -2.26520926e-01 5.26746035e-01 8.27745080e-01
4.71842103e-02 3.97751778e-02 -3.25623631e-01 -7.59332478e-01
3.86616401e-02 -8.62774968e-01 3.43172818e-01 -5.63048124e-01
-1.12953746e+00 8.67676616e-01 3.19935203e-01 -7.52443433e-01
-5.21659292e-02 -2.69313902e-01 -3.92000407e-01 1.20271397e+00
-8.92236054e-01 -9.90674257e-01 -2.91761816e-01 5.92422128e-01
-1.94873363e-02 -5.49326986e-02 7.90014744e-01 4.95708287e-01
-4.66201514e-01 8.16188455e-01 -1.58020541e-01 6.74614161e-02
5.16009152e-01 -8.56499732e-01 1.61354870e-01 8.47624004e-01
-3.73916770e-03 5.87216139e-01 6.39403999e-01 -8.32171738e-01
-7.71471798e-01 -9.86526549e-01 4.32760656e-01 -1.59239769e-01
1.38197422e-01 -2.29165271e-01 -1.00426280e+00 4.10644233e-01
-2.08947808e-01 -1.16019592e-01 5.92178881e-01 -3.48116338e-01
-2.48084724e-01 -1.49446070e-01 -1.97324300e+00 4.17429060e-01
7.22557008e-01 -3.80864620e-01 -5.83067536e-01 -1.58924550e-01
7.46598020e-02 -3.07486504e-01 -1.04047465e+00 4.32571411e-01
7.00187147e-01 -1.13183630e+00 9.66348350e-01 -3.51243019e-02
-7.32771680e-02 -1.56689048e-01 3.13324690e-01 -1.12933302e+00
-1.87367797e-01 -6.59581602e-01 5.57719707e-01 1.39924490e+00
2.20725566e-01 -8.91534865e-01 8.21355402e-01 1.15760851e+00
1.84537306e-01 -4.76729453e-01 -1.33723688e+00 -7.40435183e-01
7.42492154e-02 -9.92226526e-02 7.61384308e-01 8.32323849e-01
-3.96716297e-01 -1.96552172e-01 -2.31031671e-01 2.17844874e-01
8.50995183e-01 -1.91366076e-01 4.96181995e-01 -1.31296539e+00
-6.39795419e-03 -3.35168570e-01 -8.82728815e-01 4.04208452e-02
6.00312389e-02 -1.03983486e+00 -3.42402339e-01 -9.52383935e-01
2.08264902e-01 -4.62685883e-01 4.91098501e-02 4.05651778e-01
6.14770837e-02 4.57406968e-01 2.50938088e-01 1.02887407e-01
4.41962242e-01 2.55959541e-01 1.08263326e+00 1.39565662e-01
2.96604517e-03 -9.73962173e-02 -5.11098564e-01 6.12846136e-01
6.99096680e-01 -6.75863385e-01 -3.49083602e-01 -3.18531364e-01
-3.17015916e-01 1.84194883e-03 6.05282545e-01 -1.18024194e+00
-7.61933252e-02 4.01722580e-01 5.04598022e-01 -2.74538904e-01
2.28465751e-01 -9.23153996e-01 8.57991159e-01 6.05580032e-01
7.13536069e-02 2.85499245e-01 3.37530166e-01 2.35981718e-01
5.15637323e-02 -6.41666770e-01 1.24505234e+00 -1.42125681e-01
-9.51079503e-02 3.45752954e-01 -3.01986516e-01 -6.54527172e-02
1.10153627e+00 -5.45165300e-01 1.70525655e-01 -2.54496783e-01
-9.42937016e-01 -3.71164560e-01 6.56456292e-01 2.54438937e-01
4.61222738e-01 -8.50567460e-01 -6.36607766e-01 3.46261322e-01
-4.10069048e-01 -1.38067171e-01 5.51769376e-01 9.89093006e-01
-7.94336081e-01 1.80542633e-01 -5.79405844e-01 -3.51423204e-01
-1.53959370e+00 4.71077472e-01 5.45777202e-01 -1.07571393e-01
-7.45481193e-01 6.99472189e-01 1.32544696e-01 -2.14954391e-01
-2.68108025e-02 -5.84746115e-02 -1.30342603e-01 2.08303943e-01
4.69222426e-01 5.28204203e-01 4.72810537e-01 -1.02558041e+00
-4.90803510e-01 5.44633806e-01 -1.03248388e-01 -3.30271333e-01
1.20666957e+00 -9.99884382e-02 -2.89050817e-01 -1.81548491e-01
1.17086446e+00 2.23052099e-01 -9.32399929e-01 2.15885222e-01
-1.53604224e-01 -5.71877837e-01 2.80028343e-01 -7.67850101e-01
-1.39721441e+00 6.26826286e-01 1.13986242e+00 -1.26670133e-02
1.10322428e+00 -2.07747638e-01 7.13291943e-01 -5.54942369e-01
5.14937878e-01 -7.40212262e-01 -4.47993517e-01 -2.51327842e-01
1.00663471e+00 -6.76347673e-01 7.30866045e-02 -6.42725468e-01
-3.98876488e-01 9.28006589e-01 3.19938898e-01 4.47863787e-02
6.70893133e-01 2.85351783e-01 1.31879166e-01 -4.18099970e-01
3.45230728e-01 2.21250564e-01 3.58808726e-01 8.42016160e-01
2.27653235e-01 9.96859521e-02 -5.97284913e-01 6.54438376e-01
-4.99031991e-01 -1.36347190e-01 3.14260036e-01 5.63262284e-01
1.52183428e-01 -1.17835963e+00 -9.34818387e-01 5.50778151e-01
-8.08977902e-01 1.08597815e-01 -3.79241198e-01 6.95576906e-01
4.40339476e-01 6.91676557e-01 1.32327721e-01 -1.34618329e-02
3.76757145e-01 1.89238966e-01 7.49136567e-01 -3.66086036e-01
-7.04108953e-01 -1.93653703e-01 -8.50582868e-02 -4.53309923e-01
-3.41217279e-01 -9.57637370e-01 -9.51743007e-01 -4.25667524e-01
-2.51990110e-01 1.28963754e-01 8.92188311e-01 8.07658553e-01
4.64694023e-01 2.94582814e-01 3.26520979e-01 -8.21742654e-01
-5.07747948e-01 -9.55890954e-01 -7.89720714e-01 7.32482255e-01
-9.17512327e-02 -8.76993954e-01 -2.48106852e-01 1.34188738e-02] | [14.136228561401367, -1.9231795072555542] |
8946d903-a858-4aba-86fd-1fab5fe78054 | early-prediction-of-the-risk-of-icu-mortality | 2212.00554 | null | https://arxiv.org/abs/2212.00554v2 | https://arxiv.org/pdf/2212.00554v2.pdf | Early prediction of the risk of ICU mortality with Deep Federated Learning | Intensive Care Units usually carry patients with a serious risk of mortality. Recent research has shown the ability of Machine Learning to indicate the patients' mortality risk and point physicians toward individuals with a heightened need for care. Nevertheless, healthcare data is often subject to privacy regulations and can therefore not be easily shared in order to build Centralized Machine Learning models that use the combined data of multiple hospitals. Federated Learning is a Machine Learning framework designed for data privacy that can be used to circumvent this problem. In this study, we evaluate the ability of deep Federated Learning to predict the risk of Intensive Care Unit mortality at an early stage. We compare the predictive performance of Federated, Centralized, and Local Machine Learning in terms of AUPRC, F1-score, and AUROC. Our results show that Federated Learning performs equally well as the centralized approach and is substantially better than the local approach, thus providing a viable solution for early Intensive Care Unit mortality prediction. In addition, we show that the prediction performance is higher when the patient history window is closer to discharge or death. Finally, we show that using the F1-score as an early stopping metric can stabilize and increase the performance of our approach for the task at hand. | ['Korbinian Randl', 'Ioanna Miliou', 'Lena Mondrejevski', 'Núria Lladós Armengol'] | 2022-12-01 | null | null | null | null | ['mortality-prediction', 'icu-mortality'] | ['medical', 'medical'] | [-2.48840258e-01 1.95210353e-01 -3.84315550e-01 -5.27029037e-01
-9.29656327e-01 -3.67476434e-01 9.17383805e-02 8.05321574e-01
-6.99650526e-01 7.56219029e-01 4.89838004e-01 -5.29160678e-01
-5.10066330e-01 -6.00568295e-01 -1.86109155e-01 -6.54641032e-01
-3.56854558e-01 5.06422281e-01 -4.46696579e-01 3.01868230e-01
-3.28608453e-01 5.70868611e-01 -9.11200106e-01 6.04360342e-01
6.37854576e-01 1.01645589e+00 -5.95768034e-01 5.47925770e-01
2.49247074e-01 1.08146250e+00 -4.51336324e-01 -4.73122925e-01
6.20341897e-01 -5.20821035e-01 -7.39982128e-01 -6.20610774e-01
7.20633566e-02 -5.94476461e-01 -1.79255351e-01 6.56415284e-01
7.19198525e-01 -1.26386687e-01 3.57585996e-01 -1.04397070e+00
8.70124549e-02 6.26262486e-01 9.29526612e-02 -4.89704534e-02
1.71244279e-01 1.41635686e-01 8.52142692e-01 -4.22198355e-01
4.46068674e-01 6.82308137e-01 9.23795640e-01 6.69933021e-01
-1.24045384e+00 -3.86326730e-01 -1.50489062e-01 -2.53662854e-01
-1.20993710e+00 -3.70992631e-01 2.95295894e-01 -4.71558481e-01
5.54001749e-01 6.72962725e-01 4.03523743e-01 8.36200476e-01
3.27336669e-01 3.20101112e-01 8.56008947e-01 -2.58426905e-01
4.19809103e-01 1.88789666e-01 3.11782718e-01 7.50914991e-01
5.00701249e-01 2.12731272e-01 -3.72744560e-01 -9.47933555e-01
2.59750158e-01 7.47327030e-01 -5.11407554e-01 -6.29446268e-01
-1.13093042e+00 8.56530070e-01 3.37246001e-01 1.80135459e-01
-4.75330323e-01 -1.47525162e-01 6.46046162e-01 3.93967718e-01
4.11411375e-01 5.62016904e-01 -9.28113997e-01 -1.57057717e-01
-9.11381662e-01 -5.10452911e-02 1.10892689e+00 3.81646603e-01
4.19958532e-01 -6.16065919e-01 -4.12162066e-01 4.80119407e-01
-2.05818504e-01 -3.90679166e-02 4.66812491e-01 -1.22497761e+00
3.52934599e-01 8.07685018e-01 2.51459867e-01 -7.12888479e-01
-5.88196397e-01 -4.59221900e-01 -9.57570493e-01 3.00624579e-01
5.41999280e-01 -7.32348204e-01 -2.66985029e-01 1.70807314e+00
1.96416974e-01 -3.10268551e-01 4.48469073e-01 6.35980844e-01
3.13255817e-01 1.48052558e-01 2.02909529e-01 -4.30210143e-01
1.12699652e+00 -5.94921052e-01 -6.36552453e-01 6.38876483e-02
1.22660339e+00 -2.58197878e-02 4.53481555e-01 3.81796867e-01
-8.62523615e-01 2.87012130e-01 -4.58100826e-01 1.93540469e-01
-2.99781561e-01 -7.64725134e-02 5.24851799e-01 7.41367519e-01
-8.16872001e-01 9.41517889e-01 -1.05651462e+00 -4.32101607e-01
7.76939511e-01 5.09935141e-01 -6.13699913e-01 -1.24288514e-01
-8.61319244e-01 8.36796939e-01 2.32233211e-01 -3.63501936e-01
-5.59724033e-01 -9.23687577e-01 -6.40681565e-01 3.47425520e-01
2.02555224e-01 -1.11855364e+00 1.08929777e+00 -7.54974067e-01
-7.00012088e-01 8.37967813e-01 1.81605786e-01 -9.52335298e-01
9.52719986e-01 -3.59942228e-01 -1.06566623e-01 9.83841568e-02
-2.38706693e-01 1.04156956e-01 2.61527717e-01 -7.61489213e-01
-8.43598664e-01 -6.63970113e-01 -3.21685791e-01 2.99039204e-02
-5.40880322e-01 5.73638082e-02 3.22367311e-01 -3.73633415e-01
-3.79664123e-01 -5.80863774e-01 -6.05414689e-01 1.73972487e-01
-2.51875026e-03 1.18650682e-01 6.63771927e-01 -6.68796957e-01
1.48127007e+00 -2.31366110e+00 -2.82399386e-01 2.16928065e-01
5.57395995e-01 2.24572659e-01 2.17829034e-01 5.92871666e-01
-1.41371533e-01 4.68521178e-01 -3.07112157e-01 -2.20347151e-01
-2.92823553e-01 1.79311931e-01 8.17878768e-02 6.37314439e-01
-1.68445900e-01 8.36768925e-01 -6.66486144e-01 -4.75265801e-01
6.97084442e-02 3.44052970e-01 -6.78031087e-01 5.25667608e-01
2.18074962e-01 5.60636699e-01 -4.48956102e-01 3.81830394e-01
1.29496887e-01 -5.04289150e-01 4.89099175e-01 3.08560789e-01
5.73601276e-02 2.83146109e-02 -5.66324890e-01 1.30547667e+00
-2.15010077e-01 1.24582618e-01 2.78872609e-01 -7.42054105e-01
9.11419868e-01 7.32621968e-01 1.08610177e+00 -1.95308283e-01
2.64143050e-01 9.51066837e-02 -1.45432577e-01 -4.88533229e-01
-2.86617488e-01 -1.55228719e-01 -5.07086031e-02 5.95909655e-01
-3.55580986e-01 6.04555130e-01 -6.12658620e-01 -1.30536715e-02
1.64355266e+00 -4.57914472e-01 7.40568578e-01 -2.44834572e-01
3.13114345e-01 -1.30245268e-01 8.28601062e-01 7.98919439e-01
-4.28228170e-01 6.51782036e-01 6.31494462e-01 -9.70095336e-01
-7.23646462e-01 -7.78669477e-01 -2.47238159e-01 7.83611894e-01
-4.63824481e-01 -3.17956060e-01 -5.85698664e-01 -1.10580492e+00
3.18883210e-01 7.31319308e-01 -6.32136762e-01 -4.56008941e-01
-4.36803371e-01 -7.11954117e-01 4.97745961e-01 4.72808152e-01
1.51339516e-01 -6.71550095e-01 -1.18505132e+00 2.21754402e-01
-1.99661657e-01 -5.83641112e-01 -3.58061194e-01 4.07388628e-01
-1.08737206e+00 -1.39009237e+00 -7.68209636e-01 -4.00925666e-01
4.80966479e-01 -2.76235759e-01 9.96590972e-01 1.44535750e-01
-2.16494858e-01 4.43835169e-01 -2.10846558e-01 -4.64988738e-01
-5.25374293e-01 1.24404952e-01 -4.32941839e-02 2.14549467e-01
6.32034183e-01 -2.66325533e-01 -9.80927110e-01 1.07145146e-01
-7.35061288e-01 -2.97759563e-01 1.61926955e-01 9.13776457e-01
1.80671111e-01 -2.83887237e-01 7.43787885e-01 -1.18423116e+00
6.68276787e-01 -7.77022183e-01 -3.17674458e-01 2.75759250e-01
-1.24994910e+00 -6.28410429e-02 8.05800498e-01 5.27707227e-02
-6.60873532e-01 3.58099103e-01 1.38989612e-01 -5.38203776e-01
-4.24238920e-01 3.88979763e-01 7.32913986e-02 2.78845072e-01
7.27479696e-01 -3.23920757e-01 4.39351827e-01 -8.30754101e-01
-1.03601038e-01 8.65231633e-01 3.17269742e-01 -1.63280994e-01
8.63835290e-02 3.67825627e-01 9.61793885e-02 -1.60959452e-01
-7.30217278e-01 -5.74957013e-01 -4.26069319e-01 1.90160930e-01
8.63555253e-01 -9.77790833e-01 -8.93319964e-01 -1.38914762e-02
-8.20596814e-01 -2.13815421e-01 -6.00329757e-01 6.38837218e-01
-4.93424952e-01 2.61904836e-01 -4.76548046e-01 -9.20576155e-01
-8.46680701e-01 -6.35876417e-01 5.85784018e-01 -2.50063241e-01
-5.84243953e-01 -1.14097524e+00 3.40536475e-01 2.34957382e-01
4.51763690e-01 8.92621100e-01 1.13582039e+00 -1.28988528e+00
-1.61653787e-01 -6.69540048e-01 1.97523326e-01 2.78712153e-01
4.35976177e-01 -3.54818612e-01 -9.24120665e-01 -6.86917067e-01
1.51246414e-01 -9.77028310e-02 6.29516900e-01 2.73475051e-01
1.22403467e+00 -8.31583619e-01 -4.08090442e-01 7.35789478e-01
1.39720905e+00 5.54630570e-02 3.37289840e-01 2.60738999e-01
2.58650005e-01 7.19252527e-01 3.60804707e-01 1.03732836e+00
2.76348829e-01 2.15431735e-01 4.76015478e-01 -2.19678834e-01
4.72777009e-01 -6.65370226e-02 4.18429971e-02 3.20088178e-01
1.72704950e-01 -1.20975032e-01 -1.35770643e+00 5.56759357e-01
-2.10973382e+00 -8.32442760e-01 5.50038554e-02 2.64626241e+00
7.14951634e-01 -4.20274079e-01 2.64453918e-01 -1.42055765e-01
5.31922996e-01 -2.95271307e-01 -5.96385837e-01 -7.08177745e-01
7.08755851e-02 -3.28492344e-04 7.23026633e-01 8.77249837e-02
-9.77191687e-01 1.54164329e-01 6.83190536e+00 -1.43415362e-01
-9.61526275e-01 8.82198960e-02 1.13868248e+00 -4.35953408e-01
1.04489468e-01 -1.51869491e-01 -9.60356593e-02 4.72688019e-01
1.28214395e+00 -3.45548928e-01 2.97137171e-01 1.05587268e+00
2.12150693e-01 2.77513653e-01 -1.57914948e+00 1.03457057e+00
-3.24648142e-01 -1.28587377e+00 -2.29798481e-01 2.74923414e-01
5.61175466e-01 1.82704061e-01 -9.21396837e-02 -6.84017614e-02
4.66585964e-01 -1.11417544e+00 1.58600688e-01 8.29750359e-01
9.86595571e-01 -7.62233198e-01 1.06686354e+00 6.66670799e-01
-5.82007945e-01 -6.79243267e-01 4.91340496e-02 -9.10849050e-02
-1.52774259e-01 4.46813911e-01 -9.91491199e-01 4.54432994e-01
8.50777149e-01 4.37153935e-01 -3.88766587e-01 1.13817465e+00
5.62652588e-01 5.33094287e-01 -2.66661555e-01 3.34479630e-01
4.40722555e-02 6.80470914e-02 2.54257679e-01 9.48177159e-01
3.77527595e-01 2.29495615e-01 1.24205694e-01 4.39929128e-01
-1.69992477e-01 4.21916783e-01 -1.09237146e+00 8.56672004e-02
1.76286444e-01 8.58232796e-01 2.46189199e-02 -1.32329956e-01
-4.81225312e-01 6.76776409e-01 3.23750317e-01 1.20999165e-01
-2.73131579e-01 -1.84390366e-01 9.83212471e-01 3.41172993e-01
1.40583262e-01 2.87659287e-01 -5.22448778e-01 -1.19756711e+00
-2.24389896e-01 -1.07295907e+00 1.17795694e+00 -1.77482978e-01
-1.33042431e+00 5.55864751e-01 -4.45950836e-01 -1.18207610e+00
-5.59324324e-01 -3.72268885e-01 -5.17814755e-01 7.89575994e-01
-1.16451967e+00 -5.94468832e-01 -1.93550438e-01 8.06220055e-01
-1.84327915e-01 -2.72709399e-01 1.37885809e+00 8.94738957e-02
-6.96305513e-01 7.37719059e-01 4.34536457e-01 3.88229132e-01
8.43224049e-01 -9.59707916e-01 -2.42095903e-01 6.76866412e-01
-3.90812993e-01 6.16543174e-01 3.05073202e-01 -6.09164119e-01
-1.21986294e+00 -1.36372066e+00 1.04671514e+00 -7.53134668e-01
2.23439083e-01 -3.39597836e-02 -8.22032869e-01 7.64944911e-01
-5.09835407e-02 2.15157151e-01 1.24292123e+00 1.96836293e-01
-1.84125915e-01 -3.88357460e-01 -1.66330826e+00 3.69664967e-01
5.09396970e-01 -4.08223361e-01 -4.20048654e-01 3.68698567e-01
6.08329773e-01 6.63206354e-02 -1.36030090e+00 4.90786165e-01
5.21368325e-01 -1.17746437e+00 6.22862756e-01 -1.16883838e+00
2.02200532e-01 3.40388745e-01 -1.79848373e-01 -9.24056888e-01
-5.66174507e-01 -1.00593233e+00 -2.53986329e-01 9.75031376e-01
4.98110473e-01 -1.06078923e+00 9.66789186e-01 1.42554486e+00
4.05576497e-01 -9.71658945e-01 -1.13783014e+00 -4.80136454e-01
2.81142116e-01 -1.54651469e-03 8.96069407e-01 1.07467437e+00
2.55255848e-01 -1.37336895e-01 -4.34740961e-01 1.83418185e-01
6.47067308e-01 1.61250949e-01 5.30271053e-01 -1.35463834e+00
-1.49058625e-01 -2.31394693e-01 -3.51315886e-01 -1.28662050e-01
-1.61123574e-01 -9.58431602e-01 -4.47321296e-01 -1.40373540e+00
3.05152744e-01 -5.25387764e-01 -9.31944907e-01 9.49892879e-01
-6.85096979e-02 -4.02283400e-01 8.39188248e-02 3.38004589e-01
-3.48273307e-01 1.87303707e-01 6.64507091e-01 1.86161444e-01
-3.00724566e-01 2.85825431e-01 -8.60872269e-01 4.89021599e-01
1.10053885e+00 -7.74040937e-01 -8.06808621e-02 -2.93934405e-01
-1.26032993e-01 5.42653799e-01 4.19090152e-01 -7.93850183e-01
1.93243533e-01 -1.48874640e-01 4.08027828e-01 -1.54278561e-01
-2.29145866e-02 -1.36158049e+00 2.96002656e-01 9.31121349e-01
-5.75132966e-01 1.40124068e-01 -1.43307954e-01 5.97257018e-01
-2.48198342e-02 1.11639328e-01 7.52342224e-01 -2.82824039e-01
1.10035516e-01 5.47973633e-01 -2.93245286e-01 9.07028615e-02
1.22886527e+00 3.21397372e-02 -1.96735099e-01 -3.41393381e-01
-8.22930455e-01 4.31674033e-01 7.45797515e-01 1.34131119e-01
5.51099300e-01 -9.14083958e-01 -8.34375799e-01 3.23813826e-01
2.79063255e-01 -1.94249809e-01 -3.81531417e-02 9.86254930e-01
-5.51060379e-01 4.51191634e-01 7.23055750e-02 -2.75896221e-01
-1.30698431e+00 1.02537894e+00 6.10612094e-01 -2.32175514e-01
-9.38052773e-01 3.80789399e-01 1.05000988e-01 -2.80822992e-01
5.96499562e-01 -2.65956014e-01 1.75991178e-01 -1.44755885e-01
6.13606691e-01 5.38418651e-01 1.94795609e-01 -2.49050945e-01
-5.96873343e-01 -1.22465633e-01 -2.27403454e-02 3.35045278e-01
1.47485602e+00 -1.72080714e-02 -3.14377904e-01 2.70588726e-01
1.47505677e+00 -1.33451164e-01 -1.06918752e+00 -1.90203506e-02
1.24967203e-01 -5.16449451e-01 -6.32094592e-02 -1.08786142e+00
-1.05038822e+00 6.32165194e-01 9.31411266e-01 1.44334093e-01
1.19713044e+00 -1.51286110e-01 6.59547329e-01 3.86733949e-01
3.42272937e-01 -5.74738026e-01 -6.65810347e-01 6.81686252e-02
4.08827007e-01 -1.26405263e+00 -1.38138041e-01 1.75979093e-01
-7.00903356e-01 9.55801845e-01 1.67624369e-01 3.18335801e-01
8.42007458e-01 2.98609525e-01 4.72397953e-01 -2.48789750e-02
-1.08018064e+00 4.62097168e-01 -8.19759965e-02 4.24759358e-01
2.24464968e-01 3.08395177e-01 -2.69149870e-01 7.53365099e-01
1.00790165e-01 2.58842200e-01 2.89276063e-01 1.04482162e+00
-2.83310115e-01 -1.14701128e+00 -1.97389215e-01 9.36192095e-01
-1.06259644e+00 -1.67019684e-02 -6.43815100e-01 1.76456630e-01
1.27852308e-02 8.74304771e-01 -1.16670482e-01 -1.64708272e-01
2.83053368e-01 6.25841260e-01 -1.76115081e-01 -5.45393884e-01
-1.07507694e+00 -3.55012268e-01 3.71694118e-02 -7.97323287e-01
2.09895503e-02 -6.17270350e-01 -1.03209794e+00 -2.27612510e-01
6.76241592e-02 4.76776212e-01 4.38377470e-01 5.17938912e-01
7.94975340e-01 1.68375120e-01 8.96727383e-01 1.10412925e-01
-1.02142382e+00 -3.71472389e-01 -5.80527067e-01 5.36756575e-01
8.69618714e-01 1.34741500e-01 -4.40622896e-01 -1.78040579e-01] | [6.199338912963867, 6.511131763458252] |
f1d2aee9-006e-4a59-bbb7-796250fbf85f | an-analytics-of-culture-modeling-subjectivity | 2211.07460 | null | https://arxiv.org/abs/2211.07460v1 | https://arxiv.org/pdf/2211.07460v1.pdf | An Analytics of Culture: Modeling Subjectivity, Scalability, Contextuality, and Temporality | There is a bidirectional relationship between culture and AI; AI models are increasingly used to analyse culture, thereby shaping our understanding of culture. On the other hand, the models are trained on collections of cultural artifacts thereby implicitly, and not always correctly, encoding expressions of culture. This creates a tension that both limits the use of AI for analysing culture and leads to problems in AI with respect to cultural complex issues such as bias. One approach to overcome this tension is to more extensively take into account the intricacies and complexities of culture. We structure our discussion using four concepts that guide humanistic inquiry into culture: subjectivity, scalability, contextuality, and temporality. We focus on these concepts because they have not yet been sufficiently represented in AI research. We believe that possible implementations of these aspects into AI research leads to AI that better captures the complexities of culture. In what follows, we briefly describe these four concepts and their absence in AI research. For each concept, we define possible research challenges. | ['Marcel Worring', 'Julia Noordegraaf', 'Tobias Blanke', 'Melvin Wevers', 'Nanne van Noord'] | 2022-11-14 | null | null | null | null | ['culture'] | ['speech'] | [ 2.46373773e-01 -3.87308970e-02 -1.53495744e-01 -4.40536886e-02
1.40111148e-01 -8.04002583e-01 9.02645111e-01 2.80449092e-01
-3.93357456e-01 5.51088750e-01 8.56165946e-01 -1.84384227e-01
8.15896224e-03 -5.63856661e-01 -2.99582362e-01 -4.53788340e-01
3.10385853e-01 1.66351661e-01 -1.38594136e-01 -4.51525092e-01
6.28711939e-01 -6.86588138e-02 -1.60658348e+00 1.93110451e-01
8.89123082e-01 2.89336383e-01 8.67628753e-02 4.50669914e-01
-2.68002778e-01 1.28199792e+00 -7.63181448e-01 -4.48361218e-01
-3.29583466e-01 -9.55155730e-01 -7.79211164e-01 3.44426095e-01
-7.62695149e-02 7.82892574e-03 4.50004160e-01 9.58943009e-01
1.34953156e-01 -4.07865614e-01 6.76004052e-01 -8.72446060e-01
-1.17487621e+00 7.24425793e-01 -1.63868457e-01 1.04297146e-01
5.82300842e-01 1.17113300e-01 6.42237544e-01 -4.11420792e-01
9.30033684e-01 1.13674724e+00 7.33641267e-01 3.19707960e-01
-1.01789343e+00 -6.12154782e-01 2.60788769e-01 7.57824332e-02
-1.37341177e+00 -6.04296148e-01 7.91729510e-01 -9.01451051e-01
8.29634488e-01 2.59568185e-01 1.61117566e+00 1.29853690e+00
1.20615184e-01 5.84261060e-01 1.41780496e+00 -8.71611893e-01
2.14794159e-01 7.00928867e-01 -3.04672569e-02 1.98636129e-01
5.51648200e-01 -7.60562569e-02 -5.38013697e-01 -6.02118261e-02
8.35866809e-01 -3.72951597e-01 -2.85509169e-01 -2.07301810e-01
-1.36858833e+00 7.35571563e-01 -9.73919332e-02 1.08248138e+00
-1.84574291e-01 6.72088563e-02 3.57451051e-01 3.20524901e-01
2.70164192e-01 8.57838333e-01 1.91983640e-01 -8.63362730e-01
-6.33059382e-01 2.34956294e-01 9.46840346e-01 8.32068622e-01
3.70329946e-01 9.91088226e-02 5.78963459e-01 8.39994371e-01
4.56659734e-01 4.60971028e-01 2.64300346e-01 -1.09879720e+00
-3.15586150e-01 7.47468710e-01 2.31533214e-01 -1.36117995e+00
-1.76659003e-01 -4.19466019e-01 -3.97171974e-01 -1.10901646e-01
3.67120773e-01 8.47736672e-02 -5.93588054e-01 1.69981575e+00
5.76103181e-02 -8.02825615e-02 7.40127712e-02 1.04429746e+00
3.91477376e-01 3.54744226e-01 3.91748846e-01 -2.63144851e-01
1.27993608e+00 -5.46773612e-01 -8.94080400e-01 -6.23341382e-01
8.43076646e-01 -8.20159972e-01 1.23805332e+00 4.76090044e-01
-1.24147713e+00 9.04110223e-02 -1.32070041e+00 -3.56790833e-02
-6.02341771e-01 -4.71202195e-01 9.83474791e-01 1.09480989e+00
-1.05502748e+00 8.56804550e-02 -6.79636538e-01 -1.04269397e+00
-2.13524383e-02 -1.07790306e-01 1.88280672e-01 -2.00506561e-02
-1.06885433e+00 1.19060445e+00 3.24006408e-01 3.43675792e-01
-1.44015864e-01 -3.30024987e-01 -6.97981596e-01 -5.62929690e-01
3.31421286e-01 -6.36801839e-01 9.98079598e-01 -1.96809673e+00
-1.26114225e+00 1.10820198e+00 -1.50463479e-02 6.78590164e-02
3.31093252e-01 -1.39385119e-01 -8.25876176e-01 -2.10207358e-01
-6.59302846e-02 1.62722796e-01 1.63665906e-01 -1.78133142e+00
-5.14319599e-01 -2.28992030e-01 -2.57319175e-02 1.58140555e-01
-5.43844581e-01 5.32455981e-01 -1.70757234e-01 -4.81354833e-01
-1.01317100e-01 -9.93819177e-01 5.17497957e-02 -1.87298670e-01
3.27668577e-01 2.95159101e-01 3.54967445e-01 -4.11182106e-01
1.71304417e+00 -2.15416622e+00 2.60044366e-01 1.87278748e-01
3.44051898e-01 1.03260905e-01 4.51050177e-02 9.18457627e-01
5.58940887e-01 4.26277429e-01 -2.29707897e-01 1.10790484e-01
1.37159809e-01 4.05638814e-01 -1.01116344e-01 3.72337788e-01
1.39290512e-01 7.80050576e-01 -1.20439303e+00 -4.48802203e-01
-7.27816373e-02 8.59857798e-01 -5.30113280e-01 -4.72238481e-01
-3.83275971e-02 2.68177658e-01 -2.91061610e-01 6.87758088e-01
2.01916903e-01 -8.43867958e-02 8.83444428e-01 5.73874116e-01
-5.97319663e-01 3.02771181e-01 -9.75903809e-01 1.41384017e+00
-1.21065520e-01 8.70139480e-01 6.96697012e-02 -4.95890290e-01
9.48407054e-01 4.65996504e-01 2.27092966e-01 -8.43007386e-01
2.01612413e-01 5.62759757e-01 4.83228117e-01 -6.44062042e-01
6.91368580e-01 -6.24482274e-01 -1.46749765e-01 8.32368910e-01
-3.69908422e-01 -3.59435737e-01 1.08270749e-01 -2.21712291e-02
6.01898432e-01 4.49786931e-01 6.22996867e-01 -7.24626482e-01
3.17309886e-01 4.17309731e-01 7.89702773e-01 3.81209493e-01
-4.35455441e-01 2.64112473e-01 4.58902150e-01 -6.64611995e-01
-1.26714706e+00 -6.05365813e-01 -1.06485300e-01 7.44473279e-01
3.35908793e-02 -6.45348251e-01 -6.87119126e-01 -5.10905981e-02
-1.56218871e-01 1.11297071e+00 -7.18506753e-01 -1.35456860e-01
-6.22673273e-01 -6.73550308e-01 4.19540852e-01 3.86706680e-01
7.65653253e-02 -1.03159976e+00 -1.33448601e+00 3.02873194e-01
-5.87341525e-02 -7.76734173e-01 3.30760509e-01 -2.33897373e-01
-6.30056262e-01 -8.38827550e-01 -5.01382828e-01 -6.81684554e-01
4.18801099e-01 3.83608550e-01 1.33393013e+00 4.67704505e-01
1.23900130e-01 7.75262356e-01 -5.35958529e-01 -9.94181156e-01
-9.09103155e-01 -6.63281158e-02 -1.91066608e-01 -5.22664070e-01
1.22510302e+00 -6.30884647e-01 -2.78268158e-01 1.59454912e-01
-1.15237081e+00 2.98996121e-01 2.38393977e-01 3.86593103e-01
-8.64297077e-02 -1.43463597e-01 5.00057280e-01 -9.09337819e-01
7.94118106e-01 -6.90392792e-01 -2.60752868e-02 4.27293092e-01
-4.61366117e-01 -3.57625693e-01 2.09750801e-01 -5.25705397e-01
-8.82918477e-01 -5.03224194e-01 5.59303045e-01 1.50831893e-01
-1.41704187e-01 8.30303669e-01 3.28989089e-01 1.34685367e-01
6.51339710e-01 1.19483173e-01 2.56446749e-01 -1.34992478e-02
1.21122867e-01 8.07361603e-01 3.87821309e-02 -8.78529310e-01
3.48171979e-01 3.66564542e-01 -4.97247636e-01 -1.37357497e+00
-4.59230632e-01 -4.43631895e-02 -7.36526489e-01 -7.05778062e-01
7.79699683e-01 -8.46087277e-01 -3.51418406e-01 2.03988850e-01
-6.63353086e-01 -3.73869210e-01 -2.23712787e-01 7.18255937e-01
-6.14328265e-01 1.41449928e-01 -4.74465907e-01 -1.26208627e+00
1.37263909e-01 -1.00798047e+00 5.24400532e-01 1.15247093e-01
-1.14618123e+00 -1.31609046e+00 1.74496412e-01 4.45663214e-01
5.23190558e-01 6.54800653e-01 1.17156315e+00 -9.75810960e-02
-2.60174036e-01 8.28873962e-02 2.70422965e-01 3.25203054e-02
2.03851238e-01 3.75281036e-01 -8.82741153e-01 -2.78116390e-02
3.27949136e-01 -4.62052405e-01 5.88869229e-02 -8.62017944e-02
-2.05075629e-02 -2.32216224e-01 -2.82749720e-02 9.42785889e-02
1.69017124e+00 6.91867590e-01 6.94981694e-01 8.72453094e-01
4.34117705e-01 1.02507317e+00 4.34082001e-01 3.83738130e-01
7.86826849e-01 3.10850769e-01 -2.89375663e-01 1.99654609e-01
2.09817767e-01 -1.45878140e-02 2.65580475e-01 1.43984699e+00
-6.39070868e-01 -7.53428489e-02 -1.46719265e+00 7.70852447e-01
-1.83260548e+00 -9.06367362e-01 -3.05535793e-01 1.99424279e+00
8.17886412e-01 5.35366796e-02 3.01551789e-01 1.36098534e-01
3.20860893e-01 8.66342932e-02 -1.00597374e-01 -1.02572489e+00
-3.27898771e-01 -3.02120388e-01 -4.95516136e-02 5.12754440e-01
-3.97673547e-01 1.02565658e+00 7.40625238e+00 -2.61466175e-01
-1.08863699e+00 -1.46125495e-01 3.25550526e-01 -2.08804801e-01
-8.51419330e-01 1.63410529e-01 -1.09652296e-01 2.81250894e-01
7.21195042e-01 -2.02063888e-01 6.16691887e-01 2.76939988e-01
2.38066509e-01 -3.44730496e-01 -9.59840059e-01 5.02336502e-01
4.94251370e-01 -8.81317616e-01 -1.73843861e-01 2.39508554e-01
6.83712423e-01 -3.70741785e-01 1.81733202e-02 1.85580142e-02
4.36543733e-01 -1.01036847e+00 1.19052863e+00 3.71547103e-01
2.94009566e-01 -7.08418727e-01 4.50471789e-01 -6.55299891e-03
-4.79608715e-01 -1.43342450e-01 -2.48901650e-01 -8.42987537e-01
1.75789490e-01 1.01628691e-01 -3.38884890e-01 7.64728636e-02
5.39889872e-01 5.64332366e-01 -2.74544865e-01 6.64456725e-01
5.11833876e-02 6.16554737e-01 -2.40136445e-01 -3.59963447e-01
5.02363741e-01 -5.58926702e-01 3.69509310e-01 1.35357940e+00
2.98837841e-01 3.94344032e-01 -1.74215704e-01 9.67996597e-01
7.00611949e-01 2.22125947e-01 -9.53465104e-01 -7.77665555e-01
5.92434406e-01 5.54400265e-01 -8.70524406e-01 -1.48653507e-01
-8.08420956e-01 6.67025864e-01 1.66449174e-02 3.71360540e-01
-5.08309901e-01 2.84055740e-01 8.77153516e-01 1.86338022e-01
-1.35360155e-02 -7.44656742e-01 -5.76421142e-01 -9.48408425e-01
-1.46421671e-01 -1.44383729e+00 1.26568541e-01 -6.34033918e-01
-9.12791252e-01 1.60128072e-01 -1.14570111e-02 -5.48535109e-01
-3.81816238e-01 -4.12945569e-01 5.30134439e-02 6.46384597e-01
-9.50277984e-01 -1.44542253e+00 -2.56527394e-01 1.55525906e-02
1.24229848e-01 2.46919543e-01 1.10114455e+00 -4.39979024e-02
-3.34938556e-01 1.41727805e-01 5.60256466e-02 -1.44528389e-01
5.60424864e-01 -8.61890376e-01 2.64429033e-01 5.91561258e-01
-1.43282056e-01 9.21089351e-01 9.57643867e-01 -8.08712542e-01
-1.55076087e+00 -9.18394923e-02 1.18822098e+00 -8.58272314e-01
9.68440294e-01 -4.89326596e-01 -7.88203359e-01 8.74018788e-01
6.99094653e-01 -9.19234991e-01 1.31601274e+00 4.07671958e-01
-3.56507331e-01 4.20452505e-01 -9.26643610e-01 1.15747130e+00
1.12704849e+00 -4.86128360e-01 -6.81654990e-01 -1.00659624e-01
5.59477270e-01 8.37612972e-02 -1.25187457e+00 -9.29017067e-02
1.07861316e+00 -1.14242649e+00 5.51701665e-01 -2.64604598e-01
5.72344720e-01 -2.93952018e-01 -2.73925483e-01 -9.69944954e-01
-7.99969971e-01 -5.94285548e-01 3.97020131e-01 1.45301962e+00
3.48972917e-01 -9.17153895e-01 3.26152861e-01 9.89036620e-01
-7.60449097e-03 -4.70563591e-01 -3.51038992e-01 -2.82648712e-01
5.05098879e-01 -4.81016725e-01 8.70967746e-01 1.77883351e+00
6.57614768e-01 2.07351521e-01 -2.03436568e-01 -4.32236224e-01
3.16824108e-01 -2.03101262e-01 6.89314783e-01 -1.26061332e+00
4.62185666e-02 -7.49670446e-01 -7.10083485e-01 -7.56342635e-02
-3.93596679e-01 -3.07756841e-01 -2.46702388e-01 -1.58732975e+00
2.85040110e-01 -3.48724067e-01 5.77072091e-02 -1.59812555e-01
1.26666322e-01 2.96864450e-01 5.28887093e-01 3.36429566e-01
-1.67599916e-01 -1.79293584e-02 1.20781612e+00 3.19244087e-01
-4.26056594e-01 -1.01453972e+00 -1.34820497e+00 8.12134981e-01
9.80958521e-01 -1.61818475e-01 -6.38218880e-01 -7.51817226e-01
8.78451169e-01 -4.59050268e-01 4.62743118e-02 -1.05586004e+00
8.64949226e-02 -7.47101367e-01 3.11855853e-01 1.55796140e-01
2.99129874e-01 -9.60589826e-01 7.27407098e-01 2.69966662e-01
-1.90073654e-01 3.69492143e-01 2.87006915e-01 5.38294837e-02
-1.50165498e-01 -1.31964713e-01 4.67611939e-01 -6.25160038e-01
-6.95181906e-01 -8.49163055e-01 -8.61208200e-01 -3.36403935e-03
1.11772120e+00 -9.49229121e-01 -2.93193489e-01 -3.93498689e-01
-3.58232945e-01 9.68084931e-02 1.44632626e+00 6.40547097e-01
1.53836057e-01 -1.15504336e+00 -5.69270194e-01 1.51953071e-01
3.36348414e-01 -2.19078854e-01 -2.02492927e-03 7.25759923e-01
-7.86607563e-01 4.82326597e-01 -3.48677754e-01 -3.43223184e-01
-9.08573866e-01 6.05480969e-01 1.46077946e-01 5.68505287e-01
-6.40788496e-01 4.20892626e-01 2.32207894e-01 -6.03683814e-02
-1.00061037e-01 -2.09726930e-01 -2.62385577e-01 2.45925218e-01
4.65016305e-01 3.03352892e-01 -5.71215749e-01 -1.04085147e+00
-3.09223086e-01 6.53737605e-01 1.64127514e-01 -5.62399209e-01
1.27459383e+00 -3.65838528e-01 -5.42977273e-01 1.23410380e+00
6.42275810e-01 6.15006573e-02 -7.37898946e-01 3.27331722e-02
4.34053183e-01 -7.64233649e-01 -2.04296604e-01 -1.12493265e+00
-3.43443125e-01 5.85136294e-01 2.35090479e-01 4.96650577e-01
1.03666019e+00 -1.88464060e-01 3.16459537e-01 -1.34560347e-01
3.73469770e-01 -1.56235683e+00 -2.58742660e-01 3.40483040e-01
7.71408439e-01 -5.30683994e-01 1.41070679e-01 -3.61701012e-01
-1.05046821e+00 9.97382045e-01 5.23295105e-01 7.06270859e-02
4.46698576e-01 4.63440478e-01 5.83062470e-01 -2.05286086e-01
-8.62485528e-01 -1.42346755e-01 -2.08242536e-01 7.97803462e-01
1.24956632e+00 2.51641065e-01 -8.79388094e-01 8.15332234e-02
-5.27690589e-01 4.71074790e-01 8.16337824e-01 1.36744666e+00
-5.08800507e-01 -1.05040038e+00 -7.26019263e-01 -1.14045046e-01
-4.95448947e-01 1.91146776e-01 -1.23578501e+00 1.15167689e+00
2.77234674e-01 8.81944656e-01 2.49417171e-01 -4.32193398e-01
1.55121554e-02 2.65084177e-01 6.61726475e-01 -2.22355008e-01
-1.04860854e+00 2.50888318e-01 5.87463558e-01 -2.34025493e-01
-7.68114090e-01 -9.82090294e-01 -7.94943213e-01 -6.29440546e-01
3.74099649e-02 5.39207347e-02 8.22106063e-01 7.46295631e-01
2.37889230e-01 2.18120739e-01 -7.71480799e-02 -3.62810165e-01
5.43262698e-02 -7.46211469e-01 -4.43241239e-01 1.52533561e-01
-2.27664830e-04 -4.92325425e-01 -9.19636562e-02 8.60653892e-02] | [9.087401390075684, 6.320952892303467] |
f5f21622-7b74-4c08-ad65-43ee9db26759 | learning-a-target-sample-re-generator-for | 1707.08645 | null | http://arxiv.org/abs/1707.08645v1 | http://arxiv.org/pdf/1707.08645v1.pdf | Learning a Target Sample Re-Generator for Cross-Database Micro-Expression Recognition | In this paper, we investigate the cross-database micro-expression recognition
problem, where the training and testing samples are from two different
micro-expression databases. Under this setting, the training and testing
samples would have different feature distributions and hence the performance of
most existing micro-expression recognition methods may decrease greatly. To
solve this problem, we propose a simple yet effective method called Target
Sample Re-Generator (TSRG) in this paper. By using TSRG, we are able to
re-generate the samples from target micro-expression database and the
re-generated target samples would share same or similar feature distributions
with the original source samples. For this reason, we can then use the
classifier learned based on the labeled source samples to accurately predict
the micro-expression categories of the unlabeled target samples. To evaluate
the performance of the proposed TSRG method, extensive cross-database
micro-expression recognition experiments designed based on SMIC and CASME II
databases are conducted. Compared with recent state-of-the-art cross-database
emotion recognition methods, the proposed TSRG achieves more promising results. | ['Yuan Zong', 'Zhen Cui', 'Xiaohua Huang', 'Wenming Zheng', 'Guoying Zhao'] | 2017-07-26 | null | null | null | null | ['micro-expression-recognition'] | ['computer-vision'] | [ 1.83166847e-01 -3.83879930e-01 -2.00521842e-01 -6.63081288e-01
-9.16959047e-01 -3.32825273e-01 3.10727537e-01 3.22081894e-02
-2.28902414e-01 7.30805993e-01 -2.83712864e-01 3.66798729e-01
3.60890418e-01 -7.25917578e-01 -2.45896012e-01 -8.52085888e-01
4.27230269e-01 1.87379166e-01 -2.89474577e-01 -1.48148477e-01
4.06069793e-02 4.44473892e-01 -1.90747488e+00 4.99548614e-01
9.61568892e-01 1.37435234e+00 -2.62082368e-01 3.80375147e-01
-2.93224186e-01 8.17409158e-01 -1.05848157e+00 -5.51261365e-01
1.98291484e-02 -8.03011417e-01 -3.78829896e-01 3.11168730e-01
2.94060688e-02 2.00164199e-01 5.66777587e-02 1.25515652e+00
5.13233960e-01 2.14563102e-01 8.01898718e-01 -1.37835181e+00
-3.82308811e-01 4.19099294e-02 -7.62752712e-01 -1.12424351e-01
3.84415716e-01 -5.12735188e-01 6.53118253e-01 -9.99682605e-01
3.95336181e-01 1.13365448e+00 3.42000455e-01 6.55116618e-01
-1.14012575e+00 -9.93554652e-01 -1.17018096e-01 2.32295692e-01
-1.68843150e+00 -6.29393160e-01 9.19203222e-01 -1.79797024e-01
4.83867764e-01 3.88090163e-01 5.20921230e-01 1.10573590e+00
9.95173818e-04 1.09364367e+00 1.45046842e+00 -6.50707901e-01
3.57458144e-01 6.18537009e-01 1.64751578e-02 4.72816616e-01
-2.47734100e-01 -8.35967362e-02 -4.30756032e-01 -3.37377340e-01
2.20558137e-01 -6.24817088e-02 -3.92729253e-01 -2.60103971e-01
-7.36822486e-01 6.21023238e-01 -1.54595658e-01 6.05741262e-01
-3.10589015e-01 -5.42052746e-01 7.38388419e-01 4.63803500e-01
7.27125168e-01 1.18109800e-01 -3.32050472e-01 -2.31037199e-01
-6.74936116e-01 -4.12878133e-02 8.05701256e-01 7.34914780e-01
9.82328355e-01 1.41304925e-01 -9.09932330e-02 1.37857974e+00
1.16782226e-01 3.94210398e-01 9.44100320e-01 -2.98832536e-01
2.99443185e-01 7.24546611e-01 1.21353731e-01 -1.27659500e+00
2.95051262e-02 -3.25734884e-01 -9.76859868e-01 9.79066268e-03
2.91094571e-01 -3.24037910e-01 -6.77101195e-01 1.76507246e+00
3.92546952e-01 3.32404643e-01 7.56944835e-01 6.26865327e-01
5.65701663e-01 1.01567745e+00 2.32459735e-02 -4.60364699e-01
9.92519617e-01 -7.95514643e-01 -8.23419034e-01 -1.21894315e-01
8.00603151e-01 -6.59160614e-01 1.05959570e+00 5.43318808e-01
-5.23637414e-01 -6.32362843e-01 -1.23009193e+00 5.68585336e-01
-3.14139754e-01 8.21286261e-01 4.60757315e-01 7.68539965e-01
-3.08994532e-01 1.56428441e-01 -3.85063708e-01 -2.04778031e-01
2.62975246e-01 2.67057449e-01 -6.42542720e-01 -2.23240227e-01
-1.11307967e+00 6.80124938e-01 2.84153938e-01 1.95600808e-01
-4.97393340e-01 -3.28581244e-01 -8.05235445e-01 1.07768411e-02
2.08391801e-01 8.73463973e-02 1.17850220e+00 -1.81690919e+00
-1.79704642e+00 1.07418239e+00 -3.76014054e-01 1.37627870e-01
1.73190996e-01 1.13501899e-01 -9.21917558e-01 -1.18396915e-01
-2.09427625e-02 1.08148113e-01 7.28418827e-01 -1.14914119e+00
-6.13797724e-01 -6.64709270e-01 -5.22277832e-01 1.07939847e-01
-2.86341339e-01 1.03648894e-01 -4.47039336e-01 -5.72749496e-01
-1.21233016e-01 -8.39955986e-01 2.20820054e-01 -3.89644593e-01
-3.59942079e-01 -1.56602770e-01 8.87769461e-01 -1.65075570e-01
1.00550175e+00 -2.56248331e+00 1.30026285e-02 4.66759324e-01
-2.98085630e-01 4.85112131e-01 -1.55924276e-01 2.63193641e-02
-3.07608128e-01 -2.06996724e-01 -4.24226448e-02 -2.07261071e-01
-1.81283548e-01 2.07069710e-01 -2.17183724e-01 3.58418465e-01
1.48370758e-01 3.53763342e-01 -6.68608487e-01 -2.47251600e-01
1.25815682e-02 2.48970211e-01 -3.96219036e-03 6.67360783e-01
1.47946654e-02 1.11339219e-01 -6.07415795e-01 7.23092437e-01
8.43199611e-01 1.08129896e-01 3.10439467e-01 -3.82736892e-01
2.63612479e-01 -3.95700544e-01 -1.04732990e+00 1.11337006e+00
-7.08010614e-01 4.85657215e-01 7.29523599e-03 -1.39146459e+00
1.50776935e+00 3.27907830e-01 4.30215836e-01 -8.29940438e-01
3.80407959e-01 2.30310500e-01 -1.15388751e-01 -3.85349959e-01
2.59312302e-01 -6.19849265e-01 -3.12559932e-01 3.13747495e-01
5.30005023e-02 1.33015096e-01 6.93032369e-02 -3.02781135e-01
5.51761448e-01 -2.29255050e-01 4.38514829e-01 7.06626251e-02
9.05862629e-01 -2.19006717e-01 9.99453187e-01 2.28123873e-01
-2.49465346e-01 2.79214263e-01 7.01041281e-01 -1.01040848e-01
-5.99364281e-01 -8.04013312e-01 5.04060090e-03 9.03468013e-01
-3.79362293e-02 -3.47731858e-01 -8.68183911e-01 -8.68953943e-01
-2.27862701e-01 6.16304159e-01 -7.58704543e-01 -3.08506995e-01
8.97264406e-02 -9.22240913e-01 8.45572412e-01 3.18350881e-01
7.20141411e-01 -8.86180282e-01 -3.05623710e-01 1.99245155e-01
-2.24324450e-01 -1.01300347e+00 -2.60603428e-01 6.43064231e-02
-5.03710806e-01 -1.07734013e+00 -6.52200997e-01 -8.17535520e-01
7.31715024e-01 4.82249893e-02 8.25054288e-01 -2.29935989e-01
-2.04107150e-01 -4.23718020e-02 -5.65485656e-01 -5.25300562e-01
-6.79493666e-01 -3.09112906e-01 4.35451530e-02 8.05967152e-01
8.09776604e-01 -3.52899432e-01 -1.47125855e-01 4.51337785e-01
-8.94888461e-01 -1.71541035e-01 6.16757035e-01 1.10333037e+00
9.41079497e-01 2.78547794e-01 9.42779124e-01 -1.03693593e+00
8.29161882e-01 -5.00394285e-01 -5.79801440e-01 6.61028862e-01
-5.14574289e-01 -1.12477116e-01 9.20422673e-01 -6.67956829e-01
-1.31736910e+00 1.63525313e-01 -2.24418536e-01 -7.09004521e-01
-2.70012856e-01 7.14027703e-01 -7.35742033e-01 1.23008311e-01
3.64157796e-01 5.10661960e-01 1.00158446e-01 -3.64742130e-01
2.75822692e-02 1.09812009e+00 4.86398488e-01 -5.27278244e-01
9.12232473e-02 8.96576494e-02 -2.85603017e-01 -7.41361797e-01
-6.58123553e-01 -2.80383110e-01 -2.34138712e-01 -8.14471170e-02
6.81988418e-01 -9.59031343e-01 -4.85409260e-01 1.00620151e+00
-9.57318246e-01 -1.56141026e-02 1.96436401e-02 4.36167747e-01
-3.29209775e-01 1.78454503e-01 -3.64574760e-01 -9.68949139e-01
-3.30368191e-01 -1.19036210e+00 1.09310818e+00 4.31286126e-01
-1.70325264e-01 -7.96461642e-01 1.45173788e-01 1.45471737e-01
8.21469948e-02 3.71105880e-01 9.35432732e-01 -1.00615776e+00
3.05728644e-01 -6.07930899e-01 -2.82161850e-02 7.71037757e-01
5.88343918e-01 1.83259934e-01 -9.60460484e-01 -2.86417035e-03
2.79841512e-01 -5.93416631e-01 2.29161099e-01 -2.37056926e-01
1.22852254e+00 -1.50985450e-01 -2.11796433e-01 4.09363627e-01
1.23538756e+00 6.23739302e-01 7.38490701e-01 -6.33644015e-02
2.55960226e-01 7.19747961e-01 1.17482197e+00 7.20040500e-01
8.85742754e-02 8.15507293e-01 -2.96307892e-01 -1.60497174e-01
6.72027111e-01 -2.02450648e-01 5.01175463e-01 8.37104440e-01
4.23315883e-01 -4.24149185e-01 -5.46957731e-01 3.16535503e-01
-1.68076301e+00 -9.62000966e-01 2.38381445e-01 2.06281424e+00
9.64064240e-01 -3.02731484e-01 -3.76392831e-03 3.20616156e-01
7.60601997e-01 2.49911449e-03 -6.77684128e-01 -6.00893617e-01
-2.69929647e-01 4.17877138e-01 -1.33256942e-01 5.52225336e-02
-9.90219116e-01 6.39939129e-01 5.67647314e+00 1.18082654e+00
-1.61228681e+00 -2.78536510e-02 1.15658796e+00 -8.77861828e-02
3.90642360e-02 -3.53991866e-01 -6.13083839e-01 5.39972186e-01
9.36949968e-01 -6.05281472e-01 -5.74278238e-04 1.10471070e+00
-4.85851342e-04 -2.51490384e-01 -1.07687855e+00 1.44241643e+00
4.47211206e-01 -8.38800311e-01 -1.28640994e-01 -3.12817127e-01
4.96400177e-01 -4.05696571e-01 -4.62648831e-02 6.19697034e-01
-5.08181527e-02 -8.46302927e-01 2.47497022e-01 4.33991373e-01
8.47887695e-01 -1.12554753e+00 9.85276580e-01 2.94090867e-01
-8.90173912e-01 1.19438030e-01 -4.05427724e-01 2.66217142e-01
-2.31658131e-01 8.21233392e-01 -5.79937041e-01 6.62111461e-01
3.87927234e-01 6.32383645e-01 -4.61293668e-01 4.77861017e-01
-2.44861953e-02 5.30369163e-01 -1.37955680e-01 -1.55013755e-01
-2.00757369e-01 -5.23597479e-01 7.30423722e-03 1.15825784e+00
6.20127857e-01 2.53814667e-01 4.51540761e-02 6.94052100e-01
-1.84334114e-01 6.84114754e-01 -4.92968917e-01 -1.77836061e-01
2.90399849e-01 1.24576461e+00 -2.67740995e-01 -6.62421882e-01
-4.00248319e-01 1.25027144e+00 3.43930513e-01 3.44222397e-01
-8.13105941e-01 -6.68515086e-01 8.06429386e-01 -3.29623938e-01
7.95871094e-02 3.81768435e-01 2.05423608e-01 -1.50060272e+00
1.22946441e-01 -1.18859768e+00 4.24274743e-01 -7.96110868e-01
-1.49267304e+00 8.85255873e-01 -1.70222640e-01 -1.32033753e+00
-4.56167281e-01 -4.36903924e-01 -5.85537553e-01 8.07965279e-01
-1.05828285e+00 -8.42216611e-01 -3.99272680e-01 7.32929826e-01
2.66155541e-01 -3.91097665e-01 1.21041608e+00 4.48695391e-01
-8.96116614e-01 1.03678370e+00 4.71611619e-01 3.72698933e-01
8.47654521e-01 -7.14541435e-01 -3.75360876e-01 6.29188657e-01
2.05147684e-01 3.88145089e-01 3.81528944e-01 -2.92888254e-01
-1.28713465e+00 -1.16875553e+00 4.87149864e-01 2.78626740e-01
3.43586206e-01 -5.22956610e-01 -1.02827895e+00 3.53163213e-01
-1.27461143e-02 2.41547957e-01 1.18701780e+00 -1.88547559e-02
-4.00450766e-01 -5.74179173e-01 -1.42744148e+00 3.99339259e-01
3.33062023e-01 -6.19101167e-01 -3.23192656e-01 -1.53873826e-03
-8.24387670e-02 -2.07102329e-01 -9.53715444e-01 4.99747396e-01
6.84201300e-01 -1.09512305e+00 4.21124607e-01 -4.77229387e-01
2.79913664e-01 -2.51802593e-01 -4.88900602e-01 -1.74109113e+00
2.04982057e-01 -1.43568814e-01 2.02082425e-01 1.50963438e+00
2.94148982e-01 -6.26102030e-01 8.05993617e-01 6.77561164e-01
3.03511918e-01 -8.56373012e-01 -9.96840656e-01 -6.99686706e-01
-1.48124412e-01 -3.33031446e-01 7.68091083e-01 9.44393218e-01
4.59547341e-01 5.00314951e-01 -4.35174227e-01 -2.42540270e-01
2.84292519e-01 4.58616078e-01 9.31665003e-01 -7.64550805e-01
-4.08585757e-01 -1.51887536e-01 -6.25734925e-01 -7.32503891e-01
6.94324434e-01 -6.77961946e-01 2.15881720e-01 -6.53608561e-01
2.00144172e-01 -4.90867376e-01 -4.31796581e-01 3.80389184e-01
-3.48747015e-01 2.67496943e-01 9.87009518e-03 -9.52136591e-02
-4.25274611e-01 1.01258469e+00 8.03413451e-01 -1.35947421e-01
-2.75698245e-01 1.37983888e-01 -6.86998487e-01 3.97894859e-01
8.63829315e-01 -3.86763066e-01 -5.98865688e-01 1.38375759e-01
-3.75434339e-01 2.58249521e-01 -1.39422894e-01 -8.40442955e-01
-1.71126023e-01 -2.36178324e-01 3.89268160e-01 -3.10907096e-01
3.21692824e-01 -6.36151433e-01 1.95854127e-01 -8.09721276e-03
-2.43581370e-01 -6.02158569e-02 4.06519353e-01 4.18947041e-01
-8.09870780e-01 -1.76179722e-01 9.34437156e-01 1.46702051e-01
-6.36531651e-01 1.52450070e-01 -7.61527061e-01 -1.03541680e-01
1.28301692e+00 -1.80298179e-01 -4.04851288e-02 -5.31639338e-01
-6.40429139e-01 -1.26421511e-01 3.92618358e-01 4.83048797e-01
7.15456903e-01 -1.54323423e+00 -5.51287711e-01 4.20010328e-01
6.60009563e-01 -2.92623281e-01 2.50106931e-01 6.91128135e-01
8.84556174e-02 3.38460989e-02 -3.47308308e-01 -3.77645552e-01
-1.68353498e+00 5.01418173e-01 5.07188797e-01 -3.16365153e-01
2.33044937e-01 5.10361075e-01 2.86566198e-01 -5.07198215e-01
-1.14714630e-01 -2.40331739e-02 -1.71844587e-01 1.69992700e-01
6.66135311e-01 2.15592366e-02 1.62769422e-01 -8.02374244e-01
-3.08416545e-01 2.90579528e-01 -2.25522041e-01 -9.82366428e-02
1.12273455e+00 1.29183337e-01 -1.45167023e-01 6.29251778e-01
1.68061781e+00 1.63168237e-01 -5.68045557e-01 -1.67237133e-01
-1.26242086e-01 -6.40926421e-01 -2.11806864e-01 -6.57328367e-01
-1.24467373e+00 6.69159472e-01 8.70140016e-01 -2.69802809e-01
1.58446324e+00 -2.08286971e-01 4.11335170e-01 2.78491437e-01
3.32985878e-01 -1.29883802e+00 1.01798549e-01 1.00161098e-01
7.15781987e-01 -1.11154640e+00 -3.62606108e-01 -5.67478836e-01
-9.33910370e-01 1.09868789e+00 1.04269266e+00 8.29337314e-02
4.64042902e-01 3.54723752e-01 4.29475933e-01 4.48594205e-02
-8.07671130e-01 2.32279107e-01 -6.82587177e-03 5.49623191e-01
5.65983534e-01 2.12443560e-01 -1.79004446e-01 9.36163425e-01
1.32881433e-01 2.25538224e-01 3.72822523e-01 7.44745910e-01
-5.96109871e-03 -1.50961173e+00 -4.67725515e-01 5.89173853e-01
-5.51596582e-01 4.04998153e-01 -5.75270891e-01 5.20452678e-01
-5.11733294e-02 9.93289888e-01 2.44126916e-02 -8.44140887e-01
5.01138031e-01 3.20635080e-01 4.77055579e-01 -3.71327788e-01
-3.17486882e-01 5.89310713e-02 3.01378489e-01 -3.42695415e-01
-5.65276444e-01 -5.87709963e-01 -1.21544981e+00 -2.66030636e-02
-4.57038671e-01 6.22304678e-01 4.42410916e-01 8.26883376e-01
6.41907394e-01 1.76868841e-01 1.10867405e+00 -4.55912024e-01
-6.20371342e-01 -1.06036270e+00 -8.99303973e-01 8.01880300e-01
-9.53958184e-02 -6.69217050e-01 -1.90571725e-01 -4.30129580e-02] | [13.660445213317871, 1.783782958984375] |
4bd9949a-8514-460b-a365-84ceac9d0da7 | axm-net-cross-modal-context-sharing-attention | 2101.08238 | null | https://arxiv.org/abs/2101.08238v3 | https://arxiv.org/pdf/2101.08238v3.pdf | AXM-Net: Implicit Cross-Modal Feature Alignment for Person Re-identification | Cross-modal person re-identification (Re-ID) is critical for modern video surveillance systems. The key challenge is to align cross-modality representations induced by the semantic information present for a person and ignore background information. This work presents a novel convolutional neural network (CNN) based architecture designed to learn semantically aligned cross-modal visual and textual representations. The underlying building block, named AXM-Block, is a unified multi-layer network that dynamically exploits the multi-scale knowledge from both modalities and re-calibrates each modality according to shared semantics. To complement the convolutional design, contextual attention is applied in the text branch to manipulate long-term dependencies. Moreover, we propose a unique design to enhance visual part-based feature coherence and locality information. Our framework is novel in its ability to implicitly learn aligned semantics between modalities during the feature learning stage. The unified feature learning effectively utilizes textual data as a super-annotation signal for visual representation learning and automatically rejects irrelevant information. The entire AXM-Net is trained end-to-end on CUHK-PEDES data. We report results on two tasks, person search and cross-modal Re-ID. The AXM-Net outperforms the current state-of-the-art (SOTA) methods and achieves 64.44\% Rank@1 on the CUHK-PEDES test set. It also outperforms its competitors by $>$10\% in cross-viewpoint text-to-image Re-ID scenarios on CrossRe-ID and CUHK-SYSU datasets. | ['Syed Safwan Khalid', 'Josef Kittler', 'Muhammad Awais', 'Ammarah Farooq'] | 2021-01-19 | null | null | null | null | ['cross-view-person-re-identification', 'nlp-based-person-retrival', 'person-search'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 1.28166854e-01 -4.19570148e-01 -3.05400282e-01 -3.64399880e-01
-7.75625706e-01 -4.46637481e-01 7.70495832e-01 -3.02796423e-01
-7.92543709e-01 3.97330254e-01 5.14446557e-01 2.27620780e-01
6.32333308e-02 -3.86335820e-01 -7.14836061e-01 -3.84017318e-01
2.36979425e-01 4.85214025e-01 1.28555104e-01 -2.08853438e-01
-2.93342263e-01 2.56392837e-01 -1.76647079e+00 7.33640075e-01
4.30619836e-01 1.04935789e+00 2.57630255e-02 5.98567426e-01
1.66775420e-01 5.95010042e-01 -3.03604335e-01 -8.22747886e-01
3.71684670e-01 -2.92802095e-01 -7.79365659e-01 2.42747992e-01
1.19545233e+00 -3.08056027e-01 -8.33863497e-01 8.77440751e-01
7.93119729e-01 2.32747972e-01 5.62916040e-01 -1.33804941e+00
-8.85347009e-01 1.29794642e-01 -6.94234490e-01 4.90401208e-01
4.60530996e-01 2.63278127e-01 7.79750168e-01 -1.06818974e+00
5.93040586e-01 1.33506489e+00 9.26880300e-01 9.89687026e-01
-1.29133999e+00 -6.50557101e-01 4.06124383e-01 4.87480491e-01
-1.72065294e+00 -4.88561630e-01 5.95260918e-01 -5.04882812e-01
1.26456666e+00 3.24777961e-01 5.80123603e-01 1.51259506e+00
-2.15447396e-01 9.90894556e-01 7.52862155e-01 -3.07862788e-01
-4.00687069e-01 2.31648251e-01 1.28256336e-01 8.38924646e-01
1.46579117e-01 3.72534335e-01 -7.88666725e-01 3.23357545e-02
4.83946621e-01 3.90481979e-01 -3.55567485e-01 -4.31943268e-01
-1.22304463e+00 4.70300376e-01 5.88718414e-01 2.66919494e-01
3.59766632e-02 1.09335028e-01 7.60580778e-01 2.65336484e-01
2.07272857e-01 1.42874435e-01 -5.97615719e-01 1.65464878e-01
-7.68312156e-01 2.30522230e-01 2.44110107e-01 1.02035809e+00
3.94647002e-01 -1.38884038e-01 -8.43665063e-01 9.86750722e-01
2.88539588e-01 8.63204062e-01 4.41914260e-01 -4.02102113e-01
6.15944326e-01 5.88761032e-01 3.02518308e-02 -8.09219360e-01
-4.37504619e-01 -7.08409488e-01 -1.04954910e+00 -9.47298780e-02
2.86987901e-01 1.24773094e-02 -1.18380129e+00 1.92521751e+00
1.76454529e-01 3.25905472e-01 2.19905414e-02 1.06443453e+00
1.30716801e+00 1.70056537e-01 4.65549678e-01 1.62443921e-01
1.75988126e+00 -1.32098675e+00 -4.73702401e-01 -2.47355223e-01
4.70996052e-01 -5.18332779e-01 7.79801667e-01 -1.50983110e-01
-7.63657391e-01 -1.01647270e+00 -8.41247261e-01 -2.20776305e-01
-7.07000375e-01 5.62685370e-01 1.33402705e-01 7.10132480e-01
-1.15852714e+00 3.55061516e-02 -3.79854202e-01 -6.64917290e-01
5.22478819e-01 3.76056045e-01 -8.52467418e-01 -2.92123646e-01
-1.20637691e+00 8.41395378e-01 5.54057121e-01 3.01862448e-01
-9.19717789e-01 -6.12240434e-01 -1.09196556e+00 1.07897982e-01
4.61106509e-01 -1.15615213e+00 7.47548223e-01 -1.20235038e+00
-1.14169562e+00 1.30750012e+00 -3.79079700e-01 -3.65613133e-01
7.18244612e-01 -1.29006892e-01 -7.32873261e-01 1.83109105e-01
3.65880191e-01 8.22437823e-01 1.03494620e+00 -1.29243088e+00
-8.74522984e-01 -4.69431788e-01 -1.23967588e-01 2.27802157e-01
-4.48670059e-01 5.38278334e-02 -1.12487674e+00 -9.96963799e-01
-3.38143855e-01 -1.09525859e+00 1.29303291e-01 -1.18082985e-01
-3.89950782e-01 -3.00803930e-01 7.12067485e-01 -6.75325334e-01
1.02392328e+00 -2.00607514e+00 3.09471905e-01 2.20069259e-01
3.32769662e-01 4.08872485e-01 -4.48889852e-01 1.70840338e-01
-3.32883298e-01 -1.77901790e-01 1.50605127e-01 -8.19580674e-01
-4.34457278e-03 -1.66643709e-01 -4.85511124e-02 5.80171168e-01
8.76929536e-02 1.36605072e+00 -7.42152333e-01 -4.36304659e-01
4.73806113e-01 6.54655874e-01 -3.47612709e-01 2.23055109e-01
1.39233768e-01 3.91894042e-01 -7.74325207e-02 8.89341414e-01
7.30760396e-01 -5.58102727e-01 2.22661257e-01 -7.15411365e-01
3.33380662e-02 -4.65151161e-01 -8.81381452e-01 1.82960773e+00
-2.09146932e-01 6.80812657e-01 -2.62594342e-01 -9.12442744e-01
5.37555337e-01 2.82581061e-01 4.74093109e-01 -9.86418486e-01
1.34158328e-01 -9.20087695e-02 -6.13643527e-01 -5.58865368e-01
5.59892476e-01 2.91387379e-01 -2.61694223e-01 2.85166018e-02
4.19481844e-01 9.04914439e-01 9.46410820e-02 1.87672883e-01
6.37027204e-01 6.00583330e-02 2.02077359e-01 -1.93958133e-01
1.02557981e+00 -8.31163749e-02 4.80486929e-01 9.89506006e-01
-3.07320625e-01 5.86587608e-01 -5.40534966e-02 -8.66932154e-01
-9.03137088e-01 -1.14478827e+00 -4.77016941e-02 1.50979352e+00
4.60268021e-01 -6.58497036e-01 -4.43436444e-01 -9.83259082e-01
8.86883736e-02 2.69527733e-01 -1.08348405e+00 -1.63820133e-01
-4.45135623e-01 -6.59960389e-01 6.88974857e-01 7.92601466e-01
8.87754202e-01 -6.84680104e-01 -2.70979375e-01 -1.75688550e-01
-4.67018962e-01 -1.31519759e+00 -1.00402331e+00 -2.90844321e-01
-1.66240230e-01 -1.34601784e+00 -9.09326077e-01 -9.02243733e-01
7.38978863e-01 6.21427238e-01 1.19977522e+00 5.99337183e-02
-4.09817249e-01 1.01688278e+00 -3.25459808e-01 1.13764599e-01
2.21599698e-01 1.66397989e-01 1.04227275e-01 4.52885419e-01
8.36975515e-01 1.19643390e-01 -6.87203348e-01 5.12740731e-01
-6.56257510e-01 6.50338829e-02 4.26493347e-01 1.19708669e+00
6.80977881e-01 -2.47229531e-01 3.33216757e-01 -4.23532754e-01
2.22503901e-01 -2.97405362e-01 -4.03303802e-01 8.91876042e-01
-2.50152498e-01 -5.25919497e-02 3.82542521e-01 -5.46440065e-01
-1.10138512e+00 2.69986123e-01 1.40608788e-01 -7.10307181e-01
-2.09645018e-01 1.26695529e-01 -4.10681158e-01 -1.14247434e-01
4.22164321e-01 3.93078893e-01 -3.15856248e-01 -4.29589152e-01
3.88701022e-01 5.42236745e-01 9.76425052e-01 -3.85741204e-01
8.03081274e-01 7.23332524e-01 -2.14398429e-01 -7.48057008e-01
-8.81949842e-01 -7.21736908e-01 -8.23399544e-01 -3.24380487e-01
1.27171016e+00 -1.44595408e+00 -9.19651866e-01 6.86347306e-01
-1.03966081e+00 -5.11941500e-02 -8.55599269e-02 2.45685190e-01
-1.40792415e-01 3.83305013e-01 -2.05885485e-01 -4.17662650e-01
-4.39310461e-01 -1.20110905e+00 1.35532081e+00 4.29539979e-01
1.23164535e-01 -8.81962955e-01 -1.07570432e-01 6.75728083e-01
3.05978537e-01 9.72902924e-02 2.38082305e-01 -6.44401133e-01
-4.43613052e-01 -3.48206550e-01 -6.69788122e-01 2.42678579e-02
6.73398823e-02 -5.89340210e-01 -1.17160332e+00 -8.15072834e-01
-7.12589264e-01 -3.60914618e-01 1.45876992e+00 2.21443862e-01
1.19310868e+00 -1.18998609e-01 -6.32673264e-01 9.69794095e-01
1.24896991e+00 -3.49268615e-01 4.93834287e-01 5.16168892e-01
1.11788285e+00 5.20616055e-01 2.67537534e-01 3.00885737e-01
7.96998680e-01 1.16003656e+00 1.67466655e-01 -2.50179440e-01
-4.10636574e-01 -3.50775540e-01 2.94372708e-01 4.31322977e-02
-7.36686885e-02 -2.60919422e-01 -6.87090456e-01 5.94882905e-01
-2.18990493e+00 -1.31105435e+00 1.72616646e-01 2.18075514e+00
3.90099436e-01 -3.03489923e-01 4.08018082e-01 -4.49992955e-01
9.57274735e-01 8.28979611e-02 -4.55002904e-01 3.51066113e-01
-4.97965664e-01 -2.37771153e-01 6.76574051e-01 3.45740438e-01
-1.68647993e+00 1.08088577e+00 5.10387230e+00 8.60266447e-01
-9.16922808e-01 3.81516278e-01 4.81388599e-01 -2.67906547e-01
1.91590581e-02 -5.59070587e-01 -1.16076446e+00 4.80226606e-01
5.21942616e-01 2.81779259e-01 2.51749903e-01 6.81315124e-01
-2.10211396e-01 1.98521748e-01 -1.20193768e+00 1.70673227e+00
8.04000199e-01 -1.39067960e+00 1.98003978e-01 -2.40377977e-01
7.66325235e-01 8.84076953e-02 3.14466298e-01 3.35719556e-01
9.92080867e-02 -1.17065012e+00 7.73650169e-01 7.16184735e-01
1.32820702e+00 -6.86600029e-01 8.99719238e-01 -2.50904530e-01
-1.84134901e+00 -2.42715403e-01 -1.22068912e-01 6.53451860e-01
1.77386284e-01 -2.29186669e-01 -5.31201899e-01 6.87046289e-01
1.24002790e+00 1.00246334e+00 -1.10022533e+00 9.14978981e-01
1.53209329e-01 -1.73438847e-01 -8.35897550e-02 4.67513412e-01
4.78022918e-02 3.46768469e-01 5.26013374e-01 1.69655633e+00
4.14111763e-02 -2.72513300e-01 4.47340012e-01 5.87410450e-01
-1.34154931e-01 -3.63790631e-01 -3.37832063e-01 4.28525984e-01
1.45116836e-01 1.02308977e+00 -3.26635271e-01 -5.54040790e-01
-7.32480168e-01 1.58813822e+00 3.92870575e-01 5.76383054e-01
-9.80453968e-01 -5.68251535e-02 8.30084324e-01 -1.31097049e-01
6.26741469e-01 1.92561541e-02 1.20678850e-01 -1.60729587e+00
3.22715677e-02 -8.27197015e-01 1.01988280e+00 -5.89969099e-01
-1.81675458e+00 9.77579176e-01 1.24109425e-01 -1.31613314e+00
-2.73647368e-01 -8.06521893e-01 -1.04758151e-01 9.35321808e-01
-1.72563839e+00 -2.07948542e+00 -7.45814800e-01 1.31285691e+00
6.83989882e-01 -7.97154903e-01 6.72878742e-01 5.24659634e-01
-6.67013645e-01 1.32977879e+00 -1.05850594e-02 5.73650360e-01
9.54665482e-01 -9.33400273e-01 2.69211859e-01 1.02650619e+00
6.71032593e-02 5.83544135e-01 2.93362826e-01 -7.95444548e-01
-1.29871809e+00 -1.50981009e+00 8.97114873e-01 -7.92385101e-01
2.92683065e-01 -4.07549024e-01 -6.05641067e-01 7.61714160e-01
1.55333504e-01 5.00520766e-01 8.77933979e-01 5.76825067e-02
-9.76597309e-01 -2.50040978e-01 -9.78971422e-01 5.66762745e-01
1.42737436e+00 -9.45592880e-01 -4.64416832e-01 1.89844161e-01
4.80368555e-01 -3.61005604e-01 -5.89672506e-01 4.06851888e-01
8.19753289e-01 -8.34857523e-01 1.49547732e+00 -9.68476534e-01
-1.58003747e-01 -6.09149337e-01 -2.34644890e-01 -9.25072372e-01
-6.25383735e-01 -3.23880345e-01 -1.13637350e-01 1.12037671e+00
1.70304805e-01 -5.42083085e-01 6.16513312e-01 6.17618024e-01
1.41881883e-01 -2.41538852e-01 -1.06189144e+00 -6.82167709e-01
-4.74118859e-01 -3.46776307e-01 4.37927693e-01 9.05309200e-01
-4.53359812e-01 3.21951687e-01 -8.76485825e-01 4.09565330e-01
9.66046870e-01 1.19390348e-02 8.62272918e-01 -1.03609359e+00
-2.53319472e-01 -3.47142130e-01 -5.20974934e-01 -1.10302234e+00
3.94869328e-01 -1.06214523e+00 -3.90582263e-01 -1.18617141e+00
6.39812529e-01 -1.08550988e-01 -5.62026978e-01 5.64831495e-01
-2.75252342e-01 6.98458552e-01 5.35553634e-01 2.48524949e-01
-1.13731098e+00 6.42801344e-01 7.56664991e-01 -6.45309865e-01
-1.50509983e-01 -2.08891079e-01 -5.57682991e-01 4.13679481e-01
5.31772256e-01 -1.18309408e-01 -2.85251647e-01 -6.60839260e-01
-1.02721691e-01 -4.16862637e-01 1.09862947e+00 -9.40436900e-01
4.69369620e-01 1.38017565e-01 1.06894183e+00 -7.87227392e-01
4.56590891e-01 -1.06219602e+00 1.85143277e-01 1.75366014e-01
-4.86995816e-01 2.49723330e-01 2.93757468e-01 8.63069594e-01
-8.11913833e-02 2.09729671e-01 6.83240175e-01 -8.47987980e-02
-1.19935167e+00 6.64446414e-01 6.77367598e-02 3.30113322e-02
1.04412067e+00 -3.67361546e-01 -5.05286515e-01 -1.08309194e-01
-7.83189893e-01 4.39617962e-01 4.34802681e-01 9.09907520e-01
8.24389696e-01 -1.60158122e+00 -7.60903060e-01 4.19162393e-01
6.95198178e-01 -5.08901715e-01 8.98558795e-01 6.79716051e-01
-9.31835175e-02 6.25631750e-01 -3.05074662e-01 -7.93691039e-01
-1.63589239e+00 7.65625000e-01 7.04448640e-01 -2.18164042e-01
-5.99466920e-01 9.33519006e-01 4.56157506e-01 -2.82975495e-01
4.40037638e-01 3.02517742e-01 -3.69205683e-01 3.71818133e-02
9.49699044e-01 2.42218703e-01 -1.19209439e-01 -1.30939376e+00
-6.81943715e-01 7.53783584e-01 -2.32822299e-01 -2.13100738e-03
1.04780817e+00 -4.40562457e-01 3.30060609e-02 -1.12963632e-01
1.42048788e+00 -2.98958570e-01 -1.32469070e+00 -6.67178929e-01
-1.91755325e-01 -7.14323223e-01 -1.49352670e-01 -9.69502747e-01
-1.16243923e+00 4.65042204e-01 1.18013263e+00 -3.56938452e-01
1.08615243e+00 2.81514347e-01 6.07687175e-01 3.49530101e-01
5.39430827e-02 -1.17617035e+00 3.54352623e-01 4.26423758e-01
9.12541509e-01 -1.78235316e+00 -5.91699891e-02 -1.44780353e-01
-8.93249393e-01 9.76417482e-01 1.03422201e+00 3.12712282e-01
5.00422001e-01 -2.67751247e-01 -7.52637610e-02 -1.61282524e-01
-4.10804808e-01 -6.73731148e-01 9.60157037e-01 1.01854479e+00
1.21366382e-01 9.34502296e-03 4.05134797e-01 6.48390055e-01
1.98163882e-01 -1.52822778e-01 -3.18160146e-01 4.98694509e-01
7.10506290e-02 -9.12121415e-01 -4.70724314e-01 1.90666169e-01
-3.06708485e-01 -1.44488707e-01 -3.17960113e-01 7.76155531e-01
5.86582363e-01 9.44274783e-01 1.10847838e-01 -6.19039714e-01
3.15679818e-01 -1.28603846e-01 4.37379241e-01 -1.33297652e-01
-7.08700597e-01 -1.60891026e-01 7.23795965e-02 -6.78618133e-01
-7.03486323e-01 -5.68257689e-01 -5.55361271e-01 -2.45845482e-01
3.68839167e-02 -1.60324633e-01 1.92858338e-01 8.87570620e-01
6.19249880e-01 5.58607280e-01 2.24365935e-01 -9.28309739e-01
-1.35263398e-01 -7.69101143e-01 -8.74528065e-02 7.34862447e-01
6.26526535e-01 -9.18519795e-01 6.25693277e-02 2.56320298e-01] | [14.675957679748535, 0.9285708069801331] |
a44eb307-f0fd-4d74-9d77-08a4108ad7f7 | minimum-latency-deep-online-video | 2212.02073 | null | https://arxiv.org/abs/2212.02073v1 | https://arxiv.org/pdf/2212.02073v1.pdf | Minimum Latency Deep Online Video Stabilization | We present a novel camera path optimization framework for the task of online video stabilization. Typically, a stabilization pipeline consists of three steps: motion estimating, path smoothing, and novel view rendering. Most previous methods concentrate on motion estimation, proposing various global or local motion models. In contrast, path optimization receives relatively less attention, especially in the important online setting, where no future frames are available. In this work, we adopt recent off-the-shelf high-quality deep motion models for the motion estimation to recover the camera trajectory and focus on the latter two steps. Our network takes a short 2D camera path in a sliding window as input and outputs the stabilizing warp field of the last frame in the window, which warps the coming frame to its stabilized position. A hybrid loss is well-defined to constrain the spatial and temporal consistency. In addition, we build a motion dataset that contains stable and unstable motion pairs for the training. Extensive experiments demonstrate that our approach significantly outperforms state-of-the-art online methods both qualitatively and quantitatively and achieves comparable performance to offline methods. | ['Shuaicheng Liu', 'Bing Zeng', 'Zhen Liu', 'Zhuofan Zhang'] | 2022-12-05 | null | null | null | null | ['video-stabilization'] | ['computer-vision'] | [-1.93567008e-01 -4.45101410e-01 -4.79366153e-01 -1.21677659e-01
-6.12426460e-01 -5.29451132e-01 4.31563973e-01 -1.74045324e-01
-4.55726773e-01 4.36101735e-01 3.12851638e-01 -1.74603626e-01
3.81871849e-01 -2.81990409e-01 -8.41101706e-01 -8.42508972e-01
4.99102287e-02 -1.36465907e-01 6.33081138e-01 -7.59551898e-02
3.84778291e-01 4.31593567e-01 -1.09424007e+00 -1.47648379e-01
5.71063340e-01 8.72460485e-01 2.66463369e-01 8.25444341e-01
3.74029815e-01 9.54504430e-01 7.29382262e-02 -1.67383209e-01
2.55842388e-01 -5.47695041e-01 -6.84673250e-01 2.55901694e-01
5.56517780e-01 -8.43128324e-01 -8.34387124e-01 9.34331894e-01
5.19564748e-01 6.09713197e-01 8.19554254e-02 -9.67918932e-01
-2.20125139e-01 2.64604688e-01 -7.79169679e-01 3.51061702e-01
5.14438331e-01 4.77737606e-01 7.01403320e-01 -9.82581854e-01
9.57838595e-01 1.02817786e+00 4.88034368e-01 7.17780232e-01
-1.06938493e+00 -3.57696325e-01 5.24693489e-01 2.97100663e-01
-1.11081910e+00 -6.52744174e-01 8.58502686e-01 -4.90410417e-01
7.17394650e-01 -7.19892755e-02 7.10817099e-01 1.01576245e+00
3.78938109e-01 8.09763312e-01 1.66207403e-01 -2.19733879e-01
1.05200201e-01 -4.38499570e-01 -1.06653042e-01 7.41943896e-01
-3.41551602e-01 4.44903933e-02 -6.90901577e-01 6.89467043e-02
1.10228276e+00 -9.50325746e-03 -7.22261965e-01 -6.38723195e-01
-1.59539306e+00 4.65982527e-01 1.54818788e-01 -5.48330769e-02
-3.79144430e-01 3.48523200e-01 3.53485793e-01 6.13488294e-02
4.36616391e-01 -3.38319540e-02 -2.69340038e-01 -3.95315737e-01
-1.15510571e+00 3.83859873e-01 5.20688891e-01 9.97612715e-01
6.92114830e-01 1.39361963e-01 -2.90977359e-01 4.02717173e-01
2.65898526e-01 1.59595624e-01 2.46500552e-01 -1.50102258e+00
6.51914656e-01 1.01261862e-01 6.60167634e-01 -1.28031540e+00
-2.00131625e-01 1.34596806e-02 -8.02752376e-01 1.52280271e-01
4.77960259e-01 -2.03103483e-01 -5.78224719e-01 1.85927129e+00
7.26133525e-01 8.42831671e-01 -3.58779728e-01 1.30947220e+00
4.17475015e-01 8.64876986e-01 -2.24266842e-01 -5.90755641e-01
7.75104463e-01 -1.58155608e+00 -8.07287216e-01 -2.13649467e-01
3.48414630e-01 -8.55296850e-01 8.80006433e-01 1.85411111e-01
-1.47776365e+00 -5.51396310e-01 -9.18130457e-01 -3.52719188e-01
4.08229887e-01 2.53993481e-01 2.80560702e-01 -4.48389128e-02
-1.11229038e+00 9.27562416e-01 -1.38027465e+00 -2.89781928e-01
2.90538799e-02 2.24084169e-01 -4.02557433e-01 2.02694405e-02
-8.83730471e-01 5.60913146e-01 -6.23771921e-02 3.54352266e-01
-1.07461953e+00 -6.33809805e-01 -1.03728151e+00 -2.88413204e-02
4.72866058e-01 -1.07919776e+00 1.32711422e+00 -1.02311897e+00
-2.00214887e+00 6.95460320e-01 -5.11129022e-01 -3.24703187e-01
9.12484109e-01 -4.27674979e-01 -4.86133300e-04 2.17207968e-01
1.92401800e-02 5.98499656e-01 1.13810647e+00 -9.06121492e-01
-6.80493414e-01 3.90877239e-02 9.81628895e-02 3.35046321e-01
-1.75705999e-01 2.30876237e-01 -1.25657582e+00 -8.77647698e-01
1.12043522e-01 -1.05789244e+00 -3.95114332e-01 3.79617453e-01
-2.72730529e-01 2.15328500e-01 7.31471062e-01 -7.25414038e-01
1.68135285e+00 -2.16929364e+00 6.71163857e-01 -1.63925439e-01
1.96954921e-01 1.84908986e-01 -2.50815880e-02 2.29384005e-01
-1.46995664e-01 -3.18849981e-01 -4.19145450e-02 -8.92796814e-01
-2.70684987e-01 -2.49103233e-01 -4.56229895e-01 7.91039705e-01
-9.10818502e-02 6.44233346e-01 -9.74226594e-01 -3.64824474e-01
4.52219844e-01 4.76266384e-01 -6.94969893e-01 4.43765402e-01
-8.49290714e-02 8.92145276e-01 -2.42129654e-01 4.42891747e-01
5.57438433e-01 -2.78353781e-01 9.13056880e-02 -1.68818831e-01
-4.63299543e-01 1.44639835e-01 -1.26882005e+00 2.24342299e+00
-2.65364528e-01 7.85024405e-01 2.21308291e-01 -4.65576977e-01
4.38410372e-01 1.94848001e-01 7.38729358e-01 -2.05607533e-01
2.39690408e-01 -3.15340981e-02 -3.83961111e-01 -5.34674466e-01
8.00940990e-01 3.74384701e-01 4.04245704e-01 2.45254427e-01
-2.43680894e-01 1.43561512e-01 1.73128709e-01 1.81268767e-01
8.50547791e-01 7.45026052e-01 3.02864462e-01 6.21994492e-03
7.80275047e-01 -1.43784985e-01 8.92230332e-01 2.13543534e-01
-6.21819496e-01 1.12707877e+00 5.65097451e-01 -5.37018478e-01
-1.15335274e+00 -7.18478858e-01 4.37808454e-01 8.46465826e-01
6.86839938e-01 -5.55570960e-01 -9.33824897e-01 -3.11578453e-01
-3.79051208e-01 2.79862970e-01 -4.61624146e-01 -4.32827696e-02
-1.13251150e+00 -2.89811134e-01 -1.27909482e-01 6.24933898e-01
3.13495040e-01 -8.27677727e-01 -6.21818483e-01 3.83659333e-01
-4.28142726e-01 -1.22306108e+00 -1.39547861e+00 -5.70513725e-01
-1.07426834e+00 -9.30203497e-01 -8.68034720e-01 -8.63020003e-01
5.99391997e-01 7.82286108e-01 8.76451075e-01 2.28299052e-01
2.82790780e-01 1.50229529e-01 -1.29470944e-01 3.70341182e-01
-8.57726783e-02 -6.36646077e-02 1.56618729e-01 3.48397702e-01
-1.81175515e-01 -4.39152688e-01 -1.13033605e+00 5.27403533e-01
-6.93375766e-01 3.25889438e-01 5.61721474e-02 7.32511044e-01
7.86089480e-01 -3.98059189e-01 -1.40116751e-01 -3.18476707e-01
2.05148011e-02 -3.98136526e-01 -7.63195693e-01 -4.61583808e-02
-1.62488848e-01 -1.23631209e-01 6.83533669e-01 -5.85888207e-01
-9.24717546e-01 3.80337596e-01 -1.50857894e-02 -1.01220429e+00
2.93586999e-01 1.92760706e-01 -2.36523554e-01 -9.85985175e-02
2.44867489e-01 1.02395311e-01 9.75767002e-02 -2.05456555e-01
2.86942065e-01 -7.38080367e-02 8.71818900e-01 -2.38421753e-01
8.01370740e-01 8.95056069e-01 -4.03571576e-02 -6.45237327e-01
-6.79383993e-01 -6.15394950e-01 -7.78713882e-01 -4.83612269e-01
9.15175855e-01 -1.11781812e+00 -8.77987742e-01 7.43000627e-01
-1.30605459e+00 -5.85767686e-01 4.63581569e-02 5.88142037e-01
-7.44292438e-01 7.39839256e-01 -7.16201901e-01 -4.64176387e-01
-2.45852336e-01 -1.48216760e+00 1.17694545e+00 2.13063687e-01
-1.59177944e-01 -9.66463029e-01 3.32018077e-01 -3.52871120e-02
9.80583131e-02 2.88143426e-01 2.45268807e-01 3.38238925e-01
-1.01671994e+00 -1.77645400e-01 8.43391418e-02 1.16235770e-01
-1.16440743e-01 4.64930683e-01 -6.12523854e-01 -4.54434723e-01
-9.62374453e-03 8.16547871e-02 9.08652604e-01 8.38719368e-01
1.16520941e+00 -1.51911885e-01 -3.73713315e-01 1.22625160e+00
1.21927941e+00 3.60294580e-02 7.77417243e-01 5.48034787e-01
9.94918287e-01 4.73019630e-01 1.02000308e+00 4.78158504e-01
5.53569317e-01 9.73203182e-01 5.71794152e-01 1.32544845e-01
8.52395594e-02 -2.88889080e-01 7.25694120e-01 8.79666924e-01
-3.32825959e-01 -3.74273658e-01 -6.43267095e-01 5.42384565e-01
-2.39687848e+00 -1.17652547e+00 -1.85037136e-01 2.47836828e+00
5.89673817e-01 6.63281977e-02 1.21215321e-01 -2.20416650e-01
8.37758243e-01 7.30043352e-01 -6.70953989e-01 1.72831893e-01
-1.83385294e-02 -4.52926636e-01 3.44142795e-01 9.87217069e-01
-1.36739445e+00 1.07025480e+00 5.61998606e+00 5.54542243e-01
-1.21559680e+00 -4.57004830e-02 6.59593344e-01 -3.85954469e-01
-1.18068300e-01 2.57009357e-01 -7.04218984e-01 6.51536226e-01
7.71054387e-01 -7.48465583e-03 4.26348418e-01 7.17407107e-01
1.00253892e+00 -1.41043693e-01 -1.15340090e+00 1.16658998e+00
-3.95564064e-02 -1.58485842e+00 -1.30535781e-01 -2.67021239e-01
8.36355627e-01 -1.46344006e-01 -3.24769244e-02 -2.44067430e-01
-1.33106545e-01 -6.12595558e-01 1.12391472e+00 7.76710033e-01
6.86929286e-01 -7.75348067e-01 3.22719336e-01 3.38295102e-01
-1.48369098e+00 5.20098545e-02 -2.47496620e-01 -5.93509637e-02
6.55891657e-01 3.67925316e-01 3.80781591e-02 4.71067518e-01
6.46228611e-01 1.24375021e+00 -2.88925558e-01 1.05346262e+00
-3.15212488e-01 3.86217773e-01 -8.52485299e-02 4.61097270e-01
2.86742508e-01 -5.74986756e-01 7.47905850e-01 8.75709236e-01
4.21310365e-01 2.21457854e-01 2.31130213e-01 4.53773230e-01
9.56500135e-03 -9.14633796e-02 -3.75290036e-01 4.73068982e-01
3.04223061e-01 1.21450806e+00 -4.74530011e-01 -3.86218488e-01
-5.93620777e-01 1.39470899e+00 3.26712877e-01 4.60058093e-01
-1.18846262e+00 -9.01733711e-02 9.54003990e-01 1.69543177e-01
2.88463563e-01 -6.10997975e-01 6.13721125e-02 -1.69909859e+00
1.88419014e-01 -5.94519258e-01 1.72251284e-01 -7.74494112e-01
-6.63715065e-01 4.56785381e-01 -2.80438542e-01 -1.75736177e+00
-4.67251480e-01 -2.22992718e-01 -9.55716729e-01 7.50214458e-01
-1.39294243e+00 -9.64476824e-01 -5.85874915e-01 5.77709138e-01
8.59843850e-01 1.98609099e-01 1.34446129e-01 3.98563951e-01
-9.46174860e-01 4.16155905e-01 2.55272329e-01 5.11751464e-03
1.11429620e+00 -9.75326240e-01 6.86316252e-01 1.40311718e+00
-3.66963178e-01 5.92314303e-01 8.21669996e-01 -5.00419378e-01
-1.57682335e+00 -1.09383202e+00 8.11974823e-01 -5.63366294e-01
7.71410584e-01 -1.48438245e-01 -8.79509032e-01 8.05467784e-01
1.86272755e-01 3.49932551e-01 3.97172794e-02 -7.01366961e-01
1.46147221e-01 -1.36752501e-02 -5.20157039e-01 9.96193409e-01
1.14374781e+00 -2.81115919e-01 2.36463107e-04 1.23679422e-01
7.57775545e-01 -8.71706605e-01 -5.17302632e-01 2.19370481e-02
7.27865398e-01 -1.08265376e+00 1.02932656e+00 -4.70320463e-01
7.43678689e-01 -6.49821341e-01 1.44784272e-01 -1.03819406e+00
-4.08032417e-01 -1.48332453e+00 -5.05144656e-01 1.15818930e+00
7.14989230e-02 -3.65125202e-02 8.99418175e-01 5.99235177e-01
-2.33132288e-01 -8.36323738e-01 -6.82763278e-01 -3.93869966e-01
-1.51949555e-01 -3.30705702e-01 1.22729219e-01 7.74671912e-01
-2.27181286e-01 1.49451405e-01 -8.00310493e-01 3.46789032e-01
6.04771018e-01 2.13123396e-01 1.14101779e+00 -4.97163743e-01
-2.69086272e-01 -3.05969685e-01 -2.22533405e-01 -1.74744153e+00
3.45475465e-01 -3.30193698e-01 2.37555891e-01 -1.11276555e+00
1.33137643e-01 2.89310873e-01 6.85230792e-02 -1.51361957e-01
-5.48019469e-01 -4.16194499e-02 2.52146631e-01 5.51134884e-01
-6.29327595e-01 7.66047299e-01 1.31322491e+00 1.02821887e-01
-5.26170731e-01 1.61731645e-01 -1.36352912e-01 1.02968299e+00
4.64610308e-01 -1.69757321e-01 -3.59261304e-01 -7.69670844e-01
-4.30476442e-02 4.54552263e-01 4.13400590e-01 -8.90249789e-01
5.63114285e-01 -4.81848121e-01 1.35607734e-01 -7.72721469e-01
3.68277252e-01 -7.02016294e-01 1.37545794e-01 2.82438278e-01
-2.13683143e-01 4.73727793e-01 1.14956098e-02 7.62414634e-01
-2.36053348e-01 4.46107239e-02 1.04381573e+00 6.76960275e-02
-9.11070108e-01 9.19068754e-01 -1.90353081e-01 -6.97619095e-02
1.05756807e+00 -2.03568369e-01 -2.39245147e-02 -6.16392791e-01
-7.50600040e-01 3.65800411e-01 9.99554396e-01 4.26048219e-01
5.97628951e-01 -1.32887959e+00 -4.94045287e-01 -2.60847826e-02
-2.48798445e-01 2.44828984e-01 4.85995859e-01 1.19977868e+00
-1.03140736e+00 1.22125134e-01 6.12907223e-02 -7.77668834e-01
-1.14420617e+00 7.89344788e-01 3.87552530e-01 -1.35117665e-01
-8.53191853e-01 7.79799342e-01 3.78822893e-01 2.11296305e-01
2.93844521e-01 -4.59183186e-01 -1.63788840e-01 -3.37666459e-02
7.44177699e-01 5.77998221e-01 -2.79065788e-01 -7.87831724e-01
-3.48368168e-01 9.07278895e-01 9.43490118e-03 -2.34765574e-01
1.12377024e+00 -7.16631114e-01 -1.94715559e-01 4.30405587e-01
1.23689342e+00 -1.05202310e-02 -2.06484032e+00 -4.33275141e-02
-2.74307996e-01 -8.26807439e-01 -3.16126160e-02 1.40987709e-01
-1.17728591e+00 7.98293412e-01 3.08091193e-01 -3.96742433e-01
1.17116678e+00 -4.31568146e-01 1.07505727e+00 9.29368213e-02
2.51306564e-01 -1.14366853e+00 1.10065512e-01 6.75148547e-01
7.33388364e-01 -1.28734696e+00 7.70644248e-02 -5.20875514e-01
-5.74648023e-01 1.22955334e+00 6.89037919e-01 -2.88345188e-01
5.08753538e-01 8.88647046e-03 9.95171070e-03 3.09526563e-01
-9.09540832e-01 1.87476680e-01 2.08495453e-01 2.37198114e-01
4.06181276e-01 -4.70176131e-01 -1.86419249e-01 1.65683687e-01
2.08595708e-01 4.80730236e-02 8.10498655e-01 8.88869524e-01
-3.20307851e-01 -8.01718414e-01 -2.76377350e-01 -3.14695865e-01
-4.20602292e-01 -1.05391182e-01 -4.03221883e-02 6.18981481e-01
-3.17790449e-01 7.49761343e-01 2.70702690e-02 -2.40469217e-01
2.92017043e-01 -4.51393485e-01 2.87455291e-01 -2.51757354e-01
-3.57113272e-01 4.08352435e-01 -7.04020262e-02 -1.28889740e+00
-5.97733378e-01 -8.53950977e-01 -1.12377787e+00 -6.42154813e-01
-1.56776145e-01 -3.17135811e-01 2.63810813e-01 6.41395330e-01
3.60191196e-01 4.30386394e-01 9.43540454e-01 -1.57564509e+00
-3.69994968e-01 -5.31999707e-01 -1.71412081e-01 3.53175670e-01
8.10584188e-01 -5.43962538e-01 -3.70425582e-01 5.45639277e-01] | [10.557856559753418, -1.374179482460022] |
b161b879-13b6-4786-9b2a-2e571e1a6a45 | physion-evaluating-physical-scene | 2306.15668 | null | https://arxiv.org/abs/2306.15668v1 | https://arxiv.org/pdf/2306.15668v1.pdf | Physion++: Evaluating Physical Scene Understanding that Requires Online Inference of Different Physical Properties | General physical scene understanding requires more than simply localizing and recognizing objects -- it requires knowledge that objects can have different latent properties (e.g., mass or elasticity), and that those properties affect the outcome of physical events. While there has been great progress in physical and video prediction models in recent years, benchmarks to test their performance typically do not require an understanding that objects have individual physical properties, or at best test only those properties that are directly observable (e.g., size or color). This work proposes a novel dataset and benchmark, termed Physion++, that rigorously evaluates visual physical prediction in artificial systems under circumstances where those predictions rely on accurate estimates of the latent physical properties of objects in the scene. Specifically, we test scenarios where accurate prediction relies on estimates of properties such as mass, friction, elasticity, and deformability, and where the values of those properties can only be inferred by observing how objects move and interact with other objects or fluids. We evaluate the performance of a number of state-of-the-art prediction models that span a variety of levels of learning vs. built-in knowledge, and compare that performance to a set of human predictions. We find that models that have been trained using standard regimes and datasets do not spontaneously learn to make inferences about latent properties, but also that models that encode objectness and physical states tend to make better predictions. However, there is still a huge gap between all models and human performance, and all models' predictions correlate poorly with those made by humans, suggesting that no state-of-the-art model is learning to make physical predictions in a human-like way. Project page: https://dingmyu.github.io/physion_v2/ | ['Kevin A. Smith', 'Judith E Fan', 'Daniel LK Yamins', 'Joshua B. Tenenbaum', 'Chuang Gan', 'Daniel Bear', 'Zhenfang Chen', 'Mingyu Ding', 'Hsiao-Yu Tung'] | 2023-06-27 | null | null | null | null | ['video-prediction', 'scene-understanding'] | ['computer-vision', 'computer-vision'] | [ 1.81869984e-01 -4.42374796e-02 -3.67378712e-01 -1.88125625e-01
1.44122494e-02 -4.53464955e-01 9.07474399e-01 2.76477516e-01
9.05642286e-02 7.76515484e-01 3.58846784e-02 -2.07414180e-01
-2.32876822e-01 -9.60749626e-01 -1.07669926e+00 -8.55725706e-01
-9.94388536e-02 4.63555187e-01 6.77647710e-01 -3.62955071e-02
2.90800035e-01 6.76893473e-01 -1.71489370e+00 5.21168768e-01
3.93191397e-01 1.06863081e+00 1.20102271e-01 1.01141608e+00
1.83076188e-01 1.06468225e+00 -3.83911282e-02 3.43783982e-02
2.54374415e-01 -4.53256398e-01 -8.61609519e-01 -2.81704336e-01
5.36314785e-01 -2.59600312e-01 -6.29305243e-01 6.18636668e-01
-8.63549039e-02 1.63720161e-01 1.08537757e+00 -1.23276615e+00
-1.11560953e+00 2.66007166e-02 -8.09031129e-02 1.61339208e-01
2.92663783e-01 7.37702787e-01 1.11232960e+00 -2.34262884e-01
5.52713573e-01 1.17457581e+00 7.47893333e-01 5.82642436e-01
-1.51392674e+00 -4.00013596e-01 2.38315955e-01 3.99897903e-01
-9.53826666e-01 -2.68611699e-01 5.01802027e-01 -8.50018680e-01
1.12721229e+00 3.48596364e-01 8.08008850e-01 1.13512993e+00
6.85693204e-01 3.68148208e-01 1.10998690e+00 -2.88931638e-01
3.47409099e-01 1.33863583e-01 -1.57941293e-04 8.03749382e-01
3.64283383e-01 4.91174787e-01 -7.59554982e-01 -1.52036607e-01
1.03622937e+00 -1.80192336e-01 -3.21821868e-01 -6.25685692e-01
-1.45579493e+00 2.42142022e-01 1.80132270e-01 -1.93616822e-02
-2.56248713e-01 6.74807012e-01 1.47990510e-01 8.15597698e-02
2.06658304e-01 8.50944221e-01 -9.57804680e-01 -2.41835564e-01
-4.33194786e-01 4.50283945e-01 8.76329422e-01 5.00336289e-01
1.03369606e+00 -2.70381778e-01 1.05542585e-01 4.04958010e-01
1.75797805e-01 6.44491017e-01 3.08799714e-01 -1.28444695e+00
-7.53212273e-02 4.07384694e-01 5.35053313e-01 -9.25372124e-01
-2.10651666e-01 3.09112042e-01 -4.61115569e-01 4.58388120e-01
5.27366877e-01 5.55264242e-02 -1.00658822e+00 1.79436147e+00
2.86524236e-01 5.94283342e-01 -7.40155950e-02 8.86242151e-01
7.45919287e-01 9.69366431e-01 3.57318014e-01 -1.52895600e-02
7.80016959e-01 -7.94822216e-01 5.36450520e-02 -1.94907099e-01
6.11964524e-01 -5.25469542e-01 1.13396418e+00 4.12707537e-01
-1.06403804e+00 -9.69274461e-01 -8.88102710e-01 -2.50936532e-03
-4.49927628e-01 -3.06904823e-01 9.14207816e-01 4.30583805e-01
-8.86701822e-01 1.14965892e+00 -1.11573303e+00 -5.33175111e-01
7.61730447e-02 4.38241571e-01 -3.24207962e-01 4.54323769e-01
-9.21854675e-01 1.15653467e+00 3.24043691e-01 -3.32071304e-01
-1.10308874e+00 -1.02700555e+00 -4.45785522e-01 4.64613922e-02
1.65540546e-01 -7.80183852e-01 1.18207049e+00 -1.01588535e+00
-1.49670458e+00 6.53614879e-01 -1.11216344e-01 -2.84858316e-01
3.73035580e-01 -2.49126121e-01 -3.46553057e-01 2.13137329e-01
-4.06615615e-01 9.37327981e-01 4.79751796e-01 -1.57615173e+00
-3.10571641e-01 2.23505616e-01 3.03983122e-01 1.39793083e-01
-6.49515763e-02 -3.13318253e-01 1.04111461e-02 -6.00525066e-02
1.07640643e-02 -1.14577448e+00 -5.10386303e-02 5.34486771e-01
-4.17435944e-01 -9.40311551e-02 7.50809610e-01 -3.13710868e-01
6.68230057e-01 -1.93370509e+00 4.53384370e-02 5.29578440e-02
2.77406752e-01 1.29941836e-01 -6.55313134e-02 5.00437021e-01
-1.29764795e-01 4.71440792e-01 8.84464569e-03 3.93462032e-02
6.30005896e-02 2.95912981e-01 -4.94072437e-01 5.07085502e-01
1.71181753e-01 9.09923375e-01 -9.29271698e-01 -4.75192636e-01
5.71359098e-01 4.24980342e-01 -5.12190938e-01 1.62671879e-01
-8.22323501e-01 6.35145605e-01 -4.98790830e-01 1.83619335e-01
3.38426232e-01 -7.16553926e-01 1.93515345e-01 -2.63107587e-02
-2.24193200e-01 6.02339089e-01 -1.00713074e+00 8.95476878e-01
-2.49033138e-01 6.58093572e-01 -5.72577178e-01 -1.09642172e+00
6.30080462e-01 3.21741343e-01 7.96693027e-01 -5.96542239e-01
-2.34076649e-01 -1.45536408e-01 2.54542559e-01 -7.02637076e-01
2.22072139e-01 -3.66762042e-01 1.99926674e-01 5.30063689e-01
-1.57176256e-01 -4.20936406e-01 -7.29685053e-02 1.12617746e-01
1.29672253e+00 5.85427940e-01 2.17898920e-01 -1.61839142e-01
1.60684183e-01 7.17758983e-02 4.47106034e-01 9.46653247e-01
-6.02111556e-02 4.10871446e-01 5.79464436e-01 -6.29819930e-01
-1.50923884e+00 -1.48432922e+00 -2.02887192e-01 1.05336916e+00
5.98825932e-01 -3.56201649e-01 -3.84305507e-01 -2.23048881e-01
2.51895756e-01 6.70242488e-01 -1.02120948e+00 -4.93398190e-01
-3.42824697e-01 -6.64844573e-01 1.80356532e-01 6.35222793e-01
1.71457931e-01 -1.49719477e+00 -7.92504489e-01 1.14598699e-01
1.90268382e-01 -1.05227804e+00 1.33848712e-01 3.37370597e-02
-8.40507984e-01 -1.06572449e+00 1.28773212e-01 -2.24481687e-01
4.64064956e-01 1.00191914e-01 1.28642380e+00 5.75772047e-01
-3.62526953e-01 6.44320846e-01 -2.75619179e-01 -3.86246592e-01
-6.02297664e-01 -2.41943032e-01 3.41369987e-01 -1.42187253e-01
8.23381022e-02 -6.63255394e-01 -8.07099342e-01 5.40603518e-01
-6.05044425e-01 6.38712108e-01 4.19762135e-01 4.79095876e-01
5.31259716e-01 -4.44336608e-02 1.30807698e-01 -7.13497877e-01
-9.37248245e-02 -4.74165410e-01 -3.66927683e-01 5.59022248e-01
-4.56693679e-01 1.31156832e-01 7.66332448e-01 -7.97233760e-01
-1.20482814e+00 9.20666605e-02 3.28776330e-01 -1.89511016e-01
-6.52633190e-01 1.61327198e-01 1.23060443e-01 8.77940655e-03
6.92211211e-01 3.22186351e-01 -3.04046690e-01 -4.10083503e-01
2.08609909e-01 4.88679558e-02 4.04332101e-01 -1.13101912e+00
6.84044182e-01 5.63376009e-01 3.71833235e-01 -8.93742919e-01
-9.33396995e-01 -2.06233770e-01 -8.68517458e-01 -3.27033371e-01
8.65828812e-01 -5.23866415e-01 -1.12211668e+00 5.15659630e-01
-9.22684789e-01 -1.05921257e+00 -4.42428112e-01 6.08951092e-01
-9.13598478e-01 3.92898083e-01 -7.41156578e-01 -7.71560133e-01
3.00965518e-01 -8.40142250e-01 7.33996630e-01 2.40759909e-01
-4.79387820e-01 -1.29888153e+00 2.44278163e-01 2.34405726e-01
3.68870676e-01 1.94186807e-01 1.22506487e+00 -2.53750801e-01
-8.88318658e-01 2.30926469e-01 -3.34924698e-01 1.50563061e-01
1.86320499e-01 5.91176152e-01 -8.43373120e-01 -3.75555493e-02
-3.58305246e-01 -5.63012838e-01 7.17529118e-01 4.75076556e-01
1.38250327e+00 -1.89030617e-01 -5.57754755e-01 5.07335484e-01
1.35053897e+00 1.64657459e-01 7.25082934e-01 1.18870072e-01
5.86151063e-01 4.73305792e-01 2.98092186e-01 3.51723433e-01
1.08386941e-01 6.16901398e-01 3.64159942e-01 1.21793903e-01
-1.83840364e-01 -3.30117017e-01 4.73156810e-01 4.87334639e-01
-6.66495860e-01 -3.31220388e-01 -1.01422799e+00 3.87430906e-01
-1.88941777e+00 -1.14093733e+00 -2.35182598e-01 2.20231295e+00
9.51352000e-01 4.37518477e-01 -1.51881354e-03 -3.51720721e-01
4.08546180e-01 -5.70503362e-02 -1.04452884e+00 -3.05939436e-01
8.38970095e-02 2.31318727e-01 5.26684225e-01 5.56192577e-01
-7.50036418e-01 8.92350733e-01 7.39407635e+00 4.76807058e-01
-1.45484936e+00 -1.02001190e-01 8.15688014e-01 -1.08911254e-01
-2.53127992e-01 4.32745844e-01 -6.45445704e-01 4.73809510e-01
1.15822327e+00 2.35756598e-02 4.84081417e-01 6.40208364e-01
2.96291739e-01 -4.44217920e-01 -1.53397930e+00 4.75353152e-01
-4.14847851e-01 -1.49430716e+00 1.75406843e-01 1.13677464e-01
7.35578418e-01 1.17734179e-01 1.24106798e-02 1.02966093e-01
6.34021401e-01 -1.22339499e+00 9.00061607e-01 1.07659113e+00
6.26277804e-01 9.61026028e-02 1.58018678e-01 4.21282858e-01
-1.06749976e+00 2.19235018e-01 -5.86053789e-01 -5.94163835e-01
5.24125360e-02 4.05647755e-01 -5.06837130e-01 -7.82241598e-02
7.54960418e-01 7.24171877e-01 -5.74155509e-01 9.38964605e-01
-2.83584625e-01 1.04693723e+00 -5.11953235e-01 -1.58414677e-01
-1.38028860e-01 9.01451427e-03 1.74828216e-01 1.09284091e+00
9.62193161e-02 3.79990190e-01 9.90322158e-02 8.95048082e-01
1.69304192e-01 -1.82038352e-01 -4.13209289e-01 -4.80887480e-02
2.06579983e-01 8.72947216e-01 -7.19692886e-01 -4.57607538e-01
-5.74837029e-01 4.72058475e-01 4.08448279e-01 4.11984473e-01
-1.10566163e+00 4.93144155e-01 8.50316405e-01 5.16673565e-01
3.37840021e-01 -4.99412775e-01 -3.27868164e-01 -1.00379395e+00
-1.02395825e-01 -3.46972436e-01 -1.17899969e-01 -1.27371240e+00
-1.42327964e+00 -1.23953082e-01 2.29980469e-01 -8.21810663e-01
1.46762684e-01 -1.07024443e+00 -6.41977668e-01 4.74986285e-01
-1.14976859e+00 -9.93028164e-01 -3.17428321e-01 2.98881769e-01
-3.47536281e-02 2.69095272e-01 8.06178927e-01 -3.95300984e-01
-1.91153944e-01 3.65819409e-02 1.88753933e-01 -7.43179098e-02
6.27592802e-01 -1.17147791e+00 4.54705983e-01 2.32965976e-01
2.00603366e-01 5.34434080e-01 1.06316471e+00 -7.76749492e-01
-1.23388076e+00 -6.88147604e-01 4.33692068e-01 -9.76043105e-01
8.25045884e-01 -3.66601169e-01 -1.07179010e+00 8.46101999e-01
-1.12700105e-01 4.38472718e-01 5.29749572e-01 2.27143571e-01
-5.40985644e-01 8.41177180e-02 -6.77867889e-01 5.43742716e-01
1.34596813e+00 -5.65466344e-01 -4.19717818e-01 4.95995224e-01
5.49947083e-01 -1.76238000e-01 -9.64014232e-01 4.21568125e-01
9.36651587e-01 -1.21917677e+00 1.20801091e+00 -1.18128932e+00
8.00167561e-01 -1.96600229e-01 -1.50330871e-01 -1.04655099e+00
-6.91115081e-01 -1.60469905e-01 -3.83864284e-01 7.33506322e-01
5.76750696e-01 -6.54844522e-01 8.75950038e-01 1.19462883e+00
4.79225516e-02 -1.00813591e+00 -6.00060523e-01 -9.23577309e-01
3.02665114e-01 -4.43868399e-01 3.61741811e-01 9.54714656e-01
8.43271706e-03 1.60046101e-01 -3.85606021e-01 2.15853244e-01
2.89559811e-01 3.42606932e-01 8.48195672e-01 -1.22830653e+00
-7.26023674e-01 -2.67354012e-01 -6.04949951e-01 -1.19023347e+00
8.47111568e-02 -3.32200259e-01 -5.04740290e-02 -1.48400807e+00
5.54478228e-01 -7.39952028e-01 -4.85774785e-01 5.43050647e-01
-8.18221085e-03 1.06559590e-01 2.18020722e-01 4.43670183e-01
-7.99610734e-01 4.55170006e-01 1.35973048e+00 8.87101665e-02
-2.66544092e-02 -3.31027985e-01 -3.15204859e-01 9.06902492e-01
8.58885884e-01 -3.21923763e-01 -3.28745484e-01 -1.63303420e-01
5.36573827e-01 1.42709702e-01 1.07918203e+00 -1.32040894e+00
2.85432450e-02 -1.01233065e+00 7.24882126e-01 -5.12086339e-02
7.57556140e-01 -5.01068354e-01 3.88156652e-01 4.73648638e-01
-5.06044328e-01 -2.58027315e-01 2.07392782e-01 5.63439548e-01
3.05452913e-01 -6.27021631e-03 8.61030757e-01 -3.18230361e-01
-8.59362185e-01 3.34550768e-01 -4.29626495e-01 -5.80265410e-02
1.07999420e+00 -4.03706133e-01 -6.06284201e-01 -3.96601647e-01
-8.59652281e-01 -2.28501752e-01 9.14138556e-01 2.31543317e-01
2.59834290e-01 -8.50482583e-01 -5.12468517e-01 -1.96111605e-01
1.01468019e-01 -3.70295346e-01 2.42081374e-01 5.92754662e-01
-6.23751402e-01 3.27847183e-01 -3.79110992e-01 -5.83538294e-01
-9.89296198e-01 6.52296484e-01 5.59216797e-01 -6.53884560e-02
-5.73671460e-01 7.65998065e-01 7.13322401e-01 -3.02702069e-01
-4.06272471e-01 -5.72383583e-01 3.62587452e-01 -6.93410039e-01
1.14045560e-01 1.13824986e-01 -4.65468466e-01 -5.80061138e-01
-1.96409121e-01 7.45748222e-01 1.18334860e-01 2.17934921e-01
1.40340185e+00 1.32160053e-01 1.60797760e-01 8.71441841e-01
7.43305445e-01 -9.42278728e-02 -1.91883934e+00 -2.77697183e-02
-3.46540779e-01 -4.03936177e-01 -2.53687829e-01 -8.76489282e-01
-7.21980155e-01 8.47220600e-01 3.39170128e-01 3.98635566e-01
6.12620890e-01 4.28566039e-01 7.24738479e-01 5.67859650e-01
3.99411827e-01 -1.08901691e+00 3.53587717e-01 4.04948562e-01
5.81923783e-01 -1.16930997e+00 4.45592880e-01 -5.72981060e-01
-3.84031832e-01 9.13614392e-01 1.01672494e+00 -2.48321369e-01
9.13203657e-01 1.37908369e-01 -3.25354517e-01 -2.60419071e-01
-1.28176689e+00 1.58477306e-01 3.29822153e-01 4.22674000e-01
5.06118834e-01 3.29055548e-01 3.44945014e-01 -3.11411526e-02
-3.05708051e-01 -6.50304370e-03 3.76437902e-01 6.74965143e-01
-6.02163196e-01 -8.83692145e-01 -1.18312664e-01 7.51680613e-01
-1.01109415e-01 8.40407703e-03 -4.35872555e-01 7.84920514e-01
4.85080630e-01 5.47983527e-01 1.57385379e-01 -4.11735773e-01
-2.96566933e-02 5.67422211e-02 8.83106411e-01 -7.22172081e-01
-1.43895403e-01 -6.83452427e-01 1.76341385e-02 -5.35827875e-01
-6.43259764e-01 -9.06511724e-01 -1.40580142e+00 -6.30021274e-01
-1.92163423e-01 -4.05792966e-02 2.97961801e-01 1.14191961e+00
2.63639003e-01 2.74965316e-01 4.31514345e-02 -1.05548644e+00
-2.16195717e-01 -6.47234261e-01 -4.21636879e-01 8.80704761e-01
3.73192012e-01 -1.02319908e+00 -4.59530473e-01 6.07329488e-01] | [8.384525299072266, 0.8642739653587341] |
ad84a1c4-2505-47c5-aab5-9f8331e629b7 | effcrn-an-efficient-convolutional-recurrent | 2306.02778 | null | https://arxiv.org/abs/2306.02778v1 | https://arxiv.org/pdf/2306.02778v1.pdf | EffCRN: An Efficient Convolutional Recurrent Network for High-Performance Speech Enhancement | Fully convolutional recurrent neural networks (FCRNs) have shown state-of-the-art performance in single-channel speech enhancement. However, the number of parameters and the FLOPs/second of the original FCRN are restrictively high. A further important class of efficient networks is the CRUSE topology, serving as reference in our work. By applying a number of topological changes at once, we propose both an efficient FCRN (FCRN15), and a new family of efficient convolutional recurrent neural networks (EffCRN23, EffCRN23lite). We show that our FCRN15 (875K parameters) and EffCRN23lite (396K) outperform the already efficient CRUSE5 (85M) and CRUSE4 (7.2M) networks, respectively, w.r.t. PESQ, DNSMOS and DeltaSNR, while requiring about 94% less parameters and about 20% less #FLOPs/frame. Thereby, according to these metrics, the FCRN/EffCRN class of networks provides new best-in-class network topologies for speech enhancement. | ['Tim Fingscheidt', 'Wouter Tirry', 'Maximilian Strake', 'Kristoff Fluyt', 'Bruno Defraene', 'Jan Franzen', 'Marvin Sach'] | 2023-06-05 | null | null | null | null | ['speech-enhancement'] | ['speech'] | [ 4.80156578e-02 -6.01321645e-02 9.84107330e-02 3.75028178e-02
-3.56268018e-01 -3.19673479e-01 3.93095523e-01 -1.97267801e-01
-8.76725376e-01 8.06639135e-01 1.90739527e-01 -7.89235413e-01
-4.07780439e-01 -5.67719460e-01 -6.52092516e-01 -7.75644004e-01
-3.36851776e-01 -2.53195792e-01 4.54308599e-01 -6.97780132e-01
-1.95598342e-02 7.96617091e-01 -1.48188710e+00 1.58075467e-01
6.53985202e-01 1.08384907e+00 4.09728765e-01 9.78652716e-01
1.36313155e-01 1.05320823e+00 -8.48592281e-01 -6.34179473e-01
-1.88273489e-01 -1.77695438e-01 -6.37401998e-01 -6.97157085e-01
-1.97655246e-01 -3.49016666e-01 -8.12919557e-01 7.68851280e-01
1.10960484e+00 2.89677441e-01 3.93506080e-01 -7.43705451e-01
-1.61465108e-01 1.02603018e+00 -3.51495475e-01 4.96231258e-01
-2.58672357e-01 -1.61802799e-01 5.80165744e-01 -7.19055951e-01
4.54351813e-01 1.12444413e+00 1.04337955e+00 4.35787857e-01
-6.13378763e-01 -9.18816626e-01 -1.12945490e-01 3.09786767e-01
-1.29610062e+00 -9.57827330e-01 5.54768562e-01 4.50374067e-01
1.36057794e+00 3.46878916e-01 4.35735732e-01 1.06896579e+00
-6.29420653e-02 5.38556278e-01 1.01473296e+00 -5.03248215e-01
2.20676303e-01 -1.27123356e-01 5.73994666e-02 4.37805355e-01
8.15760940e-02 3.00243884e-01 -3.55799228e-01 1.88517779e-01
1.06086075e+00 -2.98884779e-01 -5.18486798e-01 4.05336112e-01
-8.14281702e-01 6.11289918e-01 5.26877046e-01 6.50082529e-01
-2.47004539e-01 5.20806372e-01 7.76208937e-01 6.49021566e-01
2.33725950e-01 3.35218608e-01 -5.87420166e-01 -7.57748187e-01
-9.59705830e-01 -1.71950936e-01 5.85810721e-01 1.09848154e+00
4.07987535e-01 5.53906679e-01 -5.20509370e-02 1.16800439e+00
-2.05541812e-02 7.06168175e-01 4.30805176e-01 -1.00944686e+00
5.55754483e-01 3.51443440e-02 -1.81545302e-01 -7.09644377e-01
-7.02488780e-01 -8.74539435e-01 -1.47864223e+00 -2.12098937e-02
1.64807171e-01 -2.92113543e-01 -8.03713024e-01 1.50385523e+00
-3.64271820e-01 -6.84585124e-02 3.17628145e-01 7.22186029e-01
8.49395037e-01 9.51271892e-01 -2.25843847e-01 -4.08511311e-01
