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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
cc176bf3-d762-43b6-9dc8-69c75ec72956 | sequential-attention-source-identification | 2306.15886 | null | https://arxiv.org/abs/2306.15886v1 | https://arxiv.org/pdf/2306.15886v1.pdf | Sequential Attention Source Identification Based on Feature Representation | Snapshot observation based source localization has been widely studied due to its accessibility and low cost. However, the interaction of users in existing methods does not be addressed in time-varying infection scenarios. So these methods have a decreased accuracy in heterogeneous interaction scenarios. To solve this critical issue, this paper proposes a sequence-to-sequence based localization framework called Temporal-sequence based Graph Attention Source Identification (TGASI) based on an inductive learning idea. More specifically, the encoder focuses on generating multiple features by estimating the influence probability between two users, and the decoder distinguishes the importance of prediction sources in different timestamps by a designed temporal attention mechanism. It's worth mentioning that the inductive learning idea ensures that TGASI can detect the sources in new scenarios without knowing other prior knowledge, which proves the scalability of TGASI. Comprehensive experiments with the SOTA methods demonstrate the higher detection performance and scalability in different scenarios of TGASI. | ['Xuelong Li', 'Chao GAO', 'Zhen Wang', 'Dongpeng Hou'] | 2023-06-28 | null | null | null | null | ['graph-attention'] | ['graphs'] | [ 2.31563136e-01 -2.56539881e-01 -2.78432190e-01 -1.84695777e-02
-4.91092801e-01 -1.51663259e-01 5.12571931e-01 1.99177653e-01
-1.37681857e-01 7.77836323e-01 5.42178929e-01 -6.33743256e-02
-2.00535178e-01 -6.57084405e-01 -3.43112767e-01 -8.79477978e-01
-6.54869676e-01 1.97433591e-01 5.21316528e-01 -7.31453747e-02
7.98330233e-02 1.80908397e-01 -9.16877985e-01 -8.28221813e-03
7.27713883e-01 6.86120927e-01 4.46550786e-01 7.52471924e-01
-3.93115059e-02 9.33154404e-01 -9.15771127e-01 -2.46732379e-03
-1.56766102e-01 -6.81970417e-01 -2.39986002e-01 -4.00185585e-01
-5.98956585e-01 -5.84465444e-01 -4.90717232e-01 5.47134399e-01
6.45891011e-01 -8.18107799e-02 3.58126611e-01 -1.82862353e+00
-6.86409593e-01 8.22829068e-01 -8.58485699e-01 7.26347208e-01
5.52352607e-01 -1.55210420e-01 8.85846317e-01 -3.94403785e-01
3.83079529e-01 1.05750906e+00 7.06047416e-01 2.12993249e-01
-5.43804049e-01 -6.38100326e-01 4.24507231e-01 6.14430904e-01
-1.35171270e+00 -1.81000829e-01 8.04117978e-01 -2.12056577e-01
8.20931315e-01 1.09719567e-01 3.42242062e-01 1.27005804e+00
6.51715919e-02 9.18603480e-01 4.12745029e-01 -6.16018996e-02
1.03372157e-01 9.81209278e-02 -1.19473405e-01 5.86493969e-01
4.06283766e-01 -3.59269604e-02 -5.50830781e-01 -2.78490931e-01
6.56302750e-01 4.03163821e-01 -3.81827682e-01 -1.66701362e-03
-1.62826610e+00 4.91060287e-01 5.96403122e-01 6.50546968e-01
-3.92535955e-01 3.87426466e-01 5.34306407e-01 3.54839079e-02
4.24285054e-01 -1.48984149e-01 -4.08013731e-01 -3.10695440e-01
-7.19115794e-01 -4.97241229e-01 7.32678771e-01 1.08167005e+00
4.14269477e-01 9.92733911e-02 -3.29056829e-01 1.75288200e-01
4.26338017e-01 7.62768805e-01 2.03635886e-01 -3.49939585e-01
2.71648109e-01 3.95165026e-01 1.35008693e-01 -1.41363132e+00
-4.12624925e-01 -8.07811260e-01 -1.11528599e+00 -5.30574858e-01
4.48743924e-02 -4.76374865e-01 -4.32031929e-01 2.02177501e+00
4.00276810e-01 9.48929310e-01 -1.93360418e-01 7.84234941e-01
4.92068857e-01 9.74387228e-01 5.04639186e-02 -6.07388675e-01
1.12728179e+00 -7.91551352e-01 -8.16168904e-01 -2.15865374e-02
3.08412313e-01 -3.66898030e-01 4.79963273e-01 9.57913231e-03
