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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
949fe9b8-bdb1-40a2-8e77-923cc42bf46d | reinforced-iterative-knowledge-distillation | 2106.00241 | null | https://arxiv.org/abs/2106.00241v1 | https://arxiv.org/pdf/2106.00241v1.pdf | Reinforced Iterative Knowledge Distillation for Cross-Lingual Named Entity Recognition | Named entity recognition (NER) is a fundamental component in many applications, such as Web Search and Voice Assistants. Although deep neural networks greatly improve the performance of NER, due to the requirement of large amounts of training data, deep neural networks can hardly scale out to many languages in an industry setting. To tackle this challenge, cross-lingual NER transfers knowledge from a rich-resource language to languages with low resources through pre-trained multilingual language models. Instead of using training data in target languages, cross-lingual NER has to rely on only training data in source languages, and optionally adds the translated training data derived from source languages. However, the existing cross-lingual NER methods do not make good use of rich unlabeled data in target languages, which is relatively easy to collect in industry applications. To address the opportunities and challenges, in this paper we describe our novel practice in Microsoft to leverage such large amounts of unlabeled data in target languages in real production settings. To effectively extract weak supervision signals from the unlabeled data, we develop a novel approach based on the ideas of semi-supervised learning and reinforcement learning. The empirical study on three benchmark data sets verifies that our approach establishes the new state-of-the-art performance with clear edges. Now, the NER techniques reported in this paper are on their way to become a fundamental component for Web ranking, Entity Pane, Answers Triggering, and Question Answering in the Microsoft Bing search engine. Moreover, our techniques will also serve as part of the Spoken Language Understanding module for a commercial voice assistant. We plan to open source the code of the prototype framework after deployment. | ['Daxin Jiang', 'Xianglin Zuo', 'Wanli Zuo', 'Linjun Shou', 'Jian Pei', 'Ming Gong', 'Shining Liang'] | 2021-06-01 | null | null | null | null | ['cross-lingual-ner'] | ['natural-language-processing'] | [-2.39024788e-01 -1.62696078e-01 -2.89931267e-01 -4.75428969e-01
-9.92026567e-01 -8.54715228e-01 4.17534977e-01 -2.45715097e-01
-8.68800819e-01 8.97497237e-01 1.05957530e-01 -5.53329110e-01
7.98318312e-02 -7.23615170e-01 -6.28107607e-01 -7.34055713e-02
3.31323266e-01 6.13542020e-01 7.74192438e-02 -5.82477689e-01
-2.80491680e-01 4.73809958e-01 -1.36925828e+00 1.75251171e-01
1.05015767e+00 7.75331736e-01 2.80182749e-01 3.43137562e-01
-8.09151411e-01 7.07302630e-01 -5.50449669e-01 -3.86028022e-01
3.10329229e-01 -2.20024198e-01 -9.44102526e-01 -2.67069131e-01
2.85809219e-01 -2.83497006e-01 -1.12931579e-01 1.07610142e+00
8.33768070e-01 1.14214048e-01 2.89757922e-02 -9.86952722e-01
-6.22472465e-01 9.75501299e-01 -2.36798033e-01 1.02714682e-02
3.83967489e-01 -1.55914932e-01 1.00418448e+00 -8.65311205e-01
7.26811469e-01 8.27156901e-01 7.03061223e-01 7.22476661e-01
-7.53374696e-01 -7.65207589e-01 -1.35852275e-02 -1.49045046e-03
-1.28047872e+00 -6.38921440e-01 6.56248331e-01 -4.92870174e-02
1.06424904e+00 1.73321471e-01 3.87578458e-02 1.13487828e+00
-4.81468558e-01 8.77013445e-01 1.13832486e+00 -6.23057127e-01
7.46191815e-02 6.34947658e-01 1.35254577e-01 6.77126646e-01
-9.13139582e-02 9.56399068e-02 -5.21896541e-01 -4.66680117e-02
5.72379887e-01 -6.84392825e-02 -4.37445849e-01 -1.75398618e-01
-1.15459740e+00 7.26205945e-01 3.86023521e-01 6.82193041e-01
-4.28805292e-01 -3.02033901e-01 3.79977226e-01 4.23800319e-01
1.75125092e-01 4.60356861e-01 -9.48516726e-01 -3.32092851e-01
-1.03143775e+00 -4.21890855e-01 1.42373824e+00 1.16314566e+00
8.73134792e-01 1.29721761e-01 2.41580307e-01 1.19692433e+00
7.69467046e-03 6.13719046e-01 8.87722135e-01 -7.95081377e-01
6.12238646e-01 6.34032905e-01 7.71004707e-02 -3.14967364e-01
-2.02858761e-01 -2.90794224e-01 -6.79206669e-01 -1.31942302e-01
3.87880415e-01 -6.37129128e-01 -9.15387511e-01 1.76835227e+00
3.80160481e-01 -1.62008822e-01 3.03771555e-01 8.19778264e-01
8.89291883e-01 5.59764862e-01 2.07767985e-03 -8.66358355e-02
1.32158625e+00 -1.13717782e+00 -7.57690728e-01 -2.12433532e-01
6.89079583e-01 -8.01242948e-01 1.22948086e+00 2.61702120e-01
-7.38133490e-01 -5.97904742e-01 -7.77332664e-01 -4.47351858e-02
-9.56700981e-01 3.02225828e-01 6.93560898e-01 7.46421456e-01
-9.49742377e-01 4.53846812e-01 -6.45054221e-01 -7.45238721e-01
-8.90195891e-02 5.32532156e-01 -6.07040524e-01 -1.13114603e-01
-1.48149800e+00 8.25544477e-01 5.21926701e-01 2.79940754e-01
-5.63004494e-01 -3.74349862e-01 -8.60222042e-01 1.10434955e-02
6.90720439e-01 -3.03953737e-01 1.39254034e+00 -9.88693237e-01
-1.64530480e+00 7.32753277e-01 1.94396190e-02 -3.69571179e-01
3.75835925e-01 -4.82832640e-01 -8.44969690e-01 -1.68283015e-01
1.89552546e-01 6.01340055e-01 3.55263382e-01 -9.95976985e-01
-8.64144564e-01 -1.39409721e-01 1.67565376e-01 7.81531185e-02
-6.66712761e-01 1.95868626e-01 -5.93593478e-01 -3.25753808e-01
-4.77023423e-01 -7.86228895e-01 -2.45601565e-01 -6.50290489e-01
-3.88106465e-01 -3.82264525e-01 6.30324364e-01 -7.86692619e-01
1.19776320e+00 -2.02234030e+00 -2.46529132e-01 2.55715381e-02
-1.28182441e-01 6.02352023e-01 -2.88539410e-01 4.94121909e-01
-4.44160700e-02 2.56953657e-01 -1.35844648e-01 1.32960141e-01
1.64406538e-01 2.40196064e-01 -2.58938074e-01 -1.31906092e-01
2.50126153e-01 6.90511584e-01 -1.02288210e+00 -5.11890173e-01
3.67140435e-02 3.84057820e-01 -3.97783607e-01 3.03144783e-01
-6.55101240e-02 2.34661803e-01 -5.26215136e-01 6.70257270e-01
3.60693038e-01 -1.57254040e-01 3.57692242e-01 -1.48238167e-01
-3.30215454e-01 5.71322858e-01 -1.35175681e+00 1.87821555e+00
-9.71837878e-01 2.86755502e-01 2.15876237e-01 -8.27357948e-01
8.83150816e-01 5.65759242e-01 3.44470561e-01 -7.33508527e-01
-9.00828466e-02 6.50363266e-01 -1.92540720e-01 -3.81828994e-01
5.34507036e-01 -1.02212742e-01 -2.59222865e-01 4.23459440e-01
4.76816118e-01 1.60163879e-01 5.05563796e-01 -1.22289602e-02
1.15834701e+00 2.19855055e-01 3.12789649e-01 1.92310989e-01
6.54829919e-01 1.36558443e-01 6.73522115e-01 6.86331630e-01
-1.90685689e-01 2.18409702e-01 -7.80042782e-02 -2.51653820e-01
-8.64760399e-01 -8.31095159e-01 -3.22949551e-02 1.38370907e+00
-2.29825422e-01 -4.45486397e-01 -9.34623718e-01 -1.18871331e+00
-3.44270945e-01 7.31023014e-01 1.06322274e-01 1.76705882e-01
-6.82428896e-01 -4.28759515e-01 9.31512773e-01 5.19655645e-01
6.68353736e-01 -1.62937260e+00 -2.31395110e-01 4.33321476e-01
-2.88791060e-01 -1.41947830e+00 -6.08822107e-01 6.94374442e-01
-5.59542775e-01 -8.76064420e-01 -7.89697587e-01 -9.99825060e-01
2.51562446e-01 1.89276151e-02 1.24428535e+00 -1.75129935e-01
-3.51319239e-02 4.22342211e-01 -4.92644161e-01 -2.82745093e-01
-4.52445716e-01 6.92801476e-01 2.22834989e-01 -2.14290947e-01
8.70996416e-01 -4.41811621e-01 -2.11210802e-01 3.71898264e-01
-9.24134493e-01 -4.04043704e-01 9.04856682e-01 9.70376432e-01
5.74390769e-01 7.39567503e-02 7.61040807e-01 -1.02713764e+00
6.92085743e-01 -4.59475070e-01 -6.45285249e-01 5.38537681e-01
-4.70667213e-01 2.89525479e-01 8.02418888e-01 -4.96298075e-01
-1.13885164e+00 2.27473751e-01 -4.03300315e-01 -2.99299031e-01
-4.50453162e-01 7.67936230e-01 -4.65302616e-01 2.26542175e-01
5.89921713e-01 1.79138258e-01 -5.09232283e-01 -9.04977381e-01
5.69639504e-01 1.25020921e+00 5.08094132e-01 -5.90959787e-01
6.35320425e-01 4.52002771e-02 -7.71387041e-01 -8.30211163e-01
-9.58572209e-01 -7.90799797e-01 -5.08786082e-01 2.77590841e-01
7.52950132e-01 -9.41115797e-01 -4.84433621e-01 1.60427496e-01
-1.20795166e+00 -3.15997183e-01 -4.18717265e-01 5.36026597e-01
-1.89300507e-01 2.79896140e-01 -7.04160810e-01 -6.56105995e-01
-5.72782815e-01 -9.91293252e-01 8.76225829e-01 4.83781010e-01
-5.83704375e-02 -9.18179572e-01 2.84947127e-01 4.43914503e-01
5.20377696e-01 -4.32053357e-01 6.90063059e-01 -1.37678504e+00
-3.33565772e-01 -1.73194423e-01 -3.26560810e-02 7.74297118e-01
2.99873710e-01 -3.49387556e-01 -9.57678914e-01 -8.07767063e-02
-8.11364315e-03 -6.45732522e-01 4.57428396e-01 -3.32634389e-01
8.23202550e-01 -1.00578696e-01 -7.53773600e-02 5.16301155e-01
1.44104910e+00 1.67072281e-01 2.72752553e-01 5.00196993e-01
6.93562508e-01 8.31250727e-01 4.82652903e-01 6.37623742e-02
4.44362730e-01 4.79186267e-01 -6.80965558e-02 -2.63728082e-01
-1.61372140e-01 -4.93301898e-01 5.90804935e-01 1.44389534e+00
7.51540661e-02 -1.64395943e-01 -1.04344809e+00 6.15322292e-01
-1.55598211e+00 -8.35682213e-01 1.28623188e-01 2.17414665e+00
1.19558144e+00 -6.92102313e-02 -1.67986721e-01 -3.50269079e-01
9.29808080e-01 -1.65126994e-01 -5.54521143e-01 -3.14131856e-01
-6.44928664e-02 5.08347571e-01 5.23792386e-01 3.39247435e-01
-1.14040267e+00 1.36852694e+00 5.53401136e+00 9.60109234e-01
-1.36062348e+00 3.08406174e-01 1.58186302e-01 2.96787798e-01
-1.28280029e-01 4.23173569e-02 -1.24896741e+00 2.59998471e-01
1.35639977e+00 -7.22197592e-02 6.48445547e-01 1.19828546e+00
-7.78544098e-02 3.09525043e-01 -1.18633759e+00 1.01116848e+00
-1.28184065e-01 -9.41812634e-01 -4.10197377e-02 7.50115663e-02
5.60776412e-01 5.85367680e-01 -3.69246095e-01 9.37363386e-01
7.44064212e-01 -8.42423320e-01 3.33333611e-01 2.37201646e-01
8.88706088e-01 -7.59842217e-01 9.76518214e-01 6.37781262e-01
-1.03192329e+00 2.06432298e-01 -4.41100597e-01 3.59433770e-01
3.23802501e-01 4.88597602e-01 -1.02519166e+00 6.89248323e-01
7.37494767e-01 3.35040450e-01 -2.69893795e-01 8.23893845e-01
-4.26128119e-01 6.56898320e-01 -5.91569066e-01 -8.96485429e-03
1.51700929e-01 -1.20586745e-01 2.69276232e-01 1.31994605e+00
4.27584261e-01 -2.53938913e-01 3.78365040e-01 6.88080311e-01
-5.87387383e-01 5.16312838e-01 -7.25666285e-01 -5.39491773e-01
4.42401409e-01 1.46746051e+00 -4.41982448e-01 -3.37469280e-01
-7.52083540e-01 1.18453419e+00 4.89566624e-01 4.78669375e-01
-6.10377729e-01 -7.98831820e-01 3.03535879e-01 -9.74127725e-02
3.38823140e-01 -1.55287445e-01 1.48815125e-01 -1.42148292e+00
9.78419930e-02 -1.38730991e+00 3.72948915e-01 -5.39283633e-01
-1.58054113e+00 1.16555727e+00 -4.68937397e-01 -1.06987536e+00
-6.48658812e-01 -6.69570386e-01 -2.77918190e-01 9.48017001e-01
-1.84827864e+00 -1.13295221e+00 -1.29520178e-01 8.23573768e-01
6.36761665e-01 -4.36268955e-01 1.18722308e+00 6.79102123e-01
-5.30088902e-01 6.30497158e-01 2.29195356e-01 6.64389551e-01
1.14569819e+00 -1.20646119e+00 2.36386880e-01 7.27922022e-01
7.63965487e-01 9.96878862e-01 1.18127801e-01 -5.16596735e-01
-1.29535055e+00 -8.49030375e-01 1.20991123e+00 -3.56831074e-01
9.68467176e-01 -3.53114486e-01 -8.26779723e-01 7.14365363e-01
5.38388669e-01 -2.31900532e-02 1.02504158e+00 4.51976418e-01
-3.50440383e-01 -2.77297080e-01 -8.82218122e-01 4.35312033e-01
7.99303174e-01 -8.17402542e-01 -8.49669456e-01 2.43691102e-01
8.24511945e-01 -2.19809070e-01 -9.44040716e-01 3.60265762e-01
2.42399454e-01 -7.79848814e-01 6.03063524e-01 -7.40706265e-01
1.38633028e-01 -2.87231117e-01 -2.21922576e-01 -1.38294494e+00
2.41878048e-01 -6.53227806e-01 1.75410852e-01 1.80136728e+00
7.20011175e-01 -6.61614835e-01 6.84105933e-01 5.88900506e-01
-2.37772427e-02 -3.42535436e-01 -7.82714546e-01 -8.06923211e-01
4.74093109e-03 -5.24741173e-01 6.36018634e-01 1.11282504e+00
-7.87469596e-02 6.99628949e-01 -2.00544387e-01 1.33258387e-01
2.30656222e-01 6.35955706e-02 6.98767245e-01 -1.06966162e+00
-3.41763228e-01 -1.65112074e-02 2.71164011e-02 -1.11328149e+00
4.95694786e-01 -9.93603408e-01 2.06103146e-01 -1.38394439e+00
-1.81930289e-02 -7.17706144e-01 -5.96424818e-01 8.56416285e-01
8.12489316e-02 9.05857980e-02 2.23055065e-01 6.62605464e-02
-6.63260043e-01 4.42202628e-01 7.59672225e-01 -1.31249338e-01
-3.55402321e-01 -2.09906921e-02 -7.34775424e-01 9.05918419e-01
6.89015627e-01 -5.73594570e-01 -2.54602820e-01 -6.12429738e-01
1.09794892e-01 -2.85645816e-02 -1.30641177e-01 -8.11733246e-01
5.13052881e-01 1.56825073e-02 1.22545987e-01 -3.57616633e-01
1.04766622e-01 -1.04998767e+00 -2.28153557e-01 -2.30146870e-02
-2.10134685e-01 -2.84509622e-02 1.25463188e-01 2.10596874e-01
-6.61693871e-01 -4.55197722e-01 3.61863226e-01 -3.00877213e-01
-1.12038827e+00 1.18399709e-01 -1.23813674e-01 3.68106335e-01
6.03755832e-01 6.72974670e-03 -9.30448472e-02 -3.98623496e-01
-6.77608490e-01 3.27504158e-01 3.31166208e-01 6.24163449e-01
2.67938614e-01 -1.24614096e+00 -4.30080593e-01 1.29532903e-01
1.32975549e-01 -2.69907415e-01 -7.34414905e-02 4.87180620e-01
-3.05803776e-01 6.36672139e-01 -5.74595183e-02 -2.11776078e-01
-8.34263384e-01 5.36452115e-01 3.51911753e-01 -6.78585887e-01
-1.44732922e-01 5.57313621e-01 -8.70864764e-02 -1.18898439e+00
3.42741519e-01 -7.46337473e-02 -2.62046903e-01 1.74909539e-03
4.03101802e-01 7.84449205e-02 3.66637260e-01 -4.72345859e-01
-3.54708701e-01 2.62913853e-01 -1.44170776e-01 -2.78713733e-01
1.43534756e+00 -9.38358679e-02 -4.96903285e-02 4.86999154e-01
1.11929047e+00 5.67687273e-01 -6.70276523e-01 -3.53393078e-01
5.51278770e-01 2.08120137e-01 -4.27944213e-02 -1.05710697e+00
-1.05607879e+00 9.81678367e-01 6.09553874e-01 4.98471670e-02
1.10968053e+00 -9.01275873e-02 1.19207788e+00 1.00475156e+00
7.33746767e-01 -1.42898142e+00 -3.00408304e-01 7.98143327e-01
5.11751652e-01 -1.46176887e+00 -6.80566847e-01 -2.03333735e-01
-6.78272963e-01 1.05674446e+00 6.21386766e-01 1.88936025e-01
6.50084794e-01 4.66085166e-01 5.69561541e-01 3.56113687e-02
-4.64889437e-01 -6.26626194e-01 2.23990023e-01 5.13486564e-01
9.18812275e-01 -1.00961335e-01 -3.42836499e-01 9.33606327e-01
-2.37801626e-01 1.23239473e-01 2.27757126e-01 9.59531188e-01
-2.25275919e-01 -1.80591142e+00 -1.66601345e-01 1.18854478e-01
-7.83407629e-01 -6.08514130e-01 -4.52062517e-01 8.66006255e-01
1.76442772e-01 1.07919800e+00 -3.99054676e-01 -3.19487303e-01
5.97109497e-01 5.35420954e-01 -3.38957235e-02 -8.44078422e-01
-8.47448766e-01 6.56060353e-02 4.26979244e-01 -5.71170807e-01
-4.68556195e-01 -1.59178972e-01 -1.33280885e+00 2.72585405e-03
-5.46725512e-01 6.28705442e-01 9.44523513e-01 8.96144986e-01
4.21745390e-01 3.06193650e-01 5.46625614e-01 -3.94897878e-01
-8.95450413e-01 -9.72827792e-01 -5.49845397e-01 2.71497667e-01
-6.35594279e-02 -2.55163699e-01 -2.63735712e-01 7.12142661e-02] | [10.001861572265625, 9.671125411987305] |
fc718eb0-62da-49d7-837c-962b9559ff42 | tracking-motion-and-proxemics-using-thermal | 1511.08166 | null | http://arxiv.org/abs/1511.08166v1 | http://arxiv.org/pdf/1511.08166v1.pdf | Tracking Motion and Proxemics using Thermal-sensor Array | Indoor tracking has all-pervasive applications beyond mere surveillance, for
example in education, health monitoring, marketing, energy management and so
on. Image and video based tracking systems are intrusive. Thermal array sensors
on the other hand can provide coarse-grained tracking while preserving privacy
of the subjects. The goal of the project is to facilitate motion detection and
group proxemics modeling using an 8 x 8 infrared sensor array. Each of the 8 x
8 pixels is a temperature reading in Fahrenheit. We refer to each 8 x 8 matrix
as a scene. We collected approximately 902 scenes with different configurations
of human groups and different walking directions. We infer direction of motion
of a subject across a set of scenes as left-to-right, right-to-left, up-to-down
and down-to-up using cross-correlation analysis. We used features from
connected component analysis of each background subtracted scene and performed
Support Vector Machine classification to estimate number of instances of human
subjects in the scene. | ['Chandrayee Basu', 'Anthony Rowe'] | 2015-11-25 | null | null | null | null | ['motion-detection'] | ['computer-vision'] | [ 2.38379821e-01 -5.97507894e-01 2.01210350e-01 -3.93974423e-01
5.42989420e-03 -6.00479662e-01 2.79680640e-01 -1.02650933e-01
-3.83255243e-01 5.32584846e-01 4.23617035e-01 -1.28877580e-01
-1.18336782e-01 -6.94312751e-01 -3.90930086e-01 -7.93461204e-01
-2.63828665e-01 -1.92673177e-01 1.27296150e-01 1.06115364e-01
-3.46893375e-03 4.42250758e-01 -1.49130464e+00 1.79077104e-01
2.19314426e-01 9.94917154e-01 -1.92318901e-01 1.09596920e+00
4.91956532e-01 5.20480871e-01 -6.65933371e-01 1.52395010e-01
3.86007249e-01 -3.64157200e-01 1.11454673e-01 1.57607064e-01
5.81203878e-01 -3.38712245e-01 -3.02297950e-01 8.76572013e-01
4.16577548e-01 2.84501851e-01 4.47058380e-01 -1.49005461e+00
-1.48449168e-01 -2.80880362e-01 -1.05692625e+00 2.01174930e-01
8.42308879e-01 -6.57203421e-02 1.46159604e-01 -2.50328451e-01
4.65087980e-01 1.01401603e+00 8.67591321e-01 2.63878167e-01
-1.08865547e+00 -9.01405692e-01 1.44554496e-01 1.76039472e-01
-1.47014976e+00 -3.01629454e-01 9.01734293e-01 -6.36618197e-01
7.82011747e-01 9.14236546e-01 1.04433846e+00 9.93467808e-01
4.31373328e-01 9.48222876e-02 1.22572875e+00 -3.66289735e-01
3.29629004e-01 2.45053619e-01 1.48246333e-01 6.10011995e-01
5.82465410e-01 -4.31040898e-02 -4.38269347e-01 -2.94937819e-01
7.03578830e-01 2.48255894e-01 -1.85600638e-01 -4.17332649e-01
-1.23247516e+00 4.87795353e-01 2.58083880e-01 2.62881637e-01
-1.99684858e-01 2.90963888e-01 -1.48406159e-02 6.99464232e-02
4.31182146e-01 -1.35836571e-01 -4.43419188e-01 7.65919089e-02
-8.76199901e-01 2.03190640e-01 5.94966292e-01 9.06663656e-01
6.24249876e-01 -2.43331790e-01 8.42546150e-02 2.87217319e-01
6.40779853e-01 7.48235106e-01 2.60767192e-01 -9.31800604e-01
4.01210546e-01 6.32768929e-01 3.53022873e-01 -1.50649691e+00
-7.92484522e-01 1.40239984e-01 -9.48244631e-01 2.83484071e-01
5.41052938e-01 -5.45249581e-01 -5.58630049e-01 1.45126355e+00
7.36812234e-01 3.25321615e-01 -4.52305675e-01 9.60464418e-01
4.97851789e-01 6.15891874e-01 2.57843345e-01 -3.94397616e-01
1.84388721e+00 -3.06044161e-01 -8.44672620e-01 -3.12795103e-01
2.46083111e-01 -5.74765027e-01 5.42388856e-01 2.99239635e-01
-3.94096822e-01 -6.65848553e-01 -8.79633904e-01 4.54954624e-01
-5.11859119e-01 8.22352320e-02 3.37275147e-01 1.37186742e+00
-8.22712958e-01 2.90014148e-01 -1.03007030e+00 -7.50862420e-01
3.72852981e-02 4.60683018e-01 -4.67651606e-01 1.23214886e-01
-7.21805513e-01 6.09969199e-01 -4.28588316e-03 4.71634306e-02
-1.20318495e-01 -5.02078593e-01 -9.78351355e-01 -5.83169639e-01
2.04386309e-01 -6.47145629e-01 5.00820816e-01 -8.54479790e-01
-1.19947147e+00 8.81973743e-01 -3.40378940e-01 -1.32633194e-01
6.45774245e-01 -4.43118334e-01 -9.61504281e-01 -1.45494223e-01
8.72198716e-02 3.28927606e-01 8.92286420e-01 -9.00617063e-01
-4.03227925e-01 -9.14914906e-01 -5.41982949e-01 3.65517884e-01
-5.88458329e-02 3.18284512e-01 -6.34793565e-02 -5.01864910e-01
4.20055836e-01 -1.24053884e+00 -3.14510912e-01 1.54791713e-01
-6.25321984e-01 2.97104508e-01 1.27984357e+00 -8.41590405e-01
1.12962270e+00 -2.06280613e+00 -2.70317405e-01 3.11489284e-01
1.31680116e-01 -1.93135560e-01 4.47191983e-01 1.81889474e-01
-1.00622579e-01 -2.14894101e-01 2.59285212e-01 -1.37615785e-01
-7.52530023e-02 -1.53823733e-01 2.54036784e-01 1.36727607e+00
-4.51548934e-01 2.84892797e-01 -6.71950996e-01 -4.70551640e-01
9.24801886e-01 5.67027152e-01 -2.41954759e-01 1.87162121e-04
3.56369436e-01 6.65090144e-01 -5.07062495e-01 5.24168253e-01
8.51331532e-01 2.81164169e-01 -6.63943682e-03 -1.49658397e-01
-3.91347289e-01 -3.16523224e-01 -1.56511199e+00 1.46749389e+00
-4.64047305e-03 9.44672465e-01 2.46084556e-01 -4.93107706e-01
6.95692360e-01 4.43472594e-01 9.79727089e-01 -6.15587115e-01
1.74811184e-01 -4.14533645e-01 -2.58747935e-01 -6.94377482e-01
6.49416506e-01 5.61887249e-02 -2.86875784e-01 2.97933131e-01
-7.40723789e-01 5.54016292e-01 -4.56343442e-01 -1.10872574e-01
1.22731042e+00 1.48612618e-01 4.80680823e-01 -4.32577342e-01
3.68810952e-01 -7.35367760e-02 5.18347025e-01 5.83738208e-01
-7.71086454e-01 4.02688235e-01 -2.29118288e-01 -5.75614631e-01
-7.81547785e-01 -1.05719090e+00 9.47485715e-02 1.20507371e+00
3.02425206e-01 -2.52874851e-01 -7.31660008e-01 4.21930887e-02
2.87834462e-02 5.76772571e-01 -6.19083643e-01 6.61815405e-02
-5.08503199e-01 -7.44269252e-01 3.29124004e-01 4.08882052e-01
8.58291030e-01 -7.54101694e-01 -1.37587476e+00 -6.02465235e-02
-2.60132641e-01 -1.06999254e+00 -4.81364369e-01 -7.19439164e-02
-6.54921591e-01 -1.05999124e+00 -3.78344953e-01 -3.79211247e-01
6.90019786e-01 4.93951470e-01 8.67036700e-01 -3.18362951e-01
-8.04251671e-01 7.99409032e-01 1.11108320e-02 -6.24460816e-01
3.49803090e-01 -7.08690405e-01 5.69573820e-01 1.17741331e-01
7.02613592e-01 -2.59209126e-01 -1.01049840e+00 5.75629771e-01
-4.65940163e-02 -2.54251629e-01 -1.83759362e-01 -1.27260894e-01
3.16347092e-01 3.60956281e-01 -2.44359165e-01 -4.32672739e-01
9.56114978e-02 -4.34085190e-01 -7.65564203e-01 5.88007160e-02
5.18111326e-02 -5.90925932e-01 1.15037315e-01 -5.06229043e-01
-9.08175766e-01 5.99126816e-01 4.97898132e-01 -1.59664288e-01
-7.24239588e-01 -3.07970852e-01 -4.11001831e-01 3.91771793e-02
6.33231163e-01 -6.25760332e-02 -4.30406779e-01 -4.27375734e-02
2.82448381e-01 5.32208681e-01 6.12456799e-01 -7.17930719e-02
8.42494845e-01 7.99539864e-01 1.61146462e-01 -1.57536387e+00
-4.62779589e-02 -8.52506757e-01 -6.99670553e-01 -8.43573093e-01
1.54800117e+00 -1.09445143e+00 -8.87744427e-01 6.26799107e-01
-7.44244576e-01 -1.39305905e-01 1.79809436e-01 5.48450768e-01
-1.09129094e-01 1.03116818e-01 -2.83279449e-01 -1.50390840e+00
-1.99740931e-01 -6.08841479e-01 1.13827837e+00 4.09344733e-01
-6.64387465e-01 -8.70921671e-01 1.57666475e-01 5.41400552e-01
2.53559262e-01 7.86482871e-01 6.03344999e-02 -1.56295653e-02
-4.23276544e-01 -5.62633872e-01 2.22761288e-01 -2.17485487e-01
3.90410930e-01 7.13041946e-02 -1.23142910e+00 -3.36584628e-01
3.62391531e-01 4.40826982e-01 2.51854867e-01 1.03278995e+00
8.54789495e-01 -4.14454103e-01 -7.68367827e-01 4.93051440e-01
1.43739319e+00 5.32330155e-01 8.54673028e-01 3.40031385e-01
1.04236543e+00 6.96879447e-01 4.62448508e-01 4.86153543e-01
1.73864499e-01 8.42935026e-01 1.29548505e-01 5.26612811e-02
2.66554117e-01 3.30411904e-02 5.14520109e-01 1.40873194e-01
-2.55781442e-01 -2.58425772e-01 -7.24030912e-01 1.93961456e-01
-1.55149662e+00 -1.26057529e+00 -5.79829097e-01 2.50067687e+00
-7.20829368e-02 -9.56534296e-02 3.99055749e-01 2.97410667e-01
9.27419126e-01 1.69370741e-01 -3.25221002e-01 -1.24501579e-01
5.05247235e-01 -2.84963518e-01 9.19463336e-01 3.41043979e-01
-1.37018335e+00 2.78159142e-01 6.11678028e+00 9.31539237e-02
-8.41762125e-01 -1.73896730e-01 5.56221724e-01 -2.93289304e-01
3.74866962e-01 -3.81110698e-01 -6.59117043e-01 5.64638436e-01
9.13940251e-01 3.42280328e-01 2.87533671e-01 8.08119416e-01
4.10872877e-01 -7.72161067e-01 -1.03321433e+00 1.16497624e+00
3.70194181e-03 -6.30340576e-01 -1.02876389e+00 2.50312656e-01
5.95738232e-01 -2.56579190e-01 -3.85728441e-02 -2.22855031e-01
1.29331738e-01 -7.84409642e-01 4.73746955e-01 3.96187663e-01
5.90096414e-01 -5.39520502e-01 2.26393223e-01 4.04293478e-01
-1.79732549e+00 1.35858580e-01 -1.38859972e-01 -4.79068339e-01
-2.59596519e-02 3.59205842e-01 -5.45085549e-01 2.23498121e-01
1.12807405e+00 4.46466088e-01 -4.27285254e-01 7.68726110e-01
3.62825066e-01 2.91697592e-01 -6.30369306e-01 -1.48426771e-01
-2.95925796e-01 -5.01028776e-01 6.55545235e-01 1.16491342e+00
4.94924158e-01 4.39764947e-01 2.08435699e-01 4.69518840e-01
4.17604059e-01 -4.44424093e-01 -1.16153896e+00 5.88721931e-01
5.58057308e-01 1.16606522e+00 -1.17781866e+00 -1.79330125e-01
-4.46942240e-01 1.01772845e+00 -6.15082026e-01 4.69996154e-01
-8.52507472e-01 -3.39340299e-01 9.16699111e-01 4.97669369e-01
6.08262643e-02 -4.13237125e-01 -3.00647587e-01 -7.91569352e-01
-8.01092386e-02 -5.14699876e-01 4.34063196e-01 -7.50406861e-01
-9.09536898e-01 -5.35245277e-02 2.14190915e-01 -1.34287524e+00
-8.88009667e-02 -5.37645638e-01 -6.26714885e-01 7.26946712e-01
-4.20408428e-01 -7.33456910e-01 -8.88028145e-01 8.54829371e-01
2.42737755e-01 5.32972440e-03 7.73738980e-01 3.34800839e-01
-7.08007038e-01 3.93848002e-01 6.75094873e-02 5.11611760e-01
4.81861264e-01 -1.03824830e+00 2.46658981e-01 1.08043158e+00
-2.29439493e-02 5.99143326e-01 9.25649941e-01 -8.89418185e-01
-1.64312959e+00 -1.15167904e+00 5.63132346e-01 -9.11924481e-01
4.41468656e-01 -6.26010597e-01 -3.51014048e-01 9.47917342e-01
1.73518702e-01 2.47389778e-01 9.00286913e-01 6.68750331e-02
3.18066142e-02 -2.84815609e-01 -1.58796763e+00 5.74087322e-01
8.93262327e-01 -4.17416781e-01 -1.84016496e-01 2.32183307e-01
4.26236242e-02 -4.93953377e-01 -7.57965088e-01 -7.68117085e-02
8.47169459e-01 -1.31271076e+00 1.29851007e+00 -2.57782526e-02
-2.98706234e-01 -5.89334249e-01 -3.96568030e-01 -8.26597393e-01
-5.46766281e-01 -4.54770833e-01 1.47709548e-01 1.15772259e+00
-3.06108445e-01 -5.88144422e-01 1.20961547e+00 9.91607606e-01
2.46576786e-01 8.48735943e-02 -1.16847730e+00 -6.56707525e-01
-5.34413755e-01 -7.04184115e-01 2.31305957e-01 9.30453002e-01
6.68504983e-02 1.50150418e-01 -6.31746888e-01 6.05253637e-01
1.09392047e+00 -1.89939857e-01 8.53331029e-01 -1.21660113e+00
-1.31178349e-01 -3.22201811e-02 -8.83758485e-01 -4.60689396e-01
-5.12203336e-01 -7.30041638e-02 1.08280718e-01 -1.42775607e+00
5.40901087e-02 -1.45897657e-01 7.86695108e-02 1.62173808e-01
8.28149989e-02 3.67682427e-01 -6.35840818e-02 -2.56253183e-01
-5.24915278e-01 -1.39582053e-01 7.21615016e-01 -4.95361052e-02
-3.85748506e-01 1.96846619e-01 -2.25765005e-01 9.09402668e-01
6.36418283e-01 -3.54604214e-01 -3.77495080e-01 -9.41390172e-02
-5.48679032e-04 1.85226724e-01 5.44990301e-01 -1.58891916e+00
3.41357321e-01 -1.62346259e-01 1.10694671e+00 -6.77772760e-01
7.68890679e-01 -1.60856354e+00 8.97050679e-01 9.19843614e-01
1.25592202e-01 2.29567125e-01 2.56348491e-01 8.22948217e-01
4.63281989e-01 4.08473134e-01 4.96949077e-01 -3.30159247e-01
-7.28566945e-01 -9.83468890e-02 -6.37497425e-01 -3.96091282e-01
1.61777365e+00 -6.34082854e-01 -3.13685358e-01 -4.66383040e-01
-5.04579365e-01 -1.18762078e-02 5.38094699e-01 3.28121454e-01
4.24269497e-01 -1.56757474e+00 -3.21307272e-01 4.17850912e-01
-8.56553018e-03 -3.80328655e-01 3.02498311e-01 6.18386507e-01
-5.32951713e-01 3.38512301e-01 -3.02221477e-01 -8.18236828e-01
-2.09669113e+00 4.95191485e-01 3.71861815e-01 1.76925778e-01
-7.05041289e-01 7.62519062e-01 3.44476476e-02 1.05586834e-01
3.54261369e-01 -4.78970408e-01 -9.39658582e-02 -1.21473735e-02
6.93619013e-01 1.04362345e+00 -2.59802997e-01 -9.52709675e-01
-9.85628963e-01 9.63950753e-01 5.29529393e-01 -5.74925244e-02
8.65404308e-01 -3.36371005e-01 -2.93255001e-02 6.60031021e-01
1.04977024e+00 1.02635667e-01 -1.14760804e+00 3.83958995e-01
-1.10120654e-01 -6.32101297e-01 -2.25278795e-01 -4.24580097e-01
-8.38745356e-01 4.64470297e-01 1.41955328e+00 3.78978282e-01
1.25909352e+00 -3.56093526e-01 2.54041582e-01 8.41797367e-02
6.92600191e-01 -1.18178749e+00 -1.29761353e-01 -1.25143677e-01
1.55683815e-01 -1.31291914e+00 4.73279357e-01 -4.75850731e-01
-5.45871019e-01 7.14938760e-01 3.88574958e-01 -7.18734413e-02
7.51629233e-01 4.76961941e-01 1.17659740e-01 -1.06640846e-01
-2.04245016e-01 1.46011397e-01 4.22199994e-01 9.10777986e-01
4.90520626e-01 5.10375082e-01 4.15766984e-01 -2.16497909e-02
-3.29301447e-01 -2.51416545e-02 1.77549481e-01 1.19263828e+00
-5.50115526e-01 -4.14795518e-01 -1.31227243e+00 4.71848339e-01
-3.60077053e-01 4.65508401e-01 -2.46297926e-01 8.00562024e-01
3.51447612e-01 1.51905477e+00 3.28860402e-01 -7.68695772e-01
2.84800172e-01 -5.08534238e-02 5.21563053e-01 -1.42247856e-01
-5.53032279e-01 1.53983563e-01 1.40651837e-01 -7.18466222e-01
-5.46531498e-01 -1.01079881e+00 -8.13136816e-01 -5.15569091e-01
2.96519324e-02 -3.45330656e-01 8.83603752e-01 6.46614075e-01
1.64426155e-02 5.42674422e-01 3.53191406e-01 -1.00319004e+00
2.73752838e-01 -8.27524722e-01 -8.14056337e-01 3.85855585e-01
5.18710196e-01 -3.99046183e-01 -1.53261319e-01 3.12287331e-01] | [6.9697394371032715, 0.36830049753189087] |
aa82ef1c-9775-4a14-9323-d550272a2043 | saliency-detection-and-quantization-index | 2302.11361 | null | https://arxiv.org/abs/2302.11361v2 | https://arxiv.org/pdf/2302.11361v2.pdf | HDR image watermarking using saliency detection and quantization index modulation | High-dynamic range (HDR) images are circulated rapidly over the internet with risks of being exploited for unauthorized usage. To protect these images, some HDR image based watermarking (HDR-IW) methods were put forward. However, they inherited the same problem faced by conventional IW methods for standard dynamic range (SDR) images, where only trade-offs among conflicting requirements are managed instead of simultaneous improvement. In this paper, a novel saliency (eye-catching object) detection based trade-off independent HDR-IW is proposed, to simultaneously improve robustness, imperceptibility and payload. First, the host image goes through our proposed salient object detection model to produce a saliency map, which is, in turn, exploited to segment the foreground and background of the host image. Next, the binary watermark is partitioned into the foregrounds and backgrounds using the same mask and scrambled using a random permutation algorithm. Finally, the watermark segments are embedded into selected bit-plane of the corresponding host segments using quantized indexed modulation. Experimental results suggest that the proposed work outperforms state-of-the-art methods in terms of improving the conflicting requirements. | ['Vishnu Monn Baskaran', 'KokSheik Wong', 'Minoru Kuribayashi', 'Ahmed Khan'] | 2023-02-22 | null | null | null | null | ['saliency-detection'] | ['computer-vision'] | [ 9.25423801e-01 -1.98601365e-01 -4.58709717e-01 2.50535995e-01
-2.27457911e-01 -4.28338826e-01 3.39202732e-01 -7.56930411e-02
-2.86688179e-01 6.36961877e-01 -4.80860732e-02 -2.23161191e-01
-1.82209194e-01 -6.42079175e-01 -1.48868933e-01 -8.98034751e-01
-2.98470736e-01 -5.87962568e-01 9.13071871e-01 -2.73088038e-01
7.33110011e-01 5.31441987e-01 -1.38374257e+00 -1.06022954e-01
8.12581360e-01 1.04698300e+00 3.63938928e-01 4.44031775e-01
1.54192492e-01 8.66404533e-01 -5.75341582e-01 -1.00565031e-01
5.36112547e-01 -5.43159962e-01 -3.32709253e-01 3.63263726e-01
-1.31290674e-01 -4.87610072e-01 -6.43867254e-01 1.38996041e+00
4.31989312e-01 -7.60332346e-02 2.78844923e-01 -1.30433667e+00
-1.15287375e+00 3.88380527e-01 -1.38706076e+00 4.31752443e-01
3.30758870e-01 9.47119072e-02 6.71232581e-01 -6.53968036e-01
6.83324695e-01 1.09733164e+00 1.88956428e-02 5.02859592e-01
-1.00879526e+00 -9.19575155e-01 1.22972824e-01 4.12997574e-01
-1.45690048e+00 -3.26062828e-01 1.22857642e+00 1.46162063e-01
3.12668324e-01 6.23544753e-01 7.20422089e-01 4.75083917e-01
3.18587601e-01 5.99445105e-01 1.34261692e+00 -3.81530702e-01
7.47474656e-02 2.33580470e-01 -1.14685912e-02 5.72589159e-01
5.80380857e-01 1.20913349e-01 -5.12977600e-01 -5.08102514e-02
9.22160566e-01 -7.50989129e-04 -7.95131207e-01 -4.66176748e-01
-1.54087746e+00 4.46494639e-01 4.52304572e-01 4.43403572e-01
-9.63960886e-02 -1.24066316e-01 1.02171525e-01 2.87615240e-01
9.60025638e-02 4.66959327e-02 -8.97245109e-02 3.46302003e-01
-1.07330489e+00 -1.33725807e-01 2.93732196e-01 7.68776894e-01
5.64329624e-01 1.11636885e-01 -3.07937083e-03 6.06470227e-01
8.00965130e-01 6.05143547e-01 7.12199211e-01 -6.22443020e-01
3.46665651e-01 4.97495294e-01 1.49426177e-01 -1.59375250e+00
1.00513389e-02 -2.80970365e-01 -9.25389707e-01 4.88063723e-01
4.54960354e-02 2.33857617e-01 -1.05770922e+00 1.31773627e+00
2.89246202e-01 1.47008136e-01 3.00581366e-01 1.09124625e+00
5.84169686e-01 1.00924039e+00 1.89315081e-02 -4.07410443e-01
1.46260393e+00 -7.12609351e-01 -9.01811182e-01 -2.06619725e-01
-1.53066561e-01 -1.05594516e+00 5.96473157e-01 2.64514357e-01
-1.03277898e+00 -5.66212177e-01 -1.58718717e+00 1.08378917e-01
-3.33018512e-01 -1.28638268e-01 2.16372851e-02 8.55518460e-01
-9.28817093e-01 2.06325546e-01 -3.01895708e-01 -5.96932992e-02
1.49066910e-01 3.06507260e-01 -1.93386421e-01 3.45960520e-02
-1.26380539e+00 7.58006930e-01 7.60191858e-01 1.57091394e-01
-7.01970518e-01 -2.89331108e-01 -8.64633679e-01 -9.99423265e-02
3.05642366e-01 -9.87398028e-02 5.14993548e-01 -1.10755408e+00
-1.55252945e+00 7.94105828e-01 1.71058834e-01 -4.20991391e-01
4.71832722e-01 3.32531214e-01 -7.91856766e-01 5.34653366e-01
-2.22588196e-01 6.30476356e-01 1.52830338e+00 -1.56361747e+00
-7.64754355e-01 -1.23227589e-01 -1.52092859e-01 1.17048450e-01
-1.46506384e-01 2.20437944e-01 -2.52531856e-01 -9.78767574e-01
5.29777765e-01 -8.37990463e-01 -3.85878384e-02 1.45144939e-01
-4.44193482e-01 4.21365231e-01 1.61870897e+00 -8.54688048e-01
1.56630552e+00 -2.57874775e+00 -8.71904287e-03 3.51540238e-01
2.26043120e-01 5.51965535e-01 -1.98383942e-01 1.79475084e-01
-3.08195204e-01 2.04424933e-01 -4.37176526e-01 3.81366044e-01
-2.66782403e-01 -3.27232718e-01 -4.44802910e-01 8.84078383e-01
2.25700349e-01 5.53479552e-01 -8.11411262e-01 -8.15870106e-01
4.52482581e-01 5.36401153e-01 7.36544356e-02 -1.06070012e-01
3.62549648e-02 4.05852765e-01 -3.84762317e-01 9.18232024e-01
1.22861814e+00 4.16743495e-02 -3.09358183e-02 -2.69799858e-01
-3.57088745e-01 -2.99947351e-01 -1.34200263e+00 9.40487206e-01
5.80415130e-03 6.27116799e-01 -9.74383652e-02 -6.59601092e-01
1.24921584e+00 1.03377536e-01 4.26838458e-01 -7.83877373e-01
5.50477132e-02 4.97566164e-01 5.81052788e-02 -3.41256589e-01
7.76789546e-01 1.15606867e-01 -5.95181212e-02 2.01688483e-01
-5.45615137e-01 2.38453582e-01 -4.15872037e-02 7.82984719e-02
7.64372766e-01 -1.51709067e-02 5.36929965e-01 -1.33757666e-01
1.00784051e+00 -1.56042710e-01 5.21183729e-01 3.81541312e-01
-6.25496984e-01 6.23901367e-01 5.18410131e-02 -1.14309594e-01
-1.12373877e+00 -1.02952230e+00 -1.45926878e-01 4.54381406e-01
1.13217711e+00 2.57264227e-01 -4.56042588e-01 -4.99055773e-01
-1.84606791e-01 5.93890667e-01 -2.77786046e-01 -1.38219431e-01
-6.73746705e-01 -5.69309950e-01 4.07674253e-01 -2.39938900e-01
1.14419675e+00 -1.09343255e+00 -1.14457202e+00 2.32240930e-01
-1.45236298e-01 -9.79227304e-01 -6.49164438e-01 -2.48218015e-01
-7.35835254e-01 -9.33360755e-01 -1.28683197e+00 -1.08009028e+00
6.90224111e-01 1.08807552e+00 3.52770984e-01 2.47467786e-01
-3.07283968e-01 3.82572822e-02 -6.15568519e-01 2.62560346e-03
-3.83023143e-01 -3.12720031e-01 -3.69525820e-01 4.28386033e-01
1.70693174e-02 -3.59335661e-01 -9.66100037e-01 6.21465683e-01
-1.34589875e+00 1.00576416e-01 8.89545321e-01 5.75444579e-01
4.79966909e-01 4.20031011e-01 6.30225539e-01 -2.65172720e-01
2.58854121e-01 -2.75029927e-01 -7.19409049e-01 3.63482177e-01
-4.03407186e-01 -2.53120035e-01 3.52018714e-01 -6.99562907e-01
-9.22523439e-01 -2.73616225e-01 3.64262670e-01 -1.95513368e-01
9.76887420e-02 -8.75093490e-02 -3.58060688e-01 -5.09732187e-01
2.86681727e-02 7.86662757e-01 4.93727438e-02 3.27675790e-02
4.19543147e-01 1.07081330e+00 5.46118021e-01 1.78461105e-01
1.40337813e+00 6.37596965e-01 -9.99082346e-03 -1.02218497e+00
-3.02742515e-02 -3.68294567e-01 -3.26010771e-02 -3.13973963e-01
8.72301698e-01 -7.93236911e-01 -3.37652862e-01 8.23527932e-01
-9.66916323e-01 2.48154894e-01 -1.08489245e-01 2.51397938e-01
-3.58532131e-01 8.17570448e-01 -5.24206221e-01 -9.40065920e-01
-3.49667996e-01 -1.12405276e+00 8.44874799e-01 6.11534297e-01
2.39354461e-01 -4.76949066e-01 -3.26826274e-01 1.55301645e-01
6.61439657e-01 4.06515628e-01 9.26893890e-01 -3.29758137e-01
-1.08908868e+00 -7.68447667e-02 -7.76903152e-01 8.16410631e-02
4.51764733e-01 -4.53331172e-02 -5.29775441e-01 -4.21507090e-01
2.10421175e-01 3.56103122e-01 6.78071201e-01 1.98428929e-01
9.79765892e-01 -5.09118915e-01 -3.68357360e-01 3.64753932e-01
1.67471695e+00 6.59359574e-01 1.11109281e+00 8.54981661e-01
4.09276634e-01 4.91565019e-01 1.01730263e+00 2.97691584e-01
8.80647451e-02 6.09115839e-01 5.30038953e-01 -3.78038555e-01
-2.85587728e-01 -9.97150242e-02 5.55061281e-01 6.57613695e-01
3.18576276e-01 -3.92319888e-01 -3.73078048e-01 5.20250797e-01
-1.50970614e+00 -9.42879021e-01 2.18585283e-02 2.16637039e+00
7.57444978e-01 2.88347661e-01 -3.33006531e-02 6.11730039e-01
1.31236696e+00 7.33167648e-01 -4.68668520e-01 -3.13447475e-01
-2.23701119e-01 -1.69270635e-01 9.01250958e-01 2.36677945e-01
-1.02788711e+00 6.73055410e-01 5.11896801e+00 1.11065519e+00
-1.43159080e+00 -3.92490178e-02 5.62689185e-01 3.45473975e-01
-3.94119263e-01 9.87209082e-02 -6.37031734e-01 6.99891806e-01
4.88009632e-01 -2.49392986e-01 4.07097667e-01 4.32140619e-01
2.43124872e-01 -3.59682292e-01 -2.56881237e-01 8.88795376e-01
1.21497236e-01 -8.29261541e-01 8.67954940e-02 1.52422160e-01
7.00823247e-01 -7.73610473e-01 6.50552750e-01 -3.51420879e-01
-2.57012337e-01 -4.13916677e-01 9.19125795e-01 1.39502928e-01
8.39405417e-01 -8.56614888e-01 6.49749339e-01 -2.89386865e-02
-1.32489491e+00 -3.25519323e-01 -3.28477204e-01 4.58063513e-01
2.78589487e-01 6.11958802e-01 -4.13985938e-01 5.94400108e-01
5.01937628e-01 6.21108055e-01 -6.59680188e-01 1.09970534e+00
-1.24676861e-01 1.44516468e-01 -6.91167871e-03 9.83894542e-02
1.80410728e-01 -1.60683557e-01 1.05446720e+00 9.06250954e-01
4.33169633e-01 3.47808525e-02 -1.09601542e-01 6.24695480e-01
3.30840312e-02 1.02587812e-01 -6.33884370e-01 7.10012093e-02
4.90500212e-01 1.04775715e+00 -1.07087219e+00 -3.71226132e-01
-2.92488933e-01 1.16071856e+00 -6.62619889e-01 3.53826195e-01
-1.06874096e+00 -9.78000700e-01 2.19538122e-01 2.93476701e-01
6.54540479e-01 -3.67375553e-01 -1.89916447e-01 -1.04696238e+00
-1.33894846e-01 -1.03608429e+00 7.64856711e-02 -6.80039227e-01
-8.20434272e-01 2.78092355e-01 -5.21654002e-02 -1.83248782e+00
4.53360945e-01 -1.97627082e-01 -3.48258227e-01 6.76443696e-01
-2.26329160e+00 -1.07127357e+00 -2.54338741e-01 5.16318500e-01
6.57261491e-01 1.85112227e-02 2.15977281e-01 2.78676450e-01
-5.29463828e-01 3.96075785e-01 7.51239955e-02 7.99819753e-02
6.15671337e-01 -7.78389335e-01 2.48581871e-01 1.41004062e+00
-2.51268089e-01 5.74713707e-01 5.86689353e-01 -8.66797924e-01
-1.37373149e+00 -1.05900252e+00 6.02128208e-01 5.83576858e-01
6.37052536e-01 1.04607329e-01 -8.72108400e-01 2.30510637e-01
3.73661190e-01 -1.40107915e-01 4.85314369e-01 -1.21157813e+00
-2.80141860e-01 -2.06953943e-01 -1.66848743e+00 8.25781882e-01
6.58985555e-01 -2.43962809e-01 -6.08938336e-01 -4.15338010e-01
1.04245603e+00 -1.80099353e-01 -6.96090877e-01 4.53938901e-01
6.23593986e-01 -8.56862307e-01 1.19055903e+00 4.13309038e-01
2.24069178e-01 -1.01579404e+00 -2.39679128e-01 -7.26076424e-01
-1.39266491e-01 -8.47902596e-01 -1.37362450e-01 1.28345144e+00
4.69294861e-02 -8.44500065e-01 3.55991393e-01 9.61193293e-02
3.90991718e-01 -2.79153973e-01 -9.44829166e-01 -6.43297195e-01
-7.00450897e-01 3.16658050e-01 5.49119413e-01 8.79462183e-01
-1.42027751e-01 -2.65718728e-01 -5.77377319e-01 5.10917425e-01
1.08397174e+00 8.51172283e-02 2.86758184e-01 -7.75017917e-01
4.07844298e-02 -2.73616523e-01 -6.85567141e-01 -8.55317473e-01
-3.80990624e-01 -5.18691003e-01 4.75927554e-02 -1.12399518e+00
3.18998992e-02 -5.15749693e-01 -5.23350298e-01 1.79962695e-01
-2.27613345e-01 7.27011144e-01 6.39209390e-01 4.77109343e-01
-5.51668167e-01 2.56296992e-01 1.35518587e+00 -3.12307775e-01
-3.48602742e-01 -2.95177042e-01 -6.34857774e-01 5.55151761e-01
9.40194368e-01 -4.47241902e-01 -3.26917470e-01 -4.71506082e-02
-2.42689416e-01 2.39105329e-01 4.85917538e-01 -1.14676619e+00
5.70651069e-02 -2.06408694e-01 4.67728138e-01 -7.39021778e-01
7.21751899e-03 -1.21085930e+00 2.84557223e-01 9.14265454e-01
-2.34959245e-01 -1.00987796e-02 -9.17599071e-04 8.16696167e-01
-1.74845994e-01 -3.20314318e-01 1.38793063e+00 7.88026750e-02
-9.60390627e-01 7.94734713e-03 -4.78181362e-01 -3.43119234e-01
1.65551054e+00 -8.42365563e-01 -4.38616037e-01 -8.41225088e-02
-1.66859403e-01 -1.74507976e-01 7.43560612e-01 4.10407990e-01
1.06864417e+00 -1.07476163e+00 -5.33190310e-01 4.43973273e-01
4.25936608e-03 -4.32204813e-01 2.91430116e-01 5.95525503e-01
-7.46688783e-01 1.61752537e-01 -4.83569533e-01 -4.01834428e-01
-1.42740178e+00 9.13441956e-01 -1.02077365e-01 -2.58040816e-01
-6.19389594e-01 1.98913962e-01 -1.89120457e-01 3.76812428e-01
9.03883278e-02 -1.28838420e-01 -3.75540495e-01 7.98472483e-03
7.61506021e-01 4.97279972e-01 -4.23939496e-01 -9.11840796e-01
-3.99341136e-01 1.03886986e+00 -1.19085044e-01 -2.39398792e-01
9.39022303e-01 -6.69945955e-01 -3.12426180e-01 -1.83796138e-02
1.23484063e+00 2.47082040e-01 -1.04756284e+00 -1.50747910e-01
1.69650838e-02 -1.07565343e+00 3.82249467e-02 -5.52196920e-01
-1.18939185e+00 4.94475722e-01 1.03557384e+00 5.77670753e-01
1.52369022e+00 -4.76158351e-01 1.22314692e+00 -3.37027997e-01
6.04196906e-01 -6.20855391e-01 2.47210562e-01 -2.08552197e-01
5.41405976e-01 -1.02580881e+00 1.29005358e-01 -5.17799973e-01
-5.75579047e-01 1.12177980e+00 1.31234884e-01 -3.11685473e-01
4.53201979e-01 1.22386232e-01 -2.06664354e-02 2.94046402e-01
-1.13409042e-01 -2.05669284e-01 9.02601182e-02 9.12184358e-01
-1.73847347e-01 -2.06643865e-01 -6.34622335e-01 -4.04947177e-02
2.82352000e-01 1.58013287e-03 8.56263816e-01 1.10242808e+00
-9.21648920e-01 -9.93113399e-01 -7.77398169e-01 -6.93418682e-02
-6.82648957e-01 2.77821603e-03 -3.85129489e-02 6.40262127e-01
1.49963051e-01 1.11318564e+00 -2.73361653e-01 -3.65959585e-01
5.78234345e-02 -4.89764005e-01 1.60461202e-01 -2.17423290e-02
-1.58145621e-01 1.55505389e-01 -3.84415448e-01 -2.53028482e-01
-5.44760466e-01 -2.71666259e-01 -1.09635723e+00 -1.39675289e-01
-5.23161352e-01 -7.81245530e-03 6.87136471e-01 3.72375935e-01
1.02680221e-01 5.43188274e-01 1.08012724e+00 -7.97964811e-01
-3.55078280e-01 -3.85399461e-01 -7.19863296e-01 8.57845396e-02
7.65005648e-01 -4.19964224e-01 -7.04998016e-01 1.98395371e-01] | [4.363544464111328, 8.013874053955078] |
2fefd10b-4d40-41c7-8541-6a1a38b14076 | spatial-temporal-super-resolution-of | 2106.11485 | null | https://arxiv.org/abs/2106.11485v3 | https://arxiv.org/pdf/2106.11485v3.pdf | Spatial-Temporal Super-Resolution of Satellite Imagery via Conditional Pixel Synthesis | High-resolution satellite imagery has proven useful for a broad range of tasks, including measurement of global human population, local economic livelihoods, and biodiversity, among many others. Unfortunately, high-resolution imagery is both infrequently collected and expensive to purchase, making it hard to efficiently and effectively scale these downstream tasks over both time and space. We propose a new conditional pixel synthesis model that uses abundant, low-cost, low-resolution imagery to generate accurate high-resolution imagery at locations and times in which it is unavailable. We show that our model attains photo-realistic sample quality and outperforms competing baselines on a key downstream task -- object counting -- particularly in geographic locations where conditions on the ground are changing rapidly. | ['Stefano Ermon', 'David B. Lobell', 'Marshall Burke', 'Chenlin Meng', 'William Zhang', 'Nicholas Lai', 'Dingjie Wang', 'Yutong He'] | 2021-06-22 | null | http://proceedings.neurips.cc/paper/2021/hash/ead81fe8cfe9fda9e4c2093e17e4d024-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/ead81fe8cfe9fda9e4c2093e17e4d024-Paper.pdf | neurips-2021-12 | ['object-counting'] | ['computer-vision'] | [ 6.07489824e-01 -5.33896923e-01 -3.18158984e-01 -1.97926879e-01
-1.13390338e+00 -4.66452897e-01 7.27727413e-01 1.94701210e-01
-6.43605530e-01 1.04902363e+00 3.22673947e-01 -3.28723013e-01
1.92722633e-01 -1.31565535e+00 -6.90363526e-01 -7.23789930e-01
-1.73658729e-01 2.59572744e-01 -8.40234682e-02 -8.12862813e-02
-1.51999444e-01 5.33239245e-01 -1.66906333e+00 -9.24763680e-02
9.27371979e-01 4.62479144e-01 6.31116033e-01 6.71872377e-01
3.14914197e-01 2.56517440e-01 -2.39247769e-01 -1.69852991e-02
4.36510444e-01 -1.65245593e-01 -4.77028161e-01 2.85091549e-01
6.93772078e-01 -8.19110870e-01 -2.17056841e-01 1.07047343e+00
3.46904695e-01 1.92588359e-01 5.11827886e-01 -7.28388309e-01
-4.34037060e-01 3.07509661e-01 -1.17186332e+00 3.51821721e-01
-9.92889628e-02 6.03149295e-01 8.38353813e-01 -3.53639603e-01
3.84363413e-01 1.18536210e+00 9.45510030e-01 -1.24771506e-01
-1.54779601e+00 -8.60364795e-01 2.25073159e-01 -3.94077420e-01
-1.65141857e+00 -3.71099681e-01 -9.40890908e-02 -5.88945687e-01
9.37743247e-01 3.18275094e-01 5.11761665e-01 9.09800053e-01
-8.12341049e-02 4.42892909e-01 1.05299628e+00 4.22943793e-02
-3.69056202e-02 -5.58339596e-01 -4.11255240e-01 3.41674030e-01
5.06101310e-01 2.94047296e-01 -3.95448834e-01 -2.58378595e-01
1.05451655e+00 2.76199073e-01 -1.99838459e-01 4.96241570e-01
-1.49645305e+00 6.94498181e-01 6.77013516e-01 -1.85012370e-02
-8.64798963e-01 3.46004009e-01 -1.73858255e-01 -2.88938046e-01
9.36768591e-01 2.04585344e-01 -3.02399039e-01 -1.04511380e-01
-1.35521126e+00 5.41325510e-01 -4.84861769e-02 6.54282033e-01
8.64272892e-01 8.27263072e-02 -4.78392094e-02 7.78529882e-01
-1.51483312e-01 1.30712783e+00 -7.85432160e-02 -1.10484278e+00
6.25583708e-01 2.51124114e-01 6.72652185e-01 -1.28047729e+00
-3.65470946e-01 -3.39806318e-01 -1.29857337e+00 4.81632072e-03
2.24817976e-01 -2.16956869e-01 -9.90641534e-01 1.72303307e+00
3.24525863e-01 5.66126466e-01 -2.73056179e-01 1.05711651e+00
2.76255012e-01 1.06268060e+00 6.31403089e-01 -9.53670666e-02
1.30446613e+00 -3.89901787e-01 -2.96049416e-01 -8.60467613e-01
1.85325414e-01 -2.70977288e-01 1.13691556e+00 9.17861983e-02
-6.12382412e-01 -3.96855652e-01 -7.68917024e-01 1.02148913e-01
-2.14156106e-01 1.73570544e-01 9.22143638e-01 3.34264398e-01
-9.11502481e-01 6.04444504e-01 -9.75745916e-01 -4.50508684e-01
7.40769386e-01 -1.74650505e-01 -4.85000312e-01 -2.66347915e-01
-8.98361921e-01 6.55347764e-01 1.51188299e-01 2.59537250e-01
-8.02234828e-01 -8.00950110e-01 -9.47237372e-01 2.02329725e-01
8.72601420e-02 -4.65634763e-01 1.01931071e+00 -6.72426581e-01
-6.17836237e-01 8.74395490e-01 -1.80737302e-01 -4.44206148e-01
4.40881878e-01 -1.23573229e-01 -3.54573131e-01 -9.39906165e-02
8.36366475e-01 1.03274965e+00 6.13799572e-01 -8.86787355e-01
-1.13963282e+00 -5.93333006e-01 -2.68431246e-01 4.85160679e-01
-2.39080623e-01 -1.08271532e-01 -2.25153059e-01 -7.26783156e-01
-9.13656205e-02 -8.39162707e-01 -6.35562360e-01 1.78525671e-01
-2.66629364e-02 3.50762457e-01 6.08353913e-01 -9.99961376e-01
9.35629308e-01 -2.31045794e+00 -1.92914456e-01 -1.42420810e-02
-1.52237430e-01 2.42063463e-01 -4.09835875e-01 1.82553679e-01
3.09343457e-01 5.22123456e-01 -7.02066839e-01 1.44024268e-02
-4.49876934e-01 2.93210715e-01 -4.57891673e-01 4.45057303e-01
4.66268867e-01 7.72184491e-01 -1.18281829e+00 -3.85032773e-01
4.78275895e-01 5.33641875e-01 -6.42193928e-02 -9.98355299e-02
-4.04260069e-01 6.06031775e-01 -4.37946498e-01 7.20460296e-01
8.77321899e-01 -2.89867967e-01 1.09641284e-01 2.35773951e-01
-4.12467957e-01 1.98874295e-01 -1.12574017e+00 1.42182684e+00
-5.00523210e-01 6.07839406e-01 2.11193055e-01 -5.43295622e-01
5.49019516e-01 -1.07262902e-01 2.27091342e-01 -9.65335727e-01
-3.53707135e-01 -7.66957775e-02 -4.28093225e-01 -3.11060488e-01
8.32792282e-01 -2.91320920e-01 -2.11444467e-01 3.51916969e-01
-7.38089740e-01 -3.08623105e-01 2.01340392e-01 -4.47812164e-03
8.58060956e-01 1.99823111e-01 6.21074200e-01 -1.86438233e-01
-1.12540483e-01 6.63250685e-01 6.32189512e-01 7.44540036e-01
-1.28563598e-01 7.50986576e-01 4.85215299e-02 -5.21366835e-01
-1.20059693e+00 -1.12253332e+00 -2.82348126e-01 1.06169915e+00
-1.21757492e-01 1.80172950e-01 -3.09078485e-01 1.12820789e-02
2.51990199e-01 6.99446738e-01 -3.84377629e-01 2.88332999e-01
-1.88762352e-01 -1.40029764e+00 5.79022110e-01 4.73526806e-01
9.13196564e-01 -8.16648066e-01 -7.31711090e-01 2.50476032e-01
-7.32192099e-01 -1.11828208e+00 -1.88200131e-01 -3.52560669e-01
-1.04575241e+00 -8.27866852e-01 -6.65963829e-01 -1.04805060e-01
5.42605996e-01 8.70072663e-01 1.27046633e+00 1.55110331e-02
-5.84708691e-01 -5.01472950e-02 -2.11490124e-01 -2.72053868e-01
5.44768646e-02 1.44225881e-01 -1.23465709e-01 1.16291363e-02
2.83177227e-01 -5.33819556e-01 -6.08348131e-01 8.23948979e-02
-9.14024174e-01 2.40007341e-01 6.17128670e-01 6.34956837e-01
9.72122729e-01 4.22824025e-01 4.00428981e-01 -5.60752928e-01
3.89066935e-02 -7.68400729e-01 -8.81686509e-01 8.20201859e-02
-3.25160697e-02 -1.73215315e-01 5.77439554e-02 -1.48122385e-01
-1.15464175e+00 3.92948061e-01 1.03247404e-01 1.72510803e-01
-3.00023794e-01 7.00695932e-01 9.15013552e-02 2.68581510e-01
8.34754527e-01 2.40790859e-01 -3.44320834e-01 -4.66249645e-01
1.96506858e-01 9.56908584e-01 9.15790737e-01 -3.80414426e-01
8.54399800e-01 9.55861866e-01 7.57482424e-02 -1.40846014e+00
-9.58759785e-01 -6.12492859e-01 -3.85187685e-01 -1.10304132e-02
6.40600860e-01 -1.77052271e+00 -2.43644118e-01 7.00886846e-01
-7.24765480e-01 -7.62795866e-01 -7.54128546e-02 4.68158394e-01
-1.39439255e-01 2.05322564e-01 -1.10406250e-01 -9.74821448e-01
-3.57348025e-01 -7.07290828e-01 1.44747365e+00 1.61500171e-01
1.18262783e-01 -4.90579277e-01 4.63565392e-03 2.29523957e-01
4.24263626e-01 5.92555940e-01 4.56134617e-01 5.93687832e-01
-8.70583534e-01 -1.66795716e-01 -9.44207728e-01 1.62114978e-01
2.90603489e-01 1.15357906e-01 -7.72417128e-01 -2.20439702e-01
-6.40064538e-01 -3.05352390e-01 1.25206566e+00 8.56815994e-01
1.29417980e+00 -2.91526347e-01 -4.94889855e-01 6.05698049e-01
1.71775210e+00 -4.10457879e-01 8.78749490e-01 1.50164232e-01
3.98906976e-01 4.68637764e-01 9.36110616e-01 6.16805255e-01
5.24421871e-01 5.78711569e-01 5.35059035e-01 -3.13786626e-01
-1.44664329e-02 -2.68276185e-01 -6.30920306e-02 -1.74558774e-01
-3.86024922e-01 -1.04748763e-01 -1.08156896e+00 1.18220437e+00
-1.96214259e+00 -1.35607374e+00 -4.98072684e-01 2.62470818e+00
7.53778696e-01 -3.20319772e-01 8.72922316e-02 -4.72043842e-01
7.01222658e-01 4.72839266e-01 -5.74956417e-01 5.07655144e-01
-4.29613411e-01 1.54003808e-02 1.28676605e+00 4.65146005e-01
-1.44179666e+00 1.03819752e+00 6.89421654e+00 5.71865737e-01
-1.06613576e+00 2.66901284e-01 8.40416908e-01 -3.00639570e-01
-1.91876858e-01 -1.35178998e-01 -7.77703822e-01 3.96057039e-01
9.31388915e-01 1.54687360e-01 4.39268470e-01 7.03720272e-01
8.66697609e-01 -6.81799948e-01 -3.39506060e-01 9.39909637e-01
-3.21068794e-01 -1.54337037e+00 -1.71345174e-01 2.91345745e-01
9.33908641e-01 6.21166706e-01 7.57993758e-02 -9.85195264e-02
9.09706056e-01 -1.08627212e+00 4.22621787e-01 2.97602564e-01
1.28925443e+00 -6.57538235e-01 4.54490215e-01 4.94281590e-01
-1.35625744e+00 -4.32635136e-02 -3.84449452e-01 -4.02225852e-01
2.98898876e-01 7.66688585e-01 -4.18593496e-01 1.59854144e-01
9.52654421e-01 7.05340803e-01 -3.53968114e-01 1.00756717e+00
-2.82943994e-01 7.18062699e-01 -7.78185725e-01 4.85761195e-01
4.03404504e-01 -2.18528181e-01 2.75230795e-01 1.18573809e+00
6.46325827e-01 4.88199115e-01 3.13812822e-01 8.01783323e-01
-1.17705785e-01 -2.26147011e-01 -8.81547749e-01 -9.17240605e-02
6.66434348e-01 1.27496195e+00 -4.46708828e-01 -1.79435775e-01
-3.11447233e-01 7.55519509e-01 1.63282886e-01 3.77792984e-01
-6.71707869e-01 8.33385661e-02 1.10155201e+00 3.75482798e-01
1.36005938e-01 -5.89063525e-01 -3.61470997e-01 -1.16916764e+00
-2.21299648e-01 -7.17226565e-01 3.81537408e-01 -9.72650230e-01
-1.03567147e+00 -3.83569412e-02 1.49732322e-01 -1.10330403e+00
-2.54111379e-01 -2.00870246e-01 -3.56160104e-01 1.22664523e+00
-1.90629625e+00 -1.39178908e+00 -6.79366767e-01 3.41209948e-01
4.48080778e-01 2.74581105e-01 7.47892737e-01 2.84255058e-01
-4.82367814e-01 -2.61496514e-01 2.31465772e-01 -1.42720342e-02
3.26066852e-01 -8.31877530e-01 1.06228924e+00 1.27429092e+00
2.25358526e-03 1.72625586e-01 4.61056143e-01 -8.42887700e-01
-9.84881997e-01 -1.72985363e+00 9.82541502e-01 1.67503171e-02
4.48421210e-01 -3.80411595e-02 -6.29037917e-01 8.05764437e-01
-4.69868094e-01 -9.73997787e-02 4.07740206e-01 1.25920653e-01
-3.57933760e-01 -4.22132045e-01 -1.26742291e+00 7.17422724e-01
8.32887888e-01 -4.58089232e-01 4.16007526e-02 5.98397613e-01
4.90336567e-01 9.99120399e-02 -6.01333022e-01 4.10735279e-01
5.03661692e-01 -6.36032224e-01 1.15518129e+00 -2.28948399e-01
4.72814381e-01 -4.08537805e-01 -3.76596123e-01 -1.27684784e+00
-7.08174825e-01 -7.60813430e-02 5.55382550e-01 9.31529522e-01
3.18605304e-01 -5.18276036e-01 6.23614728e-01 5.85858107e-01
2.26407915e-01 3.16660106e-01 -7.80483186e-01 -7.05714583e-01
-8.32606032e-02 -4.41445112e-01 7.96762586e-01 8.11333776e-01
-6.57652497e-01 1.17175236e-01 -5.45196474e-01 8.46943498e-01
9.02369142e-01 2.39449978e-01 8.38802338e-01 -1.35799050e+00
1.79126784e-02 -2.35665724e-01 -1.93084925e-01 -8.17057908e-01
-8.87086987e-02 -1.74619794e-01 1.95004374e-01 -1.66393995e+00
4.42568928e-01 -7.43216991e-01 2.60047913e-01 8.59346747e-01
-4.59398419e-01 6.92082822e-01 -2.31608450e-01 2.78723031e-01
-1.17698871e-01 2.93882221e-01 8.68023634e-01 -2.53350735e-01
4.72631957e-03 -3.56726311e-02 -7.52445996e-01 5.45333326e-01
7.50021935e-01 -5.24473786e-01 5.01184687e-02 -8.54660749e-01
3.24741542e-01 -6.35729521e-05 7.71291137e-01 -1.07330418e+00
-4.45213169e-01 -1.01074910e+00 4.73447084e-01 -8.43110383e-01
3.63334447e-01 -7.43744910e-01 5.82434833e-01 4.25571054e-01
1.30889220e-02 -3.82802218e-01 2.56678134e-01 6.14968538e-01
9.51483548e-02 8.96934420e-03 9.29408133e-01 -4.57396686e-01
-1.02140951e+00 6.59852743e-01 -2.79439121e-01 -4.74786758e-02
8.92324746e-01 1.17337339e-01 -3.75871778e-01 -3.81905884e-01
-5.90358451e-02 4.08230156e-01 8.10304761e-01 3.18535477e-01
1.00387365e-01 -1.02452350e+00 -1.35473311e+00 2.16429457e-01
2.42957383e-01 4.12415564e-01 4.50896919e-01 4.56973583e-01
-7.63636589e-01 2.29404300e-01 -1.43030301e-01 -5.98945796e-01
-1.03453994e+00 8.52814615e-02 2.03312188e-01 -3.15701276e-01
-6.32114112e-01 5.82938671e-01 2.51019783e-02 -6.26228690e-01
-3.11223954e-01 -2.14476258e-01 2.41943255e-01 6.06163442e-02
9.87507880e-01 3.21201771e-01 -1.89777434e-01 -9.03126240e-01
2.49733757e-02 3.95173728e-01 5.87078512e-01 -2.89775580e-01
1.56771684e+00 -4.28067207e-01 -3.10760224e-05 2.72674352e-01
6.56892121e-01 -4.81340289e-01 -1.71717966e+00 -2.84268975e-01
-3.11493903e-01 -1.10536206e+00 6.02093637e-01 -6.04925036e-01
-9.03015554e-01 8.88995707e-01 4.37561959e-01 -4.83812802e-02
1.07737660e+00 -2.97019541e-01 6.88664794e-01 3.89598101e-01
6.65887415e-01 -1.03302908e+00 -5.30840814e-01 8.30669403e-02
7.18836427e-01 -1.50347769e+00 3.96847963e-01 -2.37148002e-01
-5.46226263e-01 4.30699140e-01 2.18120858e-01 1.03341855e-01
2.98882246e-01 2.76153415e-01 -2.66003460e-01 8.80259350e-02
-4.45920229e-01 -7.77287245e-01 -2.01213062e-01 1.07076120e+00
-5.75323543e-03 6.61390960e-01 1.15112640e-01 -1.66842848e-01
1.11674495e-01 2.41085589e-01 3.83536130e-01 6.34043515e-01
-7.39448607e-01 -5.91895163e-01 -7.13889718e-01 8.80090714e-01
-1.83769718e-01 -4.91534233e-01 9.70429033e-02 5.33684611e-01
2.99078673e-02 8.96010816e-01 5.22771716e-01 3.18442099e-02
-6.31468371e-02 -3.65073383e-01 1.37322739e-01 -7.21027911e-01
-1.38565928e-01 -9.39475670e-02 2.38887504e-01 -5.53651631e-01
-7.51358151e-01 -1.06510997e+00 -6.95784271e-01 -7.98332036e-01
-3.84684838e-02 -5.05828321e-01 7.85230815e-01 8.24783981e-01
4.65229630e-01 -8.59884322e-02 5.23907542e-01 -1.29700911e+00
-4.77954745e-01 -1.13893938e+00 -9.21993256e-01 2.34533265e-01
4.11909282e-01 -5.17477810e-01 -3.95275727e-02 1.38878822e-01] | [9.438685417175293, -1.3646252155303955] |
f2392ac3-04ec-4d15-b908-ae7140446670 | species-interactions-reproduce-abundance | 2305.19154 | null | https://arxiv.org/abs/2305.19154v4 | https://arxiv.org/pdf/2305.19154v4.pdf | Species interactions reproduce abundance correlation patterns in microbial communities | During the last decades macroecology has identified broad-scale patterns of abundances and diversity of microbial communities and put forward some potential explanations for them. However, these advances are not paralleled by a full understanding of the underlying dynamical processes. In particular, abundance fluctuations over metagenomic samples are found to be correlated, but reproducing these through appropriate models remains still an open task. The present paper tackles this problem and points to species interactions as a necessary mechanism to account for them. Specifically, we discuss several possibilities to include interactions in population models and recognize Lotka-Volterra constants as successful ansatz. We design a Bayesian inference algorithm to obtain sets of interaction constants able to reproduce the experimental correlation distributions much better than the state-of-the-art attempts. Importantly, the model still reproduces single-species, experimental, macroecological patterns previously detected in the literature, concerning the abundance fluctuations across both species and communities. Endorsed by the agreement with the observed phenomenology, our analysis provides insights on the properties of microbial interactions, and suggests their sparsity as a necessary feature to balance the emergence of different patterns. | ['José A. Cuesta', 'Miguel Ángel Muñoz', 'Matteo Sireci', 'Aniello Lampo', 'José Camacho-Mateu'] | 2023-05-30 | null | null | null | null | ['bayesian-inference'] | ['methodology'] | [ 1.46360472e-01 -3.68960321e-01 9.75951701e-02 1.64238378e-01
3.83096695e-01 -5.61256111e-01 1.03799725e+00 6.80424392e-01
-5.23571134e-01 8.91527653e-01 4.68343608e-02 -2.89608926e-01
-6.03824615e-01 -7.34202921e-01 -5.76770961e-01 -1.41941166e+00
-5.40505886e-01 7.47047126e-01 3.63804489e-01 -4.15598005e-01
3.16488117e-01 6.10931098e-01 -1.98554027e+00 7.05254599e-02
6.97976470e-01 4.63947356e-01 3.97397667e-01 5.77414095e-01
-3.23933363e-01 1.60301477e-01 -2.17970192e-01 -1.82672694e-01
6.42713308e-02 -3.45026135e-01 -3.86624217e-01 -2.36374661e-02
-1.56998023e-01 2.76796252e-01 1.61115751e-01 8.83046508e-01
1.52547121e-01 -1.71197534e-01 1.24434257e+00 -6.25781536e-01
-1.99284539e-01 6.58146560e-01 -5.64053774e-01 1.89444438e-01
-3.63170281e-02 3.30674648e-01 8.01203549e-01 -6.64614916e-01
5.90316594e-01 1.28038800e+00 4.93657261e-01 1.44348919e-01
-1.66262829e+00 -2.88171023e-01 -1.20499559e-01 1.10977232e-01
-1.48103607e+00 -4.15619224e-01 3.80532116e-01 -9.71718371e-01
8.75303388e-01 3.06893915e-01 1.14341855e+00 1.35207176e+00
4.67355698e-01 -5.18749580e-02 1.21207607e+00 -6.64701462e-01
3.71921629e-01 1.44651264e-01 3.43353935e-02 2.62740403e-01
1.14513612e+00 2.91296959e-01 -4.43957150e-01 -4.90219265e-01
6.99912608e-01 3.63361253e-03 -4.46627736e-01 -4.83081967e-01
-1.23425078e+00 6.05342448e-01 -5.60155399e-02 8.17954361e-01
-5.30409694e-01 -5.02391569e-02 4.52205271e-01 2.45582797e-02
3.59933347e-01 5.13986290e-01 -6.17064416e-01 -2.45649181e-02
-5.52424729e-01 2.18682930e-01 1.01225495e+00 1.55503988e-01
8.77850890e-01 -3.10029119e-01 4.46559638e-01 5.77787280e-01
4.98200148e-01 6.92860186e-01 3.07447523e-01 -5.21937430e-01
-2.12200731e-01 5.42181253e-01 2.81859666e-01 -8.79523635e-01
-3.59784484e-01 -5.24212480e-01 -1.06114638e+00 9.89408866e-02
8.89334500e-01 1.51595801e-01 -1.49021912e-02 1.55901182e+00
3.83086145e-01 -9.40036923e-02 2.39370409e-02 4.43431258e-01
-3.63699831e-02 4.98576462e-01 1.27783775e-01 -6.75151765e-01
1.36350691e+00 -1.69034421e-01 -5.19936860e-01 1.22015163e-01
1.58788353e-01 -6.18616402e-01 6.24308407e-01 3.87689084e-01
-6.77493572e-01 -1.07218750e-01 -6.10349536e-01 8.70075047e-01
-1.71269432e-01 -2.30930984e-01 7.07502484e-01 7.06137419e-01
-8.53021085e-01 1.04144681e+00 -1.10418975e+00 -1.05092430e+00
-3.98866892e-01 1.78822652e-01 -2.48848155e-01 4.21922684e-01
-8.23227108e-01 8.19780946e-01 4.25246447e-01 3.25740427e-01
-5.25457621e-01 -3.19731623e-01 -2.10315958e-01 1.94938350e-02
1.35612234e-01 -8.66021156e-01 7.44090736e-01 -6.54269636e-01
-1.52152073e+00 7.83426642e-01 -1.56564191e-01 -3.26411933e-01
5.05849719e-01 3.04101944e-01 -1.37337089e-01 5.72196878e-02
-3.95553976e-01 2.68421084e-01 3.17619801e-01 -1.48596597e+00
-8.60986412e-02 -2.99132794e-01 -3.59120548e-01 -3.13049227e-01
-3.39130193e-01 -1.99796036e-01 4.92248386e-01 -2.17387930e-01
2.49322787e-01 -7.54403889e-01 -3.94818395e-01 -3.10530901e-01
-1.58870779e-02 4.42324728e-02 -4.59755585e-02 -5.13962135e-02
9.02381539e-01 -1.98641419e+00 4.96991903e-01 2.97783464e-01
7.13152736e-02 -3.99084389e-02 -7.80825242e-02 1.28570724e+00
-8.03412870e-02 1.69468313e-01 -3.02455455e-01 -1.36044756e-01
-4.94695380e-02 4.09196436e-01 -1.83673739e-01 8.12154949e-01
1.84634104e-01 2.25076526e-01 -8.90558183e-01 -2.01444417e-01
4.36062485e-01 6.49104416e-01 -5.95639646e-01 1.20395891e-01
-2.18763500e-01 6.32674992e-01 -2.15480164e-01 2.03195021e-01
6.99960351e-01 -2.15935245e-01 7.03804195e-01 6.77573085e-02
-7.77502418e-01 1.61850587e-01 -1.12361908e+00 8.23958576e-01
-3.52834836e-02 2.70004421e-01 2.44988173e-01 -1.19340503e+00
9.84169543e-01 1.96058482e-01 6.15868568e-01 -1.66902959e-01
3.43440741e-01 7.20616639e-01 7.13987887e-01 -5.30452788e-01
2.32065290e-01 -5.64304054e-01 5.07322550e-01 1.75636321e-01
7.86430612e-02 1.41053215e-01 3.20802659e-01 -3.78044903e-01
7.84886718e-01 -2.72315666e-02 6.56097054e-01 -9.63020563e-01
7.68923342e-01 -6.03173524e-02 2.99779717e-02 9.10328329e-01
5.78191318e-02 3.25400591e-01 5.29207170e-01 -3.59232247e-01
-1.22471571e+00 -7.06176102e-01 -6.91353202e-01 6.55742526e-01
1.82711497e-01 -2.27209717e-01 -6.83371186e-01 3.90398920e-01
1.17678724e-01 1.85046554e-01 -7.96939671e-01 7.38548464e-04
-2.82416135e-01 -1.43952000e+00 3.83911550e-01 -3.12848657e-01
-2.46601528e-03 -9.29257989e-01 -6.37846708e-01 4.85397726e-01
8.18115100e-02 -8.60545576e-01 3.81556004e-01 6.63015962e-01
-1.05026507e+00 -1.46546519e+00 -4.99095380e-01 -1.32734537e-01
3.10812473e-01 1.49608865e-01 9.59312081e-01 2.32308209e-01
-4.14999872e-01 1.24383055e-01 -3.20688844e-01 -8.41419846e-02
-9.72839475e-01 -2.58856844e-02 5.96012592e-01 -1.55020542e-02
4.57789570e-01 -1.09008515e+00 -6.64227962e-01 3.65625888e-01
-9.47830200e-01 -3.70567828e-01 6.61585450e-01 7.85875976e-01
2.66240865e-01 -3.07744276e-02 3.70858043e-01 -3.83997917e-01
2.98740506e-01 -6.51152372e-01 -6.30032063e-01 6.82884827e-02
-7.43762076e-01 3.28110278e-01 6.77242458e-01 -5.00908196e-01
-8.21181118e-01 -1.43930331e-01 -1.48297593e-01 2.79241651e-01
-6.13286316e-01 4.51721877e-01 6.48001879e-02 8.61647055e-02
7.33456194e-01 5.28067648e-01 2.50972152e-01 -6.24040544e-01
-1.33375525e-01 5.99043608e-01 5.31857125e-02 -8.96398067e-01
5.26684880e-01 6.59259379e-01 3.12069714e-01 -1.39764631e+00
-3.06787461e-01 -6.04515254e-01 -5.84101319e-01 -1.52115449e-01
7.06332326e-01 -7.34959722e-01 -1.20772982e+00 4.68474269e-01
-1.23708010e+00 -2.92197406e-01 -2.14288428e-01 6.73096657e-01
-3.56372714e-01 5.78290880e-01 -6.24509633e-01 -1.53617156e+00
-1.55295759e-01 -1.10093141e+00 9.22329903e-01 -1.04708016e-01
-1.65157944e-01 -1.26789474e+00 7.21807778e-01 -2.87163913e-01
4.94331807e-01 2.78553486e-01 9.82005477e-01 -4.23955709e-01
-5.06581843e-01 1.80392250e-01 1.04753084e-01 -5.05612902e-02
2.59628981e-01 7.24820375e-01 -8.04676116e-01 -2.52584308e-01
2.22546726e-01 4.28593308e-01 9.73930418e-01 5.32394588e-01
3.88844997e-01 -1.45033956e-01 -4.36195880e-01 2.94210404e-01
1.52520072e+00 6.30333275e-02 2.85903007e-01 4.89446878e-01
6.14680946e-02 1.06088650e+00 -3.10128145e-02 7.19310939e-01
-7.44279921e-02 7.17338026e-01 6.22312546e-01 3.49138409e-01
3.67909551e-01 4.37323302e-02 4.32101339e-01 1.16580129e+00
-5.07738531e-01 -1.83578193e-01 -9.53581154e-01 3.32859129e-01
-1.95375466e+00 -1.31780601e+00 -8.94088745e-01 2.50927353e+00
5.01609027e-01 7.17056394e-02 3.11590701e-01 1.06280036e-01
6.66120768e-01 2.38687601e-02 -1.34967268e-01 -1.37542039e-01
-5.81556737e-01 9.68280882e-02 5.23598909e-01 5.88135302e-01
-6.57384038e-01 3.75500411e-01 7.20225716e+00 3.56046140e-01
-1.15875626e+00 -3.17488052e-02 1.90916210e-01 1.97580680e-01
-3.90476942e-01 1.90463364e-01 -7.35812485e-01 3.03740501e-01
1.03832150e+00 -2.14979604e-01 5.77565908e-01 2.89751321e-01
5.09392202e-01 -3.38839918e-01 -9.65585828e-01 6.18866324e-01
-3.10273409e-01 -1.08299041e+00 3.68967988e-02 6.90510213e-01
5.72809637e-01 1.34855002e-01 -4.18823808e-01 -1.05134837e-01
1.59168150e-02 -7.33829498e-01 6.04105353e-01 6.37585342e-01
1.50166690e-01 -4.32570606e-01 7.39531815e-01 6.33291960e-01
-1.09581041e+00 -4.60019223e-02 -5.89760780e-01 -7.40331888e-01
7.66037479e-02 1.01325977e+00 -6.52763247e-01 5.92285037e-01
4.46032166e-01 5.36088407e-01 -4.62927312e-01 1.07193518e+00
1.18547365e-01 6.01486742e-01 -6.40372872e-01 -4.35757577e-01
1.30299060e-02 -7.81685650e-01 6.20926797e-01 1.23494148e+00
6.89661920e-01 -2.61003107e-01 -3.15084159e-01 1.01117790e+00
7.43685663e-01 4.88687187e-01 -5.86070299e-01 -3.74362290e-01
3.27532470e-01 1.11536241e+00 -9.80787516e-01 -1.75796673e-01
-1.32111400e-01 5.02125502e-01 1.64468840e-01 1.12185754e-01
-3.34786594e-01 3.13362569e-01 9.10288155e-01 4.39099252e-01
3.30883801e-01 -5.08077204e-01 7.40708485e-02 -1.50627553e+00
-3.33656013e-01 -5.91645956e-01 -3.22934203e-02 -1.74671769e-01
-1.32153106e+00 5.40003061e-01 2.16794953e-01 -9.04786766e-01
-1.63715035e-01 -9.66104984e-01 -4.84359801e-01 7.91879594e-01
-1.22398353e+00 -3.50532919e-01 -1.33354172e-01 6.12509251e-02
5.06023914e-02 2.64001656e-02 1.02080178e+00 7.04169199e-02
-5.41644573e-01 -3.52793783e-01 7.45373428e-01 -7.34172165e-01
4.84945804e-01 -1.22549868e+00 1.57569647e-01 5.47056615e-01
-5.42207770e-02 9.65684116e-01 1.40099299e+00 -5.97765684e-01
-1.34793484e+00 -4.60030198e-01 7.49920964e-01 -2.61327595e-01
1.32987905e+00 -4.84390944e-01 -1.02533507e+00 -5.02384678e-02
4.15381253e-01 -3.84987593e-01 6.82716489e-01 2.37064138e-01
-1.90543413e-01 7.64254257e-02 -6.65506184e-01 5.18573999e-01
1.02929485e+00 -1.86950564e-01 -4.41296577e-01 2.95499325e-01
1.81724787e-01 3.22823465e-01 -8.11091065e-01 3.05229485e-01
7.87462711e-01 -1.33542538e+00 9.19222534e-01 -4.06125814e-01
3.16431075e-01 -2.77715087e-01 -2.92006969e-01 -1.13466775e+00
-4.15807486e-01 -3.27163756e-01 2.08428815e-01 1.09553170e+00
2.46287301e-01 -9.68939781e-01 3.45064074e-01 -9.13741216e-02
2.20640019e-01 -3.01547378e-01 -1.01859140e+00 -8.80198836e-01
3.05127144e-01 2.10072339e-01 4.06316519e-01 9.66422677e-01
1.47224694e-01 2.32452035e-01 -2.74456948e-01 2.28489578e-01
8.85198474e-01 -1.07981406e-01 7.07546234e-01 -1.71440709e+00
-7.44970024e-01 -8.67390156e-01 -3.60864878e-01 -5.08846462e-01
9.40938964e-02 -5.39108872e-01 2.10959651e-03 -1.05766106e+00
3.73839557e-01 -3.02360088e-01 -1.10125616e-01 -2.18351781e-01
-1.31898955e-01 -7.83139393e-02 -1.06602713e-01 4.48021859e-01
2.99439617e-02 5.59024334e-01 8.18845332e-01 1.36379614e-01
-1.72824740e-01 -1.77060023e-01 -4.33183461e-01 5.26741385e-01
6.44392610e-01 -4.84750569e-01 1.83299333e-02 1.41580090e-01
6.03214800e-01 -5.78201674e-02 3.73563379e-01 -1.11173463e+00
-6.02797195e-02 -4.11079794e-01 -1.06324684e-02 -2.04691574e-01
1.38241649e-01 -9.17286813e-01 8.65328074e-01 8.80502403e-01
2.19420604e-02 -1.09423630e-01 7.03446269e-02 5.96245766e-01
-7.49113485e-02 -5.70479870e-01 7.46258259e-01 -2.65365511e-01
-6.44540712e-02 -8.04287046e-02 -9.40150619e-01 -3.98470134e-01
6.44176543e-01 -2.22507995e-02 -4.21541780e-01 6.05874322e-02
-6.83980048e-01 -3.29432726e-01 9.24348950e-01 -5.45700230e-02
-1.64885819e-01 -7.95792460e-01 -7.09416330e-01 -7.67049417e-02
1.61079481e-01 -4.55858439e-01 3.22204113e-01 1.17961991e+00
-8.22698951e-01 3.76018167e-01 -1.47877619e-01 -9.19387460e-01
-1.09770870e+00 7.52689779e-01 6.88766658e-01 -2.89496452e-01
-1.72206774e-01 2.67429560e-01 3.07176441e-01 -3.87427837e-01
-3.09649646e-01 -2.91724682e-01 -1.68145254e-01 3.00957173e-01
4.21369821e-01 2.19051033e-01 2.20376290e-02 -4.72281337e-01
-4.62457478e-01 6.51813328e-01 3.44490588e-01 6.38864711e-02
1.54119849e+00 -2.75388837e-01 -5.42572439e-01 1.03584146e+00
7.05087185e-01 1.30358756e-01 -1.05917561e+00 1.68340385e-01
1.77951530e-01 -3.64155143e-01 -5.82896292e-01 -2.57680088e-01
-4.33492571e-01 8.49087477e-01 4.31237966e-01 8.19421768e-01
6.32994115e-01 4.04094532e-02 -2.05776095e-01 6.30292714e-01
6.21178746e-01 -3.62513781e-01 -1.55846089e-01 4.35782492e-01
8.36684287e-01 -1.05493796e+00 -1.13210576e-02 -5.23827314e-01
1.27987891e-01 1.36423123e+00 1.54337391e-01 -2.69110173e-01
5.99698365e-01 2.59688258e-01 -2.05264255e-01 -2.99867094e-02
-9.68896747e-01 -4.05134737e-01 -2.94523835e-01 4.58034337e-01
7.44608700e-01 2.14037508e-01 -9.31986094e-01 3.10312118e-02
-1.14261312e-02 -2.80283242e-01 7.68118024e-01 4.86729741e-01
-7.27741182e-01 -1.60073066e+00 -6.97760701e-01 5.99605069e-02
-2.82707334e-01 -4.89575863e-02 -4.39736426e-01 8.49030495e-01
2.75022127e-02 8.36118162e-01 1.42568991e-01 -1.52886525e-01
2.14473337e-01 2.17769533e-01 6.41914904e-01 -3.41299564e-01
-7.29890466e-01 4.25159276e-01 -7.84834102e-02 2.04709738e-01
-7.86019325e-01 -1.08205152e+00 -6.41437709e-01 -4.75586534e-01
-5.07432938e-01 6.21701181e-01 7.23686099e-01 8.78687859e-01
1.09988995e-01 3.54040354e-01 5.05005419e-01 -1.17603230e+00
-6.37730181e-01 -1.15519142e+00 -9.56928015e-01 5.07821083e-01
1.16215095e-01 -7.62896597e-01 -9.49519217e-01 -1.24037288e-01] | [5.837083339691162, 4.288794994354248] |
7a60906c-610c-4967-b4ad-6d377a0ca42e | detecting-can-masquerade-attacks-with-signal | 2201.02665 | null | https://arxiv.org/abs/2201.02665v2 | https://arxiv.org/pdf/2201.02665v2.pdf | Detecting CAN Masquerade Attacks with Signal Clustering Similarity | Vehicular Controller Area Networks (CANs) are susceptible to cyber attacks of different levels of sophistication. Fabrication attacks are the easiest to administer -- an adversary simply sends (extra) frames on a CAN -- but also the easiest to detect because they disrupt frame frequency. To overcome time-based detection methods, adversaries must administer masquerade attacks by sending frames in lieu of (and therefore at the expected time of) benign frames but with malicious payloads. Research efforts have proven that CAN attacks, and masquerade attacks in particular, can affect vehicle functionality. Examples include causing unintended acceleration, deactivation of vehicle's brakes, as well as steering the vehicle. We hypothesize that masquerade attacks modify the nuanced correlations of CAN signal time series and how they cluster together. Therefore, changes in cluster assignments should indicate anomalous behavior. We confirm this hypothesis by leveraging our previously developed capability for reverse engineering CAN signals (i.e., CAN-D [Controller Area Network Decoder]) and focus on advancing the state of the art for detecting masquerade attacks by analyzing time series extracted from raw CAN frames. Specifically, we demonstrate that masquerade attacks can be detected by computing time series clustering similarity using hierarchical clustering on the vehicle's CAN signals (time series) and comparing the clustering similarity across CAN captures with and without attacks. We test our approach in a previously collected CAN dataset with masquerade attacks (i.e., the ROAD dataset) and develop a forensic tool as a proof of concept to demonstrate the potential of the proposed approach for detecting CAN masquerade attacks. | ['Michael D. Iannacone', 'Robert A. Bridges', 'Pablo Moriano'] | 2022-01-07 | null | null | null | null | ['time-series-clustering'] | ['time-series'] | [ 2.55348057e-01 -5.47067672e-02 4.36916873e-02 6.53569922e-02
-6.47647560e-01 -1.15700626e+00 7.34622240e-01 4.07454930e-02
-1.31254405e-01 2.77095437e-01 -3.00304681e-01 -8.66497993e-01
1.43567100e-01 -8.14518690e-01 -9.46854770e-01 -5.62951088e-01
-5.01978099e-01 -2.27897111e-02 7.19469547e-01 -2.35368595e-01
4.72888619e-01 9.79506314e-01 -1.49448800e+00 2.03102797e-01
1.36363760e-01 7.52131164e-01 -3.88123125e-01 8.83765996e-01
4.12717372e-01 7.67674565e-01 -1.26678431e+00 -3.79537046e-01
4.94553387e-01 -1.56404495e-01 5.03305951e-03 -2.16666952e-01
2.70047843e-01 -6.63666070e-01 -7.61750400e-01 1.02734697e+00
-2.03489345e-02 -1.95896268e-01 3.93008500e-01 -2.09611988e+00
1.05252922e-01 3.98743242e-01 -5.10208428e-01 5.52712202e-01
2.90654689e-01 6.40148699e-01 4.22101408e-01 -5.48970327e-02
7.59574533e-01 1.38124895e+00 5.32124460e-01 4.53644484e-01
-1.18026435e+00 -1.37660348e+00 -3.22981536e-01 4.89994138e-01
-1.32867098e+00 -6.11530900e-01 9.89874303e-01 -3.29673529e-01
7.34018505e-01 4.39148456e-01 3.64930570e-01 1.32126915e+00
5.15417814e-01 4.48864847e-01 7.59525418e-01 -6.44483194e-02
4.10738200e-01 -7.88506046e-02 1.87036037e-01 1.52626604e-01
6.33685827e-01 5.15898705e-01 2.52670031e-02 -5.04269600e-01
1.07205130e-01 -2.28403043e-02 2.39941608e-02 2.05539525e-01
-9.53352392e-01 8.78615856e-01 -9.76952091e-02 1.49688885e-01
-2.57438481e-01 6.76718533e-01 8.63832235e-01 5.85321546e-01
-3.00662428e-01 4.49975491e-01 -2.14633018e-01 -4.06872749e-01
-6.96984649e-01 8.25998262e-02 5.97000122e-01 9.14934635e-01
5.44661582e-01 4.14576441e-01 2.86934137e-01 -2.00624883e-01
1.45345509e-01 8.08525562e-01 6.44574910e-02 -8.25667143e-01
3.03453535e-01 1.20941021e-01 -2.01325342e-01 -1.50357580e+00
-1.97098970e-01 1.12071738e-01 -2.34332487e-01 5.74631512e-01
4.17288363e-01 -3.90333980e-01 -6.78088546e-01 1.59629345e+00
1.46822602e-01 8.23217452e-01 1.17833160e-01 4.56843644e-01
3.21542829e-01 5.55853784e-01 5.52721731e-02 -3.29962730e-01
1.26573443e+00 4.23974060e-02 -8.56366456e-01 3.02130878e-01
7.02543795e-01 -6.62639201e-01 2.94284314e-01 3.26481611e-01
-5.29112101e-01 -3.07621598e-01 -1.39571667e+00 8.73373926e-01
-5.63175023e-01 -4.61358964e-01 4.07712497e-02 1.36185002e+00
-5.75470865e-01 3.37099105e-01 -1.07927036e+00 -1.86907984e-02
2.68539190e-01 3.22155833e-01 -2.29993194e-01 1.89421743e-01
-1.28919482e+00 7.80956745e-01 4.51387875e-02 -2.31133014e-01
-1.19084680e+00 -8.12044442e-01 -7.01641738e-01 -3.62637132e-01
4.52836186e-01 3.06998193e-01 9.05680835e-01 -7.07173288e-01
-1.05754578e+00 4.19356644e-01 1.54653803e-01 -9.66846168e-01
5.45728028e-01 1.46420151e-01 -1.31624579e+00 6.43426895e-01
-5.40611483e-02 2.44346648e-01 1.25843394e+00 -1.13608241e+00
-4.44178939e-01 -8.50910991e-02 4.18612920e-03 -1.05195868e+00
-1.49765611e-01 4.69098926e-01 1.35677218e-01 -4.79748189e-01
-4.29745317e-01 -1.27273333e+00 1.48643047e-01 -4.48488593e-01
-6.63380265e-01 2.67809808e-01 1.80744553e+00 -4.31891054e-01
1.46181738e+00 -2.44834733e+00 -9.26743567e-01 7.32463777e-01
3.82457882e-01 6.81403220e-01 -1.74812615e-01 7.06197083e-01
-2.43944645e-01 5.32405138e-01 3.65750283e-01 3.32226217e-01
-4.74914797e-02 1.07672103e-01 -8.55039120e-01 1.15358710e+00
1.66424930e-01 6.80531502e-01 -7.45256960e-01 1.46766186e-01
4.49579537e-01 2.34698102e-01 -4.16066557e-01 -1.94924787e-01
-5.46429008e-02 1.18880101e-01 -5.30125797e-01 5.27467430e-01
7.73041606e-01 5.74399471e-01 1.25547377e-02 -4.42698091e-01
-3.19751322e-01 1.49936348e-01 -1.01082087e+00 2.46448234e-01
7.78204501e-02 1.40336061e+00 -5.33534121e-03 -8.87222111e-01
8.46331835e-01 5.10835946e-01 4.80787128e-01 -7.15576291e-01
5.08522689e-01 2.29864977e-02 4.79198217e-01 -4.24503654e-01
2.66599715e-01 4.04326022e-01 -3.51282746e-01 5.84747612e-01
-4.19926375e-01 2.20738426e-01 8.76364857e-02 3.10310453e-01
2.01129103e+00 -6.94204509e-01 -6.50051087e-02 1.37303853e-02
4.56489623e-01 2.33415782e-01 4.41819251e-01 6.51032209e-01
-4.57242668e-01 -2.58067995e-02 9.44198966e-01 -1.62850827e-01
-1.25462818e+00 -1.15688372e+00 2.16925725e-01 2.85168201e-01
2.91099310e-01 -7.66292155e-01 -6.74355030e-01 -7.67929375e-01
3.19539726e-01 9.74818945e-01 -5.61240673e-01 -5.66174686e-01
-7.46924639e-01 -1.28830120e-01 1.61611533e+00 3.00523907e-01
3.11263025e-01 -4.71241057e-01 -8.20574462e-01 3.76608878e-01
2.72069544e-01 -1.70898914e+00 -2.56248653e-01 -2.45005786e-01
-3.59682143e-01 -1.53422523e+00 4.83153731e-01 -1.53740436e-01
4.06483740e-01 5.90471268e-01 5.71000636e-01 2.60224223e-01
-4.67307478e-01 5.82629025e-01 -3.70670676e-01 -7.43997753e-01
-9.83508646e-01 -5.80071151e-01 3.23543161e-01 3.93966258e-01
6.20470524e-01 -6.29520357e-01 -2.19458252e-01 7.65124619e-01
-9.36890185e-01 -7.13622868e-01 1.54393762e-01 1.51814008e-02
7.65062720e-02 6.79932714e-01 5.54191530e-01 -4.38378572e-01
4.46272939e-01 -8.06144238e-01 -9.57407415e-01 -3.86832148e-01
-2.60422111e-01 -4.18110311e-01 8.09336483e-01 -8.14863622e-01
-2.96940714e-01 -2.64988154e-01 1.70155048e-01 -8.38227332e-01
-5.57277203e-01 1.34153903e-01 -3.59051079e-01 -3.25798839e-01
5.85694849e-01 9.62141063e-03 3.79411459e-01 2.66022414e-01
2.00171545e-01 4.56853658e-01 6.14226580e-01 -2.58406878e-01
1.58718264e+00 9.14726079e-01 2.61960626e-01 -1.37649274e+00
3.69978726e-01 -3.11819613e-01 -2.36856658e-02 -9.63278174e-01
8.43818784e-01 -9.13925648e-01 -1.10883546e+00 5.22950709e-01
-1.09187126e+00 4.73530889e-02 3.73646498e-01 3.82342786e-01
-1.63985401e-01 6.24013007e-01 -5.66405535e-01 -8.28987420e-01
3.42800379e-01 -1.08322084e+00 7.73813725e-01 -1.28824711e-01
-3.96852940e-01 -6.56446695e-01 -1.18187867e-01 2.43668631e-01
3.55858147e-01 9.33957458e-01 7.83676684e-01 -1.00924778e+00
-7.25574613e-01 -6.87594891e-01 6.02006391e-02 7.72372782e-02
-3.17870006e-02 6.92595541e-01 -8.84058595e-01 -3.12141120e-01
9.23757553e-02 2.33922184e-01 7.58944005e-02 1.62957937e-01
5.03235519e-01 -2.88374692e-01 -5.48197567e-01 2.97331244e-01
1.23336446e+00 6.47400856e-01 1.21670163e+00 4.15389746e-01
4.85452205e-01 4.43316221e-01 6.63957059e-01 2.54640698e-01
-1.69044867e-01 9.76860225e-01 7.03547359e-01 4.52553570e-01
4.50219288e-02 -1.18056528e-01 1.11327553e+00 1.65458098e-01
9.75806937e-02 -2.24639252e-01 -7.30231702e-01 2.24512994e-01
-1.28584146e+00 -1.47364700e+00 -7.91225851e-01 1.96814573e+00
1.76852588e-02 5.99059761e-01 4.14635986e-01 4.99677211e-01
1.08446848e+00 1.25256166e-01 -3.38765115e-01 -5.38935542e-01
8.11018571e-02 5.30590303e-02 1.17582345e+00 2.66671389e-01
-1.05011106e+00 7.99181283e-01 5.30571318e+00 9.77090120e-01
-1.27798259e+00 -1.15208730e-01 1.20109364e-01 3.38613312e-03
-3.35681774e-02 1.39069170e-01 -6.49580002e-01 8.86926830e-01
1.64454138e+00 -2.93525755e-01 1.59889311e-01 6.71974480e-01
8.36508214e-01 1.69754714e-01 -1.01822913e+00 7.37353861e-01
2.23306511e-02 -1.44319677e+00 -2.47661516e-01 5.36518693e-01
1.69865116e-01 -1.45678923e-01 -5.17249405e-02 2.08791614e-01
2.85294920e-01 -8.63156080e-01 8.53012443e-01 5.45674972e-02
5.04545212e-01 -9.65970695e-01 6.21717453e-01 5.76799922e-02
-1.55177879e+00 -1.67209923e-01 1.93194211e-01 -5.59829213e-02
4.64978874e-01 2.67598510e-01 -7.89436460e-01 2.77023673e-01
2.73696214e-01 4.65951204e-01 -5.84085643e-01 7.51541615e-01
-5.54780774e-02 1.17706478e+00 -6.13431394e-01 1.25849485e-01
3.58198613e-01 2.40591858e-02 1.10720336e+00 1.04931915e+00
3.04956198e-01 4.79923002e-02 -6.48918748e-02 6.31063461e-01
4.04999137e-01 -6.38405561e-01 -1.01421857e+00 -2.54807264e-01
1.06777418e+00 8.82917106e-01 -9.42901075e-01 -6.31657019e-02
-3.97077620e-01 2.66894698e-01 -6.60802662e-01 4.19854760e-01
-1.40486932e+00 -6.36249661e-01 1.16418433e+00 3.46672088e-01
5.45649290e-01 -6.63238227e-01 -4.07091305e-02 -4.80708629e-01
-7.98644945e-02 -9.53115284e-01 9.82137695e-02 -3.98611426e-01
-9.13655818e-01 1.73758984e-01 7.07512125e-02 -1.86806774e+00
-3.69716585e-01 -3.96797538e-01 -9.94717777e-01 2.10462846e-02
-1.02251589e+00 -9.08596158e-01 -1.64209142e-01 9.79785681e-01
1.93622619e-01 -3.84793788e-01 1.90809906e-01 5.15331984e-01
-4.33881372e-01 7.83148170e-01 -2.46336922e-01 6.03050828e-01
5.14231205e-01 -5.18815577e-01 6.27038896e-01 1.15620899e+00
-2.59770781e-01 3.85709316e-01 9.65804160e-01 -9.74958479e-01
-2.08968830e+00 -1.26690710e+00 2.55145371e-01 -4.17976528e-01
1.34530938e+00 -4.05344754e-01 -6.18532240e-01 7.35705972e-01
-1.26377031e-01 -2.16957673e-01 5.92729330e-01 -8.24741960e-01
-5.42086959e-01 5.64336777e-02 -1.15412271e+00 7.60431290e-01
5.74433088e-01 -6.96601152e-01 -4.42102969e-01 -1.04531541e-01
5.79942703e-01 9.22002643e-02 -6.19861424e-01 2.35919535e-01
5.65078199e-01 -1.01571310e+00 9.59055781e-01 -4.71149951e-01
-4.87982499e-04 -8.48174870e-01 -3.80299926e-01 -6.79172873e-01
8.96858051e-02 -1.25213325e+00 -1.60363972e-01 1.42284334e+00
2.24484820e-02 -8.05473447e-01 5.98502100e-01 2.97089815e-01
-1.51076838e-01 1.82939336e-01 -1.08126283e+00 -1.19899428e+00
-2.51893133e-01 -1.14564896e+00 5.79355776e-01 1.02682829e+00
-3.44451070e-02 1.70755442e-02 -2.46753737e-01 8.64427149e-01
7.59302080e-01 -5.94414055e-01 1.26487756e+00 -9.04716134e-01
9.49514583e-02 -5.64851403e-01 -1.31988811e+00 -3.21485221e-01
1.84088528e-01 -4.19717580e-01 -1.46148458e-01 -2.38204911e-01
-6.46569133e-01 -1.50617927e-01 3.30415994e-01 6.03297055e-02
4.42791760e-01 4.44043636e-01 4.10873979e-01 1.86270311e-01
-3.12625289e-01 -8.35319608e-02 3.68387848e-01 -1.88177779e-01
-5.54933660e-02 7.47147277e-02 -7.32641527e-03 6.34266615e-01
7.81125188e-01 -7.19410241e-01 -5.03259778e-01 3.33346814e-01
1.43481240e-01 1.99165568e-01 6.85209334e-01 -1.29720533e+00
1.49195880e-01 5.29470779e-02 -1.80639952e-01 -6.60936832e-01
1.24843260e-02 -1.25070488e+00 6.61989212e-01 7.30659485e-01
2.13481784e-02 4.10272151e-01 4.67217684e-01 9.54982042e-01
-1.19305246e-01 1.97025716e-01 6.48024619e-01 5.88310957e-01
-7.43022203e-01 8.17844123e-02 -1.39384127e+00 -2.37422019e-01
1.77121043e+00 -4.22356904e-01 -7.25859225e-01 -6.14163041e-01
-3.79974216e-01 3.55508775e-02 5.81215739e-01 4.87569004e-01
6.34087503e-01 -1.17752528e+00 -5.01364172e-01 2.36645848e-01
-1.64377398e-03 -1.08390629e+00 1.30634308e-01 9.00813758e-01
-5.29443204e-01 1.91219464e-01 -2.76211619e-01 -6.79507971e-01
-1.67946720e+00 8.43971729e-01 4.26468924e-02 1.81947038e-01
-7.17093408e-01 -4.41561006e-02 -4.20861334e-01 2.02437252e-01
8.30560848e-02 -1.64815485e-01 -1.98174901e-02 1.05312303e-01
7.57014632e-01 6.99489892e-01 3.41434181e-02 -7.77477860e-01
-6.28696561e-01 3.26533824e-01 1.29276495e-02 -2.76272409e-02
7.58733809e-01 2.65862662e-02 8.25420842e-02 6.68316483e-02
1.42921209e+00 3.51646692e-01 -1.02201462e+00 4.25809294e-01
1.60854384e-02 -4.77960885e-01 -1.19104832e-01 -1.86026648e-01
-1.08798754e+00 3.91036630e-01 5.36936760e-01 8.54552567e-01
8.03056300e-01 -1.12919219e-01 1.08174396e+00 2.74597883e-01
5.70366919e-01 -7.67420948e-01 6.00724556e-02 2.61241972e-01
1.48326844e-01 -5.21175146e-01 -2.13792846e-01 -4.39893931e-01
-2.85933703e-01 1.24972999e+00 2.12469697e-01 -5.66461384e-01
8.75380933e-01 7.85387337e-01 1.37549430e-01 -4.04605806e-01
-7.20994771e-01 6.13653362e-02 -2.79262811e-01 8.61477613e-01
-4.45334285e-01 1.67098969e-01 -6.77203014e-02 2.54042327e-01
-3.59592348e-01 -4.46365267e-01 1.18643928e+00 1.00575435e+00
-3.07213098e-01 -8.54062617e-01 -9.26039517e-01 4.18614954e-01
-4.11282003e-01 3.12572718e-01 -5.06426573e-01 1.10518599e+00
-5.90494461e-02 1.60915124e+00 1.21216632e-01 -1.16136467e+00
2.56825298e-01 -2.01070577e-01 -1.54581860e-01 -5.85052669e-02
-3.99851501e-01 -1.95548892e-01 4.29947555e-01 -9.51696396e-01
-2.21207768e-01 -1.02012634e+00 -1.25398731e+00 -8.76729667e-01
-1.49496004e-01 3.57727483e-02 6.62905157e-01 1.13668418e+00
4.69870359e-01 5.87707877e-01 9.52434242e-01 -7.62155175e-01
-2.36777633e-01 -4.97145236e-01 -4.93010640e-01 6.00121140e-01
6.59493029e-01 -8.19166601e-01 -8.71131361e-01 5.62634654e-02] | [5.382391929626465, 7.509516716003418] |
ed3f6979-08c3-4ac1-88f8-3a000368f45d | cutpaste-self-supervised-learning-for-anomaly | 2104.04015 | null | https://arxiv.org/abs/2104.04015v1 | https://arxiv.org/pdf/2104.04015v1.pdf | CutPaste: Self-Supervised Learning for Anomaly Detection and Localization | We aim at constructing a high performance model for defect detection that detects unknown anomalous patterns of an image without anomalous data. To this end, we propose a two-stage framework for building anomaly detectors using normal training data only. We first learn self-supervised deep representations and then build a generative one-class classifier on learned representations. We learn representations by classifying normal data from the CutPaste, a simple data augmentation strategy that cuts an image patch and pastes at a random location of a large image. Our empirical study on MVTec anomaly detection dataset demonstrates the proposed algorithm is general to be able to detect various types of real-world defects. We bring the improvement upon previous arts by 3.1 AUCs when learning representations from scratch. By transfer learning on pretrained representations on ImageNet, we achieve a new state-of-theart 96.6 AUC. Lastly, we extend the framework to learn and extract representations from patches to allow localizing defective areas without annotations during training. | ['Tomas Pfister', 'Jinsung Yoon', 'Kihyuk Sohn', 'Chun-Liang Li'] | 2021-04-08 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Li_CutPaste_Self-Supervised_Learning_for_Anomaly_Detection_and_Localization_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Li_CutPaste_Self-Supervised_Learning_for_Anomaly_Detection_and_Localization_CVPR_2021_paper.pdf | cvpr-2021-1 | ['one-class-classifier'] | ['methodology'] | [ 6.01589680e-01 3.99006128e-01 2.26307929e-01 -3.06627870e-01
-8.45018089e-01 -1.01921052e-01 4.25236225e-01 2.74282336e-01
-1.30708605e-01 1.91830657e-02 -2.52947927e-01 -8.06749314e-02
3.96272063e-01 -9.52793837e-01 -1.17443728e+00 -6.28028214e-01
-3.17705572e-01 3.59962881e-01 4.77368683e-01 -1.33825734e-01
3.00315738e-01 5.91261446e-01 -1.84505367e+00 8.11019540e-01
8.85833263e-01 1.29708874e+00 -3.51535648e-01 9.32919979e-01
-3.31099741e-02 9.02419209e-01 -7.83805192e-01 -7.05375895e-02
4.57403213e-01 -3.87379557e-01 -7.04014599e-01 6.26039386e-01
6.97993457e-01 -5.24144948e-01 -4.08672482e-01 8.73879373e-01
2.08058298e-01 -2.03966439e-01 8.93535733e-01 -1.23716974e+00
-8.52848589e-01 1.08000197e-01 -7.74124920e-01 6.35755658e-01
1.81501746e-01 4.46328700e-01 9.14276421e-01 -8.27633262e-01
3.44957292e-01 8.81560802e-01 6.67326510e-01 7.03059852e-01
-1.19046307e+00 -3.54439586e-01 1.87440366e-01 1.62784159e-01
-1.00502992e+00 -2.78483182e-01 8.56646657e-01 -5.47531486e-01
1.20600069e+00 8.63574073e-02 4.39653933e-01 1.27248383e+00
3.86952698e-01 1.09824777e+00 5.33790529e-01 -4.52970266e-01
2.79296070e-01 -4.21040833e-01 1.03058562e-01 9.76741910e-01
3.74302000e-01 2.93115340e-02 -3.09464663e-01 -2.54760265e-01
6.35209799e-01 5.24990320e-01 -4.17674519e-02 -5.06784976e-01
-6.99408472e-01 8.97790670e-01 5.38104117e-01 2.19007596e-01
-5.94634473e-01 2.57331431e-01 5.86245358e-01 7.10664213e-01
6.47713602e-01 4.91675675e-01 -3.72338295e-01 1.54716566e-01
-9.24136698e-01 1.22880181e-02 3.92188936e-01 5.78985572e-01
8.02031219e-01 3.78260463e-01 -1.24112539e-01 8.95292342e-01
1.12749778e-01 3.40596199e-01 6.32782161e-01 -5.27987599e-01
1.10248722e-01 9.09607112e-01 -4.22342718e-01 -7.20148921e-01
-2.99907386e-01 -5.65892160e-01 -8.74616027e-01 7.72102237e-01
2.38470003e-01 6.63672611e-02 -1.77839947e+00 1.15376484e+00
2.60587912e-02 3.49117190e-01 5.86681142e-02 4.19188380e-01
3.44782233e-01 5.01572549e-01 -2.29013145e-01 2.03966036e-01
7.65144706e-01 -9.78886127e-01 -3.13394547e-01 -3.88282657e-01
1.00434864e+00 -4.47033763e-01 1.18295348e+00 8.52523029e-01
-8.78537655e-01 -4.59476322e-01 -1.33411407e+00 1.26840547e-01
-3.84430230e-01 2.56718904e-01 3.84601027e-01 4.84493434e-01
-1.13479531e+00 5.97804248e-01 -1.30308282e+00 -2.96498775e-01
1.04689372e+00 2.48195320e-01 -4.66158003e-01 -3.12098414e-01
-3.25377256e-01 3.51322740e-01 2.02627659e-01 -1.87030181e-01
-1.45078647e+00 -6.68768048e-01 -1.07151628e+00 -1.02703944e-01
3.04431438e-01 -2.33220935e-01 1.01312947e+00 -1.43118680e+00
-9.49829996e-01 1.25472748e+00 1.03230365e-01 -8.58920515e-01
3.23685914e-01 -5.39311886e-01 -7.43194878e-01 2.66092241e-01
3.25051725e-01 3.59485984e-01 1.12079144e+00 -1.29018021e+00
-6.47290945e-01 -4.55813915e-01 -3.57083380e-01 -4.86532241e-01
-2.16590121e-01 -1.23821430e-01 -3.23074579e-01 -8.50012779e-01
5.18653810e-01 -6.70528233e-01 -3.26637536e-01 1.72953665e-01
-7.12806284e-01 -4.90791947e-02 1.09325492e+00 -6.91078007e-01
8.72398973e-01 -2.39979386e+00 -3.55995268e-01 5.47751248e-01
3.41062874e-01 3.27389628e-01 -4.58765626e-01 1.88472837e-01
-3.71047288e-01 -5.54387607e-02 -7.45565295e-01 -3.45005214e-01
-4.84038234e-01 4.32667494e-01 -4.13461626e-01 6.22031987e-01
9.58357930e-01 8.45576286e-01 -6.26347959e-01 -2.49085901e-03
-4.02457714e-02 -3.02265733e-02 -9.00817096e-01 4.70831394e-01
-4.37596112e-01 2.12140545e-01 -2.68950850e-01 1.13125598e+00
6.88278556e-01 -1.39230758e-01 -3.62613529e-01 4.22738701e-01
3.66422027e-01 5.65006249e-02 -8.59274983e-01 1.97600138e+00
-1.81736112e-01 5.86561263e-01 -2.96424061e-01 -1.51286542e+00
1.08505714e+00 3.02663688e-02 5.52008688e-01 -6.86268389e-01
-1.24443591e-01 2.04571828e-01 1.51151508e-01 -6.20029211e-01
2.30635498e-02 3.30025971e-01 -4.24320670e-03 6.12975180e-01
5.85758328e-01 1.92239597e-01 -7.32138529e-02 1.94251746e-01
1.91985011e+00 -1.36793882e-01 2.81196982e-02 -6.26018643e-02
3.66324067e-01 -1.01373225e-01 4.46090698e-01 8.54909480e-01
-1.22459538e-01 1.10421312e+00 8.57941449e-01 -1.00473118e+00
-1.00301909e+00 -1.32247877e+00 7.79529335e-03 1.04919648e+00
-2.52518266e-01 -1.94766328e-01 -6.20246530e-01 -1.48962629e+00
6.06733896e-02 5.97647727e-01 -1.08018625e+00 -5.89544833e-01
-5.50095618e-01 -7.70182550e-01 5.37397802e-01 8.00192773e-01
4.15568739e-01 -1.27115166e+00 -4.93546516e-01 -5.51778711e-02
3.37830961e-01 -7.82329798e-01 -2.93905418e-02 3.17494661e-01
-9.40040648e-01 -1.39753151e+00 -3.30709636e-01 -8.70378971e-01
1.14906430e+00 -2.87273824e-02 1.19762230e+00 5.31698883e-01
-7.72157907e-01 5.30753791e-01 -4.96096730e-01 -5.25132239e-01
-5.28763294e-01 -1.05926044e-01 -1.00444190e-01 2.37420827e-01
5.70874751e-01 -6.82699621e-01 -5.71758389e-01 1.01310775e-01
-1.27821946e+00 -5.82782924e-01 8.89265060e-01 9.87075448e-01
8.13016951e-01 1.74021604e-03 4.34647948e-01 -1.18114579e+00
2.99317002e-01 -7.43474007e-01 -2.03747824e-01 -7.46961460e-02
-5.15708506e-01 2.47740582e-01 4.61886138e-01 -2.86513388e-01
-7.45859444e-01 3.47060472e-01 -5.60434997e-01 -6.17198110e-01
-7.59566903e-01 -8.11259747e-02 -1.79893412e-02 6.66128919e-02
1.18164754e+00 4.04644221e-01 1.86494276e-01 -4.36701149e-01
2.47414172e-01 4.44262207e-01 8.79661381e-01 -4.18965250e-01
8.83154452e-01 7.24382281e-01 -3.41895878e-01 -7.51270056e-01
-8.88441563e-01 -5.44737637e-01 -7.88683712e-01 1.40586391e-01
7.00761020e-01 -8.36223185e-01 -4.95636312e-04 7.04534173e-01
-7.23906875e-01 -3.72991830e-01 -8.62641215e-01 -7.80625343e-02
-5.22639275e-01 3.75400960e-01 -6.08452022e-01 -6.25125468e-01
-3.99066329e-01 -7.81256914e-01 1.14410341e+00 -4.96962182e-02
9.07360166e-02 -6.76406860e-01 3.32640886e-01 4.20408472e-02
2.37107128e-01 5.81954956e-01 6.38605356e-01 -1.31770301e+00
-5.36679029e-01 -5.63221097e-01 1.71801280e-02 9.83007967e-01
2.13596627e-01 7.41064921e-02 -1.13812637e+00 -3.71836126e-01
-5.86548029e-03 -5.64321101e-01 1.53404105e+00 1.43565342e-01
1.81717336e+00 -2.32701138e-01 -3.67540061e-01 7.32738912e-01
1.30370665e+00 7.33710155e-02 1.06815195e+00 3.77441257e-01
6.26258910e-01 1.54986277e-01 4.54435408e-01 4.04083103e-01
-1.11138888e-01 9.93597880e-02 8.70813549e-01 -4.56637293e-01
-8.60057920e-02 -2.93499172e-01 3.55198622e-01 1.69299528e-01
2.79962178e-03 -1.56197518e-01 -1.12958622e+00 9.12209213e-01
-1.64674413e+00 -7.49831080e-01 2.44967975e-02 1.92617786e+00
3.67822737e-01 4.49594021e-01 1.41864896e-01 4.42623168e-01
4.45493549e-01 3.15470509e-02 -7.82161176e-01 -4.17038053e-01
-5.83623052e-02 6.18287861e-01 2.59551883e-01 9.62843001e-03
-1.31782019e+00 9.11571860e-01 6.52537775e+00 4.50863779e-01
-1.10369217e+00 -6.35213479e-02 8.98552239e-01 -5.31825237e-02
-7.18472004e-02 -3.08006167e-01 -2.02677727e-01 4.00592625e-01
8.11099112e-01 5.43649316e-01 -8.30694735e-02 1.22389054e+00
-3.40492964e-01 -8.02186411e-03 -1.21294558e+00 7.59387314e-01
5.29688418e-01 -1.20747602e+00 2.50055164e-01 1.11951806e-01
9.66821253e-01 2.54499018e-01 7.42738396e-02 3.74688029e-01
3.09851646e-01 -1.09614205e+00 1.82721674e-01 5.26259482e-01
6.33925736e-01 -8.26066971e-01 9.50706601e-01 1.27250791e-01
-5.44249773e-01 -4.00641888e-01 -4.78548199e-01 -3.45181930e-03
-3.70068699e-01 7.70169795e-01 -1.03470898e+00 2.63245612e-01
8.92238140e-01 7.98333645e-01 -8.78813803e-01 1.13457179e+00
-3.31337482e-01 9.29630160e-01 -2.06217825e-01 6.81505680e-01
1.60633460e-01 2.43056536e-01 3.05156916e-01 1.13419700e+00
3.92685443e-01 -2.65935630e-01 2.28758812e-01 9.58460212e-01
-2.07488865e-01 -8.35721791e-02 -1.02010393e+00 1.26063928e-01
-2.18453273e-01 9.99705553e-01 -6.98412895e-01 -1.03525154e-01
-5.48701823e-01 1.56978905e+00 3.34541351e-01 2.22142160e-01
-6.04009032e-01 -4.20840591e-01 7.83746839e-01 4.10370409e-01
4.94588256e-01 4.81482036e-02 -2.93677390e-01 -1.13232660e+00
2.69046485e-01 -8.30495238e-01 6.92827463e-01 -3.97586763e-01
-1.47793186e+00 5.77459753e-01 -5.74769735e-01 -1.26382542e+00
-3.91968608e-01 -8.28988492e-01 -1.24582338e+00 2.35420153e-01
-1.29301965e+00 -1.24067175e+00 -3.99495453e-01 6.57580197e-01
7.28157163e-01 -5.21377861e-01 7.91189909e-01 3.22976522e-02
-6.09502375e-01 8.47440422e-01 -7.71275163e-02 6.53187752e-01
5.19191802e-01 -1.52389205e+00 9.14396644e-01 1.44931626e+00
4.85426635e-01 3.81516628e-02 3.04352015e-01 -6.50940716e-01
-7.62829185e-01 -1.29407072e+00 8.95901993e-02 -6.01005912e-01
4.48437333e-01 -5.07082582e-01 -1.46839714e+00 1.01274776e+00
2.25663688e-02 8.02298248e-01 5.55886328e-01 7.49724433e-02
-5.56875646e-01 7.64115481e-03 -1.38964999e+00 1.00910515e-01
1.13886034e+00 -3.83747429e-01 -6.13977432e-01 3.57772440e-01
5.93577147e-01 -3.15796852e-01 -4.71328616e-01 5.76357603e-01
-3.65870073e-02 -1.18398869e+00 8.07542503e-01 -9.05314863e-01
6.74771965e-01 -2.65133917e-01 1.45109743e-02 -1.34743023e+00
-1.40394628e-01 -2.92643845e-01 -4.94595826e-01 9.30246890e-01
6.09925210e-01 -6.61928952e-01 1.03293371e+00 1.98111370e-01
-6.90445840e-01 -9.84524190e-01 -7.21221030e-01 -6.29433095e-01
7.52380490e-02 -6.68195009e-01 3.78017873e-01 7.22365260e-01
-3.58178914e-01 -2.26126403e-01 -1.35265350e-01 6.80749953e-01
4.80873346e-01 -2.15235204e-01 7.72086501e-01 -1.33800566e+00
-2.18481407e-01 -5.00847772e-02 -1.16801989e+00 -8.60610366e-01
5.09062037e-02 -6.88242853e-01 1.77690059e-01 -1.22484553e+00
-5.13755605e-02 -3.11916798e-01 -6.81220174e-01 7.70448327e-01
-9.49546397e-02 4.25554812e-01 -5.89768112e-01 1.04534090e-01
-6.98829770e-01 4.59084779e-01 7.74093986e-01 -3.82013202e-01
-1.08769342e-01 1.73068210e-01 -6.75625741e-01 7.36510098e-01
9.59016323e-01 -5.50802588e-01 -3.35223258e-01 -3.12444717e-01
-1.14880972e-01 -6.58163011e-01 4.11072463e-01 -1.48170412e+00
6.21993002e-03 3.14077795e-01 7.97503769e-01 -5.28675020e-01
-6.75212219e-02 -5.61116755e-01 -7.28361070e-01 5.19490480e-01
-2.12296993e-01 1.56221148e-02 3.42673540e-01 8.79728675e-01
-3.06740195e-01 -1.03536695e-01 8.95799160e-01 -1.57891408e-01
-8.23027015e-01 4.94495720e-01 -3.15817893e-01 1.87132567e-01
1.32894945e+00 -2.32708052e-01 -3.04873198e-01 -9.95957777e-02
-9.71637368e-01 7.36559853e-02 7.42336333e-01 5.51676095e-01
1.01905680e+00 -1.15596235e+00 -7.21553266e-01 1.03805816e+00
5.26082456e-01 2.36455098e-01 3.17336768e-01 3.79696786e-01
-8.45449448e-01 -1.94400609e-01 -4.94999349e-01 -1.03575063e+00
-9.45065141e-01 6.78489029e-01 5.04709601e-01 -2.07885012e-01
-1.01862431e+00 9.88790333e-01 2.27282196e-01 -3.35902661e-01
1.03933103e-01 -5.08464575e-01 3.40109356e-02 -4.97097850e-01
6.32757545e-01 1.10838212e-01 5.29554367e-01 -3.03554714e-01
-1.76699504e-01 2.25284591e-01 -5.23976207e-01 3.18541318e-01
1.48862505e+00 4.49801952e-01 4.59329598e-02 2.90173978e-01
1.14016879e+00 -1.27549976e-01 -1.40494561e+00 -1.57314807e-01
1.91048786e-01 -5.34166336e-01 3.07742059e-02 -6.18620932e-01
-1.53809464e+00 7.86925673e-01 1.02191818e+00 2.94799507e-01
1.42474699e+00 3.37060511e-01 7.62164950e-01 5.18735468e-01
-1.04897194e-01 -9.91997898e-01 7.93636978e-01 2.12579280e-01
9.00969863e-01 -1.48410022e+00 -3.39569360e-01 -1.55756876e-01
-7.02326179e-01 1.07924032e+00 1.02325213e+00 -7.44430482e-01
7.38354325e-01 3.22250277e-01 1.56791627e-01 -5.98157704e-01
-7.00216174e-01 -2.60469496e-01 3.74322623e-01 9.79225576e-01
2.39582151e-01 -1.78601742e-01 3.78908426e-01 4.64198560e-01
1.84639990e-01 -3.83954912e-01 6.76052332e-01 1.15011883e+00
-6.86202526e-01 -9.87973332e-01 -1.02530889e-01 1.11014104e+00
-6.56129301e-01 1.37237102e-01 -4.67924953e-01 5.86272359e-01
3.52848381e-01 6.41774833e-01 6.00871742e-01 -4.68427032e-01
4.90557790e-01 1.91514403e-01 1.43367678e-01 -9.10573840e-01
-1.74729869e-01 -2.96331018e-01 -2.49736860e-01 -1.05044556e+00
1.42016917e-01 -6.44702315e-01 -1.20756352e+00 2.62218714e-01
-1.63320035e-01 -2.59334117e-01 3.23067516e-01 8.69990826e-01
6.17083311e-01 7.30463862e-01 7.68788099e-01 -6.03244007e-01
-3.31381947e-01 -9.96214271e-01 -6.70762181e-01 7.57980049e-01
7.19783664e-01 -4.79660064e-01 -5.14075398e-01 3.40278372e-02] | [7.678961753845215, 2.0502963066101074] |
f923fe93-6bf3-481e-9e65-6bcd8a630a63 | chat-or-learn-a-data-driven-robust-question | null | null | https://aclanthology.org/2020.lrec-1.672 | https://aclanthology.org/2020.lrec-1.672.pdf | Chat or Learn: a Data-Driven Robust Question-Answering System | We present a voice-based conversational agent which combines the robustness of chatbots and the utility of question answering (QA) systems. Indeed, while data-driven chatbots are typically user-friendly but not goal-oriented, QA systems tend to perform poorly at chitchat. The proposed chatbot relies on a controller which performs dialogue act classification and feeds user input either to a sequence-to-sequence chatbot or to a QA system. The resulting chatbot is a spoken QA application for the Google Home smart speaker. The system is endowed with general-domain knowledge from Wikipedia articles and uses coreference resolution to detect relatedness between questions. We present our choices of data sets for training and testing the components, and present the experimental results that helped us optimize the parameters of the chatbot. In particular, we discuss the appropriateness of using the SQuAD dataset for evaluating end-to-end QA, in the light of our system{'}s behavior. | ['Andrei Popescu-Belis', 'Gabriel Luthier'] | 2020-05-01 | null | null | null | lrec-2020-5 | ['dialogue-act-classification'] | ['natural-language-processing'] | [-1.6022468e-01 6.3929409e-01 5.4199457e-01 -6.0152799e-01
-1.0963888e+00 -8.0407947e-01 6.2790012e-01 -2.4002707e-01
-3.8229439e-01 7.2484857e-01 4.7103471e-01 -2.9384381e-01
-1.6117336e-01 -4.5902279e-01 3.9576672e-02 -4.2215574e-01
2.0447625e-01 1.0814419e+00 4.6426135e-01 -9.7343910e-01
3.2285899e-01 -3.0597324e-02 -1.2638733e+00 5.7796651e-01
9.2895269e-01 6.7909169e-01 4.3515944e-01 1.2169900e+00
-3.5852507e-01 1.6198994e+00 -9.2746609e-01 -6.2827009e-01
-2.1017328e-01 -7.9863554e-01 -1.7607449e+00 -1.6623308e-01
4.5823582e-02 -2.5192851e-01 -1.8076770e-02 5.7761830e-01
6.4342803e-01 2.9190865e-01 4.0840474e-01 -1.3907105e+00
-2.5620523e-01 8.2271969e-01 6.9193608e-01 6.2822148e-02
9.1096318e-01 4.5200843e-01 1.2903709e+00 -3.4319434e-01
8.5670722e-01 1.3333992e+00 4.8008013e-01 1.0183365e+00
-7.4277157e-01 1.2802306e-01 -3.9569703e-01 3.8268089e-01
-8.7187296e-01 -8.6538279e-01 6.2304163e-01 -2.6151249e-01
1.0675913e+00 4.9650016e-01 1.2585606e-01 1.0767914e+00
-1.9466951e-01 8.7955332e-01 6.0399753e-01 -4.9057510e-01
3.9589083e-01 3.7373710e-01 4.2136142e-01 6.7992395e-01
-7.3930955e-01 -5.2417880e-01 -5.2734238e-01 -4.4298798e-01
2.8097156e-01 -7.0769936e-01 -3.5811719e-01 -2.8818473e-01
-9.2677468e-01 9.7606820e-01 1.2848289e-01 5.2047729e-01
-4.2098272e-01 -1.4165407e-01 6.2411314e-01 8.3431756e-01
1.4477674e-02 8.8142127e-01 -5.1653516e-01 -8.6511105e-01
-1.3851571e-01 4.1191304e-01 1.7151380e+00 8.8885170e-01
3.6147568e-01 -3.9505935e-01 -3.9857653e-01 1.2122885e+00
2.8500071e-01 1.8587738e-01 4.3092480e-01 -1.5124568e+00
6.6237444e-01 6.7612731e-01 2.7814040e-01 -4.9017176e-01
-5.7279885e-01 4.3045145e-01 -4.7353745e-02 -7.4382029e-02
9.2629910e-01 -4.7255903e-01 1.4081147e-02 1.5066674e+00
3.1298068e-01 -5.6934506e-01 5.0205398e-01 9.3720227e-01
1.0558141e+00 4.4077435e-01 -9.3910331e-03 -1.5949127e-01
1.6244234e+00 -1.1464027e+00 -8.4299707e-01 -1.0035437e-02
9.8698312e-01 -7.0231634e-01 1.4699774e+00 2.7091363e-01
-1.2815362e+00 -1.8598574e-01 -4.6860465e-01 -2.4362506e-01
4.0122770e-02 -1.4347418e-01 1.2356176e-01 6.1353004e-01
-1.0692103e+00 2.3824175e-01 -4.8025411e-01 -7.7505189e-01
-2.6283023e-01 1.3394260e-01 -1.4939994e-01 1.7666684e-01
-1.3593831e+00 1.4176060e+00 -5.3566646e-02 -1.3675471e-01
-7.6720089e-01 -1.4958784e-01 -5.9147120e-01 2.7721992e-01
4.8346543e-01 -6.2579924e-01 2.2876050e+00 -8.9624858e-01
-2.3000093e+00 6.0448968e-01 3.9887182e-02 -4.7149566e-01
4.6867594e-01 -1.5407580e-01 -7.6109655e-02 5.7382464e-01
-9.1121793e-02 3.4374362e-01 2.5650692e-01 -8.2766598e-01
-6.6364783e-01 -1.8158774e-01 6.0308355e-01 4.9163973e-01
1.7690490e-01 3.2600677e-01 -1.8055350e-01 2.0788717e-01
-3.3248737e-01 -7.0559645e-01 8.6418202e-04 -5.1067817e-01
-3.3850017e-01 -7.4754065e-01 6.9631785e-01 -6.2663108e-01
1.0455492e+00 -1.8849337e+00 1.3289163e-01 -2.1834010e-01
-1.8599778e-01 2.7907300e-01 -2.8633505e-01 1.0778631e+00
2.2806522e-01 -2.5136617e-01 -1.8057796e-01 -2.7419913e-01
2.0037340e-01 1.9785979e-01 3.4871546e-03 2.1960452e-02
1.1325290e-01 7.0668393e-01 -9.8458970e-01 -6.2767059e-01
2.2094049e-01 7.2843168e-04 -6.4186800e-01 1.0461540e+00
-7.1713358e-01 3.9770496e-01 -5.6814116e-01 2.4814743e-01
-4.4623766e-02 -4.0678479e-02 4.5512730e-01 2.8935608e-01
-1.9963525e-01 1.1406914e+00 -6.7633080e-01 1.7883421e+00
-7.0643580e-01 6.2611240e-01 5.1802433e-01 -7.5540733e-01
8.9188164e-01 8.6979669e-01 7.1264051e-02 -5.9229314e-01
2.6811731e-01 6.4420968e-02 2.9988194e-01 -1.0839252e+00
6.1049998e-01 8.7026700e-02 -1.1022293e-01 8.4098220e-01
3.6915949e-01 -5.6072503e-01 1.2427502e-01 3.7219754e-01
1.3589872e+00 4.2279467e-02 2.8668708e-01 -1.3611427e-01
8.6487436e-01 2.8685755e-01 -1.0324097e-01 6.0908902e-01
-4.9060732e-01 3.7120351e-01 8.8511598e-01 -1.2848447e-01
-7.0790380e-01 -6.2384111e-01 3.7778434e-01 1.6035414e+00
-1.5367542e-01 -4.3417180e-01 -1.3544114e+00 -7.9801726e-01
-6.7209226e-01 1.1820996e+00 -1.9721811e-01 -3.8365595e-02
-5.7779312e-01 -7.3716849e-02 9.8259550e-01 -1.4129770e-01
4.1822883e-01 -1.6161381e+00 -7.6790816e-01 4.3671057e-01
-8.7224275e-01 -9.9278575e-01 -3.8611692e-01 4.5670737e-02
-5.9405726e-01 -1.4702339e+00 -2.5062183e-01 -8.8683879e-01
-9.5646217e-02 6.2648498e-02 1.4108056e+00 2.2394072e-01
2.2555278e-01 8.8173664e-01 -7.8949767e-01 -2.1337147e-01
-1.1687028e+00 5.0785828e-01 -3.5597748e-01 -2.0391883e-01
3.4221855e-01 -5.7797414e-01 -3.3673778e-01 6.2238038e-01
-4.4905764e-01 -2.0276576e-01 4.9409427e-02 9.1443032e-01
-6.5550423e-01 -6.1410648e-01 7.0961440e-01 -7.9812640e-01
1.2469354e+00 -3.4495050e-01 -2.8571650e-01 4.3554780e-01
-6.1967254e-02 1.3920373e-01 5.9176296e-01 -3.4729600e-02
-1.3192158e+00 8.4717415e-02 -5.9760171e-01 3.0931023e-01
-3.5413396e-01 2.4988341e-01 -4.6259537e-01 2.3249274e-01
1.1973250e+00 9.9687718e-02 4.0742314e-01 -5.7526147e-01
6.1908114e-01 1.3410031e+00 5.8002406e-01 -6.8769008e-01
1.3076538e-01 -1.8512447e-01 -8.6982268e-01 -1.1377831e+00
-4.1226792e-01 -6.2698936e-01 -3.7297073e-01 -4.1252550e-01
8.0964589e-01 -4.0373176e-01 -1.6242790e+00 2.1039744e-01
-1.5483243e+00 -6.5872610e-01 -1.0992203e-01 2.6795264e-02
-1.0945916e+00 4.6570623e-01 -9.1076601e-01 -1.1718531e+00
-5.6015861e-01 -1.0629505e+00 6.2744564e-01 1.5162793e-01
-6.2192291e-01 -8.4411794e-01 3.3438119e-01 1.0200874e+00
4.7178543e-01 -4.6754587e-01 7.6154906e-01 -1.3397291e+00
-2.6843441e-01 2.6977500e-02 1.5218413e-01 3.1990999e-01
3.1467792e-02 -3.5843533e-02 -1.1124189e+00 3.5046941e-01
2.4901418e-01 -8.0777520e-01 -6.3690387e-02 -8.2434051e-02
2.9706314e-01 -7.3332351e-01 1.6649923e-01 -2.8709748e-01
7.0600688e-01 3.1062332e-01 6.7908829e-01 3.3252022e-01
1.7971244e-01 1.3780146e+00 7.2390896e-01 2.9751560e-01
8.2261837e-01 9.9105239e-01 5.1429069e-01 6.0378009e-01
1.4593753e-01 2.4397381e-02 5.4451758e-01 8.3448875e-01
-3.2160595e-02 -3.0725658e-01 -1.0754998e+00 5.5642188e-01
-2.0623541e+00 -1.1864148e+00 -2.5743490e-01 1.7071936e+00
1.0747113e+00 -2.4969533e-01 5.7675242e-01 -7.3345423e-02
5.1189727e-01 -1.0481162e-01 -2.2090323e-02 -7.6566643e-01
2.5905690e-01 -1.8806644e-02 -2.7367702e-01 9.1368252e-01
-7.2919744e-01 1.0286707e+00 6.1316953e+00 2.7255881e-01
-7.3012060e-01 3.0596584e-01 1.4684772e-01 2.3581898e-01
1.4217508e-01 -2.2544007e-01 -2.5087354e-01 2.3529060e-01
1.3766366e+00 -1.8327993e-01 5.3711069e-01 1.1107597e+00
3.3640480e-01 -2.6200575e-01 -1.3181589e+00 5.1169789e-01
-5.7778586e-02 -1.1736213e+00 -4.3855411e-01 -5.3678513e-01
-2.7092364e-01 1.9150534e-01 -5.2790362e-01 5.2453893e-01
6.8657929e-01 -8.9271986e-01 4.2849928e-01 3.0514315e-01
1.2754038e-01 -5.4192871e-01 8.8597935e-01 7.9581654e-01
-4.9740043e-01 -4.7796220e-02 -9.4497874e-02 -2.5467253e-01
3.4473097e-01 -3.0763790e-01 -1.6242974e+00 3.7852353e-01
5.2377510e-01 -1.4471269e-01 -2.7206397e-01 8.0148029e-01
-3.8828331e-01 7.5284803e-01 -2.6735827e-01 -7.4230009e-01
1.7644480e-01 -2.5062466e-01 6.3945526e-01 1.2660838e+00
-1.4689450e-01 4.8957965e-01 -7.8640215e-02 5.8055019e-01
1.3123520e-01 1.7739318e-01 -3.6106175e-01 4.7549984e-04
6.2697291e-01 1.2804904e+00 -1.0346488e-01 -2.3416547e-01
-2.7429938e-01 9.3668687e-01 4.2994648e-01 7.4971244e-02
-3.6353609e-01 -6.2426478e-01 8.2808614e-01 -1.8729618e-01
9.4149068e-02 4.3198564e-03 9.1866061e-02 -9.4996154e-01
-1.0492172e-01 -1.4073368e+00 5.5115300e-01 -1.0200061e+00
-1.1800008e+00 9.6961927e-01 -2.7225468e-01 -7.4387509e-01
-9.2762733e-01 -5.1248443e-01 -7.9514450e-01 7.4280840e-01
-8.7450206e-01 -7.5475067e-01 -2.3525049e-01 7.3564142e-01
7.2432196e-01 -1.6727443e-01 1.2708833e+00 5.4437080e-03
-1.6770329e-01 3.1863317e-01 -3.0057085e-01 4.7080350e-01
7.6157159e-01 -1.3329666e+00 4.5606723e-01 1.2825723e-01
3.4177791e-02 5.9263021e-01 1.1829597e+00 -6.6256523e-02
-1.3291190e+00 -3.6902902e-01 1.2326145e+00 -8.5279745e-01
7.3744905e-01 -3.3248025e-01 -9.6054029e-01 5.7918900e-01
7.9242456e-01 -7.3744541e-01 6.4833587e-01 2.6433140e-01
-2.8643370e-01 1.0163647e-01 -1.2340430e+00 5.7265121e-01
7.0079666e-01 -8.3517873e-01 -1.3504417e+00 6.2292266e-01
7.6831228e-01 -4.7383651e-01 -6.4978880e-01 -3.2162511e-01
2.5503841e-01 -1.2446016e+00 6.4891094e-01 -9.3501306e-01
2.6792079e-01 -1.0034308e-01 -1.3857172e-01 -1.4496169e+00
1.2223115e-01 -9.7623849e-01 2.9552898e-01 1.2923982e+00
5.6725168e-01 -3.5936221e-01 7.3056465e-01 8.9117587e-01
-2.3932429e-01 -1.1846632e-01 -1.0949931e+00 -4.1165441e-01
-3.9015468e-02 -3.4506637e-01 3.5471851e-01 8.0692828e-01
9.8826879e-01 1.2180808e+00 -2.3499413e-01 2.2366301e-03
9.3887486e-03 -2.3405351e-01 1.0045429e+00 -1.0577525e+00
-3.2611978e-01 -2.7111903e-01 -2.1875707e-02 -1.0461900e+00
-5.0719488e-02 -4.7430512e-01 6.7981440e-01 -1.3869716e+00
-3.4906185e-01 -2.2145975e-01 5.1893324e-01 3.6045101e-01
-6.9439366e-02 -2.7375326e-01 3.6635268e-01 1.3715209e-01
-9.5133859e-01 4.7134778e-01 9.7103631e-01 -7.9641799e-03
-5.1686788e-01 3.6330581e-01 -5.1930428e-01 6.2854260e-01
9.0147132e-01 -2.0311959e-01 -3.7428582e-01 -2.6060560e-01
7.8853972e-02 6.4026505e-01 7.2265022e-02 -7.1680957e-01
6.0675019e-01 -4.7052670e-02 -7.5114250e-01 -1.1422875e-01
4.4309121e-01 -6.8770927e-01 -3.2475272e-01 3.2395461e-01
-7.8247333e-01 -9.5566183e-02 -2.3731825e-01 -4.4532262e-02
-3.2822001e-01 -5.9910214e-01 6.6217953e-01 -3.0057082e-01
-3.7900379e-01 -4.1251203e-01 -9.3785900e-01 3.0811748e-01
6.7048639e-01 1.7177671e-01 -5.3672796e-01 -1.1634587e+00
-6.9383413e-01 5.7266080e-01 1.4233574e-01 4.2065600e-01
1.6265327e-01 -7.1156502e-01 -7.5673062e-01 -4.3449169e-01
2.7677813e-01 -2.1351247e-01 1.1485430e-01 6.4727271e-01
-6.6557598e-01 5.5131370e-01 -1.9436459e-01 -5.3658110e-01
-1.5498083e+00 2.2342294e-01 7.6987267e-01 -1.2912737e-01
-3.1493205e-01 7.4552274e-01 -2.4694282e-01 -1.1337041e+00
5.2773339e-01 -7.7891603e-02 -5.8112597e-01 2.2126688e-01
6.9168818e-01 4.0072873e-01 2.5340310e-01 -3.1962594e-01
-3.8440531e-01 -2.9619858e-01 1.8286915e-01 -6.7836952e-01
1.2155933e+00 -3.9741120e-01 -1.7756537e-01 5.0149584e-01
7.2995102e-01 -1.7708506e-01 -7.9710501e-01 -1.9983499e-01
5.2805293e-01 4.2213865e-02 -6.5186846e-01 -1.1203899e+00
-3.2768044e-01 8.9892310e-01 6.8604201e-02 9.2708832e-01
5.5406302e-01 1.2561414e-01 6.1036891e-01 1.1930742e+00
4.2617229e-01 -1.1733974e+00 2.6287246e-01 1.0548937e+00
1.1285329e+00 -1.0455829e+00 -6.1780131e-01 -3.5020360e-01
-1.2251174e+00 1.1751934e+00 6.5426326e-01 8.9591697e-02
2.9597409e-02 1.9096395e-02 8.4950107e-01 -3.3177930e-01
-1.2053202e+00 -2.0717768e-01 -2.4776718e-01 8.3683670e-01
6.3508254e-01 -1.6974299e-01 -3.4165505e-01 6.3246268e-01
-6.4662790e-01 -5.1549524e-03 7.4228030e-01 8.6898601e-01
-7.0954645e-01 -1.1319782e+00 -3.9330590e-01 -1.6861229e-01
-3.0117866e-01 2.5413984e-02 -1.0420102e+00 8.0966735e-01
-5.8110529e-01 1.7327603e+00 -1.6715604e-01 -1.9073913e-01
7.2385812e-01 6.6785312e-01 1.5827993e-01 -6.6688615e-01
-1.3793106e+00 -4.5382187e-01 1.0377687e+00 -6.6161227e-01
-5.2393866e-01 -6.8293864e-01 -1.1122291e+00 -2.7585751e-01
-3.4434557e-01 9.7955924e-01 6.5039921e-01 1.1462222e+00
2.3606156e-01 1.5319391e-01 6.7090976e-01 -3.9184245e-01
-7.1526110e-01 -1.4844816e+00 -2.2071663e-02 3.3075470e-01
2.3586588e-01 1.7371807e-02 -2.8638062e-01 -4.5783170e-03] | [12.758782386779785, 7.971164703369141] |
bc21bb62-d4ff-46e4-ae8b-408ec0e6ca6c | unlocking-the-power-of-deep-pico-extraction | 2005.06601 | null | https://arxiv.org/abs/2005.06601v1 | https://arxiv.org/pdf/2005.06601v1.pdf | Unlocking the Power of Deep PICO Extraction: Step-wise Medical NER Identification | The PICO framework (Population, Intervention, Comparison, and Outcome) is usually used to formulate evidence in the medical domain. The major task of PICO extraction is to extract sentences from medical literature and classify them into each class. However, in most circumstances, there will be more than one evidences in an extracted sentence even it has been categorized to a certain class. In order to address this problem, we propose a step-wise disease Named Entity Recognition (DNER) extraction and PICO identification method. With our method, sentences in paper title and abstract are first classified into different classes of PICO, and medical entities are then identified and classified into P and O. Different kinds of deep learning frameworks are used and experimental results show that our method will achieve high performance and fine-grained extraction results comparing with conventional PICO extraction works. | ['Shaochun Li', 'Zefang Tang', 'Tengteng Zhang', 'Yiqin Yu', 'Jing Mei', 'Xiang Zhang'] | 2020-04-30 | null | null | null | null | ['pico'] | ['natural-language-processing'] | [ 2.84437239e-01 9.61344540e-02 -6.07240021e-01 -2.11699903e-01
-5.72695434e-01 -3.05237353e-01 4.61304843e-01 8.57500672e-01
-2.73436517e-01 1.01593971e+00 5.30402482e-01 -2.36190438e-01
-2.34917447e-01 -9.64526117e-01 -2.97529787e-01 -4.98334736e-01
3.23533237e-01 3.91191006e-01 5.47243841e-02 3.42643470e-01
2.31089339e-01 5.45561731e-01 -1.07195735e+00 4.47798789e-01
1.04433382e+00 7.89275944e-01 1.30929291e-01 4.34726864e-01
-4.60981309e-01 9.81901467e-01 -9.70048726e-01 -4.44012105e-01
-3.23308378e-01 -2.95932710e-01 -9.74455893e-01 2.83458140e-02
-1.16181955e-01 -1.83353931e-01 -1.98547527e-01 1.09545076e+00
6.63133860e-01 -5.40726721e-01 1.02966011e+00 -8.77095640e-01
-5.18238246e-01 9.29203689e-01 -3.31046641e-01 1.77357867e-01
2.90533841e-01 -2.64191240e-01 8.58947396e-01 -5.72075725e-01
8.31535816e-01 9.30966198e-01 6.07571840e-01 4.81713951e-01
-6.60789907e-01 -7.78840721e-01 -1.13327742e-01 8.12379196e-02
-1.33132911e+00 -1.71315551e-01 4.45761323e-01 -3.88020873e-01
6.90247774e-01 2.33390510e-01 6.41252160e-01 9.61195171e-01
6.73536241e-01 9.82790589e-01 8.60456288e-01 -2.65195906e-01
2.10048974e-01 1.63356662e-01 3.60226482e-01 4.92191464e-01
6.94248915e-01 -4.54833090e-01 -3.60335976e-01 -1.33462980e-01
2.62011796e-01 1.72107458e-01 -1.99431807e-01 3.37620854e-01
-1.26715577e+00 3.70311141e-01 2.58996755e-01 5.18768609e-01
-6.04183018e-01 -4.01330441e-01 7.45032072e-01 -3.23777854e-01
3.56949091e-01 5.26967943e-01 -6.46672726e-01 4.43020463e-02
-1.06254780e+00 2.76018053e-01 1.02984858e+00 1.04348993e+00
1.09461822e-01 -5.52674830e-01 -5.93528092e-01 8.85311007e-01
8.11601356e-02 3.39468747e-01 7.33009875e-01 -4.34505641e-01
7.57840574e-01 1.12173104e+00 -2.03676615e-02 -1.34260035e+00
-7.97867835e-01 -4.31823194e-01 -1.16685236e+00 -4.88795877e-01
-1.31697312e-01 -4.01471436e-01 -1.05930758e+00 1.16353869e+00
3.25808436e-01 -6.77383840e-02 2.97250509e-01 4.34223443e-01
1.70557916e+00 7.01274097e-01 4.30632830e-01 -4.20001835e-01
1.93111551e+00 -7.18208313e-01 -1.31927097e+00 -7.24523738e-02
7.20687926e-01 -7.13264704e-01 2.07736075e-01 1.60853043e-01
-8.39138627e-01 -1.55740350e-01 -1.19125724e+00 -1.46785110e-01
-6.45522714e-01 4.22283679e-01 5.87971449e-01 3.60159487e-01
-2.38991797e-01 3.94857734e-01 -6.90633416e-01 -1.10697053e-01
8.38582039e-01 2.06556767e-01 -4.31153148e-01 5.11630662e-02
-1.40833938e+00 7.76103377e-01 7.85444081e-01 1.27415001e-01
-6.23537779e-01 -7.40657985e-01 -8.91031444e-01 3.72826219e-01
3.42431694e-01 -1.18870795e+00 8.79571617e-01 -1.19004011e-01
-8.59270155e-01 1.02371860e+00 -2.27151990e-01 -5.65603435e-01
9.76779684e-02 -1.22838549e-01 -7.22355008e-01 3.03080261e-01
4.16553557e-01 2.85577923e-01 5.02552152e-01 -7.96731949e-01
-1.05955684e+00 -4.26487833e-01 -2.75080372e-02 2.59033233e-01
-4.07165527e-01 4.26225901e-01 -5.56518018e-01 -7.88654089e-01
1.38388485e-01 -5.33079982e-01 -2.62958109e-01 -4.73736495e-01
-9.00213599e-01 -6.25358045e-01 3.81134838e-01 -8.80073071e-01
1.84649312e+00 -1.89045656e+00 -9.44503024e-02 -9.52936038e-02
8.84008586e-01 4.73859191e-01 4.74098295e-01 1.32667720e-01
6.67180121e-02 7.05753922e-01 -2.17754334e-01 1.29392920e-02
-2.23871529e-01 1.15912400e-01 4.29523140e-02 1.42275348e-01
5.15881658e-01 7.47368932e-01 -9.99984026e-01 -1.27990270e+00
-1.90624237e-01 3.11032832e-01 -2.05841690e-01 5.55500463e-02
1.93896979e-01 1.64616913e-01 -9.60635126e-01 8.91484499e-01
6.07559323e-01 -3.77697229e-01 2.44370833e-01 -4.13916677e-01
-3.44817224e-03 8.31694841e-01 -9.98874187e-01 9.49999034e-01
-4.44070339e-01 5.31128407e-01 -1.63379923e-01 -1.04085815e+00
6.62682950e-01 6.49279714e-01 6.59450114e-01 -3.30558628e-01
2.74161339e-01 3.15509737e-01 -5.76014444e-02 -8.48461270e-01
3.17142069e-01 -4.72338572e-02 -2.54828244e-01 -1.53488815e-01
1.72264293e-01 -4.12881561e-02 5.64376712e-01 5.51918373e-02
1.23177171e+00 -3.30919683e-01 8.85568678e-01 -1.75972767e-02
6.47269487e-01 1.55908555e-01 1.23999643e+00 7.51840830e-01
-2.50896126e-01 6.28268838e-01 9.32429433e-01 -2.37709492e-01
-8.80650878e-01 -6.99299812e-01 -7.32414901e-01 1.48355603e-01
-1.96152717e-01 -7.49610901e-01 -7.15566456e-01 -9.61503983e-01
-1.81903750e-01 2.20867783e-01 -6.15675449e-01 5.53335957e-02
-5.96538603e-01 -1.26244736e+00 7.03627110e-01 4.61999416e-01
6.06382489e-01 -1.34444833e+00 -5.24861395e-01 3.87112260e-01
-2.75037318e-01 -1.22698438e+00 -3.56672823e-01 5.96208014e-02
-6.52212143e-01 -1.37833416e+00 -8.92897308e-01 -1.08034885e+00
7.70631135e-01 -2.92126507e-01 9.39167321e-01 -2.12577462e-01
-3.91738117e-01 -2.53381550e-01 -4.17049408e-01 -8.56475115e-01
-4.69210505e-01 3.28302443e-01 -1.30452111e-01 -2.62572557e-01
6.35563850e-01 -3.44930947e-01 -6.26980007e-01 -2.61069536e-01
-1.02582276e+00 9.55843553e-02 8.88331413e-01 8.40948105e-01
6.65763378e-01 3.67160618e-01 7.02880681e-01 -1.21013045e+00
8.65702569e-01 -5.97898841e-01 -2.39702627e-01 4.34565902e-01
-7.16986716e-01 -1.37153462e-01 5.12487590e-01 -3.60316843e-01
-7.51583755e-01 -4.05136086e-02 -1.89500600e-01 1.30249843e-01
-3.35439980e-01 1.00128067e+00 -4.45664346e-01 6.55219734e-01
3.04851830e-01 1.34822935e-01 -3.82414877e-01 -4.25386399e-01
-3.76646244e-03 1.16121769e+00 2.65080452e-01 -8.56126025e-02
3.15036893e-01 1.90466285e-01 8.65041371e-03 -6.95952713e-01
-1.22427046e+00 -4.54338133e-01 -5.88484526e-01 1.93082646e-01
1.27862298e+00 -8.10089767e-01 -7.04242945e-01 3.74167591e-01
-1.30110765e+00 4.29391026e-01 -1.58636235e-02 6.16870224e-01
6.10355437e-02 2.79083014e-01 -7.18522847e-01 -4.58084285e-01
-8.41861308e-01 -1.13931942e+00 1.03072774e+00 6.31387234e-01
-3.22195917e-01 -7.69892991e-01 -3.70893031e-02 4.05782670e-01
-1.63461715e-01 3.95548642e-01 1.14699852e+00 -1.03754592e+00
-7.15548871e-04 -5.48932433e-01 -3.79261166e-01 7.11636022e-02
5.58606863e-01 7.74516910e-02 -4.10046041e-01 4.15916979e-01
1.83179110e-01 2.48639047e-01 7.46060729e-01 4.93810952e-01
1.37765086e+00 -7.60686338e-01 -9.54331696e-01 4.30152446e-01
1.20199585e+00 5.36969006e-01 5.05625129e-01 2.88696915e-01
7.61262059e-01 5.00112176e-01 5.28256059e-01 1.54274613e-01
5.29779851e-01 4.19494867e-01 -1.13846764e-01 -5.50659411e-02
-2.58340061e-01 -5.30128479e-02 9.11433026e-02 9.60076571e-01
2.90544748e-01 -2.44591609e-01 -1.14536524e+00 9.38271105e-01
-1.52113521e+00 -7.99503982e-01 -2.43048206e-01 1.72538435e+00
1.39468348e+00 3.74003321e-01 -2.34209403e-01 2.31844962e-01
7.82742918e-01 4.13602106e-02 -2.21610591e-01 -2.47302115e-01
-2.21131250e-01 3.55883241e-01 4.59057271e-01 -1.09316632e-01
-1.27262497e+00 5.83178461e-01 6.00387526e+00 1.01987147e+00
-1.12108374e+00 3.53800766e-02 6.61625564e-01 2.11786106e-01
1.09284997e-01 -1.92237362e-01 -1.30438590e+00 6.95945203e-01
8.44641626e-01 -3.90292615e-01 -5.25029421e-01 6.00437045e-01
2.32118160e-01 1.19821839e-02 -9.40399349e-01 9.26744699e-01
-5.38230613e-02 -1.57725799e+00 1.47903919e-01 3.06579079e-02
6.25635326e-01 -1.70205683e-01 -3.96878421e-01 2.49857619e-01
-1.42037207e-02 -1.10442519e+00 2.78971493e-01 4.29362446e-01
5.17195582e-01 -5.39104760e-01 1.15919495e+00 2.17393622e-01
-1.10483491e+00 -5.41287363e-02 -2.86487609e-01 3.75932097e-01
8.02937597e-02 1.09846652e+00 -1.05315125e+00 1.00342143e+00
6.56648159e-01 9.38301086e-01 -5.33202231e-01 1.33086562e+00
-3.45471859e-01 7.26777315e-01 -1.74054533e-01 -3.89173806e-01
1.71492212e-02 1.22092210e-01 4.65128809e-01 1.29416382e+00
1.68551758e-01 2.95445919e-01 1.32346243e-01 7.27866769e-01
-4.99491245e-01 4.84802485e-01 -4.65188771e-01 -3.51182967e-01
4.42614108e-01 1.23918056e+00 -8.93106878e-01 -8.30805540e-01
-2.41928235e-01 5.74981809e-01 -7.30292201e-02 -9.69110057e-02
-6.88876033e-01 -9.43299592e-01 1.77092806e-01 -1.10945061e-01
1.23051658e-01 2.72326648e-01 -6.43667400e-01 -1.28090847e+00
1.44984990e-01 -6.80319190e-01 5.47227383e-01 -4.17917848e-01
-1.32007980e+00 4.92783129e-01 -6.84261695e-02 -1.32471251e+00
7.58273974e-02 -5.58699548e-01 -3.86850208e-01 6.65731728e-01
-1.20494580e+00 -9.34307754e-01 -2.38152836e-02 -2.42842082e-03
6.03064358e-01 -6.07422329e-02 8.53358388e-01 7.28601217e-01
-9.83650923e-01 4.04732078e-01 1.67297423e-01 8.01161289e-01
6.61105573e-01 -1.24343872e+00 -1.01420917e-01 5.40499389e-01
-1.13911122e-01 8.06432724e-01 4.53265876e-01 -1.02463686e+00
-6.51281714e-01 -1.14434803e+00 1.62090063e+00 -4.49371248e-01
5.74163020e-01 1.74435861e-02 -6.61260545e-01 3.44356716e-01
1.03830799e-01 -1.51981577e-01 8.70474756e-01 1.51787298e-02
7.61052445e-02 -1.95210397e-01 -1.12100470e+00 6.32203043e-01
6.82454169e-01 -3.99159163e-01 -1.03790152e+00 6.55324996e-01
7.26469755e-01 -6.12252355e-01 -1.35086906e+00 8.01797867e-01
4.71718103e-01 -8.74375850e-02 8.30260396e-01 -8.71852636e-01
8.88239980e-01 -3.76782596e-01 3.93499792e-01 -9.61923063e-01
7.45846480e-02 -1.83677182e-01 -2.85541356e-01 1.56337678e+00
7.95399666e-01 -3.69253367e-01 4.80306923e-01 3.89780015e-01
2.98007801e-02 -1.12394857e+00 -6.79530084e-01 -2.72449434e-01
7.40917102e-02 -1.48747876e-01 8.05791855e-01 1.19940400e+00
2.24395752e-01 8.43126535e-01 -9.86995101e-02 2.95703202e-01
3.81803721e-01 2.81462908e-01 2.08134681e-01 -1.32056808e+00
2.74397552e-01 -5.01838505e-01 -3.04223776e-01 -5.71385801e-01
-5.58586977e-02 -8.92996490e-01 -6.08289763e-02 -2.15860128e+00
6.11098051e-01 -4.89532828e-01 -4.54859704e-01 5.81148148e-01
-5.53736806e-01 -1.29012406e-01 -8.58820155e-02 3.01762193e-01
-4.54689831e-01 2.22482294e-01 1.00215578e+00 -4.74168211e-01
-2.86427528e-01 2.34296098e-01 -6.95896387e-01 7.58924603e-01
8.03020954e-01 -9.48482394e-01 1.45148396e-01 -1.41668111e-01
3.67602587e-01 4.65586260e-02 6.16821274e-02 -9.53959405e-01
1.88350931e-01 -7.19880089e-02 4.45693284e-01 -9.79544938e-01
-2.15932399e-01 -6.90763652e-01 -9.10504013e-02 6.84338450e-01
-3.86978239e-01 -2.41625056e-01 5.76272346e-02 2.53094167e-01
-3.99337500e-01 -5.41515112e-01 2.82356113e-01 -1.24680847e-01
-2.88961738e-01 1.85488805e-01 -4.48901623e-01 1.72690656e-02
7.06611753e-01 2.20097110e-01 -4.96437579e-01 2.48116568e-01
-9.45414305e-01 4.62176055e-01 -1.17346294e-01 3.40471923e-01
5.72481215e-01 -1.20010948e+00 -8.60424280e-01 -5.37454128e-01
2.30491549e-01 3.12693745e-01 2.12501228e-01 9.78998363e-01
-6.29606307e-01 7.92992294e-01 4.16221321e-02 -3.29546750e-01
-1.32431316e+00 7.29878068e-01 1.67632520e-01 -8.82805169e-01
-6.83411896e-01 2.66265959e-01 2.51085490e-01 -3.17747176e-01
1.40594617e-01 -5.61214447e-01 -1.01973140e+00 3.97758216e-01
7.17408240e-01 4.85342145e-02 1.68665379e-01 -5.37047684e-01
-5.81104338e-01 2.69491076e-01 -2.72177696e-01 4.31299180e-01
1.46844137e+00 2.31761754e-01 -4.43576902e-01 2.22112656e-01
1.12035620e+00 2.74622709e-01 -2.32622579e-01 1.40854735e-02
2.51223117e-01 2.87319213e-01 -1.65099934e-01 -7.46150792e-01
-8.97060812e-01 6.85044944e-01 4.30681437e-01 3.13721836e-01
1.18801284e+00 1.53957069e-01 8.99209619e-01 1.92322090e-01
-1.04803935e-01 -1.08324099e+00 -6.03023291e-01 3.31540376e-01
6.28424942e-01 -1.01416802e+00 4.65446532e-01 -6.92375302e-01
-4.21772957e-01 1.06243634e+00 3.07903975e-01 6.66029230e-02
8.07465136e-01 3.07420224e-01 -2.62742192e-01 -4.54317063e-01
-3.94837141e-01 -5.20807840e-02 5.54063976e-01 3.14597815e-01
4.77254093e-01 2.19810396e-01 -1.11702538e+00 1.20614445e+00
-2.13367015e-01 2.09271014e-01 4.32062209e-01 9.32663441e-01
-1.97023958e-01 -1.12601399e+00 -2.56633073e-01 9.11413312e-01
-1.32677114e+00 -3.09701860e-01 -4.15346563e-01 6.50047600e-01
4.05236453e-01 7.89813221e-01 -2.28513747e-01 -2.02937648e-01
2.47944951e-01 1.13957736e-03 8.88252109e-02 -8.97712171e-01
-8.12139869e-01 -4.57113273e-02 3.65930676e-01 -7.82794133e-02
-7.18959033e-01 -4.80524749e-01 -1.48073602e+00 2.38123789e-01
-3.13627392e-01 3.25194329e-01 5.19278407e-01 1.27895081e+00
4.34097201e-01 1.09847307e+00 3.78980130e-01 -2.34051030e-02
3.28332968e-02 -1.03269744e+00 -1.73782289e-01 3.61247033e-01
3.36369365e-01 -4.88663167e-01 -8.06230754e-02 1.90350652e-01] | [8.494970321655273, 8.721845626831055] |
9e68b7d4-8ce5-4730-97fd-1dbdbd6ebbad | a-deep-content-based-model-for-persian-rumor | null | null | https://doi.org/10.1145/3487289 | https://dl.acm.org/doi/abs/10.1145/3487289 | A Deep Content-Based Model for Persian Rumor Verification | During the development of social media, there has been a transformation in social communication. Despite their positive applications in social interactions and news spread, it also provides an ideal platform for spreading rumors. Rumors can endanger the security of society in normal or critical situations. Therefore, it is important to detect and verify the rumors in the early stage of their spreading. Many research works have focused on social attributes in the social network to solve the problem of rumor detection and verification, while less attention has been paid to content features. The social and structural features of rumors develop over time and are not available in the early stage of rumor. Therefore, this study presented a content-based model to verify the Persian rumors on Twitter and Telegram early. The proposed model demonstrates the important role of content in spreading rumors and generates a better-integrated representation for each source rumor document by fusing its semantic, pragmatic, and syntactic information. First, contextual word embeddings of the source rumor are generated by a hybrid model based on ParsBERT and parallel CapsNets. Then, pragmatic and syntactic features of the rumor are extracted and concatenated with embeddings to capture the rich information for rumor verification. Experimental results on real-world datasets demonstrated that the proposed model significantly outperforms the state-of-the-art models in the early rumor verification task. Also, it can enhance the performance of the classifier from 2% to 11% on Twitter and from 5% to 23% on Telegram. These results validate the model's effectiveness when limited content information is available. | ['Arash Sharifi', 'Mohammad-Reza Feizi-Derakhshi', 'Zoleikha Jahanbakhsh-Nagadeh'] | 2020-11-29 | null | null | null | acm-transactions-on-asian-and-low-resource | ['rumour-detection'] | ['natural-language-processing'] | [-4.22337562e-01 -1.28541499e-01 -3.52000594e-01 -1.95375353e-01
2.58972049e-02 -9.59594175e-02 9.87915814e-01 5.38448334e-01
-3.55505019e-01 4.41697359e-01 6.68229043e-01 -7.46904090e-02
1.97098807e-01 -9.12155688e-01 -4.47846688e-02 -3.67063314e-01
-1.27841860e-01 4.68407601e-01 3.95583689e-01 -8.27824771e-01
6.05917394e-01 2.94036120e-01 -1.47354233e+00 4.23781335e-01
7.35096157e-01 9.10137117e-01 1.51275054e-01 2.18065679e-01
-5.08631706e-01 1.06671262e+00 -7.44020581e-01 -2.32668504e-01
-2.85771251e-01 -5.43936729e-01 -8.38429809e-01 1.34007230e-01
-4.28133219e-01 -4.03969169e-01 -4.07259792e-01 1.19025087e+00
4.62380499e-01 -3.65940742e-02 4.47155893e-01 -1.20798290e+00
-8.86976659e-01 9.96298313e-01 -6.51140571e-01 5.22903442e-01
6.18828833e-01 -4.20190692e-01 8.75528097e-01 -1.13051879e+00
5.59887111e-01 1.29861867e+00 6.23676240e-01 2.99639076e-01
-4.83336866e-01 -9.10530031e-01 -1.31037176e-01 7.73573518e-01
-1.19523859e+00 -1.43593624e-01 9.73034799e-01 -5.57200968e-01
4.61364418e-01 1.91724062e-01 7.79591501e-01 1.26354671e+00
3.49829555e-01 7.70853519e-01 1.24674010e+00 -2.80934107e-02
-1.55289277e-01 6.73846126e-01 5.22525251e-01 6.82983279e-01
2.82641929e-02 -5.07666729e-02 -6.07195079e-01 -4.52365160e-01
3.03282529e-01 3.54219347e-01 -2.98120528e-01 4.29835290e-01
-9.49305534e-01 1.32113254e+00 6.02825880e-01 4.82512236e-01
-7.37226903e-01 -8.37609529e-01 6.34129941e-01 4.37611699e-01
9.57905471e-01 3.39919299e-01 1.29241392e-01 -1.01179406e-01
-9.48508978e-01 2.86882669e-01 1.00408661e+00 4.71068949e-01
4.90746200e-01 -1.40726447e-01 1.98094383e-01 1.12293291e+00
2.31474593e-01 4.40323293e-01 8.37327361e-01 6.51924536e-02
2.37669468e-01 9.62223649e-01 1.14246130e-01 -1.74381495e+00
-6.78472757e-01 -5.26568830e-01 -9.26163614e-01 -3.18587154e-01
-1.18795335e-02 -2.98824850e-02 -4.45645750e-01 1.10172296e+00
5.09911954e-01 1.75805002e-01 -1.71491936e-01 1.13572049e+00
1.04201686e+00 8.91769469e-01 -3.88678938e-01 -3.41184080e-01
1.66922164e+00 -7.79581845e-01 -1.17892909e+00 -1.94398403e-01
1.32049099e-01 -1.14549780e+00 6.30577624e-01 7.65906572e-02
-5.72422743e-01 -2.18195453e-01 -1.14324653e+00 2.65600652e-01
-3.99827570e-01 -3.81206512e-01 2.52964258e-01 5.04928231e-01
-4.24856454e-01 4.55978483e-01 -3.13208848e-01 -4.65303153e-01
3.68054152e-01 -2.28671804e-01 -9.92154256e-02 -1.54238313e-01
-1.89994121e+00 1.15507472e+00 3.82104874e-01 8.67536068e-02
-9.68475044e-01 -2.37647504e-01 -5.00836492e-01 -1.48633093e-01
6.07640035e-02 -2.84199208e-01 1.02521002e+00 -8.02850127e-01
-1.32475877e+00 6.68510675e-01 -2.28833547e-03 -5.16823530e-01
6.41202927e-01 -7.69330040e-02 -9.51368392e-01 1.75259307e-01
8.75193477e-02 -1.47741109e-01 1.05781949e+00 -9.55226898e-01
-6.53794050e-01 -5.02258301e-01 -1.66503996e-01 1.32202789e-01
-5.18543363e-01 6.69264913e-01 -1.10486401e-02 -4.72377002e-01
2.90057391e-01 -8.56753469e-01 2.94554800e-01 -4.52417225e-01
-3.77104670e-01 -1.48536891e-01 1.20068765e+00 -1.00026858e+00
1.30111873e+00 -1.96335483e+00 -8.81787464e-02 -5.62814400e-02
4.42063481e-01 4.81117308e-01 3.33490700e-01 8.24625731e-01
1.67479649e-01 1.89428926e-01 -2.17995476e-02 -1.77911595e-01
-2.39264533e-01 -5.88326044e-02 -5.79775393e-01 7.57509887e-01
-1.37300298e-01 4.81776416e-01 -9.82450962e-01 -4.35225338e-01
-5.58756776e-02 5.61258078e-01 -2.28936896e-01 4.69931781e-01
1.42062053e-01 2.83063471e-01 -4.43041474e-01 4.52486247e-01
8.26289713e-01 -2.16984496e-01 7.03731999e-02 1.49519891e-01
-6.80756792e-02 5.97225845e-01 -7.90532768e-01 6.59609318e-01
-4.51867849e-01 6.46650076e-01 1.64698333e-01 -6.13065839e-01
1.32297635e+00 4.26960677e-01 2.76426703e-01 -5.17507493e-01
6.33923233e-01 1.93042368e-01 -1.85837895e-01 -8.31617594e-01
8.32222104e-01 -5.45755923e-01 -1.52836949e-01 7.61930943e-01
-4.32047695e-01 1.66619167e-01 -3.26196939e-01 3.57815444e-01
8.08516860e-01 -6.53731287e-01 5.26676536e-01 -1.58246964e-01
6.73371077e-01 2.39085872e-02 5.00802457e-01 3.01031739e-01
-2.34824821e-01 4.10144448e-01 3.75074357e-01 -2.56373018e-01
-7.70767033e-01 -5.23212433e-01 -3.05586576e-01 1.05845654e+00
3.93840998e-01 -2.12962255e-01 -6.29290640e-01 -6.86707139e-01
1.23919092e-01 7.97707796e-01 -7.95551538e-01 -1.81671560e-01
-3.42095703e-01 -1.15773976e+00 6.15313768e-01 1.38147548e-01
7.67440915e-01 -1.15883839e+00 -9.62084085e-02 3.83314818e-01
-5.92362046e-01 -1.01523042e+00 -3.50755990e-01 -4.71700817e-01
-6.12372220e-01 -1.24351168e+00 -5.68990946e-01 -6.79740608e-01
4.94861364e-01 6.06703222e-01 4.07544553e-01 2.73353994e-01
1.61753044e-01 -3.18055600e-01 -7.43598461e-01 -4.03699726e-01
-8.47832918e-01 -1.08794488e-01 2.09661499e-01 3.94657046e-01
3.58965486e-01 -3.88695002e-01 -2.69969523e-01 7.10194349e-01
-8.33991051e-01 -5.16542904e-02 2.51132369e-01 9.37966347e-01
-3.02474141e-01 1.34503171e-01 1.02972436e+00 -1.04587924e+00
1.27539849e+00 -1.21421313e+00 -5.18186912e-02 -3.92955840e-01
-8.37427795e-01 -3.15868109e-01 6.13711715e-01 -4.50482368e-01
-1.16492331e+00 -1.10109472e+00 -2.04778954e-01 -6.98606074e-02
1.76546350e-01 9.30755913e-01 2.73471594e-01 1.96330383e-01
7.17784226e-01 3.11282396e-01 4.50141877e-01 -6.94942772e-01
1.50476754e-01 1.53106844e+00 -5.34230284e-02 1.58888530e-02
5.63673615e-01 5.43464720e-01 -5.72246373e-01 -1.15663183e+00
-8.71765494e-01 -8.91359746e-01 -1.65604070e-01 -2.27229372e-01
4.56633270e-01 -6.02353990e-01 -5.46274960e-01 9.48858857e-01
-1.40145445e+00 6.35749519e-01 3.26100647e-01 4.03032631e-01
6.00512028e-02 6.55924082e-01 -1.09052920e+00 -8.97159994e-01
-5.79034805e-01 -7.62472510e-01 1.44714460e-01 1.11594843e-02
-2.84852415e-01 -1.00366044e+00 1.37312829e-01 6.62947834e-01
7.79759526e-01 -9.39894021e-02 8.17426324e-01 -1.26137805e+00
-2.25415844e-02 -4.34099525e-01 -6.08069599e-01 4.08192128e-01
3.02961051e-01 -3.59338194e-01 -8.67239952e-01 -2.13174760e-01
5.26960611e-01 -1.32701352e-01 8.75089526e-01 -3.76007736e-01
5.09268761e-01 -7.53490686e-01 -1.99395806e-01 2.35280283e-02
8.89021456e-01 -6.79269135e-02 4.20656413e-01 5.20455182e-01
5.26534677e-01 7.87632942e-01 6.85070157e-01 7.34249651e-01
5.26512921e-01 4.53118980e-01 7.36278772e-01 3.92425925e-01
-2.07282789e-02 -3.95708531e-01 6.67924762e-01 1.41286099e+00
5.68347014e-02 -1.35460854e-01 -9.09750462e-01 8.33539546e-01
-1.64530575e+00 -1.19917107e+00 -3.58143717e-01 1.96850395e+00
9.37877476e-01 4.77816239e-02 1.76492363e-01 1.95357248e-01
1.10018241e+00 6.12837374e-01 -1.64724141e-02 -4.40215498e-01
-1.81510463e-01 -5.17040551e-01 1.80490926e-01 5.93161702e-01
-7.84658611e-01 6.19657993e-01 5.48012590e+00 4.52173591e-01
-1.34177887e+00 5.51015079e-01 -8.45719315e-03 1.78241700e-01
-2.48876587e-01 -2.14127928e-01 -9.47820306e-01 7.70568013e-01
9.42030549e-01 -4.09047335e-01 4.00093675e-01 6.41961157e-01
4.48967516e-01 1.72766432e-01 -4.53744739e-01 7.35524178e-01
5.51470101e-01 -1.30121827e+00 -6.86654076e-02 -3.79997417e-02
7.22453654e-01 3.88929069e-01 -3.37819643e-02 3.31987053e-01
-4.04327996e-02 -8.13759089e-01 5.80293953e-01 1.90283090e-01
1.24898680e-01 -8.31177890e-01 1.50915945e+00 8.28059196e-01
-5.35976589e-01 -2.05966726e-01 -4.92203563e-01 -2.19875231e-01
5.05826950e-01 7.64988840e-01 -1.66662276e+00 4.68254060e-01
3.41516495e-01 8.41604590e-01 -3.33765239e-01 9.30925250e-01
-4.51220989e-01 8.35966408e-01 -2.57257018e-02 -4.63681966e-01
3.57574761e-01 -1.34080118e-02 9.02830005e-01 1.41328359e+00
1.93792045e-01 -5.62609732e-02 1.03654794e-01 7.27039039e-01
-1.58226386e-01 5.02105534e-01 -4.77741599e-01 -3.58744860e-01
5.61911404e-01 1.14118135e+00 -3.39985013e-01 -2.82991588e-01
-1.07116081e-01 8.09822500e-01 1.94640845e-01 -5.96188083e-02
-9.97383893e-01 -3.35438818e-01 4.72691357e-01 7.10192800e-01
8.27158764e-02 5.68909608e-02 -1.20534115e-01 -1.06978106e+00
-2.23937005e-01 -9.55912650e-01 3.91187817e-01 -3.10766101e-01
-1.50413108e+00 8.98956001e-01 -1.53232470e-01 -1.03169501e+00
-1.59371838e-01 -5.84582090e-02 -7.25327611e-01 8.00195754e-01
-1.63879144e+00 -1.05104244e+00 -4.30883616e-01 4.83902782e-01
4.98913616e-01 -2.17834711e-01 7.17171490e-01 4.18724656e-01
-5.31148970e-01 2.17141226e-01 8.29066187e-02 3.10271442e-01
5.54010451e-01 -6.20779097e-01 1.02685884e-01 5.29690564e-01
-3.67752403e-01 8.28951299e-01 1.00417626e+00 -8.17488074e-01
-1.03601193e+00 -1.02757525e+00 1.43057358e+00 -1.46958038e-01
1.08762980e+00 8.21189862e-03 -1.31901324e+00 4.75636244e-01
3.37994277e-01 -3.79646003e-01 4.40498292e-01 3.70639324e-01
-4.91138607e-01 1.48222849e-01 -1.44248486e+00 3.29846948e-01
5.08084834e-01 -7.60101497e-01 -1.17889047e+00 4.16362703e-01
4.40718800e-01 -1.33911029e-01 -5.90630054e-01 -3.43240127e-02
3.44947070e-01 -9.91463661e-01 4.99766856e-01 -6.90627813e-01
5.86344063e-01 -5.11920303e-02 2.41307952e-02 -1.53785813e+00
-2.39346877e-01 -4.60531265e-01 -6.60781795e-03 1.31985307e+00
2.52447516e-01 -1.02950275e+00 3.31394643e-01 -3.44077766e-01
7.96296671e-02 -5.85157573e-01 -8.24767470e-01 -5.38378000e-01
-1.39508933e-01 -3.22738737e-01 6.24649048e-01 1.42841053e+00
5.97676218e-01 4.85920399e-01 -7.03275084e-01 1.99792311e-01
5.05227327e-01 2.33808458e-01 4.59980130e-01 -1.31391263e+00
2.70598471e-01 -4.46790248e-01 -4.74641234e-01 -9.91194308e-01
1.36583507e-01 -1.09485662e+00 -4.87799197e-02 -1.46934044e+00
4.50296849e-01 -3.08854312e-01 -3.36144477e-01 6.51840270e-02
-1.45758763e-01 3.00040059e-02 -6.78848382e-03 5.82424641e-01
-1.40953884e-01 7.80082822e-01 1.15121531e+00 -2.15756103e-01
-1.23689488e-01 1.36225060e-01 -6.65857136e-01 8.37159932e-01
8.65883470e-01 -7.49857485e-01 -1.09724902e-01 5.54340072e-02
3.10975522e-01 2.43758008e-01 2.85523236e-01 -4.50667769e-01
3.31474602e-01 -1.35472760e-01 -3.43597740e-01 -5.57814956e-01
3.82555753e-01 -4.92221713e-01 -3.07841033e-01 2.88172215e-01
-1.29608184e-01 1.15812823e-01 -7.13155419e-02 6.80679917e-01
-4.03952479e-01 -2.48141199e-01 9.69790637e-01 7.35557973e-02
-6.03858173e-01 7.93880150e-02 -6.46685064e-01 5.98476119e-02
8.87089550e-01 1.99642569e-01 -6.17927015e-01 -6.89200759e-01
-5.69066525e-01 6.03743941e-02 4.14749444e-01 8.79119694e-01
1.07120359e+00 -1.15240288e+00 -1.10877943e+00 3.44085783e-01
1.26927599e-01 -6.44081712e-01 3.19390476e-01 1.30278111e+00
-4.36628014e-01 1.74441427e-01 -3.78163248e-01 -2.09888712e-01
-1.41468000e+00 5.14187157e-01 5.59802055e-02 1.81078352e-02
-5.83966315e-01 6.56970441e-01 -2.99492359e-01 -4.22496647e-01
-4.04368192e-02 1.06978998e-01 -7.30484843e-01 7.14952409e-01
1.01729679e+00 5.94380558e-01 -3.97522701e-03 -1.10495484e+00
-3.89710754e-01 -1.65749237e-01 -5.38438678e-01 3.99869867e-02
1.34997582e+00 -3.60922009e-01 -7.37416685e-01 5.59881508e-01
1.23331070e+00 3.49145919e-01 -3.34441245e-01 -5.68054557e-01
-9.06513631e-02 -5.18027782e-01 4.14407372e-01 -5.49468577e-01
-9.12344992e-01 8.82862031e-01 8.42123330e-02 6.43751740e-01
4.92078811e-01 -7.12425485e-02 1.17780626e+00 -7.68324509e-02
4.18293476e-01 -8.37348282e-01 1.04768403e-01 7.64392078e-01
1.01543427e+00 -1.14867437e+00 6.88642189e-02 -3.87711763e-01
-8.56776237e-01 9.75772500e-01 6.55403212e-02 -9.26955417e-02
8.97862375e-01 -9.65034962e-02 1.24835081e-01 -4.41639662e-01
-4.69972521e-01 1.51306570e-01 1.56280831e-01 2.09982872e-01
3.76400411e-01 2.23529920e-01 -6.71727359e-01 9.21147346e-01
-2.58903861e-01 -1.86800972e-01 8.58875513e-01 7.12402642e-01
-9.56982553e-01 -7.50634551e-01 -5.72916925e-01 2.46169105e-01
-6.71853721e-01 -1.31344125e-01 -5.65251529e-01 5.99524796e-01
-1.40798882e-01 1.32252061e+00 -2.36358404e-01 -8.36424470e-01
-1.03504404e-01 1.13109723e-01 -8.14672709e-02 -6.59494936e-01
-6.97154641e-01 -3.36739182e-01 5.31275153e-01 -1.22456610e-01
-1.74988121e-01 -6.70324028e-01 -1.24084103e+00 -9.85805392e-01
-5.75564563e-01 5.63584566e-01 7.56761491e-01 1.13651633e+00
2.58282661e-01 3.34066749e-01 1.07541633e+00 -4.70790327e-01
-1.01604319e+00 -1.40217042e+00 -7.67349005e-01 5.06130517e-01
4.51795608e-01 -7.46204317e-01 -6.67262614e-01 -5.88845551e-01] | [8.245616912841797, 10.190134048461914] |
dc6f673f-0705-427b-9171-c0bffea96425 | deep-reinforcement-learning-for-interference | 2305.07069 | null | https://arxiv.org/abs/2305.07069v1 | https://arxiv.org/pdf/2305.07069v1.pdf | Deep Reinforcement Learning for Interference Management in UAV-based 3D Networks: Potentials and Challenges | Modern cellular networks are multi-cell and use universal frequency reuse to maximize spectral efficiency. This results in high inter-cell interference. This problem is growing as cellular networks become three-dimensional with the adoption of unmanned aerial vehicles (UAVs). This is because the strength and number of interference links rapidly increase due to the line-of-sight channels in UAV communications. Existing interference management solutions need each transmitter to know the channel information of interfering signals, rendering them impractical due to excessive signaling overhead. In this paper, we propose leveraging deep reinforcement learning for interference management to tackle this shortcoming. In particular, we show that interference can still be effectively mitigated even without knowing its channel information. We then discuss novel approaches to scale the algorithms with linear/sublinear complexity and decentralize them using multi-agent reinforcement learning. By harnessing interference, the proposed solutions enable the continued growth of civilian UAVs. | ['H. Vincent Poor', 'Walid Saad', 'Hongliang Zhang', 'Xingqin Lin', 'Mojtaba Vaezi'] | 2023-05-11 | null | null | null | null | ['multi-agent-reinforcement-learning'] | ['methodology'] | [-3.27467993e-02 2.01056451e-01 -1.14191525e-01 6.48485243e-01
-6.18014820e-02 -6.83474243e-01 1.93894818e-01 -2.26437435e-01
-4.17755693e-01 1.60553551e+00 -2.21401200e-01 -4.41779852e-01
-4.46067929e-01 -8.90734196e-01 -4.75109905e-01 -1.00958681e+00
-7.12261319e-01 1.79035529e-01 -1.75999865e-01 -4.87866819e-01
-1.41244039e-01 5.60865998e-01 -9.51836228e-01 -4.96256739e-01
8.36661220e-01 1.08964157e+00 1.22591212e-01 1.03372931e+00
3.58026356e-01 8.51972282e-01 -1.05912185e+00 3.49360913e-01
6.17892563e-01 -4.61707294e-01 -6.40278518e-01 5.90870559e-01
-4.64223444e-01 -7.05636442e-01 -6.63359940e-01 6.26557648e-01
5.95482945e-01 -1.87674806e-01 4.07290101e-01 -1.71022952e+00
-2.88866997e-01 4.12856966e-01 -8.82793546e-01 8.13653544e-02
1.21900208e-01 1.07125118e-02 7.94851780e-01 -3.70976105e-02
3.17220241e-01 7.21169889e-01 4.96575475e-01 1.45214513e-01
-9.11850810e-01 -6.50718510e-01 2.82392383e-01 5.76812327e-02
-1.31502509e+00 -3.16421360e-01 2.98257232e-01 -1.51799098e-01
8.81867290e-01 2.68825173e-01 1.13805938e+00 5.97807229e-01
2.56672770e-01 4.86497223e-01 6.55750513e-01 -6.63692415e-01
3.30216199e-01 -3.16823095e-01 -6.92154825e-01 7.04624951e-01
5.12120545e-01 2.18839630e-01 -3.58926393e-02 -1.57942384e-01
1.06633162e+00 -1.73353329e-01 -6.64206624e-01 -3.63532245e-01
-1.18536806e+00 5.78935862e-01 4.05606031e-01 2.32699394e-01
-7.12088168e-01 4.74891424e-01 3.00737005e-02 7.70814776e-01
4.89883013e-02 6.83793187e-01 -5.50135434e-01 1.33846745e-01
-6.76449239e-01 1.28560796e-01 8.93120408e-01 1.31127048e+00
6.54023945e-01 3.85635853e-01 2.81689614e-01 6.10435545e-01
3.07397038e-01 6.19520605e-01 -8.35390240e-02 -1.34348857e+00
6.38655424e-02 2.86847174e-01 3.77247572e-01 -9.23202336e-01
-8.82593632e-01 -1.31634974e+00 -1.32809770e+00 2.56336361e-01
4.79229838e-01 -1.22688854e+00 -5.29699624e-01 1.58688891e+00
1.94784403e-01 1.78703964e-01 3.95982474e-01 8.26543808e-01
-3.41841340e-01 7.01848507e-01 -3.44871551e-01 -8.41062903e-01
8.49468231e-01 -9.80936885e-01 -5.95626235e-01 -2.04148650e-01
6.51255250e-01 -5.41194499e-01 1.65598944e-01 2.45541826e-01
-8.48580599e-01 3.50930430e-02 -1.47477973e+00 9.08004940e-01
-1.22990325e-01 -1.97158441e-01 4.84710395e-01 8.00287068e-01
-1.02902234e+00 1.41720042e-01 -5.31961679e-01 -4.72499579e-01
5.48992932e-01 9.38857138e-01 1.07933536e-01 1.44782746e-02
-1.21725643e+00 5.27365923e-01 1.40578836e-01 -7.17835948e-02
-6.35728598e-01 -6.44167960e-01 -3.95454973e-01 6.31426647e-02
6.64643407e-01 -9.15246487e-01 1.15264130e+00 -1.15800834e+00
-1.63467896e+00 -7.10872188e-02 5.97228110e-01 -6.55776680e-01
2.42817044e-01 2.64757797e-02 -2.13792920e-01 3.12748909e-01
-1.76481202e-01 4.13996160e-01 5.41156292e-01 -1.23933578e+00
-1.03940725e+00 -7.24976584e-02 7.53349602e-01 3.85312855e-01
-4.76224363e-01 -3.47374856e-01 1.21609643e-02 -5.25179744e-01
-3.49905998e-01 -1.31640017e+00 -6.22822702e-01 -2.19517067e-01
-4.14585382e-01 3.96614701e-01 1.01826668e+00 9.03837208e-04
9.35612738e-01 -1.65338445e+00 3.47625881e-01 2.02212587e-01
4.17660505e-01 1.44651443e-01 -2.41395965e-01 7.42730260e-01
4.69079852e-01 8.06545839e-04 2.40596056e-01 3.19730908e-01
-4.22804564e-01 3.84633839e-01 -2.12873966e-02 4.88501310e-01
-2.10073337e-01 5.05399168e-01 -9.32393610e-01 6.34956434e-02
-4.09291759e-02 2.30898708e-01 -6.80593908e-01 5.61252385e-02
-4.50928211e-02 6.09380007e-01 -6.09067142e-01 6.13861263e-01
7.15453923e-01 -2.57740945e-01 5.73150396e-01 -6.24899119e-02
-3.47373664e-01 -6.25282347e-01 -1.01701140e+00 1.26625359e+00
-6.18948519e-01 4.52603132e-01 5.57723165e-01 -1.25176418e+00
1.29902855e-01 5.84552348e-01 1.05068314e+00 -5.14680743e-01
4.01686400e-01 2.33509853e-01 2.52911776e-01 5.14598191e-02
-1.38935849e-01 -4.40581888e-02 1.13358133e-01 2.95956343e-01
3.76613289e-02 -1.19801629e-02 2.58433551e-01 1.97635427e-01
1.43951499e+00 -3.83982539e-01 6.16784334e-01 -4.10519481e-01
4.69462335e-01 -7.26184770e-02 6.76679254e-01 8.72481704e-01
-2.86465347e-01 -1.74959272e-01 6.20667994e-01 -2.96283782e-01
-7.98004627e-01 -6.25991404e-01 4.73703682e-01 6.39827251e-01
2.28375599e-01 -5.54428659e-02 -9.09559846e-01 -3.12784284e-01
-3.32219973e-02 1.56476274e-01 -1.88647777e-01 -6.78423941e-02
9.72069353e-02 -1.01413155e+00 4.61990386e-01 -3.20282532e-03
6.98352039e-01 -3.47516239e-01 -8.47318947e-01 6.63893998e-01
4.90607135e-02 -1.31924748e+00 -2.62582272e-01 3.28879446e-01
-2.42023155e-01 -1.16290104e+00 -1.10535514e+00 -4.84538287e-01
5.21936178e-01 7.29837537e-01 6.84146583e-01 3.74746710e-01
-2.96181142e-01 7.28689730e-01 -3.21087241e-01 -5.06922901e-01
-1.17832772e-01 2.67621994e-01 4.16605175e-01 -8.89260471e-02
-4.34811473e-01 -8.53479505e-01 -7.16064632e-01 2.90577412e-01
-5.67948759e-01 2.07094681e-02 6.78569555e-01 8.75663161e-01
3.60089481e-01 5.82933545e-01 1.16935480e+00 -4.51825380e-01
5.41331828e-01 -5.08186996e-01 -9.23513412e-01 1.74289554e-01
-2.82249540e-01 -3.71223360e-01 9.12749767e-01 4.25838977e-02
-5.67121148e-01 2.79814042e-02 3.17014813e-01 -1.28139379e-02
8.62856582e-02 3.88893485e-01 -1.11381775e-02 -6.97033405e-01
3.25385869e-01 -6.43859580e-02 5.00289500e-02 5.28626621e-01
6.99200481e-02 8.12797606e-01 -3.29543144e-01 -1.69946939e-01
9.04449940e-01 4.25932288e-01 5.90979338e-01 -1.48758769e+00
-8.17873955e-01 -2.47344542e-02 -2.63824254e-01 -6.07961893e-01
4.28081155e-01 -1.18413353e+00 -1.12342620e+00 5.07687867e-01
-1.07571113e+00 -5.22617459e-01 9.53673497e-02 4.99572515e-01
-6.86112344e-01 3.08683991e-01 -6.14884257e-01 -1.00554538e+00
-1.26580611e-01 -1.05932748e+00 4.93972689e-01 3.19975764e-01
1.82262033e-01 -7.24334121e-01 -7.25853369e-02 1.56101361e-01
7.34874368e-01 2.49753535e-01 7.44543135e-01 1.29721820e-01
-9.15208280e-01 -3.98736186e-02 -2.46507525e-01 6.82460070e-02
4.03721094e-01 -3.60562861e-01 -4.91853058e-01 -7.07834482e-01
-3.57324600e-01 -2.29219735e-01 2.60653526e-01 7.14684725e-01
8.48868728e-01 -5.18850863e-01 -4.22647148e-01 5.30214250e-01
1.41738641e+00 4.96341646e-01 4.10133332e-01 4.54939336e-01
4.25435185e-01 3.36630702e-01 4.96313930e-01 1.08338082e+00
3.17055196e-01 5.74069083e-01 9.57896650e-01 -3.02201539e-01
1.58026263e-01 4.40850943e-01 -1.31292995e-02 5.66587567e-01
-3.77503514e-01 -9.61820543e-01 -6.74813330e-01 1.08828828e-01
-1.89979041e+00 -8.06394219e-01 1.68515638e-01 2.05907583e+00
2.92315811e-01 -6.18038289e-02 2.81003624e-01 2.36952886e-01
2.97038615e-01 -2.24477798e-01 -5.32031715e-01 4.22951542e-02
-2.18300849e-01 -3.67657840e-01 1.10124481e+00 5.96565962e-01
-1.14660287e+00 6.93816662e-01 6.05922937e+00 4.86812115e-01
-9.18423355e-01 -2.81632960e-01 6.31455421e-01 -1.58031657e-01
1.68805927e-01 -2.46180058e-01 -1.49490431e-01 2.44318798e-01
8.39405715e-01 -3.03135276e-01 9.52466190e-01 2.04753146e-01
5.22925317e-01 -4.44831431e-01 -4.94464517e-01 9.31507885e-01
-4.17935878e-01 -1.40702128e+00 -9.37167406e-02 6.44654155e-01
8.09550583e-01 -3.85690331e-02 -1.11228354e-01 5.91363944e-02
2.85345525e-01 -6.23399377e-01 2.05055997e-01 2.98116475e-01
5.31723261e-01 -1.39186168e+00 7.63207912e-01 4.62427884e-01
-9.79153395e-01 -6.13618791e-01 -2.34322190e-01 -6.00864291e-01
1.77030921e-01 7.52400279e-01 -6.06706679e-01 8.02595973e-01
1.84434250e-01 4.26440656e-01 1.08628787e-01 1.16711986e+00
1.78018212e-01 3.34705383e-01 -3.87747645e-01 -3.43255959e-02
4.23300177e-01 -3.83264810e-01 6.98781967e-01 6.59703970e-01
7.84287393e-01 3.12735856e-01 6.81445956e-01 4.87284623e-02
-1.73562914e-01 -1.07533559e-01 -7.60074735e-01 -5.51245287e-02
5.25724769e-01 1.31199300e+00 -8.68170917e-01 -8.95108283e-03
-8.60585928e-01 1.10236096e+00 1.30298510e-01 6.33834124e-01
-8.70558441e-01 -5.29823244e-01 1.01395953e+00 -2.97313314e-02
1.05279960e-01 -6.14184022e-01 1.13215037e-01 -7.28771985e-01
-6.27873480e-01 -9.58690047e-01 -7.25260377e-02 -2.59271294e-01
-9.07704473e-01 5.28372228e-01 -5.57463527e-01 -1.28775454e+00
-9.95993391e-02 -4.16109800e-01 -9.16582197e-02 2.16489866e-01
-1.65940297e+00 -1.09301102e+00 -1.26058668e-01 3.81947756e-01
4.52221185e-01 -5.44781923e-01 9.96964633e-01 3.67873043e-01
-6.98300600e-01 3.10218900e-01 3.56932372e-01 -1.12935625e-01
4.24516320e-01 -1.04282534e+00 -3.94708157e-01 6.59378886e-01
-1.99535966e-01 6.29555136e-02 7.56457567e-01 -3.03275675e-01
-1.81055081e+00 -9.04690504e-01 -3.40483859e-02 5.48963845e-01
7.74586558e-01 -1.59791127e-01 4.50849235e-02 2.88366586e-01
7.98722804e-01 -6.33806363e-03 1.00823402e+00 -1.56778749e-02
2.89592266e-01 -2.24027038e-01 -9.26837921e-01 9.87797976e-01
9.67208028e-01 1.74093843e-01 5.53969979e-01 4.95916247e-01
6.17757797e-01 -6.82608783e-02 -6.20744348e-01 1.28521547e-01
5.04563332e-01 -8.19630444e-01 6.72775090e-01 -3.08703631e-01
-2.23488912e-01 -4.18089420e-01 -2.96413422e-01 -1.91305983e+00
-5.75725019e-01 -1.08666396e+00 -1.74143463e-01 6.35945678e-01
3.72536510e-01 -6.59368217e-01 7.69966960e-01 -1.01786360e-01
2.45567486e-02 -7.14446366e-01 -9.20774221e-01 -9.38460588e-01
-9.25858915e-02 1.63570434e-01 3.67808700e-01 7.82408357e-01
4.15681481e-01 7.72081316e-01 -7.17010677e-01 6.37910783e-01
9.48094666e-01 -1.79620326e-01 6.97445154e-01 -1.14166701e+00
-4.41518277e-01 -1.72641903e-01 -4.89513218e-01 -1.12719917e+00
3.65440771e-02 -3.56928319e-01 -2.27519438e-01 -1.26921093e+00
-2.68604577e-01 -4.23640817e-01 -2.09544644e-01 3.08871537e-01
4.07461494e-01 2.99761176e-01 2.30577812e-01 -1.84762582e-01
-8.00469875e-01 5.60144901e-01 1.37001944e+00 -1.83589995e-01
-2.35599890e-01 2.33324587e-01 -7.24461377e-01 6.59194767e-01
1.27060473e+00 7.77709410e-02 -8.06098759e-01 -5.21368980e-01
2.45438069e-01 4.96124834e-01 5.66137247e-02 -1.59743690e+00
3.27834159e-01 -3.12022358e-01 4.78859723e-01 1.32252108e-02
2.64384627e-01 -1.29845250e+00 -3.17498595e-02 8.42469752e-01
1.67183682e-01 -1.95183292e-01 2.30783343e-01 1.02992392e+00
4.73781303e-03 2.86806315e-01 9.81378376e-01 2.60882229e-02
-2.18701303e-01 3.84225518e-01 -1.45485747e+00 -1.80851534e-01
1.59754336e+00 -7.13048726e-02 -2.75861025e-01 -8.82098854e-01
-4.32407230e-01 7.48636425e-01 2.37033963e-01 -3.42252292e-03
2.83759773e-01 -9.34085131e-01 -4.24860626e-01 -3.09410319e-02
-3.03159207e-01 -5.43012261e-01 4.06769812e-01 9.84684169e-01
-5.60389757e-01 6.02007806e-01 -4.86387491e-01 -2.14935467e-01
-1.15655959e+00 3.59454811e-01 6.89184189e-01 -2.64072716e-01
2.15923116e-02 6.18470728e-01 1.47360101e-01 1.23989269e-01
9.02071893e-02 7.72997588e-02 -1.93026230e-01 -2.18970403e-02
4.47479665e-01 5.44760466e-01 -1.97092354e-01 -2.01780960e-01
-2.32208654e-01 4.39611167e-01 -2.07157478e-01 3.33816148e-02
1.25602388e+00 -5.91797113e-01 -1.97134167e-01 -2.80280203e-01
7.27963209e-01 4.08777073e-02 -1.22589195e+00 1.51102230e-01
-1.42303258e-01 -2.14640647e-01 5.18089652e-01 -7.49037623e-01
-1.20936954e+00 1.57822549e-01 5.12189150e-01 7.51997650e-01
1.38084197e+00 -5.36250472e-01 7.66242504e-01 7.29022443e-01
9.82789159e-01 -1.21930766e+00 -8.86490270e-02 7.49958575e-01
5.65760076e-01 -8.46397102e-01 3.88274416e-02 -6.25406086e-01
-1.42326191e-01 1.11711228e+00 5.80912352e-01 -3.28406692e-02
7.61592269e-01 6.13523543e-01 -1.73297189e-02 7.84358829e-02
-7.38611698e-01 -6.59482002e-01 -7.28079975e-01 9.43890393e-01
2.68658847e-01 1.89652592e-01 -2.09899247e-01 2.27359116e-01
3.47442061e-01 -8.21317434e-02 1.07714629e+00 9.94831502e-01
-8.84713590e-01 -1.44027495e+00 -1.88971534e-01 6.76318467e-01
-4.60583210e-01 1.89856589e-01 -1.59212485e-01 6.56051576e-01
2.85861455e-02 1.35918868e+00 -2.12285176e-01 -1.77895963e-01
3.02901983e-01 -6.80697381e-01 6.67094409e-01 -3.42925698e-01
1.72266424e-01 2.23692313e-01 1.83741525e-01 -3.62736881e-01
-3.91345441e-01 -2.03033984e-01 -1.24713588e+00 -2.99356878e-01
-3.29569936e-01 2.26455644e-01 1.14978507e-01 9.01318550e-01
6.95613265e-01 9.87994134e-01 9.72663701e-01 -8.26896369e-01
-3.75024229e-01 -4.40753907e-01 -9.18889165e-01 -8.00330758e-01
7.85347342e-01 -7.94916391e-01 -1.21969119e-01 -2.84039915e-01] | [5.8680243492126465, 1.6039283275604248] |
80b04517-b085-47f0-981f-41401865cb9f | soccernet-v2-a-dataset-and-benchmarks-for | 2011.13367 | null | https://arxiv.org/abs/2011.13367v3 | https://arxiv.org/pdf/2011.13367v3.pdf | SoccerNet-v2: A Dataset and Benchmarks for Holistic Understanding of Broadcast Soccer Videos | Understanding broadcast videos is a challenging task in computer vision, as it requires generic reasoning capabilities to appreciate the content offered by the video editing. In this work, we propose SoccerNet-v2, a novel large-scale corpus of manual annotations for the SoccerNet video dataset, along with open challenges to encourage more research in soccer understanding and broadcast production. Specifically, we release around 300k annotations within SoccerNet's 500 untrimmed broadcast soccer videos. We extend current tasks in the realm of soccer to include action spotting, camera shot segmentation with boundary detection, and we define a novel replay grounding task. For each task, we provide and discuss benchmark results, reproducible with our open-source adapted implementations of the most relevant works in the field. SoccerNet-v2 is presented to the broader research community to help push computer vision closer to automatic solutions for more general video understanding and production purposes. | ['Marc Van Droogenbroeck', 'Thomas B. Moeslund', 'Bernard Ghanem', 'Kamal Nasrollahi', 'Jacob V. Dueholm', 'Meisam J. Seikavandi', 'Silvio Giancola', 'Anthony Cioppa', 'Adrien Deliège'] | 2020-11-26 | null | null | null | null | ['action-spotting', 'camera-shot-boundary-detection', 'camera-shot-segmentation', 'replay-grounding'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [ 3.66962850e-01 -5.39912768e-02 -3.56124371e-01 -2.08169341e-01
-7.06014812e-01 -8.50187123e-01 2.28055060e-01 -3.10912997e-01
-3.71853948e-01 5.23485601e-01 4.33072448e-01 1.23029657e-01
1.43354490e-01 -1.11411147e-01 -1.02906644e+00 -4.90832508e-01
-1.11946449e-01 3.10506761e-01 8.80752265e-01 -6.26024187e-01
3.10140848e-01 -1.23844169e-01 -1.46579635e+00 9.93226826e-01
-2.49667559e-02 1.25026000e+00 1.40928134e-01 1.07029855e+00
5.35366535e-01 1.36830878e+00 -7.34318018e-01 -6.70279324e-01
3.68139297e-01 -3.36300433e-01 -9.15097713e-01 3.40882182e-01
6.93856597e-01 -3.06811035e-01 -5.93481958e-01 9.52355802e-01
9.64987725e-02 4.17711496e-01 1.21725872e-01 -1.71787751e+00
-6.58766776e-02 8.85772765e-01 -4.82769638e-01 6.64067507e-01
9.18904543e-01 2.76453257e-01 1.00881195e+00 -5.59925377e-01
1.02583015e+00 1.00221384e+00 8.82472038e-01 3.82243663e-01
-6.85139596e-01 -7.91944385e-01 1.51048154e-01 8.82058501e-01
-1.26468527e+00 -5.85355759e-01 5.17149687e-01 -6.10307515e-01
7.23415494e-01 3.43004674e-01 1.10494995e+00 1.71320605e+00
-2.47825198e-02 1.43602967e+00 5.46078086e-01 -1.20358892e-01
6.20533712e-02 -2.30876669e-01 3.91364306e-01 5.82485259e-01
-1.43524513e-01 -6.73822686e-02 -1.02643418e+00 2.56043553e-01
9.02742088e-01 -3.53677392e-01 -4.80518907e-01 -4.88152802e-01
-1.58672845e+00 7.21176207e-01 -8.52440447e-02 -4.92931344e-02
-1.06067501e-01 5.18516302e-01 1.10403967e+00 2.96037078e-01
1.10464938e-01 3.99282247e-01 -3.15436244e-01 -8.51314604e-01
-1.15417182e+00 9.02982473e-01 9.32243288e-01 1.34015143e+00
2.26031214e-01 4.28311899e-02 -2.04605028e-01 5.99789619e-01
-5.79558462e-02 -5.82910441e-02 9.04759243e-02 -1.81052291e+00
6.32943571e-01 1.54343724e-01 2.65756547e-01 -1.26912200e+00
-2.64186054e-01 1.56141454e-02 -9.29732323e-02 -2.87627488e-01
5.11829734e-01 -7.00351521e-02 -5.59762537e-01 1.18014896e+00
1.01856343e-01 7.86588550e-01 8.81800652e-02 1.45047033e+00
1.08840489e+00 6.99863315e-01 -2.15072706e-01 -3.64582688e-02
1.60812783e+00 -1.64179015e+00 -8.56879771e-01 -2.91451216e-01
3.25718790e-01 -7.08013058e-01 7.07524002e-01 9.26834762e-01
-1.06066942e+00 -7.31125593e-01 -1.33421719e+00 -1.35842636e-01
-1.45239502e-01 1.11073196e-01 7.19620764e-01 5.57015538e-01
-8.36983085e-01 2.23716453e-01 -8.21803570e-01 -6.15199208e-01
1.47764862e-01 -6.84618205e-02 -4.09177780e-01 -1.82652459e-01
-1.35306454e+00 7.05901921e-01 8.54350626e-01 1.37247801e-01
-1.69912422e+00 -6.75584972e-01 -1.10434306e+00 -2.79708236e-01
1.15950930e+00 -2.43711665e-01 1.80300355e+00 -1.29164708e+00
-1.42061961e+00 9.65206683e-01 4.96537268e-01 -9.81424689e-01
7.43243456e-01 -6.25749409e-01 -4.30339009e-01 6.37271762e-01
2.63984770e-01 8.48116159e-01 8.15503716e-01 -9.57273304e-01
-9.76630092e-01 4.57679927e-02 5.93505383e-01 2.53499150e-01
1.56221062e-01 4.11099672e-01 -1.21560347e+00 -9.67601180e-01
-1.96040943e-01 -9.33708072e-01 -6.77032247e-02 -2.60285854e-01
-1.85647935e-01 1.32800952e-01 9.17893946e-01 -9.17923033e-01
1.14962339e+00 -2.22680879e+00 4.08174485e-01 -1.89153239e-01
1.07954755e-01 -2.24511791e-02 9.47475135e-02 2.88839072e-01
8.64668041e-02 -4.25105482e-01 6.16916083e-02 -2.96148788e-02
3.98119129e-02 3.85016799e-01 -4.05905932e-01 4.83742714e-01
-1.59495443e-01 6.41292036e-01 -1.08808947e+00 -6.84668422e-01
2.88661689e-01 -7.07963184e-02 -7.71400034e-01 -9.42924768e-02
-3.34450036e-01 2.80600220e-01 -1.73827052e-01 9.28849280e-01
3.40205729e-01 3.44433904e-01 8.12679678e-02 -3.60891372e-01
-9.10912529e-02 -2.63362467e-01 -1.38467693e+00 2.43080640e+00
8.03479031e-02 1.04979050e+00 3.70094538e-01 -1.24980140e+00
4.55947846e-01 5.36083043e-01 4.37341332e-01 -2.25318596e-01
2.67319113e-01 -2.93347389e-01 -2.08726212e-01 -8.50524604e-01
1.16322625e+00 1.09259777e-01 -5.85315347e-01 1.35514602e-01
6.18845582e-01 -2.56396592e-01 9.14384544e-01 6.89613402e-01
1.01999748e+00 5.81691742e-01 1.23380512e-01 -1.00111336e-01
4.03059393e-01 5.79094708e-01 7.08994389e-01 1.07635736e+00
-6.47758603e-01 8.18110287e-01 6.77356958e-01 -6.25302792e-01
-1.03274798e+00 -7.81739473e-01 2.61154413e-01 1.57002759e+00
6.43648386e-01 -8.62366259e-01 -1.03282642e+00 -6.02700591e-01
-3.24220389e-01 5.25571525e-01 -6.67477369e-01 1.37483761e-01
-7.19497979e-01 -4.56576675e-01 9.39228952e-01 8.27282310e-01
6.04950964e-01 -1.11323118e+00 -6.89250767e-01 3.62694472e-01
-8.29313457e-01 -1.90244126e+00 -6.46653473e-01 -2.86298841e-01
-4.44473147e-01 -1.49062586e+00 -6.71415389e-01 -7.33911157e-01
-1.02730975e-01 4.39387023e-01 1.19272876e+00 -5.45211174e-02
-5.45489728e-01 8.46586525e-01 -9.28997397e-01 -4.04648453e-01
-3.05689096e-01 -1.42882496e-01 1.07121719e-02 9.43283364e-02
4.12683368e-01 -4.70310217e-03 -2.85275698e-01 7.89910913e-01
-7.53882945e-01 3.97980601e-01 -9.80357602e-02 4.81846660e-01
3.33566636e-01 -5.25732189e-02 1.38765678e-01 -5.11468709e-01
2.45914236e-01 -4.48843300e-01 -5.18341362e-01 8.45963135e-02
5.45625925e-01 -6.66010499e-01 5.26654795e-02 -3.48269999e-01
-1.12121403e+00 1.84428543e-01 6.49843290e-02 -4.64270264e-01
-2.68117458e-01 1.44625545e-01 1.58209801e-01 1.91806592e-02
6.74541056e-01 2.87470758e-01 -3.86056870e-01 -3.46547514e-01
4.81037945e-01 3.14550996e-01 1.15709651e+00 -7.17791319e-01
2.23809600e-01 6.13274634e-01 -6.58753753e-01 -9.42641795e-01
-1.13344216e+00 -9.26252663e-01 -6.15822673e-01 -9.41087425e-01
1.33643711e+00 -1.19796574e+00 -9.47156489e-01 4.40827250e-01
-1.18672705e+00 -3.95661652e-01 -7.16097280e-02 5.28512418e-01
-1.23326004e+00 5.55838883e-01 -9.56709146e-01 -4.85489130e-01
1.32751182e-01 -1.28975356e+00 1.01488280e+00 2.01381788e-01
-1.99753001e-01 -6.35335803e-01 1.80857718e-01 1.18579352e+00
-4.10394877e-01 3.09780210e-01 1.53571796e-02 -6.02603197e-01
-5.51221728e-01 -2.01453000e-01 -1.22239038e-01 5.19480288e-01
-6.46339238e-01 6.65050279e-03 -7.35310256e-01 -1.33722097e-01
-1.70105234e-01 -6.95686936e-01 1.01176810e+00 4.92441028e-01
1.20536137e+00 1.35627180e-01 -8.25173035e-02 6.33608937e-01
9.51079369e-01 1.84127450e-01 6.29064202e-01 5.81487000e-01
4.91119623e-01 4.83124733e-01 1.07536805e+00 5.36161602e-01
2.58340448e-01 8.59939516e-01 3.75087410e-01 1.71824604e-01
-6.11360446e-02 -3.67039740e-01 7.01811135e-01 7.14519262e-01
-7.21131980e-01 -2.18546957e-01 -7.34510958e-01 7.09280789e-01
-2.33654475e+00 -1.33102000e+00 -1.51030585e-01 1.50214744e+00
4.42190707e-01 3.42650652e-01 4.50409412e-01 2.91591942e-01
8.56414795e-01 3.81559551e-01 -5.65616880e-03 -1.77257419e-01
-1.26654223e-01 -1.29751161e-01 5.26024044e-01 1.89795732e-01
-1.63648391e+00 1.26268816e+00 6.93560457e+00 1.12642133e+00
-4.66667205e-01 3.33189636e-01 7.14035705e-02 -2.53687382e-01
5.75448155e-01 7.53626376e-02 -8.63173485e-01 1.89738244e-01
5.67490041e-01 6.76100031e-02 3.53130877e-01 1.08054197e+00
2.98469633e-01 -3.72012556e-01 -1.24482787e+00 1.23821700e+00
6.88938320e-01 -1.68003774e+00 -2.66110659e-01 -5.77798069e-01
6.82794869e-01 -8.40599388e-02 -4.42809135e-01 6.78499460e-01
4.17504162e-01 -7.20097303e-01 1.46259344e+00 3.87742609e-01
2.52362430e-01 -6.91316128e-01 8.32285941e-01 8.35179761e-02
-1.33220351e+00 -2.78557599e-01 -3.75082672e-01 -3.34254205e-01
7.00826585e-01 -2.83898294e-01 -5.48525989e-01 4.83697295e-01
1.15597010e+00 1.20919061e+00 -5.59608459e-01 1.15601587e+00
-2.71737337e-01 6.06412768e-01 -6.38281703e-02 3.96349430e-01
5.04988909e-01 8.27539340e-02 7.68796027e-01 1.52863097e+00
-2.17283461e-02 3.13539624e-01 7.25045860e-01 4.19155270e-01
1.69545397e-01 -1.81546673e-01 -3.58894616e-01 -5.76704927e-02
-6.37349710e-02 1.23622048e+00 -1.17819011e+00 -6.81858063e-01
-3.67010713e-01 1.04876232e+00 -2.80223161e-01 3.23227555e-01
-1.46565962e+00 -2.88187504e-01 6.02650225e-01 9.42143574e-02
5.23694038e-01 -2.92872041e-01 2.36158267e-01 -1.42473662e+00
-3.32645118e-01 -1.12603867e+00 7.46372938e-01 -1.23940694e+00
-8.16364050e-01 2.63166428e-01 5.38806617e-01 -1.05880213e+00
-1.62449986e-01 -6.93657517e-01 -3.04231137e-01 -4.55819100e-01
-7.42579281e-01 -1.01355433e+00 -5.43149650e-01 7.64855266e-01
1.49412036e+00 -1.96542829e-01 2.04246476e-01 4.09894526e-01
-4.41253990e-01 2.86215216e-01 -3.94401431e-01 5.34977198e-01
7.50707388e-01 -8.07964385e-01 4.45882559e-01 7.45206654e-01
5.68950951e-01 -2.36563951e-01 1.19915223e+00 -4.17621404e-01
-1.23458016e+00 -6.52456224e-01 3.08500435e-02 -8.04670095e-01
1.11373675e+00 -6.34738922e-01 -4.19448346e-01 1.22100186e+00
2.89387494e-01 -3.16036433e-01 4.87892389e-01 -2.70719528e-01
-1.53861806e-01 6.94303066e-02 -6.03719354e-01 4.24673945e-01
1.16793597e+00 -2.73014158e-01 -1.04423070e+00 4.09140021e-01
7.61400342e-01 -7.58551002e-01 -6.89511418e-01 2.77383536e-01
6.98432803e-01 -1.01670504e+00 1.02603734e+00 -6.41655028e-01
5.88364363e-01 -2.65001327e-01 -3.33081216e-01 -7.37926483e-01
2.37717152e-01 -6.29215956e-01 1.96400270e-01 8.80333245e-01
1.44586980e-01 4.75178182e-01 9.61819172e-01 1.97624892e-01
-4.59830731e-01 1.51004139e-02 -7.78487742e-01 -5.84928930e-01
-6.58125758e-01 -1.18485522e+00 -1.15013719e-01 8.40700090e-01
3.62769514e-01 -1.12959370e-01 -9.35514331e-01 -1.09285384e-01
6.36341751e-01 -2.76423126e-01 1.20817733e+00 -7.81102598e-01
-5.15436947e-01 -5.05603589e-02 -9.72347975e-01 -1.41733086e+00
7.84704015e-02 -5.64554334e-01 4.24041420e-01 -1.06185961e+00
3.86233062e-01 3.36381197e-01 -9.52386931e-02 3.23686600e-01
4.41818237e-01 9.92632627e-01 6.91845775e-01 -1.15255276e-02
-1.48543882e+00 1.75095439e-01 1.09825206e+00 -2.52294600e-01
1.01774029e-01 -1.65720955e-02 -2.45862424e-01 1.17223322e+00
2.26833180e-01 -5.09046316e-01 -2.59464592e-01 -3.90812486e-01
1.73727706e-01 5.13321579e-01 6.52526081e-01 -1.35684144e+00
3.26844305e-01 -2.27287158e-01 1.49984345e-01 -6.76014364e-01
7.03100741e-01 -5.16373217e-01 5.02664559e-02 3.31283897e-01
-4.01471049e-01 3.18909958e-02 3.08974892e-01 7.94312000e-01
-6.02271378e-01 -2.97234446e-01 3.99979502e-01 -5.36550045e-01
-1.70613205e+00 -4.65234220e-02 -9.04547393e-01 3.91450256e-01
1.49675632e+00 -5.17022550e-01 -1.04098476e-01 -6.32091522e-01
-1.30051291e+00 4.49310869e-01 2.46488005e-01 9.16308165e-01
5.96841633e-01 -1.04472840e+00 -6.84464514e-01 -1.47507742e-01
2.29224220e-01 -4.48943079e-01 7.03078926e-01 1.00187814e+00
-1.24966729e+00 3.22359443e-01 -6.56975925e-01 -8.21915388e-01
-1.34550345e+00 5.95204771e-01 5.67591116e-02 1.00131921e-01
-8.64888787e-01 9.44690704e-01 2.43249059e-01 7.51711130e-02
5.76572120e-01 -4.69834507e-01 -3.62112164e-01 1.55112341e-01
8.85102749e-01 6.40597939e-01 -1.06919251e-01 -6.89781129e-01
-3.45783919e-01 3.06185663e-01 -8.56796354e-02 -1.95119828e-01
1.05382609e+00 -9.85383242e-02 2.45506302e-01 4.92482424e-01
5.88619947e-01 -5.46929896e-01 -1.65097988e+00 1.26063049e-01
1.20093420e-01 -4.86526579e-01 -2.07599830e-02 -5.55014670e-01
-1.14015305e+00 5.48472285e-01 2.42545336e-01 -2.53050160e-02
9.13026512e-01 1.23212941e-01 8.67036164e-01 5.19309938e-01
5.39345324e-01 -1.65343153e+00 4.38378513e-01 5.62661231e-01
9.52371001e-01 -1.15938497e+00 6.05894849e-02 -4.03223008e-01
-1.23960102e+00 1.14530587e+00 7.44086385e-01 -3.61214399e-01
2.54401118e-01 3.93139988e-01 1.93758458e-01 -2.85645396e-01
-6.89200699e-01 4.48996276e-02 1.37080932e-02 5.73508799e-01
-1.14728816e-01 -1.19464464e-01 -2.69474983e-01 1.02901471e+00
-2.88404226e-01 3.21093023e-01 1.05276442e+00 9.83909726e-01
-5.10655522e-01 -5.39365172e-01 -6.42254055e-01 -2.22609356e-01
-6.09748006e-01 1.59624010e-01 -4.45264250e-01 1.11333644e+00
4.19659346e-01 9.95047927e-01 1.81090191e-01 -4.31004912e-01
4.05998915e-01 -1.49643406e-01 6.11503243e-01 -4.31671143e-01
-7.08369851e-01 9.70685259e-02 4.52456802e-01 -9.04058158e-01
-8.49382639e-01 -5.59323311e-01 -8.98517132e-01 -2.55018353e-01
-8.91555175e-02 1.30770162e-01 6.22236133e-01 8.90724719e-01
-3.04544985e-01 4.16807681e-01 -2.66681731e-01 -1.31568682e+00
6.18121326e-02 -7.56980062e-01 -6.58342123e-01 4.62523341e-01
-7.02467188e-02 -8.54977667e-01 -9.39540341e-02 7.53673136e-01] | [7.937693119049072, 0.20294694602489471] |
926fb142-a711-42f2-9e2c-5a7b6fc4e662 | int-hrl-towards-intention-based-hierarchical | 2306.11483 | null | https://arxiv.org/abs/2306.11483v1 | https://arxiv.org/pdf/2306.11483v1.pdf | Int-HRL: Towards Intention-based Hierarchical Reinforcement Learning | While deep reinforcement learning (RL) agents outperform humans on an increasing number of tasks, training them requires data equivalent to decades of human gameplay. Recent hierarchical RL methods have increased sample efficiency by incorporating information inherent to the structure of the decision problem but at the cost of having to discover or use human-annotated sub-goals that guide the learning process. We show that intentions of human players, i.e. the precursor of goal-oriented decisions, can be robustly predicted from eye gaze even for the long-horizon sparse rewards task of Montezuma's Revenge - one of the most challenging RL tasks in the Atari2600 game suite. We propose Int-HRL: Hierarchical RL with intention-based sub-goals that are inferred from human eye gaze. Our novel sub-goal extraction pipeline is fully automatic and replaces the need for manual sub-goal annotation by human experts. Our evaluations show that replacing hand-crafted sub-goals with automatically extracted intentions leads to a HRL agent that is significantly more sample efficient than previous methods. | ['Andreas Bulling', 'Stefan Leutenegger', 'Mihai Bâce', 'Florian Strohm', 'Simon Schaefer', 'Anna Penzkofer'] | 2023-06-20 | null | null | null | null | ['hierarchical-reinforcement-learning', 'montezumas-revenge'] | ['methodology', 'playing-games'] | [-1.98595878e-02 6.98011458e-01 1.21845435e-02 -1.81101635e-01
-8.23756456e-01 -5.01403689e-01 7.02368200e-01 -1.21761657e-01
-8.57925296e-01 8.26651752e-01 4.60383624e-01 -2.97456592e-01
-1.09134614e-03 -3.93921047e-01 -2.67056167e-01 -3.09313089e-01
-1.48701876e-01 7.74092197e-01 3.22503895e-01 -5.40782988e-01
3.95913810e-01 -9.72825885e-02 -2.01731753e+00 6.09518737e-02
4.04390544e-01 9.47211206e-01 2.06078812e-01 8.20215344e-01
5.44990480e-01 1.70812058e+00 -4.57222670e-01 -1.90654248e-01
5.88152051e-01 -7.89012194e-01 -7.77149558e-01 1.26873270e-01
5.21434247e-02 -7.40314364e-01 -1.82433903e-01 8.10375690e-01
3.99450868e-01 4.09863651e-01 4.63883609e-01 -1.32020843e+00
-1.68248907e-01 5.22408009e-01 -5.22930682e-01 7.63781741e-02
4.06484187e-01 8.47642004e-01 1.57332945e+00 -1.70921400e-01
8.13907743e-01 1.07307994e+00 5.31545639e-01 8.97760868e-01
-1.09669483e+00 -5.82833409e-01 4.97065112e-02 2.15855330e-01
-8.44094157e-01 -8.07070017e-01 5.62247157e-01 -6.10921085e-01
1.37471855e+00 -1.28351346e-01 7.96847224e-01 1.01591444e+00
8.36814661e-03 1.08805633e+00 1.22950315e+00 -2.71128953e-01
5.16991377e-01 -3.31440955e-01 3.06113642e-02 1.08359933e+00
-2.37475708e-02 8.02973628e-01 -8.81919086e-01 6.21460266e-02
6.37268364e-01 -3.81347507e-01 1.39268395e-02 -3.50509644e-01
-1.09086287e+00 1.17192173e+00 2.46318713e-01 -1.38763770e-01
-8.12120736e-01 2.75501043e-01 3.27941358e-01 2.02075824e-01
1.44067839e-01 9.12612915e-01 -3.87490243e-01 -6.61403298e-01
-6.73735201e-01 5.16612291e-01 7.15315640e-01 8.85828733e-01
8.47346783e-01 1.06848963e-01 -3.67370754e-01 5.15098929e-01
4.32621628e-01 1.62935436e-01 7.74296463e-01 -1.53087592e+00
2.38641292e-01 8.90825927e-01 5.23677349e-01 -5.65677762e-01
-9.42134142e-01 -4.15929586e-01 -6.22058567e-03 9.46611166e-01
7.50219643e-01 -4.91031587e-01 -5.66329479e-01 1.80995011e+00
3.25480908e-01 -4.61273551e-01 1.41139045e-01 8.64927053e-01
5.27123928e-01 8.62405822e-02 1.12039365e-01 -1.33190855e-01
1.36401129e+00 -1.04041982e+00 -3.85164469e-01 -9.88372505e-01
7.90940762e-01 -1.91128980e-02 1.25819862e+00 6.62593961e-01
-9.71002817e-01 -1.96463048e-01 -9.17843282e-01 -1.97131559e-01
1.91801131e-01 8.52602795e-02 1.08871055e+00 3.87228906e-01
-1.13571036e+00 2.91348487e-01 -5.86925030e-01 -7.04912245e-02
7.12356985e-01 5.66082358e-01 -2.97320187e-01 2.09090412e-01
-7.95676768e-01 9.68730390e-01 6.30820990e-01 -3.24588716e-01
-1.28174436e+00 -1.55323774e-01 -1.11931443e+00 5.63680343e-02
1.08051407e+00 -3.38851690e-01 1.97252417e+00 -1.06149960e+00
-1.77068472e+00 1.08257842e+00 1.91783443e-01 -7.59775043e-01
5.03637731e-01 -9.70354229e-02 2.77887970e-01 -1.43049642e-01
3.35742354e-01 1.06904924e+00 9.05819952e-01 -7.28547692e-01
-1.33773410e+00 -4.13328201e-01 4.55241650e-01 6.98331416e-01
2.08623677e-01 1.18741401e-01 -1.62361071e-01 -2.64437273e-02
-5.10642886e-01 -1.18712020e+00 -2.98512697e-01 -3.85359079e-01
-2.02791885e-01 -9.09758866e-01 1.95353225e-01 -5.69996178e-01
1.03998256e+00 -1.94338679e+00 2.32233778e-02 -2.74269372e-01
8.31494391e-01 -1.94737297e-02 -2.47147068e-01 -5.41691333e-02
2.71450758e-01 -3.63708168e-01 1.30535915e-01 -4.58689898e-01
5.01064062e-01 -2.33496772e-03 1.30517427e-02 4.78564799e-01
-1.49833962e-01 1.03332591e+00 -1.15265286e+00 -3.40652257e-01
1.92814425e-01 -3.69438946e-01 -8.86761844e-01 4.05718684e-01
-7.09543228e-01 3.93012464e-01 -1.63032547e-01 2.62289077e-01
-4.88916159e-01 -3.77607226e-01 6.32564873e-02 5.24883986e-01
-2.05854625e-01 8.69112074e-01 -6.91259742e-01 1.55683529e+00
-9.09591764e-02 6.73230231e-01 -1.29051223e-01 -4.38888609e-01
5.50102055e-01 1.19394816e-01 4.59075153e-01 -1.09284306e+00
4.44147319e-01 -2.16624543e-01 4.84513730e-01 -1.73352808e-01
6.07642651e-01 -2.15590015e-01 -3.53622168e-01 9.94311988e-01
2.41340294e-01 -2.33092159e-02 6.27954960e-01 -8.29716623e-02
1.55032182e+00 5.34468710e-01 9.31954563e-01 7.72859976e-02
1.04013093e-01 6.31235242e-01 7.53397822e-01 1.02499330e+00
-4.91955310e-01 1.35160789e-01 8.73282313e-01 -6.34591579e-01
-9.18253422e-01 -3.57861400e-01 7.30854392e-01 1.88167012e+00
-3.75424623e-01 -6.57030225e-01 -5.99139452e-01 -8.39402676e-01
-3.75194192e-01 1.14919746e+00 -8.28726947e-01 -1.66577473e-01
-2.61143088e-01 -4.71276492e-01 4.98282760e-01 2.19372183e-01
4.73882943e-01 -1.66892445e+00 -1.35830176e+00 2.80968994e-01
-6.40879199e-03 -1.04997611e+00 -4.01331753e-01 3.38201940e-01
-3.94181937e-01 -1.48725104e+00 -1.06580704e-01 -3.43591273e-01
2.65071422e-01 -1.08715266e-01 1.51433635e+00 6.07546642e-02
-2.55802304e-01 3.93899530e-01 -2.91985095e-01 -6.49268270e-01
-4.42856640e-01 -1.26849160e-01 1.43000960e-01 -4.71117526e-01
9.20564175e-01 -1.85059339e-01 -2.32126698e-01 8.41094106e-02
-9.29878727e-02 6.12400293e-01 5.24803340e-01 7.59497404e-01
2.32053727e-01 -9.79704317e-03 3.36645693e-01 -8.13634574e-01
1.00913191e+00 -3.79511803e-01 -1.16555786e+00 -2.67640859e-01
-8.12740386e-01 4.46468294e-01 2.61592597e-01 -2.59793162e-01
-9.33011532e-01 3.09483916e-01 7.60399997e-02 -1.24240099e-02
-4.11391437e-01 2.59051561e-01 3.27534080e-01 2.53080368e-01
1.04807568e+00 5.67334853e-02 2.32132569e-01 -1.82231534e-02
4.02799040e-01 4.72992480e-01 3.60041082e-01 -4.12547916e-01
3.63172323e-01 1.24411233e-01 5.60007431e-02 -4.63010788e-01
-1.15604115e+00 -3.55817944e-01 -2.80244797e-01 -4.57234234e-01
8.20907176e-01 -1.09534287e+00 -1.64525914e+00 3.98439199e-01
-7.32317626e-01 -1.21335626e+00 -7.28666723e-01 3.75763208e-01
-1.16296244e+00 -1.45410925e-01 -5.80566943e-01 -8.98422599e-01
-2.51806289e-01 -1.05581307e+00 8.69840145e-01 4.11337018e-01
-5.84149420e-01 -6.10347390e-01 3.75742942e-01 5.84227145e-01
-1.59963574e-02 -2.09235877e-01 5.02098203e-01 -9.72176433e-01
-4.46053833e-01 2.31677927e-02 -4.47682105e-02 -2.89676581e-02
-9.03657451e-02 -6.26366377e-01 -8.25215578e-01 5.45783676e-02
-7.37156346e-02 -9.65946913e-01 4.70597893e-01 2.83169419e-01
2.29021743e-01 -4.77940440e-01 1.40995920e-01 1.24305777e-01
9.40358758e-01 1.69174939e-01 4.57413852e-01 5.89704216e-01
5.04053712e-01 8.18130851e-01 7.67361104e-01 7.79535234e-01
7.54615009e-01 6.72261298e-01 6.17984653e-01 3.85381550e-01
3.14039439e-02 -3.23987693e-01 7.50008404e-01 -1.99235573e-01
-2.98016369e-01 -3.78889591e-02 -9.95500922e-01 3.91287178e-01
-2.05665779e+00 -1.10760403e+00 1.68297306e-01 1.93256319e+00
9.90312219e-01 4.20466691e-01 8.33617270e-01 -1.17677681e-01
2.58269101e-01 6.31110519e-02 -9.67433393e-01 -4.95025218e-02
3.78885776e-01 7.80771077e-02 5.23387074e-01 6.79303765e-01
-8.82240593e-01 1.44708312e+00 5.99703169e+00 5.90967417e-01
-5.80202222e-01 2.03914672e-01 3.01900536e-01 -4.31539685e-01
2.73287356e-01 -6.89916536e-02 -9.69654083e-01 5.28063253e-02
8.78606081e-01 -9.76158530e-02 9.60549295e-01 9.98895168e-01
1.81687891e-01 -5.87911844e-01 -1.28555131e+00 9.78396475e-01
8.67114589e-02 -1.00309300e+00 -7.37203240e-01 5.33661306e-01
3.78894478e-01 4.48975891e-01 -4.10319492e-02 1.05197072e+00
1.56004965e+00 -1.12758434e+00 9.26110327e-01 4.85236585e-01
4.65382129e-01 -7.00089931e-01 4.46057200e-01 8.80831420e-01
-5.84318221e-01 -5.32768786e-01 -1.31925836e-01 -7.77320206e-01
-4.04241502e-01 -1.64901942e-01 -1.23866248e+00 -2.34997720e-01
3.61289620e-01 8.06864798e-01 -7.19930410e-01 6.51381195e-01
-7.69647121e-01 5.80892205e-01 -3.60606998e-01 -4.11644489e-01
6.17846310e-01 4.75320453e-03 4.21720773e-01 3.21187139e-01
-8.77906457e-02 2.07534254e-01 2.61413157e-01 9.56529856e-01
4.40899357e-02 -2.37238213e-01 -6.67317986e-01 -4.63461369e-01
3.41185741e-02 1.24796462e+00 -4.18166041e-01 -2.16584802e-02
-2.54460424e-01 6.67429745e-01 7.74815321e-01 -3.99427414e-02
-4.75386232e-01 1.36663631e-01 7.48924673e-01 7.60577247e-02
2.19192609e-01 -7.62443990e-02 2.40346640e-02 -8.63177836e-01
-4.98097628e-01 -1.28457475e+00 4.78686929e-01 -1.09304821e+00
-9.13777053e-01 4.81053591e-01 -3.26649278e-01 -1.01686049e+00
-1.17691505e+00 -6.76988959e-01 -3.09822649e-01 5.80945194e-01
-1.20643258e+00 -8.33709598e-01 -1.25681728e-01 5.29213309e-01
7.62578666e-01 -5.37294388e-01 7.72827268e-01 -4.57616359e-01
-4.86244351e-01 9.28862169e-02 -6.26857221e-01 2.80690521e-01
4.31868345e-01 -1.41724348e+00 3.82030278e-01 5.85786104e-01
1.33037046e-01 1.32885262e-01 7.85840631e-01 -6.70999050e-01
-9.00550425e-01 -6.49242103e-01 5.80995500e-01 -6.53247416e-01
5.89679778e-01 -2.07799882e-01 -1.17281929e-01 1.11776447e+00
3.41062814e-01 -2.77677506e-01 7.02908456e-01 4.59592223e-01
-1.73917457e-01 3.07773590e-01 -8.32282603e-01 9.67377126e-01
1.04516745e+00 -5.01234174e-01 -1.01832330e+00 2.32928827e-01
3.38729292e-01 -4.82393414e-01 -1.88767701e-01 -2.15842530e-01
3.12544465e-01 -1.35270548e+00 6.21201813e-01 -8.48354340e-01
3.29170078e-01 -3.54296803e-01 2.60376096e-01 -1.30519867e+00
-4.65198904e-01 -9.95538652e-01 -1.72952771e-01 4.94144589e-01
1.87439382e-01 -1.39886796e-01 9.88932729e-01 6.94307983e-01
-1.41238198e-02 -2.99414843e-01 -6.99255347e-01 -3.84137869e-01
-4.68246311e-01 -5.46051264e-01 3.17993373e-01 5.43848693e-01
3.87155265e-01 1.06072485e+00 -5.47664046e-01 -2.76747465e-01
8.87587368e-01 -1.53240576e-01 1.12612665e+00 -1.59155715e+00
-5.03924608e-01 -6.35849416e-01 -1.82598904e-01 -1.00984097e+00
4.34596121e-01 -6.31631792e-01 6.37084961e-01 -1.46941066e+00
-2.29860237e-03 -3.09615523e-01 -2.57431366e-03 1.12579107e+00
2.34299991e-02 6.44502789e-02 1.35755852e-01 1.70177147e-01
-1.43783653e+00 5.15292108e-01 1.11452889e+00 2.02408969e-01
-3.34957957e-01 9.91842672e-02 -1.03635991e+00 1.09382045e+00
7.97619581e-01 -7.25274146e-01 -4.89345282e-01 9.77175161e-02
8.45997393e-01 1.41360596e-01 3.19311261e-01 -1.26621938e+00
5.93971372e-01 -3.88131857e-01 -2.37734225e-02 -2.50454694e-01
3.35138589e-01 -6.06022418e-01 -2.97223836e-01 3.71185690e-01
-6.39434576e-01 -1.34675130e-01 1.23775765e-01 5.84279776e-01
3.15442860e-01 -5.01918852e-01 5.82093239e-01 -6.31801903e-01
-1.01354361e+00 7.60269016e-02 -9.39213097e-01 5.34913480e-01
1.12683654e+00 -1.11767083e-01 -3.60426098e-01 -8.04064810e-01
-5.79092920e-01 4.45674896e-01 4.10713613e-01 1.68203562e-01
1.60863951e-01 -7.53221750e-01 -5.93223453e-01 3.22197452e-02
1.54002652e-01 -4.39753272e-02 -1.34288624e-01 7.05825508e-01
-2.11160690e-01 3.22886884e-01 -5.83916008e-01 -1.45001754e-01
-1.21686351e+00 4.36770052e-01 5.51919103e-01 -7.72852838e-01
-6.10362530e-01 8.73558342e-01 2.11939901e-01 -4.76193756e-01
3.63420755e-01 -4.49089035e-02 -5.98264456e-01 9.85608995e-02
6.79300427e-01 2.23683134e-01 -2.22775236e-01 -5.74090064e-01
-5.68640977e-02 -2.75817454e-01 -1.52988642e-01 -3.47906440e-01
1.52987885e+00 5.00175692e-02 2.08122283e-01 3.02711308e-01
1.99146524e-01 -1.45108312e-01 -2.09383559e+00 -2.61684328e-01
3.43439877e-01 -3.23614590e-02 3.51482272e-01 -1.01096129e+00
-5.54066002e-01 6.69649184e-01 2.06074610e-01 1.86883762e-01
8.28857422e-01 -8.00082237e-02 4.14070636e-01 8.55898559e-01
7.07139730e-01 -1.34463263e+00 3.80989671e-01 8.89676690e-01
6.18873060e-01 -1.42840481e+00 -1.06755178e-02 4.51183170e-01
-1.18715107e+00 7.79397607e-01 9.68607008e-01 -1.04048684e-01
-1.74313765e-02 9.88317057e-02 -1.55826434e-01 -5.74738503e-01
-1.25005591e+00 -8.84281874e-01 1.51427299e-01 8.26420367e-01
8.66364222e-03 -1.42424749e-02 2.61707276e-01 8.46953630e-01
-4.66533303e-01 1.75422966e-01 6.86587751e-01 9.89346683e-01
-7.75883019e-01 -7.61807323e-01 4.19332366e-03 6.09774530e-01
-2.97933638e-01 -1.66816577e-01 -4.28757340e-01 7.54704237e-01
5.54940514e-02 1.10545766e+00 1.60180315e-01 -4.21035498e-01
1.08861625e-01 1.68189168e-01 6.64920330e-01 -1.12851739e+00
-7.46656060e-01 -1.14683673e-01 5.24623930e-01 -1.03925538e+00
-3.38994831e-01 -9.10619736e-01 -1.34871316e+00 -2.68970102e-01
-3.57945748e-02 -2.05788705e-02 2.18489438e-01 1.39777744e+00
2.41310090e-01 4.66908485e-01 5.12152612e-02 -9.77277398e-01
-5.96902609e-01 -1.00004268e+00 -5.01962364e-01 2.57062197e-01
2.94997066e-01 -8.47952008e-01 -4.10119742e-01 -1.32441774e-01] | [3.911484479904175, 1.4744102954864502] |
f16429fd-700f-47ad-a248-47b6dc63acd2 | monte-carlo-tree-search-with-sampled | 1704.05963 | null | http://arxiv.org/abs/1704.05963v1 | http://arxiv.org/pdf/1704.05963v1.pdf | Monte Carlo Tree Search with Sampled Information Relaxation Dual Bounds | Monte Carlo Tree Search (MCTS), most famously used in game-play artificial
intelligence (e.g., the game of Go), is a well-known strategy for constructing
approximate solutions to sequential decision problems. Its primary innovation
is the use of a heuristic, known as a default policy, to obtain Monte Carlo
estimates of downstream values for states in a decision tree. This information
is used to iteratively expand the tree towards regions of states and actions
that an optimal policy might visit. However, to guarantee convergence to the
optimal action, MCTS requires the entire tree to be expanded asymptotically. In
this paper, we propose a new technique called Primal-Dual MCTS that utilizes
sampled information relaxation upper bounds on potential actions, creating the
possibility of "ignoring" parts of the tree that stem from highly suboptimal
choices. This allows us to prove that despite converging to a partial decision
tree in the limit, the recommended action from Primal-Dual MCTS is optimal. The
new approach shows significant promise when used to optimize the behavior of a
single driver navigating a graph while operating on a ride-sharing platform.