1.24273336e+00 -1.12858307e+00 -9.39689934e-01 1.84542909e-01
6.03677273e-01 -9.96960580e-01 6.91293836e-01 4.17047620e-01
-1.49909055e+00 -6.38498604e-01 -1.04550290e+00 1.39331311e-01
-3.41441572e-01 3.45658004e-01 4.27392066e-01 1.03240144e+00
-1.79188240e+00 8.99158835e-01 -6.10956430e-01 9.58499014e-02
1.39094844e-01 6.33405149e-01 -1.01707660e-01 2.30030373e-01
-1.36203527e+00 7.36746252e-01 3.27427089e-01 3.66199404e-01
-6.37380362e-01 -7.75383770e-01 -4.98983532e-01 4.24427092e-01
2.01542974e-01 -2.35470161e-01 1.20941281e+00 -6.19627714e-01
-2.06090045e+00 1.71455488e-01 2.70839512e-01 -9.34825957e-01
3.81848812e-01 -1.22922435e-01 -8.83246601e-01 4.59445566e-01
-8.05387497e-01 5.90993464e-01 7.77631223e-01 -8.41845512e-01
-4.99889880e-01 1.53194323e-01 1.62377208e-01 5.05088195e-02
-5.49399912e-01 5.52255392e-01 -4.10357356e-01 -8.71886015e-01
4.73758811e-03 -6.64251685e-01 -2.46010646e-01 -2.30347112e-01
-2.74788678e-01 7.82076344e-02 1.13151371e+00 -7.91289449e-01
1.38883817e+00 -2.18505669e+00 -5.12083806e-02 1.45944908e-01
3.02828103e-01 1.08366776e+00 -2.18615755e-01 2.30977237e-02
-2.12537855e-01 3.69551688e-01 7.65939197e-03 -4.36975896e-01
-3.94210331e-02 1.08484864e-01 4.11488377e-02 2.64550090e-01
-1.67478025e-02 8.69344890e-01 -5.85615814e-01 -2.03390181e-01
3.29540581e-01 1.12134922e+00 -5.84783256e-01 -6.05553351e-02
3.70966673e-01 -1.23805419e-01 -5.18717729e-02 3.37033898e-01
9.12441552e-01 1.19203232e-01 1.07705288e-01 -5.09581983e-01
-4.77098584e-01 3.90036434e-01 -1.04392970e+00 1.36546075e+00
-9.36995745e-01 1.02768362e+00 1.95665434e-01 -8.78575206e-01
1.07063627e+00 6.50538445e-01 2.87402719e-01 -1.11030304e+00
2.92480469e-01 6.16903782e-01 6.19355813e-02 -1.33435532e-01
8.55587900e-01 1.44298986e-01 1.85990810e-01 3.11091006e-01
3.75012338e-01 3.79143953e-01 2.61757553e-01 1.41536102e-01
1.32016802e+00 -4.05737668e-01 -4.77158204e-02 -3.63438368e-01
5.01823545e-01 -1.02143657e+00 4.25548941e-01 8.52731168e-01
-4.12560850e-01 4.79745597e-01 5.94255030e-01 -2.21082777e-01
-1.62230814e+00 -1.00801468e+00 -5.33832908e-02 8.93363893e-01
-2.79754966e-01 -2.93762535e-01 -8.13951790e-01 -1.38363123e-01
-9.11133885e-01 2.75735170e-01 2.20104642e-02 -1.03235595e-01
-1.03214967e+00 -7.09260345e-01 1.54928315e+00 6.66356742e-01
1.03643334e+00 -1.23614264e+00 -8.40161681e-01 4.25010711e-01
-1.85418829e-01 -1.37656724e+00 -3.07242185e-01 4.13407177e-01
-8.80345881e-01 -4.83446270e-01 -1.27783895e+00 -6.94395065e-01
2.51671761e-01 -1.26878899e-02 1.12188268e+00 1.38545081e-01
8.77046958e-02 -8.94072726e-02 -7.98132837e-01 1.97493643e-01
-6.33061767e-01 3.46589684e-01 -2.34631263e-02 -2.80433178e-01
-3.31832349e-01 -7.91633666e-01 -6.01387680e-01 2.88785458e-01
-7.61547387e-01 -1.09263450e-01 7.74600863e-01 8.09126973e-01
2.30258137e-01 3.08828503e-01 6.94067895e-01 -3.51592153e-01
4.59119290e-01 1.91414833e-01 -5.25379717e-01 9.53089297e-02
-4.03110772e-01 1.08378902e-01 9.99551594e-01 -4.74572569e-01
-1.14962602e+00 -3.91676128e-01 -8.40350688e-01 -5.57875395e-01
6.13223203e-02 1.92811638e-01 4.85306280e-03 -2.01079771e-01
3.80247891e-01 1.35309041e-01 -1.05232358e-01 -5.65898061e-01
2.27358833e-01 7.88594544e-01 5.60189724e-01 -2.08926395e-01
4.05292332e-01 7.79959112e-02 -2.85718292e-01 -1.24559891e+00
3.13155055e-02 -3.37519854e-01 -1.72513425e-01 -2.09162027e-01
6.30636871e-01 -8.27931583e-01 -9.90008414e-01 9.36533391e-01
-1.32514751e+00 -4.71613944e-01 -4.03747499e-01 5.57033777e-01
-4.30525810e-01 2.95638978e-01 -1.10008097e+00 -1.08236301e+00
-9.27003145e-01 -1.23908925e+00 6.22379422e-01 2.01395512e-01
3.64660293e-01 -8.09082687e-01 -5.42436957e-01 -6.03382587e-02
1.14190769e+00 -1.67757869e-01 7.67742693e-01 -1.79307953e-01
-2.62169868e-01 2.33631656e-01 -6.59168959e-01 6.64564073e-01
-2.19596758e-01 -4.95834686e-02 -1.02235866e+00 -5.05765259e-01
9.66143981e-02 2.51430180e-02 1.11676741e+00 7.77779460e-01
1.03111911e+00 -3.28621924e-01 8.48829672e-02 8.80306363e-01
1.40060914e+00 6.22116923e-01 1.33085287e+00 9.22767818e-02
2.62096733e-01 -3.10641602e-02 -3.85339488e-03 5.04738569e-01
-1.31932944e-01 7.13180423e-01 3.71418148e-01 -4.30369139e-01
-4.28051174e-01 2.08439454e-01 5.98246336e-01 1.59617186e+00
-5.60453415e-01 -5.43263733e-01 -5.12713432e-01 3.30241352e-01
-1.43660045e+00 -9.59539592e-01 6.90286933e-03 1.93632054e+00
6.64138615e-01 4.17641252e-01 3.09278071e-02 8.17695498e-01
9.57937658e-01 3.80626321e-01 -3.06470126e-01 -9.66639519e-01
-5.35377860e-01 7.74812639e-01 7.53023744e-01 4.23140228e-01
-8.23508501e-01 7.06895292e-01 5.83122540e+00 1.30905426e+00
-1.08565497e+00 2.25173816e-01 6.82273567e-01 -1.65360287e-01
7.94592779e-03 -4.92712080e-01 -7.85919428e-01 3.24600190e-01
1.77296853e+00 3.57102841e-01 6.61170840e-01 3.39162201e-01
2.03234583e-01 3.74516010e-01 -2.20936224e-01 1.01366782e+00
-1.92150429e-01 -1.57852900e+00 -1.76024795e-01 -6.16705641e-02
4.75712627e-01 2.30071500e-01 1.68452144e-01 3.18659723e-01
-1.95299666e-02 -9.43782091e-01 6.11303568e-01 3.35555971e-01
1.20998657e+00 -1.36716974e+00 8.86619031e-01 -1.24051824e-01
-1.57322037e+00 -2.82744169e-01 -5.96081495e-01 2.63052166e-01
2.05161273e-01 9.35429931e-01 -2.00892523e-01 6.55300021e-01
1.01507795e+00 3.70079726e-01 -2.35024765e-01 8.63736093e-01
6.29766360e-02 6.38506889e-01 -5.52343071e-01 -3.25715691e-01
5.12213051e-01 2.08129689e-01 4.92634922e-01 1.57153368e+00
4.42808539e-01 4.71344329e-02 -7.43213773e-01 4.83737290e-01
-3.92827481e-01 -1.94281176e-01 -1.69179961e-02 1.68091953e-01
6.29487514e-01 1.17917097e+00 -6.96920335e-01 -2.08292395e-01
7.98314884e-02 8.41565967e-01 9.70515385e-02 3.92351627e-01
-1.08064675e+00 -1.04187584e+00 6.52869105e-01 -1.94708586e-01
6.00587249e-01 -8.35177079e-02 1.97439358e-01 -7.86527097e-01
1.46709131e-02 -8.85562778e-01 -7.30969682e-02 -7.44632840e-01
-4.09983426e-01 1.18560255e+00 -4.43584025e-01 -9.85862017e-01
5.50404005e-02 -7.49643683e-01 -3.86841327e-01 6.35097086e-01
-1.89198017e+00 -6.05937123e-01 -9.89970863e-02 6.19011045e-01
4.34910774e-01 -3.23421746e-01 7.66666651e-01 9.28931177e-01
-5.42528629e-01 1.27424812e+00 4.64068711e-01 2.11704567e-01
1.78209573e-01 -8.17152858e-01 6.38819039e-01 7.76787937e-01
-2.94919699e-01 4.49026197e-01 5.39390862e-01 4.13922071e-02
-1.23225570e+00 -1.09283721e+00 8.49621654e-01 6.75370276e-01
4.37993526e-01 -2.72696614e-01 -7.00120687e-01 9.89233702e-02
3.83486062e-01 -2.28812210e-02 4.67163101e-02 -5.22989035e-02
-3.04456234e-01 -1.50933221e-01 -1.07577443e+00 8.40100706e-01
1.18821239e+00 -5.91177404e-01 2.20071077e-01 -8.31615999e-02
1.36232162e+00 -5.15497506e-01 -1.12553322e+00 4.04865503e-01
4.64668214e-01 -1.35479259e+00 1.08970809e+00 6.85370043e-02
3.28093916e-01 -4.07862961e-02 -1.51477188e-01 -1.15676606e+00
-8.67869854e-02 -1.19153190e+00 -1.99205339e-01 1.02535617e+00
6.41026080e-01 -7.32630432e-01 7.47068346e-01 -2.50958532e-01
-7.58486152e-01 -8.72954130e-01 -1.44417250e+00 -1.16397262e+00
1.06851190e-01 -5.72666645e-01 5.49154103e-01 4.31685686e-01
-1.94271490e-01 -1.48123400e-02 -4.93245304e-01 -1.24240018e-01
1.74008608e-01 -7.31531918e-01 4.74553891e-02 -7.66102195e-01
-1.52085334e-01 -7.58629978e-01 -4.18466777e-01 -1.16955125e+00
-7.52438307e-02 -3.26984882e-01 -2.53755778e-01 -1.13613594e+00
-1.99754477e-01 -7.02221274e-01 -4.51417595e-01 2.62219638e-01
4.25286293e-01 3.98377091e-01 3.49801987e-01 -2.17055187e-01
-4.36508566e-01 4.22021568e-01 1.08593726e+00 3.31594683e-02
-2.85383493e-01 -2.37594265e-02 -2.56726623e-01 6.26481771e-01
9.71080065e-01 -1.02568157e-01 -4.12945002e-01 -3.42683524e-01
1.95950329e-01 2.34130621e-01 2.96577752e-01 -1.43253040e+00
4.63388085e-01 6.72722340e-01 6.32315991e-04 -7.04275370e-01
8.62380862e-01 -6.73783600e-01 3.25418003e-02 6.57740831e-01
-1.78650483e-01 4.48386967e-01 3.74873996e-01 -1.18573736e-02
-1.98174000e-01 -2.19564527e-01 1.08211803e+00 2.11767823e-01
-5.65550447e-01 6.54257908e-02 -6.58567011e-01 4.49505113e-02
3.67366523e-01 -2.71817148e-01 -6.95376575e-01 -4.70944852e-01
-6.59856915e-01 -4.11892682e-01 -2.19515651e-01 1.15680858e-01
9.41182852e-01 -1.06979096e+00 -6.15596414e-01 1.83046311e-01
-6.01923823e-01 -3.19062144e-01 7.28014350e-01 9.50249195e-01
-7.29189396e-01 6.56736016e-01 -3.43251824e-01 -2.61456817e-01
-1.16454124e+00 2.00273082e-01 7.96730340e-01 -5.96286833e-01
-8.35686505e-01 9.25646305e-01 -4.77637887e-01 -3.55544508e-01
3.77343327e-01 -4.54278618e-01 -1.51456863e-01 -2.30089780e-02
5.36894262e-01 9.66695428e-01 4.21137899e-01 -4.95043337e-01
-1.11266315e-01 2.80204773e-01 -9.41439867e-02 -2.02014551e-01
1.36652601e+00 3.76165807e-02 -5.52867278e-02 -2.28740633e-01
1.50678968e+00 -3.38035941e-01 -1.10158169e+00 -8.75430778e-02
-2.41607100e-01 1.60699800e-01 3.87349755e-01 -7.88058043e-01
-1.57308102e+00 9.60575402e-01 1.01321089e+00 1.26062512e-01
1.72128999e+00 -3.43793124e-01 1.05159843e+00 3.11302602e-01
2.69494087e-01 -1.25973916e+00 6.93269670e-02 9.73221898e-01
7.49381244e-01 -4.15508986e-01 -1.65834680e-01 -1.83043376e-01
-1.99044496e-01 1.15195072e+00 1.02180742e-01 1.22060142e-01
7.96583712e-01 8.31615031e-01 -8.95582065e-02 1.15067020e-01
-7.04564571e-01 -3.42171043e-02 -3.23109746e-01 4.77538735e-01
4.48858827e-01 -4.55567576e-02 -2.60363013e-01 2.59137750e-01
-3.94335270e-01 -2.64881462e-01 6.80261910e-01 5.80401719e-01
-4.48198259e-01 -9.35046554e-01 -1.52603552e-01 3.42542827e-01
-5.23834109e-01 -4.47444111e-01 2.82989770e-01 8.29045177e-01
-1.05977274e-01 1.08104944e+00 2.08936647e-01 -6.11160934e-01
4.94854093e-01 -4.63652432e-01 8.79618227e-02 3.64414334e-01
-1.04436183e+00 2.09061220e-01 3.89933258e-01 -3.54675770e-01
-5.07950902e-01 -2.70192862e-01 -1.25192893e+00 -9.17092025e-01
-7.42823362e-01 1.76413760e-01 9.13554311e-01 7.49700308e-01
3.98056269e-01 1.17833602e+00 6.83179259e-01 -6.66018784e-01
-3.01574707e-01 -9.41138268e-01 -6.73545301e-01 -2.24052131e-01
4.41330492e-01 -3.04888129e-01 -3.36975098e-01 -4.38853025e-01] | [14.946374893188477, 5.9444580078125] |
51b92982-d843-4de2-be8a-c250198e5270 | physics-driven-fire-modeling-from-multi-view | 1804.05261 | null | http://arxiv.org/abs/1804.05261v1 | http://arxiv.org/pdf/1804.05261v1.pdf | Physics-driven Fire Modeling from Multi-view Images | Fire effects are widely used in various computer graphics applications such
as visual effects and video games. Modeling the shape and appearance of fire
phenomenon is challenging as the underlying effects are driven by complex laws
of physics. State-of-the-art fire modeling techniques rely on sophisticated
physical simulations which require intensive parameter tuning, or use
simplifications which produce physically invalid results. In this paper, we
present a novel method of reconstructing physically valid fire models from
multi-view stereo images. Our method, for the first time, provides plausible
estimation of physical properties (e.g., temperature, density) of a fire volume
using RGB cameras. This allows for a number of novel phenomena such as global
fire illumination effects. The effectiveness and usefulness of our method are
tested by generating fire models from a variety of input data, and applying the
reconstructed fire models for realistic illumination of virtual scenes. | ['Yong-Liang Yang', 'Garoe Dorta', 'Luca Benedetti', 'Dmitry Kit'] | 2018-04-14 | null | null | null | null | ['physical-simulations'] | ['miscellaneous'] | [ 3.08135778e-01 -7.63580322e-01 5.07125199e-01 1.55045643e-01
1.32191852e-01 -7.02573597e-01 8.42427850e-01 -1.64360538e-01
-2.26656243e-01 9.68843400e-01 -2.90026724e-01 -2.05562115e-01
-1.59383267e-02 -1.23702371e+00 -5.86087227e-01 -8.99021149e-01
8.41174871e-02 3.08675319e-01 9.01487172e-01 -3.01293075e-01
3.32318425e-01 9.40285504e-01 -2.03852177e+00 9.37981531e-02
6.23257399e-01 6.82903647e-01 2.27432638e-01 8.94140065e-01
-3.16146798e-02 4.44885075e-01 -4.27100211e-01 4.43441905e-02
2.09814370e-01 -4.92020100e-01 -3.77081275e-01 4.50406559e-02
3.69259752e-02 -4.52974826e-01 -2.74141103e-01 6.52637541e-01
1.99512392e-01 4.29125905e-01 9.32303667e-01 -1.17858350e+00
-1.96379274e-01 -3.23449492e-01 -8.20213854e-01 1.39733374e-01
3.76838028e-01 4.20837313e-01 3.24329972e-01 -6.17959738e-01
7.53000259e-01 1.21987164e+00 5.95388949e-01 6.06750309e-01
-1.38143432e+00 -6.43203199e-01 -4.75783721e-02 -9.56369862e-02
-1.63716745e+00 -3.28229219e-01 7.65487492e-01 -4.00749028e-01
1.12733459e+00 7.13312864e-01 1.10547829e+00 9.73105490e-01
5.94211936e-01 4.00773995e-02 1.55507481e+00 -3.05415362e-01
6.07141137e-01 1.01340553e-02 -2.31549308e-01 6.42302096e-01
2.43428811e-01 6.20669782e-01 -3.65003228e-01 -5.62482774e-01
1.46443093e+00 -6.86492473e-02 -4.90135282e-01 -3.08128506e-01
-9.00957346e-01 5.01489103e-01 2.39363149e-01 -1.50345460e-01
-3.54447395e-01 4.86151010e-01 4.63622535e-04 -3.36529374e-01
6.64483607e-01 -8.12740717e-03 -3.94355804e-01 -1.93527937e-01
-8.50743055e-01 7.20128000e-01 7.10825741e-01 5.70503712e-01
8.15695941e-01 6.50018379e-02 2.85246074e-01 6.64345324e-01
6.03509188e-01 1.13313258e+00 -2.37478957e-01 -1.11659062e+00
-2.13133425e-01 1.92846686e-01 4.32908922e-01 -6.36590779e-01
-2.63306737e-01 6.77001029e-02 -8.52882028e-01 1.09100795e+00
3.27214062e-01 2.66242791e-02 -1.16819179e+00 1.33403671e+00
5.98213196e-01 5.07860780e-01 -2.19978347e-01 9.19192731e-01
6.88840926e-01 8.51671576e-01 -3.77069972e-02 -3.37591767e-01
9.62536037e-01 -2.91834503e-01 -4.60682005e-01 9.15218070e-02
-1.11664765e-01 -7.73412526e-01 9.75798011e-01 4.70769495e-01
-1.20483017e+00 -2.32816413e-01 -6.39700174e-01 3.96405876e-01
-2.46310383e-01 -6.42780781e-01 7.07119465e-01 6.71449602e-01
-9.69403803e-01 7.13179648e-01 -9.32057440e-01 -4.72246349e-01
5.85825108e-02 2.57480353e-01 3.98071967e-02 1.57986403e-01
-9.59595323e-01 7.69940019e-01 -1.03251673e-01 -1.29403798e-02
-8.69723320e-01 -8.39634478e-01 -5.16960859e-01 -4.48979512e-02
1.68236062e-01 -1.06518054e+00 1.12648499e+00 -6.61209047e-01
-1.63836944e+00 6.65435433e-01 -1.74357727e-01 2.05451742e-01
5.25648475e-01 6.08595274e-02 -3.10994029e-01 -3.04301698e-02
-4.30749416e-01 3.24941188e-01 5.89401960e-01 -1.98685491e+00
-1.86358362e-01 -2.25348085e-01 2.49645486e-01 2.54872262e-01
3.98596764e-01 2.99138408e-02 -9.98632908e-02 -2.76890635e-01
-2.10385267e-02 -9.55567896e-01 -3.90341073e-01 4.57757980e-01
-1.22558385e-01 5.60368299e-01 9.88570273e-01 -2.60542423e-01
9.37971294e-01 -1.82046545e+00 1.02065325e-01 3.42858166e-01
-5.41017056e-02 2.37931423e-02 3.88468385e-01 5.72206080e-01
3.45281601e-01 3.96584868e-01 -5.34822166e-01 -9.26727802e-02
-3.42474669e-01 3.53384048e-01 -3.44873130e-01 5.13400137e-01
-9.79649574e-02 3.11306626e-01 -5.89686811e-01 -3.53506833e-01
7.21886873e-01 1.16837931e+00 -6.19915187e-01 1.56471893e-01
-4.33294743e-01 8.14194441e-01 -3.97935957e-01 2.37769067e-01
1.01203907e+00 1.15962736e-01 -2.22054586e-01 6.90202462e-03
-6.58240855e-01 1.13402791e-02 -1.32639492e+00 1.27742529e+00
-6.71035230e-01 1.93618119e-01 2.87079215e-01 1.30367577e-02
7.69221902e-01 3.07978272e-01 4.55507070e-01 -4.88367558e-01
2.62748450e-01 -2.75268480e-02 -1.89154491e-01 -3.33135515e-01
4.17715520e-01 -8.02250803e-01 3.48060578e-01 7.43154764e-01
-4.77451056e-01 -8.44061673e-01 -1.16789646e-01 -9.82384458e-02
7.98425853e-01 5.86999834e-01 1.73427477e-01 -3.89111787e-01
3.16287786e-01 2.01494992e-01 3.38756680e-01 6.12148046e-01
8.55365619e-02 8.15195799e-01 -1.23162813e-01 -4.72429961e-01
-1.20929480e+00 -1.22628021e+00 -3.20515275e-01 3.95129949e-01
5.75324774e-01 -2.32241109e-01 -6.81325495e-01 1.00387864e-01
-7.01652328e-03 7.76929677e-01 -5.93514740e-01 -1.68347940e-01
-5.14558971e-01 -9.72169876e-01 2.96581626e-01 2.40896836e-01
5.11313081e-01 -1.27387774e+00 -1.29419482e+00 -1.32971397e-02
4.32768129e-02 -9.31636035e-01 1.84340000e-01 -1.94823265e-01
-1.10127914e+00 -1.12003052e+00 -5.24073064e-01 2.13832155e-01
3.93563658e-01 4.83303189e-01 1.28308189e+00 6.42069638e-01
-7.48720706e-01 5.25313795e-01 -2.44179070e-01 -3.85411918e-01
-5.22401750e-01 -5.86259604e-01 -1.63257681e-02 -1.94749832e-01
-1.92079440e-01 -8.20531130e-01 -7.60837913e-01 4.15992290e-01
-1.21029770e+00 5.72850764e-01 -1.36434525e-01 2.93870628e-01
6.47245824e-01 3.10265899e-01 -4.04051870e-01 -8.32309723e-01
4.97436166e-01 -2.19906315e-01 -7.57573068e-01 2.25142762e-01
-2.25212350e-01 -8.02292824e-02 6.29732549e-01 -3.09445858e-01
-1.60027730e+00 -1.03014857e-01 1.62107218e-02 -3.74823093e-01
-5.29549658e-01 1.39074445e-01 1.11849867e-02 -2.53269225e-01
7.56785452e-01 1.46453872e-01 -5.21518648e-01 -3.32291782e-01
7.80214965e-02 3.32412243e-01 2.51097471e-01 -9.23449159e-01
7.40969241e-01 1.10056686e+00 5.06309807e-01 -1.23663461e+00
-1.41342416e-01 -1.01796858e-01 -6.88120842e-01 -8.07335913e-01
5.27820230e-01 -3.44935864e-01 -9.15138304e-01 7.89990842e-01
-1.21137166e+00 -6.03515565e-01 -1.10858031e-01 4.38963622e-01
-6.39928401e-01 1.57982841e-01 -3.83714378e-01 -1.39412224e+00
-9.11306441e-02 -1.04673779e+00 1.13939679e+00 4.14481521e-01
-4.98348624e-02 -1.12662315e+00 6.54708743e-01 -7.57417679e-02
4.86184716e-01 6.09869242e-01 7.76715100e-01 9.44880068e-01
-7.89289415e-01 3.47996116e-01 -3.32147218e-02 -3.67123634e-01
2.89808780e-01 1.04672682e+00 -1.34273994e+00 -9.16846916e-02
4.75086011e-02 8.85747373e-02 8.63359094e-01 7.67167509e-01
1.08702815e+00 3.24522913e-01 -4.91676331e-01 8.95239651e-01
1.95214832e+00 2.73054600e-01 9.06051099e-01 1.13792703e-01
7.05421507e-01 3.41012508e-01 3.27287853e-01 1.00750685e+00
-1.16322666e-01 8.45145643e-01 7.41876960e-01 -1.90983400e-01
-2.38406315e-01 1.23924771e-02 1.36342850e-02 4.24939662e-01
-1.08560646e+00 -4.09697801e-01 -9.23173010e-01 1.07679814e-01
-1.66456521e+00 -1.22717583e+00 -5.41830599e-01 2.65205956e+00
4.58918303e-01 -1.50089571e-02 1.00550219e-01 1.38874471e-01
6.41414285e-01 2.85153016e-02 -5.42613089e-01 -5.42080104e-01
-2.15889290e-02 5.37595689e-01 5.94431579e-01 6.90062582e-01
-7.79284537e-01 7.90273905e-01 7.21412659e+00 2.99291193e-01
-1.27693868e+00 1.02495477e-01 3.19530189e-01 -1.80141196e-01
-7.55032122e-01 3.77355188e-01 -4.43625003e-01 7.98611119e-02
7.05475390e-01 -6.53569624e-02 8.78912330e-01 1.37918532e-01
7.05503225e-01 -7.73954391e-01 -3.92979890e-01 8.93505335e-01
-1.69968843e-01 -1.15985286e+00 7.62059018e-02 1.73961997e-01
6.33795619e-01 -1.27356291e-01 -2.13797629e-01 -3.47666234e-01
5.31391203e-01 -9.88496423e-01 7.15025485e-01 8.11402261e-01
8.48524630e-01 -8.18695128e-01 1.46706358e-01 4.70228970e-01
-1.25725520e+00 3.78299803e-01 -4.56552774e-01 -3.15981954e-01
5.87486625e-01 7.31975377e-01 -1.62818298e-01 4.36412305e-01
8.10109079e-01 3.20277095e-01 -2.60663092e-01 1.29163563e+00
-6.11008033e-02 4.88742232e-01 -7.38001764e-01 1.47418842e-01
-7.75246322e-02 -5.64210892e-01 4.33976620e-01 9.89929914e-01
3.17919672e-01 4.36324894e-01 -2.01928660e-01 1.27454972e+00
4.41264808e-01 -6.54859766e-02 -9.32470381e-01 2.93449312e-01
1.24465898e-01 1.23016024e+00 -1.14622235e+00 2.53431741e-02
-1.74338907e-01 8.09888840e-01 -2.78035104e-02 4.98705864e-01
-1.15799141e+00 2.98571914e-01 7.74332821e-01 5.84174097e-01
-5.24794497e-03 -5.20933092e-01 -3.23361993e-01 -1.15529072e+00
-2.89544702e-01 -2.12671801e-01 -2.11343095e-01 -1.26036322e+00
-9.46634889e-01 4.32596803e-01 3.53226066e-01 -9.98458147e-01
1.02683261e-01 -6.59741402e-01 -8.34727228e-01 9.07519937e-01
-1.27541339e+00 -1.09731150e+00 -6.61613643e-01 7.86285400e-01
3.17266673e-01 4.38629061e-01 1.15219164e+00 -5.37977964e-02
-3.16566586e-01 -2.71821111e-01 2.18158543e-01 -7.75032043e-01
2.99299121e-01 -8.33027303e-01 4.50842679e-01 5.40800333e-01
-1.43090144e-01 2.84295946e-01 1.10175824e+00 -8.00945461e-01
-1.35907590e+00 -7.03236938e-01 1.30727872e-01 -2.05141678e-01
2.88688183e-01 -3.42190355e-01 -8.19058359e-01 1.38470814e-01
1.09040685e-01 1.15622237e-01 4.76590276e-01 -9.42089260e-02
-1.44164458e-01 2.61353880e-01 -1.47950661e+00 6.60559595e-01
1.17731047e+00 -4.35450286e-01 5.80850281e-02 6.64931163e-02
-2.20699143e-02 -3.69907022e-01 -5.06008923e-01 3.60070348e-01
8.94341528e-01 -1.39107752e+00 1.07358098e+00 -3.74786466e-01
3.91531736e-01 -5.09143114e-01 -4.69950698e-02 -1.36085284e+00
-3.16021353e-01 -3.58887076e-01 -1.14955492e-02 9.36440885e-01
3.68481502e-02 -4.14711654e-01 3.88307840e-01 1.00056911e+00
2.34356344e-01 -3.45529467e-01 -7.59018660e-01 -5.96426427e-01
-7.87557289e-02 -4.52389240e-01 5.21526217e-01 8.16596270e-01
-3.35565537e-01 -2.35003054e-01 -2.15348542e-01 3.14752519e-01
8.48706663e-01 1.79554299e-01 5.88220954e-01 -1.53313625e+00
-4.27105904e-01 -2.38273442e-01 -8.77475664e-02 -3.61142039e-01
-5.61578851e-03 -2.68331885e-01 6.52857572e-02 -1.56672084e+00
4.89987105e-01 -7.45303392e-01 6.34680763e-02 -7.06205145e-02
3.49357761e-02 2.91968852e-01 1.09252453e-01 2.03292191e-01
2.19284818e-01 5.86731195e-01 1.44794309e+00 2.01780081e-01
-2.06190929e-01 2.74654906e-02 7.29432777e-02 9.61829782e-01
8.99444044e-01 -5.31532288e-01 -6.46145105e-01 -3.15565616e-01
3.82239819e-01 7.54118040e-02 9.16753531e-01 -9.46375072e-01
-1.92928419e-01 -7.25008190e-01 4.16493297e-01 -3.54017258e-01
8.82305086e-01 -9.98484612e-01 8.10811460e-01 3.39220256e-01
3.32674533e-01 1.82986721e-01 6.46787405e-01 3.78950626e-01
2.51971364e-01 -2.98219979e-01 9.45543230e-01 -5.84173143e-01
-3.35121721e-01 2.27350846e-01 -5.94174862e-01 -7.32766166e-02
9.46148932e-01 -3.25229794e-01 -4.54192340e-01 -3.78033370e-01
-4.79306728e-01 -5.71051657e-01 1.11111033e+00 2.77509801e-02
6.69475853e-01 -1.18362439e+00 -4.51747507e-01 2.33639106e-01
-3.65275472e-01 -1.27200782e-03 2.57565707e-01 3.75271708e-01
-1.16919887e+00 -1.81168299e-02 -4.00487721e-01 -5.94726503e-01
-1.47248447e+00 4.52857524e-01 5.72375596e-01 5.76569177e-02
-7.22032785e-01 4.59914386e-01 7.87478030e-01 -4.20953453e-01
-5.25454760e-01 -3.03114504e-01 2.07295462e-01 -6.85080945e-01
3.87345046e-01 4.29860026e-01 -1.31324291e-01 -7.62986243e-01
-4.67621207e-01 1.00965571e+00 5.66829920e-01 -6.13258302e-01
1.36804283e+00 3.52771371e-03 -6.76907077e-02 6.72232032e-01
2.79638290e-01 -1.00369282e-01 -1.57692909e+00 4.08497155e-01
-8.13336134e-01 -9.21917140e-01 4.16321486e-01 -7.65712559e-01
-1.09754550e+00 9.86844778e-01 4.86611545e-01 7.66157210e-02
1.18290901e+00 -2.08779901e-01 4.74383831e-01 4.72194701e-03
7.79393613e-01 -8.86423528e-01 -5.83608985e-01 3.98533821e-01
8.33868027e-01 -7.32032955e-01 2.22104728e-01 -9.32337999e-01
-1.38475120e-01 1.08948898e+00 6.18489802e-01 -2.19525829e-01
1.03110540e+00 8.96339655e-01 -5.90194762e-03 -3.01050007e-01
-7.86899924e-01 -2.40789130e-02 -2.49633208e-01 5.11048317e-01
4.21284556e-01 1.02561623e-01 -1.51699409e-01 -2.46644318e-01
1.84513316e-01 2.35280335e-01 7.10762620e-01 1.15879011e+00
-5.01866281e-01 -9.18589354e-01 -9.06780422e-01 1.70153365e-01
-8.43906403e-02 -7.06167147e-02 -5.40782988e-01 7.89267302e-01
1.48239672e-01 7.66410828e-01 1.24523409e-01 -1.60406649e-01
3.40711087e-01 -1.21887706e-01 1.01551497e+00 -3.03046346e-01
-6.25387430e-01 3.33507568e-01 -1.41762704e-01 -3.00136447e-01
-8.02978456e-01 -7.95224428e-01 -1.38993120e+00 -7.50564218e-01
-5.06339967e-01 -3.79030794e-01 8.24430466e-01 8.81684303e-01
-2.47564036e-02 4.91321683e-01 3.70559186e-01 -1.28154552e+00
2.46326223e-01 -7.03645527e-01 -9.31731045e-01 5.86636186e-01
-7.97685981e-03 -1.22783852e+00 -5.30420780e-01 1.89149559e-01] | [9.50438404083252, -3.1039605140686035] |
56680ce2-7e91-4cf2-9a58-a5d2142fc4b0 | fa3l-at-semeval-2017-task-3-a-three | null | null | https://aclanthology.org/S17-2048 | https://aclanthology.org/S17-2048.pdf | FA3L at SemEval-2017 Task 3: A ThRee Embeddings Recurrent Neural Network for Question Answering | In this paper we present ThReeNN, a model for Community Question Answering, Task 3, of SemEval-2017. The proposed model exploits both syntactic and semantic information to build a single and meaningful embedding space. Using a dependency parser in combination with word embeddings, the model creates sequences of inputs for a Recurrent Neural Network, which are then used for the ranking purposes of the Task. The score obtained on the official test data shows promising results. | ['Ludovica Pannitto', 'Giuseppe Attardi', 'Antonio Carta', 'Andrea Madotto', 'Federico Errica'] | 2017-08-01 | null | null | null | semeval-2017-8 | ['question-similarity'] | ['natural-language-processing'] | [ 1.78085137e-02 2.19596863e-01 7.73085281e-02 -2.13938579e-01
-6.32660508e-01 -3.35544705e-01 5.51027536e-01 5.67556500e-01
-6.74234569e-01 3.60087395e-01 6.75089419e-01 -5.15794337e-01
-1.28068253e-01 -6.88723564e-01 -2.69317299e-01 -7.20557943e-02
6.11774959e-02 6.10226989e-01 6.08816803e-01 -4.27083492e-01
2.63918102e-01 -1.89943790e-01 -1.21936619e+00 5.58286369e-01
7.97038257e-01 7.67172277e-01 1.14102371e-01 9.88538444e-01
-5.43511033e-01 1.08751917e+00 -7.62831271e-01 -5.94569385e-01
-2.72054404e-01 -3.63197237e-01 -1.24474883e+00 -6.33085907e-01
3.34639966e-01 -1.41345009e-01 -3.30635101e-01 5.74147642e-01
3.89733046e-01 2.48138338e-01 4.91128683e-01 -6.55715466e-01
-9.49463725e-01 9.43808079e-01 3.46075237e-01 4.57033902e-01
6.52913213e-01 -1.03438787e-01 1.86653340e+00 -1.08962524e+00
7.55063057e-01 1.14587247e+00 2.91709721e-01 6.69228792e-01
-1.00081396e+00 -1.71076894e-01 -4.24203575e-02 5.72839439e-01
-8.91612470e-01 -1.03735663e-01 8.30645740e-01 -2.88787693e-01
1.39512134e+00 3.58768582e-01 5.53931534e-01 1.24974895e+00
-1.51005015e-02 7.23716974e-01 7.23620236e-01 -5.25864780e-01
3.50388318e-01 4.16309834e-02 9.77566838e-01 5.40323377e-01
1.77678406e-01 -4.33653831e-01 -4.96283978e-01 -5.68686247e-01
3.80459465e-02 -2.47762531e-01 -1.14793621e-01 -2.35195786e-01
-8.45276535e-01 1.15104485e+00 6.94114864e-01 6.26322329e-01
-4.82633144e-01 2.02650264e-01 3.84957522e-01 3.75292510e-01
5.04123688e-01 7.33632445e-01 -3.09067607e-01 -4.34750766e-02
-5.37645876e-01 4.25584644e-01 1.13397205e+00 1.37015209e-01
2.23130926e-01 -4.90671843e-01 -5.41188300e-01 8.94370854e-01
6.57539189e-01 1.41414702e-01 5.16146839e-01 -7.72426367e-01
7.63193309e-01 9.79296505e-01 -9.47585702e-02 -9.14768577e-01
-4.85160202e-01 -4.43395942e-01 -3.00064892e-01 -5.00108004e-01
3.73827547e-01 -1.51111826e-01 -5.41370630e-01 1.55431581e+00
1.29137337e-01 1.71016961e-01 4.13052648e-01 7.70030737e-01
1.20033514e+00 6.64506972e-01 3.39562476e-01 4.02794927e-01
1.44556439e+00 -1.30381608e+00 -6.97890878e-01 -2.41354674e-01
8.09046626e-01 -4.80805367e-01 8.92711759e-01 1.01452537e-01
-8.18188488e-01 -3.04699451e-01 -1.02559817e+00 -3.57984424e-01
-5.89772642e-01 -4.06503165e-03 2.88271159e-01 3.27959955e-01
-1.33446407e+00 2.43558511e-01 -4.44688648e-01 -6.32913709e-01
1.63387969e-01 6.70602918e-02 -1.60003677e-01 -2.02743322e-01
-1.54356098e+00 1.02685785e+00 4.22775716e-01 -2.36222185e-02
-4.43064958e-01 -3.71429533e-01 -8.46654356e-01 2.85111070e-01
-1.32570088e-01 -8.21018577e-01 1.20069301e+00 -3.78336936e-01
-1.26312339e+00 6.63381398e-01 -2.36144409e-01 -7.81209409e-01
-6.22538999e-02 -3.24238986e-01 -3.40176374e-01 3.67552131e-01
1.05895661e-01 4.28345889e-01 5.13962626e-01 -6.80330753e-01
-3.32079262e-01 -2.15165615e-01 4.30012971e-01 3.75974663e-02
-5.39156973e-01 3.39211583e-01 -3.40820611e-01 -2.41080105e-01
-1.28047168e-01 -5.28713703e-01 -3.46576303e-01 -6.79092765e-01
-1.84185654e-01 -1.03841233e+00 6.96979702e-01 -1.20897996e+00
1.52400053e+00 -1.98564684e+00 5.12476981e-01 2.76141673e-01
2.76812464e-01 2.98129857e-01 -5.92746973e-01 6.22986019e-01
-1.14414170e-01 3.52261543e-01 -2.07057282e-01 -4.26968932e-01
-9.07196403e-02 1.40191317e-01 -2.50036776e-01 -1.55684382e-01
6.17009819e-01 1.15969765e+00 -1.13068795e+00 -2.57036775e-01
-1.03580579e-01 3.80237252e-01 -6.46262288e-01 5.42313039e-01
-5.08209825e-01 -1.24000780e-01 -6.22409880e-01 2.07435817e-01
8.52584839e-02 -3.49071592e-01 3.46593171e-01 1.55003607e-01
1.33342639e-01 9.48711872e-01 -6.08609855e-01 1.62259698e+00
-5.81207275e-01 4.42382693e-01 -4.27031785e-01 -9.05054450e-01
9.83397841e-01 5.27222097e-01 5.37335537e-02 -6.48645520e-01
8.03532451e-02 2.74482220e-01 1.21450044e-01 -7.45545685e-01
4.67108846e-01 4.53901947e-01 -1.53717279e-01 7.43939340e-01
4.48645443e-01 7.49704167e-02 4.42632347e-01 6.11737370e-01
1.64667916e+00 -6.08681887e-02 1.38945669e-01 -2.17536584e-01
8.91668618e-01 -1.36727750e-01 -1.43693695e-02 4.74500418e-01
-2.74902694e-02 6.23685539e-01 6.73554122e-01 -5.91891289e-01
-1.06656682e+00 -1.00259161e+00 5.54816574e-02 1.25602365e+00
-3.41202617e-01 -7.04440594e-01 -7.58658350e-01 -1.04674768e+00
-6.11663759e-02 9.69619930e-01 -8.30582142e-01 -1.14964537e-01
-7.32873380e-01 -3.17665130e-01 4.47440982e-01 3.51560414e-01
4.94232811e-02 -1.35717332e+00 -5.02264738e-01 5.69726646e-01
-3.35446149e-01 -1.15417325e+00 -2.29694620e-01 2.66229510e-01
-8.08010161e-01 -1.16576445e+00 -4.14975673e-01 -1.04282689e+00
4.06654358e-01 -2.78169680e-02 1.34029913e+00 4.48876858e-01
-2.06607431e-02 2.47901693e-01 -7.82810271e-01 -6.26533478e-02
-1.64262056e-01 6.67714536e-01 -4.51925337e-01 -1.05091631e-01
6.20089948e-01 -4.76577729e-01 -4.32047069e-01 -2.48870060e-01
-8.61125231e-01 -3.84446114e-01 3.66650254e-01 9.57816482e-01
3.19055319e-02 -8.18900287e-01 8.72525573e-01 -7.77999938e-01
1.18580580e+00 -7.03781903e-01 -2.62854785e-01 4.62197781e-01
-2.62259632e-01 4.12567437e-01 6.36479735e-01 -2.64374558e-02
-7.92745292e-01 -5.11898696e-01 -6.04444146e-01 1.42007038e-01
1.69186011e-01 7.55320370e-01 1.48056298e-01 2.70598233e-01
8.57674003e-01 9.05987546e-02 -3.14379394e-01 -7.00789034e-01
5.24158537e-01 8.55158210e-01 2.61294752e-01 -3.98664534e-01
7.10446775e-01 -1.21999234e-01 -5.60536027e-01 -7.58539557e-01
-1.08789933e+00 -6.84611320e-01 -3.43980432e-01 4.28703986e-03
1.09362340e+00 -5.97855866e-01 -3.61320913e-01 -6.57529682e-02
-1.66517627e+00 5.06810844e-02 -2.46220812e-01 2.92455554e-01
2.57961936e-02 1.72114700e-01 -6.40180290e-01 -7.40013897e-01
-7.50201881e-01 -8.20843637e-01 6.31848633e-01 2.71317214e-01
-5.04946828e-01 -1.22633493e+00 5.16199648e-01 9.02032852e-01
5.84768534e-01 -1.03860423e-01 1.06026208e+00 -1.43839455e+00
-3.92598838e-01 -2.20306650e-01 -2.43938133e-01 5.15262961e-01
-2.81961679e-01 -1.69800207e-01 -8.39866698e-01 1.28483042e-01
-1.94241896e-01 -4.86841381e-01 1.10978067e+00 -5.34693003e-02
1.01814306e+00 -1.51481152e-01 6.75539225e-02 -9.24484525e-03
1.21801507e+00 -2.49084815e-01 4.84229296e-01 3.98506969e-01
4.93023455e-01 6.96391702e-01 1.65693890e-02 1.88000381e-01
9.14993286e-01 4.63197023e-01 4.36901271e-01 4.18818980e-01
-1.61726832e-01 -3.23757678e-01 4.86894339e-01 1.47271097e+00
1.14705309e-01 -2.19930485e-01 -1.11887562e+00 7.93384910e-01
-1.71414995e+00 -8.47540021e-01 -2.87608057e-01 1.70019078e+00
6.89732373e-01 1.06913097e-01 -2.65219480e-01 -5.57988882e-02
4.14509207e-01 4.78118569e-01 -7.51553625e-02 -1.01486981e+00
-9.21657458e-02 9.13697720e-01 -1.78712681e-02 6.04072452e-01
-9.29527521e-01 7.91419804e-01 6.69509268e+00 4.41355616e-01
-5.90405464e-01 5.46293557e-01 2.62963295e-01 2.54295170e-01
-6.89531505e-01 8.43175724e-02 -4.25946414e-01 2.14888066e-01
1.38418591e+00 -8.60724598e-02 2.11454615e-01 5.91988087e-01
-1.45064443e-01 9.51527990e-03 -1.00920236e+00 4.05084282e-01
4.49136585e-01 -1.46289480e+00 2.40251288e-01 -3.22133213e-01
4.30153131e-01 4.19111133e-01 -2.17230201e-01 7.37048745e-01
4.52879280e-01 -1.06926501e+00 2.95978636e-01 4.60850775e-01
2.77619027e-02 -6.35525286e-01 1.19968629e+00 3.01639378e-01
-8.96946192e-01 -1.09512873e-01 -3.00837696e-01 -1.58688560e-01
8.88395831e-02 2.60974050e-01 -7.95088291e-01 7.08972454e-01
5.31342149e-01 5.80291808e-01 -9.46959794e-01 9.00250018e-01
-8.86472404e-01 1.08493793e+00 -1.20182574e-01 -6.12501860e-01
4.28471774e-01 8.56111497e-02 4.77594554e-01 1.23269486e+00
9.51162577e-02 -2.50283808e-01 -4.13114438e-03 6.94973648e-01
-5.61001539e-01 6.02079093e-01 -3.99681479e-01 -2.45115429e-01
3.24712962e-01 1.21834195e+00 -1.28631175e-01 -4.23377752e-01
-3.46719652e-01 8.80110502e-01 9.24512088e-01 3.40262204e-01
-5.25128424e-01 -4.33973640e-01 2.21243769e-01 -2.38040045e-01
5.02794325e-01 -1.40287384e-01 -1.28806308e-01 -1.34993148e+00
1.92118570e-01 -7.71594942e-01 6.04802787e-01 -5.43485045e-01
-1.32522726e+00 7.54700720e-01 -3.31775069e-01 -3.79875422e-01
-4.03198123e-01 -5.36256075e-01 -9.65583324e-01 1.06197405e+00
-1.70068598e+00 -1.10716641e+00 -1.04475170e-01 1.75044492e-01
4.37719315e-01 -3.08592111e-01 1.27686667e+00 2.57925242e-01
-5.53810477e-01 4.25753444e-01 7.18843415e-02 4.42213118e-01
4.13772911e-01 -1.37353218e+00 7.31511652e-01 6.61374390e-01
4.49135661e-01 6.48324966e-01 5.52522480e-01 -3.69289696e-01
-1.16293514e+00 -9.51233089e-01 1.87850881e+00 -8.40059221e-01
1.10810494e+00 -3.37283075e-01 -9.84539390e-01 3.95238191e-01
5.07786930e-01 -1.22602463e-01 9.21201408e-01 4.34865505e-01
-7.16509461e-01 2.25209326e-01 -8.29971671e-01 3.62635553e-01
6.04592264e-01 -7.83147395e-01 -1.41412652e+00 1.74360126e-01
1.27683651e+00 -2.57812645e-02 -8.96184742e-01 -6.16211742e-02
4.21330810e-01 -6.93488479e-01 9.10200775e-01 -9.14298952e-01
6.33422911e-01 4.44206148e-02 -1.53189629e-01 -1.16955650e+00
-2.17760280e-01 -2.69269973e-01 -3.22713733e-01 8.93533885e-01
8.28570545e-01 -6.18797243e-01 4.85083014e-01 2.86121756e-01
5.83952405e-02 -8.42255116e-01 -1.19304287e+00 -4.60382134e-01
1.54343024e-01 -3.64554048e-01 4.08721328e-01 6.50560260e-01
1.66922376e-01 1.01143062e+00 1.51839450e-01 -2.53971089e-02
5.09241045e-01 -9.96259749e-02 3.03124577e-01 -1.36047912e+00
-2.39517897e-01 -3.58142406e-01 -2.91680694e-01 -7.50025809e-01
3.40838999e-01 -1.16198444e+00 -8.69193003e-02 -2.00414276e+00
1.99994564e-01 1.71351638e-02 -6.84279740e-01 3.05235445e-01
-5.19602776e-01 1.62844360e-01 3.39342862e-01 -1.24909170e-01
-9.92141247e-01 5.54974258e-01 6.54992938e-01 -2.41263002e-01
8.30701739e-02 -2.80856192e-01 -7.09086657e-01 4.46786642e-01
1.05032444e+00 -6.08190656e-01 -4.84656496e-03 -8.31713140e-01
3.77218068e-01 -2.06388623e-01 2.74711818e-01 -8.50836515e-01
1.39517978e-01 3.36387932e-01 -2.09849700e-02 -3.83703798e-01
1.85251087e-01 -5.63152015e-01 -4.37511981e-01 6.69685721e-01
-7.51746237e-01 8.85947198e-02 -2.02824324e-01 3.22158068e-01
-3.71045321e-01 -6.89771354e-01 1.48572743e-01 1.89336509e-01
-3.42964381e-01 5.00410749e-03 -1.98903844e-01 3.50269228e-01
5.59825242e-01 3.69013906e-01 -3.28721553e-01 -2.64961094e-01
-7.39716053e-01 5.54986954e-01 -1.36464298e-01 7.66820908e-01
6.93055391e-01 -1.35634184e+00 -1.12534916e+00 -2.23981604e-01
4.38079923e-01 -3.43425781e-01 1.49230391e-01 5.32405078e-01
-5.39008856e-01 5.96415818e-01 1.67554766e-01 -1.86373994e-01
-1.03752339e+00 5.30859455e-02 1.12292133e-01 -1.01898015e+00
-1.96674094e-01 8.44775379e-01 -5.35116315e-01 -6.55193031e-01
1.40263801e-02 -4.34809595e-01 -1.01776528e+00 2.18290687e-01
6.79111004e-01 1.00280114e-01 1.06694885e-01 -3.04475963e-01
-4.96426433e-01 1.21127687e-01 2.96637416e-02 -1.71406448e-01
1.66454005e+00 2.49307409e-01 -5.01460969e-01 3.07631910e-01
1.21862364e+00 -1.35632560e-01 -3.01896483e-01 -5.92866480e-01
6.78502738e-01 6.32630363e-02 -9.10466611e-02 -7.40679324e-01
-5.04129171e-01 1.07490492e+00 3.08377951e-01 3.91184628e-01
5.87009251e-01 3.29201549e-01 1.01032305e+00 7.46564686e-01
-1.43922940e-01 -1.02798748e+00 3.00876379e-01 1.11251307e+00
8.56709182e-01 -9.14307296e-01 -4.26047266e-01 1.82666361e-01
-3.65106732e-01 1.29874098e+00 5.48852324e-01 -3.53824407e-01
4.57555830e-01 -2.52256602e-01 -8.56748596e-02 -2.65983999e-01
-1.18964911e+00 -3.73392791e-01 4.66589928e-01 5.33433795e-01
6.07874691e-01 3.84063795e-02 -8.52005064e-01 9.33975577e-01
-1.73688978e-01 -2.09719002e-01 3.67128819e-01 7.18169153e-01
-7.95009971e-01 -1.47490001e+00 3.67584452e-02 3.28423917e-01
-5.22367299e-01 -2.76466101e-01 -5.54386377e-01 2.40531802e-01
6.91547915e-02 1.16157448e+00 -3.11063249e-02 -4.89757150e-01
4.97852772e-01 5.13052106e-01 1.36851862e-01 -9.68953967e-01
-1.08896375e+00 -7.22038269e-01 5.68781674e-01 -4.20133978e-01
-2.96214223e-01 -3.92007709e-01 -1.13805008e+00 1.46753877e-01
-1.16625898e-01 6.67825282e-01 7.00378597e-01 1.04914570e+00
3.47871095e-01 5.85662007e-01 5.32061994e-01 -7.48586059e-02
-7.40213692e-01 -1.33580351e+00 1.17634252e-01 4.54531789e-01
8.24023318e-03 -2.59041160e-01 -4.41256106e-01 -2.80797124e-01] | [11.252071380615234, 8.031338691711426] |
f88ddd76-c362-4828-b325-d4b12825eee9 | semantic-scene-completion-via-integrating | 2104.03640 | null | https://arxiv.org/abs/2104.03640v2 | https://arxiv.org/pdf/2104.03640v2.pdf | Semantic Scene Completion via Integrating Instances and Scene in-the-Loop | Semantic Scene Completion aims at reconstructing a complete 3D scene with precise voxel-wise semantics from a single-view depth or RGBD image. It is a crucial but challenging problem for indoor scene understanding. In this work, we present a novel framework named Scene-Instance-Scene Network (\textit{SISNet}), which takes advantages of both instance and scene level semantic information. Our method is capable of inferring fine-grained shape details as well as nearby objects whose semantic categories are easily mixed-up. The key insight is that we decouple the instances from a coarsely completed semantic scene instead of a raw input image to guide the reconstruction of instances and the overall scene. SISNet conducts iterative scene-to-instance (SI) and instance-to-scene (IS) semantic completion. Specifically, the SI is able to encode objects' surrounding context for effectively decoupling instances from the scene and each instance could be voxelized into higher resolution to capture finer details. With IS, fine-grained instance information can be integrated back into the 3D scene and thus leads to more accurate semantic scene completion. Utilizing such an iterative mechanism, the scene and instance completion benefits each other to achieve higher completion accuracy. Extensively experiments show that our proposed method consistently outperforms state-of-the-art methods on both real NYU, NYUCAD and synthetic SUNCG-RGBD datasets. The code and the supplementary material will be available at \url{https://github.com/yjcaimeow/SISNet}. | ['Hongsheng Li', 'Xiaogang Wang', 'Kwan-Yee Lin', 'Chao Zhang', 'Xuesong Chen', 'Yingjie Cai'] | 2021-04-08 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Cai_Semantic_Scene_Completion_via_Integrating_Instances_and_Scene_In-the-Loop_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Cai_Semantic_Scene_Completion_via_Integrating_Instances_and_Scene_In-the-Loop_CVPR_2021_paper.pdf | cvpr-2021-1 | ['3d-semantic-scene-completion'] | ['computer-vision'] | [ 2.9006752e-01 2.9319942e-02 2.5665006e-01 -7.2168189e-01
-6.5325296e-01 -4.3884712e-01 5.1090491e-01 -7.3251552e-03
4.2914860e-02 4.9273080e-01 1.8658106e-01 8.0360234e-02
-9.4116807e-02 -1.1288859e+00 -8.3978844e-01 -4.4304064e-01
4.1119438e-01 4.4685897e-01 5.0005239e-01 -1.5089756e-01
4.6979137e-02 4.7120288e-01 -1.7603320e+00 3.0627084e-01
7.9033685e-01 1.0763496e+00 9.9230570e-01 5.1670253e-01
-3.4459886e-01 5.6883699e-01 -2.4132704e-02 9.8495267e-02
3.2175210e-01 -9.2296287e-02 -6.0711014e-01 3.5214749e-01
4.5644283e-01 -5.0091320e-01 -3.3423415e-01 1.0387274e+00
2.1108222e-01 2.7576879e-01 2.4313095e-01 -8.8937241e-01
-1.7531182e-01 -9.0083420e-02 -5.1710212e-01 -2.6359904e-01
6.3968712e-01 5.4878518e-02 7.2166187e-01 -1.0555336e+00
7.1351379e-01 1.2497250e+00 3.1239110e-01 2.2074382e-01
-9.8420990e-01 -5.6200927e-01 4.9767321e-01 1.8279305e-02
-1.3346721e+00 -2.5883397e-01 1.2126331e+00 -2.3974188e-01
7.2783130e-01 3.4854740e-01 8.8829654e-01 7.6014733e-01
-2.1539108e-01 6.9184905e-01 1.2212023e+00 -7.9329185e-02
4.9627826e-01 -1.1439834e-01 1.2795211e-01 7.5692314e-01
1.1454488e-01 -4.1457288e-02 -5.2991974e-01 2.2108684e-01
1.1787479e+00 6.2629646e-01 -2.7656922e-01 -5.5732310e-01
-1.3594253e+00 3.0905902e-01 9.1678214e-01 7.1787708e-02
-5.8007801e-01 3.3529609e-01 -3.5427008e-02 -2.6789317e-01
5.3939658e-01 1.0452137e-01 -5.4186845e-01 -4.6635367e-02
-7.6542497e-01 3.0087230e-01 4.0397567e-01 1.1778197e+00
1.2188790e+00 -1.5521273e-01 1.4871861e-01 7.7020067e-01
3.4232411e-01 6.5428591e-01 -1.5666746e-01 -1.2571740e+00
5.0011569e-01 9.0222150e-01 4.4441167e-02 -7.9631674e-01
-4.3950698e-01 -4.2543325e-01 -8.0842203e-01 1.5021034e-01
2.2320119e-01 3.5653311e-01 -1.2859148e+00 1.3030475e+00
8.4084529e-01 6.2011731e-01 -1.2545869e-01 1.1073883e+00
1.2654713e+00 6.3295072e-01 -2.0515865e-02 3.1768769e-01
1.4662153e+00 -1.0094270e+00 -4.1483650e-01 -4.7594431e-01
9.4148353e-02 -5.8126485e-01 1.2525499e+00 1.8911213e-01
-8.8309520e-01 -6.7791748e-01 -9.3954945e-01 -5.5832207e-01
-2.9529506e-01 -1.1024903e-01 8.3202136e-01 1.3587870e-01
-8.4029317e-01 2.1427573e-01 -9.3782240e-01 -3.2110435e-01
6.5022612e-01 -7.3080696e-02 -5.4468060e-01 -8.4384584e-01
-7.9965615e-01 2.9046002e-01 4.8757410e-01 1.2234563e-01
-1.1721778e+00 -9.3874580e-01 -1.2165229e+00 -3.6845472e-02
6.3677835e-01 -1.1209847e+00 1.0765131e+00 -3.3450440e-01
-1.1592171e+00 7.8877765e-01 -4.8881757e-01 1.0966362e-01
3.1744510e-01 -1.5019946e-01 3.7732799e-02 2.2340591e-01
4.0410310e-01 5.9186536e-01 5.4961431e-01 -1.8431648e+00
-6.5786546e-01 -7.4417394e-01 4.8592791e-01 5.0579363e-01
2.9444864e-01 -6.5360737e-01 -8.9169401e-01 -4.5804134e-01
8.0080163e-01 -5.3224093e-01 -3.7563133e-01 1.2540281e-01
-5.9706759e-01 1.2960251e-01 7.5727868e-01 -7.3118317e-01
7.6913053e-01 -2.1434767e+00 1.8263158e-01 7.5209543e-02
4.9514711e-01 -2.3262644e-01 -2.5297426e-02 2.5651094e-01
1.0363090e-01 -1.8895857e-01 -5.3506869e-01 -7.9347718e-01
-3.6842335e-02 4.6000144e-01 -1.2369988e-01 3.5860580e-01
-2.2781292e-01 1.0536861e+00 -1.1804119e+00 -2.2560531e-01
8.6243725e-01 7.4042416e-01 -6.6530937e-01 2.9843417e-01
-4.9850702e-01 9.3491280e-01 -8.9518309e-01 9.2172956e-01
1.0827355e+00 -2.7618596e-01 -5.0256349e-02 -3.8308269e-01
-5.9789572e-02 2.2719562e-01 -1.3620975e+00 2.5343273e+00
-5.9949273e-01 1.7085905e-01 2.3915526e-01 -7.8500688e-01
7.9736525e-01 -8.2711190e-02 3.7158984e-01 -9.8415768e-01
-6.4463586e-02 1.5388751e-01 -7.7791643e-01 -2.4938035e-01
5.2558464e-01 -2.5706902e-01 -1.6292195e-01 7.9520732e-02
-1.7025179e-01 -7.8732663e-01 -2.9193705e-01 2.8019848e-01
7.7116948e-01 5.1230681e-01 2.1956518e-01 -3.2486852e-02
4.5802477e-01 2.2865504e-01 5.8439153e-01 4.4243926e-01
1.5392321e-01 8.9821517e-01 -1.7622976e-02 -4.0112823e-01
-1.0017611e+00 -1.3942487e+00 -1.3800342e-01 5.0823081e-01
8.2007456e-01 -4.0841344e-01 -6.8216461e-01 -2.8728855e-01
1.2804213e-03 8.0998713e-01 -5.6342798e-01 2.1691705e-01
-3.7915751e-01 -1.7784466e-01 -1.2978235e-01 4.9515074e-01
8.9958620e-01 -8.8883781e-01 -5.9145671e-01 6.1861224e-02
-4.4506857e-01 -1.3535376e+00 -2.5683403e-01 -6.8262131e-03
-1.0166093e+00 -1.0904264e+00 -4.7916406e-01 -5.2821499e-01
8.4305519e-01 7.7246159e-01 1.1315019e+00 3.7738997e-02
-3.2417718e-01 4.9910501e-01 -3.9676329e-01 -2.2210503e-01
4.4482987e-02 -2.0616062e-01 -1.6048332e-01 -1.3949811e-01
2.3356650e-02 -9.2821568e-01 -8.6209697e-01 2.0554341e-01
-1.0055017e+00 8.8130522e-01 2.8304681e-01 4.1015267e-01
1.2345008e+00 3.4531009e-01 -6.6413932e-02 -9.1349602e-01
-1.3770418e-01 -5.5417216e-01 -5.1332569e-01 2.0931229e-02
-5.3350382e-02 -2.0805995e-01 5.7219160e-01 2.6489949e-01
-1.3640648e+00 2.4391024e-01 -2.4821058e-01 -7.3801339e-01
-7.1450239e-01 1.7293592e-01 -6.0429370e-01 1.9274041e-01
2.7261344e-01 5.0297773e-01 -4.4515225e-01 -7.6108080e-01
4.9007410e-01 2.5164044e-01 5.6794363e-01 -7.7192402e-01
8.5290569e-01 9.8173624e-01 1.1230962e-01 -8.6584646e-01
-1.1211616e+00 -7.4851125e-01 -8.8528532e-01 -2.9783559e-01
1.0207205e+00 -1.2696233e+00 -3.7243053e-01 4.9666706e-01
-1.0864391e+00 -6.1679476e-01 -4.2767459e-01 2.6076800e-01
-5.9666204e-01 1.7242904e-01 -3.5176975e-01 -5.3712004e-01
7.0203818e-02 -1.0981308e+00 1.6803958e+00 2.6166847e-01
2.7500689e-01 -9.8614520e-01 -1.4824453e-01 7.1882355e-01
-1.3044359e-02 6.2787461e-01 5.8589417e-01 1.9237903e-01
-1.2056890e+00 3.1294830e-02 -4.8909059e-01 9.7672932e-02
3.9711824e-01 -4.4944507e-01 -1.2073029e+00 3.6210753e-02
1.7197880e-01 6.2074577e-03 7.5385004e-01 4.3239698e-01
1.3019605e+00 1.8691979e-02 -1.9199383e-01 1.0255567e+00
1.8477625e+00 -7.3540464e-02 6.8884784e-01 2.5159892e-01
1.2799312e+00 5.1839936e-01 8.0057836e-01 5.2781367e-01
9.1672206e-01 7.2279912e-01 7.9217356e-01 -2.3309132e-01
-5.6139642e-01 -7.6051211e-01 -2.6438591e-01 6.8211561e-01
-1.2584935e-01 4.2890158e-02 -9.0232968e-01 4.9683625e-01
-1.7928073e+00 -7.4803227e-01 -3.7074435e-01 2.0305176e+00
4.3595144e-01 -6.2730215e-02 -2.4420050e-01 9.8031528e-02
4.2110407e-01 3.0563354e-01 -7.6127839e-01 1.5968795e-01
-2.6385111e-03 1.4375165e-01 3.1564718e-01 7.8908771e-01
-7.7616727e-01 1.0666974e+00 4.2687688e+00 8.5601199e-01
-7.0678300e-01 2.9093406e-01 7.1688986e-01 -2.0091578e-02
-6.6143858e-01 1.2626037e-01 -6.2419617e-01 2.2717139e-01
1.2065142e-01 2.5370657e-01 5.9790319e-01 9.6421462e-01
3.2405955e-01 -4.4360539e-01 -1.0267423e+00 1.2881482e+00
-1.4997989e-01 -1.4236585e+00 1.5049595e-01 7.3329292e-02
8.7160337e-01 7.5750157e-02 -4.0948841e-01 4.3095537e-02
2.3954560e-01 -7.7643967e-01 1.0017645e+00 7.4466962e-01
1.0104486e+00 -5.3250909e-01 3.5968432e-01 6.6661781e-01
-1.7629677e+00 1.3345301e-01 -4.0390953e-01 -2.0843664e-01
3.4057450e-01 9.8293853e-01 -5.5477214e-01 1.0339971e+00
9.1514581e-01 1.1222707e+00 -3.7453797e-01 8.4039897e-01
-3.9319122e-01 1.5353209e-01 -4.0476269e-01 4.0606743e-01
1.9549248e-01 -3.9736354e-01 5.2249342e-01 8.1082201e-01
2.3343778e-01 6.3861626e-01 4.5718732e-01 1.1601841e+00
1.2901255e-01 -2.5687271e-01 -4.7214878e-01 4.9278840e-01
4.4131711e-01 1.2724749e+00 -1.0302411e+00 -5.6054491e-01
-2.2212498e-01 1.2022761e+00 3.5463116e-01 5.3617138e-01
-7.5823754e-01 -1.0359355e-01 8.6359167e-01 3.7282395e-01
2.9722485e-01 -5.2849782e-01 -7.5248915e-01 -1.4404092e+00
2.6145062e-01 -2.3668829e-01 4.7627438e-02 -1.1367911e+00
-9.3675536e-01 3.9477032e-01 1.4960647e-01 -1.1759808e+00
3.6178711e-01 -3.9376494e-01 -4.2607760e-01 8.5295391e-01
-1.8637186e+00 -1.4120086e+00 -1.1571906e+00 9.6823168e-01
9.1215891e-01 6.2017238e-01 6.1362690e-01 3.0136949e-01
-3.1322110e-01 -1.4669296e-01 -1.7284425e-01 -8.9667268e-02
8.2210653e-02 -1.1645403e+00 3.5943088e-01 8.2974446e-01
-3.8438518e-02 4.2073944e-01 3.9569682e-01 -8.3055663e-01
-1.4463861e+00 -1.3318204e+00 4.4354713e-01 -5.5192137e-01
1.6635086e-01 -6.1934137e-01 -6.6245908e-01 4.3599930e-01
-5.4418433e-01 1.2884051e-01 1.1467859e-01 -2.4649698e-01
-3.2484192e-01 -2.5308970e-01 -1.2555622e+00 5.0480646e-01
1.6493958e+00 -7.1581662e-01 -3.7580827e-01 4.0258515e-01
1.2259685e+00 -8.4453166e-01 -8.1811535e-01 4.6632379e-01
2.8683081e-01 -1.3123460e+00 1.4399370e+00 7.8204442e-03
6.5485328e-01 -6.9750911e-01 -7.0184338e-01 -9.9160582e-01
-2.7764717e-01 -1.3141762e-02 -1.8794881e-01 1.0026591e+00
-1.4886872e-01 -5.2351713e-01 7.8978783e-01 7.6839292e-01
-5.3096247e-01 -8.2840729e-01 -8.4345877e-01 -4.3736088e-01
-3.9703569e-01 -9.4791514e-01 9.7282773e-01 8.6502939e-01
-6.1220700e-01 4.2732544e-02 1.1799085e-02 5.1373166e-01
1.1557304e+00 4.9032933e-01 9.6443242e-01 -1.0541250e+00
-1.0514835e-01 -1.4124328e-01 -4.3694207e-01 -1.4372246e+00
-1.5507470e-01 -8.2835150e-01 -2.1053055e-02 -2.2688091e+00
1.6716057e-01 -9.0143889e-01 -4.5188475e-02 4.6565798e-01
-1.9704336e-01 3.9170662e-01 2.0768575e-01 9.0038881e-02
-7.4375021e-01 8.4561902e-01 1.7633758e+00 5.9196290e-02
-2.0817280e-01 -1.1270616e-01 -6.6211128e-01 6.9705665e-01
6.5320820e-01 -2.0338561e-01 -6.2597305e-01 -5.9000856e-01
-9.8777369e-02 2.0652880e-01 8.7260896e-01 -1.1009276e+00
1.1718924e-01 -2.7373394e-01 5.3220087e-01 -8.6869729e-01
8.7938815e-01 -1.0128816e+00 4.9510518e-01 1.7140199e-01
2.5847754e-01 -4.7057402e-01 1.4666952e-01 7.5479180e-01
-1.5142187e-01 3.0042204e-01 6.1048079e-01 -6.4589816e-01
-1.1928678e+00 6.8395114e-01 4.3556160e-01 -1.0607737e-01
1.0465534e+00 -5.9054214e-01 -6.4837292e-02 -2.1285914e-01
-7.9852235e-01 3.1493813e-01 8.9097649e-01 3.7217760e-01
1.0995419e+00 -1.3063724e+00 -4.7186914e-01 4.2457503e-01
2.9152277e-01 1.0947733e+00 8.8035905e-01 6.3565671e-01
-6.7904264e-01 4.2258596e-01 4.8351847e-02 -1.0084696e+00
-1.0164585e+00 3.6731523e-01 2.0936996e-01 3.6806889e-02
-1.0214579e+00 9.1128862e-01 1.0359334e+00 -9.2597336e-01
4.3556146e-02 -6.4357471e-01 2.3270896e-01 -4.2211837e-01
4.4277233e-01 3.3570918e-01 -4.8939731e-02 -6.7294794e-01
-3.6198428e-01 9.2766792e-01 3.8044921e-01 1.2255383e-01
1.5668859e+00 -4.9759236e-01 -2.8180286e-01 4.7351122e-01
1.1505255e+00 -9.9935181e-02 -1.6128972e+00 -3.9232394e-01
-6.3414669e-01 -9.8208165e-01 1.8264388e-01 -6.4304948e-01
-1.0817776e+00 8.7255269e-01 3.5317367e-01 -2.0361069e-01
1.3374259e+00 3.8766977e-01 9.0708667e-01 1.9455321e-02
1.0517958e+00 -6.9484144e-01 5.9322529e-02 4.0066743e-01
8.2097805e-01 -1.1778158e+00 1.9304931e-01 -9.0332526e-01
-4.0059939e-01 8.7704867e-01 6.3831049e-01 -2.0466973e-01
7.9848427e-01 3.5841867e-02 -3.1065530e-01 -6.0368353e-01
-1.9313830e-01 -2.3403364e-01 3.4933555e-01 5.5365181e-01
-1.6599242e-01 4.0859240e-01 5.0416255e-01 6.0593832e-01
-2.9653054e-01 -3.1062165e-02 2.1628335e-01 8.0480450e-01
-4.7021213e-01 -8.4625286e-01 -4.8470038e-01 3.3091494e-01
3.0105582e-01 -1.3343425e-01 4.5971293e-02 4.9871284e-01
4.0133983e-01 8.0640388e-01 4.6182238e-03 -3.7552783e-01
5.4077762e-01 -3.7531790e-01 4.8569259e-01 -9.6275496e-01
2.8845064e-02 3.6103290e-02 -9.9260360e-02 -1.2441658e+00
-4.7022849e-01 -5.5158675e-01 -1.5657278e+00 -2.9343480e-01
1.3039774e-01 -2.8563756e-01 8.8647521e-01 8.7759566e-01
3.9040521e-01 8.2517397e-01 5.9962088e-01 -1.4228065e+00
4.5379555e-01 -5.3809881e-01 -7.0184582e-01 4.3447235e-01
4.1772547e-01 -7.5968969e-01 -2.1514931e-01 -2.3650074e-02] | [8.536799430847168, -2.824820041656494] |
f38eac06-0f51-4ef9-ac9e-11353038743a | patient-independent-epileptic-seizure | 2011.09581 | null | https://arxiv.org/abs/2011.09581v1 | https://arxiv.org/pdf/2011.09581v1.pdf | Patient-independent Epileptic Seizure Prediction using Deep Learning Models | Objective: Epilepsy is one of the most prevalent neurological diseases among humans and can lead to severe brain injuries, strokes, and brain tumors. Early detection of seizures can help to mitigate injuries, and can be used to aid the treatment of patients with epilepsy. The purpose of a seizure prediction system is to successfully identify the pre-ictal brain stage, which occurs before a seizure event. Patient-independent seizure prediction models are designed to offer accurate performance across multiple subjects within a dataset, and have been identified as a real-world solution to the seizure prediction problem. However, little attention has been given for designing such models to adapt to the high inter-subject variability in EEG data. Methods: We propose two patient-independent deep learning architectures with different learning strategies that can learn a global function utilizing data from multiple subjects. Results: Proposed models achieve state-of-the-art performance for seizure prediction on the CHB-MIT-EEG dataset, demonstrating 88.81% and 91.54% accuracy respectively. Conclusions: The Siamese model trained on the proposed learning strategy is able to learn patterns related to patient variations in data while predicting seizures. Significance: Our models show superior performance for patient-independent seizure prediction, and the same architecture can be used as a patient-specific classifier after model adaptation. We are the first study that employs model interpretation to understand classifier behavior for the task for seizure prediction, and we also show that the MFCC feature map utilized by our models contains predictive biomarkers related to interictal and pre-ictal brain states. | ['Clinton Fookes', 'Sridha Sridharan', 'Simon Denman', 'Tharindu Fernando', 'Theekshana Dissanayake'] | 2020-11-18 | null | null | null | null | ['seizure-prediction'] | ['medical'] | [-6.19902415e-03 -3.76527846e-01 1.40851825e-01 -4.70946521e-01
-9.40174460e-01 -1.95782542e-01 2.89602935e-01 1.72533356e-02
-3.13914299e-01 6.49851739e-01 3.07825267e-01 1.35186523e-01
-6.06468737e-01 -2.26677313e-01 -3.49242479e-01 -8.21901381e-01
-7.19426215e-01 5.76322317e-01 -1.54438755e-02 -1.98297828e-01
1.75593793e-01 6.72628939e-01 -1.23427141e+00 5.40505886e-01
8.37332845e-01 1.11699212e+00 3.66209030e-01 4.50049102e-01
3.52509081e-01 5.58398843e-01 -9.09071267e-01 4.55432475e-01
1.26856744e-01 -5.77777386e-01 -5.08348286e-01 -2.97350734e-01
-1.87794477e-01 8.03646222e-02 -2.50706613e-01 7.76766002e-01
1.02302158e+00 -3.15690428e-01 8.46401811e-01 -1.15129471e+00
-3.23427200e-01 6.30710602e-01 -2.05533147e-01 6.17165804e-01
2.19713271e-01 4.22164351e-02 4.83085424e-01 -6.79557204e-01
1.29259020e-01 5.09578228e-01 5.76968789e-01 8.11477244e-01
-1.01562071e+00 -1.09942794e+00 -2.95528054e-01 8.26842546e-01
-1.56855178e+00 -3.52701008e-01 8.08024704e-01 -7.31127262e-01
1.07354748e+00 3.66975069e-01 1.14457774e+00 1.28849125e+00
8.47035885e-01 7.21004546e-01 1.35535443e+00 -3.62057574e-02
3.68289083e-01 -1.63885042e-01 3.44235837e-01 8.86810645e-02
-9.27365571e-02 2.06147760e-01 -1.09940350e+00 -1.46373779e-01
2.16018155e-01 4.66855578e-02 -8.25107574e-01 -4.60086465e-02
-1.12587154e+00 6.45147204e-01 4.25595254e-01 7.28968263e-01
-6.61502779e-01 -9.88996774e-02 3.19956630e-01 3.19138497e-01
5.15273392e-01 7.08124399e-01 -8.58337402e-01 -4.44495291e-01
-1.36898410e+00 5.44762611e-02 6.11224234e-01 4.09514368e-01
3.54015261e-01 1.88544720e-01 -2.53506124e-01 7.74285674e-01
-7.74821341e-02 4.53385860e-01 1.39379883e+00 -1.48761123e-01
9.58753154e-02 6.36148155e-01 -3.88794959e-01 -4.70539838e-01
-1.10684729e+00 -9.70389426e-01 -9.21932578e-01 -1.06332064e-01
7.03971535e-02 -1.83817208e-01 -9.73748326e-01 1.75008202e+00
-5.17167091e-01 6.66332841e-01 2.20809475e-01 6.05706096e-01
7.02923357e-01 1.98940963e-01 1.84760354e-02 -3.38881224e-01
1.24507523e+00 -3.82293910e-01 -6.13064110e-01 -4.02169853e-01
7.25327849e-01 -2.81854630e-01 7.82745719e-01 6.89956963e-01
-7.14090943e-01 -7.43293166e-02 -8.26636791e-01 6.50963545e-01
-3.43111575e-01 2.40497902e-01 2.87845433e-01 5.38550973e-01
-9.91510689e-01 4.32380706e-01 -1.34060502e+00 -4.66454327e-01
6.28825784e-01 8.45037401e-01 -4.31291550e-01 3.37176085e-01
-1.04373097e+00 1.30559146e+00 3.67605180e-01 -5.43825217e-02
-9.25330639e-01 -9.01677847e-01 -2.85150975e-01 2.18246773e-01
-2.38522932e-01 -3.36761832e-01 7.37581074e-01 -8.61541450e-01
-1.28286242e+00 5.92200518e-01 -2.90159136e-01 -8.26657116e-01
5.67446761e-02 -6.08554110e-02 -6.66031301e-01 2.61854380e-02
-7.29331523e-02 3.86472106e-01 6.74129903e-01 -6.40518069e-01
-4.58653897e-01 -7.59667158e-01 -9.18618441e-01 -1.57676041e-02
-5.08564770e-01 2.69513756e-01 1.51006863e-01 -8.02496016e-01
1.11094095e-01 -9.54068959e-01 -2.50434875e-02 -7.26437867e-01
-1.57268882e-01 -1.92609355e-01 8.98868501e-01 -1.07822120e+00
1.23972273e+00 -1.96585274e+00 2.55741686e-01 2.13310689e-01
1.37701944e-01 1.39378205e-01 -1.63897082e-01 1.56209260e-01
-3.79490554e-01 -1.87767759e-01 -4.76929992e-01 1.11174650e-01
-2.81408221e-01 -1.30535379e-01 -2.83628672e-01 3.69168609e-01
2.18336612e-01 9.60360706e-01 -4.76905704e-01 2.27660194e-01
9.25690681e-02 6.12683952e-01 -3.66213650e-01 3.82075995e-01
2.90458411e-01 8.73644352e-01 -2.92285085e-01 5.18614054e-01
3.58917534e-01 -1.29818708e-01 -1.20972708e-01 -1.94793925e-01
2.86193788e-01 2.17091694e-01 -7.06977248e-01 1.55254483e+00
-3.12830180e-01 7.56993234e-01 -3.93365949e-01 -1.37854695e+00
9.46468651e-01 4.62305009e-01 8.33720982e-01 -7.97241449e-01
2.62074351e-01 5.66388488e-01 7.55118608e-01 -5.71149528e-01
-3.92319679e-01 -1.35425022e-02 1.45802144e-02 1.71463028e-01
2.74316132e-01 -1.25636950e-01 3.49752158e-02 -2.56316572e-01
1.37233639e+00 -3.59347463e-01 5.45212626e-01 -5.74782729e-01
5.67151010e-01 -3.58021379e-01 7.63423443e-01 4.57408845e-01
-1.60271004e-01 5.62941372e-01 2.45479137e-01 -5.73835373e-01
-3.40227753e-01 -8.70416284e-01 -6.15820706e-01 5.03124058e-01
-1.74205080e-01 -9.31012481e-02 -7.57469952e-01 -3.18043649e-01
-2.74158329e-01 7.14658678e-01 -6.15122795e-01 -5.34697235e-01
-5.00539482e-01 -1.52012360e+00 4.35759127e-01 6.54886901e-01
3.95659834e-01 -1.19109428e+00 -1.00385106e+00 4.70480144e-01
-9.16765854e-02 -8.74658644e-01 -1.62427157e-01 9.18036282e-01
-7.59371340e-01 -1.27783728e+00 -7.97163963e-01 -8.31665218e-01
3.09989005e-01 -6.78886890e-01 7.78016508e-01 -7.71702677e-02
-4.16833997e-01 3.13794017e-01 -3.14738333e-01 -5.58463097e-01
-8.72479528e-02 1.74914211e-01 4.69919145e-01 1.71688989e-01
7.79668212e-01 -8.66724551e-01 -8.14888418e-01 1.78637028e-01
-6.71452582e-01 -2.10093528e-01 6.63104653e-01 1.01535630e+00
6.29473984e-01 -2.61652563e-02 1.03286195e+00 -3.43331903e-01
9.05902445e-01 -5.38172185e-01 -2.63277441e-01 3.23643088e-01
-5.68241417e-01 6.15667105e-02 7.51580179e-01 -5.39198101e-01
-3.65526885e-01 3.08663428e-01 -1.63679510e-01 -2.54729301e-01
-3.18172455e-01 5.50310493e-01 -2.17717260e-01 -3.75069708e-01
5.17091274e-01 7.79330254e-01 -2.69735992e-01 -4.15116549e-01
-4.59270626e-01 8.05544794e-01 4.17264044e-01 -1.69522502e-02
2.84987211e-01 7.13280961e-02 1.58152267e-01 -8.37202549e-01
-3.87304515e-01 -5.82964301e-01 -7.14047492e-01 7.23406747e-02
9.03779626e-01 -1.00923443e+00 -3.07806969e-01 8.00327301e-01
-8.66834819e-01 -3.81483555e-01 -1.66349728e-02 9.51813936e-01
-6.89854383e-01 -9.09271836e-02 -3.14684570e-01 -5.27826846e-01
-7.18435347e-01 -1.43246150e+00 8.70049059e-01 1.21393502e-01
-4.46629971e-01 -9.83078539e-01 2.34318286e-01 -3.09125930e-02
5.88849723e-01 1.42452329e-01 1.05422449e+00 -1.48410892e+00
-3.14569324e-01 -1.80281684e-01 3.72258902e-01 2.83771902e-01
4.00814742e-01 -5.97860992e-01 -1.08453488e+00 -5.66019237e-01
3.09148759e-01 -1.85642987e-01 6.93963051e-01 7.10929394e-01
1.33692086e+00 -1.42709725e-02 -4.25964504e-01 8.85428965e-01
1.15270603e+00 7.33621120e-01 4.94886726e-01 3.09914529e-01
1.52433619e-01 2.22326547e-01 -1.23364046e-01 3.13660711e-01
1.91700235e-01 7.02844918e-01 1.74097583e-01 1.64491802e-01
-2.30880439e-01 2.85482466e-01 3.60073775e-01 1.04750419e+00
1.82518870e-01 -1.97846796e-02 -1.27996993e+00 6.96503997e-01
-1.52184165e+00 -8.30678046e-01 -9.17081628e-03 2.19113159e+00
5.35821676e-01 -1.61818311e-01 -6.67091310e-02 3.57547730e-01
1.95029199e-01 -3.95194530e-01 -6.95650160e-01 -1.64082795e-02
-4.11682904e-01 6.97605431e-01 3.43576461e-01 9.31757390e-02
-8.90984118e-01 6.29455984e-01 6.60588789e+00 6.48413837e-01
-1.74427450e+00 2.76539207e-01 4.99758005e-01 -4.55382675e-01
3.27895820e-01 -3.82868648e-01 -4.72407401e-01 7.52832353e-01
1.26518500e+00 -3.85305911e-01 5.40286958e-01 4.20150846e-01
5.12115359e-01 5.60513102e-02 -1.26193595e+00 1.32549238e+00
3.43409210e-01 -1.11467314e+00 -1.38991579e-01 -3.10905334e-02
6.63092196e-01 4.79830325e-01 5.56963086e-02 1.39681682e-01
-2.98299819e-01 -1.30770195e+00 3.81797582e-01 8.51909161e-01
5.54787517e-01 -7.41348147e-01 8.99824440e-01 5.66217184e-01
-1.14476252e+00 -4.41812068e-01 3.71599048e-02 1.44329891e-01
-5.21005318e-03 3.76231909e-01 -9.77523744e-01 2.03611404e-01
9.86784220e-01 9.50885952e-01 -9.77070391e-01 1.49731576e+00
-7.19345286e-02 8.63693595e-01 -3.25450063e-01 -2.35865545e-02
-1.51980221e-01 1.26273349e-01 5.81018507e-01 9.26668108e-01
1.02625120e+00 1.30773425e-01 -4.63910177e-02 5.89877009e-01
2.33297884e-01 2.83842653e-01 -3.67511004e-01 6.28568828e-02
2.88629949e-01 9.69775259e-01 -7.07870603e-01 -7.74502680e-02
-2.64612824e-01 7.76192904e-01 2.92724520e-01 2.70041853e-01
-4.29842651e-01 -3.97090465e-01 5.04572988e-01 -6.00525886e-02
-2.31109969e-02 1.65780902e-01 -7.11182058e-01 -1.29119897e+00
1.09991580e-02 -1.00589037e+00 1.64518163e-01 -7.39370584e-01
-1.10938942e+00 1.09199846e+00 -1.38254762e-01 -1.27361679e+00
-6.10371053e-01 -7.49101877e-01 -9.63814020e-01 9.61137295e-01
-1.42490649e+00 -1.02059793e+00 -2.34693423e-01 9.30621266e-01
4.96595711e-01 -8.89133453e-01 1.16607940e+00 1.83907643e-01
-5.34263790e-01 4.70633566e-01 2.67317623e-01 3.28054689e-02
5.58291972e-01 -1.03124964e+00 -2.22990081e-01 8.86367261e-01
3.26574594e-01 3.64083856e-01 5.89445055e-01 -5.40221989e-01
-1.06730640e+00 -1.00905418e+00 7.76366293e-01 -3.32763910e-01
7.26011455e-01 -2.32224166e-01 -1.05403376e+00 4.91242141e-01
1.61269695e-01 -1.07339583e-01 1.01777244e+00 -2.01753184e-01
2.87171096e-01 -3.51744771e-01 -1.03535354e+00 2.24361315e-01
6.28193021e-01 -5.55000961e-01 -8.07236433e-01 3.57710749e-01
-2.49165595e-02 -1.87846929e-01 -9.50358093e-01 6.01117373e-01
3.43561858e-01 -8.49706590e-01 6.01601243e-01 -4.20605958e-01
-4.84936573e-02 6.04213029e-02 -2.25758348e-02 -1.93567085e+00
-2.76169032e-01 -4.64345425e-01 -2.22788498e-01 5.32566130e-01
4.51229960e-01 -8.03264141e-01 7.61184692e-01 5.46074390e-01
-2.69762546e-01 -1.07479811e+00 -1.22807646e+00 -8.03990185e-01
3.60801041e-01 -5.68889976e-01 8.29052031e-01 5.79116166e-01
2.78674603e-01 2.56905751e-03 -1.89046100e-01 2.84467250e-01
3.74615818e-01 -7.81643987e-02 1.46588981e-01 -1.45465446e+00
-8.85735899e-02 -6.70013547e-01 -8.14726710e-01 -3.46713871e-01
3.62673551e-01 -1.21281040e+00 -7.81347305e-02 -1.40477908e+00
3.18966955e-01 -2.82233566e-01 -8.41084480e-01 7.24870026e-01
2.72187330e-02 3.26065302e-01 2.59692105e-03 1.81477115e-01
7.37298802e-02 5.09858727e-01 7.34212518e-01 -2.62551218e-01
-3.88226241e-01 3.18199039e-01 -6.00992858e-01 6.87289894e-01
1.06194770e+00 -5.83411992e-01 -4.95881796e-01 -2.55297959e-01
-1.93722814e-01 1.81598797e-01 3.15515488e-01 -1.48546648e+00
3.55634332e-01 6.29022941e-02 6.87977612e-01 -3.88477623e-01
3.34849983e-01 -8.86878431e-01 2.46236876e-01 8.44306469e-01
-1.62894562e-01 4.27986458e-02 2.55554080e-01 2.62265921e-01
-4.13946420e-01 9.28090960e-02 7.92376757e-01 1.95313111e-01
-6.64166927e-01 4.83890265e-01 -7.37992227e-01 1.14942409e-01
1.15656495e+00 -1.08744927e-01 1.24615535e-01 -2.95390725e-01
-9.15152311e-01 -1.73043907e-02 -1.04229622e-01 5.94793558e-01
7.50027716e-01 -1.20195234e+00 -8.72464180e-01 6.59790754e-01
2.07177803e-01 -5.30279100e-01 1.14438131e-01 1.10575545e+00
-2.47624904e-01 7.22342253e-01 -5.68827689e-01 -7.28025496e-01
-1.22415340e+00 5.03644422e-02 7.89405346e-01 -2.60659426e-01
-6.39606953e-01 9.04148281e-01 1.07331723e-01 -9.69997793e-02
2.44873166e-01 -5.48876822e-01 -4.95634764e-01 5.24629243e-02
5.79723299e-01 3.53856422e-02 7.16436923e-01 -7.26806164e-01
-6.92989171e-01 6.10235274e-01 8.44549909e-02 2.38910735e-01
1.78217399e+00 3.32309157e-01 -6.95098713e-02 3.87836874e-01
1.09128666e+00 -4.54204828e-01 -8.64907622e-01 1.88342884e-01
2.40290105e-01 1.56179711e-01 3.70386779e-01 -1.39170027e+00
-1.39592588e+00 1.05253386e+00 1.41165578e+00 -1.14223063e-01
1.55508196e+00 -8.56533572e-02 6.66025817e-01 4.69128042e-01
6.39456868e-01 -9.29738343e-01 -1.31376266e-01 3.83497685e-01
1.03661144e+00 -1.12340796e+00 -3.35353434e-01 2.98124552e-01
-5.16374111e-01 1.19419265e+00 4.27226812e-01 -2.06331328e-01
9.86635566e-01 4.98881161e-01 5.29685132e-02 -1.25868142e-01
-7.67324269e-01 5.35542853e-02 7.52246857e-01 7.74583995e-01
4.07206714e-01 2.94890732e-01 -3.79367948e-01 1.30612147e+00
-5.04916728e-01 1.20587021e-01 3.17316413e-01 6.24933183e-01
-2.13219061e-01 -1.10398293e+00 -2.28087887e-01 1.07883132e+00
-4.68831688e-01 -2.54394591e-01 -3.35978895e-01 5.34899235e-01
1.88049227e-01 8.45602453e-01 -7.34827444e-02 -5.24302959e-01
1.94689974e-01 5.12731493e-01 4.52707022e-01 -5.79192042e-01
-8.38233650e-01 -3.06594372e-02 -4.62167531e-01 -4.82508630e-01
-2.38450006e-01 -6.84722483e-01 -1.29858947e+00 5.36224484e-01
-1.18233502e-01 2.55524218e-01 5.91890454e-01 1.14722860e+00
5.85051060e-01 8.22588086e-01 5.32639205e-01 -7.05488741e-01
-3.99054468e-01 -1.22978592e+00 -9.48497832e-01 4.18263465e-01
4.56054002e-01 -6.42341614e-01 -4.97606277e-01 7.16279028e-03] | [13.222028732299805, 3.551008701324463] |
9cf1ff9f-b535-4abd-ac4e-f328e2678d89 | unbiased-scene-graph-generation-from-biased | 2002.11949 | null | https://arxiv.org/abs/2002.11949v3 | https://arxiv.org/pdf/2002.11949v3.pdf | Unbiased Scene Graph Generation from Biased Training | Today's scene graph generation (SGG) task is still far from practical, mainly due to the severe training bias, e.g., collapsing diverse "human walk on / sit on / lay on beach" into "human on beach". Given such SGG, the down-stream tasks such as VQA can hardly infer better scene structures than merely a bag of objects. However, debiasing in SGG is not trivial because traditional debiasing methods cannot distinguish between the good and bad bias, e.g., good context prior (e.g., "person read book" rather than "eat") and bad long-tailed bias (e.g., "near" dominating "behind / in front of"). In this paper, we present a novel SGG framework based on causal inference but not the conventional likelihood. We first build a causal graph for SGG, and perform traditional biased training with the graph. Then, we propose to draw the counterfactual causality from the trained graph to infer the effect from the bad bias, which should be removed. In particular, we use Total Direct Effect (TDE) as the proposed final predicate score for unbiased SGG. Note that our framework is agnostic to any SGG model and thus can be widely applied in the community who seeks unbiased predictions. By using the proposed Scene Graph Diagnosis toolkit on the SGG benchmark Visual Genome and several prevailing models, we observed significant improvements over the previous state-of-the-art methods. | ['Jiaxin Shi', 'Jianqiang Huang', 'Hanwang Zhang', 'Yulei Niu', 'Kaihua Tang'] | 2020-02-27 | unbiased-scene-graph-generation-from-biased-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Tang_Unbiased_Scene_Graph_Generation_From_Biased_Training_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Tang_Unbiased_Scene_Graph_Generation_From_Biased_Training_CVPR_2020_paper.pdf | cvpr-2020-6 | ['unbiased-scene-graph-generation'] | ['computer-vision'] | [ 4.00270551e-01 5.28213918e-01 -2.20839947e-01 -3.96090835e-01
-6.33732259e-01 -3.89671415e-01 8.73772442e-01 8.08279514e-02
1.07143797e-01 1.01135445e+00 7.33652830e-01 -4.95800108e-01
-8.91207829e-02 -1.07246375e+00 -1.25037754e+00 -7.38787055e-01
3.15621197e-01 5.69661260e-01 1.31803483e-01 -1.73727617e-01
2.24040419e-01 5.80829121e-02 -1.40953767e+00 2.75200188e-01
9.32014942e-01 5.88037789e-01 2.35903502e-01 4.21033412e-01
1.71173990e-01 1.09933901e+00 -7.03018904e-01 -9.03035700e-01
-1.81964993e-01 -6.03747845e-01 -8.38557720e-01 2.78146658e-02
5.95575988e-01 -6.41319931e-01 -3.83982867e-01 1.11881602e+00
3.42230320e-01 1.14321150e-01 8.73132765e-01 -1.39790702e+00
-8.96456182e-01 9.41453576e-01 -6.66097462e-01 8.25349838e-02
5.00443578e-01 3.08123529e-01 1.33841765e+00 -5.69907546e-01
7.20233619e-01 1.72234285e+00 3.38943332e-01 5.59907496e-01
-1.32199109e+00 -5.71406424e-01 6.52721226e-01 4.36512291e-01
-1.05029356e+00 -1.24594420e-01 8.44022870e-01 -4.86583710e-01
7.92729080e-01 5.65849781e-01 6.27832472e-01 1.82332087e+00
3.52868557e-01 9.72686589e-01 1.16371787e+00 -1.94150925e-01
2.18750983e-01 -2.10907355e-01 1.03466511e-01 6.07625306e-01
6.10862136e-01 2.77653068e-01 -7.71412373e-01 -1.87905505e-01
4.90550905e-01 -3.69760484e-01 -5.04815102e-01 -4.30219561e-01
-1.24942172e+00 8.85199904e-01 7.19527900e-01 -4.30463195e-01
-4.14774269e-01 5.93157232e-01 3.77154462e-02 -3.73613864e-01
6.05223477e-01 2.32215315e-01 -3.23949486e-01 2.22887680e-01
-7.83247113e-01 5.93733490e-01 6.75198793e-01 1.08668113e+00
6.55660212e-01 -1.54584691e-01 -5.73955417e-01 4.42800969e-01
5.20866275e-01 5.53203285e-01 -8.32476467e-02 -7.79760122e-01
4.57760066e-01 3.26521069e-01 1.63310319e-01 -1.20362079e+00
-1.75412074e-01 -3.98501486e-01 -1.02878976e+00 -4.69288528e-02
4.17142481e-01 -1.50254488e-01 -1.15465546e+00 1.88167238e+00
3.55553240e-01 3.34473193e-01 -2.06874654e-01 1.11116767e+00
9.86709774e-01 5.73279262e-01 2.59326041e-01 -6.84715286e-02
1.28890622e+00 -8.89771461e-01 -7.06653595e-01 -6.44681573e-01
4.34470832e-01 -3.62544417e-01 1.41291296e+00 4.92202759e-01
-7.11669445e-01 -2.05719501e-01 -8.95699501e-01 -1.40770316e-01
-1.55598477e-01 -9.27417129e-02 8.07057917e-01 5.49540818e-01
-8.79544377e-01 4.23581332e-01 -5.32056630e-01 -4.89986867e-01
6.84484065e-01 -1.62947834e-01 -9.29875746e-02 -4.90032971e-01
-1.22145510e+00 7.45070934e-01 4.28113312e-01 -9.84296054e-02
-1.45198464e+00 -8.47945213e-01 -8.24185491e-01 8.28946475e-03
8.48232925e-01 -1.21413183e+00 1.01971304e+00 -8.70059073e-01
-1.04623842e+00 8.62475097e-01 -3.34284723e-01 -5.42446733e-01
8.10959518e-01 -5.57670593e-01 -1.29916847e-01 -2.66214639e-01
4.91589755e-01 5.56304812e-01 9.25823510e-01 -1.64888430e+00
-4.91659492e-01 -4.55574811e-01 3.70355695e-01 4.17813480e-01
2.30636671e-01 -1.49264708e-01 -3.88232440e-01 -6.39018536e-01
-1.02739558e-02 -7.68762827e-01 -1.05317608e-01 -3.07569414e-01
-1.03248870e+00 -2.14208364e-01 5.24880886e-01 -7.72876620e-01
1.30613196e+00 -1.94576299e+00 6.71865642e-02 -7.09529147e-02
3.35059911e-01 -3.09262097e-01 2.42136195e-01 3.73846889e-01
-9.29180160e-02 3.98904949e-01 -2.19463050e-01 -1.31486550e-01
2.22275145e-02 3.65376532e-01 -6.35215819e-01 4.46850449e-01
9.21184793e-02 9.08885300e-01 -1.47246480e+00 -5.67781985e-01
1.39172271e-01 1.57091871e-01 -7.61656463e-01 1.76242799e-01
-5.79807281e-01 4.24657971e-01 -2.71737605e-01 4.85557020e-01
7.35590041e-01 -5.23745120e-01 2.59106874e-01 -3.58086199e-01
1.77468404e-01 5.32936215e-01 -1.07035649e+00 1.58006966e+00
-1.64125681e-01 3.97344351e-01 -4.10106391e-01 -7.21336424e-01
5.94756126e-01 2.35814089e-03 -2.00549632e-01 -3.88825029e-01
9.26211327e-02 -1.61410227e-01 -5.31915873e-02 -3.98703754e-01
6.14260375e-01 -2.69115210e-01 -1.01085134e-01 4.47230898e-02
-5.72760887e-02 -2.16631010e-01 -9.99933258e-02 6.71545267e-01
1.09910190e+00 5.03392398e-01 5.91097414e-01 -3.24714601e-01
8.33154321e-02 2.24632949e-01 5.38398564e-01 9.49369192e-01
-6.58652261e-02 7.94984698e-01 9.50074017e-01 -4.95085977e-02
-6.92978561e-01 -1.28334296e+00 3.19515735e-01 1.01729369e+00
5.85607886e-01 -4.40833211e-01 -6.30363882e-01 -1.11669743e+00
5.54196276e-02 1.56974661e+00 -8.01609516e-01 -2.91538507e-01
-2.74271101e-01 -9.30424213e-01 5.49915075e-01 5.45438468e-01
4.95041579e-01 -9.93155837e-01 -2.70823270e-01 7.97971860e-02
-5.67419887e-01 -9.53267992e-01 -2.62840599e-01 -2.03859046e-01
-5.13419449e-01 -1.10999978e+00 -3.13765794e-01 -1.51185051e-01
5.75918674e-01 3.42224449e-01 1.56766665e+00 1.70403719e-02
2.11818308e-01 2.56595671e-01 -3.36278439e-01 -7.09968686e-01
-2.74733573e-01 -4.14804369e-01 -2.97661841e-01 7.61710405e-02
1.06327616e-01 -2.52423346e-01 -8.67470205e-01 1.44964576e-01
-5.69442570e-01 4.96675640e-01 4.25938636e-01 9.48286474e-01
6.28505588e-01 1.04903810e-01 5.46381950e-01 -1.34069192e+00
4.30864483e-01 -7.43216395e-01 -3.79608393e-01 2.14132160e-01
-7.56124675e-01 -9.19561535e-02 4.67953503e-01 -1.86663747e-01
-1.57653844e+00 -4.57532734e-01 -9.41850320e-02 -3.60345900e-01
-1.61210701e-01 4.88842219e-01 -7.64000356e-01 7.06074715e-01
7.69693494e-01 -3.50684412e-02 -5.07342041e-01 -1.75804898e-01
7.07025766e-01 2.90043354e-01 5.79561114e-01 -6.99972451e-01
4.13520366e-01 8.22307706e-01 8.02443847e-02 -4.98311758e-01
-1.21032095e+00 -2.51417398e-01 -8.20395648e-02 -3.23687434e-01
1.02731168e+00 -9.31006193e-01 -4.91010994e-01 4.09398645e-01
-1.31397808e+00 -5.54013550e-01 -1.78476721e-01 2.57615805e-01
-4.79222238e-01 3.40657979e-01 -3.16996604e-01 -9.21729088e-01
-4.31515835e-03 -7.68308401e-01 1.25008368e+00 4.39782739e-02
-2.97008067e-01 -9.66781914e-01 -1.15694366e-01 6.12529874e-01
-1.18000351e-01 4.62593466e-01 9.58876669e-01 -3.45087349e-01
-7.79366493e-01 6.29964620e-02 -5.71788788e-01 2.73356080e-01
3.25101018e-02 9.29341167e-02 -1.03490210e+00 -1.11061491e-01
-2.19276696e-01 -2.05404580e-01 9.44390237e-01 5.73082268e-01
1.15887511e+00 -5.98838687e-01 -5.94728589e-01 4.07888621e-01
1.56841505e+00 -1.47189915e-01 8.83782625e-01 1.19119696e-01
1.27829862e+00 5.99139512e-01 9.53313112e-01 5.22221684e-01
6.84396684e-01 3.35719943e-01 9.93787050e-01 1.15201874e-02
-4.77785021e-01 -9.07409668e-01 2.97650844e-01 8.48843455e-02
-1.97631732e-01 -9.88959670e-01 -9.10249472e-01 8.19017291e-01
-1.93152249e+00 -8.76499772e-01 -7.45783269e-01 2.10096049e+00
6.18242621e-01 1.67836830e-01 -3.07455182e-01 -2.28942648e-01
7.13077962e-01 4.02332067e-01 -4.86142665e-01 -5.99933788e-02
-2.49414250e-01 -5.75396754e-02 7.11703300e-01 5.44764221e-01
-9.68805730e-01 1.12287652e+00 5.39475107e+00 9.62245524e-01
-6.97684765e-01 1.39550865e-01 8.20391119e-01 1.00239575e-01
-8.22205961e-01 5.43619573e-01 -8.85814190e-01 5.24971604e-01
3.93608302e-01 -1.69051304e-01 4.82979655e-01 7.70360827e-01
2.63045967e-01 -2.12099463e-01 -1.26202166e+00 8.56470644e-01
2.01974079e-01 -1.21388400e+00 5.71076989e-01 8.82826820e-02
8.80650640e-01 -2.01731965e-01 -2.14149177e-01 3.30583125e-01
9.61489141e-01 -1.14397454e+00 1.19359350e+00 4.36834663e-01
6.60593629e-01 -2.95661718e-01 5.53889573e-01 6.21863902e-01
-7.97886848e-01 2.57373661e-01 -3.67158949e-01 -1.84279084e-01
2.40554273e-01 8.52353990e-01 -8.90556991e-01 9.66071308e-01
7.80997574e-01 6.99978054e-01 -5.20713151e-01 7.10824907e-01
-9.18775916e-01 9.98993814e-01 2.47099460e-03 -9.10979658e-02
-1.92560107e-02 -1.33761764e-01 8.69550288e-01 1.07094014e+00
4.02892053e-01 8.05368051e-02 -1.44033164e-01 1.10694420e+00
-2.04685897e-01 -6.26800060e-02 -8.73164058e-01 3.34364474e-01
2.82685995e-01 9.19891536e-01 -5.26988983e-01 -5.91833293e-01
-3.64906788e-01 1.05261743e+00 3.50911856e-01 5.37406504e-01
-1.23652458e+00 2.85064548e-01 4.50544506e-01 3.53520423e-01
2.37131417e-01 3.54125351e-01 -5.38466454e-01 -1.23255217e+00
-2.31368378e-01 -8.48416328e-01 6.26993001e-01 -1.30511498e+00
-1.47718513e+00 1.17457427e-01 4.39356089e-01 -8.52281451e-01
-7.11935833e-02 -4.15309608e-01 -6.73869789e-01 9.25889313e-01
-1.23974645e+00 -1.43471920e+00 -5.06914198e-01 5.22228181e-01
5.06914794e-01 5.46768486e-01 4.16966110e-01 -6.96398243e-02
-4.57028031e-01 2.39694700e-01 -4.46573496e-01 -1.48125276e-01
8.31546903e-01 -1.56317163e+00 4.44680423e-01 1.30511355e+00
1.03320614e-01 5.41910589e-01 1.09131050e+00 -1.16152036e+00
-1.07829762e+00 -1.32048535e+00 9.74243581e-01 -7.92695343e-01
7.06858635e-01 -3.19867998e-01 -8.00519824e-01 1.03507996e+00
2.86344528e-01 -1.56478465e-01 1.88181177e-01 3.65581900e-01
-5.66901267e-01 -7.25233927e-03 -9.66389954e-01 9.57083404e-01
1.58506930e+00 -6.11166395e-02 -6.81894660e-01 3.89061421e-01
9.31064129e-01 -4.51119214e-01 -1.71392113e-01 4.75443572e-01
2.55575150e-01 -1.15521491e+00 9.04418945e-01 -7.15475321e-01
8.65313888e-01 -3.81836772e-01 -3.02140415e-01 -1.53902757e+00
-4.28379446e-01 -4.45580065e-01 -1.32658184e-01 1.40330541e+00
3.48837912e-01 -7.08574533e-01 7.73300290e-01 4.41310495e-01
-3.52385432e-01 -5.66087842e-01 -7.12916076e-01 -6.69968426e-01
-5.94936013e-02 -6.73669457e-01 7.86210299e-01 9.43319857e-01
-6.01787984e-01 6.09355271e-01 -6.94683850e-01 5.23300946e-01
9.85949814e-01 1.86026484e-01 1.00261223e+00 -9.79041100e-01
-6.03766024e-01 -1.40218168e-01 -8.54251683e-02 -1.17033255e+00
-1.50969848e-01 -7.78687179e-01 3.02782208e-01 -2.06114149e+00
5.44580936e-01 -1.78371310e-01 -6.34666681e-02 3.21668059e-01
-7.59854853e-01 -6.78728074e-02 -2.69820355e-02 -5.28833941e-02
-3.89284343e-01 5.78194916e-01 1.54643250e+00 -2.82906473e-01
1.79871827e-01 -2.01859429e-01 -1.17978871e+00 1.01502585e+00
5.91725826e-01 -5.17383993e-01 -6.95489526e-01 -3.19718003e-01
7.23815799e-01 7.45541155e-02 1.01135767e+00 -4.28364486e-01
-1.34064406e-01 -4.36896652e-01 6.12972528e-02 -5.53071201e-01
9.90913436e-02 -4.23153996e-01 3.24266583e-01 1.18591152e-01
-6.00509644e-02 -4.19453293e-01 -1.23986587e-01 1.02819908e+00
2.46241298e-02 -2.11128369e-02 3.82881552e-01 -1.81312531e-01
-7.70873725e-01 9.56163853e-02 -1.00318259e-02 2.72709787e-01
8.03521395e-01 2.23437157e-02 -8.41830134e-01 -5.50557315e-01
-5.44072747e-01 1.60797834e-01 5.08183300e-01 1.77012250e-01
5.87499201e-01 -1.17677605e+00 -8.72127533e-01 -2.73605734e-01
3.25393051e-01 3.43145937e-01 4.77819502e-01 7.77548254e-01
-2.09302709e-01 1.71386991e-02 2.34888464e-01 -5.65207541e-01
-1.18445241e+00 8.04496884e-01 1.51233673e-02 -5.10595381e-01
-4.89496082e-01 9.75112498e-01 1.06905866e+00 -7.11072832e-02
1.11397244e-02 -4.46683109e-01 -8.39975476e-02 -4.58548851e-02
5.10953330e-02 4.00135100e-01 -1.38920337e-01 -3.97231281e-01
-4.07494724e-01 4.41727275e-03 2.13340729e-01 -2.51200069e-02
8.27987552e-01 -3.12370241e-01 -1.43262953e-01 4.29751366e-01
6.45079017e-01 2.03852400e-01 -1.26199889e+00 2.23205805e-01
-3.31917107e-01 -6.80185735e-01 3.61208655e-02 -1.12200952e+00
-6.26288712e-01 7.56565213e-01 6.66396245e-02 3.07425290e-01
1.15839946e+00 4.60148156e-01 3.86335760e-01 -1.66681737e-01
5.65385103e-01 -7.13805735e-01 -4.98003699e-02 1.46966860e-01
1.24041474e+00 -1.25776684e+00 2.69791067e-01 -8.30663621e-01
-8.67467940e-01 2.85107195e-01 6.61155105e-01 -2.66964346e-01
3.55139524e-01 -2.35301703e-01 -2.73638606e-01 -6.00217462e-01
-9.08540964e-01 -3.31805021e-01 5.12838423e-01 6.06459796e-01
3.51798534e-01 5.49367726e-01 -3.88046324e-01 8.05864573e-01
-3.93694758e-01 -1.71559364e-01 6.43130422e-01 3.28237474e-01
5.87267764e-02 -6.31361306e-01 -3.71584892e-01 6.83923662e-01
-4.11226064e-01 -4.11962897e-01 -4.93610770e-01 1.01973844e+00
3.85578454e-01 1.11478603e+00 -2.15504229e-01 -3.03576201e-01
4.23016876e-01 -2.43341058e-01 5.90088189e-01 -7.18338847e-01
-2.28358388e-01 1.68436356e-02 4.78222489e-01 -7.51385808e-01
-4.91433650e-01 -7.26468623e-01 -9.44064617e-01 -3.35957855e-01
-3.30239624e-01 -3.29723448e-01 2.35796466e-01 8.21754038e-01
1.53166935e-01 7.72140265e-01 8.53942186e-02 -4.45975870e-01
-2.86248654e-01 -9.61119354e-01 -4.36255187e-01 7.45414913e-01
1.08031698e-01 -8.59374225e-01 -5.72000861e-01 2.56373912e-01] | [10.333995819091797, 1.8191927671432495] |
e0785a77-06ee-4113-aff5-c8894d641f0f | a-dataless-faceswap-detection-approach-using | 2212.02571 | null | https://arxiv.org/abs/2212.02571v1 | https://arxiv.org/pdf/2212.02571v1.pdf | A Dataless FaceSwap Detection Approach Using Synthetic Images | Face swapping technology used to create "Deepfakes" has advanced significantly over the past few years and now enables us to create realistic facial manipulations. Current deep learning algorithms to detect deepfakes have shown promising results, however, they require large amounts of training data, and as we show they are biased towards a particular ethnicity. We propose a deepfake detection methodology that eliminates the need for any real data by making use of synthetically generated data using StyleGAN3. This not only performs at par with the traditional training methodology of using real data but it shows better generalization capabilities when finetuned with a small amount of real data. Furthermore, this also reduces biases created by facial image datasets that might have sparse data from particular ethnicities. | ['Julian Togelius', 'Nasir Memon', 'Anubhav Jain'] | 2022-12-05 | null | null | null | null | ['face-swapping'] | ['computer-vision'] | [-1.20568303e-02 4.11417991e-01 2.54874304e-02 -5.76954365e-01
-3.29594672e-01 -4.80044842e-01 8.66470218e-01 -8.25629711e-01
-2.85585821e-01 8.69123459e-01 2.64409602e-01 -1.90867181e-03
3.03401798e-01 -9.90910590e-01 -5.67280233e-01 -6.45654857e-01
7.04673082e-02 4.27094311e-01 -1.31497711e-01 -4.62302715e-01
2.77554728e-02 7.22787619e-01 -1.80084991e+00 2.25076765e-01
5.45619547e-01 8.28774333e-01 -3.04042518e-01 3.97580266e-01
2.63689552e-02 5.01086771e-01 -9.88851190e-01 -6.18616939e-01
7.87119865e-01 -6.15541995e-01 -4.78320688e-01 6.99797273e-02
9.95133638e-01 -6.64248049e-01 -2.24866167e-01 8.61200988e-01
8.56610239e-01 -3.11465740e-01 1.37846485e-01 -1.29819715e+00
-4.56770748e-01 2.34075084e-01 -8.02499056e-01 -1.15844525e-01
2.60067731e-01 4.85743552e-01 2.35210299e-01 -7.53559113e-01
7.43619502e-01 1.55604219e+00 9.22514260e-01 1.15212381e+00
-1.18556273e+00 -1.45502365e+00 -4.99745250e-01 -2.71545082e-01
-1.44576836e+00 -8.84475112e-01 7.87081778e-01 -3.10798585e-01
5.43716192e-01 7.10225180e-02 9.51766372e-01 1.43891776e+00
-2.79887795e-01 4.93395865e-01 1.24319971e+00 -4.59242612e-01
-3.60246524e-02 3.25595766e-01 -5.82835317e-01 9.39155042e-01
1.92246288e-01 3.76850039e-01 -3.83705288e-01 -1.64173737e-01
9.86788094e-01 -3.47884119e-01 -1.33547723e-01 -2.64396906e-01
-8.24911833e-01 1.00869429e+00 1.66715235e-01 2.51719147e-01
-2.15423658e-01 2.50606149e-01 4.23954338e-01 4.78242189e-01
4.69422907e-01 4.79462266e-01 -1.48931593e-01 1.17286280e-01
-1.44668758e+00 2.43779927e-01 7.08586156e-01 5.74764431e-01
9.42563415e-01 6.81973994e-01 1.27004966e-01 8.64077330e-01
-4.41934355e-03 5.85751891e-01 7.52014875e-01 -1.05756688e+00
8.52867216e-02 5.15342534e-01 8.61726999e-02 -1.22418630e+00
-1.59816712e-01 -1.01374499e-01 -5.40302634e-01 7.98997223e-01
7.30995119e-01 -6.30652547e-01 -1.27216911e+00 1.84322071e+00
4.67565924e-01 1.75257117e-01 -1.61263019e-01 6.47108972e-01
7.30375886e-01 3.88936490e-01 -2.84011275e-01 2.55952299e-01
1.05980456e+00 -4.57269728e-01 -6.41816020e-01 -6.33749440e-02
5.54805100e-01 -7.41948485e-01 1.12039101e+00 4.15842831e-01
-9.85125422e-01 -3.65981609e-01 -9.81080413e-01 1.85454860e-01
-4.15412486e-01 5.94364367e-02 8.40169311e-01 1.58345401e+00
-1.46898258e+00 3.32941979e-01 -1.72869533e-01 -5.82027614e-01
9.57841158e-01 6.67462885e-01 -6.25891268e-01 4.11230838e-03
-1.08365500e+00 7.83734977e-01 6.67689294e-02 2.08273325e-02
-1.21387017e+00 -7.96973586e-01 -7.35746145e-01 -2.32405245e-01
3.02353203e-01 -3.89019072e-01 7.95091271e-01 -1.86063719e+00
-1.62232411e+00 1.27669656e+00 -2.80498099e-02 -2.85002857e-01
8.71807694e-01 -5.74272871e-02 -6.01926267e-01 4.55513364e-03
-2.83134699e-01 1.27036619e+00 1.21575224e+00 -1.13824236e+00
-1.86160460e-01 -2.76250601e-01 1.20467536e-01 -3.61099303e-01
-6.11079156e-01 2.65821457e-01 1.04946801e-02 -8.10298264e-01
-5.03445506e-01 -1.05553293e+00 1.25891402e-01 1.51283145e-01
-2.12230176e-01 1.26490831e-01 1.10856462e+00 -5.42243063e-01
8.09165120e-01 -1.96343184e+00 -3.70522738e-01 5.39919138e-01
1.34647995e-01 7.99395919e-01 -4.69722927e-01 2.55031914e-01
-3.18978280e-01 3.31495434e-01 4.92188931e-02 -1.81041900e-02
-3.79086941e-01 6.87399507e-02 -4.38341163e-02 4.49355572e-01
2.59708971e-01 6.61515772e-01 -7.15520382e-01 -2.25191072e-01
4.28846329e-02 5.41504860e-01 -7.59263635e-01 -5.67488447e-02
5.82319424e-02 3.44355792e-01 8.40125531e-02 7.71111131e-01
1.04193676e+00 2.18212411e-01 2.65194178e-01 3.13273631e-02
-1.75991841e-02 -2.00941324e-01 -9.99033749e-01 1.35880101e+00
-3.30183685e-01 8.32511127e-01 2.50772059e-01 -6.72279119e-01
1.08999002e+00 1.43360451e-01 3.13115090e-01 -6.92358375e-01
3.74690384e-01 2.46504217e-01 9.93385911e-02 -4.35897112e-01
2.60453671e-01 -3.73301357e-01 2.93922424e-01 4.72118556e-01
9.86635610e-02 -1.17411181e-01 -3.68921692e-03 6.25992287e-03
9.91331995e-01 7.10175261e-02 -1.34552330e-01 -4.17186111e-01
4.68765169e-01 -4.19748500e-02 6.59484029e-01 3.98854494e-01
-1.96677506e-01 7.00908065e-01 6.15469217e-01 -7.75214970e-01
-1.22324407e+00 -7.30670273e-01 7.11613297e-02 8.30061197e-01
-3.88240993e-01 -4.02058251e-02 -1.19727921e+00 -8.09778929e-01
1.81243226e-01 2.43039623e-01 -1.04524648e+00 -1.64068237e-01
-5.32436252e-01 -8.30013275e-01 1.18353617e+00 1.98587790e-01
8.64223242e-01 -1.01820910e+00 -4.78273392e-01 -1.92831114e-01
1.64067477e-01 -8.87120306e-01 -2.35434175e-01 -5.92038095e-01
-4.60852504e-01 -1.14367151e+00 -1.08727932e+00 -8.19555879e-01
7.77008653e-01 6.83327988e-02 1.22022724e+00 3.31537217e-01
-5.34855485e-01 9.05677602e-02 -1.15658261e-01 -6.14902675e-01
-4.92906332e-01 -1.84630647e-01 2.18630150e-01 3.90258431e-01
6.41676128e-01 -3.87893885e-01 -7.00139046e-01 3.29536527e-01
-7.83332109e-01 -2.00404674e-01 5.32966435e-01 8.98253202e-01
-1.41772211e-01 -1.06846221e-01 8.70888889e-01 -1.16862464e+00
5.63370049e-01 -3.71417791e-01 -5.95265865e-01 -1.29893906e-02
-3.46965551e-01 -2.43695095e-01 2.68360019e-01 -5.95146418e-01
-1.26744115e+00 -4.83602993e-02 -1.87333390e-01 -5.08641303e-01
-2.55071878e-01 -2.38407612e-01 -1.38651967e-01 -5.88089705e-01
8.99157703e-01 -1.83014199e-01 5.75493813e-01 -2.19962031e-01
1.30346686e-01 6.95195198e-01 2.70922959e-01 -3.56835812e-01
8.45633864e-01 8.16668868e-01 2.19072048e-02 -9.34546888e-01
-1.45325795e-01 3.05509657e-01 -3.91841680e-01 -4.74462032e-01
4.00292188e-01 -8.97733867e-01 -7.68073380e-01 9.15241063e-01
-6.66154206e-01 -4.83043045e-01 -2.53046036e-01 2.40162224e-01
-1.10214412e-01 6.79906234e-02 -3.39253753e-01 -7.18636036e-01
-1.56460211e-01 -8.74687254e-01 7.85061002e-01 2.70253152e-01
-5.31593084e-01 -8.40303004e-01 6.92291632e-02 2.99951404e-01
6.05154991e-01 6.63039684e-01 5.44599950e-01 -1.23729594e-01
-2.93103218e-01 -4.07299787e-01 -2.20497638e-01 5.10620296e-01
4.03790206e-01 4.13171113e-01 -1.31985044e+00 -3.93520832e-01
-1.84972763e-01 -6.33921444e-01 7.15722263e-01 1.68458745e-01
1.11870694e+00 -3.52425754e-01 -1.24073029e-01 6.18179440e-01
1.22022510e+00 8.15600306e-02 8.86564970e-01 1.72696784e-01
6.55639052e-01 1.06432068e+00 3.24990571e-01 3.55361164e-01
8.78353715e-02 4.50171679e-01 4.92620140e-01 -5.32663763e-01
-5.12553275e-01 -2.88017660e-01 2.98971355e-01 -4.54402305e-02
-1.70105487e-01 2.30294652e-02 -8.74543846e-01 6.42458558e-01
-1.24405229e+00 -1.02613449e+00 8.44518393e-02 2.12406111e+00
7.56075799e-01 -1.67007014e-01 5.72138190e-01 -3.94657115e-03
7.89359272e-01 4.97610085e-02 -4.85803872e-01 -5.67055643e-01
-2.97129899e-01 6.97928548e-01 5.86611092e-01 2.13028356e-01
-6.89327478e-01 1.17892957e+00 7.51773977e+00 6.43614411e-01
-1.59473717e+00 1.18025787e-01 7.93120682e-01 -5.69141865e-01
-3.65720391e-01 -3.56336772e-01 -5.99959195e-01 6.55082345e-01
7.65646577e-01 1.97107524e-01 4.34260398e-01 6.82137668e-01
3.69661570e-01 -1.41940907e-01 -7.71776497e-01 1.05830097e+00
3.56364846e-01 -1.32277286e+00 2.76299268e-01 3.06235015e-01
7.50110269e-01 -1.66795224e-01 3.27353567e-01 -5.57518639e-02
5.78914940e-01 -1.40062690e+00 5.70770144e-01 1.88484311e-01
1.06560016e+00 -9.13213074e-01 6.10823750e-01 5.32362871e-02
-4.54209596e-01 -1.46994933e-01 -3.46272409e-01 -8.49541575e-02
-4.26198244e-01 5.97628653e-01 -1.08616042e+00 -2.21196502e-01
8.38144243e-01 1.86948001e-01 -5.63894331e-01 7.34151959e-01
-1.90455437e-01 5.30351579e-01 -3.98689985e-01 8.11567605e-02
1.23566635e-01 -1.90340087e-01 1.55298293e-01 8.73460174e-01
5.87596238e-01 -3.56064916e-01 -4.30305779e-01 5.48862100e-01
-3.51054996e-01 -2.26548817e-02 -9.71490920e-01 -6.67421594e-02
3.86820018e-01 1.19849515e+00 -5.05232036e-01 -3.73667479e-01
-3.56983602e-01 8.40846896e-01 3.71013992e-02 1.78446412e-01
-5.38265824e-01 -1.78660884e-01 1.05892277e+00 4.31979001e-01
1.25589818e-01 1.75090656e-01 -8.21611285e-02 -9.49621737e-01
-3.77692759e-01 -1.30258679e+00 2.43873551e-01 -7.04699695e-01
-8.75032127e-01 6.29073262e-01 -2.50509858e-01 -9.46378171e-01
-3.34887266e-01 -6.33543789e-01 -6.35265350e-01 9.00912762e-01
-1.26057422e+00 -1.24468124e+00 -4.29780155e-01 7.94437826e-01
2.61697531e-01 -5.68291664e-01 9.30987835e-01 4.49169070e-01
-5.89190900e-01 1.07845771e+00 -1.89366549e-01 3.90583426e-01
1.00662196e+00 -7.51736045e-01 4.62145567e-01 7.62980402e-01
-1.23683095e-01 5.49938083e-01 6.07761502e-01 -6.58403933e-01
-1.03782427e+00 -8.02809179e-01 5.18357933e-01 -3.28106940e-01
2.67479569e-01 -5.75303257e-01 -6.42315567e-01 5.48189163e-01
3.38920504e-01 3.05741699e-03 7.56316721e-01 1.08898669e-01
-4.87749934e-01 -5.06178081e-01 -1.88087034e+00 6.22432292e-01
1.04494584e+00 -4.31705952e-01 -7.05643892e-02 2.24686280e-01
-1.86464757e-01 -1.45221978e-01 -6.02685511e-01 4.54361767e-01
9.98611569e-01 -1.30954945e+00 7.96484411e-01 -5.94262064e-01
2.80139714e-01 -1.16304271e-01 3.50366920e-01 -1.52764857e+00
-9.35075730e-02 -8.25323343e-01 1.96214885e-01 1.44954348e+00
2.78570205e-01 -8.33631217e-01 1.47583008e+00 6.48598850e-01
4.34284836e-01 -3.44729215e-01 -7.40492940e-01 -6.87501848e-01
1.49105877e-01 8.43215138e-02 1.07944167e+00 1.29885566e+00
-4.31732744e-01 -1.47548050e-01 -7.33793318e-01 -2.27792040e-01
5.56947231e-01 -2.37236962e-01 1.25612950e+00 -1.13686848e+00
7.29754893e-03 -3.82728010e-01 -6.44366205e-01 -3.43249649e-01
1.88293666e-01 -4.55420941e-01 -3.62540692e-01 -7.43638396e-01
-1.14059411e-01 -5.72138190e-01 1.15221515e-01 7.18569040e-01
2.03189671e-01 9.57407892e-01 7.75429606e-03 -1.68855622e-01
4.11869377e-01 1.63855553e-01 1.29869950e+00 3.32295075e-02
7.03598335e-02 -3.54438990e-01 -8.33917499e-01 8.47127676e-01
9.84741807e-01 -3.10940266e-01 -5.33815920e-01 -3.91602814e-01
2.41313577e-01 -4.94268745e-01 3.37122738e-01 -1.06982851e+00
-4.08196092e-01 -5.79602718e-02 9.02823329e-01 5.17526269e-02
3.92203212e-01 -6.18917286e-01 4.33833957e-01 5.58622897e-01
6.09125048e-02 -1.73546910e-01 2.72530019e-01 8.32879543e-02
-5.26917242e-02 -1.53812289e-01 1.24451447e+00 -4.51448500e-01
-7.80134261e-01 1.51289001e-01 -4.25148070e-01 -1.08737238e-01
1.22453237e+00 -5.44630110e-01 -1.57650098e-01 -5.94177902e-01
-3.52316916e-01 -1.12475187e-01 9.14199948e-01 4.84565645e-01
3.49557728e-01 -1.37055814e+00 -6.82269037e-01 7.80237794e-01
-1.39989763e-01 -4.15686846e-01 1.84058789e-02 3.52969170e-01
-9.81873572e-01 -7.31121451e-02 -7.57521510e-01 -2.92364866e-01
-1.58733296e+00 2.84371495e-01 5.15946448e-01 4.34371591e-01
-3.85880888e-01 1.08656967e+00 2.15613618e-01 -3.67799461e-01
-7.33857751e-02 2.28802711e-01 -1.83333039e-01 2.99631000e-01
7.09835768e-01 3.14261883e-01 1.52785718e-01 -7.75155485e-01
-2.24537224e-01 3.73531848e-01 -5.12373932e-02 -1.81879830e-02
1.25910449e+00 2.22492442e-01 8.34933072e-02 -1.93677559e-01
1.02964151e+00 4.38077718e-01 -1.34598482e+00 9.78760347e-02
-3.44402850e-01 -9.79483306e-01 -1.03748761e-01 -7.54179120e-01
-1.59168446e+00 8.12648475e-01 8.21373284e-01 1.54914051e-01
1.16195619e+00 -3.71026158e-01 7.93695807e-01 -1.75251509e-03
4.32840526e-01 -1.11551189e+00 1.32986888e-01 1.97988316e-01
7.21354127e-01 -1.19446182e+00 -1.69251338e-01 -3.30596507e-01
-4.91961539e-01 1.01040530e+00 6.66134238e-01 -1.43660754e-01
5.69815099e-01 3.94209087e-01 5.20869076e-01 -2.00309917e-01
-3.25979531e-01 -1.67910621e-01 -2.51134366e-01 8.73191059e-01
2.70261139e-01 -8.10188940e-05 8.50083679e-03 -3.18805158e-01
-5.03852367e-01 2.56520212e-01 5.26407778e-01 6.90501451e-01
-1.96018189e-01 -1.40701878e+00 -6.51108086e-01 5.50374508e-01
-7.50817001e-01 1.41954690e-01 -8.50565612e-01 1.04201150e+00
4.57454860e-01 7.02446461e-01 1.42627448e-01 -4.74833041e-01
7.54820853e-02 1.09587938e-01 7.26294696e-01 -4.57939506e-01