-4.16066378e-01 -4.81640577e-01 -8.79731357e-01 4.47543442e-01
-2.64675707e-01 2.00100780e-01 4.27020609e-01 4.53413010e-01
-9.68050599e-01 -6.56299107e-03 -7.80601442e-01 -7.67049372e-01
6.14276156e-02 1.88687012e-01 5.32417074e-02 2.10899815e-01
-1.56579888e+00 4.34119612e-01 3.54626358e-01 1.19304493e-01
-1.01429486e+00 -4.21959490e-01 -6.20546401e-01 2.89837807e-01
5.58669686e-01 -5.13067901e-01 9.09713566e-01 -8.47489417e-01
-1.18799591e+00 -7.03972951e-03 -5.02005339e-01 -4.73153323e-01
3.69363219e-01 1.89040452e-02 -7.76863456e-01 2.00986028e-01
4.20376331e-01 -8.91721174e-02 8.73904824e-01 -1.14181471e+00
-8.79180729e-01 -3.04004904e-02 1.53105706e-01 1.13031328e-01
-5.48692763e-01 7.34328246e-03 -7.30982661e-01 -6.05817318e-01
-2.82932460e-01 -4.81102496e-01 -1.12704068e-01 -4.16446269e-01
-4.42734957e-01 -2.20611840e-01 9.64671612e-01 -7.50112772e-01
1.51462221e+00 -1.98757267e+00 -6.97531849e-02 2.41209567e-01
3.09429199e-01 2.00046211e-01 -7.75504038e-02 9.22807455e-01
3.47298354e-01 -4.17832993e-02 -8.02193806e-02 -3.46310586e-01
-2.44176194e-01 -3.88008356e-02 -3.97939771e-01 2.85198033e-01
1.56511297e-03 6.41096890e-01 -1.43033135e+00 -7.65866339e-01
8.45891014e-02 6.50578856e-01 -2.49659583e-01 4.39614087e-01
-8.50711912e-02 7.74764955e-01 -6.31398737e-01 4.25710112e-01
5.53753495e-01 -7.85076618e-01 4.05745894e-01 -8.01557526e-02
-4.78756987e-02 9.07878280e-02 -9.75159168e-01 1.37621343e+00
-5.25635481e-01 5.19938946e-01 -9.75876078e-02 -8.47653568e-01
5.31296432e-01 5.75447679e-01 7.62923360e-01 -5.48406005e-01
3.61626633e-02 -6.54457584e-02 -9.19840932e-02 -5.43669522e-01
6.83797747e-02 3.46626937e-01 -1.72217563e-01 6.79495871e-01
-2.30678782e-01 9.78631616e-01 1.28654227e-01 5.75170994e-01
1.41269457e+00 -3.54593098e-02 5.22380590e-01 1.63586468e-01
6.30983829e-01 -2.57463694e-01 7.16093838e-01 8.92811239e-01
-3.42626005e-01 2.75000483e-01 3.69192898e-01 -4.49383669e-02
-5.90598464e-01 -1.10489082e+00 4.30213004e-01 1.03799629e+00
6.38980687e-01 -4.54162598e-01 -5.64505517e-01 -8.72310877e-01
-1.61809593e-01 5.03861189e-01 -6.96958780e-01 -1.59036085e-01
-5.33288300e-01 -7.22291052e-01 4.69201535e-01 5.30420005e-01
7.78627455e-01 -1.03027678e+00 -6.21251047e-01 4.16161954e-01
-5.63147068e-01 -1.07322705e+00 -6.35014176e-01 -3.18934739e-01
-2.24303469e-01 -1.08882618e+00 -7.62607813e-01 -6.90532923e-01
5.69187224e-01 7.47564554e-01 7.11867452e-01 2.30051950e-01
-4.61548083e-02 3.65101933e-01 -5.26744187e-01 -2.09565759e-01
-1.02598667e-01 2.92989701e-01 1.68103203e-01 4.22854602e-01
2.48279735e-01 -6.05111957e-01 -7.74756730e-01 3.20088089e-01
-6.69618785e-01 -6.05995692e-02 6.33038998e-01 4.97388750e-01
7.06965029e-02 2.76971161e-01 1.09290862e+00 -6.09870672e-01
6.52442038e-01 -1.24183357e+00 -2.84647495e-01 4.87666667e-01
-3.93620163e-01 -1.52740836e-01 7.61938512e-01 -2.80093044e-01
-1.10935390e+00 -3.51143599e-01 -7.63002783e-02 -6.85417727e-02
-1.72013760e-01 5.47871113e-01 -2.26908296e-01 2.39532992e-01
2.21087992e-01 5.50258040e-01 -3.56541216e-01 -2.39891216e-01
-1.90008301e-02 7.65827358e-01 1.44747004e-01 3.18539590e-02
8.72372687e-01 6.47495985e-01 -2.06429482e-01 -8.39157939e-01
-4.99855131e-01 -7.33127594e-01 -2.65626073e-01 -2.42315575e-01
6.81781292e-01 -8.75145435e-01 -1.04317653e+00 4.83643860e-01
-1.42198765e+00 -1.10808402e-01 3.52299064e-01 4.58076984e-01
-1.16033524e-01 6.71359241e-01 -6.40841424e-01 -1.13825655e+00
-4.21155632e-01 -8.98483098e-01 1.05107331e+00 2.93483555e-01
2.98125558e-02 -1.14294851e+00 1.50084898e-01 -1.69126555e-01
5.75735450e-01 2.11283222e-01 6.07963502e-01 -6.25922978e-01
-7.67965376e-01 -2.17460617e-01 -5.80269694e-01 -3.72296363e-01
6.26430571e-01 -2.21753284e-01 -7.78017104e-01 -4.27839696e-01
-7.48777464e-02 4.02618706e-01 5.04701018e-01 3.00215185e-01