Numerical experiments on a real dataset of 7,000 trips in New Jersey suggest
that Primal-Dual MCTS improves upon standard MCTS by producing deeper decision
trees and exhibits a reduced sensitivity to the size of the action space. | ['Warren B. Powell', 'Lina Al-Kanj', 'Daniel R. Jiang'] | 2017-04-20 | null | null | null | null | ['game-of-go'] | ['playing-games'] | [-1.50129214e-01 3.03162545e-01 -5.30198991e-01 -8.51583034e-02
-7.01323092e-01 -5.43084919e-01 2.03530297e-01 -3.91751155e-02
-4.97220904e-01 1.01470625e+00 8.57106224e-03 -9.59843040e-01
-4.10110176e-01 -1.07171953e+00 -5.88255525e-01 -7.55111516e-01
4.13756743e-02 9.47819233e-01 4.14406717e-01 -5.08099973e-01
3.40178519e-01 5.59767962e-01 -1.63733172e+00 -1.07581116e-01
9.70657945e-01 7.31645703e-01 1.32882223e-01 5.71075439e-01
-2.16178045e-01 5.85727870e-01 -4.71100807e-01 -6.21475816e-01
6.06728911e-01 -5.53890586e-01 -9.70633566e-01 1.28678799e-01
-2.49182239e-01 -3.69878829e-01 -3.84623647e-01 1.08163035e+00
2.42979318e-01 5.63006520e-01 3.95227075e-01 -1.50474083e+00
3.02146882e-01 7.02987850e-01 -6.98256552e-01 4.27358747e-01
2.24166110e-01 5.01309037e-01 9.89683270e-01 -5.45812119e-03
5.19766986e-01 1.17242658e+00 3.78280282e-01 5.09291112e-01
-1.26601350e+00 -4.09808576e-01 2.93049514e-01 6.01191282e-01
-1.45715761e+00 -4.72750850e-02 4.46382314e-01 1.45771271e-02
9.68436420e-01 3.79175156e-01 9.38249826e-01 4.08239990e-01
3.79708350e-01 8.74796987e-01 1.26625514e+00 -2.26322994e-01
7.69174218e-01 -7.75779481e-04 -5.28726242e-02 6.91607296e-01
2.80990034e-01 4.94105399e-01 -4.35558259e-01 -4.99750286e-01
2.48677790e-01 -2.62758225e-01 9.90769360e-03 -6.12873375e-01
-3.54217321e-01 1.17562878e+00 3.03705543e-01 -2.03947246e-01
-6.40898943e-01 2.93926716e-01 4.25245136e-01 2.05593124e-01
5.60099006e-01 1.39789939e-01 -2.18618453e-01 -7.36940980e-01
-7.98312247e-01 7.89311171e-01 1.01506007e+00 6.96733654e-01
8.36327434e-01 -9.23187062e-02 -3.25539678e-01 3.78317684e-01
-7.08739981e-02 3.03126067e-01 3.85318097e-04 -1.41421878e+00
6.30146682e-01 4.56928849e-01 7.25088120e-01 -5.07410884e-01
-3.80126029e-01 -4.71129417e-01 -2.59798884e-01 5.35757184e-01
7.03457952e-01 -4.32255566e-01 -8.19603622e-01 1.68024206e+00
6.45578623e-01 1.34667352e-01 -2.23199427e-01 7.91121244e-01
-3.33350331e-01 7.54245937e-01 -1.24416508e-01 -3.08421820e-01
1.04599822e+00 -7.07607687e-01 -5.06917894e-01 -4.69374478e-01
9.54589725e-01 -7.00262636e-02 8.28546226e-01 5.84253907e-01
-1.27907479e+00 -8.89511257e-02 -8.83146763e-01 4.80440170e-01
-2.86210030e-01 -5.65213919e-01 8.79883826e-01 1.09496486e+00
-9.87717867e-01 6.42163515e-01 -8.27114642e-01 -7.61594772e-02
5.35186648e-01 3.61579329e-01 1.40446171e-01 -4.61255848e-01
-1.06203866e+00 1.21338296e+00 5.22338867e-01 6.59921616e-02
-9.86847579e-01 -5.23189664e-01 -6.08595610e-01 3.27966362e-01
1.27708149e+00 -5.90224624e-01 1.44942677e+00 -4.91959482e-01
-1.51087403e+00 3.58466387e-01 -3.66962880e-01 -9.34423745e-01
6.68392360e-01 2.59236962e-01 6.90684244e-02 -8.08254927e-02
2.60989964e-01 3.95164490e-01 4.16133434e-01 -8.37284863e-01
-1.13485777e+00 -4.66235548e-01 4.32495683e-01 6.02510691e-01
1.96372896e-01 -2.24902421e-01 -4.72686380e-01 6.96835741e-02
-7.84220323e-02 -1.30607080e+00 -9.88432944e-01 -4.89886880e-01
-2.41243303e-01 -5.98737955e-01 1.09170988e-01 -2.75180787e-01
1.42830133e+00 -1.71143174e+00 1.24747321e-01 6.50448740e-01
-8.52600709e-02 1.04840323e-02 1.47415295e-01 6.18580401e-01
4.48384047e-01 6.77911565e-02 -2.64995247e-01 -1.68295562e-01
1.64095700e-01 8.38813007e-01 -9.72572267e-02 6.39832914e-01
-4.71159041e-01 7.23062158e-01 -9.64048266e-01 -2.81370521e-01
3.75890136e-01 -5.13316810e-01 -8.76110494e-01 -2.08347559e-01
-4.00068969e-01 3.27544883e-02 -6.46130621e-01 2.17020959e-01
7.15711176e-01 2.10068181e-01 1.86439767e-01 6.03250921e-01
-2.41176754e-01 3.90993983e-01 -1.50105286e+00 1.31277108e+00
-4.44156647e-01 3.39785933e-01 1.15955293e-01 -1.20063758e+00
3.92534167e-01 -1.75100312e-01 5.01289368e-01 -6.68490529e-01
2.24437967e-01 2.46873498e-02 -5.61581403e-02 -2.96029389e-01
6.53726816e-01 -3.36329699e-01 -3.42968583e-01 5.47968030e-01
-5.22903979e-01 -2.91439295e-01 5.43475389e-01 3.17334771e-01
1.09962416e+00 -2.22863555e-02 2.59064764e-01 -3.11607599e-01
4.23878729e-01 5.86816311e-01 6.76434934e-01 1.11793363e+00
-3.54686260e-01 -1.73682064e-01 1.01060426e+00 -3.27554286e-01
-7.81824708e-01 -9.11590457e-01 6.37677237e-02 1.05158687e+00
4.17919338e-01 -3.27017337e-01 -9.00830567e-01 -6.06750727e-01
1.75819352e-01 1.37117386e+00 -7.16648996e-01 -3.39464724e-01
-4.93518978e-01 -7.28271246e-01 1.05166294e-01 2.90920347e-01
6.95807338e-01 -6.47708118e-01 -8.22367728e-01 5.06105661e-01
-3.75338435e-01 -9.03495550e-01 -5.00002444e-01 1.60303086e-01
-9.04825747e-01 -1.07016754e+00 -3.81485015e-01 -3.39878164e-02
5.29872298e-01 3.89540464e-01 8.83601964e-01 -5.41307367e-02
9.21571851e-02 2.72220910e-01 -8.59089047e-02 -3.01110983e-01
-5.66589952e-01 2.00118393e-01 -2.25142501e-02 -1.02529094e-01
5.40443420e-01 -2.76759893e-01 -5.42503357e-01 4.98236001e-01
-4.21544909e-01 -1.36086727e-02 3.87510270e-01 7.61242747e-01
6.46512091e-01 7.62774408e-01 3.33972573e-01 -8.89530480e-01
8.57313275e-01 -6.81353390e-01 -1.14317679e+00 7.15143532e-02
-8.95780444e-01 4.61552382e-01 6.09037519e-01 -1.88965157e-01
-1.13738012e+00 -7.12219551e-02 -3.60017014e-03 -2.02468783e-01
-3.47901955e-02 6.39755487e-01 1.22975037e-02 -9.77400970e-03
6.53827190e-01 1.30203739e-01 1.03860341e-01 -2.68845171e-01
5.66896856e-01 4.74098057e-01 4.82019424e-01 -6.18273795e-01
4.89923865e-01 5.94206631e-01 3.53911430e-01 -4.17532027e-01
-7.42236972e-01 -5.96313119e-01 -1.38901561e-01 -4.90207255e-01
5.05956888e-01 -4.90579724e-01 -1.23200011e+00 6.85480535e-02
-4.40099627e-01 -5.18120527e-01 -4.48402941e-01 2.70107627e-01
-9.22548056e-01 1.55124485e-01 -1.58379927e-01 -1.31847084e+00
3.31647485e-01 -9.98615563e-01 4.71586943e-01 4.80589122e-01
1.25955582e-01 -7.27260709e-01 3.62148285e-02 5.91853738e-01
1.63140953e-01 1.81032181e-01 6.10288739e-01 -2.61975110e-01
-5.83029330e-01 -2.94377923e-01 3.27572852e-01 -6.19647354e-02
-1.79345965e-01 -2.51167387e-01 -3.75048429e-01 -3.34728211e-01
-7.43216053e-02 8.93688276e-02 5.97907186e-01 7.17841446e-01
1.04522002e+00 -5.58533192e-01 -6.54943526e-01 3.06142062e-01
1.55698073e+00 5.37387311e-01 5.43939948e-01 5.13824224e-01
8.14721659e-02 6.40730441e-01 1.12118196e+00 6.57082260e-01
5.29396832e-01 7.57677615e-01 5.69026470e-01 2.36664787e-01
4.82017547e-01 -4.37158108e-01 3.06848228e-01 -2.69533396e-01
-6.64271861e-02 -2.69883126e-01 -8.62003267e-01 7.80483902e-01
-1.92730641e+00 -1.15162790e+00 -7.68087879e-02 2.52894616e+00
5.18362701e-01 7.99322963e-01 7.24590003e-01 2.63251603e-01
5.19416094e-01 -1.57837391e-01 -1.01741755e+00 -1.15535963e+00
3.68826419e-01 2.35539213e-01 1.38586473e+00 6.31536543e-01
-4.17516023e-01 1.00256300e+00 6.51266050e+00 1.24287927e+00
-4.33309734e-01 1.05208457e-01 6.85995460e-01 -4.94843990e-01
-2.59564996e-01 2.87060469e-01 -9.05264258e-01 4.82600331e-01
1.42314720e+00 -7.50928104e-01 1.00141084e+00 9.09884751e-01
7.20366180e-01 -1.03486776e+00 -7.30030179e-01 5.22758663e-01
-4.35930789e-01 -1.45011818e+00 -3.93575341e-01 5.79345345e-01
9.08590376e-01 2.46397331e-02 -7.95373693e-02 5.20305634e-01
1.05931616e+00 -7.34223366e-01 7.25226521e-01 -1.57928392e-02
3.02414447e-01 -1.67807758e+00 7.78017044e-01 8.58135939e-01
-1.11680567e+00 -6.05093718e-01 -3.16243291e-01 -2.92765796e-01
5.53107381e-01 2.74172872e-01 -9.81228828e-01 5.36093652e-01
6.12030566e-01 7.84759074e-02 -1.14742063e-01 1.12100661e+00
-2.55413711e-01 7.90367484e-01 -6.91032231e-01 -3.35993439e-01
8.20099175e-01 -6.10709965e-01 5.92505991e-01 6.59403741e-01
-4.95332144e-02 3.45125943e-01 2.59799391e-01 8.31995904e-01
4.40930843e-01 -3.24262559e-01 -6.89245284e-01 1.68020114e-01
6.20239913e-01 7.08822191e-01 -8.58419955e-01 -3.29278767e-01
-2.22401530e-01 5.76173484e-01 1.47465974e-01 3.75669867e-01
-1.07248497e+00 -2.53534675e-01 1.03980064e+00 1.20973811e-01
4.16846991e-01 -2.29186162e-01 -6.34445250e-01 -3.42527837e-01
-3.89223546e-02 -6.67201579e-01 6.13945603e-01 -5.10611057e-01
-5.12599707e-01 5.13800867e-02 5.73764265e-01 -9.67355132e-01
-7.36160696e-01 -1.70647204e-01 -7.39378154e-01 1.03552449e+00
-1.31952477e+00 -1.70271382e-01 1.50857314e-01 4.57439840e-01
5.91509581e-01 2.58483768e-01 9.29201990e-02 -1.04663014e-01
-5.60470223e-01 4.32039529e-01 3.40192616e-01 -5.87677240e-01
-2.53052652e-01 -1.32817268e+00 5.62591314e-01 9.56696332e-01
-1.98159516e-01 6.72958717e-02 1.18638802e+00 -6.43065810e-01
-1.36934888e+00 -5.34506381e-01 4.56163943e-01 1.14425318e-02
6.88416779e-01 -9.77861285e-02 -5.37220180e-01 5.72436690e-01
-1.28977582e-01 -5.17893076e-01 2.24469379e-01 1.25479937e-01
5.08992374e-01 -1.42692149e-01 -1.36372125e+00 9.50110435e-01
1.07372439e+00 -2.02682897e-01 -3.21329534e-01 3.25202882e-01
2.17627615e-01 -6.12488151e-01 -4.25363868e-01 -1.33082673e-01
1.60248473e-01 -1.11747742e+00 7.22903371e-01 -6.56199872e-01
-2.51587629e-01 -2.59337068e-01 3.41222614e-01 -1.59148920e+00
-1.20397888e-01 -1.18001771e+00 -2.44122948e-02 5.73331594e-01
4.29297358e-01 -9.52226281e-01 1.26600826e+00 9.20703173e-01
-1.89813897e-01 -1.10698938e+00 -1.53700399e+00 -1.02691925e+00
2.21748441e-01 -6.23059809e-01 7.39362657e-01 9.57982913e-02
4.36090007e-02 -1.26442492e-01 -1.98022485e-01 2.24901497e-01
9.19508934e-01 1.42856210e-01 8.44530642e-01 -8.63679051e-01
-3.15579325e-01 -5.42095721e-01 -1.10778786e-01 -1.44483626e+00
1.69325560e-01 -6.74017847e-01 2.26426437e-01 -1.69242382e+00
-1.33998170e-01 -5.72353661e-01 6.67269081e-02 2.71549165e-01
-4.04786412e-03 -3.06075215e-01 2.82684993e-02 -2.11435482e-01
-3.69270772e-01 4.91686851e-01 1.26455319e+00 6.56426847e-02
-4.28087384e-01 6.67174160e-01 -8.24121296e-01 5.86488843e-01
9.27384734e-01 -7.90516436e-01 -5.96346915e-01 1.53304473e-01
3.35328549e-01 5.78118205e-01 -5.95434904e-05 -7.02998817e-01
3.26685935e-01 -8.23282361e-01 -5.08188367e-01 -9.85451162e-01
3.94949526e-01 -6.81677997e-01 3.92138243e-01 9.17599201e-01
-2.23416284e-01 1.34953316e-02 2.15978906e-01 9.87115204e-01
1.73938736e-01 -5.76769173e-01 7.12719858e-01 -1.49318516e-01
-8.15866172e-01 2.63551146e-01 -9.92821693e-01 1.54293984e-01
1.32417870e+00 -6.06258512e-01 8.51347744e-02 -6.38738275e-01
-7.49150395e-01 9.64496374e-01 2.87131995e-01 -3.32903527e-02
2.49955416e-01 -1.04698622e+00 -5.95229149e-01 -1.99725240e-01
-1.41773447e-01 -6.43693879e-02 3.86925727e-01 7.98602223e-01
-4.79672909e-01 5.45211613e-01 4.31924313e-02 -3.11972022e-01
-1.09845889e+00 7.04247773e-01 6.24362409e-01 -7.00986028e-01
-6.91407859e-01 5.86310387e-01 -2.29498997e-01 9.45980400e-02
3.91981713e-02 -3.17490876e-01 1.47685856e-01 -1.27906352e-01
4.25098747e-01 8.43224764e-01 -1.47368880e-02 -2.43759096e-01
-1.85353845e-01 5.62646426e-02 1.46116465e-01 -5.99135816e-01
9.57045972e-01 -4.43442702e-01 3.21811646e-01 1.18730910e-01
7.19574392e-01 -2.86516368e-01 -1.43877172e+00 -9.72237140e-02
1.34861737e-01 -7.14132667e-01 3.79729152e-01 -5.95165014e-01
-1.13970184e+00 3.63461137e-01 2.43068904e-01 6.09728634e-01
9.83353734e-01 -1.89244404e-01 8.07373643e-01 4.58105445e-01
8.65450621e-01 -1.47699249e+00 -6.83770657e-01 2.93352664e-01
2.54287273e-01 -8.12403560e-01 1.89770162e-02 -1.74862102e-01
-8.72765303e-01 6.52324378e-01 4.47802454e-01 -1.06120341e-01
4.28960085e-01 7.34289587e-02 -5.93150437e-01 -1.38519987e-01
-1.09462428e+00 -6.22177720e-01 -3.01839501e-01 5.11199355e-01
-5.90621352e-01 3.83764654e-01 -7.40120232e-01 1.10559762e-01
-3.08362812e-01 1.63319975e-01 9.34643924e-01 1.01695108e+00
-5.24973273e-01 -9.31061924e-01 -4.01536971e-01 7.41410911e-01
-1.77768975e-01 1.99025437e-01 1.13690610e-03 8.80124688e-01
3.66322696e-02 1.30050814e+00 1.34400472e-01 -8.50332603e-02
3.62121880e-01 -7.51849078e-03 5.04576504e-01 -3.84258807e-01
-5.54199278e-01 -3.04437280e-01 5.70900738e-01 -8.84835422e-01
-6.51775673e-02 -9.49967027e-01 -1.21164036e+00 -9.97060537e-01
-5.30645669e-01 6.52251422e-01 6.49472594e-01 1.18356085e+00
1.75736725e-01 3.06129217e-01 8.45088005e-01 -5.54349899e-01
-8.49928558e-01 -4.68603998e-01 -7.27027535e-01 -6.88867643e-02
3.43195647e-02 -9.50125277e-01 -6.06707811e-01 -7.71265745e-01] | [4.30506706237793, 2.3433966636657715] |
91c2226a-a3e7-4850-afbd-763b455d30be | unsupervised-data-augmentation-for-aspect | null | null | https://aclanthology.org/2022.coling-1.586 | https://aclanthology.org/2022.coling-1.586.pdf | Unsupervised Data Augmentation for Aspect Based Sentiment Analysis | Recent approaches to Aspect-based Sentiment Analysis (ABSA) take a co-extraction approach to this span-level classification task, performing the subtasks of aspect term extraction (ATE) and aspect sentiment classification (ASC) simultaneously. In this work, we build on recent progress in applying pre-training to this co-extraction task with the introduction of an adaptation of Unsupervised Data Augmentation in semi-supervised learning. As originally implemented, UDA cannot accommodate span-level classification since it relies on advanced data augmentation techniques, such as back-translation, that alter the sequence lengths of the original data and cause index mismatches. We introduce an adaptation of UDA using Masked Language Model (MLM) unmasking that accommodates this index-match constraint and test the approach on standard ABSA benchmark datasets. We show that simple augmentations applied to modest-sized datasets along with consistency training lead to competitive performance with the current ABSA state-of-the-art in the restaurant and laptop domains using only 75% of the training data. | ['Sahil Badyal', 'Adam Faulkner', 'David Z. Chen'] | null | null | null | null | coling-2022-10 | ['term-extraction', 'aspect-based-sentiment-analysis'] | ['natural-language-processing', 'natural-language-processing'] | [ 7.94251084e-01 2.04720590e-02 -1.53766572e-01 -6.66115284e-01
-1.23866224e+00 -9.09792483e-01 9.02529836e-01 6.17115140e-01
-4.95066941e-01 3.21944118e-01 3.17521363e-01 -7.49395728e-01
3.51929039e-01 -6.57506406e-01 -6.44045055e-01 -2.96240777e-01
5.34335487e-02 6.80943191e-01 1.14128746e-01 -6.33676350e-01
2.30793998e-01 3.13293189e-02 -1.62273431e+00 8.94942880e-01
2.89826959e-01 9.20730114e-01 -6.14827931e-01 8.38969350e-01
-5.35599589e-01 5.23746192e-01 -6.19951725e-01 -6.40314400e-01
4.95150089e-01 -3.00319195e-01 -9.16770220e-01 2.87621170e-01
4.05655652e-01 1.82382017e-01 4.38193023e-01 4.86001134e-01
3.17014903e-01 -6.52145743e-02 4.95674372e-01 -1.25838685e+00
-2.69048125e-01 6.65296078e-01 -6.70811236e-01 4.02194299e-02
4.40677404e-01 -1.89967960e-01 1.38134944e+00 -9.68602896e-01
4.79848891e-01 8.45515370e-01 1.01534462e+00 2.42118940e-01
-1.48959434e+00 -7.09019154e-02 2.99229354e-01 -2.44502783e-01
-9.57085609e-01 -6.13686085e-01 5.60731053e-01 -2.90747017e-01
1.80974042e+00 5.53262889e-01 7.25940168e-01 7.82416582e-01
-6.85742199e-02 9.64435101e-01 1.33605838e+00 -8.54574561e-01
3.62085521e-01 2.67254800e-01 5.18387616e-01 4.85657305e-01
4.00416255e-01 -2.19014168e-01 -7.62511730e-01 -6.16629481e-01
-8.39332491e-02 -3.51853937e-01 2.92379200e-01 -4.94369954e-01
-1.17476153e+00 8.67122769e-01 -2.77069062e-01 1.16913043e-01
-2.48159721e-01 -3.60208482e-01 8.96108627e-01 6.62877381e-01
7.18412220e-01 7.21781492e-01 -1.24431539e+00 -2.85129935e-01
-1.13139868e+00 3.00132364e-01 1.29949951e+00 1.21985853e+00
7.26136982e-01 4.83311936e-02 -2.06836581e-01 7.21080065e-01
5.26455417e-02 2.65136391e-01 6.69624507e-01 -3.66123617e-01
4.78041589e-01 8.93824697e-01 -1.59455135e-01 -3.50295156e-01
-4.95423883e-01 -3.81807327e-01 -2.01648861e-01 2.53670603e-01
3.92338753e-01 -1.65473834e-01 -1.25525713e+00 1.38832510e+00
6.28151476e-01 -2.76594728e-01 1.67849720e-01 2.53619045e-01
4.52248454e-01 1.82670057e-01 -8.97997990e-02 -1.61037385e-01
1.94703364e+00 -1.18548334e+00 -4.57399458e-01 -4.85057026e-01
1.08480620e+00 -1.26934218e+00 1.22657967e+00 5.01556456e-01
-9.53013062e-01 -1.84703171e-01 -1.51779974e+00 -1.68199614e-01
-9.82786298e-01 -6.01664931e-02 9.78045285e-01 9.90587592e-01
-9.46436584e-01 2.67749846e-01 -8.63273859e-01 -3.17364395e-01
3.38739544e-01 3.89746308e-01 -3.24713647e-01 2.79614944e-02
-7.84961700e-01 6.06498003e-01 2.32604146e-01 -4.02542561e-01
-1.97013542e-01 -9.96679783e-01 -1.19390249e+00 -3.88992369e-01
6.35598719e-01 -8.00892472e-01 1.60148382e+00 -1.04476118e+00
-1.36067462e+00 1.22951651e+00 -2.89886445e-01 -4.82108176e-01
8.96243975e-02 -4.25535023e-01 -4.36505288e-01 -3.19436818e-01
2.04916224e-01 2.55991817e-01 1.03411055e+00 -1.01302743e+00
-7.07987368e-01 -4.82818961e-01 3.01399916e-01 1.69697836e-01
-2.99395084e-01 3.78639042e-01 -3.48742038e-01 -9.80630577e-01
-7.28800446e-02 -1.13737404e+00 -4.68659997e-01 -5.87371588e-01
-4.40445065e-01 5.55650517e-02 8.12485695e-01 -4.12348300e-01
1.27825093e+00 -1.98226666e+00 -1.41348600e-01 3.49872857e-01
-1.37721285e-01 2.98314154e-01 -3.94807786e-01 6.04894042e-01
-3.68516326e-01 -1.09076165e-02 -6.41236961e-01 -8.56612682e-01
6.42588884e-02 2.01841310e-01 -4.11952943e-01 1.33537203e-01
3.63622725e-01 9.00140703e-01 -5.67829430e-01 -3.06104839e-01
-2.31146783e-01 2.36187235e-01 -8.68666530e-01 1.77238584e-01
-5.97140133e-01 -9.36829522e-02 -1.05902649e-01 8.17498028e-01
5.26973844e-01 -1.29976720e-01 2.10301310e-01 -2.52489205e-02
-1.30649343e-01 1.00511932e+00 -1.20029950e+00 1.89193475e+00
-6.59172177e-01 4.16527033e-01 1.31756188e-02 -9.15474117e-01
6.95049107e-01 2.56465316e-01 5.64364552e-01 -4.13668245e-01
4.80702445e-02 3.02367270e-01 -6.23717122e-02 -1.76651523e-01
8.47828627e-01 -2.08704725e-01 -4.78385091e-01 8.66346359e-01
1.73150614e-01 -2.88994431e-01 5.34983575e-01 2.40111202e-01
1.17908823e+00 3.65561813e-01 6.74821258e-01 -3.69857997e-01
6.87538922e-01 5.61699331e-01 2.89385974e-01 6.12047672e-01
2.44517922e-01 7.68025398e-01 6.02513254e-01 -6.07047796e-01
-1.34632456e+00 -4.70524997e-01 -7.77391344e-02 1.54318690e+00
-6.91403568e-01 -1.13164854e+00 -7.63098240e-01 -1.15522933e+00
-1.16277598e-01 8.45886528e-01 -6.73486352e-01 6.07238486e-02
-6.48122072e-01 -1.32016122e+00 4.44612145e-01 5.21330059e-01
7.84921125e-02 -9.38635528e-01 -4.42496270e-01 2.09442228e-01
2.04078063e-01 -1.38001180e+00 -4.82725680e-01 6.73913360e-01
-7.21599400e-01 -9.23765063e-01 -1.62268549e-01 -6.63464427e-01
5.06108463e-01 1.24615349e-01 1.70426011e+00 -1.12646108e-03
-3.75479460e-01 2.58281142e-01 -5.24048448e-01 -6.80746496e-01
-5.57527244e-01 6.10667229e-01 -1.38183624e-01 -6.53276592e-02
9.84557629e-01 -7.73736477e-01 -2.74650574e-01 3.61329801e-02
-1.31370890e+00 7.68106803e-02 6.57018363e-01 6.81819618e-01
7.15482116e-01 -3.04515451e-01 3.12119782e-01 -1.71059704e+00
5.05454063e-01 -2.77027160e-01 -5.37366807e-01 6.39831871e-02
-1.08198643e+00 2.71637738e-01 5.23814797e-01 -2.81328529e-01
-6.43402219e-01 2.14620203e-01 -3.70376647e-01 1.15287162e-01
-1.79393038e-01 7.35585153e-01 -2.42539570e-01 1.65186465e-01
5.88967621e-01 2.27659747e-01 -7.94495568e-02 -5.67249477e-01
5.43966115e-01 7.90836215e-01 1.28489971e-01 -3.82954210e-01
7.99589753e-01 4.98409122e-01 1.59156602e-02 -6.93523109e-01
-1.17010009e+00 -7.36800671e-01 -8.19893479e-01 4.45513546e-01
5.11823833e-01 -1.03184485e+00 -7.10064098e-02 4.36593950e-01
-7.60630548e-01 -2.25109421e-02 -7.02719450e-01 5.94931431e-02
-6.01036370e-01 2.33803585e-01 -4.00892615e-01 -5.20304501e-01
-6.81501806e-01 -9.90808427e-01 1.39986217e+00 -3.14509630e-01
-6.16842568e-01 -7.29692578e-01 4.31387812e-01 6.09151423e-01
5.26609838e-01 1.33483231e-01 1.08433795e+00 -1.38653386e+00
-3.22036415e-01 -4.68277991e-01 2.37339959e-01 3.04108232e-01
2.44620860e-01 -2.99858421e-01 -1.15482950e+00 -2.37706766e-01
5.01378104e-02 -2.94154078e-01 8.03074718e-01 -8.78731459e-02
5.50578117e-01 -3.74194652e-01 9.69814658e-02 6.13566339e-01
1.26369619e+00 -2.42532060e-01 3.19567978e-01 9.11634862e-01
6.03718042e-01 4.96172726e-01 7.07124114e-01 3.07253033e-01
4.30160671e-01 7.01297343e-01 9.44108889e-02 -9.82526839e-02
-1.32423967e-01 -9.58429649e-02 5.81815839e-01 1.01367652e+00
2.83698738e-01 1.39302880e-01 -7.29956329e-01 7.51227200e-01
-1.66742373e+00 -4.69898939e-01 -1.16173029e-01 2.16643620e+00
1.16294551e+00 5.23171961e-01 5.23109257e-01 5.02712786e-01
-8.77110958e-02 3.94241273e-01 -3.26307029e-01 -1.05574477e+00
-1.53531373e-01 4.89735126e-01 5.42806029e-01 5.58708310e-01
-1.43443465e+00 8.19786549e-01 6.40963030e+00 6.37804151e-01
-6.96449757e-01 2.96627253e-01 4.85433966e-01 -2.26343304e-01
-5.99646688e-01 2.51681626e-01 -8.86511862e-01 -3.89319547e-02
1.30065656e+00 5.92518933e-02 4.54941630e-01 1.08914769e+00
-3.36009026e-01 3.54540854e-04 -1.21389723e+00 5.31342983e-01
4.72922653e-01 -1.13224018e+00 1.78193733e-01 -4.39137556e-02
8.36378038e-01 1.72989488e-01 -4.47600447e-02 6.07575178e-01
1.91820472e-01 -6.56801403e-01 4.53840703e-01 1.49335647e-02
7.04192102e-01 -7.29756534e-01 8.28650236e-01 5.22401594e-02
-1.23612273e+00 2.76308805e-01 -2.41850354e-02 -1.42094955e-01
7.62938485e-02 7.41645753e-01 -9.47040081e-01 6.79424286e-01
4.52284276e-01 6.25028849e-01 -7.65884757e-01 4.08962727e-01
-7.12096915e-02 5.84912896e-01 -4.11906421e-01 1.70830786e-01
2.56111592e-01 -9.57155451e-02 7.89723754e-01 1.46422327e+00
-1.20551087e-01 -4.74717706e-01 2.26934671e-01 1.57789260e-01
-2.76730862e-02 4.63068157e-01 -6.40086830e-01 -1.89216673e-01
-1.43508822e-01 1.38207483e+00 -6.78213716e-01 -5.87522447e-01
-9.06244934e-01 8.20101202e-01 2.45105907e-01 3.59401554e-02
-3.77464414e-01 -3.87086093e-01 9.21605885e-01 1.37500986e-01
8.16284657e-01 -1.65355399e-01 -6.34883344e-01 -1.37146807e+00
3.14812541e-01 -1.56294262e+00 8.00256133e-01 -2.89162368e-01
-1.16126406e+00 9.09529209e-01 -7.78004900e-02 -1.30311131e+00
-6.96859300e-01 -6.51876271e-01 -4.06490982e-01 8.15332651e-01
-1.60983658e+00 -1.54237866e+00 3.15560102e-01 2.95122236e-01
7.60916471e-01 -3.27842832e-01 1.26448977e+00 2.19246730e-01
-1.90183505e-01 8.63133371e-01 -1.75213963e-01 6.33650320e-03
6.85024619e-01 -1.54286611e+00 1.01899004e+00 9.39912379e-01
5.38300216e-01 7.82683730e-01 8.34798276e-01 -5.43352365e-01
-1.60622442e+00 -1.04722238e+00 1.24964666e+00 -1.07840073e+00
1.03692913e+00 -9.48966205e-01 -7.97045827e-01 9.85808909e-01
4.35593843e-01 -3.42187807e-02 1.17283332e+00 6.94387257e-01
-8.04297268e-01 -4.96612713e-02 -8.05916131e-01 4.90283757e-01
7.57427096e-01 -7.54467130e-01 -8.34516168e-01 2.97757596e-01
9.28406060e-01 -3.60932410e-01 -7.74955690e-01 5.16253531e-01
5.21141291e-01 -7.28940904e-01 7.97170341e-01 -1.14051437e+00
4.02329057e-01 -3.92101407e-01 -3.77808690e-01 -1.10883236e+00
2.38594532e-01 -8.08317304e-01 -3.02450448e-01 1.50390196e+00
9.61508632e-01 -4.65616286e-01 7.90150046e-01 5.95664263e-01
-6.31127954e-02 -9.77838516e-01 -7.87287116e-01 -4.55538481e-01
-1.06403612e-01 -7.55228937e-01 7.53378928e-01 8.09790969e-01
2.07390770e-01 8.14711928e-01 -1.62167981e-01 1.53792817e-02
3.65507215e-01 5.94339430e-01 9.16115642e-01 -7.92289138e-01
-5.67416251e-01 -2.85086960e-01 -3.38283151e-01 -6.16277635e-01
-1.54581696e-01 -9.27960336e-01 -1.16453908e-01 -1.13328886e+00
1.77025318e-01 -2.24236771e-01 -2.81186730e-01 7.25397170e-01
-4.17882323e-01 5.28586745e-01 -1.94607731e-02 6.35044426e-02
-6.29360676e-01 2.84088403e-01 4.41900522e-01 -1.62346974e-01
-3.55416834e-01 3.26089770e-01 -1.00946045e+00 7.49686718e-01
7.46464014e-01 -6.87590659e-01 -3.93278480e-01 -2.56001681e-01
6.92132890e-01 -6.40053272e-01 -3.26651543e-01 -7.35769689e-01
-1.36235341e-01 3.23772281e-01 2.78357845e-02 -6.35597527e-01
2.60352731e-01 -8.58931839e-01 -3.59205067e-01 1.71974242e-01
-4.54328179e-01 6.42479241e-01 6.68521821e-01 3.70781690e-01
-2.19435349e-01 -9.87978354e-02 2.84568846e-01 -1.05361357e-01
-3.70580733e-01 -4.20745052e-02 -5.38957715e-01 3.14054430e-01
7.12689877e-01 1.13250509e-01 -2.33933285e-01 -1.05248637e-01
-5.72490692e-01 -1.83350950e-01 6.06178224e-01 5.32431483e-01
3.37954573e-02 -9.74076092e-01 -5.61629295e-01 6.27109706e-01
5.07742167e-01 -1.15134165e-01 -2.49078825e-01 8.39050710e-01
-1.96522355e-01 4.92120892e-01 2.17715338e-01 -3.31042171e-01
-1.49325430e+00 8.03263128e-01 -1.63438916e-01 -8.32644939e-01
-3.60165179e-01 5.35164654e-01 -2.79002547e-01 -8.81614208e-01
-1.04434043e-01 -3.28864366e-01 -2.04610437e-01 2.37381712e-01
6.18800163e-01 -1.99479491e-01 8.55366111e-01 -4.62961048e-01
-4.97260153e-01 3.29413235e-01 -4.26487267e-01 -4.47977155e-01
1.43828762e+00 -3.76077704e-02 -2.06385747e-01 6.15904391e-01
1.19668853e+00 4.98170763e-01 -4.72772449e-01 -4.49638247e-01
3.78667414e-01 -1.45435572e-01 -9.00177956e-02 -9.17748749e-01
-7.45143771e-01 5.20230651e-01 7.19725341e-02 3.88189048e-01
1.27404690e+00 -6.14322051e-02 7.78341889e-01 5.76044798e-01
1.15879871e-01 -1.04681599e+00 -2.89006442e-01 6.63277447e-01
5.12969136e-01 -1.30679071e+00 4.08567488e-01 -4.16283041e-01
-6.80043161e-01 6.96158588e-01 3.21817070e-01 -3.24635580e-02
8.22483599e-01 7.13422418e-01 5.58984756e-01 -1.94092676e-01
-1.06935859e+00 -3.71217191e-01 3.95754099e-01 3.02317411e-01
6.08134031e-01 -1.05869755e-01 -4.05487925e-01 6.58925414e-01
-5.59877813e-01 -2.53985435e-01 4.01940912e-01 1.64002776e+00
7.57624358e-02 -1.57862163e+00 -2.08918065e-01 6.33737266e-01
-9.23077345e-01 -7.67200172e-01 -3.74723464e-01 7.86210895e-01
-2.44510472e-01 8.52200747e-01 -1.26141459e-01 -3.52981418e-01
4.56342399e-01 6.15912914e-01 1.86103925e-01 -8.94478977e-01
-1.00426662e+00 2.29411453e-01 7.10396588e-01 -5.04845917e-01
-6.78199351e-01 -1.05591500e+00 -8.34170401e-01 4.31177206e-02
-1.88655868e-01 2.26066440e-01 9.15997088e-01 1.21622527e+00
6.16616547e-01 5.81474066e-01 5.56585908e-01 -4.52707976e-01
-4.43249673e-01 -1.07209694e+00 -2.41794467e-01 5.95293939e-01
4.63980734e-01 -1.46239087e-01 -2.42194623e-01 2.32124433e-01] | [11.379034042358398, 6.8027238845825195] |
a872f570-cba2-4635-8599-05a3a675e3bd | stars-are-all-you-need-a-distantly-supervised | 2305.01710 | null | https://arxiv.org/abs/2305.01710v1 | https://arxiv.org/pdf/2305.01710v1.pdf | Stars Are All You Need: A Distantly Supervised Pyramid Network for Document-Level End-to-End Sentiment Analysis | In this paper, we propose document-level end-to-end sentiment analysis to efficiently understand aspect and review sentiment expressed in online reviews in a unified manner. In particular, we assume that star rating labels are a "coarse-grained synthesis" of aspect ratings across in the review. We propose a Distantly Supervised Pyramid Network (DSPN) to efficiently perform Aspect-Category Detection, Aspect-Category Sentiment Analysis, and Rating Prediction using only document star rating labels for training. By performing these three related sentiment subtasks in an end-to-end manner, DSPN can extract aspects mentioned in the review, identify the corresponding sentiments, and predict the star rating labels. We evaluate DSPN on multi-aspect review datasets in English and Chinese and find that with only star rating labels for supervision, DSPN can perform comparably well to a variety of benchmark models. We also demonstrate the interpretability of DSPN's outputs on reviews to show the pyramid structure inherent in document level end-to-end sentiment analysis. | ['John P. Lalor', 'Yixing Chen', 'Wenchang Li'] | 2023-05-02 | null | null | null | null | ['aspect-category-detection'] | ['natural-language-processing'] | [ 7.29579628e-02 2.20228612e-01 -5.47549367e-01 -9.90284503e-01
-1.07085431e+00 -8.85634661e-01 6.28990889e-01 2.81224251e-01
-2.02589005e-01 3.61597203e-02 7.46290624e-01 -4.91996884e-01
3.13120395e-01 -6.43263817e-01 -3.76189530e-01 -7.34670907e-02
5.02204537e-01 3.82296920e-01 -3.09988469e-01 -5.72587788e-01
5.17320216e-01 -7.42456838e-02 -1.35534286e+00 9.33875144e-01
6.53862059e-01 1.25717843e+00 -2.70696580e-01 7.87979364e-01
-3.02560836e-01 1.00087953e+00 -7.62748063e-01 -1.01267982e+00
2.31571585e-01 -1.34564474e-01 -5.33951879e-01 5.28958678e-01
3.68350357e-01 -1.27498314e-01 3.39205682e-01 9.67198908e-01
1.18453547e-01 -3.75685841e-02 9.62043822e-01 -7.62767375e-01
-1.03722882e+00 7.75534332e-01 -8.47873986e-01 -2.16331482e-01
3.36898804e-01 -2.37541124e-01 2.05333233e+00 -1.42221510e+00
4.10356581e-01 1.06072056e+00 6.51507616e-01 1.48998439e-01
-6.35518730e-01 -2.38372147e-01 6.22570097e-01 -2.73226529e-01
-6.44673645e-01 -1.80863887e-01 7.47076929e-01 -2.20684126e-01
1.23870850e+00 1.74873732e-02 7.11502910e-01 4.65678215e-01
4.55734283e-01 1.16426003e+00 8.10418844e-01 -1.51650503e-01
3.39075595e-01 4.20169473e-01 5.69016695e-01 5.00054240e-01
1.34896293e-01 -5.61596990e-01 -7.37690985e-01 -9.06950906e-02
-8.52625892e-02 9.77055654e-02 1.79547861e-01 -4.22359779e-02
-8.73958468e-01 1.02526283e+00 3.32820565e-01 -2.24444613e-01
-4.33417976e-01 -2.21857160e-01 7.04462111e-01 5.05948603e-01
1.16285098e+00 5.72547317e-01 -1.20619094e+00 -6.93565682e-02
-8.39470744e-01 -4.94647250e-02 1.06860721e+00 9.85928833e-01
8.10177803e-01 7.32616633e-02 -1.96941152e-01 1.11681020e+00
4.27812755e-01 7.18282402e-01 5.44842660e-01 -7.20320344e-01
5.91534555e-01 1.10164905e+00 -6.81026205e-02 -8.13093424e-01
-3.34247172e-01 -6.87941194e-01 -5.75506151e-01 1.02960721e-01
-3.14527512e-01 -3.31946313e-01 -1.12307405e+00 1.05890191e+00
1.66878238e-01 -6.98808789e-01 3.97005677e-01 6.02498114e-01
1.26169395e+00 7.75940776e-01 1.40619613e-02 5.21889739e-02
1.76868963e+00 -1.47132182e+00 -5.71323335e-01 -5.60705125e-01
8.93258154e-01 -8.77073705e-01 1.47325134e+00 5.22079170e-01
-9.37307000e-01 -5.98203897e-01 -1.13308227e+00 -4.03945655e-01
-6.49827659e-01 7.01289535e-01 7.57829010e-01 4.65666562e-01
-1.12456536e+00 2.64673114e-01 -4.97988552e-01 2.43638884e-02
3.22321564e-01 2.42840618e-01 -3.08106869e-01 6.27836436e-02
-9.58710611e-01 5.74624062e-01 -4.18835849e-01 -4.04613465e-03
-7.54646420e-01 -8.69704545e-01 -1.16066432e+00 2.58514732e-01
3.74683663e-02 -5.52248657e-01 1.81713283e+00 -1.19625080e+00
-1.28640819e+00 9.84872162e-01 -5.08204043e-01 -5.88046089e-02
-3.38844031e-01 -3.49930465e-01 -4.52359319e-01 1.02179885e-01
4.44209903e-01 5.76060653e-01 8.14101458e-01 -1.13722253e+00
-8.97866786e-01 -4.51809198e-01 5.44563055e-01 6.97070479e-01
-6.27403498e-01 3.53321284e-01 -6.22126639e-01 -5.47131121e-01
-1.12533294e-01 -8.08174074e-01 -4.37416404e-01 -2.92762756e-01
-3.38759929e-01 -4.33628529e-01 5.28375566e-01 -4.57859397e-01
9.49381292e-01 -2.16410232e+00 -3.09677243e-01 1.33327737e-01
2.99700707e-01 -4.81881537e-02 -4.42754090e-01 2.43397161e-01
-3.26383337e-02 5.35984822e-02 -1.64220154e-01 -8.57803047e-01
1.67121857e-01 -9.44168568e-02 -5.99741876e-01 -2.21523251e-02
2.66204536e-01 1.18202388e+00 -7.77776599e-01 -1.53902158e-01
-3.72778237e-01 4.37649906e-01 -6.71057284e-01 2.00885221e-01
-3.27397287e-01 -3.24531972e-01 -4.31725979e-01 1.00145745e+00
3.39825362e-01 -5.49982131e-01 -1.03816591e-01 -3.73947889e-01
1.79067507e-01 8.66560817e-01 -5.62900960e-01 1.25598598e+00
-9.16272998e-01 6.93530440e-01 8.21515769e-02 -5.21698058e-01
1.13660550e+00 1.95538685e-01 1.44829214e-01 -6.30390704e-01
9.18258801e-02 1.72473133e-01 -2.41265520e-01 4.67932895e-02
1.20645034e+00 -3.44008088e-01 -5.62584937e-01 1.03096640e+00
5.32390065e-02 -6.60999715e-01 4.93865848e-01 4.55303639e-01
9.63947296e-01 -2.24767953e-01 3.04863393e-01 -2.16354296e-01
6.37256026e-01 3.16735327e-01 3.01518738e-01 4.66734111e-01
3.32633927e-02 7.60851026e-01 8.46873999e-01 -5.53539097e-01
-9.54461753e-01 -8.66234422e-01 1.33156106e-01 1.62228131e+00
-2.23369777e-01 -9.14565146e-01 -3.05904567e-01 -1.13527274e+00
-1.11169830e-01 6.24255121e-01 -8.30760241e-01 -7.28048012e-02
-1.49751201e-01 -9.26731646e-01 -1.63745997e-03 7.49767303e-01
1.36997074e-01 -1.17801619e+00 -7.94353336e-02 7.16813430e-02
4.89231795e-02 -1.12548554e+00 -6.45321965e-01 6.04680777e-01
-8.07342529e-01 -7.85587013e-01 -5.78281283e-01 -9.53688204e-01
8.95191193e-01 5.31876445e-01 1.83330262e+00 -2.50952572e-01
2.44998857e-01 3.75009984e-01 -6.93868756e-01 -7.48563468e-01
-2.10210577e-01 3.86051029e-01 -2.66308665e-01 -2.01886505e-01
9.54347491e-01 -3.18236858e-01 -8.29496980e-01 2.35964149e-01
-8.10083985e-01 -3.03315997e-01 7.48637557e-01 4.67257947e-01
7.40635931e-01 -1.34554148e-01 6.48122787e-01 -1.40060067e+00
1.13793075e+00 -3.02833050e-01 -3.67236197e-01 4.12961431e-02
-9.19343710e-01 -1.63189575e-01 9.66236711e-01 3.07011809e-02
-9.46505249e-01 -1.90378755e-01 -3.98872107e-01 2.30940402e-01
2.99794614e-01 9.34781134e-01 -4.63265553e-02 4.90799457e-01
4.81036633e-01 1.34866893e-01 -1.14084601e-01 -3.11081678e-01
6.45523369e-01 1.02178645e+00 1.79376394e-01 -1.25663370e-01
5.70779979e-01 5.27898371e-01 -4.85546112e-01 -3.53650570e-01
-1.94373941e+00 -8.09993386e-01 -4.28572237e-01 1.22733064e-01
7.68149912e-01 -1.62957752e+00 -3.13842952e-01 2.26286709e-01
-9.75584924e-01 -8.17164313e-03 -6.49557531e-01 1.14176139e-01
-2.80668885e-01 8.70585367e-02 -9.84532058e-01 -4.63882089e-01
-1.11815560e+00 -1.12905705e+00 1.56787086e+00 1.79336235e-01
-5.39945364e-01 -1.14078760e+00 2.49754101e-01 7.44461715e-01
2.75519520e-01 -4.54120427e-01 9.04349387e-01 -8.13941538e-01
2.40405602e-03 -5.13245881e-01 -3.44907224e-01 9.06521797e-01
2.01568604e-01 4.85931262e-02 -1.05277896e+00 -8.49953480e-03
1.47439435e-01 -4.00213182e-01 9.50793207e-01 3.19863111e-01
7.29284585e-01 -3.24878752e-01 2.67055124e-01 4.60725516e-01
1.25021005e+00 -5.33835143e-02 2.71398723e-01 3.38183999e-01
8.43351245e-01 7.76178300e-01 1.03404295e+00 4.29510325e-01
7.42054224e-01 5.48994802e-02 2.64662951e-01 -4.12022114e-01
8.25645477e-02 -1.18652947e-01 8.58066797e-01 1.46326649e+00
4.18095052e-01 -1.24049224e-01 -4.54899132e-01 8.67432535e-01
-1.45450485e+00 -4.60440338e-01 -5.33753112e-02 1.48753500e+00
8.53172898e-01 4.10782754e-01 1.14825517e-02 4.87612997e-04
2.37408727e-01 6.11317396e-01 -5.67094028e-01 -1.17233419e+00
-1.80063218e-01 2.24645302e-01 3.30365628e-01 4.62632686e-01
-1.05633461e+00 1.01672852e+00 6.43802166e+00 5.33679843e-01
-8.13482285e-01 1.06388517e-01 1.10033751e+00 -2.04890430e-01
-9.02182221e-01 -3.32919508e-02 -1.11012256e+00 -1.13091335e-01
9.11863267e-01 -1.11381479e-01 -1.60458669e-01 1.37690973e+00
1.94894060e-01 1.71679914e-01 -9.48702872e-01 4.88558292e-01
4.92745280e-01 -1.03763223e+00 3.63344014e-01 -2.21702680e-01
1.21924675e+00 3.11702549e-01 3.48550439e-01 5.34025550e-01
6.62419736e-01 -9.14523482e-01 4.79808003e-01 -2.09445089e-01
7.07449794e-01 -9.73268926e-01 9.84096408e-01 -1.31433323e-01
-1.34891212e+00 -3.04850619e-02 -7.39439845e-01 -1.69280753e-01
2.76110232e-01 1.12313163e+00 -7.48597503e-01 3.50342393e-01
6.11574948e-01 1.45942008e+00 -7.06503510e-01 2.36776501e-01
-7.52138495e-01 5.70051730e-01 1.16045438e-01 -2.96891630e-01
5.72252452e-01 -3.05607527e-01 6.30081892e-02 1.16548967e+00
2.56213456e-01 -1.73695073e-01 -9.50176865e-02 3.19741428e-01
-3.99514049e-01 4.76982057e-01 -4.21763569e-01 -1.83364958e-01
-1.72365218e-01 1.86159909e+00 -7.08167315e-01 -5.89888215e-01
-8.36189687e-01 7.21186996e-01 1.80883735e-01 3.08365464e-01
-3.37057620e-01 -5.47719836e-01 9.15483475e-01 -1.56661257e-01
8.22831094e-01 1.90795526e-01 -8.38171124e-01 -1.37335336e+00
2.89756835e-01 -1.19310045e+00 2.79097915e-01 -1.03457832e+00
-1.44419599e+00 9.38220739e-01 -6.19768858e-01 -1.43493009e+00
-1.69643432e-01 -1.03520274e+00 -7.99855590e-01 7.21700430e-01
-1.88610899e+00 -1.18978512e+00 5.81304580e-02 4.47671622e-01
8.94934058e-01 -3.18574637e-01 7.35721171e-01 9.36107431e-03
-5.85731082e-02 6.29434586e-01 -3.57226990e-02 2.62558758e-01
7.65319705e-01 -1.65787768e+00 9.08446074e-01 6.47614419e-01
2.52863705e-01 6.63070023e-01 5.79877257e-01 -5.53724825e-01
-1.15138996e+00 -1.12617004e+00 1.30535328e+00 -7.36493170e-01
9.43179190e-01 -4.61801082e-01 -4.89591926e-01 6.53811395e-01
4.43116784e-01 -2.53361017e-01 1.15851665e+00 7.05874681e-01
-8.61223936e-01 -5.08935094e-01 -6.29762948e-01 6.65274382e-01
4.92411077e-01 -8.54490817e-01 -7.90112138e-01 3.85008633e-01
1.10676777e+00 2.31928900e-02 -8.88655305e-01 5.09654462e-01
6.12521172e-01 -7.70704508e-01 6.86537087e-01 -5.66994548e-01
1.34268844e+00 -3.76158923e-01 -2.50797927e-01 -1.60417366e+00
-8.78837854e-02 -1.71542242e-01 7.62766898e-02 1.17732167e+00
1.26129293e+00 -2.45144352e-01 1.00388706e+00 4.58678842e-01
-2.73721874e-01 -1.15916252e+00 -2.09142342e-01 -1.02599777e-01
1.52209938e-01 -5.90548158e-01 5.99308133e-01 3.86575401e-01
3.18298131e-01 1.07668400e+00 -1.65533379e-01 -1.29675418e-01
1.63591549e-01 6.60429120e-01 7.01838851e-01 -1.03676093e+00
-3.96543711e-01 -3.97277147e-01 4.40809764e-02 -1.29223394e+00
2.84501076e-01 -8.41167152e-01 2.28852302e-01 -1.97120225e+00
4.00750190e-01 -5.18465228e-02 -5.15813231e-01 4.92369950e-01
-4.87656146e-01 5.60807586e-01 5.20305075e-02 9.60840434e-02
-1.09312022e+00 6.71880960e-01 1.43242347e+00 -4.58267212e-01
-1.03010803e-01 3.18429649e-01 -1.79435968e+00 8.69623303e-01
7.28521764e-01 -3.84790629e-01 -6.88854814e-01 -4.79298383e-01
1.16449678e+00 -3.37987781e-01 -6.20341420e-01 -4.26051766e-01
3.54318544e-02 2.16850907e-01 3.89368862e-01 -1.05602443e+00
2.96120375e-01 -6.07241035e-01 -9.05722082e-01 -4.51696478e-02
-6.95798874e-01 5.60793817e-01 -4.45880890e-02 3.35946530e-01
-7.73246348e-01 -1.64522871e-01 2.79815495e-01 -2.08635360e-01
-3.45946461e-01 2.70809323e-01 -5.54620028e-01 1.28011152e-01
4.18771237e-01 9.82152969e-02 -4.79648083e-01 -8.18402946e-01
-3.43557984e-01 2.67408937e-01 3.84702832e-01 5.09937823e-01
6.60172045e-01 -9.44683969e-01 -6.90894485e-01 1.61587447e-01
5.87139845e-01 8.54714438e-02 3.97873931e-02 4.11663353e-01
-1.16187148e-01 4.17939395e-01 3.32714885e-01 -3.57001811e-01
-1.22208738e+00 4.21368122e-01 -5.96265756e-02 -9.60927844e-01
-9.94855836e-02 8.41298103e-01 5.38798571e-01 -1.14933026e+00
-4.84907553e-02 -3.85588348e-01 -7.39137232e-01 4.70897973e-01
6.35546923e-01 -2.88044482e-01 3.75294089e-01 -4.93494928e-01
-1.63230091e-01 9.55547333e-01 -6.38468206e-01 -2.56468952e-01
1.42721820e+00 -2.83075452e-01 -3.17886949e-01 4.72275674e-01
1.33967388e+00 4.35903102e-01 -9.87699628e-01 -1.20503262e-01
-3.10021918e-02 1.15557499e-01 -5.80482483e-02 -1.06002223e+00
-1.31609261e+00 8.59469652e-01 -2.06582770e-01 2.67528594e-01
1.28312504e+00 1.41350508e-01 8.30475807e-01 6.76247537e-01
-1.97347343e-01 -1.26704597e+00 1.58508882e-01 9.53675807e-01
6.50341928e-01 -1.34882677e+00 4.00087088e-01 -4.40821856e-01
-1.33953524e+00 7.70283282e-01 6.25144541e-01 -3.25463295e-01
9.22437310e-01 4.09722984e-01 7.35508680e-01 -4.39705193e-01
-1.36183059e+00 -1.13361008e-01 5.66341400e-01 3.47958863e-01
7.28908658e-01 3.63152958e-02 -1.91189036e-01 8.80313754e-01
-5.94475567e-01 -4.06293571e-01 7.58040845e-01 6.76756978e-01
-5.07716835e-01 -1.12038922e+00 1.51149035e-01 9.09179568e-01
-9.71021652e-01 -6.95527613e-01 -5.71113288e-01 1.96911082e-01
-4.43708926e-01 1.20431685e+00 2.51392219e-02 -4.13697928e-01
5.01755357e-01 -1.38387918e-01 -4.07545984e-01 -1.14187133e+00
-1.23559940e+00 1.37446329e-01 4.25850660e-01 -4.13603187e-01
-4.23644364e-01 -3.51672769e-01 -9.94451404e-01 1.08835548e-01
1.89608894e-02 4.93960679e-01 9.50410008e-01 1.01647365e+00
6.57799363e-01 4.79761630e-01 1.09292579e+00 -3.44602793e-01
-5.77178359e-01 -1.07387030e+00 -7.15448380e-01 3.50374013e-01
3.56843352e-01 1.72144994e-01 -7.02612936e-01 5.48719242e-02] | [11.435355186462402, 6.689326286315918] |
16da1bd5-a606-465d-acfc-aa19b9add6f9 | improving-target-speaker-extraction-with | 2301.06277 | null | https://arxiv.org/abs/2301.06277v1 | https://arxiv.org/pdf/2301.06277v1.pdf | Improving Target Speaker Extraction with Sparse LDA-transformed Speaker Embeddings | As a practical alternative of speech separation, target speaker extraction (TSE) aims to extract the speech from the desired speaker using additional speaker cue extracted from the speaker. Its main challenge lies in how to properly extract and leverage the speaker cue to benefit the extracted speech quality. The cue extraction method adopted in majority existing TSE studies is to directly utilize discriminative speaker embedding, which is extracted from the pre-trained models for speaker verification. Although the high speaker discriminability is a most desirable property for speaker verification task, we argue that it may be too sophisticated for TSE. In this study, we propose that a simplified speaker cue with clear class separability might be preferred for TSE. To verify our proposal, we introduce several forms of speaker cues, including naive speaker embedding (such as, x-vector and xi-vector) and new speaker embeddings produced from sparse LDA-transform. Corresponding TSE models are built by integrating these speaker cues with SepFormer (one SOTA speech separation model). Performances of these TSE models are examined on the benchmark WSJ0-2mix dataset. Experimental results validate the effectiveness and generalizability of our proposal, showing up to 9.9% relative improvement in SI-SDRi. Moreover, with SI-SDRi of 19.4 dB and PESQ of 3.78, our best TSE system significantly outperforms the current SOTA systems and offers the top TSE results reported till date on the WSJ0-2mix. | ['Huan Zhou', 'Ziqing Du', 'Xucheng Wan', 'Kai Liu'] | 2023-01-16 | null | null | null | null | ['target-speaker-extraction', 'speech-separation', 'speaker-verification'] | ['audio', 'speech', 'speech'] | [ 1.77645072e-01 -4.65341546e-02 -2.23606631e-01 -2.78259188e-01
-1.33351910e+00 -3.23472083e-01 5.49316585e-01 -1.80018857e-01
-1.42925635e-01 3.67841423e-01 4.73200679e-01 -2.56519675e-01
-1.42206624e-01 5.18329628e-02 -2.48336971e-01 -1.08934951e+00
2.99853855e-03 -2.67884210e-02 -1.14047714e-01 -1.48617670e-01
1.35166505e-02 5.65773726e-01 -1.61214221e+00 -6.58413628e-03
8.60207438e-01 1.15787697e+00 2.36333504e-01 5.88890374e-01
-3.50921974e-02 1.96029738e-01 -8.22619081e-01 -1.88044339e-01
2.57398874e-01 -5.06353498e-01 -1.80533066e-01 1.34441897e-01
4.70491946e-01 -1.59716964e-01 -2.44632497e-01 9.43526864e-01
8.79523337e-01 2.38901544e-02 7.23419726e-01 -1.16543996e+00
-4.11839157e-01 1.03500783e+00 -5.23227930e-01 3.65039051e-01
2.50849128e-01 -7.30323419e-02 1.19221854e+00 -1.36838090e+00
3.67025938e-03 1.27607203e+00 4.33521092e-01 5.65974653e-01
-1.21038997e+00 -1.01549590e+00 1.60418525e-01 8.28133002e-02
-1.70044720e+00 -1.27553880e+00 1.18831122e+00 -2.10635304e-01
7.57018864e-01 6.44002318e-01 3.33786219e-01 1.15243733e+00
-2.84038812e-01 1.00667953e+00 1.21584201e+00 -5.81879377e-01
1.24083608e-01 5.75478017e-01 4.33858603e-01 3.15771312e-01
3.57357115e-02 2.82641828e-01 -1.03520656e+00 -1.66007221e-01
6.54171184e-02 -3.38984191e-01 -5.91399968e-01 -3.85267474e-02
-1.06057978e+00 6.78285241e-01 1.82463929e-01 4.98451769e-01
-3.17373157e-01 -3.59797627e-01 1.98866174e-01 2.45213434e-01
4.34696048e-01 1.54372677e-01 -1.08542390e-01 6.11388776e-03
-1.39914382e+00 -2.04096455e-02 5.71965814e-01 1.06814325e+00
3.08488518e-01 8.00035894e-01 -3.93312842e-01 1.16618967e+00
7.02567160e-01 9.90292549e-01 7.00310946e-01 -3.26896340e-01
4.84800458e-01 7.61436820e-02 -1.97054863e-01 -5.92374325e-01
1.68594148e-03 -8.52095783e-01 -7.00692475e-01 -1.16878189e-01
5.45537248e-02 1.90240815e-02 -9.00942624e-01 1.71361244e+00
2.45477736e-01 3.47892135e-01 4.59756911e-01 8.78376305e-01
9.61152434e-01 8.50288033e-01 -1.82582304e-01 -4.45060998e-01
1.48684597e+00 -9.24071968e-01 -9.71875846e-01 -3.61824483e-01
2.38041550e-01 -9.54318166e-01 8.46373796e-01 4.27454948e-01
-7.81316519e-01 -5.96816838e-01 -1.24858272e+00 4.07834798e-01
3.31178680e-03 6.57354832e-01 1.14177115e-01 1.15674484e+00
-9.65319574e-01 -7.61394724e-02 -5.44596910e-01 -2.66689241e-01
4.52408679e-02 3.48847449e-01 -3.38723183e-01 1.07904106e-01
-1.23635542e+00 5.60753345e-01 9.28791761e-02 2.74160415e-01
-8.96714211e-01 -6.14970863e-01 -8.66237700e-01 3.35152715e-01
2.27191836e-01 -2.26062432e-01 1.04635096e+00 -6.53680384e-01
-1.97738898e+00 5.62154293e-01 -6.50754869e-01 -5.73021114e-01
9.71786603e-02 -1.21592291e-01 -9.78011429e-01 1.44487187e-01
-1.76211586e-03 3.53503674e-01 1.35397089e+00 -1.48731518e+00
-5.04933596e-01 -3.26324433e-01 -5.80112636e-01 6.10221364e-02
-7.52226472e-01 4.03038800e-01 -3.41821462e-01 -8.20827544e-01
3.55786949e-01 -1.16492093e+00 2.49953866e-01 -5.10469139e-01
-6.89355552e-01 -2.43361309e-01 1.00305712e+00 -1.08847427e+00
1.50111055e+00 -2.59392071e+00 -7.53463730e-02 2.46293932e-01
-1.57689061e-02 5.14461756e-01 -5.69630377e-02 2.49109074e-01
-2.34694645e-01 -1.38084695e-01 -2.11084574e-01 -7.25106657e-01
2.67831802e-01 -3.09148669e-01 -5.85433006e-01 5.21127224e-01
2.60934621e-01 4.75329965e-01 -3.64706397e-01 -3.98769051e-01
2.81733036e-01 7.69608796e-01 -4.37360376e-01 2.93727815e-01
5.25688529e-01 2.67347753e-01 -2.09888518e-01 6.99362159e-01
8.92894924e-01 3.11741412e-01 6.04619309e-02 -3.54981542e-01
-8.06421414e-02 7.28140116e-01 -1.39703119e+00 1.27571821e+00
-3.42862338e-01 7.53854811e-01 4.92275059e-01 -7.41298437e-01
1.14011097e+00 6.67127967e-01 2.24056333e-01 -3.10576946e-01
1.03448108e-01 3.72342020e-01 3.54023099e-01 -1.55728191e-01
3.09847206e-01 -4.04761940e-01 -1.32112294e-01 9.57225189e-02
3.07666600e-01 -1.03452057e-01 -1.88089430e-01 1.75934225e-01
5.75235963e-01 -4.17098790e-01 3.20327848e-01 -5.36663830e-01
8.78069937e-01 -6.43321753e-01 6.45065129e-01 4.10115898e-01
-5.06290555e-01 5.49896896e-01 5.08607961e-02 5.15234470e-01
-5.69303453e-01 -1.13848686e+00 -3.33609492e-01 9.93349612e-01
-9.57569927e-02 -4.06534463e-01 -7.85723150e-01 -3.70795310e-01
-1.60957411e-01 9.98956263e-01 -1.59067154e-01 -1.22090071e-01
-7.09682703e-01 -5.79997420e-01 7.65752614e-01 4.17398304e-01
1.65712342e-01 -6.26689196e-01 -8.97758827e-03 1.32529989e-01
-1.81297034e-01 -1.26097417e+00 -8.34512055e-01 2.80078053e-01
-5.27251959e-01 -3.50314617e-01 -9.36805725e-01 -7.39343762e-01
4.05665874e-01 6.91379189e-01 4.20638949e-01 -3.90864819e-01
1.89870387e-01 2.46874914e-01 -3.48615348e-01 -4.30993885e-01
-6.43856883e-01 1.05589941e-01 5.88130116e-01 6.01392329e-01
2.99611419e-01 -4.85339463e-01 -2.68728971e-01 2.79882729e-01
-6.22687280e-01 -2.75383949e-01 5.97500503e-01 8.75788748e-01
1.45253405e-01 9.96970013e-02 9.51228499e-01 -2.51734167e-01
6.08601213e-01 -2.09056035e-01 -4.12054241e-01 8.06187242e-02
-6.23769701e-01 -1.01983023e-03 5.96744120e-01 -4.44273978e-01
-1.20500767e+00 -4.54096124e-02 -4.49811280e-01 -5.68029165e-01
-2.02561468e-01 3.46886486e-01 -7.13586092e-01 1.39432281e-01
2.88855910e-01 5.92415929e-01 -2.19659992e-02 -5.92638433e-01
2.16048077e-01 1.23888099e+00 2.86597818e-01 -4.17037874e-01
9.63576913e-01 1.61873519e-01 -5.24018168e-01 -1.44662380e+00
-4.81096148e-01 -7.38558352e-01 -3.60896021e-01 1.52615481e-03
5.03048718e-01 -1.19504094e+00 -5.04725456e-01 4.19143796e-01
-8.57285917e-01 2.62633204e-01 2.43704747e-02 9.08941448e-01
-9.85395983e-02 4.26847160e-01 -4.82487857e-01 -1.28496897e+00
-4.82092917e-01 -1.46647131e+00 1.16390526e+00 9.33130756e-02
-2.15030536e-01 -6.18208170e-01 -1.80180997e-01 5.83093226e-01
6.64912105e-01 -5.41726947e-01 5.77344298e-01 -1.04604018e+00
-3.47770452e-01 -9.15300623e-02 9.54600275e-02 6.26008391e-01
5.51103532e-01 -1.49354488e-01 -1.68306744e+00 -6.25954330e-01
2.93188661e-01 2.28944555e-01 9.23208356e-01 4.82070774e-01
6.00920737e-01 -2.59431869e-01 -4.39987987e-01 3.82448494e-01
1.01494730e+00 3.14004660e-01 2.81525165e-01 -1.46668583e-01
5.56825340e-01 6.22023284e-01 4.05369133e-01 2.35650718e-01
1.25392929e-01 9.75284040e-01 -1.27118260e-01 1.24445148e-01
-5.80344856e-01 -1.27069026e-01 1.09174144e+00 1.69846296e+00
2.65384078e-01 -2.74995595e-01 -6.57332242e-01 6.63661420e-01
-1.23827291e+00 -8.50875854e-01 1.02384090e-01 2.36631465e+00
6.87034369e-01 1.93421781e-01 1.96349800e-01 6.62033260e-01
8.27527821e-01 3.38224888e-01 -3.82702142e-01 -1.10788144e-01
-3.31155837e-01 1.89053521e-01 2.25378036e-01 6.77937269e-01
-1.01674831e+00 7.41809428e-01 5.44257975e+00 1.09813452e+00
-1.54342830e+00 2.09703937e-01 1.88651428e-01 -2.71456808e-01
-1.12367779e-01 -2.54614681e-01 -1.22963631e+00 5.29751599e-01
1.36356103e+00 -1.83377519e-01 2.38855824e-01 6.17236316e-01
3.13705385e-01 2.38881290e-01 -1.18485606e+00 1.17469549e+00
5.60075045e-01 -7.32754529e-01 -1.28084093e-01 2.10817724e-01
3.24460328e-01 -1.70009330e-01 3.07193756e-01 4.94226724e-01
-2.27042571e-01 -8.06816161e-01 9.27070320e-01 5.12816720e-02
6.96450293e-01 -6.91307545e-01 5.37830591e-01 2.67045766e-01
-1.28204310e+00 -1.54500395e-01 8.75873119e-02 4.89275783e-01
1.72427863e-01 6.54039443e-01 -1.26109636e+00 7.39286482e-01
4.82541710e-01 4.24321651e-01 -3.60691011e-01 8.15594852e-01
-2.16249570e-01 1.22902858e+00 -5.04696310e-01 5.21337949e-02
-9.68205705e-02 -2.44294293e-02 1.16757476e+00 1.50892425e+00
7.26581514e-01 -3.17246705e-01 -5.88027798e-02 7.03472853e-01
8.29559863e-02 2.25690931e-01 -4.78676289e-01 -6.51116520e-02
8.58831882e-01 1.15697098e+00 -4.31090087e-01 -2.88680017e-01
-1.65611312e-01 8.13453197e-01 -6.98258355e-02 4.66952115e-01
-8.31995428e-01 -4.09266263e-01 9.27496135e-01 -4.49986057e-03
7.18355536e-01 -2.09447905e-01 -1.29156420e-02 -1.17611384e+00
1.65672615e-01 -1.16170967e+00 3.09350062e-02 -3.24920386e-01
-1.27996790e+00 9.41428185e-01 -1.28466874e-01 -1.47077072e+00
-2.68816024e-01 -5.45403957e-01 -5.32494307e-01 1.04192626e+00
-1.64846838e+00 -1.23325491e+00 2.42776245e-01 4.10696179e-01
7.20374227e-01 -5.15898824e-01 8.51180375e-01 4.94405895e-01
-9.00557578e-01 1.17809570e+00 1.95953086e-01 8.10992494e-02
7.95797110e-01 -9.72654819e-01 1.40597999e-01 1.08409178e+00
3.32718253e-01 9.20582652e-01 8.16462219e-01 -1.88355953e-01
-1.59195697e+00 -8.75092089e-01 9.91358221e-01 -1.36327744e-01
5.28730452e-01 -7.28219151e-01 -8.48553240e-01 4.79513973e-01
3.26331645e-01 -1.45384341e-01 9.54237103e-01 6.17775433e-02
-4.33267087e-01 -5.05079567e-01 -9.52366650e-01 4.99951005e-01
7.03389525e-01 -7.62126029e-01 -7.81153738e-01 -2.88185418e-01
7.97293127e-01 1.25065865e-02 -7.05168605e-01 2.65220582e-01
3.82571638e-01 -8.23287606e-01 1.12965417e+00 -1.07578032e-01
-3.02304059e-01 -5.77358484e-01 -6.58171654e-01 -1.57645965e+00
-1.49397016e-01 -7.27770865e-01 -1.04408368e-01 1.98213756e+00
6.22333944e-01 -7.96158731e-01 4.17775810e-01 2.74024040e-01
-2.58992732e-01 -2.79878557e-01 -1.25605607e+00 -1.19692802e+00
-1.41980708e-01 -5.46136022e-01 7.41633773e-01 9.29038346e-01
3.88405146e-03 6.94252968e-01 -5.95729470e-01 6.43453002e-01
7.61820316e-01 4.80446517e-02 7.41716862e-01 -1.22004056e+00
-2.68526554e-01 -5.38349926e-01 -2.06489697e-01 -1.15287209e+00
4.52172637e-01 -1.05335617e+00 2.02372700e-01 -1.12417328e+00
-1.22152925e-01 -3.65730286e-01 -6.72941566e-01 1.79493621e-01
-2.87556469e-01 -1.47090152e-01 4.78042960e-01 8.11381713e-02
-2.71391938e-03 9.58176911e-01 7.38139868e-01 -3.51527452e-01
-4.00978744e-01 1.49051264e-01 -8.70738447e-01 4.84449655e-01
6.94646597e-01 -4.63693589e-01 -3.23792100e-01 -1.85924899e-02
-6.46782815e-01 1.97478533e-01 1.18495956e-01 -1.10215139e+00
2.44483158e-01 2.27843881e-01 -2.58146301e-02 -6.14250124e-01
9.45583403e-01 -8.29147995e-01 -6.82135224e-02 1.78014666e-01
-2.88423777e-01 -6.58555686e-01 3.09527397e-01 4.58657980e-01
-5.87534606e-01 -1.26856551e-01 7.29108214e-01 6.34656727e-01
-3.04288030e-01 -8.41272250e-02 -3.99265826e-01 -3.43448371e-01
6.93469405e-01 -1.40131459e-01 6.25823066e-02 -3.38099658e-01
-5.76507568e-01 -1.93027690e-01 -2.50091881e-01 4.70216185e-01
6.31583333e-01 -1.47933769e+00 -9.40985024e-01 5.41826725e-01
2.67107129e-01 -5.80811024e-01 2.86901712e-01 1.04721713e+00
3.89529765e-01 8.41315329e-01 2.10505038e-01 -6.96515441e-01
-1.59966671e+00 4.50582594e-01 1.20397666e-02 1.31698549e-01
-5.02009928e-01 9.70259428e-01 2.94473618e-01 -4.22145218e-01
3.19439620e-01 -4.97979015e-01 -1.64676085e-01 6.77302703e-02
5.85825443e-01 3.44740272e-01 4.02616203e-01 -1.10179484e+00
-7.14233100e-01 3.32703948e-01 -2.96459384e-02 -3.18475246e-01
1.21784329e+00 -3.00187230e-01 1.23181209e-01 5.16465545e-01
1.23298168e+00 8.49230766e-01 -9.71235514e-01 -5.37800550e-01
7.40276799e-02 -4.27345812e-01 3.98424387e-01 -5.57673097e-01
-9.57547784e-01 1.07477295e+00 7.86128461e-01 2.97781736e-01
1.11590707e+00 6.06429428e-02 8.77504289e-01 -3.82701540e-03
1.47129908e-01 -7.94190884e-01 -2.05377057e-01 2.40485504e-01
1.10553265e+00 -1.18221378e+00 -2.05721721e-01 -5.03437579e-01
-6.29176974e-01 8.09406757e-01 3.44495684e-01 3.10898691e-01
8.15156639e-01 3.57760489e-01 2.38509417e-01 1.28559873e-01
-4.53606129e-01 -2.40285531e-01 6.92948580e-01 3.26005578e-01
5.46731114e-01 3.29684466e-01 6.08115830e-02 1.00798762e+00
-5.48534811e-01 -6.61326349e-01 2.06133068e-01 5.05604684e-01
-5.04293263e-01 -1.12952542e+00 -8.39843214e-01 1.96543306e-01
-3.35085213e-01 -2.47053683e-01 -2.84326047e-01 3.17405164e-01
-3.47233146e-01 1.37346435e+00 -3.52998853e-01 -4.51763332e-01
3.05308610e-01 4.63244855e-01 1.10270254e-01 -4.38503265e-01
-4.34308410e-01 6.96673751e-01 1.25809968e-01 4.94760163e-02
-3.89141351e-01 -7.41293728e-01 -1.09267211e+00 2.92785279e-02
-7.45688498e-01 2.61601567e-01 8.71932328e-01 8.31693113e-01
4.33190018e-01 6.50759637e-01 9.26582038e-01 -8.47600698e-01
-6.49449527e-01 -1.23217392e+00 -7.35589862e-01 1.96806956e-04
7.85489500e-01 -7.37017155e-01 -7.50166357e-01 8.30037333e-03] | [14.734637260437012, 5.949735164642334] |
0793ebc1-e454-42ae-9a7e-2e2a4269fc76 | taking-modality-free-human-identification-as | 2010.00975 | null | https://arxiv.org/abs/2010.00975v2 | https://arxiv.org/pdf/2010.00975v2.pdf | Taking Modality-free Human Identification as Zero-shot Learning | Human identification is an important topic in event detection, person tracking, and public security. There have been numerous methods proposed for human identification, such as face identification, person re-identification, and gait identification. Typically, existing methods predominantly classify a queried image to a specific identity in an image gallery set (I2I). This is seriously limited for the scenario where only a textual description of the query or an attribute gallery set is available in a wide range of video surveillance applications (A2I or I2A). However, very few efforts have been devoted towards modality-free identification, i.e., identifying a query in a gallery set in a scalable way. In this work, we take an initial attempt, and formulate such a novel Modality-Free Human Identification (named MFHI) task as a generic zero-shot learning model in a scalable way. Meanwhile, it is capable of bridging the visual and semantic modalities by learning a discriminative prototype of each identity. In addition, the semantics-guided spatial attention is enforced on visual modality to obtain representations with both high global category-level and local attribute-level discrimination. Finally, we design and conduct an extensive group of experiments on two common challenging identification tasks, including face identification and person re-identification, demonstrating that our method outperforms a wide variety of state-of-the-art methods on modality-free human identification. | ['Jian Cheng', 'Yao Zhao', 'Shuai Zheng', 'Zhenfeng Zhu', 'Xingxing Zhang', 'Zhizhe Liu'] | 2020-10-02 | null | null | null | null | ['gait-identification'] | ['computer-vision'] | [ 3.68118227e-01 -6.01766109e-01 -3.40793878e-01 -3.15585822e-01
-8.39472234e-01 -5.13490617e-01 7.39023328e-01 -8.37808996e-02
-3.90354931e-01 5.57066798e-01 1.94029659e-01 2.49467716e-01
5.36623299e-02 -5.69834828e-01 -2.99955934e-01 -7.02551961e-01
3.05971503e-01 2.72455394e-01 -1.05426395e-02 4.79299622e-03
8.04707706e-02 3.91878217e-01 -1.90903330e+00 -5.90410493e-02
4.35411751e-01 1.17522860e+00 -7.17699155e-02 3.01001221e-01
9.74563062e-02 4.64849442e-01 -4.57042366e-01 -6.41620696e-01
7.65996575e-02 -5.12465119e-01 -9.27631199e-01 2.90728599e-01
5.18226147e-01 -2.70180672e-01 -4.50337589e-01 1.19023764e+00
8.44418406e-01 3.22641015e-01 6.60941005e-01 -1.66251779e+00
-7.63243377e-01 1.97690334e-02 -7.13255525e-01 2.69370049e-01
8.88399661e-01 1.83648854e-01 5.88986397e-01 -1.03892887e+00
5.12014925e-01 1.36353445e+00 7.25968778e-01 7.77499139e-01
-9.68515098e-01 -9.49283004e-01 6.87641203e-02 6.68618023e-01
-1.94794393e+00 -8.90504241e-01 7.55959868e-01 -4.58057880e-01
5.45744479e-01 3.27580035e-01 4.33842748e-01 1.28281844e+00
-5.38459122e-01 7.94332445e-01 8.35789025e-01 -4.62013990e-01
7.06875995e-02 2.12203950e-01 3.11184049e-01 5.28111756e-01
1.75527140e-01 4.72940132e-02 -8.66291463e-01 -3.12323034e-01
5.76821864e-01 3.01904261e-01 -4.04133320e-01 -2.71119714e-01
-1.32109475e+00 5.83020091e-01 2.71784663e-01 2.40209922e-01
-3.82801890e-01 -1.56597078e-01 6.87009215e-01 5.65090440e-02
3.24872226e-01 -9.76492539e-02 1.54938549e-01 1.08253948e-01
-7.55336165e-01 2.01138761e-02 5.81802785e-01 9.20502782e-01
8.58558297e-01 -1.01302028e-01 -5.73660195e-01 9.73905325e-01
2.18145683e-01 6.61232233e-01 5.11369169e-01 -6.92243993e-01
2.57385701e-01 4.60974455e-01 3.44813764e-01 -1.10213780e+00
-1.74960524e-01 -3.94255631e-02 -1.10021925e+00 -9.08700377e-02
2.98917949e-01 1.10739870e-02 -7.53595054e-01 1.94804835e+00
4.91696596e-01 6.26010537e-01 1.87398225e-01 1.07880163e+00
1.08316255e+00 3.60167861e-01 4.75059301e-01 -1.88825592e-01
1.90734708e+00 -7.96735227e-01 -8.37387323e-01 -3.43952388e-01
2.48965740e-01 -5.93870997e-01 7.34810770e-01 -1.67914629e-01
-6.56657696e-01 -8.88379514e-01 -8.38859677e-01 2.41687134e-01
-5.33204734e-01 1.66542351e-01 4.04566437e-01 9.42917943e-01
-9.71211493e-01 1.23110734e-01 -3.15605849e-01 -9.97139752e-01
4.10056740e-01 2.89633989e-01 -7.81945705e-01 -2.82217920e-01
-1.41345322e+00 6.02225602e-01 4.29742783e-01 1.25581352e-02
-8.61146331e-01 -3.83414537e-01 -1.19961166e+00 -2.47135237e-02
4.37541157e-01 -8.16261590e-01 7.48063326e-01 -9.08196032e-01
-1.10414505e+00 1.47916961e+00 -6.05578542e-01 -2.17801973e-01
2.48973817e-01 7.04216361e-02 -8.66225004e-01 2.20009714e-01
4.24467772e-01 5.75910091e-01 1.08621097e+00 -1.35992432e+00
-6.95795178e-01 -7.90808558e-01 -3.52659822e-02 3.61377567e-01
-6.91356659e-01 5.29190838e-01 -8.52447629e-01 -6.53985083e-01
-2.10318714e-01 -7.88612008e-01 3.39973599e-01 -3.74921300e-02
-4.81428564e-01 -3.94964367e-01 8.93241048e-01 -8.50429833e-01
1.27186382e+00 -2.29244232e+00 1.07897624e-01 6.93007037e-02
6.25828207e-02 3.67800385e-01 -1.18846312e-01 3.20788324e-01
-6.13567196e-02 8.05982500e-02 -2.00834110e-01 -6.61234558e-01
9.33835506e-02 -1.74269870e-01 -1.55131370e-01 6.11960411e-01
-6.59186840e-02 9.67625797e-01 -8.51593852e-01 -7.37797499e-01
2.40654767e-01 5.62475502e-01 8.52203295e-02 2.62581289e-01
6.06401384e-01 5.50064385e-01 -3.61744702e-01 1.09114480e+00
6.36956513e-01 -2.78798461e-01 1.03646955e-02 -5.18693149e-01
1.10088341e-01 -6.12664878e-01 -1.29598844e+00 1.61024714e+00
-2.11961359e-01 2.85096228e-01 1.48099348e-01 -1.02840400e+00
7.04099715e-01 6.17308736e-01 6.14077687e-01 -5.66308439e-01
-8.75866041e-02 -1.32848239e-02 -4.14802969e-01 -3.36220205e-01
4.36474204e-01 1.28017977e-01 -3.07966888e-01 4.49839830e-01
1.47810027e-01 8.08750749e-01 1.21555313e-01 6.46297038e-02
6.24704182e-01 -5.80507666e-02 5.36075830e-01 -2.84180604e-02
9.46603596e-01 -2.53270328e-01 6.71107471e-01 8.11040342e-01
-8.30093682e-01 7.44695723e-01 -2.47710809e-01 -3.59258026e-01
-8.47877443e-01 -8.67016673e-01 -1.37239486e-01 1.26796722e+00
6.29439354e-01 -2.63561517e-01 -8.69011581e-01 -6.02466285e-01
-2.25877799e-02 2.69032657e-01 -6.65951669e-01 -3.02015424e-01
-3.24693590e-01 -7.96948493e-01 8.30450118e-01 5.58524787e-01
1.12652755e+00 -1.04317725e+00 -4.15261447e-01 -1.01528607e-01
-5.29394567e-01 -1.28365362e+00 -1.07828259e+00 -5.70810318e-01
-1.03941202e-01 -1.29947603e+00 -1.13537705e+00 -1.08511007e+00
5.55049121e-01 6.95638835e-01 8.63109231e-01 2.90085673e-02
-4.71806258e-01 9.84415591e-01 -2.84546584e-01 -8.40119347e-02
1.28326148e-01 -1.84214801e-01 5.23632288e-01 8.55888486e-01
8.35029900e-01 -3.68161090e-02 -8.18016469e-01 6.46809995e-01
-5.79600930e-01 -3.88662368e-01 2.13525370e-01 9.88478243e-01
4.25075352e-01 -4.48818952e-02 6.95050657e-01 -3.63565773e-01
5.01257241e-01 -3.96413237e-01 -7.05764145e-02 7.08639026e-01
-2.70763934e-01 -4.62517709e-01 3.88590276e-01 -5.69242656e-01
-1.15507174e+00 2.60687470e-01 6.96431547e-02 -6.17148042e-01
-5.22333622e-01 1.30174667e-01 -5.82681000e-01 -3.79230112e-01
4.75879818e-01 5.63811958e-01 1.15424074e-01 -3.68939310e-01
1.83251590e-01 9.57765818e-01 8.81382406e-01 -2.84548849e-01
9.57824707e-01 5.36798835e-01 -2.93450147e-01 -9.90442753e-01
-5.97467780e-01 -1.04826355e+00 -7.59928644e-01 -3.74100327e-01
1.07755554e+00 -1.15811658e+00 -9.48643565e-01 9.04540479e-01
-9.61034656e-01 2.41871044e-01 -1.53959366e-02 2.48641640e-01
-3.37749630e-01 7.64788032e-01 -2.94939935e-01 -1.04967237e+00
-5.19954681e-01 -9.70452666e-01 1.39911020e+00 5.10187209e-01
-1.57143041e-01 -9.10467744e-01 -1.35164052e-01 4.69976455e-01
3.26370388e-01 1.34405747e-01 4.36197579e-01 -7.47300684e-01
-2.68512577e-01 -4.33325857e-01 -4.96329397e-01 -9.35783535e-02
4.11363244e-01 -6.43914104e-01 -1.23468769e+00 -6.25553370e-01
-2.94863313e-01 -4.48921859e-01 8.69454384e-01 1.99641466e-01
9.43215787e-01 -2.00379148e-01 -6.42606080e-01 7.26118982e-01
1.12341142e+00 2.68065393e-01 4.50755864e-01 2.71494746e-01
9.46848929e-01 6.52874410e-01 6.10033691e-01 5.37134290e-01
4.78794754e-01 1.09388149e+00 1.81514192e-02 -1.72500595e-01
-2.38271207e-01 -2.90364146e-01 9.27861929e-02 1.22041486e-01
-1.91283137e-01 -2.21122220e-01 -8.85624528e-01 5.72273016e-01
-1.99941587e+00 -1.31080568e+00 4.80184793e-01 2.57648945e+00
4.56041515e-01 -6.09118164e-01 5.16887963e-01 1.34193450e-01
1.40726972e+00 2.46062860e-01 -6.88507915e-01 4.33834940e-01
-3.01547259e-01 -3.36274892e-01 2.95062691e-01 1.84228927e-01
-1.73058486e+00 9.31195021e-01 5.82165051e+00 9.03584361e-01
-8.25714231e-01 2.49663606e-01 5.79484701e-01 3.66409987e-01
2.18233645e-01 -3.28098446e-01 -1.11390495e+00 6.55936003e-01
6.94288254e-01 -3.35408211e-01 3.92753184e-01 6.81838632e-01
-9.32794809e-02 8.59159678e-02 -1.06654787e+00 1.87433076e+00
6.29959524e-01 -9.82928634e-01 1.68452814e-01 -7.73077458e-02
3.41018677e-01 -7.39459693e-01 9.32628810e-02 2.99768984e-01
-2.20008492e-01 -1.09037757e+00 4.49871063e-01 5.92204809e-01
1.23977566e+00 -7.35081077e-01 7.38444149e-01 1.47492886e-01
-1.83279145e+00 -2.20248893e-01 -1.56123608e-01 2.95274317e-01
2.81059384e-01 -2.20067039e-01 -2.52985597e-01 7.40458727e-01
8.89094710e-01 8.15984905e-01 -5.98505616e-01 1.10831523e+00
2.54753590e-01 2.98892647e-01 -5.89583479e-02 3.97350997e-01
-1.88475117e-01 -9.71545205e-02 5.11238635e-01 1.11593652e+00
3.83297205e-01 1.89697176e-01 5.37196755e-01 5.71596920e-01
-1.36263937e-01 1.39091834e-01 -6.73231423e-01 2.03977346e-01
7.71529078e-01 1.09691107e+00 -4.78062004e-01 -4.43599224e-01
-4.99466002e-01 1.50549948e+00 -1.92411654e-02 5.79399228e-01
-7.49915123e-01 -3.95070672e-01 9.16022718e-01 -1.49950594e-01
3.23798954e-02 2.11188048e-01 1.58603936e-01 -1.37107861e+00
-5.38512766e-02 -7.90165246e-01 8.58343184e-01 -4.99157637e-01
-1.49741542e+00 5.19042552e-01 1.46143854e-01 -1.24921143e+00
-4.87171769e-01 -3.43506694e-01 -5.71660221e-01 1.08613312e+00
-1.33763254e+00 -1.60510623e+00 -7.90346384e-01 1.11891842e+00
5.03808320e-01 -5.61945081e-01 1.03885281e+00 6.48886800e-01
-9.79596853e-01 1.08893859e+00 -3.04211280e-03 5.42677879e-01
9.29425538e-01 -8.19198012e-01 2.63692468e-01 9.27006483e-01
-1.24092745e-02 6.91184282e-01 3.94430846e-01 -6.21748924e-01
-1.53996789e+00 -1.24920869e+00 8.12683403e-01 -3.88894379e-01
2.78422385e-01 -2.52873868e-01 -9.33270514e-01 5.55905819e-01
-2.73318850e-02 2.15177238e-01 7.82458007e-01 -8.91671628e-02
-4.34892416e-01 -1.99913025e-01 -1.20477819e+00 3.69211376e-01
1.38596845e+00 -1.01306248e+00 -2.24441186e-01 4.01835531e-01
2.17378274e-01 5.99521846e-02 -8.72879863e-01 4.52168792e-01
5.32359600e-01 -6.16501689e-01 1.60961688e+00 -5.19277573e-01
-4.04944152e-01 -4.66349363e-01 -2.13329569e-01 -8.64111364e-01
-4.31320041e-01 -4.49867547e-01 -2.42780179e-01 1.62600255e+00
-3.65086734e-01 -6.73903108e-01 6.45212710e-01 6.34255528e-01
3.53268415e-01 -1.13432601e-01 -9.77513731e-01 -9.03095424e-01
-4.98185754e-01 5.59623428e-02 5.77280879e-01 1.00508583e+00
-2.03826293e-01 2.30015770e-01 -8.82513046e-01 3.31612289e-01
1.22356033e+00 1.71289578e-01 6.60807669e-01 -1.34807646e+00
1.67409927e-01 -3.04452807e-01 -7.98682392e-01 -7.59185493e-01
4.42610472e-01 -7.13066578e-01 -1.31828517e-01 -1.28494418e+00
7.65196085e-01 -7.70512596e-02 -4.31138009e-01 4.77690786e-01
-4.65881765e-01 6.46435797e-01 1.96002334e-01 7.76988924e-01
-8.53456438e-01 5.35371780e-01 5.08999765e-01 -4.53816801e-01
-5.70052564e-02 2.50696559e-02 -6.99992359e-01 5.98965049e-01
5.64540803e-01 8.08746144e-02 -2.23218128e-01 -6.58337921e-02
-6.88304305e-01 -1.28802639e-02 8.95495117e-01 -1.08674753e+00
5.40252030e-01 3.15115675e-02 5.55224419e-01 -3.47886533e-01
6.30799174e-01 -6.17819130e-01 2.45713413e-01 2.30781019e-01
-1.27246991e-01 -6.87516928e-02 -2.08780672e-02 8.39162946e-01
-4.86578286e-01 -2.89275050e-02 8.38326395e-01 -1.43346190e-01
-1.65699983e+00 7.24287212e-01 2.96519045e-02 1.94691699e-02
1.26830196e+00 -5.46140134e-01 -2.69766539e-01 -3.77637655e-01
-6.69764817e-01 2.41063565e-01 5.82413554e-01 5.92212379e-01
7.76966393e-01 -1.60465431e+00 -7.77361035e-01 2.60331541e-01
6.18390799e-01 -7.26498485e-01 6.60626888e-01 6.19637728e-01
1.64987996e-01 4.96417940e-01 -3.66757244e-01 -5.66334486e-01
-1.57115555e+00 9.19630885e-01 4.07749683e-01 1.61162376e-01
-5.89513242e-01 6.12945259e-01 4.64727312e-01 -1.92083120e-01
5.17149866e-01 8.69357765e-01 -6.11382961e-01 2.52115607e-01
1.07759309e+00 3.90932739e-01 -3.93191159e-01 -1.51345265e+00
-8.20993781e-01 9.53767776e-01 2.16181800e-01 6.96187615e-02
5.55615664e-01 -5.15651882e-01 6.70862868e-02 2.31497690e-01
1.20013154e+00 -4.35509950e-01 -1.02519763e+00 -5.63918948e-01
-4.97736000e-02 -6.26393616e-01 -1.96913972e-01 -4.66326058e-01
-9.60297704e-01 7.16309249e-01 1.12011123e+00 -3.33539918e-02
1.21267760e+00 -3.51715088e-03 7.34679282e-01 3.17123413e-01
6.68757379e-01 -9.88935649e-01 4.01848406e-02 1.04014106e-01
7.10627139e-01 -1.74505877e+00 -2.87765086e-01 -4.54078615e-01
-7.66268849e-01 7.12256730e-01 6.50867164e-01 4.98717934e-01
5.58132231e-01 -2.98248768e-01 -8.23452622e-02 1.23271354e-01
1.00907469e-02 -7.25388885e-01 4.67016459e-01 9.23052490e-01
9.65096578e-02 -6.58951104e-02 1.42079532e-01 7.11343527e-01
3.97000343e-01 -9.54227075e-02 -2.25918993e-01 8.02443206e-01
-3.41639549e-01 -8.59984934e-01 -7.96532691e-01 9.23834890e-02
-2.55415291e-01 1.80080794e-02 -3.62588018e-01 5.94156086e-01
-2.83547095e-03 1.23352134e+00 -3.45623074e-03 -4.66889650e-01
2.60099858e-01 3.97914141e-01 2.47233376e-01 -3.63337010e-01
-3.17554057e-01 -3.06351036e-01 -5.21887690e-02 -4.89035100e-01
-7.07920134e-01 -6.51341736e-01 -7.39867747e-01 -3.26498926e-01
3.25402319e-02 -2.33018268e-02 2.27636307e-01 8.48004699e-01
4.53541011e-01 7.67954141e-02 5.20061255e-01 -1.02636838e+00
-4.80543554e-01 -7.28130579e-01 -7.13705599e-01 1.06376994e+00
1.75873831e-01 -9.84542966e-01 -2.06135079e-01 2.92726636e-01] | [14.607255935668945, 0.9910005331039429] |
8c01c433-3de0-418b-b7b5-63aa2ca41881 | challenging-environments-for-traffic-sign | 1902.06857 | null | https://arxiv.org/abs/1902.06857v2 | https://arxiv.org/pdf/1902.06857v2.pdf | Challenging Environments for Traffic Sign Detection: Reliability Assessment under Inclement Conditions | State-of-the-art algorithms successfully localize and recognize traffic signs over existing datasets, which are limited in terms of challenging condition type and severity. Therefore, it is not possible to estimate the performance of traffic sign detection algorithms under overlooked challenging conditions. Another shortcoming of existing datasets is the limited utilization of temporal information and the unavailability of consecutive frames and annotations. To overcome these shortcomings, we generated the CURE-TSD video dataset and hosted the first IEEE Video and Image Processing (VIP) Cup within the IEEE Signal Processing Society. In this paper, we provide a detailed description of the CURE-TSD dataset, analyze the characteristics of the top performing algorithms, and provide a performance benchmark. Moreover, we investigate the robustness of the benchmarked algorithms with respect to sign size, challenge type and severity. Benchmarked algorithms are based on state-of-the-art and custom convolutional neural networks that achieved a precision of 0.55 and a recall of 0.32, F0.5 score of 0.48 and F2 score of 0.35. Experimental results show that benchmarked algorithms are highly sensitive to tested challenging conditions, which result in an average performance drop of 0.17 in terms of precision and a performance drop of 0.28 in recall under severe conditions. The dataset is publicly available at https://github.com/olivesgatech/CURE-TSD. | ['Min-Hung Chen', 'Tariq Alshawi', 'Ghassan AlRegib', 'Dogancan Temel'] | 2019-02-19 | null | null | null | null | ['traffic-sign-detection'] | ['computer-vision'] | [-1.92463081e-02 -6.67348385e-01 -1.29443541e-01 -1.60749748e-01
-8.24127793e-01 -5.68160772e-01 4.24184442e-01 -2.88066894e-01
-4.46090162e-01 6.52502716e-01 -1.30725428e-01 -3.23224634e-01
-1.48827925e-01 -3.16452175e-01 -5.43539464e-01 -6.05812728e-01
-3.19370031e-01 -4.00342420e-02 7.01701403e-01 1.01915926e-01
1.01064332e-01 4.64348465e-01 -2.00462675e+00 3.52805525e-01
6.21726692e-01 1.36403704e+00 -1.07417531e-01 9.20738399e-01
2.74576843e-01 5.81846416e-01 -9.80474710e-01 -3.00712228e-01
4.18006510e-01 -8.92674923e-02 -2.38191187e-01 -1.04873948e-01
9.75968122e-01 -6.09522581e-01 -6.95757270e-01 6.02688670e-01
8.34844530e-01 -1.64171308e-01 3.05011123e-01 -1.72219229e+00
-2.40887240e-01 -4.04686667e-02 -4.18314308e-01 6.94129467e-01
1.05800979e-01 6.01861596e-01 7.97468245e-01 -7.10183382e-01
7.48712897e-01 7.62395561e-01 8.05046380e-01 4.17839348e-01
-8.71446133e-01 -7.95175612e-01 9.03236028e-03 6.80882573e-01
-1.40531230e+00 -6.61700010e-01 4.86591607e-01 -4.23007876e-01
7.73757339e-01 1.96561128e-01 3.35303366e-01 1.01447666e+00
-2.61451244e-01 8.33128393e-01 1.13217199e+00 -1.52439550e-01
1.52057543e-01 -2.64274869e-02 2.54967064e-01 5.72993755e-01
4.83348459e-01 1.87705040e-01 -5.50479293e-01 5.24536036e-02
3.56266707e-01 -4.49865371e-01 -2.86883831e-01 -2.04481333e-01
-1.19719326e+00 3.00593555e-01 2.86472768e-01 3.37839156e-01
-4.03599769e-01 2.53484309e-01 7.72917330e-01 3.26057404e-01
1.19535096e-01 -1.46661531e-02 -3.83057535e-01 -5.92790663e-01
-1.15382266e+00 2.52621025e-01 6.49995148e-01 7.80532777e-01
1.34991124e-01 2.95736820e-01 -4.98133868e-01 6.30539656e-01
-2.26823352e-02 8.60664606e-01 1.68379575e-01 -9.84820843e-01
5.60899258e-01 2.60191381e-01 1.44454837e-01 -9.94906843e-01
-4.17626530e-01 -5.89529395e-01 -4.39763308e-01 3.91558647e-01
7.84494996e-01 -2.05477729e-01 -1.07110500e+00 1.60424972e+00
-2.23035086e-02 5.37972927e-01 -2.53086090e-01 1.16217387e+00
9.03394759e-01 2.55034447e-01 1.39230400e-01 -1.54283121e-02
1.23961544e+00 -6.67400479e-01 -5.69554269e-01 -2.95895636e-02
5.75755179e-01 -6.97313070e-01 9.51658010e-01 4.57706720e-01
-9.17292893e-01 -6.11556828e-01 -1.04118228e+00 5.08362472e-01
-3.97013217e-01 5.89580178e-01 4.04217124e-01 9.68344152e-01
-1.07065034e+00 1.99695900e-01 -6.31786406e-01 -6.82639122e-01
6.11225605e-01 2.12390646e-01 -2.78393000e-01 -1.42914355e-01
-9.50306773e-01 8.66191804e-01 7.42594600e-02 1.76700816e-01
-8.52169573e-01 -9.08742428e-01 -3.12475324e-01 -3.27943116e-01
3.10021579e-01 -1.54382527e-01 1.15944958e+00 -5.13918579e-01
-9.71588671e-01 9.20298398e-01 -1.33957416e-01 -7.47563779e-01
9.44452703e-01 -1.61363736e-01 -1.02985024e+00 2.93816805e-01
3.39770429e-02 4.64299262e-01 5.40374517e-01 -9.74653602e-01
-8.56125772e-01 1.01544321e-01 -7.11897910e-02 -2.89560586e-01
1.57916341e-02 2.60236412e-01 -7.86521614e-01 -2.81254083e-01
-2.42508680e-01 -9.72924232e-01 3.09493870e-01 2.50667065e-01
-4.27167118e-02 1.12502307e-01 1.18000352e+00 -6.96860909e-01
1.22775543e+00 -2.15165639e+00 -7.27803230e-01 9.59297195e-02
2.25352690e-01 8.91334057e-01 -3.04572225e-01 1.69462234e-01
1.01160906e-01 3.82955670e-02 6.30711317e-02 3.89622562e-02
1.94919392e-01 9.33861956e-02 -5.74478582e-02 6.71257794e-01
1.05329640e-01 8.28817308e-01 -7.75187552e-01 -2.77973294e-01
5.64949870e-01 5.26735723e-01 -2.54845768e-01 -6.60888702e-02
1.89921424e-01 2.58094221e-01 -3.44355702e-01 1.04312992e+00
7.78695762e-01 6.03729673e-02 -2.03382298e-01 -5.97950041e-01
-3.91623110e-01 1.80569738e-01 -1.27785647e+00 1.02639103e+00
-3.16166967e-01 1.27685952e+00 -3.54832113e-02 -8.53555799e-01
8.15507233e-01 3.94150406e-01 6.48249209e-01 -9.64091122e-01
2.70080149e-01 5.91472805e-01 1.48843050e-01 -6.96021020e-01
2.49829099e-01 3.99424702e-01 1.60576373e-01 2.06648111e-01
-5.58515675e-02 1.83700740e-01 8.27046096e-01 1.40037701e-01
1.50768781e+00 -2.08081976e-01 -2.13995859e-01 1.29660249e-01
6.15452528e-01 -2.07030714e-01 4.62661326e-01 8.56938899e-01
-9.13202822e-01 6.46602273e-01 5.71886957e-01 -2.56056309e-01
-9.66649711e-01 -1.22238219e+00 -3.16556394e-01 7.18732536e-01
5.33984415e-03 -2.25966930e-01 -5.36864400e-01 -5.60709655e-01
4.42707837e-02 5.31504691e-01 -5.27597010e-01 -4.40159291e-02
-7.02653646e-01 -7.23699689e-01 1.22082794e+00 6.95193350e-01
1.00267553e+00 -8.34153593e-01 -8.19867134e-01 -3.94157544e-02
-2.38216490e-01 -1.76858044e+00 -2.55708218e-01 -3.69494349e-01
-4.00810152e-01 -1.35747612e+00 -8.58056128e-01 -3.95073920e-01
2.78954178e-01 2.16198623e-01 1.01431179e+00 9.77849364e-02
-3.87467116e-01 4.15406168e-01 -3.71834338e-01 -3.38603109e-01
-2.09696904e-01 -1.42774373e-01 4.44057994e-02 2.41129056e-01
4.00394708e-01 -3.92966449e-01 -8.38819027e-01 7.77028978e-01
-6.32036984e-01 -4.72406268e-01 6.30664468e-01 5.58971643e-01
2.28997573e-01 -2.07227975e-01 4.46900904e-01 -1.68654576e-01
4.85921413e-01 -2.46504173e-01 -6.92739248e-01 1.47131413e-01
-4.95519131e-01 -4.04064119e-01 1.27848029e-01 -3.17798585e-01
-8.90673578e-01 -1.54437706e-01 1.88087486e-02 -5.15286744e-01
-4.80372876e-01 9.54565629e-02 1.32182270e-01 -2.70419896e-01
6.81778371e-01 1.63673088e-01 -1.09608978e-01 -3.60639840e-01
6.44431040e-02 8.05000544e-01 8.18181574e-01 -3.21156561e-01
7.44131923e-01 5.81477880e-01 9.46170539e-02 -9.94714499e-01
-6.23619258e-01 -6.36185527e-01 -3.76427442e-01 -8.28067482e-01
6.19885862e-01 -8.51366103e-01 -8.10189366e-01 8.24380159e-01
-7.12280571e-01 -2.36868009e-01 -8.69977325e-02 6.32082939e-01
-3.62094700e-01 4.24038857e-01 -3.38392824e-01 -7.53175080e-01
-2.24791422e-01 -9.64960158e-01 9.18706596e-01 2.78556794e-01
-1.29654735e-01 -5.40784299e-01 -2.39718288e-01 7.14697540e-01
8.80951047e-01 5.22538960e-01 9.32875648e-02 -4.29706693e-01
-6.03932381e-01 -6.57465219e-01 -6.58590496e-01 4.90706384e-01
-1.84714869e-01 1.49029881e-01 -1.09183621e+00 -2.39698663e-01
-6.77158594e-01 -3.37454617e-01 9.55248237e-01 5.06275654e-01
8.01983893e-01 1.46057472e-01 -2.84655422e-01 4.00076061e-01
1.11280882e+00 1.03775106e-01 7.08790123e-01 4.90739763e-01
2.25021899e-01 3.14900815e-01 6.82730675e-01 3.01616848e-01
3.12676392e-02 9.40467000e-01 3.31891716e-01 6.83327615e-02
-7.29454041e-01 -2.73531936e-02 3.87542844e-01 1.96119428e-01
-2.13999227e-01 -4.46942329e-01 -1.04390466e+00 8.19328308e-01
-1.55558479e+00 -1.19599354e+00 -6.45891905e-01 2.25137591e+00
2.60306507e-01 3.57254088e-01 4.77540880e-01 5.36392987e-01
8.37384880e-01 1.95801139e-01 -2.88655847e-01 -1.47470400e-01
-3.97548705e-01 5.57071269e-02 7.75867462e-01 6.00535013e-02
-1.16873026e+00 6.85085714e-01 5.99261951e+00 6.22744560e-01
-1.48094022e+00 1.37080416e-01 1.82902157e-01 -3.35903108e-01
5.77064455e-01 -3.41962814e-01 -7.62783289e-01 7.99440861e-01
1.23377931e+00 -3.78094502e-02 7.29922354e-02 4.99512583e-01
5.98833382e-01 -2.33253077e-01 -5.80482483e-01 8.94706786e-01
1.63511708e-01 -1.15859509e+00 -4.49163496e-01 -6.61967546e-02
6.17980242e-01 6.34650767e-01 1.49679497e-01 3.20412934e-01
-1.98775575e-01 -7.52147079e-01 7.55748630e-01 3.81932437e-01
1.10194719e+00 -3.26819867e-01 9.87180889e-01 -1.35927603e-01
-1.32053554e+00 -1.64757252e-01 2.47945003e-02 5.15926704e-02
5.27833104e-01 5.92436254e-01 -7.16405213e-01 3.06809336e-01
9.56224501e-01 5.04005134e-01 -6.54997945e-01 1.69355714e+00
-2.07264036e-01 1.11134124e+00 -6.93174243e-01 -1.08549818e-02
2.68493295e-01 4.19886053e-01 7.61239171e-01 1.54969728e+00
3.92667353e-01 -2.27166951e-01 3.73189338e-02 5.62566102e-01
8.59277323e-02 -2.15666562e-01 -2.33132333e-01 2.23172769e-01
4.64182138e-01 1.10021925e+00 -5.04383147e-01 -3.66456091e-01
-4.59101439e-01 5.60135782e-01 -3.91387135e-01 5.84413230e-01
-1.32081330e+00 -3.54559541e-01 6.76632524e-01 3.70276272e-01
5.69865584e-01 -2.61874378e-01 -2.58724511e-01 -1.01713622e+00
5.69107234e-01 -6.50599062e-01 4.84593183e-01 -8.11023593e-01
-1.01763856e+00 4.12495852e-01 3.00658005e-03 -1.57646871e+00
9.27111804e-02 -7.28669882e-01 -7.07663596e-01 5.47754109e-01
-1.52996075e+00 -1.17846298e+00 -7.72748172e-01 4.78528053e-01
3.52878571e-01 -3.56129289e-01 3.85460675e-01 9.00528014e-01
-6.66642427e-01 9.88057613e-01 3.64265144e-02 2.84167022e-01
7.49252439e-01 -7.75505126e-01 2.92730719e-01 1.08951879e+00
-2.64611632e-01 5.61640672e-02 7.44904101e-01 -3.33786339e-01
-9.57270086e-01 -1.16788626e+00 9.11333025e-01 -2.97877431e-01
8.96367610e-01 -1.32785991e-01 -6.08687520e-01 3.74621034e-01
-1.87366441e-01 4.79872853e-01 2.04898983e-01 -2.42903143e-01
-5.11950254e-01 -4.38122183e-01 -1.18266869e+00 4.81654674e-01
1.08883226e+00 -3.20297331e-01 -2.08311185e-01 3.18079107e-02
-2.33245473e-02 -4.36118186e-01 -6.81376696e-01 5.69079340e-01
9.45366740e-01 -1.04615271e+00 9.85474586e-01 -2.79382169e-01
-6.64925799e-02 -7.04421878e-01 -2.05949903e-01 -7.17539012e-01
-4.76257550e-03 -2.92580575e-01 -3.15443993e-01 1.33181512e+00
5.30055463e-01 -6.62126720e-01 7.85453320e-01 4.65940535e-01
-1.56558558e-01 -4.96044576e-01 -1.20477819e+00 -1.22413325e+00
-3.50584328e-01 -8.45807791e-01 2.36061171e-01 5.50826788e-01
-4.62997079e-01 -2.48737022e-01 -4.05772179e-01 1.58946678e-01
7.06132948e-01 -1.70366138e-01 9.68627691e-01 -9.00755048e-01
4.96001868e-03 -5.53453326e-01 -9.96667325e-01 -7.35832632e-01
-2.28655741e-01 -5.03195107e-01 -9.14264843e-02 -1.35125291e+00
-8.19587559e-02 -2.48404875e-01 -6.31985664e-01 4.90423650e-01
6.02950975e-02 7.49728262e-01 3.83934617e-01 1.52851969e-01
-8.26780081e-01 2.73167714e-02 6.64343536e-01 -1.23747289e-01
-3.10739093e-02 1.28349677e-01 -2.19225571e-01 4.24907655e-01
1.31203914e+00 -2.12199494e-01 -8.89640003e-02 -3.57355267e-01
-2.97556937e-01 -3.52671593e-01 9.34862852e-01 -1.55094123e+00
1.51113749e-01 2.15415284e-01 2.52819717e-01 -8.34884942e-01
3.42871994e-01 -6.09964848e-01 -5.96735552e-02 5.79283774e-01
-8.68024379e-02 -2.03867868e-01 5.27473807e-01 3.53001624e-01
-2.73013085e-01 2.01652259e-01 7.41515100e-01 4.08375442e-01
-1.15897930e+00 1.01382919e-01 -4.63749737e-01 2.23961085e-01
1.07096756e+00 -6.01444244e-01 -5.75538039e-01 -4.40747470e-01
-2.28998542e-01 2.67963380e-01 1.58258736e-01 7.12858319e-01
3.48286837e-01 -1.22628891e+00 -9.56651509e-01 7.32246637e-02
3.41218382e-01 -7.88815081e-01 5.35519242e-01 1.39515030e+00
-7.23110855e-01 6.88566208e-01 -4.32987332e-01 -8.42007935e-01
-1.59160483e+00 5.44594042e-02 5.66401482e-01 -8.98183063e-02
-4.47340816e-01 4.57631350e-01 -3.87338728e-01 -1.00520998e-01
4.15698767e-01 -3.01373631e-01 3.04732006e-02 -6.30134568e-02
5.95178366e-01 8.14582825e-01 2.97580302e-01 -8.28784108e-01
-6.50700808e-01 7.38025963e-01 2.10362136e-01 1.44630363e-02
1.19756317e+00 2.09735334e-03 5.31477273e-01 -7.63838366e-02
1.17227936e+00 -1.77858099e-01 -1.12967026e+00 2.54370589e-02
-9.50974822e-02 -6.06660426e-01 1.80524796e-01 -1.16493118e+00
-1.31654060e+00 6.76416636e-01 1.07575035e+00 -2.32068673e-02
1.36547506e+00 -1.70023978e-01 9.29021657e-01 2.18733758e-01
5.10961562e-02 -1.08641648e+00 -2.72354990e-01 4.78364319e-01
7.15643346e-01 -1.14700854e+00 -1.84402943e-01 -3.32245648e-01
-5.42159975e-01 8.74123514e-01 6.14047050e-01 1.33306563e-01
5.77838480e-01 4.36820894e-01 3.31114531e-01 -1.66250110e-01
-7.13409841e-01 -5.74552298e-01 3.86590362e-01 7.83319890e-01
3.88248831e-01 -2.23279502e-02 -5.26678085e-01 2.29612768e-01
-3.34612727e-02 4.35734093e-01 2.87741959e-01 9.87083852e-01
-3.10422391e-01 -6.15977287e-01 -4.22541142e-01 3.58976364e-01
-4.87861186e-01 3.74588966e-01 -4.07970577e-01 9.72773850e-01
3.52560103e-01 1.16050446e+00 -5.14962599e-02 -4.59392697e-01
8.25079381e-01 1.29813269e-01 1.91197038e-01 2.20236078e-01
-3.34311754e-01 -2.94971824e-01 5.99733591e-01 -7.43637741e-01
-4.49533403e-01 -8.55357349e-01 -9.89031553e-01 -4.68109936e-01
-1.68245196e-01 -1.77770793e-01 6.04916334e-01 6.73804641e-01
5.31024098e-01 5.21001339e-01 2.64727563e-01 -6.17426157e-01
-2.83927500e-01 -8.44406486e-01 -4.32552904e-01 6.27270937e-01
2.88763821e-01 -7.52509356e-01 -2.98120201e-01 5.33310771e-02] | [8.008650779724121, -0.8316307663917542] |
1cf8d1b3-1fc0-4136-a228-781d81f8047e | residual-expansion-algorithm-fast-and | 1705.09549 | null | http://arxiv.org/abs/1705.09549v1 | http://arxiv.org/pdf/1705.09549v1.pdf | Residual Expansion Algorithm: Fast and Effective Optimization for Nonconvex Least Squares Problems | We propose the residual expansion (RE) algorithm: a global (or near-global)
optimization method for nonconvex least squares problems. Unlike most existing
nonconvex optimization techniques, the RE algorithm is not based on either
stochastic or multi-point searches; therefore, it can achieve fast global
optimization. Moreover, the RE algorithm is easy to implement and successful in
high-dimensional optimization. The RE algorithm exhibits excellent empirical
performance in terms of k-means clustering, point-set registration, optimized
product quantization, and blind image deblurring. | ['Daiki Ikami', 'Toshihiko Yamasaki', 'Kiyoharu Aizawa'] | 2017-05-26 | residual-expansion-algorithm-fast-and-1 | http://openaccess.thecvf.com/content_cvpr_2017/html/Ikami_Residual_Expansion_Algorithm_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Ikami_Residual_Expansion_Algorithm_CVPR_2017_paper.pdf | cvpr-2017-7 | ['blind-image-deblurring'] | ['computer-vision'] | [-3.42371285e-01 -5.39308786e-01 -2.78214008e-01 2.69446410e-02
-1.30809796e+00 -2.60491222e-01 1.61984667e-01 -2.36852959e-01
-3.12200367e-01 7.44456351e-01 3.52201909e-01 7.94926435e-02
-4.05273080e-01 -2.91105092e-01 -5.34571648e-01 -9.69738483e-01
1.16940089e-01 4.69426811e-01 -5.45057915e-02 9.61459894e-03
5.23190856e-01 5.72026432e-01 -7.96914756e-01 -5.69837093e-01
1.12801075e+00 8.75560939e-01 1.69699579e-01 4.52656418e-01
4.11653280e-01 3.74178559e-01 -3.15853596e-01 5.33829890e-02
2.81234115e-01 -2.41120189e-01 -6.07604086e-01 1.79695949e-01
3.57064813e-01 -2.53092796e-01 -2.26744220e-01 1.32242608e+00
7.47078657e-01 5.01036048e-01 7.82070756e-01 -8.59866083e-01
-1.13191545e+00 -9.93558913e-02 -7.83092499e-01 5.07052690e-02
4.12161380e-01 -7.92586133e-02 7.57534802e-01 -1.47640157e+00
3.88596088e-01 1.09553134e+00 1.04359460e+00 -7.24125504e-02
-1.03325748e+00 -3.22957307e-01 -3.06006491e-01 7.77101964e-02
-2.07453036e+00 -6.89954400e-01 6.69497430e-01 -3.98769587e-01
6.19353890e-01 5.06928742e-01 4.13546681e-01 1.47387370e-01
1.84858739e-01 5.97203314e-01 8.69714379e-01 -2.69151062e-01
3.79467070e-01 -1.89919293e-01 -1.58328697e-01 6.18533313e-01
2.24440560e-01 8.46830383e-02 -4.78199571e-01 -4.79873478e-01
1.16137934e+00 -1.00395635e-01 -7.41684675e-01 -3.90703201e-01
-1.62908661e+00 9.08067048e-01 4.13994104e-01 2.78905392e-01
-6.28747582e-01 3.92669201e-01 2.80253142e-02 -8.54456648e-02
5.73095202e-01 5.09077311e-01 -1.38473436e-01 -1.96863100e-01
-1.32053089e+00 1.53111324e-01 5.89605987e-01 8.75511587e-01
7.97390342e-01 3.02908868e-01 -3.38914506e-02 9.77903962e-01
6.83861732e-01 1.03809607e+00 6.13099813e-01 -1.41689837e+00
4.45127457e-01 1.49398699e-01 3.46551269e-01 -1.81849635e+00
-4.84971523e-01 -3.45249236e-01 -1.14779687e+00 -2.65159439e-02
-3.47768441e-02 -9.62825492e-02 -5.15559137e-01 1.13196802e+00
3.78879756e-01 5.57523429e-01 -5.13115712e-02 1.29812396e+00
7.45864749e-01 7.87024677e-01 -5.34405768e-01 -5.77270687e-01
7.07506776e-01 -1.29116321e+00 -1.13989186e+00 -2.22826704e-01
6.02040946e-01 -9.32498455e-01 5.56209385e-01 2.33127460e-01
-1.25477827e+00 -1.61218986e-01 -8.77046287e-01 -1.44278869e-01
9.86557379e-02 3.25948268e-01 2.56057858e-01 5.22579491e-01
-1.27276838e+00 5.25050759e-01 -8.16735566e-01 -1.59231052e-02
1.94335610e-01 4.04101133e-01 -3.95227611e-01 -2.81689823e-01
-6.64337456e-01 7.83531189e-01 -3.95785682e-02 4.57357228e-01
-4.40119028e-01 -5.38661718e-01 -8.60732079e-01 -9.87600237e-02
1.48794845e-01 -5.61565757e-01 5.22253811e-01 -4.07411188e-01
-1.88542616e+00 5.95043242e-01 -8.77155483e-01 8.02988559e-02
2.98688680e-01 -3.55207413e-01 -2.23663300e-01 2.45310739e-01
1.51011959e-01 2.48920858e-01 1.09162438e+00 -1.22093308e+00
4.06221636e-02 -3.09026480e-01 -8.52465808e-01 5.87694049e-01
-2.29475021e-01 1.48828954e-01 -6.57128334e-01 -9.81251776e-01
7.59119928e-01 -1.06013668e+00 -5.67305386e-01 9.79014486e-02
-5.11658907e-01 1.67128652e-01 6.45384550e-01 -8.31352532e-01
1.42460489e+00 -2.35796928e+00 3.68885905e-01 5.19436717e-01
2.16566920e-01 3.70849594e-02 -1.29505292e-01 1.98939070e-01
-7.42727444e-02 2.20732182e-01 -4.66833562e-01 -5.18911898e-01
-3.48482095e-02 -5.81340715e-02 8.27072561e-02 1.27292073e+00
-5.07144809e-01 9.77468014e-01 -8.68215084e-01 -4.23378378e-01
4.26373720e-01 5.12103498e-01 -6.31911933e-01 -1.72407389e-01
6.13989592e-01 5.28455079e-01 -5.57127357e-01 7.18708932e-01
9.13129151e-01 -4.94833410e-01 -3.91201079e-01 -4.44167465e-01
-2.23780185e-01 -2.91485876e-01 -1.64776051e+00 1.76209521e+00
-1.08799301e-01 6.29876375e-01 4.50059921e-01 -9.41299140e-01
9.37612832e-01 4.62053269e-01 9.36404586e-01 -3.77744198e-01
-3.02234944e-02 5.67608416e-01 -8.00927341e-01 -2.09756643e-01
6.93591475e-01 1.05453074e-01 2.17548281e-01 3.21136743e-01
-4.38309073e-01 -5.00274062e-01 -1.03157826e-01 -1.64326116e-01
7.67801464e-01 -4.31345463e-01 3.57557744e-01 -5.81398070e-01
6.25036061e-01 1.05413504e-01 7.10522175e-01 5.31222284e-01
-3.03567171e-01 1.19517493e+00 -4.60973293e-01 -1.85181305e-01
-8.67039144e-01 -7.70281851e-01 -3.48292381e-01 5.69565535e-01
7.92645514e-01 -4.12875742e-01 -5.94536424e-01 -9.51964259e-02
-1.06853873e-01 3.86551768e-01 9.91712362e-02 -3.06365080e-02
-6.46249235e-01 -9.19663727e-01 1.83180675e-01 1.31715894e-01
6.19186342e-01 -4.92148966e-01 8.46163034e-02 2.76233405e-01
-5.03533781e-01 -1.06094301e+00 -1.37821531e+00 -4.98585373e-01
-1.18659568e+00 -1.05934942e+00 -1.21642900e+00 -1.00142479e+00
9.85709012e-01 6.78294659e-01 8.38585794e-01 2.03197479e-01
2.99348682e-02 5.72465003e-01 -1.89053848e-01 4.07907099e-01
7.14045316e-02 -3.72077048e-01 3.97683024e-01 5.21089025e-02
8.64448324e-02 -2.38747850e-01 -7.72191882e-01 8.43400300e-01
-5.63753605e-01 -5.97531259e-01 2.82974184e-01 8.79207671e-01
1.13642216e+00 4.49633926e-01 1.54803723e-01 -5.94400465e-02
8.82345915e-01 -2.98496187e-01 -6.22447848e-01 2.68889159e-01
-6.93688393e-01 -3.12381953e-01 2.93061286e-01 -2.48127952e-01
-6.90268755e-01 8.79290476e-02 -2.01057643e-02 -6.23115063e-01
4.04928654e-01 5.88766754e-01 2.12600634e-01 -9.94374335e-01
5.65454125e-01 4.41505909e-01 1.75344031e-02 -4.90309149e-01
5.38098395e-01 5.88714719e-01 5.71725547e-01 -2.20249385e-01
9.09339845e-01 6.67188525e-01 -7.68366084e-03 -1.10717535e+00
-3.39667141e-01 -8.35596979e-01 -6.81006849e-01 1.96117908e-02
8.86593819e-01 -9.90674257e-01 -5.82318366e-01 6.98782265e-01
-1.05593443e+00 -1.99599355e-01 -1.68255165e-01 7.90643990e-01
-7.23691761e-01 9.23277795e-01 -4.97292042e-01 -5.68171203e-01
-4.12010878e-01 -1.44407117e+00 1.10722661e+00 1.16515473e-01
-1.08538680e-01 -1.36177671e+00 1.25814036e-01 3.21743846e-01
4.69030470e-01 6.80052787e-02 1.88965261e-01 -5.97666316e-02
-4.85344410e-01 -4.49933767e-01 -1.83015168e-01 3.78497511e-01
2.56265372e-01 -1.27329752e-01 -1.63026690e-01 -6.32559478e-01
3.77198040e-01 3.46190073e-02 2.62321234e-01 1.01184344e+00
9.82049584e-01 -5.97061098e-01 -4.10922498e-01 1.03711498e+00
1.61974418e+00 -1.39782578e-01 7.67059445e-01 3.40280086e-01
8.69140744e-01 5.18115833e-02 5.52927613e-01 4.42683548e-01
6.27786100e-01 8.84134352e-01 2.45004967e-01 -2.12648392e-01
-2.17731148e-01 9.92542952e-02 3.58232886e-01 1.29483676e+00
2.39158645e-02 6.00935565e-03 -9.56964552e-01 7.17273295e-01
-1.96322143e+00 -7.74764478e-01 -6.11118197e-01 2.40548134e+00
7.00706005e-01 -7.84377694e-01 -3.22837353e-01 -1.08231746e-01
1.07228339e+00 1.03607669e-01 -5.02569377e-01 -3.23999599e-02
-3.90005380e-01 -3.01230550e-01 1.02438593e+00 8.03167164e-01
-1.15233493e+00 9.38071489e-01 8.29671383e+00 9.94847119e-01
-7.45211840e-01 2.82871544e-01 3.09259146e-01 1.38264373e-01
-2.89249927e-01 -2.36238614e-01 -6.17292643e-01 4.94625181e-01
3.87265503e-01 -2.70392388e-01 8.24511290e-01 6.12797737e-01
4.87068981e-01 -6.85536191e-02 -4.76491451e-01 1.59472680e+00
2.16739371e-01 -1.48161960e+00 -3.68562043e-01 1.24160558e-01
1.28098190e+00 2.26303697e-01 1.51887760e-01 -4.34223086e-01
2.46000156e-01 -1.15926719e+00 4.79931921e-01 5.09997845e-01
7.67630160e-01 -5.25829792e-01 6.47794664e-01 3.35789979e-01
-1.14401412e+00 -1.91371739e-02 -4.46196258e-01 4.10200089e-01
6.97899163e-01 1.05032587e+00 2.78700492e-03 3.10493886e-01
7.57932723e-01 1.01000130e+00 -1.34921819e-01 1.58474696e+00
-1.86193615e-01 4.35098201e-01 -7.69836485e-01 2.26398975e-01
1.86866507e-01 -7.64448524e-01 1.15113890e+00 8.21972787e-01
6.94070935e-01 3.40270072e-01 2.66704470e-01 6.19244814e-01
2.30560094e-01 4.20871168e-01 -2.40951732e-01 2.42271438e-01
7.09433973e-01 8.65488291e-01 -3.21036398e-01 -7.30958581e-02
-2.74563849e-01 1.16301501e+00 2.22012238e-03 8.46537530e-01
-5.18957019e-01 -4.34472173e-01 6.01692379e-01 -1.00401275e-01
3.90017360e-01 -7.11085796e-01 -5.56263447e-01 -1.32748079e+00
-7.98103362e-02 -7.28052378e-01 8.85891840e-02 -9.69877362e-01
-1.24460340e+00 3.26478332e-01 -3.73100668e-01 -1.39990878e+00
-7.84878805e-02 -3.31198871e-01 -4.89488423e-01 8.47549319e-01
-1.49399662e+00 -6.89718902e-01 -2.79017270e-01 9.43776786e-01
1.81742653e-01 -5.83303422e-02 6.06107831e-01 4.15405631e-01
-6.41679883e-01 5.65552354e-01 9.83742893e-01 -2.28504673e-01
7.81197071e-01 -1.01858783e+00 2.01759487e-02 9.32541668e-01
-3.05489898e-01 6.75455570e-01 5.65911353e-01 -7.56131113e-01
-1.72379315e+00 -9.13529038e-01 7.23301411e-01 -6.35804981e-02
6.67762756e-01 4.75518912e-01 -9.35991168e-01 5.07014096e-01
-7.01006725e-02 2.79787570e-01 6.14884794e-01 -2.34876335e-01
1.54584154e-01 4.43717986e-02 -1.38371551e+00 3.84230703e-01
6.71702325e-01 -3.69897276e-01 -2.36816555e-01 9.02455151e-01
3.51410806e-01 -6.92684472e-01 -1.29120564e+00 4.56903160e-01
2.79840287e-02 -4.83385384e-01 1.38442564e+00 1.53646156e-01
6.31301180e-02 -5.29365301e-01 -2.91919917e-01 -1.39151609e+00
-8.15540373e-01 -1.12464666e+00 -3.78721207e-01 9.56548333e-01
1.88313201e-01 -8.37927043e-01 7.70005405e-01 5.36429584e-01
-1.63553134e-01 -4.85793978e-01 -1.30080926e+00 -9.15763080e-01
6.44785613e-02 -1.63885146e-01 4.44510192e-01 1.06166506e+00
-6.41605705e-02 -1.46767020e-01 -6.76835299e-01 5.24170220e-01
1.09201622e+00 8.66187457e-03 5.82391262e-01 -7.73568094e-01
-4.32185382e-02 -6.41137838e-01 -3.16294581e-01 -1.72731543e+00
2.44792718e-02 -7.54790962e-01 4.14087594e-01 -1.78949809e+00
-1.29825503e-01 -7.14965522e-01 -3.41031497e-04 2.12237552e-01
-4.69069362e-01 6.45065784e-01 -1.28039017e-01 1.01295197e+00
-6.49936318e-01 9.38666523e-01 1.37555182e+00 -8.55183527e-02
-3.76022696e-01 -8.49873796e-02 -4.52920198e-01 6.36302590e-01
5.71116924e-01 -4.61762369e-01 2.52439757e-03 -7.49232292e-01
2.28664741e-01 -1.04247509e-02 2.00035587e-01 -8.03677738e-01
6.45923197e-01 -1.42721891e-01 2.33376529e-02 -6.01016879e-01
3.18640411e-01 -8.58344734e-01 2.78168380e-01 1.36724889e-01
2.01806054e-01 3.23778540e-02 -2.07446977e-01 5.86848438e-01
-5.43217897e-01 -4.54675287e-01 8.15738320e-01 3.78594510e-02
-5.25972307e-01 5.29659688e-01 -3.33516270e-01 -6.59276545e-02
9.18403089e-01 -4.88973022e-01 -1.29641518e-01 -6.32439375e-01
-6.64531231e-01 3.77495289e-01 7.15334177e-01 1.51972920e-02
8.26558411e-01 -1.72090173e+00 -7.84976125e-01 1.24069611e-02
-3.18565756e-01 2.21108556e-01 1.16534933e-01 1.36897171e+00
-8.48150313e-01 3.36274832e-01 5.90733171e-01 -7.57973433e-01
-1.01385915e+00 4.92163837e-01 4.77093726e-01 9.90202799e-02
-5.92647076e-01 9.42959428e-01 -1.43252864e-01 -6.91653192e-01
4.50600404e-03 2.99426883e-01 -9.39786658e-02 -3.55775803e-01
5.87195754e-01 1.01212132e+00 -1.43113583e-01 -1.08935428e+00
-4.43066329e-01 1.39347363e+00 5.26019692e-01 -9.71396863e-02
1.26691628e+00 -6.85717404e-01 -7.51925945e-01 -7.56489336e-02
1.70064056e+00 2.74166763e-01 -1.04914272e+00 -2.13208154e-01
-2.37151355e-01 -8.31945777e-01 7.54816771e-01 -1.63801745e-01
-1.26042759e+00 3.01723540e-01 4.76281732e-01 -4.32877615e-02
1.09094143e+00 -2.56169975e-01 1.02870739e+00 3.89446169e-01
3.25122565e-01 -1.14612186e+00 1.14716522e-01 5.56941509e-01
1.14337838e+00 -1.21735668e+00 6.61570549e-01 -5.91286004e-01
-3.66675943e-01 1.03681791e+00 2.39788182e-02 -3.37019444e-01
1.02655256e+00 -1.26787769e-02 1.26716837e-01 -2.80363709e-01
9.86549705e-02 2.46798813e-01 6.90649569e-01 4.79097933e-01
3.96600412e-03 -2.23366275e-01 -3.55170101e-01 -3.47556770e-02
-1.06427781e-01 -3.35282204e-03 3.05883169e-01 6.24450684e-01
-4.90521729e-01 -6.92081273e-01 -9.72685993e-01 2.10322767e-01
-2.01709926e-01 -3.59565943e-01 1.79969132e-01 2.44192556e-01
-4.68489468e-01 1.20311821e+00 -2.26281196e-01 -8.29410553e-02
1.06224008e-01 -6.26066327e-01 2.19788581e-01 -2.76341021e-01
-1.00654706e-01 4.06612039e-01 -3.35442483e-01 -8.69108558e-01
-5.77294350e-01 -6.97848976e-01 -1.11077642e+00 -4.75669295e-01
-8.49098325e-01 2.77675390e-01 8.51709008e-01 9.51697767e-01
6.19551480e-01 -4.93455194e-02 8.98419738e-01 -1.10277331e+00
-4.69838589e-01 -4.33440089e-01 -7.19954312e-01 1.42121881e-01
4.20923382e-01 -5.48329115e-01 -7.91985273e-01 -8.52940157e-02] | [11.659388542175293, -2.6389853954315186] |
0140a08e-cba8-43b8-9f58-280dee0ffee9 | synthetic-propaganda-embeddings-to-train-a | null | null | https://aclanthology.org/D19-5023 | https://aclanthology.org/D19-5023.pdf | Synthetic Propaganda Embeddings To Train A Linear Projection | This paper presents a method of detecting fine-grained categories of propaganda in text. Given a sentence, our method aims to identify a span of words and predict the type of propaganda used. To detect propaganda, we explore a method for extracting features of propaganda from contextualized embeddings without fine-tuning the large parameters of the base model. We show that by generating synthetic embeddings we can train a linear function with ReLU activation to extract useful labeled embeddings from an embedding space generated by a general-purpose language model. We also introduce an inference technique to detect continuous spans in sequences of propaganda tokens in sentences. A result of the ensemble model is submitted to the first shared task in fine-grained propaganda detection at NLP4IF as Team Stalin. In this paper, we provide additional analysis regarding our method of detecting spans of propaganda with synthetically generated representations. | ['Mehdi Ghanimifard', 'Adam Ek'] | 2019-11-01 | null | null | null | ws-2019-11 | ['propaganda-detection'] | ['natural-language-processing'] | [ 3.42218548e-01 1.34496585e-01 2.00712569e-02 -3.07372004e-01
-9.80337262e-01 -6.78332031e-01 1.24437392e+00 4.51461196e-01
-5.09114206e-01 4.92360741e-01 1.17980254e+00 -4.79317248e-01
8.94569457e-02 -9.02458727e-01 -5.19947350e-01 -4.44475740e-01
-2.32604459e-01 1.74816549e-01 -2.97190994e-01 -4.50062692e-01
9.49520051e-01 4.69283044e-01 -8.65981102e-01 7.88415253e-01
7.47394025e-01 1.83371753e-01 -9.17797387e-02 1.08855414e+00
-4.39012319e-01 1.17181313e+00 -1.30771160e+00 -8.98612887e-02
-4.25465479e-02 -7.34533072e-01 -9.81433094e-01 -1.75354943e-01
5.63644886e-01 -1.26030713e-01 -3.25079829e-01 8.51192057e-01
6.96741521e-01 1.28748110e-02 1.19835901e+00 -5.94821513e-01
-1.07641006e+00 1.27831447e+00 -2.93288946e-01 7.58210659e-01
6.45341814e-01 -7.69509375e-02 8.90829206e-01 -1.10243297e+00
1.03626883e+00 1.64941859e+00 7.29135811e-01 6.08914137e-01
-1.59543836e+00 -3.03655982e-01 -1.58177644e-01 4.07250486e-02
-1.04930162e+00 -3.78175527e-01 8.41264606e-01 -8.93327236e-01
1.30268919e+00 2.00872049e-01 4.78459567e-01 1.84648728e+00
5.13356566e-01 5.05853474e-01 1.21631479e+00 -8.19473684e-01
2.81353116e-01 1.50961727e-01 5.30897677e-01 9.91930902e-01
8.87299404e-02 2.89511234e-02 -5.60590684e-01 -8.08049917e-01
3.91931683e-01 -3.44160974e-01 -2.22046785e-02 3.84450972e-01
-1.18501031e+00 1.61092281e+00 3.39150101e-01 7.04658210e-01
-3.26043576e-01 5.07622540e-01 8.54476273e-01 3.03428501e-01
1.22287941e+00 1.01761627e+00 -7.14397877e-02 -1.45073682e-01
-8.95685256e-01 4.93236870e-01 7.46454298e-01 2.50506729e-01
4.32372570e-01 1.67079821e-01 -5.79007328e-01 7.46357977e-01
-1.53170317e-01 3.29763681e-01 5.99313319e-01 -1.43423989e-01
6.13337696e-01 3.20425034e-01 8.07484463e-02 -1.29690421e+00
-5.05626857e-01 -1.15654953e-02 -3.68758619e-01 1.46914288e-01
2.91020721e-01 -4.21039045e-01 -8.28842223e-01 1.49011564e+00
7.85625726e-02 -2.97118843e-01 -3.13170344e-01 5.77308416e-01
4.31065947e-01 1.02183557e+00 1.09322816e-01 2.83875335e-02
1.36794484e+00 -7.35487282e-01 -5.69482565e-01 -2.07814038e-01
1.22161126e+00 -6.98988199e-01 1.00812781e+00 3.96738667e-03
-5.94988465e-01 -3.83057982e-01 -8.55183780e-01 -1.88182257e-02
-5.08802414e-01 -1.74901739e-01 3.42019200e-01 5.24137557e-01
-6.20255291e-01 7.28732646e-01 -3.42710704e-01 -4.29350704e-01
4.07059848e-01 -5.34317076e-01 -3.03587765e-01 3.56384516e-01
-1.29212642e+00 1.31598401e+00 4.81797963e-01 -1.84714720e-01
-9.76248562e-01 -7.42803752e-01 -1.11567700e+00 -1.71010435e-01
-2.66117990e-01 -2.96641201e-01 7.95241058e-01 -8.60168040e-01
-6.71070576e-01 1.02069938e+00 -7.44299870e-03 -5.08211195e-01
2.33954862e-01 -3.51273984e-01 -5.05573928e-01 1.42867997e-01
5.30926883e-01 1.75332472e-01 1.29256880e+00 -7.08141029e-01
-8.51508137e-03 -1.47347376e-01 -1.31841496e-01 -4.80437696e-01
-4.88009304e-01 8.45835149e-01 9.97011125e-01 -1.04955065e+00
-5.93523026e-01 -3.59539568e-01 -2.31604517e-01 -4.51953769e-01
-7.56177187e-01 -6.36146665e-01 5.42450190e-01 -9.85836446e-01
1.54165518e+00 -1.91079521e+00 1.77619815e-01 -2.47610477e-03
3.33118975e-01 1.26133218e-01 -6.15067005e-01 1.02863276e+00
-2.67006129e-01 8.21510792e-01 -8.00361857e-02 1.47536784e-01
1.20759867e-01 -1.72510788e-01 -8.94189298e-01 6.99178576e-01
6.24763012e-01 7.52787113e-01 -1.22116005e+00 -4.72165525e-01
-9.27644372e-02 3.35483223e-01 -5.22049963e-01 1.50138855e-01
-2.51520813e-01 -1.34354651e-01 -5.49024284e-01 1.17460221e-01
3.93720716e-01 3.82153504e-02 5.96569963e-02 1.53445512e-01
-3.42229277e-01 8.78570795e-01 -5.57651520e-01 1.40023446e+00
-5.97658396e-01 1.05613160e+00 -2.81517178e-01 -8.13879430e-01
8.43170643e-01 5.53674549e-02 -2.06859574e-01 -3.47229749e-01
2.41549119e-01 -5.94831705e-02 -1.22600660e-01 -7.24829912e-01
6.47765219e-01 -4.02719796e-01 -7.89279640e-01 1.25663173e+00
5.01987413e-02 -1.03368228e-02 2.62337118e-01 5.66224515e-01
1.88591313e+00 -2.59537131e-01 6.15781724e-01 -4.34300035e-01
1.90401509e-01 4.95962620e-01 1.44236058e-03 1.12995470e+00
-9.05853361e-02 1.73329040e-01 6.13870859e-01 -1.02064359e+00
-1.34533274e+00 -1.25498867e+00 -2.85291433e-01 1.51239145e+00
-6.44134581e-01 -8.77140164e-01 -4.95869577e-01 -1.19889462e+00
-1.20057296e-02 1.08987188e+00 -1.26304901e+00 1.82789620e-02
-9.63979483e-01 -6.55859530e-01 1.01678324e+00 3.81836027e-01
-2.42741510e-01 -1.50831223e+00 -8.61810923e-01 4.92876083e-01
-5.51160090e-02 -4.34591711e-01 -3.86026442e-01 3.62577826e-01
-2.90598750e-01 -1.05914211e+00 -3.12306434e-01 -8.19283843e-01
5.79701602e-01 -3.31468917e-02 9.73968685e-01 -1.23661980e-01
-9.00430143e-01 -1.01770936e-02 -6.64290488e-01 -2.20920920e-01
-1.07735729e+00 -7.31584663e-03 1.25617608e-01 -5.99816561e-01
4.06274021e-01 -1.99438408e-01 -5.07741980e-03 -5.52632391e-01
-7.80670345e-01 -1.08308785e-01 8.73603001e-02 1.19265413e+00
-3.39739561e-01 -3.02332491e-01 7.45199978e-01 -1.02920890e+00
1.56201792e+00 -6.62692606e-01 4.27259840e-02 7.37437885e-03
-1.25991508e-01 4.43481416e-01 9.06602740e-01 -4.11729753e-01
-8.98502648e-01 -4.89629596e-01 -1.98006347e-01 2.33507484e-01
-1.79870814e-01 4.51326817e-01 6.87135637e-01 4.70827520e-01
1.55581009e+00 4.25160080e-02 -2.14828447e-01 -5.31288505e-01
9.81536925e-01 8.83896828e-01 7.62117282e-02 -4.34025139e-01
9.05189157e-01 2.42799699e-01 -6.11876070e-01 -9.42459941e-01
-1.06721807e+00 -3.82550180e-01 -5.91236532e-01 8.52097049e-02
8.33556950e-01 -4.59973931e-01 5.74600250e-02 -3.86303179e-02
-1.80835497e+00 -1.28528759e-01 -3.34887207e-01 2.17940614e-01
-5.12010098e-01 3.73150975e-01 -8.48798275e-01 -9.13926601e-01
-6.35200799e-01 -2.76815057e-01 9.82466161e-01 -3.00063431e-01
-9.38959181e-01 -1.25672472e+00 7.73132682e-01 -5.91210648e-02
3.32264841e-01 5.80367684e-01 1.28214729e+00 -1.03514600e+00
3.99155289e-01 -2.06894547e-01 -2.15296581e-01 2.45415494e-01
2.92529941e-01 4.17246446e-02 -9.25686479e-01 2.66947821e-02
-9.22685489e-02 -6.74692810e-01 1.17343593e+00 1.10717520e-01
8.91920745e-01 -1.10908401e+00 -2.20941573e-01 1.18285351e-01
1.18537474e+00 -3.67905080e-01 3.92507046e-01 6.46605372e-01
3.99362475e-01 7.64385581e-01 4.65674371e-01 6.71128809e-01
-5.26998997e-01 2.03233704e-01 1.42557919e-01 3.44938129e-01
9.70984530e-03 -5.04533947e-01 8.91400099e-01 6.07432604e-01
2.43254334e-01 -2.18106911e-01 -9.03474152e-01 9.26382840e-01
-1.39692843e+00 -1.91512179e+00 -1.60720706e-01 1.42533576e+00
1.01617217e+00 6.18497171e-02 8.78453329e-02 6.33042082e-02
8.57290626e-01 9.33990002e-01 1.39898375e-01 -1.55788577e+00
9.24762245e-03 7.07743943e-01 1.02374732e-01 7.57873714e-01
-1.24865913e+00 1.07471895e+00 7.19464350e+00 1.03926063e+00
-7.62201607e-01 3.43108624e-01 1.92983851e-01 -2.32507110e-01
-6.97149038e-01 -2.02558234e-01 -6.24062002e-01 4.79744792e-01
1.27908969e+00 -4.82359707e-01 2.74566531e-01 8.44230175e-01
4.71485585e-01 3.43158662e-01 -1.02959037e+00 3.01650077e-01
4.75306273e-01 -1.99868989e+00 2.15532318e-01 -2.55996007e-02
6.64273500e-01 1.63944080e-01 -2.72475004e-01 4.21501994e-01
6.02620423e-01 -1.42752671e+00 7.24429965e-01 1.67817637e-01
5.51486135e-01 -6.57187343e-01 4.68707919e-01 5.72677970e-01
-5.54620147e-01 -1.15183361e-01 -6.81545019e-01 -4.85231340e-01
4.71292526e-01 7.76441514e-01 -1.44370639e+00 -4.67763003e-03
-2.15908699e-03 4.98553693e-01 -6.50278807e-01 2.93831527e-01
-7.26193249e-01 1.05884504e+00 5.10380790e-02 -6.13669634e-01
5.93672633e-01 2.43223459e-01 7.03533471e-01 2.09989190e+00
2.04629585e-01 -3.96179199e-01 -3.71371512e-03 1.15529370e+00
-5.14230765e-02 1.56696990e-01 -1.14154172e+00 -7.11767375e-01
5.50151765e-01 1.30577350e+00 -2.37321630e-01 -3.59732091e-01
3.54194105e-01 9.40741777e-01 7.96418667e-01 -1.14574777e-02
-8.27027559e-01 -7.26407230e-01 6.08029783e-01 1.53460801e-01
5.45081459e-02 -3.03027749e-01 -1.00818865e-01 -1.06004965e+00
-2.11262703e-01 -5.43036282e-01 3.06627780e-01 -5.54451525e-01
-1.68724597e+00 6.51855111e-01 -1.35042951e-01 -4.24365878e-01
-7.47842371e-01 -6.93995059e-01 -1.33990896e+00 1.03776526e+00
-6.27437472e-01 -8.43950629e-01 2.33706906e-01 -9.42461640e-02
9.15685356e-01 -2.67427135e-02 1.23880458e+00 -4.38451767e-01
-2.38546714e-01 1.10905305e-01 -3.79348658e-02 6.58346117e-01
7.32777238e-01 -1.22783506e+00 8.05741787e-01 8.44813228e-01
4.34860915e-01 9.39743817e-01 1.12636971e+00 -7.62721121e-01
-9.36521232e-01 -1.25535512e+00 1.56434500e+00 -7.84949720e-01
1.40443468e+00 -7.92011380e-01 -6.46837831e-01 5.27246118e-01
5.02969980e-01 -3.56467187e-01 8.21117401e-01 4.57154214e-01
-1.10375905e+00 7.06941009e-01 -1.16715503e+00 5.04787028e-01
1.14968669e+00 -8.21093976e-01 -1.43829584e+00 8.55512500e-01
6.74974144e-01 3.99166137e-01 -4.11401242e-01 -3.02325726e-01
4.90562856e-01 -5.95193446e-01 8.27830434e-01 -1.40487015e+00
1.12474048e+00 2.41994023e-01 -4.57095578e-02 -1.85403717e+00
-6.85711741e-01 -3.90926719e-01 -1.24385215e-01 1.12425721e+00
2.22600088e-01 -4.86367673e-01 3.28032196e-01 -3.96466762e-01
5.73259778e-02 -4.56202298e-01 -1.09428549e+00 -9.53235924e-01
5.79210162e-01 -3.57332438e-01 -7.36159980e-02 9.90655541e-01
4.73380238e-01 6.85896814e-01 -4.11169320e-01 -1.42888606e-01
6.40562236e-01 7.68910199e-02 5.73653519e-01 -7.94689894e-01
-1.79721773e-01 -3.21065158e-01 -4.61702883e-01 -5.97680509e-01
5.98352253e-01 -1.24194372e+00 1.98841110e-01 -1.76006508e+00
3.89408857e-01 -1.06497109e-01 -8.95256028e-02 2.59900600e-01
5.20972908e-02 1.46336764e-01 1.63067132e-01 -1.20584816e-01
-2.19063014e-01 3.85846436e-01 5.03926635e-01 -5.40988624e-01
1.20987147e-01 -7.33072340e-01 -7.74753273e-01 8.99247825e-01
1.04542208e+00 -1.06370819e+00 5.89403808e-02 -6.30534589e-01
4.39395666e-01 -4.16666269e-01 7.32231259e-01 -4.36704278e-01
-2.36706346e-01 -4.11877960e-01 4.62391794e-01 -6.68632269e-01
7.76642486e-02 2.11690992e-01 -5.27999103e-01 7.81695127e-01
-7.63592958e-01 1.05340175e-01 6.97604474e-03 5.66600561e-01
1.77661076e-01 -7.35185981e-01 5.13837814e-01 -1.63954571e-01
-4.96108770e-01 -4.53524262e-01 -1.24829197e+00 4.81373370e-01
7.76188314e-01 2.41158739e-01 -1.00496876e+00 1.35533782e-02
-6.56738460e-01 -2.84980953e-01 3.20867896e-01 5.20349801e-01
7.42615700e-01 -1.38960254e+00 -1.28061450e+00 2.15824936e-02
1.36523113e-01 -9.83206570e-01 -4.32272553e-02 4.36490655e-01
-6.96732342e-01 2.42181435e-01 -3.80556397e-02 2.87573691e-02
-1.25882006e+00 6.96912348e-01 -2.15280011e-01 -4.63739455e-01
-8.46663058e-01 9.03349817e-01 -7.49685168e-02 -3.98189306e-01
-3.43690276e-01 2.78389528e-02 -9.31681022e-02 3.31148237e-01
1.17997670e+00 4.50786710e-01 -4.26386029e-01 -4.65371162e-01
-6.14137888e-01 4.35910821e-02 -2.17323929e-01 -2.56488323e-01
1.52639782e+00 3.41940433e-01 -1.91315830e-01 5.26342630e-01
1.50634861e+00 6.97232068e-01 -3.69955063e-01 2.06695318e-01
2.33135924e-01 -4.79046404e-01 -2.48051509e-01 -8.08837950e-01
1.44490823e-01 9.41266835e-01 1.00826547e-01 6.27977490e-01
2.67022163e-01 3.75239342e-01 5.45651793e-01 5.60562134e-01
2.40008533e-01 -1.23194587e+00 3.38456184e-01 8.78500104e-01
1.35759521e+00 -7.40619540e-01 6.95901690e-03 2.52778102e-02
-2.95188695e-01 1.41742194e+00 -3.46939191e-02 -8.41818988e-01
3.54558378e-01 2.79353827e-01 -2.66012680e-02 -6.71622574e-01
-9.53241646e-01 9.05649140e-02 6.52421117e-02 7.20304191e-01
7.28148758e-01 1.51161820e-01 -8.32879007e-01 2.77079374e-01
-3.18905830e-01 -5.46298504e-01 6.98160410e-01 9.00689960e-01
-1.09013307e+00 -8.03983092e-01 -5.89985490e-01 5.09358585e-01
-4.63618129e-01 -6.32846415e-01 -1.03506410e+00 5.37921369e-01
6.14041686e-02 1.12005949e+00 1.38579577e-01 -4.84426498e-01
-2.78450195e-02 3.27154011e-01 5.95845520e-01 -1.13611102e+00
-1.04124904e+00 -3.97942632e-01 8.92544210e-01 -4.37423021e-01
-7.08753914e-02 -3.49193245e-01 -1.04450524e+00 -5.39470017e-01
-1.57913521e-01 4.30267721e-01 6.45610213e-01 7.54447579e-01
1.79384753e-01 3.50314111e-01 7.97468662e-01 -9.09836769e-01
-1.23042428e+00 -1.46232378e+00 -4.08081561e-01 5.55596113e-01
1.49584740e-01 -2.83621132e-01 -7.13326812e-01 -1.80928737e-01] | [8.43415355682373, 10.676199913024902] |
1492d654-abc0-4faa-810c-a18129fef4cb | integrating-incremental-speech-recognition | null | null | https://aclanthology.org/W12-1638 | https://aclanthology.org/W12-1638.pdf | Integrating Incremental Speech Recognition and POMDP-Based Dialogue Systems | null | ['Peter A. Heeman', 'Ethan O. Selfridge', 'Jason D. Williams', 'Iker Arizmendi'] | 2012-07-01 | null | null | null | ws-2012-7 | ['dialogue-management'] | ['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] | [-7.275010585784912, 3.617556571960449] |
234fe5aa-0265-4852-a64a-f1001f26751e | separability-and-scatteredness-s-s-ratio | 2305.10219 | null | https://arxiv.org/abs/2305.10219v1 | https://arxiv.org/pdf/2305.10219v1.pdf | Separability and Scatteredness (S&S) Ratio-Based Efficient SVM Regularization Parameter, Kernel, and Kernel Parameter Selection | Support Vector Machine (SVM) is a robust machine learning algorithm with broad applications in classification, regression, and outlier detection. SVM requires tuning the regularization parameter (RP) which controls the model capacity and the generalization performance. Conventionally, the optimum RP is found by comparison of a range of values through the Cross-Validation (CV) procedure. In addition, for non-linearly separable data, the SVM uses kernels where a set of kernels, each with a set of parameters, denoted as a grid of kernels, are considered. The optimal choice of RP and the grid of kernels is through the grid-search of CV. By stochastically analyzing the behavior of the regularization parameter, this work shows that the SVM performance can be modeled as a function of separability and scatteredness (S&S) of the data. Separability is a measure of the distance between classes, and scatteredness is the ratio of the spread of data points. In particular, for the hinge loss cost function, an S&S ratio-based table provides the optimum RP. The S&S ratio is a powerful value that can automatically detect linear or non-linear separability before using the SVM algorithm. The provided S&S ratio-based table can also provide the optimum kernel and its parameters before using the SVM algorithm. Consequently, the computational complexity of the CV grid-search is reduced to only one time use of the SVM. The simulation results on the real dataset confirm the superiority and efficiency of the proposed approach in the sense of computational complexity over the grid-search CV method. | ['Soosan Beheshti', 'Mahdi Shamsi'] | 2023-05-17 | null | null | null | null | ['outlier-detection'] | ['methodology'] | [-3.42204601e-01 -4.05461520e-01 -4.23247665e-01 -2.33134940e-01
-3.09126973e-01 -3.63700926e-01 1.95777357e-01 4.28143591e-01
-3.91334862e-01 6.40111685e-01 -4.54363734e-01 -3.98820639e-01
-4.24333960e-01 -6.62648141e-01 -2.17070878e-01 -1.02106953e+00
1.73176881e-02 1.73528433e-01 3.63369197e-01 -1.53196201e-01
6.20947897e-01 7.07772732e-01 -1.48873198e+00 -1.08099870e-01
1.42754054e+00 1.36069393e+00 -2.73170555e-03 1.63839132e-01
-7.34590590e-02 1.47039607e-01 -6.56619549e-01 1.02636918e-01
3.94447654e-01 -3.23222041e-01 -2.21049294e-01 2.26144996e-02
-9.95478854e-02 4.64319512e-02 -1.32318623e-02 1.12424910e+00
1.33514762e-01 4.30926204e-01 8.85068536e-01 -1.40239942e+00
-2.76127070e-01 -5.90631291e-02 -5.08755624e-01 4.79103565e-01
3.42348665e-02 -1.67913064e-02 6.47936106e-01 -8.93953979e-01
8.94776881e-02 6.35415554e-01 6.05383754e-01 -1.04705326e-01
-1.06502199e+00 -6.09003246e-01 -1.38797462e-01 2.16209173e-01
-1.67517638e+00 1.48663253e-01 8.41874361e-01 -7.53026783e-01
4.96449769e-01 4.96890336e-01 5.96885204e-01 2.68930137e-01
3.14137906e-01 2.54817694e-01 1.07331228e+00 -5.63541472e-01
6.48640931e-01 7.81574667e-01 5.08469343e-01 7.35690653e-01
5.54140508e-01 -1.03879303e-01 -1.48178682e-01 -5.82253754e-01
7.95182228e-01 4.38582674e-02 -3.42771828e-01 -6.20248139e-01
-6.84777081e-01 1.13915503e+00 3.42034221e-01 2.01004058e-01
8.58362839e-02 -6.17526293e-01 6.24892056e-01 1.63596690e-01
2.17104882e-01 5.13719022e-01 -3.20279777e-01 1.00672603e-01
-1.00238562e+00 -1.19319027e-02 6.56801879e-01 4.92727607e-01
6.02523208e-01 2.24785045e-01 5.28232940e-02 9.38144267e-01
5.50099090e-02 4.21372741e-01 7.63620973e-01 -1.46009922e-01
6.80723548e-01 1.00668800e+00 2.26116821e-01 -1.27804148e+00
-4.24453139e-01 -5.50808847e-01 -7.85113156e-01 3.89859557e-01
7.25611746e-01 -4.10698429e-02 -4.85488027e-01 1.11794269e+00
5.34979045e-01 8.78021196e-02 1.32668540e-01 1.08584607e+00
4.90740210e-01 5.20597219e-01 -2.47538194e-01 -4.69012231e-01
1.11014128e+00 -7.26367831e-01 -4.29194152e-01 -3.46800778e-04
6.77564740e-01 -6.20949566e-01 1.28285086e+00 3.74546170e-01
-3.98095191e-01 -4.40973938e-01 -1.37015510e+00 6.58692062e-01
-3.97352993e-01 7.38307238e-01 3.46403629e-01 5.59435964e-01
-2.43412465e-01 5.80407560e-01 -7.70742476e-01 -6.30381182e-02
7.84841031e-02 3.25677067e-01 -4.38265860e-01 3.23814064e-01
-1.04239011e+00 1.02450967e+00 6.35734141e-01 2.94827580e-01
1.66056573e-01 -4.08470899e-01 -7.21783638e-01 9.10304636e-02
1.52103147e-02 1.11591443e-02 3.71759266e-01 -1.00957799e+00
-1.25735724e+00 4.53620404e-01 1.04915373e-01 -3.74420643e-01
4.39294100e-01 1.49310023e-01 -4.89831269e-01 1.66143566e-01
-1.72558263e-01 -4.12868410e-01 9.31981027e-01 -8.70371163e-01
-5.10052085e-01 -3.76847446e-01 -4.37980294e-01 1.94108769e-01
-4.40188557e-01 8.87434259e-02 7.58414492e-02 -4.62228596e-01
4.10076022e-01 -8.62401426e-01 -1.33076280e-01 -1.47382408e-01
-4.59970891e-01 -1.56284526e-01 8.20609689e-01 -7.52383709e-01
1.53724456e+00 -2.51428413e+00 -1.79342598e-01 8.75387967e-01
-2.36117721e-01 3.55685532e-01 4.93485659e-01 2.78180778e-01
-2.27753103e-01 -2.00495169e-01 -3.86645347e-01 4.78094995e-01
-3.41764778e-01 -1.28372654e-01 -2.85678536e-01 8.63036990e-01
1.80024594e-01 1.18702315e-01 -5.90791702e-01 -4.00451183e-01
3.31887752e-01 1.77388683e-01 -1.41339123e-01 6.30994290e-02
3.17357272e-01 9.54128280e-02 -6.39061928e-01 4.42117959e-01
7.80042768e-01 -3.03114682e-01 -7.87772611e-02 -2.29435399e-01
-1.32659808e-01 -3.56501520e-01 -1.78086233e+00 6.50797606e-01
-2.27585182e-01 5.74152827e-01 -2.18771502e-01 -1.21070969e+00
1.64432764e+00 -2.44652107e-02 2.98730373e-01 -2.43912846e-01
2.18921751e-01 5.67931533e-01 4.54327576e-02 -4.48518753e-01
1.37063876e-01 1.88604757e-01 7.78301731e-02 7.62745813e-02
-3.34841758e-01 1.18872598e-01 7.17979670e-02 -3.84064853e-01
6.09905660e-01 -3.72679681e-01 8.21600676e-01 -4.89766181e-01
9.85755980e-01 2.07384840e-01 4.56536204e-01 1.59719989e-01
-1.32002354e-01 3.49462986e-01 7.31799126e-01 -4.24861103e-01
-9.24260259e-01 -8.39889169e-01 -8.66513073e-01 4.44521397e-01
4.40141231e-01 1.30381748e-01 -6.18813396e-01 -6.07864141e-01
4.09047484e-01 7.66017914e-01 -4.63918269e-01 -4.38881278e-01
-5.20608604e-01 -8.83909881e-01 1.97802991e-01 3.12516868e-01
3.94279420e-01 -6.49852097e-01 -5.96598089e-01 -2.18431607e-01
3.28844845e-01 -7.50038385e-01 -3.50618392e-01 3.40278834e-01
-1.06193984e+00 -1.33834434e+00 -4.28561419e-01 -6.91001117e-01
1.09116971e+00 2.19092444e-02 3.07503223e-01 1.67887732e-01
-2.05802456e-01 -2.27596030e-01 -3.42201948e-01 -7.04692230e-02
-3.04923266e-01 -3.67714316e-02 1.68698192e-01 1.74059585e-01
6.02513030e-02 -4.13529664e-01 -5.70700824e-01 6.86216891e-01
-6.49936557e-01 -4.45943922e-01 2.70417899e-01 1.16561973e+00
6.21068716e-01 4.96637315e-01 6.24069750e-01 -5.06636858e-01
8.08271766e-01 -5.32583535e-01 -1.03956652e+00 3.65855753e-01
-8.36809456e-01 1.23491473e-01 1.10054064e+00 -6.05262995e-01
-6.66601241e-01 -3.66318375e-02 4.63307232e-01 -4.72573668e-01
1.04496919e-01 5.13435602e-01 -5.61599322e-02 -4.56943631e-01
8.52987587e-01 3.50484610e-01 2.84927964e-01 -2.46064261e-01
-7.42000639e-02 8.72889102e-01 2.18599588e-01 -3.32323909e-01
6.74661458e-01 9.00292695e-02 2.85606474e-01 -9.21584249e-01
-3.66802722e-01 -5.73526978e-01 -5.84650695e-01 -1.92972884e-01
3.10689032e-01 -4.24553484e-01 -7.47824252e-01 3.61306041e-01
-5.95564544e-01 1.90615639e-01 -9.83647406e-02 6.66846216e-01
-3.66717905e-01 4.04944241e-01 -1.89219564e-01 -1.09285724e+00
-3.78826946e-01 -1.25125515e+00 5.15049577e-01 4.53331798e-01
-1.50330231e-01 -9.98921633e-01 -3.15455139e-01 2.35814929e-01
1.01792157e-01 3.96857083e-01 1.10730922e+00 -1.15117180e+00
-1.13642707e-01 -7.91232288e-01 -7.62087554e-02 6.16680503e-01
1.22315086e-01 2.42304981e-01 -4.44531143e-01 -4.90404695e-01
1.54986173e-01 8.86397436e-02 4.62566048e-01 3.72790635e-01
1.11155510e+00 -4.59667176e-01 -1.94383472e-01 7.88751185e-01
1.62893009e+00 6.02884531e-01 4.24800783e-01 7.09230840e-01
3.84826899e-01 1.93790540e-01 9.69092607e-01 6.56696796e-01
-1.55775607e-01 7.43461847e-01 1.79152682e-01 -1.46167949e-01
5.40838659e-01 3.26289833e-02 1.02220751e-01 4.50586885e-01
7.55574778e-02 2.87325919e-01 -9.31654871e-01 4.95937839e-02
-1.75482428e+00 -7.50552177e-01 -1.02946848e-01 2.87088513e+00
5.91951430e-01 2.19588891e-01 3.00348490e-01 6.44158125e-01
9.11722898e-01 -2.22421661e-01 -6.92056358e-01 -6.25958920e-01
1.40484087e-02 -1.07471794e-01 6.55053854e-01 3.51384729e-01
-9.60899591e-01 4.33900386e-01 5.43546152e+00 1.18606949e+00
-1.60977268e+00 -5.13814569e-01 5.90435684e-01 1.16468608e-01
2.02500001e-01 -1.46971848e-02 -8.95293117e-01 9.57236707e-01
3.48802596e-01 -2.54752964e-01 3.55561078e-01 1.25733018e+00
2.28628680e-01 -5.19180715e-01 -6.47040606e-01 9.49687362e-01
-1.54486187e-02 -1.06278539e+00 -2.60638058e-01 -1.34437323e-01
5.39671421e-01 -5.24454713e-01 6.67101368e-02 -6.05371632e-02
-3.79019767e-01 -7.96503723e-01 4.71132517e-01 4.87600207e-01
5.34302771e-01 -9.36597705e-01 1.08374155e+00 5.09471953e-01
-1.07205057e+00 -5.47826588e-01 -3.40181172e-01 2.10939348e-01
-3.03037077e-01 8.01311135e-01 -8.91963363e-01 3.30926627e-01
4.86016512e-01 1.74274281e-01 -4.69086856e-01 1.28572345e+00
1.63779095e-01 4.42934126e-01 -4.88286555e-01 -3.02630067e-01
-5.36604002e-02 -8.99147272e-01 6.22882783e-01 9.86438513e-01
3.65205914e-01 5.36438562e-02 2.87935019e-01 6.59060538e-01
6.64537907e-01 6.87745869e-01 -2.69232959e-01 1.88936755e-01
7.94837654e-01 9.61223543e-01 -8.29236627e-01 -1.95545889e-02
-1.65406123e-01 5.80368102e-01 3.27176124e-01 3.24789762e-01
-7.32550800e-01 -6.84235990e-01 3.97702724e-01 2.62098670e-01
3.57654840e-01 -1.65753201e-01 -6.88092113e-01 -8.30351293e-01
4.86730456e-01 -6.37396693e-01 5.76967061e-01 -4.32991982e-01
-1.17500949e+00 6.82455063e-01 1.08776949e-01 -1.70888340e+00
-2.73401663e-03 -5.60916305e-01 -7.26476908e-01 1.01661479e+00
-1.05929399e+00 -5.98279655e-01 -3.42295289e-01 4.72148567e-01
2.58215487e-01 -5.71132183e-01 6.17875159e-01 -1.01672381e-01
-9.98764753e-01 8.25423658e-01 5.75947940e-01 4.69662137e-02
3.67213249e-01 -9.35116112e-01 -3.82289261e-01 6.58586860e-01
-3.28540504e-01 6.35025859e-01 1.00757778e+00 -5.46485603e-01
-8.83065701e-01 -6.08716488e-01 4.44147229e-01 1.25722602e-01
6.89642906e-01 -1.30684313e-03 -1.08494568e+00 2.32500010e-04
-6.35154188e-01 3.99636269e-01 6.34790659e-01 -5.56517094e-02
-6.26200810e-02 -4.85480756e-01 -1.52878404e+00 3.22633207e-01
8.71158615e-02 -2.01422349e-01 -3.12167197e-01 2.96125323e-01
3.26051623e-01 -4.89356101e-01 -1.06382132e+00 5.31126738e-01
5.12352765e-01 -1.07803905e+00 9.12714541e-01 -4.18720961e-01
3.00222486e-02 -5.27046978e-01 4.58374098e-02 -1.26805377e+00
-1.87391981e-01 5.45875244e-02 -1.83745772e-02 9.26176786e-01
4.81804699e-01 -1.09568286e+00 7.39017725e-01 5.74291706e-01
2.66469300e-01 -1.45995855e+00 -1.02402484e+00 -1.08282030e+00
-2.67189920e-01 1.17148951e-01 4.24472004e-01 1.00725341e+00
2.37196416e-01 -2.34801561e-01 1.16236314e-01 2.89296389e-01
4.59797829e-01 2.22816005e-01 5.39430082e-01 -1.32965958e+00
-1.74659386e-01 -3.42695683e-01 -8.32579613e-01 -3.72216314e-01
-7.30593875e-03 -7.64373362e-01 -2.96066910e-01 -1.11814010e+00
-2.07017824e-01 -9.02858913e-01 -2.97454357e-01 1.96900830e-01
-1.92344636e-01 -4.53134984e-01 -1.28276974e-01 5.16013682e-01
1.55837014e-01 3.61852586e-01 1.03657389e+00 2.45456338e-01
-8.83069098e-01 5.78965187e-01 -2.48703882e-01 8.18358779e-01
9.19146538e-01 -3.06205601e-01 -4.92902458e-01 3.89523417e-01
-3.73942479e-02 2.66415387e-01 1.86022878e-01 -9.87262666e-01
1.34783760e-01 -5.32437742e-01 5.70899069e-01 -3.56829941e-01
1.28294528e-01 -9.32720840e-01 1.27517343e-01 4.73050088e-01
-1.38072908e-01 1.92170776e-02 1.14460126e-01 5.63992262e-01
-2.21545875e-01 -4.88112330e-01 9.87688661e-01 4.26879644e-01
-3.64274681e-01 -4.60937843e-02 -1.94384530e-03 -1.63358912e-01
1.47815371e+00 -8.43614876e-01 -1.02820866e-01 1.04841888e-01
-5.73094189e-01 2.61890218e-02 4.04754132e-01 1.71491846e-01
5.79339385e-01 -1.29643452e+00 -2.30003655e-01 5.49106121e-01
2.39527211e-01 -3.11352462e-01 4.73475307e-02 9.47897911e-01
-6.80786729e-01 3.07230502e-01 -1.46196038e-01 -6.39351130e-01
-1.37912071e+00 4.24525142e-01 5.61436951e-01 -2.10807775e-03
-4.84490037e-01 4.36311096e-01 -3.85376066e-01 -7.77998641e-02
2.03103185e-01 -4.07341659e-01 -4.63825494e-01 6.70367926e-02
2.94299692e-01 7.72374451e-01 1.93633676e-01 -4.57987934e-01
-4.34990644e-01 5.45986354e-01 3.15100729e-01 2.04915985e-01
9.81986821e-01 2.44798586e-01 -1.66699886e-01 4.46598530e-01
1.17918777e+00 9.53156054e-02 -9.94960129e-01 1.71067752e-02
-1.12071978e-02 -6.85185075e-01 -3.57881784e-02 -5.07110596e-01
-8.56172919e-01 3.29901785e-01 5.84541321e-01 2.94738829e-01
1.10567296e+00 -3.79964828e-01 4.73860383e-01 1.98528320e-01
2.53121674e-01 -1.30439377e+00 -2.58784235e-01 2.23676249e-01
7.80412495e-01 -1.19088578e+00 1.71484187e-01 -6.75886393e-01
-8.38403165e-01 1.46824384e+00 7.10556507e-01 -3.71601731e-01
8.79672229e-01 2.24653855e-01 -6.91128075e-02 1.19494162e-01
-1.37538284e-01 2.55963564e-01 5.89762509e-01 1.97926804e-01
5.45314588e-02 2.26978555e-01 -7.72340834e-01 8.92071903e-01
-3.59953552e-01 -1.70593902e-01 1.14881799e-01 4.83670175e-01
-8.09434772e-01 -7.40413070e-01 -6.66181087e-01 6.60583794e-01
1.72030315e-01 1.39061213e-01 1.37883246e-01 8.32521379e-01
9.93762612e-02 8.54043126e-01 -2.95704659e-02 -3.81555468e-01
5.05939245e-01 1.77240837e-02 1.84067577e-01 -1.67913109e-01
-3.34986180e-01 -4.58985001e-01 -1.43079475e-01 -2.10110411e-01
3.49097610e-01 -5.15729785e-01 -1.51869178e+00 -2.31181547e-01
-9.38512027e-01 5.29539049e-01 8.90658736e-01 1.07268822e+00
4.88521159e-02 -8.84053782e-02 1.12267077e+00 -1.38915718e-01
-1.10553753e+00 -5.56694269e-01 -9.73264635e-01 1.87319934e-01
1.81402951e-01 -9.66765285e-01 -9.84239697e-01 -4.07342792e-01] | [8.272941589355469, 4.059072494506836] |
72eb793c-fc74-4922-8470-bab1de6584dc | cross-modal-pattern-propagation-for-rgb-t | null | null | http://openaccess.thecvf.com/content_CVPR_2020/html/Wang_Cross-Modal_Pattern-Propagation_for_RGB-T_Tracking_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Wang_Cross-Modal_Pattern-Propagation_for_RGB-T_Tracking_CVPR_2020_paper.pdf | Cross-Modal Pattern-Propagation for RGB-T Tracking | Motivated by our observations on RGB-T data that pattern correlations are high-frequently recurred across modalities also along sequence frames, in this paper, we propose a cross-modal pattern-propagation (CMPP) tracking framework to diffuse instance patterns across RGB-T data on spatial domain as well as temporal domain. To bridge RGB-T modalities, the cross-modal correlations on intra-modal paired pattern-affinities are derived to reveal those latent cues between heterogenous modalities. Through the correlations, the useful patterns may be mutually propagated between RGB-T modalities so as to fulfill inter-modal pattern-propagation. Further, considering the temporal continuity of sequence frames, we adopt the spirit of pattern propagation to dynamic temporal domain, in which long-term historical contexts are adaptively correlated and propagated into the current frame for more effective information inheritance. Extensive experiments demonstrate that the effectiveness of our proposed CMPP, and the new state-of-the-art results are achieved with the significant improvements on two RGB-T object tracking benchmarks.