-6.98263228e-01 -2.61489421e-01 3.55715722e-01 -5.09845018e-01
-5.66183507e-01 -5.97636521e-01 -6.71877265e-01 -1.04614687e+00
-1.33782789e-01 -1.48480669e-01 6.31350219e-01 7.00188756e-01
4.55317110e-01 5.98223843e-02 7.89827168e-01 -9.80453968e-01
-1.36667462e-02 -9.80535269e-01 -5.19941092e-01 4.94043529e-01
2.24537835e-01 -6.99147820e-01 -2.66630262e-01 -8.90363380e-02] | [12.831899642944336, 0.651921272277832] |
22e17121-977b-43ea-bfda-c22dc4953a80 | spiking-neural-networks-for-visual-place | 2109.06452 | null | https://arxiv.org/abs/2109.06452v2 | https://arxiv.org/pdf/2109.06452v2.pdf | Spiking Neural Networks for Visual Place Recognition via Weighted Neuronal Assignments | Spiking neural networks (SNNs) offer both compelling potential advantages, including energy efficiency and low latencies and challenges including the non-differentiable nature of event spikes. Much of the initial research in this area has converted deep neural networks to equivalent SNNs, but this conversion approach potentially negates some of the advantages of SNN-based approaches developed from scratch. One promising area for high-performance SNNs is template matching and image recognition. This research introduces the first high-performance SNN for the Visual Place Recognition (VPR) task: given a query image, the SNN has to find the closest match out of a list of reference images. At the core of this new system is a novel assignment scheme that implements a form of ambiguity-informed salience, by up-weighting single-place-encoding neurons and down-weighting "ambiguous" neurons that respond to multiple different reference places. In a range of experiments on the challenging Nordland, Oxford RobotCar, SPEDTest, Synthia, and St Lucia datasets, we show that our SNN achieves comparable VPR performance to state-of-the-art and classical techniques, and degrades gracefully in performance with an increasing number of reference places. Our results provide a significant milestone towards SNNs that can provide robust, energy-efficient, and low latency robot localization. | ['Tobias Fischer', 'Michael Milford', 'Somayeh Hussaini'] | 2021-09-14 | null | null | null | null | ['template-matching', 'visual-place-recognition'] | ['computer-vision', 'computer-vision'] | [ 3.73967260e-01 -3.56551319e-01 8.84991363e-02 -1.66687384e-01
-6.61112130e-01 -5.15566647e-01 6.12373471e-01 1.17243282e-01
-8.09739947e-01 7.56528735e-01 -1.90828498e-02 2.01010585e-01
-2.18597814e-01 -7.08749592e-01 -1.03464639e+00 -7.56944239e-01
-2.14563742e-01 4.01484221e-01 6.83018565e-01 -4.20386374e-01
6.02856696e-01 8.64780903e-01 -1.91674280e+00 1.32504463e-01
3.34934831e-01 1.07131660e+00 7.04127908e-01 2.45823696e-01
1.72019541e-01 6.39178574e-01 -4.93095070e-01 4.69777048e-01
4.68711138e-01 -2.89978415e-01 -5.44476867e-01 -5.91300905e-01
3.39440346e-01 3.42275888e-01 -3.99671406e-01 1.13158429e+00
6.94821060e-01 4.77920294e-01 3.30126315e-01 -1.59384286e+00
-5.81750929e-01 4.31588978e-01 -3.65830094e-01 3.93548280e-01
2.38099337e-01 3.50147605e-01 6.95995331e-01 -8.27530801e-01
8.72641981e-01 1.26033223e+00 9.17680144e-01 5.70265353e-01
-1.45137691e+00 -7.61796057e-01 -8.24978575e-03 2.72656113e-01
-1.54435468e+00 -7.62902021e-01 5.06649673e-01 -2.50106025e-02
1.70821476e+00 -2.50085831e-01 7.34382510e-01 1.20815933e+00
4.18961793e-01 5.44470787e-01 9.20381784e-01 -7.09959641e-02
7.43860185e-01 -4.88618344e-01 -2.24296704e-01 3.66362244e-01
3.42257291e-01 5.03518760e-01 -9.18979824e-01 -1.08771078e-01
9.30825233e-01 7.12376758e-02 -2.16867402e-01 -7.12493539e-01
-1.52567053e+00 4.33045506e-01 1.11518037e+00 3.98776382e-01
-5.14311731e-01 7.36949682e-01 3.21053445e-01 1.12076014e-01
-3.90814930e-01 8.01442921e-01 -1.35286614e-01 -1.25371680e-01
-9.08096850e-01 4.06523675e-01 5.88172019e-01 9.16627586e-01
8.93243968e-01 2.47753799e-01 -9.60520729e-02 7.05751121e-01
2.25055069e-01 5.75791717e-01 6.44997478e-01 -1.23741210e+00
2.89691776e-01 3.94280434e-01 2.74004266e-02 -1.18484259e+00
-7.00043499e-01 -3.54500383e-01 -6.35279477e-01 6.66673422e-01
1.28196031e-01 2.73268968e-01 -1.27779388e+00 1.93636024e+00
-2.89776891e-01 1.75703153e-01 3.58920634e-01 9.95443404e-01
4.87235576e-01 5.85594594e-01 3.81478518e-02 4.02711213e-01
1.06143010e+00 -9.87870514e-01 -2.70708561e-01 -8.65979254e-01
2.23218337e-01 -3.10008645e-01 5.88492990e-01 1.30829722e-01
-1.03139675e+00 -3.82789999e-01 -1.47759295e+00 -2.04168469e-01
-6.82524860e-01 -3.23156685e-01 7.17602193e-01 9.52743366e-02
-1.55170298e+00 6.97198510e-01 -9.36653733e-01 -8.17053199e-01
5.43314815e-01 7.24144280e-01 -5.45155942e-01 -8.76457915e-02
-8.40022504e-01 1.19130158e+00 5.78810990e-01 -1.49634909e-02
-9.89139140e-01 -3.08762670e-01 -1.04846847e+00 1.75201610e-01
2.75354460e-02 -4.15787548e-01 1.18517745e+00 -9.31619346e-01
-1.11848879e+00 7.08920836e-01 -2.80002505e-01 -8.43383312e-01
-1.39586359e-01 3.20991784e-01 -2.53430456e-01 5.81586622e-02
4.14983392e-01 1.59739375e+00 3.02797765e-01 -1.18189871e+00
-7.11365402e-01 -4.30669218e-01 -2.12977156e-01 3.97363663e-01
2.83932090e-01 -7.26942793e-02 -1.81388110e-01 -4.63583052e-01
5.39313674e-01 -9.58000481e-01 -4.57505852e-01 2.96032965e-01
1.47487819e-01 1.02409702e-02 5.32623529e-01 -6.86711371e-02
2.77778894e-01 -2.37088847e+00 4.88717481e-02 2.68810183e-01
1.18236262e-02 -1.04139268e-01 -5.32581508e-01 2.55543590e-01
8.09232816e-02 -2.62232512e-01 -4.73788917e-01 -2.34255716e-02
1.14907943e-01 4.59697545e-01 -3.26179445e-01 5.73762238e-01
3.13118875e-01 9.56078112e-01 -1.21594870e+00 -1.07311174e-01
1.09658189e-01 4.67613816e-01 -3.51752609e-01 -4.60216850e-01
-1.43655375e-01 1.94627017e-01 1.15368098e-01 8.27149510e-01
5.17733812e-01 -1.63863003e-02 -5.81473969e-02 -7.77726024e-02
-4.09790397e-01 4.04246092e-01 -1.16287851e+00 2.12353158e+00
-1.09143905e-01 1.03254545e+00 1.18931748e-01 -7.76970446e-01
1.16038930e+00 -8.85506794e-02 3.69576871e-01 -1.37676680e+00
2.45672613e-01 6.56028330e-01 2.31125280e-01 2.64744520e-01
7.40066588e-01 3.97489071e-01 -2.92430162e-01 1.83893666e-01
4.01937008e-01 1.58645764e-01 2.40359306e-01 -9.94644091e-02
1.52380776e+00 2.91834950e-01 3.50029945e-01 -6.13767862e-01
-5.62818013e-02 2.96717197e-01 5.63702285e-01 9.54705119e-01
-5.40958464e-01 7.03856230e-01 1.30059808e-01 -5.44811666e-01
-1.02407861e+00 -1.29398882e+00 -5.87667413e-02 9.80162203e-01
6.33592606e-01 1.09374553e-01 -4.65275258e-01 9.93128046e-02
-5.43106114e-03 5.20086229e-01 -6.24121010e-01 -3.43823969e-01
-7.04635322e-01 -5.63298523e-01 9.23408449e-01 8.47357333e-01
8.01565886e-01 -1.61756277e+00 -1.52402067e+00 3.78048748e-01
-3.47001180e-02 -9.69440162e-01 -6.89123347e-02 1.22562408e+00
-7.84037948e-01 -6.89727187e-01 -5.75229764e-01 -1.44077635e+00
6.55207217e-01 3.98681074e-01 9.88476813e-01 -1.84649631e-01
-2.02599734e-01 -1.41984344e-01 -1.40744103e-02 -2.82050699e-01
1.93057612e-01 8.55832472e-02 2.26881087e-01 -5.51413715e-01
5.84218562e-01 -7.12509274e-01 -5.82178652e-01 2.76756287e-01
-6.56150520e-01 -2.64314294e-01 7.18297780e-01 8.74484241e-01
9.46105123e-01 -4.09464002e-01 5.83855271e-01 -2.21130714e-01
4.00486231e-01 -4.58632946e-01 -7.10961223e-01 -9.52635109e-02
-5.31849444e-01 3.78058374e-01 5.71978450e-01 -4.17362869e-01
-4.15587544e-01 5.33687532e-01 -8.63406211e-02 -3.80292118e-01
-2.27977931e-01 1.81256160e-01 1.84478506e-01 -6.20541692e-01
1.10923874e+00 4.90256697e-01 -2.00107262e-01 5.80603220e-02
8.07062984e-02 1.78535938e-01 1.08140767e+00 -4.56359148e-01
4.80594844e-01 5.75343251e-01 1.09038942e-01 -4.21171427e-01
3.47032882e-02 -4.95758861e-01 -3.70983839e-01 3.50438990e-02
5.39107978e-01 -8.60880196e-01 -7.60135055e-01 5.25632322e-01
-1.22321641e+00 -5.97977579e-01 -4.12591606e-01 1.97491139e-01
-8.88173223e-01 -1.73269257e-01 -3.98530126e-01 -5.71193039e-01
-1.81427136e-01 -1.20086873e+00 8.50584567e-01 5.38422585e-01
-2.74913698e-01 -4.24063653e-01 3.55749249e-01 -4.67906237e-01
7.52713740e-01 1.72378197e-01 4.83478904e-01 -7.53314793e-01
-7.83706069e-01 9.59211290e-02 -4.49888766e-01 -4.02135938e-01
-2.14700624e-01 -5.71098268e-01 -1.08198881e+00 -2.80006796e-01
-2.38246381e-01 -4.18244720e-01 1.09774446e+00 4.20591921e-01
7.63915241e-01 -3.93084511e-02 -8.24577868e-01 8.43942165e-01
1.74118936e+00 5.59066534e-01 9.35702860e-01 9.19772446e-01
2.73370117e-01 2.52875894e-01 2.97214299e-01 1.30836889e-01
4.40536171e-01 6.04610205e-01 8.51891398e-01 6.23295270e-02
-1.23347893e-01 -1.79632232e-01 2.85348922e-01 2.71623462e-01
2.31287137e-01 -1.63533777e-01 -1.05890739e+00 1.14600849e+00
-2.16170764e+00 -1.05298841e+00 4.41183329e-01 2.15220118e+00
5.03101587e-01 1.18396416e-01 -2.22066253e-01 -5.37221972e-03
6.74608886e-01 2.25565881e-01 -8.62433553e-01 -5.01215816e-01
-4.13875610e-01 3.66507769e-01 9.40929830e-01 1.08574979e-01
-1.02804840e+00 9.24822807e-01 6.57928419e+00 5.15565634e-01
-1.22643006e+00 -1.24318212e-01 1.08596109e-01 -1.29010469e-01
5.50933853e-02 -2.21613497e-01 -7.40474701e-01 3.48584950e-01
1.05783415e+00 1.92656275e-02 1.06546676e+00 9.03037786e-01
-3.49370390e-01 -4.33312386e-01 -1.30477417e+00 1.34881341e+00
2.33018830e-01 -1.65658998e+00 -3.49636734e-01 -1.69939175e-01
6.60665631e-01 8.94697607e-01 1.82902336e-01 3.45172286e-01
5.85515797e-01 -1.16793931e+00 1.05815411e+00 2.54592448e-01
4.92423385e-01 -6.85945094e-01 7.22670496e-01 1.30702749e-01
-1.38291764e+00 -3.04701120e-01 -6.74012363e-01 -2.71775305e-01
-7.17426389e-02 -3.74970958e-02 -7.41981328e-01 1.23949461e-01
1.09259033e+00 6.43947899e-01 -4.50247169e-01 1.53539562e+00
9.49004889e-02 -2.80180007e-01 -7.66266346e-01 -4.05370921e-01
5.48976600e-01 3.08818221e-01 5.07951736e-01 1.10885715e+00
4.41433579e-01 -8.43061656e-02 -1.93967484e-02 1.17139351e+00
-2.21032307e-01 -5.18178523e-01 -9.86538351e-01 3.67558569e-01
1.09909689e+00 1.13084102e+00 -1.19483423e+00 -6.49112165e-02
1.71799175e-02 1.04274809e+00 4.71366435e-01 2.73007512e-01
-6.13585651e-01 -6.91643894e-01 6.39156580e-01 -3.03351998e-01
5.04721642e-01 -3.26041311e-01 -6.99778140e-01 -5.31763732e-01
-4.25426774e-02 -5.51973462e-01 3.47326025e-02 -1.00922585e+00
-9.01462495e-01 8.13284397e-01 -3.80370945e-01 -1.02359819e+00
-3.38732868e-01 -6.65223956e-01 -4.55623955e-01 7.60363519e-01
-1.62704170e+00 -7.54783273e-01 -3.74715000e-01 5.16076744e-01
4.58433151e-01 6.06208015e-03 9.36500013e-01 2.75602669e-01
-1.05632074e-01 4.58941281e-01 1.82518691e-01 8.29488412e-02
5.47688782e-01 -1.04487932e+00 9.15967107e-01 9.20655310e-01
2.23375499e-01 5.71723998e-01 6.28302097e-01 -4.61803585e-01
-1.36724079e+00 -1.17969120e+00 8.00679862e-01 -3.80298644e-01
4.36270893e-01 -4.58303481e-01 -6.47606492e-01 5.61791539e-01
2.37516612e-01 2.92051375e-01 1.06653199e-01 -2.43398771e-01
-4.29799974e-01 -1.28604800e-01 -1.42647707e+00 8.67812216e-01
1.34236026e+00 -4.97056514e-01 -7.28227973e-01 -1.38908520e-01
4.35502589e-01 -4.67662066e-01 -2.85727382e-01 3.82152975e-01
6.23690486e-01 -8.78327191e-01 1.01069307e+00 9.62459892e-02
-4.12733033e-02 -7.47854948e-01 -5.43027103e-01 -1.44702101e+00
-5.88745296e-01 -3.24024081e-01 3.35863948e-01 7.32299089e-01
3.91577780e-01 -7.01100767e-01 7.05736518e-01 3.42541516e-01
-3.91283125e-01 -4.58407640e-01 -1.33085287e+00 -1.02810812e+00
-3.55165362e-01 -1.90175205e-01 3.46374720e-01 5.47087133e-01
5.98075874e-02 1.10495619e-01 2.24605247e-01 2.78677016e-01
6.31839752e-01 -1.14202395e-01 2.33114049e-01 -1.19834948e+00
1.49170354e-01 -6.12103522e-01 -1.01701272e+00 -8.28644931e-01
-1.56674851e-02 -9.81306076e-01 9.19830441e-01 -1.80561888e+00
2.18714308e-02 -4.51034307e-01 -6.05234861e-01 9.27045465e-01
6.56766057e-01 5.26591539e-01 2.61199534e-01 2.25857943e-01
-1.00133073e+00 3.27477515e-01 4.79352415e-01 -2.16030568e-01
-2.08687842e-01 -7.22854555e-01 -4.41561550e-01 5.43295741e-01
8.21119130e-01 -7.48055339e-01 -2.20622256e-01 -6.63451254e-01
1.95956960e-01 -4.24819767e-01 7.08717108e-01 -1.80699503e+00
9.80414152e-01 4.26627249e-02 6.56872094e-01 -5.73653698e-01
4.76513952e-01 -9.36190665e-01 2.38642260e-01 6.52488410e-01
-2.08586246e-01 4.71358567e-01 5.62750399e-01 5.41073740e-01
-7.30868103e-03 -1.95116345e-02 9.11505222e-01 -3.71128052e-01
-1.44219708e+00 -4.06140015e-02 -6.03889525e-01 5.81956729e-02
1.00492489e+00 -8.12044322e-01 -7.32179940e-01 2.32115626e-01
-2.61802733e-01 2.58346826e-01 7.98354089e-01 4.81231958e-01
8.12557280e-01 -1.36987734e+00 -1.99403346e-01 4.07703906e-01
1.59310743e-01 1.75943986e-01 -2.43181810e-01 6.78325891e-01
-6.54449701e-01 5.85694432e-01 -1.17597604e+00 -8.66321325e-01
-5.25386631e-01 5.26776254e-01 5.09219050e-01 2.49394789e-01
-5.80056548e-01 8.67603898e-01 7.88625970e-04 -5.61570346e-01
4.72827047e-01 -4.11659628e-01 8.58839322e-03 -1.38083056e-01
2.82878339e-01 3.14600408e-01 2.94494659e-01 -6.69510067e-01
-8.46778810e-01 3.58243883e-01 2.43782163e-01 -3.56659800e-01
1.39681554e+00 9.29522961e-02 -7.93325081e-02 3.72837335e-01
8.62690687e-01 -6.30183697e-01 -1.42037344e+00 2.43108105e-02
3.67969304e-01 -7.04499483e-02 -1.92768394e-03 -8.05590034e-01
-8.71441543e-01 6.78820610e-01 8.71570647e-01 -3.32709700e-01
1.09520042e+00 9.60608572e-03 7.20437050e-01 9.20482993e-01
9.92164552e-01 -1.09943843e+00 -1.87053252e-02 9.64862168e-01
8.17340553e-01 -9.57678795e-01 -5.72062492e-01 3.31364125e-01
-3.25809062e-01 9.13812518e-01 8.58869970e-01 -7.75432587e-01
9.45629552e-02 7.28102565e-01 -8.92751366e-02 -2.55621225e-01
-7.97088742e-01 -3.10237288e-01 -1.78984910e-01 1.10934246e+00
-2.30469972e-01 -2.13384449e-01 2.39632562e-01 4.87991899e-01
1.49497500e-04 -1.35927997e-03 2.98664033e-01 1.18628538e+00
-8.98410618e-01 -5.99966645e-01 -3.51070702e-01 2.71407634e-01
-1.17601082e-01 -3.60561967e-01 -2.36308068e-01 6.33303523e-01
6.67120442e-02 5.93530357e-01 3.18777978e-01 -4.71403807e-01
3.90993595e-01 -4.57571782e-02 5.39420247e-01 -4.33162630e-01
-7.52516806e-01 -2.95815766e-01 -1.34435415e-01 -1.00717235e+00
-3.59577507e-01 -6.63532257e-01 -1.75140226e+00 -9.47221443e-02
3.35454717e-02 -2.42698789e-01 8.73126090e-01 7.29451120e-01
6.75122976e-01 5.56457579e-01 3.21655683e-02 -1.61784041e+00
-2.59337574e-01 -5.33684611e-01 -2.00327173e-01 5.35670016e-03
4.94210243e-01 -8.60756159e-01 -1.45154953e-01 -2.11118177e-01] | [7.666865348815918, -1.7744299173355103] |
942ce375-c716-4696-b963-2bc50dd94c8c | mononeuralfusion-online-monocular-neural-3d | 2209.15153 | null | https://arxiv.org/abs/2209.15153v1 | https://arxiv.org/pdf/2209.15153v1.pdf | MonoNeuralFusion: Online Monocular Neural 3D Reconstruction with Geometric Priors | High-fidelity 3D scene reconstruction from monocular videos continues to be challenging, especially for complete and fine-grained geometry reconstruction. The previous 3D reconstruction approaches with neural implicit representations have shown a promising ability for complete scene reconstruction, while their results are often over-smooth and lack enough geometric details. This paper introduces a novel neural implicit scene representation with volume rendering for high-fidelity online 3D scene reconstruction from monocular videos. For fine-grained reconstruction, our key insight is to incorporate geometric priors into both the neural implicit scene representation and neural volume rendering, thus leading to an effective geometry learning mechanism based on volume rendering optimization. Benefiting from this, we present MonoNeuralFusion to perform the online neural 3D reconstruction from monocular videos, by which the 3D scene geometry is efficiently generated and optimized during the on-the-fly 3D monocular scanning. The extensive comparisons with state-of-the-art approaches show that our MonoNeuralFusion consistently generates much better complete and fine-grained reconstruction results, both quantitatively and qualitatively. | ['Hongbo Fu', 'Ying Shan', 'Tai-Jiang Mu', 'Yan-Pei Cao', 'Shi-Sheng Huang', 'Zi-Xin Zou'] | 2022-09-30 | null | null | null | null | ['3d-scene-reconstruction'] | ['computer-vision'] | [-1.25754505e-01 8.47566202e-02 3.08802575e-01 -4.86098647e-01
-3.22168320e-01 -3.04669440e-01 5.85877120e-01 -3.81704062e-01
-1.99675217e-01 7.24416077e-01 1.54368147e-01 -4.25677225e-02
8.83715972e-02 -1.04955566e+00 -1.02108896e+00 -6.69258654e-01
8.54698271e-02 5.54689407e-01 3.76215838e-02 -1.24442190e-01
2.89871097e-01 9.25413132e-01 -1.83181751e+00 2.09370032e-01
6.34893477e-01 1.05899572e+00 4.67714161e-01 6.78020954e-01
-5.13599515e-01 7.61059225e-01 -1.00727782e-01 -2.05420315e-01
4.45468783e-01 -2.34857023e-01 -6.76531136e-01 1.05925851e-01
8.36337030e-01 -7.61179864e-01 -6.53157413e-01 8.74803483e-01
5.03403366e-01 9.90203768e-02 5.81815243e-01 -6.40240014e-01
-4.88555193e-01 4.08877172e-02 -3.30338240e-01 -1.71062440e-01
3.40918362e-01 -2.59159058e-02 6.64251566e-01 -1.36484361e+00
9.91031945e-01 1.46473527e+00 5.81610799e-01 5.46650469e-01
-1.16356802e+00 -4.51233327e-01 3.08312148e-01 -1.94205537e-01
-1.30454767e+00 -4.45432812e-01 9.01174128e-01 -4.72335428e-01
9.66291189e-01 2.67124444e-01 1.18067420e+00 3.23359430e-01
2.11386308e-01 6.68337762e-01 1.06375062e+00 -1.07553788e-01
1.90166295e-01 -1.72155306e-01 -2.89221495e-01 1.02870929e+00
1.96918827e-02 6.85395122e-01 -6.62181556e-01 9.16816071e-02
1.86042500e+00 2.40617082e-01 -6.21957421e-01 -6.83211982e-01
-1.06282067e+00 7.24766254e-01 7.68614590e-01 3.08801774e-02
-4.05201375e-01 6.88763380e-01 2.15990469e-01 1.89562008e-01
7.68732309e-01 4.07965481e-01 -3.71793240e-01 -3.02138962e-02
-1.04405081e+00 5.44327497e-01 6.34413540e-01 9.55501437e-01
1.02463055e+00 4.99312788e-01 1.16862953e-01 6.54714644e-01
4.32813436e-01 3.89658123e-01 8.52983892e-02 -1.38903224e+00
6.37799725e-02 6.25344455e-01 -9.06395912e-02 -9.88715470e-01
-2.24412918e-01 -5.76625288e-01 -9.34158862e-01 6.44990206e-01
1.43082498e-03 3.48782420e-01 -1.03445876e+00 1.38209450e+00
6.27804518e-01 2.10716888e-01 -2.95836300e-01 1.23247612e+00
1.47091138e+00 5.37177861e-01 -4.50998098e-01 8.57838616e-03
8.21799040e-01 -7.89892495e-01 -5.40422976e-01 2.71716654e-01
4.82327104e-01 -4.87352759e-01 9.18381751e-01 3.31896931e-01
-1.41619873e+00 -3.99071842e-01 -7.33244002e-01 -5.37277520e-01
-7.11074704e-03 -1.46784425e-01 8.10797989e-01 2.96244085e-01
-1.25518179e+00 7.57206798e-01 -7.54331172e-01 1.41483471e-01
6.59538209e-01 2.98973978e-01 -6.04996979e-01 -2.59561181e-01
-7.28286445e-01 5.02107263e-01 1.14632018e-01 -1.25993066e-03
-1.07423079e+00 -1.09155035e+00 -1.07185853e+00 -2.15918705e-01
9.41174403e-02 -1.27417135e+00 1.05196357e+00 -5.59905767e-01
-1.73220897e+00 1.22112823e+00 1.47154871e-02 -2.94079423e-01
4.74555194e-01 -2.94108659e-01 7.44141042e-01 3.40319037e-01
-2.44962007e-01 1.00550592e+00 7.11492062e-01 -1.54315639e+00
8.05618893e-03 -5.19469798e-01 3.22114170e-01 6.17915571e-01
3.28958392e-01 -5.06789744e-01 -4.98721361e-01 -6.37986064e-01
6.17152154e-01 -5.14409363e-01 -3.67347151e-01 7.55936086e-01
-2.74537146e-01 3.13100547e-01 7.02851951e-01 -7.78027117e-01
4.42072988e-01 -1.75306523e+00 5.86035550e-01 -3.27634364e-01
4.44206059e-01 -1.20627642e-01 9.23306644e-02 7.84160048e-02
2.70368368e-03 -3.23817842e-02 -3.37218165e-01 -8.14382970e-01
-2.69791782e-01 3.39328766e-01 -4.05336618e-01 6.47836208e-01
-2.32735649e-02 1.18591964e+00 -8.02477956e-01 -3.16127062e-01
6.50067389e-01 1.03389418e+00 -1.08169663e+00 5.92460036e-01
-2.41592616e-01 9.30125177e-01 -3.00962180e-01 7.69718468e-01
8.26245248e-01 -2.64712393e-01 -2.90390462e-01 -3.66496325e-01
-3.30693275e-01 2.66060710e-01 -7.98965514e-01 2.36084890e+00
-4.07556683e-01 6.14077687e-01 5.27737498e-01 -8.16815138e-01
8.93925011e-01 1.70700565e-01 5.64697504e-01 -8.52720022e-01
3.25134099e-01 8.21848884e-02 -6.82700574e-01 -1.92880947e-02
6.03653789e-01 -4.61346596e-01 2.44605362e-01 3.23248893e-01
1.48490727e-01 -1.23965907e+00 -3.92987937e-01 1.62712738e-01
5.58843553e-01 7.55819559e-01 1.43975198e-01 -2.37492502e-01
3.72956336e-01 -7.50696063e-02 2.74750173e-01 3.90212685e-01
2.73526758e-01 1.04815316e+00 -3.45208906e-02 -6.96086466e-01
-1.19260299e+00 -1.13135755e+00 -2.48853743e-01 3.71017903e-01
3.23874801e-01 -5.67014702e-02 -7.44522750e-01 -5.27203502e-03
-9.02410075e-02 3.29930276e-01 -6.40376687e-01 2.05357417e-01
-6.56356573e-01 -3.02650094e-01 1.38977945e-01 2.13910207e-01
5.68623662e-01 -1.12666774e+00 -7.38643110e-01 2.59227276e-01
-4.43154797e-02 -1.08489454e+00 -1.12849362e-01 2.24690102e-02
-1.47472322e+00 -9.33807909e-01 -9.98227596e-01 -5.21746576e-01
7.74789274e-01 5.20774364e-01 1.34221780e+00 4.08477783e-01
-5.87708890e-01 3.08138132e-01 3.90033908e-02 1.34051393e-03
-1.08682759e-01 -4.50605959e-01 -1.72993198e-01 -2.94916749e-01
-5.24373889e-01 -1.04576671e+00 -7.72518694e-01 3.32528740e-01
-8.48986328e-01 6.99474335e-01 1.72966599e-01 8.66170943e-01
1.22690713e+00 -3.62994522e-01 7.06894770e-02 -9.97894764e-01
1.19087711e-01 -1.86222181e-01 -8.62802386e-01 -2.49438256e-01
-1.58401236e-01 -5.38860075e-02 5.52334309e-01 -5.93483485e-02
-1.19503891e+00 5.97503874e-03 -5.46026766e-01 -1.07201529e+00
-1.96812913e-01 3.22726399e-01 7.73194134e-02 -6.49182022e-01
5.29210627e-01 6.01143003e-01 -2.88200118e-02 -6.31563127e-01
5.23114741e-01 -8.01994652e-02 5.34276843e-01 -4.40614194e-01
8.30394328e-01 9.69484091e-01 1.57347843e-01 -9.05983686e-01
-8.77710402e-01 -2.30239481e-01 -1.01119328e+00 -3.35383207e-01
8.84283304e-01 -1.15064740e+00 -6.15748882e-01 7.17082679e-01
-1.24617600e+00 -8.74360561e-01 -6.06956780e-01 4.21113372e-01
-1.12433195e+00 4.68162477e-01 -8.63706827e-01 -8.17972839e-01
-4.41676944e-01 -1.28905547e+00 1.42121327e+00 -1.47892460e-01
1.24485411e-01 -9.79045808e-01 9.18534398e-02 2.56709307e-01
1.84325472e-01 5.98844767e-01 7.88711131e-01 7.03092217e-01
-1.40114129e+00 2.21299455e-01 -5.10209382e-01 1.07245401e-01
-3.97938192e-02 -1.61577433e-01 -1.15608907e+00 -1.31022289e-01
2.31934577e-01 -3.73455763e-01 7.53481567e-01 6.81885183e-01
1.52422643e+00 -1.04839377e-01 3.26926783e-02 1.60111833e+00
1.47060907e+00 -1.91829905e-01 7.74067163e-01 -3.29431817e-02
1.10070634e+00 6.74588501e-01 5.61817408e-01 4.21810180e-01
4.34278995e-01 6.87030375e-01 9.74380851e-01 -2.66397625e-01
-5.03728390e-01 -3.33106577e-01 -4.00410034e-02 1.11904681e+00
-2.67308384e-01 5.00941277e-02 -2.58314967e-01 2.92818278e-01
-1.50283217e+00 -7.41602004e-01 -9.48433578e-02 2.05902863e+00
6.36694431e-01 -3.02771717e-01 -2.30638862e-01 1.53949738e-01
2.53501713e-01 2.82886058e-01 -7.24033237e-01 -3.28315973e-01
-2.89728552e-01 9.38516557e-02 2.95040518e-01 6.22808397e-01
-6.52248800e-01 1.12963116e+00 6.66531181e+00 6.64824605e-01
-1.21786165e+00 1.46462470e-01 6.35968268e-01 -4.03760344e-01
-7.66179800e-01 -1.69541091e-01 -6.14006162e-01 1.55826034e-02
3.18306237e-01 1.29454926e-01 6.45255506e-01 6.41361356e-01
1.72326684e-01 -1.07521102e-01 -9.06089783e-01 1.63082206e+00
-1.34355268e-02 -2.00486159e+00 4.16307807e-01 3.04168910e-01
9.61658239e-01 3.31748426e-01 -7.42083862e-02 -1.19109917e-02
2.28080258e-01 -1.30248547e+00 1.20262825e+00 7.66124308e-01
1.25840533e+00 -6.89520538e-01 1.46083221e-01 5.58855832e-01
-1.22049117e+00 2.94947952e-01 -7.22280860e-01 -1.09079204e-01
5.94002187e-01 8.79733503e-01 -3.47778738e-01 4.69896674e-01
8.14243317e-01 1.02221370e+00 1.26268402e-01 1.11058760e+00
-3.83250937e-02 -4.00308594e-02 -1.95528209e-01 2.55711198e-01
1.20597154e-01 -4.38935101e-01 5.92167795e-01 8.48848403e-01
4.16734278e-01 4.75161523e-01 -5.99667057e-02 1.42482626e+00
-2.63467640e-01 2.00246111e-01 -9.33934033e-01 -1.77409481e-02
5.21014584e-03 1.04922152e+00 -5.43062806e-01 -2.18680754e-01
-1.50686055e-01 8.81554425e-01 7.06122518e-01 3.05553645e-01
-5.21101415e-01 1.72693148e-01 7.59605885e-01 3.68366480e-01
1.90944418e-01 -7.36880958e-01 -5.93635082e-01 -1.36176157e+00
-2.90548116e-01 -1.05784938e-01 -2.66321987e-01 -1.34655058e+00
-9.81207669e-01 7.50180721e-01 -6.48713037e-02 -8.82357001e-01
-1.25029430e-01 -7.14549601e-01 -3.13990265e-01 9.51563537e-01
-1.84033537e+00 -8.74784946e-01 -6.93199456e-01 7.60640979e-01
5.64634979e-01 5.49966320e-02 6.82445467e-01 2.65534908e-01
1.71864063e-01 1.80520162e-01 -3.59599292e-01 -1.81412354e-01
1.00464240e-01 -9.23887908e-01 6.58829153e-01 4.84967798e-01
8.53939950e-02 3.80800366e-01 4.18812215e-01 -8.06939065e-01
-1.75488567e+00 -1.20740974e+00 2.38647521e-01 -2.84401029e-01
-1.15823336e-01 -4.15408701e-01 -9.80192482e-01 3.77632260e-01
-2.13961348e-01 2.63902545e-01 3.37508708e-01 -3.74504864e-01
-2.26392388e-01 2.69869298e-01 -1.40820229e+00 6.65856481e-01
1.70768952e+00 -7.31832027e-01 -2.91976333e-01 2.83847094e-01
9.86133039e-01 -7.77669668e-01 -9.02449548e-01 4.09280241e-01
5.83133280e-01 -1.42034602e+00 1.29119110e+00 -1.31794317e-02
6.96041405e-01 -1.49064258e-01 -4.16333437e-01 -1.18954241e+00
-1.84882849e-01 -5.89129984e-01 -4.09405470e-01 4.35664356e-01
-1.73005894e-01 -5.50364614e-01 1.08149230e+00 2.62744635e-01
-6.28656745e-01 -1.08470821e+00 -8.80484164e-01 -3.85337591e-01
3.46139930e-02 -5.91736674e-01 6.39482439e-01 7.83472359e-01
-7.12701142e-01 -1.82998389e-01 -3.59085709e-01 -1.45797953e-01
8.85949016e-01 4.80454773e-01 9.00350809e-01 -1.28882694e+00
-3.04359287e-01 -2.89913982e-01 -4.60303128e-01 -1.90519822e+00
1.33297920e-01 -9.73760247e-01 1.02369010e-01 -1.68807328e+00
-2.92814877e-02 -7.05330729e-01 3.64410609e-01 -3.37851271e-02
4.45053071e-01 7.24878967e-01 1.41861409e-01 2.71127820e-01
-1.04532279e-01 1.05219400e+00 2.12566113e+00 3.10513616e-01
-1.43755525e-01 -3.75232428e-01 -3.42916638e-01 8.22579265e-01
2.57837355e-01 -1.60909638e-01 -5.41061699e-01 -8.17870378e-01
3.31199855e-01 6.02980912e-01 6.70643151e-01 -6.72764480e-01
-5.75837381e-02 -2.26270497e-01 4.13480401e-01 -1.21949315e+00
1.07803750e+00 -5.90963185e-01 2.25581646e-01 3.07520628e-01
2.94617005e-03 -1.42281696e-01 1.59825966e-01 5.12115002e-01
-2.28702083e-01 9.59306657e-02 9.71503675e-01 -7.34621286e-01
-6.59255266e-01 1.11756742e+00 -6.07623756e-02 8.79210979e-02
5.03269076e-01 -6.41160607e-01 1.16614074e-01 -4.42589521e-01
-8.93600464e-01 -3.65629047e-01 9.53510821e-01 1.05423652e-01
1.34175646e+00 -1.38156617e+00 -6.19990826e-01 5.64695418e-01
-2.30357245e-01 7.84251630e-01 5.65423131e-01 5.22938192e-01
-1.18457460e+00 5.42083800e-01 -3.60030353e-01 -1.01845741e+00
-1.06156743e+00 1.59679443e-01 8.15295458e-01 3.57342474e-02
-1.04358399e+00 1.26050651e+00 1.18032455e+00 -1.09084642e+00
9.58658010e-03 -5.59032023e-01 -3.32409665e-02 -5.62621415e-01
4.24232632e-01 1.03967048e-01 1.79747492e-02 -8.40421736e-01
5.31981997e-02 1.10625780e+00 4.26151484e-01 -1.23677716e-01
1.63919759e+00 -2.71611452e-01 -1.24553390e-01 4.40474182e-01
1.21328509e+00 -2.71761864e-01 -1.78516281e+00 -1.79669037e-01
-9.27622080e-01 -1.02911222e+00 5.15938103e-01 -3.27944785e-01
-1.56371057e+00 1.38719702e+00 1.08675659e-01 -4.33784753e-01
8.38041544e-01 -1.17320284e-01 7.72072136e-01 1.72291294e-01
1.04521179e+00 -4.09543812e-01 1.04406320e-01 8.08493137e-01
1.35748041e+00 -8.36936414e-01 2.98718303e-01 -9.42642212e-01
-4.67607230e-02 1.16340494e+00 4.49628979e-01 -6.48179471e-01
8.89712155e-01 1.53984517e-01 -1.93000481e-01 -6.18127525e-01
-6.11609578e-01 2.75202811e-01 4.49164063e-01 5.78730524e-01
4.60585564e-01 5.16956188e-02 3.38333964e-01 -9.92131904e-02
-4.71583396e-01 -7.00071529e-02 5.08954048e-01 5.70123076e-01
-3.08040470e-01 -6.92039907e-01 -1.28094330e-01 2.64521480e-01
-1.18774079e-01 -1.29302427e-01 -2.08759978e-01 5.89982629e-01
-7.24578500e-02 1.59977928e-01 3.00279886e-01 4.39855978e-02
2.18073070e-01 -5.84203362e-01 1.16254961e+00 -8.91043246e-01
-4.63639021e-01 1.72116160e-01 -7.43557960e-02 -1.05537057e+00
-3.06676269e-01 -4.48446780e-01 -1.52519774e+00 -4.12441611e-01
-1.32886231e-01 -2.65684396e-01 9.81520057e-01 7.92656481e-01
2.60632008e-01 4.59250689e-01 3.63957644e-01 -1.82370782e+00
3.31944674e-02 -3.75356615e-01 -7.16684043e-01 1.66384727e-01
3.46205771e-01 -9.83373284e-01 -2.63081014e-01 -1.25933766e-01] | [9.01490592956543, -3.1224000453948975] |
0165520c-3a65-4f52-9590-52e8a7d4a5a6 | 07120194 | 0712.0194 | null | http://arxiv.org/abs/0712.0194v1 | http://arxiv.org/pdf/0712.0194v1.pdf | Computing High Accuracy Power Spectra with Pico | This paper presents the second release of Pico (Parameters for the Impatient
COsmologist). Pico is a general purpose machine learning code which we have
applied to computing the CMB power spectra and the WMAP likelihood. For this
release, we have made improvements to the algorithm as well as the data sets
used to train Pico, leading to a significant improvement in accuracy. For the 9
parameter nonflat case presented here Pico can on average compute the TT, TE
and EE spectra to better than 1% of cosmic standard deviation for nearly all
$\ell$ values over a large region of parameter space. Performing a cosmological
parameter analysis of current CMB and large scale structure data, we show that
these power spectra give very accurate 1 and 2 dimensional parameter
posteriors. We have extended Pico to allow computation of the tensor power
spectrum and the matter transfer function. Pico runs about 1500 times faster
than CAMB at the default accuracy and about 250,000 times faster at high
accuracy. Training Pico can be done using massively parallel computing
resources, including distributed computing projects such as Cosmology@Home. On
the homepage for Pico, located at http://cosmos.astro.uiuc.edu/pico, we provide
new sets of regression coefficients and make the training code available for
public use. | ['William A. Fendt', 'Benjamin D. Wandelt'] | 2007-12-02 | null | null | null | null | ['pico'] | ['natural-language-processing'] | [-6.75661325e-01 -3.12626183e-01 1.31288916e-01 -3.40081900e-02
-7.54694998e-01 -6.33092701e-01 9.64239478e-01 -2.99078584e-01
-4.01526511e-01 8.83921504e-01 -6.18855096e-02 -6.48741424e-01
3.27867791e-02 -9.01743829e-01 -3.97328228e-01 -9.72869754e-01
-2.26749390e-01 1.06296432e+00 4.24289465e-01 -6.68121725e-02
4.00621027e-01 1.91451684e-01 -9.80471909e-01 -2.74336815e-01
4.85986620e-01 9.82835352e-01 1.75086319e-01 1.07407379e+00
3.65068585e-01 1.99797422e-01 -1.69919416e-01 -3.03635716e-01
4.48394895e-01 -3.00934851e-01 -8.34086716e-01 -5.32430410e-01
-4.38549519e-02 -1.16002504e-02 -2.08512396e-01 1.10831821e+00
6.10127032e-01 2.86627352e-01 8.35582674e-01 -8.05205941e-01
-1.86240092e-01 6.03031278e-01 -1.68798625e-01 7.53243089e-01
-7.62644336e-02 1.77416906e-01 1.06138980e+00 -8.42143536e-01
3.91557336e-01 9.63677764e-01 6.86580479e-01 1.80312231e-01
-1.32108068e+00 -5.37552178e-01 -7.59861708e-01 3.89245786e-02
-1.40869641e+00 -2.82761186e-01 4.61702883e-01 -6.25208735e-01
1.37421739e+00 4.33625638e-01 7.42321491e-01 8.90101016e-01
4.69280005e-01 3.38376984e-02 1.31247318e+00 -6.40607059e-01
5.33410847e-01 3.13104361e-01 2.95231082e-02 5.67293644e-01
2.58660227e-01 3.29445690e-01 -1.47522062e-01 -6.35828018e-01
8.98742497e-01 -8.10274959e-01 1.28131449e-01 -3.55071314e-02
-1.34204626e+00 1.03829670e+00 -8.76210704e-02 4.48716164e-01
1.39768168e-01 3.20910603e-01 1.96949720e-01 1.15169592e-01
6.89089060e-01 5.53215623e-01 -8.83685708e-01 -4.41637367e-01
-6.22011185e-01 7.59190679e-01 9.65293586e-01 5.19524038e-01
7.39631772e-01 7.29970708e-02 9.32119668e-01 9.80236828e-01
4.30602401e-01 8.61944556e-01 6.33161306e-01 -1.44600272e+00
1.57864809e-01 -3.61845613e-01 4.82719600e-01 -8.42044950e-01
-8.88101161e-01 -3.21900785e-01 -6.17513120e-01 3.09344411e-01
4.81125325e-01 -4.52295035e-01 -3.85880202e-01 1.26960039e+00
5.26753187e-01 -8.46438333e-02 -8.82912800e-02 8.48015368e-01
3.52127045e-01 1.00370812e+00 -2.22593606e-01 -1.26199454e-01
1.39138412e+00 -7.30527878e-01 -1.38375178e-01 -1.34226039e-01
6.83362484e-01 -1.12166584e+00 5.82943380e-01 5.20691156e-01
-1.25524366e+00 -2.67374724e-01 -1.14352381e+00 4.81915586e-02
-3.18968415e-01 -4.14048344e-01 1.18355250e+00 8.88511777e-01
-1.12448013e+00 8.96835983e-01 -1.33998740e+00 -3.09839308e-01
-4.06877637e-01 3.25949788e-02 2.24200010e-01 5.52371025e-01
-1.16235149e+00 1.04203475e+00 6.74221277e-01 -4.86384869e-01
-3.77352327e-01 -8.10239673e-01 -3.68946493e-01 -7.07141757e-02
-1.35969803e-01 -6.41889334e-01 1.55692804e+00 -5.67762017e-01
-1.42193830e+00 6.06868505e-01 3.25598150e-01 -3.08746666e-01
8.52893814e-02 1.69825450e-01 -7.09905505e-01 2.54231900e-01
-3.00369337e-02 2.93981314e-01 5.53933047e-02 -6.81186259e-01
-2.24184290e-01 -3.02236415e-02 -2.24663943e-01 -9.21024382e-02
7.99526870e-02 5.19973099e-01 -2.39021719e-01 -5.02835989e-01
3.26160133e-01 -1.08738792e+00 -9.62299332e-02 -5.53666532e-01
-5.28870756e-03 -1.39856577e-01 5.48604950e-02 -1.00644875e+00
6.58369362e-01 -1.67149544e+00 1.72170758e-01 2.63543695e-01
-1.67412594e-01 -9.08635184e-02 3.14613611e-01 4.80485827e-01
-2.64955699e-01 3.71025771e-01 -2.78228641e-01 -4.87225689e-02
3.41035724e-01 -9.09765139e-02 1.66262358e-01 6.17303908e-01
-4.97047812e-01 3.14843416e-01 -4.87045288e-01 -3.13940704e-01
2.42307559e-01 3.79683077e-01 -7.60885715e-01 -1.59487665e-01
-3.63521367e-01 7.71067858e-01 -2.00330272e-01 3.49480689e-01
7.74997294e-01 -4.71812755e-01 1.42012253e-01 3.18418026e-01
-4.48975414e-01 5.36020696e-01 -1.09517515e+00 1.47692978e+00
-6.16559505e-01 5.96049547e-01 1.62932158e-01 -7.59701550e-01
7.76222169e-01 4.76188391e-01 6.49337709e-01 -6.65552557e-01
1.01098921e-02 4.39945430e-01 1.73505381e-01 -2.66076952e-01
3.85556430e-01 -6.22059226e-01 5.78895062e-02 6.82397544e-01
1.94214314e-01 -4.32614952e-01 1.64926574e-01 2.67428935e-01
6.75169408e-01 2.17867538e-01 2.90335804e-01 -8.42249334e-01
3.22113007e-01 -3.32318544e-02 5.28724611e-01 3.87228817e-01
1.18611030e-01 3.63632590e-01 6.50032282e-01 -4.29701865e-01
-1.91082382e+00 -1.09995377e+00 -9.62613702e-01 9.17471826e-01
-3.12228888e-01 -8.64763856e-01 -6.73313379e-01 2.69118905e-01
-2.24364385e-01 9.42861617e-01 -1.41046986e-01 3.45156014e-01
-5.60411990e-01 -1.81124353e+00 2.53990412e-01 2.44815961e-01
3.37244540e-01 -7.93564320e-01 -6.00575693e-02 -2.50537872e-01
-2.46049896e-01 -8.93128693e-01 -1.72384530e-01 -9.08731520e-02
-7.23186255e-01 -6.46885514e-01 -3.66566688e-01 -4.21715260e-01
8.02369118e-02 -4.21918571e-01 9.81251359e-01 1.93337336e-01
-3.51327658e-01 6.77467957e-02 -2.94831485e-01 -2.40702704e-01
-7.10923910e-01 7.43581876e-02 2.82046884e-01 -5.69541395e-01
1.17359705e-01 -9.59157407e-01 -4.59026694e-01 4.36417937e-01
-4.17927712e-01 2.14600936e-01 -1.38974562e-01 5.40525675e-01
7.91614503e-02 2.34208703e-01 2.35513195e-01 -3.94296199e-01
-6.62645511e-03 -7.67752349e-01 -1.49792147e+00 -2.74699003e-01
-7.50433624e-01 1.53845236e-01 3.83699507e-01 -4.76700887e-02
-1.10987413e+00 -4.61018652e-01 -7.13904023e-01 4.24063891e-01
1.12807721e-01 3.85795146e-01 1.71650752e-01 -8.84967968e-02
7.52335250e-01 -1.95796564e-01 -3.92905295e-01 -9.45869327e-01
9.84739214e-02 6.36833370e-01 3.01053047e-01 -9.26534593e-01
8.44273448e-01 1.32749289e-01 1.00747965e-01 -7.96756506e-01
-6.58716142e-01 -4.86447036e-01 -6.58877730e-01 -7.32567906e-02
1.02224600e+00 -8.42758894e-01 -7.09466755e-01 4.08242315e-01
-8.01354051e-01 -5.49541235e-01 3.65652412e-01 9.95524764e-01
-7.51162708e-01 4.07613993e-01 -7.41308928e-01 -7.86354661e-01
-2.62957871e-01 -8.76954615e-01 6.64296269e-01 2.46994138e-01
-7.09302500e-02 -1.49678099e+00 6.83424413e-01 4.29812133e-01
5.10365963e-01 1.98845357e-01 1.02600062e+00 -4.78651583e-01
-5.47388911e-01 -8.08700398e-02 6.17056200e-03 1.48826733e-01
-5.68347275e-01 1.57449991e-01 -8.15822601e-01 -5.28079987e-01
2.01096892e-01 -1.79950431e-01 7.25270391e-01 3.48331362e-01
8.43010008e-01 -4.98531759e-02 -3.03246677e-01 9.09244955e-01
1.65918314e+00 3.31828231e-03 2.19371229e-01 5.64809680e-01
2.05829218e-01 1.38613403e-01 1.91317052e-01 5.72751820e-01
1.19577475e-01 7.89092481e-01 -1.25240255e-02 4.79524165e-01
1.80340916e-01 2.85104245e-01 3.96931767e-01 1.46850264e+00
-6.46782637e-01 3.86704206e-01 -1.26649773e+00 2.27356002e-01
-1.61681521e+00 -1.07010221e+00 -4.08833236e-01 2.33273149e+00
9.21152532e-01 3.25714499e-01 2.42437467e-01 -5.09545028e-01
4.10991251e-01 4.90682535e-02 -1.26514077e-01 -4.91744190e-01
2.88395416e-02 5.42901039e-01 7.45816529e-01 1.18855858e+00
-8.61273706e-01 5.96544385e-01 7.34864187e+00 6.74570918e-01
-8.85643959e-01 8.01322520e-01 4.21932995e-01 -5.30827940e-01
-3.68590266e-01 3.41139048e-01 -7.77979255e-01 6.07652009e-01
1.64308906e+00 -4.79636520e-01 9.69820857e-01 8.73788476e-01
1.95902824e-01 -4.55459923e-01 -5.76038241e-01 8.42180848e-01
-2.21036047e-01 -1.35770226e+00 -9.65424955e-01 3.17092240e-01
8.65169406e-01 9.51605678e-01 -4.31319684e-01 4.65378463e-01
3.96388710e-01 -7.42613316e-01 7.09113836e-01 4.51956898e-01
5.72920084e-01 -8.69439721e-01 6.73638463e-01 4.50355202e-01
-7.87992597e-01 2.38896742e-01 -1.13653302e+00 -2.06479341e-01
3.61146986e-01 8.66855323e-01 -7.14869738e-01 5.03653467e-01
8.75305831e-01 2.29994491e-01 -7.01601744e-01 9.89825547e-01
9.32300314e-02 8.92836094e-01 -7.59260595e-01 -2.07009941e-01
-1.42926797e-02 -6.15514457e-01 5.41857362e-01 1.03281045e+00
4.70458835e-01 4.74905707e-02 -1.34892330e-01 9.86020923e-01
4.36418086e-01 2.16907845e-03 2.86220433e-03 4.80121654e-03
3.05381924e-01 1.59198880e+00 -9.41471338e-01 -4.57197964e-01
-4.78598416e-01 2.99743563e-01 1.63837299e-01 2.26424128e-01
-8.92434537e-01 -3.16569299e-01 5.14599264e-01 -2.91707143e-02
5.21136165e-01 -9.24705267e-01 -3.27478081e-01 -1.47833884e+00
-2.93739468e-01 -4.78359580e-01 1.52461514e-01 -1.09034455e+00
-1.19472933e+00 2.76525766e-01 4.42751616e-01 -3.50101650e-01
-5.07782459e-01 -1.08404744e+00 -8.17648768e-01 1.27731574e+00
-7.45240748e-01 -5.73689878e-01 2.66336888e-01 4.09061909e-02
4.96848933e-02 -3.73509735e-01 9.57355201e-01 2.51141280e-01
-3.88622105e-01 -5.46017624e-02 7.35598445e-01 -5.10322638e-02
5.54589033e-01 -1.48130143e+00 8.56504738e-01 5.87401092e-01
2.20516715e-02 5.02322018e-01 1.22277427e+00 -8.42783272e-01
-1.13389301e+00 -4.45626855e-01 8.11366439e-01 -5.90488672e-01
1.28629375e+00 -7.03692436e-01 -6.13040626e-01 9.80787039e-01
5.02235711e-01 -5.66869751e-02 7.45513439e-01 4.62376952e-01
-8.49697441e-02 2.23184779e-01 -9.36408401e-01 3.45946670e-01
5.54588556e-01 -3.56829792e-01 -5.61917961e-01 9.94679749e-01
6.67079329e-01 -6.46228790e-02 -1.32814336e+00 3.20658647e-02
5.86504459e-01 -1.09711850e+00 9.70562518e-01 -3.34139973e-01
-1.23837933e-01 -1.00759156e-01 -3.77940267e-01 -1.23248541e+00
-5.83987117e-01 -8.44229519e-01 3.09565008e-01 9.88264859e-01
5.51511347e-01 -1.02326930e+00 4.70844954e-01 3.60626638e-01
9.53591689e-02 -3.50966781e-01 -1.13304794e+00 -8.34521294e-01
9.88642037e-01 -6.29292130e-01 4.01175708e-01 8.77204299e-01
1.53266892e-01 2.38256156e-01 -2.18547016e-01 4.28742200e-01
8.56015027e-01 1.58985965e-02 4.06408519e-01 -1.16474175e+00
-1.00014210e+00 -4.32061136e-01 -2.05030646e-02 -4.97356623e-01
-1.35343045e-01 -1.26610243e+00 -4.06663626e-01 -9.37265396e-01
2.68844604e-01 -5.83136201e-01 1.77849770e-01 1.42898681e-02
2.85369664e-01 5.06496668e-01 -3.53637636e-02 2.09800094e-01
-1.27024725e-01 9.50195268e-02 7.84173131e-01 4.32129145e-01
3.76835018e-01 -5.92994466e-02 -1.18554629e-01 1.00178206e+00
1.33807981e+00 -3.70416552e-01 1.07416302e-01 -1.54779807e-01
7.40380764e-01 1.73615605e-01 4.02064651e-01 -1.16326547e+00
-2.76600689e-01 -1.81393623e-01 7.61859357e-01 -4.81797069e-01
5.44028759e-01 5.20735793e-02 3.50535959e-01 3.11335593e-01
1.27958491e-01 1.69969708e-01 1.70050785e-01 -2.12460488e-01
4.63423789e-01 -9.04827356e-01 1.17910445e+00 -5.56475759e-01
-3.43997404e-02 7.22586215e-02 -7.42453218e-01 7.27003440e-02
5.32789111e-01 8.13005090e-01 -4.29577172e-01 -2.64188945e-01
-9.51617002e-01 -1.48231953e-01 7.98882067e-01 -3.98067087e-02
-3.30774784e-01 -1.42607200e+00 -5.76694787e-01 7.40970448e-02
-4.07069504e-01 -5.16062021e-01 1.13836244e-01 9.94872272e-01
-1.08747721e+00 7.34683990e-01 -2.43957937e-01 -3.81662220e-01
-8.26789558e-01 7.22395480e-01 8.01328003e-01 4.27207798e-02
-7.32878506e-01 5.23464739e-01 -1.77983373e-01 -7.19998598e-01
-4.56823379e-01 -4.95812781e-02 4.12138671e-01 -4.01612699e-01
8.33856165e-01 5.60091257e-01 8.76099104e-04 -4.87602293e-01
-1.60367221e-01 5.84210157e-01 3.82993698e-01 -7.66970098e-01
1.39774418e+00 -2.41693959e-01 -6.26954854e-01 7.44876266e-01
1.18656683e+00 3.97597998e-01 -9.60438371e-01 6.21616654e-02
4.73697111e-02 -1.99025869e-01 2.51055658e-01 -1.02117682e+00
-6.54768884e-01 8.04011583e-01 2.94048011e-01 5.90484798e-01
4.42102402e-01 5.40962100e-01 6.51114464e-01 2.92547107e-01
5.36354661e-01 -1.26215684e+00 -2.89018899e-01 4.64200914e-01
9.26968932e-01 -1.06003642e+00 3.81101221e-01 -1.13262378e-01
-1.52213931e-01 1.23004091e+00 3.80406260e-01 -1.94420591e-01
9.80509877e-01 3.60417962e-01 -1.26013711e-01 5.44597358e-02
-1.01690471e+00 2.62200922e-01 1.20539546e-01 2.95360029e-01
4.99780566e-01 3.97685319e-01 -6.19246006e-01 3.53705406e-01
-9.37624395e-01 -6.30133986e-01 7.66604960e-01 4.60015297e-01
-8.38411331e-01 -1.36629486e+00 -8.11359644e-01 1.68896377e-01
-4.41267669e-01 -2.72028416e-01 1.10856965e-01 7.28274345e-01
-3.58158424e-02 6.12081707e-01 2.99398839e-01 -5.24746291e-02
-6.40669823e-01 6.02917492e-01 7.64111578e-01 -3.74590039e-01
-1.16915256e-01 5.20878196e-01 5.45397162e-01 -2.76898801e-01
-2.61134177e-01 -1.15711534e+00 -1.43246782e+00 -8.76437962e-01
-9.68514606e-02 4.12563026e-01 1.22834158e+00 9.42211211e-01
-3.24372649e-01 8.91379192e-02 4.47399974e-01 -1.26598954e+00
-5.84464908e-01 -1.35763848e+00 -8.68945599e-01 -2.54122801e-02
-8.08444247e-02 -6.56709313e-01 -7.88549721e-01 -2.65318304e-01] | [7.043764591217041, 3.5328826904296875] |
8db2ffff-237e-45ff-b5b4-1c40f6aca994 | learning-to-count-everything | 2104.08391 | null | https://arxiv.org/abs/2104.08391v1 | https://arxiv.org/pdf/2104.08391v1.pdf | Learning To Count Everything | Existing works on visual counting primarily focus on one specific category at a time, such as people, animals, and cells. In this paper, we are interested in counting everything, that is to count objects from any category given only a few annotated instances from that category. To this end, we pose counting as a few-shot regression task. To tackle this task, we present a novel method that takes a query image together with a few exemplar objects from the query image and predicts a density map for the presence of all objects of interest in the query image. We also present a novel adaptation strategy to adapt our network to any novel visual category at test time, using only a few exemplar objects from the novel category. We also introduce a dataset of 147 object categories containing over 6000 images that are suitable for the few-shot counting task. The images are annotated with two types of annotation, dots and bounding boxes, and they can be used for developing few-shot counting models. Experiments on this dataset shows that our method outperforms several state-of-the-art object detectors and few-shot counting approaches. Our code and dataset can be found at https://github.com/cvlab-stonybrook/LearningToCountEverything. | ['Minh Hoai', 'Thu Nguyen', 'Udbhav Sharma', 'Viresh Ranjan'] | 2021-04-16 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Ranjan_Learning_To_Count_Everything_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Ranjan_Learning_To_Count_Everything_CVPR_2021_paper.pdf | cvpr-2021-1 | ['object-counting'] | ['computer-vision'] | [-3.10669392e-02 -4.87974286e-01 -9.23803598e-02 -4.28825647e-01
-3.74240279e-01 -4.61257279e-01 6.53350234e-01 4.85235214e-01
-1.01495767e+00 6.79221630e-01 -2.56186724e-01 1.37841493e-01
4.26006407e-01 -8.76616955e-01 -7.76738167e-01 -4.62998211e-01
9.53413546e-02 6.90201402e-01 7.17601120e-01 2.68817425e-01
3.94449800e-01 3.16536546e-01 -1.73119426e+00 2.24010974e-01
3.82760346e-01 7.27537513e-01 3.84844542e-01 9.51238871e-01
-3.45761269e-01 1.02679408e+00 -8.75231445e-01 -3.98813307e-01
-4.98470441e-02 -4.14755940e-01 -5.80280423e-01 -2.35088229e-01
7.43328869e-01 -4.42178130e-01 -1.76136926e-01 1.17499518e+00
3.75011921e-01 3.48529249e-01 9.48911011e-01 -1.19593072e+00
-6.47202015e-01 2.97882169e-01 -8.10641289e-01 1.02716064e+00
8.40395987e-02 2.47210518e-01 6.47735476e-01 -1.15127683e+00
5.30735612e-01 1.06444478e+00 4.30540234e-01 8.24376225e-01
-1.04024267e+00 -7.43389487e-01 1.07286945e-01 2.57176876e-01
-1.57252312e+00 -3.44632924e-01 2.08775803e-01 -6.24229491e-01
9.98454630e-01 3.02840490e-02 8.09679449e-01 8.65763545e-01
-2.18637675e-01 6.58296347e-01 7.94695318e-01 -6.34383440e-01
6.57710314e-01 1.88600481e-01 6.29769862e-01 8.38824093e-01
6.10563874e-01 -1.41468763e-01 -2.59794384e-01 6.91858977e-02
7.10279226e-01 3.55139315e-01 2.82189250e-01 -3.02821636e-01
-1.17991948e+00 9.37728524e-01 5.95533192e-01 5.57490587e-01
-5.64707071e-02 5.51302195e-01 4.72111762e-01 -2.51376927e-01
5.47408938e-01 4.68188599e-02 -1.38960898e-01 -2.72276476e-02
-8.74166012e-01 1.33265376e-01 9.43011105e-01 1.14945459e+00
7.95339286e-01 -1.30690113e-01 -6.51196599e-01 8.87925863e-01
-3.40200126e-01 4.23793077e-01 2.00605273e-01 -7.89298296e-01
2.52424777e-01 4.30932999e-01 2.30796129e-01 -6.66684389e-01
-4.21407849e-01 -4.57936823e-02 -8.02542269e-01 1.15982071e-01
5.84332526e-01 9.03009325e-02 -1.22456789e+00 1.52075291e+00
3.89145792e-01 6.39923215e-01 -4.84503567e-01 8.74117851e-01
9.75563884e-01 6.74279392e-01 4.14650083e-01 -1.54776469e-01
1.62298930e+00 -1.00964510e+00 -3.89816970e-01 -5.30695558e-01
5.77773571e-01 -1.69549912e-01 1.11645508e+00 4.43170816e-02
-8.99104953e-01 -5.80921471e-01 -9.56460297e-01 -1.79192483e-01
-8.68262649e-01 1.27329051e-01 6.00127220e-01 5.06413102e-01
-7.29086041e-01 4.45254594e-01 -7.72942424e-01 -7.60018349e-01
7.63302326e-01 2.28361487e-02 6.09813556e-02 -2.60012031e-01
-7.60869980e-01 1.06584787e+00 6.88035369e-01 -4.70955312e-01
-9.67273891e-01 -4.20915633e-01 -8.50797236e-01 3.18706214e-01
4.05653656e-01 -7.18750119e-01 1.43036354e+00 -5.85765958e-01
-6.11211061e-01 1.26907516e+00 -4.55540150e-01 -4.44723576e-01
4.30940211e-01 1.25667676e-02 -1.59956753e-01 1.43769979e-01
5.63067079e-01 6.65310264e-01 5.51509738e-01 -1.13965392e+00
-9.43603754e-01 -3.04916561e-01 1.85189731e-02 -1.38481170e-01
-3.94117355e-01 3.18030238e-01 -6.11806750e-01 -4.18289423e-01
-5.04463971e-01 -4.20732558e-01 -5.15407249e-02 2.08428681e-01
-2.22061843e-01 -4.93667454e-01 4.84860867e-01 -9.50513035e-02
1.06055856e+00 -1.95997739e+00 -2.44693309e-01 -3.65012556e-01
4.60780561e-01 3.47353548e-01 -1.91530406e-01 3.61277461e-02
1.76050618e-01 6.49464726e-02 -2.75684267e-01 -6.14877760e-01
-2.93336004e-01 2.50497907e-01 -2.51628421e-02 5.55514216e-01
3.51664662e-01 9.19528842e-01 -1.38245296e+00 -9.22132134e-01
5.37706792e-01 3.23470891e-01 -1.69242889e-01 1.16228908e-01
-1.70540720e-01 4.57572639e-01 -2.96073034e-02 8.28977883e-01
6.34060740e-01 -4.79795098e-01 -3.57220501e-01 1.70636475e-01
-1.84311867e-01 -3.92053813e-01 -1.10342145e+00 1.33938479e+00
-3.52456987e-01 4.29223865e-01 -5.72784662e-01 -8.64973426e-01
8.45856786e-01 -2.37490684e-01 1.07176483e-01 -5.90319872e-01
5.11123180e-01 1.02218263e-01 -9.86030102e-02 -3.69259775e-01
3.60805988e-01 -2.88370311e-01 -2.08645001e-01 4.31511700e-01
5.07815599e-01 -1.12745911e-01 1.09514046e+00 2.40302891e-01
1.27320147e+00 -3.63136709e-01 8.49514365e-01 2.16662064e-01
2.13583186e-01 2.38963328e-02 5.78280210e-01 1.35184705e+00
-5.39736152e-01 8.07178617e-01 2.83816874e-01 -7.13764429e-01
-1.44892442e+00 -1.02968478e+00 -2.19159111e-01 1.50210893e+00
1.69991747e-01 -1.57163188e-01 -6.41120672e-01 -3.56004119e-01
3.66029963e-02 9.60765541e-01 -9.67764854e-01 1.05391964e-01
-5.18185854e-01 -6.99492395e-01 3.89359593e-01 8.89896929e-01
4.59590465e-01 -1.22716677e+00 -9.28392589e-01 9.98853892e-02
-6.09823465e-02 -1.16374314e+00 -5.56060672e-01 3.38678211e-01
-6.14659786e-01 -1.20576978e+00 -1.00445831e+00 -7.74508715e-01
6.09859586e-01 5.75802207e-01 1.29151011e+00 3.84535432e-01
-8.02208483e-01 2.68443406e-01 -3.19087565e-01 -9.05828357e-01
-2.06300810e-01 -3.12035941e-02 -2.46437592e-03 -2.49981433e-01
1.08284688e+00 -1.91810310e-01 -5.21026552e-01 1.82196707e-01
-6.50051594e-01 -1.86966091e-01 1.93720847e-01 7.06285417e-01
6.24052048e-01 -3.14647675e-01 3.89432192e-01 -1.00544119e+00
3.86939108e-01 -7.12334692e-01 -7.37129033e-01 3.37337017e-01
-5.39078191e-03 -3.67268920e-01 7.02948987e-01 -7.52692521e-01
-6.74651563e-01 2.22661331e-01 3.95504117e-01 -6.68905139e-01
-2.65134037e-01 -5.42167872e-02 3.28051835e-01 1.78246975e-01
9.44436789e-01 1.43609062e-01 -4.91466850e-01 -2.82575667e-01
4.03840363e-01 4.85983193e-01 6.34546280e-01 -2.17303976e-01
5.94945550e-01 6.88144445e-01 -7.63402954e-02 -8.84401083e-01
-1.20609379e+00 -1.22175252e+00 -1.09351373e+00 -4.35967147e-01
9.84260440e-01 -8.36370230e-01 -7.10423589e-01 4.54464614e-01
-1.39063835e+00 -3.17362607e-01 -5.09221196e-01 3.64630282e-01
-5.97427428e-01 1.58992484e-01 -5.95038176e-01 -1.15199161e+00
-3.58696133e-01 -4.24845368e-01 1.15552175e+00 6.15326464e-01
-1.15301833e-01 -7.69531012e-01 3.50907862e-01 -1.26511678e-01
1.46655083e-01 9.13485959e-02 6.36948705e-01 -8.37999165e-01
-5.09280503e-01 -6.18187964e-01 -5.54320395e-01 2.50770003e-01
-2.26082608e-01 1.11083992e-01 -1.09992981e+00 -2.27104753e-01
-1.80411860e-01 -5.51584899e-01 1.41410613e+00 4.48570371e-01
1.46149445e+00 -2.07067057e-02 -5.36419034e-01 5.57768464e-01
1.60791802e+00 1.19905174e-02 4.56535697e-01 1.08127773e-01
6.81566596e-01 1.61147043e-01 6.47083879e-01 6.84721351e-01
2.74355590e-01 3.43349725e-01 4.31613743e-01 1.47689581e-02
2.87350565e-02 -2.19854951e-01 -3.87172401e-01 6.69257879e-01
-4.12407160e-01 -3.73956412e-01 -1.08193231e+00 7.63268173e-01
-1.88665056e+00 -1.41000330e+00 -2.62185261e-02 2.26637220e+00
5.15388608e-01 1.16951317e-01 4.37129736e-01 -9.86861959e-02
1.21086228e+00 6.51732981e-02 -6.42703652e-01 -1.78816587e-01
1.01720922e-01 2.68769562e-01 5.72069764e-01 1.38505906e-01
-1.30691862e+00 9.86370385e-01 6.09289980e+00 6.50463760e-01
-5.54682016e-01 4.32196677e-01 6.36667728e-01 -2.88150966e-01
6.15835786e-01 -1.01734191e-01 -1.19057941e+00 6.79099560e-01
7.87679374e-01 -3.40607226e-01 4.18149590e-01 1.09830713e+00
-1.62055045e-01 -6.22628629e-01 -1.13646507e+00 1.05504143e+00
3.84386659e-01 -1.20316124e+00 -4.73278165e-02 -3.05270970e-01
6.17684364e-01 1.38904631e-01 -3.31550479e-01 7.41149008e-01
5.15888393e-01 -9.93420482e-01 7.64667511e-01 7.50185370e-01
1.02053213e+00 -5.20659149e-01 7.66692996e-01 7.44525194e-01
-1.33487129e+00 -2.69635797e-01 -1.04450965e+00 -2.99540699e-01
8.39844272e-02 4.27238405e-01 -7.26431012e-01 -2.23841190e-01
8.73054683e-01 5.86983263e-01 -8.01357985e-01 1.72034502e+00
-2.41892755e-01 2.68286079e-01 -3.41494322e-01 -5.50996184e-01
3.33523285e-03 1.12875484e-01 1.55415207e-01 1.61469054e+00
3.29875976e-01 3.68726134e-01 1.67953432e-01 1.12222123e+00
-2.93943226e-01 -2.46996265e-02 -7.95502901e-01 1.75612152e-01
8.46012354e-01 1.41542459e+00 -1.33517480e+00 -9.11241174e-01
-6.23051643e-01 8.98733139e-01 8.89478207e-01 7.94020668e-02
-1.02252972e+00 -4.31126148e-01 8.47959593e-02 1.06549233e-01
4.14496601e-01 1.71444956e-02 -1.63750052e-01 -1.27605343e+00
-1.88093096e-01 -8.05529058e-02 6.15570188e-01 -8.77159715e-01
-1.62488770e+00 8.79159868e-02 2.66387671e-01 -1.11891437e+00
6.47456944e-02 -6.37927651e-01 -1.00836968e+00 6.28520072e-01
-1.20969343e+00 -1.01466393e+00 -8.82094622e-01 3.98295939e-01
6.66195512e-01 1.45349894e-02 7.74646640e-01 2.57698536e-01
-4.68959272e-01 2.99844354e-01 8.29641968e-02 5.97347260e-01
5.41666389e-01 -1.26796055e+00 6.85942769e-01 6.45648122e-01
3.20551723e-01 4.97049928e-01 4.96667147e-01 -5.71749806e-01
-6.06205344e-01 -1.28339565e+00 8.53772879e-01 -8.28639150e-01
5.30973971e-01 -6.70943141e-01 -1.04840314e+00 6.59684896e-01
-3.28293383e-01 7.76033163e-01 4.23469126e-01 9.71818622e-03
-4.29405481e-01 2.92402923e-01 -1.22279978e+00 3.43994379e-01
1.14923930e+00 -3.61212492e-01 -5.62144756e-01 5.03187299e-01
4.14888620e-01 -2.08600014e-01 -2.57655442e-01 8.76183361e-02
5.04837513e-01 -8.93604636e-01 9.41705108e-01 -6.59590542e-01
4.30633545e-01 -2.78311163e-01 6.58827275e-02 -1.09304500e+00
-5.35864830e-01 4.34355140e-01 -4.19598222e-01 9.72993851e-01
-3.86681110e-02 -2.95013100e-01 7.33533621e-01 2.35565066e-01
2.16031462e-01 -3.14300418e-01 -1.01676309e+00 -1.16116452e+00
2.86310077e-01 -6.73724413e-02 3.55527014e-01 7.95357764e-01
-1.44611135e-01 5.25581241e-01 -2.67189711e-01 -1.71112627e-01
7.64698625e-01 -1.10972598e-01 9.46473539e-01 -1.31227493e+00
-6.50735572e-03 -2.47942641e-01 -7.38899589e-01 -7.02063501e-01
-1.45636469e-01 -9.18840289e-01 2.93654978e-01 -1.65250015e+00
1.13898706e+00 -7.62459785e-02 -2.73067594e-01 4.74945486e-01
-4.25627440e-01 5.58866978e-01 4.95640874e-01 3.00450563e-01
-1.05567551e+00 2.46360481e-01 1.02863729e+00 -3.80166560e-01
7.14331493e-02 3.45092639e-02 -2.40127355e-01 8.89189482e-01
9.05185819e-01 -8.44128013e-01 6.94995522e-02 -2.61327207e-01
-1.52106583e-01 -1.11434124e-01 7.52188861e-01 -1.37834191e+00
5.63541472e-01 -1.47553951e-01 6.94235921e-01 -7.92186975e-01
3.76805723e-01 -5.33056438e-01 -2.96915025e-01 5.40210903e-01
-5.54455668e-02 -3.37502182e-01 4.73200679e-02 7.51182437e-01
1.13112770e-01 -8.20683002e-01 1.23705471e+00 -6.48971796e-01
-9.99497831e-01 3.42001557e-01 -1.53572619e-01 4.09572542e-01
1.37518740e+00 -4.65833068e-01 -7.21454918e-01 -1.80941448e-02
-7.58434176e-01 1.98037013e-01 5.29991925e-01 2.12305307e-01
4.82811183e-01 -1.21843493e+00 -6.99727654e-01 -1.50068223e-01
5.95618725e-01 -1.03865694e-02 1.13985330e-01 6.07285738e-01
-4.62690264e-01 3.49764019e-01 -3.48081231e-01 -7.25208938e-01
-1.22473884e+00 1.08982539e+00 2.59464145e-01 -7.73629546e-02
-5.84693789e-01 9.43780184e-01 3.48460913e-01 -2.75405824e-01
2.22034663e-01 -3.18727136e-01 -3.68977129e-01 1.48749188e-01
9.38991308e-01 5.67734063e-01 -1.85688347e-01 -5.09931445e-01
-3.99082780e-01 4.55801964e-01 -8.90182406e-02 2.01042891e-01
1.24725020e+00 1.37361377e-01 1.89893514e-01 1.23206091e+00
1.06149578e+00 -4.59315121e-01 -1.14710104e+00 -2.47973651e-01
3.06248665e-03 -7.72590220e-01 -4.64972436e-01 -5.94619513e-01
-8.02806497e-01 1.01666379e+00 6.79456234e-01 9.84381735e-02
6.50055826e-01 4.34127808e-01 2.43115515e-01 4.60748643e-01
7.05131531e-01 -1.18893647e+00 4.37735617e-01 7.63578892e-01
4.27264005e-01 -1.60831070e+00 4.78587188e-02 -2.04507515e-01
-4.59531546e-01 1.01453972e+00 8.49261522e-01 -3.54636550e-01
4.68725622e-01 3.06904495e-01 -4.03579235e-01 -3.66634399e-01
-5.74437857e-01 -7.08405495e-01 -9.86576080e-02 7.97465742e-01
3.08554709e-01 1.43900543e-01 -1.36957094e-01 4.29732919e-01
2.31757700e-01 3.46194923e-01 6.81435168e-01 1.07452583e+00
-1.00513387e+00 -2.32667699e-01 -3.97096366e-01 9.49432790e-01
-1.66109204e-01 -1.90045312e-01 -3.12123179e-01 8.52572203e-01
2.94043660e-01 6.96362019e-01 5.83885193e-01 3.50193344e-02
3.65705550e-01 5.84976152e-02 5.52283406e-01 -1.18024027e+00
-2.31909335e-01 -4.55980361e-01 -2.20050395e-01 -2.30602428e-01
-4.54168290e-01 -5.41164577e-01 -1.13538396e+00 -6.13558054e-01
-4.98142153e-01 -3.72255504e-01 4.76158708e-01 6.25879943e-01
-3.43421191e-01 4.47801292e-01 2.27327645e-01 -1.00578713e+00
-3.35038543e-01 -1.21833801e+00 -8.33981872e-01 4.31697607e-01
1.13422483e-01 -9.37809706e-01 -4.93723333e-01 1.42940685e-01] | [8.990392684936523, 0.5606868267059326] |
b53b3c1d-8840-4a10-b376-5463313f2ba3 | a-general-framework-combining-generative | 2101.10553 | null | https://arxiv.org/abs/2101.10553v1 | https://arxiv.org/pdf/2101.10553v1.pdf | A General Framework Combining Generative Adversarial Networks and Mixture Density Networks for Inverse Modeling in Microstructural Materials Design | Microstructural materials design is one of the most important applications of inverse modeling in materials science. Generally speaking, there are two broad modeling paradigms in scientific applications: forward and inverse. While the forward modeling estimates the observations based on known parameters, the inverse modeling attempts to infer the parameters given the observations. Inverse problems are usually more critical as well as difficult in scientific applications as they seek to explore the parameters that cannot be directly observed. Inverse problems are used extensively in various scientific fields, such as geophysics, healthcare and materials science. However, it is challenging to solve inverse problems, because they usually need to learn a one-to-many non-linear mapping, and also require significant computing time, especially for high-dimensional parameter space. Further, inverse problems become even more difficult to solve when the dimension of input (i.e. observation) is much lower than that of output (i.e. parameters). In this work, we propose a framework consisting of generative adversarial networks and mixture density networks for inverse modeling, and it is evaluated on a materials science dataset for microstructural materials design. Compared with baseline methods, the results demonstrate that the proposed framework can overcome the above-mentioned challenges and produce multiple promising solutions in an efficient manner. | ['Ankit Agrawal', 'Alok Choudhary', 'Wei-keng Liao', 'Arindam Paul', 'Dipendra Jha', 'Zijiang Yang'] | 2021-01-26 | null | null | null | null | ['geophysics'] | ['miscellaneous'] | [ 3.39665651e-01 -8.02808404e-02 2.44361386e-01 -1.10199004e-01
-4.23586100e-01 -4.47234064e-01 5.59781313e-01 -2.50811845e-01
-1.61401540e-01 8.45302403e-01 -6.81296065e-02 -3.13462973e-01
-3.88104767e-01 -9.20585871e-01 -8.71137559e-01 -1.02974796e+00
4.14400220e-01 8.09432805e-01 -1.44030321e-02 -2.17383325e-01
4.44896668e-01 3.53614032e-01 -1.27635288e+00 -1.09079637e-01
9.72966373e-01 1.13565528e+00 4.28811342e-01 3.16469938e-01
-1.87240303e-01 2.59034306e-01 -2.16955766e-01 -1.04013652e-01
1.05625987e-01 -2.00324401e-01 -4.30051148e-01 -7.55610093e-02
-2.12061659e-01 -1.27613202e-01 -1.46159172e-01 1.18811429e+00
3.54434729e-01 2.12286234e-01 1.16145718e+00 -1.02044487e+00
-1.13298011e+00 2.19182268e-01 -7.41932929e-01 7.60943219e-02
-4.55720723e-02 1.58932969e-01 4.87526715e-01 -9.35702860e-01
3.14206153e-01 1.28305590e+00 3.68265510e-01 3.41482371e-01
-1.27686262e+00 -7.43044674e-01 3.09251755e-01 4.29598182e-01
-1.25534248e+00 -4.46284562e-01 9.77025926e-01 -6.04689717e-01
3.11644465e-01 1.74197376e-01 4.10457015e-01 1.09137595e+00
4.13991123e-01 3.94657910e-01 1.37194276e+00 -1.46513924e-01
3.59319508e-01 2.51985133e-01 -1.98697835e-01 2.40675330e-01
9.21498835e-02 8.97727534e-02 -3.75070684e-02 4.93365005e-02
1.10903084e+00 2.91811734e-01 -3.07603061e-01 -1.72301665e-01
-1.13637137e+00 7.44085431e-01 5.36665320e-01 1.99319229e-01
-4.62465346e-01 3.77366729e-02 -9.57585722e-02 1.05108432e-01
4.37875688e-01 3.69328588e-01 -1.20934345e-01 4.02444489e-02
-5.48928320e-01 2.18585804e-01 5.88885307e-01 9.03286755e-01
8.22046399e-01 8.91967341e-02 3.86312753e-01 8.74018133e-01
7.62512267e-01 8.31868887e-01 3.07001740e-01 -6.84922099e-01
5.29052258e-01 3.29810739e-01 1.28860414e-01 -1.24891198e+00
-1.10412754e-01 -5.62364876e-01 -1.35613501e+00 1.48067698e-01
3.70631814e-01 3.92604172e-02 -1.07863879e+00 1.82016134e+00
4.90775764e-01 5.39214015e-01 -3.01590413e-02 1.16990447e+00
8.14308167e-01 1.01851809e+00 7.45333731e-02 -1.83111608e-01
1.25568223e+00 -6.36422396e-01 -6.99860156e-01 -4.51652616e-01
-9.18912292e-02 -9.33622837e-01 9.20516610e-01 2.60190547e-01
-1.03906870e+00 -4.63495225e-01 -8.69490564e-01 -4.17659618e-02
-2.81111628e-01 -4.30081114e-02 4.03731406e-01 1.77873731e-01
-6.88401103e-01 6.10903382e-01 -9.13787365e-01 -1.99199133e-02
1.49857640e-01 4.52638835e-01 -1.40131384e-01 -3.36192921e-02
-1.18937421e+00 7.08708882e-01 1.35224815e-02 6.36737764e-01
-8.22467983e-01 -8.02677035e-01 -5.37757397e-01 5.12047745e-02
5.71058393e-01 -6.80705011e-01 8.02415431e-01 -2.98325866e-01
-1.50600863e+00 4.32620704e-01 -1.52033061e-01 1.21882088e-01
5.18075287e-01 -1.66359439e-01 -4.87449050e-01 -2.82306552e-01
-5.59633505e-03 5.10713421e-02 9.10480976e-01 -1.50129235e+00
-9.71228331e-02 -5.17161787e-01 6.01680726e-02 3.18150520e-02
-3.19731653e-01 -5.91342151e-01 -3.11837345e-01 -5.81681132e-01
5.60307622e-01 -1.02980757e+00 -4.01995450e-01 1.37652814e-01
-7.96380281e-01 1.84355542e-01 1.19092548e+00 -7.65511036e-01
8.05503368e-01 -1.86987925e+00 5.62431335e-01 2.83208370e-01
4.09441739e-01 1.10893838e-01 1.24053046e-01 3.90566111e-01
-2.79012490e-02 1.01069599e-01 -4.71048951e-01 -1.89462543e-01
-8.71301964e-02 -3.20660509e-02 -3.50353241e-01 4.95862544e-01
-1.15808941e-01 8.94919574e-01 -5.51393270e-01 -4.78564382e-01
4.27911013e-01 6.27850175e-01 -4.60180849e-01 4.24810261e-01
-2.79672414e-01 1.06450498e+00 -9.16822970e-01 4.75892603e-01
7.23260522e-01 -4.83184814e-01 -1.48826078e-01 -4.82279718e-01
-1.78290310e-03 -1.67267680e-01 -1.38160062e+00 1.46456087e+00
-7.45942175e-01 2.65850425e-01 1.26782387e-01 -1.38026237e+00
1.01096320e+00 3.38118166e-01 4.34646219e-01 -3.72855574e-01
2.50182033e-01 2.41303563e-01 1.07447498e-01 -7.69497335e-01
1.01535752e-01 -6.66289091e-01 3.01736332e-02 5.95691800e-01
-4.42051977e-01 -3.52747798e-01 -1.64784342e-01 -2.23632082e-01
7.67771721e-01 1.37751982e-01 8.59248564e-02 -1.20274633e-01
9.29164112e-01 -2.82350570e-01 7.07211196e-01 2.32508555e-01
3.10195237e-01 6.37019098e-01 1.89310893e-01 -4.23137486e-01
-1.04618537e+00 -1.34003460e+00 -7.05959126e-02 3.55156660e-01
4.08175021e-01 3.22560579e-01 -6.86703384e-01 -2.73583472e-01
-1.89287439e-02 6.22350216e-01 -6.31975412e-01 -2.66138047e-01
-6.58265710e-01 -7.71471560e-01 -1.21985167e-01 4.53735173e-01
4.94466096e-01 -1.11272156e+00 -7.14626610e-02 2.67354131e-01
-1.66331127e-01 -1.23759437e+00 -4.43204165e-01 -3.05534244e-01
-9.80006099e-01 -8.57724130e-01 -8.19843233e-01 -5.82882106e-01
9.32706416e-01 2.08683267e-01 8.63477886e-01 -2.86205113e-02
-2.55525887e-01 8.61703046e-03 -1.67097244e-02 -4.39193785e-01
-5.39486468e-01 2.06610728e-02 1.32633671e-01 2.81236351e-01
9.22785606e-03 -1.01480746e+00 -8.63661945e-01 5.81957281e-01
-1.05718124e+00 2.49712154e-01 9.37655389e-01 7.20381737e-01
8.20192754e-01 1.23023860e-01 8.57072115e-01 -1.05993330e+00
5.54759920e-01 -8.07474017e-01 -5.00830054e-01 3.61193568e-01
-5.84177256e-01 2.44353801e-01 1.07793450e+00 -7.94445097e-01
-1.19687223e+00 -1.87685624e-01 -1.85312986e-01 -7.13574529e-01
-1.93601876e-01 6.22852027e-01 -4.77417588e-01 -7.71860704e-02
2.73466676e-01 3.62631649e-01 9.42109600e-02 -6.32715940e-01
1.90561384e-01 6.07928753e-01 4.34280276e-01 -8.53443444e-01
1.08949542e+00 4.48731571e-01 3.44577968e-01 -9.10636663e-01
-6.77458823e-01 -1.21936329e-01 -3.02151173e-01 -2.21368775e-01
8.82610559e-01 -4.68561620e-01 -8.64170074e-01 4.33227837e-01
-9.73162651e-01 1.05708122e-01 -1.78894684e-01 5.35775423e-01
-6.25463426e-01 3.42085063e-01 -3.66066217e-01 -7.04756796e-01
-2.45885491e-01 -1.60891509e+00 9.04111028e-01 3.45714986e-01
-6.20260239e-02 -1.36257696e+00 -1.03488132e-01 4.19807285e-01
5.92577159e-01 4.06445265e-01 1.26611912e+00 -1.97714642e-01
-8.69286418e-01 -1.83437496e-01 -1.71623439e-01 1.32246926e-01
4.37560469e-01 -2.25475311e-01 -7.15614080e-01 -2.05185384e-01
5.18466532e-01 -5.72643802e-02 3.66693527e-01 5.20963192e-01
1.39032269e+00 -2.50934482e-01 -3.77270907e-01 7.03614473e-01
1.24834812e+00 3.63367707e-01 6.12641275e-01 1.62000865e-01
7.83137381e-01 7.04272568e-01 4.01996434e-01 2.51573145e-01
2.51806438e-01 5.87573469e-01 5.01841307e-01 -1.54212892e-01
1.55821458e-01 -3.33656371e-01 -1.54003218e-01 9.76660728e-01
-2.44507357e-01 -3.57398272e-01 -8.41692150e-01 3.54095280e-01
-1.69796813e+00 -7.30600238e-01 -2.59606123e-01 2.07424593e+00
5.74090004e-01 -2.56553348e-02 -4.57998544e-01 1.08758971e-01
8.21557879e-01 1.65903494e-02 -1.14369738e+00 3.07941996e-02
1.89847305e-01 7.71685913e-02 1.67894885e-01 3.01078558e-01
-6.20145440e-01 5.60268223e-01 5.54594803e+00 8.37075591e-01
-1.22669303e+00 1.26053646e-01 7.69210935e-01 1.91850349e-01
-5.40033579e-01 8.45495909e-02 -5.12834370e-01 8.52914870e-01
6.14813447e-01 -7.01809973e-02 5.32815993e-01 5.44162929e-01
2.11187243e-01 -3.53387259e-02 -1.23840797e+00 9.86150444e-01
-3.02682310e-01 -1.36670017e+00 -2.15785690e-02 2.70368963e-01
7.04129934e-01 -4.49998766e-01 3.40892732e-01 4.91615944e-02
-6.21998087e-02 -1.16592288e+00 6.55399978e-01 6.33131623e-01
8.54026914e-01 -5.45251250e-01 4.91751373e-01 7.76910722e-01
-1.17441487e+00 1.12072781e-01 -3.31590384e-01 -2.93912459e-02
5.87611258e-01 9.19597208e-01 -4.87379670e-01 3.58848751e-01
6.19582772e-01 6.58381462e-01 1.17444135e-01 7.78940678e-01
-3.01460236e-01 4.57559735e-01 -3.99350435e-01 -1.04876846e-01
-2.02452660e-01 -8.10732305e-01 5.06098986e-01 5.45039892e-01
6.90327883e-01 3.45010012e-01 3.72290574e-02 1.45237339e+00
-3.55811566e-02 -9.62536782e-02 -4.77075130e-01 -7.32489973e-02
4.66091156e-01 1.17642522e+00 -7.33628631e-01 -4.91218790e-02
-4.45516817e-02 4.70245481e-01 1.47379965e-01 5.85267067e-01
-8.27822089e-01 -1.45200482e-02 5.12823462e-01 3.49324793e-01
-1.00904718e-01 -3.09081316e-01 -4.15915281e-01 -1.01007450e+00
1.42205939e-01 -7.18237996e-01 -1.57811970e-01 -6.82873547e-01
-1.40532458e+00 4.35591072e-01 7.77643099e-02 -1.40430534e+00
-2.12517440e-01 -5.85357904e-01 -7.27138817e-01 1.12058616e+00
-1.34880137e+00 -1.01171446e+00 -4.18613613e-01 3.64002526e-01
4.81385380e-01 9.35926288e-02 5.59573472e-01 3.01813811e-01
-5.87775290e-01 2.41575852e-01 4.02705878e-01 -2.20597222e-01
4.06011373e-01 -1.00058842e+00 3.11838895e-01 4.90413547e-01
-1.87746465e-01 6.43064678e-01 8.95509899e-01 -6.25517249e-01
-1.68848574e+00 -8.51966143e-01 1.28575265e-01 -7.77421966e-02
5.86554348e-01 -4.03441489e-01 -1.18368578e+00 4.82789248e-01
-1.98051691e-01 2.03482375e-01 4.05173004e-01 -2.11744130e-01
2.19166800e-01 -9.77931991e-02 -1.29740763e+00 4.33456212e-01
8.45461905e-01 -3.37075114e-01 -2.58870929e-01 3.64332020e-01
4.45106536e-01 -6.08843923e-01 -1.13368201e+00 4.42156494e-01
5.41917443e-01 -7.78201163e-01 1.26139224e+00 -3.26344818e-01
8.59648108e-01 -3.85508269e-01 -1.30329296e-01 -1.51761460e+00
-1.77189693e-01 -3.28593969e-01 -2.70087481e-01 1.26344848e+00
3.72602016e-01 -1.00001967e+00 9.92227793e-01 6.85884774e-01
-5.70850335e-02 -1.11254120e+00 -8.87754023e-01 -8.12420607e-01
2.17750311e-01 -2.28327885e-01 8.57795477e-01 6.97146952e-01
-7.24450529e-01 3.38001072e-01 -4.09870923e-01 3.88723254e-01
9.70234394e-01 3.85698318e-01 6.63715541e-01 -1.31161082e+00
-5.92013478e-01 -1.51666671e-01 -3.17485631e-01 -1.22970617e+00
1.20085202e-01 -5.61734676e-01 4.30722117e-01 -1.60649323e+00
1.11243553e-01 -1.06013179e+00 -1.80347860e-01 -1.08434595e-01
-1.86727792e-01 -4.28028293e-02 -2.85989679e-02 4.93091106e-01
1.83463052e-01 8.58592987e-01 1.65315664e+00 -3.11000466e-01
5.59199490e-02 2.45977372e-01 -7.17272758e-01 7.32425749e-01
7.30074108e-01 -4.51659441e-01 -7.17233479e-01 -6.80399120e-01
6.39589876e-02 4.26161021e-01 4.13509339e-01 -8.18982720e-01
2.45434552e-01 -6.88483298e-01 3.77745241e-01 -4.10620660e-01
6.42606497e-01 -1.05085599e+00 7.45887756e-01 3.46057296e-01
2.58917585e-02 -3.43518853e-02 -4.69127148e-02 7.27583587e-01
-1.54794976e-01 -2.62333900e-01 8.43391538e-01 -1.34397417e-01
-3.21785778e-01 6.85281456e-01 5.97864464e-02 -3.27848247e-03
1.18769884e+00 -1.01246417e-01 -3.42900217e-01 -4.70654130e-01
-7.03832567e-01 1.08002707e-01 5.06181896e-01 4.62651104e-01
8.71941566e-01 -1.28158855e+00 -6.16726935e-01 2.95470297e-01
-2.65127808e-01 4.92869556e-01 6.30380034e-01 9.01871860e-01
-3.12413514e-01 1.24021955e-01 -1.72199979e-02 -7.89249182e-01
-6.87408328e-01 5.78445554e-01 2.49531850e-01 -2.96030790e-01
-5.41444659e-01 6.72339201e-01 8.82226288e-01 -3.96121681e-01
-1.39886200e-01 -2.95481011e-02 -1.14262037e-01 -4.81604218e-01
2.04494685e-01 5.81202686e-01 -5.88690676e-02 -6.41612649e-01
-1.52147233e-01 1.02323127e+00 -5.55393621e-02 -1.05408140e-01
1.63820004e+00 -1.71094477e-01 -1.94944188e-01 4.75100726e-01
1.21723390e+00 -2.30350137e-01 -1.18539953e+00 -3.83675218e-01
-3.15896481e-01 -3.63282591e-01 1.55969664e-01 -4.47677851e-01
-1.39727032e+00 1.27475739e+00 5.71089745e-01 2.21933156e-01
9.93013144e-01 -2.36614123e-02 9.91935134e-01 2.51441926e-01
3.12468499e-01 -9.96748567e-01 5.04463129e-02 1.64149716e-01
1.14211833e+00 -1.24105954e+00 2.14988440e-01 -6.33478761e-01
-1.78691581e-01 1.03510177e+00 6.52500451e-01 -1.92005739e-01
1.09658003e+00 3.13690037e-01 -4.73513193e-02 -2.57221907e-01
-3.47685724e-01 4.88955557e-01 4.15269583e-01 4.17190969e-01
2.66094506e-01 -2.80297715e-02 1.29160926e-01 4.79180634e-01
-2.17567489e-01 -1.20218351e-01 3.04864973e-01 8.05229485e-01
-2.15692267e-01 -1.17887592e+00 -8.03255260e-01 6.66017413e-01
-1.94220349e-01 2.96864718e-01 9.75721180e-02 6.14710450e-01
2.67859455e-02 8.86415720e-01 -1.12075008e-01 -4.03252065e-01
3.30647409e-01 -4.41096723e-01 2.03329146e-01 -4.59006757e-01
1.24979913e-01 -5.09770913e-03 -4.97824520e-01 -2.40572855e-01
-3.19314271e-01 -6.09945893e-01 -1.29448473e+00 -5.14877200e-01
-4.55417395e-01 2.36505568e-01 1.05311084e+00 1.18590629e+00
2.20520824e-01 5.98517537e-01 8.10624719e-01 -1.04651248e+00
-5.69894373e-01 -9.56496477e-01 -5.92252672e-01 4.20714229e-01
3.38780880e-01 -1.10477853e+00 -3.83699208e-01 3.44792008e-02] | [6.844232559204102, 3.3001317977905273] |
138cb551-d44d-436d-8c5a-3c5f03e4435f | learning-material-aware-local-descriptors-for | 1810.08729 | null | http://arxiv.org/abs/1810.08729v1 | http://arxiv.org/pdf/1810.08729v1.pdf | Learning Material-Aware Local Descriptors for 3D Shapes | Material understanding is critical for design, geometric modeling, and
analysis of functional objects. We enable material-aware 3D shape analysis by
employing a projective convolutional neural network architecture to learn
material- aware descriptors from view-based representations of 3D points for
point-wise material classification or material- aware retrieval. Unfortunately,
only a small fraction of shapes in 3D repositories are labeled with physical
mate- rials, posing a challenge for learning methods. To address this
challenge, we crowdsource a dataset of 3080 3D shapes with part-wise material
labels. We focus on furniture models which exhibit interesting structure and
material variabil- ity. In addition, we also contribute a high-quality expert-
labeled benchmark of 115 shapes from Herman-Miller and IKEA for evaluation. We
further apply a mesh-aware con- ditional random field, which incorporates
rotational and reflective symmetries, to smooth our local material predic-
tions across neighboring surface patches. We demonstrate the effectiveness of
our learned descriptors for automatic texturing, material-aware retrieval, and
physical simulation. The dataset and code will be publicly available. | ['Kavita Bala', 'Siddhartha Chaudhuri', 'Siddhant Ranade', 'Balazs Kovacs', 'Melinos Averkiou', 'Hubert Lin', 'Vladimir G. Kim', 'Evangelos Kalogerakis'] | 2018-10-20 | null | null | null | null | ['material-classification'] | ['computer-vision'] | [-2.45645016e-01 -2.76515007e-01 5.92209287e-02 -2.29585603e-01
-1.00150132e+00 -9.02772963e-01 5.75881600e-01 4.63834763e-01
4.29762214e-01 2.74492264e-01 1.80941999e-01 -6.06210297e-03
-4.00323212e-01 -1.16841340e+00 -1.18814504e+00 -3.56516659e-01
-1.21907435e-01 9.19781208e-01 1.32174967e-02 -2.35474512e-01
3.18110615e-01 1.56644690e+00 -1.37914491e+00 4.92256731e-01
2.37016290e-01 1.32500982e+00 -1.33495450e-01 3.47224861e-01
-2.31774062e-01 1.74972504e-01 -2.01550797e-01 -3.77181917e-01
5.85108459e-01 4.68881428e-01 -6.95967615e-01 1.25521362e-01
1.02297497e+00 -3.07488084e-01 -4.15089041e-01 4.95820522e-01
7.17807889e-01 2.13498902e-02 1.17516887e+00 -1.01029170e+00
-9.83019054e-01 1.66472450e-01 -3.70537370e-01 -5.40289462e-01
3.39272827e-01 1.84502974e-01 9.76388633e-01 -1.36440456e+00
9.64723170e-01 1.34908450e+00 8.27756226e-01 4.33770001e-01
-1.06676841e+00 -3.95004690e-01 3.94323654e-02 -4.57189918e-01
-1.45555782e+00 -3.27678889e-01 1.55216837e+00 -6.11327589e-01
1.05485308e+00 1.83107197e-01 8.31729889e-01 6.55707598e-01
4.37915862e-01 7.37273633e-01 5.93121767e-01 -5.52166738e-02
3.32954586e-01 -1.69173747e-01 -4.02272671e-01 8.67814422e-01
4.84715775e-02 -9.18830037e-02 -5.19464374e-01 -5.42373061e-01
1.35008311e+00 2.88372695e-01 1.25891760e-01 -1.09511292e+00
-1.22485626e+00 2.25293785e-01 6.76793635e-01 -1.18367851e-01
-4.45209563e-01 5.03318846e-01 1.35023832e-01 9.98982936e-02
8.56217563e-01 6.90066218e-01 -6.93533123e-01 9.66162682e-02
-6.76620185e-01 6.80523932e-01 7.13989794e-01 1.37438989e+00
8.76924872e-01 1.76260039e-01 1.45505155e-02 8.25987279e-01
3.72303605e-01 1.17339396e+00 -6.18465006e-01 -1.31066132e+00
2.26026252e-01 9.16935086e-01 2.23394185e-01 -1.17935979e+00
-5.30219853e-01 -1.85517043e-01 -7.13831186e-01 5.05659699e-01
5.59767559e-02 3.84328485e-01 -9.36131299e-01 1.16365135e+00
5.24384737e-01 -1.87541470e-01 -3.41312379e-01 9.57062781e-01
1.31489205e+00 2.54209638e-01 -1.75820798e-01 4.97272760e-01
9.70787823e-01 -4.11151856e-01 1.16692908e-01 5.14886856e-01
4.07886952e-01 -1.00729048e+00 8.14259529e-01 1.68869287e-01
-1.56474257e+00 -2.32655182e-01 -8.32508028e-01 -1.04843497e-01
-3.78345191e-01 -1.19672355e-03 7.72420108e-01 1.90192237e-01
-8.78074825e-01 8.54661047e-01 -6.36832118e-01 2.05902848e-02
9.36190069e-01 4.25074697e-01 -3.67036402e-01 -1.09351635e-01
-5.66108108e-01 4.34199989e-01 -4.85694915e-01 -2.18783990e-01
-1.01014125e+00 -1.44733727e+00 -7.68129230e-01 -1.57925874e-01
-3.38038206e-02 -8.57948422e-01 9.02820051e-01 -1.99570790e-01
-1.28866160e+00 1.24089134e+00 9.52043310e-02 1.31949529e-01
4.10312712e-01 -7.88525045e-02 -2.05921337e-01 2.28624925e-01
-4.70028967e-02 6.97056770e-01 7.35935628e-01 -1.78093636e+00
2.53495485e-01 -2.09675625e-01 2.68721163e-01 3.97120137e-03
5.22166006e-02 -4.61425036e-01 -2.52411216e-01 -7.78168142e-01
4.82142210e-01 -6.76308572e-01 -4.30691242e-02 8.42691302e-01
-4.17113096e-01 -1.95902169e-01 7.72276342e-01 -2.97910213e-01
2.70831913e-01 -2.15281105e+00 6.74346322e-03 3.93247962e-01
4.15891737e-01 -3.14663261e-01 -3.25987577e-01 7.39259958e-01
7.07618818e-02 1.68802351e-01 -7.04910085e-02 -4.32282507e-01
2.56722569e-01 -1.78692147e-01 -3.56268793e-01 5.84960997e-01
6.64941967e-01 1.10823131e+00 -7.32619762e-01 -3.35766733e-01
5.79102874e-01 6.73020184e-01 -7.30597854e-01 -2.15099193e-03
-7.15997577e-01 3.30801666e-01 -7.55795836e-01 1.36733377e+00
1.00017726e+00 -3.49458337e-01 -5.20383358e-01 -6.37134314e-01
3.21289757e-03 1.92260712e-01 -1.10631561e+00 1.89385819e+00
-7.37161100e-01 3.71212482e-01 3.02075148e-01 -4.85181421e-01
1.20021236e+00 3.88941616e-02 1.13847804e+00 -5.13727367e-01
8.03199634e-02 1.96385071e-01 -7.87146151e-01 -1.44833446e-01
6.55651748e-01 1.31079443e-02 -3.21579576e-02 5.55153847e-01
-3.18983734e-01 -9.03322875e-01 -5.94959855e-01 1.96568146e-01
1.18738282e+00 4.48621452e-01 -4.49128181e-01 -5.11498332e-01
2.13183120e-01 1.22402661e-01 1.55406266e-01 4.60046500e-01
3.09571594e-01 1.02680218e+00 9.35733877e-03 -9.18508589e-01
-1.37393296e+00 -1.67628098e+00 -3.33500177e-01 5.97221494e-01
4.20098394e-01 -4.30687517e-01 -2.91065812e-01 -2.75544524e-01
7.33339190e-01 1.28702074e-01 -5.11528075e-01 -1.40715718e-01
-8.44028473e-01 1.18023776e-01 2.44782455e-02 5.99432707e-01
1.72039926e-01 -1.02454531e+00 -2.72893876e-01 1.16567709e-01
4.53869671e-01 -7.81983852e-01 -2.78346747e-01 -7.08283335e-02
-8.86575103e-01 -9.86352086e-01 -7.30264664e-01 -7.62318611e-01
8.45244169e-01 3.87563735e-01 1.66670895e+00 2.97323465e-01
-4.38227236e-01 1.09814966e+00 -1.22279815e-01 -3.78875583e-01
-5.59550524e-01 1.11601509e-01 5.88486642e-02 -3.88044953e-01
-3.88077460e-02 -8.23004425e-01 -8.23370934e-01 4.03657645e-01
-6.28401101e-01 1.12097934e-01 4.11583930e-01 3.30201358e-01
1.06382966e+00 -2.75348514e-01 3.09218675e-01 -3.84018004e-01
1.93750322e-01 -3.86225849e-01 -5.96540749e-01 2.95432210e-01
5.19619249e-02 1.35645848e-02 3.88480842e-01 -4.61057007e-01
-4.94676173e-01 1.49631158e-01 -4.72604372e-02 -8.58281672e-01
-2.39148870e-01 7.37491325e-02 -2.36264825e-01 -5.25903821e-01
5.39886653e-01 -8.12938288e-02 -1.92293987e-01 -7.40744889e-01
4.41903621e-01 3.23983788e-01 1.25285819e-01 -1.20965588e+00
1.00627840e+00 7.25411415e-01 5.30814230e-01 -9.93571639e-01
-4.83113348e-01 -1.76248744e-01 -5.75915754e-01 -4.82708633e-01
5.86628258e-01 -9.66884792e-01 -1.01274061e+00 4.92807359e-01
-1.23450160e+00 -4.47209477e-01 -5.87079525e-01 7.86696300e-02
-9.47806001e-01 6.05047233e-02 -6.05879366e-01 -6.45600021e-01
-5.62187791e-01 -1.13281560e+00 1.95989776e+00 -2.95590073e-01
-1.55561835e-01 -8.45677137e-01 -1.71133444e-01 2.06665456e-01
5.82661569e-01 4.68465775e-01 1.20628095e+00 -1.51354056e-02
-1.27816296e+00 -3.05324584e-01 -3.26677948e-01 -1.66301399e-01
3.12812090e-01 4.50067483e-02 -1.00001681e+00 -2.05506191e-01
-4.44049418e-01 -3.34218711e-01 4.23458874e-01 4.15152192e-01
1.47148657e+00 1.47428721e-01 -2.87284553e-01 5.89589834e-01
1.27470899e+00 -3.00631434e-01 1.18247993e-01 -8.39619190e-02
9.93320584e-01 4.94741768e-01 2.51070917e-01 6.59397066e-01
2.67163098e-01 6.20581269e-01 6.24239922e-01 -5.19182794e-02
-4.80860054e-01 -4.22809482e-01 -2.78639972e-01 8.95245373e-01
-2.04333320e-01 -1.54138997e-01 -1.03168273e+00 4.98677015e-01
-1.28474009e+00 -5.68779826e-01 -9.63855684e-02 1.99163914e+00
4.93221313e-01 1.32141486e-01 -9.28894058e-02 -2.59374231e-01
4.16711956e-01 2.29898617e-02 -8.96725953e-01 -1.86258014e-02
-2.05989644e-01 2.68015921e-01 5.62258482e-01 2.69523799e-01
-9.88141477e-01 8.28193545e-01 6.38266516e+00 7.81450629e-01
-1.16345489e+00 -2.08535597e-01 5.14951050e-01 -1.63479134e-01
-1.04656994e+00 -3.11693817e-01 -5.75389922e-01 -3.25695090e-02
3.96905690e-01 1.37979835e-01 4.94303256e-01 9.79274690e-01
2.42823027e-02 3.01223904e-01 -1.21004915e+00 1.23350108e+00
-4.15327959e-02 -1.97944951e+00 3.81771237e-01 1.91908270e-01
1.01672542e+00 1.75839782e-01 5.31342961e-02 -1.65647835e-01
2.22564593e-01 -9.56052363e-01 1.01435864e+00 9.66909289e-01
1.03284562e+00 -5.72466075e-01 5.77572770e-02 -8.83769020e-02
-1.46946478e+00 4.54428643e-01 -4.21514153e-01 2.40270138e-01
1.29984334e-01 7.55818963e-01 -5.33150911e-01 6.67215466e-01
5.66842198e-01 1.03211415e+00 -3.28024089e-01 1.04316473e+00
5.05979419e-01 1.66247308e-01 -6.09923363e-01 -1.23934224e-01
-1.42989844e-01 6.33746013e-02 6.92047298e-01 5.93216300e-01
2.30220452e-01 -4.79443073e-02 2.42871732e-01 1.39937067e+00
-4.76713330e-01 9.19253007e-02 -8.57348144e-01 -1.59069270e-01
6.31766140e-01 1.08095992e+00 -9.71794307e-01 -7.79089779e-02
-8.39135051e-02 4.98931825e-01 2.74858236e-01 2.30586901e-01
-3.92028958e-01 6.14418536e-02 6.98394895e-01 6.31000102e-01
3.11597109e-01 -8.19211006e-01 -7.08649635e-01 -7.35723615e-01
2.15333462e-01 -2.30725408e-01 -4.63753551e-01 -1.13566041e+00
-1.87205219e+00 1.23158634e-01 2.23659471e-01 -1.56180441e+00
3.54197383e-01 -9.35002983e-01 -3.40953022e-01 7.93228209e-01
-1.12579477e+00 -1.57603395e+00 -2.18878508e-01 2.41058335e-01
3.91051918e-01 -2.28785947e-01 8.14713299e-01 2.52871543e-01
3.01837891e-01 3.01233560e-01 6.53060675e-02 -2.41268962e-03
5.17246902e-01 -9.85095739e-01 7.59807646e-01 -1.14182197e-01
-3.18507925e-02 4.90472287e-01 4.52583075e-01 -9.57898200e-01
-2.28858256e+00 -1.19777501e+00 2.28186876e-01 -8.09019446e-01
4.04340357e-01 -5.91248512e-01 -9.03876126e-01 2.09883928e-01
-4.16852295e-01 2.52931476e-01 2.79254019e-01 -7.82090202e-02
-4.42246199e-01 -5.12728095e-02 -1.30898285e+00 5.79416692e-01
1.58090818e+00 -7.47783482e-01 -2.42641538e-01 6.54002964e-01
8.97798002e-01 -7.31118321e-01 -1.45470428e+00 5.46059966e-01
7.42093325e-01 -4.76259291e-01 1.55249143e+00 -5.64918101e-01
7.71609545e-01 -6.88589886e-02 -5.34980714e-01 -9.96172965e-01
-3.60466510e-01 -3.43976110e-01 -7.31730312e-02 1.10408247e+00
2.00753748e-01 -2.15373605e-01 1.16271210e+00 8.36376548e-01
-5.64725518e-01 -1.06818116e+00 -7.33340383e-01 -8.74457181e-01
5.77385724e-01 -5.36358535e-01 1.17827690e+00 1.01118040e+00
-6.95057094e-01 -3.57050568e-01 2.94984430e-02 7.83094987e-02
7.73997962e-01 7.68472373e-01 8.84323061e-01 -1.41352701e+00
5.57392836e-02 -5.94179094e-01 -5.29103339e-01 -1.14070475e+00
2.49149799e-01 -1.07277334e+00 -7.41351396e-02 -1.66695213e+00
-4.02199402e-02 -1.09710932e+00 -5.25009669e-02 3.04619253e-01
5.90039611e-01 3.22497755e-01 -1.74488097e-01 7.79497698e-02
-6.50761724e-01 9.21764195e-01 1.74745643e+00 -5.71669102e-01
-1.07725985e-01 -1.58530891e-01 -2.45526597e-01 6.00806236e-01
7.91309834e-01 -1.63310513e-01 -6.73502088e-02 -8.16003978e-01
5.30562699e-01 8.33843052e-02 6.21649981e-01 -9.81500864e-01
-5.73431961e-02 -3.40600848e-01 6.68533385e-01 -1.22340775e+00
9.54588652e-01 -1.04421556e+00 3.69656771e-01 -8.03645700e-02
-2.98650771e-01 7.43576959e-02 3.18380475e-01 3.78968447e-01
3.79154742e-01 1.03595093e-01 4.59877729e-01 -3.27567875e-01
-2.28929013e-01 9.40400124e-01 8.24973732e-03 9.79418382e-02
6.36153519e-01 -8.23758245e-02 -3.53089541e-01 -1.37590379e-01
-3.78200978e-01 -1.08232029e-01 1.24304748e+00 3.96691084e-01
1.04626524e+00 -1.79391634e+00 -5.18495262e-01 3.91861111e-01
3.07102352e-01 4.81864661e-01 5.39396524e-01 1.18297145e-01
-8.76113057e-01 3.34610082e-02 -2.03600585e-01 -7.41907597e-01
-7.60330260e-01 2.96570808e-01 4.12200958e-01 4.18363541e-01
-7.95827448e-01 6.79152548e-01 1.13233283e-01 -8.86009157e-01
1.02243364e-01 -4.54633623e-01 5.85326254e-01 -6.44530356e-01
-2.54058302e-03 2.76672751e-01 3.15903604e-01 -6.67823613e-01
-4.76404816e-01 1.17434812e+00 3.23176712e-01 2.19070315e-01
1.70441937e+00 3.81347954e-01 -1.34885564e-01 4.94631678e-01
1.38245106e+00 2.57069349e-01 -1.34922981e+00 -3.16803038e-01
-4.73797470e-01 -5.08478880e-01 4.51900065e-03 -6.69515789e-01
-1.05705273e+00 9.82777417e-01 3.49289745e-01 1.20230712e-01
5.36773860e-01 6.02823198e-01 8.46941411e-01 5.40895581e-01
8.00861001e-01 -8.03353727e-01 2.36844420e-01 4.07916248e-01
1.47463596e+00 -1.18726945e+00 3.67280483e-01 -6.49657845e-01
1.76724553e-01 1.18508029e+00 4.41429317e-01 -5.51601410e-01
1.19080651e+00 3.05639446e-01 -2.50765681e-01 -9.25371706e-01
-5.05631685e-01 3.51819187e-01 6.85363352e-01 6.11140013e-01
2.43963182e-01 2.98046261e-01 6.26589179e-01 2.14905605e-01
-3.06974024e-01 -3.63910943e-01 -1.01677723e-01 1.28970110e+00
-2.98372626e-01 -9.93011057e-01 -3.90807420e-01 6.84470892e-01
6.78299461e-03 1.25121072e-01 -4.48761523e-01 4.24936265e-01
-1.53105676e-01 2.25171372e-01 2.74439186e-01 -4.06856000e-01
5.08356988e-01 -2.77008891e-01 7.65740037e-01 -5.18799007e-01
-3.40793520e-01 -1.44930661e-01 -2.12341268e-02 -5.64892054e-01
-3.46077532e-01 -5.88027894e-01 -1.17395604e+00 -5.05498886e-01
1.07719049e-01 -4.12012607e-01 9.24150705e-01 6.34949148e-01
7.84259319e-01 2.79283017e-01 7.75443792e-01 -1.51310062e+00
-1.82467073e-01 -4.41203028e-01 -5.00269294e-01 7.30880857e-01
1.13091327e-01 -1.11117077e+00 -1.59632877e-01 -2.04448253e-01] | [8.610685348510742, -3.6465096473693848] |
8920b1e7-1462-4f42-92c3-1fdc4fd557e6 | hyperexpan-taxonomy-expansion-with-hyperbolic | 2109.10500 | null | https://arxiv.org/abs/2109.10500v1 | https://arxiv.org/pdf/2109.10500v1.pdf | HyperExpan: Taxonomy Expansion with Hyperbolic Representation Learning | Taxonomies are valuable resources for many applications, but the limited coverage due to the expensive manual curation process hinders their general applicability. Prior works attempt to automatically expand existing taxonomies to improve their coverage by learning concept embeddings in Euclidean space, while taxonomies, inherently hierarchical, more naturally align with the geometric properties of a hyperbolic space. In this paper, we present HyperExpan, a taxonomy expansion algorithm that seeks to preserve the structure of a taxonomy in a more expressive hyperbolic embedding space and learn to represent concepts and their relations with a Hyperbolic Graph Neural Network (HGNN). Specifically, HyperExpan leverages position embeddings to exploit the structure of the existing taxonomies, and characterizes the concept profile information to support the inference on unseen concepts during training. Experiments show that our proposed HyperExpan outperforms baseline models with representation learning in a Euclidean feature space and achieves state-of-the-art performance on the taxonomy expansion benchmarks. | ['Nanyun Peng', 'Te-Lin Wu', 'Muhao Chen', 'Mingyu Derek Ma'] | 2021-09-22 | null | https://aclanthology.org/2021.findings-emnlp.353 | https://aclanthology.org/2021.findings-emnlp.353.pdf | findings-emnlp-2021-11 | ['taxonomy-expansion'] | ['natural-language-processing'] | [-9.56551954e-02 4.27145213e-01 -3.74514014e-01 -2.56042510e-01
2.21052747e-02 -8.81053388e-01 4.85494018e-01 5.72271109e-01
-2.56349087e-01 5.15020117e-02 4.90899980e-01 -6.70117795e-01
-6.25329196e-01 -1.15827596e+00 -3.00249439e-02 -5.55245757e-01
-3.06268334e-01 8.72401178e-01 2.06542656e-01 -2.72301704e-01
2.53275901e-01 5.32849550e-01 -1.55324388e+00 -1.24258839e-01
8.28902662e-01 8.16870093e-01 -2.36701801e-01 3.46342623e-01
-1.12459235e-01 4.77596313e-01 -2.49924108e-01 -5.11063874e-01
2.63450503e-01 -5.12655750e-02 -9.37463164e-01 -1.52562767e-01
4.69342202e-01 -3.14502001e-01 -9.82653499e-01 9.96190429e-01
-5.07898182e-02 3.76864731e-01 1.08481383e+00 -1.62174869e+00
-1.24118161e+00 5.69462359e-01 -2.19515190e-01 1.57170802e-01
2.39412934e-01 -2.38836333e-01 1.79552686e+00 -1.01468492e+00
5.24456859e-01 1.21620202e+00 9.26492929e-01 3.28392446e-01
-1.26262665e+00 -5.77864230e-01 1.48499861e-01 4.72166479e-01
-1.74802613e+00 5.03329754e-01 7.67173469e-01 -6.12713456e-01
9.98323917e-01 1.36995047e-01 1.00268817e+00 1.20509028e+00
4.29944657e-02 2.79096931e-01 4.38806564e-01 -1.58436924e-01
4.48929459e-01 1.03224799e-01 5.98067760e-01 7.13019431e-01
6.31640077e-01 7.88879022e-02 -3.83743912e-01 -4.43654180e-01
6.45668149e-01 2.37580851e-01 -2.18807399e-01 -1.04358339e+00
-7.25870132e-01 1.37091744e+00 8.99337053e-01 1.56933010e-01
-1.02850817e-01 -1.67672470e-01 5.17076373e-01 1.08553089e-01
2.67975122e-01 1.07851696e+00 -1.28470331e-01 1.29663929e-01
-5.92523277e-01 2.57977724e-01 7.40736485e-01 1.29306316e+00
6.23084247e-01 -2.50088304e-01 3.12720120e-01 6.63665295e-01
1.82581574e-01 -1.83318108e-01 7.71383286e-01 -6.96322978e-01
1.82645693e-01 1.28742075e+00 -6.10452294e-01 -9.96407151e-01
-6.67519510e-01 -5.18132269e-01 -8.79112184e-01 -2.81339496e-01
6.66130409e-02 4.12627339e-01 -4.73049909e-01 1.42211592e+00
4.56336766e-01 1.83372185e-01 1.20571531e-01 6.79910600e-01
7.63510108e-01 5.51614165e-01 -1.23465046e-01 2.53042698e-01
1.41834939e+00 -9.01079118e-01 -1.65014416e-01 5.46412803e-02
8.37323368e-01 -7.67704844e-02 1.42824936e+00 2.89975971e-01
-4.03936118e-01 -3.37346852e-01 -1.35967612e+00 -4.71999258e-01
-8.83302987e-01 -4.16203409e-01 9.85264122e-01 7.85820365e-01
-9.50885773e-01 6.46909177e-01 -7.34631777e-01 -7.35119104e-01
2.88973927e-01 3.08679760e-01 -2.64609158e-01 -2.51734674e-01
-1.18382394e+00 6.01989627e-01 9.89550471e-01 -5.15790164e-01
-5.54387689e-01 -9.76991415e-01 -1.08225107e+00 4.78458017e-01
3.46137166e-01 -6.14247203e-01 1.02050138e+00 3.90436053e-01
-1.17107320e+00 5.28093815e-01 6.02115989e-01 -5.97313702e-01
-1.81325778e-01 8.85133073e-02 -3.82732093e-01 2.76167154e-01
-2.35090643e-01 6.65436924e-01 2.95476824e-01 -8.93275976e-01
-7.35716403e-01 -5.34370124e-01 4.71296608e-01 3.88434798e-01
-1.27379167e+00 -5.39761424e-01 -1.73376292e-01 -5.10443509e-01
5.72448254e-01 -1.00256419e+00 -1.24628723e-01 9.94228870e-02
-3.68092984e-01 -8.25150549e-01 8.61741960e-01 -8.24117940e-03
1.49120426e+00 -2.09593868e+00 5.00887454e-01 1.91851959e-01
8.71231318e-01 -1.79910704e-01 -1.32961020e-01 8.13051939e-01
4.52528410e-02 1.01729952e-01 -2.05250874e-01 -1.84060350e-01
3.90711218e-01 5.37946880e-01 -7.08053350e-01 5.58060348e-01
-2.08023980e-01 7.31392205e-01 -1.26396430e+00 -3.83105725e-01
1.45850062e-01 4.91837651e-01 -9.19531226e-01 2.69289732e-01
-2.19278529e-01 -2.78451681e-01 -1.53274298e-01 4.64499354e-01
3.83116573e-01 -3.04117888e-01 4.89115149e-01 -1.10113680e-01
3.03330481e-01 5.17216921e-01 -9.89513516e-01 1.73153043e+00
-4.66214508e-01 4.51931059e-01 -7.45700181e-01 -9.65041220e-01
9.86493528e-01 3.72962840e-02 5.78955531e-01 -1.14037625e-01
6.43881932e-02 1.78350613e-01 7.67352879e-02 -1.15844309e-01
9.72280383e-01 -1.76179130e-02 -1.10453889e-01 4.37542111e-01
2.52650082e-01 -3.42852890e-01 7.39023760e-02 6.04151964e-01
1.04230344e+00 -1.81494281e-01 5.73155105e-01 -3.56669456e-01
2.37536699e-01 -1.07637912e-01 3.88306737e-01 4.68703240e-01
7.41333365e-02 4.32923734e-01 4.92143452e-01 -5.21486104e-01
-1.09480298e+00 -1.44474137e+00 -4.69796807e-01 1.34737134e+00
3.31882894e-01 -1.12423849e+00 -4.12232459e-01 -9.02086198e-01
1.30862206e-01 9.56443667e-01 -8.93436015e-01 -5.68693876e-01
-2.12257624e-01 -4.89394784e-01 6.07459366e-01 9.14908469e-01
-7.04184920e-02 -5.92102945e-01 -6.33600950e-01 8.02042112e-02
2.08008364e-01 -9.62688684e-01 -5.56191564e-01 1.79267570e-01
-1.00213253e+00 -1.38934851e+00 -2.03383684e-01 -7.93973625e-01
4.63272691e-01 3.41315269e-01 1.03123093e+00 -8.88969377e-02
-4.26781446e-01 4.23994035e-01 -6.66673124e-01 -3.38040777e-02
5.18726883e-03 6.32636487e-01 4.66379672e-01 -5.24610758e-01
9.16699886e-01 -1.01576579e+00 -5.24262130e-01 1.14829764e-01
-1.09694266e+00 -3.16416740e-01 1.98006347e-01 9.90061402e-01
3.71361703e-01 3.91963184e-01 2.94961959e-01 -9.91088331e-01
7.85291910e-01 -7.19274342e-01 -6.51178300e-01 1.51905522e-01
-1.17779672e+00 1.94568619e-01 9.48337793e-01 -5.86748004e-01
-4.87986237e-01 -2.76364863e-01 1.22351937e-01 -4.98496890e-01
7.33860582e-02 7.62069643e-01 -1.97751913e-02 4.56895269e-02
9.51934814e-01 1.55334219e-01 -3.42204005e-01 -5.25744319e-01
9.27337348e-01 7.47042418e-01 7.49833405e-01 -5.84150851e-01
1.04464281e+00 3.86486351e-01 2.67729700e-01 -8.57206643e-01
-1.01792300e+00 -9.68011558e-01 -9.00735676e-01 4.89456266e-01
5.35236835e-01 -5.79181194e-01 -7.48909295e-01 -4.91477132e-01
-9.16429460e-01 1.87306121e-01 -5.48894942e-01 5.19117773e-01
-3.97544771e-01 4.97920543e-01 -6.95130110e-01 -5.68303943e-01
-4.40120071e-01 -6.76334381e-01 1.01938462e+00 -2.88020670e-02
-3.73950332e-01 -1.29282355e+00 4.38728213e-01 -2.99392287e-02
6.80774450e-02 1.39242820e-02 1.60850346e+00 -1.07301128e+00
-5.55467188e-01 -3.18449289e-01 -2.80633360e-01 -4.13093865e-02
7.49254748e-02 -2.65741587e-01 -8.57554078e-01 -5.95083952e-01
-2.44598389e-01 -4.23219472e-01 7.22564936e-01 -1.76432088e-01
1.46880758e+00 -5.92005610e-01 -3.59133631e-01 1.15340483e+00
1.46917856e+00 -9.69871134e-02 4.02887970e-01 4.04054463e-01
8.30960512e-01 6.59957886e-01 3.07851046e-01 5.20864725e-01
5.23437083e-01 5.73256075e-01 6.51260257e-01 5.01062274e-01