7.59380698e-01 -5.79375207e-01 -9.01990116e-01 1.87964544e-01
1.45855260e+00 5.14028907e-01 3.39789689e-01 2.31613200e-02
6.92907751e-01 2.32119411e-01 6.02007508e-01 6.29648566e-01
6.09264433e-01 6.57909572e-01 5.64682901e-01 -3.08912378e-02
-4.33986671e-02 -4.20192599e-01 4.88064766e-01 6.62026227e-01
2.09584147e-01 -9.81821954e-01 -8.58757675e-01 7.55853176e-01
-1.93526912e+00 -1.21381199e+00 -1.00717880e-01 2.17197561e+00
5.13610780e-01 9.76847187e-02 2.84626752e-01 9.36038420e-02
1.01887763e+00 3.60447109e-01 -4.79860485e-01 1.59695357e-01
2.10543305e-01 -5.90638638e-01 3.98382932e-01 4.33765531e-01
-8.11687410e-01 5.16957879e-01 5.32691050e+00 8.41892660e-01
-1.05313945e+00 5.57272017e-01 3.83297950e-01 1.14262439e-01
-1.64482504e-01 -1.58490852e-01 -6.92026556e-01 1.13775790e+00
9.52411413e-01 -3.25716913e-01 1.96313530e-01 6.39118731e-01
4.67809439e-01 -4.97721657e-02 -6.52534008e-01 9.03713048e-01
2.18649864e-01 -1.19570518e+00 -1.17482983e-01 -1.95989802e-01
4.36736971e-01 1.24918520e-01 -2.40121216e-01 6.01201355e-02
1.66962937e-01 -4.30924803e-01 4.03466552e-01 4.90505517e-01
6.98678076e-01 -7.06336319e-01 9.15210783e-01 4.73606437e-01
-1.68107760e+00 -1.93502530e-01 -5.82993254e-02 1.21945836e-01
6.47788107e-01 8.30134034e-01 -1.25479412e+00 8.97248209e-01
6.43717527e-01 8.61920238e-01 -3.47116888e-01 1.27554071e+00
-2.52122104e-01 9.94409204e-01 -5.21167159e-01 -3.47038358e-01
6.58922642e-02 6.50156140e-02 7.86369920e-01 1.30219698e+00
5.08932114e-01 -1.09395169e-01 3.12004328e-01 3.97253364e-01
9.12572592e-02 -1.40697286e-01 -7.85874307e-01 2.86051817e-02
6.32186413e-01 1.07926273e+00 -8.73351038e-01 -3.48246008e-01
-3.80891174e-01 1.16483390e+00 -1.02569219e-02 4.26350206e-01
-1.27248061e+00 -6.39464915e-01 3.92543554e-01 2.34331548e-01
3.47735733e-01 -4.15624738e-01 -6.19723164e-02 -1.12513518e+00
-1.99599266e-01 -3.67419481e-01 5.44355035e-01 -7.86638677e-01
-1.09510303e+00 7.83167481e-01 -5.21644875e-02 -1.21360445e+00
-1.96577772e-01 1.42107338e-01 -9.36424434e-01 4.80336696e-01
-1.46210384e+00 -1.09743547e+00 -4.54937637e-01 6.43618107e-01
4.43298161e-01 5.68097681e-02 4.92735624e-01 5.07801890e-01
-7.80767202e-01 6.85746551e-01 2.06538871e-01 2.76277930e-01
6.21475220e-01 -9.74640727e-01 3.80523920e-01 1.21467292e+00
-5.28205112e-02 6.45089388e-01 6.61767066e-01 -1.01336300e+00
-1.29046786e+00 -1.22663951e+00 1.16323888e+00 -1.99370846e-01
6.51772618e-01 -4.07302886e-01 -7.17260122e-01 6.64433599e-01
4.28767920e-01 -3.49934995e-01 5.98359466e-01 -1.35156170e-01
-2.34851748e-01 -3.09732109e-01 -1.05987561e+00 6.08775914e-01
1.34835041e+00 -4.37470973e-01 -2.66026109e-01 5.21351874e-01
8.97850811e-01 3.59510742e-02 -4.24580544e-01 1.61365926e-01
2.13901639e-01 -7.95434177e-01 8.83627594e-01 -2.52499342e-01
-8.83343145e-02 -5.53782463e-01 3.04475307e-01 -1.12089360e+00
-4.95656729e-01 -6.53490663e-01 -4.30564016e-01 1.52336621e+00
1.24372356e-01 -9.26531076e-01 4.44640934e-01 -6.67643100e-02
2.45620340e-01 -3.77772748e-01 -8.96200836e-01 -7.82174647e-01
-9.02323782e-01 -2.08772823e-01 9.42836344e-01 7.27388620e-01
2.20162347e-01 4.42434758e-01 -9.19758916e-01 7.39741683e-01
5.84401548e-01 3.16574693e-01 5.73482454e-01 -9.94743109e-01
-3.66805077e-01 4.55284342e-02 -2.44076774e-01 -9.46186721e-01
-1.84751466e-01 -5.19670129e-01 1.99203074e-01 -1.79204297e+00
3.13810170e-01 -5.82919419e-01 -6.23215675e-01 4.02095824e-01
-3.53970081e-01 9.52385813e-02 -1.45102590e-01 4.41591352e-01
-1.19892240e+00 5.06970584e-01 5.64261794e-01 5.48279891e-03
-1.28163531e-01 1.37541577e-01 -4.37124282e-01 3.30055416e-01
1.06446624e+00 -5.29315770e-01 -8.42727065e-01 -5.66814661e-01
1.13065399e-01 3.30446064e-01 4.53541368e-01 -1.19898391e+00