| [' Jian Yang', ' Xiaoya Zhang', ' Tong Zhang', ' Ling Zhou', ' Zhen Cui', ' Chunyan Xu', 'Chaoqun Wang'] | 2020-06-01 | null | null | null | cvpr-2020-6 | ['rgb-t-tracking'] | ['computer-vision'] | [ 1.79114610e-01 -5.76380193e-01 -2.99199998e-01 -2.46581137e-01
-3.06802511e-01 -6.11306846e-01 6.54886901e-01 -4.42957669e-01
1.35736624e-02 5.05002141e-01 4.34299290e-01 1.26157954e-01
-5.07829189e-01 -5.13309658e-01 -6.84930444e-01 -9.18335855e-01
-8.22233185e-02 -2.35584110e-01 8.02600741e-01 -1.46546081e-01
6.64187642e-03 2.71977842e-01 -1.50155711e+00 6.27371311e-01
3.61600518e-01 1.13497353e+00 1.30179450e-01 4.01621819e-01
-1.77110165e-01 1.16782415e+00 -5.00621535e-02 -3.46881390e-01
1.57138452e-01 -3.57007653e-01 -5.37065685e-01 5.12505889e-01
4.09561872e-01 8.90534073e-02 -7.88243771e-01 9.83585179e-01
-4.02677134e-02 2.84662008e-01 2.70302236e-01 -1.35543621e+00
-6.50729716e-01 4.31183010e-01 -1.01317465e+00 2.81110287e-01
4.15024310e-01 6.27722263e-01 6.81675315e-01 -8.21563423e-01
9.35554147e-01 1.31045270e+00 7.00286865e-01 3.89266670e-01
-8.78760040e-01 -6.12188697e-01 7.06814766e-01 4.55581546e-01
-1.02950823e+00 -1.17152721e-01 1.08487165e+00 -4.30433691e-01
3.90335441e-01 3.91155511e-01 1.05193520e+00 1.07067943e+00
2.22582176e-01 1.06692457e+00 1.23601031e+00 -7.41911381e-02
-1.83135629e-01 -2.71212906e-01 3.16455178e-02 6.63450241e-01
-3.11269045e-01 2.65079528e-01 -1.19366753e+00 -3.13376039e-02
7.44510412e-01 4.41598386e-01 -4.80259567e-01 -3.48671168e-01
-1.82362258e+00 1.05626725e-01 7.21723616e-01 5.97369373e-01
-4.55283165e-01 3.01690489e-01 2.70509720e-01 3.33262943e-02
2.24537194e-01 -2.00661600e-01 -5.33907712e-01 -2.78106064e-01
-7.28370249e-01 2.36107130e-02 1.22598909e-01 1.26471865e+00
8.57738376e-01 -2.18320832e-01 -5.21047771e-01 7.51790226e-01
6.69307172e-01 5.68603158e-01 3.49638700e-01 -9.88241315e-01
5.95396876e-01 7.60812998e-01 8.38454738e-02 -1.25815487e+00
-3.50286305e-01 -3.69261950e-01 -1.01017642e+00 -1.32910758e-01
5.31776071e-01 1.31273493e-01 -8.94686520e-01 1.87290335e+00
6.87839329e-01 6.15335524e-01 -1.75771073e-01 1.06763923e+00
6.67888522e-01 7.47243047e-01 1.41736224e-01 -4.90672499e-01
1.41040719e+00 -9.13538218e-01 -7.04657674e-01 7.85462111e-02
2.34367251e-02 -8.97627294e-01 9.44240808e-01 6.46896884e-02
-9.03745949e-01 -8.92314553e-01 -6.31823242e-01 2.47611463e-01
-2.46528178e-01 -2.97532499e-01 6.71332419e-01 2.56050318e-01
-6.60014093e-01 4.17623043e-01 -9.22885239e-01 -4.02589232e-01
5.57556570e-01 -9.97587070e-02 -3.55701596e-01 -3.11700761e-01
-1.03757143e+00 4.69065547e-01 4.24728572e-01 4.14820820e-01
-7.98171282e-01 -1.15406108e+00 -5.06123424e-01 -5.83419800e-01
4.48627949e-01 -8.54589820e-01 7.51301527e-01 -7.18926430e-01
-1.24412847e+00 6.42225862e-01 -3.96599859e-01 1.62526760e-02
5.51448345e-01 -1.63103621e-02 -6.44556701e-01 1.85483724e-01
1.72006395e-02 6.48378372e-01 7.10924745e-01 -1.51984298e+00
-1.15119922e+00 -3.40194553e-01 1.00370631e-01 3.20989966e-01
-3.16259444e-01 -1.31378040e-01 -1.15027201e+00 -8.74622762e-01
5.61182022e-01 -1.07650256e+00 -2.00395659e-02 3.98286402e-01
-3.83573443e-01 -2.04574689e-01 1.22910559e+00 -3.96414757e-01
1.15521801e+00 -2.26500201e+00 5.53908110e-01 1.35740995e-01
3.63018103e-02 -2.08459347e-01 1.69862285e-01 9.05566514e-02
1.75753191e-01 -2.74618477e-01 -2.64005452e-01 -2.79315233e-01
1.38125224e-02 6.93790197e-01 -4.12050575e-01 4.82545972e-01
1.85628645e-02 1.09564281e+00 -1.26328409e+00 -5.12677372e-01
4.12910521e-01 3.29595029e-01 -5.25743738e-02 1.19278833e-01
-3.97525162e-01 1.01187897e+00 -7.01163888e-01 9.57618296e-01
7.08422899e-01 -4.13932443e-01 -9.36731789e-03 -7.10266948e-01
-2.51713008e-01 -2.86956221e-01 -8.71997416e-01 2.25899005e+00
1.52514298e-02 5.33668160e-01 -4.57763612e-01 -4.45245236e-01
4.60924029e-01 3.74886617e-02 7.99352229e-01 -7.81081736e-01
-7.31196553e-02 1.43763032e-02 -1.99802548e-01 -7.90664077e-01
5.56969523e-01 -6.67363107e-02 -1.00573571e-02 1.44554645e-01
-1.18345343e-01 3.11972380e-01 1.49862200e-01 9.96004716e-02
7.69905210e-01 7.20561147e-01 -1.85330912e-01 1.26794446e-02
6.75050139e-01 1.37088984e-01 8.89549792e-01 5.62992036e-01
-3.47284377e-01 6.82226062e-01 1.71869591e-01 -4.30696607e-01
-6.98276281e-01 -1.06841314e+00 -7.24717081e-02 9.65568185e-01
7.86061287e-01 -2.21894011e-01 -5.19725159e-02 -6.92733049e-01
-3.80526893e-02 1.67514279e-01 -9.15906429e-01 7.49019608e-02
-6.92858219e-01 -7.05350399e-01 5.16734838e-01 5.96299469e-01
8.84665906e-01 -9.55701590e-01 -6.01677477e-01 2.37470388e-01
-4.53746617e-01 -1.20818460e+00 -6.17271841e-01 -1.14085570e-01
-8.63602698e-01 -9.83500063e-01 -1.03913510e+00 -4.89942044e-01
3.73214215e-01 4.10972029e-01 9.64849651e-01 -3.23872315e-03
-6.53409809e-02 6.55749023e-01 -5.53006470e-01 1.95132524e-01
6.57201856e-02 -4.79052782e-01 -9.70442817e-02 4.73699987e-01
1.12439238e-01 -7.67979980e-01 -9.43235397e-01 5.96333146e-01
-8.40388894e-01 2.38936931e-01 4.18564737e-01 7.65387952e-01
9.09978330e-01 2.23348528e-01 -2.84980293e-02 -4.28447604e-01
-2.93688923e-01 -4.87149060e-01 -4.18180346e-01 7.30085492e-01
-5.40331788e-02 -2.45921865e-01 1.68898210e-01 -7.14591742e-01
-1.42252743e+00 -4.40646708e-02 3.47836137e-01 -1.05172002e+00
-1.71150908e-01 2.75721639e-01 -2.09069565e-01 -1.10294975e-01
1.77620649e-01 7.04702556e-01 -4.51012939e-01 -2.83108473e-01
6.94520056e-01 -1.29623532e-01 7.32189119e-01 -7.04933107e-01
1.01244986e+00 9.53228116e-01 1.73802420e-01 -3.82385761e-01
-8.29410076e-01 -5.06761611e-01 -6.43736601e-01 -7.23212659e-01
1.19534659e+00 -7.87565112e-01 -6.72371268e-01 8.53802860e-01
-9.63409722e-01 -2.37047449e-01 -1.79420218e-01 3.49405318e-01
-5.37909508e-01 5.75440049e-01 -6.63787544e-01 -6.50456965e-01
2.64187127e-01 -1.04597819e+00 1.09064317e+00 5.48036039e-01
1.45677865e-01 -1.07506490e+00 3.50735933e-02 1.62719622e-01
1.08973436e-01 2.92602271e-01 6.52267754e-01 2.37865657e-01
-1.13315380e+00 4.64529008e-01 -3.41979980e-01 -3.36139679e-01
5.40653467e-01 8.76615755e-03 -8.41471136e-01 1.18178584e-01
-1.26693964e-01 6.96299374e-02 7.92419553e-01 3.72385651e-01
1.05817366e+00 -8.26819316e-02 -5.88862836e-01 7.54732251e-01
1.23981774e+00 2.04608619e-01 5.21213770e-01 5.31349421e-01
1.15149713e+00 7.36407697e-01 1.07059884e+00 3.58493775e-01
5.55132985e-01 9.80979383e-01 2.80668467e-01 1.54928237e-01
-4.20572668e-01 -4.32534307e-01 2.35219508e-01 8.97870719e-01
-2.98531294e-01 -1.26780160e-02 -6.26016557e-01 5.71229458e-01
-2.23797679e+00 -1.18168640e+00 -3.03532571e-01 1.72514176e+00
8.26499999e-01 -1.18480109e-01 2.92620748e-01 -1.45748392e-01
6.46949410e-01 4.21628237e-01 -6.40482128e-01 7.76222229e-01
-5.53579509e-01 -5.12744546e-01 4.43907052e-01 9.90051329e-02
-1.08421195e+00 5.68076074e-01 5.56116009e+00 1.12318480e+00
-9.94015455e-01 3.39312345e-01 5.19090950e-01 -3.80934738e-02
-4.35712695e-01 8.41090530e-02 -5.31173348e-01 7.97800243e-01
5.80295804e-04 1.50247946e-01 2.86805421e-01 2.45129526e-01
1.32597880e-02 -1.61998525e-01 -8.37890625e-01 1.24185121e+00
-2.34801322e-01 -1.35564387e+00 -7.60714412e-02 -9.62431058e-02
9.54164386e-01 -1.54210269e-01 2.88694441e-01 -4.25732918e-02
2.49893889e-01 -4.06277269e-01 1.11693358e+00 9.87430632e-01
3.06214601e-01 -5.01840115e-01 2.81122625e-01 1.12447992e-01
-1.93177390e+00 -1.09287696e-02 -2.37246856e-01 4.11550432e-01
5.46405971e-01 5.34442484e-01 3.75475287e-02 1.22767103e+00
1.06476665e+00 1.71655190e+00 -5.08229494e-01 9.67980921e-01
-1.07482649e-01 3.65437031e-01 -4.63594675e-01 3.24444324e-01
2.58774072e-01 -1.71496302e-01 6.48216903e-01 1.11233759e+00
3.50992709e-01 2.93841809e-01 2.15566799e-01 8.27127278e-01
3.51463377e-01 -5.70577204e-01 -3.36386442e-01 3.60309541e-01
3.92012686e-01 1.08805466e+00 -8.84201348e-01 -2.74765849e-01
-5.39901912e-01 9.15335417e-01 -8.03227276e-02 6.86630666e-01
-1.19081795e+00 4.81327981e-01 6.33777797e-01 -2.48821333e-01
5.22988796e-01 -2.23668516e-01 -2.96159476e-01 -1.28960001e+00
2.31434643e-01 -6.39856815e-01 6.94253206e-01 -1.10562992e+00
-1.68176925e+00 5.02671897e-01 1.43081203e-01 -2.06004977e+00
2.72010952e-01 -3.11381131e-01 -2.35350654e-01 5.54549932e-01
-1.76540124e+00 -1.61499369e+00 -4.87420678e-01 1.20983148e+00
5.05102932e-01 3.94945890e-02 4.66717519e-02 5.19171774e-01
-3.55167866e-01 4.42238897e-01 -2.25805044e-01 -3.49481702e-02
3.71755272e-01 -8.86167645e-01 -2.45456323e-01 1.15363395e+00
3.18314195e-01 7.84210443e-01 4.29454535e-01 -8.04100871e-01
-1.66725898e+00 -1.27589047e+00 1.27768576e-01 -3.76442552e-01
9.06808257e-01 1.96408078e-01 -8.37332129e-01 5.74451089e-01
3.50202382e-01 2.66830504e-01 4.47335869e-01 -6.50064349e-02
-6.34686828e-01 -2.85113811e-01 -8.27109635e-01 6.25087500e-01
1.55670607e+00 -6.19219422e-01 -6.57242179e-01 1.17307253e-01
7.80064464e-01 -7.12523520e-01 -1.06245267e+00 5.97430170e-01
7.97511339e-01 -1.04342771e+00 1.14186013e+00 -2.38432437e-01
3.92465681e-01 -1.06755602e+00 -4.77049857e-01 -8.97528410e-01
-3.73585016e-01 -6.15550160e-01 -4.08518165e-01 1.46287251e+00
4.90425713e-02 -3.57492834e-01 6.35462940e-01 4.44657564e-01
-1.53957814e-01 -5.61274648e-01 -1.04572237e+00 -5.62403917e-01
-4.11576688e-01 -6.81563318e-01 6.82615638e-01 9.69388783e-01
-2.18868226e-01 -4.87700671e-01 -7.03380823e-01 3.84052634e-01
9.46412325e-01 5.90129673e-01 6.88062251e-01 -5.67592382e-01
-5.76904416e-01 -5.34231305e-01 -5.38126409e-01 -1.59249377e+00
-2.39857167e-01 -4.56457496e-01 2.28204712e-01 -9.89330411e-01
4.30699319e-01 -7.44448245e-01 -7.69723117e-01 4.40012962e-01
-4.93163317e-01 5.47276974e-01 5.99956393e-01 6.36235356e-01
-1.01029539e+00 8.40477645e-01 1.70470071e+00 -3.21879089e-01
5.75834326e-02 -2.85021842e-01 -1.77746117e-01 6.14161611e-01
1.29502475e-01 -4.17604983e-01 -5.28706431e-01 -5.18999755e-01
7.48849809e-02 1.99124381e-01 5.24643362e-01 -1.02246320e+00
4.99733001e-01 -4.81224507e-01 5.06111383e-01 -1.05996096e+00
4.97873634e-01 -1.21302438e+00 7.62502551e-01 3.14972848e-01
-1.57884955e-01 -8.41468871e-02 1.27463087e-01 1.06252694e+00
-3.44742119e-01 4.55538005e-01 5.80184579e-01 -6.34102523e-02
-1.24644101e+00 7.50756800e-01 -5.00281081e-02 -1.60397276e-01
1.11714113e+00 -5.33282340e-01 -4.84811664e-01 -1.21420510e-01
-8.86853695e-01 4.26161945e-01 5.60466588e-01 7.20712006e-01
6.24373376e-01 -1.72014463e+00 -1.97557732e-01 2.55387500e-02
4.04489338e-01 2.14262363e-02 9.06962693e-01 1.29038489e+00
-6.93442896e-02 9.43132862e-02 -3.39525849e-01 -1.38600349e+00
-1.18021166e+00 5.57130396e-01 3.99685651e-01 -1.72159135e-01
-9.44141448e-01 9.22493577e-01 5.52928150e-01 -1.12429991e-01
1.58793554e-01 -4.51845556e-01 -1.92704290e-01 -3.19035351e-02
5.82372546e-01 2.12574080e-01 -4.92389858e-01 -9.37260628e-01
-4.91045862e-01 1.31812060e+00 2.61256665e-01 -6.73378035e-02
1.13218403e+00 -7.06847489e-01 -2.38595665e-01 8.49197268e-01
8.75557601e-01 -1.10107221e-01 -1.76887047e+00 -6.70411170e-01
2.12416351e-02 -9.00672853e-01 -4.44158822e-01 -5.47566772e-01
-1.57457042e+00 4.28475469e-01 6.83582783e-01 8.58193487e-02
1.51376045e+00 2.13507086e-01 7.57687747e-01 -1.71651304e-01
7.27839410e-01 -4.55501676e-01 3.75527531e-01 3.92958492e-01
6.38130248e-01 -9.94959831e-01 -5.24820127e-02 -6.01630270e-01
-7.27182984e-01 1.01002455e+00 5.80718577e-01 2.36886337e-01
7.69012213e-01 4.84274887e-02 1.04898205e-02 -3.50103289e-01
-7.50572503e-01 -3.65521133e-01 4.89974767e-01 8.18284035e-01
4.05418649e-02 -1.01923600e-01 1.37168512e-01 2.12623209e-01
3.42180759e-01 1.53810129e-01 -3.64982001e-02 9.73257720e-01
-3.48016284e-02 -1.05919015e+00 -5.33987641e-01 -6.46954849e-02
-1.35546654e-01 2.42234975e-01 -9.59667489e-02 7.70255148e-01
5.69765389e-01 9.01991785e-01 -1.59772113e-01 -4.46573347e-01
3.29023927e-01 -1.05627015e-01 7.36913443e-01 -6.24455512e-03
-5.01506031e-01 4.06581521e-01 -1.59637138e-01 -8.69741499e-01
-1.38100958e+00 -8.77654195e-01 -1.08653748e+00 -2.93763012e-01
-1.09071191e-02 -4.03059989e-01 3.12589519e-02 1.09045494e+00
3.11290592e-01 8.50417197e-01 5.51786005e-01 -8.28384697e-01
2.80148208e-01 -8.32705438e-01 -5.31654298e-01 7.04163730e-01
4.69226032e-01 -1.01840937e+00 -6.04015030e-02 4.79372054e-01] | [9.364490509033203, 0.17459428310394287] |
3c1cee1a-d44c-4bf6-bd14-f6999eee40fa | aspect-sentiment-triplet-extraction-using | 2108.06107 | null | https://arxiv.org/abs/2108.06107v1 | https://arxiv.org/pdf/2108.06107v1.pdf | Aspect Sentiment Triplet Extraction Using Reinforcement Learning | Aspect Sentiment Triplet Extraction (ASTE) is the task of extracting triplets of aspect terms, their associated sentiments, and the opinion terms that provide evidence for the expressed sentiments. Previous approaches to ASTE usually simultaneously extract all three components or first identify the aspect and opinion terms, then pair them up to predict their sentiment polarities. In this work, we present a novel paradigm, ASTE-RL, by regarding the aspect and opinion terms as arguments of the expressed sentiment in a hierarchical reinforcement learning (RL) framework. We first focus on sentiments expressed in a sentence, then identify the target aspect and opinion terms for that sentiment. This takes into account the mutual interactions among the triplet's components while improving exploration and sample efficiency. Furthermore, this hierarchical RLsetup enables us to deal with multiple and overlapping triplets. In our experiments, we evaluate our model on existing datasets from laptop and restaurant domains and show that it achieves state-of-the-art performance. The implementation of this work is publicly available at https://github.com/declare-lab/ASTE-RL. | ['Soujanya Poria', 'Navonil Majumder', 'Tapas Nayak', 'Samson Yu Bai Jian'] | 2021-08-13 | null | null | null | null | ['aspect-sentiment-triplet-extraction'] | ['natural-language-processing'] | [-3.37621905e-02 -3.50783542e-02 -3.27066094e-01 -5.97143173e-01
-1.10265017e+00 -8.49152863e-01 3.89455557e-01 4.09100473e-01
-1.05563976e-01 6.24178529e-01 4.48391467e-01 -3.88663739e-01
2.17283770e-01 -7.44439006e-01 -6.24656558e-01 -5.89126229e-01
1.87386014e-02 5.76220334e-01 -1.67290822e-01 -5.28176248e-01
3.95918608e-01 -2.17476025e-01 -1.52868247e+00 7.63938427e-01
7.00991809e-01 1.25689375e+00 -3.24381977e-01 4.88106608e-01
-2.70841897e-01 9.67938542e-01 -6.45692647e-01 -7.28669524e-01
3.42516489e-02 -3.37326169e-01 -8.13909113e-01 8.55138972e-02
1.51104927e-02 1.19448796e-01 5.15308082e-01 1.06328332e+00
3.65442425e-01 -8.43017101e-02 4.99382377e-01 -1.53349841e+00
-5.92432022e-01 9.11856771e-01 -7.56981611e-01 -1.46689087e-01
4.74327385e-01 -1.11612126e-01 1.86796021e+00 -9.72240508e-01
4.75569665e-01 1.13161397e+00 5.34390450e-01 2.99866199e-01
-9.69432712e-01 -5.97633064e-01 6.93085372e-01 -2.35510245e-03
-6.52442753e-01 -2.59949476e-01 9.29179072e-01 -3.06447297e-01
1.21272218e+00 3.28641564e-01 1.06620479e+00 1.02278852e+00
1.51348665e-01 1.41483653e+00 1.57447970e+00 -2.90516883e-01
3.43792439e-01 5.28047383e-01 7.89420068e-01 5.98114073e-01
1.81333989e-01 -2.35166222e-01 -9.11483943e-01 -3.96340936e-01
-6.83148578e-02 -2.92559028e-01 3.61710973e-02 -2.49071717e-01
-9.10475671e-01 9.93672252e-01 1.90784171e-01 9.38190073e-02
-4.94964480e-01 -1.28963664e-01 4.17946905e-01 3.23388904e-01
7.53828406e-01 6.51582003e-01 -1.03019595e+00 -1.57286361e-01
-3.27127993e-01 2.88552701e-01 1.24193799e+00 8.34980249e-01
9.08330917e-01 -4.75919545e-01 -7.12559074e-02 8.21567595e-01
4.60599005e-01 4.66122866e-01 3.75870943e-01 -5.43895125e-01
6.33917689e-01 1.13997018e+00 1.74897432e-01 -8.90118301e-01
-3.75342876e-01 -1.63251758e-01 -3.82545024e-01 -7.47297034e-02
-1.87076673e-01 -5.30254960e-01 -6.66850150e-01 1.57371151e+00
6.78882599e-01 -2.82669276e-01 4.39060062e-01 6.21733308e-01
9.16012704e-01 5.45835912e-01 -2.37107258e-02 1.49800582e-03
1.88421285e+00 -1.27587700e+00 -6.97838664e-01 -6.83935344e-01
5.79942882e-01 -8.60009134e-01 1.08513474e+00 3.72862846e-01
-9.07666445e-01 2.78699235e-03 -9.63157594e-01 3.63743976e-02
-6.11817598e-01 4.15489495e-01 8.93151402e-01 3.16784650e-01
-8.33191633e-01 3.56573820e-01 -7.00591028e-01 9.54191834e-02
1.75080314e-01 4.94922966e-01 -1.56368732e-01 2.99486458e-01
-1.36415601e+00 6.59956574e-01 8.23583156e-02 1.59790069e-01
-2.87311643e-01 -4.55063522e-01 -1.03567600e+00 5.25252484e-02
5.21889985e-01 -7.67501950e-01 1.46919620e+00 -1.14287984e+00
-1.37997210e+00 7.48622715e-01 -5.09859264e-01 -1.61860242e-01
-1.39479280e-01 -4.92079824e-01 -4.01912302e-01 -1.78551912e-01
4.08057332e-01 3.77471834e-01 6.03790104e-01 -1.35249209e+00
-8.65331471e-01 -4.67821002e-01 5.92938960e-01 5.82832932e-01
-5.22698641e-01 2.06517279e-01 -4.90139544e-01 -3.36967856e-01
-1.03580661e-01 -1.11762500e+00 -3.38416427e-01 -7.32297659e-01
-6.12756848e-01 -4.98250008e-01 4.69168842e-01 -1.71893626e-01
1.26312470e+00 -1.96183491e+00 2.92106788e-03 3.36694568e-01
4.89737466e-02 -7.61684701e-02 -2.81040579e-01 4.46255505e-01
-8.86840001e-02 1.69225596e-02 -1.89478502e-01 -5.40435433e-01
2.11598754e-01 8.07531998e-02 -4.34382349e-01 -1.28536209e-01
5.53751469e-01 9.44438994e-01 -9.41421986e-01 -4.73662734e-01
-2.82681048e-01 3.11949283e-01 -7.92009234e-01 3.06733817e-01
-5.31149447e-01 2.38186270e-01 -8.76029611e-01 8.16992760e-01
3.95104229e-01 -5.00087321e-01 3.74847531e-01 -3.36817175e-01
2.82012485e-02 8.56451631e-01 -1.14693058e+00 1.06992459e+00
-5.38767874e-01 2.87141919e-01 -2.00192183e-01 -6.97740495e-01
1.00289428e+00 1.20554082e-01 6.06309772e-01 -5.83324373e-01
2.10004732e-01 2.85457373e-01 -3.43287349e-01 -5.40910125e-01
7.85743415e-01 -1.62877217e-01 -4.98020053e-01 8.31005692e-01
2.18004696e-02 -1.64904922e-01 7.30993927e-01 2.04774916e-01
8.13827872e-01 2.72694170e-01 5.03646195e-01 -1.48831725e-01
5.75474083e-01 9.51657891e-02 5.09961605e-01 1.91044092e-01
1.30625337e-01 1.27309322e-01 1.14074397e+00 -4.72384185e-01
-5.55049181e-01 -5.47034383e-01 2.40458757e-01 1.23859322e+00
1.70087844e-01 -9.81058180e-01 -4.60854530e-01 -9.98637021e-01
-1.97926819e-01 7.74957180e-01 -9.17501569e-01 9.51629356e-02
-4.02450621e-01 -1.07067013e+00 3.74005586e-02 4.32990015e-01
3.20032746e-01 -1.17851079e+00 -5.15813112e-01 2.59317993e-03
-5.69134176e-01 -1.12691200e+00 -2.75320262e-01 5.31875253e-01
-5.19679129e-01 -1.12279952e+00 -1.30059332e-01 -6.64656222e-01
7.25621760e-01 -4.78954762e-02 1.63067412e+00 -2.39907876e-02
2.52582043e-01 9.94970798e-02 -5.64668596e-01 -6.65057957e-01
-1.06724240e-01 1.18905835e-01 -2.26657629e-01 9.44342613e-02
7.74180591e-01 -5.14397621e-01 -7.05766797e-01 1.55274108e-01
-6.94914162e-01 1.05957100e-02 6.08676851e-01 8.24308753e-01
8.23879600e-01 -1.06894098e-01 3.17945451e-01 -1.38074291e+00
8.84358048e-01 -6.30380869e-01 -5.95800936e-01 2.90301114e-01
-7.54763901e-01 2.10458606e-01 3.99873525e-01 -7.92645141e-02
-9.20546949e-01 7.55749270e-03 -2.61307716e-01 1.05133124e-01
7.53943101e-02 8.75058591e-01 -1.55838653e-02 5.91434717e-01
2.77029932e-01 3.74978259e-02 -4.68116403e-01 -1.39517456e-01
3.30928504e-01 6.67492568e-01 -9.30262357e-02 -7.59573877e-01
3.73490185e-01 4.36616510e-01 -2.86516815e-01 -2.99823403e-01
-1.67647898e+00 -7.28058457e-01 -1.88503310e-01 -3.35884057e-02
5.51463366e-01 -1.00417161e+00 -8.65499496e-01 2.52945065e-01
-1.04969943e+00 -9.27756056e-02 -3.16449642e-01 2.28955626e-01
-4.93371516e-01 -4.13571149e-02 -5.81936657e-01 -1.04015350e+00
-7.09611177e-01 -1.38034129e+00 1.57174122e+00 3.70989084e-01
-4.71721411e-01 -9.67784882e-01 4.65292722e-01 5.22197962e-01
2.11248636e-01 8.99039507e-02 8.19295406e-01 -9.25262272e-01
-3.35837483e-01 -1.42971098e-01 8.21827278e-02 3.54891360e-01
2.60382861e-01 9.21870545e-02 -9.80951130e-01 1.77913569e-02
1.61544800e-01 -6.29873991e-01 8.18797171e-01 2.35814273e-01
6.18157685e-01 -5.79433799e-01 -3.27880234e-02 2.31040090e-01
1.30157089e+00 -3.00600044e-02 1.13877922e-01 7.49646842e-01
5.00317454e-01 7.80837357e-01 1.07305539e+00 5.07454395e-01
8.30553889e-01 4.14044112e-01 3.24808985e-01 -1.87589452e-01
4.54910368e-01 -1.96638748e-01 7.53887355e-01 1.05078709e+00
1.64121300e-01 -3.00411910e-01 -7.23125637e-01 6.17811739e-01
-2.08483577e+00 -6.19718075e-01 -9.83095989e-02 1.51876616e+00
1.05698025e+00 3.50864738e-01 1.38429806e-01 -1.01696039e-02
3.73365432e-01 3.45995098e-01 -5.61913908e-01 -7.38425255e-01
-1.75178632e-01 8.71677175e-02 -1.14110829e-02 5.13683856e-01
-1.23464310e+00 1.01307094e+00 5.66978884e+00 4.62358505e-01
-9.35338378e-01 -1.46923900e-01 8.58860016e-01 -8.33300799e-02
-7.72959590e-01 1.86666742e-01 -1.03852010e+00 4.85715596e-03
7.24060178e-01 -1.57011479e-01 2.25264907e-01 1.06628060e+00
-1.77987233e-01 -1.84308544e-01 -1.07611156e+00 3.99724603e-01
1.37252301e-01 -9.44324613e-01 1.52686208e-01 -1.85685903e-01
9.87719059e-01 9.76202786e-02 2.15489388e-01 4.19176191e-01
6.28546953e-01 -7.07067609e-01 6.86878383e-01 2.09344149e-01
2.78037399e-01 -7.72803664e-01 9.65963781e-01 8.16265773e-03
-1.28920019e+00 -1.33635104e-01 2.30858121e-02 -1.53754996e-02
9.08786729e-02 8.91915560e-01 -7.30658770e-01 6.44882798e-01
8.57998312e-01 1.06803632e+00 -4.72802371e-01 3.67572039e-01
-7.02036202e-01 5.62224388e-01 -2.27135241e-01 -5.56394637e-01
3.90976161e-01 -4.26251262e-01 4.77246284e-01 1.04188633e+00
1.66916192e-01 -1.80670600e-02 1.17120691e-01 5.94873428e-01
-1.41395465e-01 3.48666668e-01 -4.65943307e-01 -2.93003440e-01
5.74918054e-02 1.74957252e+00 -7.27699637e-01 -5.29148936e-01
-5.59460521e-01 6.68068171e-01 4.56910759e-01 3.00944209e-01
-7.81992137e-01 -1.79357082e-01 8.06040108e-01 -5.14660716e-01
6.00783885e-01 2.80531585e-01 -3.13006461e-01 -1.46474004e+00
3.80843043e-01 -1.39619684e+00 5.25683105e-01 -7.09075391e-01
-1.42118144e+00 1.05330265e+00 -2.38138318e-01 -1.29889798e+00
-4.02086258e-01 -7.12836504e-01 -5.82337618e-01 5.58135450e-01
-1.86302984e+00 -1.23270237e+00 4.85620163e-02 5.31038702e-01
6.18641376e-01 2.36188546e-02 1.05397964e+00 5.99996857e-02
-5.63411415e-01 2.96490371e-01 -3.18649769e-01 1.27466530e-01
5.75864911e-01 -1.56024098e+00 5.05266011e-01 5.19072294e-01
3.51664126e-01 8.19240630e-01 7.49697089e-01 -5.29273093e-01
-1.32267177e+00 -8.18069518e-01 1.21720397e+00 -7.42700994e-01
1.06685030e+00 -3.78520459e-01 -4.55383331e-01 7.77486503e-01
5.22667885e-01 -4.09930319e-01 1.11559367e+00 7.33879387e-01
-6.19596303e-01 -1.35449782e-01 -5.83206594e-01 7.93856263e-01
4.95670438e-01 -4.77575779e-01 -6.62623405e-01 4.94368315e-01
7.24215925e-01 -3.92198443e-01 -7.38560081e-01 5.02759814e-01
6.20574713e-01 -1.18415058e+00 5.85517168e-01 -6.88211381e-01
9.58624244e-01 -2.79118448e-01 -2.33484656e-01 -1.49842167e+00
1.65364891e-01 -3.93841803e-01 -2.42684677e-01 1.36418235e+00
1.17749286e+00 -4.94998842e-01 6.60651863e-01 5.05145669e-01
1.01321958e-01 -1.45303011e+00 -4.37367678e-01 -1.52120680e-01
-2.22875819e-01 -3.90366137e-01 8.75574648e-01 7.57383943e-01
3.57070446e-01 1.10605228e+00 -2.21325263e-01 1.03075802e-01
2.69696087e-01 1.02220201e+00 6.12414181e-01 -8.34981501e-01
-5.27745426e-01 -4.16892558e-01 2.05888242e-01 -8.09741020e-01
3.19814116e-01 -7.02916861e-01 1.52430445e-01 -1.57293272e+00
4.15826678e-01 -3.99766475e-01 -6.30980253e-01 7.58781493e-01
-5.52735507e-01 1.37950942e-01 2.85696778e-02 -2.48529725e-02
-1.05825949e+00 6.73107743e-01 1.04631162e+00 -2.70915210e-01
-2.60668188e-01 3.14003676e-01 -1.38303792e+00 1.04701614e+00
1.18207395e+00 -5.99979460e-01 -3.68381202e-01 -3.06841850e-01
9.85899568e-01 -2.61636078e-01 -2.01000869e-01 -3.89525056e-01
1.44786537e-01 -5.34811243e-02 1.13559373e-01 -7.45538652e-01
4.46738511e-01 -6.71891510e-01 -4.77284729e-01 2.10742727e-01
-5.80794156e-01 5.70648193e-01 2.82131225e-01 3.22722733e-01
-5.97139180e-01 -5.33264354e-02 2.48484343e-01 -1.24837302e-01
-3.81327987e-01 6.31040782e-02 -1.31876960e-01 1.88520104e-01
7.46657610e-01 4.73970890e-01 -4.16281641e-01 -4.03512865e-01
-5.61634600e-01 5.65789163e-01 1.71571627e-01 5.53105116e-01
5.94388425e-01 -1.20298266e+00 -6.31710291e-01 1.54555559e-01
3.82507384e-01 -1.56689569e-01 -9.35979784e-02 8.62136602e-01
1.62373602e-01 2.52103776e-01 5.60875721e-02 -3.06191742e-01
-1.53212500e+00 3.68781179e-01 1.05828479e-01 -1.07224762e+00
-8.21338501e-03 9.56525922e-01 2.17503205e-01 -8.02295148e-01
-2.41769813e-02 -3.24896157e-01 -7.55853117e-01 4.06767488e-01
3.17097425e-01 -1.12484768e-01 5.89210093e-02 -5.43718874e-01
-5.30948818e-01 7.14239776e-01 -1.93781585e-01 -3.08728158e-01
1.53676426e+00 -1.57974921e-02 -4.90643561e-01 6.50242627e-01
1.19292223e+00 4.16060030e-01 -6.39573276e-01 -9.47567970e-02
1.94258794e-01 -6.77711666e-02 -2.67517000e-01 -8.57131004e-01
-1.09326947e+00 5.83867788e-01 -5.24374768e-02 3.99696440e-01
1.28704941e+00 2.22926944e-01 8.13888609e-01 5.65308571e-01
1.17156319e-01 -1.06011319e+00 1.35041296e-01 7.01313555e-01
7.65224278e-01 -1.34213853e+00 3.46310772e-02 -3.65943968e-01
-1.17657936e+00 8.59587550e-01 7.80178249e-01 -1.04539745e-01
6.73313618e-01 4.32339311e-01 3.91459912e-01 -5.74141204e-01
-1.34216094e+00 -4.56113070e-01 3.58053088e-01 5.14738001e-02
7.54520714e-01 1.21208057e-01 -3.79619241e-01 8.34330738e-01
-5.79959095e-01 -2.13307142e-01 2.48051271e-01 1.02435231e+00
-1.88114032e-01 -1.41459894e+00 -1.03433644e-02 4.74211425e-01
-6.65396988e-01 -6.11738443e-01 -7.74571657e-01 4.15538192e-01
-3.02739744e-03 1.13964748e+00 -2.87923336e-01 -4.60520238e-01
7.05464363e-01 3.97518538e-02 -3.68887326e-03 -5.83384275e-01
-1.00211227e+00 2.66444474e-01 6.41798615e-01 -7.08270550e-01
-7.44038105e-01 -7.02978611e-01 -1.16125810e+00 1.60212636e-01
-4.24132705e-01 7.08037794e-01 8.27961922e-01 9.01202142e-01
4.36649293e-01 4.91529465e-01 9.37801063e-01 -4.09072399e-01
-4.15606737e-01 -7.66595960e-01 -3.60170186e-01 4.47956055e-01
4.13443893e-01 -4.05513555e-01 -3.32984954e-01 -1.26947671e-01] | [11.491742134094238, 6.6636433601379395] |
6e3785aa-8c3f-4cde-939a-c54bd497309c | importance-attribution-in-neural-networks-by | 2302.03132 | null | https://arxiv.org/abs/2302.03132v1 | https://arxiv.org/pdf/2302.03132v1.pdf | Importance attribution in neural networks by means of persistence landscapes of time series | We propose and implement a method to analyze time series with a neural network using a matrix of area-normalized persistence landscapes obtained through topological data analysis. We include a gating layer in the network's architecture that is able to identify the most relevant landscape levels for the classification task, thus working as an importance attribution system. Next, we perform a matching between the selected landscape functions and the corresponding critical points of the original time series. From this matching we are able to reconstruct an approximate shape of the time series that gives insight into the classification decision. We test this technique with input data from a dataset of electrocardiographic signals. | ['Oriol Pujol', 'Carles Casacuberta', 'Aina Ferrà'] | 2023-02-06 | null | null | null | null | ['topological-data-analysis'] | ['graphs'] | [ 2.20219865e-01 -1.40976191e-01 -8.30664784e-02 -4.19399202e-01
-1.10853766e-03 -5.71568191e-01 5.86977541e-01 7.87170231e-01
-4.26259547e-01 7.33811915e-01 1.14110261e-01 -3.15103948e-01
-8.05653632e-01 -1.13633299e+00 -4.03144240e-01 -6.53240621e-01
-7.70614803e-01 3.26675147e-01 3.87716919e-01 -4.91022289e-01
3.99750680e-01 8.26802313e-01 -1.55237424e+00 4.92027283e-01
7.21028268e-01 1.00968397e+00 -1.46845460e-01 5.25096714e-01
4.21987325e-02 3.90322179e-01 -7.03272641e-01 2.46593744e-01
1.48112342e-01 -6.47570431e-01 -6.67527378e-01 -6.50293112e-01
1.29529312e-01 5.21761477e-01 -1.09031312e-01 1.13443899e+00
3.39546502e-01 5.20993620e-02 1.02006960e+00 -7.52008975e-01
6.03498667e-02 5.22425771e-01 8.82800296e-02 7.08181798e-01
2.17898548e-01 -7.10549951e-02 9.32441950e-01 -4.40576851e-01
1.12832201e+00 8.76530468e-01 9.16940272e-01 1.49410646e-02
-1.66855907e+00 3.96655574e-02 -3.63349885e-01 4.60805237e-01
-1.23510277e+00 -1.62560001e-01 1.20711482e+00 -6.20576382e-01
6.08551860e-01 4.24170583e-01 1.33412921e+00 8.04568112e-01
6.79759741e-01 -3.24896798e-02 1.33681905e+00 -3.95180851e-01
5.59979498e-01 -1.35090277e-01 4.29056793e-01 5.03669500e-01
1.36172041e-01 9.75466892e-02 -2.90981144e-01 -3.40216875e-01
7.49262810e-01 -1.17308713e-01 -3.34443480e-01 -4.75768507e-01
-1.01299071e+00 5.94038844e-01 8.86661410e-01 1.00043058e+00
-7.14918554e-01 1.35085389e-01 4.91382360e-01 7.19809175e-01
5.47829390e-01 6.70372546e-01 -2.30821386e-01 1.04998812e-01
-8.61807227e-01 2.02851683e-01 6.30668700e-01 -2.11244658e-01
7.83246696e-01 -2.28998989e-01 -5.89603670e-02 2.67208815e-01
-2.37600529e-03 1.57762602e-01 3.48446310e-01 -5.29539168e-01
3.52418497e-02 1.04599547e+00 -2.02927217e-01 -1.36915672e+00
-8.47065628e-01 -7.51774311e-01 -1.08445215e+00 4.36476648e-01
6.82792664e-01 1.51929229e-01 -4.30320144e-01 1.76484561e+00
1.90434441e-01 4.17449623e-02 -1.48627013e-01 5.67426980e-01
9.02118981e-02 5.67389131e-01 -3.93512733e-02 -3.23362559e-01
1.32499182e+00 1.08577482e-01 -5.72029769e-01 4.47205007e-01
5.55091739e-01 -5.20857535e-02 8.97594512e-01 4.84648257e-01
-7.72486150e-01 -6.05502665e-01 -1.36579001e+00 6.01256490e-01
-8.03794086e-01 1.74356386e-01 9.29119065e-02 2.33673349e-01
-1.17997015e+00 1.45221138e+00 -7.69604981e-01 -4.25165653e-01
2.24381402e-01 2.33164608e-01 -3.76509696e-01 8.93838525e-01
-1.35428631e+00 1.16350198e+00 6.53682590e-01 2.87496090e-01
-2.71752208e-01 -4.49918807e-01 -3.47132981e-01 1.42540440e-01
-3.56044263e-01 -3.73752505e-01 3.53880614e-01 -1.11704373e+00
-1.07590866e+00 7.00443506e-01 1.30636841e-01 -5.67282200e-01
5.29395223e-01 5.11909187e-01 -4.11023527e-01 3.81355822e-01
-2.20601499e-01 1.09096900e-01 8.01795304e-01 -9.79347289e-01
-2.17732061e-02 -5.28266251e-01 -2.64301151e-01 -3.35445195e-01
-5.14974952e-01 -4.96599257e-01 3.93874019e-01 -6.93548322e-01
2.86216557e-01 -5.34936488e-01 -2.91988462e-01 -6.00432903e-02
-2.31747523e-01 1.73221752e-02 5.88856101e-01 -6.09357297e-01
1.59912717e+00 -2.15950251e+00 5.86340308e-01 9.44619834e-01
4.76984799e-01 -5.94843887e-02 2.58636903e-02 6.71037674e-01
-5.33395469e-01 3.65111046e-02 -5.35444319e-01 3.15755159e-01
-4.27334666e-01 -1.23837665e-01 -2.84538269e-01 6.68372810e-01
2.03052789e-01 8.22144866e-01 -6.42694652e-01 -2.07900017e-01
1.56364739e-01 4.53434914e-01 -2.40347132e-01 -1.85731351e-02
-2.43567452e-01 4.99897152e-01 -4.38381106e-01 1.76132783e-01
4.26757276e-01 -6.96833357e-02 4.86864567e-01 -3.02533388e-01
-2.39492774e-01 1.58506468e-01 -8.06569278e-01 1.36449659e+00
-2.21955791e-01 9.85446155e-01 -3.81970763e-01 -1.20458817e+00
1.43847954e+00 3.49586904e-02 6.69902086e-01 -7.43489921e-01
3.24184865e-01 3.90807450e-01 1.79769427e-01 -2.52271265e-01
1.79011784e-02 -8.89452174e-02 7.22659426e-03 5.15828788e-01
-4.47423160e-02 3.31466675e-01 1.08766250e-01 -3.92124295e-01
1.27629316e+00 -1.03846923e-01 3.46463263e-01 -1.04161763e+00
7.83405542e-01 -5.18508963e-02 3.55435908e-02 4.41684008e-01
-1.09877154e-01 2.83658713e-01 1.12487137e+00 -1.28311062e+00
-1.11430192e+00 -1.03620851e+00 -4.95689124e-01 4.13948774e-01
-1.03121534e-01 -2.48185858e-01 -8.75806153e-01 -3.32870185e-01
-8.62022266e-02 3.40775400e-01 -1.25761521e+00 -4.77673322e-01
-7.58604228e-01 -6.86070979e-01 5.77273428e-01 7.45337307e-02
7.93359987e-03 -1.34556437e+00 -1.34788752e+00 3.94089043e-01
1.48163131e-02 -2.81837702e-01 2.40780517e-01 5.76319218e-01
-1.34991360e+00 -1.10075784e+00 -2.79320180e-01 -5.16472101e-01
5.70638001e-01 -6.72005832e-01 1.05457199e+00 2.50870347e-01
-4.82419848e-01 -1.28178939e-01 -1.57369867e-01 -7.60657787e-02
-7.63121128e-01 2.68977433e-01 -4.55681719e-02 2.72807598e-01
-5.83444051e-02 -1.10157216e+00 -7.01802075e-01 2.29525656e-01
-9.95411873e-01 -2.59332389e-01 2.74582684e-01 5.84297836e-01
5.58750749e-01 2.07057551e-01 6.74067974e-01 -4.07889754e-01
9.79955733e-01 -3.67026299e-01 -7.08234847e-01 3.61575305e-01
-5.70286810e-01 4.09850091e-01 9.38999534e-01 -4.18330818e-01
-3.79786074e-01 -2.74976641e-02 -4.25934754e-02 -2.02780902e-01
-2.49007512e-02 6.54717505e-01 2.00040802e-01 -2.87869275e-01
1.05093682e+00 2.48295158e-01 1.15057692e-01 -5.20832837e-01
2.16034129e-01 2.68038929e-01 3.92257303e-01 -3.14621776e-01
3.82393688e-01 4.95997190e-01 4.64202791e-01 -7.96648681e-01
6.97107334e-03 -1.05649218e-01 -1.09009945e+00 -7.29134500e-01
7.09939539e-01 -4.19334099e-02 -9.10065770e-01 2.09441349e-01
-1.31128037e+00 -1.74524650e-01 -6.88985527e-01 1.51834428e-01
-7.57062972e-01 4.09508683e-02 -2.70182222e-01 -7.91224003e-01
-2.98057437e-01 -7.75405765e-01 6.74487650e-01 -1.29301429e-01
-2.76050001e-01 -1.29340398e+00 7.56923020e-01 -6.95929706e-01
5.79467952e-01 8.53456140e-01 1.52837431e+00 -5.38148701e-01
-3.74319702e-01 -2.84249783e-01 1.64029971e-01 1.31093606e-01
-2.13597715e-02 2.62115419e-01 -7.96335995e-01 -5.68847060e-02
2.17766210e-01 3.74537706e-01 1.11999583e+00 6.31791055e-01
1.02553272e+00 -7.21641034e-02 -4.80600983e-01 6.23993576e-01
1.62130964e+00 2.80694276e-01 7.72737443e-01 4.15536433e-01
2.07248226e-01 9.94072378e-01 2.26306483e-01 2.45503172e-01
-2.79581845e-01 8.16425920e-01 3.49860579e-01 6.02808744e-02
1.40507475e-01 -1.85275257e-01 2.19154790e-01 6.79039538e-01
-3.40803087e-01 1.01642713e-01 -1.26804566e+00 2.66390145e-01
-1.95745361e+00 -1.17009664e+00 -2.40013391e-01 2.32930636e+00
4.58772451e-01 4.79476392e-01 4.03760463e-01 5.52337110e-01
6.99491322e-01 2.70464540e-01 -3.98112863e-01 -5.51527083e-01
-1.71903655e-01 3.50067496e-01 1.68596014e-01 5.07938683e-01
-9.44255114e-01 3.58168185e-01 7.35437965e+00 4.72023308e-01
-1.29103708e+00 -2.18717188e-01 6.45280004e-01 2.80062348e-01
-3.90496105e-01 -8.28807428e-03 1.12949267e-01 2.73684591e-01
1.33034742e+00 -3.15102488e-01 3.62880111e-01 4.80416507e-01
2.72948176e-01 -2.96903756e-02 -1.00659978e+00 6.97554469e-01
-3.73601645e-01 -1.63446355e+00 1.71159327e-01 9.51994509e-02
1.00410484e-01 -9.09741223e-03 -1.88725188e-01 -3.15320998e-01
-5.95302284e-01 -1.04868209e+00 8.13553572e-01 1.25646067e+00
6.72278464e-01 -6.74490333e-01 6.18918300e-01 2.53076881e-01
-1.34684217e+00 -3.49346399e-01 -1.41132161e-01 -2.08439335e-01
-3.97375189e-02 7.46962309e-01 -6.30113423e-01 5.25783420e-01
3.43460083e-01 9.13058460e-01 -8.13669503e-01 1.09894800e+00
1.94171458e-01 3.60225886e-01 -2.92904705e-01 -4.18833494e-01
-1.09797232e-01 -5.67292094e-01 9.20435429e-01 1.12884128e+00
3.45671743e-01 -3.26277286e-01 -4.37770009e-01 1.20521748e+00
2.63933003e-01 2.71065682e-01 -1.04761088e+00 -8.50591734e-02
1.04678243e-01 1.16084766e+00 -1.48327303e+00 -2.50272334e-01
5.39482236e-01 5.55136979e-01 3.70193392e-01 1.76022217e-01
-4.31718439e-01 -6.03757262e-01 5.42932153e-01 2.53565729e-01
-5.55904657e-02 -2.89577395e-01 -4.17521775e-01 -9.42067146e-01
5.72531670e-02 -3.88602287e-01 2.23613918e-01 -7.23637879e-01
-9.92319942e-01 1.15908837e+00 2.13172473e-03 -1.42874992e+00
-2.60344923e-01 -7.19423115e-01 -8.52857888e-01 9.65837777e-01
-9.57728505e-01 -4.19966280e-01 -1.12031162e-01 4.01681691e-01
-1.76199734e-01 2.82487813e-02 9.28868413e-01 1.34056553e-01
-2.22957775e-01 -2.22146392e-01 1.45866200e-01 -1.53411329e-01
8.56937841e-02 -1.52699220e+00 5.23233354e-01 5.65966129e-01
1.88065991e-01 7.11680293e-01 8.32000077e-01 -6.35721982e-01
-9.73935306e-01 -7.58153379e-01 8.49894166e-01 -6.49117172e-01
7.24621713e-01 -4.71536696e-01 -1.14480007e+00 -1.52541175e-02
-3.19875404e-02 -2.79941633e-02 3.89198512e-01 -1.48809254e-01
-6.84671849e-02 -4.18142140e-01 -9.77570713e-01 2.63836563e-01
8.48719001e-01 -7.76819646e-01 -6.14160597e-01 5.39503386e-03
2.68922657e-01 2.03440636e-01 -1.02465820e+00 2.75320560e-01
8.43135297e-01 -1.27812290e+00 8.30484629e-01 -5.56863189e-01
3.38167161e-01 -3.08960170e-01 2.56859869e-01 -1.39175630e+00
-2.86077291e-01 -5.33905864e-01 9.90828425e-02 5.01825094e-01
3.96616340e-01 -8.67400050e-01 4.50372726e-01 -3.64723325e-01
4.29814279e-01 -9.89347756e-01 -1.31550932e+00 -5.65621912e-01
-4.57468331e-02 -2.55342275e-01 4.21008766e-01 7.76084721e-01
9.41752940e-02 1.02652594e-01 1.41693115e-01 -2.39471465e-01
5.44278443e-01 1.78301945e-01 -2.62576481e-03 -1.72799563e+00
4.77153957e-02 -9.29110289e-01 -8.98222029e-01 1.15314864e-01
3.11622489e-02 -1.05019414e+00 -3.55070144e-01 -1.21674502e+00
-3.23178828e-01 -4.44115728e-01 -8.22330773e-01 2.49677554e-01
3.92078757e-01 2.08078772e-01 -2.38738637e-02 3.54865253e-01
2.27984469e-02 3.81647468e-01 8.97986352e-01 1.51112616e-01
-5.04962802e-01 -9.34724286e-02 -1.64662227e-01 5.54838896e-01
8.26017141e-01 -5.72199941e-01 -2.22182348e-01 2.99756587e-01
6.10729098e-01 2.22747073e-01 6.01292372e-01 -1.31646585e+00
1.30928345e-02 1.04751647e-01 4.70915228e-01 -4.54323739e-01
-3.70003516e-03 -8.62840474e-01 6.58186853e-01 1.05539572e+00
-5.97633839e-01 4.43816096e-01 2.30286255e-01 4.28441733e-01
-3.53942126e-01 -1.47929303e-02 7.24297464e-01 3.04523230e-01
-2.96010017e-01 -1.12937585e-01 -3.74850661e-01 -3.64648312e-01
7.51904309e-01 -2.80086428e-01 -2.81807423e-01 -1.31940097e-01
-9.49139416e-01 -2.68572658e-01 4.86845583e-01 2.17969209e-01
5.48294306e-01 -1.35724902e+00 -4.37448651e-01 5.39325833e-01
8.71521160e-02 -9.65816736e-01 1.76170468e-01 8.78494263e-01
-9.04650033e-01 2.46288106e-01 -1.01947117e+00 -7.34965444e-01
-9.63061154e-01 4.57960755e-01 8.48844826e-01 -2.26087943e-01
-6.16955817e-01 -6.40441626e-02 -4.37063038e-01 -1.71116084e-01
-2.29890987e-01 -7.42853940e-01 -8.04928541e-01 6.60747409e-01
3.68150383e-01 2.88634568e-01 1.37684792e-01 -6.60068035e-01
-7.32134461e-01 9.28275049e-01 7.91672528e-01 -3.00460577e-01
1.58443534e+00 2.03369662e-01 -6.01379573e-01 9.02433217e-01
1.26734769e+00 -1.30975544e-01 -8.88306737e-01 2.98546612e-01
3.44948173e-01 -1.41939536e-01 -1.23349860e-01 -6.47488534e-01
-6.84641302e-01 8.75235319e-01 9.86367583e-01 9.94068980e-01
1.43073773e+00 -3.78195286e-01 5.28444201e-02 3.96446764e-01
3.50793689e-01 -7.91810453e-01 -1.24012753e-01 3.14369977e-01
9.91047800e-01 -3.94316405e-01 -1.28340796e-01 4.62898836e-02
4.29907776e-02 1.81563389e+00 -2.10066035e-01 -4.76079702e-01
9.62293446e-01 2.65784729e-02 2.46332474e-02 -5.77283382e-01
-7.74405837e-01 -1.76423356e-01 6.56582117e-01 3.03772062e-01
2.03159794e-01 2.79125214e-01 -7.74898708e-01 2.90599108e-01
-2.55793899e-01 -6.15092181e-02 4.22540337e-01 5.81037998e-01
-5.30827641e-01 -1.12083662e+00 -2.32780948e-01 3.27609241e-01
-3.66722196e-02 2.23618910e-01 -6.03263617e-01 6.43723607e-01
1.11828623e-02 2.10745826e-01 1.17956892e-01 -6.38990343e-01
5.89840710e-01 4.33078617e-01 3.78259391e-01 -7.43807778e-02
-7.54804730e-01 -2.41744354e-01 -9.30349976e-02 -6.15317166e-01
-3.42322022e-01 -3.88228834e-01 -9.71944988e-01 7.54656941e-02
-2.64526950e-03 2.68184572e-01 9.61145759e-01 6.38033450e-01
4.40376014e-01 7.96007335e-01 7.37770975e-01 -8.63127470e-01
-1.21706083e-01 -7.14597285e-01 -5.50938725e-01 4.08392102e-01
6.41408443e-01 -5.25883198e-01 -3.10541272e-01 -2.31592625e-01] | [7.501455307006836, 3.741959571838379] |
bc6ba7dd-bad1-423c-8f6f-1bf135a55ceb | reconstruction-attack-on-instance-encoding | null | null | https://aclanthology.org/2021.emnlp-main.154 | https://aclanthology.org/2021.emnlp-main.154.pdf | Reconstruction Attack on Instance Encoding for Language Understanding | A private learning scheme TextHide was recently proposed to protect the private text data during the training phase via so-called instance encoding. We propose a novel reconstruction attack to break TextHide by recovering the private training data, and thus unveil the privacy risks of instance encoding. We have experimentally validated the effectiveness of the reconstruction attack with two commonly-used datasets for sentence classification. Our attack would advance the development of privacy preserving machine learning in the context of natural language processing. | ['Yuan Hong', 'Shangyu Xie'] | null | null | null | null | emnlp-2021-11 | ['sentence-classification'] | ['natural-language-processing'] | [ 6.56090796e-01 4.35498178e-01 -1.47111028e-01 -6.09182537e-01
-9.09990430e-01 -9.21497941e-01 3.76343489e-01 4.43408370e-01
-5.76919854e-01 7.07059920e-01 1.77859098e-01 -6.71024561e-01
1.50879756e-01 -1.06503379e+00 -8.94747615e-01 -9.71582115e-01
-1.16651721e-01 -7.75382593e-02 -1.50260299e-01 1.89474970e-02
3.39167923e-01 3.76834244e-01 -1.01796424e+00 7.28024065e-01
4.81045514e-01 8.56211901e-01 -5.15010536e-01 5.81164837e-01
1.32975981e-01 4.82041419e-01 -6.80239916e-01 -1.06140506e+00
7.59629250e-01 -3.22055161e-01 -9.97887433e-01 -2.94697940e-01
1.54706120e-01 -3.35073501e-01 -5.96086979e-01 1.34510720e+00
3.97880942e-01 -2.44313762e-01 7.78303146e-02 -1.19128788e+00
-4.05282170e-01 9.44462299e-01 -2.60553420e-01 -6.43994510e-02
3.26819658e-01 1.22163825e-01 8.71638060e-01 -1.13814235e-01
5.91143370e-01 7.37342358e-01 5.08664846e-01 7.87742913e-01
-1.05147791e+00 -9.66328084e-01 -4.59286511e-01 1.26130387e-01
-1.55552959e+00 -4.73545462e-01 7.66607165e-01 -6.10771999e-02
6.86081409e-01 7.67261505e-01 1.61836803e-01 1.17790508e+00
6.03004515e-01 9.86659586e-01 1.23814940e+00 -5.14441967e-01
3.29891682e-01 4.97312814e-01 1.60528988e-01 7.14877367e-01
6.92664862e-01 1.69011816e-01 -4.12248582e-01 -9.45903063e-01
8.87187049e-02 -7.15496317e-02 -6.20654583e-01 -3.69980603e-01
-7.45836854e-01 1.03443158e+00 -2.15424057e-02 1.69131458e-01
2.58314997e-01 -7.78975785e-02 1.09643245e+00 7.54881740e-01
6.39604270e-01 3.47160041e-01 -7.09879220e-01 2.35207558e-01
-3.62613380e-01 1.03301995e-01 1.26087749e+00 7.48636425e-01
5.03782094e-01 -5.46522200e-01 6.38896078e-02 1.92444593e-01
6.86837733e-02 3.05034608e-01 5.74487805e-01 -4.88250494e-01
9.77898657e-01 2.14551434e-01 -3.25004607e-02 -1.14341211e+00
1.53228208e-01 -1.94263205e-01 -8.47474217e-01 -3.43895346e-01
3.62312943e-01 -3.80895585e-01 -1.04673505e-01 1.58815873e+00
3.11334580e-01 8.98856148e-02 7.02018678e-01 4.92397219e-01
1.42165810e-01 6.30482435e-01 -4.76681553e-02 -1.55750915e-01
1.26327920e+00 -5.28719902e-01 -7.80315816e-01 5.85671701e-03
1.23026037e+00 -1.83136299e-01 7.44500816e-01 5.87922215e-01
-7.45217919e-01 1.56466737e-01 -1.36384463e+00 -2.06844151e-01
-3.94887060e-01 -1.59753248e-01 9.57475185e-01 1.47260082e+00
-4.65446234e-01 5.53816259e-01 -7.20191956e-01 1.28726721e-01
7.17324257e-01 6.90556347e-01 -1.04702902e+00 -3.00203096e-02
-1.61103702e+00 2.93436080e-01 7.22953856e-01 -4.08988930e-02
-6.39917672e-01 -3.77733648e-01 -9.90789592e-01 1.64414927e-01
2.58074671e-01 -3.57291251e-01 9.73248601e-01 -6.73184633e-01
-1.39273477e+00 1.34529960e+00 3.44706386e-01 -1.21849144e+00
5.39296031e-01 7.64065981e-02 -5.64595222e-01 2.99001008e-01
-3.58186841e-01 -2.04875350e-01 1.00319922e+00 -9.74257886e-01
-3.83797705e-01 -9.24329579e-01 -1.03411712e-01 -2.79398471e-01
-7.05035925e-01 -1.07577793e-01 3.90623480e-01 -7.04284072e-01
-1.64360866e-01 -7.41239727e-01 -3.18118423e-01 -2.00476095e-01
-4.70257819e-01 1.25457868e-01 1.31648874e+00 -5.71721911e-01
1.06885314e+00 -2.55067134e+00 -3.18952411e-01 3.54972959e-01
1.02237705e-02 7.20653117e-01 5.92581593e-02 6.39778197e-01
-5.65338619e-02 4.48626220e-01 -5.12538671e-01 -5.81345201e-01
-8.90738741e-02 3.91774714e-01 -9.89502728e-01 1.09411275e+00
-2.41996840e-01 7.47630119e-01 -6.47308290e-01 -2.26518184e-01
-7.14624524e-02 3.73267084e-01 -5.59066832e-01 2.93698728e-01
-3.20354968e-01 1.77789748e-01 -7.38893569e-01 8.28921702e-03
1.12534356e+00 -2.99581997e-02 4.73167330e-01 3.47704083e-01
4.22724664e-01 5.83434343e-01 -8.92750025e-01 1.60353088e+00
-1.46684811e-01 4.86126184e-01 1.27450094e-01 -7.31762528e-01
8.26122224e-01 6.11714542e-01 1.56509668e-01 -2.15381980e-01
5.13080895e-01 -8.89235288e-02 -5.25899410e-01 -5.63632250e-01
4.48057771e-01 -1.39515936e-01 -5.62803566e-01 9.11040783e-01
-1.79231539e-01 2.00405017e-01 -9.08683479e-01 1.84122637e-01
9.28897917e-01 -3.60416561e-01 2.74720252e-01 -1.71035618e-01
8.14881623e-01 -5.39050639e-01 5.01740336e-01 6.93014324e-01
-3.40725154e-01 1.61917537e-01 7.54344404e-01 -6.31129622e-01
-6.77964091e-01 -3.99958700e-01 -1.55321866e-01 6.74173057e-01
-1.54758468e-01 -6.19380474e-01 -1.04691482e+00 -1.69737363e+00
2.77352989e-01 5.66825688e-01 -5.28677166e-01 -5.70073605e-01
-4.07008469e-01 -3.97934437e-01 1.29135072e+00 2.09212061e-02
8.15119147e-01 -5.45877337e-01 -3.71905178e-01 -2.56840110e-01
-9.82369184e-02 -1.33993781e+00 -4.02328402e-01 1.89447775e-01
-8.34526360e-01 -1.15079260e+00 4.88554500e-02 -5.17789662e-01
7.45344341e-01 1.63764760e-01 3.01577419e-01 3.16507429e-01
-9.49980468e-02 1.05995722e-01 -2.08163127e-01 -4.61931258e-01
-8.93448830e-01 2.83610702e-01 -1.45518631e-01 4.59438205e-01
6.72425926e-01 -2.68143088e-01 -4.35906619e-01 -5.77674024e-02
-1.29714608e+00 -2.36858875e-01 -1.01963259e-01 8.35936427e-01
2.23594278e-01 5.13666093e-01 3.48645359e-01 -1.71836710e+00
5.05127549e-01 -4.36275870e-01 -7.52952278e-01 3.88548255e-01
-4.40855801e-01 3.33369702e-01 9.33847904e-01 -2.38718927e-01
-9.46780920e-01 1.46213144e-01 -5.20976722e-01 5.55366138e-03
-2.63682336e-01 2.23148778e-01 -6.99590385e-01 -5.00757635e-01
3.54916632e-01 3.07148427e-01 1.74333841e-01 -5.70420504e-01
2.06529066e-01 1.27288580e+00 3.78104001e-01 -5.80919087e-01
6.20264232e-01 6.92022443e-01 2.06960127e-01 -4.69638109e-01
-6.72391236e-01 -8.69778022e-02 -2.12956175e-01 6.16534770e-01
5.15468657e-01 -7.27021277e-01 -1.02247226e+00 8.40150118e-01
-9.48246539e-01 4.84901741e-02 -1.96734309e-01 2.42819801e-01
-6.10199451e-01 1.07318532e+00 -7.08119869e-01 -7.49766350e-01
-6.90672755e-01 -9.10033226e-01 8.20416570e-01 -2.79135793e-01
2.83479482e-01 -1.02566183e+00 -2.74523273e-02 6.82415485e-01
-1.34808302e-01 3.62897784e-01 9.63726640e-01 -1.26945436e+00
-3.70353997e-01 -1.00507903e+00 2.83084214e-01 6.17902458e-01
-1.03651799e-01 -4.31096822e-01 -1.29397929e+00 -6.28549755e-01
9.10394073e-01 -3.42276514e-01 7.39163458e-01 -4.89062786e-01
1.65969348e+00 -1.00166941e+00 -1.79478660e-01 9.42311168e-01
1.47294199e+00 -3.84246483e-02 9.71982837e-01 1.30560815e-01
4.64322120e-01 7.31617987e-01 3.66659373e-01 7.35915542e-01
5.64638823e-02 3.81190866e-01 3.01534653e-01 3.29739720e-01
9.12163556e-01 -6.92899644e-01 3.04545224e-01 3.31861287e-01
6.86916113e-01 -5.62212467e-01 -5.20692766e-01 2.13471517e-01
-1.67214000e+00 -9.83788490e-01 6.95485482e-03 2.25369835e+00
1.05498099e+00 8.46147537e-04 -3.65705073e-01 5.37984967e-01
4.29036885e-01 3.42070639e-01 1.03111818e-01 -9.64439631e-01
-7.57128298e-02 3.93544734e-01 1.07195497e+00 4.85956162e-01
-1.36662912e+00 1.05395615e+00 6.18542147e+00 7.06099510e-01
-1.23570406e+00 2.26679474e-01 6.53076768e-01 3.13642949e-01
-3.94982100e-01 1.06016412e-01 -8.39031518e-01 4.17318344e-01
1.08346009e+00 -5.09762108e-01 2.35587254e-01 9.74815965e-01
-4.38415051e-01 3.08338761e-01 -1.00321877e+00 6.27321184e-01
1.80935282e-02 -1.34090781e+00 -1.34241611e-01 2.08392948e-01
2.42294550e-01 -4.31867570e-01 1.55161649e-01 5.88564835e-02
1.17189117e-01 -7.68404305e-01 2.53676951e-01 -1.32180870e-01
7.95587480e-01 -1.15703666e+00 6.97068810e-01 7.43898809e-01
-4.01496172e-01 -2.86492050e-01 -5.77385128e-01 -6.23788536e-02
-2.48528734e-01 3.70546252e-01 -9.61946011e-01 6.81928456e-01
2.21947536e-01 4.02440041e-01 -4.38499302e-01 4.34842855e-01
-4.59100366e-01 8.05730700e-01 -1.99903488e-01 2.97260769e-02
2.00815469e-01 2.06008926e-02 5.65072417e-01 1.19949329e+00
-1.29161850e-01 3.14713180e-01 -1.02622710e-01 4.69227344e-01
-4.75251108e-01 1.13436736e-01 -1.14697981e+00 -2.53072500e-01
2.74250448e-01 6.61035478e-01 -2.04079688e-01 3.81169207e-02
-5.81358597e-02 1.34023702e+00 8.37894455e-02 -5.21583073e-02
-3.26766193e-01 -5.72707176e-01 7.54809141e-01 -1.77535936e-01
3.08756500e-01 -3.44109684e-01 -4.39978749e-01 -1.49516261e+00
1.21914074e-01 -9.33186769e-01 8.52477968e-01 8.67538229e-02
-1.25252903e+00 3.73435557e-01 -2.38217115e-01 -9.05217528e-01
-9.16789323e-02 -3.68098199e-01 -3.96686792e-01 7.28145838e-01
-1.40752125e+00 -1.03177261e+00 6.13972545e-01 1.04748845e+00
-1.84134081e-01 -3.90976191e-01 1.45052803e+00 5.33628603e-03
-5.52865863e-01 1.41935134e+00 2.83480436e-01 3.28021765e-01
5.18859863e-01 -7.04304218e-01 7.70052671e-01 9.42188680e-01
2.99675316e-01 8.75109017e-01 6.20890796e-01 -6.59004807e-01
-1.77941549e+00 -9.92654443e-01 1.32200408e+00 -2.76922464e-01
2.78890371e-01 -1.09655476e+00 -1.05347431e+00 9.14272726e-01
2.20256731e-01 2.07995802e-01 1.30934310e+00 -2.93414652e-01
-9.44948792e-01 -1.37286022e-01 -2.02756476e+00 1.75990984e-01
3.71865362e-01 -1.21763873e+00 -5.32388210e-01 4.01165396e-01
1.23100972e+00 -3.74654263e-01 -6.85093701e-01 1.31620660e-01
5.02708912e-01 -6.95874870e-01 7.42495298e-01 -1.00001025e+00
-3.05926856e-02 2.04211056e-01 -3.27082396e-01 -7.23373890e-01
3.70462656e-01 -1.21035051e+00 -3.30881268e-01 1.20145822e+00
2.43651748e-01 -1.16661155e+00 1.22703159e+00 8.05335462e-01
6.08951449e-01 -2.38165557e-01 -1.35782635e+00 -5.47348380e-01
2.22456366e-01 -3.62011999e-01 1.04034472e+00 1.32461047e+00
3.00307631e-01 -2.70550787e-01 -6.82068169e-01 4.98579532e-01
7.99693227e-01 -2.38669768e-01 8.02437305e-01 -8.10000300e-01
-5.87663114e-01 4.54904675e-01 -6.29757106e-01 -6.77473545e-01
7.21726418e-01 -1.05891728e+00 -3.74190241e-01 -4.24145669e-01
-1.42721890e-03 -3.84771466e-01 -3.25600356e-01 5.81852555e-01
-5.20439744e-02 -2.72903349e-02 -3.84622253e-02 -6.72954023e-02
-9.43936259e-02 4.93103564e-01 7.95577466e-01 -5.76438755e-02
1.78866878e-01 4.02583301e-01 -1.01016331e+00 1.98458686e-01
1.01146460e+00 -1.04772973e+00 -5.35168886e-01 -3.04919392e-01
2.90858686e-01 1.06625758e-01 3.15415829e-01 -6.78859293e-01
1.38072565e-01 -8.77799913e-02 -2.28270292e-01 -2.41334215e-01
-1.82105750e-01 -1.41371465e+00 1.64225280e-01 6.46408319e-01
-7.76514649e-01 -2.15052351e-01 1.36158839e-01 6.77623332e-01
-4.98269722e-02 -2.89453655e-01 6.93890929e-01 8.66303518e-02
-2.83037782e-01 2.74426162e-01 -1.11315951e-01 -2.49600913e-02
1.28080857e+00 7.83673897e-02 -4.91692990e-01 -1.66097879e-01
-4.23852533e-01 1.01809077e-01 6.92857683e-01 2.09253624e-01
7.16521919e-01 -7.11062610e-01 -3.94101232e-01 9.30453539e-01
9.08869952e-02 -5.34079552e-01 2.82868922e-01 1.52995378e-01
-4.35681552e-01 3.62564474e-01 -1.17872491e-01 2.82776207e-02
-1.77039778e+00 1.04792988e+00 2.14046881e-01 -3.35333645e-01
-9.27684784e-01 8.38883519e-01 -1.90875515e-01 -2.49067262e-01
6.57808900e-01 9.40385647e-03 2.84857363e-01 -4.35483903e-01
9.29115891e-01 4.18797322e-02 3.57661307e-01 -2.69016802e-01
-1.32475048e-01 -7.92009681e-02 -3.03641766e-01 -1.18349098e-01
1.11420023e+00 -2.37615988e-01 -4.77085531e-01 -2.31988579e-01
1.92163551e+00 2.63980955e-01 -7.16473222e-01 -2.81747937e-01
9.14920121e-02 -1.00177824e+00 3.51999234e-03 -3.73540759e-01
-9.15972948e-01 1.05069184e+00 5.11237025e-01 3.29217017e-01
1.14363778e+00 -5.11297047e-01 1.43420255e+00 5.66852212e-01
9.71712530e-01 -7.02643812e-01 -6.52452946e-01 3.25214177e-01
3.38088214e-01 -9.22655821e-01 1.44191179e-03 -5.01528442e-01
-7.37824559e-01 8.14357758e-01 5.96699901e-02 7.25106522e-02
7.78563023e-01 5.76710582e-01 2.89869793e-02 1.11133099e-01
-8.52130055e-01 7.12190211e-01 -2.60006815e-01 4.84844685e-01
-1.52564541e-01 1.21611558e-01 -6.28223181e-01 1.01032770e+00
4.07023691e-02 1.49141818e-01 4.77778912e-01 1.37031615e+00
-5.64170778e-02 -1.84689629e+00 -1.32620871e-01 -1.19651690e-01
-1.24830008e+00 2.09164806e-02 -7.28808343e-01 2.63504952e-01
-9.34881270e-02 7.80643404e-01 -4.19164300e-01 -6.90761983e-01
-4.33097314e-03 2.18942344e-01 2.42424324e-01 -5.47952890e-01
-1.23830020e+00 -5.82197070e-01 1.90304250e-01 -3.68331432e-01
-4.20508683e-02 -3.78568530e-01 -1.16990316e+00 -7.19741642e-01
-4.52706397e-01 6.01325393e-01 7.88601696e-01 5.52854478e-01
6.46434486e-01 -4.79066700e-01 1.37479544e+00 1.23068832e-01
-1.09348667e+00 -1.24832287e-01 -7.40153730e-01 4.42315757e-01
6.08934700e-01 2.39659563e-01 -5.60591400e-01 -2.19129682e-01] | [5.892279624938965, 7.020198822021484] |
5b8de74a-9c6b-4d34-81c5-0d1e7628d5d6 | security-vulnerability-detection-using-deep | 2105.02388 | null | https://arxiv.org/abs/2105.02388v1 | https://arxiv.org/pdf/2105.02388v1.pdf | Security Vulnerability Detection Using Deep Learning Natural Language Processing | Detecting security vulnerabilities in software before they are exploited has been a challenging problem for decades. Traditional code analysis methods have been proposed, but are often ineffective and inefficient. In this work, we model software vulnerability detection as a natural language processing (NLP) problem with source code treated as texts, and address the automated software venerability detection with recent advanced deep learning NLP models assisted by transfer learning on written English. For training and testing, we have preprocessed the NIST NVD/SARD databases and built a dataset of over 100,000 files in $C$ programming language with 123 types of vulnerabilities. The extensive experiments generate the best performance of over 93\% accuracy in detecting security vulnerabilities. | ['Shaoen Wu', 'Noah Ziems'] | 2021-05-06 | null | null | null | null | ['vulnerability-detection'] | ['miscellaneous'] | [-1.44701466e-01 -2.28601560e-01 -2.95395821e-01 -2.66991198e-01
-9.13716018e-01 -9.24699068e-01 1.71261281e-01 4.33977723e-01
-1.64524376e-01 1.88505232e-01 6.18578047e-02 -1.21665108e+00
2.31568813e-01 -8.91129792e-01 -6.80230677e-01 7.16268495e-02
-6.35393620e-01 -3.43910784e-01 2.50246078e-01 -1.43291742e-01
8.77590835e-01 7.61116445e-02 -1.13017166e+00 5.76266170e-01
8.71334195e-01 6.41587973e-01 -3.02258760e-01 8.62831473e-01
-6.08088791e-01 1.08895445e+00 -9.52975392e-01 -8.05100381e-01
3.25013757e-01 4.81819570e-01 -1.15685320e+00 -6.26032174e-01
3.06437403e-01 -2.71650463e-01 -2.51461148e-01 1.81757760e+00
1.61312640e-01 -7.42699921e-01 3.14267576e-01 -1.32557583e+00
-9.97912288e-01 1.05861151e+00 -1.07541251e+00 4.71028447e-01
7.19280005e-01 1.27068102e-01 1.03561461e+00 -7.87060022e-01
3.38638902e-01 1.16836679e+00 8.56375396e-01 6.55831099e-01
-9.61002350e-01 -8.81907284e-01 -1.57287702e-01 3.11714020e-02
-1.31217468e+00 3.15637589e-02 9.60613012e-01 -9.79987264e-01
1.84436393e+00 -4.27743167e-01 -6.74794242e-02 1.10271621e+00
7.55444765e-01 4.58831519e-01 6.27676964e-01 -7.64638245e-01
-1.07109351e-02 1.89173967e-01 1.01643109e+00 1.00619340e+00
3.65438312e-01 1.24830060e-01 7.29859294e-03 -7.92425156e-01
-2.79546273e-03 -1.09784551e-01 8.56292024e-02 1.54636636e-01
-6.31871104e-01 1.25135756e+00 1.55971944e-01 4.11937445e-01
5.29428385e-02 -4.61468436e-02 1.14720583e+00 8.55892658e-01
2.42799178e-01 4.88972455e-01 -8.49967599e-01 -3.18715096e-01
-6.23111904e-01 1.42724067e-01 8.65742445e-01 8.75612020e-01
6.35580182e-01 3.51859987e-01 5.36016524e-01 5.65272570e-01
7.35250354e-01 4.55689073e-01 5.44946969e-01 -3.40577066e-01
9.94641066e-01 1.10051596e+00 -3.57213467e-01 -1.44554234e+00
-1.29701287e-01 2.73575932e-02 -4.70637947e-01 6.25180006e-01
9.47350487e-02 -3.71661305e-01 -6.66332960e-01 1.42773688e+00
-2.70307541e-01 -2.62911707e-01 3.10701221e-01 1.50724631e-02
6.69536054e-01 7.87299991e-01 3.46559942e-01 1.21941924e-01
1.30589676e+00 -4.87651408e-01 -3.64400327e-01 -1.54944956e-01
9.24044728e-01 -8.02943647e-01 1.21331131e+00 9.16153789e-01
-5.30442774e-01 -4.47666049e-01 -1.43070471e+00 9.26250443e-02
-8.93684804e-01 -8.00084472e-02 4.79645103e-01 1.54896951e+00
-1.12308013e+00 3.26866567e-01 -5.90854883e-01 7.60357529e-02
4.45607066e-01 3.81626934e-01 -3.31900001e-01 1.83807060e-01
-1.19497406e+00 6.08875573e-01 6.47879958e-01 -4.01277751e-01
-1.23704207e+00 -5.83136141e-01 -9.21422303e-01 1.37457818e-01
3.85062218e-01 3.57359797e-01 1.23827255e+00 -6.19225800e-01
-9.82838929e-01 9.82881546e-01 2.63495594e-01 -5.71374178e-01
-1.81386814e-01 -3.32855344e-01 -8.35911989e-01 -2.00357363e-01
-1.89810812e-01 -2.62275159e-01 7.66834974e-01 -8.43235016e-01
-4.24649596e-01 -1.92125052e-01 3.10892045e-01 -9.02677417e-01
-9.49322820e-01 9.64290202e-01 1.44114658e-01 -5.55663824e-01
-5.60935676e-01 -4.51416433e-01 -1.07767992e-01 -6.95636332e-01
-4.77106899e-01 -4.19542432e-01 8.16963732e-01 -1.22527564e+00
1.73692155e+00 -2.16798472e+00 -3.24675143e-01 3.79739076e-01
4.38518316e-01 7.46814311e-01 -4.76536423e-01 3.44671458e-01
-4.96257901e-01 7.61895478e-01 -4.42356348e-01 1.61796302e-01
1.91056833e-01 -4.65613097e-01 -8.86148512e-01 1.80685967e-01
2.81641275e-01 8.40453029e-01 -6.16294622e-01 -5.77859998e-01
-2.22194508e-01 1.33575991e-01 -5.86351752e-01 2.16218695e-01
-3.39207858e-01 -4.65612769e-01 -7.06568956e-01 1.11967516e+00
7.32421637e-01 1.40845597e-01 -1.09568670e-01 4.87171054e-01
-1.37035936e-01 2.18539774e-01 -8.93576920e-01 1.39510441e+00
-4.35684532e-01 7.47027695e-01 -8.71913657e-02 -8.53078723e-01
1.23410404e+00 2.78827667e-01 1.63806399e-04 -5.72914004e-01
1.44562855e-01 2.45936155e-01 2.00151503e-01 -1.05974102e+00
3.55579585e-01 1.95372909e-01 -7.13132679e-01 7.80289114e-01
2.29495037e-02 8.27568620e-02 -1.05463833e-01 3.52968574e-01
1.68640673e+00 -2.32603014e-01 2.44600832e-01 -3.78049165e-01
1.04127967e+00 1.86034605e-01 5.76065302e-01 5.30738294e-01
-3.13210070e-01 -8.70300233e-02 1.12112856e+00 -8.06204259e-01
-1.16854906e+00 -8.90301943e-01 7.44395778e-02 9.24277425e-01
-5.00754416e-01 -6.41724169e-01 -9.66492951e-01 -1.20373738e+00
-1.94162101e-01 5.77158570e-01 -4.44098413e-01 -3.40738773e-01
-7.81114161e-01 -7.89296448e-01 1.33347273e+00 6.76758230e-01
3.54037344e-01 -1.40729928e+00 -3.79163057e-01 1.01286396e-01
2.28942633e-01 -8.60274911e-01 -2.65933514e-01 1.80406749e-01
-4.95752603e-01 -1.42218339e+00 -9.46962386e-02 -1.21855271e+00
4.17733282e-01 -2.32448995e-01 1.23623526e+00 3.92579198e-01
-7.05712616e-01 -1.07552968e-01 -5.71568727e-01 -5.11335492e-01
-9.82309580e-01 1.24153867e-01 2.07163021e-02 -6.38370872e-01
9.76969302e-01 -2.00965732e-01 1.42980635e-01 -1.39597505e-01
-8.67753744e-01 -1.06875050e+00 5.11796415e-01 5.55801332e-01
-1.40948221e-01 6.34014547e-01 6.97074175e-01 -1.01449800e+00
1.15621352e+00 -6.33917153e-01 -1.12006807e+00 3.23018968e-01
-5.26163638e-01 2.16366827e-01 1.04081106e+00 -2.35789195e-01
-1.12754476e+00 4.69627837e-03 -6.76669598e-01 -1.99115485e-01
-2.26805627e-01 9.88030553e-01 -4.70594078e-01 -4.91589874e-01
9.83133614e-01 1.01465724e-01 -3.35424513e-01 -4.54962403e-01
6.10012934e-02 8.46951365e-01 3.78827095e-01 -9.11693990e-01
1.24586058e+00 -3.65956664e-01 -6.91409528e-01 -7.44017839e-01
-1.39411747e-01 -1.37827992e-01 -4.47529048e-01 1.85821787e-01
8.52202296e-01 -6.10934556e-01 -6.35861516e-01 8.30432713e-01
-1.33431745e+00 -3.24387103e-02 4.05572742e-01 -4.10594270e-02
2.25784164e-02 1.06612909e+00 -7.86903679e-01 -7.20473468e-01
-6.64461434e-01 -1.43392265e+00 5.06316900e-01 -4.86727394e-02
-3.29932421e-02 -7.57905722e-01 6.21819258e-01 8.03783834e-02
4.40368742e-01 2.56370485e-01 1.78376794e+00 -8.60372007e-01
-2.24528238e-01 -6.32067442e-01 -2.44613826e-01 6.11660957e-01
2.41615940e-02 7.12303579e-01 -8.97358358e-01 -2.61363775e-01
1.83999643e-01 -5.99356532e-01 5.15221596e-01 -1.66532710e-01
1.44315040e+00 -2.58691907e-01 -2.22486332e-01 3.31355035e-01
1.74840689e+00 6.38971090e-01 7.33626664e-01 4.87491667e-01
5.64716458e-01 6.64469957e-01 2.43702754e-01 2.13799521e-01
3.39778751e-04 -1.24258012e-01 5.81325591e-01 5.18897414e-01
5.79907656e-01 -1.71851903e-01 6.74392402e-01 8.04132760e-01
2.66453147e-01 -1.47418857e-01 -1.78944755e+00 5.52675962e-01
-1.18855155e+00 -7.45788991e-01 -8.68275464e-02 1.90513778e+00
7.97550619e-01 5.57783723e-01 2.27975426e-03 3.43915880e-01
5.57267070e-01 2.34098226e-01 -3.13436866e-01 -8.76204908e-01
6.70566931e-02 3.94551426e-01 3.26141536e-01 4.51825410e-01
-1.31864440e+00 9.64642227e-01 6.63961363e+00 7.19512641e-01
-9.44701910e-01 8.63360167e-02 4.92977142e-01 5.42429447e-01
-3.55705202e-01 -5.72866276e-02 -7.30457723e-01 3.15626800e-01
1.41439021e+00 -4.48883891e-01 1.98681891e-01 1.30786729e+00
-3.84425044e-01 2.92520165e-01 -9.64067578e-01 6.93289220e-01
1.39140338e-01 -1.05624294e+00 -1.27000064e-01 -1.48477346e-01
4.47440475e-01 1.28950566e-01 2.83894032e-01 8.69652689e-01
5.75247407e-01 -1.25113213e+00 3.23111385e-01 4.97606806e-02
5.57084084e-01 -1.05390954e+00 1.00408316e+00 4.74514842e-01
-1.33862352e+00 -7.05950022e-01 -5.26318789e-01 5.42582832e-02
-3.79516542e-01 3.22395533e-01 -6.29381478e-01 2.74531990e-01
1.08306742e+00 5.27197659e-01 -9.52738106e-01 7.13461876e-01
-2.43274271e-01 6.89874887e-01 1.54964983e-01 -3.31783175e-01
2.19963849e-01 3.12423587e-01 1.91501111e-01 1.28085792e+00
1.11511484e-01 -1.47207290e-01 6.64240122e-02 1.10719919e+00
-2.64122128e-01 9.35411081e-02 -8.65297079e-01 -6.19525254e-01
2.83927381e-01 9.45614100e-01 -5.01087189e-01 -1.24388590e-01
-9.35670197e-01 1.93942711e-02 3.43273371e-01 1.17229395e-01
-8.41633022e-01 -1.07436264e+00 4.91615176e-01 -1.89928547e-01
-1.91514529e-02 -3.51879478e-01 -3.28376919e-01 -1.31328523e+00
3.67062718e-01 -1.39825356e+00 5.36060095e-01 -2.56862998e-01
-1.53188479e+00 1.05052984e+00 -1.71899453e-01 -1.19680631e+00
-1.41029560e-03 -1.03571439e+00 -1.18809605e+00 9.99967337e-01
-1.35366082e+00 -8.85390162e-01 2.23790541e-01 5.25816977e-01
5.79191744e-01 -9.92097020e-01 7.67071962e-01 5.34663320e-01
-6.34456098e-01 1.02881014e+00 -1.52953729e-01 9.02555585e-01
2.01632783e-01 -1.20520639e+00 1.10460329e+00 1.37967157e+00
-1.67823255e-01 1.18339753e+00 3.61473471e-01 -1.04326904e+00
-1.48424637e+00 -1.07162285e+00 7.48411775e-01 -6.32678270e-01
1.24345887e+00 -5.81815004e-01 -1.33195770e+00 6.27899885e-01
4.19498533e-01 -2.91708112e-01 7.48111665e-01 3.19596976e-02
-1.03302252e+00 3.85953337e-01 -1.32955980e+00 2.17696443e-01
4.65856552e-01 -1.14496315e+00 -1.05160511e+00 1.65875137e-01
1.12102199e+00 7.28992373e-02 -8.54637563e-01 1.77730635e-01
2.01909244e-01 -5.47580421e-01 9.59493876e-01 -8.73560965e-01
6.94598496e-01 -2.46286184e-01 -2.65711427e-01 -7.00088024e-01
-9.15647745e-02 -4.15132076e-01 -1.22410059e-01 1.48444104e+00
7.69145131e-01 -5.66677332e-01 8.37246120e-01 6.11962855e-01
3.89555134e-02 -2.91751444e-01 -7.64046133e-01 -6.67582929e-01
8.14482927e-01 -7.51232803e-01 7.19245315e-01 1.21670783e+00
5.26343286e-01 -1.89939126e-01 -6.25011399e-02 6.17551565e-01
7.03892291e-01 -1.44687101e-01 4.11244303e-01 -1.18885231e+00
-2.70905375e-01 -5.19746661e-01 -4.73916799e-01 -1.70217842e-01
7.86026716e-01 -7.11095572e-01 -5.95395453e-03 -6.97815239e-01
3.45027804e-01 1.14941388e-01 -2.93921113e-01 9.29268420e-01
-7.86934793e-02 -2.88963109e-01 -3.18184197e-01 -1.04475789e-01
-1.88411817e-01 9.67370868e-02 -8.14610869e-02 -7.18291044e-01
-1.28987189e-02 -6.56615347e-02 -5.50123453e-01 1.02296972e+00
8.61592531e-01 -7.17007399e-01 -3.37211043e-01 -6.42389357e-01
7.01833725e-01 1.26866132e-01 7.22795399e-03 -9.27993894e-01
6.32670447e-02 7.80514181e-02 -1.15379483e-01 -5.33353209e-01
-5.78232825e-01 -5.82212687e-01 -5.25885105e-01 8.45971525e-01
-5.12492657e-01 5.36565661e-01 6.15578115e-01 3.79480511e-01
-3.27484876e-01 -8.38405073e-01 7.01262236e-01 -2.02281311e-01
-1.06884050e+00 3.38552892e-01 -7.08315849e-01 1.25459164e-01
1.04374170e+00 1.18498698e-01 -7.39585340e-01 4.56887037e-01
-2.36058772e-01 1.17621690e-01 3.98188621e-01 9.14635777e-01
9.61467326e-01 -7.52713084e-01 -5.57721436e-01 4.31309611e-01
4.11337614e-01 -4.73670125e-01 1.98449805e-01 -1.35888383e-01
-8.78041983e-01 3.56531799e-01 -5.05915105e-01 -7.64312819e-02
-1.58187473e+00 1.22319198e+00 2.19178334e-01 -3.04939538e-01
-4.68187660e-01 7.55948365e-01 -3.13325673e-01 -8.10731113e-01
4.15996939e-01 -3.01417232e-01 -6.37457788e-01 -4.15067047e-01
9.14877176e-01 1.36516571e-01 1.02544695e-01 -3.55717540e-01
-5.85540295e-01 6.95994437e-01 -4.99117106e-01 3.44348818e-01
1.49842310e+00 5.33287346e-01 -5.23613870e-01 -1.03611708e-01
1.49071670e+00 1.85977310e-01 -4.02841061e-01 -1.98157921e-01
6.78299785e-01 -4.73900795e-01 -1.40831649e-01 -7.72611260e-01
-1.07452130e+00 1.33639193e+00 7.01884925e-01 6.23405695e-01
1.01275969e+00 -1.91477358e-01 7.53932655e-01 6.45941734e-01
6.51923299e-01 -6.41061962e-01 2.24014863e-01 8.44672084e-01
4.67447877e-01 -1.32621467e+00 -3.10802102e-01 -1.98035300e-01
-2.15890825e-01 1.47547019e+00 9.56684768e-01 -3.36607426e-01
8.81528795e-01 8.06662560e-01 1.23496756e-01 -4.55986083e-01
-5.32929957e-01 5.94554186e-01 -9.15009603e-02 8.13784182e-01
5.56183577e-01 -1.97830319e-01 -1.10655241e-01 9.94096100e-01
5.76307357e-04 -2.34083921e-01 9.57979918e-01 1.30848825e+00
-5.91196597e-01 -1.31323421e+00 -4.68333453e-01 4.04837191e-01
-9.74331617e-01 -4.22300965e-01 -5.64275980e-01 5.68658113e-01
-2.99583543e-02 1.10266078e+00 -5.84832668e-01 -8.98218989e-01
2.31594980e-01 2.10881740e-01 -2.89853245e-01 -8.25281441e-01
-1.00725937e+00 -4.69477296e-01 -1.02207989e-01 -4.11030740e-01
2.13890582e-01 -3.97798538e-01 -1.17218029e+00 -3.41417700e-01
-9.93063748e-02 3.96318644e-01 5.11028826e-01 5.89741707e-01
1.83012873e-01 4.31407630e-01 7.73734629e-01 -2.22646877e-01
-8.66535246e-01 -6.96653605e-01 -3.55464458e-01 1.07609995e-01
4.60839808e-01 -3.13614368e-01 -5.67313015e-01 2.32464820e-01] | [7.046045303344727, 7.776185035705566] |
b5e87b6b-38f9-4800-b0ff-565c74be4cb0 | setgen-scalable-and-efficient-template | 2211.05622 | null | https://arxiv.org/abs/2211.05622v1 | https://arxiv.org/pdf/2211.05622v1.pdf | SETGen: Scalable and Efficient Template Generation Framework for Groupwise Medical Image Registration | Template generation is a crucial step of groupwise image registration which deforms a group of subjects into a common space. Existing traditional and deep learning-based methods can generate high-quality template images. However, they suffer from substantial time costs or limited application scenarios like fixed group size. In this paper, we propose an efficient groupwise template generative framework based on variational autoencoder models utilizing the arithmetic property of latent representation of input images. We acquire the latent vectors of each input and use the average vector to construct the template through the decoder. Therefore, the method can be applied to groups of any scale. Secondly, we explore a siamese training scheme that feeds two images to the shared-weight twin networks and compares the distances between inputs and the generated template to prompt the template to be close to the implicit center. We conduct experiments on 3D brain MRI scans of groups of different sizes. Results show that our framework can achieve comparable and even better performance to baselines, with runtime decreased to seconds. | ['Albert C. S. Chung', 'Ziyi He'] | 2022-11-10 | null | null | null | null | ['medical-image-registration'] | ['medical'] | [ 4.50270951e-01 2.32496262e-01 1.61533728e-01 -5.66493332e-01
-9.36015964e-01 -2.05166548e-01 4.92311835e-01 -5.93431771e-01
-4.09725308e-01 4.37457323e-01 3.42976749e-01 2.73607373e-01
-6.45775869e-02 -5.30567229e-01 -6.08773112e-01 -9.83603776e-01
6.91244453e-02 5.57150483e-01 7.11914599e-02 6.41140789e-02
6.62702322e-02 2.22127065e-01 -1.07390213e+00 2.10691392e-01
6.97373867e-01 8.38030815e-01 3.82371873e-01 9.73279774e-02
3.84931386e-01 1.15103222e-01 -5.25895655e-01 -3.85863870e-01
5.72727621e-01 -6.37239814e-01 -5.62988758e-01 3.00785184e-01
6.62774265e-01 -5.50691068e-01 -4.70153421e-01 1.15542555e+00
9.31421876e-01 2.32514113e-01 8.98709655e-01 -1.12668598e+00
-8.13267767e-01 5.99685013e-01 -6.24371469e-01 1.33240119e-01
-5.40662520e-02 1.72577687e-02 6.26410782e-01 -8.67147923e-01
8.44602406e-01 9.10840631e-01 5.14961600e-01 8.90781641e-01
-1.32209706e+00 -8.30354571e-01 -9.39298943e-02 9.85741541e-02
-1.45494938e+00 -6.22525573e-01 9.29991603e-01 -5.03984451e-01
7.34022677e-01 -7.65771568e-02 7.23653793e-01 1.18907893e+00
5.54426610e-01 4.71172899e-01 1.17664289e+00 -1.39068305e-01
9.03679524e-03 -3.65466595e-01 -3.08693886e-01 5.88406205e-01
3.54798511e-02 7.02100918e-02 -4.56141859e-01 -1.43016845e-01
1.18683660e+00 5.36772907e-02 -4.54594344e-01 -4.13109541e-01
-1.59680223e+00 1.01821423e+00 6.72228515e-01 5.09929359e-01
-7.15303600e-01 8.33649412e-02 -1.27607258e-02 7.26793408e-02
4.16167468e-01 2.54754484e-01 8.77864361e-02 4.72282857e-01
-1.34662306e+00 2.86966205e-01 3.00602764e-01 7.87920654e-01
5.04058361e-01 1.30480230e-01 -4.18882608e-01 7.86175549e-01
4.53052640e-01 4.19104725e-01 8.60373855e-01 -9.61749434e-01
3.80042285e-01 2.90533841e-01 -4.31432426e-01 -1.10779035e+00
-3.06097507e-01 -4.44281250e-01 -1.13933814e+00 5.27904667e-02
2.27491871e-01 -2.14236706e-01 -1.01761699e+00 1.89466763e+00
4.43317860e-01 5.83075643e-01 -2.44703948e-01 1.07481265e+00
7.79797852e-01 3.97879720e-01 -3.02483290e-01 -2.78617024e-01
1.22177112e+00 -9.26567674e-01 -7.35907376e-01 -2.37259775e-01
2.74297625e-01 -7.75044680e-01 6.71573222e-01 -1.45769417e-02
-1.36070037e+00 -5.00920355e-01 -1.08740878e+00 -6.73824847e-02
1.28022537e-01 1.08384766e-01 3.61601204e-01 4.36490655e-01
-1.27182615e+00 7.16804564e-01 -1.05081558e+00 -8.42742249e-03
6.59511983e-01 5.35481274e-01 -5.41475654e-01 1.05541013e-01
-9.21511054e-01 7.54105151e-01 1.04794145e-01 2.25034729e-01
-8.79541814e-01 -9.55353558e-01 -8.87437522e-01 -1.43186972e-01
-6.16428107e-02 -9.36398029e-01 8.83102715e-01 -8.49406242e-01
-1.46259546e+00 1.01777422e+00 -1.44591480e-01 -2.87125170e-01
5.92651308e-01 1.99981317e-01 -3.56714329e-04 1.15204841e-01
2.77698427e-01 9.59912598e-01 1.18828845e+00 -9.09372389e-01
8.00799951e-02 -5.45781732e-01 -4.62364495e-01 3.27674419e-01
-8.14113021e-02 8.47472530e-03 -4.09571737e-01 -8.90141547e-01
4.90900695e-01 -1.08610475e+00 -3.82271707e-01 1.22112416e-01
-3.44518960e-01 -1.37845606e-01 2.92974025e-01 -8.41821671e-01
7.26207554e-01 -2.14962387e+00 4.61113781e-01 2.97726810e-01
5.32131314e-01 -1.32411525e-01 -1.77176371e-01 -6.51976019e-02
-3.29706937e-01 -1.51232973e-01 -4.05851305e-01 -4.12840068e-01
-7.40504414e-02 7.96474591e-02 -7.82786235e-02 7.70944357e-01
-9.10154358e-02 1.07193911e+00 -6.34957016e-01 -6.97789967e-01
-8.10182765e-02 6.73752725e-01 -8.91772628e-01 5.08189976e-01
4.11852092e-01 6.42696261e-01 -2.44442776e-01 1.40186742e-01
6.57380283e-01 -2.37426758e-01 2.98682749e-01 -6.86298192e-01
2.73135841e-01 -1.77084766e-02 -9.30027127e-01 2.04713440e+00
-1.24182411e-01 3.95019799e-01 -7.32217059e-02 -1.14344144e+00
8.30466330e-01 4.20439452e-01 7.31481850e-01 -5.84424198e-01
3.88398319e-01 1.58396959e-01 3.04046601e-01 -3.04706782e-01
-7.01967031e-02 -3.30486208e-01 9.10613015e-02 9.10187542e-01
3.81140232e-01 -8.94884467e-02 -1.41334087e-01 7.16912374e-02
7.32460678e-01 1.11754172e-01 8.48633796e-02 -2.42426917e-01
2.20796347e-01 -6.14308059e-01 5.13293266e-01 5.40803373e-01
-2.85113901e-01 1.04359472e+00 2.26107419e-01 -4.08099174e-01
-1.25974965e+00 -1.28562307e+00 -5.11183888e-02 5.54502070e-01
-1.20632589e-01 5.44315986e-02 -1.04855216e+00 -5.03802180e-01
-4.17194277e-01 3.27825010e-01 -9.50472713e-01 -2.30091944e-01
-8.14547420e-01 -7.94646740e-01 3.23503822e-01 4.02103573e-01
5.11110425e-01 -1.15848660e+00 -5.89644194e-01 2.98057497e-01
-4.50380534e-01 -1.11656260e+00 -9.98131335e-01 -3.08924049e-01
-1.07927680e+00 -8.25896621e-01 -1.31251574e+00 -1.10117805e+00
1.16440868e+00 6.70724958e-02 7.22126186e-01 7.74865150e-02
-2.49146774e-01 -4.00977246e-02 8.29247013e-02 -7.02587515e-03
-1.61125794e-01 7.04165995e-02 1.36467353e-01 3.43196392e-01
-5.13353534e-02 -8.54971349e-01 -9.46392655e-01 3.63296419e-01
-8.66636693e-01 3.51634175e-01 5.31098843e-01 1.00947249e+00
7.54461884e-01 -3.90420675e-01 2.93036550e-01 -6.55038536e-01
7.67862439e-01 -2.77603805e-01 -4.14992750e-01 2.59821832e-01
-3.85319889e-01 1.49072856e-01 1.92838728e-01 -6.61065876e-01
-8.33359003e-01 8.57294872e-02 -7.88347945e-02 -6.67570829e-01
2.20093265e-01 2.71731585e-01 -1.13877989e-01 -7.01350346e-02
5.10318518e-01 4.12598312e-01 3.74857008e-01 -1.62314236e-01
3.14998090e-01 3.22376102e-01 5.86320698e-01 -4.20187771e-01
8.71418118e-01 5.46404243e-01 -2.87499636e-01 -3.26437801e-01
-4.53065962e-01 2.11815699e-03 -9.86975312e-01 -2.91305542e-01
1.07757616e+00 -7.84909070e-01 -4.32602763e-01 6.00179911e-01
-1.09243727e+00 -4.07524407e-01 -1.00640813e-02 7.95840561e-01
-8.00705731e-01 2.40819365e-01 -5.32674611e-01 7.30139539e-02
-7.69705057e-01 -1.49612987e+00 1.22433913e+00 1.57097563e-01
-1.87135756e-01 -8.08975458e-01 1.64565623e-01 3.39851677e-01
4.21516865e-01 3.47968161e-01 7.33428717e-01 -5.00383854e-01
-4.74359572e-01 -9.10138711e-02 -1.38828577e-02 1.05751112e-01
1.61463588e-01 -4.30995226e-01 -7.08521187e-01 -4.42144722e-01
3.38581413e-01 -6.73552230e-02 6.29838645e-01 6.63330376e-01
1.14746332e+00 -2.90239215e-01 -3.58061075e-01 9.22023654e-01
1.15767646e+00 1.80840507e-01 8.47446859e-01 4.86871451e-02
5.76691270e-01 5.35188854e-01 1.81981809e-02 1.61394477e-01
3.03200841e-01 8.74030888e-01 -8.51808116e-02 -1.34444535e-01
-3.67691189e-01 3.66529748e-02 2.76527256e-01 1.41014540e+00
-2.48395711e-01 1.92862481e-01 -7.47682631e-01 6.04296029e-01
-1.61375034e+00 -1.07459021e+00 2.85363078e-01 2.07561207e+00
1.05485749e+00 -2.64901310e-01 -1.40302524e-01 -2.08159953e-01
7.67441869e-01 2.53071934e-01 -4.93420392e-01 -1.76853184e-02
2.13259086e-01 5.59703469e-01 1.81420997e-01 3.31707448e-01
-8.38611126e-01 8.65035713e-01 6.52343369e+00 6.87628746e-01
-1.35143399e+00 5.71765423e-01 5.07637680e-01 -3.01841885e-01
-2.06665337e-01 -3.87565702e-01 -3.43936026e-01 4.21644598e-01
7.35017002e-01 -3.49601299e-01 6.44753873e-01 3.96609157e-01
1.01782866e-01 3.45954746e-01 -1.06757522e+00 1.08611655e+00
4.62088943e-01 -1.30279303e+00 -2.65115197e-03 2.24293470e-01
9.04074132e-01 9.75438580e-02 2.94856012e-01 -6.62133843e-02
5.39016053e-02 -1.11399722e+00 5.80603182e-01 6.92969143e-01
8.84361327e-01 -4.99306381e-01 6.12746060e-01 1.57793939e-01
-8.22561681e-01 5.97432554e-01 -6.71360672e-01 4.47532773e-01
2.40086392e-01 3.27191353e-01 -7.64260650e-01 3.69114161e-01
5.65202057e-01 4.91986185e-01 -4.30159926e-01 9.27017510e-01
-1.70199722e-01 2.65299320e-01 -3.22824925e-01 2.72707731e-01
6.42074943e-02 -3.37022305e-01 2.53394395e-01 5.08937359e-01
6.43711805e-01 2.53780961e-01 4.75446694e-02 1.11673176e+00
-1.87200055e-01 2.24486724e-01 -6.97313309e-01 2.75934875e-01
2.45659783e-01 1.30722535e+00 -8.22421968e-01 -2.51076430e-01
-2.68728316e-01 1.16120422e+00 3.84353101e-01 3.25029254e-01
-1.02273285e+00 -2.24427968e-01 4.38362360e-01 -1.22139174e-02
3.36898655e-01 -3.53825718e-01 -1.85855314e-01 -1.25365806e+00
4.45155799e-02 -9.24291432e-01 8.70375037e-02 -8.45713377e-01
-1.21575773e+00 1.02163696e+00 4.27940898e-02 -1.16195416e+00
-3.91949683e-01 -1.79152191e-01 -7.00431168e-01 9.73595500e-01
-8.83276463e-01 -1.19191039e+00 -3.90796632e-01 9.12411630e-01
2.77226597e-01 -2.19103992e-01 8.78288209e-01 4.65469241e-01
-4.07929182e-01 8.43579113e-01 -8.53071958e-02 4.98010844e-01
8.05860937e-01 -8.16980660e-01 5.15126348e-01 8.23880613e-01
4.02641445e-01 8.43238652e-01 3.20378453e-01 -7.50565946e-01
-9.52587903e-01 -1.04630208e+00 8.47366154e-01 -2.23133877e-01
4.85184103e-01 -4.47465360e-01 -8.57446671e-01 8.55964065e-01
5.58837056e-01 9.25238952e-02 7.35423028e-01 -2.95561880e-01
-9.23188627e-02 1.19495159e-03 -1.11334515e+00 8.54082823e-01
1.13065648e+00 -5.44069409e-01 -7.48527050e-01 4.11333382e-01
5.28304398e-01 -7.08380699e-01 -1.25335908e+00 2.00327784e-01
7.86706388e-01 -7.52885103e-01 1.00399518e+00 -3.72940898e-01
5.15548408e-01 -2.00665653e-01 3.06322165e-02 -1.59248447e+00
-4.38890904e-01 -5.96609533e-01 3.51378173e-01 9.11856115e-01
3.59262377e-01 -6.46064758e-01 7.43123949e-01 7.20507383e-01
-5.15412875e-02 -7.46475995e-01 -1.23258591e+00 -4.72447544e-01
3.36114347e-01 -1.65081710e-01 8.40942383e-01 9.65351820e-01
-3.34651768e-01 9.56107378e-02 -4.43386555e-01 1.13209240e-01
9.40482557e-01 2.23266318e-01 4.86285180e-01 -8.36684585e-01
-1.86661407e-01 -3.84102017e-01 -4.99610066e-01 -7.64059305e-01
2.17081442e-01 -1.40749681e+00 7.87814334e-02 -1.41702509e+00
2.50242203e-01 -2.68954933e-01 -2.43977502e-01 6.06073439e-01
-1.46627247e-01 6.57678604e-01 1.24356791e-01 2.89471596e-01
-1.17798321e-01 6.74421847e-01 1.64261353e+00 -2.62736589e-01
-1.02895554e-02 -2.28550866e-01 -5.61307073e-01 4.31555808e-01
8.74186814e-01 -6.99124455e-01 -4.19373453e-01 -6.41653895e-01
-1.17643245e-01 1.89596117e-02 2.63645798e-01 -8.89350653e-01
2.93375969e-01 3.92543636e-02 6.80859208e-01 -2.84390956e-01
1.51834905e-01 -5.90535522e-01 4.16170895e-01 3.54428202e-01
-3.88318956e-01 1.24624543e-01 -1.80313304e-01 1.49916932e-01
-2.38308106e-02 -2.00390980e-01 8.79140139e-01 -2.91940961e-02
1.61153134e-02 8.07817340e-01 -1.17934875e-01 5.26207946e-02
1.07197165e+00 -2.72345364e-01 4.53876071e-02 -2.94981807e-01
-9.59848523e-01 -2.87273806e-02 2.48257399e-01 4.42826897e-01
9.64952350e-01 -1.73167098e+00 -9.49244678e-01 6.56171918e-01
-1.87694475e-01 -1.37356907e-01 3.28036308e-01 1.22696900e+00
-4.98856336e-01 1.89347416e-01 -5.39364040e-01 -7.44578660e-01
-1.17702663e+00 3.42338711e-01 6.89835787e-01 -1.37158230e-01
-9.34481263e-01 7.89762437e-01 4.70409483e-01 -4.91589010e-01
-6.65289015e-02 -1.66109934e-01 -2.01261714e-01 8.24492872e-02
5.90465248e-01 -1.33885652e-01 2.40124762e-02 -9.34700072e-01
-3.41855496e-01 8.37638319e-01 -1.97942659e-01 -3.38928968e-01
1.65079641e+00 1.96748257e-01 -1.89439282e-01 -5.79983965e-02
1.44739628e+00 -2.20206842e-01 -1.23754525e+00 -3.07758927e-01
-4.57917780e-01 -3.15665692e-01 1.56639770e-01 -2.82063067e-01
-1.81091857e+00 8.74544024e-01 1.02031875e+00 -4.61493224e-01
9.91238236e-01 4.08465900e-02 9.29833114e-01 -6.46958128e-02
4.74299073e-01 -8.11000645e-01 4.01123427e-02 -1.46479812e-02
1.11488259e+00 -9.94502366e-01 -1.16503071e-02 -1.04708165e-01
-5.95995605e-01 9.24509883e-01 4.05444622e-01 -5.63771367e-01
7.19424605e-01 1.39484003e-01 2.34767050e-01 -3.96810710e-01
-5.52162528e-01 1.05181001e-01 6.89907312e-01 7.61061430e-01
3.78620267e-01 1.17857978e-01 -3.48938555e-01 5.43008685e-01
-3.93109679e-01 -3.18108276e-02 2.88476765e-01 5.08718610e-01
2.26080529e-02 -1.21414912e+00 -2.69730061e-01 6.38216019e-01
-4.80841160e-01 -2.62860477e-01 5.74277267e-02 5.41437984e-01
1.09444976e-01 4.71376181e-01 1.92158580e-01 -1.76172733e-01
1.03150308e-01 1.06657907e-01 9.56133246e-01 -5.40298104e-01
-5.82090795e-01 4.41594571e-01 -4.84420776e-01 -7.76384234e-01
-6.34939134e-01 -8.27310622e-01 -1.11699200e+00 -6.06336892e-02
-2.55669802e-01 -7.05305487e-02 4.99522626e-01 8.91032696e-01
2.99971223e-01 4.04907644e-01 5.90477884e-01 -1.06544626e+00
-4.72885102e-01 -9.96898532e-01 -4.62337077e-01 5.71750641e-01
-4.67266887e-02 -7.10659504e-01 -9.60993543e-02 2.53349900e-01] | [13.985038757324219, -2.5141360759735107] |
9a78d54c-5dae-4feb-b49d-07e9fe3c048d | personality-testing-of-gpt-3-limited-temporal | 2306.04308 | null | https://arxiv.org/abs/2306.04308v1 | https://arxiv.org/pdf/2306.04308v1.pdf | Personality testing of GPT-3: Limited temporal reliability, but highlighted social desirability of GPT-3's personality instruments results | To assess the potential applications and limitations of chatbot GPT-3 Davinci-003, this study explored the temporal reliability of personality questionnaires applied to the chatbot and its personality profile. Psychological questionnaires were administered to the chatbot on two separate occasions, followed by a comparison of the responses to human normative data. The findings revealed varying levels of agreement in the chatbot's responses over time, with some scales displaying excellent while others demonstrated poor agreement. Overall, Davinci-003 displayed a socially desirable and pro-social personality profile, particularly in the domain of communion. However, the underlying basis of the chatbot's responses, whether driven by conscious self-reflection or predetermined algorithms, remains uncertain. | ['Ljubisa Bojic', 'Bojana M. Dinic', 'Bojana Bodroza'] | 2023-06-07 | null | null | null | null | ['chatbot', 'chatbot'] | ['methodology', 'natural-language-processing'] | [-6.23425305e-01 6.33421183e-01 9.25805122e-02 -5.01025736e-01
-1.28803074e-01 -4.52036291e-01 3.34537208e-01 4.63729799e-02
-3.78421336e-01 7.39418745e-01 2.57450402e-01 2.74960883e-02
-2.52190411e-01 -3.89653206e-01 4.25649971e-01 -4.15930569e-01
-6.98924204e-03 5.49685419e-01 -6.78671449e-02 -5.84805131e-01
5.91238022e-01 8.93288031e-02 -8.18582594e-01 1.67723939e-01
7.90857673e-01 4.70464617e-01 1.51549190e-01 7.53107548e-01
3.15251574e-02 1.23540771e+00 -1.06344867e+00 -9.67796385e-01
-1.23332620e-01 -6.34697199e-01 -1.14021528e+00 2.81008720e-01
-1.35680497e-01 -6.22736096e-01 -7.21257403e-02 8.41045141e-01
1.92902818e-01 3.25511634e-01 3.62223983e-01 -1.47251296e+00
-9.54208255e-01 5.56128919e-01 1.47791147e-01 2.76587829e-02
6.96608067e-01 5.97868800e-01 9.83209074e-01 -3.21331114e-01
7.78398812e-01 1.32664728e+00 9.50115204e-01 7.68400490e-01
-1.51926708e+00 -7.17064798e-01 -5.74881852e-01 1.56164944e-01
-9.56890464e-01 -4.63860542e-01 5.20737469e-01 -5.84994435e-01
9.18726802e-01 -1.53653203e-02 1.03074038e+00 1.10948133e+00
4.04053986e-01 3.25515419e-02 1.68191075e+00 1.01791270e-01
3.26337785e-01 9.36864674e-01 1.97855443e-01 1.20675541e-01
1.38266593e-01 -3.97222042e-02 -5.15636444e-01 -6.79157734e-01
8.14208508e-01 -4.44256991e-01 4.42689985e-01 1.76772788e-01
-5.01532972e-01 9.87286687e-01 1.49340749e-01 5.16121447e-01
-7.94472992e-01 -1.82361275e-01 8.11190665e-01 8.00548077e-01
4.72115427e-01 6.67436898e-01 2.16561346e-03 -1.27954423e+00
-2.24297598e-01 2.46751159e-01 1.14257407e+00 8.41495454e-01
6.16361260e-01 -1.10585466e-01 -8.75446796e-02 9.85596299e-01
1.87180713e-01 -4.57433201e-02 2.46800795e-01 -1.44293106e+00
7.83883780e-02 6.45135641e-01 5.40533006e-01 -1.10627496e+00
-5.67755163e-01 2.50899076e-01 -1.43926665e-01 2.73155481e-01
5.64500034e-01 -3.11303288e-01 2.51257986e-01 1.52255785e+00
9.48951393e-02 -8.34810913e-01 3.64522729e-03 6.09919429e-01
5.55019677e-01 -1.19358368e-01 5.79950333e-01 -4.04887319e-01
1.09841406e+00 -3.33183259e-01 -7.22352505e-01 -2.12977365e-01
5.38134992e-01 -6.66594207e-01 1.36951208e+00 1.49872378e-01
-1.26484728e+00 -4.19397235e-01 -4.85918701e-01 7.10974410e-02
1.83805168e-01 -1.99097931e-01 6.37305856e-01 1.32677341e+00
-1.37969851e+00 8.81720245e-01 -2.93628961e-01 -1.19884884e+00
8.58612731e-02 3.07370245e-01 -3.67191464e-01 4.26057577e-01
-1.00097632e+00 1.25797391e+00 4.14491631e-02 -9.01939124e-02
-3.35926443e-01 -4.88552749e-02 -2.85346389e-01 1.46594405e-01
-1.81569651e-01 -1.36289805e-01 1.61875331e+00 -7.69451141e-01
-2.06488061e+00 1.03201401e+00 6.94817007e-02 -2.50860989e-01
4.57700282e-01 4.08309609e-01 -3.36366296e-01 3.64200443e-01
4.72090036e-01 3.09022129e-01 3.90364498e-01 -8.84066880e-01
-3.98753107e-01 -1.82204425e-01 2.14228094e-01 3.12905908e-01
-3.88039827e-01 5.99378586e-01 5.44534326e-01 2.95050710e-01
-1.95250615e-01 -7.17266023e-01 -9.29831266e-02 -3.99773180e-01
2.36606851e-01 -6.98558688e-01 5.06251931e-01 -6.38972223e-01
8.06408346e-01 -2.25640559e+00 -8.75484407e-01 -2.89934855e-02
5.50994337e-01 -5.13173863e-02 -5.10313995e-02 1.11438727e+00
3.63495797e-01 1.78258896e-01 4.76989836e-01 -1.88187569e-01
2.19773009e-01 1.54863775e-01 4.26184148e-01 7.30687380e-01
1.40915662e-01 7.53336012e-01 -1.20398724e+00 -6.20496392e-01
7.85166845e-02 -2.52132684e-01 -4.68368620e-01 5.37684143e-01
2.72289723e-01 4.47778434e-01 -8.78916234e-02 3.58552873e-01
5.49799979e-01 3.95762324e-02 5.52372158e-01 7.89505124e-01
-3.19527119e-01 8.18523228e-01 -1.27092808e-01 7.93801248e-01
-3.06701474e-02 6.70264721e-01 6.64129078e-01 -4.62049484e-01
1.60850489e+00 7.59359539e-01 6.04194283e-01 -6.33032441e-01
4.31622744e-01 8.68724808e-02 6.47571862e-01 -8.30514252e-01
9.79065359e-01 -6.99771941e-01 -1.74668372e-01 8.96591067e-01
1.42693534e-01 -2.56848991e-01 1.20606184e-01 2.12110758e-01
1.21926606e+00 -2.39635184e-01 4.59602207e-01 -4.33026820e-01
-6.14093021e-02 3.57437670e-01 6.47657335e-01 7.56029546e-01
-1.17823088e+00 -1.22949667e-01 1.09445691e+00 5.96231148e-02
-9.26616907e-01 -7.05751121e-01 1.67059883e-01 1.09879816e+00
-7.11395741e-02 -3.57800037e-01 -8.52376461e-01 -5.72714150e-01
-1.29525200e-01 1.06318843e+00 -6.48019552e-01 -2.63198435e-01
6.25947416e-02 -1.45849288e-01 6.62743866e-01 6.61235824e-02
5.61822474e-01 -1.46860397e+00 -6.69817448e-01 7.47940361e-01
-4.58896816e-01 -1.04989052e+00 -2.14162380e-01 -2.23416343e-01
-9.43768501e-01 -8.64466846e-01 1.09279552e-03 -3.63893390e-01
1.42633423e-01 4.87171859e-01 8.89359832e-01 4.26837742e-01
2.90623605e-01 6.43874228e-01 -4.87947762e-01 -2.91417003e-01
-9.20583546e-01 2.17881911e-02 1.57543689e-01 -5.04494905e-01
7.96227574e-01 -9.83051062e-01 -2.68598557e-01 4.87411082e-01
-9.54805538e-02 -4.41629946e-01 2.87409365e-01 7.46084750e-01
-9.27510738e-01 -2.14614093e-01 9.76727605e-01 -8.52045119e-01
1.34773481e+00 -8.79011750e-01 2.37183690e-01 -2.98852742e-01
-7.59306133e-01 -1.12133634e+00 6.20441854e-01 -4.66631889e-01
-1.35569632e+00 -5.40241539e-01 3.55410017e-02 1.79642364e-01
-2.33339712e-01 3.21177065e-01 3.12753141e-01 -2.66583562e-01
9.12741840e-01 -2.09972635e-01 7.32926726e-01 -1.68211386e-01
-1.40365988e-01 1.11711133e+00 2.90270954e-01 -9.15809155e-01
4.56518799e-01 1.49162728e-02 -6.74811125e-01 -1.02534676e+00
-3.01584274e-01 -5.26242137e-01 -6.27256989e-01 -8.13453257e-01
7.00257599e-01 -5.49121201e-01 -1.70730615e+00 3.29429537e-01
-1.05449116e+00 -7.63036847e-01 -1.30520016e-01 4.51200455e-01
-7.65723646e-01 2.38172159e-01 -1.06700492e+00 -1.45311129e+00
-3.31208646e-01 -6.83109581e-01 1.33861378e-01 2.36682326e-01
-1.33032846e+00 -9.80179250e-01 2.80933112e-01 7.33110845e-01
4.15491700e-01 -2.76161760e-01 8.02282035e-01 -7.95794725e-01
2.35397488e-01 -3.91703576e-01 -2.06009954e-01 3.11893553e-01
4.03262079e-01 -1.03281699e-01 -6.87737286e-01 -6.73330098e-04
4.52832013e-01 -8.91332746e-01 -4.83330846e-01 -1.81491613e-01
-6.12050071e-02 -6.55703843e-01 3.39787960e-01 -3.84524077e-01
8.11283827e-01 4.79665875e-01 8.15250456e-01 5.66913784e-01
-1.32281318e-01 1.00198758e+00 8.45823646e-01 6.59893394e-01
5.46252668e-01 3.72421175e-01 1.92765638e-01 4.48698670e-01
5.45705914e-01 -2.32433096e-01 7.41777599e-01 6.07275426e-01
-2.83646673e-01 4.83035654e-01 -4.94692713e-01 5.24287701e-01
-1.61940467e+00 -1.37755442e+00 -4.26297396e-01 1.73870850e+00
4.78810757e-01 9.84218121e-02 1.02383053e+00 -2.34068424e-01
8.71147811e-01 8.12188461e-02 -4.11023766e-01 -1.27883995e+00
3.58584166e-01 1.08236820e-02 1.14433698e-01 2.32811928e-01
-1.76010989e-02 9.21946049e-01 7.98715639e+00 9.46039800e-03
-6.61454260e-01 3.56133312e-01 3.70662510e-01 1.92939490e-02
2.55052149e-01 2.76669919e-01 -6.32219017e-02 2.78652787e-01
1.25706410e+00 -7.81867087e-01 6.10861123e-01 1.08033502e+00
7.98790991e-01 -3.29796016e-01 -7.49974132e-01 2.84485281e-01
-2.67914981e-01 -6.68014228e-01 -1.04997098e+00 3.35488170e-01
2.96095371e-01 -3.21451008e-01 -2.07265884e-01 8.43532205e-01
7.73306608e-01 -6.35068953e-01 8.11949432e-01 1.02326706e-01
5.63920915e-01 -3.27373743e-01 7.10301220e-01 4.44835484e-01
-4.33191299e-01 -1.61299154e-01 -4.88869160e-01 -1.15987194e+00
1.66962072e-01 -3.13929498e-01 -1.42687583e+00 -3.38379703e-02
7.79598594e-01 4.20913428e-01 -2.36253783e-01 4.66707259e-01
-1.88391775e-01 9.61008787e-01 1.45823687e-01 -7.07074463e-01
2.33256802e-01 -6.49665296e-01 4.71609235e-01 8.95175576e-01
5.62399589e-02 3.76749009e-01 -2.74076015e-01 1.12978029e+00
2.06515655e-01 1.73884913e-01 -4.41241592e-01 -4.67287660e-01
9.39409912e-01 1.49473536e+00 -7.09659874e-01 -3.87881756e-01
-2.54434139e-01 8.31011832e-01 6.65535927e-01 2.47863770e-01
-4.01999414e-01 2.92652920e-02 1.09214282e+00 1.97227329e-01
-2.69255966e-01 -3.14026564e-01 -9.62935686e-01 -5.52680492e-01
-2.74844378e-01 -9.19969916e-01 1.72283500e-01 -8.30979645e-01
-1.53424323e+00 1.16323523e-01 -1.93117663e-01 -7.70396471e-01
-1.29515380e-01 -1.11959457e-01 -1.14989758e+00 9.18340504e-01
-1.64998323e-01 -7.45800436e-01 -2.97385126e-01 2.52808988e-01
1.65667087e-01 6.19854257e-02 8.90033066e-01 -2.41282210e-01
-2.91857719e-01 4.59647417e-01 -3.98102961e-02 -4.63372543e-02
9.74186718e-01 -1.04722333e+00 3.17548364e-01 -3.98975015e-02
-9.52154398e-01 9.61440504e-01 9.15995121e-01 -7.87120223e-01
-9.73931313e-01 -4.07182217e-01 9.49654698e-01 -4.40908641e-01
1.09054708e+00 -2.77682632e-01 -9.04225826e-01 7.54100144e-01
6.03059709e-01 -1.08391273e+00 8.14284146e-01 6.33185446e-01
1.32038549e-01 2.28380680e-01 -1.82633352e+00 7.07972884e-01
1.00936651e+00 -7.62935340e-01 -6.21057093e-01 1.85848370e-01
6.35970309e-02 6.40400648e-02 -1.34709716e+00 -7.58283257e-01
7.05099285e-01 -1.46845400e+00 1.94410637e-01 -4.36164856e-01
5.99575281e-01 6.41717076e-01 2.82509863e-01 -1.33735359e+00
-8.89222324e-01 -9.16728675e-01 6.87164843e-01 1.47751153e+00
-1.50614738e-01 -1.07862103e+00 9.35484827e-01 1.41491878e+00
-1.14246115e-01 -2.65745431e-01 -8.90004218e-01 -7.27279961e-01
2.52152830e-01 -1.19547367e-01 4.24416006e-01 1.05092621e+00
1.29901087e+00 4.31313545e-01 -4.85098600e-01 -4.67181623e-01
4.83777463e-01 -5.87017179e-01 1.18243742e+00 -1.35268497e+00
-1.50871769e-01 -4.76464391e-01 -5.51587939e-01 -4.70435292e-01
1.59700871e-01 -4.51693863e-01 1.46637887e-01 -1.45232677e+00
2.78179556e-01 -2.97924906e-01 5.14102578e-01 3.01781386e-01
-1.35604501e-01 -1.12636082e-01 4.74421620e-01 3.47363710e-01
-3.58405977e-01 5.70856214e-01 1.05226314e+00 5.53292274e-01
-5.73367417e-01 5.03615774e-02 -1.23236883e+00 4.12625283e-01
1.29143846e+00 -5.31913459e-01 -2.78925449e-01 3.62758726e-01
-1.34839162e-01 5.61168373e-01 2.36962140e-01 -7.54270196e-01
1.53490692e-01 -6.76053345e-01 -1.88980326e-01 5.14783338e-02
5.34455359e-01 -4.06420410e-01 1.30771980e-01 4.17626232e-01
-1.72298670e-01 4.19721790e-02 -4.47463356e-02 -1.43412892e-02
1.79432198e-01 -5.14876127e-01 7.55698383e-01 -3.39896441e-01
-1.04222268e-01 -3.59302253e-01 -1.49573004e+00 -1.43372685e-01
8.59842598e-01 -6.80738688e-01 -3.74825090e-01 -1.18846405e+00
-5.23706675e-01 6.34563491e-02 9.54157710e-01 7.93295354e-02
2.33466029e-01 -1.05727828e+00 -4.65603828e-01 -4.18226451e-01
-7.43681341e-02 -9.54425573e-01 5.04155517e-01 1.36709595e+00
-2.88911104e-01 1.17929704e-01 -7.21212685e-01 -1.09017760e-01
-1.35893810e+00 9.54083651e-02 2.06955791e-01 7.15831369e-02
-6.93187475e-01 2.94432789e-01 -2.07797185e-01 -4.25239027e-01
-1.98737770e-01 4.55917984e-01 -2.13856563e-01 -3.72034833e-02
3.19654077e-01 7.93267667e-01 -4.20458347e-01 -7.72419691e-01
-2.91078836e-02 -5.57818770e-01 7.47518614e-02 -5.82760513e-01
1.16405249e+00 -5.17549634e-01 -4.80867118e-01 8.06674600e-01
8.11516821e-01 -2.98767388e-01 -9.02971327e-01 1.69678360e-01
2.26573348e-01 -9.87291813e-01 -8.95639122e-01 -5.54268122e-01
-6.21896796e-02 4.73075211e-01 -7.54291192e-02 8.96214545e-01
4.61274058e-01 -6.42042607e-02 5.68144202e-01 3.30483973e-01
6.12143695e-01 -1.65903902e+00 2.42611781e-01 7.33693600e-01
9.11621690e-01 -9.03412759e-01 -3.86480510e-01 -1.99298546e-01
-1.34502053e+00 1.07786524e+00 8.83812308e-01 -2.89836854e-01
2.07745269e-01 -3.10572773e-01 4.00758415e-01 -4.96282399e-01
-1.07820714e+00 1.79148838e-01 -7.06397057e-01 1.08977509e+00
8.63399863e-01 3.76001298e-01 -1.13988328e+00 7.91957796e-01
-1.06525242e+00 1.26079485e-01 1.16702664e+00 7.18915701e-01
-4.89589602e-01 -9.30214882e-01 -4.66144532e-01 5.81069887e-01
-2.13113308e-01 4.63857651e-01 -1.18305206e+00 8.63120794e-01
-3.46737981e-01 1.73219287e+00 -1.41293541e-01 -7.73601174e-01
3.87842566e-01 3.34512711e-01 2.29459211e-01 -7.90365338e-01
-1.42194307e+00 -2.89023578e-01 8.59415889e-01 -3.18648666e-01
-2.74056435e-01 -1.22051251e+00 -9.09859657e-01 -1.09842205e+00
-1.29433110e-01 6.05822623e-01 3.12178850e-01 9.84964788e-01
2.37813629e-02 -2.59520143e-01 8.44785571e-01 -5.79872012e-01
-8.33913624e-01 -1.74290800e+00 -1.16216588e+00 6.60764992e-01
-3.51049334e-01 -4.95167404e-01 -3.71245623e-01 -4.45947021e-01] | [12.517544746398926, 7.796350002288818] |
7d32f035-9274-4c61-a187-8a620e0d2144 | remote-sensing-image-classification | 1410.5358 | null | http://arxiv.org/abs/1410.5358v3 | http://arxiv.org/pdf/1410.5358v3.pdf | Remote sensing image classification exploiting multiple kernel learning | We propose a strategy for land use classification which exploits Multiple
Kernel Learning (MKL) to automatically determine a suitable combination of a
set of features without requiring any heuristic knowledge about the
classification task. We present a novel procedure that allows MKL to achieve
good performance in the case of small training sets. Experimental results on
publicly available datasets demonstrate the feasibility of the proposed
approach. | ['Paolo Napoletano', 'Claudio Cusano', 'Raimondo Schettini'] | 2014-10-20 | null | null | null | null | ['remote-sensing-image-classification'] | ['miscellaneous'] | [-5.04790153e-03 -4.14022624e-01 -5.56388259e-01 -4.91777092e-01
-8.45402837e-01 -3.27083647e-01 6.81928396e-01 2.93326467e-01
-8.26762140e-01 9.02887702e-01 -3.58887285e-01 -6.79879189e-01
-4.77537245e-01 -1.08178747e+00 -2.28580803e-01 -8.45774829e-01
-3.42205763e-01 -1.20601833e-01 3.94353509e-01 -3.82229835e-02
2.45064974e-01 7.77875781e-01 -1.47318792e+00 -1.02206124e-02
1.14706922e+00 8.00422966e-01 4.98526603e-01 7.68172383e-01
5.18071763e-02 7.02676594e-01 -1.22657321e-01 6.05570041e-02
3.65594178e-01 -4.84301709e-02 -8.02996039e-01 -8.55894876e-04
2.42395535e-01 -1.49210379e-01 1.54897541e-01 8.44876587e-01
1.18567504e-01 1.49664313e-01 1.19804060e+00 -1.07865536e+00
-6.57640994e-02 1.42380252e-01 -4.76152867e-01 1.54840767e-01
-5.29297329e-02 -2.44146317e-01 9.81584489e-01 -7.26508498e-01
8.50206539e-02 7.59344816e-01 9.28005159e-01 -1.28270626e-01
-1.34958124e+00 -3.15033555e-01 2.82673799e-02 4.00196314e-01
-1.93525183e+00 -3.65155607e-01 6.51509285e-01 -4.75111187e-01
9.42965984e-01 3.92064303e-01 5.22887766e-01 6.50331751e-02
3.66049856e-01 7.77776361e-01 1.54858458e+00 -7.26718426e-01
2.89936751e-01 5.67991972e-01 5.93728304e-01 8.94732058e-01
2.53114432e-01 -8.91842544e-02 2.28205055e-01 -7.62337923e-01
8.32611740e-01 -1.23867899e-01 1.66900344e-02 -4.75080669e-01
-1.05172074e+00 9.35263276e-01 1.31861210e-01 4.30187583e-01
-6.13089323e-01 -7.37079466e-03 1.54324204e-01 5.13841093e-01
6.22850060e-01 2.97198206e-01 -7.30018616e-01 3.04258525e-01
-9.99657273e-01 1.00752167e-01 8.10010195e-01 6.58358991e-01
1.29304242e+00 -3.18184435e-01 2.72845626e-01 8.22164178e-01
1.80618867e-01 7.25491941e-01 2.71736205e-01 -2.11005360e-01
2.72962034e-01 7.72401690e-01 2.03227341e-01 -7.51919985e-01
-3.77995789e-01 2.41619572e-01 -8.01344454e-01 3.71809483e-01
2.75850028e-01 -4.57478128e-02 -7.08679199e-01 1.19832778e+00
3.04994464e-01 4.39571083e-01 5.30751050e-01 2.00546131e-01
6.18743181e-01 7.25460887e-01 2.97428459e-01 -9.94065627e-02
9.75082219e-01 -6.16320491e-01 -3.32536012e-01 -5.71155734e-02
1.04464257e+00 -4.88723844e-01 8.44055116e-01 5.46666980e-02
-2.16985881e-01 -4.55577910e-01 -9.67005014e-01 4.95947778e-01
-7.96067834e-01 7.47979760e-01 1.11009705e+00 8.04481506e-01
-9.40033078e-01 3.98230940e-01 -7.14740098e-01 -4.34978396e-01
2.52121896e-01 5.08009017e-01 -7.24469125e-01 1.29482806e-01
-8.98992538e-01 1.05672204e+00 7.72673786e-01 2.69112647e-01
-4.25514609e-01 -2.34742910e-01 -9.30834174e-01 -1.68002725e-01
1.78400010e-01 -6.63295314e-02 9.02744532e-01 -6.98480844e-01
-1.16895902e+00 6.58410430e-01 -1.31196946e-01 -4.21373636e-01
4.42305088e-01 -2.16634929e-01 -4.28426296e-01 8.75054076e-02
3.62958349e-02 1.06567577e-01 1.03983903e+00 -9.45096254e-01
-8.20573568e-01 -1.06925126e-02 -6.54600710e-02 1.51786625e-01
-5.39117992e-01 -1.33375466e-01 9.19054523e-02 -3.80121082e-01
1.61199376e-01 -8.94252539e-01 -7.21559048e-01 -2.26961151e-02
8.05495828e-02 -4.34438169e-01 6.75817788e-01 -5.02847075e-01
1.22359216e+00 -2.07759762e+00 -2.39953049e-03 7.75378227e-01
-2.10659385e-01 4.19616282e-01 -7.72570819e-02 4.72289026e-01
3.33132297e-02 -2.10285466e-03 -4.95018780e-01 2.40114212e-01
-4.03159350e-01 4.16884959e-01 -1.29946828e-01 7.46245444e-01
2.46761829e-01 8.02338958e-01 -9.00834739e-01 -7.95884252e-01
7.23041236e-01 1.39536396e-01 -5.31710982e-02 1.97874710e-01
3.22383374e-01 -2.24236354e-01 -6.99576020e-01 7.26046026e-01
7.09385991e-01 -4.61721048e-02 -1.51623383e-01 -1.16473973e-01
-3.01147163e-01 -2.47630417e-01 -1.49559665e+00 1.01574361e+00
-6.18065000e-01 4.06729311e-01 -3.33216757e-01 -1.11008143e+00
8.31708789e-01 1.30245313e-01 3.92676145e-01 1.31021421e-02
-1.28496215e-01 3.59949440e-01 -2.76943415e-01 -6.73686087e-01
1.24276049e-01 -1.89137474e-01 -2.28426456e-01 2.12709472e-01
1.01570055e-01 -1.23349696e-01 -5.55579178e-03 -3.93711925e-01
1.02246618e+00 -1.68397322e-01 1.32347167e+00 -7.00474501e-01
8.53781104e-01 5.30789010e-02 1.61028191e-01 1.00179362e+00
-3.78621817e-01 -7.92306885e-02 4.69868295e-02 -8.51158142e-01
-6.91067815e-01 -9.59596574e-01 -4.04990345e-01 9.91442442e-01
-1.07721992e-01 -1.60512924e-01 -2.82688797e-01 -1.00158167e+00
3.65980238e-01 5.78330576e-01 -7.79642105e-01 1.53492168e-01
-4.35065508e-01 -1.34409201e+00 5.98525822e-01 3.25310290e-01
7.04853773e-01 -8.47399294e-01 -7.04084396e-01 4.65767980e-02
6.18842319e-02 -8.85555029e-01 1.33616969e-01 3.30606222e-01
-9.94866729e-01 -1.09572566e+00 -4.95678008e-01 -9.05144453e-01
9.34971392e-01 6.46228611e-01 6.15105152e-01 3.93857695e-02
-2.02422798e-01 1.64611369e-01 -4.58085120e-01 -1.70437209e-02
-4.86173868e-01 5.02608001e-01 -1.10102363e-01 3.26098576e-02
3.82645965e-01 -4.84259635e-01 -6.36268780e-02 2.36525521e-01
-8.02606702e-01 -1.06801717e-02 1.16739845e+00 6.53479755e-01
2.60332048e-01 5.68229377e-01 5.29425442e-01 -7.78301418e-01
4.51971143e-01 -6.71710730e-01 -7.23662078e-01 8.21467340e-01
-5.79840481e-01 4.09444839e-01 3.56523782e-01 -3.88623804e-01
-8.61366808e-01 6.64699852e-01 7.14331269e-02 1.09564133e-01
-6.08011961e-01 8.45994949e-01 -6.09574951e-02 -7.23727524e-01
6.64845169e-01 7.06649780e-01 -1.22071631e-01 -6.11528099e-01
3.84700507e-01 8.72078955e-01 1.90907285e-01 -4.68156427e-01
8.78987014e-01 3.99807364e-01 9.22595188e-02 -1.33442795e+00
-4.47151095e-01 -8.63456190e-01 -1.25885534e+00 -1.17644906e-01
4.46019381e-01 -9.54924226e-01 -3.60127985e-01 6.32511318e-01
-5.83534062e-01 -2.43685067e-01 2.20845178e-01 7.77559936e-01
-7.13855386e-01 2.25425690e-01 -9.05994847e-02 -8.88282597e-01
-2.62353063e-01 -8.73470545e-01 8.05831194e-01 1.18691914e-01
1.93129381e-04 -1.31922185e+00 1.39890671e-01 -5.77288046e-02
2.77615547e-01 1.03845716e-01 1.01092494e+00 -7.01593518e-01
-2.73128122e-01 -5.73048294e-01 -3.44024062e-01 2.97696680e-01
9.02705908e-01 -5.86008932e-03 -8.13062429e-01 -1.58374399e-01
-4.18026626e-01 -1.29287183e-01 1.06763959e+00 3.15859079e-01
8.92571032e-01 -2.41577283e-01 -5.38339853e-01 4.61440533e-01
1.73172283e+00 -1.06367722e-01 5.08119404e-01 5.29168308e-01
5.56290329e-01 3.57557267e-01 8.26722622e-01 3.98412377e-01
4.79447216e-01 3.73946637e-01 4.02188441e-03 -5.45042217e-01
4.09839898e-01 1.24681115e-01 4.73582074e-02 3.59646618e-01
-3.11052024e-01 3.80192846e-01 -1.21175396e+00 6.28163815e-01
-2.16846347e+00 -8.74359965e-01 -1.21769970e-02 2.09007907e+00
8.21479619e-01 -8.52363929e-02 7.93843716e-02 1.71886891e-01
3.31974447e-01 4.97172624e-02 -2.85893053e-01 -1.43249810e-01
-9.36052725e-02 1.60491034e-01 1.14978409e+00 8.25827360e-01
-1.81930721e+00 1.35947657e+00 8.15374088e+00 1.12069952e+00
-8.09669197e-01 -1.47246584e-01 1.30282879e-01 5.99525094e-01
2.60073561e-02 3.44961474e-04 -9.33850527e-01 1.05216969e-02
6.34459794e-01 -3.68870050e-01 4.27932516e-02 9.49187398e-01
1.35001674e-01 -7.37779140e-01 -5.06865263e-01 7.79332757e-01
-2.16590419e-01 -1.07985651e+00 1.82901189e-01 -6.61991090e-02
5.03692746e-01 -3.09720542e-02 -2.50114679e-01 1.70816898e-01
3.84460658e-01 -5.69410145e-01 2.69470513e-01 6.07892990e-01
3.83312434e-01 -8.44391406e-01 7.99175560e-01 5.86233795e-01
-1.53977644e+00 -6.34616762e-02 -3.71723920e-01 -3.56889486e-01
-4.78268832e-01 4.74155575e-01 -1.27630019e+00 6.48997843e-01
2.80228257e-01 7.37817168e-01 -1.12504113e+00 1.20720589e+00
-3.22186172e-01 6.54752493e-01 -5.94674349e-01 -2.06701636e-01
3.97493750e-01 -2.02322677e-01 1.98622331e-01 1.45215821e+00
1.25094518e-01 1.79180369e-01 4.58532810e-01 2.29697004e-01
7.19056904e-01 7.45123863e-01 -9.94048893e-01 8.31273273e-02
2.54445583e-01 1.34190750e+00 -1.04673469e+00 -2.80893564e-01
-6.59596264e-01 1.10781479e+00 5.94782114e-01 3.23156446e-01
-3.15790623e-01 -4.48803604e-01 5.83107531e-01 -2.42222548e-01
3.65965486e-01 -6.05406165e-01 -6.41785637e-02 -1.15687799e+00
8.38316232e-02 -2.39388853e-01 5.53467691e-01 -1.86281130e-01
-1.15115845e+00 3.67420644e-01 5.88044941e-01 -1.41016006e+00
-3.30947280e-01 -6.36535466e-01 -4.50928450e-01 8.88750613e-01
-1.86264813e+00 -1.61512923e+00 2.78919991e-02 5.46646297e-01
2.29086101e-01 -3.00009519e-01 1.24971676e+00 -6.24847151e-02
-2.74850249e-01 3.98111343e-01 4.42481190e-01 1.47158489e-01
5.92382193e-01 -1.23049891e+00 2.01584682e-01 8.50863457e-01
1.73079204e-02 1.91568732e-01 5.77886581e-01 -6.03691339e-01
-1.01024950e+00 -1.12120259e+00 1.02596736e+00 -1.82101235e-01
7.08879471e-01 -1.43157139e-01 -7.87352979e-01 5.19816756e-01
-2.29903743e-01 2.22765043e-01 1.32811499e+00 3.00304770e-01
-3.20121646e-01 -1.36160374e-01 -1.23116148e+00 2.17471212e-01
3.47250760e-01 -5.11963725e-01 -5.57622671e-01 1.18306704e-01
4.01485860e-02 6.94538206e-02 -8.86206448e-01 5.99350274e-01
7.41786897e-01 -4.31492299e-01 8.95657837e-01 -3.66845816e-01
-2.40702361e-01 -6.54596269e-01 -4.34416413e-01 -1.30896401e+00
-4.51773643e-01 4.21666875e-02 1.57161459e-01 7.73345172e-01
4.91971880e-01 -7.48524487e-01 4.43529904e-01 5.83060324e-01
3.21248472e-01 -4.82023716e-01 -1.04572880e+00 -1.04284847e+00
-9.11432877e-03 -3.86188239e-01 5.13012409e-01 1.08659935e+00
1.03989713e-01 -2.16628298e-01 -5.24854720e-01 6.31186306e-01
7.45910585e-01 3.79583776e-01 7.01797664e-01 -1.45342362e+00
-3.77094075e-02 -2.87897646e-01 -8.76860440e-01 -3.80092025e-01
5.37171304e-01 -7.56974995e-01 -1.72151085e-02 -1.38133931e+00
2.82141924e-01 -7.71636844e-01 -5.60937166e-01 1.10439098e+00
-5.48099160e-01 1.35285288e-01 -1.31279841e-01 2.50016659e-01
-4.91114616e-01 3.60894501e-01 4.99045908e-01 -1.18326470e-01
-6.10497653e-01 4.33195084e-01 -5.02050757e-01 7.07859278e-01
9.35227156e-01 -5.57150185e-01 -2.51867324e-01 1.02206297e-01
-3.73272896e-02 -3.68067235e-01 4.27165210e-01 -1.19459355e+00
-3.31294313e-02 -7.62085140e-01 4.09935832e-01 -6.31417632e-01
-9.04237404e-02 -1.12231255e+00 2.24064231e-01 6.47528887e-01
-2.70730287e-01 -1.96108893e-01 4.34449911e-01 4.16828215e-01
1.18472986e-02 -3.38517457e-01 7.64026999e-01 1.65621396e-02
-1.33190310e+00 9.57342163e-02 -7.26870358e-01 -9.03914690e-01
1.40771973e+00 -1.33842311e-03 2.46971563e-01 1.28566578e-01
-7.35688090e-01 2.97497988e-01 3.96608531e-01 3.52158666e-01
5.34634888e-01 -1.52996945e+00 -8.23433101e-01 5.23842931e-01
6.39283419e-01 -6.44080579e-01 -2.97475219e-01 4.06711757e-01
-5.47120035e-01 3.95713896e-01 -3.32290858e-01 -4.28060710e-01
-1.50992405e+00 2.76624411e-01 4.45730954e-01 -2.31276289e-01
-5.07904172e-01 3.72465521e-01 -2.91589886e-01 -4.86505419e-01
-2.67590135e-01 -4.75240976e-01 -4.91807044e-01 3.07179093e-01
6.87898338e-01 3.77842933e-01 1.95496008e-02 -8.54854763e-01
-5.41339397e-01 5.86105943e-01 -1.35178432e-01 -1.35532990e-01
1.47074580e+00 2.85415957e-03 -2.09766731e-01 7.94046819e-01
1.26520157e+00 -3.88302118e-01 -8.42042506e-01 -5.36359847e-01
2.32445627e-01 -6.78620100e-01 2.65115559e-01 -4.20546383e-01
-5.32292008e-01 2.77230114e-01 8.40669155e-01 1.17543563e-01
1.25766516e+00 1.30732059e-02 8.99719819e-02 1.11154902e+00
5.82843959e-01 -9.81139958e-01 -5.57053626e-01 2.81661689e-01
5.87170720e-01 -1.49511659e+00 2.08726004e-01 -5.21610439e-01
-2.89253503e-01 1.45964336e+00 2.77745247e-01 -2.47159019e-01
1.49597132e+00 1.34125799e-01 2.29479268e-01 5.86267672e-02
-4.22814488e-01 -9.26909387e-01 2.45118514e-01 5.06561935e-01
9.66868401e-02 5.29037893e-01 -4.35978383e-01 8.51630941e-02
2.62905568e-01 6.79750890e-02 2.13498071e-01 1.33183718e+00
-8.22495162e-01 -1.14803553e+00 -2.92780191e-01 7.86643863e-01
4.35109586e-02 -1.24850072e-01 -2.74790406e-01 7.82382786e-01
-4.23942618e-02 7.08120227e-01 -3.89903903e-01 -4.23100471e-01
2.16520026e-01 2.69212842e-01 4.46945101e-01 -5.48612654e-01
-2.28811368e-01 -9.25294682e-03 -1.18942626e-01 -2.98392385e-01
-8.74391377e-01 -9.00655806e-01 -9.79007244e-01 -8.22852366e-03
-5.81365228e-01 2.24525735e-01 4.30258125e-01 1.24386775e+00
5.30198552e-02 -3.45313787e-01 9.94199216e-01 -8.02683592e-01
-5.12033343e-01 -8.54181945e-01 -8.72531950e-01 6.96761310e-02
7.33298779e-01 -6.42259836e-01 -3.14301729e-01 1.07332766e-01] | [7.97214412689209, 3.939903974533081] |
b0cddb37-76ae-4bd2-8c72-787c87d866d5 | three-dimensional-reconstruction-of-human | null | null | http://openaccess.thecvf.com/content_CVPR_2020/html/Fieraru_Three-Dimensional_Reconstruction_of_Human_Interactions_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Fieraru_Three-Dimensional_Reconstruction_of_Human_Interactions_CVPR_2020_paper.pdf | Three-Dimensional Reconstruction of Human Interactions | Understanding 3d human interactions is fundamental for fine grained scene analysis and behavioural modeling. However, most of the existing models focus on analyzing a single person in isolation, and those who process several people focus largely on resolving multi-person data association, rather than inferring interactions. This may lead to incorrect, lifeless 3d estimates, that miss the subtle human contact aspects--the essence of the event--and are of little use for detailed behavioral understanding. This paper addresses such issues and makes several contributions: (1) we introduce models for interaction signature estimation (ISP) encompassing contact detection, segmentation, and 3d contact signature prediction; (2) we show how such components can be leveraged in order to produce augmented losses that ensure contact consistency during 3d reconstruction; (3) we construct several large datasets for learning and evaluating 3d contact prediction and reconstruction methods; specifically, we introduce CHI3D, a lab-based accurate 3d motion capture dataset with 631 sequences containing 2,525 contact events, 728,664 ground truth 3d poses, as well as FlickrCI3D, a dataset of 11,216 images, with 14,081 processed pairs of people, and 81,233 facet-level surface correspondences within 138,213 selected contact regions. Finally, (4) we present models and baselines to illustrate how contact estimation supports meaningful 3d reconstruction where essential interactions are captured. Models and data are made available for research purposes at http://vision.imar.ro/ci3d.
| [' Cristian Sminchisescu', ' Vlad Olaru', ' Alin-Ionut Popa', ' Elisabeta Oneata', ' Mihai Zanfir', 'Mihai Fieraru'] | 2020-06-01 | null | null | null | cvpr-2020-6 | ['contact-detection'] | ['robots'] | [ 4.93292809e-02 -3.13982755e-01 -1.08341217e-01 -3.03827673e-01
-5.99133253e-01 -3.98001075e-01 5.20539403e-01 -1.92650527e-01
-4.48052555e-01 3.69416833e-01 5.86128891e-01 2.03873530e-01
-1.38090536e-01 -4.75430876e-01 -7.43224561e-01 -1.42911196e-01
-3.26321304e-01 8.97876799e-01 2.73028463e-01 -9.40001383e-02
2.66148113e-02 6.14869297e-01 -1.55663240e+00 1.43360808e-01
5.79223216e-01 8.00521970e-01 -2.94810236e-01 9.33387220e-01
5.44482708e-01 2.42620915e-01 -4.49504822e-01 -4.88487065e-01
3.78185064e-01 -3.02098632e-01 -6.23110354e-01 3.51603806e-01
9.27119315e-01 -7.80145407e-01 -4.52256233e-01 3.86825502e-01
6.36634827e-01 2.00573936e-01 9.23501611e-01 -1.49706793e+00
-9.71520171e-02 -1.55884132e-01 -7.50255406e-01 5.70514798e-02
9.90265310e-01 5.48059940e-01 8.60109568e-01 -8.68953645e-01
8.02289486e-01 1.48629904e+00 1.08540666e+00 5.68256795e-01
-1.32613266e+00 -4.94130194e-01 4.70082238e-02 1.57255784e-01
-1.43882525e+00 -5.91098070e-01 6.88815355e-01 -6.91288352e-01
1.06359303e+00 3.86129856e-01 1.22413373e+00 1.63901019e+00
-7.73913274e-03 8.71038198e-01 8.51935148e-01 -6.67925850e-02
-2.40905032e-01 -1.43029481e-01 4.84248161e-01 6.67234242e-01
4.65313941e-01 1.79177940e-01 -7.14922607e-01 -4.63302553e-01
1.01937675e+00 -4.38423194e-02 -2.85150930e-02 -4.59989399e-01
-1.25696969e+00 3.55183452e-01 1.33949116e-01 -1.64301500e-01
-3.71962547e-01 1.38625592e-01 1.72831863e-01 2.85145119e-02
4.20002401e-01 -1.00277081e-01 -1.89689413e-01 -4.15122926e-01
-6.62976861e-01 1.09483194e+00 7.83670008e-01 1.06304801e+00
6.84170902e-01 -4.01298493e-01 -1.66792050e-01 7.81842828e-01
1.68294281e-01 7.68940568e-01 -3.86758029e-01 -1.43833280e+00
4.44906741e-01 6.70392811e-01 4.75676864e-01 -1.29168677e+00
-6.84970140e-01 8.13283473e-02 -6.10663354e-01 5.73743135e-03
5.13823271e-01 -1.96280673e-01 -5.20845413e-01 1.56773400e+00
4.95983660e-01 3.19630891e-01 -6.77884042e-01 1.06144500e+00
8.07348311e-01 4.97816913e-02 -8.72312412e-02 2.93129146e-01
1.29834318e+00 -6.80385888e-01 -3.75935942e-01 -5.10296702e-01
4.82240915e-01 -5.09075642e-01 1.07390904e+00 1.12567440e-01
-1.35173547e+00 -6.65986180e-01 -6.84579849e-01 -2.14594051e-01
-5.02615422e-02 -5.30324765e-02 6.73879623e-01 4.12982196e-01
-8.82293999e-01 3.88142973e-01 -7.49478459e-01 -8.00676167e-01
5.23624420e-01 3.56092066e-01 -3.94773871e-01 -1.51797254e-02
-9.15559888e-01 9.81029987e-01 -5.05899549e-01 7.96744078e-02
-6.84855700e-01 -7.15045571e-01 -8.09712052e-01 -5.44666111e-01
3.47793788e-01 -1.21111751e+00 1.04824519e+00 -3.96545559e-01
-7.05977738e-01 1.27307856e+00 -5.00058770e-01 -1.98467091e-01
9.65079725e-01 -7.53757656e-01 -1.69498935e-01 1.94890797e-01
3.02676886e-01 8.23427856e-01 4.33491141e-01 -1.46070433e+00
-4.05967861e-01 -5.87203205e-01 -2.02801883e-01 5.40253758e-01
2.54849911e-01 1.57074586e-01 -8.02655816e-01 -5.07352710e-01
-9.95854512e-02 -1.04417288e+00 -2.17809677e-01 3.37340713e-01
-6.45045459e-01 -2.82196980e-02 6.08510196e-01 -8.74367237e-01
7.95480847e-01 -1.87195957e+00 1.97692454e-01 3.18431973e-01
4.58916068e-01 2.02706866e-02 -1.66837927e-02 5.04780531e-01
3.64983618e-01 -9.77724092e-04 -2.38026917e-01 -9.12139237e-01
1.30222693e-01 -6.37260079e-02 5.08368798e-02 7.39144027e-01
-1.62151847e-02 1.25297844e+00 -5.31260848e-01 -5.19654870e-01
6.60144746e-01 5.88690400e-01 -7.91735649e-01 2.69601315e-01
9.85472351e-02 5.99420428e-01 -4.01354671e-01 8.68032455e-01
5.99311829e-01 -3.57564241e-02 -1.22006789e-01 -2.79579818e-01
8.39141831e-02 -5.94718866e-02 -1.34393418e+00 1.51666820e+00
5.14089204e-02 5.10285378e-01 1.96995959e-01 -4.50419605e-01
5.62161267e-01 -4.82285097e-02 1.02877462e+00 -3.58295739e-01
2.82877177e-01 -1.65921494e-01 -4.24219310e-01 -4.61660743e-01
4.75407422e-01 3.30077320e-01 -3.77968401e-01 5.79576969e-01
-3.75041723e-01 -1.17085710e-01 -2.10467666e-01 1.51887596e-01
1.10488486e+00 2.96745509e-01 3.63113768e-02 -3.80210066e-03
1.77682023e-02 2.67099440e-01 4.70282108e-01 8.73589098e-01
-5.87994874e-01 7.30053842e-01 1.62290171e-01 -4.25634801e-01
-1.10116017e+00 -1.49218667e+00 -1.26688004e-01 5.19165099e-01
5.55415034e-01 -4.00602669e-01 -8.54280591e-01 -3.26900870e-01
3.67778033e-01 3.78748655e-01 -5.91027200e-01 -7.84922540e-02
-7.40291119e-01 -5.08490741e-01 7.92448521e-01 6.40028059e-01
7.86546946e-01 -8.38378310e-01 -5.59157133e-01 -1.80387959e-01
-7.66178608e-01 -1.21620834e+00 -7.32429504e-01 -5.64144075e-01
-6.45261586e-01 -1.47894406e+00 -5.21593273e-01 -3.97340626e-01
5.54025650e-01 6.87583923e-01 1.33517385e+00 2.61867106e-01
-6.27365112e-01 1.12279224e+00 -1.92316115e-01 -1.46661088e-01
1.38151035e-01 -3.91088754e-01 5.92269659e-01 -9.96434391e-02
8.59851897e-01 -6.00628614e-01 -6.05673969e-01 7.47667015e-01
3.78722325e-02 5.55481948e-02 2.30996951e-01 4.55460817e-01
5.48574746e-01 -1.73225015e-01 2.42774822e-02 -5.86833596e-01
4.67136323e-01 -2.69416124e-01 -4.95155640e-02 -1.32381236e-02
-9.31664705e-02 -6.77779436e-01 -1.76307976e-01 -3.59597206e-01
-1.10779846e+00 -1.58751849e-02 1.04812477e-02 -4.89730686e-01
-6.64425373e-01 -2.12011516e-01 -2.05294237e-01 3.31264399e-02
8.54295015e-01 -4.41246778e-02 1.49811432e-01 -5.21395802e-01
1.74211040e-01 5.74745893e-01 6.88350677e-01 -7.40559161e-01
7.70519972e-01 8.10819864e-01 -1.44254625e-01 -1.23426640e+00
-5.43813825e-01 -6.98542774e-01 -1.23807418e+00 -5.92339635e-01
1.02984619e+00 -1.10224092e+00 -1.16550994e+00 9.58076060e-01
-9.96684611e-01 -7.55922616e-01 -1.54670849e-01 4.73934770e-01
-8.54431570e-01 3.65792662e-01 -6.94886506e-01 -1.06093991e+00
4.32408191e-02 -7.35015988e-01 1.43906605e+00 -6.59953207e-02
-1.03639710e+00 -8.00836027e-01 1.16517305e-01 9.77087855e-01
-1.12704791e-01 4.49762017e-01 4.97482449e-01 -1.44637063e-01
-5.76008379e-01 -3.32545131e-01 -1.40618801e-01 -1.40447021e-01
5.32014854e-02 -1.04228862e-01 -8.50629449e-01 -1.32993743e-01
-2.89483249e-01 -3.51963222e-01 6.60782576e-01 7.37316072e-01
7.55875051e-01 -1.38349924e-03 -6.49126709e-01 4.61142480e-01
6.66415155e-01 -3.19381386e-01 6.49633884e-01 -7.77541101e-02
1.08160913e+00 1.06070137e+00 7.38567233e-01 5.73961318e-01
9.00189579e-01 1.09040928e+00 4.86390218e-02 -1.01493284e-01
-2.54715562e-01 -5.31563461e-01 3.10380846e-01 3.47659767e-01
-6.47601783e-01 2.49705035e-02 -8.71970057e-01 4.73501712e-01
-2.02160907e+00 -1.27072442e+00 -5.20775080e-01 2.27314305e+00
4.89107519e-01 1.43175259e-01 8.23085785e-01 -5.71249314e-02
6.42719984e-01 1.69649020e-01 -6.07354283e-01 3.13954800e-01
-1.25440404e-01 -1.97516024e-01 4.75272000e-01 8.18020582e-01
-1.10874748e+00 9.14770186e-01 6.60189152e+00 3.70033532e-01
-1.24072149e-01 -1.19538158e-01 3.69402289e-01 -5.03139675e-01
-1.89490139e-01 -1.45662040e-01 -9.39021587e-01 2.53248990e-01
4.96167600e-01 2.94303834e-01 3.93934101e-01 5.65285623e-01
6.98282540e-01 -4.80691642e-01 -1.24136150e+00 1.37816608e+00
1.75891116e-01 -1.00918007e+00 -9.51717347e-02 4.23327476e-01
5.30952632e-01 -2.08215460e-01 -2.28099197e-01 -4.00493070e-02
3.57472390e-01 -1.10347688e+00 8.05286229e-01 9.92797554e-01
6.51861727e-01 -6.11751795e-01 4.16713595e-01 5.35924673e-01
-1.44012141e+00 1.83970988e-01 -7.37511292e-02 -2.96852320e-01
5.93477011e-01 4.16655958e-01 -4.33963776e-01 3.81493270e-01
1.22136402e+00 1.07433391e+00 -5.99766910e-01 8.08239698e-01
-3.94033156e-02 3.42705101e-01 -5.28581023e-01 7.90197253e-02
-4.06744033e-01 -2.65902698e-01 8.65893841e-01 1.02934825e+00
-1.15070671e-01 3.55921090e-01 3.54664654e-01 9.62649226e-01
2.88152903e-01 -4.15040672e-01 -9.18687105e-01 2.84662068e-01
7.07444489e-01 9.02119279e-01 -5.20515442e-01 -9.90542248e-02
-4.21077728e-01 1.20186567e+00 2.39028096e-01 5.43698847e-01
-9.61862564e-01 2.53740340e-01 1.29003537e+00 7.35299826e-01
-2.48908728e-01 -6.78992987e-01 -5.01998663e-01 -1.20509350e+00
1.71643615e-01 -7.29150653e-01 2.31660783e-01 -7.59681165e-01
-1.53505754e+00 -8.75427127e-02 6.11366451e-01 -1.10775828e+00
-3.29794258e-01 -4.27206904e-01 -4.62476730e-01 7.34401882e-01
-5.74663877e-01 -1.47716069e+00 -6.82381630e-01 7.36236453e-01
4.37363595e-01 3.28231066e-01 4.14547563e-01 4.26420301e-01
-4.71671700e-01 6.13353908e-01 -7.12422192e-01 1.99555546e-01
7.48623669e-01 -9.44064796e-01 8.26874673e-01 5.09762526e-01
-1.81134585e-02 6.32156551e-01 6.00833714e-01 -1.15604889e+00
-1.49338758e+00 -9.66225922e-01 9.91389751e-01 -1.42732179e+00
1.84778214e-01 -7.31776834e-01 -6.46398544e-01 1.06078088e+00
-4.19968784e-01 -2.49290302e-01 6.17712140e-01 2.52092123e-01
-6.77699863e-04 2.57100701e-01 -1.29547143e+00 1.11000896e+00
2.18794155e+00 -5.34487605e-01 -4.94613886e-01 6.47531673e-02
3.52806062e-01 -5.31533837e-01 -8.20132315e-01 3.19280446e-01
1.01201665e+00 -1.23903346e+00 1.54120493e+00 -4.71696466e-01
2.12890338e-02 -8.20896551e-02 -3.09963107e-01 -7.84712017e-01
-3.58655810e-01 -3.76920253e-01 -2.61072189e-01 8.95052850e-01
-1.25410020e-01 -3.81335050e-01 1.06495512e+00 1.11801386e+00
-5.46041131e-02 -7.24452376e-01 -7.16554403e-01 -6.20423257e-01
-2.66489387e-01 -8.52606952e-01 3.88743728e-01 7.36951828e-01
-4.37288314e-01 1.45333797e-01 -7.40164757e-01 4.41579670e-02
1.08491099e+00 -1.60607874e-01 1.51084435e+00 -1.38561618e+00
-3.50578278e-01 -3.14721972e-01 -3.87959778e-01 -1.44714010e+00
1.38801336e-01 -3.24345797e-01 -4.89721373e-02 -1.54267681e+00
4.39604938e-01 -4.90139127e-01 5.61858237e-01 3.02209437e-01
-2.24634260e-03 3.88548285e-01 -4.04630154e-02 4.87137318e-01
-5.71373582e-01 5.00256777e-01 1.30619931e+00 -1.77623592e-02
-2.62930661e-01 2.03832731e-01 -5.14767051e-01 1.00598192e+00
4.40830022e-01 2.99469884e-02 -3.72996867e-01 -2.02060908e-01
-2.03441918e-01 -1.09132104e-01 1.40920496e+00 -1.10643804e+00
1.02705054e-01 -3.77762794e-01 8.59714508e-01 -9.54324126e-01
9.57086802e-01 -7.26493955e-01 5.42292118e-01 3.47424358e-01
-1.65390089e-01 -1.35739386e-01 9.98430848e-02 5.73821008e-01
4.33624059e-01 2.52255678e-01 4.13079798e-01 -2.98710793e-01
-7.88318872e-01 5.45863092e-01 -2.86862642e-01 2.92315632e-01
1.14511776e+00 -7.78360605e-01 -8.80192667e-02 -5.62615097e-01
-6.04302943e-01 4.13051039e-01 8.84149015e-01 5.27966261e-01
6.86814606e-01 -1.44393647e+00 -7.15467274e-01 2.52449512e-01
4.00468633e-02 3.74698415e-02 5.96072853e-01 8.24796855e-01
-2.19772309e-01 1.84798509e-01 -1.13633446e-01 -8.23513031e-01
-1.46234858e+00 -9.45264753e-03 4.21760917e-01 2.50316709e-01
-8.49203289e-01 7.70317972e-01 9.99960974e-02 -6.49458051e-01
2.38976821e-01 -4.33396176e-02 1.23388276e-01 -1.57991678e-01
4.51001734e-01 8.38190913e-01 -3.47516537e-01 -1.16460383e+00
-6.70894027e-01 1.08547497e+00 3.01292837e-01 -3.00828308e-01
1.05814159e+00 -4.81962979e-01 2.12560147e-01 5.22643864e-01
9.00042474e-01 2.64797937e-02 -1.66153729e+00 -9.23511162e-02
-2.99287677e-01 -7.12775648e-01 -5.18488407e-01 -5.72217822e-01
-4.53381658e-01 6.54568017e-01 4.05318409e-01 -2.91715205e-01
6.61289394e-01 4.69618738e-01 1.00982237e+00 1.75664335e-01
7.96178043e-01 -1.06820941e+00 2.42669955e-01 3.75494778e-01
8.43493700e-01 -1.17826998e+00 1.91456556e-01 -7.03624725e-01
-7.37080872e-01 3.89683247e-01 9.23941135e-01 -1.76738948e-01
6.38506770e-01 1.69240624e-01 -2.39942610e-01 -4.99443382e-01
-5.07373095e-01 -1.97168931e-01 4.07529682e-01 1.00779939e+00
3.57634239e-02 7.74004981e-02 1.59653023e-01 8.05722475e-01
-3.76351863e-01 -1.23259671e-01 1.48413450e-01 8.60739410e-01
-1.89549819e-01 -9.30890679e-01 -4.97125387e-01 6.65872097e-01
1.31487221e-01 3.20564538e-01 -8.12820196e-01 8.57068896e-01
1.53352663e-01 1.04480898e+00 3.14404070e-01 -8.01608562e-01
7.31950283e-01 -2.02257201e-01 7.86790371e-01 -4.82287705e-01
-4.19381440e-01 -1.35422185e-01 5.88412285e-01 -9.50992882e-01
-4.62848127e-01 -1.16569304e+00 -1.19365966e+00 -1.03560627e+00
5.92215769e-02 -3.71792287e-01 1.83558483e-02 8.96201432e-01
5.18900573e-01 1.84151679e-02 3.39944750e-01 -1.45712721e+00
-1.38806760e-01 -7.11234987e-01 -5.99189103e-01 6.90090895e-01
1.64239421e-01 -1.12587357e+00 -4.04101878e-01 1.96867093e-01] | [7.0147294998168945, -0.9650317430496216] |
0c7ecff7-9397-4db8-8583-a01a3428e884 | causalvlr-a-toolbox-and-benchmark-for-visual | 2306.17462 | null | https://arxiv.org/abs/2306.17462v1 | https://arxiv.org/pdf/2306.17462v1.pdf | CausalVLR: A Toolbox and Benchmark for Visual-Linguistic Causal Reasoning | We present CausalVLR (Causal Visual-Linguistic Reasoning), an open-source toolbox containing a rich set of state-of-the-art causal relation discovery and causal inference methods for various visual-linguistic reasoning tasks, such as VQA, image/video captioning, medical report generation, model generalization and robustness, etc. These methods have been included in the toolbox with PyTorch implementations under NVIDIA computing system. It not only includes training and inference codes, but also provides model weights. We believe this toolbox is by far the most complete visual-linguitic causal reasoning toolbox. We wish that the toolbox and benchmark could serve the growing research community by providing a flexible toolkit to re-implement existing methods and develop their own new causal reasoning methods. Code and models are available at https://github.com/HCPLab-SYSU/Causal-VLReasoning. The project is under active development by HCP-Lab's contributors and we will keep this document updated. | ['Liang Lin', 'Guanbin Li', 'Weixing Chen', 'Yang Liu'] | 2023-06-30 | null | null | null | null | ['visual-question-answering', 'video-captioning', 'causal-inference', 'medical-report-generation', 'causal-inference'] | ['computer-vision', 'computer-vision', 'knowledge-base', 'medical', 'miscellaneous'] | [-2.51718342e-01 4.11659092e-01 -5.59227169e-01 -4.13482368e-01
-4.97111589e-01 -5.32836556e-01 7.99835086e-01 2.86671566e-03
2.40592167e-01 7.11894155e-01 9.13828969e-01 -7.42621839e-01
-9.38595384e-02 -5.48045218e-01 -6.27741933e-01 -3.25841963e-01
-7.91881084e-02 4.23046380e-01 1.34770721e-01 -7.91571438e-02
2.56447881e-01 2.93069899e-01 -1.56595337e+00 9.09116209e-01
4.89342213e-01 4.86491621e-01 -6.71956092e-02 8.79128635e-01
1.09512992e-02 1.58651781e+00 -3.06008458e-01 -6.58587158e-01
-5.15825748e-01 -4.30432826e-01 -1.11550081e+00 -6.14965856e-01
5.59709847e-01 -1.35982379e-01 -4.89801586e-01 9.71809745e-01
5.36734581e-01 -1.06548935e-01 6.56255960e-01 -1.62541604e+00
-1.34291220e+00 9.13393497e-01 -5.39611518e-01 5.79182327e-01
8.91949654e-01 3.16042781e-01 9.76504862e-01 -7.18863010e-01
9.98102665e-01 1.95137441e+00 4.12959635e-01 8.52697551e-01
-9.32679713e-01 -8.02556217e-01 3.60132337e-01 8.54966819e-01
-1.20235813e+00 -3.68171424e-01 6.04482591e-01 -7.25820303e-01
1.20328093e+00 6.35067165e-01 5.77779591e-01 1.78436220e+00
1.58352748e-01 9.93313551e-01 1.14300334e+00 -2.09717989e-01
-2.79807542e-02 -1.91942155e-01 7.97084644e-02 1.22030842e+00
-2.09367737e-01 4.38027054e-01 -8.28378856e-01 -4.18387622e-01
8.98338735e-01 -5.82278848e-01 -2.17500284e-01 -1.26933053e-01
-1.56041741e+00 9.98406291e-01 5.12986600e-01 8.29675123e-02
-1.14691764e-01 8.37142408e-01 4.28754061e-01 -1.14980442e-02
3.41891438e-01 -6.66467845e-02 -1.08071499e-01 -1.36005476e-01
-4.13547903e-01 4.84998941e-01 4.98139352e-01 8.14619660e-01
2.10417638e-04 1.95136696e-01 -6.03044569e-01 7.20026255e-01
6.25331759e-01 4.85393316e-01 9.86447185e-02 -1.23909473e+00
4.61136997e-02 3.50664228e-01 9.42991860e-03 -9.27954257e-01
-4.85764772e-01 2.49315351e-01 -6.57337129e-01 1.49253830e-01
1.13403052e-01 -2.31016055e-01 -8.07399750e-01 1.51095831e+00
3.91461462e-01 4.35400844e-01 -1.28110170e-01 1.17818773e+00
1.59624577e+00 5.29792249e-01 6.47151709e-01 -2.36637175e-01
1.51230931e+00 -7.32251167e-01 -1.06725335e+00 4.74001840e-02
3.67048204e-01 -9.69562769e-01 1.17379832e+00 3.90964150e-01
-1.09551167e+00 -2.36301631e-01 -9.08843458e-01 -3.51042718e-01
-3.24532837e-01 9.20079201e-02 1.19749129e+00 4.16006088e-01
-1.04742181e+00 2.83525378e-01 -7.11914957e-01 -5.70924819e-01
5.08243680e-01 -2.08248451e-01 -2.34937310e-01 -6.67559654e-02
-1.69371200e+00 1.01293886e+00 1.62776962e-01 -1.29708603e-01
-1.31530070e+00 -9.19512153e-01 -8.02194595e-01 -4.74730641e-01
3.66422325e-01 -1.11667705e+00 1.28042376e+00 -5.49781382e-01
-1.02838588e+00 9.34511721e-01 -1.82851031e-01 -2.89725333e-01
7.33125865e-01 -2.94746101e-01 -5.96052706e-01 4.98359092e-03
-2.38150358e-02 7.58786857e-01 6.53320074e-01 -1.26812100e+00
-3.16627562e-01 -1.58310920e-01 1.00385472e-01 2.28275999e-01
1.64724782e-01 6.38213098e-01 -4.96532261e-01 -6.51879251e-01
-7.34033823e-01 -5.65873444e-01 -7.28992969e-02 1.60172820e-01
-6.81241393e-01 -5.97657561e-01 8.93377066e-01 -5.58657706e-01
1.17778671e+00 -1.81353319e+00 7.51452148e-02 -2.31395900e-01
3.49329174e-01 1.15137920e-01 -1.39977545e-01 4.05726105e-01
-6.56417191e-01 4.19877946e-01 1.49270982e-01 1.84899122e-02
-1.44407395e-02 1.75199613e-01 -6.89004958e-01 6.10900044e-01
2.99439609e-01 1.16480935e+00 -1.34880197e+00 -9.29633379e-01
7.58295894e-01 6.83989167e-01 -2.53684312e-01 1.59481749e-01
-6.48748279e-01 4.98401582e-01 -3.64535481e-01 5.65627933e-01
3.30147624e-01 -3.75087708e-01 4.30632800e-01 -3.72171640e-01
-2.04149112e-01 2.83278257e-01 -1.08829618e+00 1.60397005e+00
-2.04682097e-01 7.56496191e-01 -3.61358941e-01 -5.35230577e-01
6.47982538e-01 5.83843350e-01 1.63741395e-01 -5.43133914e-01
1.05766848e-01 -4.07783061e-01 -4.25797999e-02 -1.00470138e+00
1.67613670e-01 1.03139021e-01 9.01708454e-02 3.55127752e-01
-1.26100913e-01 6.43582344e-02 1.90985203e-01 6.15788579e-01
1.10441935e+00 4.29138541e-01 5.34662068e-01 -2.22904518e-01
3.45190823e-01 2.67369032e-01 2.51822352e-01 6.65024757e-01
-2.74071991e-01 3.85295630e-01 8.76321375e-01 -5.91704071e-01
-9.09544885e-01 -1.35353100e+00 -3.39044839e-01 1.02237356e+00
-8.65648761e-02 -7.03281105e-01 -4.61829662e-01 -5.69811761e-01
7.48316944e-02 1.43086469e+00 -9.63362515e-01 8.85320455e-02
-2.61772722e-01 -9.06663835e-01 8.56500030e-01 7.49970376e-01
1.34765536e-01 -1.35332048e+00 -4.85080361e-01 -1.89780653e-01
-3.61905903e-01 -8.18454564e-01 -9.88678187e-02 -3.91240001e-01
-3.93711686e-01 -1.21819556e+00 -2.34586775e-01 -1.83936834e-01
3.35021675e-01 -2.52220988e-01 1.32800245e+00 2.04593420e-01
-7.39975572e-01 5.74598312e-01 -3.16908240e-01 -4.78056461e-01
-5.53137302e-01 -7.72041976e-01 -1.61271784e-02 -4.50470179e-01
1.41219914e-01 -2.15493754e-01 -5.63546956e-01 9.93508846e-02
-4.72279429e-01 3.82432371e-01 -1.09899938e-01 6.25351846e-01
5.42708397e-01 -2.75986195e-01 2.86035389e-01 -9.89570796e-01
9.65516746e-01 -6.58928335e-01 -5.71222544e-01 4.15277541e-01
-4.19011295e-01 3.23482789e-02 1.39021367e-01 -2.53611445e-01
-1.47140503e+00 -2.08802834e-01 -3.29771340e-01 -6.05960608e-01
-2.19033867e-01 5.57454169e-01 1.60937324e-01 4.10756052e-01
9.76443946e-01 -3.66735697e-01 -3.27004999e-01 -1.93995938e-01
1.09560478e+00 3.31429750e-01 7.54612863e-01 -5.92600226e-01
3.34759951e-01 6.37262583e-01 -1.82824284e-01 -3.87143403e-01
-7.43274808e-01 -1.30265668e-01 -3.39782208e-01 -6.35867655e-01
1.18871152e+00 -9.20103848e-01 -1.24828470e+00 -2.88497172e-02
-1.55520129e+00 -6.75355613e-01 1.40217319e-01 2.10653141e-01
-4.22464043e-01 1.36545792e-01 -5.87265968e-01 -7.66300201e-01
-1.93398327e-01 -1.07023799e+00 9.32181418e-01 4.06027772e-02
-6.57921672e-01 -1.24751973e+00 1.71471730e-01 4.27233845e-01
-4.19803075e-02 4.82275337e-01 1.13892949e+00 -1.14566505e-01
-4.00847495e-01 2.95048207e-01 -5.73943555e-01 -2.01632991e-01
-1.42087892e-01 6.62678719e-01 -1.03388929e+00 3.20223093e-01
-6.06338263e-01 -5.65851331e-01 8.53906453e-01 6.42027020e-01
1.50869024e+00 -1.58790439e-01 -8.32806766e-01 4.16038483e-01
1.12437952e+00 1.33426979e-01 7.01509953e-01 -1.60514459e-01
1.11172175e+00 6.60495937e-01 7.13044167e-01 3.87773663e-01
8.00152600e-01 5.61841667e-01 7.07855284e-01 -2.48707592e-01
-7.45983303e-01 -2.34440386e-01 2.67582595e-01 4.51445639e-01
-4.37172621e-01 -4.15677905e-01 -1.35307574e+00 5.34592032e-01
-2.32379413e+00 -1.25332761e+00 -1.02495277e+00 1.61136341e+00
9.67204511e-01 -4.52539623e-01 -1.00419372e-01 -4.65526760e-01
5.85864127e-01 2.38027632e-01 -3.83892924e-01 -7.70341635e-01
-1.74752418e-02 -1.59700170e-01 2.24123985e-01 8.21843088e-01
-1.10876894e+00 1.18546510e+00 7.32373381e+00 8.75203311e-01
-9.23617005e-01 5.81526875e-01 6.44741774e-01 -1.07749283e-01
-6.66624427e-01 -3.98957543e-02 -5.98648608e-01 1.18494779e-01
1.08934939e+00 -3.40199202e-01 4.83128458e-01 7.53832996e-01
5.07521272e-01 -1.60627946e-01 -1.15848327e+00 9.24419880e-01
8.68231058e-02 -1.99654281e+00 7.12022558e-02 -4.19935942e-01
4.28292602e-01 9.88701284e-02 2.90542264e-02 6.04536273e-02
1.15356088e+00 -1.40898514e+00 8.02925885e-01 6.77475989e-01
1.02455485e+00 -5.94940305e-01 3.60087305e-01 -3.44365031e-01
-9.50836778e-01 9.43489820e-02 -2.79500514e-01 -6.89678192e-02
3.32888484e-01 5.58792472e-01 -7.32021332e-01 4.60237443e-01
1.13598061e+00 9.18642223e-01 -5.80270767e-01 6.65669441e-01
-7.54519999e-01 7.64289320e-01 3.93144548e-01 -1.86504424e-01
-2.61047930e-01 4.65365291e-01 6.59753799e-01 1.49353921e+00
-2.42260769e-01 1.94331214e-01 -5.96792810e-02 1.26512349e+00
1.86913207e-01 -6.03176057e-02 -7.81510234e-01 -9.27853584e-02
3.50830644e-01 1.07391715e+00 -5.42692304e-01 -4.88256067e-01
-3.54037970e-01 6.95091486e-01 5.04048049e-01 1.78459659e-01
-1.44595349e+00 2.88653672e-01 7.34672964e-01 -2.02972651e-01
-2.40273535e-01 2.48214472e-02 -3.94602925e-01 -1.02265060e+00
-6.78854108e-01 -8.61382663e-01 8.81646633e-01 -1.63217175e+00
-1.49578595e+00 4.58823115e-01 5.75315416e-01 -6.39219701e-01
-3.02686751e-01 -7.65599370e-01 -4.31184322e-01 7.43887126e-01
-1.08112216e+00 -1.33097243e+00 -3.55614960e-01 9.02648032e-01
3.80063772e-01 -4.75792494e-03 1.10290110e+00 2.23454595e-01
-6.39231563e-01 2.53420949e-01 -6.67007446e-01 1.59157097e-01
8.48539412e-01 -1.26804626e+00 4.62412149e-01 7.78954089e-01
1.60622261e-02 8.82609785e-01 9.91090775e-01 -1.12768567e+00
-1.21864605e+00 -1.16482103e+00 7.75393069e-01 -1.02447927e+00
1.35540843e+00 -4.30316120e-01 -6.48373783e-01 1.06294262e+00
8.92623842e-01 -1.43627552e-02 5.41526556e-01 3.79258901e-01
-5.75921834e-01 2.96842337e-01 -8.03658187e-01 1.11996877e+00
1.19139564e+00 -4.30033147e-01 -7.04163790e-01 9.05445635e-01
9.79685783e-01 -6.28939688e-01 -7.89806783e-01 2.86046505e-01
3.30944926e-01 -9.06364620e-01 1.17178380e+00 -4.82650101e-01
8.70004535e-01 -6.14277244e-01 1.17301293e-01 -1.11058950e+00
-4.60884720e-01 -5.44058442e-01 -1.70484021e-01 1.30190837e+00
4.18292880e-01 -4.36357349e-01 2.79773265e-01 3.82384598e-01
-6.25084192e-02 -4.55142766e-01 -8.16587865e-01 -3.54716390e-01
3.17124389e-02 -1.15731931e+00 3.72262806e-01 1.29479301e+00
1.78093910e-01 4.73828197e-01 -5.27354717e-01 3.19307327e-01
7.84711301e-01 -6.51527792e-02 4.51678514e-01 -8.00698102e-01
-1.14368588e-01 -3.94811958e-01 -9.84625369e-02 -1.09104507e-01
3.13063622e-01 -1.04781258e+00 -2.28081301e-01 -2.02541447e+00
3.13322455e-01 -1.94105878e-01 5.58743216e-02 1.18946373e+00
-2.70366728e-01 2.07166627e-01 8.11531320e-02 6.00198582e-02
-4.81451958e-01 2.84410447e-01 1.41610229e+00 -2.37002060e-01
1.49687409e-01 -4.90804046e-01 -6.78489804e-01 7.78620303e-01
6.79224372e-01 -4.92111713e-01 -7.16311693e-01 -3.96840453e-01
4.24667418e-01 2.58438468e-01 1.25058711e+00 -4.91963089e-01
5.28843626e-02 -5.67595780e-01 4.53783274e-01 -5.67236662e-01
1.78887054e-01 -2.12351292e-01 4.75893855e-01 4.56814975e-01
-4.76587892e-01 2.63941914e-01 5.07485211e-01 2.30230764e-01
8.51445794e-02 7.64355883e-02 4.94932711e-01 -2.37295687e-01
-1.14620602e+00 3.53825837e-03 -4.82555717e-01 -1.02073466e-02
9.81129467e-01 3.49236429e-01 -1.17826700e+00 -4.41606283e-01
-7.61863828e-01 3.64864022e-01 -2.48511638e-02 9.37654853e-01
7.43232250e-01 -1.46124184e+00 -9.32202101e-01 -5.84701657e-01
3.21793258e-01 -5.77086747e-01 5.28263152e-01 9.86681819e-01
-5.39846718e-01 5.09045362e-01 -1.59476981e-01 -4.90505964e-01
-1.49557221e+00 9.47221100e-01 1.81663811e-01 3.61379460e-02
-7.03328133e-01 1.08007920e+00 3.07383835e-01 -2.67998785e-01
1.07966594e-01 -1.52123451e-01 -3.91968131e-01 -9.41387489e-02
8.64425957e-01 1.57458812e-01 -3.37192714e-01 -4.01941478e-01
-7.95625925e-01 9.88746881e-02 8.19558576e-02 -3.66642587e-02
1.00559473e+00 1.21993542e-01 -4.27934110e-01 6.79703116e-01
6.87256575e-01 -9.72267389e-02 -1.03540766e+00 4.51049328e-01
-3.26813936e-01 -2.32556298e-01 2.39968672e-01 -1.47257018e+00
-8.74632299e-01 9.02374685e-01 7.87794173e-01 -9.17778537e-02
7.80331075e-01 4.54708040e-01 -1.30883500e-01 -3.03617656e-01
1.30442351e-01 -6.35043740e-01 1.46621794e-01 3.22020680e-01
1.64295018e+00 -9.98359501e-01 3.39406252e-01 -6.36447430e-01
-9.67004061e-01 9.07268822e-01 6.37680173e-01 -3.24235149e-02
5.31070352e-01 6.08210921e-01 4.37351406e-01 -6.27430916e-01
-1.37948835e+00 -2.96875805e-01 4.86168236e-01 1.03429723e+00
9.85549986e-01 5.39371312e-01 -3.48438144e-01 3.03711891e-01
-1.66064650e-01 2.11141005e-01 5.35130203e-01 2.98288137e-01
3.92828077e-01 -1.07452810e+00 -5.86752176e-01 1.78736106e-01
-2.83016920e-01 -6.15643680e-01 -4.62296903e-01 9.62896526e-01
2.45852634e-01 1.06592786e+00 1.70883656e-01 -3.66225839e-01
2.56991714e-01 -1.00073569e-01 7.30783463e-01 -4.78962511e-01
-4.13338691e-01 -4.87885997e-02 6.88240409e-01 -1.05217838e+00
-5.78156888e-01 -8.64412367e-01 -1.61443090e+00 -6.24316990e-01
3.81941348e-01 -4.59030002e-01 4.62758273e-01 6.63083851e-01
2.20444053e-01 9.52295423e-01 -3.66718531e-01 -4.31761175e-01
3.20284843e-01 -7.32933104e-01 1.84452515e-02 3.11247498e-01
-3.47021297e-02 -1.04222167e+00 -3.93946981e-03 4.03554291e-01] | [10.538769721984863, 1.6316741704940796] |
c267e0d2-ae8a-440d-bde5-ad0ba16921ee | robust-video-object-tracking-using-particle | 1509.08182 | null | http://arxiv.org/abs/1509.08182v1 | http://arxiv.org/pdf/1509.08182v1.pdf | Robust video object tracking using particle filter with likelihood based feature fusion and adaptive template updating | A robust algorithm solution is proposed for tracking an object in complex
video scenes. In this solution, the bootstrap particle filter (PF) is
initialized by an object detector, which models the time-evolving background of
the video signal by an adaptive Gaussian mixture. The motion of the object is
expressed by a Markov model, which defines the state transition prior. The
color and texture features are used to represent the object, and a marginal
likelihood based feature fusion approach is proposed. A corresponding object
template model updating procedure is developed to account for possible scale
changes of the object in the tracking process. Experimental results show that
our algorithm beats several existing alternatives in tackling challenging
scenarios in video tracking tasks. | ['Bin Liu', 'Yi Dai'] | 2015-09-28 | null | null | null | null | ['video-object-tracking'] | ['computer-vision'] | [ 1.80092245e-01 -4.98632997e-01 8.44937377e-03 4.74290438e-02
-1.42684162e-01 -3.49113673e-01 6.78953409e-01 -4.55359787e-01
-5.12133181e-01 6.87537193e-01 -4.66256261e-01 3.76022786e-01
9.52108130e-02 -5.15349209e-01 -3.66556555e-01 -1.09355497e+00
7.00691640e-02 5.35862803e-01 9.46123123e-01 4.05968964e-01
1.57207802e-01 7.71059573e-01 -1.58287132e+00 -9.53508988e-02
3.81327152e-01 1.08730638e+00 7.91085601e-01 1.14183462e+00
-6.96237907e-02 8.00049961e-01 -6.63508296e-01 -7.26232454e-02
3.07041377e-01 -3.49576324e-01 -1.10133834e-01 8.62880707e-01
3.91483247e-01 -2.43405253e-01 -3.20571125e-01 1.42309833e+00
3.21905799e-02 4.13596928e-01 6.78329229e-01 -1.32132363e+00
2.82340739e-02 -7.70345554e-02 -4.74410117e-01 5.12585461e-01
3.51843834e-01 -4.01140265e-02 3.07125688e-01 -1.08113086e+00
5.84465027e-01 1.47110105e+00 5.27357459e-01 3.87746423e-01
-9.34715986e-01 -4.27012652e-01 2.46000186e-01 3.17116886e-01
-1.46202815e+00 -2.58155197e-01 6.75454259e-01 -6.69124782e-01
3.90919238e-01 2.11579055e-01 1.05447793e+00 8.01445544e-01
8.54639888e-01 9.34733033e-01 9.97701287e-01 -4.52957869e-01
3.19349676e-01 6.23389985e-03 1.47944376e-01 7.16968715e-01
5.63215792e-01 3.55039269e-01 -4.12614048e-01 -6.19780123e-01
9.61470783e-01 8.29886943e-02 -7.76442438e-02 -6.82403684e-01
-8.56860578e-01 5.28939486e-01 -2.93039709e-01 1.12171285e-01
-6.53877676e-01 2.30077267e-01 -4.58492674e-02 -1.73282325e-01
3.10406685e-01 -4.68574524e-01 -1.58013523e-01 3.23934592e-02
-1.03519392e+00 5.38331270e-01 7.54882574e-01 1.01907158e+00
3.68678004e-01 2.17967629e-01 -4.84956503e-01 -8.99155345e-03
1.09817350e+00 1.05750561e+00 9.58869383e-02 -1.14501357e+00
-2.73809969e-01 3.49627554e-01 8.32949460e-01 -8.29110920e-01
-2.21765339e-02 -4.18030500e-01 -3.82243752e-01 6.85023606e-01
3.68080914e-01 3.85137647e-02 -9.36400950e-01 1.20018995e+00
8.99087548e-01 8.21558058e-01 -1.69069976e-01 6.67909324e-01
5.40321231e-01 6.65721476e-01 6.35856241e-02 -7.79041469e-01
1.51594603e+00 -6.90007687e-01 -1.30618370e+00 4.33791317e-02
-4.18081492e-01 -1.09923351e+00 -1.07997775e-01 6.21282399e-01
-1.11215782e+00 -1.09453285e+00 -7.80788541e-01 7.32031107e-01
1.06508605e-01 2.02879697e-01 1.04702525e-01 7.27991521e-01
-8.63547504e-01 3.43599886e-01 -1.27703094e+00 -3.87458891e-01
-3.67547222e-03 3.92940819e-01 1.10657714e-01 1.17338948e-01
-5.26704550e-01 9.59959924e-01 5.89179933e-01 4.25356090e-01
-1.15558684e+00 -1.55892611e-01 -5.56669533e-01 -2.73494691e-01
1.99027419e-01 -7.42841482e-01 1.12718666e+00 -9.79052246e-01
-1.68299794e+00 5.64315379e-01 -4.37141985e-01 -3.51505756e-01
7.58922279e-01 -3.62143129e-01 -4.68151003e-01 2.73906559e-01
-8.57060254e-02 1.00156151e-01 1.72100198e+00 -1.47642994e+00
-1.08683562e+00 -1.21647462e-01 -4.66963768e-01 3.52976292e-01
2.71627933e-01 5.45275927e-01 -8.93165886e-01 -6.59682453e-01
1.67476013e-01 -1.09039021e+00 -4.20233369e-01 8.81842747e-02
7.95093849e-02 -1.03208587e-01 1.37960279e+00 -5.81752777e-01
1.24496436e+00 -2.22837663e+00 9.32861343e-02 2.36238435e-01
7.24124163e-02 2.55733579e-01 2.39833221e-01 6.16345778e-02
4.24348503e-01 -9.45052087e-01 5.01992144e-02 -3.34583044e-01
-2.45742127e-01 3.03616107e-01 -2.02278778e-01 8.55385721e-01
-4.35346216e-02 4.63011801e-01 -7.75077760e-01 -7.82808006e-01
4.81226385e-01 7.73028612e-01 -1.60841674e-01 4.42959994e-01
-2.49242321e-01 6.76459730e-01 -7.73235857e-01 4.69624311e-01
1.14200962e+00 -9.23518166e-02 2.19013914e-02 -1.60629228e-01
-3.62100184e-01 -5.52438080e-01 -1.85905540e+00 1.06510413e+00
5.38254678e-01 3.49259764e-01 5.71122229e-01 -4.72649574e-01
9.18440521e-01 3.44396383e-01 6.52523458e-01 2.57451117e-01
3.69921327e-01 -2.13449657e-01 5.17875627e-02 -5.31202972e-01
5.22176743e-01 5.36384992e-02 4.04587418e-01 -4.23282236e-02
5.83938174e-02 3.38474326e-02 2.81207651e-01 -5.15906960e-02
9.34323251e-01 6.97281420e-01 4.25302178e-01 -2.28246316e-01
7.52794027e-01 -5.04668579e-02 9.01831746e-01 1.02496374e+00
-5.39920032e-01 2.85470754e-01 -5.64826787e-01 -4.49329853e-01
-5.69929898e-01 -1.24530923e+00 -2.87858486e-01 6.06988490e-01
4.94957477e-01 9.41453408e-03 -6.95657134e-01 -1.85694247e-01
-8.39520246e-02 2.55669475e-01 -4.13100094e-01 1.44417793e-01
-6.34835899e-01 -9.14380372e-01 -2.43478537e-01 3.24396878e-01
4.40090805e-01 -9.30966258e-01 -1.08990943e+00 7.81470716e-01
-1.40743190e-02 -1.33293176e+00 -2.96086371e-01 -2.23365918e-01
-1.05708742e+00 -1.17151427e+00 -5.32805502e-01 -5.52729726e-01
5.90296626e-01 3.37884724e-01 7.97140956e-01 1.30707502e-01
-4.07398134e-01 9.77193594e-01 -2.40962595e-01 -4.95094568e-01
-7.59719968e-01 -1.04739964e+00 2.86891699e-01 5.47394931e-01
2.91833192e-01 2.10931376e-01 -5.34923434e-01 4.74771321e-01
-6.27144635e-01 -2.28851452e-01 3.18132102e-01 5.96279025e-01
6.85536265e-01 4.74088967e-01 -3.73500586e-01 -2.11224213e-01
1.81401417e-01 -1.34163439e-01 -1.25209641e+00 4.09162879e-01
-1.08978085e-01 -4.97321904e-01 5.18875457e-02 -8.89463663e-01
-1.41589200e+00 5.33666134e-01 3.68762374e-01 -8.48782778e-01
-3.45649153e-01 -5.42227961e-02 -1.66736186e-01 -4.40350622e-01
7.69740045e-02 4.56836492e-01 1.37030423e-01 -4.48886424e-01
1.14437588e-01 2.63085485e-01 6.96429074e-01 -5.15652180e-01
1.05762827e+00 9.03728962e-01 2.68249571e-01 -1.03550684e+00
-6.08718455e-01 -7.18197167e-01 -7.51606584e-01 -9.43700433e-01
1.08867002e+00 -9.85902190e-01 -7.47900963e-01 7.67725170e-01
-1.21114266e+00 1.31163746e-01 -3.13932419e-01 1.20791447e+00
-7.23564625e-01 6.40611887e-01 -6.16392195e-01 -1.56415153e+00
5.21654449e-03 -1.03353298e+00 9.44568515e-01 5.00958681e-01
1.59243315e-01 -1.05411792e+00 2.80481488e-01 1.08586207e-01
2.28078470e-01 2.57121831e-01 5.19088171e-02 -3.52573097e-01
-1.16512871e+00 -4.70080972e-01 1.86053798e-01 3.03828567e-01
-6.89306762e-03 5.24727702e-01 -7.87933707e-01 -6.80499911e-01
5.20991325e-01 6.50569141e-01 5.43264568e-01 1.06284940e+00
4.73562241e-01 2.67994434e-01 -8.31964076e-01 3.04509789e-01
1.30917943e+00 7.71341503e-01 1.21137105e-01 2.48693585e-01
4.08815920e-01 1.37881890e-01 1.20457470e+00 8.17391157e-01
7.95615278e-03 7.87226439e-01 4.38325644e-01 1.24060959e-01
-1.44518256e-01 3.72570902e-01 6.21698797e-01 4.01282132e-01
-2.31396824e-01 -2.90602416e-01 -4.91670161e-01 2.12848466e-02
-2.08320260e+00 -1.24321628e+00 -4.62375760e-01 2.07829070e+00
1.03267960e-01 2.30798990e-01 5.39587066e-02 -2.04054728e-01
1.15202856e+00 -2.90584207e-01 -3.98046404e-01 5.17443240e-01
-9.41410065e-02 -1.54787049e-01 3.76161158e-01 5.91921389e-01
-1.34340608e+00 7.48312652e-01 7.32986164e+00 5.93885601e-01
-5.76667488e-01 1.89509451e-01 -2.29585499e-01 1.67247489e-01
5.81894398e-01 8.25030357e-02 -1.46740556e+00 5.81538796e-01
5.46134591e-01 -2.70832688e-01 -7.35297683e-04 6.64174795e-01
4.23062980e-01 -4.26827133e-01 -7.17079282e-01 1.01504195e+00
1.92546517e-01 -9.22311127e-01 -1.09566808e-01 2.47870058e-01
5.07291377e-01 -7.89270476e-02 7.45860711e-02 1.19383661e-02
4.34372067e-01 -8.95372555e-02 1.04923725e+00 1.14236343e+00
-9.46429744e-02 -4.72725868e-01 5.87663174e-01 5.02988696e-01
-1.66555965e+00 -2.63512611e-01 -4.01425213e-01 -3.57463919e-02
5.73283851e-01 4.34646308e-01 -6.14074647e-01 3.59986693e-01
8.14770222e-01 4.82156843e-01 -3.72402400e-01 1.79306304e+00
7.21949711e-02 4.82331991e-01 -4.83857661e-01 2.10416049e-01
-3.59137394e-02 -6.44011497e-01 1.29179740e+00 1.13760066e+00
5.70044219e-01 1.23132832e-01 9.10955608e-01 5.46573520e-01
5.90517044e-01 -5.50961867e-02 -2.96593666e-01 3.93283308e-01
3.14802408e-01 1.32817328e+00 -1.19636393e+00 -6.22245729e-01
-4.73464638e-01 9.19367552e-01 -1.92108586e-01 4.82571006e-01
-9.97045338e-01 4.08402115e-01 4.12578553e-01 -9.09747630e-02
9.23856914e-01 -3.16373348e-01 6.29341483e-01 -1.15345049e+00
-1.18025675e-01 -3.94025803e-01 4.68543947e-01 -8.24093223e-01
-1.14264369e+00 5.15425205e-01 2.70263940e-01 -1.58603740e+00
-1.18868612e-01 -7.62633622e-01 -6.53889239e-01 8.02009761e-01
-1.21937656e+00 -1.07783484e+00 -2.25168049e-01 7.47327030e-01
6.49715006e-01 -4.26780909e-01 4.26646113e-01 -2.85021085e-02
-4.86482829e-01 -3.80595356e-01 3.49690050e-01 -6.74160197e-02
3.17777932e-01 -8.93338501e-01 -3.64224799e-02 1.23482144e+00
4.75059561e-02 5.21426320e-01 1.21948063e+00 -1.19705105e+00
-1.57592928e+00 -9.54036534e-01 3.81433725e-01 -6.17282748e-01
7.42158115e-01 -2.36327201e-01 -8.97508502e-01 5.61260343e-01
1.94873288e-01 4.39906716e-01 6.00941896e-01 -5.95926762e-01
4.40969914e-01 2.53759325e-01 -1.06809151e+00 1.21703364e-01
5.24300039e-01 6.78038597e-02 -7.76864469e-01 1.95560932e-01
2.97572985e-02 -5.73899329e-01 -6.90396667e-01 5.72960317e-01
5.69487095e-01 -6.10648394e-01 9.51445878e-01 -4.39821810e-01
-1.02170265e+00 -1.06566155e+00 -2.22536996e-01 -6.56413198e-01
-8.67342174e-01 -6.26539588e-01 -8.66184771e-01 8.89522254e-01
-4.60934132e-01 -1.85316786e-01 9.00488794e-01 3.03230405e-01
2.16683820e-01 -8.53285119e-02 -1.23218775e+00 -1.01478493e+00
-7.35275149e-01 -3.43057513e-01 -3.00848465e-02 3.02319944e-01
-7.46096730e-01 -9.69248340e-02 -4.59382087e-01 6.35493457e-01
1.52061629e+00 2.00762138e-01 7.08305717e-01 -1.58419466e+00
-2.93849558e-01 -1.14081360e-01 -6.20590985e-01 -1.02801085e+00
3.43324423e-01 -2.30788484e-01 2.49636680e-01 -1.26638365e+00
4.84567463e-01 3.61008011e-02 -3.66244495e-01 -5.66021860e-01
-4.04162437e-01 1.20791942e-01 3.87179136e-01 2.37491190e-01
-8.75209153e-01 4.52131927e-01 1.00690269e+00 8.70698839e-02
1.14084750e-01 6.66151345e-01 4.35756922e-01 1.18241274e+00
3.52171510e-01 -6.72009647e-01 1.16674108e-02 8.09772983e-02
-4.00575429e-01 2.20375091e-01 5.22461832e-01 -1.27822447e+00
4.40970004e-01 -2.57328331e-01 8.28410506e-01 -1.21263492e+00
7.20743775e-01 -1.49310899e+00 8.24792325e-01 8.85473669e-01
4.12025303e-01 7.78433681e-02 -3.40786092e-02 1.02291477e+00
-1.11889750e-01 -5.10849059e-01 1.06263232e+00 -5.13624474e-02
-8.49461079e-01 5.42417586e-01 -9.74884391e-01 -5.84715366e-01
1.34909964e+00 -6.08672857e-01 3.23878467e-01 -1.17037587e-01
-1.27941453e+00 -2.95469668e-02 3.81305486e-01 3.65627080e-01
7.39515185e-01 -1.37491727e+00 -6.60235167e-01 3.43979150e-01
-3.00396055e-01 -4.95114058e-01 1.52164325e-01 8.36750686e-01
-2.49781370e-01 -5.81323653e-02 -1.51083708e-01 -9.62788820e-01
-1.80402875e+00 7.16558635e-01 3.56559575e-01 -1.57578468e-01
-6.13457918e-01 7.69167066e-01 1.82660192e-01 3.97725850e-01
2.80284345e-01 -3.12583447e-02 -4.98121947e-01 -6.50271401e-02
8.71573508e-01 5.52826941e-01 -5.16219735e-01 -1.23569679e+00
-2.69031674e-01 7.27549732e-01 1.32368907e-01 -1.51002437e-01
8.39156568e-01 -4.51552570e-01 -1.22966222e-01 6.08578265e-01
4.60029989e-01 -4.86622266e-02 -1.83925951e+00 -4.90777910e-01
8.48008022e-02 -8.12456906e-01 4.71579507e-02 -2.75079429e-01
-7.23761320e-01 2.01555178e-01 1.13880134e+00 1.19013205e-01
8.52602065e-01 -1.86040074e-01 2.00026199e-01 1.43315524e-01
6.14283383e-01 -9.34846103e-01 -9.33854878e-02 6.14297688e-01
5.76999724e-01 -9.52636361e-01 2.75974751e-01 -5.33846796e-01
-2.25916728e-01 1.14806998e+00 4.48952466e-01 -3.25426936e-01
1.03977692e+00 8.27383757e-01 1.86745562e-02 4.89899702e-02
-7.81411052e-01 -3.94922614e-01 3.96344125e-01 6.14713669e-01
-2.07065530e-02 -1.54052421e-01 -6.26480505e-02 8.37423205e-02
4.35684144e-01 2.03029856e-01 2.01420024e-01 1.31102073e+00
-9.56425548e-01 -8.71672094e-01 -1.22524691e+00 -1.67539343e-02
-5.93120575e-01 4.26574349e-01 2.18562961e-01 6.13914251e-01
2.75069714e-01 8.73363078e-01 1.89369097e-01 1.98739290e-01
2.32014269e-01 1.05018336e-02 8.76929998e-01 -3.03762943e-01
-2.50902057e-01 1.02530885e+00 -4.42736000e-01 -5.70824385e-01
-1.00718951e+00 -1.20210481e+00 -1.11154962e+00 3.91069114e-01
-6.25444531e-01 2.54949689e-01 5.87931275e-01 7.37610519e-01
-1.72713280e-01 3.52781445e-01 4.44545716e-01 -1.15136087e+00
-3.30171913e-01 -7.88241625e-01 -8.05952907e-01 1.03748053e-01
3.15388620e-01 -1.19662952e+00 -3.06054950e-01 5.05235910e-01] | [6.592889785766602, -2.0042612552642822] |
d8d234a4-1d18-457c-9f2c-f0f68e4b3945 | regret-guarantees-for-adversarial-online | 2302.05765 | null | https://arxiv.org/abs/2302.05765v1 | https://arxiv.org/pdf/2302.05765v1.pdf | Regret Guarantees for Adversarial Online Collaborative Filtering | We investigate the problem of online collaborative filtering under no-repetition constraints, whereby users need to be served content in an online fashion and a given user cannot be recommended the same content item more than once. We design and analyze a fully adaptive algorithm that works under biclustering assumptions on the user-item preference matrix, and show that this algorithm exhibits an optimal regret guarantee, while being oblivious to any prior knowledge about the sequence of users, the universe of items, as well as the biclustering parameters of the preference matrix. We further propose a more robust version of the algorithm which addresses the scenario when the preference matrix is adversarially perturbed. We then give regret guarantees that scale with the amount by which the preference matrix is perturbed from a biclustered structure. To our knowledge, these are the first results on online collaborative filtering that hold at this level of generality and adaptivity under no-repetition constraints. | ['Claudio Gentile', 'Mark Herbster', 'Fabio Vitale', 'Stephen Pasteris'] | 2023-02-11 | null | null | null | null | ['collaborative-filtering'] | ['miscellaneous'] | [ 1.02060147e-01 -3.07272989e-02 -4.47299927e-02 -1.48432612e-01
-2.62858063e-01 -1.28691888e+00 1.49123156e-02 8.93778875e-02
-6.27540052e-01 5.64893603e-01 3.90920818e-01 -5.60100079e-01
-6.72050536e-01 -9.81999815e-01 -8.98781538e-01 -6.97078884e-01
-4.57309514e-01 6.17132008e-01 1.82613492e-01 -3.88885200e-01
1.43822253e-01 3.56988996e-01 -1.14501834e+00 3.27942908e-01
7.24667728e-01 1.02740943e+00 -2.05571860e-01 8.34758162e-01
1.48311630e-01 1.13085963e-01 -1.93030879e-01 -7.18096554e-01
9.52441633e-01 -4.33289468e-01 -7.81301618e-01 4.91633147e-01
2.51986951e-01 -3.67823780e-01 -4.37119633e-01 1.18894148e+00
3.42642725e-01 4.03253734e-01 2.90636361e-01 -9.20766175e-01
-4.98876184e-01 1.10181808e+00 -3.29411983e-01 1.44393682e-01
4.65717614e-01 -4.16397750e-01 1.25498056e+00 -3.78691077e-01
7.34345794e-01 7.53047407e-01 5.21424711e-01 4.86821175e-01
-1.32256770e+00 -4.49641228e-01 8.18844795e-01 -9.11776870e-02
-9.57495570e-01 -2.50469446e-01 3.22328150e-01 -1.63202047e-01
1.90982670e-02 6.87267005e-01 6.18868530e-01 5.84613860e-01
-1.10789381e-01 8.12288046e-01 7.93076694e-01 -3.30271751e-01
6.74439311e-01 7.38319159e-02 1.09745972e-01 4.09871668e-01
5.79394758e-01 -7.11939931e-02 -3.67214829e-01 -6.82318509e-01
5.00523448e-01 2.89706677e-01 -4.09300983e-01 -9.39721048e-01
-5.85131228e-01 6.92076623e-01 1.52733296e-01 2.12520927e-01
-3.80815893e-01 -1.25941977e-01 1.33560762e-01 1.13322878e+00
2.98415869e-01 3.47362310e-01 -6.76294684e-01 1.87672734e-01
-8.18221569e-01 4.02262449e-01 1.23552668e+00 1.05595720e+00
3.42050552e-01 -1.98671922e-01 -2.03090515e-02 3.23593915e-01
-4.85048816e-02 4.05312419e-01 9.27391872e-02 -1.15872741e+00
5.25596857e-01 1.07676290e-01 1.01823854e+00 -7.58057237e-01
-1.89668760e-01 -6.78806126e-01 -4.94864762e-01 1.09525612e-02
7.49253631e-01 -7.63438880e-01 -3.72635633e-01 1.96475220e+00
3.71307045e-01 3.96525115e-02 -3.75494361e-01 9.22723889e-01
-7.82996863e-02 3.43319923e-01 -4.93324459e-01 -7.66175151e-01
9.19639647e-01 -4.34973687e-01 -4.54844415e-01 1.33493140e-01
5.28623462e-01 -5.90472221e-01 5.79891741e-01 6.80166483e-01
-1.46418297e+00 1.29034877e-01 -7.97178030e-01 5.97799361e-01
9.61221755e-02 -4.50232148e-01 6.18879080e-01 1.20602417e+00
-8.02183390e-01 6.17087722e-01 -5.96547067e-01 -1.59713730e-01
3.35046351e-01 4.30988282e-01 -2.08945736e-01 -2.36597285e-01
-9.65693831e-01 3.33294541e-01 4.15754355e-02 -1.67672619e-01
-6.08837426e-01 -7.37493932e-01 -8.10501724e-02 2.91666687e-01
8.01136076e-01 -8.08123052e-01 1.17109406e+00 -1.41315448e+00
-1.24480867e+00 1.64582297e-01 6.33853376e-02 -5.56569993e-01
9.36906636e-01 -1.58458903e-01 -3.57651263e-01 -5.42347394e-02
-2.97787815e-01 -5.87973535e-01 7.40441382e-01 -1.22146511e+00
-8.97130847e-01 -6.00013375e-01 5.51632702e-01 2.79032320e-01
-5.99631250e-01 -9.14241299e-02 -3.88204187e-01 -7.04692066e-01
3.09461921e-01 -1.08035028e+00 -6.90539837e-01 1.63128942e-01
-5.65957092e-02 2.89684504e-01 2.33792946e-01 -2.18898430e-01
1.24171805e+00 -2.05429149e+00 1.50716454e-01 7.66459167e-01
-1.47736892e-01 -5.63055985e-02 -2.18213767e-01 8.66848826e-01
2.58830249e-01 3.49956870e-01 -1.60562117e-02 -5.41414991e-02
9.97204185e-02 2.81517833e-01 -4.46804672e-01 7.30296552e-01
-9.53029096e-01 3.05303007e-01 -9.02177036e-01 2.41965607e-01
-3.43337625e-01 -2.92049527e-01 -1.23600471e+00 7.38845346e-03
-3.07519495e-01 6.35614991e-02 -6.16367519e-01 -5.18836156e-02
9.14451659e-01 -4.68770005e-02 8.56658459e-01 1.82818592e-01
2.37790607e-02 -1.65698633e-01 -1.82407236e+00 1.42349899e+00
-4.79061365e-01 -1.85312778e-01 8.41918588e-01 -7.52043366e-01
8.00162628e-02 4.74651098e-01 7.05769956e-01 -1.75209165e-01
2.23256513e-01 -2.45112609e-02 -2.96237487e-02 -2.99548864e-01
3.60779464e-01 -2.70126671e-01 -1.50052756e-01 9.56952512e-01
-1.06168687e-01 6.23380601e-01 1.62187591e-01 7.12218761e-01
1.12462842e+00 -4.65488315e-01 -1.17313107e-02 -2.16707513e-01
1.68454617e-01 -4.33136970e-01 7.20385134e-01 1.51837158e+00
1.82178751e-01 2.59167671e-01 3.82928818e-01 -2.49806553e-01
-1.04926360e+00 -1.01040471e+00 8.69342219e-03 1.47290218e+00
2.79960811e-01 -3.18789661e-01 -4.51158911e-01 -7.98187733e-01
4.33861226e-01 5.47771394e-01 -8.24164093e-01 1.97990045e-01
-6.10624440e-03 -6.31712198e-01 -4.60084192e-02 2.72194803e-01
1.67972580e-01 -2.80971587e-01 9.89318341e-02 4.51420367e-01
4.89459485e-02 -7.03920424e-01 -1.09142590e+00 -2.92044640e-01
-1.00289023e+00 -1.11501455e+00 -2.57712424e-01 -3.97131056e-01
9.71317589e-01 4.39230055e-01 6.42129302e-01 8.36076494e-03
1.67821854e-01 7.12264419e-01 -5.01957953e-01 -7.82593042e-02
-1.06830552e-01 -1.85926005e-01 2.01917633e-01 6.87743425e-01
-2.33148634e-01 -7.24478126e-01 -8.66903961e-01 5.62739611e-01
-1.12013447e+00 -4.31217492e-01 4.58418317e-02 5.12292504e-01
2.51954734e-01 3.93536597e-01 5.70394397e-01 -1.56485772e+00
7.27975011e-01 -8.37561846e-01 -6.60451889e-01 2.93798208e-01
-5.67522824e-01 -7.78364837e-02 1.07687020e+00 -4.42890137e-01
-1.12039280e+00 3.00358504e-01 3.61996815e-02 2.89824083e-02
5.06196171e-02 4.43620294e-01 -4.37778562e-01 -2.39434198e-01
5.82230330e-01 1.27404466e-01 -1.73770845e-01 -8.14078033e-01
8.43447447e-01 4.99871790e-01 2.23794505e-01 -6.28177047e-01
8.31571519e-01 8.55840564e-01 1.36998206e-01 -4.33776319e-01
-7.75372326e-01 -6.92217112e-01 -4.85147625e-01 4.80426773e-02
-6.07026890e-02 -5.91974139e-01 -9.67769444e-01 1.45763606e-01
-4.46118027e-01 -1.18796811e-01 -2.10155234e-01 4.15534794e-01
-6.25299692e-01 8.00557315e-01 -5.39989293e-01 -1.14950311e+00
-6.46091253e-02 -2.74852395e-01 -2.80447900e-01 -1.37430489e-01
2.04678491e-01 -8.65222871e-01 -9.01443735e-02 3.12121540e-01
4.43279237e-01 -1.90354303e-01 4.93811995e-01 -9.03553963e-01
-4.79596466e-01 -5.75822592e-01 4.24424171e-01 5.10508306e-02
6.91401511e-02 -3.31630796e-01 -4.29977588e-02 -8.49424601e-01
-4.09456193e-02 -5.18836118e-02 6.18306875e-01 2.08848178e-01
1.24603939e+00 -1.19681776e+00 -1.82834342e-01 4.15907413e-01
1.38476467e+00 -3.57861519e-02 2.58209050e-01 5.38057722e-02
1.98843986e-01 3.38590533e-01 5.17242253e-01 1.03366458e+00
8.22233558e-02 5.32382786e-01 4.23913270e-01 4.56952482e-01
6.51607871e-01 -2.95002490e-01 1.53168052e-01 3.94074380e-01
-1.33573443e-01 -5.73300123e-01 6.62464201e-02 6.86906934e-01
-2.14009428e+00 -1.20665455e+00 1.92558706e-01 2.75131130e+00
8.24427724e-01 -6.24130992e-03 6.04470789e-01 1.12715445e-01
6.41865551e-01 -2.13066965e-01 -7.45110691e-01 -4.53573465e-01
-2.79958472e-02 9.75674018e-02 1.09118795e+00 7.50987411e-01
-9.12380815e-01 5.51708639e-01 6.49399090e+00 5.02872407e-01
-5.36736369e-01 2.84344196e-01 1.61177650e-01 -7.91112483e-01
-7.35422134e-01 1.75449967e-01 -5.10115445e-01 5.27300715e-01
7.24772274e-01 -4.16444033e-01 1.05555820e+00 7.44579375e-01
2.51132578e-01 1.27671883e-01 -8.73191833e-01 3.22572619e-01
-1.11956485e-01 -1.19150555e+00 -8.58662501e-02 1.62502453e-01
1.19892323e+00 -1.49232030e-01 1.50038406e-01 -2.94081420e-01
9.77298558e-01 -2.93834686e-01 6.93981946e-01 4.64568526e-01
1.88613564e-01 -9.90409791e-01 3.49896044e-01 6.78773820e-01
-8.22316170e-01 -8.34557474e-01 -4.25335675e-01 -1.81795552e-01
3.77430648e-01 6.26612782e-01 -1.52106866e-01 6.59274757e-01
5.31392157e-01 1.74492255e-01 8.48026574e-02 1.50887465e+00
9.34614688e-02 9.51197147e-01 -6.17735565e-01 1.58251017e-01
-8.57185647e-02 -3.72819990e-01 6.45626605e-01 8.68454874e-01
4.17967021e-01 6.47567570e-01 5.07311821e-01 1.21177575e-02
-4.21155214e-01 3.99622679e-01 -2.01265678e-01 1.86571077e-01
7.06528306e-01 1.04107738e+00 -4.55656230e-01 -3.19652483e-02
-5.90852618e-01 1.05836749e+00 6.51203096e-02 4.11417186e-01
-2.97297895e-01 -1.93002641e-01 7.48830974e-01 5.04977405e-01
8.30143332e-01 -1.94970801e-01 -2.94521689e-01 -1.26207876e+00
-1.06968880e-01 -7.15997696e-01 9.93167341e-01 -1.78094089e-01
-1.67346978e+00 -5.24410419e-02 -4.05560225e-01 -9.21238005e-01
1.41038716e-01 -1.86840534e-01 -5.12196004e-01 4.78100568e-01
-7.45227695e-01 -6.07654452e-01 3.65564317e-01 9.01195645e-01
6.83791637e-02 -6.01872895e-03 6.17603898e-01 4.63364422e-02
-2.14227393e-01 8.38565946e-01 1.04259896e+00 -2.26338863e-01
4.22392398e-01 -1.11232305e+00 -1.48498088e-01 1.01167047e+00
1.46094903e-01 8.22967529e-01 1.00425637e+00 -5.44710398e-01
-1.75277781e+00 -1.10693264e+00 5.49428403e-01 -2.34201446e-01
8.58220339e-01 -5.79558551e-01 -4.83791322e-01 7.80499279e-01
-1.58805296e-01 2.87372991e-02 1.02685654e+00 5.51311672e-01
-2.97412604e-01 -3.12662363e-01 -1.37254250e+00 7.35852778e-01
1.46641672e+00 -1.44823655e-01 -2.33535320e-01 4.09355402e-01
4.43044305e-01 -2.08534777e-01 -9.78871226e-01 1.34647172e-02
8.63996267e-01 -6.39341891e-01 7.96728671e-01 -1.15315413e+00
-3.27661157e-01 -1.96147799e-01 -2.02725396e-01 -1.17924762e+00
-8.50975513e-01 -9.58007812e-01 -2.44073689e-01 7.25231469e-01
5.87387562e-01 -6.75890625e-01 1.05823100e+00 9.87927735e-01
2.41255969e-01 -3.93150091e-01 -9.92028773e-01 -8.42740238e-01
2.54567534e-01 -9.20709372e-02 5.10598958e-01 7.57582068e-01
3.90692383e-01 -3.13510448e-02 -8.64760339e-01 5.05388021e-01
7.14390516e-01 5.70725262e-01 6.86865926e-01 -1.18735349e+00
-8.75814319e-01 -3.80407013e-02 2.92447835e-01 -1.21978378e+00
-2.03116432e-01 -7.01406360e-01 -3.55343819e-01 -1.26488197e+00
2.05372438e-01 -6.47779226e-01 -5.13985217e-01 2.05874324e-01
1.29000083e-01 1.28605133e-02 2.96027124e-01 1.17555432e-01
-9.78327096e-01 9.79705304e-02 1.17826056e+00 2.57374316e-01
-3.61635685e-01 9.43926275e-01 -1.21013832e+00 4.39713866e-01
7.25832641e-01 -3.80177736e-01 -6.69474721e-01 -1.94253042e-01
6.00147009e-01 4.11302298e-01 -2.88735926e-01 -3.33325833e-01
4.40074503e-01 -6.21626556e-01 1.01460710e-01 -2.52833843e-01
-1.24672860e-01 -1.08588791e+00 5.15548050e-01 5.25625765e-01
-8.67753386e-01 -2.59734035e-01 -3.31486106e-01 1.27949989e+00
6.18688524e-01 -4.34245706e-01 7.17217565e-01 -1.70977667e-01
6.09076098e-02 7.19160438e-01 -8.93090785e-01 2.78446134e-02
8.61676633e-01 -1.22595780e-01 1.22773545e-02 -9.17242885e-01
-1.31624937e+00 3.81925195e-01 6.88395321e-01 1.11215487e-01
1.45869628e-01 -1.04486668e+00 -5.83962798e-01 1.09316014e-01
-2.04807863e-01 -7.02979147e-01 5.96266031e-01 3.70200783e-01
1.84704974e-01 1.09796152e-01 -3.37997116e-02 4.34345454e-01
-1.37360716e+00 1.13863742e+00 1.69507533e-01 9.44669917e-03
-4.28099602e-01 6.31105185e-01 -4.25917767e-02 -1.28693715e-01
3.38801801e-01 3.25096548e-01 8.14754292e-02 -4.97464016e-02
5.81059635e-01 4.57659364e-01 6.10524975e-02 -1.50661245e-01
4.98860367e-02 5.58994301e-02 -4.40114260e-01 -3.22236955e-01
1.26540494e+00 -6.75151110e-01 6.41035335e-03 -3.67373116e-02
5.86530268e-01 7.46127129e-01 -1.31291556e+00 -6.12429380e-01
-3.71483415e-01 -9.60744858e-01 -1.01889551e-01 -8.51595998e-01
-1.24521232e+00 7.53680840e-02 2.72231191e-01 5.89773059e-01
1.10200167e+00 -3.50498110e-01 6.01862609e-01 4.41394508e-01
8.19010079e-01 -1.25661325e+00 -2.69733101e-01 2.78980821e-01
5.00531673e-01 -3.80371600e-01 -6.40763640e-02 -4.06153232e-01
-4.15495753e-01 7.02998757e-01 7.37615768e-03 -3.46429855e-01
1.16327739e+00 1.72745958e-01 -5.10025084e-01 4.66431826e-01
-9.31043804e-01 -2.55227536e-01 -5.42513132e-02 3.76653641e-01
6.22490197e-02 3.53933275e-01 -8.70363295e-01 1.07129204e+00
1.08269237e-01 1.05492815e-01 7.56045282e-01 9.17146564e-01
-7.92497754e-01 -1.49725032e+00 -3.24701816e-01 6.84229970e-01
-8.91479254e-01 -3.75611032e-03 -3.63620460e-01 -1.26291528e-01
-3.99351828e-02 1.18313193e+00 -1.73608169e-01 -2.93928415e-01
3.30900133e-01 -8.41014460e-02 7.03918099e-01 -4.59067523e-01
-6.67842090e-01 2.71149147e-02 2.62670219e-01 -3.76575887e-01
4.34481874e-02 -7.73351252e-01 -7.94897079e-01 -7.36408174e-01
-5.60820878e-01 7.14298069e-01 2.14280769e-01 7.67972529e-01
6.35416627e-01 -2.17107669e-01 1.30288947e+00 -2.65916497e-01
-1.03234637e+00 -4.82483029e-01 -1.07068217e+00 7.14625955e-01
-1.52360806e-02 -1.02777563e-01 -4.58722919e-01 -1.63655773e-01] | [4.601436138153076, 3.4043312072753906] |
ce210dde-e54e-442b-b6d2-2756a2c89fba | phocal-a-multi-modal-dataset-for-category | 2205.08811 | null | https://arxiv.org/abs/2205.08811v1 | https://arxiv.org/pdf/2205.08811v1.pdf | PhoCaL: A Multi-Modal Dataset for Category-Level Object Pose Estimation with Photometrically Challenging Objects | Object pose estimation is crucial for robotic applications and augmented reality. Beyond instance level 6D object pose estimation methods, estimating category-level pose and shape has become a promising trend. As such, a new research field needs to be supported by well-designed datasets. To provide a benchmark with high-quality ground truth annotations to the community, we introduce a multimodal dataset for category-level object pose estimation with photometrically challenging objects termed PhoCaL. PhoCaL comprises 60 high quality 3D models of household objects over 8 categories including highly reflective, transparent and symmetric objects. We developed a novel robot-supported multi-modal (RGB, depth, polarisation) data acquisition and annotation process. It ensures sub-millimeter accuracy of the pose for opaque textured, shiny and transparent objects, no motion blur and perfect camera synchronisation. To set a benchmark for our dataset, state-of-the-art RGB-D and monocular RGB methods are evaluated on the challenging scenes of PhoCaL. | ['Benjamin Busam', 'Nassir Navab', 'Sven Meier', 'Lorenzo Garattoni', 'Rahul Parthasarathy Srikanth', 'Siyuan Shen', 'Yitong Li', 'HyunJun Jung', 'Pengyuan Wang'] | 2022-05-18 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Wang_PhoCaL_A_Multi-Modal_Dataset_for_Category-Level_Object_Pose_Estimation_With_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Wang_PhoCaL_A_Multi-Modal_Dataset_for_Category-Level_Object_Pose_Estimation_With_CVPR_2022_paper.pdf | cvpr-2022-1 | ['transparent-objects', '6d-pose-estimation'] | ['computer-vision', 'computer-vision'] | [ 5.14604934e-02 9.41903964e-02 2.51508415e-01 -4.31351006e-01
-8.75609994e-01 -6.59616470e-01 5.64667821e-01 -6.51494861e-02
-1.35862604e-01 3.92138898e-01 -2.53399581e-01 2.65571058e-01
-1.08570429e-02 -3.54092032e-01 -8.30553949e-01 -6.65598214e-01
1.69840887e-01 1.11337316e+00 5.11518002e-01 1.74386539e-02
1.88458025e-01 8.10120642e-01 -1.79556859e+00 1.26490176e-01
2.73043543e-01 1.44366217e+00 3.42317671e-01 5.30869126e-01
2.64187843e-01 4.56690282e-01 -1.68387070e-01 -6.48713470e-01
5.88022470e-01 2.79037923e-01 -6.20189071e-01 3.81218880e-01
1.03826785e+00 -3.99520189e-01 6.73908219e-02 8.46286774e-01
5.78377485e-01 -1.91751257e-01 7.53475368e-01 -1.51986241e+00
-4.38812822e-01 -8.88151005e-02 -6.64114177e-01 -6.09499753e-01
9.04640019e-01 2.19430402e-01 7.49243796e-01 -9.81732309e-01
6.56545997e-01 1.38214815e+00 9.13498223e-01 3.65147769e-01
-1.10968864e+00 -2.98668504e-01 -8.92818347e-02 1.32106468e-01
-1.32814968e+00 -3.09918016e-01 8.76689851e-01 -5.22475481e-01
8.34936440e-01 4.21841264e-01 6.25984430e-01 1.10199702e+00
1.52449943e-02 4.18637723e-01 1.41392338e+00 -3.05511177e-01
1.98002264e-01 4.07477379e-01 -1.44704357e-01 5.89960515e-01
4.79865789e-01 -1.28007665e-01 -7.46938467e-01 -4.88959923e-02
6.70225680e-01 4.37279046e-02 -3.78062755e-01 -1.21209943e+00
-1.71599460e+00 8.43243971e-02 5.92058778e-01 -2.91546255e-01
-3.51888061e-01 3.09048206e-01 5.18619120e-02 -1.71266362e-01
2.57556647e-01 2.84279317e-01 -6.90920234e-01 -1.34576201e-01
-1.66731715e-01 1.41156390e-01 6.75681293e-01 1.47582459e+00
8.41306508e-01 -4.94406283e-01 2.13786960e-01 6.57636404e-01
9.72320735e-01 1.22859538e+00 -9.33045596e-02 -1.09053838e+00
5.14588237e-01 7.68124223e-01 6.20637596e-01 -9.21519756e-01
-5.33646107e-01 8.96221623e-02 -4.09960389e-01 1.89650550e-01
4.11454111e-01 5.00611186e-01 -7.19755948e-01 1.06503773e+00
8.48902762e-01 -2.65369266e-01 4.28340547e-02 1.26295865e+00
1.25601315e+00 2.49924466e-01 -4.78441089e-01 1.57178286e-02
1.65292239e+00 -6.21568501e-01 -3.46339941e-01 -3.48936230e-01
3.97830963e-01 -1.08375394e+00 1.01261580e+00 6.31753922e-01
-1.00508428e+00 -2.69852042e-01 -7.79523671e-01 -3.62313539e-01
-2.10210949e-01 3.82595986e-01 6.86895370e-01 8.13339055e-01
-8.43726575e-01 -1.86987203e-02 -6.21562779e-01 -6.75515592e-01
1.79763258e-01 4.23555881e-01 -8.09969783e-01 -3.90859872e-01
-5.14713228e-01 1.13374448e+00 3.79243582e-01 2.83184618e-01
-8.18786681e-01 -8.35096598e-01 -7.91007459e-01 -7.41950810e-01
3.37403536e-01 -6.14198208e-01 1.18324924e+00 -3.43972027e-01
-1.67490506e+00 1.64358556e+00 3.19726616e-01 -3.43997627e-02
5.60984850e-01 -4.49963331e-01 1.37159705e-01 2.56136507e-01
-1.35488927e-01 7.07581341e-01 6.49888158e-01 -1.88659549e+00
-4.38143611e-01 -9.26076949e-01 -5.52788451e-02 4.07618880e-01
-4.27232645e-02 -7.00569227e-02 -5.28176665e-01 1.35701835e-01
7.48610377e-01 -1.26114833e+00 2.52903670e-01 5.20577908e-01
-5.80355167e-01 1.12795997e-02 1.01855755e+00 -5.00897169e-01
-2.50274986e-02 -2.00732517e+00 1.60498485e-01 -1.64833829e-01
-9.94603932e-02 -2.61993796e-01 2.07666323e-01 5.87055236e-02
3.07302237e-01 -5.58508098e-01 8.41983482e-02 -8.65410984e-01
4.19326246e-01 1.21897146e-01 -4.59352359e-02 1.20397031e+00
-1.97964370e-01 7.24659681e-01 -6.83532536e-01 -6.62814260e-01
6.71235025e-01 6.23944104e-01 -4.69160497e-01 4.35162306e-01
-4.08070832e-01 5.89520991e-01 -2.04889983e-01 1.36590302e+00
1.12063658e+00 -5.52688949e-02 -2.07910597e-01 -7.18939662e-01
-9.01801512e-02 -8.04935247e-02 -1.46067631e+00 2.03877258e+00
-4.50001359e-01 5.20634413e-01 3.59414607e-01 -2.86358297e-01
1.08008862e+00 1.90423474e-01 7.07551897e-01 -5.45042932e-01
2.84311891e-01 4.83605891e-01 -6.11818016e-01 -5.18112004e-01
7.22700715e-01 6.68051988e-02 -1.17386006e-01 -8.22555348e-02
-1.09114923e-01 -9.75570977e-01 -4.97064143e-01 -1.42237827e-01
7.91303396e-01 5.09186804e-01 1.26200030e-02 -1.42304614e-01
3.23439449e-01 1.15482084e-01 2.59916306e-01 2.65529871e-01
-1.57449812e-01 9.84721839e-01 -8.83008763e-02 -3.76237571e-01
-1.03752470e+00 -1.27828670e+00 -4.34231579e-01 6.19166434e-01
6.59389794e-01 1.22414783e-01 -2.75461644e-01 -2.11358368e-01
4.04607475e-01 1.31223053e-01 -3.90308738e-01 1.99488148e-01
-3.18532497e-01 -6.74462676e-01 4.51894552e-02 2.10094944e-01
7.62866139e-01 -6.98504567e-01 -9.87285852e-01 -1.74540728e-01
-1.66270405e-01 -1.64356542e+00 1.13167979e-01 1.95784137e-01
-6.71351314e-01 -1.27404404e+00 -7.56907821e-01 -6.75784767e-01
4.86315131e-01 5.28027117e-01 1.37559509e+00 -5.92831254e-01
-3.45496535e-01 1.01613772e+00 -4.50201064e-01 -4.65087891e-01
-2.78999954e-02 -2.55372554e-01 2.72729158e-01 1.40460255e-02
1.09328233e-01 -2.58701950e-01 -7.04221785e-01 9.79060233e-01
-3.95198733e-01 1.39582783e-01 4.81032491e-01 5.04939444e-02
7.57709265e-01 -5.92392027e-01 -2.29039952e-01 -2.69918200e-02
-4.36920166e-01 -1.41343892e-01 -1.12395597e+00 1.74354970e-01
2.51047332e-02 -3.93789262e-01 -1.93317816e-01 -4.88275021e-01
-1.01695049e+00 6.28510535e-01 3.73327345e-01 -4.56248939e-01
-3.62240970e-01 -2.09048614e-01 -4.77844834e-01 -4.80942816e-01
7.90280819e-01 -1.52620405e-01 -1.69979095e-01 -5.49056411e-01
4.28491592e-01 9.42281365e-01 7.87451088e-01 -6.43556654e-01
8.60834837e-01 9.35996115e-01 3.29099953e-01 -1.02082789e+00
-7.09339917e-01 -9.02229786e-01 -9.57301795e-01 -4.75360632e-01
9.17328358e-01 -1.09430194e+00 -1.09979773e+00 7.02435315e-01
-1.53261018e+00 -4.30520475e-01 -2.10645106e-02 6.38686240e-01
-8.96616817e-01 2.54469752e-01 -1.68071181e-01 -1.10907507e+00
-8.23292509e-02 -1.21450472e+00 2.01495147e+00 4.43421025e-03
1.42832056e-01 -3.58994663e-01 -6.92861155e-02 8.16652656e-01
-4.67773154e-02 6.67482018e-01 1.92158282e-01 1.14853293e-01
-1.24390507e+00 -2.00428069e-01 -3.93176436e-01 -2.80513950e-02
6.28915504e-02 -3.75099555e-02 -1.33214629e+00 -8.65790918e-02
-3.28490622e-02 -7.28340626e-01 2.52259076e-01 1.57340720e-01
7.77685463e-01 1.61260262e-01 -3.36043775e-01 7.52505481e-01
1.48060155e+00 -2.86599994e-01 7.01480687e-01 5.41817963e-01
1.03892589e+00 7.19711483e-01 1.18463659e+00 5.55049956e-01
7.69572616e-01 1.29763889e+00 1.30937159e+00 2.04424635e-01
-1.29042640e-01 4.07078832e-01 2.83751905e-01 4.59252477e-01
-2.61609107e-01 -8.88930447e-03 -1.09374070e+00 3.82697821e-01
-1.47706485e+00 -4.22009498e-01 -8.88886631e-01 2.47492218e+00
6.49770439e-01 -2.00540766e-01 2.19228081e-02 2.71034420e-01
3.99574816e-01 -2.54050374e-01 -3.96953493e-01 1.90303147e-01
-2.71705329e-01 -4.60659295e-01 7.67472088e-01 1.72365516e-01
-1.05185449e+00 6.47407889e-01 5.18054676e+00 3.16905737e-01
-8.62669289e-01 6.82743266e-02 1.06904097e-01 1.89399645e-02
-1.65717825e-02 -3.48011792e-01 -1.06492674e+00 1.96567997e-01
5.68440735e-01 8.25223207e-01 1.42983228e-01 1.14615893e+00
-8.40311795e-02 -5.51522851e-01 -1.30697465e+00 1.63540924e+00
2.24395320e-01 -8.94081414e-01 -5.49040973e-01 2.51109034e-01
8.61365557e-01 4.23726350e-01 7.96472281e-02 -1.21574312e-01
3.43253464e-02 -6.40447021e-01 1.21935916e+00 5.84272563e-01
7.17286348e-01 -2.60032207e-01 7.82015860e-01 2.13585585e-01
-1.25986588e+00 -6.81993505e-03 -5.44126451e-01 3.68344426e-01
1.73221201e-01 6.47141397e-01 -9.70339179e-01 6.50195181e-01
1.15867007e+00 6.55699968e-01 -7.76392817e-01 1.21586561e+00
1.57895759e-02 -1.97017759e-01 -6.95371091e-01 -1.01323668e-02
-2.19587117e-01 -1.46620765e-01 3.56782794e-01 8.21057141e-01
3.73028070e-01 4.45638001e-02 1.60295919e-01 5.71123958e-01
3.65964361e-02 -1.51196182e-01 -5.61224163e-01 4.03602600e-01
3.10651869e-01 1.53335667e+00 -7.74879158e-01 1.93618581e-01
-2.08068728e-01 1.05701137e+00 -4.49552312e-02 -1.13502055e-01
-9.08586979e-01 2.50420421e-01 5.81673861e-01 2.22091451e-01
1.78252444e-01 -6.05052888e-01 -3.08564901e-01 -1.08355319e+00
4.16141123e-01 -5.71273685e-01 -1.16216891e-01 -1.48972106e+00
-1.17841327e+00 3.20578158e-01 1.54381141e-01 -1.46715665e+00
1.69555396e-01 -1.15846217e+00 2.99883068e-01 5.49399555e-01
-1.57194042e+00 -1.73082674e+00 -1.16336298e+00 6.07020199e-01
2.52092808e-01 2.03235745e-01 7.40442455e-01 2.57884353e-01
3.37142833e-02 4.97355238e-02 6.46108994e-04 -3.63226026e-01
8.12394857e-01 -1.25349700e+00 -2.06948757e-01 2.68263459e-01
7.40568386e-03 1.80560052e-01 7.47091889e-01 -4.43095386e-01
-2.40733242e+00 -1.07902825e+00 2.44045258e-01 -1.25120723e+00
4.11661804e-01 -7.11799324e-01 -4.66502219e-01 5.69895267e-01
-3.21984380e-01 4.66748774e-01 1.13287777e-01 -2.33915552e-01
-2.94809252e-01 -3.77154589e-01 -1.49604046e+00 1.22410171e-01