1.94004551e-02 -7.29489326e-01 -1.35805354e-01 1.19529486e+00
3.18615705e-01 -2.21182376e-01 -9.94627178e-01 6.27049208e-01
-1.59645247e+00 -6.73151672e-01 1.34766236e-01 2.15424919e+00
7.40198076e-01 5.68664335e-02 4.23925132e-01 2.18954250e-01
2.55730331e-01 4.63713035e-02 -6.73660338e-01 -2.08365202e-01
1.03591040e-01 2.46998698e-01 4.04019594e-01 4.58026618e-01
-8.93708408e-01 1.06433046e+00 6.42064476e+00 5.18670142e-01
-6.21622682e-01 -3.42102535e-02 -2.43527919e-01 1.40822262e-01
-4.86988962e-01 8.93453285e-02 -7.46106327e-01 -1.04384094e-01
8.85487854e-01 -8.25560749e-01 3.49354565e-01 1.41413677e+00
-8.26053739e-01 7.09021449e-01 -1.56697059e+00 1.01953113e+00
1.28101066e-01 -1.33533359e+00 3.94666851e-01 5.69669664e-01
4.80832100e-01 -9.72986743e-02 4.78073388e-01 7.84122586e-01
5.60721695e-01 -1.19656241e+00 1.61441535e-01 -1.25045314e-01
9.96884584e-01 -8.21014106e-01 6.73825264e-01 1.49638876e-01
-1.70916510e+00 -4.19252396e-01 -9.80232835e-01 -1.58534214e-01
-2.66890913e-01 3.80683318e-03 -1.36302292e+00 6.19010210e-01
4.79955435e-01 8.33952665e-01 -8.17563057e-01 1.12870717e+00
-3.32195938e-01 5.58353305e-01 -2.48598129e-01 -2.42929861e-01
4.05381680e-01 -3.90481472e-01 4.24212217e-01 1.18909717e+00
2.83172756e-01 2.48444214e-01 3.09721202e-01 9.43221509e-01
-2.84650236e-01 2.72555918e-01 -8.14595461e-01 -3.44954103e-01
9.55309749e-01 1.27993822e+00 -5.52857935e-01 -1.79128140e-01
-3.87672067e-01 7.23015845e-01 6.37187481e-01 2.80519903e-01
-6.63594842e-01 -7.18306243e-01 7.98300385e-01 2.53420800e-01
4.29861881e-02 -3.86135906e-01 -2.42397025e-01 -1.18308318e+00
-2.74962187e-01 -6.14471674e-01 9.36125576e-01 -4.86260951e-01
-1.53749359e+00 8.12100410e-01 2.54389107e-01 -1.11951256e+00
-4.07632053e-01 -9.72028494e-01 -4.49131221e-01 4.91458595e-01
-9.51511025e-01 -1.20779145e+00 -4.59616125e-01 3.83428931e-01
4.31711942e-01 -2.55408168e-01 1.24224448e+00 -9.34055001e-02
-4.45404321e-01 8.82231534e-01 3.51729214e-01 7.77167305e-02
1.69902995e-01 -1.79972243e+00 6.34570420e-01 5.04606962e-01
5.60529590e-01 1.00861382e+00 4.89285916e-01 -4.29365456e-01
-1.16559637e+00 -1.20463288e+00 6.96592331e-01 -5.30457437e-01
1.15053248e+00 -8.03143144e-01 -1.29835689e+00 9.17702794e-01
-9.20948759e-02 -2.47108385e-01 1.10051239e+00 7.98869550e-01
-1.26172423e+00 1.14245102e-01 -8.38225543e-01 7.51664698e-01
1.63689959e+00 -7.35453367e-01 -1.12338305e+00 5.42358100e-01
1.32425940e+00 -6.52084723e-02 -1.24265218e+00 1.58641800e-01
5.72138250e-01 -5.89132965e-01 1.23257148e+00 -1.00662172e+00
2.39645138e-01 -6.45990297e-02 -4.01404500e-01 -1.40089214e+00
-8.55099857e-01 -4.39910263e-01 -5.46437323e-01 6.72754407e-01
1.13449514e-01 -7.22434163e-01 1.09984612e+00 2.06189245e-01
-1.39506727e-01 -9.20457780e-01 -9.22187448e-01 -1.32629168e+00
5.40218234e-01 -8.77068341e-02 9.84172940e-01 1.09172904e+00
6.66725457e-01 7.78023183e-01 1.48720160e-01 4.22105193e-01
7.72485614e-01 1.79062977e-01 5.07307589e-01 -1.88066137e+00
-2.31323123e-01 -5.19143760e-01 -1.01358747e+00 -1.13221288e+00
3.87323260e-01 -1.35683513e+00 -3.83797199e-01 -1.58500791e+00
2.14235216e-01 -4.80900347e-01 -2.92075992e-01 3.09034765e-01
2.78247669e-02 1.29019655e-02 -2.92503852e-02 3.50998551e-01
-5.54861128e-01 1.10347044e+00 6.42262816e-01 -1.58035174e-01
-1.21065423e-01 -3.11290532e-01 -8.23897839e-01 5.73245525e-01
6.03022575e-01 -4.48073208e-01 -1.04755688e+00 -2.31483176e-01
2.79517710e-01 -5.18037617e-01 5.40788770e-02 -1.06450236e+00
6.23852372e-01 1.63100846e-02 -8.65874514e-02 -5.65397382e-01
5.51122248e-01 -8.83604109e-01 -1.95568040e-01 2.47085586e-01
-3.94323409e-01 3.51597250e-01 -4.95689968e-03 8.82692456e-01
-2.73235254e-02 -4.07286614e-01 5.58646739e-01 4.04678732e-01
-6.23199165e-01 7.55192757e-01 -4.38600220e-02 2.39876628e-01
8.24635863e-01 -3.71754646e-01 -4.39049006e-01 -2.03952968e-01
-3.49309564e-01 2.34688163e-01 6.74222469e-01 5.90308249e-01
8.21437240e-01 -1.54325569e+00 -2.55338848e-01 -1.32360654e-02
8.30971062e-01 1.02868348e-01 -6.62459359e-02 1.10496797e-01
-4.91834372e-01 6.96739733e-01 -1.39753848e-01 -3.54994923e-01
-1.02720857e+00 1.26031411e+00 2.64454812e-01 -2.47196972e-01
-1.07707977e+00 8.74936342e-01 5.01710117e-01 -8.27215612e-01
7.39253402e-01 -4.68436986e-01 -3.91617060e-01 -7.25260600e-02
5.44064879e-01 3.93549085e-01 -7.98290074e-02 -2.91214675e-01
-2.50724822e-01 3.82300079e-01 -3.62611294e-01 -3.84317189e-02
1.31733906e+00 1.94675267e-01 2.78690141e-02 3.44008595e-01
1.49443185e+00 -4.04370397e-01 -9.05728757e-01 -5.62236667e-01
2.86729157e-01 -5.50629795e-01 -1.42803982e-01 -2.25978181e-01
-4.90763307e-01 1.00694871e+00 2.99914002e-01 4.47844595e-01
9.09513652e-01 1.77018195e-01 6.70336783e-01 7.39495218e-01
3.45809430e-01 -9.40749824e-01 4.96206492e-01 5.98699629e-01
8.80318820e-01 -5.79834938e-01 2.13123485e-01 -6.78231359e-01
-3.34039569e-01 1.35035479e+00 6.78044319e-01 -3.59433740e-02
7.06069171e-01 -2.77529895e-01 -4.96990770e-01 -5.04531801e-01
-6.82467759e-01 -6.77226484e-02 4.67562109e-01 1.15415597e+00
2.66387433e-01 7.31978267e-02 3.74288149e-02 7.21158206e-01
-8.88348043e-01 -5.84844530e-01 5.51755965e-01 5.10323822e-01
-5.59614837e-01 -9.26520646e-01 -2.22263541e-02 3.06402087e-01
4.47687298e-01 -3.49643171e-01 -6.21187806e-01 1.01130784e+00
8.82226229e-02 8.62727642e-01 2.59835899e-01 -6.26110196e-01
3.57902020e-01 2.73418631e-02 3.89763594e-01 -1.07086945e+00
-5.87898307e-02 -4.91432160e-01 -2.58996785e-01 -3.59006941e-01
4.31247681e-01 -4.69176322e-01 -1.21012866e+00 -4.63824809e-01
-2.52455235e-01 3.93975139e-01 3.66722226e-01 4.84135598e-01
2.43497580e-01 4.40597624e-01 3.61123055e-01 -2.08392695e-01
-1.25289559e+00 -1.02904332e+00 -1.06865716e+00 4.93340671e-01
-1.21599920e-01 -1.03579772e+00 -4.43511099e-01 -4.56982434e-01] | [9.241908073425293, 7.919550895690918] |
07bac234-e0d8-48c0-ac61-ca4fb65850fd | chemlambda-universality-and-self | 1403.8046 | null | http://arxiv.org/abs/1403.8046v1 | http://arxiv.org/pdf/1403.8046v1.pdf | Chemlambda, universality and self-multiplication | We present chemlambda (or the chemical concrete machine), an artificial
chemistry with the following properties: (a) is Turing complete, (b) has a
model of decentralized, distributed computing associated to it, (c) works at
the level of individual (artificial) molecules, subject of reversible, but
otherwise deterministic interactions with a small number of enzymes, (d)
encodes information in the geometrical structure of the molecules and not in
their numbers, (e) all interactions are purely local in space and time. This is
part of a larger project to create computing, artificial chemistry and
artificial life in a distributed context, using topological and graphical
languages. | ['Marius Buliga', 'Louis H. Kauffman'] | 2014-03-31 | null | null | null | null | ['artificial-life'] | ['miscellaneous'] | [ 6.25374243e-02 2.86002696e-01 2.27563158e-01 1.64765239e-01
2.67869323e-01 -1.25223148e+00 1.34363985e+00 4.82548237e-01
-2.07684666e-01 1.14104450e+00 9.31886807e-02 -5.68804920e-01
-1.45653576e-01 -1.14725780e+00 -7.35328197e-01 -9.43444192e-01
-6.46114290e-01 1.04919934e+00 3.30319673e-01 -2.31347397e-01
2.67459273e-01 1.01631570e+00 -1.37437952e+00 1.11688785e-01
7.59129882e-01 6.50109529e-01 4.61418293e-02 1.14207518e+00
2.03076582e-02 1.24962246e+00 -4.17911619e-01 -2.84459472e-01
2.31068999e-01 -9.36279118e-01 -1.07060003e+00 -3.40279013e-01
-3.00357372e-01 2.14170843e-01 -4.76186991e-01 8.33420932e-01
8.14287290e-02 -2.79156454e-02 9.70825911e-01 -7.64601648e-01
-5.69114566e-01 4.43308651e-01 1.78404674e-01 -4.96429831e-01
2.69527733e-01 4.71060276e-01 5.55775046e-01 -5.98468542e-01
1.05881107e+00 1.26766813e+00 1.87233225e-01 4.00820464e-01
-1.53393495e+00 2.93290652e-02 -3.70492816e-01 -4.03500050e-01
-1.58328974e+00 -5.27727842e-01 3.48811336e-02 -4.22494113e-01
1.27889979e+00 5.39502323e-01 9.33150887e-01 2.40475506e-01
7.30894029e-01 -2.45425656e-01 1.33856606e+00 -6.84888780e-01
1.02406549e+00 -1.85374528e-01 -4.00898814e-01 8.48762453e-01
7.42499828e-01 1.01632081e-01 -5.63602805e-01 -4.03607965e-01
9.89511847e-01 2.09693890e-03 1.31404653e-01 -6.25386059e-01
-1.72602892e+00 4.38700736e-01 9.82887447e-02 7.21797884e-01
-5.09606481e-01 5.06945729e-01 1.88439518e-01 4.18528855e-01
-2.97249928e-02 7.43968964e-01 -5.54975688e-01 -6.82624727e-02
-1.57400295e-01 2.12478489e-01 1.57494974e+00 8.35021257e-01
1.00856328e+00 -2.11138651e-01 3.44117135e-01 -2.01106191e-01
2.87381202e-01 3.32242072e-01 2.96820223e-01 -1.29320157e+00
-9.10629926e-04 5.72919250e-01 5.35682857e-01 -7.18942344e-01
-3.40724438e-01 2.91046411e-01 -9.72699285e-01 4.51594859e-01
4.16325003e-01 -2.12748677e-01 -6.65467083e-01 1.56043243e+00
3.18091601e-01 -2.95988500e-01 4.46361899e-01 5.32093048e-01
4.72591162e-01 1.11517310e+00 2.62522370e-01 -3.36774111e-01
1.20939708e+00 -3.32302481e-01 -6.37081623e-01 3.07420731e-01
9.21992242e-01 -4.38749582e-01 2.94320881e-01 5.35180688e-01
-1.34670258e+00 8.55251215e-03 -1.19485569e+00 -3.94213408e-01
-1.08740330e+00 -2.39920393e-01 1.34317172e+00 9.67680633e-01
-1.64979529e+00 6.59183502e-01 -6.74202383e-01 -3.23958248e-01
-1.84904113e-01 6.26646459e-01 -4.42148447e-01 4.02357839e-02
-1.03237462e+00 1.00994384e+00 6.35816813e-01 -1.04139969e-01
-9.89871025e-01 -1.40504420e-01 -5.02709329e-01 -9.50458273e-02
-2.39962097e-02 -1.19094157e+00 7.15822935e-01 -7.72794902e-01
-1.88857985e+00 7.69402564e-01 -2.52847280e-02 -1.68295979e-01
4.84114230e-01 5.52106678e-01 -8.27532187e-02 2.56700665e-01
-2.95844674e-01 7.45716870e-01 -5.29291965e-02 -1.22443449e+00
-2.27030367e-01 -5.91269076e-01 6.83489889e-02 2.23127633e-01
1.41860217e-01 -1.32192388e-01 1.12585656e-01 -1.76119775e-01
3.82494718e-01 -7.91774869e-01 -5.45493543e-01 1.19488783e-01
-2.87178993e-01 -3.89613569e-01 5.52802980e-01 -2.18097866e-01
4.87472028e-01 -1.83455861e+00 5.24130404e-01 4.10524726e-01
5.23655415e-01 4.05628160e-02 3.61311257e-01 1.23508954e+00
-8.27828124e-02 5.79058051e-01 -3.16251814e-01 3.61876398e-01
-1.84790641e-02 2.32856736e-01 1.08664520e-01 7.16510594e-01
-1.45519599e-01 9.65766370e-01 -9.83883381e-01 -5.41320086e-01
1.62623644e-01 1.32246599e-01 -3.68696004e-01 -1.23633236e-01
-6.38706625e-01 5.84351361e-01 -4.98893231e-01 3.72723341e-01
3.10305387e-01 -2.21591011e-01 8.51850212e-01 4.34036881e-01
-7.21199989e-01 2.74516523e-01 -1.30234754e+00 1.49901795e+00
-2.99777836e-02 1.96033463e-01 1.99459046e-01 -5.61190605e-01
8.38001907e-01 7.50347793e-01 6.22885823e-01 -2.62322217e-01
-1.65183797e-01 4.20273453e-01 2.31937364e-01 -1.14539012e-01
1.67326704e-01 -5.91432862e-02 1.49353847e-01 8.44370604e-01
-2.17252061e-01 -9.44747329e-01 4.74552572e-01 6.24621987e-01
1.39376235e+00 5.41152284e-02 3.13257784e-01 -7.29565382e-01
4.06250834e-01 -1.00497812e-01 3.34439903e-01 8.11136246e-01
6.53826296e-02 2.81285476e-02 5.73385239e-01 -7.35535502e-01
-1.72429657e+00 -8.88313293e-01 2.41176672e-02 6.07558668e-01
4.53807145e-01 -6.00463986e-01 -9.80635881e-01 2.43883580e-01
-1.47877991e-01 3.33499521e-01 -3.25230896e-01 1.11959428e-01
-4.33367610e-01 -5.16295731e-01 4.70991910e-01 1.37565730e-05
5.18145919e-01 -1.01974261e+00 -4.73009884e-01 3.52939993e-01
2.80527562e-01 -5.98079443e-01 -1.74699172e-01 5.70633709e-01
-1.05196059e+00 -9.39274371e-01 -3.14758241e-01 -8.29724669e-01
8.22080135e-01 -4.94889021e-02 9.17746842e-01 2.59251744e-01
-2.97831327e-01 1.31845623e-01 4.00926061e-02 -3.99191409e-01
-5.14020920e-01 -3.29391122e-01 -1.86028518e-03 -1.91223755e-01
-2.30784461e-01 -8.40078115e-01 -5.95149696e-01 7.54486695e-02
-1.02650642e+00 1.82491750e-01 2.88149029e-01 5.96253812e-01
4.38837022e-01 2.83803672e-01 1.85429022e-01 -8.18980575e-01
3.08117837e-01 -2.55520761e-01 -6.40599608e-01 3.42027247e-01
-2.91503668e-01 1.38823137e-01 8.04874122e-01 -2.76937652e-02
-7.36313641e-01 4.09202188e-01 5.03814101e-01 7.95703173e-01
-3.23480576e-01 2.80087709e-01 -5.72109640e-01 -2.04595327e-01
6.60417914e-01 2.40186512e-01 -3.82769778e-02 -1.99119106e-01
6.56456828e-01 3.90208960e-01 3.95701766e-01 -1.05693412e+00
3.95297140e-01 6.33189857e-01 7.51124978e-01 -1.17756212e+00
4.70125586e-01 1.03122227e-01 -1.01131511e+00 -1.45955279e-01
7.40819693e-01 -7.83484578e-01 -1.05838978e+00 7.63048112e-01
-1.27151120e+00 -4.24478978e-01 -4.83528316e-01 3.15523684e-01
-6.06007755e-01 4.27992344e-01 -8.24795067e-01 -1.11503172e+00
5.47945034e-03 -6.48378551e-01 5.64807773e-01 9.13771465e-02
-3.34101111e-01 -1.02506542e+00 3.39340568e-01 -1.94929585e-01
4.60368127e-01 7.04068065e-01 1.42475379e+00 -4.52085912e-01
-1.09803772e+00 -2.12317869e-01 5.76510057e-02 -1.58651128e-01
1.94506839e-01 6.08766854e-01 -4.64172482e-01 -2.92626210e-02
-3.68438452e-01 -3.53428274e-01 3.97130758e-01 -1.13414098e-02
8.09857488e-01 -5.80420256e-01 -3.12785238e-01 1.03132106e-01
1.39722371e+00 4.62110519e-01 1.09011555e+00 -6.97890744e-02
3.30111235e-01 7.42531776e-01 -3.61086041e-01 4.06126827e-01
2.09968835e-01 1.72911257e-01 2.27115661e-01 3.50256674e-02
2.49818824e-02 -1.41337886e-01 3.02338213e-01 1.12000501e+00
-5.05163431e-01 -2.07059562e-01 -9.35607910e-01 4.76095498e-01
-1.75343156e+00 -9.07895267e-01 -4.68969494e-01 2.24475265e+00
1.08619833e+00 -2.15454161e-01 1.53325960e-01 -5.02204709e-02
8.48485708e-01 -3.48732442e-01 -5.01504898e-01 -8.63764226e-01
-4.45740283e-01 6.85382843e-01 8.64348352e-01 3.86137873e-01
-5.00230253e-01 1.04982007e+00 8.00143814e+00 4.52754885e-01
-6.55212522e-01 -9.83196646e-02 6.90837562e-01 4.15273786e-01
-5.12654006e-01 5.93044460e-01 2.44142041e-02 5.06496012e-01
1.37046599e+00 -1.52878761e-01 1.12812352e+00 4.03157443e-01
6.50281981e-02 -2.07790807e-01 -1.22688460e+00 4.00097579e-01
-9.48886037e-01 -1.73659432e+00 1.60185799e-01 3.15925002e-01
1.14423716e+00 -4.16522240e-03 -3.85037929e-01 -3.05783063e-01
1.10150242e+00 -1.34536862e+00 8.42565775e-01 7.95132816e-01
7.33692944e-01 -8.06699216e-01 2.73789495e-01 6.16502225e-01
-7.79094398e-01 9.89452377e-02 -2.37218291e-01 -4.25959945e-01
-3.21706504e-01 5.55574775e-01 -4.45987195e-01 6.08498037e-01
3.16971153e-01 6.68929815e-02 -2.26068169e-01 7.69299150e-01
-1.50133669e-01 4.80446160e-01 -6.69761956e-01 -7.01492965e-01
1.46295622e-01 -8.29459071e-01 7.94373155e-02 1.06732035e+00
6.88705444e-02 7.61247218e-01 -1.32318825e-01 8.91058326e-01
-1.10382393e-01 1.92198791e-02 -7.82249391e-01 -6.11304045e-01
6.14463031e-01 8.98634136e-01 -9.74130034e-01 -6.85054719e-01
2.57706251e-02 5.21001875e-01 -3.87313038e-01 4.33267355e-01
7.41019398e-02 -7.81734407e-01 3.97611380e-01 1.54291168e-01
2.65001833e-01 -8.88917863e-01 -4.66377318e-01 -7.98874676e-01
-4.36542869e-01 -6.92250133e-01 2.83383504e-02 -8.59835565e-01
-7.07594216e-01 2.13491898e-02 -4.94158864e-01 -2.82728493e-01
-2.06162885e-01 -6.20000124e-01 -5.48597515e-01 8.92989337e-01
-7.06292748e-01 -8.11228156e-01 3.45437288e-01 5.65533876e-01
-5.92373490e-01 1.34006366e-01 1.38143849e+00 -3.10313106e-01
-9.33271274e-02 -3.73679072e-01 7.62875617e-01 -1.60876140e-01
9.77280140e-02 -1.50556362e+00 4.61063236e-01 4.61760223e-01
-3.20194900e-01 9.84882772e-01 6.67087078e-01 -8.01121414e-01
-2.16340733e+00 -7.85345197e-01 1.35689592e+00 -4.99838114e-01
5.19801199e-01 -7.12527156e-01 -4.80147809e-01 3.14476639e-01
3.34186047e-01 -1.87660366e-01 3.90419751e-01 -2.27601349e-01
-1.69807747e-01 -6.76560476e-02 -1.56345046e+00 6.45194590e-01
9.58643913e-01 -5.47539532e-01 -1.54284105e-01 9.57478583e-01
8.68322134e-01 -1.86018661e-01 -1.16951334e+00 -1.09648198e-01
5.15525639e-01 -7.68846512e-01 6.94219470e-01 -6.31363034e-01
1.41883329e-01 -6.16422474e-01 7.99091607e-02 -7.77546346e-01
-4.61620659e-01 -1.43579841e+00 3.78919616e-02 8.19817781e-01
2.98339754e-01 -1.21148729e+00 5.01944184e-01 8.54187906e-01
5.81754632e-02 -2.05543458e-01 -1.10991764e+00 -7.15486288e-01
3.44147742e-01 4.77217644e-01 9.35811222e-01 1.02656209e+00
4.53742683e-01 1.59619302e-01 -4.39578593e-02 -6.13743216e-02
1.92676947e-01 1.71083789e-02 6.98849559e-01 -1.24831879e+00
-2.73194402e-01 -2.94991136e-01 -6.44876540e-01 -8.53550136e-01
-4.38325763e-01 -8.02313983e-01 -9.35145617e-02 -1.53535032e+00
7.41710365e-02 -4.68771398e-01 6.66296110e-02 4.45852995e-01
8.94065559e-01 -1.99445024e-01 -2.44539529e-01 3.86384845e-01
-6.19623840e-01 2.72244394e-01 1.16563976e+00 9.38208029e-02
1.03454307e-01 -5.19949079e-01 -4.91045505e-01 2.10744753e-01
6.32435977e-01 -3.44926864e-01 -4.04345021e-02 2.11635843e-01
6.71635926e-01 3.27955037e-01 1.69608802e-01 -7.74555564e-01
4.63271976e-01 -6.27868712e-01 4.27897871e-01 -1.97915241e-01
2.30135992e-01 -7.57395804e-01 9.42083299e-01 1.07262993e+00
-2.36595497e-01 1.40809715e-01 -3.27536345e-01 4.00873750e-01
2.94886827e-01 -1.98184401e-01 6.74816310e-01 -8.36645007e-01
-8.60154703e-02 5.08097596e-02 -9.25498545e-01 -4.11820948e-01
1.06352687e+00 -1.34738833e-01 -6.78602397e-01 -4.83540930e-02
-3.85924369e-01 -1.24663502e-01 1.31228507e+00 -5.79792321e-01
-6.00516088e-02 -9.94251430e-01 -5.51494360e-01 -2.03933731e-01
-3.00761491e-01 2.47657776e-01 -2.97398537e-01 2.40303352e-01
-1.39025629e+00 7.95015931e-01 -6.68077618e-02 -2.65677422e-01
-4.72425610e-01 8.26283038e-01 2.83030093e-01 5.47724701e-02
-4.66325343e-01 4.06471789e-01 6.47863671e-02 -5.11466205e-01
-1.47572547e-01 -7.66245201e-02 4.36498880e-01 -4.51639950e-01
3.99910897e-01 3.63158643e-01 5.39172068e-02 -4.35046434e-01
-3.29764128e-01 1.74520105e-01 3.95396411e-01 -3.42316300e-01
1.52787948e+00 1.30563125e-01 -1.29666448e+00 3.15807849e-01
6.54133022e-01 6.69622719e-02 -7.85893261e-01 1.96922705e-01
2.07904298e-02 -4.76768147e-03 -3.96431416e-01 -9.44013476e-01
-2.87127286e-01 5.28832674e-01 3.16486619e-02 6.73472583e-01
5.84703445e-01 -8.21561962e-02 3.44083697e-01 7.99478889e-01
6.44440114e-01 -1.16183865e+00 -1.32398441e-01 5.03115177e-01
4.84546751e-01 -1.73668012e-01 2.00005263e-01 -3.30745459e-01
-8.27399790e-02 1.41886806e+00 -2.46708348e-01 -1.07290044e-01
4.97328579e-01 3.39198858e-01 -5.24332404e-01 -4.81615573e-01
-1.16661227e+00 1.35714054e-01 -4.21367943e-01 6.65152431e-01
6.30855620e-01 2.85205632e-01 -6.80293143e-01 -3.28448236e-01
-7.22430795e-02 -4.37699743e-02 6.05249763e-01 1.39393151e+00
-6.69429898e-01 -1.28837013e+00 -2.87157893e-01 -2.23152906e-01
-2.04198718e-01 -2.68236306e-02 -1.27316332e+00 6.12090170e-01
3.59242022e-01 9.50774789e-01 -2.51992375e-01 2.95986950e-01
-1.14317328e-01 1.20214276e-01 7.58761764e-01 -4.21509534e-01
-5.06842375e-01 -3.73054802e-01 2.87959009e-01 -3.11770737e-01
-2.85754561e-01 -7.44081318e-01 -1.62049007e+00 -9.15245831e-01
-2.12692872e-01 7.77754962e-01 1.27522445e+00 7.24442124e-01
6.64081872e-01 1.92151695e-01 6.31524265e-01 -7.23352194e-01
-2.38186821e-01 -6.20370328e-01 -9.29281473e-01 -1.11521453e-01
3.87643315e-02 1.56863704e-01 -3.06546330e-01 2.60116756e-01] | [5.618916034698486, 4.253814697265625] |
90b93f3b-c70d-4a7b-b782-988ca9c79647 | human-from-blur-human-pose-tracking-from | 2303.17209 | null | https://arxiv.org/abs/2303.17209v1 | https://arxiv.org/pdf/2303.17209v1.pdf | Human from Blur: Human Pose Tracking from Blurry Images | We propose a method to estimate 3D human poses from substantially blurred images. The key idea is to tackle the inverse problem of image deblurring by modeling the forward problem with a 3D human model, a texture map, and a sequence of poses to describe human motion. The blurring process is then modeled by a temporal image aggregation step. Using a differentiable renderer, we can solve the inverse problem by backpropagating the pixel-wise reprojection error to recover the best human motion representation that explains a single or multiple input images. Since the image reconstruction loss alone is insufficient, we present additional regularization terms. To the best of our knowledge, we present the first method to tackle this problem. Our method consistently outperforms other methods on significantly blurry inputs since they lack one or multiple key functionalities that our method unifies, i.e. image deblurring with sub-frame accuracy and explicit 3D modeling of non-rigid human motion. | ['Martin R. Oswald', 'Marc Pollefeys', 'Otmar Hilliges', 'Jie Song', 'Denys Rozumnyi', 'Yiming Zhao'] | 2023-03-30 | null | null | null | null | ['pose-tracking', 'deblurring'] | ['computer-vision', 'computer-vision'] | [ 4.86690521e-01 1.10687412e-01 3.47015858e-01 -8.14686622e-03
-4.45362478e-01 -5.01748443e-01 5.36685646e-01 -6.46678090e-01
-3.10632825e-01 5.74016690e-01 3.60905349e-01 1.24102272e-01
-5.13026444e-03 -2.38979220e-01 -9.03391480e-01 -4.88221318e-01
4.36694086e-01 2.88788050e-01 1.56722814e-01 -7.49864653e-02
2.71905631e-01 2.88902670e-01 -1.39569199e+00 4.50785942e-02
7.89445519e-01 8.24291646e-01 3.22226822e-01 9.38901007e-01
4.24891651e-01 9.89566207e-01 -2.48188838e-01 -9.04322639e-02
2.25273266e-01 -5.49272060e-01 -1.12212980e+00 6.43140972e-01
7.11127818e-01 -8.71516883e-01 -5.84897637e-01 1.01717830e+00
3.94356638e-01 2.83358157e-01 4.39842939e-01 -8.72416854e-01
-8.51653636e-01 -1.54944479e-01 -7.36363232e-01 8.82325172e-02
7.81797767e-01 1.30726591e-01 3.66259277e-01 -8.22320580e-01
8.69230747e-01 1.26316845e+00 7.36877739e-01 6.80172563e-01
-1.45355165e+00 9.87841487e-02 -3.32521312e-02 1.78284287e-01
-1.27645028e+00 -5.80090642e-01 6.92407191e-01 -6.46534264e-01
6.37128711e-01 3.97270352e-01 6.67133868e-01 1.04269850e+00
3.10377982e-02 6.34442568e-01 1.18441689e+00 -4.40676481e-01
1.78724974e-01 -1.11251712e-01 7.37232640e-02 7.78494954e-01
1.66568816e-01 1.44942328e-01 -6.20233297e-01 -1.62458375e-01
1.27523303e+00 -2.10053381e-02 -8.35043907e-01 -5.19866049e-01
-1.38962793e+00 2.85051227e-01 2.46339902e-01 4.23367843e-02
-5.31485260e-01 4.18239057e-01 -2.04810500e-01 8.12794939e-02
6.32056117e-01 3.50157052e-01 -2.60517776e-01 -1.42309666e-01
-1.09505737e+00 3.72381479e-01 6.37022018e-01 7.53020823e-01
7.39900410e-01 -1.09228544e-01 3.46894637e-02 5.94589353e-01
1.81190863e-01 5.27991056e-01 2.82509863e-01 -1.55020559e+00
1.33701906e-01 -2.94018760e-02 7.83549428e-01 -1.02201688e+00
-1.67430490e-01 -1.64979115e-01 -7.32892275e-01 2.25474104e-01
5.95246196e-01 2.23808698e-02 -1.07798755e+00 1.70581985e+00
3.72108430e-01 4.55642462e-01 -1.83748364e-01 1.45075881e+00
4.59504366e-01 4.24661279e-01 -4.47654098e-01 -1.23637550e-01
1.46630299e+00 -1.00717020e+00 -9.59529459e-01 -4.04385537e-01
1.01470374e-01 -8.74345183e-01 7.05411494e-01 3.27346951e-01
-1.41413105e+00 -5.31689167e-01 -8.28634918e-01 -4.91791248e-01
1.79995939e-01 4.23727632e-02 1.46661967e-01 3.53257626e-01
-1.15773928e+00 7.53336132e-01 -9.98760164e-01 -3.92127275e-01
-4.08804566e-02 1.36822313e-01 -4.75936413e-01 -3.80371690e-01
-1.10075080e+00 1.19084024e+00 -1.36784881e-01 1.69333160e-01
-8.77117991e-01 -5.78019440e-01 -1.01200342e+00 -2.28616208e-01
3.63881916e-01 -1.40931606e+00 1.29386234e+00 -1.00427890e+00
-1.56873274e+00 8.69960189e-01 -5.18987834e-01 -2.75062978e-01
9.11324501e-01 -7.23845840e-01 2.20066383e-01 2.60029465e-01
-9.08625647e-02 5.31318903e-01 1.26778793e+00 -1.48027885e+00
-1.10918105e-01 -2.76616007e-01 -3.43354642e-02 4.77048218e-01
1.07267454e-01 -2.04330802e-01 -7.54767299e-01 -9.17255163e-01
2.86777109e-01 -1.19662714e+00 -2.29200557e-01 4.22564968e-02
-1.74623758e-01 6.35418355e-01 8.50184560e-01 -1.28397250e+00
1.10405636e+00 -2.01779103e+00 8.47406030e-01 -1.75441429e-01
4.04856920e-01 2.03102957e-02 -4.36592475e-02 1.90325007e-02
-3.29945683e-02 -1.42397866e-01 -4.53488261e-01 -6.69449091e-01
-2.02218175e-01 2.05413580e-01 -2.58559465e-01 7.42275476e-01
-1.16896346e-01 9.99821663e-01 -1.02681577e+00 -9.74788740e-02
4.58184212e-01 9.70568776e-01 -5.86498857e-01 3.20946991e-01
-8.13533589e-02 9.54852045e-01 -2.01886833e-01 1.08419769e-01
8.44870269e-01 -4.40035760e-01 -9.14919749e-02 -5.24926782e-01
-1.13342248e-01 -5.23523912e-02 -1.22754312e+00 2.07838535e+00
-2.84661144e-01 6.59854949e-01 4.30579036e-01 -5.41455269e-01
3.63396823e-01 3.91527414e-01 6.85839832e-01 -6.66299984e-02
5.56048825e-02 1.72550246e-01 -5.67867100e-01 -7.42985070e-01
6.75042391e-01 -9.77906212e-02 2.02646434e-01 4.85771805e-01
-1.81486681e-01 -2.71595865e-01 -3.67064655e-01 1.02681346e-01
1.09024513e+00 5.48780382e-01 7.28799924e-02 -3.45430747e-02
4.14573193e-01 7.19145872e-04 2.98973829e-01 6.31506801e-01
-1.05902262e-01 1.29335082e+00 1.44636363e-01 -6.55402958e-01
-1.20700073e+00 -9.12766099e-01 2.18924344e-01 5.24910152e-01
5.07955194e-01 -1.34424001e-01 -1.07278037e+00 -2.24760488e-01
-8.88461545e-02 4.61083710e-01 -6.93288863e-01 -2.70947404e-02
-6.95821941e-01 -5.87213635e-01 2.28007630e-01 1.30812615e-01
5.75867295e-01 -5.11159956e-01 -8.68378639e-01 9.71926153e-02
-8.03873658e-01 -1.23349190e+00 -1.03675771e+00 -3.80327880e-01
-8.27828407e-01 -1.03563499e+00 -1.38804591e+00 -6.30540609e-01
9.23191786e-01 6.12914622e-01 9.07508552e-01 3.64025459e-02
-3.99774134e-01 6.14793420e-01 -1.53151096e-03 4.33891505e-01
-3.02078307e-01 -5.52535355e-01 1.60565257e-01 2.28047684e-01
-2.60416746e-01 -3.91262174e-01 -8.30022931e-01 3.29718679e-01
-9.55218673e-01 5.79568505e-01 2.72786796e-01 9.14066911e-01
3.63321394e-01 -1.55033907e-02 -2.43910521e-01 -5.17542064e-01
4.85210896e-01 -1.04690224e-01 -4.83405948e-01 2.17027590e-01
-3.21836650e-01 3.45877349e-01 5.18248715e-02 -5.88001728e-01
-1.29714477e+00 3.49873513e-01 2.72861123e-01 -8.17469120e-01
-2.76803821e-01 9.93770361e-02 5.27903512e-02 -2.67160863e-01
7.08817780e-01 2.37464949e-01 2.46734694e-01 -7.32291698e-01
6.88276887e-01 3.66318136e-01 9.79635835e-01 -3.07174027e-01
6.80893481e-01 8.32175612e-01 -4.02416517e-05 -7.57715642e-01
-6.71144009e-01 -4.78558749e-01 -7.97572136e-01 -2.72779703e-01
1.18406391e+00 -1.04894853e+00 -6.90931797e-01 8.93549263e-01
-1.57042086e+00 -3.50792587e-01 -5.49089611e-02 5.49195051e-01
-9.18507159e-01 9.31367218e-01 -9.60290432e-01 -7.68303871e-01
-3.60585451e-02 -1.19239104e+00 1.32350218e+00 -2.12480083e-01
-3.60962421e-01 -8.42360497e-01 4.33865264e-02 5.21372914e-01
3.90320212e-01 4.02804822e-01 2.41004512e-01 4.24245119e-01
-8.56060088e-01 6.94105253e-02 -1.60176352e-01 2.22524315e-01
3.16353440e-01 -5.13919532e-01 -9.51405406e-01 -2.99764872e-01
4.60881293e-01 -1.65500806e-03 9.34110940e-01 7.47288287e-01
7.58412123e-01 -4.66313034e-01 -2.50101626e-01 7.08059967e-01
1.23400199e+00 -2.16525093e-01 7.49409139e-01 3.21596771e-01
1.07284856e+00 6.88020051e-01 3.97774756e-01 3.25600952e-01
4.78286773e-01 1.03019464e+00 2.46479303e-01 -2.04050660e-01
-5.26818097e-01 -1.14717424e-01 3.86237115e-01 5.25197804e-01
-3.76117110e-01 -2.09831018e-02 -8.10683250e-01 5.12888014e-01
-2.09591961e+00 -1.07108390e+00 -2.50929981e-01 2.32933378e+00
7.15357959e-01 -3.90116096e-01 -1.04933545e-01 -9.23375115e-02
7.59367287e-01 1.07229210e-01 -4.55610067e-01 2.13712111e-01
6.27522543e-02 -2.40406528e-01 5.67366600e-01 1.20211065e+00
-1.04869413e+00 9.22721267e-01 6.73430920e+00 3.07903051e-01
-8.49376261e-01 1.72541872e-01 4.01950836e-01 -1.61476091e-01
-2.40676045e-01 7.35766962e-02 -3.41342539e-01 4.91461694e-01
4.94870394e-01 7.63222054e-02 1.00498915e+00 3.25264722e-01
5.73126793e-01 -3.47559720e-01 -1.01246047e+00 1.20625806e+00
3.97775859e-01 -1.15130568e+00 -2.02758551e-01 1.42775103e-01
9.82388496e-01 -2.95703471e-01 1.11305512e-01 -5.49706697e-01
2.04619482e-01 -9.45183277e-01 1.00999403e+00 9.27831829e-01
7.77470589e-01 -1.68719426e-01 4.03382212e-01 4.38731611e-01
-8.27500939e-01 2.56001890e-01 4.04417366e-02 -1.91763774e-01
7.68300891e-01 6.51720285e-01 -2.38355488e-01 5.84400594e-01
6.80427730e-01 7.70798981e-01 -2.35420123e-01 9.40822899e-01
-1.71863496e-01 -1.93415000e-03 -1.76411167e-01 7.53459096e-01
-1.75114304e-01 -3.07178974e-01 8.04317415e-01 8.25793505e-01
5.99062324e-01 4.04891729e-01 -1.50866076e-01 9.36598301e-01
2.81755686e-01 -4.95730102e-01 -4.06482995e-01 4.64870542e-01
5.58893681e-02 8.36980700e-01 -3.73207688e-01 -4.84319329e-01
-3.09935868e-01 1.87327480e+00 4.42340001e-02 7.06935942e-01
-7.96370983e-01 7.18759000e-02 7.41214812e-01 2.61827230e-01
2.92730123e-01 -4.12224442e-01 -2.98863888e-01 -1.69932318e+00
8.86210129e-02 -8.28100920e-01 6.65386617e-02 -1.28317654e+00
-9.30275261e-01 6.14071727e-01 3.73001769e-02 -1.12891400e+00
-4.40717578e-01 -3.74518931e-01 -7.24100173e-02 1.17173910e+00
-1.34492397e+00 -1.01434207e+00 -6.65165961e-01 4.53877658e-01
4.41856176e-01 4.75723237e-01 5.52005768e-01 1.10697493e-01
-1.53819755e-01 -8.75379145e-02 -1.32123739e-01 -3.04899544e-01
9.42063630e-01 -1.11034679e+00 6.68050647e-01 1.05645585e+00
-3.51177305e-01 6.18268788e-01 1.13907385e+00 -7.94824183e-01
-1.44683158e+00 -8.74175072e-01 8.41942251e-01 -8.30268383e-01
3.20934504e-01 -1.04369216e-01 -9.77958977e-01 7.65091896e-01
2.10622028e-01 2.04617515e-01 -3.24274860e-02 -5.83579719e-01
-9.88600403e-02 3.83709967e-01 -1.01258004e+00 4.90436912e-01
1.18498755e+00 -5.83895683e-01 -6.38510227e-01 1.60646498e-01
6.61314845e-01 -9.76265252e-01 -6.93083942e-01 1.75561845e-01
5.30218899e-01 -8.91159832e-01 1.27693987e+00 -1.94723338e-01
4.70560580e-01 -7.37588227e-01 8.92330706e-02 -1.33054614e+00
-5.08525610e-01 -9.82961178e-01 -4.69200611e-01 7.23432899e-01
-1.62759915e-01 -4.77807641e-01 7.08104372e-01 9.69148457e-01
2.09369227e-01 -2.35079184e-01 -6.46667659e-01 -6.21208072e-01
-3.93128335e-01 -1.64161175e-01 3.44103992e-01 8.64493251e-01
-1.87864095e-01 -2.83720549e-02 -1.20882785e+00 3.77650797e-01
1.08204043e+00 -2.94724721e-02 7.55289912e-01 -7.93522179e-01
-5.10055304e-01 -1.10317431e-02 -2.06032500e-01 -1.68148196e+00
-1.87236443e-02 -3.26624662e-01 3.23554397e-01 -1.42800057e+00
2.35794917e-01 3.88470776e-02 1.94820985e-01 1.32537037e-01
-3.22435260e-01 3.03915232e-01 1.22363888e-01 5.58933020e-01
-3.17204058e-01 4.18374509e-01 1.45476365e+00 6.29548728e-02
-2.31672347e-01 -2.66184211e-01 -4.50812668e-01 7.22481251e-01
3.88877153e-01 -3.33012670e-01 -3.71604890e-01 -9.72487986e-01
-1.90970208e-02 3.77425730e-01 9.95282412e-01 -7.79414058e-01
4.23434287e-01 -1.69599339e-01 4.49159265e-01 -2.46330142e-01
5.70879757e-01 -7.41270781e-01 8.64478648e-01 4.89032328e-01
-2.87153184e-01 1.80362202e-02 8.65058042e-03 7.30462730e-01
-2.78277099e-02 -3.51742208e-02 7.67029285e-01 -3.46558362e-01
-6.90277815e-01 1.79051027e-01 -3.69568855e-01 -2.35034823e-01
6.82183564e-01 -2.08937705e-01 -2.43564397e-01 -7.26963520e-01
-1.02009928e+00 -2.02423722e-01 1.15955746e+00 4.10369009e-01
5.97299278e-01 -1.33179080e+00 -6.18731201e-01 2.03671351e-01
-3.28799963e-01 9.52030271e-02 3.48521531e-01 1.00150144e+00
-7.39372015e-01 3.15430790e-01 -6.18439876e-02 -6.00171626e-01
-1.39875066e+00 6.30669355e-01 6.05330110e-01 -5.13051525e-02
-8.21574748e-01 6.88103020e-01 4.65693176e-01 8.90238434e-02
3.66289876e-02 -1.63319021e-01 2.64937490e-01 -4.45496142e-01
7.03943372e-01 5.92326641e-01 -5.33758365e-02 -1.04079449e+00
-2.96775103e-01 9.97461259e-01 2.73048550e-01 -6.57314837e-01
1.19080198e+00 -8.13674152e-01 -2.22783610e-01 5.50447591e-02
1.10225856e+00 -3.73931713e-02 -1.91162562e+00 -2.12131187e-01
-3.96909297e-01 -8.01260173e-01 2.16764629e-01 -6.43261135e-01
-9.14238513e-01 5.69660306e-01 4.87874568e-01 -1.75917253e-01
1.20162702e+00 2.56435480e-02 8.71066213e-01 -1.58095241e-01
4.00798768e-01 -8.49756956e-01 2.11663648e-01 4.34936911e-01
1.09537315e+00 -1.00804818e+00 2.18607694e-01 -5.99230886e-01
-5.72694838e-01 9.35601115e-01 4.21153396e-01 -6.48717955e-02
3.57044697e-01 1.43177450e-01 6.35773987e-02 -1.99480038e-02
-5.56483150e-01 -1.74843073e-01 5.26874304e-01 4.57693130e-01
2.53166586e-01 -2.40831375e-01 -1.36514142e-01 7.68303275e-02
4.26765904e-02 3.21274668e-01 5.73081017e-01 7.59025276e-01
-2.70889044e-01 -9.31282997e-01 -9.58885908e-01 -2.70026863e-01
-3.04229438e-01 4.16116342e-02 -2.68696934e-01 3.32175434e-01
-4.88872416e-02 9.64102268e-01 -1.89422369e-02 -3.18035841e-01
2.61936069e-01 -4.33991775e-02 8.41515124e-01 -2.15110779e-01
-4.66521792e-02 2.51273751e-01 -1.85300767e-01 -8.20592701e-01
-5.99198580e-01 -7.11553931e-01 -7.97369719e-01 -5.13772130e-01
-1.31766647e-01 -2.78709590e-01 4.23085183e-01 7.40091145e-01
5.06327629e-01 3.13068241e-01 1.94925770e-01 -1.32378268e+00
-2.32381716e-01 -6.53261781e-01 -3.54820251e-01 7.90737092e-01
8.56787384e-01 -5.95406115e-01 -5.41691780e-01 7.02317238e-01] | [11.44452953338623, -2.547445058822632] |
0f16d9ae-0f1a-42c4-a852-f01642c7e0fb | iterative-correlation-based-feature | 2201.08959 | null | https://arxiv.org/abs/2201.08959v5 | https://arxiv.org/pdf/2201.08959v5.pdf | Few-shot Object Counting with Similarity-Aware Feature Enhancement | This work studies the problem of few-shot object counting, which counts the number of exemplar objects (i.e., described by one or several support images) occurring in the query image. The major challenge lies in that the target objects can be densely packed in the query image, making it hard to recognize every single one. To tackle the obstacle, we propose a novel learning block, equipped with a similarity comparison module and a feature enhancement module. Concretely, given a support image and a query image, we first derive a score map by comparing their projected features at every spatial position. The score maps regarding all support images are collected together and normalized across both the exemplar dimension and the spatial dimensions, producing a reliable similarity map. We then enhance the query feature with the support features by employing the developed point-wise similarities as the weighting coefficients. Such a design encourages the model to inspect the query image by focusing more on the regions akin to the support images, leading to much clearer boundaries between different objects. Extensive experiments on various benchmarks and training setups suggest that we surpass the state-of-the-art methods by a sufficiently large margin. For instance, on a recent large-scale FSC-147 dataset, we surpass the state-of-the-art method by improving the mean absolute error from 22.08 to 14.32 (35%$\uparrow$). Code has been released in https://github.com/zhiyuanyou/SAFECount. | ['Xinyi Le', 'Wenhan Luo', 'Kai Yang', 'Lei Cui', 'Xin Lu', 'Zhiyuan You'] | 2022-01-22 | null | null | null | null | ['object-counting'] | ['computer-vision'] | [ 3.02180022e-01 -4.36678350e-01 -2.19792336e-01 -2.46605188e-01
-8.20863128e-01 -4.43725675e-01 6.17955387e-01 4.27700907e-01
-6.93999052e-01 4.60511893e-01 -3.11206765e-02 2.28385359e-01
-1.49665579e-01 -7.85611272e-01 -6.45882130e-01 -7.74100482e-01
4.88604568e-02 4.03750181e-01 7.00505614e-01 4.49949019e-02
5.45092106e-01 2.82324612e-01 -1.88537574e+00 2.55184501e-01
7.89022744e-01 1.23024201e+00 5.58892429e-01 3.88316691e-01
-2.39374399e-01 6.30733550e-01 -5.96372724e-01 -2.53149807e-01
1.69124201e-01 -3.14407200e-01 -5.64410746e-01 1.66121840e-01
7.02670693e-01 -2.39437714e-01 -1.07049346e-01 1.25403607e+00
3.54496151e-01 3.18890870e-01 6.59597278e-01 -9.95091617e-01
-5.29493988e-01 1.35772094e-01 -9.93605733e-01 4.63886559e-01
1.71073511e-01 2.06582710e-01 1.08642876e+00 -1.28745115e+00
5.33236623e-01 8.96304250e-01 2.94824839e-01 2.82776564e-01
-1.12441933e+00 -6.51733279e-01 2.47079823e-02 3.60787064e-01
-1.62849760e+00 -4.16473001e-01 5.34511030e-01 -2.83864468e-01
6.40343249e-01 4.18851942e-01 5.48518419e-01 5.60951829e-01
-3.11113209e-01 7.51062572e-01 9.71109807e-01 -4.89910245e-01
3.28309774e-01 2.23852560e-01 2.38902271e-01 7.96900988e-01
4.23748434e-01 -1.93353862e-01 -3.75981599e-01 1.18511682e-02
5.29215753e-01 4.81669992e-01 -1.38813436e-01 -4.39194769e-01
-1.28913164e+00 8.11229944e-01 6.68390334e-01 5.32366335e-01
-3.37138951e-01 2.35086698e-02 4.06374156e-01 -1.60404041e-01
2.74382800e-01 4.79689419e-01 -2.56997272e-02 4.78805602e-02
-1.04598475e+00 1.78205252e-01 4.61633146e-01 9.49312150e-01
9.62220252e-01 -4.54384774e-01 -4.63539004e-01 1.01691091e+00
-1.60945430e-01 3.14896166e-01 3.33474904e-01 -8.82164598e-01
5.76693892e-01 8.98429036e-01 2.03161448e-01 -1.02937317e+00
-1.40419528e-01 -5.67608953e-01 -8.32997739e-01 1.17758177e-01
5.39387524e-01 3.31188917e-01 -8.05657983e-01 1.50532269e+00
4.39804822e-01 3.50975901e-01 -1.79023996e-01 9.92636561e-01
5.92607796e-01 6.72146797e-01 1.23236864e-03 -1.70939237e-01
1.64011443e+00 -1.14144778e+00 -2.97437191e-01 -2.93139815e-01
6.50355399e-01 -7.04095662e-01 1.39667869e+00 2.08177194e-01
-9.81157005e-01 -6.34884357e-01 -1.10848570e+00 1.45211160e-01
-5.19062817e-01 3.05700272e-01 4.20862228e-01 3.13313186e-01
-5.15443683e-01 6.16370261e-01 -4.98578668e-01 -2.95390338e-01
4.50382948e-01 2.45437189e-03 -2.20480859e-01 -3.11268866e-01
-9.12791133e-01 7.25633562e-01 7.19433546e-01 -4.26847547e-01
-7.46558905e-01 -6.65439963e-01 -8.36766005e-01 2.40513459e-01
8.00247252e-01 -4.08003926e-01 9.53369319e-01 -6.12175226e-01
-7.56477058e-01 9.27122533e-01 -2.18070567e-01 -2.78368026e-01
3.96161467e-01 -1.01326801e-01 -4.34742689e-01 3.04241925e-01
4.65632707e-01 5.71545064e-01 7.59253621e-01 -1.27729142e+00
-1.16615570e+00 -3.58131468e-01 1.32508650e-01 2.21285060e-01
-5.25804222e-01 1.07809983e-01 -8.65154982e-01 -5.55403352e-01
-4.31966484e-02 -6.88936889e-01 -1.06061436e-01 1.26010552e-01
-3.06752175e-01 -2.90092498e-01 6.74195111e-01 -7.12251514e-02
1.40201640e+00 -2.35606170e+00 -2.14459002e-01 3.30486447e-02
3.15540820e-01 3.65148783e-01 -1.84870183e-01 2.12634191e-01
8.90071839e-02 -1.27024561e-01 -3.14608485e-01 -4.90601480e-01
-1.76537573e-01 1.14899129e-01 -1.69621170e-01 5.54487228e-01
2.95316517e-01 6.89093769e-01 -1.16758001e+00 -6.87265933e-01
3.85383964e-01 1.79784447e-01 -2.48172849e-01 4.24850360e-02
4.61271368e-02 1.72130451e-01 -3.19607198e-01 6.33656502e-01
7.37230062e-01 -5.63112020e-01 -2.34103456e-01 -1.27793759e-01
-1.98611483e-01 1.75678357e-02 -1.47846615e+00 1.50536764e+00
-3.17969918e-01 2.05306977e-01 -3.65713716e-01 -9.71826673e-01
9.31797147e-01 -8.02426413e-02 4.41817164e-01 -8.20344567e-01
4.90836017e-02 3.61430973e-01 -4.49670814e-02 -2.93489397e-01
6.65863335e-01 -6.11712709e-02 -1.74956590e-01 4.62744743e-01
7.39664435e-02 1.07758269e-01 6.72023475e-01 2.61445582e-01
9.18452799e-01 -4.79630947e-01 5.55007339e-01 -2.47235969e-01
6.51094258e-01 5.40341735e-02 4.04670209e-01 8.50070298e-01
-2.25202024e-01 7.56483734e-01 3.17678303e-01 -4.22791272e-01
-9.93285358e-01 -9.96159732e-01 -2.24400252e-01 1.20377183e+00
5.76025784e-01 -5.33998549e-01 -6.39397264e-01 -6.89776599e-01
1.58008575e-01 6.84032261e-01 -7.98878491e-01 -1.24635346e-01
-5.45343816e-01 -4.47081774e-01 1.91805527e-01 6.99820101e-01
4.90923643e-01 -1.00799024e+00 -9.72902358e-01 6.52795061e-02
-1.36137074e-02 -1.03967929e+00 -6.79361701e-01 1.36957601e-01
-6.19605601e-01 -1.04336989e+00 -9.83841360e-01 -7.78329492e-01
7.43388653e-01 8.48757029e-01 1.05378926e+00 3.48620474e-01
-4.95780170e-01 5.93975596e-02 -5.22195876e-01 -2.84755558e-01
9.83837843e-02 -8.53665397e-02 6.79088607e-02 5.76619282e-02
3.01005095e-01 -1.67507961e-01 -8.76779079e-01 5.39539635e-01
-8.05122733e-01 6.05143346e-02 6.38989508e-01 9.61466670e-01
9.79740083e-01 8.39123428e-02 4.07435119e-01 -7.40865529e-01
2.68574715e-01 -4.91213500e-01 -5.71575463e-01 4.99899656e-01
-4.55115259e-01 -7.63920173e-02 6.10652328e-01 -5.22457957e-01
-6.38023973e-01 1.40704349e-01 2.60017574e-01 -4.72777814e-01
-1.62303954e-01 2.64073402e-01 1.00146726e-01 1.13570943e-01
6.28507435e-01 4.48042482e-01 -2.05236688e-01 -2.76921302e-01
3.80069435e-01 4.75764602e-01 6.51336968e-01 -4.65724319e-01
8.08083713e-01 6.45030081e-01 -1.45088091e-01 -7.81857908e-01
-9.98116612e-01 -1.03015709e+00 -6.43027484e-01 -1.47320241e-01
6.28235102e-01 -7.77041614e-01 -4.22410101e-01 1.44852147e-01
-9.25426900e-01 6.32945523e-02 -4.81706738e-01 4.62404013e-01
-4.26452875e-01 2.94085294e-01 -3.07216555e-01 -7.84305453e-01
-3.99881363e-01 -9.60343719e-01 1.16591859e+00 4.54797089e-01
-1.16501890e-01 -4.12327111e-01 -4.43297485e-03 1.26032025e-01
2.26158738e-01 6.41447082e-02 8.04920435e-01 -7.78683305e-01
-5.97679555e-01 -4.88065362e-01 -5.44800401e-01 2.33734846e-01
1.09319791e-01 -7.92168751e-02 -8.35689783e-01 -3.19219977e-01
1.64643377e-02 -2.18691334e-01 1.15528584e+00 2.24402100e-01
1.39279032e+00 -8.15756503e-04 -4.06286478e-01 3.95654827e-01
1.54874730e+00 1.50072828e-01 5.39775312e-01 2.96778560e-01
4.46196228e-01 4.25998539e-01 1.04720354e+00 6.02504969e-01
1.08673796e-01 7.09975481e-01 4.62831110e-01 -3.15023661e-02
-4.17484120e-02 -1.78581506e-01 -1.33926347e-01 5.26715994e-01
-1.23762324e-01 -7.44528323e-02 -7.57661879e-01 7.35499620e-01
-1.87547076e+00 -1.01985991e+00 -3.19314189e-02 2.55900168e+00
5.57709396e-01 2.71932214e-01 2.98272550e-01 2.08741248e-01
9.90535796e-01 3.42777669e-01 -4.74305630e-01 9.39319208e-02
-2.18659174e-02 1.13419652e-01 4.70189750e-01 4.51866873e-02
-1.17836940e+00 7.62889385e-01 5.18967009e+00 1.33829117e+00
-9.56086814e-01 8.42417330e-02 7.48450637e-01 -3.62215370e-01
9.48456302e-02 -3.40657383e-02 -9.55166996e-01 7.45567620e-01
3.98006737e-01 -3.12470883e-01 3.79860431e-01 1.04685152e+00
-1.59644157e-01 -5.47592103e-01 -1.02948141e+00 1.07709980e+00
2.01858819e-01 -1.40401602e+00 -1.04108617e-01 -1.10467263e-01
6.80663705e-01 -9.66911167e-02 6.50522485e-02 3.71820360e-01
-7.92969242e-02 -9.00324643e-01 8.75655711e-01 3.88973206e-01
9.92267668e-01 -7.49481320e-01 6.46413684e-01 5.41203856e-01
-1.53013217e+00 -2.61654943e-01 -7.01392114e-01 -5.06338775e-02
7.73071721e-02 6.73055887e-01 -5.44633806e-01 3.54873359e-01
7.88588166e-01 5.71142554e-01 -6.86133444e-01 1.27824724e+00
-7.17653930e-02 2.07177520e-01 -2.90271014e-01 -2.16057345e-01
2.83211321e-01 -1.69614881e-01 3.85534793e-01 1.26890802e+00
5.01867890e-01 2.08196521e-01 1.42521515e-01 8.69877398e-01
-1.81789607e-01 2.14186266e-01 -5.14857292e-01 1.56604663e-01
6.96094275e-01 1.53183234e+00 -1.01348412e+00 -6.87081099e-01
-5.21744192e-01 9.61822212e-01 4.67407763e-01 1.52652459e-02
-1.01925397e+00 -5.66039145e-01 3.92913729e-01 2.16743618e-01
6.26623929e-01 6.54629096e-02 -2.31077984e-01 -1.03608525e+00
2.45309278e-01 -5.38566411e-01 5.61204731e-01 -5.23427248e-01
-1.24643624e+00 4.76626396e-01 1.57508086e-02 -1.50926638e+00
3.69937439e-03 -5.54992974e-01 -8.38769734e-01 7.43773341e-01
-1.40818822e+00 -8.43164265e-01 -5.50083160e-01 5.52007854e-01
6.04684830e-01 -5.11003621e-02 6.80229723e-01 4.00738657e-01
-5.92513204e-01 5.87122381e-01 2.36489281e-01 1.44474031e-02
6.46555781e-01 -1.18072033e+00 3.22688997e-01 7.06761479e-01
4.51329052e-01 6.59001410e-01 4.28322613e-01 -3.81850749e-01
-1.07904136e+00 -1.10998833e+00 6.68439984e-01 -3.57189506e-01
6.06044054e-01 -2.63401598e-01 -1.06614769e+00 1.07932366e-01
-1.79217726e-01 5.43551445e-01 4.08959955e-01 -4.46227416e-02
-5.10266602e-01 -2.31363684e-01 -1.12091959e+00 5.03609061e-01
1.15420032e+00 -5.47324359e-01 -5.03420532e-01 1.48328245e-01
4.11790311e-01 -1.44250005e-01 -6.23331904e-01 4.18574512e-01
5.55631876e-01 -1.08415246e+00 1.06760383e+00 -3.06122601e-01
5.23635447e-01 -5.60363948e-01 -3.07455391e-01 -1.05330646e+00
-5.29064655e-01 9.64506716e-02 -2.21436709e-01 1.16128910e+00
2.37161055e-01 -4.28059489e-01 6.79373860e-01 2.54322678e-01
4.61967662e-02 -1.08880317e+00 -9.79729831e-01 -1.05446398e+00
-2.04827949e-01 -2.98099250e-01 6.20700121e-01 7.58920014e-01
1.30997971e-02 2.41972789e-01 -2.70276755e-01 -4.29049917e-02
5.63711345e-01 5.01251280e-01 7.06636965e-01 -1.05573261e+00
-3.06539923e-01 -5.95346570e-01 -4.87424046e-01 -1.09559095e+00
-3.97195727e-01 -7.74343610e-01 1.55148074e-01 -1.42414486e+00
8.43105257e-01 -5.12959182e-01 -6.49514258e-01 2.94181138e-01
-4.62520748e-01 4.96994376e-01 4.57880259e-01 4.21341866e-01
-1.07039595e+00 3.73668522e-01 1.12122726e+00 -1.80533543e-01
1.14807993e-01 -9.79099199e-02 -7.41548479e-01 7.17519045e-01
8.16392541e-01 -5.79823017e-01 -9.08518732e-02 -2.26611480e-01
-1.10260464e-01 -2.43088335e-01 5.11282146e-01 -1.22324765e+00
3.96251321e-01 -1.52246445e-01 4.75876361e-01 -6.67300045e-01
3.83459479e-01 -7.49930203e-01 -6.17074445e-02 6.37765706e-01
-2.52509743e-01 -1.45778969e-01 -4.16826755e-02 6.04122877e-01
-2.21253306e-01 -5.28819561e-01 1.00681520e+00 -1.92673489e-01
-1.02127016e+00 2.09221750e-01 2.04422414e-01 1.43984780e-01
1.37726808e+00 -4.60836858e-01 -5.85775614e-01 5.79148764e-03
-3.22770953e-01 2.37745196e-01 6.11788273e-01 2.18427494e-01
7.31594324e-01 -1.42836010e+00 -6.16022527e-01 1.76368311e-01
6.92635179e-01 1.46044577e-02 3.64585489e-01 9.54438925e-01
-2.23815575e-01 3.04264724e-01 -1.61827415e-01 -7.22969174e-01
-1.32939196e+00 7.78726101e-01 6.54654428e-02 -3.28131855e-01
-6.11078680e-01 8.55647326e-01 4.80879068e-01 5.50665557e-02
2.58592278e-01 -2.15884402e-01 -1.92554295e-01 2.44262248e-01
9.27312970e-01 5.40907681e-01 -7.56483227e-02 -6.84028029e-01
-4.18649614e-01 8.09858024e-01 -2.82041669e-01 2.66903043e-01
1.21505189e+00 9.98930037e-02 1.66759312e-01 3.55649859e-01
1.36308849e+00 -9.07500759e-02 -1.29288268e+00 -5.62162220e-01
8.10552910e-02 -9.11063492e-01 -2.98045874e-01 -5.46091437e-01
-9.95063484e-01 7.78598070e-01 7.44736671e-01 2.50936627e-01
1.16478086e+00 3.82818580e-01 6.55808806e-01 3.08405101e-01
5.00272989e-01 -1.05896032e+00 4.97762501e-01 4.95741427e-01
6.95683062e-01 -1.52723515e+00 1.23840198e-01 -3.59734058e-01
-7.91221559e-01 8.38193655e-01 4.97704476e-01 -3.71746480e-01
3.54710340e-01 1.62212506e-01 -3.70894313e-01 -4.10426617e-01
-4.73687679e-01 -4.66618210e-01 4.32316661e-01 2.06609771e-01
1.52535155e-01 1.74369782e-01 -3.34188640e-01 5.23430467e-01
1.74042881e-01 -7.49792233e-02 1.82749301e-01 8.39093268e-01
-8.54744136e-01 -6.25820220e-01 -3.70630354e-01 7.81743586e-01
-3.05287540e-01 -2.09734350e-01 -2.67188400e-02 6.66553974e-01
3.67557079e-01 7.64056683e-01 2.98147202e-01 -3.95243555e-01
4.04404193e-01 -2.92939961e-01 3.72796178e-01 -7.92087197e-01
-2.28252023e-01 -1.59160525e-01 -2.04974234e-01 -5.19037724e-01
-2.80463606e-01 -5.65812469e-01 -1.32557333e+00 -2.19571337e-01
-5.59764504e-01 -7.71888793e-02 3.88083190e-01 6.77262247e-01
1.39190227e-01 3.54125082e-01 7.22470760e-01 -8.13362420e-01
-6.44021809e-01 -8.46534133e-01 -7.72798181e-01 6.43870950e-01
1.21682599e-01 -9.46456194e-01 -4.38901246e-01 -1.68564364e-01] | [9.296103477478027, 0.9511836171150208] |
552ac85e-7eb7-4358-b2b7-8117ed2f96a1 | camera-intrinsic-blur-kernel-estimation-a | null | null | http://openaccess.thecvf.com/content_cvpr_2015/html/Mosleh_Camera_Intrinsic_Blur_2015_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2015/papers/Mosleh_Camera_Intrinsic_Blur_2015_CVPR_paper.pdf | Camera Intrinsic Blur Kernel Estimation: A Reliable Framework | This paper presents a reliable non-blind method to measure intrinsic lens blur. We first introduce an accurate camera-scene alignment framework that avoids erroneous homography estimation and camera tone curve estimation. This alignment is used to generate a sharp correspondence of a target pattern captured by the camera. Second, we introduce a Point Spread Function (PSF) estimation approach where information about the frequency spectrum of the target image is taken into account. As a result of these steps and the ability to use multiple target images in this framework, we achieve a PSF estimation method robust against noise and suitable for mobile devices. Experimental results show that the proposed method results in PSFs with more than 10 dB higher accuracy in noisy conditions compared with the PSFs generated using state-of-the-art techniques. | ['J. M. Pierre Langlois', 'Emmanuel Onzon', 'Ali Mosleh', 'Paul Green', 'Isabelle Begin'] | 2015-06-01 | null | null | null | cvpr-2015-6 | ['homography-estimation'] | ['computer-vision'] | [ 5.53188503e-01 -4.60931391e-01 5.96342683e-01 -6.58260062e-02
-6.42374516e-01 -6.02393389e-01 6.53721988e-01 -2.74734288e-01
-3.98701191e-01 7.87385583e-01 2.78077126e-01 5.45592643e-02
-1.58135071e-01 -4.08192784e-01 -6.86964393e-01 -6.21139526e-01
2.89537549e-01 -6.76986121e-04 4.47579563e-01 1.89507052e-01
5.61806977e-01 4.73100126e-01 -1.59978366e+00 -1.93345636e-01
8.25505257e-01 9.07343149e-01 5.24349272e-01 9.83667493e-01
2.93093592e-01 6.67152822e-01 -7.58644223e-01 -4.75023121e-01
3.97692204e-01 -3.78233016e-01 -2.56540298e-01 6.09592080e-01
5.80370188e-01 -6.41578555e-01 -6.06807843e-02 1.45465171e+00
4.36764270e-01 -1.15256965e-01 6.97674453e-01 -4.89055961e-01
-3.85952413e-01 -8.95782001e-03 -5.58077812e-01 -4.13309112e-02
7.26261675e-01 1.47869259e-01 3.62873197e-01 -8.96042883e-01
3.28869492e-01 7.29530096e-01 5.83639920e-01 -7.51019046e-02
-1.28048038e+00 -2.10406348e-01 -8.23344290e-01 1.33603320e-01
-1.41526628e+00 -5.83588064e-01 8.02358687e-01 -4.67807829e-01
3.50232840e-01 3.33010942e-01 5.52915096e-01 8.65003407e-01
4.15334880e-01 -1.62305217e-03 1.66102219e+00 -8.05347741e-01
6.14517704e-02 4.63423610e-01 6.78996593e-02 3.98414403e-01
5.62085748e-01 4.16709274e-01 -2.01513514e-01 -5.02152964e-02
1.12170529e+00 -2.94369817e-01 -9.63330865e-01 -3.80630612e-01
-1.35099685e+00 2.10555211e-01 7.07491562e-02 5.71238279e-01
-4.29176420e-01 1.99166462e-01 -1.66905776e-01 -3.60475220e-02
9.15739238e-02 5.83315969e-01 3.78323734e-01 -3.73760730e-01
-1.00019062e+00 5.17473780e-02 6.78867877e-01 1.00933862e+00
5.55200338e-01 4.18862626e-02 -1.51879936e-01 6.98207557e-01
3.30619365e-01 8.31431150e-01 3.57353359e-01 -1.01275730e+00
1.59737155e-01 1.58173293e-01 6.70820475e-01 -1.07281554e+00
1.33121714e-01 -6.17707849e-01 -4.83085126e-01 5.59241295e-01
6.30554318e-01 9.33652595e-02 -6.40223563e-01 1.14920318e+00
6.65735975e-02 4.13042337e-01 9.10953954e-02 1.05331206e+00
2.42720500e-01 3.95226479e-01 -7.81101346e-01 -5.94245076e-01
1.37059104e+00 -5.87210536e-01 -1.04708612e+00 1.96015090e-02
-3.14286977e-01 -1.64135087e+00 8.10554922e-01 8.00771594e-01
-1.19258165e+00 -7.60595500e-01 -1.19798756e+00 1.74105644e-01
2.18601152e-01 5.10615647e-01 -2.47869249e-02 9.91041839e-01
-1.07535267e+00 3.16509306e-01 -3.75584453e-01 -3.55037630e-01
-2.22185791e-01 1.46669790e-01 -2.59382844e-01 -7.01123774e-02
-5.43516099e-01 1.11635363e+00 1.04159422e-01 -7.51818419e-02
-4.87537801e-01 -4.21652168e-01 -5.95077097e-01 1.19893618e-01
2.12197214e-01 -6.28518403e-01 1.36083257e+00 -9.27578270e-01
-1.94509375e+00 6.84011519e-01 -4.06911284e-01 -4.34275508e-01
6.47877753e-01 -3.85616481e-01 -3.66225123e-01 5.99199653e-01
-2.61000991e-01 -2.20295966e-01 1.30977857e+00 -1.54546571e+00
-4.13157851e-01 5.09933904e-02 -2.22035900e-01 2.63460636e-01
-2.73477733e-02 1.45247176e-01 -6.34698510e-01 -4.30255085e-01
6.70287609e-02 -5.97729087e-01 1.38532415e-01 -2.16765672e-01
-2.23920166e-01 5.63088000e-01 7.48105645e-01 -8.36674929e-01
1.10343850e+00 -2.09236240e+00 -8.91219079e-02 1.31378084e-01
1.38549611e-01 4.20142382e-01 2.63453633e-01 4.95249510e-01
9.47473198e-02 -5.85895836e-01 -1.43546745e-01 -3.18589032e-01
-3.89102846e-01 -2.65884936e-01 -2.00016946e-01 7.43352771e-01
-2.93408930e-01 4.43848997e-01 -6.08046234e-01 -2.79548138e-01
7.17791855e-01 6.67686641e-01 -4.14044335e-02 4.76537555e-01
2.21729249e-01 5.44919491e-01 9.36504081e-02 4.44003999e-01
9.98595893e-01 4.12349999e-02 -2.92960741e-03 -5.34439921e-01
-3.75383139e-01 -1.69038981e-01 -1.33922708e+00 1.23431814e+00
-4.56449449e-01 7.64360964e-01 7.75323957e-02 -2.91047722e-01
1.23057544e+00 3.79576772e-01 2.10337192e-01 -3.65123630e-01
3.30984503e-01 6.05168760e-01 -1.23618565e-01 -4.66225892e-01
5.56866348e-01 -5.08142300e-02 4.34464574e-01 1.14200987e-01
4.32134122e-02 -5.22518456e-01 -4.06670719e-02 -1.18588366e-01
7.45291948e-01 -2.65261550e-02 5.89045584e-01 -2.88026661e-01
9.60425317e-01 -3.47475052e-01 2.89159846e-02 7.00703204e-01
-6.31924197e-02 1.06581140e+00 7.89733008e-02 -4.36341800e-02
-1.26979208e+00 -1.00890350e+00 -1.58532664e-01 -2.75988221e-01
3.70092660e-01 -1.18772373e-01 -1.17849624e+00 -8.02921802e-02
-3.13632697e-01 7.61047661e-01 -8.29842538e-02 2.87057638e-01
-4.48794752e-01 -4.12291288e-01 7.97271878e-02 -5.54216886e-03
9.39490616e-01 -3.76545370e-01 -7.91298270e-01 1.20613270e-01
-1.16157115e-01 -1.43184614e+00 -6.26612484e-01 -5.15966177e-01
-6.51708424e-01 -1.32402444e+00 -1.09169567e+00 -5.27538717e-01
8.46132159e-01 6.39611900e-01 8.87624025e-01 -3.36981505e-01
-1.33942366e-01 8.28496933e-01 -2.28828326e-01 -4.23539393e-02
-5.89145124e-01 -7.71261930e-01 -1.53022073e-02 6.17648244e-01
2.41239786e-01 -5.36978722e-01 -6.61936879e-01 4.54351693e-01
-7.10087299e-01 4.60370891e-02 7.44197428e-01 5.13032615e-01
2.24964440e-01 3.42817992e-01 -1.12869732e-01 -5.62093437e-01
7.30135560e-01 1.71477780e-01 -1.34093428e+00 5.07838838e-02
-7.62035906e-01 -2.43824422e-01 5.30611098e-01 -4.46502477e-01
-1.43680656e+00 1.91260457e-01 3.20610791e-01 -5.41633189e-01
-3.01600784e-01 1.18756495e-01 -6.57080486e-02 -5.63991725e-01
9.31932986e-01 4.93040115e-01 3.17678601e-01 -5.21440268e-01
2.25415541e-04 1.03236055e+00 1.17375517e+00 -1.62999704e-01
1.10768104e+00 4.55550253e-01 2.69063801e-01 -1.16387010e+00
-1.71628579e-01 -8.13721001e-01 -4.84222412e-01 -5.94311178e-01
8.43992651e-01 -9.00859237e-01 -7.45254397e-01 1.03373063e+00
-1.35569382e+00 1.21519990e-01 1.25004634e-01 1.00405943e+00
-6.20696485e-01 7.89752185e-01 -4.16788608e-01 -1.12666011e+00
-4.20155264e-02 -1.28111041e+00 1.01792014e+00 5.13096988e-01
1.55580759e-01 -8.17096055e-01 -7.98959956e-02 4.99412030e-01
6.58966660e-01 2.18780264e-02 9.62339863e-02 1.53998733e-01
-1.21580720e+00 -3.65279108e-01 -3.06716383e-01 5.86365461e-01
4.75481182e-01 -3.19899954e-02 -1.20379210e+00 -2.31074080e-01
7.65690684e-01 4.18876648e-01 3.68794650e-01 5.81781507e-01
4.84565586e-01 -1.96315840e-01 -3.82260047e-03 6.20563984e-01
2.03830242e+00 1.66464448e-01 1.13455176e+00 3.55583876e-01
4.32006836e-01 3.68997037e-01 7.14885533e-01 1.75664082e-01
-1.92913800e-01 1.17497885e+00 3.71054381e-01 -1.09418437e-01
-5.39332330e-01 8.02580416e-02 2.77951360e-01 4.60317552e-01
-1.38718814e-01 -2.26598874e-01 -6.16863370e-01 5.33398211e-01
-1.47317588e+00 -9.99061167e-01 -7.62351096e-01 2.81845045e+00
5.71635127e-01 3.51123139e-02 -7.48915672e-02 2.73161352e-01
8.68938029e-01 -2.18733117e-01 1.39437944e-01 -2.01584801e-01
-1.87955469e-01 8.91131908e-02 9.61697102e-01 1.03802323e+00
-8.85743976e-01 4.90287930e-01 6.40895176e+00 5.53486466e-01
-1.19121742e+00 -9.71349776e-02 8.38765427e-02 4.17585224e-01
-9.08016115e-02 -2.67047118e-02 -5.80605447e-01 7.30193436e-01
7.69099355e-01 -2.45787263e-01 6.54967070e-01 4.31504965e-01
5.01548409e-01 -7.55244076e-01 -6.86451435e-01 1.30856276e+00
4.24948424e-01 -9.30423677e-01 -3.65294963e-01 3.07989895e-01
6.55089319e-01 -4.52566564e-01 1.14596702e-01 -7.88675249e-01
-1.81267187e-01 -7.32371449e-01 6.28435135e-01 9.59627211e-01
7.88631916e-01 -5.41206777e-01 8.94169569e-01 1.98131293e-01
-8.30886960e-01 2.10141852e-01 -3.81470591e-01 -3.77936698e-02
3.26703310e-01 9.54234064e-01 -1.26492000e+00 6.77015126e-01
2.95554876e-01 3.01982701e-01 -5.10907888e-01 1.74198079e+00
-2.78761476e-01 4.99524921e-01 -1.84927985e-01 6.60995468e-02
-2.58995742e-01 -6.27239287e-01 1.01918340e+00 1.02183330e+00
8.68077159e-01 -6.50451407e-02 -3.61947924e-01 9.49717045e-01
3.99452984e-01 6.57317694e-04 -8.21158946e-01 2.33633995e-01
2.85687834e-01 1.17915082e+00 -5.36265969e-01 -7.86420405e-02
-5.41297376e-01 1.29398096e+00 -4.63498414e-01 5.15508533e-01
-6.98512077e-01 -3.61081868e-01 2.95507669e-01 2.01179683e-01
3.06658924e-01 -4.45236206e-01 -2.03736961e-01 -1.07739317e+00
1.66779727e-01 -8.79396915e-01 -3.46930772e-01 -1.27787519e+00
-9.25281823e-01 5.69015384e-01 4.18491382e-03 -1.65253472e+00
-2.99629331e-01 -7.89856434e-01 -4.99349773e-01 1.36449993e+00
-1.43004680e+00 -1.07784092e+00 -4.65666890e-01 5.25878608e-01
4.11778837e-01 -2.57590413e-01 6.87183917e-01 1.69663712e-01
1.72214702e-01 1.05807357e-01 3.52297574e-01 -2.18019515e-01
7.98549712e-01 -1.39239740e+00 1.20910816e-02 1.45076156e+00
-8.35736394e-02 7.54747570e-01 1.13161838e+00 -6.16520166e-01
-1.25342035e+00 -5.77725291e-01 8.75925779e-01 -4.30213571e-01
4.19350743e-01 -1.61919028e-01 -6.55790985e-01 3.56362671e-01
5.03581464e-01 -1.33614317e-01 2.82136351e-01 -5.94217598e-01
-3.69021785e-03 -3.30662400e-01 -1.07772934e+00 3.41651857e-01
3.49245816e-01 -6.16245568e-01 -6.95283949e-01 4.33671586e-02
2.55840212e-01 -6.59860194e-01 -5.65839171e-01 2.81217068e-01
5.74951053e-01 -1.64432585e+00 8.81872237e-01 7.28640020e-01
6.45028939e-03 -8.50562334e-01 -1.58254787e-01 -1.17111325e+00
-2.33825803e-01 -1.04861307e+00 1.70919061e-01 1.18514299e+00
1.30717888e-01 -7.81356931e-01 3.56376886e-01 2.53952712e-01
3.29242274e-02 7.26717487e-02 -6.34371638e-01 -8.29862535e-01
-9.29454207e-01 1.64735485e-02 1.49562657e-01 5.40054500e-01
-3.20761919e-01 2.94178069e-01 -7.08011448e-01 4.72415417e-01
1.20112026e+00 -2.60879882e-02 1.02648795e+00 -1.10743737e+00
-6.28026485e-01 -1.70324855e-02 -7.71703005e-01 -9.70117390e-01
-3.96399528e-01 3.15081538e-03 -1.93089973e-02 -1.26135743e+00
1.50505200e-01 1.45200536e-01 1.74391940e-01 -6.13020837e-01
-1.58007652e-01 4.96526748e-01 1.47748634e-01 4.02197212e-01
6.81342781e-02 5.81018738e-02 1.13028312e+00 4.33585972e-01
-7.30260164e-02 4.44075167e-01 -3.95741522e-01 6.67699993e-01
4.98869956e-01 -2.38006450e-02 -5.11116266e-01 -3.46560240e-01
-2.11486280e-01 9.02502462e-02 6.13004386e-01 -1.59794521e+00
3.08220983e-01 2.05943510e-01 4.87753838e-01 -4.11010176e-01
5.15925646e-01 -1.25596595e+00 6.60702288e-01 4.02448654e-01
1.68592855e-01 -1.68278873e-01 -1.34204933e-03 6.17009223e-01
-3.18197817e-01 -5.82420886e-01 1.02786744e+00 -1.49683177e-01
-3.12456638e-01 -4.74908561e-01 -2.34101474e-01 -6.66254997e-01
1.00016761e+00 -6.11309826e-01 -6.56561494e-01 -7.75757492e-01
-3.10724139e-01 -6.35326087e-01 9.53531384e-01 -1.25652745e-01
6.76258624e-01 -1.07230186e+00 -4.92949933e-01 3.51701081e-01
-1.31972089e-01 -4.62633938e-01 2.13580448e-02 9.86403704e-01
-1.02360940e+00 4.59665507e-01 -1.88515663e-01 -7.36279666e-01
-1.58541787e+00 4.97846931e-01 5.17630816e-01 1.68919206e-01
-3.94736469e-01 4.42634523e-01 -6.26148582e-02 3.23294252e-01
8.98940414e-02 -2.00399175e-01 -3.05490881e-01 -5.64697266e-01
7.37212718e-01 4.17150378e-01 6.40538186e-02 -8.44507217e-01
-1.01823593e-02 1.04476678e+00 3.76863539e-01 -5.62472224e-01
9.24138784e-01 -5.05243301e-01 -1.18451715e-01 3.47709268e-01
9.32899594e-01 9.42002833e-01 -1.13585818e+00 -5.08454405e-02
-1.89706177e-01 -1.07277560e+00 2.72230953e-01 -8.57333362e-01
-4.89973456e-01 6.73158705e-01 9.00408030e-01 2.90452272e-01
1.40675676e+00 -4.98348027e-01 4.26872402e-01 -1.35266498e-01
4.44425583e-01 -8.70695174e-01 4.18647192e-03 -4.77540120e-02
8.17629039e-01 -8.61461401e-01 4.46896106e-02 -8.53347301e-01
-3.38120997e-01 1.35828066e+00 2.51888841e-01 -3.86352800e-02
2.28982896e-01 2.43905276e-01 2.41155759e-01 1.29408196e-01
-9.43871886e-02 -4.42854345e-01 4.22367036e-01 7.75164902e-01
1.83696091e-01 -2.60701716e-01 -4.83248413e-01 -8.40590894e-02
-7.47466311e-02 2.40003049e-01 1.08747530e+00 5.77603817e-01
-7.01232970e-01 -1.17093658e+00 -1.10808718e+00 -7.46608675e-02
-6.49036109e-01 -4.63077426e-02 -2.80829281e-01 5.82781553e-01
-1.79157645e-01 1.13944793e+00 -2.41156653e-01 -1.62969932e-01
3.47821355e-01 -2.25711152e-01 7.69051015e-01 -3.00640315e-01
-2.54662603e-01 7.37444937e-01 1.06848106e-01 -4.71906394e-01
-5.80652058e-01 -4.99603659e-01 -5.65011978e-01 5.60517481e-04
-4.86833572e-01 2.99597271e-02 8.76676500e-01 5.32025158e-01
9.46845338e-02 2.44201139e-01 7.69057333e-01 -7.64014781e-01
-4.06293690e-01 -9.45229828e-01 -7.18532264e-01 3.20861191e-01
5.57410955e-01 -4.72674072e-01 -7.00310290e-01 4.17154908e-01] | [11.552196502685547, -2.67510986328125] |
dacffbb5-4ec3-41f1-b8a3-543786fd0e64 | visual-discourse-parsing | 1903.02252 | null | https://arxiv.org/abs/1903.02252v4 | https://arxiv.org/pdf/1903.02252v4.pdf | Discourse Parsing in Videos: A Multi-modal Appraoch | Text-level discourse parsing aims to unmask how two sentences in the text are related to each other. We propose the task of Visual Discourse Parsing, which requires understanding discourse relations among scenes in a video. Here we use the term scene to refer to a subset of video frames that can better summarize the video. In order to collect a dataset for learning discourse cues from videos, one needs to manually identify the scenes from a large pool of video frames and then annotate the discourse relations between them. This is clearly a time consuming, expensive and tedious task. In this work, we propose an approach to identify discourse cues from the videos without the need to explicitly identify and annotate the scenes. We also present a novel dataset containing 310 videos and the corresponding discourse cues to evaluate our approach. We believe that many of the multi-discipline AI problems such as Visual Dialog and Visual Storytelling would greatly benefit from the use of visual discourse cues. | ['Arjun R. Akula', 'Song-Chun Zhu'] | 2019-03-06 | null | null | null | null | ['visual-storytelling'] | ['natural-language-processing'] | [ 4.96202558e-01 1.19860910e-01 -8.39062929e-02 -4.15848255e-01
-7.30510652e-01 -8.15998316e-01 7.33319581e-01 3.27625483e-01
-1.34725437e-01 5.17894804e-01 6.34531796e-01 -2.34582081e-01
3.70742798e-01 -3.01815540e-01 -8.74852836e-01 -5.89356482e-01
4.06255014e-02 1.92318261e-01 3.58146340e-01 -1.22006655e-01
5.18688858e-01 8.19188654e-02 -1.60680091e+00 1.09947920e+00
2.60436505e-01 6.33489907e-01 6.42568171e-01 8.11047852e-01
-3.70449692e-01 1.59144580e+00 -8.46276760e-01 -1.44658566e-01
-2.45933473e-01 -8.06310654e-01 -1.12400448e+00 6.46132469e-01
6.51082933e-01 -6.80861413e-01 -1.34294912e-01 9.14953291e-01
4.38444726e-02 3.35903257e-01 6.36182070e-01 -1.27503312e+00
-2.06046820e-01 6.69851542e-01 -6.13318980e-01 4.88889843e-01
9.59499359e-01 -8.70453194e-02 9.41520274e-01 -7.87947953e-01
1.28759015e+00 1.48777008e+00 2.20876746e-03 3.94067377e-01
-9.50002849e-01 -2.93172866e-01 4.51012284e-01 4.92790461e-01
-7.85690129e-01 -6.38274014e-01 1.07096577e+00 -9.03018177e-01
6.28985047e-01 4.17139858e-01 6.19675636e-01 1.29391420e+00
-4.32307094e-01 1.13771808e+00 9.60166752e-01 -7.01161325e-01
3.99674587e-02 4.60324250e-02 1.14170350e-01 6.73343956e-01
-3.24069262e-01 -3.87010485e-01 -7.63961136e-01 2.17270598e-01
7.79400587e-01 -2.03987002e-01 -3.55486959e-01 -4.01157975e-01
-1.39527512e+00 7.22216666e-01 2.04153895e-01 4.23472553e-01
-2.34375581e-01 1.69056192e-01 4.50981945e-01 1.78102311e-02
3.04383129e-01 4.45106059e-01 1.52212813e-01 -3.72218460e-01
-7.66620874e-01 2.23921061e-01 6.56684399e-01 8.53684723e-01
7.36271143e-01 -5.53294979e-02 -3.80814701e-01 5.43743432e-01
1.49732903e-01 1.60154864e-01 -1.12064987e-01 -1.50801754e+00
6.66249275e-01 7.16098070e-01 9.20957327e-02 -1.40809190e+00
-3.20636898e-01 6.15298748e-01 -2.25324631e-01 -1.15966415e-02
8.55774462e-01 -1.21371329e-01 -6.67010963e-01 1.50994492e+00
5.69792807e-01 1.15281664e-01 1.35202259e-02 1.11176741e+00
1.35473442e+00 8.74483287e-01 2.36690715e-01 -3.89328510e-01
1.71749794e+00 -8.51054907e-01 -9.56771672e-01 -3.13797653e-01
6.74950838e-01 -9.00961101e-01 1.16070151e+00 1.86652988e-02
-1.15712690e+00 -4.60121185e-01 -1.00248396e+00 -5.09523869e-01
-2.55602568e-01 1.74656138e-01 5.42056024e-01 5.04734516e-02
-6.58256531e-01 1.49793997e-01 -7.36736953e-01 -3.95893663e-01
3.42740029e-01 -1.62926003e-01 -3.50932449e-01 -5.15452363e-02
-8.42063665e-01 8.09599102e-01 4.66144174e-01 -4.49802838e-02
-9.75003481e-01 -3.13485622e-01 -1.18077099e+00 -1.15243107e-01
7.22386122e-01 -2.48759508e-01 1.24554729e+00 -1.37113059e+00
-1.14599311e+00 1.22881138e+00 -4.86328393e-01 -1.88868046e-01
2.97995538e-01 -3.31742913e-01 4.42949794e-02 9.31196868e-01
1.76196575e-01 8.22068155e-01 8.74104202e-01 -1.47960317e+00
-7.38176048e-01 -2.16401622e-01 7.42189944e-01 3.91475052e-01
-5.35347760e-02 4.22025442e-01 -6.67596519e-01 -6.10298395e-01
-1.24552757e-01 -8.46360028e-01 2.46470988e-01 -1.17234729e-01
-5.20834982e-01 -2.79450864e-01 1.26795447e+00 -8.61671448e-01
9.38233197e-01 -2.27857518e+00 4.34893370e-01 -3.25887412e-01
3.51035029e-01 -5.16006164e-02 1.16275668e-01 4.34325308e-01
5.06897420e-02 -9.17259827e-02 -1.78495366e-02 -1.09503508e-01
-3.98290724e-01 2.00958192e-01 -3.56651038e-01 4.99490231e-01
1.96033284e-01 8.58902395e-01 -1.14165747e+00 -1.01471376e+00
3.90759468e-01 4.51439053e-01 -3.74919564e-01 3.54483604e-01
-6.53051555e-01 8.25227439e-01 -4.93200421e-01 5.38976908e-01
8.90363753e-02 -4.78287697e-01 3.57803464e-01 -3.69672775e-01
-1.40243873e-01 2.51681894e-01 -8.27608347e-01 1.79854488e+00
8.78188834e-02 1.63323247e+00 1.61887497e-01 -1.19456851e+00
6.07013166e-01 4.31135505e-01 4.27564859e-01 -5.54460585e-01
3.48039418e-01 -2.64673799e-01 -2.06213787e-01 -1.29231775e+00
6.20062888e-01 1.48498714e-01 -2.69680113e-01 5.72503805e-01
-1.09658860e-01 -8.86501465e-03 7.00595677e-01 3.37469220e-01
6.62329912e-01 3.37531328e-01 3.66852045e-01 4.37601507e-02
4.82480198e-01 4.47555929e-01 2.71346539e-01 4.77001518e-01
-1.37912318e-01 4.81829494e-01 8.95984650e-01 -4.49669033e-01
-1.05070484e+00 -6.47030175e-01 3.51222545e-01 1.38075459e+00
3.00865889e-01 -5.85123062e-01 -8.19484055e-01 -4.61034924e-01
-4.22748566e-01 5.26356399e-01 -6.86678708e-01 4.91757065e-01
-9.84930992e-01 -1.65566817e-01 1.36117354e-01 4.73923296e-01
2.83322334e-01 -1.03343666e+00 -1.09992206e+00 -2.35809878e-01
-7.84791648e-01 -1.54037511e+00 -3.47599417e-01 -4.07502912e-02
-4.39049959e-01 -1.48279631e+00 -5.54290056e-01 -1.09726560e+00
8.39948475e-01 5.07903039e-01 1.09036124e+00 1.49622127e-01
-2.51806647e-01 5.27493954e-01 -6.17720187e-01 -1.44089982e-01
-5.26263714e-01 -3.09858769e-01 -4.94558334e-01 -1.24711156e-01
2.90597409e-01 -2.10609600e-01 -4.07065719e-01 5.16185574e-02
-8.43510389e-01 6.20704174e-01 6.89573539e-03 5.11352360e-01
5.01332581e-01 -3.96627098e-01 7.56538212e-02 -9.72109497e-01
4.21351165e-01 -3.50108653e-01 -2.87092924e-01 4.84537870e-01
5.27766883e-01 -1.45121083e-01 2.62717038e-01 -4.81794983e-01
-9.96637523e-01 4.16291595e-01 5.23798645e-01 -6.67674124e-01
-2.17231989e-01 2.88475126e-01 -8.95380750e-02 3.49808842e-01
3.84194374e-01 -7.88434669e-02 -2.18526199e-01 -1.52094081e-01
7.74185181e-01 5.38875163e-01 7.49171913e-01 -6.69922650e-01
3.38540077e-01 7.28495300e-01 2.82639284e-02 -1.32888949e+00
-1.00037634e+00 -4.15558666e-01 -8.55059683e-01 -8.24097455e-01
1.52986991e+00 -9.18871582e-01 -9.01046097e-01 -1.80983603e-01
-1.60881126e+00 -3.81658942e-01 7.60865724e-03 2.55433708e-01
-8.46992493e-01 5.06846488e-01 -4.85254616e-01 -7.86662102e-01
1.75723121e-01 -1.15485430e+00 1.11799157e+00 2.14875594e-01
-6.34661973e-01 -8.23420942e-01 -2.25443169e-01 8.03357303e-01
-2.21122861e-01 9.13640678e-01 8.95163774e-01 -4.44164157e-01
-6.48776710e-01 3.68780375e-01 -3.64288956e-01 -1.92751989e-01
1.98365033e-01 2.92377144e-01 -9.84798133e-01 -2.53313873e-02
1.25101164e-01 -4.44025129e-01 5.34327805e-01 5.59852183e-01
9.42973912e-01 -3.17421108e-01 -3.17619205e-01 3.25411946e-01
9.30041373e-01 6.66201591e-01 2.82114297e-01 3.79640102e-01
1.04929793e+00 1.31911814e+00 8.88869107e-01 1.87027648e-01
4.90146160e-01 7.11139500e-01 2.91125417e-01 -2.04594687e-01
-2.73182660e-01 -2.87384242e-01 4.00970966e-01 6.18182838e-01
-8.74731839e-02 -3.49459410e-01 -1.02528346e+00 5.35222948e-01
-2.08122993e+00 -1.28251219e+00 -2.21484616e-01 1.41101360e+00
7.08841622e-01 -1.26595795e-01 2.74078131e-01 4.40636091e-02
8.70099127e-01 4.26101059e-01 -3.10327798e-01 -2.64335811e-01
-1.72952905e-01 -4.71140474e-01 -7.89491460e-02 5.71819782e-01
-1.32588279e+00 9.34175432e-01 6.33943033e+00 3.08843583e-01
-1.10248685e+00 -3.71404253e-02 8.72242391e-01 -2.67813057e-01
-4.13320996e-02 5.97520173e-02 -4.00679797e-01 3.09050173e-01
7.07833529e-01 7.10759014e-02 1.64533451e-01 7.56022692e-01
4.15623337e-01 -6.27013385e-01 -1.45520937e+00 1.01048899e+00
4.33769941e-01 -1.61565745e+00 -4.48546000e-02 -4.22158867e-01
6.00108802e-01 -6.01079941e-01 -2.80369371e-01 -1.76146865e-01
4.41450179e-02 -1.16861796e+00 8.30559909e-01 3.43566865e-01
6.16686642e-01 -3.82142067e-01 2.07022667e-01 1.61741138e-01
-1.27548373e+00 1.84644580e-01 8.24301690e-02 -3.56940031e-01
5.28022468e-01 -7.55086308e-03 -9.98965442e-01 1.28417745e-01
6.88667297e-01 1.02349830e+00 -5.99317372e-01 5.45569181e-01
-3.51015836e-01 4.55894955e-02 -5.31489141e-02 -2.38716602e-01
1.03959419e-01 -3.34929116e-02 6.23363495e-01 1.27990866e+00
1.03662647e-01 3.35696906e-01 3.83390814e-01 7.15692520e-01
1.16061874e-01 -9.98514295e-02 -7.56054580e-01 -3.03389519e-01
4.27603930e-01 1.03416073e+00 -1.22860885e+00 -4.88866299e-01
-5.25938928e-01 8.25369298e-01 2.35164449e-01 3.81747961e-01
-7.57869065e-01 -9.88820195e-02 4.23315525e-01 1.92524552e-01
3.20551872e-01 -3.27297181e-01 5.42007126e-02 -1.16794002e+00
2.54332945e-02 -9.53514874e-01 4.45058107e-01 -1.26391995e+00
-7.31065273e-01 4.74617720e-01 4.13314193e-01 -1.31138361e+00
-5.92142522e-01 -5.41395605e-01 -4.62781161e-01 2.45692089e-01
-1.19758427e+00 -8.63205910e-01 -5.87291002e-01 5.70125282e-01
1.01342630e+00 2.38143161e-01 5.24564385e-01 7.50628933e-02
-4.22895193e-01 -1.42667413e-01 -2.60883689e-01 4.17157322e-01
6.75711095e-01 -1.06329429e+00 -1.85055569e-01 8.39348316e-01
4.64699477e-01 3.84464622e-01 1.18318248e+00 -4.78584260e-01
-1.27091074e+00 -5.28327286e-01 7.27832377e-01 -6.47345781e-01
6.56071723e-01 -4.41194683e-01 -7.56667912e-01 8.21942806e-01
7.30092227e-01 -4.57529664e-01 6.15365088e-01 -1.77707270e-01
-1.48470864e-01 2.96309590e-01 -5.30445755e-01 6.96537077e-01
9.18197453e-01 -7.97538459e-01 -1.05313885e+00 4.60498422e-01
6.72765493e-01 -8.08696926e-01 -5.62429190e-01 7.30439052e-02
3.73770386e-01 -8.59310925e-01 1.18892372e+00 -6.68682277e-01
1.15378809e+00 -4.57119703e-01 -2.88490862e-01 -7.76602566e-01
3.35887522e-01 -5.06178319e-01 -3.73852961e-02 1.50759792e+00
6.74780756e-02 3.49726677e-01 7.61568129e-01 6.00592077e-01
1.34342328e-01 -2.96131879e-01 -7.44832814e-01 -2.11450294e-01
-2.41300866e-01 -2.56841272e-01 5.41556589e-02 1.13442588e+00
5.28261960e-01 6.25580966e-01 -4.71293718e-01 -1.66971013e-01
2.76338309e-01 4.55155998e-01 1.02779675e+00 -9.19093668e-01
-6.00857139e-02 -2.49134302e-01 -2.49150321e-01 -1.07534599e+00
2.54880518e-01 -5.73220074e-01 2.20432594e-01 -1.59982681e+00
3.93937379e-01 9.04159173e-02 3.74266297e-01 2.23861977e-01
-1.73669592e-01 2.03441247e-01 6.16208076e-01 3.80531162e-01
-1.02387500e+00 2.04034552e-01 1.46403706e+00 -2.63812810e-01
-2.28129700e-01 -6.00562513e-01 -5.34602225e-01 9.89230871e-01
4.86634582e-01 -3.75022501e-01 -5.80379784e-01 -5.01957536e-01
4.52859364e-02 6.11916482e-01 4.14914995e-01 -4.94458199e-01
2.68472344e-01 -4.63061959e-01 4.37451035e-01 -9.24656153e-01
7.72776425e-01 -6.95312798e-01 1.61499605e-02 2.30250522e-01
-5.80531776e-01 2.08511293e-01 1.10538803e-01 4.49913889e-01
-4.90558863e-01 -1.80764630e-01 3.99551243e-01 -3.42358738e-01
-1.12563050e+00 -4.81294960e-01 -6.43023074e-01 2.11820364e-01
1.54123878e+00 -2.01981008e-01 -8.34146023e-01 -5.83741486e-01
-9.01761591e-01 4.01629686e-01 5.14032423e-01 5.59130847e-01
6.78549349e-01 -1.13124871e+00 -4.93083864e-01 -2.56545812e-01
8.87062773e-02 1.32308856e-01 3.26146334e-01 7.17365742e-01
-6.82807267e-01 3.80509466e-01 -4.27907854e-01 -9.12466943e-01
-1.84424448e+00 5.90411365e-01 -1.56979978e-01 1.10642791e-01
-8.22574914e-01 7.40192950e-01 5.69656312e-01 4.09298748e-01
4.04604793e-01 -3.23874533e-01 -6.76715195e-01 5.94655097e-01
8.03763866e-01 1.74126342e-01 -6.55814648e-01 -1.07700384e+00
-4.22293901e-01 6.39051259e-01 1.74338564e-01 -9.76609811e-02
1.24254549e+00 -5.25907755e-01 -1.91033483e-01 8.39375317e-01
1.31628811e+00 -8.77709612e-02 -1.52878010e+00 -4.50611115e-02
3.18161517e-01 -5.71981907e-01 -2.87760705e-01 -2.43232056e-01
-7.82034695e-01 9.97305632e-01 1.14529364e-01 4.74934667e-01
1.08907366e+00 4.39958066e-01 4.04386193e-01 2.11676404e-01
4.62080278e-02 -1.04310632e+00 5.56434512e-01 3.64946812e-01
9.37324822e-01 -1.43452644e+00 2.17320517e-01 -6.32070065e-01
-9.57174301e-01 1.43538702e+00 5.61896145e-01 -1.64488573e-02
9.33986306e-02 1.38713464e-01 2.47587085e-01 -4.87495512e-01
-7.26492584e-01 -3.38960052e-01 3.53162855e-01 6.68512821e-01
5.77892721e-01 -7.65148774e-02 -1.03847481e-01 1.47625178e-01
-1.74305543e-01 -2.37971380e-01 7.18327701e-01 9.53283310e-01
-5.65680921e-01 -8.98821175e-01 -5.67769229e-01 2.85895169e-02
-4.72896069e-01 2.37584189e-01 -8.18539798e-01 9.12302494e-01
-1.47536144e-01 1.15935636e+00 3.71967643e-01 -1.19760267e-01
6.29776195e-02 5.18824570e-02 5.51799357e-01 -5.95852315e-01
-3.10195923e-01 2.01783970e-01 5.22749722e-01 -6.40792668e-01
-1.40219867e+00 -8.10706139e-01 -1.34517467e+00 -2.02165857e-01
2.87003815e-01 -1.42187610e-01 4.29532886e-01 1.09116173e+00
-3.51533964e-02 7.01665580e-01 1.93903685e-01 -1.06429863e+00
5.89588583e-01 -7.19386339e-01 -4.97635864e-02 7.51284778e-01
5.58737218e-01 -8.12260330e-01 -1.51169315e-01 7.68111765e-01] | [10.642853736877441, 0.8940608501434326] |
5845c567-22ae-4e27-acbc-184e07680371 | a-contextual-combinatorial-semi-bandit | 2206.08144 | null | https://arxiv.org/abs/2206.08144v2 | https://arxiv.org/pdf/2206.08144v2.pdf | A Contextual Combinatorial Semi-Bandit Approach to Network Bottleneck Identification | Bottleneck identification is a challenging task in network analysis, especially when the network is not fully specified. To address this task, we develop a unified online learning framework based on combinatorial semi-bandits that performs bottleneck identification in parallel with learning the specifications of the underlying network. Within this framework, we adapt and study various combinatorial semi-bandit methods such as epsilon-greedy, LinUCB, BayesUCB, NeuralUCB, and Thompson Sampling. In addition, our framework is capable of using contextual information in the form of contextual bandits. Finally, we evaluate our framework on the real-world application of road networks and demonstrate its effectiveness in different settings. | ['Morteza Haghir Chehreghani', 'Niklas Åkerblom', 'Fazeleh Hoseini'] | 2022-06-16 | null | null | null | null | ['thompson-sampling'] | ['methodology'] | [ 6.50998801e-02 -1.31820679e-01 -9.92837548e-01 -6.63958266e-02
-6.54045343e-01 -5.79831481e-01 2.93815643e-01 -1.87209055e-01
-3.02831829e-01 1.31472266e+00 1.81171000e-01 -1.02107418e+00
-8.23972285e-01 -6.95113659e-01 -9.88248050e-01 -3.55806828e-01
1.43392580e-02 6.81816101e-01 4.13795561e-01 1.65822744e-01
2.69340962e-01 4.23240095e-01 -8.20662677e-01 -8.09640437e-02
9.45811331e-01 1.01888216e+00 5.32670319e-02 7.58866370e-01
-1.53918073e-01 9.28223193e-01 -1.92210957e-01 -5.06485939e-01
2.27102220e-01 -2.83357441e-01 -9.87656176e-01 -2.74959635e-02
9.28441808e-02 -6.27594173e-01 -5.25072455e-01 8.25609982e-01
1.46268174e-01 2.50594378e-01 7.73695707e-01 -1.15544999e+00
1.16736650e-01 1.24558580e+00 -5.91954291e-01 5.79143643e-01
-5.23148850e-02 -3.69208716e-02 1.35378957e+00 -3.37535292e-01
-4.69535068e-02 1.50448179e+00 6.73710346e-01 1.15017094e-01
-1.80265737e+00 -7.23553061e-01 6.52078390e-01 3.23306948e-01
-1.06890035e+00 -6.38676703e-01 8.02908063e-01 -3.38097692e-01
4.67189550e-01 3.95054400e-01 6.93268359e-01 1.20346045e+00
-5.18364429e-01 1.21517086e+00 9.67040360e-01 -5.27830780e-01
4.64947790e-01 -1.75317258e-01 6.03300631e-01 4.35616940e-01
4.49505538e-01 2.26827458e-01 -5.77190280e-01 -5.08864880e-01
9.20134246e-01 1.53709371e-02 6.91917837e-02 -3.88325810e-01
-9.60455954e-01 9.65322614e-01 7.71991387e-02 -4.84975249e-01
-4.01101023e-01 6.80680752e-01 4.86532778e-01 2.30974361e-01
9.09375668e-01 -9.03554037e-02 -4.44087386e-01 -2.37051398e-01
-1.12737310e+00 -1.46880075e-02 1.10280681e+00 1.20176005e+00
9.67404068e-01 1.45197399e-02 -4.60352868e-01 9.02238488e-01
3.11570257e-01 4.01879787e-01 -2.21376598e-01 -1.16868818e+00
6.63731873e-01 8.63539204e-02 3.76967162e-01 -5.35928071e-01
-3.08779389e-01 -5.90900540e-01 -8.49588037e-01 -4.59250212e-01
5.47906280e-01 -4.72330332e-01 -5.75182140e-01 1.64594340e+00
3.53598297e-01 6.61531031e-01 -3.69362831e-01 4.80986983e-01
1.79580245e-02 5.80785275e-01 -2.25324288e-01 -3.51374358e-01
7.14866400e-01 -1.35602570e+00 -5.64193428e-01 -2.33327553e-01
4.33242917e-01 -2.38693357e-01 7.51558661e-01 4.20067757e-01
-8.85497153e-01 -1.57723337e-01 -6.42678201e-01 5.16315639e-01
-1.61239162e-01 4.11168812e-03 8.89688373e-01 1.16578388e+00
-9.96873975e-01 4.12199885e-01 -7.36481905e-01 -2.60318995e-01
4.14964497e-01 2.07043678e-01 3.24720562e-01 -2.16022164e-01
-1.08045924e+00 4.83310699e-01 6.43992841e-01 4.52240080e-01
-1.12553883e+00 -8.80448639e-01 -5.88414490e-01 3.24876785e-01
1.19138265e+00 -6.35246277e-01 1.36472201e+00 -9.32893991e-01
-1.86457431e+00 -1.63885932e-02 -2.04948470e-01 -8.76878262e-01
8.06915760e-01 -3.08844954e-01 4.33567464e-02 -2.81334966e-02
-3.67051139e-02 -1.73056368e-02 8.20100307e-01 -7.60099769e-01
-6.22750461e-01 -1.77359469e-02 4.45004731e-01 -7.83117712e-02
-2.54376143e-01 6.20445348e-02 -4.25554782e-01 -5.30032635e-01
-2.92450666e-01 -8.35798562e-01 -4.31984812e-01 -4.62616146e-01
-9.60038126e-01 -1.73670948e-01 5.11079907e-01 -4.60079163e-01
1.47883534e+00 -1.96054459e+00 -1.45804957e-01 8.06801856e-01
-4.12611328e-02 -3.40259932e-02 2.57209055e-02 6.58514678e-01
9.58167911e-02 4.05617326e-01 1.14026211e-01 -2.02156469e-01
1.60097390e-01 2.02086017e-01 -3.34493876e-01 5.12593985e-01
-1.91409618e-01 6.84138358e-01 -8.82446706e-01 -5.28078914e-01
3.09782416e-01 -2.73992538e-01 -7.91013241e-01 -1.09603420e-01
-2.80483723e-01 1.69123635e-01 -5.87242246e-01 7.09195554e-01
4.02664006e-01 -4.14963573e-01 3.75661284e-01 1.59388930e-01
-1.08161569e-01 2.16262266e-01 -1.31081486e+00 1.11774313e+00
-5.87912560e-01 6.03021562e-01 1.82487905e-01 -1.47452676e+00
3.74190629e-01 1.80879205e-01 3.57908189e-01 -1.91551179e-01
-1.94749415e-01 1.10659897e-01 -2.06425831e-01 -1.64460748e-01
2.52060313e-02 2.56341100e-01 1.52543038e-01 7.76472569e-01
-9.88413990e-02 3.49003464e-01 5.56234360e-01 3.16100389e-01
1.17235911e+00 -2.55668074e-01 4.31676686e-01 -4.60601330e-01
4.18745279e-01 -2.41874188e-01 4.43734676e-01 1.56991470e+00
-2.53618985e-01 -5.42400628e-02 1.16205931e+00 -3.22548091e-01
-9.78327811e-01 -1.12178898e+00 5.87882660e-03 1.47219241e+00
-1.34722844e-01 -1.64751559e-01 -5.85941613e-01 -8.25555563e-01
3.87363255e-01 5.71795404e-01 -5.73489487e-01 1.01836108e-01
-1.47580639e-01 -6.35712802e-01 3.72430146e-01 2.84408510e-01
3.57204556e-01 -5.32128811e-01 -9.96840149e-02 3.91857505e-01
-1.96970657e-01 -1.09297180e+00 -6.86525047e-01 1.25229239e-01
-9.15901542e-01 -1.05695438e+00 -5.13599634e-01 -3.09267133e-01
4.31097925e-01 4.20410186e-01 1.00143194e+00 -3.83006185e-01
-1.17538041e-02 4.38119471e-01 -2.20650300e-01 -2.67745525e-01
-4.06648479e-02 6.11614764e-01 -1.24756418e-01 3.22410434e-01
-1.05401583e-01 -5.89652538e-01 -6.87203765e-01 5.10355115e-01
-4.62776959e-01 4.99391891e-02 4.81486589e-01 1.00577009e+00
1.77411243e-01 1.96517035e-01 7.51264274e-01 -1.22769642e+00
7.30181575e-01 -6.00082636e-01 -1.08040392e+00 4.43617433e-01
-4.55019206e-01 5.49550727e-02 3.51932764e-01 -6.54366553e-01
-9.59843457e-01 -3.32033858e-02 4.17753220e-01 -2.47740746e-01
2.27844074e-01 9.77264643e-01 1.19490564e-01 -1.70564532e-01
2.52653748e-01 -2.31952537e-02 -1.15474910e-01 -5.14707804e-01
3.36122245e-01 7.26613879e-01 7.57521912e-02 -8.78197849e-01
4.54094261e-01 4.44896817e-01 -6.97466433e-02 -9.96598065e-01
-1.02363002e+00 -5.56757927e-01 -3.75325084e-01 -5.61157346e-01
4.66789715e-02 -7.66776025e-01 -1.10869551e+00 3.20543051e-01
-9.46158230e-01 -8.48516345e-01 6.13478646e-02 6.18793249e-01
-8.57795596e-01 2.82344937e-01 -6.62539542e-01 -1.53420687e+00
-7.78359473e-02 -9.84983087e-01 4.72041845e-01 -3.66511643e-02
-7.70074874e-02 -1.18932211e+00 5.12310676e-02 3.34889591e-01
3.59670192e-01 -1.45823315e-01 9.49649155e-01 -6.16824865e-01
-8.42831850e-01 1.82685312e-02 -5.68220615e-01 3.02919209e-01
-4.16069850e-03 3.13304394e-01 -5.68128645e-01 -3.05737048e-01
-6.36711359e-01 -3.26336324e-01 1.03823996e+00 1.04682922e+00
1.22161305e+00 -5.24766445e-01 -2.95110583e-01 6.22523010e-01
1.44012809e+00 -4.15238552e-02 2.08456799e-01 2.65665531e-01
3.70094568e-01 5.80570996e-01 3.84577572e-01 7.41946697e-01
2.21220136e-01 3.93546164e-01 5.46427727e-01 1.20022915e-01
3.87735993e-01 -5.19533396e-01 1.93758681e-01 4.04250383e-01
2.39842072e-01 -1.26460433e-01 -1.03984737e+00 6.18909717e-01
-2.30323553e+00 -8.61644268e-01 8.77915248e-02 2.12095857e+00
7.53288865e-01 5.01418173e-01 6.73410594e-01 3.00257057e-02
1.02907956e+00 1.23205818e-01 -7.76711524e-01 -2.73969382e-01
1.75104678e-01 -1.34429395e-01 1.14490640e+00 6.50234997e-01
-1.12144244e+00 1.07420838e+00 7.52597570e+00 1.06710327e+00
-6.00757599e-01 1.03022262e-01 9.30443823e-01 -2.79756784e-01
1.45026565e-01 2.29557335e-01 -7.87495136e-01 6.13277614e-01
1.19468403e+00 -5.66037893e-02 1.10260224e+00 8.23790789e-01
6.10281408e-01 -3.66023511e-01 -1.12005734e+00 7.75428534e-01
-6.17530704e-01 -1.50811720e+00 -2.05344945e-01 1.92288488e-01
7.50472903e-01 1.62380189e-01 -6.47455379e-02 2.93472350e-01
1.06634974e+00 -6.72942281e-01 7.30332971e-01 5.28128088e-01
6.30164444e-01 -1.01260173e+00 3.98958087e-01 5.00368893e-01
-8.59967768e-01 -4.83682305e-01 -3.23640347e-01 -3.21609974e-01
1.32657871e-01 8.80907297e-01 -6.64720416e-01 4.17147070e-01
5.40231407e-01 7.36361027e-01 -1.32227391e-01 1.46286035e+00
-1.85037315e-01 1.16029918e+00 -5.91258049e-01 -2.92361498e-01
4.34723884e-01 -3.80975604e-01 3.34677428e-01 1.11955738e+00
4.54419143e-02 -3.77300292e-01 3.72140735e-01 7.02062666e-01
3.67320217e-02 7.68166557e-02 -4.25873429e-01 8.83191898e-02
8.14015865e-01 8.07453811e-01 -6.47504330e-01 -3.22497100e-01
-3.67669135e-01 2.95237601e-01 5.15556514e-01 9.54667151e-01
-9.00178671e-01 -2.29600057e-01 4.49475259e-01 -2.69667506e-01
5.06393909e-01 -7.10317791e-02 -2.18995631e-01 -9.90001559e-01
-2.33250007e-01 -5.56481242e-01 4.33431000e-01 -4.33358312e-01
-1.29303873e+00 -2.36454129e-01 3.78411114e-01 -3.99483919e-01
-1.15710668e-01 -5.48250735e-01 -7.23358989e-01 6.73909009e-01
-1.58894098e+00 -8.96108508e-01 2.46781796e-01 6.99684620e-01
5.01812875e-01 3.78040075e-02 2.20223635e-01 2.50131011e-01
-1.31273592e+00 3.39704752e-01 7.78054297e-01 1.77232102e-01
4.48298693e-01 -1.21173728e+00 1.65537164e-01 8.44915152e-01
-1.77111268e-01 4.59434777e-01 6.72527790e-01 -5.72418749e-01
-1.11112678e+00 -1.14663446e+00 1.89576507e-01 8.80235806e-02
1.21301270e+00 -6.56426191e-01 -1.79345891e-01 8.97476196e-01
-3.61657441e-02 2.21663974e-02 5.50586581e-01 8.08740258e-01
-1.91759124e-01 -6.08029842e-01 -7.05447435e-01 6.43931568e-01
1.19029665e+00 -5.74935675e-01 4.57101762e-02 5.38616955e-01
8.58536005e-01 -8.35931450e-02 -6.15197122e-01 1.03250153e-01
6.75732017e-01 -8.74859631e-01 9.46722031e-01 -9.23665702e-01
-8.39051753e-02 2.80858725e-01 -6.56832336e-03 -1.19071770e+00
-2.94573724e-01 -1.23162258e+00 -4.83972490e-01 1.21666169e+00
6.06832504e-01 -9.63621676e-01 1.14276683e+00 2.95557737e-01
2.04517081e-01 -4.84362602e-01 -1.32705688e+00 -9.26883340e-01
-2.27134839e-01 -6.35875344e-01 5.87899029e-01 3.75968665e-01
1.33766845e-01 2.28965059e-01 -8.86193693e-01 -6.83700346e-05
9.59900141e-01 1.57242104e-01 8.76331091e-01 -1.21857917e+00
-3.76074463e-01 -6.19786978e-01 2.01652274e-01 -1.41545570e+00
6.49102032e-01 -4.03526992e-01 -1.27493948e-01 -1.06802583e+00
4.81086224e-01 -4.76437300e-01 -5.24853349e-01 1.49953246e-01
1.47713274e-01 -2.15114117e-01 -1.91988498e-01 -8.52785725e-03
-9.77406025e-01 4.61091727e-01 7.77965784e-01 -2.89490372e-01
-2.06391662e-01 7.32662976e-01 -6.99124157e-01 5.03072262e-01
8.54274213e-01 -4.39411581e-01 -6.16914749e-01 -2.28638440e-01
4.99137819e-01 4.68483508e-01 3.30471039e-01 -4.81452823e-01
4.87069756e-01 -4.31090295e-01 -1.29415870e-01 -9.47492719e-01
1.00383900e-01 -8.47388327e-01 -7.36747086e-02 5.72793424e-01
-5.59389889e-01 -1.78398430e-01 -3.40926722e-02 1.18008709e+00
1.10608689e-01 -2.41709799e-01 3.74949008e-01 -1.55934191e-03
-2.51815081e-01 4.67884153e-01 -5.45028150e-01 2.42924601e-01
7.73074448e-01 -1.15869366e-01 -4.33326960e-01 -6.52694345e-01
-6.57200634e-01 7.44756460e-01 -1.44795224e-01 -3.42411429e-01
3.71538877e-01 -1.12931645e+00 -4.39919978e-01 -5.52142784e-02
-1.33381799e-01 -5.46973012e-02 2.40005046e-01 1.16887414e+00
-3.90193045e-01 9.36508358e-01 2.55622298e-01 -5.12925148e-01
-7.26902783e-01 6.67005002e-01 2.63747841e-01 -4.82547194e-01
-1.19497240e-01 4.03614759e-01 2.39011385e-02 -4.33230639e-01
6.77473605e-01 -4.66033697e-01 1.61316767e-01 1.19316481e-01
3.67864639e-01 9.98342156e-01 -3.91686022e-01 1.59388825e-01
-1.62353650e-01 -5.28145134e-02 4.02747430e-02 -2.59560049e-01
1.20887661e+00 -5.65156162e-01 -3.60667735e-01 5.84036887e-01
7.84050524e-01 -2.52326906e-01 -1.48739290e+00 -6.33816123e-01
4.58861589e-01 -2.86049128e-01 2.89291620e-01 -4.98987079e-01
-8.48660409e-01 4.28218395e-01 -2.10283473e-01 6.82946265e-01
7.73205996e-01 -2.28410408e-01 1.40963107e-01 6.27084851e-01
3.73275101e-01 -1.32884502e+00 -1.73199311e-01 7.04797149e-01
2.75297791e-01 -1.13479328e+00 -1.63639970e-02 -3.86223555e-01
-2.96650380e-02 1.07013988e+00 2.55593747e-01 -1.43117040e-01
1.04749274e+00 -1.13558047e-01 -6.61230743e-01 3.13831449e-01
-1.05578315e+00 -3.76332641e-01 -1.02502838e-01 3.24924350e-01
-1.67091280e-01 1.91585779e-01 -7.19232112e-02 3.55893373e-01
2.09815651e-01 -2.34079361e-02 5.94071865e-01 6.53162241e-01
-3.55541885e-01 -7.66688704e-01 -3.97121996e-01 7.98209667e-01
-4.98038471e-01 -3.17837656e-01 -1.94967553e-01 7.28771687e-01
-8.58318865e-01 1.10696244e+00 -4.36455011e-02 -6.06016219e-02
1.03109755e-01 -1.31024703e-01 2.03713223e-01 -4.42935377e-01
-3.67798209e-02 4.93542612e-01 4.72478598e-01 -5.15151262e-01
-4.81804579e-01 -7.57370889e-01 -5.02091926e-03 -7.01219082e-01
-6.83957100e-01 4.31553394e-01 5.01532674e-01 1.05227220e+00
-2.86141187e-02 5.28526962e-01 8.94819021e-01 -7.44169831e-01
-8.86379242e-01 -9.96877193e-01 -9.10886049e-01 -3.71820122e-01
7.17693210e-01 -7.39144802e-01 -4.50819463e-01 -5.71649194e-01] | [4.537153244018555, 3.268362283706665] |
8bd56555-a12a-4142-8907-e72f49202e15 | learning-compositional-shape-priors-for-few | 2106.06440 | null | https://arxiv.org/abs/2106.06440v2 | https://arxiv.org/pdf/2106.06440v2.pdf | Learning Compositional Shape Priors for Few-Shot 3D Reconstruction | The impressive performance of deep convolutional neural networks in single-view 3D reconstruction suggests that these models perform non-trivial reasoning about the 3D structure of the output space. Recent work has challenged this belief, showing that, on standard benchmarks, complex encoder-decoder architectures perform similarly to nearest-neighbor baselines or simple linear decoder models that exploit large amounts of per-category data. However, building large collections of 3D shapes for supervised training is a laborious process; a more realistic and less constraining task is inferring 3D shapes for categories with few available training examples, calling for a model that can successfully generalize to novel object classes. In this work we experimentally demonstrate that naive baselines fail in this few-shot learning setting, in which the network must learn informative shape priors for inference of new categories. We propose three ways to learn a class-specific global shape prior, directly from data. Using these techniques, we are able to capture multi-scale information about the 3D shape, and account for intra-class variability by virtue of an implicit compositional structure. Experiments on the popular ShapeNet dataset show that our method outperforms a zero-shot baseline by over 40%, and the current state-of-the-art by over 10%, in terms of relative performance, in the few-shot setting. | ['Eugene Belilovsky', 'Anders Eriksson', 'Mahsa Baktashmotlagh', 'Sarah Parisot', 'Stavros Tsogkas', 'Mateusz Michalkiewicz'] | 2021-06-11 | null | null | null | null | ['single-view-3d-reconstruction'] | ['computer-vision'] | [ 2.86665678e-01 3.78088415e-01 -1.40847206e-01 -5.97297490e-01
-8.27363372e-01 -7.93176651e-01 1.03126609e+00 1.76824120e-04
-1.83444545e-01 2.26230606e-01 5.22799969e-01 -7.76112750e-02
1.20843731e-01 -8.87368023e-01 -1.10796452e+00 -5.41875124e-01
1.50950059e-01 9.25265074e-01 3.87820154e-01 -9.85168070e-02
2.04168335e-01 6.01684213e-01 -1.74661601e+00 4.50434506e-01
-3.38902324e-03 9.79408205e-01 9.33673680e-02 5.73195755e-01
-2.08445191e-01 2.93304831e-01 -9.10148695e-02 -5.28230965e-01
3.57727677e-01 -1.15403779e-01 -6.38976157e-01 3.35303217e-01
9.24555838e-01 -6.42740190e-01 -2.98776448e-01 9.55160499e-01
5.39094627e-01 1.83099329e-01 1.01468313e+00 -7.83655643e-01
-9.87918913e-01 2.36980841e-01 -1.45023033e-01 -9.11299884e-02
1.38709068e-01 3.82443070e-01 1.26603734e+00 -1.19909191e+00
9.76212800e-01 1.41834617e+00 8.32578957e-01 7.44429827e-01
-1.68359423e+00 -2.30709389e-01 2.50654757e-01 -3.16177011e-02
-1.23457491e+00 -6.09970212e-01 7.36017346e-01 -5.59184313e-01
1.38318515e+00 3.33265774e-02 7.30310678e-01 1.28030360e+00
-7.00774640e-02 7.75485396e-01 7.64918208e-01 -1.75536767e-01
3.95618200e-01 -1.55434459e-01 -5.77050122e-03 7.08802640e-01
2.37284407e-01 1.54281303e-01 -3.21866781e-01 -3.00431456e-02
8.42220664e-01 -1.78991761e-02 -3.12493625e-03 -1.01710618e+00
-1.15971863e+00 7.53057837e-01 5.53179324e-01 1.77150652e-01
-3.71088013e-02 2.84422517e-01 3.41661841e-01 1.58535466e-01
6.78241014e-01 3.53587776e-01 -6.10085070e-01 -5.27408570e-02
-7.78852880e-01 3.91462684e-01 9.01003063e-01 1.13747299e+00
6.55299127e-01 1.22202896e-01 -1.17668174e-01 7.25061655e-01
2.97051460e-01 4.77590650e-01 2.18305007e-01 -1.05252242e+00
2.23910913e-01 3.62489581e-01 -1.45065159e-01 -6.22729897e-01
-2.67640144e-01 -3.33645701e-01 -6.94989026e-01 3.81266385e-01
4.71125871e-01 3.15724224e-01 -1.34907377e+00 1.83418858e+00
2.54107773e-01 1.94425166e-01 -3.93728353e-02 8.34130406e-01
8.96312833e-01 4.21487749e-01 3.42073180e-02 2.79667556e-01
1.21517301e+00 -5.07668972e-01 -1.27733974e-02 -4.19076353e-01
2.06848949e-01 -5.82649887e-01 1.20510173e+00 1.69106081e-01
-1.22335720e+00 -5.99496841e-01 -1.02969980e+00 -3.69621277e-01
-4.16584790e-01 -2.90229321e-01 5.86519301e-01 5.22009611e-01
-1.00763118e+00 9.24350083e-01 -8.97353768e-01 -5.74451506e-01
8.45102668e-01 2.18621567e-01 -4.41865623e-01 -3.71379465e-01
-5.37065744e-01 8.85675550e-01 1.77245840e-01 -3.70719433e-01
-1.19000506e+00 -1.04283381e+00 -1.08982956e+00 1.62489206e-01
3.57884765e-01 -1.08491445e+00 1.36898685e+00 -5.63050210e-01
-1.34049022e+00 1.31070232e+00 -1.03599750e-01 -2.75934249e-01
3.05230975e-01 5.44738807e-02 3.04254562e-01 -2.24403590e-02
-5.83349615e-02 1.03890657e+00 8.80387843e-01 -1.58475804e+00
-1.33937418e-01 -5.69643855e-01 1.43690631e-01 1.51089415e-01
1.64079055e-01 -4.91131306e-01 -4.02867734e-01 -5.33025980e-01