4.40672010e-01 -1.83862001e-01 1.31489500e-01 -5.46181858e-01
4.74330336e-02 -1.02870488e+00 3.07730585e-01 8.19113553e-01
-3.30043137e-02 9.43108201e-02 -7.16010556e-02 1.29778397e+00
7.16476962e-02 2.34542325e-01 1.72190025e-01 1.36158824e-01
-9.17710364e-01 6.51468396e-01 -6.13347232e-01 -4.57692426e-03
1.21585727e+00 -1.95882376e-02 -2.20874086e-01 -7.38362372e-01
-2.18437403e-01 3.79400492e-01 2.73871273e-01 5.01467645e-01
6.59126103e-01 -1.27420056e+00 -5.42728662e-01 1.88361213e-01
1.63267359e-01 -4.67633694e-01 4.64805096e-01 1.12072408e+00
-1.57899484e-01 4.77229238e-01 5.86439222e-02 -4.57354248e-01
-1.08273673e+00 8.47850144e-01 1.55170187e-01 -4.45134282e-01
-3.13575178e-01 7.19127715e-01 1.13557734e-01 1.35548294e-01
2.75562048e-01 5.62566556e-02 -3.28221887e-01 4.80938442e-02
6.76998496e-01 5.11161983e-01 -2.33942181e-01 -5.52745819e-01
-7.15868056e-01 2.21656129e-01 1.72007233e-01 5.89551553e-02
1.07677662e+00 -5.80843568e-01 3.95907909e-02 3.57885689e-01
1.21955287e+00 2.18618676e-01 -1.13233387e+00 -4.88935232e-01
1.00011513e-01 -4.59156781e-01 -2.54424423e-01 -8.08068812e-01
-1.05552983e+00 8.37758780e-01 4.25711006e-01 5.56861877e-01
1.23310304e+00 -1.70976698e-01 1.27579701e+00 2.95588765e-02
8.30165029e-01 -5.65534532e-01 9.05391648e-02 2.81316817e-01
6.09973490e-01 -1.16001582e+00 -5.19844234e-01 -3.46226305e-01
-5.37029445e-01 7.44939566e-01 5.23302078e-01 2.34229714e-01
8.04071009e-01 2.93621451e-01 -2.34001115e-01 -2.48294622e-02
-6.42324507e-01 -4.17421520e-01 -1.56152815e-01 8.36783886e-01
1.14969723e-01 -9.91151407e-02 -2.13150561e-01 4.95021075e-01
5.54147303e-01 1.01443894e-01 2.68189520e-01 7.75425017e-01
-3.38140249e-01 -9.37033474e-01 -5.84206283e-02 2.82112420e-01
-2.83910662e-01 -1.58133328e-01 -7.55911618e-02 4.34038460e-01
2.17392176e-01 1.33702075e+00 -1.09543137e-01 -6.11414969e-01
-9.91088003e-02 -3.32938969e-01 -1.38015315e-01 -5.12075186e-01
-3.37303758e-01 -7.54208937e-02 -9.28684920e-02 -4.29641962e-01
-3.11870277e-01 -4.96170700e-01 -1.45031154e+00 -4.17843044e-01
-2.21257851e-01 7.04594254e-01 2.05142871e-01 8.60247552e-01
8.58059406e-01 6.39600933e-01 9.90452826e-01 -5.62884808e-01
-7.64661431e-02 -8.37381959e-01 -3.70380968e-01 2.63427496e-01
5.95462322e-01 -6.54238522e-01 -3.19089651e-01 -1.91841990e-01] | [6.624197483062744, 2.185577392578125] |
8bc4e90f-e1e5-46f6-afd3-f1e23fdf96c3 | diffustereo-high-quality-human-reconstruction | 2207.08000 | null | https://arxiv.org/abs/2207.08000v2 | https://arxiv.org/pdf/2207.08000v2.pdf | DiffuStereo: High Quality Human Reconstruction via Diffusion-based Stereo Using Sparse Cameras | We propose DiffuStereo, a novel system using only sparse cameras (8 in this work) for high-quality 3D human reconstruction. At its core is a novel diffusion-based stereo module, which introduces diffusion models, a type of powerful generative models, into the iterative stereo matching network. To this end, we design a new diffusion kernel and additional stereo constraints to facilitate stereo matching and depth estimation in the network. We further present a multi-level stereo network architecture to handle high-resolution (up to 4k) inputs without requiring unaffordable memory footprint. Given a set of sparse-view color images of a human, the proposed multi-level diffusion-based stereo network can produce highly accurate depth maps, which are then converted into a high-quality 3D human model through an efficient multi-view fusion strategy. Overall, our method enables automatic reconstruction of human models with quality on par to high-end dense-view camera rigs, and this is achieved using a much more light-weight hardware setup. Experiments show that our method outperforms state-of-the-art methods by a large margin both qualitatively and quantitatively. | ['Yebin Liu', 'Jingxiang Sun', 'Hongwen Zhang', 'Zerong Zheng', 'Ruizhi Shao'] | 2022-07-16 | null | null | null | null | ['3d-human-reconstruction', 'stereo-matching-1'] | ['computer-vision', 'computer-vision'] | [ 2.75793653e-02 -4.97929752e-02 4.79933232e-01 -2.00807020e-01