1.13334727e+00 -5.36617994e-01 -4.42762375e-01 6.66343510e-01
6.03362858e-01 -1.13453412e+00 -1.14332783e+00 6.40932143e-01
6.81082010e-01 -1.22187030e+00 1.38122380e+00 4.41829294e-01
-1.60945430e-02 -5.98842025e-01 -7.48929977e-01 -7.79848874e-01
3.09578061e-01 -4.69951183e-01 -2.19667554e-01 1.22034466e+00
-4.71627377e-02 -4.68565762e-01 9.62531149e-01 5.82149565e-01
-4.09203142e-01 -3.50240976e-01 -1.06424999e+00 -8.76144409e-01
-6.34397388e-01 -6.14612579e-01 5.23545504e-01 6.66146696e-01
-5.86746395e-01 -3.81227210e-02 -1.90022171e-01 6.97743356e-01
9.90645230e-01 2.77538806e-01 1.42654788e+00 -1.65462887e+00
7.67390653e-02 -1.84875261e-02 -8.20055902e-01 -1.00765085e+00
-8.42888579e-02 -5.37905514e-01 2.14279249e-01 -1.59055936e+00
1.56755045e-01 -7.20224202e-01 5.04843295e-01 1.71559200e-01
5.05248189e-01 9.35602665e-01 -6.15072856e-03 2.82062620e-01
-9.04679537e-01 6.31919622e-01 1.13523388e+00 2.88764071e-02
1.84940681e-01 -5.01274094e-02 -9.91245285e-02 7.40868986e-01
3.12390953e-01 -2.25624502e-01 -3.21020260e-02 -5.76858222e-01
6.44226313e-01 -2.21608937e-01 9.16644275e-01 -1.16092455e+00
8.39502178e-03 -1.65076479e-01 3.01082999e-01 -1.25559402e+00
1.20489740e+00 -1.46493924e+00 4.43695605e-01 3.36108744e-01
3.11637580e-01 -1.28419459e-01 -5.80140688e-02 4.27647978e-01
3.03893924e-01 8.52785185e-02 8.96210074e-01 -2.74143606e-01
-7.71155536e-01 3.96637648e-01 4.18638617e-01 -2.90297747e-01
1.24932027e+00 -6.62581027e-01 -3.36297542e-01 2.21061893e-02
-3.34881425e-01 7.22579584e-02 1.10233486e+00 4.65060055e-01
6.54833972e-01 -1.42359102e+00 -5.11081219e-01 -1.01731561e-01
7.33101904e-01 6.99918687e-01 -1.16030909e-02 9.41885829e-01
-9.24294889e-01 5.28778911e-01 -1.05408147e-01 -1.42693448e+00
-1.45830810e+00 2.28787556e-01 4.05866325e-01 5.18721461e-01
-3.40667158e-01 9.09644783e-01 -8.79528299e-02 -1.01548970e+00
4.55082625e-01 -7.43723631e-01 3.25375736e-01 -1.45874443e-02
1.85690358e-01 5.00206113e-01 4.80433464e-01 -9.95258331e-01
-7.36827075e-01 1.06061447e+00 6.63058758e-01 -5.20865768e-02
1.37271357e+00 -4.05562967e-01 -3.12358677e-01 6.93979323e-01
1.05615568e+00 -2.03409418e-01 -1.51110566e+00 -1.30953640e-01
-5.88204004e-02 -7.42237985e-01 -5.06491289e-02 -7.44043291e-01
-6.73550546e-01 7.75105238e-01 9.56751883e-01 9.36146379e-02
7.97728300e-01 4.92367744e-01 1.86211944e-01 5.43401182e-01
1.14601254e+00 -9.12386715e-01 4.60258394e-01 4.82152492e-01
1.24014807e+00 -1.78041446e+00 2.54466861e-01 -7.12247133e-01
-4.34857965e-01 9.96311367e-01 6.34783566e-01 2.26795077e-01
2.52906621e-01 9.04482529e-02 5.58860824e-02 -3.68547678e-01
-8.73298012e-03 -8.99216682e-02 4.85710055e-01 1.09569216e+00
1.64052010e-01 -6.75441399e-02 4.93576974e-01 -2.98660472e-02
-2.92686850e-01 -4.13194060e-01 2.39482582e-01 8.79701078e-01
-3.77359241e-01 -6.59595013e-01 -1.14171863e+00 -1.17649673e-03
8.82690400e-02 4.41261739e-01 -3.37890714e-01 8.91323388e-01
1.15478151e-01 7.88379669e-01 -2.08243844e-03 -4.33196463e-02
5.69713414e-01 -3.76282185e-01 1.11707699e+00 -6.01185441e-01
-3.46667469e-01 -1.16738796e-01 1.99014530e-01 -8.45998704e-01
-8.36829245e-01 -7.86201656e-01 -9.40640867e-01 -2.36288663e-02
-5.29400706e-01 -4.87967610e-01 1.50496018e+00 6.94106460e-01
3.61968912e-02 -1.95344612e-02 5.31134725e-01 -1.73751688e+00
-3.31577122e-01 -9.57102895e-01 -6.07673764e-01 3.36539239e-01
2.85640717e-01 -1.02587187e+00 -2.83285499e-01 4.96791899e-02] | [7.22430419921875, -2.4037723541259766] |
9f1c37ff-467f-4cd0-8a62-95ef14f0f666 | a-baseline-temporal-tagger-for-all-languages | null | null | https://aclanthology.org/D15-1063 | https://aclanthology.org/D15-1063.pdf | A Baseline Temporal Tagger for all Languages | null | ['Jannik Str{\\"o}tgen', 'Michael Gertz'] | 2015-09-01 | null | null | null | emnlp-2015-9 | ['timex-normalization', 'temporal-information-extraction'] | ['natural-language-processing', 'natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.197238922119141, 3.749377489089966] |
d89e4816-fff5-4bd3-b26c-a2c909e2c54a | fast-learning-of-multidimensional-hawkes | 2212.06081 | null | https://arxiv.org/abs/2212.06081v1 | https://arxiv.org/pdf/2212.06081v1.pdf | Fast Learning of Multidimensional Hawkes Processes via Frank-Wolfe | Hawkes processes have recently risen to the forefront of tools when it comes to modeling and generating sequential events data. Multidimensional Hawkes processes model both the self and cross-excitation between different types of events and have been applied successfully in various domain such as finance, epidemiology and personalized recommendations, among others. In this work we present an adaptation of the Frank-Wolfe algorithm for learning multidimensional Hawkes processes. Experimental results show that our approach has better or on par accuracy in terms of parameter estimation than other first order methods, while enjoying a significantly faster runtime. | ['Manuela Veloso', 'Tucker Balch', 'Vamsi K. Potluru', 'Mohsen Ghassemi', 'Niccolò Dalmasso', 'Renbo Zhao'] | 2022-12-12 | null | null | null | null | ['epidemiology'] | ['medical'] | [-2.05097690e-01 -2.44060144e-01 1.92467317e-01 5.58567885e-03
-6.07127190e-01 -4.13254261e-01 1.20819116e+00 5.78759909e-01
-1.93488061e-01 6.22504711e-01 4.61163610e-01 -2.74211556e-01
-4.75197971e-01 -9.32905078e-01 -3.88850063e-01 -9.91495252e-01
-1.37039736e-01 8.23287368e-01 3.93228948e-01 -1.41898468e-01
-6.48220628e-02 2.27557912e-01 -1.47106612e+00 -4.50676447e-03
4.31760401e-01 5.76351583e-01 -2.30346978e-01 9.18620467e-01
-8.41320902e-02 7.07393169e-01 -3.44106376e-01 -4.72228348e-01
7.09912181e-02 -5.14849067e-01 -2.25429416e-01 -3.21120590e-01
-3.98240536e-01 8.29741731e-02 -3.28734100e-01 4.92669612e-01
6.57215238e-01 3.71732682e-01 1.04532146e+00 -1.34913731e+00
-3.58097821e-01 5.92648685e-01 -8.34994912e-01 5.68147838e-01
1.55778438e-01 -1.38939828e-01 8.32009315e-01 -8.01588118e-01
2.43683934e-01 1.28756142e+00 8.14359367e-01 1.75935701e-01
-1.10707319e+00 -2.92455941e-01 -8.01866502e-02 1.73648700e-01
-9.81998563e-01 3.02204520e-01 5.09960055e-01 -5.28719127e-01
6.75668776e-01 2.44353741e-01 5.82572758e-01 1.50910175e+00
5.17273962e-01 1.03475893e+00 1.04059243e+00 -3.43920618e-01
5.58864832e-01 -1.49670556e-01 2.53368527e-01 1.21972010e-01
3.31493020e-01 1.11028790e-01 -5.29103100e-01 -7.41822064e-01
7.45458722e-01 2.98459381e-01 4.62511361e-01 -8.12670588e-02
-1.07990694e+00 9.67452586e-01 -1.49040326e-01 4.50695872e-01
-8.39835048e-01 2.64379382e-01 3.15798908e-01 -4.75267097e-02
7.26937175e-01 -1.78609207e-01 -1.55505314e-01 -3.38831961e-01
-7.67953873e-01 6.22141778e-01 8.33151460e-01 6.01950347e-01
1.48674548e-01 -2.28213772e-01 -4.87207711e-01 4.18565899e-01
2.08949596e-01 7.84045517e-01 3.54455024e-01 -2.42342651e-01
3.54340643e-01 1.28302857e-01 6.24144316e-01 -1.11269569e+00
-7.99549341e-01 -3.32777292e-01 -1.25593984e+00 -1.55701682e-01
4.81938690e-01 -4.23166245e-01 -5.80036819e-01 1.52166152e+00
7.93744981e-01 8.84984076e-01 -6.57381937e-02 4.52793539e-01
1.90276220e-01 1.19487798e+00 4.44818527e-01 -4.69237328e-01
1.25156033e+00 -5.41160941e-01 -8.84003580e-01 3.76684368e-01
1.00361131e-01 -9.20694828e-01 7.39832401e-01 5.92536867e-01
-1.04190540e+00 -3.82303953e-01 -3.46680224e-01 3.96377325e-01
-2.15741247e-01 -4.47104573e-01 6.62544727e-01 6.61968887e-01
-8.38373065e-01 6.24873340e-01 -9.35317576e-01 -4.73255128e-01
1.89133480e-01 -1.94150001e-01 2.24948496e-01 2.63015270e-01
-1.30774045e+00 5.80007076e-01 8.84189308e-02 -2.57915825e-01
-8.94493282e-01 -1.08169305e+00 -2.53990889e-01 1.54915407e-01
1.96599752e-01 -7.32805371e-01 1.28549588e+00 -2.85731196e-01
-1.40056634e+00 3.46316397e-01 -4.59947921e-02 -7.38877535e-01
9.17533398e-01 -3.25059861e-01 -8.04879904e-01 -1.07462734e-01
-9.29061994e-02 -8.55452940e-02 9.85188723e-01 -7.88017392e-01
-7.93124974e-01 -2.87916094e-01 -4.62394774e-01 1.62983090e-01
-1.81876555e-01 3.74977738e-02 1.78759217e-01 -7.60592580e-01
-3.30563515e-01 -1.06307304e+00 -2.59725153e-01 -4.64160442e-01
-2.70479232e-01 -6.09027684e-01 5.71570635e-01 -4.03975755e-01
1.35122311e+00 -1.87317407e+00 9.31930020e-02 3.59517664e-01
5.40075414e-02 1.90261975e-01 2.11012110e-01 1.29185867e+00
1.17415011e-01 -1.91921413e-01 5.53090358e-05 -5.04079163e-01
2.37022594e-01 2.33152464e-01 -5.62609673e-01 4.89445359e-01
-1.15283787e-01 6.29794121e-01 -1.06276798e+00 -2.95399904e-01
3.04390609e-01 5.09055614e-01 -2.67225742e-01 -8.03323165e-02
-1.49970815e-01 4.09893155e-01 -3.66141498e-01 4.57565933e-02
3.94775927e-01 -3.35380614e-01 -8.40244070e-02 5.62774837e-01
-2.64746070e-01 -1.22875191e-01 -1.49195254e+00 9.50120211e-01
-5.99010050e-01 3.76533598e-01 -4.91534889e-01 -7.50544071e-01
8.22271466e-01 6.53496504e-01 7.57267237e-01 -3.55210096e-01
1.23841219e-01 -1.41668916e-01 -2.20213868e-02 -2.61225700e-01
5.30121505e-01 -6.81404531e-01 -1.91298753e-01 8.49981844e-01
-1.17363170e-01 1.91176623e-01 2.77884871e-01 -3.67934890e-02
1.04489613e+00 -1.89096272e-01 4.34153110e-01 -4.80279744e-01
3.38836730e-01 -1.57715484e-01 1.61976278e-01 1.00505984e+00
1.65455759e-01 2.28678808e-01 6.49776757e-01 -5.02191663e-01
-1.30378616e+00 -1.30289137e+00 -9.62545425e-02 1.07707250e+00
-3.28067094e-02 -2.52185613e-01 -6.40310705e-01 -8.46958905e-02
3.31988782e-02 1.00771844e+00 -9.70474303e-01 -1.08046234e-01
-3.73927325e-01 -1.31428385e+00 5.49122810e-01 4.77651149e-01
8.19217563e-02 -1.18774033e+00 -6.64023638e-01 5.84287167e-01
4.56525460e-02 -6.15045607e-01 -3.39906216e-01 -1.79660693e-01
-8.46855462e-01 -8.26175392e-01 -1.00671864e+00 -2.34412596e-01
-2.91771628e-02 1.39369950e-01 1.01290238e+00 -5.18250704e-01
-3.98090303e-01 5.64790368e-01 -3.81484210e-01 -1.01350284e+00
-5.75202465e-01 -2.25772206e-02 2.10180074e-01 5.01790226e-01
4.03254002e-01 -5.57627201e-01 -6.91774607e-01 3.16700071e-01
-1.41225410e+00 -1.93233669e-01 2.07283691e-01 5.53921282e-01
5.10303736e-01 4.49650526e-01 6.43659949e-01 -7.32849360e-01
9.53166902e-01 -1.04338849e+00 -4.93069828e-01 1.67331249e-02
-3.35490942e-01 8.55000615e-02 5.54184198e-01 -5.34266710e-01
-1.31635904e+00 -3.42913628e-01 1.86875137e-03 -9.96158049e-02
-3.18732560e-01 2.32709572e-01 2.80531555e-01 7.20221639e-01
4.64963764e-01 1.28663465e-01 -3.40384334e-01 -3.51240486e-01
4.40842748e-01 3.10271382e-01 2.13015661e-01 -3.55665028e-01
8.26066494e-01 9.26567793e-01 4.79793131e-01 -9.55386460e-01
-7.39645302e-01 -7.99968302e-01 -2.80864656e-01 -2.92809069e-01
9.94490027e-01 -6.74941242e-01 -7.10345685e-01 7.56963551e-01
-8.94348383e-01 -3.35235059e-01 -3.98683459e-01 4.51853752e-01
-7.54532278e-01 -8.52419809e-03 -7.67433405e-01 -1.13092339e+00
-2.60879099e-01 -3.72983187e-01 1.03727138e+00 3.24684590e-01
-2.45053962e-01 -1.72208786e+00 8.57219875e-01 -2.18485802e-01
2.82100052e-01 4.76164132e-01 8.96699488e-01 -6.16131604e-01
1.08476784e-02 -4.25256968e-01 2.37479135e-01 -1.86026499e-01
6.02252269e-03 8.31240639e-02 -6.44345939e-01 1.10700294e-01
1.98854096e-02 4.11803603e-01 4.75671351e-01 6.10240817e-01
6.64613247e-01 -1.45175606e-01 -3.68039072e-01 1.07350005e-02
1.54966879e+00 1.13372669e-01 5.24535239e-01 2.26436585e-01
2.80894846e-01 6.11207247e-01 5.40184021e-01 1.12298632e+00
3.02655399e-01 4.88763541e-01 3.91452946e-02 2.25398187e-02
4.10750091e-01 -7.16888547e-01 1.32630080e-01 5.35839021e-01
-1.45857498e-01 -4.66975808e-01 -1.08990121e+00 5.67978024e-01
-2.26881194e+00 -1.69030440e+00 -8.28190386e-01 2.29966259e+00
4.81048912e-01 5.53725585e-02 6.86203063e-01 3.64631489e-02
8.18197906e-01 7.29103833e-02 -4.28567678e-01 -4.95703876e-01
-1.71354756e-01 2.21712053e-01 4.63519335e-01 4.83700275e-01
-1.15714753e+00 5.58720529e-01 6.70303679e+00 7.61894405e-01
-5.63098609e-01 3.66925478e-01 5.04347682e-01 -5.97498044e-02
-2.09951550e-01 -3.89504462e-01 -8.75469565e-01 6.47281528e-01
1.21253037e+00 -5.60743690e-01 7.77364825e-04 4.78122562e-01
6.24946773e-01 -1.49580628e-01 -6.92604601e-01 9.64367986e-01
-5.77839389e-02 -9.91725981e-01 -1.63758636e-01 1.17317699e-01
1.16733754e+00 -2.03069180e-01 2.90259540e-01 1.49860904e-01
5.89456797e-01 -7.56462395e-01 5.30845225e-01 9.04308259e-01
-2.27828130e-01 -1.03827965e+00 7.45748818e-01 7.63782620e-01
-9.69788790e-01 -1.10549115e-01 -3.85964364e-01 -2.53262103e-01
7.31447816e-01 1.02731824e+00 -6.36194289e-01 4.23563182e-01
5.92355132e-01 5.29439569e-01 3.68324108e-03 1.25168002e+00
3.53010036e-02 8.51810694e-01 -5.22552848e-01 -4.01805878e-01
4.75368768e-01 -3.13668042e-01 5.68897307e-01 1.06044149e+00
8.54237258e-01 1.77807391e-01 -4.50540185e-02 3.86945665e-01
3.93323570e-01 3.01931977e-01 -4.39194649e-01 -3.82952504e-02
1.35717437e-01 1.15828824e+00 -9.81291056e-01 -5.28661013e-01
-2.34530047e-01 7.82173455e-01 -2.67626196e-01 1.44342616e-01
-1.07523918e+00 2.62641281e-01 7.17236817e-01 3.49069655e-01
3.96307141e-01 -3.07399988e-01 -3.29396725e-01 -9.71238077e-01
-4.11130190e-01 -3.22093546e-01 5.98508477e-01 -5.14470100e-01
-1.55317211e+00 3.35405201e-01 3.10824394e-01 -1.20673335e+00
-3.29055756e-01 -4.51119065e-01 -8.64050329e-01 8.09066474e-01
-1.06044781e+00 -8.04104507e-01 -5.42611256e-02 6.90838695e-01
4.78325158e-01 9.12272483e-02 7.14557111e-01 1.87999204e-01
-4.03569043e-01 -8.81818756e-02 6.44033551e-01 -5.92112958e-01
4.74317044e-01 -1.47254229e+00 6.13257647e-01 5.68997264e-01
3.74491125e-01 2.59121537e-01 1.30174696e+00 -7.37830698e-01
-1.09516323e+00 -8.74405444e-01 9.67751145e-01 -6.95651770e-01
1.02097809e+00 -3.67617041e-01 -7.83936322e-01 3.40782017e-01
2.39867255e-01 -5.18624246e-01 9.82032299e-01 1.60735369e-01
8.66263434e-02 1.04299553e-01 -8.06811213e-01 7.02754498e-01
7.11592972e-01 3.51405330e-02 -4.63431150e-01 6.66819096e-01
2.97559500e-01 -3.13059315e-02 -6.71542943e-01 -5.86021915e-02
2.30001524e-01 -1.04864347e+00 9.50739086e-01 -7.91686118e-01
3.93328249e-01 -1.45379201e-01 1.71410292e-01 -1.44346428e+00
-5.79874039e-01 -1.00004995e+00 -4.12739545e-01 1.18160343e+00
1.32598922e-01 -4.41572428e-01 4.10729825e-01 3.64530683e-01
3.94897431e-01 -4.75651592e-01 -9.38396215e-01 -7.30528712e-01
2.17061520e-01 -5.61274827e-01 6.44361734e-01 4.92912412e-01
-1.55081376e-01 3.28254312e-01 -8.65314364e-01 2.71431923e-01
9.82628167e-01 1.06670246e-01 7.29612589e-01 -1.54323900e+00
-5.09806037e-01 -4.86398727e-01 -2.53575623e-01 -5.66860378e-01
-1.91497475e-01 -5.05730391e-01 -7.44558796e-02 -1.54322910e+00
2.52699465e-01 -3.68605077e-01 -4.87577856e-01 -1.42419636e-01
-3.23203206e-01 2.15835571e-01 -1.11970184e-02 1.76090688e-01
-6.26246452e-01 5.68082869e-01 7.48202503e-01 4.84639168e-01
-1.94756702e-01 5.72532952e-01 -3.72443378e-01 7.58806944e-01
9.57060575e-01 -6.05632305e-01 -4.43309635e-01 6.04469739e-02
5.20720363e-01 1.69008821e-01 3.43851388e-01 -9.79172409e-01
7.62716159e-02 -1.28033414e-01 2.43662804e-01 -6.39760852e-01
2.60093719e-01 -6.12835705e-01 6.29639268e-01 5.54419398e-01
-3.73151362e-01 4.08221900e-01 1.53556556e-01 1.16451430e+00
-7.01216012e-02 -1.29514337e-01 6.32634044e-01 4.02880739e-03
-2.41863936e-01 4.94912922e-01 -6.85018063e-01 1.03474110e-02
1.44821954e+00 4.39649969e-01 -8.13805126e-03 -8.06346118e-01
-8.17999661e-01 -2.57624853e-02 1.00061223e-01 7.17380643e-01
1.58978060e-01 -1.29257119e+00 -7.24384129e-01 -1.51448637e-01
-1.87695116e-01 -5.96506178e-01 8.57488632e-01 1.01505673e+00
-3.65512311e-01 4.63416070e-01 9.87834930e-02 -5.08688569e-01
-1.15958524e+00 8.59048724e-01 2.84802355e-02 -8.01079929e-01
-7.85739243e-01 4.14420426e-01 2.14548424e-01 1.59029998e-02
-1.36117646e-02 -6.45598397e-02 -1.78404540e-01 4.46670115e-01
9.28171039e-01 1.00794184e+00 -1.91940695e-01 -4.52729434e-01
-1.12777553e-01 2.82718927e-01 2.51785010e-01 -4.14253503e-01
1.44012403e+00 -1.49510697e-01 1.68431342e-01 1.14222050e+00
6.68210924e-01 -2.05715448e-01 -1.23026633e+00 -2.76520491e-01
2.21310735e-01 -1.24867432e-01 -3.13553125e-01 -5.15541494e-01
-1.63692936e-01 1.03096807e+00 4.43232030e-01 1.08371627e+00
7.75080204e-01 2.32832223e-01 9.89613593e-01 -1.83737606e-01
5.46623170e-01 -1.09926867e+00 6.48823977e-02 2.88092107e-01
4.64365840e-01 -8.14864039e-01 -4.38284129e-01 -8.12769681e-02
-8.88881624e-01 9.13250268e-01 -2.18553096e-01 -6.40729606e-01
1.12046301e+00 2.62976617e-01 -1.64653599e-01 1.85743123e-01
-9.80499029e-01 -1.37344897e-01 2.08695561e-01 5.79983294e-01
2.04609245e-01 3.48120272e-01 -5.26568055e-01 4.34888393e-01
-5.75282499e-02 -1.08709499e-01 4.91687536e-01 6.28751457e-01
-4.98667270e-01 -1.00764656e+00 -5.85783243e-01 3.81196409e-01
-6.64002001e-01 -3.35483849e-02 4.44423333e-02 6.13999188e-01
-5.14618084e-02 1.00850964e+00 2.64316410e-01 2.95299470e-01
3.61552775e-01 2.28278130e-01 3.13344657e-01 -2.49180987e-01
-4.93233263e-01 2.19522759e-01 -2.67999947e-01 -1.14336178e-01
-3.75504971e-01 -1.19251335e+00 -9.10525501e-01 -4.31372970e-01
1.52182296e-01 1.67451054e-01 5.28998017e-01 9.75790501e-01
2.73498863e-01 5.24881065e-01 6.11169517e-01 -7.30866611e-01
-7.10291207e-01 -9.98749316e-01 -8.80368471e-01 4.13163722e-01
2.46208217e-02 -6.99345231e-01 -1.93080246e-01 -5.63378744e-02] | [6.919283866882324, 3.462026596069336] |
80185037-c7ea-4a63-ba5e-b94d3977d67d | are-quantum-computers-practical-yet-a-case | 2205.04490 | null | https://arxiv.org/abs/2205.04490v2 | https://arxiv.org/pdf/2205.04490v2.pdf | Are Quantum Computers Practical Yet? A Case for Feature Selection in Recommender Systems using Tensor Networks | Collaborative filtering models generally perform better than content-based filtering models and do not require careful feature engineering. However, in the cold-start scenario collaborative information may be scarce or even unavailable, whereas the content information may be abundant, but also noisy and expensive to acquire. Thus, selection of particular features that improve cold-start recommendations becomes an important and non-trivial task. In the recent approach by Nembrini et al., the feature selection is driven by the correlational compatibility between collaborative and content-based models. The problem is formulated as a Quadratic Unconstrained Binary Optimization (QUBO) which, due to its NP-hard complexity, is solved using Quantum Annealing on a quantum computer provided by D-Wave. Inspired by the reported results, we contend the idea that current quantum annealers are superior for this problem and instead focus on classical algorithms. In particular, we tackle QUBO via TTOpt, a recently proposed black-box optimizer based on tensor networks and multilinear algebra. We show the computational feasibility of this method for large problems with thousands of features, and empirically demonstrate that the solutions found are comparable to the ones obtained with D-Wave across all examined datasets. | ['Evgeny Frolov', 'Ivan Oseledets', 'Rafael Ballester-Ripoll', 'Andrei Chertkov', 'Artyom Nikitin'] | 2022-05-09 | null | null | null | null | ['tensor-networks'] | ['methodology'] | [-1.23061366e-01 -2.52587050e-01 2.21142266e-03 -2.98503280e-01
-6.58063293e-01 -7.69028068e-01 4.37493324e-01 4.33649182e-01
-6.43013835e-01 6.75406873e-01 -1.01770572e-01 -2.54528731e-01
-8.38502765e-01 -9.90314245e-01 -4.28057164e-01 -8.53776515e-01
-1.36489823e-01 7.29404807e-01 -1.87565297e-01 -6.16165996e-01
3.79454970e-01 4.60620761e-01 -1.59841740e+00 -1.68890730e-02
1.15844035e+00 1.21047664e+00 -4.53773737e-02 4.47983563e-01
-6.95703700e-02 1.82098523e-01 -1.45956948e-01 -9.18174088e-01
6.60185397e-01 -3.29372019e-01 -8.05121779e-01 -8.05542916e-02
2.93916136e-01 3.76260370e-01 -1.96181938e-01 1.33567226e+00
4.59268242e-01 5.61505198e-01 4.44057852e-01 -9.73452032e-01
-5.17532408e-01 5.95212996e-01 2.94184480e-02 4.88900393e-02
3.28210920e-01 -2.60858294e-02 1.56935883e+00 -8.64553213e-01
8.40110779e-01 7.59004056e-01 4.20574129e-01 1.98271498e-01
-1.52608740e+00 -2.32300267e-01 -6.32975176e-02 6.29698634e-01
-1.51992953e+00 -2.82573223e-01 6.46890402e-01 -4.79097992e-01
4.49667841e-01 5.53306639e-01 9.32653308e-01 8.81208420e-01
1.90112516e-01 3.80170196e-01 1.33982897e+00 -4.17741120e-01
5.34921646e-01 3.24064702e-01 2.82261580e-01 6.60786033e-01
4.32618558e-01 2.60132223e-01 -8.07293296e-01 -4.83543456e-01
-8.74049366e-02 8.56264904e-02 -2.37056822e-01 -5.23278594e-01
-1.15752113e+00 1.06349277e+00 5.76251268e-01 4.47945863e-01
-5.69205880e-01 -2.24962413e-01 6.84937015e-02 6.08060479e-01
5.38438022e-01 9.64559317e-01 -3.72149646e-01 -3.47912945e-02
-9.64468539e-01 4.48667198e-01 1.15247238e+00 5.46945035e-01
9.97772813e-01 -4.69387889e-01 -1.37633421e-02 3.16354573e-01
3.50245684e-01 4.15304303e-01 1.58096299e-01 -5.71894586e-01
2.15793669e-01 3.84016097e-01 3.95051003e-01 -1.21026874e+00
-6.47874296e-01 -9.08718467e-01 -9.05350626e-01 -4.61960398e-02
6.66691065e-01 -3.13774526e-01 -3.27195674e-01 1.32737696e+00
6.67921960e-01 -3.19646984e-01 8.99443217e-03 1.30310476e+00
4.72245872e-01 3.44103038e-01 -2.57741213e-01 -4.53114629e-01
1.26617503e+00 -8.60419810e-01 -7.28178263e-01 2.25339949e-01
7.52970278e-01 -1.01339126e+00 5.03423214e-01 8.54152679e-01
-8.06309342e-01 -2.66873717e-01 -1.01540148e+00 2.24658266e-01
-5.19266188e-01 6.35451749e-02 1.05733550e+00 1.13494217e+00
-7.72438407e-01 1.13964236e+00 -4.21782434e-01 -3.39412957e-01
6.01212122e-02 5.66215277e-01 -3.39504391e-01 -7.27254823e-02
-1.29498303e+00 8.55068207e-01 4.43175375e-01 4.94986027e-01
-3.25087011e-01 -3.70337993e-01 -9.88067687e-02 4.72371019e-02
7.58148730e-01 -8.88056338e-01 7.62772381e-01 -8.97923648e-01
-1.91066372e+00 2.84482747e-01 1.22869492e-01 -4.50938731e-01
4.32953775e-01 -9.30493325e-02 -2.99680263e-01 -1.30778365e-02
-3.30198348e-01 -7.06666261e-02 7.58193195e-01 -7.45906293e-01
-3.40740293e-01 -4.35541153e-01 3.92918468e-01 7.30344579e-02
-4.46873456e-01 -7.94664547e-02 3.07005248e-03 -3.01650494e-01
4.25561786e-01 -1.23705542e+00 -7.51836658e-01 -3.61534923e-01
-3.96749318e-01 -1.45289108e-01 8.44671577e-02 -2.61737704e-01
1.14288449e+00 -1.80785263e+00 5.73800266e-01 8.01029861e-01
2.53468841e-01 1.65140107e-01 1.87906325e-02 9.32597518e-01
1.10002778e-01 1.08559519e-01 8.81758034e-02 -1.25235453e-01
2.96918571e-01 3.30152959e-02 -2.73184478e-02 8.84008288e-01
-1.98134258e-01 6.54909670e-01 -7.73672760e-01 -3.55010808e-01
5.52036837e-02 7.52967373e-02 -9.62882459e-01 -8.52704868e-02
-2.56390184e-01 7.01147020e-01 -6.56967759e-01 5.81877530e-01
8.55245233e-01 -4.34637398e-01 4.30137962e-01 -3.98773283e-01
-3.37360442e-01 1.25595570e-01 -1.64479554e+00 1.82907212e+00
-4.53856647e-01 1.73717916e-01 1.39768839e-01 -1.14332557e+00
7.88928926e-01 8.73734951e-02 6.01945758e-01 -6.31503940e-01
2.81730771e-01 4.40348208e-01 3.29486161e-01 -3.65398914e-01
7.46774852e-01 -1.51371643e-01 -1.72585975e-02 4.74244475e-01
2.99622267e-01 -1.63768455e-01 4.71644968e-01 2.15830386e-01
1.06533849e+00 -7.32218325e-02 1.38059825e-01 -4.52378958e-01
6.04904354e-01 2.03756943e-01 4.09674466e-01 1.08830714e+00
1.43703923e-01 3.35829467e-01 3.27896506e-01 -2.81457633e-01
-7.59722650e-01 -5.89285612e-01 -3.23424906e-01 9.21050370e-01
2.94389278e-01 -1.07810974e+00 -4.83535588e-01 -5.85152984e-01
-6.86185583e-02 4.05877948e-01 -4.07342374e-01 -4.17145086e-04
-1.19605616e-01 -1.20614243e+00 -1.30480796e-01 -4.39732164e-01
1.10701220e-02 -3.93713772e-01 -2.89606243e-01 5.26804268e-01
-3.05516310e-02 -8.42501044e-01 4.24458422e-02 3.64173204e-01
-8.60069036e-01 -1.03991008e+00 -3.93916100e-01 1.19668014e-01
4.28960025e-01 1.15512490e-01 1.01949918e+00 1.03324808e-01
-2.74701089e-01 1.61009535e-01 -9.38222528e-01 -6.94723353e-02
-1.71542421e-01 3.83033276e-01 2.34612122e-01 4.89122003e-01
2.55696595e-01 -4.49627250e-01 -7.39792287e-01 3.26753914e-01
-7.94485390e-01 -1.40391216e-01 6.62451148e-01 1.05460930e+00
4.65296328e-01 2.81807274e-01 3.06888461e-01 -9.56006587e-01
5.07955968e-01 -5.16258419e-01 -8.69152844e-01 2.21319735e-01
-8.78872037e-01 3.46542746e-01 9.38678205e-01 -1.12240165e-01
-7.02393174e-01 2.33953789e-01 -2.50040591e-01 1.22102328e-01
1.33234188e-01 7.96509445e-01 8.44875649e-02 -7.15084195e-01
7.84111023e-01 -9.90428217e-03 -2.88097262e-01 -6.78606749e-01
5.82616150e-01 6.04075134e-01 -3.11816216e-01 -5.63346744e-01
1.00727260e+00 5.19867897e-01 5.02295256e-01 -8.58585358e-01
-1.01349378e+00 -4.39725310e-01 -6.46275699e-01 -2.58523881e-01
6.28165841e-01 -5.26920557e-01 -1.14535546e+00 -2.30657496e-02
-8.67523074e-01 3.89997035e-01 -1.51237756e-01 8.79121900e-01
-3.68345171e-01 5.62132895e-01 -2.24241629e-01 -9.29466903e-01
-3.15197617e-01 -1.21855795e+00 4.87196296e-01 2.02744707e-01
2.09497988e-01 -6.99156165e-01 1.10807694e-01 6.33991838e-01
5.27150452e-01 -3.71541120e-02 7.75549591e-01 -8.24570656e-01
-7.98340857e-01 -3.69951576e-01 -5.47756180e-02 6.79203793e-02
-2.97837883e-01 -9.08636898e-02 -5.86999655e-01 -3.82837564e-01
-9.99643840e-03 -1.23580359e-01 6.87322497e-01 -1.02819517e-01
7.83599079e-01 -7.78760612e-02 2.09576320e-02 5.24991035e-01
1.63988936e+00 -3.08135986e-01 1.44254595e-01 2.45432884e-01
4.02041554e-01 7.07982302e-01 8.38620484e-01 7.19187856e-01
2.49052033e-01 1.07333446e+00 5.61174214e-01 2.29667559e-01
3.85559618e-01 1.87327549e-01 4.24818657e-02 1.11292136e+00
-3.87960225e-01 -1.01685025e-01 -6.41382933e-01 4.46092077e-02
-2.10204720e+00 -8.25646520e-01 -6.16063416e-01 2.42581701e+00
6.51388586e-01 -4.57066968e-02 1.11818761e-01 -7.13308677e-02
4.32865649e-01 -1.40600666e-01 -1.77951485e-01 -3.79125267e-01
-2.79080868e-01 5.53591430e-01 6.81865692e-01 3.24569702e-01
-1.07375443e+00 6.63103819e-01 5.29876089e+00 9.47546065e-01
-8.97816479e-01 5.59747696e-01 1.62009105e-01 -3.39120686e-01
-2.77829021e-01 4.68973219e-01 -5.97757578e-01 4.98260140e-01
1.03259158e+00 -1.49546534e-01 7.36731231e-01 6.35622203e-01
1.21448211e-01 -4.10444260e-01 -9.12133515e-01 1.12167108e+00
-1.64845183e-01 -1.37005079e+00 -3.68055195e-01 1.81384429e-01
1.02987051e+00 1.73635706e-01 3.80340181e-02 1.97423380e-02
1.08762152e-01 -7.32856393e-01 6.36446357e-01 9.05611992e-01
-9.45678912e-03 -7.51744926e-01 8.03111494e-01 2.93388009e-01
-7.98953235e-01 -2.69418091e-01 -6.47623837e-01 -7.51666129e-02
1.80060461e-01 1.18697715e+00 -1.67457536e-01 1.22114170e+00
5.45510948e-01 4.20846850e-01 -4.48892027e-01 1.55596566e+00
-8.15989152e-02 5.76066554e-01 -5.35527527e-01 -5.97210109e-01
4.14065480e-01 -9.21318948e-01 7.43063748e-01 7.42536306e-01
4.84142035e-01 -4.38952036e-02 1.52647838e-01 7.59199142e-01
9.90230963e-02 8.29939187e-01 -2.53488809e-01 -2.10769713e-01
-6.67538643e-02 1.46896207e+00 -6.65420234e-01 -1.35681957e-01
-2.77860284e-01 8.04953754e-01 2.86968499e-01 1.82606161e-01
-5.82092285e-01 -4.07820851e-01 5.21064341e-01 -4.65958416e-01
5.08437753e-01 -4.35475826e-01 -1.42978474e-01 -1.62137282e+00
-1.95510779e-03 -8.74924958e-01 2.29147285e-01 -1.71101972e-01
-1.35974693e+00 6.22045636e-01 -4.07061070e-01 -1.33577037e+00
-1.40581233e-03 -6.29173219e-01 1.56181501e-02 8.89311492e-01
-1.34178710e+00 -6.39746130e-01 4.13912870e-02 5.32069445e-01
-1.27610460e-01 1.75714105e-01 8.85848463e-01 5.60041428e-01
-5.27139366e-01 3.12790424e-01 8.87519658e-01 -5.93733370e-01
6.27773821e-01 -1.32612836e+00 -1.71104819e-01 7.50942171e-01
5.10447145e-01 7.69425094e-01 1.09987593e+00 -4.37845737e-01
-2.33709764e+00 -5.96617997e-01 9.76812601e-01 -1.84613138e-01
1.02858269e+00 -7.13793933e-01 -4.27121103e-01 -9.73441228e-02
-9.83536243e-03 7.85991475e-02 9.73832905e-01 6.13829076e-01
-7.12831318e-02 -1.78768545e-01 -9.96967316e-01 3.58356208e-01
1.12523890e+00 -4.51556623e-01 -2.44879469e-01 9.98521686e-01
2.25709721e-01 -7.79610053e-02 -1.17454875e+00 -2.09483504e-02
4.65510219e-01 -1.09368479e+00 6.95900679e-01 -6.73897564e-01
-4.10907576e-03 -2.93699205e-01 -2.72493184e-01 -1.42320406e+00
-2.59715289e-01 -9.46932197e-01 1.70309708e-01 7.64756262e-01
5.42069137e-01 -6.71970785e-01 7.33588636e-01 4.83636826e-01
1.85730457e-01 -6.64163411e-01 -1.19428074e+00 -7.60336816e-01
-1.76220745e-01 -4.62693304e-01 5.55376589e-01 8.80874991e-01
1.71457380e-01 2.75531143e-01 -6.75067246e-01 1.30178809e-01
6.44367099e-01 7.07633018e-01 6.40355051e-01 -1.59140730e+00
-7.58198798e-01 -1.67113215e-01 -3.91081125e-01 -9.00731742e-01
-7.39400238e-02 -1.20341110e+00 -3.00882369e-01 -1.09016502e+00
-8.72752815e-02 -8.52582693e-01 -3.21037471e-01 5.08370139e-02
1.24566905e-01 3.36419880e-01 3.55226636e-01 2.31955945e-01
-8.27900648e-01 6.04218662e-01 1.16820002e+00 -6.65835813e-02
-1.72229320e-01 3.00124288e-01 -4.65635985e-01 2.91425139e-01
4.18870181e-01 -6.83873117e-01 9.52156633e-02 -4.84374836e-02
1.22172570e+00 1.62285194e-01 9.15940106e-02 -8.77963305e-01
5.91274977e-01 -1.65211260e-01 -8.18158463e-02 -4.52320546e-01
4.65408832e-01 -9.85986531e-01 3.51035237e-01 3.73052806e-01
-1.53420359e-01 -1.78192586e-01 -5.04246175e-01 6.81596339e-01
-1.58651844e-01 -7.74384916e-01 4.06779230e-01 -5.77407368e-02
-3.62348288e-01 3.59878808e-01 -4.05416608e-01 -4.18990761e-01
7.48092055e-01 2.23658651e-01 -2.61001736e-01 -2.71172851e-01
-1.11943841e+00 -3.89496684e-02 2.26223975e-01 3.21786068e-02
2.25707844e-01 -1.11316884e+00 -6.42739892e-01 -7.46874586e-02
1.27704129e-01 -6.35942817e-01 5.31547844e-01 1.48003364e+00
-3.29822272e-01 5.26910186e-01 1.76091529e-02 -2.90924639e-01
-9.24523592e-01 7.64051259e-01 2.39603132e-01 -2.63476253e-01
-2.10008398e-01 7.57969141e-01 -5.28028071e-01 -4.36248332e-01
-2.80440331e-01 3.94766964e-02 -1.72399864e-01 4.93829072e-01
1.61002725e-01 4.75647539e-01 6.26566172e-01 -6.05582297e-01
-3.66779238e-01 4.67474431e-01 1.24593928e-01 -1.52005972e-02
1.39408565e+00 -1.43265858e-01 -5.16442895e-01 1.34218827e-01
1.19533527e+00 2.18157977e-01 -3.49843770e-01 -4.27579701e-01
9.50975996e-03 -6.82126045e-01 3.58598441e-01 -6.56863570e-01
-9.41426218e-01 5.02292871e-01 6.93696618e-01 6.61722064e-01
9.22317088e-01 -1.56773612e-01 5.12175918e-01 9.41224337e-01
6.99702263e-01 -1.19798791e+00 -3.70779872e-01 5.19972563e-01
5.79096556e-01 -1.19397771e+00 2.08122164e-01 -5.06305635e-01
-2.28802279e-01 1.31738770e+00 1.80452957e-03 -2.14453474e-01
8.41599882e-01 -4.29673463e-01 -3.14818531e-01 -2.80471891e-01
-7.19987452e-01 -6.84593737e-01 4.96164441e-01 -5.89648969e-02
2.01694459e-01 1.88075438e-01 -1.03460765e+00 4.44121987e-01
-3.56425285e-01 -1.80457368e-01 5.40882826e-01 7.18508422e-01
-1.09780923e-01 -1.68060815e+00 -3.90074462e-01 6.42632186e-01
-5.19734681e-01 -2.32518867e-01 -2.56917983e-01 2.82539815e-01
2.68302649e-01 1.36175108e+00 -5.17538369e-01 -6.55456483e-01
9.30337608e-02 -2.50266269e-02 5.18956244e-01 -4.47566420e-01
-1.05136526e+00 -1.30519375e-01 2.82204032e-01 -6.30619705e-01
-5.70114255e-01 -6.71406448e-01 -5.13265133e-01 -2.81924397e-01
-1.02934635e+00 8.09127450e-01 1.22727942e+00 1.08505356e+00
4.70591843e-01 5.13118282e-02 9.59170640e-01 -7.73154140e-01
-9.32243109e-01 -6.57959223e-01 -7.92621076e-01 2.37327665e-01
3.83820944e-02 -8.44830513e-01 -4.91832405e-01 -6.71465695e-01] | [5.914766311645508, 4.830348491668701] |
ef6a596a-22fd-4d8d-ba09-faf6761a2126 | targeted-syntactic-evaluation-of-language | 1808.09031 | null | http://arxiv.org/abs/1808.09031v1 | http://arxiv.org/pdf/1808.09031v1.pdf | Targeted Syntactic Evaluation of Language Models | We present a dataset for evaluating the grammaticality of the predictions of
a language model. We automatically construct a large number of minimally
different pairs of English sentences, each consisting of a grammatical and an
ungrammatical sentence. The sentence pairs represent different variations of
structure-sensitive phenomena: subject-verb agreement, reflexive anaphora and
negative polarity items. We expect a language model to assign a higher
probability to the grammatical sentence than the ungrammatical one. In an
experiment using this data set, an LSTM language model performed poorly on many
of the constructions. Multi-task training with a syntactic objective (CCG
supertagging) improved the LSTM's accuracy, but a large gap remained between
its performance and the accuracy of human participants recruited online. This
suggests that there is considerable room for improvement over LSTMs in
capturing syntax in a language model. | ['Tal Linzen', 'Rebecca Marvin'] | 2018-08-27 | targeted-syntactic-evaluation-of-language-1 | https://aclanthology.org/D18-1151 | https://aclanthology.org/D18-1151.pdf | emnlp-2018-10 | ['ccg-supertagging'] | ['natural-language-processing'] | [ 5.21175936e-02 5.81803620e-01 8.72878656e-02 -8.66094291e-01
-9.67007697e-01 -6.23760641e-01 5.00455618e-01 2.15833977e-01
-6.54958248e-01 6.70443714e-01 5.42954564e-01 -6.52880609e-01
2.39877596e-01 -8.48153651e-01 -7.75050879e-01 -3.19694519e-01
-5.92633560e-02 8.11487734e-01 -2.47061476e-02 -2.76246399e-01
1.82421684e-01 1.53781548e-02 -9.39617276e-01 8.86843443e-01
6.76200628e-01 6.84110224e-01 7.19689131e-01 3.43796819e-01
-3.70632619e-01 9.38964784e-01 -5.67491353e-01 -8.05864871e-01
-2.76305042e-02 -2.69275010e-01 -1.22681510e+00 -4.10151154e-01
9.93544281e-01 -2.60949135e-02 7.57651255e-02 1.03125381e+00
3.30020577e-01 -4.28105295e-02 3.15442324e-01 -7.73842633e-01
-5.79140484e-01 1.05445278e+00 -1.71003237e-01 5.23399115e-01
4.92037535e-01 -1.22883245e-01 1.74695301e+00 -7.73855448e-01
7.42686987e-01 1.85777545e+00 7.58442819e-01 6.39654458e-01
-1.44232786e+00 -5.33812404e-01 2.44471401e-01 -8.90194774e-02
-8.11994016e-01 -4.73405510e-01 5.00267446e-01 -2.54392028e-01
1.68990719e+00 6.37123287e-02 4.03841645e-01 1.44651210e+00
7.96701610e-01 4.78467494e-01 1.37003911e+00 -4.37443048e-01
-1.78909510e-01 4.85790633e-02 3.89830440e-01 7.09195673e-01
2.15098187e-01 2.07968861e-01 -5.34951389e-01 -4.29927051e-01
3.85186911e-01 -6.16073728e-01 -5.75738847e-02 1.04580134e-01
-1.20879817e+00 8.35882425e-01 4.06605512e-01 8.17385256e-01
-2.68140435e-01 2.08071128e-01 5.80212235e-01 7.41172373e-01
5.16500711e-01 6.50198579e-01 -8.55772018e-01 1.66042656e-01
-5.50594270e-01 4.97810811e-01 8.91629457e-01 6.61187291e-01
6.11968219e-01 3.27627473e-02 9.20799468e-03 8.41332972e-01
5.48520148e-01 3.89729023e-01 7.83032358e-01 -1.01926994e+00
1.00630403e+00 3.38802636e-01 6.42205104e-02 -1.18790746e+00
-8.18253994e-01 -1.35960996e-01 -5.25139049e-02 -4.53627914e-01
8.28592777e-01 -1.06636956e-01 -3.58710796e-01 2.39252281e+00
-2.16085374e-01 -5.36320686e-01 9.61104184e-02 6.17893040e-01
7.28121102e-01 6.24897420e-01 8.51513028e-01 -2.53208399e-01
1.56710625e+00 -2.80023545e-01 -6.15306437e-01 -1.17769337e+00
1.29953253e+00 -7.04875171e-01 1.51720035e+00 -2.84387004e-02
-1.49118650e+00 -5.07423937e-01 -7.13039696e-01 -3.28950375e-01
-8.07528198e-02 -3.05978090e-01 6.84442818e-01 4.10213500e-01
-1.25469601e+00 6.34995282e-01 -5.80359697e-01 -6.21720195e-01
-3.26833688e-02 3.60156715e-01 -6.68313980e-01 4.56469841e-02
-1.59348536e+00 1.58090878e+00 7.47035086e-01 -4.96537378e-03
-4.39244181e-01 -1.86920688e-01 -1.14983749e+00 1.74849242e-01
-2.56237328e-01 -5.82493424e-01 1.56909990e+00 -1.45602465e+00
-8.49127769e-01 1.77371037e+00 -3.03391993e-01 -4.69370693e-01
-1.52475074e-01 -8.02550018e-02 -3.38962525e-01 -1.54586136e-01
2.64175832e-01 6.98264122e-01 4.82162654e-01 -9.03484821e-01
-5.65064311e-01 -6.46420121e-01 -3.46950069e-02 3.63606453e-01
-2.32302040e-01 6.70679927e-01 6.15209579e-01 -5.27642310e-01
2.28503153e-01 -6.96332097e-01 2.45448634e-01 -5.43130219e-01
-3.53988737e-01 -8.99616480e-01 2.18954608e-01 -8.78673673e-01
9.59309220e-01 -1.97890496e+00 1.61524415e-02 -1.85136184e-01
-5.52154006e-03 -6.23836964e-02 -5.01951039e-01 4.99284893e-01
-3.12293082e-01 5.22326827e-01 -2.97494102e-02 -5.00632107e-01
7.42648020e-02 3.19617659e-01 -4.85708833e-01 3.41914594e-01
1.50797129e-01 1.19177759e+00 -9.02313590e-01 -4.85298634e-01
-3.64885777e-01 -1.41809270e-01 -4.74906057e-01 8.04111511e-02
-4.68712628e-01 3.90006825e-02 -1.99026465e-01 3.39568466e-01
3.39559585e-01 -3.55144441e-02 6.14759147e-01 -2.14830823e-02
6.76568598e-02 1.40659618e+00 -4.09164965e-01 1.48715556e+00
-5.50723433e-01 5.34669042e-01 2.08411068e-01 -1.00504148e+00
7.92439699e-01 6.46448016e-01 -2.23296851e-01 -1.02787268e+00
2.13696927e-01 5.63155711e-01 8.22030365e-01 -6.01536214e-01
3.09434175e-01 -7.65162170e-01 -3.87766778e-01 7.83216536e-01
2.69748807e-01 -1.67189389e-02 6.67751059e-02 1.24766819e-01
9.26324606e-01 7.95573443e-02 2.84687430e-01 -7.59944022e-01
2.31076822e-01 -2.55314261e-01 7.76394188e-01 8.31847727e-01
-1.84742168e-01 1.21197909e-01 5.65318048e-01 -6.37219250e-01
-9.00670886e-01 -1.14184105e+00 -1.69812307e-01 1.60417283e+00
-4.94545102e-01 -4.76457387e-01 -8.54676723e-01 -7.12457061e-01
-1.76730469e-01 1.26169634e+00 -3.55005771e-01 -2.65143335e-01
-1.09461689e+00 -7.97352850e-01 7.10123956e-01 4.97701257e-01
1.76912904e-01 -1.70361173e+00 -4.66686040e-01 5.54733276e-01
-6.16543591e-01 -1.20477009e+00 -2.91744232e-01 2.75798440e-01
-1.05272090e+00 -7.75458336e-01 2.55476266e-01 -1.24371850e+00
4.10376728e-01 -1.96596429e-01 1.77093875e+00 3.03296089e-01
3.46417934e-01 -7.15189613e-03 2.33532805e-02 -3.32788706e-01
-1.02680326e+00 1.45928115e-01 3.88181880e-02 -4.60018903e-01
7.93429911e-01 -5.38890958e-01 -8.41082335e-02 -1.30500808e-01
-5.12665749e-01 -4.89766188e-02 3.35569292e-01 8.19579661e-01
-2.96092108e-02 -6.16644084e-01 6.85358763e-01 -1.06982565e+00
9.88842070e-01 -4.71660733e-01 -1.70208499e-01 2.06757471e-01
-2.78604835e-01 1.28703266e-01 6.78476512e-01 -3.15548062e-01
-1.18419933e+00 -1.65043250e-01 -2.53921926e-01 3.75143647e-01
-1.52398780e-01 6.53005123e-01 -1.85899720e-01 3.29062313e-01
6.34400427e-01 -5.28508872e-02 -7.89628476e-02 -4.86548513e-01
6.46408126e-02 4.09966141e-01 3.88982654e-01 -1.03539479e+00
3.58133346e-01 -6.11688346e-02 -3.24972451e-01 -6.33166075e-01
-1.23153806e+00 1.94213882e-01 -4.71749842e-01 3.18585962e-01
7.68256068e-01 -8.36457849e-01 -6.42903149e-01 3.63979548e-01
-1.59634650e+00 -6.24755561e-01 7.61119183e-03 4.18140709e-01
-5.17706752e-01 2.14971483e-01 -1.13070989e+00 -3.74454796e-01
-4.25116360e-01 -8.50750268e-01 9.85588312e-01 -3.99954706e-01
-8.95969272e-01 -1.45361316e+00 1.57734938e-02 2.12115258e-01
4.56325442e-01 -2.07890674e-01 1.55025434e+00 -1.15013194e+00
-1.30338475e-01 -2.97564026e-02 -6.47373274e-02 1.51550591e-01
7.01098517e-02 -2.96924651e-01 -1.02212262e+00 -3.52683544e-01
5.25183558e-01 -7.69443631e-01 8.90153885e-01 4.24094498e-01
6.54936731e-01 -7.20199823e-01 -2.82757044e-01 1.87999204e-01
1.13912082e+00 7.37692937e-02 6.29901469e-01 2.21900925e-01
4.80118841e-01 8.58426690e-01 3.07029426e-01 -2.86891639e-01
7.56056666e-01 5.43157041e-01 7.33337104e-02 2.58469760e-01
1.84310824e-01 -3.58781427e-01 9.36591148e-01 7.44755089e-01
6.45054460e-01 -6.31881297e-01 -1.04984379e+00 5.61910093e-01
-1.53280377e+00 -1.06550217e+00 -3.01216900e-01 2.07778692e+00
1.10450029e+00 4.31246102e-01 -1.56533018e-01 -1.52010813e-01
5.95286727e-01 5.14807880e-01 -5.72642218e-03 -1.08401954e+00
-2.25849539e-01 3.14230472e-01 1.53111652e-01 9.69555020e-01
-1.08063102e+00 1.51363587e+00 6.79793167e+00 4.09985870e-01
-1.19053972e+00 1.73407987e-01 6.43415451e-01 2.38544732e-01
-5.24997771e-01 1.20147005e-01 -8.87759566e-01 4.37354773e-01
1.60227561e+00 -1.43475085e-01 2.81919777e-01 4.19012517e-01
3.67811650e-01 -2.40050957e-01 -1.50470424e+00 2.65228629e-01
9.28062871e-02 -9.25392210e-01 1.79538339e-01 -7.93636963e-03
1.15537927e-01 3.78407449e-01 -2.13215947e-01 3.83099467e-01
3.49438310e-01 -1.03970599e+00 9.18567181e-01 1.47573143e-01
3.98200631e-01 -1.76171690e-01 5.30231893e-01 6.61538720e-01
-7.45969415e-01 -4.34096530e-02 -4.74272102e-01 -6.31614447e-01
3.26100856e-01 7.29478449e-02 -7.28302956e-01 -2.81409442e-01
5.67222536e-01 3.43246132e-01 -7.80279338e-01 4.04384345e-01
-4.62477803e-01 8.09659779e-01 -3.70031685e-01 -2.37548307e-01
2.34294534e-01 -9.28402394e-02 5.87564170e-01 1.35459197e+00
-2.08537076e-02 -3.41042727e-02 2.32393995e-01 1.00083697e+00
-2.68098921e-01 2.50606179e-01 -9.28622007e-01 -5.05212545e-02
6.62726045e-01 1.16587293e+00 -4.51446652e-01 -3.53541851e-01
-6.31846964e-01 7.24091470e-01 8.35859954e-01 1.49905190e-01
-3.46880168e-01 7.52974376e-02 4.03078020e-01 1.54833287e-01
-1.70736194e-01 -2.35710934e-01 -4.40979183e-01 -1.05428469e+00
2.18837470e-01 -7.35527813e-01 4.83309090e-01 -9.23392534e-01
-1.47625446e+00 5.81620693e-01 -5.41416928e-02 -4.08655554e-01
-7.40340948e-01 -8.39800775e-01 -7.13159621e-01 1.04118049e+00
-1.12330770e+00 -1.14510059e+00 2.55235136e-01 9.09577534e-02
3.89791459e-01 6.81185201e-02 1.25025427e+00 9.59208384e-02
-3.00884247e-01 4.33217764e-01 -4.39580530e-01 2.22774982e-01
7.27728307e-01 -1.32830966e+00 9.04569328e-01 6.43478334e-01
2.73697436e-01 9.93622720e-01 7.62710035e-01 -5.71142793e-01
-8.17669988e-01 -8.91477823e-01 1.97601986e+00 -6.81656659e-01
8.03308070e-01 -5.94960868e-01 -1.14764953e+00 1.48342550e+00
5.31056762e-01 -2.63330400e-01 5.14047325e-01 3.40776950e-01
-3.36761653e-01 2.22036228e-01 -1.14287174e+00 4.82949615e-01
1.11307108e+00 -7.88287222e-01 -1.30938637e+00 5.94772398e-01
6.73451424e-01 -2.85868824e-01 -6.68577373e-01 2.52729207e-01
2.80762166e-01 -8.83973718e-01 6.39177501e-01 -1.17639506e+00
5.39712906e-01 3.38387549e-01 -2.63763607e-01 -1.44633639e+00
-5.54038584e-01 -1.21378452e-01 1.90941215e-01 1.07630372e+00
7.18401372e-01 -8.42535257e-01 3.07430059e-01 9.80591238e-01
-2.62225270e-01 -4.68415052e-01 -1.09364593e+00 -6.91235185e-01
8.52117419e-01 -2.59699821e-01 1.15673013e-01 9.56602573e-01
3.72342765e-01 8.41774762e-01 2.50060588e-01 -1.95458025e-01
3.36822420e-01 2.08784565e-01 3.54823060e-02 -1.46738565e+00
-7.92205036e-02 -3.03237706e-01 1.99475773e-02 -7.23385930e-01
1.07988346e+00 -1.46848881e+00 -7.63492808e-02 -1.58182800e+00
2.92715043e-01 -2.35078081e-01 9.45891216e-02 7.97214746e-01
-1.88614786e-01 -1.74428776e-01 1.59429163e-01 1.19912927e-03
-3.70104998e-01 2.24954039e-01 8.22891414e-01 3.78044099e-02
6.12735413e-02 -3.03518325e-01 -9.03134227e-01 1.08395934e+00
9.01938617e-01 -7.21645296e-01 8.40132404e-03 -9.46709454e-01
5.55888653e-01 1.46959513e-01 3.57953519e-01 -4.56926465e-01
-2.61549056e-01 2.59459782e-02 2.41721720e-01 1.34324476e-01
1.90649316e-01 -3.99533600e-01 -3.66742432e-01 6.04659617e-01
-7.28874803e-01 6.23070121e-01 4.48778450e-01 -1.87480310e-03
-1.50721334e-02 -3.67341578e-01 8.76235723e-01 -7.33851790e-01
-4.82875258e-01 -3.07859093e-01 -7.23771095e-01 5.53142369e-01
3.26521635e-01 8.67721587e-02 -4.48808342e-01 -2.33172983e-01
-8.49090040e-01 -8.95712525e-03 3.36975843e-01 6.64247036e-01
4.75931972e-01 -1.14101863e+00 -7.64256597e-01 1.06102012e-01
-1.56995997e-01 -4.21245128e-01 -3.72516811e-01 6.46446288e-01
-1.59628943e-01 5.74763179e-01 -1.69041231e-02 -1.86162964e-01
-1.09625590e+00 1.57214910e-01 8.75478804e-01 -3.68434638e-01
-3.49572778e-01 9.89552975e-01 3.33634436e-01 -7.99982071e-01
-1.05847806e-01 -5.87327957e-01 -8.36087242e-02 9.99544263e-02
3.21801335e-01 -2.35022187e-01 1.60502508e-01 -9.76261795e-01
-3.17254812e-01 1.04099669e-01 -3.25406134e-01 -1.31408975e-01
1.18114901e+00 -3.40473168e-02 -4.44535434e-01 9.21961486e-01
1.23371887e+00 -5.95475882e-02 -5.48393488e-01 -2.35648751e-01
6.99288368e-01 -1.03064902e-01 -1.84157103e-01 -8.40115666e-01
-6.25094771e-01 8.37517500e-01 1.94070771e-01 2.00651497e-01
3.75074178e-01 2.82762825e-01 9.68240917e-01 7.09344566e-01
5.61084807e-01 -1.03414989e+00 -2.03205302e-01 9.56744492e-01
1.07505155e+00 -1.14273059e+00 -4.34740841e-01 -2.13859126e-01
-3.94605964e-01 1.04465675e+00 9.15018976e-01 -2.81702191e-01
2.52755314e-01 1.44179866e-01 1.88538376e-02 -5.80638766e-01
-1.02774823e+00 2.50356048e-01 -4.12363820e-02 2.49093637e-01
1.19977987e+00 8.26652050e-02 -7.80217469e-01 5.89643240e-01
-7.16022074e-01 -5.20560980e-01 4.25530434e-01 6.02258921e-01
-5.28629184e-01 -1.07512832e+00 -9.11399499e-02 6.12472773e-01
-7.69247651e-01 -6.14746332e-01 -6.97551608e-01 8.17698777e-01
-2.01436639e-01 7.92070329e-01 5.43506920e-01 4.50375490e-02
2.21320003e-01 6.41420841e-01 6.27489269e-01 -1.13689601e+00
-8.84459913e-01 -2.58233309e-01 6.83106184e-01 -7.25743234e-01
-2.21955746e-01 -9.66925979e-01 -1.54275095e+00 -5.96638396e-02
-2.65126556e-01 7.57162869e-02 2.47201443e-01 1.31567836e+00
8.55088159e-02 -1.32077843e-01 2.26013914e-01 -6.29927099e-01
-7.25833118e-01 -1.36428881e+00 -4.30377930e-01 7.95995951e-01
-9.68785305e-03 -3.28317940e-01 -4.02439505e-01 -1.43651098e-01] | [10.617230415344238, 9.183589935302734] |
91dd9f2f-3b49-48c0-a523-01e6277fc60e | machine-learning-diffusion-monte-carlo-energy | 2205.04547 | null | https://arxiv.org/abs/2205.04547v2 | https://arxiv.org/pdf/2205.04547v2.pdf | Machine Learning Diffusion Monte Carlo Energies | We present two machine learning methodologies that are capable of predicting diffusion Monte Carlo (DMC) energies with small datasets (~60 DMC calculations in total). The first uses voxel deep neural networks (VDNNs) to predict DMC energy densities using Kohn-Sham density functional theory (DFT) electron densities as input. The second uses kernel ridge regression (KRR) to predict atomic contributions to the DMC total energy using atomic environment vectors as input (we used atom centred symmetry functions, atomic environment vectors from the ANI models, and smooth overlap of atomic positions). We first compare the methodologies on pristine graphene lattices, where we find the KRR methodology performs best in comparison to gradient boosted decision trees, random forest, gaussian process regression, and multilayer perceptrons. In addition, KRR outperforms VDNNs by an order of magnitude. Afterwards, we study the generalizability of KRR to predict the energy barrier associated with a Stone-Wales defect. Lastly, we move from 2D to 3D materials and use KRR to predict total energies of liquid water. In all cases, we find that the KRR models are more accurate than Kohn-Sham DFT and all mean absolute errors are less than chemical accuracy. | ['Isaac Tamblyn', 'Jaron T. Krogel', 'Kevin Ryczko'] | 2022-05-09 | null | null | null | null | ['total-energy'] | ['miscellaneous'] | [ 7.96649307e-02 -3.58184576e-01 -2.20536664e-01 -2.96370517e-02
-8.17076325e-01 7.34586045e-02 8.10003936e-01 3.59315306e-01
-7.46054292e-01 1.29874897e+00 3.35235178e-01 -6.49412692e-01
-7.62551799e-02 -1.21458793e+00 -6.73089981e-01 -1.27460349e+00
-3.60742629e-01 5.43318450e-01 1.41869664e-01 -2.20750272e-01
8.35786343e-01 8.05945635e-01 -1.27666223e+00 4.68773872e-01
1.04226494e+00 1.01061654e+00 -2.19207883e-01 5.45154572e-01
8.84744227e-02 9.44756448e-01 3.90857607e-02 1.19815553e-02
5.45999268e-03 -1.64403170e-01 -8.83051991e-01 -9.74793971e-01
1.29319221e-01 -6.53458685e-02 -4.71660316e-01 7.96079397e-01
6.86492205e-01 4.32188481e-01 1.57501447e+00 -5.56579769e-01
-1.14951825e+00 6.37148499e-01 -4.82705861e-01 -3.28359753e-02
2.94854790e-01 3.14449638e-01 7.79829860e-01 -1.06734467e+00
6.08518958e-01 9.89970505e-01 9.20322537e-01 7.70182788e-01
-1.51184762e+00 -5.11282563e-01 -4.63456511e-01 4.09142137e-01
-1.32458067e+00 -2.46235922e-01 7.08277702e-01 -6.66512609e-01
2.08975315e+00 1.15451314e-01 5.80853999e-01 1.33916080e+00
8.33784223e-01 1.09129772e-01 1.28406119e+00 -8.33391488e-01
9.35333788e-01 -1.81496233e-01 4.73431051e-01 5.52759469e-01
5.69383800e-01 4.26652431e-01 -3.02181363e-01 -5.96967280e-01
5.60716689e-01 -5.25809750e-02 3.26506607e-02 -1.98316276e-01
-1.12932074e+00 1.24173141e+00 7.41289973e-01 3.04717034e-01
-8.56046915e-01 5.87796569e-01 6.28193974e-01 -6.33153394e-02
5.36953986e-01 6.27335131e-01 -3.47931862e-01 -4.97590518e-03
-9.02740002e-01 6.26720726e-01 7.93120265e-01 1.78826749e-01
5.33597410e-01 1.61486149e-01 1.32229462e-01 5.90608716e-01
3.55709344e-01 2.83053547e-01 5.02838433e-01 -9.09937739e-01
3.37263882e-01 8.40295553e-02 2.59035856e-01 -5.38949728e-01
-4.67892140e-01 3.81245017e-02 -1.15811968e+00 4.33192909e-01
1.70704544e-01 -2.68526077e-01 -8.00014079e-01 9.57618952e-01
1.47593562e-02 -5.54896295e-01 9.04031545e-02 6.70479298e-01
9.24798489e-01 9.77850735e-01 2.90853858e-01 -3.27044815e-01
8.02726567e-01 -9.56322253e-01 -1.83817312e-01 1.30701020e-01
7.89455473e-01 -6.66331351e-02 4.98673201e-01 4.47569281e-01
-1.16418862e+00 -1.99845031e-01 -9.27233815e-01 -3.79153103e-01
-6.21522486e-01 -2.65884131e-01 1.01748574e+00 7.62192130e-01
-1.06697285e+00 1.46740675e+00 -1.12293613e+00 1.49065098e-02
4.10641819e-01 5.35598397e-01 -2.86179721e-01 -1.46408424e-01
-1.13422298e+00 1.02319539e+00 4.20803607e-01 -2.19385177e-01
-6.91502631e-01 -8.65352631e-01 -6.92412436e-01 -2.32007504e-01
-2.01702580e-01 -5.44039369e-01 6.96266532e-01 -5.09838939e-01
-1.56420565e+00 5.66696048e-01 -2.76819199e-01 -7.06453085e-01
2.40326047e-01 2.22950414e-01 -4.03125852e-01 -2.68181246e-02
-4.24860716e-02 8.00739348e-01 2.35108659e-01 -1.12773633e+00
1.97268665e-01 -3.37799460e-01 -5.49227834e-01 -2.37525567e-01
1.68996572e-01 -1.64264888e-01 5.76037228e-01 -7.44976848e-02
1.75746545e-01 -5.87229133e-01 -7.80157685e-01 -5.08071721e-01
-6.36664093e-01 -3.48562926e-01 1.06661253e-01 -7.54263520e-01
8.90230656e-01 -1.52744758e+00 2.95840859e-01 5.81911862e-01
3.37189138e-01 -1.04607604e-01 1.40980080e-01 7.65753925e-01
-3.60340148e-01 2.81191081e-01 -4.38012093e-01 -2.56526414e-02
-3.94576974e-02 -1.17175423e-01 2.01798588e-01 5.16053438e-01
7.21434951e-02 9.03247297e-01 -5.64940453e-01 -3.15822773e-02
2.21209258e-01 7.67969787e-01 -8.35234106e-01 -4.98057187e-01
-3.98130625e-01 2.32622638e-01 -1.41998738e-01 7.05358446e-01
7.35865414e-01 -3.06198359e-01 -2.34607924e-02 -2.65634596e-01
-6.57401204e-01 5.39326489e-01 -5.80883026e-01 1.24195397e+00
-2.69417554e-01 2.58033961e-01 -4.13517118e-01 -1.10310483e+00
8.96468341e-01 1.45190701e-01 6.27056599e-01 -8.84487867e-01
-9.88052487e-02 6.79941118e-01 2.23899975e-01 -2.05647469e-01
3.97954762e-01 -5.42247534e-01 1.66088581e-01 3.61804307e-01
7.47403875e-02 -2.97627263e-02 -8.98425505e-02 -3.55231725e-02
1.22059262e+00 -2.15488270e-01 5.30771792e-01 -6.26783669e-01
5.07658601e-01 1.87080324e-01 -4.02450524e-02 8.08051527e-01
-5.57516888e-03 2.88811773e-01 5.06083786e-01 -7.65955985e-01
-1.37238646e+00 -9.45206821e-01 -3.62820894e-01 6.76475406e-01
-3.64307523e-01 -3.68923455e-01 -6.93664670e-01 -2.90226728e-01
2.19843894e-01 1.15991604e+00 -6.53332949e-01 -6.31343499e-02
-4.24384356e-01 -1.35724699e+00 1.75728738e-01 5.57091773e-01
3.36597741e-01 -1.31912482e+00 -1.57599956e-01 2.41109625e-01
3.28824669e-01 -4.79139537e-01 1.50232673e-01 6.94310248e-01
-9.82908070e-01 -7.13900566e-01 -7.53409386e-01 -3.00375164e-01
2.74145007e-01 -2.36697569e-01 8.44024897e-01 -1.40525922e-01
-6.07922256e-01 6.14631325e-02 -9.83380601e-02 1.54654440e-02
-6.74566746e-01 1.49613544e-01 2.43557870e-01 -8.27594280e-01
5.76018572e-01 -9.22198057e-01 -9.32263553e-01 -3.68998349e-01
-2.63465881e-01 -4.24840823e-02 4.30496067e-01 6.10277474e-01
6.89167500e-01 2.09176570e-01 3.13206136e-01 -7.10316896e-01
7.68611968e-01 -7.06094921e-01 -3.07140440e-01 -2.26405874e-01
-8.31682622e-01 3.53318274e-01 7.27124751e-01 -1.42127261e-01
-7.31965363e-01 -7.12734386e-02 -6.29317343e-01 -9.64368228e-03
-2.01953605e-01 4.22224373e-01 2.64531374e-01 -4.05545920e-01
1.12006378e+00 2.05357611e-01 -2.39313304e-01 -5.34216702e-01
1.49472237e-01 4.92822677e-01 6.72014356e-02 -8.99378598e-01
2.04520300e-01 4.19197738e-01 4.50787604e-01 -9.84366059e-01
-2.58548111e-01 5.61041618e-03 -8.68782401e-01 1.09862648e-01
1.29448116e+00 -6.06805146e-01 -1.04700267e+00 3.97131503e-01
-1.20090783e+00 -4.10928905e-01 -2.46172801e-01 6.81670606e-01
-5.50292552e-01 4.68922079e-01 -1.10401738e+00 -1.08121514e+00
-7.73585320e-01 -1.20868039e+00 7.12856233e-01 -2.17980295e-02
-1.33474708e-01 -1.28448272e+00 2.38785118e-01 2.86361188e-01
6.29139841e-01 5.80886066e-01 1.69603002e+00 -6.12625539e-01
-3.49089712e-01 -1.68139622e-01 -1.83207303e-01 1.60103440e-01
-8.01114738e-02 6.89442754e-02 -7.28235841e-01 -2.14240819e-01
-2.88565904e-01 -2.55536467e-01 1.38556004e+00 9.01582718e-01
1.32438505e+00 -2.98249811e-01 -4.34167266e-01 2.72968471e-01
1.65775943e+00 3.86748135e-01 9.01486993e-01 3.56131971e-01
8.48484218e-01 2.33575389e-01 -3.28701884e-01 2.07939848e-01
1.87886506e-01 3.01290303e-01 3.30367625e-01 1.09946609e-01
2.28143647e-01 -1.13064535e-01 4.84587848e-01 6.63830876e-01
-1.01217651e+00 -1.20937303e-02 -1.18405879e+00 4.99916300e-02
-1.43666923e+00 -1.22134221e+00 -7.19797730e-01 2.27008796e+00
7.62136459e-01 1.48502305e-01 1.93063855e-01 1.68658793e-01
3.94957453e-01 1.28058344e-01 -6.80751204e-01 -7.87032604e-01
2.58947432e-01 7.11387455e-01 9.04836416e-01 5.71474433e-01
-8.51651132e-01 7.22800791e-01 7.22609806e+00 8.74922395e-01
-1.16216612e+00 3.66606265e-01 7.42586613e-01 -2.27152899e-01
-3.81398857e-01 -9.19059068e-02 -6.39300168e-01 6.07078910e-01
1.28195131e+00 4.18750644e-01 8.31549764e-01 7.63225853e-01
8.99889991e-02 -2.38252282e-01 -1.05209363e+00 8.23078513e-01
-5.08746624e-01 -1.86226010e+00 2.61246990e-02 5.84892631e-01
9.79166389e-01 4.79092449e-01 -4.79198210e-02 4.12930220e-01
4.32561040e-01 -1.47346783e+00 3.77529860e-01 9.04962420e-01
8.94697964e-01 -9.25999582e-01 5.50587177e-01 2.72864491e-01
-5.78231752e-01 1.79717332e-01 -6.80195332e-01 -3.37848723e-01
1.12350527e-02 9.10539687e-01 -5.46927631e-01 3.56167883e-01
4.70858783e-01 4.79009479e-01 -1.91741139e-01 6.13377988e-01
2.85145223e-01 6.17684186e-01 -1.51519388e-01 -4.26061720e-01
4.79487956e-01 -5.65311372e-01 1.34115964e-01 1.02948797e+00
4.19246733e-01 4.04824652e-02 -1.87441647e-01 1.13550377e+00
1.04827443e-02 1.43615410e-01 -4.52098101e-01 -1.87487509e-02
3.21187109e-01 7.97942877e-01 -6.31464422e-01 7.69843906e-03
-2.05347508e-01 9.76220965e-01 4.15668398e-01 4.17262107e-01
-5.38543463e-01 -2.12686241e-01 5.68203807e-01 2.76687503e-01
2.98930615e-01 -4.15474832e-01 -1.08251758e-01 -8.63279343e-01
-3.72343183e-01 -2.69738048e-01 -1.08256660e-01 -7.47021317e-01
-1.72351408e+00 1.43010825e-01 -1.28520772e-01 -1.14338435e-01
-1.35412499e-01 -1.19981301e+00 -1.07504380e+00 1.06735766e+00
-1.44670904e+00 -5.33959568e-01 2.50554323e-01 2.62901545e-01
6.09932393e-02 -8.06340128e-02 1.05632102e+00 3.17786820e-02
-5.76287031e-01 -2.29737964e-02 8.66732240e-01 -1.34640127e-01
3.01965714e-01 -1.41722703e+00 7.77462482e-01 8.01846199e-03
-3.21059495e-01 7.06039488e-01 6.08941376e-01 -9.53278542e-01
-1.55229175e+00 -1.12949848e+00 6.56803846e-01 -1.96251586e-01
5.42593956e-01 -2.67151684e-01 -9.14357603e-01 3.14327389e-01
-4.79726382e-02 2.66637564e-01 8.00684571e-01 9.45402756e-02
-1.49555728e-01 4.29301918e-01 -1.33791745e+00 3.49286288e-01
9.80292380e-01 -6.14642978e-01 -1.27621189e-01 8.04368198e-01
6.24660671e-01 -5.42800277e-02 -1.21258354e+00 2.75790006e-01
4.69404191e-01 -1.26850927e+00 1.31520975e+00 -9.42601144e-01
8.11598480e-01 3.24254096e-01 -4.35648799e-01 -1.19393051e+00
-5.90074480e-01 -1.16025098e-01 -2.41015568e-01 6.47882521e-01
7.03586340e-01 -8.39636505e-01 7.01723933e-01 7.69842565e-01
4.50369045e-02 -1.13655519e+00 -1.30283499e+00 -6.21225238e-01
9.65226710e-01 -4.50597346e-01 4.53396171e-01 8.40312243e-01
5.75032085e-03 1.77728295e-01 -1.99616045e-01 -1.12297200e-01
6.21616185e-01 -2.91372597e-01 -1.47618562e-01 -1.37585831e+00
-1.26060158e-01 -3.30427498e-01 -2.31161997e-01 -4.92301553e-01
3.87197286e-01 -1.28897095e+00 -3.41731995e-01 -1.76571095e+00
3.52346838e-01 -4.72000241e-01 -3.40529799e-01 1.60725176e-01
4.23549742e-01 1.72634646e-02 -3.10632706e-01 2.36282080e-01
-2.70129591e-01 7.56237864e-01 8.88342202e-01 -2.23022550e-01
-2.64101505e-01 -1.18151747e-01 -3.89980257e-01 6.94394112e-01
8.64420652e-01 -5.47477365e-01 3.86797458e-01 1.70107022e-01
5.29910505e-01 9.48551949e-03 7.04702795e-01 -1.20492744e+00
9.92060304e-02 -3.91106196e-02 8.68818223e-01 -5.86658716e-01
3.49217504e-01 -3.98749650e-01 -2.46286746e-02 7.29860902e-01
-2.36171037e-01 -2.99456626e-01 -3.03148571e-02 4.70901757e-01
4.59745556e-01 -6.67093337e-01 8.49329829e-01 -5.04337847e-01
-1.92325562e-01 2.86112905e-01 -7.95125604e-01 -4.27674860e-01
6.94006205e-01 -2.21025214e-01 -3.25900644e-01 -3.87401506e-02
-7.27059662e-01 -2.87351161e-01 5.95805168e-01 -4.08030868e-01
4.69811648e-01 -1.43541920e+00 -3.26726794e-01 2.15552188e-02
-2.76903182e-01 -3.05449992e-01 4.97024894e-01 9.90398586e-01
-8.60571682e-01 7.12891459e-01 -2.96645369e-02 -3.27029794e-01
-8.18694174e-01 6.58717573e-01 6.11685991e-01 -2.98398942e-01
-5.51905692e-01 7.09231913e-01 -4.76356179e-01 -6.90867066e-01
-3.74435127e-01 -4.88331079e-01 3.76311429e-02 -2.58163989e-01
2.97515154e-01 8.98219824e-01 2.86422223e-01 -4.23784226e-01
-4.61798638e-01 5.07354856e-01 -1.55799776e-01 3.83921945e-03
1.70329845e+00 4.59875315e-01 -4.18314606e-01 4.58290488e-01
1.40369785e+00 4.61494625e-02 -1.05022144e+00 1.58780202e-01
-8.56668055e-02 2.00796261e-01 4.54225779e-01 -8.02590847e-01
-5.96059561e-01 1.08015430e+00 6.97195649e-01 1.11059636e-01
4.73226488e-01 -6.26863167e-02 7.23884165e-01 7.06516087e-01
3.70536119e-01 -1.03718758e+00 -2.56662816e-01 5.31596541e-01
6.52166545e-01 -1.12038612e+00 2.91152328e-01 -1.29911885e-01
-6.08597994e-01 1.43628728e+00 2.27050483e-01 -3.92303586e-01
6.51054919e-01 -8.28175768e-02 -9.47171688e-01 -5.20200729e-01
-6.36622012e-01 2.45875046e-01 1.68042451e-01 5.76477289e-01
6.30930662e-01 2.56032348e-01 -2.52275914e-01 5.41341901e-01
-1.75613403e-01 -1.47109702e-01 3.82203549e-01 6.46184862e-01
-6.60650492e-01 -1.09856164e+00 -2.92040944e-01 7.91590452e-01
-1.53088987e-01 -5.34974158e-01 -2.89771110e-01 7.94641793e-01
1.35145828e-01 5.94754040e-01 1.81584418e-01 -7.35978037e-02
-1.15962885e-01 4.86407608e-01 8.22336137e-01 -3.85574639e-01
-4.64800209e-01 -4.23982978e-01 9.24331993e-02 -3.87325764e-01
-2.81883299e-01 -6.15656078e-01 -1.47380400e+00 -7.03237534e-01
-4.13391739e-01 2.25255221e-01 8.79217148e-01 1.01249456e+00
2.28577450e-01 3.35164905e-01 4.06672180e-01 -1.35388982e+00
-4.80380327e-01 -9.60804939e-01 -7.46912122e-01 -2.37190947e-01
1.51196763e-01 -5.30602634e-01 -5.38656294e-01 -6.17745996e-01] | [5.261703014373779, 5.324174880981445] |
e87d0861-0d16-4724-a209-2dae9fc55549 | deep-learning-based-dereverberation-of | 2008.03339 | null | https://arxiv.org/abs/2008.03339v1 | https://arxiv.org/pdf/2008.03339v1.pdf | Deep Learning Based Dereverberation of Temporal Envelopesfor Robust Speech Recognition | Automatic speech recognition in reverberant conditions is a challenging task as the long-term envelopes of the reverberant speech are temporally smeared. In this paper, we propose a neural model for enhancement of sub-band temporal envelopes for dereverberation of speech. The temporal envelopes are derived using the autoregressive modeling framework of frequency domain linear prediction (FDLP). The neural enhancement model proposed in this paper performs an envelop gain based enhancement of temporal envelopes and it consists of a series of convolutional and recurrent neural network layers. The enhanced sub-band envelopes are used to generate features for automatic speech recognition (ASR). The ASR experiments are performed on the REVERB challenge dataset as well as the CHiME-3 dataset. In these experiments, the proposed neural enhancement approach provides significant improvements over a baseline ASR system with beamformed audio (average relative improvements of 21% on the development set and about 11% on the evaluation set in word error rates for REVERB challenge dataset). | ['Sriram Ganapathy', 'Rohit Kumar', 'Anirudh Sreeram', 'Anurenjan Purushothaman'] | 2020-08-07 | null | null | null | null | ['robust-speech-recognition'] | ['speech'] | [ 2.36551881e-01 -1.36047333e-01 7.16418147e-01 -4.50984508e-01
-1.32847857e+00 -3.66130412e-01 1.81815892e-01 -3.32483739e-01
-4.07322764e-01 5.65185308e-01 6.51112139e-01 -5.44635773e-01
4.38427106e-02 -2.53899395e-01 -6.41381741e-01 -8.02129865e-01
-2.56157339e-01 -5.45496702e-01 -1.80268645e-01 -4.82141644e-01
-3.22332680e-01 5.02707899e-01 -1.43126822e+00 4.98641521e-01
7.14170992e-01 1.35176194e+00 4.06639874e-01 1.44248676e+00
6.05972588e-01 7.42443144e-01 -1.07500684e+00 1.93618005e-03
1.54336635e-02 -3.97750735e-01 -3.39436293e-01 -1.23908676e-01
2.11986229e-01 -3.44277173e-01 -7.44252980e-01 7.59823143e-01
1.18814981e+00 6.96019292e-01 5.08644819e-01 -5.54613531e-01
-6.51253402e-01 9.17686701e-01 -1.22054815e-01 6.43465340e-01
9.15979967e-02 -1.44341305e-01 8.07007432e-01 -1.41751790e+00
-1.52879685e-01 1.15642679e+00 8.38404655e-01 4.50274915e-01
-8.02421629e-01 -7.08878219e-01 -7.84633160e-02 5.83763063e-01
-1.32992423e+00 -1.15750456e+00 9.33061540e-01 3.54791619e-02
1.76815653e+00 4.75862682e-01 1.70313254e-01 1.26731694e+00
3.43633704e-02 6.62682712e-01 9.04675484e-01 -5.32100320e-01
4.58382592e-02 -1.03559256e-01 3.72519791e-01 5.67838773e-02
-6.55932605e-01 9.17911649e-01 -8.06963623e-01 1.19932242e-01
1.24858893e-01 -6.59722865e-01 -7.70324767e-01 1.00935805e+00
-6.24308884e-01 3.50699931e-01 6.22235477e-01 4.66461539e-01
-7.55838752e-01 7.58100748e-02 2.97687173e-01 4.38551456e-01
9.23296928e-01 3.02583069e-01 -5.05279303e-01 -3.84768307e-01
-1.06652784e+00 -1.75867905e-03 4.20813084e-01 5.91062546e-01
-3.53208259e-02 9.53396738e-01 -3.80925655e-01 1.53870201e+00
3.62089306e-01 5.52461505e-01 6.71442389e-01 -5.09920180e-01
5.41820884e-01 -6.17239118e-01 1.31827742e-01 -5.56439817e-01
-3.81013811e-01 -1.12165761e+00 -1.09278905e+00 5.49561679e-02
-6.68731928e-02 -5.89957058e-01 -1.13461471e+00 1.74453008e+00
7.37304389e-02 3.17949682e-01 5.40950537e-01 8.79650474e-01
9.27820325e-01 1.48257768e+00 -2.03707710e-01 -4.26685452e-01
1.13512290e+00 -1.29579079e+00 -1.39264858e+00 -1.91326495e-02
4.59713787e-02 -1.07565868e+00 7.38527596e-01 5.94911158e-01
-1.31657052e+00 -1.07126951e+00 -1.33878827e+00 -1.43753380e-01
-4.05239284e-01 3.50324184e-01 -2.77046531e-01 9.16692674e-01
-1.39819658e+00 7.40061879e-01 -5.86005390e-01 4.55136567e-01
-9.10184011e-02 -1.54568041e-02 -3.77257243e-02 3.66796702e-01
-1.62743378e+00 6.16402388e-01 2.19960451e-01 5.53949058e-01
-1.05065334e+00 -1.13590169e+00 -9.64873552e-01 4.01017338e-01
-2.78838158e-01 -1.64665371e-01 1.66075277e+00 -7.39043534e-01
-1.75764441e+00 2.35133588e-01 -3.55776995e-01 -9.38614726e-01
2.44119465e-01 -6.09505177e-01 -1.15476060e+00 -1.57838598e-01
-9.02382672e-01 6.86614811e-02 1.14311051e+00 -8.58740091e-01
-5.53496301e-01 -1.20083317e-01 -6.62021637e-01 1.74712494e-01
-3.47193182e-01 3.80801201e-01 2.80808628e-01 -1.27276301e+00
-5.99486940e-02 -4.94921654e-01 -4.69102710e-02 -8.33602190e-01
-2.34856173e-01 -1.15989223e-01 9.31499839e-01 -1.76618874e+00
1.34805310e+00 -2.59698105e+00 -1.83141842e-01 -1.37684986e-01
-3.05299997e-01 7.01977730e-01 -2.69985855e-01 3.70265022e-02
-5.82613766e-01 -1.07886054e-01 -8.96192268e-02 -6.45490885e-01
1.90274537e-01 -5.86683691e-01 -8.65361631e-01 2.36172065e-01
4.12222296e-01 4.26261336e-01 -5.35178781e-01 2.08295956e-01
2.24158719e-01 1.47960782e+00 -1.91271976e-01 6.13749385e-01
2.36280009e-01 2.55249858e-01 4.74882185e-01 1.94805622e-01
7.96787739e-01 5.60897708e-01 -2.96792924e-01 -4.03567046e-01
-1.59979999e-01 8.88800442e-01 -7.02467680e-01 1.23232877e+00
-9.04207230e-01 1.06460571e+00 2.14446887e-01 -3.76749605e-01
1.18752861e+00 1.03327560e+00 -1.61891788e-01 -7.43982375e-01
6.21503666e-02 4.76449043e-01 2.82461107e-01 -1.20125316e-01
6.47270024e-01 -2.49004573e-01 4.82842624e-01 2.10715964e-01
1.41601458e-01 -1.43690482e-01 -3.94090652e-01 -3.87895226e-01
1.02536583e+00 -1.75191909e-01 7.14909434e-02 1.30915686e-01
6.19476914e-01 -9.16148603e-01 2.76871413e-01 4.16791975e-01
-2.01411664e-01 8.09951425e-01 -3.00059170e-01 -2.76811086e-02
-1.18391061e+00 -1.24956179e+00 -2.43818343e-01 1.20311582e+00
-6.56511545e-01 -2.79506117e-01 -8.87301505e-01 -1.08047172e-01
-5.74294686e-01 1.19245589e+00 -2.51198649e-01 -3.02738607e-01
-8.17566276e-01 -6.11026764e-01 8.28220546e-01 7.32484519e-01
4.56675023e-01 -1.22657084e+00 1.42535353e-02 5.83225310e-01
-3.73010784e-01 -1.27470100e+00 -7.31850743e-01 4.41271663e-01
-4.28923428e-01 -1.40965730e-01 -1.09843874e+00 -8.34193707e-01
1.34458840e-01 2.04290763e-01 8.00266266e-01 -3.69680911e-01
-4.19263504e-02 -2.36514267e-02 -6.08197749e-01 -6.48513675e-01
-7.04114556e-01 -1.32936999e-01 2.13823229e-01 3.04671407e-01
-1.34775564e-01 -8.62512112e-01 -6.07221365e-01 3.32876414e-01
-5.35969973e-01 -2.12164357e-01 3.61348063e-01 9.43953931e-01
5.33456504e-01 2.56921321e-01 1.00595033e+00 6.07969575e-02
1.04139328e+00 -1.09370939e-01 -6.35738850e-01 5.80094196e-02
-2.02557489e-01 -2.71581084e-01 7.11181283e-01 -6.36955798e-01
-1.77046716e+00 -5.27403891e-01 -8.35168839e-01 -3.54725480e-01
7.57047627e-03 4.20822352e-01 -1.82934582e-01 3.32938820e-01
8.16797256e-01 3.44479501e-01 -6.39350951e-01 -8.37124765e-01
7.64106289e-02 1.47942317e+00 6.91751540e-01 -8.75392184e-02
4.74187165e-01 -2.86893100e-01 -5.55599034e-01 -1.22909379e+00
-5.85730672e-01 -3.39654237e-01 -2.86706574e-02 -2.48290300e-01
4.91004050e-01 -9.37040865e-01 -4.02688950e-01 8.14975619e-01
-1.43613565e+00 -5.47966897e-01 -1.59020573e-01 8.72510970e-01
-5.78931332e-01 6.30862564e-02 -1.05438590e+00 -1.50265253e+00
-1.00477421e+00 -1.01763034e+00 8.05869341e-01 1.65631831e-01
-9.70698074e-02 -4.50850248e-01 9.16675925e-02 2.18370095e-01
1.01413202e+00 -1.84900329e-01 5.71603358e-01 -6.30369902e-01
-1.22799553e-01 -8.91817585e-02 2.11122870e-01 1.06661928e+00
2.67037421e-01 -1.86695158e-01 -1.72509146e+00 -5.16211331e-01
4.17233467e-01 -3.27997617e-02 7.66955316e-01 6.49772286e-01
1.20124269e+00 -3.53803843e-01 2.49505982e-01 7.35884070e-01
7.04015911e-01 7.85134792e-01 8.27947259e-01 -2.72476614e-01
1.58386528e-01 4.91286308e-01 5.05811155e-01 3.08277458e-01
-3.01362157e-01 7.21781015e-01 -1.78152665e-01 -2.16905326e-01
-7.66571999e-01 -7.78475776e-02 8.03668857e-01 1.50933754e+00
-7.00697750e-02 -3.33793908e-01 -7.94779956e-01 7.69367158e-01
-1.14187300e+00 -9.00706768e-01 1.22365132e-01 2.11320543e+00
9.90233123e-01 -9.00094584e-02 -5.26760630e-02 6.39938414e-01
7.62504995e-01 2.25846559e-01 -4.59219992e-01 -7.95374930e-01
-3.60089958e-01 7.09763587e-01 -5.58086745e-02 9.13248241e-01
-9.21562672e-01 6.74704015e-01 5.89244175e+00 1.00802100e+00
-1.20460713e+00 3.50055724e-01 1.00721812e+00 -1.34147659e-01
2.35074788e-01 -7.73482144e-01 -6.75050616e-01 3.36278714e-02
1.82802463e+00 -3.89882624e-01 7.05773413e-01 5.44682682e-01
5.90922356e-01 5.59601724e-01 -8.53557408e-01 1.03477144e+00
-1.64925531e-02 -6.90765738e-01 -5.23103535e-01 -1.29256800e-01
5.63759685e-01 3.03608388e-01 7.50202298e-01 6.01783872e-01
-2.04472730e-04 -1.40513372e+00 8.17783535e-01 4.24662173e-01
9.82931256e-01 -1.18240178e+00 6.94818497e-01 8.88643935e-02
-1.21822417e+00 -1.14743635e-01 -2.40206584e-01 2.92481691e-01
2.50973463e-01 7.17489779e-01 -1.32164466e+00 4.34068173e-01
8.59083116e-01 1.82737932e-01 1.86056904e-02 1.07835591e+00
-2.55890101e-01 1.21032095e+00 -2.30055258e-01 1.63315773e-01
-2.89725792e-02 2.07281515e-01 9.00008440e-01 1.64224970e+00
5.72490752e-01 1.07315764e-01 -8.16262066e-01 7.04992294e-01
-3.10459882e-01 -2.22575385e-03 -1.47195116e-01 -1.96091488e-01
3.68517101e-01 1.07742798e+00 1.44707844e-01 -1.96847960e-01
5.68527281e-02 7.07114458e-01 -5.69365453e-04 9.57417488e-01
-1.09307122e+00 -9.98954117e-01 6.60360456e-01 -3.62819642e-01
5.82114398e-01 -1.62885845e-01 2.07957309e-02 -5.53511560e-01
1.38588980e-01 -1.08925760e+00 -1.00250743e-01 -1.15204656e+00
-1.07215023e+00 1.47456741e+00 -5.01616895e-01 -9.55514014e-01
-4.95172888e-01 -6.14465594e-01 -4.09520090e-01 1.45448411e+00
-1.62718439e+00 -8.43079627e-01 -3.31813730e-02 4.23711449e-01
8.75015199e-01 -2.06270546e-01 9.77930069e-01 5.03475547e-01
-3.65916312e-01 9.81700540e-01 1.23471983e-01 7.84435198e-02
5.34120023e-01 -1.08318543e+00 9.74355638e-01 1.07937574e+00
2.15632677e-01 4.01657045e-01 8.83466959e-01 -2.97296435e-01
-6.82947695e-01 -1.33296657e+00 1.04899263e+00 1.76639169e-01
4.28526580e-01 -5.17982304e-01 -1.19691873e+00 4.55787778e-01
4.62680250e-01 5.07300459e-02 6.80125535e-01 5.00285905e-03
-4.42565858e-01 -3.74631345e-01 -8.47855866e-01 4.29446369e-01
6.64794028e-01 -8.23306859e-01 -7.60094762e-01 1.49314106e-01
1.33091319e+00 -5.34735322e-01 -7.54589558e-01 5.97705364e-01
3.70832175e-01 -7.99353242e-01 1.05198693e+00 -5.89794457e-01
9.92887020e-02 -3.27202350e-01 -5.95494211e-01 -1.87668884e+00
-5.72953969e-02 -1.23579454e+00 -3.55372190e-01 1.50105584e+00
8.56219888e-01 -5.47851861e-01 2.85545010e-02 -3.49095813e-03
-5.64766288e-01 -4.79159147e-01 -9.82522845e-01 -9.37029719e-01
-1.45467624e-01 -8.53603661e-01 6.27829671e-01 3.20151210e-01
-2.03109428e-01 2.27254197e-01 -4.84647632e-01 5.06544232e-01
3.26033413e-01 -4.14281100e-01 1.59442648e-01 -3.16693515e-01
-5.72319508e-01 -1.92788467e-01 -7.83869326e-02 -1.07536602e+00
9.70231295e-02 -5.63235819e-01 6.43030405e-01 -1.11976695e+00
-5.36588728e-01 1.23311609e-01 -8.30682993e-01 9.90512595e-03
-1.95261717e-01 -5.12855910e-02 2.44465649e-01 -6.17960811e-01
8.41466710e-02 1.01175606e+00 7.56444097e-01 -2.13666514e-01
-3.90033484e-01 3.99342924e-01 -2.29756191e-01 6.44411266e-01
8.74664366e-01 -2.96249717e-01 -2.82487214e-01 -2.31599852e-01
-3.02884340e-01 4.81557429e-01 1.07609995e-01 -1.24293852e+00
1.69503242e-01 6.30203187e-01 3.50219607e-01 -8.94976616e-01
9.16063607e-01 -5.92867553e-01 2.54126370e-01 1.57830968e-01
-6.74854577e-01 -1.52116179e-01 6.10955536e-01 4.13900107e-01
-5.85414231e-01 1.11443080e-01 1.04110467e+00 5.05942643e-01
-1.26212612e-01 -6.24721274e-02 -5.30446827e-01 -2.89340466e-01
2.78026193e-01 2.37395614e-01 -2.13031486e-01 -7.47198105e-01
-9.34464872e-01 -5.20399511e-01 -7.03520715e-01 4.47793782e-01
1.03112662e+00 -1.35333574e+00 -1.15536630e+00 2.13987023e-01
-4.10414398e-01 -3.81100774e-01 7.36916959e-01 6.43119097e-01
-1.68869812e-02 5.01323044e-01 1.20304421e-01 -1.78345397e-01
-1.52605033e+00 4.12934154e-01 1.04916406e+00 -1.99733749e-01
-5.12593567e-01 1.31698227e+00 2.86516368e-01 -1.01389997e-01
6.65087104e-01 -5.25570869e-01 -2.36391038e-01 -2.09407970e-01
1.15824294e+00 6.13496661e-01 6.74528420e-01 -8.13880444e-01
-2.05429479e-01 9.58920717e-02 -1.08629301e-01 -6.16495907e-01
1.66501522e+00 -2.24672183e-01 2.34899506e-01 5.10507047e-01
1.48246694e+00 1.90575644e-01 -1.04446101e+00 -2.98383147e-01
-1.74632639e-01 -6.61241934e-02 6.26064777e-01 -1.28973818e+00
-8.60330343e-01 1.18607116e+00 8.81731331e-01 3.40158373e-01
1.66566038e+00 -4.67952073e-01 1.11852193e+00 1.87732011e-01
-1.23744301e-01 -1.01012003e+00 -2.47202069e-03 8.90705049e-01
1.71355784e+00 -6.43077314e-01 -6.30796254e-01 -7.07827508e-02
-2.83197343e-01 1.13543093e+00 2.34822109e-01 9.91286933e-02
7.40917265e-01 5.90018451e-01 3.34814876e-01 4.30675298e-01
-8.92620087e-01 -1.87266227e-02 5.12031019e-01 5.89924932e-01
6.60583377e-01 1.12165421e-01 1.83898151e-01 9.41055357e-01
-1.04725909e+00 -4.55341786e-01 1.64462626e-01 3.49227428e-01
-3.32391918e-01 -6.53792620e-01 -8.71849298e-01 -1.03383541e-01
-6.79908931e-01 -4.82482165e-01 -3.26826543e-01 3.10128741e-02
-5.43771505e-01 1.72721994e+00 1.84834667e-03 -4.83262658e-01
7.12187231e-01 3.03341717e-01 2.15730488e-01 -3.52639519e-02
-8.88620377e-01 6.98941410e-01 4.20444727e-01 -2.02365741e-01
1.14134081e-01 -3.07785600e-01 -1.04664111e+00 1.42162099e-01
-4.47174281e-01 1.30027741e-01 1.13606131e+00 6.24591410e-01
2.39120334e-01 1.42288852e+00 1.02573955e+00 -7.24764884e-01
-8.09481323e-01 -1.56054604e+00 -6.20675266e-01 1.06404565e-01
1.11988139e+00 -8.63403305e-02 -6.13844633e-01 1.61136568e-01] | [14.998860359191895, 5.994904041290283] |
5d0025dc-d9f9-4ae0-8549-11a0719a1cc8 | a-comparison-study-the-impact-of-age-and | null | null | https://dl.acm.org/doi/10.1145/3469877.3490576 | https://dl.acm.org/doi/10.1145/3469877.3490576 | A comparison study: the impact of age and gender distribution on age estimation | Age estimation from a single facial image is a challenging and attractive research area in the computer vision community. Several facial datasets annotated with age and gender attributes became available in the literature. However, one major drawback is that these datasets do not consider the label distribution during data collection. Therefore, the models training on these datasets inevitably have bias for the age having least number of images. In this work, we analyze the age and gender distribution of previous datasets and publish an Uniform Age and Gender Dataset (UAGD) which has almost equal number of female and male images in each age. In addition, we investigate the impact of age and gender distribution on age estimation by comparing DEX CNN model trained on several different datasets. Our experiments show that UAGD dataset has good performance for age estimation task and also it is suitable for being an evaluation benchmark. | ['Guoliang Chen', 'Qiuming Luo', 'Chang Kong'] | 2022-01-10 | null | null | null | mmasia-2022-1 | ['age-estimation', 'age-estimation'] | ['computer-vision', 'miscellaneous'] | [-4.41917002e-01 6.26319125e-02 -1.32027432e-01 -7.98147798e-01
2.00656548e-01 -2.76264455e-02 6.34825945e-01 1.25513181e-01
-7.10509777e-01 7.02192545e-01 -1.03766344e-01 2.43213907e-01
2.39319950e-01 -8.54611158e-01 -1.92748979e-01 -8.64825785e-01
-3.40367071e-02 5.05728602e-01 -2.51998454e-01 2.88040161e-01
2.61327803e-01 2.34818146e-01 -1.93257701e+00 -3.33530813e-01
8.18358362e-01 1.38675547e+00 -3.19187701e-01 2.09193960e-01
-1.09568210e-02 3.48113328e-01 -6.19377196e-01 -8.28801334e-01
1.80734709e-01 -1.49617463e-01 -6.19068325e-01 -1.40836090e-01
8.76945019e-01 -6.16235852e-01 -6.79842755e-02 1.05851114e+00
8.51536930e-01 -3.51142824e-01 1.01982176e+00 -1.95549238e+00
-5.83061814e-01 6.04967475e-01 -1.29137361e+00 -1.14738885e-02
1.14611916e-01 -4.11784381e-01 5.64038396e-01 -8.01061213e-01
3.93930227e-01 1.42216623e+00 6.07185781e-01 9.28831756e-01
-4.99745160e-01 -1.43357110e+00 6.10489994e-02 2.77112663e-01
-1.59806633e+00 -3.48632574e-01 5.90366304e-01 -4.51056331e-01
1.82678133e-01 -1.06474198e-01 5.97813785e-01 1.27808845e+00
3.74114364e-02 5.63677669e-01 1.51415241e+00 -2.58458376e-01
3.11594065e-02 1.25877023e-01 1.35415122e-01 7.16423929e-01
3.45032632e-01 2.48459335e-02 -5.47815919e-01 -8.95947043e-04
5.62959850e-01 -1.40429214e-01 3.04021329e-01 -2.92491436e-01
-7.20600665e-01 7.16705441e-01 9.87608433e-02 -2.56246068e-02
-1.24525450e-01 -2.90241893e-02 6.03020966e-01 2.13066220e-01
6.72388434e-01 -7.42640495e-02 -5.01279652e-01 -1.99200973e-01
-8.62035513e-01 4.59637582e-01 4.23232108e-01 7.73688495e-01
5.84205031e-01 -5.85981011e-02 1.61898255e-01 8.90899837e-01
3.38815093e-01 5.55623233e-01 5.68412542e-01 -8.79132509e-01
1.86403051e-01 7.10550845e-01 -2.82013446e-01 -8.87712836e-01
-5.06746054e-01 -8.15923512e-02 -8.68071139e-01 4.23741639e-01
9.18238461e-01 -1.26134679e-01 -1.00975442e+00 1.96246088e+00
3.15487176e-01 -3.74463759e-02 -1.73233464e-01 5.97032905e-01
1.22768080e+00 1.76987648e-01 7.00503111e-01 -3.92558932e-01
1.59511054e+00 -5.56691349e-01 -7.16702819e-01 -1.52653933e-01
4.64764774e-01 -7.88260460e-01 6.73959136e-01 5.91020286e-01
-9.22648787e-01 -5.03533840e-01 -7.56639481e-01 1.37110427e-01
-6.39133275e-01 4.59830225e-01 8.96201670e-01 9.79452252e-01
-9.38449621e-01 4.85507071e-01 -4.57309693e-01 -6.61241472e-01
7.55059242e-01 5.80237031e-01 -7.21242964e-01 8.25232416e-02
-1.02760649e+00 8.30437303e-01 2.53199697e-01 -1.51538089e-01
-6.01818502e-01 -7.85762608e-01 -9.74381328e-01 -3.34214658e-01
-1.88712075e-01 -6.00062311e-01 1.15688336e+00 -1.20205104e+00
-9.18678701e-01 1.46492255e+00 -5.37641943e-02 -2.98059791e-01
6.94302261e-01 -1.39709204e-01 -4.30691093e-01 -1.45540044e-01
-6.42126873e-02 1.12773514e+00 1.06016445e+00 -9.23469424e-01
-7.26154566e-01 -1.13001823e+00 -1.83684587e-01 8.76338184e-02
-6.49175882e-01 3.98314893e-01 -1.62596658e-01 -4.41959709e-01
5.40839555e-03 -8.06192756e-01 2.03702047e-01 1.23795196e-01
1.41582072e-01 -6.00392878e-01 9.09700871e-01 -7.69052267e-01
1.53308928e+00 -1.96319914e+00 5.32055981e-02 7.05112517e-02
4.44090277e-01 6.46709278e-02 1.49286419e-01 3.79092693e-02
-2.24236086e-01 1.61153376e-01 2.93560535e-01 -4.42661941e-01
-1.21241972e-01 1.75741255e-01 3.29582810e-01 4.96759385e-01
3.63197513e-02 3.43411356e-01 -5.81777215e-01 -1.20862508e+00
-1.03493422e-01 3.05436105e-01 -3.30015272e-01 3.70585859e-01
3.90026540e-01 4.72945303e-01 -3.31366926e-01 1.06507897e+00
1.25038707e+00 3.71099144e-01 -5.04957251e-02 -3.66243660e-01
2.44675651e-02 -5.96793830e-01 -1.00394928e+00 1.21250105e+00
-3.45183343e-01 2.61217028e-01 -1.07395649e-03 -6.93933845e-01
1.12844145e+00 1.47195891e-01 6.92248046e-01 -4.75708842e-01
8.37676704e-01 2.21201062e-01 2.47611314e-01 -3.80270004e-01
4.50952828e-01 -6.51953667e-02 -1.57510281e-01 4.45101023e-01
2.88322121e-01 5.59892729e-02 5.12707353e-01 8.72283636e-05
3.84410471e-01 1.19160131e-01 2.26810634e-01 -2.02211961e-01
6.82163000e-01 -7.61076033e-01 8.61098230e-01 1.69589028e-01
-8.26081812e-01 6.87548220e-01 6.76504254e-01 -7.65758097e-01
-1.23156679e+00 -1.09606767e+00 -4.64649528e-01 1.16372383e+00
-2.13827148e-01 -3.97780180e-01 -1.01640618e+00 -8.98983717e-01
1.18912347e-01 -6.18302114e-02 -1.09631288e+00 -5.70256002e-02
-2.36454487e-01 -1.08643460e+00 6.44466698e-01 7.80612350e-01
8.82946372e-01 -1.06877780e+00 -2.99956948e-01 -4.21817988e-01
9.46744904e-03 -1.15996194e+00 -2.52373159e-01 -4.23376024e-01
-8.31609488e-01 -1.31444061e+00 -9.89100456e-01 -8.44178915e-01
9.97280836e-01 -4.52831477e-01 1.29819512e+00 6.48669377e-02
-2.53561020e-01 3.04065466e-01 -3.04793984e-01 -1.03044820e+00
-8.02733600e-02 4.77311879e-01 3.53595436e-01 2.43843123e-01
9.35725391e-01 -6.27271771e-01 -8.62743974e-01 3.67635041e-01
-3.93406659e-01 -1.61501735e-01 5.51431000e-01 4.98355985e-01
-2.26246696e-02 -8.72229710e-02 8.52959991e-01 -9.09548640e-01
2.89916158e-01 -3.98882151e-01 -4.57591802e-01 3.98605429e-02
-9.48809624e-01 -2.17737705e-01 1.70344055e-01 -3.29572707e-01
-1.10915351e+00 -2.40215138e-02 -1.48941621e-01 -1.24725454e-01
-2.65156209e-01 9.75635126e-02 -3.27375501e-01 6.70268312e-02
3.38883728e-01 -2.18393892e-01 4.64336962e-01 -4.88166630e-01
-1.98241964e-01 8.77088547e-01 3.21804792e-01 -7.98175752e-01
6.64642811e-01 2.88462967e-01 1.98453292e-01 -6.56286180e-01
-6.23971403e-01 -1.80503622e-01 -7.59948373e-01 -5.64011097e-01
7.89662719e-01 -1.05377543e+00 -9.93542492e-01 1.44539034e+00
-6.99243307e-01 1.48488209e-01 3.82804900e-01 5.00449896e-01
-2.94978529e-01 3.61929268e-01 -5.43655217e-01 -9.53478396e-01
-5.97839534e-01 -9.13911819e-01 1.06235421e+00 5.99315822e-01
-4.13852364e-01 -7.91780591e-01 -2.73375303e-01 3.80814195e-01
2.67081976e-01 4.34182763e-01 7.19040036e-01 -2.53429115e-01
1.89012781e-01 -2.33572930e-01 -4.96904939e-01 4.06339437e-01
6.83436915e-02 5.45214653e-01 -1.17261088e+00 -1.98262289e-01
-5.68130493e-01 -6.67074144e-01 8.27112257e-01 4.62257624e-01
1.60147369e+00 7.65875429e-02 -1.28564462e-01 3.70522678e-01
1.30440736e+00 -7.48506142e-03 8.05671036e-01 1.59198344e-01
6.12382710e-01 8.89324248e-01 9.07738328e-01 7.51374662e-01
5.06648600e-01 2.79705942e-01 5.09963691e-01 -4.06258041e-03
1.84835926e-01 -6.56128079e-02 4.44066525e-02 6.16900802e-01
-6.47675514e-01 2.34431788e-01 -8.83156598e-01 5.93545794e-01
-1.48490036e+00 -8.15071225e-01 1.63839366e-02 2.19090796e+00
9.40993905e-01 -1.49846405e-01 6.06402397e-01 3.71511430e-01
8.12863469e-01 1.04789890e-01 -2.15251952e-01 -7.97278404e-01
-5.99646717e-02 4.30093676e-01 4.83692884e-01 -1.25808120e-01
-1.28550327e+00 6.43025577e-01 6.64666271e+00 6.64635122e-01
-1.13054836e+00 -1.44823834e-01 1.18097329e+00 5.55662718e-03
4.27460730e-01 -4.95026827e-01 -1.07335913e+00 6.22511268e-01
7.49671280e-01 -2.55188584e-01 -3.74101597e-04 9.97757316e-01
-9.79938135e-02 -3.97611290e-01 -9.50245202e-01 1.39174485e+00
1.94841012e-01 -3.21661144e-01 -1.29782230e-01 2.17263326e-01
6.28866851e-01 -5.87937057e-01 5.28161943e-01 3.43991309e-01
7.95889553e-03 -1.25186300e+00 5.16797066e-01 2.76298881e-01
1.01105142e+00 -1.14625740e+00 1.06024778e+00 -3.22229341e-02
-9.74842727e-01 -2.87049562e-01 -3.01744252e-01 -4.09879863e-01
-3.34339321e-01 3.23717326e-01 -4.54126596e-01 2.83660263e-01
1.22054482e+00 6.14147961e-01 -1.09302199e+00 8.23703945e-01
1.14355586e-01 2.21897006e-01 -1.19921088e-01 1.01436891e-01
-1.42746106e-01 -3.54073703e-01 -4.38269526e-01 6.34467304e-01
6.41230226e-01 -1.72365040e-01 -2.11841404e-01 1.85937047e-01
1.02134235e-03 3.58735085e-01 -5.79597712e-01 -9.87049118e-02
5.42273223e-01 1.55035257e+00 -5.46969593e-01 -2.96936429e-04
-5.79253972e-01 4.75932181e-01 2.48021170e-01 -8.98203626e-02
-8.37090015e-01 -8.41127411e-02 1.00398242e+00 2.52318501e-01
-1.61594674e-01 -4.12902012e-02 -2.48660311e-01 -8.03433001e-01
-2.29973748e-01 -6.87439620e-01 7.04106212e-01 -7.35419154e-01
-1.39278519e+00 3.22395653e-01 1.98387831e-01 -1.04107451e+00
-1.65974960e-01 -9.24588501e-01 -4.97283787e-01 6.62539899e-01
-1.19227302e+00 -1.58981156e+00 -7.40751505e-01 4.20839041e-01
1.55385569e-01 -4.86777574e-01 7.55137682e-01 6.90801382e-01
-7.68252134e-01 1.03799772e+00 -4.57949340e-01 3.90044808e-01
1.26374006e+00 -1.29092896e+00 -1.28742188e-01 2.38185272e-01
-5.16618013e-01 5.63900173e-01 7.57833064e-01 -4.23081875e-01
-7.82915413e-01 -8.61355245e-01 1.11637926e+00 -5.88309467e-01
2.28613645e-01 -1.99585333e-01 -4.36852604e-01 6.46366775e-01
4.09383088e-01 -5.51699512e-02 6.95290685e-01 4.13240641e-01
-4.65160340e-01 -4.63510573e-01 -1.46481836e+00 4.67416972e-01
1.00157142e+00 2.64007840e-02 -1.41507313e-01 -3.11698079e-01
-6.05340935e-02 -4.14551288e-01 -1.24892688e+00 8.18105638e-01
9.82979953e-01 -1.15645540e+00 7.53399670e-01 -3.57457280e-01
7.90113330e-01 -1.15153985e-02 2.58996755e-01 -1.07865500e+00
6.79026097e-02 1.46628857e-01 -2.63270497e-01 1.92935109e+00
1.16717197e-01 -5.57811618e-01 1.13454545e+00 7.05301166e-01
3.96693975e-01 -9.78755951e-01 -5.47475457e-01 -5.27346373e-01
4.34478223e-01 1.03324227e-01 9.74524140e-01 8.19168985e-01
-4.94816571e-01 2.39084303e-01 -3.87149006e-01 -2.12664962e-01
8.57541561e-01 -1.89999476e-01 9.32181239e-01 -1.60228097e+00
5.48404098e-01 -4.04818505e-01 -7.96532333e-01 9.70533341e-02
3.36262673e-01 -2.63840854e-01 -5.05730867e-01 -1.08992076e+00
5.12009919e-01 -4.49916542e-01 -2.29935080e-01 3.94941509e-01
-2.01107979e-01 6.88292980e-01 -1.40401527e-01 -2.90764779e-01
-2.68948823e-01 4.42088246e-01 1.24252141e+00 -8.60903263e-02
3.66903841e-01 1.07120238e-01 -6.49399102e-01 9.60686326e-01
1.03811145e+00 -2.00622335e-01 -4.32304889e-01 1.66288521e-02
2.35451296e-01 -5.16121328e-01 -4.64295112e-02 -1.01528740e+00
-1.06652334e-01 -5.19922897e-02 9.07306850e-01 -8.11139226e-01
3.50100458e-01 -6.99911892e-01 -5.16900942e-02 2.52566010e-01
2.26935282e-01 3.93867165e-01 -9.97626111e-02 -3.21827084e-02
-3.32058966e-01 -2.57012039e-01 9.82356429e-01 -2.03565106e-01
-9.64867294e-01 9.26821768e-01 3.33638489e-02 1.13727726e-01
1.22854996e+00 -4.06304926e-01 -2.48480067e-01 -2.69165277e-01
-6.65635347e-01 2.91782767e-01 7.21415520e-01 5.89625895e-01
3.17233145e-01 -1.62537718e+00 -7.42283463e-01 1.50326386e-01
5.16536891e-01 -3.19333106e-01 2.30497628e-01 7.40106523e-01
-4.67613727e-01 -1.30430967e-01 -9.03799295e-01 -4.86947238e-01
-1.89004934e+00 4.33884412e-01 1.90844595e-01 7.01593831e-02
3.43982577e-01 8.45319271e-01 2.52091169e-01 -6.09898925e-01
3.10801953e-01 2.58320183e-01 -7.97747374e-01 7.17885435e-01
4.66510445e-01 5.48746049e-01 -1.72558725e-01 -9.30317104e-01
-3.13893288e-01 8.11792970e-01 -2.91703194e-01 1.32991150e-01
1.17594182e+00 -1.69259176e-01 -4.48417723e-01 2.51381874e-01
1.09231091e+00 -2.77925372e-01 -9.34286177e-01 -6.29329914e-03
-8.80579725e-02 -6.00596011e-01 -3.07359338e-01 -4.66431946e-01
-1.56613028e+00 9.08726692e-01 9.78527665e-01 -1.60748929e-01
1.20887506e+00 -1.07521877e-01 5.51747859e-01 -2.96088755e-01
3.36823463e-01 -1.63413870e+00 1.38293356e-01 3.60580593e-01
5.54498255e-01 -1.55591333e+00 2.97713339e-01 -4.95584011e-01
-6.23024046e-01 9.89784718e-01 1.39273071e+00 2.68826276e-01
9.10898566e-01 4.40235203e-03 4.08262640e-01 -5.14937304e-02
-3.09449613e-01 -2.45880425e-01 6.00496382e-02 8.79132271e-01
7.89965034e-01 -6.92477301e-02 -9.70523655e-01 7.23252892e-01
-6.09862208e-01 1.57729357e-01 3.40930194e-01 6.80782199e-01
-1.24754980e-01 -1.47301579e+00 -2.99112529e-01 6.07160807e-01
-9.63520348e-01 3.06938916e-01 -4.13674921e-01 9.82279718e-01
6.77149534e-01 6.48382604e-01 3.06523353e-01 -3.27203840e-01
-1.54500604e-01 2.38305882e-01 1.01562810e+00 -3.49688768e-01
-3.84890974e-01 -5.64573288e-01 1.05360411e-01 -2.49642417e-01
-5.15547037e-01 -8.47784877e-01 -8.32559168e-01 -6.91596627e-01
-3.66121829e-02 1.12418063e-01 7.85806179e-01 5.37234724e-01
-3.42543684e-02 7.76284048e-03 7.27601171e-01 -5.63361466e-01
-3.70857060e-01 -1.27371085e+00 -1.01636875e+00 6.29818976e-01
-3.18806350e-01 -1.08572209e+00 -2.80828267e-01 1.20930403e-01] | [13.558060646057129, 0.979480504989624] |
edb864ed-6f01-44fd-93e0-40519562cc08 | on-advances-in-text-generation-from-images | 2205.11686 | null | https://arxiv.org/abs/2205.11686v2 | https://arxiv.org/pdf/2205.11686v2.pdf | On Advances in Text Generation from Images Beyond Captioning: A Case Study in Self-Rationalization | Combining the visual modality with pretrained language models has been surprisingly effective for simple descriptive tasks such as image captioning. More general text generation however remains elusive. We take a step back and ask: How do these models work for more complex generative tasks, i.e. conditioning on both text and images? Are multimodal models simply visually adapted language models, or do they combine they reason jointly over modalities? We investigate these questions in the context of self-rationalization (jointly generating task labels/answers and free-text explanations) of three tasks: (i) visual question answering in VQA-X, (ii) visual commonsense reasoning in VCR, and (iii) visual-textual entailment in e-SNLI-VE. We show that recent unimodal advances, CLIP image representations and scaling of language models, do not consistently improve self-rationalization in multimodal tasks. We find that no single model type works universally best across tasks, datasets, and finetuning data sizes. Our findings motivate the need for novel general backbones approach that move text generation from images and text beyond image captioning. | ['Ana Marasović', 'Alan W Black', 'Florian Metze', 'Yonatan Bisk', 'Akshita Bhagia', 'Shruti Palaskar'] | 2022-05-24 | null | null | null | null | ['visual-commonsense-reasoning'] | ['reasoning'] | [ 4.63721454e-01 5.38567305e-01 -7.72817507e-02 -5.44688702e-01
-9.03429985e-01 -8.36486042e-01 1.11901462e+00 -8.75210837e-02
-3.71397883e-01 7.68160880e-01 6.40509665e-01 -6.70183241e-01
2.51750082e-01 -4.16760564e-01 -1.03100812e+00 -4.60705429e-01
6.90437436e-01 7.69665301e-01 -1.72496915e-01 -2.80558020e-01
1.64246708e-01 6.04102723e-02 -1.31727576e+00 9.56263363e-01
6.03781879e-01 6.60755217e-01 2.20864415e-01 1.03419805e+00
-2.86237627e-01 1.31226480e+00 -3.32056701e-01 -9.17697132e-01
-2.78730392e-01 -9.75630581e-01 -9.96951878e-01 3.06629300e-01
9.19004202e-01 -3.08885336e-01 -3.99391800e-01 8.31886292e-01
3.16901535e-01 -8.04313943e-02 1.23264408e+00 -1.52479267e+00
-1.54401731e+00 7.54425526e-01 -5.28353631e-01 8.36869255e-02
3.98403645e-01 6.19782507e-01 1.23771751e+00 -8.95017505e-01
9.78458762e-01 1.79809248e+00 3.38948131e-01 1.10945499e+00
-1.77973533e+00 -2.28298843e-01 2.71092743e-01 2.97377646e-01
-9.88374531e-01 -4.92997885e-01 5.14811158e-01 -6.70049012e-01
1.03336704e+00 4.26382571e-01 3.49949270e-01 1.73315179e+00
-1.83262751e-02 1.05849767e+00 1.41365314e+00 -3.97924542e-01
-1.11061625e-01 4.46727127e-01 -4.58990857e-02 7.46386886e-01
2.66251534e-01 -1.79025605e-02 -6.56391799e-01 1.65644258e-01
8.09272230e-01 -4.20416921e-01 -3.10379684e-01 -3.20936114e-01
-1.59995985e+00 1.23972452e+00 3.40113401e-01 3.32301185e-02
-2.74411649e-01 7.24662006e-01 3.15016896e-01 1.52961090e-01
1.71821594e-01 6.70186520e-01 -1.20589815e-01 2.89779872e-01
-7.76039064e-01 3.81341755e-01 6.21575713e-01 9.64790225e-01
6.06018066e-01 3.74280721e-01 -7.12042511e-01 6.22179329e-01
3.90507400e-01 9.18282092e-01 1.86071157e-01 -1.13974583e+00
6.52014613e-01 2.31060654e-01 -2.74830405e-02 -7.06176758e-01
-3.08718830e-01 5.34147434e-02 -6.82248950e-01 1.67423546e-01
6.53455257e-01 -2.83657849e-01 -1.15463018e+00 2.10112286e+00
-1.65784672e-01 -5.78841507e-01 3.16704452e-01 1.06152511e+00
1.19571114e+00 8.19118023e-01 6.00765705e-01 -2.63812039e-02
1.64278769e+00 -7.17396498e-01 -7.54118085e-01 -6.34832561e-01
3.48123789e-01 -7.98412919e-01 1.59427929e+00 1.55594768e-02
-1.44429469e+00 -5.21081388e-01 -8.00650358e-01 -6.80030048e-01
-4.61195529e-01 1.89111605e-01 6.65361524e-01 5.46689689e-01
-1.35699308e+00 2.83318665e-03 -2.73433119e-01 -6.21956646e-01
4.10757363e-01 4.08287533e-02 -3.19804341e-01 -7.88053498e-02
-1.08671689e+00 9.55044150e-01 2.41717696e-01 -1.94037884e-01
-1.25405681e+00 -4.09300447e-01 -1.02661061e+00 -6.31981269e-02
3.30602944e-01 -1.49394107e+00 1.24241555e+00 -1.34164417e+00
-1.19419742e+00 1.39219260e+00 -2.55993217e-01 -5.24546027e-01
5.59254348e-01 6.20590411e-02 1.00242913e-01 2.19966829e-01
1.26608819e-01 1.46318424e+00 1.10770369e+00 -1.85967720e+00
-9.32482481e-02 -2.85843819e-01 2.38821357e-01 4.90079850e-01
6.13332307e-03 -2.28252411e-01 -2.52096891e-01 -5.13150752e-01
-2.80476391e-01 -9.16380048e-01 5.21097332e-02 -5.08129820e-02
-6.10765874e-01 -3.41105670e-01 3.66602391e-01 -7.17171788e-01
6.34364486e-01 -1.93487287e+00 3.83930445e-01 -2.32804179e-01
2.33877137e-01 -2.12802455e-01 -4.93498862e-01 4.44358855e-01
-3.05368870e-01 5.29306352e-01 9.17146280e-02 -4.78286356e-01
5.22330046e-01 3.54567736e-01 -8.89359236e-01 8.47216174e-02
4.60213482e-01 1.67752826e+00 -6.34669483e-01 -8.92613888e-01
2.47087225e-01 4.47940767e-01 -5.85125983e-01 2.60976255e-01
-7.83554256e-01 2.78864354e-01 -1.21747836e-01 3.34909499e-01
2.70225585e-01 -7.13073075e-01 -9.08706523e-03 -4.26019281e-01
2.17861488e-01 -6.22054935e-02 -5.68464220e-01 1.55450976e+00
-2.57490307e-01 1.04197037e+00 -8.89736339e-02 -7.29806960e-01
3.78070384e-01 2.59241074e-01 -2.85606474e-01 -8.99710119e-01
6.46933243e-02 -1.50594026e-01 5.65406308e-02 -8.81226122e-01
4.76747036e-01 -6.00253642e-01 -1.32266149e-01 5.51769257e-01
3.04910481e-01 -5.31624556e-01 2.08937868e-01 6.46327257e-01
5.59655190e-01 3.20051253e-01 7.87420571e-02 1.44721687e-01
1.24793947e-01 4.23292488e-01 -2.80671120e-01 1.21589506e+00
-3.55521888e-02 8.99670541e-01 7.99882174e-01 -8.51157680e-02
-1.39141071e+00 -1.30307579e+00 1.81113183e-01 1.64744747e+00
1.07169561e-01 -2.47423857e-01 -7.84676790e-01 -5.17469823e-01
-1.12805329e-02 1.33443570e+00 -1.00396013e+00 -4.93196510e-02
-2.97149003e-01 -6.27769589e-01 7.42325902e-01 6.40433073e-01
2.43849501e-01 -1.37587976e+00 -4.74065542e-01 -3.12077761e-01
-5.23552477e-01 -1.27129698e+00 -4.99320924e-01 -2.11713603e-04
-6.25699341e-01 -6.55947089e-01 -1.00923240e+00 -7.02633560e-01
5.62889755e-01 1.19370922e-01 1.60107231e+00 -9.74082276e-02
-2.09079236e-01 1.17935300e+00 -1.46352686e-02 -3.65839779e-01
-7.59098649e-01 -2.32186332e-01 -4.32004988e-01 -4.55955528e-02
-1.01407245e-02 1.22093437e-02 -5.62161744e-01 3.24055217e-02
-1.10255945e+00 6.17754221e-01 8.38051617e-01 8.15480292e-01
4.41692144e-01 -8.11326742e-01 3.46054018e-01 -9.42467332e-01
9.16089356e-01 -3.89576882e-01 -4.56489250e-02 6.49729490e-01
-2.46691301e-01 3.73001158e-01 2.67090619e-01 -4.64869797e-01
-1.45071101e+00 -7.06859380e-02 6.94564357e-02 -4.76523370e-01
-4.40322787e-01 2.57681966e-01 9.52048078e-02 4.16245431e-01
8.94112289e-01 4.06945348e-01 6.61300942e-02 1.21700376e-01
1.28034925e+00 1.53561741e-01 8.03306758e-01 -6.99715316e-01
6.69698656e-01 6.08702004e-01 -1.40185608e-02 -8.88153493e-01
-8.01776826e-01 6.47871420e-02 -2.01673537e-01 -1.78215995e-01
1.64226782e+00 -9.85243678e-01 -8.77766073e-01 -1.46120772e-01
-1.62081373e+00 -6.80121303e-01 -3.50013852e-01 6.94548860e-02
-1.04415572e+00 3.41361970e-01 -5.36576271e-01 -8.07368875e-01
-2.71853119e-01 -1.01647556e+00 1.25872564e+00 1.59624368e-01
-4.49740559e-01 -1.13185465e+00 -6.95438311e-02 7.66016126e-01
3.44857931e-01 1.50354579e-01 1.43075716e+00 -4.86780077e-01
-8.16817045e-01 2.83050090e-01 -7.07125068e-01 1.90441236e-01
-5.83103299e-01 -4.17452864e-02 -1.01987171e+00 1.11441880e-01
-2.80030847e-01 -1.13301039e+00 1.19127452e+00 4.01938140e-01
1.09915113e+00 -4.40048993e-01 -1.16731644e-01 4.54072386e-01
1.32899892e+00 -1.22076027e-01 8.37271690e-01 -7.39634112e-02
8.30335677e-01 8.90113950e-01 1.76142275e-01 -2.93308161e-02
7.56951571e-01 4.51692402e-01 5.23000836e-01 -3.97827655e-01
-7.08668232e-01 -5.28840303e-01 5.16000092e-01 8.33788589e-02
-8.55904594e-02 -6.73032880e-01 -9.03831244e-01 5.40181160e-01
-2.02239823e+00 -1.32359529e+00 -3.21919471e-01 1.53167748e+00
7.03270435e-01 -2.98205972e-01 -1.02776363e-02 -4.63774621e-01
5.49567521e-01 2.74256200e-01 -4.86251622e-01 -5.18909276e-01
-5.46979666e-01 -1.86849967e-01 2.05118462e-01 5.42525172e-01
-8.72183442e-01 1.05432773e+00 6.50060606e+00 6.78269506e-01
-9.32165861e-01 2.76728839e-01 9.54876542e-01 8.25519394e-03
-9.27449584e-01 1.18555732e-01 -5.24384022e-01 -9.58070606e-02
6.37531340e-01 2.82365620e-01 5.26759088e-01 6.65396094e-01
-6.87108263e-02 -3.81967910e-02 -1.45416629e+00 1.31695569e+00
7.65446723e-01 -1.54534566e+00 7.15122163e-01 -7.96447694e-02
7.88500190e-01 -1.49769813e-01 5.65196931e-01 4.05519962e-01
4.22573835e-01 -1.56886160e+00 1.15227294e+00 6.76713586e-01
9.92985666e-01 -1.22579560e-01 2.78201342e-01 1.02994993e-01
-5.48150659e-01 7.53390640e-02 -8.44104216e-02 2.07583010e-01
3.35970193e-01 -2.76665211e-01 -1.03575826e+00 6.79653138e-02
3.19150895e-01 2.49419227e-01 -9.16767299e-01 3.05454701e-01
-3.94331843e-01 3.80268961e-01 2.59178668e-01 -2.44543836e-01
3.32093716e-01 1.61594257e-01 4.84523386e-01 1.37883663e+00
-2.89062578e-02 -5.73982969e-02 -3.09309214e-01 1.51566875e+00
-3.19676870e-03 -2.81956177e-02 -7.31262743e-01 -4.76475418e-01
-1.24572121e-01 1.23016930e+00 -6.83894455e-01 -4.46939111e-01
-4.40152138e-01 1.12978435e+00 4.32356238e-01 9.09656882e-01
-9.35682595e-01 1.36933863e-01 1.51201367e-01 1.82855010e-01
1.91907763e-01 -1.86290473e-01 -4.64548677e-01 -1.42279088e+00
-4.37538147e-01 -9.62460458e-01 4.73704100e-01 -1.79778171e+00
-1.49945581e+00 4.92179662e-01 1.37249693e-01 -4.08918470e-01
-7.08259165e-01 -1.03730035e+00 -4.04574871e-01 6.48396969e-01
-1.22534359e+00 -1.60711694e+00 -3.77843469e-01 8.65652740e-01
7.46482968e-01 1.27502251e-02 6.35601521e-01 -2.70833939e-01
-1.95208058e-01 2.87796021e-01 -4.66759980e-01 -3.23712677e-02
8.89317691e-01 -1.55565035e+00 1.24687545e-01 6.53437138e-01
5.57132244e-01 5.68114460e-01 1.02339208e+00 -5.15079916e-01
-1.47685993e+00 -7.22801089e-01 7.22573400e-01 -1.01008868e+00
6.27562284e-01 -6.27036393e-01 -6.42398596e-01 9.94024873e-01
9.63418543e-01 -4.91370499e-01 5.25060296e-01 -2.07679309e-02
-5.95726073e-01 3.71794909e-01 -6.41238451e-01 1.08469701e+00
1.00711322e+00 -8.10901225e-01 -7.35236943e-01 5.70620239e-01
7.83023357e-01 -1.62197620e-01 -2.73994505e-01 1.89597651e-01
3.73803109e-01 -8.28583360e-01 1.19781971e+00 -1.10947657e+00
1.04294538e+00 -9.71531570e-02 -3.96745950e-01 -1.05577385e+00
-2.04860911e-01 -5.39732397e-01 2.18218178e-01 1.26755476e+00
8.00756335e-01 -2.30450779e-01 4.36638892e-01 6.87869906e-01
6.89879134e-02 -2.17587411e-01 -6.22120321e-01 -1.92895532e-01
2.80977249e-01 -5.44371605e-01 6.02282658e-02 8.04439187e-01
-1.30496725e-01 1.09574091e+00 -4.64200050e-01 -3.85551900e-01
5.99109411e-01 1.28261939e-01 1.02814329e+00 -6.45945847e-01
-3.67287695e-01 -6.02228105e-01 6.68158308e-02 -9.81676877e-01
3.56380999e-01 -1.02330256e+00 1.06648862e-01 -1.95050728e+00
7.07052469e-01 2.53637671e-01 3.73494238e-01 5.33920646e-01
-2.32065499e-01 4.51036572e-01 5.34142792e-01 1.73807785e-01
-9.39914227e-01 4.24031109e-01 1.55363989e+00 -2.17642114e-01
2.80937016e-01 -6.38858378e-01 -1.08724177e+00 7.65799820e-01
4.45591301e-01 1.43888518e-02 -6.99403822e-01 -8.04933906e-01
7.27783263e-01 2.88228512e-01 1.06935096e+00 -3.32530499e-01
-1.54524308e-03 -4.06818748e-01 8.06244433e-01 -3.26769859e-01
4.87271637e-01 -4.13592041e-01 -3.38117480e-02 3.02696619e-02
-8.83970439e-01 1.97199807e-01 1.79829359e-01 7.42544115e-01
5.69604672e-02 -1.00951411e-01 6.03401780e-01 -4.43207175e-01
-7.92967319e-01 -7.93385580e-02 -4.49128032e-01 4.57644641e-01
8.59440565e-01 -1.73111126e-01 -9.43054378e-01 -9.91222978e-01
-1.05328250e+00 1.56752139e-01 4.45594966e-01 5.82763731e-01
7.50752926e-01 -1.29682302e+00 -9.60801899e-01 -3.66354883e-01
3.53676081e-01 -4.97492492e-01 4.03278559e-01 8.59350145e-01
-2.74469167e-01 7.26314485e-01 -8.70074481e-02 -8.57344568e-01
-1.19070518e+00 8.53711605e-01 3.19744229e-01 1.60651598e-02
-1.24611877e-01 8.64224374e-01 1.06192088e+00 -9.00085345e-02
3.37047279e-02 -2.10189879e-01 -1.74390763e-01 2.02900797e-01
2.80247420e-01 -1.55537173e-01 -6.44769490e-01 -8.34101737e-01
-1.57807335e-01 4.68918741e-01 3.25864479e-02 -7.92440236e-01
8.98244202e-01 -4.71532792e-01 1.80752657e-04 6.10062540e-01
9.18134034e-01 -4.12888348e-01 -1.24492824e+00 1.08821593e-01
-4.38916534e-01 -5.40077910e-02 -3.86578947e-01 -1.08559608e+00
-6.65169299e-01 1.32891655e+00 2.70703286e-01 3.09729069e-01
7.65735686e-01 7.39695072e-01 1.98485658e-01 2.61637390e-01
-2.61739999e-01 -7.70120978e-01 7.77004659e-01 3.47781301e-01
1.47997618e+00 -1.39224374e+00 -1.74517080e-01 4.83612763e-03
-1.56790018e+00 9.94930446e-01 7.44220793e-01 5.15411943e-02
-1.04389556e-01 -2.51339197e-01 8.97295848e-02 -3.97533387e-01
-1.11663640e+00 -6.52720392e-01 6.45240068e-01 7.59203911e-01
6.27703011e-01 6.23687096e-02 5.63662797e-02 5.74541688e-01
-1.58145100e-01 -2.89354026e-01 4.92249191e-01 3.25780898e-01
-2.74527043e-01 -5.22659361e-01 -5.65286994e-01 1.72558382e-01
-2.57894039e-01 -3.68561924e-01 -8.86344552e-01 1.10370302e+00
5.38395457e-02 9.23937500e-01 1.26286313e-01 -3.30599174e-02
8.74299705e-02 4.47279781e-01 9.31296647e-01 -5.53381979e-01
-2.50372797e-01 6.78581223e-02 3.14630479e-01 -3.80359203e-01
-4.51314479e-01 -6.01646662e-01 -1.03404593e+00 -1.56693995e-01
5.55624217e-02 -2.83363670e-01 6.89846694e-01 8.77320111e-01
2.17721775e-01 4.14792389e-01 -2.14168981e-01 -7.57292509e-01
-5.28188586e-01 -7.51935720e-01 -1.57550678e-01 9.18521404e-01
3.98693472e-01 -3.01605523e-01 -4.11380649e-01 6.38164461e-01] | [10.851419448852539, 1.7017357349395752] |
9895030e-348c-4750-8e4f-d4d6f6b12c02 | client-recruitment-for-federated-learning-in | 2304.14663 | null | https://arxiv.org/abs/2304.14663v1 | https://arxiv.org/pdf/2304.14663v1.pdf | Client Recruitment for Federated Learning in ICU Length of Stay Prediction | Machine and deep learning methods for medical and healthcare applications have shown significant progress and performance improvement in recent years. These methods require vast amounts of training data which are available in the medical sector, albeit decentralized. Medical institutions generate vast amounts of data for which sharing and centralizing remains a challenge as the result of data and privacy regulations. The federated learning technique is well-suited to tackle these challenges. However, federated learning comes with a new set of open problems related to communication overhead, efficient parameter aggregation, client selection strategies and more. In this work, we address the step prior to the initiation of a federated network for model training, client recruitment. By intelligently recruiting clients, communication overhead and overall cost of training can be reduced without sacrificing predictive performance. Client recruitment aims at pre-excluding potential clients from partaking in the federation based on a set of criteria indicative of their eventual contributions to the federation. In this work, we propose a client recruitment approach using only the output distribution and sample size at the client site. We show how a subset of clients can be recruited without sacrificing model performance whilst, at the same time, significantly improving computation time. By applying the recruitment approach to the training of federated models for accurate patient Length of Stay prediction using data from 189 Intensive Care Units, we show how the models trained in federations made up from recruited clients significantly outperform federated models trained with the standard procedure in terms of predictive power and training time. | ['Bart De Moor', 'Wouter Verbeke', 'Lyse Naomi Wamba Momo', 'Vincent Scheltjens'] | 2023-04-28 | null | null | null | null | ['length-of-stay-prediction'] | ['medical'] | [-9.44944248e-02 3.12151730e-01 -1.38627827e-01 -5.02541959e-01
-6.77926302e-01 -4.02845472e-01 1.28530964e-01 3.39899808e-01
-6.64515197e-01 9.26954389e-01 -6.30470663e-02 -5.91701686e-01
-6.80379093e-01 -8.60141456e-01 -4.54942048e-01 -7.70550609e-01
-2.04759181e-01 1.06783926e+00 -5.06506741e-01 3.19116771e-01
-3.44126940e-01 7.40123391e-01 -1.27026880e+00 7.07323492e-01
6.67407274e-01 1.18230534e+00 -2.43739247e-01 5.80319881e-01
-2.17271835e-01 8.93817961e-01 -7.16663480e-01 -4.67122585e-01
6.87446594e-01 -2.53231287e-01 -9.35549378e-01 -5.13116755e-02
-4.41597067e-02 -4.89917725e-01 -7.45803863e-02 5.52669108e-01
8.39629412e-01 -7.26398975e-02 1.50305256e-01 -1.11505938e+00
2.21396908e-01 7.10211694e-01 2.08496228e-01 -1.10777490e-01
2.11120807e-02 3.80562395e-02 7.22666264e-01 -5.67517936e-01
5.75159311e-01 6.71205163e-01 6.84852898e-01 6.32718801e-01
-1.21098435e+00 -6.62257493e-01 -2.33013391e-01 -8.71641263e-02
-1.16566467e+00 -5.70075989e-01 3.68349731e-01 -3.51466924e-01
8.17846358e-01 6.23168826e-01 3.81484717e-01 6.99822068e-01
-1.79137006e-01 5.00378132e-01 7.83222854e-01 -4.91795361e-01
5.07627249e-01 6.38371825e-01 -7.37546906e-02 4.58818376e-01
3.05340976e-01 6.59057498e-02 -4.52300459e-01 -8.13794971e-01
4.22781587e-01 4.26640362e-01 -5.28654084e-02 -3.52461368e-01
-9.03893828e-01 9.20742452e-01 1.26224130e-01 3.68674368e-01
-7.24757731e-01 -2.58145690e-01 6.13811970e-01 4.67246205e-01
4.48820263e-01 3.02106768e-01 -9.46478784e-01 -1.39852822e-01
-1.00054395e+00 7.75595848e-03 1.26747000e+00 7.31954575e-01
7.79455125e-01 -3.17644298e-01 -6.40992001e-02 6.39544964e-01
-1.20537706e-01 1.07063770e-01 6.20888293e-01 -9.47788358e-01
4.80640471e-01 1.09578180e+00 2.72444356e-02 -6.22111380e-01
-5.50977051e-01 -5.24985313e-01 -8.48588824e-01 4.22416218e-02
5.37073314e-01 -6.83278978e-01 -4.80408490e-01 1.46038496e+00
6.52198732e-01 -2.62831986e-01 2.13874429e-01 6.56874537e-01
2.31554121e-01 1.75573289e-01 -6.51446357e-02 -3.67271394e-01
1.14342833e+00 -6.73347890e-01 -5.75147986e-01 3.09198320e-01
1.08724964e+00 -5.93184769e-01 3.17965478e-01 5.38379312e-01
-1.10487998e+00 2.59571392e-02 -4.75396156e-01 4.37926739e-01
-3.39935660e-01 -6.17172383e-02 8.37118506e-01 1.02901876e+00
-8.80939841e-01 8.33039761e-01 -9.61033702e-01 -3.72679979e-01
8.41504872e-01 1.13472593e+00 -5.27469933e-01 -1.35181203e-01
-7.44574606e-01 6.96061134e-01 1.43855855e-01 -1.26565784e-01
-4.61185366e-01 -8.83517206e-01 -3.57040465e-01 3.21877599e-01
1.73981413e-01 -9.95049238e-01 1.26358461e+00 -1.05960095e+00
-1.11763465e+00 7.23409235e-01 2.21585050e-01 -7.17472732e-01
1.00627005e+00 1.39517993e-01 -2.43431270e-01 6.09787367e-02
-2.15980679e-01 -1.68158468e-02 3.45120639e-01 -9.71822262e-01
-1.12193108e+00 -7.31469393e-01 -2.38222495e-01 -3.20571475e-02
-7.03913331e-01 1.42524317e-01 -2.61738300e-02 -1.70007423e-01
-2.30878934e-01 -7.52352715e-01 -6.16333008e-01 -7.95144141e-02
-3.37827243e-02 3.20395194e-02 7.79752374e-01 -5.90935886e-01
1.04490471e+00 -1.87688792e+00 -3.99407893e-01 6.70242250e-01
4.18839812e-01 3.60201001e-01 9.85437855e-02 5.49908638e-01
1.54066801e-01 1.25652373e-01 -2.77296677e-02 -4.27065045e-01
-1.51011407e-01 5.10104895e-01 1.70918331e-01 5.49371958e-01
-1.81827679e-01 5.24901688e-01 -5.13057768e-01 -7.66170204e-01
1.32249981e-01 4.20400888e-01 -8.02543223e-01 4.59516317e-01
7.56700933e-02 5.51390946e-01 -6.21093094e-01 6.58006907e-01
4.16113138e-01 -4.53572899e-01 6.69216156e-01 1.79070786e-01
2.20701188e-01 2.01205760e-01 -1.05998695e+00 1.38142502e+00
-7.15890944e-01 -1.07421853e-01 5.92076182e-01 -1.28529370e+00
9.35370684e-01 8.04411173e-01 1.24544573e+00 -5.06021619e-01
4.81619328e-01 5.43053746e-01 6.97255805e-02 -5.17457008e-01
-6.25651255e-02 -2.35706106e-01 1.56884864e-01 8.35368812e-01
1.46021470e-01 7.93725789e-01 -1.71758398e-01 1.23844752e-02
1.30436158e+00 -4.80711401e-01 2.20753819e-01 -1.44404054e-01
4.61873919e-01 5.97588085e-02 6.20925665e-01 6.40671968e-01
-2.52038896e-01 2.33676374e-01 2.52991140e-01 -9.58951592e-01
-9.67296958e-01 -5.66922367e-01 -1.28512278e-01 1.07303858e+00
-6.01130307e-01 -6.24047555e-02 -7.33263552e-01 -9.13629234e-01
3.40264916e-01 3.39715034e-01 -5.44754148e-01 -1.89221348e-03
-7.28679776e-01 -8.10926735e-01 4.24643636e-01 2.67306864e-01
-8.88371617e-02 -1.07730460e+00 -1.11959779e+00 5.34848154e-01
-8.46824050e-02 -8.88782144e-01 -2.04001769e-01 6.13562167e-01
-1.10780501e+00 -1.28261960e+00 -7.36671627e-01 -4.40917879e-01
8.77830982e-01 -3.42274874e-01 8.93387496e-01 2.93078452e-01
-6.13971233e-01 2.48748526e-01 -1.91523433e-01 -5.65131128e-01
-6.25325084e-01 3.35504472e-01 1.93537250e-02 3.18336964e-01
4.40340459e-01 -5.78320146e-01 -7.33822346e-01 3.02565217e-01
-7.43568003e-01 -2.18172908e-01 5.38919508e-01 1.06493318e+00
2.09160879e-01 -2.72132814e-01 8.70463431e-01 -1.41690075e+00
4.85585153e-01 -6.42745793e-01 -5.86061895e-01 2.76455194e-01
-1.15264487e+00 -1.81746949e-02 1.04153609e+00 -2.42827028e-01
-9.26915467e-01 3.87535095e-01 1.25183448e-01 -4.89524305e-01
-3.26532394e-01 4.22735959e-01 -2.37952527e-02 -1.47403762e-01
7.60096431e-01 -1.25687331e-01 6.13045037e-01 -7.87188888e-01
2.72839721e-02 1.18069327e+00 1.67238310e-01 -5.30000687e-01
2.79645324e-01 5.94911933e-01 1.91547554e-02 -2.10606933e-01
-4.44805235e-01 -7.67801464e-01 -4.89102006e-01 4.11199071e-02
2.04832971e-01 -6.36057019e-01 -1.17300892e+00 -1.30562812e-01
-9.30034578e-01 -8.74334350e-02 -6.88812196e-01 6.36853814e-01
-5.02924502e-01 -3.21200565e-02 -4.39552069e-01 -1.05433154e+00
-1.01262021e+00 -7.97608137e-01 4.80564743e-01 -1.06110096e-01
-2.43189514e-01 -1.00426114e+00 9.28071514e-02 7.02676415e-01
8.31104636e-01 4.03875738e-01 9.31322753e-01 -1.33569419e+00
-4.10859108e-01 -6.14305735e-01 7.73466006e-02 3.15305442e-01
3.17715317e-01 -5.16696453e-01 -1.00488317e+00 -5.34455061e-01
1.20780624e-01 -1.25220507e-01 2.35730544e-01 1.65479958e-01
1.09635735e+00 -9.06157076e-01 -3.91687691e-01 7.29202926e-01
1.37449551e+00 2.43396699e-01 3.91654670e-02 3.55108708e-01
3.06244940e-01 9.54449177e-01 2.62039542e-01 1.08810270e+00
1.59335449e-01 3.86280686e-01 4.94013339e-01 -2.72592664e-01
4.40695971e-01 2.29835793e-01 -1.57826573e-01 5.30500114e-01
-1.26770362e-01 1.95986167e-01 -1.06152070e+00 6.21997356e-01
-2.09200883e+00 -7.87239671e-01 2.82350093e-01 2.43284178e+00
8.27059269e-01 -4.53503370e-01 4.68035787e-01 1.57052487e-01
5.09745121e-01 -6.98630512e-01 -7.07244754e-01 -6.95064843e-01
1.03377543e-01 4.11204547e-01 8.66465211e-01 6.11856803e-02
-6.93376780e-01 1.53460830e-01 5.45739269e+00 3.04217279e-01
-1.05453575e+00 2.87547201e-01 9.24426258e-01 -4.91109043e-01
1.66315958e-01 -1.09721519e-01 -5.38683116e-01 2.81281710e-01
1.45385063e+00 -3.32068235e-01 6.04410589e-01 1.12293470e+00
2.93367326e-01 2.57192016e-01 -1.20668495e+00 9.87505257e-01
-2.60710627e-01 -1.60534787e+00 -1.98659256e-01 5.18651068e-01
6.33767188e-01 2.46752903e-01 -2.53645271e-01 3.28285359e-02
3.74026984e-01 -1.11607766e+00 2.99295664e-01 4.16534662e-01
7.10986674e-01 -1.02162862e+00 1.05755317e+00 7.19370425e-01
-6.09565139e-01 -6.42269731e-01 -7.73121342e-02 7.59489164e-02
-1.02256827e-01 5.10185003e-01 -1.18776071e+00 7.10841596e-01
6.92609131e-01 -1.04728602e-01 -5.96235357e-02 1.03736198e+00
7.85834253e-01 4.62192506e-01 -4.73145813e-01 1.55494481e-01
9.54388604e-02 -4.34669033e-02 -7.14892596e-02 1.10231042e+00
2.80960560e-01 5.81227206e-02 1.98789507e-01 5.45860112e-01
-2.14580983e-01 4.92029339e-01 -4.41613138e-01 1.63947955e-01
3.28877777e-01 1.39557099e+00 -2.77343363e-01 -3.11325133e-01
-3.57664555e-01 4.68783468e-01 2.65159488e-01 5.30827418e-03
-3.18709403e-01 -2.38888443e-01 6.33661687e-01 3.23248923e-01
1.11325704e-01 3.59737396e-01 -3.92550498e-01 -8.10627878e-01
-1.77931804e-02 -1.18094635e+00 9.10078883e-01 3.63159776e-01
-1.42563379e+00 7.28919268e-01 -3.65015477e-01 -1.06481779e+00
-5.56933701e-01 -4.44007993e-01 -3.04749250e-01 1.01254642e+00
-1.21615469e+00 -1.12644207e+00 -2.67263532e-01 1.00223243e+00
-1.25310734e-01 -4.11777526e-01 1.28691483e+00 7.45694220e-01
-4.44148809e-01 7.90150523e-01 3.52000952e-01 2.37187892e-01
5.13679028e-01 -8.58386397e-01 -2.82134712e-01 2.20764264e-01
1.64707806e-02 6.77714050e-01 3.30036432e-01 -2.89113253e-01
-1.38754189e+00 -1.25870860e+00 1.10782897e+00 -2.03382537e-01
2.61960596e-01 -3.38247061e-01 -6.58752561e-01 5.36762893e-01
-7.83737674e-02 3.77990514e-01 1.14758265e+00 3.01710963e-01
5.87228537e-02 -4.68385547e-01 -1.63464952e+00 3.05339750e-02
5.40617049e-01 -3.26957256e-01 -1.54581685e-02 6.73809469e-01
2.01085404e-01 -1.41822398e-01 -1.15823042e+00 3.11932117e-01
5.50007403e-01 -9.39619899e-01 5.66656828e-01 -9.68563557e-01
-1.57839015e-01 3.29181254e-01 -3.90272811e-02 -8.10588419e-01
4.42285910e-02 -1.04113209e+00 -1.58033624e-01 9.61246073e-01
5.07822096e-01 -9.60080087e-01 1.54602730e+00 1.32878816e+00
2.40228087e-01 -1.14796722e+00 -1.28894186e+00 -7.21064985e-01
9.61917713e-02 -5.07134907e-02 9.07117128e-01 9.51625168e-01
2.85127908e-01 -1.14093080e-01 -3.39894503e-01 -2.14473352e-01
6.86907172e-01 2.02022254e-01 9.02648926e-01 -1.33494186e+00
-5.04319191e-01 -1.50316373e-01 -2.75029987e-01 -1.61071345e-01
-2.14214787e-01 -9.09368455e-01 -3.30087215e-01 -1.44405103e+00
1.76660523e-01 -1.01489878e+00 -4.67224330e-01 8.77347589e-01
3.63531500e-01 -2.97825903e-01 1.96136832e-01 4.18319255e-01
-3.04167420e-01 -1.00848135e-02 7.30175436e-01 1.61792904e-01
-3.88514698e-01 5.05213380e-01 -4.75714624e-01 3.95707846e-01
1.01837146e+00 -7.27332532e-01 -1.10982925e-01 -1.95435971e-01
-7.83224180e-02 4.79508281e-01 1.68768123e-01 -8.19107771e-01
5.33611417e-01 -3.17286663e-02 3.89328718e-01 -2.33426005e-01
1.81581587e-01 -1.35861719e+00 5.31620562e-01 8.87754977e-01
-2.47133777e-01 7.88104460e-02 -3.89345922e-02 3.13288778e-01
-1.22006148e-01 -8.08288306e-02 6.76187932e-01 -2.83134788e-01
2.41877720e-01 4.69557643e-01 -1.96048826e-01 -2.98536390e-01
1.22508609e+00 -2.43665949e-01 -1.07547835e-01 -1.86515510e-01
-9.98165846e-01 2.24435836e-01 2.74841249e-01 4.85848188e-02
2.55970955e-01 -7.34632432e-01 -6.15589797e-01 3.65622520e-01
-1.36598527e-01 3.52899469e-02 1.42810673e-01 1.07708514e+00
-4.54741299e-01 6.31167650e-01 -1.15406685e-01 -3.53117496e-01
-1.42048013e+00 5.56958139e-01 4.73311245e-01 -4.87461835e-01
-6.43622935e-01 4.14794654e-01 -2.37566084e-01 -6.93314016e-01
5.73199451e-01 3.51261497e-02 2.45530888e-01 -1.77157167e-02
2.84178257e-01 5.36335945e-01 6.95752382e-01 -3.97807300e-01
-4.80306745e-01 -1.67217150e-01 -1.15296818e-01 1.72265485e-01
1.66340280e+00 1.25914663e-01 -1.24321476e-01 -1.33912742e-01
1.28604245e+00 -2.38688394e-01 -8.39386582e-01 -2.22473294e-01
2.56157249e-01 -3.87902379e-01 -1.05535604e-01 -1.02026427e+00
-1.47257388e+00 5.98542035e-01 7.72916675e-01 1.07240185e-01
1.16296303e+00 -1.31597832e-01 7.33189046e-01 4.02022839e-01
6.10735655e-01 -9.61920679e-01 -5.36193371e-01 -1.40390784e-01
2.20735610e-01 -1.15751600e+00 -8.31859559e-02 9.94521081e-02
-3.71838272e-01 1.10466540e+00 3.04132670e-01 2.18018308e-01
7.96262205e-01 4.66076821e-01 3.60304266e-01 -1.88029602e-01
-9.18513417e-01 3.09763134e-01 -1.59315929e-01 4.06913608e-01
2.20304400e-01 2.47741386e-01 -3.28518301e-01 8.62262607e-01
-3.71381156e-02 2.87031829e-01 1.83078721e-01 1.05552125e+00
-9.41156223e-02 -1.47117376e+00 -3.85508597e-01 9.46890295e-01
-1.01241267e+00 1.57307208e-01 -2.63406217e-01 4.56812620e-01
3.90435368e-01 9.82473195e-01 -7.47720525e-02 -2.50752233e-02
4.04927075e-01 2.38069907e-01 8.08113888e-02 -5.34667790e-01
-1.40806234e+00 3.65414657e-02 2.07088768e-01 -6.43572748e-01
-1.53510123e-01 -5.82437038e-01 -1.22029805e+00 -2.85956949e-01
-5.14553070e-01 5.31931162e-01 1.08064091e+00 8.01241934e-01
8.14115584e-01 2.46604607e-01 9.35771644e-01 -4.21952188e-01
-1.28080535e+00 -6.27701104e-01 -6.07865930e-01 4.32786584e-01
3.43793482e-01 -4.52977121e-02 -3.75675976e-01 -4.32179822e-03] | [6.151136875152588, 6.457866191864014] |
8fdf7abc-73a7-413c-b92b-5768c047a57d | tap-a-comprehensive-data-repository-for | 2304.08640 | null | https://arxiv.org/abs/2304.08640v1 | https://arxiv.org/pdf/2304.08640v1.pdf | TAP: A Comprehensive Data Repository for Traffic Accident Prediction in Road Networks | Road safety is a major global public health concern. Effective traffic crash prediction can play a critical role in reducing road traffic accidents. However, Existing machine learning approaches tend to focus on predicting traffic accidents in isolation, without considering the potential relationships between different accident locations within road networks. To incorporate graph structure information, graph-based approaches such as Graph Neural Networks (GNNs) can be naturally applied. However, applying GNNs to the accident prediction problem faces challenges due to the lack of suitable graph-structured traffic accident datasets. To bridge this gap, we have constructed a real-world graph-based Traffic Accident Prediction (TAP) data repository, along with two representative tasks: accident occurrence prediction and accident severity prediction. With nationwide coverage, real-world network topology, and rich geospatial features, this data repository can be used for a variety of traffic-related tasks. We further comprehensively evaluate eleven state-of-the-art GNN variants and two non-graph-based machine learning methods using the created datasets. Significantly facilitated by the proposed data, we develop a novel Traffic Accident Vulnerability Estimation via Linkage (TRAVEL) model, which is designed to capture angular and directional information from road networks. We demonstrate that the proposed model consistently outperforms the baselines. The data and code are available on GitHub (https://github.com/baixianghuang/travel). | ['Kai Shu', 'Bryan Hooi', 'Baixiang Huang'] | 2023-04-17 | null | null | null | null | ['severity-prediction'] | ['computer-vision'] | [ 2.57137641e-02 1.34745032e-01 -6.53808057e-01 -1.43777624e-01
-5.76980472e-01 1.71902087e-02 3.01407725e-01 3.61200541e-01
-1.94067851e-01 6.80606604e-01 4.17284191e-01 -9.14804220e-01
-6.56719565e-01 -1.39445126e+00 -6.49259448e-01 -3.90202403e-01
-1.57631233e-01 5.07336020e-01 4.99358416e-01 -5.85831881e-01
-1.28363311e-01 6.95189953e-01 -1.13944566e+00 -2.94405445e-02
1.01693881e+00 6.14366591e-01 -1.54129893e-01 3.10619682e-01
5.48211411e-02 7.33017683e-01 -4.01707739e-02 -7.71540761e-01
3.47450316e-01 3.07813194e-02 -5.73434472e-01 -5.61170161e-01
1.86708763e-01 5.48486970e-02 -1.31717491e+00 7.80319870e-01
5.83591461e-01 2.82019973e-01 6.27515852e-01 -1.68020892e+00
-2.97895133e-01 5.65969288e-01 -4.40271020e-01 3.62337112e-01
6.43660799e-02 2.97020912e-01 9.75471139e-01 -4.08995837e-01
2.69588917e-01 1.13999152e+00 8.35316896e-01 3.28966230e-01
-9.43191826e-01 -8.24240685e-01 1.27450064e-01 6.38763070e-01
-1.35964477e+00 -1.75521150e-01 1.09207165e+00 -4.96919900e-01
9.81370091e-01 4.31317747e-01 6.00838006e-01 1.07916641e+00
1.67174205e-01 4.31444585e-01 4.89511639e-01 7.48260096e-02
-2.20482558e-01 -4.86963540e-01 3.10288519e-01 7.61274219e-01
5.25737762e-01 2.84517884e-01 -1.61036581e-01 -1.22698292e-01
4.61518526e-01 5.60161293e-01 -1.22005582e-01 -9.26143080e-02
-9.87836123e-01 8.37165236e-01 1.18119323e+00 -1.47630736e-01
-6.73962831e-01 2.72093594e-01 6.13028228e-01 4.48950455e-02
6.01065099e-01 -3.24450172e-02 -2.67080307e-01 -4.17761616e-02
-4.57921445e-01 8.33607912e-02 5.22203267e-01 8.23682070e-01
8.42151701e-01 -1.45682003e-02 -1.11057289e-01 9.59434330e-01
4.57368232e-02 4.32551295e-01 -1.87311739e-01 -5.95595956e-01
1.05394042e+00 8.90388131e-01 -7.33797312e-01 -1.51362669e+00
-8.10806274e-01 -3.09903204e-01 -1.38925564e+00 -2.39782646e-01
2.78023124e-01 -3.63583297e-01 -6.89189017e-01 1.69577241e+00
4.38977629e-01 6.33083105e-01 -4.76610780e-01 5.26614487e-01
1.08712447e+00 4.48648691e-01 3.35376173e-01 9.20764133e-02
1.11493921e+00 -8.94448578e-01 -5.26001632e-01 -2.43324414e-01
1.00598097e+00 -1.62363201e-01 7.60616779e-01 -1.18923403e-01
-6.94813848e-01 -4.55183953e-01 -3.62101674e-01 9.80321988e-02
-7.60708988e-01 -2.90981531e-01 5.04814744e-01 4.96614933e-01
-8.90535772e-01 3.85734677e-01 -7.59570062e-01 -4.26690459e-01
8.17701936e-01 3.13792735e-01 -4.44075257e-01 -5.74180722e-01
-1.58613229e+00 8.43924701e-01 2.98397958e-01 3.07535380e-01
-3.50590527e-01 -8.50388229e-01 -1.04454565e+00 -1.93853546e-02
7.46745884e-01 -9.87085879e-01 5.22468209e-01 -1.73622537e-02
-6.23957336e-01 2.99502283e-01 -1.50592595e-01 -4.30317789e-01
4.83934432e-01 3.52915451e-02 -8.31057727e-01 -2.74537578e-02
2.45550171e-01 2.87200123e-01 2.66778618e-01 -1.02255297e+00
-5.01520157e-01 -2.83111721e-01 -2.04846673e-02 -1.23981513e-01
-3.92360330e-01 -1.24935545e-01 -6.25954509e-01 -7.89821208e-01
-3.07365388e-01 -1.05107629e+00 -5.14547110e-01 -2.06707656e-01
-9.46279109e-01 -4.76743549e-01 6.38166368e-01 -7.33582854e-01
1.82131910e+00 -1.77609122e+00 -1.41459495e-01 6.31855071e-01
6.05584085e-01 5.92057049e-01 -3.11286211e-01 8.31649661e-01
-3.27889442e-01 1.92159221e-01 -4.18882757e-01 -9.34696272e-02
-2.52593338e-01 2.55781770e-01 8.08292031e-02 2.72867322e-01
9.50666517e-02 1.27486169e+00 -1.05330658e+00 -4.99147445e-01
4.45189446e-01 4.98998433e-01 -4.92635906e-01 -6.50211349e-02
2.81837314e-01 3.77821088e-01 -6.62367582e-01 5.17925918e-01
5.69197893e-01 -1.85197871e-02 1.69138871e-02 -2.62962818e-01
1.44837856e-01 4.11874980e-01 -7.36453712e-01 1.14868033e+00
-4.02258724e-01 5.14070511e-01 -3.33242238e-01 -1.51653516e+00
7.96715140e-01 2.00786963e-01 7.39323914e-01 -8.26982021e-01
-1.35233119e-01 -5.76682948e-02 1.28098994e-01 -7.85220861e-01
-4.05697674e-02 2.36212149e-01 -8.86530504e-02 1.84829921e-01
-4.43345219e-01 3.04939270e-01 2.97664970e-01 3.81127477e-01
1.60536158e+00 -6.96970165e-01 9.47716087e-02 7.49670342e-02
4.17354167e-01 1.07519321e-01 5.68859994e-01 5.64277828e-01
-3.20151180e-01 4.49427426e-01 7.16092885e-01 -5.93096793e-01
-7.79668093e-01 -1.11620533e+00 -7.26544932e-02 8.11263800e-01
-3.68058756e-02 -5.00377655e-01 -5.54365337e-01 -9.37859058e-01
2.38504618e-01 5.71466029e-01 -5.76638758e-01 -5.25946081e-01
-8.99891436e-01 -8.23715091e-01 7.32300997e-01 5.10224700e-01
4.49495614e-01 -8.41479361e-01 4.26919580e-01 1.25375763e-01
-5.11998773e-01 -1.15039623e+00 -4.53151226e-01 -5.19453883e-01
-6.36692643e-01 -1.53270066e+00 -2.78237730e-01 -6.02556407e-01
6.02664828e-01 6.74883187e-01 1.24743390e+00 5.79175234e-01
-3.30691069e-01 2.00268447e-01 -1.69621468e-01 -4.73624796e-01
-2.44764015e-01 5.68897605e-01 1.26602918e-01 2.03560051e-02
4.41781551e-01 -7.53949702e-01 -7.92410970e-01 5.20466566e-01
-6.89007401e-01 1.97764486e-01 5.78090668e-01 4.37526613e-01
4.47601467e-01 3.69523555e-01 1.17174459e+00 -1.00757980e+00
6.43362463e-01 -1.17959428e+00 -2.14169651e-01 5.60389273e-02
-5.93690693e-01 -3.56797636e-01 6.10740602e-01 5.93250431e-02
-7.01153398e-01 -1.77097186e-01 -5.51622570e-01 -2.20113605e-01
-4.01180625e-01 8.97071540e-01 -3.05006206e-01 1.15872443e-01
4.88027900e-01 -2.09069252e-01 3.16780934e-04 -3.07446033e-01
2.86936730e-01 5.53782344e-01 4.27982062e-01 -3.04357827e-01
8.91207814e-01 2.81037092e-01 5.53322673e-01 -9.17445540e-01
-6.46234095e-01 -7.02749312e-01 -7.28963673e-01 -4.65889037e-01
8.63573134e-01 -8.68418276e-01 -7.20743835e-01 3.16412836e-01
-1.00241435e+00 -3.91537219e-01 1.29023671e-01 5.62160969e-01
-2.11022824e-01 3.75957370e-01 -4.97241348e-01 -5.83445489e-01
-2.05022961e-01 -8.70177150e-01 6.35599554e-01 -3.11501622e-01
2.72790603e-02 -1.28375185e+00 1.41303658e-01 7.16581166e-01
3.85426253e-01 6.19033992e-01 1.28230298e+00 -5.97479105e-01
-5.43817937e-01 -5.54108381e-01 -6.24388874e-01 1.18514039e-01
4.49957728e-01 -8.93889219e-02 -4.95862961e-01 1.21681891e-01
-8.36784542e-01 1.81832790e-01 1.20455170e+00 5.87478876e-01
1.40862477e+00 -4.28900212e-01 -8.10196459e-01 4.91555274e-01
9.82168853e-01 -1.76991954e-01 8.08042824e-01 1.38626426e-01
1.52131748e+00 1.03252542e+00 3.04077595e-01 -2.53628995e-02
1.00919807e+00 8.34266186e-01 6.27481937e-01 -3.65026116e-01
-4.74430978e-01 -5.42128623e-01 -2.24105075e-01 8.42226624e-01
-4.26299900e-01 -5.61040699e-01 -1.53524899e+00 6.16940677e-01
-2.23997116e+00 -1.21320772e+00 -9.15034056e-01 2.07265806e+00
2.57025808e-01 3.79831791e-01 6.28879547e-01 2.65270919e-01
9.15852308e-01 8.53574499e-02 -4.66371894e-01 -3.85533720e-02
7.07672611e-02 -1.77653044e-01 7.54559338e-01 3.07882309e-01
-1.06258559e+00 8.18639100e-01 5.58889294e+00 1.03947723e+00
-7.17230737e-01 1.01475552e-01 6.83083475e-01 -3.45414318e-02
-3.19819450e-01 -2.31210098e-01 -4.05281514e-01 5.49251556e-01
1.19129741e+00 -8.91589448e-02 5.21387815e-01 3.99178505e-01
6.81198120e-01 4.01043296e-01 -6.97242081e-01 8.17368567e-01
-3.52307916e-01 -1.46038687e+00 1.94479376e-01 2.11555973e-01
4.78633463e-01 6.30912960e-01 -9.28078815e-02 3.80260825e-01
3.44292819e-01 -1.10771656e+00 -4.53543290e-02 7.93124914e-01
7.40886033e-01 -1.03211987e+00 6.60368383e-01 3.42377901e-01
-1.64256942e+00 -2.79910654e-01 -1.89235315e-01 5.23521891e-03
6.69841707e-01 8.80104363e-01 -6.95831180e-01 1.08981562e+00
6.93202138e-01 1.12689018e+00 -7.31253684e-01 1.31253469e+00
-2.62615621e-01 1.06032741e+00 -2.77068794e-01 2.84656823e-01
5.67043126e-02 -2.15824366e-01 5.91657698e-01 1.14510584e+00
3.01264703e-01 2.41455838e-01 6.80802241e-02 6.04737103e-01
-2.98928887e-01 -4.31436114e-03 -1.27899230e+00 3.48477721e-01
5.99072814e-01 1.20850909e+00 -4.09762770e-01 -5.03483675e-02
-8.39169145e-01 1.67066306e-01 5.13312697e-01 5.28352439e-01
-1.01379693e+00 -5.18490851e-01 8.91732395e-01 6.09270215e-01
-1.43378600e-01 -2.95994550e-01 -2.92313367e-01 -8.29708576e-01
-1.28587380e-01 -4.67913955e-01 6.21420681e-01 -5.18398404e-01
-1.58560765e+00 5.84383488e-01 3.89029294e-01 -1.27068901e+00
5.78903221e-02 -5.09443462e-01 -1.16050231e+00 6.76538587e-01
-1.62413645e+00 -1.44436705e+00 -3.45359445e-01 7.83578217e-01
2.24035487e-01 -2.01978952e-01 1.41821519e-01 8.47473383e-01
-1.19743621e+00 4.76949483e-01 -2.67493129e-01 5.05576611e-01
4.73661274e-01 -8.68624866e-01 9.63789284e-01 7.42380083e-01
-1.08612306e-01 3.47760171e-01 1.73338354e-01 -1.07744324e+00
-1.16819227e+00 -1.92898667e+00 1.05613530e+00 -7.37288177e-01
9.36544836e-01 -3.14234734e-01 -1.02881384e+00 8.07953596e-01
-2.71787673e-01 2.05185562e-01 6.25600219e-01 2.25480378e-01
-1.33359164e-01 -3.38243186e-01 -7.48815358e-01 7.35985994e-01
1.73365641e+00 -5.40280402e-01 -8.97296593e-02 6.04728997e-01
8.05003226e-01 1.33946938e-02 -8.09231460e-01 6.21168852e-01
1.95485607e-01 -6.91687644e-01 1.39155841e+00 -8.51449907e-01
2.52206087e-01 -4.97554354e-02 3.46413732e-01 -1.53779161e+00
-5.99750817e-01 -3.46527159e-01 -1.59016941e-02 1.15932548e+00
6.46667778e-01 -1.14712274e+00 6.78093433e-01 8.26225221e-01
-5.23546576e-01 -8.77345562e-01 -1.00379384e+00 -7.36214519e-01
1.65195554e-01 -9.41020250e-01 9.77704406e-01 8.08568299e-01
-5.12405932e-02 4.30263668e-01 -5.04702985e-01 2.92478710e-01
6.27475917e-01 -4.76862371e-01 9.05649781e-01 -1.64652574e+00
2.44972557e-01 -6.59198940e-01 -7.16504812e-01 -6.17925942e-01
4.87194836e-01 -1.22092104e+00 -4.06616002e-01 -1.98650360e+00
-6.27661422e-02 -7.32620239e-01 -3.81837398e-01 7.26007223e-01
-3.90450329e-01 9.67277884e-02 -9.38408226e-02 1.13802887e-02
-3.89787078e-01 5.66573143e-01 1.10438848e+00 -1.85074762e-01
-9.52067412e-03 4.71702695e-01 -6.53549075e-01 6.01998925e-01
1.11082947e+00 -6.07701540e-01 -6.62512243e-01 -4.75269616e-01
3.98497611e-01 1.70383975e-01 6.59514427e-01 -9.01233971e-01
4.42454159e-01 -3.21160764e-01 -2.00364411e-01 -5.30239701e-01
1.39051661e-01 -8.74520719e-01 3.04940432e-01 4.60919112e-01
-2.22536266e-01 3.68531227e-01 2.39679381e-01 1.02188981e+00
-1.34130701e-01 5.30155241e-01 2.31350273e-01 3.21966022e-01
-4.38789040e-01 1.15265739e+00 -2.74314374e-01 9.12776142e-02
1.07571840e+00 -9.86792147e-02 -7.37278581e-01 -4.48504776e-01
-4.23784465e-01 6.19430125e-01 -2.37388294e-02 6.20577753e-01
6.93153739e-01 -1.65713501e+00 -9.16953027e-01 1.34130433e-01
3.32303107e-01 5.94437048e-02 6.46460354e-01 1.23173392e+00
-2.75720477e-01 4.98992205e-01 -8.06054007e-03 -2.15563178e-01
-9.47782040e-01 9.22076583e-01 1.88955382e-01 -2.27585971e-01
-7.04203069e-01 4.02817786e-01 1.91346973e-01 -8.65052879e-01
-1.00384079e-01 -5.08343168e-02 -3.37627113e-01 -2.29120940e-01
1.81634799e-01 9.89663184e-01 2.06103817e-01 -9.70001876e-01
-3.85537922e-01 5.34290552e-01 4.17616993e-01 5.44567704e-01
1.30303729e+00 -2.16642395e-01 -1.27404690e-01 5.48200728e-03
1.27136624e+00 -1.84895128e-01 -7.37519741e-01 -2.58545011e-01
-3.75155583e-02 -4.29719955e-01 1.29324310e-02 -3.74346673e-01
-1.60911524e+00 9.77708340e-01 1.23141266e-01 5.80263913e-01
1.09843624e+00 8.08782950e-02 1.33985484e+00 3.60498250e-01
1.76565766e-01 -6.87133551e-01 -2.77637333e-01 3.61782789e-01
8.43248427e-01 -1.46230996e+00 -2.88934708e-01 -9.12413657e-01
-5.65052569e-01 9.18556392e-01 7.46283829e-01 -3.08299363e-02
1.17012048e+00 -3.10817957e-01 -2.96842158e-01 -4.80622739e-01
-7.10887909e-01 -5.12484670e-01 7.50122368e-01 8.79447460e-01
1.19034380e-01 3.04743111e-01 -4.88241017e-02 5.20062745e-01
-2.07451358e-01 -2.00394005e-01 1.12045541e-01 4.18571264e-01
-1.80707425e-01 -1.11070108e+00 6.17363490e-02 1.04726326e+00
-1.84362844e-01 -2.26926655e-01 -2.09272325e-01 7.35975325e-01
-2.35065576e-02 1.28721070e+00 -4.63715941e-02 -7.73627937e-01
7.15944648e-01 -3.38833362e-01 -2.58060038e-01 -4.98208672e-01
-3.13389748e-01 -6.21502101e-01 5.14488101e-01 -6.76795065e-01
-1.09791718e-01 -4.82734948e-01 -1.10894966e+00 -9.21487033e-01
-4.89022136e-02 -6.33844733e-03 2.58975863e-01 8.72705042e-01
6.95365906e-01 7.56729543e-01 6.55014217e-01 -7.05947697e-01
2.22442701e-01 -6.08782530e-01 -3.52590680e-01 5.21904886e-01
3.46637666e-01 -8.18215191e-01 -1.50838628e-01 -5.09967566e-01] | [6.473294258117676, 2.0650875568389893] |
f060111b-c648-45a1-a1c1-e5e5b4b3c1ef | provable-subspace-identification-under-post | 2210.07532 | null | https://arxiv.org/abs/2210.07532v1 | https://arxiv.org/pdf/2210.07532v1.pdf | Provable Subspace Identification Under Post-Nonlinear Mixtures | Unsupervised mixture learning (UML) aims at identifying linearly or nonlinearly mixed latent components in a blind manner. UML is known to be challenging: Even learning linear mixtures requires highly nontrivial analytical tools, e.g., independent component analysis or nonnegative matrix factorization. In this work, the post-nonlinear (PNL) mixture model -- where unknown element-wise nonlinear functions are imposed onto a linear mixture -- is revisited. The PNL model is widely employed in different fields ranging from brain signal classification, speech separation, remote sensing, to causal discovery. To identify and remove the unknown nonlinear functions, existing works often assume different properties on the latent components (e.g., statistical independence or probability-simplex structures). This work shows that under a carefully designed UML criterion, the existence of a nontrivial null space associated with the underlying mixing system suffices to guarantee identification/removal of the unknown nonlinearity. Compared to prior works, our finding largely relaxes the conditions of attaining PNL identifiability, and thus may benefit applications where no strong structural information on the latent components is known. A finite-sample analysis is offered to characterize the performance of the proposed approach under realistic settings. To implement the proposed learning criterion, a block coordinate descent algorithm is proposed. A series of numerical experiments corroborate our theoretical claims. | ['Xiao Fu', 'Qi Lyu'] | 2022-10-14 | null | null | null | null | ['speech-separation'] | ['speech'] | [ 3.57563049e-01 -9.51492563e-02 -3.31406116e-01 3.71881872e-02
-5.51691234e-01 -6.34809852e-01 4.52478230e-01 -3.37200344e-01
-2.61691324e-02 7.71980286e-01 1.54682219e-01 -5.63870430e-01
-7.12535143e-01 -4.13609520e-02 -5.55538297e-01 -1.43969381e+00
-1.29617199e-01 2.70071447e-01 -5.03185868e-01 1.47137269e-01
3.74707468e-02 5.75350344e-01 -1.30910516e+00 -3.74935895e-01
1.22520173e+00 5.64961851e-01 1.48342848e-01 4.37819630e-01
6.22393303e-02 4.73113537e-01 -3.69863749e-01 3.04456837e-02
3.95201206e-01 -6.50029063e-01 -6.08651713e-02 5.52694976e-01
4.07831967e-02 -5.59209362e-02 -2.01709554e-01 1.39751601e+00
3.73382568e-01 7.98265710e-02 9.62829113e-01 -1.43362498e+00
-5.32922208e-01 6.98877156e-01 -6.37608111e-01 1.00992285e-02
8.47244561e-02 -1.41481146e-01 5.95687747e-01 -1.02255666e+00
9.41379890e-02 1.05750024e+00 5.59398770e-01 2.68379867e-01
-1.32600284e+00 -9.69245076e-01 2.34065473e-01 -4.37581986e-02
-1.63637435e+00 -9.02638793e-01 1.12340999e+00 -8.24488103e-01
1.03146628e-01 3.98267776e-01 3.60382706e-01 9.54649925e-01
-1.20363431e-02 6.52172446e-01 1.38479543e+00 -3.66542697e-01
3.63194942e-01 2.02733040e-01 3.25408310e-01 4.97460335e-01
5.64087510e-01 -2.59817299e-02 -3.84115040e-01 -4.21145767e-01
8.78330588e-01 -4.67546582e-02 -5.62171519e-01 -6.25593483e-01
-1.06683779e+00 6.25226378e-01 -1.23147592e-01 3.60681832e-01
-4.12055790e-01 -4.27299231e-01 -2.64268726e-01 2.15017498e-01
5.34279644e-01 2.26959988e-01 -8.29228014e-02 2.11364716e-01
-1.11034667e+00 -1.00557409e-01 8.60461354e-01 9.32573974e-01
4.50123042e-01 4.83056396e-01 -5.71388891e-03 8.15333843e-01
6.61544561e-01 7.34567463e-01 3.95899773e-01 -7.69633591e-01
3.62186998e-01 2.46802479e-01 4.20875430e-01 -8.73208463e-01
-3.30663532e-01 -8.00396681e-01 -1.42614317e+00 6.12091720e-02
6.49513602e-01 -3.18070382e-01 -7.53375411e-01 1.94467163e+00
3.26340318e-01 6.95696533e-01 2.19514444e-01 8.47599328e-01
4.76160169e-01 4.67360705e-01 -3.37603122e-01 -1.02775085e+00
9.87511158e-01 -4.53562349e-01 -9.69238162e-01 -1.23478308e-01
-1.26995891e-01 -7.86782444e-01 6.12382948e-01 6.24233246e-01
-1.02056682e+00 -1.92340061e-01 -1.03930342e+00 5.90135157e-01
2.13736743e-01 3.64878565e-01 4.37020451e-01 9.72342312e-01
-9.16001678e-01 1.20391816e-01 -9.43496943e-01 -1.96832605e-02
1.40802875e-01 3.30520153e-01 -4.06865716e-01 -2.90836245e-01
-8.79980087e-01 3.59456152e-01 1.11005164e-03 5.77152729e-01
-9.33772027e-01 -4.75609392e-01 -6.81172013e-01 1.84532832e-02
3.81933600e-01 -5.69149852e-01 7.14248240e-01 -7.89481163e-01
-1.49765289e+00 2.55979896e-01 -4.98172194e-01 -2.00003445e-01
6.49018466e-01 -2.14065507e-01 -3.59629095e-01 2.08567500e-01
6.38181269e-02 -1.47308320e-01 1.55912030e+00 -1.41900063e+00
-1.03514932e-01 -3.65880847e-01 -1.14572212e-01 1.69401899e-01
-4.91434842e-01 -7.88600743e-02 -4.27465945e-01 -8.55334222e-01
7.95111358e-01 -1.03089511e+00 -3.79583925e-01 -4.28232342e-01
-7.27910936e-01 1.16706267e-01 4.26149428e-01 -6.95756674e-01
1.13393867e+00 -2.21858096e+00 3.94890159e-01 4.83232588e-01
4.51345503e-01 -1.80897918e-02 1.21686071e-01 2.84397930e-01
-5.15392780e-01 -1.00381635e-01 -7.31406689e-01 -4.13830161e-01
-1.01988865e-02 -1.59214348e-01 -2.63003558e-01 1.09326482e+00
3.13790552e-02 4.46591765e-01 -5.97499609e-01 -1.29469320e-01
2.12242410e-01 5.70809364e-01 -4.00807649e-01 1.59144878e-01
1.59659088e-01 9.02504921e-01 -2.03889415e-01 6.32677972e-01
7.78671741e-01 -3.11646879e-01 3.71591389e-01 -1.90003052e-01
-1.61909625e-01 -1.76605806e-01 -1.72895169e+00 1.17734253e+00
-1.38479754e-01 4.82262313e-01 6.37353003e-01 -1.42565727e+00
5.19930720e-01 6.48757219e-01 8.47167909e-01 2.73641129e-03
6.90339357e-02 1.92395687e-01 3.82816821e-01 -4.92872536e-01
-8.18580315e-02 -2.41122782e-01 2.15195701e-01 2.16545284e-01
-3.19209434e-02 3.48776281e-01 1.03096351e-01 9.26173106e-02
9.43356514e-01 -2.29853332e-01 4.81975615e-01 -5.86713612e-01
5.69899559e-01 -4.05175745e-01 7.20642984e-01 8.25062752e-01
2.31933910e-02 2.52888203e-01 3.35389793e-01 5.81763923e-01
-7.74292529e-01 -1.14697039e+00 -4.01693106e-01 5.96344471e-01
1.65501744e-01 4.53501754e-02 -6.84504271e-01 3.21133109e-03
-1.35266528e-01 5.65142214e-01 -2.24759072e-01 -5.84987663e-02
-3.04136842e-01 -1.27681434e+00 4.65516001e-01 3.90554406e-02
1.62658453e-01 -3.48218918e-01 7.53417145e-04 1.76482543e-01
-8.38865116e-02 -9.84780967e-01 -2.84894437e-01 2.90696770e-01
-8.95302951e-01 -9.84160423e-01 -9.67262983e-01 -6.44429624e-01
1.02145088e+00 4.81337368e-01 4.58220899e-01 -4.70528096e-01
3.77912112e-02 5.42049408e-01 -5.01912786e-03 -1.53605165e-02
-3.83974195e-01 -3.57191414e-01 6.45209491e-01 5.64812720e-01
-9.47589725e-02 -9.48888838e-01 -2.88034558e-01 4.44340587e-01
-9.38638270e-01 7.14521036e-02 8.96793842e-01 7.18368769e-01
4.05749708e-01 8.00499201e-01 6.07577682e-01 -5.40829659e-01
7.21603155e-01 -7.57628560e-01 -6.56545937e-01 2.35178083e-01
-7.00716853e-01 1.16266675e-01 5.35827041e-01 -9.54662144e-01
-9.86737669e-01 6.48167208e-02 1.86369583e-01 -6.70122981e-01
-3.28593105e-01 7.53011823e-01 -8.79811823e-01 3.78316455e-02
2.23050818e-01 3.90015185e-01 -5.75965270e-02 -6.80017769e-01
4.46253300e-01 4.28772509e-01 7.60975838e-01 -5.68190217e-01
1.15426743e+00 5.25118291e-01 1.04443422e-02 -1.23454654e+00
-5.01156569e-01 -7.47276545e-01 -5.40855050e-01 -2.79546231e-01
5.29150426e-01 -9.30568933e-01 -5.76760113e-01 5.53697586e-01
-7.73925185e-01 -8.35680068e-02 8.55226070e-02 9.42311287e-01
-2.60138988e-01 5.42795181e-01 -3.86443287e-01 -1.42791045e+00
1.25499934e-01 -9.93397713e-01 5.71150064e-01 9.45094302e-02
8.41431040e-03 -8.57021868e-01 -2.11330261e-02 2.89701819e-01
5.77077605e-02 -9.45153832e-02 9.00431871e-01 -5.43379903e-01
-3.46639156e-01 -1.92929327e-01 7.63324499e-02 2.45011091e-01
5.65612316e-01 -1.61706492e-01 -9.04537678e-01 -3.63417655e-01
6.11316860e-01 3.25418651e-01 8.25295627e-01 8.50055695e-01
7.57247806e-01 -6.20072782e-01 -4.04644161e-01 5.64657807e-01
1.02835548e+00 4.32437330e-01 2.85161287e-01 -1.55400530e-01
8.90106082e-01 6.48199081e-01 -4.63276217e-03 6.62849009e-01
1.02206655e-01 3.84258240e-01 2.43845075e-01 -3.59391011e-02
2.67031610e-01 1.76710412e-01 5.55242121e-01 1.20913696e+00
-3.19401957e-02 -4.20553654e-01 -7.08878160e-01 3.85893613e-01
-1.83469200e+00 -9.04051065e-01 -2.16379687e-01 2.54019666e+00
8.26535106e-01 8.88921097e-02 6.66287020e-02 4.75194395e-01
9.68838394e-01 -1.75281674e-01 -8.25849295e-01 5.29057920e-01
-4.43044156e-01 -1.57933414e-01 5.61797798e-01 7.64914393e-01
-1.19206297e+00 4.32228982e-01 5.42379570e+00 7.42090642e-01
-1.04715729e+00 1.81981787e-01 1.56483382e-01 -3.78434034e-03
-2.44547591e-01 -6.69912994e-02 -6.64261341e-01 4.38876510e-01
5.42333484e-01 -1.29666418e-01 6.16216958e-01 4.38839585e-01
5.70834100e-01 7.52138719e-02 -7.84739196e-01 1.20524704e+00
1.90000817e-01 -7.04829335e-01 -2.97801495e-01 3.28932285e-01
7.73302972e-01 -4.07971919e-01 3.23670179e-01 -3.59793082e-02
2.24930108e-01 -8.56688797e-01 7.30859101e-01 8.48919630e-01
6.84455633e-01 -6.36309087e-01 3.28564972e-01 6.86461270e-01
-1.03338814e+00 -1.68641657e-01 -1.16470203e-01 -9.37703773e-02
3.27688992e-01 1.17154860e+00 -4.80391622e-01 5.95176578e-01
1.81498572e-01 6.67480648e-01 -3.24838579e-01 1.28201902e+00
-2.87967533e-01 1.15202701e+00 -4.66068000e-01 4.74534690e-01
-1.28434762e-01 -8.28993261e-01 1.13485718e+00 8.29804838e-01
4.41779584e-01 2.05044270e-01 1.52792320e-01 8.25063646e-01
2.28905663e-01 1.12155162e-01 -4.29353982e-01 -3.49751770e-01
3.47200960e-01 1.12395704e+00 -9.40173805e-01 -2.27860883e-01
-2.80478626e-01 7.76364744e-01 -1.69744343e-01 8.88086379e-01
-6.57434404e-01 1.91181004e-01 6.93976283e-01 2.33032666e-02
1.98336452e-01 -6.13665164e-01 -3.29123348e-01 -1.46765935e+00
1.67448282e-01 -9.40962970e-01 5.20856725e-03 -3.19718808e-01
-1.42791796e+00 2.56621748e-01 1.81354240e-01 -1.37996268e+00
-2.82052368e-01 -3.63985598e-01 -5.22573292e-01 8.21940541e-01
-1.14225376e+00 -8.10853779e-01 1.78758930e-02 6.88860059e-01
3.13229650e-01 -3.70289803e-01 4.99257475e-01 5.91809869e-01
-1.02576339e+00 3.53914589e-01 6.31832004e-01 -1.47528931e-01
5.63681722e-01 -1.10186136e+00 -2.99312711e-01 1.30435765e+00
2.03846827e-01 1.07168710e+00 9.34212744e-01 -7.77787268e-01
-1.60520792e+00 -8.92064214e-01 4.25568104e-01 -6.88066557e-02
7.71455646e-01 -5.09397030e-01 -9.80774879e-01 3.91172469e-01
6.51031639e-03 -2.21275851e-01 8.32434118e-01 -7.78554678e-02
-1.52835011e-01 7.80017823e-02 -7.64909148e-01 6.24539196e-01
8.63290548e-01 -2.90825307e-01 -3.28785479e-01 3.72027636e-01
3.91006589e-01 -3.82135920e-02 -6.43281639e-01 5.01843214e-01
6.10178649e-01 -6.95172131e-01 1.14434469e+00 -6.67304397e-01
5.50289676e-02 -6.03327870e-01 -3.22240949e-01 -1.26829731e+00
-4.26061153e-01 -8.39560151e-01 -5.26510417e-01 1.36836362e+00
2.47870192e-01 -7.67209232e-01 6.12864435e-01 6.34087503e-01
-9.76606682e-02 -3.75670195e-01 -9.75447118e-01 -1.01022887e+00
-1.15449689e-01 -6.63035870e-01 1.90932825e-01 1.29481041e+00
-2.33152077e-01 3.68752390e-01 -6.93032801e-01 8.64545643e-01
1.03291881e+00 -3.73302847e-02 5.06782770e-01 -1.33838058e+00
-5.50703347e-01 -6.57665551e-01 -1.14454545e-01 -1.14956295e+00
4.83362556e-01 -7.43692219e-01 1.50492415e-01 -1.31744409e+00
-7.04673529e-02 -7.02967346e-01 -4.13323939e-01 9.75180417e-02
-3.14502269e-01 -3.33868861e-02 -1.64404497e-01 4.58694935e-01
-1.66703969e-01 6.00166559e-01 8.04217041e-01 -2.92362511e-01
-4.36538190e-01 4.84279513e-01 -7.98805296e-01 9.59184229e-01
6.85111403e-01 -4.26778793e-01 -7.60610044e-01 -2.56640334e-02
1.27064556e-01 3.62913400e-01 4.07286495e-01 -9.92932200e-01
3.76375109e-01 -1.90210402e-01 2.59588748e-01 -4.38479632e-01
3.97152096e-01 -9.95282054e-01 4.41504210e-01 1.93610579e-01
-9.41379666e-02 -4.37866449e-01 1.15905750e-04 8.37070167e-01
3.43128573e-04 -3.32217991e-01 5.07251859e-01 3.18332762e-01
-1.17410019e-01 1.29199922e-01 -6.53658748e-01 -2.13736251e-01
6.18003428e-01 5.59847280e-02 9.71491113e-02 -6.35052204e-01
-8.75029802e-01 -1.82184167e-02 8.80406201e-02 7.15110973e-02
5.29584587e-01 -1.26288915e+00 -7.78331518e-01 4.02423680e-01
-1.49813592e-01 -1.76608622e-01 3.62140954e-01 1.43614483e+00
2.70557791e-01 4.17449623e-01 2.92243510e-01 -6.46051586e-01
-1.09412038e+00 6.24698997e-01 1.23996578e-01 9.34027880e-02
-5.63619196e-01 7.49390662e-01 5.12631178e-01 -1.80331782e-01
2.52558857e-01 -1.96614683e-01 -3.46368879e-01 2.44215533e-01
4.26275611e-01 5.76645851e-01 -1.28909171e-01 -7.60377288e-01
-2.59471774e-01 3.59646112e-01 3.56151372e-01 -5.01158118e-01
1.15350640e+00 -3.83476824e-01 -3.01356137e-01 6.53986871e-01
8.70667815e-01 5.92556179e-01 -1.30812931e+00 -3.28768611e-01
4.08430584e-02 -1.10126436e-01 5.99032789e-02 -2.72229165e-01
-9.50036824e-01 6.38500273e-01 5.21417081e-01 3.17726374e-01
1.14602768e+00 -8.62888396e-02 6.83255941e-02 3.70710731e-01
1.60573870e-01 -5.35123527e-01 -2.83080935e-01 1.22168019e-01
7.89885342e-01 -1.00604296e+00 -2.12538354e-02 -4.53994870e-01
-2.18824625e-01 7.72437334e-01 2.36450911e-01 7.89374262e-02
8.77855718e-01 3.87428910e-01 -1.72072232e-01 1.03358604e-01
-3.19222152e-01 -1.82341084e-01 6.33613586e-01 3.74472111e-01
2.89826304e-01 2.49641821e-01 -1.02392063e-01 9.51416731e-01
-1.69994175e-01 -4.48928028e-01 4.74565655e-01 6.20551825e-01
-1.52639627e-01 -8.47252667e-01 -1.06102192e+00 5.60262024e-01
-3.24715525e-01 -2.39997640e-01 -2.98304111e-01 4.40039694e-01
-2.39121299e-02 1.35670900e+00 -3.43260348e-01 4.53976206e-02
-7.42664039e-02 1.92018002e-01 4.13357705e-01 -4.64762211e-01
2.54357547e-01 9.71335411e-01 -4.27437514e-01 -5.22538312e-02
-7.82037914e-01 -8.86370480e-01 -8.54509294e-01 1.07383460e-01
-6.42508090e-01 4.14128453e-01 6.39501691e-01 1.22284305e+00
-1.13225579e-01 5.13451219e-01 5.65977871e-01 -7.85495460e-01
-3.81474704e-01 -1.05422950e+00 -9.18951452e-01 -1.10742658e-01
6.12112880e-01 -9.85832930e-01 -8.67757916e-01 2.67013729e-01] | [7.804302215576172, 4.2917938232421875] |
848bb017-efd2-408b-a599-325a0d4c473c | time-series-anomaly-detection-detection-of | 1708.03665 | null | http://arxiv.org/abs/1708.03665v1 | http://arxiv.org/pdf/1708.03665v1.pdf | Time Series Anomaly Detection; Detection of anomalous drops with limited features and sparse examples in noisy highly periodic data | Google uses continuous streams of data from industry partners in order to
deliver accurate results to users. Unexpected drops in traffic can be an
indication of an underlying issue and may be an early warning that remedial
action may be necessary. Detecting such drops is non-trivial because streams
are variable and noisy, with roughly regular spikes (in many different shapes)
in traffic data. We investigated the question of whether or not we can predict
anomalies in these data streams. Our goal is to utilize Machine Learning and
statistical approaches to classify anomalous drops in periodic, but noisy,
traffic patterns. Since we do not have a large body of labeled examples to
directly apply supervised learning for anomaly classification, we approached
the problem in two parts. First we used TensorFlow to train our various models
including DNNs, RNNs, and LSTMs to perform regression and predict the expected
value in the time series. Secondly we created anomaly detection rules that
compared the actual values to predicted values. Since the problem requires
finding sustained anomalies, rather than just short delays or momentary
inactivity in the data, our two detection methods focused on continuous
sections of activity rather than just single points. We tried multiple
combinations of our models and rules and found that using the intersection of
our two anomaly detection methods proved to be an effective method of detecting
anomalies on almost all of our models. In the process we also found that not
all data fell within our experimental assumptions, as one data stream had no
periodicity, and therefore no time based model could predict it. | ['Stephen T. Edwards', 'Paolo M. Piselli', 'Jason M. Gurevitch', 'Dominique T. Shipmon'] | 2017-08-11 | null | null | null | null | ['anomaly-classification'] | ['computer-vision'] | [ 3.39998864e-02 -4.35599893e-01 6.85284883e-02 -3.20450693e-01
-5.45729935e-01 -6.18489087e-01 5.36790133e-01 4.98198032e-01
-1.60900921e-01 6.05928838e-01 -4.91631702e-02 -9.08900559e-01
-1.33865267e-01 -8.06629479e-01 -5.41152418e-01 -3.51611167e-01
-4.88561898e-01 3.53171319e-01 6.14385724e-01 -2.33303621e-01
4.70863163e-01 7.94393361e-01 -1.65209723e+00 5.24223030e-01
5.35801709e-01 1.20454109e+00 -6.59622192e-01 8.98571968e-01
-5.60854495e-01 1.10979569e+00 -9.63083744e-01 -2.47442219e-02
4.10004020e-01 -3.45037609e-01 -6.43905163e-01 1.23223260e-01
1.97826862e-01 -3.10625643e-01 -3.38237435e-01 6.67346060e-01
-8.68787840e-02 1.49954513e-01 3.53475839e-01 -1.74047816e+00
-6.76960051e-02 1.59801453e-01 -5.63421130e-01 1.17246807e+00
3.28984857e-01 5.55784583e-01 8.44241858e-01 -4.04996306e-01
7.04498664e-02 8.29347908e-01 7.16857791e-01 1.88142523e-01
-1.21150279e+00 -6.21198893e-01 3.95308077e-01 2.86812156e-01
-9.97866094e-01 -4.71791148e-01 5.28771281e-01 -5.46578825e-01
1.11243308e+00 3.18749219e-01 6.49853468e-01 1.13926673e+00
1.94659904e-01 4.22886044e-01 6.49501383e-01 -4.46745241e-03
2.02938065e-01 4.39948170e-03 1.30252331e-01 1.57601878e-01
3.24075580e-01 9.33678821e-02 -3.08164775e-01 -5.98839164e-01
4.38231111e-01 3.10017586e-01 2.28152201e-02 3.64014894e-01
-9.41183984e-01 6.65608644e-01 3.67554128e-02 5.18830717e-01
-6.67463839e-01 1.73308462e-01 6.32840455e-01 7.11503863e-01
6.40968680e-01 5.37969768e-01 -6.14211082e-01 -6.63163900e-01
-8.27363491e-01 1.91283450e-01 8.35186243e-01 3.55319291e-01
4.11349446e-01 5.16656220e-01 1.34366661e-01 5.34787595e-01
-1.68634921e-01 1.96639687e-01 5.33545554e-01 -8.64963353e-01
5.79037488e-01 5.51334500e-01 2.99000114e-01 -1.15457070e+00
-4.39023644e-01 -1.48489922e-01 -2.38629565e-01 2.34193265e-01
1.08138192e+00 -3.97569418e-01 -8.96106482e-01 1.35181248e+00
-8.19875747e-02 4.59117889e-01 -1.46193370e-01 4.75655377e-01
-1.46157578e-01 8.65747273e-01 7.98024610e-02 -2.65853673e-01
7.10508704e-01 1.23091917e-02 -7.08637178e-01 -1.65462047e-01
8.95620763e-01 -7.75596678e-01 9.19341624e-01 4.80445325e-01
-9.35471535e-01 -2.00063258e-01 -7.55204499e-01 6.91596448e-01
-6.38511539e-01 -6.88977540e-01 4.29951400e-01 3.28894556e-01
-7.12695122e-01 9.60150719e-01 -1.00454879e+00 -3.66114229e-01
3.47320110e-01 2.77925730e-01 -8.95000696e-02 2.54667908e-01
-9.69058216e-01 7.93955266e-01 1.56226397e-01 2.26624776e-04
-4.28660452e-01 -5.18101871e-01 -6.08364105e-01 -6.25060797e-02
3.10004443e-01 2.12979287e-01 1.48841441e+00 -1.02490628e+00
-6.98123634e-01 4.25324202e-01 -4.87048626e-01 -7.20311463e-01
3.13049614e-01 -3.37524414e-02 -1.04908681e+00 -2.32332528e-01
8.16008002e-02 -2.38086656e-01 6.94799304e-01 -8.10780823e-01
-1.18854189e+00 -3.27993959e-01 -2.89542526e-01 -3.73746336e-01
-5.37792034e-02 1.86144069e-01 1.77885145e-01 -4.52162772e-01
1.59259751e-01 -5.96865773e-01 -2.00741082e-01 -3.84869665e-01
-2.38736585e-01 -2.78690994e-01 1.08444917e+00 -6.68891490e-01
1.46602428e+00 -2.09623408e+00 -9.36865807e-01 7.38265455e-01
1.28266737e-01 2.90347219e-01 8.29335749e-02 5.74071646e-01
-2.93554395e-01 3.49855304e-01 -1.43908247e-01 1.66230112e-01
-2.78247625e-01 4.32941735e-01 -7.14360356e-01 4.98709291e-01
4.85359281e-01 4.18146074e-01 -9.31075096e-01 9.93761886e-03
1.06924810e-01 2.30491273e-02 -2.45052129e-01 1.98509499e-01
-2.59019852e-01 1.29446641e-01 -2.95511603e-01 7.84376860e-01
2.05660820e-01 7.70961568e-02 -3.88789982e-01 3.99756610e-01
-2.70111233e-01 5.98618150e-01 -1.13487852e+00 5.60063243e-01
-3.13860655e-01 1.18693936e+00 -2.35208169e-01 -1.32859480e+00
9.00230050e-01 4.68572468e-01 9.67567801e-01 -1.12070954e+00
-1.11995339e-01 4.90544617e-01 3.63511354e-01 -7.49187112e-01
2.02928290e-01 2.14154609e-02 2.01592833e-01 7.73434341e-01
-4.35549408e-01 4.93067652e-01 4.03817385e-01 -1.60650179e-01
1.91884804e+00 -3.81109506e-01 -2.37852737e-01 2.92014778e-01
2.82537311e-01 2.32266188e-01 5.71687102e-01 7.64871061e-01
-1.93483472e-01 4.49837267e-01 9.57846999e-01 -8.21340978e-01
-1.11095762e+00 -1.06812000e+00 1.20837661e-02 8.83053184e-01
-2.81542867e-01 -2.73317516e-01 -2.70496517e-01 -7.79006422e-01
1.22827977e-01 1.05882049e+00 -3.12867761e-01 -2.16641113e-01
-7.26849020e-01 -8.32255006e-01 5.17536223e-01 5.65959454e-01
8.98090228e-02 -1.23504174e+00 -5.85483968e-01 6.07319295e-01
-2.63032150e-02 -1.19385624e+00 -2.29378179e-01 3.68669540e-01
-9.43060458e-01 -1.41041625e+00 7.01397881e-02 -1.21141456e-01
4.98453945e-01 2.15821326e-01 1.17909908e+00 1.59659460e-01
-2.34994933e-01 3.69621277e-01 -2.35288113e-01 -8.90044034e-01
-5.47167778e-01 -3.21187943e-01 1.34253845e-01 1.77784815e-01
9.48061824e-01 -9.04326081e-01 -3.59429538e-01 4.78401870e-01
-1.02275372e+00 -9.94151771e-01 1.68116540e-01 2.25448787e-01
1.94968194e-01 3.80083144e-01 7.80523956e-01 -6.96808815e-01
1.16609108e+00 -9.87362146e-01 -5.46435177e-01 -3.44006509e-01
-5.71436703e-01 -8.75698328e-02 8.06316674e-01 -4.90541875e-01
-4.99465078e-01 -2.12484166e-01 -2.78420091e-01 -4.79948342e-01
-6.35623038e-01 3.65453571e-01 3.72683614e-01 4.61445451e-01
8.39316010e-01 1.85862496e-01 2.28574313e-02 -4.32124764e-01
-3.31097484e-01 8.58943522e-01 3.92307311e-01 -3.59911561e-01
6.11754358e-01 4.64834005e-01 -2.79846787e-01 -9.81347799e-01
-5.85699379e-01 -6.43485069e-01 -4.07476425e-01 -2.54922211e-01
3.81538302e-01 -4.78931099e-01 -6.88839436e-01 3.69054377e-01
-1.16139197e+00 -3.13817829e-01 -6.37406766e-01 3.17677885e-01
-3.18749964e-01 7.02582598e-02 -5.32927811e-01 -1.14380658e+00
1.54894754e-01 -6.66514933e-01 4.71231759e-01 -9.74929929e-02
-7.09563017e-01 -9.46735799e-01 7.41192251e-02 -1.36682346e-01
7.87176669e-01 4.71162498e-01 8.23764861e-01 -1.44515431e+00
-3.21682602e-01 -8.56792271e-01 -5.98024651e-02 5.20540476e-01
6.49164438e-01 4.38338667e-01 -8.55565071e-01 2.40447409e-02
2.53471844e-02 2.34417275e-01 4.51755464e-01 2.03256011e-01
1.56688941e+00 -5.10576546e-01 -1.53202698e-01 8.85397568e-02
1.08708334e+00 7.98500896e-01 7.38784134e-01 4.66543674e-01
4.78006899e-01 5.63635945e-01 3.24299932e-01 4.74483907e-01
7.34439865e-02 5.85317135e-01 4.87877786e-01 -4.56334762e-02
4.75680470e-01 8.49413965e-03 7.04023838e-01 1.81951478e-01
-8.01487938e-02 -2.31095672e-01 -1.21423471e+00 8.14393401e-01
-1.80653882e+00 -1.50850379e+00 -5.74215531e-01 2.50883532e+00
3.92691821e-01 8.73984933e-01 7.86053836e-01 4.26485598e-01
6.15025103e-01 -6.56033158e-02 -6.64757609e-01 -9.52840507e-01
3.54384452e-01 6.40449896e-02 6.02245867e-01 1.44696295e-01
-8.94436955e-01 3.21712166e-01 6.84835291e+00 1.65390342e-01
-1.38100195e+00 -3.19590062e-01 6.31215334e-01 -4.62191671e-01
2.33359095e-02 -1.77685067e-01 -5.00381649e-01 1.02411747e+00
1.79822910e+00 -1.38880610e-01 2.11685002e-01 6.81230307e-01
7.64402390e-01 -1.37407497e-01 -1.17136598e+00 6.89292312e-01
-4.94488597e-01 -1.08114862e+00 -1.61901399e-01 3.52528989e-01
3.63922983e-01 3.86438817e-01 -3.10219705e-01 2.54566371e-01
4.07418549e-01 -1.05243409e+00 2.72687018e-01 6.60675883e-01
-2.85095396e-03 -8.22328806e-01 7.79270291e-01 3.85015845e-01
-9.05301630e-01 -3.16239715e-01 6.52745441e-02 -4.53450143e-01
2.81139553e-01 7.83882618e-01 -6.60709500e-01 -8.96719247e-02
7.35845745e-01 6.18302047e-01 -3.94049674e-01 1.32440996e+00
2.53480077e-01 1.09509706e+00 -8.05067301e-01 1.86437011e-01
4.61227775e-01 1.54360933e-02 7.58634031e-01 1.23366606e+00
4.60233480e-01 -2.01866299e-01 1.44621134e-01 6.66239202e-01
4.35127646e-01 -2.35024735e-01 -9.61892188e-01 -1.99042723e-01
4.11476970e-01 8.03994417e-01 -7.29901075e-01 -2.14763090e-01
-7.63071418e-01 3.84755641e-01 -2.39728853e-01 5.28908074e-01
-7.49094486e-01 -5.11373043e-01 1.11964548e+00 8.09260190e-01
1.02657504e-01 -3.01949531e-01 -3.73771995e-01 -9.24826980e-01
3.43111604e-01 -9.74931180e-01 5.61472535e-01 -4.00766075e-01
-1.42350066e+00 4.88109916e-01 -4.79023419e-02 -1.36631227e+00
-7.27025926e-01 -4.98140514e-01 -1.30143332e+00 7.68734217e-01
-1.10656226e+00 -1.53772190e-01 -1.46745175e-01 7.24647284e-01
5.08932352e-01 -1.15241662e-01 5.15642226e-01 3.89123917e-01
-6.11422420e-01 1.97481677e-01 -7.09557682e-02 3.06395441e-01
3.83504987e-01 -1.28456831e+00 7.45584607e-01 1.11699450e+00
2.63262630e-01 3.16521764e-01 9.50996876e-01 -7.24653423e-01
-8.17502737e-01 -9.84281301e-01 7.83343911e-01 -6.16735399e-01
1.33004045e+00 1.04050003e-01 -1.46780407e+00 8.45874786e-01
-2.93985155e-04 1.49694547e-01 6.02153718e-01 6.81760535e-02
-1.21832006e-02 -1.58466086e-01 -9.89662826e-01 4.22121465e-01
7.11687684e-01 -3.39809418e-01 -3.51112008e-01 3.03217024e-01
3.20083529e-01 -1.58169627e-01 -5.55639446e-01 2.24629700e-01
2.15920389e-01 -1.08541799e+00 5.19225895e-01 -1.02586448e+00
2.15890452e-01 -2.69842505e-01 -1.80558875e-01 -1.28732550e+00
-1.12180570e-02 -4.89026964e-01 -4.40704346e-01 1.25403953e+00
6.62794232e-01 -8.57977808e-01 7.51576364e-01 9.83970642e-01
-1.07412234e-01 -8.28158736e-01 -7.72974193e-01 -9.03682351e-01
-2.36030281e-01 -1.21670067e+00 7.24641204e-01 7.10166812e-01
-1.23751603e-01 -5.79213612e-02 -7.16794133e-02 1.30093813e-01
2.89768130e-01 -2.61992097e-01 6.18227005e-01 -1.28987670e+00
2.93467324e-02 -6.74109101e-01 -7.41922021e-01 -4.55255151e-01
-1.98401049e-01 -4.07325476e-01 -1.36400297e-01 -1.28082645e+00
-5.93854845e-01 -4.73334283e-01 -5.60857952e-01 6.68874502e-01
2.50746340e-01 6.23959973e-02 -2.71146894e-01 1.58731222e-01
-2.52802908e-01 -1.27672572e-02 3.84343475e-01 7.32873976e-02
-5.79516649e-01 8.46687138e-01 -6.63507164e-01 8.39725673e-01
1.15203285e+00 -5.45914888e-01 -3.68446082e-01 -1.49857877e-02
2.27112070e-01 -2.16922447e-01 5.88007867e-01 -1.14611816e+00
1.40784755e-01 -2.63230354e-01 5.26861250e-01 -6.33994341e-01
1.90982241e-02 -9.93144572e-01 -1.44237727e-01 2.74484247e-01
-2.44008213e-01 6.20926261e-01 4.67721581e-01 5.35723388e-01
-3.19048703e-01 -3.22510861e-02 4.14136976e-01 -1.46153435e-01
-7.91037858e-01 2.34497607e-01 -9.42887545e-01 1.58944696e-01
1.06771600e+00 -3.58706146e-01 -1.01454385e-01 -8.57747555e-01
-8.51827502e-01 4.24356073e-01 1.88812912e-01 5.88545620e-01
4.65337723e-01 -1.23829925e+00 -5.17857671e-01 3.17400724e-01
-5.99749610e-02 -2.46063635e-01 -2.98266321e-01 8.61735523e-01
-4.27829474e-01 3.22182953e-01 -6.49813563e-02 -7.38405704e-01
-9.87322330e-01 4.37403262e-01 6.39401972e-01 -2.39239689e-02
-6.54886842e-01 1.97541550e-01 -6.32050693e-01 -3.44022252e-02
4.99856323e-01 -3.95092428e-01 -2.78550554e-02 2.91679889e-01
7.66023755e-01 5.56810141e-01 4.58068758e-01 -3.80143464e-01
-3.13706964e-01 4.72706854e-02 -1.64706483e-01 5.75008579e-02
1.32231688e+00 2.23840829e-02 7.79647976e-02 9.34910774e-01
1.11489081e+00 -2.92577744e-01 -1.09564245e+00 -9.28467885e-02
4.79314804e-01 -5.36896288e-01 -1.36146829e-01 -4.92994279e-01
-1.14057589e+00 7.57576764e-01 4.78974611e-01 1.35221219e+00
1.10964978e+00 -1.12244882e-01 1.01595700e+00 4.02740359e-01
-1.21539431e-02 -1.20177388e+00 -9.65429172e-02 5.55285573e-01
4.50124532e-01 -1.22117758e+00 -3.05701166e-01 2.21922278e-01
-5.39715350e-01 1.27241147e+00 8.00060153e-01 -1.51054084e-01
7.40131080e-01 3.47716630e-01 1.94017053e-01 -4.29496527e-01
-1.11466098e+00 -3.74292701e-01 -1.11260666e-02 4.54710692e-01
4.37047064e-01 -3.14455003e-01 1.59229394e-02 -1.95191447e-02
-1.11676469e-01 1.68074016e-02 7.90485203e-01 9.65721548e-01
-6.12536132e-01 -8.19942236e-01 -4.60924327e-01 1.14795136e+00
-9.35025454e-01 3.57050031e-01 -4.44462270e-01 7.80868769e-01
1.36043102e-01 1.28020513e+00 7.20237255e-01 -5.04373431e-01
7.83253312e-01 3.20046306e-01 -2.82757968e-01 -3.12505215e-01
-3.08574051e-01 -2.33621135e-01 1.67727828e-01 -9.53019381e-01
1.04154862e-01 -8.04248214e-01 -1.22184432e+00 -6.03951097e-01
5.00883125e-02 1.28785491e-01 6.33366764e-01 1.28367102e+00
2.68209606e-01 5.46854019e-01 9.55415368e-01 -6.23468578e-01
-3.97090226e-01 -7.32853711e-01 -5.08187354e-01 5.97000420e-01
8.48297298e-01 -3.28415185e-01 -7.23088741e-01 -1.73336670e-01] | [7.387280464172363, 2.657933473587036] |
d1a1583b-d2c4-4f87-a884-a7bb2e26fb3e | meta-learning-based-knowledge-extrapolation-1 | 2302.05640 | null | https://arxiv.org/abs/2302.05640v1 | https://arxiv.org/pdf/2302.05640v1.pdf | Meta-Learning Based Knowledge Extrapolation for Temporal Knowledge Graph | In the last few years, the solution to Knowledge Graph (KG) completion via learning embeddings of entities and relations has attracted a surge of interest. Temporal KGs(TKGs) extend traditional Knowledge Graphs (KGs) by associating static triples with timestamps forming quadruples. Different from KGs and TKGs in the transductive setting, constantly emerging entities and relations in incomplete TKGs create demand to predict missing facts with unseen components, which is the extrapolation setting. Traditional temporal knowledge graph embedding (TKGE) methods are limited in the extrapolation setting since they are trained within a fixed set of components. In this paper, we propose a Meta-Learning based Temporal Knowledge Graph Extrapolation (MTKGE) model, which is trained on link prediction tasks sampled from the existing TKGs and tested in the emerging TKGs with unseen entities and relations. Specifically, we meta-train a GNN framework that captures relative position patterns and temporal sequence patterns between relations. The learned embeddings of patterns can be transferred to embed unseen components. Experimental results on two different TKG extrapolation datasets show that MTKGE consistently outperforms both the existing state-of-the-art models for knowledge graph extrapolation and specifically adapted KGE and TKGE baselines. | ['You Dou', 'Zhen Huang', 'Fenglong Su', 'Chengjin Xu', 'Zhongwu Chen'] | 2023-02-11 | null | null | null | null | ['knowledge-graph-embedding'] | ['graphs'] | [-4.32411224e-01 7.00782955e-01 -7.24402785e-01 -3.14897858e-02
-2.27488019e-02 -4.23998833e-01 7.28138328e-01 3.85410815e-01
-5.20075262e-02 8.83184850e-01 3.32215935e-01 -4.35628563e-01
-4.93179709e-01 -1.33897054e+00 -1.22460139e+00 -3.05158079e-01
-5.72433650e-01 6.44136965e-01 4.32587385e-01 -3.89871478e-01
-3.57596904e-01 7.38227740e-02 -1.11011839e+00 8.80682692e-02
1.00646079e+00 7.60001242e-01 -3.45811814e-01 4.80809987e-01
-2.71751523e-01 1.19484937e+00 -1.39389550e-02 -9.48520124e-01
7.69856349e-02 -8.45144689e-03 -9.71120298e-01 -3.84637296e-01
4.15194094e-01 -2.94519484e-01 -1.36371291e+00 5.68884254e-01
2.20657185e-01 3.32661361e-01 4.93321180e-01 -1.67682278e+00
-1.70879126e+00 1.13772857e+00 -2.15833560e-01 2.30393291e-01
4.27861929e-01 -2.11229786e-01 1.25272918e+00 -1.14006174e+00
1.08748579e+00 9.87751424e-01 9.74340677e-01 3.60024393e-01
-1.09243214e+00 -2.46714830e-01 4.90215868e-01 8.37171316e-01
-1.56159639e+00 -2.02303194e-02 6.89316690e-01 -3.34959835e-01
1.48576236e+00 -1.96041226e-01 1.00149441e+00 1.25266826e+00
-3.11770122e-02 7.19004154e-01 3.95215571e-01 -4.55550492e-01
1.17851242e-01 1.98621973e-02 3.90978158e-01 9.37037051e-01
5.50386488e-01 9.43432003e-02 -1.14513767e+00 -1.09496854e-01
6.22569144e-01 8.46782327e-02 -4.29491758e-01 -6.61296248e-01
-1.26438999e+00 6.59693003e-01 8.49542737e-01 7.59044960e-02
-3.87627810e-01 4.55651999e-01 3.89876425e-01 4.51615810e-01
8.17656994e-01 -5.46875112e-02 -7.85888910e-01 1.05030306e-01
-4.38651174e-01 9.87639651e-02 1.12870526e+00 1.48953688e+00
8.71224403e-01 -6.42585084e-02 -7.02781603e-02 4.97693747e-01
1.37716413e-01 1.18583038e-01 1.97917089e-01 -3.46020639e-01
8.10463846e-01 1.16173708e+00 5.07019907e-02 -1.04005921e+00
-2.02723339e-01 -3.84751856e-01 -6.16482198e-01 -6.16450906e-01
1.67888165e-01 7.82751888e-02 -1.14468718e+00 1.61071813e+00
6.89948320e-01 8.32001686e-01 1.82603121e-01 3.96531492e-01
1.07765234e+00 7.07615077e-01 7.88027570e-02 -5.51960096e-02
8.74019563e-01 -1.09691167e+00 -7.55875468e-01 -9.90893319e-02
1.05855989e+00 6.41123652e-02 6.35436714e-01 -7.22224340e-02
-6.42732024e-01 -2.98175037e-01 -9.28130805e-01 -2.90449768e-01
-1.24034941e+00 -2.80906409e-01 9.47077751e-01 1.20843090e-01
-1.41875446e+00 9.07044172e-01 -8.68894279e-01 -6.56530201e-01
3.16231906e-01 3.06792170e-01 -5.51583350e-01 -3.71681094e-01
-1.79878521e+00 9.57357824e-01 1.18342745e+00 4.12158906e-01
-7.31735826e-01 -1.04117572e+00 -1.16704476e+00 1.02718212e-01
8.00906241e-01 -9.43477452e-01 5.39079249e-01 -1.68950722e-01
-9.72572088e-01 5.67297339e-01 -4.15376760e-02 -5.75858772e-01
2.33702525e-01 -3.07243168e-01 -1.07469749e+00 -6.40010461e-02
1.15536270e-03 4.33057129e-01 6.76885366e-01 -1.11804795e+00
-3.51106256e-01 -3.42133492e-01 3.35323587e-02 6.11962490e-02
-6.40403569e-01 -7.22847044e-01 -6.80347085e-01 -4.17565405e-01
-1.18120335e-01 -7.88324475e-01 2.95115203e-01 -3.48817319e-01
-5.82419693e-01 -7.19527423e-01 1.11838317e+00 -7.86581933e-01
1.65560269e+00 -1.65991604e+00 5.09838164e-01 1.40736669e-01
5.81264853e-01 4.08642411e-01 -1.20929345e-01 1.10957766e+00
-1.82427973e-01 1.10905254e-02 2.60201186e-01 -1.30062908e-01
8.82565603e-02 7.51794398e-01 -4.68858540e-01 -2.30643470e-02
3.14010233e-02 1.77535570e+00 -1.53174102e+00 -5.29390395e-01
2.05191430e-02 3.48179132e-01 -2.49980971e-01 -9.88472486e-04
-6.10449851e-01 -1.70066804e-01 -3.92095745e-01 9.46094573e-01
4.88374084e-01 -7.28004336e-01 6.98766828e-01 -4.33733463e-01
3.37495118e-01 9.06106830e-02 -6.63886905e-01 1.90788627e+00
-2.05711007e-01 3.61810058e-01 -8.03069055e-01 -1.06653297e+00
7.63495684e-01 4.19320643e-01 3.88046801e-01 -3.86893511e-01
-2.70882845e-01 1.33396685e-01 -4.57770973e-01 -6.17302001e-01
9.30062592e-01 2.29531020e-01 7.21905306e-02 1.39134303e-01
7.50490010e-01 4.11614835e-01 2.75672823e-01 1.13108552e+00
1.50110734e+00 6.07679248e-01 1.79579958e-01 2.71057308e-01
-1.37732551e-02 6.11268654e-02 6.54443145e-01 4.75957096e-01
-1.36006519e-01 -6.87842071e-02 5.85979164e-01 -6.48940980e-01
-8.63383234e-01 -1.51046085e+00 4.72036004e-01 8.72799218e-01
2.42348120e-01 -1.03850019e+00 3.53049603e-03 -1.27259290e+00
6.12424135e-01 7.34101772e-01 -1.03467774e+00 -5.51988482e-01
-5.36285281e-01 -5.43410718e-01 3.67356449e-01 9.09298003e-01
3.46991062e-01 -8.41906071e-01 5.31335115e-01 5.29165864e-01
-1.64176583e-01 -1.43940222e+00 -3.62699836e-01 -5.75828552e-02
-9.25989032e-01 -1.39734888e+00 -3.03938359e-01 -8.40965629e-01
7.03397512e-01 1.45837799e-01 1.32155073e+00 -1.18160978e-01
-1.10250272e-01 8.99955332e-01 -8.11646938e-01 -5.63016310e-02
-5.79709522e-02 1.87832817e-01 2.72692889e-01 -4.96301726e-02
5.48006773e-01 -1.06797969e+00 -6.05364561e-01 -9.37778875e-02
-7.69285679e-01 4.24816646e-02 4.15510386e-01 8.83976817e-01
4.95078802e-01 1.63472146e-01 1.00552511e+00 -1.20185924e+00
4.86595541e-01 -1.00516498e+00 -3.13043892e-01 8.96245122e-01
-1.11237228e+00 1.63327560e-01 5.97813547e-01 -6.06915295e-01
-9.55042243e-01 -4.03087974e-01 5.84240496e-01 -8.33137691e-01
7.29896426e-01 1.14445877e+00 1.07033625e-01 -1.82993859e-01
6.26286030e-01 2.85545200e-01 -5.14860094e-01 -2.64167398e-01
1.19997108e+00 5.78287654e-02 6.01562858e-01 -6.11934960e-01
1.01515353e+00 3.45681965e-01 -1.67785212e-02 -4.34873492e-01
-9.00575042e-01 -4.52587783e-01 -6.33156240e-01 -3.20679814e-01
4.43688452e-01 -1.01042080e+00 -5.56927085e-01 3.01178187e-01
-1.09098232e+00 -4.89118576e-01 -6.46508694e-01 4.74413097e-01
-1.89243421e-01 4.64100987e-01 -8.05615902e-01 -5.00933468e-01
-2.39132375e-01 -7.93374628e-02 7.53205240e-01 9.10853371e-02
1.28786400e-01 -1.63039887e+00 2.80851305e-01 1.76422611e-01
1.74834952e-01 5.99129558e-01 1.07476616e+00 -5.33787668e-01
-1.04893267e+00 -2.22399250e-01 -2.15650052e-01 1.37565911e-01
3.90705436e-01 -7.01767877e-02 -5.95509589e-01 -1.33138493e-01
-9.32485521e-01 -4.70834315e-01 1.17003322e+00 1.34644052e-02
5.68833649e-01 -5.06230295e-01 -9.81437445e-01 5.29466987e-01
1.62693977e+00 -1.43555626e-01 6.44254565e-01 1.53703451e-01
1.25256109e+00 3.31358910e-01 4.24829304e-01 9.12331119e-02
1.14504206e+00 3.79337698e-01 3.93857479e-01 3.87848556e-01
-3.63920510e-01 -1.16542029e+00 3.19469512e-01 1.39083314e+00
-3.09603035e-01 -5.08314252e-01 -8.51969957e-01 1.21448290e+00
-2.45103717e+00 -8.21720958e-01 -1.19615190e-01 1.82083046e+00
9.40614879e-01 -9.53960493e-02 -1.84276089e-01 -1.28968850e-01
7.74622083e-01 2.50718892e-01 -6.46486223e-01 -7.52838776e-02
-1.36559978e-01 1.13099873e-01 5.39231181e-01 5.26672482e-01
-7.95650065e-01 1.30662501e+00 5.18814754e+00 6.78286970e-01
-4.69728440e-01 2.39295065e-01 -1.58794031e-01 1.23312101e-01
-5.49020946e-01 5.32126129e-01 -8.35529506e-01 2.82996953e-01
1.06303453e+00 -6.08362377e-01 5.51059663e-01 5.86161971e-01
-4.08624023e-01 1.63884684e-01 -1.31796014e+00 8.03735912e-01
-5.91144450e-02 -1.58791590e+00 3.02223772e-01 -8.75342637e-02
1.03058684e+00 1.52234033e-01 -1.16813630e-01 8.96815419e-01
9.80243683e-01 -6.96004987e-01 2.00881436e-01 8.65106344e-01
8.92350376e-01 -2.79167503e-01 5.21011889e-01 8.81125852e-02
-1.64670098e+00 -1.06381346e-02 -2.96990752e-01 1.40978768e-01
4.04824674e-01 7.64011979e-01 -1.18917429e+00 1.60824704e+00
6.43989980e-01 1.19609773e+00 -5.73234081e-01 6.38558805e-01
-6.16872370e-01 7.51196623e-01 -3.14639270e-01 1.31273851e-01
7.02485740e-02 -2.19406277e-01 3.65033269e-01 8.17225575e-01
3.42801094e-01 1.33564129e-01 -1.14479303e-01 7.33451188e-01
-5.99813402e-01 -2.08529606e-01 -9.82437134e-01 -4.88502145e-01
7.90996552e-01 1.16278851e+00 -3.94962013e-01 -4.88002151e-01
-6.96209610e-01 1.10261393e+00 1.02456164e+00 9.06647921e-01
-7.59642601e-01 -2.83083439e-01 2.26962432e-01 4.65386659e-02
6.41035974e-01 -2.28670344e-01 3.60824168e-01 -1.58799589e+00
3.21634054e-01 7.05973536e-04 9.99355018e-01 -8.80520999e-01
-1.53286767e+00 3.93194973e-01 2.16417074e-01 -9.24320161e-01
-1.08746879e-01 -3.59362185e-01 -3.87508839e-01 5.69945574e-01
-1.59013462e+00 -1.80545008e+00 -3.27042222e-01 7.52881229e-01
-2.54970223e-01 8.65127668e-02 7.09617853e-01 2.37957239e-01
-4.23726290e-01 5.14293313e-01 1.37839049e-01 3.25008005e-01
4.88121539e-01 -1.41523504e+00 6.72344923e-01 8.24672878e-01
4.20975119e-01 6.72202289e-01 3.64971220e-01 -1.25479555e+00
-1.66747487e+00 -1.69724858e+00 1.27692652e+00 -7.28818774e-01
1.16947687e+00 -3.88498753e-01 -1.08863187e+00 1.64882815e+00
-9.99493059e-03 7.92819321e-01 4.23067629e-01 6.47146046e-01
-8.01161706e-01 -3.01606089e-01 -7.24075854e-01 6.05994463e-01
1.69153380e+00 -7.21494138e-01 -7.85698175e-01 4.68503118e-01
1.25893831e+00 -3.76259059e-01 -1.45745206e+00 5.32206774e-01
3.77509475e-01 -3.30204457e-01 1.02958190e+00 -9.71171260e-01
1.71320707e-01 -3.70043695e-01 1.29330844e-01 -1.64733291e+00
-2.53248543e-01 -6.78187370e-01 -1.39913356e+00 1.30574381e+00
3.81218284e-01 -9.44792688e-01 1.09514236e+00 4.87628788e-01
-1.20896250e-01 -8.67501974e-01 -9.22489524e-01 -1.21968091e+00
-2.07689136e-01 -1.88600704e-01 6.20324671e-01 1.42762518e+00
5.75282574e-01 5.35051286e-01 -5.69070935e-01 4.37762082e-01
7.72538304e-01 2.07666188e-01 5.71708202e-01 -1.18351710e+00
-1.19921789e-01 3.14424545e-01 -7.82762825e-01 -8.13980877e-01
2.64100134e-01 -1.34522831e+00 -6.62954271e-01 -2.12788773e+00
-4.45866287e-02 -3.89405012e-01 -4.57983226e-01 8.30003858e-01
-3.55623037e-01 -2.83314139e-01 -2.98918694e-01 -6.15035333e-02
-9.24599290e-01 1.13697946e+00 9.95656848e-01 -3.58019352e-01
-2.96042144e-01 -5.30292690e-01 -3.98211896e-01 2.16729775e-01
3.79543662e-01 -3.49428654e-01 -9.98812616e-01 -3.70507449e-01
8.89639437e-01 5.29081896e-02 5.19969583e-01 -4.91277695e-01
7.96986878e-01 1.01571947e-01 1.06395133e-01 -6.15017116e-01
3.88058990e-01 -6.92353487e-01 3.84659171e-01 -1.90365873e-02
1.37222856e-01 -2.14110762e-01 1.73249424e-01 1.48814940e+00
-2.54263461e-01 3.75111550e-01 -3.56254667e-01 7.07750022e-02
-1.30328786e+00 8.71411502e-01 4.74617392e-01 2.08900288e-01
1.10086286e+00 -2.97305286e-01 -7.96806276e-01 -2.70025462e-01
-1.23413050e+00 7.65231609e-01 5.02156802e-02 6.24200463e-01
1.00515366e+00 -1.75433767e+00 -4.65691596e-01 -4.07416373e-01
6.37431204e-01 2.34213933e-01 4.67498362e-01 8.05311203e-01
-1.00767359e-01 3.35310191e-01 4.10559356e-01 -2.49322690e-02
-6.66969121e-01 1.06269431e+00 7.38313347e-02 -6.39519095e-01
-9.13012981e-01 8.95422161e-01 -1.95238933e-01 -6.97311103e-01
-3.13369930e-02 -5.11613429e-01 4.17860635e-02 5.82496496e-03
-1.20202802e-01 5.56069314e-01 6.49850219e-02 -1.60138950e-01
-3.24237347e-01 1.68322295e-01 -4.10625815e-01 2.86864191e-01
1.50106442e+00 -1.27576083e-01 -1.14356928e-01 5.27128935e-01
1.08628309e+00 -1.70945585e-01 -1.01294184e+00 -8.92248631e-01
3.52171659e-01 -2.22576678e-01 -1.86304554e-01 -8.11634421e-01
-9.88797009e-01 2.01928735e-01 -2.00662494e-01 3.29513162e-01
7.83534050e-01 3.52410316e-01 1.02449346e+00 4.35257524e-01
7.22682059e-01 -1.02402449e+00 5.19609675e-02 4.35462922e-01
6.64835632e-01 -9.72022116e-01 1.13896839e-01 -6.70896709e-01
-3.75301600e-01 9.61826980e-01 6.54940426e-01 4.21903618e-02
1.00915480e+00 -2.23630726e-01 -5.53811610e-01 -6.83777332e-01
-1.21403539e+00 -2.83483922e-01 2.31651798e-01 9.68896151e-01
-1.90432608e-01 2.81391472e-01 -4.21500728e-02 5.27493775e-01
9.90813971e-02 5.32071412e-01 3.48580122e-01 9.59717274e-01
3.20954155e-03 -1.23700440e+00 3.01608831e-01 5.88999689e-01
2.31326625e-01 -3.42360586e-01 -4.63594109e-01 1.10258472e+00
1.06400326e-01 7.25079775e-01 -3.73860240e-01 -8.92172396e-01
3.85836452e-01 3.52250189e-01 6.23748124e-01 -6.38404608e-01
-1.43087775e-01 -9.34150934e-01 5.02375603e-01 -5.17343462e-01
-3.51808965e-01 -2.75959760e-01 -1.07903242e+00 -4.96628046e-01
-6.14578187e-01 3.71952564e-01 -1.32715246e-02 7.24125206e-01
6.90562963e-01 5.90591490e-01 1.87539697e-01 -2.28876784e-01
-1.11724727e-01 -9.66516256e-01 -1.01419246e+00 3.79391432e-01
3.83275673e-02 -8.91723275e-01 -1.84840783e-01 1.04510717e-01] | [8.667867660522461, 7.956736087799072] |
9ebef11b-64ca-49a4-9d1a-22393ddca78b | hyperspectral-unmixing-based-on-clustered | 1812.10788 | null | http://arxiv.org/abs/1812.10788v1 | http://arxiv.org/pdf/1812.10788v1.pdf | Hyperspectral Unmixing Based on Clustered Multitask Networks | Hyperspectral remote sensing is a prominent research topic in data
processing. Most of the spectral unmixing algorithms are developed by adopting
the linear mixing models. Nonnegative matrix factorization (NMF) and its
developments are used widely for estimation of signatures and fractional
abundances in the SU problem. Sparsity constraints was added to NMF, and was
regularized by $ L_ {q} $ norm. In this paper, at first hyperspectral images
are clustered by fuzzy c- means method, and then a new algorithm based on
sparsity constrained distributed optimization is used for spectral unmixing. In
the proposed algorithm, a network including clusters is employed. Each pixel in
the hyperspectral images considered as a node in this network. The proposed
algorithm is optimized with diffusion LMS strategy, and then the update
equations for fractional abundance and signature matrices are obtained.