3.64049703e-01 -9.30289686e-01 -9.75351259e-02 4.49312240e-01
-2.08836496e-01 -3.56786162e-01 5.05077481e-01 -3.86903733e-01
2.92710721e-01 -2.10329747e+00 2.33029798e-01 -3.48533913e-02
1.78750858e-01 7.50925615e-02 -2.53483266e-01 1.96075156e-01
8.25398639e-02 1.65867805e-01 -5.23584902e-01 -6.75078750e-01
4.29510355e-01 5.32444417e-01 -5.04448354e-01 4.14980292e-01
4.27626133e-01 1.22878206e+00 -8.69765878e-01 -5.77985542e-03
2.72603989e-01 6.17777765e-01 -8.49693179e-01 2.17707559e-01
-4.82289016e-01 2.74735928e-01 6.36234432e-02 6.28010690e-01
5.36163509e-01 -4.98531878e-01 1.46687804e-02 -2.58617014e-01
2.45073020e-01 3.48672867e-01 -9.22646344e-01 2.28307271e+00
-3.94787848e-01 4.73511159e-01 -1.61036074e-01 -1.18199027e+00
9.51347888e-01 1.71600029e-01 2.79388070e-01 -3.60177517e-01
1.16706111e-01 1.71307012e-01 -1.15076341e-01 -3.51544827e-01
1.65925711e-01 -5.83057702e-01 -5.28396107e-02 5.13261497e-01
5.52780032e-01 -6.21647894e-01 -1.34101242e-01 3.41381654e-02
1.08158112e+00 5.14278471e-01 1.91822425e-01 -1.62640199e-01
-4.25204560e-02 -1.78682134e-01 3.76913309e-01 7.50987053e-01
-2.99697649e-02 1.00648701e+00 2.88063675e-01 -7.56953776e-01
-1.61432731e+00 -1.53311265e+00 -2.21971795e-01 8.91203880e-01
1.39392270e-02 -5.53028062e-02 -3.65579516e-01 -6.20442808e-01
2.55189836e-01 9.81538594e-01 -6.90973282e-01 -2.78689057e-01
-3.94091576e-01 -5.43177247e-01 2.35765636e-01 5.78856289e-01
1.01569273e-01 -8.32984746e-01 -6.12278759e-01 2.30458051e-01
1.71997219e-01 -1.32188451e+00 -2.06138238e-01 1.60822928e-01
-1.10045958e+00 -1.00201666e+00 -8.07972372e-01 -8.17626715e-01
7.16633856e-01 1.57736734e-01 1.38191688e+00 -1.28923208e-01
-5.13049722e-01 6.09725893e-01 -8.06244612e-02 -2.59226590e-01
-3.82216334e-01 -3.54675315e-02 -3.01315248e-01 -1.35552824e-01
6.46804392e-01 -1.11222589e+00 -4.88559693e-01 -6.02324344e-02
-7.49950349e-01 1.87461063e-01 7.84342706e-01 8.40277433e-01
6.38571620e-01 -5.84740102e-01 4.90201443e-01 -8.63937080e-01
1.63135588e-01 -4.59587783e-01 -4.08036113e-01 1.17498942e-01
-3.64920735e-01 4.36732858e-01 5.21628678e-01 -3.97778392e-01
-1.12811935e+00 2.10090354e-01 -3.49740714e-01 -7.78175712e-01
-6.87879145e-01 5.76913208e-02 -2.04647154e-01 2.37428099e-01
7.59958208e-01 3.65513563e-01 1.16320677e-01 -7.99889266e-01
7.38760531e-01 2.81244338e-01 5.88599801e-01 -7.52025306e-01
7.84546793e-01 8.34880173e-01 1.99205637e-01 -7.14750051e-01
-1.11361289e+00 -3.16376239e-01 -1.04267716e+00 9.59716588e-02
1.00631046e+00 -1.08820450e+00 -3.62742603e-01 1.91897660e-01
-1.36625803e+00 -4.29230899e-01 -6.00295484e-01 1.56746164e-01
-1.05043983e+00 2.68506765e-01 -5.04206061e-01 -4.21144336e-01
-1.62675381e-01 -1.02043629e+00 1.48142636e+00 -2.95539409e-01
-1.05065092e-01 -9.81472373e-01 -3.77897322e-02 9.75437611e-02
3.92166048e-01 3.08257431e-01 1.38924015e+00 -6.52173936e-01
-8.11417341e-01 -9.48917642e-02 -3.88913721e-01 2.59661347e-01
-1.60086587e-01 -4.33221281e-01 -1.21378982e+00 -1.86960235e-01
-4.21648286e-03 -4.97072607e-01 1.04795086e+00 2.67954767e-01
1.33351517e+00 -2.15932369e-01 -1.55794784e-01 7.47164369e-01
1.55138206e+00 -3.19733739e-01 4.71128047e-01 -2.61357337e-01
7.12245882e-01 4.44986075e-01 -8.34507793e-02 3.39033693e-01
5.62978029e-01 6.91125751e-01 6.10206485e-01 3.99305195e-01
-6.72454059e-01 -4.98287469e-01 1.75010134e-02 7.69846320e-01
-1.69275314e-01 5.12182973e-02 -8.13019931e-01 8.16806853e-01
-1.73138797e+00 -1.02927041e+00 3.28286588e-01 2.22702837e+00
9.07657683e-01 2.89984107e-01 -7.82399550e-02 -2.79986858e-01
3.62510920e-01 1.61600411e-01 -8.38958323e-01 -2.84187883e-01
3.06608118e-02 3.49827200e-01 3.94449621e-01 4.70617265e-01
-1.03978896e+00 8.90925169e-01 6.56365013e+00 5.88402212e-01
-7.90537834e-01 1.13075800e-01 5.41124523e-01 -1.71528295e-01
-6.81295395e-01 6.28201365e-02 -8.16621244e-01 5.67742214e-02
6.88090682e-01 6.85172006e-02 5.80238104e-01 9.24034536e-01
-3.09264034e-01 2.27703065e-01 -1.67989457e+00 1.08267760e+00
4.90439266e-01 -1.53900671e+00 1.81721300e-01 1.42277256e-01
8.91510427e-01 4.16323572e-01 4.63848673e-02 3.77550423e-01
5.18261373e-01 -9.72489893e-01 7.82042146e-01 7.36482620e-01
9.43617284e-01 -3.97330999e-01 2.33712107e-01 5.26624203e-01
-9.76257443e-01 1.30398899e-01 -7.58776784e-01 -1.40761407e-02
2.74763316e-01 5.05893886e-01 -8.01360190e-01 1.35060355e-01
4.75479633e-01 9.81962621e-01 -4.59962666e-01 9.83947039e-01
-3.44925195e-01 3.25640053e-01 -3.89331967e-01 1.85722038e-01
2.30056539e-01 1.80111244e-01 7.14900970e-01 9.99278963e-01
4.23827380e-01 2.16237828e-01 1.16510458e-01 1.25266349e+00
-3.45615655e-01 -3.49792749e-01 -1.01786256e+00 2.32623592e-01
2.82081932e-01 8.95681858e-01 -6.95065379e-01 -4.68474895e-01
-5.00260413e-01 8.15123916e-01 5.69788396e-01 2.23745972e-01
-4.72712934e-01 -5.75149432e-02 6.51580691e-01 1.73017994e-01
8.78937304e-01 -4.56028074e-01 -6.12440109e-01 -1.28404486e+00
-6.73792288e-02 -6.16979837e-01 1.29196256e-01 -8.97723973e-01
-1.61664963e+00 2.24081799e-01 1.17211767e-01 -1.08883274e+00
-3.66594672e-01 -8.52836788e-01 -3.35387170e-01 5.35349786e-01
-1.39931738e+00 -1.38405752e+00 -1.17272370e-01 3.73973221e-01
8.09982002e-01 -1.90430254e-01 1.02468169e+00 7.19844401e-02
2.41537943e-01 3.32505167e-01 -1.71395335e-02 5.18664606e-02
4.17519957e-01 -1.27170670e+00 8.16663444e-01 6.03069901e-01
6.89284801e-01 3.99951339e-01 6.06936932e-01 -6.12538636e-01
-1.67513609e+00 -9.51708615e-01 8.64332139e-01 -8.79092216e-01
4.58315134e-01 -6.88393414e-01 -1.05212235e+00 8.51556063e-01
-1.79419238e-02 4.31697607e-01 6.11887515e-01 2.27731943e-01
-9.86259997e-01 2.02726439e-01 -1.08502007e+00 4.60324168e-01
1.69232869e+00 -6.90519631e-01 -1.04954922e+00 2.82167375e-01
8.33673298e-01 -3.08161288e-01 -9.27471757e-01 3.99994373e-01
7.19160855e-01 -9.38685954e-01 1.32393932e+00 -8.34889710e-01
6.20441437e-01 -9.16813221e-03 -6.69138730e-01 -1.32310700e+00
-3.41490060e-01 -2.01026365e-01 -3.51042837e-01 7.58616984e-01
1.84502497e-01 -1.14187263e-01 8.17946017e-01 7.72955060e-01
-4.52446252e-01 -7.73152769e-01 -9.56970811e-01 -8.23525429e-01
3.90638411e-01 -6.44049644e-01 6.42047465e-01 5.90782642e-01
-4.20918614e-01 6.60645604e-01 -2.37561971e-01 -8.82997662e-02
8.75236869e-01 3.47618103e-01 8.48347127e-01 -1.44661832e+00
-4.50575531e-01 -6.99827135e-01 -7.01656044e-01 -1.33127272e+00
3.90079677e-01 -1.21083546e+00 1.42604828e-01 -1.73479950e+00
2.94816673e-01 -3.13852191e-01 1.16555043e-01 6.03394628e-01
2.31968477e-01 5.13489902e-01 3.95244986e-01 1.33801356e-01
-5.15447259e-01 6.55561149e-01 1.29960644e+00 -4.20128047e-01
1.59372479e-01 -6.76005334e-02 -5.24806440e-01 7.97669232e-01
3.67641658e-01 -3.32025111e-01 -3.71755511e-01 -8.28134656e-01
2.17649758e-01 -3.12811702e-01 6.89423859e-01 -7.51121581e-01
1.86775327e-01 5.06739244e-02 5.67677140e-01 -7.62589216e-01
7.65297890e-01 -8.12259316e-01 -4.85407338e-02 2.44136661e-01
-3.86389077e-01 -4.74866539e-01 2.06030369e-01 7.57815480e-01
1.47004411e-01 -2.05206960e-01 7.29867697e-01 -4.90709394e-01
-1.06990600e+00 6.11165822e-01 1.24046497e-01 2.80918241e-01
7.61299908e-01 -4.16043967e-01 -1.37837514e-01 -3.08598161e-01
-1.18746698e+00 -3.76169652e-01 5.92499137e-01 6.19562268e-01
6.99990213e-01 -1.52758443e+00 -7.42534876e-01 4.13727999e-01
3.92166436e-01 1.92192376e-01 1.48649916e-01 3.98809284e-01
-5.87949678e-02 2.37897396e-01 -3.05954784e-01 -9.89568949e-01
-7.65980959e-01 6.27416193e-01 3.35764736e-01 1.61747724e-01
-1.01555848e+00 9.01268601e-01 4.17822778e-01 -7.59696543e-01
2.10386008e-01 -3.47096980e-01 2.14854196e-01 -5.09704016e-02
3.04204911e-01 -8.42320845e-02 1.79762751e-01 -6.12839878e-01
-1.91636056e-01 9.02044594e-01 1.98631704e-01 -4.47356552e-02
1.73063016e+00 -7.59477541e-02 1.05528578e-01 7.68940926e-01
1.49954677e+00 -5.37657499e-01 -1.76678002e+00 -5.04255712e-01
-7.86596164e-02 -6.57611251e-01 -3.04331668e-02 -7.24340856e-01
-7.25340068e-01 1.29567540e+00 3.61902565e-01 -6.96611479e-02
6.40750885e-01 5.00559986e-01 5.79149604e-01 5.98734140e-01
4.21096534e-01 -7.43872821e-01 1.62705094e-01 8.34215879e-01
9.20356333e-01 -1.44393015e+00 -5.45787588e-02 -1.24640398e-01
-4.65980411e-01 1.16974771e+00 3.91565502e-01 -4.42214549e-01
9.43687141e-01 -5.28054759e-02 -3.78969640e-01 -4.04249489e-01
-9.43152070e-01 -3.87278318e-01 6.95047319e-01 8.11238825e-01
2.01441795e-01 -7.14969821e-04 3.98985505e-01 5.65852463e-01
-2.63151824e-01 -2.82501280e-01 3.52636933e-01 5.61926246e-01
-6.25123620e-01 -8.74881744e-01 1.97493769e-02 4.80246633e-01
-3.92388627e-02 6.82771951e-02 -2.31616691e-01 7.59908140e-01
1.49004772e-01 3.22411090e-01 3.81675661e-01 -7.99123347e-02
4.14857745e-01 5.10947883e-01 9.15771782e-01 -1.04568136e+00
8.55142251e-03 4.20267358e-02 -3.39571689e-03 -5.86007714e-01
-3.85169357e-01 -8.32435846e-01 -1.01802993e+00 -2.40161836e-01
1.33035854e-01 -6.09088302e-01 7.02959538e-01 1.11530411e+00
4.46911007e-01 3.21350098e-01 2.28107110e-01 -1.32665992e+00
-9.18246984e-01 -7.30216205e-01 -2.73538172e-01 7.64930010e-01
3.04260880e-01 -8.69888067e-01 -3.49077493e-01 1.38695642e-01] | [8.496231079101562, -3.142235517501831] |
d796108a-a3e6-430d-9e02-5a8ab0249a54 | digital-and-physical-world-attacks-on-remote | 2110.11525 | null | https://arxiv.org/abs/2110.11525v1 | https://arxiv.org/pdf/2110.11525v1.pdf | Digital and Physical-World Attacks on Remote Pulse Detection | Remote photoplethysmography (rPPG) is a technique for estimating blood volume changes from reflected light without the need for a contact sensor. We present the first examples of presentation attacks in the digital and physical domains on rPPG from face video. Digital attacks are easily performed by adding imperceptible periodic noise to the input videos. Physical attacks are performed with illumination from visible spectrum LEDs placed in close proximity to the face, while still being difficult to perceive with the human eye. We also show that our attacks extend beyond medical applications, since the method can effectively generate a strong periodic pulse on 3D-printed face masks, which presents difficulties for pulse-based face presentation attack detection (PAD). The paper concludes with ideas for using this work to improve robustness of rPPG methods and pulse-based face PAD. | ['Adam Czajka', 'Kevin W. Bowyer', 'Patrick Flynn', 'Nathan Vance', 'Jeremy Speth'] | 2021-10-21 | null | null | null | null | ['face-presentation-attack-detection'] | ['computer-vision'] | [ 6.02204680e-01 3.47495556e-01 6.21118784e-01 -7.61028752e-02
-3.68643522e-01 -6.54263675e-01 3.15191299e-01 -5.24384737e-01
-1.60773963e-01 7.33210266e-01 1.70897901e-01 -8.60281140e-02
2.91572273e-01 -4.35981989e-01 -4.59895372e-01 -8.46383393e-01
-1.45528480e-01 -3.39442581e-01 1.49133489e-01 1.21683329e-01
2.14790389e-01 8.38309348e-01 -1.42312145e+00 5.20942286e-02
2.86523879e-01 1.07841563e+00 -5.58418691e-01 9.56400394e-01
1.99781150e-01 5.74229062e-01 -1.22089386e+00 -4.93619710e-01
4.05676812e-01 -9.18310523e-01 4.94676121e-02 9.36190505e-03
7.71469414e-01 -7.98651874e-01 -5.54673135e-01 7.64182925e-01
1.02213538e+00 -3.60420436e-01 4.81022954e-01 -1.24448931e+00
-3.95598173e-01 -5.56613617e-02 -8.09482455e-01 1.08631015e-01
8.96519303e-01 3.99745584e-01 -3.16823006e-01 -6.54491186e-01
5.60976505e-01 1.17131209e+00 7.23471045e-01 1.02241135e+00
-1.07594824e+00 -8.24915648e-01 -7.97724962e-01 1.36435166e-01
-1.31738150e+00 -9.10559058e-01 1.06624889e+00 -3.49760195e-03
5.87889194e-01 8.36703122e-01 4.30933952e-01 9.71465886e-01
4.09231484e-01 -1.88552365e-01 1.60154867e+00 -4.93578911e-01
-5.77001981e-02 4.89362627e-01 -3.67152125e-01 6.83880627e-01
2.17045635e-01 2.44811594e-01 -6.09691799e-01 -6.29179955e-01
9.20354664e-01 -5.90976298e-01 -9.66487706e-01 1.22643255e-01
-2.97701091e-01 2.33201802e-01 -3.41780365e-01 3.09956372e-01
-1.54330775e-01 2.05518275e-01 1.38373688e-01 2.97738284e-01
2.75323093e-01 2.77560204e-01 2.44209260e-01 -3.04086685e-01
-6.95734203e-01 -5.38718343e-01 1.24687505e+00 5.41855156e-01
1.29133053e-02 2.29774445e-01 -2.17466831e-01 4.45822418e-01
3.17626178e-01 1.08647883e+00 -8.28136578e-02 -8.93496454e-01
7.25085810e-02 -2.39048257e-01 2.62568474e-01 -1.11288309e+00
-9.66232046e-02 5.29278755e-01 -2.53783554e-01 7.77758539e-01
5.43904245e-01 -6.40647173e-01 -7.93655872e-01 1.27545738e+00
3.69643390e-01 7.20875621e-01 1.55485803e-02 9.05519247e-01
1.23658609e+00 4.37607169e-01 -2.49960974e-01 -9.85040367e-01
1.47230315e+00 -5.77595457e-03 -1.30445433e+00 3.02303791e-01
-6.13542832e-02 -1.14821625e+00 6.92336380e-01 6.47714615e-01
-1.21419287e+00 -3.96276861e-01 -1.17682815e+00 2.09528610e-01
1.40053138e-01 -1.05353512e-01 -1.39657274e-01 1.80615675e+00
-1.20567465e+00 4.70189691e-01 -4.31239784e-01 6.94304705e-02
2.83329487e-01 3.87087405e-01 -3.99318337e-01 5.16242683e-02
-9.73124385e-01 1.00210452e+00 -7.15195239e-01 2.87340373e-01
-1.27770007e-01 -1.01519883e+00 -7.66845465e-01 -2.79618323e-01
7.58825019e-02 -4.16113660e-02 7.63399243e-01 -5.06066084e-01
-2.34901953e+00 1.00821495e+00 -2.82683253e-01 -3.57564807e-01
6.42252445e-01 -1.13916479e-01 -9.77077425e-01 1.03985357e+00
-8.36390078e-01 -4.43805940e-02 1.48522341e+00 -1.01008248e+00
3.61927569e-01 -2.14845419e-01 -5.35475373e-01 -8.43845606e-02
-1.31038457e-01 5.94598472e-01 -5.78472354e-02 -6.71125725e-02
-2.49948770e-01 -5.20764828e-01 3.84838104e-01 4.23235893e-01
-1.39547423e-01 1.92475006e-01 1.54492068e+00 -8.23491871e-01
9.68918025e-01 -2.09340239e+00 -8.34727347e-01 3.10578227e-01
6.41792789e-02 8.48642826e-01 -9.20087397e-02 5.17021477e-01
-7.47264549e-02 5.68882897e-02 -5.26268110e-02 -7.20046908e-02
-1.40278354e-01 4.71321750e-04 -6.24374270e-01 1.05519676e+00
-1.04506798e-01 7.42799580e-01 -3.66352439e-01 -3.27271074e-01
4.05748218e-01 1.13046491e+00 2.30189696e-01 2.18042269e-01
5.12188375e-01 5.45221567e-01 1.16658971e-01 5.41372240e-01
1.27621853e+00 5.75735271e-01 2.55701505e-02 -3.49488050e-01
2.96066608e-02 -3.27594914e-02 -9.67576802e-01 7.81901062e-01
-2.39706278e-01 8.17699373e-01 2.88763195e-01 -2.29179144e-01
1.32685256e+00 8.25717747e-01 2.61018932e-01 -7.60831475e-01
2.72699803e-01 9.06704888e-02 -1.21923201e-01 -9.41129386e-01
-1.78259984e-01 -7.80277967e-01 4.80268806e-01 6.86388254e-01
-3.49734515e-01 -2.48168185e-01 -3.06365639e-01 -6.98374212e-02
8.97073627e-01 -9.95678082e-02 -3.19359861e-02 -1.77692756e-01
7.44643927e-01 -8.46297860e-01 1.18927926e-01 4.84421015e-01
-3.56992304e-01 7.83535838e-01 6.96361542e-01 -2.33607292e-01
-5.56811154e-01 -8.43821406e-01 -4.59637582e-01 -2.52828132e-02
2.91144848e-01 -2.26673558e-01 -7.69517720e-01 -5.67650735e-01
1.45620823e-01 3.73102814e-01 -4.63499546e-01 1.03342272e-01
-6.03533745e-01 -3.52216423e-01 9.38529074e-01 2.30005205e-01
2.69009352e-01 -1.00387573e+00 -9.14290369e-01 4.93583865e-02
-7.71556795e-02 -1.27232742e+00 -4.54170793e-01 -6.85027063e-01
-4.38898623e-01 -1.33227396e+00 -1.05239534e+00 -3.69047672e-01
8.53710115e-01 -1.49548113e-01 7.63825178e-01 5.30137718e-02
-1.05086422e+00 1.20681989e+00 -4.30149212e-02 -5.24204135e-01
-6.89222753e-01 -1.38061786e+00 2.99306929e-01 4.73246485e-01
5.86626112e-01 -7.61782587e-01 -8.77193391e-01 5.12837648e-01
-4.56406206e-01 -7.00633287e-01 -3.50615531e-02 1.26294613e-01
5.08815004e-03 -1.13372408e-01 3.22948158e-01 -6.81298792e-01
7.83634067e-01 3.32445681e-01 -5.94250977e-01 3.01063471e-02
-1.25016747e-02 -5.33137858e-01 4.80325222e-01 -6.75946295e-01
-1.29127729e+00 -2.63424724e-01 -4.56460901e-02 -6.33542538e-01
-3.73511910e-01 -3.10912222e-01 1.39283352e-02 -1.02584076e+00
1.09668136e+00 2.52859920e-01 5.81339121e-01 -2.63856322e-01
-8.23716000e-02 7.53524899e-01 7.89414763e-01 -8.91631562e-03
9.64931488e-01 6.96593702e-01 4.90237057e-01 -1.50744021e+00
2.91071415e-01 -5.33972420e-02 -2.53780782e-01 -6.53988123e-01
3.76197398e-01 -3.35905224e-01 -1.37348151e+00 7.09615350e-01
-9.45434928e-01 -2.11986840e-01 -9.70034301e-02 5.25346041e-01
-3.84407699e-01 9.87061679e-01 -8.03193152e-01 -1.28544414e+00
-3.53321791e-01 -5.19777775e-01 6.23256087e-01 4.90096569e-01
-8.33512768e-02 -1.01169705e+00 8.83284584e-02 2.00776100e-01
7.46806800e-01 7.79406190e-01 4.54145491e-01 2.20072612e-01
-1.87631220e-01 -5.47452748e-01 1.29120171e-01 4.47633773e-01
5.28163731e-01 2.14261368e-01 -1.64900088e+00 -2.47386053e-01
8.46095622e-01 -2.90751368e-01 3.64195853e-01 5.42071581e-01
9.19353902e-01 -4.92744386e-01 -2.40925297e-01 4.88043487e-01
1.32316506e+00 4.34002995e-01 1.40766847e+00 -7.16949761e-01
1.13889992e-01 7.71462202e-01 1.22408755e-01 3.43611240e-01
-6.89829707e-01 3.99666578e-01 9.96763352e-03 -4.55126494e-01
-5.31050801e-01 5.04689068e-02 4.44191247e-01 -8.66116285e-02
-3.08113217e-01 -2.24382505e-01 -2.70256907e-01 -1.65977240e-01
-6.48977399e-01 -9.07390594e-01 -3.55512887e-01 2.51159716e+00
7.81721473e-01 -4.05279398e-01 5.78517467e-02 2.84415901e-01
9.53284800e-01 -1.03042997e-01 -1.92640260e-01 -6.18997037e-01
5.90016842e-02 8.21326911e-01 5.52905083e-01 6.60501063e-01
-7.52167106e-01 2.53084600e-01 7.29470062e+00 2.13073358e-01
-1.54558885e+00 -2.00268421e-02 4.73219723e-01 -2.52913117e-01
-9.88993347e-02 -3.70740801e-01 -6.74410045e-01 6.61115229e-01
8.83934140e-01 -2.52535403e-01 7.40941167e-02 6.71035871e-02
2.74328262e-01 -5.38161397e-01 -8.92835557e-01 1.48793542e+00
6.05009139e-01 -9.12122309e-01 -4.32026774e-01 1.72284737e-01
1.92978512e-02 -8.57214093e-01 3.33948493e-01 -5.12869179e-01
-6.63349211e-01 -1.11325085e+00 -2.82450855e-01 4.58085507e-01
1.46946454e+00 -6.98787868e-01 1.84139550e-01 -1.57259062e-01
-7.49512792e-01 3.10848296e-01 -5.20414710e-01 3.27399112e-02
3.72699797e-01 4.21723425e-01 -7.96563327e-01 2.90292390e-02
2.15319321e-01 -7.34160915e-02 -6.14145100e-02 1.21120703e+00
-4.00875896e-01 6.76089585e-01 -7.62322545e-01 9.80731025e-02
-4.45364207e-01 -3.35496515e-02 9.20595586e-01 1.01492906e+00
4.69039142e-01 7.19479084e-01 -7.07421541e-01 8.11850965e-01
1.58872426e-01 -1.31330386e-01 -8.35663557e-01 1.60954982e-01
1.23445340e-01 1.38999748e+00 -6.42797649e-01 7.99401775e-02
-4.23531651e-01 1.02848792e+00 -1.00355887e+00 5.06506085e-01
-6.62019968e-01 -9.41249251e-01 4.71357584e-01 5.21998525e-01
1.40983433e-01 1.16459817e-01 8.49527046e-02 -6.79097414e-01
2.36666083e-01 -6.01500273e-01 3.52166176e-01 -9.31363404e-01
-1.04259694e+00 5.24167418e-01 -1.92681462e-01 -1.17542183e+00
-1.21774502e-01 -6.73405111e-01 -9.16788995e-01 1.29477215e+00
-1.47899783e+00 -4.62912410e-01 -3.89703363e-01 9.17923152e-01
-2.70891041e-01 3.88006955e-01 9.58892345e-01 -5.69837680e-03
1.19431585e-01 8.78903925e-01 -4.77194965e-01 4.52277102e-02
1.02869117e+00 -7.26814270e-01 1.87942311e-01 9.00763571e-01
-2.33765915e-01 2.58762568e-01 7.81480074e-01 -5.52854717e-01
-1.74573553e+00 -2.59552330e-01 7.47451961e-01 -3.80503148e-01
7.30425268e-02 -4.89704669e-01 -8.82442594e-01 -6.80638254e-02
5.44400513e-01 1.35521650e-01 8.09574187e-01 -7.50666499e-01
-3.54464382e-01 -2.31173918e-01 -1.93697774e+00 3.43037635e-01
5.38794994e-01 -6.39884651e-01 -3.79191726e-01 3.59441400e-01
-3.82198170e-02 -5.56311846e-01 -8.31652105e-01 2.39606082e-01
7.05162406e-01 -9.26819801e-01 7.40949750e-01 1.62241638e-01
-8.78792182e-02 -2.26813972e-01 5.92494190e-01 -9.38635468e-01
1.92430481e-01 -1.74925172e+00 2.13413071e-02 1.06835866e+00
-4.68487516e-02 -1.29569399e+00 6.86349094e-01 1.10598469e+00
4.86226112e-01 1.24003783e-01 -9.59909320e-01 -8.25282156e-01
-4.25127387e-01 4.23865914e-02 1.53931230e-01 7.33098865e-01
6.66983485e-01 -2.75279760e-01 -7.19651818e-01 4.57957923e-01
8.91479969e-01 -2.26056159e-01 6.76724494e-01 -9.30085361e-01
-3.58226508e-01 2.51033843e-01 -7.39635050e-01 -5.68266869e-01
-4.64724451e-01 -5.36666587e-02 1.80969015e-02 -8.64567995e-01
-3.82676005e-01 2.77588814e-01 8.88246149e-02 3.71495157e-01
2.38187760e-01 1.04012191e+00 1.48012087e-01 -3.09447259e-01
5.35079658e-01 8.03385824e-02 1.24283206e+00 4.02535737e-01
-3.87301594e-01 2.48863455e-02 -4.59331244e-01 4.63573009e-01
6.27974927e-01 -4.12661642e-01 -5.14643192e-01 3.81252855e-01
-1.28087148e-01 5.28548896e-01 3.62053126e-01 -1.14228702e+00
3.51778060e-01 3.14722568e-01 7.29193509e-01 -3.60490739e-01
5.84733009e-01 -7.60270357e-01 3.43755215e-01 7.35820949e-01
1.94239065e-01 -4.73233491e-01 3.79764557e-01 1.80333763e-01
1.51389450e-01 -1.50516376e-01 1.15462494e+00 -2.11148068e-01
3.01831104e-02 -6.79733828e-02 -5.95259607e-01 -2.25367412e-01
9.81556654e-01 -5.90458393e-01 -9.68119025e-01 -8.66933584e-01
-8.83211076e-01 -3.64329666e-01 3.52936566e-01 -5.02477176e-02
9.81006444e-01 -7.62439489e-01 -7.02630162e-01 9.60469663e-01
-4.93892789e-01 -9.68715489e-01 4.66986626e-01 1.20427835e+00
-8.95780981e-01 5.78668937e-02 -2.62830973e-01 -4.22405094e-01
-2.07901788e+00 5.11437237e-01 7.03722477e-01 7.75661647e-01
-1.15584016e+00 6.85914278e-01 3.99122424e-02 6.24139130e-01
2.89838433e-01 3.04582596e-01 -2.38398373e-01 -2.29788691e-01
1.35017228e+00 5.33160448e-01 7.82046244e-02 -2.55743146e-01
-2.78170884e-01 9.77035642e-01 3.60329419e-01 -2.57276922e-01
8.80049825e-01 -2.11161301e-01 -3.44665423e-02 -8.64285231e-02
1.11359525e+00 5.74321508e-01 -1.24216819e+00 1.43452540e-01
-7.09243596e-01 -1.12477660e+00 -1.81427777e-01 -9.62335408e-01
-1.15985942e+00 9.94857490e-01 8.01492929e-01 3.25894445e-01
1.44241333e+00 -2.19675511e-01 6.70121551e-01 4.13320474e-02
2.88302422e-01 -8.68698955e-01 1.29965097e-01 -5.27504444e-01
9.55276191e-01 -2.96594381e-01 3.63022573e-02 -1.09509182e+00
-3.81623775e-01 1.49701965e+00 1.99434459e-01 -2.11171415e-02
8.89752030e-01 9.35060561e-01 6.88048959e-01 -4.65953685e-02
-4.66823846e-01 1.95190057e-01 1.85798898e-01 1.00380659e+00
1.25215456e-01 -4.08262968e-01 -6.77397728e-01 -6.22351840e-03
1.00667760e-01 3.71498942e-01 1.11112738e+00 8.45955431e-01
-2.58074820e-01 -1.00304139e+00 -7.40540564e-01 1.44851372e-01
-9.97356057e-01 1.85328886e-01 -6.23508632e-01 6.87073052e-01
-3.47528547e-01 1.06233323e+00 -4.15761536e-03 8.51475224e-02
3.77215922e-01 3.68199587e-01 1.07914591e+00 5.92077076e-02
-3.13489527e-01 5.25573134e-01 3.38981971e-02 -7.59011507e-01
-5.68719566e-01 -3.38368535e-01 -7.35909224e-01 -3.59122664e-01
8.11098889e-02 -5.78241423e-02 5.83321631e-01 4.32643533e-01
2.68670410e-01 -2.88018018e-01 9.48367655e-01 -5.83584428e-01
-2.64150411e-01 -8.14188719e-01 -1.19014394e+00 2.10846797e-01
4.50370550e-01 -2.65599191e-01 -9.48540509e-01 1.03016868e-01] | [13.883305549621582, 2.7539944648742676] |
d3b82271-17bd-48b1-b6ab-8c9a13498012 | knowledge-aware-neural-collective-matrix | 2206.13255 | null | https://arxiv.org/abs/2206.13255v1 | https://arxiv.org/pdf/2206.13255v1.pdf | Knowledge-aware Neural Collective Matrix Factorization for Cross-domain Recommendation | Cross-domain recommendation (CDR) can help customers find more satisfying items in different domains. Existing CDR models mainly use common users or mapping functions as bridges between domains but have very limited exploration in fully utilizing extra knowledge across domains. In this paper, we propose to incorporate the knowledge graph (KG) for CDR, which enables items in different domains to share knowledge. To this end, we first construct a new dataset AmazonKG4CDR from the Freebase KG and a subset (two domain pairs: movies-music, movie-book) of Amazon Review Data. This new dataset facilitates linking knowledge to bridge within- and cross-domain items for CDR. Then we propose a new framework, KG-aware Neural Collective Matrix Factorization (KG-NeuCMF), leveraging KG to enrich item representations. It first learns item embeddings by graph convolutional autoencoder to capture both domain-specific and domain-general knowledge from adjacent and higher-order neighbours in the KG. Then, we maximize the mutual information between item embeddings learned from the KG and user-item matrix to establish cross-domain relationships for better CDR. Finally, we conduct extensive experiments on the newly constructed dataset and demonstrate that our model significantly outperforms the best-performing baselines. | ['Haiping Lu', 'Jianmo Ni', 'Jun Ma', 'Yan Ge', 'Li Zhang'] | 2022-06-27 | null | null | null | null | ['general-knowledge'] | ['miscellaneous'] | [-6.37496471e-01 -1.32992044e-01 -8.46330941e-01 -5.11392832e-01
-2.66124070e-01 -5.86670101e-01 8.08393061e-02 3.15784276e-01
-1.16958544e-01 6.11470222e-01 5.46475708e-01 3.65439896e-03
-5.16291142e-01 -1.08571517e+00 -8.17087471e-01 1.47658130e-02
-5.73814549e-02 4.82066005e-01 9.83398110e-02 -4.09466296e-01
-6.87183961e-02 1.14632018e-01 -9.58815932e-01 5.45013845e-01
9.68044341e-01 1.10419929e+00 2.59823769e-01 -8.26670080e-02
-4.39000189e-01 6.33992791e-01 -1.81033850e-01 -8.79524648e-01
3.44846308e-01 -1.87220215e-03 -8.45439374e-01 1.53562486e-01
5.66616096e-02 -4.46768880e-01 -5.06114304e-01 7.32986808e-01
1.81316108e-01 7.04883635e-01 6.75726235e-01 -1.25172615e+00
-2.11960053e+00 9.61072743e-01 -5.65981686e-01 7.47670755e-02
4.50862497e-01 -2.43404016e-01 1.47241139e+00 -1.01000786e+00
7.39230156e-01 9.63106930e-01 6.38294041e-01 2.70345569e-01
-1.12432563e+00 -6.37581527e-01 7.01865733e-01 3.19779605e-01
-1.37971258e+00 3.33894819e-01 7.72833288e-01 -2.16737121e-01
1.34376490e+00 -3.97073984e-01 7.99709082e-01 9.83732522e-01
-3.66632015e-01 8.41742814e-01 3.77257079e-01 -2.33532161e-01
3.03746271e-03 4.94247437e-01 4.29081619e-01 2.83046514e-01
4.83218133e-01 -2.95447946e-01 -6.79450095e-01 -1.76046073e-01
1.00997365e+00 4.90799814e-01 -2.88665682e-01 -7.19352305e-01
-9.38177884e-01 1.06432509e+00 8.92616987e-01 2.93089867e-01
-5.09129882e-01 -2.47910619e-01 1.94268391e-01 6.07523382e-01
4.78838831e-01 5.77017665e-01 -8.07293594e-01 3.35389107e-01
-1.41738370e-01 2.53249198e-01 8.46800864e-01 1.29872334e+00
1.03898978e+00 -3.68993282e-01 2.23194480e-01 1.25727630e+00
6.23127103e-01 5.83689287e-02 6.97412848e-01 -4.48170960e-01
5.46719253e-01 1.00533354e+00 6.03417493e-02 -1.30753911e+00
-2.70414501e-01 -7.03979015e-01 -4.44582194e-01 -6.72010362e-01
-9.54630822e-02 -2.03992659e-03 -3.65850836e-01 1.57474911e+00
4.52310652e-01 4.55734074e-01 1.94931492e-01 9.81598735e-01
1.17205226e+00 5.05245745e-01 4.06326093e-02 1.47364020e-01
1.10533869e+00 -1.32407761e+00 -6.32210612e-01 -1.21408857e-01
8.97826791e-01 -3.81312072e-01 9.91445899e-01 3.82209510e-01
-7.29502201e-01 -8.20304215e-01 -1.00603127e+00 -3.08022082e-01
-8.95090938e-01 -5.22500486e-04 9.14596736e-01 1.69874176e-01
-6.25798702e-01 6.28234804e-01 -1.42335698e-01 -3.54548275e-01
5.43049812e-01 4.99199301e-01 -6.74180865e-01 -6.06407583e-01
-1.87079394e+00 6.49405897e-01 4.55850184e-01 -4.41803217e-01
-3.17421645e-01 -1.09641135e+00 -1.06137991e+00 1.58136740e-01
2.77396351e-01 -9.35852408e-01 8.94978821e-01 -1.08628333e+00
-1.22426081e+00 5.91068745e-01 2.03336969e-01 -3.34302813e-01
-3.49672288e-01 -5.03859520e-01 -1.17088521e+00 -2.66676277e-01
1.77325219e-01 5.54699361e-01 6.25384271e-01 -1.11204958e+00
-6.41435981e-01 -3.98594707e-01 5.93514204e-01 3.79696727e-01
-7.62268186e-01 -3.89809251e-01 -8.75307858e-01 -8.65677416e-01
-3.01734179e-01 -7.45558977e-01 -9.29369554e-02 -3.05560857e-01
-4.13336530e-02 -5.24312675e-01 3.18980724e-01 -5.69616914e-01
1.60276866e+00 -2.14312100e+00 1.31006211e-01 3.74788344e-01
3.02635670e-01 1.95229098e-01 -5.51584601e-01 7.28648663e-01
-1.12195298e-01 -1.38778597e-01 4.84489381e-01 -2.65346080e-01
2.15870500e-01 5.36354482e-01 -1.30070850e-01 -1.25296123e-04
2.15912417e-01 1.27302217e+00 -1.04845119e+00 2.51559075e-02
-4.54501361e-02 4.45048958e-01 -9.13808584e-01 1.22194551e-01
-2.52520144e-01 5.65537661e-02 -5.16282856e-01 6.52699292e-01
9.71237659e-01 -8.01108241e-01 5.14937937e-01 -5.04146218e-01
5.37901282e-01 3.57416064e-01 -1.20504963e+00 2.09209633e+00
-4.63774502e-01 -5.23587167e-02 -3.32970589e-01 -9.47340906e-01
1.01413095e+00 -2.09601715e-01 5.48517048e-01 -9.57412124e-01
-9.28617939e-02 1.03959337e-01 -2.41773501e-01 -3.64685804e-01
8.07705045e-01 3.25138308e-02 -1.36875272e-01 5.72980821e-01
5.54730177e-01 5.60845435e-01 -6.58722827e-03 6.59679472e-01
9.46676672e-01 -1.15825161e-01 4.40126032e-01 2.87288390e-02
3.07526201e-01 -2.27194801e-01 3.56394649e-01 2.96286464e-01
1.06891366e-02 2.33665958e-01 2.47606948e-01 -3.82381588e-01
-6.33865952e-01 -1.20409274e+00 5.31286895e-02 1.40082932e+00
5.16138971e-01 -6.15583301e-01 -6.57105073e-02 -1.15862334e+00
8.97979259e-01 5.81716418e-01 -8.08380485e-01 -5.14004529e-01
-2.20952407e-01 -3.37207109e-01 -5.50341345e-02 9.72891748e-01
1.36794597e-01 -9.25249577e-01 8.76178443e-01 3.67383838e-01
2.11913008e-02 -1.09399271e+00 -8.02087188e-01 1.21806897e-01
-6.29924893e-01 -1.10543358e+00 -6.46496832e-01 -1.09621167e+00
4.70696986e-01 6.85694575e-01 1.53223920e+00 -1.60473794e-01
1.47245571e-01 6.26744926e-01 -8.91714513e-01 -1.55513291e-03
4.31812555e-01 1.01159208e-01 4.84237969e-01 9.56499949e-02
1.17062056e+00 -7.50580430e-01 -7.33118832e-01 5.85907102e-01
-7.95146644e-01 -5.34742117e-01 5.15651524e-01 7.31536925e-01
6.86070621e-01 1.19201154e-01 1.24593627e+00 -1.14182734e+00
1.09814155e+00 -1.37692463e+00 -4.10207734e-02 2.46072829e-01
-6.89393044e-01 -1.16374850e-01 6.83309793e-01 -5.35755098e-01
-8.96730959e-01 -3.51461440e-01 -2.34696898e-03 -6.85047805e-01
4.82080653e-02 9.12288547e-01 -3.85679275e-01 1.07527219e-01
5.70938408e-01 -1.27740055e-02 -3.62306654e-01 -8.59808326e-01
9.69750166e-01 7.08536148e-01 3.43822986e-01 -5.85771978e-01
4.41139579e-01 1.89011022e-01 -7.36690700e-01 -1.78183928e-01
-1.10072267e+00 -1.00432730e+00 -6.94626093e-01 2.28423208e-01
5.61120629e-01 -1.31226766e+00 -4.04777110e-01 -6.44470528e-02
-7.93068826e-01 -1.55063514e-02 -4.61663723e-01 6.01459861e-01
-1.85278296e-01 4.17761892e-01 -5.53280890e-01 -1.09941684e-01
-1.39716342e-01 -3.75034481e-01 7.34130442e-01 3.75747412e-01
-2.18260922e-02 -1.45104551e+00 3.88404965e-01 4.51537967e-01
2.79167324e-01 -3.58768135e-01 9.39645886e-01 -1.10541117e+00
-3.42629403e-01 -2.02167511e-01 -6.05065346e-01 6.26641393e-01
5.31391382e-01 -6.98457420e-01 -3.46874118e-01 -2.71945119e-01
-6.87860727e-01 -3.75218511e-01 9.11036253e-01 1.35747403e-01
1.06811738e+00 -1.88763544e-01 -5.53636372e-01 3.81116569e-01
1.42994690e+00 -7.43148699e-02 3.17721933e-01 2.88990974e-01
8.84856820e-01 4.81163919e-01 7.73734808e-01 6.95229471e-01
9.79645908e-01 7.83725619e-01 2.66675532e-01 1.46644533e-01
-7.75875002e-02 -6.88106418e-01 4.29464191e-01 1.16839147e+00
2.57508844e-01 -3.77105027e-01 -3.25519294e-01 8.91251683e-01
-2.09610367e+00 -6.18214905e-01 5.43262511e-02 1.99522579e+00
7.34328389e-01 1.11126155e-02 4.07400489e-01 -4.18843120e-01
6.59752429e-01 -1.84170872e-01 -8.10331762e-01 -3.87332022e-01
-9.85818580e-02 2.28937507e-01 2.62311816e-01 2.71708071e-01
-1.04507816e+00 9.27755117e-01 5.22164345e+00 8.84847522e-01
-5.09820342e-01 1.35702476e-01 -2.66555301e-03 -2.65835315e-01
-6.64924741e-01 -3.29713315e-01 -9.41402972e-01 5.64706802e-01
6.70813382e-01 -2.17425138e-01 4.74242628e-01 1.06389475e+00
-3.94942433e-01 5.36695540e-01 -9.82354283e-01 9.42505538e-01
2.17480883e-01 -1.35316944e+00 2.69554794e-01 9.48789194e-02
1.06646824e+00 8.28689709e-02 1.63634986e-01 9.80120957e-01
7.39828885e-01 -7.25684404e-01 -1.11731023e-01 5.40779233e-01
5.31820059e-01 -9.73905206e-01 6.77502334e-01 1.90322533e-01
-1.44803405e+00 -1.62645370e-01 -9.10168409e-01 1.18642012e-02
2.57043242e-01 7.52948105e-01 -6.73120797e-01 9.61521864e-01
7.64916897e-01 1.60113502e+00 -3.92196804e-01 7.50432789e-01
-1.41469866e-01 2.36504436e-01 3.15580592e-02 1.24131285e-01
2.45842993e-01 -2.79482931e-01 2.85666576e-03 1.07726347e+00
3.15598637e-01 1.34927407e-01 3.79131109e-01 6.89950228e-01
-6.17910147e-01 4.46885943e-01 -6.52987182e-01 -3.56136233e-01
6.05788469e-01 1.34548378e+00 -4.50938344e-02 -2.43421659e-01
-1.08946383e+00 1.22509325e+00 8.91288817e-01 3.97739798e-01
-8.25673223e-01 -3.10311377e-01 1.25962794e+00 3.16716582e-01
7.75146425e-01 3.22343111e-02 2.25668848e-01 -1.46634090e+00
-9.71302539e-02 -5.83466768e-01 7.82641411e-01 -7.46006668e-01
-2.40984416e+00 2.45880157e-01 -3.80354613e-01 -1.26686192e+00
1.91382319e-01 -8.29327881e-01 -2.61121124e-01 8.52286994e-01
-1.91206717e+00 -1.47608614e+00 -4.56700400e-02 1.09205401e+00
2.49405578e-01 -3.83927196e-01 7.68669069e-01 7.59117603e-01
-4.23733145e-01 1.01154280e+00 4.42931473e-01 1.45410314e-01
1.04655397e+00 -1.18671882e+00 4.05017734e-01 6.11183792e-02
3.15214783e-01 1.17901576e+00 -1.73851356e-01 -9.41331267e-01
-1.34591699e+00 -1.42995238e+00 7.80908287e-01 -6.12954736e-01
8.73259425e-01 -1.62981302e-01 -1.18915284e+00 9.75611925e-01
2.13101774e-01 2.88721293e-01 1.54697120e+00 9.52946901e-01
-8.51237237e-01 -1.16745703e-01 -9.38680828e-01 1.96293399e-01
1.40120089e+00 -6.31114125e-01 -7.62016892e-01 3.38650078e-01
1.26629353e+00 1.75932840e-01 -1.49998415e+00 1.58374578e-01
5.49542367e-01 -5.98673940e-01 1.20132887e+00 -1.37861276e+00
4.60251659e-01 -2.49432832e-01 -4.61144924e-01 -1.76204145e+00
-1.05126071e+00 -1.06442580e-02 -7.32573390e-01 1.49704301e+00
5.31325817e-01 -5.51762342e-01 8.24725211e-01 4.87486124e-01
-2.01478839e-01 -9.28223133e-01 -3.13692391e-01 -9.58542764e-01
9.28993598e-02 -5.24052978e-01 1.00264502e+00 1.49334252e+00
5.81169188e-01 5.29665649e-01 -4.60813940e-01 2.09167749e-01
1.98556110e-01 4.71038848e-01 7.36240208e-01 -1.35527253e+00
-6.95566952e-01 -1.11331083e-01 -2.90257812e-01 -1.45015836e+00
2.36493871e-01 -1.36931419e+00 -7.46159554e-01 -1.63592744e+00
2.66893804e-01 -6.58625007e-01 -1.15397215e+00 5.01005590e-01
-3.07694525e-01 1.28888175e-01 -2.07647551e-02 7.34315515e-02
-1.27619147e+00 5.29905558e-01 1.40726078e+00 3.11776511e-02
-2.86802471e-01 -2.69154251e-01 -1.45061505e+00 3.31121564e-01
5.52985489e-01 -2.46376738e-01 -6.42151952e-01 -6.03677869e-01
8.49970818e-01 -4.10211861e-01 1.52759282e-02 -4.38002616e-01
1.81902185e-01 -5.50404051e-03 6.87151670e-01 -3.84382933e-01
2.35374242e-01 -1.06140888e+00 6.60003051e-02 -3.02386552e-01
-3.40870380e-01 -6.73620254e-02 1.20503753e-01 1.16959286e+00
-3.41978699e-01 -4.28837612e-02 2.90062010e-01 -4.28020619e-02
-1.18318486e+00 6.69837773e-01 1.87591314e-01 5.05686365e-02
9.60091293e-01 4.90820035e-03 -3.56510669e-01 -3.54033440e-01
-1.01626432e+00 5.23989081e-01 3.73469591e-01 9.89332139e-01
6.52530789e-01 -2.02840495e+00 -2.89584339e-01 3.33566844e-01
7.46364057e-01 -3.25054109e-01 7.53802478e-01 4.29773688e-01
7.14982376e-02 4.74471688e-01 -3.54518980e-01 -3.63493562e-02
-7.38285422e-01 1.16568601e+00 -1.13190874e-01 -6.76381111e-01
-2.84563750e-01 1.28521216e+00 3.65493178e-01 -9.42516744e-01
-4.40912470e-02 -1.56347483e-01 -6.28146708e-01 3.85500848e-01
4.97252434e-01 2.29400292e-01 2.34107487e-02 -4.52912033e-01
-3.39788765e-01 4.41447735e-01 -6.94834471e-01 4.75537807e-01
1.44455767e+00 -3.63843024e-01 9.96242836e-02 7.41587505e-02
1.42297125e+00 1.30819127e-01 -8.44948947e-01 -9.89851296e-01
-2.58265257e-01 -7.47892320e-01 -5.42130657e-02 -9.54619408e-01
-1.27495635e+00 5.70601583e-01 1.48874760e-01 2.31978998e-01
1.12205350e+00 2.87797034e-01 1.17224026e+00 3.00231814e-01
4.76879388e-01 -1.14857912e+00 4.29740757e-01 3.49594265e-01
8.35539699e-01 -1.24021649e+00 2.30326094e-02 -3.96196812e-01
-1.09106159e+00 8.37094665e-01 1.01322186e+00 -4.31524485e-01
1.18430698e+00 -3.20869565e-01 -2.83398896e-01 -1.74206823e-01
-9.08848524e-01 -4.42926824e-01 6.42542779e-01 9.70159411e-01
3.36181939e-01 2.35088497e-01 -1.34382054e-01 1.73134446e+00
1.40659750e-01 2.92709947e-01 5.39281406e-02 6.33278072e-01
-3.55448902e-01 -1.50054419e+00 3.55039418e-01 7.80619979e-01
-8.27720016e-02 -2.86395758e-01 -5.24981618e-01 6.40171170e-01
4.70635563e-01 8.44828129e-01 -3.61168198e-02 -7.64257431e-01
5.14626622e-01 -1.32750452e-01 3.27129066e-01 -8.98880482e-01
-5.96156597e-01 -2.03844994e-01 -8.38502403e-03 -5.67275167e-01
-2.22384676e-01 -3.93567860e-01 -1.13855398e+00 -4.28544313e-01
-5.73520482e-01 2.24378794e-01 1.67172685e-01 6.33280933e-01
9.97494876e-01 5.24455965e-01 5.27373970e-01 -3.36863220e-01
-1.39847979e-01 -8.05399239e-01 -1.35743284e+00 7.25129128e-01
-1.13950931e-01 -9.69821393e-01 1.09266639e-01 -2.32444033e-01] | [10.262395858764648, 5.636030197143555] |
caedcf28-1f5a-425f-a4f6-fedd6a192e96 | deep-diacritization-efficient-hierarchical | 2011.00538 | null | https://arxiv.org/abs/2011.00538v1 | https://arxiv.org/pdf/2011.00538v1.pdf | Deep Diacritization: Efficient Hierarchical Recurrence for Improved Arabic Diacritization | We propose a novel architecture for labelling character sequences that achieves state-of-the-art results on the Tashkeela Arabic diacritization benchmark. The core is a two-level recurrence hierarchy that operates on the word and character levels separately---enabling faster training and inference than comparable traditional models. A cross-level attention module further connects the two, and opens the door for network interpretability. The task module is a softmax classifier that enumerates valid combinations of diacritics. This architecture can be extended with a recurrent decoder that optionally accepts priors from partially diacritized text, which improves results. We employ extra tricks such as sentence dropout and majority voting to further boost the final result. Our best model achieves a WER of 5.34%, outperforming the previous state-of-the-art with a 30.56% relative error reduction. | ['Mohamed Gabr', 'Muhammad N. ElNokrashy', 'Badr AlKhamissi'] | 2020-11-01 | null | https://aclanthology.org/2020.wanlp-1.4 | https://aclanthology.org/2020.wanlp-1.4.pdf | coling-wanlp-2020-12 | ['arabic-text-diacritization'] | ['natural-language-processing'] | [ 3.91959727e-01 3.86956960e-01 -2.03902990e-01 -5.59789956e-01
-9.22580183e-01 -6.26045048e-01 4.30373728e-01 2.39524320e-01
-7.31088519e-01 6.32862628e-01 2.47308746e-01 -7.66632438e-01
5.15468657e-01 -6.88179612e-01 -6.27863169e-01 -7.63483405e-01
-5.19155376e-02 4.52505559e-01 2.91152835e-01 -5.83251595e-01
1.93482384e-01 1.66003346e-01 -1.04219508e+00 8.29226971e-01
1.04633307e+00 6.34415030e-01 7.78104441e-05 9.16210413e-01
1.08178750e-01 1.03054655e+00 -8.52890015e-01 -8.45176160e-01
-6.80623427e-02 -4.35371608e-01 -9.70997155e-01 -2.60529757e-01
3.99075449e-01 -4.36571181e-01 -4.42555904e-01 7.58606791e-01
3.60743284e-01 -6.09000698e-02 5.65727413e-01 -3.62294585e-01
-1.04965949e+00 1.17147005e+00 -5.22572756e-01 1.18037045e-01
6.24471083e-02 1.30372751e-03 1.33580399e+00 -7.30140567e-01
2.29051813e-01 1.35003376e+00 7.47422278e-01 6.16040289e-01
-1.13349462e+00 -4.25908148e-01 3.90742242e-01 2.53995985e-01
-1.08795738e+00 -3.93691152e-01 2.41726026e-01 -6.03912286e-02
1.53822601e+00 2.68874705e-01 3.78976703e-01 9.01290417e-01
2.75721580e-01 1.18075526e+00 8.12761247e-01 -6.54824317e-01
-4.09925170e-02 -2.75567144e-01 6.03193998e-01 9.08643365e-01
5.15946038e-02 -4.25868899e-01 -4.64418173e-01 2.39032805e-01
4.37936991e-01 -6.11204684e-01 -1.20810112e-02 5.22345841e-01
-8.43918741e-01 8.91694784e-01 4.48446602e-01 4.78178561e-02
-1.93965599e-01 3.21998388e-01 6.57552600e-01 1.80141881e-01
4.35598403e-01 4.59537208e-01 -5.44061542e-01 -2.81071126e-01
-9.52060699e-01 -9.21461061e-02 8.80065143e-01 8.49250317e-01
3.97402734e-01 2.68718630e-01 -3.42703134e-01 9.20774102e-01
5.03249109e-01 1.86436489e-01 3.10377747e-01 -6.20348215e-01
7.50655949e-01 4.02800173e-01 -1.36224091e-01 -2.42464632e-01
-2.76414305e-01 -4.47557211e-01 -5.82654834e-01 1.86869547e-01
6.26712799e-01 -5.42982459e-01 -1.40983546e+00 1.48277605e+00
-1.64996520e-01 -6.27033636e-02 7.98259974e-02 8.42848122e-01
5.41120708e-01 9.07287657e-01 9.00125951e-02 1.60259783e-01
1.55853975e+00 -1.41378748e+00 -7.96102583e-01 -6.20179534e-01
7.84804821e-01 -7.48777151e-01 1.16595399e+00 5.65083742e-01
-1.36238027e+00 -2.60372221e-01 -1.50668263e+00 -4.86364067e-01
-2.55661011e-01 5.49796402e-01 7.53720999e-01 9.60189164e-01
-1.04954541e+00 7.27408171e-01 -1.24808705e+00 1.47853166e-01
2.58827806e-01 5.57623863e-01 -9.79346186e-02 1.78116113e-01
-1.35639918e+00 1.34379220e+00 6.87714875e-01 7.50678062e-01
-6.89406753e-01 -4.58953440e-01 -1.08697331e+00 1.73647180e-01
3.00828484e-03 -2.48406172e-01 1.42334020e+00 -8.85349393e-01
-2.21722054e+00 8.58861864e-01 -3.17812413e-01 -7.23142505e-01
2.87151635e-01 -6.19818032e-01 -3.50113213e-01 -4.92459442e-03
-3.12966406e-01 6.42382860e-01 4.81976777e-01 -7.55837500e-01
-7.02828169e-01 1.90884452e-02 1.08436957e-01 3.03210288e-01
-1.66711882e-01 2.40201980e-01 -8.37962985e-01 -7.26687849e-01
-1.60724018e-03 -9.81876373e-01 -2.97798872e-01 -6.15146816e-01
-5.23344100e-01 -3.46399695e-01 4.98128831e-01 -1.21011364e+00
1.52262366e+00 -1.82432103e+00 2.28563353e-01 5.20050898e-02
-2.96772510e-01 6.19364977e-01 -1.66145429e-01 2.20727473e-01
-3.39345820e-02 9.65051427e-02 -5.30113995e-01 -7.00639904e-01
6.02463000e-02 2.76247799e-01 -1.46147162e-01 4.83034641e-01
1.05851412e+00 1.00632346e+00 -7.18023896e-01 -2.71623842e-02
-1.13226742e-01 4.89895672e-01 -5.13611019e-01 -2.16995344e-01
-5.16772687e-01 1.36535494e-02 1.04554079e-01 5.83331704e-01
6.26855791e-01 -1.63046077e-01 4.92163807e-01 1.77359760e-01
-1.77830197e-02 1.13660562e+00 -8.18520129e-01 1.74517334e+00
-3.99754256e-01 8.39434028e-01 -1.78120479e-01 -7.60183692e-01
8.87999177e-01 2.54968464e-01 -5.59546351e-01 -4.69180882e-01
3.13863665e-01 2.88867742e-01 4.27216589e-01 -1.55185282e-01
7.72476137e-01 1.44129962e-01 -2.38945633e-01 3.27045351e-01
8.41311663e-02 7.91833773e-02 4.07778680e-01 1.00367919e-01
1.02219677e+00 4.97732013e-01 -5.93285784e-02 -4.02070761e-01
5.69933891e-01 -1.94217935e-01 5.43359160e-01 6.42967403e-01
1.70483887e-01 6.01626098e-01 8.87254238e-01 -3.37998420e-01
-1.17566073e+00 -9.73747194e-01 -1.64151922e-01 1.37697947e+00
-1.23584352e-01 -4.40212190e-01 -1.05834866e+00 -7.60799706e-01
-2.66022891e-01 9.29712355e-01 -7.51560211e-01 -1.27484892e-02
-1.18888664e+00 -1.17640948e+00 1.17724872e+00 6.93203866e-01
4.63407218e-01 -1.14502823e+00 -1.69132873e-01 4.03977573e-01
-1.92905858e-01 -1.01704466e+00 -2.90806621e-01 7.87089765e-01
-6.39871299e-01 -4.90096390e-01 -5.60667694e-01 -1.17541683e+00
5.51225841e-01 -6.48167610e-01 1.13627994e+00 8.27333331e-02
9.12699252e-02 -8.44775617e-01 -4.20164973e-01 -2.33258739e-01
-8.54414701e-01 6.58391654e-01 -2.76818007e-01 -2.50543237e-01
4.70727652e-01 5.74183837e-02 -1.92781582e-01 -1.56805754e-01
-6.57124519e-01 3.63578647e-02 5.92797101e-01 1.06429398e+00
1.40548378e-01 -4.56107050e-01 5.77622056e-01 -1.23249555e+00
4.68274385e-01 -1.29294515e-01 -6.26855373e-01 3.96241158e-01
-3.64476353e-01 5.22303581e-01 6.79695427e-01 -2.09194958e-01
-1.24529576e+00 9.37929004e-02 -7.07446158e-01 4.65624452e-01
-9.83018279e-02 6.99971378e-01 -1.94542646e-01 3.42676103e-01
6.62831187e-01 -8.17607418e-02 -8.97320434e-02 -4.96538341e-01
4.95593846e-01 1.02532053e+00 7.68041492e-01 -3.77940625e-01
2.73723304e-01 3.95281725e-02 -3.02355438e-01 -8.50561202e-01
-1.09787798e+00 -9.13468674e-02 -7.46477067e-01 3.26638609e-01
9.70401645e-01 -1.04353070e+00 -8.34283233e-01 8.69250774e-01
-1.29957736e+00 -7.63085842e-01 1.43068239e-01 1.94704235e-01
-2.14758068e-01 5.42577446e-01 -1.60250008e+00 -6.67060971e-01
-5.18238485e-01 -1.29432893e+00 8.67400885e-01 3.25859994e-01
-2.33725086e-01 -9.40763056e-01 -2.61301816e-01 3.26411098e-01
2.39875108e-01 -1.56305060e-01 8.60537827e-01 -5.35030067e-01
-3.51225644e-01 -1.55335382e-01 -1.96036890e-01 6.47325516e-01
-1.59752771e-01 1.42792732e-01 -1.20920038e+00 -3.61737013e-01
-3.97006214e-01 -3.76883328e-01 1.17224634e+00 1.80929422e-01
8.42402637e-01 -1.75292119e-01 -2.48593651e-02 4.54361975e-01
1.02139771e+00 1.19773604e-01 1.10708153e+00 6.02533400e-01
6.01673126e-01 2.24423558e-01 4.60778236e-01 2.12904096e-01
5.03205240e-01 4.36996222e-01 4.29760784e-01 -2.42385954e-01
-2.07450449e-01 -6.16200641e-03 7.63391018e-01 9.96419966e-01
1.59153044e-01 -5.63412130e-01 -8.79799604e-01 6.47718310e-01
-1.89147937e+00 -6.48786306e-01 -3.95789295e-01 1.90567029e+00
1.37639594e+00 4.41856593e-01 -7.05405250e-02 2.03773707e-01
6.48435473e-01 9.49900299e-02 -3.37114334e-01 -1.17978978e+00
-2.83756047e-01 7.13647664e-01 7.35411108e-01 1.01922691e+00
-1.21631908e+00 1.64794636e+00 6.74855328e+00 8.85274112e-01
-1.00942743e+00 -7.27054775e-02 7.88483977e-01 1.73612341e-01
-2.78096527e-01 3.53287384e-02 -1.31087220e+00 4.08747345e-01
1.41094518e+00 5.92001975e-01 3.07319105e-01 3.79743636e-01
-2.29059253e-02 -1.50255308e-01 -9.92074609e-01 2.75504947e-01
3.46338123e-01 -1.44159198e+00 -4.16425802e-03 -2.29442790e-01
6.82098806e-01 2.78019488e-01 2.68242531e-03 2.82444209e-01
7.82222986e-01 -1.20912194e+00 8.49774361e-01 2.72156775e-01
8.09933662e-01 -1.06062436e+00 9.05084074e-01 1.14894025e-01
-8.13179910e-01 2.69003566e-02 -1.89201713e-01 -2.58325636e-01
3.01520050e-01 2.68265992e-01 -8.62350047e-01 4.71585453e-01
3.96951050e-01 6.49345338e-01 -5.93047500e-01 9.22045290e-01
-1.08908570e+00 1.16882873e+00 -4.19793725e-01 -3.50211799e-01
7.78099835e-01 -2.01575711e-01 7.96555206e-02 1.90647316e+00
-2.67211109e-01 -1.90950349e-01 -2.71692395e-01 4.39974248e-01
-1.52620509e-01 -2.80725121e-01 2.26985916e-01 1.63444415e-01
3.82547319e-01 1.04218495e+00 -6.79444790e-01 -4.29588020e-01
-2.94390827e-01 1.48697245e+00 8.59329998e-01 2.80752897e-01
-1.00145185e+00 -1.01981759e+00 5.86102664e-01 -4.73727107e-01
7.75952756e-01 -4.58135217e-01 -6.04216933e-01 -1.10517395e+00
-6.74448013e-02 -1.01313639e+00 3.12600136e-01 -3.76836061e-01
-9.61764455e-01 7.50491798e-01 -5.38674653e-01 -5.87069333e-01
-2.98518777e-01 -1.03042066e+00 -4.85767543e-01 1.23174119e+00
-1.65339267e+00 -1.33335447e+00 2.76603401e-01 -3.62880416e-02
5.41931987e-01 -3.11949532e-02 1.01551139e+00 2.87842214e-01
-8.70916069e-01 1.22720551e+00 1.74316049e-01 5.39473295e-01
6.35808051e-01 -1.38460636e+00 1.23689473e+00 1.34260130e+00
3.99540439e-02 6.74566329e-01 4.05079901e-01 -5.03041387e-01
-1.04047394e+00 -9.87404048e-01 1.39756548e+00 -5.12929797e-01
8.09216082e-01 -6.82059169e-01 -9.97071922e-01 8.78224134e-01
4.52580273e-01 -4.09080297e-01 5.98359048e-01 3.12689513e-01
-4.22596276e-01 2.77788997e-01 -5.96131086e-01 9.23012137e-01
6.13835633e-01 -5.63368559e-01 -5.57395816e-01 1.07181296e-01
8.24278474e-01 -7.31916845e-01 -7.10877895e-01 1.93540141e-01
5.19880474e-01 -6.33313477e-01 6.18016124e-01 -7.18052030e-01
6.12489164e-01 -3.27154338e-01 4.37663831e-02 -1.26971805e+00
-4.25897986e-01 -9.78328168e-01 -2.00924605e-01 9.92474675e-01
1.06075215e+00 -3.33338737e-01 8.83557737e-01 3.60833436e-01
-6.51145458e-01 -7.59850144e-01 -6.80396020e-01 -5.72574615e-01
4.60672677e-01 -3.63459826e-01 2.52241910e-01 7.68797636e-01
4.19024110e-01 8.27862322e-01 -3.51646304e-01 4.18189704e-01
1.70488819e-01 -3.52065623e-01 9.40805227e-02 -6.35523796e-01
-4.30958807e-01 -6.01261258e-01 -5.62866069e-02 -1.58869708e+00
4.49779004e-01 -1.05184996e+00 3.56914490e-01 -1.54272532e+00
-2.31907353e-01 -3.17484051e-01 5.25974520e-02 8.92566800e-01
-5.69566131e-01 5.81673265e-01 1.58652321e-01 -2.55682170e-01
-4.48392987e-01 4.55387294e-01 8.23844194e-01 -1.21434117e-02
-2.65391976e-01 -2.26627871e-01 -6.64866984e-01 6.52818620e-01
1.07225668e+00 -1.83087394e-01 1.65231183e-01 -1.17004907e+00
2.00512365e-01 -2.74667621e-01 -2.51983136e-01 -6.99019372e-01
2.25645810e-01 2.98828304e-01 2.35412762e-01 -5.68471253e-01
3.57842624e-01 -9.11694318e-02 -5.75175941e-01 5.81479371e-01
-4.71875668e-01 2.31892928e-01 5.73579133e-01 7.81526566e-02
-9.57229882e-02 -5.02998531e-01 9.42509532e-01 4.97101136e-02
-5.05690873e-01 -1.80475518e-01 -8.88803601e-01 3.31261382e-02
5.66519260e-01 -9.25356969e-02 -5.34612954e-01 -1.01474427e-01
-6.48006797e-01 3.29131931e-01 4.17433083e-02 4.22304869e-01
1.92438260e-01 -8.48890781e-01 -1.11200762e+00 3.87710661e-01
-2.18928233e-01 3.78850289e-02 -2.13336214e-01 5.13980150e-01
-1.13617039e+00 4.37511951e-01 3.69302854e-02 -2.73113966e-01
-1.10633755e+00 -1.70775548e-01 3.54930282e-01 -3.77698004e-01
-4.86594647e-01 1.23969626e+00 -4.88745838e-01 -4.65593815e-01
3.78745586e-01 -4.06263888e-01 -1.09840080e-01 -3.88727523e-02
6.02702498e-01 2.65046954e-01 4.85586286e-01 -5.01486182e-01
-3.55476409e-01 7.74362832e-02 -6.40745163e-01 -3.39838803e-01
1.26683986e+00 9.83535871e-02 -3.46662670e-01 2.85152197e-01
9.37097073e-01 -7.61641636e-02 -1.39495170e+00 -1.83217287e-01
3.87752146e-01 9.85657126e-02 -2.07483699e-03 -1.20027268e+00
-5.45665145e-01 1.00505209e+00 4.68051210e-02 -4.59251180e-02
7.58148015e-01 -3.07566881e-01 8.78160834e-01 6.01345122e-01
-3.10529619e-01 -1.17740881e+00 -1.31204367e-01 1.51712823e+00
5.53036213e-01 -1.00057888e+00 -3.00051957e-01 -3.08952779e-01
-8.57513905e-01 1.20988357e+00 5.17473578e-01 -4.08075929e-01
3.07907850e-01 8.22329521e-01 2.34025836e-01 2.54742563e-01
-8.64103377e-01 -1.53093696e-01 2.54466325e-01 2.46176630e-01
1.03597677e+00 5.24116047e-02 -3.62415016e-01 5.89336932e-01
-4.19432968e-01 -4.43665773e-01 6.39644921e-01 7.60374725e-01
-4.16166037e-01 -1.40005171e+00 -5.60045093e-02 3.04254264e-01
-8.15281630e-01 -7.38079131e-01 -1.93643898e-01 5.06032825e-01
-2.02951312e-01 1.05385613e+00 4.64958221e-01 -2.57459581e-01
2.50212789e-01 1.94378465e-01 5.50708950e-01 -7.16651976e-01
-1.10761619e+00 1.79298446e-01 6.74837887e-01 -1.85873732e-01
2.93497629e-02 -5.88506699e-01 -1.61248791e+00 -3.91509622e-01
-7.18837559e-01 1.23381183e-01 5.96363068e-01 1.04327774e+00
1.86532110e-01 6.82895780e-01 2.01868802e-01 -6.38819814e-01
-6.39675736e-01 -1.25809920e+00 -2.85368115e-01 -9.77715179e-02
4.73499149e-01 1.17078915e-01 -2.26772018e-02 1.29898176e-01] | [10.829487800598145, 10.311607360839844] |
956514ee-0154-4852-a7c3-7274f1ba6f3d | surgical-gesture-recognition-based-on | 2105.00460 | null | https://arxiv.org/abs/2105.00460v1 | https://arxiv.org/pdf/2105.00460v1.pdf | Surgical Gesture Recognition Based on Bidirectional Multi-Layer Independently RNN with Explainable Spatial Feature Extraction | Minimally invasive surgery mainly consists of a series of sub-tasks, which can be decomposed into basic gestures or contexts. As a prerequisite of autonomic operation, surgical gesture recognition can assist motion planning and decision-making, and build up context-aware knowledge to improve the surgical robot control quality. In this work, we aim to develop an effective surgical gesture recognition approach with an explainable feature extraction process. A Bidirectional Multi-Layer independently RNN (BML-indRNN) model is proposed in this paper, while spatial feature extraction is implemented via fine-tuning of a Deep Convolutional Neural Network(DCNN) model constructed based on the VGG architecture. To eliminate the black-box effects of DCNN, Gradient-weighted Class Activation Mapping (Grad-CAM) is employed. It can provide explainable results by showing the regions of the surgical images that have a strong relationship with the surgical gesture classification results. The proposed method was evaluated based on the suturing task with data obtained from the public available JIGSAWS database. Comparative studies were conducted to verify the proposed framework. Results indicated that the testing accuracy for the suturing task based on our proposed method is 87.13%, which outperforms most of the state-of-the-art algorithms. | ['Benny Lo', 'Ruoxi Wang', 'Dandan Zhang'] | 2021-05-02 | null | null | null | null | ['surgical-gesture-recognition'] | ['medical'] | [ 1.98101372e-01 1.92556456e-01 -4.97450083e-01 -3.26314479e-01
-3.11718702e-01 -1.40312046e-01 3.46570939e-01 -4.86593992e-01
-6.82942986e-01 4.97407198e-01 5.61783075e-01 -4.50164795e-01
-4.86351073e-01 -3.62972170e-01 -4.25477028e-01 -1.12541473e+00
-5.72245521e-03 -3.86771224e-02 -1.74272433e-01 -1.77877679e-01
4.01244879e-01 6.79799378e-01 -1.16062391e+00 4.31989521e-01
5.09458005e-01 9.17943716e-01 5.43455243e-01 4.78441864e-01
-1.44996047e-01 8.21496785e-01 -4.05140907e-01 4.23312515e-01
2.54407197e-01 -5.37439227e-01 -7.53132641e-01 -3.30798626e-01
-1.29343256e-01 -4.03113663e-01 -4.67537522e-01 8.95647287e-01
5.73528826e-01 1.98220506e-01 4.75879520e-01 -7.70181179e-01
-3.76667529e-01 5.59133112e-01 3.08368094e-02 1.33159935e-01
-2.34739766e-01 2.24332854e-01 4.77310061e-01 -6.83822513e-01
9.08499539e-01 8.14320385e-01 3.30773890e-01 9.30889845e-01
-3.00255120e-01 -7.53569603e-01 2.09389940e-01 2.85642534e-01
-1.14284003e+00 1.36928827e-01 7.86843359e-01 -3.24706078e-01
7.54991591e-01 4.71739531e-01 7.78132021e-01 1.07441723e+00
9.96790707e-01 6.89882338e-01 9.66535628e-01 -3.51871282e-01
2.87135337e-02 -2.72845179e-01 -3.89160588e-02 9.75694895e-01
2.96090186e-01 4.70094383e-01 -3.65549654e-01 5.19566178e-01
1.31107938e+00 5.32211661e-01 -5.66350639e-01 -4.08666402e-01
-1.52821386e+00 6.95713639e-01 1.23307788e+00 9.64850664e-01
-5.31861424e-01 3.73245358e-01 2.76978642e-01 -1.19798789e-02
-1.68858126e-01 5.52035093e-01 -2.81189799e-01 -1.63546428e-01
-4.68386203e-01 -4.15677607e-01 6.14421308e-01 9.25393283e-01
1.42630786e-01 1.19181298e-01 -3.66027147e-01 3.05805713e-01
4.71914709e-01 1.89900063e-02 1.17860496e+00 -4.33563739e-01
3.11280549e-01 7.44292140e-01 -3.00115645e-01 -8.68302763e-01
-1.09711540e+00 -5.73874593e-01 -1.10783780e+00 4.79435325e-01
1.25889167e-01 -2.26680189e-01 -1.40302837e+00 1.37086701e+00
1.46957844e-01 1.41468376e-01 9.54192132e-02 1.41333604e+00
1.36923695e+00 1.98386699e-01 2.67447859e-01 -1.35018125e-01
1.14233804e+00 -1.18408251e+00 -1.03613067e+00 -3.01465411e-02
9.32011068e-01 -4.93824929e-01 8.53264630e-01 2.56735682e-01
-4.02240753e-01 -5.66583455e-01 -1.17952192e+00 -1.36114249e-03
-4.00518298e-01 6.46263540e-01 1.08764887e+00 4.39409405e-01
-9.80101466e-01 5.39141178e-01 -1.25033486e+00 -4.56950009e-01
3.34080547e-01 7.53347039e-01 -6.89180255e-01 5.91584966e-02
-8.60569298e-01 1.07889175e+00 4.82146442e-01 9.72097397e-01
-1.00460577e+00 -1.86922789e-01 -8.48625243e-01 -1.58920586e-01
2.03898549e-01 -6.71293497e-01 8.98461163e-01 -8.67428303e-01
-1.58553016e+00 5.64209580e-01 2.38844007e-02 -8.42681229e-02
3.99745613e-01 -2.15347037e-01 -2.36545473e-01 9.91714075e-02
-3.20122272e-01 5.55359602e-01 4.66965497e-01 -1.07945669e+00
-3.58021080e-01 -4.73760486e-01 -3.97323929e-02 6.22864127e-01
-2.37321571e-01 -3.29733998e-01 -5.33808410e-01 -7.84753501e-01
5.43633163e-01 -1.22464240e+00 -4.98233706e-01 3.34256589e-02
-4.70895529e-01 1.03558935e-01 7.16311038e-01 -8.14712226e-01
1.30578494e+00 -1.98085082e+00 3.95300865e-01 2.40602434e-01
1.59653768e-01 2.57340580e-01 -6.18637353e-02 1.64244354e-01
-8.00219029e-02 -6.30398793e-03 -1.14508554e-01 1.86551705e-01
-4.66044694e-01 4.63678896e-01 2.59532362e-01 5.57262123e-01
-1.77728862e-01 9.67197716e-01 -7.10677147e-01 -3.42776269e-01
6.47368431e-01 3.61916780e-01 -3.61501813e-01 3.85343671e-01
2.27321982e-01 9.01779890e-01 -6.91667914e-01 8.78963768e-01
1.44207224e-01 7.22977147e-02 2.23921865e-01 -3.85989994e-01
-1.03186779e-01 -1.48096234e-01 -6.62118495e-01 2.42849374e+00
-5.50353706e-01 5.09039342e-01 5.72497211e-02 -9.65742588e-01
9.68660116e-01 3.70528966e-01 7.30529487e-01 -5.46055496e-01
6.94327533e-01 2.72911102e-01 5.73853254e-01 -1.05294502e+00
2.17415452e-01 -1.58394486e-01 1.14830863e-02 -9.54714790e-02
8.61425996e-02 1.41113907e-01 -4.23141807e-01 -3.39060813e-01
9.52757716e-01 4.37315404e-01 4.95924234e-01 -3.07237387e-01
4.87375587e-01 3.26730698e-01 4.82659459e-01 6.13686800e-01
-3.76833469e-01 4.72613513e-01 1.38015479e-01 -7.49718249e-01
-4.32963461e-01 -5.14849782e-01 1.15960725e-02 5.52512228e-01
5.91486990e-01 1.45249203e-01 -5.99300683e-01 -9.18285906e-01
-2.81293601e-01 4.04504091e-01 -1.13583708e+00 -3.87948573e-01
-8.28297317e-01 -8.25928569e-01 3.10718149e-01 9.02213454e-01
5.83762467e-01 -1.54502308e+00 -9.70305383e-01 2.13507682e-01
1.04620241e-01 -8.81843686e-01 -2.04416141e-01 4.18415159e-01
-1.08562291e+00 -1.30029953e+00 -7.34307468e-01 -1.12697911e+00
9.99850154e-01 1.40668496e-01 2.22283214e-01 3.78043830e-01
-6.43378496e-01 5.48038855e-02 -5.73870659e-01 -4.27674830e-01
-1.40974104e-01 2.84155279e-01 -1.87123746e-01 -1.88907415e-01
3.50905180e-01 -1.04339764e-01 -9.41052437e-01 3.69852334e-01
-9.71457541e-01 4.08787042e-01 1.32769895e+00 1.22759187e+00
5.69049776e-01 -5.15517175e-01 1.69614881e-01 -6.55631602e-01
4.63164151e-01 -3.22619617e-01 -2.17433482e-01 1.38215765e-01
-6.58915460e-01 8.85309726e-02 3.24927092e-01 -4.03650850e-01
-1.01878393e+00 3.17636847e-01 -1.38427734e-01 -5.00357270e-01
-3.14107955e-01 8.11188698e-01 -3.64802741e-02 -3.82379681e-01
3.68776470e-01 1.54886171e-01 2.48258814e-01 -2.79916465e-01
1.72293708e-01 6.70309603e-01 5.74123859e-01 -7.76702240e-02
3.26318145e-01 3.33122313e-01 1.61736459e-01 -5.44188023e-01
-3.22455466e-01 -4.55613613e-01 -8.22992742e-01 -5.11695385e-01
1.29879284e+00 -6.16492569e-01 -8.35808098e-01 3.78427833e-01
-9.54924226e-01 -4.72574800e-01 1.80413708e-01 1.26328182e+00
-4.85518008e-01 -1.71300992e-01 -7.00981438e-01 -4.42644209e-01
-4.73089397e-01 -1.55809414e+00 8.43293965e-01 4.42279220e-01
-9.95006263e-02 -8.79964292e-01 -2.18309090e-01 1.73013046e-01
6.74714208e-01 6.04107440e-01 8.17221999e-01 -7.76878417e-01
-7.27084041e-01 -5.13574123e-01 -1.43718079e-01 3.33982036e-02
3.84820163e-01 -4.45660323e-01 -6.52740657e-01 -1.88050196e-01
-8.81327596e-03 1.46890312e-01 7.41374791e-01 7.40434766e-01
1.48786318e+00 -2.56011844e-01 -6.46334350e-01 1.04435742e+00
1.54729807e+00 6.74245775e-01 6.51908815e-01 6.51568413e-01
1.01910949e+00 2.80485570e-01 7.54069746e-01 2.14147381e-02
1.24044865e-01 5.61267138e-01 9.89426970e-01 -3.65279377e-01
-3.49020928e-01 8.01424358e-06 -6.03861399e-02 8.20509076e-01
-6.57970786e-01 1.86928168e-01 -9.66823161e-01 2.59055376e-01
-2.00376964e+00 -6.80190086e-01 9.12172068e-03 1.76611352e+00
4.18554187e-01 -1.49653018e-01 -6.61438465e-01 -1.82047099e-01
3.33904743e-01 8.57109949e-02 -5.34698725e-01 -2.51651466e-01
3.33772600e-01 2.76007056e-01 7.49849319e-01 4.30035472e-01
-1.28384936e+00 9.44523573e-01 5.36852264e+00 4.76406127e-01
-1.62373066e+00 -8.51008072e-02 2.96501637e-01 -1.98711023e-01
9.74152014e-02 -4.91413146e-01 -3.68059099e-01 1.57908395e-01
3.19154590e-01 3.29446614e-01 2.76918679e-01 9.56832051e-01
3.70763630e-01 -8.37469846e-03 -9.87020910e-01 1.00347829e+00
2.04311937e-01 -1.40623868e+00 6.24416918e-02 8.06949362e-02
5.64654589e-01 -3.79624479e-02 -1.21752605e-01 3.34992170e-01
-2.27006897e-01 -1.42517567e+00 2.48267412e-01 7.73443401e-01
9.07511115e-01 -4.08983111e-01 1.40833259e+00 3.05848718e-01
-1.16172826e+00 -3.43751818e-01 -8.08784887e-02 1.86345443e-01
-2.67251372e-01 -2.35939622e-01 -1.04686582e+00 7.58778751e-01
5.12596905e-01 7.40789115e-01 -2.19740435e-01 1.07375669e+00
-4.38744694e-01 2.44262338e-01 3.50786448e-02 -5.06223917e-01
5.27984679e-01 -1.51162278e-02 3.79959553e-01 1.12665284e+00
6.02682531e-01 4.06360656e-01 -8.32441449e-02 6.35944843e-01
5.76487556e-02 1.62199512e-01 -5.90163469e-01 1.20668702e-01
1.44732343e-02 1.38012981e+00 -8.68446350e-01 -2.14264207e-02
-1.92227736e-01 9.06300545e-01 -6.84218332e-02 4.27036673e-01
-6.70342982e-01 -4.88882095e-01 4.79294956e-01 -3.38843286e-01
2.27482300e-02 -3.03793043e-01 -6.04130983e-01 -9.50525463e-01
1.77404042e-02 -6.68036401e-01 2.58805305e-01 -7.36116290e-01
-4.42067116e-01 8.56326222e-01 -1.47398248e-01 -1.65323591e+00
-3.20668519e-01 -1.12581682e+00 -6.33930743e-01 8.98790121e-01
-1.44894159e+00 -1.32239723e+00 -1.05924332e+00 7.28255033e-01
6.42464101e-01 -1.78587034e-01 1.16855979e+00 -5.60909882e-02
-5.20030558e-01 4.34299886e-01 -2.57130533e-01 4.29181367e-01
4.56072897e-01 -8.35156024e-01 -4.02834535e-01 8.08871984e-01
-2.11560786e-01 8.32513928e-01 3.00586581e-01 -6.27358615e-01
-1.61392534e+00 -9.44863439e-01 3.14503461e-01 -5.05010411e-02
2.50296533e-01 1.91309676e-01 -4.92067218e-01 7.64226496e-01
6.07237779e-02 2.41336390e-01 6.05768204e-01 -2.33856142e-01
4.08565223e-01 1.01398125e-01 -1.01694155e+00 6.16281748e-01
1.25578916e+00 -7.73877054e-02 -5.92789829e-01 1.42467886e-01
4.48242188e-01 -9.79697049e-01 -7.31171906e-01 9.15944576e-01
8.18397760e-01 -5.99606991e-01 6.89200163e-01 -8.21780503e-01
6.21337295e-01 -2.97383815e-01 2.15612575e-02 -1.36113656e+00
-1.84476823e-01 -1.36026859e-01 1.78551838e-01 2.54962623e-01
4.58576411e-01 -4.11930889e-01 8.10303450e-01 6.75820529e-01
-6.45012915e-01 -1.14939070e+00 -9.65317845e-01 -4.15368289e-01
-2.26145297e-01 -2.73152858e-01 3.81478369e-01 1.02196705e+00
3.32399547e-01 -3.15717936e-01 -2.30982184e-01 1.51460975e-01
6.17901608e-02 5.27265035e-02 5.77127457e-01 -7.53588259e-01
5.18728979e-02 -7.28618979e-01 -7.01900423e-01 -8.06281269e-01
-2.09402785e-01 -1.00558782e+00 3.80197376e-01 -2.02766085e+00
-2.74701603e-02 -3.56554389e-01 -7.20120907e-01 8.49934220e-01
-5.84914126e-02 1.62381604e-02 6.81954250e-02 3.32251430e-01
-2.95141879e-02 4.17921036e-01 1.77067792e+00 -1.53001964e-01
-3.72178257e-01 9.37919691e-02 -4.60607201e-01 6.33197963e-01
8.37301195e-01 -2.48365089e-01 -2.30399638e-01 -5.02468288e-01
-4.85962987e-01 3.93571228e-01 2.32729584e-01 -1.03895414e+00
5.00582159e-01 -2.76257843e-01 5.64163983e-01 -4.58562642e-01
2.01869935e-01 -1.17580605e+00 3.63285653e-02 1.10271001e+00
-4.98979270e-01 -7.97876045e-02 2.28803784e-01 3.73059154e-01
-4.20202285e-01 7.17423484e-02 3.48440349e-01 -1.68795884e-01
-1.08279276e+00 4.43331689e-01 -2.42161885e-01 -7.63066292e-01
1.18602049e+00 -2.63066918e-01 -1.45975336e-01 -7.57258683e-02
-1.03596818e+00 -1.17527217e-01 -1.25362530e-01 7.18854845e-01
9.49835002e-01 -1.51384544e+00 -3.81156683e-01 3.27628762e-01
1.78572372e-01 1.60144240e-01 3.87555480e-01 1.20370364e+00
-1.01645637e+00 7.83610165e-01 -5.83925605e-01 -5.22507071e-01
-1.16559517e+00 2.81060666e-01 6.10700727e-01 -3.28480035e-01
-8.04109752e-01 8.62965524e-01 1.40515253e-01 -4.08496439e-01
5.41334093e-01 -8.40942025e-01 -7.31944978e-01 -3.86367023e-01
2.96820074e-01 -1.63628593e-01 1.45820588e-01 -4.76978838e-01
-5.33926010e-01 6.06669247e-01 2.73649096e-01 5.80234379e-02
1.34766829e+00 2.89189935e-01 1.30857721e-01 7.89329484e-02
1.11434162e+00 -4.85303432e-01 -1.02591133e+00 1.59547374e-01
-2.53983736e-01 -2.38203511e-01 3.67504299e-01 -1.26919937e+00
-1.38267004e+00 8.37473214e-01 9.56542492e-01 -4.80745882e-01
1.25054908e+00 -2.62765139e-01 5.65501332e-01 2.80133218e-01
5.74704051e-01 -7.98379362e-01 -2.35000253e-01 1.87638015e-01
1.27217209e+00 -1.39548683e+00 -2.05603912e-01 -1.48203447e-01
-7.69538820e-01 1.56192374e+00 8.78196180e-01 -1.97442248e-01
7.72153020e-01 1.61440820e-01 5.28226376e-01 -3.80232990e-01
-5.54148518e-02 -1.12441741e-01 6.66637897e-01 2.55278766e-01
5.02504826e-01 3.34712237e-01 -7.54681170e-01 1.01422143e+00
-1.02050267e-01 3.91917735e-01 2.43261501e-01 1.08669901e+00
-3.63843888e-01 -6.23609424e-01 -5.12733832e-02 4.01380509e-01
-2.98744887e-01 -7.48101845e-02 -1.15854450e-01 1.36345983e+00
4.90154400e-02 6.78391695e-01 -2.40621343e-01 -8.42033565e-01
3.72153342e-01 -3.71863060e-02 2.82350212e-01 -4.82531816e-01
-6.68113589e-01 1.01328589e-01 -2.99397670e-02 -1.13489616e+00
-6.18726373e-01 -2.05650553e-01 -1.66328096e+00 2.83149213e-01
-2.26487800e-01 -9.85694602e-02 1.16397047e+00 1.13103747e+00
1.82172637e-02 1.19996870e+00 2.83900946e-01 -1.07824540e+00
-2.49731228e-01 -1.22115135e+00 -3.69655937e-01 1.50532275e-01
3.58405411e-01 -7.01539814e-01 -2.25815997e-01 -2.07926318e-01] | [14.096013069152832, -3.3569281101226807] |
6f073397-cc9f-4d33-8bd1-55a1cfcc0e6a | probabilistic-color-constancy | 2005.02730 | null | https://arxiv.org/abs/2005.02730v1 | https://arxiv.org/pdf/2005.02730v1.pdf | Probabilistic Color Constancy | In this paper, we propose a novel unsupervised color constancy method, called Probabilistic Color Constancy (PCC). We define a framework for estimating the illumination of a scene by weighting the contribution of different image regions using a graph-based representation of the image. To estimate the weight of each (super-)pixel, we rely on two assumptions: (Super-)pixels with similar colors contribute similarly and darker (super-)pixels contribute less. The resulting system has one global optimum solution. The proposed method achieves competitive performance, compared to the state-of-the-art, on INTEL-TAU dataset. | ['Jarno Nikkanen', 'Alexandros Iosifidis', 'Uygar Tuna', 'Moncef Gabbouj', 'Jenni Raitoharju', 'Firas Laakom'] | 2020-05-06 | null | null | null | null | ['color-constancy'] | ['computer-vision'] | [ 3.24591309e-01 -4.11905110e-01 5.92392571e-02 -4.01921749e-01
-3.64226401e-01 -5.14047265e-01 3.28190833e-01 6.55193850e-02
-4.42063540e-01 5.93507051e-01 -1.74513564e-01 -2.97435876e-02
3.35534513e-01 -4.89319921e-01 -4.26051319e-01 -9.97259617e-01
1.03449315e-01 6.63324399e-03 4.35292333e-01 9.68754813e-02
4.46506917e-01 4.07221854e-01 -1.44353974e+00 1.59939051e-01
9.19989884e-01 9.19041455e-01 1.63469434e-01 8.60469997e-01
-1.29513681e-01 7.32545376e-01 -4.37425882e-01 -2.00946569e-01
4.29812729e-01 -7.49986172e-01 -4.89776194e-01 3.12444627e-01
8.82487357e-01 6.89215139e-02 -1.69885844e-01 1.52545381e+00
1.00024901e-01 2.59599537e-01 6.52924180e-01 -1.33091259e+00
-8.93302679e-01 -3.45529639e-03 -1.23495936e+00 2.72567689e-01
-1.51234105e-01 -1.51723459e-01 1.09449446e+00 -7.79082060e-01
5.41294754e-01 1.03282475e+00 1.21388800e-01 4.15565103e-01
-1.51224935e+00 -2.64896214e-01 4.19522643e-01 2.10684240e-01
-1.33471477e+00 -2.60422468e-01 1.09544134e+00 -1.71027169e-01
4.88821417e-01 5.83288133e-01 4.77341682e-01 3.46061945e-01
2.93160886e-01 7.50398040e-01 1.83371103e+00 -6.99771762e-01
5.38149536e-01 1.29686207e-01 2.11129472e-01 9.18767512e-01
4.34801072e-01 -1.23239987e-01 -6.79569185e-01 -7.29657263e-02
5.47613740e-01 -5.14240041e-02 -2.47141778e-01 -5.93727887e-01
-1.00571239e+00 5.59639931e-01 7.04096079e-01 2.59125620e-01
-2.57037818e-01 5.11279881e-01 -1.08449437e-01 2.94496771e-02
5.81067145e-01 2.78668851e-01 -1.99868739e-01 1.33867040e-01
-9.90950584e-01 -9.40666646e-02 5.37348866e-01 7.78690994e-01
1.15481901e+00 2.02166766e-01 -3.22219789e-01 7.61467874e-01
2.79513478e-01 5.00241399e-01 1.49165571e-01 -1.19989371e+00
1.43134862e-01 4.99801189e-01 2.95765281e-01 -1.04256868e+00
-4.10128564e-01 -4.31934893e-01 -8.09880435e-01 7.15126514e-01
4.44921911e-01 -5.22835143e-02 -1.30494225e+00 1.69208050e+00
3.79614145e-01 6.42773211e-02 -7.54207745e-02 9.56740022e-01
3.30425352e-01 7.23511457e-01 -3.23488447e-03 -1.88924402e-01
1.24765277e+00 -1.19167769e+00 -5.61886370e-01 -3.71916503e-01
-3.38269860e-01 -9.65928555e-01 9.86108184e-01 6.96530044e-01
-1.00454426e+00 -6.12503648e-01 -1.20776343e+00 2.05950998e-02
-4.48206574e-01 1.62992194e-01 8.30975235e-01 1.07986701e+00
-1.32488775e+00 3.35757315e-01 -4.80083942e-01 -2.34948978e-01
5.54098599e-02 -1.43263554e-02 -1.55825809e-01 -3.49500328e-01
-4.59372640e-01 6.41539812e-01 3.97143424e-01 5.78145981e-02
-6.58109605e-01 -4.08186227e-01 -6.42276347e-01 -2.21842200e-01
4.03243452e-02 -4.50857192e-01 7.45685995e-01 -1.53418863e+00
-1.44389927e+00 9.59580660e-01 -5.20327151e-01 6.49498999e-02
3.43720973e-01 -3.63883898e-02 -4.15443659e-01 2.39020392e-01
-1.77274182e-01 4.92873967e-01 1.10338163e+00 -2.01425266e+00
-7.25791454e-01 -4.55822527e-01 3.52472216e-02 3.69438052e-01
-1.00373603e-01 1.15369052e-01 -1.00446081e+00 -4.72479969e-01
5.17978370e-01 -9.81275916e-01 -2.15214297e-01 2.13601038e-01
-6.32430196e-01 1.51260927e-01 8.16188395e-01 -6.43866181e-01
9.58744884e-01 -2.27509999e+00 6.12936802e-02 5.11045396e-01
4.02267843e-01 -1.64700225e-01 -1.38708785e-01 -3.51885743e-02
-1.66555077e-01 -2.72816986e-01 -4.62687284e-01 -5.10476887e-01
-1.06641121e-01 3.77352685e-01 1.07216470e-01 8.47956717e-01
1.44069552e-01 4.08625215e-01 -1.10601676e+00 -6.32299006e-01
3.95401329e-01 4.72806334e-01 -4.34158146e-01 1.61394626e-01
1.20081464e-02 3.73581648e-01 -2.05392390e-01 7.47147799e-01
1.34902203e+00 -9.15207192e-02 2.83507317e-01 -4.38789248e-01
-5.33395648e-01 -4.11847353e-01 -1.39786685e+00 1.53291249e+00
-2.82811254e-01 7.65471935e-01 7.71403015e-02 -6.34707332e-01
8.54359448e-01 -3.96967620e-01 4.43035662e-01 -7.87482500e-01
-4.21234965e-02 -2.00683512e-02 -2.92153843e-02 -1.03476189e-01
9.33848917e-01 2.24493463e-02 2.53531963e-01 5.67966342e-01
-7.80128464e-02 -2.54739255e-01 3.74556661e-01 2.42715657e-01
7.54992545e-01 2.38547742e-01 1.29993156e-01 -5.14853001e-01
4.66213048e-01 -2.23807916e-01 5.37578046e-01 6.64576054e-01
-5.23192704e-01 8.31684530e-01 5.71963727e-01 -1.07463174e-01
-9.39562500e-01 -1.34330535e+00 -7.53415525e-02 1.30286610e+00
7.39378810e-01 -1.05680995e-01 -8.34480643e-01 -3.74210954e-01
4.92852821e-04 8.43310773e-01 -8.39108288e-01 1.08507499e-01
-3.67774725e-01 -9.32231188e-01 -1.79901555e-01 2.61882186e-01
6.37197018e-01 -6.00802720e-01 -7.74631381e-01 -2.36165822e-01
-1.32987067e-01 -9.72607434e-01 -6.15221262e-01 4.01594430e-01
-6.97969317e-01 -8.87606323e-01 -8.82774770e-01 -5.70573688e-01
1.14766407e+00 8.41897845e-01 1.31269026e+00 8.32083747e-02
-4.00049478e-01 5.00002086e-01 -3.14320147e-01 -1.12488978e-01
2.82085501e-02 -4.07340348e-01 -3.18143457e-01 4.15728390e-01
-2.65583936e-02 4.82507758e-02 -9.67379570e-01 2.32514068e-01
-9.57945168e-01 2.82772362e-01 6.13813639e-01 6.64822817e-01
1.08257473e+00 2.01918721e-01 -3.62448394e-01 -1.02426541e+00
3.30863535e-01 -1.26997754e-01 -6.52794003e-01 5.69361985e-01
-7.42423952e-01 5.35246253e-01 4.40448165e-01 -4.71167900e-02
-1.40320802e+00 2.94153750e-01 5.83992839e-01 -2.13700399e-01
5.76940700e-02 -1.83187742e-02 -8.28966275e-02 -4.56551820e-01
4.26061183e-01 1.39685169e-01 -4.92336810e-01 -3.61031830e-01
9.46333230e-01 2.05418274e-01 9.44139242e-01 -6.97860181e-01
8.60075295e-01 8.56440127e-01 3.30702752e-01 -6.61802649e-01
-6.61777675e-01 -7.69736469e-01 -6.40062451e-01 -4.56891537e-01
8.77182126e-01 -5.97253382e-01 -3.84972960e-01 5.23882866e-01
-1.01125610e+00 -1.89959303e-01 -6.72607720e-02 1.82452962e-01
-2.42847666e-01 4.66538131e-01 -1.37437060e-01 -1.18538237e+00
-1.30506337e-01 -9.05723333e-01 9.05557930e-01 7.04729259e-01
3.99347216e-01 -9.62707818e-01 3.52839500e-01 -5.21168411e-02
4.44554657e-01 5.10291398e-01 6.36843145e-01 2.36305445e-01
-5.07167399e-01 -4.64972630e-02 -9.70502973e-01 4.02217716e-01
2.38638043e-01 4.49332535e-01 -1.22323489e+00 -1.85105309e-01
-3.06017309e-01 -1.89964324e-02 1.23241127e+00 4.18760449e-01
1.27210248e+00 -2.78808195e-02 3.85368839e-02 8.05000365e-01
2.04968905e+00 -9.79979560e-02 7.50127971e-01 2.00240061e-01
8.38751018e-01 5.18979549e-01 5.14449835e-01 5.75199783e-01
4.17970806e-01 5.33855975e-01 7.37796128e-01 -6.10777318e-01
-3.40543777e-01 3.69708277e-02 2.25720391e-01 4.32812601e-01
-3.83369058e-01 -2.01124609e-01 -4.90926236e-01 4.06005621e-01
-1.76766193e+00 -7.25733280e-01 -6.29983842e-01 2.41822076e+00
7.91847527e-01 -3.51612233e-02 -2.12528035e-02 -4.03741747e-02
8.47449124e-01 3.63704532e-01 -6.12989306e-01 -5.47134936e-01
-4.40074205e-01 3.38445514e-01 1.04722285e+00 4.62864906e-01
-1.06453717e+00 7.87166893e-01 7.20225477e+00 4.81840789e-01
-1.06201172e+00 1.29957080e-01 1.01696026e+00 2.16461018e-01
-4.35719311e-01 1.64244279e-01 -2.49994665e-01 4.22049880e-01
4.13709939e-01 -4.94473651e-02 6.97571635e-01 8.40291619e-01
2.93888189e-02 -7.56110251e-01 -7.75628388e-01 1.09682703e+00
5.92434525e-01 -7.17000842e-01 -1.70419961e-01 -3.27193171e-01
1.29983604e+00 1.96347549e-03 4.87420410e-01 -3.29937071e-01
4.70071465e-01 -6.26918972e-01 9.78332520e-01 8.81459773e-01
8.20425868e-01 -6.61768258e-01 4.51988757e-01 -4.93325293e-01
-1.40189362e+00 2.56460488e-01 -6.76341414e-01 3.15696955e-01
-1.96530342e-01 8.70741606e-01 -1.08455703e-01 5.56275487e-01
7.99281776e-01 3.85149151e-01 -9.77194190e-01 1.28588355e+00
-4.16732788e-01 2.79947490e-01 -1.82031631e-01 1.94178056e-02
1.50344327e-01 -5.18601835e-01 1.61696538e-01 1.38758826e+00
6.32241815e-02 -1.85074583e-01 7.67592490e-02 1.04759145e+00
-1.30520150e-01 -8.39084163e-02 -1.25989690e-01 2.31123969e-01
4.60608974e-02 1.72546649e+00 -1.11922657e+00 -3.79915148e-01
-5.53512216e-01 1.66857302e+00 2.89114565e-01 8.26652646e-01
-6.46015346e-01 -3.44716728e-01 5.93296111e-01 -3.77476633e-01
1.26073569e-01 -4.59164977e-01 -6.85543001e-01 -1.02377856e+00
-1.71471223e-01 -4.14859354e-01 5.12533128e-01 -1.22354639e+00
-1.18656433e+00 3.75810355e-01 -2.37483859e-01 -1.30274868e+00
4.12698001e-01 -8.91687870e-01 -7.21100271e-01 9.72069502e-01
-2.02116585e+00 -1.07948244e+00 -7.70106614e-01 7.71626234e-01
1.86057165e-01 2.68891871e-01 6.32757008e-01 -9.65910554e-02
-5.11947036e-01 3.58552456e-01 4.54242259e-01 -1.79985851e-01
8.94747436e-01 -1.85832548e+00 1.82822630e-01 1.34185648e+00
3.11945111e-01 4.71635789e-01 8.98612261e-01 -2.53979206e-01
-1.55722857e+00 -9.59315121e-01 5.03114402e-01 -1.90224379e-01
5.02759576e-01 -5.26030362e-02 -4.62410152e-01 5.29966727e-02
5.53979695e-01 1.56514198e-01 5.75312734e-01 6.14015050e-02
-7.10037887e-01 -3.37425321e-01 -1.24416828e+00 5.73955655e-01
5.48728824e-01 -4.01346862e-01 -5.71124367e-02 2.36012176e-01
3.50679934e-01 -2.70844609e-01 -2.71877706e-01 -2.21532166e-01
5.69495201e-01 -1.42450285e+00 8.30511808e-01 -2.17207521e-01
2.77286649e-01 -6.55399919e-01 -2.51499593e-01 -1.43622589e+00
-6.20739937e-01 -3.82637411e-01 1.45118460e-01 8.49731565e-01
2.84962922e-01 -5.56125879e-01 6.66951120e-01 8.57093215e-01
-6.58661574e-02 -2.18581825e-01 -9.31431949e-01 -4.92037624e-01
-3.61181885e-01 -2.82578498e-01 2.46833384e-01 7.69446969e-01
-3.19340438e-01 -1.55553222e-01 -5.21852076e-01 3.17138314e-01
1.10759389e+00 3.12672645e-01 5.07663667e-01 -9.98953760e-01
-3.03399414e-01 -5.61138093e-01 -4.17387784e-01 -5.38092315e-01
-2.00222105e-01 -5.24109185e-01 1.47545829e-01 -1.55960143e+00
7.40317881e-01 -2.72098064e-01 -9.28863049e-01 5.05844533e-01
-5.69355845e-01 9.05739129e-01 3.02572250e-01 5.94662949e-02
-9.42913115e-01 2.19424263e-01 1.19671094e+00 -3.95792693e-01
-6.93981051e-02 -3.51696134e-01 -7.12091208e-01 5.34367561e-01
7.98148751e-01 -2.13469371e-01 -8.23216662e-02 -4.60618228e-01
3.20724219e-01 -6.48934722e-01 3.02663893e-01 -9.10375714e-01
9.14511532e-02 -4.42091495e-01 6.63661122e-01 -4.32231426e-01
3.36632788e-01 -9.10701036e-01 -1.28710598e-01 3.83071095e-01
-8.24611112e-02 1.20577820e-01 1.91987842e-01 5.27781546e-01
-5.17505370e-02 -3.30243200e-01 1.23207891e+00 -5.05934730e-02
-1.09242845e+00 1.44254183e-02 -1.09271348e-01 -5.45551218e-02
9.07408953e-01 -3.62807602e-01 -5.96822083e-01 -3.72874737e-01
-2.16962278e-01 -7.39374608e-02 8.22960138e-01 8.94021839e-02
7.05006182e-01 -1.32476532e+00 -6.17045760e-01 1.80288672e-01
4.23548549e-01 -6.54186904e-01 3.08966190e-01 6.34545326e-01
-7.87171304e-01 3.45650613e-02 -4.24947768e-01 -5.01082659e-01
-1.33803904e+00 4.36092943e-01 1.97451308e-01 1.23203686e-02
-1.67849377e-01 1.00064981e+00 4.00265843e-01 8.79136473e-02
-1.11847237e-01 -1.80108175e-01 3.64480056e-02 -3.40121657e-01
4.98770386e-01 4.88222152e-01 -8.02516490e-02 -8.12695324e-01
-4.29974318e-01 7.94325709e-01 2.26031587e-01 -3.30513567e-01
1.08219182e+00 -1.89659998e-01 -5.63975155e-01 5.67899644e-01
1.07391071e+00 1.73826426e-01 -1.43000281e+00 -1.92503616e-01
-3.36226583e-01 -1.10004604e+00 5.74033976e-01 -9.21120405e-01
-1.40413320e+00 5.80482125e-01 1.16030526e+00 2.44868606e-01
1.50533187e+00 -8.50635096e-02 9.42762196e-02 6.13927059e-02
3.62751365e-01 -1.44999886e+00 1.59559637e-01 2.33700499e-01
5.87719142e-01 -1.26238358e+00 4.03579891e-01 -4.48178113e-01
-9.60467517e-01 1.17590427e+00 4.22219604e-01 -2.45606109e-01
4.47079688e-01 -3.97117846e-02 1.50759757e-01 -1.52446017e-01
-3.53414863e-01 -5.93259037e-01 7.32734799e-01 7.48863995e-01
6.39343500e-01 4.14711714e-01 -2.42190599e-01 -2.82094717e-01
2.33917966e-01 -5.70405424e-01 4.38356608e-01 6.92214370e-01
-5.43458343e-01 -9.81702924e-01 -6.76704645e-01 2.72984713e-01
-1.22664295e-01 -2.32453600e-01 -6.29131436e-01 5.37458599e-01
2.03032166e-01 1.15910041e+00 2.13666692e-01 -1.68937057e-01
6.97657987e-02 -2.34082535e-01 8.17295730e-01 -2.61215121e-01
-1.65754944e-01 2.26551414e-01 -2.11932078e-01 -8.40773940e-01
-7.87252247e-01 -5.94411075e-01 -1.10039508e+00 -2.69822061e-01
-1.42891571e-01 -9.85015184e-02 9.71272647e-01 3.01351368e-01
-2.49228645e-02 6.35289192e-01 1.03471160e+00 -9.71856833e-01
7.15530664e-02 -6.11517608e-01 -1.10909677e+00 5.02603412e-01
1.80061400e-01 -5.28922260e-01 -4.24744576e-01 2.60853976e-01] | [10.464164733886719, -2.557286262512207] |
ae4bd6e1-f5ef-46d8-9c9d-0f4a99698abb | robust-learning-through-cross-task-1 | 2006.04096 | null | https://arxiv.org/abs/2006.04096v1 | https://arxiv.org/pdf/2006.04096v1.pdf | Robust Learning Through Cross-Task Consistency | Visual perception entails solving a wide set of tasks, e.g., object detection, depth estimation, etc. The predictions made for multiple tasks from the same image are not independent, and therefore, are expected to be consistent. We propose a broadly applicable and fully computational method for augmenting learning with Cross-Task Consistency. The proposed formulation is based on inference-path invariance over a graph of arbitrary tasks. We observe that learning with cross-task consistency leads to more accurate predictions and better generalization to out-of-distribution inputs. This framework also leads to an informative unsupervised quantity, called Consistency Energy, based on measuring the intrinsic consistency of the system. Consistency Energy correlates well with the supervised error (r=0.67), thus it can be employed as an unsupervised confidence metric as well as for detection of out-of-distribution inputs (ROC-AUC=0.95). The evaluations are performed on multiple datasets, including Taskonomy, Replica, CocoDoom, and ApolloScape, and they benchmark cross-task consistency versus various baselines including conventional multi-task learning, cycle consistency, and analytical consistency. | ['Zhangjie Cao', 'Oğuzhan Kar', 'Teresa Yeo', 'Rohan Suri', 'Nikhil Cheerla', 'Leonidas Guibas', 'Alexander Sax', 'Jitendra Malik', 'Amir Zamir'] | 2020-06-07 | null | null | null | null | ['surface-normals-estimation'] | ['computer-vision'] | [ 6.85841665e-02 -2.22368225e-01 8.22926015e-02 -4.94257331e-01
-7.92082369e-01 -4.10995275e-01 7.44224966e-01 2.37053514e-01
-4.37705427e-01 9.13036644e-01 -3.60683888e-01 1.16073474e-01
-3.79349113e-01 -3.14969599e-01 -8.20601463e-01 -7.51975954e-01
5.53333312e-02 2.45113686e-01 5.61680138e-01 3.31745774e-01
4.95666385e-01 1.25122949e-01 -1.67206573e+00 1.56691641e-01
9.64521348e-01 1.44716048e+00 5.19071221e-01 5.00810444e-01
4.01674062e-01 5.62446296e-01 -5.11164188e-01 -4.22032028e-01
1.45478398e-01 -2.40406826e-01 -6.74265146e-01 -1.01926811e-01
7.31868207e-01 1.53583288e-01 2.53237695e-01 1.34396946e+00
2.99924642e-01 1.12341642e-01 1.05525815e+00 -1.53947663e+00
-6.24403894e-01 -1.56217366e-01 -5.96766353e-01 1.30393878e-01
1.26286641e-01 1.38618112e-01 1.06695569e+00 -9.61628020e-01
4.96502310e-01 1.28917086e+00 7.53966093e-01 2.54616737e-01
-1.24737990e+00 -5.71604371e-01 1.33445889e-01 4.52696949e-01
-1.35585845e+00 -1.77192032e-01 4.89262402e-01 -8.00537527e-01
7.05787420e-01 -2.83833686e-02 3.75037313e-01 1.31856465e+00
1.04211903e+00 4.09237683e-01 1.84075284e+00 -3.79950762e-01
4.37778383e-01 3.75393361e-01 1.41386688e-01 6.53106272e-01
6.30812883e-01 1.38264850e-01 -7.45072842e-01 1.42437458e-01
6.63296521e-01 -3.13874155e-01 -1.60593733e-01 -5.11962175e-01
-1.14357805e+00 4.27985460e-01 3.05671304e-01 1.06569238e-01
-7.88879916e-02 2.63133757e-02 4.42727149e-01 2.95561254e-01
5.06777227e-01 2.07315326e-01 -2.84161687e-01 2.03332573e-01
-6.11829102e-01 3.23790938e-01 4.84691799e-01 9.51005697e-01
7.68417180e-01 -1.32490993e-01 -4.26025003e-01 8.47597778e-01
4.09087956e-01 6.71417892e-01 4.92912531e-01 -9.95223641e-01
3.56578290e-01 2.63917297e-01 2.12807938e-01 -8.61463428e-01
-6.83310926e-01 -4.86706555e-01 -1.21158886e+00 6.43185318e-01
5.13884902e-01 4.58262712e-02 -6.70802534e-01 1.92272830e+00
1.88223153e-01 -1.42515585e-01 1.08097672e-01 8.37219954e-01
5.33209026e-01 2.50293434e-01 1.12114303e-01 -3.20548654e-01
1.32257283e+00 -9.69204783e-01 -7.52112687e-01 -4.51544315e-01
7.10771605e-02 -7.24809408e-01 1.14968050e+00 8.47780526e-01
-8.01898658e-01 -1.25409389e+00 -1.18348920e+00 1.25527978e-01
-3.13318551e-01 1.74470887e-01 3.76214415e-01 3.66702467e-01
-8.94798398e-01 6.61821663e-01 -6.22141898e-01 -3.41239840e-01
3.74370843e-01 -2.91894265e-02 -4.51859057e-01 2.35233288e-02
-6.84481919e-01 1.23454094e+00 7.33288229e-01 -2.70318203e-02
-9.64654446e-01 -3.54723036e-01 -5.91138363e-01 -2.02190861e-01
2.56789416e-01 -7.59673297e-01 1.05121553e+00 -8.19675982e-01
-1.08294046e+00 8.76806378e-01 -2.18265161e-01 -2.81398505e-01
7.46041179e-01 -3.45734388e-01 -3.77350479e-01 -1.95989549e-01
3.99072766e-01 7.30344415e-01 1.13290823e+00 -1.40233433e+00
-7.13145256e-01 -5.31613767e-01 -1.07100740e-01 1.83908522e-01
-1.42149791e-01 -3.89891058e-01 -3.99050683e-01 -4.81765836e-01
3.81513029e-01 -1.05949986e+00 2.52491891e-01 2.02561781e-01
-4.67747360e-01 -3.76209348e-01 5.01562655e-01 -6.57144070e-01
9.46357369e-01 -2.01824570e+00 1.22145854e-01 2.07220599e-01
1.28535762e-01 -2.33462542e-01 5.93741275e-02 2.63731945e-02
1.16718456e-01 -6.15624599e-02 -1.08865730e-01 -5.91085017e-01
-1.24308847e-01 2.16012791e-01 -1.77603904e-02 5.23959577e-01
2.18024716e-01 6.78153217e-01 -7.62413442e-01 -7.06080198e-01
2.59187430e-01 -5.00840740e-03 -1.25188500e-01 2.22417697e-01
-2.47948349e-01 6.59600556e-01 -1.95743352e-01 6.02530718e-01
7.17749894e-01 -6.53370857e-01 1.45323053e-01 -3.83194566e-01
-6.33084923e-02 -2.06536919e-01 -1.32319200e+00 1.76905429e+00
-5.69000006e-01 7.63941109e-01 -5.40299892e-01 -8.90866101e-01
9.70956802e-01 4.23763394e-02 1.21044712e-02 -1.03379023e+00
-1.11128159e-01 2.83478588e-01 6.74691126e-02 -5.55735111e-01
3.29652190e-01 2.03012705e-01 4.73528691e-02 1.99139848e-01
2.62267590e-01 -2.48986766e-01 1.09406285e-01 -1.04983017e-01
6.53299689e-01 4.31682497e-01 7.15622962e-01 -5.13628721e-01
4.78239506e-01 -3.28127801e-01 3.93716782e-01 8.35805655e-01
-3.63122433e-01 4.41525578e-01 4.41929907e-01 -2.49552697e-01
-1.13017333e+00 -1.24182498e+00 -6.28976703e-01 7.30881214e-01
4.06031311e-01 -8.11254382e-02 -4.61795151e-01 -5.78106463e-01
1.36170104e-01 2.94133693e-01 -7.55811751e-01 -7.61307403e-02
3.51286791e-02 -5.02320826e-01 2.68699288e-01 6.47824764e-01
8.81018519e-01 -8.32922578e-01 -7.36609519e-01 -5.96776307e-02
-3.66490513e-01 -1.64223683e+00 -1.57031432e-01 2.98903883e-01
-8.58389854e-01 -1.20086300e+00 -4.36889708e-01 -5.72743952e-01
4.31595802e-01 2.39031702e-01 1.09778309e+00 -2.68479973e-01
-1.07107453e-01 4.22905952e-01 -5.36125824e-02 -4.39568788e-01
-1.93725139e-01 -3.60064894e-01 4.14373159e-01 3.66950333e-02
-2.04129871e-02 -3.52088124e-01 -4.74303365e-01 5.79989851e-01
-5.98201692e-01 2.30018981e-02 6.04813099e-01 1.12942898e+00
1.00020242e+00 -1.45818666e-01 6.96213484e-01 -7.05238819e-01
4.75446224e-01 -2.15650588e-01 -9.40125823e-01 6.67552888e-01
-1.15407097e+00 2.82721579e-01 5.33552289e-01 -4.31014091e-01
-1.18906665e+00 -3.65261957e-02 3.35602671e-01 -5.57502627e-01
-1.87293082e-01 9.30926651e-02 1.89959276e-02 -2.74385065e-01
8.52157235e-01 8.31485093e-02 -2.09848344e-01 -1.77848980e-01
9.18203667e-02 3.19410980e-01 6.89708710e-01 -5.30967772e-01
5.12494743e-01 4.23011303e-01 4.83210504e-01 -7.79785812e-01
-1.08661211e+00 -5.54948807e-01 -6.55324578e-01 -5.36570609e-01
1.03515279e+00 -1.06602609e+00 -8.40318620e-01 7.83872068e-01
-1.45878327e+00 -2.79085696e-01 1.84751734e-01 6.88279450e-01
-6.92503512e-01 4.17508483e-01 -2.22202446e-02 -1.00214803e+00
-1.91567525e-01 -8.88859510e-01 1.03969502e+00 1.82901353e-01
-1.55459508e-01 -1.30409420e+00 1.42255187e-01 1.77226275e-01
3.19954194e-02 3.67618054e-01 8.50277603e-01 -5.54445684e-01
-5.85596383e-01 1.98445842e-01 -5.48973143e-01 7.24772394e-01
1.36982664e-01 6.64364770e-02 -1.55153430e+00 -3.75892967e-01
7.24921972e-02 -7.24269807e-01 8.65119278e-01 6.31312370e-01
1.27518511e+00 -7.58106187e-02 -2.97879249e-01 4.68081117e-01
1.73919201e+00 -6.53018653e-02 5.07390797e-01 3.12123805e-01
5.31613946e-01 7.01243281e-01 9.36159074e-01 5.85008919e-01
4.00850803e-01 7.49924541e-01 5.62462091e-01 3.14975381e-01
-4.79484200e-02 5.97338080e-02 2.31163517e-01 5.03824651e-01
-1.98679045e-01 -2.51900643e-01 -8.59650195e-01 2.85395116e-01
-1.97985053e+00 -9.05487180e-01 -3.05175394e-01 2.32747507e+00
6.27269268e-01 5.87787032e-01 2.13314388e-02 6.73673451e-02
6.55529022e-01 -1.18146941e-01 -7.12550163e-01 -1.49480447e-01
-2.34008700e-01 8.08426514e-02 2.70410568e-01 3.95336598e-01
-1.18306077e+00 5.89415193e-01 6.21738720e+00 8.71669948e-01
-8.14371347e-01 3.55670571e-01 7.06874192e-01 2.89324284e-01
6.62040636e-02 -1.33179858e-01 -9.25657392e-01 5.25140882e-01
6.08659148e-01 -6.72692806e-02 5.65413721e-02 9.46104646e-01
-3.74420062e-02 -7.05797613e-01 -1.42546308e+00 1.24039471e+00
3.08804393e-01 -9.80827808e-01 -2.56458759e-01 -1.97578624e-01
1.15141892e+00 -1.47606641e-01 1.88944817e-01 1.43237129e-01
-1.08909542e-02 -8.03249896e-01 9.94034529e-01 7.78563619e-01
8.84009778e-01 -3.37101609e-01 7.60179102e-01 4.08516347e-01
-1.25053978e+00 -1.34535044e-01 -4.69394654e-01 9.66173112e-02
-1.64317489e-01 7.50539124e-01 -3.96518797e-01 6.73816621e-01
1.08673978e+00 8.25400770e-01 -9.15545166e-01 1.13641167e+00
-2.31286809e-01 -1.13757327e-01 -2.21208911e-02 -1.22136222e-02
-7.15213940e-02 -1.60951495e-01 2.67729282e-01 1.15137708e+00
2.85482049e-01 -3.97725940e-01 1.11928038e-01 8.22200060e-01
2.58459896e-01 -1.37325242e-01 -7.45097041e-01 5.73080242e-01
4.77812529e-01 1.26889408e+00 -5.32957315e-01 -3.94059539e-01
-4.74063665e-01 1.06268346e+00 5.05566776e-01 2.74233788e-01
-9.25768495e-01 -1.51589483e-01 5.04191756e-01 -4.68692005e-01
9.66553167e-02 -2.48173565e-01 -6.35083616e-01 -9.30131793e-01
4.55778688e-01 -6.94703519e-01 3.43407959e-01 -9.75920916e-01
-1.58844388e+00 5.69470704e-01 2.26705879e-01 -1.49693406e+00
-2.14690804e-01 -9.23005223e-01 -5.65630019e-01 8.97738039e-01
-1.80382621e+00 -1.02068079e+00 -9.91759777e-01 7.22348988e-01
4.33129311e-01 -2.55665213e-01 6.04816139e-01 7.20028877e-02
-4.97629613e-01 6.14943862e-01 1.07410669e-01 -3.06229770e-01
1.04359567e+00 -1.29757893e+00 7.52631351e-02 8.76359761e-01
1.59121826e-01 1.99461475e-01 5.97282529e-01 -5.37387669e-01
-8.84124756e-01 -1.23626685e+00 4.59693849e-01 -6.59831762e-01
6.73080921e-01 -2.39676401e-01 -9.87030625e-01 4.67989236e-01
3.92841585e-02 1.71436757e-01 2.75871456e-01 -3.01751331e-03
-6.76203132e-01 -4.92710739e-01 -8.52186203e-01 2.12193936e-01
1.16434681e+00 -8.05885851e-01 -5.13967812e-01 6.92247152e-01
4.23386276e-01 -4.34050053e-01 -8.25292528e-01 4.78263617e-01
8.19446981e-01 -1.39068866e+00 7.71667421e-01 -2.70896077e-01
4.68523234e-01 -6.08448423e-02 -3.71435374e-01 -1.16627431e+00
-3.74781251e-01 -1.31162524e-01 -1.04721330e-01 1.04923987e+00
3.15898687e-01 -6.70650601e-01 1.65769696e-01 1.16776310e-01
-3.11206162e-01 -5.45872629e-01 -1.08076549e+00 -1.42415822e+00
-8.48682970e-02 -5.62922418e-01 -4.22027968e-02 6.84767187e-01
-3.50139856e-01 3.37153047e-01 -4.42795694e-01 4.40068156e-01
1.06794572e+00 2.83517968e-02 7.24352419e-01 -1.45408523e+00
-5.23061275e-01 -3.35081100e-01 -4.03666586e-01 -9.15688276e-01
2.24556923e-01 -6.97463155e-01 3.27738643e-01 -1.38344765e+00
3.66751730e-01 -4.84156609e-01 -3.55605334e-01 3.77157807e-01
-1.47758588e-01 2.39742950e-01 -1.97952464e-02 4.87039536e-01
-8.45019877e-01 4.25937414e-01 1.35773969e+00 -1.04812449e-02
5.88938743e-02 8.92461687e-02 -1.77932948e-01 9.13328111e-01
7.84431756e-01 -3.45804572e-01 -4.21721488e-01 -2.91229457e-01
2.66975433e-01 -5.04747108e-02 8.42196107e-01 -1.61089087e+00
1.01152904e-01 -2.23403960e-01 5.58937907e-01 -3.56936365e-01
3.58301431e-01 -8.11496019e-01 8.94053653e-02 5.76878905e-01
-2.43671164e-01 2.60057747e-01 3.31065655e-01 7.94408143e-01
-1.76198527e-01 -3.46674383e-01 1.11679971e+00 4.24852483e-02
-1.11411178e+00 1.39251444e-02 1.92593545e-01 3.08654189e-01
1.09635234e+00 -2.92814583e-01 -6.60627067e-01 -1.60925955e-01
-5.81772447e-01 1.04190305e-01 1.78601667e-01 4.60869849e-01
6.21537507e-01 -1.41011930e+00 -5.66378653e-01 9.53388140e-02
6.04546428e-01 -2.64776081e-01 2.75110215e-01 8.70554209e-01
-6.14664219e-02 5.01703978e-01 -6.11161947e-01 -1.40344465e+00
-1.23903692e+00 3.98999661e-01 2.84470201e-01 -1.55726209e-01
-1.78551838e-01 7.31943905e-01 3.04983944e-01 8.12993348e-02
3.47930193e-01 -6.71912014e-01 -5.09773642e-02 9.47142243e-02
2.20923394e-01 4.62383330e-01 7.40691647e-02 -2.00107917e-01
-4.36838329e-01 1.06625700e+00 3.05478901e-01 5.76971807e-02
7.77059913e-01 -2.32520103e-01 1.18317446e-02 9.49252248e-01
9.17827368e-01 -4.19427276e-01 -1.64043999e+00 -3.71021956e-01
1.39881447e-01 -5.02223074e-01 -2.98344284e-01 -8.12096179e-01
-5.20072639e-01 1.00537694e+00 1.06281340e+00 2.66340852e-01
1.01820278e+00 4.82400954e-02 -1.02559604e-01 6.30402923e-01
5.60162961e-01 -1.30981481e+00 8.59240353e-01 5.91645896e-01
1.16311502e+00 -1.92998791e+00 2.36470088e-01 -1.94143653e-01
-9.96338964e-01 9.87069666e-01 1.10411990e+00 1.57473147e-01
5.57644129e-01 5.70461713e-02 -3.12028438e-01 5.19980602e-02
-8.43293846e-01 -1.49020791e-01 7.21729875e-01 7.39695251e-01
1.76651970e-01 -2.75400579e-02 6.99727144e-03 2.66002595e-01
1.89590156e-01 -1.56258255e-01 1.99095249e-01 4.21138585e-01
-4.11087751e-01 -5.99960923e-01 -1.69558674e-01 3.78760189e-01
5.45893388e-04 1.39220029e-01 5.79887927e-02 9.37295437e-01
3.29093307e-01 9.19290304e-01 2.02412322e-01 -3.86667430e-01
1.90758735e-01 1.39820978e-01 7.61172116e-01 -4.63907242e-01
-1.08509049e-01 -1.34858117e-01 -1.05789781e-01 -5.90622663e-01
-7.16848969e-01 -6.19296789e-01 -9.34897006e-01 -4.69827913e-02
-3.29661340e-01 -4.00406629e-01 6.01535559e-01 1.07200229e+00
7.29702339e-02 6.16008818e-01 5.73441625e-01 -6.54998422e-01
-6.42595351e-01 -1.14712822e+00 -5.33578098e-01 5.35749316e-01
4.24832225e-01 -1.02099645e+00 -3.72836292e-01 6.33524954e-02] | [9.512909889221191, 1.7529308795928955] |
0a0d6929-4e52-48ef-8225-8e5e91f33f6e | ma-copie-adore-le-v-elo-analyse-des-besoins-r | null | null | https://aclanthology.org/2019.jeptalnrecital-court.19 | https://aclanthology.org/2019.jeptalnrecital-court.19.pdf | Ma copie adore le v\'elo : analyse des besoins r\'eels en correction orthographique sur un corpus de dict\'ees d'enfants (A corpus analysis to define the needs of dyslexic children in terms of spelling correction) | Cet article pr{\'e}sente la constitution d{'}un corpus de textes produits, sur des donn{\'e}es lors de dict{\'e}es, par des enfants paralys{\'e}s c{\'e}r{\'e}braux (PC) ou dysorthographiques, son annotation en termes d{'}erreurs orthographiques, et enfin son analyse quantitative. Cette analyse de corpus a pour objectif de d{\'e}finir des besoins r{\'e}els en mati{\`e}re de correction orthographique, et ce pour les personnes souffrant de troubles du langage {\'e}crit comme pour le grand public. Notre {\'e}tude sugg{\`e}re que les correcteurs orthographiques ne r{\'e}pondent que partiellement {\`a} ces besoins. | ['Jean-Yves Antoine', 'Marion Crochetet', 'Emmanuelle Lopez', 'Samuel Pouplin', 'Celine Arbizu'] | 2019-07-01 | null | null | null | jeptalnrecital-2019-7 | ['spelling-correction'] | ['natural-language-processing'] | [ 8.96762013e-02 3.57125163e-01 6.14037275e-01 -4.15109217e-01