-5.28011084e-01 -4.33150113e-01 4.97283548e-01 -5.50351143e-01
-5.32027960e-01 3.05907816e-01 3.28432441e-01 1.84668303e-01
2.85937548e-01 -8.31862628e-01 -6.30706966e-01 -4.52566862e-01
4.46779877e-01 7.63732255e-01 6.41571581e-01 -2.30841696e-01
6.89366041e-03 7.10283816e-01 -1.66829216e+00 2.71116674e-01
4.94790912e-01 8.87202799e-01 3.27553898e-01 9.14869368e-01
4.24381196e-01 9.64658558e-01 9.75177623e-03 -4.63996083e-01
5.85510790e-01 -2.73899823e-01 -5.95951736e-01 3.72575462e-01
1.00665212e+00 -9.48600054e-01 -7.36735702e-01 7.64205754e-01
8.17629397e-01 -4.38218303e-02 3.93794984e-01 -7.64378250e-01
-1.46704614e-01 -1.49772584e-01 -5.84557474e-01 -1.25583991e-01
7.24246323e-01 4.20150757e-01 5.76104522e-01 -1.05769169e+00
9.53676999e-01 1.39689183e+00 6.48143888e-01 7.18109369e-01
-1.32852161e+00 -6.67840958e-01 -1.55795231e-01 4.33027931e-02
-1.31456685e+00 -7.63442934e-01 8.93038154e-01 -4.53471333e-01
9.62388217e-01 -1.81531832e-02 1.24807465e+00 1.13625348e+00
3.21464896e-01 6.01543844e-01 1.25869024e+00 -1.21099263e-01
2.14765832e-01 -1.22579761e-01 -4.36758488e-01 1.00210464e+00
2.74865385e-02 3.72791678e-01 -1.01999164e+00 -2.02034697e-01
1.63669264e+00 2.43861705e-01 -3.50664794e-01 -7.71462202e-01
-1.41001749e+00 6.26671970e-01 4.80003655e-01 -1.81838408e-01
-4.53363240e-01 3.01593274e-01 -4.33864305e-03 5.55226952e-02
5.15689194e-01 -5.71370907e-02 -1.67884361e-02 -5.67161851e-02
-1.04532766e+00 2.36718282e-01 6.48781717e-01 9.88673091e-01
9.04898226e-01 9.17388797e-02 3.09098274e-01 6.80112541e-01
3.66475344e-01 8.56723607e-01 1.48185724e-02 -1.73876536e+00
3.04063082e-01 5.27069569e-01 1.01598622e-02 -1.07186198e+00
-4.15342838e-01 -3.27068776e-01 -1.33601439e+00 5.80126762e-01
2.46298462e-01 4.22289334e-02 -7.83766448e-01 1.49668980e+00
6.23667538e-01 1.51826993e-01 -1.12245262e-01 1.22893775e+00
7.28664517e-01 4.36288953e-01 -7.43247151e-01 1.96118578e-01
1.18337071e+00 -9.80071306e-01 -2.38013923e-01 -5.21034300e-01
2.00613394e-01 -7.40362465e-01 8.12342167e-01 6.02458835e-01
-1.53221440e+00 -5.12709081e-01 -8.82306576e-01 -5.22103786e-01
2.30645940e-01 -2.86374450e-01 6.33319795e-01 2.73897141e-01
-1.56106806e+00 5.24000049e-01 -9.09753025e-01 -5.06114364e-01
1.51219487e-01 3.42356145e-01 -7.13568807e-01 -6.32604301e-01
-7.73881257e-01 7.06902266e-01 -7.20891729e-02 7.09442422e-02
-9.27478611e-01 -6.11758649e-01 -1.01173794e+00 -3.85324955e-01
2.60512084e-01 -1.65256512e+00 1.06563711e+00 -5.53576469e-01
-1.65445948e+00 1.21094036e+00 -1.78493083e-01 -2.36600250e-01
7.91505873e-01 -2.51884639e-01 1.22621231e-01 5.51101923e-01
-4.73547243e-02 1.03689599e+00 8.69065046e-01 -1.30422330e+00
-5.40459037e-01 -7.18520939e-01 1.55082494e-01 4.36293840e-01
-4.39653806e-02 -1.82427809e-01 -9.80032563e-01 -4.51180190e-01
4.04962331e-01 -1.04665494e+00 -5.29416203e-01 5.31173944e-01
-1.82593405e-01 4.60376263e-01 5.88498116e-01 -6.60286307e-01
6.43729091e-01 -1.93940580e+00 5.11369526e-01 1.74000099e-01
7.96950936e-01 -1.82511792e-01 9.89229307e-02 9.93889943e-02
2.51539052e-01 -7.87214279e-01 -1.71808675e-01 -8.68025780e-01
-3.52813929e-01 1.29828155e-01 1.34072438e-01 6.62958026e-01
-4.38604295e-01 8.63483131e-01 -7.79191971e-01 -4.66421157e-01
6.69762552e-01 8.64123285e-01 -9.33552802e-01 4.70152289e-01
2.08974659e-01 6.46712244e-01 -6.81675971e-02 6.08574390e-01
7.04103172e-01 -4.67761934e-01 2.84544174e-02 -3.73471022e-01
-2.95155067e-02 -9.60128754e-02 -1.40464258e+00 2.45805120e+00