Simulation results based on defined performance metrics illustrate advantage of
the proposed algorithm in spectral unmixing of hyperspectral data compared with
other methods. | ['Sara Khoshsokhan', 'Roozbeh Rajabi', 'Hadi Zayyani'] | 2018-12-27 | null | null | null | null | ['hyperspectral-unmixing'] | ['computer-vision'] | [ 6.54421329e-01 -7.28742123e-01 -5.41894957e-02 3.63652036e-02
6.18099496e-02 -6.16354346e-01 9.28592905e-02 -1.82631105e-01
-3.71974438e-01 6.57156467e-01 -9.12883580e-02 -1.66312382e-01
-7.77084649e-01 -7.58276880e-01 1.58971269e-02 -1.34795320e+00
-1.63526893e-01 2.73762077e-01 -4.69693810e-01 -1.62036698e-02
1.61432892e-01 7.02411294e-01 -1.55484104e+00 -2.44070031e-02
1.36473727e+00 9.43604708e-01 4.72214341e-01 4.98924226e-01
-2.07959622e-01 5.94475687e-01 -4.02012259e-01 3.67196470e-01
7.33742476e-01 -4.94875878e-01 -2.77335763e-01 7.44376421e-01
1.64074793e-01 4.77820821e-02 -7.17953742e-02 1.79571939e+00
5.18630803e-01 6.26796663e-01 7.29232132e-01 -1.28947282e+00
-3.50351185e-01 9.36964214e-01 -1.28975916e+00 2.36924082e-01
-4.60565716e-01 -2.33040020e-01 6.89067066e-01 -8.03257883e-01
3.41617703e-01 1.20542991e+00 5.15211821e-01 -4.74070525e-03
-1.00837898e+00 -9.26235735e-01 1.99172378e-01 2.28437796e-01
-1.74399579e+00 -3.03932488e-01 6.98052347e-01 -5.44526160e-01
5.37366152e-01 5.29149652e-01 8.99864137e-01 -3.71811002e-01
-1.66122481e-01 4.88064557e-01 1.28443527e+00 -5.71897507e-01
2.44735599e-01 -2.35171273e-01 3.46705675e-01 4.02230799e-01
7.62987256e-01 -9.49475095e-02 -3.80988307e-02 -4.21802491e-01
5.34050405e-01 4.55330104e-01 -5.11813581e-01 -1.27195418e-01
-1.14960325e+00 8.23641598e-01 3.78181458e-01 4.15711343e-01
-8.22117269e-01 -2.52824724e-01 7.88796041e-03 3.63570213e-01
5.45976877e-01 7.46822283e-02 3.57149690e-02 6.47600412e-01
-1.45460951e+00 -2.37497941e-01 4.14598495e-01 5.70749044e-01
9.89833653e-01 4.39456344e-01 2.58689970e-01 1.03894103e+00
6.98631048e-01 1.12672067e+00 5.30492544e-01 -9.92398202e-01
2.01679453e-01 6.92308664e-01 2.19257116e-01 -1.38368928e+00
-5.30595422e-01 -6.56196535e-01 -1.41720080e+00 2.14551821e-01
-1.57578394e-01 -5.30944765e-01 -9.33020234e-01 1.16540325e+00
3.19748551e-01 6.31877661e-01 2.89175957e-01 1.12326992e+00
2.97768235e-01 1.26657403e+00 -3.77623700e-02 -1.10037529e+00
7.90220737e-01 -9.11205709e-01 -1.03150904e+00 -2.23857909e-02
1.01945698e-01 -9.06307578e-01 1.40251100e-01 6.59732044e-01
-6.72023952e-01 -3.41191709e-01 -1.13504601e+00 7.74599671e-01
-1.57842070e-01 3.18117082e-01 5.82784653e-01 8.04123998e-01
-9.90532339e-01 5.78402698e-01 -6.71106219e-01 -4.26352590e-01
1.50146753e-01 5.84760666e-01 -1.64052337e-01 -2.77639091e-01
-8.89789701e-01 3.54636937e-01 7.14657247e-01 6.61935568e-01
-4.72960830e-01 -2.31348291e-01 -4.74389344e-01 -1.20193705e-01
1.31408796e-01 -4.83965129e-01 6.30869448e-01 -1.55929160e+00
-1.28701544e+00 3.62979025e-01 -2.92526901e-01 -1.89450979e-01
-1.68136843e-02 3.55321348e-01 -8.57659698e-01 3.68352205e-01
2.09746119e-02 2.31043622e-02 1.16732538e+00 -1.38559783e+00
-8.69156659e-01 -5.90045810e-01 -4.39252406e-01 5.36145270e-01
-5.72528422e-01 3.55640873e-02 1.52964815e-01 -7.03576624e-01
1.03327680e+00 -1.03961217e+00 -7.23236263e-01 -4.18871224e-01
-1.38024554e-01 3.71941239e-01 1.04270017e+00 -7.93686152e-01
1.38042927e+00 -2.15736699e+00 3.30458581e-01 1.10997832e+00
6.76763877e-02 4.69440818e-01 -1.91982150e-01 4.52608556e-01
-4.03622627e-01 -6.57872390e-03 -7.20217705e-01 1.26789480e-01
-3.06727648e-01 1.09973542e-01 -1.00697018e-01 8.45104277e-01
-4.51113999e-01 2.92058941e-02 -9.40153599e-01 -1.97259247e-01
3.19594324e-01 3.29046965e-01 -3.33326869e-03 -4.44994271e-02
-2.79979650e-02 3.01011026e-01 -3.04840803e-01 7.62023270e-01
1.39686346e+00 -3.02047879e-01 6.00917816e-01 -5.05655289e-01
-5.45333385e-01 -1.03509068e+00 -1.99109042e+00 1.46036804e+00
5.86938299e-02 1.31582692e-01 8.33707213e-01 -1.19458807e+00
8.07883203e-01 4.47745740e-01 9.04673636e-01 1.54534832e-01
2.44692773e-01 4.48282689e-01 3.16725791e-01 -5.77123165e-01
4.74223971e-01 -3.09362888e-01 8.76114905e-01 7.32873976e-01
-2.75542051e-01 -1.21481515e-01 5.50458670e-01 7.54388124e-02
3.16057801e-01 -3.69136691e-01 2.64016688e-01 -6.72454178e-01
9.90423322e-01 3.04658353e-01 5.88728666e-01 4.22901899e-01
-1.23023447e-02 -3.89183648e-02 -3.59275669e-01 5.95403276e-02
-6.65079832e-01 -4.73757982e-01 -7.50862509e-02 7.58752584e-01
3.43825966e-01 3.78194898e-02 -4.06925559e-01 3.82623784e-02
7.42890686e-02 3.62989992e-01 -2.27378607e-01 3.25750053e-01
1.24329058e-02 -1.60633647e+00 2.59215315e-03 -1.65688306e-01
6.23347223e-01 -5.01991987e-01 9.61421430e-02 3.39823216e-01
-8.34074840e-02 -4.92280424e-01 -1.00490533e-01 -8.35758671e-02
-1.26611853e+00 -1.00567937e+00 -8.01833093e-01 -7.56409764e-01
1.13261151e+00 1.06853592e+00 2.76788682e-01 -2.63774805e-02
-6.21122047e-02 1.72476038e-01 -5.27307868e-01 -3.42662603e-01
-2.70370871e-01 -2.16946453e-01 4.00251955e-01 7.20096946e-01
8.66862312e-02 -7.43791342e-01 -6.85643613e-01 2.22526670e-01
-1.53348696e+00 -2.60534614e-01 5.95112264e-01 6.90732658e-01
7.42361009e-01 1.16337037e+00 3.60993087e-01 -9.25242245e-01
7.00111210e-01 -5.51091433e-01 -7.35689759e-01 3.39556336e-01
-8.46758842e-01 -5.02717018e-01 4.84292179e-01 -1.79061219e-01
-1.14106214e+00 2.66146541e-01 5.73869228e-01 -5.62623322e-01
1.69623241e-01 1.14183176e+00 -3.10521305e-01 -5.16298175e-01
6.14251018e-01 2.53328800e-01 4.22751307e-01 -3.29249442e-01
4.57899123e-01 1.13388503e+00 1.69337228e-01 -2.60233462e-01
1.01332235e+00 8.64884138e-01 1.79336995e-01 -1.24136734e+00
-5.60265541e-01 -9.68021214e-01 -4.76871759e-01 -3.24687421e-01
6.21481419e-01 -1.11741006e+00 -4.14475352e-01 6.89345837e-01
-5.72162092e-01 1.26209959e-01 -3.03439070e-02 9.55530047e-01
4.10607383e-02 7.61601627e-01 -4.07965750e-01 -1.25405490e+00
-5.22687435e-01 -9.67405021e-01 2.23980546e-01 3.58861625e-01
4.50845778e-01 -8.68639588e-01 -7.41465762e-02 3.07816148e-01
3.80661488e-01 1.61841407e-01 6.90739870e-01 -2.81255990e-01
-3.14719856e-01 -1.10475622e-01 -3.92020971e-01 7.00183332e-01
5.56122005e-01 3.74143988e-01 -6.59537077e-01 -6.77201092e-01
3.14952612e-01 3.05263996e-01 8.76696229e-01 8.34954619e-01
7.32081234e-01 -2.35689759e-01 -3.71307164e-01 7.70913601e-01
1.93712103e+00 6.23753548e-01 2.90894568e-01 2.18542531e-01
8.73775840e-01 6.27265751e-01 6.26582921e-01 6.93052769e-01
-3.40302467e-01 -3.88714910e-01 5.66601574e-01 -2.28843808e-01
5.72589993e-01 4.70833093e-01 1.30983040e-01 1.33146262e+00
-3.32538009e-01 -4.12292808e-01 -7.74679184e-01 3.01042289e-01
-2.17486978e+00 -1.22245896e+00 -6.03024602e-01 2.05212760e+00
3.80728185e-01 -7.57797539e-01 -2.41416737e-01 6.05219722e-01
1.32115042e+00 4.19684649e-01 -6.17519438e-01 3.27817738e-01
-6.68336153e-01 4.40786704e-02 9.33661103e-01 6.93790257e-01
-1.10320222e+00 6.49547160e-01 5.50997877e+00 8.69025290e-01
-1.29556072e+00 1.24349557e-01 3.91464829e-01 1.35874609e-02
-3.60359043e-01 7.58394822e-02 -2.85778940e-01 2.61663973e-01
4.70305532e-01 -2.16692150e-01 8.31664562e-01 2.03804120e-01
7.73562610e-01 -3.24328274e-01 -5.00765909e-03 1.33746207e+00
1.82755619e-01 -9.28799033e-01 2.51375049e-01 -6.94932118e-02
1.45349300e+00 3.42417024e-02 -1.10509004e-02 -5.71335614e-01
2.83172131e-01 -7.79524565e-01 3.25892061e-01 6.37794793e-01
6.71044886e-01 -9.46282625e-01 6.33242488e-01 5.02882898e-01
-1.33167255e+00 -3.94868046e-01 -7.08546519e-01 -4.99202535e-02
1.46231115e-01 1.17297590e+00 -2.84169257e-01 9.34913456e-01
2.31333569e-01 1.01581097e+00 -1.35541819e-02 1.24789345e+00
7.09163351e-03 7.35222578e-01 -5.80258608e-01 2.41366953e-01
2.72937894e-01 -1.22478771e+00 6.15834713e-01 8.65506649e-01
8.17798913e-01 6.01345599e-01 4.07655627e-01 3.96647066e-01
1.89587325e-01 7.66588032e-01 -2.93212652e-01 -5.88212550e-01
5.89294732e-01 1.55797076e+00 -9.49994564e-01 -5.26708603e-01
-1.22659013e-01 7.59986639e-01 -5.36915302e-01 6.37876391e-01
-3.96019518e-01 -1.98185638e-01 4.79629636e-01 -1.26572192e-01
1.36985015e-02 -3.15660089e-01 9.69270021e-02 -1.28210771e+00
-5.55831194e-01 -9.93757248e-01 5.80806136e-01 -7.80192435e-01
-1.28835654e+00 4.91185695e-01 -2.30660468e-01 -1.45803535e+00
4.03651863e-01 -6.30444229e-01 -3.05437595e-01 1.22838831e+00
-1.74013901e+00 -8.29567671e-01 -5.14113247e-01 8.95802677e-01
7.13119730e-02 -4.77947146e-01 6.07353389e-01 4.33073133e-01
-9.09020960e-01 -2.93252975e-01 1.00559795e+00 -3.42549145e-01
4.83108431e-01 -8.25290322e-01 -7.29950726e-01 1.45003569e+00
-6.04755506e-02 6.92938268e-01 5.40460348e-01 -8.48473012e-01
-1.25702238e+00 -1.26850748e+00 3.03145081e-01 7.65951872e-01
8.32511425e-01 5.61163425e-01 -4.63381618e-01 4.16201591e-01
4.99802649e-01 -5.36394678e-03 1.26155663e+00 -4.65679646e-01
1.52284622e-01 -3.73684585e-01 -1.23886633e+00 2.02297345e-01
5.62950552e-01 -1.24205224e-01 1.62082851e-01 7.90094137e-01
2.45721087e-01 7.50238150e-02 -9.99177516e-01 2.92507023e-01
2.06547409e-01 -6.60502791e-01 7.30757415e-01 -3.86683971e-01
-1.72001481e-01 -1.14932442e+00 -3.45442772e-01 -1.67152488e+00
-8.82334769e-01 -7.19497919e-01 2.92088687e-01 9.24088955e-01
3.49561751e-01 -7.34707832e-01 6.64506018e-01 -2.54278593e-02
1.26719981e-01 1.95053086e-01 -4.01935786e-01 -5.46698809e-01
-4.78528559e-01 7.49164373e-02 6.97178662e-01 1.47121811e+00
-1.21052101e-01 2.17989326e-01 -3.50223273e-01 7.60658681e-01
1.08459663e+00 3.15047413e-01 1.69832900e-01 -1.57599258e+00
-2.03896746e-01 -4.45340812e-01 1.51676945e-02 -6.37767255e-01
2.82444060e-01 -1.10940766e+00 -2.01064438e-01 -1.58751309e+00
3.80075350e-02 -4.30060089e-01 -5.38931370e-01 1.32925376e-01
-2.40449771e-01 3.10836375e-01 1.76389366e-01 7.50990570e-01
5.34406453e-02 1.48466736e-01 1.06382096e+00 -5.60050428e-01
-4.19334561e-01 5.74214868e-02 -5.01402080e-01 5.54999173e-01
9.02536452e-01 -1.59598082e-01 -7.54114509e-01 -5.13947070e-01
2.97741592e-01 3.25209081e-01 -4.13252890e-01 -1.01311183e+00
3.98645788e-01 -5.06349385e-01 2.29854226e-01 -5.23872733e-01
1.44360155e-01 -1.41616118e+00 8.06283534e-01 4.08249378e-01
4.67177331e-02 -1.22111216e-02 -2.15800479e-02 6.03585482e-01
-4.53959972e-01 -4.77059633e-01 8.65394175e-01 -2.79250145e-01
-8.11945200e-01 6.51553631e-01 -5.38882315e-01 -7.52391696e-01
9.71599996e-01 -3.49673718e-01 -1.92454189e-01 -1.62306711e-01
-1.09943128e+00 2.49910489e-01 2.64004022e-01 -3.30503404e-01
4.99913365e-01 -1.26782477e+00 -9.19748902e-01 2.04696789e-01
-1.00440361e-01 -1.24049947e-01 6.08247876e-01 1.07905531e+00
-1.06039119e+00 -4.14412878e-02 -2.53365457e-01 -4.02095050e-01
-1.31361580e+00 3.64711434e-01 4.73413289e-01 8.93703178e-02
3.45104486e-01 9.33858097e-01 -3.95480216e-01 -3.61326784e-01
-2.65438467e-01 1.33733585e-01 -5.61245918e-01 5.81329465e-01
5.58802485e-01 7.33264387e-01 1.88315306e-02 -1.08425701e+00
-3.05441637e-02 6.92428768e-01 5.23305058e-01 -3.80618215e-01
1.36840224e+00 -4.16194826e-01 -1.36184561e+00 2.71064103e-01
7.95181751e-01 1.89628363e-01 -6.56815827e-01 -3.63224685e-01
-3.74830216e-01 -5.49364567e-01 4.00282681e-01 -5.34041882e-01
-1.41942358e+00 6.34886742e-01 8.51676643e-01 3.73901218e-01
1.47718108e+00 -1.13756573e+00 3.13130200e-01 3.15318912e-01
2.43052945e-01 -1.08010972e+00 -7.27459610e-01 3.41967881e-01
3.43380362e-01 -1.04067504e+00 3.79973233e-01 -5.82765877e-01
-2.91729808e-01 1.19117069e+00 3.52404177e-01 -5.12780547e-02
1.09990501e+00 1.32634848e-01 4.17494088e-01 3.82053517e-02
1.59816463e-02 -2.68758476e-01 -1.14420742e-01 3.85761738e-01
3.84780169e-01 3.24949712e-01 -6.51456118e-01 7.77312070e-02
2.17358142e-01 -1.07447892e-01 5.14183462e-01 8.23769093e-01
-9.61379349e-01 -1.02076423e+00 -1.07392621e+00 8.09480786e-01
-2.17822865e-01 -2.26819485e-01 -7.36970231e-02 -6.92189857e-03
3.73394847e-01 1.40807319e+00 -1.18584909e-01 -4.25644070e-01
-6.78959787e-02 -4.02314402e-02 2.01240107e-01 -5.48052192e-01
-3.11058432e-01 5.12757540e-01 -4.37466800e-01 4.57715727e-02
-1.35407412e+00 -7.18017340e-01 -1.23820353e+00 -2.91441917e-01
-8.18776488e-01 8.72287393e-01 7.73567617e-01 4.96757179e-01
7.85187185e-02 1.96864292e-01 9.67807591e-01 -8.75643849e-01
-4.95845348e-01 -9.80817080e-01 -1.58503139e+00 1.24338508e-01
-8.16240907e-03 -2.10512772e-01 -7.25820959e-01 2.53679246e-01] | [10.070609092712402, -2.0452301502227783] |
f9ac094d-b6b8-4b88-b7d8-f8a029cef841 | nested-grassmanns-for-dimensionality | 2010.14589 | null | https://arxiv.org/abs/2010.14589v3 | https://arxiv.org/pdf/2010.14589v3.pdf | Nested Grassmannians for Dimensionality Reduction with Applications | In the recent past, nested structures in Riemannian manifolds has been studied in the context of dimensionality reduction as an alternative to the popular principal geodesic analysis (PGA) technique, for example, the principal nested spheres. In this paper, we propose a novel framework for constructing a nested sequence of homogeneous Riemannian manifolds. Common examples of homogeneous Riemannian manifolds include the $n$-sphere, the Stiefel manifold, the Grassmann manifold and many others. In particular, we focus on applying the proposed framework to the Grassmann manifold, giving rise to the nested Grassmannians (NG). An important application in which Grassmann manifolds are encountered is planar shape analysis. Specifically, each planar (2D) shape can be represented as a point in the complex projective space which is a complex Grass-mann manifold. Some salient features of our framework are: (i) it explicitly exploits the geometry of the homogeneous Riemannian manifolds and (ii) the nested lower-dimensional submanifolds need not be geodesic. With the proposed NG structure, we develop algorithms for the supervised and unsupervised dimensionality reduction problems respectively. The proposed algorithms are compared with PGA via simulation studies and real data experiments and are shown to achieve a higher ratio of expressed variance compared to PGA. | ['Baba C. Vemuri', 'Chun-Hao Yang'] | 2020-10-27 | null | null | null | null | ['motion-segmentation'] | ['computer-vision'] | [-1.49336979e-01 6.85378760e-02 4.98399943e-01 -1.26891181e-01
-1.18255980e-01 -6.35860384e-01 4.03684616e-01 -4.46037948e-01
-2.02734515e-01 4.20399100e-01 2.42751706e-02 -2.53213972e-01
-6.44921184e-01 -6.14343047e-01 -2.90810615e-01 -9.93473768e-01
-2.82188654e-01 6.36273203e-03 1.22261085e-02 -2.83761472e-01
3.86128336e-01 8.26207221e-01 -1.38864541e+00 -9.13271964e-01
1.09312391e+00 5.14834106e-01 7.35095441e-02 6.27009034e-01
1.04643367e-01 1.10838540e-01 -2.56857695e-03 -4.34654832e-01
3.73431861e-01 -2.65828699e-01 -9.50765431e-01 6.31795824e-01
3.16443890e-02 1.20491147e-01 -4.60591048e-01 1.20282662e+00
1.00162201e-01 4.29405183e-01 8.09511125e-01 -1.20333731e+00
-2.64239490e-01 1.02959320e-01 -6.25815034e-01 2.80061737e-02
-4.72840816e-02 -3.08340639e-01 9.72467840e-01 -1.31694508e+00
5.79545379e-01 1.42062473e+00 1.03955857e-01 3.80945385e-01
-8.98736060e-01 6.54296130e-02 -6.05481900e-02 9.05926749e-02
-1.81907427e+00 -2.06615403e-01 1.01087368e+00 -5.98742902e-01
3.56393367e-01 5.61071515e-01 1.85523018e-01 1.32676378e-01
2.67556280e-01 5.20033836e-01 8.42481554e-01 -4.30662483e-01
1.06834479e-01 -1.17400242e-03 4.23163384e-01 8.05031180e-01
4.13606137e-01 -3.88923883e-01 3.02726299e-01 -2.82499678e-02
5.93317449e-01 1.64720803e-01 -1.18187308e-01 -9.40273464e-01
-1.28353286e+00 8.04802001e-01 2.70116180e-01 5.87977588e-01
-1.89452812e-01 -2.95536697e-01 -8.01457614e-02 6.21929113e-03
2.08476067e-01 1.59018397e-01 2.03370854e-01 7.77629204e-03
-1.54405057e-01 1.69526160e-01 7.82176852e-01 1.17263937e+00
1.04436946e+00 1.76069379e-01 4.80811477e-01 6.75498545e-01
7.07224905e-01 4.35875833e-01 3.95140871e-02 -9.13379967e-01
7.68764853e-01 9.17317152e-01 1.59993127e-01 -1.57386911e+00
-6.43355370e-01 -1.70248434e-01 -9.08095062e-01 3.94257866e-02
3.43548208e-01 -1.67716667e-01 -3.70510891e-02 1.50014555e+00
5.88790357e-01 1.40073942e-02 3.72504741e-01 9.99640405e-01
3.19602340e-01 3.97273839e-01 -4.46541548e-01 -7.92410597e-02
1.21045363e+00 -6.10784233e-01 -5.08219659e-01 3.33444655e-01
9.90813673e-01 -8.51775646e-01 9.02125537e-01 1.21630922e-01
-8.90263855e-01 -2.11549938e-01 -1.22630858e+00 1.06012300e-01
-1.68594509e-01 3.88392061e-01 2.37954304e-01 7.77599394e-01
-1.10335994e+00 8.83228064e-01 -8.16231489e-01 -7.97602177e-01
-1.83291510e-01 3.14144850e-01 -5.08419752e-01 -3.70661877e-02
-7.62459099e-01 6.67442799e-01 7.41155744e-02 4.36218500e-01
-6.30765080e-01 -6.53888732e-02 -8.83201003e-01 -2.52130657e-01
1.51358396e-01 -2.90276617e-01 5.69150627e-01 -3.25320572e-01
-1.46785486e+00 6.65863097e-01 -7.93029666e-02 1.09162636e-01
2.39805862e-01 -8.37424472e-02 -4.66958553e-01 2.33337209e-01
-8.33599120e-02 4.40389514e-02 7.81819403e-01 -8.33371520e-01
-4.63177383e-01 -7.84159303e-01 5.17565846e-01 5.21996439e-01
-4.85748082e-01 -1.81576274e-02 -1.18651874e-01 -5.20796835e-01
5.80466866e-01 -1.46147871e+00 -2.58938193e-01 -4.86502469e-01
-5.94496310e-01 -4.21790481e-01 1.02648175e+00 -4.66977090e-01
1.34300840e+00 -2.32161283e+00 8.56213331e-01 3.34508181e-01
2.54541308e-01 -4.35566716e-02 8.35588723e-02 5.09774566e-01
-1.53945312e-01 1.52132049e-01 -4.23479468e-01 -2.25766882e-01
-6.19042367e-02 1.34472713e-01 8.31935331e-02 1.04306841e+00
2.13692024e-01 2.78163195e-01 -8.42659950e-01 -4.40486848e-01
2.27413431e-01 3.46726209e-01 -4.38185990e-01 9.30042937e-02
5.30753553e-01 5.48800766e-01 -6.84928775e-01 2.55117565e-01
8.82507384e-01 2.18694538e-01 8.13033357e-02 -3.66740137e-01
-5.75914800e-01 -9.81555209e-02 -1.46658564e+00 1.52297950e+00
-1.32789463e-01 2.88665116e-01 3.95368278e-01 -1.00571477e+00
1.12665379e+00 1.54214039e-01 4.71415251e-01 1.15106836e-01
2.84099609e-01 2.40569681e-01 2.32811213e-01 -3.93232107e-01
5.83330333e-01 6.17110692e-02 5.84323816e-02 4.22029585e-01
-1.78399190e-01 1.22226328e-02 4.41713244e-01 2.77046859e-01
7.95636117e-01 -1.71456975e-03 -6.99762627e-02 -1.10880542e+00
1.22032559e+00 -3.77207667e-01 5.48008025e-01 -7.70304278e-02
7.81417787e-02 5.59738874e-01 5.46012878e-01 -1.97130945e-02
-9.69805181e-01 -1.26516163e+00 -3.43029499e-01 3.07688504e-01
4.81770307e-01 -3.79966319e-01 -1.16699445e+00 -4.53168601e-01
-3.26864511e-01 5.85927546e-01 -2.85862386e-01 -2.13174328e-01
-5.59981942e-01 -8.11755955e-01 1.53325126e-01 1.21361893e-02
6.63199663e-01 -3.97025228e-01 -2.89276421e-01 -4.05475982e-02
6.65866584e-02 -9.94946659e-01 -9.10907090e-01 -5.65944731e-01
-1.30142748e+00 -1.36823046e+00 -5.74517012e-01 -8.13951313e-01
8.97040188e-01 9.45805490e-01 4.31500316e-01 -2.22249866e-01
-2.70239234e-01 9.79827881e-01 -3.65023166e-01 1.96514621e-01
-2.17109591e-01 2.49797967e-03 8.89437020e-01 8.88501048e-01
1.97071835e-01 -6.77939534e-01 -7.67112255e-01 9.92555261e-01
-1.20759606e+00 -1.85804993e-01 2.94105053e-01 3.50930899e-01
4.35967565e-01 2.60954767e-01 4.49368417e-01 -3.84194881e-01
4.83323216e-01 -6.36680424e-01 -7.07299173e-01 -1.48654198e-02
-4.45627481e-01 1.91082731e-01 4.89973456e-01 1.68585122e-01
-9.15139019e-01 -6.20033145e-02 3.17842633e-01 -4.20315623e-01
1.28817275e-01 3.93819749e-01 -7.57198691e-01 -4.15511489e-01
3.97823036e-01 1.14582285e-01 3.45183969e-01 -8.20362210e-01
4.70945626e-01 6.83099031e-01 2.50265189e-02 -5.03194332e-01
1.06659174e+00 7.66091704e-01 5.65524042e-01 -1.52036834e+00
-2.86605418e-01 -5.70437849e-01 -1.24116421e+00 -2.14399710e-01
1.01570916e+00 -3.84368539e-01 -7.19157159e-01 3.64327371e-01
-9.81584728e-01 4.11189497e-01 7.99660906e-02 8.04659963e-01
-7.38136172e-01 1.04666150e+00 -6.04506433e-01 -8.98916066e-01
-2.00926244e-01 -1.19114733e+00 6.19496405e-01 1.67481825e-01
1.03443407e-01 -1.28943765e+00 9.40803215e-02 8.47915038e-02
-5.85631877e-02 3.88492554e-01 1.01066685e+00 -2.23747417e-01
-4.86433595e-01 -2.37031221e-01 7.18647391e-02 5.80891013e-01
6.13707304e-01 4.85006087e-02 -3.59685212e-01 -4.98302996e-01
6.92451119e-01 4.45986539e-01 4.65021767e-02 1.16709305e-03
6.02700293e-01 -2.00326681e-01 -1.91572443e-01 5.08266032e-01
1.20153737e+00 3.72797787e-01 6.36778593e-01 1.89366445e-01
1.04436088e+00 1.10259426e+00 8.34148943e-01 2.48064339e-01
3.56475323e-01 6.31537974e-01 4.09290969e-01 2.18353480e-01
4.83510405e-01 1.29937887e-01 4.01601762e-01 1.80576229e+00
-3.10529798e-01 3.42697680e-01 -7.17527807e-01 4.98624414e-01
-1.81538785e+00 -5.32521546e-01 -5.59748352e-01 2.65280819e+00
1.23874381e-01 -3.93065631e-01 1.07418239e-01 1.24760576e-01
1.08064973e+00 1.54177742e-02 -3.53616089e-01 -4.16635185e-01
-3.31223845e-01 -4.01520997e-01 4.39798683e-01 6.41601205e-01
-1.23280370e+00 5.50233960e-01 5.24861765e+00 6.00927114e-01
-6.26464248e-01 -5.73402941e-02 1.74072117e-01 3.07791054e-01
-1.59811020e-01 9.46102738e-02 -8.09784651e-01 1.60863116e-01
7.77224958e-01 -4.09082204e-01 4.66172606e-01 7.79777408e-01
3.31547886e-01 8.04198608e-02 -9.90522146e-01 1.25513256e+00
1.06637433e-01 -7.54353285e-01 2.91726679e-01 4.93982166e-01
6.53298736e-01 -3.56521487e-01 2.07886964e-01 -9.40751880e-02
-1.06096163e-01 -6.52051687e-01 2.06778467e-01 4.76472765e-01
4.66468245e-01 -1.02793670e+00 4.12102610e-01 2.20251039e-01
-1.45164371e+00 1.95962906e-01 -5.75987518e-01 1.00416996e-01
-1.91662628e-02 3.27251703e-01 -4.78081375e-01 1.03774714e+00
3.75282407e-01 1.01228774e+00 -5.05896628e-01 1.01730168e+00
2.78682798e-01 3.36426556e-01 -2.84896374e-01 -8.08197632e-02
6.52204454e-01 -1.48365641e+00 1.09526026e+00 8.07244241e-01
6.37752473e-01 4.35134500e-01 -2.34036297e-01 7.34421074e-01
2.33760208e-01 6.66249990e-01 -9.78396714e-01 -5.06359898e-02
-1.39284395e-02 1.59796286e+00 -7.34036744e-01 1.12238407e-01
-7.68475831e-01 8.40132177e-01 3.63095514e-02 3.69619399e-01
-4.33795035e-01 -8.00631702e-01 1.06488478e+00 1.70535818e-02
1.61011443e-01 -7.02311695e-01 1.92360133e-01 -1.29693758e+00
1.40943676e-01 -5.18109679e-01 2.67742634e-01 -3.50157887e-01
-1.02283978e+00 6.15168989e-01 -7.11968169e-02 -1.67949784e+00
-2.05865260e-02 -8.46292257e-01 -5.58614969e-01 9.55001414e-01
-8.25490355e-01 -5.68674922e-01 -1.81953520e-01 8.99514437e-01
1.92564547e-01 -2.37883464e-01 6.27017677e-01 3.03454459e-01
-7.39892304e-01 7.20757619e-02 6.64595485e-01 1.24490924e-01
2.97267474e-02 -1.42842996e+00 3.18441331e-01 9.42311466e-01
7.73766683e-03 8.84031415e-01 5.16654074e-01 -3.76259387e-01
-2.09115171e+00 -1.28710020e+00 3.04335952e-01 -4.40275311e-01
8.91456008e-01 -3.50677371e-01 -8.20990682e-01 6.36869133e-01
-4.81737331e-02 -4.34302211e-01 5.92350304e-01 -1.31532177e-01
-4.06319164e-02 -2.15598419e-01 -1.06114447e+00 9.99434829e-01
1.10036767e+00 -3.06448549e-01 -4.14135754e-01 4.83216554e-01
6.40625715e-01 4.74808030e-02 -1.08640766e+00 3.55900228e-01
1.56206250e-01 -6.64095700e-01 6.35477006e-01 -6.62496448e-01
-2.27049083e-01 -8.06587934e-01 -4.41865772e-01 -1.27684915e+00
-2.96086341e-01 -9.34057593e-01 2.94369519e-01 1.09877813e+00
-1.26110092e-02 -7.12886274e-01 6.42660975e-01 6.17883086e-01
-2.00885937e-01 -5.60932994e-01 -1.17092490e+00 -9.52877462e-01
2.07281664e-01 -7.89116994e-02 2.66221017e-01 7.61736810e-01
2.35799238e-01 3.90209258e-01 -3.03678721e-01 3.76524001e-01
1.04545796e+00 -6.22723885e-02 6.46830738e-01 -1.26902211e+00
4.36012857e-02 -2.58658081e-01 -9.05433714e-01 -1.08095980e+00
-1.02613764e-02 -1.06941772e+00 -2.00960189e-01 -1.09249818e+00
-1.75852731e-01 -5.04111409e-01 -1.02121636e-01 -6.02622032e-01
2.48012483e-01 -1.47010043e-01 1.25437692e-01 2.86460042e-01
-4.85260636e-01 7.87987828e-01 1.12637591e+00 1.42759293e-01
-3.53107631e-01 1.75423205e-01 -4.75117862e-01 8.35294664e-01
7.44944155e-01 4.26067077e-02 -4.40526277e-01 -4.08084542e-01
4.18016268e-03 -7.61265755e-02 -8.24663043e-02 -9.44701612e-01
-7.80914277e-02 1.54509276e-01 -4.82983083e-01 -5.19249797e-01
2.94229537e-01 -8.69444728e-01 3.06357235e-01 2.68378645e-01
2.67080754e-01 2.27720708e-01 -1.87715143e-01 7.18782067e-01
-1.75686896e-01 -4.89412546e-01 9.09967721e-01 2.76895566e-03
-3.69681537e-01 4.67644572e-01 -2.94372499e-01 -7.94388354e-02
1.27729285e+00 -1.45245150e-01 -2.66990438e-02 -1.48160174e-01
-9.42253709e-01 1.72656551e-01 4.82825071e-01 5.02348244e-01
7.62700617e-01 -1.62638903e+00 -5.06976485e-01 -3.27597149e-02
1.24491431e-01 -6.20696023e-02 3.01608324e-01 1.24386287e+00
-7.19789922e-01 7.30419219e-01 -2.54107006e-02 -7.66360462e-01
-1.22714889e+00 7.84580529e-01 4.71471220e-01 1.98599562e-01
-5.82074106e-01 1.81248814e-01 7.54106998e-01 -5.30241668e-01
-1.40710756e-01 -1.49982005e-01 -5.84235191e-01 -1.34291485e-01
3.34593803e-01 8.59836459e-01 -7.85193443e-02 -1.35947919e+00
-3.37262005e-01 1.03926957e+00 1.36731014e-01 -1.43595919e-01
1.04646587e+00 -6.85325086e-01 -7.08976209e-01 6.06225908e-01
1.67609584e+00 4.18310277e-02 -8.49200606e-01 -1.67493463e-01
3.43534321e-01 -4.40841854e-01 -3.05011511e-01 5.48119426e-01
-9.61374819e-01 1.06260753e+00 4.95810330e-01 4.04174268e-01
7.52293408e-01 -1.18144825e-01 3.31071585e-01 5.44529796e-01
7.13778853e-01 -9.99587476e-01 -1.75160185e-01 4.18526441e-01
1.07299876e+00 -5.91819167e-01 -3.96925837e-01 -7.16779649e-01
-4.50444490e-01 1.29660106e+00 3.51869076e-01 -5.07960975e-01
1.05841875e+00 -5.80462992e-01 -2.70751357e-01 -2.07677528e-01
-5.35567403e-02 -1.02675930e-01 3.59870762e-01 3.05279046e-01
1.87140793e-01 2.08574831e-01 -7.11515605e-01 3.99899781e-01
-1.30595639e-01 -6.06432974e-01 8.70520473e-01 7.49274611e-01
-4.77028102e-01 -9.09575701e-01 -6.63136601e-01 2.55766362e-01
-3.73420149e-01 1.84686571e-01 -1.10730387e-01 6.65850520e-01
-3.40015531e-01 1.22009051e+00 -3.71500179e-02 -5.68219483e-01
5.68075120e-01 -1.50333390e-01 3.30622345e-01 -5.99793315e-01
1.84596509e-01 4.86732535e-02 -1.07459746e-01 -5.26185513e-01
-4.22023356e-01 -1.08661807e+00 -1.07920063e+00 -1.86904728e-01
-1.34788215e-01 6.42341256e-01 6.48907244e-01 8.19442928e-01
3.87094587e-01 -1.01434179e-01 1.23409915e+00 -7.48057783e-01
-4.63134825e-01 -9.09405529e-01 -1.37681150e+00 5.00944316e-01
8.03353563e-02 -9.63229299e-01 -7.94408083e-01 -1.48405924e-01] | [7.7716875076293945, 4.1582770347595215] |
7c751fdb-6f35-4f42-9f3a-c3beefc798f2 | indudonet-an-interpretable-dual-domain | 2109.05298 | null | https://arxiv.org/abs/2109.05298v1 | https://arxiv.org/pdf/2109.05298v1.pdf | InDuDoNet: An Interpretable Dual Domain Network for CT Metal Artifact Reduction | For the task of metal artifact reduction (MAR), although deep learning (DL)-based methods have achieved promising performances, most of them suffer from two problems: 1) the CT imaging geometry constraint is not fully embedded into the network during training, leaving room for further performance improvement; 2) the model interpretability is lack of sufficient consideration. Against these issues, we propose a novel interpretable dual domain network, termed as InDuDoNet, which combines the advantages of model-driven and data-driven methodologies. Specifically, we build a joint spatial and Radon domain reconstruction model and utilize the proximal gradient technique to design an iterative algorithm for solving it. The optimization algorithm only consists of simple computational operators, which facilitate us to correspondingly unfold iterative steps into network modules and thus improve the interpretablility of the framework. Extensive experiments on synthesized and clinical data show the superiority of our InDuDoNet. Code is available in \url{https://github.com/hongwang01/InDuDoNet}.%method on the tasks of MAR and downstream multi-class pelvic fracture segmentation. | ['Yefeng Zheng', 'Deyu Meng', 'Kai Ma', 'Jiawei Chen', 'Haimiao Zhang', 'Yuexiang Li', 'Hong Wang'] | 2021-09-11 | null | null | null | null | ['metal-artifact-reduction'] | ['medical'] | [ 8.60652775e-02 3.13232869e-01 -1.18851140e-01 -4.41371143e-01
-8.42418253e-01 -6.96273372e-02 8.82674977e-02 -1.06506862e-01
-2.44295686e-01 6.50847912e-01 1.98529810e-01 -5.79397261e-01
-4.03173387e-01 -7.86899745e-01 -7.26805687e-01 -6.35765970e-01
1.60446674e-01 4.95010704e-01 1.57310888e-01 -1.37742490e-01
-2.26552356e-02 5.04681647e-01 -8.18863750e-01 2.89505035e-01
1.27448082e+00 1.02078283e+00 2.53392190e-01 5.20911254e-02
3.20556127e-02 8.93612206e-01 -3.83160822e-02 -1.08818531e-01
1.88691333e-01 -5.50293267e-01 -9.68551576e-01 1.23935984e-02
-2.98233867e-01 -4.63130087e-01 -4.64106709e-01 9.27946925e-01
6.57942057e-01 -2.40160115e-02 5.90426862e-01 -8.40308905e-01
-5.84059477e-01 7.07513034e-01 -8.47976148e-01 2.54067272e-01
-9.55854133e-02 6.91639185e-02 7.20466495e-01 -1.13829303e+00
4.08106297e-01 9.86185193e-01 9.61799800e-01 5.02593577e-01
-9.87316072e-01 -6.39011383e-01 -5.36724320e-03 1.23324588e-01
-1.30419075e+00 -2.43495792e-01 9.87752259e-01 -3.67350847e-01
2.76026160e-01 3.46713185e-01 6.43051147e-01 7.98654079e-01
1.95712283e-01 1.03029799e+00 1.02758336e+00 -2.76713401e-01
1.85937554e-01 -2.22989708e-01 1.92985445e-01 1.00660133e+00
1.94616228e-01 -6.04057647e-02 -1.49793103e-01 -6.09193742e-02
1.10432196e+00 1.19247831e-01 -4.46738213e-01 -3.47369730e-01
-9.89747703e-01 7.71002114e-01 7.67188072e-01 3.58661294e-01
-4.03238118e-01 2.21358791e-01 3.95884812e-01 -8.12665001e-02
7.14103580e-01 2.55232990e-01 -6.81389943e-02 1.01739079e-01
-8.30220580e-01 1.17823988e-01 3.09223145e-01 6.57705069e-01
5.60484290e-01 -2.02763993e-02 -3.12192570e-02 1.00815105e+00
4.00057465e-01 1.76362231e-01 5.22362828e-01 -8.49020839e-01
4.86434549e-01 7.19189346e-01 -1.67811990e-01 -9.67938364e-01
-6.18136168e-01 -5.99917412e-01 -1.04182410e+00 1.42058671e-01
4.60295260e-01 -2.72040159e-01 -1.06639695e+00 1.38416231e+00
4.04638320e-01 2.06013009e-01 -3.09941709e-01 1.19913280e+00
9.33297992e-01 3.49127412e-01 6.73987046e-02 -8.71884152e-02
1.26139998e+00 -1.00903201e+00 -8.18213582e-01 -3.07648778e-01
8.15035224e-01 -4.88129050e-01 1.12433660e+00 3.10093075e-01
-1.31973779e+00 -3.78553510e-01 -1.17342746e+00 -1.74348816e-01
3.42408158e-02 2.71331400e-01 7.07752466e-01 4.78818834e-01
-7.71954894e-01 6.46470785e-01 -1.28180718e+00 2.42084771e-01
9.61615562e-01 4.45552766e-01 -5.65833859e-02 -1.72551051e-01
-1.08178794e+00 8.15668344e-01 3.48283827e-01 6.77626431e-01
-8.67319584e-01 -6.68896258e-01 -8.61400187e-01 -1.49908304e-01
4.49894965e-01 -8.22586834e-01 1.32665062e+00 -8.05735767e-01
-1.37935734e+00 6.37212038e-01 -9.19906944e-02 -2.13821009e-01
9.55556870e-01 -3.38172913e-01 -4.35542800e-02 3.02403688e-01
2.05613136e-01 2.19740391e-01 4.90661442e-01 -1.35016382e+00
-3.42784077e-01 -5.56484818e-01 2.21993346e-02 1.99209154e-01
-3.59283924e-01 -1.94859996e-01 -6.63477898e-01 -9.30445015e-01
5.69688797e-01 -9.01750088e-01 -5.98731935e-01 2.52538115e-01
-5.11051297e-01 -9.40090492e-02 4.88216609e-01 -8.84108603e-01
1.30568957e+00 -2.18281603e+00 -1.52684553e-02 2.41324350e-01
4.98650998e-01 1.75373033e-01 2.44524091e-01 -8.01054910e-02
-3.60717356e-01 4.37532887e-02 -8.18589926e-01 -3.52652937e-01
-3.15242708e-01 2.03627720e-01 1.49253393e-02 6.86316073e-01
1.80897489e-01 1.01966476e+00 -8.68731558e-01 -7.08182871e-01
2.89385587e-01 3.14317286e-01 -6.07356250e-01 1.08244412e-01
-8.34118649e-02 8.45784307e-01 -9.93528128e-01 5.02846599e-01
9.49371815e-01 -4.73944634e-01 1.55789509e-01 -2.52580941e-01
-7.38532096e-02 1.71407998e-01 -1.16804659e+00 1.97924936e+00
-5.77960491e-01 1.46788001e-01 1.26099244e-01 -1.16331983e+00
7.03079104e-01 3.10380340e-01 8.02204311e-01 -7.44395912e-01
2.62556523e-01 5.91070473e-01 -7.77094066e-02 -7.06531763e-01
7.79738128e-02 -4.83796060e-01 2.41226166e-01 3.55026573e-01
-4.07565445e-01 3.50670032e-02 -1.95780471e-01 -5.49211316e-02
8.97932410e-01 3.05484474e-01 1.01323366e-01 -4.58334059e-01
6.11384571e-01 9.39572304e-02 7.49417186e-01 5.79780161e-01
-4.80799600e-02 9.34492350e-01 4.03188705e-01 -5.14900804e-01
-8.16364527e-01 -1.12967861e+00 -4.51710403e-01 5.78054190e-01
4.09966469e-01 -1.16371714e-01 -8.62709880e-01 -6.23228133e-01
-2.17929393e-01 5.96974969e-01 -7.08579123e-01 -2.22491622e-01
-9.52669978e-01 -1.18082678e+00 4.94907945e-01 8.98677826e-01
6.23422563e-01 -9.39108491e-01 -5.55913210e-01 2.80235171e-01
-4.33543772e-01 -7.75900185e-01 -4.35540259e-01 1.48095861e-01
-1.24585319e+00 -9.88901079e-01 -8.25332999e-01 -7.90267169e-01
9.50004518e-01 6.97001070e-02 8.61434162e-01 4.35017914e-01
-2.33612344e-01 5.67362942e-02 -1.68942764e-01 -3.15972507e-01
-3.25147778e-01 -4.41164486e-02 -2.99769878e-01 -9.42964479e-02
-1.23656541e-01 -6.49323165e-01 -8.94828856e-01 2.62444168e-01
-1.08394623e+00 4.36354846e-01 6.20187044e-01 1.06406760e+00
6.71176076e-01 -5.26103899e-02 5.47751307e-01 -1.13148415e+00
6.09112442e-01 -5.21384239e-01 -2.27025524e-01 1.17623582e-01
-7.05932736e-01 1.07255494e-02 4.91005212e-01 -5.99651411e-02
-1.36403918e+00 6.88029975e-02 -5.80888212e-01 -2.78485507e-01
2.04628706e-02 7.61849344e-01 -1.10920183e-01 8.03788826e-02
5.86744428e-01 2.28972778e-01 2.40254670e-01 -5.81394851e-01
2.17324153e-01 5.36008358e-01 3.95058841e-01 -7.15273023e-01
6.29817903e-01 8.24127853e-01 -6.24769703e-02 -5.08932590e-01
-9.23254728e-01 -1.23957224e-01 -6.02385521e-01 -1.40494570e-01
9.03400540e-01 -9.19980466e-01 -5.14753163e-01 5.24571240e-01
-1.11435223e+00 -3.65667939e-01 -2.53624380e-01 4.81062680e-01
-5.42100251e-01 7.09923744e-01 -9.07140434e-01 -5.14988244e-01
-4.46494669e-01 -1.41878116e+00 8.35189879e-01 1.86529845e-01
1.96471321e-03 -1.03934944e+00 -2.15481684e-01 5.23528755e-01
1.63553789e-01 4.39768136e-01 1.09151375e+00 -4.52057928e-01
-4.73283738e-01 2.78468663e-03 -3.56194317e-01 3.61562490e-01
1.46703064e-01 -5.17031431e-01 -8.58033001e-01 -1.26012176e-01
4.92846400e-01 -1.61722288e-01 7.26579070e-01 7.17828095e-01
1.88111949e+00 -8.19434505e-03 -3.79718155e-01 8.78241539e-01
1.36967969e+00 2.34007329e-01 6.30141854e-01 5.17673910e-01
8.91878366e-01 4.09367621e-01 5.60772598e-01 4.93101150e-01
5.07802844e-01 5.53351045e-01 5.98644316e-01 -6.36108994e-01
-1.43575966e-01 -1.17735900e-01 -1.66179568e-01 8.53206933e-01
-2.64621288e-01 7.81549737e-02 -1.15089524e+00 5.54821372e-01
-2.02392554e+00 -4.34597313e-01 -4.09393728e-01 1.82693291e+00
7.85198450e-01 2.14782983e-01 -1.59091309e-01 7.00835809e-02
6.38343751e-01 6.68773204e-02 -6.74077630e-01 -3.90107892e-02
2.07348138e-01 1.81449071e-01 2.69498587e-01 3.68435323e-01
-9.77320433e-01 4.84437793e-01 5.06724215e+00 1.00142562e+00
-1.00309622e+00 3.54076594e-01 1.15262997e+00 1.81472152e-01
-5.23604751e-01 -9.98034403e-02 -2.59533584e-01 4.55734968e-01
4.95607525e-01 4.20486294e-02 1.97544575e-01 6.88607872e-01
6.49200857e-01 -4.49608713e-02 -9.18499172e-01 8.12141120e-01
-1.77834436e-01 -1.36095834e+00 -1.90767050e-01 -1.48825697e-03
4.57554728e-01 -2.14940379e-03 -2.46024076e-02 1.77398831e-01
-1.07773319e-01 -1.01074290e+00 7.72185802e-01 4.13229406e-01
6.47085428e-01 -6.07261717e-01 6.28897071e-01 3.54681939e-01
-1.01039994e+00 -4.97387014e-02 -1.71689689e-01 1.10609904e-01
2.34639242e-01 7.52776682e-01 -5.91965735e-01 9.15411532e-01
7.42333710e-01 6.79074883e-01 -3.86655003e-01 1.06406939e+00
-2.28235275e-01 5.96223474e-01 -2.69100249e-01 4.10448402e-01
2.02203438e-01 -3.94054085e-01 5.28050005e-01 1.02343202e+00
2.68221110e-01 3.21573377e-01 5.21843880e-02 1.14181340e+00
-1.04885185e-02 1.62167335e-03 -3.14358920e-01 3.93261284e-01
1.68900311e-01 1.11383867e+00 -9.68291283e-01 -1.88074440e-01
-3.66586357e-01 7.13052511e-01 2.21893206e-01 2.35198155e-01
-1.17518640e+00 -1.27374202e-01 1.75133362e-01 3.39813232e-01
1.07690498e-01 -2.30680645e-01 -1.12758231e+00 -1.06181586e+00
3.38804930e-01 -7.55088925e-01 5.38719118e-01 -5.55586576e-01
-1.09391248e+00 6.51570857e-01 6.51824325e-02 -1.32732344e+00
1.39144048e-01 -5.47252357e-01 -6.77723348e-01 8.10169399e-01
-1.63012421e+00 -1.12519407e+00 -4.33595508e-01 4.27349001e-01
6.31724715e-01 2.59490162e-01 4.40494865e-01 6.42137587e-01
-7.78264284e-01 4.16907281e-01 2.32015848e-01 2.62477398e-01
3.14265162e-01 -1.17481768e+00 4.53917608e-02 8.48567426e-01
-4.05647308e-01 5.27568996e-01 4.59949881e-01 -6.53935671e-01
-1.24900937e+00 -1.09284842e+00 4.31466073e-01 -2.65278339e-01
5.07696033e-01 -1.00291088e-01 -9.47889924e-01 6.48160696e-01
-5.82805276e-02 2.27981269e-01 5.99250913e-01 -9.42673087e-02
2.99085975e-01 4.23196563e-03 -9.91633892e-01 6.29609764e-01
1.10499465e+00 -1.33337051e-01 -5.39764822e-01 4.40119117e-01
5.83949864e-01 -8.57115090e-01 -9.41453099e-01 7.52213597e-01
2.77079552e-01 -1.03860247e+00 1.02207732e+00 -2.68935591e-01
7.50935495e-01 -2.64773607e-01 3.12379390e-01 -9.86026585e-01
-2.63793737e-01 -2.63973475e-01 -7.87380189e-02 9.11046326e-01
3.68596733e-01 -7.31153011e-01 8.47118795e-01 7.51584470e-01
-6.72039807e-01 -1.29023850e+00 -9.98506606e-01 -5.20678699e-01
3.31013381e-01 -5.64738452e-01 5.05707145e-01 9.49163079e-01
-9.19505209e-02 5.83939664e-02 -2.16452375e-01 2.97956198e-01
5.19870877e-01 1.75403789e-01 2.89778739e-01 -9.34738815e-01
-2.90595859e-01 -3.49934638e-01 2.67776251e-01 -1.19584310e+00
-1.91651344e-01 -9.99293268e-01 2.31430620e-01 -1.80091214e+00
1.74902439e-01 -7.14532495e-01 -3.45844924e-01 7.24280894e-01
-3.35437626e-01 1.94625899e-01 -3.87816466e-02 3.88968587e-01
-3.01035494e-01 8.76947045e-01 1.70543480e+00 1.18834069e-02
-1.32337660e-01 1.22142725e-01 -8.29504728e-01 9.03588474e-01
6.96205258e-01 -5.58916271e-01 -4.14412260e-01 -8.51904273e-01
-4.80627939e-02 3.78501117e-01 4.90807801e-01 -8.92588496e-01
-1.76766294e-03 -6.49377853e-02 4.04235303e-01 -3.89875382e-01
1.24975882e-01 -8.15322399e-01 1.70050889e-01 7.71298885e-01
-1.12681955e-01 2.92134248e-02 2.53759533e-01 4.37877983e-01
-2.41237760e-01 -3.37745875e-01 9.33243215e-01 -2.70319045e-01
-5.02377868e-01 5.30598164e-01 -1.95304647e-01 7.15210196e-03
7.70503700e-01 -3.14253837e-01 1.17039066e-02 -7.94289410e-02
-8.66806030e-01 3.92339617e-01 1.46189854e-01 1.29965067e-01
7.39183784e-01 -1.28382373e+00 -6.54659569e-01 2.29786653e-02
-1.97817400e-01 7.42894292e-01 5.46418726e-01 1.31504393e+00
-9.82235253e-01 2.24337384e-01 1.12284506e-02 -6.46004915e-01
-6.89784765e-01 2.23174676e-01 6.79457545e-01 -4.64961439e-01
-1.04476774e+00 7.35432923e-01 5.77243388e-01 -5.19800186e-01
-7.47319609e-02 -3.23761791e-01 -1.13703869e-01 -3.25693488e-01
1.84398651e-01 3.57492268e-01 1.83262154e-01 -2.89858997e-01
-4.32721496e-01 3.80931079e-01 -2.01069161e-01 1.94290996e-01
1.76831841e+00 -9.50279608e-02 -1.58144549e-01 5.68835028e-02
9.92696404e-01 -3.31903070e-01 -1.29790342e+00 -1.63294956e-01
4.76373360e-02 -3.73957753e-01 1.49752751e-01 -6.20938241e-01
-1.31891870e+00 8.40001166e-01 8.16935599e-01 -2.08375737e-01
1.34883034e+00 -1.46872416e-01 1.04681408e+00 -4.40973081e-02
8.41129422e-02 -9.84220445e-01 1.43806756e-01 2.97760159e-01
8.63864601e-01 -1.17874670e+00 1.88596517e-01 -6.62441790e-01
-4.52542186e-01 1.15799069e+00 7.22802877e-01 -1.83170721e-01
7.56069839e-01 1.79726183e-01 1.35354787e-01 -4.62114722e-01
-4.85040694e-02 2.05081895e-01 2.62764454e-01 1.43565834e-01
5.15198708e-01 -1.60378411e-01 -7.77109623e-01 7.78190911e-01
1.11768462e-01 2.09711105e-01 3.51948112e-01 9.77518976e-01
-1.57056481e-01 -9.82381880e-01 -2.53425658e-01 3.98218781e-01
-5.10624170e-01 -8.96053687e-02 2.10183531e-01 7.82227695e-01
1.49223834e-01 5.48083246e-01 -3.45748425e-01 -1.08260550e-01
4.32439983e-01 -3.45736772e-01 2.85294741e-01 -5.68066657e-01
-4.17590261e-01 2.49660701e-01 1.66017991e-02 -4.39385712e-01
-3.05899262e-01 -6.57903731e-01 -1.69683564e+00 9.62511003e-02
-3.46531183e-01 7.29275569e-02 4.36400086e-01 1.20491207e+00
1.54417813e-01 8.79553616e-01 5.93365788e-01 -7.16579497e-01
-5.39618254e-01 -7.63161063e-01 -4.99761492e-01 3.48906666e-01
2.19737798e-01 -6.90494120e-01 -1.51056610e-02 -1.14273019e-01] | [13.633004188537598, -2.5477066040039062] |
3b0c1833-8a1c-4e08-a5ea-0bdf74087f01 | towards-realistic-symmetry-based-completion | 2201.01858 | null | https://arxiv.org/abs/2201.01858v1 | https://arxiv.org/pdf/2201.01858v1.pdf | Towards realistic symmetry-based completion of previously unseen point clouds | 3D scanning is a complex multistage process that generates a point cloud of an object typically containing damaged parts due to occlusions, reflections, shadows, scanner motion, specific properties of the object surface, imperfect reconstruction algorithms, etc. Point cloud completion is specifically designed to fill in the missing parts of the object and obtain its high-quality 3D representation. The existing completion approaches perform well on the academic datasets with a predefined set of object classes and very specific types of defects; however, their performance drops significantly in the real-world settings and degrades even further on previously unseen object classes. We propose a novel framework that performs well on symmetric objects, which are ubiquitous in man-made environments. Unlike learning-based approaches, the proposed framework does not require training data and is capable of completing non-critical damages occurring in customer 3D scanning process using e.g. Kinect, time-of-flight, or structured light scanners. With thorough experiments, we demonstrate that the proposed framework achieves state-of-the-art efficiency in point cloud completion of real-world customer scans. We benchmark the framework performance on two types of datasets: properly augmented existing academic dataset and the actual 3D scans of various objects. | ['Mykola Maksymenko', 'Volodymyr Karpiv', 'Vladyslav Selotkin', 'Rostyslav Hryniv', 'Oles Dobosevych', 'Taras Rumezhak'] | 2022-01-05 | null | null | null | null | ['point-cloud-completion'] | ['computer-vision'] | [ 2.94345349e-01 -1.82507262e-01 5.27159035e-01 -4.09143388e-01
-7.54820049e-01 -4.15548682e-01 1.85054809e-01 6.06819652e-02
5.94928209e-03 1.33353204e-01 -4.72574592e-01 -9.84134153e-02
-1.53835848e-01 -8.08720231e-01 -9.53803897e-01 -5.15823245e-01
9.49085429e-02 1.26231503e+00 5.42335093e-01 -5.79937249e-02
6.45756125e-01 9.53720570e-01 -1.80590713e+00 1.20876925e-02
7.23779082e-01 9.48762596e-01 4.99291867e-01 3.45017195e-01
-3.99992853e-01 -1.90525189e-01 -3.90241265e-01 -2.51862854e-01
5.13172984e-01 4.48085845e-01 -3.72563452e-01 7.38908529e-01
3.63648117e-01 -3.06229383e-01 -5.98840490e-02 8.16268146e-01
3.94041568e-01 -7.70583153e-02 5.47609985e-01 -1.18201041e+00
-3.99846256e-01 -4.75115359e-01 -1.01043522e+00 -2.18049437e-01
5.97652674e-01 1.94504395e-01 3.91490519e-01 -1.60016203e+00
4.41051781e-01 1.32371724e+00 6.99092925e-01 3.42298716e-01
-8.43976200e-01 -6.56300962e-01 -1.10115848e-01 7.78993964e-02
-1.22523880e+00 -1.73010036e-01 9.81543481e-01 -3.32546979e-01
5.71509302e-01 1.68554291e-01 4.37791765e-01 6.47027433e-01
7.07608312e-02 6.38619065e-01 9.78104115e-01 -2.97465354e-01
3.54514867e-01 -9.27845910e-02 8.12390149e-02 5.07384181e-01
3.39536995e-01 4.94444510e-03 -4.61519033e-01 -4.32418436e-01
9.10990536e-01 6.16668940e-01 -1.70060620e-01 -8.21870208e-01
-1.09767246e+00 3.83954763e-01 7.12110028e-02 -2.69613981e-01
-6.96288347e-01 -2.14159712e-01 -1.29929753e-02 -1.65427593e-03
4.30731803e-01 -3.70747715e-01 -5.01614213e-01 6.96812570e-02
-8.52630258e-01 2.53698021e-01 5.02019644e-01 1.46068525e+00
9.61280942e-01 -1.19436249e-01 4.06011850e-01 7.97882020e-01
3.86285126e-01 7.03535736e-01 -4.99943160e-02 -7.23138392e-01
8.28375101e-01 7.27298081e-01 5.02416253e-01 -6.94705844e-01
-2.33053714e-01 -3.51484686e-01 -5.59176147e-01 4.56157029e-01
8.66537392e-02 3.72707397e-01 -1.14059794e+00 8.70587409e-01
1.03942740e+00 3.92531604e-01 -2.40401968e-01 1.01837802e+00
8.27666521e-01 4.32380676e-01 -3.97091717e-01 -2.50237375e-01
1.13509274e+00 -6.82353854e-01 -5.83651662e-01 -9.56156850e-02
-2.69511063e-03 -1.00635004e+00 1.29376328e+00 7.80824959e-01
-1.43567431e+00 -5.44674575e-01 -8.32208693e-01 1.52118906e-01
1.70198992e-01 -2.42374793e-01 4.42588508e-01 4.74688530e-01
-5.65571964e-01 5.71622550e-01 -9.02756035e-01 -2.38349885e-01
5.92282355e-01 4.40924942e-01 -4.57418621e-01 -8.05177987e-01
-3.38657498e-01 6.50294125e-01 -9.39624757e-02 3.13307345e-01
-1.09720099e+00 -8.96581233e-01 -3.60200971e-01 -2.73493975e-01
4.68997777e-01 -6.07074380e-01 1.01856065e+00 -1.90479532e-01
-1.16798532e+00 1.01869702e+00 -1.34654030e-01 2.78182119e-01
6.09939337e-01 -5.40295482e-01 -2.41221696e-01 2.01084971e-01
-3.97009822e-03 5.87299019e-02 1.08712888e+00 -1.74152589e+00
-4.09843057e-01 -1.02737248e+00 -9.59176719e-02 2.81436920e-01
1.11063465e-01 9.90864355e-03 -6.55438840e-01 -1.17140353e-01
1.01320136e+00 -9.01567161e-01 -2.23854735e-01 4.19985026e-01
-3.56940418e-01 -1.49607867e-01 1.36115932e+00 -3.81776124e-01
2.01588944e-01 -2.11799574e+00 -2.89617538e-01 1.75012961e-01
-8.01163688e-02 2.53037959e-01 -5.49686588e-02 7.16351151e-01
3.45659838e-03 1.63270049e-02 -3.24953437e-01 -7.09885240e-01
-2.26815164e-01 5.25562763e-01 -1.67043805e-01 7.10333943e-01
3.50524560e-02 3.64609718e-01 -6.25727236e-01 -6.19430780e-01
4.58725899e-01 6.14892066e-01 -4.00378972e-01 2.89968908e-01
-5.76053858e-02 6.49227858e-01 -6.16072536e-01 1.46138465e+00
1.37882698e+00 4.31625955e-02 -4.69971031e-01 -1.81849822e-01
7.23456442e-02 -2.26248816e-01 -1.65237176e+00 2.07133341e+00
-4.23835397e-01 -1.52472094e-01 5.64044595e-01 -9.06535864e-01
1.18693113e+00 3.44490290e-01 7.68795490e-01 -3.50270599e-01
-2.10035771e-01 2.84173936e-01 -5.81985831e-01 -8.78100514e-01
3.89705926e-01 -3.15677941e-01 2.82531232e-01 5.53521276e-01
-4.14935946e-01 -5.59121430e-01 -5.02418160e-01 6.72469437e-02
1.34352231e+00 1.41294077e-01 -2.34904438e-01 1.49205372e-01
3.36098820e-01 1.55408800e-01 5.45288682e-01 4.53161985e-01
-3.61872651e-02 1.25977314e+00 -2.65986472e-01 -4.54198480e-01
-1.15303898e+00 -1.48767734e+00 -3.82783145e-01 3.47476900e-01
5.94493151e-01 8.85061324e-02 -3.63890558e-01 -4.26003188e-01
2.38236979e-01 5.22415221e-01 -2.80316383e-01 1.08086519e-01
-7.14404225e-01 -5.52855074e-01 -2.09809989e-02 2.82427102e-01
4.85356659e-01 -9.79756117e-01 -6.51994050e-01 2.09749997e-01
-7.47886598e-02 -1.37834513e+00 -9.68597904e-02 -2.95684308e-01
-1.58661377e+00 -1.40447509e+00 -4.20239419e-01 -7.20992506e-01
1.09867632e+00 8.47722948e-01 1.26212466e+00 4.73454565e-01
-4.55617845e-01 9.04510975e-01 -4.57216501e-01 -6.43819869e-01
-3.20079215e-02 -5.90107262e-01 1.20283753e-01 8.18552449e-02
2.61432528e-01 -8.52239490e-01 -8.00235748e-01 6.59318745e-01
-1.18410444e+00 -2.63566047e-01 5.06070554e-01 5.20620883e-01
9.36476111e-01 5.16319633e-01 1.50672629e-01 -7.50828803e-01
2.38752559e-01 -4.96033639e-01 -5.37743986e-01 3.62777128e-03
-4.27703798e-01 -4.81133878e-01 1.07752942e-01 -4.54692513e-01
-1.06125891e+00 2.73105800e-01 -2.51492262e-02 -8.92381370e-01
-3.58847886e-01 6.35498911e-02 -3.06695372e-01 -1.59736171e-01
4.44421262e-01 3.02891463e-01 -8.00449401e-02 -1.04752123e+00
-1.03623003e-01 5.39735436e-01 8.23579431e-01 -6.58637226e-01
1.45325673e+00 1.05993330e+00 2.69362301e-01 -8.22564363e-01
-4.21112299e-01 -8.23269367e-01 -9.84259963e-01 -2.41434202e-01
3.15398633e-01 -7.96919227e-01 -6.11223519e-01 4.46177840e-01
-1.39120579e+00 3.23047459e-01 -2.61295915e-01 3.59625280e-01
-5.13806939e-01 5.97815335e-01 -3.52919072e-01 -9.85057831e-01
-4.54177648e-01 -1.02447200e+00 1.60855281e+00 -3.66179012e-02
5.55073380e-01 -4.93383110e-01 -1.40787214e-01 7.23978460e-01
4.41015437e-02 5.72657049e-01 9.47473645e-01 -1.66938528e-01
-1.01509225e+00 -5.51836908e-01 6.92328960e-02 2.79752493e-01
1.81406051e-01 -1.00489501e-02 -8.40216994e-01 -3.70858908e-01
5.21394432e-01 -1.29746586e-01 2.90952653e-01 6.36047795e-02
1.24932444e+00 1.22000888e-01 -3.94717932e-01 4.34660941e-01
1.78864563e+00 6.21127822e-02 8.43117356e-01 1.22588709e-01
5.69236755e-01 7.87514448e-01 1.32081437e+00 7.73434997e-01
2.60015875e-01 5.47747076e-01 1.28964651e+00 -1.61845796e-02
1.71758041e-01 -8.97790492e-02 -2.90693864e-02 6.46227360e-01
-2.77708679e-01 9.02076885e-02 -1.06324267e+00 5.94605923e-01
-1.54160380e+00 -6.41400397e-01 -8.59058976e-01 2.43652129e+00
3.03507775e-01 1.78007305e-01 -2.42700025e-01 5.38025439e-01
8.85220885e-01 -4.35944319e-01 -7.40795851e-01 1.78584289e-02
1.70724988e-01 4.87802863e-01 3.05347264e-01 1.22511573e-01
-4.45094228e-01 5.20842791e-01 5.74976110e+00 3.82365197e-01
-6.97012722e-01 3.90709043e-01 -1.43067196e-01 -1.51512966e-01
-3.82773906e-01 3.35474797e-02 -5.62672019e-01 2.77193367e-01
2.34732822e-01 3.56336832e-01 1.52993158e-01 8.50510478e-01
2.83700585e-01 -2.91872025e-01 -1.13907766e+00 1.38531613e+00
3.94056290e-02 -9.28665519e-01 -2.11247936e-01 1.42737821e-01
5.78892469e-01 5.23026139e-02 -6.76150545e-02 3.65934223e-02
-4.64594364e-02 -8.04990351e-01 7.01703370e-01 4.49869275e-01
6.40536666e-01 -4.75270867e-01 6.03068948e-01 9.21885073e-01
-8.95394623e-01 5.91616780e-02 -6.35712147e-01 1.27380937e-01
4.39557701e-01 1.10024142e+00 -1.01540124e+00 5.78766704e-01
9.65634286e-01 1.20989941e-01 -3.83324862e-01 1.37593246e+00
3.92771885e-03 4.54328477e-01 -6.02915525e-01 5.34846902e-01
-3.59116420e-02 -2.82491177e-01 7.64377534e-01 5.53531170e-01
4.84568059e-01 5.63036859e-01 1.25903904e-01 8.36935282e-01
1.65127605e-01 -1.18457727e-01 -6.16107166e-01 5.92328966e-01
5.28831244e-01 1.24108493e+00 -5.60789883e-01 -5.85205480e-02
-4.84024405e-01 8.82267118e-01 -1.61825549e-02 3.03073198e-01
-7.85920799e-01 1.75383747e-01 3.68842661e-01 7.74005413e-01
3.25575382e-01 -5.88465750e-01 -5.23524225e-01 -1.07356822e+00
6.51077211e-01 -5.31758010e-01 2.04539690e-02 -1.18681502e+00
-1.26646423e+00 1.90294474e-01 2.56616645e-03 -1.60832727e+00
3.24457288e-01 -2.99205393e-01 -7.64982343e-01 7.71230340e-01
-1.57408381e+00 -1.17597222e+00 -8.99567902e-01 9.90479231e-01
9.02077615e-01 1.16634741e-01 5.64865649e-01 5.07829428e-01
-1.59666091e-01 -2.11726710e-01 1.19733252e-01 -4.32646304e-01
4.44880486e-01 -7.77074993e-01 4.01126057e-01 6.22946024e-01
-1.64311588e-01 3.07680696e-01 7.79125571e-01 -9.08778787e-01
-2.05467272e+00 -9.08559442e-01 3.89529854e-01 -5.65947652e-01
-4.31612469e-02 -5.95890403e-01 -1.01014745e+00 5.87684870e-01
-4.52391863e-01 5.25190175e-01 2.22643450e-01 -1.59720466e-01
2.51306649e-02 -1.26272544e-01 -1.73212659e+00 -9.17114913e-02
1.35560071e+00 -1.55648202e-01 -7.83443868e-01 7.87213862e-01
4.92249489e-01 -8.67525816e-01 -8.08112919e-01 7.22415090e-01
3.81779104e-01 -1.15744555e+00 1.30298448e+00 -3.90147805e-01
2.78774470e-01 -3.23256880e-01 -3.57189298e-01 -8.97822082e-01
-5.99963814e-02 -3.75179052e-01 -1.93812586e-02 1.11276591e+00
-8.66570622e-02 -4.81168330e-01 1.28718150e+00 8.52567971e-01
-6.27193928e-01 -6.86853707e-01 -1.03248823e+00 -7.28120685e-01
-3.35038126e-01 -7.07414269e-01 8.43044460e-01 7.24185884e-01
-7.69273102e-01 -1.58121064e-01 -6.45615011e-02 8.37511301e-01
1.27434003e+00 3.93651038e-01 1.20691192e+00 -1.69910502e+00
3.02990489e-02 6.01097703e-01 -3.51366222e-01 -9.36806083e-01
-3.22446823e-01 -3.11693817e-01 4.28521410e-02 -1.64506030e+00
-6.17361926e-02 -1.16684306e+00 2.27389678e-01 2.06306994e-01
1.93675473e-01 1.88541502e-01 5.79961129e-02 7.83489704e-01
-3.12789559e-01 5.56086481e-01 1.48580861e+00 -1.79019105e-02
-1.54531345e-01 5.22552133e-01 -1.41091630e-01 7.94227421e-01
4.99048859e-01 -7.63234377e-01 -4.37815517e-01 -7.97608435e-01
3.89242172e-03 3.48362833e-01 5.45565903e-01 -1.03900588e+00
1.77260056e-01 -3.38288575e-01 3.18228990e-01 -1.54349792e+00
8.90107334e-01 -1.47027636e+00 5.67881107e-01 3.08788598e-01
5.53679109e-01 8.89261588e-02 -4.81174849e-02 7.77792871e-01
-2.41661556e-02 -4.97869343e-01 6.77534878e-01 -4.97324526e-01
-2.67169267e-01 6.95026040e-01 4.01500851e-01 -1.58332542e-01
1.12014937e+00 -8.84819329e-01 9.71548855e-02 -1.17874824e-01
-7.06240773e-01 9.96948704e-02 7.82422602e-01 4.27315861e-01
1.20737982e+00 -1.28302813e+00 -7.97070801e-01 3.84488553e-01
9.31631848e-02 9.60415304e-01 3.36819947e-01 5.52069426e-01
-5.31683147e-01 6.55267611e-02 -9.96831581e-02 -1.28264618e+00
-1.31282389e+00 4.89444733e-01 -8.37131143e-02 9.97997373e-02
-9.86036479e-01 4.93180543e-01 1.15939468e-01 -5.67015171e-01
1.94985017e-01 -4.86937851e-01 3.06007951e-01 -5.13372898e-01
1.46877751e-01 5.42794168e-01 6.69900119e-01 -4.98023182e-01
-3.20721775e-01 1.08245838e+00 1.39218017e-01 1.02906592e-01
1.68420112e+00 -2.32720956e-01 2.25210384e-01 1.96298689e-01
8.27600062e-01 -6.92016929e-02 -1.19245172e+00 -4.99645978e-01
-3.65719885e-01 -9.73828197e-01 -1.19125500e-01 -4.21247900e-01
-1.17579710e+00 9.83024955e-01 6.64651632e-01 -6.13601729e-02
1.03060806e+00 1.29035294e-01 1.08760619e+00 7.64709786e-02
1.03144276e+00 -8.69241416e-01 3.61847699e-01 8.72807056e-02
1.15941894e+00 -1.23993957e+00 3.81854892e-01 -8.39006901e-01
-3.39010090e-01 1.04586780e+00 5.47375321e-01 -3.20804715e-01
7.34483957e-01 9.89432931e-02 -2.58784026e-01 -7.54119098e-01
-4.25412714e-01 1.22913539e-01 -1.97879702e-01 8.41667354e-01
-4.37925279e-01 -1.69406548e-01 4.47770208e-02 1.88589931e-01
-7.66016096e-02 -5.09934798e-02 5.82438648e-01 1.34558403e+00
-5.18674791e-01 -9.79027152e-01 -1.16329563e+00 4.04717833e-01
-2.26274744e-01 3.28838468e-01 1.03840820e-01 9.94966567e-01
1.21407598e-01 8.64831984e-01 1.13848813e-01 -8.15379471e-02
8.80584776e-01 -6.66019693e-02 5.86344659e-01 -8.62815261e-01
-3.68052512e-01 5.56202345e-02 -2.73298234e-01 -5.83627284e-01
-4.90374267e-01 -1.03810883e+00 -1.47268558e+00 -1.60629198e-01
-6.32482290e-01 -1.92516461e-01 1.23647285e+00 7.70391524e-01
2.37847343e-01 -4.25411128e-02 1.00177383e+00 -1.16034985e+00
-7.34809875e-01 -8.26107860e-01 -7.96399891e-01 7.03254521e-01
1.83174491e-01 -1.06131244e+00 -3.76886874e-01 1.87020153e-02] | [8.350464820861816, -3.1627163887023926] |
ec5bf21b-16d0-42ac-a6db-bfc96d82ff93 | creating-unbiased-public-benchmark-datasets | 2107.01905 | null | https://arxiv.org/abs/2107.01905v1 | https://arxiv.org/pdf/2107.01905v1.pdf | Creating Unbiased Public Benchmark Datasets with Data Leakage Prevention for Predictive Process Monitoring | Advances in AI, and especially machine learning, are increasingly drawing research interest and efforts towards predictive process monitoring, the subfield of process mining (PM) that concerns predicting next events, process outcomes and remaining execution times. Unfortunately, researchers use a variety of datasets and ways to split them into training and test sets. The documentation of these preprocessing steps is not always complete. Consequently, research results are hard or even impossible to reproduce and to compare between papers. At times, the use of non-public domain knowledge further hampers the fair competition of ideas. Often the training and test sets are not completely separated, a data leakage problem particular to predictive process monitoring. Moreover, test sets usually suffer from bias in terms of both the mix of case durations and the number of running cases. These obstacles pose a challenge to the field's progress. The contribution of this paper is to identify and demonstrate the importance of these obstacles and to propose preprocessing steps to arrive at unbiased benchmark datasets in a principled way, thus creating representative test sets without data leakage with the aim of levelling the playing field, promoting open science and contributing to more rapid progress in predictive process monitoring. | ['Jochen De Weerdt', 'Hans Weytjens'] | 2021-07-05 | null | null | null | null | ['predictive-process-monitoring'] | ['time-series'] | [ 0.5361042 -0.10617581 -0.15949924 -0.1268553 -0.58109355 -0.6818739
0.9207862 0.70629627 -0.16933696 0.6760449 -0.01220533 -0.39388826
-0.41790187 -0.97370553 -0.32986903 -0.64424735 0.09005237 0.7902962
-0.05720818 0.43261412 0.6321238 0.4358001 -1.6562351 0.2243367
0.70037824 0.8610642 -0.1659228 0.38023487 -0.26171887 0.89039093
-0.7013866 -0.43009406 0.34050372 -0.4035894 -0.76110667 0.18850094
-0.10067595 0.24465281 0.10739957 0.7987662 0.32110438 -0.10769612
0.58266777 -1.3857516 -0.21057859 0.55009204 -0.64868915 0.14642102
0.31111217 0.29351318 0.8839121 -0.15289266 0.621831 0.7933723
0.34341472 0.39417967 -1.1932218 -0.65571785 0.08981466 0.25865927
-1.0719715 -0.27646858 0.599599 -0.68395185 0.5699662 0.4118964
0.7343725 1.4659476 0.19414887 0.69393766 1.3631561 -0.3456885
0.727576 0.23980653 0.08667916 0.13754429 0.71745837 -0.01031634
-0.70812595 -0.1732259 0.4968576 0.19384196 -0.256675 -0.51889056
-1.1432843 0.48985073 -0.32836053 0.56448215 -0.4099749 -0.27954707
0.47406974 0.4953754 0.23270826 0.8878703 -0.45375624 -0.6816345
-1.0103204 0.46285748 1.3226571 0.7446631 0.45214748 -0.37538633
-0.2176494 0.33727816 0.04918414 0.2564524 0.4587158 -0.6966542
0.48045567 1.0514958 0.09688295 -0.95886976 -0.12948447 -0.39630193
-0.65302694 0.1122667 0.8677354 -0.0061332 -0.53671503 1.343628
-0.0630146 -0.22033665 -0.11093573 0.67992836 0.21587974 0.57693243
0.33518082 -0.38417473 1.2964009 -0.6397955 -0.65214884 -0.2842925
0.44074768 -0.6204484 0.9778416 0.8978695 -1.1767198 -0.2998268
-0.8672473 0.4519366 -0.4004861 -0.15381704 0.83050525 0.6510256
-0.45896485 0.97130233 -0.99977887 -0.2702074 0.7408902 0.15944691
-0.35337397 -0.2945216 -0.7445253 0.93025225 0.16004129 -0.0622453
-0.71782476 -0.9159183 -0.49069056 0.16092433 0.89045376 -0.4566343
1.1812156 -0.77067375 -1.1605164 0.658149 -0.12637137 -0.38459006
0.92834026 -0.22614425 -0.3819275 -0.29979074 0.05279733 -0.10323439
0.5993387 -1.1979805 -0.8798132 -0.6310205 -0.42363843 -0.06297716
-0.1612775 -0.05399025 -0.4301001 -0.21373445 -0.09231067 -0.5592722
-0.40243837 -0.5085577 -0.29623237 -0.21993868 0.77070713 -0.3986762
1.3327066 -1.9177886 0.02664632 0.28113362 0.3954849 0.12451503
0.31553638 0.6861845 0.05962566 0.30280596 -0.18447445 -0.30760235
0.07353441 0.24317585 -0.42157492 0.3668701 0.4079775 0.575728
-0.9773934 -0.3154017 0.4186561 0.03002205 0.05611651 0.13215767
-0.28256664 0.6135925 -0.5728939 0.76939857 0.191046 -0.11030291
0.25303388 0.33040452 -0.35756803 0.34149477 -1.4078872 1.2181995
-0.35695815 0.5747585 -0.16784996 -0.93021536 1.0758389 0.35956472
0.7209222 -0.6856589 0.17375152 0.14517649 0.0977129 -0.3591315
0.35863405 -0.14226156 -0.03301613 0.7160532 -0.24334174 -0.08102427
0.605523 -0.32336253 1.4832814 0.0214403 0.20854762 -0.14707008
0.38106054 0.10500018 0.7114453 0.54307216 -0.23511809 0.4724416
1.016613 -0.55049163 -1.1267698 -0.95334786 -0.03344245 0.5501736
-0.17856322 -0.66206884 -0.51348263 -0.45884103 -0.03760201 0.6915016
-0.5388974 -0.18737377 -0.28582588 -0.8123039 0.29336512 0.4273506
0.22545876 -1.0687613 -0.94430614 0.34889442 0.15010133 -1.0325837
0.20764804 0.37635475 -0.94112635 -1.4854738 -0.2821745 -0.22678179
0.44136977 -0.118112 1.2992128 -0.29046488 -0.34204352 0.03992097
-0.2805089 -0.9233915 -0.6716733 0.11803287 -0.3046625 -0.05072874
0.7409905 -0.526988 -0.3587715 0.3861983 -0.77773994 -0.0779027
0.89192426 0.26796472 0.48546356 0.4287839 0.2965123 -0.8916035
0.7125194 -0.5188315 -0.5489472 0.24971752 -1.148459 -0.01285892
0.46121874 -0.33566815 -0.84344214 -0.30367932 0.53600854 -0.20373885
-0.5684763 0.39718902 -0.36758992 0.6499736 0.5182326 0.10890809
0.14537357 -0.3635159 -0.11198005 0.5019299 0.21508254 -0.49498641
0.56792146 0.45488513 0.11681388 -0.5401883 -0.5503067 -0.61533505
-0.33875686 -0.1896787 0.5410263 -0.44240093 -0.66117185 0.27736294
-0.6596698 -0.34667557 -0.6716011 0.18914956 -0.49898058 0.15967318
-0.44719183 -0.9568278 -0.31594193 -1.1594087 0.74220544 0.23951846
-0.77594817 -0.88275605 0.04870706 0.62715673 0.38083208 0.44009656
0.7020125 -0.82287246 -0.6217723 -0.6597969 0.07734536 0.21627176
0.28736636 0.33084527 -0.9194119 0.05787823 0.17052504 0.02429753
0.45888397 0.17920819 1.2722303 0.03816041 -0.47741374 -0.06884642
1.2123262 0.25544134 0.66997087 0.5921502 0.44544223 0.9895131
0.5370611 0.43817538 0.14172783 0.40538117 0.17602156 0.16621538
0.3242686 -0.20700398 0.15861079 0.34678867 -0.45611477 -0.14569294
-1.2512999 0.53621525 -1.7051101 -1.2441789 -0.4114015 2.477945
0.74112356 0.43698272 0.07292591 0.85994136 0.5183354 -0.2019238
-0.18116015 -0.411715 0.110264 0.21462257 0.50156724 -0.04568858
-1.0309244 0.34892246 5.591755 0.65540195 -0.945132 -0.12260842
0.6618482 -0.19582914 -0.13998823 0.07862016 -0.62765574 0.5659261
1.3923409 -0.31056678 0.24204864 0.802175 0.44147763 -0.35613546
-1.4731632 0.97331065 -0.22987147 -1.1497338 -0.14798452 0.3666434
0.6006513 -0.18075673 -0.23551391 0.23393561 0.37388012 -1.3079923
0.72712696 0.49706376 0.13759007 -0.7770784 0.82293254 0.30674553
-0.72725 -0.26595768 -0.06148366 -0.4967424 -0.00959998 0.97275954
-0.8164733 0.76187056 0.7180421 0.51740944 -0.40582758 1.1780945
-0.11522304 0.82954806 -0.02586562 -0.13094336 0.04649244 -0.3037143
0.4274668 0.9899482 0.08815088 -0.5051216 0.05472753 0.96400917
0.16750357 0.03763732 -0.59471387 -0.545561 0.36585292 1.2188585
-0.94120073 0.0545623 -0.36062294 0.5627496 -0.02262308 -0.11380596
-0.4816398 0.1741633 0.6702187 0.43844473 -0.03179258 -0.19999623
-0.81033796 -0.73519593 0.1532384 -0.976754 0.40786922 -0.21982227
-1.4343426 0.14392857 -0.05775917 -0.9033701 -0.06795502 -0.4428561
-0.61332566 0.8663961 -1.3179362 -0.7377176 -0.45798635 0.12478501
0.44845062 0.03139132 0.78744495 0.2532134 -0.6711906 0.09646676
-0.01538002 -0.12478741 0.79954237 -1.3500304 0.21460377 0.47537306
0.08619153 0.41894493 0.87358123 -0.57213 -1.560786 -0.9051846
0.7350456 -1.0126481 0.9221423 -0.32370314 -0.844259 0.3473563
-0.13178961 -0.47362936 0.8815829 0.4149766 -0.0810295 -0.18046816
-0.89863634 0.6668116 0.6861896 -0.1994238 -0.6219248 -0.00683261
-0.00611457 -0.09803148 -1.0125073 0.35881168 0.33489656 -0.99539524
0.4127761 -0.41106027 0.7580844 -0.16773012 0.15433946 -0.8979314
0.07243101 -0.59731716 -0.11837334 1.4886497 0.41916087 -0.38206086
0.95300066 0.86948407 0.16921553 -0.939393 -0.6345332 -0.67995685
0.01592456 -0.6712945 0.53240114 1.0056353 -0.02627051 0.39299488
0.07220174 -0.1869439 0.48325756 0.24050552 0.87453806 -1.6877179
-0.17525864 -0.7786805 -0.44403017 -0.14337371 -0.4350598 -0.4911936
-0.27989376 -1.5379809 0.35991177 -0.532616 -0.12714769 0.218534
-0.03221223 -0.22013052 0.07645844 0.46255204 -0.45657644 0.23036224
1.2640284 -0.02888583 -0.2862085 0.28478003 -0.8364221 0.5952362
0.99136496 -0.34052813 -0.52940416 0.12695614 0.30339563 -0.18802561
0.4612096 -1.2189652 0.1585563 -0.22422056 0.40509707 -0.38039792
-0.05021762 -0.9589515 0.51918644 0.44329414 -0.30708006 0.05201494
0.08609462 0.71902937 -0.22067316 -0.25003934 0.29084715 -0.39032525
-0.53501076 0.33700532 -0.3302452 -0.05735386 1.481291 -0.49940482
-0.126559 -0.08791105 -0.47006765 0.31074348 0.60678285 0.48015392
-0.11030603 -0.5646266 -0.53646183 0.07427777 0.20947504 0.40766972
0.2106769 0.980383 -0.3047918 0.27708814 -0.17628284 -0.4056028
-1.0953649 0.6764883 0.01143285 -0.74754536 -0.7094207 0.28291065
-0.3452574 -0.0425255 0.30941936 -0.23476635 -0.10089824 0.2641252
0.71205 0.582741 0.31115863 0.20588242 -0.12361986 -0.06150169
-0.0380047 0.01661834 1.4553477 0.12956122 -0.19220017 0.80898917
0.463744 0.15274218 -1.2299742 -0.00758937 0.5263416 -0.45726264
-0.17894591 -0.8986823 -0.7921295 0.6596293 0.26508123 0.58722746
0.9846036 0.07036423 0.23571633 0.07991585 0.39816853 -1.2778263
-0.0475622 0.11797921 0.5675527 -1.0986297 0.05360699 -0.43410647
-0.6153988 0.9430509 0.44970948 0.27039647 0.3179615 0.5775375
-0.01545374 -0.32776177 -0.932707 0.09951065 -0.18037438 0.524871
0.4797069 0.17749155 -0.46061513 0.6667419 -0.36445868 0.23943368
0.47534022 1.202609 -0.1350251 -1.4482816 -0.44433358 0.78836876
-0.6693356 0.26071855 -0.53977025 0.926831 0.18410571 1.0548899
0.19138618 -0.14458719 0.56655324 0.25178567 0.38805288 -0.5152101
-0.48576498 -0.21340491 0.19057216 -0.42288628 -0.18576567 -0.9735219
-0.7613135 -0.72057617 -0.25337905 0.2523095 0.8029239 0.97634476
0.34746325 0.7073577 0.244379 -0.48423877 -0.72336435 -0.95023644
-0.43408296 0.55295527 -0.37062317 -0.72554415 -0.01041697 -0.03799322] | [8.653228759765625, 6.049703121185303] |
2ca97b8e-f288-4c69-9fde-857d6ccb8461 | texttovec-deep-contextualized-neural | 1810.03947 | null | http://arxiv.org/abs/1810.03947v4 | http://arxiv.org/pdf/1810.03947v4.pdf | textTOvec: Deep Contextualized Neural Autoregressive Topic Models of Language with Distributed Compositional Prior | We address two challenges of probabilistic topic modelling in order to better
estimate the probability of a word in a given context, i.e., P(word|context):
(1) No Language Structure in Context: Probabilistic topic models ignore word
order by summarizing a given context as a "bag-of-word" and consequently the
semantics of words in the context is lost. The LSTM-LM learns a vector-space
representation of each word by accounting for word order in local collocation
patterns and models complex characteristics of language (e.g., syntax and
semantics), while the TM simultaneously learns a latent representation from the
entire document and discovers the underlying thematic structure. We unite two
complementary paradigms of learning the meaning of word occurrences by
combining a TM (e.g., DocNADE) and a LM in a unified probabilistic framework,
named as ctx-DocNADE. (2) Limited Context and/or Smaller training corpus of
documents: In settings with a small number of word occurrences (i.e., lack of
context) in short text or data sparsity in a corpus of few documents, the
application of TMs is challenging. We address this challenge by incorporating
external knowledge into neural autoregressive topic models via a language
modelling approach: we use word embeddings as input of a LSTM-LM with the aim
to improve the word-topic mapping on a smaller and/or short-text corpus. The
proposed DocNADE extension is named as ctx-DocNADEe.
We present novel neural autoregressive topic model variants coupled with
neural LMs and embeddings priors that consistently outperform state-of-the-art
generative TMs in terms of generalization (perplexity), interpretability (topic
coherence) and applicability (retrieval and classification) over 6 long-text
and 8 short-text datasets from diverse domains. | ['Hinrich Schütze', 'Yatin Chaudhary', 'Florian Buettner', 'Pankaj Gupta'] | 2018-10-09 | texttovec-deep-contextualized-neural-1 | https://openreview.net/forum?id=rkgoyn09KQ | https://openreview.net/pdf?id=rkgoyn09KQ | iclr-2019-5 | ['information-extraction'] | ['natural-language-processing'] | [ 2.48458058e-01 2.59737939e-01 -3.21987033e-01 -3.82712245e-01
-6.90833449e-01 -3.85638744e-01 1.00131214e+00 1.18276052e-01
-5.21079957e-01 4.81118262e-01 5.54720879e-01 -3.22545975e-01
-2.86296278e-01 -8.98874283e-01 -7.41933346e-01 -9.23447728e-01
1.41894117e-01 7.91528821e-01 1.70737244e-02 2.59950086e-02
1.09056279e-01 -1.23957828e-01 -1.31215727e+00 2.07127661e-01
8.40548635e-01 6.56349659e-01 8.80379379e-01 2.35457599e-01
-5.85583389e-01 2.63515830e-01 -5.65011621e-01 -7.44071677e-02
-3.05677980e-01 3.49023975e-02 -6.58461154e-01 1.58770204e-01
1.92582995e-01 4.97312918e-02 -2.14813724e-01 6.29892170e-01
1.30011976e-01 4.37497079e-01 9.99081790e-01 -8.99896026e-01
-8.21701467e-01 9.60815907e-01 -3.59868526e-01 6.91891462e-02
-1.85282886e-01 -3.78065586e-01 1.30019176e+00 -1.19713795e+00
5.27665675e-01 1.50881624e+00 4.03541237e-01 2.96039492e-01
-1.53035903e+00 -4.62238491e-01 8.32373559e-01 2.07378026e-02
-1.53483665e+00 8.12295731e-03 7.53674150e-01 -4.62410808e-01
1.26619613e+00 -1.14718102e-01 4.08992738e-01 1.82913935e+00
4.99568969e-01 7.10954547e-01 8.05361509e-01 -5.32515168e-01
4.88223195e-01 3.61245096e-01 3.90642226e-01 3.03980876e-02
1.92908496e-01 -3.12509954e-01 -7.39201367e-01 -3.82059246e-01
4.35959369e-01 1.80271760e-01 -2.75021009e-02 -3.48697543e-01
-1.10101628e+00 1.33884108e+00 4.96418253e-02 4.82775986e-01
-6.45106971e-01 2.51368344e-01 3.04899693e-01 -1.79897681e-01
9.43701208e-01 1.48631305e-01 -6.29380107e-01 -1.81225538e-02
-1.11597705e+00 3.97223800e-01 8.29609931e-01 8.80631089e-01
8.80261660e-01 2.16859266e-01 -8.46098587e-02 1.13241124e+00
5.61551630e-01 6.83822870e-01 1.04809308e+00 -2.73745507e-01
4.47357744e-01 1.09344274e-01 -2.42974702e-02 -8.18168998e-01
-2.31565699e-01 -4.96103138e-01 -5.90222836e-01 -4.18214858e-01
-2.71282285e-01 -2.49909475e-01 -1.21637976e+00 2.23325944e+00
1.91015229e-01 4.30609435e-01 2.14567497e-01 3.67889047e-01
6.65858209e-01 1.26835036e+00 5.70957720e-01 -1.34783924e-01
1.67814207e+00 -6.89939439e-01 -7.85684884e-01 -8.04816604e-01
5.63507259e-01 -6.17513895e-01 1.13639200e+00 3.07432592e-01
-6.47054017e-01 -4.66015548e-01 -7.21830547e-01 -9.68268439e-02
-8.89084876e-01 2.63534009e-01 3.81631762e-01 4.82597172e-01
-9.31294978e-01 1.25938863e-01 -8.70791674e-01 -5.05089402e-01
1.05341785e-01 2.93168962e-01 -1.59475580e-01 -2.25268662e-01
-1.57323694e+00 9.10242200e-01 9.51542020e-01 -1.23534724e-01
-9.18236494e-01 -9.70659316e-01 -1.05287611e+00 3.97493601e-01
4.05735880e-01 -7.64759362e-01 9.33163226e-01 -5.09831250e-01
-1.28077865e+00 3.68735731e-01 -5.75986207e-01 -5.54635406e-01
-2.29293317e-01 -6.53692663e-01 -2.88165748e-01 -2.96432078e-01
2.00259909e-01 7.90891409e-01 1.02877760e+00 -1.27272522e+00
-5.94241083e-01 -2.71750987e-01 -3.37607473e-01 3.81251097e-01
-7.17055798e-01 -1.60786167e-01 -3.00438046e-01 -9.63400662e-01
1.43821716e-01 -9.02946115e-01 2.54756976e-02 -7.74507642e-01
-3.67024541e-01 -7.39513040e-01 1.22713304e+00 -5.73002458e-01
1.19424760e+00 -2.12305713e+00 2.89613217e-01 1.21632338e-01
5.04277050e-02 7.95210525e-02 -1.96776599e-01 7.09539175e-01
-1.15936711e-01 1.11654773e-01 -2.69948214e-01 -9.68674183e-01
2.10875317e-01 7.22234428e-01 -1.13220823e+00 1.77200362e-01
3.26768160e-01 7.76022375e-01 -7.65658975e-01 -4.11627322e-01
1.84427217e-01 9.75631535e-01 -4.37994748e-01 1.57438472e-01
-5.38244426e-01 -9.77759510e-02 -4.29027110e-01 1.29138276e-01
4.41939771e-01 -1.81898132e-01 3.47386837e-01 7.92053938e-02
3.00449785e-02 6.72506452e-01 -1.18569279e+00 1.69163823e+00
-1.02778542e+00 6.83766901e-01 -3.39065969e-01 -9.51680481e-01
9.91945744e-01 6.03831470e-01 1.90626085e-01 5.37389629e-02
-1.00074664e-01 -6.84934994e-03 -2.28281274e-01 -1.88943818e-01
8.80813241e-01 -4.85238671e-01 -1.53973669e-01 6.85245633e-01
6.68662071e-01 -2.14991681e-02 -4.56315326e-03 3.01232994e-01
6.51186705e-01 8.02483931e-02 2.64301598e-01 -3.95046890e-01
-1.75775886e-02 -3.80782515e-01 4.41396803e-01 8.81006896e-01
5.23508251e-01 4.64160591e-01 3.34755689e-01 -1.46194950e-01
-9.42997932e-01 -1.21485078e+00 -4.64703619e-01 1.66727901e+00
-3.17260236e-01 -7.20456362e-01 -3.62481892e-01 -3.93719822e-01
-1.35840103e-01 1.32988119e+00 -8.24896455e-01 -1.90377444e-01
-4.49029416e-01 -1.03756881e+00 2.20800906e-01 5.85233152e-01
-1.57432839e-01 -1.24357748e+00 -3.60002816e-01 4.43230093e-01
-2.44961917e-01 -1.29185390e+00 -3.01080346e-01 5.43625236e-01
-9.60195780e-01 -2.54986823e-01 -6.60869122e-01 -6.27817869e-01
3.10050756e-01 4.02261205e-02 1.18236125e+00 -8.09747696e-01
2.19989493e-01 4.77277070e-01 -3.89261752e-01 -6.65887177e-01
-1.27817929e-01 3.15810591e-01 2.57631212e-01 3.97791974e-02
6.57072127e-01 -7.36392677e-01 -2.09034950e-01 2.26036217e-02
-1.28304172e+00 -1.75914183e-01 7.46487379e-01 1.14618385e+00
4.67719048e-01 2.22414181e-01 4.92244393e-01 -1.01889288e+00
8.65205884e-01 -9.61453021e-01 -2.69975007e-01 1.57327116e-01
-6.14554524e-01 1.74163148e-01 2.69791096e-01 -1.03845727e+00
-1.26751065e+00 -3.80952239e-01 8.95237029e-02 -4.07376677e-01
-1.98979065e-01 1.02091074e+00 -1.38428152e-01 8.00706983e-01
4.59760755e-01 6.03754818e-01 -4.70241576e-01 -5.91699421e-01
7.02662289e-01 5.40722907e-01 2.80842304e-01 -7.90469110e-01
4.34753925e-01 3.20306927e-01 -3.25657576e-01 -1.16318488e+00
-8.30539882e-01 -6.23725474e-01 -5.51422179e-01 4.14084107e-01
8.02862525e-01 -1.09317327e+00 6.09201379e-03 8.55082422e-02
-1.47469044e+00 -1.01794168e-01 -3.00026089e-01 7.49186218e-01
-4.23767775e-01 1.68197870e-01 -3.39575917e-01 -9.13457155e-01
-4.57131743e-01 -9.13732290e-01 1.41684449e+00 -1.29099950e-01
-5.07220745e-01 -1.59120166e+00 3.13431114e-01 -2.30706066e-01
5.45403123e-01 -1.73041373e-01 1.30927265e+00 -1.13520634e+00
-1.81370780e-01 -1.59751549e-01 -7.12434277e-02 4.02830809e-01
1.22369088e-01 -2.50138760e-01 -1.21591198e+00 -2.13913903e-01
2.67067790e-01 -8.28689779e-04 1.12716532e+00 8.09615672e-01
8.16647291e-01 -4.51019168e-01 -5.93818307e-01 1.02491900e-01
1.32049394e+00 -1.68491050e-03 3.89472395e-01 7.60683119e-02
5.89152515e-01 7.48097062e-01 3.22355509e-01 4.37400997e-01
3.79909545e-01 6.52703464e-01 9.92072746e-02 1.83649674e-01
1.60681352e-01 -4.94893879e-01 5.96772909e-01 9.56643224e-01
3.71097147e-01 -6.07087433e-01 -1.04810083e+00 1.00480044e+00
-1.73123860e+00 -4.47486639e-01 2.82346278e-01 2.03511405e+00
8.50084066e-01 2.42680833e-01 -2.91638851e-01 -2.35572204e-01
7.04995334e-01 4.32171971e-01 -3.25604141e-01 -3.64867002e-01
-5.43954037e-02 2.06256583e-01 4.69587184e-02 5.59662938e-01
-1.05867577e+00 1.32276487e+00 5.39848280e+00 1.18682492e+00
-1.09770691e+00 4.30259943e-01 3.37250739e-01 7.78747052e-02
-4.94292676e-01 2.12554187e-01 -1.21801329e+00 4.56349522e-01
1.24858391e+00 -7.77359083e-02 -1.51019588e-01 1.05320299e+00
1.48497999e-01 1.37780355e-02 -1.15351009e+00 6.80014312e-01
3.13221604e-01 -1.09234631e+00 4.70927536e-01 2.93637663e-01
6.60525441e-01 2.55544901e-01 4.27073419e-01 6.44358933e-01
3.21931779e-01 -1.06869423e+00 6.10554218e-01 4.47284669e-01
4.81178522e-01 -4.65265840e-01 8.85790944e-01 4.63722110e-01
-8.73606026e-01 6.65475577e-02 -5.70074618e-01 3.03594440e-01
4.02507156e-01 8.53963315e-01 -1.04459107e+00 4.54326451e-01
5.15273988e-01 6.56144083e-01 -1.54818624e-01 5.27134776e-01
-3.08913857e-01 1.04119384e+00 -5.72755396e-01 -1.31904697e-02
5.38733184e-01 -6.33656383e-02 7.95421422e-01 1.55148530e+00
4.53090429e-01 -1.69223472e-01 1.60414055e-01 1.24146736e+00
6.91704859e-04 1.65970802e-01 -4.97494936e-01 -8.90383571e-02
6.05906487e-01 1.04718459e+00 -5.07212818e-01 -4.85259026e-01
-1.83633164e-01 5.65665841e-01 1.57380432e-01 7.43738830e-01
-4.47055936e-01 -9.14487243e-02 6.34129465e-01 -1.49819791e-01
7.80255735e-01 -4.49867994e-01 -2.75838941e-01 -1.01381218e+00
-2.79127099e-02 -3.07807416e-01 3.08029711e-01 -7.27553248e-01
-1.40733039e+00 8.60166728e-01 5.17719984e-01 -7.21616089e-01
-6.49583578e-01 -6.41038537e-01 -8.55174422e-01 1.22558570e+00
-1.45650768e+00 -1.51596379e+00 2.24461257e-01 2.81113088e-01
9.42043245e-01 -2.49747202e-01 1.03876793e+00 -2.76391298e-01
-2.81408936e-01 2.33321115e-01 3.56850415e-01 -3.64045680e-01
7.40874231e-01 -1.37204885e+00 4.14916992e-01 5.36846876e-01
5.39257944e-01 1.07191026e+00 8.25495660e-01 -6.91776752e-01
-1.06838500e+00 -1.29763210e+00 1.34886205e+00 -6.45761549e-01
8.18003774e-01 -1.01054215e+00 -1.20307326e+00 9.31589007e-01
1.70439571e-01 -3.32927316e-01 8.59700739e-01 7.03317404e-01
-3.80261898e-01 2.45234177e-01 -5.97918630e-01 5.50611198e-01
2.31877506e-01 -7.83907473e-01 -1.13907611e+00 4.09817040e-01
1.25937140e+00 -1.61692705e-02 -6.58116877e-01 2.48495862e-03
4.36968654e-01 -2.20192626e-01 9.03764307e-01 -6.21751845e-01
2.85920024e-01 2.13455018e-02 -3.88146996e-01 -1.49312305e+00
-1.96097255e-01 -3.64961267e-01 -3.31232965e-01 1.35231507e+00
5.01302421e-01 -7.00507462e-01 5.70552766e-01 3.15901428e-01
-5.76071963e-02 -8.11661780e-01 -1.02657044e+00 -6.95087135e-01
5.44067800e-01 -9.23787355e-01 4.07190889e-01 8.70928705e-01
-6.67357296e-02 5.44608831e-01 -4.57540333e-01 3.58785391e-01
2.37483814e-01 -6.74393848e-02 3.59590769e-01 -1.24242437e+00
-1.45957649e-01 -1.28553808e-01 -2.76023895e-01 -1.17549253e+00
6.43634617e-01 -6.76921010e-01 1.81330070e-01 -1.47690248e+00
2.19229385e-01 -2.86116987e-01 -4.28694606e-01 5.45985579e-01
-2.28213623e-01 -3.09459269e-01 -9.77304392e-03 1.95636854e-01
-3.45755547e-01 1.05658269e+00 4.07158971e-01 -5.23559749e-02
-2.96246797e-01 -3.08507290e-02 -5.25029659e-01 6.08816147e-01
5.08057714e-01 -8.46033216e-01 -5.18904328e-01 -4.27392066e-01
2.14531764e-01 -3.78382862e-01 3.26898754e-01 -5.97099364e-01
3.01398426e-01 1.94513388e-02 8.52495208e-02 -7.68808424e-01
7.78880119e-01 -6.95420980e-01 -2.15401784e-01 -2.33297907e-02
-4.72480863e-01 -1.41382158e-01 2.89488465e-01 9.65038419e-01
-3.49260569e-01 -4.14191902e-01 1.00806132e-01 -1.43198008e-02
-5.57048321e-01 4.71413247e-02 -7.14040935e-01 -2.07032710e-02
4.68296349e-01 4.60744761e-02 -1.19878493e-01 -4.30680275e-01
-7.83153892e-01 -3.02627217e-02 -1.41112581e-01 7.70579934e-01
5.46025693e-01 -1.18500674e+00 -8.01782489e-01 1.33935317e-01
2.16781273e-01 1.79185107e-01 4.10074979e-01 4.03593898e-01
5.02191365e-01 9.89693761e-01 3.85068655e-01 -7.80868053e-01
-8.76289368e-01 3.42924237e-01 1.11458972e-02 -5.14137864e-01
-6.01324975e-01 8.22234511e-01 1.03762233e+00 -5.42032480e-01
9.99618843e-02 -5.96133053e-01 -3.14453334e-01 2.22708791e-01
2.67030418e-01 -8.02885517e-02 5.93562238e-02 -7.12403893e-01
-5.47049530e-02 4.37074721e-01 -3.98524463e-01 -3.91899288e-01
1.54970169e+00 -2.36054733e-01 -5.50756045e-02 9.80503678e-01
1.08208239e+00 -3.41462255e-01 -1.02174473e+00 -7.55592585e-01
2.18764186e-01 -4.18352634e-02 5.01875937e-01 -5.98943233e-01
-4.21277195e-01 1.05634463e+00 5.11420846e-01 4.16003764e-02
6.51511252e-01 3.77438933e-01 5.90511203e-01 4.53545630e-01
-1.37714460e-01 -1.10752904e+00 1.10717234e-03 8.95723224e-01
9.54004109e-01 -1.00646567e+00 -8.06230307e-02 -2.96407521e-01
-7.11017251e-01 1.03491330e+00 1.46349654e-01 -1.76552124e-02
1.00225711e+00 -6.51389286e-02 -2.88932055e-01 -3.25618058e-01
-1.09213459e+00 4.37306017e-02 6.78488672e-01 5.57767212e-01
4.94859993e-01 1.63161993e-01 -9.03256908e-02 1.03618824e+00
-4.40235436e-01 -4.62276220e-01 2.70694613e-01 7.47529984e-01
-4.02200758e-01 -1.04937971e+00 -3.38947862e-01 3.38506341e-01
-3.76556754e-01 -5.73046863e-01 5.37281483e-03 6.97930932e-01
1.03633389e-01 8.24673891e-01 2.84654856e-01 -4.33915779e-02
-4.19601314e-02 7.05735385e-01 -2.61657119e-01 -1.07124698e+00
-1.63579300e-01 4.28028107e-01 -1.11279674e-01 -1.62430063e-01
-2.36115053e-01 -6.91652060e-01 -7.96834171e-01 3.28437984e-01
-5.62527478e-01 2.91030645e-01 1.17775714e+00 1.40644300e+00
4.52326715e-01 6.42623544e-01 1.96749508e-01 -7.85869420e-01
-4.78931308e-01 -1.59041178e+00 -8.50462377e-01 4.85427678e-02
6.74627051e-02 -8.13027143e-01 -4.17607039e-01 2.50409186e-01] | [10.442458152770996, 6.952220916748047] |
b983090a-3bff-4717-8ad9-609cb0deeb54 | cgpart-a-part-segmentation-dataset-based-on | 2103.14098 | null | https://arxiv.org/abs/2103.14098v2 | https://arxiv.org/pdf/2103.14098v2.pdf | Learning Part Segmentation through Unsupervised Domain Adaptation from Synthetic Vehicles | Part segmentations provide a rich and detailed part-level description of objects. However, their annotation requires an enormous amount of work, which makes it difficult to apply standard deep learning methods. In this paper, we propose the idea of learning part segmentation through unsupervised domain adaptation (UDA) from synthetic data. We first introduce UDA-Part, a comprehensive part segmentation dataset for vehicles that can serve as an adequate benchmark for UDA (https://qliu24.github.io/udapart). In UDA-Part, we label parts on 3D CAD models which enables us to generate a large set of annotated synthetic images. We also annotate parts on a number of real images to provide a real test set. Secondly, to advance the adaptation of part models trained from the synthetic data to the real images, we introduce a new UDA algorithm that leverages the object's spatial structure to guide the adaptation process. Our experimental results on two real test datasets confirm the superiority of our approach over existing works, and demonstrate the promise of learning part segmentation for general objects from synthetic data. We believe our dataset provides a rich testbed to study UDA for part segmentation and will help to significantly push forward research in this area. | ['Alan Yuille', 'Weichao Qiu', 'Jiteng Mu', 'Xiaoding Yuan', 'Qihao Liu', 'Mengqi Guo', 'Zizhang Li', 'Zhishuai Zhang', 'Adam Kortylewski', 'Qing Liu'] | 2021-03-25 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Liu_Learning_Part_Segmentation_Through_Unsupervised_Domain_Adaptation_From_Synthetic_Vehicles_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Liu_Learning_Part_Segmentation_Through_Unsupervised_Domain_Adaptation_From_Synthetic_Vehicles_CVPR_2022_paper.pdf | cvpr-2022-1 | ['geometric-matching'] | ['computer-vision'] | [ 6.19753897e-02 1.50726840e-01 -3.18286389e-01 -5.45429766e-01
-6.87419593e-01 -7.97733843e-01 5.39765596e-01 -4.60644096e-01
5.60719147e-03 5.02994061e-01 -1.96388975e-01 -1.78464383e-01
4.77357388e-01 -8.40604007e-01 -1.16925263e+00 -4.39492971e-01
2.98754960e-01 7.75782645e-01 7.55596638e-01 -1.69927537e-01
-2.13948593e-01 7.10563958e-01 -1.48572266e+00 1.48230761e-01
8.07258904e-01 9.75849390e-01 3.98156226e-01 1.98826537e-01
-2.89027303e-01 3.73585761e-01 -5.09893239e-01 -4.25491512e-01
7.12227881e-01 -4.91166532e-01 -1.03676367e+00 6.07756257e-01
4.05525982e-01 -4.30867583e-01 -2.74571210e-01 8.25237393e-01
1.75607637e-01 -6.38573468e-02 8.26844513e-01 -1.49721134e+00
-3.88050914e-01 5.21306753e-01 -5.58995306e-01 -2.58010685e-01
-4.07246687e-02 3.93588275e-01 6.83478177e-01 -7.13005245e-01
9.43634212e-01 1.03813386e+00 9.69559610e-01 8.20769191e-01
-1.07015538e+00 -6.70756042e-01 9.77463126e-02 -2.85122953e-02
-1.19654167e+00 -3.55522543e-01 1.10792243e+00 -4.74005967e-01
3.76014739e-01 -8.25968534e-02 9.37288761e-01 9.64975059e-01
-2.99123555e-01 1.49039054e+00 9.03375864e-01 -1.54277086e-01
3.95250022e-01 2.81684268e-02 1.77865639e-01 4.67720151e-01
4.27504182e-01 -1.36886507e-01 2.81049639e-01 2.81699300e-01
8.16756248e-01 1.23960905e-01 -6.83054551e-02 -1.05508423e+00
-1.12649918e+00 6.68723702e-01 6.15980685e-01 2.46766210e-02
-4.01824713e-01 1.45655915e-01 3.00425082e-01 4.58780443e-03
3.10263425e-01 1.18875787e-01 -7.19281852e-01 1.78723395e-01
-7.70341277e-01 3.91491979e-01 7.09418833e-01 1.37712038e+00
9.95348454e-01 -1.25322074e-01 -1.62041157e-01 1.07516706e+00
4.50045913e-01 6.29840374e-01 3.04678321e-01 -1.19165325e+00
2.86066025e-01 7.71782696e-01 2.50741899e-01 -1.35545537e-01
-3.67134362e-01 -1.21352203e-01 -6.49281800e-01 2.48557776e-01
6.28293931e-01 -7.55910799e-02 -1.39074981e+00 1.44892764e+00
5.71653724e-01 4.89624813e-02 -6.40548393e-03 9.42636073e-01
7.77883410e-01 4.34761435e-01 3.73750702e-02 4.32571322e-01
1.18901455e+00 -1.37471735e+00 -3.39035273e-01 -3.11210275e-01
6.86929047e-01 -6.45924866e-01 1.15830910e+00 4.75396030e-03
-8.99177492e-01 -8.00515950e-01 -1.02986884e+00 1.16657898e-01
-5.24444342e-01 3.39833230e-01 8.96272540e-01 6.10249162e-01
-7.86615789e-01 3.14296633e-01 -9.94117677e-01 -4.75288898e-01
1.24061120e+00 1.33382514e-01 -2.52724409e-01 -4.13728207e-01
-7.93101370e-01 3.59451771e-01 4.79419917e-01 -8.36285055e-02
-1.02937114e+00 -4.81911749e-01 -1.09285402e+00 -4.69668537e-01
4.21695322e-01 -7.38056719e-01 1.60280967e+00 -8.79405379e-01
-1.25960779e+00 1.10231411e+00 2.08958775e-01 -5.72861016e-01
6.67976797e-01 -7.12702423e-02 -7.69749507e-02 1.31691664e-01
4.48330581e-01 1.24614370e+00 5.87341189e-01 -1.73500490e+00
-4.61144418e-01 -1.77660391e-01 1.20301403e-01 -7.55593255e-02
2.00130582e-01 -4.43766296e-01 -9.22406316e-01 -5.90817928e-01
1.12214215e-01 -1.03760600e+00 -3.59068155e-01 2.54047602e-01
-7.09938705e-01 -1.99206591e-01 1.12922204e+00 -2.21229210e-01
5.19409657e-01 -2.04579020e+00 -3.64214033e-01 -4.69975024e-02
7.19400570e-02 6.03622317e-01 -2.75720567e-01 2.66530216e-01
1.10462636e-01 5.32695502e-02 -8.07169020e-01 -4.43955302e-01
1.69620395e-01 4.56058025e-01 -7.30508417e-02 3.85872126e-01
4.48710322e-01 1.39843845e+00 -6.58407509e-01 -6.52777016e-01
2.63154656e-01 2.51692623e-01 -4.58091140e-01 2.13798031e-01
-7.13782132e-01 4.55894321e-01 -7.07154691e-01 8.97262692e-01
1.09488976e+00 -2.21986309e-01 -3.02279770e-01 -1.58944353e-01
1.10894583e-01 -1.64427720e-02 -9.53476906e-01 1.79401696e+00
-9.57797170e-02 5.79967618e-01 -6.95370734e-02 -1.25877345e+00
7.60844171e-01 -1.81319982e-01 6.89511001e-01 -6.16075516e-01
1.94515154e-01 -2.59790346e-02 -5.01295067e-02 -5.25370181e-01
2.00736254e-01 -7.72165088e-03 -3.19158107e-01 4.84077930e-01
1.27569005e-01 -6.66697085e-01 4.45327222e-01 1.94394514e-01
9.73032594e-01 6.10942125e-01 -1.33546395e-02 -1.10028349e-01
2.87207842e-01 5.66227496e-01 6.07322395e-01 5.75355351e-01
-4.26932067e-01 9.35034156e-01 3.42309833e-01 -5.21485925e-01
-1.26908052e+00 -1.10296392e+00 -3.58063012e-01 5.82801163e-01
5.96382201e-01 1.51245221e-02 -1.21733797e+00 -1.14096391e+00
3.18583921e-02 5.49060643e-01 -7.43470728e-01 -1.63923614e-02
-4.73819762e-01 -5.75916946e-01 5.61859310e-01 9.73613977e-01
1.04736757e+00 -1.23458719e+00 -5.08434415e-01 -7.95201659e-02
-1.49731874e-01 -1.41857314e+00 -5.03873646e-01 3.23929526e-02
-8.25820446e-01 -1.22871172e+00 -9.19227898e-01 -1.07125211e+00
7.09876001e-01 6.11995935e-01 1.34830987e+00 1.19936958e-01
-3.43484581e-01 4.39654440e-01 -4.79794383e-01 -6.34150922e-01
-5.18633842e-01 -8.12661950e-04 -2.12836936e-01 -1.65803745e-01
3.68853450e-01 -3.84161770e-01 -6.56649709e-01 7.81795740e-01
-1.05418277e+00 3.09079617e-01 1.06751132e+00 5.23360491e-01
1.11788917e+00 1.73302982e-02 5.05957425e-01 -1.07279682e+00
1.47875743e-02 -4.55287427e-01 -7.36046016e-01 9.22682285e-02
-1.79933384e-01 -9.96235684e-02 3.91757101e-01 -3.84192377e-01
-9.60001528e-01 4.28909659e-01 -3.10825974e-01 -4.86480623e-01
-6.66406512e-01 3.04837041e-02 -6.87121630e-01 7.95935690e-02
4.36334372e-01 2.98639387e-01 -3.88872251e-02 -5.68756461e-01
5.78039646e-01 5.65943301e-01 6.34803832e-01 -4.97544348e-01
1.01086020e+00 6.83708787e-01 -2.21437961e-01 -6.88490629e-01
-8.17975521e-01 -6.40690506e-01 -1.08364272e+00 -1.46858037e-01
1.01392484e+00 -9.45151865e-01 -1.06333934e-01 6.63669765e-01
-1.00754631e+00 -1.13583112e+00 -6.71854794e-01 1.80523977e-01
-8.60348463e-01 1.91568792e-01 -3.72871399e-01 -3.66868675e-01
-2.00722665e-02 -1.11476946e+00 1.37408340e+00 2.60565102e-01
1.75336286e-01 -8.05635750e-01 -7.78048066e-03 7.17153966e-01
2.36782841e-02 3.95970017e-01 4.73283589e-01 -7.45668292e-01
-8.56333375e-01 -3.43494684e-01 -3.39384586e-01 5.41000307e-01
4.42063212e-02 -6.88477457e-02 -8.93087924e-01 1.26053244e-01
-3.08332831e-01 -5.73243022e-01 8.38076711e-01 3.83099169e-01
1.39584064e+00 5.34283705e-02 -7.57417440e-01 6.01498306e-01
1.14542902e+00 7.48222247e-02 6.72648609e-01 1.93259388e-01
8.52701187e-01 3.56220692e-01 1.01263809e+00 1.43705502e-01
4.92122382e-01 7.40683794e-01 4.27087486e-01 -3.69796038e-01
-7.43908405e-01 -4.94910330e-01 3.46322693e-02 4.56577182e-01
1.38426960e-01 -2.74556756e-01 -1.02278709e+00 8.23183417e-01
-1.59908271e+00 -7.46763110e-01 -4.12370265e-01 1.82655013e+00
6.12568319e-01 2.99628079e-01 4.64873463e-01 -4.85765748e-02
6.06324613e-01 -2.10318610e-01 -6.84203625e-01 1.21691622e-01
-2.23945417e-02 -3.68676223e-02 5.30690253e-01 -6.97938679e-03
-1.46168792e+00 1.21816289e+00 5.88136292e+00 9.44250286e-01
-8.74079108e-01 1.03572115e-01 6.95431590e-01 4.63892251e-01
-2.07481742e-01 -1.41567945e-01 -7.04255342e-01 5.58939576e-01
6.04392767e-01 1.68712944e-01 -3.06075271e-02 1.27124166e+00
4.23313528e-02 -1.00521639e-01 -1.08617687e+00 6.97504222e-01
-3.80800590e-02 -1.27486491e+00 -8.34029317e-02 1.38354465e-01
1.19494247e+00 4.77222174e-01 -3.17002743e-01 5.43532610e-01
4.32892084e-01 -7.42287219e-01 7.75904357e-01 3.17420721e-01
6.31356895e-01 -5.53705573e-01 8.08080554e-01 5.20470083e-01
-1.30200100e+00 2.86535561e-01 -6.39113188e-01 2.68075377e-01
1.49531543e-01 3.60567123e-01 -1.06200480e+00 5.07361650e-01
6.39380753e-01 7.72670567e-01 -8.47863495e-01 1.42659843e+00
-2.83404291e-01 8.31588686e-01 -4.72047955e-01 1.76514253e-01
2.26057351e-01 -3.03363085e-01 1.99979141e-01 1.13180149e+00
-3.18896100e-02 4.42442931e-02 3.09234798e-01 1.10909021e+00
-5.22899449e-01 -1.70085236e-01 -8.17537427e-01 -1.15805648e-01
3.69442731e-01 1.30008054e+00 -1.34617662e+00 -5.43136477e-01
-4.98361319e-01 1.02963638e+00 -6.53729867e-03 3.23405057e-01
-1.02948260e+00 -3.13267261e-01 6.95824981e-01 2.54876435e-01
6.78306997e-01 -1.57150209e-01 -4.05395031e-01 -9.75519538e-01
-6.13590591e-02 -6.01385891e-01 2.33550481e-02 -8.21429491e-01
-1.23594689e+00 6.00448489e-01 3.38909626e-01 -1.74344409e+00
-1.74263656e-01 -6.57343686e-01 -6.48276269e-01 4.20757681e-01
-1.43977773e+00 -1.49659407e+00 -5.93043029e-01 4.00345832e-01
8.90455484e-01 1.61853246e-02 4.01906818e-01 2.03733757e-01
-5.92592120e-01 4.87200826e-01 3.85284331e-03 6.53714418e-01
3.51926476e-01 -1.14620638e+00 9.84114170e-01 7.24600375e-01
4.01013404e-01 7.65735731e-02 3.60833883e-01 -6.48489416e-01
-1.11813998e+00 -1.54394495e+00 7.78724849e-02 -8.00839186e-01
3.49475056e-01 -6.00139260e-01 -8.40681672e-01 8.46907496e-01
-7.08593354e-02 5.28579473e-01 4.48273063e-01 -5.41523159e-01
-8.37615803e-02 2.42470484e-02 -1.25228631e+00 4.91169959e-01
1.32039261e+00 -1.52048841e-01 -5.64065397e-01 3.94613028e-01
8.54542077e-01 -3.94747764e-01 -6.82734787e-01 7.51736403e-01
4.48358059e-01 -9.91863430e-01 9.77462649e-01 -3.35526705e-01
3.31550002e-01 -5.44721127e-01 -9.41537321e-02 -1.12681508e+00
4.09270972e-02 -1.82089016e-01 -3.66315097e-02 1.40838993e+00
2.91261017e-01 -3.97672206e-01 1.15545905e+00 5.09377718e-01
-5.27082980e-01 -8.84827852e-01 -4.16760236e-01 -1.02051580e+00
2.44542196e-01 -7.91086495e-01 8.35257173e-01 6.73946977e-01
-7.83368349e-01 1.12339318e-01 3.46365571e-01 1.72901910e-03
6.28331721e-01 4.79541481e-01 1.28146815e+00 -1.17757165e+00
-7.88249671e-02 -5.11704862e-01 -6.72438622e-01 -1.44873524e+00
2.19005644e-01 -8.92665684e-01 4.06570017e-01 -1.82754707e+00
1.89634740e-01 -6.90138221e-01 1.32643402e-01 6.52622998e-01
3.39548364e-02 9.17548180e-01 3.02534588e-02 3.33629280e-01
-8.56080115e-01 7.41250813e-01 1.61350775e+00 -4.19612825e-01
-1.10728480e-01 5.22693098e-01 -5.17761767e-01 8.63192976e-01
9.56802249e-01 -3.82374138e-01 -4.58307475e-01 -2.67433763e-01
-6.12137616e-01 -3.51031274e-01 4.39526737e-01 -1.21988153e+00
-1.82472706e-01 -5.17134443e-02 5.66418171e-01 -9.00511742e-01
3.08978647e-01 -9.03397083e-01 -6.53868243e-02 2.76250094e-01
1.15948021e-01 -5.15408397e-01 3.98110867e-01 5.52703023e-01
-1.99205086e-01 -1.39669552e-01 8.25278103e-01 -8.97269621e-02
-1.06948566e+00 6.11596107e-01 -8.71122032e-02 2.72319973e-01
1.30767858e+00 -4.78553712e-01 8.63504410e-02 -1.40609995e-01
-5.12797177e-01 4.84451890e-01 1.00383902e+00 3.74931931e-01
4.94837582e-01 -1.44518816e+00 -5.15154839e-01 3.93022060e-01
4.70032603e-01 6.49834692e-01 1.65777698e-01 5.70863605e-01
-7.17330754e-01 4.07677472e-01 -2.43727013e-01 -8.63676548e-01
-8.80042732e-01 6.95413113e-01 3.28912050e-01 2.06971064e-01
-8.77139330e-01 8.71681154e-01 4.77956027e-01 -7.29877234e-01
4.21119705e-02 -5.60777307e-01 1.09839551e-01 -2.71503180e-01
1.81756541e-01 4.33218572e-03 -2.54849214e-02 -9.09839749e-01
-3.09242815e-01 7.45842576e-01 9.73895565e-02 1.14093043e-01
1.21148980e+00 -7.55998120e-02 3.86063248e-01 2.46808156e-01
1.13788342e+00 -1.41952157e-01 -1.99385118e+00 -1.14492930e-01
1.64618231e-02 -2.65371889e-01 -3.88041466e-01 -8.28127563e-01
-1.24759781e+00 8.52793515e-01 4.65499490e-01 1.56059742e-01
9.72714543e-01 5.36627650e-01 9.79517341e-01 4.78505403e-01
5.75904667e-01 -9.59487140e-01 1.55095503e-01 3.13010395e-01
7.65011966e-01 -1.57282984e+00 -2.09295437e-01 -5.99924147e-01
-7.30220020e-01 8.28880727e-01 8.07357907e-01 -1.57547966e-01
6.11554742e-01 2.68937171e-01 4.08360034e-01 -2.04007432e-01
-1.89105749e-01 -7.02108920e-01 2.76520342e-01 1.00197613e+00
3.49581651e-02 4.51554358e-02 -5.06693013e-02 7.46302783e-01
-1.80378139e-01 5.88072054e-02 3.12478662e-01 8.38467717e-01
-2.92392045e-01 -1.38260281e+00 -2.52749056e-01 3.34312916e-01
-6.71800449e-02 4.83547121e-01 -6.06288731e-01 1.28523576e+00
4.81139213e-01 6.08285844e-01 2.27377042e-02 -2.22031608e-01
4.36044276e-01 1.69955343e-01 4.77008730e-01 -7.44134367e-01
2.07485288e-01 -1.43261224e-01 -1.66468367e-01 -5.93124270e-01
-5.15819788e-01 -7.36521065e-01 -1.52129626e+00 2.29178235e-01
-1.07685961e-01 9.03051645e-02 7.19109952e-01 9.87117708e-01
1.62898898e-01 5.37703991e-01 6.09576166e-01 -1.32524443e+00
-1.04566604e-01 -1.03723478e+00 -4.35919464e-01 7.17287064e-01
8.10654834e-02 -8.09828222e-01 1.36659518e-01 3.21896195e-01] | [9.245656967163086, 0.5163228511810303] |
0317fec0-f52d-4ea3-8d4c-aba2761f5a9b | da-gan-instance-level-image-translation-by | 1802.06454 | null | http://arxiv.org/abs/1802.06454v1 | http://arxiv.org/pdf/1802.06454v1.pdf | DA-GAN: Instance-level Image Translation by Deep Attention Generative Adversarial Networks (with Supplementary Materials) | Unsupervised image translation, which aims in translating two independent
sets of images, is challenging in discovering the correct correspondences
without paired data. Existing works build upon Generative Adversarial Network
(GAN) such that the distribution of the translated images are indistinguishable
from the distribution of the target set. However, such set-level constraints
cannot learn the instance-level correspondences (e.g. aligned semantic parts in
object configuration task). This limitation often results in false positives
(e.g. geometric or semantic artifacts), and further leads to mode collapse
problem. To address the above issues, we propose a novel framework for
instance-level image translation by Deep Attention GAN (DA-GAN). Such a design
enables DA-GAN to decompose the task of translating samples from two sets into
translating instances in a highly-structured latent space. Specifically, we
jointly learn a deep attention encoder, and the instancelevel correspondences
could be consequently discovered through attending on the learned instance
pairs. Therefore, the constraints could be exploited on both set-level and
instance-level. Comparisons against several state-ofthe- arts demonstrate the
superiority of our approach, and the broad application capability, e.g, pose
morphing, data augmentation, etc., pushes the margin of domain translation
problem. | ['Jianlong Fu', 'Chang Wen Chen', 'Shuang Ma', 'Tao Mei'] | 2018-02-18 | null | null | null | cvpr-2018 | ['deep-attention', 'deep-attention'] | ['computer-vision', 'natural-language-processing'] | [ 5.50038993e-01 3.11180294e-01 -6.10893182e-02 -2.96934694e-01
-1.03414011e+00 -7.80649602e-01 7.73268580e-01 -6.60334349e-01
1.57839321e-02 8.46848845e-01 -2.23242883e-02 -3.78280580e-02
4.58024703e-02 -9.39617455e-01 -1.24211514e+00 -9.08959210e-01
5.51258683e-01 7.27448404e-01 -2.31715724e-01 -1.18203789e-01
2.73340102e-02 3.28852147e-01 -1.46216595e+00 3.27811360e-01
1.04639661e+00 8.49031985e-01 1.46001831e-01 2.01645672e-01
-2.44823754e-01 3.87349933e-01 -7.16519892e-01 -4.39039707e-01
4.91611898e-01 -9.01642919e-01 -5.99499643e-01 4.68907058e-01
6.24140859e-01 -3.37792516e-01 -1.67176828e-01 1.41237903e+00
2.70676881e-01 7.30960304e-03 7.52320051e-01 -1.50956655e+00
-1.27553999e+00 3.59167278e-01 -6.17007494e-01 -3.98304909e-01
2.37981409e-01 4.00632590e-01 8.47666621e-01 -9.96630311e-01
6.63972616e-01 1.33616567e+00 2.79564947e-01 7.06699431e-01
-1.38900101e+00 -7.95642912e-01 1.63898647e-01 -5.03386296e-02
-1.16689706e+00 -3.16554636e-01 1.13585079e+00 -5.43289185e-01
3.44146162e-01 2.62121141e-01 5.43069780e-01 1.55810595e+00
4.70601097e-02 6.62085474e-01 1.44277704e+00 -2.68624246e-01
2.55100101e-01 1.72584042e-01 -5.79061508e-01 4.60406929e-01
2.16144562e-01 2.43318319e-01 -3.72066945e-01 1.48432583e-01
1.10644841e+00 1.01273216e-01 -2.08838552e-01 -4.60561067e-01
-1.51006532e+00 6.79832697e-01 6.56679451e-01 1.44334182e-01
-2.50426590e-01 1.15090171e-02 7.15965107e-02 3.79633993e-01
3.68164510e-01 6.86741292e-01 -4.22378957e-01 3.49662483e-01
-7.07128108e-01 2.49881685e-01 3.03330898e-01 1.42525804e+00
9.63652134e-01 1.54726654e-01 -3.15374136e-01 4.26965564e-01
2.00023532e-01 7.84667075e-01 4.13664699e-01 -7.29601145e-01
6.66552782e-01 5.40792525e-01 -1.77099195e-03 -9.85456586e-01
1.72652364e-01 -4.91950631e-01 -1.08029544e+00 3.07943523e-01
3.67782235e-01 -2.25131288e-02 -1.17061687e+00 2.18591523e+00
2.96195060e-01 2.23097131e-01 4.57609296e-02 9.98560488e-01
6.21713102e-01 5.47171772e-01 -8.71975571e-02 -1.62623614e-01
1.22135139e+00 -8.70185971e-01 -7.16786265e-01 -3.70381057e-01
8.40217546e-02 -7.51180470e-01 1.41857088e+00 4.09102477e-02
-1.15996981e+00 -8.18894148e-01 -9.11037028e-01 -2.14680787e-02
-2.70226926e-01 1.00098796e-01 4.10964727e-01 2.37509266e-01
-8.75294387e-01 3.32041115e-01 -6.82178915e-01 -1.04332902e-01
6.61469579e-01 3.83917570e-01 -4.85944003e-01 -7.44701475e-02
-1.10285056e+00 5.88106751e-01 3.04949135e-01 1.21013701e-01
-9.89487052e-01 -6.50655746e-01 -8.93726766e-01 2.87960339e-02
3.29306841e-01 -1.00529897e+00 8.17849815e-01 -1.53349125e+00
-1.55849349e+00 9.76104021e-01 4.27493006e-02 -1.63136385e-02
7.22161293e-01 -6.43590912e-02 -1.77035198e-01 -1.20601803e-01
3.86223733e-01 7.79339910e-01 1.27237272e+00 -1.66514790e+00
-3.33841771e-01 -5.13960958e-01 1.51432246e-01 2.04019621e-01
-2.73967654e-01 -3.22489262e-01 -3.39881212e-01 -8.03803325e-01
2.43540272e-01 -9.95509207e-01 -1.24270260e-01 -4.24846588e-03
-6.66505456e-01 2.57883638e-01 7.49368370e-01 -6.03275001e-01
5.92079461e-01 -2.20664048e+00 5.99357784e-01 2.44246349e-02
1.22970864e-01 -5.82209006e-02 -3.27473342e-01 1.19251214e-01
-2.32619435e-01 1.66570306e-01 -4.57980782e-01 -3.25563133e-01
-1.56808756e-02 3.43244791e-01 -6.10390127e-01 3.40730816e-01
6.29698157e-01 1.33289278e+00 -9.67272520e-01 -2.87741810e-01
7.97029287e-02 3.67284447e-01 -6.62592530e-01 5.79173148e-01
-4.88969117e-01 1.24789047e+00 -5.48986852e-01 5.43734908e-01
8.10143054e-01 -2.54701734e-01 4.54022475e-02 -4.51033175e-01
2.58880705e-01 1.19990431e-01 -7.49102116e-01 1.95984435e+00
-5.42903841e-01 4.20397192e-01 -2.59791583e-01 -1.13811994e+00
9.61576939e-01 2.23104998e-01 3.27241659e-01 -6.83933735e-01
3.89163122e-02 2.10691273e-01 1.38475701e-01 -3.98914993e-01
2.58898228e-01 -3.64970714e-01 -1.95277572e-01 4.12597448e-01
1.45193502e-01 -2.13860765e-01 -4.07870322e-01 -1.19531125e-01
6.19353354e-01 5.35617769e-01 7.04837441e-02 -1.89512715e-01
3.72990042e-01 -2.22051628e-02 6.48904741e-01 4.91028309e-01
9.24829766e-02 9.07239318e-01 5.32478392e-01 -2.86179036e-01
-1.23257184e+00 -1.18363416e+00 9.66181010e-02 6.79655373e-01
4.60741967e-01 1.68907538e-01 -9.81480002e-01 -8.97332728e-01
-8.58055800e-02 6.38886929e-01 -8.48710895e-01 -5.34414709e-01
-6.78069234e-01 -4.66650307e-01 3.05628687e-01 5.10837734e-01
6.59528971e-01 -1.15963697e+00 -1.61100253e-01 -4.21296209e-02
-3.31672817e-01 -1.22537255e+00 -6.66268528e-01 -7.89470077e-02
-8.47885966e-01 -8.64337444e-01 -6.15277410e-01 -8.68451118e-01
1.17657208e+00 3.88623029e-02 1.16490078e+00 -1.65186107e-01
-9.17381793e-03 4.95801605e-02 -2.28460968e-01 -2.11226568e-01
-5.87936163e-01 -2.62722634e-02 1.17914721e-01 2.98415631e-01
1.65465102e-01 -8.87800336e-01 -5.74404061e-01 5.02675176e-01
-1.08858287e+00 3.40183496e-01 8.78118753e-01 1.15390980e+00
9.04204130e-01 1.05086556e-02 7.35044241e-01 -1.07313633e+00
3.60644847e-01 -3.75141978e-01 -5.76061964e-01 3.59286159e-01
-3.82621050e-01 1.36793256e-01 8.46243680e-01 -7.00531065e-01
-9.96414363e-01 9.68899429e-02 1.25972018e-01 -9.36669171e-01
-2.97644645e-01 2.14478254e-01 -9.76473808e-01 1.24632090e-01
6.36654258e-01 4.58109468e-01 7.66638592e-02 -3.18238854e-01
5.98470867e-01 3.01807255e-01 7.87266195e-01 -8.89647067e-01
1.33910131e+00 5.04179478e-01 -4.16939706e-02 -1.90300763e-01
-8.50383222e-01 1.45228043e-01 -8.20066869e-01 1.53187841e-01
9.47610319e-01 -8.99666846e-01 -1.57724589e-01 3.73394787e-01
-1.28923213e+00 -2.08514050e-01 -5.97172916e-01 1.79859132e-01
-8.91104877e-01 -3.43956761e-02 -2.53501564e-01 -3.30622315e-01
-4.67520207e-03 -1.42385316e+00 1.40528929e+00 1.87229514e-01
6.50419015e-03 -9.26843882e-01 -1.50422499e-01 3.88221085e-01
2.24611670e-01 6.12368822e-01 9.62596714e-01 -5.48540473e-01
-9.77434576e-01 -1.18804917e-01 -1.45841613e-01 3.74582499e-01
4.98646080e-01 -1.23936161e-01 -9.49793041e-01 -3.59710932e-01
1.83607593e-01 -3.34179640e-01 4.98259038e-01 2.04950422e-01
1.39120758e+00 -6.11332595e-01 -2.02418208e-01 9.48012352e-01
1.31066394e+00 1.13734990e-01 8.96885097e-01 1.56748250e-01
1.09756589e+00 5.51581860e-01 6.07750714e-01 6.35856912e-02
1.11549109e-01 8.13737035e-01 5.43759525e-01 -4.66633528e-01
-2.80464798e-01 -6.43344343e-01 2.55960137e-01 6.28216028e-01
-8.82117730e-03 -3.22645724e-01 -6.00538552e-01 3.84525627e-01
-1.70088828e+00 -7.69376874e-01 3.89475673e-01 2.16194320e+00
9.26512182e-01 -7.06171468e-02 -5.36014214e-02 -2.03454360e-01
9.26149964e-01 -8.35704966e-04 -8.54277134e-01 2.43216716e-02
-1.18663013e-01 2.59234101e-01 3.52021396e-01 3.51845503e-01
-9.51837838e-01 9.70898628e-01 5.34527445e+00 7.90426016e-01
-1.15734971e+00 2.47119874e-01 7.26603806e-01 4.95211296e-02
-8.23681176e-01 8.00909027e-02 -5.07331431e-01 6.22522771e-01
2.93152273e-01 -1.01746343e-01 5.75027049e-01 7.04271376e-01
-1.07132748e-01 5.18388212e-01 -1.44658828e+00 9.59163606e-01
1.53288558e-01 -1.22894239e+00 5.06401420e-01 3.31734180e-01
1.11023450e+00 -4.58054334e-01 5.75206101e-01 3.32143158e-01
2.61852115e-01 -1.16450942e+00 9.06047106e-01 4.47419167e-01
1.39517665e+00 -7.03300774e-01 5.89722157e-01 3.50772768e-01
-8.26275110e-01 2.39660531e-01 -2.66977012e-01 2.05177769e-01
1.37049377e-01 4.10614908e-01 -4.83492494e-01 7.52970278e-01
4.07326847e-01 7.39885330e-01 -2.44473875e-01 3.39315206e-01
-4.81449217e-01 2.76384383e-01 5.42378984e-03 4.32076216e-01
8.92167613e-02 -4.71206337e-01 5.98540068e-01 6.17085040e-01
4.20112669e-01 -1.63474940e-02 1.63904667e-01 1.64190388e+00
-4.74683851e-01 -3.06545615e-01 -7.13910580e-01 -5.64283244e-02
3.92377406e-01 1.10385609e+00 -4.94370937e-01 -3.17216396e-01
-2.48694122e-01 1.31819284e+00 3.21802706e-01 5.10604978e-01
-1.13443232e+00 -1.98924333e-01 8.14747632e-01 1.79636523e-01
1.92905501e-01 2.70549487e-02 -5.86487412e-01 -1.41903591e+00
3.79237264e-01 -1.13839769e+00 -7.28475675e-02 -6.84132934e-01
-1.52070272e+00 7.46492028e-01 -1.37146711e-01 -1.67574739e+00
-2.09806770e-01 -4.12744790e-01 -7.01648414e-01 9.65980172e-01
-1.37166858e+00 -1.51624382e+00 -4.50091928e-01 6.49772286e-01
5.32571912e-01 -2.00877324e-01 7.52911568e-01 3.60157669e-01
-4.80202138e-01 9.73760724e-01 -7.45303258e-02 2.53265500e-01
7.18295038e-01 -1.11907291e+00 5.07350326e-01 8.66553485e-01
2.81856328e-01 6.32560194e-01 5.16387045e-01 -6.00385129e-01
-1.40539694e+00 -1.42375076e+00 3.62678349e-01 -8.19401443e-01
4.32205230e-01 -5.31290829e-01 -1.00132442e+00 8.16560507e-01
7.93524161e-02 1.95359334e-01 4.05416399e-01 -2.77884334e-01
-5.34116030e-01 1.89888272e-02 -1.30775380e+00 6.76945269e-01
1.26290464e+00 -6.81012511e-01 -5.61707497e-01 4.17556554e-01
9.64147747e-01 -6.08592033e-01 -7.61877000e-01 4.37957555e-01
3.01787853e-01 -8.80666673e-01 1.08676684e+00 -9.61318851e-01
9.37765419e-01 -4.61364657e-01 -1.29440844e-01 -1.43499064e+00
-2.94404119e-01 -6.67097807e-01 5.79216592e-02 1.38686419e+00
3.40971231e-01 -7.33902097e-01 6.15444481e-01 5.56681037e-01
-2.34775305e-01 -6.57960594e-01 -9.11743045e-01 -8.92831445e-01
3.89551669e-01 6.17244579e-02 1.08088076e+00 1.18663847e+00
-5.05632699e-01 3.52915555e-01 -5.03767371e-01 3.35426092e-01
6.38211012e-01 4.23717409e-01 9.76895154e-01 -8.09164941e-01
-5.79961419e-01 -4.30719256e-01 -4.80682164e-01 -1.07007635e+00
4.25927937e-01 -9.01112378e-01 1.36085495e-01 -1.26490474e+00
2.29175642e-01 -6.27296209e-01 -1.44815534e-01 4.92759287e-01
-4.07203436e-01 3.79764289e-01 1.16715379e-01 5.17865539e-01
-1.91828266e-01 8.44552636e-01 1.77674162e+00 -3.22818547e-01
3.42858583e-03 -1.33394599e-01 -8.67872238e-01 5.23713350e-01
7.10741699e-01 -4.96531218e-01 -5.68525553e-01 -6.86457694e-01
7.94566572e-02 2.30336506e-02 5.88349223e-01 -7.25119233e-01
-5.67993857e-02 -4.13677543e-01 4.39766884e-01 -2.59599596e-01
1.73914164e-01 -9.87560511e-01 4.32258368e-01 2.34876081e-01
-3.98907483e-01 -1.81361333e-01 3.14473286e-02 6.60158813e-01
-4.49009448e-01 2.50660211e-01 7.52309144e-01 -1.91327393e-01
-3.62046450e-01 6.69676960e-01 3.79926115e-01 2.13845566e-01
1.09640849e+00 -3.16962242e-01 -2.83434272e-01 -1.98015586e-01
-5.84092975e-01 -5.07944748e-02 8.13611627e-01 5.38070798e-01
4.41588461e-01 -1.79827166e+00 -7.43945062e-01 6.03471398e-01
3.01113188e-01 6.17510557e-01 1.20982908e-01 6.64036632e-01
-1.32959530e-01 3.46731655e-02 -4.77599025e-01 -7.77792394e-01
-7.29758680e-01 7.86044002e-01 3.72303694e-01 -1.59240484e-01
-4.93101984e-01 8.37942421e-01 8.89009535e-01 -6.18539035e-01
-1.94701508e-01 -1.87282175e-01 2.95562148e-01 -2.41122618e-01
2.48923659e-01 -1.95664942e-01 -1.17896207e-01 -6.42872393e-01
-1.82072818e-01 7.49296308e-01 4.03240696e-02 -8.63447711e-02
1.17110026e+00 1.18092317e-02 -1.78475931e-01 1.89762473e-01
1.32072544e+00 2.61211116e-03 -1.53878689e+00 -2.88990438e-01
-4.54137981e-01 -7.75343597e-01 -3.81167442e-01 -6.20230734e-01
-1.28076410e+00 9.10381675e-01 3.30288768e-01 -2.37434268e-01
1.23692763e+00 2.08889380e-01 8.27696681e-01 5.64248897e-02
4.47572231e-01 -6.36319578e-01 2.92157561e-01 1.56612843e-01
1.10086584e+00 -1.38958204e+00 -3.11524391e-01 -4.95178968e-01
-5.97460449e-01 8.70291233e-01 9.52808440e-01 -3.25935900e-01
2.65283465e-01 1.14074849e-01 -3.84177789e-02 -1.71971083e-01
-4.06906068e-01 -5.17435931e-02 4.94255334e-01 8.71012032e-01
1.58203065e-01 2.25070819e-01 7.69846737e-02 5.37405968e-01
-3.13462287e-01 -2.27391377e-01 1.92950040e-01 6.30900979e-01
3.16958800e-02 -1.20697141e+00 -3.90911430e-01 2.02904835e-01
-9.09917355e-02 -4.99179736e-02 -4.65133786e-01 8.06615412e-01
3.32131505e-01 4.18770611e-01 2.12539554e-01 -4.96372253e-01
3.85844052e-01 -1.96238264e-01 6.99444294e-01 -7.02277482e-01
-2.89586991e-01 4.04097326e-02 -4.16001141e-01 -4.67732131e-01
-4.13413763e-01 -5.64332485e-01 -9.55691218e-01 -1.09464094e-01
-3.55509728e-01 -1.21838830e-01 3.81370664e-01 9.38547313e-01
5.45208752e-01 7.15546846e-01 8.72932613e-01 -7.73041010e-01
-6.77403867e-01 -8.88097703e-01 -3.79895031e-01 9.01193559e-01
3.44823897e-01 -7.29711592e-01 -4.41315949e-01 3.96633655e-01] | [11.68026351928711, -0.37211790680885315] |
4aaba07b-91ca-4b2b-8eb5-f6260f92896e | on-text-style-transfer-via-style-masked | 2210.06394 | null | https://arxiv.org/abs/2210.06394v1 | https://arxiv.org/pdf/2210.06394v1.pdf | On Text Style Transfer via Style Masked Language Models | Text Style Transfer (TST) is performable through approaches such as latent space disentanglement, cycle-consistency losses, prototype editing etc. The prototype editing approach, which is known to be quite successful in TST, involves two key phases a) Masking of source style-associated tokens and b) Reconstruction of this source-style masked sentence conditioned with the target style. We follow a similar transduction method, in which we transpose the more difficult direct source to target TST task to a simpler Style-Masked Language Model (SMLM) Task, wherein, similar to BERT \cite{bert}, the goal of our model is now to reconstruct the source sentence from its style-masked version. We arrive at the SMLM mechanism naturally by formulating prototype editing/ transduction methods in a probabilistic framework, where TST resolves into estimating a hypothetical parallel dataset from a partially observed parallel dataset, wherein each domain is assumed to have a common latent style-masked prior. To generate this style-masked prior, we use "Explainable Attention" as our choice of attribution for a more precise style-masking step and also introduce a cost-effective and accurate "Attribution-Surplus" method of determining the position of masks from any arbitrary attribution model in O(1) time. We empirically show that this non-generational approach well suites the "content preserving" criteria for a task like TST, even for a complex style like Discourse Manipulation. Our model, the Style MLM, outperforms strong TST baselines and is on par with state-of-the-art TST models, which use complex architectures and orders of more parameters. | ['Maunendra Sankar Desarkar', 'Suvodip Dey', 'Pooja Shekar', 'Sharan Narasimhan'] | 2022-10-12 | null | null | null | null | ['text-style-transfoer'] | ['natural-language-processing'] | [ 7.60903478e-01 4.24643338e-01 3.89274210e-02 -4.85184938e-01
-1.04641521e+00 -8.70785713e-01 1.19370282e+00 -2.55171955e-01
-2.89115161e-01 8.52462292e-01 4.74663883e-01 -3.81451875e-01
1.54234797e-01 -4.01233524e-01 -8.38039577e-01 -5.67932963e-01
4.87645686e-01 9.84615564e-01 -6.27472103e-02 -4.42956239e-01
4.25967634e-01 3.08368146e-01 -1.03963006e+00 4.88085359e-01
6.30850255e-01 4.18612123e-01 3.08559150e-01 8.65789354e-01
-3.16399634e-01 6.98128045e-01 -7.98985183e-01 -6.21935010e-01
4.91629615e-02 -7.10492969e-01 -1.28734732e+00 1.60048634e-01
3.09330612e-01 -1.59854759e-02 1.21587524e-02 8.48025024e-01
1.79248899e-01 2.42303237e-02 9.79893327e-01 -1.01864529e+00
-8.15629959e-01 9.04186487e-01 -4.64167655e-01 -2.16460183e-01
5.05720973e-01 6.57761842e-02 1.05982244e+00 -9.65845704e-01
8.55220616e-01 1.86740422e+00 4.08360660e-01 7.84549892e-01
-1.98877025e+00 -4.06587243e-01 2.03692362e-01 -2.40056247e-01
-1.13454270e+00 -5.96032858e-01 8.06068897e-01 -4.34345186e-01
8.27574790e-01 5.07408857e-01 1.87305704e-01 1.70973659e+00
2.12323107e-02 8.66449594e-01 1.53738105e+00 -7.94839978e-01
1.58522844e-01 5.00366807e-01 4.90306243e-02 4.30339158e-01
-1.75947770e-01 1.59543995e-02 -6.12855256e-01 -3.44005197e-01
6.14109814e-01 -4.63712484e-01 -1.14246115e-01 -1.46156192e-01
-1.54258573e+00 7.88243115e-01 7.72743300e-02 2.47067660e-01
-6.57362193e-02 1.61545858e-01 3.64130020e-01 6.64111495e-01
7.83192694e-01 4.55074370e-01 -4.64325577e-01 -5.17022088e-02
-1.18341637e+00 5.12940943e-01 8.53938460e-01 1.25440300e+00
7.65298069e-01 3.08080912e-02 -5.53713977e-01 6.93808556e-01
2.98056155e-01 3.32968146e-01 2.40857244e-01 -1.15159965e+00
5.96608758e-01 1.48024008e-01 2.41643235e-01 -5.48944712e-01
8.43059421e-02 -2.84069210e-01 -7.53036201e-01 3.03679407e-01
4.44550782e-01 -3.90257090e-02 -7.54592359e-01 2.24769831e+00
1.06312782e-01 -1.04899518e-01 9.97902267e-03 6.20252669e-01
1.36254445e-01 6.93999112e-01 7.80258328e-02 -2.90906399e-01
1.51788080e+00 -7.60634542e-01 -5.94975948e-01 -3.38843495e-01
5.22365630e-01 -1.07201147e+00 1.51146507e+00 4.64692205e-01
-1.31574082e+00 -5.44124901e-01 -9.59530950e-01 -4.57151949e-01
-7.04921782e-02 1.86378598e-01 4.36535776e-01 6.21598125e-01
-1.24681258e+00 8.65595222e-01 -6.10539675e-01 -3.67298752e-01
2.84763694e-01 1.28484428e-01 -3.36877733e-01 1.59694597e-01
-1.03856623e+00 1.13433373e+00 2.33658150e-01 -1.06769718e-01
-1.02499747e+00 -7.33006716e-01 -7.74436057e-01 -8.37768614e-02
4.66341436e-01 -1.18623447e+00 1.44628310e+00 -1.48769367e+00
-1.87309325e+00 1.19407165e+00 -6.94866002e-01 -2.42865354e-01
7.84798801e-01 -2.44868338e-01 -5.66348732e-02 -3.18009645e-01
3.22714776e-01 7.31318772e-01 1.24683726e+00 -1.59120846e+00
-5.81464209e-02 -1.27415270e-01 6.46532932e-03 2.10755587e-01
9.95570421e-02 4.26392049e-01 -1.09969541e-01 -1.04040217e+00
-1.16321675e-01 -1.22822475e+00 -4.65631187e-02 -1.96529198e-02
-8.24899256e-01 -2.31198341e-01 5.67444384e-01 -7.26969838e-01
1.03245020e+00 -2.05520487e+00 7.76753426e-01 -1.23452581e-01
2.68871248e-01 -1.09036610e-01 -1.36067078e-01 4.89535034e-01
-2.38891646e-01 2.90532976e-01 -6.88355982e-01 -1.17158043e+00
4.42783207e-01 3.59196633e-01 -6.57554030e-01 2.49240324e-01
4.45198923e-01 8.22037041e-01 -8.18094015e-01 -3.81105810e-01
-6.50412887e-02 1.86898157e-01 -7.24444747e-01 4.23682839e-01
-6.53113186e-01 5.65431476e-01 -9.04859975e-02 1.42012060e-01
4.06339496e-01 -1.25292614e-01 3.80489230e-01 -2.12015770e-02
-1.31400406e-01 5.99477410e-01 -1.17202175e+00 2.13673329e+00
-4.67472702e-01 7.18196154e-01 1.16971664e-01 -6.86959267e-01
8.07022929e-01 3.53903264e-01 -2.32330635e-01 -1.44760627e-02
7.55221471e-02 6.83594868e-02 -3.00880373e-01 -1.41758904e-01
8.25568020e-01 -6.14655912e-01 -4.99198735e-01 8.25966299e-01
2.02734992e-01 -2.32928887e-01 -1.75946340e-01 4.76802707e-01
8.72213662e-01 6.81341290e-01 2.07607150e-01 -5.65974593e-01
4.07162935e-01 -2.28873193e-02 4.94653583e-01 8.09225500e-01
1.14588127e-01 7.03620315e-01 8.68139863e-01 -2.91026589e-02
-1.34845901e+00 -1.13386238e+00 2.02897146e-01 1.12694275e+00
-3.48151416e-01 -3.02315503e-01 -1.02560830e+00 -6.31252348e-01
-2.26799831e-01 1.41667545e+00 -8.01836312e-01 -4.30647507e-02
-8.63209367e-01 -4.98246342e-01 7.79789388e-01 2.50225306e-01
2.47217119e-01 -1.12012267e+00 -1.06232531e-01 1.96015954e-01
-4.15212929e-01 -7.91774929e-01 -7.84953415e-01 1.27815142e-01
-6.80794954e-01 -4.27132428e-01 -5.76592088e-01 -6.09457076e-01
4.96972919e-01 -1.75858438e-01 1.40966690e+00 -2.00941622e-01
2.31373549e-01 9.16904435e-02 -6.33149296e-02 -4.24512565e-01
-1.06216741e+00 1.57433808e-01 4.36776280e-02 1.91487908e-01
3.83890085e-02 -6.79455459e-01 -1.57670468e-01 -9.21976753e-04
-1.09273541e+00 4.38582778e-01 5.53182364e-01 9.80200231e-01
2.13694364e-01 -4.69306856e-01 3.93161058e-01 -1.40609050e+00
6.30462587e-01 -3.67601395e-01 -3.50162446e-01 2.98564106e-01
-6.32396519e-01 4.69345003e-01 5.58547854e-01 -6.17059112e-01
-1.46781242e+00 -1.03477031e-01 -1.09069943e-02 -4.61552113e-01
-3.28541160e-01 5.28755374e-02 -5.77637672e-01 4.67040747e-01
7.28785276e-01 2.71367610e-01 6.97442368e-02 -6.76519215e-01
8.21595967e-01 5.79624951e-01 5.45101345e-01 -1.03498960e+00
1.01108587e+00 3.69877994e-01 -9.82403755e-02 -5.57469487e-01
-8.72537374e-01 8.49900618e-02 -7.39090502e-01 2.96610147e-01
9.89083111e-01 -7.82146156e-01 -4.52627897e-01 3.58277261e-01
-1.71811736e+00 -5.88412285e-01 -3.77724856e-01 -1.38127021e-02
-7.49733031e-01 5.49102247e-01 -6.85551882e-01 -7.32550263e-01
-7.55385309e-02 -1.13982737e+00 1.28084052e+00 -3.44104856e-01
-8.90862465e-01 -1.23409796e+00 6.52091056e-02 2.62766987e-01
4.43256944e-01 1.73966080e-01 1.37967718e+00 -5.48008263e-01
-4.97931808e-01 2.77687550e-01 -2.10742652e-01 5.14940500e-01
8.23680833e-02 -1.80762991e-01 -1.35232449e+00 -1.95462331e-01
3.81486058e-01 -1.72940403e-01 7.46189475e-01 1.55677423e-01
8.39747965e-01 -5.18027723e-01 -1.03957869e-01 5.47851861e-01
1.29871500e+00 -1.40828609e-01 5.67976534e-01 5.33101186e-02
8.88861716e-01 6.81415319e-01 2.11750418e-01 8.16145986e-02
4.59135920e-01 8.12646985e-01 -4.03100029e-02 -1.24445729e-01
-2.36062393e-01 -3.18007767e-01 7.19255984e-01 8.03347111e-01
1.34832352e-01 -3.76767009e-01 -4.55035001e-01 4.83840346e-01
-1.70744276e+00 -1.04772127e+00 -7.81236738e-02 2.14509439e+00
1.21664250e+00 3.69884044e-01 1.19447790e-01 1.20498076e-01
5.25135994e-01 2.26860210e-01 -2.44888350e-01 -8.68876219e-01
-2.40521818e-01 2.10888430e-01 1.95262730e-01 9.45788920e-01
-8.40715408e-01 1.13097262e+00 6.61778259e+00 9.07900333e-01
-7.96317756e-01 4.54840392e-01 6.11922562e-01 7.39105865e-02
-8.47859442e-01 5.15902579e-01 -6.75228536e-01 4.56454068e-01
1.02073753e+00 -1.24154404e-01 8.95284235e-01 5.01679361e-01
3.48453641e-01 4.24825698e-02 -1.73852026e+00 6.66952372e-01
1.53090000e-01 -1.36509585e+00 3.16100746e-01 -8.89894664e-02
5.32489777e-01 -3.30448061e-01 -5.27545661e-02 4.31994766e-01
5.61836362e-01 -8.77686799e-01 1.30129135e+00 5.63099027e-01
1.06167042e+00 -2.71488935e-01 1.28834784e-01 4.31856155e-01
-8.11293483e-01 2.36046404e-01 -1.38014853e-01 -2.21490070e-01
2.71527946e-01 4.93108660e-01 -7.66453564e-01 6.30976796e-01
1.13750711e-01 5.29135525e-01 -4.18401301e-01 6.22182488e-02
-3.41362268e-01 8.22874308e-01 2.47131567e-04 2.42510170e-01
1.02684302e-02 -2.86327213e-01 9.65133548e-01 1.47360981e+00
6.62930831e-02 -1.29181877e-01 5.14771529e-02 1.61931527e+00
-9.42826495e-02 -2.56411016e-01 -6.44179881e-01 1.48719668e-01
5.31929433e-01 1.00196707e+00 -5.54715514e-01 -7.75560737e-01
-1.00867175e-01 1.51856792e+00 3.51041198e-01 3.96155715e-01
-7.76378989e-01 -1.14208773e-01 5.47828376e-01 -1.41949385e-01
1.01098821e-01 -1.53961718e-01 -6.75971508e-01 -1.37635708e+00
1.40787894e-02 -1.04410994e+00 3.56354401e-04 -1.06994700e+00
-1.37984705e+00 6.91959620e-01 3.26162279e-01 -9.78278458e-01
-3.61480951e-01 -3.58287275e-01 -7.38497674e-01 1.46682131e+00
-1.35295641e+00 -1.44027185e+00 2.28898391e-01 4.94411528e-01
9.51500535e-01 -1.13124987e-02 1.05004239e+00 -7.09458888e-02
-4.52458650e-01 5.96449316e-01 -1.47559255e-01 -1.35136306e-01
9.23490584e-01 -1.77075708e+00 7.29107141e-01 1.01216424e+00
3.38796042e-02 8.65924597e-01 1.05713964e+00 -7.87911594e-01
-1.30250394e+00 -1.13069522e+00 1.35826135e+00 -1.12915373e+00
7.28636026e-01 -9.15981352e-01 -8.81602228e-01 1.16778696e+00
5.28125823e-01 -6.74932063e-01 3.65483105e-01 9.89116430e-02
-5.07471800e-01 2.17256233e-01 -8.87678802e-01 8.57781351e-01
1.11887741e+00 -7.75723279e-01 -1.08230698e+00 3.29531133e-01
1.01616848e+00 -3.22644085e-01 -6.09098673e-01 -1.78357288e-01
3.29402387e-01 -6.78383231e-01 7.79727697e-01 -6.33272350e-01
4.93252933e-01 -3.67450625e-01 -2.57251859e-01 -1.45854640e+00
-4.62420583e-01 -1.18939424e+00 2.98732370e-01 1.74755561e+00
5.10447025e-01 -5.44913828e-01 3.71154934e-01 4.84165758e-01
-2.88670182e-01 -1.22546956e-01 -8.88412356e-01 -7.08527088e-01
4.40265507e-01 -4.52023685e-01 6.55922592e-01 9.92979109e-01
-3.07586759e-01 8.88297439e-01 -4.92942989e-01 -4.44767512e-02
7.78597176e-01 1.43702522e-01 7.30162442e-01 -1.13408470e+00
-7.70123720e-01 -6.45457625e-01 4.88894254e-01 -9.96903539e-01
4.05283540e-01 -9.98528123e-01 1.44665763e-01 -1.19050300e+00
1.73664466e-01 -5.00716031e-01 1.51707567e-02 3.15604836e-01
-3.02264839e-01 1.95664596e-02 3.35261822e-01 3.36245269e-01
-3.06658149e-01 4.49553609e-01 1.09089947e+00 -1.71997294e-01
-6.22554012e-02 -8.28171223e-02 -1.02279866e+00 6.60018742e-01
5.32313049e-01 -7.99591243e-01 -3.85256588e-01 -5.64269066e-01
2.21677855e-01 3.13926846e-01 5.40035605e-01 -2.57867903e-01
-1.54239684e-01 -2.65029609e-01 2.10657135e-01 -1.97259873e-01
3.80509436e-01 -3.62990946e-01 4.14242238e-01 2.32841447e-01
-7.69024849e-01 2.47838825e-01 3.28433439e-02 4.73520786e-01
2.00539142e-01 -2.21808285e-01 5.85460782e-01 -3.05616796e-01
-2.07632810e-01 -1.92851886e-01 -6.24223053e-01 5.37839197e-02
4.56540763e-01 -8.91624987e-02 -1.95331052e-01 -1.94037974e-01
-7.87394643e-01 -3.14972043e-01 6.65791631e-01 4.26400185e-01
1.20211229e-01 -1.36118650e+00 -9.36187029e-01 1.37238756e-01
-8.72569457e-02 4.33046222e-02 -7.11614564e-02 7.34694242e-01
-1.29515961e-01 2.16604501e-01 2.96464060e-02 -5.94037890e-01
-9.93220270e-01 6.92296267e-01 3.01887263e-02 -3.96082610e-01
-4.65825349e-01 7.60241091e-01 5.24987400e-01 -5.48133254e-01
-1.15872391e-01 -3.03547174e-01 2.00292572e-01 -1.57018363e-01
3.40778828e-01 2.21103966e-01 -8.32053181e-03 -5.06560206e-01
-1.07391424e-01 3.21947306e-01 -8.12588111e-02 -6.89053774e-01
1.03553319e+00 -4.79122102e-01 -3.73099744e-01 8.88165474e-01
9.77013409e-01 2.39201620e-01 -1.25896525e+00 -3.52540076e-01
2.33995810e-01 -1.68689132e-01 -1.81057438e-01 -9.03655648e-01
-2.72759795e-01 8.85693789e-01 9.00536925e-02 8.44401196e-02
9.32680666e-01 -5.19195497e-02 5.36534727e-01 -1.42510782e-03
1.79039091e-01 -7.64430404e-01 1.80342756e-02 5.27778327e-01
1.16674912e+00 -9.32980955e-01 -1.74749434e-01 -2.69365042e-01
-7.76854098e-01 8.74750257e-01 1.82910591e-01 -2.19376162e-01
2.89526880e-01 2.52043247e-01 5.05977273e-02 -1.42222736e-03
-1.02280509e+00 2.40815237e-01 2.77894765e-01 4.17848676e-01
4.40472186e-01 3.05474877e-01 -1.35861948e-01 5.07391453e-01
-4.93820727e-01 -1.43767804e-01 6.14778519e-01 7.36656606e-01
-5.31536490e-02 -1.54366755e+00 -4.66361314e-01 1.13977224e-01
-2.91676015e-01 -3.03142428e-01 -5.34918666e-01 6.60387516e-01
-1.35704607e-03 8.78479898e-01 -1.44850373e-01 -2.00398937e-01
1.91236079e-01 4.54840660e-01 6.50612533e-01 -9.01494205e-01
-7.35809982e-01 3.22760433e-01 4.02338177e-01 -3.40031296e-01
-3.23162526e-01 -8.87999833e-01 -7.56721795e-01 -5.14249861e-01
-1.11428849e-01 3.54042314e-02 5.04189134e-01 1.40187943e+00
2.02961653e-01 4.81414467e-01 6.78858221e-01 -1.26402438e+00
-7.38279581e-01 -1.20622826e+00 -3.25023353e-01 6.71345174e-01
3.76159400e-01 -4.77763116e-01 -2.94409961e-01 4.58438158e-01] | [11.5960693359375, 9.54311466217041] |
8b055f45-0011-460c-ba7f-95eca659fbff | machine-learning-with-tree-tensor-networks-cp | 2305.19440 | null | https://arxiv.org/abs/2305.19440v1 | https://arxiv.org/pdf/2305.19440v1.pdf | Machine learning with tree tensor networks, CP rank constraints, and tensor dropout | Tensor networks approximate order-$N$ tensors with a reduced number of degrees of freedom that is only polynomial in $N$ and arranged as a network of partially contracted smaller tensors. As suggested in [arXiv:2205.15296] in the context of quantum many-body physics, computation costs can be further substantially reduced by imposing constraints on the canonical polyadic (CP) rank of the tensors in such networks. Here we demonstrate how tree tensor networks (TTN) with CP rank constraints and tensor dropout can be used in machine learning. The approach is found to outperform other tensor-network based methods in Fashion-MNIST image classification. A low-rank TTN classifier with branching ratio $b=4$ reaches test set accuracy 90.3\% with low computation costs. Consisting of mostly linear elements, tensor network classifiers avoid the vanishing gradient problem of deep neural networks. The CP rank constraints have additional advantages: The number of parameters can be decreased and tuned more freely to control overfitting, improve generalization properties, and reduce computation costs. They allow us to employ trees with large branching ratios which substantially improves the representation power. | ['Thomas Barthel', 'Hao Chen'] | 2023-05-30 | null | null | null | null | ['tensor-networks'] | ['methodology'] | [ 6.99980035e-02 2.61868209e-01 -3.97217095e-01 -4.48141098e-01
-3.65743965e-01 -5.17982006e-01 4.27598804e-01 -1.57022536e-01
-3.75531614e-01 5.14305770e-01 -1.76069438e-01 -4.58823055e-01
-5.09913683e-01 -7.53000557e-01 -6.65105104e-01 -8.66768360e-01
-5.87490559e-01 5.33758461e-01 1.99650899e-01 -3.39135647e-01
3.61905694e-01 6.75517857e-01 -1.28789246e+00 3.86045754e-01
6.57953799e-01 1.33486569e+00 -3.63561332e-01 5.16317725e-01
1.50657237e-01 1.02592826e+00 -2.23400984e-02 -7.35766709e-01
4.89323616e-01 1.27162337e-01 -1.02168977e+00 -3.43028098e-01
8.11275542e-01 -2.12561011e-01 -9.86576021e-01 1.16207659e+00
1.63784146e-01 2.11047947e-01 4.60698813e-01 -9.84894335e-01
-5.88645458e-01 9.83103633e-01 -3.18775684e-01 3.07688415e-01
-3.29000920e-01 -1.79407131e-02 1.67257547e+00 -5.77185690e-01
4.66529965e-01 1.30333567e+00 6.62327111e-01 4.81949866e-01
-1.64548957e+00 -5.48343718e-01 -2.13318586e-01 3.45056534e-01
-1.15640461e+00 -2.95832396e-01 8.19517732e-01 -4.32538509e-01
1.14962327e+00 5.01784384e-01 5.68346322e-01 5.57048738e-01
2.15459853e-01 2.88074523e-01 9.82346535e-01 -3.33541393e-01
2.63412632e-02 -2.68917620e-01 5.59047937e-01 1.32426631e+00
3.65586489e-01 -3.72645296e-02 -7.24047124e-01 -3.64068419e-01
6.66596532e-01 -4.24884669e-02 -8.05994682e-03 -5.59560180e-01
-1.60273314e+00 1.04817653e+00 9.82561290e-01 2.82792211e-01
2.80776080e-02 9.27389145e-01 6.15843594e-01 2.65424162e-01
3.16738605e-01 8.30610096e-01 -5.35644114e-01 4.42178436e-02
-4.57668126e-01 3.74127060e-01 6.17198110e-01 7.33871460e-01
8.40446234e-01 3.33596796e-01 3.56478244e-02 7.81274259e-01
-9.66116339e-02 4.35701609e-01 1.91303030e-01 -1.47403908e+00
5.11285901e-01 5.02917886e-01 -2.25754187e-01 -9.15001869e-01
-4.78496760e-01 -4.49537456e-01 -1.16135907e+00 -8.21178779e-03
5.59101045e-01 2.99764723e-01 -8.63249958e-01 1.54741275e+00
8.43026638e-02 -4.14904356e-01 -2.95411617e-01 7.81407773e-01
4.79560196e-01 3.58537436e-01 -1.81514427e-01 2.12312371e-01
1.51808643e+00 -7.23148108e-01 -3.99219334e-01 6.15145117e-02
1.23798466e+00 -6.24075651e-01 8.46092284e-01 5.54204106e-01
-9.66962636e-01 -3.96222388e-03 -1.12306428e+00 -4.07273591e-01
-2.39275798e-01 1.70477442e-02 1.67560291e+00 7.56104112e-01
-1.11814964e+00 1.32260263e+00 -8.85520101e-01 2.00860158e-01
3.79730344e-01 1.01470590e+00 -4.47419018e-01 -2.56957799e-01
-1.03917563e+00 7.21868813e-01 4.78710890e-01 5.68463206e-01
-6.60732031e-01 -6.41604781e-01 -5.39100409e-01 -1.10358693e-01
6.47597313e-02 -6.61139250e-01 1.03651083e+00 -3.95582855e-01
-1.28515768e+00 6.49539649e-01 1.30419493e-01 -4.13590431e-01
2.04108153e-02 2.50824869e-01 -4.77400608e-02 3.38162929e-01
-7.61883035e-02 6.42125010e-01 6.36608422e-01 -3.89654487e-01
-3.23548391e-02 -5.09462833e-01 4.25938249e-01 -1.07785761e-01
-4.29050028e-01 -7.66447484e-02 -1.66280605e-02 -3.40963304e-01
9.07450855e-01 -1.50292945e+00 -3.71916801e-01 -1.79113328e-01
-5.54877162e-01 -3.60178590e-01 5.74636817e-01 -3.00764114e-01
6.31013572e-01 -1.87543368e+00 4.08097774e-01 3.53125989e-01
7.67184973e-01 -1.68312505e-01 -1.92095488e-01 2.57771403e-01
-3.51945430e-01 3.83796513e-01 -8.30309168e-02 3.57854366e-02
-2.50743125e-02 4.61155862e-01 -4.35290672e-03 5.10811150e-01
1.21740602e-01 5.88299334e-01 -6.20230436e-01 -2.73059487e-01
-2.03495711e-01 2.49648780e-01 -1.02943671e+00 -5.81034660e-01
-2.57851958e-01 2.59616911e-01 -4.99406844e-01 6.43524170e-01
5.36670983e-01 -5.02793133e-01 2.39675462e-01 -6.89356565e-01
2.44213715e-02 5.98334670e-01 -9.03904915e-01 1.49458146e+00
-1.57713011e-01 6.08051598e-01 7.73292407e-02 -1.34262919e+00
5.97482562e-01 1.38694495e-01 6.12730145e-01 -5.41343272e-01
6.62709996e-02 4.20723617e-01 7.51658857e-01 -1.71829522e-01
5.17466009e-01 -3.06081325e-01 -5.46965189e-02 3.45817477e-01
2.48424187e-01 -3.17483038e-01 5.85182548e-01 3.61398607e-01
1.34611917e+00 -3.46506685e-01 -5.82706094e-01 -5.27940929e-01
3.19943190e-01 -2.83966754e-02 5.85097909e-01 6.80624068e-01
-5.67628406e-02 5.96377850e-02 9.50255454e-01 -8.62787127e-01
-1.44270003e+00 -7.69659460e-01 -5.84195077e-01 1.24275041e+00
-3.54567468e-01 -7.04918206e-01 -4.87700880e-01 -3.46997082e-01
1.18450761e-01 3.50236654e-01 -5.17750502e-01 -1.75952464e-01
-9.77269888e-01 -1.28091919e+00 6.09412134e-01 3.46911222e-01
3.98690730e-01 -6.35190725e-01 1.25101417e-01 -5.04798815e-02
-2.02736557e-01 -1.30721259e+00 -1.14575855e-01 4.73439783e-01
-1.55147374e+00 -8.52055967e-01 -2.30100185e-01 -4.35274184e-01
6.51905954e-01 6.16463646e-02 9.11593497e-01 1.00725591e-01
-2.96628088e-01 -8.48753974e-02 -8.09733272e-02 2.20154628e-01
-2.45061085e-01 2.45201096e-01 5.14757514e-01 -3.01430494e-01
1.60496220e-01 -7.62014449e-01 -5.23352087e-01 1.38129249e-01
-7.75685728e-01 -2.03662038e-01 4.95421886e-01 1.11202765e+00
2.16832027e-01 2.26634324e-01 -2.41059810e-01 -5.99078298e-01
1.52187914e-01 5.45125455e-02 -7.15614140e-01 -2.15061739e-01
-6.36057079e-01 6.40848279e-01 6.75503194e-01 -3.61803681e-01
-4.24393058e-01 -2.26093173e-01 3.24183762e-01 -3.29521239e-01
6.40136540e-01 5.53046167e-01 4.40594941e-01 -7.22907960e-01
6.70640826e-01 -1.68657854e-01 -1.50308669e-01 -6.09445632e-01
3.73847514e-01 4.18332294e-02 4.97056320e-02 -1.02321982e+00
6.90944195e-01 3.48562181e-01 9.72926974e-01 -8.47728372e-01
-1.07686174e+00 -4.13021153e-05 -8.42108846e-01 9.07182321e-02
6.97712302e-01 -8.08107316e-01 -1.30628109e+00 1.70909896e-01
-1.07727122e+00 -3.58662158e-02 -6.28138334e-02 8.82997215e-01
-4.11405116e-01 3.94259214e-01 -1.28246427e+00 -4.60272849e-01
-2.44611248e-01 -1.26717508e+00 5.64637005e-01 -2.78193146e-01
1.48733482e-01 -6.40343606e-01 -4.22896802e-01 7.64013231e-01
3.55890661e-01 -4.34451364e-02 1.56854510e+00 -3.22149307e-01
-1.08593047e+00 -3.57447922e-01 -4.41905797e-01 5.59955776e-01
-4.13596869e-01 -5.54996468e-02 -8.08488071e-01 -2.77339429e-01
-1.44199982e-01 -2.74340481e-01 1.05025637e+00 2.98928529e-01
1.42625260e+00 -5.37603557e-01 -7.44750872e-02 6.40944362e-01
1.10513210e+00 -2.24305481e-01 3.03780228e-01 6.93035796e-02
1.22784877e+00 6.08346999e-01 -2.04345018e-01 4.94328365e-02
2.04638392e-01 6.06102169e-01 4.39893365e-01 4.67095673e-01
4.13900167e-02 1.34780571e-01 2.69884557e-01 1.45057976e+00
-6.79752827e-01 4.25397635e-01 -1.09424949e+00 1.34239823e-01
-1.54779160e+00 -9.12364960e-01 -6.00423694e-01 2.14201713e+00
7.78686523e-01 4.72801834e-01 2.89390385e-02 1.80967212e-01
4.14659619e-01 1.51745319e-01 -4.66200322e-01 -6.45923972e-01
-3.42434784e-03 3.81466269e-01 9.91780937e-01 5.30176103e-01
-9.12989736e-01 8.56665015e-01 6.66771078e+00 7.17353940e-01
-9.65495288e-01 2.85419255e-01 4.89777088e-01 -3.86750668e-01
-2.80451685e-01 2.26568028e-01 -7.43956864e-01 -3.23801592e-04
1.18538332e+00 3.14434856e-01 8.41784000e-01 7.69647717e-01
-1.14119835e-01 2.49232262e-01 -1.37167215e+00 1.07742655e+00
-3.76307994e-01 -1.64027727e+00 7.91888088e-02 4.37926739e-01
5.28401732e-01 6.71405554e-01 2.96722770e-01 3.10118765e-01
2.63358176e-01 -1.07609737e+00 5.19750416e-01 3.37284863e-01
7.72324860e-01 -7.38046110e-01 4.97276574e-01 1.59073379e-02
-8.86056721e-01 -2.33612686e-01 -7.63580978e-01 -3.05365473e-01
-2.94585943e-01 9.28021014e-01 -4.51774597e-01 2.65098602e-01
8.97015572e-01 3.45057189e-01 -5.75457036e-01 6.05468154e-01
1.91386566e-01 5.92754066e-01 -7.54711092e-01 -4.13071871e-01
5.29312015e-01 -7.14302421e-01 4.37225461e-01 6.39963865e-01
9.02087763e-02 2.67743558e-01 -1.48377083e-02 8.79199684e-01
-3.29707772e-01 -1.22294761e-01 -6.36959851e-01 -4.96694118e-01
7.24819601e-02 1.30223775e+00 -6.17587268e-01 -2.23566934e-01
9.20899585e-02 5.97123981e-01 5.35747349e-01 1.17026269e-01
-3.68899643e-01 -4.75582272e-01 4.87697482e-01 3.06035634e-02
4.33439642e-01 -7.00350046e-01 -3.01973522e-01 -1.54791069e+00
1.81827933e-01 -6.69419706e-01 4.93781194e-02 -4.58762676e-01
-1.23321939e+00 6.70425594e-01 -1.04822174e-01 -7.68999577e-01
2.94303354e-02 -1.28744280e+00 -1.00846201e-01 6.23547256e-01
-9.21693623e-01 -1.02086222e+00 3.32895130e-01 3.21630418e-01
-3.34669948e-01 -1.35410507e-03 9.88739610e-01 4.78869706e-01
-8.53068352e-01 4.77582514e-01 2.84665138e-01 3.29916388e-01
5.92822619e-02 -1.35834217e+00 1.43979132e-01 4.83583003e-01
4.20537740e-02 9.62975442e-01 7.07209945e-01 -2.52611488e-01
-1.95667791e+00 -7.73991466e-01 8.66656482e-01 -4.32122231e-01
1.16895401e+00 -7.42403328e-01 -6.17152452e-01 7.20037639e-01
-2.72404432e-01 5.41307330e-01 7.14445829e-01 8.55689883e-01
-1.01560891e+00 -4.84345496e-01 -9.77988243e-01 5.99073410e-01
1.10300672e+00 -8.72299612e-01 -1.02730151e-02 1.05593312e+00
8.43926430e-01 -2.22269744e-01 -1.37405217e+00 4.54148203e-01
5.37910044e-01 -8.14160943e-01 9.91975486e-01 -8.48770320e-01
2.54374266e-01 7.88079854e-03 -4.76046145e-01 -8.36165309e-01
-4.65085000e-01 -8.34591091e-01 -1.04084611e-01 4.96333301e-01
4.91476536e-01 -6.90278947e-01 1.04737270e+00 8.29439640e-01
-1.61337912e-01 -6.92615032e-01 -1.28200161e+00 -8.59444976e-01
3.85712743e-01 -5.13829768e-01 2.43213981e-01 1.02113891e+00
2.22460523e-01 6.13304257e-01 -2.36716256e-01 8.92704353e-03
1.03828895e+00 -1.23257317e-01 1.98705480e-01 -1.44019163e+00
-3.67183030e-01 -4.98665303e-01 -6.02174461e-01 -9.01068807e-01
1.70584649e-01 -1.51260805e+00 -4.67474729e-01 -6.67374671e-01
2.14382991e-01 -1.02289629e+00 -3.46550822e-01 6.30704463e-01
6.25698686e-01 4.88786936e-01 2.64142722e-01 4.19654399e-01
-3.78843337e-01 5.75893998e-01 1.39028442e+00 -1.98014572e-01
3.43904614e-01 -2.20737785e-01 -3.26499313e-01 8.68548691e-01
6.40778363e-01 -6.89653575e-01 3.68974730e-02 -5.25607109e-01
6.15136921e-01 1.60171479e-01 3.78322393e-01 -9.82748210e-01
2.76905689e-02 9.56637114e-02 1.89755067e-01 -5.06655872e-01
7.45853364e-01 -3.74471217e-01 -1.54655263e-01 6.46858811e-01
-3.54473919e-01 2.14230821e-01 -1.10088229e-01 3.44359905e-01
5.00660576e-02 -5.12800753e-01 6.91939592e-01 -3.42251122e-01
1.29832961e-02 4.76020634e-01 -2.14806825e-01 -2.28269890e-01
4.17446434e-01 2.43349895e-01 -4.71964598e-01 5.74046485e-02
-7.56796360e-01 -1.63118795e-01 5.28589562e-02 -5.90706691e-02
1.59612969e-01 -1.46162724e+00 -3.36570293e-01 -6.40080720e-02
-9.50201303e-02 -1.08374201e-01 2.69682646e-01 1.02518749e+00
-8.89839351e-01 9.39108253e-01 -6.06160536e-02 -8.19204628e-01
-7.54480243e-01 3.33222598e-01 5.75094342e-01 -3.86288106e-01
-6.10961854e-01 1.07964647e+00 -1.91992149e-01 -9.64592993e-01
-3.48610468e-02 -6.46228790e-01 3.09588373e-01 -1.40786022e-01
1.68154150e-01 4.62706566e-01 3.26729149e-01 -6.20493233e-01
-3.23657513e-01 5.19066870e-01 -3.83625627e-01 -1.78324971e-02
1.41564941e+00 5.84963441e-01 -9.18092549e-01 3.34154367e-01
1.54312491e+00 -2.98078716e-01 -5.68370044e-01 -3.25810462e-01
1.18733153e-01 -1.74500644e-01 4.44603920e-01 -4.93743509e-01
-1.20369995e+00 9.78620112e-01 4.86780286e-01 4.29703385e-01
5.19465208e-01 4.38254662e-02 7.32970297e-01 1.24047804e+00
3.98798645e-01 -8.95546913e-01 2.15091169e-01 8.56400907e-01
6.91885650e-01 -9.33476269e-01 1.93041906e-01 -4.57637459e-01
-6.36759102e-02 1.42949617e+00 3.45646441e-01 -3.86734605e-01
7.52267182e-01 -2.10504964e-01 -5.40236890e-01 -7.22557783e-01
-7.55689502e-01 1.07531324e-01 4.74221677e-01 1.26551926e-01
6.35081589e-01 3.31091225e-01 -1.86241254e-01 -6.49352744e-02
-6.80048823e-01 -4.29790795e-01 7.75274813e-01 5.15348375e-01
-2.93998599e-01 -1.29385018e+00 -2.39607453e-01 9.05358434e-01
-6.09351933e-01 -4.68769103e-01 1.50692821e-01 5.77361107e-01
-3.04345577e-03 7.85129726e-01 -2.46576760e-02 -4.51547354e-01
5.44941165e-02 1.12457231e-01 9.83943939e-01 -4.54540551e-01
-3.98800701e-01 -2.34951079e-01 3.40938210e-01 -7.63305783e-01
-3.42772037e-01 -6.09388649e-01 -1.08294046e+00 -9.26494360e-01
-5.25272131e-01 2.34770820e-01 8.83426964e-01 8.85387182e-01
3.36741686e-01 2.83564299e-01 4.77314711e-01 -9.43212926e-01
-9.98663068e-01 -1.02444708e+00 -7.53172874e-01 1.35668620e-01
2.65645295e-01 -7.31965184e-01 -6.30377769e-01 -3.40793163e-01] | [5.9344024658203125, 4.986341953277588] |
98580531-9af6-4838-b7d5-2d31e151acd5 | a-comparative-analysis-on-bangla-handwritten | null | null | https://scholar.google.com/citations?view_op=view_citation&hl=en&user=zQKHA64AAAAJ&citation_for_view=zQKHA64AAAAJ:d1gkVwhDpl0C | https://ieeexplore.ieee.org/abstract/document/9152905/ | A Comparative Analysis on Bangla Handwritten Digit Recognition with Data Augmentation and Non-Augmentation Process | Determination of Bangla handwritten digit is a
momentous image classification task. Though object recognition
technology is getting smarter day by day, still Bangla handwritten
digit recognition remains inconclusive. Researchers are becoming
more concerned about handwritten digit recognition for it’s
educational and advantageous importance. But it is a matter
of trouble that the improvement in Bangla handwritten digit
recognition is significantly less as compared to the other languages. To improve the performance of the Bangla handwritten
digit recognition system, we have designed a model, in which
all basic Bangla digits have been classified. Furthermore, we
have also demonstrated Densenet121 architecture in our system.
For recognizing Bangla handwriting digits, we proposed CNN
(Convolution Neural Network) model. Our system has been
experimented on the NumtaDB dataset for recognizing Bangla
digit both with augmentation and non-augmentation. | ['Atiqul Islam Chowdhury', 'Mahim Anzum Haque Pantho', 'Refat E Ferdous', 'MD Abdullah Al Nasim'] | 2020-06-26 | null | null | null | international-congress-on-human-computer | ['handwritten-digit-recognition'] | ['computer-vision'] | [-3.97916585e-01 -3.88886392e-01 1.19875848e-01 -8.29995453e-01
1.16066508e-01 -6.88101470e-01 7.00344801e-01 -3.67868751e-01
-5.97045660e-01 5.36272526e-01 2.17456087e-01 -7.53735483e-01
1.40758991e-01 -9.44149494e-01 -1.95434526e-01 -6.56191587e-01
4.90576208e-01 4.00980026e-01 -4.12271842e-02 -2.42088825e-01
5.21225154e-01 1.36198545e+00 -9.48145688e-01 5.89459002e-01
3.15569103e-01 7.83354759e-01 -3.84906046e-02 9.19772387e-01
-1.73822477e-01 1.33798432e+00 -7.39588678e-01 -5.70669889e-01
4.73358154e-01 -4.68273848e-01 -9.69329894e-01 4.23436940e-01
3.23174804e-01 -8.85513842e-01 -9.74911451e-01 1.01582730e+00
7.12360084e-01 -3.30880761e-01 7.31650829e-01 -6.87870443e-01
-1.34094906e+00 5.21968067e-01 -7.40806520e-01 3.99790138e-01
-4.20994043e-01 5.65533899e-02 3.11982632e-01 -9.45443034e-01
3.54849398e-01 1.21980059e+00 5.33345640e-01 6.79227233e-01
-2.13736668e-01 -6.02719545e-01 -1.85269490e-01 3.86286288e-01
-1.37294829e+00 -1.97721735e-01 6.38527870e-01 -4.13339257e-01
1.17320633e+00 5.72505742e-02 3.59530210e-01 5.84869206e-01
2.70880073e-01 9.40901220e-01 1.22991168e+00 -7.01681852e-01
-2.48465121e-01 2.50530124e-01 7.07497776e-01 5.78707159e-01
3.05887640e-01 -1.78375423e-01 2.70708323e-01 6.88864052e-01
1.07917082e+00 1.77928776e-01 1.97412178e-01 3.86549324e-01
-1.19259191e+00 6.23358190e-01 3.46687019e-01 6.31457627e-01
-3.48471016e-01 -8.24422687e-02 4.78483975e-01 5.52045703e-01
-1.55716628e-01 6.60380647e-02 -4.07394677e-01 -2.34300375e-01
-6.74387336e-01 1.88088398e-02 6.12665057e-01 9.44036603e-01
2.09833264e-01 4.74383771e-01 -6.27207980e-02 1.23090231e+00
1.45182461e-01 3.51422757e-01 6.56169534e-01 -2.32423112e-01
4.21974570e-01 8.52943718e-01 -1.18573599e-01 -1.10354090e+00
-1.26416460e-01 -2.30994120e-01 -1.51787937e+00 6.22536480e-01
6.90156817e-01 -3.37429434e-01 -1.53611994e+00 6.45307004e-01
-2.90018290e-01 -6.35037303e-01 3.90076965e-01 1.27167511e+00
9.85475838e-01 1.01309645e+00 -1.44585103e-01 3.10283422e-01
1.41153526e+00 -8.63704801e-01 -7.50572443e-01 1.73819125e-01
2.66965181e-01 -1.14861810e+00 7.46873915e-01 8.12360048e-01
-6.96136832e-01 -8.96650553e-01 -1.05336308e+00 -3.46409231e-01
-5.55029988e-01 8.04637551e-01 8.25316489e-01 1.13477290e+00
-6.10083103e-01 3.07301402e-01 -6.21242046e-01 -7.01290667e-01
7.83222377e-01 5.37500381e-01 -4.25117314e-01 -3.63114357e-01
-7.00005710e-01 1.08686197e+00 6.05079710e-01 3.90736222e-01
-8.24932754e-01 1.95584774e-01 -3.43455851e-01 -1.60596579e-01
-3.39306682e-01 2.73601860e-01 1.00068879e+00 -9.90674734e-01
-1.37443638e+00 1.02910376e+00 2.40468487e-01 -4.25033391e-01
5.25144458e-01 2.63975173e-01 -8.50908756e-01 -2.80616730e-01
-5.54050446e-01 6.79787934e-01 5.63221335e-01 -6.33763373e-01
-8.47202122e-01 -7.41614342e-01 -1.99801400e-01 1.18745072e-02
-5.65557778e-01 4.39405501e-01 -2.93914199e-01 -9.12303925e-01
3.27235639e-01 -4.15434510e-01 -1.38159692e-01 -5.00177667e-02
-1.59327000e-01 -3.06761503e-01 1.48470199e+00 -1.07780302e+00
8.16469431e-01 -2.12458706e+00 -5.98384857e-01 1.35510892e-01
-1.26388013e-01 6.17743671e-01 -2.75196612e-01 1.29644617e-01
3.74054052e-02 -1.40291482e-01 4.40322496e-02 2.94014186e-01
-2.97899544e-01 3.13923925e-01 -3.30412745e-01 3.28111231e-01
5.22543013e-01 6.32141888e-01 -1.89072803e-01 -3.09965104e-01
3.12854916e-01 5.62319815e-01 -6.90953806e-02 4.44558859e-02
3.02152246e-01 2.21009269e-01 -3.16255450e-01 1.30643702e+00
1.04701602e+00 1.89804807e-01 1.28264548e-02 -1.82277456e-01
-4.65278745e-01 -2.73455065e-02 -1.30983508e+00 9.87771273e-01
-1.94345817e-01 9.91775095e-01 -3.96723181e-01 -1.28728664e+00
1.56380749e+00 3.64686072e-01 -3.06166261e-01 -8.35608959e-01
5.43241084e-01 4.16805148e-01 6.19790316e-01 -6.75777197e-01
6.18308365e-01 1.31609678e-01 3.01116377e-01 3.47308993e-01
1.10628225e-01 1.17464848e-01 1.28790602e-01 -1.10849515e-01
7.29368150e-01 -1.44211814e-01 1.48223191e-01 -2.17817977e-01
6.93628490e-01 3.75737369e-01 1.36013538e-01 6.07491612e-01
-1.75234988e-01 7.13025331e-01 3.87049556e-01 -1.02156186e+00
-1.51963246e+00 -3.22671235e-01 -4.16896611e-01 5.57150602e-01
-3.35909486e-01 6.48971736e-01 -5.21962881e-01 -5.80774426e-01
3.11649628e-02 1.53005198e-01 -3.88233691e-01 2.84745067e-01
-6.32542193e-01 -9.63396609e-01 1.09743536e+00 8.81285310e-01
1.53610492e+00 -1.27055204e+00 -1.98614702e-01 2.70034790e-01
6.37510777e-01 -1.03532505e+00 -3.09995741e-01 3.65648627e-01
-9.85150993e-01 -8.34908545e-01 -1.39371860e+00 -1.51869190e+00
7.57549465e-01 7.61199817e-02 4.63678449e-01 -2.36806437e-01
-6.68161571e-01 -1.13663450e-01 -3.99656713e-01 -6.97404027e-01
-5.00455737e-01 1.47607839e-02 -1.29818857e-01 -1.74314771e-02
9.04571176e-01 -7.99429268e-02 -4.20798957e-01 1.98307902e-01
-5.66694617e-01 -2.70602286e-01 1.11840260e+00 6.20575011e-01
8.61589387e-02 4.44240719e-01 7.06542492e-01 -7.29653835e-01
7.50283241e-01 -4.46458086e-02 -7.97598064e-01 2.45274156e-01
-4.37041253e-01 -3.15430045e-01 7.96583652e-01 -4.53057796e-01
-1.04923415e+00 2.19974741e-01 -4.19781119e-01 1.30513743e-01
-6.16016567e-01 4.48047012e-01 -2.32524052e-01 -1.99869588e-01
4.65225697e-01 6.10905886e-01 -4.08588573e-02 -7.16622591e-01
-1.25720888e-01 1.74206841e+00 6.13962948e-01 -1.66172460e-01
1.50127947e-01 1.21132545e-01 -1.38512015e-01 -1.12725675e+00
-8.60094056e-02 -2.44701892e-01 -8.68300438e-01 -2.16850892e-01
8.38041246e-01 -8.41565013e-01 -8.97432625e-01 1.39360297e+00
-1.43647420e+00 1.53585553e-01 1.87089249e-01 5.69742978e-01
2.75299877e-01 4.14176106e-01 -9.10691917e-01 -8.75224471e-01
-2.65875638e-01 -1.04686642e+00 2.60464042e-01 2.70316362e-01
1.83011353e-01 -7.43141472e-01 -2.51402348e-01 2.42981091e-01
4.72458601e-01 1.39735397e-02 9.66746151e-01 -6.43042982e-01
-5.63125789e-01 -7.83376276e-01 -8.01398635e-01 1.08925915e+00
6.67489529e-01 2.51316458e-01 -9.03029263e-01 -7.08292201e-02
-7.09383190e-02 -3.60321432e-01 8.24499130e-01 1.36895522e-01
1.30075765e+00 -4.77699161e-01 3.15202475e-01 3.10304552e-01
1.47533476e+00 1.10519278e+00 1.04730177e+00 4.34712946e-01
6.21027231e-01 2.37454280e-01 3.82394612e-01 5.25869668e-01
1.61002159e-01 1.26966923e-01 2.66936328e-02 -1.27056628e-01
-4.17031229e-01 2.40978137e-01 9.46760699e-02 1.09806371e+00
-6.32217467e-01 -1.57929003e-01 -1.25849521e+00 7.44384944e-01
-1.16772413e+00 -7.83798397e-01 -4.64807063e-01 1.70262015e+00
8.78964007e-01 -8.79676715e-02 1.70397371e-01 6.25810862e-01
7.80689895e-01 -5.41762188e-02 -4.30738807e-01 -8.64892006e-01
-4.33361351e-01 4.10338461e-01 5.37377179e-01 2.04994068e-01
-1.22269738e+00 9.18640316e-01 5.72340298e+00 4.05942440e-01
-1.87327409e+00 -2.90728122e-01 9.93892431e-01 7.54678071e-01
4.27352101e-01 -5.67864835e-01 -1.04189634e+00 3.10541391e-01
5.39354622e-01 1.45568460e-01 2.63014108e-01 1.04006505e+00
3.12978821e-03 1.22061275e-01 -8.95170152e-01 1.00277019e+00
-7.03881402e-03 -1.42460608e+00 6.54381573e-01 -1.01423338e-01
6.81475043e-01 -4.20425534e-01 1.68522447e-01 2.30987906e-01
2.79098392e-01 -1.36742663e+00 3.69463950e-01 2.88133174e-01
6.98318124e-01 -1.03601599e+00 1.17619562e+00 3.87995094e-01
-5.58550417e-01 -6.24827556e-02 -8.32463264e-01 -3.36722136e-01
-6.97333515e-01 2.22394183e-01 -9.86652911e-01 2.70984303e-02
7.11128831e-01 8.26667905e-01 -6.29907608e-01 9.80777800e-01
2.76166856e-01 6.89067662e-01 1.34457245e-01 -6.16898894e-01
8.32435727e-01 -2.27080345e-01 -3.14806193e-01 1.56350625e+00
4.05645370e-01 5.71218073e-01 -4.53368515e-01 4.75303829e-01
-3.23058039e-01 1.99220970e-01 -4.86410320e-01 -6.75438762e-01
1.93470448e-01 1.13769615e+00 -9.69518840e-01 -4.03465092e-01
-5.79110265e-01 1.12488687e+00 -2.85109073e-01 5.91277294e-02
-4.50698107e-01 -1.09972417e+00 2.88708270e-01 -1.40687183e-01
3.97401005e-01 -3.61118615e-01 -6.87095582e-01 -9.69031036e-01
-1.02783248e-01 -1.19749832e+00 1.21973358e-01 -5.66516221e-01
-1.09051001e+00 6.66188300e-01 -8.46390367e-01 -1.01657259e+00
2.24092826e-01 -1.30081618e+00 -5.02335310e-01 1.11083043e+00
-1.17274988e+00 -1.18038893e+00 -4.75482464e-01 7.38908350e-01
8.10678482e-01 -7.97415495e-01 7.86306381e-01 7.64124691e-01
-6.24941587e-01 7.02236950e-01 3.51184905e-01 1.11489952e+00
4.94713306e-01 -9.50507343e-01 4.89426732e-01 9.44732428e-01
1.94269754e-02 5.18370509e-01 1.75869584e-01 -4.28835303e-01
-1.47220469e+00 -1.35421491e+00 8.71799469e-01 -1.85694262e-01
2.74547428e-01 -1.82635300e-02 -8.73264551e-01 7.76824236e-01
3.63070726e-01 1.06870187e-02 3.02356780e-01 -4.26674604e-01
-2.78071553e-01 -3.18270713e-01 -1.43527436e+00 3.33092809e-01
2.92780042e-01 -3.28117281e-01 -5.97230971e-01 2.34113723e-01
-1.86262801e-02 -2.14885429e-01 -6.92682147e-01 6.32925481e-02
7.64930248e-01 -6.20914102e-01 6.76878572e-01 -8.88819993e-01
6.60583258e-01 -4.30672079e-01 -4.85954314e-01 -7.06922650e-01
-3.89092714e-01 1.20199092e-01 3.54209989e-01 1.39611709e+00
1.31598085e-01 -5.90095937e-01 1.03267431e+00 2.48952568e-01
1.44615173e-01 -5.35145819e-01 -3.92060429e-01 -9.74151611e-01
5.24423301e-01 3.55176330e-02 6.56507552e-01 1.39805841e+00
-5.53013027e-01 1.82205603e-01 -5.35139024e-01 2.54642069e-01
3.64497811e-01 -1.07157119e-01 4.80731577e-01 -9.26678717e-01
-5.96798435e-02 -4.41102386e-01 -9.00432348e-01 -1.04096496e+00
-5.93040347e-01 -7.16957331e-01 -3.33933204e-01 -1.64069474e+00
1.35052934e-01 -4.28792179e-01 -2.82612413e-01 6.91330135e-01
4.41624641e-01 5.21464586e-01 1.05140254e-01 1.37902692e-01
7.49695078e-02 5.12934849e-03 1.57383740e+00 -5.60173333e-01
1.70634501e-02 9.68954414e-02 -4.15739119e-01 4.67974812e-01
9.15895998e-01 -1.16410159e-01 7.47227818e-02 -8.92854571e-01
-3.84182185e-01 -8.51088539e-02 1.70692727e-01 -1.06006491e+00
4.77360576e-01 -5.27487248e-02 1.01285446e+00 -9.66397882e-01
-2.66872067e-02 -7.53369451e-01 -2.62015373e-01 6.88182652e-01
-3.55899841e-01 -5.86410798e-02 2.65546978e-01 -2.32688889e-01
-4.76370662e-01 -3.95221859e-01 1.13704383e+00 -2.00414404e-01
-1.10757709e+00 1.47231400e-01 -7.99625218e-01 -7.04970181e-01
9.50558066e-01 -7.35178351e-01 -2.33040124e-01 -1.06534228e-01
-6.36724293e-01 -3.08372766e-01 -1.17627248e-01 3.46865922e-01
9.81191814e-01 -1.27526081e+00 -9.61613297e-01 2.02418029e-01
-2.00169519e-01 -1.52160779e-01 8.69392604e-02 2.03636438e-01
-1.38876748e+00 8.08964729e-01 -1.03609014e+00 -1.79044336e-01
-1.46006763e+00 2.94681877e-01 3.46838504e-01 2.43926227e-01
-4.06647176e-01 8.22932541e-01 -6.79654598e-01 -5.45617819e-01
6.50490820e-01 -3.22142661e-01 -5.13565600e-01 -1.36768281e-01
6.87615097e-01 5.33194780e-01 3.32979888e-01 -3.28422636e-01
-3.91408831e-01 6.06577933e-01 -7.23899424e-01 2.38640979e-01
1.55354321e+00 4.82384115e-01 -5.00030100e-01 6.87004626e-02
1.03116536e+00 -1.68703824e-01 -6.96735382e-01 1.66196376e-01
4.12830710e-02 -6.03469074e-01 7.04962015e-02 -1.41957510e+00
-1.25735033e+00 1.30932224e+00 1.02614617e+00 6.81832954e-02
1.13783586e+00 -6.20647252e-01 6.48707747e-01 1.33556640e+00
2.77142912e-01 -1.04435110e+00 -1.12860873e-01 1.00277054e+00
1.02616763e+00 -1.45635629e+00 -5.04586473e-02 1.42562762e-01
-7.31192291e-01 1.84837425e+00 7.33234227e-01 -5.29210150e-01
7.50602067e-01 4.92869258e-01 5.95687866e-01 2.02480763e-01
-2.40830615e-01 1.65111125e-01 -8.19977373e-02 7.98101366e-01
7.92276740e-01 -9.08681154e-02 -3.77975017e-01 6.80667043e-01
-7.70274177e-02 3.07575077e-01 1.02850699e+00 1.04271936e+00
-5.65854967e-01 -1.13295782e+00 -6.05132222e-01 5.64584255e-01
-7.32617080e-01 -5.31222932e-02 -6.03176236e-01 1.01103842e+00
1.43026471e-01 5.33302009e-01 3.18617582e-01 -3.65569681e-01
3.76891434e-01 -1.09318145e-01 7.35295355e-01 -3.64834070e-01
-5.77456892e-01 -4.85491008e-01 -1.79571763e-01 2.70684451e-01
-2.27222577e-01 -3.11089575e-01 -1.16584218e+00 -7.74825692e-01
-5.98237738e-02 -2.08974630e-01 1.15807092e+00 6.04078948e-01
-9.88359824e-02 4.41865683e-01 7.23580360e-01 -7.08154365e-02
-6.77723765e-01 -1.14114606e+00 -7.80644238e-01 -1.63855940e-01
1.12524591e-01 2.17580169e-01 3.21704984e-01 4.20082271e-01] | [11.828791618347168, 2.6816489696502686] |
0138ecf5-b01c-4f55-9887-73df2eb90ff5 | unsupervised-vision-language-grammar | null | null | https://openreview.net/forum?id=N0n_QyQ5lBF | https://openreview.net/pdf?id=N0n_QyQ5lBF | Unsupervised Vision-Language Grammar Induction with Shared Structure Modeling | We introduce a new task, unsupervised vision-language (VL) grammar induction. Given an image-caption pair, the goal is to extract a shared hierarchical structure for both image and language simultaneously. We argue that such structured output, grounded in both modalities, is a clear step towards the high-level understanding of multimodal information. Besides challenges existing in conventional visually grounded grammar induction tasks, VL grammar induction requires a model to capture contextual semantics and perform a fine-grained alignment. To address these challenges, we propose a novel method, CLIORA, which constructs a shared vision-language constituency tree structure with context-dependent semantics for all possible phrases in different levels of the tree. It computes a matching score between each constituent and image region, trained via contrastive learning. It integrates two levels of fusion, namely at feature-level and at score-level, so as to allow fine-grained alignment. We introduce a new evaluation metric for VL grammar induction, CCRA, and show a 3.3% improvement over a strong baseline on Flickr30k Entities. We also evaluate our model via two derived tasks, i.e., language grammar induction and phrase grounding, and improve over the state-of-the-art for both. | ['Tinne Tuytelaars', 'Zilong Zheng', 'Wenjuan Han', 'Bo Wan'] | 2021-09-29 | null | null | null | iclr-2022-4 | ['phrase-grounding'] | ['natural-language-processing'] | [ 5.58557630e-01 3.65601599e-01 -9.83342081e-02 -5.07908344e-01
-1.42374933e+00 -7.91081429e-01 6.10399961e-01 1.85636163e-01
-2.34391421e-01 3.43175530e-01 5.10898232e-01 -3.53755176e-01
3.85306716e-01 -7.32148409e-01 -1.05701935e+00 -6.89453006e-01
1.55592725e-01 5.00407040e-01 2.31325582e-01 -3.64358872e-02
-1.06448727e-02 4.42914739e-02 -1.53220987e+00 8.41116309e-01
8.53328943e-01 8.90823662e-01 6.38960898e-01 7.08857358e-01
-2.61946857e-01 1.19347727e+00 5.67859970e-02 -5.99151194e-01
-1.99346058e-02 -3.99810970e-01 -1.09459007e+00 4.43919182e-01
1.22388768e+00 -9.80397463e-02 -9.56329796e-03 1.07608211e+00
1.39023319e-01 -2.14023605e-01 4.62738454e-01 -1.19503975e+00
-7.91415274e-01 6.26988411e-01 -6.10728383e-01 -3.57163072e-01
5.22298157e-01 2.50566542e-01 1.78358746e+00 -1.18222857e+00
9.20122266e-01 1.32884121e+00 2.84801841e-01 5.78649342e-01
-1.33452225e+00 -3.87109578e-01 6.05054975e-01 3.40376943e-02
-1.21150184e+00 -2.37620443e-01 6.58018589e-01 -7.04353213e-01
9.03540492e-01 -6.32936209e-02 4.05113012e-01 1.07030594e+00
-7.87745938e-02 1.18371403e+00 1.24820852e+00 -7.07935333e-01
-3.77075896e-02 -2.11733431e-01 1.93855315e-01 1.23018134e+00
7.65910000e-02 -1.11829164e-02 -7.03523219e-01 1.13471970e-01
7.66718626e-01 -4.37016696e-01 -1.40404850e-01 -7.07698226e-01
-1.46292901e+00 9.21815038e-01 5.92271745e-01 2.06964817e-02
-2.12270498e-01 4.47347879e-01 1.35589778e-01 -1.18242435e-01
4.30763476e-02 2.28071198e-01 -2.60413408e-01 3.39261591e-01
-7.95660019e-01 2.60450184e-01 6.83754444e-01 1.21520734e+00
1.07504582e+00 -4.54457283e-01 -4.97603118e-01 5.07129431e-01
8.02995145e-01 8.59887660e-01 -2.72165895e-01 -7.66076922e-01
7.79974282e-01 7.70045757e-01 -2.88995743e-01 -8.53215337e-01
-1.98508978e-01 -1.18382759e-01 -7.37741351e-01 3.31596956e-02
3.53239894e-01 1.93482712e-02 -1.30331671e+00 1.86915624e+00
1.30906314e-01 2.28144705e-01 1.52138203e-01 8.58647943e-01
1.27921271e+00 5.22136688e-01 5.09980440e-01 6.39875159e-02
1.58733833e+00 -1.24448514e+00 -3.86362910e-01 -4.34016377e-01
5.67770958e-01 -5.61281681e-01 1.53471327e+00 2.62660861e-01
-1.05742085e+00 -5.02028704e-01 -7.19757736e-01 -3.99492979e-01
-2.07192540e-01 1.66957527e-01 5.41865110e-01 2.60997415e-01
-1.28654277e+00 -1.75464720e-01 -5.59218943e-01 -5.57726741e-01
2.48738810e-01 2.40223318e-01 -5.40766835e-01 -2.21655965e-01
-9.70199347e-01 4.87350166e-01 4.36213076e-01 -1.32668793e-01
-1.11280680e+00 -2.40143582e-01 -1.25427175e+00 -1.50229961e-01
6.12750649e-01 -1.24344063e+00 1.17800450e+00 -9.29970205e-01
-9.64074492e-01 1.58708572e+00 -3.82118821e-01 -4.93682504e-01
-4.97754011e-03 -2.45408759e-01 -2.05173567e-02 4.11433846e-01
2.72724152e-01 1.12828565e+00 9.91953492e-01 -1.75057244e+00
-9.17201281e-01 -4.59822714e-01 2.67393351e-01 4.84189570e-01
2.52020627e-01 2.22909465e-01 -1.07230043e+00 -5.19969344e-01
1.01551279e-01 -9.68199849e-01 -2.64788419e-01 -4.37397331e-01
-5.08603573e-01 -3.59302521e-01 3.43903780e-01 -8.54299605e-01
1.15150058e+00 -1.89302707e+00 4.95566517e-01 1.71372235e-01
3.83069396e-01 -2.65515685e-01 -4.16615516e-01 3.21678758e-01
1.64502740e-01 2.50472963e-01 -4.32321787e-01 -6.77562118e-01
3.87421548e-02 4.50491726e-01 -4.82355326e-01 -1.05819851e-01
6.04600370e-01 1.38030338e+00 -1.17165577e+00 -8.16641927e-01
1.36432827e-01 1.60611436e-01 -5.92071950e-01 3.06607366e-01
-7.45404482e-01 4.93961275e-01 -4.66833800e-01 8.70876789e-01
3.01552176e-01 -6.28585100e-01 3.46641302e-01 -4.32749867e-01
9.13596451e-02 1.18688136e-01 -9.56980407e-01 2.04337835e+00
-4.88475978e-01 3.41598988e-01 -8.31790417e-02 -6.62313879e-01
7.88609147e-01 8.52898806e-02 1.10901028e-01 -7.68168211e-01
-2.44488075e-01 1.63729507e-02 -3.76947343e-01 -5.71170807e-01
4.17007565e-01 2.13713460e-02 -6.32097960e-01 1.26624793e-01
4.33390349e-01 -2.63128281e-01 1.21322013e-01 6.80358648e-01
9.41792548e-01 5.05377471e-01 3.07594538e-01 3.77467088e-02
6.10660911e-01 1.15850352e-01 2.19717532e-01 1.10173583e+00
6.82768524e-02 8.30250919e-01 4.25842702e-01 -3.42817158e-01
-8.23286116e-01 -1.24018586e+00 3.65345031e-01 1.50078464e+00
3.23346853e-01 -6.86510980e-01 -7.10522294e-01 -1.01393580e+00
-1.85748145e-01 6.50113046e-01 -6.39003754e-01 3.97297859e-01
-6.78693891e-01 -4.36099172e-01 4.80308682e-01 5.57344735e-01
6.03874087e-01 -1.32815695e+00 -2.34192282e-01 -6.43372238e-02
-5.82589149e-01 -1.92060804e+00 -5.13495088e-01 -5.13731465e-02
-5.26695848e-01 -1.07419968e+00 -1.50062159e-01 -1.12934065e+00
6.35821581e-01 1.38600722e-01 1.92065549e+00 1.74264312e-01
-8.06161538e-02 8.51853669e-01 -5.29598475e-01 -3.01556498e-01
-3.61104906e-01 -3.93334255e-02 -5.44131041e-01 1.38065130e-01
3.42891306e-01 -1.91155508e-01 -3.97017211e-01 4.76702023e-03
-7.27235794e-01 6.22064233e-01 8.84490609e-01 7.78888583e-01
1.13596189e+00 -4.24826354e-01 6.79500625e-02 -7.00430751e-01
3.27142566e-01 -3.21511738e-02 -6.29039407e-01 7.13940680e-01
-1.13900527e-01 3.39423835e-01 2.68912435e-01 1.06737085e-01
-1.00116229e+00 6.02381349e-01 2.29535028e-02 -3.22633505e-01
-5.04304171e-01 6.26843452e-01 -5.90144932e-01 -6.72630221e-02
3.15305054e-01 4.20094341e-01 -2.94054270e-01 -2.17921510e-01
1.12644815e+00 3.69348079e-01 9.36606586e-01 -9.24609542e-01
8.32794309e-01 4.76871878e-01 -5.91369718e-02 -7.45239854e-01
-1.34155393e+00 -8.25965703e-01 -9.21072662e-01 -2.84622423e-02
1.43042159e+00 -1.32428503e+00 -3.17202300e-01 1.86009869e-01
-1.26598001e+00 -4.67443347e-01 -2.37731576e-01 8.69844481e-02
-9.50544477e-01 3.66326332e-01 -2.42917076e-01 -5.27800083e-01
-3.41576636e-01 -1.12131977e+00 1.78043044e+00 -4.72904108e-02
1.66769415e-01 -9.72618461e-01 5.10325804e-02 7.09092617e-01
-4.39021826e-01 3.33990335e-01 9.09446716e-01 -3.78989935e-01
-1.17147708e+00 3.56777072e-01 -5.88628829e-01 2.62764275e-01
-1.83572173e-01 -3.85592952e-02 -9.15930510e-01 -7.99409002e-02
-3.66311520e-01 -7.11348355e-01 1.16716146e+00 2.55629450e-01
9.94429827e-01 -1.47901207e-01 -2.24982992e-01 6.42137289e-01
1.59481668e+00 -2.93442369e-01 7.85745323e-01 2.12611288e-01
1.27667999e+00 6.36258900e-01 7.14366913e-01 -1.10447355e-01
1.04581475e+00 7.63254881e-01 7.01322436e-01 -4.43988144e-01
-4.59239155e-01 -6.92144871e-01 3.90893906e-01 6.65799618e-01
-1.05485864e-01 -2.13465706e-01 -1.24570096e+00 7.34883368e-01
-2.13670015e+00 -9.23071861e-01 -1.92322999e-01 1.99625909e+00
7.46459961e-01 -6.89536408e-02 1.39966220e-01 -5.25753021e-01
5.19909799e-01 1.28183693e-01 -2.36543894e-01 -1.67081535e-01
-4.09754336e-01 1.18738219e-01 3.70128661e-01 6.88327909e-01
-1.40823841e+00 1.51704836e+00 5.75951338e+00 6.07012391e-01
-8.15109909e-01 -1.01001263e-02 4.03587639e-01 3.25203180e-01
-5.19085407e-01 3.53668928e-01 -1.06740141e+00 -6.64727688e-02
4.81998950e-01 2.79146492e-01 2.62010068e-01 4.88458186e-01
6.06708564e-02 -9.23978314e-02 -1.22770190e+00 1.17395377e+00
4.31812137e-01 -1.40197587e+00 6.32510900e-01 1.47566885e-01
8.05999517e-01 4.11728561e-01 -1.18988931e-01 2.70943075e-01
7.01283455e-01 -1.28021049e+00 7.27475405e-01 5.80396414e-01
8.80691230e-01 -4.12579149e-01 5.15044987e-01 2.41714239e-01
-1.53803086e+00 1.18696451e-01 4.33372781e-02 1.41698092e-01
3.47761482e-01 3.09992433e-01 -8.21861804e-01 8.99363399e-01
5.83584905e-01 6.51305437e-01 -8.22463214e-01 6.40464306e-01
-8.73332500e-01 6.02409840e-01 -3.19608711e-02 2.64502764e-01
5.26781201e-01 -3.53571884e-02 4.73714918e-01 1.38532472e+00
-7.26831555e-02 1.00633288e-02 7.82552481e-01 8.43397677e-01
-1.68416828e-01 2.36051962e-01 -7.80839384e-01 1.31615466e-02
2.25508690e-01 1.42859459e+00 -5.86249590e-01 -3.77885759e-01
-6.81111217e-01 9.93073821e-01 5.47854185e-01 4.46867138e-01
-8.40992451e-01 1.43214881e-01 2.95563757e-01 -1.09954730e-01
5.21573067e-01 -4.64353144e-01 -1.83924973e-01 -1.41690123e+00
7.69673884e-02 -1.05317461e+00 6.86710119e-01 -1.02617097e+00
-1.29917300e+00 5.98059535e-01 9.26832203e-04 -1.04312921e+00
-3.25522661e-01 -6.63313150e-01 -2.43592173e-01 8.11879694e-01
-1.74113202e+00 -2.22517943e+00 -5.51905215e-01 8.23057532e-01
5.80711067e-01 7.62104020e-02 9.29074109e-01 -1.53704837e-01
-2.06384927e-01 4.33705717e-01 -5.23966730e-01 5.18302798e-01
6.11808777e-01 -1.59069908e+00 5.33594668e-01 1.05970645e+00
9.93617773e-01 5.02402604e-01 3.55573773e-01 -5.90307474e-01
-1.40507746e+00 -1.33469284e+00 1.10489368e+00 -9.94379282e-01
7.31878638e-01 -7.49890983e-01 -6.64958239e-01 8.66100073e-01
3.12661529e-01 3.66865024e-02 5.48163176e-01 3.84600341e-01
-6.70981586e-01 1.33151054e-01 -6.40174747e-01 7.01772034e-01
1.30081844e+00 -8.37689459e-01 -9.14469004e-01 3.49236608e-01
1.04592514e+00 -3.90256077e-01 -6.89410090e-01 6.60020173e-01
4.47068036e-01 -8.38426709e-01 9.94399190e-01 -7.01273680e-01
5.54660678e-01 -5.18716276e-01 -5.57517886e-01 -7.72306085e-01
-3.35305393e-01 -5.82411528e-01 -8.74482989e-02 1.15223837e+00
6.08418822e-01 1.25920698e-02 7.20607817e-01 3.15596819e-01
-1.09784886e-01 -6.29460216e-01 -4.70166743e-01 -5.61025321e-01
-4.43130732e-02 -7.11405754e-01 1.50762588e-01 7.07177222e-01
-1.62354648e-01 1.01818275e+00 -1.96488142e-01 6.25481129e-01
9.30240631e-01 4.69326556e-01 1.01873219e+00 -9.72968936e-01
-2.14071929e-01 -1.63533121e-01 -4.45542932e-01 -1.27651680e+00
3.12265009e-01 -1.07537675e+00 2.60446906e-01 -2.04026508e+00
5.25728941e-01 -2.16932982e-01 -3.11221719e-01 7.75326192e-01
-3.40197802e-01 3.57601643e-01 4.50842083e-01 2.18456447e-01
-1.26823807e+00 3.75396550e-01 1.27914000e+00 -2.69667685e-01
-7.28319734e-02 -6.21481657e-01 -8.79433155e-01 9.92549181e-01
4.71062511e-01 -9.53575820e-02 -5.06693125e-01 -6.97964311e-01
3.62218112e-01 -4.98264134e-02 8.83478582e-01 -7.80754328e-01
2.80697912e-01 -2.53922284e-01 -7.19301850e-02 -8.53474855e-01
5.13846502e-02 -6.14155471e-01 -3.43095899e-01 -1.14284448e-01
-3.12902391e-01 -3.44398692e-02 8.79365727e-02 6.16870165e-01
-4.31148648e-01 9.98058766e-02 3.76073182e-01 -3.54085237e-01
-1.40546930e+00 3.79422247e-01 2.64220059e-01 3.48092973e-01
7.80663013e-01 -1.86268017e-01 -2.49314576e-01 -3.72266501e-01
-1.12989688e+00 5.47817349e-01 5.54465890e-01 5.01601398e-01
8.82160246e-01 -1.23169303e+00 -9.33922648e-01 1.03033096e-01
8.03760767e-01 2.24657685e-01 7.05838576e-02 7.14186668e-01
-1.97648913e-01 4.89147365e-01 1.82575524e-01 -1.12723494e+00
-1.51466084e+00 4.16311741e-01 1.63379401e-01 -7.10957646e-01
-5.85620105e-01 1.05075836e+00 8.63030314e-01 -3.93021166e-01
2.67814159e-01 -5.86166441e-01 -2.63553351e-01 -5.79092726e-02
4.24046874e-01 -6.16250873e-01 -7.72808492e-03 -9.24669743e-01
-3.32215846e-01 1.03430533e+00 6.78096665e-03 -2.46941984e-01
9.26323652e-01 -4.63388741e-01 -3.26128840e-01 3.75486165e-01
7.53285766e-01 4.06817235e-02 -1.23622441e+00 -4.90089148e-01
2.68568695e-01 -1.71240777e-01 -8.44958276e-02 -1.00285077e+00
-6.75813079e-01 9.38200116e-01 3.93565506e-01 8.01086985e-03
1.29674530e+00 6.31058156e-01 5.02797544e-01 4.97252941e-01
5.19944668e-01 -6.64075136e-01 4.16052461e-01 8.74536157e-01
1.00903296e+00 -1.56503093e+00 -2.80798078e-01 -7.45480776e-01
-9.77381706e-01 7.63781488e-01 5.36118150e-01 5.06748259e-02
5.86874001e-02 8.04168507e-02 1.31778881e-01 -4.18585956e-01
-8.30601096e-01 -9.91044044e-01 8.45782697e-01 7.01819181e-01
4.36197758e-01 1.06838241e-01 1.33614004e-01 4.35020208e-01
-1.13470731e-02 -8.98405015e-02 1.02480523e-01 6.04833245e-01
-5.04694641e-01 -1.16964841e+00 -2.23543316e-01 5.75187914e-02
-2.27566019e-01 -6.34027123e-01 -5.89129806e-01 8.62928271e-01
3.01987886e-01 9.09729481e-01 -3.22951339e-02 -3.21431696e-01
2.95252711e-01 6.42603934e-02 8.29635322e-01 -1.08782625e+00
-2.01273888e-01 1.70929432e-01 3.44273031e-01 -7.58148134e-01
-8.99116516e-01 -5.29701591e-01 -1.36142802e+00 5.92578173e-01
-1.13280028e-01 -1.59328043e-01 4.65645909e-01 1.00589991e+00
3.00637901e-01 2.85011232e-01 2.60617048e-01 -6.09506965e-01
-1.53682500e-01 -4.64966595e-01 -2.41995543e-01 7.87662685e-01
2.61507511e-01 -3.59371305e-01 -1.85096040e-02 6.57463193e-01] | [10.570443153381348, 1.489418864250183] |
e21111c5-5645-4925-99b7-b63fe2c85666 | sloper4d-a-scene-aware-dataset-for-global-4d | 2303.09095 | null | https://arxiv.org/abs/2303.09095v2 | https://arxiv.org/pdf/2303.09095v2.pdf | SLOPER4D: A Scene-Aware Dataset for Global 4D Human Pose Estimation in Urban Environments | We present SLOPER4D, a novel scene-aware dataset collected in large urban environments to facilitate the research of global human pose estimation (GHPE) with human-scene interaction in the wild. Employing a head-mounted device integrated with a LiDAR and camera, we record 12 human subjects' activities over 10 diverse urban scenes from an egocentric view. Frame-wise annotations for 2D key points, 3D pose parameters, and global translations are provided, together with reconstructed scene point clouds. To obtain accurate 3D ground truth in such large dynamic scenes, we propose a joint optimization method to fit local SMPL meshes to the scene and fine-tune the camera calibration during dynamic motions frame by frame, resulting in plausible and scene-natural 3D human poses. Eventually, SLOPER4D consists of 15 sequences of human motions, each of which has a trajectory length of more than 200 meters (up to 1,300 meters) and covers an area of more than 2,000 $m^2$ (up to 13,000 $m^2$), including more than 100K LiDAR frames, 300k video frames, and 500K IMU-based motion frames. With SLOPER4D, we provide a detailed and thorough analysis of two critical tasks, including camera-based 3D HPE and LiDAR-based 3D HPE in urban environments, and benchmark a new task, GHPE. The in-depth analysis demonstrates SLOPER4D poses significant challenges to existing methods and produces great research opportunities. The dataset and code are released at \url{http://www.lidarhumanmotion.net/sloper4d/} | ['Cheng Wang', 'Yuexin Ma', 'Siqi Shen', 'Hongwei Yi', 'Lan Xu', 'Chenglu Wen', 'Xiping Lin', 'Yitai Lin', 'Yudi Dai'] | 2023-03-16 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Dai_SLOPER4D_A_Scene-Aware_Dataset_for_Global_4D_Human_Pose_Estimation_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Dai_SLOPER4D_A_Scene-Aware_Dataset_for_Global_4D_Human_Pose_Estimation_CVPR_2023_paper.pdf | cvpr-2023-1 | ['camera-calibration', '3d-human-pose-estimation', 'human-scene-contact-detection'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [-3.72588247e-01 -3.05179864e-01 6.75123855e-02 -3.04900289e-01
-7.74388552e-01 -2.78351247e-01 2.31765926e-01 -5.44107676e-01
-5.26762903e-01 4.79798257e-01 2.55090177e-01 1.06794111e-01
1.89369082e-01 -6.44900322e-01 -6.89536572e-01 -2.84048259e-01
-1.72414541e-01 8.65790367e-01 3.52195084e-01 -3.67471069e-01
-1.79672644e-01 5.96936226e-01 -1.32624519e+00 -2.80327499e-01
3.46787959e-01 6.78639412e-01 3.10579807e-01 9.26688075e-01
5.29820859e-01 3.80984426e-01 -3.59873801e-01 -6.16053522e-01
5.56438386e-01 3.26622091e-02 -5.38219333e-01 5.17612875e-01
7.39709437e-01 -6.75551951e-01 -6.90512717e-01 7.39663184e-01
8.59804809e-01 3.94288599e-01 2.72907674e-01 -1.53823555e+00
3.47099937e-02 -1.85713321e-01 -8.66551876e-01 9.06784460e-02
8.97883296e-01 5.65458953e-01 5.53162217e-01 -1.21387124e+00
8.82773042e-01 1.49535656e+00 7.82735825e-01 4.35592711e-01
-9.54373062e-01 -7.43149638e-01 3.27930530e-03 1.42544657e-01
-1.94402039e+00 -5.49834192e-01 6.71517193e-01 -5.89103758e-01
8.77990246e-01 1.86093464e-01 1.21647274e+00 1.42903888e+00
6.41701818e-02 7.47979701e-01 4.52783406e-01 4.19948660e-02
5.84585965e-03 -3.36370528e-01 -1.92760050e-01 7.45135427e-01
3.96178126e-01 -3.57887186e-02 -8.46425414e-01 -7.74286315e-02
1.16069412e+00 1.38794824e-01 -1.53525576e-01 -6.11775875e-01
-1.75568545e+00 5.21066368e-01 4.81552124e-01 -3.22948933e-01
-2.00245529e-01 4.93841439e-01 2.93500990e-01 -1.97508633e-01
1.93919972e-01 -1.31576166e-01 -3.01014364e-01 -3.75997961e-01
-5.44169009e-01 7.89233446e-01 2.39306629e-01 1.65949523e+00
8.56275856e-01 -2.44597159e-03 1.52357191e-01 6.03357136e-01
3.10850143e-01 1.30758047e+00 5.99466302e-02 -1.56992269e+00
8.89067233e-01 4.15505528e-01 4.61091280e-01 -1.35143244e+00
-5.73504031e-01 4.05055135e-02 -9.00325418e-01 -2.47870177e-01
3.26276273e-01 -1.73704833e-01 -3.45683426e-01 1.52676511e+00
8.39188457e-01 2.95923531e-01 -4.72543836e-01 1.23392332e+00
6.52683139e-01 4.40338224e-01 -1.79212779e-01 1.77687947e-02
1.40207827e+00 -1.02966213e+00 -3.90624017e-01 -4.87928271e-01
5.12223244e-01 -6.22612894e-01 1.19511092e+00 1.49870828e-01
-1.03661752e+00 -7.57527232e-01 -6.29819036e-01 -3.21191400e-01
1.09206788e-01 9.53350142e-02 3.74885470e-01 3.56960684e-01
-7.64553189e-01 2.99124960e-02 -8.96493256e-01 -6.34696960e-01
1.91444650e-01 1.73958689e-01 -6.06771708e-01 -2.16612324e-01
-9.58428383e-01 7.80881882e-01 -4.15829942e-03 2.14043260e-01
-7.50499785e-01 -4.42710072e-01 -1.07790160e+00 -5.15050530e-01
5.95383346e-01 -1.10780716e+00 1.23834884e+00 1.45826235e-01
-1.25730860e+00 1.16642118e+00 -2.31882617e-01 -2.96247452e-01
1.03597486e+00 -8.76227379e-01 -2.06027970e-01 1.22185498e-01
4.59172279e-01 8.13354850e-01 4.86240417e-01 -1.02110100e+00
-5.79300582e-01 -5.62866747e-01 -2.25430951e-01 6.04797900e-01
1.79483056e-01 -5.24396636e-02 -1.22408211e+00 -6.18437707e-01
8.59677196e-02 -1.32555878e+00 -5.69303393e-01 1.02064967e-01
-5.90422690e-01 1.86951891e-01 6.72486782e-01 -5.14348328e-01
1.01301837e+00 -1.88975859e+00 2.55557537e-01 -3.95066431e-03
2.52621919e-01 -1.16469771e-01 1.37675077e-01 1.53463885e-01
4.52925920e-01 -2.10310489e-01 -1.49482876e-01 -8.25182080e-01
3.31301615e-02 1.97568849e-01 -3.38088050e-02 6.91045821e-01
-4.68500972e-01 1.24086833e+00 -9.18215036e-01 -6.94219410e-01
7.42203236e-01 5.21450222e-01 -5.86234510e-01 2.16593623e-01
2.06091776e-01 6.80544376e-01 -6.11082613e-01 8.41610134e-01
8.02297533e-01 -9.81361866e-02 -1.41931012e-01 -8.54381323e-02
-5.31788729e-02 -1.47355989e-01 -1.46371627e+00 2.10559011e+00
-6.81353286e-02 6.46446943e-01 -7.07351342e-02 -4.18423712e-01
5.97961903e-01 1.08671496e-02 8.06773484e-01 -4.79913086e-01
2.69192249e-01 -1.83583722e-02 -8.36738586e-01 -5.64492404e-01
7.24257231e-01 2.90224791e-01 -4.49800938e-01 1.03995651e-01
-2.72119999e-01 -3.31139088e-01 -2.09054928e-02 6.59119785e-02
1.01337612e+00 2.94705540e-01 4.75658894e-01 -3.03277858e-02
3.81133854e-01 1.01878159e-01 6.86094940e-01 4.57045197e-01
-5.26314557e-01 7.60483265e-01 -2.21829519e-01 -6.69632077e-01
-1.20497811e+00 -1.38861430e+00 8.74152929e-02 6.64291263e-01
5.75829864e-01 -5.86040437e-01 -7.62770474e-01 -9.78330821e-02
1.04391888e-01 1.84216559e-01 -2.12925583e-01 2.20972389e-01
-1.02742290e+00 -4.04313952e-01 5.47380865e-01 6.29192114e-01
1.07362473e+00 -7.96922028e-01 -1.05114257e+00 3.66336815e-02
-9.80291784e-01 -1.66956270e+00 -8.93238842e-01 -5.76662183e-01
-7.00220942e-01 -1.11894631e+00 -8.64896953e-01 -4.37840581e-01
3.84133905e-01 7.64871001e-01 1.22521055e+00 -1.51508018e-01
-4.92757827e-01 6.49789929e-01 -1.09468736e-01 -1.50589928e-01
4.39843684e-01 -1.68265775e-01 6.63465440e-01 -1.30730689e-01
3.56485993e-01 -5.63601196e-01 -8.07997406e-01 7.90969789e-01
7.26324096e-02 2.99576402e-01 1.41249225e-01 2.80321687e-01
8.60641718e-01 -1.60465211e-01 -2.34244645e-01 -3.73096615e-01
-5.34845442e-02 -3.52007359e-01 -6.81204259e-01 -2.17449158e-01
1.47307605e-01 -6.59209311e-01 2.05477490e-03 -4.51361597e-01
-8.94660592e-01 3.83359551e-01 -2.71470612e-03 -7.05833375e-01
-2.44466424e-01 -1.49740651e-02 -4.19726461e-01 1.86309963e-01
7.55843759e-01 -4.96532209e-02 -3.77402812e-01 -3.19266409e-01
4.32895571e-01 4.38430786e-01 1.03636789e+00 -6.06893301e-01
1.21754289e+00 7.69783378e-01 -1.26179317e-02 -1.15231061e+00
-6.13023698e-01 -7.62252212e-01 -1.13062739e+00 -5.80229640e-01
1.17752397e+00 -1.56201136e+00 -1.07592797e+00 7.24250853e-01
-1.09284341e+00 -6.15404665e-01 -2.33082533e-01 7.29892790e-01
-9.24812853e-01 3.71245921e-01 -5.12318075e-01 -8.17301154e-01
-1.14289895e-01 -1.22760117e+00 1.59070420e+00 -2.15609491e-01
-4.74262804e-01 -5.58504581e-01 1.31909788e-01 8.98301482e-01
-1.79184049e-01 6.41088665e-01 6.48377985e-02 2.99406946e-01
-8.92945588e-01 -1.22873694e-01 8.02103505e-02 -2.46756449e-01
-8.20091441e-02 -1.96751580e-01 -6.11670315e-01 -2.87429065e-01
-1.12424426e-01 -1.02871247e-01 1.60759196e-01 6.24430060e-01
7.96804011e-01 -1.08578332e-01 -4.61831927e-01 8.22233915e-01
1.01039267e+00 -1.90243408e-01 7.18589246e-01 4.45407957e-01
1.13128972e+00 4.94491637e-01 1.01072609e+00 7.10662186e-01
9.83057201e-01 1.30941379e+00 3.77459168e-01 1.64556786e-01
-4.92670462e-02 -6.23201191e-01 4.08258468e-01 7.55306065e-01
-5.44330060e-01 -8.56079236e-02 -1.16794682e+00 4.48018342e-01
-1.83007753e+00 -1.06681871e+00 -4.72652525e-01 2.33672929e+00
3.61874938e-01 8.17333013e-02 5.23955762e-01 -1.38290808e-01
7.45647728e-01 1.73591897e-01 -6.12591743e-01 4.92205709e-01
6.32166490e-03 -3.22288573e-01 5.97363472e-01 6.78715527e-01
-1.05877590e+00 1.12064099e+00 5.85647631e+00 5.63670397e-01
-5.92101753e-01 5.11670560e-02 1.45143658e-01 -3.72615665e-01
1.17850780e-01 -1.10838577e-01 -1.00623560e+00 4.41718042e-01
6.79348588e-01 -8.75101089e-02 3.61995429e-01 1.21624553e+00
5.86104751e-01 -2.59577304e-01 -9.58131492e-01 1.71344936e+00
-6.91208914e-02 -1.20647979e+00 -1.00827947e-01 4.91897434e-01
6.23293161e-01 2.27922142e-01 -1.01853594e-01 1.49977639e-01
3.04530203e-01 -8.12049568e-01 1.10962820e+00 4.35400903e-01
8.93516541e-01 -7.46880829e-01 3.73317868e-01 7.92924762e-01
-1.88795412e+00 9.26287174e-02 -4.78315085e-01 -1.09810114e-01
7.11602271e-01 4.46344018e-01 -4.18790370e-01 4.77002203e-01
1.34086406e+00 9.84181583e-01 -5.35637736e-01 7.56388485e-01
-1.77983344e-01 5.07472008e-02 -5.15736818e-01 3.29714328e-01
-1.99388996e-01 -3.68343592e-01 8.17779064e-01 1.17972219e+00
4.62601840e-01 6.13991857e-01 5.31342089e-01 4.17578548e-01
-5.50799593e-02 -7.85218626e-02 -8.60354125e-01 6.39277995e-01
6.71333671e-01 1.08978629e+00 -6.21401846e-01 -4.15083021e-01
-1.13758326e-01 1.22929549e+00 8.11306164e-02 2.06032753e-01
-1.25621355e+00 9.58925039e-02 1.04426193e+00 5.47157466e-01
-1.73175216e-01 -8.96328807e-01 -1.50461361e-01 -1.57850134e+00
2.59047925e-01 -6.85760438e-01 1.67399853e-01 -1.19580960e+00
-9.05004501e-01 3.51532251e-01 4.48857784e-01 -1.54359591e+00
-2.54223704e-01 -3.14630330e-01 -2.31059268e-01 5.23670316e-01
-7.78596222e-01 -1.18653858e+00 -9.50179935e-01 1.11403143e+00
8.85241568e-01 1.02694094e-01 3.98101121e-01 5.16475320e-01
-5.25958121e-01 3.19232851e-01 -5.21341324e-01 3.10884655e-01
6.62754893e-01 -8.11714709e-01 1.03892362e+00 8.52804005e-01
4.17188667e-02 4.65726256e-01 6.91436887e-01 -9.06647563e-01
-1.56126988e+00 -1.28119349e+00 9.28049326e-01 -1.26395929e+00
2.73299664e-01 -6.88099384e-01 -4.02602106e-01 1.04325902e+00
-4.32559878e-01 2.17249393e-01 2.53487766e-01 -2.40569338e-01
6.94079101e-02 4.81346101e-02 -1.05498230e+00 9.84169364e-01
2.01480961e+00 -2.90183812e-01 -4.74270433e-01 4.17313665e-01
8.13929617e-01 -1.09112906e+00 -7.21374631e-01 3.26819897e-01
5.55490017e-01 -9.78114724e-01 1.53837657e+00 -1.78478539e-01
-3.37185897e-02 -5.25369227e-01 -6.86283350e-01 -6.99770033e-01
-3.96113366e-01 -7.36002147e-01 -3.04799080e-01 6.96253419e-01
-2.54438847e-01 -2.45307177e-01 1.07498682e+00 7.73447275e-01
-2.28410438e-02 -5.01552761e-01 -1.07831609e+00 -7.60654449e-01
-4.01976943e-01 -9.64991748e-01 4.58488941e-01 5.73648810e-01
-4.72607762e-01 2.83381194e-01 -9.48220074e-01 3.34073722e-01
1.04443562e+00 -2.37347797e-01 1.80433142e+00 -1.06242228e+00
3.45877670e-02 1.26939714e-02 -7.01586008e-01 -1.65781987e+00
-2.74963100e-02 -3.63786489e-01 -1.46163153e-02 -1.26413500e+00
1.58097595e-01 -1.68965921e-01 6.80482030e-01 1.59413025e-01
-4.25211079e-02 3.69850099e-01 2.84679741e-01 4.74985987e-01
-6.04272485e-01 6.57840252e-01 1.15343022e+00 5.57825938e-02
-6.86228350e-02 1.37452364e-01 -1.44616619e-01 1.17634392e+00
3.72243643e-01 -2.63645828e-01 -4.14554328e-01 -6.81955397e-01
6.60012066e-02 1.88272610e-01 9.31626320e-01 -1.35214698e+00
3.03744465e-01 -6.35722101e-01 5.62277257e-01 -1.11839151e+00
9.73310888e-01 -8.89472604e-01 6.87590182e-01 5.58113098e-01
2.37893894e-01 4.51321214e-01 -7.90950879e-02 6.41761959e-01
1.69259325e-01 5.27018487e-01 5.89586020e-01 -4.49241340e-01
-9.96509135e-01 8.49326134e-01 -1.31297201e-01 2.72994280e-01
1.24822402e+00 -6.34107888e-01 7.13587999e-02 -4.91722494e-01
-6.29992902e-01 5.24400592e-01 7.22765028e-01 6.04799628e-01
8.32728922e-01 -1.77273929e+00 -6.64044142e-01 6.40731528e-02
2.62845457e-01 5.83906829e-01 5.90744615e-01 7.68278599e-01
-8.08789551e-01 3.30912322e-01 -1.16516769e-01 -1.13294458e+00
-1.37111282e+00 3.66681904e-01 3.29448491e-01 1.21306233e-01
-1.13914323e+00 6.89559281e-01 1.77526355e-01 -5.70871949e-01
9.15121138e-02 -2.36162245e-01 2.21756637e-01 -2.22795621e-01
4.50514346e-01 9.39354718e-01 -3.21828157e-01 -1.40750706e+00
-6.17556691e-01 1.26388323e+00 6.36380672e-01 -3.59257132e-01
1.09652460e+00 -7.89385796e-01 4.57653195e-01 3.44664484e-01
1.11709237e+00 2.08596945e-01 -1.67508101e+00 -3.39367658e-01
-3.54262263e-01 -9.32996273e-01 -5.46979249e-01 -2.96436585e-02
-8.95804584e-01 7.26574659e-01 4.76471812e-01 -6.69237018e-01
7.77625918e-01 6.08619116e-02 1.16308653e+00 5.36751330e-01
1.23983634e+00 -1.14672494e+00 4.17504817e-01 5.45864224e-01
8.66145730e-01 -1.26760709e+00 1.75382927e-01 -5.88744938e-01
-8.09697747e-01 6.77342236e-01 7.98945129e-01 -6.05812632e-02
4.40749347e-01 6.28288314e-02 3.10308952e-02 -2.06519559e-01
-2.42197350e-01 -1.30249232e-01 1.64597049e-01 8.08213711e-01
-1.65235445e-01 2.60920823e-01 3.55753064e-01 3.57573032e-01
-7.53021359e-01 -1.20158128e-01 2.55617291e-01 9.35567319e-01
-4.36351925e-01 -6.07939124e-01 -7.43894696e-01 -1.62151694e-01
2.33652115e-01 3.74065280e-01 -6.03549071e-02 1.02704823e+00
2.65205592e-01 9.36951339e-01 -7.48461261e-02 -6.22550070e-01
8.14940155e-01 -3.24197501e-01 4.16617870e-01 -5.09734452e-01
-8.83800834e-02 3.84368002e-02 -2.40423959e-02 -1.13493419e+00
-3.16440433e-01 -9.54632401e-01 -1.28298473e+00 -9.71683502e-01
9.71443132e-02 -3.76402736e-01 3.19202244e-01 6.57512605e-01
3.51674408e-01 8.11996404e-03 4.33299452e-01 -1.62223041e+00
1.75273083e-02 -6.06156051e-01 -3.84450763e-01 5.43160856e-01
1.99162498e-01 -9.20446992e-01 -8.86151567e-03 3.85275543e-01] | [7.050492286682129, -0.9239473342895508] |
76e1bc27-245a-4a29-bf20-6f1b50ec0a77 | sdc-stacked-dilated-convolution-a-unified-1 | null | null | http://openaccess.thecvf.com/content_CVPR_2019/html/Schuster_SDC_-_Stacked_Dilated_Convolution_A_Unified_Descriptor_Network_for_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Schuster_SDC_-_Stacked_Dilated_Convolution_A_Unified_Descriptor_Network_for_CVPR_2019_paper.pdf | SDC - Stacked Dilated Convolution: A Unified Descriptor Network for Dense Matching Tasks | Dense pixel matching is important for many computer vision tasks such as disparity and flow estimation. We present a robust, unified descriptor network that considers a large context region with high spatial variance. Our network has a very large receptive field and avoids striding layers to maintain spatial resolution. These properties are achieved by creating a novel neural network layer that consists of multiple, parallel, stacked dilated convolutions (SDC). Several of these layers are combined to form our SDC descriptor network. In our experiments, we show that our SDC features outperform state-of-the-art feature descriptors in terms of accuracy and robustness. In addition, we demonstrate the superior performance of SDC in state-of-the-art stereo matching, optical flow and scene flow algorithms on several famous public benchmarks.