-4.11473304e-01 -9.87864435e-01 7.78790951e-01 7.42401123e-01
-6.90562665e-01 8.10152590e-01 9.63205919e-02 -7.89861530e-02
-2.79148251e-01 -1.01518166e+00 -8.36066484e-01 -3.38888615e-01
1.11247420e-01 4.59555507e-01 5.61686680e-02 -7.81007409e-01
5.76705456e-01 3.82073671e-01 -1.32163882e+00 1.29124090e-01
7.29109704e-01 9.69624996e-01 3.12951803e-01 5.13304234e-01
2.06279889e-01 5.96731961e-01 -5.39021015e-01 -9.10032094e-01
4.77707177e-01 -6.39284611e-01 -3.88028175e-01 -4.09771740e-01
3.42808336e-01 3.33413899e-01 -4.82117176e-01 1.57212806e+00
8.58486295e-02 3.06288153e-01 1.77463281e+00 -5.46313703e-01
-1.05460393e+00 6.95539653e-01 -2.53932804e-01 9.17746246e-01
4.96933192e-01 -4.10081334e-02 1.07290828e+00 -9.34892952e-01
3.83006155e-01 6.87510967e-01 9.65094686e-01 2.63366789e-01
-5.91409922e-01 -6.62562549e-01 7.33758137e-02 -5.02952933e-01
-1.12962365e+00 -5.65219045e-01 1.11179143e-01 -6.36845529e-01
5.87102830e-01 3.20082933e-01 1.43807799e-01 8.47987533e-01
3.42099965e-01 6.77382171e-01 1.42697167e+00 -4.17054564e-01
1.31089017e-01 4.93312478e-01 -8.80183838e-03 8.15541923e-01
6.45456910e-02 2.51550019e-01 -2.13639244e-01 -1.08089723e-01
8.68278503e-01 1.84904084e-01 -9.64407474e-02 1.03391027e+00
-9.55332398e-01 8.98624301e-01 2.25894794e-01 5.26053607e-01
-2.92630345e-01 6.20760024e-02 4.37132180e-01 1.92884162e-01
7.11493045e-02 6.06097341e-01 -2.54590154e-01 -6.51362419e-01
-5.35592377e-01 4.66554612e-01 8.20257366e-01 1.27580655e+00
4.80686337e-01 -2.71515409e-03 5.67386746e-01 7.75043666e-01
4.16142382e-02 1.90959405e-03 1.76684633e-01 -7.49565244e-01
4.95271713e-01 4.41271931e-01 7.64657706e-02 -6.72746778e-01
-3.02985549e-01 -4.37567413e-01 -7.77396202e-01 1.49190202e-01
6.73100829e-01 1.60081603e-03 -2.29145568e-02 1.60691261e+00
-5.67436874e-01 -4.95189101e-01 -1.68309209e-03 6.06022835e-01
1.73965581e-02 2.54653811e-01 4.09285456e-01 -6.57159150e-01
1.70841873e+00 -4.78078797e-02 1.01120360e-01 -7.79593885e-02
3.81833404e-01 -1.24017107e+00 1.09438813e+00 8.43255401e-01
-1.24766827e+00 -4.00755167e-01 -8.95301282e-01 5.82227595e-02
-7.96064809e-02 5.82364023e-01 -2.32336462e-01 6.84791267e-01
-9.04930711e-01 8.70504320e-01 -3.95228446e-01 -3.72552812e-01
-1.37546465e-01 3.52606028e-01 -2.85378456e-01 4.75943595e-01
-7.27805674e-01 8.78486931e-01 7.02126145e-01 5.95321596e-01
-5.49751222e-01 8.99070203e-02 -1.04628646e+00 9.66899544e-02
8.17776084e-01 1.67997658e-01 1.35140073e+00 -8.83803725e-01
-1.02871048e+00 8.90344799e-01 -2.26284355e-01 4.36034091e-02
9.67664957e-01 2.24276870e-01 -9.94657040e-01 2.48876989e-01
1.83430001e-01 9.97068062e-02 3.66706371e-01 -1.13200939e+00
-1.12017965e+00 -3.95270526e-01 4.62207109e-01 3.19895655e-01
-4.82077301e-01 6.73218727e-01 -2.94377580e-02 -1.47126436e+00
4.98151690e-01 -1.15399718e+00 5.48177958e-01 1.22540385e-01
-7.56450474e-01 -9.24105048e-02 6.28445268e-01 -5.87678432e-01
1.12299585e+00 -1.97912920e+00 2.34192133e-01 2.39527270e-01
7.72457838e-01 5.23468852e-01 -1.95358530e-01 4.15971369e-01
-1.88902497e-01 -6.84077013e-03 -8.06355298e-01 8.15024227e-02
3.36709917e-01 -5.08723557e-02 2.32591972e-01 4.37751502e-01
-4.23261195e-01 2.80579090e-01 -7.35493064e-01 -2.72776157e-01
-3.08878452e-01 9.25324857e-02 -6.85952783e-01 -2.70751894e-01
-5.63850343e-01 4.37859386e-01 -8.31779599e-01 -2.07431972e-01
3.88967425e-01 -2.42780373e-01 2.88978130e-01 -1.45470932e-01
-8.81359160e-01 2.80802399e-01 -1.47258008e+00 1.09985888e+00
-3.54729258e-02 -2.35353075e-02 3.74957204e-01 -9.26313102e-01
1.60606480e+00 2.94846267e-01 6.05219364e-01 -4.67176527e-01
4.50069159e-01 7.23558187e-01 -2.25175008e-01 1.71377927e-01
7.89405406e-01 -2.04674259e-01 5.74171990e-02 1.40034926e+00
-7.87208438e-01 -1.56051561e-01 7.32151568e-01 -3.84619772e-01
1.11636722e+00 2.02577814e-01 1.67030673e-02 -6.97105408e-01
9.17031646e-01 -4.06664908e-01 3.64120215e-01 3.85004759e-01
-3.77256311e-02 1.60159186e-01 7.11768568e-01 -1.89573899e-01
-1.70991039e+00 -7.54886150e-01 -1.50872022e-01 1.18981934e+00
-4.43461537e-01 -4.12321687e-01 -7.36158729e-01 -5.99997997e-01
-3.45612407e-01 1.27182865e+00 -7.85323501e-01 -4.83410396e-02
-6.78471625e-01 -1.01675451e+00 1.05344665e+00 9.40286696e-01
6.73968673e-01 -8.86097312e-01 -7.72271395e-01 1.84839219e-01
-3.44284922e-01 -7.71470249e-01 -4.99252707e-01 3.22880736e-03
-3.14751983e-01 -7.99387157e-01 -8.50983918e-01 -3.05582345e-01
7.31389403e-01 -2.83488840e-01 1.13226676e+00 -8.61645117e-02
-2.99938262e-01 -3.64382751e-02 -7.67748892e-01 -7.80278504e-01
-1.18344545e+00 -1.42917484e-01 2.40960583e-01 -3.37417543e-01
5.89552283e-01 -5.40298104e-01 -1.08527720e+00 7.53069639e-01
-8.52943420e-01 -5.72918713e-01 4.85938460e-01 1.95536867e-01
4.38470542e-01 -7.21883103e-02 1.07672811e-01 -8.73911500e-01
6.90553308e-01 -8.27378556e-02 -1.74400076e-01 -2.14679748e-01
-4.54312474e-01 -1.06869735e-01 1.31298244e+00 5.04319549e-01
-9.94441509e-01 -8.06210458e-01 -4.47620809e-01 1.75986007e-01
-7.24010646e-01 4.71401870e-01 -3.23583156e-01 1.02575397e+00
1.06888962e+00 5.62948942e-01 -7.93373525e-01 -8.97813439e-01
4.41295467e-02 5.47939360e-01 9.10647094e-01 -9.29821193e-01
3.67045164e-01 6.79487064e-02 -3.69367003e-01 -6.61420763e-01
-7.94525743e-02 2.09562629e-01 -1.01443875e+00 1.53255031e-01
1.19637585e+00 -5.49836814e-01 -9.73113835e-01 8.47619176e-02
-1.07871318e+00 -3.43043029e-01 -1.61474794e-01 9.26685274e-01
-5.87687969e-01 1.02127612e-01 -9.58642304e-01 -4.81338441e-01
2.46456429e-01 -9.35003102e-01 1.07590532e+00 -8.40241536e-02
-4.54651564e-01 -5.29370189e-01 3.67361456e-01 1.56300843e-01
-4.17868704e-01 -2.84355551e-01 1.33387506e+00 -2.32244521e-01
-6.32088840e-01 -3.39274913e-01 -3.39969218e-01 3.91137779e-01
2.36646578e-01 1.06756352e-01 -5.54078937e-01 -2.45757222e-01
-5.41433930e-01 6.10372901e-01 3.11533153e-01 -1.60002276e-01
9.79480386e-01 -6.60281539e-01 2.30643734e-01 1.84939519e-01
1.51998913e+00 5.14979839e-01 5.70072532e-01 3.82238254e-02
-2.47506380e-01 4.18029130e-01 5.74158847e-01 3.12699407e-01
4.37786907e-01 7.39262521e-01 4.31407839e-02 8.99266541e-01
5.53976834e-01 -9.93459523e-02 4.76946145e-01 1.32267153e+00
-1.08603680e+00 -4.10722136e-01 -1.04937756e+00 6.26220226e-01
-1.30934715e+00 -4.71412927e-01 -6.36081338e-01 2.16747952e+00
7.70724237e-01 2.34651282e-01 3.62471104e-01 1.30420670e-01
8.72248769e-01 2.82787025e-01 1.95746645e-02 -7.69053936e-01
-3.91374141e-01 4.42356735e-01 8.85061696e-02 3.62778336e-01
-4.42622632e-01 7.00150728e-01 4.47748709e+00 6.89459622e-01
-8.90841424e-01 8.01493749e-02 4.30976719e-01 2.33702645e-01
-6.50715292e-01 -6.71193451e-02 -8.23964357e-01 8.35083961e-01
8.25632572e-01 9.35169756e-02 1.59132555e-01 4.63278472e-01
2.17606768e-01 -3.24996442e-01 -1.22720480e+00 6.48054481e-01
5.97401023e-01 -1.10364556e+00 -1.03593302e+00 -3.26405875e-02
3.90036672e-01 -5.22876754e-02 2.60954201e-01 3.99782777e-01
9.34567094e-01 -1.10213411e+00 1.12221253e+00 4.49944526e-01
1.43310153e+00 -5.52018285e-01 -3.75457138e-01 3.74170154e-01
-9.36300039e-01 9.81868207e-02 -5.62242866e-01 -1.45758718e-01
3.30373675e-01 2.41093561e-01 3.73256318e-02 9.04105604e-01
3.31950366e-01 2.53557742e-01 -7.96083331e-01 6.54079378e-01
-5.03842354e-01 7.64264524e-01 -5.10452151e-01 -3.82179290e-01
4.78140056e-01 -9.13565218e-01 7.62494743e-01 1.10982275e+00
7.18671918e-01 1.07697964e+00 -4.02440310e-01 6.10176444e-01
-1.96098760e-02 6.56517029e-01 -2.18558740e-02 -2.19405740e-01
-1.38602540e-01 1.48861796e-01 -1.47707820e+00 7.57336244e-02
-2.61377156e-01 9.15881753e-01 2.71607012e-01 -1.19063631e-01
-5.57765186e-01 -7.74866223e-01 2.45488733e-01 9.73159447e-02
-9.02572945e-02 -5.03572166e-01 -2.40811467e-01 -8.45809937e-01
-1.16350122e-01 -8.79764020e-01 1.81580529e-01 -8.09701622e-01
-8.77921641e-01 7.13403344e-01 1.63279101e-01 -7.39954174e-01
-4.38854843e-01 -6.64752960e-01 -2.31521994e-01 5.83639801e-01
-4.50084031e-01 -3.91413331e-01 5.94270647e-01 6.05014682e-01
1.65073812e-01 -5.58178008e-01 1.20648694e+00 5.81968546e-01
-6.25002682e-01 9.29942057e-02 6.53509617e-01 3.47266734e-01
6.02164686e-01 -1.36507952e+00 3.47383946e-01 6.99284613e-01
3.73889714e-01 7.13614881e-01 8.26108992e-01 -1.07031524e+00
-3.88796866e-01 -9.36303437e-01 1.39274704e+00 -5.48188269e-01
4.92868185e-01 -1.89352661e-01 -5.90002358e-01 1.74905646e+00
-1.83252804e-02 -6.46138847e-01 4.49098200e-01 5.65009415e-02
-2.28191614e-01 4.03556079e-02 -6.59417570e-01 9.55551863e-01
4.12598133e-01 -3.51214141e-01 -4.81197417e-01 4.52078879e-01
-1.84427798e-01 -2.41348192e-01 -1.15927625e+00 4.20054138e-01
1.33324727e-01 -1.60414016e+00 1.54401854e-01 -3.16501528e-01
9.01833117e-01 -3.57163399e-01 1.52555704e-01 -9.34791148e-01
4.23053741e-01 -5.72364271e-01 5.97627819e-01 9.26899016e-01
2.03685299e-01 -3.34521055e-01 4.69205946e-01 3.54380757e-01
-5.03475249e-01 4.19863731e-01 -1.12445986e+00 -1.14938110e-01
7.11966991e-01 -1.09729803e+00 4.00354207e-01 8.09939444e-01
-3.70172888e-01 -2.78783590e-02 -1.20901123e-01 -2.82279581e-01
1.58736423e-01 -1.89461097e-01 3.16546142e-01 -1.04709089e+00
-8.49865317e-01 -7.54188538e-01 -1.96760461e-01 -1.28794110e+00
-1.84005722e-02 -9.77009773e-01 -6.08832717e-01 -1.10823119e+00
-2.24873051e-02 -4.93363202e-01 -9.69579890e-02 3.26046050e-01
1.73687905e-01 7.70456851e-01 6.39255822e-01 3.33377838e-01
-5.44794023e-01 3.57395589e-01 1.39441228e+00 5.83810031e-01
2.72086084e-01 4.95018251e-02 -8.28462481e-01 1.30234599e+00
3.40189636e-01 -5.65689266e-01 -1.61817536e-01 -5.61864972e-01
1.23150861e+00 1.02554470e-01 2.71869928e-01 -6.82161808e-01
2.41425276e-01 -1.28635183e-01 8.45295846e-01 -6.02895439e-01
7.43427873e-01 -2.57519275e-01 1.05594441e-01 -6.84836581e-02
-2.72287071e-01 5.52954018e-01 1.25717252e-01 1.48849338e-01
-1.95640642e-02 -4.84144717e-01 9.37076092e-01 -6.84254169e-01
-6.10268414e-01 5.91409206e-01 -6.53106391e-01 4.13854450e-01
8.16721499e-01 -4.77799982e-01 -3.65317464e-02 -4.64666449e-02
-9.88887727e-01 -8.22648033e-02 4.45966214e-01 -1.09841466e-01
6.73954040e-02 -7.77536929e-01 -1.31137943e+00 2.86699057e-01
-3.48505192e-02 -1.18061565e-01 9.84682262e-01 7.28946269e-01
-1.34353924e+00 1.49942473e-01 -3.82207423e-01 -4.43393260e-01
-1.19368947e+00 3.09428543e-01 -2.09738389e-02 4.41868931e-01
-1.11112428e+00 1.22940481e+00 -1.86542436e-01 -3.57622564e-01
-5.29958189e-01 8.35740194e-02 4.83029895e-02 1.25094224e-03
2.51193484e-03 9.58970368e-01 -1.07758060e-01 -1.01728988e+00
8.67574513e-02 5.54024279e-01 -1.83287710e-01 -5.67124844e-01
8.78317237e-01 -4.16856915e-01 -9.07890439e-01 2.26554379e-01
9.30544019e-01 6.64626241e-01 -1.35923648e+00 7.58567929e-01
-4.61238176e-01 -2.87469745e-01 -9.66685832e-01 -9.70096111e-01
-5.57075679e-01 7.89121091e-01 2.41802812e-01 -1.86396062e-01
4.48971570e-01 -1.88906863e-01 3.91658515e-01 5.52562118e-01
5.65628171e-01 -1.54116285e+00 -3.48143399e-01 7.81596720e-01
9.73116577e-01 -5.36140263e-01 -1.17089234e-01 -2.58513749e-01
-5.36167204e-01 1.27392554e+00 -4.22652006e-01 -1.75161347e-01
6.92207336e-01 -5.16720526e-02 1.80706888e-01 -3.51429015e-01
9.61522833e-02 -1.01260789e-01 1.09820262e-01 1.19290106e-01
8.17251086e-01 -3.34878266e-02 -1.13216579e+00 1.00395703e+00
-1.19489467e+00 -2.71730553e-02 6.99227095e-01 8.39878559e-01
-8.21967497e-02 -1.06570578e+00 -5.99781811e-01 9.85523015e-02
-4.73495394e-01 -1.22339793e-01 -4.40368801e-01 8.59635651e-01
5.64288557e-01 8.00391614e-01 1.10633522e-01 -6.73983432e-03
8.19409847e-01 6.08054698e-02 4.27277297e-01 -6.49613738e-01
-1.25881219e+00 5.79252362e-01 -1.20608155e-02 1.02721430e-01
-4.29629982e-01 -9.12550032e-01 -7.27485776e-01 -9.30725515e-01
1.33643061e-01 4.41815466e-01 3.20095628e-01 1.33498287e+00
-4.21892613e-01 5.25489271e-01 -8.43685791e-02 -4.58979517e-01
-7.98647523e-01 -6.98551178e-01 -1.51464403e+00 3.46402675e-01
-3.33750039e-01 -3.93948376e-01 -6.11908920e-02 -1.40199929e-01] | [14.102850914001465, 13.317082405090332] |
7eb156c9-4a15-409c-995d-1d6739e6f493 | using-multiple-losses-for-accurate-facial-age | 2106.09393 | null | https://arxiv.org/abs/2106.09393v1 | https://arxiv.org/pdf/2106.09393v1.pdf | using multiple losses for accurate facial age estimation | Age estimation is an essential challenge in computer vision. With the advances of convolutional neural networks, the performance of age estimation has been dramatically improved. Existing approaches usually treat age estimation as a classification problem. However, the age labels are ambiguous, thus make the classification task difficult. In this paper, we propose a simple yet effective approach for age estimation, which improves the performance compared to classification-based methods. The method combines four classification losses and one regression loss representing different class granularities together, and we name it as Age-Granularity-Net. We validate the Age-Granularity-Net framework on the CVPR Chalearn 2016 dataset, and extensive experiments show that the proposed approach can reduce the prediction error compared to any individual loss. The source code link is https://github.com/yipersevere/age-estimation. | ['Tapio Elomaa', 'Heikki Huttunen', 'Yi Zhou'] | 2021-06-17 | null | null | null | null | ['age-estimation', 'age-estimation'] | ['computer-vision', 'miscellaneous'] | [-2.06121802e-01 -6.60129860e-02 -3.53484124e-01 -7.19525218e-01
-4.96197551e-01 -3.77433514e-03 5.98371029e-01 5.25343895e-01
-6.20470047e-01 8.12810481e-01 7.01511651e-02 -1.53371328e-02
-7.84626417e-03 -8.71225417e-01 -4.06417638e-01 -7.15656459e-01
4.49840091e-02 3.34968686e-01 -9.47021991e-02 3.54339898e-01
2.13757142e-01 -4.43410613e-02 -1.65993023e+00 -2.17654333e-01
1.44738817e+00 1.48547769e+00 -2.38942295e-01 2.11937353e-01
-1.21760458e-01 6.73208117e-01 -5.57738841e-01 -8.07343245e-01
2.59380843e-02 -2.83606853e-02 -5.67286611e-01 -4.54566836e-01
5.49178600e-01 -5.83443940e-01 -3.31191093e-01 1.00472498e+00
6.01969898e-01 -1.91781193e-01 9.60592151e-01 -1.59233594e+00
-6.73923790e-01 6.98923349e-01 -8.77794921e-01 1.51036575e-01
-2.70152148e-02 -3.85520726e-01 8.60306144e-01 -7.28444517e-01
5.23988418e-02 1.25829244e+00 1.01262498e+00 7.45971620e-01
-8.06583464e-01 -1.16696680e+00 2.48440638e-01 4.97253239e-01
-1.45125115e+00 -1.71110615e-01 5.28933167e-01 -5.52959204e-01
2.99077451e-01 -9.14356634e-02 6.72179759e-01 1.20016873e+00
5.41763157e-02 1.12429798e+00 1.22780681e+00 -1.26448482e-01
7.07927570e-02 -3.06595117e-01 4.62858409e-01 6.86982691e-01
5.29434144e-01 2.02312112e-01 -4.76565570e-01 9.65426862e-03
4.37779635e-01 3.79266620e-01 5.13890721e-02 -1.88837901e-01
-8.97195160e-01 7.22344935e-01 5.99558890e-01 -1.48818165e-01
-1.86997920e-01 3.48146319e-01 4.74995047e-01 6.14401586e-02
1.07878411e+00 1.27427042e-01 -5.13860166e-01 -2.14746073e-01
-9.29786325e-01 5.18718362e-01 4.59760964e-01 5.92041254e-01
4.35010254e-01 -2.30292961e-01 -2.27423936e-01 1.12342334e+00
1.52754113e-01 3.21777731e-01 3.54003161e-01 -9.08561885e-01
3.31760198e-01 6.48389935e-01 -1.25928655e-01 -6.98038101e-01
-5.53221166e-01 -4.79231894e-01 -1.10295856e+00 1.88029781e-01
7.13313460e-01 -2.04090565e-01 -1.08509600e+00 1.91393363e+00
4.70314741e-01 2.96042234e-01 -3.60989869e-01 4.66717958e-01
9.20603573e-01 3.64965737e-01 4.72120851e-01 -2.22046852e-01
1.32816207e+00 -8.98415148e-01 -6.81428313e-01 -1.24307275e-01
5.20851910e-01 -4.27580595e-01 5.22224486e-01 7.01505423e-01
-9.33844447e-01 -4.01648045e-01 -1.06389296e+00 4.40794788e-02
-2.95398921e-01 2.53495544e-01 9.06143725e-01 4.25211847e-01
-7.29722142e-01 8.24088156e-01 -7.89728284e-01 -2.50807285e-01
9.33795869e-01 2.95665652e-01 -3.10844034e-01 -7.08177164e-02
-1.24854434e+00 8.07912946e-01 5.98378897e-01 4.11966518e-02
-5.56270361e-01 -1.02720726e+00 -8.59855294e-01 -6.92066103e-02
2.12885037e-01 -7.88298488e-01 1.47923565e+00 -8.53415728e-01
-8.91824841e-01 9.55459416e-01 -4.00817767e-02 -5.76425076e-01
8.76572371e-01 -7.38897145e-01 -1.93607837e-01 -2.35485345e-01
7.03931414e-03 7.48914063e-01 9.18843567e-01 -9.44736540e-01
-1.08710325e+00 -6.38371170e-01 -1.02071062e-01 -1.36643089e-02
-6.95828557e-01 1.84954509e-01 -3.11233759e-01 -7.47392893e-01
-1.04163036e-01 -6.26173019e-01 -6.34832829e-02 2.48150110e-01
-4.72650193e-02 -7.41569281e-01 5.36796808e-01 -1.09516656e+00
1.68252277e+00 -1.93157244e+00 4.28457744e-02 -2.39637956e-01
5.47173798e-01 2.26319805e-01 3.19197059e-01 -5.23396023e-02
-1.23381346e-01 6.33771494e-02 -3.20360392e-01 -7.05017567e-01
-1.50014460e-01 -1.31374642e-01 1.15034744e-01 2.92027175e-01
4.10905108e-02 6.60303831e-01 -9.18137848e-01 -7.73055315e-01
-1.94417045e-03 5.34034789e-01 -2.65935808e-01 2.85267532e-01
8.10646117e-02 2.13348120e-01 -3.96945149e-01 9.08588350e-01
7.99339056e-01 5.91623560e-02 -2.11720601e-01 -3.34266722e-02
1.38739854e-01 -9.98697728e-02 -7.71489441e-01 1.40290332e+00
-3.95667762e-01 2.79638529e-01 -3.40322882e-01 -9.89940464e-01
1.15871763e+00 1.17803067e-02 5.37849486e-01 -6.06185257e-01
3.08635324e-01 2.53458202e-01 -8.99299458e-02 -1.17701732e-01
2.29083121e-01 -1.97915621e-02 -2.45119944e-01 7.24145174e-02
2.62984019e-02 2.98886895e-01 2.43476719e-01 5.40091060e-02
9.38953161e-01 2.31480375e-01 4.04248893e-01 5.01401536e-02
4.84066099e-01 -6.06822908e-01 1.05677104e+00 5.38844645e-01
-6.11022055e-01 6.58147693e-01 6.03169978e-01 -6.60716593e-01
-9.77424622e-01 -1.34142554e+00 -4.26510215e-01 9.18034136e-01
-7.01798499e-02 -5.50897002e-01 -8.72684121e-01 -1.09579384e+00
4.37231451e-01 3.24370086e-01 -8.84364665e-01 -3.18668395e-01
-4.43433285e-01 -9.90004778e-01 4.56469327e-01 8.70620370e-01
6.82765245e-01 -9.87495482e-01 -1.74470440e-01 -9.11150724e-02
-2.37180397e-01 -8.61045897e-01 -2.08770990e-01 -2.44329587e-01
-1.01335859e+00 -9.52569544e-01 -1.15420640e+00 -5.50396323e-01
8.63944113e-01 -3.84429932e-01 1.24456251e+00 3.47511262e-01
-3.94304752e-01 1.00758240e-01 -4.35318619e-01 -8.50536168e-01
-3.64541970e-02 4.79329437e-01 3.06430638e-01 -6.98689744e-02
5.41265905e-01 -6.48607850e-01 -9.90186453e-01 7.90108647e-03
-3.65260512e-01 1.06394216e-01 5.64561427e-01 7.96472967e-01
3.61642957e-01 -2.45499667e-02 9.52020764e-01 -7.80528247e-01
4.10201758e-01 -5.07688105e-01 -5.36707044e-01 1.55937657e-01
-1.11722553e+00 -7.75242969e-02 3.18035573e-01 -4.20217127e-01
-1.01701856e+00 -6.03592545e-02 -3.14647108e-01 -2.68207461e-01
-1.06802784e-01 3.52013111e-01 -1.64546296e-01 2.62370110e-01
2.23501161e-01 -2.80414950e-02 5.26240468e-02 -6.88375413e-01
8.81467760e-02 8.02878082e-01 4.46990937e-01 -7.24624693e-01
6.40943229e-01 7.13067651e-02 -4.94645862e-03 -3.83858621e-01
-1.34962547e+00 -1.69837758e-01 -4.58723992e-01 -3.98898065e-01
6.76292300e-01 -1.03826094e+00 -7.66580105e-01 1.32003772e+00
-8.46699297e-01 -2.71165818e-01 9.45130736e-02 4.07282710e-01
-4.26740080e-01 2.94420123e-01 -7.04066873e-01 -8.90081167e-01
-8.03352416e-01 -7.00708449e-01 8.47973824e-01 6.33532286e-01
-2.54505515e-01 -8.56478155e-01 -2.48943597e-01 4.94748950e-01
1.56590641e-01 4.44710255e-01 6.53186262e-01 -4.72095668e-01
-1.10600464e-01 -3.76788914e-01 -5.49677014e-01 5.54251850e-01
2.78802980e-02 1.74466163e-01 -8.69953930e-01 -2.02342704e-01
-5.45683503e-01 -4.26376820e-01 1.28101921e+00 6.27692103e-01
1.85121226e+00 -1.30161876e-02 -3.34526449e-01 6.97470725e-01
1.19757152e+00 1.76316556e-02 6.77058756e-01 4.20923769e-01
8.82422626e-01 6.63878262e-01 1.02995038e+00 6.29462063e-01
6.77731097e-01 5.20811141e-01 4.70703125e-01 4.00414243e-02
-1.14938878e-02 -2.36002937e-01 -7.38531724e-02 6.45091891e-01
-4.04620677e-01 1.14034899e-01 -1.09775579e+00 7.00165927e-01
-2.06414342e+00 -6.93026781e-01 7.89628997e-02 2.19537139e+00
1.03376818e+00 3.32916260e-01 3.76331359e-01 2.79542327e-01
8.33001971e-01 2.41016522e-01 -6.33381784e-01 -4.22050864e-01
3.13275397e-01 1.28603444e-01 4.66598600e-01 1.50479734e-01
-1.38347661e+00 6.06477499e-01 5.67636061e+00 1.08484721e+00
-7.92455912e-01 7.92069212e-02 1.12651074e+00 -5.94284050e-02
2.61799663e-01 -3.73952866e-01 -9.99855280e-01 9.51124310e-01
6.30344748e-01 -3.87909323e-01 4.12912220e-02 8.75562251e-01
-1.61704987e-01 -1.77248821e-01 -1.21764195e+00 1.12708366e+00
9.99273267e-04 -8.42021048e-01 -2.34877214e-01 1.99823510e-02
3.96757662e-01 -3.87225211e-01 3.65820497e-01 4.64770019e-01
2.28054136e-01 -9.96675193e-01 6.23787880e-01 6.48421228e-01
8.16338360e-01 -1.15385616e+00 9.46540594e-01 2.28419811e-01
-1.14853525e+00 -3.58447552e-01 -1.97727755e-01 -3.16020608e-01
-2.34151501e-02 9.78107870e-01 -3.88647646e-01 5.44042528e-01
1.26352215e+00 1.00862384e+00 -8.41433585e-01 1.36296988e+00
-3.64639074e-01 6.52396202e-01 -5.59160262e-02 1.08538963e-01
-2.92462319e-01 -1.82246804e-01 3.88084240e-02 7.50475943e-01
7.02649355e-01 -6.22271262e-02 -1.07009858e-01 4.94233519e-01
-3.17222714e-01 -8.95737670e-03 -2.23455697e-01 4.57880087e-03
5.51746964e-01 1.28923798e+00 -4.88304049e-01 -2.18166485e-01
-2.76722729e-01 7.30915844e-01 7.20182359e-01 -8.32589567e-02
-8.24965119e-01 -3.37581664e-01 7.18453586e-01 -2.86612734e-02
9.30802897e-02 1.48043828e-02 -3.98522794e-01 -9.45457101e-01
8.10486898e-02 -5.19769609e-01 7.02024877e-01 -5.68332136e-01
-1.61035693e+00 5.93428724e-02 6.04066290e-02 -1.14149845e+00
-3.99541147e-02 -5.25508583e-01 -5.66962123e-01 5.50574183e-01
-1.30886447e+00 -1.34160602e+00 -4.82227147e-01 7.34151006e-02
4.47878450e-01 -7.33071789e-02 5.12420475e-01 5.58470726e-01
-8.25983286e-01 1.04951918e+00 1.67633772e-01 3.44232887e-01
9.70361471e-01 -1.47185779e+00 2.80255020e-01 6.40751302e-01
-4.95502561e-01 3.03275526e-01 6.32426322e-01 -6.21768773e-01
-4.64148015e-01 -1.12808335e+00 1.10764956e+00 -4.54870194e-01
5.09897828e-01 -4.33409274e-01 -9.10507023e-01 4.58456427e-01
-1.00930847e-01 -7.32356980e-02 5.74335098e-01 5.57828784e-01
-6.23041213e-01 -5.19200623e-01 -1.21693134e+00 3.97274554e-01
1.18191648e+00 4.31600437e-02 -5.60657561e-01 -3.96716706e-02
7.60581613e-01 -3.62458527e-01 -1.17489862e+00 8.41597080e-01
9.79993045e-01 -8.93569052e-01 1.13537920e+00 -2.81884611e-01
9.64956403e-01 8.14447105e-02 3.45697403e-01 -1.24898374e+00
-1.72128633e-01 1.23733431e-01 -7.49754786e-01 1.73051679e+00
2.31269345e-01 -7.97455907e-01 8.31313550e-01 5.23135066e-01
9.77414176e-02 -1.24849677e+00 -7.86895394e-01 -8.63352060e-01
4.66547489e-01 -1.75204471e-01 7.52802908e-01 8.80407453e-01
-3.60552281e-01 1.20227493e-01 -5.11301875e-01 -8.76321867e-02
9.90824163e-01 -2.04252735e-01 4.46176797e-01 -1.77362454e+00
-1.85423587e-02 -6.43802881e-01 -5.63463092e-01 -3.44507426e-01
2.70834357e-01 -6.11924887e-01 -1.14570089e-01 -1.64642143e+00
4.14500803e-01 -6.32261515e-01 -7.67941475e-01 5.61224341e-01
-7.16510057e-01 1.84901685e-01 1.23200230e-01 -6.09511323e-02
-6.43493295e-01 6.93542421e-01 8.75188231e-01 -2.05410495e-01
2.57631063e-01 3.09349597e-01 -7.22552657e-01 8.30755353e-01
1.18449795e+00 -4.40718710e-01 -2.58942187e-01 -1.67153269e-01
2.15305269e-01 -3.08077157e-01 1.22273974e-01 -1.23078859e+00
-1.38349444e-01 5.42604923e-02 8.09816301e-01 -6.60504282e-01
2.63330847e-01 -4.92564976e-01 -4.09512036e-02 6.86665595e-01
-1.59713924e-01 -1.01815976e-01 -4.86433506e-02 3.52809787e-01
-1.91331878e-01 -9.00202617e-02 7.63375640e-01 2.00445533e-01
-7.59155989e-01 9.57561433e-01 1.89105093e-01 -2.18644701e-02
1.04303181e+00 -1.31558655e-02 -4.83961850e-01 -1.72956556e-01
-6.38102472e-01 6.58980370e-01 4.80176628e-01 5.49868405e-01
5.60358346e-01 -1.62321043e+00 -8.73738408e-01 -2.24284679e-01
4.56148893e-01 9.29277688e-02 4.44647312e-01 8.38508844e-01
-3.58358651e-01 -1.00678921e-01 -2.48401880e-01 -5.27978659e-01
-1.60693181e+00 6.59057021e-01 2.66372114e-01 -4.37186569e-01
-3.70537102e-01 1.11902511e+00 4.07377541e-01 -5.21211386e-01
5.48994899e-01 -2.18710236e-04 -5.55275559e-01 3.85021180e-01
7.01870561e-01 6.01773620e-01 -2.08994716e-01 -1.97760999e-01
-3.12656611e-01 6.20557487e-01 -4.44927514e-01 3.10151160e-01
1.28182173e+00 -1.07993878e-01 -1.70781180e-01 5.32838941e-01
9.42879856e-01 -5.73992491e-01 -1.13986886e+00 -3.43151242e-01
5.37201017e-02 -4.72426474e-01 -1.13681085e-01 -7.68820584e-01
-1.20983732e+00 8.79717469e-01 8.79222274e-01 2.57590711e-02
1.34552348e+00 6.66583106e-02 9.93931413e-01 -4.12153192e-02
3.04104835e-01 -1.28476572e+00 1.62931345e-02 3.68116289e-01
6.59458399e-01 -1.54432929e+00 2.88728803e-01 -3.79304320e-01
-2.45134488e-01 7.26848960e-01 1.12835646e+00 7.15677962e-02
7.22489655e-01 -6.02981262e-02 4.08474468e-02 2.51101196e-01
-5.43758869e-01 -1.17048286e-01 3.12899888e-01 6.42692447e-01
4.58352774e-01 2.54663914e-01 -8.91691983e-01 9.35215592e-01
-2.49228016e-01 2.30668336e-01 1.25377893e-01 5.90455532e-01
-2.98241496e-01 -1.34866142e+00 -2.42087215e-01 1.06316388e+00
-7.94238269e-01 2.52724886e-02 -1.48599580e-01 5.58813095e-01
4.50274169e-01 6.35841131e-01 3.14434826e-01 -4.69139099e-01
3.44212935e-03 4.87964377e-02 5.68177879e-01 -2.53192246e-01
-2.84457386e-01 -4.74554628e-01 2.85370320e-01 -3.79540294e-01
-2.72896558e-01 -8.38292956e-01 -1.11516798e+00 -4.86051440e-01
-8.82940814e-02 -1.58384591e-01 6.30908370e-01 7.92770743e-01
3.80752757e-02 5.47876954e-01 6.91538811e-01 -5.08955956e-01
-5.21291494e-01 -9.92566645e-01 -6.09595537e-01 3.57594162e-01
2.24977687e-01 -1.12094545e+00 -3.81859541e-01 -1.76036060e-01] | [13.627915382385254, 0.8772023916244507] |
9ae6d909-7da7-40d4-ba4c-94205c1f21eb | pix2vex-image-to-geometry-reconstruction | 1903.11149 | null | https://arxiv.org/abs/1903.11149v2 | https://arxiv.org/pdf/1903.11149v2.pdf | Pix2Vex: Image-to-Geometry Reconstruction using a Smooth Differentiable Renderer | The long-coveted task of reconstructing 3D geometry from images is still a standing problem. In this paper, we build on the power of neural networks and introduce Pix2Vex, a network trained to convert camera-captured images into 3D geometry. We present a novel differentiable renderer ($DR$) as a forward validation means during training. Our key insight is that $DR$s produce images of a particular appearance, different from typical input images. Hence, we propose adding an image-to-image translation component, converting between these rendering styles. This translation closes the training loop, while allowing to use minimal supervision only, without needing any 3D model as ground truth. Unlike state-of-the-art methods, our $DR$ is $C^\infty$ smooth and thus does not display any discontinuities at occlusions or dis-occlusions. Through our novel training scheme, our network can train on different types of images, where previous work can typically only train on images of a similar appearance to those rendered by a $DR$. | ['Daniel Cohen-Or', 'Oliver Deussen', 'Amit H. Bermano', 'Felix Petersen'] | 2019-03-26 | null | null | null | null | ['3d-object-reconstruction'] | ['computer-vision'] | [ 3.17317635e-01 4.08566564e-01 2.52431184e-01 -4.77247894e-01
-3.81009430e-01 -6.67449355e-01 5.15304029e-01 -4.37913448e-01
-1.84203699e-01 4.04590517e-01 -4.65131044e-01 -5.22362769e-01
2.76305139e-01 -9.93276834e-01 -1.36981535e+00 -3.19000304e-01
8.83634761e-02 3.42375576e-01 1.86510593e-01 -2.15480343e-01
1.24412604e-01 7.39980996e-01 -1.50716734e+00 2.85190314e-01
6.59498394e-01 1.10310662e+00 -2.71577854e-02 7.69355953e-01
-1.93433419e-01 5.90995848e-01 -4.80795205e-01 -6.63155556e-01
7.06292212e-01 -2.79799044e-01 -7.38104820e-01 3.52285981e-01
1.07162642e+00 -6.56128943e-01 -2.62266070e-01 9.90044475e-01
1.88080162e-01 -1.19693190e-01 6.32248223e-01 -9.92676198e-01
-7.95008183e-01 4.90251072e-02 -4.81736034e-01 -3.15639734e-01
4.54572797e-01 4.07878570e-02 5.87488949e-01 -9.65917587e-01
9.39358473e-01 1.23615670e+00 9.29532766e-01 6.25539720e-01
-1.49046528e+00 -4.87016231e-01 3.01701397e-01 -2.92349428e-01
-1.32304657e+00 -3.66941899e-01 1.07075715e+00 -5.74351907e-01
8.38938832e-01 3.04896712e-01 6.76167607e-01 1.02278256e+00
-7.77153745e-02 7.18281388e-01 1.23330677e+00 -5.75677395e-01
1.24929026e-01 3.17484647e-01 -3.35761994e-01 9.32755411e-01
-1.31538466e-01 2.49561876e-01 -1.50397807e-01 2.80092359e-01
1.45568037e+00 -2.97613274e-02 -4.84824419e-01 -7.13532746e-01
-9.67843294e-01 5.57925344e-01 6.47814989e-01 1.07925981e-02
-1.30659097e-03 3.87460083e-01 3.73401567e-02 4.87392366e-01
5.44107974e-01 3.78360838e-01 -3.39427412e-01 1.09562501e-01
-8.13214004e-01 2.30685383e-01 5.94893157e-01 1.13158190e+00
1.06478286e+00 3.65325361e-01 3.09805602e-01 6.49953306e-01
3.47632319e-02 4.76199120e-01 8.15066695e-02 -1.32466674e+00
2.69782782e-01 7.41379082e-01 1.35768488e-01 -1.03604984e+00
-1.17000677e-01 -3.19579601e-01 -8.84514153e-01 8.70471895e-01
3.92278016e-01 6.10410376e-03 -1.01278865e+00 1.54515326e+00
3.25062066e-01 2.91355001e-03 -1.70639932e-01 7.09454596e-01
7.05927968e-01 4.41183209e-01 -3.99861336e-01 1.83605179e-01
9.31394994e-01 -8.45954597e-01 -3.21981639e-01 -1.78418010e-01
3.74111861e-01 -6.90742671e-01 1.39835036e+00 5.37047088e-01
-1.43005967e+00 -7.69223511e-01 -1.06448710e+00 -3.34131777e-01
-4.91175443e-01 5.65178096e-02 4.99601394e-01 6.13489091e-01
-1.41283596e+00 9.04008090e-01 -7.07209587e-01 -2.14685932e-01
3.94171357e-01 2.24025533e-01 -5.50747335e-01 -4.48081270e-03
-7.53077328e-01 9.87264633e-01 7.85981268e-02 1.02419533e-01
-6.52334094e-01 -7.38606572e-01 -8.90280306e-01 -3.41933757e-01
2.72307456e-01 -9.06410158e-01 1.08407307e+00 -1.31313300e+00
-1.91285288e+00 1.42073774e+00 -2.00822409e-02 -1.66663587e-01
9.62962687e-01 -2.31828377e-01 -1.47331193e-01 1.68401107e-01
-2.13800117e-01 8.10158908e-01 1.13668585e+00 -1.81862259e+00
-2.86982596e-01 -1.87402785e-01 5.02047122e-01 1.46145880e-01
1.08028583e-01 -2.22307235e-01 -6.43131614e-01 -6.81083322e-01
1.67880207e-01 -8.02039802e-01 -1.61236465e-01 5.08065403e-01
-5.64031482e-01 1.71137691e-01 8.13208878e-01 -4.72004831e-01
5.90433896e-01 -2.11993861e+00 2.02450052e-01 2.02928632e-01
2.60534585e-01 3.31106305e-01 -1.66619554e-01 3.32868040e-01
-3.43432099e-01 2.40517616e-01 -5.17841637e-01 -7.46873140e-01
2.56437063e-02 2.15385079e-01 -3.26379269e-01 2.99925297e-01
4.89937127e-01 7.76252687e-01 -6.27190411e-01 -2.19747663e-01
5.17902017e-01 7.19915628e-01 -7.77856052e-01 4.96706039e-01
-4.39684451e-01 6.95507526e-01 -1.27529576e-01 4.45016295e-01
8.65072012e-01 -2.57145643e-01 -6.85077086e-02 -2.42659897e-01
-2.37490416e-01 -7.79030398e-02 -1.18920779e+00 2.13267446e+00
-6.19567990e-01 6.36711240e-01 3.01815748e-01 -1.22793901e+00
1.07619095e+00 1.96566302e-02 2.18753770e-01 -6.56780243e-01
2.85273403e-01 2.00124592e-01 -3.87293547e-01 -4.58195895e-01
3.92809927e-01 -1.64005503e-01 1.52162850e-01 4.29827690e-01
-9.48689282e-02 -7.23096251e-01 -1.26574606e-01 1.36749735e-02
7.84241915e-01 6.78570092e-01 -1.42701432e-01 -1.80924460e-01
4.57045943e-01 -9.23803002e-02 2.82373697e-01 4.92144138e-01
2.07720339e-01 1.12010789e+00 4.37853426e-01 -7.54221380e-01
-1.19408345e+00 -1.15518677e+00 -2.21841559e-01 5.81054091e-01
2.00368941e-01 -4.16638628e-02 -9.70607698e-01 -5.90718448e-01
-8.13173056e-02 5.51198065e-01 -7.79563785e-01 9.92185920e-02
-8.85377109e-01 -2.68976450e-01 4.18136626e-01 3.00941914e-01
5.64516902e-01 -9.92825866e-01 -7.77323723e-01 -1.99039467e-02
2.37719059e-01 -1.20133340e+00 -2.76239365e-01 1.04250811e-01
-8.85398388e-01 -1.11415172e+00 -8.50131631e-01 -8.50625277e-01
1.00772452e+00 1.09505482e-01 1.47523654e+00 1.60555780e-01
-2.72719234e-01 3.67009491e-01 -1.43074125e-01 -2.59079814e-01
-6.18999898e-01 -1.61610112e-01 -2.20093638e-01 1.06751710e-01
-1.89970613e-01 -9.89776492e-01 -7.64729500e-01 2.36149281e-01
-1.10533261e+00 4.25279588e-01 4.31657195e-01 6.05818808e-01
7.26060569e-01 -2.14263842e-01 2.65841242e-02 -1.10145438e+00
2.53489524e-01 1.08228885e-01 -7.53255188e-01 3.28298248e-02
-4.01891619e-01 9.70428586e-02 9.46962953e-01 -3.23692739e-01
-8.41295063e-01 1.69300064e-01 -3.42497468e-01 -8.89445126e-01
-4.36307460e-01 1.23091996e-01 -1.76259294e-01 -2.22125232e-01
7.18721330e-01 6.23678863e-02 1.35652050e-01 -6.93951905e-01
6.83642507e-01 2.74635375e-01 7.85970211e-01 -5.55342197e-01
1.00158226e+00 7.70271838e-01 1.62836704e-02 -6.85936034e-01
-6.15224719e-01 1.06215224e-01 -8.36217344e-01 -1.10693581e-01
7.76192665e-01 -7.09330559e-01 -6.61786377e-01 5.10738015e-01
-1.32620859e+00 -7.26422906e-01 -6.14116132e-01 7.23447744e-03
-7.84419119e-01 1.93496332e-01 -5.73908448e-01 -4.63925987e-01
-1.15888089e-01 -1.21107888e+00 1.17868602e+00 -1.30933551e-02
7.32014254e-02 -9.36664999e-01 1.39031047e-03 8.20662528e-02
4.52661216e-01 7.13467121e-01 9.68019247e-01 1.57592595e-01
-8.15479934e-01 -2.22068861e-01 -5.06627798e-01 6.76847696e-01
1.30649552e-01 1.96819708e-01 -1.25556636e+00 -2.32052624e-01
5.22617586e-02 -3.17089230e-01 6.10168576e-01 2.33378068e-01
1.47517550e+00 -2.09017143e-01 -3.47990803e-02 1.21105230e+00
1.52732456e+00 2.19604507e-01 7.68811107e-01 3.05683881e-01
9.32380736e-01 5.29592872e-01 -5.71614765e-02 -1.08235246e-02
4.15171802e-01 7.48537898e-01 7.11270452e-01 -6.70544386e-01
-3.21984202e-01 -3.97740692e-01 2.38165081e-01 5.78166425e-01
-3.12171519e-01 -4.32016961e-02 -7.49627292e-01 1.68713316e-01
-1.47352099e+00 -6.58421278e-01 -1.38255507e-01 2.23774576e+00
7.58563578e-01 1.36949390e-01 -1.22630864e-01 1.34955660e-01
4.38087106e-01 1.89961612e-01 -6.12677693e-01 -6.77233040e-01
-4.94192205e-02 5.30528963e-01 4.10873622e-01 7.22378850e-01
-1.07423604e+00 9.22357559e-01 6.75978184e+00 5.32370329e-01
-1.54789221e+00 -2.41654009e-01 7.43872762e-01 3.20825875e-02
-5.33410788e-01 -6.25363216e-02 -4.11436707e-01 1.57258570e-01
4.00729388e-01 3.09760809e-01 6.14949644e-01 8.63651752e-01
9.28188339e-02 7.97158256e-02 -1.39550173e+00 1.11488962e+00
2.87731051e-01 -1.47596073e+00 1.49654791e-01 -9.85669643e-02
6.47450030e-01 -1.42159179e-01 2.11749405e-01 1.37451589e-01
5.72018862e-01 -1.18780017e+00 1.02438557e+00 4.93066579e-01
1.26732123e+00 -3.95468801e-01 2.09375516e-01 3.90174955e-01
-8.84479880e-01 3.95552754e-01 -2.43349671e-01 -4.84953150e-02
3.58195782e-01 5.42479873e-01 -4.24791217e-01 6.42492592e-01
8.98043871e-01 7.91412771e-01 -6.13752127e-01 5.06835639e-01
-1.99452892e-01 1.44804344e-02 -3.63357157e-01 5.65128148e-01
5.68247475e-02 -4.47687358e-01 2.94348955e-01 9.81303632e-01
4.46203917e-01 7.28280395e-02 9.59741771e-02 1.18810344e+00
-2.12179497e-01 -5.92527725e-02 -9.37886238e-01 3.76842201e-01
-1.33795710e-03 8.97608578e-01 -6.48350894e-01 -4.25059199e-01
-3.41163397e-01 1.33908725e+00 5.21639109e-01 4.22655880e-01
-8.06707442e-01 -2.60206252e-01 3.88634175e-01 3.92848849e-01
4.26372468e-01 -3.03753316e-01 -1.79611206e-01 -1.36899972e+00
2.97544926e-01 -7.64421999e-01 -1.16559833e-01 -1.09632146e+00
-1.15080357e+00 9.93313611e-01 -1.90348849e-02 -1.47060204e+00
-1.92273855e-01 -8.81295919e-01 -5.89436531e-01 9.22114372e-01
-1.76206172e+00 -1.20543385e+00 -4.40266401e-01 7.02775896e-01
3.52059215e-01 2.47703120e-01 8.65398228e-01 3.11830699e-01
-4.26427573e-01 5.25497139e-01 -4.14504819e-02 2.22914711e-01
6.02450013e-01 -1.24745309e+00 5.72108209e-01 7.73658395e-01
8.32367018e-02 5.35079122e-01 5.53441286e-01 -1.61903858e-01
-1.41929388e+00 -1.10955417e+00 4.29373831e-01 -5.61383307e-01
1.93869308e-01 -4.85466838e-01 -9.26453888e-01 8.54384065e-01
2.92722464e-01 3.16011310e-01 2.49549255e-01 -1.84870720e-01
-5.94195485e-01 -8.40048790e-02 -1.19239163e+00 6.59138083e-01
1.38014209e+00 -6.26212120e-01 -3.12266827e-01 1.01249158e-01
7.04204440e-01 -8.42898965e-01 -8.03826392e-01 3.85662794e-01
6.05560362e-01 -1.47549450e+00 1.35480464e+00 -3.67674172e-01
6.65040076e-01 -2.89234787e-01 -1.62216499e-01 -1.32866001e+00
2.08494756e-02 -4.99994099e-01 1.03433043e-01 9.30250406e-01
3.12717378e-01 -5.73451340e-01 9.97619927e-01 6.79863632e-01
-4.19225603e-01 -6.57212615e-01 -7.03164637e-01 -8.47569108e-01
4.14012879e-01 -5.10517597e-01 6.55044198e-01 1.06032455e+00
-4.19568300e-01 1.88756343e-02 -2.96773881e-01 4.26073894e-02
5.43270886e-01 3.18223953e-01 1.20002961e+00 -1.04299664e+00
-4.97670501e-01 -5.57147503e-01 -1.45840764e-01 -1.43210542e+00
1.66901983e-02 -8.33256125e-01 -2.02124454e-02 -1.28102493e+00
-3.50864887e-01 -7.89362311e-01 2.74837464e-01 3.60514462e-01
1.04272403e-01 5.72179854e-01 2.15116307e-01 1.29537150e-01
-2.40753405e-02 5.20008445e-01 1.61788738e+00 -9.22702253e-02
-1.69017330e-01 -2.31554925e-01 -4.92434233e-01 1.06476843e+00
6.49034858e-01 -1.00186117e-01 -4.99886841e-01 -1.02347732e+00
2.07403809e-01 2.15439443e-02 5.36639333e-01 -7.84452140e-01
-3.30446571e-01 -4.96357344e-02 3.89697134e-01 -3.82185608e-01
5.33919454e-01 -8.85291517e-01 5.05803049e-01 9.82611105e-02
-1.75856709e-01 1.95021614e-01 2.06415877e-01 2.28838444e-01
-1.16058059e-01 -5.99785037e-02 8.84036243e-01 -4.94030774e-01
-4.72456634e-01 4.33681846e-01 5.40613793e-02 -1.14525408e-01
7.87348151e-01 -5.23232281e-01 -1.57815307e-01 -3.07001352e-01
-7.71995306e-01 -3.02827686e-01 1.11395156e+00 2.05485031e-01
5.95357895e-01 -1.33383942e+00 -3.72826934e-01 6.06081843e-01
-5.59321977e-02 7.12582529e-01 1.94516525e-01 3.81046772e-01
-1.01500845e+00 6.95923300e-05 -1.97449312e-01 -8.16774905e-01
-8.42183530e-01 5.85623860e-01 6.39036834e-01 -8.08933452e-02
-8.75877202e-01 8.17455947e-01 3.69944334e-01 -7.00021803e-01
1.59415424e-01 -6.75407588e-01 1.73719689e-01 -3.65370363e-01
2.74698585e-01 -1.22323014e-01 1.55265659e-01 -5.67047000e-01
-1.63790137e-02 1.04775572e+00 1.49299756e-01 -1.82802171e-01
1.35767376e+00 4.81029898e-02 6.10497110e-02 2.52277255e-01
1.42998469e+00 -6.38483912e-02 -1.69530547e+00 -1.17060892e-01
-5.42956233e-01 -7.31210411e-01 -8.07904229e-02 -5.49870610e-01
-1.38018000e+00 1.17380893e+00 5.35568714e-01 3.81940991e-01
1.04604805e+00 -1.70654312e-01 5.42941451e-01 1.50290355e-01
4.14839387e-01 -7.96693325e-01 1.01767235e-01 3.88861984e-01
1.20076895e+00 -1.31445754e+00 -1.30107954e-01 -6.80108786e-01
-2.40867987e-01 1.10202742e+00 5.40934861e-01 -5.51976621e-01
7.16385484e-01 2.46272370e-01 3.54370385e-01 -3.31693619e-01
-2.72466481e-01 -9.66223143e-03 2.28706926e-01 7.17329562e-01
3.71550709e-01 -1.77744493e-01 2.21467003e-01 -3.30632180e-02
-4.77248073e-01 -2.31622551e-02 4.63891357e-01 9.17855799e-01
1.14425041e-01 -1.28036296e+00 -3.09182823e-01 1.20304629e-01
-1.94666177e-01 -4.85536419e-02 -3.09606582e-01 9.85093772e-01
3.52788270e-01 4.19776082e-01 3.34679812e-01 -3.30774933e-01
6.35425389e-01 -5.65938354e-02 7.51610279e-01 -4.46344435e-01
-3.24304998e-01 -2.38752868e-02 -1.69319272e-01 -7.51336575e-01
-5.25104225e-01 -5.09098709e-01 -9.04774666e-01 -4.27813858e-01
-3.22486041e-03 -3.14837843e-01 6.63897336e-01 5.68204582e-01
3.75768453e-01 4.54977721e-01 8.09251010e-01 -1.37867665e+00
-8.74738395e-02 -5.97780108e-01 -4.21044111e-01 7.45632172e-01
5.17115593e-01 -6.58788919e-01 -3.58983457e-01 5.39992094e-01] | [9.250553131103516, -3.04294753074646] |
e459717e-89e9-4eb5-b058-118b59e935ad | accurate-urban-road-centerline-extraction | 1508.06163 | null | http://arxiv.org/abs/1508.06163v2 | http://arxiv.org/pdf/1508.06163v2.pdf | Accurate Urban Road Centerline Extraction from VHR Imagery via Multiscale Segmentation and Tensor Voting | It is very useful and increasingly popular to extract accurate road
centerlines from very-high-resolution (VHR) re- mote sensing imagery for
various applications, such as road map generation and updating etc. There are
three shortcomings of current methods: (a) Due to the noise and occlusions
(owing to vehicles and trees), most road extraction methods bring in
heterogeneous classification results; (b) Morphological thinning algorithm is
widely used to extract road centerlines, while it pro- duces small spurs around
the centerlines; (c) Many methods are ineffective to extract centerlines around
the road intersections. To address the above three issues, we propose a novel
method to ex- tract smooth and complete road centerlines via three techniques:
the multiscale joint collaborative representation (MJCR) & graph cuts (GC),
tensor voting (TV) & non-maximum suppression (NMS) and fitting based connection
algorithm. Specifically, a MJCR-GC based road area segmentation method is
proposed by incorporating mutiscale features and spatial information. In this
way, a homogenous road segmentation result is achieved. Then, to obtain a
smooth and correct road centerline network, a TV-NMS based centerline
extraction method is introduced. This method not only extracts smooth road
centerlines, but also connects the discontinuous road centerlines. Finally, to
overcome the ineffectiveness of current methods in the road intersection, a
fitting based road centerline connection algorithm is proposed. As a result, we
can get a complete road centerline network. Extensive experiments on two
datasets demonstrate that our method achieves higher quantitative results, as
well as more satisfactory visual performances by comparing with state-of-the-
art methods. | ['Guangliang Cheng', 'Feiyun Zhu', 'Shiming Xiang', 'Chunhong Pan'] | 2015-08-25 | null | null | null | null | ['road-segementation'] | ['computer-vision'] | [ 3.18421125e-01 -4.57102031e-01 -1.42937496e-01 -1.09630562e-01
-3.65759373e-01 -2.69137353e-01 6.00850642e-01 5.94804548e-02
-3.30630094e-01 6.66696608e-01 4.41269353e-02 -4.79668856e-01
-2.89907336e-01 -1.46606469e+00 -4.13718492e-01 -4.77350354e-01
1.50706589e-01 1.64441317e-01 7.78081000e-01 -4.76541370e-01
5.36792278e-01 8.85617316e-01 -1.52436221e+00 -3.23704928e-01
1.41228306e+00 6.48475885e-01 1.78370282e-01 3.82326990e-01
-5.47984481e-01 2.43812203e-01 -4.09028493e-02 -6.76082894e-02
3.06438625e-01 -1.55077159e-01 -6.94932699e-01 3.29043269e-01
2.26894408e-01 -5.57069600e-01 -2.07930058e-01 1.07120550e+00
1.10322712e-02 1.72262698e-01 9.23815072e-01 -1.01465976e+00
-1.44972116e-01 1.00015320e-01 -1.40934479e+00 -6.49844036e-02
5.93303442e-02 -5.78148700e-02 8.22899222e-01 -9.67437923e-01
5.82567215e-01 1.34226739e+00 7.18635380e-01 -1.47178560e-01
-1.07108104e+00 -8.03679407e-01 2.64579713e-01 1.05806492e-01
-1.76247144e+00 -2.18629763e-01 1.07912886e+00 -3.82567286e-01
3.06905061e-01 3.63552779e-01 6.84943199e-01 2.47056484e-01
2.60002334e-02 6.58302724e-01 1.14542675e+00 -1.21124707e-01
-1.95384547e-01 -1.89928204e-01 2.46178731e-01 7.39637554e-01
2.79385418e-01 -1.97532233e-02 1.48888201e-01 3.44451554e-02
1.18583751e+00 1.42757654e-01 -4.15675193e-01 -9.22854096e-02
-1.23460376e+00 6.99606359e-01 6.93629205e-01 3.93490165e-01
-5.00959754e-01 -1.52622789e-01 2.29487225e-01 -9.92215052e-02
4.23200399e-01 -1.17702708e-01 -1.80641878e-02 2.08081827e-01
-1.17525864e+00 3.14077258e-01 3.07054013e-01 7.32419968e-01
1.44567871e+00 -1.21974228e-02 2.37449124e-01 1.02060747e+00
3.33418190e-01 7.69933820e-01 -1.71023998e-02 -8.94757509e-01
6.69334948e-01 7.99250603e-01 9.72992480e-02 -1.72202992e+00
-4.51500148e-01 -2.34199345e-01 -1.23566949e+00 2.15013742e-01
1.91917300e-01 -1.58978730e-01 -9.26900923e-01 9.46154892e-01
4.66470033e-01 3.18821520e-01 -1.64978087e-01 7.76459396e-01
8.74827445e-01 9.84324634e-01 -3.95097882e-02 -1.91698492e-01
1.24682093e+00 -7.27516890e-01 -8.66214097e-01 -6.99594989e-02
4.12401766e-01 -9.20950353e-01 8.16940069e-01 2.36474767e-01
-6.21336758e-01 -5.71637392e-01 -9.67578351e-01 8.23249966e-02
-4.34530586e-01 3.36862236e-01 6.04138553e-01 2.99262702e-01
-6.68017626e-01 4.47202265e-01 -5.23522794e-01 -1.86572418e-01
4.27541077e-01 -2.36576814e-02 -3.12127560e-01 -1.69579163e-01
-1.00309706e+00 5.79458475e-01 2.32495815e-01 5.27151346e-01
-2.72452354e-01 -4.97521192e-01 -8.84206831e-01 -3.19850743e-01
4.97342914e-01 -3.55117977e-01 5.81069350e-01 -5.67311823e-01
-1.25525093e+00 5.00613511e-01 -3.54960978e-01 -4.19287235e-02
6.80174947e-01 -8.65000039e-02 -5.77828348e-01 2.46218324e-01
2.73131073e-01 5.51772892e-01 4.88364339e-01 -1.52960706e+00
-1.08077443e+00 -4.52578723e-01 -4.14098531e-01 2.77659655e-01
-1.03611844e-02 -1.95149973e-01 -4.67207372e-01 -5.66436648e-01
8.09521019e-01 -5.94517052e-01 -4.45546180e-01 -4.24484760e-02
-8.49772990e-01 -1.81014568e-01 1.31384814e+00 -7.64324009e-01
1.49110329e+00 -2.02625704e+00 -2.08321467e-01 7.34206796e-01
3.86043161e-01 4.63278532e-01 2.45486740e-02 3.89275014e-01
2.05160990e-01 4.97605950e-01 -8.21413040e-01 2.28839770e-01
-5.63053727e-01 1.52951330e-01 -2.26649761e-01 4.76111501e-01
1.53097540e-01 5.60686767e-01 -7.60403872e-01 -8.85776758e-01
5.89462876e-01 5.16005814e-01 -1.90905354e-03 -1.00448586e-01
1.69871986e-01 3.01213115e-01 -8.65492463e-01 7.21830547e-01
1.28907168e+00 3.85116726e-01 -4.36078608e-01 -5.60557723e-01
-8.61726820e-01 -3.11776102e-01 -1.65971732e+00 1.15358329e+00
-1.96620077e-01 3.79484594e-01 2.25256994e-01 -8.31095874e-01
1.56111217e+00 -5.25227152e-02 4.23789114e-01 -5.92684329e-01
1.22204162e-02 4.91851181e-01 -3.51009548e-01 -5.85614264e-01
6.99904561e-01 8.79528075e-02 2.58790582e-01 8.47905800e-02
-7.69602060e-01 -3.93145531e-01 1.05789609e-01 1.58062220e-01
4.99983162e-01 2.07010761e-01 -9.51310769e-02 -2.17935517e-01
9.35167313e-01 4.41995829e-01 8.24647427e-01 1.94464758e-01
-5.44966795e-02 6.35277033e-01 2.15392619e-01 -4.27539557e-01
-9.68957961e-01 -7.81117558e-01 -3.95971209e-01 1.03052013e-01
6.72434270e-01 -6.20108023e-02 -6.80852830e-01 -2.08318308e-01
-5.11774868e-02 2.53592849e-01 -2.98852772e-01 3.40007365e-01
-7.87895858e-01 -7.36106455e-01 3.74276608e-01 2.70341843e-01
1.37311780e+00 -7.18761444e-01 -1.48385152e-01 3.74397904e-01
-4.51999515e-01 -1.05003929e+00 -2.44347274e-01 -7.67843127e-01
-1.02502096e+00 -1.06281769e+00 -9.33851421e-01 -6.68344021e-01
7.58002579e-01 1.06712103e+00 4.20108646e-01 2.73823887e-01
-1.65343001e-01 -2.78534472e-01 -3.41273010e-01 1.01051219e-01
2.48659924e-02 2.24185307e-02 -4.35223699e-01 3.27853620e-01
5.63918091e-02 -5.54515600e-01 -6.98503494e-01 6.42872632e-01
-6.32238686e-01 3.90431523e-01 7.28764236e-01 5.42636037e-01
9.17457044e-01 5.51014841e-01 4.80843008e-01 -9.72720265e-01
4.84606653e-01 -4.52022612e-01 -7.25715518e-01 5.51202595e-02
-6.56684399e-01 -4.78043050e-01 5.32515883e-01 8.67999252e-03
-1.15893269e+00 2.43486669e-02 -2.96053886e-01 -1.36058286e-01
-2.03990772e-01 5.33940434e-01 -1.03015073e-01 -1.82312042e-01
3.15820932e-01 3.96292686e-01 -2.10297685e-02 -4.65076894e-01
4.57843602e-01 9.66825247e-01 5.86306870e-01 -1.71789020e-01
1.15301490e+00 8.23479950e-01 2.48632833e-01 -1.22610211e+00
-3.47179651e-01 -6.12867534e-01 -6.86546266e-01 -3.66371840e-01
9.06394958e-01 -7.73000479e-01 -5.27101457e-01 7.71192968e-01
-1.08966100e+00 2.54387334e-02 2.12482706e-01 6.36741936e-01
-2.25992352e-01 6.12537444e-01 -5.54815412e-01 -1.00740373e+00
-3.96062046e-01 -1.07828093e+00 8.23675513e-01 5.93070209e-01
2.90907472e-01 -7.39330053e-01 -1.54821783e-01 3.80150765e-01
3.87758493e-01 6.35614872e-01 8.69667590e-01 2.33607158e-01
-8.80016148e-01 -7.57107511e-02 -8.88797641e-01 8.94733295e-02
1.85930148e-01 6.91978037e-01 -6.18443549e-01 1.94402739e-01
-5.54596484e-01 4.05118972e-01 1.00168097e+00 5.05537927e-01
8.93739223e-01 -1.60361096e-01 -5.75510979e-01 6.87259853e-01
1.58627367e+00 1.27652586e-01 1.07339895e+00 5.63952684e-01
1.07752109e+00 9.55152214e-01 1.08548260e+00 1.33461535e-01
6.52810991e-01 4.20207709e-01 4.32501018e-01 -6.63216770e-01
-1.28404930e-01 -4.66003746e-01 8.88398066e-02 7.51343429e-01
-4.24095988e-01 1.07048117e-01 -7.58239865e-01 7.23951221e-01
-1.89286971e+00 -9.04669523e-01 -1.19478047e+00 2.23360276e+00
3.92644525e-01 7.48843849e-02 1.44437566e-01 4.71698493e-01
1.11528623e+00 3.66002858e-01 -3.91346157e-01 -2.75638312e-01
-3.39115798e-01 -1.91875443e-01 8.75791013e-01 6.60603702e-01
-1.05152059e+00 1.14879572e+00 4.80180788e+00 1.24095178e+00
-1.10083079e+00 -1.37908772e-01 4.96432096e-01 8.81812871e-01
-5.15954137e-01 2.62437046e-01 -8.27121317e-01 2.23472849e-01
3.68259847e-02 1.57964155e-01 2.12608472e-01 4.87086952e-01
6.83190167e-01 -4.97655571e-01 1.98831216e-01 8.66200566e-01
-4.33631599e-01 -1.38748372e+00 1.87570348e-01 1.15905777e-01
7.44143069e-01 -5.88544123e-02 -3.83260608e-01 -1.34158254e-01
1.47948533e-01 -7.66110063e-01 2.94367790e-01 6.63236916e-01
9.52998817e-01 -9.77186978e-01 6.29682958e-01 4.35653538e-01
-1.91381478e+00 1.25331283e-01 -3.77730370e-01 1.43123344e-01
3.55407953e-01 1.03739965e+00 -4.62903023e-01 1.06072438e+00
5.50701141e-01 9.85373199e-01 -4.06745881e-01 1.20675743e+00
-3.17432940e-01 4.04775590e-01 -3.52002919e-01 1.59291014e-01
4.33024555e-01 -9.15641367e-01 6.83269262e-01 1.16447771e+00
1.88931152e-01 1.78245842e-01 2.77542949e-01 8.83239031e-01
2.93653131e-01 6.24666989e-01 -7.23233640e-01 2.92304516e-01
7.18903840e-01 1.56932271e+00 -1.03822362e+00 -2.21885756e-01
-2.75117934e-01 5.65973520e-01 -8.06030408e-02 5.52705944e-01
-4.85953182e-01 -9.49150562e-01 2.27791145e-01 4.17261392e-01
-5.98683618e-02 -5.17605066e-01 -4.08398032e-01 -8.76850009e-01
-6.34573773e-03 -4.35491562e-01 1.66929998e-02 -7.17674434e-01
-8.28697205e-01 3.40666950e-01 -4.97723138e-03 -1.47894180e+00
3.95876527e-01 -4.46243808e-02 -9.26878154e-01 9.17637467e-01
-2.15846801e+00 -1.32754827e+00 -6.28021836e-01 6.33102715e-01
4.91963446e-01 2.41075397e-01 3.32602143e-01 4.04897928e-01
-7.63349712e-01 -9.97604206e-02 6.23018257e-02 2.65238255e-01
2.45084748e-01 -8.16762447e-01 2.45473802e-01 1.06195021e+00
-4.00404274e-01 4.81195807e-01 2.56443113e-01 -9.31744814e-01
-1.13065588e+00 -1.27349770e+00 7.45911121e-01 5.27427793e-01
4.65621263e-01 1.81909338e-01 -1.15277374e+00 2.79897481e-01
-2.70378441e-01 -1.46641478e-01 3.21046039e-02 -2.12987006e-01
-4.64219163e-04 -4.58021402e-01 -1.15130329e+00 7.86760688e-01
8.28068256e-01 -9.72688869e-02 -3.63927722e-01 1.08432040e-01
4.51255202e-01 -9.61483344e-02 -8.32579851e-01 8.39670300e-01
5.92583299e-01 -1.09068251e+00 9.10187602e-01 3.55488628e-01
3.42511803e-01 -9.81821895e-01 2.23420739e-01 -1.02988029e+00
-2.22065940e-01 -4.50041473e-01 5.07284701e-01 1.61964357e+00
3.40854019e-01 -9.39025998e-01 5.36847711e-01 3.71070914e-02
-1.94418043e-01 -7.80858219e-01 -6.62462354e-01 -4.99738842e-01
-2.01566339e-01 -2.70177394e-01 7.13094711e-01 1.06066895e+00
-5.35604954e-01 2.14852437e-01 -1.01442732e-01 2.55981505e-01
7.46738136e-01 3.46661896e-01 1.01691878e+00 -1.50899541e+00
6.25622988e-01 -6.97352648e-01 -1.89248323e-01 -1.25582683e+00
-1.40982747e-01 -4.74670768e-01 2.06663683e-02 -1.95164979e+00
-6.67711645e-02 -1.16674244e+00 2.90847987e-01 1.98043108e-01
-5.32900617e-02 7.91188926e-02 -2.22787544e-01 6.46178365e-01
9.36564710e-03 5.74556172e-01 1.75542545e+00 3.64351757e-02
-4.29395884e-01 1.27163276e-01 -4.00096595e-01 8.85548711e-01
8.61720920e-01 -7.88657665e-02 -3.41066122e-01 -1.71206191e-01
5.14042936e-02 1.01078719e-01 2.51354843e-01 -8.79389703e-01
2.14044616e-01 -4.59806234e-01 1.20268390e-01 -1.24919307e+00
1.29084617e-01 -6.18934989e-01 1.50569394e-01 4.08023238e-01
3.20358276e-01 -1.56980291e-01 6.52339077e-04 6.27078474e-01
-3.68179172e-01 -8.29515532e-02 8.73666525e-01 -6.70951977e-02
-8.94590318e-01 4.30278152e-01 -3.94553483e-01 -2.65434921e-01
1.07590592e+00 -6.77121341e-01 -5.09557903e-01 -1.84863254e-01
-1.58776462e-01 7.00567603e-01 4.64192897e-01 1.18656397e-01
9.07406271e-01 -1.17530143e+00 -9.67329025e-01 1.66280121e-01
6.81134090e-02 4.73550618e-01 4.49007720e-01 9.84865189e-01
-1.03236735e+00 7.03346431e-02 -1.12964272e-01 -6.16930902e-01
-1.15424967e+00 1.48977507e-02 1.81285068e-01 -1.10407369e-02
-8.76350999e-01 3.20348710e-01 2.35648192e-02 -4.55871850e-01
-4.00409877e-01 -1.98366120e-01 -7.44886160e-01 1.98531568e-01
3.53690505e-01 8.95786583e-01 -2.43240580e-01 -1.17677319e+00
-1.59017593e-01 1.51128578e+00 1.79744631e-01 -2.06978202e-01
1.17360222e+00 -4.95871753e-01 -1.90657869e-01 2.19319537e-01
8.69370818e-01 1.62182674e-01 -1.03143489e+00 -1.63264006e-01
-2.88446844e-01 -6.91552997e-01 4.00925636e-01 -1.85173333e-01
-1.25116313e+00 1.02949381e+00 3.84839714e-01 3.09603810e-01
1.06107152e+00 -5.12624443e-01 1.27141595e+00 2.80392300e-02
4.04886633e-01 -1.20547628e+00 -5.56676507e-01 3.11116636e-01
6.23073101e-01 -9.87060010e-01 3.31589609e-01 -1.16623151e+00
-4.86930609e-01 1.34591472e+00 4.88546878e-01 -2.49676183e-01
7.21066654e-01 -9.21904817e-02 -7.61235133e-02 -2.62999892e-01
1.03254035e-01 -5.90855420e-01 6.90878555e-02 4.39968288e-01
1.78963151e-02 2.24990368e-01 -6.41771436e-01 -2.04655677e-02
-1.06328323e-01 -1.10507727e-01 6.30967438e-01 6.91895902e-01
-9.61954176e-01 -9.30338919e-01 -7.04566836e-01 5.43932736e-01
6.40308345e-03 1.35358170e-01 2.74431482e-02 8.62340868e-01
2.51132905e-01 1.22738791e+00 1.78041682e-01 -5.53042233e-01
3.82300109e-01 -6.94097996e-01 -1.40570149e-01 -3.20407838e-01
-6.45044595e-02 3.87693197e-01 1.40512243e-01 -3.50073904e-01
-5.32367885e-01 -5.76911211e-01 -1.66299200e+00 -3.91752720e-01
-7.24189401e-01 1.18317097e-01 8.09096873e-01 9.49290872e-01
1.43153220e-01 1.91250712e-01 1.09377611e+00 -8.08546126e-01
3.11534494e-01 -6.72407448e-01 -7.80627370e-01 1.65844798e-01
1.34415403e-01 -8.08121324e-01 -3.97375375e-01 -1.12592325e-01] | [8.6122465133667, -1.6918890476226807] |
a80ab26b-37c2-418b-99b3-87faca635a67 | clear-memory-augmented-auto-encoder-for | 2208.03879 | null | https://arxiv.org/abs/2208.03879v2 | https://arxiv.org/pdf/2208.03879v2.pdf | Clear Memory-Augmented Auto-Encoder for Surface Defect Detection | In surface defect detection, due to the extreme imbalance in the number of positive and negative samples, positive-samples-based anomaly detection methods have received more and more attention. Specifically, reconstruction-based methods are the most popular. However, existing methods are either difficult to repair abnormal foregrounds or reconstruct clear backgrounds. Therefore, we propose a clear memory-augmented auto-encoder (CMA-AE). At first, we propose a novel clear memory-augmented module (CMAM), which combines the encoding and memoryencoding in a way of forgetting and inputting, thereby repairing abnormal foregrounds and preserving clear backgrounds. Secondly, a general artificial anomaly generation algorithm (GAAGA) is proposed to simulate anomalies that are as realistic and feature-rich as possible. At last, we propose a novel multi scale feature residual detection method (MSFR) for defect segmentation, which makes the defect location more accurate. Extensive comparison experiments demonstrate that CMA-AE achieves state-of-the-art detection accuracy and shows great potential in industrial applications. | ['Bin Li', 'Wenyong Yu', 'Lixin Tang', 'Tongzhi Niu', 'Wei Luo'] | 2022-08-08 | null | null | null | null | ['defect-detection'] | ['computer-vision'] | [ 4.42767680e-01 -3.15531850e-01 3.78403425e-01 -5.38829677e-02
-4.55912828e-01 2.85191774e-01 2.72715420e-01 2.04099864e-01
-5.71951568e-02 4.15241539e-01 -3.42299759e-01 7.35430885e-03
1.99137971e-01 -1.13228631e+00 -5.25916576e-01 -9.64649975e-01
3.01914155e-01 3.81109267e-02 8.35160315e-01 -1.77551210e-02
4.86872196e-01 5.66924691e-01 -1.96325684e+00 3.42356294e-01
1.30321109e+00 1.07296872e+00 2.96261907e-01 5.66804469e-01
-4.32377338e-01 7.02551663e-01 -7.62472034e-01 -2.58168787e-01
1.21803053e-01 -6.42069101e-01 -4.39120889e-01 4.99813378e-01
1.29630983e-01 -3.61013949e-01 -4.28063959e-01 1.15501571e+00
4.92727518e-01 -1.29171116e-02 4.62352574e-01 -1.15202188e+00
-5.46972036e-01 -5.20359678e-03 -1.00646782e+00 4.42554682e-01
1.00161776e-01 2.14176610e-01 3.52341056e-01 -1.06822205e+00
3.15269142e-01 1.18496919e+00 4.48384345e-01 6.75702214e-01
-8.76623213e-01 -4.56987768e-01 3.21797639e-01 4.54474032e-01
-1.22664428e+00 -2.61356115e-01 9.72798765e-01 -3.51254433e-01
6.26303911e-01 4.61020410e-01 7.67185450e-01 7.69452512e-01
3.39853168e-01 1.20445240e+00 7.40127683e-01 -5.70371807e-01
3.86219323e-01 -2.57105768e-01 7.94002041e-03 9.11934555e-01
6.67724609e-01 -2.36488283e-01 -5.38449824e-01 -1.50412291e-01
9.47730184e-01 4.82463539e-01 -4.83313531e-01 -2.98302323e-01
-1.10847187e+00 6.03685617e-01 -3.37059773e-03 2.50487864e-01
-4.99262393e-01 -1.23976044e-01 5.22300243e-01 2.19798282e-01
5.17145336e-01 2.05828756e-01 -1.04187533e-01 9.89593416e-02
-8.32844853e-01 2.73155216e-02 1.52961329e-01 4.56934661e-01
6.28123462e-01 5.32751560e-01 -1.81636646e-01 1.14064717e+00
3.58346641e-01 5.07966042e-01 7.38080025e-01 -6.42150283e-01
9.98882428e-02 9.71901357e-01 -2.29067206e-01 -1.20016932e+00
-2.60935370e-02 -4.50961053e-01 -1.15324724e+00 5.82900345e-01
-8.41559097e-02 2.84730256e-01 -1.17102063e+00 1.21718085e+00
3.76604468e-01 6.61029041e-01 1.09345645e-01 7.17368245e-01
4.32975352e-01 7.35107422e-01 -2.27951631e-01 -4.04034853e-01
9.66339111e-01 -9.47601676e-01 -7.69395947e-01 -3.97357017e-01
5.18510938e-01 -6.62395120e-01 1.02803242e+00 7.56958246e-01
-1.05513048e+00 -4.34146881e-01 -1.20449233e+00 3.07182312e-01
-1.11683123e-01 3.04223150e-01 5.85653782e-01 4.60982442e-01
-6.62318051e-01 5.10486364e-01 -1.14432359e+00 -1.25133321e-01
4.75309223e-01 4.03889082e-02 -2.79159248e-01 -2.78072566e-01
-6.82610869e-01 3.74547064e-01 3.34756225e-01 3.45702618e-01
-8.16565454e-01 -3.15977097e-01 -8.87267768e-01 -2.62169719e-01
2.79338509e-01 -3.80901843e-01 8.98610234e-01 -9.47052419e-01
-1.32212484e+00 7.14721382e-01 -1.36071786e-01 -2.28568852e-01
3.21307927e-01 -2.60838777e-01 -7.42677093e-01 2.26830199e-01
6.18166476e-03 1.58731773e-01 1.11869514e+00 -1.30940032e+00
-8.86616349e-01 -5.74377537e-01 -3.61751258e-01 -8.00473839e-02
-5.17024934e-01 -1.09303035e-01 -4.00221586e-01 -1.21702576e+00
7.91087985e-01 -3.23943228e-01 -3.82275343e-01 2.28892118e-01
-5.33956707e-01 1.38105124e-01 1.08403206e+00 -8.36954057e-01
1.68243814e+00 -2.45961428e+00 1.90905124e-01 2.94363827e-01
1.70027360e-01 6.36945784e-01 -4.98100705e-02 3.54659185e-02
-1.38870135e-01 -2.40399107e-01 -8.73751938e-01 -4.42970127e-01
-4.52205509e-01 4.36316788e-01 -2.78986216e-01 3.20297539e-01
5.22334218e-01 2.98892081e-01 -7.12549686e-01 -5.74759126e-01
3.64069611e-01 1.64644837e-01 -6.39717340e-01 3.31860960e-01
-8.64229947e-02 4.01880026e-01 -4.74461585e-01 1.22394407e+00
1.03451109e+00 -5.19073717e-02 -3.30282092e-01 1.54305212e-02
-1.68959890e-02 -3.58592451e-01 -1.49997699e+00 1.50653100e+00
-3.13049823e-01 2.46892348e-01 1.96804747e-01 -9.24558222e-01
1.11381972e+00 -1.79354586e-02 3.55493218e-01 -7.65872300e-01
-1.00271106e-01 6.00446343e-01 -1.89571276e-01 -5.94325840e-01
4.92184192e-01 8.89229178e-02 3.15391898e-01 2.05166832e-01
-2.84925044e-01 1.78049773e-01 7.16658607e-02 5.03274240e-02
1.37540162e+00 -1.74838960e-01 2.04470798e-01 5.05387783e-04
1.02310300e+00 -2.94795930e-01 1.05444968e+00 3.71929228e-01
-1.23417847e-01 9.80512857e-01 3.65393251e-01 -4.14822489e-01
-6.58403516e-01 -9.20306921e-01 -5.22500314e-02 3.13024044e-01
4.26639587e-01 -8.18698034e-02 -7.85395205e-01 -7.09226429e-01
-4.44112904e-02 7.15873122e-01 -4.41860527e-01 -5.25096416e-01
-7.43493259e-01 -1.02489638e+00 2.80705780e-01 6.33050501e-01
9.25403953e-01 -1.20037746e+00 -6.85754895e-01 1.59729436e-01
-3.76198329e-02 -5.94392419e-01 -1.39335811e-01 -1.35056734e-01
-1.15456235e+00 -1.09933603e+00 -8.05274844e-01 -8.23619902e-01
1.05045974e+00 3.04748833e-01 7.21004367e-01 6.64871812e-01
-6.50211692e-01 2.83344686e-01 -5.40059626e-01 -2.67782092e-01
-4.24973339e-01 -4.07357424e-01 -7.90560916e-02 4.87429678e-01
2.35932812e-01 -5.24512231e-01 -6.18713081e-01 1.06426291e-01
-1.18252432e+00 -1.07402638e-01 9.86475050e-01 1.02663815e+00
9.01056707e-01 2.69599378e-01 5.89989305e-01 -8.44902158e-01
3.18819523e-01 -4.04713899e-01 -4.87756699e-01 1.44790456e-01
-6.47687256e-01 -1.77912131e-01 5.05744398e-01 -3.94378811e-01
-1.22408521e+00 -5.42483293e-02 -4.17088419e-01 -6.60236418e-01
-2.10547328e-01 -9.67770442e-03 -5.50375938e-01 5.95150690e-04
3.17118496e-01 6.35988772e-01 5.26760072e-02 -5.65724611e-01
-1.63064882e-01 6.44573867e-01 7.69297063e-01 -3.56442004e-01
7.17288554e-01 4.84227896e-01 -3.93078141e-02 -9.44823265e-01
-4.07086164e-01 -3.82419735e-01 -3.95834595e-01 -1.26978457e-01
7.05016375e-01 -6.77427053e-01 -1.00376487e-01 1.32628191e+00
-1.12220252e+00 -3.08710188e-02 -5.04142642e-01 2.43751943e-01
-2.54389018e-01 7.56943166e-01 -6.68000042e-01 -1.04860008e+00
-3.56665701e-01 -1.09313250e+00 1.11598647e+00 3.17272991e-01
3.15047890e-01 -6.53772771e-01 -1.87009841e-01 2.41043530e-02
3.21518660e-01 4.90893811e-01 9.80609119e-01 -5.00746846e-01
-8.07133198e-01 -2.35757902e-01 -1.44644111e-01 6.90150380e-01
2.44808555e-01 3.21587618e-03 -8.86552036e-01 -1.76811293e-01
2.72356570e-01 1.82888567e-01 1.07394338e+00 1.93416089e-01
1.55989158e+00 -1.19168930e-01 -6.13959968e-01 3.69659841e-01
1.29285359e+00 5.30589223e-01 1.09885633e+00 4.78068233e-01
6.94734812e-01 2.32274756e-01 1.00056398e+00 6.49765313e-01
-6.51220530e-02 3.83740216e-01 6.48784697e-01 -2.71310270e-01
-2.12217256e-01 2.21861266e-02 3.57818097e-01 1.18830037e+00
2.36154586e-01 -3.81105900e-01 -6.31753147e-01 7.98750758e-01
-1.81620109e+00 -7.11781323e-01 -3.89796644e-01 2.32869816e+00
5.69860041e-01 4.19851631e-01 -2.52837658e-01 7.69348145e-01
9.41528738e-01 4.99276631e-02 -5.86991608e-01 -1.52032703e-01
-2.30517372e-01 3.22310418e-01 1.07284598e-01 1.95886105e-01
-9.53059852e-01 5.43328106e-01 5.45534182e+00 1.11685264e+00
-9.73681450e-01 8.08586478e-02 6.98794901e-01 2.67030805e-01
-3.66194457e-01 -1.33521348e-01 -4.94684547e-01 8.27841341e-01
3.49089354e-01 3.15350831e-01 -2.43188255e-02 9.37778413e-01
-2.03529909e-01 -3.05157751e-01 -6.38231933e-01 1.09166598e+00
3.10111344e-01 -1.06889284e+00 3.61647546e-01 -2.38581568e-01
4.54107970e-01 -6.73986554e-01 -6.54005930e-02 9.94075835e-02
-3.37937921e-01 -4.28655744e-01 5.69079399e-01 7.60054886e-01
6.78289890e-01 -8.19105446e-01 1.00220084e+00 2.08608627e-01
-1.11981487e+00 -3.68171841e-01 -5.26446581e-01 2.70844072e-01
1.70511529e-01 1.15682268e+00 -2.52787203e-01 5.39027750e-01
6.62199974e-01 6.36768341e-01 -6.72971606e-01 1.32309937e+00
-8.06100592e-02 7.52904117e-01 -1.33970320e-01 2.89545268e-01
-2.77269900e-01 -1.16416492e-01 6.88089311e-01 8.99944365e-01
6.25596225e-01 1.66933045e-01 2.47793198e-01 7.05682516e-01
2.57469833e-01 6.71531782e-02 -4.39968616e-01 7.23710582e-02
3.59112084e-01 9.72602963e-01 -9.54487920e-01 -2.07259119e-01
-4.75228041e-01 1.47445166e+00 -1.65574938e-01 6.76104054e-02
-6.36575222e-01 -7.20658064e-01 6.83788776e-01 2.22512275e-01
1.91264749e-01 2.99214758e-02 -2.18819618e-01 -1.14199996e+00
1.94181070e-01 -8.83330882e-01 4.68487352e-01 -5.74574292e-01
-1.15249765e+00 4.70444888e-01 -4.13262814e-01 -1.35344815e+00
2.27365896e-01 -5.74312866e-01 -8.49197805e-01 4.34987158e-01
-1.45960224e+00 -9.72932875e-01 -6.16739213e-01 4.40076977e-01
1.07514346e+00 -3.77958715e-01 6.63272619e-01 7.39166141e-01
-1.10594857e+00 5.71868658e-01 -8.79433453e-02 1.24180000e-02
3.95579994e-01 -1.04115450e+00 4.60801423e-01 1.44540310e+00
-1.28944367e-01 2.55161256e-01 4.41931516e-01 -9.90864158e-01
-1.20048892e+00 -1.18030071e+00 1.61153644e-01 3.94522473e-02
1.55784369e-01 -3.18404846e-02 -1.50727582e+00 4.43178803e-01
-1.30123392e-01 3.08710396e-01 2.08036676e-01 -6.46221817e-01
8.90253335e-02 -2.44011268e-01 -1.43122387e+00 4.40051615e-01
9.88836586e-01 -9.69574973e-02 -3.78367007e-01 1.23109467e-01
7.56849706e-01 -4.66685891e-01 -3.62281650e-01 7.83742785e-01
-2.37239976e-04 -1.18867874e+00 7.82432437e-01 5.18936664e-02
3.02238673e-01 -7.28323936e-01 -1.94444835e-01 -1.08091640e+00
-2.89339989e-01 -1.86487794e-01 -5.72117746e-01 1.43393660e+00
1.33045958e-02 -9.39026356e-01 7.79509842e-01 6.88478397e-03
-6.66024983e-01 -1.01262379e+00 -8.83153677e-01 -7.00179040e-01
-4.09619451e-01 -1.81099713e-01 6.07694089e-01 8.06377351e-01
-5.01590788e-01 -3.13602597e-01 -1.53308153e-01 4.72772092e-01
4.62465972e-01 4.12734598e-02 5.24820149e-01 -1.39288318e+00
-3.07197034e-01 -2.41761014e-01 -7.78375328e-01 -8.86838138e-01
-2.12567046e-01 -4.78761017e-01 2.39491120e-01 -1.44095194e+00
-3.23304981e-02 -5.73612869e-01 -4.47440892e-01 3.26005787e-01
-3.78311515e-01 1.49006411e-01 -5.11062205e-01 1.18596040e-01
-4.72096771e-01 6.92549944e-01 1.07931995e+00 -1.43783629e-01
-8.04097503e-02 4.60735746e-02 -3.18878293e-01 8.74782801e-01
6.23812616e-01 -3.23400140e-01 -3.21069479e-01 -3.92224520e-01
-1.58747032e-01 -1.61004901e-01 2.95389086e-01 -1.47592902e+00
7.56550133e-02 -8.96170065e-02 5.00588536e-01 -8.27920318e-01
2.19875097e-01 -6.21521592e-01 -8.89212936e-02 7.59554565e-01
2.56010294e-01 3.31056923e-01 2.51147389e-01 9.19115305e-01
-4.60908860e-01 -4.37260598e-01 9.14539576e-01 -1.82303518e-01
-9.99555528e-01 3.23278636e-01 -3.89598191e-01 -1.34781733e-01
1.22352481e+00 -4.61242616e-01 -1.95151165e-01 1.88697293e-01
-3.81756037e-01 3.71302031e-02 5.85388660e-01 3.10711741e-01
1.19586778e+00 -1.30880260e+00 -6.50874436e-01 8.27007353e-01
2.36495867e-01 3.65489960e-01 5.80587089e-01 7.06609488e-01
-8.98945630e-01 -3.52216959e-01 -1.18860386e-01 -7.11225331e-01
-1.19271946e+00 4.93494242e-01 1.57541171e-01 -5.75998984e-02
-9.85670865e-01 9.34311271e-01 1.69821650e-01 -2.59525254e-02
7.23787919e-02 -1.43043250e-01 -1.64063394e-01 -2.70509779e-01
7.77214468e-01 7.85160840e-01 2.19129771e-01 -3.99279386e-01
-2.21533984e-01 5.23458421e-01 -1.42772406e-01 3.57106596e-01
1.11948466e+00 -7.95356035e-02 -3.72802675e-01 3.25214028e-01
5.21095872e-01 2.83239901e-01 -1.00948453e+00 -4.17373367e-02
1.24621466e-02 -9.13430691e-01 4.03148718e-02 -4.07388061e-01
-1.49698782e+00 1.05482924e+00 1.05362999e+00 1.99418187e-01
1.63215959e+00 -4.84369427e-01 1.34462440e+00 1.06749371e-01
4.07527119e-01 -9.86680806e-01 5.66374898e-01 7.12404251e-02
6.69930518e-01 -9.94748294e-01 -1.03914768e-01 -6.71239734e-01
-3.44056696e-01 1.05030489e+00 1.23015869e+00 -1.26590714e-01
4.75632638e-01 3.67358744e-01 -4.14953679e-02 -1.28577739e-01
-4.31386113e-01 6.64966404e-02 -7.82432407e-02 5.48690438e-01
-3.64812389e-02 -1.98912174e-01 -3.07866007e-01 6.17163122e-01
5.06591320e-01 -3.11129689e-01 5.36030114e-01 1.25098050e+00
-7.52837300e-01 -1.02555442e+00 -5.11735737e-01 7.25032508e-01
-4.42025393e-01 1.90926492e-01 2.16710642e-02 3.30402792e-01
3.78893644e-01 7.69737244e-01 2.28531092e-01 -4.72729892e-01
2.89612323e-01 -4.14430685e-02 2.04105303e-01 -6.69388294e-01
-5.99210411e-02 -1.42441943e-01 -3.67923528e-01 -6.07748091e-01
-7.13071302e-02 -6.75589800e-01 -1.50610328e+00 5.36359921e-02
-7.99042463e-01 5.14672473e-02 3.85504752e-01 7.90274262e-01
5.32618821e-01 9.65817750e-01 5.58926940e-01 -5.24907827e-01
-1.97637558e-01 -7.46369064e-01 -7.39130080e-01 2.97309011e-01
3.50739896e-01 -9.00563896e-01 -4.42093462e-01 -1.59975380e-01] | [7.517828941345215, 1.9540423154830933] |
c5cdfbd7-f759-4322-bc02-3f03f4e11434 | stochastic-compartmental-models-of-covid-19 | 2206.13907 | null | https://arxiv.org/abs/2206.13907v1 | https://arxiv.org/pdf/2206.13907v1.pdf | Stochastic compartmental models of COVID-19 pandemic must have temporally correlated uncertainties | Compartmental models are an important quantitative tool in epidemiology, enabling us to forecast the course of a communicable disease. However, the model parameters, such as the infectivity rate of the disease, are riddled with uncertainties, which has motivated the development and use of stochastic compartmental models. Here, we first show that a common stochastic model, which treats the uncertainties as white noise, is fundamentally flawed since it erroneously implies that greater parameter uncertainties will lead to the eradication of the disease. Then, we present a principled modeling of the uncertainties based on reasonable assumptions on the contacts of each individual. Using the central limit theorem and Doob's theorem on Gaussian Markov processes, we prove that the correlated Ornstein-Uhlenbeck process is the appropriate tool for modeling uncertainties in the infectivity rate. We demonstrate our results using a compartmental model of the COVID-19 pandemic and the available US data from the Johns Hopkins University COVID-19 database. In particular, we show that the white noise stochastic model systematically underestimates the severity of the Omicron variant of COVID-19, whereas the Ornstein-Uhlenbeck model correctly forecasts the course of this variant. Moreover, using an SIS model of sexually transmitted disease, we derive an exact closed-form solution for the asymptotic distribution of infected individuals. This analytic result shows that the white noise model underestimates the severity of the pandemic because of unrealistic noise-induced transitions. Our results strongly support the need for temporal correlations in modeling of uncertainties in compartmental models of infectious disease. | ['Mohammad Farazmand', 'Konstantinos Mamis'] | 2022-06-23 | null | null | null | null | ['epidemiology'] | ['medical'] | [ 4.65056896e-02 -2.75012225e-01 2.17641637e-01 2.10428536e-01
-6.71772733e-02 -3.86699289e-01 5.43776035e-01 1.99853629e-01
-4.46589828e-01 1.02451634e+00 1.17313173e-02 -6.69867456e-01
-5.07502139e-01 -7.80961335e-01 -5.07090092e-01 -1.00261140e+00
-2.57678926e-01 9.24830198e-01 1.39044169e-02 -7.61510432e-02
-3.24486613e-01 4.64101732e-01 -9.99948382e-01 -5.18203795e-01
9.11335289e-01 5.24067581e-01 2.04726145e-01 9.25794482e-01
-1.75754175e-01 3.27825993e-01 -8.06354284e-01 -6.09719828e-02
2.80539133e-02 -4.71050590e-01 -1.09137826e-01 -4.74489123e-01
-9.29537058e-01 -3.29866618e-01 -8.37472156e-02 8.28844309e-01
2.53465325e-01 -2.02607155e-01 1.30334878e+00 -1.39851880e+00
-2.41255492e-01 2.34600902e-01 -4.11629617e-01 2.53341019e-01
-7.71183744e-02 7.80251473e-02 1.64466903e-01 -1.35706335e-01
5.17209470e-01 1.36482155e+00 1.10644507e+00 5.98458946e-01
-1.37218237e+00 -2.83478945e-01 -2.46746046e-03 -4.56506252e-01
-1.53597021e+00 -1.12983771e-02 1.73386917e-01 -8.25705707e-01
7.16683030e-01 3.07634115e-01 7.28288412e-01 1.14840233e+00
9.79086041e-01 5.73958866e-02 1.10235441e+00 -1.83172271e-01
4.67347771e-01 -5.66447750e-02 1.71137825e-01 1.00495823e-01
1.02735710e+00 4.24710989e-01 3.67229939e-01 -7.36902654e-01
9.07112122e-01 6.42886102e-01 -8.70215222e-02 -2.18432188e-01
-5.93730986e-01 7.90260136e-01 -2.11153224e-01 2.62993902e-01
-7.80177474e-01 3.51974785e-01 2.13852581e-02 -7.40075856e-02
6.83317184e-01 -2.47445911e-01 -4.65957642e-01 5.35570122e-02
-6.29264653e-01 2.90698618e-01 8.87537658e-01 5.34529567e-01
6.54532164e-02 3.52308340e-02 -2.02600628e-01 2.84391493e-01
5.84152222e-01 1.60892820e+00 -3.24285179e-01 -7.10216045e-01
-7.69627243e-02 1.02734165e-02 7.44853914e-01 -5.58784902e-01
-7.27122545e-01 -4.64984596e-01 -1.05553341e+00 -3.22823972e-02
6.92764342e-01 -6.65633202e-01 -9.75250423e-01 1.92723358e+00
1.74603403e-01 1.60774127e-01 3.23068760e-02 1.80301860e-01
-4.62065004e-02 8.24527144e-01 2.52755433e-01 -8.44650269e-01
1.09377074e+00 6.60527274e-02 -1.10705483e+00 3.97988170e-01
4.50222909e-01 -4.87149805e-01 2.97173887e-01 6.01221249e-02
-9.22601938e-01 1.43880576e-01 -3.85961294e-01 9.34365690e-01
-2.12877572e-01 -5.92522025e-01 3.07304233e-01 8.80661130e-01
-9.44337904e-01 4.63770390e-01 -1.17166352e+00 -7.53431737e-01
-1.20036282e-01 5.57650905e-03 2.16466755e-01 2.32768953e-02
-1.24403298e+00 7.97064781e-01 -2.31337354e-01 4.16822672e-01
-9.10013080e-01 -7.69499481e-01 -3.07274073e-01 1.16228715e-01
9.24026072e-02 -9.10537243e-01 9.49699819e-01 -3.70942235e-01
-8.97077501e-01 2.88940817e-01 -2.68191218e-01 -5.54393291e-01
6.83543742e-01 3.30556571e-01 -4.56033647e-01 1.42936602e-01
-4.64567542e-02 -2.81106651e-01 5.82575023e-01 -1.37955463e+00
-2.59839207e-01 -4.73171681e-01 -3.84128630e-01 -1.13177411e-01
3.82459491e-01 1.51882440e-01 7.69581273e-02 -4.91004735e-01
-2.98362106e-01 -1.10436654e+00 -4.42966670e-01 -3.49494785e-01
-2.53213227e-01 1.72125101e-01 3.22961658e-01 -6.30365789e-01
1.12421167e+00 -1.94585657e+00 -3.24577421e-01 3.51515979e-01
3.84968594e-02 2.63098087e-02 1.78080842e-01 9.41890717e-01
2.88048863e-01 3.88716161e-01 -4.62972581e-01 -3.11963558e-01
-1.26507610e-01 5.55593491e-01 -3.16553742e-01 9.25514877e-01
-1.61210622e-03 4.48218316e-01 -9.10716712e-01 -1.59523740e-01
-3.05628143e-02 8.70273471e-01 -3.29358160e-01 1.54575678e-02
-5.03842421e-02 5.47216773e-01 -3.14318866e-01 1.46914989e-01
9.72249150e-01 -3.21144462e-01 2.32194901e-01 4.88473266e-01
-2.49661818e-01 -3.31574172e-01 -9.33112800e-01 4.43100989e-01
-4.27727342e-01 4.67325561e-02 2.36949474e-01 -6.43970609e-01
3.60787898e-01 5.38892627e-01 6.30360365e-01 1.38681466e-02
2.09681779e-01 2.41137624e-01 1.71800688e-01 -2.35135928e-01
8.06049630e-02 -7.11132824e-01 1.35182351e-01 7.00171173e-01
-3.15167367e-01 -1.71341002e-01 9.89151150e-02 1.72696680e-01
9.97010767e-01 -1.69108018e-01 6.59088418e-02 -4.46258843e-01
-8.61140341e-02 3.34906913e-02 6.58260345e-01 8.96442711e-01
-2.44201824e-01 2.71748960e-01 8.16190660e-01 -2.14607295e-04
-9.67900515e-01 -1.61893821e+00 -4.64545608e-01 1.63130015e-01
-7.32161179e-02 2.11445227e-01 -7.63289571e-01 4.31229472e-02
1.13463521e-01 7.65419662e-01 -1.01615012e+00 -1.12833753e-01
-1.60511240e-01 -1.50858438e+00 4.27297145e-01 2.48483997e-02
-3.19285393e-02 -5.37436783e-01 -5.07005811e-01 2.51127362e-01
-4.49440256e-02 -6.92713618e-01 -1.99772179e-01 8.82904455e-02
-7.79425323e-01 -1.37149024e+00 -1.00648940e+00 -7.74438307e-02
9.16454732e-01 4.86385003e-02 5.29777229e-01 7.57172257e-02
-1.37226984e-01 7.47746885e-01 7.20009804e-02 -6.29130661e-01
-8.01080346e-01 -6.49736524e-01 4.11506742e-01 -2.29215503e-01
2.34098732e-01 1.91775645e-04 -7.19492912e-01 2.63063908e-01
-9.91073012e-01 -6.00439906e-01 6.97039953e-03 5.94616830e-01
4.64790136e-01 4.34032649e-01 6.66780889e-01 -7.30824471e-01
7.81086326e-01 -7.13859081e-01 -7.22134888e-01 3.48605424e-01
-6.94042981e-01 1.34301065e-02 3.34317267e-01 -4.12105322e-01
-1.22956383e+00 -3.53492975e-01 1.05302520e-01 7.42001401e-04
5.45771904e-02 5.01179695e-01 3.26355964e-01 3.76845837e-01
1.49493292e-01 2.32142910e-01 3.22277427e-01 -4.97743011e-01
-2.12802470e-01 7.71061599e-01 2.03940228e-01 -3.55754972e-01
5.48078537e-01 9.43083763e-01 3.81489784e-01 -1.19827187e+00
-2.13994250e-01 -3.29084307e-01 -9.40920115e-02 -2.03005046e-01
8.41677070e-01 -8.88574243e-01 -1.02405572e+00 9.10850704e-01
-1.11731017e+00 -5.26346922e-01 -2.63175666e-01 8.09619367e-01
-8.60666215e-01 1.66299269e-01 -8.99699688e-01 -1.78307641e+00
5.93031906e-02 -9.72745836e-01 6.78334117e-01 -2.17848103e-02
-1.52172893e-01 -1.54659736e+00 6.08952224e-01 -9.92048010e-02
7.16838181e-01 3.91399294e-01 1.06777108e+00 -6.08815193e-01
-9.73148048e-02 -2.04912037e-01 2.64613256e-02 1.85582601e-02
2.62040496e-01 4.47156161e-01 -4.88328397e-01 -2.56423593e-01
4.07244533e-01 6.45214081e-01 5.08415520e-01 1.08312976e+00
1.75144657e-01 -2.96542287e-01 -6.66971087e-01 2.93544680e-01
1.55547583e+00 5.06794393e-01 4.24175054e-01 -6.36309013e-02
-1.99170504e-02 7.91449130e-01 2.89125770e-01 7.66243100e-01
3.70617926e-01 2.33464882e-01 3.15712869e-01 1.37064546e-01
3.78098637e-01 -2.84202427e-01 2.15843201e-01 5.59725642e-01
-1.63274780e-01 -4.80811864e-01 -1.13763833e+00 5.69628179e-01
-1.53983796e+00 -1.16433799e+00 -5.92211306e-01 2.80738473e+00
5.85942149e-01 -3.48460907e-03 4.42933112e-01 -3.42605293e-01
7.90509880e-01 -3.98583204e-01 -2.75678188e-01 -4.23254400e-01
-2.89824426e-01 -1.88705131e-01 9.56841290e-01 8.44201386e-01
-5.33478200e-01 1.86021715e-01 7.80041647e+00 4.64617461e-01
-7.21033573e-01 2.10650280e-01 5.84074676e-01 2.30451956e-01
-4.21004623e-01 -1.39909729e-01 -7.98994660e-01 8.42377186e-01
1.44749606e+00 -5.21161675e-01 2.25234985e-01 -2.01590173e-02
9.08301413e-01 -5.13566852e-01 -4.78557974e-01 2.93541253e-01
-4.33651865e-01 -7.53358424e-01 -3.05796653e-01 3.97250623e-01
8.20757568e-01 -1.23965017e-01 2.64721304e-01 1.44484444e-02
7.58236229e-01 -8.30500960e-01 1.36855885e-01 9.76504803e-01
4.77644354e-01 -1.04178596e+00 1.07300270e+00 6.02703691e-01
-8.04865420e-01 2.22627476e-01 -2.46731296e-01 3.52843404e-02
6.28075719e-01 9.27745640e-01 -5.92011929e-01 1.41240120e-01
3.03092927e-01 -8.91996399e-02 1.69094726e-01 1.17008913e+00
1.07579455e-01 7.26259112e-01 -7.01631308e-01 1.26061708e-01
-5.47695234e-02 -3.87783825e-01 6.20833635e-01 9.28159297e-01
6.93815112e-01 4.21888717e-02 -3.78550947e-01 7.90313065e-01
6.56325340e-01 -2.06362352e-01 -6.36011064e-01 -4.19392228e-01
3.51623207e-01 5.51598012e-01 -7.37960994e-01 -1.33676901e-01
-1.33241653e-01 5.98681271e-01 -5.60950577e-01 8.85592043e-01
-7.84986556e-01 -2.04623789e-01 8.08806360e-01 2.95312643e-01
3.57432425e-01 -3.40654999e-01 -1.34512663e-01 -8.75894427e-01
-4.85342920e-01 -4.08653408e-01 6.32503331e-02 -4.48379606e-01
-1.20739043e+00 5.16620457e-01 2.66533524e-01 -7.42001772e-01
-7.02190697e-01 -4.29647297e-01 -7.11011648e-01 1.28472221e+00
-1.19487262e+00 -6.38779163e-01 4.35733169e-01 3.77851218e-01
-1.14556767e-01 4.91250217e-01 6.64067566e-01 -1.16196118e-01
-5.43875813e-01 -1.80723906e-01 1.02223325e+00 -3.50960612e-01
3.06709945e-01 -9.46344912e-01 2.66325802e-01 5.82340896e-01
-8.97235990e-01 9.85460043e-01 1.44871497e+00 -1.41853786e+00
-1.11530721e+00 -9.77208138e-01 9.40660715e-01 -4.49203193e-01
9.93525803e-01 -2.13694200e-01 -6.89868569e-01 3.66222054e-01
-9.90357548e-02 -4.43556041e-01 6.48551822e-01 -3.45661759e-01
1.96638554e-01 3.22527826e-01 -1.57070720e+00 5.63013852e-01
5.23328185e-01 -3.40824246e-01 -2.07172707e-01 5.01999736e-01
7.54023015e-01 3.24638665e-01 -8.32724571e-01 3.09551567e-01
5.82045734e-01 -7.20588863e-01 8.79961967e-01 -7.84701228e-01
-2.45179206e-01 -6.72261491e-02 -1.53389484e-01 -1.50886476e+00
-6.99165687e-02 -6.12518370e-01 1.58223301e-01 9.25210118e-01
3.14732194e-01 -1.14955115e+00 1.93475366e-01 6.55116379e-01
6.02889597e-01 -4.77327943e-01 -1.10032284e+00 -1.42255414e+00
4.86290455e-01 -2.62305647e-01 4.29473579e-01 3.70540023e-01
-2.66979814e-01 -3.83137494e-01 -4.91023779e-01 3.96088421e-01
1.17857766e+00 -3.76386613e-01 2.04641715e-01 -1.47011316e+00
-1.98965505e-01 -3.68906967e-02 8.26308317e-03 -4.59276438e-01
-4.18673120e-02 1.37476757e-01 2.60130763e-01 -1.56944394e+00
4.05816406e-01 -4.49078381e-01 -1.30204469e-01 -3.31663877e-01
-1.26010850e-01 -2.25198627e-01 -6.57304144e-03 2.19078556e-01
1.25126749e-01 3.85388345e-01 8.75462353e-01 2.19076321e-01
-2.53359526e-01 6.35153830e-01 -3.07062894e-01 8.14945996e-01
7.12365091e-01 -7.03028381e-01 -4.84032363e-01 3.10531724e-02
4.40729082e-01 3.02416444e-01 4.02830362e-01 -6.65051162e-01
-1.31445266e-02 -5.06819189e-01 2.69854795e-02 -8.18760931e-01
3.99507612e-01 -1.00975919e+00 9.56856072e-01 1.31982255e+00
9.55025852e-02 6.04281984e-02 2.90153712e-01 1.12993038e+00
5.10179162e-01 -2.47812077e-01 6.11424685e-01 6.36245124e-03
4.90864575e-01 3.34853977e-01 -1.18495488e+00 1.52601004e-01
9.45373535e-01 1.60909459e-01 -6.14266515e-01 -7.48948932e-01
-8.54087949e-01 2.38990173e-01 6.57344699e-01 -7.44653121e-02
2.74071962e-01 -7.13841021e-01 -5.93754768e-01 -2.39941683e-02
-4.59102422e-01 -5.51058054e-01 6.37671173e-01 1.21825016e+00
-6.46071196e-01 7.18722463e-01 1.25667468e-01 -5.52757502e-01
-9.74363446e-01 8.95034611e-01 4.45910662e-01 -2.98213214e-01
8.92741829e-02 2.07928434e-01 3.77490014e-01 -2.79433727e-01
-7.81540126e-02 -3.03267688e-01 1.92999452e-01 3.36960331e-02
6.63966954e-01 6.86170697e-01 -4.70917642e-01 -7.11709499e-01
-5.14543414e-01 4.55232859e-01 3.54356706e-01 -3.25481415e-01
1.21695101e+00 -6.80618525e-01 -1.85002476e-01 8.56377900e-01
6.24389231e-01 -5.74311651e-02 -1.28629112e+00 1.21484265e-01
-2.18748197e-01 7.80970529e-02 -2.79025882e-01 -6.99102819e-01
-3.59503776e-01 8.70534778e-01 6.05078518e-01 5.65058410e-01
7.50982702e-01 -2.19510376e-01 4.36795741e-01 -3.55255343e-02
4.98323679e-01 -7.40231335e-01 -7.71584928e-01 4.07621175e-01
6.32752419e-01 -6.95671380e-01 -1.88833550e-01 -3.88231456e-01
-3.43143851e-01 5.11042655e-01 -1.75663158e-01 4.55657728e-02
1.14995909e+00 6.58568740e-01 -3.05418968e-02 9.97329876e-02
-9.06305850e-01 -2.01263145e-01 -3.67370456e-01 1.07493436e+00
8.83298069e-02 4.39408034e-01 -5.94404161e-01 5.96857607e-01
4.40163046e-01 2.33583301e-01 8.99775267e-01 7.15530455e-01
-4.89068657e-01 -8.25670123e-01 -8.47742736e-01 3.82325590e-01
-5.62061906e-01 -2.46470332e-01 -1.31688580e-01 7.41005063e-01
-8.37775990e-02 1.07769156e+00 3.76638800e-01 3.71156961e-01
5.02699949e-02 8.08143839e-02 3.75645578e-01 -3.85814726e-01
-3.06239694e-01 3.10757667e-01 -2.30159685e-01 1.35672733e-01
-1.52233630e-01 -7.02854037e-01 -1.24973476e+00 -8.31801891e-01
-1.76114812e-01 3.82695138e-01 7.96751142e-01 9.49595332e-01
2.59966850e-01 3.59296441e-01 5.36776841e-01 -2.55403817e-01
-9.60575402e-01 -7.78807342e-01 -1.13150692e+00 -1.65338039e-01
4.11348492e-01 -6.64105654e-01 -1.06760764e+00 -2.65748709e-01] | [5.976702690124512, 4.341594219207764] |
e8067ae7-1e73-4922-8625-bbf536432aa6 | scenario-discovery-via-rule-extraction | 1910.01713 | null | https://arxiv.org/abs/1910.01713v2 | https://arxiv.org/pdf/1910.01713v2.pdf | REDS: Rule Extraction for Discovering Scenarios | Scenario discovery is the process of finding areas of interest, known as scenarios, in data spaces resulting from simulations. For instance, one might search for conditions, i.e., inputs of the simulation model, where the system is unstable. Subgroup discovery methods are commonly used for scenario discovery. They find scenarios in the form of hyperboxes, which are easy to comprehend. Given a computational budget, results tend to get worse as the number of inputs of the simulation model and the cost of simulations increase. We propose a new procedure for scenario discovery from few simulations, dubbed REDS. A key ingredient is using an intermediate machine learning model to label data for subsequent use by conventional subgroup discovery methods. We provide statistical arguments why this is an improvement. In our experiments, REDS reduces the number of simulations required by 50--75\% on average, depending on the quality measure. It is also useful as a semi-supervised subgroup discovery method and for discovering better scenarios from third-party data, when a simulation model is not available. | ['Klemens Böhm', 'Vadim Arzamasov'] | 2019-10-03 | null | null | null | null | ['subgroup-discovery'] | ['methodology'] | [ 1.27403855e-01 -2.08752621e-02 -2.03800842e-01 -2.77645499e-01
-7.02876985e-01 -6.80334508e-01 7.66824126e-01 6.16447330e-01
-1.70190230e-01 1.04103029e+00 -4.71267253e-02 -9.05352712e-01
-4.64139462e-01 -9.89188731e-01 -8.28101814e-01 -6.56809390e-01
-4.59678173e-01 5.26310802e-01 8.39524791e-02 6.51963474e-03
3.03644508e-01 6.29758060e-01 -1.85794806e+00 1.10093758e-01
6.75989091e-01 4.34914291e-01 3.13351005e-02 4.07072395e-01
5.23507185e-02 1.79322079e-01 -6.66194260e-01 1.65984824e-01
3.34492296e-01 -5.40882826e-01 -5.87256610e-01 9.90063474e-02
-4.28429335e-01 4.40412201e-02 6.51129633e-02 8.63144219e-01
3.66976500e-01 3.35370481e-01 6.43005073e-01 -1.59833646e+00
7.47382820e-01 9.34204519e-01 -5.39728105e-01 1.71052411e-01
5.82014441e-01 3.00245672e-01 4.13585991e-01 -5.82454920e-01
6.25675976e-01 1.06611729e+00 5.01128852e-01 2.30806768e-02
-1.35796916e+00 -7.69265652e-01 3.96937802e-02 4.94372519e-03
-1.69804478e+00 -4.47286844e-01 6.17918313e-01 -4.18157876e-01
7.30863452e-01 5.90741515e-01 3.72168839e-01 7.90903330e-01
-6.03613928e-02 2.42214233e-01 1.14570999e+00 -5.68296790e-01
8.38306308e-01 4.63581026e-01 -9.67697650e-02 3.43973100e-01
6.16450012e-01 3.95002484e-01 -2.24192157e-01 -7.62877584e-01
6.29701793e-01 -5.55953309e-02 -1.94664180e-01 -5.11944056e-01
-1.25563729e+00 8.81785274e-01 -8.39699656e-02 1.88240260e-01
-2.63378650e-01 -3.44415754e-02 2.62034416e-01 5.79185069e-01
3.95437390e-01 8.70236278e-01 -5.65259397e-01 -5.51931076e-02
-1.02844751e+00 4.30906534e-01 9.10090566e-01 7.17163026e-01
7.07867622e-01 -3.46155554e-01 1.03467003e-01 1.67212039e-01
1.63610637e-01 1.41357094e-01 1.56886980e-01 -7.36008406e-01
5.37509978e-01 7.73195922e-01 6.22481346e-01 -4.79996562e-01
-5.06634831e-01 -2.11908266e-01 -5.79639852e-01 2.45355755e-01
7.69305527e-01 -4.07546192e-01 -6.96117878e-01 1.61780047e+00
4.94074553e-01 2.16935277e-01 3.11314426e-02 6.56234860e-01
2.84906149e-01 7.62149453e-01 6.97055310e-02 -7.39657700e-01
1.07673848e+00 -1.93366721e-01 -2.85502166e-01 1.76187620e-01
7.44705141e-01 -5.79136312e-01 9.74554241e-01 3.82090807e-01
-8.23698819e-01 -2.80419022e-01 -9.56659019e-01 1.28732133e+00
-6.23265386e-01 -2.00452343e-01 6.78045452e-01 7.35437572e-01
-5.66435933e-01 8.85551929e-01 -6.62191570e-01 -5.38644969e-01
8.94527063e-02 4.35132921e-01 -6.82307929e-02 1.49994940e-01
-1.22563863e+00 7.60957420e-01 5.16304612e-01 -2.63251305e-01
-9.93860483e-01 -8.21339428e-01 -7.33528197e-01 1.23031057e-01
5.64380050e-01 -3.86769265e-01 1.01019228e+00 -3.62689316e-01
-9.45849776e-01 4.91436899e-01 -1.19439758e-01 -2.51224577e-01
7.22356975e-01 4.57088351e-01 -6.59723163e-01 -1.25131413e-01
-5.02358638e-02 -2.36605089e-02 8.91790152e-01 -1.34249878e+00
-4.44995761e-01 1.63914971e-02 2.88893759e-01 -1.08446911e-01
2.85592765e-01 3.53733540e-01 -1.93739295e-01 -3.68363351e-01
-3.89253795e-02 -8.61487627e-01 -1.04469848e+00 -5.02208769e-01
-5.54018855e-01 -1.38497517e-01 5.97975969e-01 -2.16102093e-01
1.40879285e+00 -1.91282022e+00 -2.18295217e-01 1.01924574e+00
-9.40765254e-03 3.98616083e-02 2.34418690e-01 7.74387717e-01
-5.52433729e-01 5.86335063e-01 -2.74437785e-01 3.08799714e-01
-5.49877547e-02 -1.20555356e-01 -3.58212620e-01 5.14602363e-01
4.84860167e-02 3.29144031e-01 -7.23068118e-01 -4.06972408e-01
5.25319457e-01 -4.06792760e-01 -1.91439003e-01 3.10021460e-01
-3.49021196e-01 6.58270776e-01 -4.79959816e-01 2.14073539e-01
5.07575929e-01 -5.14540374e-01 2.47075066e-01 1.03060886e-01
-2.91702300e-01 1.96231842e-01 -1.78056967e+00 1.10237253e+00
-4.80854034e-01 1.48156807e-01 -3.40890080e-01 -1.19659936e+00
8.85286272e-01 4.20045316e-01 4.46180373e-01 -2.57099152e-01
1.85968056e-01 1.15257137e-01 -2.48374813e-03 -2.73064494e-01
-7.72503624e-03 -1.48995548e-01 -4.91110325e-01 9.61346447e-01
-4.46470737e-01 -2.52140254e-01 5.38313985e-01 1.31912082e-01
1.25130618e+00 -3.51860672e-01 8.60207438e-01 -4.40872580e-01
2.91543752e-01 1.67826340e-01 4.36439812e-01 9.56691265e-01
2.79418349e-01 1.56890079e-01 7.46848047e-01 -4.32095438e-01
-1.09362948e+00 -9.84751701e-01 -1.05755858e-01 4.60100830e-01
1.55936003e-01 -4.97243762e-01 -4.36591595e-01 -6.73151612e-01
7.86975399e-02 8.63995731e-01 -5.02908409e-01 1.64986290e-02
-3.44101250e-01 -9.44168568e-01 4.01761383e-02 2.76955634e-01
2.06993446e-01 -9.64952290e-01 -7.46936321e-01 3.02144319e-01
1.82551667e-01 -7.31941521e-01 1.87956288e-01 5.67922771e-01
-1.01987147e+00 -1.19647038e+00 -3.15672040e-01 -1.46911934e-01
9.07248795e-01 -5.61922900e-02 1.05368149e+00 1.51870936e-01
-2.46385604e-01 -5.45583814e-02 -3.72547090e-01 -4.82120007e-01
-5.54596543e-01 -1.14612542e-01 4.31459308e-01 -3.26737732e-01
1.54047340e-01 -7.69981086e-01 -2.90971100e-01 7.41339028e-01
-7.68122315e-01 -1.46468222e-01 5.60414076e-01 4.68646199e-01
6.62421167e-01 7.14752495e-01 7.26074338e-01 -1.03355253e+00
6.58830166e-01 -9.24754620e-01 -9.60982203e-01 2.72616863e-01
-8.78011048e-01 3.56300712e-01 7.92247057e-01 -4.61322486e-01
-8.11398923e-01 1.61227733e-01 4.13665295e-01 -3.76610219e-01
-6.29638255e-01 8.13081861e-01 -3.50827098e-01 1.64009884e-01
8.28963101e-01 -2.84477890e-01 -1.56717673e-01 -5.52488804e-01
9.00340229e-02 5.42737484e-01 4.09393758e-02 -5.28788030e-01
1.09462810e+00 2.78775930e-01 5.45865111e-02 -7.26011395e-01
-3.54293466e-01 -5.94804108e-01 -3.86762321e-01 -4.14708436e-01
7.55003244e-02 -5.91726005e-01 -6.45119786e-01 -1.01489328e-01
-6.11506939e-01 -3.71906996e-01 -3.63778085e-01 6.39445007e-01
-5.88582218e-01 1.18938282e-01 2.37295926e-01 -7.93772340e-01
1.58115029e-01 -9.02039587e-01 5.49414635e-01 3.13084006e-01
-6.46065831e-01 -7.53107607e-01 6.90963939e-02 -2.01554477e-01
6.37997538e-02 6.62175834e-01 1.09025943e+00 -1.00164807e+00
-4.30773765e-01 -3.52069557e-01 -3.26823108e-02 -3.45185250e-01
1.41154125e-01 2.36842018e-02 -7.64056683e-01 -3.86833519e-01
-2.20533267e-01 1.83801368e-01 4.42378640e-01 3.21055114e-01
1.58410704e+00 -5.29645264e-01 -7.59004533e-01 2.57180154e-01
1.37558877e+00 5.01031160e-01 4.75077361e-01 1.61945254e-01
9.71261039e-02 6.69158638e-01 9.70098555e-01 8.83084834e-01
-2.43397444e-01 6.82957947e-01 1.70982942e-01 -3.21227729e-01
5.39163888e-01 -2.62942195e-01 1.12008929e-01 2.98732877e-01
1.59280360e-01 -2.31547937e-01 -1.23394239e+00 6.57604516e-01
-1.83781695e+00 -9.62330639e-01 9.51348629e-04 2.66912103e+00
6.09919429e-01 5.19270360e-01 4.77392018e-01 3.97970766e-01
8.54508281e-01 -1.12763979e-01 -7.18528509e-01 -4.65660393e-01
3.10468704e-01 2.07027227e-01 5.04780293e-01 2.93717086e-01
-9.10708249e-01 3.24942201e-01 6.46351862e+00 9.28167224e-01
-7.76157141e-01 -2.99175233e-01 7.06963539e-01 -3.89805362e-02
-3.36437285e-01 2.79834121e-01 -5.13952851e-01 6.79236889e-01
1.46908629e+00 -7.32303977e-01 1.88175812e-01 7.89606273e-01
7.35467255e-01 -7.12612808e-01 -1.45652688e+00 8.30986321e-01
-4.13410366e-01 -1.42410898e+00 -2.93470025e-01 2.16690913e-01
6.04578853e-01 -3.69921476e-01 -3.53109479e-01 1.66635334e-01
5.76210439e-01 -1.22819602e+00 3.64864349e-01 4.21230823e-01
7.86190391e-01 -1.03229547e+00 8.36212039e-01 6.00518405e-01
-1.11561835e+00 3.86005454e-02 3.77286114e-02 -2.76221901e-01
1.06824957e-01 8.20360839e-01 -1.15458179e+00 6.13159835e-01
5.45756519e-01 7.79503882e-02 -5.77586710e-01 1.49450409e+00
5.61223552e-02 1.09362233e+00 -8.72773647e-01 -2.39310250e-01
3.04764863e-02 -1.92153320e-01 5.44529259e-01 9.91621792e-01
3.42204839e-01 1.21360101e-01 2.96419829e-01 8.18841875e-01
2.59755939e-01 -1.79726794e-01 -1.02889395e+00 -1.05583511e-01
7.64467716e-01 9.15888011e-01 -1.15279663e+00 -2.56574392e-01
-1.36685744e-01 1.76522374e-01 -4.48796034e-01 4.22482163e-01
-8.22413445e-01 -5.65298915e-01 4.06924307e-01 3.70718032e-01
1.36286691e-01 -3.60059813e-02 -3.95330101e-01 -7.69098163e-01
-4.15468141e-02 -1.09142053e+00 4.99370575e-01 -3.81708175e-01
-1.08943617e+00 1.97749063e-01 7.91179299e-01 -1.71061707e+00
-8.46131206e-01 -1.79290697e-01 -9.34766352e-01 7.21875072e-01
-7.58616149e-01 -3.73607635e-01 -1.97957054e-01 5.70523500e-01
3.06057036e-01 -1.10751860e-01 7.35596776e-01 -1.61347445e-02
-5.88130474e-01 3.88676047e-01 2.37382606e-01 -4.25605059e-01
4.43003237e-01 -1.08918703e+00 4.61728632e-01 8.97542536e-01
1.27145231e-01 5.54993510e-01 1.34128594e+00 -6.44546807e-01