-5.54923475e-01 3.63567740e-01 4.96324152e-01 -1.45019501e-01
7.14937210e-01 4.29729857e-02 4.46314573e-01 -7.08871484e-01
6.35644794e-02 2.29376003e-01 -3.57415587e-01 9.93971527e-02
6.89290404e-01 -1.36777431e-01 9.80122909e-02 4.76640940e-01
8.81868079e-02 -4.31908458e-01 -1.44616425e-01 5.22188842e-01
1.13312638e+00 1.82608470e-01 -3.55052352e-02 -1.61087558e-01
4.48080331e-01 -1.61214426e-01 6.37099385e-01 4.58569974e-01
9.29122865e-02 1.17192757e+00 -7.81445801e-02 -5.87382257e-01
-1.48794687e+00 -1.13952637e+00 3.22311282e-01 4.66009468e-01
2.35957757e-01 -4.65712219e-01 -7.33217835e-01 -2.19783515e-01
1.86955836e-02 6.85422197e-02 -4.70633954e-01 1.86988726e-01
-6.36103511e-01 -2.06162199e-01 3.60813230e-01 6.37503862e-01
9.04550791e-01 -6.33419096e-01 -8.78974080e-01 2.24519577e-02
-2.28547066e-01 -1.33663893e+00 -7.06198752e-01 -4.46825512e-02
-1.09563529e+00 -9.77213621e-01 -1.19024515e+00 -5.32406271e-01
7.93138444e-01 6.45546675e-01 1.25155413e+00 -1.53389037e-01
-2.63785809e-01 6.66109204e-01 1.68930709e-01 3.05013984e-01
-2.46807531e-01 -3.09155554e-01 1.82522446e-01 -4.72115278e-02
-8.98323953e-02 -9.80981350e-01 -1.13630033e+00 4.71166819e-01
-8.91986787e-01 7.01203167e-01 4.46067512e-01 7.31820643e-01
7.36866891e-01 -3.46854270e-01 -4.79593307e-01 -6.47927403e-01
1.03038639e-01 9.25678015e-02 -8.78300548e-01 -1.83105841e-01
-4.86387044e-01 6.49867440e-03 4.61921722e-01 -2.23557770e-01
-1.09268320e+00 5.01957834e-01 -1.33077562e-01 -8.99911642e-01
8.79637673e-02 -9.67420638e-02 6.60793297e-03 -4.19741482e-01
5.57301044e-01 3.60996336e-01 3.13762128e-01 -4.51251715e-01
5.04463434e-01 3.65087509e-01 9.09740984e-01 -2.70321310e-01
8.13833416e-01 1.10531175e+00 2.39727482e-01 -6.28241360e-01
-6.44993365e-01 -6.38423920e-01 -1.04896045e+00 -5.79558134e-01
8.93873751e-01 -1.47183704e+00 -7.59909630e-01 1.03914726e+00
-1.27280772e+00 -5.34990132e-01 -2.02244386e-01 4.71034437e-01
-8.64376843e-01 5.69197357e-01 -1.00995195e+00 -5.04632056e-01
-3.46568018e-01 -1.08713651e+00 1.51870716e+00 -1.31620923e-02
-1.45657152e-01 -9.57438827e-01 1.28396705e-01 6.79279029e-01
3.03565115e-01 1.84115008e-01 2.31700599e-01 4.41680580e-01
-1.09305108e+00 1.14756957e-01 -2.61688501e-01 3.28136474e-01
-2.81025499e-01 -3.86740357e-01 -1.05196798e+00 -5.62582016e-01
9.79136974e-02 -4.72275972e-01 8.06888163e-01 4.42852587e-01
7.26425529e-01 1.94395885e-01 -1.88129440e-01 1.20290565e+00
1.40859175e+00 -3.66620749e-01 8.20438683e-01 2.57251501e-01
1.12055874e+00 4.72015589e-01 3.13990116e-01 6.41723454e-01
6.60406470e-01 8.99718285e-01 4.21322376e-01 -4.61330771e-01
-5.06281674e-01 -3.45348865e-01 4.49935108e-01 1.12396049e+00
-3.28461736e-01 6.67162687e-02 -8.63118887e-01 3.43351454e-01
-1.85362649e+00 -1.00740445e+00 -1.60992637e-01 2.32018781e+00
5.69853961e-01 -1.18463496e-02 1.92919150e-01 -4.29713465e-02
4.68629241e-01 3.24064106e-01 -6.13730848e-01 1.03661291e-01
-2.29090944e-01 -4.81824577e-02 6.76882446e-01 6.71051085e-01
-6.44938946e-01 9.18001890e-01 6.63323593e+00 6.34214997e-01
-8.76346707e-01 1.15431167e-01 4.11534488e-01 -4.13049757e-01
-4.36454356e-01 -1.34238631e-01 -7.87874997e-01 1.76763892e-01
6.02757752e-01 1.19283296e-01 6.49431229e-01 7.91608930e-01
1.98017105e-01 -3.12274605e-01 -1.10303390e+00 1.66568267e+00
3.69437069e-01 -1.51962912e+00 -1.00944331e-02 5.15515327e-01
9.20871735e-01 3.20128262e-01 -8.85027051e-02 -4.36594456e-01
5.55737078e-01 -5.94854474e-01 8.65143061e-01 7.35125124e-01
9.04640913e-01 -9.03010368e-01 4.17188585e-01 5.98557889e-01
-1.16371560e+00 2.24308193e-01 -5.13142288e-01 6.38415068e-02
6.26544118e-01 1.02557540e+00 -1.61397859e-01 4.70368892e-01
9.64340031e-01 9.81230617e-01 -4.07789409e-01 7.63569593e-01