| [' Didier Stricker', ' Christian Unger', ' Oliver Wasenmuller', 'Rene Schuster'] | 2019-06-01 | null | null | null | cvpr-2019-6 | ['stereo-matching'] | ['computer-vision'] | [ 7.13646784e-02 -9.52755153e-01 -2.26288944e-01 -3.55856925e-01
-2.07829848e-01 -2.20770687e-01 7.18629837e-01 -2.71004528e-01
-5.12851357e-01 5.25602520e-01 3.92194629e-01 3.23423184e-02
-2.30729277e-03 -9.32318628e-01 -5.47396302e-01 -6.80544674e-01
-1.30398944e-01 -1.73772112e-01 6.10981584e-01 -2.71156609e-01
6.34834051e-01 7.93228388e-01 -1.86253464e+00 2.66863793e-01
8.96075189e-01 1.17184424e+00 4.59963530e-02 6.26489520e-01
-6.70374259e-02 8.88306260e-01 -2.46734664e-01 -2.46355787e-01
6.08992457e-01 -1.11104980e-01 -8.31604362e-01 -3.89516532e-01
1.27107406e+00 -5.27783155e-01 -7.96424985e-01 1.02985775e+00
6.19464934e-01 4.04754162e-01 5.50187588e-01 -1.26375961e+00
-7.77301431e-01 -2.28790790e-02 -7.03706443e-01 4.87374038e-01
4.29797471e-01 2.68622547e-01 9.38595653e-01 -9.17998791e-01
6.89890265e-01 1.53132617e+00 7.92751789e-01 4.27730352e-01
-1.03501153e+00 -8.18437219e-01 -3.45705897e-02 3.06734920e-01
-1.32247615e+00 -5.68894029e-01 7.51048028e-01 -5.23711860e-01
1.27127266e+00 -5.44227436e-02 6.18184030e-01 8.60196650e-01
3.14679652e-01 6.11876845e-01 8.55789065e-01 -1.48274124e-01
-7.36213997e-02 -4.70017970e-01 1.67898461e-01 8.91983330e-01
2.90934324e-01 5.62893867e-01 -5.43002248e-01 7.99414143e-02
1.18091822e+00 1.35543123e-01 -3.57291371e-01 -2.90123910e-01
-1.30493915e+00 7.41379619e-01 9.40264642e-01 1.47752002e-01
-1.90902755e-01 6.42451525e-01 5.34505308e-01 1.73061803e-01
1.55059218e-01 2.57005811e-01 -9.56577957e-02 -1.90807819e-01
-6.66423798e-01 3.20428401e-01 4.99859035e-01 9.03083861e-01
1.11438501e+00 1.12451538e-01 -2.71849126e-01 9.58036005e-01
3.34066190e-02 7.40924954e-01 4.41499144e-01 -1.45253372e+00
5.00617266e-01 4.53481436e-01 -1.82357550e-01 -1.49169803e+00
-2.95486122e-01 -1.68642610e-01 -1.25092614e+00 4.35795993e-01
1.94556743e-01 2.51526088e-01 -6.92314208e-01 1.69310141e+00
2.18576696e-02 8.13102126e-01 -6.88093081e-02 1.07591546e+00
1.07342803e+00 5.67479730e-01 4.17353436e-02 3.91490251e-01
9.70739484e-01 -1.18925571e+00 -4.00118381e-01 -1.90086842e-01
4.09999579e-01 -8.52772534e-01 6.09175980e-01 -1.53040931e-01
-9.77212131e-01 -1.04266024e+00 -9.94008601e-01 -4.93518651e-01
-4.47315514e-01 -3.11732471e-01 1.08274829e+00 4.00373250e-01
-1.20569003e+00 8.93297315e-01 -5.09159029e-01 -3.03215414e-01
6.63903117e-01 4.36934024e-01 -5.69848418e-01 -1.84763923e-01
-1.05006456e+00 7.41790414e-01 4.76715751e-02 2.41533238e-02
-4.91457224e-01 -8.36001217e-01 -1.24747753e+00 1.12255648e-01
-3.42578679e-01 -1.02742279e+00 6.53044701e-01 -6.07584834e-01
-1.37964249e+00 1.05259979e+00 -5.41125298e-01 -2.88273782e-01
4.00825232e-01 -1.42378911e-01 -4.27856058e-01 3.26747149e-01
3.06326330e-01 1.19748783e+00 8.32639277e-01 -7.02212811e-01
-9.82593298e-01 -1.29469797e-01 -5.69259338e-02 -1.23391636e-01
-4.40911502e-02 6.34656101e-02 -4.95750487e-01 -6.79966807e-01
1.13548517e-01 -6.70318961e-01 -1.98156998e-01 5.02310216e-01
1.05261296e-01 -2.69520104e-01 9.17145848e-01 8.72965753e-02
1.14570355e+00 -2.12098098e+00 -6.79518357e-02 1.70287862e-01
3.78290147e-01 5.49919546e-01 -4.52386916e-01 -3.47883739e-02
-5.74685149e-02 -1.60908148e-01 1.52240833e-03 -3.13363224e-01
-1.15761593e-01 9.43971202e-02 -3.69473249e-01 7.62297869e-01
3.52946252e-01 8.99378479e-01 -9.52604175e-01 -7.04652131e-01
7.43556559e-01 6.14056766e-01 -7.74617255e-01 -7.55405053e-03
4.16983277e-01 3.38526756e-01 -3.66271049e-01 5.66916764e-01
1.17658830e+00 -2.67650902e-01 -4.24268961e-01 -1.96806043e-01
-3.73721957e-01 1.91990107e-01 -1.21341538e+00 1.95928574e+00
-4.91315454e-01 1.14580035e+00 -2.38116622e-01 -8.14660490e-01
1.05280924e+00 -1.62101701e-01 6.22159839e-01 -1.01673245e+00
1.21416762e-01 5.36821783e-01 -1.58361152e-01 -2.36388534e-01
5.04667461e-01 3.99200112e-01 2.91810811e-01 2.42923856e-01
1.58518672e-01 -1.11418143e-01 3.68480414e-01 8.14661267e-04
9.93043602e-01 -1.19516842e-01 2.28422970e-01 -6.75107479e-01
1.12358630e+00 -3.03979933e-01 8.93591225e-01 8.20311427e-01
-7.74691761e-01 8.60554934e-01 2.51176059e-01 -1.00333750e+00
-7.66415358e-01 -9.32162106e-01 -4.45953280e-01 7.65489042e-01
6.08029127e-01 -3.07096243e-01 -3.18514794e-01 -4.02364939e-01
5.84555686e-01 -2.77801275e-01 -7.00678885e-01 -1.03069693e-01
-1.00634003e+00 9.35467426e-03 6.60809875e-01 7.94698417e-01
1.16952932e+00 -1.07570839e+00 -7.48814344e-01 1.99531764e-01
-8.04696977e-02 -1.31430936e+00 -9.46158469e-01 -2.92238533e-01
-6.79861188e-01 -1.38651073e+00 -7.57452250e-01 -1.16785693e+00
4.51915443e-01 8.50013614e-01 1.29763484e+00 2.30221108e-01
-5.54747283e-01 2.36205712e-01 -4.68424894e-02 4.69521768e-02
1.71145320e-01 -1.17349915e-01 -1.34296760e-01 -6.51997104e-02
5.80093682e-01 -7.35288382e-01 -1.14846611e+00 4.79339510e-01
-7.42381334e-01 -9.99750495e-02 1.28339544e-01 9.11475539e-01
4.04681325e-01 -5.09635329e-01 1.23630911e-01 -3.67844343e-01
4.37205583e-01 9.26524848e-02 -9.43898320e-01 5.31458035e-02
-3.63281697e-01 1.36105016e-01 5.73869169e-01 -3.23852628e-01
-8.35172474e-01 8.20871964e-02 -1.28473312e-01 -5.34689963e-01
-1.44127488e-01 -2.74315953e-01 4.01966751e-01 -8.99642110e-01
6.29267931e-01 1.21736899e-01 8.21380392e-02 -2.30068043e-01
2.23046333e-01 5.89097917e-01 8.75228047e-01 -6.93686724e-01
6.50025010e-01 6.95371926e-01 5.26961923e-01 -6.56706572e-01
-6.03097856e-01 -8.43768239e-01 -7.91473448e-01 -2.31058016e-01
7.06252217e-01 -1.06950009e+00 -1.20109200e+00 1.02933931e+00
-1.46328998e+00 -8.42417628e-02 -9.47173759e-02 6.18235469e-01
-6.87131107e-01 2.91090101e-01 -7.99683452e-01 -2.62198716e-01
-4.60210741e-01 -1.41559386e+00 1.24189973e+00 4.81759310e-01
6.58266023e-02 -1.27022827e+00 3.38993639e-01 1.69252453e-03
8.58231187e-01 2.52901912e-01 4.71092373e-01 9.24048796e-02
-8.27541709e-01 1.46756113e-01 -8.51154149e-01 4.51885045e-01
1.96124792e-01 2.81951994e-01 -1.09681630e+00 -2.69040883e-01
-6.78750932e-01 -2.49693960e-01 1.51755834e+00 6.33888602e-01
1.35614872e+00 3.10251981e-01 -3.54738533e-01 1.49394333e+00
1.69619656e+00 1.06166489e-01 9.87175524e-01 4.46405649e-01
7.75226831e-01 3.92416716e-01 1.85080692e-01 4.22828496e-01
3.13463360e-01 5.43755949e-01 3.64084840e-01 -2.72205055e-01
-5.88038623e-01 -3.63619961e-02 1.90492999e-03 4.52844203e-01
-3.70887935e-01 9.35217887e-02 -7.97187030e-01 4.83159661e-01
-1.89400840e+00 -1.22595930e+00 -2.50851899e-01 1.99149966e+00
4.67964500e-01 -5.35940900e-02 -2.77138472e-01 -1.38873845e-01
8.47986937e-01 6.19226336e-01 -2.71934032e-01 -4.97859925e-01
-5.37473261e-01 6.48037195e-01 7.00835943e-01 6.55476034e-01
-1.55989027e+00 1.12036562e+00 7.19051027e+00 6.16769552e-01
-1.24849534e+00 -2.89327323e-01 4.33396816e-01 2.66554803e-01
-7.40741119e-02 -1.34193778e-01 -8.53346646e-01 4.05503273e-01
2.10818067e-01 -3.80755439e-02 3.42147440e-01 6.28243387e-01
8.81064907e-02 -9.75814983e-02 -1.12506795e+00 1.60612583e+00
1.53049566e-02 -1.98484373e+00 5.11793755e-02 -3.44538689e-03
1.05320811e+00 3.03333580e-01 6.10338859e-02 1.54343560e-01
2.35102892e-01 -1.06305277e+00 3.15859795e-01 4.12965417e-01
9.17528450e-01 -4.57329184e-01 7.29838073e-01 -4.05257881e-01
-1.59366858e+00 -1.68294720e-02 -6.69172704e-01 -2.17784181e-01
-4.58073840e-02 4.08678293e-01 3.41172457e-01 2.68828809e-01
7.84094930e-01 1.55375504e+00 -6.06633604e-01 1.41994357e+00
7.08960220e-02 -8.90256464e-02 -2.07206085e-01 1.65527344e-01
5.81262350e-01 -3.66743356e-01 2.39497781e-01 1.44227207e+00
1.87649846e-01 -1.81667805e-01 2.60229222e-02 9.16735768e-01
-1.85723901e-01 -7.93134868e-02 -8.47182572e-01 6.77846968e-01
5.04786372e-01 1.03265440e+00 -2.82861561e-01 -3.90387535e-01
-7.22642422e-01 9.30248022e-01 4.23800677e-01 3.49482358e-01
-6.00760400e-01 -8.36468220e-01 1.46953523e+00 -3.07215691e-01
3.18767309e-01 -1.49013713e-01 -2.03202829e-01 -1.29752398e+00
-1.94329787e-02 -3.86002511e-01 3.17133188e-01 -5.46046972e-01
-1.44451487e+00 4.67505842e-01 -3.96507829e-01 -1.57816911e+00
-1.93513855e-01 -9.70351219e-01 -9.44884598e-01 1.10284960e+00
-2.24653029e+00 -7.53812671e-01 -8.59310627e-01 1.09611619e+00
2.53722459e-01 -2.91990370e-01 6.90086663e-01 6.07670009e-01
-5.81741035e-01 4.59742039e-01 3.53364088e-02 5.38435102e-01
1.05084503e+00 -9.92078662e-01 7.54591584e-01 9.14849639e-01
-1.67977303e-01 6.34809196e-01 5.15143201e-02 -1.62820607e-01
-1.09196794e+00 -1.09580815e+00 9.08775091e-01 2.74025388e-02
5.42061388e-01 -3.96841951e-02 -8.09791148e-01 3.29566568e-01
1.49815947e-01 9.53795314e-01 2.54706383e-01 -2.09372398e-02
-8.58262181e-01 -5.29513001e-01 -1.04205048e+00 3.15590560e-01
1.48846209e+00 -7.12820947e-01 -4.19432342e-01 -6.39777957e-03
4.48494583e-01 -4.41164374e-01 -6.45243585e-01 4.90172863e-01
7.95862615e-01 -1.45253372e+00 1.42576492e+00 -5.66734552e-01
5.35424352e-01 -2.06266731e-01 -2.30960652e-01 -1.01981926e+00
-7.67119348e-01 -6.40107453e-01 1.06984429e-01 9.73151147e-01
-1.56956762e-01 -9.57068861e-01 7.31084943e-01 2.88386643e-01
-2.39685029e-01 -3.57686222e-01 -9.13319528e-01 -9.07386661e-01
2.05067396e-01 -2.62521744e-01 6.89757168e-01 8.45088363e-01
-2.53627986e-01 -7.79441325e-04 -2.09083945e-01 -1.77269891e-01
7.84533083e-01 3.69029105e-01 7.60155499e-01 -1.37114513e+00
2.76812941e-01 -8.69367003e-01 -1.02388310e+00 -1.47945416e+00
5.06994426e-01 -7.18631208e-01 -1.28040075e-01 -1.30473876e+00
1.09952554e-01 -4.73163038e-01 -4.66154188e-01 2.17348903e-01
-1.59613505e-01 4.17675912e-01 2.88944453e-01 2.76971489e-01
-6.94477737e-01 4.88118589e-01 1.65287805e+00 -2.73070574e-01
-1.58714920e-01 -2.16380715e-01 -2.89098620e-01 5.96573710e-01
6.45429671e-01 -3.27979624e-02 -1.06397137e-01 -7.32392609e-01
-3.27600449e-01 -2.57872432e-01 5.78734159e-01 -1.22483063e+00
4.10570771e-01 -1.87271699e-01 4.50860262e-01 -5.26930153e-01
1.86574876e-01 -6.37053251e-01 -2.88653046e-01 6.37563944e-01
-1.87792838e-01 9.76326019e-02 1.69511899e-01 1.41729653e-01
-6.94931448e-01 6.92488551e-02 1.08162010e+00 -2.42271647e-02
-1.35015166e+00 6.85993195e-01 1.18872598e-01 2.14808345e-01
8.54062080e-01 -4.31633413e-01 -8.37366819e-01 -1.83834359e-01
1.38809070e-01 3.47688735e-01 3.29929620e-01 5.98894775e-01
8.41676414e-01 -1.71022999e+00 -6.40616238e-01 6.59001052e-01
2.76462734e-01 -3.81191671e-02 1.95008472e-01 6.61523223e-01
-1.18118346e+00 7.90965557e-01 -7.53166080e-01 -8.02361369e-01
-9.73514199e-01 4.11796153e-01 5.12708545e-01 6.42133579e-02
-7.41450846e-01 7.59625733e-01 3.97466689e-01 -1.50441721e-01
3.28464955e-01 -4.00502056e-01 -1.28552899e-01 -2.57275432e-01
6.64010882e-01 6.55314147e-01 -2.13831037e-01 -8.24729085e-01
-6.40990078e-01 1.34472632e+00 2.63637662e-01 4.85825390e-01
1.17073274e+00 -9.23127979e-02 -2.30959713e-01 -3.31156075e-01
1.76959574e+00 -3.19967896e-01 -1.50462675e+00 -4.02078122e-01
-4.31637406e-01 -1.08468640e+00 1.53194219e-01 -1.47385910e-01
-1.48897326e+00 1.04119742e+00 6.97835863e-01 -3.46320570e-01
1.25252283e+00 -3.18035871e-01 9.62976158e-01 4.12761867e-01
2.88944185e-01 -1.05480421e+00 -2.42909510e-02 9.08940077e-01
6.76272452e-01 -1.63726842e+00 -1.55958310e-01 -5.98864257e-01
-1.56069264e-01 1.46046770e+00 7.95294344e-01 -5.87732852e-01
9.55988646e-01 3.28366399e-01 1.35197639e-01 -3.10095306e-02
-6.22448146e-01 -6.41688108e-01 5.20893276e-01 8.13812077e-01
3.78171533e-01 -2.16592565e-01 -1.46586195e-01 -3.84439677e-01
8.49634185e-02 7.60252625e-02 1.11887150e-01 7.62560368e-01
-4.70365852e-01 -9.21303630e-01 -6.34571761e-02 4.51765023e-02
-3.20764393e-01 -2.27786005e-01 8.70071352e-02 7.94085741e-01
2.20136434e-01 6.87766671e-01 5.22795975e-01 -3.90435219e-01
3.36766511e-01 -5.28554082e-01 4.73866493e-01 -2.87769129e-03
-5.63266516e-01 -2.60373533e-01 -3.03224623e-01 -1.23554099e+00
-8.24277937e-01 -3.64208698e-01 -9.82722521e-01 -8.67380381e-01
1.70501888e-01 -3.96869630e-01 2.28707656e-01 5.48283219e-01
4.77736890e-01 2.70462811e-01 8.47919226e-01 -9.20955420e-01
-1.13607228e-01 -6.05461955e-01 -5.47019243e-01 9.36258376e-01
7.63040602e-01 -8.39932024e-01 -3.83791417e-01 -6.20820560e-02] | [8.868021965026855, -2.08463716506958] |
ec09579f-4d13-4fbb-a612-aa3d4f3d2b75 | super-fan-integrated-facial-landmark | 1712.02765 | null | http://arxiv.org/abs/1712.02765v2 | http://arxiv.org/pdf/1712.02765v2.pdf | Super-FAN: Integrated facial landmark localization and super-resolution of real-world low resolution faces in arbitrary poses with GANs | This paper addresses 2 challenging tasks: improving the quality of low
resolution facial images and accurately locating the facial landmarks on such
poor resolution images. To this end, we make the following 5 contributions: (a)
we propose Super-FAN: the very first end-to-end system that addresses both
tasks simultaneously, i.e. both improves face resolution and detects the facial
landmarks. The novelty or Super-FAN lies in incorporating structural
information in a GAN-based super-resolution algorithm via integrating a
sub-network for face alignment through heatmap regression and optimizing a
novel heatmap loss. (b) We illustrate the benefit of training the two networks
jointly by reporting good results not only on frontal images (as in prior work)
but on the whole spectrum of facial poses, and not only on synthetic low
resolution images (as in prior work) but also on real-world images. (c) We
improve upon the state-of-the-art in face super-resolution by proposing a new
residual-based architecture. (d) Quantitatively, we show large improvement over
the state-of-the-art for both face super-resolution and alignment. (e)
Qualitatively, we show for the first time good results on real-world low
resolution images. | ['Georgios Tzimiropoulos', 'Adrian Bulat'] | 2017-12-07 | super-fan-integrated-facial-landmark-1 | http://openaccess.thecvf.com/content_cvpr_2018/html/Bulat_Super-FAN_Integrated_Facial_CVPR_2018_paper.html | http://openaccess.thecvf.com/content_cvpr_2018/papers/Bulat_Super-FAN_Integrated_Facial_CVPR_2018_paper.pdf | cvpr-2018-6 | ['face-hallucination'] | ['computer-vision'] | [ 2.83577174e-01 2.81277090e-01 1.85420215e-01 -5.20776808e-01
-1.30723298e+00 -2.10148856e-01 4.41496253e-01 -8.57003033e-01
-9.28898901e-02 5.24851382e-01 5.18387079e-01 5.63638568e-01
4.75434447e-03 -5.69776356e-01 -7.25479901e-01 -2.99632996e-01
-1.02242909e-01 4.82265472e-01 -1.34856906e-02 -3.98461133e-01
-1.07180901e-01 7.42036402e-01 -1.64922965e+00 3.30926329e-01
5.08179903e-01 7.95902312e-01 -4.86872308e-02 3.75116557e-01
3.84230793e-01 3.80893230e-01 -1.10844240e-01 -6.23238981e-01
5.66634655e-01 -4.81485337e-01 -8.27103138e-01 2.56840169e-01
1.32242954e+00 -6.32277489e-01 -1.21906370e-01 1.14367378e+00
7.41123617e-01 -4.63995226e-02 3.18386048e-01 -6.58956349e-01
-7.11727619e-01 2.45200813e-01 -1.28863835e+00 6.51090592e-02
4.41852152e-01 -1.59154445e-01 6.63381457e-01 -1.34304476e+00
8.59928548e-01 1.68160772e+00 8.11721802e-01 9.47083056e-01
-1.40664184e+00 -9.20949399e-01 -5.08688502e-02 -1.23873599e-01
-1.68296647e+00 -1.14892781e+00 7.77924597e-01 -2.46743426e-01
6.10854268e-01 -4.15892042e-02 1.43734589e-01 1.00213528e+00
-2.93557227e-01 2.66537696e-01 1.19829822e+00 -3.34040403e-01
-5.09383045e-02 -2.41595507e-01 -6.05971694e-01 1.00348639e+00
-2.75851618e-02 2.00216979e-01 -6.81651115e-01 2.35021804e-02
1.36544323e+00 -8.63960013e-02 -2.66519845e-01 -2.18926251e-01
-8.17065656e-01 5.40167212e-01 5.66365182e-01 3.41696918e-01
-4.17728007e-01 -2.28347462e-02 -1.18402362e-01 -7.05746189e-02
8.50889921e-01 2.08561286e-01 -2.57830828e-01 1.37594700e-01
-1.32053077e+00 2.37328172e-01 1.88685432e-01 6.03323996e-01
6.94670022e-01 2.69213170e-01 7.66453594e-02 1.11268663e+00
2.65587062e-01 4.46428120e-01 1.65467456e-01 -1.24998868e+00
4.14476335e-01 1.16397657e-01 1.53800175e-01 -8.97403479e-01
-5.59058130e-01 -4.09236878e-01 -8.45133424e-01 5.95541418e-01
3.39202166e-01 -1.75235227e-01 -1.03114164e+00 2.23950481e+00
4.79516059e-01 4.12733406e-01 -1.38315365e-01 1.07919836e+00
8.57690632e-01 3.30941558e-01 -2.24583000e-01 -4.19236213e-01
1.48945916e+00 -8.88715863e-01 -7.37032652e-01 -1.46404535e-01
-2.20297024e-01 -8.12359095e-01 8.87022972e-01 3.29724938e-01
-1.69600260e+00 -8.56598735e-01 -8.32662642e-01 -2.52830952e-01
-1.17033897e-02 1.96269065e-01 3.10017258e-01 3.52981389e-01
-1.64427137e+00 4.72372323e-01 -8.11667979e-01 -4.13697958e-01
7.78941035e-01 3.04397225e-01 -8.37626398e-01 -2.15316936e-01
-8.79871786e-01 7.06925809e-01 -2.93618917e-01 6.69325516e-02
-8.83774817e-01 -1.00066769e+00 -8.46876323e-01 5.63358404e-02
4.41939205e-01 -7.81434298e-01 1.07291043e+00 -1.03968537e+00
-1.58676434e+00 1.02649808e+00 -5.37276268e-01 -1.71776302e-02
4.46365863e-01 -4.03711230e-01 -5.48351705e-01 3.79329711e-01
1.48465142e-01 8.75039756e-01 1.02426863e+00 -1.61292946e+00
-5.33379018e-01 -7.99843132e-01 -2.57073462e-01 7.39534050e-02
-1.82165757e-01 3.06751281e-01 -6.33032024e-01 -5.54804981e-01
-4.47947811e-03 -6.10177934e-01 -1.08644115e-02 1.51978612e-01
-1.67756543e-01 8.52222294e-02 7.91193187e-01 -9.59938586e-01
8.28412771e-01 -1.98885965e+00 2.65237957e-01 -7.44965449e-02
4.18091208e-01 3.13513577e-01 -3.96280557e-01 -1.94837496e-01
-4.49072361e-01 1.64107993e-01 -2.72947103e-01 -8.30966711e-01
-2.70103186e-01 -2.43499026e-01 -2.18641490e-01 4.94924217e-01
3.89145225e-01 8.39436352e-01 -6.10723138e-01 -3.49081576e-01
7.06816539e-02 1.17265761e+00 -5.16284823e-01 8.11386704e-02
9.41281021e-02 7.31136382e-01 -7.49848709e-02 7.07869947e-01
1.05189908e+00 -1.96122572e-01 4.42059189e-02 -6.11356556e-01
-1.14436179e-01 -2.00938791e-01 -1.14693582e+00 1.88745236e+00
-4.64707196e-01 3.73739958e-01 5.32018185e-01 -3.12130183e-01
8.15751374e-01 4.52434987e-01 5.96317053e-01 -9.46311176e-01
-2.78696567e-01 5.95660172e-02 -3.36624593e-01 -3.09526891e-01
3.49318624e-01 -2.74032176e-01 5.20529091e-01 4.50781405e-01
2.88026094e-01 2.41097078e-01 -8.18099231e-02 -3.63000557e-02
6.44336045e-01 4.61804688e-01 1.98394418e-01 -1.16510019e-01
4.71055269e-01 -6.22586250e-01 5.17203331e-01 1.25085995e-01
-1.04148686e-02 1.18058813e+00 3.44916403e-01 -3.24804902e-01
-1.27420211e+00 -1.06678188e+00 -3.35734606e-01 1.18748951e+00
-2.05296382e-01 -3.99277717e-01 -1.05978227e+00 -5.22951126e-01
-1.98322207e-01 -1.73771959e-02 -1.10365570e+00 3.81976932e-01
-7.40206301e-01 -6.95128500e-01 4.46922719e-01 7.04504311e-01
6.63174927e-01 -6.94740236e-01 -2.45619386e-01 -2.62978613e-01
-1.52395248e-01 -1.47510636e+00 -6.10867858e-01 -4.51715082e-01
-6.54309392e-01 -8.21705520e-01 -7.62793362e-01 -6.24682307e-01
7.70026386e-01 2.91646451e-01 1.21536458e+00 3.49100679e-02
-3.67169350e-01 3.16581845e-01 4.13967706e-02 8.17366689e-02
5.96617758e-02 -1.47812560e-01 1.75325647e-01 2.89648473e-01
-1.48673728e-01 -7.89527357e-01 -6.90238178e-01 4.58482683e-01
-6.02991641e-01 4.62122932e-02 6.75927937e-01 6.68241918e-01
8.46347690e-01 6.80645183e-02 5.33506274e-01 -8.53978217e-01
1.92984492e-01 -2.51055807e-01 -6.84866250e-01 1.32760018e-01
-4.44570154e-01 -9.48946178e-03 5.80003738e-01 -2.01677382e-02
-1.34603405e+00 1.48752794e-01 -3.98996621e-01 -5.98611653e-01
-1.48732260e-01 -1.37963787e-01 -3.05347830e-01 -2.80861139e-01
6.62554622e-01 -1.46899089e-01 1.78395882e-01 -8.16866755e-01
4.34213817e-01 3.02902877e-01 9.84130859e-01 -3.70326608e-01
1.06736267e+00 9.42641139e-01 3.88580889e-01 -7.51208305e-01
-9.41632867e-01 -1.24311000e-01 -9.98710811e-01 -2.68131960e-02
9.23390090e-01 -1.14323390e+00 -6.83178842e-01 1.66259602e-01
-8.89841616e-01 -3.49731594e-01 2.16811169e-02 9.03453752e-02
-5.83840251e-01 1.40919000e-01 -6.38871014e-01 -6.84263110e-01
-3.74184549e-01 -1.12314737e+00 1.60262144e+00 4.69607860e-01
1.04651310e-01 -8.53593290e-01 1.26203507e-01 5.10056496e-01
5.67956507e-01 5.36592007e-01 4.04137373e-01 1.34269549e-02
-5.35354733e-01 1.60252854e-01 -5.66550791e-01 1.17279543e-02
1.61446050e-01 2.03203149e-02 -1.38617229e+00 -4.71462667e-01
-2.52540350e-01 -3.98473561e-01 9.31487799e-01 3.47825795e-01
1.04628527e+00 -2.18541369e-01 -5.50144985e-02 1.06020832e+00
1.47843921e+00 -3.41523677e-01 8.33783388e-01 -1.13705201e-02
8.84284437e-01 8.08654785e-01 5.10116160e-01 1.65761486e-01
4.02006805e-01 1.34738684e+00 4.20439392e-01 -6.00116909e-01
-6.70925319e-01 -4.47795421e-01 2.45100707e-01 2.00999931e-01
-5.67287803e-01 4.29138958e-01 -5.66430092e-01 4.47641760e-01
-1.81507766e+00 -1.17093277e+00 3.40139002e-01 2.18792105e+00
9.00573075e-01 -4.07415986e-01 4.13213015e-01 -3.17080587e-01
6.21473730e-01 2.81408548e-01 -4.90888447e-01 6.44883439e-02
-1.38798684e-01 6.53778911e-01 1.31595775e-01 7.79193401e-01
-1.05801940e+00 1.08147335e+00 6.34790134e+00 8.11300337e-01
-1.19784045e+00 3.73954445e-01 8.55082333e-01 -4.46200222e-01
-8.25519562e-02 -4.83680665e-01 -9.69476879e-01 1.62857071e-01
6.64774299e-01 1.71144471e-01 7.48610198e-01 7.44287550e-01
1.36199966e-01 1.11713894e-01 -1.06070638e+00 1.20301163e+00
4.77244645e-01 -1.18318808e+00 5.15739759e-03 1.72805190e-01
9.93723452e-01 -1.48668125e-01 4.56978619e-01 -3.26691903e-02
1.90021455e-01 -1.63315344e+00 5.34338415e-01 6.41425073e-01
1.42486799e+00 -1.02768803e+00 5.54957449e-01 -2.14250460e-01
-1.39977646e+00 1.15054421e-01 -4.67138529e-01 4.41174150e-01
1.46772921e-01 3.67466062e-01 -3.82088095e-01 6.46565616e-01
8.82812619e-01 6.85585082e-01 -7.63377309e-01 5.28656721e-01
-2.25885689e-01 2.22999424e-01 -1.69675037e-01 1.05590570e+00
-1.57881945e-01 -3.85658443e-02 3.50524992e-01 1.13461637e+00
2.21559182e-01 2.11559445e-01 -1.44761026e-01 1.02692366e+00
-4.12933469e-01 -1.71000466e-01 -4.16538566e-01 4.46762204e-01
4.18192297e-01 1.77720082e+00 -3.79392296e-01 9.28229764e-02
-3.38254362e-01 9.41014826e-01 5.76824784e-01 4.71344173e-01
-6.66729569e-01 4.20409888e-02 8.04199696e-01 4.56569880e-01
4.86194998e-01 -8.13640933e-03 -1.73295841e-01 -8.96842480e-01
1.05288856e-01 -8.03433418e-01 2.52346873e-01 -8.82970691e-01
-1.21789718e+00 9.44712937e-01 -8.98819640e-02 -8.94654334e-01
-5.31839311e-01 -3.26041043e-01 -3.20443630e-01 1.05013955e+00
-1.70424020e+00 -1.53217816e+00 -5.41213155e-01 7.74789631e-01
4.63408530e-01 -2.52268046e-01 7.08360553e-01 4.13779080e-01
-6.78077698e-01 1.01624668e+00 -2.69340068e-01 2.79210120e-01
9.94691491e-01 -9.76880252e-01 5.47861516e-01 1.01055908e+00
1.68898687e-01 7.10189104e-01 4.37983483e-01 -5.24313211e-01
-1.30611002e+00 -1.25081992e+00 5.33663750e-01 -5.36014080e-01
1.60865694e-01 -4.67137873e-01 -9.13727641e-01 6.61091149e-01
7.50728175e-02 4.68124628e-01 3.23144615e-01 1.77736729e-01
-5.72614908e-01 -3.15086842e-01 -1.67130065e+00 5.51524818e-01
1.34225249e+00 -6.04042470e-01 -2.12055236e-01 6.44287765e-02
4.58661824e-01 -6.59231603e-01 -1.02068484e+00 5.36422372e-01
8.56612921e-01 -1.29719329e+00 1.27596450e+00 -3.46978605e-01
6.94496751e-01 -4.26717192e-01 -1.82178214e-01 -1.14565825e+00
-4.98968422e-01 -7.42479086e-01 -1.39152810e-01 1.31459928e+00
2.12707862e-01 -4.11309659e-01 7.46600330e-01 3.89795005e-01
2.48086765e-01 -1.05557084e+00 -1.01135659e+00 -3.49945337e-01
-1.33575588e-01 7.90805742e-02 7.96726406e-01 8.67309213e-01
-4.55144793e-01 3.33619535e-01 -5.92756867e-01 3.03177238e-01
1.04386294e+00 1.20190261e-02 8.83391678e-01 -1.14544201e+00
-4.26338427e-02 -3.87224644e-01 -2.60356694e-01 -8.09580922e-01
2.04775468e-01 -4.67224598e-01 -1.09791219e-01 -1.20240819e+00
5.68357229e-01 -5.48083521e-02 1.17614761e-01 6.06959820e-01
-1.88723430e-01 1.06448817e+00 1.43475816e-01 1.78808048e-01
-5.28416216e-01 4.59525347e-01 1.24743199e+00 5.34047425e-01
6.45665824e-02 -2.89879024e-01 -1.10484076e+00 7.74642110e-01
4.34094638e-01 -3.18808407e-02 -2.08524108e-01 -4.89386797e-01
1.05416454e-01 2.18120471e-01 4.48586762e-01 -8.87994707e-01
3.42785381e-02 4.23201770e-02 9.96364355e-01 -3.43980432e-01
6.47234440e-01 -7.45913267e-01 1.73149914e-01 -1.07940175e-01
-2.92878121e-01 5.07701300e-02 3.39409202e-01 4.21449065e-01
5.02441637e-03 4.42636520e-01 1.45066822e+00 -9.87368152e-02
-5.57934880e-01 5.42680264e-01 5.49309671e-01 5.17843701e-02
7.44895995e-01 -2.57698923e-01 -5.28977811e-01 -5.44163883e-01
-7.55950987e-01 -4.80324887e-02 7.85378158e-01 4.24895644e-01
7.22964048e-01 -1.47653198e+00 -1.12768579e+00 5.69003940e-01
-2.39102706e-01 1.49817606e-02 3.78054529e-01 8.79718065e-01
-1.68637037e-01 2.88896054e-01 -4.46286172e-01 -5.21657169e-01
-1.59469604e+00 3.04347605e-01 6.33993208e-01 -1.64772749e-01
-6.15013063e-01 9.12828803e-01 4.64101672e-01 -2.35547975e-01
1.50291726e-01 2.98088163e-01 -3.14976364e-01 -5.69095984e-02
9.91352856e-01 3.46053004e-01 1.26925796e-01 -1.07360339e+00
-3.82603616e-01 1.30477881e+00 -2.52141327e-01 -3.93611372e-01
1.61410403e+00 -2.90735453e-01 -1.65584683e-01 -2.74348948e-02
1.22709477e+00 3.59134316e-01 -1.62209153e+00 -3.64004701e-01
-4.79540676e-01 -8.81272256e-01 1.06130466e-01 -9.84763443e-01
-1.63407171e+00 6.98212385e-01 9.42675829e-01 -4.15836066e-01
1.43404818e+00 1.10126682e-01 6.19730115e-01 -2.58505166e-01
5.02457857e-01 -8.15627456e-01 2.69652426e-01 1.59505248e-01
1.16016924e+00 -1.25381398e+00 2.83279896e-01 -4.96320426e-01
-5.85540295e-01 1.01475346e+00 6.95418835e-01 -4.14903974e-03
4.74061996e-01 3.58067364e-01 9.70127899e-03 -3.44040304e-01
-6.17177367e-01 -3.68230999e-01 5.19739687e-01 7.02391565e-01
6.32387877e-01 -2.26844817e-01 2.47534990e-01 4.89398032e-01
-2.80480385e-01 2.90263481e-02 1.61280379e-01 4.06925797e-01
-2.37731636e-01 -8.46824348e-01 -6.79588139e-01 1.02999516e-01
-6.60247922e-01 -1.12849688e-02 -4.31570947e-01 7.66562521e-01
8.52918550e-02 8.53479385e-01 1.73094898e-01 -3.52537215e-01
3.60579222e-01 -2.24435940e-01 8.49180341e-01 -4.80565935e-01
-2.78889686e-01 2.52116472e-01 -1.57196343e-01 -1.05615282e+00
-3.62396419e-01 -5.57179272e-01 -9.71814454e-01 -4.11699176e-01
4.79360931e-02 -2.23870590e-01 5.70215046e-01 6.37262702e-01
6.62496030e-01 4.93074656e-01 8.51297081e-01 -1.36331058e+00
-2.67688662e-01 -1.07952332e+00 -5.61989844e-01 4.18340862e-01
6.61642134e-01 -8.48880529e-01 -2.60444909e-01 2.17931699e-02] | [12.801054954528809, -0.1183665469288826] |
08d1844c-3c91-4ef8-a6ad-154f7e02e294 | subspace-hybrid-beamforming-for-head-worn | 2303.08967 | null | https://arxiv.org/abs/2303.08967v1 | https://arxiv.org/pdf/2303.08967v1.pdf | Subspace Hybrid Beamforming for Head-worn Microphone Arrays | A two-stage multi-channel speech enhancement method is proposed which consists of a novel adaptive beamformer, Hybrid Minimum Variance Distortionless Response (MVDR), Isotropic-MVDR (Iso), and a novel multi-channel spectral Principal Components Analysis (PCA) denoising. In the first stage, the Hybrid-MVDR performs multiple MVDRs using a dictionary of pre-defined noise field models and picks the minimum-power outcome, which benefits from the robustness of signal-independent beamforming and the performance of adaptive beamforming. In the second stage, the outcomes of Hybrid and Iso are jointly used in a two-channel PCA-based denoising to remove the 'musical noise' produced by Hybrid beamformer. On a dataset of real 'cocktail-party' recordings with head-worn array, the proposed method outperforms the baseline superdirective beamformer in noise suppression (fwSegSNR, SDR, SIR, SAR) and speech intelligibility (STOI) with similar speech quality (PESQ) improvement. | ['Thomas Lunner', 'Vladimir Tourbabin', 'Jacob Donley', 'Patrick A. Naylor', 'Pierre Guiraud', 'Alastair H. Moore', 'Sina Hafezi'] | 2023-03-15 | null | null | null | null | ['speech-enhancement'] | ['speech'] | [ 4.82533902e-01 -3.90698880e-01 6.56379580e-01 -1.73865799e-02
-1.26596236e+00 -3.88460219e-01 3.48286688e-01 -2.32450128e-01
-4.78543043e-01 2.74576902e-01 1.16247642e+00 -1.13606356e-01
-5.50442159e-01 -9.54813957e-02 -1.60657391e-01 -1.36970508e+00
9.01138708e-02 -5.59768319e-01 -6.76437393e-02 -3.12174231e-01
-1.03067428e-01 4.05449420e-01 -1.46619987e+00 2.97236264e-01
6.95687175e-01 1.01069033e+00 6.32514477e-01 9.02420998e-01
4.27503556e-01 4.22226906e-01 -5.81955731e-01 -2.63361126e-01
4.02943581e-01 -4.39847261e-01 2.49831185e-01 1.14894006e-02
4.82818075e-02 -9.91992876e-02 -4.23985928e-01 1.21805096e+00
1.43991745e+00 4.33767319e-01 3.05721611e-01 -5.72994649e-01
-1.56204194e-01 4.35703933e-01 -6.21713638e-01 3.67671400e-01
5.31238914e-01 5.54948896e-02 8.29148233e-01 -1.17800105e+00
1.96159884e-01 1.44082224e+00 9.45180058e-01 2.25377262e-01
-1.28672767e+00 -8.75651300e-01 -3.52434397e-01 5.73879257e-02
-1.38403463e+00 -1.13822591e+00 9.48578298e-01 -1.12822674e-01
8.93086374e-01 7.00420797e-01 3.27405006e-01 1.12987554e+00
-1.26546279e-01 6.39621198e-01 1.41113782e+00 -4.47917819e-01
3.55631322e-01 -1.93450108e-01 -1.16875932e-01 -4.67907824e-02
7.63891563e-02 5.84951699e-01 -7.98019469e-01 -2.48933032e-01
4.79047298e-01 -5.42388678e-01 -7.63596773e-01 1.74129874e-01
-1.04297340e+00 3.51439983e-01 -2.50066787e-01 6.17479622e-01
-7.56093979e-01 -4.07528400e-01 1.18758842e-01 3.99506569e-01
4.54959184e-01 2.26979300e-01 -4.88082230e-01 1.98882259e-02
-1.12474036e+00 1.94259897e-01 6.39481843e-01 7.09189951e-01
-6.50593499e-03 4.79346216e-01 -5.39804280e-01 1.43090045e+00
8.70554268e-01 9.99436259e-01 5.13680816e-01 -7.80446887e-01
5.68298042e-01 -4.37964201e-01 7.08418712e-02 -8.62194061e-01
-4.32640642e-01 -1.12991738e+00 -1.09778309e+00 2.00554073e-01
-2.72107329e-02 -4.27313983e-01 -7.80260324e-01 1.65916932e+00
3.03801805e-01 1.76046193e-01 3.01891685e-01 1.04776263e+00
8.08807075e-01 7.42459714e-01 -3.60926078e-03 -9.08719718e-01
1.20862615e+00 -6.15710914e-01 -1.35888195e+00 -2.05837578e-01
-2.02701941e-01 -1.15581477e+00 3.27716202e-01 9.53591347e-01
-1.31517637e+00 -7.55719900e-01 -1.09259713e+00 2.92109400e-01
2.72874624e-01 6.13068268e-02 7.24959821e-02 1.27620435e+00
-1.30607343e+00 1.64704338e-01 -5.45374453e-01 1.76363647e-01
2.80620232e-02 2.06332177e-01 -4.67321098e-01 -8.72332603e-02
-9.42170143e-01 7.77504683e-01 -1.29188463e-01 4.07832712e-01
-7.48220742e-01 -6.39592767e-01 -9.15896356e-01 1.73720848e-02
-1.11183427e-01 -4.14387733e-01 9.85700130e-01 -8.24516952e-01
-1.85213232e+00 2.67071456e-01 -5.01664519e-01 -2.94934392e-01
2.52445906e-01 -4.76888657e-01 -1.10275757e+00 -1.18344594e-02
-3.70019406e-01 -1.74647141e-02 1.36624765e+00 -1.41770804e+00
-4.81523752e-01 -4.68629152e-01 -8.71718347e-01 5.15807688e-01
9.44030937e-03 4.96960998e-01 -2.83385813e-01 -1.27576721e+00
6.35254502e-01 -4.62905824e-01 -2.40949318e-01 -6.13022506e-01
-3.98898870e-01 3.52362424e-01 5.10401547e-01 -1.51966906e+00
1.54307532e+00 -2.88076806e+00 3.80485475e-01 4.12917972e-01
-2.03869626e-01 5.82482755e-01 -4.81275082e-01 1.50354534e-01
-4.89710599e-01 -3.27765942e-01 -2.59284109e-01 -6.28084958e-01
-9.81143266e-02 -1.50628150e-01 -2.63793617e-02 7.92720735e-01
-1.90041587e-01 1.70260176e-01 -6.62316144e-01 -1.98749676e-01
3.94805700e-01 1.03363574e+00 -4.58418459e-01 4.46303576e-01
6.78144872e-01 4.74487126e-01 1.55110180e-01 4.97099906e-01
1.20131075e+00 8.93787384e-01 4.67503816e-02 -5.26000142e-01
-3.79809380e-01 6.61352426e-02 -1.85981023e+00 1.69546366e+00
-4.81271744e-01 4.61977899e-01 1.04525864e+00 -3.61059457e-01
9.66546357e-01 7.78555512e-01 2.50502139e-01 -7.11658895e-01
4.74506766e-02 3.84681374e-01 2.71588236e-01 -4.73253727e-01
1.25198543e-01 -2.54484683e-01 4.54290390e-01 1.07895099e-01
3.28007072e-01 -4.70038243e-02 -3.08552265e-01 -2.14671642e-01
1.20361018e+00 -3.12133372e-01 3.88226539e-01 -2.66064465e-01
7.75887489e-01 -9.53038871e-01 6.78817689e-01 7.10319102e-01
-2.71015048e-01 9.60266113e-01 -3.00313115e-01 4.11093742e-01
-8.01850855e-01 -1.30831599e+00 -1.21959157e-01 1.10340846e+00
-2.81392723e-01 -9.59776416e-02 -7.86648631e-01 2.47703806e-01
-3.04544508e-01 1.06913912e+00 -4.63994183e-02 1.30006626e-01
-5.67144871e-01 -9.33036864e-01 3.62537503e-01 9.39740911e-02
2.66291708e-01 -8.38801861e-01 -1.16894931e-01 3.77033442e-01
-3.56425881e-01 -8.17431509e-01 -6.30915046e-01 4.93951052e-01
-5.88984072e-01 -3.75492156e-01 -7.97207892e-01 -5.52789211e-01
4.62116212e-01 4.12406355e-01 2.75022596e-01 -5.23545325e-01
2.71721154e-01 3.70135933e-01 -5.90068758e-01 -3.93335938e-01
-4.77113813e-01 -7.82626987e-01 3.60074639e-01 5.96714914e-01
-4.98411879e-02 -8.32592189e-01 -6.77978635e-01 2.46064499e-01
-7.41095066e-01 -2.69977599e-01 9.75507915e-01 8.81624162e-01
3.52899402e-01 5.35714030e-01 5.57518780e-01 4.84837778e-03
1.15886080e+00 -2.55043507e-01 -3.33575130e-01 -1.59600675e-01
-3.13242435e-01 -1.85398504e-01 2.68274188e-01 -6.15378201e-01
-1.83110189e+00 5.35375997e-02 -8.73762250e-01 -2.32137278e-01
-2.69522071e-01 3.38197023e-01 -8.65267992e-01 -5.13103567e-02
6.04515135e-01 5.01194298e-01 -8.25121328e-02 -9.38535094e-01
2.91346967e-01 1.14773679e+00 8.81581366e-01 1.37924850e-01
8.90676379e-01 2.53878474e-01 -1.88276038e-01 -1.27931237e+00
-2.22069606e-01 -9.25178647e-01 -4.29958373e-01 -8.07421952e-02
6.95627034e-01 -1.10662758e+00 -2.28381008e-01 9.17757332e-01
-1.25941443e+00 -1.62382815e-02 -1.05175115e-02 9.20266271e-01
-1.71452552e-01 5.29549181e-01 -4.30014789e-01 -1.47413802e+00
-7.03314364e-01 -1.13065219e+00 8.48615110e-01 4.04312275e-02
-2.60188848e-01 -5.47477305e-01 1.25052109e-01 3.57270986e-01
8.16878438e-01 -2.82558531e-01 4.35616463e-01 -7.04393506e-01
1.24728002e-01 -2.52874605e-02 3.23411524e-01 8.95063579e-01
1.18415534e-01 -7.02004313e-01 -1.44739032e+00 -4.18560714e-01
7.39928901e-01 4.82062370e-01 5.73098540e-01 1.03744161e+00
5.79773784e-01 -3.66769999e-01 3.40558924e-02 7.27086067e-01
1.29169941e+00 6.51979327e-01 9.87024009e-01 -4.56289984e-02
1.39991030e-01 2.61449724e-01 2.89009869e-01 5.30990362e-01
-2.45368958e-01 6.45298481e-01 8.03628042e-02 -2.34294564e-01
-5.91482341e-01 1.24017559e-01 6.43404007e-01 1.32301283e+00
-7.75215626e-02 -4.29976076e-01 -3.71571004e-01 4.75507975e-01
-1.29686117e+00 -1.02143514e+00 -2.51849741e-01 2.35804391e+00
7.45391726e-01 -8.65921229e-02 -1.03821710e-01 8.18143845e-01
7.07939923e-01 2.51902044e-01 -1.65740967e-01 -5.12266636e-01
-7.95583308e-01 7.04800665e-01 5.12920320e-01 7.01054275e-01
-1.09361684e+00 2.74420977e-01 5.74354887e+00 1.11398566e+00
-7.39409685e-01 6.00497663e-01 2.23152891e-01 -4.50515673e-02
-1.60597414e-01 -5.56852341e-01 -3.70429724e-01 3.00846159e-01
1.02386367e+00 2.64045238e-01 8.06411982e-01 2.91500777e-01
8.06219757e-01 -1.99488774e-01 -4.67054427e-01 1.37572205e+00
2.25162148e-01 -5.22116840e-01 -5.18540084e-01 -1.76726446e-01
3.84764522e-01 -1.50968373e-01 2.13768154e-01 -1.47863895e-01
-3.56735364e-02 -8.20978105e-01 9.51628745e-01 6.25578344e-01
6.96070611e-01 -7.21537530e-01 7.12124407e-01 2.84821898e-01
-1.01389372e+00 -5.61241031e-01 1.45278648e-02 3.78613889e-01
2.11372390e-01 8.95908535e-01 -4.59908932e-01 6.80270135e-01
6.22472584e-01 3.78680937e-02 1.38954267e-01 1.33903193e+00
-4.51907128e-01 7.99534798e-01 -3.14155906e-01 3.34661126e-01
-3.91560107e-01 -1.01480886e-01 1.35696113e+00 1.72236121e+00
3.56323391e-01 5.81580043e-01 -5.84496975e-01 3.70471418e-01
3.22623879e-01 -1.47201659e-04 -3.29109132e-02 4.51614082e-01
6.30494654e-01 1.29301322e+00 -3.38842049e-02 3.10492098e-01
-3.31020832e-01 9.03197229e-01 -7.32156157e-01 6.66371107e-01
-3.80284756e-01 -4.98237729e-01 7.86230147e-01 -2.04887822e-01
4.54204679e-01 -2.24835694e-01 -4.40801263e-01 -6.16348922e-01
-1.01624891e-01 -1.32462287e+00 5.46202697e-02 -1.00883043e+00
-9.46136713e-01 6.60142839e-01 -3.72496516e-01 -1.15523052e+00
-7.76201636e-02 -3.86373162e-01 -4.78290021e-01 1.39503467e+00
-1.22053790e+00 -8.91399264e-01 3.08002532e-01 6.70028031e-01
6.44303560e-01 -2.12543249e-01 7.96764970e-01 7.91857183e-01
-4.89927083e-01 7.57200241e-01 3.58485222e-01 -1.10145301e-01
5.25809228e-01 -9.80307877e-01 7.82869384e-02 1.39790714e+00
-1.37533456e-01 6.00802362e-01 1.23918986e+00 -6.51041985e-01
-1.66432083e+00 -1.05007517e+00 8.52594674e-01 2.66174525e-01
3.79138887e-01 -4.12314206e-01 -6.84521258e-01 1.03486411e-01
4.27809477e-01 -3.68131042e-01 8.25447857e-01 -2.69553035e-01
-1.48209721e-01 -3.92297953e-01 -1.29599762e+00 4.50450957e-01
9.70735490e-01 -4.02126640e-01 -6.83357358e-01 3.36301848e-02
5.80184519e-01 -3.02377641e-01 -7.40047038e-01 3.99662226e-01
6.28808141e-01 -8.68037522e-01 1.18938065e+00 9.43551213e-02
-2.49978110e-01 -4.95755523e-01 -5.09707630e-01 -1.68783796e+00
-7.18564212e-01 -1.29728770e+00 1.15514034e-02 1.52345622e+00
5.16178906e-01 -5.08771062e-01 -7.18230829e-02 7.15720430e-02
-4.95695174e-01 -1.06199741e-01 -1.13683653e+00 -6.79871738e-01
-6.39868855e-01 -8.69168639e-01 1.73183769e-01 4.84417796e-01
-1.93796888e-01 9.01329443e-02 -6.49357915e-01 6.94888949e-01
9.49660897e-01 -7.44673550e-01 3.06952119e-01 -7.94819772e-01
-6.43537343e-01 -3.23589236e-01 -1.25084817e-01 -9.62325096e-01
-3.16600233e-01 -3.24357480e-01 3.88283968e-01 -1.39392030e+00
-2.31918409e-01 1.30632043e-01 -5.66934288e-01 -2.00370830e-02
-2.25464344e-01 -1.46613091e-01 1.29356191e-01 -2.45383531e-01
-1.60797417e-01 4.02674347e-01 9.21271205e-01 3.81090585e-03
-5.70035934e-01 3.78069997e-01 -7.23083854e-01 6.41150951e-01
2.90751755e-01 -4.18153226e-01 -2.43337691e-01 -3.68682534e-01
-2.16470972e-01 5.28221607e-01 6.81462735e-02 -1.22153950e+00
5.39884090e-01 4.17961866e-01 3.67544293e-01 -6.14818215e-01
7.53857553e-01 -8.15017939e-01 6.29279733e-01 2.98702478e-01
-2.52157569e-01 -2.84634650e-01 2.28821471e-01 6.07241452e-01
-2.32135952e-01 9.79822576e-02 1.05552995e+00 2.78834432e-01
-2.69716352e-01 -3.13687176e-01 -6.98854446e-01 -5.25435567e-01
3.75041783e-01 -2.41237953e-01 7.59740695e-02 -5.79554796e-01
-8.68202090e-01 -4.01989907e-01 -4.57967162e-01 1.66916937e-01
9.94321823e-01 -1.04363108e+00 -1.17494190e+00 3.71594757e-01
-5.69819927e-01 -3.64277989e-01 6.67627037e-01 1.16626918e+00
1.22134104e-01 2.72497475e-01 4.53338772e-02 -3.11697274e-01
-1.46626544e+00 2.56701648e-01 3.38347882e-01 -9.41470265e-03
-3.98903072e-01 1.25033736e+00 -4.79100132e-03 8.69933367e-02
4.64362025e-01 2.72434228e-03 -4.04145479e-01 -1.32549733e-01
8.46465349e-01 6.92998052e-01 4.87841636e-01 -1.07612681e+00
-2.02131540e-01 2.71454632e-01 4.27915037e-01 -7.27538049e-01
1.54283452e+00 -4.66526747e-01 -1.39051184e-01 5.69250323e-02
1.12339938e+00 7.56251037e-01 -8.64223599e-01 -1.55205160e-01
-1.24219097e-01 -5.22086263e-01 7.14089870e-01 -1.26796770e+00
-1.03943157e+00 7.20963657e-01 1.15997398e+00 -8.91987905e-02
1.97155261e+00 -5.81048608e-01 6.50973737e-01 -2.33497247e-01
8.32828432e-02 -9.64830935e-01 -1.80405691e-01 2.09575891e-01
1.32963037e+00 -6.21539891e-01 -1.72250554e-01 -3.33836854e-01
-4.31845844e-01 8.67164671e-01 -2.06776410e-01 3.22741330e-01
9.32448685e-01 7.00053215e-01 2.53807604e-01 3.21537256e-01
-2.95992166e-01 -2.92478859e-01 4.29802865e-01 9.00368989e-01
2.61769354e-01 6.87767193e-02 -3.38197589e-01 1.15482068e+00
-2.66331643e-01 -4.30357248e-01 2.88343383e-03 5.38209617e-01
-5.60845137e-01 -8.75599980e-01 -1.00331819e+00 2.56780833e-01
-6.28222227e-01 -4.41651434e-01 -2.06082277e-02 6.54890165e-02
1.51635706e-01 1.89385378e+00 -2.28072762e-01 -7.36671805e-01
7.51296043e-01 -2.45135613e-02 1.43507585e-01 -1.63327441e-01
-9.25385416e-01 1.13958347e+00 2.79379845e-01 -3.66202682e-01
-3.87101859e-01 -9.75424170e-01 -5.75754762e-01 1.19776517e-01
-5.54763436e-01 -5.28935604e-02 1.08066952e+00 8.48384798e-01
2.98714072e-01 6.74585640e-01 7.50476062e-01 -8.08382332e-01
-5.57856619e-01 -1.33804095e+00 -8.34193408e-01 4.59300280e-01
6.97138190e-01 -3.33945155e-01 -7.53645003e-01 -1.92966759e-02] | [14.990835189819336, 5.8549957275390625] |
71da79a9-dc81-4b54-849e-0979166dae95 | progress-measures-for-grokking-via | 2301.05217 | null | https://arxiv.org/abs/2301.05217v2 | https://arxiv.org/pdf/2301.05217v2.pdf | Progress measures for grokking via mechanistic interpretability | Neural networks often exhibit emergent behavior, where qualitatively new capabilities arise from scaling up the amount of parameters, training data, or training steps. One approach to understanding emergence is to find continuous \textit{progress measures} that underlie the seemingly discontinuous qualitative changes. We argue that progress measures can be found via mechanistic interpretability: reverse-engineering learned behaviors into their individual components. As a case study, we investigate the recently-discovered phenomenon of ``grokking'' exhibited by small transformers trained on modular addition tasks. We fully reverse engineer the algorithm learned by these networks, which uses discrete Fourier transforms and trigonometric identities to convert addition to rotation about a circle. We confirm the algorithm by analyzing the activations and weights and by performing ablations in Fourier space. Based on this understanding, we define progress measures that allow us to study the dynamics of training and split training into three continuous phases: memorization, circuit formation, and cleanup. Our results show that grokking, rather than being a sudden shift, arises from the gradual amplification of structured mechanisms encoded in the weights, followed by the later removal of memorizing components. | ['Tom Lieberum', 'Jacob Steinhardt', 'Jess Smith', 'Lawrence Chan', 'Neel Nanda'] | 2023-01-12 | null | null | null | null | ['memorization'] | ['natural-language-processing'] | [ 5.67000270e-01 2.32242405e-01 3.41437072e-01 -5.20332903e-03
2.08370328e-01 -7.69758582e-01 1.02391636e+00 2.24197358e-01
-4.15403247e-01 6.09734833e-01 2.41743103e-01 -2.99683422e-01
-4.03432995e-01 -6.51131272e-01 -1.02678835e+00 -8.57904315e-01
-3.21548790e-01 1.27646923e-01 1.65388193e-02 -4.87827957e-01
3.82389307e-01 3.15070152e-01 -1.47342336e+00 1.41727269e-01
9.69543457e-01 5.00412822e-01 -1.74442768e-01 6.17365837e-01
8.24900642e-02 8.12875569e-01 -4.16290760e-01 -7.94234499e-02
3.24412137e-01 -9.39976454e-01 -7.87457526e-01 -9.09047425e-02
1.84166014e-01 -2.71259844e-02 -3.33934069e-01 8.27603221e-01
4.18169014e-02 2.48123612e-02 7.43865311e-01 -1.18156791e+00
-9.62859452e-01 9.79546845e-01 -3.43890786e-01 5.05541921e-01
-7.03264400e-02 2.53953636e-01 9.66532588e-01 -7.15130687e-01
5.89695334e-01 7.86240220e-01 8.47567916e-01 5.30766487e-01
-1.65803611e+00 -5.92699766e-01 8.60445350e-02 -9.89141837e-02
-1.17458344e+00 -6.69414461e-01 7.15425193e-01 -7.36859500e-01
9.26800549e-01 -1.62432529e-02 1.18069708e+00 8.30082238e-01
3.39510053e-01 4.17094558e-01 1.03736472e+00 -6.92330062e-01
3.29914212e-01 -2.44399137e-03 2.24395484e-01 1.05241323e+00
6.27073407e-01 2.46810719e-01 -7.15029538e-01 4.94185369e-03
9.65067148e-01 1.10069141e-02 -5.70715368e-01 -4.41166848e-01
-1.13498044e+00 4.69543189e-01 3.18296731e-01 4.12202418e-01
-2.04623163e-01 4.14669335e-01 2.11103037e-02 8.37179780e-01
2.58002073e-01 1.06020629e+00 -4.43416655e-01 -4.59571369e-02
-7.98619151e-01 -1.40598014e-01 6.09620929e-01 3.99205863e-01
7.63083398e-01 1.32488191e-01 4.75287914e-01 3.99793357e-01
-8.47048014e-02 1.82273746e-01 7.12147415e-01 -9.35728133e-01
-5.08976430e-02 8.05255949e-01 -1.98543027e-01 -6.10038936e-01
-4.89499241e-01 -7.59014964e-01 -9.01782393e-01 2.11150929e-01
6.05059028e-01 -2.16323063e-01 -7.26932526e-01 2.19892216e+00
-2.33250320e-01 -9.44666490e-02 -4.83827628e-02 4.26085949e-01
4.47693691e-02 4.62847233e-01 -2.00526938e-01 -2.29043245e-01
9.59081411e-01 -5.36370814e-01 -2.54143983e-01 -1.08689470e-02
8.09413195e-01 -5.66997826e-02 1.11652696e+00 3.61995190e-01
-1.24597085e+00 -2.41493255e-01 -1.36786079e+00 2.62700468e-01
-4.26527262e-01 -1.75275832e-01 7.16133833e-01 5.11647522e-01
-1.26000619e+00 1.14143538e+00 -7.71390557e-01 -4.29146081e-01
3.63582581e-01 4.71701890e-01 -3.64040196e-01 4.10674393e-01
-9.59207058e-01 7.35880673e-01 1.48367599e-01 -1.09617531e-01
-6.94170952e-01 -7.28452384e-01 -4.79886919e-01 4.41885740e-01
-5.65092750e-02 -9.28455114e-01 1.12478364e+00 -1.32516885e+00
-1.49910569e+00 8.18788171e-01 -4.94703092e-02 -5.05608499e-01
5.16249463e-02 -2.51670796e-02 -2.97944732e-02 2.66594198e-02
-3.54539365e-01 3.91818941e-01 1.05979121e+00 -1.22128892e+00
-1.66172281e-01 -3.66354883e-01 -2.10786629e-02 -5.58790416e-02
-6.46535575e-01 -4.10672575e-01 5.45606136e-01 -5.67434907e-01
3.44490260e-01 -9.03983712e-01 1.01400651e-01 -3.08774531e-01
-9.86008644e-02 1.54712349e-01 1.80589437e-01 -2.63168514e-01
1.13033879e+00 -2.23087621e+00 4.22187746e-01 4.71133888e-01
7.68716216e-01 -1.62316307e-01 -2.90026009e-01 6.40426099e-01
-3.38767916e-01 3.59117240e-01 -4.31313872e-01 -2.34239027e-02
-6.34036437e-02 -1.65012814e-02 -6.99551940e-01 4.63411361e-01
3.82657707e-01 1.04182470e+00 -8.95402074e-01 1.48836434e-01
-3.48283052e-01 2.79378176e-01 -6.43030047e-01 -2.83859581e-01
6.33294322e-03 4.03148144e-01 3.50086428e-02 1.32981881e-01
1.88290298e-01 -5.20791709e-01 3.03316176e-01 2.59559929e-01
-3.63191873e-01 4.70231235e-01 -5.99663079e-01 1.40224004e+00
-3.18171501e-01 9.97201562e-01 -5.17951362e-02 -1.37530208e+00
5.91793120e-01 1.68035120e-01 4.47945863e-01 -6.36747539e-01
3.39581043e-01 2.85176247e-01 6.16190493e-01 -1.94918141e-01
1.09126195e-01 -2.77454793e-01 9.03124437e-02 1.10607195e+00
4.50085104e-01 -8.81012902e-02 2.03850225e-01 3.03101063e-01
1.49657989e+00 -1.96329281e-01 2.06441805e-01 -6.93890274e-01
3.14797200e-02 -5.69163412e-02 1.65787205e-01 8.32281530e-01
8.75867680e-02 1.18023701e-01 8.36566329e-01 -5.10214031e-01
-1.34019172e+00 -1.22723186e+00 5.01565635e-02 1.16709685e+00
-1.46280482e-01 -5.00795066e-01 -8.70823503e-01 2.05883496e-02
-1.95348620e-01 3.45049828e-01 -1.04567623e+00 -8.94218266e-01
-6.17746651e-01 -9.23751473e-01 7.24884570e-01 3.65666807e-01
5.51236153e-01 -9.83430147e-01 -1.04840016e+00 9.15623158e-02
3.46058041e-01 -5.14586389e-01 -2.70329893e-01 6.66216314e-01
-1.31158292e+00 -9.50486243e-01 -5.40465057e-01 -7.81388938e-01
9.79088545e-01 1.49812087e-01 9.19952631e-01 3.33833277e-01
-4.21116918e-01 3.39375466e-01 -3.78309377e-02 -1.23991139e-01
-3.71733010e-01 3.29196870e-01 3.50746721e-01 -3.20625007e-01
-1.52554542e-01 -1.38063145e+00 -5.53691626e-01 6.34515658e-02
-1.03173923e+00 1.81278855e-01 8.40501785e-01 8.35722923e-01
6.28336286e-03 1.34018764e-01 5.67814469e-01 -7.46589780e-01
1.09442997e+00 -3.56088221e-01 -2.95124233e-01 3.14678371e-01
-9.28761721e-01 8.20241272e-01 8.28783274e-01 -7.43767083e-01
-7.67328262e-01 -2.61766702e-01 4.55025464e-01 1.08772784e-01
1.23957016e-01 4.88746315e-01 3.37910086e-01 -1.90848872e-01
9.22985256e-01 6.46795452e-01 -5.01643633e-04 -3.74327041e-02
4.46684778e-01 8.94642323e-02 4.58898306e-01 -8.58844757e-01
1.01210082e+00 4.60040152e-01 -1.68448925e-01 -9.33547378e-01
-4.17331338e-01 2.43122518e-01 -4.84375596e-01 -1.99322537e-01
3.62726241e-01 -5.36834002e-01 -8.42433035e-01 5.75893641e-01
-9.23372090e-01 -8.12042594e-01 -8.16469669e-01 1.05503790e-01
-4.49229419e-01 -6.50168061e-02 -7.31915474e-01 -6.17913663e-01
-2.41808727e-01 -5.24885297e-01 3.95021915e-01 2.47556105e-01
-4.67894495e-01 -8.57004285e-01 5.18355608e-01 -4.12285089e-01
7.37690568e-01 -1.19841278e-01 1.50264931e+00 -4.14691627e-01
-4.89472568e-01 3.02049611e-02 6.78161308e-02 9.72847221e-04
1.53174940e-02 2.05355152e-01 -7.17118561e-01 -1.57015771e-01
1.40483469e-01 -1.31910086e-01 1.19721937e+00 2.00477347e-01
8.37427557e-01 -4.69184428e-01 -2.10485727e-01 7.08819807e-01
1.11094975e+00 2.73615718e-01 4.83667105e-01 3.24257910e-01
2.21785635e-01 5.47793865e-01 -6.56557977e-01 2.01369196e-01
-2.17321012e-02 2.87807453e-02 -2.72319876e-02 1.81175411e-01
1.68442458e-01 -4.83152211e-01 5.41688979e-01 1.25573921e+00
-5.22071481e-01 1.55807108e-01 -1.01703203e+00 3.10825676e-01
-1.70387566e+00 -1.13289618e+00 3.95338863e-01 2.20391011e+00
1.05260336e+00 5.20968735e-01 2.90973131e-02 2.42316306e-01
4.58577245e-01 -7.47470632e-02 -6.17913544e-01 -3.77418339e-01
-1.71522319e-01 6.16138458e-01 3.51593047e-01 5.26167274e-01
-3.92298549e-01 7.69118249e-01 7.05225372e+00 2.28984490e-01
-1.35753131e+00 -1.17264837e-02 5.04522383e-01 -1.09105438e-01
-5.69565773e-01 1.76696241e-01 -1.19157627e-01 1.66309178e-01
9.16909039e-01 -3.17968369e-01 1.08605754e+00 2.95026302e-01
-3.30471471e-02 -5.56896403e-02 -1.22704518e+00 6.19817317e-01
-2.20285937e-01 -1.47866321e+00 7.23123699e-02 6.91673309e-02
7.81408429e-01 -1.30546108e-01 2.24127114e-01 2.42238447e-01
5.61081707e-01 -1.15821826e+00 8.47383082e-01 6.86777294e-01
6.84728801e-01 -3.48655015e-01 -2.23180708e-02 4.27146941e-01
-1.01374292e+00 -4.27911013e-01 -2.42637563e-02 -6.35919571e-01
-2.80461013e-01 5.57303846e-01 -5.79153895e-01 -3.08743209e-01
4.03214306e-01 4.27502841e-01 -6.26482964e-01 8.29543591e-01
-2.23066092e-01 7.56797969e-01 -3.26158941e-01 -1.39745936e-01
-3.04566085e-01 -4.12050605e-01 4.01864529e-01 7.85223305e-01
2.93525934e-01 2.00385988e-01 -7.65349805e-01 1.24052262e+00
-3.28759283e-01 -2.81167775e-01 -7.09294200e-01 -5.26390433e-01
4.58391756e-01 1.02489960e+00 -1.19581389e+00 -1.77424863e-01
1.07134774e-01 9.01889026e-01 4.80882615e-01 5.32083035e-01
-5.70192456e-01 -5.51215470e-01 5.27604520e-01 3.16707909e-01
1.37415230e-01 -6.52193010e-01 -4.42116827e-01 -1.22370875e+00
6.82794079e-02 -6.07487679e-01 -4.04629230e-01 -5.65375090e-01
-7.34392047e-01 4.75093901e-01 -1.53259650e-01 -5.86121082e-01
-1.37740746e-01 -4.50755358e-01 -1.01129746e+00 2.59610623e-01
-8.48448575e-01 -4.22536254e-01 -1.93419382e-01 3.39949816e-01
1.14208922e-01 1.62000414e-02 7.36498833e-01 -4.81782369e-02
-6.72643542e-01 6.26565993e-01 3.47542971e-01 4.87265214e-02
1.87605709e-01 -1.21123207e+00 5.71861923e-01 7.86324024e-01
3.77106458e-01 1.27825129e+00 7.68002450e-01 -2.03255937e-01
-1.14116895e+00 -4.68820393e-01 7.06906319e-01 -3.89817506e-01
8.89522493e-01 -8.00821364e-01 -7.58584261e-01 6.94383085e-01
3.24340880e-01 -3.13322783e-01 6.05044246e-01 3.08148742e-01
-6.21257663e-01 -7.03735054e-02 -6.67912900e-01 7.96007872e-01
1.47617078e+00 -7.23559499e-01 -5.10068297e-01 -1.49593309e-01
7.25445092e-01 3.69245380e-01 -4.80947912e-01 1.63054302e-01
7.21478045e-01 -9.07320499e-01 5.63907385e-01 -8.32223415e-01
7.34385252e-01 1.05312401e-02 2.58266032e-01 -1.64903617e+00
-4.33252990e-01 -9.81377721e-01 -1.56573564e-01 8.23871851e-01
6.99261367e-01 -9.50896084e-01 4.98841584e-01 3.53488833e-01
-3.36639136e-02 -6.99082136e-01 -6.20798707e-01 -6.27901852e-01
3.56183022e-01 1.17206806e-02 4.27935928e-01 1.02856207e+00
8.42646241e-01 6.07912898e-01 1.37219653e-01 -4.09087926e-01
2.42337525e-01 -6.68267021e-03 3.80988628e-01 -1.33907402e+00
-5.12555063e-01 -8.62509131e-01 -2.12007567e-01 -9.57104027e-01
-1.88744709e-01 -8.85732591e-01 5.65756038e-02 -9.12192345e-01
1.91485509e-01 -2.92618454e-01 -3.57488096e-01 5.18137515e-01
3.04966092e-01 -6.47846833e-02 7.42230862e-02 4.31780010e-01
-4.14719164e-01 6.24004543e-01 8.43272924e-01 4.46575768e-02
-5.11214077e-01 -2.39533737e-01 -9.67449188e-01 7.11566746e-01
1.01839530e+00 -5.03205001e-01 -5.97674489e-01 -4.07375962e-01
1.21110332e+00 -1.56348750e-01 4.05624419e-01 -1.32757223e+00
3.32207620e-01 1.42095923e-01 3.28223407e-01 4.23337847e-01
-2.19223481e-02 -4.47140098e-01 3.58457938e-02 8.37354779e-01
-7.10000873e-01 4.76479024e-01 1.33778408e-01 3.98839295e-01
1.57137513e-01 -3.18767317e-02 5.96424878e-01 -9.40850750e-02
-1.66711971e-01 -1.09441198e-01 -1.06461442e+00 1.45328298e-01
7.72817969e-01 -2.14466572e-01 -6.78825617e-01 -4.31448787e-01
-8.88207912e-01 -8.67612883e-02 6.56780541e-01 -1.52692320e-02
4.98325199e-01 -9.95455205e-01 -3.28599751e-01 4.65621769e-01
-3.10930282e-01 -2.56440312e-01 -1.23533927e-01 1.03619707e+00
-4.67691213e-01 2.74513602e-01 -5.18406391e-01 -2.70157367e-01
-6.75654829e-01 2.83595860e-01 6.08563960e-01 -1.37335211e-01
-4.27348107e-01 8.77146959e-01 2.40605921e-01 -2.55238950e-01
9.59601719e-03 -6.65856063e-01 1.57180637e-01 1.27792135e-01
1.87636599e-01 8.89611915e-02 -7.90004507e-02 1.54059693e-01
-1.59660310e-01 4.66816992e-01 -4.10270765e-02 -4.26065296e-01
1.56519604e+00 1.63156018e-01 -3.57227832e-01 5.54028451e-01
8.43523383e-01 -1.35136386e-02 -1.30729365e+00 -1.25938430e-01
-2.57775150e-02 2.63440907e-01 -4.70729977e-01 -4.83733237e-01
-8.54830027e-01 9.81160104e-01 3.28199238e-01 6.85152829e-01
1.10967255e+00 3.71616259e-02 4.00958061e-01 8.30170751e-01
2.39364758e-01 -9.80213165e-01 3.04811358e-01 7.41042137e-01
6.61883771e-01 -5.30894339e-01 -2.45693922e-01 1.50067389e-01
-5.38789816e-02 1.18230271e+00 4.82945263e-01 -5.49851000e-01
7.74499476e-01 3.37792069e-01 -4.62656081e-01 -3.59506190e-01
-1.04257476e+00 7.70970359e-02 -1.54089287e-01 3.08538586e-01
4.00836796e-01 -1.39433667e-01 -5.83582707e-02 5.46778202e-01
-6.12082958e-01 -8.49049091e-02 7.17514694e-01 8.87695849e-01
-8.53122354e-01 -7.28157699e-01 3.29419263e-02 4.81062114e-01
-6.34890869e-02 -3.72712880e-01 -8.06258559e-01 7.27223754e-01
1.71270564e-01 3.52612048e-01 2.23311484e-01 -7.46415734e-01
-7.80343935e-02 4.20252174e-01 8.79707456e-01 -4.77838606e-01
-3.60975176e-01 -6.02184653e-01 -3.91957819e-01 -5.85600547e-02
-1.11372903e-01 -6.23884201e-01 -1.22963536e+00 -4.62855190e-01
-2.24639654e-01 2.75897861e-01 4.17712629e-01 1.04770505e+00
4.61295038e-01 6.96218014e-01 4.09193814e-01 -7.79759169e-01
-5.81347883e-01 -8.50283980e-01 -4.58380431e-01 3.04474950e-01
5.66480041e-01 -4.59351838e-01 -6.92796707e-01 1.69354007e-01] | [8.195107460021973, 3.3445687294006348] |
256bc3fc-d1b3-4ae1-96a3-017e7eefe0b6 | user-generated-text-corpus-for-evaluating | 2104.03523 | null | https://arxiv.org/abs/2104.03523v1 | https://arxiv.org/pdf/2104.03523v1.pdf | User-Generated Text Corpus for Evaluating Japanese Morphological Analysis and Lexical Normalization | Morphological analysis (MA) and lexical normalization (LN) are both important tasks for Japanese user-generated text (UGT). To evaluate and compare different MA/LN systems, we have constructed a publicly available Japanese UGT corpus. Our corpus comprises 929 sentences annotated with morphological and normalization information, along with category information we classified for frequent UGT-specific phenomena. Experiments on the corpus demonstrated the low performance of existing MA/LN methods for non-general words and non-standard forms, indicating that the corpus would be a challenging benchmark for further research on UGT. | ['Eiichiro Sumita', 'Taro Watanabe', 'Masao Utiyama', 'Shohei Higashiyama'] | 2021-04-08 | null | https://aclanthology.org/2021.naacl-main.438 | https://aclanthology.org/2021.naacl-main.438.pdf | naacl-2021-4 | ['lexical-normalization'] | ['natural-language-processing'] | [ 1.78240985e-01 -7.14495704e-02 4.47853096e-02 -2.38208771e-01
-8.69211674e-01 -7.51984954e-01 4.99004006e-01 3.41371775e-01
-6.22976065e-01 8.84614885e-01 4.23866928e-01 -5.34713984e-01
5.06549954e-01 -6.71890736e-01 -1.22586131e-01 -4.98182327e-01
3.03670824e-01 4.81501520e-01 2.54473805e-01 -4.51062590e-01
3.30843478e-01 -1.34944515e-02 -1.12404978e+00 3.26747745e-01
1.29517651e+00 4.77391362e-01 3.76609713e-01 6.07407808e-01
-4.53585893e-01 4.76348430e-01 -1.23013699e+00 -9.67908263e-01
9.91215259e-02 -3.76847416e-01 -1.00565290e+00 5.67830727e-02
5.64307034e-01 1.47089317e-01 -1.15668466e-02 1.24290073e+00
7.81079888e-01 3.70341361e-01 8.14156532e-01 -7.15317190e-01
-1.10389197e+00 1.25342357e+00 -4.55257833e-01 4.86100405e-01
3.59965384e-01 -4.72366624e-02 1.22811949e+00 -9.77534890e-01
9.06691372e-01 1.30887520e+00 4.68010634e-01 8.31978798e-01
-8.42128158e-01 -2.14906469e-01 1.30230054e-01 5.56885637e-02
-1.29185915e+00 -3.28571856e-01 3.99115145e-01 -1.38467401e-01
1.37853837e+00 6.26230359e-01 3.03061634e-01 1.15273726e+00
1.24832138e-01 9.79248166e-01 7.99303114e-01 -8.98233414e-01
-1.94799693e-04 1.09752916e-01 5.85969567e-01 5.25827348e-01
4.34497535e-01 -7.55004227e-01 -4.87735355e-03 -2.63619721e-02
5.75936854e-01 -8.24189782e-01 -2.81345725e-01 5.66891670e-01
-1.34499288e+00 8.39044929e-01 -2.80689448e-01 7.84227848e-01
-1.09091699e-01 -3.13628763e-01 8.64723682e-01 3.17020506e-01
6.07360184e-01 3.81828576e-01 -7.94951737e-01 -3.31784397e-01
-4.20813590e-01 1.83049589e-01 8.20787609e-01 1.49480212e+00
3.61516982e-01 2.49502227e-01 -4.78322715e-01 1.36320078e+00
1.54618192e-02 7.61419654e-01 1.06162870e+00 -3.36530805e-01
9.68183041e-01 6.85783803e-01 -2.44339734e-01 -6.70722723e-01
-3.54801655e-01 -1.14063673e-01 -6.89432144e-01 -5.74927568e-01
7.51623929e-01 -2.23078698e-01 -9.54058349e-01 1.70340216e+00
1.53184757e-01 -8.54793429e-01 2.94397295e-01 3.54471862e-01
1.37449181e+00 9.73402619e-01 2.72162110e-01 -6.13480151e-01
1.63077712e+00 -8.83177042e-01 -1.32877052e+00 -2.93056071e-01
1.04291439e+00 -1.48048699e+00 1.84637880e+00 2.49250401e-02
-8.93527567e-01 -3.70897830e-01 -8.33305776e-01 -4.49233621e-01
-8.44251335e-01 5.03565192e-01 4.78595704e-01 1.15613294e+00
-7.13816822e-01 2.62712091e-01 -3.37707520e-01 -8.90514910e-01
1.80270188e-02 -1.54512942e-01 -3.10088992e-01 1.94023654e-01
-1.46737480e+00 9.62798178e-01 6.17383957e-01 6.75154198e-03
-2.61544377e-01 -7.13156685e-02 -1.01134956e+00 -3.72191668e-01
3.64627600e-01 -4.14925277e-01 1.53186333e+00 -3.27349097e-01
-1.40206444e+00 1.08928633e+00 -2.36481547e-01 -2.21212372e-01
1.39505208e-01 -2.06273884e-01 -7.20450878e-01 -2.82706767e-01
3.63261700e-01 1.30714715e-01 5.25972068e-01 -7.31977820e-01
-7.12542415e-01 -3.11990023e-01 -3.80600452e-01 3.51846665e-01
-6.53739393e-01 7.54809499e-01 -3.77457440e-01 -1.12115490e+00
9.47168618e-02 -6.18155837e-01 -1.16312811e-02 -9.83320594e-01
-7.68984437e-01 -7.85750747e-01 6.99323773e-01 -1.18229222e+00
1.93513131e+00 -1.73845160e+00 -2.23911092e-01 -5.77502213e-02
-2.01221764e-01 2.27558866e-01 -1.24929070e-01 5.01272857e-01
5.80810159e-02 6.64319873e-01 -2.18698382e-01 -2.84747750e-01
2.28091344e-01 4.67129290e-01 8.11254829e-02 4.82691750e-02
6.30790666e-02 1.26149261e+00 -9.85759258e-01 -7.96329319e-01
-2.57750251e-03 -2.72052705e-01 -7.26459622e-02 5.68181649e-02
-2.18248948e-01 1.17754862e-01 -3.54712158e-01 9.05781269e-01
3.69852304e-01 5.06727219e-01 1.33322641e-01 -4.78723720e-02
-2.93815732e-01 8.34014416e-01 -9.07613695e-01 1.15138197e+00
-2.89183944e-01 4.70830411e-01 -2.93930650e-01 -4.10750449e-01
7.94673622e-01 4.58925962e-01 -3.36454093e-01 -4.62912977e-01
6.25954986e-01 5.04097223e-01 1.59900516e-01 -7.59886622e-01
1.04020953e+00 -1.65239833e-02 -7.55024731e-01 3.82944524e-01
2.85691351e-01 -5.52131116e-01 1.08502507e+00 1.18603602e-01
8.54230821e-01 -1.61685959e-01 8.72768044e-01 -6.32335007e-01
5.97658157e-01 7.26803020e-02 5.70270002e-01 8.28712404e-01
-3.01147401e-01 7.73850143e-01 3.88505787e-01 -2.61414438e-01
-1.07164049e+00 -8.92249227e-01 -1.32369831e-01 1.21678948e+00
-4.07400459e-01 -7.61362076e-01 -1.29352796e+00 -1.05460262e+00
-5.74830174e-01 1.33989358e+00 -5.28410316e-01 1.96022004e-01
-1.06167746e+00 -1.43819022e+00 6.73885643e-01 4.63859975e-01
3.77135783e-01 -1.42097974e+00 1.85261324e-01 4.49831486e-01
-6.93259239e-01 -1.23824620e+00 -1.11773205e+00 2.89280027e-01
-6.64920151e-01 -9.53360021e-01 -5.99031210e-01 -1.24623978e+00
6.62164390e-01 3.58162642e-01 1.08545554e+00 -2.14013994e-01
1.27172172e-01 7.36820921e-02 -8.18819284e-01 -6.70817852e-01
-9.45068896e-01 3.07984233e-01 2.02948108e-01 -3.66943091e-01
5.86606681e-01 -2.12555125e-01 3.36032480e-01 3.64566028e-01
-9.25678849e-01 -2.73521543e-01 3.67021471e-01 7.01364875e-01
5.13584077e-01 1.73777565e-01 4.58109558e-01 -1.17183089e+00
1.19673538e+00 -9.53429043e-02 -3.24581623e-01 4.11050141e-01
-1.63305953e-01 -2.00749174e-01 5.35855532e-01 -7.84789801e-01
-1.41242993e+00 -3.51263314e-01 -5.32017887e-01 5.92100203e-01
-6.60180897e-02 6.14400208e-01 -9.69001770e-01 2.37833202e-01
7.59058654e-01 2.90660828e-01 -4.88622993e-01 -8.41416299e-01
4.30452317e-01 8.81808102e-01 6.28190339e-01 -7.85304546e-01
8.25375676e-01 -3.98327291e-01 -6.30182862e-01 -1.11906457e+00
-9.36642170e-01 -4.24764246e-01 -8.54156435e-01 1.23780064e-01
7.55770683e-01 -4.02318507e-01 -1.50632774e-02 7.21186042e-01
-1.35568547e+00 -2.67249644e-01 -3.53156656e-01 2.48497859e-01
-1.03997841e-01 8.25415194e-01 -1.10748172e+00 -6.65767729e-01
-7.94980168e-01 -9.43691313e-01 7.58918464e-01 2.29899257e-01
-7.44125128e-01 -1.10417652e+00 -1.28133278e-02 3.32313597e-01
1.71804443e-01 -1.53562263e-01 1.23458266e+00 -1.00860310e+00
2.50594556e-01 -4.35121715e-01 1.97944604e-02 5.58207989e-01
4.13433760e-01 1.84655979e-01 -7.59755015e-01 6.58100843e-02
2.66585559e-01 -1.13115206e-01 5.57320058e-01 3.13279688e-01
6.99221730e-01 -7.24269629e-01 1.30604610e-01 1.46252677e-01
7.77763426e-01 4.02245820e-01 7.16328800e-01 5.79955339e-01
8.98577034e-01 5.67846715e-01 7.48353004e-01 1.61001712e-01
5.48862159e-01 3.89077544e-01 -9.69668999e-02 9.67321917e-02
-1.97871998e-01 -2.96414882e-01 6.78865612e-01 1.75099039e+00
-1.81103289e-01 -7.39868402e-01 -7.51892626e-01 6.10629857e-01
-1.57773304e+00 -7.36813426e-01 -7.87936330e-01 2.00096941e+00
1.24167943e+00 3.76896173e-01 9.80278999e-02 2.30622888e-01
9.16133344e-01 4.98758107e-02 2.46543571e-01 -7.48758554e-01
-8.95375133e-01 -9.65157524e-02 2.95426458e-01 2.20878929e-01
-1.31843603e+00 1.46264589e+00 7.19736147e+00 1.40570295e+00
-5.77315509e-01 2.48574808e-01 5.50082505e-01 5.42248070e-01
-2.15657845e-01 -3.69414032e-01 -1.10214341e+00 2.08310410e-01
1.03681505e+00 -6.15743458e-01 1.78232826e-02 9.36051071e-01
1.80571705e-01 -1.86826177e-02 -5.65351427e-01 7.65386224e-01
3.24866265e-01 -7.08568215e-01 5.42076290e-01 -8.75587463e-02
8.18507135e-01 -1.50637046e-01 -3.09190661e-01 5.97430766e-01
2.24290371e-01 -4.16169941e-01 9.89441812e-01 -3.55408788e-01
7.02528059e-01 -5.34423351e-01 1.27971423e+00 2.06455573e-01
-1.19269276e+00 4.84389007e-01 -7.54592955e-01 -3.97063419e-02
2.45829850e-01 4.63771850e-01 -9.30364251e-01 6.24855220e-01
5.51030219e-01 5.49551427e-01 -1.09346819e+00 7.90688276e-01
-6.59980059e-01 1.05016184e+00 -3.52430165e-01 -6.19638860e-01
1.17043503e-01 -2.67188102e-01 9.74134505e-01 1.78539228e+00
3.97122115e-01 -5.28627820e-02 -3.32812779e-02 3.71672511e-01
-1.68314576e-01 1.16394043e+00 -5.03079772e-01 -3.00696075e-01
4.44678485e-01 1.37348509e+00 -1.19297647e+00 -7.16066301e-01
-4.86754835e-01 1.00009561e+00 3.97929192e-01 1.53517783e-01
-6.22996390e-01 -6.15341783e-01 5.72714627e-01 -1.97217241e-01
3.01887654e-03 -3.04538250e-01 -6.87831938e-01 -1.54970372e+00
1.94381610e-01 -9.07406271e-01 8.75406384e-01 -5.12320697e-01
-1.74792087e+00 1.06936431e+00 -1.30380303e-01 -1.31510115e+00
1.42269626e-01 -8.26748133e-01 -5.80542266e-01 6.59005761e-01
-1.05377495e+00 -1.10485709e+00 2.62175232e-01 3.66197884e-01
1.01608729e+00 -1.13667995e-01 9.54839647e-01 2.17855737e-01
-1.17171848e+00 8.66780996e-01 3.27430665e-02 3.34719419e-01
7.49756157e-01 -1.64061260e+00 9.53012586e-01 1.25768411e+00
1.48283750e-01 9.63606894e-01 6.75015986e-01 -1.14766252e+00
-8.86258900e-01 -1.11439478e+00 1.66701722e+00 -7.46016860e-01
9.57493722e-01 -5.04536033e-01 -1.08417070e+00 7.05363691e-01
4.45489287e-01 -5.83618283e-01 7.53763437e-01 8.18123594e-02
-1.33151621e-01 3.98228884e-01 -1.01099896e+00 9.96659994e-01
1.14377403e+00 -1.48842454e-01 -1.10517287e+00 3.36235046e-01
9.62624431e-01 -3.86888236e-01 -7.23984182e-01 1.69382423e-01
-6.35334151e-03 -1.43153116e-01 5.10835230e-01 -4.18406099e-01
-4.65443730e-02 -3.86580974e-01 -1.38636097e-01 -1.61641347e+00
-4.67139959e-01 -9.53389823e-01 3.30534101e-01 1.87369740e+00
9.12930965e-01 -4.87981230e-01 1.88861355e-01 6.08144283e-01
-6.43675089e-01 1.91772077e-02 -8.81786942e-01 -9.95759904e-01
2.35728040e-01 -7.68020093e-01 5.45644760e-01 9.73172128e-01
4.98985440e-01 8.97930682e-01 -3.63313347e-01 -2.53854930e-01
3.22975993e-01 -3.05232406e-01 3.29587668e-01 -9.03720021e-01
2.99916089e-01 -8.56843710e-01 -1.32201657e-01 -7.63207972e-01
-3.34647670e-02 -1.06819022e+00 9.92227048e-02 -1.60464799e+00
2.33407900e-01 9.04617980e-02 2.10010037e-01 8.02648306e-01
-7.84570694e-01 3.90635610e-01 -2.01915517e-01 -2.18521506e-02
-5.16479969e-01 7.32737720e-01 1.33079207e+00 -8.55347961e-02
-2.51312286e-01 -1.06379762e-01 -6.39342010e-01 7.65007317e-01
1.03536034e+00 -5.03579676e-01 6.23619854e-02 -4.74147946e-01
-8.32567811e-02 -6.81794167e-01 -7.03689456e-01 -4.76172835e-01
-1.54681832e-01 -4.08156633e-01 3.41298163e-01 -8.47005248e-01
-3.17353219e-01 -2.14237660e-01 -4.40550297e-01 1.00145750e-01
2.10869405e-02 3.86184663e-01 3.05723667e-01 -1.61070183e-01
-1.01820089e-01 -7.06374824e-01 5.60736895e-01 -3.13316435e-01
-7.33930528e-01 5.26635014e-02 -1.19472253e+00 4.72405463e-01
5.67155898e-01 -2.28876859e-01 -4.72688287e-01 -1.12085082e-01
-3.51277292e-01 7.18952790e-02 3.46951723e-01 5.81867158e-01
5.20491660e-01 -1.22083068e+00 -9.58212376e-01 1.11304060e-01
4.23584729e-01 -2.81749964e-01 -1.62762776e-01 5.23659468e-01
-6.67791486e-01 3.89947444e-01 -4.81653325e-02 1.63543597e-01
-1.38277233e+00 5.68879604e-01 2.55702250e-02 -3.23319554e-01
-4.44112986e-01 7.35917211e-01 5.57347387e-02 -9.33566511e-01
1.04995988e-01 -4.80392754e-01 -7.18125105e-01 3.08424860e-01
8.53423238e-01 5.08505523e-01 3.05580080e-01 -9.53565657e-01
-8.74747112e-02 5.73848896e-02 -2.80179739e-01 -5.52296527e-02
9.67843711e-01 -5.35571992e-01 -6.50346041e-01 6.56476200e-01
9.80299532e-01 4.19104695e-01 -1.04549721e-01 -2.71386802e-01
6.76028073e-01 -1.66004449e-01 -2.56396323e-01 -7.78357863e-01
-5.48071325e-01 4.69445258e-01 -9.81273949e-02 3.38625193e-01
8.96095812e-01 -5.58406077e-02 1.10213339e+00 5.49216568e-01
3.81630063e-01 -1.61786473e+00 -2.15237662e-01 1.30652595e+00
1.10257673e+00 -1.04119337e+00 -3.47833157e-01 -9.82075989e-01
-6.54961944e-01 1.30056190e+00 8.45672429e-01 5.76459289e-01
2.99174696e-01 1.95571914e-01 5.76986074e-01 1.23234130e-01
-3.12902778e-01 -4.90290463e-01 3.84384811e-01 6.59962118e-01
8.29934299e-01 2.33005807e-01 -1.19865656e+00 1.01094759e+00
-7.80884027e-01 -8.43239546e-01 8.12879503e-01 8.96723449e-01
-4.74593341e-01 -1.46866834e+00 -5.15468895e-01 6.72700286e-01
-7.34711587e-01 -6.37143373e-01 -6.12199068e-01 1.00770557e+00
-1.55159414e-01 1.08270836e+00 -4.92977262e-01 -2.16967493e-01
5.55490196e-01 2.25407496e-01 4.76318389e-01 -1.00654745e+00
-6.81469381e-01 5.10075748e-01 7.26001263e-01 -4.20401338e-03
-1.88774895e-02 -1.07551992e+00 -1.06863236e+00 -1.88604146e-01
-3.64933401e-01 2.91216671e-01 3.37112486e-01 7.88792968e-01
-5.42959750e-01 3.57308865e-01 2.78773606e-01 -5.61667681e-01
-2.58165389e-01 -1.74153256e+00 -7.65619159e-01 5.55358946e-01
-4.58504975e-01 -2.47381657e-01 -6.21374965e-01 2.62122720e-01] | [10.584929466247559, 10.270169258117676] |
18baaaeb-3d8d-4e82-b5a3-c8fc90c06823 | optimal-and-robust-category-level-perception | 2206.12498 | null | https://arxiv.org/abs/2206.12498v2 | https://arxiv.org/pdf/2206.12498v2.pdf | Optimal and Robust Category-level Perception: Object Pose and Shape Estimation from 2D and 3D Semantic Keypoints | We consider a category-level perception problem, where one is given 2D or 3D sensor data picturing an object of a given category (e.g., a car), and has to reconstruct the 3D pose and shape of the object despite intra-class variability (i.e., different car models have different shapes). We consider an active shape model, where -for an object category- we are given a library of potential CAD models describing objects in that category, and we adopt a standard formulation where pose and shape are estimated from 2D or 3D keypoints via non-convex optimization. Our first contribution is to develop PACE3D* and PACE2D*, the first certifiably optimal solvers for pose and shape estimation using 3D and 2D keypoints, respectively. Both solvers rely on the design of tight (i.e., exact) semidefinite relaxations. Our second contribution is to develop outlier-robust versions of both solvers, named PACE3D# and PACE2D#. Towards this goal, we propose ROBIN, a general graph-theoretic framework to prune outliers, which uses compatibility hypergraphs to model measurements' compatibility. We show that in category-level perception problems these hypergraphs can be built from the winding orders of the keypoints (in 2D) or their convex hulls (in 3D), and many outliers can be filtered out via maximum hyperclique computation. The last contribution is an extensive experimental evaluation. Besides providing an ablation study on simulated datasets and on the PASCAL3D+ dataset, we combine our solver with a deep keypoint detector, and show that PACE3D# improves over the state of the art in vehicle pose estimation in the ApolloScape datasets, and its runtime is compatible with practical applications. We release our code at https://github.com/MIT-SPARK/PACE. | ['Luca Carlone', 'Heng Yang', 'Jingnan Shi'] | 2022-06-24 | null | null | null | null | ['vehicle-pose-estimation'] | ['computer-vision'] | [-1.30442232e-01 3.37761194e-01 1.63686693e-01 -1.52338907e-01
-9.89747345e-01 -8.54198635e-01 3.13206315e-01 1.13997437e-01
-4.77042533e-02 -5.45994518e-03 -2.47700959e-01 -2.40447782e-02
-1.70234695e-01 -4.72373277e-01 -1.42343724e+00 -4.45137829e-01
-2.47148752e-01 8.49554598e-01 1.96719438e-01 -1.40356183e-01
2.66698927e-01 6.99139595e-01 -1.74039459e+00 -2.17939645e-01
9.15409505e-01 1.18880343e+00 6.28596619e-02 6.44884348e-01
3.05971026e-01 2.86057759e-02 -3.64717066e-01 -3.98577690e-01
7.98627853e-01 2.55500108e-01 -2.79057175e-01 5.50366402e-01
1.10272586e+00 -4.85209301e-02 -1.63054645e-01 1.41322410e+00
2.76953787e-01 4.13690247e-02 5.70348084e-01 -1.95158613e+00
-3.54335219e-01 -1.15354121e-01 -5.87260008e-01 -4.94886309e-01
3.78927231e-01 3.14253479e-01 8.72120559e-01 -1.35804284e+00
6.33889675e-01 1.45014501e+00 9.89907801e-01 2.44852662e-01
-1.45214486e+00 -3.54182452e-01 4.55458373e-01 1.84280008e-01
-1.64483249e+00 -2.93789834e-01 9.27011013e-01 -7.01424122e-01
7.39119828e-01 3.83625478e-01 6.02044702e-01 1.03841817e+00
1.77147910e-01 8.42189074e-01 8.88696074e-01 1.15245886e-01
4.56385761e-01 -7.11192638e-02 2.88741261e-01 6.48535073e-01
6.18812919e-01 3.20322037e-01 -4.11173671e-01 -4.58347648e-01
4.58587050e-01 -8.66583735e-02 -1.57201961e-01 -1.23077273e+00
-1.20754266e+00 7.29220629e-01 3.65719676e-01 -4.65668976e-01
-1.29367366e-01 9.38847959e-02 2.25317121e-01 3.05747986e-01
3.80492747e-01 3.81166577e-01 -4.88491327e-01 8.35772827e-02
-4.15550590e-01 5.94860554e-01 9.45492625e-01 1.55366719e+00
8.99024129e-01 -9.91207212e-02 3.23439777e-01 5.19092679e-01
4.07508165e-01 9.16196883e-01 -2.61949807e-01 -1.19577456e+00
5.15601456e-01 4.95437354e-01 4.10500407e-01 -1.41082680e+00
-5.45725167e-01 -2.39800066e-01 -7.44673252e-01 6.29763186e-01
3.85513186e-01 1.37335762e-01 -1.15754199e+00 1.61372924e+00
5.67594409e-01 4.49667513e-01 -3.12193353e-02 1.09202433e+00
5.94892681e-01 3.94741595e-01 -4.39148575e-01 -1.04826458e-01
1.03655136e+00 -7.16726243e-01 -2.93648034e-01 -4.53035861e-01
4.94468510e-01 -6.24578714e-01 8.08672428e-01 6.74439251e-01
-9.43313897e-01 -6.32748485e-01 -1.20351779e+00 5.31565733e-02
-4.06870663e-01 -2.11436395e-02 3.76888692e-01 3.89853001e-01
-8.63888264e-01 4.30518270e-01 -9.35891986e-01 -2.40893662e-01
2.74641782e-01 3.91056865e-01 -5.22198141e-01 -1.22525640e-01
-5.48156261e-01 9.28881109e-01 2.27812693e-01 2.80604184e-01
-1.01514161e+00 -8.26908529e-01 -1.15918565e+00 -5.05800068e-01
8.39497507e-01 -6.58248305e-01 1.03396094e+00 -6.25706434e-01
-1.02995682e+00 9.17861819e-01 -7.38806874e-02 -3.00430268e-01
7.78821170e-01 -2.93058008e-01 -1.41109750e-01 -1.46543980e-01
1.39572203e-01 6.11593723e-01 9.75086570e-01 -1.74544406e+00
-3.82781208e-01 -6.44990385e-01 -3.68593037e-02 1.96589217e-01
5.28127134e-01 -4.30147439e-01 -7.81374872e-01 -3.66280645e-01
7.30805755e-01 -1.33886707e+00 -4.68539298e-01 2.25478381e-01
-7.82371104e-01 -1.32157177e-01 7.86121249e-01 -5.65404058e-01
4.48123157e-01 -2.35968733e+00 2.52657801e-01 7.53996313e-01
4.22578305e-01 -4.37560752e-02 -2.90542334e-01 1.09157652e-01
-6.01123571e-02 4.91776615e-02 -2.65699297e-01 -6.37984931e-01
4.10128713e-01 5.63845634e-01 -4.12121624e-01 1.02474999e+00
1.32886544e-01 6.88833177e-01 -8.46991897e-01 -9.33765322e-02
3.44668210e-01 1.32238582e-01 -6.20940566e-01 9.98627916e-02
-3.63556206e-01 3.23020965e-01 -3.93232226e-01 8.03438008e-01
1.36049736e+00 1.46070853e-01 -1.60727903e-01 -4.61360544e-01
2.68534222e-03 -6.95103630e-02 -1.94342554e+00 1.70950985e+00
-1.13061756e-01 2.01674148e-01 4.03407305e-01 -9.20121849e-01
9.62582767e-01 -1.80501193e-01 6.58402562e-01 -5.40548980e-01
1.72367003e-02 2.92079538e-01 -3.08194220e-01 -3.05085272e-01
4.82918859e-01 4.73451346e-01 -3.83594334e-01 -1.49144217e-01
-1.47560954e-01 -7.75844693e-01 1.06642954e-01 1.58581421e-01
1.10602582e+00 -5.93815558e-03 7.98891187e-02 -2.81919628e-01
1.27691165e-01 1.66126892e-01 7.67923415e-01 8.07902396e-01
-7.32446387e-02 9.13778305e-01 6.33759737e-01 -2.30369970e-01
-1.05058992e+00 -1.44164526e+00 -2.62756854e-01 3.59739572e-01
5.92690289e-01 -3.64569277e-01 -4.22812879e-01 -7.35903561e-01
6.66528702e-01 5.71865320e-01 -5.85400820e-01 -1.11873753e-01
-4.79240954e-01 -2.91570514e-01 1.74279258e-01 3.87744486e-01
1.46691874e-01 -3.42240661e-01 -3.16350549e-01 -2.12890119e-03
7.89683685e-02 -1.18832934e+00 -5.41767240e-01 2.39021704e-01
-7.14158416e-01 -1.36126757e+00 -1.51313037e-01 -6.32321179e-01
8.92885089e-01 2.88708985e-01 1.26563478e+00 -1.73276477e-02
-2.02700406e-01 7.76700854e-01 -1.96920425e-01 -6.24689221e-01
-2.05640674e-01 -4.49615628e-01 4.68548238e-01 3.26119885e-02
1.40541598e-01 -5.80387831e-01 -2.47294202e-01 6.52063727e-01
-6.02905512e-01 -2.15036422e-01 3.63021642e-01 6.08312070e-01
1.17280352e+00 -6.35218620e-02 -1.73482131e-02 -6.75080776e-01
2.05716476e-01 -4.66110349e-01 -1.21037018e+00 -1.01946391e-01
-3.23389292e-01 -6.67337403e-02 4.36631680e-01 -4.29129273e-01
-2.85778344e-01 6.65760577e-01 5.13043404e-02 -1.06857169e+00
-2.84812063e-01 2.75420815e-01 -6.19588256e-01 -2.73635268e-01
6.95657611e-01 -1.69951886e-01 8.02340880e-02 -4.54703361e-01
3.87612253e-01 3.42352837e-01 7.66670585e-01 -7.66406894e-01
1.25225925e+00 5.08980453e-01 2.49177203e-01 -8.04145575e-01
-7.46590197e-01 -7.10045516e-01 -6.64886951e-01 -2.11626723e-01
5.10328770e-01 -1.03709996e+00 -9.49839890e-01 5.25252700e-01
-1.23221123e+00 -1.99168414e-01 -3.35731834e-01 3.93505931e-01
-8.62964749e-01 4.64367568e-01 -1.07385963e-02 -7.56350756e-01
3.39358598e-01 -1.21866870e+00 1.39794374e+00 -3.67645115e-01
3.17672379e-02 -6.41599238e-01 1.05716638e-01 1.27219290e-01
-2.30770826e-01 6.84946835e-01 4.51417893e-01 -6.50085986e-01
-8.68533552e-01 -3.21395367e-01 -6.79058358e-02 4.52097774e-01
-3.15936089e-01 -3.67081910e-02 -7.90040314e-01 -5.02165139e-01
-4.56627272e-02 -1.40146866e-01 6.14671886e-01 2.43942201e-01
1.13744402e+00 -3.73261273e-01 -4.73234147e-01 8.20629537e-01
1.33665395e+00 -1.84021965e-01 4.56319928e-01 7.11487606e-02
8.99271071e-01 5.12104213e-01 9.66328084e-01 3.41316313e-01
5.77151537e-01 8.95639002e-01 1.05317974e+00 1.64168686e-01
7.75087103e-02 -2.53843874e-01 4.12857533e-01 7.50027657e-01
1.39910117e-01 -1.54366875e-02 -9.60829973e-01 6.32805049e-01
-2.07877159e+00 -7.02912390e-01 -4.90382910e-01 2.46511292e+00
3.01891953e-01 2.07774580e-01 1.83943167e-01 -1.18393600e-01
5.75023770e-01 -2.24541903e-01 -1.01159823e+00 -1.09387353e-01
-3.91638875e-02 -2.15572894e-01 8.28213096e-01 6.80640578e-01
-1.21376622e+00 5.49095988e-01 5.56676531e+00 5.18679500e-01
-6.49539351e-01 -6.60425723e-02 2.93959290e-01 -6.54465845e-03
-2.05649629e-01 1.39015494e-02 -7.99190640e-01 3.40868592e-01
5.23127556e-01 -1.53215528e-01 4.04643953e-01 1.22529697e+00
-4.11925046e-03 -1.00780591e-01 -1.56746614e+00 1.23353314e+00
2.28084207e-01 -1.11262619e+00 -3.73019785e-01 2.68110424e-01
7.94647455e-01 2.92083323e-01 1.05927743e-01 1.60710067e-01
3.79446745e-01 -8.53806436e-01 1.10946023e+00 4.17072803e-01
6.31503105e-01 -7.47462988e-01 5.64669013e-01 4.00661379e-01
-1.28895485e+00 7.88855702e-02 -6.22964144e-01 2.08246857e-01
-3.62593755e-02 7.10911870e-01 -6.27870262e-01 7.24172294e-01
8.49494457e-01 8.13457131e-01 -5.09573281e-01 1.33911192e+00
1.10426173e-02 2.53519267e-01 -8.49965930e-01 2.55065769e-01
-1.09229349e-01 -2.86609501e-01 1.11414921e+00 9.89304721e-01
3.61368626e-01 3.29989791e-02 7.63138473e-01 1.00384164e+00
8.72078612e-02 -3.28413010e-01 -8.15348148e-01 5.77383280e-01
4.95759964e-01 1.16528070e+00 -4.10245121e-01 -1.12768263e-01
-3.48955691e-01 7.43406594e-01 2.29177505e-01 3.69515955e-01
-1.00054109e+00 -2.25699574e-01 1.22675169e+00 8.92850980e-02
4.11215514e-01 -5.57084799e-01 -6.29319489e-01 -1.13788295e+00
5.13381422e-01 -9.55015063e-01 1.54654592e-01 -7.09275961e-01
-1.52447832e+00 2.95290440e-01 2.01668710e-01 -1.57717156e+00
-1.42732963e-01 -9.96377230e-01 -3.80275160e-01 6.17911398e-01
-1.05444849e+00 -1.01517797e+00 -4.88442272e-01 6.65002048e-01
2.12118223e-01 2.88474739e-01 4.06168401e-01 1.01070531e-01
-3.46988618e-01 5.07359505e-01 1.28854997e-02 -9.39623564e-02
3.94641221e-01 -1.40211308e+00 6.42053604e-01 9.22271729e-01
1.52899921e-01 3.22542340e-01 9.84087408e-01 -7.30901301e-01
-2.04195595e+00 -1.36517012e+00 4.68991071e-01 -9.00284648e-01
6.61733389e-01 -7.79497385e-01 -7.65479982e-01 8.66567016e-01
-5.22623420e-01 3.18183601e-01 -9.68998894e-02 2.90171336e-02
-4.87159044e-01 -1.86244473e-01 -1.28326809e+00 5.81156313e-01
1.50211740e+00 -3.28929305e-01 -5.33710003e-01 6.42819285e-01
7.35207617e-01 -1.06627679e+00 -7.31599629e-01 6.34917796e-01
1.47379547e-01 -8.60143781e-01 1.22611618e+00 -3.39005411e-01
-2.60643482e-01 -8.04911196e-01 -4.94890571e-01 -1.42699289e+00
-8.62250477e-02 -6.69310927e-01 -3.12821597e-01 8.73744488e-01
2.22228751e-01 -7.13442028e-01 7.49677658e-01 6.36025727e-01
-7.45785594e-01 -5.89763284e-01 -1.20146286e+00 -1.21543396e+00
-1.27766833e-01 -1.01887131e+00 5.05518675e-01 6.57000780e-01
-4.25973952e-01 2.21099705e-02 -1.77519381e-01 9.75397289e-01
9.78937149e-01 1.30506918e-01 1.29733074e+00 -1.37366319e+00
-1.98388368e-01 -7.29084760e-02 -9.59376276e-01 -1.31699789e+00
2.45882347e-01 -8.53291094e-01 4.65612143e-01 -1.14621854e+00
-5.58491684e-02 -5.09872258e-01 1.91882566e-01 4.43735510e-01
2.06017807e-01 1.18773878e-01 3.77783656e-01 -9.99498088e-03
-6.59108281e-01 4.63670582e-01 1.09703588e+00 -3.87419283e-01
-1.68626919e-01 2.09300518e-01 -4.91934299e-01 9.77675021e-01
5.66923857e-01 -4.04539734e-01 -2.02164754e-01 -3.85044277e-01
4.05026734e-01 -4.13694270e-02 1.02536488e+00 -1.10705805e+00
2.93331951e-01 -2.67218858e-01 2.52807975e-01 -9.83523846e-01
6.77646935e-01 -1.20439589e+00 3.44323128e-01 1.85647085e-01
1.55242026e-01 2.48553336e-01 3.02071154e-01 7.99317002e-01
-1.31575577e-02 2.28578132e-02 6.45447969e-01 8.34895596e-02
-8.83199155e-01 6.86939776e-01 -4.18164134e-02 2.37164229e-01
1.16016769e+00 -4.88801569e-01 -1.92006111e-01 -2.69491941e-01
-8.24796379e-01 4.73900646e-01 7.63923645e-01 6.43425047e-01
8.13796878e-01 -1.49670827e+00 -6.93448365e-01 3.89692456e-01
5.62054753e-01 5.44410229e-01 1.29507795e-01 8.55452120e-01
-2.45213807e-01 9.70813334e-02 1.09571598e-01 -1.18732011e+00
-1.12250674e+00 7.36485600e-01 3.38659823e-01 2.12817505e-01
-4.01920885e-01 7.54167974e-01 3.35885823e-01 -9.04312491e-01
5.79460442e-01 -7.36790359e-01 4.41869229e-01 -2.33706594e-01
1.13919973e-01 5.41102052e-01 3.66077155e-01 -8.86072338e-01
-6.31602705e-01 8.35366547e-01 2.60003924e-01 1.89874321e-01
1.07040358e+00 -7.71241635e-02 -1.42034590e-02 3.65296572e-01
1.28288090e+00 2.49909833e-01 -1.63133574e+00 6.59347326e-02
-6.64969087e-02 -5.83551526e-01 -1.90059841e-01 -5.21779001e-01
-9.77313161e-01 3.39974493e-01 5.45969605e-01 1.27254263e-01
8.08217704e-01 3.10650676e-01 4.30964023e-01 6.00434601e-01
6.42987669e-01 -1.07170916e+00 2.29018331e-02 7.24392772e-01
1.30877244e+00 -1.26048315e+00 4.47492162e-03 -7.06275284e-01
-4.01389003e-01 9.56741393e-01 7.52917826e-01 -4.25089657e-01
6.77424848e-01 2.85284966e-01 9.24338307e-03 -2.84773618e-01
-4.81681556e-01 -3.21259350e-01 6.84472740e-01 7.81672120e-01
-3.48997355e-01 2.60427207e-01 3.34402710e-01 4.13513780e-01
-4.27558690e-01 -4.86799002e-01 4.25695479e-01 6.66137636e-01
-3.94935161e-01 -7.56518960e-01 -7.30397642e-01 4.05621976e-01
3.04215789e-01 3.10714453e-01 -4.00322646e-01 8.09215128e-01
4.99794334e-01 1.00365722e+00 1.43782988e-01 -5.41242301e-01
9.03995931e-01 -3.83948296e-01 4.70296085e-01 -6.39680445e-01
-2.63853520e-02 -1.13906331e-01 6.90563545e-02 -1.15234864e+00
-1.00779764e-01 -9.60188866e-01 -1.09300292e+00 -3.99004102e-01
-2.48038128e-01 -7.51626864e-02 7.66664624e-01 6.60114288e-01
6.23469770e-01 3.75264734e-02 6.66490018e-01 -1.18190455e+00
-7.21650720e-01 -4.88033593e-01 -5.79581857e-01 4.59325641e-01
5.74547946e-01 -8.92910361e-01 -6.39708698e-01 -1.50616780e-01] | [7.602764129638672, -2.6636240482330322] |
d809706d-75ce-4414-9299-98e958fe78b2 | task-adaptive-network-for-image-restoration | null | null | https://openaccess.thecvf.com/content/WACV2022W/VAQ/html/Zhou_Task_Adaptive_Network_for_Image_Restoration_With_Combined_Degradation_Factors_WACVW_2022_paper.html | https://openaccess.thecvf.com/content/WACV2022W/VAQ/papers/Zhou_Task_Adaptive_Network_for_Image_Restoration_With_Combined_Degradation_Factors_WACVW_2022_paper.pdf | Task Adaptive Network for Image Restoration With Combined Degradation Factors | Existing methods have achieved excellent performance on image restoration, but most of them are designed for one type of degradation. However, the weather is complex in the real world. So networks designed for single tasks are usually difficult to apply. Therefore, we propose a task-adaptive attention module to enable the network to restore images with multiple degradation factors. The task-adaptive attention module mainly includes three parts: Task-Adaptive sub-network, Task Channel Attention, and Task Operation Attention. To evaluate the model, we construct a mixed degradation factors dataset that combines three degradation factors of rain, haze, and raindrop. The experimental results show that our method not only better restores images with mixed degradation factors, but also show competitive results compared to the state-of-the-art models of each task. | ['Congduan Li', 'Wantong Liao', 'Minyi Lin', 'Chaktou Leong', 'Jingyuan Zhou'] | 2022-02-15 | null | null | null | ieee-cvf-winter-conference-on-applications-of-3 | ['single-image-haze-removal', 'single-image-deraining'] | ['computer-vision', 'computer-vision'] | [ 7.18599232e-03 -6.40714824e-01 2.62664229e-01 -4.31309044e-01
-4.01584446e-01 1.42429203e-01 2.66152769e-01 -6.27958238e-01
-2.05091417e-01 6.22712195e-01 5.38618326e-01 -3.07466984e-01
1.18743010e-01 -4.92409468e-01 -5.28613925e-01 -9.67554867e-01
1.41330808e-01 -4.50810827e-02 3.11376691e-01 -3.69817168e-01
8.17070995e-03 2.28504822e-01 -1.60234416e+00 4.32573110e-01
1.49747968e+00 8.14293742e-01 7.79190183e-01 7.15227485e-01
-1.61873698e-02 8.02428246e-01 -9.45227563e-01 -4.54019718e-02
1.17786050e-01 -4.04496193e-01 -3.65774184e-01 3.68752539e-01
4.02859747e-01 -7.09301591e-01 -7.25733578e-01 1.29595780e+00
8.45247269e-01 2.12637037e-01 3.70103687e-01 -9.87668812e-01
-1.45237589e+00 3.39303851e-01 -5.34145772e-01 6.94360077e-01
1.79306790e-02 3.85873079e-01 5.64084947e-01 -9.40559149e-01
6.59352019e-02 1.68002486e+00 3.54139388e-01 5.71513176e-01
-8.37034881e-01 -6.32535577e-01 5.19487679e-01 6.97311938e-01
-9.66607153e-01 -3.87156218e-01 5.89433134e-01 -2.07907215e-01
7.48905540e-01 -5.62839881e-02 4.67122287e-01 9.42819357e-01
4.67772871e-01 8.28172505e-01 1.30366099e+00 1.48841083e-01
-3.48754078e-02 -3.96288484e-01 5.39783053e-02 1.05263360e-01
9.17054340e-02 3.03620938e-02 -1.94067463e-01 4.63807136e-01
7.97903776e-01 2.10121125e-01 -8.93338978e-01 1.67168915e-01
-1.04548907e+00 4.75006044e-01 9.12108600e-01 1.09075688e-01
-6.07861638e-01 5.63065782e-02 1.60468608e-01 4.99223083e-01
8.59658480e-01 6.42675757e-02 -3.31811368e-01 4.02043015e-01
-7.37237275e-01 2.48617277e-01 3.05228829e-01 4.81162727e-01
5.43180764e-01 4.34553444e-01 -4.46467757e-01 1.10625267e+00
4.04570550e-01 7.11740911e-01 4.14943963e-01 -8.25509608e-01
5.13301373e-01 1.83081537e-01 3.28859210e-01 -8.35190296e-01
-4.15120572e-01 -6.34048104e-01 -1.37321866e+00 2.83614218e-01
-7.71272182e-02 -1.69395357e-01 -1.56441224e+00 1.50975955e+00
1.54619738e-02 1.72397852e-01 7.81553239e-02 1.49727690e+00
1.08672798e+00 1.03184903e+00 8.44926089e-02 -4.42471415e-01
1.15011668e+00 -1.44456816e+00 -1.26483285e+00 -5.34701705e-01
-2.11356506e-01 -8.35833073e-01 1.50813639e+00 4.64056730e-01
-9.18318748e-01 -8.67185414e-01 -1.06369734e+00 -1.23566456e-01
2.33520996e-02 8.27205107e-02 4.32257235e-01 3.24663401e-01
-1.14297354e+00 3.77791703e-01 -4.61107254e-01 -1.92479804e-01
5.71001828e-01 7.20469421e-03 -1.73256189e-01 -7.73316205e-01
-1.28295183e+00 1.00409639e+00 -1.60967337e-03 5.72476566e-01
-1.51616073e+00 -3.90737832e-01 -7.10754871e-01 2.66679287e-01
5.48264384e-02 -8.08742642e-01 1.14449179e+00 -1.11994982e+00
-1.32115519e+00 2.82117724e-01 -2.95097917e-01 -1.55992687e-01
3.06524396e-01 -6.32775605e-01 -7.42678463e-01 6.65567741e-02
-4.83113267e-02 3.65769237e-01 1.19526339e+00 -1.72572446e+00
-5.59572756e-01 -2.02981457e-01 3.12511981e-01 5.03960669e-01
-4.36681986e-01 2.60711778e-02 -5.00948906e-01 -7.64716744e-01
6.00143559e-02 -2.67148972e-01 -3.00712556e-01 3.22020687e-02
-3.13183010e-01 5.85684218e-02 1.07931352e+00 -1.13130164e+00
1.09377098e+00 -2.29426479e+00 2.08128840e-01 -4.30740833e-01
2.84305751e-01 6.40324354e-01 -4.69294161e-01 1.19160496e-01
-1.84562847e-01 1.92484736e-01 -4.13773954e-01 -5.04216254e-01
-2.36225605e-01 4.89102513e-01 -3.60913247e-01 4.51376110e-01
8.55887905e-02 5.71563423e-01 -7.30875373e-01 -1.69087932e-01
1.02152027e-01 6.55802131e-01 -1.74733594e-01 6.90735340e-01
-2.38711700e-01 4.64187562e-01 -2.34974667e-01 8.02053332e-01
1.08143508e+00 -1.39238089e-01 -1.22646719e-01 -4.58438903e-01
3.70937623e-02 1.19781159e-01 -8.24475229e-01 1.34247458e+00
-4.41481501e-01 5.72603047e-01 4.31654841e-01 -6.87277257e-01
8.13174486e-01 3.37419987e-01 -1.90761477e-01 -9.07930851e-01
5.76883629e-02 5.15184179e-02 2.39297956e-01 -1.04611659e+00
3.91438782e-01 -2.80846626e-01 5.76056182e-01 1.40611246e-01
-6.97200373e-02 2.67440844e-02 7.28094205e-02 7.68569857e-02
1.01293373e+00 9.17704105e-02 -1.61289066e-01 -1.53060362e-01
5.00140071e-01 -4.72538531e-01 8.80710185e-01 5.80753088e-01
-4.93202627e-01 9.35503840e-01 2.46968269e-01 -5.19492626e-01
-8.70965302e-01 -9.10012007e-01 1.20634660e-01 1.02097750e+00
6.04898214e-01 1.07809432e-01 -6.61657274e-01 -5.23735940e-01
-1.33051500e-01 4.60192084e-01 -6.59107447e-01 -2.65608996e-01
-5.71005046e-01 -1.23654211e+00 8.88556093e-02 3.92749667e-01
1.08698463e+00 -1.32054448e+00 -3.80037501e-02 2.79402584e-02
-5.67884803e-01 -9.74639714e-01 -6.61059141e-01 -1.39884472e-01
-7.53150761e-01 -1.03579950e+00 -9.84345138e-01 -8.65790606e-01
6.62158966e-01 1.05221975e+00 1.12156570e+00 4.21411783e-01
1.23366773e-01 -3.85941863e-02 -4.96984839e-01 -3.52430582e-01
2.97543481e-02 -2.11687103e-01 -1.08895391e-01 3.07835281e-01
-1.25988334e-01 -8.06283116e-01 -8.87879074e-01 4.16336805e-01
-1.18881142e+00 -1.66241243e-01 8.47958028e-01 9.70871449e-01
3.47372264e-01 3.36958766e-01 7.18249381e-01 -5.78357041e-01
9.01901424e-01 -5.09392679e-01 -1.94012195e-01 2.28942826e-01
-5.15444994e-01 -2.96008795e-01 7.50587106e-01 -4.68413174e-01
-1.43910348e+00 -3.66505414e-01 -1.77638680e-01 -5.69478869e-01
-2.97879130e-01 6.57923639e-01 -7.37841606e-01 -1.22973092e-01
5.03828585e-01 3.24213833e-01 -1.07589483e-01 -8.91011953e-01
1.44357696e-01 7.78813720e-01 7.95729578e-01 -2.10259840e-01
9.64397848e-01 3.68441641e-01 -4.59847927e-01 -5.70644259e-01
-1.08018267e+00 -1.50752231e-01 -1.85587183e-01 -4.49838161e-01
8.47922921e-01 -1.31430626e+00 -3.16734642e-01 1.19548094e+00
-1.20549226e+00 -6.33705378e-01 1.17822066e-01 3.86567712e-01
-8.66314545e-02 5.54976344e-01 -7.95269608e-01 -6.92500889e-01
-4.47561562e-01 -1.32124722e+00 6.94867790e-01 6.64156556e-01
8.10900033e-01 -6.99207544e-01 -2.65243292e-01 4.86838013e-01
9.09662485e-01 -1.36444554e-01 9.08195436e-01 1.26299083e-01
-6.70597434e-01 1.53719485e-01 -7.29964077e-01 7.86974728e-01
2.43953034e-01 -2.31800392e-01 -1.03591406e+00 -4.84127074e-01
1.73645899e-01 -2.08007455e-01 1.51561093e+00 4.73873019e-01
1.36168659e+00 -1.58553839e-01 2.39274185e-02 8.29570532e-01
1.34520137e+00 1.36862010e-01 1.30210328e+00 4.57211792e-01
7.28737473e-01 3.70570838e-01 6.76421106e-01 6.14991412e-02
6.96079016e-01 3.29076827e-01 1.08622432e+00 -5.04396439e-01
-5.66533923e-01 2.36410394e-01 5.97703516e-01 8.68259132e-01
-2.27417275e-01 -6.43620312e-01 -6.99689388e-01 9.85387743e-01
-1.77064562e+00 -7.36095488e-01 -2.36595720e-01 1.84727979e+00
7.25770652e-01 6.62169829e-02 -3.32957923e-01 2.98015457e-02
7.68765569e-01 5.77849388e-01 -6.12312138e-01 -5.71740139e-03
-5.20734012e-01 -1.84616119e-01 2.11164460e-01 5.34528434e-01
-1.16774070e+00 8.35403919e-01 6.59858894e+00 6.18642807e-01
-1.01489353e+00 2.84824222e-01 5.66142619e-01 -9.48333591e-02
-3.81587833e-01 -1.71652243e-01 -4.42277133e-01 7.68330932e-01
5.98203778e-01 1.87747002e-01 6.47394657e-01 3.94654423e-01
5.49183130e-01 -4.20687459e-02 -3.55023652e-01 7.18379498e-01
2.42996126e-01 -6.25151634e-01 3.20238411e-01 -2.26871088e-01
7.38122404e-01 1.11712903e-01 1.51778445e-01 3.93591046e-01
3.78216892e-01 -1.09193468e+00 4.50553387e-01 7.36709118e-01
5.27018368e-01 -5.82294643e-01 1.20156062e+00 2.87332773e-01
-8.13733101e-01 -3.75062883e-01 -7.22509742e-01 -2.06891626e-01
5.02547860e-01 1.05554545e+00 1.28732128e-02 6.71146452e-01
1.25465095e+00 9.27729428e-01 -4.88871932e-01 1.48028553e+00
-6.93338454e-01 7.73304939e-01 1.73487917e-01 6.38484538e-01
4.30743657e-02 -1.44320428e-01 4.60169971e-01 1.04264975e+00
3.44843835e-01 1.76487118e-01 2.14197576e-01 3.26406032e-01
-1.89631417e-01 -3.70778471e-01 -2.14451924e-01 3.28326732e-01
2.02960804e-01 1.22936392e+00 -1.95360154e-01 -3.03973883e-01
-4.27160233e-01 1.18799543e+00 2.00943232e-01 7.90286243e-01
-8.90575707e-01 -4.64038163e-01 1.14951015e+00 -1.60933480e-01
5.09685278e-01 -3.73549640e-01 -2.13858485e-01 -1.27754068e+00
1.56086341e-01 -1.11999774e+00 4.94389758e-02 -1.43080115e+00
-1.53922248e+00 9.08586979e-01 -3.66020530e-01 -1.00406301e+00
6.41870320e-01 -4.45011199e-01 -9.67452526e-01 1.15270579e+00
-2.17039394e+00 -1.16145027e+00 -9.60951507e-01 7.10004151e-01
7.76196301e-01 2.28543719e-03 4.54427779e-01 7.04462349e-01
-9.75924492e-01 2.88917959e-01 1.41539037e-01 -1.55723244e-01
1.07755435e+00 -1.18774474e+00 3.75356644e-01 1.40631258e+00
-5.30838192e-01 3.70294809e-01 7.36165404e-01 -6.91927910e-01
-1.13091242e+00 -1.31554878e+00 6.92139506e-01 2.30265692e-01
3.06511492e-01 -8.59756544e-02 -1.37600541e+00 4.14357662e-01
5.22173882e-01 1.83597356e-01 9.83754694e-02 -5.50587922e-02
-5.18254340e-01 -5.47257423e-01 -1.01984298e+00 5.47641397e-01
1.01311958e+00 -3.71461719e-01 -4.40731019e-01 6.03728950e-01
1.15486038e+00 -5.04280984e-01 -6.86279953e-01 4.75019813e-01
5.44619337e-02 -1.03084469e+00 1.04408288e+00 -4.45599675e-01
6.33903444e-01 -7.38859355e-01 -3.07558268e-01 -1.86354089e+00
-7.76661456e-01 -2.06289023e-01 -3.04661900e-01 1.24730480e+00
1.18618049e-01 -6.45836353e-01 1.13842487e-01 -5.55875152e-02
-7.95787632e-01 -5.87139130e-01 -6.80687189e-01 -6.47904217e-01
-9.76771489e-02 2.96796355e-02 7.50963867e-01 7.78491855e-01
-7.71658123e-01 4.30121094e-01 -9.30276990e-01 6.09135628e-01
6.29537344e-01 1.29034221e-01 5.27623534e-01 -1.05981660e+00
-1.35926455e-01 -2.56214827e-01 4.77078930e-02 -1.21600890e+00
-5.71220703e-02 -1.99156240e-01 4.72274214e-01 -2.17850518e+00
2.76864201e-01 -7.09837154e-02 -4.19448555e-01 6.70977950e-01
-8.65466595e-01 3.18751216e-01 1.45094752e-01 5.15035272e-01
-5.66871285e-01 1.02349341e+00 1.79990745e+00 -3.23896170e-01
4.47405912e-02 -1.29581481e-01 -1.00464427e+00 5.11905432e-01
9.51419234e-01 -2.38424510e-01 -4.13563788e-01 -1.20462096e+00
-1.59599736e-01 -2.34211564e-01 4.35224265e-01 -1.13312042e+00
1.35088816e-01 -2.77803510e-01 4.99187887e-01 -4.53224570e-01
3.48413587e-01 -5.81673741e-01 -1.26241505e-01 2.41222739e-01
-6.64763600e-02 6.08519325e-03 -1.07380021e-02 7.45281935e-01
-5.02463281e-01 2.17564955e-01 9.46324527e-01 -1.77690253e-01
-8.58986974e-01 5.33958375e-01 -4.78049189e-01 -2.74998695e-01
4.69139487e-01 1.88939542e-01 -9.97194827e-01 -8.04724634e-01
-7.10988462e-01 6.22701228e-01 3.39933813e-01 6.34580910e-01
8.45622301e-01 -1.17477620e+00 -1.10690963e+00 4.88136671e-02
-2.35056132e-01 1.23684689e-01 8.53960872e-01 8.08864534e-01
-4.41214800e-01 -9.33457166e-02 -4.02639598e-01 -2.22227722e-01
-1.13964379e+00 6.50708616e-01 5.53140998e-01 -1.73835084e-01
-5.47909558e-01 7.17987716e-01 5.65838814e-01 -3.14125240e-01
2.28309482e-01 -1.18910015e-01 -4.98155057e-01 -1.11458451e-01
9.23934460e-01 2.40536809e-01 4.91958559e-02 -4.51566279e-01
-7.53025664e-03 3.62986147e-01 -2.17576921e-01 2.89490968e-01
1.41791344e+00 -5.28934836e-01 -4.02921677e-01 7.83803612e-02
4.97620374e-01 -5.29206812e-01 -1.60145819e+00 -4.45037425e-01
-6.83462143e-01 -7.33696580e-01 3.42315048e-01 -1.20650661e+00
-1.54644620e+00 1.18561029e+00 8.41059446e-01 1.96948543e-01
1.87477064e+00 -4.35451627e-01 8.78591478e-01 1.86944261e-01
1.00148156e-01 -6.95249021e-01 2.42049798e-01 7.68112659e-01
1.27512372e+00 -1.12739348e+00 4.17395644e-02 -3.41127366e-01
-5.64601660e-01 7.60531902e-01 1.09018862e+00 -1.50091156e-01
6.94839299e-01 4.03056517e-02 5.65853238e-01 -1.02030702e-01
-7.64502406e-01 -4.24594909e-01 2.07181588e-01 9.79831278e-01
2.27953181e-01 -8.22821707e-02 -3.06150049e-01 7.56644368e-01
3.49179894e-01 1.14853969e-02 6.40078843e-01 6.75251484e-01
-8.66243422e-01 -8.44701469e-01 -6.12911463e-01 4.17252332e-01
-5.02463520e-01 -3.90590221e-01 -1.26450911e-01 3.91750485e-01
3.01255494e-01 1.23330414e+00 -1.47120968e-01 -5.49099565e-01
5.03116190e-01 -3.81165832e-01 9.84745026e-02 -3.09018731e-01
-5.69009721e-01 4.77842540e-02 -3.95753458e-02 -5.42765439e-01
-6.33481920e-01 -2.80538619e-01 -6.51803732e-01 -3.30058068e-01
-2.08134890e-01 -1.18388481e-01 4.30518121e-01 1.00840676e+00
4.34494108e-01 1.09619236e+00 7.13585138e-01 -1.20245814e+00
-2.99668700e-01 -1.38255537e+00 -7.21513689e-01 2.50545293e-01
7.27323532e-01 -7.49257624e-01 -4.39592153e-01 1.13947056e-01] | [11.04465389251709, -2.9183356761932373] |
56ce3fe1-b24f-476b-a6f0-2ac03d81a195 | a-self-supervised-miniature-one-shot-texture | 2306.08814 | null | https://arxiv.org/abs/2306.08814v1 | https://arxiv.org/pdf/2306.08814v1.pdf | A Self-Supervised Miniature One-Shot Texture Segmentation (MOSTS) Model for Real-Time Robot Navigation and Embedded Applications | Determining the drivable area, or free space segmentation, is critical for mobile robots to navigate indoor environments safely. However, the lack of coherent markings and structures (e.g., lanes, curbs, etc.) in indoor spaces places the burden of traversability estimation heavily on the mobile robot. This paper explores the use of a self-supervised one-shot texture segmentation framework and an RGB-D camera to achieve robust drivable area segmentation. With a fast inference speed and compact size, the developed model, MOSTS is ideal for real-time robot navigation and various embedded applications. A benchmark study was conducted to compare MOSTS's performance with existing one-shot texture segmentation models to evaluate its performance. Additionally, a validation dataset was built to assess MOSTS's ability to perform texture segmentation in the wild, where it effectively identified small low-lying objects that were previously undetectable by depth measurements. Further, the study also compared MOSTS's performance with two State-Of-The-Art (SOTA) indoor semantic segmentation models, both quantitatively and qualitatively. The results showed that MOSTS offers comparable accuracy with up to eight times faster inference speed in indoor drivable area segmentation. | ['William R. Norris', 'Zheyu Zhou', 'Chirag Rastogi', 'Yu Chen'] | 2023-06-15 | null | null | null | null | ['navigate', 'robot-navigation'] | ['reasoning', 'robots'] | [ 2.76274651e-01 1.08466573e-01 2.98749715e-01 -5.61732829e-01
-4.74432826e-01 -4.99570638e-01 4.98962879e-01 -4.25701439e-02
-4.62181985e-01 7.39644408e-01 -6.61479115e-01 -6.27916992e-01
-2.22438842e-01 -1.23458505e+00 -6.86332345e-01 -5.38400769e-01
4.44589406e-02 8.40108752e-01 7.84435570e-01 -4.57849264e-01
3.99052471e-01 3.48130524e-01 -1.90499628e+00 -1.86961398e-01
1.32892942e+00 1.01925969e+00 8.87615621e-01 6.07795358e-01
-2.64710158e-01 2.31119797e-01 -2.43219465e-01 1.21135198e-01
3.31378371e-01 1.31075352e-01 -8.27993274e-01 -7.69589990e-02
5.83659589e-01 -2.76127785e-01 -8.32448155e-02 8.78406227e-01
6.49124458e-02 4.71422791e-01 6.35924399e-01 -1.12069178e+00
2.07930699e-01 -6.80190548e-02 -3.19627553e-01 5.82688339e-02
5.54038763e-01 -1.80903729e-02 4.26249146e-01 -6.10419750e-01
6.84199333e-01 1.03842890e+00 1.00400686e+00 -2.73833089e-02
-8.48598659e-01 -3.29452336e-01 1.81948736e-01 2.54229456e-01
-1.51731825e+00 -3.23690861e-01 6.98821723e-01 -2.68997639e-01
9.48410988e-01 3.13787371e-01 4.48804647e-01 7.88937211e-01
4.24351096e-01 5.52081347e-01 1.25976551e+00 -3.19821417e-01
7.53374457e-01 1.09770782e-01 9.27023217e-02 1.14899087e+00
3.97594064e-01 -6.02193251e-02 -8.16994086e-02 4.93853867e-01
4.86106575e-01 -1.29638597e-01 2.21129715e-01 -8.04879069e-01
-9.86600161e-01 5.29577076e-01 7.95800865e-01 1.81409717e-01
4.50909324e-02 4.54914570e-03 1.40261844e-01 -4.44261953e-02
4.53021765e-01 5.23896925e-02 -3.47082704e-01 -3.44006211e-01
-9.65357423e-01 8.16718563e-02 7.24673271e-01 1.19833302e+00
1.19636166e+00 -1.37146845e-01 4.12773192e-01 8.64030778e-01
5.10425031e-01 1.03152263e+00 1.29380345e-01 -1.17435694e+00
7.19295859e-01 4.99519140e-01 2.17055559e-01 -1.24916732e+00
-6.69236839e-01 3.92366871e-02 -5.16092181e-01 2.36871719e-01
5.98319232e-01 1.29753023e-01 -1.43495274e+00 1.06134379e+00
3.95776242e-01 -2.98888296e-01 -1.63847767e-02 8.98066163e-01
7.87583292e-01 4.96572971e-01 -2.40098611e-01 5.16173482e-01
1.07044685e+00 -1.15566587e+00 -4.51787591e-01 -7.60839581e-01
7.22453654e-01 -5.82801223e-01 1.09901845e+00 2.58487821e-01
-3.67844373e-01 -6.12924159e-01 -1.29425120e+00 -2.59777278e-01
-9.32091296e-01 1.30361766e-01 7.91459501e-01 1.15496874e+00
-1.01711273e+00 4.68482524e-01 -1.06107438e+00 -7.78970182e-01
4.23818320e-01 3.76265436e-01 -4.32521909e-01 -3.56192350e-01
-9.65568125e-01 9.62721109e-01 3.21947336e-01 1.94537207e-01
-7.28099465e-01 -1.67383045e-01 -1.21400416e+00 -5.36296964e-01
5.71444511e-01 -4.51738209e-01 9.90332961e-01 -3.75622928e-01
-1.62466931e+00 8.81399035e-01 -4.12290454e-01 -4.99589115e-01
6.89689994e-01 -2.59110421e-01 -4.95638624e-02 1.72301248e-01
7.07496881e-01 7.76533008e-01 2.81834900e-01 -1.56551015e+00
-1.00040841e+00 -6.31503284e-01 2.35546634e-01 4.21564072e-01
3.77053976e-01 -8.99612188e-01 -6.80384040e-01 1.41869530e-01
6.37352109e-01 -1.16078591e+00 -5.31821072e-01 -5.11212572e-02
-6.29945576e-01 2.93534398e-01 1.27467299e+00 -5.99060178e-01
6.94939852e-01 -1.98098159e+00 -3.05068165e-01 5.78008711e-01
-2.34986559e-01 -2.14485690e-01 3.24822128e-01 1.72575563e-01
7.83965647e-01 -7.89688453e-02 -7.12377608e-01 -5.03994048e-01
9.56057906e-02 6.58692420e-01 2.52056211e-01 5.17960608e-01
-5.37270725e-01 6.17485940e-01 -9.35884774e-01 -5.48072219e-01
1.11524308e+00 2.96939820e-01 -2.94633150e-01 -1.47736832e-01
-2.78330687e-02 4.11105990e-01 -5.93031704e-01 1.03177118e+00
8.01885247e-01 1.16939001e-01 1.36755168e-01 -3.92279141e-02
-5.52845836e-01 1.86747000e-01 -1.18776894e+00 1.82510746e+00
-8.07610929e-01 8.91064286e-01 1.05663531e-01 -8.02754879e-01
1.07728648e+00 -3.50578696e-01 2.83403128e-01 -1.31892276e+00
1.75289467e-01 6.71141565e-01 -4.46985394e-01 -4.52712089e-01
9.50682700e-01 4.29928452e-01 -2.72229999e-01 -1.82106391e-01
-2.90743858e-01 -5.69904685e-01 2.44892180e-01 -1.28240392e-01
9.42980587e-01 6.11056864e-01 4.81350422e-02 -6.57464206e-01
5.54578543e-01 5.61336339e-01 2.52009630e-01 9.25392866e-01
-3.18117380e-01 6.63777351e-01 -2.59178966e-01 -4.16637808e-01
-1.01559174e+00 -1.23633385e+00 -2.50087470e-01 7.67719328e-01
1.11652732e+00 -9.56950150e-03 -1.12205732e+00 -4.74749714e-01
-7.54823163e-02 6.18998051e-01 -4.83146727e-01 3.86259913e-01
-4.81853187e-01 -5.64107597e-01 2.48841867e-01 4.27985847e-01
1.18071830e+00 -7.04545081e-01 -1.20283198e+00 1.36842191e-01
-5.79573274e-01 -1.36891127e+00 2.85551101e-01 5.28950930e-01
-6.72397494e-01 -1.24182904e+00 -4.35573041e-01 -8.41705561e-01
7.90037572e-01 7.15711296e-01 8.20095003e-01 -1.94205746e-01
-3.37132394e-01 5.61278224e-01 -3.78597081e-01 -2.74149835e-01
9.16520357e-02 3.33056659e-01 -1.39233470e-01 -5.76358974e-01
5.94108775e-02 -2.65940934e-01 -6.56169653e-01 8.61338139e-01
-2.89388239e-01 3.10855001e-01 3.81536692e-01 4.18888837e-01
8.61246049e-01 4.19709384e-01 -8.90319794e-02 -8.86954427e-01
-4.01699394e-02 -4.94812399e-01 -6.98630810e-01 -1.44546889e-02
-5.63746214e-01 -1.73146948e-01 4.58307445e-01 3.15141916e-01
-1.34121752e+00 1.01692826e-01 -3.01879197e-01 2.60482907e-01
-5.99347234e-01 8.06093737e-02 -2.96874136e-01 -3.76764059e-01
4.46515948e-01 6.59746453e-02 -1.18159980e-01 -3.42926502e-01
2.84977823e-01 7.05819547e-01 6.19053602e-01 -6.26733482e-01
6.53056026e-01 9.01743889e-01 2.93800049e-02 -1.32210696e+00
-5.80472231e-01 -6.92165792e-01 -1.12243915e+00 -3.92308474e-01
7.18521118e-01 -7.20099926e-01 -1.91128597e-01 6.13149941e-01
-6.29266083e-01 -9.78683233e-01 -2.59820260e-02 1.11181207e-01
-7.89527833e-01 1.89566731e-01 -3.36954623e-01 -9.80155587e-01
3.19473706e-02 -1.22971094e+00 1.24307418e+00 3.95395696e-01
-1.14056759e-01 -1.08150077e+00 -6.43192679e-02 8.74752522e-01
4.55019265e-01 3.35213602e-01 4.17495459e-01 2.48693153e-02
-7.28177249e-01 2.02045380e-03 -2.44917601e-01 -2.09781244e-01
3.03276747e-01 -1.87445432e-01 -1.21155441e+00 1.29872099e-01
-3.55609030e-01 1.07486591e-01 7.03842759e-01 4.33918625e-01
9.32084203e-01 7.56551996e-02 -6.86738431e-01 6.07777715e-01
1.43815327e+00 7.15583384e-01 8.53285313e-01 1.11800528e+00
7.57125735e-01 6.28654718e-01 1.29426908e+00 -1.45040914e-01
8.07274222e-01 7.80455172e-01 6.67367280e-01 -1.17413357e-01
-9.85595956e-02 -2.17220888e-01 -2.68946569e-02 5.92616975e-01
-8.61466676e-02 -3.52930903e-01 -1.15388548e+00 5.81321955e-01
-1.82890046e+00 -5.31693995e-01 -4.35096473e-01 2.01542687e+00
1.29479200e-01 4.71828073e-01 -3.18638951e-01 4.46768075e-01
4.59162533e-01 1.11443877e-01 -3.29839230e-01 -5.03360927e-01
2.85721794e-02 3.97905596e-02 1.00495422e+00 8.38564515e-01
-1.49265993e+00 1.15854633e+00 5.83844805e+00 1.08349943e+00
-8.82344723e-01 -3.38253602e-02 6.42212033e-01 6.05648458e-01
-6.03507683e-02 -1.61348388e-01 -7.02519894e-01 4.58762616e-01
8.31409633e-01 6.09586477e-01 2.64934778e-01 1.19127834e+00
2.96372175e-01 -1.14377987e+00 -4.37317878e-01 7.53489852e-01
-3.64696383e-02 -9.93282616e-01 -4.20498103e-01 -2.76061110e-02
7.92096913e-01 1.71566203e-01 -1.19549640e-01 1.48123130e-01
7.23868981e-02 -8.71879697e-01 1.03500736e+00 3.50719333e-01
6.79982185e-01 -7.01057315e-01 1.01057351e+00 4.03519064e-01
-1.67639863e+00 2.15693880e-02 -3.38852227e-01 -2.73263663e-01
1.67686254e-01 5.19337833e-01 -1.05100894e+00 8.57326448e-01
9.36185241e-01 5.42012393e-01 -7.84511328e-01 1.04669809e+00
-1.93577424e-01 9.48502570e-02 -6.91723108e-01 -1.92022979e-01
5.74084163e-01 -4.54661578e-01 1.63490832e-01 1.27815950e+00
3.45387161e-01 -3.29806894e-01 4.26955819e-01 5.62917888e-01
5.21688163e-01 -1.81944013e-01 -8.97633433e-01 5.50053954e-01
5.84299266e-01 1.20654392e+00 -1.63498580e+00 -3.03662956e-01
4.47905883e-02 1.10042310e+00 -7.64668584e-02 2.64424086e-01
-7.94277012e-01 -6.56123281e-01 2.94469416e-01 4.02911961e-01
1.96428642e-01 -6.95402920e-01 -7.61158288e-01 -3.88286531e-01
-1.43646710e-02 -1.03101365e-01 -2.25828886e-01 -8.78561914e-01
-5.99950850e-01 3.44603300e-01 1.86589696e-02 -1.07624662e+00
1.20818235e-01 -8.65866244e-01 -2.69781739e-01 4.21201676e-01
-1.57393730e+00 -1.25086021e+00 -9.12019968e-01 4.17290807e-01
8.69553387e-01 2.71214187e-01 7.95229435e-01 4.28272545e-01
-5.75603604e-01 1.36623383e-01 4.02412146e-01 -1.60452634e-01
1.40082747e-01 -1.34884465e+00 4.07665312e-01 8.71639371e-01
-2.88504690e-01 4.28277552e-01 7.88011670e-01 -9.28023934e-01
-1.30375063e+00 -1.10148382e+00 3.09629709e-01 -2.46926293e-01
1.47269711e-01 -4.58530575e-01 -5.40160239e-01 3.77684653e-01
-4.56689298e-01 -2.83048868e-01 5.98146878e-02 -5.37663978e-03
2.00694367e-01 -6.80000111e-02 -1.50490391e+00 6.32259905e-01
1.31656969e+00 -3.89985055e-01 -4.15853530e-01 -2.80750263e-02
3.09909940e-01 -9.51485813e-01 -4.34315532e-01 6.02131903e-01
7.20837712e-01 -1.33286977e+00 1.10611236e+00 6.87835336e-01
2.09115386e-01 -4.33663696e-01 -4.97354358e-01 -1.01213396e+00
6.82664141e-02 1.09639093e-01 2.02764735e-01 9.73449349e-01
4.10905242e-01 -7.63392150e-01 1.15267670e+00 5.83508968e-01
-6.65114880e-01 -4.59732115e-01 -1.15127158e+00 -9.39946532e-01
-4.68606383e-01 -6.90419614e-01 2.44650573e-01 6.76844716e-01
-4.89200056e-01 -2.30888762e-02 1.77765608e-01 2.86203265e-01
7.27658749e-01 6.16946034e-02 9.55158412e-01 -1.09971726e+00
6.53758943e-01 -2.02098250e-01 -5.28776705e-01 -1.06797540e+00
-1.67450234e-01 -5.67989349e-01 7.76470602e-01 -2.27113152e+00
-2.71585435e-01 -1.08197737e+00 1.96889251e-01 3.15168411e-01
2.31499031e-01 3.98680300e-01 -3.79132241e-01 3.45027894e-02
-9.44513440e-01 4.43017870e-01 9.97380555e-01 -2.35725522e-01
-3.20013821e-01 -7.60449320e-02 -7.91749358e-02 1.04626811e+00
9.66869175e-01 -4.90402728e-02 -6.05909228e-01 -3.07342976e-01
-5.29792868e-02 -3.41885179e-01 2.90539920e-01 -1.61039162e+00
3.05106401e-01 -9.75904465e-02 4.01594132e-01 -1.04640663e+00
6.60885811e-01 -8.31316650e-01 1.33786559e-01 5.60524046e-01
3.93769681e-01 -3.21798354e-01 3.79378617e-01 5.24672091e-01
7.03096837e-02 -2.49164641e-01 7.02610493e-01 -3.84746552e-01
-1.48300445e+00 -8.93213227e-02 -7.27640927e-01 -1.57112554e-01
1.14940834e+00 -1.06158650e+00 -1.66329771e-01 4.68833335e-02
-3.22106034e-01 3.55068326e-01 8.77432823e-01 3.72566193e-01
8.19478691e-01 -8.73207569e-01 1.90891653e-01 2.99545527e-01
9.49700847e-02 6.16909862e-01 3.76008004e-01 6.87224329e-01
-1.35988009e+00 6.59149766e-01 -3.55968475e-01 -1.02058578e+00
-1.08469105e+00 3.91681232e-02 4.69842821e-01 8.33027288e-02
-1.03082108e+00 6.61017239e-01 8.99536982e-02 -1.05179274e+00
6.84723556e-02 -5.78672707e-01 -1.17639946e-02 -1.88256815e-01
-4.76226257e-03 8.20123076e-01 1.47098944e-01 -7.26034462e-01
-6.47209942e-01 9.55799878e-01 4.79879588e-01 -1.87505811e-01
8.46246839e-01 -8.90520215e-01 1.01435892e-01 8.02175999e-01
8.53411198e-01 -1.05887249e-01 -1.28451109e+00 3.31169218e-01
2.33714566e-01 -5.88544786e-01 7.29686096e-02 -7.26754963e-01
-6.38633490e-01 6.52766705e-01 9.56939101e-01 2.43834313e-02
5.58898807e-01 -3.55565041e-01 8.34081590e-01 8.24338555e-01
1.39557505e+00 -1.59061921e+00 -1.55798018e-01 8.12647223e-01
2.48643026e-01 -1.64865887e+00 -1.71956234e-02 -7.91701555e-01
-2.35984087e-01 1.05298030e+00 6.20019257e-01 8.36915746e-02
4.69165683e-01 2.88370222e-01 1.08125500e-01 -3.27543795e-01
3.45827520e-01 -1.68039545e-01 6.18749112e-02 8.98000896e-01
-1.91086099e-01 3.43621224e-01 1.86763227e-01 2.19955556e-02
-6.82149112e-01 -2.21068457e-01 3.52363378e-01 1.23194182e+00
-9.83936012e-01 -6.16006970e-01 -6.54845893e-01 4.50974494e-01
1.90913588e-01 2.87988663e-01 -4.95197512e-02 1.09425354e+00
4.78220522e-01 1.31528974e+00 1.86843187e-01 -4.91544455e-01
3.13521206e-01 -2.96419680e-01 4.20008749e-01 -4.21011955e-01
7.95113295e-02 -3.41160715e-01 3.66253555e-01 -8.44323695e-01
-4.28476065e-01 -4.69843894e-01 -1.62611949e+00 -3.11860621e-01
-1.44638196e-01 -7.82116801e-02 1.19492733e+00 1.23935878e+00
-5.88495890e-03 5.87255657e-01 5.23215890e-01 -1.22339439e+00
6.08023226e-01 -5.48851609e-01 -5.44956326e-01 -8.03883895e-02
3.88973244e-02 -1.09205985e+00 -2.25154519e-01 -2.28807956e-01] | [8.399362564086914, -2.207371473312378] |
2ee6659a-3981-4b4a-a6bb-fbd4ccf551fd | supporting-search-engines-with-knowledge-and | 2102.06762 | null | https://arxiv.org/abs/2102.06762v1 | https://arxiv.org/pdf/2102.06762v1.pdf | Supporting search engines with knowledge and context | Search engines leverage knowledge to improve information access. In order to effectively leverage knowledge, search engines should account for context, i.e., information about the user and query. In this thesis, we aim to support search engines in leveraging knowledge while accounting for context. In the first part of this thesis, we study how to make structured knowledge more accessible to the user when the search engine proactively provides such knowledge as context to enrich search results. As a first task, we study how to retrieve descriptions of knowledge facts from a text corpus. Next, we study how to automatically generate knowledge fact descriptions. And finally, we study how to contextualize knowledge facts, that is, to automatically find facts related to a query fact. In the second part of this thesis, we study how to improve interactive knowledge gathering. We focus on conversational search, where the user interacts with the search engine to gather knowledge over large unstructured knowledge repositories. We focus on multi-turn passage retrieval as an instance of conversational search. We propose to model query resolution as a term classification task and propose a method to address it. In the final part of this thesis, we focus on search engine support for professional writers in the news domain. We study how to support such writers create event-narratives by exploring knowledge from a corpus of news articles. We propose a dataset construction procedure for this task that relies on existing news articles to simulate incomplete narratives and relevant articles. We study the performance of multiple rankers, lexical and semantic, and provide insights into the characteristics of this task. | ['Nikos Voskarides'] | 2021-02-12 | null | null | null | null | ['conversational-search'] | ['natural-language-processing'] | [ 2.34547868e-01 4.57086533e-01 -4.76373583e-01 1.19496293e-01
-1.01353014e+00 -8.74063492e-01 9.04281676e-01 4.96762902e-01
-4.68505949e-01 1.06315827e+00 7.58884728e-01 -2.11794451e-01
-6.51175976e-01 -8.41005147e-01 -5.00909388e-01 9.73951668e-02
4.59751308e-01 7.10922360e-01 5.15811145e-01 -6.16113424e-01
7.37656295e-01 1.89957440e-01 -1.74399590e+00 7.01970875e-01
8.22777092e-01 4.60453033e-01 6.71925008e-01 6.09319508e-01
-5.75753033e-01 9.76193547e-01 -8.44685614e-01 -4.30617660e-01
-3.61152858e-01 -5.09851217e-01 -1.67116833e+00 -1.95374832e-01
1.85036883e-01 1.40447810e-01 1.32487074e-01 7.71407962e-01
5.46397567e-01 4.12976295e-01 5.83337188e-01 -8.01790416e-01
-3.31652850e-01 1.28644204e+00 2.36475378e-01 6.75248384e-01
1.07285225e+00 -4.77263808e-01 1.02526867e+00 -6.35121047e-01
1.40478444e+00 8.83904874e-01 3.76692861e-01 4.72078651e-01
-8.05417597e-01 -1.87842369e-01 1.33973673e-01 4.95147407e-01
-1.14740086e+00 -6.37329519e-01 7.37728417e-01 -5.03370583e-01
1.25873744e+00 6.81699634e-01 6.91012025e-01 1.21807563e+00
-2.44918585e-01 7.57251918e-01 6.21965170e-01 -9.37089205e-01
2.87149940e-02 6.33019865e-01 5.10863066e-01 4.03234750e-01
3.81398611e-02 -4.53345746e-01 -9.42379713e-01 -4.14625227e-01
3.74374479e-01 -5.63704014e-01 -3.73194605e-01 2.87388444e-01
-1.05402195e+00 6.28518045e-01 -2.11342961e-01 8.05570781e-01
-7.18499959e-01 -3.50578874e-01 3.34196091e-01 2.27025986e-01
3.38931948e-01 1.27245641e+00 -4.52851504e-01 -6.36694372e-01
-7.43043900e-01 4.85044926e-01 1.62735057e+00 1.04344499e+00
5.40770471e-01 -7.46469438e-01 -5.01787066e-01 1.03422379e+00
-6.33599535e-02 1.95017949e-01 7.77755499e-01 -1.02693963e+00
4.57858056e-01 7.33968079e-01 3.90154988e-01 -1.00842190e+00
-1.38982773e-01 -5.99648952e-01 -2.75229034e-03 -7.33656347e-01
1.47085175e-01 -3.30592506e-02 -3.60273898e-01 1.54605711e+00
3.04977894e-01 -1.61733031e-01 3.90367150e-01 6.28013134e-01
1.02087212e+00 5.33626497e-01 7.12564727e-03 -7.37001181e-01
1.68548191e+00 -8.56585503e-01 -9.83707607e-01 -2.02543721e-01
9.17666137e-01 -9.62400317e-01 1.18876731e+00 2.36806855e-01
-1.24321127e+00 -1.54772192e-01 -6.55358911e-01 -1.48883194e-01
-6.95591688e-01 1.06862269e-01 3.25078696e-01 2.17986554e-01
-7.53252387e-01 4.56345618e-01 -4.34289813e-01 -6.42895103e-01
-2.89680064e-01 -8.87574032e-02 1.37612164e-01 8.41186196e-02
-1.74274635e+00 9.92000580e-01 6.26477778e-01 -4.26470101e-01
-3.80327672e-01 -8.14113736e-01 -4.50277150e-01 2.59694219e-01
1.10134292e+00 -1.01934361e+00 1.52330661e+00 -7.29855478e-01
-1.25606334e+00 7.12966681e-01 -3.50762129e-01 -2.15381756e-01
1.04424380e-01 -4.01215553e-01 -4.86927956e-01 2.58722365e-01
6.39615655e-01 -9.46120396e-02 2.88777411e-01 -1.29214060e+00
-9.41210687e-01 -1.25599802e-01 4.49204266e-01 5.25985658e-01
-2.93325394e-01 2.85314828e-01 -7.55797148e-01 -5.10275841e-01
-2.74399906e-01 -5.68802118e-01 2.53654301e-01 -1.08390057e+00
-6.49848700e-01 -4.53570247e-01 6.88583791e-01 -6.76324844e-01
1.85978758e+00 -1.67520916e+00 1.60574913e-01 4.29553837e-01
1.24674521e-01 3.90675999e-02 2.68566370e-01 1.01057315e+00
4.64329720e-01 4.32577312e-01 1.88682854e-01 -5.41298427e-02
6.05881400e-03 3.15563411e-01 -6.65474176e-01 -6.11341476e-01
-4.14687246e-01 8.06515455e-01 -1.07768619e+00 -8.18912327e-01
-4.67479974e-01 1.47640631e-01 -3.63500088e-01 -7.94043094e-02
-7.08171725e-01 6.36129305e-02 -9.79349911e-01 4.56104606e-01
-3.70694041e-01 -4.92552876e-01 1.76473960e-01 -6.46631494e-02
-3.20636362e-01 7.18112469e-01 -1.01772511e+00 1.60374093e+00
-7.69822896e-01 5.89797854e-01 -9.93294828e-03 -7.41651714e-01
5.02339959e-01 5.48930883e-01 3.10587883e-01 -6.49229467e-01
-6.42282814e-02 3.04745793e-01 -5.08514166e-01 -9.41579819e-01
9.41686094e-01 -2.67627854e-02 -1.98189154e-01 7.82886922e-01
-2.47744113e-01 7.30229020e-02 6.41199887e-01 5.04984140e-01
1.28115857e+00 -1.99425355e-01 3.61184418e-01 -1.03058033e-01
5.53041816e-01 6.24619722e-01 1.82145219e-02 1.09822035e+00
5.58844745e-01 -1.88572228e-01 3.60263765e-01 -1.94418862e-01
-5.99424422e-01 -3.80701095e-01 1.41711891e-01 1.45413566e+00
1.97473258e-01 -8.16686928e-01 -7.23749399e-01 -6.02420151e-01
-2.38562584e-01 1.06814408e+00 -3.84244919e-01 -9.14840400e-02
-6.22665584e-01 -2.80982614e-01 6.82266355e-01 1.32140040e-01
5.85701823e-01 -1.24882567e+00 -9.36021090e-01 4.90589082e-01
-9.77481127e-01 -1.08053660e+00 -2.44971573e-01 4.80319373e-02
-3.88017356e-01 -1.18628860e+00 -2.85307974e-01 -6.07682168e-01
6.00572899e-02 -7.84669630e-03 1.50329292e+00 2.44058430e-01
-7.47922361e-02 1.05823851e+00 -8.49131227e-01 -4.82395977e-01
-5.63672423e-01 7.30581343e-01 -2.36830890e-01 -3.91273379e-01
3.70778203e-01 -5.11312068e-01 -2.98814714e-01 3.69933218e-01
-1.11085463e+00 8.23108330e-02 5.77568769e-01 6.42940938e-01
5.32226264e-01 2.12585300e-01 6.19773090e-01 -1.08785558e+00
1.37740803e+00 -6.70577407e-01 -1.22998737e-01 7.88398445e-01
-7.76670098e-01 3.12206388e-01 1.27559528e-01 -4.86124218e-01
-1.44642162e+00 -3.43516529e-01 7.94054940e-03 2.11006463e-01
1.26716215e-02 1.36102808e+00 -2.50982065e-02 3.68232071e-01
1.04840934e+00 3.27922851e-01 -4.67791289e-01 -5.51726818e-01
4.47103083e-01 7.01953173e-01 4.37878698e-01 -1.12916136e+00
2.83949882e-01 1.72250032e-01 -2.16649204e-01 -1.12932038e+00
-1.08779812e+00 -9.12055552e-01 -2.64574885e-01 -4.72140282e-01
5.83690047e-01 -3.57199907e-01 -6.98475301e-01 -4.76863593e-01
-1.18764758e+00 -3.29738289e-01 -4.46642697e-01 4.20236707e-01
-6.98263824e-01 1.53596386e-01 -3.03880513e-01 -7.05555320e-01
-4.31040704e-01 -6.41118765e-01 9.86840010e-01 2.27670595e-01
-5.98772228e-01 -1.08883071e+00 3.23645949e-01 6.92956865e-01
3.00554127e-01 -1.92993190e-02 9.83835220e-01 -1.31255841e+00
-6.11979842e-01 8.23724804e-06 1.98918015e-01 -3.80814373e-01
1.16875134e-01 -2.49672011e-01 -6.69226706e-01 2.47883931e-01
-7.08640739e-02 -2.05021426e-01 5.26976466e-01 -9.54090729e-02
6.86598241e-01 -9.01590824e-01 -7.25517333e-01 -7.85027519e-02
1.01100910e+00 3.15726489e-01 4.45859879e-01 8.84985507e-01
2.43177906e-01 9.54087496e-01 7.79408038e-01 3.45248431e-01
4.19763923e-01 9.19937849e-01 -2.69123912e-01 6.99675322e-01
-1.67742550e-01 -3.42385799e-01 -1.70310661e-01 5.99669814e-01
-3.01085919e-01 -1.92377821e-01 -1.09845984e+00 7.10650325e-01
-1.98560154e+00 -1.43906009e+00 1.54411167e-01 1.64999771e+00
1.43802440e+00 3.97154242e-02 5.48027158e-02 -2.95720808e-02
4.48113084e-01 -2.01264262e-01 -9.22630951e-02 -1.60639048e-01
4.98133227e-02 2.54154112e-02 -5.54576926e-02 8.52048755e-01
-5.14694750e-01 1.05689120e+00 5.32767200e+00 1.04291999e+00
-7.37404227e-01 1.33663386e-01 -9.81693119e-02 -1.28331557e-01
-6.28295720e-01 3.18097293e-01 -1.11799514e+00 2.47383490e-01
7.89553702e-01 -7.07647741e-01 4.32911456e-01 8.15158010e-01
4.48932536e-02 -4.37986463e-01 -1.11057639e+00 5.13598442e-01
2.79999942e-01 -1.70572591e+00 2.40510061e-01 2.17231493e-02
6.78435087e-01 -5.46989262e-01 -3.20862979e-01 4.89629686e-01
1.71880648e-01 -6.90317094e-01 4.97428417e-01 9.67737377e-01
2.65753776e-01 -4.73720610e-01 6.74487412e-01 7.51355529e-01
-9.67370689e-01 -8.50313455e-02 2.10816264e-01 -5.08789439e-03
3.10481906e-01 5.43616235e-01 -1.16343355e+00 7.79139340e-01
4.69292015e-01 2.12962389e-01 -2.90511996e-01 8.80350828e-01
-2.14944154e-01 3.38825762e-01 -2.73643821e-01 -4.30983633e-01
-8.54739919e-03 3.38127106e-01 9.41148877e-01 1.40352833e+00
2.79566228e-01 5.03618598e-01 2.71568030e-01 8.38438392e-01
-1.41916960e-01 4.16518688e-01 -6.60138309e-01 -9.45752300e-03
8.76160860e-01 8.45511436e-01 -6.92950726e-01 -6.81488097e-01
-1.02781756e-02 8.23365033e-01 2.10949913e-01 4.54411328e-01
-1.72545061e-01 -5.58945656e-01 1.29452407e-01 2.45182022e-01
-1.38876308e-02 1.55733272e-01 1.08451404e-01 -9.76324379e-01
2.72439718e-01 -9.21329558e-01 5.83942115e-01 -9.00323749e-01
-9.74294186e-01 5.39068639e-01 6.60803556e-01 -6.17633939e-01
-7.86692977e-01 1.45518333e-02 -2.74826884e-01 7.40132570e-01
-1.45785797e+00 -1.00477648e+00 -1.86004147e-01 5.98693967e-01
6.83331668e-01 8.82418752e-02 7.65576124e-01 1.93650216e-01
-9.22263265e-02 1.14177205e-01 -1.36754259e-01 -1.52516728e-02
7.51264989e-01 -1.08517957e+00 -3.09746325e-01 5.89740574e-01
3.81405771e-01 1.14649582e+00 9.05625403e-01 -1.22483075e+00
-1.25758648e+00 -5.75833499e-01 1.58655775e+00 -7.68543422e-01
6.89756572e-01 1.94131613e-01 -8.55710268e-01 6.02906764e-01
1.74047247e-01 -8.02809536e-01 9.32860136e-01 5.24260581e-01
-9.42853168e-02 2.07400367e-01 -8.00930142e-01 5.72012424e-01
1.09322894e+00 -8.64264548e-01 -1.34436417e+00 4.51504439e-01
7.33369529e-01 -5.81877530e-01 -7.92227387e-01 2.73875892e-02
5.54478943e-01 -4.52766210e-01 9.37716246e-01 -6.05025828e-01
4.25776422e-01 -1.84528887e-01 -1.46963894e-01 -1.25320768e+00
4.17199992e-02 -7.45908082e-01 -3.19226593e-01 1.34890676e+00
8.36348772e-01 -3.66188914e-01 4.10424471e-01 7.28664279e-01
-2.65830576e-01 -5.40133476e-01 -9.59142148e-01 -7.57550597e-01
-2.60166645e-01 -4.47955042e-01 5.02884090e-01 1.06785488e+00
7.25433528e-01 7.78433323e-01 2.00353302e-02 7.81668425e-02
7.38115907e-02 3.52039248e-01 6.30167305e-01 -1.36417711e+00
-3.01526994e-01 -5.85194767e-01 3.40245217e-01 -9.45405245e-01
1.43637910e-01 -8.23851049e-01 -9.78073180e-02 -1.92251074e+00
3.91705185e-01 -1.45165175e-01 1.96284309e-01 4.58961219e-01
-1.43181548e-01 -2.65604347e-01 -5.21741398e-02 5.72605669e-01
-7.75203645e-01 5.76407090e-02 1.11250198e+00 -2.30837464e-02
-7.22081959e-01 2.30289567e-02 -1.01465547e+00 4.53077823e-01
5.52064776e-01 -3.83379459e-01 -6.82974219e-01 -2.21258461e-01
1.12232280e+00 4.23604667e-01 4.13742602e-01 -6.12948596e-01
9.12267983e-01 -3.60964686e-01 -1.39913857e-01 -4.28041995e-01
2.86179572e-01 -7.55491316e-01 2.81091571e-01 2.06362270e-02
-7.51599610e-01 -1.20915091e-02 2.07010746e-01 3.76643032e-01
-4.26457793e-01 -6.82011127e-01 -1.21360473e-01 -4.08408642e-01
-6.67759240e-01 -3.66942197e-01 -6.91058934e-01 4.50294614e-01
7.18731463e-01 -1.16421126e-01 -5.56315958e-01 -4.90692198e-01
-9.26536083e-01 3.16117913e-01 9.16389525e-02 2.14205757e-01
4.47222769e-01 -1.02337086e+00 -4.64181423e-01 -4.11382020e-01
3.19540143e-01 -2.54797876e-01 -1.27495844e-02 6.83232069e-01
-6.48651645e-02 8.58834267e-01 1.54605985e-01 5.25336340e-02
-1.41941833e+00 4.75439012e-01 1.54370395e-02 -5.88090718e-01
-3.24563473e-01 6.72933340e-01 -4.65291679e-01 1.48848072e-01
3.54676157e-01 -2.82170355e-01 -8.42186511e-01 5.93099177e-01
6.20552719e-01 2.88977653e-01 1.58196494e-01 -2.58902997e-01
-2.23724470e-01 4.44202960e-01 -4.80590276e-02 -6.62436128e-01
1.00570893e+00 -4.69915211e-01 -2.09061518e-01 4.62845176e-01
7.53300369e-01 5.95239580e-01 -1.57342240e-01 -4.75466877e-01
5.72240651e-01 -1.59078538e-01 -1.07969437e-02 -1.36766529e+00
-2.25026965e-01 4.77837063e-02 -2.06331804e-01 8.45488608e-01
1.01592469e+00 5.77820241e-01 5.65302491e-01 1.09949505e+00
2.13284522e-01 -1.35307086e+00 1.91900879e-01 8.75737131e-01
9.37972903e-01 -9.09003019e-01 1.04631066e-01 -4.40906852e-01
-7.01580524e-01 1.23420310e+00 2.93174773e-01 7.74808109e-01
4.93018299e-01 -1.66317895e-02 -1.11445179e-02 -9.07591105e-01
-7.79489517e-01 -5.11951804e-01 4.71708298e-01 2.42839724e-01
3.06192249e-01 -3.36066931e-01 -6.01319253e-01 9.77379739e-01
-7.12890208e-01 1.79805592e-01 3.87748897e-01 1.28406274e+00
-6.69595957e-01 -1.24879885e+00 -3.09641331e-01 4.20835733e-01
-7.22094476e-01 -2.78696775e-01 -9.36474204e-01 7.15128839e-01
-9.00787786e-02 1.25967431e+00 -3.31259042e-01 -2.53609300e-01
6.98331773e-01 5.62336087e-01 1.30276740e-01 -8.64610195e-01
-8.56031835e-01 -1.16865508e-01 9.21701252e-01 -3.14155519e-01
-7.46152818e-01 -6.00533664e-01 -9.30705190e-01 -1.26354638e-02
-4.93126512e-01 1.01671147e+00 6.75570011e-01 1.23559439e+00
5.48813462e-01 5.45264959e-01 1.37144849e-01 -1.84514046e-01
-2.23316371e-01 -9.26277697e-01 -8.64077210e-02 1.77072108e-01
1.37250617e-01 -6.49672866e-01 -2.60075957e-01 3.06315750e-01] | [12.106457710266113, 7.920120716094971] |
c58ba8e3-92e9-4694-84a4-a609e0af2995 | spectral-processing-and-optimization-of | 2107.07379 | null | https://arxiv.org/abs/2107.07379v1 | https://arxiv.org/pdf/2107.07379v1.pdf | Spectral Processing and Optimization of Static and Dynamic 3D Geometries | Geometry processing of 3D objects is of primary interest in many areas of computer vision and graphics, including robot navigation, 3D object recognition, classification, feature extraction, etc. The recent introduction of cheap range sensors has created a great interest in many new areas, driving the need for developing efficient algorithms for 3D object processing. Previously, in order to capture a 3D object, expensive specialized sensors were used, such as lasers or dedicated range images, but now this limitation has changed. The current approaches of 3D object processing require a significant amount of manual intervention and they are still time-consuming making them unavailable for use in real-time applications. The aim of this thesis is to present algorithms, mainly inspired by the spectral analysis, subspace tracking, etc, that can be used and facilitate many areas of low-level 3D geometry processing (i.e., reconstruction, outliers removal, denoising, compression), pattern recognition tasks (i.e., significant features extraction) and high-level applications (i.e., registration and identification of 3D objects in partially scanned and cluttered scenes), taking into consideration different types of 3D models (i.e., static and dynamic point clouds, static and dynamic 3D meshes). | ['Gerasimos Arvanitis'] | 2021-07-15 | null | null | null | null | ['3d-object-recognition'] | ['computer-vision'] | [ 3.62595439e-01 -5.22159994e-01 3.46365154e-01 -3.28642845e-01
-1.84792250e-01 -4.34236228e-01 3.85289520e-01 3.14878702e-01
-4.66204822e-01 1.37850955e-01 -5.21293521e-01 -5.25780991e-02
-3.77107173e-01 -7.92949200e-01 -4.07487482e-01 -5.96954525e-01
-1.13698184e-01 7.67286420e-01 4.43443209e-01 -2.25921702e-02
7.27815747e-01 1.56864202e+00 -2.21257687e+00 -4.31373835e-01
6.14935040e-01 1.16346383e+00 2.76137114e-01 3.72857332e-01
-4.48229373e-01 -3.58623028e-01 -2.99402326e-01 -7.40722194e-02
2.62227595e-01 -1.54816508e-01 -2.40985915e-01 5.84070265e-01
1.09838955e-01 3.99958864e-02 6.06281795e-02 1.20066285e+00
4.21647102e-01 3.53423208e-01 7.82359898e-01 -9.17016029e-01
-1.35341704e-01 -3.01175684e-01 -6.57227755e-01 -3.15089613e-01
4.98074859e-01 -2.06236571e-01 2.67949939e-01 -1.33441901e+00
7.32510448e-01 1.37534630e+00 5.87854087e-01 1.96117222e-01
-1.24798465e+00 -3.62250477e-01 -2.78036982e-01 4.02807266e-01
-1.42308486e+00 -3.53139460e-01 1.11473143e+00 -4.71493781e-01
6.66114450e-01 5.44894695e-01 6.98358834e-01 7.08228827e-01
2.77909338e-01 3.67718875e-01 1.15050697e+00 -5.37336290e-01
4.22567993e-01 2.47215673e-01 2.51637101e-01 2.85932153e-01
5.20064950e-01 2.09022872e-02 -3.50813985e-01 -2.30644360e-01
8.72727096e-01 1.29860342e-01 -1.84545204e-01 -8.54964793e-01
-9.50747490e-01 6.41315162e-01 -2.73871876e-04 4.76644665e-01
-5.11535287e-01 -3.70834529e-01 2.61751860e-01 2.83690751e-01
5.33949375e-01 2.21428379e-01 -3.38779509e-01 -1.37338400e-01
-5.58944702e-01 2.54701912e-01 5.89656830e-01 1.09600663e+00
9.57873046e-01 -3.55340093e-02 5.89030206e-01 8.38514626e-01
5.82406759e-01 9.55375373e-01 3.72755796e-01 -6.38337255e-01
1.07764043e-01 8.58956873e-01 -7.50609338e-02 -1.28482735e+00
-6.94272518e-01 -3.28093283e-02 -9.68138814e-01 5.63650191e-01
1.13927938e-01 5.67973673e-01 -8.42915416e-01 9.21550572e-01
8.20431769e-01 -5.55643551e-02 -1.65943459e-01 9.39311564e-01
7.84400582e-01 4.93061692e-01 -6.62088215e-01 -5.06585717e-01
1.10715377e+00 1.16135336e-01 -6.40720844e-01 -1.86711885e-02
3.26674044e-01 -1.15067887e+00 7.97848225e-01 6.32032335e-01
-1.04091585e+00 -6.84312701e-01 -8.08768570e-01 2.01049656e-01
-4.61016476e-01 -2.39008710e-01 2.15818554e-01 5.64936280e-01
-5.26959836e-01 6.00129604e-01 -9.07120526e-01 -5.52999318e-01
3.14176172e-01 4.55431312e-01 -5.87525249e-01 -3.14083338e-01
-5.87519765e-01 9.47748780e-01 2.50894815e-01 3.10350299e-01
-4.60934900e-02 -3.51220280e-01 -7.43203640e-01 -3.96833628e-01
5.27972162e-01 -4.08221096e-01 5.10101438e-01 -3.43724787e-01
-1.47271776e+00 1.24660063e+00 -1.55238912e-01 8.58773198e-03
3.94003481e-01 -3.00027370e-01 -2.18284726e-01 4.70641293e-02
-1.49975583e-01 -2.27303118e-01 1.12308550e+00 -1.24466765e+00
-2.92525500e-01 -1.11243534e+00 -6.24266624e-01 2.71887630e-01
6.69764355e-02 1.20266490e-01 -3.65049601e-01 -3.69983703e-01
1.06033409e+00 -9.89065766e-01 -3.60564470e-01 1.73579767e-01
-2.79881477e-01 -2.04368874e-01 1.21098173e+00 -4.68301684e-01
5.22392631e-01 -2.35558558e+00 3.84814143e-01 8.36659014e-01
-7.76054617e-03 2.61468858e-01 3.59948099e-01 4.15387869e-01
-2.98506524e-02 -2.49435753e-02 -3.56338680e-01 -3.59626561e-01
-1.98172748e-01 1.82608455e-01 -8.03562067e-03 9.08676922e-01
-3.70924696e-02 2.43989542e-01 -4.51100647e-01 -4.41086948e-01
8.29724610e-01 5.53723156e-01 -1.70383796e-01 2.43891999e-02
6.55724257e-02 5.44697702e-01 -6.95321977e-01 9.95595634e-01
9.30734873e-01 2.90987670e-01 -3.83066565e-01 -2.41144538e-01
-5.34830034e-01 -2.38955125e-01 -1.90202034e+00 1.60768008e+00
-2.61736780e-01 3.38625491e-01 3.28908324e-01 -1.14352667e+00
1.49693000e+00 1.76843405e-01 7.75912464e-01 -5.53289652e-01
3.74770254e-01 6.51477158e-01 -2.87287980e-01 -5.28484881e-01
5.34467876e-01 8.21412262e-03 -3.07787172e-02 2.28519678e-01
-5.30245841e-01 -8.73334706e-01 -2.00623393e-01 -3.77950400e-01
6.33451939e-01 5.24975732e-02 4.05390412e-01 -1.06686190e-01
8.09777558e-01 2.67807811e-01 2.13781968e-01 1.59683377e-01
3.98009419e-02 6.59522772e-01 -2.49254167e-01 -2.87506282e-01
-9.86466944e-01 -8.60353708e-01 -3.87584358e-01 1.94036379e-01
5.73688388e-01 1.22081995e-01 -2.99975574e-01 1.61106717e-02
3.67516369e-01 3.62289906e-01 -5.08949570e-02 -1.96168572e-01
-8.94461930e-01 -5.22186995e-01 -2.14169636e-01 4.27066796e-02
3.96620244e-01 -8.07837129e-01 -1.00451231e+00 3.39351833e-01
4.32594657e-01 -1.05942595e+00 2.03176320e-01 7.21800402e-02
-1.66479754e+00 -1.07657063e+00 -5.15519679e-01 -5.61168492e-01
6.54684782e-01 6.67186499e-01 1.04629421e+00 1.27205908e-01
-6.90062344e-01 6.37861073e-01 -4.34237123e-01 -6.73135400e-01
-1.67257935e-01 -3.39818716e-01 4.36892271e-01 1.53633608e-02
2.76303470e-01 -7.55494952e-01 -1.33393899e-01 7.14641869e-01
-8.45960319e-01 -4.08169270e-01 7.05069602e-01 5.08561075e-01
1.23055923e+00 2.56064147e-01 3.41732241e-02 -6.35839939e-01
3.93088102e-01 6.88587427e-02 -9.58988190e-01 -1.13758504e-01
-3.54785860e-01 -3.74637306e-01 3.49078000e-01 -4.22869831e-01
-6.21872962e-01 2.95040160e-01 -3.46175522e-01 -7.62873650e-01
-6.74429357e-01 4.18662399e-01 -4.78450328e-01 -3.92150491e-01
6.98171794e-01 3.62384111e-01 3.73365462e-01 -8.74374211e-01
1.13892607e-01 7.59337306e-01 3.58850151e-01 -1.12102449e-01
1.33986568e+00 6.22470617e-01 6.82830572e-01 -1.67691958e+00
-9.98218581e-02 -8.33490252e-01 -1.05220377e+00 -3.86808574e-01
6.01489365e-01 -4.75015551e-01 -5.40140629e-01 6.22073770e-01
-1.22114420e+00 3.39826524e-01 -5.06580174e-01 7.73031950e-01
-7.23921120e-01 4.56100851e-01 -2.95353830e-02 -1.00545418e+00
-3.41148186e-03 -1.05565143e+00 1.03835392e+00 1.70123383e-01
-1.49846897e-01 -7.09595382e-01 -6.15575239e-02 1.50319144e-01
1.02726415e-01 5.87040782e-01 9.61272001e-01 -2.51204401e-01
-6.90659344e-01 -5.75433314e-01 1.08975828e-01 1.72580376e-01
2.73749620e-01 1.54033020e-01 -7.75217712e-01 -1.15419157e-01
4.48628634e-01 3.61754186e-02 1.34198830e-01 2.95009583e-01
1.01515007e+00 3.67755413e-01 -3.92958671e-01 5.15842915e-01
1.49462068e+00 4.80460674e-01 7.13012516e-01 1.80314466e-01
4.66926605e-01 8.12750399e-01 1.07087064e+00 4.93759453e-01
-2.27557138e-01 9.34874892e-01 7.63991177e-01 3.70750949e-02
1.77674264e-01 3.52448523e-01 -7.51798600e-02 8.37445498e-01
-6.31481767e-01 1.55587286e-01 -8.10871124e-01 8.58581811e-02
-1.52820325e+00 -8.04538250e-01 -7.04512060e-01 2.61533594e+00
2.12319568e-01 1.90634757e-01 -7.22271279e-02 8.76498878e-01
7.12072968e-01 -1.67409420e-01 -6.32828236e-01 -4.27602857e-01
7.78385531e-03 2.49705702e-01 3.89746159e-01 2.21244648e-01
-9.09724534e-01 4.83670324e-01 4.83603954e+00 6.45306051e-01
-1.24595904e+00 -3.15405428e-01 -2.08799019e-01 3.98782015e-01
-3.09547652e-02 -3.33291978e-01 -7.41163492e-01 3.45747292e-01
4.60531205e-01 -2.89381705e-02 1.07477523e-01 9.58196640e-01
2.94640630e-01 -4.28450108e-01 -9.84528661e-01 1.54352999e+00
1.45040333e-01 -8.27164054e-01 -1.18972912e-01 3.31561983e-01
3.51047784e-01 -2.76323587e-01 -9.95004550e-02 -2.60116130e-01
-6.33148789e-01 -6.97387099e-01 7.31796265e-01 7.27547824e-01
5.35277843e-01 -7.74354160e-01 8.03454220e-01 6.73257411e-01
-1.02840459e+00 3.69630307e-01 -5.87154210e-01 5.66273034e-02
4.37350005e-01 1.14206851e+00 -3.75452459e-01 8.32997143e-01
8.71954322e-01 5.93813181e-01 -2.50519246e-01 1.44309199e+00
1.39341101e-01 4.72784191e-02 -6.59887493e-01 -2.17275843e-01
-1.45719364e-01 -7.16181338e-01 1.01275599e+00 6.16465032e-01
5.36548018e-01 4.11696374e-01 8.46378505e-02 7.44746089e-01
3.26943219e-01 3.42582583e-01 -8.11392128e-01 2.78703988e-01
3.09446752e-01 9.54784095e-01 -1.12200260e+00 -4.26934846e-02
-4.04896408e-01 7.62298286e-01 -4.93449837e-01 -3.01613789e-02
-4.04695362e-01 -7.60803998e-01 6.98461175e-01 3.81982774e-01
9.28067341e-02 -1.00691211e+00 -5.96604466e-01 -8.38213623e-01
4.06013913e-02 -4.49960083e-01 5.55242710e-02 -5.15745342e-01
-1.05247903e+00 2.16028526e-01 3.19431096e-01 -1.55180550e+00
-2.25558177e-01 -8.14684331e-01 -3.62804353e-01 8.73628974e-01
-1.27302265e+00 -4.53404605e-01 -4.81182724e-01 8.63532543e-01
5.88977456e-01 7.95459747e-02 7.45080471e-01 3.04073274e-01
-1.63016081e-01 -9.46850330e-02 3.44451308e-01 -3.38977337e-01
4.72849160e-01 -8.08794439e-01 -8.36026762e-03 6.62580431e-01
1.00114562e-01 2.70836622e-01 7.30438471e-01 -6.00234568e-01
-2.09882355e+00 -7.23603666e-01 6.35344267e-01 -2.71572977e-01
1.66372716e-01 -5.69234073e-01 -1.07265615e+00 7.70953000e-02
-8.21040869e-01 -9.92568210e-02 3.27881426e-01 -2.35539511e-01
2.97144592e-01 -1.36315227e-01 -1.43589509e+00 3.35490108e-01
1.21047127e+00 -1.79524437e-01 -7.38095582e-01 2.56069660e-01
2.11496741e-01 -6.36681020e-01 -9.03940618e-01 5.44969738e-01
4.92521733e-01 -1.11987019e+00 1.22615874e+00 3.12666744e-02
-3.02353323e-01 -4.16443229e-01 -1.20614991e-01 -9.65149641e-01
-1.12307526e-01 -2.44361714e-01 6.20579123e-02 8.25639904e-01
-1.36843368e-01 -8.57735395e-01 9.05388951e-01 6.33335710e-01
-5.96865773e-01 -4.38212544e-01 -1.27394402e+00 -1.00062096e+00
-5.52529633e-01 -6.15271926e-01 3.94505918e-01 8.14220130e-01
-5.39437592e-01 2.22824998e-02 3.46214846e-02 3.31500322e-01
1.00204110e+00 6.84647977e-01 1.00321984e+00 -2.04778886e+00
2.65199304e-01 -4.63561922e-01 -1.11847651e+00 -9.22017753e-01
-1.54501438e-01 -5.18621325e-01 1.34934084e-02 -1.47857690e+00
-5.50164759e-01 -6.82098806e-01 5.19635618e-01 -1.57992944e-01
3.86820406e-01 3.64549726e-01 7.88053051e-02 4.91097152e-01
8.25672075e-02 5.82155645e-01 1.21678925e+00 1.07801966e-01
-5.52956641e-01 5.11854470e-01 -4.20387164e-02 1.01887250e+00
5.76206028e-01 -2.72525042e-01 -7.82196075e-02 -2.19177872e-01
1.68783203e-01 -1.59684479e-01 3.36069524e-01 -1.14110208e+00
2.04944819e-01 -2.78237253e-01 5.16687870e-01 -1.16101408e+00
7.00972438e-01 -1.54830897e+00 3.49763125e-01 5.60134172e-01
5.30792356e-01 -2.64739674e-02 1.02819555e-01 4.16912198e-01
-2.79718131e-01 -5.72739899e-01 9.26034987e-01 -3.11037421e-01
-8.04633021e-01 3.16834927e-01 -3.46701115e-01 -4.78509516e-01
1.30855024e+00 -9.99068677e-01 4.50479895e-01 3.25093493e-02
-8.68978560e-01 -1.23848222e-01 7.86944568e-01 -1.68089513e-02
1.09955263e+00 -1.26694787e+00 -5.20004094e-01 5.53865790e-01
2.73846462e-02 6.99912131e-01 2.49509186e-01 7.05917895e-01
-5.63722253e-01 7.22431466e-02 -2.53987342e-01 -1.10243547e+00
-1.63340533e+00 4.27924246e-01 -7.07418621e-02 2.80396342e-01
-7.35034883e-01 4.74381536e-01 -4.76275921e-01 -2.76108116e-01
2.15286031e-01 -6.03487968e-01 -3.04326028e-01 1.58869684e-01
2.43076354e-01 8.54097664e-01 4.72989529e-01 -8.76022041e-01
-5.22798061e-01 1.66696751e+00 4.76299584e-01 2.23416790e-01
1.43513036e+00 -7.51715824e-02 -2.06500754e-01 6.69459343e-01
9.84594345e-01 -2.99656522e-02 -5.46868086e-01 -1.89966410e-01
1.66495949e-01 -9.05561924e-01 -2.57990062e-02 -2.51030400e-02
-6.37194037e-01 1.02900326e+00 7.38443315e-01 5.38980544e-01
1.17347074e+00 2.29333639e-01 4.04831290e-01 5.63285768e-01
9.00140524e-01 -1.02608240e+00 -1.75864920e-01 6.56337857e-01
1.13953197e+00 -1.03609884e+00 4.24910843e-01 -8.23854327e-01
7.35653192e-02 1.44072652e+00 9.63325799e-02 -1.09861568e-01
8.70166361e-01 1.23279750e-01 -1.87479585e-01 -2.98775107e-01
7.98305273e-02 -3.05646032e-01 2.89481491e-01 6.50497317e-01
1.67645179e-02 -1.39977008e-01 -2.93458223e-01 -6.06801361e-02
-8.01858604e-02 -2.64466703e-01 4.59543556e-01 1.25568557e+00
-7.19014049e-01 -1.07134557e+00 -1.19824433e+00 5.31243742e-01
-1.98068796e-03 6.29075766e-01 -3.03679526e-01 9.93756115e-01
6.90666065e-02 5.96210420e-01 1.76538691e-01 -2.64296710e-01
9.99868095e-01 4.30295207e-02 6.10344052e-01 -4.96882766e-01
-1.01830699e-01 1.72126129e-01 -2.43502691e-01 -7.52213180e-01
-7.36841917e-01 -1.04593694e+00 -1.09058130e+00 -1.38112292e-01
-5.42364419e-01 -6.87591285e-02 1.49557948e+00 6.11031830e-01
2.95700524e-02 -7.19540641e-02 8.61358941e-01 -1.40285933e+00
-4.52581495e-01 -6.67951405e-01 -1.00019717e+00 3.88434500e-01
-1.29295856e-01 -1.07836342e+00 -4.67635930e-01 -6.63337037e-02] | [8.125543594360352, -2.781708240509033] |
9e5ff328-3600-4614-a631-8978f2a72736 | potential-based-reward-shaping-for-learning | 2302.10720 | null | https://arxiv.org/abs/2302.10720v2 | https://arxiv.org/pdf/2302.10720v2.pdf | Learning to Play Text-based Adventure Games with Maximum Entropy Reinforcement Learning | Text-based games are a popular testbed for language-based reinforcement learning (RL). In previous work, deep Q-learning is commonly used as the learning agent. Q-learning algorithms are challenging to apply to complex real-world domains due to, for example, their instability in training. Therefore, in this paper, we adapt the soft-actor-critic (SAC) algorithm to the text-based environment. To deal with sparse extrinsic rewards from the environment, we combine it with a potential-based reward shaping technique to provide more informative (dense) reward signals to the RL agent. We apply our method to play difficult text-based games. The SAC method achieves higher scores than the Q-learning methods on many games with only half the number of training steps. This shows that it is well-suited for text-based games. Moreover, we show that the reward shaping technique helps the agent to learn the policy faster and achieve higher scores. In particular, we consider a dynamically learned value function as a potential function for shaping the learner's original sparse reward signals. | ['Sophie Fellenz', 'Rati Devidze', 'Weichen Li'] | 2023-02-21 | null | null | null | null | ['text-based-games'] | ['playing-games'] | [-2.32806981e-01 1.09743483e-01 -1.38506949e-01 1.47400603e-01
-7.80054212e-01 -3.99729520e-01 4.85928595e-01 2.38274917e-01
-1.00249815e+00 1.12854195e+00 -6.23182580e-02 -2.00119451e-01
-1.46137536e-01 -7.79984474e-01 -7.57355988e-01 -8.39145958e-01
-1.57695860e-01 6.67752862e-01 3.66404146e-01 -8.42615604e-01
2.94970840e-01 3.18353564e-01 -1.44011462e+00 -1.11093059e-01
8.92472267e-01 6.85507834e-01 4.79929209e-01 7.38300681e-01
-1.44621342e-01 1.28441513e+00 -8.88555706e-01 -1.14614114e-01
5.06902039e-01 -8.52832198e-01 -4.70144898e-01 -8.46893266e-02
-3.06898266e-01 -3.60972822e-01 -1.04839921e-01 1.04401958e+00
7.57195473e-01 6.16326451e-01 2.87371337e-01 -1.16146111e+00
2.49365252e-02 6.81407392e-01 -5.19525111e-01 6.42610639e-02
3.80238652e-01 3.15935910e-01 1.01948440e+00 -5.74488223e-01
6.61255002e-01 1.42823017e+00 1.51061043e-01 9.23125923e-01
-1.15061903e+00 -5.14716148e-01 3.16140682e-01 2.97648787e-01
-7.93757796e-01 -1.45857692e-01 8.11877489e-01 -1.01757245e-02
7.75815189e-01 -5.75893298e-02 1.09687269e+00 9.51884389e-01
1.73907056e-01 1.21034813e+00 1.30174661e+00 -5.51283419e-01
8.79004300e-01 1.19072713e-01 -8.00454259e-01 6.50232732e-01
-1.65354073e-01 6.20711863e-01 -4.60194111e-01 -1.27288043e-01
9.58606541e-01 -1.12328596e-01 2.50100583e-01 -7.19643533e-01
-8.68489265e-01 1.33784974e+00 4.68033880e-01 2.63262004e-01
-7.03631639e-01 6.08335912e-01 3.89633864e-01 6.94114923e-01
2.22964078e-01 9.62691367e-01 -2.33517259e-01 -6.75396681e-01
-7.03657508e-01 7.53426373e-01 7.98677385e-01 3.09810698e-01
7.17127204e-01 5.00405014e-01 -3.25622201e-01 9.98914063e-01
2.77724385e-01 4.81257170e-01 6.81625307e-01 -1.18181229e+00
3.15197021e-01 2.05954194e-01 3.14840287e-01 -4.69995677e-01
-3.42866242e-01 -4.93636400e-01 -4.50075626e-01 8.86699498e-01
6.36288822e-01 -5.85834026e-01 -6.11840069e-01 1.66729891e+00
4.25274223e-01 1.07417189e-01 2.22768515e-01 1.00756335e+00
1.97477669e-01 6.77544296e-01 -1.30239591e-01 -3.08756202e-01
8.32465351e-01 -1.04700494e+00 -6.10189676e-01 -3.73894423e-01
6.61040843e-01 -3.68788540e-01 1.20088744e+00 6.91078246e-01
-1.36662281e+00 -2.99902320e-01 -7.48406351e-01 6.50268555e-01
-7.90587291e-02 -3.26340348e-01 4.33255732e-01 5.44595182e-01
-1.09628689e+00 9.00240302e-01 -7.04214811e-01 -5.73016889e-02
3.14401299e-01 4.88795489e-01 1.08255178e-01 6.16759956e-02
-1.38735700e+00 9.67509568e-01 3.17342609e-01 -4.41568732e-01
-1.15999210e+00 -2.75911063e-01 -7.30727851e-01 9.58570391e-02
9.16422427e-01 -3.53181064e-01 1.75663090e+00 -1.34862638e+00
-2.49988747e+00 3.21739137e-01 3.80449653e-01 -6.28450155e-01
7.09684730e-01 1.95219237e-02 2.67441362e-01 2.24297807e-01
-8.65090564e-02 6.44708633e-01 1.30689991e+00 -1.13209605e+00
-6.54779017e-01 -3.47882174e-02 4.79416579e-01 5.30517459e-01
-7.74819404e-02 -5.85360266e-02 1.03103355e-01 -7.16702640e-01
-5.21654129e-01 -8.72357309e-01 -8.07696998e-01 -1.24516271e-01
1.57677203e-01 -2.80903518e-01 4.20465320e-01 -9.84926615e-03
7.97760546e-01 -1.94699967e+00 2.91682929e-01 3.07894737e-01
1.98489711e-01 5.24433970e-01 -5.62706530e-01 6.82329416e-01
1.77701414e-01 -2.34210625e-01 -4.41247597e-02 -1.80346340e-01
1.65386021e-01 5.92501342e-01 -7.49283955e-02 1.84036091e-01
2.33941153e-01 1.05766153e+00 -1.37076306e+00 -2.37376288e-01
1.40200198e-01 4.89941426e-02 -8.16745162e-01 4.59122062e-01
-6.34436071e-01 5.63059449e-01 -7.02539682e-01 8.73793364e-02
1.88487872e-01 1.92218900e-01 2.14266062e-01 7.63359010e-01
5.84052429e-02 5.51843829e-02 -1.29678750e+00 1.59199190e+00
-7.03891277e-01 2.94080406e-01 3.34049582e-01 -1.19946456e+00
1.17978215e+00 9.75692198e-02 7.05539644e-01 -1.28961682e+00
1.14135511e-01 2.54971623e-01 2.91264445e-01 -3.72462004e-01
4.99852300e-01 -4.92288738e-01 4.56936583e-02 7.42262423e-01
3.36220115e-02 -5.42011917e-01 5.17338514e-01 1.84552535e-01
1.24499059e+00 2.38309175e-01 3.40207160e-01 -2.33388413e-02
3.98844302e-01 -1.02516510e-01 5.09104967e-01 9.97240543e-01
-2.17899084e-01 8.34580138e-02 8.64094794e-01 -2.60034621e-01
-1.00418818e+00 -7.00300097e-01 5.37822247e-01 1.50032842e+00
-1.06657468e-01 -3.58979076e-01 -5.26398778e-01 -8.76402736e-01
1.23762116e-01 5.25037110e-01 -4.75011408e-01 -3.65958482e-01
-6.33575976e-01 -4.43804115e-01 3.11809510e-01 3.40329379e-01
2.05565944e-01 -1.55743217e+00 -1.05502045e+00 8.10067713e-01
2.37865970e-01 -6.48983538e-01 -3.36998999e-01 4.89877462e-01
-8.33881915e-01 -8.46462727e-01 -1.14467227e+00 -5.44643044e-01
3.80605429e-01 -1.24775231e-01 1.01558506e+00 6.10851170e-03
1.03467695e-01 5.99705219e-01 -6.01091325e-01 -5.19928455e-01
-7.73106277e-01 -1.41525954e-01 -3.08505278e-02 -2.18325555e-01
-1.49820551e-01 -4.16007519e-01 -3.92490506e-01 1.99027196e-01
-1.06090939e+00 -1.81603387e-01 5.81682146e-01 1.32860577e+00
3.65608692e-01 2.59854663e-02 9.74515319e-01 -7.28432059e-01
1.28994334e+00 -3.51421803e-01 -9.86414909e-01 -1.42539099e-01
-5.45748174e-01 5.08619964e-01 1.09446383e+00 -8.43988836e-01
-7.08536208e-01 1.20361261e-01 -3.33817124e-01 -3.60425383e-01
1.57498196e-01 5.03477871e-01 4.47090149e-01 -2.80380070e-01
8.84501100e-01 2.48058304e-01 3.65921438e-01 -5.77921346e-02
3.48482966e-01 3.13216746e-01 -1.76344991e-01 -8.40277672e-01
8.58150125e-01 -4.68630567e-02 1.02025688e-01 -5.50014675e-01
-5.35304427e-01 -2.49881849e-01 -3.38424556e-02 -4.15171146e-01
2.42157415e-01 -6.60021305e-01 -1.00751960e+00 2.74243325e-01
-7.18621075e-01 -1.04976118e+00 -9.36221600e-01 5.86975157e-01
-1.15679550e+00 1.62964389e-01 -4.61922824e-01 -1.19383168e+00
-9.46273431e-02 -1.31128180e+00 7.38180220e-01 2.98897445e-01
2.90416837e-01 -1.06412864e+00 5.30314207e-01 -1.40522867e-01
6.25473142e-01 1.03447728e-01 4.10848558e-01 -4.80945617e-01
-1.70719877e-01 2.20733836e-01 2.54846424e-01 2.27710858e-01
-2.11015165e-01 -3.69360119e-01 -6.31353676e-01 -5.45364738e-01
2.01969445e-02 -9.97222483e-01 7.60036230e-01 3.36974561e-01
9.76399839e-01 -3.74770999e-01 4.57027435e-01 1.54862538e-01
1.26457202e+00 3.69051456e-01 4.94778931e-01 6.58187032e-01
1.65889457e-01 4.89831358e-01 9.60190892e-01 9.63597596e-01
1.95499420e-01 8.12165976e-01 7.54362166e-01 -7.77695179e-02
3.57219249e-01 -3.90031070e-01 8.44227374e-01 5.16301632e-01
-9.55997258e-02 1.52221009e-01 -6.53454840e-01 2.42695093e-01
-2.30128717e+00 -1.06542909e+00 5.45237124e-01 2.20160460e+00
1.22147644e+00 2.47455895e-01 8.08098912e-01 1.65774837e-01
2.77070016e-01 8.46706033e-02 -8.33814144e-01 -7.08153069e-01
1.34277329e-01 6.27488375e-01 4.05812234e-01 4.88127381e-01
-6.47365034e-01 1.11973441e+00 6.04323769e+00 1.12397528e+00
-1.21552885e+00 -1.73523724e-02 3.90963078e-01 -3.18758637e-01
-1.60011709e-01 -2.28166252e-01 -3.76062721e-01 3.37091863e-01
9.67271745e-01 -2.87041336e-01 1.07894468e+00 8.99949670e-01
5.72090209e-01 -3.97947073e-01 -7.91891515e-01 8.99185121e-01
-3.34968805e-01 -1.12409425e+00 -4.11552846e-01 7.53870979e-02
7.77115405e-01 7.26376027e-02 8.03689584e-02 7.80756116e-01
8.60125422e-01 -1.05772066e+00 7.93258250e-01 1.90025508e-01
5.03926635e-01 -1.08370602e+00 6.53461218e-01 7.04157889e-01
-9.27713811e-01 -4.60826069e-01 -5.43774128e-01 -3.25170338e-01
-2.05515146e-01 2.71250904e-01 -8.67600143e-01 2.86726564e-01
2.23879933e-01 5.46125948e-01 -2.90382713e-01 1.18825150e+00
-4.63805497e-01 6.15343571e-01 -2.82501876e-01 -6.93724632e-01
8.22420478e-01 -3.67792398e-01 5.36812127e-01 8.20490956e-01
1.68124765e-01 -1.45375699e-01 5.83709896e-01 7.69149303e-01
8.53604972e-02 3.51228625e-01 -8.06656420e-01 -1.59674257e-01
5.08067645e-02 1.07970941e+00 -5.45640588e-01 -1.03925198e-01
-3.67179848e-02 6.42857552e-01 6.43086314e-01 3.87863636e-01
-5.68199337e-01 -3.69803548e-01 4.53774124e-01 -7.78937042e-02
3.82816434e-01 -3.12427908e-01 3.84325594e-01 -1.10834849e+00
-4.02507126e-01 -1.30024481e+00 1.50569215e-01 -5.16963363e-01
-1.11769235e+00 3.43317479e-01 -1.64518371e-01 -1.14336765e+00
-8.19081724e-01 -3.65722477e-01 -5.29463530e-01 5.81616402e-01
-1.79947293e+00 -3.57310385e-01 2.21945286e-01 9.41332638e-01
5.91964006e-01 -4.90987062e-01 7.04685390e-01 -5.93010969e-02
-3.00429910e-01 4.46652621e-01 5.21769583e-01 -4.08403017e-02
6.06713414e-01 -1.67131257e+00 6.61882162e-02 3.33541691e-01
2.77938843e-01 -4.61372845e-02 8.34087729e-01 -3.27876180e-01
-1.61131191e+00 -7.30572224e-01 -2.68070158e-02 2.05419034e-01
8.93452525e-01 -3.33999902e-01 -6.38363004e-01 4.36941907e-02
5.97096197e-02 -9.96660162e-03 2.77770728e-01 -1.67116806e-01
3.74782160e-02 -2.72083342e-01 -1.18913651e+00 6.04312599e-01
5.28839648e-01 -2.39565685e-01 -3.39885592e-01 2.50372529e-01
3.86272967e-01 -5.53444803e-01 -5.83181560e-01 -2.48055831e-01
2.32214510e-01 -8.38904083e-01 7.03292966e-01 -7.00413704e-01
3.64770710e-01 -4.99163643e-02 2.08320618e-01 -2.08737946e+00
-2.71049321e-01 -9.66327608e-01 -2.12426767e-01 4.95675474e-01
1.89077213e-01 -7.56193697e-01 1.02857852e+00 6.78199232e-02
3.33861351e-01 -9.39463615e-01 -1.05967307e+00 -9.90389466e-01
4.36944187e-01 -3.32937181e-01 3.38260621e-01 5.67322493e-01
3.22783947e-01 2.05347359e-01 -5.66336334e-01 -6.41564071e-01
5.30747235e-01 -1.65197313e-01 8.35266232e-01 -9.98608947e-01
-8.70655715e-01 -5.61707795e-01 -5.09842858e-02 -1.14204741e+00
1.13528587e-01 -6.33923471e-01 2.59932905e-01 -1.19561386e+00
-1.77205443e-01 -6.35755122e-01 -3.12668264e-01 4.23850238e-01
-1.89813986e-01 -1.15733154e-01 8.30988169e-01 -1.43292740e-01
-8.79010618e-01 1.02111566e+00 1.79200315e+00 -1.26104146e-01
-5.73254168e-01 3.31737995e-01 -5.07071435e-01 3.67408395e-01
1.12315953e+00 -7.31673300e-01 -5.15839159e-01 -6.21950952e-04
6.52375519e-01 5.69477022e-01 9.42569524e-02 -8.12037110e-01
1.86093912e-01 -6.11515582e-01 -1.42735448e-02 2.91199666e-02
4.26738679e-01 -6.79542005e-01 -5.97965181e-01 8.96564960e-01
-5.88891804e-01 4.37724497e-03 1.94331497e-01 5.36480129e-01
-2.67583698e-01 -6.10018790e-01 9.32566643e-01 -4.53156471e-01
-2.70432204e-01 1.54027924e-01 -9.06478167e-01 4.53794122e-01
1.00048113e+00 -4.07251492e-02 7.93989003e-02 -9.80486870e-01
-6.56206191e-01 4.78396982e-01 2.12647155e-01 8.25277492e-02
8.13329577e-01 -1.25113225e+00 -7.21756816e-01 9.87926424e-02
-1.22721627e-01 -1.21535867e-01 -2.25222766e-01 6.73080146e-01
-2.63527185e-01 7.39616677e-02 -5.80964625e-01 -2.83208460e-01
-9.20558810e-01 3.66019577e-01 6.44091964e-01 -8.08189154e-01
-5.73756397e-01 4.88088191e-01 -1.24679431e-01 -4.60812837e-01
4.38266933e-01 -2.66308397e-01 -3.33474755e-01 2.55692117e-02
5.16347408e-01 3.11760366e-01 -2.14949429e-01 7.66991526e-02
9.97637808e-02 2.95623958e-01 5.22006601e-02 -6.85812056e-01
1.53198218e+00 2.82420069e-01 4.08565193e-01 3.32924664e-01
6.12415254e-01 -1.08925916e-01 -1.67305243e+00 -3.74211520e-01
4.00504731e-02 -4.44631904e-01 9.83908325e-02 -7.17954457e-01
-1.12049294e+00 7.92351007e-01 4.84687239e-01 5.19477963e-01
1.01665330e+00 -3.98430884e-01 4.93317991e-01 7.81631052e-01
5.98614573e-01 -1.48061955e+00 6.76140666e-01 8.11228812e-01
7.84156740e-01 -1.12527418e+00 -3.03059220e-01 4.46339935e-01
-9.69945133e-01 1.30588508e+00 4.87904519e-01 -4.56705213e-01
2.76970118e-01 2.39277676e-01 1.13418624e-01 1.34589270e-01
-1.00812316e+00 -5.44537604e-01 -1.92751333e-01 7.71071017e-01
1.25515265e-02 -5.31680174e-02 -2.58585632e-01 9.66640785e-02
-6.14407919e-02 4.18199822e-02 8.50279808e-01 1.05292940e+00
-7.25522339e-01 -1.68244946e+00 -4.69931006e-01 4.09893990e-01
-3.84376049e-01 8.26599225e-02 -3.11471373e-01 5.18266141e-01
-3.33305627e-01 1.00224900e+00 -2.88284838e-01 -6.50784820e-02
3.02312076e-01 -2.94193596e-01 7.06654727e-01 -7.56482303e-01
-1.05367804e+00 2.01913252e-01 -1.18402794e-01 -9.55637097e-01
-3.56851846e-01 -4.88733828e-01 -1.39906263e+00 -1.36607781e-01
-8.00737888e-02 4.87897187e-01 6.77234471e-01 9.28795636e-01
7.94894174e-02 5.51226616e-01 9.46854472e-01 -8.89313698e-01
-1.11563087e+00 -7.64223874e-01 -8.02383840e-01 2.49612018e-01
3.97895187e-01 -7.85442233e-01 -1.42848939e-01 -6.81574404e-01] | [3.9415717124938965, 1.8335745334625244] |
9261c057-6e3a-41c3-80fd-37a1a29fba99 | neural-surface-reconstruction-of-dynamic | 2206.15258 | null | https://arxiv.org/abs/2206.15258v2 | https://arxiv.org/pdf/2206.15258v2.pdf | Neural Surface Reconstruction of Dynamic Scenes with Monocular RGB-D Camera | We propose Neural-DynamicReconstruction (NDR), a template-free method to recover high-fidelity geometry and motions of a dynamic scene from a monocular RGB-D camera. In NDR, we adopt the neural implicit function for surface representation and rendering such that the captured color and depth can be fully utilized to jointly optimize the surface and deformations. To represent and constrain the non-rigid deformations, we propose a novel neural invertible deforming network such that the cycle consistency between arbitrary two frames is automatically satisfied. Considering that the surface topology of dynamic scene might change over time, we employ a topology-aware strategy to construct the topology-variant correspondence for the fused frames. NDR also further refines the camera poses in a global optimization manner. Experiments on public datasets and our collected dataset demonstrate that NDR outperforms existing monocular dynamic reconstruction methods. | ['Juyong Zhang', 'Yan Wang', 'Xuetao Feng', 'Wanquan Feng', 'Hongrui Cai'] | 2022-06-30 | null | null | null | null | ['rgb-d-reconstruction'] | ['computer-vision'] | [ 1.31335825e-01 -1.94718674e-01 5.52921891e-02 -4.37979996e-01
-4.15333152e-01 -8.41491759e-01 3.22418571e-01 -9.43709612e-01
-1.21409610e-01 4.89388764e-01 1.58070043e-01 4.53854166e-02
-2.74582505e-02 -7.96029031e-01 -1.01654065e+00 -6.49729192e-01
3.90988678e-01 2.15638950e-01 2.25820974e-01 -2.61258427e-02
-4.03428636e-02 9.31468010e-01 -1.44138634e+00 -7.50100985e-02
5.81011832e-01 9.40119267e-01 3.40213746e-01 5.95155716e-01
2.32421249e-01 5.18168688e-01 -6.60949126e-02 -6.77414984e-02
7.33525634e-01 -1.24958456e-01 -6.31740630e-01 5.70496500e-01
7.29090393e-01 -7.48252809e-01 -9.25855994e-01 1.04045129e+00
4.39838707e-01 3.77590507e-01 2.38297600e-02 -5.85372269e-01
-5.53664029e-01 -2.45697975e-01 -6.54891849e-01 -1.84074417e-01
6.95937276e-01 2.06909612e-01 4.64419901e-01 -1.10773289e+00
1.20554674e+00 1.44469476e+00 3.95262718e-01 5.61404705e-01
-1.05875814e+00 -4.68665242e-01 5.51549196e-01 -6.90609589e-02
-1.58018661e+00 -5.59345245e-01 1.30719745e+00 -2.92300880e-01
6.00500464e-01 2.93237954e-01 9.31939006e-01 8.62784863e-01
5.18455617e-02 3.50301743e-01 7.48514473e-01 -1.12982333e-01
-9.08315852e-02 -7.27097273e-01 -5.10069132e-01 8.03168952e-01
6.11043759e-02 3.71503562e-01 -6.60422981e-01 -3.37283090e-02
1.56558228e+00 3.21985155e-01 -6.37383223e-01 -6.64718211e-01
-1.41586256e+00 2.54618019e-01 3.54444683e-01 4.64367401e-03
-3.10664207e-01 3.83405417e-01 -8.61863717e-02 1.73642099e-01
5.14369071e-01 1.62997440e-01 -6.10196710e-01 -4.03117426e-02
-5.40552616e-01 1.68827772e-01 4.36656147e-01 1.06273162e+00
1.08633602e+00 3.56469065e-01 1.94099337e-01 5.90221226e-01
5.37250996e-01 6.73745632e-01 2.47818470e-01 -1.72777593e+00
2.51217425e-01 7.04562485e-01 2.12856352e-01 -1.33404815e+00
-2.08591565e-01 -5.37020601e-02 -8.68048429e-01 2.94815074e-03
7.66931996e-02 -5.21255918e-02 -9.80926335e-01 1.58543193e+00
8.28365982e-01 5.34989893e-01 -2.45274261e-01 1.23742104e+00
7.92202652e-01 5.05994439e-01 -6.49190307e-01 -3.56928736e-01
8.59607995e-01 -6.73772812e-01 -7.47082412e-01 -1.18362352e-01
5.51977754e-02 -6.42379105e-01 8.07966292e-01 3.46560091e-01
-1.24323583e+00 -3.88695121e-01 -1.01767647e+00 -4.92500186e-01
3.02061737e-01 7.03963041e-02 5.20537019e-01 4.78868745e-02
-1.00808489e+00 5.63712358e-01 -1.26653600e+00 1.22921199e-01
-1.26540950e-02 4.22211468e-01 -4.64687198e-01 -3.58707398e-01
-7.28565514e-01 3.01559657e-01 1.82032600e-01 5.47406197e-01
-9.41158116e-01 -6.39203489e-01 -9.40509737e-01 -4.49341089e-01
5.59773147e-01 -1.02126515e+00 9.39501166e-01 -7.88123429e-01
-2.10073113e+00 1.02838147e+00 -3.82440537e-01 2.35050425e-01
7.33097196e-01 -3.28726768e-01 -2.90048216e-02 2.75562286e-01
-2.89120823e-01 4.52525556e-01 7.85711467e-01 -1.45389950e+00
-1.59344643e-01 -3.94937843e-01 3.33513439e-01 5.59114397e-01
1.02417609e-02 -3.54556561e-01 -1.16088414e+00 -6.93080723e-01
9.49953437e-01 -1.10295248e+00 -3.74534726e-01 6.10942304e-01
-5.07834673e-01 3.95098716e-01 1.07925904e+00 -7.13309944e-01
7.98863113e-01 -1.94202423e+00 7.65331328e-01 9.55973491e-02
1.90973833e-01 -1.62958369e-01 -2.73600500e-03 -1.62615225e-01
1.05085485e-01 -2.91252166e-01 -2.58736312e-01 -5.10227740e-01
-2.88676769e-01 6.00916028e-01 -1.81304455e-01 7.72346258e-01
-1.47569895e-01 9.73785818e-01 -7.57258892e-01 -2.28483826e-01
4.86180842e-01 8.21689963e-01 -7.02132344e-01 5.19046307e-01
-3.74664515e-01 9.87683713e-01 -6.36191130e-01 8.65852237e-01
9.48256552e-01 -2.42320627e-01 1.71861544e-01 -5.42908907e-01
-2.81546295e-01 4.84972708e-02 -1.54593134e+00 2.53981328e+00
-3.34914148e-01 3.93358499e-01 3.68172348e-01 -5.91733873e-01
8.88609886e-01 7.77745843e-02 7.65828669e-01 -5.80776930e-01
2.32509196e-01 1.13610230e-01 -4.44653243e-01 -4.44262743e-01
6.21467352e-01 1.49717152e-01 2.90384710e-01 1.67299032e-01
-4.54322368e-01 -3.70983541e-01 -3.86508256e-01 -2.01237559e-01
6.45857394e-01 8.82640481e-01 -1.29460216e-01 6.25491515e-02
6.95356429e-01 -2.94590890e-01 1.12320578e+00 1.10559858e-01
2.03449443e-01 1.02225292e+00 -1.17692672e-01 -8.13549221e-01
-9.69931304e-01 -1.09064507e+00 -3.13951448e-02 5.86750209e-01
6.77955508e-01 6.77019581e-02 -5.21484911e-01 -1.19443968e-01
-1.10226311e-01 -2.09327340e-02 -5.31056583e-01 -2.74041742e-02
-1.17886889e+00 -5.14698982e-01 -8.72407258e-02 3.40949625e-01
6.22866750e-01 -6.36934221e-01 -6.52797699e-01 1.88183501e-01
-3.06053370e-01 -1.41877210e+00 -8.38223696e-01 -3.21768343e-01
-1.01553881e+00 -1.13842142e+00 -5.89518011e-01 -5.39826751e-01
8.40792418e-01 5.59804618e-01 8.71827722e-01 2.91984886e-01
-7.24231675e-02 5.17920375e-01 2.90514342e-02 4.22761321e-01
-2.01863229e-01 -3.77447695e-01 4.69077229e-02 2.27901027e-01
-3.50445479e-01 -9.89316583e-01 -9.41629469e-01 6.21535838e-01
-9.11702335e-01 4.61149067e-01 8.93289596e-02 3.72141808e-01
1.19271386e+00 -3.08205485e-01 -8.75506252e-02 -4.62352395e-01
-1.63907737e-01 -9.46218669e-02 -9.15986300e-01 8.19880366e-02
-1.77596569e-01 -1.37755364e-01 2.72272587e-01 -7.20617473e-01
-1.19635463e+00 5.43315411e-01 6.19194470e-02 -1.22849894e+00
8.85943323e-02 2.32313499e-01 -5.16078115e-01 -4.72824782e-01
2.91910022e-01 4.10175025e-01 -1.97125360e-01 -5.86327255e-01
4.20302391e-01 1.58521309e-02 9.04458165e-01 -8.00983548e-01
1.14413261e+00 1.02245855e+00 1.55846104e-01 -6.66009068e-01
-6.84084356e-01 -1.98325127e-01 -9.80582952e-01 -3.57962728e-01
9.52478230e-01 -1.21142805e+00 -8.95009756e-01 8.05943608e-01
-1.43542254e+00 -5.11924148e-01 -1.64410576e-01 5.29233277e-01
-6.40611231e-01 6.19095922e-01 -6.06727540e-01 -5.03377795e-01
-2.72537917e-01 -1.24381053e+00 1.39263678e+00 1.16963260e-01
3.89140576e-01 -7.97210455e-01 2.28053421e-01 1.41231433e-01
6.07688464e-02 9.67578053e-01 3.13221514e-01 5.86049914e-01
-1.40847898e+00 1.30278572e-01 -1.02301486e-01 1.19341999e-01
2.91097492e-01 2.37321466e-01 -7.52853990e-01 -4.15151626e-01
1.23144336e-01 8.85676891e-02 4.17358309e-01 4.36632037e-01
1.40705204e+00 -3.91301095e-01 -1.40980095e-01 1.64911842e+00
1.60022211e+00 2.70933568e-01 6.50568783e-01 3.11791450e-01
1.47152472e+00 4.21117544e-01 3.87257040e-01 5.34376502e-01
7.01945603e-01 9.85058248e-01 6.66591644e-01 4.36157286e-02
-8.88160393e-02 -3.05662841e-01 3.83367479e-01 1.17693460e+00
-4.11664724e-01 1.66731514e-02 -5.47645807e-01 2.23760098e-01
-1.82055318e+00 -5.82264662e-01 -6.96116015e-02 2.19488120e+00
8.00180316e-01 -3.68890256e-01 -2.63659090e-01 -3.36717099e-01
6.89907908e-01 3.55211258e-01 -1.05938184e+00 1.49264053e-01
-3.62946898e-01 -1.51999861e-01 4.58957911e-01 6.99143529e-01
-6.86759293e-01 9.15757179e-01 6.08942652e+00 2.70800769e-01
-1.39045131e+00 1.82636473e-02 3.33571374e-01 -3.26848596e-01
-7.02664614e-01 1.06932014e-01 -4.90049183e-01 3.26218694e-01
4.12929177e-01 -1.31500155e-01 7.88212538e-01 5.71860671e-01
4.31590825e-01 4.48726639e-02 -8.77824724e-01 1.25330722e+00
-4.68191691e-02 -1.59716344e+00 6.40299246e-02 2.25399971e-01
1.08039391e+00 1.03513174e-01 -1.41771525e-01 -3.05187196e-01
2.52506346e-01 -7.17130840e-01 8.43810499e-01 1.01875901e+00
9.76096869e-01 -6.33287013e-01 1.03442088e-01 3.04211378e-01
-1.43284655e+00 1.28167167e-01 -4.10269439e-01 1.11563325e-01
5.10712326e-01 4.89422143e-01 8.37187320e-02 8.59222591e-01
8.02348197e-01 1.15713429e+00 -1.63528964e-01 6.37612462e-01
-8.32551047e-02 -3.39681730e-02 -4.40860689e-01 6.89886928e-01
-1.80594310e-01 -6.85919702e-01 7.35867381e-01 5.59893012e-01
4.53917891e-01 5.75850725e-01 3.67157191e-01 1.04467106e+00
-2.17967361e-01 -2.13928133e-01 -4.81051862e-01 3.12021911e-01
5.43438911e-01 1.19476140e+00 -3.89823258e-01 1.90505795e-02
-2.86309034e-01 1.30521357e+00 2.26680100e-01 4.98914719e-01
-7.42369890e-01 3.03193629e-01 7.93209493e-01 1.25873148e-01
1.50436372e-01 -7.68258870e-01 -2.51934249e-02 -1.75710297e+00
3.68942946e-01 -6.62917733e-01 6.62290528e-02 -1.07235229e+00
-7.83861101e-01 5.04416347e-01 -6.38937354e-02 -1.36528051e+00
-1.41770348e-01 -3.15485895e-01 -3.98360670e-01 6.81683362e-01
-1.71845567e+00 -1.30602241e+00 -7.97742903e-01 9.10473824e-01
4.51341778e-01 2.39592224e-01 3.94658506e-01 2.86545157e-01
-5.53051591e-01 1.79797515e-01 -1.73748717e-01 -3.62363458e-02
4.64206457e-01 -9.02738810e-01 3.79222542e-01 9.67785299e-01
-1.69954762e-01 6.53535068e-01 3.01207453e-01 -6.77463293e-01
-2.08199763e+00 -1.16454852e+00 -1.81939527e-02 -3.50184053e-01
2.29508951e-01 -1.88824981e-01 -1.11313844e+00 7.75978386e-01
-2.32100546e-01 5.15152097e-01 2.28878409e-02 -6.30856693e-01
-1.81148320e-01 -1.45392075e-01 -1.06995881e+00 5.36611259e-01
1.70291996e+00 -7.35070646e-01 -2.29208529e-01 3.09875727e-01
1.23449564e+00 -1.36578965e+00 -1.07785046e+00 4.38630491e-01
6.91147029e-01 -9.72997725e-01 1.26435256e+00 -1.08962514e-01
3.04728866e-01 -6.95359349e-01 -4.93101984e-01 -8.57765913e-01
-6.56789243e-02 -1.32149255e+00 -5.82260072e-01 9.12484109e-01
-3.91927928e-01 -5.56902885e-01 9.72774506e-01 7.75387466e-01
-4.06981766e-01 -8.15800428e-01 -1.08636248e+00 -5.65265179e-01
-2.16697514e-01 -3.23935330e-01 8.73962760e-01 1.10092664e+00
-1.04333127e+00 -3.36193889e-01 -6.09441519e-01 5.69872081e-01
6.93976402e-01 2.99879879e-01 1.08291769e+00 -1.09346592e+00
-3.71550918e-01 1.62671328e-01 -3.63580227e-01 -1.72439075e+00
2.84784168e-01 -4.95541066e-01 -3.58404446e-04 -1.26249039e+00
-1.24644153e-01 -3.65988255e-01 1.25360787e-01 1.93587080e-01
2.62220390e-02 3.10095489e-01 -5.20526282e-02 5.04627168e-01
-1.73967749e-01 8.22875261e-01 1.97998846e+00 2.54334927e-01
-3.93755049e-01 -4.39573526e-01 -2.35903651e-01 9.95222926e-01
1.67377889e-01 -2.36242563e-01 -3.98720562e-01 -9.44855094e-01
3.95017356e-01 4.59412932e-01 5.11348128e-01 -7.36194909e-01
2.11870730e-01 -6.96653366e-01 4.10017341e-01 -9.27551866e-01
5.93653619e-01 -9.36127245e-01 6.96211755e-01 2.18825445e-01
2.90744994e-02 3.44626397e-01 -5.58389947e-02 7.70284593e-01
3.98323080e-03 3.65870714e-01 7.60186076e-01 -1.86000839e-01
-6.76892042e-01 1.07273078e+00 2.65194267e-01 1.00995402e-03
7.79267669e-01 -4.51388299e-01 -5.89819774e-02 -3.40952903e-01
-6.73877120e-01 3.21764387e-02 1.08099091e+00 4.87888545e-01
9.48272467e-01 -1.53006446e+00 -3.73577893e-01 3.74739468e-01
-1.35430411e-01 9.59390342e-01 4.31599200e-01 6.68698013e-01
-8.91331851e-01 -1.02176197e-01 1.82774868e-02 -8.23183060e-01
-8.93844783e-01 1.97102532e-01 9.01911855e-01 2.32506678e-01
-1.01232529e+00 6.26998782e-01 3.63529801e-01 -7.79522598e-01
-3.32455300e-02 -4.11235362e-01 2.46578723e-01 -5.49910188e-01
3.16160053e-01 4.19517696e-01 -4.92498279e-02 -8.61090243e-01
-2.62163967e-01 1.19205761e+00 3.08249116e-01 -1.19785115e-01
1.54838419e+00 -5.75745583e-01 -3.05007726e-01 2.97837645e-01
1.36940086e+00 8.87908041e-02 -1.96947086e+00 -3.29537362e-01
-6.47887588e-01 -8.43434036e-01 2.07334459e-01 -1.42770559e-01
-1.51080918e+00 4.77901101e-01 3.36586386e-01 -4.74820107e-01
1.17413604e+00 -3.18095237e-01 1.01781595e+00 5.04617333e-01
5.46475232e-01 -8.84999454e-01 1.47831142e-01 5.53186357e-01
1.07842338e+00 -8.08509529e-01 2.81206220e-01 -6.86266899e-01
-5.89755457e-03 1.35090399e+00 8.18181336e-01 -3.86107087e-01
5.75580537e-01 1.55140832e-01 -5.12788780e-02 -2.04312131e-01
-4.52336550e-01 3.21523577e-01 5.26957572e-01 5.09081960e-01
2.59053446e-02 -2.68439949e-01 1.66906431e-01 -6.94661364e-02
-1.15049578e-01 6.03869325e-03 5.96784055e-01 8.34287465e-01
-1.03841811e-01 -1.01533151e+00 -4.12592500e-01 -1.88016444e-01
-4.78002504e-02 3.30610514e-01 -1.50730059e-01 6.72462344e-01
1.08309053e-01 5.21436512e-01 1.10647582e-01 -2.96855330e-01
4.67039019e-01 -4.47363108e-01 7.32294858e-01 -5.33275306e-01
-1.40589952e-01 3.68664145e-01 -1.78177118e-01 -1.20030844e+00
-7.94011831e-01 -6.89888060e-01 -1.32044327e+00 -3.08660656e-01
-2.37929165e-01 -5.02936125e-01 5.83676517e-01 8.64145935e-01
5.60180545e-01 3.72774303e-01 9.55443263e-01 -1.33672917e+00
-9.11247060e-02 -3.48312348e-01 -3.17104489e-01 3.46334457e-01
6.20668411e-01 -7.80091405e-01 -3.44784379e-01 3.37601393e-01] | [8.965178489685059, -2.768359422683716] |
ff7f3eca-84b5-4fb4-bee8-9f62b4d7c7ad | early-heart-disease-prediction-using-hybrid | 2208.08882 | null | https://arxiv.org/abs/2208.08882v2 | https://arxiv.org/pdf/2208.08882v2.pdf | Early heart disease prediction using hybrid quantum classification | The rate of heart morbidity and heart mortality increases significantly which affect the global public health and world economy. Early prediction of heart disease is crucial for reducing heart morbidity and mortality. This paper proposes two quantum machine learning methods i.e. hybrid quantum neural network and hybrid random forest quantum neural network for early detection of heart disease. The methods are applied on the Cleveland and Statlog datasets. The results show that hybrid quantum neural network and hybrid random forest quantum neural network are suitable for high dimensional and low dimensional problems respectively. The hybrid quantum neural network is sensitive to outlier data while hybrid random forest is robust on outlier data. A comparison between different machine learning methods shows that the proposed quantum methods are more appropriate for early heart disease prediction where 96.43% and 97.78% area under curve are obtained for Cleveland and Statlog dataset respectively. | ['Gerhard Hellstern', 'Hanif Heidari'] | 2022-08-17 | null | null | null | null | ['disease-prediction'] | ['medical'] | [-2.76231080e-01 8.85376483e-02 -2.87809223e-01 -1.98543981e-01
-6.95963979e-01 1.64458528e-01 2.13493668e-02 5.88902771e-01
-2.63884366e-01 9.84155476e-01 -1.77947491e-01 -2.16097906e-01
-2.24721268e-01 -1.30717123e+00 5.46582639e-02 -6.36821628e-01
-1.21482305e-01 8.41344476e-01 1.77103803e-01 9.93549228e-02
2.88571894e-01 3.10076207e-01 -1.07308161e+00 7.69863650e-03
8.84671271e-01 7.42215872e-01 -7.81372428e-01 1.32057810e+00
5.02806187e-01 5.80375016e-01 -5.68885244e-02 -1.24993235e-01
4.65073079e-01 -8.59689593e-01 -5.58449328e-01 -8.77907515e-01
1.85481861e-01 -2.69540548e-01 -4.68712747e-01 9.66378689e-01
1.38294566e+00 -3.76440585e-02 1.23753905e+00 -1.09539545e+00
-4.86366808e-01 2.63630092e-01 6.66329563e-02 5.58043599e-01
2.26921421e-02 -7.07000792e-02 9.36871409e-01 -4.54202980e-01
2.57504791e-01 7.83954322e-01 1.10887945e+00 5.14278889e-01
-1.25710917e+00 -5.90676546e-01 -1.55465007e+00 5.93947232e-01
-1.55030417e+00 -5.64885698e-02 4.86695707e-01 -5.09341717e-01
1.10675752e+00 2.42026970e-01 7.43472397e-01 3.80318224e-01
7.62218118e-01 -7.71419257e-02 1.21404386e+00 -6.87355697e-01
2.57691741e-01 -1.74328033e-02 7.53870904e-01 1.23783731e+00
4.48038459e-01 1.09512508e+00 -4.60932970e-01 -7.47114420e-01
8.12791526e-01 3.92745018e-01 2.84027815e-01 -5.91799095e-02
-1.28191686e+00 9.99769747e-01 4.40831602e-01 -2.00879909e-02
-4.39296454e-01 1.28643870e-01 6.11246645e-01 3.06606680e-01
4.59819473e-02 -2.92305090e-02 -5.60189128e-01 -1.06995881e-01
-7.52437294e-01 2.89510399e-01 5.22714615e-01 4.98280287e-01
6.85412824e-01 -2.70899147e-01 -1.46035150e-01 5.69375575e-01
4.25551653e-01 1.06368315e+00 1.75977409e-01 -1.07119620e+00
-1.07021108e-01 3.98255289e-01 -1.81776613e-01 -8.52486193e-01
-1.06078625e+00 -3.32100928e-01 -1.56575561e+00 8.85464996e-02
4.92231041e-01 -3.77947241e-01 -8.53907168e-01 1.36372447e+00
2.16677964e-01 6.67082310e-01 5.47769427e-01 4.96193171e-01
1.10708237e+00 7.26924598e-01 1.81707069e-01 -3.66747081e-01
1.37356114e+00 -3.17502052e-01 -5.58581531e-01 4.23106879e-01
9.74850416e-01 -7.03368723e-01 2.97014415e-01 2.21088365e-01
-5.13521135e-01 -4.01814699e-01 -8.42253625e-01 1.12029620e-01
-1.92436948e-01 -1.23819187e-02 4.61884499e-01 1.07611012e+00
-4.35496598e-01 1.11728537e+00 -9.09464240e-01 -5.15072703e-01
2.78068691e-01 5.24107277e-01 2.64228191e-02 2.70331144e-01
-1.70835471e+00 8.82296860e-01 4.38061982e-01 -2.37346202e-01
-1.60002485e-01 -1.34438470e-01 -2.78338671e-01 -8.24093819e-02
-5.45037210e-01 -1.35651898e+00 9.19555306e-01 -1.04818404e-01
-1.41373217e+00 8.15876007e-01 3.52433652e-01 -7.85170078e-01
2.34986186e-01 9.07285810e-02 -4.83432829e-01 1.53996974e-01
1.24189198e-01 4.67969179e-01 4.81865019e-01 1.82678580e-01
-5.60225785e-01 -6.88304245e-01 -9.77644265e-01 7.57082552e-02
3.25555444e-01 -1.20073408e-01 6.80881560e-01 -2.70778656e-01
5.42504728e-01 -1.37245560e+00 -3.07264745e-01 -3.34721059e-01
-5.29147387e-01 -5.60919285e-01 5.44896960e-01 -4.17739511e-01
1.34886384e+00 -1.65101266e+00 -2.45513946e-01 4.30329561e-01
4.00818229e-01 5.05900942e-02 6.35218382e-01 3.45318556e-01
-2.17287228e-01 1.65396050e-01 1.19942455e-02 8.08171213e-01
-1.49296165e-01 3.41137171e-01 3.12616169e-01 6.48691297e-01
-1.68696404e-01 7.62152195e-01 -5.92737913e-01 -8.26397657e-01
1.69658065e-01 3.76440376e-01 -6.16228759e-01 -1.95313811e-01
6.61602318e-01 3.64404380e-01 -4.87599820e-01 6.38212323e-01
4.36087668e-01 -3.01252037e-01 -3.16579401e-01 -1.82448462e-01
1.50611758e-01 -9.00695473e-02 -1.16224444e+00 9.30924177e-01
3.29594254e-01 3.24604362e-01 -1.10768306e+00 -8.27537000e-01
9.74597573e-01 6.74058080e-01 3.41477633e-01 -7.13431716e-01
8.30872208e-02 3.55452359e-01 3.58574957e-01 -7.02064395e-01
-3.11755985e-01 -9.57120895e-01 -6.51039407e-02 3.86447310e-01
-4.89247963e-02 2.97400147e-01 -1.27518386e-01 1.72222227e-01
1.09460425e+00 -2.54811794e-01 7.33971477e-01 -2.74538726e-01
2.81710088e-01 -3.42039275e-03 1.10914207e+00 1.11410427e+00
-1.17791116e+00 6.31154537e-01 6.19930625e-01 -1.04431248e+00
-1.21481419e+00 -1.48577023e+00 -7.56971478e-01 4.04535472e-01
-2.09715776e-02 -4.16318387e-01 -3.94027263e-01 -5.50614953e-01
-1.25628293e-01 2.40180284e-01 -3.26849431e-01 -4.91253793e-01
-3.72408509e-01 -1.42591727e+00 7.15981960e-01 2.20400468e-01
5.89983106e-01 -8.56362820e-01 -6.90372646e-01 1.41243905e-01
-6.22702651e-02 -4.00041580e-01 3.03188324e-01 2.24597096e-01
-1.47460198e+00 -1.27356815e+00 -7.08517313e-01 -4.72226739e-01
6.68147504e-02 -5.26184678e-01 1.20399988e+00 -2.44809166e-02
-6.77061737e-01 -2.82973439e-01 -1.97735861e-01 -3.93565953e-01
-8.16908419e-01 7.41891190e-02 4.39261109e-01 -5.04138172e-01
4.60065842e-01 -4.17101771e-01 -9.84434664e-01 -1.34087846e-01
-3.75783332e-02 -3.04859132e-02 8.05922031e-01 1.18232465e+00
4.74663973e-01 6.04509041e-02 8.66570592e-01 -1.07147479e+00
2.58223325e-01 -3.36738110e-01 -3.18806410e-01 2.17665639e-02
-1.27390158e+00 2.96917766e-01 2.38880634e-01 9.10100862e-02
-2.67229021e-01 3.41035396e-01 -1.98546410e-01 7.66803622e-02
-1.63576573e-01 1.82353899e-01 7.64359355e-01 5.70284314e-02
1.40496004e+00 -2.71919817e-01 -1.06117725e-01 -1.99618071e-01
-1.81212664e-01 8.76855016e-01 -5.33009470e-02 3.29956063e-03
4.62268472e-01 3.03164031e-02 9.10246968e-01 -1.08217132e+00
-6.49120629e-01 -4.55314785e-01 -7.79929996e-01 -3.20623219e-02
1.47833359e+00 -7.18547165e-01 -6.93406701e-01 4.89091337e-01
-8.29410732e-01 2.33718514e-01 -8.67605060e-02 1.00997651e+00
-5.30622184e-01 3.74201000e-01 -1.06781733e+00 -9.00549769e-01
-9.52235341e-01 -8.10081661e-01 6.37623847e-01 4.52186435e-01
-9.91800204e-02 -9.58914816e-01 5.83637595e-01 9.90401357e-02
1.30415127e-01 5.44520259e-01 1.31219769e+00 -6.29333973e-01
-4.98442292e-01 -6.46833837e-01 -4.52866226e-01 1.47379473e-01
-1.82143778e-01 1.54455736e-01 -8.10077667e-01 1.87643364e-01
-4.91255343e-01 8.33514892e-03 6.18973434e-01 7.88985312e-01
-1.59589443e-02 -1.36213554e-02 -2.06929728e-01 2.98382759e-01
1.41092730e+00 1.02701172e-01 6.62497044e-01 1.73141778e-01
3.93307090e-01 -2.11316884e-01 2.64981121e-01 4.91878361e-01
2.62222528e-01 3.10029447e-01 1.10340752e-02 4.47679423e-02
7.81698897e-02 2.04776675e-01 3.22890840e-02 9.16716158e-01
-2.49395460e-01 6.38880491e-01 -1.58846796e+00 -2.59504421e-04
-1.79297924e+00 -1.42129099e+00 -9.91658509e-01 2.41291666e+00
4.46542025e-01 3.88719201e-01 4.11582768e-01 3.18593025e-01
7.87546039e-01 -5.56614518e-01 -3.79699349e-01 -6.53267741e-01
1.01485193e-01 5.72297513e-01 6.49236023e-01 2.41853163e-01
-1.21116078e+00 3.97767216e-01 6.33658218e+00 3.08366895e-01
-7.67624557e-01 2.75842130e-01 8.08236420e-01 3.88643354e-01
7.75235295e-01 4.13088441e-01 -4.48499769e-01 1.89669043e-01
1.61256468e+00 -3.45617563e-01 1.96513720e-02 4.85308498e-01
2.66663939e-01 -1.50322944e-01 -7.28222966e-01 1.38373888e+00
-6.86100900e-01 -1.26387644e+00 -1.57658920e-01 -7.01225400e-02
5.24341822e-01 6.76478207e-01 -3.55876088e-01 4.13437337e-01
-2.83709675e-01 -9.11645770e-01 -1.55716717e-01 1.13275051e+00
6.90209270e-01 -9.37386274e-01 1.26311731e+00 5.49314678e-01
-7.05348790e-01 -7.83636719e-02 -3.69902760e-01 -4.20814782e-01
5.71809150e-02 8.51613641e-01 -7.91820347e-01 2.85597295e-01
8.17952812e-01 6.11044407e-01 -6.66517675e-01 1.15733635e+00
2.73345053e-01 8.06394279e-01 -8.50023746e-01 -2.21082643e-01
-2.32362479e-01 -2.98523188e-01 3.70478898e-01 6.00675464e-01
3.37410659e-01 9.72481146e-02 9.39619690e-02 6.35199606e-01
3.40260357e-01 6.46021247e-01 -8.04541290e-01 1.77423686e-01
2.79680770e-02 8.80842328e-01 -5.50273478e-01 -6.56580269e-01
-1.61574051e-01 7.07198560e-01 -2.07469761e-01 -1.34364441e-01
-8.48374844e-01 -7.06871927e-01 -2.15635806e-01 3.91737884e-03
-5.50489545e-01 2.56223857e-01 -7.36277521e-01 -1.27325583e+00
-7.67908812e-01 -2.50813007e-01 7.33488619e-01 -4.86966461e-01
-1.19769990e+00 4.40290093e-01 -3.08407754e-01 -1.33014441e+00
-3.24681014e-01 -5.01198888e-01 -6.24873102e-01 7.54599869e-01
-9.04295087e-01 -7.16277719e-01 1.27437204e-01 2.08956271e-01
-2.19585389e-01 -2.73957521e-01 1.46036565e+00 4.43472266e-01
-9.28451419e-01 1.72957867e-01 4.75735247e-01 2.68321991e-01
7.85364032e-01 -1.40752304e+00 -9.11302716e-02 5.45381606e-01
-5.76404035e-02 3.81859511e-01 7.48643994e-01 -8.39026868e-01
-8.82824838e-01 -7.45324552e-01 1.26321208e+00 -4.03965354e-01
1.79792434e-01 3.29269677e-01 -3.87601227e-01 1.92978039e-01
-2.14094728e-01 4.34398293e-01 1.11309314e+00 2.96831697e-01
-1.18073449e-02 -1.49828732e-01 -1.42464423e+00 1.51149228e-01
1.69774443e-01 -7.35014141e-01 -4.96009022e-01 8.45706642e-01
-1.48765042e-01 -5.68058603e-02 -1.41432631e+00 7.86338091e-01
1.08824229e+00 -1.56397653e+00 9.36887622e-01 -5.81449866e-01
2.66500674e-02 -9.57606286e-02 -2.32274234e-02 -5.13589025e-01
-4.23179358e-01 -5.19562840e-01 -6.84532374e-02 4.81171668e-01
6.74342453e-01 -8.42678249e-01 1.13185763e+00 7.18421936e-01
4.82083827e-01 -6.68822587e-01 -1.41083777e+00 -3.35454047e-01
3.16487372e-01 -2.61868745e-01 -2.41394401e-01 9.05397296e-01
-1.15181647e-01 9.63871181e-01 -5.17997980e-01 1.35624766e-01
1.13784623e+00 8.98565054e-02 5.30866504e-01 -1.84223354e+00
-1.46140352e-01 3.06992297e-05 -1.11196303e+00 2.28641834e-02
-6.27200961e-01 -1.02381229e+00 -7.02557266e-01 -1.22645319e+00
7.13052452e-01 -5.72725594e-01 -8.18387270e-01 1.01639949e-01
-2.41614074e-01 4.90221381e-01 -1.54232100e-01 3.95278841e-01
-2.01136455e-01 1.69542700e-01 7.84294844e-01 3.26041520e-01
-3.73567671e-01 6.14816010e-01 9.71599966e-02 7.77599454e-01
1.22457397e+00 -1.04333591e+00 1.97292566e-02 3.40721637e-01
4.87901688e-01 7.12559283e-01 5.51408947e-01 -1.53268683e+00
-5.46525791e-02 1.17871128e-01 5.40795803e-01 -7.22706020e-01
-1.33415669e-01 -3.88863564e-01 1.51239440e-01 1.41893446e+00
1.58662349e-02 1.31652966e-01 -5.99099338e-01 8.07948709e-01
1.04015931e-01 -1.85902297e-01 1.34536517e+00 -5.74358040e-03
-3.64999361e-02 3.12677503e-01 -2.48705998e-01 -3.19170244e-02
1.05179322e+00 -1.53917059e-01 -2.43600771e-01 -1.51605874e-01
-1.37618077e+00 -2.31244639e-01 1.52724147e-01 -1.07619941e-01
4.36949104e-01 -1.49787962e+00 -7.68336892e-01 3.69266391e-01
1.00504555e-01 -4.80684698e-01 4.01045740e-01 1.30856657e+00
-1.25639880e+00 5.46409428e-01 -4.22771275e-01 -7.05178857e-01
-1.23638117e+00 2.57518739e-01 7.81163633e-01 -2.02441350e-01
-7.06247151e-01 4.91603494e-01 -6.31659746e-01 -4.92566973e-01
-1.94235802e-01 -3.57634008e-01 2.35723257e-02 -3.65445316e-01
-3.32283345e-03 1.08418679e+00 -2.72048235e-01 -8.43614459e-01
-4.20530379e-01 8.11275482e-01 2.18329564e-01 1.99417427e-01
1.01361334e+00 7.92505443e-02 -3.26148659e-01 7.71132529e-01
1.17005754e+00 -3.96375567e-01 -1.81598783e-01 4.52341847e-02
3.96822780e-01 2.80095994e-01 4.12345171e-01 -8.22069824e-01
-3.47478628e-01 1.29195225e+00 1.88656211e+00 5.90454698e-01
9.94353056e-01 -2.39935100e-01 1.07853389e+00 7.22907484e-01
-1.62536781e-02 -1.20754898e+00 -5.61020374e-01 3.23655993e-01
-1.53451711e-01 -1.53127360e+00 3.06669563e-01 -2.24417284e-01
-2.74621665e-01 1.62239325e+00 1.54988626e-02 -5.29958606e-01
1.26861954e+00 -3.94242346e-01 2.51320958e-01 -2.45461330e-01
-8.07503998e-01 -1.75958946e-01 1.04912266e-01 2.39363626e-01
5.89110255e-01 4.75862652e-01 -6.18154228e-01 1.79773331e-01
5.59396744e-02 5.34994483e-01 4.81861115e-01 1.01349342e+00
-8.18393886e-01 -1.16990602e+00 -3.66379976e-01 1.17609882e+00
-7.95747280e-01 -2.53253877e-01 -2.18303546e-01 3.29441994e-01
4.15994316e-01 4.53262508e-01 -3.12133372e-01 -3.13842922e-01
4.18070331e-02 8.58822763e-01 4.04842734e-01 -4.38209623e-01
-2.06954688e-01 7.78202936e-02 4.27946411e-02 -1.75777510e-01
-4.24807310e-01 -7.02460706e-01 -1.79367578e+00 -2.76187986e-01
-7.14991271e-01 2.24763513e-01 6.12506986e-01 7.09215581e-01
4.83049572e-01 2.28114128e-01 3.55280071e-01 -1.27391368e-01
-7.74480999e-01 -9.53084111e-01 -9.79157686e-01 2.09499136e-01
5.49424291e-01 -6.89024091e-01 -1.92322686e-01 -7.16676563e-02] | [14.11827564239502, 3.3437511920928955] |
62d8146b-fd27-4e11-9703-10479c89a96c | on-the-audio-visual-synchronization-for-lip | 2303.00502 | null | https://arxiv.org/abs/2303.00502v1 | https://arxiv.org/pdf/2303.00502v1.pdf | On the Audio-visual Synchronization for Lip-to-Speech Synthesis | Most lip-to-speech (LTS) synthesis models are trained and evaluated under the assumption that the audio-video pairs in the dataset are perfectly synchronized. In this work, we show that the commonly used audio-visual datasets, such as GRID, TCD-TIMIT, and Lip2Wav, can have data asynchrony issues. Training lip-to-speech with such datasets may further cause the model asynchrony issue -- that is, the generated speech and the input video are out of sync. To address these asynchrony issues, we propose a synchronized lip-to-speech (SLTS) model with an automatic synchronization mechanism (ASM) to correct data asynchrony and penalize model asynchrony. We further demonstrate the limitation of the commonly adopted evaluation metrics for LTS with asynchronous test data and introduce an audio alignment frontend before the metrics sensitive to time alignment for better evaluation. We compare our method with state-of-the-art approaches on conventional and time-aligned metrics to show the benefits of synchronization training. | ['Brian Mak', 'Zhe Niu'] | 2023-03-01 | null | null | null | null | ['audio-visual-synchronization', 'audio-visual-synchronization', 'lip-to-speech-synthesis', 'speech-synthesis'] | ['audio', 'computer-vision', 'computer-vision', 'speech'] | [-1.29837602e-01 -2.03190446e-01 -2.34158561e-01 -2.34632179e-01
-1.05514252e+00 -5.07160485e-01 7.34886646e-01 1.62418857e-02
-2.27855109e-02 3.83968145e-01 1.93665355e-01 -1.61280751e-01
2.46941775e-01 7.35446736e-02 -7.19545662e-01 -5.88453531e-01
-1.39624089e-01 3.75717103e-01 5.18641353e-01 2.47371614e-01
2.42751818e-02 3.18581700e-01 -2.03526521e+00 4.74517614e-01
5.19717276e-01 7.88879275e-01 1.38727248e-01 1.08976078e+00
-1.41912356e-01 1.00623667e+00 -1.12212467e+00 2.85745654e-02
9.06954985e-03 -4.93221521e-01 -6.38979256e-01 -1.23413138e-01
6.14445984e-01 -2.88407654e-01 -2.77965307e-01 6.19018555e-01
9.92724478e-01 -3.81030552e-02 2.75880992e-01 -2.15392017e+00
2.75261939e-01 4.45866138e-01 -2.10055023e-01 1.47788942e-01
5.46126962e-01 3.22751135e-01 6.67020679e-01 -7.96663880e-01
8.35668683e-01 1.36831355e+00 6.58013761e-01 6.21277511e-01
-1.14730227e+00 -9.22538161e-01 -7.12652057e-02 5.96426368e-01
-1.46359146e+00 -1.29934680e+00 6.93593800e-01 -4.53345597e-01
9.53601956e-01 3.57683808e-01 4.47933704e-01 1.38586092e+00
-2.58022070e-01 8.37764025e-01 8.60735178e-01 -4.25922573e-01
1.85532808e-01 -2.61674542e-02 -3.80990267e-01 3.97869088e-02
-4.71593469e-01 2.63691634e-01 -1.22394371e+00 -3.38335298e-02
4.10892516e-01 -7.64374316e-01 -1.35647640e-01 -1.35452107e-01
-1.49295473e+00 2.82394201e-01 -4.41097826e-01 2.44244426e-01
3.96333709e-02 2.81922787e-01 1.01097226e+00 5.26844680e-01
4.69028085e-01 -2.18252242e-02 -3.61352980e-01 -7.03325212e-01
-1.45672667e+00 1.69576734e-01 6.74648881e-01 1.19062066e+00
2.67045736e-01 4.17775065e-01 -2.53059119e-01 7.20577717e-01
6.01255357e-01 4.12601680e-01 7.77082980e-01 -1.05835557e+00
6.61309123e-01 -1.08678184e-01 -6.21089488e-02 -4.79690909e-01
-1.57334045e-01 2.57236362e-01 -4.70961004e-01 3.72029781e-01
4.67513323e-01 8.37238133e-02 -6.91438675e-01 1.85239124e+00
4.39634621e-01 8.52179646e-01 2.81742483e-01 6.90302372e-01
9.99586046e-01 6.27364397e-01 -2.15858705e-02 -6.44672215e-01
1.04820859e+00 -9.60534930e-01 -1.24879074e+00 2.09372774e-01
6.47723317e-01 -1.38925946e+00 1.03401935e+00 2.35664144e-01
-1.35487270e+00 -8.83759737e-01 -1.05298972e+00 6.97225183e-02
-3.64588350e-02 1.33112505e-01 -2.76674889e-02 4.81053650e-01
-1.18467176e+00 5.71250975e-01 -1.03268492e+00 -4.21271652e-01
-1.62983403e-01 3.05288076e-01 -3.56647700e-01 4.92551476e-01
-1.09509718e+00 5.63722432e-01 5.89569239e-03 -2.40845636e-01
-1.21584487e+00 -8.05078506e-01 -7.31674254e-01 -3.06183666e-01
2.34948888e-01 -2.42194980e-01 1.79004705e+00 -1.13814723e+00
-2.12131858e+00 6.76443875e-01 -4.30667669e-01 -3.90540749e-01
8.58644605e-01 -5.85595407e-02 -7.80528843e-01 2.37537801e-01
2.49269847e-02 8.19390714e-01 1.20220339e+00 -1.12923825e+00
-5.48269153e-01 2.83630192e-01 -4.21470612e-01 1.39120504e-01
6.66468143e-02 5.37053585e-01 -6.61622703e-01 -6.41241789e-01
-2.85817474e-01 -9.99831140e-01 4.57783490e-01 -5.36886007e-02
-4.83187854e-01 -2.07113206e-01 1.21048737e+00 -5.18617570e-01
1.35952234e+00 -2.44280863e+00 -5.08343466e-02 -1.39398351e-01
-1.46182597e-01 3.59827071e-01 -2.99718708e-01 6.93227947e-01
-3.26943874e-01 1.26863584e-01 4.62049305e-01 -9.96406198e-01
-7.31880218e-02 2.17614368e-01 -4.21154410e-01 6.06792808e-01
-1.26483098e-01 3.71934742e-01 -6.89616203e-01 -9.45375562e-01
3.43699485e-01 4.32763636e-01 -2.97665089e-01 7.15863883e-01
-2.76670545e-01 7.04308391e-01 2.48507008e-01 4.81645584e-01
4.92287815e-01 2.50762016e-01 9.40863695e-03 -2.87355512e-01
-4.98869687e-01 6.48095906e-01 -1.48304963e+00 1.66157043e+00
-6.53412163e-01 1.32447743e+00 1.81237221e-01 -4.36786115e-01
4.94417250e-01 1.22106600e+00 7.46872187e-01 -7.29160666e-01
-7.10524768e-02 2.65764982e-01 5.59646748e-02 -6.76690221e-01
4.20497507e-01 4.09071207e-01 3.55071515e-01 4.92373466e-01
1.87197849e-01 -3.80868435e-01 1.77699372e-01 1.45133540e-01
6.52396917e-01 1.61274672e-01 2.53963564e-02 1.89460907e-02
4.13138807e-01 -3.91285866e-01 4.46297377e-01 3.36293995e-01
-4.33978945e-01 9.52958882e-01 5.68679810e-01 1.17086038e-01
-1.31509078e+00 -9.68011320e-01 1.76861614e-01 9.68350410e-01
-7.18318149e-02 -1.15741563e+00 -8.91911983e-01 -3.21519673e-01
-3.31742734e-01 4.85993087e-01 -1.90107852e-01 1.90189958e-01
-4.24058944e-01 1.73213169e-01 1.14783287e+00 4.05063391e-01
7.83901215e-02 -8.81725907e-01 -5.71349025e-01 3.26107323e-01
-4.09174293e-01 -1.53577185e+00 -8.11421275e-01 -1.50815919e-01
-5.66390097e-01 -1.03538895e+00 -5.91153741e-01 -4.64877963e-01
1.91425577e-01 2.48792946e-01 9.55538511e-01 -6.17727377e-02
-7.67050534e-02 4.20556128e-01 -2.34299690e-01 -5.14891922e-01
-1.11340714e+00 -2.56935328e-01 4.29270267e-01 1.47096276e-01
-2.97497272e-01 -5.13189316e-01 -3.92365277e-01 8.23915720e-01
-9.29081082e-01 3.18375826e-01 -1.46012962e-01 4.93635833e-01
5.11443138e-01 -1.46667272e-01 5.74562669e-01 -1.09209955e-01
4.66461182e-01 -2.05555275e-01 -7.85858750e-01 9.22627151e-02
-4.38106775e-01 -4.56040502e-02 6.05324328e-01 -1.09300649e+00
-7.03705490e-01 3.26807164e-02 -1.17719747e-01 -1.00755227e+00
-1.88156202e-01 1.92775056e-01 -2.93371588e-01 1.23991497e-01
4.12060827e-01 2.84450483e-02 2.40326285e-01 -4.07214671e-01
1.94465905e-01 9.71509159e-01 6.56202257e-01 -4.64822382e-01
4.96397823e-01 4.24997270e-01 -1.60632625e-01 -1.04487991e+00
9.76628996e-03 -5.73493898e-01 -4.05318290e-01 -5.16644299e-01
5.71889758e-01 -1.21349061e+00 -9.23461914e-01 9.35302615e-01
-1.60562003e+00 -7.67319500e-01 -3.64277393e-01 7.32146323e-01
-9.33982790e-01 5.16425848e-01 -4.24442053e-01 -8.28786194e-01
-6.04394265e-02 -1.47168291e+00 1.30444539e+00 4.50963937e-02
-3.70490104e-01 -7.13350594e-01 3.78450930e-01 1.88173354e-01
4.20267582e-01 -7.03828409e-02 3.89840811e-01 -6.43770039e-01
-2.26004720e-01 2.14913085e-01 7.56198317e-02 4.18408126e-01
3.78816128e-02 9.10973668e-01 -1.36840391e+00 -3.28098536e-01
-2.25086629e-01 -3.26935500e-01 6.42434880e-02 5.07228196e-01
8.00795734e-01 -3.00697714e-01 -1.55219227e-01 3.67359877e-01
8.68076324e-01 1.69397786e-01 6.22946501e-01 -5.90192936e-02
4.30904448e-01 6.74457610e-01 7.23052859e-01 6.31332457e-01
4.19462830e-01 1.43122113e+00 3.08117807e-01 -2.23873004e-01
-7.34191656e-01 -3.61792862e-01 8.74646783e-01 1.29905736e+00
1.98656365e-01 -3.95943999e-01 -6.84332013e-01 6.70943201e-01
-1.85758305e+00 -1.01820076e+00 -2.70836532e-01 2.55740929e+00
1.08931506e+00 -2.59253401e-02 3.80540341e-01 5.31013072e-01
8.65759432e-01 2.62017697e-01 -7.77977109e-02 -6.01340115e-01
-6.80435747e-02 4.17647213e-02 2.26716489e-01 6.99404716e-01
-7.70506263e-01 1.02647901e+00 6.51340342e+00 9.82670844e-01
-1.68080866e+00 4.46331203e-01 5.18986844e-02 -3.09859276e-01
6.13685362e-02 9.57578048e-02 -6.26376927e-01 6.32752061e-01
1.49358487e+00 -2.69936979e-01 3.71371865e-01 5.56397676e-01
8.80239069e-01 -9.20361578e-02 -1.61310053e+00 1.22185802e+00
-1.02596082e-01 -1.12710726e+00 -3.78300905e-01 -1.54453635e-01
5.42876780e-01 2.68048257e-01 -2.91856695e-02 4.06471789e-02
-2.59898249e-02 -1.00410104e+00 1.38232553e+00 2.68353552e-01
1.26049364e+00 -3.38779569e-01 3.21068943e-01 4.84250411e-02
-1.68163073e+00 3.94096971e-01 1.21160887e-01 2.78274566e-01
3.73659551e-01 2.21767411e-01 -1.19018090e+00 5.93677461e-01
4.78022099e-01 5.58442235e-01 -5.30572712e-01 1.04383385e+00
-1.87121391e-01 7.17276633e-01 -4.94198233e-01 3.25070828e-01
-1.16342425e-01 6.75641358e-01 6.82398915e-01 1.10650504e+00
4.04512316e-01 -6.04035139e-01 -1.36763886e-01 3.93980980e-01
2.80607760e-01 2.28721067e-01 -6.95530355e-01 -4.12921002e-03
9.10449862e-01 7.80818462e-01 -2.69615710e-01 -3.24747652e-01
-3.94022167e-01 7.90736318e-01 -3.44420731e-01 3.57705295e-01
-1.10809839e+00 -9.39260423e-02 8.63508165e-01 1.41091034e-01
1.31863616e-02 -3.46556008e-01 2.83069909e-01 -1.06298363e+00
8.82932451e-04 -1.22097015e+00 1.91895455e-01 -1.21830595e+00
-6.85987473e-01 5.41712105e-01 1.46571979e-01 -1.71270430e+00
-7.76198566e-01 8.09415802e-02 -7.48358786e-01 5.12285590e-01
-1.46646309e+00 -1.13116026e+00 -2.42035672e-01 9.74136710e-01
8.34091187e-01 -2.44772255e-01 7.22850919e-01 6.02976441e-01
-3.94183517e-01 9.01851237e-01 -1.54739227e-02 -1.75848857e-01
1.36999035e+00 -8.96428764e-01 3.36249799e-01 8.17970693e-01
2.83551127e-01 6.10476956e-02 1.04529989e+00 -3.00842583e-01
-1.27088583e+00 -9.31083143e-01 1.10036385e+00 -2.28964150e-01
7.40143836e-01 -5.32865882e-01 -7.26946950e-01 4.80211258e-01
5.00537157e-01 1.75613508e-01 5.48160195e-01 -4.21284974e-01
-4.53595459e-01 -3.78843039e-01 -7.07854867e-01 6.94047272e-01
6.71977043e-01 -9.35582817e-01 -2.08101660e-01 3.81389648e-01
8.20051730e-01 -6.70471191e-01 -7.65544474e-01 1.93925008e-01
6.84025943e-01 -1.14421844e+00 7.73123443e-01 -2.33469963e-01
-9.66273844e-02 -4.79965568e-01 -7.00424761e-02 -1.19446540e+00
6.55543983e-01 -1.49150383e+00 -3.68092135e-02 1.66479814e+00
1.14830166e-01 -4.38464522e-01 4.02016997e-01 7.19165131e-02
-1.80141032e-01 4.67656069e-02 -1.71438026e+00 -1.07691669e+00
-4.29487675e-01 -8.06530833e-01 6.17631257e-01 8.24986219e-01
2.46755317e-01 1.70423418e-01 -5.23853064e-01 3.14254105e-01
2.22897321e-01 -3.54173660e-01 1.29935288e+00 -8.78272891e-01
2.11619772e-03 -3.51576358e-01 -4.34541970e-01 -6.80439949e-01
3.34668398e-01 -3.16588044e-01 2.60954797e-01 -8.65280926e-01
-4.47350979e-01 -3.61379594e-01 1.55006960e-01 3.40691745e-01
2.76544958e-01 -3.74020189e-02 4.84155267e-01 4.75836098e-01
-5.77959239e-01 4.18444335e-01 9.77740347e-01 -2.64680944e-02
-4.16956991e-01 1.47554547e-01 3.79112870e-01 3.65315497e-01
4.67959970e-01 -6.76496446e-01 -5.91079295e-01 -3.36348414e-01
-2.22948287e-02 5.29899955e-01 3.50535661e-01 -1.42147875e+00
6.67855144e-01 1.60975084e-02 -4.83417958e-01 -5.54164052e-01
5.66846073e-01 -9.00050104e-01 5.51924229e-01 3.12051713e-01
-3.69245648e-01 3.08939964e-01 5.30896187e-01 1.78489432e-01
-6.24453604e-01 8.11677426e-02 7.79551804e-01 6.49964333e-01
-2.60779887e-01 2.05170438e-01 -5.61215878e-01 3.77740599e-02
1.03802121e+00 -1.95575375e-02 -4.62534636e-01 -8.05686593e-01
-5.17939806e-01 7.05309212e-02 4.35292661e-01 6.34450793e-01
3.98443967e-01 -1.36499524e+00 -6.96337819e-01 3.91058087e-01
2.54208148e-01 -2.29920670e-01 3.19489539e-01 1.17241919e+00
-3.49558860e-01 4.50317532e-01 -1.58727169e-03 -1.07366443e+00
-2.00430608e+00 3.66906494e-01 3.81728321e-01 -1.27552867e-01
-2.63082057e-01 3.98915738e-01 -7.24332184e-02 -1.19539080e-02
6.65835381e-01 -6.11135781e-01 9.89919752e-02 2.39193752e-01
2.76323944e-01 3.90891731e-01 2.47348189e-01 -9.05520797e-01
-5.35966516e-01 4.35041547e-01 3.27299863e-01 -7.61309266e-01
7.36269414e-01 -2.42708638e-01 1.97745442e-01 6.05241776e-01
1.10765803e+00 4.23425853e-01 -1.35770512e+00 -1.04272198e-02
-2.10956007e-01 -2.57473379e-01 -1.52584255e-01 -4.25898612e-01
-8.26807201e-01 1.12393093e+00 7.86961138e-01 4.24336821e-01
9.42962408e-01 -1.43787324e-01 6.02291644e-01 -2.55777329e-01
-1.05007761e-03 -1.30446017e+00 3.22839618e-01 3.66351396e-01
7.95214474e-01 -1.01245463e+00 -2.98186630e-01 -2.51178592e-01
-7.10521221e-01 1.28499126e+00 3.93273085e-01 5.17948806e-01
4.22627598e-01 8.29300880e-01 5.26332855e-01 3.38171422e-01
-1.27197790e+00 -1.19193286e-01 1.06007360e-01 7.61009932e-01
3.21864963e-01 -1.94928616e-01 -6.15755916e-02 2.35629499e-01
-1.05915479e-01 1.57384276e-01 5.47422230e-01 8.74742270e-01
2.38733575e-01 -1.28205013e+00 -5.30881226e-01 -5.09927630e-01
-4.98268664e-01 -9.63214636e-02 -1.96665660e-01 8.86153221e-01
-8.86118039e-02 1.24273360e+00 2.20728204e-01 -4.48662102e-01
1.83066770e-01 1.52926400e-01 3.69204313e-01 -3.02757621e-01
-7.75550961e-01 5.28568983e-01 1.82753712e-01 -8.76223385e-01
-7.76509702e-01 -6.45017147e-01 -1.16065753e+00 -3.78247559e-01
-3.69046986e-01 1.08829789e-01 1.05791461e+00 8.55172992e-01
6.56031728e-01 5.47433794e-01 8.82337630e-01 -1.05367565e+00
-3.10361326e-01 -1.00437593e+00 -1.99798688e-01 3.87594551e-01
6.55328512e-01 -4.04363900e-01 -6.81456804e-01 4.45070624e-01] | [14.51187801361084, 5.200313568115234] |
6320dea6-bde8-46c2-940f-027c3435ddb5 | efficient-human-pose-estimation-via-3d-event | 2206.04511 | null | https://arxiv.org/abs/2206.04511v2 | https://arxiv.org/pdf/2206.04511v2.pdf | Efficient Human Pose Estimation via 3D Event Point Cloud | Human Pose Estimation (HPE) based on RGB images has experienced a rapid development benefiting from deep learning. However, event-based HPE has not been fully studied, which remains great potential for applications in extreme scenes and efficiency-critical conditions. In this paper, we are the first to estimate 2D human pose directly from 3D event point cloud. We propose a novel representation of events, the rasterized event point cloud, aggregating events on the same position of a small time slice. It maintains the 3D features from multiple statistical cues and significantly reduces memory consumption and computation complexity, proved to be efficient in our work. We then leverage the rasterized event point cloud as input to three different backbones, PointNet, DGCNN, and Point Transformer, with two linear layer decoders to predict the location of human keypoints. We find that based on our method, PointNet achieves promising results with much faster speed, whereas Point Transfomer reaches much higher accuracy, even close to previous event-frame-based methods. A comprehensive set of results demonstrates that our proposed method is consistently effective for these 3D backbone models in event-driven human pose estimation. Our method based on PointNet with 2048 points input achieves 82.46mm in MPJPE3D on the DHP19 dataset, while only has a latency of 12.29ms on an NVIDIA Jetson Xavier NX edge computing platform, which is ideally suitable for real-time detection with event cameras. Code is available at https://github.com/MasterHow/EventPointPose. | ['Kaiwei Wang', 'Lei Sun', 'Kailun Yang', 'Yaozu Ye', 'Hao Shi', 'Jiaan Chen'] | 2022-06-09 | null | null | null | null | ['2048'] | ['playing-games'] | [-2.46068001e-01 -4.70766336e-01 9.08522829e-02 -8.25917497e-02
-6.81071818e-01 -1.45749420e-01 2.50249952e-01 1.64827660e-01
-7.89569199e-01 4.30610359e-01 -3.41187753e-02 1.31021842e-01
2.19306901e-01 -7.67938018e-01 -9.35756803e-01 -2.72365749e-01
-4.03049707e-01 5.77184260e-01 5.68111956e-01 -1.07836187e-01
-9.30832252e-02 7.04107881e-01 -1.68915987e+00 1.73595011e-01
1.62307322e-01 1.33457470e+00 -3.81006487e-02 8.47535312e-01
4.31901962e-01 4.94030982e-01 -5.84536433e-01 -3.25331837e-01
4.22913074e-01 3.60476449e-02 -3.15364033e-01 -1.67368695e-01
5.25971413e-01 -7.54388094e-01 -6.23826087e-01 3.86017352e-01
1.04032576e+00 5.83492182e-02 6.61260411e-02 -1.55389476e+00
1.46544769e-01 -1.87954232e-01 -6.88925445e-01 3.18330288e-01
6.91025555e-01 3.90588939e-01 6.60574496e-01 -1.19867563e+00
7.88217008e-01 1.21871841e+00 1.04870951e+00 3.17341864e-01
-7.88949668e-01 -8.36754560e-01 -1.11184917e-01 5.37848055e-01
-1.54753244e+00 -1.61143720e-01 6.26794219e-01 -8.46165046e-02
1.41045296e+00 1.50602683e-01 1.40098321e+00 1.40136981e+00
6.46820664e-01 9.01858091e-01 7.05451429e-01 -8.93475413e-02
1.80504084e-01 -4.60669160e-01 -1.89147919e-01 7.33235538e-01
1.77269414e-01 3.34492266e-01 -1.14862478e+00 -2.31701970e-01
1.07476449e+00 4.17379946e-01 -2.36817256e-01 -4.46521580e-01
-1.50952995e+00 5.41735649e-01 6.75808668e-01 -3.57855350e-01
-6.93507791e-01 7.36736774e-01 4.68150616e-01 -1.38682788e-02
3.54247808e-01 -2.78224170e-01 -5.51473260e-01 -5.70843041e-01
-8.16080928e-01 6.16546035e-01 5.75433612e-01 1.04451311e+00
2.66431183e-01 -3.02290857e-01 9.27925855e-02 1.79709956e-01
3.06186587e-01 8.42124283e-01 -4.50682733e-03 -8.85175109e-01
5.12974560e-01 5.68749428e-01 2.47764662e-01 -1.20064783e+00
-8.24055016e-01 -1.65113285e-01 -7.19787538e-01 3.51393998e-01
3.99837524e-01 -2.42911845e-01 -6.80344462e-01 1.39329791e+00
6.97734654e-01 4.29930568e-01 -4.43707198e-01 1.25522614e+00
6.80473089e-01 7.86909282e-01 1.54921591e-01 2.52180338e-01
1.51617336e+00 -5.87710142e-01 -2.98770249e-01 -2.02422500e-01
3.41754019e-01 -5.71695745e-01 7.73937941e-01 6.36017919e-01
-1.02780175e+00 -5.87803364e-01 -1.09416974e+00 -3.66529942e-01
-5.17142341e-02 1.90749168e-01 8.42926323e-01 3.43323290e-01
-9.12246585e-01 4.26697642e-01 -1.56650734e+00 -7.03780770e-01
3.36669922e-01 5.64153910e-01 -4.55980897e-01 1.36585906e-01
-1.01143491e+00 7.40733504e-01 2.65418082e-01 3.62052560e-01
-7.01202512e-01 -7.15058625e-01 -8.60106468e-01 -2.49034930e-02
3.75354588e-01 -1.01460385e+00 1.25620866e+00 -2.20210493e-01
-1.40586603e+00 6.69924974e-01 -2.29383439e-01 -7.10066736e-01
7.10146904e-01 -7.18260884e-01 -2.19753489e-01 4.55535233e-01
5.76801412e-02 9.87949252e-01 5.69840193e-01 -7.81924009e-01
-8.44467282e-01 -5.92364311e-01 -1.50133967e-01 3.13245624e-01
-5.83906844e-02 3.46282497e-02 -9.35199857e-01 -4.54034060e-01
2.05855355e-01 -1.18884897e+00 -2.51905650e-01 5.73422670e-01
-3.70691389e-01 -1.45267114e-01 6.58831656e-01 -5.33428192e-01
9.64062572e-01 -2.04616952e+00 -2.53954493e-02 2.13314116e-01
2.03804091e-01 -9.80085041e-03 3.12786311e-01 4.88618821e-01
4.19645943e-02 -5.22956133e-01 6.17266782e-02 -6.36612117e-01
5.73638305e-02 -2.11364068e-02 -2.14220881e-01 7.56245792e-01
1.02772027e-01 1.08937943e+00 -7.68674135e-01 -5.77868938e-01
7.99392760e-01 9.69584465e-01 -6.58261299e-01 -9.64114517e-02
9.50348675e-02 2.82478392e-01 -2.70828336e-01 7.97571301e-01
6.02131963e-01 -3.59912544e-01 -2.49088444e-02 -3.88449848e-01
-1.67738125e-01 3.81709775e-03 -1.40306866e+00 2.15450239e+00
-1.46459982e-01 6.25864804e-01 -2.52237409e-01 -3.60892951e-01
5.44238329e-01 2.42082104e-01 8.87364447e-01 -7.52515972e-01
3.64655167e-01 3.45497159e-03 -4.51581270e-01 -1.74815863e-01
6.41198337e-01 2.48759791e-01 -3.88355792e-01 1.33450404e-01
4.81047034e-02 2.99775988e-01 -1.73160091e-01 1.70818955e-01
1.41734385e+00 3.83677334e-01 3.71311069e-01 2.31007904e-01
-8.09625983e-02 2.68327206e-01 5.76485276e-01 5.95738471e-01
-4.31762397e-01 8.46303284e-01 1.72039300e-01 -6.00634992e-01
-1.04502869e+00 -1.46076238e+00 -1.27582982e-01 7.47957587e-01
5.07091105e-01 -8.26418638e-01 -4.38735723e-01 -2.87950605e-01
1.61363885e-01 2.49472648e-01 -2.56900251e-01 -3.74161936e-02
-8.83596480e-01 -6.17157280e-01 5.21497726e-01 8.91075790e-01
7.02910364e-01 -9.53598142e-01 -1.56050706e+00 3.03452402e-01
-2.19739646e-01 -1.24318194e+00 -3.71478736e-01 2.46608183e-02
-7.99030900e-01 -1.19415331e+00 -6.31723523e-01 -5.34093559e-01
3.22601140e-01 2.37140551e-01 1.09878302e+00 -1.68074757e-01
-6.26663446e-01 7.21034110e-01 -3.15794349e-01 -5.84402740e-01
5.56766748e-01 -1.15397885e-01 1.87219694e-01 -2.08157748e-01
6.62881613e-01 -4.71564174e-01 -1.13864410e+00 3.23776871e-01
-4.23818678e-01 1.36004612e-01 4.55039144e-01 4.18522507e-01
8.70004594e-01 -5.97336367e-02 -9.96296778e-02 -1.34336069e-01
2.78434716e-02 -3.39185387e-01 -6.29571676e-01 -2.23420843e-01
-7.45385215e-02 -4.82170314e-01 2.63241619e-01 -2.73218364e-01
-9.06856179e-01 4.76245135e-01 -2.01780200e-01 -6.93663359e-01
-1.93721473e-01 1.01685867e-01 5.57700954e-02 4.90629598e-02
6.19957983e-01 1.63426884e-02 -6.07784726e-02 -2.25251749e-01
1.23279013e-01 2.48866037e-01 8.04404795e-01 -4.43838865e-01
5.05996644e-01 1.04404569e+00 1.36387840e-01 -7.98386395e-01
-2.98226267e-01 -6.29641712e-01 -6.71788335e-01 -4.76752728e-01
9.47900176e-01 -1.30696261e+00 -1.30568862e+00 7.17330575e-01
-1.47568798e+00 -2.32680559e-01 -2.76972324e-01 7.18043268e-01
-5.65796018e-01 1.23256385e-01 -7.57172048e-01 -6.90251887e-01
-4.20949370e-01 -8.73779953e-01 1.63941753e+00 9.52775702e-02
-3.54039282e-01 -5.02575815e-01 3.04275248e-02 -2.02168450e-02
-1.24015950e-01 7.35830605e-01 6.03687018e-02 -7.89802149e-02
-8.95466924e-01 -4.88234729e-01 -1.33916035e-01 -2.52244681e-01
-3.78390759e-01 -2.01367781e-01 -6.44050360e-01 -4.38177347e-01
-8.92621186e-03 -1.35779396e-01 5.26885211e-01 6.50691152e-01
1.08413172e+00 2.43149176e-01 -7.47098088e-01 8.72752309e-01
1.38591635e+00 1.70193948e-02 6.96339905e-01 3.78229350e-01
7.65101433e-01 1.24115035e-01 7.81945407e-01 9.82507467e-01
7.25311100e-01 9.58763838e-01 5.65036058e-01 -7.03104287e-02
-2.49718651e-01 -3.36835891e-01 4.48815227e-01 4.71720040e-01
-2.80159861e-01 -3.09362918e-01 -1.07986319e+00 3.34392965e-01
-1.97898090e+00 -9.76144314e-01 -2.52285361e-01 2.22854233e+00
3.85295540e-01 3.51128757e-01 3.07224214e-01 -7.90331606e-03
6.56735957e-01 1.00350678e-01 -6.66335464e-01 1.93892151e-01
1.77389503e-01 4.66799855e-01 6.80892348e-01 -8.30987468e-03
-1.16675329e+00 7.32436061e-01 5.43113041e+00 4.60774601e-01
-9.57517445e-01 1.77874461e-01 1.88010633e-01 -8.90206158e-01
5.24642527e-01 -1.88509345e-01 -1.15780067e+00 6.39144182e-01
9.25698876e-01 7.22188205e-02 -4.75105569e-02 8.82676065e-01
2.96632618e-01 -2.58313864e-01 -1.23678339e+00 1.58350348e+00
3.80929862e-03 -1.19926810e+00 -2.29124770e-01 1.67216465e-01
2.14054331e-01 2.88780391e-01 -1.45514548e-01 1.04340121e-01
-8.20464715e-02 -7.12926745e-01 9.98129725e-01 4.32154506e-01
7.47732818e-01 -1.00224376e+00 5.88990629e-01 2.86521643e-01
-1.58998108e+00 1.22308150e-01 -3.33288819e-01 -1.59674793e-01
7.15566039e-01 5.18767118e-01 -7.79105484e-01 3.53836417e-01
1.26333737e+00 7.76772499e-01 -4.32844222e-01 1.19208455e+00
-1.50879756e-01 2.83970475e-01 -9.83056605e-01 -2.52371803e-02
-1.50027126e-01 3.34935516e-01 6.26947820e-01 1.01266980e+00
4.99363482e-01 3.61337006e-01 2.80682266e-01 4.57688183e-01
9.94919613e-02 -3.09694022e-01 -4.39195573e-01 6.30367756e-01
5.61508775e-01 1.10423517e+00 -9.25349653e-01 -1.72071919e-01
-4.47021306e-01 1.33324039e+00 1.72054812e-01 -1.77331436e-02
-1.40925324e+00 -4.38136190e-01 6.42590344e-01 3.27424705e-01
4.28802192e-01 -8.55497956e-01 -1.84611827e-01 -1.15716076e+00
3.60670358e-01 -5.97526312e-01 5.25056362e-01 -8.99415970e-01
-9.93902504e-01 4.61426347e-01 1.81725711e-01 -1.51838541e+00
-2.13870734e-01 -8.05210531e-01 -3.77337188e-01 4.06826526e-01
-1.01738667e+00 -1.08336532e+00 -7.12471783e-01 9.85824168e-01
3.89023006e-01 2.83060044e-01 5.74882865e-01 5.52404642e-01
-4.65932131e-01 5.08564711e-01 -2.96628088e-01 2.08204567e-01
6.14548981e-01 -1.08025551e+00 9.21884000e-01 8.74701738e-01
1.90927118e-01 4.43033993e-01 5.48646450e-01 -1.00671327e+00
-1.93229520e+00 -9.46922481e-01 8.11484873e-01 -8.38246047e-01
2.42152065e-01 -6.44881904e-01 -4.43821013e-01 8.84851038e-01
-1.48424432e-01 2.18862578e-01 3.89700770e-01 2.82161385e-02
5.55545874e-02 -1.17556192e-01 -9.58132923e-01 7.13998497e-01
1.45510459e+00 -2.79166579e-01 -4.29891586e-01 5.03689766e-01
6.12636089e-01 -9.95114148e-01 -7.81398892e-01 4.13109839e-01
6.11227632e-01 -1.10217094e+00 1.42645884e+00 -7.36506563e-03
1.18432328e-01 -3.68540466e-01 -6.88051358e-02 -7.68380225e-01
-2.23215863e-01 -4.65555280e-01 -5.37453830e-01 5.94867647e-01
-1.80608481e-01 -4.70914006e-01 1.25750649e+00 5.36306083e-01
-7.76591673e-02 -8.27908814e-01 -1.25297940e+00 -7.21341670e-01
-6.35718167e-01 -9.60619867e-01 4.48305666e-01 2.37715781e-01
-2.75594622e-01 6.41542524e-02 -3.38582844e-01 4.66406018e-01
1.00027120e+00 -9.39245075e-02 1.07765675e+00 -9.77039993e-01
-9.32919905e-02 -2.93285679e-02 -7.83657610e-01 -1.25971174e+00
-2.41960511e-01 -5.24167776e-01 5.93603514e-02 -1.37758553e+00
-1.36397928e-01 -1.98427081e-01 -2.41560838e-03 4.40886557e-01
1.98716414e-03 5.18551886e-01 2.46753320e-01 1.37951031e-01
-7.70002782e-01 6.63396716e-01 7.93365300e-01 2.28222445e-01
-1.35643333e-01 -1.47834629e-01 2.52695121e-02 9.36743975e-01
6.09184027e-01 -4.02228475e-01 -1.30329728e-01 -5.31161547e-01
1.81200325e-01 1.68785423e-01 1.13773823e+00 -1.80429292e+00
7.61670113e-01 2.84198880e-01 1.01421285e+00 -1.20413411e+00
8.52129936e-01 -9.75676358e-01 5.40882707e-01 6.81141019e-01
2.68402994e-01 6.33593678e-01 3.13022196e-01 6.29559398e-01
7.84409493e-02 4.76036698e-01 4.13349479e-01 -9.34168845e-02
-1.21636581e+00 7.15484798e-01 2.41963584e-02 -1.42396078e-01
1.24752963e+00 -4.66730982e-01 8.95777903e-03 -1.55990914e-01
-5.53365886e-01 1.99112132e-01 4.92245972e-01 4.65921611e-01
1.01123524e+00 -1.58676398e+00 -5.86033583e-01 1.80165157e-01
-4.17892449e-02 2.44690508e-01 4.78405327e-01 7.94102728e-01
-9.73997772e-01 4.10013586e-01 -2.82839507e-01 -1.15987504e+00
-1.18399668e+00 3.23014349e-01 1.01124369e-01 6.02194434e-03
-1.18799865e+00 9.45329785e-01 -1.70330435e-01 -1.44462541e-01
3.22628707e-01 -3.60521466e-01 3.92341554e-01 -6.81380108e-02
4.41359013e-01 7.41506636e-01 1.83884099e-01 -6.24711692e-01
-7.57510185e-01 7.53615379e-01 1.81273013e-01 -1.98730841e-01
1.13404751e+00 -2.62019294e-03 2.50495642e-01 2.62175083e-01
1.17575991e+00 -2.56266713e-01 -1.66818488e+00 1.68613642e-02
-3.11767101e-01 -4.99819487e-01 -3.10049821e-02 -4.58574504e-01
-9.26668525e-01 8.08005869e-01 9.57789660e-01 -3.20418805e-01
1.15115213e+00 -7.33171850e-02 1.43544900e+00 2.29102209e-01
1.08402383e+00 -9.43156421e-01 9.74330828e-02 2.93400675e-01
7.44892478e-01 -1.20981753e+00 2.63000995e-01 -4.07578140e-01
-3.73124957e-01 9.45237696e-01 8.30119669e-01 -4.08207387e-01
7.24889815e-01 3.31710935e-01 -2.67098099e-01 -4.88207370e-01
-7.18356490e-01 -1.32012814e-01 1.32119715e-01 4.54668403e-01
2.76338886e-02 -4.33922075e-02 1.67420562e-02 4.01152432e-01
-3.96831125e-01 2.10484743e-01 -2.21797302e-02 1.27594423e+00
-2.61195034e-01 -7.31090128e-01 -5.77822089e-01 2.62576491e-01
-2.88700938e-01 1.38263199e-02 1.55233115e-01 1.11201990e+00
2.90716141e-01 6.04328096e-01 4.61879760e-01 -5.49187183e-01
8.12407196e-01 -1.90463066e-01 6.71933889e-01 -3.56234848e-01
-6.81643069e-01 -1.27248257e-01 -1.51289493e-01 -1.30682027e+00
-2.52240449e-01 -8.64418805e-01 -1.50511360e+00 -6.41644239e-01
9.91592463e-03 -3.47210437e-01 6.91446364e-01 5.69991946e-01
8.87245417e-01 2.77281761e-01 1.13777786e-01 -1.41644514e+00
-1.09248668e-01 -4.55583304e-01 -4.49500114e-01 3.06906790e-01
8.72183815e-02 -8.61167789e-01 -1.07057214e-01 -4.45521735e-02] | [7.216372489929199, -0.9016264081001282] |
7c77e79b-8ff6-47c1-b42c-5c36252537b5 | face-from-depth-for-head-pose-estimation-on | 1712.05277 | null | http://arxiv.org/abs/1712.05277v2 | http://arxiv.org/pdf/1712.05277v2.pdf | Face-from-Depth for Head Pose Estimation on Depth Images | Depth cameras allow to set up reliable solutions for people monitoring and
behavior understanding, especially when unstable or poor illumination
conditions make unusable common RGB sensors. Therefore, we propose a complete
framework for the estimation of the head and shoulder pose based on depth
images only. A head detection and localization module is also included, in
order to develop a complete end-to-end system. The core element of the
framework is a Convolutional Neural Network, called POSEidon+, that receives as
input three types of images and provides the 3D angles of the pose as output.