-1.02449453e+00 -1.10918140e+00 7.41040647e-01 -3.78151298e-01
7.15362668e-01 -5.06989479e-01 -7.74964392e-01 2.04010025e-01
-3.60464603e-01 -2.69613236e-01 6.67468965e-01 3.15762371e-01
-4.22869772e-02 -4.70898859e-02 -1.28726733e+00 7.62108803e-01
8.07767034e-01 -2.63672262e-01 -4.35968399e-01 4.49188799e-01
5.02355456e-01 7.07449466e-02 -8.57557237e-01 3.02062660e-01
1.21608727e-01 -6.82344556e-01 8.32808495e-01 -7.58525193e-01
-1.20684244e-01 -5.84488094e-01 -2.33068261e-02 -1.52933431e+00
1.95561439e-01 -1.16223693e+00 -6.53616935e-02 1.08821034e+00
8.94232392e-01 -4.91909832e-01 7.18587101e-01 5.40713787e-01
5.22518933e-01 -6.09818876e-01 -7.73651838e-01 -1.16778290e+00
-9.34671909e-02 -6.49442732e-01 1.06973815e+00 9.29837048e-01
2.81991094e-01 1.21550880e-01 -5.90567142e-02 2.66120762e-01
8.90768945e-01 3.88618171e-01 8.11582506e-01 -1.42857838e+00
-1.59303024e-01 -2.41902187e-01 -1.59558356e-01 -3.76202703e-01
-3.73146683e-02 -4.66064304e-01 -5.22904582e-02 -1.20846295e+00
6.43253475e-02 -8.61059487e-01 -1.10407665e-01 3.88612777e-01
4.41394858e-02 -3.37872177e-01 -8.63165408e-02 1.42281741e-01
-5.51303148e-01 1.43164009e-01 4.36780959e-01 2.11012438e-01
-6.12396538e-01 7.11988926e-01 -5.37186444e-01 9.75168884e-01
1.17094159e+00 -6.65115595e-01 -4.61069584e-01 5.54462135e-01
2.84451336e-01 4.77856368e-01 3.20513666e-01 -1.02695453e+00
1.71045333e-01 -5.35380185e-01 3.14635068e-01 -8.10504436e-01
-1.63846418e-01 -8.45288754e-01 8.26737463e-01 4.07884747e-01
-3.14543515e-01 6.07408844e-02 3.39094281e-01 6.58177793e-01
8.33911598e-02 -4.59948629e-01 6.25502229e-01 -1.48709379e-02
-6.92534566e-01 4.14635353e-02 -7.44116664e-01 -1.26085222e-01
1.31311822e+00 -2.40038797e-01 -7.84871692e-04 -5.61259985e-01
-8.12816262e-01 4.66314942e-01 4.80027676e-01 2.02789232e-01
3.51624906e-01 -1.06268573e+00 -6.25879705e-01 3.33145559e-02
4.61126655e-01 -1.22047514e-01 -1.25979960e-01 6.10933185e-01
-2.08572239e-01 3.16989243e-01 6.50032610e-02 -3.26722383e-01
-1.25598752e+00 7.19853103e-01 8.90223905e-02 -2.54694492e-01
-6.02599263e-01 3.86268675e-01 -1.01180255e-01 -4.28126752e-01
2.92022139e-01 -3.29890281e-01 3.42196226e-02 2.77300864e-01
6.33959711e-01 5.30800164e-01 3.79012257e-01 -3.41807120e-02
-6.55724347e-01 1.36523306e-01 1.73481569e-01 -2.26260304e-01
1.18911111e+00 -2.17833072e-02 2.08729655e-01 5.34516156e-01
8.96368504e-01 -4.32612225e-02 -1.10253930e+00 -6.33304045e-02
3.20130378e-01 -3.97708207e-01 -1.15195528e-01 -8.92573178e-01
-7.26120412e-01 3.98982525e-01 5.09658456e-01 4.85244602e-01
1.07792282e+00 2.95955122e-01 -6.61767498e-02 5.95426798e-01
7.15501666e-01 -9.56916332e-01 -3.18890095e-01 5.93304932e-02
7.89457500e-01 -1.21001828e+00 2.28986055e-01 -3.11274111e-01
-4.63891953e-01 8.82528782e-01 3.21841776e-01 1.54504150e-01
8.98434103e-01 5.16654372e-01 -2.96503365e-01 -3.20711106e-01
-8.88051331e-01 -2.43840113e-01 -5.11058867e-02 5.65540254e-01
-1.96548834e-01 3.19293171e-01 -3.38726968e-01 7.17662871e-01
-1.39604583e-01 -1.87504441e-01 6.60130501e-01 7.90001214e-01
-5.09307384e-01 -8.94362152e-01 -5.41152000e-01 9.18749988e-01
-1.59536917e-02 1.13083825e-01 -3.11693937e-01 9.65328455e-01
1.02985121e-01 1.42746460e+00 1.31033972e-01 -3.64303768e-01
3.40906322e-01 4.88378992e-03 5.27602062e-02 -7.79727638e-01
-3.88286173e-01 -2.18000397e-01 5.44933558e-01 -7.35814214e-01
-1.81172714e-01 -9.10320699e-01 -1.02811491e+00 -5.27110279e-01
-4.27190900e-01 7.21104741e-01 5.80447435e-01 1.12523186e+00
1.36068255e-01 3.23592663e-01 1.11317647e+00 -5.40550232e-01
-5.13221979e-01 -6.65338218e-01 -6.01552248e-01 4.89723086e-02
-3.86787169e-02 -8.02730560e-01 -8.02913725e-01 5.06185666e-02] | [7.312282562255859, 4.136934757232666] |
1167403f-59ed-4421-bfa9-1f91fe0f232b | bandwidth-scalable-fully-mask-based-deep-fcrn | 2205.04276 | null | https://arxiv.org/abs/2205.04276v2 | https://arxiv.org/pdf/2205.04276v2.pdf | Bandwidth-Scalable Fully Mask-Based Deep FCRN Acoustic Echo Cancellation and Postfiltering | Although today's speech communication systems support various bandwidths from narrowband to super-wideband and beyond, state-of-the art DNN methods for acoustic echo cancellation (AEC) are lacking modularity and bandwidth scalability. Our proposed DNN model builds upon a fully convolutional recurrent network (FCRN) and introduces scalability over various bandwidths up to a fullband (FB) system (48 kHz sampling rate). This modular approach allows joint wideband (WB) pre-training of mask-based AEC and postfilter stages with dedicated losses, followed by a separate training of them on FB data. A third lightweight blind bandwidth extension stage is separately trained on FB data, flexibly allowing to extend the WB postfilter output towards higher bandwidths until reaching FB. Thereby, higher frequency noise and echo are reliably suppressed. On the ICASSP 2022 Acoustic Echo Cancellation Challenge blind test set we report a competitive performance, showing robustness even under highly delayed echo and dynamic echo path changes. | ['Tim Fingscheidt', 'Pejman Mowlaee', 'Zhengyang Li', 'Karim Haddad', 'Rasmus Kongsgaard Olsson', 'Ernst Seidel'] | 2022-05-09 | null | null | null | null | ['bandwidth-extension', 'acoustic-echo-cancellation', 'bandwidth-extension', 'acoustic-echo-cancellation'] | ['audio', 'medical', 'speech', 'speech'] | [ 1.47024289e-01 -4.96573523e-02 5.09710491e-01 -2.42923334e-01
-8.83416593e-01 -4.56614435e-01 4.78388190e-01 -3.02406073e-01
-7.23288059e-01 2.03408509e-01 6.39328122e-01 -6.24888361e-01
-1.78325981e-01 -1.38141975e-01 -3.30814928e-01 -6.07843697e-01
-5.24454355e-01 -1.06955700e-01 3.98541391e-01 -3.44890773e-01
-2.64557987e-01 5.90078712e-01 -1.61186194e+00 2.67674476e-01
5.64162493e-01 9.51317787e-01 1.96948946e-01 1.36539352e+00
4.20144424e-02 3.79046798e-01 -9.14275289e-01 -3.62914264e-01
3.96971613e-01 -1.95424691e-01 -2.52159417e-01 -3.74784231e-01
4.83687192e-01 -7.16048121e-01 -4.36483949e-01 8.16580355e-01
1.20532930e+00 3.31607610e-01 1.95317805e-01 -6.61560357e-01
-1.93739712e-01 9.62955236e-01 3.24169993e-02 2.19066292e-01
1.80647656e-01 8.21271092e-02 7.35435426e-01 -1.09890985e+00
8.79284814e-02 1.10462248e+00 1.17887115e+00 6.34370327e-01
-1.22085035e+00 -9.14350629e-01 -1.74364969e-02 -1.47630554e-02
-8.95989954e-01 -1.26736271e+00 7.52735913e-01 -9.06297341e-02
1.39373708e+00 2.89308310e-01 6.81522250e-01 1.34942555e+00
-4.15662080e-01 4.59191710e-01 6.88134909e-01 -4.65379030e-01
3.20783108e-01 -1.85098037e-01 2.41192384e-03 1.07678227e-01
-2.43201077e-01 5.62437356e-01 -8.18149805e-01 -1.77082390e-01
6.87654495e-01 -4.48263079e-01 -7.61653721e-01 1.78987548e-01
-8.05615783e-01 4.63699371e-01 2.20787793e-01 3.77855003e-01
-3.62555563e-01 4.43777651e-01 5.62172413e-01 6.85816467e-01
4.02142793e-01 3.27176929e-01 -5.04845023e-01 -2.10972160e-01
-1.44243526e+00 1.12950251e-01 1.07253850e+00 7.47230172e-01
3.63386899e-01 9.39941764e-01 -9.23743993e-02 1.40110397e+00
4.43441033e-01 6.83055401e-01 7.80637681e-01 -8.43759835e-01
4.19623107e-01 -5.41388214e-01 6.23604171e-02 -5.22936642e-01
-7.69108474e-01 -1.03961754e+00 -9.32523847e-01 3.17664854e-02
2.02940151e-01 -4.38810885e-01 -1.26355362e+00 1.68230987e+00
2.46644631e-01 3.20821404e-01 1.17689066e-01 1.28921187e+00
7.73855209e-01 5.23079276e-01 -2.47127414e-02 -6.81469217e-02
1.25629306e+00 -8.97312462e-01 -9.00854647e-01 -3.90031099e-01
1.68129712e-01 -1.05541575e+00 6.32264912e-01 5.74435949e-01
-1.24291551e+00 -6.98478997e-01 -1.13570154e+00 -7.02863410e-02
-1.02073424e-01 3.39411832e-02 4.17096943e-01 1.28571880e+00
-1.62456965e+00 5.18962085e-01 -6.25761330e-01 1.29196346e-01
-2.22962722e-01 4.41371411e-01 -1.79194808e-01 1.49843380e-01
-1.65067279e+00 6.07527018e-01 -1.06262751e-01 6.06413782e-01
-1.04622805e+00 -1.25464439e+00 -8.13139796e-01 3.06756645e-01
-4.60564010e-02 -3.93040210e-01 1.52458823e+00 -8.67287874e-01
-2.14509225e+00 3.15319985e-01 1.85241118e-01 -9.97917116e-01
5.74369371e-01 -5.99983692e-01 -1.21443379e+00 1.99758843e-01
-4.86480266e-01 4.93162453e-01 1.50684226e+00 -9.11573231e-01
-3.64921004e-01 2.29766428e-01 -2.36545742e-01 -6.36286587e-02
-3.96401405e-01 -3.51319611e-02 -2.95848310e-01 -9.67492342e-01
2.76371598e-01 -4.31087047e-01 -6.40857369e-02 -3.51823956e-01
-1.52023762e-01 3.35802466e-01 5.92911482e-01 -1.09854460e+00
1.38371789e+00 -2.44565272e+00 -8.99106190e-02 3.28517765e-01
3.81448641e-02 7.99759746e-01 -5.02162814e-01 4.84029800e-01
-2.88920492e-01 -2.66224086e-01 -1.73955306e-01 -7.08161831e-01
2.07145914e-01 -1.29498616e-01 -4.59683746e-01 4.81928617e-01
1.35881484e-01 3.28181267e-01 -5.06215513e-01 2.69250333e-01
2.88306415e-01 9.37295139e-01 -8.36420119e-01 2.70375550e-01
3.56880844e-01 2.29796976e-01 2.93825150e-01 4.07991618e-01
1.05344689e+00 4.34212148e-01 9.33974609e-03 -3.40807199e-01
-2.65783697e-01 5.46638966e-01 -1.34431899e+00 1.73016775e+00
-7.89898157e-01 7.76887834e-01 1.11574137e+00 -6.69263661e-01
1.10998440e+00 7.72861600e-01 1.47342950e-01 -7.87001967e-01
-1.63249835e-01 7.72504568e-01 2.95199811e-01 -2.29140580e-01
5.24379432e-01 -2.99634308e-01 3.97374839e-01 2.07347676e-01
5.65800428e-01 -2.93496549e-01 -2.91036874e-01 -8.52854773e-02
1.20153010e+00 -1.76494509e-01 -2.89864093e-01 -4.34132457e-01
7.39292324e-01 -8.88783038e-01 2.92537600e-01 8.59024107e-01
-3.33460867e-01 7.11841822e-01 -1.05308227e-01 -1.27497286e-01
-1.01500595e+00 -1.08234501e+00 -2.39566088e-01 1.42048013e+00
-2.72757441e-01 -2.64648706e-01 -8.01968813e-01 -3.87473740e-02
-1.25624659e-02 4.59274828e-01 -9.14800912e-02 -1.55240625e-01
-8.24453235e-01 -1.97857380e-01 1.07601798e+00 1.62551090e-01
5.57218671e-01 -8.82066429e-01 -5.24496138e-01 7.05612421e-01
-5.23647219e-02 -1.10937715e+00 -7.04262257e-01 4.91895318e-01
-5.78566372e-01 -3.33118409e-01 -1.20173764e+00 -6.16943181e-01
-4.92713712e-02 7.21761584e-02 1.00210106e+00 -1.44308269e-01
-2.80519426e-01 5.38224280e-01 -2.39137679e-01 -1.87545300e-01
-5.69768429e-01 -2.59010196e-02 3.30446929e-01 4.59544547e-02
-2.33068302e-01 -8.59769583e-01 -1.11742413e+00 2.32437745e-01
-8.81652117e-01 -2.88274646e-01 6.20269418e-01 8.98317754e-01
-2.64817532e-02 -3.50890279e-01 9.54572260e-01 -3.29076439e-01
7.96542704e-01 -9.39180553e-02 -7.30029106e-01 -1.59330636e-01
-2.22739682e-01 -9.34396088e-02 6.58358932e-01 -6.89542890e-01
-1.29325950e+00 -2.24558324e-01 -8.85399282e-01 -4.56745714e-01
3.14506367e-02 1.33899942e-01 1.23632746e-02 -1.86674342e-01
6.87544465e-01 2.32154533e-01 -8.38439241e-02 -8.43779385e-01
2.78536439e-01 8.06624651e-01 7.83541739e-01 -1.43372640e-01
5.32823801e-01 4.89407443e-02 -3.13716441e-01 -1.06475651e+00
-2.28905663e-01 -6.34806991e-01 1.32953199e-02 -3.07296544e-01
4.05001611e-01 -1.01185572e+00 -7.43274987e-01 7.55954206e-01
-9.69673336e-01 -5.74046791e-01 -8.82269293e-02 8.23016763e-01
-2.07136914e-01 1.42811611e-01 -9.21171188e-01 -1.13216066e+00
-7.20439196e-01 -9.00316715e-01 9.07693863e-01 4.52377228e-03
-6.17319234e-02 -6.76514149e-01 1.25045612e-01 -9.50043425e-02
1.37077940e+00 -5.37783921e-01 4.27717119e-01 -6.32000625e-01
-2.05257431e-01 -9.39733982e-02 1.06686503e-01 6.36485577e-01
-1.75639570e-01 -3.64401668e-01 -1.60070586e+00 -5.94537556e-01
2.69991934e-01 -7.94851705e-02 1.23551881e+00 5.52806735e-01
5.93616486e-01 -3.03665400e-01 1.26079679e-01 9.67950404e-01
1.04698527e+00 2.59189177e-02 6.24071300e-01 -1.42185673e-01
1.53612748e-01 4.47899371e-01 -2.49479383e-01 3.35228324e-01
-2.15410203e-01 5.15229166e-01 3.34454417e-01 -2.63524532e-01
-7.39074230e-01 -8.22615102e-02 5.33127666e-01 1.14172626e+00
1.21732533e-01 -2.60941923e-01 -6.61198080e-01 6.27408206e-01
-1.11768866e+00 -6.82071686e-01 6.17335998e-02 2.35545897e+00
8.97552788e-01 2.23839171e-02 -8.42638910e-02 4.07471389e-01
5.69522083e-01 2.98428774e-01 -4.00714666e-01 -6.90721035e-01
-8.80038217e-02 6.54488564e-01 6.43137693e-01 8.57773721e-01
-1.00602722e+00 7.17295051e-01 6.84403706e+00 8.71366203e-01
-1.27611482e+00 2.07592458e-01 1.97448373e-01 -3.05930287e-01
-4.57361490e-01 -4.80432391e-01 -5.90593100e-01 3.71210300e-03
1.48786628e+00 5.33404648e-01 7.46243477e-01 4.30980355e-01
-4.41588834e-02 2.38766849e-01 -7.44437158e-01 8.65975976e-01
-2.90861845e-01 -9.44308341e-01 -3.71698231e-01 -1.57146573e-01
2.31178224e-01 3.32577974e-01 2.22173721e-01 5.28361857e-01
8.86815935e-02 -8.94092798e-01 8.41964006e-01 4.42950636e-01
1.05148208e+00 -8.18858981e-01 4.24657851e-01 9.19200554e-02
-1.18142879e+00 -3.36218357e-01 -3.54595870e-01 2.31376499e-01
2.22248495e-01 9.42603052e-01 -7.72689879e-01 3.92386198e-01
7.13380158e-01 5.27324528e-02 2.14912179e-05 1.07433689e+00
-2.81021744e-02 8.64852965e-01 -5.43852270e-01 2.92363852e-01
1.88304633e-01 2.53827721e-01 8.01590562e-01 1.76053488e+00
5.50868750e-01 8.85591730e-02 -4.69653189e-01 4.67910945e-01
-1.37168154e-01 -2.59739488e-01 7.41388649e-02 1.82121053e-01
6.85915530e-01 1.16800821e+00 -2.07889840e-01 3.23838205e-03
-3.12166423e-01 9.95180368e-01 8.79286677e-02 8.67667317e-01
-7.06524014e-01 -7.25210547e-01 9.53558385e-01 -2.27725580e-01
5.55628002e-01 -2.05305129e-01 2.62413025e-01 -8.08804154e-01
-1.72220245e-01 -8.90806496e-01 -2.97690537e-02 -4.48863745e-01
-1.05004573e+00 7.11448789e-01 -5.58572233e-01 -9.31740165e-01
-3.69377822e-01 -5.74969471e-01 -3.70840371e-01 1.28221345e+00
-1.68196189e+00 -7.64784932e-01 1.82548925e-01 6.07500792e-01
3.43497425e-01 -1.79500535e-01 1.10043097e+00 8.74258697e-01
-9.10849646e-02 9.60168123e-01 1.18192337e-01 8.32959488e-02
7.87195265e-01 -1.06632376e+00 7.22845852e-01 9.72920656e-01
-3.70811038e-02 7.32780397e-01 8.67998421e-01 -3.29509169e-01
-1.24016988e+00 -8.48010480e-01 7.33410835e-01 4.32605982e-01
6.55547738e-01 -8.39695394e-01 -1.04744172e+00 8.08884427e-02
1.78870559e-01 1.64628953e-01 5.29866874e-01 7.05748200e-02
-6.65174186e-01 -4.78160083e-01 -9.64838862e-01 4.94137585e-01
9.16132689e-01 -8.50223780e-01 -1.61942050e-01 -1.77334160e-01
9.35093045e-01 -4.85998422e-01 -7.17054129e-01 4.55057651e-01
8.13904285e-01 -1.43312514e+00 1.21571004e+00 -1.16315946e-01
-1.31078735e-01 -5.80227673e-02 -1.09503463e-01 -1.44644105e+00
-9.58801731e-02 -1.42220747e+00 -3.05570751e-01 1.20119429e+00
4.55936939e-01 -8.32558870e-01 3.03290755e-01 2.99554288e-01
-7.45232642e-01 -1.09740466e-01 -1.31279004e+00 -6.73695624e-01
-1.78452134e-01 -9.73355532e-01 6.20371819e-01 3.63384753e-01
-1.09186165e-01 -2.05531985e-01 -6.45351112e-01 4.88351017e-01
5.44195056e-01 -4.50113982e-01 2.90285647e-01 -8.31845224e-01
-7.74470508e-01 -6.58898771e-01 -1.12573124e-01 -1.42605281e+00
-2.08436325e-01 -4.12459970e-01 3.91306132e-01 -1.02243137e+00
-7.14762092e-01 -3.11877847e-01 -3.13413471e-01 1.45679368e-02
9.57094133e-02 3.40082169e-01 1.81152031e-01 -6.30112141e-02
-1.80303808e-02 4.56205904e-01 1.00776839e+00 5.17148338e-02
-3.55370760e-01 2.12370962e-01 -7.76549131e-02 5.36368251e-01
4.86855537e-01 -1.93008155e-01 -3.74956816e-01 -5.79635680e-01
-3.21064773e-03 4.34517324e-01 3.38456661e-01 -1.24163854e+00
5.93922615e-01 6.55013204e-01 2.55012244e-01 -3.22614431e-01
6.14504695e-01 -7.44862378e-01 3.85907233e-01 5.08283019e-01
-4.40011233e-01 -5.50430298e-01 6.38971627e-01 5.40399849e-01
-3.55251223e-01 -4.14629616e-02 7.34254301e-01 2.38472730e-01
-3.24952096e-01 -2.93308999e-02 -7.23831892e-01 -6.92536086e-02
-2.04355866e-02 -5.73256006e-03 -2.10058168e-01 -6.69798374e-01
-9.93231714e-01 -1.24589622e-01 -4.07429308e-01 2.99719512e-01
6.59306049e-01 -1.07856655e+00 -8.53658080e-01 8.59829068e-01
-3.87431383e-01 -4.00256336e-01 8.06241393e-01 7.65011787e-01
-3.35135043e-01 5.34800410e-01 2.45671377e-01 -4.65255857e-01
-9.32207525e-01 8.69332254e-02 9.46610570e-01 -5.76608665e-02
-9.20664668e-01 1.28778505e+00 3.20380181e-03 -4.11272198e-01
8.06184709e-01 -4.26227421e-01 -3.01741641e-02 -1.07150152e-01
7.14214087e-01 4.51778471e-01 6.40786409e-01 -3.64052325e-01
-3.61180395e-01 3.22746426e-01 2.44461715e-01 -7.26111352e-01
1.21080136e+00 -2.86666900e-01 1.46749780e-01 9.96684358e-02
1.23274434e+00 3.11151654e-01 -1.35696769e+00 -2.62052983e-01
-4.51651178e-02 -8.09599012e-02 3.93374264e-01 -1.11212170e+00
-1.06049407e+00 9.87783611e-01 9.50585902e-01 3.09962720e-01
1.56519186e+00 -4.19268996e-01 7.86268830e-01 8.98029953e-02
3.80984508e-02 -1.00705814e+00 -1.46079108e-01 7.05585241e-01
1.09034216e+00 -6.50198460e-01 -5.25729597e-01 5.19155748e-02
-2.08031654e-01 1.37251294e+00 -4.47696708e-02 -9.86080840e-02
9.47437823e-01 7.64650881e-01 4.42867577e-01 1.24372438e-01
-5.57680070e-01 -5.08715995e-02 4.98535067e-01 6.29539192e-01
5.82038224e-01 -6.98913708e-02 7.12888688e-02 7.84176290e-01
-4.27997082e-01 -4.81948316e-01 1.87458977e-01 2.46413514e-01
-5.04329801e-01 -8.54106486e-01 -5.58968127e-01 1.61455274e-01
-6.77421093e-01 -5.49549103e-01 1.21023268e-01 5.10572970e-01
-2.07402542e-01 1.34448385e+00 2.75268137e-01 -2.13359654e-01
3.65691066e-01 2.44142085e-01 1.00466304e-01 -1.07469760e-01
-1.21503639e+00 8.70026112e-01 3.50893497e-01 -7.58845031e-01
-2.32860610e-01 -3.25752437e-01 -1.07980537e+00 -1.83077604e-01
-5.51248372e-01 1.38578981e-01 1.05428231e+00 5.92520058e-01
2.39299566e-01 1.02446759e+00 4.82036352e-01 -1.08190823e+00
-7.27486491e-01 -1.33186483e+00 -7.42465615e-01 -1.60057489e-02
1.27334356e+00 -1.03459015e-01 -7.30235636e-01 -1.84622630e-01] | [15.073112487792969, 5.949801921844482] |
b908ce39-1a80-4a54-84cb-40ed4d891d5d | robustness-of-convolutional-neural-networks | 2108.01995 | null | https://arxiv.org/abs/2108.01995v1 | https://arxiv.org/pdf/2108.01995v1.pdf | Robustness of convolutional neural networks to physiological ECG noise | The electrocardiogram (ECG) is one of the most widespread diagnostic tools in healthcare and supports the diagnosis of cardiovascular disorders. Deep learning methods are a successful and popular technique to detect indications of disorders from an ECG signal. However, there are open questions around the robustness of these methods to various factors, including physiological ECG noise. In this study we generate clean and noisy versions of an ECG dataset before applying Symmetric Projection Attractor Reconstruction (SPAR) and scalogram image transformations. A pretrained convolutional neural network is trained using transfer learning to classify these image transforms. For the clean ECG dataset, F1 scores for SPAR attractor and scalogram transforms were 0.70 and 0.79, respectively, and the scores decreased by less than 0.05 for the noisy ECG datasets. Notably, when the network trained on clean data was used to classify the noisy datasets, performance decreases of up to 0.18 in F1 scores were seen. However, when the network trained on the noisy data was used to classify the clean dataset, the performance decrease was less than 0.05. We conclude that physiological ECG noise impacts classification using deep learning methods and careful consideration should be given to the inclusion of noisy ECG signals in the training data when developing supervised networks for ECG classification. | ['P. J. Aston', 'N. A. S. Smith', 'A. Sundar', 'P. M. Harris', 'J. Venton'] | 2021-08-02 | null | null | null | null | ['ecg-classification'] | ['medical'] | [ 4.00318950e-01 -2.33188663e-02 6.02836370e-01 -3.94095123e-01
-8.28305185e-01 -5.53474486e-01 4.46369871e-02 9.20275152e-02
-4.49081779e-01 5.39507747e-01 2.14759100e-04 -3.37968826e-01
-9.79583412e-02 -5.87083697e-01 -5.45114934e-01 -8.40191245e-01
-2.25612655e-01 1.08016161e-02 -3.75474423e-01 -8.49431679e-02
-2.96464026e-01 5.22565126e-01 -9.57374156e-01 4.72096592e-01
6.25607312e-01 1.06307387e+00 -2.69692391e-01 1.01376319e+00
4.55822736e-01 4.95174795e-01 -1.09591377e+00 -2.57376194e-01
3.07528198e-01 -1.06882012e+00 -4.07064110e-01 -2.84606487e-01
3.10718358e-01 -2.17253923e-01 -2.83839494e-01 1.04368699e+00
1.22535372e+00 -2.54203409e-01 3.46638918e-01 -6.81884170e-01
-3.72897506e-01 6.29353523e-01 -1.54774934e-01 6.04359090e-01
1.77668422e-01 3.70154977e-01 2.77830601e-01 -6.42695069e-01
4.47295725e-01 6.11835837e-01 1.18477595e+00 3.45173150e-01
-1.49005032e+00 -6.89608872e-01 -5.27903020e-01 4.44566496e-02
-1.27694571e+00 -3.45590889e-01 7.25720525e-01 -4.06059146e-01
6.13295734e-01 4.37316775e-01 9.14145172e-01 1.04565024e+00
5.81491053e-01 3.63608822e-02 1.17598653e+00 -2.74796247e-01
1.99441984e-01 5.17888106e-02 1.00445651e-01 1.98150173e-01
3.56220901e-01 1.65383190e-01 -1.42246544e-01 -3.14920306e-01
8.01506639e-01 -1.94864407e-01 -3.49074125e-01 2.96807915e-01
-1.29273701e+00 5.74933767e-01 4.05269533e-01 5.02390981e-01
-6.58368647e-01 9.84741598e-02 8.33102942e-01 7.11898565e-01
3.29325169e-01 8.84268701e-01 -3.54780644e-01 -2.54557610e-01
-8.86192203e-01 4.56722379e-02 5.05386829e-01 1.64205939e-01
1.24714570e-02 4.45350677e-01 -1.24463618e-01 7.45194495e-01
-2.23769873e-01 5.17288983e-01 7.03616381e-01 -9.17291462e-01
9.61048752e-02 2.49701053e-01 -1.75128520e-01 -1.11461103e+00
-7.61974275e-01 -1.03069174e+00 -1.14500690e+00 2.34087948e-02
5.34805596e-01 -5.46852350e-01 -8.67576897e-01 1.47848594e+00
-2.07333546e-03 2.24425778e-01 3.53269994e-01 9.72010076e-01
9.95708942e-01 3.08502734e-01 -8.15442875e-02 -2.07418919e-01
1.21302974e+00 -6.26403689e-02 -8.09068561e-01 -6.05670661e-02
6.79763138e-01 -6.87807620e-01 9.16322529e-01 6.16581917e-01
-1.03166246e+00 -8.62652779e-01 -1.20108867e+00 3.62513632e-01
5.00924625e-02 2.03412384e-01 3.66245694e-02 9.58762705e-01
-9.26930189e-01 1.08194327e+00 -1.01799881e+00 -2.72649229e-01
5.33030570e-01 4.22985375e-01 -3.97260755e-01 4.95574586e-02
-1.31326568e+00 9.51736152e-01 6.54467121e-02 3.31480294e-01
-6.23765945e-01 -5.67206025e-01 -6.62609637e-01 1.55120492e-01
-3.51937175e-01 -5.13633311e-01 8.12173545e-01 -1.04770124e+00
-1.20869768e+00 8.38399708e-01 2.79949278e-01 -7.65594065e-01
5.89943290e-01 -2.74429798e-01 -7.43022442e-01 3.49687427e-01
-2.91675441e-02 1.50946140e-01 7.89245903e-01 -8.04363191e-01
-7.73630664e-02 -3.22828025e-01 -3.98728460e-01 -1.24135818e-02
2.88997311e-02 3.29107381e-02 1.39346331e-01 -6.60508931e-01
3.57943475e-01 -1.02626371e+00 -1.27525061e-01 -2.40880653e-01
-1.27599433e-01 3.47803324e-01 4.42890108e-01 -9.06891048e-01
1.05085897e+00 -2.49291754e+00 -1.69303104e-01 5.70969164e-01
3.69511038e-01 3.85331899e-01 -1.41026750e-01 1.89751368e-02
-4.54816103e-01 2.39831358e-01 -2.98492908e-01 2.44530514e-01
-5.70317924e-01 5.47996126e-02 -7.13349283e-02 7.75960803e-01
3.46696168e-01 8.62561584e-01 -7.41099179e-01 8.25596377e-02
3.07814747e-01 7.71398604e-01 -2.78582752e-01 1.10398553e-01
5.43623745e-01 8.57618511e-01 6.74075410e-02 1.45297274e-01
5.93534172e-01 -9.93392915e-02 2.51850039e-01 -3.74280870e-01
2.10677579e-01 2.32111663e-01 -9.66574788e-01 1.39833581e+00
-2.02606037e-01 8.84187162e-01 -2.38715380e-01 -1.10016572e+00
1.10633385e+00 7.36027300e-01 6.19241118e-01 -6.43555045e-01
2.57817686e-01 3.48697603e-01 8.96198452e-01 -6.05093837e-01
-2.99942911e-01 -5.48073292e-01 7.93271586e-02 4.62869674e-01
-4.01207842e-02 -2.29225472e-01 -2.76440084e-01 -1.39174864e-01
1.29168963e+00 -2.49484748e-01 1.05031822e-02 -3.53791982e-01
2.82899767e-01 -2.85470963e-01 4.70959485e-01 8.02750111e-01
-1.79284260e-01 1.21637142e+00 7.08792865e-01 -9.58810389e-01
-8.38059962e-01 -9.63358462e-01 -5.17180204e-01 3.11311156e-01
-2.65715837e-01 -3.03692430e-01 -9.53440130e-01 -5.69533348e-01
-2.01632172e-01 4.44065183e-01 -6.42177522e-01 -7.03796625e-01
-6.21744692e-01 -1.12275338e+00 1.30050051e+00 6.51936054e-01
4.10059333e-01 -1.22834384e+00 -1.03870094e+00 3.31264794e-01
-3.36522818e-01 -9.27900076e-01 -1.07592773e-02 4.25771505e-01
-1.04155946e+00 -1.04532969e+00 -8.62687528e-01 -6.25900805e-01
5.79454780e-01 -4.46455985e-01 8.26873481e-01 2.40757912e-01
-4.87381965e-01 1.95833668e-01 -1.25002831e-01 -6.95827961e-01
-6.66171193e-01 -9.51722637e-02 1.77772522e-01 -5.89532778e-03
2.01307386e-01 -5.20162880e-01 -7.81538546e-01 1.13059849e-01
-7.12626040e-01 -3.63193840e-01 3.73473942e-01 1.08430254e+00
5.30817926e-01 -4.01434153e-02 9.30050015e-01 -9.18629289e-01
9.37987149e-01 -2.34486163e-01 -2.96967197e-02 -2.79940069e-01
-6.09795928e-01 -1.38906226e-01 6.81558311e-01 -5.09516239e-01
-4.57684070e-01 4.55528386e-02 -4.72540021e-01 -4.59581196e-01
-2.35285997e-01 6.52420878e-01 2.40510896e-01 1.08583763e-01
1.20914876e+00 -5.54681756e-02 2.20348373e-01 -3.01667124e-01
-2.23588169e-01 5.47568321e-01 7.06389427e-01 -1.19761266e-01
4.38418239e-01 3.05763334e-01 4.87792604e-02 -8.60563993e-01
-4.92234945e-01 -1.26534238e-01 -5.33124328e-01 -2.01767370e-01
8.87938797e-01 -6.84585989e-01 -1.61361709e-01 4.25276190e-01
-7.94113278e-01 -3.41348588e-01 -5.54532766e-01 7.24200010e-01
-3.59948516e-01 2.57005960e-01 -5.85664690e-01 -5.02001882e-01
-7.81862497e-01 -1.09454811e+00 4.94697779e-01 -1.59951225e-01
-7.94638693e-01 -8.36417317e-01 1.18703395e-01 2.49043293e-02
6.19133472e-01 9.64193165e-01 1.00053120e+00 -8.45339894e-01
4.65050429e-01 -5.17328620e-01 1.67252466e-01 8.31308305e-01
5.42476833e-01 -2.82074392e-01 -1.22956705e+00 -3.80220264e-01
5.42098582e-01 -9.35656279e-02 6.09959185e-01 4.75552201e-01
1.08075941e+00 -1.04763471e-01 7.08179846e-02 6.88759089e-01
9.59170401e-01 5.97379506e-01 9.83892798e-01 1.20539606e-01
4.51890796e-01 3.06026995e-01 -6.47297278e-02 1.59854516e-02
-2.49993250e-01 2.22585291e-01 -6.67168051e-02 -7.04365849e-01
-1.83040842e-01 1.78218052e-01 2.11228207e-01 6.79944038e-01
-2.27554753e-01 2.60959953e-01 -1.03192019e+00 3.48678112e-01
-1.29220498e+00 -8.97150159e-01 -4.01310295e-01 2.15770650e+00
6.80032313e-01 2.81253636e-01 2.51074135e-02 8.12716544e-01
6.42975569e-01 -2.86296755e-01 -7.01462805e-01 -5.27794600e-01
-3.30280453e-01 7.55840898e-01 7.35066307e-04 2.41880938e-02
-1.03366244e+00 1.21438913e-01 6.50619602e+00 -1.30164072e-01
-1.85766459e+00 -3.22554074e-02 9.16028142e-01 -7.91283771e-02
3.24520588e-01 -4.10732239e-01 2.28111267e-01 4.19640124e-01
1.42490244e+00 6.16546534e-02 1.56333745e-01 4.55370784e-01
3.25767398e-01 7.43655711e-02 -9.79973972e-01 1.29043198e+00
1.03293527e-02 -1.02622902e+00 -4.58670199e-01 -1.63752854e-01
5.77527642e-01 7.56561989e-03 2.00394168e-01 1.01337567e-01
-5.24552703e-01 -1.38109004e+00 4.06344712e-01 6.16714895e-01
1.15765357e+00 -8.39019835e-01 1.28751862e+00 6.39158562e-02
-6.02051616e-01 6.18707687e-02 -2.80204654e-01 -1.60864785e-01
-1.12593852e-01 7.64099121e-01 -8.80397618e-01 2.20545948e-01
9.28532243e-01 6.49327159e-01 -5.57672739e-01 9.15691257e-01
-6.91380203e-02 1.13610375e+00 -2.47543067e-01 5.50233126e-01
-1.41069442e-01 9.48477536e-03 4.37217563e-01 1.23336661e+00
3.67347866e-01 1.92199618e-01 -1.73182085e-01 6.78989172e-01
-2.25831699e-02 -5.63298613e-02 -6.05364203e-01 1.26383081e-01
-8.59327242e-02 1.13278258e+00 -8.42616200e-01 -4.35030580e-01
-1.54416218e-01 7.64532149e-01 -3.61850768e-01 4.53562737e-01
-7.14191079e-01 -8.09140444e-01 3.00512135e-01 4.84219104e-01
-1.42736107e-01 3.92475009e-01 -8.30965579e-01 -8.71476710e-01
9.84100550e-02 -1.24730086e+00 3.99758309e-01 -6.46532059e-01
-1.06564891e+00 8.97984743e-01 -2.42524892e-01 -1.33190870e+00
-2.40654871e-01 -3.35008174e-01 -8.58472228e-01 1.15477347e+00
-9.45001245e-01 -3.17186087e-01 -3.65461618e-01 4.52114403e-01
1.48306936e-01 -6.34284168e-02 1.21190989e+00 3.92912656e-01
-2.57932097e-01 5.38822830e-01 -2.65709199e-02 5.65757930e-01
7.83870578e-01 -1.12880111e+00 5.09163320e-01 7.08864152e-01
1.27971023e-01 7.90691614e-01 6.60635710e-01 -4.83974725e-01
-8.28183115e-01 -1.17571628e+00 6.25847816e-01 -3.36152673e-01
-7.95062557e-02 5.82315065e-02 -1.20795095e+00 3.72448206e-01
9.32790861e-02 3.80567461e-01 9.23783481e-01 -2.12814689e-01
5.55397198e-03 -2.70908117e-01 -1.33123028e+00 2.39083007e-01
3.97974521e-01 -5.92371464e-01 -7.03918457e-01 6.12865910e-02
1.68603845e-02 -5.28532147e-01 -1.23818266e+00 6.62366152e-01
9.70319986e-01 -7.76724041e-01 8.02180588e-01 -4.95060891e-01
3.10207039e-01 -2.55523622e-01 2.13817000e-01 -1.57870722e+00
-2.76375651e-01 -4.08736646e-01 2.49556869e-01 5.26153922e-01
5.68888485e-01 -7.79816508e-01 4.85185504e-01 4.16833699e-01
-1.02360301e-01 -7.48495758e-01 -8.79328668e-01 -4.40595180e-01
1.47732735e-01 -4.38871175e-01 2.69056648e-01 9.40195560e-01
8.37189481e-02 1.77994356e-01 -1.78197846e-01 6.52472749e-02
2.59016663e-01 -2.85279036e-01 2.90982425e-01 -1.26842678e+00
-1.69060454e-01 2.06276271e-02 -6.54439747e-01 -2.88575832e-02
-3.74843329e-01 -1.06909013e+00 -1.33793950e-01 -1.45392644e+00
-1.85585439e-01 -3.09888572e-01 -7.92981386e-01 3.97065222e-01
-2.05282867e-01 7.38684118e-01 2.05005050e-01 1.08453505e-01
1.58442393e-01 -9.52657536e-02 9.06127751e-01 -4.02481444e-02
-5.02939939e-01 2.86359876e-01 -7.57218659e-01 7.40827620e-01
1.25245392e+00 -7.95684397e-01 -3.39913607e-01 -1.68299004e-01
5.93965799e-02 1.64755106e-01 4.04199928e-01 -1.38552713e+00
-2.04110831e-01 6.00217998e-01 1.22092891e+00 -1.48100153e-01
2.09877595e-01 -5.78281701e-01 6.01550460e-01 1.01343191e+00
-5.47792614e-01 3.68273258e-01 6.10442340e-01 8.92404839e-02
-1.23054467e-01 1.50705308e-01 9.44637597e-01 -2.30992660e-01
2.02427685e-01 -2.56675392e-01 -7.63977170e-01 1.88330919e-01
6.32119238e-01 -3.51256043e-01 2.11652055e-01 -5.32726526e-01
-1.20342159e+00 -3.83371979e-01 6.64238408e-02 1.62881672e-01
8.86678636e-01 -1.16744697e+00 -9.19969559e-01 6.18571579e-01
-1.24206632e-01 -9.22660232e-02 2.58358419e-01 1.18813622e+00
-7.71907508e-01 1.45854205e-01 -4.77351427e-01 -8.36569130e-01
-1.36719322e+00 1.08208112e-01 7.98181295e-01 9.03176516e-03
-1.04070103e+00 5.47174931e-01 -1.19028330e-01 -1.67411983e-01
1.66905001e-01 -6.63652182e-01 -2.09833264e-01 -6.03496246e-02
5.90581298e-01 4.78169203e-01 6.12343371e-01 -4.80388224e-01
-3.81609440e-01 5.70950329e-01 1.88433975e-01 -9.87737924e-02
1.41241646e+00 2.81740516e-01 1.48426294e-01 4.97752100e-01
1.07976973e+00 -2.10015863e-01 -6.82434738e-01 1.53279886e-01
-2.24411830e-01 -4.97542843e-02 5.29507063e-02 -1.08850026e+00
-1.24865079e+00 1.16252470e+00 1.35892510e+00 3.54141384e-01
1.30527520e+00 -4.65122819e-01 4.61818933e-01 2.97817349e-01
1.80996321e-02 -7.21030653e-01 -1.21565849e-01 1.79348126e-01
9.57556725e-01 -7.67986774e-01 -9.50277150e-02 8.13187212e-02
-8.59154344e-01 1.28168142e+00 1.79767475e-01 -4.40460265e-01
6.33340180e-01 3.77918571e-01 6.37916863e-01 -2.45979831e-01
-2.43913472e-01 3.35477948e-01 2.49108553e-01 7.21585929e-01
6.73676312e-01 -4.37663542e-03 -2.78808177e-01 8.12534630e-01
-4.35913116e-01 1.27271861e-01 6.43712997e-01 8.46071422e-01
9.83922854e-02 -7.76209950e-01 -5.76654792e-01 9.56914544e-01
-9.76411283e-01 -9.44039747e-02 -3.86632085e-01 4.56154495e-01
3.11388016e-01 8.19571912e-01 7.67743662e-02 -5.12037337e-01
4.94933218e-01 4.56645101e-01 2.52405465e-01 -7.16555059e-01
-1.11603677e+00 1.72773868e-01 -4.96696346e-02 -2.87102699e-01
-1.70432016e-01 -5.59098125e-01 -1.20465016e+00 2.81427264e-01
-3.07799369e-01 1.19447187e-01 5.38238168e-01 6.43371344e-01
3.28374177e-01 8.56320441e-01 4.00695980e-01 -5.60151815e-01
-6.08714461e-01 -1.08560967e+00 -5.24758160e-01 5.57383239e-01
6.48265123e-01 -4.77283597e-02 -4.38814700e-01 3.11286598e-01] | [14.352516174316406, 3.2539567947387695] |
464e600d-1e2d-48da-8533-2db3d73c5c6d | cold-concurrent-loads-disaggregator-for-non | 2106.02352 | null | https://arxiv.org/abs/2106.02352v1 | https://arxiv.org/pdf/2106.02352v1.pdf | COLD: Concurrent Loads Disaggregator for Non-Intrusive Load Monitoring | The modern artificial intelligence techniques show the outstanding performances in the field of Non-Intrusive Load Monitoring (NILM). However, the problem related to the identification of a large number of appliances working simultaneously is underestimated. One of the reasons is the absence of a specific data. In this research we propose the Synthesizer of Normalized Signatures (SNS) algorithm to simulate the aggregated consumption with up to 10 concurrent loads. The results show that the synthetic data provides the models with at least as a powerful identification accuracy as the real-world measurements. We have developed the neural architecture named Concurrent Loads Disaggregator (COLD) which is relatively simple and easy to understand in comparison to the previous approaches. Our model allows identifying from 1 to 10 appliances working simultaneously with mean F1-score 78.95%. The source code of the experiments performed is available at https://github.com/arx7ti/cold-nilm. | ['Elena Gryazina', 'Dmitrii Kriukov', 'Ilia Kamyshev'] | 2021-06-04 | null | null | null | null | ['non-intrusive-load-monitoring', 'non-intrusive-load-monitoring', 'non-intrusive-load-monitoring'] | ['knowledge-base', 'miscellaneous', 'time-series'] | [-6.54915646e-02 -1.77328855e-01 -1.96565047e-01 -5.35464466e-01
-3.38679612e-01 -3.62184912e-01 4.41998214e-01 -9.06875283e-02
-7.42271319e-02 7.17730641e-01 -4.97155637e-02 -6.49916530e-02
-2.48164311e-01 -8.00929904e-01 -4.91682202e-01 -8.01944673e-01
-1.78290635e-01 7.08817303e-01 -1.57761768e-01 -5.36416955e-02
-1.36193231e-01 3.39377314e-01 -1.93513668e+00 7.96500593e-03
8.43916297e-01 1.08691192e+00 1.40355676e-01 3.61625314e-01
9.94472131e-02 7.99579263e-01 -9.45517719e-01 -1.73784152e-01
3.91471148e-01 -3.00136715e-01 -5.28697431e-01 -1.70578986e-01
6.17725179e-02 -5.61573923e-01 -9.43477228e-02 1.24922693e+00
5.27992427e-01 8.94812644e-02 4.42257196e-01 -1.93528044e+00
-4.63600636e-01 1.19363606e+00 -5.63667536e-01 1.95346355e-01
4.45842385e-01 -4.46852781e-02 6.32856369e-01 -1.61687322e-02
1.00821823e-01 9.41446722e-01 5.14494181e-01 2.06578821e-01
-1.19225073e+00 -1.04089487e+00 -2.72960868e-02 4.82067645e-01
-1.60492420e+00 -2.41653726e-01 7.87001908e-01 -1.73528895e-01
8.71556997e-01 7.44304597e-01 2.02640250e-01 1.38994718e+00
-3.11695337e-01 8.09989750e-01 1.13719165e+00 -4.28550273e-01
3.27218950e-01 4.59221095e-01 5.96038282e-01 1.30934566e-01
6.20657444e-01 6.37739077e-02 6.29000291e-02 -2.26603836e-01
6.51179180e-02 1.07027695e-01 -8.94428864e-02 3.65577102e-01
-6.39549792e-01 8.09309840e-01 2.79322412e-04 7.20108688e-01
-4.85749990e-01 -1.82309970e-01 7.33620822e-01 2.48881653e-01
4.41378355e-01 3.29929799e-01 -6.16462171e-01 -1.90459117e-01
-9.99241769e-01 3.70383891e-03 1.02691126e+00 9.33175087e-01
4.61640567e-01 4.94758487e-01 2.63774037e-01 6.49190605e-01
2.42296327e-02 6.50969684e-01 8.29657972e-01 -7.75535047e-01
1.58643275e-01 7.97638834e-01 1.21968105e-01 -9.45045531e-01
-7.34571576e-01 -3.36596519e-01 -1.11242545e+00 1.69762224e-01
1.86593220e-01 -2.54005045e-01 -3.16612571e-01 1.92664003e+00
2.39539370e-01 2.77646780e-01 -1.68920055e-01 4.83184814e-01
4.88773316e-01 8.19872141e-01 2.46285364e-01 -4.06504273e-01
1.35893762e+00 -6.19322598e-01 -1.11313009e+00 3.54542822e-01
3.53497624e-01 -6.30683601e-01 9.43937778e-01 5.20673156e-01
-7.28752673e-01 -5.36436439e-01 -9.88329530e-01 5.41844189e-01
-8.49925220e-01 4.20448035e-02 5.85168302e-01 1.01091480e+00
-6.48600996e-01 5.51445544e-01 -6.77032411e-01 -6.06487751e-01
1.47769719e-01 4.78583574e-01 -9.26723182e-02 5.13684392e-01
-1.19566882e+00 8.01013529e-01 9.66337979e-01 -4.07674760e-01
-5.14374614e-01 -5.10213614e-01 -5.10068476e-01 3.83495465e-02
2.42888242e-01 -1.90126717e-01 1.25827837e+00 -1.00933290e+00
-1.30811942e+00 4.52172935e-01 1.43617108e-01 -5.11001408e-01
6.13687634e-01 -7.75114894e-02 -1.01403069e+00 -2.83686191e-01
2.45768875e-02 3.33145298e-02 4.21856314e-01 -1.13396001e+00
-8.45490873e-01 -4.78388101e-01 -1.79680154e-01 -4.25347507e-01
-6.35865808e-01 1.05225429e-01 1.28753230e-01 -5.72298050e-01
-4.10896212e-01 -7.99095690e-01 3.76753241e-01 -8.95983577e-01
-7.80909181e-01 -3.53704721e-01 1.16551852e+00 -7.91027188e-01
1.51401365e+00 -1.97880125e+00 -4.69476074e-01 2.59974450e-01
-2.47388899e-01 4.29632902e-01 8.13086778e-02 4.99761283e-01
-4.67036933e-01 -4.17519659e-02 -7.20289052e-02 -5.35390496e-01
5.89563608e-01 2.22701922e-01 -1.86602578e-01 3.94641042e-01
-4.29392934e-01 6.88371360e-01 -7.00659931e-01 -1.39130503e-01
6.38490081e-01 2.90029943e-01 1.76618472e-01 4.65904325e-01
-2.72975206e-01 2.51535684e-01 -3.20257097e-01 7.75651157e-01
8.97976518e-01 -1.68253824e-01 2.68598169e-01 -4.72599387e-01
-4.30067964e-02 -6.09839782e-02 -1.62405264e+00 1.19292235e+00
-3.86185080e-01 2.40334705e-01 -1.79848701e-01 -1.08473337e+00
7.91937113e-01 6.52127743e-01 7.36328900e-01 -9.34165776e-01
6.23300612e-01 1.33794889e-01 -3.05779397e-01 -6.44808292e-01
1.48872510e-01 3.74459922e-01 -1.85742021e-01 7.55522013e-01
-3.58878374e-02 4.20538723e-01 6.00402594e-01 -2.42391512e-01
1.08146441e+00 -1.74856007e-01 6.71312571e-01 -2.85077810e-01
5.43488979e-01 -3.37995589e-01 6.10713243e-01 6.33127928e-01
-3.43322486e-01 -9.57674608e-02 4.08964343e-02 -5.06748021e-01
-1.06406391e+00 -8.69133294e-01 -1.46498978e-01 1.11860991e+00
-1.72736391e-01 -1.12819754e-01 -1.00692165e+00 -3.29026729e-01
1.12015747e-01 1.41972816e+00 -4.05424029e-01 1.99643313e-03
-3.69508117e-01 -1.19320965e+00 8.01565289e-01 3.69449735e-01
6.04992092e-01 -1.16402686e+00 -6.69964731e-01 1.09650239e-01
-2.50141412e-01 -1.30605006e+00 -1.23287991e-01 4.22048897e-01
-4.78102058e-01 -1.09456158e+00 -1.09899819e-01 -4.68428046e-01
2.45176047e-01 -9.92215127e-02 1.26797616e+00 -2.52442211e-01
-2.80748844e-01 -1.00700937e-01 -3.80872101e-01 -7.56559670e-01
-7.51654923e-01 2.94924617e-01 4.90156472e-01 9.51487273e-02
1.08689022e+00 -1.15457428e+00 -3.11099201e-01 4.78981256e-01
-8.88586283e-01 -2.42565021e-01 3.31861049e-01 2.07671717e-01
1.74979806e-01 6.22865558e-01 8.99969876e-01 -9.16868687e-01
7.30622411e-01 -8.27592313e-01 -1.01650715e+00 6.49086758e-02
-9.85163748e-01 -3.00118297e-01 9.94879961e-01 -6.91512406e-01
-7.71224678e-01 -1.84943695e-02 -1.91047028e-01 -3.38995129e-01
-9.65833485e-01 -2.45256588e-01 -3.74457359e-01 4.63476241e-01
5.37504196e-01 2.01093897e-01 -2.99834758e-01 -8.01347733e-01
1.31164849e-01 9.72637951e-01 4.07842010e-01 -5.17157435e-01
7.11563766e-01 1.83506981e-01 -1.75449029e-01 -7.41942286e-01
-4.15363401e-01 -3.99527490e-01 -4.03117061e-01 -1.13334721e-02
5.80526173e-01 -5.79684675e-01 -1.41149569e+00 7.07361937e-01
-8.60572517e-01 -2.41298139e-01 -4.45867985e-01 2.59093553e-01
-3.60813051e-01 4.06080097e-01 -6.07320249e-01 -1.25752294e+00
-7.33905017e-01 -1.16963971e+00 7.25603104e-01 3.38064909e-01
-5.82415104e-01 -7.15616345e-01 -9.46643203e-02 2.23613441e-01
5.80920577e-01 5.85925817e-01 8.66032004e-01 -1.14727652e+00
-2.78915375e-01 -5.05770862e-01 4.81862016e-02 4.34313953e-01
3.80838454e-01 1.00468434e-02 -1.32383490e+00 -2.78984845e-01
2.35851005e-01 -1.02717489e-01 4.60905358e-02 2.10056141e-01
1.24750257e+00 -4.74234164e-01 -1.44835725e-01 4.19849277e-01
1.78472865e+00 7.83723354e-01 5.34925997e-01 4.17747974e-01
2.87284315e-01 3.26188117e-01 9.43281054e-02 7.40063787e-01
3.70068520e-01 7.15872765e-01 5.90523064e-01 9.80394855e-02
3.19750845e-01 -2.02632621e-01 2.70589560e-01 9.76471126e-01
1.67456463e-01 -5.23271322e-01 -5.01742661e-01 4.22511935e-01
-1.85467124e+00 -1.10416257e+00 -8.51783380e-02 2.06842494e+00
4.36291009e-01 1.62095636e-01 7.09207177e-01 7.74705291e-01
8.16416919e-01 -6.46059439e-02 -6.05068266e-01 -5.79380751e-01
7.91772828e-02 4.80491146e-02 6.44038975e-01 3.08089048e-01
-1.10396242e+00 1.85694471e-01 6.10882711e+00 9.54921067e-01
-7.38605082e-01 3.30255657e-01 6.17545843e-01 -4.25291173e-02
2.88390547e-01 -6.75123274e-01 -7.75007546e-01 1.26947761e+00
1.49451661e+00 -1.90847814e-01 7.95157731e-01 1.03574908e+00
5.39802909e-01 -1.57856405e-01 -1.09505796e+00 1.07104182e+00
-8.19671806e-03 -5.63297927e-01 4.82676774e-02 1.09731436e-01
7.28468299e-01 1.55935943e-01 -1.41075686e-01 2.26393387e-01
5.61296403e-01 -6.52196884e-01 4.68352377e-01 6.17497563e-01
4.25867021e-01 -8.23737681e-01 1.05077851e+00 7.19956875e-01
-1.10340798e+00 -5.33530891e-01 -2.08119918e-02 -5.01350798e-02
7.65022263e-02 7.98883915e-01 -6.26760602e-01 7.24103689e-01
8.10703456e-01 7.51129314e-02 -3.98790449e-01 6.99771702e-01
4.04925086e-02 8.01344156e-01 -9.25557733e-01 -1.12256706e-01
-1.89553306e-01 -3.66123587e-01 2.40627006e-01 1.08158088e+00
4.23374504e-01 -2.83832341e-01 2.21477672e-01 9.13031399e-01
-1.54310362e-02 1.81854725e-01 -5.83244026e-01 1.40339360e-01
7.32221186e-01 1.33418298e+00 -5.17585278e-01 -4.01940286e-01
-3.98646772e-01 8.12694252e-01 -1.66577965e-01 1.47028416e-01
-9.11026001e-01 -3.15708309e-01 5.38871467e-01 -5.90562858e-02
-4.79211062e-02 2.54539341e-01 -2.75379628e-01 -8.96903455e-01
5.92411235e-02 -1.09300077e+00 5.83210230e-01 -5.51084757e-01
-1.58623576e+00 5.89942515e-01 2.60009646e-01 -9.73844588e-01
-4.90224242e-01 -5.24738014e-01 -8.08324337e-01 6.43843114e-01
-1.09446752e+00 -1.23616731e+00 -5.38907349e-01 5.52219748e-01
4.38236564e-01 -3.53018761e-01 1.27153635e+00 7.85759509e-01
-1.04213679e+00 7.20061481e-01 5.46055377e-01 1.09967723e-01
2.80151337e-01 -1.37397981e+00 4.60877359e-01 7.00271130e-01
-6.29881173e-02 2.44777292e-01 9.76630092e-01 -4.30812389e-01
-1.12476373e+00 -9.95487750e-01 6.36497259e-01 -3.69066268e-01
6.33155525e-01 -3.62527221e-01 -7.45979071e-01 7.49916136e-01
3.13527226e-01 -4.36603785e-01 8.49671125e-01 -6.32755905e-02
-2.24681586e-01 -5.26349664e-01 -1.63178551e+00 1.32368684e-01
5.62579989e-01 -2.73810804e-01 -3.74574363e-01 4.69699025e-01
4.24206674e-01 1.96809381e-01 -1.01977122e+00 2.30583429e-01
4.02612627e-01 -1.20155191e+00 7.22202897e-01 -1.31236017e-01
-3.28132659e-01 -4.11303550e-01 -6.50363863e-01 -1.13108873e+00
-2.36718103e-01 -5.14917135e-01 -6.07673705e-01 1.66690838e+00
1.47509009e-01 -1.05541337e+00 4.73082662e-01 5.99928677e-01
4.27142411e-01 -3.76460254e-01 -7.37924993e-01 -9.52294946e-01
-3.37355316e-01 -5.42432070e-01 1.42188370e+00 1.14020979e+00
-5.26204370e-02 8.46648738e-02 -3.72178078e-01 3.19024980e-01
8.54065597e-01 6.16395548e-02 6.21846497e-01 -1.42452908e+00
-3.78974438e-01 -3.42100412e-01 -4.30632591e-01 -2.85704732e-01
1.93947762e-01 -6.60988331e-01 -5.05802810e-01 -9.94755149e-01
1.39724270e-01 -1.42128810e-01 -6.83433056e-01 4.96288538e-01
4.03927177e-01 3.57202441e-01 1.67990327e-01 -4.92050461e-02
-3.38043213e-01 8.34554285e-02 3.42296004e-01 -2.17867345e-01
-5.26993610e-02 2.87375867e-01 -6.23434424e-01 8.53553355e-01
1.64452839e+00 -3.44999701e-01 -4.84934747e-01 -4.29401994e-02
-2.22695455e-01 -4.54947352e-01 2.47026719e-02 -1.21682429e+00
-2.23985463e-02 7.22751543e-02 4.82772738e-01 -7.61756897e-01
2.65336245e-01 -1.32031941e+00 8.68500531e-01 5.16700864e-01
2.99006868e-02 4.84743267e-01 1.34967268e-01 8.81756619e-02
1.51850849e-01 -3.96486551e-01 6.97428584e-01 -3.08333069e-01
-7.30706930e-01 6.34578168e-02 -1.75572246e-01 -4.10800815e-01
1.15697956e+00 3.43831629e-02 -4.21924561e-01 -4.54704940e-01
-5.67391753e-01 1.60570189e-01 3.31978321e-01 5.23979127e-01
-2.56127536e-01 -1.31785381e+00 -4.19164062e-01 2.29617804e-01
-1.68898553e-02 -5.55275023e-01 9.70473513e-02 4.25368488e-01
-2.73645610e-01 5.63923538e-01 -1.85576797e-01 -3.47820550e-01
-1.12447667e+00 9.37157571e-01 4.32297468e-01 -3.32204282e-01
-4.18551505e-01 -2.26567574e-02 -3.25625211e-01 -4.15250599e-01
5.67707479e-01 -1.85727268e-01 -2.32131481e-01 1.44694805e-01
7.81281710e-01 1.10813677e+00 1.68817863e-01 -7.77434170e-01
-2.31565863e-01 2.75859058e-01 1.66936249e-01 5.71361840e-01
1.47896862e+00 -1.98177218e-01 -1.83248699e-01 7.15417624e-01
1.29828691e+00 -4.31591809e-01 -4.04904008e-01 2.58084927e-02
1.12840511e-01 -2.43047655e-01 -1.12501107e-01 -1.14605975e+00
-1.12398684e+00 3.92047554e-01 1.32595575e+00 9.11574185e-01
1.38703084e+00 -4.65643138e-01 9.14252520e-01 3.36913258e-01
6.22219503e-01 -1.21023595e+00 -5.83370805e-01 -5.22515960e-02
3.87888044e-01 -1.17911935e+00 -1.30253360e-01 -6.31299987e-02
-1.51081741e-01 8.03431928e-01 4.85192806e-01 5.46565801e-02
7.36586094e-01 6.89058781e-01 1.97526455e-01 -1.55083379e-02
-4.38419610e-01 -1.67624667e-01 -3.35947692e-01 6.96874022e-01
2.08542347e-01 6.21401668e-01 -2.67712027e-01 5.72085917e-01
-3.51126164e-01 1.60503283e-01 3.46863180e-01 5.48725843e-01
-8.26090947e-02 -1.13384378e+00 -5.13405919e-01 6.09403908e-01
-7.09876597e-01 4.67650801e-01 8.64123553e-03 7.82598972e-01
4.04211640e-01 1.35508513e+00 5.26043847e-02 -6.01070225e-01
4.65311468e-01 5.00979543e-01 2.37838225e-03 1.03608645e-01
-5.66056967e-01 -2.03802213e-01 -5.09397686e-03 -5.61941147e-01
-3.97384524e-01 -5.15691936e-01 -7.76157260e-01 -8.47111285e-01
-2.87556291e-01 1.73893139e-01 8.52257729e-01 6.20723963e-01
2.97935873e-01 6.11338139e-01 9.92753386e-01 -7.19234884e-01
-7.01035738e-01 -1.36503279e+00 -9.77922440e-01 1.02736878e+00
5.32855392e-02 -6.03342891e-01 -5.06027877e-01 -9.32375863e-02] | [6.006244659423828, 2.5967018604278564] |
c99e501c-2df1-477f-97e2-b98a4edc43ca | fmas-fast-multi-objective-supernet | 2303.16322 | null | https://arxiv.org/abs/2303.16322v1 | https://arxiv.org/pdf/2303.16322v1.pdf | FMAS: Fast Multi-Objective SuperNet Architecture Search for Semantic Segmentation | We present FMAS, a fast multi-objective neural architecture search framework for semantic segmentation. FMAS subsamples the structure and pre-trained parameters of DeepLabV3+, without fine-tuning, dramatically reducing training time during search. To further reduce candidate evaluation time, we use a subset of the validation dataset during the search. Only the final, Pareto non-dominated, candidates are ultimately fine-tuned using the complete training set. We evaluate FMAS by searching for models that effectively trade accuracy and computational cost on the PASCAL VOC 2012 dataset. FMAS finds competitive designs quickly, e.g., taking just 0.5 GPU days to discover a DeepLabV3+ variant that reduces FLOPs and parameters by 10$\%$ and 20$\%$ respectively, for less than 3$\%$ increased error. We also search on an edge device called GAP8 and use its latency as the metric. FMAS is capable of finding 2.2$\times$ faster network with 7.61$\%$ MIoU loss. | ['Brett H. Meyer', 'Warren J. Gross', 'Olivier Therrien', 'Marihan Amein', 'Zhuoran Xiong'] | 2023-03-28 | null | null | null | null | ['architecture-search'] | ['methodology'] | [ 3.73518988e-02 -5.78755438e-02 -1.55628040e-01 -4.59698588e-01
-8.88743341e-01 -6.73884094e-01 -3.42994720e-01 1.64232165e-01
-8.79427731e-01 7.64425457e-01 -6.76896036e-01 -4.40321118e-01
-3.89263660e-01 -7.40510106e-01 -9.92214799e-01 -3.90576690e-01
1.17773704e-01 6.16450369e-01 3.54523510e-01 2.58334786e-01
2.95609862e-01 3.87921393e-01 -1.66396403e+00 2.07447380e-01
7.66761124e-01 1.60097158e+00 2.96475917e-01 6.29848421e-01
2.00542547e-02 2.26634577e-01 -8.23165178e-01 -4.08378869e-01
5.06463885e-01 -6.73847571e-02 -9.71965313e-01 -3.87173653e-01
7.88294613e-01 -8.20247307e-02 -9.46152136e-02 9.32592154e-01
6.67870641e-01 2.36477911e-01 5.43211699e-02 -1.14806592e+00
3.36828157e-02 8.56270194e-01 -5.90355635e-01 4.26364332e-01
-5.90073287e-01 2.42276549e-01 1.03382230e+00 -6.07366025e-01
5.39836824e-01 9.83903289e-01 6.74577296e-01 5.58503330e-01
-1.17107713e+00 -1.02603245e+00 8.66701230e-02 1.64183795e-01
-1.71485186e+00 -5.72589695e-01 4.49910462e-01 -5.11251092e-02
1.48125935e+00 4.50222522e-01 6.60163164e-01 6.43481672e-01
-5.84102385e-02 8.62815678e-01 8.09363663e-01 -1.85370415e-01
4.90594894e-01 -1.06345952e-01 4.86460209e-01 1.06261480e+00
3.89333665e-01 -2.58308230e-03 -5.04033327e-01 -1.20809264e-01
4.04855579e-01 -5.60139298e-01 -1.01071365e-01 1.21486954e-01
-5.10363102e-01 8.20122838e-01 4.54570442e-01 -1.97736192e-02
-1.89143032e-01 7.47068226e-01 4.99739200e-01 8.66415575e-02
1.85719043e-01 9.66554701e-01 -8.84739459e-01 -4.07766104e-01
-1.19323564e+00 2.41970569e-01 4.57922250e-01 7.78316915e-01
8.06135774e-01 2.89834648e-01 -1.72941804e-01 9.94630575e-01
6.43034950e-02 2.61320204e-01 2.75668412e-01 -1.33516216e+00
2.58620590e-01 8.05438340e-01 -1.46165013e-01 -7.94570565e-01
-5.67518592e-01 -1.09869635e+00 -3.74599814e-01 2.34167904e-01
2.66654134e-01 -4.05249834e-01 -1.13409317e+00 1.66809869e+00
2.53007293e-01 -1.40654659e-02 -2.58646965e-01 7.68322527e-01
7.55868793e-01 4.57724869e-01 -2.43968681e-01 -9.24742818e-02
1.23927283e+00 -1.31756699e+00 5.25837839e-02 -5.34852743e-01
6.60092175e-01 -6.31646335e-01 1.05288863e+00 6.29491508e-01
-1.40840161e+00 -4.38497514e-01 -1.43964636e+00 1.99031994e-01
-8.21478143e-02 2.25925758e-01 6.98110580e-01 8.99130702e-01
-1.19552958e+00 9.67116833e-01 -1.08999670e+00 2.47161999e-01
8.90973270e-01 9.09878075e-01 3.43399048e-01 1.18657179e-01
-7.14836180e-01 3.36928159e-01 5.30241489e-01 3.28990002e-03
-9.35780764e-01 -1.02886951e+00 -5.03297031e-01 3.21812570e-01
5.46848655e-01 -5.13519347e-01 1.06358862e+00 -8.45084608e-01
-1.29701734e+00 6.29122555e-01 -8.47173259e-02 -8.25579941e-01
3.19141716e-01 -1.59992620e-01 -1.95898116e-01 -2.03911979e-02
-1.14770919e-01 1.07253480e+00 5.94137192e-01 -9.35853481e-01
-7.89258957e-01 -3.03705484e-01 1.76885083e-01 -8.97460207e-02
-4.75957990e-01 -7.34744146e-02 -9.15174186e-01 -4.00805563e-01
-7.43967074e-04 -1.15658009e+00 -3.90209109e-01 -1.76265687e-02
-3.71869028e-01 -7.63584077e-02 6.70558035e-01 -2.68444598e-01
1.25874984e+00 -1.82606697e+00 -2.01725230e-01 4.52200621e-01
3.77996206e-01 5.29720664e-01 -4.03607905e-01 -5.63114703e-01
2.82488883e-01 3.86381924e-01 -1.00933373e-01 -2.92902261e-01
-4.69324738e-02 8.02363008e-02 3.09031814e-01 2.56918937e-01
-3.79307568e-02 9.90879059e-01 -3.27201873e-01 -3.85014147e-01
-1.20217673e-01 3.67204458e-01 -9.25373673e-01 -4.69905317e-01
-4.04676914e-01 -1.25699759e-01 -3.86608154e-01 7.18626618e-01
6.48788214e-01 -6.21094882e-01 6.35669827e-02 -3.13205779e-01
-8.54390115e-02 2.35427737e-01 -1.28723431e+00 1.88586283e+00
-6.58401132e-01 8.22556436e-01 2.56721050e-01 -9.81435478e-01
1.01694810e+00 -2.88061619e-01 2.48329580e-01 -1.00715613e+00
6.25231206e-01 5.99255681e-01 1.25857648e-02 8.92861560e-02
4.03698206e-01 6.12819254e-01 1.71984453e-02 3.18331838e-01
3.12919915e-02 2.05484524e-01 4.12191778e-01 -3.40798050e-01
1.37597156e+00 1.60446521e-02 -3.53731990e-01 -7.01112270e-01
1.60336643e-01 4.79167193e-01 8.42167616e-01 7.45577753e-01
-2.37435728e-01 3.09477359e-01 2.70551026e-01 -6.05437219e-01
-7.70938516e-01 -6.85219944e-01 -1.37429371e-01 1.29527724e+00
2.77432054e-01 -2.48318687e-01 -1.07856607e+00 -5.71660101e-01
-1.88192904e-01 8.14819753e-01 -4.31844085e-01 -1.55204460e-01
-8.24264824e-01 -9.18449283e-01 9.24891293e-01 6.20258689e-01
7.15514600e-01 -1.00067604e+00 -1.30813003e+00 1.76010117e-01
2.05358177e-01 -8.27653468e-01 -4.02506590e-01 6.18100941e-01
-1.00816643e+00 -9.19753313e-01 -4.28893834e-01 -9.36010122e-01
6.13723695e-01 -2.10235476e-01 1.26597297e+00 4.46178138e-01
-8.51383805e-01 -3.44941974e-01 4.70078476e-02 -2.35159010e-01
-3.36657353e-02 3.69946599e-01 -9.57348049e-02 -5.13069212e-01
1.97420776e-01 -3.47220778e-01 -8.28268051e-01 4.16252792e-01
-3.45588058e-01 -1.82276834e-02 4.07199383e-01 9.48516905e-01
1.06682444e+00 3.77675146e-01 3.25698137e-01 -5.87135077e-01
3.85177493e-01 7.73700280e-03 -1.16819072e+00 1.67501003e-01
-1.03066850e+00 2.01250747e-01 5.78498125e-01 -3.56743246e-01
-5.39796770e-01 1.73907652e-01 -5.16831018e-02 -9.14792657e-01
1.04672045e-01 3.79210591e-01 1.33712023e-01 -2.74616212e-01
9.38408494e-01 -9.80631635e-02 -3.25587004e-01 -5.80282867e-01
-5.88577464e-02 2.12910354e-01 4.16699201e-01 -5.59141278e-01
3.03663969e-01 4.67786565e-02 4.77320142e-02 -4.10591513e-01
-6.27586126e-01 -2.04658449e-01 2.03141514e-02 2.26100795e-02
5.77811182e-01 -6.78230584e-01 -1.25613272e+00 2.71726966e-01
-9.49981153e-01 -5.19708753e-01 -1.85774207e-01 3.04949522e-01
-2.05837429e-01 -3.80956382e-01 -5.02811968e-01 -5.31793237e-01
-1.03135395e+00 -1.62748361e+00 8.15508008e-01 4.87234235e-01
-3.27455938e-01 -5.60359716e-01 -4.80780721e-01 7.21611977e-01
4.67525035e-01 2.47407511e-01 8.57250988e-01 -5.24904490e-01
-6.30942166e-01 -2.10710578e-02 -4.94037181e-01 3.62078428e-01
-4.13547546e-01 -1.78678885e-01 -9.03560579e-01 -4.83781129e-01
-2.33611733e-01 -4.02682543e-01 1.02252495e+00 7.27005124e-01
1.66575789e+00 -2.16533288e-01 -4.77160394e-01 1.04291034e+00
1.69945419e+00 7.61069775e-01 1.92829654e-01 5.17385602e-01
6.22646153e-01 1.99977964e-01 5.07441223e-01 3.47180754e-01
-2.35169455e-02 5.73786318e-01 6.56875610e-01 -1.18063174e-01
-1.52284116e-01 6.52722940e-02 -1.68092504e-01 4.87193584e-01
1.06485084e-01 -4.20264959e-01 -1.20827985e+00 6.61427438e-01
-1.50008976e+00 -2.85980254e-01 1.41328752e-01 1.96629202e+00
8.24634790e-01 5.77373087e-01 -1.67314857e-02 9.26094577e-02
6.76830828e-01 -3.11696269e-02 -1.11505270e+00 -6.35523200e-01
-3.99129577e-02 1.01266420e+00 1.02847278e+00 5.26809156e-01
-9.76055980e-01 1.14055741e+00 5.04675341e+00 1.33468139e+00
-1.38238370e+00 6.62479736e-03 1.20036042e+00 -8.97400618e-01
-1.14000939e-01 -1.58373013e-01 -9.50213730e-01 3.18145126e-01
1.05615830e+00 -2.51908670e-03 6.97206438e-01 1.10156846e+00
-4.05497104e-02 -5.72374612e-02 -9.54437256e-01 1.30452216e+00
-1.52478695e-01 -1.90616655e+00 -3.55908930e-01 -1.98161080e-02
8.35147142e-01 4.23776329e-01 1.01641379e-01 2.52740592e-01
3.30095500e-01 -1.34054112e+00 9.45057809e-01 -1.99825123e-01
8.90333891e-01 -1.24257219e+00 6.46701932e-01 8.95187333e-02
-1.06534672e+00 -1.94017470e-01 -2.48751491e-01 2.93625534e-01
-5.05671911e-02 4.52521116e-01 -7.43092716e-01 2.58290507e-02
1.29534578e+00 -9.41692665e-02 -6.07893169e-01 1.01229405e+00
2.23965943e-01 6.34805679e-01 -8.29759002e-01 -4.17999476e-01
5.14612496e-01 1.31679460e-01 4.85347837e-01 9.64947283e-01
3.35690826e-01 -1.20133504e-01 1.09809570e-01 8.86716664e-01
-4.23537105e-01 -6.21550679e-02 3.77095163e-01 -8.71841684e-02
6.53253675e-01 1.05966759e+00 -1.22208822e+00 -1.64850757e-01
3.20876747e-01 9.10821140e-01 1.22182667e-01 -2.48902254e-02
-1.26025057e+00 -6.36664271e-01 5.78380764e-01 -9.84741524e-02
4.79101956e-01 1.21382982e-01 -7.32351363e-01 -3.46695662e-01
1.71168983e-01 -8.88851881e-01 2.27233067e-01 -3.49397391e-01
-5.38160682e-01 1.01884651e+00 -3.50906700e-01 -5.99598527e-01
1.09642208e-01 -5.77689469e-01 -2.40862310e-01 7.78454959e-01
-1.06979942e+00 -6.32202446e-01 -3.90696257e-01 1.83768943e-01
7.76797831e-01 -2.07427591e-01 5.92261374e-01 5.69304466e-01
-8.13383818e-01 1.31248307e+00 -4.87684011e-02 -2.93685138e-01
1.64580926e-01 -9.02348518e-01 7.06654131e-01 5.36360025e-01
1.34025421e-02 5.39564848e-01 6.05714560e-01 -3.91993016e-01
-1.36909807e+00 -1.11974335e+00 3.79226416e-01 3.92562039e-02
3.63270909e-01 -3.33627403e-01 -6.09580040e-01 1.70681775e-01
-1.00275852e-01 4.52480465e-02 3.81946623e-01 7.02744126e-02
-1.79594055e-01 -4.20890301e-01 -1.31578720e+00 6.92122161e-01
1.42716336e+00 1.00523874e-01 2.94292271e-01 1.09439254e-01
9.67091918e-01 -6.40443802e-01 -8.10258567e-01 6.71063840e-01
6.75236881e-01 -7.58375585e-01 9.39909160e-01 -2.48185217e-01
1.58573136e-01 -5.49642816e-02 -3.05388063e-01 -1.06440330e+00
-9.14032385e-02 -7.82594025e-01 9.30755341e-05 7.74248123e-01
1.14523566e+00 -6.09329164e-01 1.51220059e+00 4.25761163e-01
-3.76716167e-01 -1.45764387e+00 -1.10202289e+00 -7.91812658e-01
4.24498022e-02 -6.50365889e-01 7.19342351e-01 5.93904912e-01
-7.71450460e-01 2.50187039e-01 1.04706042e-01 -6.33631721e-02
5.45583427e-01 1.56309411e-01 3.18596691e-01 -1.09129393e+00
-7.66595185e-01 -9.27876353e-01 -4.14829329e-02 -9.75489736e-01
-1.17083639e-01 -7.43371665e-01 5.78620397e-02 -1.11568892e+00
-1.24708712e-01 -9.25690949e-01 -4.32797015e-01 8.10502589e-01
2.04669654e-01 5.39814949e-01 2.19152331e-01 -1.62740722e-01
-5.65021336e-01 2.49597535e-01 9.47008312e-01 -3.58778298e-01
-5.25410533e-01 -1.50903866e-01 -7.36392736e-01 8.30798149e-01
1.09897101e+00 -5.37229359e-01 -5.72288871e-01 -9.38361645e-01
2.53957063e-01 -5.39224483e-02 9.68545899e-02 -1.30415475e+00
8.61773193e-02 -2.03880712e-01 4.66447055e-01 -5.84610820e-01
5.57606578e-01 -4.83080328e-01 1.56939000e-01 7.29431868e-01
-3.56729239e-01 2.38550708e-01 8.56825471e-01 1.39205217e-01
1.64730757e-01 -4.31526780e-01 1.16904211e+00 1.64504070e-02
-7.53699362e-01 3.30434829e-01 1.06579483e-01 1.87462434e-01
1.02428150e+00 -4.65451211e-01 -4.61317211e-01 2.91521996e-01
-5.85589528e-01 4.88832831e-01 2.89082497e-01 3.89477462e-01
4.32955831e-01 -7.54353404e-01 -2.90344596e-01 1.72717664e-02
-3.00309479e-01 2.26710021e-01 3.84821326e-01 3.91584396e-01
-1.04059875e+00 5.71254134e-01 -6.12860471e-02 -6.51475608e-01
-1.50974047e+00 1.51041478e-01 6.03476405e-01 -2.66546786e-01
-3.48545402e-01 1.76008940e+00 -3.35975252e-02 -2.99562097e-01
3.92872721e-01 -1.13697968e-01 1.97115645e-01 -1.24641404e-01
2.48185471e-01 9.13128138e-01 3.76717776e-01 -1.18142754e-01
-7.24657953e-01 3.85603815e-01 -1.02409363e-01 -2.51156181e-01
1.31590533e+00 2.98683882e-01 7.06510916e-02 -1.08660243e-01
1.45097733e+00 -2.85805374e-01 -1.32125759e+00 7.23153055e-02
-9.88216922e-02 -1.97941568e-02 6.93171501e-01 -1.26530290e+00
-1.73789132e+00 5.63832164e-01 1.06775951e+00 -2.79814601e-01
1.41767621e+00 -6.31761970e-03 1.11901116e+00 4.46921796e-01
4.64613557e-01 -1.41159236e+00 -1.37106404e-01 3.36864442e-01
5.77054620e-01 -8.72665763e-01 2.92516928e-02 -2.72019058e-01
-1.04688331e-01 8.46315026e-01 1.07275295e+00 -1.20562769e-01
3.90310466e-01 4.71540600e-01 -2.43075844e-02 -3.56711566e-01
-9.04434144e-01 2.85219103e-01 1.28956884e-01 3.93975899e-02
-5.02298698e-02 1.87556177e-01 -3.29101503e-01 6.13717496e-01
-5.36005259e-01 -8.83845240e-02 8.76558386e-03 9.37340617e-01
-4.43666905e-01 -8.14060926e-01 -6.42115921e-02 7.54239559e-01
-6.96953535e-01 -2.14513361e-01 -1.98309019e-01 5.35775363e-01
3.71849179e-01 9.78427052e-01 4.15087372e-01 -6.28102779e-01
2.76120275e-01 -2.04111516e-01 3.80404323e-01 -1.36139244e-01
-9.21489537e-01 4.91765290e-02 3.90097558e-01 -8.83696020e-01
1.57937393e-01 -1.51598111e-01 -1.52101386e+00 -4.37064320e-01
-5.20273209e-01 1.26747191e-01 1.09536147e+00 6.18982077e-01
7.38154531e-01 7.92846382e-01 3.33537221e-01 -6.13718748e-01
-3.06333810e-01 -4.77156043e-01 4.13407795e-02 -3.62577558e-01
-7.37825930e-02 -6.60978675e-01 -2.08315358e-01 -4.27251101e-01] | [8.571768760681152, 3.0673108100891113] |
b1d8d801-43ef-49d2-b8dc-c248d875fa8f | an-end-to-end-and-accurate-ppg-based | 2105.00594 | null | https://arxiv.org/abs/2105.00594v2 | https://arxiv.org/pdf/2105.00594v2.pdf | An End-to-End and Accurate PPG-based Respiratory Rate Estimation Approach Using Cycle Generative Adversarial Networks | Respiratory rate (RR) is a clinical sign representing ventilation. An abnormal change in RR is often the first sign of health deterioration as the body attempts to maintain oxygen delivery to its tissues. There has been a growing interest in remotely monitoring of RR in everyday settings which has made photoplethysmography (PPG) monitoring wearable devices an attractive choice. PPG signals are useful sources for RR extraction due to the presence of respiration-induced modulations in them. The existing PPG-based RR estimation methods mainly rely on hand-crafted rules and manual parameters tuning. An end-to-end deep learning approach was recently proposed, however, despite its automatic nature, the performance of this method is not ideal using the real world data. In this paper, we present an end-to-end and accurate pipeline for RR estimation using Cycle Generative Adversarial Networks (CycleGAN) to reconstruct respiratory signals from raw PPG signals. Our results demonstrate a higher RR estimation accuracy of up to 2$\times$ (mean absolute error of 1.9$\pm$0.3 using five fold cross validation) compared to the state-of-th-art using a identical publicly available dataset. Our results suggest that CycleGAN can be a valuable method for RR estimation from raw PPG signals. | ['Amir M. Rahmani', 'Amir Hosein Afandizadeh Zargari', 'Rui Cao', 'Seyed Amir Hossein Aqajari'] | 2021-05-03 | null | null | null | null | ['photoplethysmography-ppg', 'respiratory-rate-estimation'] | ['medical', 'medical'] | [ 2.15194821e-01 5.74593283e-02 1.86552212e-01 -1.41434640e-01
-8.48853528e-01 -3.49315733e-01 7.03175217e-02 -2.96518683e-01
-2.77958870e-01 8.53958845e-01 4.19385470e-02 -1.49222314e-01
1.62230313e-01 -6.09907925e-01 -4.50489014e-01 -8.69509816e-01
-8.08291808e-02 -9.85583887e-02 -1.71180084e-01 1.53611019e-01
5.84225655e-02 5.54098248e-01 -1.01670599e+00 2.54794098e-02
6.23918295e-01 1.06227911e+00 -2.14843482e-01 1.13631284e+00
2.15288162e-01 6.17422402e-01 -9.75362659e-01 -5.29053360e-02
3.40759665e-01 -1.03438687e+00 -4.42069858e-01 -5.32391250e-01
1.70418695e-01 -3.83655846e-01 -5.03737509e-01 3.48531157e-01
1.22125435e+00 3.85810435e-02 3.03313345e-01 -8.63282979e-01
-1.48901958e-02 3.65719020e-01 -2.48160884e-01 4.20628548e-01
1.38943896e-01 6.77194476e-01 5.54523885e-01 -4.68708515e-01
5.54075539e-02 5.68249345e-01 9.69564021e-01 8.34224343e-01
-1.28864241e+00 -6.66999042e-01 -7.66809762e-01 -2.49731913e-02
-1.39339530e+00 -5.20411611e-01 9.13916767e-01 -2.82575548e-01
1.04370236e+00 5.10108292e-01 7.16977656e-01 1.19530988e+00
4.44758654e-01 -2.63715466e-03 1.49249542e+00 -2.24146411e-01
1.49148777e-01 1.32891685e-01 -4.76143688e-01 4.03715402e-01
9.57535505e-02 1.91277742e-01 -5.63255966e-01 -1.12402067e-01
7.25319386e-01 -8.24832544e-02 -6.80800855e-01 3.56369317e-01
-1.18075407e+00 4.60754961e-01 3.79406661e-01 3.49301279e-01
-5.80487728e-01 6.65308475e-01 3.22738409e-01 1.91926673e-01
2.73892462e-01 4.46677834e-01 -2.61251599e-01 -5.52095115e-01
-1.00353706e+00 -1.64534971e-01 8.56036842e-01 2.10047558e-01
8.72160196e-02 2.43204355e-01 -4.16161031e-01 9.21785533e-01
4.45210934e-01 6.30094886e-01 7.64884114e-01 -1.05536640e+00
1.29445612e-01 2.86251843e-01 2.67426334e-02 -8.93661439e-01
-6.39981329e-01 -5.21238863e-01 -9.37722206e-01 -4.73857485e-02
3.87119889e-01 -1.57752231e-01 -5.47561526e-01 1.57714438e+00
1.99718282e-01 4.06283826e-01 -3.82586271e-02 1.03379726e+00
1.00053072e+00 2.65838057e-01 1.06866792e-01 -3.60180169e-01
1.21530640e+00 -4.42352980e-01 -7.23212659e-01 -5.96499294e-02
1.31909788e-01 -4.53375578e-01 1.00535882e+00 3.44834656e-01
-9.76450861e-01 -6.11806214e-01 -9.99253333e-01 2.29248315e-01
1.02962472e-01 -4.15779948e-02 1.96931794e-01 1.08719206e+00
-9.25289631e-01 9.79873240e-01 -9.93545115e-01 -1.80142507e-01
4.33190405e-01 3.53836596e-01 -1.79855257e-01 2.07561076e-01
-1.32885563e+00 7.35957980e-01 -1.98716879e-01 4.66095895e-01
-5.92451096e-01 -7.88873672e-01 -5.15920579e-01 -2.08077103e-01
-2.28258539e-02 -7.47127235e-01 9.17161882e-01 -3.65954727e-01
-1.98773956e+00 8.00626814e-01 -1.16809480e-01 -4.92602557e-01
7.68528402e-01 -6.98089749e-02 -5.57148397e-01 4.41754222e-01
-5.14528811e-01 1.47463351e-01 9.52711940e-01 -6.55227900e-01
3.05715412e-01 -2.36782178e-01 -4.59644228e-01 -6.56250864e-02
1.00550829e-02 -3.43776532e-02 9.37529095e-03 -4.17754650e-01
4.73965108e-02 -9.48471069e-01 1.07933290e-01 4.38062586e-02
-2.13948935e-01 8.64172429e-02 4.67830688e-01 -7.67844558e-01
1.15305114e+00 -1.81339014e+00 -3.32601041e-01 -1.85167044e-02
3.94592553e-01 4.55906510e-01 2.56861210e-01 2.99429357e-01
-7.55424052e-02 3.31869781e-01 -4.34020281e-01 -2.60343790e-01
-2.67145157e-01 -1.22913726e-01 -2.22517010e-02 8.44327986e-01
5.56652956e-02 1.10816967e+00 -8.58641088e-01 -2.78713375e-01
6.79563820e-01 9.23931897e-01 -1.13928996e-01 6.83371246e-01
2.32764512e-01 1.01718068e+00 -2.42829379e-02 5.35229921e-01
3.65342379e-01 -1.97621241e-01 8.58730450e-02 -4.10307854e-01
-3.33477892e-02 5.07070243e-01 -8.09788346e-01 1.53007877e+00
-5.61626494e-01 8.44488800e-01 -3.02862078e-01 -7.13052213e-01
1.11488676e+00 7.82589316e-01 7.86074221e-01 -6.89160466e-01
2.63992041e-01 3.01241010e-01 3.30595195e-01 -7.88479626e-01
-2.93265253e-01 -6.01418495e-01 1.29343569e-01 5.38694441e-01
-2.93359995e-01 -5.50641060e-01 -4.11242098e-01 -3.90041769e-01
1.29241574e+00 2.46723473e-01 1.76848084e-01 -2.39307526e-02
5.33108056e-01 -4.37300414e-01 4.07884181e-01 6.79666340e-01
-5.30463398e-01 1.16017628e+00 2.15026274e-01 -3.81501168e-01
-6.69240654e-01 -7.13820219e-01 -2.82452077e-01 1.42314553e-01
-3.91771793e-01 -4.01211902e-02 -6.99161530e-01 -3.60783279e-01
-1.13816179e-01 6.71540856e-01 -5.53124905e-01 -2.83325195e-01
-7.45069027e-01 -9.70752537e-01 1.06511796e+00 5.55260420e-01
6.55423462e-01 -1.42477310e+00 -9.87533987e-01 5.02518475e-01
-4.20092881e-01 -1.12109125e+00 -2.23698452e-01 3.13796178e-02
-9.94889319e-01 -1.04173028e+00 -8.74650180e-01 1.67851210e-01
1.63514674e-01 -1.76499993e-01 1.07535362e+00 -1.26874939e-01
-8.21393788e-01 4.41056341e-01 -6.25593355e-03 -6.62580192e-01
-5.50704122e-01 -9.26377773e-02 1.37330800e-01 1.70208290e-02
3.43016297e-01 -9.01842058e-01 -1.30930710e+00 3.09768379e-01
-5.01271784e-01 -2.52149850e-01 4.26401824e-01 3.86716485e-01
7.56682158e-01 -4.64974612e-01 7.96545208e-01 -6.68291926e-01
6.87123239e-01 -4.57774162e-01 -4.04827148e-01 -3.02026659e-01
-8.24453175e-01 -8.96027163e-02 7.08617210e-01 -3.17731500e-01
-4.05613393e-01 -2.02522472e-01 -1.88659549e-01 -7.25766480e-01
-6.68978766e-02 9.51630846e-02 1.82778448e-01 7.28213321e-03
8.78503621e-01 2.98933268e-01 3.24630529e-01 -1.67887390e-01
2.29024701e-02 7.16052473e-01 6.98790371e-01 -1.76803961e-01
6.44771516e-01 3.78270060e-01 4.89579588e-01 -9.17372048e-01
-4.48048234e-01 -3.60611647e-01 -4.37219352e-01 -3.98031324e-01
9.79707539e-01 -8.73788834e-01 -1.05085182e+00 6.57615304e-01
-8.03974688e-01 -6.45910800e-01 -4.79230821e-01 6.32053435e-01
-5.33130109e-01 4.17397439e-01 -3.41792345e-01 -1.02436042e+00
-1.09686005e+00 -7.68212736e-01 6.69601262e-01 3.16397548e-01
-4.57875252e-01 -9.36225832e-01 3.75262946e-01 6.09583199e-01
1.05197334e+00 9.04786050e-01 2.69855917e-01 -1.96526244e-01
-2.76245773e-01 -4.36545491e-01 4.61698920e-02 6.73108697e-01
4.10774440e-01 -1.51361972e-01 -1.48162878e+00 -1.67019412e-01
4.75352973e-01 -1.73587441e-01 4.47654724e-01 4.79707390e-01
1.51727068e+00 -1.41510502e-01 6.19001351e-02 7.42106616e-01
1.39934397e+00 2.18289137e-01 1.00099361e+00 -7.25376830e-02
7.97127903e-01 2.94450492e-01 1.04101174e-01 4.34699982e-01
2.32856572e-02 5.63526511e-01 4.93095130e-01 -1.20802209e-01
-3.62413347e-01 -5.60416169e-02 2.73781657e-01 5.57871163e-01
-2.77018219e-01 -3.06167871e-01 -6.36952221e-01 2.90656537e-01
-1.12988138e+00 -9.01457727e-01 -5.33795357e-01 2.63965392e+00
1.03715515e+00 -1.74763292e-01 1.50938809e-01 3.35096359e-01
5.71287811e-01 1.92420959e-01 -7.52696812e-01 -5.63722312e-01
3.01412702e-01 7.18722224e-01 5.66438019e-01 6.01420514e-02
-6.65049732e-01 1.25286207e-01 6.44683790e+00 -5.07611111e-02
-1.64042473e+00 7.03220889e-02 6.40765309e-01 -1.74706757e-01
-7.56544769e-02 -3.80400300e-01 -4.40553814e-01 8.64295781e-01
1.76433277e+00 1.22694470e-01 5.32805860e-01 3.47544223e-01
7.64452338e-01 -1.05786011e-01 -9.70395446e-01 1.31720185e+00
3.58102508e-02 -8.66961122e-01 -8.17311883e-01 4.33388129e-02
9.55514982e-02 2.25509465e-01 -2.31874496e-01 -5.09001687e-02
-4.62198347e-01 -1.52687788e+00 3.68580632e-02 6.43986821e-01
1.37066483e+00 -4.79682803e-01 6.73823774e-01 1.15725949e-01
-8.71912837e-01 3.20893317e-01 -6.72566965e-02 2.40548432e-01
1.38082067e-02 9.06393886e-01 -1.11737597e+00 2.96943665e-01
4.69572365e-01 3.72041702e-01 -2.69089967e-01 1.01203048e+00
-4.44144309e-01 1.05793500e+00 -5.75886846e-01 5.08619845e-02
-4.05606508e-01 6.68522939e-02 5.98116159e-01 1.09293330e+00
3.48727554e-01 3.69930357e-01 -5.67155004e-01 1.28632700e+00
-2.86372483e-01 -2.32406586e-01 -5.98590314e-01 -1.09098367e-01
3.40162069e-01 1.33533061e+00 -4.68285739e-01 1.15407169e-01
-3.43213111e-01 9.11933601e-01 -3.81722867e-01 6.21301346e-02
-1.00646508e+00 -6.57507181e-01 4.80421096e-01 4.38025355e-01
-1.10884793e-01 1.81852393e-02 -3.22347313e-01 -8.58901203e-01
8.07958841e-02 -8.41697514e-01 2.64812678e-01 -7.64423192e-01
-8.98561299e-01 6.01636887e-01 -1.48973599e-01 -1.28621054e+00
-4.45143729e-01 -2.28248730e-01 -7.79350042e-01 1.41794133e+00
-1.69280422e+00 -4.23904896e-01 -9.55571890e-01 3.39248061e-01
2.94851154e-01 3.23602945e-01 1.08776712e+00 2.85034508e-01
-5.72171628e-01 6.12474322e-01 -2.89278507e-01 -4.84545194e-02
7.07540512e-01 -1.29616249e+00 3.24967533e-01 8.02647114e-01
-2.99497336e-01 5.35008371e-01 6.88239276e-01 -2.44409069e-01
-1.68314922e+00 -1.09225142e+00 7.06481457e-01 -6.38345122e-01
1.53418481e-01 -7.39855617e-02 -8.71350706e-01 1.56714812e-01
3.20519619e-02 6.48950100e-01 8.35289180e-01 -6.65474296e-01
-2.29122769e-02 -4.03690636e-01 -1.57115936e+00 1.41617641e-01
5.84630013e-01 -5.31157374e-01 -4.48130310e-01 7.57430792e-02
2.93527752e-01 -6.37754321e-01 -1.36361313e+00 4.86848712e-01
7.02818155e-01 -9.12207544e-01 8.58309150e-01 1.45743608e-01
2.51944423e-01 -4.21621680e-01 2.14582786e-01 -1.15152943e+00
9.76359844e-02 -1.06282508e+00 -3.55855823e-01 1.06546474e+00
-3.12511832e-03 -9.50830996e-01 6.91737294e-01 9.98035967e-01
-1.98980905e-02 -6.12043381e-01 -7.62341678e-01 -5.23297727e-01
-2.60035813e-01 -5.67746878e-01 2.60830075e-01 6.74185872e-01
-1.34135842e-01 1.41410708e-01 -3.49800915e-01 -5.81892282e-02
6.80314898e-01 -7.63373524e-02 6.48546576e-01 -1.06380963e+00
-5.19014835e-01 -2.27772355e-01 -2.16682866e-01 -3.40174854e-01
-3.28265399e-01 -6.84436858e-01 2.19921276e-01 -1.53896654e+00
-1.77425534e-01 -3.31445277e-01 -7.67988086e-01 4.58211958e-01
-3.38544279e-01 7.41271675e-01 -1.16526790e-01 8.75117332e-02
1.52639061e-01 2.43723184e-01 1.11134303e+00 2.07864329e-01
-6.24666631e-01 2.48248309e-01 -5.51524937e-01 2.03949913e-01
1.03838027e+00 -6.95931792e-01 -4.53502983e-01 1.86397716e-01
-1.24600809e-02 3.08366746e-01 4.76020128e-01 -1.19275808e+00
-1.91257730e-01 1.59400672e-01 5.68162024e-01 -2.87810534e-01
2.85117656e-01 -6.97668612e-01 6.16030335e-01 7.11647987e-01
-1.57240570e-01 -1.34150252e-01 3.44022155e-01 3.46344113e-01
3.94068323e-02 2.40609678e-03 1.08527768e+00 -3.77603859e-01
-2.50730254e-02 2.31768563e-01 -2.91168094e-01 3.27581495e-01
6.22821331e-01 -3.10612917e-01 -3.43902439e-01 -4.20479447e-01
-3.07185322e-01 -2.39421755e-01 1.41110972e-01 2.08352476e-01
6.68925703e-01 -9.33550060e-01 -7.50533104e-01 1.79097146e-01
4.92531583e-02 6.02482669e-02 2.77560145e-01 1.35746849e+00
-6.82547629e-01 3.02549720e-01 -5.69528714e-02 -6.86816096e-01
-1.15976131e+00 4.74722460e-02 1.03799677e+00 -7.03062788e-02
-9.24382985e-01 6.79353356e-01 -3.41742337e-01 1.37336582e-01
3.47530954e-02 -3.16154927e-01 -1.92423955e-01 -2.17045635e-01
4.74429429e-01 5.22697508e-01 3.05262804e-01 -7.31757805e-02
-4.84455675e-01 5.70415854e-01 6.03658259e-01 -1.17688082e-01
1.28990209e+00 -8.68728682e-02 -4.67436276e-02 6.87278509e-01
1.20877492e+00 1.25362888e-01 -1.14598608e+00 2.38751367e-01
-5.04344702e-01 -4.46060956e-01 2.07060978e-01 -1.08506680e+00
-1.38370216e+00 1.01408398e+00 1.02757132e+00 1.52565092e-01
1.32020581e+00 -4.06506777e-01 8.51939142e-01 2.45348923e-02
2.61594772e-01 -7.34763503e-01 9.80950743e-02 -3.35922301e-01
8.28201175e-01 -9.87485468e-01 -8.49103183e-02 -3.13728675e-02
-4.89909351e-01 1.08991134e+00 2.59980381e-01 -3.75863202e-02
3.99097979e-01 5.25022186e-02 2.99716949e-01 -3.94334607e-02
-4.86543030e-01 2.39093944e-01 1.95339561e-01 5.98506749e-01
8.25445354e-01 -3.04282960e-02 -5.01371026e-01 3.21734935e-01
-6.18989542e-02 3.27273190e-01 5.78836858e-01 5.76864421e-01
-3.73547971e-02 -9.77936149e-01 -1.98901996e-01 5.46685040e-01
-1.24984217e+00 -6.11382835e-02 -8.08307752e-02 6.83525443e-01
-2.75557011e-01 1.23492289e+00 -2.22687155e-01 -1.64220542e-01
3.92501295e-01 4.38326985e-01 6.67306662e-01 -3.98334771e-01
-8.91293049e-01 -1.06938183e-01 -3.38825919e-02 -7.55531788e-01
-5.65751135e-01 -4.56432939e-01 -1.29133713e+00 -2.18807422e-02
5.73662296e-02 -1.31166592e-01 1.02803040e+00 5.62984824e-01
4.96978134e-01 7.65055358e-01 8.78559709e-01 -5.15801251e-01
-4.14861947e-01 -1.15286815e+00 -3.78599674e-01 3.90825391e-01
4.09953922e-01 -1.81330100e-01 -7.83876002e-01 -5.74306101e-02] | [13.941012382507324, 2.9520440101623535] |
3e14f19f-44c0-4d90-8d2c-7bb7c2fb84f4 | meta-referential-games-to-learn-compositional-1 | 2207.08012 | null | https://arxiv.org/abs/2207.08012v1 | https://arxiv.org/pdf/2207.08012v1.pdf | Meta-Referential Games to Learn Compositional Learning Behaviours | Human beings use compositionality to generalise from past experiences to actual or fictive, novel experiences. To do so, we separate our experiences into fundamental atomic components. These atomic components can then be recombined in novel ways to support our ability to imagine and engage with novel experiences. We frame this as the ability to learn to generalise compositionally. And, we will refer to behaviours making use of this ability as compositional learning behaviours (CLBs). A central problem to learning CLBs is the resolution of a binding problem (BP) (by learning to, firstly, segregate the supportive stimulus components from the observation of multiple stimuli, and then, combine them in a single episodic experience). While it is another feat of intelligence that human beings perform with ease, it is not the case for state-of-the-art artificial agents. Thus, in order to build artificial agents able to collaborate with human beings, we propose to develop a novel benchmark to investigate agents' abilities to exhibit CLBs by solving a domain-agnostic version of the BP. We take inspiration from the language emergence and grounding framework of referential games and propose a meta-learning extension of referential games, entitled Meta-Referential Games, and use this framework to build our benchmark, that we name Symbolic Behaviour Benchmark (S2B). While it has the potential to test for more symbolic behaviours, rather than solely CLBs, in the present paper, though, we solely focus on the single-agent language grounding task that tests for CLBs. We provide baseline results for it, using state-of-the-art RL agents, and show that our proposed benchmark is a compelling challenge that we hope will spur the research community towards developing more capable artificial agents. | ['James Alfred Walker', 'Sondess Missaoui', 'Kevin Denamganaï'] | 2022-07-16 | meta-referential-games-to-learn-compositional | https://openreview.net/forum?id=ffS_Y258dZs | https://openreview.net/pdf?id=ffS_Y258dZs | null | ['systematic-generalization'] | ['reasoning'] | [ 2.76753694e-01 4.91851479e-01 2.60020286e-01 6.17438667e-02
-1.40550643e-01 -6.23547852e-01 1.18201125e+00 1.07959114e-01
-6.09685302e-01 9.96248662e-01 9.48273912e-02 -2.77911305e-01
-4.21146154e-01 -1.01904893e+00 -7.24744260e-01 -5.98917186e-01
-3.27512801e-01 6.33055806e-01 4.16164190e-01 -9.30848718e-01
2.11373910e-01 4.05845165e-01 -1.94007802e+00 2.83565879e-01
9.11002040e-01 3.81796271e-01 4.02456760e-01 4.49261665e-01
2.40004193e-02 1.17929697e+00 -7.52381384e-01 -3.16082388e-01
1.53002471e-01 -1.02902436e+00 -1.09050012e+00 -1.20018736e-01
-2.62165070e-01 4.25454080e-02 7.11371228e-02 7.37231016e-01
4.18143600e-01 4.50211972e-01 5.49583793e-01 -1.48191118e+00
-6.98277354e-01 7.18638718e-01 2.21906111e-01 -9.39439423e-03
8.30043256e-01 4.49630439e-01 8.78176987e-01 -2.67329454e-01
9.65928376e-01 1.11222804e+00 6.10696554e-01 1.07704914e+00
-1.29467261e+00 -3.53153467e-01 2.32013032e-01 9.98369306e-02
-9.46572363e-01 -3.37779611e-01 6.47763491e-01 -2.68946975e-01
1.33147871e+00 3.38962018e-01 1.16431177e+00 1.45953798e+00
8.17725062e-03 8.01278234e-01 1.65838897e+00 -6.07257068e-01
6.53701246e-01 2.84764711e-02 -2.26478979e-01 5.33582866e-01
2.38703862e-01 7.12797523e-01 -6.69323385e-01 2.25848351e-02
8.06514561e-01 -2.62116134e-01 -1.09290995e-01 -5.04986525e-01
-1.23778725e+00 7.71867096e-01 4.43131983e-01 8.30584466e-01
-5.89267075e-01 3.68677258e-01 1.96530640e-01 7.10479558e-01
-1.93143323e-01 1.18457866e+00 -1.37597591e-01 -3.24777782e-01
-5.35902441e-01 6.21357262e-01 1.00268471e+00 4.93230641e-01
6.41078174e-01 1.35641217e-01 5.03304116e-02 5.43401539e-01
1.08323842e-01 1.64544076e-01 8.82452011e-01 -1.12875903e+00
-3.06245655e-01 6.95163488e-01 -2.10889559e-02 -6.39668763e-01
-5.43860257e-01 -4.94947940e-01 -4.61632699e-01 5.25096476e-01
2.19215944e-01 -9.05162618e-02 -5.02507150e-01 2.27723765e+00
1.74122259e-01 5.00022829e-01 6.45960212e-01 8.75804067e-01
7.84218431e-01 6.45424843e-01 1.04610436e-01 -4.01277483e-01
1.14368320e+00 -7.89216042e-01 -3.01829308e-01 -3.43327165e-01
6.56537056e-01 -3.10848653e-02 1.06827962e+00 4.25214827e-01
-1.46340251e+00 -3.25548649e-01 -1.24624562e+00 3.19956511e-01
-4.86422837e-01 -9.91921842e-01 1.08240032e+00 4.54743147e-01
-1.30970705e+00 6.69861138e-01 -6.32978320e-01 -7.63749659e-01
-7.42305219e-02 2.65347362e-01 -2.81486928e-01 4.48516190e-01
-1.53509521e+00 1.27957332e+00 6.92833960e-01 -2.61115789e-01
-1.06838822e+00 -3.76867801e-01 -8.19557488e-01 -1.67897984e-01
7.01723874e-01 -1.18130231e+00 1.41996539e+00 -1.28067899e+00
-1.75982952e+00 1.12494636e+00 1.73625663e-01 -6.89695060e-01
2.69082069e-01 2.90829301e-01 -3.95728201e-01 -1.19406588e-01
-1.00177050e-01 7.72545397e-01 4.78388250e-01 -1.74153471e+00
-6.46981716e-01 7.14063793e-02 7.33759522e-01 2.60350764e-01
1.81782573e-01 -5.93255721e-02 2.14312896e-01 -6.19808018e-01
7.19674630e-03 -1.02680969e+00 -1.30403072e-01 -5.55553079e-01
6.76214173e-02 -3.68464977e-01 2.40549490e-01 2.18062662e-02
7.25543559e-01 -2.09445786e+00 8.13860118e-01 7.95741752e-02
3.27889949e-01 1.86738208e-01 -3.86465967e-01 7.55144060e-01
4.76845130e-02 6.67027012e-02 -3.66395354e-01 -1.99407324e-01
5.60537279e-01 6.64237857e-01 -4.71477002e-01 -7.40321130e-02
2.19910711e-01 1.36228609e+00 -1.27411556e+00 -1.35432035e-01
-4.85674217e-02 1.75172746e-01 -6.12645388e-01 2.40442380e-01
-6.86740041e-01 4.91860390e-01 -1.80322006e-01 1.79266736e-01
6.93805888e-02 -6.07320517e-02 2.46088743e-01 4.55522686e-01
-1.38995200e-01 4.46920097e-01 -1.17534029e+00 1.77501404e+00
-4.18046445e-01 3.18404764e-01 -1.75885737e-01 -1.12812614e+00
7.95724571e-01 4.72806960e-01 1.22923099e-01 -1.19995522e+00
3.31761986e-02 3.64651620e-01 4.41872805e-01 -4.82813686e-01
3.69650662e-01 -6.63517833e-01 -3.57547432e-01 7.99011171e-01
1.33596614e-01 -4.22246724e-01 4.25217479e-01 1.29464403e-01
1.19439709e+00 4.04745311e-01 5.72656691e-01 -1.46281719e-01
4.50764209e-01 -2.86120195e-02 4.40299153e-01 9.53809738e-01
-1.11689605e-01 -1.62146136e-03 4.71520424e-01 -6.04951680e-01
-8.22878301e-01 -1.41177177e+00 2.67841190e-01 1.29343331e+00
2.75061101e-01 -3.57060730e-01 -7.13928759e-01 -3.51245731e-01
-3.13584447e-01 1.07114887e+00 -7.80602634e-01 -3.96217912e-01
-6.68254495e-01 -5.03681719e-01 7.43342876e-01 2.15851218e-01
6.68299079e-01 -2.17619133e+00 -1.45648897e+00 3.86926472e-01
5.32510579e-02 -5.31860352e-01 3.41885537e-01 5.14808178e-01
-5.57500660e-01 -7.65967548e-01 -4.23699468e-01 -8.55495334e-01
7.16397688e-02 -8.17771405e-02 1.42763770e+00 6.32336318e-01
1.27035856e-01 7.20864713e-01 -6.01538718e-01 -4.23447371e-01
-7.31135368e-01 -2.71840803e-02 9.37218219e-02 -5.15446603e-01
2.85068378e-02 -1.04382551e+00 -2.21820846e-01 1.64341033e-01
-1.33191550e+00 3.38546038e-01 5.85138559e-01 9.75291371e-01
1.28931507e-01 1.30199000e-01 7.39417553e-01 -6.87576532e-01
9.98644233e-01 -4.96144652e-01 -3.81366342e-01 2.93252558e-01
-3.25310320e-01 2.70832807e-01 5.36113858e-01 -6.26614571e-01
-8.59993398e-01 -4.41585839e-01 -4.72161733e-02 3.39706987e-01
-4.83247161e-01 8.62604737e-01 4.20725793e-02 1.98909380e-02
8.28918993e-01 6.14973366e-01 -4.15038550e-03 1.60113320e-01
5.71086526e-01 2.28206858e-01 6.66908085e-01 -1.05373561e+00
8.29836905e-01 9.54357311e-02 8.14734772e-02 -5.73954761e-01
-5.00154495e-01 6.38191998e-02 -3.64653856e-01 -1.49492621e-01
7.70642519e-01 -4.00851935e-01 -1.04387987e+00 2.74700761e-01
-1.01882255e+00 -9.52606142e-01 -6.96237266e-01 9.67357010e-02
-1.23960102e+00 -7.79531375e-02 -4.49422717e-01 -8.84450197e-01
1.48238808e-01 -1.02738714e+00 6.39023960e-01 2.52702296e-01
-6.75796390e-01 -1.11614490e+00 5.56545377e-01 -7.25057572e-02
5.39928555e-01 4.76807535e-01 1.02770269e+00 -8.63745451e-01
-4.64343488e-01 4.96170312e-01 4.13202375e-01 -1.64515555e-01
-4.44951765e-02 -4.43157732e-01 -8.16350400e-01 -3.49880427e-01
1.26403064e-01 -7.33256102e-01 6.53040290e-01 -1.68870240e-01
2.54762143e-01 -3.51950973e-01 -1.21604919e-01 3.88109058e-01
1.20966351e+00 5.22373259e-01 9.39419270e-01 1.07495081e+00
-6.63587600e-02 8.97714794e-01 4.09928739e-01 2.68807769e-01
6.45850480e-01 8.67068052e-01 3.98738146e-01 6.45914897e-02
-7.18735904e-02 -2.29424268e-01 4.64218616e-01 6.57373786e-01
-3.93712074e-01 -1.27998665e-01 -9.19893324e-01 4.89882350e-01
-2.12149024e+00 -1.48292935e+00 4.39258993e-01 1.88987446e+00
1.22137189e+00 1.37677625e-01 3.91680986e-01 1.89449862e-01
3.31377834e-01 -4.36665118e-02 -3.44423860e-01 -7.76790142e-01
-3.53528202e-01 6.28397882e-01 -4.97087955e-01 5.62886357e-01
-6.28002822e-01 1.21416688e+00 5.89016485e+00 5.15490592e-01
-1.15977907e+00 7.56821409e-02 7.84974545e-02 1.41159490e-01
-5.22785664e-01 2.51135472e-02 -1.28017187e-01 1.80424944e-01
9.26846206e-01 -2.52506793e-01 1.20030963e+00 4.42023695e-01
-2.27153659e-01 -8.92661884e-02 -1.44864392e+00 6.71057105e-01
8.97827819e-02 -1.24437189e+00 8.46864283e-02 -3.93882841e-02
6.90017939e-01 -3.00766587e-01 2.20373765e-01 7.77623177e-01
6.05470240e-01 -1.43456709e+00 1.05916524e+00 5.24627745e-01
1.87313974e-01 -6.19266808e-01 5.22932470e-01 6.37578070e-01
-9.37891543e-01 -2.47495607e-01 -8.70976970e-02 -6.75892532e-01
9.92286131e-02 -2.92394310e-01 -5.99959791e-01 5.48581958e-01
4.06896651e-01 2.52113968e-01 -5.29360712e-01 9.63477910e-01
-4.64667618e-01 1.54530197e-01 -1.07524790e-01 -3.03446025e-01
3.85713160e-01 9.60266814e-02 6.17516041e-01 7.65593052e-01
3.65973115e-01 4.93133456e-01 3.92929912e-02 1.26139474e+00
2.50112563e-01 -1.56625137e-01 -7.97950447e-01 -3.75315212e-02
4.87643898e-01 9.71496582e-01 -8.42543423e-01 -2.82906294e-01
-2.30356619e-01 9.99891758e-01 4.66619402e-01 3.49691302e-01
-6.77423358e-01 -3.87606099e-02 4.12560344e-01 -7.19899833e-02
6.66469187e-02 -2.82781869e-01 -6.77470118e-02 -1.07376599e+00
-2.87380546e-01 -1.48599136e+00 1.98169008e-01 -9.22838748e-01
-1.27368295e+00 7.97757506e-01 2.32620895e-01 -6.99335098e-01
-8.27364385e-01 -3.71975213e-01 -6.27597570e-01 5.66762626e-01
-1.31068897e+00 -1.32408965e+00 -3.05850595e-01 7.13059783e-01
2.98876971e-01 -2.51676589e-01 1.19096684e+00 -4.51072723e-01
-4.42256667e-02 3.34021986e-01 -5.46481550e-01 -2.69822001e-01
3.44637424e-01 -1.33384502e+00 5.59673429e-01 6.20981693e-01
4.90529686e-01 1.05901039e+00 8.62822056e-01 -5.66510320e-01
-1.19934630e+00 -6.30715787e-01 7.68691480e-01 -5.06269932e-01
6.10323370e-01 -4.50154841e-01 -9.17831540e-01 7.29259133e-01
4.21058029e-01 -3.47363174e-01 5.63310444e-01 8.97139087e-02
-3.16796452e-01 4.19430375e-01 -9.95520234e-01 1.14933896e+00
1.56252325e+00 -4.23654258e-01 -1.50121558e+00 -7.19179586e-02
8.29214036e-01 -1.32976323e-01 -5.03371894e-01 3.95907879e-01
4.47481394e-01 -1.32712233e+00 1.05166209e+00 -6.91457868e-01
2.92627007e-01 -5.27473271e-01 -2.05264315e-01 -1.64711118e+00
-4.71065491e-01 -8.42579544e-01 3.33032943e-02 1.10926640e+00
3.67539108e-01 -1.11953402e+00 4.93510187e-01 3.92573625e-01
-2.34108970e-01 -4.62524831e-01 -9.99645472e-01 -1.05859816e+00
4.11530316e-01 -4.72363681e-01 8.25375319e-01 1.14461899e+00
6.08233809e-01 3.13200116e-01 -4.51396853e-02 -1.03852376e-01
2.85414636e-01 1.92635119e-01 8.26862335e-01 -1.33852220e+00
-9.01358306e-01 -8.73280227e-01 -3.27866584e-01 -5.76267898e-01
5.09869039e-01 -1.10810184e+00 1.01054549e-01 -1.40760803e+00
1.56762451e-01 -5.19162118e-01 -2.37050027e-01 5.49074650e-01
6.10325485e-02 2.50315726e-01 5.32801688e-01 2.64859498e-01
-9.48180556e-01 5.14998615e-01 1.18051720e+00 -3.90039682e-02
-4.29691851e-01 -4.45114315e-01 -8.74048829e-01 6.78721607e-01
7.87694752e-01 -2.38894224e-01 -6.61093831e-01 -1.08465731e-01
8.37481141e-01 4.70070839e-02 7.47247279e-01 -1.23163760e+00
4.12833512e-01 -5.07477820e-01 -6.99018240e-02 2.53635138e-01
2.85155267e-01 -6.49443328e-01 5.37779927e-01 8.34856153e-01
-3.33575517e-01 1.20351389e-01 3.11925828e-01 1.01351552e-01
-1.01098664e-01 -4.72285688e-01 3.47628206e-01 -4.49736983e-01
-1.03672731e+00 -4.06564951e-01 -7.60655344e-01 1.85586795e-01
1.22949696e+00 -5.23519635e-01 -4.24506336e-01 -4.94179606e-01
-8.04718971e-01 -1.15252519e-02 8.65268111e-01 2.76525617e-01
5.97147882e-01 -1.35982847e+00 -5.88092446e-01 1.87065706e-01
2.71585941e-01 -2.38698199e-01 -3.70408148e-02 5.96170843e-01
-4.20126647e-01 3.35787535e-01 -6.67418659e-01 -2.45661736e-01
-7.26750374e-01 8.56501222e-01 5.19754112e-01 -2.29701668e-01
-4.90258873e-01 7.66210556e-01 3.80821586e-01 -4.89203751e-01
5.52944280e-02 -2.56466061e-01 -3.33773375e-01 -8.99221599e-02
4.24422920e-01 -5.73280416e-02 -1.71278715e-01 -4.56505090e-01
-3.14163893e-01 3.85512024e-01 2.92664051e-01 -6.15329802e-01
1.51183307e+00 1.32170960e-01 -3.45113814e-01 6.60911918e-01
3.93651634e-01 -1.00296512e-01 -1.00047994e+00 -6.46569356e-02
2.08657309e-01 -6.23250492e-02 -4.77383703e-01 -1.08173883e+00
-3.46659869e-01 3.77419084e-01 2.35175431e-01 7.12454140e-01
1.12648404e+00 2.49949142e-01 4.88402933e-01 5.58124900e-01
9.13724005e-01 -7.40403652e-01 5.47721684e-01 7.61568666e-01
1.15358782e+00 -7.69023418e-01 -3.38020444e-01 1.06740296e-01
-6.19758844e-01 8.70394647e-01 5.44218242e-01 -2.49595895e-01
-2.58930512e-02 4.16789681e-01 -1.64895177e-01 -3.64094704e-01
-1.20658696e+00 -5.85252762e-01 2.67679747e-02 1.20306671e+00
3.17059070e-01 -1.46789312e-01 -2.94403881e-01 6.47548914e-01
-6.50600672e-01 4.68260273e-02 4.97647285e-01 1.09588778e+00
-4.22386795e-01 -1.51139164e+00 -2.79141068e-01 -1.45323783e-01
-4.03608307e-02 -4.80252951e-02 -7.52444506e-01 1.10247886e+00
4.31792110e-01 9.04445529e-01 -6.79089651e-02 -3.12518150e-01
3.85243058e-01 1.90676287e-01 9.90088224e-01 -7.55734086e-01
-9.29969251e-01 -4.82476294e-01 1.14869863e-01 -5.10472298e-01
-8.39651227e-01 -6.99505091e-01 -1.46954024e+00 -4.65133011e-01
2.34321162e-01 2.30242267e-01 1.86402828e-01 1.10623360e+00
-6.07264265e-02 6.33102834e-01 8.12760592e-02 -1.03329587e+00
-2.61848986e-01 -6.65232897e-01 -3.58253658e-01 7.51602829e-01
2.39686295e-01 -8.95338356e-01 -4.34354514e-01 -1.18735842e-01] | [4.237679481506348, 1.395906686782837] |
a25d14df-9836-494c-907e-95fb567179bf | deblurring-masked-autoencoder-is-better | 2306.08249 | null | https://arxiv.org/abs/2306.08249v2 | https://arxiv.org/pdf/2306.08249v2.pdf | Deblurring Masked Autoencoder is Better Recipe for Ultrasound Image Recognition | Masked autoencoder (MAE) has attracted unprecedented attention and achieves remarkable performance in many vision tasks. It reconstructs random masked image patches (known as proxy task) during pretraining and learns meaningful semantic representations that can be transferred to downstream tasks. However, MAE has not been thoroughly explored in ultrasound imaging. In this work, we investigate the potential of MAE for ultrasound image recognition. Motivated by the unique property of ultrasound imaging in high noise-to-signal ratio, we propose a novel deblurring MAE approach that incorporates deblurring into the proxy task during pretraining. The addition of deblurring facilitates the pretraining to better recover the subtle details presented in the ultrasound images, thus improving the performance of the downstream classification task. Our experimental results demonstrate the effectiveness of our deblurring MAE, achieving state-of-the-art performance in ultrasound image classification. Overall, our work highlights the potential of MAE for ultrasound image recognition and presents a novel approach that incorporates deblurring to further improve its effectiveness. | ['Qicheng Lao', 'Kang Li', 'Jun Gao', 'Qingbo Kang'] | 2023-06-14 | null | null | null | null | ['deblurring'] | ['computer-vision'] | [ 6.28027439e-01 9.15580802e-03 3.33850354e-01 -3.58054966e-01
-9.67136145e-01 -2.21375972e-01 4.08581585e-01 -4.29921120e-01
-2.44767189e-01 3.46040845e-01 5.01107395e-01 -1.34644032e-01
-2.02889830e-01 -4.75271851e-01 -8.92822146e-01 -1.24094093e+00
-1.24009714e-01 -2.33002752e-01 4.92512994e-02 5.52233048e-02
-1.68622494e-01 3.13347280e-01 -1.53551948e+00 5.13921678e-01
8.00005198e-01 1.15541017e+00 4.24036860e-01 7.71037221e-01
4.59797591e-01 9.96692955e-01 -6.84072256e-01 -9.91103146e-03
6.05569780e-02 -6.96256995e-01 -6.42361820e-01 9.64352712e-02
4.55815852e-01 -6.31466925e-01 -7.42444754e-01 1.03234422e+00
8.94097209e-01 1.15766719e-01 6.01441681e-01 -6.31628513e-01
-9.19344187e-01 6.75641656e-01 -3.88456494e-01 5.67095876e-01
-6.17158934e-02 -2.06179976e-01 4.61053014e-01 -1.07982647e+00
2.42384061e-01 8.31875682e-01 7.42394924e-01 6.13668740e-01
-8.88866842e-01 -7.70962894e-01 -2.72592813e-01 3.89214694e-01
-1.35133839e+00 -4.47611660e-01 8.95394206e-01 -4.52152103e-01
4.80218440e-01 2.19437197e-01 3.45814914e-01 1.09401274e+00
3.02642763e-01 1.01837587e+00 1.34567893e+00 -4.36349452e-01
2.03143939e-01 -2.63843149e-01 -2.39288453e-02 9.59497511e-01
-2.79163737e-02 2.07273677e-01 -6.77422822e-01 6.33159727e-02
1.01803505e+00 5.42253256e-03 -7.17523813e-01 -1.88807491e-02
-1.08766186e+00 6.13194764e-01 7.38602519e-01 3.91822040e-01
-6.27817035e-01 1.00212209e-01 3.93316485e-02 7.41230994e-02
4.88610655e-01 4.44435865e-01 5.46507686e-02 2.10357130e-01
-1.08561242e+00 -4.31438446e-01 3.72480035e-01 3.15632790e-01
4.79483157e-01 3.71020913e-01 -5.12439609e-01 1.07933891e+00
-2.69268360e-02 4.45126921e-01 8.02684128e-01 -7.12215781e-01
2.86703017e-02 2.20017321e-02 -2.00840384e-01 -1.16676044e+00
-3.71611476e-01 -8.72973740e-01 -1.10923362e+00 1.60441529e-02
-2.26130802e-02 -1.09114602e-01 -1.34809804e+00 1.58224881e+00
1.95622757e-01 9.71113443e-01 4.57214564e-01 1.37135291e+00
1.25544739e+00 7.16917157e-01 -7.18191490e-02 1.71583786e-03
1.41990685e+00 -1.01370263e+00 -7.87942648e-01 -2.96006531e-01
2.00059399e-01 -6.73427165e-01 5.93964458e-01 4.15873170e-01
-9.33753252e-01 -7.59640872e-01 -1.24206853e+00 1.59736037e-01
2.72524714e-01 3.82130861e-01 5.30635059e-01 5.76611638e-01
-9.75798130e-01 5.93596041e-01 -1.18577659e+00 -3.58070619e-02
6.34736359e-01 2.31851041e-01 -3.44597131e-01 -5.94840467e-01
-1.13609111e+00 8.45826447e-01 7.53231347e-02 5.13530850e-01
-1.22594666e+00 -7.98898697e-01 -1.11022711e+00 1.84803292e-01
3.43460709e-01 -4.45245475e-01 1.08741081e+00 -6.00234687e-01
-1.51415789e+00 5.88321269e-01 3.73207629e-02 -4.65218991e-01
3.18496913e-01 -3.09847951e-01 -5.17227650e-01 6.71215057e-01
5.41144842e-03 5.62987566e-01 1.37179959e+00 -1.32911742e+00
-2.67843846e-02 -2.06281066e-01 -2.69020885e-01 1.51284300e-02
-3.08481634e-01 -2.67828137e-01 -5.04667222e-01 -1.28256202e+00
4.79972124e-01 -8.42016697e-01 -8.89673904e-02 -1.99017182e-01
-1.86008736e-01 2.12347507e-01 7.28817403e-01 -9.47147250e-01
1.18440771e+00 -2.28040814e+00 2.10373521e-01 2.39737071e-02