-1.19778872e-01 -1.74610868e-01 -5.59902132e-01 5.98791122e-01
2.01342385e-02 -1.88645631e-01 3.50364566e-01 9.57244992e-01
4.59590495e-01 3.66141170e-01 4.93646935e-02 9.20215666e-01
-2.62928940e-03 -4.10194695e-01 -9.00414944e-01 6.63537741e-01
1.36298746e-01 1.29847610e+00 -5.50769150e-01 -4.68677551e-01
-4.22614962e-01 1.73970175e+00 4.00295734e-01 4.95532304e-02
-6.02801681e-01 1.95130244e-01 6.44135952e-01 4.33542699e-01
1.04448244e-01 -5.39172530e-01 -4.69325744e-02 -1.62605178e+00
5.73179089e-02 -8.96828413e-01 1.42583519e-01 -1.15179837e+00
-1.24295080e+00 6.37499511e-01 -3.06337684e-01 -1.28360248e+00
-6.23662770e-01 -4.17376131e-01 -3.35635804e-02 9.49600577e-01
-1.54243040e+00 -1.17114818e+00 -8.81909907e-01 1.23197484e+00
4.45892990e-01 2.34327130e-02 6.47730052e-01 4.27608877e-01
-6.07495978e-02 1.61002457e-01 -3.30403186e-02 -8.30110684e-02
8.08232009e-01 -1.04924142e+00 7.48994648e-01 8.84733379e-01
-4.97391857e-02 4.30538267e-01 3.08149189e-01 -7.02045918e-01
-1.70004606e+00 -9.13654268e-01 7.23300934e-01 -5.73267579e-01
1.78286523e-01 -5.71158707e-01 -4.49554861e-01 4.85170871e-01
1.57839820e-01 1.96113124e-01 3.02691847e-01 -1.63393095e-01
-3.69105518e-01 -5.81700727e-02 -1.15541911e+00 5.84212184e-01
1.53580081e+00 -8.42480540e-01 -3.35028112e-01 1.51299551e-01
5.30924737e-01 -9.24357474e-01 -7.55645335e-01 1.84778228e-01
8.27895701e-01 -1.76348138e+00 1.21475029e+00 2.11230218e-01
6.17433488e-01 -4.03392226e-01 -2.18337297e-01 -1.11365879e+00
-6.68443203e-01 -6.96493030e-01 -2.11916924e-01 7.12240040e-01
-9.27054510e-02 -3.48655105e-01 1.11434865e+00 6.02551937e-01
-3.54512893e-02 -5.36521196e-01 -9.07290757e-01 -7.17678249e-01
-4.40022141e-01 -4.63623106e-01 1.87268972e-01 6.48802757e-01
-3.03051174e-01 3.21643651e-01 -7.89789498e-01 1.13480091e-01
1.26416552e+00 2.85627581e-02 1.19058585e+00 -1.10533476e+00
-5.64341009e-01 -1.95422061e-02 -5.96514940e-01 -1.69060993e+00
-1.51208058e-01 -5.43513894e-01 -8.33066776e-02 -1.49886978e+00
4.08714175e-01 -7.03633204e-02 3.12850565e-01 -2.20905561e-02
2.65722126e-01 6.36739850e-01 2.39974797e-01 5.03707886e-01
-5.03183484e-01 5.93990862e-01 1.37951529e+00 1.33786157e-01
-3.43035087e-02 -2.87589043e-01 -2.40304127e-01 1.08623183e+00
1.98337674e-01 -2.24640384e-01 -3.57063890e-01 -7.31562257e-01
2.81023771e-01 2.90952086e-01 7.73277521e-01 -1.42112589e+00
6.06255829e-01 6.92536980e-02 5.77867508e-01 -8.00726950e-01
9.60186422e-01 -9.57939982e-01 5.90167403e-01 4.22698945e-01
3.87887955e-02 3.49623322e-01 -1.07580118e-01 5.75691700e-01
-1.38798580e-01 5.15140653e-01 1.12146735e+00 -4.63432252e-01
-7.43218482e-01 5.87697625e-01 -5.98112158e-02 -7.26248473e-02
7.95606434e-01 -4.88924026e-01 -4.01299559e-02 -6.33436322e-01
-6.41917169e-01 3.75356711e-02 1.23845792e+00 1.94057956e-01
9.37020719e-01 -1.48153591e+00 -5.79767585e-01 5.33315778e-01
5.77019155e-02 4.15716857e-01 4.74075645e-01 7.15833545e-01
-8.60397518e-01 2.59587675e-01 -3.12446892e-01 -9.17321563e-01
-1.34067285e+00 1.90909743e-01 5.04442513e-01 -2.28165731e-01
-1.05814385e+00 8.35264802e-01 4.86933649e-01 -4.99025613e-01
1.33257046e-01 -7.36041069e-02 4.60431993e-01 -3.64235133e-01
5.55559158e-01 4.60986674e-01 -9.24380198e-02 -8.75847936e-01
-2.31624052e-01 9.78894114e-01 1.77572265e-01 -5.24990261e-01
1.40757847e+00 -4.53875929e-01 -4.01028469e-02 2.43659541e-01
1.09381270e+00 -6.83030440e-03 -1.71888804e+00 -3.73192638e-01
-9.78563428e-01 -7.89109051e-01 4.00031060e-01 -4.88503933e-01
-1.38454211e+00 9.33644772e-01 5.53205729e-01 -5.22906303e-01
1.09464335e+00 8.66893083e-02 1.17211092e+00 2.56816119e-01
9.47612047e-01 -1.03254008e+00 1.76108792e-01 5.16654551e-01
7.52236426e-01 -1.10784566e+00 1.13449581e-01 -5.79615295e-01
-4.13567275e-01 9.37848866e-01 7.02171385e-01 -2.62128860e-01