Moreover, a Face-from-Depth component based on a Deterministic Conditional GAN
model is able to hallucinate a face from the corresponding depth image. We
empirically demonstrate that this positively impacts the system performances.
We test the proposed framework on two public datasets, namely Biwi Kinect Head
Pose and ICT-3DHP, and on Pandora, a new challenging dataset mainly inspired by
the automotive setup. Experimental results show that our method overcomes
several recent state-of-art works based on both intensity and depth input data,
running in real-time at more than 30 frames per second. | ['Rita Cucchiara', 'Roberto Vezzani', 'Matteo Fabbri', 'Guido Borghi', 'Simone Calderara'] | 2017-12-12 | null | null | null | null | ['head-detection', 'head-pose-estimation'] | ['computer-vision', 'computer-vision'] | [-1.00603402e-01 1.64603084e-01 3.02081615e-01 -7.24821866e-01
-6.26231611e-01 -2.34047756e-01 4.56902921e-01 -5.22042215e-01
-7.53367603e-01 4.71357733e-01 -2.04643477e-02 2.20652580e-01
4.37480479e-01 -8.43074322e-01 -7.92162418e-01 -7.37573624e-01
2.47104079e-01 5.77750742e-01 -1.30748870e-02 -1.14475880e-02
-1.60216928e-01 4.57603216e-01 -2.12773538e+00 -9.12059322e-02
4.73726183e-01 1.21742773e+00 1.16504520e-01 6.95410907e-01
3.59764427e-01 4.71598804e-01 -5.27049780e-01 -5.44541478e-01
1.22717366e-01 -1.98651776e-01 -2.71800965e-01 4.99306649e-01
6.30839705e-01 -9.30690289e-01 -3.46980542e-01 7.84687698e-01
1.08119357e+00 -1.37532234e-01 2.85419911e-01 -1.31543708e+00
-9.46111605e-03 -2.94126626e-02 -4.03757155e-01 -3.06454986e-01
9.34647381e-01 6.30964637e-01 4.33494985e-01 -7.75559127e-01
3.77912074e-01 1.35834992e+00 5.77671170e-01 9.70398366e-01
-7.45714903e-01 -6.20054603e-01 7.48860389e-02 3.65145594e-01
-1.47445202e+00 -6.94503665e-01 6.90823317e-01 -8.51022974e-02
6.79125190e-01 1.61155969e-01 7.85440981e-01 1.72083282e+00
-3.28540534e-01 8.71520460e-01 1.11492765e+00 -2.42312148e-01
3.86139899e-01 1.19236998e-01 -1.09834082e-01 7.10568547e-01
1.53131470e-01 -2.03899089e-02 -6.55279756e-01 1.64933652e-01
7.05272913e-01 2.43890598e-01 -5.20029604e-01 -3.17574888e-01
-6.96429431e-01 5.56233525e-01 5.90856552e-01 -5.54105677e-02
-4.86057609e-01 7.23630488e-02 6.20353632e-02 -1.66497558e-01
3.97770435e-01 -6.34828508e-01 -2.89855748e-01 -3.61934781e-01
-7.53666162e-01 3.40179712e-01 8.89533460e-01 8.37921143e-01
6.91846490e-01 -3.39917272e-01 -1.16370216e-01 4.74531054e-01
6.71354532e-01 9.66309011e-01 2.68546671e-01 -6.91447556e-01
5.22829413e-01 7.05904901e-01 1.70876130e-01 -7.47839510e-01
-9.05811429e-01 -1.36639982e-01 -7.00478792e-01 3.19331735e-01
5.50326109e-01 -2.85871714e-01 -7.78348327e-01 1.68358874e+00
7.42367327e-01 1.40227497e-01 -2.85815775e-01 1.37007225e+00
1.02018380e+00 2.45017186e-01 -1.87466294e-01 -8.32735598e-02
1.63377619e+00 -7.94017255e-01 -7.63036788e-01 -5.06646037e-01
3.30390602e-01 -2.44811431e-01 1.11541343e+00 6.41047835e-01
-1.04882896e+00 -6.37470663e-01 -8.39483261e-01 -2.13203549e-01
-1.51231900e-01 5.43414772e-01 3.73844534e-01 1.23553574e+00
-1.26087093e+00 2.85846412e-01 -1.19706082e+00 -6.18366003e-01
3.18756610e-01 5.75897574e-01 -6.36884809e-01 -2.68148005e-01
-7.91217387e-01 7.88616419e-01 9.82559565e-03 6.31935477e-01
-8.31956327e-01 -2.04280540e-01 -9.38430011e-01 -1.48308441e-01
1.81182638e-01 -8.73049915e-01 1.23077393e+00 -6.78960145e-01
-1.98018646e+00 1.05107701e+00 -2.67918140e-01 -2.11234555e-01
8.74830723e-01 -5.81401408e-01 -5.80705814e-02 2.48874143e-01
-1.97933733e-01 6.15533650e-01 9.04853225e-01 -1.07494128e+00
-2.70883143e-01 -1.16265988e+00 1.83009859e-02 2.23664671e-01
-5.43757737e-01 -1.36780456e-01 -1.13051331e+00 1.35057038e-02
-7.82474652e-02 -9.73907709e-01 1.87593281e-01 2.32854947e-01
-5.78341126e-01 -1.25783265e-01 7.20041811e-01 -8.94366682e-01
6.46445215e-01 -1.85531855e+00 2.76458949e-01 3.46283577e-02
1.62390277e-01 1.30336016e-01 2.67974615e-01 -7.39397183e-02
7.04699084e-02 -6.21424139e-01 -1.83328524e-01 -1.28454769e+00
1.35304347e-01 1.69775978e-01 3.75012964e-01 9.11051810e-01
-8.31417367e-02 7.87792265e-01 -4.02730018e-01 -2.54655063e-01
5.81065238e-01 1.17835605e+00 -4.46625978e-01 6.91876173e-01
-1.83161851e-02 7.91417539e-01 -3.04325558e-02 7.67560124e-01
1.16244245e+00 8.89017805e-02 1.37206629e-01 -1.01015009e-01
7.02381432e-02 5.69359139e-02 -1.25355053e+00 2.04817963e+00
-5.05611837e-01 2.01765016e-01 2.58950830e-01 -6.21527493e-01
8.49575818e-01 4.40420151e-01 1.27830803e-01 -8.08642924e-01
4.73208457e-01 3.84820183e-03 -5.36872447e-01 -8.30953121e-01
1.10683396e-01 8.78358781e-02 9.49406400e-02 3.13490361e-01
3.77496481e-02 1.56504303e-01 -2.50942290e-01 -1.91378653e-01
1.03375340e+00 3.66069257e-01 -1.01107858e-01 3.04675788e-01
6.81194365e-01 -8.18271399e-01 2.94372112e-01 3.40991646e-01
-2.23270223e-01 1.10911822e+00 3.96686018e-01 -4.41968530e-01
-6.82886839e-01 -9.85686183e-01 1.76526174e-01 9.66777861e-01
-3.79731320e-03 -3.38918507e-01 -1.32111537e+00 -7.45530069e-01
-1.81507170e-01 3.73125404e-01 -8.68640006e-01 1.12485856e-01
-3.78122985e-01 -8.57380331e-01 4.58106399e-01 5.74102402e-01
7.61041701e-01 -1.09903145e+00 -9.59819794e-01 -2.10599214e-01
-3.62039357e-01 -1.59928477e+00 -4.58670892e-02 -9.00887698e-02
-5.91175318e-01 -1.05832362e+00 -1.04156590e+00 -2.61460602e-01
6.35269880e-01 5.01118228e-02 9.31693971e-01 2.77441423e-02
-2.76257932e-01 5.95646024e-01 -2.37611502e-01 -2.92577624e-01
1.59820989e-01 1.56021401e-01 1.40576899e-01 3.77957851e-01
5.27764499e-01 -5.64353824e-01 -8.93134534e-01 2.74618894e-01
-5.77122509e-01 -4.93624136e-02 5.20484507e-01 2.78585494e-01
1.99461013e-01 -4.43294525e-01 1.83798403e-01 -5.29448986e-01
1.00213066e-01 -2.16654286e-01 -6.69323206e-01 -8.30992684e-02
-1.80988207e-01 -1.91751152e-01 2.95857996e-01 -9.00549442e-02
-1.03896785e+00 5.73845327e-01 -7.10161567e-01 -3.73837292e-01
-6.92065179e-01 -1.93300813e-01 -8.71012926e-01 1.68211937e-01
6.00266755e-01 2.06773475e-01 2.10754260e-01 -6.71875358e-01
2.89916724e-01 9.92853522e-01 8.69605780e-01 -2.14959607e-01
6.57120526e-01 8.56639147e-01 -2.06546366e-01 -9.13232028e-01
-6.22835219e-01 -4.97732550e-01 -9.12909985e-01 -4.20140386e-01
9.77164805e-01 -1.26806331e+00 -1.37386632e+00 1.00159478e+00
-1.32769299e+00 -1.72434747e-01 2.08183900e-01 3.38236898e-01
-5.39175808e-01 2.27722913e-01 -4.49341357e-01 -1.16994131e+00
-4.64943916e-01 -1.21080911e+00 1.65111041e+00 5.40422976e-01
1.11231133e-01 -6.93711638e-01 -1.58858210e-01 6.21957958e-01
3.01287830e-01 4.96836245e-01 -3.89716551e-02 -1.09506339e-01
-4.53855515e-01 -4.11731482e-01 -7.29863718e-02 2.80582666e-01
-5.78085184e-02 -3.43673408e-01 -1.69967616e+00 -4.00548100e-01
2.09298819e-01 -5.41908741e-01 7.00458646e-01 3.76796931e-01
1.10922062e+00 -1.58958539e-01 -1.72286525e-01 8.10375869e-01
1.14105558e+00 -4.15439934e-01 9.32057142e-01 2.83585131e-01
9.10325527e-01 7.70266771e-01 2.79714376e-01 8.57634962e-01
8.00562263e-01 1.12695527e+00 9.19450283e-01 -1.76904052e-01
-7.79060721e-02 -5.80030791e-02 6.61077023e-01 3.56079817e-01
-3.56146425e-01 -2.54534632e-01 -7.06586778e-01 1.76789731e-01
-1.73213327e+00 -6.81448817e-01 -1.96537524e-01 2.40662289e+00
6.27974570e-01 4.51180860e-02 6.26063228e-01 3.95898759e-01
5.20083845e-01 -1.26723617e-01 -5.77910841e-01 1.78897038e-01
6.66342527e-02 4.02275801e-01 2.48044878e-01 3.95254403e-01
-1.04206765e+00 7.53600299e-01 4.86909962e+00 2.13830858e-01
-1.20401514e+00 1.89442128e-01 4.79783416e-01 -4.37995553e-01
1.93180770e-01 -7.25111723e-01 -9.61788654e-01 5.00377119e-01
1.03962171e+00 6.86441481e-01 3.45932364e-01 1.08619714e+00
3.96614760e-01 -3.13192427e-01 -1.14114940e+00 1.44239199e+00
4.67873871e-01 -5.47036767e-01 -4.95572448e-01 1.77139804e-01
1.41717315e-01 4.74664308e-02 9.63029340e-02 1.72671452e-01
-3.13503623e-01 -1.13786840e+00 8.13695788e-01 6.38772905e-01
8.88886929e-01 -1.00026238e+00 9.28168476e-01 5.96088290e-01
-1.00127959e+00 6.21208437e-02 -2.71985650e-01 -1.67047992e-01
1.33911043e-01 6.67746663e-01 -8.97697866e-01 4.15222496e-01
9.66794372e-01 5.38142681e-01 -8.24076056e-01 8.36874068e-01
-7.58516967e-01 2.40102038e-01 -5.55889249e-01 1.20060861e-01
-3.23562741e-01 1.78646389e-02 1.89232305e-01 1.08074272e+00
3.13647926e-01 -1.69136543e-02 -1.57231301e-01 7.07410574e-01
-6.42322451e-02 -3.15649718e-01 -4.93185431e-01 6.15768790e-01
2.01174736e-01 1.51602709e+00 -5.72715700e-01 -1.20051563e-01
-3.23033750e-01 1.43445277e+00 1.95959687e-01 1.72919616e-01
-1.09507442e+00 -1.08407095e-01 7.68968940e-01 2.29232311e-01
3.92543137e-01 -1.72579542e-01 -1.99783534e-01 -1.25019026e+00
4.06474769e-01 -8.85931909e-01 8.56688544e-02 -7.86772847e-01
-8.67548704e-01 7.05157578e-01 -2.03462020e-01 -1.04926205e+00
-6.13186657e-01 -7.80255497e-01 -3.65152299e-01 8.73981714e-01
-1.49502063e+00 -1.48711526e+00 -1.22383344e+00 9.74821687e-01
3.15831393e-01 2.64660090e-01 1.02145231e+00 5.44392049e-01
-9.74829972e-01 8.85377884e-01 -4.79587287e-01 1.90846622e-01
6.49564087e-01 -1.13171804e+00 5.68973660e-01 7.65463531e-01
-1.26299992e-01 3.55133653e-01 6.81429982e-01 -2.48329297e-01
-1.78041685e+00 -9.03729022e-01 5.45876563e-01 -7.68722415e-01
-2.43853077e-01 -9.13857281e-01 -5.60849488e-01 6.01224065e-01
1.88718997e-02 3.28169793e-01 3.25985372e-01 -9.45703238e-02
-1.08698942e-01 -1.75979361e-01 -1.40172696e+00 2.55682856e-01
1.25764585e+00 -3.82550329e-01 -2.12661639e-01 2.43257880e-01
2.78339714e-01 -6.89149022e-01 -3.09060335e-01 2.52325714e-01
7.47630954e-01 -1.62798798e+00 9.75700498e-01 1.12372734e-01
1.54232711e-01 -2.86691278e-01 -2.02523787e-02 -1.00511301e+00
3.30119729e-01 -5.43242574e-01 -4.32334363e-01 1.18456054e+00
-1.77322820e-01 -4.04315084e-01 1.12752676e+00 8.32252085e-01
2.87584335e-01 -5.81830621e-01 -1.09564281e+00 -4.24303561e-01
-5.49231827e-01 -8.20873260e-01 6.00358546e-01 2.50287950e-01
-3.94653916e-01 3.33926439e-01 -4.50177222e-01 4.54668075e-01
8.74132752e-01 -2.03313306e-01 1.35536313e+00 -1.17667520e+00
-3.12488824e-01 1.19966595e-02 -7.30613708e-01 -1.25939226e+00
1.76226884e-01 -2.19823152e-01 1.40190959e-01 -1.23349214e+00
6.98009208e-02 9.66332704e-02 2.86091089e-01 5.61506331e-01
-1.58919141e-01 6.28776312e-01 2.10877389e-01 -1.26699865e-01
-5.54544866e-01 7.66774535e-01 7.37681091e-01 1.31247044e-01
-1.08080462e-01 3.54887694e-01 -4.33843285e-01 1.01017261e+00
3.44380677e-01 -2.28900000e-01 -1.23940296e-01 -4.80793327e-01
-1.76892411e-02 -1.67241320e-02 7.58786738e-01 -1.41807580e+00
2.85223693e-01 3.81707400e-01 7.29588449e-01 -6.28585160e-01
8.62282395e-01 -8.81257236e-01 -6.35284409e-02 5.23306668e-01
2.73500174e-01 -7.66875073e-02 8.31242576e-02 3.75881642e-01
1.65893286e-01 1.27683237e-01 7.66439915e-01 5.02450541e-02
-5.03831208e-01 3.09705615e-01 4.95841354e-02 -2.98926264e-01
9.89918530e-01 -1.95899308e-01 -7.95435086e-02 -7.19498813e-01
-6.70748651e-01 -6.35133386e-02 5.00849903e-01 3.83122146e-01
7.12493479e-01 -1.32459629e+00 -6.18889928e-01 6.90209627e-01
1.32900447e-01 2.98105955e-01 3.13133627e-01 8.76292825e-01
-5.38349926e-01 2.68771112e-01 -2.54585177e-01 -9.08847034e-01
-1.40276027e+00 1.91132978e-01 4.64672416e-01 1.22819737e-01
-6.40824020e-01 9.04649615e-01 5.97431883e-02 -6.03258610e-01
6.82226658e-01 -4.11260247e-01 -2.62562543e-01 6.31437898e-02
9.81276155e-01 2.77379662e-01 4.51942891e-01 -8.40908408e-01
-6.42542243e-01 6.82275176e-01 3.04885149e-01 -4.22385514e-01
1.44733858e+00 -3.81125420e-01 5.12095392e-02 2.39404812e-01
1.24342072e+00 -1.59794316e-01 -1.67968488e+00 3.53787504e-02
-4.03038651e-01 -4.83719379e-01 -3.06377243e-02 -6.25249505e-01
-1.30495083e+00 1.07656229e+00 1.15690410e+00 -1.87664092e-01
1.32519627e+00 -5.66214286e-02 7.63364732e-01 2.43584931e-01
5.85466206e-01 -7.08541751e-01 7.86729902e-02 2.62133092e-01
8.05746019e-01 -1.44710648e+00 -1.30028069e-01 -4.07984912e-01
-4.71880883e-01 1.10836399e+00 7.94623971e-01 1.93364173e-01
3.03638995e-01 3.04913789e-01 1.01150319e-01 -5.75002655e-02
-2.47394726e-01 -4.90468144e-01 8.53670463e-02 8.85963023e-01
2.74772853e-01 -5.04073426e-02 2.42030963e-01 8.78809333e-01
-4.44656044e-01 1.80232778e-01 2.80430973e-01 7.47320533e-01
-2.24543959e-01 -9.81964290e-01 -7.67558813e-01 -3.28964025e-01
-2.66952306e-01 4.31892008e-01 -3.64133835e-01 6.81444526e-01
3.24531108e-01 1.04258168e+00 1.63668826e-01 -4.44729239e-01
5.36392391e-01 -3.32471505e-02 6.73478007e-01 -4.56108719e-01
-5.04271448e-01 -1.10649332e-01 -2.29752749e-01 -1.00878489e+00
-4.42251652e-01 -5.03445208e-01 -1.12309241e+00 -4.93000805e-01
-1.29889384e-01 -3.86749476e-01 1.09048510e+00 1.08315635e+00
2.18745336e-01 3.45483184e-01 7.60014474e-01 -1.59263527e+00
-4.17069942e-01 -1.28714335e+00 -4.66138959e-01 5.21346390e-01
3.56416106e-01 -7.20461369e-01 -2.24815533e-01 8.41573551e-02] | [13.654485702514648, 0.2764734923839569] |
17cc45b4-0040-4d59-9538-15f2c5e9b697 | discourse-segmentation-for-building-a-rst | null | null | https://aclanthology.org/W17-3610 | https://aclanthology.org/W17-3610.pdf | Discourse Segmentation for Building a RST Chinese Treebank | null | ['Mikel Iruskieta', 'Shuyuan Cao', 'Iria da Cunha', 'Chuan Wang', 'Nianwen Xue'] | 2017-09-01 | null | null | null | ws-2017-9 | ['discourse-segmentation'] | ['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] | [-7.210562229156494, 3.7606616020202637] |
e3bf34a2-b180-4e9c-9dbe-4e9a75c613ad | multimodal-dialogue-state-tracking-1 | 2206.07898 | null | https://arxiv.org/abs/2206.07898v1 | https://arxiv.org/pdf/2206.07898v1.pdf | Multimodal Dialogue State Tracking | Designed for tracking user goals in dialogues, a dialogue state tracker is an essential component in a dialogue system. However, the research of dialogue state tracking has largely been limited to unimodality, in which slots and slot values are limited by knowledge domains (e.g. restaurant domain with slots of restaurant name and price range) and are defined by specific database schema. In this paper, we propose to extend the definition of dialogue state tracking to multimodality. Specifically, we introduce a novel dialogue state tracking task to track the information of visual objects that are mentioned in video-grounded dialogues. Each new dialogue utterance may introduce a new video segment, new visual objects, or new object attributes, and a state tracker is required to update these information slots accordingly. We created a new synthetic benchmark and designed a novel baseline, Video-Dialogue Transformer Network (VDTN), for this task. VDTN combines both object-level features and segment-level features and learns contextual dependencies between videos and dialogues to generate multimodal dialogue states. We optimized VDTN for a state generation task as well as a self-supervised video understanding task which recovers video segment or object representations. Finally, we trained VDTN to use the decoded states in a response prediction task. Together with comprehensive ablation and qualitative analysis, we discovered interesting insights towards building more capable multimodal dialogue systems. | ['Steven C. H. Hoi', 'Nancy F. Chen', 'Hung Le'] | 2022-06-16 | multimodal-dialogue-state-tracking | https://aclanthology.org/2022.naacl-main.248 | https://aclanthology.org/2022.naacl-main.248.pdf | naacl-2022-7 | ['dialogue-state-tracking'] | ['natural-language-processing'] | [ 2.99599230e-01 2.10337773e-01 -2.62209922e-01 -4.64337826e-01
-6.25842273e-01 -8.37556779e-01 1.01991892e+00 -3.78583781e-02
-3.88325810e-01 7.68374979e-01 6.15799367e-01 4.68600392e-02
4.64786142e-01 -5.56327045e-01 -5.70179760e-01 -5.49533010e-01
1.22505881e-01 5.19914567e-01 4.06748831e-01 -6.42222285e-01
2.19933257e-01 1.54935911e-01 -1.34702122e+00 7.70689487e-01
5.92014015e-01 8.12414765e-01 3.33302885e-01 9.36264336e-01
-3.95797849e-01 9.01564538e-01 -1.01351094e+00 -2.31871486e-01
-1.48042098e-01 -8.83775771e-01 -1.13727319e+00 5.50837874e-01
5.83422296e-02 -6.45669401e-01 -8.40611815e-01 8.92773211e-01
4.06274915e-01 5.34215510e-01 5.69803476e-01 -1.49239802e+00
-3.60192478e-01 4.95204866e-01 -1.45504907e-01 1.64891481e-02
1.17577159e+00 6.57020152e-01 6.97529197e-01 -7.00069427e-01
1.04431510e+00 1.50243831e+00 1.61379859e-01 1.13318002e+00
-1.05592501e+00 -1.93377689e-01 4.42634255e-01 2.79024780e-01
-8.24747145e-01 -8.40095222e-01 7.74757206e-01 -2.28284582e-01
1.00785804e+00 4.08633530e-01 8.40871274e-01 1.70018792e+00
-2.86754429e-01 1.41516817e+00 8.91574621e-01 -4.13475186e-01
1.66722417e-01 2.95474440e-01 -1.81870386e-01 7.74733722e-01
-7.13212550e-01 -2.57029772e-01 -8.55673254e-01 1.39222026e-01
7.08559155e-01 -2.53138870e-01 -4.51239139e-01 -5.45136988e-01
-1.71953964e+00 7.21716225e-01 1.51362926e-01 1.45472437e-01
-1.91511735e-02 -2.10088268e-01 7.15060949e-01 4.40921396e-01
1.19255424e-01 3.62710208e-01 -3.60260189e-01 -4.89047348e-01
-5.29663563e-01 1.83625966e-01 1.13843274e+00 1.00382113e+00
6.61919475e-01 9.39882733e-03 -8.60808253e-01 8.90203714e-01
1.90844342e-01 3.29351515e-01 4.44808513e-01 -1.24266720e+00
5.60084522e-01 9.31495786e-01 2.25467682e-01 -7.14063227e-01
-6.13626361e-01 4.64528739e-01 -7.33268261e-01 -3.36664051e-01
5.86126626e-01 -2.97375113e-01 -7.48099506e-01 1.90342307e+00
5.82923353e-01 -7.11781159e-02 4.22042251e-01 9.97408986e-01
1.28937292e+00 1.02962673e+00 1.41314128e-02 -4.93225306e-01
1.40955925e+00 -1.06425214e+00 -1.04817235e+00 1.04161918e-01
5.94679475e-01 -4.75756437e-01 1.24687696e+00 2.14442849e-01
-1.13618028e+00 -6.12685204e-01 -5.63854635e-01 4.54740971e-03
-4.02765363e-01 1.24008097e-01 4.90325361e-01 4.07300234e-01
-9.71788883e-01 1.84245169e-01 -6.88774347e-01 -8.12484026e-01
-1.72898144e-01 2.03853816e-01 -3.72116983e-01 6.98407516e-02
-1.58406961e+00 8.36672008e-01 5.29021263e-01 2.24706084e-01
-1.08606827e+00 -3.61844525e-02 -1.32370651e+00 2.04226491e-03
6.81607366e-01 -6.78172827e-01 1.55183399e+00 -9.47968304e-01
-2.02275491e+00 7.57086158e-01 -1.29551619e-01 -2.46893778e-01
3.15277934e-01 2.41645902e-01 -5.36735535e-01 5.16292512e-01
-1.09264113e-01 1.01924419e+00 9.17686224e-01 -1.29533291e+00
-8.73400450e-01 -3.23756546e-01 6.33548319e-01 7.41254985e-01
-3.58141154e-01 -5.56610823e-02 -8.88087690e-01 -3.64717454e-01
-5.95577173e-02 -1.00806236e+00 1.24848904e-02 -2.23518774e-01
-6.30839050e-01 -3.46857965e-01 9.61728096e-01 -6.51170790e-01
1.16588473e+00 -1.96182728e+00 5.45325696e-01 -2.10120335e-01
6.76513538e-02 2.69293487e-01 -3.35414827e-01 5.70400655e-01
2.18753174e-01 -3.11913341e-01 -7.28400797e-02 -3.21225822e-01
2.56836534e-01 1.64903015e-01 -9.87600163e-02 2.38384947e-01
2.38077760e-01 8.85523677e-01 -1.06948459e+00 -6.41891003e-01
3.69678736e-01 1.75939083e-01 -6.24110937e-01 5.44984341e-01
-6.24681592e-01 7.76594877e-01 -3.02649826e-01 8.14107060e-01
2.77111828e-01 -1.96975768e-01 3.31267148e-01 -6.75665379e-01
-1.69412732e-01 2.31381521e-01 -1.04303861e+00 2.36099887e+00
-4.14581686e-01 7.60539651e-01 2.16451898e-01 -9.69882250e-01
7.17089355e-01 5.00632286e-01 5.61498642e-01 -6.88841760e-01
1.26869371e-02 -4.47797120e-01 -2.98835725e-01 -9.46338534e-01
9.18320894e-01 2.74638474e-01 -6.36787117e-01 4.09107894e-01
4.17690128e-01 -1.65053114e-01 4.81676251e-01 5.24746299e-01
1.10162532e+00 3.29294473e-01 2.10985169e-01 4.83768255e-01
6.50924385e-01 2.96560049e-01 3.26810181e-01 7.26917624e-01
-4.82198805e-01 4.39045548e-01 8.33314896e-01 -1.59486175e-01
-8.07069421e-01 -8.57332468e-01 9.76958796e-02 1.64907110e+00
4.88485217e-01 -3.87692302e-01 -7.05458343e-01 -1.01406538e+00
-3.74238789e-01 6.07155740e-01 -5.19084096e-01 -3.61482441e-01
-4.88551259e-01 -4.68814373e-01 4.30741012e-01 2.25814655e-01
7.20165670e-01 -1.46539402e+00 -4.10010606e-01 1.32178813e-01
-7.31216133e-01 -1.27751255e+00 -7.38388002e-01 -8.67324024e-02
-4.12587345e-01 -1.11339986e+00 -8.82309139e-01 -8.82655799e-01
5.57108402e-01 4.11189012e-02 1.03057170e+00 -2.91166812e-01
-1.63960412e-01 1.02650607e+00 -5.91337740e-01 3.15524250e-01
-8.90194893e-01 1.07283577e-01 6.31327322e-03 3.15495551e-01
5.64104617e-02 1.59999758e-01 -6.70133471e-01 5.13974607e-01
-8.81506145e-01 4.07447517e-01 2.91986525e-01 9.97172832e-01
3.02003957e-02 -4.20257360e-01 5.32359958e-01 -7.02054203e-01
8.04095566e-01 -3.36064339e-01 -2.69797683e-01 5.95435679e-01
9.55706090e-02 7.85899758e-02 3.63369107e-01 -5.49747288e-01
-1.48922956e+00 5.80030143e-01 -9.54907238e-02 -1.70971557e-01
-4.29599702e-01 4.04003531e-01 -3.22284997e-01 2.64301151e-01
5.70579529e-01 6.26433372e-01 1.84472725e-01 -7.94730484e-02
5.35964668e-01 7.49233484e-01 7.12567270e-01 -6.32894516e-01
3.35665554e-01 2.40996122e-01 -4.39123213e-01 -1.11703777e+00
-4.40097779e-01 -5.92733383e-01 -7.16699362e-01 -6.13894820e-01
1.02071536e+00 -9.24493134e-01 -1.10405242e+00 5.69875956e-01
-1.26810455e+00 -5.66719949e-01 -1.61066249e-01 1.83428258e-01
-8.73850346e-01 7.04754293e-01 -5.84624827e-01 -8.51964116e-01
2.78502125e-02 -1.20678651e+00 1.10123563e+00 3.67213845e-01
-1.18428916e-01 -9.39884603e-01 2.10658118e-01 4.60239023e-01
1.71126276e-01 7.60138854e-02 5.31185627e-01 -7.71914244e-01
-4.92453396e-01 -4.75882441e-02 -1.47985980e-01 4.63833362e-02
1.54657722e-01 8.86704996e-02 -7.90290058e-01 -3.10919851e-01
-2.02086627e-01 -8.00747097e-01 7.38907635e-01 2.02145040e-01
7.95527995e-01 -5.94622672e-01 -2.87875593e-01 2.85729855e-01
5.81432223e-01 5.01795650e-01 5.66675901e-01 9.72673148e-02
6.27106905e-01 8.92681301e-01 9.13974583e-01 7.59486794e-01
7.39118516e-01 1.09865737e+00 3.74880046e-01 -9.19082090e-02
-6.60614967e-02 -2.76960820e-01 8.30789030e-01 7.13666677e-01
8.86462852e-02 -4.69342262e-01 -5.60264885e-01 4.04120505e-01
-2.10799813e+00 -1.05868626e+00 3.47429037e-01 1.88859487e+00
1.14877796e+00 -1.23280406e-01 4.32480186e-01 -3.12531799e-01
9.12853658e-01 3.14265817e-01 -7.85402298e-01 -1.40627712e-01
-9.91384909e-02 -5.28889120e-01 -6.41726702e-02 3.25512409e-01
-1.23793018e+00 1.22234070e+00 5.55077887e+00 4.96262193e-01
-8.34915638e-01 -9.22836661e-02 3.41098011e-01 -3.91635261e-02
-8.72610956e-02 -7.81617612e-02 -8.77246082e-01 3.20563674e-01
7.73129404e-01 2.45120730e-02 4.73649621e-01 4.63648379e-01
3.15392703e-01 -2.98663318e-01 -1.15600681e+00 9.48602080e-01
2.70634532e-01 -1.28547704e+00 2.00716004e-01 -3.38031888e-01
4.41773683e-01 -5.26267827e-01 -3.92162614e-03 7.72754490e-01
3.20640653e-01 -6.27223253e-01 3.16383898e-01 6.41620457e-01
8.84599805e-01 -4.03338581e-01 3.05655956e-01 3.16400290e-01
-1.16130793e+00 -7.99388662e-02 -2.96193063e-02 3.10152858e-01
3.88134986e-01 -4.62724388e-01 -1.26303959e+00 2.49928921e-01
4.91968215e-01 7.15211630e-01 -4.60515648e-01 7.89780736e-01
-6.39058463e-03 2.89170474e-01 -1.37235135e-01 -1.88147157e-01
2.37192601e-01 3.09581757e-02 7.09936559e-01 1.31999314e+00
1.33252740e-01 1.96778014e-01 2.88487941e-01 7.60398149e-01
-1.50876753e-02 1.24883465e-02 -5.66683769e-01 -1.47223517e-01
4.30638433e-01 1.41470921e+00 -6.63235009e-01 -5.09202659e-01
-5.54399908e-01 1.34934461e+00 -5.12348488e-02 4.70837831e-01
-8.55806768e-01 -2.40856677e-01 5.85600913e-01 -1.74004465e-01
1.38846263e-01 -2.08863262e-02 6.42199397e-01 -1.64604795e+00
-4.01925564e-01 -1.07998860e+00 5.16997159e-01 -8.11590672e-01
-8.46870244e-01 6.02581680e-01 1.35901541e-01 -1.25457752e+00
-7.07900405e-01 -4.24756855e-01 -3.77987415e-01 3.31921399e-01
-1.19025278e+00 -1.17775905e+00 -5.83134592e-01 1.15321267e+00
1.16984487e+00 -5.01751482e-01 9.74812984e-01 -2.27642469e-02
-9.12713826e-01 6.15676939e-01 -9.67721418e-02 3.88152659e-01
1.05448973e+00 -1.31697655e+00 2.33444273e-01 4.73787725e-01
8.39306712e-02 3.49415004e-01 7.47334361e-01 -6.17893994e-01
-1.57870591e+00 -8.38232040e-01 4.71757919e-01 -4.88199353e-01
5.13530076e-01 -4.53070641e-01 -8.65337551e-01 5.45342088e-01
5.02406240e-01 -3.28337401e-01 3.69263798e-01 -2.15817034e-01
1.98154271e-01 1.86929792e-01 -9.98241067e-01 7.13044643e-01
9.87310708e-01 -6.34683728e-01 -6.28886700e-01 5.06044686e-01
7.36714303e-01 -8.70983243e-01 -8.51943254e-01 2.48491988e-01
4.64082032e-01 -9.78255749e-01 9.97546852e-01 -8.99027407e-01
3.12680840e-01 -1.23055823e-01 -7.89426640e-02 -1.34363699e+00
8.98386538e-02 -9.46329236e-01 -4.77742821e-01 1.41225827e+00
2.11368605e-01 4.69283313e-02 8.31188798e-01 8.12124431e-01
-3.46250296e-01 -2.52232492e-01 -8.52663696e-01 -2.88955033e-01
-6.55795217e-01 5.46394624e-02 4.31666523e-01 8.46941888e-01
5.34925640e-01 6.23131573e-01 -6.35925114e-01 -1.19630851e-01
1.20449580e-01 1.29293099e-01 1.14333260e+00 -7.02571929e-01
-7.25650415e-02 -2.95829654e-01 -1.15106404e-01 -1.64971232e+00
3.87091875e-01 -4.95969117e-01 2.89822906e-01 -1.60466039e+00
1.96362928e-01 -3.56257297e-02 4.31695394e-02 3.86650801e-01
-1.00552708e-01 1.65684440e-04 3.50208342e-01 1.54654026e-01
-1.30374396e+00 9.72444534e-01 1.52130318e+00 -5.04729688e-01
-5.60624421e-01 3.34694028e-01 -1.18641511e-01 5.43123543e-01
4.12612915e-01 5.76212443e-02 -4.39130366e-01 4.15551253e-02
3.03815082e-02 9.65001464e-01 3.29508275e-01 -6.75355017e-01
5.04312992e-01 -2.31250256e-01 2.05883071e-01 -6.88307941e-01
8.72193336e-01 -6.59299135e-01 -3.40808511e-01 2.48167619e-01
-7.05829978e-01 -8.86378363e-02 3.74887250e-02 6.90597832e-01
-3.52827370e-01 -2.72232294e-01 4.42220062e-01 -3.62529248e-01
-1.43098807e+00 1.19189285e-01 -6.27755463e-01 1.29645001e-02
1.04881823e+00 -2.82429487e-01 -4.85066354e-01 -7.84519970e-01
-1.18789899e+00 4.97144938e-01 3.13924670e-01 6.94165111e-01
7.83980668e-01 -1.28762507e+00 -4.98824120e-01 5.69562912e-02
3.68057519e-01 -8.81558508e-02 5.10272801e-01 7.21365213e-01
-2.21361250e-01 4.15532947e-01 -3.74147713e-01 -8.55303466e-01
-1.29400516e+00 5.38930893e-01 2.32021064e-01 -4.52319868e-02
-4.15242791e-01 6.72588050e-01 3.54819417e-01 -6.72174394e-01
6.23337090e-01 -2.53102034e-01 -7.18472004e-01 5.01838207e-01
3.30233723e-01 1.77157879e-01 -3.44960153e-01 -8.33571196e-01
-1.66386031e-02 5.56262322e-02 -4.72448952e-03 -2.03078002e-01
8.89076173e-01 -7.06459045e-01 2.25891858e-01 7.19592631e-01
8.48501205e-01 -5.09493828e-01 -1.67107344e+00 -3.46877337e-01
-3.08324128e-01 -2.67523319e-01 -3.70019913e-01 -9.11285281e-01
-9.09627020e-01 5.85323393e-01 5.51076591e-01 4.16115284e-01
9.51516330e-01 1.36067659e-01 7.86049068e-01 7.14857042e-01
1.12508930e-01 -1.30260539e+00 7.42726743e-01 8.07582438e-01
1.05252302e+00 -1.59133482e+00 -3.44224632e-01 -2.61356169e-03
-1.34095156e+00 1.13551426e+00 1.04043782e+00 6.04850829e-01
1.02795117e-01 -2.01982662e-01 5.30753471e-02 -1.18749224e-01
-9.42594111e-01 -4.11551982e-01 1.87702879e-01 7.36593425e-01
1.90858305e-01 -3.28521162e-01 1.28840536e-01 5.16778946e-01
1.20544098e-01 -1.18704721e-01 6.06771290e-01 8.01484287e-01
-4.04446036e-01 -9.83662784e-01 -3.58863473e-01 2.69254714e-01
9.10321325e-02 1.20511100e-01 -4.76115823e-01 8.29423845e-01
-3.97875279e-01 9.83398318e-01 2.34145060e-01 -4.79434162e-01
5.32229006e-01 2.16351211e-01 5.84115684e-01 -8.06176245e-01
-5.85493028e-01 1.35952458e-01 3.88620555e-01 -6.38575256e-01
-7.09966183e-01 -7.76245892e-01 -1.20041132e+00 -7.99009874e-02
-2.99032152e-01 1.50681809e-01 6.05411172e-01 8.58822107e-01
1.90410167e-01 5.83466351e-01 8.20961714e-01 -1.03343856e+00
-4.50972915e-01 -1.12414193e+00 -3.19177538e-01 5.08677840e-01
3.33061844e-01 -6.77291751e-01 -8.26527476e-02 3.73051137e-01] | [10.90459156036377, 1.2040150165557861] |
091567e9-0d9f-4300-b88a-054facc27132 | routing-by-spontaneous-synchronization | 2305.13914 | null | https://arxiv.org/abs/2305.13914v1 | https://arxiv.org/pdf/2305.13914v1.pdf | Routing by spontaneous synchronization | Selective attention allows to process stimuli which are behaviorally relevant, while attenuating distracting information. However, it is an open question what mechanisms implement selective routing, and how they are engaged in dependence on behavioral need. Here we introduce a novel framework for selective processing by spontaneous synchronization. Input signals become organized into 'avalanches' of synchronized spikes which propagate to target populations. Selective attention enhances spontaneous synchronization and boosts signal transfer by a simple disinhibition of a control population, without requiring changes in synaptic weights. Our framework is fully analytically tractable and provides a complete understanding of all stages of the routing mechanism, yielding closed-form expressions for input-output correlations. Interestingly, although gamma oscillations can naturally occur through a recurrent dynamics, we can formally show that the routing mechanism itself does not require such oscillatory activity and works equally well if synchronous events would be randomly shuffled over time. Our framework explains a large range of physiological findings in a unified framework and makes specific predictions about putative control mechanisms and their effects on neural dynamics. | ['Udo Ernst', 'Maik Schünemann'] | 2023-05-23 | null | null | null | null | ['open-question'] | ['natural-language-processing'] | [ 6.33622408e-01 -2.40827620e-01 8.45546350e-02 1.16874970e-01
-4.74084318e-02 -8.72904956e-01 6.74715340e-01 2.61203557e-01
-6.75302505e-01 9.98837352e-01 4.64198798e-01 1.57417044e-01
-2.13024244e-01 -5.35591960e-01 -6.60827219e-01 -1.23143446e+00
-9.34868604e-02 -2.83300653e-02 5.90557039e-01 -3.58762056e-01
5.89275002e-01 4.74218607e-01 -1.62127328e+00 2.58667409e-01
6.43705666e-01 6.63498640e-01 3.82580072e-01 7.58434832e-01
2.54208326e-01 5.72427928e-01 -6.18923426e-01 1.01484567e-01
-6.13378212e-02 -9.04611409e-01 -5.19075215e-01 -2.40897849e-01
-2.63621509e-02 4.06765431e-01 -2.86767006e-01 8.97280812e-01
5.23029804e-01 3.72876748e-02 7.65930116e-01 -8.19208682e-01
-8.81665885e-01 5.08582652e-01 -2.49451607e-01 1.14957047e+00
3.58865224e-02 2.77063787e-01 7.56523967e-01 -6.44120157e-01
6.45773351e-01 8.94173324e-01 5.71451604e-01 7.22195208e-01
-1.90440142e+00 -5.39671361e-01 3.17491502e-01 -2.79468566e-01
-1.15943336e+00 -6.12610579e-01 4.93337691e-01 -3.35330099e-01
1.21467948e+00 3.04857999e-01 1.16034341e+00 1.06453979e+00
9.33172941e-01 3.46861422e-01 1.27934766e+00 -2.36982480e-01
4.48818922e-01 -1.15013227e-01 4.37939167e-01 6.52244017e-02
7.06536174e-01 -1.04872258e-02 -9.26092207e-01 -1.21211775e-01
1.19063413e+00 -3.30641344e-02 -6.28006995e-01 -6.67504817e-02
-1.34593570e+00 5.00970244e-01 2.63353258e-01 6.32444680e-01
-5.22035599e-01 4.75115061e-01 1.18620180e-01 5.29872417e-01
1.78282008e-01 7.67146707e-01 -3.97162437e-01 5.57670370e-02
-1.00745523e+00 1.24517865e-01 4.74166095e-01 5.21881044e-01
6.96184278e-01 2.04540953e-01 -2.82923758e-01 5.56956172e-01
-4.72693630e-02 5.67072570e-01 8.15726876e-01 -1.11151445e+00
-1.88118428e-01 1.42429218e-01 1.28434896e-01 -9.11740661e-01
-6.00301325e-01 -5.94544888e-01 -7.01517880e-01 2.01765463e-01
6.49228573e-01 -2.78884828e-01 -4.36601877e-01 2.40015888e+00
-4.35795486e-01 1.16937552e-02 -2.45802611e-01 7.18635798e-01
-6.96129818e-03 5.38832784e-01 2.49714419e-01 -8.15260589e-01
1.06522262e+00 -2.97366202e-01 -8.55981827e-01 -4.26097304e-01
7.36887679e-02 -4.26129907e-01 8.03947747e-01 1.36645734e-02
-1.53621542e+00 -1.01308830e-01 -7.75850296e-01 1.81474477e-01
-2.90231496e-01 -3.89910609e-01 7.35378087e-01 1.97329953e-01
-1.51903689e+00 7.68694401e-01 -8.26429665e-01 -7.80902684e-01
3.16558480e-01 6.56019926e-01 -7.31129348e-02 7.92231023e-01
-1.02152479e+00 9.25654888e-01 -9.19560492e-02 -1.08222188e-02
-7.59187341e-01 -5.65393567e-01 -1.63029164e-01 2.59183705e-01
-3.68894100e-01 -1.12745094e+00 1.06472051e+00 -1.22808528e+00
-1.36855102e+00 9.77163315e-01 -5.30910850e-01 -4.78784680e-01
-2.32293069e-01 2.50016540e-01 -8.72215554e-02 3.04799289e-01
1.84424929e-02 5.64427137e-01 7.19216406e-01 -9.48709309e-01
-8.58485177e-02 -4.41874206e-01 -2.00180605e-01 2.61816442e-01
-1.12969078e-01 2.73640245e-01 2.48633370e-01 -9.40860868e-01
1.31726697e-01 -8.65610003e-01 -2.68382877e-01 7.92385172e-03
9.39939469e-02 2.63470560e-01 1.19030833e-01 2.06227943e-01
1.14915287e+00 -2.08931375e+00 3.64464283e-01 1.17476573e-02
3.26390684e-01 -1.18529625e-01 -1.91338211e-01 6.34626687e-01
-2.68348753e-01 2.13896960e-01 -4.23048824e-01 2.40642384e-01
-3.31677735e-01 -2.35658474e-02 -4.17091519e-01 5.27325809e-01
5.15938640e-01 1.01264811e+00 -8.61995637e-01 3.94647457e-02
-3.09690893e-01 5.66811323e-01 -6.71999454e-01 -1.38035223e-01
2.10370421e-01 6.03573501e-01 -2.15181053e-01 2.87801653e-01
3.39186281e-01 -4.01938796e-01 1.13989100e-01 1.44849136e-01
-7.47627795e-01 6.09712422e-01 -7.87100315e-01 1.15166152e+00
-1.38397664e-01 9.89240348e-01 2.15961531e-01 -1.19103813e+00
6.08598113e-01 3.13859582e-01 1.24595121e-01 -7.56617367e-01
3.59387308e-01 3.14407140e-01 5.72714567e-01 -1.26837775e-01
5.21264151e-02 -3.19974512e-01 -8.80856737e-02 7.67631888e-01
2.13265732e-01 4.51983847e-02 2.58035243e-01 1.47544011e-01
1.37062705e+00 -4.17088181e-01 3.33742231e-01 -9.81061935e-01
6.31312206e-02 -1.26184136e-01 3.90611410e-01 1.02558148e+00
-4.60123301e-01 5.82405686e-01 6.59173906e-01 -5.29321693e-02
-9.13326979e-01 -1.24966574e+00 -1.45466492e-01 1.26472700e+00
3.88881356e-01 -2.59527490e-02 -7.27041602e-01 3.75694782e-01
-2.77141005e-01 4.97658104e-01 -1.02610636e+00 -7.00031340e-01
-5.56359172e-01 -1.24058938e+00 3.84587198e-01 2.99698919e-01
-5.93961105e-02 -1.58168077e+00 -1.12216592e+00 4.50317323e-01
3.54040489e-02 -6.38798773e-01 -6.33997917e-01 8.67427766e-01
-1.10898948e+00 -7.05932319e-01 -7.74237156e-01 -7.01925457e-01
7.82492757e-01 5.57283103e-01 9.44941163e-01 -7.28752986e-02
-4.06491607e-01 2.22729415e-01 2.73246795e-01 -3.83534402e-01
7.76209757e-02 8.73146430e-02 2.78766125e-01 -1.75302267e-01
3.44170868e-01 -1.22828066e+00 -8.26405883e-01 1.21279053e-01
-8.29804480e-01 -3.08964372e-01 4.72002059e-01 5.37541568e-01
6.78111613e-01 -4.05359209e-01 8.97796035e-01 -5.84242821e-01
1.00843835e+00 -4.25838202e-01 -3.11327726e-01 -2.16243371e-01
-3.11019272e-01 3.63062471e-01 4.63171393e-01 -6.44169927e-01
-1.02196336e+00 -1.92587733e-01 3.03439736e-01 1.26457572e-01
-2.56207474e-02 2.03939006e-01 2.70830333e-01 -2.73952961e-01
8.11534941e-01 6.92112148e-01 1.56552456e-02 6.77889735e-02
5.60217835e-02 -8.81768689e-02 3.84657025e-01 -3.58752906e-01
3.89470011e-01 7.20274389e-01 -1.94650516e-01 -8.21332335e-01
-6.39475524e-01 -1.30367890e-01 -5.56146443e-01 -1.27793431e-01
6.25732780e-01 -6.70379221e-01 -7.71266460e-01 5.48865795e-01
-1.30877149e+00 -5.63069522e-01 -4.49865252e-01 5.20308495e-01
-8.78007591e-01 -3.76150101e-01 -9.32474971e-01 -8.91231358e-01
-2.13699013e-01 -7.22750664e-01 6.63746834e-01 4.13659185e-01
-5.74208319e-01 -9.19711053e-01 3.79404038e-01 -6.40220702e-01
8.97602499e-01 -2.84719467e-01 8.64756167e-01 -4.54012245e-01
-5.47315001e-01 2.03101113e-01 -1.50025383e-01 -2.83905536e-01
5.68115860e-02 2.30310678e-01 -9.10135865e-01 9.23189297e-02
2.94455856e-01 -2.98876017e-02 1.14937460e+00 9.17947888e-01
7.61130273e-01 -3.04854929e-01 -5.60322404e-01 3.60610992e-01
1.35683167e+00 1.29853010e-01 5.42712927e-01 1.46863097e-02
-1.11929718e-02 6.22084320e-01 -2.50723928e-01 1.13729663e-01
-1.41806841e-01 2.81622171e-01 2.32272327e-01 2.09167957e-01
-2.21065450e-02 4.91399504e-02 4.05710310e-01 1.03894734e+00
-4.40135241e-01 -7.65651986e-02 -5.09145737e-01 7.61243045e-01
-1.65145791e+00 -1.55899644e+00 -2.49622643e-01 2.47250581e+00
1.11692178e+00 3.60157460e-01 1.59538947e-02 -9.06840190e-02
9.01463211e-01 1.13254832e-02 -6.70535803e-01 -5.33183813e-01
-6.42529070e-01 3.92821610e-01 6.14944100e-01 5.80143511e-01
-4.71185178e-01 8.64131987e-01 8.72920418e+00 2.69004285e-01
-1.29708481e+00 2.35599548e-01 4.00098324e-01 -5.06976724e-01
-5.23506105e-01 -8.07952434e-02 -8.22673023e-01 5.85545063e-01
1.25451589e+00 -6.57725513e-01 6.14555597e-01 -5.04925214e-02
6.76128030e-01 -2.93362886e-01 -7.67170072e-01 5.72370410e-01
-2.03501612e-01 -1.47955215e+00 -3.09775807e-02 2.06722412e-02
5.87609351e-01 2.20986605e-01 9.42170992e-02 -6.58869594e-02
2.47237638e-01 -7.97538221e-01 6.24507487e-01 7.78807580e-01
3.56878102e-01 -4.36399102e-01 1.69220418e-01 4.54316467e-01
-8.03705633e-01 -3.68902981e-02 -4.92156982e-01 -6.03627741e-01
1.73650235e-01 6.63634300e-01 2.83232301e-01 -5.43923557e-01
4.98434782e-01 4.30725992e-01 -3.85135442e-01 1.15632558e+00
1.65043753e-02 5.57428360e-01 -4.41728115e-01 -2.73955047e-01
-1.68369040e-01 -2.78082173e-02 7.76013732e-01 1.28670084e+00
3.26474905e-01 3.48707736e-01 -8.33526194e-01 1.11650538e+00
1.61856279e-01 -1.22966565e-01 -7.65724421e-01 -1.26574248e-01
6.64240122e-01 1.12909901e+00 -1.19420052e+00 -4.72827852e-02
-1.03733875e-01 1.10893297e+00 4.35107827e-01 6.51316524e-01
-6.32951140e-01 -4.31546420e-01 7.09725022e-01 2.86000818e-01
2.40397349e-01 -3.32350910e-01 -3.91424894e-01 -1.27048445e+00
-2.37280115e-01 -3.20167691e-01 -1.41696751e-01 -8.08156133e-01
-1.17253625e+00 6.33700907e-01 -3.20472121e-01 -6.82597041e-01
8.21321458e-02 -2.98241973e-01 -7.91110158e-01 8.73367369e-01
-1.17724693e+00 -2.81649649e-01 5.81433661e-02 6.67213321e-01
3.97183985e-01 3.96741450e-01 7.59833336e-01 2.76406854e-02
-5.43599665e-01 1.05487317e-01 7.50460923e-02 -4.87128466e-01
5.68343759e-01 -9.69105244e-01 1.46049783e-01 7.61749148e-01
9.88338888e-02 1.15989447e+00 1.01943552e+00 -4.51897919e-01
-1.02828145e+00 -8.02850366e-01 1.16101539e+00 -3.94208848e-01
9.43989277e-01 -5.50861299e-01 -9.75549936e-01 4.77455139e-01
4.87243235e-01 6.18742732e-03 8.06203246e-01 -1.24737822e-01
-1.89443141e-01 2.13019401e-01 -7.18651474e-01 8.00478578e-01
1.39215875e+00 -5.10899127e-01 -6.55883491e-01 7.11369291e-02
5.76618314e-01 2.99549878e-01 -2.85852909e-01 -3.57367322e-02
6.27695024e-01 -1.10618293e+00 6.33944750e-01 -5.41719913e-01
3.44401486e-02 -2.39779457e-01 1.62767954e-02 -1.36730576e+00
-7.29940534e-01 -7.44615495e-01 1.58637419e-01 8.98067415e-01
6.41531587e-01 -9.65859711e-01 1.80111259e-01 3.52036536e-01
-9.00381524e-03 -4.45091099e-01 -8.26452017e-01 -6.44870698e-01
3.02076012e-01 -6.84879571e-02 -2.14125216e-01 6.76750422e-01
5.56644320e-01 4.25619781e-01 1.99278429e-01 -8.74520987e-02
4.20838058e-01 -1.45083785e-01 -2.34423518e-01 -1.16842580e+00
-1.78392395e-01 -9.19871032e-01 -2.53009975e-01 -9.82205868e-01
-9.68828332e-03 -7.54136324e-01 4.97314110e-02 -1.17588341e+00
3.51824045e-01 6.04203679e-02 -7.23972678e-01 2.45694131e-01
-1.08296365e-01 5.24832964e-01 -3.26815955e-02 6.30202651e-01
-5.23082614e-01 4.64469612e-01 1.04511619e+00 4.25288588e-01
-4.12624836e-01 -9.76561978e-02 -1.07626927e+00 7.41017878e-01
9.49979603e-01 -6.46403015e-01 -3.83039832e-01 -4.34399068e-01
5.78915298e-01 -1.48800656e-01 3.91557634e-01 -9.81769443e-01
4.98051018e-01 -2.71082461e-01 1.92336336e-01 -3.63894142e-02
1.79684907e-01 -3.12889308e-01 -3.91254798e-02 5.62582135e-01
-7.14466870e-01 2.41831303e-01 3.73119026e-01 6.52371526e-01
6.08961880e-02 -7.60976374e-02 1.13970518e+00 -1.99365258e-01
-3.31693999e-02 6.33025616e-02 -1.62169969e+00 2.12460890e-01
9.06849682e-01 -3.64802033e-01 -6.04790747e-01 -3.68183434e-01
-1.07976973e+00 -2.75861353e-01 5.59347332e-01 -1.66144967e-02
2.39513993e-01 -1.27899826e+00 -3.34524363e-01 2.53279597e-01
-2.90360272e-01 -7.28109360e-01 1.43129215e-01 1.18505740e+00
-1.52642623e-01 6.60039067e-01 -5.84033132e-01 -3.67886215e-01
-7.00385094e-01 5.74775934e-01 4.38234568e-01 1.47622809e-01
-1.12714581e-01 1.10614121e+00 6.39055133e-01 3.17886919e-01
-2.53837407e-01 -3.79228920e-01 -2.66873658e-01 2.82105237e-01
6.96675837e-01 -7.00069740e-02 -1.88904554e-01 -4.47345912e-01
-4.42314506e-01 6.93948746e-01 7.48905689e-02 -3.04155141e-01
1.13238573e+00 -5.48650086e-01 -4.67796445e-01 9.64123785e-01
7.74733365e-01 9.61028710e-02 -1.31074595e+00 1.39422476e-01
-2.75962502e-01 1.00590743e-01 -2.33173221e-01 -4.85517859e-01
-9.91567194e-01 1.02978158e+00 9.83527303e-02 6.45575941e-01
1.16233718e+00 2.02388465e-01 2.52293438e-01 1.55359492e-01
4.85445589e-01 -8.46608818e-01 2.64315128e-01 4.90203053e-01
9.80434954e-01 -5.24367154e-01 -5.26480675e-01 -3.98285419e-01
-3.71787429e-01 8.88960302e-01 4.81483132e-01 -7.99658000e-01
7.04403460e-01 5.98763347e-01 -3.25988144e-01 -2.33404949e-01
-1.47262466e+00 -3.12672466e-01 -2.57824630e-01 8.07165146e-01
7.79738367e-01 -2.54574716e-01 -7.40819037e-01 7.27821887e-01
-3.40921469e-02 7.82029256e-02 7.87014663e-01 7.54721463e-01
-8.56091082e-01 -6.56358540e-01 -1.31531462e-01 4.28286016e-01
-6.95782423e-01 -4.78406519e-01 -5.03086686e-01 5.67995608e-01
-1.20110013e-01 6.37490451e-01 4.72660333e-01 8.79202038e-02
-2.36650463e-02 6.59990385e-02 8.12968552e-01 -8.41204166e-01
-5.90418458e-01 3.53830069e-01 -4.04779464e-01 -3.92943591e-01
-6.12970054e-01 -1.08324873e+00 -1.43163157e+00 -1.41353935e-01
-1.58193201e-01 1.12492479e-02 4.37832028e-02 8.86840463e-01
6.13237798e-01 7.09836185e-01 2.32157439e-01 -8.83038700e-01
-6.14267848e-02 -5.80831409e-01 -6.96863592e-01 1.01579286e-01
5.12401819e-01 -5.76415360e-01 -7.11698353e-01 2.50158817e-01] | [8.156854629516602, 3.08337664604187] |
88824a5a-3e47-4454-98d0-ce4c00cbca0b | end-to-end-audiovisual-speech-recognition | 1802.06424 | null | http://arxiv.org/abs/1802.06424v2 | http://arxiv.org/pdf/1802.06424v2.pdf | End-to-end Audiovisual Speech Recognition | Several end-to-end deep learning approaches have been recently presented
which extract either audio or visual features from the input images or audio
signals and perform speech recognition. However, research on end-to-end
audiovisual models is very limited. In this work, we present an end-to-end
audiovisual model based on residual networks and Bidirectional Gated Recurrent
Units (BGRUs). To the best of our knowledge, this is the first audiovisual
fusion model which simultaneously learns to extract features directly from the
image pixels and audio waveforms and performs within-context word recognition
on a large publicly available dataset (LRW). The model consists of two streams,
one for each modality, which extract features directly from mouth regions and
raw waveforms. The temporal dynamics in each stream/modality are modeled by a
2-layer BGRU and the fusion of multiple streams/modalities takes place via
another 2-layer BGRU. A slight improvement in the classification rate over an
end-to-end audio-only and MFCC-based model is reported in clean audio
conditions and low levels of noise. In presence of high levels of noise, the
end-to-end audiovisual model significantly outperforms both audio-only models. | ['Pingchuan Ma', 'Maja Pantic', 'Stavros Petridis', 'Georgios Tzimiropoulos', 'Feipeng Cai', 'Themos Stafylakis'] | 2018-02-18 | null | null | null | null | ['lipreading'] | ['computer-vision'] | [ 2.32866943e-01 -2.15534985e-01 1.13207251e-01 -3.02956134e-01
-1.57743037e+00 -3.70112479e-01 7.14513361e-01 -2.84025278e-02
-4.21176493e-01 4.25359249e-01 4.31961864e-01 -1.74275503e-01
3.45103562e-01 -1.47064880e-01 -8.05948496e-01 -7.30005980e-01
-1.17559835e-01 -3.54947627e-01 7.92792216e-02 1.27393141e-01
-5.93733899e-02 1.47113994e-01 -1.93033957e+00 8.74907553e-01
1.59273297e-01 1.56464100e+00 2.26738065e-01 1.53885293e+00
3.45809720e-02 1.05099022e+00 -5.58353364e-01 6.58567026e-02
-2.08718497e-02 -5.53768396e-01 -3.00866306e-01 4.57469700e-03
6.34635627e-01 -3.71702880e-01 -6.30031526e-01 4.57245976e-01
1.09678400e+00 1.94192976e-01 5.09608328e-01 -9.82045174e-01
-5.66538811e-01 4.60436702e-01 -2.65666813e-01 1.85546100e-01
6.67429328e-01 3.44922483e-01 1.07856333e+00 -1.33818138e+00
5.23421228e-01 1.34997153e+00 5.00826180e-01 3.39395165e-01
-1.25312376e+00 -7.05580413e-01 2.69304991e-01 5.03386319e-01
-1.26387954e+00 -9.72089589e-01 9.37093914e-01 -2.85080969e-01
1.43781877e+00 1.97661072e-01 5.58886707e-01 1.51048172e+00
1.39447704e-01 1.04387140e+00 9.67372298e-01 -4.58821446e-01
1.85085505e-01 -2.73060706e-03 -1.58892214e-01 1.02999002e-01
-7.30189979e-01 4.09977198e-01 -1.19145465e+00 5.75979240e-02
3.21134299e-01 -2.98360050e-01 -3.25815260e-01 2.77700603e-01
-8.67049158e-01 6.80261850e-01 2.57667661e-01 3.20009738e-01
-5.69020152e-01 4.74883676e-01 6.90976560e-01 6.19503915e-01
6.05293989e-01 -3.67752463e-01 -2.65753925e-01 -2.89586157e-01
-1.37432837e+00 -1.10532276e-01 5.33095300e-01 4.91593152e-01
2.83257127e-01 6.97103679e-01 -3.75088274e-01 1.08196282e+00
6.10275149e-01 6.77918077e-01 7.10021257e-01 -5.50713360e-01
4.48642939e-01 -2.93675661e-01 -2.43333220e-01 -6.20149195e-01
-1.58433333e-01 -3.97328347e-01 -4.46298510e-01 4.11605746e-01
-1.14988863e-01 -1.14259869e-01 -1.25146961e+00 1.61471117e+00
1.13537103e-01 6.92610145e-01 2.32043058e-01 8.91691506e-01
1.39470482e+00 1.22112560e+00 1.90392047e-01 -2.85131276e-01
1.22265708e+00 -7.96303272e-01 -1.01293147e+00 7.16327429e-02
-9.19194520e-03 -1.10954654e+00 7.92098761e-01 7.12044597e-01
-1.39961088e+00 -9.77281928e-01 -1.02723789e+00 -3.22851181e-01
-3.75324219e-01 3.47991765e-01 3.68009768e-02 4.78177994e-01
-1.33430505e+00 1.77521631e-01 -7.22110212e-01 -1.26925319e-01
3.95000428e-02 3.04175526e-01 -3.66367638e-01 4.62736711e-02
-1.24891686e+00 4.67405677e-01 -1.11630641e-01 1.49523661e-01
-1.47066808e+00 -5.31082034e-01 -1.02006817e+00 -9.13560539e-02
1.24935195e-01 -3.08063000e-01 1.43921149e+00 -1.19358897e+00
-1.84382272e+00 5.73582888e-01 -5.98157763e-01 -7.63582170e-01
3.52042764e-01 -3.37288648e-01 -5.88559866e-01 4.01443839e-01
-3.01253706e-01 7.56545126e-01 1.37840271e+00 -1.25236261e+00
-6.89062953e-01 -1.51660770e-01 -5.51330626e-01 1.64061442e-01
-4.49299663e-02 2.67027795e-01 -6.17484629e-01 -8.37101161e-01
-3.69047552e-01 -5.84705770e-01 4.20965105e-01 -4.82506789e-02
-1.53984681e-01 -1.73497632e-01 9.74965930e-01 -1.16109967e+00
1.11506987e+00 -2.62164164e+00 1.03169821e-01 -1.00163296e-01
-2.60531485e-01 3.12299341e-01 -5.18274486e-01 5.11751652e-01
-2.82807767e-01 -1.67848855e-01 5.23852855e-02 -9.77497220e-01
8.21889192e-03 -9.58152115e-02 -6.74758315e-01 4.51565832e-01
3.76692206e-01 6.80161953e-01 -4.98256385e-01 -2.40132347e-01
3.72885406e-01 1.33578205e+00 -3.96323293e-01 4.58595276e-01
4.38626334e-02 2.66421765e-01 2.29786173e-01 6.97116017e-01
4.59041148e-01 4.41228598e-01 -2.51029044e-01 -1.06587499e-01
-7.93033615e-02 5.67920268e-01 -1.02074075e+00 1.75637245e+00
-7.76542246e-01 1.20880949e+00 4.47646111e-01 -6.39805079e-01
8.83601725e-01 1.06959236e+00 2.43382663e-01 -1.02490485e+00
2.02266917e-01 3.05168003e-01 -3.82726818e-01 -4.48168784e-01
4.31467772e-01 -1.63640290e-01 1.26202866e-01 1.13873124e-01
6.53301179e-01 2.67891455e-02 -2.05838159e-01 4.32074033e-02
8.43297482e-01 1.15391530e-01 -1.00922041e-01 6.12464666e-01
5.80028117e-01 -7.24356353e-01 2.61436105e-01 5.56211352e-01
-8.04660171e-02 1.07473195e+00 1.62599996e-01 9.58296210e-02
-8.01315725e-01 -1.36096621e+00 1.76567689e-01 1.47619331e+00
-3.07656080e-01 -4.35035318e-01 -4.47245240e-01 -6.63283318e-02
-7.61047602e-02 6.86122835e-01 -4.87200439e-01 -1.42192394e-01
-2.96017200e-01 -7.52496123e-02 9.42680538e-01 6.17954433e-01
2.07499579e-01 -1.27476287e+00 -2.69136012e-01 3.34538013e-01
-3.02098542e-01 -1.26744008e+00 -5.40669203e-01 2.74759084e-01
-5.19373178e-01 -4.79006916e-01 -9.42334473e-01 -7.79957056e-01
-4.13007028e-02 5.68762645e-02 6.82593524e-01 -4.90189672e-01
-4.95691508e-01 7.29044378e-01 -4.37213987e-01 -4.37993556e-01
-2.11943060e-01 -3.95837069e-01 -5.75651899e-02 6.33014262e-01
1.06306903e-01 -6.07431531e-01 -5.62527537e-01 -2.87160166e-02
-9.85700727e-01 -3.76439959e-01 4.14617181e-01 8.08001578e-01
6.69600964e-01 -2.00448126e-01 6.57126427e-01 2.94346106e-03
6.46559298e-01 -3.91576350e-01 -3.85062248e-01 -2.01203674e-01
-4.11958173e-02 -3.17557842e-01 5.12598515e-01 -7.85787344e-01
-1.16796064e+00 2.09274471e-01 -4.62835819e-01 -8.92714858e-01
-3.01708400e-01 4.90903854e-01 -8.24665427e-02 3.36814731e-01
3.77257824e-01 4.11622196e-01 -1.37154222e-01 -6.63766265e-01
5.28088510e-01 1.20797634e+00 9.57257867e-01 -1.15647316e-01
3.51140231e-01 3.81968826e-01 -5.25195539e-01 -1.19434214e+00
-2.16240004e-01 -6.54242098e-01 -2.51227617e-01 -4.77938861e-01
9.36636746e-01 -1.47002482e+00 -5.88402569e-01 5.86459637e-01
-1.12571692e+00 -4.32771206e-01 -2.41150334e-01 8.07442307e-01
-6.46556616e-01 1.97133869e-01 -8.40957344e-01 -1.35528684e+00
-4.88619000e-01 -1.20638061e+00 1.34145820e+00 2.77948957e-02
-1.52851164e-01 -6.56436503e-01 -1.02079296e-02 1.91985235e-01
5.29651463e-01 -5.26306108e-02 4.41478759e-01 -5.10823190e-01
-3.21301758e-01 -3.56755078e-01 1.23626411e-01 6.83320761e-01
-9.03397202e-02 1.62940219e-01 -1.84376836e+00 -3.95885646e-01
7.59304827e-03 -5.11826634e-01 1.37521708e+00 6.82736278e-01
7.97993124e-01 -1.88327104e-01 1.46315709e-01 4.86497521e-01
1.12789190e+00 4.45895225e-01 7.36961424e-01 -2.96519369e-01
3.82727891e-01 4.79550421e-01 3.32506150e-01 4.10675973e-01
9.78774130e-02 7.19971716e-01 4.88083005e-01 -3.25915098e-01
-8.12556922e-01 -3.37599516e-01 1.14314747e+00 9.36596274e-01
2.31014669e-01 -5.12033343e-01 -5.46079695e-01 8.48603666e-01
-1.47470748e+00 -9.19116676e-01 2.67531902e-01 2.25308466e+00
7.65707910e-01 3.82585824e-02 3.36295068e-01 4.20331389e-01
5.61109781e-01 3.11977834e-01 -4.87958878e-01 -5.70589244e-01
-2.42667064e-01 4.65916127e-01 -1.08314224e-03 7.25799799e-01
-1.04787183e+00 7.06661403e-01 6.23950195e+00 9.77153003e-01
-1.62487090e+00 1.57917202e-01 4.91095632e-01 -7.39075303e-01
-3.03346999e-02 -2.57717252e-01 -6.40740931e-01 2.69919097e-01
1.50526345e+00 4.52596754e-01 4.58067775e-01 4.74214762e-01
4.54206228e-01 1.01816456e-03 -1.08558810e+00 1.33260202e+00
3.51619661e-01 -8.63152146e-01 -7.47169778e-02 2.60873865e-02
1.90402836e-01 3.95814925e-01 5.54587305e-01 3.23027641e-01
-1.06519915e-01 -1.17779386e+00 1.08216131e+00 4.91237730e-01
1.19459438e+00 -8.59341085e-01 5.21004856e-01 -3.00972704e-02
-1.55986631e+00 -1.83650285e-01 1.12232476e-01 2.38002047e-01
3.02310467e-01 2.30360240e-01 -7.86716163e-01 3.29873264e-01
8.61330807e-01 6.87155187e-01 -2.94827163e-01 9.73504245e-01
-2.03718051e-01 1.12737167e+00 -3.97781819e-01 2.56873041e-01
2.22680449e-01 4.15436357e-01 6.40462637e-01 1.88133919e+00
4.43225861e-01 -1.91505387e-01 -5.25851883e-02 4.95927006e-01
-6.77300468e-02 8.39168504e-02 -6.20928645e-01 -2.63408840e-01
2.68748224e-01 1.02489948e+00 -2.72296786e-01 -2.14248747e-01
-4.52330619e-01 9.82958317e-01 -2.66477644e-01 7.68788457e-01
-8.35356295e-01 -6.31798387e-01 7.14851499e-01 -1.72624707e-01
7.59251595e-01 -1.24671444e-01 4.54378426e-02 -8.99218082e-01
8.14382955e-02 -8.41506958e-01 3.52564126e-01 -1.02269661e+00
-1.11586607e+00 7.43921995e-01 -4.39188451e-01 -1.24304545e+00
-6.04739010e-01 -5.48183382e-01 -5.41432381e-01 9.09071267e-01
-1.66780162e+00 -1.10198605e+00 -5.54650277e-02 9.20547605e-01
8.28627706e-01 -2.98973888e-01 7.55645394e-01 4.42467421e-01
-1.88159555e-01 8.40402186e-01 1.66265488e-01 2.32658252e-01
9.24515605e-01 -8.23161662e-01 2.29656935e-01 7.27174103e-01
3.94332439e-01 2.63200313e-01 7.15901732e-01 -4.14346039e-01
-1.52517104e+00 -9.91621196e-01 9.48279083e-01 3.20988148e-02
4.27399188e-01 -7.48454809e-01 -9.79282200e-01 4.89320874e-01
6.90009832e-01 1.94395587e-01 7.61751473e-01 -2.18508109e-01
-6.23549163e-01 -2.65630484e-01 -7.94628859e-01 2.93225765e-01
4.73886222e-01 -1.29745221e+00 -4.67142820e-01 -1.27683923e-01
9.06653285e-01 -1.83766261e-01 -6.80018187e-01 2.18245238e-01
7.09711790e-01 -7.59733498e-01 1.04817820e+00 -3.76555771e-01
2.05750525e-01 -3.91727298e-01 -6.54636443e-01 -1.29131877e+00
2.47318700e-01 -9.49361622e-01 -1.93281561e-01 1.53225267e+00
4.43768561e-01 -4.11930352e-01 1.27767831e-01 -1.25530183e-01
-7.92333335e-02 -3.57650489e-01 -1.28398824e+00 -7.09439754e-01
-2.72800207e-01 -1.13127017e+00 -4.66528721e-02 3.67868572e-01
-3.28563787e-02 4.88518775e-01 -7.55044818e-01 6.85827956e-02
3.40497375e-01 -1.23719402e-01 4.81299102e-01 -8.10037732e-01
-4.80142176e-01 -2.90268302e-01 -4.73761261e-01 -9.28283453e-01
6.44342676e-02 -7.65800118e-01 3.44092935e-01 -1.50398910e+00
-3.46152514e-01 4.30873185e-01 -5.19780099e-01 4.67538118e-01
2.32177198e-01 4.11984891e-01 4.93545383e-01 -1.20658435e-01
-4.14052039e-01 8.19513798e-01 5.13807952e-01 -3.56229931e-01
-4.27509606e-01 -1.72196403e-01 -2.68617183e-01 3.26581061e-01
3.98241013e-01 -3.62219393e-01 -3.76415014e-01 -2.36079082e-01
-8.92797858e-02 4.10232276e-01 4.04147863e-01 -8.86597037e-01
3.57567221e-01 5.24724483e-01 5.85735917e-01 -7.89740741e-01
1.10278749e+00 -6.06554627e-01 5.42766228e-02 8.70373026e-02
-6.49646223e-01 -2.62141109e-01 5.20442486e-01 7.92972028e-01
-8.15316617e-01 3.99301440e-01 7.32344747e-01 1.89810976e-01
-4.17857766e-01 -8.70127827e-02 -9.56117749e-01 -3.47421438e-01
6.53790474e-01 -1.31165430e-01 3.01849190e-02 -9.94358897e-01
-1.29553580e+00 -7.36058876e-02 -2.84535080e-01 5.78660905e-01
1.02383983e+00 -1.45872307e+00 -1.02648127e+00 4.30684328e-01
-1.68675691e-01 -4.65050220e-01 4.34881389e-01 8.46943080e-01
9.21514183e-02 5.71173251e-01 5.01353852e-02 -8.00076544e-01
-1.70275831e+00 4.97714311e-01 2.21908003e-01 1.02205560e-01
-4.52830881e-01 9.98072505e-01 1.74531281e-01 5.60598932e-02
1.00398207e+00 -3.46403033e-01 -2.89684504e-01 5.66969931e-01
8.59617412e-01 3.02365154e-01 2.23101631e-01 -1.01088750e+00
-2.85840839e-01 6.24200106e-01 8.12966097e-03 -8.04353058e-01
1.47171068e+00 -2.04237729e-01 4.79863018e-01 9.42233324e-01
1.53996348e+00 -8.62297416e-02 -1.38510883e+00 -3.97818804e-01
-5.31672478e-01 -1.74470440e-01 4.64220315e-01 -9.38476503e-01
-1.21863341e+00 1.55770278e+00 1.05819011e+00 -1.08332103e-02
1.59722412e+00 -1.04956642e-01 8.40268016e-01 6.41421322e-03
-1.41744286e-01 -1.14366162e+00 3.30666929e-01 5.12512207e-01
1.15760601e+00 -8.49080324e-01 -5.94202399e-01 3.23378816e-02
-8.31660509e-01 1.06536674e+00 4.40839417e-02 1.26230959e-02
7.04096854e-01 4.69093621e-01 3.26552838e-01 2.48340786e-01
-1.22233558e+00 -5.20424187e-01 5.11392832e-01 5.94445825e-01
5.65416276e-01 -2.04827353e-01 2.54094392e-01 5.20939767e-01
1.75550333e-04 7.42281089e-03 1.05223037e-01 8.18265975e-01
-2.53755629e-01 -7.34262288e-01 -7.01255918e-01 -1.18267745e-01
-7.17426717e-01 -2.66673476e-01 -6.79749250e-01 3.96631867e-01
-1.70917839e-01 1.46710408e+00 -4.76216385e-03 -5.73210180e-01
3.71125370e-01 6.05959058e-01 2.40311533e-01 -9.84069258e-02
-9.12928224e-01 9.53561723e-01 3.99844982e-02 -6.84311390e-01
-1.74576610e-01 -5.98757744e-01 -1.34728694e+00 2.71226197e-01
-2.15594023e-01 -1.91569000e-01 1.03958964e+00 6.40946209e-01
3.93269390e-01 8.69077325e-01 6.84062600e-01 -1.29868984e+00
-2.06521913e-01 -1.16169620e+00 -6.33308232e-01 2.64634937e-01
1.12765801e+00 -3.69404584e-01 -5.87682545e-01 3.28777075e-01] | [14.378901481628418, 5.157271385192871] |
aec68f01-80b7-4cbe-ace3-d3ff04114cd9 | challenging-deep-image-descriptors-for | 1909.08866 | null | https://arxiv.org/abs/1909.08866v1 | https://arxiv.org/pdf/1909.08866v1.pdf | Challenging deep image descriptors for retrieval in heterogeneous iconographic collections | This article proposes to study the behavior of recent and efficient state-of-the-art deep-learning based image descriptors for content-based image retrieval, facing a panel of complex variations appearing in heterogeneous image datasets, in particular in cultural collections that may involve multi-source, multi-date and multi-view Permission to make digital | ['Valérie Gouet-Brunet', 'Dimitri Gominski', 'Liming Chen', 'Martyna Poreba'] | 2019-09-19 | null | null | null | null | ['content-based-image-retrieval'] | ['computer-vision'] | [-1.92648560e-01 -9.26164269e-01 -3.50602329e-01 -3.44893456e-01
-7.80228019e-01 -6.52230382e-01 8.51134479e-01 5.83259225e-01
-8.26964974e-01 2.18211398e-01 2.21885473e-01 2.00124383e-01
-6.90820694e-01 -6.22725606e-01 -2.14951798e-01 -6.21946156e-01
-2.65485257e-01 3.40035617e-01 -1.12982169e-01 -6.16487265e-01
7.27918088e-01 9.18611884e-01 -1.96455669e+00 4.61399764e-01
3.98083217e-02 1.30444264e+00 1.53985336e-01 6.89902484e-01
-2.40049973e-01 4.46756810e-01 -8.31304193e-01 -6.49377346e-01
4.34735358e-01 3.36947113e-01 -7.00265765e-01 2.26695035e-02
1.22650254e+00 -6.61554098e-01 -8.15651357e-01 1.09190476e+00
8.76856446e-01 -3.87566909e-02 9.50793087e-01 -1.09623206e+00
-1.25747097e+00 -7.15224147e-02 -5.43773651e-01 7.10414112e-01
6.27646446e-01 -5.90278469e-02 7.31823862e-01 -9.79656577e-01
1.11101973e+00 1.32396495e+00 5.60753644e-01 -1.33823559e-01
-6.30103350e-01 -3.28058302e-01 -2.25021869e-01 5.67632258e-01
-1.69091332e+00 -3.15873027e-01 6.10347271e-01 -4.48723108e-01
8.42476845e-01 1.74168840e-01 6.37163639e-01 1.21740746e+00
4.19728994e-01 7.93823957e-01 8.61442685e-01 -5.23393393e-01
-1.79674312e-01 -2.46090814e-02 1.55101523e-01 5.80800533e-01
4.66518342e-01 -1.49484202e-01 -7.25583792e-01 -4.01036829e-01
5.68615198e-01 3.22186738e-01 -7.07973354e-03 -5.15517235e-01
-8.57001662e-01 1.02689314e+00 2.81770468e-01 8.81560266e-01
-7.20263064e-01 -1.26345277e-01 7.16060340e-01 7.35707998e-01
4.77194399e-01 3.31260741e-01 -3.96962613e-01 -1.85875461e-01
-1.19577444e+00 5.58511257e-01 6.64606810e-01 1.12680864e+00
7.10419655e-01 -1.55676961e-01 5.01313654e-04 1.02410352e+00
1.37459457e-01 5.78910828e-01 7.00841248e-01 -6.50845587e-01
2.56282091e-01 4.02378142e-01 -2.55401939e-01 -1.49088264e+00
-2.77052432e-01 -1.15554303e-01 -9.82265949e-01 -2.67134577e-01
-5.81863150e-02 5.33011854e-01 -8.83600593e-01 8.36455405e-01
4.33349609e-03 -5.22017121e-01 -3.53650361e-01 9.35247242e-01
1.12228131e+00 4.61199939e-01 4.88932915e-02 1.37326851e-01
1.30889761e+00 -7.53385425e-01 -3.99292707e-01 3.51338953e-01
3.04054648e-01 -1.04814351e+00 5.56552827e-01 4.17974979e-01
-1.20355988e+00 -8.80700946e-01 -1.11995363e+00 -3.03579032e-01
-1.24054313e+00 -2.93231811e-02 4.79088277e-01 4.80323106e-01
-1.42047250e+00 3.99742424e-01 5.44115640e-02 -8.27351987e-01
4.26385552e-01 1.61207095e-01 -8.46504450e-01 -6.33780658e-01
-1.09489357e+00 8.99423540e-01 3.42953354e-01 -1.95828333e-01
-9.13172424e-01 -6.96431339e-01 -7.22552359e-01 8.30542445e-02
1.19670726e-01 -4.07084197e-01 7.65581429e-01 -1.28351903e+00
-1.00708282e+00 1.59489787e+00 8.82731155e-02 -8.09028745e-02
5.87134063e-01 -3.08161169e-01 -7.22462535e-01 5.45658946e-01
3.46197002e-02 7.04355597e-01 1.27723753e+00 -1.21015465e+00
-1.58230126e-01 -4.38966900e-01 9.73158553e-02 1.01766750e-01
-5.97382724e-01 3.36385876e-01 -7.63495862e-01 -5.74884057e-01
-3.51926565e-01 -8.41975927e-01 3.23029011e-01 3.22640449e-01
6.65318370e-02 -3.94182354e-01 1.08195603e+00 -4.11019742e-01
8.34233761e-01 -2.24930429e+00 2.88848251e-01 1.29659340e-01
9.76943523e-02 5.77546179e-01 -5.73466718e-01 1.08725047e+00
5.66705763e-02 2.94479787e-01 1.98999956e-01 -2.06024006e-01
-3.83443609e-02 1.47291735e-01 -4.08553928e-02 6.38173163e-01
-2.34983280e-01 6.94952548e-01 -5.88684976e-01 -8.96090925e-01
4.15968060e-01 5.83768666e-01 1.07530482e-01 1.76870435e-01
4.10774171e-01 -2.24665329e-01 -6.06406927e-01 1.16197395e+00
1.01704860e+00 -2.20765829e-01 -3.03097162e-02 -3.35144222e-01
-1.48510575e-01 -6.73819065e-01 -7.79019415e-01 1.96209109e+00
-2.04580665e-01 1.19267285e+00 -5.15282666e-03 -1.26857734e+00
2.67315686e-01 9.68440846e-02 9.73431289e-01 -1.20241010e+00
3.92708153e-01 1.56170815e-01 -4.79862899e-01 -5.16972840e-01
1.21466875e+00 5.86978137e-01 -7.36498535e-02 1.33032709e-01
4.54845309e-01 1.43397287e-01 4.07122254e-01 3.02993685e-01
5.62317908e-01 -2.75787413e-01 1.70595869e-01 -5.51566601e-01
7.47295916e-01 -2.54193932e-01 -2.73252100e-01 1.15341425e+00
-4.43832219e-01 7.86120236e-01 -5.85831236e-04 -9.84273791e-01
-1.40399861e+00 -9.01962280e-01 -5.55630267e-01 1.31460381e+00
1.48782283e-01 3.70952189e-02 -5.13555288e-01 -1.32639214e-01
2.88118631e-01 -2.92673737e-01 -5.75546443e-01 8.02086592e-02
-3.78966898e-01 -3.39296699e-01 4.82501924e-01 -1.02685373e-02
7.97793448e-01 -1.12226713e+00 -4.24616098e-01 -6.55773282e-03
7.77030438e-02 -1.26745045e+00 -3.98551345e-01 -2.91130513e-01
-5.52341700e-01 -1.38396287e+00 -1.53850150e+00 -1.10954440e+00
3.24239403e-01 5.69598496e-01 1.69041336e+00 5.00991702e-01
-7.54637897e-01 1.21469724e+00 -5.13968825e-01 -4.26984251e-01
7.46086240e-02 3.88258100e-01 -2.38089278e-01 3.06952745e-02
6.24836326e-01 -2.07360983e-02 -9.43439007e-01 -5.45854047e-02
-1.52339756e+00 -1.02278209e+00 4.74019438e-01 6.86722040e-01
5.80994785e-01 -7.87298009e-02 -1.23573333e-01 -3.36990535e-01
9.47218716e-01 -7.23384857e-01 -4.71575797e-01 5.91270566e-01
-7.05477059e-01 -3.52070361e-01 2.55930126e-01 -4.87308264e-01
-5.73504567e-01 -3.43707770e-01 2.03528225e-01 -6.57714188e-01
-3.86951119e-01 6.37979507e-01 4.88366067e-01 -7.11082757e-01
4.13288355e-01 7.02250421e-01 -3.12436044e-01 -3.27319711e-01
4.60236609e-01 7.33541191e-01 3.33396405e-01 -4.38042402e-01
6.30968571e-01 6.71240389e-01 1.22983538e-01 -1.31128013e+00
-1.21869266e-01 -9.53289688e-01 -8.42489839e-01 -3.04736018e-01
8.80110681e-01 -1.26273727e+00 -5.45352995e-01 1.06578755e+00
-9.54904437e-01 3.21177125e-01 2.84315467e-01 1.69729963e-01
-3.33329976e-01 4.95165944e-01 -6.50966167e-01 -1.59915760e-01
-6.69215560e-01 -1.02287364e+00 1.40695822e+00 3.42640817e-01
2.53618479e-01 -1.08696020e+00 1.43733859e-01 3.39235008e-01
8.89028490e-01 5.29731624e-02 7.72715747e-01 -6.12463951e-01
-7.24682987e-01 -6.68605506e-01 -3.67248356e-01 2.31562585e-01
-1.13088481e-01 3.70478988e-01 -7.13731408e-01 -6.52387381e-01
-3.54272217e-01 -6.78325415e-01 8.91218305e-01 4.84488666e-01
1.25376892e+00 2.01017827e-01 -2.39910096e-01 5.35292208e-01
2.03576756e+00 9.96410847e-03 8.33993435e-01 9.06898677e-01
4.49436069e-01 2.22983733e-01 1.14904292e-01 5.43090165e-01
4.12737519e-01 8.06084514e-01 3.16437155e-01 1.11733899e-01
-8.49611089e-02 2.00275302e-01 -2.75537133e-01 8.47415149e-01
4.99369316e-02 -5.79727411e-01 -7.93209434e-01 7.28274763e-01
-1.37816918e+00 -9.99894321e-01 1.11229353e-01 1.73571026e+00
3.45303029e-01 -3.61360520e-01 -3.98958102e-03 -2.14145437e-01
5.47880530e-01 7.43196428e-01 -3.24487001e-01 -4.77111369e-01
-5.91707945e-01 1.67097598e-01 8.04050803e-01 -2.89672226e-01
-1.37169576e+00 7.53555417e-01 7.75971174e+00 1.10593140e+00
-1.43211246e+00 3.31676632e-01 4.50251073e-01 -2.30986550e-02
-1.89684913e-01 -5.52548409e-01 -3.30550015e-01 4.85773087e-01
7.74026275e-01 -1.01955220e-01 2.22928926e-01 9.85815823e-01
-5.78735471e-01 -3.14619511e-01 -6.55245066e-01 1.60664046e+00
5.69740415e-01 -1.61281824e+00 4.61931944e-01 1.46105647e-01
9.62427258e-01 5.71886122e-01 4.59282756e-01 1.75276220e-01
-3.72905642e-01 -7.66361892e-01 6.65327132e-01 4.69527662e-01
7.94598997e-01 -5.39193809e-01 9.42743421e-01 -4.10677224e-01
-1.04962754e+00 -6.06964342e-02 -4.81788874e-01 5.04768014e-01
-3.82944375e-01 2.50803679e-01 3.92336607e-01 7.66236603e-01
9.55614865e-01 7.25197911e-01 -9.41763163e-01 1.36680555e+00
6.09762669e-01 -2.57175833e-01 -1.93502456e-02 -1.38262268e-02
6.02701187e-01 1.94917068e-01 3.39829385e-01 1.23432338e+00
3.35980862e-01 -2.38694608e-01 6.67091832e-02 1.48576349e-01
-3.27684253e-01 5.50296009e-01 -9.80799258e-01 -1.06987685e-01
2.67655790e-01 1.27395618e+00 -6.67300045e-01 -5.46593487e-01
-6.14416897e-01 1.23593032e+00 1.15517437e-01 5.34897864e-01
-3.02846909e-01 -1.97488159e-01 5.54590583e-01 1.36091709e-01
5.27519822e-01 -5.27910233e-01 4.46315169e-01 -1.28659654e+00
-3.86239626e-02 -9.39997256e-01 5.57324529e-01 -1.16820443e+00
-1.61491191e+00 6.37516379e-01 3.53697807e-01 -1.13553822e+00
-1.62031010e-01 -8.24927151e-01 -1.52044773e-01 6.67715430e-01
-1.72122931e+00 -1.37324858e+00 -3.32862347e-01 1.25648284e+00
6.80316150e-01 -9.64732647e-01 9.30948019e-01 7.99704373e-01
1.19604170e-01 7.21337616e-01 8.10018539e-01 4.37017620e-01
7.56495655e-01 -5.87296307e-01 9.97633338e-02 2.78130293e-01
4.99777973e-01 5.02645433e-01 2.08029211e-01 -1.77268554e-02
-1.80601537e+00 -3.28010321e-01 4.22351658e-01 -1.63564578e-01
5.55605412e-01 7.33219311e-02 -7.32668161e-01 1.96328297e-01
7.09018707e-01 1.31948143e-01 5.34699023e-01 -7.67159462e-02
-7.65607953e-01 -2.51246572e-01 -1.16785741e+00 3.17965329e-01
5.41274965e-01 -1.16464674e+00 -2.02909738e-01 3.53851616e-01
-1.33200616e-01 -3.01644474e-01 -1.16315699e+00 1.57975912e-01
7.75272131e-01 -8.94796908e-01 1.34877753e+00 -6.87441826e-01
4.07248378e-01 3.48660558e-01 -4.62325752e-01 -1.01066279e+00
-7.38859713e-01 -2.55799621e-01 9.06659290e-02 6.84614539e-01
-2.36958385e-01 -3.61951500e-01 6.17508292e-01 2.91439235e-01
1.42522335e-01 -3.97131741e-01 -1.20039606e+00 -7.51296580e-01
2.74601042e-01 7.48493522e-02 4.52609032e-01 8.05826664e-01
-5.12768626e-01 -2.87288278e-01 -2.67436266e-01 -2.29990825e-01
6.82297230e-01 3.24777812e-02 5.36029816e-01 -1.37021101e+00
5.35097539e-01 -8.30845892e-01 -1.14141083e+00 -4.57352281e-01
1.86037660e-01 -3.03498596e-01 -5.41764200e-01 -1.31234741e+00
3.19302380e-01 -6.18078783e-02 -6.45893097e-01 -9.26973969e-02
1.51118591e-01 3.12402785e-01 5.83596826e-01 8.38981867e-01
-1.01117992e+00 3.86443526e-01 1.05160999e+00 -6.91699207e-01
5.38481355e-01 -7.52090037e-01 -2.52337575e-01 3.63802224e-01
4.57340956e-01 -5.43045282e-01 -2.55730927e-01 -6.43421531e-01
3.24634284e-01 -1.48004785e-01 2.36595765e-01 -1.12479758e+00
1.52169898e-01 3.09849530e-02 6.05624437e-01 -5.92863381e-01
4.19783860e-01 -9.63363647e-01 -4.48272079e-02 3.67337495e-01
-3.91968906e-01 5.39708614e-01 2.27906048e-01 6.76899910e-01
-8.89791131e-01 -3.79054725e-01 5.54146707e-01 -7.01345801e-01
-1.38413012e+00 6.97898388e-01 -5.63564777e-01 1.02746852e-01
8.23971987e-01 -3.87433201e-01 -4.35330570e-01 -7.13214397e-01
-3.56960058e-01 -3.01024497e-01 4.06512856e-01 1.13779831e+00
7.01371133e-01 -1.36955535e+00 -7.93631077e-01 3.77157480e-02
5.85656106e-01 -7.81595111e-01 6.24346435e-01 2.66935945e-01
-9.47466671e-01 8.74397933e-01 -6.69803739e-01 -6.31265283e-01
-1.31106353e+00 6.48047864e-01 4.14834708e-01 -1.64285898e-01
-2.84455478e-01 7.87082672e-01 -3.10272604e-01 1.16217948e-01
1.83677033e-01 3.25237274e-01 -3.05032045e-01 6.57599747e-01
4.96301919e-01 2.28413716e-01 1.80674076e-01 -1.15377855e+00
-4.59677607e-01 9.45754349e-01 -3.83005261e-01 1.72127843e-01
1.04141617e+00 -2.49964550e-01 -2.35325500e-01 2.94760585e-01
2.17493629e+00 -6.54637277e-01 -3.44124287e-01 -2.59379357e-01
-3.64821494e-01 -6.02118909e-01 3.89284581e-01 -5.19756317e-01
-1.34230423e+00 6.16282940e-01 1.46038663e+00 4.00353640e-01
1.16807830e+00 -2.19653115e-01 9.19479132e-01 6.83927655e-01
7.26660490e-01 -1.49083221e+00 1.95423216e-01 3.31058800e-01
1.06636000e+00 -1.63854456e+00 7.44635165e-01 2.67254293e-01
-3.11063826e-01 1.42145514e+00 1.80503651e-01 -1.93268448e-01
1.30672216e+00 -4.26503718e-01 1.71017364e-01 -6.50165677e-01
-2.89281279e-01 -1.32647648e-01 5.45646548e-01 7.46921539e-01
2.12145478e-01 -1.68880165e-01 -4.92030859e-01 -6.40029967e-01
4.11213607e-01 7.80138075e-02 3.65826666e-01 1.14874542e+00
4.50310707e-02 -1.17425108e+00 -3.07991982e-01 2.30007768e-01
-9.21260297e-01 -4.51431960e-01 -3.89834046e-01 9.49798048e-01
1.09079495e-01 8.00304055e-01 3.06365013e-01 -1.42601892e-01
1.41254425e-01 -1.74965799e-01 5.67888737e-01 9.72023234e-02
-8.61209512e-01 -2.41507590e-01 -3.10029566e-01 -6.69979632e-01
-9.81563568e-01 -5.08281529e-01 -1.63158029e-01 -5.14300525e-01
2.70402908e-01 -3.33889425e-01 1.24687338e+00 6.22852504e-01
7.51199245e-01 -1.11687325e-01 5.20821452e-01 -1.07117438e+00
-2.79794067e-01 -9.22002614e-01 -6.81334019e-01 1.08607960e+00
4.40753818e-01 -6.70718253e-01 -1.33117229e-01 3.85204493e-03] | [10.701857566833496, 0.4902758002281189] |
e3ddf6e6-2729-4975-a167-e68668f9bb6e | net2vec-deep-learning-for-the-network | 1705.03881 | null | http://arxiv.org/abs/1705.03881v1 | http://arxiv.org/pdf/1705.03881v1.pdf | Net2Vec: Deep Learning for the Network | We present Net2Vec, a flexible high-performance platform that allows the
execution of deep learning algorithms in the communication network. Net2Vec is
able to capture data from the network at more than 60Gbps, transform it into
meaningful tuples and apply predictions over the tuples in real time. This
platform can be used for different purposes ranging from traffic classification
to network performance analysis.
Finally, we showcase the use of Net2Vec by implementing and testing a
solution able to profile network users at line rate using traces coming from a
real network. We show that the use of deep learning for this case outperforms
the baseline method both in terms of accuracy and performance. | ['Alberto Garcia-Duran', 'Saverio Niccolini', 'Felipe Huici', 'Jose Mendes', 'Filipe Manco', 'Roberto Gonzalez', 'Mathias Niepert'] | 2017-05-10 | null | null | null | null | ['traffic-classification'] | ['miscellaneous'] | [-6.31816804e-01 -1.36769384e-01 -1.39809266e-01 -5.42769372e-01
-1.46836445e-01 -3.24091762e-01 5.63275933e-01 1.88178062e-01
-2.88774729e-01 7.28584528e-01 -1.27326250e-01 -8.91380727e-01
-2.07205206e-01 -1.31263638e+00 -5.08745193e-01 -2.18904838e-01
-5.93559504e-01 1.29735732e+00 5.22663176e-01 -4.51540917e-01
-2.82788903e-01 1.27725375e+00 -1.32321322e+00 7.26883531e-01
-3.08731437e-01 1.36046493e+00 -2.63210684e-01 9.53218818e-01
-4.45479125e-01 9.11974549e-01 -1.12297523e+00 -5.87299407e-01
5.39502323e-01 4.38882321e-01 -7.98921049e-01 -2.50945896e-01
9.38374083e-03 -4.88353878e-01 -8.09466958e-01 4.89508331e-01
6.93749309e-01 -1.81896925e-01 -2.00612936e-02 -1.54342139e+00
3.65490228e-01 1.02879131e+00 1.66184157e-01 5.04181802e-01
1.39094666e-01 7.64157623e-02 8.44339490e-01 -4.83894706e-01
6.20842099e-01 1.45482183e+00 5.01503944e-01 -5.63840233e-02
-1.29135036e+00 -8.51675749e-01 -2.78538704e-01 4.22621310e-01
-1.11304629e+00 -5.27247787e-01 3.04784298e-01 -3.35493743e-01
1.27368963e+00 6.68782115e-01 6.61652446e-01 1.37961888e+00
-2.03728136e-02 5.15372336e-01 4.53730464e-01 -2.13610366e-01
2.26876199e-01 6.05839670e-01 2.64161944e-01 4.65930283e-01
-1.08669192e-01 1.93974569e-01 -2.30356812e-01 -3.03881258e-01
7.49502003e-01 -1.06622763e-02 4.52363603e-02 1.35125900e-02
-1.01945162e+00 5.17163336e-01 4.88289982e-01 4.76280570e-01
-6.56692088e-01 5.82610250e-01 1.06877494e+00 8.53538454e-01
2.87747920e-01 4.25822139e-01 -6.09083712e-01 -7.24882722e-01
-8.83995891e-01 2.32565384e-02 1.39231169e+00 1.17920566e+00
7.54009962e-01 5.09391606e-01 -2.85274416e-01 4.33628112e-01
-1.07936770e-01 5.09438396e-01 3.49341959e-01 -8.61793101e-01
2.60742873e-01 4.15603101e-01 -3.22725832e-01 -1.08326614e+00
-5.75293839e-01 -4.77549285e-01 -7.26172626e-01 2.36671165e-01
7.21674711e-02 -6.42530143e-01 -3.62165153e-01 1.10548913e+00
-6.27263039e-02 6.32565022e-01 -9.89206731e-02 6.87778652e-01
4.34275240e-01 8.88846815e-01 -2.59821583e-02 -4.73786481e-02
1.01334429e+00 -8.24272573e-01 -3.82329613e-01 3.44366610e-01
6.15054727e-01 -6.45631969e-01 5.94848633e-01 6.15351140e-01
-1.08149564e+00 -7.83689797e-01 -9.88344133e-01 3.57525736e-01
-6.23093903e-01 -2.23490670e-01 8.74818563e-01 1.00449967e+00
-1.27328634e+00 8.30509841e-01 -8.57480943e-01 -6.50322616e-01
3.60148162e-01 8.41195583e-01 -3.28122169e-01 3.82450223e-01
-9.36871707e-01 5.41846037e-01 6.11759603e-01 -1.67240277e-01
-6.56366348e-01 -8.12070787e-01 -2.27595448e-01 6.96392655e-01
2.56075859e-01 -5.15149891e-01 1.16150427e+00 -7.78601587e-01
-1.49420619e+00 2.55109727e-01 2.10344240e-01 -9.98572767e-01
6.29371881e-01 -1.15967542e-01 -1.12249053e+00 -2.42732512e-03
-5.34311116e-01 -7.66866282e-02 2.37537086e-01 -8.38337302e-01
-6.07419193e-01 -6.30469397e-02 1.38063088e-01 -6.87094033e-01
-6.41344666e-01 1.38207823e-01 -8.63934875e-01 -1.11245707e-01
-6.72794342e-01 -7.32354581e-01 -3.89589339e-01 -4.10564363e-01
-6.85705066e-01 1.25336042e-02 1.26477039e+00 -1.55737594e-01
9.71484065e-01 -1.72641861e+00 -5.13715386e-01 9.54190850e-01
3.92983496e-01 9.68112469e-01 -2.21332565e-01 7.65729070e-01
-2.02864841e-01 2.84774214e-01 7.02631712e-01 -2.64661342e-01
9.30445492e-02 3.05151314e-01 -4.54309285e-01 1.16138153e-01
-1.84538767e-01 9.49062049e-01 -6.45744026e-01 -2.28099689e-01
6.24044538e-01 6.37306273e-01 -3.77536833e-01 4.76922393e-01
-4.18260515e-01 -4.46972065e-02 -2.23334983e-01 2.27219298e-01
7.52788782e-01 -2.18339324e-01 5.60622394e-01 -4.04114127e-01
3.70917059e-02 2.88205057e-01 -9.95594680e-01 1.15233374e+00
-1.01056099e+00 1.28333938e+00 -1.05340496e-01 -1.03904772e+00
1.21094477e+00 3.68313938e-01 7.83275545e-01 -8.97898495e-01
4.86240983e-01 1.48188978e-01 -2.44761527e-01 -4.68677759e-01
2.10863218e-01 3.79292130e-01 2.24516451e-01 7.33964562e-01
2.59536445e-01 4.96188849e-01 1.66713476e-01 -3.04594659e-03
1.40400982e+00 -8.92851949e-01 -1.07053757e-01 -1.70761608e-02
5.77709079e-01 -2.73281902e-01 6.50279671e-02 7.11340725e-01
1.78509086e-01 7.00411946e-03 1.00551474e+00 -9.79063809e-01
-1.12487936e+00 -1.20690596e+00 4.59914692e-02 1.15575159e+00
-3.37891608e-01 -4.46640462e-01 -3.45168352e-01 -8.10343325e-01
1.78750411e-01 7.32142985e-01 -2.79772550e-01 1.36943549e-01
-5.13102770e-01 -5.03321409e-01 7.14907348e-01 4.06025767e-01
2.67593831e-01 -1.09027469e+00 -2.06900567e-01 5.39604843e-01
4.58721340e-01 -1.41610575e+00 3.01693380e-01 7.06652105e-02
-4.82290238e-01 -9.58358586e-01 2.08700716e-01 -3.21048707e-01
-1.61225691e-01 -1.23887718e-01 1.66261101e+00 3.90155464e-01
-5.13041914e-01 2.11095467e-01 -3.24455202e-01 -3.17347974e-01
-4.86796319e-01 4.54753965e-01 1.04060777e-01 2.81838894e-01
6.76353693e-01 -1.10309935e+00 -8.37042108e-02 2.41942063e-01
-7.09746242e-01 -5.12761295e-01 5.00944912e-01 1.78143531e-01
3.30145299e-01 2.27565914e-01 3.97338837e-01 -1.55179811e+00
7.16860294e-01 -8.39168131e-01 -8.42406809e-01 1.75937675e-02
-6.18586838e-01 -3.78582031e-02 9.35445786e-01 -8.54980573e-03
-2.95469284e-01 -4.69241470e-01 -8.07978153e-01 -6.10719264e-01
-2.58997023e-01 2.14048162e-01 -2.56630123e-01 7.06313271e-03
5.99534452e-01 -5.16939275e-02 -3.51216421e-02 -3.80643725e-01
2.39232734e-01 9.71246958e-01 3.23449671e-01 -3.67215365e-01
6.60007477e-01 5.45001626e-01 3.37901980e-01 -1.21249616e+00
-2.90088326e-01 -4.19311017e-01 -4.85916317e-01 -1.93908259e-01
1.59361064e-01 -5.34182072e-01 -1.38605237e+00 -1.04427151e-01
-1.32177269e+00 -5.11738002e-01 -4.88601744e-01 3.51260543e-01
-4.50857520e-01 -1.09214127e-01 -7.86384046e-01 -3.97751242e-01
-5.33471107e-01 -1.09689283e+00 4.24478680e-01 -1.38399228e-01
1.05656721e-01 -1.38707125e+00 -6.20548837e-02 -1.68890908e-01
9.81033444e-01 2.31877089e-01 8.43063116e-01 -1.43849289e+00
-7.58774877e-01 -6.34757459e-01 -6.84333622e-01 5.16336024e-01
-3.26091558e-01 4.08622682e-01 -1.06802297e+00 -2.56071448e-01
-5.14356673e-01 -4.46231216e-02 4.46458340e-01 -1.41412511e-01
1.75771213e+00 -2.61536181e-01 -4.57659900e-01 1.17645419e+00
1.78607523e+00 -4.13789181e-03 8.81446362e-01 1.97642609e-01
5.74439764e-01 2.24079788e-01 1.35955006e-01 7.71040738e-01
-7.01041892e-02 9.39210176e-01 8.08038175e-01 -2.76071608e-01
-6.00753315e-02 6.20344505e-02 5.46538793e-02 4.36495960e-01
1.61447227e-01 -7.89449275e-01 -9.17701483e-01 1.90193370e-01
-1.78089976e+00 -1.19918799e+00 -4.63065743e-01 2.04822135e+00
-1.56359076e-01 6.57820225e-01 4.23100233e-01 1.93326548e-01
3.56108189e-01 -4.53705117e-02 -1.02469139e-01 -1.06237173e+00
2.70426750e-01 6.67675495e-01 8.34123313e-01 6.22126818e-01
-6.67132139e-01 8.73635650e-01 6.87715864e+00 7.47125745e-01
-1.57411325e+00 9.41042900e-02 5.05831957e-01 -9.66615304e-02
-2.33672693e-01 -2.22074792e-01 -7.80722260e-01 5.05379260e-01
2.08208680e+00 -3.18392992e-01 4.20732260e-01 9.79329169e-01
3.32918525e-01 6.32164717e-01 -1.28232729e+00 1.13222039e+00
-3.89687955e-01 -1.87666607e+00 2.69152403e-01 1.54242754e-01
-1.57198571e-02 5.34303188e-01 -5.19975200e-02 6.15436316e-01
3.42972785e-01 -1.26693535e+00 7.46364072e-02 6.74087524e-01
8.68393719e-01 -1.15307140e+00 1.25419009e+00 -7.42241293e-02
-7.89039552e-01 -1.39295176e-01 -3.99567962e-01 7.81401843e-02
2.97566980e-01 9.93329048e-01 -1.57033479e+00 4.51578915e-01
3.69334668e-01 4.89830077e-01 -5.56234896e-01 1.15722501e+00
4.77205664e-01 7.88942397e-01 -4.17285651e-01 -4.04106528e-02
2.49659523e-01 5.31770773e-02 3.38499069e-01 1.80188227e+00
4.76979584e-01 -3.63424778e-01 2.25970134e-01 6.18939161e-01
-2.42218643e-01 1.71030343e-01 -4.78154361e-01 2.15866446e-01
7.11044729e-01 1.34595513e+00 -4.17118609e-01 -4.41731155e-01
-3.61114770e-01 7.69470215e-01 2.31520608e-01 3.11393231e-01
-8.06757092e-01 -7.41114140e-01 1.17242885e+00 6.32189095e-01
3.17650646e-01 -2.32374460e-01 -3.87223512e-02 -7.39282489e-01
-2.59593129e-01 -5.05657017e-01 4.64105606e-02 -5.44766307e-01
-1.05503023e+00 1.29975629e+00 -4.71450448e-01 -9.06197131e-01
-4.75441962e-01 -8.21360528e-01 -7.10922480e-01 8.04638565e-01
-1.27564240e+00 -7.97940135e-01 -5.81929266e-01 9.35654819e-01
2.32344002e-01 -7.44150758e-01 1.13556027e+00 8.60356390e-01
-8.24915171e-01 8.92993927e-01 1.78369135e-01 3.11754048e-01
3.48409772e-01 -1.32245028e+00 7.99852550e-01 4.62350994e-01
3.76485050e-01 9.59933177e-02 7.30583489e-01 -4.09533530e-02
-1.38268447e+00 -1.34054708e+00 6.04680121e-01 -1.68160036e-01
8.05421889e-01 -5.65324008e-01 -6.24186337e-01 1.02587414e+00
2.68700838e-01 4.48404074e-01 9.41756070e-01 2.91206360e-01
-3.32313836e-01 -8.54400396e-01 -1.02453208e+00 4.08440888e-01
5.08035123e-01 -4.04906154e-01 4.61339384e-01 6.59625947e-01
8.81601274e-01 -1.30250692e-01 -1.26094520e+00 -1.92169011e-01
4.36913908e-01 -1.62272954e+00 1.00463152e+00 -9.16493356e-01
-1.60486564e-01 2.84421772e-01 -2.25735798e-01 -1.24944699e+00
-1.95801228e-01 -8.64516258e-01 -4.80394155e-01 1.14943850e+00
5.14208794e-01 -6.11151755e-01 1.21253562e+00 2.41880178e-01
5.74415969e-03 -5.88908017e-01 -7.04615772e-01 -7.83771873e-01
-4.23000187e-01 -1.11222005e+00 1.40731657e+00 3.95919979e-01
-1.75593272e-01 2.54384905e-01 -2.69925117e-01 2.28235617e-01
1.33520603e-01 1.18054589e-02 1.26465261e+00 -1.32071197e+00
-5.17107964e-01 -4.73701894e-01 -1.15587533e+00 -8.95784318e-01
3.82829189e-01 -8.30023766e-01 -7.67043114e-01 -1.06048703e+00
-6.55429840e-01 -9.38238502e-01 -4.74640429e-01 1.28021330e-01
8.46597612e-01 2.01663822e-01 2.10145071e-01 -1.06195308e-01
-4.81121987e-01 -4.95551713e-02 4.92149740e-01 6.28115386e-02
-2.11135983e-01 3.59362602e-01 -2.70950615e-01 3.44756514e-01
1.02476859e+00 -1.40621379e-01 -3.37392122e-01 -5.98942935e-01
1.10410750e-01 5.35401404e-01 1.75818503e-01 -1.66149485e+00
2.09206611e-01 -2.91514602e-02 4.55082268e-01 -3.99286270e-01
5.49425662e-01 -1.45417154e+00 1.59448460e-01 3.28275293e-01
-1.54529482e-01 2.70852983e-01 2.28128016e-01 4.23997492e-01
-1.43038541e-01 1.33777961e-01 5.29333830e-01 2.73163050e-01
-7.88160026e-01 7.69795477e-01 -4.17863995e-01 -2.70117760e-01
1.12987518e+00 1.85519800e-01 -5.22817254e-01 -6.38477445e-01
-9.04158771e-01 9.74698290e-02 2.25481883e-01 2.94446886e-01
5.50541341e-01 -1.37080920e+00 -7.02894151e-01 6.36444747e-01
-7.92118013e-02 -8.46593201e-01 1.38663454e-02 5.64330816e-01
-1.13557625e+00 7.34123766e-01 -6.50487602e-01 -4.46663857e-01
-1.05211258e+00 8.69800866e-01 5.16479731e-01 -2.88975835e-01
-5.82948267e-01 5.30681372e-01 -8.02887738e-01 -2.55433738e-01
4.41349864e-01 1.08800732e-01 -1.54843450e-01 5.73760718e-02
7.83426285e-01 7.63029456e-01 7.06018806e-01 -3.04184854e-01
-3.39330792e-01 -1.97669595e-01 7.29655251e-02 6.29090294e-02
1.41124237e+00 1.02060400e-01 -2.68179446e-01 2.93353379e-01
1.76038241e+00 2.42186915e-02 -6.19942129e-01 -1.88635811e-01
-1.73297413e-02 -6.25249863e-01 1.80876911e-01 -5.16559005e-01
-1.74244964e+00 1.09749198e+00 7.39236474e-01 8.68479967e-01
1.02493215e+00 -4.38052237e-01 1.16836143e+00 6.26027465e-01
5.51014543e-01 -7.87891090e-01 -2.33400941e-01 5.39500654e-01
1.44603178e-01 -7.80054212e-01 -3.93844247e-01 -3.90664190e-01
-2.58180827e-01 1.72053468e+00 3.09014380e-01 -2.55827755e-01
1.31308043e+00 7.31293738e-01 2.25166127e-01 -3.70873928e-01
-1.37557089e+00 -2.81331450e-01 -3.26693028e-01 9.14881706e-01
3.05075943e-01 3.98491383e-01 2.64775276e-01 2.85148174e-01
-4.46182936e-01 -5.89971095e-02 1.11209679e+00 2.43460402e-01
-3.90900463e-01 -1.36369193e+00 -1.86820626e-02 8.90956640e-01
-4.31995720e-01 1.83391929e-01 -2.48640850e-02 7.38703549e-01
-3.86532582e-02 9.57021713e-01 6.33353233e-01 -1.17736363e+00
6.39169991e-01 -4.29527871e-02 -2.72740871e-01 -4.46256101e-01
-9.09140766e-01 -4.28395152e-01 4.38495576e-01 -1.26276207e+00
3.57116401e-01 -1.78341568e-01 -9.11633849e-01 -1.28414440e+00
2.61675596e-01 3.80418807e-01 9.75023091e-01 5.69394231e-01
4.75945950e-01 8.17968309e-01 1.04649365e+00 -5.17613828e-01
-2.53936887e-01 -6.88557625e-01 -9.41649139e-01 2.36382976e-01
3.60874802e-01 -1.45884380e-01 -1.22489482e-01 -4.31560040e-01] | [5.07667875289917, 7.230299472808838] |
7da6b93a-5dee-447b-9442-079e204ffa00 | unconstrained-scene-generation-with-locally | 2104.00670 | null | https://arxiv.org/abs/2104.00670v1 | https://arxiv.org/pdf/2104.00670v1.pdf | Unconstrained Scene Generation with Locally Conditioned Radiance Fields | We tackle the challenge of learning a distribution over complex, realistic, indoor scenes. In this paper, we introduce Generative Scene Networks (GSN), which learns to decompose scenes into a collection of many local radiance fields that can be rendered from a free moving camera. Our model can be used as a prior to generate new scenes, or to complete a scene given only sparse 2D observations. Recent work has shown that generative models of radiance fields can capture properties such as multi-view consistency and view-dependent lighting. However, these models are specialized for constrained viewing of single objects, such as cars or faces. Due to the size and complexity of realistic indoor environments, existing models lack the representational capacity to adequately capture them. Our decomposition scheme scales to larger and more complex scenes while preserving details and diversity, and the learned prior enables high-quality rendering from viewpoints that are significantly different from observed viewpoints. When compared to existing models, GSN produces quantitatively higher-quality scene renderings across several different scene datasets. | ['Joshua M. Susskind', 'Graham W. Taylor', 'Nitish Srivastava', 'Miguel Angel Bautista', 'Terrance DeVries'] | 2021-04-01 | null | http://openaccess.thecvf.com//content/ICCV2021/html/DeVries_Unconstrained_Scene_Generation_With_Locally_Conditioned_Radiance_Fields_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/DeVries_Unconstrained_Scene_Generation_With_Locally_Conditioned_Radiance_Fields_ICCV_2021_paper.pdf | iccv-2021-1 | ['scene-generation'] | ['computer-vision'] | [ 3.50643754e-01 -3.36952478e-01 3.98240298e-01 -6.78483546e-01
-5.87000191e-01 -6.98936701e-01 7.00053334e-01 -4.47275072e-01
3.25836062e-01 6.52294934e-01 4.05062914e-01 -4.84055914e-02
-1.37945055e-03 -9.44463909e-01 -8.14153969e-01 -8.28849733e-01
2.92707175e-01 4.35035825e-01 6.54683039e-02 -1.48886368e-01
-2.51543134e-01 6.95540786e-01 -1.78655744e+00 4.63853747e-01
5.18486083e-01 6.60486698e-01 6.45061016e-01 9.80651975e-01
6.59840032e-02 9.95318055e-01 -3.95356923e-01 -4.79804799e-02
4.05053526e-01 -3.99444401e-01 -3.74747872e-01 4.75143492e-01
9.89532113e-01 -7.82681167e-01 -4.62969214e-01 8.81684244e-01
2.07860738e-01 2.26579368e-01 5.93143165e-01 -1.04736614e+00
-7.65166521e-01 -1.33266509e-01 -3.84853333e-01 -1.95606902e-01
5.52103937e-01 8.71838704e-02 8.44707429e-01 -7.27972746e-01
6.91163301e-01 1.39420295e+00 4.81627971e-01 5.62124670e-01
-1.55076313e+00 -3.01799566e-01 4.35186774e-01 -9.65420827e-02
-1.08607554e+00 -5.11019945e-01 9.12549675e-01 -4.19060707e-01
6.82066619e-01 4.52147663e-01 6.99294448e-01 1.26666749e+00
2.79092818e-01 3.56818378e-01 1.14650428e+00 -1.81590959e-01
2.60802418e-01 7.95873031e-02 -4.62656856e-01 4.39292014e-01
1.79702863e-01 2.99201701e-02 -4.88257885e-01 -2.26941958e-01
1.26498008e+00 3.03115189e-01 -4.07583147e-01 -6.54028952e-01
-1.22113228e+00 5.82936823e-01 5.47210336e-01 -1.39692202e-01
-3.73947710e-01 2.51358986e-01 -3.15580755e-01 -4.12186645e-02
5.61348021e-01 4.19440776e-01 -4.15487945e-01 2.34027401e-01
-6.31455779e-01 3.94888252e-01 6.21744692e-01 1.15201306e+00
9.26617980e-01 2.73071349e-01 1.47736371e-01 7.27373123e-01
3.95613760e-01 9.92647648e-01 -2.24445447e-01 -1.55547929e+00
2.64605790e-01 1.57223254e-01 2.45350957e-01 -1.07857168e+00
-1.85209513e-01 -3.31695765e-01 -1.10096514e+00 3.59623551e-01
8.29317942e-02 -5.30726425e-02 -1.07199323e+00 1.90019083e+00
4.87194479e-01 2.44224295e-01 -1.05025820e-01 9.60600793e-01
9.63101268e-01 9.76252675e-01 -4.18096453e-01 -1.97771676e-02
1.07228482e+00 -8.87424529e-01 -4.81532633e-01 -4.19829577e-01
-7.84839541e-02 -1.01904261e+00 9.27445710e-01 5.03199577e-01
-1.14699543e+00 -7.28986442e-01 -7.18704760e-01 -3.34159791e-01
7.13619143e-02 -1.54264048e-01 9.05688465e-01 4.46420461e-01
-1.31675887e+00 1.39840111e-01 -9.03887868e-01 -2.14020103e-01
2.75339693e-01 -6.05697185e-02 -2.87549108e-01 -8.30066264e-01
-5.24316072e-01 5.28727174e-01 -2.34874979e-01 -1.73807684e-02
-1.43517995e+00 -9.27652836e-01 -1.16786742e+00 -3.54432724e-02
5.36929779e-02 -1.34988320e+00 1.13057673e+00 -7.24762201e-01
-1.47147453e+00 5.89353085e-01 -4.36499029e-01 1.43554464e-01
2.37782672e-01 -2.35051140e-01 -2.83648401e-01 1.16770029e-01
3.54676321e-02 5.64634800e-01 6.76536918e-01 -1.94890404e+00
-2.78495282e-01 -2.16338038e-01 6.44961178e-01 4.71932501e-01
2.07086504e-01 -3.35112423e-01 -4.54602718e-01 -3.73208940e-01
4.26071525e-01 -9.12362397e-01 -5.86750805e-01 1.16509713e-01
-4.81152683e-01 5.48011780e-01 8.26529920e-01 -3.66996497e-01
4.67830479e-01 -2.01616621e+00 3.37247670e-01 2.25018952e-02
3.50203216e-01 -2.92286098e-01 -3.24362993e-01 4.68788415e-01
-1.69256050e-02 -2.57912666e-01 -3.14651290e-03 -7.93294787e-01
-7.83388093e-02 5.56194127e-01 -4.66210663e-01 4.21235085e-01
-5.06125502e-02 6.14127517e-01 -9.45208788e-01 -7.54957497e-02
6.29912794e-01 1.06233358e+00 -9.72366929e-01 4.00595963e-01
-5.01141250e-01 1.07717967e+00 -2.89256483e-01 3.91300321e-01
9.94421363e-01 -5.86182594e-01 2.60138273e-01 -2.98379779e-01
7.23947361e-02 3.92559320e-01 -1.31802607e+00 2.00392103e+00
-8.48854005e-01 7.24861503e-01 3.46900076e-01 -4.06289071e-01
7.42483854e-01 5.24246804e-02 4.78757322e-01 -5.82874358e-01
-9.52327698e-02 -2.59919018e-01 -4.71802205e-01 -1.94063216e-01
5.96976280e-01 -3.20816368e-01 1.02347843e-01 3.64471883e-01
-3.03373840e-02 -7.33916044e-01 -8.15385766e-03 3.25148255e-01
8.40274692e-01 3.81077290e-01 6.66452125e-02 -3.43637727e-02
1.56228706e-01 -3.67558777e-01 5.75890958e-01 6.82971716e-01
5.74152946e-01 1.09710491e+00 -3.84400673e-02 -6.39005005e-01
-1.13199878e+00 -1.47729194e+00 -1.95065156e-01 8.87263596e-01
3.06600392e-01 -4.80973393e-01 -3.10044765e-01 -1.45932091e-02
-2.20646814e-01 7.10933864e-01 -3.28682870e-01 1.58083126e-01
-4.83020723e-01 -7.48961568e-01 -1.74874663e-01 2.85599798e-01
4.08367306e-01 -6.31645620e-01 -6.33532166e-01 -2.80020405e-02
-4.74820524e-01 -1.54575741e+00 -2.29326606e-01 -3.09617538e-02
-6.40103996e-01 -8.39465141e-01 -4.89481449e-01 -5.06988168e-01
8.99659991e-01 7.79633462e-01 1.68690383e+00 -8.22854936e-02
-5.09824216e-01 8.09262395e-01 6.27722442e-02 -1.89455330e-01
-4.18305695e-01 -5.35502970e-01 -1.02137281e-02 1.16657376e-01
-2.21209571e-01 -1.16130185e+00 -7.34680533e-01 2.68886894e-01
-9.94698822e-01 6.04146540e-01 1.75108343e-01 5.41004598e-01
8.41382861e-01 1.76733300e-01 -2.22904757e-01 -9.15177226e-01
-1.81349106e-02 -4.88085330e-01 -8.06238294e-01 5.01368530e-02
-2.47603543e-02 -1.75408006e-01 8.61448407e-01 -1.79524526e-01
-1.65905023e+00 2.17426606e-02 -3.72013077e-02 -3.75206888e-01
-5.92537761e-01 -5.93550876e-02 -4.49469179e-01 -4.44211401e-02
5.68212032e-01 1.81534648e-01 -5.70532918e-01 -5.74002028e-01
5.60735703e-01 -4.37552854e-02 5.60125947e-01 -8.12264085e-01
9.71114814e-01 9.95632946e-01 4.04516309e-01 -8.65844309e-01
-1.09837842e+00 -2.55459130e-01 -6.16353452e-01 -2.47636829e-02
8.68983209e-01 -1.28351116e+00 -5.25895476e-01 4.23978239e-01
-1.24790609e+00 -5.90136409e-01 -3.86954069e-01 4.46504176e-01
-5.97823918e-01 1.68755934e-01 -5.75957298e-01 -7.00124502e-01
2.42758334e-01 -9.77886319e-01 1.50223029e+00 2.23303542e-01
1.43217772e-01 -1.22723901e+00 3.85702163e-01 3.48280430e-01
4.29608792e-01 5.29765964e-01 8.78823519e-01 5.87075591e-01
-1.25541985e+00 1.76115274e-01 -1.45056754e-01 3.72732133e-01
5.25111437e-01 1.46588475e-01 -1.27476788e+00 -3.11382413e-01
2.53461957e-01 -2.64789402e-01 7.38844275e-01 6.25334144e-01
1.32005310e+00 -2.61246711e-01 -2.15140596e-01 1.24009383e+00
1.67219663e+00 5.71752191e-02 7.02874839e-01 -1.30776241e-01
1.12012136e+00 5.45952141e-01 2.04076357e-02 5.76864839e-01
7.04201698e-01 6.80805266e-01 7.19988167e-01 -3.57815653e-01
-4.16500121e-01 -3.53813797e-01 7.77431801e-02 8.37403536e-01
-1.37834758e-01 -6.07985020e-01 -6.90734506e-01 4.48574364e-01
-1.52189922e+00 -1.13320780e+00 -2.53175706e-01 2.02858639e+00
4.61065412e-01 -2.48192549e-01 -4.30395037e-01 -3.10134381e-01
2.15039209e-01 5.79845011e-01 -6.65528238e-01 -1.43960118e-01
-3.73550773e-01 -2.22881418e-02 1.38566539e-01 7.39825785e-01
-7.99845278e-01 5.77301383e-01 7.05428505e+00 2.63475835e-01
-1.03290772e+00 4.12913188e-02 6.55245721e-01 -3.86835784e-01
-1.11618960e+00 1.30282298e-01 -7.31374204e-01 1.07649915e-01
5.02292752e-01 2.94614881e-01 7.40313053e-01 7.31793880e-01
1.00255653e-01 -1.19188271e-01 -1.19308567e+00 1.22885084e+00
3.55176538e-01 -1.36459374e+00 2.52319574e-01 1.68126300e-01
1.38004935e+00 1.26105055e-01 2.83232778e-01 -4.74677980e-02
9.90675092e-01 -1.09710145e+00 5.02024114e-01 7.04856813e-01
7.33209848e-01 -5.11790872e-01 1.21943116e-01 4.18109894e-01
-1.17528188e+00 2.63699830e-01 -6.38551056e-01 -1.83420077e-01
5.56794524e-01 8.82405460e-01 -4.27925289e-01 6.81557834e-01
9.05973852e-01 8.61188352e-01 -4.41969186e-01 7.24484503e-01
-3.17635387e-01 2.60672450e-01 -5.42477012e-01 5.69258332e-01
-9.81533621e-03 -3.45500082e-01 4.07451749e-01 8.00465584e-01
4.83332723e-01 1.77956864e-01 3.11503798e-01 1.06183469e+00
-1.26747484e-03 -4.04254645e-01 -8.53146434e-01 5.93876660e-01
1.44071206e-01 1.44756532e+00 -4.34025884e-01 -2.79231757e-01
-4.13002312e-01 1.01298165e+00 2.19286323e-01 8.71157646e-01
-8.11454237e-01 3.71685356e-01 1.03305566e+00 2.39901498e-01
3.83425087e-01 -6.09997272e-01 -1.48961991e-01 -1.60497928e+00
1.07660778e-02 -7.13275194e-01 -1.63916647e-01 -1.38564205e+00
-1.28394353e+00 7.02534080e-01 2.11212754e-01 -1.14148211e+00
-2.83587277e-01 -6.12705469e-01 -5.02157867e-01 8.59242558e-01
-1.64570439e+00 -1.41477001e+00 -8.00236702e-01 7.83162594e-01
5.52328110e-01 2.25439757e-01 9.32616889e-01 2.17693269e-01
-1.19519889e-01 -1.84487358e-01 2.90386111e-01 -2.82397956e-01
5.82123816e-01 -1.24658716e+00 3.91929477e-01 9.72785652e-01
4.50358123e-01 6.21055067e-01 7.08789706e-01 -2.99243748e-01
-1.42302144e+00 -1.15943122e+00 3.69386315e-01 -8.00325513e-01
4.87802364e-02 -6.66408002e-01 -5.91824889e-01 9.08458650e-01
2.79338568e-01 4.01674896e-01 8.34602594e-01 1.50112793e-01
-6.76096201e-01 -1.95713744e-01 -1.05684769e+00 5.32729864e-01
1.30292583e+00 -5.49138546e-01 7.17702648e-03 3.69712681e-01
7.20252335e-01 -7.56751180e-01 -6.89270437e-01 2.48855814e-01
5.36787570e-01 -1.48603129e+00 1.38891232e+00 -4.07120496e-01
7.24567175e-01 -4.39397126e-01 -7.22106934e-01 -1.44471717e+00
-5.76331377e-01 -4.63734776e-01 -7.79993683e-02 1.09123147e+00
-6.93306848e-02 -4.39002216e-01 4.41770613e-01 6.35118604e-01
-1.86432049e-01 -2.91498363e-01 -5.76321661e-01 -4.68868703e-01
-1.71443045e-01 -5.69740057e-01 9.13824201e-01 8.30649674e-01
-1.03301334e+00 4.38775659e-01 -6.30052567e-01 5.90713382e-01
1.06002486e+00 6.40214205e-01 1.16825008e+00 -1.15812171e+00
-7.42408156e-01 3.31764147e-02 -1.93184897e-01 -1.60707259e+00
5.36537431e-02 -5.03159106e-01 1.38028309e-01 -1.88743997e+00
3.23653221e-01 -5.49222946e-01 1.43054262e-01 1.04272380e-01
-3.41179334e-02 3.97573858e-01 2.65979588e-01 9.10583586e-02
-5.60500383e-01 7.14038670e-01 1.61653864e+00 1.82472961e-03
1.03593677e-01 -7.84776062e-02 -6.58487558e-01 1.07287657e+00
5.06860912e-01 -2.12355822e-01 -9.54322815e-01 -1.14126444e+00
5.69411993e-01 8.76729116e-02 7.36494660e-01 -8.97761345e-01
8.17045569e-03 -5.99736392e-01 8.30669284e-01 -6.68295503e-01
8.83034468e-01 -9.34068382e-01 7.06813335e-01 -2.75370181e-01
3.64088006e-02 -1.02506680e-02 1.26135483e-01 5.70692837e-01
-1.10053033e-01 2.86064893e-01 6.94962144e-01 -4.24031824e-01
-5.18209100e-01 4.94534194e-01 -1.60044074e-01 -6.80867061e-02
7.06626356e-01 -1.85714883e-03 -3.49145830e-01 -8.58388722e-01
-4.86466885e-01 -1.07185796e-01 8.88081372e-01 4.03348893e-01
6.07961059e-01 -1.59350288e+00 -6.18954480e-01 4.41442013e-01
8.85949358e-02 7.48953521e-01 7.09596157e-01 2.65101790e-01
-6.12695873e-01 2.24530384e-01 -7.04199895e-02 -9.01864231e-01
-1.13589585e+00 3.82156402e-01 3.40433598e-01 -1.84689716e-01
-7.30381072e-01 8.46403182e-01 1.17681134e+00 -8.07793856e-01
-2.80374587e-01 -5.59126258e-01 2.24564373e-01 -5.60869575e-01
5.41412354e-01 1.30949005e-01 -1.91437662e-01 -7.31190860e-01
-2.12888032e-01 8.36477458e-01 3.31516594e-01 -4.66664955e-02
1.60854042e+00 -4.60098833e-01 -2.05524623e-01 4.87776190e-01
1.13814700e+00 3.54620457e-01 -1.85664713e+00 -3.10209155e-01
-9.96385455e-01 -8.97284925e-01 6.79589063e-02 -5.89559972e-01
-1.21571469e+00 8.74247491e-01 2.04947740e-01 -3.63998041e-02
1.19257653e+00 1.29950896e-01 6.59792125e-01 2.70549089e-01
7.46177971e-01 -4.24639374e-01 1.93461463e-01 5.37308335e-01
8.78094137e-01 -1.19539225e+00 1.35886490e-01 -5.40329754e-01
-2.87883699e-01 9.77780402e-01 6.30136430e-01 -1.11751825e-01
6.70254290e-01 5.80574155e-01 4.18771803e-02 -2.50750512e-01
-9.31535542e-01 1.75897349e-02 4.24954385e-01 9.40327942e-01
2.63682604e-01 7.14748427e-02 8.80171537e-01 -4.41358052e-02
-4.43755239e-01 -4.41008866e-01 5.94195843e-01 6.37746871e-01
-1.23474658e-01 -8.44561040e-01 -4.59254414e-01 1.18601874e-01
-4.65691984e-02 -9.36710611e-02 3.22605632e-02 3.76002789e-01
2.91108757e-01 9.94147062e-01 2.34430403e-01 -1.23486333e-01
1.83427975e-01 -5.15309215e-01 8.53351772e-01 -8.80222976e-01
1.13054648e-01 2.92573929e-01 -1.23810567e-01 -8.81304920e-01
-7.62345016e-01 -5.90392172e-01 -6.73042357e-01 -4.93852526e-01
4.51387018e-02 -3.26076508e-01 5.69312990e-01 6.58954501e-01
3.64289135e-01 7.84906507e-01 7.85808980e-01 -1.30698788e+00
9.44335237e-02 -4.86408651e-01 -6.28990889e-01 5.44923425e-01
7.51136720e-01 -5.24270833e-01 -4.31338519e-01 4.71069485e-01] | [9.316012382507324, -3.082720994949341] |
df3f84ee-c6df-4b5f-9fa0-b83efb1b8620 | minimax-bayes-reinforcement-learning | 2302.10831 | null | https://arxiv.org/abs/2302.10831v1 | https://arxiv.org/pdf/2302.10831v1.pdf | Minimax-Bayes Reinforcement Learning | While the Bayesian decision-theoretic framework offers an elegant solution to the problem of decision making under uncertainty, one question is how to appropriately select the prior distribution. One idea is to employ a worst-case prior. However, this is not as easy to specify in sequential decision making as in simple statistical estimation problems. This paper studies (sometimes approximate) minimax-Bayes solutions for various reinforcement learning problems to gain insights into the properties of the corresponding priors and policies. We find that while the worst-case prior depends on the setting, the corresponding minimax policies are more robust than those that assume a standard (i.e. uniform) prior. | ['Emilio Jorge', 'Divya Grover', 'Hannes Eriksson', 'Christos Dimitrakakis', 'Thomas Kleine Buening'] | 2023-02-21 | null | null | null | null | ['decision-making-under-uncertainty', 'decision-making-under-uncertainty'] | ['medical', 'reasoning'] | [ 3.45049679e-01 1.69341803e-01 -4.71741468e-01 -5.58433533e-01
-7.56199956e-01 -5.78020632e-01 4.95672971e-01 3.23162466e-01
-9.02684391e-01 9.90127683e-01 -1.13430284e-01 -7.30300903e-01
-7.33577311e-01 -7.66647041e-01 -4.64623183e-01 -8.51630390e-01
2.23055750e-01 5.55121839e-01 1.14606030e-01 -1.60376102e-01
6.14496529e-01 3.18605185e-01 -1.30536616e+00 -4.18303788e-01
7.88845956e-01 1.00434911e+00 1.18412167e-01 5.66924155e-01
1.05500191e-01 3.12994272e-01 -6.23894870e-01 -3.79271924e-01
3.54907274e-01 -2.34632045e-01 -6.17556214e-01 2.73398727e-01
3.20018679e-02 -5.32296658e-01 -2.01619212e-02 1.38792324e+00
3.39547962e-01 3.50703806e-01 1.03093910e+00 -1.26898551e+00
-2.99782991e-01 6.12064004e-01 -5.05174637e-01 4.44544256e-01
2.85715133e-01 1.81248054e-01 1.03284025e+00 4.88617755e-02
2.02146173e-01 1.38682461e+00 2.79594421e-01 3.47258270e-01
-1.37575138e+00 -4.42507803e-01 3.06530029e-01 -6.99550882e-02
-1.34060740e+00 -5.51737547e-01 3.40350270e-01 -3.47422063e-01
2.75184155e-01 -2.13179886e-02 3.70076805e-01 1.01252389e+00
6.84529722e-01 7.12835968e-01 1.52025914e+00 -5.97639561e-01
7.92774558e-01 2.52725899e-01 1.92375720e-01 3.08550715e-01
8.38018656e-01 4.76361960e-01 -3.15424860e-01 -3.57892036e-01
7.96097636e-01 -4.01458517e-02 -1.49095982e-01 -5.06071270e-01
-8.80498588e-01 1.21829069e+00 -7.36038238e-02 9.97096449e-02
-5.85044563e-01 4.69313979e-01 4.70459200e-02 5.78081787e-01
5.23509691e-03 5.24100959e-01 -3.13550949e-01 -2.67626718e-02
-6.07097507e-01 7.35021234e-01 1.06364715e+00 7.23145425e-01
6.32272243e-01 1.02367073e-01 -3.18601549e-01 4.52607781e-01
6.79893672e-01 4.47923928e-01 -1.31092919e-02 -1.34929514e+00
4.80259567e-01 -4.01786059e-01 9.84363377e-01 -7.84115672e-01
-2.16078222e-01 -4.63415295e-01 -3.24670583e-01 6.73288226e-01
9.62782025e-01 -5.91311276e-01 -6.52095199e-01 1.78488910e+00
9.12607014e-02 -3.17025423e-01 1.30463699e-02 8.45165789e-01
-3.24444264e-01 4.18330789e-01 2.42669493e-01 -6.09497190e-01
1.06366992e+00 -1.18955962e-01 -9.61842835e-01 -6.10873282e-01
2.11545631e-01 -6.23255849e-01 7.23000228e-01 5.99074304e-01
-8.60412598e-01 1.13694869e-01 -1.15793312e+00 5.63595712e-01
1.18800215e-01 -4.37228978e-01 5.74034810e-01 1.08851194e+00
-8.20740342e-01 7.06969142e-01 -5.96386850e-01 -2.91382462e-01
-3.79919522e-02 2.13761717e-01 1.99315831e-01 3.27568538e-02
-1.25594854e+00 1.28701758e+00 6.00639224e-01 -5.66851608e-02
-9.76650000e-01 -1.85321838e-01 -6.49589360e-01 6.10429011e-02
1.05239546e+00 -5.91828525e-01 1.87359166e+00 -8.30013812e-01
-1.78579712e+00 2.94000357e-01 2.42546517e-02 -5.55338323e-01
7.16009378e-01 -3.44197154e-02 1.46214694e-01 -2.40326654e-02
5.74100614e-02 3.28292340e-01 9.89690661e-01 -1.17167509e+00
-7.34330475e-01 -2.14925945e-01 4.16103303e-01 4.83542770e-01
2.29639798e-01 2.56456677e-02 1.62169427e-01 -4.79270786e-01
4.69359607e-02 -9.08696353e-01 -7.27435887e-01 3.59164886e-02
-3.14101875e-01 -2.78132260e-01 9.47622284e-02 -3.18910778e-01
1.20757079e+00 -1.89616609e+00 -3.65421295e-01 4.43264633e-01
-2.14488685e-01 -3.67531091e-01 2.74856448e-01 4.21012461e-01
2.38950163e-01 6.04073107e-02 -3.36221904e-01 5.67543954e-02
5.08290768e-01 3.76544058e-01 -3.26510757e-01 7.48470724e-01
-1.39209107e-01 2.89441079e-01 -9.12899137e-01 -3.71032029e-01
2.28134468e-01 -1.29287317e-01 -5.48903227e-01 -1.18829072e-01
-1.97766155e-01 9.15287137e-02 -9.67080116e-01 3.32162529e-01
3.03765088e-01 -6.63332120e-02 5.33200502e-01 4.35634077e-01
-1.00479722e-01 3.69622618e-01 -1.72590804e+00 1.11893976e+00
-1.69687659e-01 4.46026176e-01 3.51310641e-01 -1.21094561e+00
6.54965878e-01 2.64663070e-01 4.42077935e-01 -2.56794482e-01
3.44936877e-01 4.96759042e-02 2.25052357e-01 -2.77125895e-01
3.20842385e-01 -8.23068380e-01 -3.38655680e-01 7.77070045e-01
-2.06947759e-01 -4.45099920e-01 2.37947136e-01 -1.81084558e-01
9.05202985e-01 -8.09463300e-03 7.70892620e-01 -7.57002592e-01
-2.12538213e-01 -1.19474307e-01 8.38030994e-01 1.39905941e+00
-3.49745691e-01 1.77708492e-01 7.27378845e-01 -2.85227615e-02
-6.60045743e-01 -1.00200737e+00 -5.15734792e-01 8.42805684e-01
9.51297954e-02 -1.15162998e-01 -7.02843010e-01 -4.63314325e-01
2.71211416e-01 1.20717680e+00 -3.78286123e-01 -3.66923818e-03
1.27245083e-01 -1.01151216e+00 5.57644479e-02 3.18749517e-01
1.51187405e-01 -5.22627115e-01 -9.83063698e-01 4.20963138e-01
9.05126184e-02 -7.52899230e-01 -4.04464036e-01 4.39857453e-01
-8.71638060e-01 -1.02233541e+00 -4.76139098e-01 1.08237423e-01
2.99909055e-01 -6.43847231e-03 1.05177546e+00 -4.80321437e-01
2.45622769e-01 8.24858189e-01 -2.98780464e-02 -7.76130915e-01
-3.89992476e-01 -4.69420224e-01 2.67954648e-01 -1.47913113e-01
4.25819039e-01 -2.19348148e-01 -4.15115237e-01 2.36455932e-01
-7.71089077e-01 -5.39350688e-01 6.25505984e-01 6.93448663e-01
4.89860177e-01 5.41418850e-01 5.59559464e-01 -7.94382989e-01
1.08995628e+00 -4.31187898e-01 -1.06662679e+00 3.83936018e-01
-8.43425453e-01 4.57441330e-01 1.42203301e-01 -3.76241714e-01
-1.30132043e+00 -1.26187965e-01 1.49794802e-01 -1.79529041e-02
-4.04613733e-01 5.49461961e-01 -8.96699503e-02 7.93383792e-02
4.57856745e-01 -2.80301332e-01 -1.51062295e-01 -3.71012837e-01
8.46936107e-02 5.85418642e-01 8.27192366e-02 -1.11970627e+00
3.43561471e-01 -3.78246903e-02 8.35767761e-02 -8.00688446e-01
-1.07203078e+00 -1.80905219e-02 -2.71956772e-01 -1.48185035e-02
9.62197244e-01 -4.69508946e-01 -7.75764942e-01 1.86152861e-03
-7.15946138e-01 -3.76495987e-01 -3.69381279e-01 7.62151539e-01
-1.16187477e+00 3.32241654e-01 -2.74213493e-01 -1.39805996e+00
3.85756791e-01 -1.21946418e+00 6.07681870e-01 1.99016005e-01
-3.66575658e-01 -1.00571215e+00 -1.69939566e-02 1.58132717e-01
2.14004397e-01 8.81621540e-02 8.50946903e-01 -6.19754910e-01
-5.37538350e-01 -2.01738372e-01 2.30168924e-01 1.16875701e-01
5.18288650e-02 4.94691618e-02 -6.05996430e-01 -2.94057488e-01
4.26544309e-01 -2.09232733e-01 6.94216371e-01 8.07820201e-01
9.90980685e-01 -5.56444228e-01 -1.23143919e-01 7.54677951e-02
1.44743001e+00 8.48960280e-01 2.08084121e-01 3.63296807e-01
-1.74073577e-01 6.19735420e-01 8.80824208e-01 7.80809104e-01
3.90099347e-01 5.64799428e-01 3.56151879e-01 7.31687486e-01
7.67803848e-01 -2.06441998e-01 2.97055870e-01 1.12160761e-02
7.05821365e-02 -3.40887040e-01 -9.69854474e-01 1.29161075e-01
-1.94620180e+00 -1.06054664e+00 4.07730848e-01 2.56804752e+00
1.06553316e+00 5.34214735e-01 1.51500762e-01 6.87646419e-02
8.33317041e-01 1.06997870e-01 -6.57461047e-01 -5.58342040e-01
8.32008496e-02 -5.78818284e-02 8.32662344e-01 6.99329436e-01
-1.15510178e+00 3.92540455e-01 8.19815445e+00 8.37016821e-01
-6.61579370e-01 8.20773318e-02 6.41991317e-01 -1.13993317e-01
-3.50457460e-01 2.79579908e-01 -1.00144112e+00 6.12304568e-01
9.92590427e-01 -4.67208415e-01 3.71276528e-01 7.19097793e-01
3.85783195e-01 -6.80970430e-01 -1.23194146e+00 5.40785015e-01
-6.05350316e-01 -8.14707518e-01 -2.92441010e-01 2.47331515e-01
7.30918467e-01 -3.74075919e-01 -4.37777303e-02 1.83029488e-01
1.08740282e+00 -9.14599240e-01 8.78968894e-01 5.71220458e-01
5.35213649e-01 -1.00854349e+00 6.99003577e-01 7.33797848e-01
-3.20306331e-01 -4.76084113e-01 -5.28019547e-01 -3.77950370e-01
8.31596106e-02 6.99827969e-01 -5.25567651e-01 1.88118353e-01
5.35337150e-01 2.59099692e-01 7.19936639e-02 1.28987193e+00
-4.17222112e-01 5.90945780e-01 -6.47937000e-01 -3.00682664e-01
4.61010545e-01 -4.58575219e-01 5.96909404e-01 1.00275564e+00
3.56946886e-01 2.55062193e-01 4.33046818e-01 6.96052134e-01
3.88435155e-01 -1.74893230e-01 -5.37756026e-01 -7.57901222e-02
7.06715524e-01 6.62531912e-01 -6.51731372e-01 -7.12451115e-02
-3.44267040e-01 1.20993115e-01 3.61910351e-02 6.36621714e-01
-3.73454958e-01 -2.89234549e-01 7.01293349e-01 8.15178677e-02
2.07077459e-01 -2.94756204e-01 -3.57011527e-01 -9.07409132e-01
-1.66847378e-01 -9.88496125e-01 6.10944808e-01 -4.70941067e-01
-1.29082036e+00 -2.84957767e-01 7.88154423e-01 -8.26679468e-01
-5.07145047e-01 -7.94619083e-01 -5.89775741e-01 7.43028224e-01
-1.23103464e+00 -1.58356994e-01 5.82057476e-01 3.20525616e-01
2.96357960e-01 1.23197831e-01 4.60870445e-01 -3.18365425e-01
-6.09535456e-01 1.81472808e-01 5.38167477e-01 -1.84044927e-01
6.10359728e-01 -1.53827393e+00 -4.62893277e-01 6.80175662e-01
-1.51034608e-01 5.59920907e-01 1.36807954e+00 -4.87763524e-01
-1.38964915e+00 -4.23531204e-01 3.85402352e-01 -1.90767378e-01
7.86595225e-01 2.59774238e-01 -5.39567649e-01 8.59367073e-01
1.80446357e-01 -5.20907760e-01 4.68883604e-01 2.98556030e-01
-3.89125831e-02 -1.08660795e-01 -1.29530311e+00 6.21796846e-01
4.99429137e-01 -2.96124250e-01 -9.20289576e-01 1.74694553e-01
2.55946815e-01 -8.67567435e-02 -7.23336756e-01 2.96642721e-01
4.69179302e-01 -8.17071915e-01 7.33392060e-01 -6.13155663e-01
-4.29270929e-03 -2.64775634e-01 -4.98937249e-01 -1.47622097e+00
-2.59482801e-01 -9.48844135e-01 1.92887098e-01 8.57820451e-01
3.20814162e-01 -8.79247665e-01 6.27500355e-01 1.04924893e+00
2.84004629e-01 -7.22778678e-01 -1.08109927e+00 -9.24688101e-01
3.85134280e-01 -6.67349100e-01 3.60401362e-01 6.95768774e-01
-6.39371723e-02 1.87866494e-01 -2.62275279e-01 1.87014267e-02
1.04849946e+00 1.19390614e-01 2.41578698e-01 -1.17291570e+00
-6.04900539e-01 -6.80349886e-01 -1.58012018e-01 -1.19188797e+00
9.75247622e-02 -2.18436494e-01 3.99651825e-01 -1.36164474e+00
3.03545799e-02 -6.87216580e-01 -5.75596571e-01 3.07877153e-01
-1.49503782e-01 -5.48973620e-01 1.64790645e-01 -2.49911308e-01
-5.48470974e-01 5.34315169e-01 1.14828348e+00 3.04346830e-02
-1.25036880e-01 6.72963858e-01 -1.03672028e+00 9.22506928e-01
1.01865637e+00 -6.28184795e-01 -5.46107173e-01 -3.12381029e-01
5.05846083e-01 6.42207861e-01 1.51907220e-01 -4.36854124e-01
2.31403306e-01 -1.08489943e+00 1.95016310e-01 -4.17499810e-01
2.25005761e-01 -5.98208308e-01 -6.66336194e-02 3.96045357e-01
-5.89560747e-01 -2.08537858e-02 -1.07123502e-01 9.33323741e-01
6.94081262e-02 -1.10238373e+00 9.50892508e-01 -3.02705765e-01
-5.58192551e-01 1.21200703e-01 -8.93637538e-01 6.71485364e-02
9.64866698e-01 -8.05150643e-02 -5.32700494e-02 -9.38445687e-01
-8.26043308e-01 3.63318235e-01 4.39257950e-01 -6.69274479e-02
4.49582279e-01 -9.53423142e-01 -5.77119112e-01 -3.26722950e-01
-3.01491946e-01 -2.21237078e-01 -2.67414153e-01 6.05116546e-01
-3.19291130e-02 5.49835861e-01 -5.18945158e-02 -3.11199248e-01
-5.53478956e-01 4.32564497e-01 5.40234625e-01 -2.27889970e-01
-1.55925840e-01 5.70570469e-01 2.13890389e-01 -5.76687930e-03
4.60600644e-01 -1.10661983e-01 -6.07461147e-02 6.91827685e-02
5.40466368e-01 5.06005108e-01 -4.01824147e-01 3.58108915e-02
-1.94837138e-01 2.57672608e-01 3.04959379e-02 -6.61334157e-01
1.07998455e+00 -4.46458369e-01 2.08629623e-01 7.32173800e-01
4.06867445e-01 -3.30806226e-01 -1.62560403e+00 -2.98958600e-01
3.13085943e-01 -6.93274975e-01 2.66544163e-01 -6.01369381e-01
-6.12604260e-01 8.44632268e-01 3.63098830e-01 5.04959941e-01
8.66171896e-01 -3.82369787e-01 -5.35539165e-02 9.30050790e-01
7.39843547e-01 -1.35001504e+00 -3.92865956e-01 4.15940136e-01
8.52040887e-01 -1.18469346e+00 2.36596093e-01 3.83271948e-02
-7.53289819e-01 9.49027300e-01 2.80983716e-01 -1.99936345e-01
9.36519504e-01 2.57702768e-01 -1.70042396e-01 2.09055826e-01
-9.08896565e-01 -2.60029525e-01 7.15858536e-03 6.00955546e-01
1.74274057e-01 3.72218132e-01 -4.78902936e-01 4.89952505e-01
-2.66392320e-01 -4.31504883e-02 6.86307728e-01 1.14304006e+00
-7.82933056e-01 -9.74914193e-01 -5.43460190e-01 6.41076803e-01
-8.84147346e-01 2.91633427e-01 -9.43181589e-02 6.76910758e-01
-2.77854145e-01 1.41624916e+00 3.19978199e-03 1.78344667e-01
1.95372105e-01 1.41813299e-02 7.60984182e-01 -5.85900366e-01
2.18132846e-02 2.54535645e-01 2.42231846e-01 -4.32685107e-01
-4.76252258e-01 -9.67145443e-01 -7.82451391e-01 -4.20116097e-01
-3.00771177e-01 2.76052058e-01 6.02506816e-01 1.22019553e+00
-3.05980444e-01 1.67154089e-01 3.61802578e-01 -6.14069104e-01
-1.68384969e+00 -6.28429055e-01 -9.13697779e-01 -5.97710200e-02
5.30988276e-01 -9.40378785e-01 -4.40339774e-01 -4.41820145e-01] | [4.487454414367676, 3.0001368522644043] |
8621164b-7062-441f-ac6b-44b2e598362b | msinet-twins-contrastive-search-of-multi | 2303.07065 | null | https://arxiv.org/abs/2303.07065v1 | https://arxiv.org/pdf/2303.07065v1.pdf | MSINet: Twins Contrastive Search of Multi-Scale Interaction for Object ReID | Neural Architecture Search (NAS) has been increasingly appealing to the society of object Re-Identification (ReID), for that task-specific architectures significantly improve the retrieval performance. Previous works explore new optimizing targets and search spaces for NAS ReID, yet they neglect the difference of training schemes between image classification and ReID. In this work, we propose a novel Twins Contrastive Mechanism (TCM) to provide more appropriate supervision for ReID architecture search. TCM reduces the category overlaps between the training and validation data, and assists NAS in simulating real-world ReID training schemes. We then design a Multi-Scale Interaction (MSI) search space to search for rational interaction operations between multi-scale features. In addition, we introduce a Spatial Alignment Module (SAM) to further enhance the attention consistency confronted with images from different sources. Under the proposed NAS scheme, a specific architecture is automatically searched, named as MSINet. Extensive experiments demonstrate that our method surpasses state-of-the-art ReID methods on both in-domain and cross-domain scenarios. Source code available in https://github.com/vimar-gu/MSINet. | ['Jian Zhao', 'Yang You', 'Shanghang Zhang', 'Yuqiang Fang', 'Wei Jiang', 'Chen Chen', 'Hao Luo', 'Kai Wang', 'Jianyang Gu'] | 2023-03-13 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Gu_MSINet_Twins_Contrastive_Search_of_Multi-Scale_Interaction_for_Object_ReID_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Gu_MSINet_Twins_Contrastive_Search_of_Multi-Scale_Interaction_for_Object_ReID_CVPR_2023_paper.pdf | cvpr-2023-1 | ['person-re-identification', 'vehicle-re-identification', 'architecture-search'] | ['computer-vision', 'computer-vision', 'methodology'] | [-2.94480562e-01 -5.56697905e-01 -1.10741369e-01 -5.32178998e-01
-5.54890156e-01 -4.43546057e-01 7.14849234e-01 -3.71527344e-01
-5.58062673e-01 2.80957341e-01 1.62309274e-01 -1.94519073e-01
-3.73904496e-01 -4.92723733e-01 -6.30221665e-01 -6.71762168e-01
2.87087888e-01 4.91313010e-01 -6.43946826e-02 -2.43215322e-01
4.32882518e-01 6.09230101e-01 -1.73168337e+00 8.27160105e-03
1.20690954e+00 9.69668567e-01 6.31955683e-01 3.90641317e-02
-2.94258744e-01 3.30572635e-01 -6.16880774e-01 -7.25474834e-01
3.85712445e-01 -3.33393246e-01 -8.52834046e-01 -4.33806658e-01
4.94146109e-01 -1.39319777e-01 -6.28225267e-01 1.21074784e+00
9.74620998e-01 3.65481704e-01 7.68509448e-01 -1.20544517e+00
-1.30532789e+00 5.43988109e-01 -8.42758179e-01 7.27294266e-01
1.51606187e-01 5.02857864e-02 6.32392108e-01 -1.24206007e+00
3.69315654e-01 1.49174488e+00 3.82362247e-01 7.10281849e-01
-9.34513569e-01 -1.22399879e+00 2.89586067e-01 8.31483424e-01
-1.78733814e+00 -5.79898298e-01 8.71488154e-01 -9.68731195e-02
6.85997427e-01 4.37070668e-01 2.72270232e-01 1.17942011e+00
-1.47573516e-01 1.23120034e+00 7.80430734e-01 -3.47875416e-01
-1.30950347e-01 3.15241247e-01 2.78022319e-01 2.08633482e-01
2.39837155e-01 2.46751741e-01 -4.90183800e-01 5.14127649e-02
8.08420658e-01 -6.92475215e-02 -2.85736054e-01 -4.25144613e-01
-1.11640036e+00 4.41481233e-01 1.00012493e+00 3.95892620e-01
-3.06970865e-01 -2.14062974e-01 3.26266497e-01 2.02926636e-01
2.89488256e-01 6.10381961e-01 -1.94036797e-01 4.38758582e-01
-6.06642485e-01 1.30961522e-01 1.96559161e-01 1.20325685e+00
6.11330628e-01 -3.90605666e-02 -3.14356238e-01 1.14295638e+00
2.56701648e-01 5.21250427e-01 8.30610871e-01 -4.39252257e-01
4.78416830e-01 6.75437808e-01 6.63711354e-02 -9.80428636e-01
-3.43353897e-01 -8.55279148e-01 -1.03550446e+00 -3.63645375e-01
-6.65237308e-02 4.22154874e-01 -8.96624565e-01 1.78913140e+00
5.19842684e-01 2.56918907e-01 -4.40290570e-02 1.26575303e+00
1.06533313e+00 4.54790056e-01 2.68380105e-01 1.62770197e-01
1.45512116e+00 -1.22426164e+00 -5.59208810e-01 -2.84821182e-01
4.71753359e-01 -6.80858672e-01 1.03651893e+00 -1.31861284e-01
-1.07291937e+00 -9.29976225e-01 -7.97682583e-01 -1.06155947e-01
-6.64948285e-01 5.60951710e-01 2.55113214e-01 2.75987595e-01
-1.12148690e+00 1.46620020e-01 -4.10393387e-01 -4.60479140e-01
3.14664006e-01 4.52435344e-01 -4.58625793e-01 -1.37991354e-01
-1.36047757e+00 1.19667351e+00 4.59601045e-01 5.15746057e-01
-5.10143399e-01 -6.26841605e-01 -8.35709035e-01 1.88769996e-01
1.61554933e-01 -8.10407519e-01 1.18840003e+00 -8.65730882e-01
-1.05043077e+00 1.08514309e+00 -2.77535141e-01 -9.29889828e-02
5.12272060e-01 -3.02248120e-01 -5.21782398e-01 -2.34359324e-01
3.45910996e-01 1.04724848e+00 7.06357837e-01 -1.45961034e+00
-4.51004177e-01 -4.55866426e-01 -2.06022978e-01 5.47365546e-01
-5.70038557e-01 4.49114293e-01 -8.99741590e-01 -7.57656991e-01
-1.31620215e-02 -7.90640712e-01 5.87729104e-02 -2.15404123e-01
-3.87740999e-01 -6.35062575e-01 6.25643909e-01 -5.04680037e-01
1.22458613e+00 -2.29070807e+00 2.08288595e-01 1.45270914e-01
8.42226669e-03 5.17783403e-01 -6.12758398e-01 2.05259129e-01
-3.10564369e-01 1.29684880e-01 1.51715636e-01 -5.81938803e-01
-6.59857877e-03 1.30076203e-02 -2.26124257e-01 3.70233715e-01
1.06984511e-01 1.19484973e+00 -7.05617547e-01 -6.58172369e-01
-1.57305840e-02 2.91867435e-01 -1.61692202e-01 3.75703007e-01
2.58201927e-01 5.81892610e-01 -6.77607596e-01 7.31721282e-01
1.01006281e+00 -5.18492758e-01 -1.21914841e-01 -3.58416915e-01
-1.46432713e-01 7.82307535e-02 -1.01602590e+00 1.75899637e+00
-2.29468539e-01 2.91002929e-01 -4.19807024e-02 -1.01685464e+00
1.02047551e+00 -1.37102306e-01 5.16295880e-02 -1.37018132e+00
2.74318993e-01 1.06765725e-01 -2.73682475e-02 -5.30405402e-01
5.78434169e-01 6.24508679e-01 1.08828738e-01 3.32867414e-01
-5.27370125e-02 4.82203364e-01 3.73057611e-02 1.25244007e-01
5.81434608e-01 1.10774916e-02 1.39333740e-01 -4.79699165e-01
7.57279158e-01 -2.40870751e-02 5.86014926e-01 8.83815408e-01
-2.31476411e-01 4.04103935e-01 -2.65942037e-01 -4.41360205e-01
-9.76543486e-01 -8.11112642e-01 -4.04995084e-01 1.18298078e+00
1.08549094e+00 1.40063418e-02 -7.69660652e-01 -5.89729607e-01
1.13169223e-01 5.57772696e-01 -7.04491615e-01 -3.44485819e-01
-7.00840414e-01 -6.76414609e-01 5.13355434e-01 6.40447438e-01
9.56485569e-01 -1.14480352e+00 -1.52911037e-01 -1.86773375e-01
-2.22452566e-01 -8.18165123e-01 -9.18896377e-01 -2.83188194e-01
-4.27569896e-01 -9.48914886e-01 -1.03460419e+00 -1.19710433e+00
7.90189505e-01 7.85677135e-01 8.66783559e-01 2.67209351e-01
-4.50885028e-01 3.18938255e-01 -4.38120067e-01 -4.87598032e-01
2.48330142e-02 2.75653362e-01 3.24445128e-01 4.47231978e-02
6.27240181e-01 -2.36624748e-01 -9.42161322e-01 9.90694165e-01
-7.65338242e-01 1.72676802e-01 9.44675267e-01 1.08892238e+00
4.54804152e-01 -3.32107216e-01 7.30198383e-01 -2.00642526e-01
7.58008003e-01 -7.73789406e-01 -5.92439175e-01 6.50349081e-01
-8.72107923e-01 1.38759047e-01 2.48745099e-01 -7.03167081e-01
-1.20475960e+00 -4.30569828e-01 2.41593346e-01 -7.98146784e-01
-1.22159421e-01 4.35407192e-01 -1.88676611e-01 -2.76532829e-01
5.59870064e-01 4.97547448e-01 -7.15075284e-02 -7.85909057e-01
1.29344359e-01 7.92713106e-01 6.92982793e-01 -6.32527947e-01
8.19604218e-01 2.55115908e-02 -2.92047650e-01 -2.20399350e-01
-5.75191557e-01 -7.05550432e-01 -5.58952928e-01 -5.66235855e-02
6.60963416e-01 -1.09077239e+00 -7.45943666e-01 5.76341152e-01
-1.31308019e+00 -1.73681732e-02 2.49138176e-01 4.24438357e-01
-3.47830690e-02 2.92314827e-01 -4.59680349e-01 -6.43987954e-01
-5.43103874e-01 -1.21965301e+00 1.11729288e+00 7.33880579e-01
1.17863394e-01 -5.99299610e-01 -2.04976097e-01 4.83497083e-01
7.31640041e-01 -5.94258785e-01 7.75214672e-01 -9.47956860e-01
-6.89530075e-01 -1.88094154e-01 -8.04192841e-01 5.73734306e-02
-5.58111072e-02 -5.69055557e-01 -1.00428545e+00 -3.67911518e-01
-1.19801089e-01 -1.72679082e-01 7.20834792e-01 2.07111180e-01
1.50068831e+00 -2.10976720e-01 -7.71725535e-01 7.23785937e-01
1.04611349e+00 4.76690680e-01 6.10719442e-01 8.71632874e-01
6.52570784e-01 6.75464809e-01 8.14245462e-01 9.34063867e-02
7.17976630e-01 1.03529060e+00 3.53999197e-01 -3.25090647e-01
-2.23572895e-01 -2.10013762e-01 1.16298138e-03 6.17846549e-01
6.13631420e-02 -4.16797489e-01 -1.05989575e+00 7.31923938e-01
-2.01786327e+00 -7.93575346e-01 1.87932774e-01 1.97070205e+00
7.28070915e-01 -4.52015281e-01 -1.20480314e-01 -4.51697767e-01
1.11092412e+00 -1.50888458e-01 -9.61253881e-01 1.28816292e-01
-2.97826648e-01 -1.25071898e-01 3.96769553e-01 3.07462906e-04
-1.18504322e+00 9.34995890e-01 5.51616907e+00 1.15510869e+00
-9.78086352e-01 1.14808694e-01 5.10451078e-01 2.23821010e-02
-2.54817694e-01 -3.18355352e-01 -9.20408964e-01 5.97700179e-01
4.27243024e-01 -4.13625538e-01 3.64288062e-01 1.03791714e+00
-1.32012740e-01 2.15401798e-01 -1.13194990e+00 1.36772764e+00
2.86764979e-01 -1.16295159e+00 3.26830804e-01 -1.05420448e-01
5.45391440e-01 9.83868316e-02 3.84270638e-01 3.50691944e-01
-7.07693771e-03 -9.12295103e-01 6.97284460e-01 3.86589438e-01
7.83947229e-01 -3.66683841e-01 9.03788745e-01 3.41994762e-01
-1.29325175e+00 -2.89751172e-01 -4.11973208e-01 5.45391202e-01
-1.11986794e-01 -1.74371541e-01 -5.70896685e-01 8.68562222e-01
1.00609756e+00 5.31428576e-01 -9.65654969e-01 1.49114287e+00
2.49035638e-02 -7.54665909e-03 -1.99257776e-01 -4.33343500e-02
6.93098232e-02 -1.38277173e-01 5.17326653e-01 1.16420805e+00
3.60797733e-01 9.96992886e-02 -6.47474602e-02 9.14394259e-01
-1.69847175e-01 1.21843576e-01 -3.83545578e-01 4.75088060e-01
1.00127792e+00 9.16364610e-01 -2.75009036e-01 -2.63334662e-01
-2.17269897e-01 1.09324932e+00 5.05136013e-01 6.69950187e-01
-7.21073627e-01 -2.84536451e-01 8.11765909e-01 -2.03457266e-01
1.54330790e-01 1.36661649e-01 -1.94365963e-01 -1.17568135e+00
2.38886431e-01 -7.50750065e-01 5.68442345e-01 -9.62834001e-01
-1.65994287e+00 7.27922380e-01 2.51603752e-01 -1.28833544e+00
1.18774928e-01 -3.22941899e-01 -6.67745352e-01 1.18764949e+00
-1.88624954e+00 -1.23404682e+00 -5.17077804e-01 7.47186244e-01
7.45058417e-01 -4.57957596e-01 5.99529207e-01 7.16746390e-01
-9.33264196e-01 1.29995894e+00 2.21766442e-01 2.48084694e-01
9.61011946e-01 -7.52121031e-01 5.46405196e-01 8.09897602e-01
-1.02057740e-01 9.93834853e-01 4.32035387e-01 -5.64354777e-01
-1.57149720e+00 -1.00099897e+00 6.47824109e-01 -3.39342177e-01
4.10070300e-01 -1.66908428e-01 -1.07993555e+00 5.46466172e-01
2.24703848e-01 -1.24212250e-01 4.05882657e-01 -4.16819565e-02
-5.85612953e-01 -2.19698697e-01 -1.01974225e+00 6.61853552e-01
1.43860722e+00 -5.12648106e-01 -5.66763580e-01 1.24696158e-01
5.88153005e-01 -5.71744978e-01 -5.70043683e-01 5.65584779e-01
5.20272791e-01 -6.47740185e-01 1.22963691e+00 -4.97297823e-01
4.14101072e-02 -4.70797747e-01 8.59191269e-02 -1.27944350e+00
-6.70983315e-01 -1.72854260e-01 1.51236698e-01 1.29642522e+00
2.04623878e-01 -9.02177513e-01 4.40920502e-01 7.73313940e-01
-2.57693112e-01 -6.32720530e-01 -9.73460913e-01 -1.03158081e+00
-5.70515692e-02 1.07442886e-01 1.01955295e+00 1.05342412e+00
-2.38244310e-01 2.61093974e-01 -2.12867469e-01 3.83401036e-01
6.03840828e-01 2.29812279e-01 7.44628906e-01 -1.08796597e+00
4.26110514e-02 -9.19953287e-01 -4.51619536e-01 -1.35146976e+00
3.80094290e-01 -9.52460408e-01 -1.74250677e-01 -1.27245939e+00
4.90213901e-01 -9.10021663e-01 -6.94161057e-01 5.62303305e-01
-4.54547554e-01 1.08328313e-01 2.11190209e-01 7.85501480e-01
-6.37315154e-01 8.57043803e-01 1.21824419e+00 -2.62108743e-01
-4.45577409e-03 -1.61657125e-01 -8.65292013e-01 1.36061981e-01
7.68721581e-01 -2.43685603e-01 -4.95650977e-01 -8.96399856e-01
-1.77806228e-01 -2.98493713e-01 4.26584631e-01 -6.22740030e-01
7.15762913e-01 2.58954354e-02 5.78891456e-01 -6.51779473e-01
2.31740400e-01 -6.54307842e-01 3.67000878e-01 7.73963779e-02
-4.31339294e-01 3.52730960e-01 3.94330800e-01 4.49136853e-01
-4.10242051e-01 -9.56489593e-02 5.16170800e-01 5.16736209e-02
-9.86800492e-01 4.98743236e-01 3.17831129e-01 -3.70102942e-01
1.04664338e+00 -2.32403651e-01 -7.66875029e-01 -8.33873898e-02
-1.78544164e-01 9.46571469e-01 3.94229621e-01 8.60133469e-01
6.79885983e-01 -1.49535191e+00 -6.53992772e-01 2.79124141e-01
4.46021497e-01 -3.13044828e-03 6.05550468e-01 7.98067153e-01
-7.40210041e-02 5.44531107e-01 -1.81008592e-01 -5.13520718e-01
-1.32744050e+00 7.83599615e-01 4.14733171e-01 4.12455276e-02
-3.21763456e-01 1.22320187e+00 7.87358105e-01 -4.86841738e-01
4.98743236e-01 3.30983490e-01 -4.06715184e-01 -1.08225876e-02
7.74244785e-01 2.34433532e-01 -2.25866027e-02 -7.40167558e-01
-5.61518669e-01 7.37336278e-01 -5.43418109e-01 3.95694584e-01
9.98452067e-01 -4.16452497e-01 -9.17274430e-02 -1.56858817e-01
1.07322681e+00 -5.28688729e-01 -8.16041708e-01 -5.84661365e-01
2.61028800e-02 -6.35362923e-01 -9.07764062e-02 -9.29630339e-01
-8.84264708e-01 5.57338476e-01 8.96463454e-01 -1.06402375e-01
1.24307799e+00 2.26010531e-01 6.47390246e-01 4.97635484e-01
4.48095649e-01 -8.79329801e-01 -7.65242130e-02 2.55934536e-01
1.19635594e+00 -1.53212392e+00 -2.22501829e-01 -2.33041689e-01
-7.34507024e-01 6.81574345e-01 1.15599275e+00 1.31677032e-01
4.66452241e-01 -2.60281950e-01 1.10803843e-01 -1.97944403e-01
-4.72940743e-01 -2.40248010e-01 5.50746679e-01 4.29986447e-01
7.97396749e-02 -1.23440370e-01 -3.96319985e-01 5.48203647e-01
-1.87902655e-02 -2.54990965e-01 -1.77713871e-01 7.02951372e-01
1.91654102e-03 -1.11664569e+00 -5.07309198e-01 2.93358952e-01
-3.71632986e-02 -2.00136423e-01 -3.41949075e-01 7.32917190e-01
-3.32836173e-02 8.47711444e-01 1.11182757e-01 -5.05229354e-01
3.44091564e-01 -7.16798604e-02 2.92513520e-01 -2.43188277e-01
-4.94542658e-01 -2.00724632e-01 -1.89410284e-01 -4.57398385e-01
-3.79331797e-01 -5.70579469e-01 -8.10160518e-01 -2.99432814e-01
-6.81380332e-01 3.48163366e-01 8.16775262e-01 6.74806356e-01
9.40824151e-01 1.52560666e-01 7.04808414e-01 -1.01159346e+00
-6.07345045e-01 -1.16790617e+00 -2.95623720e-01 3.74854922e-01
3.35911155e-01 -1.02529669e+00 -1.72120839e-01 -2.27439031e-01] | [14.782042503356934, 0.9354726076126099] |
ef4dc84d-5ece-4890-8b42-1cbe062e56a2 | csdn-cross-modal-shape-transfer-dual | 2208.00751 | null | https://arxiv.org/abs/2208.00751v2 | https://arxiv.org/pdf/2208.00751v2.pdf | CSDN: Cross-modal Shape-transfer Dual-refinement Network for Point Cloud Completion | How will you repair a physical object with some missings? You may imagine its original shape from previously captured images, recover its overall (global) but coarse shape first, and then refine its local details. We are motivated to imitate the physical repair procedure to address point cloud completion. To this end, we propose a cross-modal shape-transfer dual-refinement network (termed CSDN), a coarse-to-fine paradigm with images of full-cycle participation, for quality point cloud completion. CSDN mainly consists of "shape fusion" and "dual-refinement" modules to tackle the cross-modal challenge. The first module transfers the intrinsic shape characteristics from single images to guide the geometry generation of the missing regions of point clouds, in which we propose IPAdaIN to embed the global features of both the image and the partial point cloud into completion. The second module refines the coarse output by adjusting the positions of the generated points, where the local refinement unit exploits the geometric relation between the novel and the input points by graph convolution, and the global constraint unit utilizes the input image to fine-tune the generated offset. Different from most existing approaches, CSDN not only explores the complementary information from images but also effectively exploits cross-modal data in the whole coarse-to-fine completion procedure. Experimental results indicate that CSDN performs favorably against ten competitors on the cross-modal benchmark. | ['Jing Qin', 'Jun Wang', 'Mingqiang Wei', 'Honghua Chen', 'Haoran Xie', 'Liangliang Nan', 'Zhe Zhu'] | 2022-08-01 | null | null | null | null | ['point-cloud-completion'] | ['computer-vision'] | [ 8.81856605e-02 1.59015819e-01 3.90482008e-01 -8.14805622e-04
-1.06681490e+00 -5.90110004e-01 4.60023701e-01 -9.59189609e-02
1.05694816e-01 3.03989738e-01 -1.07790604e-01 1.38641730e-01
-2.02913538e-01 -9.90803957e-01 -1.08511901e+00 -7.22016513e-01
4.02039737e-01 7.04183280e-01 2.69235522e-01 -4.09082025e-01
4.41601872e-01 9.40067470e-01 -1.55571556e+00 2.52217174e-01
9.85188723e-01 1.04071093e+00 4.06865418e-01 3.85896236e-01
-1.53140157e-01 -3.79379541e-02 -1.51700750e-01 -3.50535095e-01
5.07327080e-01 2.05227792e-01 -7.83454716e-01 3.94687563e-01
5.13376534e-01 -1.21220656e-01 -2.49589682e-01 1.02813804e+00
2.96228170e-01 1.90367594e-01 4.86805290e-01 -9.02819693e-01
-8.63482475e-01 -5.04618622e-02 -8.75926197e-01 -3.00676107e-01
3.71497273e-01 5.13242602e-01 7.38693297e-01 -1.56083286e+00
6.67967498e-01 1.34539926e+00 8.28710914e-01 3.69398355e-01
-1.47213674e+00 -5.09138584e-01 8.55565444e-02 -2.99497128e-01
-1.63631022e+00 -2.40890101e-01 1.04224455e+00 -4.31386858e-01
7.11994708e-01 1.07023574e-01 4.70204502e-01 5.13320982e-01
6.45530447e-02 2.41770983e-01 6.67494476e-01 -1.53776839e-01
2.82726213e-02 -3.10652703e-01 -2.26683229e-01 6.96362257e-01
-1.82099119e-02 4.18319762e-01 -2.21048981e-01 -3.22406888e-01
1.10800934e+00 2.26971462e-01 -4.10645187e-01 -7.32567251e-01
-1.32305956e+00 5.55284798e-01 9.17727411e-01 7.11870790e-02
-4.75771904e-01 9.58398879e-02 -1.34646103e-01 7.09407702e-02
4.96723086e-01 4.50307786e-01 -5.26006699e-01 2.73450077e-01
-7.92574406e-01 3.73997837e-01 3.28732133e-01 1.07895553e+00
1.32591033e+00 -6.42175823e-02 -6.83407485e-02 6.55992270e-01
1.96537226e-01 5.82943499e-01 -1.42008960e-01 -1.10713434e+00
7.75700033e-01 9.16823089e-01 2.59314448e-01 -1.05561006e+00
-2.43120447e-01 -3.02452713e-01 -1.06366575e+00 5.41105747e-01
8.46063048e-02 1.54285178e-01 -1.10936308e+00 1.37527156e+00
6.54783189e-01 4.06966418e-01 -1.37848169e-01 9.00849223e-01
8.61186743e-01 6.59441888e-01 -1.59478962e-01 6.41253740e-02
1.15378881e+00 -8.82765055e-01 -3.19890715e-02 -4.82115149e-02
1.69069543e-01 -9.32844639e-01 1.03739774e+00 2.13829026e-01
-1.45258117e+00 -8.69879663e-01 -8.77368927e-01 -3.40154380e-01
-1.92645252e-01 1.96459010e-01 2.25734293e-01 -1.31584808e-01
-1.05729926e+00 8.22341502e-01 -6.17400825e-01 8.33958089e-02
5.57878554e-01 3.21842730e-01 -5.89929819e-01 -3.01270396e-01
-6.43340945e-01 3.39843333e-01 2.51217574e-01 2.45422691e-01
-6.83182955e-01 -1.39749610e+00 -8.70083392e-01 2.38558501e-01
3.61423165e-01 -1.24667704e+00 7.17692316e-01 -6.44811869e-01
-1.01653314e+00 7.38094270e-01 -4.20880541e-02 1.45867541e-01
3.89281750e-01 1.48200795e-01 -3.42767797e-02 4.25782502e-02
2.43031427e-01 9.40689147e-01 1.19807351e+00 -1.84514928e+00
-7.10682154e-01 -4.45867717e-01 -4.87707816e-02 2.12804437e-01
5.08488238e-01 -5.55489242e-01 -8.48856747e-01 -8.77453148e-01
4.75967020e-01 -8.40279341e-01 -3.11943233e-01 2.10624337e-01
-3.37836534e-01 -1.83931947e-01 9.74187732e-01 -4.98936296e-01
7.15413153e-01 -2.65975904e+00 3.73375863e-01 4.90028203e-01
3.90253156e-01 -1.54470317e-02 -4.89910066e-01 2.74298459e-01
-2.99346924e-01 -3.88379395e-02 -5.60569882e-01 -8.79817128e-01
-9.34548154e-02 3.08354884e-01 -4.40082550e-01 3.92881393e-01
4.50155586e-01 9.87297773e-01 -8.37000132e-01 -1.00879706e-01
4.00132149e-01 6.18479967e-01 -7.51438439e-01 3.38603735e-01
-2.49777481e-01 6.69764996e-01 -4.59259033e-01 6.98447168e-01
1.43986511e+00 -3.53153825e-01 -3.98688734e-01 -6.94685459e-01
-2.42544517e-01 -3.48920614e-01 -1.50458241e+00 2.02433372e+00
-2.72419214e-01 -1.29038051e-01 3.54390442e-01 -5.04813373e-01
9.92092609e-01 4.57295701e-02 6.80477619e-01 -5.67342162e-01
-2.04256713e-01 1.69273987e-01 -5.19757032e-01 -2.62138546e-01
7.37074673e-01 -1.21761836e-01 -1.65741902e-03 1.26432657e-01
-1.60861149e-01 -6.08428597e-01 -1.85755730e-01 2.65514374e-01
8.22129607e-01 2.26264149e-01 -1.66447461e-01 -5.37343062e-02
6.69165730e-01 6.98550642e-02 5.61415195e-01 2.18043521e-01
4.46729302e-01 1.46475720e+00 1.66080117e-01 -5.25343955e-01
-1.27343309e+00 -1.31918728e+00 8.85779411e-02 6.07692003e-01
5.32902479e-01 -2.69178092e-01 -6.73050404e-01 -5.81194162e-01
2.54132658e-01 4.54903007e-01 -6.96553349e-01 -1.86249405e-01
-7.33658552e-01 -1.36048064e-01 9.15601179e-02 5.12152731e-01
4.06514734e-01 -1.19157147e+00 -2.52614710e-02 1.03644006e-01
-2.93199867e-01 -1.00561106e+00 -9.24629152e-01 -2.34071657e-01
-9.39701796e-01 -1.21536028e+00 -6.34288013e-01 -7.92764902e-01
1.00768280e+00 5.32113433e-01 1.14612949e+00 5.58402300e-01
-1.60238802e-01 3.54298890e-01 -2.33717710e-01 1.82150006e-01
-2.39280671e-01 -1.85502563e-02 -1.25447333e-01 3.33843470e-01
-4.07082021e-01 -9.42260325e-01 -7.30050981e-01 3.42983335e-01
-1.21691096e+00 1.97294340e-01 7.80818880e-01 7.47051477e-01
1.19633055e+00 1.12240918e-01 2.18738779e-01 -7.06171989e-01
1.65727288e-01 -4.40243930e-01 -5.97711682e-01 1.47002861e-01
-5.10758340e-01 1.01825930e-01 4.52177376e-01 -1.58279151e-01
-9.28276777e-01 2.91640699e-01 -2.62103140e-01 -1.10627294e+00
-9.37056690e-02 1.88046977e-01 -4.82002348e-01 -3.55673611e-01
3.75049770e-01 2.20386684e-01 -6.56656995e-02 -8.55639875e-01
4.81464088e-01 -7.42186382e-02 1.00863576e+00 -7.68393159e-01
1.32147646e+00 7.50374973e-01 8.85462165e-02 -3.89909446e-01
-6.13408148e-01 -5.05191386e-01 -7.79714823e-01 -2.50784121e-02
8.33500862e-01 -8.04058850e-01 -6.90418065e-01 5.28574347e-01
-1.42569983e+00 -1.86598301e-01 -6.44961655e-01 -3.79948616e-02
-6.93783581e-01 4.95075196e-01 -3.30813289e-01 -1.88404933e-01
-4.34569508e-01 -1.12983501e+00 1.71575201e+00 1.88555062e-01
3.79245251e-01 -7.16620147e-01 7.24360794e-02 4.68335807e-01
1.27917409e-01 3.52309078e-01 9.30904984e-01 1.21731490e-01
-1.15443957e+00 1.52673855e-01 -4.94978935e-01 2.94228077e-01
1.37465537e-01 -1.94007438e-03 -7.44576693e-01 -5.44796407e-01
-1.20361194e-01 2.93121263e-02 5.57779968e-01 2.48931244e-01
1.29747391e+00 -1.27412781e-01 -3.44422072e-01 1.07748544e+00
1.61988056e+00 -4.41825658e-01 9.96283531e-01 -3.92962731e-02
1.21377981e+00 3.80316913e-01 6.83723032e-01 4.69112158e-01
5.28574407e-01 7.78326571e-01 8.01045001e-01 -3.63897055e-01
-4.73744631e-01 -4.17309254e-01 9.37678367e-02 7.68319666e-01
-2.93587089e-01 1.31147400e-01 -7.32436955e-01 5.31978726e-01
-1.73484302e+00 -7.24763989e-01 -2.98967838e-01 2.11161375e+00
5.74328244e-01 -8.70525986e-02 -2.13486314e-01 -1.46465123e-01
7.72246540e-01 2.86261458e-02 -4.33820516e-01 2.73625832e-03
-1.50993466e-01 4.54204172e-01 3.24949890e-01 6.70894980e-01
-7.74430573e-01 8.81228745e-01 4.77524662e+00 1.20781374e+00
-8.63209546e-01 7.46560097e-02 3.86609465e-01 2.18904495e-01
-5.68502665e-01 2.25822955e-01 -6.08734548e-01 4.10950243e-01
2.04240799e-01 2.26748362e-01 5.93863904e-01 6.04890883e-01
1.29685789e-01 1.33947656e-02 -1.09104085e+00 1.16185009e+00
-8.23739246e-02 -1.79975164e+00 2.68670976e-01 1.71508327e-01
9.93697107e-01 1.22885689e-01 6.01537190e-02 1.22973464e-01
8.50841254e-02 -8.44780266e-01 8.57637882e-01 1.16631079e+00
1.18650854e+00 -9.47531343e-01 4.26690251e-01 4.49343920e-01
-1.69192302e+00 7.95411170e-02 -4.18963999e-01 3.88684481e-01
1.97312132e-01 6.12680912e-01 -3.49631846e-01 1.16256058e+00
7.71820128e-01 6.35962903e-01 -6.91571176e-01 9.76852119e-01
1.60067126e-01 -1.16611846e-01 -2.32738912e-01 9.74849820e-01
1.67721629e-01 -4.69371021e-01 7.19574809e-01 7.88873255e-01
5.63561022e-01 3.37190002e-01 3.11117113e-01 1.25151205e+00
-6.46146163e-02 -2.57238209e-01 -4.40981746e-01 5.88189542e-01
3.08504343e-01 1.53385246e+00 -4.91322696e-01 -2.40047112e-01
-1.90658316e-01 9.18281257e-01 4.33065593e-01 4.03550565e-01
-6.72602892e-01 -1.16405740e-01 7.03254938e-01 4.53204691e-01
5.56795299e-01 -2.33348355e-01 -5.76101601e-01 -9.89476800e-01
2.44129539e-01 -5.87956488e-01 2.32470244e-01 -1.27117324e+00
-1.36010528e+00 5.70224226e-01 -1.30630538e-01 -1.54931235e+00
1.38173163e-01 -1.62503779e-01 -7.70058990e-01 1.17677593e+00
-1.50918198e+00 -1.57415831e+00 -7.60449409e-01 7.63539255e-01
6.20303929e-01 1.39450490e-01 4.77529466e-01 3.12724531e-01
-1.32529721e-01 4.25828397e-01 -1.93464369e-01 -1.21878728e-01
5.19597948e-01 -9.87392902e-01 5.81681371e-01 7.88836896e-01
-1.60382867e-01 4.26662982e-01 2.81271338e-01 -9.23389435e-01
-1.39323318e+00 -1.40966427e+00 4.65517193e-01 -4.95226741e-01
2.36703604e-01 -5.28357327e-02 -1.26699221e+00 3.57268959e-01
-1.99162737e-01 3.93694133e-01 -2.54744887e-01 -4.69358474e-01
-2.64789045e-01 -7.81786591e-02 -1.29569590e+00 3.90532255e-01
1.11939430e+00 -3.55844587e-01 -5.43027639e-01 9.85684469e-02
1.03119934e+00 -7.30534792e-01 -1.02515042e+00 6.80207253e-01
1.12009734e-01 -8.26047838e-01 1.37207270e+00 -2.32186332e-01
4.76744741e-01 -8.41245532e-01 -1.57274544e-01 -1.33245766e+00
-7.62118042e-01 -5.21262050e-01 -1.98889747e-02 1.37138176e+00
1.33276686e-01 -4.15379465e-01 8.76525104e-01 5.18248975e-01
-7.65423775e-01 -8.43215168e-01 -9.62139010e-01 -4.51546848e-01
5.80398962e-02 -3.72487962e-01 1.12698138e+00 8.22921574e-01
-7.17321038e-01 5.14146984e-02 -5.88129088e-03 6.73358560e-01
6.88005984e-01 5.36667585e-01 9.98865366e-01 -1.26602769e+00
-1.76204562e-01 -2.64045954e-01 -1.85035244e-01 -1.11651719e+00
-1.18859306e-01 -8.83898616e-01 2.46569850e-02 -1.41027224e+00
4.25452292e-02 -8.49972546e-01 1.05315082e-01 3.35408688e-01
-2.73047298e-01 4.82217431e-01 2.16486797e-01 3.86559486e-01
-2.75210053e-01 8.17529440e-01 1.67786741e+00 -2.03855857e-01
-2.55418986e-01 9.66871306e-02 -7.44074404e-01 6.84450805e-01
3.58542681e-01 -3.20713699e-01 -1.57996595e-01 -5.23120880e-01
2.64137447e-01 1.61459282e-01 8.04088712e-01 -1.04693556e+00
3.10317576e-01 -3.84278983e-01 4.08267170e-01 -1.09352565e+00
5.82231581e-01 -1.06189370e+00 7.39709318e-01 1.05268164e-02
2.85826087e-01 1.79551840e-01 1.03363290e-01 7.05707490e-01
-5.09951822e-02 -6.67760754e-03 9.20404673e-01 -1.23249739e-01
-5.54632068e-01 9.31167662e-01 4.74267602e-01 -7.53419176e-02
8.76972020e-01 -4.16364372e-01 -1.94718525e-01 8.95122066e-02
-1.03412664e+00 2.66700596e-01 1.01506972e+00 4.15200472e-01
1.05616021e+00 -1.60335386e+00 -7.23867357e-01 6.58506334e-01
3.33165258e-01 8.75976861e-01 7.36589491e-01 7.16021836e-01
-5.11204898e-01 9.85100958e-03 4.26764823e-02 -7.69350290e-01
-9.13530767e-01 8.40367913e-01 3.75053942e-01 -1.88794792e-01
-9.92086112e-01 6.71492934e-01 7.47064948e-01 -6.23259842e-01
-4.04299982e-02 -4.96133000e-01 7.78437480e-02 -2.24937230e-01
3.10500473e-01 3.51712644e-01 3.98964286e-01 -7.89825141e-01
-6.23276904e-02 1.07885337e+00 1.35175377e-01 1.46942317e-01
1.47152960e+00 -2.60635316e-01 -2.69735813e-01 2.58383900e-03
1.17319691e+00 2.68252790e-01 -1.57173371e+00 -3.78359348e-01
-4.10836995e-01 -6.86542392e-01 -1.55796230e-01 -5.23723900e-01
-1.49703133e+00 5.55271268e-01 3.65437895e-01 -5.45005538e-02
1.17158186e+00 2.27328613e-01 9.99698699e-01 -1.33803114e-01
4.38315541e-01 -8.27887833e-01 1.42246738e-01 4.65014875e-01
1.39931941e+00 -9.54753518e-01 -2.49346485e-03 -7.46342540e-01
-3.71129066e-01 1.00659084e+00 7.22222984e-01 -4.43789631e-01
6.73125148e-01 2.90936269e-02 -4.02263433e-01 -7.61773646e-01
-4.63486522e-01 2.30165981e-02 5.89016736e-01 6.95436537e-01
-4.44367647e-01 9.41246524e-02 3.22484821e-01 5.47294617e-01
-1.06403165e-01 -1.01369984e-01 2.76423663e-01 4.90388572e-01
-2.23893449e-01 -1.04523909e+00 -7.95057237e-01 3.19382399e-01
2.45058924e-01 -2.49559656e-02 -1.53881818e-01 5.82262278e-01
6.47544444e-01 5.62141180e-01 2.41568357e-01 -4.76216197e-01
8.13947558e-01 -3.87443095e-01 2.91397482e-01 -7.03553855e-01
-6.90885186e-01 3.22299339e-02 -3.71374160e-01 -8.18089068e-01
-1.73378363e-01 -6.09865844e-01 -1.42839968e+00 -5.68558753e-01
-1.54266894e-01 -2.43414920e-02 4.04592693e-01 7.76870251e-01
7.65584052e-01 4.80658203e-01 9.41423953e-01 -1.56606853e+00
-2.22076148e-01 -6.10969543e-01 -3.45486850e-01 6.05719209e-01
3.91982526e-01 -8.49307775e-01 -4.08598781e-01 -5.32440562e-03] | [8.39142894744873, -3.624077081680298] |
6b8cb772-c72f-4b5e-af05-1905287a15e7 | improved-static-hand-gesture-classification | 2305.02039 | null | https://arxiv.org/abs/2305.02039v1 | https://arxiv.org/pdf/2305.02039v1.pdf | Improved Static Hand Gesture Classification on Deep Convolutional Neural Networks using Novel Sterile Training Technique | In this paper, we investigate novel data collection and training techniques towards improving classification accuracy of non-moving (static) hand gestures using a convolutional neural network (CNN) and frequency-modulated-continuous-wave (FMCW) millimeter-wave (mmWave) radars. Recently, non-contact hand pose and static gesture recognition have received considerable attention in many applications ranging from human-computer interaction (HCI), augmented/virtual reality (AR/VR), and even therapeutic range of motion for medical applications. While most current solutions rely on optical or depth cameras, these methods require ideal lighting and temperature conditions. mmWave radar devices have recently emerged as a promising alternative offering low-cost system-on-chip sensors whose output signals contain precise spatial information even in non-ideal imaging conditions. Additionally, deep convolutional neural networks have been employed extensively in image recognition by learning both feature extraction and classification simultaneously. However, little work has been done towards static gesture recognition using mmWave radars and CNNs due to the difficulty involved in extracting meaningful features from the radar return signal, and the results are inferior compared with dynamic gesture classification. This article presents an efficient data collection approach and a novel technique for deep CNN training by introducing ``sterile'' images which aid in distinguishing distinct features among the static gestures and subsequently improve the classification accuracy. Applying the proposed data collection and training methods yields an increase in classification rate of static hand gestures from $85\%$ to $93\%$ and $90\%$ to $95\%$ for range and range-angle profiles, respectively. | ['Murat Torlak', 'Yiorgos Makris', 'Richard Willis', 'Shiva Thiagarajan', 'Josiah Smith'] | 2023-05-03 | null | null | null | null | ['gesture-recognition'] | ['computer-vision'] | [ 4.80101675e-01 -4.23618495e-01 -1.07924473e-02 -4.59442705e-01
-5.00787079e-01 -4.63708460e-01 4.74904090e-01 -5.79451442e-01
-8.03359926e-01 6.27803206e-01 -2.69103289e-01 -2.97432661e-01
-6.02723598e-01 -7.86720276e-01 -9.55446288e-02 -1.14603531e+00
-3.20143312e-01 1.80522665e-01 -2.17133567e-01 -1.58084631e-02
1.16969928e-01 9.82622683e-01 -1.43838990e+00 3.15186679e-02
2.91310638e-01 1.31483698e+00 -2.25975066e-02 7.75437593e-01
4.26578894e-02 1.53463379e-01 -9.83193874e-01 1.52028948e-02
5.47350943e-01 -2.47861683e-01 -5.24516106e-02 -4.81702328e-01
4.34375137e-01 -5.93839109e-01 -5.51546454e-01 6.65517390e-01
1.04262161e+00 1.13713391e-01 7.07130313e-01 -8.59222531e-01
-3.24752390e-01 6.66070059e-02 -5.26768684e-01 2.24761799e-01
3.42129767e-01 4.52148914e-01 2.93702245e-01 -4.72765118e-01
4.80030090e-01 5.88050604e-01 5.81028819e-01 8.04123342e-01
-4.83479679e-01 -1.22627079e+00 -3.60154837e-01 2.80387759e-01
-1.38230348e+00 -2.64204592e-01 8.47420871e-01 -2.90255815e-01
1.11853576e+00 4.76277947e-01 6.84161246e-01 1.26310170e+00
4.16888833e-01 4.59616214e-01 9.20522511e-01 -3.58571857e-01
1.08427435e-01 -3.11339527e-01 1.24628082e-01 6.62662268e-01
2.46489719e-01 5.67577243e-01 -5.16636252e-01 8.84735361e-02
1.00626743e+00 4.24541146e-01 -5.06828129e-01 -2.16712826e-04
-1.07175064e+00 4.21860844e-01 6.83568954e-01 8.93763483e-01
-6.33495092e-01 4.04288173e-01 -6.13639690e-02 4.30847079e-01
1.47942398e-02 5.14657378e-01 -3.78178239e-01 -3.80724788e-01
-9.47891593e-01 1.54904068e-01 5.65023482e-01 8.19400191e-01
2.68390059e-01 4.23139870e-01 8.66742805e-02 4.24944967e-01
3.42059046e-01 1.15000534e+00 3.07900399e-01 -3.08284700e-01
5.63115776e-01 3.63054067e-01 4.94472161e-02 -1.01506841e+00
-9.16036367e-01 -6.73465729e-01 -1.14650869e+00 4.28060651e-01
4.27103162e-01 -4.25429970e-01 -1.39380515e+00 1.38613522e+00
-2.87829135e-02 1.07500121e-01 1.44421399e-01 1.22499669e+00
1.08652961e+00 3.30319792e-01 2.48848237e-02 -2.11995497e-01
1.18060434e+00 -8.03821459e-02 -6.68873549e-01 2.72585475e-03
3.30614954e-01 -6.30771458e-01 5.69160819e-01 4.17648494e-01
-6.11722171e-01 -4.31121051e-01 -1.23188686e+00 5.81494868e-01
-3.50737602e-01 6.68587089e-02 8.18104565e-01 1.27231050e+00
-5.19836307e-01 3.09147209e-01 -1.02159643e+00 -2.98699230e-01
6.12864792e-01 9.33662772e-01 -9.55838487e-02 -6.04082905e-02
-8.70462298e-01 5.68515182e-01 -3.43505293e-01 5.04523516e-01
-2.06741720e-01 -5.42559683e-01 -4.44669247e-01 -1.41843095e-01
-1.71418428e-01 -2.14200288e-01 9.19964254e-01 -6.03080451e-01
-1.75864422e+00 4.00354028e-01 1.21855430e-01 -2.17460021e-01
2.65119791e-01 -3.15125495e-01 -7.33964443e-01 2.78681427e-01
-5.41887343e-01 3.59627306e-01 8.66519272e-01 -8.06178570e-01
-3.57015312e-01 -7.46468544e-01 -2.01083004e-01 -8.12899545e-02
-1.07088462e-01 3.37536596e-02 1.43715918e-01 -5.69851816e-01
7.44691908e-01 -8.27521682e-01 3.70691270e-02 -6.05845004e-02
-5.50380312e-02 1.18977427e-01 1.24002802e+00 -4.03002113e-01
7.81369746e-01 -1.75857723e+00 -2.29780301e-01 4.94872093e-01
5.67692593e-02 6.95496917e-01 -6.76475093e-02 2.72551358e-01
2.28182197e-01 -2.75617450e-01 -1.69274867e-01 4.90225166e-01
-2.19800264e-01 -1.10235959e-01 -7.82790259e-02 5.83035886e-01
-7.23995343e-02 1.23513877e+00 -4.75910664e-01 1.97672602e-02
4.59948808e-01 7.27196991e-01 -1.34652913e-01 5.81083819e-02
1.74204931e-01 8.57222915e-01 -5.45646906e-01 1.20100248e+00
7.85493493e-01 2.65609205e-01 7.34198391e-02 -3.68673772e-01
-1.40015692e-01 -2.23450154e-01 -9.15836990e-01 1.47881866e+00
-6.16841257e-01 1.01699495e+00 2.09414572e-01 -9.82557058e-01
1.21209168e+00 5.48790038e-01 7.80925691e-01 -1.30261433e+00
3.99528772e-01 2.05375120e-01 1.45929217e-01 -9.97443199e-01
1.44185513e-01 -5.94196916e-01 -2.05850787e-02 5.38619220e-01
-2.24609688e-01 1.10062003e-01 -5.29191673e-01 -6.65345907e-01
1.33993912e+00 5.36539853e-02 3.34037771e-03 2.15980217e-01
2.14879543e-01 -6.65599629e-02 1.37960136e-01 7.82711923e-01
-2.06581384e-01 4.31081325e-01 -5.20157754e-01 -7.26324558e-01
-3.07809323e-01 -1.07083523e+00 -2.52738863e-01 6.54499948e-01
2.13249207e-01 5.33341646e-01 -2.58127779e-01 -3.70314211e-01
-5.56513891e-02 2.23517045e-01 -1.42071798e-01 9.24620703e-02
-1.06716335e+00 -8.90652955e-01 8.79301190e-01 8.13416123e-01
8.94447207e-01 -1.28559661e+00 -1.63881826e+00 2.71724761e-01
2.17726991e-01 -1.06241202e+00 3.39295328e-01 4.60583121e-01
-9.39985693e-01 -1.11219597e+00 -8.73589337e-01 -7.67007530e-01
4.55584437e-01 1.80720732e-01 4.84128475e-01 -5.16579710e-02
-8.50028634e-01 3.49983901e-01 -5.32362223e-01 -6.54713690e-01
4.16642517e-01 1.24520488e-01 1.90485269e-01 -2.59720892e-01
7.65649259e-01 -7.85593927e-01 -8.46375465e-01 1.90177962e-01
-7.07126141e-01 -3.33334327e-01 1.13731992e+00 6.61320984e-01
-1.85417697e-01 -3.43027681e-01 6.96599424e-01 -4.85447824e-01
4.17532861e-01 9.71316993e-02 -5.31056225e-01 4.36777435e-02
-2.18655258e-01 -1.06488548e-01 4.41293776e-01 -4.80399072e-01
-8.91698897e-01 5.50729632e-02 -3.57525170e-01 -2.88845539e-01
-5.60460865e-01 4.34962273e-01 -1.26640707e-01 -5.10386109e-01
7.60343432e-01 2.87179500e-01 -1.07172027e-01 -3.12961340e-01
8.05489346e-02 1.11135387e+00 5.99301696e-01 -1.25673175e-01
1.04542649e+00 6.20898783e-01 8.74079615e-02 -1.35975075e+00
-2.47315932e-02 -5.66492498e-01 -7.36615837e-01 -4.01020467e-01
8.45263541e-01 -4.73584056e-01 -1.10067868e+00 5.91788888e-01
-9.81139541e-01 -2.91407257e-01 1.65664613e-01 9.77366209e-01
-2.16133714e-01 2.65985895e-02 -3.26321602e-01 -1.23800933e+00