4.40368146e-01 4.83625919e-01 -3.90194684e-01 1.20542414e-01
-3.82997453e-01 -3.47904384e-01 -5.23135722e-01 -1.31641254e-01
-4.20019388e-01 1.79840937e-01 -1.53221235e-01 5.96628964e-01
4.56980675e-01 1.20341599e+00 -8.02933931e-01 -2.05498636e-01
3.28276128e-01 6.85135722e-01 -4.70616132e-01 6.17757857e-01
3.74799222e-01 7.00629056e-01 -2.86906272e-01 6.40574336e-01
7.66715884e-01 -4.58422273e-01 -2.00063344e-02 -6.43829346e-01
1.92216977e-01 -2.73317307e-01 -7.48591363e-01 1.59831774e+00
-5.50803602e-01 7.96355903e-01 2.96082646e-01 -1.37038946e+00
7.43247211e-01 4.08064425e-01 5.07306635e-01 -5.78050911e-01
3.19954991e-01 2.15848252e-01 -3.23220082e-02 -9.30922687e-01
1.24442592e-01 -5.68402588e-01 2.30934352e-01 4.10799831e-01
2.62332410e-01 -1.67770743e-01 -4.96356606e-01 -4.57541943e-02
1.11540592e+00 -1.15980275e-01 1.68785557e-01 -2.02659711e-01
5.05598187e-01 -2.66617179e-01 1.95874527e-01 1.07357717e+00
-1.47715166e-01 1.18367207e+00 -1.98804453e-01 -2.91055143e-01
-5.55311501e-01 -8.18578780e-01 -1.46524146e-01 8.77743602e-01
2.44290069e-01 2.06415966e-01 -8.63319933e-01 -6.39796674e-01
-3.26973528e-01 3.47586572e-01 -8.31343472e-01 -4.50949281e-01
-6.50221646e-01 -1.05107403e+00 5.80350876e-01 5.97199619e-01
7.54803836e-01 -9.25016880e-01 -3.23774368e-01 3.73074710e-01
-4.90016490e-01 -1.27975297e+00 -4.73289371e-01 7.68315122e-02
-7.93389618e-01 -8.73810649e-01 -1.33207107e+00 -1.09744430e+00
8.65161538e-01 5.97670674e-01 6.78671777e-01 -2.95468066e-02
-6.03999317e-01 4.73374546e-01 -5.32196224e-01 -1.14266939e-01
-7.02370822e-01 -3.95143747e-01 -6.93205670e-02 4.00659293e-01
-4.12141392e-03 -6.11236989e-01 -1.06371105e+00 3.71446103e-01
-1.26127756e+00 -9.14006867e-03 1.13990605e+00 1.17512381e+00
2.27114126e-01 1.34637892e-01 5.92295706e-01 -6.99987471e-01
3.14712942e-01 -5.63358545e-01 -7.66113251e-02 1.20836914e-01
-2.57851481e-01 1.21550657e-01 2.87789762e-01 -6.19287193e-01
-1.07806623e+00 -2.09399894e-01 -3.91369283e-01 -7.05946684e-01
-6.05482422e-02 7.45670557e-01 2.71273196e-01 -4.07984555e-01
5.44531286e-01 7.76454926e-01 4.13781166e-01 -6.30835593e-01
2.84562092e-02 8.48487020e-01 9.94123101e-01 -2.71032602e-01
6.06318891e-01 6.32906735e-01 -1.76440939e-01 -1.00901115e+00
-1.04359341e+00 -3.94481659e-01 -1.54386312e-01 -3.13325673e-01
8.93782079e-01 -1.22673738e+00 -4.94161785e-01 7.98530936e-01
-7.31015265e-01 -2.08202437e-01 3.91773209e-02 7.88495362e-01
-2.52106756e-01 6.50922298e-01 -8.24765086e-01 -4.97013062e-01
-4.12596554e-01 -1.38034225e+00 9.61449385e-01 2.46521398e-01
1.13578707e-01 -6.47414327e-01 -2.54746467e-01 6.54521108e-01
7.19858050e-01 -4.20689322e-02 7.40883648e-01 -4.30330992e-01
-5.50018668e-01 -1.51569232e-01 -4.27204281e-01 7.33557522e-01
3.21996748e-01 -8.62756491e-01 -1.19512630e+00 -4.15135145e-01
4.29219872e-01 -3.11412543e-01 1.19584501e+00 6.99105024e-01
1.44331276e+00 -1.99481398e-01 -1.71894327e-01 7.00540006e-01
1.20308256e+00 5.06524071e-02 7.55964339e-01 1.75787508e-01
5.40708244e-01 4.69443858e-01 3.88363510e-01 2.27990508e-01
9.96102914e-02 5.28943062e-01 1.85736969e-01 -4.67130154e-01
-8.30072701e-01 -3.76529694e-02 7.29505792e-02 1.16521978e+00
1.04740605e-01 -2.34972034e-02 -7.65063584e-01 6.65672600e-01
-1.49507535e+00 -6.22093499e-01 3.06607395e-01 1.71593332e+00
8.34319174e-01 -3.74009103e-01 -5.94463885e-01 -7.07003567e-03
7.76484311e-01 3.72009426e-01 -4.91582870e-01 1.45481467e-01
5.24321832e-02 6.59616113e-01 1.27215087e-01 4.14323747e-01
-1.31135559e+00 7.11743832e-01 6.37808418e+00 9.72483158e-01
-1.43856025e+00 4.02655274e-01 6.70732737e-01 3.86446357e-01
1.77417174e-01 -6.54701233e-01 -1.84669092e-01 3.66979629e-01
5.91312945e-01 4.13677573e-01 2.71416068e-01 6.58337891e-01
-3.05107562e-03 -3.10466960e-02 -8.99678946e-01 1.08618772e+00
3.88390720e-01 -1.41728532e+00 -7.14072809e-02 -3.73424500e-01
9.04383183e-01 -9.88292024e-02 3.06037188e-01 2.54004925e-01
-6.67784736e-02 -1.17297053e+00 4.56705928e-01 3.85107666e-01
9.41698968e-01 -3.47898304e-01 9.67711270e-01 1.35156527e-01
-7.28838742e-01 -4.25116345e-02 -1.14821292e-01 1.49577588e-01
-4.00444493e-02 6.77392483e-01 -9.13456500e-01 6.46117151e-01
7.46252954e-01 6.28653705e-01 -3.36534917e-01 9.47824538e-01
-2.08300367e-01 8.95259023e-01 -3.93512249e-02 3.79058778e-01
2.70885438e-01 3.47456783e-02 5.46349466e-01 1.35427678e+00
4.93552893e-01 4.67259407e-01 4.38450277e-02 6.80451989e-01
-1.07195042e-01 -3.94834101e-01 -1.57183722e-01 2.30517182e-02
3.61142963e-01 8.94522607e-01 -6.70070529e-01 -3.14445287e-01
-2.87333816e-01 1.33453906e+00 -1.30863696e-01 5.75227857e-01
-6.64759159e-01 -3.97116661e-01 4.90306079e-01 -1.74932092e-01
9.36550140e-01 -1.58802956e-01 7.99510553e-02 -1.14517295e+00
3.80781740e-02 -8.92940462e-01 3.03866386e-01 -8.56146514e-01
-1.26829636e+00 8.63081515e-01 -1.95622340e-01 -1.36003041e+00
-1.30442716e-02 -5.59993029e-01 -4.78248835e-01 6.05318725e-01
-1.87213945e+00 -1.12742352e+00 -6.39475703e-01 3.08970004e-01
7.04518139e-01 -6.57525733e-02 7.62664378e-01 3.53511184e-01
-5.71437120e-01 6.00425422e-01 2.16061696e-01 3.53212953e-01
5.01803339e-01 -8.92985582e-01 8.25397149e-02 1.08547139e+00
-1.19633041e-01 7.42112637e-01 6.66462004e-01 -5.52527785e-01
-1.42761505e+00 -1.22672725e+00 1.37843192e-02 -7.45348111e-02
5.05857050e-01 1.56828500e-02 -1.15707290e+00 4.28728521e-01
1.66934319e-02 3.90382618e-01 7.23905802e-01 -5.49010515e-01
-2.78775275e-01 -9.75159779e-02 -9.10788655e-01 3.15099776e-01
6.45605803e-01 -6.16159022e-01 -7.78686523e-01 6.97750598e-02
5.21189392e-01 -7.20661640e-01 -8.47489119e-01 7.31230080e-01
6.17780209e-01 -8.57940435e-01 1.26736963e+00 -2.98420548e-01
8.21687579e-01 -2.44980499e-01 2.37334967e-02 -1.44632661e+00
-3.83177638e-01 -5.28934658e-01 -3.23015064e-01 8.61151636e-01
8.87464583e-02 -5.13988793e-01 8.63006473e-01 2.63284177e-01
-2.89597601e-01 -6.92600608e-01 -9.40079927e-01 -6.54529691e-01
-1.21269830e-01 -3.79594535e-01 1.93943769e-01 9.09350812e-01
-3.69477123e-01 -8.37171152e-02 -6.96243703e-01 7.14513838e-01
8.81341159e-01 1.81143552e-01 1.45395562e-01 -6.78127527e-01
-5.50348341e-01 -1.08252369e-01 -3.77924383e-01 -1.49506652e+00
-2.62506232e-02 -9.41232324e-01 5.16918302e-01 -1.46359360e+00
1.96402922e-01 -3.33935916e-01 -4.64197308e-01 3.87353271e-01
-6.90232992e-01 8.72816563e-01 -1.16081731e-02 2.96276063e-01
-3.31159532e-01 6.48160875e-01 1.38813782e+00 -2.30837420e-01
8.37800056e-02 7.63127441e-03 -8.17058444e-01 4.25630301e-01
3.92511219e-01 -5.19135296e-01 -1.17558070e-01 -6.29568815e-01
-5.76052010e-01 2.57048219e-01 3.70050430e-01 -1.16636443e+00
3.12368542e-01 3.75655204e-01 5.45410454e-01 -2.17477590e-01
4.23315018e-01 -7.68203259e-01 -2.42246836e-02 5.52678287e-01
-3.16973180e-01 -6.95480049e-01 2.59448022e-01 9.11596239e-01
-5.27451336e-01 -2.66414493e-01 7.60391951e-01 -1.90060899e-01
-8.21118891e-01 -1.05449576e-02 -6.00714922e-01 -3.52684587e-01
8.01464081e-01 -2.23187894e-01 -1.41898289e-01 -4.10824239e-01
-9.07779813e-01 -9.74045917e-02 -1.65638044e-01 3.74113232e-01
1.03111434e+00 -1.06004131e+00 -8.71436834e-01 5.74553430e-01
-6.85407594e-02 -1.10910974e-01 8.42400849e-01 1.08634889e+00
-4.32039529e-01 9.63529572e-02 -6.74282163e-02 -8.10871840e-01
-1.25281465e+00 3.98403436e-01 3.91419858e-01 1.29553378e-01
-1.00023723e+00 1.20791149e+00 4.38503593e-01 -1.29717693e-01
2.87311286e-01 -2.32675761e-01 -2.17470646e-01 -2.57561177e-01
8.97079468e-01 1.59367397e-01 3.82736534e-01 -4.07982588e-01
-3.00171047e-01 6.91437602e-01 -1.44983113e-01 1.34553269e-01
1.63244987e+00 -2.14221984e-01 -1.10411130e-01 -1.43368617e-01
1.52787673e+00 -8.51268247e-02 -1.36034644e+00 -6.20676517e-01
-2.63868600e-01 -4.32290077e-01 6.45733416e-01 -8.53833914e-01
-1.32666934e+00 9.34028327e-01 9.01506841e-01 -1.00550592e-01
1.46498024e+00 1.70216173e-01 1.02784455e+00 2.01016784e-01
7.68056810e-02 -4.05050904e-01 5.06739438e-01 7.16567859e-02
1.05562901e+00 -1.42409003e+00 -6.26301989e-02 -4.77724701e-01
-6.96045458e-01 1.12119615e+00 2.83125490e-01 -6.95932060e-02
5.88321567e-01 3.92613173e-01 3.95645648e-01 -3.64420891e-01
-5.23371510e-02 -2.14486480e-01 6.19907022e-01 5.51508188e-01
4.96175960e-02 -9.44627300e-02 4.86165211e-02 8.61749291e-01
8.10252503e-02 4.20748405e-02 4.28901553e-01 1.03745675e+00
-5.01117826e-01 -6.16432905e-01 -5.81149638e-01 3.48248869e-01
-6.28043413e-01 -2.59136170e-01 7.81174079e-02 3.08352590e-01
-1.83318272e-01 1.07239294e+00 -1.66314900e-01 -4.79998469e-01
1.26466408e-01 -6.83260858e-02 4.57843632e-01 -6.05059683e-01
-2.93087006e-01 3.80530894e-01 -2.20924258e-01 -5.01126230e-01
-4.30695653e-01 -2.36264199e-01 -9.14558470e-01 2.77538717e-01
-4.36200202e-01 2.61826396e-01 7.44522095e-01 1.02248013e+00
5.30861735e-01 8.98896039e-01 6.71960711e-01 -1.07875955e+00
-5.58721840e-01 -1.09855556e+00 -4.61841792e-01 2.74371386e-01
9.20342207e-01 -6.45402431e-01 -3.02449673e-01 1.63834646e-01] | [11.427007675170898, -2.4527745246887207] |
d0f0c158-f247-4abc-8564-1412d2f631ae | iterative-adversarial-attack-on-image-guided | 2305.13208 | null | https://arxiv.org/abs/2305.13208v1 | https://arxiv.org/pdf/2305.13208v1.pdf | Iterative Adversarial Attack on Image-guided Story Ending Generation | Multimodal learning involves developing models that can integrate information from various sources like images and texts. In this field, multimodal text generation is a crucial aspect that involves processing data from multiple modalities and outputting text. The image-guided story ending generation (IgSEG) is a particularly significant task, targeting on an understanding of complex relationships between text and image data with a complete story text ending. Unfortunately, deep neural networks, which are the backbone of recent IgSEG models, are vulnerable to adversarial samples. Current adversarial attack methods mainly focus on single-modality data and do not analyze adversarial attacks for multimodal text generation tasks that use cross-modal information. To this end, we propose an iterative adversarial attack method (Iterative-attack) that fuses image and text modality attacks, allowing for an attack search for adversarial text and image in an more effective iterative way. Experimental results demonstrate that the proposed method outperforms existing single-modal and non-iterative multimodal attack methods, indicating the potential for improving the adversarial robustness of multimodal text generation models, such as multimodal machine translation, multimodal question answering, etc. | ['Richang Hong', 'WenBo Hu', 'Youze Wang'] | 2023-05-16 | null | null | null | null | ['adversarial-attack', 'adversarial-text', 'multimodal-machine-translation'] | ['adversarial', 'adversarial', 'natural-language-processing'] | [ 5.89192390e-01 1.25047728e-01 4.55272317e-01 -1.54066645e-02
-1.29381120e+00 -1.08292818e+00 1.32363641e+00 -1.83506742e-01
-8.65648016e-02 5.51613092e-01 2.77894616e-01 -2.40681082e-01
3.11208963e-01 -7.88507462e-01 -1.03516233e+00 -7.57469118e-01
6.12950742e-01 5.59975922e-01 1.31537132e-02 -6.16792977e-01
1.26690000e-01 2.01236889e-01 -1.06411791e+00 7.76779473e-01
7.71609843e-01 8.12717855e-01 -2.27276936e-01 1.05883479e+00
-2.87043005e-01 7.77082086e-01 -6.78402543e-01 -1.30315363e+00
5.24496399e-02 -7.73267150e-01 -7.72878766e-01 -1.03313848e-02
5.67349195e-01 -3.84210825e-01 -4.89363790e-01 1.26959062e+00
7.86436915e-01 -1.09774433e-01 1.04120731e+00 -1.67995405e+00
-7.66552627e-01 8.35353911e-01 -4.19556350e-01 -4.45986211e-01
7.01711357e-01 4.66552675e-01 5.72392285e-01 -8.91557515e-01
6.97188318e-01 1.60835683e+00 3.35250318e-01 9.34228182e-01
-1.13761199e+00 -5.11378884e-01 -1.20904908e-01 1.58215135e-01
-8.86130452e-01 -3.66069227e-01 1.10636389e+00 -4.06928629e-01
3.05150867e-01 5.54760098e-01 -5.46044931e-02 2.01205111e+00
2.82377273e-01 1.10342205e+00 1.06555927e+00 -3.99203211e-01
-9.08767432e-02 2.34555602e-01 -5.23769557e-01 5.05298913e-01
-4.19796854e-01 2.10790604e-01 -5.56365907e-01 -2.79299110e-01
2.98653305e-01 -3.95565063e-01 -2.68595934e-01 -1.44820318e-01
-1.52346790e+00 1.00920951e+00 2.81158715e-01 1.18139707e-01
-1.21294476e-01 1.39138177e-01 5.67703962e-01 4.10438478e-01
7.83004984e-02 4.09030527e-01 2.67675549e-01 1.67648181e-01
-6.57647192e-01 3.92059177e-01 6.32691979e-01 8.34643841e-01
3.86134535e-01 2.31720045e-01 -4.66424882e-01 5.89881361e-01
4.19687599e-01 1.11780024e+00 3.38231504e-01 -4.90772605e-01
1.10683966e+00 5.75981081e-01 -8.11587945e-02 -1.09795642e+00
-1.25606894e-01 4.51242417e-01 -1.16923952e+00 3.39088976e-01
5.20471632e-01 -5.22524178e-01 -1.00795186e+00 1.81274569e+00
3.20003211e-01 -1.23137176e-01 5.18539786e-01 8.21433187e-01
1.08363950e+00 8.20665300e-01 1.62604600e-01 1.39826685e-01
1.29778969e+00 -8.17372143e-01 -8.29785824e-01 -6.74764737e-02
1.95960328e-01 -1.26016438e+00 1.01237285e+00 2.47854501e-01
-1.19719744e+00 -5.09819806e-01 -8.76000226e-01 -6.78268522e-02
-7.97664464e-01 -1.17982671e-01 7.80549645e-02 7.63544440e-01
-6.62615955e-01 1.09209277e-01 -3.60751510e-01 -2.69690514e-01
1.81204721e-01 1.57785937e-01 -5.49212098e-01 -2.00151071e-01
-1.63652205e+00 9.10839260e-01 4.83647346e-01 1.90774709e-01
-1.16353273e+00 -2.47069314e-01 -9.73730922e-01 -2.89165884e-01
2.43194029e-01 -1.00456488e+00 9.80958819e-01 -1.16704583e+00
-1.71019876e+00 5.81305146e-01 2.23394260e-01 -4.35629129e-01
9.52460110e-01 -1.36598811e-01 -3.15316200e-01 5.34381151e-01
-3.41831565e-01 9.23377335e-01 1.71343338e+00 -1.80280221e+00
-1.03114441e-01 -3.68494362e-01 3.05420496e-02 4.18409795e-01
-3.37431043e-01 4.23960760e-02 -1.81718290e-01 -8.77545774e-01
-2.22924307e-01 -1.00272477e+00 1.10861011e-01 -3.41202050e-01
-8.89222264e-01 1.87122464e-01 1.23612690e+00 -9.16917682e-01
6.80119097e-01 -1.87273371e+00 6.33953393e-01 1.52626812e-01
-8.70228186e-02 2.36014873e-01 -3.62200499e-01 8.48191679e-01
-1.97948247e-01 3.90441000e-01 -3.07199895e-01 -5.59796274e-01
4.46756154e-01 -9.32847410e-02 -8.61263692e-01 1.37690544e-01
5.13814688e-01 1.34301507e+00 -5.52779198e-01 -6.17158771e-01
4.03139740e-01 4.99593377e-01 -2.70232335e-02 4.41421658e-01
-4.87849623e-01 7.57614553e-01 -2.55606383e-01 8.74716938e-01
6.45904124e-01 3.53551924e-01 -1.87587202e-01 -4.45682734e-01
4.16969121e-01 -7.07384348e-01 -7.50502646e-01 1.71082115e+00
-1.17264159e-01 7.10442424e-01 4.28520776e-02 -8.64458203e-01
7.43749082e-01 6.95832253e-01 9.57932731e-04 -6.47864401e-01
4.80152547e-01 3.02625913e-02 -3.28589261e-01 -7.88508534e-01
6.82321668e-01 -2.51056939e-01 -5.43900430e-01 4.92240041e-01
3.73102189e-03 -6.83103025e-01 6.63055256e-02 5.58672190e-01
8.55548799e-01 1.89512193e-01 -4.23053890e-01 5.89604795e-01
7.74887860e-01 -8.83859918e-02 -3.40924442e-01 8.13288093e-01
-9.63219553e-02 1.01389229e+00 6.36419713e-01 7.13305846e-02
-1.27599072e+00 -1.32538676e+00 3.45346570e-01 7.71650791e-01
2.14954153e-01 5.91560677e-02 -1.11088085e+00 -1.02445924e+00
-3.95416319e-01 7.71735132e-01 -6.75516605e-01 -4.61665720e-01
-4.34115231e-01 -5.84782302e-01 1.23533678e+00 1.38683796e-01
6.33339345e-01 -1.14421105e+00 9.79666412e-02 -1.04600504e-01
-8.26366186e-01 -1.28992915e+00 -5.24147868e-01 -4.57467973e-01
-4.28399205e-01 -1.02357352e+00 -1.04013419e+00 -7.06647813e-01
7.44963348e-01 -2.26566404e-01 8.34476233e-01 -1.86412662e-01
-8.57656151e-02 8.82343769e-01 -6.04373693e-01 -4.07155216e-01
-1.22878790e+00 -9.66773629e-02 -2.08064333e-01 4.94831711e-01
-2.97126621e-01 -3.46592784e-01 -3.41527343e-01 2.78510302e-01
-1.63595867e+00 2.36019582e-01 6.79317176e-01 1.09809220e+00
1.47894204e-01 -1.69621140e-01 7.28124857e-01 -6.67593837e-01
9.27182138e-01 -5.25662422e-01 -2.97076046e-01 7.15170264e-01
1.17148928e-01 -1.01535199e-02 7.83246756e-01 -8.74868274e-01
-1.40517747e+00 4.82468940e-02 -2.44735613e-01 -6.81354463e-01
-4.54171479e-01 4.14564818e-01 -5.59201837e-01 -9.55318063e-02
6.71353877e-01 5.02581537e-01 8.63166675e-02 1.44000337e-01
8.53579342e-01 4.08919364e-01 7.02369750e-01 -6.78312242e-01
1.35733664e+00 4.98350888e-01 1.46478519e-01 -5.24608374e-01
-8.78468305e-02 2.48391092e-01 -4.17701244e-01 -6.29549026e-01
1.24756837e+00 -5.39805591e-01 -6.63745940e-01 8.87504518e-01
-1.51422572e+00 -3.93342860e-02 9.94242653e-02 1.96417019e-01
-8.07153344e-01 8.07662785e-01 -5.56378186e-01 -7.52995253e-01
-5.07756889e-01 -1.36319256e+00 1.37584698e+00 1.63530022e-01
6.40095621e-02 -1.14959741e+00 2.92438883e-02 9.70373273e-01
2.17610881e-01 8.47175956e-01 1.08883011e+00 -7.44898379e-01
-6.51814878e-01 -4.23406154e-01 -1.26992300e-01 3.35661650e-01
-1.90375164e-01 9.52170342e-02 -8.62982869e-01 -1.30058169e-01
-1.64875761e-01 -8.99758577e-01 6.17255628e-01 -1.38915643e-01
5.92668056e-01 -3.49983275e-01 1.18033253e-01 2.58042395e-01
1.20850217e+00 -3.69167887e-02 9.87931728e-01 1.11625768e-01
9.21174586e-01 8.50661039e-01 6.78048074e-01 4.83206138e-02
2.30240479e-01 4.01750445e-01 1.03008866e+00 -2.53666133e-01
2.18092613e-02 -3.82477015e-01 8.11651528e-01 4.23766404e-01
3.20262045e-01 -8.08651149e-01 -8.47107410e-01 3.17076772e-01
-2.08262610e+00 -1.27923751e+00 -2.22232044e-01 1.85408652e+00
7.29650378e-01 -7.73110986e-02 -1.65689617e-01 1.20466240e-01
1.00500321e+00 8.84788856e-02 -5.62632084e-01 -6.26811981e-01
-4.09846187e-01 -1.44196033e-01 7.06872940e-02 3.63385439e-01
-1.31302285e+00 1.01702583e+00 5.39010572e+00 1.05685413e+00
-8.47231627e-01 -2.76501309e-02 5.00959575e-01 2.46227220e-01
-5.14657617e-01 -2.37040654e-01 -3.20170581e-01 3.90082479e-01
7.77429581e-01 4.94569615e-02 4.17664081e-01 4.91571039e-01
8.39865506e-02 1.14677019e-01 -1.04496193e+00 9.36195970e-01
5.95811844e-01 -1.08576739e+00 7.82052338e-01 -1.97287112e-01
9.50395346e-01 -6.42033815e-01 4.67918694e-01 3.23072404e-01
2.50536680e-01 -1.07682145e+00 8.50711286e-01 8.56620252e-01
4.90760624e-01 -9.26716030e-01 7.52287924e-01 4.39236343e-01
-8.25945914e-01 5.17979339e-02 1.14774354e-01 6.82992041e-01
5.10804892e-01 -2.47613434e-02 -7.76402056e-01 9.99000847e-01
4.04884368e-01 -4.87354510e-02 -7.23215580e-01 3.93533140e-01
-1.94563478e-01 2.42626771e-01 -4.07706723e-02 1.69932187e-01
2.28704989e-01 -9.66432393e-02 9.24966455e-01 9.95420456e-01
4.00018245e-01 -3.35990727e-01 -1.02727816e-01 9.07617569e-01
-2.24997967e-01 1.15099631e-01 -1.04363275e+00 -4.30148065e-01
-6.85517043e-02 1.23819423e+00 -4.58507180e-01 -2.90122062e-01
-1.97643816e-01 1.41561282e+00 -1.75196663e-01 5.42046905e-01
-1.34310246e+00 -4.08438474e-01 1.01157151e-01 -4.40538496e-01
3.48111824e-03 -1.70851469e-01 -2.97260255e-01 -1.16617751e+00
-8.56591836e-02 -1.47096550e+00 3.67383689e-01 -1.25199950e+00
-1.47289467e+00 6.78375781e-01 -3.45798731e-02 -1.33784139e+00
-5.12548625e-01 -5.26162982e-01 -7.33994305e-01 8.22174728e-01
-1.13708580e+00 -2.10015893e+00 -1.50019586e-01 1.27185822e+00
6.37108088e-01 -4.25872266e-01 7.03758419e-01 8.23849365e-02
-4.77183551e-01 9.80313122e-01 -6.48806468e-02 2.51884550e-01
9.70078826e-01 -1.07630312e+00 1.90559939e-01 1.17426491e+00
5.37848100e-02 2.46668264e-01 6.95942581e-01 -7.05944002e-01
-1.64982009e+00 -1.07203281e+00 2.79263645e-01 -6.18101537e-01
8.16864014e-01 -3.69729012e-01 -6.27268612e-01 5.43274820e-01
8.40747893e-01 -7.72083759e-01 6.09800220e-01 -7.69902647e-01
-3.39113146e-01 2.61762202e-01 -1.28267074e+00 1.19266677e+00
5.25024295e-01 -6.41842008e-01 -6.16109371e-01 4.15117055e-01
7.70002067e-01 -5.45148373e-01 -8.36724639e-01 4.75258797e-01
3.15967739e-01 -8.16419840e-01 1.03389013e+00 -6.79056168e-01
8.71369898e-01 -2.23786652e-01 -2.31781930e-01 -1.23533809e+00
5.81617296e-01 -9.92812335e-01 -1.86008558e-01 1.58415437e+00
5.62495589e-01 -3.97381902e-01 4.86781120e-01 6.94454968e-01
5.24503440e-02 -7.62993395e-02 -8.72571588e-01 -2.91448623e-01
2.70747006e-01 -4.50562686e-01 4.79776531e-01 9.79317486e-01
-1.25756666e-01 3.98612320e-01 -8.27217996e-01 4.79049087e-01
6.47281706e-01 -2.01740861e-01 1.11275721e+00 -4.79804695e-01
-3.58910933e-02 -4.51539457e-01 -2.86860138e-01 -5.10597885e-01
2.62794703e-01 -7.22261846e-01 8.99617523e-02 -1.33588171e+00
1.16323112e-02 2.91354269e-01 2.87680447e-01 3.03791076e-01
-4.27559763e-01 5.73911786e-01 5.23828030e-01 1.20249931e-02
-2.78999746e-01 8.53045821e-01 1.39128542e+00 -6.08111680e-01
2.65489459e-01 -1.59823194e-01 -4.19594020e-01 6.28536761e-01
8.10676634e-01 -3.31921160e-01 -5.44340372e-01 -4.45622772e-01
3.26233864e-01 2.81237930e-01 7.13270128e-01 -7.35687912e-01
2.93395579e-01 -1.74527109e-01 3.78456563e-01 -6.68294728e-01
6.14647150e-01 -9.36951935e-01 1.32927313e-01 2.19166875e-01
-4.62605298e-01 3.86365950e-02 2.83902347e-01 6.59773767e-01
-3.69385451e-01 -2.77790129e-01 6.62238657e-01 -3.24007869e-02
-3.65247756e-01 1.67534828e-01 -4.34037566e-01 -4.45483401e-02
1.15709960e+00 -5.67774661e-02 -7.37744153e-01 -7.81741619e-01
-1.03233790e+00 2.06689462e-01 2.13964045e-01 7.58419514e-01
9.31302130e-01 -1.58846545e+00 -8.99254560e-01 -1.93524286e-01
3.49436738e-02 -2.26503968e-01 5.16302526e-01 6.46907270e-01
-4.22486573e-01 5.14778681e-02 -1.20786875e-01 -5.93902588e-01
-1.44492769e+00 7.05314040e-01 3.68576676e-01 -5.23255169e-01
-2.71681529e-02 5.10731220e-01 3.70808363e-01 -5.71661830e-01
1.41887456e-01 4.01543915e-01 -1.83615848e-01 1.82627484e-01
4.25516516e-01 2.29087800e-01 -1.89373478e-01 -8.80830109e-01
1.66056171e-01 5.03867686e-01 -6.35844795e-03 -7.68815935e-01
6.20621085e-01 -3.33566427e-01 -1.39181361e-01 2.11977363e-01
1.02296305e+00 -1.31576151e-01 -9.38742876e-01 1.38942967e-03
-4.96594191e-01 -1.91419169e-01 -4.91806269e-01 -1.02239835e+00
-8.50651324e-01 1.18231165e+00 7.14810073e-01 4.01120871e-01
1.20078325e+00 -1.78379551e-01 1.06231511e+00 4.07255262e-01
-1.96176976e-01 -1.01327598e+00 8.29720438e-01 4.62198883e-01
1.60749376e+00 -1.48162758e+00 -3.79662633e-01 -1.08469300e-01
-1.01258647e+00 1.20012593e+00 7.40966618e-01 3.02837670e-01
9.68232229e-02 -8.88031162e-03 2.91067481e-01 9.27387625e-02
-4.18835402e-01 -6.84466213e-02 4.33143854e-01 7.75460720e-01
-2.29406729e-01 -2.64744341e-01 3.62315588e-02 3.92918199e-01
-8.37434679e-02 -5.79183102e-01 4.10273254e-01 7.75800765e-01
4.72629331e-02 -1.38106108e+00 -1.16522479e+00 -1.61054894e-01
-3.93513441e-01 -2.52494305e-01 -1.09489274e+00 7.05853105e-01
6.18892955e-04 1.18675637e+00 -4.66530234e-01 -6.78110659e-01
2.33915851e-01 4.12177652e-01 4.94126469e-01 1.55463284e-02
-7.64213264e-01 -5.19624464e-02 1.37358770e-01 -1.34006143e-01
-3.45025629e-01 -5.32667220e-01 -8.81681383e-01 -4.43542153e-01
-2.61547685e-01 -2.70973414e-01 1.04856920e+00 1.08882022e+00
1.57283112e-01 3.92930061e-01 8.84305239e-01 -1.05599582e+00
-4.12186980e-01 -8.73322845e-01 8.34701061e-02 7.92863369e-01
1.57853857e-01 -1.66022673e-01 -3.65662277e-01 5.15778363e-01] | [11.16129207611084, 1.2483168840408325] |
c35f0bfc-d3ca-42bb-b6d8-237bbf2ad556 | incorporating-word-and-subword-units-in | 1908.05925 | null | https://arxiv.org/abs/1908.05925v2 | https://arxiv.org/pdf/1908.05925v2.pdf | Incorporating Word and Subword Units in Unsupervised Machine Translation Using Language Model Rescoring | This paper describes CAiRE's submission to the unsupervised machine translation track of the WMT'19 news shared task from German to Czech. We leverage a phrase-based statistical machine translation (PBSMT) model and a pre-trained language model to combine word-level neural machine translation (NMT) and subword-level NMT models without using any parallel data. We propose to solve the morphological richness problem of languages by training byte-pair encoding (BPE) embeddings for German and Czech separately, and they are aligned using MUSE (Conneau et al., 2018). To ensure the fluency and consistency of translations, a rescoring mechanism is proposed that reuses the pre-trained language model to select the translation candidates generated through beam search. Moreover, a series of pre-processing and post-processing approaches are applied to improve the quality of final translations. | ['Yan Xu', 'Pascale Fung', 'Zihan Liu', 'Genta Indra Winata'] | 2019-08-16 | incorporating-word-and-subword-units-in-1 | https://aclanthology.org/W19-5327 | https://aclanthology.org/W19-5327.pdf | ws-2019-8 | ['unsupervised-machine-translation'] | ['natural-language-processing'] | [ 5.11821806e-01 -1.22788228e-01 -4.91831690e-01 -4.07221586e-01
-1.13337445e+00 -6.22937262e-01 7.54642248e-01 2.41464265e-02
-6.30153775e-01 8.21735859e-01 5.34480333e-01 -8.14146578e-01
4.12738025e-01 -6.33989990e-01 -7.45565891e-01 -3.79147917e-01
3.44683826e-01 7.38932550e-01 -3.58644813e-01 -3.35937083e-01
2.31078416e-01 9.27641243e-02 -7.07469881e-01 4.76512611e-01
1.33262920e+00 2.29504645e-01 4.19575691e-01 4.69343424e-01
-3.10072750e-01 2.47297063e-02 -2.13104293e-01 -9.12516236e-01
5.26390851e-01 -7.33569026e-01 -7.43907273e-01 -3.78982387e-02
2.34249517e-01 -2.72134393e-02 -5.14008068e-02 1.14716947e+00
6.66664004e-01 -1.01133622e-01 6.55950129e-01 -6.86312616e-01
-9.24854755e-01 9.62425649e-01 -4.56643194e-01 2.37599194e-01
1.27230167e-01 2.88974345e-01 1.30870652e+00 -1.44228578e+00
7.34362245e-01 1.20563626e+00 4.57062632e-01 5.78173041e-01
-1.27343762e+00 -4.84392345e-01 -1.08183645e-01 2.05375075e-01
-1.19520521e+00 -5.13528228e-01 3.97171825e-01 -1.55796021e-01
1.44828165e+00 1.27627507e-01 5.77442765e-01 1.31850791e+00
6.11443579e-01 6.55560672e-01 1.02063978e+00 -8.49876821e-01
1.22902796e-01 7.80349001e-02 -4.10874456e-01 4.42243755e-01
-8.17525387e-03 -5.02661057e-02 -6.42714024e-01 -1.74995154e-01
7.06844568e-01 -3.54650140e-01 4.47775908e-02 1.28765497e-02
-1.58634961e+00 8.76611590e-01 -6.55807788e-04 4.83493775e-01
-6.27446115e-01 -2.57524550e-01 4.14829522e-01 5.43587089e-01
7.30812132e-01 7.82566786e-01 -7.55499661e-01 -1.81446508e-01
-1.20165420e+00 -1.18567646e-01 5.29236615e-01 1.12578130e+00
6.82917237e-01 -1.38597235e-01 -4.52509075e-01 1.06247056e+00
1.96803913e-01 6.07850373e-01 9.21835124e-01 -4.47925955e-01
1.02264023e+00 2.97891378e-01 -1.35103509e-01 -4.96382415e-01
7.61567578e-02 -5.86177170e-01 -7.36787260e-01 -5.72701812e-01
-9.10401344e-03 -9.47461277e-02 -9.18251455e-01 1.74832785e+00
1.38257727e-01 2.92205587e-02 2.29157135e-01 8.69147420e-01
2.39912212e-01 9.18985069e-01 -3.16376463e-02 -3.89428765e-01
1.28421032e+00 -1.40393543e+00 -6.26243055e-01 -4.85802740e-01
8.66043508e-01 -1.29425800e+00 1.26010692e+00 -8.54967758e-02
-1.37807333e+00 -4.93904680e-01 -1.02362049e+00 -2.05617666e-01
-1.46925166e-01 4.16032553e-01 1.68014482e-01 3.15867186e-01
-1.03462648e+00 6.58134818e-01 -9.13206816e-01 -5.04198372e-01
2.29331609e-02 3.44093919e-01 -3.23534280e-01 -2.64292479e-01
-1.37106502e+00 1.31801236e+00 2.60352314e-01 1.62668213e-01
-5.29224038e-01 -3.78127187e-01 -8.15346003e-01 -1.42417043e-01
-2.32523203e-01 -9.71775711e-01 1.11651433e+00 -1.17286551e+00
-1.99599433e+00 1.00938261e+00 -5.49124122e-01 -4.73616451e-01
3.59195203e-01 -2.40378454e-01 -3.40898663e-01 -8.20098221e-02
2.76675671e-01 6.62222862e-01 6.47038221e-01 -5.33533573e-01
-5.98257244e-01 -1.72167867e-01 -5.70235968e-01 5.35207093e-01
-3.39951605e-01 4.71496195e-01 -4.99182463e-01 -9.06294942e-01
2.15713292e-01 -1.00068116e+00 -5.15145361e-01 -8.28341365e-01
-4.61256206e-01 -1.34963945e-01 -7.87982065e-03 -1.18794835e+00
1.30897534e+00 -1.87313259e+00 7.39632070e-01 1.42817035e-01
-3.08423579e-01 2.86873817e-01 -8.70109320e-01 7.93108821e-01
1.97492242e-01 2.38443434e-01 -4.93678749e-01 -7.49585629e-01
-3.24880667e-02 2.43818223e-01 -4.18194115e-01 2.35954240e-01
5.89476645e-01 1.25853622e+00 -7.93281376e-01 -4.75489289e-01
-8.94899070e-02 1.88695624e-01 -4.64664042e-01 2.04844549e-01
-2.48175070e-01 7.05588520e-01 -1.11439727e-01 4.94312853e-01
6.74009383e-01 2.33062536e-01 5.38103998e-01 1.77625611e-01
-2.36786902e-01 1.23818099e+00 -4.35177714e-01 2.08106780e+00
-8.62514377e-01 1.55324295e-01 -2.74612695e-01 -7.13345587e-01
9.11285222e-01 3.91603619e-01 1.09283082e-01 -8.04964364e-01
1.05670802e-01 8.16962600e-01 3.04965317e-01 -1.97302014e-01
6.10070348e-01 -2.60930747e-01 -8.72741118e-02 6.56346500e-01
3.14098209e-01 -3.00943553e-01 1.59791127e-01 -1.88974366e-01
1.00312066e+00 4.47932303e-01 2.07010970e-01 -3.83977234e-01
3.67781401e-01 2.40229741e-01 9.01948214e-01 3.14222872e-01
8.59624371e-02 7.41855860e-01 -1.33579835e-01 -2.73283839e-01
-1.51842976e+00 -9.87364888e-01 8.94175619e-02 1.10365462e+00
-2.39206418e-01 -5.69968939e-01 -8.15799236e-01 -6.07602417e-01
-3.16757470e-01 9.17562842e-01 -2.77513444e-01 -2.05930606e-01
-1.11721981e+00 -8.58365417e-01 6.11289024e-01 1.95992842e-01
1.91953838e-01 -1.01396227e+00 -2.67849024e-02 5.09361088e-01
-6.11313403e-01 -1.14312005e+00 -8.74201059e-01 2.18078747e-01
-1.14596379e+00 -3.89782429e-01 -7.87058055e-01 -1.12365103e+00
6.06259882e-01 2.38119345e-03 1.12127483e+00 -1.40568301e-01
3.92086059e-01 -4.33702886e-01 -4.35322195e-01 -1.02150239e-01
-8.31241369e-01 6.18967772e-01 3.79408389e-01 -1.18757762e-01
6.53909147e-01 -7.09094167e-01 -2.76932478e-01 1.81932434e-01
-5.30291080e-01 3.80830407e-01 1.03213847e+00 1.00833058e+00
8.06930721e-01 -5.86926222e-01 3.35896879e-01 -4.52995390e-01
8.59484494e-01 -2.49795124e-01 -4.41614330e-01 3.80328745e-01
-7.04392076e-01 1.99595958e-01 8.58616948e-01 -6.61835074e-01
-6.87453210e-01 -1.81277439e-01 -2.57313937e-01 -2.24372223e-01
2.49429062e-01 5.68938851e-01 -3.48236263e-01 3.09315503e-01
3.84904325e-01 6.24123514e-01 -3.03444326e-01 -6.16993546e-01
4.85818505e-01 9.41993296e-01 3.95342976e-01 -4.74688679e-01
1.03518963e+00 -2.86450714e-01 -4.20214713e-01 -1.74079895e-01
-5.80742419e-01 -2.44964942e-01 -1.01297176e+00 4.20690507e-01
8.94032359e-01 -9.25314128e-01 2.94247866e-01 1.67284101e-01
-1.57443666e+00 -4.16255146e-01 4.46203798e-02 9.78577971e-01
-6.19323790e-01 2.40234807e-01 -1.05084026e+00 -3.84343535e-01
-8.27478647e-01 -1.27960241e+00 1.02284443e+00 -1.74057856e-01
-3.90867949e-01 -8.05969834e-01 3.84924203e-01 3.29026669e-01
5.43603957e-01 -4.34294790e-01 1.17972994e+00 -7.64918268e-01
-4.79124039e-01 7.68914595e-02 9.34346542e-02 4.27008927e-01
2.09437013e-01 -2.33322233e-01 -3.40533942e-01 -3.95165175e-01
5.86838461e-02 -3.11195161e-02 8.11027467e-01 1.23306684e-01
2.73855627e-01 -5.58064222e-01 -1.27962753e-01 7.69762635e-01
1.15465832e+00 -8.74664113e-02 5.55694342e-01 5.51465333e-01
6.49489522e-01 4.71917838e-01 5.10554135e-01 -1.38466462e-01
5.28445959e-01 7.87051082e-01 -3.80242839e-02 7.26991147e-02
-1.82308495e-01 -6.15342796e-01 8.28878999e-01 1.71377635e+00
3.31633985e-02 -1.70990348e-01 -8.61767530e-01 6.07139170e-01
-1.70859325e+00 -6.53927565e-01 -6.58329530e-03 2.23407769e+00
1.28388011e+00 2.21983213e-02 -2.41071030e-01 -3.12680334e-01
8.80726695e-01 -1.32152945e-01 -2.41806239e-01 -9.21351254e-01
-2.87500858e-01 6.41243935e-01 5.96675873e-01 6.26938164e-01
-6.89182043e-01 1.65732193e+00 5.88946486e+00 1.01307726e+00
-1.25879061e+00 4.71158475e-01 4.22179401e-01 -1.01084802e-02
-7.26075411e-01 3.09624612e-01 -7.63346612e-01 4.16719109e-01
1.33097529e+00 -1.08283050e-01 8.11224699e-01 3.05871993e-01
3.26923579e-01 3.26819360e-01 -9.60961521e-01 6.65007353e-01
7.21665323e-02 -1.15387046e+00 1.28381625e-01 9.15136039e-02
8.90882432e-01 4.89542425e-01 7.08257109e-02 4.57843482e-01
3.57625246e-01 -9.77594256e-01 7.03415453e-01 1.14362203e-01
9.48124945e-01 -8.01418185e-01 7.06061184e-01 5.45525074e-01
-8.75656486e-01 1.75304949e-01 -6.80247903e-01 1.36418745e-01
2.84221232e-01 5.79138577e-01 -1.02280748e+00 8.99702370e-01
1.18507922e-01 5.53718507e-01 -3.54393244e-01 7.16651917e-01
-7.14487255e-01 9.98867273e-01 -1.55145556e-01 4.92816344e-02
3.94720793e-01 -6.29425645e-01 7.19660699e-01 1.38838446e+00
7.30967820e-01 -4.66628462e-01 2.08740365e-02 7.64592886e-01
-3.20750326e-01 7.39015818e-01 -2.47167587e-01 -2.57368803e-01
5.58823168e-01 1.15274322e+00 -3.71076345e-01 -3.83948892e-01
-2.83805877e-01 1.68572342e+00 4.93068039e-01 3.17294776e-01
-5.19620299e-01 -1.26618519e-01 6.92214310e-01 -1.25929475e-01
2.96280980e-01 -5.19187212e-01 -5.18947721e-01 -1.52299178e+00
1.76376164e-01 -1.17071307e+00 -9.37855318e-02 -4.33190495e-01
-1.28006518e+00 9.29226279e-01 -4.81679291e-01 -1.37686539e+00
-4.86010104e-01 -1.98440105e-01 -7.73658335e-01 1.35677409e+00
-1.55059803e+00 -1.44430590e+00 7.58569956e-01 2.62924749e-02
7.93073833e-01 -4.25829589e-01 1.02199483e+00 2.31457546e-01
-5.72956920e-01 8.49088490e-01 2.71318465e-01 1.62902310e-01
1.02552748e+00 -9.77100909e-01 1.26669300e+00 1.11185789e+00
3.93006653e-01 9.12373722e-01 6.04791939e-01 -7.70091355e-01
-1.45562947e+00 -1.43931353e+00 1.97604716e+00 -4.33806986e-01
6.28033400e-01 -6.62079990e-01 -7.51502395e-01 6.02347910e-01
7.02944815e-01 -4.79437828e-01 7.68450856e-01 1.01629324e-01
-3.81850988e-01 2.98586190e-01 -7.57962883e-01 9.20790255e-01
1.03387642e+00 -6.19147420e-01 -8.69380832e-01 3.64298970e-01
1.05589819e+00 -3.23074698e-01 -6.17497921e-01 4.57655072e-01
4.72552150e-01 -2.70476192e-01 5.31170964e-01 -8.82658541e-01
6.81165576e-01 -1.46628737e-01 -2.96848804e-01 -1.84190929e+00
-3.98487568e-01 -8.30655873e-01 1.96241140e-01 1.20474923e+00
9.52372670e-01 -4.94060725e-01 2.69483984e-01 -1.48669347e-01
-3.58599812e-01 -8.63919199e-01 -1.17168856e+00 -9.81891394e-01
6.18500173e-01 -2.48167515e-01 7.88785756e-01 9.62386370e-01
2.44225487e-01 7.74053872e-01 -4.92353499e-01 6.39940277e-02
2.24453703e-01 1.42034441e-01 5.86067975e-01 -6.01564884e-01
-5.28915644e-01 -4.95606422e-01 -1.31535586e-02 -1.11348844e+00
3.27619314e-01 -1.47370672e+00 1.18003629e-01 -1.63018692e+00
2.40396857e-01 -6.35099337e-02 -3.82950336e-01 2.87737459e-01
-3.91643316e-01 3.82052511e-01 1.73141971e-01 5.89405894e-01
-1.77653313e-01 7.98477769e-01 1.21930432e+00 -7.27051795e-02
-2.92269468e-01 7.18513411e-03 -4.76158649e-01 1.41463786e-01
9.81141686e-01 -6.82656944e-01 4.97363396e-02 -1.06585789e+00
1.86287448e-01 -4.19696197e-02 -3.24138045e-01 -4.77100492e-01
5.03308587e-02 -3.19979459e-01 6.97384402e-02 -4.41884875e-01
1.31295204e-01 -3.13348264e-01 -1.41990319e-01 4.15443540e-01
-5.18677950e-01 7.41322696e-01 1.47431746e-01 1.14489749e-01
-2.02629566e-01 -1.65972188e-01 6.22977018e-01 -8.35575983e-02
6.80599064e-02 3.86401236e-01 -4.74442810e-01 -9.31531265e-02
4.41095114e-01 8.72638077e-02 4.32191417e-02 8.24084878e-03
-2.75677919e-01 2.29607314e-01 6.20908737e-01 6.54203534e-01
4.32522714e-01 -1.63254929e+00 -1.27890766e+00 2.85256028e-01
1.55202687e-01 -3.72377217e-01 -2.64911145e-01 1.09384847e+00
-2.67588019e-01 3.89022917e-01 -1.97538421e-01 -2.92256117e-01
-9.04814124e-01 3.05320114e-01 2.26482805e-02 -6.29531145e-01
-4.30289567e-01 7.41047621e-01 -4.31646377e-01 -1.06721818e+00
-3.02318096e-01 -1.06404044e-01 2.86304772e-01 -3.55800986e-01
3.64909828e-01 -7.17178211e-02 3.83017898e-01 -7.56779015e-01
-3.84752989e-01 4.24219757e-01 -2.17811763e-01 -7.35142827e-01
1.30217528e+00 -2.90326655e-01 -4.27187383e-01 2.31997281e-01
1.01974189e+00 3.61958504e-01 -4.95737523e-01 -5.12629211e-01
3.13693076e-01 -1.60316825e-01 -2.36309439e-01 -8.93112719e-01
-4.53654438e-01 1.06660771e+00 1.30996525e-01 -6.13050997e-01
9.81956422e-01 -3.00058872e-01 1.30425704e+00 2.70256162e-01
2.98673272e-01 -1.32011271e+00 -3.38410467e-01 1.09473932e+00
7.02592671e-01 -9.73345399e-01 -5.05694866e-01 -1.81343466e-01
-5.65524578e-01 1.11636090e+00 2.95389801e-01 -5.71242236e-02
5.44896126e-02 1.21313371e-01 3.56240839e-01 4.92964447e-01
-9.97894108e-01 -9.61471498e-02 3.54530692e-01 4.03595746e-01
6.06433034e-01 2.09847093e-01 -1.14406514e+00 6.75997734e-01
-5.24157763e-01 -2.00427264e-01 1.88083440e-01 6.93194270e-01
-3.15623730e-01 -1.87780869e+00 -2.53369361e-01 1.56256303e-01
-4.40253437e-01 -9.53172863e-01 -5.54866433e-01 1.26075998e-01
5.68303578e-02 9.44892108e-01 4.29475531e-02 -7.09788561e-01
4.33440357e-02 4.94020611e-01 6.65639818e-01 -9.57902789e-01
-7.87389696e-01 3.86843473e-01 8.38156939e-02 -2.94229358e-01
6.48401538e-03 -5.49508154e-01 -9.54618633e-01 -1.50216997e-01
-2.93321788e-01 3.71996105e-01 9.67987299e-01 1.12495160e+00
6.37861013e-01 1.45172015e-01 8.69307339e-01 -6.84718728e-01
-8.77311707e-01 -1.44084048e+00 5.35434559e-02 1.71764761e-01
-1.22237273e-01 6.54327741e-04 -1.80287987e-01 -1.24617569e-01] | [11.634339332580566, 10.334495544433594] |
79e57172-b4c1-44d0-914e-8f3574b69e6b | amplitude-modulated-video-camera-light | null | null | http://openaccess.thecvf.com/content_cvpr_2016/html/Kolaman_Amplitude_Modulated_Video_CVPR_2016_paper.html | http://openaccess.thecvf.com/content_cvpr_2016/papers/Kolaman_Amplitude_Modulated_Video_CVPR_2016_paper.pdf | Amplitude Modulated Video Camera - Light Separation in Dynamic Scenes | Controlled light conditions improve considerably the performance of most computer vision algorithms. Dynamic light conditions create varying spatial changes in color and intensity across the scene. These condition, caused by a moving shadow for example, force developers to create algorithms which are robust to such variations. We suggest a computational camera which produces images that are not influenced by environmental variations in light conditions. The key insight is that many years ago, similar difficulties were already solved in radio communication; As a result each channel is immune to interference from other radio channels. Amplitude Modulated (AM) video camera separates the influence of a modulated light from other unknown light sources in the scene; Causing the AM video camera frame to appear the same - independent of the light conditions in which it was taken. We built a prototype of the AM video camera by using off the shelf hardware and tested it. AM video camera was used to demonstrate color constancy, shadow removal and contrast enhancement in real time. We show theoretically and empirically that: 1. the proposed system can produce images with similar noise levels as a standard camera. 2. The images created by such camera are almost completely immune to temporal, spatial and spectral changes in the background light. | ['Maxim Lvov', 'Rami Hagege', 'Amir Kolaman', 'Hugo Guterman'] | 2016-06-01 | null | null | null | cvpr-2016-6 | ['shadow-removal', 'color-constancy'] | ['computer-vision', 'computer-vision'] | [ 5.65453649e-01 -6.56119525e-01 4.97571468e-01 -1.25172868e-01
1.97617710e-01 -6.90177143e-01 4.24660742e-01 -4.80616331e-01
-4.73407954e-01 8.31921577e-01 -3.30975354e-01 -1.49700969e-01
4.08418357e-01 -6.07008874e-01 -7.09383130e-01 -9.19800699e-01
9.18218568e-02 -3.74868721e-01 7.69946754e-01 -9.83972698e-02
2.85169184e-01 4.26225245e-01 -2.02133584e+00 8.32978413e-02
5.15580714e-01 6.66538775e-01 7.56794572e-01 1.21565759e+00
-1.23463929e-01 9.26177561e-01 -9.67031837e-01 2.36645550e-01
5.17547429e-01 -5.93313158e-01 -9.58975926e-02 6.23657525e-01
5.02420723e-01 -3.03279936e-01 -2.98130423e-01 1.33363497e+00
1.37134507e-01 -4.75377339e-04 3.02763343e-01 -1.22490084e+00
-6.01909161e-01 -7.98882768e-02 -8.69085133e-01 3.07181209e-01
4.92206812e-01 2.40332752e-01 3.00962385e-02 -6.05576932e-01
4.53959197e-01 9.09729660e-01 4.52584803e-01 4.66051459e-01
-1.12270498e+00 -5.54958344e-01 -8.84430707e-02 2.68370420e-01
-1.18001604e+00 -4.87816870e-01 6.46219850e-01 -1.74565241e-02
6.11037016e-01 4.79425460e-01 5.91222882e-01 7.92819142e-01
7.47759700e-01 1.60093725e-01 1.70994079e+00 -7.99716234e-01
2.85629421e-01 4.86964136e-01 -4.75665368e-02 6.96024120e-01
4.85230863e-01 2.15742439e-01 -2.76355952e-01 2.31372029e-01
5.83050251e-01 -1.30890995e-01 -8.52916956e-01 -7.16793910e-02
-9.29806292e-01 1.58528164e-01 1.07690327e-01 4.59216803e-01
-1.26134306e-01 2.92501837e-01 -2.16621459e-01 5.49008608e-01
-8.56591165e-02 1.97118178e-01 -4.03036803e-01 3.96074988e-02
-5.51550806e-01 -3.59331280e-01 7.13489830e-01 8.36397529e-01
6.62939131e-01 2.04880148e-01 3.91995877e-01 6.60755396e-01
2.36459732e-01 1.23708940e+00 5.36859334e-01 -8.20577681e-01
3.86309773e-02 3.72838564e-02 2.48662621e-01 -1.10739911e+00
-3.48227262e-01 -1.71713322e-01 -4.63769108e-01 1.01044321e+00
5.08834004e-01 -3.53627235e-01 -1.18301845e+00 1.46938217e+00
2.83237725e-01 3.00971508e-01 3.53552401e-02 9.50007319e-01
5.40357888e-01 8.56953323e-01 -5.63950121e-01 -8.13953280e-01
1.09000194e+00 -5.06472588e-01 -1.13392878e+00 -1.02501839e-01
-9.22717303e-02 -1.45373511e+00 9.23496068e-01 8.52590322e-01
-8.65031660e-01 -7.81511366e-01 -1.31571341e+00 4.46815550e-01
-2.91646332e-01 -2.79348731e-01 3.86499643e-01 1.28776371e+00
-1.10449457e+00 1.38806224e-01 -4.30713892e-01 -3.77718121e-01
-1.98341072e-01 2.17329934e-01 -2.70905584e-01 -2.38326088e-01
-7.62302935e-01 1.04924357e+00 1.88767258e-02 2.34229162e-01
-6.59233749e-01 -1.88934281e-01 -3.28811705e-01 -3.82349044e-01
4.04869676e-01 -3.15935105e-01 1.03768349e+00 -1.83399343e+00
-1.79989958e+00 7.25769818e-01 -3.86829823e-01 -1.02676623e-01
3.90993267e-01 -6.58563673e-02 -9.53608692e-01 3.16822350e-01
-1.78346112e-01 -4.38451283e-02 1.25334835e+00 -1.62348354e+00
-7.22595036e-01 -1.46654382e-01 -5.39957546e-02 1.48993999e-01
2.20656574e-01 5.12858890e-02 -7.46019006e-01 -1.76003471e-01
-7.83229806e-03 -1.13586211e+00 5.49373925e-02 7.51165822e-02
-1.26747116e-01 7.56201088e-01 1.34808874e+00 -7.53541738e-02
7.10817635e-01 -2.17898655e+00 -5.30793846e-01 2.29835808e-01
-3.00674945e-01 4.45275813e-01 -1.53808566e-02 2.24435911e-01
-9.57886577e-02 -3.82285804e-01 -1.68675676e-01 2.74815619e-01
-6.47141457e-01 3.82263094e-01 2.42930930e-02 8.05149794e-01
-2.62612492e-01 1.24405608e-01 -5.66615045e-01 -4.85321403e-01
3.74224901e-01 5.93986928e-01 -8.87090415e-02 6.57879701e-03
2.26433054e-02 2.17245653e-01 -3.81307527e-02 5.91456175e-01
1.01547956e+00 3.41493577e-01 -4.48514484e-02 -8.69118422e-02
-3.93712133e-01 -6.90072358e-01 -1.43579245e+00 9.99083936e-01
-5.54266095e-01 1.35405695e+00 5.36117673e-01 -3.43158931e-01
9.12015200e-01 2.28976205e-01 4.12171409e-02 -1.02520609e+00
2.09174886e-01 2.99991071e-02 4.25492190e-02 -8.27152848e-01
5.15103698e-01 -1.60723895e-01 5.27995169e-01 2.63363659e-01
-5.63148797e-01 -3.96973401e-01 6.87109753e-02 1.23520680e-01
9.58408058e-01 -3.35762068e-03 1.91117123e-01 -1.42766565e-01
6.58309400e-01 -1.59111068e-01 4.46672082e-01 9.29444134e-01
-3.06168765e-01 5.56479990e-01 -1.62929595e-01 -2.98605710e-01
-7.57706523e-01 -1.05581343e+00 -2.44036227e-01 8.44631255e-01
7.77042270e-01 1.66067868e-01 -6.17476106e-01 1.15283079e-01
-2.41929248e-01 6.22008979e-01 -3.73689234e-01 1.42486885e-01
-4.63139087e-01 -9.23611462e-01 1.44356012e-01 -8.50331634e-02
7.67937243e-01 -7.54989207e-01 -1.11568499e+00 1.79035798e-01
1.67532399e-01 -1.25035834e+00 -1.54498413e-01 1.98342025e-01
-5.86997569e-01 -1.42693305e+00 -3.48992586e-01 -5.84172547e-01
7.19984591e-01 1.02709198e+00 9.13913012e-01 3.26973677e-01
-8.04425120e-01 8.83884192e-01 -4.79688585e-01 -7.69058228e-01
-5.19471407e-01 -1.02786934e+00 3.07304170e-02 3.36783916e-01
1.81781083e-01 -4.13113773e-01 -6.20660841e-01 4.11838442e-01
-1.07455409e+00 -7.13627599e-03 4.47777003e-01 3.46308827e-01
1.68015599e-01 7.47361422e-01 -1.84314977e-02 -8.60100687e-01
4.37573880e-01 7.20870122e-02 -9.34636414e-01 5.12750745e-02
-4.66201007e-01 -2.80455768e-01 8.16710234e-01 -3.54952604e-01
-1.58974051e+00 2.97167063e-01 5.32904565e-01 7.28191510e-02
-4.29071635e-01 -2.83020288e-01 -1.42371997e-01 -4.38567907e-01
8.10582519e-01 3.21532816e-01 -2.49642442e-04 5.10855229e-04
2.48783693e-01 8.03505242e-01 7.87610948e-01 -1.64570622e-02
1.19527876e+00 9.96908724e-01 3.06295305e-01 -1.66848898e+00
-3.49081069e-01 -3.42937022e-01 -3.30207318e-01 -6.74297810e-01
8.76235187e-01 -7.09604561e-01 -4.42554504e-01 8.56537104e-01
-1.06192493e+00 -1.97603002e-01 3.54239911e-01 5.68793178e-01
-1.45267490e-02 4.00556892e-01 -3.04888278e-01 -1.01153934e+00
2.03460500e-01 -9.55492318e-01 4.15058583e-01 8.24037969e-01
3.43641758e-01 -9.09492433e-01 -1.55191682e-02 1.83596149e-01
5.74397206e-01 2.02775940e-01 4.88308072e-01 5.92003882e-01
-9.16680038e-01 -6.86426163e-02 -1.77441299e-01 4.81310874e-01
5.52660942e-01 5.98583519e-01 -1.45650852e+00 -1.29971564e-01
3.42084110e-01 4.15224344e-01 7.50985622e-01 5.89201570e-01
5.76640964e-01 2.24807635e-01 -2.71776110e-01 6.25283182e-01
2.00979042e+00 7.76920855e-01 9.32913065e-01 5.27004600e-01
4.23505634e-01 3.13635528e-01 5.45186579e-01 8.73243660e-02
-4.13737088e-01 7.06746697e-01 5.89999676e-01 -6.15015447e-01
-2.93319166e-01 4.30900216e-01 5.18333614e-01 2.69790322e-01
-1.66065767e-01 -4.09964591e-01 -4.51284885e-01 2.01764256e-01
-1.17416751e+00 -1.14648592e+00 -9.29569840e-01 2.20313048e+00
5.89028299e-01 1.19889110e-01 -2.65585810e-01 2.87471652e-01
9.24082100e-01 -9.04005319e-02 -1.74720407e-01 -5.50619841e-01
-4.82867271e-01 2.64543355e-01 1.03637362e+00 8.14007878e-01
-9.38214362e-01 5.39999187e-01 6.68870163e+00 1.61234051e-01
-1.47176039e+00 -1.79239869e-01 2.61981934e-01 -1.92173108e-01
-6.33952394e-02 4.56192158e-02 -3.25732380e-01 7.78438151e-01
6.05537653e-01 -1.46260574e-01 4.11322176e-01 5.21003604e-01
5.54328620e-01 -1.07526290e+00 -6.89097464e-01 1.12102091e+00
4.32294309e-01 -6.86153710e-01 -3.04779291e-01 -3.62246372e-02
7.75318325e-01 -1.45965159e-01 2.20059857e-01 -4.08777267e-01
3.06904837e-02 -7.35340774e-01 3.95360798e-01 4.44531888e-01
5.29434621e-01 -6.08103991e-01 5.58704555e-01 8.10063928e-02
-1.02582848e+00 -8.63860846e-02 -5.92728376e-01 -3.15336764e-01
1.04088590e-01 6.31226957e-01 -8.07539344e-01 1.43088266e-01
6.06753588e-01 4.14983705e-02 -7.35452950e-01 1.15000629e+00
-1.74371108e-01 3.91386002e-01 -2.72467852e-01 -7.57826641e-02
1.25783950e-01 -3.99736404e-01 6.42789781e-01 1.27713096e+00
1.05765671e-01 2.25691363e-01 -1.81729108e-01 4.22052026e-01
3.82646114e-01 -2.24483490e-01 -7.84373343e-01 4.53266233e-01
1.79619282e-01 1.14399576e+00 -1.09076190e+00 -1.35427177e-01
-8.05398703e-01 1.32345939e+00 -7.57016838e-01 6.04906380e-01
-8.80986273e-01 -5.75703979e-01 4.96114254e-01 7.92465433e-02
1.87677369e-01 -2.50970751e-01 1.21098785e-02 -9.34988260e-01
-5.95600791e-02 -7.77804732e-01 -7.91922137e-02 -1.32420194e+00
-7.50507832e-01 5.48016250e-01 -3.18197697e-01 -1.31092155e+00
2.89674670e-01 -7.90244579e-01 -5.38147986e-01 8.46978009e-01
-1.54607236e+00 -5.85294366e-01 -5.91793656e-01 1.03062010e+00
7.03645110e-01 -2.18092352e-01 6.17468596e-01 1.24091566e-01
-3.21147770e-01 -1.83502380e-02 4.29157346e-01 -1.41092807e-01
8.74720573e-01 -1.14952374e+00 -2.58920521e-01 1.54118776e+00
2.86038995e-01 4.65369195e-01 1.22344601e+00 -4.11947250e-01
-1.67997766e+00 -6.65857971e-01 2.38741741e-01 -3.75769764e-01
3.72119784e-01 -1.85611024e-01 -7.33357131e-01 1.92099839e-01
4.94595289e-01 2.27319151e-02 4.11497027e-01 -4.33286756e-01
-1.04241654e-01 -5.54860055e-01 -1.14295769e+00 4.17102545e-01
4.41722542e-01 -4.02145177e-01 -5.35151005e-01 2.19022349e-01
3.14239204e-01 -2.07220420e-01 1.43943787e-01 -1.87621504e-01
8.54269147e-01 -1.55439126e+00 6.63105607e-01 2.99145073e-01
-2.04500253e-03 -8.86829913e-01 -9.63886529e-02 -1.19137919e+00
-1.16210498e-01 -8.87141943e-01 5.98142326e-01 9.49941397e-01
3.59562635e-01 -8.97739887e-01 3.54555666e-01 5.50108850e-01
1.26662955e-01 1.16124570e-01 -5.10149717e-01 -7.80658901e-01
-7.33596623e-01 -3.43976051e-01 3.12668178e-03 7.71902978e-01
-3.18642259e-01 2.67993718e-01 -5.47667801e-01 6.52017534e-01
8.03149104e-01 4.02459130e-02 9.77803946e-01 -1.13104427e+00
-5.58606744e-01 8.20305645e-02 -5.20187557e-01 -6.10945880e-01
-3.06879908e-01 -6.04335815e-02 1.53643206e-01 -1.35705602e+00
4.93077077e-02 -8.37738886e-02 -6.68535978e-02 -1.86033875e-01
-1.88285828e-01 6.18911684e-01 3.50728542e-01 5.98485097e-02
-2.57850528e-01 -2.04077646e-01 1.08058381e+00 1.85615987e-01
-2.79728085e-01 -6.78180251e-03 -2.54234105e-01 9.37030315e-01
6.38624728e-01 -3.87110978e-01 -7.11012721e-01 -4.57422495e-01
2.54011899e-01 -7.36287758e-02 1.26795426e-01 -1.33517361e+00
2.73243278e-01 -2.38361523e-01 7.39523292e-01 -2.91104853e-01
3.59691501e-01 -1.44023013e+00 4.20947582e-01 6.07666135e-01
2.83636987e-01 1.55153885e-01 2.45856851e-01 8.06797326e-01
1.20494119e-03 -3.80358219e-01 1.27800834e+00 -4.50693935e-01
-9.92217779e-01 -3.59739214e-01 -9.81343865e-01 -4.24510270e-01
1.24644709e+00 -8.65256727e-01 -6.68515503e-01 -7.15386093e-01
-1.86739907e-01 -2.97474295e-01 8.06818187e-01 1.54989481e-01
4.81928289e-01 -6.49841726e-01 -2.83726662e-01 4.74909365e-01
-2.60664433e-01 -5.49347222e-01 2.45604396e-01 6.11102223e-01
-1.19820118e+00 1.22950815e-01 -5.18618703e-01 -6.07501686e-01
-2.11806560e+00 6.26207113e-01 4.97146666e-01 6.74642742e-01
-7.26163983e-01 7.64922678e-01 2.63265371e-01 6.61928475e-01
1.07832581e-01 -3.32515687e-01 -5.93330078e-02 -3.71398330e-01
9.03607309e-01 3.09593588e-01 -8.21228698e-02 -3.91096264e-01
-2.92703956e-01 8.83183897e-01 1.62801281e-01 -2.75572270e-01
1.09353280e+00 -5.21169603e-01 -2.33364031e-02 7.15205431e-01
8.70446146e-01 5.35758674e-01 -9.86147881e-01 7.90749416e-02
-4.61963594e-01 -1.00534177e+00 2.91112632e-01 -8.11709046e-01
-1.01518857e+00 4.61102813e-01 1.06813490e+00 6.44329846e-01
1.75141156e+00 -5.47317266e-01 3.06047976e-01 3.78274888e-01
5.34335852e-01 -1.38376939e+00 9.52074602e-02 1.61876574e-01
3.36078495e-01 -1.03518510e+00 2.52962530e-01 -5.66205442e-01
-6.53508842e-01 1.45128417e+00 4.01109427e-01 -4.70479503e-02
3.66897076e-01 8.68238151e-01 7.44464040e-01 5.93741983e-02
-7.32460260e-01 -4.12298292e-01 -2.95383513e-01 1.04248238e+00
3.16169471e-01 -1.44589201e-01 -1.26146972e-01 -5.87859869e-01
1.59084380e-01 -3.93173434e-02 1.47801638e+00 9.82166708e-01
-8.52937877e-01 -8.33339512e-01 -1.13326943e+00 -1.37167305e-01
-6.74354374e-01 2.30277687e-01 -5.28025210e-01 9.84098494e-01
4.18404996e-01 1.38081503e+00 -4.59810048e-02 -3.29727083e-02
2.25989372e-01 -1.03835583e-01 7.00008392e-01 -2.41838992e-01
-1.19084649e-01 2.40999952e-01 -5.03372699e-02 -6.87787294e-01
-8.96606684e-01 -3.63651216e-01 -1.09741938e+00 -8.65724385e-02
-3.91828001e-01 -4.20158952e-02 1.07585859e+00 5.54857910e-01
-2.62251526e-01 5.67182720e-01 7.76017964e-01 -5.69359064e-01
2.95203388e-01 -5.97431362e-01 -1.02376521e+00 4.03435111e-01
7.31211543e-01 -4.67703670e-01 -7.70642936e-01 5.21378279e-01] | [10.514522552490234, -2.5910017490386963] |
edd0fa45-9eed-4797-a844-92e146f40dab | debiased-mapping-for-full-reference-image | 2302.11464 | null | https://arxiv.org/abs/2302.11464v2 | https://arxiv.org/pdf/2302.11464v2.pdf | Debiased Mapping for Full-Reference Image Quality Assessment | Mapping images to deep feature space for comparisons has been wildly adopted in recent learning-based full-reference image quality assessment (FR-IQA) models. Analogous to the classical classification task, the ideal mapping space for quality regression should possess both inter-class separability and intra-class compactness. The inter-class separability that focuses on the discrimination of images with different quality levels has been highly emphasized in existing models. However, the intra-class compactness that maintains small objective quality variance of images with the same or indistinguishable quality escapes the research attention, potentially leading to the perception-biased measures. In this paper, we reveal that such bias is mainly caused by the unsuitable subspace that the features are projected and compared in. To account for this, we develop the Debiased Mapping based quality Measure (DMM), which relies on the orthonormal bases of deep learning features formed by singular value decomposition (SVD). The SVD in deep learning feature domain, which overwhelmingly separates the quality variations with singular values and projection bases, facilitates the quality inference with dedicatedly designed distance measure. Experiments on different IQA databases demonstrate the mapping method is able to mitigate the perception bias efficiently, and the superior performance on quality prediction verifies the effectiveness of our method. The implementation will be publicly available. | ['Lingyu Zhu', 'Shiqi Wang', 'Hanwei Zhu', 'Baoliang Chen'] | 2023-02-22 | null | null | null | null | ['image-quality-assessment'] | ['computer-vision'] | [-1.21907435e-01 -3.28100622e-01 -1.08233035e-01 -3.77704978e-01
-5.90640366e-01 -2.98106164e-01 4.36256617e-01 -2.29810774e-01
-1.67519584e-01 4.28376645e-01 3.19148600e-01 3.89124416e-02
-6.51380301e-01 -7.82210410e-01 -4.28587824e-01 -9.78762388e-01
1.48430809e-01 -2.84852535e-01 -1.99684292e-01 -2.49390110e-01
3.44793230e-01 4.31093425e-01 -1.83243096e+00 1.84263751e-01
1.10847449e+00 1.57209933e+00 2.36152150e-02 1.29706755e-01
-1.11922966e-02 3.48154247e-01 -5.21037340e-01 -4.77828741e-01
5.93849003e-01 -4.42982584e-01 -5.11558890e-01 3.07205785e-02
4.93691325e-01 -4.29094374e-01 -5.47319531e-01 1.41658425e+00
6.88127160e-01 -9.63156521e-02 6.37225270e-01 -1.58099663e+00
-1.10300303e+00 1.13532275e-01 -4.66311991e-01 2.57960528e-01
3.47207099e-01 1.34237677e-01 1.01930034e+00 -1.06662095e+00
2.53710717e-01 1.30764830e+00 5.74593127e-01 8.74486417e-02
-1.04622591e+00 -5.96590459e-01 -2.14284003e-01 5.93034267e-01
-1.65343773e+00 -3.01362365e-01 9.63272214e-01 -6.23886049e-01
4.30219799e-01 4.30745631e-01 7.59790540e-01 9.46152270e-01
5.64903200e-01 4.42561179e-01 1.46310151e+00 -1.86946839e-01
5.80747575e-02 2.73558915e-01 3.37737873e-02 4.83564109e-01
3.20732921e-01 4.63643640e-01 -5.32214165e-01 1.41969785e-01
7.13690281e-01 1.10777959e-01 -6.86017156e-01 -5.52990437e-01
-1.35016298e+00 6.44579530e-01 5.41692615e-01 1.83928728e-01
-2.42431879e-01 -4.67155546e-01 3.78291965e-01 5.59895992e-01
1.77888840e-01 2.56777912e-01 -1.19031467e-01 -3.15769874e-02
-7.69394338e-01 -7.32060298e-02 2.51955986e-01 8.04377317e-01
5.58393121e-01 5.27622849e-02 -5.35630405e-01 9.24124420e-01
3.42103571e-01 7.59220839e-01 7.61736393e-01 -1.05831921e+00
2.39203095e-01 7.71958888e-01 1.15196794e-01 -1.64863741e+00
-2.95553923e-01 -7.73379505e-01 -1.14663458e+00 6.12397075e-01
3.79116952e-01 4.69496965e-01 -2.46607780e-01 1.75509536e+00
2.24048391e-01 -1.70476153e-01 3.11650466e-02 1.23766387e+00
7.59871662e-01 3.55515689e-01 -1.78972438e-01 -3.97541136e-01
1.31760561e+00 -5.68110704e-01 -8.34818900e-01 5.42010605e-01
1.92257762e-01 -7.67892182e-01 1.45559704e+00 6.07755899e-01
-8.89001966e-01 -1.13811731e+00 -1.36750078e+00 1.65706240e-02
-2.36710757e-01 1.39636680e-01 2.86330521e-01 7.70389438e-01
-8.74774694e-01 1.00297296e+00 -4.12146032e-01 -1.34779885e-01
4.81022835e-01 8.34852606e-02 -6.45803332e-01 -1.46458358e-01
-1.27271843e+00 8.60161066e-01 2.05847636e-01 3.77133191e-01
-6.39894664e-01 -5.97687483e-01 -5.59123099e-01 -2.34344434e-02
-1.03628719e-02 -5.37746727e-01 4.70674723e-01 -1.08653533e+00
-1.41341805e+00 6.67994142e-01 3.22857976e-01 1.69359948e-02
6.15311265e-01 -2.69334942e-01 -9.17408466e-01 1.12915799e-01
9.31754783e-02 3.15671682e-01 1.13763130e+00 -1.13864863e+00
-4.15529698e-01 -7.21413732e-01 3.74300256e-02 2.69351721e-01
-7.21201479e-01 -1.17655389e-01 -2.56549746e-01 -7.53809214e-01
4.58525240e-01 -6.89006269e-01 3.90443772e-01 3.70698661e-01
-2.68862814e-01 -1.75976962e-01 5.14407575e-01 -6.68787241e-01
1.35418344e+00 -2.31703448e+00 2.01830834e-01 1.50153875e-01
3.18991125e-01 7.48706087e-02 -2.71665961e-01 2.34498709e-01
-1.94156259e-01 -8.87696818e-02 7.05355629e-02 2.92138785e-01
1.51356980e-01 -1.18733682e-01 -2.37557381e-01 9.21329796e-01
1.69280961e-01 6.98015094e-01 -7.35228062e-01 -6.82813108e-01
3.69571507e-01 4.98108715e-01 -3.44278097e-01 3.09130877e-01
6.41109705e-01 3.35477501e-01 -2.93955654e-01 6.25554144e-01
1.08156776e+00 -7.51520917e-02 -3.23122114e-01 -9.88327324e-01
-1.23250619e-01 -1.63752988e-01 -1.42845726e+00 1.70674634e+00
-2.04756647e-01 4.47905630e-01 -1.05590209e-01 -9.61288810e-01
1.03197670e+00 2.79705804e-02 5.17137647e-01 -1.16367996e+00
1.27974853e-01 1.73667878e-01 1.01254374e-01 -5.61495841e-01
2.21042857e-01 -4.46916111e-02 3.22732806e-01 1.15803584e-01
5.46866581e-02 1.27746850e-01 -1.86692566e-01 -5.66685572e-02
6.53090775e-01 2.23004103e-01 3.63776386e-01 -5.65696418e-01
8.11030626e-01 -4.80562359e-01 7.00715899e-01 4.07285631e-01
-7.57324696e-01 6.61908686e-01 1.52214989e-01 -3.98942918e-01
-9.48002458e-01 -1.36157084e+00 -6.89622462e-01 7.32612550e-01
7.15373874e-01 -1.96897998e-01 -5.30303657e-01 -5.21876872e-01
5.62559031e-02 2.62001306e-01 -6.26062751e-01 -5.97318172e-01
-1.79172590e-01 -6.45064414e-01 3.32698822e-01 3.37162971e-01
6.56410396e-01 -8.84617865e-01 -4.53983873e-01 -1.88975662e-01
-1.66000590e-01 -7.31078982e-01 -3.41891706e-01 -1.96571201e-01
-6.86811268e-01 -1.15739214e+00 -7.46249378e-01 -4.80684131e-01
5.10215700e-01 4.81411457e-01 1.00106430e+00 7.99058527e-02
-2.07551435e-01 3.26690018e-01 -4.30848330e-01 5.20138070e-02
-2.60639578e-01 -5.60199380e-01 4.95187342e-01 3.63519698e-01
3.86604756e-01 -4.52032894e-01 -1.06615889e+00 5.99692762e-01
-8.47973287e-01 1.22170504e-02 7.27669597e-01 8.49137723e-01
6.52362227e-01 3.35006475e-01 5.23863196e-01 -7.35162199e-02
6.38993621e-01 -3.17074120e-01 -3.77966136e-01 3.68535191e-01
-1.02877748e+00 5.03389426e-02 6.41443074e-01 -2.74126828e-01
-7.93967187e-01 -4.51943010e-01 -6.82726409e-03 -5.63980281e-01
-1.42376482e-01 2.31399596e-01 -6.93905294e-01 -2.95594037e-01
6.42446756e-01 4.39332992e-01 1.36451632e-01 -3.01394105e-01
4.54215795e-01 7.40897775e-01 5.85616291e-01 -4.42908019e-01
9.23744738e-01 5.25128186e-01 1.19070061e-01 -6.66584253e-01
-5.82549870e-01 -2.69242913e-01 -6.32762134e-01 -4.76458490e-01
6.67391300e-01 -9.33390200e-01 -7.75068760e-01 4.84643042e-01
-9.12824869e-01 4.01383221e-01 -4.69016165e-01 7.18469381e-01
-5.13283670e-01 7.49762952e-01 -4.36805457e-01 -6.29802942e-01
-4.23843801e-01 -1.27295053e+00 8.97519052e-01 1.68391794e-01
1.21313483e-01 -6.33500814e-01 -1.48700969e-02 3.50890458e-01
4.32182938e-01 1.42887548e-01 9.19211268e-01 -2.42264256e-01
-4.13511127e-01 -4.85246107e-02 -4.86005187e-01 8.48617435e-01
2.43970707e-01 -2.99338773e-02 -1.02325130e+00 -5.20419836e-01
3.35715920e-01 -2.63322145e-01 6.38164818e-01 3.79298717e-01
1.23182857e+00 -1.43956438e-01 1.88567296e-01 8.81516993e-01
1.49208736e+00 7.31092542e-02 8.72482479e-01 4.73895550e-01
5.82499444e-01 7.13655651e-01 8.30528736e-01 4.21155542e-01
8.38760734e-02 7.81000495e-01 5.47322035e-01 -1.32734373e-01
-2.23061949e-01 -9.47471485e-02 3.83429736e-01 1.09122491e+00
-6.35002851e-02 1.34388477e-01 -5.10891497e-01 1.60665259e-01
-1.50117958e+00 -1.04679489e+00 -2.81373481e-03 2.25362897e+00
6.04352176e-01 6.03417233e-02 -1.91230439e-02 6.00859523e-01
6.66666925e-01 5.08497432e-02 -6.92907870e-01 -8.75666067e-02
-3.73743445e-01 -1.98725417e-01 2.25349888e-01 1.02746136e-01
-1.00697160e+00 2.11794943e-01 5.78210545e+00 1.16113627e+00
-1.08613062e+00 5.75272553e-02 6.57462478e-01 8.23174343e-02
-3.27921629e-01 -3.09008628e-01 -3.77614588e-01 6.19839966e-01
6.66262925e-01 -1.82892546e-01 4.13992286e-01 8.46854270e-01
3.95859331e-01 2.11064622e-01 -1.10901725e+00 1.61200547e+00
1.73346519e-01 -8.70619357e-01 3.94840270e-01 1.72008350e-01
4.45791364e-01 -3.28369617e-01 4.82243478e-01 3.32068503e-01
-4.83144790e-01 -1.12340569e+00 9.45457578e-01 7.21991897e-01
9.91890371e-01 -7.23921597e-01 7.51048028e-01 1.06103905e-01
-1.06039715e+00 -3.32098573e-01 -8.26140285e-01 -7.45289177e-02
-2.32697919e-01 8.55115533e-01 -1.43046053e-02 9.25892055e-01
8.95215034e-01 8.23873222e-01 -7.46215105e-01 9.07864273e-01
1.83322966e-01 2.00114280e-01 2.50241756e-01 3.87839496e-01
-2.98642784e-01 -7.07419097e-01 5.17246127e-01 7.66336143e-01
5.78938425e-01 -3.38722318e-02 -6.69065341e-02 1.09733093e+00
1.57719791e-01 4.18179333e-01 -5.05974352e-01 1.39574811e-01
3.48739088e-01 1.27875483e+00 -4.39800411e-01 -9.09973532e-02
-5.71207225e-01 1.02770579e+00 8.93416703e-02 2.35321999e-01
-8.56744349e-01 -5.19007802e-01 8.19511712e-01 6.29971176e-02
3.98639478e-02 -3.36312526e-03 -3.63133937e-01 -1.22979343e+00
3.74359608e-01 -1.04008937e+00 1.41361132e-01 -6.65236890e-01
-1.62043107e+00 7.34655738e-01 -2.50559211e-01 -1.93485081e+00
4.49654490e-01 -7.24070489e-01 -3.83666277e-01 9.64007854e-01
-1.71442497e+00 -9.32187498e-01 -6.44464135e-01 9.38856304e-01
1.66139066e-01 -3.36124659e-01 5.89028060e-01 5.22519171e-01
-6.36689663e-01 8.08210194e-01 1.68695435e-01 1.34365737e-01
8.24435592e-01 -1.11522245e+00 -2.50712901e-01 7.61232555e-01
9.57917422e-03 6.55795515e-01 5.79751551e-01 -2.57906377e-01
-1.56423998e+00 -9.26608622e-01 3.04271191e-01 -5.23704529e-01
5.38783252e-01 -1.18748622e-03 -1.07292974e+00 -1.59762591e-01
6.84680343e-02 2.90758848e-01 7.96090662e-01 -1.48848727e-01
-6.96976721e-01 -6.59998715e-01 -1.11562884e+00 3.19909543e-01
1.11889076e+00 -7.25380063e-01 -5.90535581e-01 6.88826814e-02
6.18392646e-01 1.31614536e-01 -1.34458041e+00 6.60758853e-01
8.70154619e-01 -1.52042329e+00 1.15231633e+00 -3.41788977e-01
4.96757776e-01 -5.61181366e-01 -5.79729795e-01 -1.10864472e+00
-6.64546013e-01 -2.68921014e-02 -6.43743053e-02 1.23515320e+00
-1.06641747e-01 -5.32970071e-01 3.47927988e-01 2.93624043e-01
-7.52850398e-02 -8.15209389e-01 -1.11357617e+00 -1.12293875e+00
1.95349917e-01 -2.62652814e-01 7.44575799e-01 9.33394313e-01
-6.39386475e-02 3.38131636e-02 -2.81769037e-01 2.80781090e-01
7.51138866e-01 3.04904908e-01 5.58141470e-01 -1.37135172e+00
-2.45793223e-01 -7.46270895e-01 -8.32702756e-01 -7.11612821e-01
-2.34523695e-02 -8.99949610e-01 -1.81098565e-01 -1.25969875e+00
4.33479100e-01 -4.05266702e-01 -8.71992946e-01 -6.92591146e-02
-1.87152177e-01 2.46760279e-01 1.26580656e-01 5.72811842e-01
-6.44126654e-01 1.10600972e+00 1.58091605e+00 -3.27493936e-01
1.14471307e-02 -2.78917789e-01 -7.47153580e-01 5.37259758e-01
5.22880495e-01 -1.58017337e-01 -5.38914263e-01 -1.91602975e-01
3.37914824e-01 -1.30365074e-01 4.22144860e-01 -1.29985702e+00
-1.55634239e-01 -2.05845013e-01 6.70678735e-01 -3.26101363e-01
1.44134983e-01 -1.03070855e+00 1.71792343e-01 4.88051236e-01
-2.58802414e-01 1.09711932e-02 -2.10114747e-01 6.60141408e-01
-5.13791084e-01 2.13019460e-01 1.02658916e+00 1.46163285e-01
-6.77702665e-01 4.27959442e-01 4.05175388e-01 7.37430854e-03
9.18307960e-01 -5.66153646e-01 -2.30645567e-01 -1.31932750e-01
-4.14456666e-01 -2.55295217e-01 5.02180755e-01 5.56823373e-01
8.97941947e-01 -1.94098365e+00 -8.43107998e-01 3.75711799e-01
4.66202199e-01 -5.78527451e-01 5.37145138e-01 1.05996609e+00
-2.36486852e-01 1.67178720e-01 -7.00686634e-01 -7.95808077e-01
-9.10949469e-01 8.66929531e-01 4.05439466e-01 6.93154261e-02
-5.71638525e-01 5.03358364e-01 3.98903936e-01 -1.47053972e-01
1.05493307e-01 -2.06297830e-01 -4.08582300e-01 7.35819787e-02
7.00576007e-01 7.80269206e-01 2.37022340e-01 -1.02261114e+00
-3.49162161e-01 8.46249163e-01 3.11567336e-01 1.87644094e-01
9.99142468e-01 -4.88640606e-01 -1.17922418e-01 5.03711462e-01
1.51614761e+00 -7.22271279e-02 -1.19322848e+00 -1.66573882e-01
-3.32592815e-01 -9.01788473e-01 2.18531534e-01 -7.16101885e-01
-1.15333056e+00 1.04015088e+00 1.33952987e+00 2.37044349e-01
1.37329948e+00 -4.02315855e-01 5.64151525e-01 9.11379755e-02
5.73306978e-01 -1.11574173e+00 2.33776629e-01 -7.51833618e-02
1.34691858e+00 -1.47701871e+00 6.11124747e-02 -1.56677932e-01
-5.42174459e-01 1.10232246e+00 5.97915709e-01 -1.34536564e-01
7.80660808e-01 -2.04582512e-01 1.44622192e-01 -4.26374003e-02
-1.34265944e-01 5.52399121e-02 6.48572743e-01 6.79884195e-01
2.49300987e-01 2.54408926e-01 -3.95199180e-01 8.60519350e-01
-2.91555166e-01 -1.19867153e-01 2.18251020e-01 3.36899430e-01
-3.90000165e-01 -6.89107180e-01 -5.71024358e-01 3.70498925e-01
-9.60983112e-02 1.67166779e-03 9.85695198e-02 6.16659582e-01
4.55864102e-01 1.11110914e+00 -3.13206576e-02 -7.74027765e-01
4.74827081e-01 -2.42574349e-01 3.91176343e-01 -5.89600690e-02
-1.55987844e-01 -8.68903175e-02 -5.58573663e-01 -1.06459296e+00
-4.39490139e-01 -3.54631096e-01 -9.06799793e-01 -3.41421723e-01
-3.49107862e-01 -7.36360103e-02 4.77535605e-01 6.85009837e-01
3.20602179e-01 3.15023988e-01 1.01213825e+00 -6.06262088e-01
-8.85163724e-01 -7.57638335e-01 -9.79533553e-01 9.43218112e-01
4.57849830e-01 -1.06106186e+00 -5.95888913e-01 -2.19008997e-01] | [11.913718223571777, -1.8095849752426147] |
00656b1b-24af-4c58-870c-179d47e77439 | ttt-ucdr-test-time-training-for-universal | 2208.09198 | null | https://arxiv.org/abs/2208.09198v3 | https://arxiv.org/pdf/2208.09198v3.pdf | Test-time Training for Data-efficient UCDR | Image retrieval under generalized test scenarios has gained significant momentum in literature, and the recently proposed protocol of Universal Cross-domain Retrieval is a pioneer in this direction. A common practice in any such generalized classification or retrieval algorithm is to exploit samples from many domains during training to learn a domain-invariant representation of data. Such criterion is often restrictive, and thus in this work, for the first time, we explore the generalized retrieval problem in a data-efficient manner. Specifically, we aim to generalize any pre-trained cross-domain retrieval network towards any unknown query domain/category, by means of adapting the model on the test data leveraging self-supervised learning techniques. Toward that goal, we explored different self-supervised loss functions~(for example, RotNet, JigSaw, Barlow Twins, etc.) and analyze their effectiveness for the same. Extensive experiments demonstrate the proposed approach is simple, easy to implement, and effective in handling data-efficient UCDR. | ['Soma Biswas', 'Aheli Saha', 'Titir Dutta', 'Abhishek Samanta', 'Soumava Paul'] | 2022-08-19 | null | null | null | null | ['sketch-based-image-retrieval'] | ['computer-vision'] | [ 1.54296547e-01 -5.55707753e-01 -3.38491648e-01 -3.84681523e-01
-1.25719225e+00 -8.67037773e-01 7.55298257e-01 2.32433796e-01
-4.67496455e-01 7.64135838e-01 -1.09491348e-01 1.16003267e-01
-7.35574782e-01 -6.16880178e-01 -4.31786239e-01 -7.12732792e-01
1.56127810e-02 7.61809349e-01 1.93970844e-01 -2.31863141e-01
4.29007411e-01 7.70010412e-01 -1.53713381e+00 -1.69968158e-01
9.29773331e-01 1.04747021e+00 2.76500672e-01 2.36303166e-01
3.49448295e-03 2.77709186e-01 -4.95964170e-01 -5.17385721e-01
4.22985494e-01 -3.46943647e-01 -8.76524031e-01 1.24253482e-01
4.71269161e-01 -1.23841194e-02 -4.26772535e-01 9.80809450e-01
8.04674029e-01 3.22008669e-01 1.17547727e+00 -9.84766781e-01
-8.86781871e-01 1.19000390e-01 -6.32927001e-01 1.51545689e-01
3.75699013e-01 -1.50030658e-01 1.04592371e+00 -8.46626639e-01
7.81793416e-01 1.03315854e+00 2.91135162e-01 6.79034233e-01
-1.24219418e+00 -8.52957785e-01 -4.39081527e-02 1.62953854e-01
-1.81450081e+00 -2.71117926e-01 1.02010238e+00 -1.15431249e-01
4.50449377e-01 9.64770615e-02 1.47080749e-01 1.12704253e+00
4.05351073e-02 1.10516262e+00 1.17789292e+00 -7.00015604e-01
1.99871853e-01 4.38756883e-01 3.38921338e-01 3.90946507e-01
2.23001853e-01 -9.76195559e-02 -3.62530619e-01 -3.62102687e-01
4.76767063e-01 5.75693883e-02 -3.46578449e-01 -9.74633098e-01
-9.03623164e-01 8.18655550e-01 3.86596978e-01 5.70443869e-01
3.92305367e-02 -1.65854976e-01 6.73631489e-01 8.49382877e-01
8.03759277e-01 5.56187391e-01 -3.99307549e-01 3.14805329e-01
-1.01541972e+00 4.42650259e-01 7.44932234e-01 9.79570806e-01
7.44704425e-01 -4.32206839e-01 -2.74861278e-03 1.33543980e+00
1.03582211e-01 5.04880309e-01 9.39450026e-01 -4.43655580e-01
4.23619419e-01 3.37557614e-01 -5.19555770e-02 -8.35158825e-01
2.23294031e-02 -4.58987892e-01 -9.34503675e-01 -1.33227408e-01
1.32887870e-01 1.57301098e-01 -9.04331565e-01 1.73659933e+00
1.85006887e-01 -3.60662937e-02 3.94286484e-01 8.24000895e-01
5.10272443e-01 3.89417231e-01 1.32858008e-01 -2.04145119e-01
9.13435042e-01 -8.93187761e-01 -4.28654790e-01 -7.61498883e-02
6.01224184e-01 -7.94490755e-01 9.71275210e-01 4.89525408e-01
-8.90769243e-01 -6.00795090e-01 -1.12382412e+00 8.39610621e-02
-6.63027585e-01 4.29718383e-02 5.35724819e-01 5.25218606e-01
-9.59926724e-01 6.72652125e-01 -2.44719103e-01 -7.04715610e-01
2.31763184e-01 2.90613741e-01 -6.15850329e-01 -7.26316988e-01
-1.29885793e+00 9.46035564e-01 7.41070807e-01 -3.07416409e-01
-8.07793319e-01 -3.54912221e-01 -6.80705190e-01 2.05384139e-02
3.08740824e-01 -6.19650841e-01 1.01088631e+00 -1.06889415e+00
-1.29674876e+00 1.11201620e+00 1.50492296e-01 -4.06741291e-01
4.54662740e-01 -3.86255771e-01 -6.72174931e-01 3.47126454e-01
2.09065855e-01 5.04775286e-01 1.08625054e+00 -1.39831388e+00
-3.21096689e-01 -6.83648825e-01 -2.04695255e-01 2.52423346e-01
-6.18217587e-01 -5.21922193e-04 -7.35752702e-01 -6.88110590e-01
6.67942092e-02 -8.51097405e-01 8.19657743e-03 6.57756254e-02
-1.19770378e-01 -6.78479671e-01 7.24614620e-01 -3.20019335e-01
1.23623884e+00 -2.32576704e+00 2.30243817e-01 4.64251369e-01
-1.45434171e-01 4.82956320e-01 -3.74590099e-01 6.30495191e-01
-1.83568880e-01 -1.40469313e-01 -4.55212057e-01 -3.96760553e-01
-9.45852026e-02 2.39679113e-01 -4.83545363e-01 6.23785734e-01
3.09482545e-01 6.01444125e-01 -8.95210862e-01 -7.68860042e-01
5.29482216e-02 3.23415637e-01 -5.40749669e-01 2.28460521e-01
-2.15230674e-01 2.22469077e-01 -7.43507326e-01 6.22546732e-01
8.00938189e-01 -1.91897675e-01 2.27527604e-01 5.59989437e-02
2.00184390e-01 -1.88753471e-01 -1.22102094e+00 2.04097462e+00
-5.25704682e-01 4.33423728e-01 -2.02463642e-01 -1.45673907e+00
1.06820154e+00 2.36630499e-01 4.14554566e-01 -8.02821100e-01
-1.48919681e-02 5.67138731e-01 -4.89203364e-01 -4.14822459e-01
4.92060393e-01 -1.45297423e-01 -2.54991293e-01 3.40809405e-01
3.71614963e-01 -7.21877739e-02 1.80391580e-01 8.15536769e-04
8.69418859e-01 1.09351911e-01 5.13216913e-01 -1.50148034e-01
7.48917937e-01 1.34179503e-01 1.28386855e-01 7.88788855e-01
-2.65059881e-02 7.89731264e-01 -4.63916473e-02 -6.56210780e-02
-8.11267197e-01 -1.04264116e+00 -4.22672033e-01 1.08710063e+00
3.72627288e-01 -5.78441918e-02 -4.00146633e-01 -8.31381083e-01
1.42876729e-01 3.11829031e-01 -3.94119561e-01 -4.29185808e-01
-4.22956288e-01 -5.19331932e-01 6.61945462e-01 1.47466585e-01
7.72029340e-01 -9.91513431e-01 -2.03565329e-01 1.53198153e-01
1.01594932e-01 -8.09322119e-01 -4.05670464e-01 3.05903852e-01
-9.78649139e-01 -1.03056526e+00 -1.34616458e+00 -1.06223607e+00
5.21216393e-01 5.40649235e-01 1.10559022e+00 -1.79247394e-01
-3.20874661e-01 7.46745169e-01 -5.86297989e-01 -1.55662792e-02
-7.75576979e-02 2.61561036e-01 1.90314680e-01 8.72282311e-02
5.41806161e-01 -3.84434313e-01 -6.60375297e-01 2.46935189e-01
-1.38730681e+00 -7.79672861e-01 7.33395815e-01 1.09744895e+00
6.72854662e-01 1.45903423e-01 7.53695965e-01 -8.80243778e-01
9.89568114e-01 -6.32370353e-01 -5.90054750e-01 6.39221430e-01
-8.10741127e-01 2.19958976e-01 6.15111709e-01 -4.67852563e-01
-1.06910276e+00 -8.87441859e-02 1.08597986e-01 -6.30411208e-01
-1.89078435e-01 5.34100473e-01 -2.17947707e-01 -3.23546350e-01
7.98855782e-01 5.84203482e-01 -2.26611793e-02 -7.13637769e-01
4.17605013e-01 7.76490569e-01 2.91046381e-01 -8.39307427e-01
9.79085088e-01 3.66513819e-01 -8.08582678e-02 -6.71989501e-01
-7.43298709e-01 -1.11405098e+00 -5.91874957e-01 1.97409213e-01
5.40782750e-01 -9.23495471e-01 -1.81365460e-01 2.64097273e-01
-8.97708774e-01 1.61642551e-01 -1.67048424e-01 4.45857853e-01
-5.57545364e-01 6.40457630e-01 -1.13654919e-01 -6.50862813e-01
-4.83503073e-01 -8.95867288e-01 1.17539322e+00 1.45044103e-01
1.80698782e-01 -1.10008705e+00 4.69493032e-01 8.51960555e-02
3.32106411e-01 -1.86774343e-01 7.90224493e-01 -1.22775102e+00
-4.79239285e-01 -5.07187724e-01 -5.37475884e-01 7.61241615e-01
1.14804089e-01 -4.42651361e-01 -8.69629502e-01 -5.95459819e-01
-6.61195144e-02 -7.45321989e-01 9.42188561e-01 -1.01274453e-01
1.35166395e+00 1.21258967e-01 -5.72077096e-01 3.92301559e-01
1.84767163e+00 1.65323783e-02 4.88128752e-01 3.04788262e-01
8.33002850e-02 4.71109211e-01 9.31589663e-01 2.72536755e-01
-5.18214218e-02 7.55904794e-01 1.20534211e-01 2.15153098e-02
-1.89152025e-02 -5.95177002e-02 7.86545873e-02 6.18752480e-01
2.24222094e-01 -4.09276783e-01 -7.19851255e-01 8.48000050e-01
-1.60699451e+00 -8.68723214e-01 5.72772622e-01 2.34640932e+00
7.99603283e-01 -1.40844271e-01 -7.69138783e-02 2.19975505e-02
5.77764034e-01 1.75772458e-01 -4.83646572e-01 -5.13801686e-02
-2.13123605e-01 5.96373141e-01 3.60910475e-01 1.64770931e-02
-1.16361356e+00 9.77024913e-01 5.79342842e+00 1.24613523e+00
-1.08926952e+00 -3.27664204e-02 2.92073816e-01 1.62436649e-01
-1.35404497e-01 -1.78249627e-01 -6.73547864e-01 2.10507989e-01
5.72563648e-01 -3.71101409e-01 1.98407948e-01 1.01943755e+00
-4.10402864e-01 4.27315608e-02 -1.14099038e+00 1.06523812e+00
4.78453904e-01 -8.35022449e-01 3.10613066e-01 -9.53350142e-02
6.53346419e-01 8.59964862e-02 2.93560892e-01 5.78608632e-01
2.15944514e-01 -5.32628417e-01 4.96365167e-02 4.76403862e-01
8.66870344e-01 -6.40963495e-01 5.27411461e-01 2.84848183e-01
-8.84566665e-01 1.10523505e-02 -5.80909789e-01 5.93978345e-01
-3.31535876e-01 3.11882704e-01 -6.78345203e-01 9.21336353e-01
5.91632605e-01 7.61994779e-01 -6.03613615e-01 1.38530099e+00
2.65101433e-01 2.28431359e-01 -2.24877611e-01 8.41576830e-02
4.31054354e-01 -1.62436083e-01 5.00349104e-01 1.30495071e+00
3.84440362e-01 -5.69403023e-02 2.19444394e-01 3.35198879e-01
-2.87812978e-01 3.36899340e-01 -9.40724313e-01 1.36472613e-01
2.01386914e-01 1.00988579e+00 -3.68066221e-01 -2.60384679e-01
-3.06741357e-01 1.28462780e+00 4.33792144e-01 5.52673638e-01
-5.30467629e-01 -5.06244898e-01 6.07475996e-01 -2.03320280e-01
2.96882868e-01 -6.47176942e-03 2.44024590e-01 -1.33913422e+00
1.17777422e-01 -8.58844578e-01 7.34260976e-01 -5.29549360e-01
-1.88005316e+00 5.48315048e-01 2.37652272e-01 -1.74518716e+00
-3.03103685e-01 -5.74766040e-01 -1.06700793e-01 8.81539464e-01
-2.13884282e+00 -9.78404462e-01 -1.51272222e-01 1.07855475e+00
6.81314826e-01 -4.62915242e-01 9.12401557e-01 5.95878184e-01
-2.79714793e-01 8.51414919e-01 8.67643118e-01 -7.26526650e-03
1.40135264e+00 -1.05444086e+00 -3.03388566e-01 5.00099480e-01
2.98970252e-01 8.51208687e-01 5.68060696e-01 -3.19744736e-01
-1.37341762e+00 -9.93751109e-01 7.05048084e-01 -5.33255078e-02
7.81453013e-01 -3.65096591e-02 -1.09334421e+00 4.03240144e-01
1.91183269e-01 2.35059664e-01 6.55583024e-01 3.78977545e-02
-6.87649667e-01 -5.63035727e-01 -1.32272875e+00 5.18883467e-01
9.29772377e-01 -7.62962937e-01 -7.95149624e-01 5.91367781e-01
4.35729444e-01 -1.14186145e-01 -8.55110645e-01 4.10613716e-01
5.00184059e-01 -6.48820877e-01 1.08871233e+00 -7.04274774e-01
2.15298101e-01 -8.34636912e-02 -4.46533591e-01 -1.34576011e+00
-9.12471712e-02 -3.38023782e-01 1.43461600e-01 1.30163264e+00
1.28946707e-01 -7.29981720e-01 6.88551068e-01 2.06051737e-01
5.76666929e-02 -3.63545209e-01 -1.01349950e+00 -1.00199878e+00
3.32378417e-01 -1.17225803e-01 2.92006552e-01 9.86211479e-01
-1.66719005e-01 1.98768735e-01 -3.40886652e-01 1.25082552e-01
7.58645833e-01 4.22169536e-01 6.90269053e-01 -1.38447857e+00
-3.42410505e-01 -2.96278179e-01 -4.02929038e-01 -1.38176465e+00
4.45170909e-01 -1.02665746e+00 -2.24533621e-02 -1.14760053e+00
2.90235102e-01 -7.72401750e-01 -7.45845556e-01 4.86326739e-02
-2.57838145e-02 3.27506900e-01 1.97746038e-01 4.89032060e-01
-9.87083018e-01 6.03321612e-01 1.21513379e+00 -3.80395323e-01
5.24791237e-03 2.01197535e-01 -6.86307311e-01 1.95633277e-01
7.29378641e-01 -4.80890095e-01 -6.43297672e-01 -3.84941190e-01
-1.92982972e-01 9.14825574e-02 3.06747317e-01 -9.95971978e-01
4.00105894e-01 2.36486405e-01 2.05041021e-01 -4.79227602e-01
3.73871714e-01 -1.18639362e+00 -1.72979325e-01 1.02211311e-01
-4.81938004e-01 -2.09825397e-01 8.90426412e-02 9.34040129e-01
-8.07235360e-01 -4.98930037e-01 7.45644033e-01 -1.93430558e-01
-9.25105214e-01 3.53391320e-01 1.62321895e-01 1.03887603e-01
1.01902246e+00 -1.12566791e-01 -1.43738911e-01 -2.15632856e-01
-6.54235303e-01 2.12827966e-01 3.57720673e-01 4.41269696e-01
6.99388862e-01 -1.30208504e+00 -4.75644141e-01 3.40995900e-02
8.38643491e-01 -2.14914814e-01 9.57359448e-02 2.70393312e-01
-2.63711572e-01 7.27180541e-01 -1.61952138e-01 -6.37120306e-01
-1.19304919e+00 7.98529625e-01 1.84561312e-01 -6.72658384e-01
-2.89224774e-01 5.75845838e-01 1.73534572e-01 -3.89673829e-01
2.05211058e-01 2.26799488e-01 -2.61963964e-01 1.55069262e-01
3.30272824e-01 1.62203349e-02 1.77629814e-01 -3.27210903e-01
-2.31499761e-01 8.41601670e-01 -5.42341530e-01 -9.34496000e-02
1.32015061e+00 -8.91944394e-02 -4.34922688e-02 2.26050690e-01
1.63648272e+00 -2.62331218e-01 -7.91074812e-01 -7.77076006e-01
2.85577923e-01 -4.95227784e-01 -7.64190108e-02 -6.99559391e-01
-9.06929314e-01 6.31592691e-01 8.29010785e-01 2.04047963e-01
1.51879013e+00 1.34804443e-01 6.25731289e-01 6.84896648e-01
6.70719922e-01 -1.03063691e+00 2.14940265e-01 2.52231449e-01
8.62833858e-01 -1.41735005e+00 7.67776743e-02 -1.30839229e-01
-3.92487139e-01 9.66412902e-01 5.93308449e-01 -4.81393248e-01
7.07929671e-01 -4.86418277e-01 -2.03313217e-01 -1.79617524e-01
-5.05839527e-01 -4.04896110e-01 4.25115556e-01 5.71788311e-01
3.63352478e-01 -1.77741334e-01 -5.63997149e-01 3.13310586e-02
4.53602523e-01 5.56200333e-02 -1.06240213e-02 1.00560141e+00
-1.73691899e-01 -1.49832046e+00 -3.10468823e-01 2.91425526e-01
-4.14064854e-01 -5.35394214e-02 -4.05799985e-01 1.11169684e+00
-1.89804420e-01 6.61141038e-01 -2.08547249e-01 3.12514720e-04
4.99678701e-01 1.27415285e-01 5.84135115e-01 -5.44385970e-01
-4.40157056e-01 -3.75839211e-02 -1.26005217e-01 -2.97508091e-01
-7.71380424e-01 -6.30903423e-01 -6.59099162e-01 2.11243987e-01
-4.95288670e-01 3.41697574e-01 6.30068958e-01 7.19970763e-01
3.02359521e-01 1.16497949e-02 9.36484933e-01 -5.54180384e-01
-8.94575536e-01 -8.13748598e-01 -7.17492044e-01 6.75756633e-01
3.71500939e-01 -7.93439984e-01 -4.63979810e-01 -2.29392305e-01] | [11.307955741882324, 1.0303281545639038] |
81f6a3d8-4a6e-420f-a595-52e2085f631a | duplex-conversation-towards-human-like | 2205.15060 | null | https://arxiv.org/abs/2205.15060v4 | https://arxiv.org/pdf/2205.15060v4.pdf | Duplex Conversation: Towards Human-like Interaction in Spoken Dialogue Systems | In this paper, we present Duplex Conversation, a multi-turn, multimodal spoken dialogue system that enables telephone-based agents to interact with customers like a human. We use the concept of full-duplex in telecommunication to demonstrate what a human-like interactive experience should be and how to achieve smooth turn-taking through three subtasks: user state detection, backchannel selection, and barge-in detection. Besides, we propose semi-supervised learning with multimodal data augmentation to leverage unlabeled data to increase model generalization. Experimental results on three sub-tasks show that the proposed method achieves consistent improvements compared with baselines. We deploy the Duplex Conversation to Alibaba intelligent customer service and share lessons learned in production. Online A/B experiments show that the proposed system can significantly reduce response latency by 50%. | ['Yongbin Li', 'Jian Sun', 'Luo Si', 'Fei Huang', 'Yuchuan Wu', 'Ting-En Lin'] | 2022-05-30 | null | null | null | null | ['spoken-dialogue-systems'] | ['speech'] | [ 7.92073831e-02 5.48853219e-01 -1.90236405e-01 -8.96335721e-01
-1.20438647e+00 -8.21235418e-01 7.95051634e-01 -4.61318046e-01
-2.75630981e-01 6.95925176e-01 3.79375994e-01 -8.19464922e-01
4.77708548e-01 -4.84374240e-02 -2.32067034e-01 -3.36108804e-01
-9.73463207e-02 9.06512499e-01 -1.60441816e-01 -8.67954731e-01
6.62768111e-02 1.88117892e-01 -1.06291246e+00 9.24769580e-01
7.62840152e-01 8.97574127e-01 2.28397548e-02 1.25945842e+00
-3.29009175e-01 1.15752935e+00 -7.30753362e-01 -2.25164443e-01
-5.50505519e-02 -5.06929100e-01 -1.32618344e+00 4.84757125e-01
-7.12042488e-03 -7.02223063e-01 -3.82066667e-01 2.07453012e-01
6.41820252e-01 2.82564878e-01 4.88170028e-01 -1.45641196e+00
9.23738331e-02 7.35387325e-01 -1.78171888e-01 -9.39768255e-02
9.90900815e-01 4.37232256e-01 1.07404387e+00 -8.36884022e-01
2.27660939e-01 1.79721367e+00 3.30350995e-01 6.31961584e-01
-1.27454293e+00 -3.78606141e-01 4.18904394e-01 -8.86356980e-02
-7.46258318e-01 -1.08424222e+00 2.29084313e-01 1.74621418e-02
1.26049876e+00 4.98955846e-01 1.79413050e-01 1.28820086e+00
-3.70718092e-01 1.58789217e+00 9.63040948e-01 -6.36208057e-01
3.60227213e-03 4.63931441e-01 3.52882326e-01 7.74900258e-01
-1.10349679e+00 -3.27505320e-01 -5.84918916e-01 -2.11482599e-01
3.21436644e-01 -5.79777837e-01 -1.86899856e-01 3.94706540e-02
-1.05375659e+00 8.19391370e-01 -9.49933678e-02 -1.18766271e-01
-3.89283746e-01 -3.73842776e-01 5.80096960e-01 8.74664843e-01
2.30260655e-01 4.02042985e-01 -7.31837153e-01 -9.94695008e-01
-4.97403322e-04 -3.96366185e-03 1.69416106e+00 1.38676572e+00
4.64942336e-01 -1.15734555e-01 -1.56447336e-01 1.31482565e+00
1.81716293e-01 5.43516278e-01 1.49640396e-01 -1.15675473e+00
8.20612788e-01 3.97739053e-01 5.15721321e-01 -1.63011789e-01
-7.88274586e-01 2.76793391e-01 -4.86820698e-01 -4.17957902e-01
5.47267973e-01 -8.64742339e-01 -7.44829416e-01 1.40680659e+00
2.58930057e-01 -2.11077988e-01 3.23263466e-01 7.84339786e-01
7.29273200e-01 8.38112414e-01 -7.47413263e-02 -2.26054937e-01
1.25704551e+00 -1.49882507e+00 -8.89088273e-01 -3.55975896e-01
1.24098706e+00 -8.59531224e-01 1.42162395e+00 6.01482093e-01
-1.02816224e+00 -1.92753658e-01 -5.54697156e-01 9.51619074e-02
-3.03402711e-02 -4.52329777e-02 9.99215603e-01 1.02774179e+00
-7.91189730e-01 -1.68998063e-01 -5.41272581e-01 -5.77767432e-01
-3.17340404e-01 4.55949008e-01 -2.16613397e-01 2.55455017e-01
-1.28949821e+00 8.68196666e-01 -6.72960132e-02 1.63943797e-01
-6.96993768e-01 -1.81701437e-01 -8.36775780e-01 2.47272011e-02
4.90446121e-01 -2.36764222e-01 2.42960453e+00 -9.06608343e-01
-2.54076314e+00 5.84126115e-01 -4.29538339e-01 -2.04553530e-01
4.36425716e-01 -3.70707124e-01 -4.29218590e-01 1.26648480e-02
-3.54664564e-01 5.51755488e-01 2.60758847e-01 -1.36141574e+00
-1.36711705e+00 7.82411080e-03 2.48383999e-01 6.28745914e-01
1.41870873e-02 -2.99591143e-02 -8.54200304e-01 2.29490101e-01
1.40208192e-02 -1.17977452e+00 9.15296376e-02 -9.91874099e-01
-6.91834569e-01 -5.30180097e-01 6.54719532e-01 -5.83131850e-01
1.15004134e+00 -1.74850833e+00 -7.90838674e-02 5.10618985e-01
-1.65097758e-01 6.00383133e-02 -4.87746924e-01 9.86989856e-01
2.65844643e-01 -2.24186435e-01 3.89078170e-01 -6.34802520e-01
3.87010843e-01 1.35533452e-01 -2.82646596e-01 3.27829574e-03
-2.10616872e-01 8.99155200e-01 -7.55245388e-01 7.00684637e-02
2.53702730e-01 -4.82079312e-02 -4.83582348e-01 6.35720313e-01
-2.22688943e-01 6.81564748e-01 -2.46541142e-01 7.45529056e-01
4.34762269e-01 -1.19107850e-01 5.57199299e-01 1.04974896e-01
-4.28252183e-02 7.39590168e-01 -7.26260602e-01 1.60433638e+00
-8.72967482e-01 5.68913460e-01 7.13324428e-01 -5.60900331e-01
6.49331212e-01 3.68013680e-01 9.86179486e-02 -1.11377466e+00
2.63032496e-01 -8.94884318e-02 6.24992847e-02 -7.43847609e-01
6.91214740e-01 3.50411922e-01 -4.39946800e-01 9.19551373e-01
-1.62239209e-01 -5.69798462e-02 6.74581379e-02 5.96457481e-01
7.95728981e-01 -4.53275889e-01 -1.00357749e-01 3.41779262e-01
4.00402993e-01 -2.52793282e-01 1.72613010e-01 1.13895237e+00
-3.21233928e-01 8.10417607e-02 7.93778121e-01 -1.32318452e-01
-6.26227796e-01 -6.60860658e-01 4.37337309e-01 2.11898851e+00
-1.38421291e-02 -7.99680874e-02 -8.37314069e-01 -7.30072796e-01
-6.83901533e-02 9.73887980e-01 -7.42749497e-02 8.66412744e-02
-4.02075261e-01 -2.92071372e-01 7.16700375e-01 3.56686890e-01
4.83367920e-01 -9.01318669e-01 7.53669515e-02 3.59601527e-01
-6.55055463e-01 -1.29384041e+00 -8.77439559e-01 2.15845406e-01
-3.70769531e-01 -7.63727546e-01 -3.65231097e-01 -9.39483821e-01
1.31454736e-01 4.98584062e-01 9.61141586e-01 -2.02461526e-01
2.31942862e-01 7.46503294e-01 -5.33780754e-01 -9.83100459e-02
-7.94912398e-01 5.36004305e-01 9.57005098e-02 1.73570946e-01
4.52103049e-01 -5.19374534e-02 -4.85044092e-01 8.58595669e-01
-1.83107942e-01 1.96944416e-01 6.65395260e-01 9.82684195e-01
-4.76287574e-01 -7.25495577e-01 8.70409846e-01 -1.11710417e+00
1.21801043e+00 -2.86805332e-01 -9.59206223e-02 5.32566190e-01
-5.59184968e-01 -1.37533203e-01 1.78172827e-01 -2.97707409e-01
-1.47033012e+00 5.11524230e-02 -2.73163199e-01 4.16432679e-01
-4.12283689e-01 5.35273850e-01 -3.55985433e-01 1.95597589e-01
3.46313477e-01 2.57443458e-01 3.73154938e-01 -2.22426996e-01
7.56309927e-01 1.63367414e+00 2.85414100e-01 -5.38845599e-01
2.38334388e-02 -1.25772681e-03 -8.23884189e-01 -1.01327550e+00
-3.15101087e-01 -8.39225173e-01 -2.24950984e-01 -3.82743776e-01
3.07640165e-01 -1.08463335e+00 -1.87022257e+00 5.67121923e-01
-1.17690587e+00 -1.02445722e+00 5.30906379e-01 3.67727309e-01
-7.95892239e-01 2.74335384e-01 -1.12506473e+00 -1.42556477e+00
-2.61275887e-01 -1.07406282e+00 1.00983071e+00 2.42829710e-01
-5.50082326e-01 -7.85661221e-01 -5.97331859e-02 8.15925062e-01
3.84863824e-01 -8.25325549e-01 7.49166310e-01 -1.14628756e+00
-3.81346703e-01 -2.23150283e-01 -1.09321170e-01 1.43531010e-01
2.96686947e-01 -3.45994443e-01 -1.21510124e+00 -4.15496022e-01
-5.10878026e-01 -9.84438658e-01 4.21180159e-01 1.27184287e-01
6.12965107e-01 -5.67019999e-01 -2.78854012e-01 5.07646380e-03
1.91046894e-01 7.41306126e-01 3.68997544e-01 1.18305162e-01
4.41988230e-01 1.05073130e+00 8.50273252e-01 6.98969722e-01
9.97503340e-01 5.95788658e-01 8.66920352e-02 -2.25314468e-01
4.48884934e-01 -2.98674792e-01 4.94017422e-01 8.53723228e-01
3.49779457e-01 -7.18153715e-01 -9.35408771e-01 2.07058072e-01
-2.15256357e+00 -7.21602738e-01 -3.98902968e-02 1.89904225e+00
8.30371201e-01 2.09564164e-01 6.87031031e-01 -2.68172860e-01
6.56719863e-01 -2.58511603e-01 -4.12473053e-01 -8.73975575e-01
9.36671793e-02 -3.72316033e-01 2.52548933e-01 1.16930401e+00
-9.56705451e-01 1.29774666e+00 6.66307831e+00 4.75179225e-01
-9.53780353e-01 -4.27876264e-02 8.86629641e-01 -1.13884315e-01
-2.27272123e-01 -3.18523943e-01 -6.99037552e-01 1.13209859e-01
1.27743649e+00 1.73085883e-01 7.71308959e-01 6.36964321e-01
6.07574046e-01 -2.75528133e-01 -1.46300590e+00 1.00905490e+00
-8.29541534e-02 -8.42845500e-01 -2.62178570e-01 -5.58036864e-02
4.59065765e-01 -6.62517250e-02 6.66088099e-03 1.05434895e+00
6.21983528e-01 -8.16984236e-01 2.96748072e-01 3.94387931e-01
3.36668760e-01 -9.51732993e-01 6.19675636e-01 5.65273166e-01
-7.78091133e-01 -3.05815190e-01 4.05020416e-01 -1.14714041e-01
5.99446416e-01 -2.86643595e-01 -1.63292897e+00 -6.59227371e-03
1.56345248e-01 -1.55603796e-01 1.93034634e-01 5.75657845e-01
1.77178122e-02 7.18348503e-01 -3.10467154e-01 -3.77864718e-01
4.64718133e-01 -2.70422190e-01 1.26269355e-01 1.55618525e+00
-2.12423757e-01 5.12224555e-01 5.16703427e-01 3.11587332e-03
-2.64204144e-01 1.73449248e-01 -4.55556720e-01 -2.80004859e-01
7.15553522e-01 1.18761134e+00 -2.04986900e-01 -3.01106602e-01
-6.55246139e-01 1.36959147e+00 4.22903597e-02 8.29800069e-01
-5.71109056e-01 -5.99387169e-01 6.56790197e-01 -3.70425165e-01
-4.95969597e-03 -1.18926935e-01 8.00106302e-02 -8.51574004e-01
-3.01110059e-01 -1.38918436e+00 1.59713715e-01 -6.20686829e-01
-1.00960398e+00 6.15437686e-01 -4.32323605e-01 -7.89137006e-01
-8.84901285e-01 -4.28616852e-01 -4.72724915e-01 7.94234753e-01
-1.10755253e+00 -1.21387815e+00 -6.83570132e-02 3.93505573e-01
1.09638727e+00 -3.48371863e-01 1.05901730e+00 2.82907724e-01
-6.51745260e-01 1.04713953e+00 1.14495918e-01 2.64870584e-01
1.03790903e+00 -1.29186738e+00 6.60192251e-01 -6.79963082e-02
-1.90238416e-01 7.82517433e-01 6.51137471e-01 -3.07163090e-01
-1.69485009e+00 -4.14778352e-01 8.13423753e-01 -3.60940903e-01
6.47368252e-01 -6.87261224e-01 -6.33426070e-01 9.95045960e-01
7.30616868e-01 -9.54710960e-01 9.23326433e-01 9.08022106e-01
-2.30526045e-01 -9.65978354e-02 -8.53574216e-01 7.33157337e-01
6.32690132e-01 -9.40204382e-01 -2.46911407e-01 5.14025569e-01
5.78840256e-01 -6.28172815e-01 -5.63992143e-01 -8.51396993e-02
8.96147847e-01 -8.61044645e-01 4.64691907e-01 -1.06942070e+00
-2.46294618e-01 5.33438206e-01 -2.61507649e-02 -1.62926948e+00
1.61963478e-01 -1.54184747e+00 4.30628425e-03 1.10806048e+00
9.48982656e-01 -8.07677090e-01 8.18680644e-01 1.24348128e+00
-2.57621795e-01 -4.72046733e-01 -7.06969500e-01 -3.27543020e-01
-2.59782344e-01 -3.92928183e-01 4.76996303e-01 6.84620023e-01
1.00647604e+00 9.96738374e-01 -6.84584796e-01 2.44638756e-01
1.12080775e-01 -1.00692935e-01 1.42487323e+00 -7.97490895e-01
-2.32423976e-01 -3.54197204e-01 3.59085202e-01 -2.25174284e+00
3.74932259e-01 -3.29061925e-01 4.56637889e-01 -1.10459900e+00
-8.74506384e-02 -3.94558877e-01 1.10384464e-01 3.32894444e-01
1.19696997e-01 -5.62201262e-01 6.16455711e-02 1.67882480e-02
-1.07880652e+00 6.17929280e-01 1.12113595e+00 -2.61728287e-01
-7.30571628e-01 5.90398848e-01 -6.30339444e-01 4.28935558e-01
6.98037922e-01 3.42419565e-01 -5.41767418e-01 -2.26915747e-01
-5.01494557e-02 7.63006866e-01 -3.41305047e-01 -1.28128782e-01
5.70650280e-01 -2.46592090e-02 -3.67316544e-01 -4.90316421e-01
6.35273457e-01 -5.77315748e-01 -8.51019263e-01 -1.96998641e-02
-8.65562201e-01 -1.17282934e-01 3.27799708e-01 4.53512073e-01
-8.01081434e-02 2.70283133e-01 1.70483068e-01 6.89054132e-02
-6.43067837e-01 -1.11248203e-01 -1.16045713e+00 -9.61379334e-02
5.83552420e-01 1.05501488e-01 -3.79014313e-01 -1.61524713e+00
-8.68337810e-01 1.06508410e+00 -2.19908878e-01 6.62910163e-01
3.61127645e-01 -9.17698205e-01 -4.75455523e-01 1.62123531e-01
3.32617462e-01 -6.23956501e-01 2.37867996e-01 8.53749633e-01
-2.01350629e-01 8.95062566e-01 2.88815409e-01 -7.05589473e-01
-1.51324296e+00 -1.65397182e-01 4.41107005e-01 3.74628715e-02
-6.02596905e-03 1.03814232e+00 -6.42899005e-03 -1.11472237e+00
1.00927234e+00 -3.34644496e-01 -9.26004909e-03 1.19740866e-01
5.56272447e-01 3.14861655e-01 2.01659366e-01 -3.58259618e-01
-1.18429966e-01 -4.01856959e-01 -6.28790081e-01 -7.79771328e-01
8.17438245e-01 -9.19499874e-01 2.53399760e-01 6.59183145e-01
1.01477206e+00 -1.89560607e-01 -1.31275892e+00 -4.02215958e-01
1.60499960e-01 -1.73725173e-01 -1.91577449e-01 -1.30022037e+00
-3.70610148e-01 6.48187876e-01 5.78631639e-01 6.86476529e-01
6.60098612e-01 -2.50835065e-03 9.21468437e-01 1.35738802e+00
4.61088389e-01 -1.48295403e+00 1.58625975e-01 9.88054514e-01
8.43748808e-01 -1.73058689e+00 -7.57871330e-01 -3.22223932e-01
-1.47729468e+00 8.68983924e-01 7.79960692e-01 7.57222235e-01
2.37595588e-01 1.34792075e-01 8.79338443e-01 1.98229719e-02
-1.33447003e+00 -3.54770511e-01 -6.46213815e-02 4.48744506e-01
6.15976155e-01 1.70662388e-01 1.79915562e-01 4.67849225e-01
7.70247728e-02 -3.91691536e-01 4.79227036e-01 1.11668229e+00
-6.54795051e-01 -1.17964113e+00 -2.43323162e-01 1.93596855e-01
2.28636295e-01 -2.18186006e-02 -9.22316551e-01 7.77608216e-01
-8.76631558e-01 1.83221996e+00 2.13861465e-04 -6.62953258e-01
8.59027505e-01 6.37085438e-01 3.07816803e-01 -5.81960320e-01
-7.37846076e-01 4.14545536e-01 1.03865397e+00 -6.67838931e-01
-4.53880094e-02 -4.81360584e-01 -1.36567891e+00 -5.10181010e-01
-4.08661544e-01 4.87494707e-01 5.76831758e-01 1.04948676e+00
4.34295475e-01 2.20400006e-01 1.27855372e+00 -8.00862908e-01
-8.16032171e-01 -1.57010865e+00 -4.89033878e-01 2.07370967e-01
5.14314711e-01 -1.46587282e-01 -3.52035582e-01 -2.63101369e-01] | [12.867080688476562, 7.947339057922363] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.