-6.81744158e-01 -8.37353587e-01 1.03193700e+00 5.10807633e-01
-3.38458091e-01 -4.82369572e-01 -2.46560544e-01 2.00665146e-01
9.49363470e-01 5.96119463e-01 9.16268706e-01 -9.44267958e-02
-8.96000564e-01 -7.41557956e-01 -3.33212316e-01 -2.73817629e-01
3.11100602e-01 -5.58991432e-01 -1.11879301e+00 -2.78922349e-01
-1.26532167e-02 1.00693181e-02 7.40345597e-01 6.76365077e-01
7.88977802e-01 -3.38559132e-03 -5.78064203e-01 9.88825440e-01
1.53864050e+00 1.08984792e+00 8.06240082e-01 2.05717444e-01
5.84419489e-01 2.97655195e-01 3.51432621e-01 3.67304236e-01
-4.10233259e-01 7.38067985e-01 3.31123292e-01 -1.36908278e-01
1.27825558e-01 5.02415001e-01 5.75615875e-02 3.38605851e-01
-6.29856944e-01 -2.26286173e-01 -9.23977196e-01 2.39738822e-01
-1.27920294e+00 -1.12533879e+00 -5.75180762e-02 2.08258033e+00
1.84508845e-01 -3.03037614e-01 -1.96162999e-01 3.96364301e-01
2.81566292e-01 1.84889629e-01 -6.11790001e-01 -1.33489922e-01
1.99338738e-02 1.20542264e+00 5.70935905e-01 1.43241197e-01
-1.21162772e+00 6.68674946e-01 5.23323488e+00 1.55491203e-01
-1.88290679e+00 -6.03932403e-02 -1.23874962e-01 -4.83886331e-01
9.23694968e-02 -7.06006885e-01 -4.75418955e-01 2.10495237e-02
6.65483236e-01 6.24882042e-01 7.13574067e-02 5.06320357e-01
-9.27371457e-02 -2.34176278e-01 -8.13226521e-01 1.46139944e+00
4.53995690e-02 -1.17377162e+00 -2.30914399e-01 2.17044219e-01
3.94075423e-01 -1.04556717e-01 2.76550829e-01 2.22382262e-01
-3.23245823e-01 -1.28546047e+00 1.43329531e-01 4.89610523e-01
1.22329736e+00 -8.09981704e-01 1.05398464e+00 2.55027622e-01
-1.18149209e+00 -2.09273145e-01 -1.42882973e-01 -2.10018665e-01
-3.32323350e-02 3.97016555e-01 -7.08484232e-01 5.00027657e-01
7.14483619e-01 3.31264399e-02 1.34577051e-01 7.51371026e-01
5.07841446e-02 3.46747994e-01 -4.18028265e-01 -6.46121025e-01
2.81782538e-01 6.98343739e-02 5.24475217e-01 1.30369806e+00
4.55873996e-01 7.16412783e-01 -2.24732280e-01 5.63490927e-01
1.95430860e-01 -3.10438782e-01 -6.73782587e-01 -3.69797763e-03
1.17827833e-01 1.10106361e+00 -8.48848641e-01 1.06174983e-01
-2.35176846e-01 8.39688480e-01 -5.95530927e-01 6.14102542e-01
-6.43178880e-01 -9.67623234e-01 8.13006163e-01 2.74724774e-02
3.57669353e-01 -7.92595148e-01 -4.76395518e-01 -7.42612481e-01
1.72002211e-01 -3.37093949e-01 1.29082993e-01 -4.21293736e-01
-7.06811726e-01 7.10379839e-01 -1.10849790e-01 -1.44976199e+00
-5.50583422e-01 -9.48066294e-01 -3.93113136e-01 8.27077508e-01
-1.52744401e+00 -1.11576748e+00 -9.24081683e-01 4.30066258e-01
3.47427547e-01 -4.20398712e-01 1.04587817e+00 2.33361781e-01
-1.30836934e-01 8.11973035e-01 2.43398413e-01 5.08301437e-01
4.23224956e-01 -6.09257102e-01 1.57366380e-01 4.97470170e-01
1.10381439e-01 5.67969084e-01 2.57827759e-01 -5.32318115e-01
-1.96312189e+00 -8.25312257e-01 3.49372745e-01 -2.41269067e-01
-3.15934420e-02 -3.31420600e-01 -4.28642094e-01 2.38529116e-01
-1.11400560e-01 -5.48725016e-02 6.46585047e-01 -1.32331654e-01
-1.72749504e-01 -2.12185055e-01 -1.36303723e+00 4.29690480e-01
1.26308250e+00 -3.41391176e-01 -2.46843427e-01 -9.83882509e-03
1.32537214e-02 -5.08769214e-01 -5.18109977e-01 6.60516977e-01
1.34447086e+00 -7.84173012e-01 9.16834533e-01 -3.85590941e-01
9.17584896e-02 -5.07642189e-03 -3.28710556e-01 -9.13684011e-01
-1.65671870e-01 -3.72850657e-01 -3.79376151e-02 4.62614834e-01
1.56286120e-01 -6.75402105e-01 1.41275370e+00 2.77640879e-01
3.49112414e-02 -8.39122832e-01 -1.34254599e+00 -8.92854393e-01
-2.61453211e-01 -6.93273783e-01 5.27271688e-01 6.46608889e-01
-7.14270398e-02 -8.57503563e-02 -1.28888786e-01 3.26005429e-01
6.22920454e-01 3.74143213e-01 6.70791507e-01 -1.36986482e+00
-1.05763301e-01 -3.70398939e-01 -9.38424349e-01 -1.00290930e+00
-2.81768769e-01 -6.40709162e-01 1.39310583e-01 -1.48178267e+00
-3.19638222e-01 -4.71588939e-01 -1.54284701e-01 3.40101153e-01
5.15604377e-01 6.61914825e-01 -9.92579758e-03 2.92334616e-01
-5.35902753e-02 2.77721047e-01 1.01537609e+00 -3.70607287e-01
-4.29839432e-01 3.17456424e-01 -1.94725230e-01 3.59251797e-01
7.51448512e-01 -2.74185240e-01 -5.83507568e-02 -4.75028247e-01
-1.23772092e-01 2.98699200e-01 5.45526564e-01 -1.51077914e+00
4.88940269e-01 -7.07277209e-02 9.31970775e-01 -6.64482117e-01
4.73150045e-01 -1.07304990e+00 2.05397889e-01 8.60752881e-01
1.30896479e-01 -2.19753563e-01 8.37222338e-02 3.70555997e-01
-5.36628254e-02 1.68456659e-01 7.29435503e-01 -1.23643212e-01
-7.79641807e-01 1.59709305e-01 -6.04951501e-01 -6.63029492e-01
8.34392786e-01 -7.92502403e-01 -2.70614058e-01 -6.11477792e-01
-5.87900996e-01 -3.33089203e-01 -1.92802146e-01 3.42740744e-01
9.22181904e-01 -1.15380657e+00 -3.71015757e-01 6.39396906e-01
-6.61176145e-02 -1.96801007e-01 3.26435804e-01 8.27248931e-01
-7.11071670e-01 9.36263978e-01 -4.75022167e-01 -7.91098714e-01
-1.37987602e+00 -1.73697934e-01 3.66111130e-01 1.57255650e-01
-6.90173268e-01 9.54857588e-01 -3.38490218e-01 -1.84795991e-01
4.11526024e-01 -5.22233129e-01 -1.27356216e-01 -1.06708765e-01
5.06581426e-01 2.87364304e-01 3.24499577e-01 -3.50587547e-01
-6.98594809e-01 1.07739651e+00 2.30701938e-01 -3.64751995e-01
1.55421805e+00 4.92742032e-01 4.63072121e-01 -5.19278506e-03
1.11021125e+00 -2.76235074e-01 -9.68941927e-01 -5.56527264e-02
-4.24512140e-02 -4.06511545e-01 1.42224291e-02 -1.12719846e+00
-1.26687181e+00 1.15418661e+00 1.40874720e+00 -1.54493690e-01
1.22705805e+00 -2.08545193e-01 9.26619411e-01 9.27587211e-01
8.41039896e-01 -8.04885983e-01 3.80245857e-02 4.31200922e-01
7.78322697e-01 -9.69316721e-01 -1.36658959e-02 -1.61432892e-01
5.37947798e-03 1.39359486e+00 4.25567389e-01 -1.84167642e-02
8.75754833e-01 6.42901480e-01 7.04023957e-01 -6.03789628e-01
2.32114736e-02 -3.27204764e-01 2.96795934e-01 1.12348449e+00
6.48560166e-01 1.02255970e-01 -2.67466336e-01 5.46911359e-01
-3.27764839e-01 1.09741889e-01 -9.00707841e-02 1.32116902e+00
-3.75671118e-01 -7.40397155e-01 -4.70986158e-01 6.74814463e-01
-3.01392764e-01 2.37132177e-01 -4.13629204e-01 1.00265169e+00
8.50775316e-02 8.39199185e-01 1.57455936e-01 -8.61027539e-01
4.13139939e-01 2.80119807e-01 9.04325306e-01 -2.32186005e-01
-7.64702320e-01 1.10944703e-01 -2.15698570e-01 -4.39118057e-01
-6.18455350e-01 -3.72117072e-01 -1.39790881e+00 4.04752493e-02
-4.17754352e-01 -2.21560329e-01 1.16811597e+00 1.19003344e+00
2.27994636e-01 5.54398000e-01 3.92701656e-01 -1.12431669e+00
-3.04201514e-01 -1.07591999e+00 -5.74633181e-01 7.94539973e-03
2.92658627e-01 -7.00498104e-01 -6.86145350e-02 -3.94831985e-01] | [6.710053443908691, 0.25581711530685425] |
d48ec71c-33e3-46dc-997c-2c29da677907 | hyperspectral-image-super-resolution-via-deep-1 | 2006.10300 | null | https://arxiv.org/abs/2006.10300v2 | https://arxiv.org/pdf/2006.10300v2.pdf | Hyperspectral Image Super-resolution via Deep Progressive Zero-centric Residual Learning | This paper explores the problem of hyperspectral image (HSI) super-resolution that merges a low resolution HSI (LR-HSI) and a high resolution multispectral image (HR-MSI). The cross-modality distribution of the spatial and spectral information makes the problem challenging. Inspired by the classic wavelet decomposition-based image fusion, we propose a novel \textit{lightweight} deep neural network-based framework, namely progressive zero-centric residual network (PZRes-Net), to address this problem efficiently and effectively. Specifically, PZRes-Net learns a high resolution and \textit{zero-centric} residual image, which contains high-frequency spatial details of the scene across all spectral bands, from both inputs in a progressive fashion along the spectral dimension. And the resulting residual image is then superimposed onto the up-sampled LR-HSI in a \textit{mean-value invariant} manner, leading to a coarse HR-HSI, which is further refined by exploring the coherence across all spectral bands simultaneously. To learn the residual image efficiently and effectively, we employ spectral-spatial separable convolution with dense connections. In addition, we propose zero-mean normalization implemented on the feature maps of each layer to realize the zero-mean characteristic of the residual image. Extensive experiments over both real and synthetic benchmark datasets demonstrate that our PZRes-Net outperforms state-of-the-art methods to a \textit{significant} extent in terms of both 4 quantitative metrics and visual quality, e.g., our PZRes-Net improves the PSNR more than 3dB, while saving 2.3$\times$ parameters and consuming 15$\times$ less FLOPs. The code is publicly available at https://github.com/zbzhzhy/PZRes-Net . | ['Jie Chen', 'Huanqiang Zeng', 'Zhiyu Zhu', 'Junhui Hou', 'Jiantao Zhou'] | 2020-06-18 | null | null | null | null | ['hyperspectral-unmixing'] | ['computer-vision'] | [ 6.17457449e-01 -4.93200421e-01 1.93143904e-01 -2.54139662e-01
-1.12495208e+00 -3.10160398e-01 1.30479112e-01 -5.70848763e-01
-2.46087134e-01 7.79355466e-01 8.09518844e-02 5.94447963e-02
-5.71847856e-01 -1.15175080e+00 -7.45616019e-01 -1.13673246e+00
4.48143892e-02 -4.87783581e-01 -1.03900269e-01 -3.75773311e-01
-6.24351278e-02 5.60684264e-01 -1.74397278e+00 2.12081045e-01
1.19890761e+00 1.33020639e+00 5.33389509e-01 4.93696421e-01
3.41829747e-01 5.32577753e-01 -2.18248889e-02 2.36036003e-01
6.16725743e-01 -2.61414915e-01 -5.57808936e-01 2.89745539e-01
5.92057645e-01 -4.88327652e-01 -6.84586287e-01 1.60350645e+00
5.37407100e-01 4.90022629e-01 1.11261927e-01 -5.57064652e-01
-1.16482270e+00 3.76908123e-01 -1.16568851e+00 2.29027539e-01
-3.87326665e-02 3.38368595e-01 9.12437260e-01 -1.16233718e+00
2.16864735e-01 9.65946972e-01 6.07758403e-01 -4.20541242e-02
-1.45620537e+00 -8.79187167e-01 6.30076975e-02 1.23394884e-01
-1.79317999e+00 -1.60275504e-01 8.53174090e-01 -2.04710141e-01
6.41812861e-01 5.02833962e-01 4.31218058e-01 5.22604287e-01
-1.31473929e-01 3.52852643e-01 1.43608522e+00 -1.05793856e-01
-2.37307876e-01 -4.99187708e-01 7.16584399e-02 6.08112931e-01
2.55528539e-01 2.93710589e-01 -4.29982692e-01 2.91271329e-01
1.13630092e+00 3.38691950e-01 -9.03363407e-01 3.85930568e-01
-1.13804328e+00 5.15079141e-01 1.04742038e+00 4.47278798e-01
-6.05065525e-01 -1.39613584e-01 -8.23316947e-02 9.22881365e-02
6.71095788e-01 3.10027033e-01 -3.02522689e-01 7.62220681e-01
-9.96690452e-01 4.26636711e-02 -8.80837664e-02 6.16594613e-01
1.23556900e+00 3.75335455e-01 -1.40234187e-01 1.12122023e+00
1.39448801e-02 6.50154710e-01 2.40029261e-01 -1.15991569e+00
2.77054012e-01 4.37819660e-01 9.69980434e-02 -9.79884446e-01
-3.76618505e-01 -7.54516006e-01 -1.79346097e+00 2.95736343e-01
-2.35781938e-01 -9.13407877e-02 -9.72366273e-01 1.68568540e+00
7.17434660e-02 4.30030316e-01 1.39863387e-01 1.26407933e+00
1.06543911e+00 1.04985154e+00 -1.54141456e-01 -4.32396084e-01
1.26569629e+00 -8.86490226e-01 -4.18663621e-01 -1.81289181e-01
-1.40257597e-01 -5.70979774e-01 1.08030438e+00 2.76909232e-01
-1.16837692e+00 -8.74629021e-01 -1.06881344e+00 -1.66418269e-01
-1.79494396e-01 4.49938744e-01 3.81588757e-01 9.32023674e-02
-1.22753739e+00 8.34769845e-01 -4.93268132e-01 7.88911507e-02
5.31388044e-01 2.05648959e-01 -4.23578560e-01 -3.73421818e-01
-1.20381284e+00 2.42030263e-01 5.44237077e-01 3.95752132e-01
-5.34345984e-01 -1.04595697e+00 -8.89617980e-01 3.35829258e-01
4.11085606e-01 -4.45978612e-01 5.85612357e-01 -7.40856409e-01
-1.29359567e+00 6.00741386e-01 -1.56684190e-01 7.98369944e-02
-3.85843264e-03 2.54820625e-04 -7.30028689e-01 3.24265957e-01
1.35783970e-01 3.95515084e-01 8.61571848e-01 -1.37150335e+00
-7.94313312e-01 -4.82669681e-01 -9.03372765e-02 3.32376003e-01
-4.04129237e-01 -2.14125738e-01 -3.63894403e-01 -8.16902280e-01
4.71063673e-01 -5.48319042e-01 -2.05345735e-01 -1.73389733e-01
-3.73261005e-01 2.03670353e-01 6.99060678e-01 -9.07456458e-01
1.05487478e+00 -2.33606839e+00 1.33210778e-01 -2.18894221e-02
4.35592532e-01 3.48428816e-01 -3.63189459e-01 -4.38542739e-02
-4.53943789e-01 2.03677684e-01 -6.27284765e-01 1.65832322e-02
-2.86346138e-01 -2.19712004e-01 -4.08212185e-01 6.77482188e-01
1.77427784e-01 7.63340890e-01 -6.93075895e-01 1.03119919e-02
4.07466173e-01 9.90592599e-01 -1.89987123e-01 2.39055231e-01
2.35594928e-01 6.32585883e-01 -3.00671488e-01 6.94419742e-01
1.31910920e+00 -4.11762238e-01 -1.82690650e-01 -7.78247356e-01
-6.48052037e-01 -1.89467892e-01 -1.14063907e+00 1.66548705e+00
-4.45480019e-01 4.71739501e-01 3.63986105e-01 -9.29961085e-01
9.98563766e-01 1.14311077e-01 6.55569077e-01 -9.55286384e-01
-1.44535899e-01 2.41962895e-01 -5.23198247e-01 -1.66634426e-01
5.90888023e-01 -2.29991898e-01 2.79703081e-01 8.36250931e-02
4.43510376e-02 9.93687543e-04 -6.03392068e-03 -2.96127617e-01
6.10918462e-01 2.04668567e-02 9.37518701e-02 -3.95375133e-01
8.41356635e-01 -3.10473055e-01 6.86502337e-01 5.55341423e-01
7.56875798e-02 8.72559965e-01 -9.19822603e-02 -4.13725078e-01
-8.97966385e-01 -1.04014945e+00 -2.75901318e-01 9.00905490e-01
4.15103465e-01 1.80511966e-01 -3.79625171e-01 1.76512934e-02
-2.02124581e-01 4.52608079e-01 -4.85347152e-01 -1.89837739e-02
-4.55060363e-01 -1.01964557e+00 2.12839052e-01 3.36269170e-01
1.34486115e+00 -8.93506348e-01 -2.83865303e-01 1.06125876e-01
-4.02037591e-01 -1.03743672e+00 -3.33747953e-01 7.23346323e-02
-6.52779222e-01 -7.41609871e-01 -7.61001468e-01 -5.71393132e-01
4.19769466e-01 1.01132965e+00 8.04590940e-01 -1.07278191e-01
-4.76922184e-01 -9.29209404e-03 -1.66364521e-01 1.71306938e-01
2.90571064e-01 -1.95025563e-01 -3.17659438e-01 3.25998694e-01
-2.20455360e-02 -1.04272127e+00 -9.40792203e-01 1.28633142e-01
-1.29980671e+00 3.70540738e-01 6.94335461e-01 1.01353502e+00
1.02314091e+00 8.70125353e-01 3.02543133e-01 -4.93702918e-01
2.90764064e-01 -3.87504727e-01 -8.44565153e-01 1.47101164e-01
-4.48330194e-01 -3.01658392e-01 6.56280339e-01 -2.11499646e-01
-1.33834875e+00 1.17729967e-02 -8.59541446e-02 -6.07309997e-01
-2.31247723e-01 6.89354718e-01 -1.79021701e-01 -3.71499926e-01
5.80375254e-01 7.24999011e-01 -2.70875335e-01 -5.70334852e-01
2.26839438e-01 5.80102086e-01 1.04852068e+00 -2.68849164e-01
1.15643799e+00 7.45547056e-01 1.11585721e-01 -1.04698455e+00
-1.13013268e+00 -5.23155689e-01 -4.43284839e-01 -1.60505161e-01
9.09070790e-01 -1.38628900e+00 -8.43431592e-01 6.73457205e-01
-7.54024565e-01 -4.77191120e-01 -2.92483538e-01 3.70797634e-01
-3.90290141e-01 3.20993036e-01 -6.60950303e-01 -5.81185460e-01
-5.97392440e-01 -1.06449807e+00 1.18340337e+00 6.56266391e-01
7.29912043e-01 -5.51554501e-01 -3.56039703e-01 3.46544206e-01
6.97192013e-01 4.15413499e-01 6.70748472e-01 4.62972462e-01
-8.75610352e-01 7.50891538e-03 -1.02662957e+00 5.59138060e-01
2.94163942e-01 -2.79160500e-01 -1.14165604e+00 -5.04135430e-01
1.29871815e-01 -2.62689114e-01 1.35434973e+00 6.59913242e-01
1.68904459e+00 -2.02783421e-01 2.16545165e-01 1.36734939e+00
2.23826146e+00 -8.62061754e-02 8.21277201e-01 2.57065684e-01
7.67582893e-01 4.11633790e-01 3.83780599e-01 4.71654594e-01
2.45601892e-01 3.63974750e-01 7.61703432e-01 -6.68972552e-01
-2.99445629e-01 1.90442741e-01 4.08907197e-02 3.74829352e-01
-5.77542841e-01 -1.35405948e-02 -4.72757250e-01 3.88748467e-01
-1.59870124e+00 -1.15445387e+00 -6.97287768e-02 2.16630292e+00
8.07394922e-01 -5.18455207e-01 -2.99108326e-01 5.40841594e-02
8.46146882e-01 7.29568005e-01 -7.50953376e-01 4.83822435e-01
-6.08143270e-01 4.52036560e-01 7.80353963e-01 5.46405613e-01
-1.34889376e+00 7.94551373e-01 4.55733967e+00 9.17595625e-01
-1.40413296e+00 1.17144458e-01 7.26895094e-01 -2.27401331e-02
-1.98269576e-01 -3.27398092e-01 -4.51302260e-01 2.19664708e-01
4.63349044e-01 -1.95788264e-01 1.01517832e+00 3.56896341e-01
1.96806312e-01 -9.16218534e-02 -2.52973109e-01 1.20213604e+00
-1.12755276e-01 -1.51205933e+00 -6.21752627e-02 8.01202133e-02
8.81286800e-01 2.84742624e-01 4.33378100e-01 2.08432421e-01
2.70933986e-01 -1.10580897e+00 4.46951687e-01 6.60241008e-01
1.48692024e+00 -9.10836875e-01 5.05573511e-01 9.76366624e-02
-1.77949548e+00 -3.30216527e-01 -5.84314406e-01 2.76468724e-01
-1.48761347e-01 8.07991147e-01 3.14520568e-01 1.14629233e+00
1.22804797e+00 9.99835551e-01 -3.84864092e-01 6.61752164e-01
-1.56547427e-01 3.30122411e-01 8.09524581e-03 8.89849365e-01
2.47133017e-01 -6.03105962e-01 4.50183421e-01 1.06554663e+00
7.93938577e-01 8.23210120e-01 3.79302979e-01 1.10454726e+00
-2.09887043e-01 -2.86595851e-01 -3.28376353e-01 1.12056360e-01
2.63637811e-01 1.71000600e+00 -4.04414713e-01 -2.27467924e-01
-3.75782609e-01 9.91798818e-01 -3.29793133e-02 7.59596348e-01
-6.38185382e-01 -5.26131928e-01 8.88541818e-01 -1.04900248e-01
4.62596148e-01 -8.99189860e-02 -2.14449883e-01 -1.30784655e+00
8.57259780e-02 -8.64081442e-01 3.77528340e-01 -1.23812497e+00
-1.22946656e+00 9.53170717e-01 -3.08050483e-01 -1.28189850e+00
4.13816184e-01 -4.79125857e-01 -2.71866471e-01 1.45573485e+00
-2.25905132e+00 -1.18203568e+00 -1.13171113e+00 8.82530332e-01
4.32729483e-01 1.07273959e-01 6.86347008e-01 2.83710361e-01
-8.13327014e-01 8.16301629e-02 2.22333059e-01 5.11405356e-02
5.08745372e-01 -9.89535272e-01 9.92172733e-02 1.15175581e+00
-4.66010481e-01 3.45111787e-01 3.31609666e-01 -4.67541069e-01
-1.42466319e+00 -1.62227023e+00 2.06781760e-01 3.89296174e-01
6.01287246e-01 2.39879429e-01 -1.29814041e+00 4.31349069e-01
1.80174053e-01 5.22398293e-01 3.80685925e-01 -3.83150071e-01
-6.10951304e-01 -6.78842485e-01 -1.15345681e+00 4.76283818e-01
9.77683842e-01 -7.52553999e-01 1.03103323e-02 3.34562272e-01
1.05886078e+00 -3.50277811e-01 -1.26508188e+00 8.64989400e-01
2.12672934e-01 -1.28776550e+00 1.30798924e+00 1.67906567e-01
6.34241879e-01 -6.45784140e-01 -5.44898033e-01 -1.27208352e+00
-9.61297691e-01 -4.34154749e-01 1.49644300e-01 8.96763563e-01
1.12744048e-01 -7.57051766e-01 2.91091710e-01 1.46325141e-01
-3.44207793e-01 -5.81532180e-01 -6.98986709e-01 -5.72980464e-01
-1.37707740e-01 -2.00756803e-01 7.73370028e-01 1.10297728e+00
-6.46566391e-01 -1.04103489e-02 -4.09175396e-01 9.71459866e-01
1.09839296e+00 4.60213721e-01 2.40228042e-01 -1.20543885e+00
-9.30653512e-02 -6.99159503e-01 -2.07307842e-02 -7.62936711e-01
9.42100286e-02 -7.18337893e-01 1.62401900e-01 -1.59377420e+00
3.80187482e-01 -9.16027501e-02 -5.75559437e-01 5.54683208e-01
-3.86546552e-01 6.96397424e-01 1.25571579e-01 3.13165635e-01
-2.09556058e-01 6.35734022e-01 1.35548759e+00 -2.93464065e-01
-2.11728454e-01 -2.92945564e-01 -9.97895598e-01 5.72736442e-01
7.97277212e-01 -3.18715652e-03 -1.28119200e-01 -5.66602528e-01
3.13709043e-02 3.33617270e-01 6.94605410e-01 -1.19623923e+00
1.76942170e-01 -3.91147494e-01 7.11282253e-01 -6.09502852e-01
5.25609851e-01 -6.23133659e-01 3.28908533e-01 2.73146648e-02
-2.36987665e-01 -5.02465606e-01 2.43553445e-01 4.17588234e-01
-4.01453465e-01 3.29518557e-01 1.21777034e+00 -2.06159770e-01
-8.47545445e-01 8.30919623e-01 1.06775701e-01 -4.90529686e-01
6.08815908e-01 -9.71987098e-02 -6.93131447e-01 -1.79095447e-01
-4.85936999e-01 8.92001837e-02 3.53411168e-01 -3.42807919e-02
7.94620216e-01 -1.22743154e+00 -1.00560343e+00 4.31515843e-01
2.87219491e-02 2.27173239e-01 1.09131825e+00 8.74598026e-01
-3.17023993e-01 1.23598292e-01 -3.48746866e-01 -5.58608651e-01
-9.51408803e-01 3.89767468e-01 5.79360366e-01 -2.01730937e-01
-8.92325163e-01 8.27057302e-01 4.75143552e-01 -4.42612678e-01
-1.20332211e-01 -4.06881049e-02 -3.06529462e-01 -1.68936998e-01
9.31159258e-01 4.17805880e-01 -7.52901137e-02 -7.55817890e-01
3.41272093e-02 7.89462030e-01 3.07478815e-01 3.25919650e-02
1.86225915e+00 -3.81558239e-01 -6.27200902e-01 1.20193109e-01
1.23541296e+00 -1.82642683e-01 -1.62502468e+00 -7.34690785e-01
-5.66676974e-01 -7.87769258e-01 6.71987236e-01 -6.99092984e-01
-1.61762393e+00 7.26822674e-01 7.21825421e-01 2.13588327e-01
2.12691426e+00 -3.24474931e-01 7.17537880e-01 2.01328620e-01
1.94918349e-01 -6.72638357e-01 -6.34405538e-02 4.67956007e-01
1.08587074e+00 -1.29919374e+00 1.82198912e-01 -4.31766152e-01
-3.59626740e-01 1.08692181e+00 4.70030040e-01 -1.88525632e-01
5.88489294e-01 -9.62951221e-03 -2.42399678e-01 -1.71317458e-01
-2.50041246e-01 -5.58953941e-01 3.98758829e-01 4.77908522e-01
4.09175754e-01 1.92203447e-01 1.70280129e-01 4.65156823e-01
4.53177094e-02 2.64748018e-02 3.57092053e-01 3.89528930e-01
-6.28784120e-01 -4.14053947e-01 -4.99633163e-01 3.70042920e-01
-2.54644871e-01 -3.91109675e-01 2.56491154e-01 4.78108287e-01
2.39297584e-01 9.67929065e-01 -9.35098678e-02 -6.02096081e-01
2.68220574e-01 -3.75579268e-01 1.90608680e-01 -3.45950723e-01
-1.71954930e-01 4.85099286e-01 -4.43706095e-01 -8.05634439e-01
-5.57504475e-01 -3.87084961e-01 -1.11379969e+00 -3.47833961e-01
-1.16066329e-01 -1.38026938e-01 3.93541455e-01 6.18066490e-01
3.67251575e-01 8.46104681e-01 1.06272566e+00 -1.34987903e+00
-2.94233769e-01 -9.50979888e-01 -1.08551157e+00 1.32336477e-02
7.04934001e-01 -3.01068097e-01 -5.25281072e-01 -8.91123712e-03] | [10.27596378326416, -1.955743670463562] |
6c9caf4b-4c39-42b6-95b9-cdb079263593 | lenet-lightweight-and-efficient-lidar | 2301.04275 | null | https://arxiv.org/abs/2301.04275v3 | https://arxiv.org/pdf/2301.04275v3.pdf | LENet: Lightweight And Efficient LiDAR Semantic Segmentation Using Multi-Scale Convolution Attention | LiDAR-based semantic segmentation is critical in the fields of robotics and autonomous driving as it provides a comprehensive understanding of the scene. This paper proposes a lightweight and efficient projection-based semantic segmentation network called LENet with an encoder-decoder structure for LiDAR-based semantic segmentation. The encoder is composed of a novel multi-scale convolutional attention (MSCA) module with varying receptive field sizes to capture features. The decoder employs an Interpolation And Convolution (IAC) mechanism utilizing bilinear interpolation for upsampling multi-resolution feature maps and integrating previous and current dimensional features through a single convolution layer. This approach significantly reduces the network's complexity while also improving its accuracy. Additionally, we introduce multiple auxiliary segmentation heads to further refine the network's accuracy. Extensive evaluations on publicly available datasets, including SemanticKITTI, SemanticPOSS, and nuScenes, show that our proposed method is lighter, more efficient, and robust compared to state-of-the-art semantic segmentation methods. Full implementation is available at https://github.com/fengluodb/LENet. | ['Ben Ding'] | 2023-01-11 | null | null | null | null | ['lidar-semantic-segmentation'] | ['computer-vision'] | [ 1.29757747e-01 1.16906367e-01 -2.96526104e-02 -7.89189577e-01
-6.95666254e-01 -2.95513511e-01 4.44993734e-01 -1.78682148e-01
-6.53713524e-01 4.21469182e-01 -4.52433191e-02 -2.77444899e-01
1.96261272e-01 -1.10882533e+00 -1.00327337e+00 -3.39910328e-01
3.31065238e-01 4.49937433e-01 7.86354184e-01 -2.56575018e-01
2.44782463e-01 5.22619307e-01 -1.77175629e+00 1.15257919e-01
1.11687803e+00 1.33120668e+00 7.60018945e-01 4.82390374e-01
-2.49366835e-01 4.27783668e-01 -9.59960595e-02 -2.41759449e-01
4.34405804e-01 1.01356737e-01 -1.00722826e+00 -1.50775895e-01
6.50440395e-01 -4.59786355e-01 -4.13435280e-01 1.05498016e+00
2.78204978e-01 1.82711631e-01 2.32291520e-01 -1.21113861e+00
-5.64673603e-01 3.60821575e-01 -4.22402978e-01 -6.66842088e-02
-8.75462070e-02 3.34475376e-02 8.41177225e-01 -9.58128691e-01
4.53275323e-01 1.40725064e+00 9.10046756e-01 3.06878030e-01
-8.79171908e-01 -8.40379655e-01 1.90226622e-02 2.46218279e-01
-1.44995010e+00 -1.53601602e-01 6.11968398e-01 -2.96374619e-01
9.74333405e-01 -8.08108002e-02 7.32028902e-01 6.39315426e-01
1.80168286e-01 9.30828989e-01 9.68274474e-01 5.79473078e-02
2.29448631e-01 -3.47831585e-02 2.30133012e-01 8.74506950e-01
1.94562390e-01 1.64298471e-02 -3.56662303e-01 3.71626824e-01
8.54849875e-01 6.66271895e-02 1.57838717e-01 -4.14891809e-01
-9.10908937e-01 1.04475737e+00 1.11998463e+00 -1.18156284e-01
-3.10055643e-01 4.39983100e-01 2.34644070e-01 -3.72035950e-01
4.40172970e-01 8.19917992e-02 -3.61063361e-01 3.66840884e-02
-9.75460410e-01 3.42813671e-01 4.79328573e-01 1.25450051e+00
1.24215245e+00 2.66226809e-02 1.12236723e-01 8.42158675e-01
2.98563302e-01 5.51091313e-01 3.37444097e-01 -1.46867633e+00
4.17245001e-01 7.25800514e-01 -4.76290882e-02 -5.85517883e-01
-4.93091941e-01 -1.38787702e-01 -3.99184346e-01 3.07471246e-01
4.77523319e-02 3.12335528e-02 -1.38933921e+00 1.34787953e+00
3.38483989e-01 3.56000036e-01 1.94988288e-02 1.18098783e+00
1.04315257e+00 5.72943628e-01 2.23680809e-01 8.28719735e-01
1.27534902e+00 -1.39863551e+00 -3.38036627e-01 -6.45325124e-01
3.50054801e-01 -5.20805299e-01 9.38809991e-01 4.11574403e-03
-8.69565070e-01 -7.38424182e-01 -1.21470022e+00 -8.48144233e-01
-7.56488800e-01 2.61783749e-01 9.95901704e-01 5.06538212e-01
-1.20389450e+00 6.01132751e-01 -1.02838373e+00 -4.92262602e-01
9.45268691e-01 4.73157912e-01 -1.23988293e-01 -2.46672839e-01
-1.01655507e+00 9.09141541e-01 7.10610688e-01 3.00575774e-02
-7.77096391e-01 -5.15640318e-01 -1.20235598e+00 -4.81135212e-02
2.09161893e-01 -6.63077533e-01 1.42356229e+00 -5.31985700e-01
-1.47423959e+00 7.36215115e-01 -3.61113667e-01 -6.90798104e-01
3.24635386e-01 -4.19878393e-01 -1.10107753e-02 3.21616858e-01
4.73846704e-01 1.63698411e+00 4.61195886e-01 -1.14177370e+00
-8.82163703e-01 -4.31309313e-01 1.70404136e-01 4.43144113e-01
1.21629059e-01 -2.21609026e-01 -8.36487949e-01 -1.87145859e-01
4.07408655e-01 -8.15471470e-01 -4.70647901e-01 2.80512989e-01
-4.08448100e-01 -6.94965050e-02 1.18692958e+00 -7.12531984e-01
6.57377779e-01 -2.08043146e+00 -2.26653010e-01 -8.68035629e-02
1.07735649e-01 3.39568585e-01 -5.01947664e-02 3.50815617e-02
3.14665526e-01 -2.45989412e-02 -7.43514538e-01 -5.97028673e-01
-1.42308250e-01 5.11029303e-01 -3.58846830e-03 3.36744696e-01
3.50049794e-01 1.19538367e+00 -7.07081318e-01 -5.95893145e-01
7.11422980e-01 7.27858305e-01 -6.50936246e-01 -9.67736319e-02
-3.24315429e-01 6.01205707e-01 -5.67048132e-01 8.35895002e-01
9.97170866e-01 -1.31249562e-01 -4.60567653e-01 -4.11116704e-02
-4.86303240e-01 4.78485256e-01 -9.46262717e-01 2.27811623e+00
-4.82671767e-01 8.09387207e-01 1.17517143e-01 -9.81264710e-01
1.02534568e+00 -2.18691871e-01 3.10703903e-01 -8.95977199e-01
3.80248964e-01 3.27433020e-01 -4.10790294e-01 -2.23028377e-01
9.03082848e-01 1.81376919e-01 -1.80621192e-01 -6.02021515e-02
2.28825971e-01 -3.92461956e-01 2.71955013e-01 7.56123811e-02
7.70965874e-01 4.71983671e-01 -8.42490271e-02 -3.20427001e-01
4.81938809e-01 4.27677929e-01 6.88032866e-01 4.21684116e-01
-2.55246758e-01 7.58024871e-01 4.80115972e-02 -2.94890851e-01
-1.08112824e+00 -1.16314614e+00 -4.96624500e-01 8.61894071e-01
7.47205138e-01 -2.44866177e-01 -9.33452547e-01 -3.19313735e-01
2.22135305e-01 7.90304005e-01 -4.39467043e-01 7.34856874e-02
-5.38865745e-01 -2.67922550e-01 6.26410186e-01 9.68289256e-01
1.18056583e+00 -1.03165936e+00 -9.87854183e-01 7.80123770e-02
-2.61331916e-01 -1.44834793e+00 -2.30969653e-01 2.29042351e-01
-9.79398608e-01 -9.86763954e-01 -3.73454779e-01 -8.55197608e-01
5.70994914e-01 3.83583277e-01 8.49613905e-01 -3.39964330e-01
-4.92742836e-01 -4.15933393e-02 -1.98663473e-01 -4.92614686e-01
7.88227394e-02 2.47988001e-01 -3.96388143e-01 -4.14908051e-01
5.45228839e-01 -4.41042006e-01 -7.52086639e-01 1.65625498e-01
-7.46793985e-01 4.98145372e-01 5.85384190e-01 5.31035304e-01
8.89459431e-01 -2.28989929e-01 1.35030523e-01 -7.49545693e-01
1.66534215e-01 -5.15149593e-01 -9.20296848e-01 -3.18178862e-01
-4.78573918e-01 -5.75904027e-02 5.07822692e-01 1.60223618e-01
-1.19307458e+00 3.53911459e-01 -4.26348180e-01 -4.31490034e-01
-3.21295917e-01 2.82085478e-01 4.51508686e-02 -2.83221424e-01
3.08330804e-01 9.24346223e-02 -1.16556145e-01 -6.37629986e-01
5.63482165e-01 7.48925269e-01 7.80374825e-01 -3.52644563e-01
5.98091185e-01 7.86965668e-01 -1.53480873e-01 -7.48736024e-01
-9.21991229e-01 -7.25086272e-01 -1.02109849e+00 -1.00674145e-01
1.09891021e+00 -1.17706609e+00 -4.63258028e-01 6.45964205e-01
-1.16783035e+00 -4.99700069e-01 -2.83987105e-01 4.69113737e-01
-7.03065217e-01 -4.46621142e-02 -6.03347659e-01 -4.43204910e-01
-3.61696064e-01 -1.43322611e+00 1.47313595e+00 6.37219489e-01
2.19014421e-01 -7.51991391e-01 -2.00985000e-01 6.94329977e-01
4.54652548e-01 1.79756463e-01 4.61916894e-01 -3.15790832e-01
-1.00083077e+00 4.46860585e-03 -7.71002889e-01 3.78145933e-01
-2.36014351e-01 -2.15853408e-01 -1.16668117e+00 2.38707289e-02
-4.30215210e-01 -3.38933855e-01 1.25026858e+00 5.21653831e-01
1.38537192e+00 6.46279082e-02 -3.67189586e-01 1.05807185e+00
1.70719433e+00 1.46313921e-01 6.42478108e-01 4.43651587e-01
9.58788216e-01 4.07070130e-01 7.14411795e-01 2.08725393e-01
9.26517308e-01 3.97251815e-01 6.79658175e-01 -2.79007167e-01
-3.52761418e-01 -3.68232459e-01 1.10554630e-02 6.44703329e-01
2.46147782e-01 1.01174906e-01 -1.08299458e+00 8.12361419e-01
-1.80546844e+00 -6.07821405e-01 -2.53277868e-01 1.80387902e+00
5.10931969e-01 1.64640807e-02 -1.76187664e-01 -5.95390312e-02
6.49355829e-01 1.38108984e-01 -9.22182739e-01 -6.53961122e-01
-1.08570056e-02 5.66601932e-01 1.07560945e+00 5.34100354e-01
-1.35237396e+00 1.54933834e+00 5.74343538e+00 7.15820730e-01
-1.15146351e+00 3.25896710e-01 4.06617016e-01 1.75548971e-01
-1.43727198e-01 -2.95244306e-02 -1.00657475e+00 3.04923296e-01
9.09243643e-01 1.05337746e-01 3.58408839e-01 1.11932039e+00
1.35728851e-01 -5.12380242e-01 -5.24438977e-01 6.25992358e-01
-1.23009011e-01 -1.39763641e+00 -2.66824722e-01 -1.21591032e-01
7.43000925e-01 9.24981534e-01 -9.27681476e-03 2.61046588e-01
4.82816041e-01 -1.02233207e+00 9.55282450e-01 3.35617423e-01
8.96283209e-01 -8.20488811e-01 6.45508230e-01 2.49942988e-01
-1.51263678e+00 -1.64658114e-01 -5.84977329e-01 -8.68950561e-02
1.90117851e-01 4.24711406e-01 -6.89205110e-01 4.31901544e-01
1.06074715e+00 9.25144315e-01 -6.29391909e-01 1.04367352e+00
-3.29715878e-01 3.58460337e-01 -5.00096738e-01 2.19983771e-01
6.39526308e-01 -3.70566279e-01 2.16546938e-01 1.21544302e+00
3.23569626e-01 3.13055329e-02 3.20635051e-01 1.22310126e+00
-9.67421979e-02 -1.97220653e-01 -4.55584884e-01 2.17682391e-01
6.30218565e-01 1.34084904e+00 -8.95786345e-01 -3.80101502e-01
-5.22937059e-01 9.53188956e-01 3.23576629e-01 3.12659383e-01
-9.12070394e-01 -6.29559696e-01 8.50017965e-01 -3.30371223e-03
6.20756626e-01 -3.95193994e-01 -7.85950065e-01 -7.68045068e-01
-7.89058879e-02 -8.21604803e-02 -3.31792943e-02 -9.90508080e-01
-6.06074691e-01 5.52351594e-01 -6.10621944e-02 -9.37642753e-01
1.81657657e-01 -5.87217391e-01 -4.69649374e-01 1.03117359e+00
-2.04522634e+00 -1.47849786e+00 -7.71611094e-01 5.86982608e-01
7.97803700e-01 2.49499187e-01 6.45298183e-01 3.26793671e-01
-4.98324752e-01 9.34920758e-02 5.62380403e-02 1.62679270e-01
2.45607480e-01 -1.13440847e+00 8.26787889e-01 8.97895694e-01
-3.18756938e-01 2.15007797e-01 3.01181853e-01 -6.23592436e-01
-1.06575274e+00 -1.68050492e+00 6.26935065e-01 -8.48638639e-02
3.87543768e-01 -3.56975108e-01 -8.02674294e-01 7.89747894e-01
4.78902049e-02 7.71639645e-02 1.88034654e-01 -3.70865226e-01
-2.36692578e-01 -5.95822074e-02 -1.25389290e+00 4.15086359e-01
1.09965885e+00 -4.98483628e-01 -4.64288324e-01 1.06552109e-01
9.62038636e-01 -7.83877134e-01 -6.76327884e-01 5.99629998e-01
5.24505436e-01 -1.06539881e+00 1.03236735e+00 1.24609910e-01
5.58129609e-01 -3.06301385e-01 -2.86472887e-01 -1.01958990e+00
-2.08289102e-01 1.63246654e-02 1.39430195e-01 7.23406553e-01
2.93504089e-01 -7.84547448e-01 9.10434365e-01 3.69565248e-01
-7.32867002e-01 -9.08590436e-01 -9.17446494e-01 -6.48192942e-01
7.47058019e-02 -6.76319480e-01 6.10861897e-01 4.71161515e-01
-4.57883775e-01 1.18328243e-01 1.57977968e-01 3.79077464e-01
6.84870899e-01 1.55408874e-01 5.01164258e-01 -1.25378191e+00
3.87829542e-01 -3.37945700e-01 -5.54252982e-01 -1.24714088e+00
2.42452413e-01 -1.05656087e+00 3.50359052e-01 -2.04074478e+00
-1.49915675e-02 -6.14949465e-01 -9.50948372e-02 6.58117950e-01
9.76139084e-02 5.99054217e-01 9.79232043e-02 8.14343020e-02
-6.48263872e-01 6.90784991e-01 1.34981501e+00 -5.83894700e-02
-1.20037943e-02 -6.70444295e-02 -5.46514034e-01 8.40559959e-01
1.21239626e+00 -3.82380545e-01 -3.55351120e-01 -6.95399106e-01
-3.62368852e-01 -2.22235873e-01 5.83789051e-01 -1.42043126e+00
3.84250849e-01 5.31213395e-02 2.01876312e-01 -1.12762249e+00
7.46138811e-01 -6.48497820e-01 -1.88701004e-01 4.33991939e-01
-6.42735735e-02 -1.23262145e-01 4.60672557e-01 3.58950168e-01
-2.47983307e-01 -8.37608948e-02 9.50374126e-01 -2.44587809e-01
-1.34733415e+00 5.75043917e-01 -1.99828237e-01 -6.07876964e-02
9.92223144e-01 -5.56020081e-01 -2.44295627e-01 -4.00025025e-02
-3.77548128e-01 5.96563518e-01 5.46216369e-01 6.11568809e-01
8.12171161e-01 -1.16342831e+00 -4.40973729e-01 3.80759776e-01
7.81680048e-02 6.86321378e-01 1.91757962e-01 6.42385721e-01
-1.07438266e+00 7.42594659e-01 -4.63529617e-01 -8.07924747e-01
-8.64024341e-01 -1.35770708e-04 3.57781470e-01 2.26106346e-01
-8.47458899e-01 1.08204818e+00 2.06971362e-01 -9.02159870e-01
1.19540267e-01 -5.61404347e-01 -1.39737815e-01 -2.22882986e-01
1.10516861e-01 5.84666431e-01 1.15780979e-02 -9.68128026e-01
-3.67476761e-01 7.25546360e-01 1.45921201e-01 2.33727712e-02
1.17338896e+00 -3.03756893e-01 3.20873037e-02 2.69308299e-01
1.26833189e+00 -7.01560020e-01 -1.75230467e+00 -2.69738376e-01
-1.26298651e-01 -4.46909755e-01 4.39029038e-01 -7.24108100e-01
-1.25653410e+00 1.16620541e+00 5.98873079e-01 -3.71435195e-01
1.02306128e+00 -2.96504106e-02 1.27207446e+00 2.85032183e-01
4.96528208e-01 -1.33692908e+00 -2.61852235e-01 8.39693546e-01
6.36460900e-01 -1.38183153e+00 -1.39344692e-01 -6.85051680e-01
-6.34558916e-01 8.29398394e-01 8.48328531e-01 -3.70095074e-01
6.53885841e-01 3.38409275e-01 1.05539322e-01 -2.30952919e-01
-1.95085526e-01 -6.09032273e-01 9.27919671e-02 5.35740256e-01
1.00361116e-01 2.57303923e-01 -5.63961193e-02 3.08803499e-01
-5.49340785e-01 4.58478965e-02 1.83833331e-01 8.74457717e-01
-6.79501355e-01 -7.57764935e-01 -2.03942046e-01 4.51473713e-01
-9.12377760e-02 -1.40362561e-01 2.44978275e-02 7.69755781e-01
4.46435899e-01 8.04888666e-01 3.99493903e-01 -3.33501577e-01
1.49202168e-01 -7.13027865e-02 1.76479056e-01 -6.40703797e-01
-2.71736175e-01 -1.69907033e-01 -4.63899150e-02 -1.01419985e+00
-3.84574175e-01 -5.47596753e-01 -1.79789996e+00 -2.19529852e-01
-1.85759217e-01 -2.15094969e-01 1.25897634e+00 9.03754771e-01
5.09788573e-01 6.90686882e-01 3.53324413e-01 -1.18549168e+00
-2.96859704e-02 -8.77855897e-01 -3.73895556e-01 -3.06072142e-02
1.09153233e-01 -8.64596963e-01 1.88604668e-01 -4.40478958e-02] | [8.194713592529297, -2.5667667388916016] |
28a02af3-7771-4f60-b87e-741988ec84a6 | lifetime-achievement-award-translating-today | null | null | https://aclanthology.org/J15-4007 | https://aclanthology.org/J15-4007.pdf | Lifetime Achievement Award: Translating Today into Tomorrow | null | ['Sheng Li'] | 2015-12-01 | null | null | null | cl-2015-12 | ['lexical-analysis'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.41171407699585, 3.5683469772338867] |
c916948b-98ea-4064-b365-78ae42e8d27e | deep-learning-based-stereo-camera-multi-video | 2303.12916 | null | https://arxiv.org/abs/2303.12916v1 | https://arxiv.org/pdf/2303.12916v1.pdf | Deep learning-based stereo camera multi-video synchronization | Stereo vision is essential for many applications. Currently, the synchronization of the streams coming from two cameras is done using mostly hardware. A software-based synchronization method would reduce the cost, weight and size of the entire system and allow for more flexibility when building such systems. With this goal in mind, we present here a comparison of different deep learning-based systems and prove that some are efficient and generalizable enough for such a task. This study paves the way to a production ready software-based video synchronization system. | ['Thierry Dutoit', 'François Cresson', 'Thierry Ravet', 'Kevin El Haddad', 'Nicolas Boizard'] | 2023-03-22 | null | null | null | null | ['video-synchronization'] | ['computer-vision'] | [-2.99223930e-01 -3.26966822e-01 2.09600881e-01 -3.90523195e-01
-6.82011396e-02 -2.86159903e-01 5.70478976e-01 -1.34247895e-02
-6.91865444e-01 4.32762027e-01 -3.75015110e-01 -2.55741328e-01
3.30502182e-01 -6.81675792e-01 -5.39045691e-01 -6.39330685e-01
9.26867351e-02 3.09190452e-01 8.01943898e-01 -1.46742091e-01
1.64141387e-01 5.27421951e-01 -1.80063009e+00 9.58895832e-02
2.46344611e-01 8.19194257e-01 6.46499693e-01 7.41012454e-01
-2.01065615e-01 8.02251339e-01 -5.84926963e-01 -2.67741531e-01
4.83591229e-01 -4.44130450e-01 -6.53974235e-01 7.97618032e-02
6.16337359e-01 -5.89506447e-01 -4.76569533e-01 8.56375873e-01
4.27459776e-01 -9.66949835e-02 1.34000614e-01 -1.22623229e+00
4.73019421e-01 2.99018741e-01 -1.83110699e-01 2.85837054e-01
3.92933071e-01 -2.21383199e-02 4.79949653e-01 -4.50350225e-01
6.36708558e-01 6.84555948e-01 5.62987864e-01 5.17211378e-01
-8.93085182e-01 -5.90100229e-01 -2.06166252e-01 4.61732149e-01
-1.26899266e+00 -7.25700498e-01 7.38860071e-01 -3.55661511e-01
1.00350106e+00 1.74772248e-01 1.05339205e+00 7.81243265e-01
3.57990235e-01 3.84029895e-01 9.36814368e-01 -5.75188875e-01
4.29080784e-01 7.29060993e-02 3.34182680e-01 5.59237599e-01
5.37374020e-01 -2.19625413e-01 -6.19836152e-01 1.26197636e-01
1.05238318e+00 2.04637364e-01 -3.62643212e-01 -4.74643499e-01
-1.06102347e+00 5.54158688e-01 4.91891652e-02 6.35971427e-01
-1.44003719e-01 3.10888529e-01 6.47494853e-01 4.92315382e-01
1.24375679e-01 7.41466507e-02 -1.95957750e-01 -3.59187067e-01
-1.28118658e+00 -4.90605757e-02 1.07970488e+00 1.11707807e+00
6.83402240e-01 1.22529730e-01 4.80581194e-01 3.62674028e-01
3.71606022e-01 3.68997693e-01 6.37557328e-01 -1.15687299e+00
3.44507322e-02 4.07279104e-01 -1.34921789e-01 -9.42842305e-01
-5.22262573e-01 1.17474996e-01 -6.85661077e-01 4.60575163e-01
5.55568218e-01 -1.91182747e-01 -4.57795143e-01 1.20225966e+00
2.02782437e-01 3.39195728e-01 -1.71122491e-01 9.40561950e-01
7.30182469e-01 5.59093297e-01 -5.32246113e-01 -1.78641260e-01
1.15874791e+00 -8.39029729e-01 -6.29411399e-01 -6.89993203e-02
4.53353077e-01 -9.87221003e-01 6.14107549e-01 6.58175230e-01
-1.03026533e+00 -7.36877739e-01 -1.31949258e+00 -4.27222252e-02
-1.99358985e-01 2.72280835e-02 6.86966062e-01 6.89300418e-01
-1.47809768e+00 5.41680992e-01 -1.12628365e+00 -7.06999063e-01
-1.73576355e-01 6.71600342e-01 -4.29243207e-01 2.17187032e-01
-7.19437838e-01 9.37641978e-01 3.38116556e-01 3.04923564e-01
-4.32289451e-01 -1.04801301e-02 -6.12538278e-01 4.30579931e-02
1.88537687e-02 -7.63557374e-01 1.46852398e+00 -1.07459235e+00
-1.87978745e+00 1.14362371e+00 -1.14661537e-01 -7.30774999e-01
6.60416961e-01 -1.47464663e-01 -7.13493228e-02 2.20586061e-01
-3.23053747e-01 6.62804306e-01 1.05696535e+00 -5.88899076e-01
-8.10998261e-01 -6.07487373e-02 2.22300991e-01 3.98399755e-02
-4.25444603e-01 1.57431141e-01 -7.47534037e-01 -5.75489588e-02
-1.26430809e-01 -1.05427766e+00 -2.23098725e-01 6.59273118e-02
8.90650973e-02 1.24386042e-01 7.43374825e-01 -1.85015291e-01
8.09588552e-01 -2.28269029e+00 6.29107505e-02 -2.83646602e-02
2.33256951e-01 5.30367315e-01 3.05314332e-01 4.01246995e-01
9.12284330e-02 -7.04739749e-01 4.61031437e-01 -2.87229717e-01
-2.84677058e-01 2.55819887e-01 -6.62289411e-02 7.17180908e-01
-4.48960632e-01 1.28252879e-01 -4.13168758e-01 -6.77887559e-01
8.15647542e-01 5.06146252e-01 -5.73947012e-01 3.29963177e-01
2.49241203e-01 3.57334465e-01 -1.34229258e-01 1.35996724e-02
6.11706078e-01 1.20460242e-02 3.72547120e-01 -1.54503271e-01
-4.91327137e-01 2.01574668e-01 -1.52982271e+00 1.55367637e+00
-5.45194507e-01 1.27790558e+00 3.31156909e-01 -9.98893797e-01
9.63464379e-01 3.96366924e-01 4.94724095e-01 -6.11018419e-01
4.69840139e-01 2.69955903e-01 -6.32275343e-02 -3.79635632e-01
6.22798681e-01 1.43886864e-01 3.16441864e-01 3.51943821e-01
2.85961628e-01 -2.62614638e-01 4.38292980e-01 2.79267016e-03
9.12394106e-01 1.59525629e-02 3.89837891e-01 -4.72965360e-01
7.83569276e-01 -2.50915829e-02 2.68786579e-01 5.75146317e-01
-2.30035707e-01 4.98598963e-01 1.80805027e-01 -8.47196877e-01
-1.23213649e+00 -5.19231439e-01 1.15400068e-01 5.40778160e-01
3.11229318e-01 -5.97952604e-01 -9.36451793e-01 -2.61003412e-02
-3.55777472e-01 6.02813363e-02 5.66362776e-03 2.70493537e-01
-6.57356143e-01 -1.28861979e-01 1.90565526e-01 3.20755064e-01
4.88443971e-01 -7.80032218e-01 -1.33723903e+00 2.20794648e-01
-1.24696065e-02 -1.30679703e+00 -2.31392413e-01 1.63489133e-01
-1.22879148e+00 -1.19716418e+00 -7.28138804e-01 -8.84538472e-01
5.21813750e-01 8.41168940e-01 9.34570312e-01 2.23109901e-01
-2.94751078e-01 5.41080236e-01 -1.82188556e-01 -3.53146225e-01
-1.61491469e-01 1.26646101e-01 1.65742487e-01 -2.71769971e-01
3.75344485e-01 -4.53576267e-01 -4.67314512e-01 2.43337825e-01
-7.59783924e-01 3.01254719e-01 2.08160177e-01 3.61295938e-01
2.38692909e-01 -9.76261348e-02 -2.19671026e-01 -5.22580922e-01
3.03644717e-01 6.19408637e-02 -1.36568773e+00 -1.16201781e-01
-3.75949472e-01 -1.18660629e-02 7.53873050e-01 -2.31208533e-01
-8.87988925e-01 5.04836082e-01 -2.00034082e-01 -3.34894240e-01
-1.40849650e-01 5.00400104e-02 2.44684413e-01 -3.30978572e-01
4.01923150e-01 3.41237456e-01 1.76294550e-01 -2.55349398e-01
5.95742166e-02 7.10146666e-01 4.11982715e-01 -9.27766860e-02
4.55707431e-01 7.37321198e-01 3.00503261e-02 -1.30691850e+00
-1.23656400e-01 -6.48523390e-01 -5.79363883e-01 -5.98027408e-01
6.11576140e-01 -1.10277593e+00 -8.73917997e-01 8.04843366e-01
-1.48997343e+00 -3.44851911e-01 -1.03650466e-02 6.69249952e-01
-6.50464892e-01 3.43889356e-01 -5.48400939e-01 -4.07314986e-01
-3.10440868e-01 -1.33975565e+00 9.06426191e-01 3.53585303e-01
-2.39458412e-01 -9.48153675e-01 4.50792283e-01 2.20713228e-01
4.45407689e-01 -2.57108212e-01 -3.81417647e-02 -3.46675187e-01
-8.16954315e-01 -2.74884433e-01 -1.34994775e-01 4.08709049e-01
-1.50620520e-01 4.01679277e-01 -9.81757522e-01 -3.29736620e-01
4.50443596e-01 6.91848025e-02 5.59295952e-01 5.55043340e-01
7.25949526e-01 1.43077299e-01 -2.76261091e-01 6.14246190e-01
1.72643602e+00 2.63789296e-01 5.58919549e-01 6.15701318e-01
3.68764937e-01 4.44560707e-01 4.16789830e-01 6.12176776e-01
2.92438954e-01 9.58275616e-01 2.74783731e-01 -1.01875871e-01
-7.20552355e-02 4.63692158e-01 6.25308871e-01 1.01125085e+00
-2.94668078e-01 3.16650607e-02 -9.89603460e-01 2.50752687e-01
-1.89952219e+00 -1.04579759e+00 -5.85416257e-01 2.30592036e+00
4.81363058e-01 1.32271513e-01 4.91559207e-02 1.73668742e-01
7.76915431e-01 6.88649938e-02 6.56043291e-02 -6.02779508e-01
3.12924087e-01 1.65370613e-01 6.06666267e-01 4.89794403e-01
-9.89285231e-01 7.59605765e-01 7.35845709e+00 4.15659100e-01
-1.62921023e+00 4.63884883e-02 2.90379077e-02 -1.77054331e-01
3.63600820e-01 2.22676508e-02 -9.43668127e-01 5.53628027e-01
1.14350951e+00 -2.85321385e-01 3.29274267e-01 1.01370323e+00
2.87169695e-01 -6.40425742e-01 -1.15651286e+00 1.42316127e+00
7.13493153e-02 -1.44754350e+00 -3.56727779e-01 8.81969631e-02
2.76713550e-01 4.24736202e-01 -4.23481196e-01 -1.44927830e-01
-8.26586932e-02 -3.34461361e-01 7.77994812e-01 2.93487668e-01
2.96061158e-01 -6.03647888e-01 1.01221192e+00 2.02178806e-01
-1.02912569e+00 2.86394686e-01 -5.55903971e-01 -5.65422952e-01
2.37206593e-01 6.59974158e-01 -5.42786181e-01 3.82540852e-01
8.65971446e-01 3.55205357e-01 -5.03349006e-01 1.14690351e+00
2.01497078e-02 1.72562733e-01 -5.44163346e-01 -8.52385759e-02
1.60644099e-01 -3.24842900e-01 2.42141977e-01 1.32136989e+00
4.38231319e-01 -2.33015150e-01 -6.66139275e-02 1.70868501e-01
3.21064532e-01 -6.62107319e-02 -9.31121767e-01 2.61985958e-01
2.39825293e-01 1.28524756e+00 -9.66099262e-01 -5.96387386e-01
-9.00642931e-01 1.14385569e+00 7.49865472e-02 -2.21319258e-01
-9.66445208e-01 -4.74833757e-01 6.33222461e-01 6.29158914e-02
2.63680816e-01 -7.40774512e-01 -5.96592315e-02 -1.47162676e+00
-1.06920861e-01 -6.55692458e-01 7.97726437e-02 -7.82924771e-01
-4.72750127e-01 7.00295806e-01 -1.83799967e-01 -1.44673347e+00
-5.39376199e-01 -7.27487743e-01 -4.91048962e-01 3.32118452e-01
-1.35610366e+00 -6.35313332e-01 -6.78166509e-01 8.65165949e-01
5.42142153e-01 -2.93746740e-01 7.11775362e-01 4.80132461e-01
-6.40283763e-01 7.73959309e-02 1.35728121e-01 3.28975320e-02
8.48955333e-01 -8.98603737e-01 1.63246989e-01 1.04345953e+00
2.95982420e-01 5.48153698e-01 9.60108995e-01 -9.35827345e-02
-1.64500415e+00 -3.91596109e-01 9.89657462e-01 -5.51662594e-02
6.35320783e-01 -2.95572639e-01 -4.62847233e-01 6.45950913e-01
7.61385322e-01 -1.02335609e-01 5.03449023e-01 -6.87564313e-02
7.11003765e-02 -6.22052431e-01 -6.80174708e-01 5.61862051e-01
5.53409457e-01 -5.52565277e-01 -5.08792818e-01 1.88122764e-01
2.18662813e-01 -6.36713386e-01 -5.66717803e-01 -2.46277586e-01
6.85304344e-01 -1.71853292e+00 7.30381846e-01 3.02185535e-01
1.77109241e-01 -2.79015958e-01 5.75504974e-02 -8.56641352e-01
-6.62554661e-03 -7.14262009e-01 1.28900975e-01 1.01177251e+00
-1.71379015e-01 -6.62813306e-01 9.43378866e-01 5.53114653e-01
-1.07913882e-01 -1.39538925e-02 -1.03023684e+00 -9.50133026e-01
-5.66252410e-01 -5.34434021e-01 1.56591684e-01 6.77458167e-01
3.14948648e-01 2.71742195e-01 -2.56568372e-01 2.74539441e-02
4.03742611e-01 2.34197304e-01 1.16997278e+00 -1.19753110e+00
-2.32940704e-01 -3.76722366e-01 -9.40187573e-01 -9.05846238e-01
-5.10685928e-02 -2.67732501e-01 -1.07438415e-01 -1.17030990e+00
6.08193725e-02 5.92441373e-02 1.20544486e-01 1.44602999e-01
4.62758750e-01 2.65840113e-01 4.77173060e-01 3.29702944e-01
-7.13864744e-01 1.39295012e-02 6.74449682e-01 2.62784064e-01
-3.14301178e-02 4.83121090e-02 -7.71210268e-02 9.31958139e-01
8.89042020e-01 -3.49479824e-01 -3.43478709e-01 -6.60492241e-01
1.51300833e-01 -1.25237331e-02 1.91040143e-01 -1.55960703e+00
8.08581114e-01 5.37135974e-02 2.06924640e-02 -3.96559387e-01
4.32452500e-01 -1.30199790e+00 4.25045162e-01 8.68326545e-01
1.80535108e-01 1.85171947e-01 7.95421749e-02 1.63570181e-01
-6.35663509e-01 -4.09414083e-01 9.68914449e-01 -1.33636400e-01
-1.03548980e+00 3.36668007e-02 -7.36606717e-01 -3.91067505e-01
1.33891976e+00 -5.41219950e-01 -1.51667416e-01 -3.54967058e-01
-3.96685779e-01 -1.67408541e-01 7.62600780e-01 5.74144311e-02
3.73431325e-01 -9.76205289e-01 -9.69667733e-02 4.11522418e-01
-1.59642607e-01 -7.63987824e-02 8.19559172e-02 8.20330501e-01
-1.30675924e+00 8.13356102e-01 -7.08204269e-01 -8.53724182e-01
-1.78119016e+00 3.91396880e-01 3.72049242e-01 6.77696839e-02
-6.77288949e-01 6.21387184e-01 -2.38873273e-01 1.70965180e-01
2.85279274e-01 -4.58677828e-01 -2.02921018e-01 1.93705127e-01
6.27969801e-01 4.79684025e-01 1.95976734e-01 -3.30607086e-01
-5.08058727e-01 7.07341075e-01 1.51964828e-01 -2.61474997e-01
1.27716041e+00 -1.72846228e-01 -2.22494379e-01 4.70728397e-01
8.64356637e-01 -6.83975518e-02 -1.26238906e+00 1.16803393e-01
-5.26220202e-02 -4.23012137e-01 2.27836758e-01 3.45200956e-01
-1.23646677e+00 9.05894399e-01 7.42694616e-01 4.13419664e-01
1.18110967e+00 -3.30236733e-01 6.87517464e-01 5.72492778e-01
8.52789462e-01 -1.23975563e+00 -1.94544688e-01 5.52489400e-01
2.83300132e-01 -1.04142332e+00 7.23216385e-02 -2.71782190e-01
-3.05111319e-01 1.67150128e+00 4.99154299e-01 -3.57881069e-01
5.86746216e-01 7.19206452e-01 1.98956430e-01 -2.93396413e-02
-7.64151096e-01 -2.00352699e-01 -2.98965245e-01 5.98849356e-01
6.67416751e-01 -2.11797521e-01 -5.83105028e-01 -1.18090115e-01
-7.80109018e-02 4.34149742e-01 9.46720243e-01 1.07307768e+00
-6.27446055e-01 -1.37293231e+00 -4.41587150e-01 -7.87144229e-02
-2.44038656e-01 2.37202868e-01 -2.40838587e-01 8.96075726e-01
-4.27106433e-02 7.69063950e-01 2.93421507e-01 -2.85181105e-01
1.65349990e-01 -1.47567660e-01 8.47566128e-01 -4.49264467e-01
-6.16265714e-01 1.53912365e-01 1.75889442e-03 -9.00308490e-01
-7.63707042e-01 -6.14684820e-01 -9.67591047e-01 -8.36489081e-01
-2.34083623e-01 -1.10500760e-01 8.98103118e-01 8.25511694e-01
3.80015045e-01 2.53490984e-01 5.42518735e-01 -1.13815498e+00
-1.67298630e-01 -5.03024101e-01 -5.74071765e-01 -9.86941680e-02
1.94200188e-01 -2.63070792e-01 -8.25750977e-02 3.73078763e-01] | [8.75910758972168, -1.5925190448760986] |
58383595-b6de-4201-9eb8-7c8266bb940b | multi-class-classification-of-vulnerabilities | 2004.00362 | null | https://arxiv.org/abs/2004.00362v1 | https://arxiv.org/pdf/2004.00362v1.pdf | Multi-Class classification of vulnerabilities in Smart Contracts using AWD-LSTM, with pre-trained encoder inspired from natural language processing | Vulnerability detection and safety of smart contracts are of paramount importance because of their immutable nature. Symbolic tools like OYENTE and MAIAN are typically used for vulnerability prediction in smart contracts. As these tools are computationally expensive, they are typically used to detect vulnerabilities until some predefined invocation depth. These tools require more search time as the invocation depth increases. Since the number of smart contracts is increasing exponentially, it is difficult to analyze the contracts using these traditional tools. Recently a machine learning technique called Long Short Term Memory (LSTM) has been used for binary classification, i.e., to predict whether a smart contract is vulnerable or not. This technique requires nearly constant search time as the invocation depth increases. In the present article, we have shown a multi-class classification, where we classify a smart contract in Suicidal, Prodigal, Greedy, or Normal categories. We used Average Stochastic Gradient Descent Weight-Dropped LSTM (AWD-LSTM), which is a variant of LSTM, to perform classification. We reduced the class imbalance (a large number of normal contracts as compared to other categories) by considering only the distinct opcode combination for normal contracts. We have achieved a weighted average Fbeta score of 90.0%. Hence, such techniques can be used to analyze a large number of smart contracts and help to improve the security of these contracts. | ['Raj Kishore', 'S. Swayamjyoti', 'Devadatta Sahoo', 'Kisor K. Sahu', 'Ajay K. Gogineni'] | 2020-03-21 | null | null | null | null | ['vulnerability-detection'] | ['miscellaneous'] | [-1.28207728e-01 -1.23155840e-01 -4.76946503e-01 -2.65321672e-01
-5.57906687e-01 -8.56328905e-01 4.39547151e-01 3.84970009e-02
-2.21598446e-01 2.78917670e-01 -2.44280532e-01 -9.80180681e-01
9.00992751e-02 -9.84517574e-01 -3.46089363e-01 -6.43674016e-01
-4.09667760e-01 4.18972999e-01 3.58888209e-01 -1.74515605e-01
4.06098485e-01 4.49869812e-01 -1.30332839e+00 7.47926652e-01
6.91777766e-01 1.11442888e+00 -4.22974497e-01 6.06494188e-01
-2.95025676e-01 8.12917173e-01 -7.85243690e-01 -6.87655032e-01
2.76071668e-01 -7.91560784e-02 -8.00037980e-01 -1.01207972e+00
-1.75204210e-03 -4.76096720e-01 -1.16046302e-01 1.32587111e+00
4.17968303e-01 -3.01405370e-01 2.36749694e-01 -1.34012604e+00
5.28622884e-03 1.32484734e+00 -5.75793028e-01 1.97664768e-01
2.02082917e-01 -1.65994652e-02 1.01128447e+00 -3.03316802e-01
2.73135185e-01 1.18401277e+00 6.88603580e-01 6.32962465e-01
-1.07877016e+00 -8.48459005e-01 -2.80819833e-02 5.00988662e-01
-1.04748833e+00 -1.46895489e-02 9.67560232e-01 -5.77426732e-01
1.40691650e+00 3.76180321e-01 2.91301250e-01 1.21384919e+00
5.58315575e-01 5.39371073e-01 8.74498963e-01 -2.38288939e-01
4.26313132e-01 1.15481600e-01 3.02214146e-01 3.59904945e-01
8.18255320e-02 5.29260099e-01 -1.77721784e-01 -8.10855031e-01
1.25005737e-01 2.99325347e-01 1.02118850e-01 1.39767826e-01
-6.24938726e-01 1.20005012e+00 6.64815009e-02 9.18002725e-01
-2.29409650e-01 3.67214561e-01 1.06790924e+00 6.40988231e-01
3.30261409e-01 4.60192859e-01 -9.65176284e-01 -5.58227479e-01
-9.17725623e-01 2.12129354e-01 1.06314707e+00 1.96819007e-01
3.19459498e-01 3.11474979e-01 5.44973128e-02 3.12817007e-01
1.84961528e-01 3.37333351e-01 6.64177656e-01 -8.52751195e-01
5.66151202e-01 9.13193405e-01 -3.99386734e-02 -1.12626171e+00
-3.85438323e-01 -4.53879714e-01 -7.76710868e-01 7.98500001e-01
3.07129771e-01 -1.81627825e-01 -5.29190779e-01 1.52604508e+00
1.39858618e-01 -3.56210731e-02 6.37225956e-02 3.20751697e-01
2.63687313e-01 4.84682918e-01 -7.64312297e-02 -1.96889982e-01
9.89211261e-01 -3.84242505e-01 -4.66679573e-01 3.73878092e-01
1.04663265e+00 -4.30670649e-01 9.48422670e-01 7.96457589e-01
-6.66786373e-01 -8.63341987e-02 -9.80314612e-01 7.38399029e-01
-7.26390839e-01 -4.07793462e-01 9.31815803e-01 1.29419541e+00
-7.74330854e-01 1.15529215e+00 -9.68415201e-01 3.82749617e-01
2.45541990e-01 5.34713626e-01 2.70326454e-02 6.88473582e-01
-1.54552758e+00 7.67319918e-01 3.63904148e-01 8.14962611e-02
-8.79327238e-01 -3.60552013e-01 -7.76160359e-01 2.57342160e-01
2.94388145e-01 2.95897573e-01 1.33706486e+00 -9.13358986e-01
-1.37972319e+00 4.40261751e-01 2.15045840e-01 -6.83655441e-01
5.46493173e-01 -3.34626026e-02 -5.09720623e-01 -2.09524333e-01
-4.49719459e-01 -2.31102601e-01 8.38003814e-01 -7.52532184e-01
-5.20224929e-01 -3.61062258e-01 4.03540283e-01 -5.60360253e-01
-6.22800231e-01 5.96109688e-01 4.85703945e-01 -4.60308373e-01
-7.44216666e-02 -9.12525654e-01 -1.14295311e-01 -7.89970875e-01
-3.88469577e-01 -4.21335012e-01 9.79973555e-01 -9.58225846e-01
1.46592915e+00 -1.98954809e+00 -1.02051206e-01 6.03780925e-01
2.41326261e-03 5.70513010e-01 9.41708311e-02 5.50529778e-01
-2.55179793e-01 4.14345711e-01 -3.23187411e-01 4.54114825e-02
3.00170720e-01 6.78511634e-02 -7.82924712e-01 4.00172710e-01
-4.18584824e-01 8.25661302e-01 -5.38728178e-01 -2.81353861e-01
6.37440607e-02 2.20844492e-01 -4.14511204e-01 1.69496939e-01
-3.93092752e-01 1.12678051e-01 -5.44153035e-01 9.82131839e-01
6.12882793e-01 2.48982385e-01 2.84485459e-01 7.33677298e-02
-5.69459237e-02 2.59777486e-01 -9.23788249e-01 1.22944665e+00
-8.29158843e-01 3.28993052e-01 -6.35508150e-02 -9.96907651e-01
9.29730654e-01 7.22161293e-01 3.81994903e-01 -7.32933998e-01
2.74021178e-01 5.14418244e-01 3.36525828e-01 -4.91028935e-01
2.65544350e-03 1.21531924e-02 -2.40270764e-01 8.75430405e-01
-5.12770057e-01 3.09277773e-01 9.84521806e-02 -8.65092278e-02
1.32576549e+00 5.02304696e-02 1.14471391e-01 -2.91932747e-02
5.62426567e-01 -3.47451419e-01 8.67162228e-01 5.21659970e-01
-9.04544219e-02 -1.11093096e-01 9.57200944e-01 -1.00633407e+00
-7.13465095e-01 -5.16750634e-01 8.26918483e-02 1.13646889e+00
-3.99595231e-01 -1.99632481e-01 -8.50694180e-01 -1.19474888e+00
3.88273746e-02 9.21886623e-01 -5.11903703e-01 -3.18166435e-01
-8.20768654e-01 -5.98418295e-01 9.85649705e-01 5.96566916e-01
3.29080939e-01 -1.41108775e+00 -1.16185534e+00 2.78361589e-01
2.79170901e-01 -6.52701080e-01 -1.56083226e-01 3.69284034e-01
-9.98436153e-01 -1.25278962e+00 -3.11220676e-01 -3.38077247e-01
2.77604192e-01 -6.08028173e-01 8.89710546e-01 3.08628261e-01
-2.62152940e-01 -1.90501213e-01 -4.30649161e-01 2.42010411e-02
-7.21269608e-01 1.29084989e-01 8.62913877e-02 -8.38011578e-02
3.76786500e-01 -8.35159063e-01 -2.55155146e-01 2.06306100e-01
-8.45067084e-01 -5.99460542e-01 5.87222338e-01 6.93724811e-01
6.82519153e-02 3.54307771e-01 5.98675609e-01 -9.68194783e-01
7.63034165e-01 -3.47069234e-01 -1.04972541e+00 3.75011951e-01
-9.74485576e-01 1.85837746e-01 1.08396661e+00 -5.41234791e-01
-9.05840993e-01 -7.93640614e-02 -3.49364907e-01 -4.93675053e-01
1.31935209e-01 6.74926996e-01 -1.81502670e-01 -3.44240993e-01
4.04426366e-01 5.27870841e-02 -4.50106710e-01 -6.29614472e-01
4.40456234e-02 8.16764534e-01 7.55133182e-02 -4.86340195e-01
7.27371335e-01 1.15820646e-01 -6.04193285e-02 -8.71652663e-02
9.98838339e-03 1.27259925e-01 -2.28261635e-01 -1.01817235e-01
4.50394660e-01 5.47132678e-02 -1.53896344e+00 7.55573928e-01
-1.35823905e+00 -5.47806144e-01 1.02400951e-01 1.48857877e-01
-1.22792952e-01 5.97693205e-01 -8.43742549e-01 -1.23216605e+00
-7.90191770e-01 -1.27235091e+00 4.73837644e-01 -1.06227674e-01
-4.75672483e-01 -1.00743639e+00 2.63495088e-01 -1.94906332e-02
5.71499169e-01 5.31887591e-01 1.27591276e+00 -1.14489305e+00
-4.35291417e-02 -7.04095304e-01 1.95324242e-01 5.62536657e-01
-4.36010994e-02 1.26897559e-01 -9.63233232e-01 -3.93383950e-01
5.07119894e-01 -2.95890778e-01 7.17547178e-01 1.18938982e-01
1.33858418e+00 -7.43516028e-01 -2.87022591e-01 5.42749286e-01
1.13617349e+00 9.87861335e-01 5.83480120e-01 4.12884980e-01
3.84547085e-01 6.65203094e-01 6.56085372e-01 4.45289522e-01
-1.76588699e-01 7.88668215e-01 6.92731023e-01 1.52624235e-01
7.07739770e-01 1.39635243e-03 7.44365752e-01 4.74358112e-01
-1.16765410e-01 1.22350059e-01 -1.28671992e+00 2.42016882e-01
-1.62414992e+00 -1.06869626e+00 -3.83015312e-02 2.27467895e+00
9.00289536e-01 4.83441085e-01 4.57349122e-02 6.80398941e-01
5.34581244e-01 1.75330102e-01 -5.43280303e-01 -1.21593475e+00
2.92291462e-01 2.67663240e-01 6.33362353e-01 5.10940015e-01
-9.86478984e-01 7.34004736e-01 5.77694035e+00 1.14658320e+00
-1.42171896e+00 2.60245740e-01 6.85724914e-01 -1.07442364e-02
-4.07269120e-01 1.76928565e-01 -5.49412906e-01 8.30352426e-01
1.21624017e+00 -5.28783426e-02 4.44840699e-01 1.27201569e+00
7.16770515e-02 -3.93718444e-02 -1.07641697e+00 7.65755773e-01
-2.49698773e-01 -1.01940835e+00 -3.13859105e-01 1.36236697e-01
3.72114837e-01 -1.83789060e-01 2.07200021e-01 3.78768414e-01
4.04550910e-01 -1.07065916e+00 5.36880076e-01 4.73232791e-02
6.70082867e-01 -1.01997185e+00 1.17106521e+00 3.04714352e-01
-1.03230393e+00 -6.22838497e-01 -2.98742186e-02 -9.16371942e-02
1.18345015e-01 7.02973723e-01 -5.95301867e-01 4.99605805e-01
7.91447580e-01 1.53995648e-01 -1.52726099e-01 6.17648363e-01
-1.28828838e-01 5.93839169e-01 -1.90294504e-01 -1.25516683e-01
1.48395017e-01 -5.92853688e-03 4.23895240e-01 9.55685854e-01
3.47943038e-01 -2.03668162e-01 6.83857352e-02 8.05826008e-01
2.71146327e-01 -2.57629186e-01 -5.21958172e-01 -3.06820035e-01
3.72649550e-01 1.04258657e+00 -7.45077610e-01 -3.68822664e-01
-2.81357408e-01 5.43586314e-01 -1.71992540e-01 2.79036552e-01
-1.01776195e+00 -8.71095955e-01 5.04429579e-01 -1.66583762e-01
2.04992726e-01 -3.97041393e-03 -5.04212499e-01 -8.71360242e-01
1.83588699e-01 -1.10006452e+00 6.81685805e-01 3.97515334e-02
-7.41139114e-01 9.62376654e-01 -7.61538967e-02 -1.17821848e+00
-8.07319999e-01 -4.16881055e-01 -9.86805737e-01 7.71915793e-01
-1.14556456e+00 -8.81350994e-01 2.75971472e-01 4.46719378e-01
3.24839681e-01 -5.57241499e-01 9.85498786e-01 3.86714071e-01
-5.14060557e-01 9.14113462e-01 2.05804165e-02 1.91920295e-01
4.42293473e-02 -1.10330045e+00 5.71037829e-01 7.34047651e-01
4.27832901e-02 8.03008676e-01 7.77028680e-01 -9.28542554e-01
-1.21675014e+00 -6.85968935e-01 8.91425133e-01 -8.27266946e-02
7.41732538e-01 -4.20667797e-01 -9.09387589e-01 6.01235688e-01
1.12936590e-02 -3.05775374e-01 5.65433860e-01 3.74343619e-02
-7.51788735e-01 -3.11457038e-01 -1.46663570e+00 3.96211326e-01
4.78492081e-01 -6.34478986e-01 -3.51290882e-01 1.36372661e-02
8.12147319e-01 -5.06555922e-02 -9.93740141e-01 5.66140354e-01
6.83061779e-01 -1.38491666e+00 6.56818151e-01 -3.88539463e-01
-2.62608137e-02 7.59460554e-02 3.62520404e-02 -8.96496475e-01
2.87744194e-01 -8.45422208e-01 -5.78283966e-01 1.19233871e+00
5.16008258e-01 -1.25763059e+00 9.35143650e-01 7.07655132e-01
6.53997362e-02 -1.09709692e+00 -1.32733440e+00 -1.19495153e+00
1.18509226e-01 -6.47086740e-01 8.56353283e-01 1.05981863e+00
5.06545603e-01 -4.55604821e-01 -2.84797013e-01 -2.21390471e-01
4.35605615e-01 5.15789866e-01 2.12794051e-01 -1.07650495e+00
-7.85001755e-01 -7.12709486e-01 -5.32827258e-01 -1.28625259e-01
4.31482732e-01 -7.36538947e-01 -4.64127511e-01 -8.63481164e-01
1.86776295e-02 -4.12212789e-01 -4.79598343e-01 1.02258766e+00
4.69689131e-01 -8.24063793e-02 -1.77444056e-01 3.16695385e-02
-4.79163788e-02 1.66624680e-01 4.41261053e-01 -4.89732802e-01
-3.86500895e-01 5.83690643e-01 -1.68718979e-01 8.38256836e-01
1.07682395e+00 -7.31272161e-01 -8.04894269e-02 -1.20561076e-02
6.93922877e-01 2.42355049e-01 6.83473516e-03 -6.57588661e-01
3.57413804e-03 -3.04810077e-01 -2.13275596e-01 -5.28290570e-01
4.35562171e-02 -7.45799601e-01 2.50416994e-01 1.27774417e+00
-3.56780738e-01 1.18463121e-01 6.94020465e-02 3.79739553e-02
-3.34651947e-01 -5.60389102e-01 4.26560313e-01 -5.62692918e-02
-3.96917611e-01 1.95140228e-01 -3.01137626e-01 -3.94238293e-01
1.00428820e+00 -7.51942396e-02 -1.30418867e-01 -3.22089970e-01
-5.25016487e-01 3.80052663e-02 4.22593474e-01 3.77608299e-01
5.97033501e-01 -1.16994011e+00 -4.50932264e-01 7.72667304e-02
-2.48979524e-01 -4.58607793e-01 2.27305681e-01 8.27612877e-01
-4.85450804e-01 4.45504218e-01 -3.65702733e-02 -1.55612335e-01
-1.75545013e+00 7.86335289e-01 3.74814093e-01 -5.98759294e-01
-6.31489515e-01 7.04152524e-01 -2.60435104e-01 -3.62366915e-01
6.23083889e-01 -4.38989192e-01 -4.32109386e-01 1.17246203e-01
6.62636578e-01 6.57857120e-01 1.98268190e-01 -2.99534678e-01
-5.83926439e-01 5.70624888e-01 -7.29586482e-02 -2.12549269e-01
1.38146496e+00 7.75129020e-01 -5.92173815e-01 3.21558446e-01
1.21078253e+00 1.03233583e-01 -8.00121486e-01 2.38088429e-01
7.05065966e-01 -5.63192368e-01 -1.10962488e-01 -8.22890460e-01
-1.38364851e+00 1.15577471e+00 7.69833028e-01 7.76918590e-01
1.07236195e+00 -1.98322192e-01 1.08918726e+00 4.29149806e-01
7.19437957e-01 -9.48858678e-01 -2.96918988e-01 5.32062471e-01
6.95269585e-01 -7.63089061e-01 -4.60826308e-01 1.33131608e-01
-3.99716645e-01 1.24786079e+00 2.34634399e-01 2.69245449e-02
3.12374204e-01 7.28493571e-01 3.26595604e-02 -1.82986781e-01
-6.86441123e-01 4.96983588e-01 -1.69286191e-01 4.36883450e-01
2.60828048e-01 1.72318950e-01 -4.11510617e-01 8.59857678e-01
-1.82647482e-01 -2.12795556e-01 5.13595223e-01 7.68800735e-01
-3.24956566e-01 -1.51533771e+00 -4.20583218e-01 4.27753747e-01
-9.16467547e-01 -2.20989183e-01 -4.14246589e-01 3.18061829e-01
1.11843556e-01 1.09357905e+00 -3.44467491e-01 -1.01030445e+00
1.06250472e-01 1.93529174e-01 3.65665965e-02 -3.57935309e-01
-1.19530940e+00 -3.44398230e-01 1.10490963e-01 -1.01043332e+00
2.16789499e-01 -4.89972442e-01 -1.29414308e+00 -5.17603815e-01
-3.06159765e-01 4.13100153e-01 8.30798388e-01 8.36588144e-01
8.00404847e-02 3.19567114e-01 9.69801664e-01 -6.43524408e-01
-1.15063441e+00 -8.91263902e-01 -6.55304193e-01 1.26983777e-01
1.85816616e-01 -5.74860811e-01 -6.86433077e-01 -4.68721628e-01] | [6.823615550994873, 7.379506587982178] |
d6ed21ed-611d-4991-a6ef-90df47cf2cda | graph-based-collaborative-ranking | 1604.03147 | null | http://arxiv.org/abs/1604.03147v3 | http://arxiv.org/pdf/1604.03147v3.pdf | Graph-based Collaborative Ranking | Data sparsity, that is a common problem in neighbor-based collaborative
filtering domain, usually complicates the process of item recommendation. This
problem is more serious in collaborative ranking domain, in which calculating
the users similarities and recommending items are based on ranking data. Some
graph-based approaches have been proposed to address the data sparsity problem,
but they suffer from two flaws. First, they fail to correctly model the users
priorities, and second, they cannot be used when the only available data is a
set of ranking instead of rating values. In this paper, we propose a novel
graph-based approach, called GRank, that is designed for collaborative ranking
domain. GRank can correctly model users priorities in a new tripartite graph
structure, and analyze it to directly infer a recommendation list. The
experimental results show a significant improvement in recommendation quality
compared to the state of the art graph-based recommendation algorithms and
other collaborative ranking techniques. | ['Haratizadeh Saman', 'Shams Bita'] | 2017-01-31 | null | null | null | null | ['collaborative-ranking'] | ['graphs'] | [-1.43018350e-01 -4.18076813e-01 -3.90332162e-01 -4.72774267e-01
-5.26548885e-02 -4.48031247e-01 1.28085881e-01 5.54749906e-01
-5.25256433e-02 4.94767219e-01 7.19133258e-01 -3.17809522e-01
-1.01642787e+00 -1.31640327e+00 -1.14548720e-01 -3.82327795e-01
-1.46477520e-01 6.33833230e-01 3.81488323e-01 -6.83140695e-01
6.91186130e-01 4.47495990e-02 -1.47681463e+00 6.76704168e-01
1.03019226e+00 7.25343287e-01 1.10871822e-01 2.69825071e-01
-4.00914490e-01 8.11533511e-01 -2.19330862e-01 -5.45867741e-01
2.46872067e-01 -5.49100816e-01 -5.15800536e-01 -1.50847360e-01
1.80308446e-01 -1.83040217e-01 -2.94239283e-01 1.14550114e+00
4.20989424e-01 4.44830507e-01 8.26160133e-01 -1.15545130e+00
-9.34731662e-01 9.28627908e-01 -5.49503505e-01 1.66782767e-01
5.69751680e-01 -8.60190213e-01 1.49276614e+00 -8.15037370e-01
2.76613086e-01 1.12691903e+00 5.60918927e-01 3.07501368e-02
-6.87771320e-01 -6.33461952e-01 5.48934996e-01 3.43586713e-01
-1.24511743e+00 8.34139064e-02 7.85002291e-01 -4.09895539e-01
4.07649249e-01 5.31839073e-01 9.82715368e-01 3.24702561e-01
-4.10073064e-02 3.86821538e-01 1.09600234e+00 -2.02281803e-01
2.10982844e-01 -1.07615612e-01 7.19637632e-01 5.40863991e-01
7.05775142e-01 -1.35124788e-01 -4.45812553e-01 -5.04696131e-01
7.11087346e-01 7.11261272e-01 -3.08955282e-01 -4.47487026e-01
-9.00742233e-01 1.05590725e+00 5.11138022e-01 4.64731008e-01
-2.67241240e-01 -4.51339304e-01 1.32751703e-01 7.01492071e-01
5.09375334e-01 3.72513294e-01 -2.58497477e-01 4.12653893e-01
-8.33792031e-01 2.10661501e-01 8.37875187e-01 7.50417650e-01
5.35184503e-01 -2.49839768e-01 -1.51393354e-01 8.90810966e-01
8.53469014e-01 8.74629095e-02 5.46980083e-01 -2.97107846e-01
4.43679839e-01 9.80052233e-01 4.57548574e-02 -1.73301399e+00
-3.70823592e-01 -6.56447172e-01 -9.05593693e-01 -8.87920782e-02
2.75795788e-01 -8.88913646e-02 -5.45518875e-01 1.03819811e+00
3.49577278e-01 3.41940701e-01 1.87256783e-02 1.17636240e+00
1.19735718e+00 6.82160795e-01 -1.90202415e-01 -6.06962204e-01
9.53293502e-01 -9.75480855e-01 -6.69193089e-01 1.91981420e-01
5.71818352e-01 -8.34347486e-01 8.57756257e-01 8.16873729e-01
-6.83334947e-01 -6.04955256e-01 -7.59528160e-01 3.26475799e-01
-3.00691843e-01 -1.70611829e-01 1.06521034e+00 6.76462293e-01
-1.02412808e+00 6.24151647e-01 -3.00096441e-03 -3.92422527e-01
-4.61031087e-02 5.26205361e-01 -5.71261011e-02 -4.74840611e-01
-1.36439276e+00 4.99172956e-01 3.91228855e-01 5.79835698e-02
-2.78988391e-01 -4.40768987e-01 -3.31757247e-01 1.71447068e-01
5.34238935e-01 -5.80894530e-01 5.99288881e-01 -8.92905056e-01
-1.24319375e+00 8.25823247e-02 2.82603562e-01 -3.12014483e-02
9.98379439e-02 -3.75039428e-02 -9.67838049e-01 -5.71396053e-01
-3.10803771e-01 -3.83936703e-01 5.73166370e-01 -1.23038518e+00
-8.35323036e-01 -5.17106712e-01 4.32571173e-01 5.88010669e-01
-8.47481370e-01 -6.55552223e-02 -6.39112175e-01 -7.34579682e-01
2.80030727e-01 -6.25671327e-01 -6.31370485e-01 -5.09705544e-01
-1.64837185e-02 -3.94848347e-01 4.53944027e-01 -3.40211540e-01
1.84777927e+00 -1.68495321e+00 2.42583498e-01 9.13343966e-01
3.50019157e-01 4.30339575e-01 -2.63154328e-01 8.45780909e-01
2.92505503e-01 7.11612329e-02 3.54772210e-01 2.39420369e-01
-3.34270746e-01 3.67042542e-01 -2.04410896e-01 1.86674401e-01
-8.30084145e-01 5.34735560e-01 -1.23983061e+00 -4.65719759e-01
8.11901465e-02 3.17960531e-01 -7.90980101e-01 2.44228601e-01
2.16556881e-02 3.18499625e-01 -8.50372910e-01 2.39558324e-01
6.38949037e-01 -2.29850665e-01 8.82087052e-01 -4.40880507e-01
-5.09830751e-03 3.51564705e-01 -1.70516193e+00 1.51655746e+00
-3.78659517e-01 -2.34214500e-01 -2.53862083e-01 -1.21361077e+00
1.16143715e+00 1.82184994e-01 5.80870390e-01 -4.47021484e-01
3.82033437e-02 1.05027750e-01 2.23739699e-01 -3.96279693e-01
5.77367425e-01 1.82200104e-01 2.07976222e-01 5.24688423e-01
-2.09718943e-01 2.75544077e-01 4.92339045e-01 5.51231384e-01
9.79395747e-01 -3.20073396e-01 3.33806127e-01 -3.63513708e-01
6.82369530e-01 1.10570356e-01 6.76430345e-01 6.19945824e-01
7.61971414e-01 5.20330489e-01 1.73076868e-01 -4.51369196e-01
-6.65157795e-01 -5.93831658e-01 3.57685953e-01 1.03047192e+00
4.21834588e-01 -1.08226919e+00 -4.86901067e-02 -9.65729058e-01
1.98383629e-01 2.54088491e-01 -3.78652424e-01 -2.07516834e-01
-2.29172483e-01 -7.50579357e-01 -2.21906379e-01 2.74210960e-01
1.49558753e-01 -5.98909557e-01 5.13050377e-01 4.60013032e-01
-1.82148769e-01 -5.02138674e-01 -5.75708747e-01 -3.21941972e-01
-1.26355708e+00 -1.24017394e+00 -5.43139279e-01 -7.37666905e-01
9.29174364e-01 9.68865395e-01 1.25787723e+00 7.65418530e-01
4.69390124e-01 2.44529054e-01 -1.12748146e+00 -7.75145218e-02
1.14701964e-01 -9.16053057e-02 1.80801511e-01 2.46023521e-01
3.25727433e-01 -7.21333683e-01 -6.89613521e-01 7.36926854e-01
-7.98326075e-01 -2.17429891e-01 5.52211702e-01 6.55280888e-01
6.74614251e-01 7.24573910e-01 8.15843463e-01 -1.59793687e+00
1.14392412e+00 -8.56863439e-01 -4.62792993e-01 4.90725100e-01
-1.25501192e+00 -1.47194535e-01 9.77800250e-01 -2.28461921e-01
-8.18894923e-01 -2.66759336e-01 -6.13350272e-02 1.59835666e-02
2.35282660e-01 1.20602202e+00 -4.53413129e-02 -2.00613186e-01
3.90478104e-01 -1.88346177e-01 -3.36177915e-01 -9.31467474e-01
3.75772476e-01 8.26605141e-01 -1.13129482e-01 -2.71191478e-01
6.05190516e-01 1.17424659e-01 8.07112008e-02 -3.81822258e-01
-1.13650811e+00 -1.01236570e+00 -4.05880988e-01 -3.55994523e-01
3.72624487e-01 -8.44286561e-01 -5.20778060e-01 6.32799044e-03
-6.30678058e-01 2.71048218e-01 2.82223932e-02 8.93747211e-01
7.40511194e-02 7.05664456e-01 -3.82181197e-01 -7.19413459e-01
-4.26449239e-01 -6.88624740e-01 1.11199401e-01 1.79581806e-01
1.57240734e-01 -1.14584613e+00 2.18563527e-01 6.62413418e-01
3.97371292e-01 -1.96869448e-01 9.56108928e-01 -8.66580546e-01
-2.28669122e-01 -2.45144233e-01 -2.04264030e-01 9.00147855e-02
3.04036558e-01 -1.63770303e-01 -3.00645772e-02 -5.50653577e-01
-3.46775144e-01 2.70974059e-02 5.40459812e-01 1.65721968e-01
1.03488147e+00 -4.48212326e-01 -3.30272108e-01 1.45496860e-01
1.54772627e+00 2.28188619e-01 5.89709044e-01 -1.68349952e-01
9.33236122e-01 4.96313572e-01 1.06261265e+00 5.33228099e-01
4.07581568e-01 7.19807386e-01 4.71950412e-01 -1.17524311e-01
-1.56789087e-02 -4.27162856e-01 2.54053809e-02 1.66802585e+00
-6.26130342e-01 -7.14656591e-01 -4.94484991e-01 2.46224985e-01
-2.47017956e+00 -8.16634297e-01 -8.80151570e-01 2.58090854e+00
4.15955245e-01 1.23018716e-02 4.03407574e-01 4.66274917e-01
8.08238924e-01 -2.28973348e-02 -1.32091239e-01 -2.94759810e-01
1.88328758e-01 7.77289867e-02 4.99004364e-01 5.53207099e-01
-6.95673168e-01 7.45432258e-01 6.21072578e+00 7.95991957e-01
-6.01753533e-01 1.19527303e-01 5.32628968e-02 2.24139586e-01
-6.99977875e-01 1.42520636e-01 -6.12174511e-01 4.16040957e-01
5.30026078e-01 -2.94608057e-01 5.01963079e-01 5.74510396e-01
2.11656809e-01 1.61766633e-01 -7.68686414e-01 9.63625491e-01
2.91351616e-01 -1.11317086e+00 4.19258267e-01 1.91775709e-01
9.68840182e-01 -3.94015014e-01 -9.96174812e-02 1.95277795e-01
8.00509334e-01 -8.37498605e-01 9.34955180e-02 7.33539402e-01
9.35680494e-02 -8.81057084e-01 9.80661750e-01 4.74799573e-01
-1.42696607e+00 -1.77659750e-01 -9.27990675e-01 -5.48516393e-01
4.60518077e-02 1.19899249e+00 -4.24630523e-01 1.07595885e+00
6.88393772e-01 1.19515610e+00 -3.29228938e-01 1.47265005e+00
-3.63032132e-01 8.69465113e-01 -1.98402163e-02 -3.77495617e-01
-5.79439513e-02 -7.82145262e-01 2.38867775e-01 7.90289521e-01
5.48474491e-01 5.08876801e-01 5.77110648e-01 4.93298657e-02
-5.85931428e-02 9.75653768e-01 -5.37791610e-01 2.53584106e-02
4.09412056e-01 1.35037875e+00 -7.50613868e-01 -2.70191550e-01
-6.70847654e-01 6.95930660e-01 3.14646244e-01 6.58021569e-02
-5.62365890e-01 -1.41130507e-01 2.91999489e-01 3.45780909e-01
2.45941401e-01 -3.19939882e-01 1.04685932e-01 -1.27628851e+00
-3.12950343e-01 -9.50662315e-01 7.94516385e-01 -5.65418065e-01
-1.62732351e+00 3.16737026e-01 -1.47921830e-01 -1.49736547e+00
2.12491378e-01 -2.91237205e-01 -6.51837945e-01 5.75756669e-01
-1.34161782e+00 -9.93557692e-01 -3.71984363e-01 1.03546691e+00
2.41423592e-01 -2.04669788e-01 6.78653955e-01 8.89597714e-01
-1.45337090e-01 5.38783073e-01 4.65830714e-01 -2.52707958e-01
6.58499062e-01 -9.77703691e-01 -1.74879029e-01 6.66057408e-01
5.56272686e-01 7.40488410e-01 4.46964115e-01 -8.03902328e-01
-1.62132096e+00 -9.61039305e-01 9.37791467e-01 -1.15513012e-01
4.58291888e-01 6.64656088e-02 -7.48055816e-01 3.12271595e-01
-1.56673267e-01 -4.32868451e-02 1.16916394e+00 6.99249566e-01
-2.41751656e-01 -2.92821348e-01 -1.02112782e+00 3.52636039e-01
1.47225761e+00 8.24116170e-02 -6.04266882e-01 5.70742369e-01
4.35058117e-01 1.86633408e-01 -1.13909686e+00 2.93454856e-01
6.31369591e-01 -1.12712133e+00 1.09674978e+00 -8.46595943e-01
1.66244730e-01 -7.28796422e-01 -2.15815559e-01 -1.60144532e+00
-1.15030265e+00 -1.06163891e-02 -9.73762050e-02 1.36323941e+00
4.10185426e-01 -4.81962562e-01 7.71716893e-01 1.97734997e-01
1.29659072e-01 -6.15860939e-01 -1.17212921e-01 -6.21984065e-01
-4.56282109e-01 -6.61836192e-02 8.63775194e-01 1.28116763e+00
1.97442234e-01 6.95511520e-01 -8.98314118e-01 8.39193910e-02
5.16754985e-01 5.12109339e-01 5.56516111e-01 -1.86889458e+00
-5.19342899e-01 1.60077913e-03 -5.68344831e-01 -1.01521170e+00
-3.89066160e-01 -1.07002020e+00 -5.71955025e-01 -2.26751471e+00
1.48826838e-01 -9.69084203e-01 -8.45619380e-01 2.59675205e-01
-1.19641930e-01 4.34357256e-01 7.98687432e-03 5.28582335e-01
-8.95005763e-01 3.19079198e-02 1.33105648e+00 5.71084842e-02
-3.46009761e-01 3.13180894e-01 -1.07803798e+00 6.62020206e-01
6.97360277e-01 -5.50236821e-01 -1.01316524e+00 -4.84900177e-01
1.01718354e+00 2.21115038e-01 -3.98800552e-01 -8.38238358e-01
5.16185224e-01 -6.21814311e-01 1.17620051e-01 -6.02792740e-01
-1.32103980e-01 -1.14350426e+00 6.58651888e-01 2.92571336e-01
-3.62080783e-01 8.85367617e-02 -7.53432333e-01 8.33956957e-01
-2.99967617e-01 -2.59386033e-01 2.82865435e-01 -1.46192953e-01
-8.73889506e-01 6.90112650e-01 -1.17802881e-01 -2.12394327e-01
6.49648607e-01 1.78893446e-03 -6.60674646e-02 -5.86557627e-01
-7.18330383e-01 2.45475560e-01 2.48840839e-01 5.09786785e-01
8.87586236e-01 -1.60767019e+00 -8.23541462e-01 -2.21726626e-01
2.96189576e-01 -3.81679237e-01 2.13954583e-01 6.17126882e-01
-1.98654130e-01 4.26945873e-02 -1.03362553e-01 6.35009557e-02
-1.48274732e+00 8.53836358e-01 -1.89453274e-01 -5.65673411e-01
-3.16498339e-01 6.50315464e-01 -1.64432287e-01 -3.79436165e-01
1.00628681e-01 9.29726064e-02 -1.23995006e+00 1.36473387e-01
5.27106047e-01 5.31426311e-01 2.41213411e-01 -5.80062151e-01
-2.76357681e-01 8.05938780e-01 -1.48902237e-01 3.40589195e-01
1.36294913e+00 -1.90187469e-01 -3.74311805e-01 1.76572919e-01
8.14485133e-01 5.38348854e-01 -1.73119649e-01 -4.91501749e-01
-1.75012916e-01 -9.41108644e-01 4.04857486e-01 -7.97538936e-01
-1.58450091e+00 3.39080781e-01 4.06855971e-01 6.54693723e-01
1.02574623e+00 -4.27309334e-01 8.13814878e-01 4.52177525e-01
6.65574551e-01 -9.89652038e-01 -1.82878211e-01 4.74341542e-01
6.84877336e-01 -1.08655906e+00 4.36397552e-01 -8.25144351e-01
-5.21486223e-01 1.10274255e+00 5.62107086e-01 -3.97612602e-01
1.22800148e+00 -2.80394495e-01 -1.68106735e-01 -3.30272496e-01
-6.26733065e-01 -4.12742376e-01 6.93049192e-01 4.73302692e-01
8.48182499e-01 1.22362889e-01 -1.24238539e+00 9.04394329e-01
1.15673311e-01 1.71217456e-01 3.26879114e-01 7.62544751e-01
-8.11467290e-01 -1.61985731e+00 -2.74993658e-01 1.06937969e+00
-3.54652941e-01 -2.46105477e-01 -5.00067651e-01 8.64905193e-02
3.68307769e-01 1.61304867e+00 -3.04610640e-01 -9.80740786e-01
4.99546260e-01 -5.30116141e-01 2.39285767e-01 -8.31829250e-01
-7.00073302e-01 -8.41985941e-02 3.11627567e-01 -3.77551049e-01
-6.45847321e-01 -2.00253487e-01 -1.15483999e+00 -5.26055634e-01
-8.40590596e-01 1.05208528e+00 5.02336800e-01 6.48591399e-01
4.62289125e-01 4.23859417e-01 9.82050896e-01 -3.72904509e-01
-2.34605268e-01 -6.38396502e-01 -1.07649982e+00 6.32607639e-01
-3.50486666e-01 -8.14685106e-01 -2.81352192e-01 -4.30973619e-01] | [10.083142280578613, 5.7137370109558105] |
e715bcf3-9d1d-4ead-8d5e-fc99b4aad399 | the-surprising-creativity-of-digital | 1803.03453 | null | https://arxiv.org/abs/1803.03453v4 | https://arxiv.org/pdf/1803.03453v4.pdf | The Surprising Creativity of Digital Evolution: A Collection of Anecdotes from the Evolutionary Computation and Artificial Life Research Communities | Biological evolution provides a creative fount of complex and subtle adaptations, often surprising the scientists who discover them. However, because evolution is an algorithmic process that transcends the substrate in which it occurs, evolution's creativity is not limited to nature. Indeed, many researchers in the field of digital evolution have observed their evolving algorithms and organisms subverting their intentions, exposing unrecognized bugs in their code, producing unexpected adaptations, or exhibiting outcomes uncannily convergent with ones in nature. Such stories routinely reveal creativity by evolution in these digital worlds, but they rarely fit into the standard scientific narrative. Instead they are often treated as mere obstacles to be overcome, rather than results that warrant study in their own right. The stories themselves are traded among researchers through oral tradition, but that mode of information transmission is inefficient and prone to error and outright loss. Moreover, the fact that these stories tend to be shared only among practitioners means that many natural scientists do not realize how interesting and lifelike digital organisms are and how natural their evolution can be. To our knowledge, no collection of such anecdotes has been published before. This paper is the crowd-sourced product of researchers in the fields of artificial life and evolutionary computation who have provided first-hand accounts of such cases. It thus serves as a written, fact-checked collection of scientifically important and even entertaining stories. In doing so we also present here substantial evidence that the existence and importance of evolutionary surprises extends beyond the natural world, and may indeed be a universal property of all complex evolving systems. | ['Westley Weimer', 'François Taddei', 'Anh Nguyen', 'Jean-Baptiste Mouret', 'David E. Moriarty', 'Sara Mitri', 'Risto Miikkulainen', 'Carlos Maestre', 'Hod Lipson', 'Richard E. Lenski', 'Laurent Keller', 'Christian Gagné', 'Antoine Frénoy', 'Stephanie Forrest', 'Kai Olav Ellefsen', 'Stephane Doncieux', 'Samuel Bernard', 'Joel Lehman', 'Thomas S. Ray', 'Simon Thibault', 'Julie Beaulieu', 'Jason Yosinski', 'Eric Shulte', 'Charles Ofria', 'Carole Knibbe', 'Stephan Fischer', 'Robert Feldt', 'Fred C. Dyer', 'Danesh Tarapore', 'Christoph Adami', 'Robert T. Pennock', 'Robert MacCurdy', 'Peter J. Bentley', 'Patryk Chrabaszcz', 'Marc Schoenauer', 'Marc Parizeau', 'Leni Le Goff', 'Karl Sims', 'Guillaume Beslon', 'Frank Hutter', 'Dusan Misevic', 'David M. Bryson', 'William F. Punch', 'Richard Watson', 'Peter Krcah', 'Nick Cheney', 'Lee Altenberg', 'Laura M. Grabowski', 'Kenneth O. Stanley', 'Jeff Clune', 'David Parsons', 'Babak Hodjat', 'Antoine Cully'] | 2018-03-09 | null | null | null | null | ['artificial-life'] | ['miscellaneous'] | [ 2.57403702e-01 1.55946359e-01 2.32341126e-01 2.51348644e-01
2.15224504e-01 -9.35472250e-01 7.69665956e-01 1.71084628e-01
-2.36042783e-01 9.51422453e-01 3.13818038e-01 -3.65734696e-01
-3.52527834e-02 -6.86682880e-01 -7.20008790e-01 -7.12347746e-01
-8.35060477e-02 2.94903189e-01 7.17810392e-02 -6.70669973e-01
5.73746860e-01 2.30011851e-01 -1.66617060e+00 -3.01504433e-01
9.14300084e-01 2.40499496e-01 1.04630865e-01 7.40584433e-01
-1.47123054e-01 8.03910494e-01 -9.49513912e-01 -8.71346474e-01
4.27573882e-02 -1.08641183e+00 -5.71500182e-01 -6.08971901e-02
-1.51770175e-01 6.30354956e-02 -1.31457141e-02 1.02539754e+00
3.41747642e-01 -3.44578892e-01 2.41632119e-01 -8.05439532e-01
-1.28953874e+00 5.82330287e-01 -3.84722918e-01 1.95578992e-01
3.61377090e-01 7.95711875e-01 7.06550837e-01 -4.07914907e-01
9.73072708e-01 9.52104270e-01 9.43426073e-01 5.14953256e-01
-1.26235533e+00 -4.05171305e-01 -2.42897823e-01 -2.28174835e-01
-1.21310592e+00 -4.59938794e-01 5.23145139e-01 -7.97053218e-01
8.00576627e-01 5.59217989e-01 1.59580994e+00 1.32713139e+00
7.82834530e-01 2.22951576e-01 9.46729422e-01 -2.98833579e-01
3.90000880e-01 2.03083605e-01 -5.04655063e-01 6.41712368e-01
8.16186726e-01 2.91417599e-01 -6.80752337e-01 -3.45109224e-01
7.01943219e-01 -1.17065854e-01 -5.86921930e-01 1.65175050e-01
-1.41547406e+00 4.56896424e-01 -9.30859614e-03 7.70815194e-01
-3.60343099e-01 4.68528569e-01 3.62662673e-01 3.76181036e-01
3.40287656e-01 1.03231895e+00 -5.67013502e-01 -8.13901961e-01
-5.34923553e-01 7.24422708e-02 9.18767571e-01 4.46011364e-01
2.46839508e-01 -5.01235984e-02 6.25597477e-01 4.41602647e-01
3.68460752e-02 1.08116105e-01 8.75137448e-01 -8.79698575e-01
-6.71799898e-01 8.13721955e-01 4.11661826e-02 -1.49301577e+00
-1.51868891e-02 -6.90487862e-01 -4.83032316e-01 6.49228454e-01
4.99275655e-01 -1.47906661e-01 -2.68836737e-01 1.82274234e+00
3.64838988e-01 -2.70974278e-01 -1.65656842e-02 8.75776350e-01
3.53549898e-01 5.84532022e-01 -1.99672788e-01 -2.63111204e-01
1.24273503e+00 -4.81632113e-01 -4.98107761e-01 -5.91772795e-02
2.49388054e-01 -8.84103477e-01 1.27643728e+00 6.22011542e-01
-1.21619964e+00 5.14952838e-02 -1.24929690e+00 2.37076968e-01
-2.62719512e-01 -7.19436765e-01 9.20598090e-01 9.69044685e-01
-9.62830365e-01 9.71050084e-01 -8.04230630e-01 -8.36430967e-01
4.62965429e-01 -1.87993124e-01 -5.79375867e-03 5.75540721e-01
-6.50994301e-01 9.79109824e-01 1.93882704e-01 -2.75534559e-02
-4.52861935e-01 -6.83327436e-01 -4.15772870e-02 -2.19196543e-01
5.46208739e-01 -1.00982738e+00 9.26311135e-01 -1.64276528e+00
-1.36998117e+00 1.09907663e+00 -1.34593487e-01 -1.74377143e-01
7.01355934e-01 1.26680091e-01 -5.86742342e-01 -1.92734838e-01
-1.07263513e-01 8.65727141e-02 5.20596266e-01 -1.37920511e+00
-1.98140845e-01 -3.46964806e-01 -1.28322691e-01 -1.45421416e-01
-2.81344056e-01 4.71041538e-02 2.44394079e-01 -8.80573630e-01
4.37139682e-02 -5.86479425e-01 -2.14268506e-01 4.61386889e-01
-2.04940885e-01 4.92751002e-02 3.76821905e-01 -4.33441401e-01
1.21063411e+00 -2.17553592e+00 1.92679524e-01 -3.53983879e-01
5.73192775e-01 -8.73578638e-02 3.56672645e-01 9.12468970e-01
3.37805569e-01 8.22644174e-01 -4.41115797e-01 3.77777874e-01
-2.65071634e-02 1.16792105e-01 -2.59769440e-01 5.11384428e-01
7.85216922e-04 1.02703047e+00 -1.31208599e+00 -1.74899757e-01
-4.75221932e-01 5.29376209e-01 -3.57711017e-01 -1.88824400e-01
-3.02542180e-01 5.01280665e-01 -3.11381668e-01 6.25109375e-01
5.53169623e-02 -5.13009727e-01 2.69805759e-01 5.05339921e-01
-6.28030300e-01 2.01752987e-02 -6.02833450e-01 1.48121774e+00
-9.29523632e-02 9.60977554e-01 -2.58050203e-01 -4.82416004e-01
7.44533956e-01 3.27473134e-01 1.37097642e-01 -3.85531425e-01
2.85701483e-01 5.68795085e-01 4.76752251e-01 -6.84002101e-01
6.98612571e-01 -5.47787905e-01 1.28444523e-01 1.06231797e+00
-3.31482738e-01 -4.78900045e-01 6.29684925e-02 -7.37706944e-02
1.30364478e+00 3.26729178e-01 6.88467920e-01 -2.85215974e-01
-2.52333358e-02 3.32900375e-01 6.63392007e-01 5.60896039e-01
-1.52965516e-01 5.26862979e-01 3.10810775e-01 -7.87197292e-01
-1.45275509e+00 -1.03610551e+00 -2.21712008e-01 5.87332904e-01
2.09875166e-01 -4.43353981e-01 -5.44428229e-01 1.86306879e-01
-1.91125005e-01 7.15830803e-01 -7.32709825e-01 -4.70910192e-01
-2.39871681e-01 -8.40244055e-01 5.48537791e-01 -1.69754356e-01
2.61490494e-01 -1.14431119e+00 -1.57159770e+00 4.10314798e-01
2.28631869e-01 -4.65116441e-01 -1.63068831e-01 -1.17624350e-01
-8.06095719e-01 -8.49074900e-01 -5.89341402e-01 -3.89020383e-01
6.06081963e-01 5.33423014e-02 1.11398959e+00 6.85217917e-01
-8.04358363e-01 1.16238855e-01 -2.95180023e-01 -5.95164895e-01
-1.02535188e+00 -3.64270419e-01 3.82089464e-04 -2.68286526e-01
1.48638338e-01 -1.14364696e+00 -4.61545169e-01 6.13780282e-02
-9.33139622e-01 -4.03250232e-02 4.58506584e-01 7.58961380e-01
1.41004890e-01 1.61903158e-01 7.43299544e-01 -8.02838147e-01
7.02324808e-01 -5.70767701e-01 -2.06097618e-01 3.65440100e-01
-6.64303839e-01 -1.70596316e-01 6.45081460e-01 -7.22863019e-01
-7.50740170e-01 -6.13242090e-01 2.72799104e-01 3.21551144e-01
-2.53028031e-02 4.77651328e-01 2.42415711e-01 -1.29309604e-02
1.06516826e+00 4.36992854e-01 8.10607970e-02 -3.07815462e-01
1.51933074e-01 4.68790591e-01 4.85603780e-01 -5.07572412e-01
8.46808195e-01 7.23546386e-01 -3.92432541e-01 -1.17646885e+00
-1.62597567e-01 4.61342067e-01 -1.62128910e-01 -5.11008441e-01
7.12639630e-01 -3.87328297e-01 -6.89703405e-01 3.70880812e-01
-1.14670348e+00 -2.57159740e-01 -9.35382247e-01 4.74796556e-02
-4.16050971e-01 1.36383131e-01 -3.52650821e-01 -1.03807414e+00
-8.30340907e-02 -7.28633046e-01 3.50594014e-01 5.06114125e-01
-9.63824272e-01 -8.65980089e-01 3.64119709e-01 4.70652282e-02
8.07200432e-01 8.56386662e-01 1.03493810e+00 -2.45347947e-01
-6.77360952e-01 -2.44426444e-01 3.97885442e-01 -1.75483078e-01
3.64322037e-01 7.08039820e-01 -8.24463487e-01 1.26743808e-01
2.95043290e-01 -1.52128369e-01 2.71210283e-01 -2.17300966e-01
5.26003659e-01 -6.97891593e-01 -2.98665434e-01 4.05664682e-01
1.56467700e+00 6.06956899e-01 5.79483747e-01 6.34456694e-01
7.83300698e-02 7.25804627e-01 -1.36694640e-01 6.14000797e-01
8.83420110e-02 3.05977523e-01 2.31574386e-01 1.46942854e-01
1.54417098e-01 -2.99319863e-01 3.75381351e-01 8.55396628e-01
-4.49767381e-01 -3.12945336e-01 -1.15979326e+00 5.84495127e-01
-1.52552772e+00 -1.40834105e+00 -2.26500526e-01 2.31988287e+00
1.05753839e+00 3.46736938e-01 2.23153815e-01 -2.71424316e-02
6.52589619e-01 -9.85418260e-02 -7.67749965e-01 -6.73813939e-01
-4.79690999e-01 -9.06375889e-03 5.44235250e-03 1.20919190e-01
-1.94137886e-01 7.57941008e-01 7.02739763e+00 2.93080479e-01
-1.18286777e+00 2.11339980e-01 4.99294102e-01 -3.00358057e-01
-9.07962620e-01 2.86103755e-01 1.15678467e-01 6.47229910e-01
7.94580042e-01 -9.02925611e-01 6.25303388e-01 5.04787624e-01
2.22308457e-01 -3.39604288e-01 -7.72446036e-01 4.86860067e-01
-1.22317031e-01 -1.46380043e+00 -2.55933344e-01 1.67739481e-01
6.59822047e-01 -9.75355878e-02 1.39833137e-01 -6.06358171e-01
6.31339550e-01 -1.22513902e+00 1.34607077e+00 5.30015409e-01
4.47730333e-01 -5.82802474e-01 1.80090174e-01 3.34928840e-01
-3.71579617e-01 -7.02280253e-02 -2.54760057e-01 -5.61110914e-01
4.08546507e-01 7.67234087e-01 -4.35147971e-01 -8.72975811e-02
7.08521128e-01 4.33748305e-01 -5.01901805e-01 1.13927758e+00
-2.07681909e-01 6.55812979e-01 -1.84451267e-01 -6.09507263e-01
-6.53168708e-02 -1.55308485e-01 1.18796957e+00 9.17232096e-01
4.66102570e-01 2.54662454e-01 -8.23470712e-01 1.32821858e+00
3.16908419e-01 -4.83449623e-02 -8.49795640e-01 -1.13627732e+00
4.65266228e-01 9.80974734e-01 -1.14834917e+00 -1.02873603e-02
-1.52223974e-01 1.10480738e+00 -1.67381510e-01 1.47729456e-01
-8.10091913e-01 -2.22483590e-01 8.83494258e-01 1.41458452e-01
-3.36551592e-02 -3.74547929e-01 -7.50884652e-01 -1.00135684e+00
5.02347276e-02 -1.02128160e+00 -3.08630288e-01 -6.59809649e-01
-1.22102177e+00 6.29414856e-01 -6.63581729e-01 -7.95594752e-01
7.04900548e-03 -4.66433913e-02 -7.05719531e-01 2.56294519e-01
-6.18717253e-01 -5.95157206e-01 -6.39460906e-02 -1.14083320e-01
2.48454586e-01 -1.94027778e-02 6.87242329e-01 -2.22909927e-01
-3.66306126e-01 3.86969298e-01 4.21081245e-01 -5.27696848e-01
2.82459527e-01 -7.00528145e-01 5.86510718e-01 6.97755933e-01
2.35014334e-01 9.23543453e-01 1.26312768e+00 -6.98413134e-01
-1.56477177e+00 -2.89156556e-01 8.05312812e-01 -6.17659628e-01
9.66079116e-01 -2.97463745e-01 -9.43873644e-01 5.06839395e-01
4.10775274e-01 -7.13111103e-01 7.13829041e-01 -2.26626173e-02
-3.67849559e-01 4.26368594e-01 -1.07951665e+00 1.12079048e+00
1.31863439e+00 -2.76880950e-01 -5.24891853e-01 2.51522988e-01
1.07242000e+00 9.75709632e-02 -6.25763178e-01 -1.53176025e-01
1.00078893e+00 -1.19149649e+00 7.21524358e-01 -5.64578593e-01
9.33466673e-01 -2.89288253e-01 2.66455468e-02 -1.03400159e+00
-1.58957258e-01 -1.03639817e+00 5.07047415e-01 1.42231405e+00
6.58533931e-01 -1.00875378e+00 4.49143082e-01 5.56778491e-01
-2.67730415e-01 -7.57387578e-01 -7.98693180e-01 -9.17878330e-01
2.64413476e-01 -1.93262845e-01 7.82239974e-01 1.52975941e+00
2.48487532e-01 -1.73325725e-02 9.06800330e-02 -4.29417163e-01
4.90649015e-01 3.01116593e-02 7.57492661e-01 -1.15948701e+00
-7.51887858e-01 -1.02271378e+00 -6.47214472e-01 -9.81576815e-02
-5.79273701e-01 -6.81647360e-01 1.18232463e-02 -9.65505481e-01
3.75614017e-01 -2.83607781e-01 3.54144841e-01 7.14699998e-02
1.47816176e-02 2.29883045e-01 7.57159665e-02 5.30683160e-01
-1.10292420e-01 2.63205498e-01 1.35219383e+00 1.85878724e-01
-3.59514236e-01 -5.68666875e-01 -1.28304446e+00 9.39766586e-01
7.09919274e-01 -5.34435987e-01 -2.10323825e-01 -3.09172571e-01
1.00561404e+00 -3.26510072e-01 6.19296432e-01 -9.31373835e-01
1.59098893e-01 -5.15515089e-01 1.41992778e-01 2.86897659e-01
5.22569716e-02 -6.21215701e-01 1.12185764e+00 9.08689260e-01
-2.73859575e-02 2.07575083e-01 1.13403246e-01 5.31059563e-01
2.42346391e-01 -1.39381707e-01 6.05272472e-01 -5.74377000e-01
-4.22818363e-01 -3.09376389e-01 -8.76537323e-01 1.82386562e-01
1.18570209e+00 -8.02258968e-01 -6.54118180e-01 -3.58326912e-01
-2.79185742e-01 -2.40022629e-01 1.53438604e+00 1.59539089e-01
4.43645716e-01 -7.73990273e-01 -6.62736952e-01 -1.93122149e-01
-3.34234387e-02 -2.53941834e-01 1.92705438e-01 7.85648882e-01
-9.19535458e-01 -2.34860584e-01 -4.95829731e-01 -2.80000359e-01
-7.45432734e-01 6.45266175e-01 2.62765795e-01 5.72934151e-01
-9.82836962e-01 1.08022392e+00 1.20070748e-01 -2.29925364e-02
-2.22853750e-01 2.00383320e-01 4.26939666e-01 8.14829860e-03
6.15751445e-01 2.26204813e-01 -5.48539519e-01 -4.49254572e-01
-2.90926069e-01 3.90219092e-01 2.75642395e-01 -3.20763648e-01
1.74776638e+00 -1.58976033e-01 -6.98755622e-01 1.08398509e+00
6.27450168e-01 1.72917560e-01 -9.91419673e-01 3.71466011e-01
-5.36849573e-02 -6.64590478e-01 -5.48829257e-01 -1.06356931e+00
-6.94090247e-01 6.79627419e-01 9.33491066e-02 6.02098167e-01
1.02409840e+00 1.94706500e-01 7.65472353e-01 6.72722310e-02
6.28474116e-01 -9.36285257e-01 1.82278767e-01 -5.96018396e-02
1.03133202e+00 -7.19959259e-01 -4.18373384e-02 8.94395709e-02
-4.23834473e-01 1.03075421e+00 2.55750149e-01 -4.19865437e-02
1.87679827e-01 3.49981308e-01 -1.95534497e-01 -3.03664386e-01
-1.08906722e+00 1.63029179e-01 -4.73211527e-01 5.46383739e-01
7.77104974e-01 -1.67686178e-03 -1.10284126e+00 5.34523010e-01
-7.34400392e-01 1.85527205e-01 1.00611997e+00 1.06277037e+00
-7.53220022e-01 -8.38615835e-01 -5.43422282e-01 -9.82463732e-02
-5.40156782e-01 6.93629980e-02 -1.06063509e+00 7.49462783e-01
3.65125775e-01 5.83989263e-01 -1.45754501e-01 -5.60530901e-01
-1.78499684e-01 -1.13807786e-02 5.57374656e-01 -2.25678071e-01
-9.27050650e-01 -3.55184138e-01 2.38003619e-02 -1.82915881e-01
-1.65053946e-03 -8.77268136e-01 -1.20563769e+00 -1.06816018e+00
-1.86438158e-01 1.12339258e-01 7.45948315e-01 7.04120159e-01
5.04984677e-01 3.81658167e-01 1.82688639e-01 -5.23354888e-01
-6.20255172e-02 -2.74751455e-01 -5.52323878e-01 1.09478585e-01
2.70790100e-01 -2.32747197e-01 -4.43355531e-01 1.52276561e-01] | [5.571763515472412, 4.202932357788086] |
160d17ae-865a-4705-b9af-45039c60a40c | improving-image-recognition-by-retrieving | 2304.05173 | null | https://arxiv.org/abs/2304.05173v1 | https://arxiv.org/pdf/2304.05173v1.pdf | Improving Image Recognition by Retrieving from Web-Scale Image-Text Data | Retrieval augmented models are becoming increasingly popular for computer vision tasks after their recent success in NLP problems. The goal is to enhance the recognition capabilities of the model by retrieving similar examples for the visual input from an external memory set. In this work, we introduce an attention-based memory module, which learns the importance of each retrieved example from the memory. Compared to existing approaches, our method removes the influence of the irrelevant retrieved examples, and retains those that are beneficial to the input query. We also thoroughly study various ways of constructing the memory dataset. Our experiments show the benefit of using a massive-scale memory dataset of 1B image-text pairs, and demonstrate the performance of different memory representations. We evaluate our method in three different classification tasks, namely long-tailed recognition, learning with noisy labels, and fine-grained classification, and show that it achieves state-of-the-art accuracies in ImageNet-LT, Places-LT and Webvision datasets. | ['Cordelia Schmid', 'Alireza Fathi', 'Ahmet Iscen'] | 2023-04-11 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Iscen_Improving_Image_Recognition_by_Retrieving_From_Web-Scale_Image-Text_Data_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Iscen_Improving_Image_Recognition_by_Retrieving_From_Web-Scale_Image-Text_Data_CVPR_2023_paper.pdf | cvpr-2023-1 | ['learning-with-noisy-labels', 'learning-with-noisy-labels'] | ['computer-vision', 'natural-language-processing'] | [ 3.16139698e-01 -2.71750093e-01 -3.39224339e-01 -3.81851673e-01
-1.00509977e+00 -1.79855600e-01 8.29225898e-01 6.71974719e-02
-7.70695806e-01 5.81842780e-01 3.05782706e-01 7.76675493e-02
-9.91040990e-02 -5.71603358e-01 -8.63451898e-01 -7.28637636e-01
3.35472226e-01 7.44530916e-01 4.30428505e-01 1.30808726e-01
5.46611905e-01 4.16431665e-01 -1.87886429e+00 6.34555638e-01
4.65215147e-01 1.28828263e+00 5.77177167e-01 4.37100530e-01
-1.08928196e-01 1.13457334e+00 -5.40863276e-01 -2.21764758e-01
-8.02276731e-02 4.68965806e-02 -8.88902366e-01 2.21451111e-02
9.28596199e-01 -4.02990133e-01 -7.14841664e-01 8.59352052e-01
5.38612187e-01 5.06880105e-01 6.86610401e-01 -1.11325598e+00
-1.20229805e+00 3.73225749e-01 -4.78042066e-01 4.55873251e-01
3.14207263e-02 -4.75736596e-02 1.00536847e+00 -1.43445587e+00
7.45560706e-01 1.35804522e+00 2.32583299e-01 5.05307913e-01
-1.02888143e+00 -7.28461742e-01 3.93790007e-01 7.07479417e-01
-1.32080185e+00 -4.70137775e-01 5.28166354e-01 -2.87965655e-01
1.33271348e+00 9.79208667e-03 1.73558325e-01 1.11525345e+00
-7.54655302e-02 1.21618855e+00 1.08138454e+00 -5.22660732e-01
-4.07924950e-02 -1.45490933e-02 6.98926568e-01 6.73747957e-01
6.05767267e-03 1.17575124e-01 -8.63692284e-01 -7.98657984e-02
4.11794156e-01 3.49422246e-01 -3.84329498e-01 -3.59657109e-01
-1.15165222e+00 7.56287873e-01 6.58479333e-01 2.10717589e-01
-2.42626667e-01 3.37728590e-01 1.59770980e-01 1.32884204e-01
5.67659497e-01 2.27428690e-01 -3.59738261e-01 1.70584932e-01
-8.65335047e-01 -9.82026663e-03 4.90638316e-01 1.08753765e+00
8.66108716e-01 -2.34961569e-01 -5.38400590e-01 1.18370950e+00
3.51740330e-01 7.27255881e-01 8.89989257e-01 -6.95579231e-01
3.84019911e-01 5.46429217e-01 -8.57465342e-02 -9.15183425e-01
-1.86984539e-01 -4.48759824e-01 -8.31730247e-01 -1.15129612e-01
2.51736082e-02 6.79469824e-01 -1.38813484e+00 1.58403111e+00
-2.28928238e-01 2.75017917e-01 -6.16839714e-03 9.69640255e-01
1.24677873e+00 7.42804050e-01 3.30278516e-01 -6.30427375e-02
1.10556912e+00 -1.64970279e+00 -4.43564326e-01 -6.17429554e-01
3.40888858e-01 -6.22359455e-01 1.09674335e+00 1.93833098e-01
-9.59984660e-01 -7.56498158e-01 -7.98462868e-01 -3.74901712e-01
-6.05176747e-01 2.06242830e-01 2.82368392e-01 6.41646609e-02
-1.04649389e+00 5.53099632e-01 -4.28674996e-01 -2.51396805e-01
5.85700631e-01 2.38136351e-01 -4.50666159e-01 -5.60898900e-01
-1.00142813e+00 1.10922909e+00 2.99628675e-01 -9.36840251e-02
-1.16196680e+00 -4.15313244e-01 -7.52404869e-01 3.16061914e-01
3.28736603e-01 -5.14046550e-01 1.17478466e+00 -1.04898024e+00
-8.44464898e-01 1.21678507e+00 -2.53036886e-01 -4.26084250e-01
-5.21772634e-03 -4.42641020e-01 -2.01465279e-01 1.39605284e-01
1.50078326e-01 1.14938724e+00 1.15456712e+00 -1.05140162e+00
-5.56598902e-01 -5.67432404e-01 -2.05423906e-01 1.95139304e-01
-5.52314579e-01 -1.05155267e-01 -9.22061980e-01 -7.66008019e-01
6.71569109e-02 -1.05491543e+00 -4.63318601e-02 3.80235799e-02
-1.52755678e-02 -2.87471414e-01 8.18304002e-01 -5.04765987e-01
8.73971224e-01 -2.22920394e+00 2.08034843e-01 -8.64636153e-02
5.94212636e-02 4.46946442e-01 -6.20266676e-01 9.69935879e-02
1.63403094e-01 1.60941854e-03 -1.19478330e-02 -5.49334049e-01
-5.55894151e-02 3.54599476e-01 -5.70232928e-01 4.63640466e-02
1.97732449e-01 1.17083800e+00 -8.34942758e-01 -6.07185543e-01
1.92977302e-02 5.08261144e-01 -2.36708820e-01 2.63950974e-01
-2.84125954e-01 9.48313251e-02 -4.28595871e-01 6.59110904e-01
3.33987832e-01 -6.63891196e-01 -1.32468671e-01 -2.13412210e-01
3.37790161e-01 1.97653294e-01 -7.08836913e-01 1.67059124e+00
-3.92667174e-01 7.39242315e-01 -2.15735286e-01 -7.81129658e-01
8.64393413e-01 -5.71133196e-02 -1.46091104e-01 -1.27455771e+00
-1.14824036e-02 9.95312929e-02 -2.94729143e-01 -4.77877825e-01
7.43531048e-01 1.97358429e-01 2.46816218e-01 5.41548967e-01
2.88646072e-01 3.33676457e-01 3.27115506e-02 2.91517049e-01
9.18667555e-01 4.79701906e-02 2.17153475e-01 -3.71038504e-02
4.08826649e-01 3.96300890e-02 1.34686828e-01 1.15141678e+00
-2.16286957e-01 6.73266947e-01 -2.20820263e-01 -5.21566153e-01
-7.34325945e-01 -8.19785714e-01 -1.77994594e-02 1.69602096e+00
3.08234245e-01 -1.45628184e-01 -1.00497447e-01 -7.60539591e-01
1.85017794e-01 6.05835676e-01 -7.04840899e-01 -4.58101094e-01
-4.46419805e-01 -6.57144606e-01 3.33951384e-01 9.25665617e-01
7.62105167e-01 -1.50111914e+00 -5.45439184e-01 -1.34461671e-01
-3.34229022e-01 -9.79625404e-01 -5.03141165e-01 3.59222502e-01
-1.03573716e+00 -1.02543437e+00 -7.68126965e-01 -1.09697652e+00
7.67635524e-01 7.20428228e-01 1.45764101e+00 3.57486546e-01
-4.51724827e-01 6.38079047e-01 -3.44776064e-01 -9.13235992e-02
-1.17378356e-02 1.30817279e-01 -2.91744858e-01 -1.37719214e-01
5.54041028e-01 -7.15679973e-02 -5.13823032e-01 3.95283431e-01
-9.06332254e-01 -1.91730186e-02 9.27385032e-01 1.29812670e+00
9.38729763e-01 -3.84474009e-01 7.20343888e-01 -7.87407279e-01
4.27136213e-01 -3.46903384e-01 -3.47578704e-01 4.58334059e-01
-6.80278182e-01 3.63696098e-01 1.26061678e-01 -6.75320029e-01
-9.90625739e-01 1.25243023e-01 7.94194192e-02 -7.64135301e-01
-1.22657098e-01 5.21587133e-01 -2.02718321e-02 -1.05246954e-01
4.69420522e-01 4.85871136e-01 -3.13525528e-01 -6.59174562e-01
5.05881548e-01 5.97816408e-01 5.32001972e-01 -5.84547639e-01
1.49117798e-01 2.82458633e-01 -3.07370365e-01 -6.22562230e-01
-1.13117218e+00 -6.78918302e-01 -4.82025653e-01 -8.02390184e-03
5.05788505e-01 -9.41848814e-01 -3.27323347e-01 4.19215679e-01
-1.08516896e+00 -2.11670265e-01 -2.34441251e-01 3.11199039e-01
-3.98333102e-01 7.19645321e-02 -6.38557732e-01 -5.54423630e-01
-4.35882658e-01 -1.31256795e+00 1.28648305e+00 1.42093018e-01
1.39699757e-01 -6.31532133e-01 1.16801150e-01 5.98663509e-01
5.24821639e-01 -5.00969648e-01 1.06948912e+00 -7.76450634e-01
-1.03226888e+00 -5.02141826e-02 -5.99818766e-01 3.08258325e-01
-2.77583033e-01 -4.42462087e-01 -1.36466277e+00 -3.70865822e-01
-2.35217094e-01 -8.21078241e-01 1.86921000e+00 1.64295316e-01
1.31423724e+00 -1.75329134e-01 -5.57101965e-01 3.92423302e-01
1.31815112e+00 1.18216619e-01 7.38406718e-01 5.32749832e-01
5.78893065e-01 4.86259222e-01 7.62822211e-01 -6.23143883e-03
1.65631592e-01 5.90983033e-01 6.17665470e-01 1.66156724e-01
-4.15874958e-01 -1.29237697e-01 1.68044657e-01 5.85463703e-01
2.39477769e-01 -2.79544920e-01 -8.21536243e-01 8.92213285e-01
-2.10292840e+00 -8.63159180e-01 2.81932384e-01 2.06347966e+00
6.90319777e-01 -3.67439874e-02 -4.18192327e-01 -2.59175718e-01
6.47816956e-01 2.93202788e-01 -6.79332733e-01 8.92919979e-06
-3.28793883e-01 2.59943873e-01 2.71143854e-01 2.69524038e-01
-1.28425062e+00 1.12453890e+00 6.54981470e+00 9.16753769e-01
-1.00198889e+00 2.73046851e-01 8.57439220e-01 -3.89826477e-01
1.42486900e-01 -3.02746922e-01 -9.67073143e-01 1.49631456e-01
6.94182575e-01 8.69563594e-02 2.31542051e-01 1.12230158e+00
-5.60940981e-01 -1.83234543e-01 -1.32398903e+00 1.14799178e+00
6.27219319e-01 -1.34082258e+00 5.00167191e-01 -2.24108279e-01
7.37235308e-01 4.90907401e-01 4.75668430e-01 5.98569095e-01
1.48310224e-02 -1.24696410e+00 5.37120283e-01 8.03866446e-01
5.45505166e-01 -5.23038208e-01 7.45729029e-01 3.26725543e-01
-7.84403205e-01 -1.46976873e-01 -6.63359046e-01 3.22797410e-02
-3.59211892e-01 3.16222727e-01 -6.06037796e-01 -2.32909359e-02
1.01914549e+00 7.65756667e-01 -1.05059648e+00 1.02984154e+00
-1.98002234e-01 3.78571242e-01 -3.57373580e-02 -1.47156194e-01
2.18456581e-01 3.78651589e-01 2.36062407e-01 1.20125425e+00
6.91178665e-02 4.32409495e-02 1.90275505e-01 7.01142311e-01
-5.81180453e-01 3.29744257e-02 -6.99375689e-01 -2.71705110e-02
4.67417538e-01 1.28730083e+00 -4.77223366e-01 -6.10557258e-01
-4.27122891e-01 9.25114870e-01 8.78882289e-01 3.78193438e-01
-5.20674825e-01 -8.85023549e-02 3.20847899e-01 -2.58315772e-01
6.41082287e-01 1.62557494e-02 6.44279197e-02 -1.02500772e+00
2.92689465e-02 -8.32650721e-01 5.84363043e-01 -1.18331850e+00
-1.40300119e+00 7.50490606e-01 -1.42652482e-01 -8.70635450e-01
-1.69628084e-01 -8.05057406e-01 -1.31805792e-01 7.49803901e-01
-1.79153037e+00 -1.27265990e+00 -5.15130281e-01 6.51231647e-01
7.60225236e-01 -4.30020392e-01 8.02821815e-01 3.57608706e-01
-2.14799091e-01 4.27914143e-01 1.94830552e-01 5.29434569e-02
1.03414309e+00 -8.21265340e-01 1.38405427e-01 5.02813756e-01
6.47892058e-01 6.64913952e-01 2.48262897e-01 -5.52788615e-01
-1.38287354e+00 -1.29063654e+00 1.03642333e+00 -5.74138463e-01
4.21562999e-01 2.30163578e-02 -1.01913798e+00 8.38104725e-01
2.36842707e-01 1.56376347e-01 4.27022785e-01 4.71692383e-02
-9.05249119e-01 2.53578126e-02 -8.89647365e-01 3.98434639e-01
1.09858775e+00 -7.29995966e-01 -9.91697371e-01 3.52682501e-01
7.20018089e-01 -1.06632620e-01 -4.41146851e-01 4.72390682e-01
5.71836472e-01 -7.02678859e-01 1.14555585e+00 -8.57783258e-01
4.85375404e-01 -9.40256268e-02 -4.90220040e-01 -1.24488771e+00
-5.29195666e-01 3.43449652e-01 -2.64043063e-01 1.01130223e+00
3.98395240e-01 -3.88019890e-01 7.52174795e-01 5.17185271e-01
-7.26941526e-02 -6.57173634e-01 -8.19104135e-01 -6.51304007e-01
-1.16748318e-01 -2.37095833e-01 2.85623968e-01 7.14457273e-01
-3.17095578e-01 7.02863693e-01 -5.19135594e-01 -6.03829045e-03
5.78421056e-01 6.47374690e-01 4.90259558e-01 -1.17948699e+00
-3.48906070e-01 -4.41949874e-01 -3.33808273e-01 -1.33899915e+00
4.90578502e-01 -1.03275537e+00 1.80221483e-01 -1.48372066e+00
7.85131991e-01 -3.30940366e-01 -7.34902799e-01 7.57319272e-01
-2.95668632e-01 6.30005419e-01 4.44951206e-01 6.32464468e-01
-1.28916073e+00 6.31002009e-01 1.02742481e+00 -7.63672352e-01
1.81695789e-01 -3.57942194e-01 -5.26991963e-01 6.78996742e-01
4.75400031e-01 -5.23610830e-01 -2.74691999e-01 -9.03189957e-01
6.14897683e-02 -2.88956314e-01 6.72329485e-01 -8.73163700e-01
5.18766701e-01 1.78903475e-01 6.72017217e-01 -1.00766432e+00
6.64315879e-01 -8.64619374e-01 -2.51308590e-01 2.44971335e-01
-7.73124278e-01 6.64405525e-02 3.05542260e-01 7.62471437e-01
-4.65383857e-01 -3.72682214e-01 6.57782674e-01 -2.39206880e-01
-1.23361659e+00 3.70226949e-01 -1.03147384e-02 1.27469659e-01
6.31777287e-01 1.72285557e-01 -9.21347737e-01 -3.06093097e-01
-8.80087852e-01 3.75558197e-01 3.78015667e-01 7.41403282e-01
1.03998053e+00 -1.41538012e+00 -4.92054939e-01 1.86891630e-01
6.62059426e-01 -3.50640506e-01 1.89724416e-01 6.21189535e-01
-6.70965435e-03 7.63489842e-01 -2.22615212e-01 -7.53823221e-01
-1.53169811e+00 9.77985799e-01 1.31249189e-01 -2.71764249e-01
-4.89842772e-01 9.06887949e-01 5.42389333e-01 -2.28932470e-01
6.76819444e-01 -2.07263082e-01 -3.13259959e-01 -7.55823450e-03
8.28031361e-01 2.99442649e-01 2.22514287e-01 -7.69047797e-01
-4.30270523e-01 6.44177496e-01 -4.96080428e-01 7.32536241e-02
1.27819884e+00 -1.01206668e-01 -6.28995672e-02 5.22952676e-01
1.16811192e+00 -5.54050207e-01 -1.01968086e+00 -7.52411664e-01
2.26067066e-01 -4.22873169e-01 2.29838043e-01 -9.83840704e-01
-1.33408213e+00 1.11082911e+00 7.63675332e-01 -2.97040492e-01
1.12791300e+00 2.16530547e-01 6.19677067e-01 9.80108440e-01
4.57235128e-01 -1.06979573e+00 4.72116530e-01 7.84303606e-01
1.08227909e+00 -1.52331090e+00 -1.11949861e-01 -7.83627853e-03
-5.33949733e-01 8.27787399e-01 7.72027791e-01 -1.17999375e-01
5.51273704e-01 -8.30819681e-02 -1.87976733e-02 -1.86142832e-01
-1.11238611e+00 -4.63223606e-01 5.54210782e-01 4.70987439e-01
2.83424497e-01 -2.22072750e-01 6.71299174e-02 5.33696234e-01
4.39940780e-01 -7.15380535e-02 -2.50812974e-02 1.06722236e+00
-7.61035144e-01 -7.90188253e-01 -3.16647887e-01 5.62560320e-01
-2.14023113e-01 -4.88103032e-01 -7.21427619e-01 7.07247376e-01
-2.21579999e-01 7.87899554e-01 1.26289830e-01 -2.46810570e-01
1.81312174e-01 5.15202582e-01 5.08429289e-01 -5.87729216e-01
-5.27815223e-01 3.21251191e-02 -3.05974949e-02 -6.85688794e-01
-6.09638870e-01 -1.84797570e-01 -9.93532598e-01 4.03800160e-02
-4.73336220e-01 -2.94902287e-02 5.38590968e-01 8.81524861e-01
6.37217045e-01 4.99769598e-01 1.04430959e-01 -8.98940206e-01
-5.32356560e-01 -1.01897192e+00 -3.74119282e-01 6.47344649e-01
2.89959490e-01 -8.95943940e-01 -2.67968863e-01 -1.13465860e-02] | [10.234721183776855, 1.784932255744934] |
7cd5675d-e0e3-4954-9118-7077b8329b01 | machine-learning-for-subgroup-discovery-under | 1902.10327 | null | http://arxiv.org/abs/1902.10327v1 | http://arxiv.org/pdf/1902.10327v1.pdf | Machine learning for subgroup discovery under treatment effect | In many practical tasks it is needed to estimate an effect of treatment on
individual level. For example, in medicine it is essential to determine the
patients that would benefit from a certain medicament. In marketing, knowing
the persons that are likely to buy a new product would reduce the amount of
spam. In this chapter, we review the methods to estimate an individual
treatment effect from a randomized trial, i.e., an experiment when a part of
individuals receives a new treatment, while the others do not. Finally, it is
shown that new efficient methods are needed in this domain. | ['Aleksey Buzmakov'] | 2019-02-27 | null | null | null | null | ['subgroup-discovery'] | ['methodology'] | [ 2.74687290e-01 3.56610566e-01 -6.80312514e-01 -5.21221817e-01
-1.96762860e-01 -4.31418568e-01 3.32513869e-01 5.82646906e-01
-6.91722214e-01 1.10055637e+00 -2.54759956e-02 -3.98317724e-01
-2.99838245e-01 -1.00799274e+00 -1.00941360e+00 -7.19141185e-01
6.90556364e-03 8.08336973e-01 -1.08764045e-01 -1.34278107e-02
3.62993091e-01 4.19893533e-01 -1.21601582e+00 3.97602051e-01
9.65491831e-01 7.04458475e-01 -8.57225209e-02 3.78825366e-01
9.74252746e-02 1.40191093e-01 -6.84419036e-01 -5.60270607e-01
3.59657556e-01 -5.44460356e-01 -6.93817258e-01 3.44380647e-01
4.89824384e-01 -4.67434645e-01 1.04959488e-01 1.26224458e+00
3.75374943e-01 6.23082705e-02 1.05813050e+00 -1.07097113e+00
-2.60669053e-01 9.69432116e-01 -6.78395927e-01 -3.46559994e-02
3.79455686e-01 3.30661274e-02 9.46095586e-01 -1.16337679e-01
5.59423327e-01 1.64048874e+00 2.39221886e-01 5.32037616e-01
-1.43995094e+00 -8.39111269e-01 2.15089962e-01 -2.09435225e-01
-7.27256894e-01 -1.45977125e-01 6.04503930e-01 -4.72309202e-01
8.33775178e-02 2.29391888e-01 6.42871678e-01 1.10090959e+00
6.04113758e-01 8.70625377e-01 1.60539842e+00 -2.45138466e-01
5.97105205e-01 5.25394261e-01 5.56836903e-01 3.00012439e-01
8.09999883e-01 4.14232790e-01 -2.37557992e-01 -6.25997424e-01
5.20440876e-01 2.53503084e-01 -1.00626446e-01 -6.16875589e-01
-6.32639468e-01 1.33969688e+00 3.61430258e-01 4.42824781e-01
-6.96684539e-01 1.18827321e-01 2.86349028e-01 3.63873839e-01
5.22232771e-01 6.18040442e-01 -7.65605628e-01 3.18409413e-01
-6.75382197e-01 6.49376988e-01 1.05057693e+00 3.39666843e-01
3.29649180e-01 -6.56036079e-01 -4.78470139e-03 4.31491077e-01
5.31275477e-03 5.64760506e-01 2.74825931e-01 -9.22183812e-01
1.47343323e-01 4.42792475e-01 3.61948222e-01 -6.75975680e-01
-3.34730297e-01 -4.00363535e-01 -8.46700132e-01 2.40253612e-01
9.03833270e-01 -3.84669036e-01 -8.04745913e-01 1.73968089e+00
4.24662173e-01 2.50950851e-03 -4.23184723e-01 7.39944756e-01
2.11788788e-01 2.10803732e-01 3.93869102e-01 -6.96366251e-01
1.47169983e+00 -3.07756662e-01 -9.11308408e-01 -2.58026540e-01
5.38819313e-01 -6.91210449e-01 7.00086176e-01 7.67520249e-01
-1.11930811e+00 -1.74597353e-01 -6.18775845e-01 6.51290894e-01
-4.64993060e-01 -2.59731561e-01 1.04724717e+00 1.07412755e+00
-3.88721436e-01 1.18264854e+00 -3.36814940e-01 -5.08633614e-01
7.53392339e-01 7.24817812e-01 -1.20538786e-01 -1.64232701e-01
-1.46901560e+00 8.26677501e-01 9.59113985e-02 -2.29431689e-01
-7.59304523e-01 -7.75809228e-01 -3.35121930e-01 1.85045823e-01
4.73282993e-01 -9.78435934e-01 1.25047386e+00 -1.00731313e+00
-1.16729844e+00 7.66687453e-01 1.64472982e-01 -6.75907612e-01
6.78656995e-01 -1.48975357e-01 4.92934436e-02 -1.10787794e-01
3.31947133e-02 5.03468692e-01 1.23071730e+00 -9.23066616e-01
-7.72862494e-01 -9.39741075e-01 1.01827905e-01 1.58173367e-02
-2.70814478e-01 -6.22355826e-02 3.03296447e-01 -4.96138304e-01
-2.84379482e-01 -1.01719832e+00 -7.20642328e-01 -2.57538885e-01
-5.19602358e-01 -4.68448758e-01 5.14257967e-01 -5.17534912e-01
7.08444357e-01 -1.97835445e+00 2.67604645e-02 4.09984648e-01
4.91552740e-01 -3.33975032e-02 1.18413135e-01 3.53593022e-01
-1.77910298e-01 4.08470154e-01 4.17533480e-02 3.89494181e-01
1.54278763e-02 -5.64586967e-02 -2.35044926e-01 6.82711542e-01
-1.32978544e-01 4.29759264e-01 -7.77775049e-01 -3.03309619e-01
3.89800332e-02 8.29116628e-02 -6.13095582e-01 -2.03749448e-01
-1.85811639e-01 3.20512354e-01 -1.04258382e+00 2.63445944e-01
8.19600880e-01 -3.73120457e-01 3.34324986e-01 8.32976997e-02
4.89775389e-01 1.57332182e-01 -1.07309783e+00 9.27903950e-01
-2.76302248e-01 1.25209659e-01 2.73079932e-01 -1.27475929e+00
4.05500889e-01 2.82344401e-01 4.83985156e-01 -2.38415539e-01
5.48972547e-01 2.44761705e-01 2.32870251e-01 -3.19654614e-01
5.55180497e-02 -7.17216730e-01 -1.73435375e-01 5.52351415e-01
-4.66985017e-01 -6.71533197e-02 2.27033615e-01 7.21763968e-02
1.22700369e+00 -6.28103733e-01 7.28944063e-01 -3.93086761e-01
2.45688960e-01 1.60827618e-02 3.99489224e-01 1.18559945e+00
-2.41974071e-01 -2.18716800e-01 6.60685778e-01 -2.46308476e-01
-7.32088029e-01 -9.55558956e-01 -4.31761235e-01 6.57609284e-01
3.49204987e-02 2.29026079e-01 -9.29644704e-01 -1.03331339e+00
5.34175038e-01 9.46004033e-01 -8.44695210e-01 -1.88520432e-01
-7.53668845e-02 -1.08700526e+00 -3.14810634e-01 9.81067717e-02
4.10096884e-01 -7.55993962e-01 -2.16390386e-01 5.47569394e-02
-1.74069609e-02 -5.47793746e-01 -5.19500256e-01 -1.35544213e-02
-1.50987291e+00 -1.09130359e+00 -8.09562147e-01 -4.59615320e-01
8.83345783e-01 1.34976223e-01 9.45934415e-01 4.99161240e-03
-3.06705356e-01 1.16975836e-01 -1.34874731e-01 -7.70086348e-01
-3.92560095e-01 -1.11470111e-01 2.93389946e-01 -7.83279762e-02
7.32476473e-01 -5.15202463e-01 -7.05165565e-01 2.51171052e-01
-5.46950877e-01 -5.41284561e-01 8.94392371e-01 9.37317789e-01
1.16418764e-01 5.97025752e-01 6.31373525e-01 -1.68170452e+00
9.98309910e-01 -4.56877947e-01 -5.62291384e-01 1.92821369e-01
-7.19084740e-01 4.68699411e-02 4.74110872e-01 -9.88484800e-01
-8.30722213e-01 5.58612108e-01 4.27415609e-01 2.91688234e-01
-2.17032120e-01 1.80587903e-01 -2.11442620e-01 9.89734828e-02
8.75431597e-01 -3.32172900e-01 1.06473744e-01 -5.98845243e-01
4.92865920e-01 9.09463167e-01 -3.48028898e-01 -4.67529386e-01
4.39163715e-01 3.50254953e-01 7.39809945e-02 -7.68641710e-01
-1.03670943e+00 -3.15427244e-01 -2.71173835e-01 -2.49957796e-02
7.42827356e-01 -4.50181723e-01 -1.36296189e+00 1.49488196e-01
-1.00074506e+00 -1.75765976e-01 -3.91173035e-01 8.50986779e-01
-5.65476120e-01 7.23816305e-02 -4.64017063e-01 -8.49148095e-01
-1.25270069e-01 -8.84453714e-01 7.64791846e-01 2.64357001e-01
-4.67244744e-01 -9.52666640e-01 -1.83126986e-01 6.16036475e-01
-2.93730926e-02 -1.61079794e-01 8.92044723e-01 -1.09515107e+00
-5.65069377e-01 -7.87600875e-01 1.07927704e-02 2.07597047e-01
3.37791115e-01 -4.76221234e-01 -5.47793031e-01 -5.36573946e-01
5.63970506e-01 -1.12029798e-01 5.97519755e-01 1.38499629e+00
9.50906217e-01 -5.01692832e-01 -7.75211394e-01 -2.13974401e-01
9.78142679e-01 5.34023106e-01 4.46619838e-01 -1.15897492e-01
3.81651260e-02 1.04416442e+00 8.15562785e-01 4.12712842e-01
-2.94819355e-01 5.19117773e-01 -7.31099444e-03 -1.00629300e-01
2.83456653e-01 -2.80210614e-01 2.11269632e-01 1.15129575e-01
2.18133569e-01 -2.76758939e-01 -1.50272891e-01 2.26480842e-01
-1.74013233e+00 -7.90500164e-01 -2.44207114e-01 2.56888437e+00
9.07426775e-01 3.92066687e-01 3.62629801e-01 -5.80425113e-02
1.05070019e+00 -4.60395902e-01 -6.97492659e-01 -2.85483122e-01
3.20739239e-01 7.38511384e-02 1.04622889e+00 5.96258819e-01
-1.00716734e+00 6.43423200e-01 7.67056465e+00 1.15823853e+00
-3.99345607e-01 -3.06072482e-03 1.10896552e+00 1.45430878e-01
-5.78391040e-03 1.70440480e-01 -1.04361629e+00 5.48499405e-01
1.00177467e+00 -4.60685730e-01 3.32606435e-01 8.35950434e-01
5.79201281e-01 -6.63970649e-01 -1.20780540e+00 5.86582482e-01
-1.40534207e-01 -7.89513707e-01 -1.55833066e-01 5.35515845e-01
6.44797504e-01 -8.80167842e-01 9.74664316e-02 -4.29761894e-02
8.03455114e-01 -8.42929661e-01 -1.88310862e-01 4.51999605e-01
6.42119944e-02 -5.28370380e-01 9.68525827e-01 6.45780802e-01
-1.93811253e-01 -8.28378126e-02 -4.80178386e-01 -3.09213489e-01
2.42139950e-01 1.28419065e+00 -9.42670941e-01 6.04471974e-02
3.29370648e-01 3.59580994e-01 -2.82307386e-01 1.09879553e+00
-3.73304069e-01 6.38783991e-01 -4.55956548e-01 -5.06847620e-01
-1.32520422e-01 -4.61799353e-01 2.84671038e-01 4.43503141e-01
2.40861699e-01 4.39719856e-01 6.64351657e-02 6.53869033e-01
-1.15936793e-01 4.43617225e-01 -8.45001698e-01 -2.83909619e-01
9.45955515e-02 9.91889179e-01 -9.06402171e-01 -3.71532381e-01
-2.07369983e-01 8.57697189e-01 -3.02283078e-01 1.08540758e-01
-3.72202426e-01 -3.00603986e-01 3.70089233e-01 5.17412484e-01
4.35117930e-02 5.33687651e-01 -5.07959962e-01 -7.03984857e-01
-4.26794082e-01 -1.05042970e+00 4.93404388e-01 -4.80156064e-01
-1.70446157e+00 -6.01980239e-02 4.21446376e-02 -9.93420839e-01
-1.21452585e-01 -6.07098222e-01 -2.13716969e-01 8.00152004e-01
-4.47495490e-01 -5.91384113e-01 5.10257185e-01 1.21125564e-01
4.18178171e-01 2.52318755e-02 3.56649816e-01 2.09841326e-01
-3.15031111e-01 2.59854585e-01 4.77848090e-02 -2.33571365e-01
6.65681601e-01 -1.13263977e+00 -2.23671257e-01 3.18884611e-01
-2.52646565e-01 8.70072246e-01 1.33431327e+00 -1.09118104e+00
-9.50901687e-01 -3.86890054e-01 1.21995687e+00 -4.02577817e-01
6.98282957e-01 -1.52850717e-01 -5.48225284e-01 4.55041170e-01
7.63300881e-02 -6.69327974e-01 5.12697279e-01 4.63185668e-01
1.70214519e-01 -2.19084755e-01 -1.63987160e+00 7.53190160e-01
7.38733411e-01 -3.29144746e-02 -6.57945335e-01 1.04285312e+00
5.32919049e-01 1.55201569e-01 -8.29812765e-01 1.00455545e-01
5.38027942e-01 -7.32404649e-01 1.04243994e+00 -1.01197708e+00
2.99056143e-01 2.30844662e-01 3.50704491e-01 -1.53663230e+00
-1.57649502e-01 -4.57822114e-01 2.63348639e-01 7.04862952e-01
4.99692559e-01 -7.69738138e-01 1.14531207e+00 9.40567911e-01
7.34825790e-01 -4.58058953e-01 -7.13986695e-01 -9.44991469e-01
3.35144699e-01 1.97320610e-01 3.01667720e-01 9.52673614e-01
2.34127820e-01 6.76206410e-01 -5.70254982e-01 3.60588916e-03
1.10255122e+00 4.91809584e-02 6.08500183e-01 -1.54302049e+00
-3.88411760e-01 -2.94298500e-01 -2.15094134e-01 -1.05924690e+00
1.24938719e-01 -5.80531359e-01 -1.57437474e-01 -1.32610643e+00
7.96713471e-01 -1.59656227e-01 -7.98592344e-02 6.16957955e-02
-2.47126654e-01 -4.41557229e-01 -5.69017045e-02 -8.24968517e-02
-5.61553910e-02 2.51677006e-01 1.45174909e+00 -2.26983100e-01
-3.52927148e-01 9.54672515e-01 -9.15574312e-01 9.26360488e-01
9.38315451e-01 -1.09340954e+00 -3.06826741e-01 5.54170072e-01
1.99473724e-01 5.23988903e-01 7.25258812e-02 -2.45935991e-01
2.95811594e-02 -5.71410120e-01 3.30169469e-01 -4.97518063e-01
1.08591117e-01 -9.49838519e-01 2.53574312e-01 1.26230204e+00
-6.21269464e-01 -4.20603096e-01 -3.12156677e-01 8.37325275e-01
1.30884454e-01 -9.28996861e-01 7.82286048e-01 -2.25818828e-01
1.37671709e-01 7.65544176e-02 -4.53335851e-01 -1.32904217e-01
9.69611168e-01 2.69898295e-01 -2.61334211e-01 -6.25081599e-01
-1.07409370e+00 1.34177104e-01 2.43873864e-01 2.40023434e-02
3.44240546e-01 -1.13668466e+00 -5.60682595e-01 -2.95924842e-01
-3.24021429e-01 -8.62124085e-01 3.66921246e-01 9.38212097e-01
-2.27644995e-01 2.52603948e-01 -1.32041723e-01 -1.26948832e-02
-1.54890776e+00 9.96849060e-01 -3.99491489e-02 -4.76396948e-01
-1.22577749e-01 6.92845225e-01 5.83258569e-01 -1.58733726e-01
1.57037809e-01 -1.18896179e-01 -2.69795001e-01 7.49307722e-02
6.02156520e-01 4.51272964e-01 -2.38127664e-01 9.33218747e-02
-2.01694846e-01 3.28188241e-01 -5.24347842e-01 -2.55638391e-01
1.26811457e+00 9.43549499e-02 -3.20022345e-01 4.03364599e-01
9.38085675e-01 5.15551083e-02 -7.61209011e-01 -6.74671307e-02
-8.21288489e-03 -8.58012080e-01 9.98453349e-02 -9.36063051e-01
-1.07295263e+00 6.81906104e-01 7.72029400e-01 6.24663591e-01
1.03166437e+00 3.45297121e-02 6.01692259e-01 4.31766272e-01
5.16074061e-01 -1.21538794e+00 -2.22384408e-01 -3.08286816e-01
5.34351826e-01 -1.26812387e+00 6.54052854e-01 -8.40661883e-01
-5.39461970e-01 5.82157314e-01 2.21550036e-02 -3.27942401e-01
1.01329255e+00 -1.94375604e-01 -4.43909168e-01 -4.03560549e-01
-5.48251510e-01 -4.97531854e-02 1.99612588e-01 4.84987348e-01
3.53467584e-01 5.80049336e-01 -1.39796114e+00 5.24904191e-01
2.60622706e-02 3.07643294e-01 7.39032149e-01 6.57444954e-01
-5.09270251e-01 -1.47432566e+00 -5.23413837e-01 1.17609107e+00
-6.04431272e-01 5.31056989e-03 -7.87805617e-01 6.88129246e-01
6.09293096e-02 1.08286071e+00 -2.76221573e-01 -1.26206204e-01
4.83436137e-01 -6.98148459e-02 6.37252152e-01 -5.76037407e-01
-5.90104461e-01 1.70852721e-01 9.82008800e-02 -3.53005290e-01
-5.57424605e-01 -9.45468605e-01 -7.27336645e-01 -6.75103903e-01
-5.76159239e-01 3.08135450e-01 5.63859344e-01 8.38918924e-01
-1.53637856e-01 3.08128357e-01 7.62673140e-01 -1.99565679e-01
-9.52321768e-01 -9.02286410e-01 -1.26666820e+00 5.07053256e-01
-1.43177673e-01 -6.72687054e-01 -8.03593636e-01 -6.89431280e-02] | [8.174763679504395, 5.282687664031982] |
def28e47-700c-42c9-94cf-b24aa04657d6 | accurate-gaze-estimation-using-an-active-gaze | 2301.13186 | null | https://arxiv.org/abs/2301.13186v1 | https://arxiv.org/pdf/2301.13186v1.pdf | Accurate Gaze Estimation using an Active-gaze Morphable Model | Rather than regressing gaze direction directly from images, we show that adding a 3D shape model can: i) improve gaze estimation accuracy, ii) perform well with lower resolution inputs and iii) provide a richer understanding of the eye-region and its constituent gaze system. Specifically, we use an `eyes and nose' 3D morphable model (3DMM) to capture the eye-region 3D facial geometry and appearance and we equip this with a geometric vergence model of gaze to give an `active-gaze 3DMM'. We show that our approach achieves state-of-the-art results on the Eyediap dataset and we present an ablation study. Our method can learn with only the ground truth gaze target point and the camera parameters, without access to the ground truth gaze origin points, thus widening the applicability of our approach compared to other methods. | ['Nick Pears', 'Hao Sun'] | 2023-01-30 | null | null | null | null | ['gaze-estimation'] | ['computer-vision'] | [ 1.92232635e-02 3.77199918e-01 -1.29533574e-01 -3.64480436e-01
-2.57575989e-01 -6.38363481e-01 6.08971834e-01 -6.07870817e-01
-2.50238270e-01 2.19468355e-01 6.38261735e-02 -1.93114325e-01
1.42131783e-02 -1.58896089e-01 -7.91774273e-01 -7.18215168e-01
1.51059434e-01 3.14798743e-01 1.70239672e-01 -1.91267133e-01
4.71828878e-01 5.23405969e-01 -2.01126838e+00 -1.24747783e-01
7.38710880e-01 1.14902830e+00 -2.65529484e-01 7.75031805e-01
2.97897130e-01 4.99460280e-01 -2.23358452e-01 -3.07194084e-01
4.00195837e-01 -4.22083765e-01 -7.49169350e-01 2.37123847e-01
1.22929168e+00 -3.99619669e-01 1.86746061e-01 7.28322804e-01
1.53847307e-01 4.67598997e-02 8.44008744e-01 -1.33383548e+00
-6.65825069e-01 -3.04030806e-01 -1.04797602e+00 1.14621639e-01
5.63310564e-01 2.56972402e-01 8.89492691e-01 -7.91605115e-01
8.30449462e-01 1.24257624e+00 6.33018136e-01 7.81101525e-01
-1.39504588e+00 -7.02856183e-01 4.06413406e-01 2.19893698e-02
-1.44217110e+00 -8.19198549e-01 7.01136231e-01 -4.62201416e-01
7.90938199e-01 1.90011665e-01 7.83172250e-01 9.62906480e-01
4.44439612e-02 5.41649699e-01 1.42699623e+00 -5.65463006e-01
-1.88218683e-01 -7.44843483e-02 8.87345001e-02 9.53027844e-01
-7.57020414e-02 2.88917840e-01 -8.08277309e-01 3.30430567e-01
1.03463042e+00 -2.44474888e-01 -3.59422356e-01 -8.38583648e-01
-8.30349565e-01 6.05547786e-01 7.97757983e-01 -2.54692584e-01
-1.37412980e-01 1.62747324e-01 -5.31131864e-01 2.51898617e-01
7.05189645e-01 3.23572218e-01 -3.75561595e-01 -1.31252080e-01
-8.81896615e-01 4.88627434e-01 4.09065545e-01 9.76054013e-01
1.13930392e+00 -2.89027631e-01 3.38792384e-01 3.58398229e-01
7.97487378e-01 8.23020816e-01 -6.65366873e-02 -1.31629109e+00
2.03921616e-01 7.86567450e-01 2.46919811e-01 -7.99627662e-01
-6.08755708e-01 1.41918346e-01 -6.50607944e-02 7.86519170e-01
8.79346371e-01 -1.99458078e-01 -1.17854714e+00 1.80541384e+00
4.02072251e-01 8.89765918e-02 -4.46666002e-01 1.10168588e+00
7.20280588e-01 1.49177954e-01 -2.21170694e-01 -3.71633470e-02
1.24575734e+00 -7.05105543e-01 -4.27171260e-01 -3.96330088e-01
5.84572852e-01 -5.57476938e-01 1.16873991e+00 3.60463649e-01
-1.22983217e+00 -2.74903089e-01 -8.15404952e-01 -4.91746575e-01
-2.92909861e-01 -1.37779275e-02 7.18626797e-01 7.90664852e-01
-1.55648005e+00 1.76364899e-01 -1.05569839e+00 -4.34352160e-01
6.22493863e-01 7.73705781e-01 -5.53817213e-01 2.02566132e-01
-5.40942848e-01 1.22493291e+00 -3.56819570e-01 4.47115116e-02
-5.15344143e-01 -8.32824707e-01 -1.02634072e+00 -2.01991901e-01
2.98950821e-01 -8.70157599e-01 1.18713391e+00 -1.23165059e+00
-1.66870284e+00 1.27296197e+00 -7.30223179e-01 1.34626776e-01
1.87665120e-01 -1.32230759e-01 9.88620613e-03 2.32113376e-01
-5.17698407e-01 1.17153907e+00 1.12867200e+00 -1.29804683e+00
-5.03655553e-01 -7.55429745e-01 2.53000110e-01 4.29886162e-01
1.13494493e-01 2.11556166e-01 -4.79778111e-01 -3.61527270e-03
8.89597982e-02 -1.21865106e+00 2.98511684e-01 4.97175515e-01
-2.16460288e-01 9.36216116e-03 7.77988374e-01 -4.65947926e-01
1.06629455e+00 -2.13655877e+00 4.56883520e-01 1.74372032e-01
6.51186049e-01 -9.11556277e-03 -6.05349094e-02 -1.66978925e-01
-2.83981353e-01 1.46037370e-01 -1.13208294e-02 -8.51206005e-01
-2.52075177e-02 -5.47219403e-02 4.07799520e-02 7.36106992e-01
4.12044644e-01 1.07605386e+00 -5.47012627e-01 -3.73655915e-01
2.42904678e-01 7.49541223e-01 -8.73832047e-01 -1.37084365e-01
-2.00174272e-01 6.61950886e-01 -7.93697312e-02 6.25659704e-01
6.99143291e-01 -3.29775751e-01 -3.19967449e-01 1.29395753e-01
-1.92202136e-01 1.60473332e-01 -7.80363619e-01 1.77203405e+00
-3.92816126e-01 1.16670966e+00 2.07631946e-01 -5.33456765e-02
6.30456448e-01 -3.99874039e-02 1.97651148e-01 -8.05400908e-01
4.85612780e-01 -1.25203550e-01 1.73461333e-01 -4.40851957e-01
3.41448307e-01 1.20771928e-02 2.93471992e-01 5.02101898e-01
2.33340129e-01 -7.40371346e-02 -4.76667434e-01 -1.45416200e-01
4.74627614e-01 5.66793323e-01 7.35447183e-02 -3.50741595e-01
3.30542833e-01 -3.01804215e-01 7.57531682e-03 8.95281732e-02
-1.43147886e-01 8.71740818e-01 7.44752526e-01 -1.87606454e-01
-8.36685777e-01 -1.08833563e+00 -3.06580275e-01 1.06625533e+00
2.33374551e-01 -1.23672813e-01 -1.10974395e+00 -6.61880732e-01
-2.84870192e-02 5.04269898e-01 -1.17789733e+00 5.53583913e-02
-4.17893648e-01 -3.62701297e-01 1.34878591e-01 3.26853484e-01
-1.35491812e-03 -6.07627630e-01 -7.70933867e-01 -7.43482649e-01
5.64337373e-02 -9.49298978e-01 -6.76835358e-01 -2.41810456e-01
-6.33780479e-01 -1.35773623e+00 -5.98046601e-01 -4.89766508e-01
9.53721344e-01 1.59711629e-01 1.04555905e+00 6.33235574e-02
-1.18892007e-01 3.88661146e-01 -1.97619632e-01 -7.40714133e-01
5.28160594e-02 1.47376120e-01 2.05880731e-01 1.26759797e-01
6.20398939e-01 -4.92204219e-01 -6.97070301e-01 3.93382460e-01
-4.33863133e-01 1.08769380e-01 2.36684576e-01 2.95300245e-01
1.97389364e-01 -6.11160815e-01 -1.64467841e-02 -8.66248190e-01
1.39468312e-01 -8.72202367e-02 -1.00781262e+00 -2.28670463e-02
-7.31159747e-01 1.19325712e-01 -9.20349211e-02 -4.36983854e-01
-9.34456408e-01 -4.28924384e-03 1.09026968e-01 -6.69573843e-01
-3.38995904e-01 -1.73847213e-01 -1.08185217e-01 -5.92986465e-01
9.11920369e-01 -3.50385666e-01 4.98537034e-01 -5.02472281e-01
5.40997088e-01 5.79258680e-01 4.52545762e-01 -2.90749222e-01
7.59586811e-01 7.45416284e-01 4.73357469e-01 -6.63867652e-01
-1.00365126e+00 -2.99686790e-01 -1.21387064e+00 -2.84377962e-01
8.44316185e-01 -8.49218428e-01 -1.34268415e+00 5.32778203e-01
-9.33853090e-01 -5.17392457e-01 1.06909893e-01 4.60166663e-01
-7.43224084e-01 -1.19097335e-02 -2.03321695e-01 -8.23030770e-01
-1.47835184e-02 -1.23371434e+00 1.48682761e+00 3.80837888e-01
-2.76957035e-01 -1.12460387e+00 6.41866401e-02 1.87053323e-01
1.74306363e-01 3.69419217e-01 6.93612874e-01 -8.68444238e-03
-7.91782975e-01 9.25755501e-03 -2.47357324e-01 4.88305502e-02
1.00719988e-01 3.41320723e-01 -1.45048594e+00 -1.52250752e-01
-1.87325507e-01 -6.58226535e-02 6.65397704e-01 6.63566232e-01
6.61299944e-01 -1.71623319e-01 -4.09460157e-01 1.22556424e+00
1.18954301e+00 -4.55619425e-01 6.15405619e-01 2.48706773e-01
9.45914567e-01 9.35419321e-01 4.11700070e-01 3.94794792e-02
7.28335083e-01 8.38066161e-01 7.47020543e-01 -2.11757898e-01
-2.28214771e-01 -2.07917705e-01 2.87262470e-01 1.93690166e-01
-4.82489049e-01 -2.42311750e-02 -1.05683160e+00 3.22991908e-01
-1.48547614e+00 -5.92902243e-01 -2.27515146e-01 2.23886013e+00
6.96416080e-01 -1.51404485e-01 4.94169533e-01 -1.75863028e-01
3.19583625e-01 -7.00283274e-02 -6.68372989e-01 -3.72194350e-01
1.17778361e-01 3.79787475e-01 5.69547117e-01 8.69234562e-01
-9.01086748e-01 1.08431292e+00 7.19943142e+00 -4.16308129e-03
-1.40216458e+00 -1.28469035e-01 2.20232919e-01 -5.19357443e-01
-2.37584293e-01 3.79736982e-02 -1.04305935e+00 3.77028555e-01
9.55543399e-01 3.55113357e-01 8.17594588e-01 3.26514184e-01
8.81653130e-02 -3.05422932e-01 -1.34068680e+00 1.14650023e+00
4.89203095e-01 -9.25279498e-01 -2.05834076e-01 7.20495880e-01
5.04372478e-01 7.52092600e-02 3.99448127e-01 -2.16552332e-01
-1.15494225e-02 -1.26686180e+00 7.64697373e-01 9.53135073e-01
1.31318474e+00 -6.01916969e-01 1.63090184e-01 3.39944452e-01
-6.19991362e-01 3.58248465e-02 1.17458299e-01 -1.95475876e-01
-1.02185704e-01 -3.81162405e-01 -6.99943900e-01 2.10184127e-01
7.78680563e-01 7.98945248e-01 -1.11826205e+00 9.45012450e-01
-2.23975882e-01 2.23149478e-01 -6.11892641e-01 1.47297218e-01
-1.59353435e-01 -2.27449924e-01 6.00940526e-01 4.39577997e-01
5.38330302e-02 -3.72084230e-02 -7.13520408e-01 1.05792356e+00
2.68434808e-02 -2.82631159e-01 -4.26097095e-01 3.00558507e-01
2.32565656e-01 1.02844465e+00 -4.56996888e-01 4.14580375e-01
-3.90690565e-01 6.15220368e-01 3.54155958e-01 6.01019382e-01
-5.07693410e-01 -1.07153438e-01 1.22778559e+00 5.18216252e-01
4.18639541e-01 -2.27255538e-01 -3.66394818e-01 -1.27876019e+00
-1.05992898e-01 -7.28208423e-01 -1.74937770e-01 -1.36502421e+00
-8.11669886e-01 4.87663537e-01 1.38675794e-01 -1.08356893e+00
-6.52268410e-01 -1.07199073e+00 -3.09081763e-01 1.25961614e+00
-1.70992482e+00 -1.46490312e+00 -4.61025357e-01 6.96990728e-01
-8.91978815e-02 2.09889606e-01 8.95123720e-01 -1.42044395e-01
-4.73433256e-01 8.18781018e-01 -2.81255871e-01 2.41516978e-02
9.31782305e-01 -1.34579349e+00 5.00956953e-01 6.50408208e-01
1.34187922e-01 8.31378937e-01 6.39750540e-01 -6.74332231e-02
-1.43478024e+00 -4.51777637e-01 7.01024234e-01 -1.60646594e+00
4.26061839e-01 -7.90442586e-01 -6.50660932e-01 1.08153200e+00
1.98903516e-01 9.63456482e-02 5.32872856e-01 5.26909649e-01
-4.79149252e-01 -1.09081216e-01 -1.00628054e+00 7.10141778e-01
1.13634026e+00 -6.36142194e-01 -5.25047362e-01 -1.67172357e-01
4.40030754e-01 -8.22038114e-01 -6.15135849e-01 1.72556400e-01
7.85296023e-01 -1.09564757e+00 9.48445499e-01 -5.58604896e-01
2.17562452e-01 -2.45558932e-01 2.29524195e-01 -1.15117383e+00
-9.76753682e-02 -7.60729253e-01 -4.84536827e-01 9.47297931e-01
3.52288455e-01 -6.98046029e-01 7.86749542e-01 7.83025026e-01
1.72594428e-01 -8.18714142e-01 -6.73296511e-01 -1.33749932e-01
1.88617006e-01 -3.93734306e-01 7.92654634e-01 6.57085121e-01
-1.19818546e-01 3.13166440e-01 -1.08955003e-01 2.89077044e-01
5.44294119e-01 -4.68732193e-02 1.03049529e+00 -1.67850399e+00
1.27672777e-01 -6.48634732e-01 -5.83545864e-01 -1.29134214e+00
3.60421717e-01 -4.70896840e-01 -1.62739709e-01 -9.64239776e-01
-5.23178875e-02 -3.41709763e-01 4.48440947e-02 5.76185286e-01
-1.00035608e-01 6.94730580e-01 1.44097626e-01 1.79623991e-01
-3.48756492e-01 1.30607799e-01 1.43150628e+00 4.77752596e-01
-1.70933992e-01 -1.55860176e-02 -8.73053074e-01 7.07487285e-01
2.00068682e-01 -2.19096005e-01 -4.31764454e-01 -6.75582945e-01
5.60838044e-01 -1.76326141e-01 6.70552433e-01 -4.34284538e-01
2.34038875e-01 -1.52603090e-01 6.63334370e-01 -3.52495492e-01
6.38043940e-01 -7.09159195e-01 -3.11250597e-01 -3.02247018e-01
-1.23487823e-01 -1.09210886e-01 3.85794193e-01 3.29417348e-01
1.95975021e-01 3.29070948e-02 8.58423054e-01 2.55090088e-01
-4.04954016e-01 3.81012052e-01 3.77501175e-02 -2.73966286e-02
7.83067763e-01 -7.13253617e-01 -3.88729364e-01 -3.53209376e-01
-7.23350406e-01 8.53380039e-02 1.54989707e+00 5.79277933e-01
1.27225012e-01 -9.87480640e-01 -2.60714322e-01 7.89370060e-01
2.74397552e-01 2.11220190e-01 -8.95780772e-02 1.06157291e+00
-5.77865541e-01 4.77459222e-01 -1.38055667e-01 -1.01254272e+00
-1.27900720e+00 6.20939374e-01 7.27172911e-01 6.54554844e-01
-4.86578226e-01 9.57354128e-01 6.18750989e-01 -2.80317336e-01
8.22085962e-02 -5.09093583e-01 -4.28571016e-01 1.06627233e-01
5.17640054e-01 1.54625595e-01 -1.32952174e-02 -1.27046776e+00
-4.74487394e-01 1.21829820e+00 1.71557635e-01 -1.28637433e-01
1.11484754e+00 -6.19814575e-01 -7.74246380e-02 4.89408731e-01
1.12705386e+00 2.66312748e-01 -1.80873096e+00 1.82908513e-02
-2.90627986e-01 -6.95380867e-01 3.13535899e-01 -7.55391121e-01
-9.53951359e-01 1.14258611e+00 6.14949346e-01 -8.50132853e-02
1.23773229e+00 1.87166825e-01 2.56937921e-01 -6.81237727e-02
1.83215022e-01 -5.13442993e-01 -3.49502444e-01 3.82788301e-01
5.63260198e-01 -1.36570036e+00 -6.18172176e-02 -3.98372978e-01
-5.19176185e-01 8.90106380e-01 6.33803010e-01 -1.36922851e-01
9.13172364e-01 -5.42747825e-02 4.07100558e-01 -4.27778244e-01
-6.85741603e-01 -5.20948112e-01 1.05923033e+00 8.39922249e-01
3.90122741e-01 -3.21999848e-01 4.74206686e-01 1.52273387e-01
-6.58081591e-01 -7.00505301e-02 3.59067321e-01 4.88213658e-01
-7.66405687e-02 -9.04013038e-01 -2.48753592e-01 1.94846675e-01
-2.94121474e-01 -5.56495823e-02 -6.03399813e-01 1.04357111e+00
1.27802134e-01 7.23034501e-01 6.80631638e-01 -1.91649884e-01
3.63520533e-01 1.82217985e-01 1.19788361e+00 -7.77137458e-01
-1.62629306e-01 9.08978730e-02 -3.38969648e-01 -7.58767724e-01
-5.60436845e-01 -9.07955408e-01 -9.02541697e-01 -5.66163361e-01
-5.41922987e-01 -5.33659935e-01 7.66589642e-01 9.54350829e-01
6.03436112e-01 7.10137188e-02 4.52530891e-01 -1.24844432e+00
1.33906707e-01 -1.02971625e+00 -5.64454079e-01 2.12329254e-01
1.09681404e+00 -1.06009781e+00 -5.66794872e-01 2.52618879e-01] | [14.115095138549805, 0.06408105790615082] |
43aef080-a26f-43c1-9a24-07d0111b3208 | unsupervised-labeled-parsing-with-deep-inside | null | null | https://aclanthology.org/D19-1161 | https://aclanthology.org/D19-1161.pdf | Unsupervised Labeled Parsing with Deep Inside-Outside Recursive Autoencoders | Understanding text often requires identifying meaningful constituent spans such as noun phrases and verb phrases. In this work, we show that we can effectively recover these types of labels using the learned phrase vectors from deep inside-outside recursive autoencoders (DIORA). Specifically, we cluster span representations to induce span labels. Additionally, we improve the model{'}s labeling accuracy by integrating latent code learning into the training procedure. We evaluate this approach empirically through unsupervised labeled constituency parsing. Our method outperforms ELMo and BERT on two versions of the Wall Street Journal (WSJ) dataset and is competitive to prior work that requires additional human annotations, improving over a previous state-of-the-art system that depends on ground-truth part-of-speech tags by 5 absolute F1 points (19{\%} relative error reduction). | ['Yi-Pei Chen', 'Patrick Verga', 'Andrew McCallum', 'Andrew Drozdov', 'Mohit Iyyer'] | 2019-11-01 | null | null | null | ijcnlp-2019-11 | ['constituency-parsing'] | ['natural-language-processing'] | [ 2.78301954e-01 5.17447412e-01 -3.67604017e-01 -4.73767012e-01
-1.44896495e+00 -9.23158348e-01 2.19832748e-01 4.03872669e-01
-4.16260302e-01 7.83216238e-01 6.79083884e-01 -3.23281139e-01
2.93117940e-01 -6.66005552e-01 -9.71650779e-01 -1.00891672e-01
-1.18207345e-02 5.91615021e-01 -6.42210543e-02 -3.37665565e-02
1.97496619e-02 8.50018188e-02 -1.13702142e+00 5.64686596e-01
5.26764691e-01 8.04450512e-01 -4.79954779e-02 7.67624021e-01
-3.91939998e-01 9.26085234e-01 -5.59483707e-01 -4.88452405e-01
-1.12816682e-02 -2.07546949e-01 -1.20895553e+00 6.29831851e-02
5.10270774e-01 -3.15146536e-01 -1.74351156e-01 6.58900380e-01
2.74788588e-01 1.60280108e-01 5.86777091e-01 -5.86647332e-01
-7.19707966e-01 1.30619109e+00 -5.25836825e-01 1.42425865e-01
2.71625876e-01 -2.73521930e-01 1.82010674e+00 -9.84083831e-01
6.37208581e-01 9.91560459e-01 1.03577435e+00 4.82203484e-01
-1.68301392e+00 -5.36523283e-01 2.19508901e-01 -1.85615882e-01
-1.27892876e+00 -4.52253550e-01 7.28938401e-01 -5.57758749e-01
1.46905208e+00 -5.78224175e-02 2.28375286e-01 1.41385019e+00
-2.17409566e-01 9.41649497e-01 8.20410132e-01 -9.11347806e-01
1.31426156e-01 -2.92155057e-01 7.46301889e-01 8.24087322e-01
2.13157564e-01 -1.05005302e-01 -4.08987612e-01 -1.68161973e-01
4.49451387e-01 -2.84450293e-01 -5.85959032e-02 3.16670425e-02
-1.03311133e+00 1.15523434e+00 9.13706273e-02 3.31789881e-01
-2.77128011e-01 4.74940211e-01 5.93682945e-01 -4.17038202e-02
6.02332056e-01 3.65601122e-01 -8.91354740e-01 -3.69284570e-01
-9.69695330e-01 1.96065065e-02 1.06709814e+00 1.20131516e+00
7.37370908e-01 6.85555264e-02 -3.38739336e-01 9.83052313e-01
7.92708099e-02 8.23880285e-02 2.93565243e-01 -1.09206295e+00
6.65376782e-01 5.06063700e-01 4.37506922e-02 -3.79001111e-01
-5.15352368e-01 -4.70751941e-01 -4.10257466e-02 -4.07994717e-01
5.79742134e-01 -3.58216017e-01 -1.01895928e+00 2.02539611e+00
9.64547694e-03 -7.16290846e-02 9.92985666e-02 3.53322834e-01
7.16795266e-01 6.69541776e-01 4.63630497e-01 -1.13198064e-01
1.62989831e+00 -1.03738868e+00 -7.30344176e-01 -4.15752351e-01
7.84720838e-01 -7.29694963e-01 1.30994534e+00 1.40988767e-01
-1.00699270e+00 -5.89294314e-01 -9.19561923e-01 -4.45196182e-01
-1.08313709e-01 4.31549430e-01 6.35409474e-01 5.80524743e-01
-8.74020576e-01 5.43682098e-01 -9.75490630e-01 -1.05015427e-01
3.04881930e-01 3.21547955e-01 -2.60026425e-01 6.26268685e-02
-1.17266500e+00 6.03189468e-01 6.67553902e-01 -5.07757187e-01
-9.39617932e-01 -8.82881045e-01 -1.21174526e+00 3.40544283e-01
2.88502991e-01 -4.49214429e-01 1.51911044e+00 -5.41056633e-01
-1.30109382e+00 1.00942695e+00 -2.20356524e-01 -5.19426107e-01
-1.25158876e-01 -7.10251510e-01 -3.71147394e-01 3.13883513e-01
2.79504836e-01 8.50679815e-01 4.60476369e-01 -1.12398076e+00
-3.53955358e-01 -3.85211259e-02 2.92293221e-01 -2.20955595e-01
-4.16007906e-01 2.14395270e-01 -3.10957164e-01 -7.84585774e-01
-1.44364506e-01 -1.04132128e+00 5.46164438e-03 -4.02769119e-01
-5.70612192e-01 -5.56168318e-01 4.36777592e-01 -1.15524423e+00
1.44699836e+00 -2.07481217e+00 8.32806155e-03 -2.27735877e-01
1.52078882e-01 -1.63257107e-01 -1.39152929e-01 5.09644091e-01
-2.69242644e-01 2.47369274e-01 -4.62899953e-01 -7.82362878e-01
2.76579499e-01 4.21797365e-01 -3.79572064e-01 2.20103905e-01
3.78145009e-01 8.55955124e-01 -6.85457349e-01 -5.87623954e-01
-1.86728910e-01 4.05358672e-01 -7.10185766e-01 1.01560168e-01
-4.91720617e-01 2.31483638e-01 -6.75423220e-02 5.61642647e-01
1.94403052e-01 -3.99123251e-01 2.43597195e-01 5.85434586e-02
3.97001859e-03 9.47050750e-01 -7.78446496e-01 2.09911489e+00
-7.30141163e-01 7.06088603e-01 -1.23084046e-01 -9.81181085e-01
7.44279683e-01 5.30530334e-01 4.45573211e-01 6.91617057e-02
3.06035608e-01 2.20172890e-02 -1.26481399e-01 -3.38270962e-01
4.39737022e-01 -1.83310822e-01 -6.47251427e-01 4.61526811e-01
5.02472043e-01 1.60390913e-01 4.24396664e-01 1.22066893e-01
1.36421978e+00 4.11111444e-01 3.17918479e-01 -3.42732817e-01
1.92151278e-01 3.83786410e-02 7.48071671e-01 5.53782940e-01
-4.33887579e-02 6.46526992e-01 5.85924208e-01 -4.57604975e-01
-1.38380921e+00 -1.06350493e+00 -4.32903469e-01 1.54748249e+00
-4.78974551e-01 -6.83994055e-01 -9.72105026e-01 -9.61095273e-01
-1.56680048e-01 1.19656718e+00 -6.48679376e-01 2.91710138e-01
-8.73088241e-01 -4.02589232e-01 1.01675522e+00 1.17646766e+00
4.55872491e-02 -1.23196590e+00 -4.94886696e-01 5.30088603e-01
-4.05643970e-01 -1.52419794e+00 -4.58052427e-01 5.89182496e-01
-6.58454716e-01 -8.83582771e-01 -3.66490066e-01 -9.13677096e-01
3.31632495e-01 -3.74834925e-01 1.63935065e+00 -2.21328616e-01
-2.96667755e-01 2.06857204e-01 -6.36482775e-01 -2.97661334e-01
-4.79202896e-01 5.06708503e-01 -1.57901406e-01 -6.12269163e-01
7.72439778e-01 -5.82987547e-01 -2.78905898e-01 -1.92383483e-01
-7.06781983e-01 3.47072515e-03 4.51161385e-01 1.02732038e+00
6.43192947e-01 -2.04491466e-01 4.48191822e-01 -1.25300217e+00
4.79630589e-01 -3.90306443e-01 -3.85788411e-01 1.22697115e-01
-3.98609877e-01 3.70159626e-01 6.78218782e-01 -2.08602235e-01
-1.03636837e+00 1.63052589e-01 -4.24543768e-01 -3.40185374e-01
-4.90718424e-01 3.85293663e-01 -1.33677140e-01 6.26842797e-01
7.31566906e-01 -1.23400226e-01 -5.06995142e-01 -5.33490956e-01
6.91059887e-01 6.13210797e-01 4.34921026e-01 -1.02840829e+00
5.03294647e-01 1.73412576e-01 -3.60734433e-01 -4.99769062e-01
-1.39620340e+00 -4.71282452e-01 -7.49333799e-01 4.17953402e-01
1.13995683e+00 -1.04158986e+00 -3.61827940e-01 -2.68859416e-01
-1.43250322e+00 -3.32839936e-01 -4.83977377e-01 5.58351696e-01
-5.68441033e-01 3.09167087e-01 -1.25717950e+00 -5.20032167e-01
-6.06988132e-01 -8.21343482e-01 1.39775014e+00 -7.22851828e-02
-7.38457024e-01 -9.55533743e-01 2.14824587e-01 4.03304428e-01
7.22180819e-03 2.80080646e-01 1.10198331e+00 -9.74842548e-01
-1.30842328e-01 -2.28975996e-01 -1.07709415e-01 4.23675179e-01
-7.99067244e-02 -2.37552688e-01 -1.22156501e+00 -9.05888826e-02
-2.15404540e-01 -6.08053088e-01 1.03196895e+00 3.51600081e-01
1.13267958e+00 -5.40189385e-01 -2.35922039e-01 3.26586932e-01
1.43261743e+00 -6.56232834e-02 3.03792626e-01 1.31462082e-01
7.92774141e-01 6.62722826e-01 1.56072885e-01 3.92404884e-01
3.39432001e-01 2.61538327e-01 -1.40831443e-02 1.95547447e-01
-2.77023733e-01 -4.60713416e-01 3.89112413e-01 1.05452561e+00
2.40785673e-01 -3.23259771e-01 -9.00621831e-01 9.30859745e-01
-1.62391174e+00 -8.08531523e-01 -1.61479060e-02 1.60645556e+00
1.15431273e+00 3.67482454e-01 2.74073742e-02 1.61454841e-01
7.48440266e-01 2.83601195e-01 -3.13111842e-01 -6.65706515e-01
1.78920656e-01 6.56148016e-01 3.81008208e-01 5.75536907e-01
-1.39366853e+00 1.19426227e+00 6.90592527e+00 7.70793438e-01
-5.38341761e-01 3.38744670e-01 6.22818708e-01 3.43355052e-02
-6.04458988e-01 2.39905864e-01 -1.14902043e+00 2.08222061e-01
1.35128868e+00 4.03722078e-01 4.15561080e-01 1.19038999e+00
-3.88839275e-01 3.32640380e-01 -1.39114618e+00 6.16406560e-01
1.61334455e-01 -1.18964076e+00 -4.88879293e-01 -5.43207563e-02
8.30786467e-01 5.07948920e-02 -1.81460842e-01 7.85258412e-01
9.01181817e-01 -1.01964974e+00 6.46402061e-01 1.84571102e-01
1.03170812e+00 -7.88994670e-01 5.08076966e-01 3.38109732e-01
-1.34334350e+00 -2.56619230e-02 -3.27804893e-01 -1.53011069e-01
2.83852756e-01 6.66151881e-01 -9.35930133e-01 1.51698455e-01
5.10360837e-01 3.51465791e-01 -2.99155325e-01 3.60252559e-01
-7.73201704e-01 1.33913577e+00 -4.36917782e-01 -1.27970139e-02
2.14036196e-01 2.59327561e-01 3.71581525e-01 1.59421015e+00
2.39769910e-02 8.78720358e-02 4.79179919e-01 1.15384769e+00
-5.38365006e-01 9.21610296e-02 -4.07512814e-01 -3.43485296e-01
6.20311022e-01 1.19185150e+00 -7.42584348e-01 -4.06283289e-01
-5.63661933e-01 8.40081453e-01 7.00733125e-01 2.48806655e-01
-7.80247688e-01 -4.57362860e-01 3.76840383e-01 -1.59613222e-01
6.48502767e-01 -2.86834508e-01 -5.22117436e-01 -1.14653182e+00
-7.30822086e-02 -5.92535675e-01 5.38957715e-01 -4.45464194e-01
-1.40306234e+00 6.30785346e-01 9.76492912e-02 -7.99211919e-01
-5.01943588e-01 -6.81712389e-01 -3.88666838e-01 7.01993644e-01
-1.19614458e+00 -1.38776028e+00 1.61929652e-01 -1.97202899e-02
8.85954559e-01 1.51138031e-03 1.32326603e+00 8.20026845e-02
-5.52314103e-01 7.86524892e-01 -3.39035168e-02 7.42969811e-01
5.01133263e-01 -1.49702477e+00 7.06056416e-01 9.84049976e-01
6.54836893e-01 7.66525805e-01 6.75145626e-01 -5.35244823e-01
-8.65389943e-01 -9.56384838e-01 1.19442570e+00 -7.65572071e-01
8.13090444e-01 -5.83409727e-01 -7.06925094e-01 1.24413729e+00
4.19613481e-01 4.87942174e-02 1.20999193e+00 7.04763412e-01
-8.48855615e-01 3.07121605e-01 -8.75217915e-01 2.41316497e-01
1.25114942e+00 -7.80353069e-01 -9.53783274e-01 2.77984053e-01
1.40706277e+00 -3.28498840e-01 -9.09344912e-01 4.48721945e-01
4.10710722e-01 -5.85497022e-01 7.49418616e-01 -6.45713985e-01
8.38694572e-01 1.30191013e-01 -5.22523522e-01 -9.95381594e-01
-3.40096503e-01 -3.95397604e-01 -4.31183577e-01 1.50226319e+00
5.75512588e-01 2.20317487e-02 8.03156674e-01 5.99424541e-01
-5.85285842e-01 -7.72491097e-01 -7.75636554e-01 -6.03832901e-01
4.26516682e-01 -6.94912314e-01 4.09634411e-01 8.57317030e-01
2.38702893e-01 7.53399730e-01 -1.56749427e-01 1.38067022e-01
5.06327927e-01 3.07918042e-01 3.32826048e-01 -1.14092147e+00
-7.88985193e-01 -2.45300606e-01 -1.43860370e-01 -1.23069870e+00
7.91533530e-01 -1.01633120e+00 9.99281034e-02 -1.48395765e+00
1.74885184e-01 -2.68292248e-01 -1.66931495e-01 8.48779142e-01
-1.90704793e-01 1.92242950e-01 4.32657897e-02 1.51391281e-02
-6.74513578e-01 3.46982092e-01 5.48110843e-01 -1.55807376e-01
6.17255829e-02 -3.20549428e-01 -8.14524055e-01 7.85824955e-01
8.70774746e-01 -7.23962605e-01 -8.83904323e-02 -6.43497944e-01
2.71071076e-01 1.67557195e-01 -8.43599513e-02 -8.63986313e-01
-1.89533576e-01 2.31483951e-01 4.65149462e-01 -6.81302130e-01
3.09954762e-01 -6.23432994e-01 -3.23206097e-01 2.65048116e-01
-7.99197316e-01 -1.15069219e-04 3.92269999e-01 5.06910622e-01
-2.25456636e-02 -6.42492115e-01 4.84355509e-01 -2.72660792e-01
-5.66132545e-01 -8.68357718e-02 -4.21409011e-01 4.39168423e-01
8.14559221e-01 2.50811666e-01 -3.55439633e-01 -2.12834954e-01
-8.29288423e-01 -3.77120823e-02 2.06942305e-01 9.04615968e-02
1.86382592e-01 -1.08007681e+00 -6.37749195e-01 -5.74328862e-02
1.88574180e-01 7.65059963e-02 -5.95171414e-02 2.64608324e-01
-5.63485742e-01 5.13146818e-01 3.26618522e-01 -3.09150517e-01
-9.07699585e-01 6.66640341e-01 -2.05919564e-01 -7.47540712e-01
-7.23107755e-01 1.12602091e+00 2.73581780e-02 -5.96445322e-01
2.70993769e-01 -4.57855642e-01 -1.97922096e-01 -6.29422069e-02
2.30507433e-01 2.60647386e-01 -6.14878051e-02 -6.39926076e-01
-2.13362575e-01 4.08281356e-01 -2.85966486e-01 -3.23479921e-01
1.29550195e+00 -4.17681225e-02 1.05564982e-01 6.70573533e-01
1.48457265e+00 3.65884900e-01 -1.33093345e+00 -3.31317365e-01
4.45352346e-01 4.06501703e-02 -1.51231185e-01 -7.43581593e-01
-5.12171626e-01 9.69340682e-01 2.44802594e-01 1.01231135e-01
6.98461473e-01 3.50941956e-01 1.18490267e+00 4.12066787e-01
1.82239875e-01 -1.11277497e+00 1.83933362e-01 6.51300669e-01
5.11396766e-01 -1.04860044e+00 -1.27902776e-01 -5.09943128e-01
-6.41508758e-01 1.13346195e+00 3.62481326e-01 -4.08760339e-01
5.88878036e-01 5.74469149e-01 -5.85974120e-02 -5.61519861e-02
-7.52311170e-01 -3.05182904e-01 2.47537836e-01 1.64570197e-01
9.17268634e-01 2.57427812e-01 -4.91002560e-01 1.03955579e+00
-5.63355565e-01 -2.82180190e-01 2.22088963e-01 1.01131892e+00
-5.81085920e-01 -1.27368653e+00 -1.32771999e-01 3.31583261e-01
-9.89479244e-01 -4.94190276e-01 -1.24680407e-01 7.34220147e-01
1.68803513e-01 1.00583625e+00 2.98359424e-01 -2.75193602e-01
7.62753561e-02 7.04221368e-01 3.82237256e-01 -1.11168385e+00
-4.61943060e-01 4.31646198e-01 5.30958056e-01 -2.11992696e-01
-3.71397883e-01 -6.19133413e-01 -1.64313316e+00 4.87610251e-02
-4.42749113e-01 1.68662861e-01 5.53901851e-01 1.02523100e+00
2.56387323e-01 8.23899925e-01 2.88306028e-01 -4.83403593e-01
-7.19794393e-01 -1.31271613e+00 -3.26368392e-01 5.40189564e-01
3.17381956e-02 -4.83299613e-01 -3.18135738e-01 4.55229938e-01] | [10.398979187011719, 9.573623657226562] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.