5.35961449e-01 4.33821499e-01 6.98603913e-02 -5.03283203e-01
-7.00915575e-01 -1.07067734e-01 1.93265483e-01 7.41966486e-01
8.31464306e-02 -4.08244878e-01 3.94695133e-01 -2.06854358e-01
-1.36845812e-01 1.68300033e-01 4.36256319e-01 8.05371583e-01
-3.41172129e-01 -9.52719569e-01 -5.01826346e-01 4.42668999e-04
6.00529499e-02 -1.72169641e-01 -3.41726601e-01 5.39021671e-01
-2.77076755e-02 5.89561939e-01 3.49386558e-02 -5.95629513e-01
5.52121520e-01 -4.41404998e-01 7.94568837e-01 -4.23184514e-01
-4.77852106e-01 1.74940810e-01 -9.38549116e-02 -1.37222886e+00
-6.37017131e-01 -6.28218591e-01 -8.00371170e-01 -8.83884132e-01
5.43411151e-02 -5.13132870e-01 3.18125129e-01 6.23157740e-01
6.06297374e-01 9.13085416e-02 5.54367125e-01 -1.43463337e+00
-4.89477105e-02 -4.53402072e-01 -8.49022627e-01 3.64218026e-01
2.18878374e-01 -5.51971078e-01 -3.58877480e-01 1.19799875e-01] | [8.760184288024902, -2.5951480865478516] |
a4a671e7-57f7-4b9d-9953-8cef1a0924b9 | contrastive-learning-of-musical | 2103.09410 | null | https://arxiv.org/abs/2103.09410v2 | https://arxiv.org/pdf/2103.09410v2.pdf | Contrastive Learning of Musical Representations | While deep learning has enabled great advances in many areas of music, labeled music datasets remain especially hard, expensive, and time-consuming to create. In this work, we introduce SimCLR to the music domain and contribute a large chain of audio data augmentations to form a simple framework for self-supervised, contrastive learning of musical representations: CLMR. This approach works on raw time-domain music data and requires no labels to learn useful representations. We evaluate CLMR in the downstream task of music classification on the MagnaTagATune and Million Song datasets and present an ablation study to test which of our music-related innovations over SimCLR are most effective. A linear classifier trained on the proposed representations achieves a higher average precision than supervised models on the MagnaTagATune dataset, and performs comparably on the Million Song dataset. Moreover, we show that CLMR's representations are transferable using out-of-domain datasets, indicating that our method has strong generalisability in music classification. Lastly, we show that the proposed method allows data-efficient learning on smaller labeled datasets: we achieve an average precision of 33.1% despite using only 259 labeled songs in the MagnaTagATune dataset (1% of the full dataset) during linear evaluation. To foster reproducibility and future research on self-supervised learning in music, we publicly release the pre-trained models and the source code of all experiments of this paper. | ['John Ashley Burgoyne', 'Janne Spijkervet'] | 2021-03-17 | null | null | null | null | ['music-auto-tagging', 'music-classification'] | ['music', 'music'] | [ 3.13494533e-01 -1.45392001e-01 -3.29682171e-01 -1.43416762e-01
-1.22254241e+00 -1.05203378e+00 2.02241018e-01 -1.83994442e-01
-4.78789955e-01 5.20327210e-01 1.46328077e-01 8.74593481e-02
-4.04097527e-01 -5.70788920e-01 -7.21293807e-01 -4.01746154e-01
-2.60276765e-01 4.61940914e-01 -2.37066239e-01 -8.62419307e-02
1.84133165e-02 7.17261806e-02 -1.60323286e+00 4.72851008e-01
5.33979237e-01 1.08730686e+00 -2.02118922e-02 6.78955197e-01
2.39518031e-01 6.67205274e-01 -8.00930083e-01 -3.41207057e-01
5.93485594e-01 -5.54030061e-01 -8.01398396e-01 -2.12571636e-01
7.93387949e-01 7.17170164e-02 -2.94297904e-01 6.46765232e-01
8.64386678e-01 1.99325427e-01 6.61953747e-01 -1.18312347e+00
-5.69247663e-01 1.41969538e+00 -4.44437414e-01 6.75032064e-02
3.53292413e-02 4.55156304e-02 1.36608827e+00 -7.58285880e-01
2.76759088e-01 7.82746315e-01 1.14066112e+00 4.82194632e-01
-1.40617681e+00 -1.28542984e+00 -1.14164427e-01 -5.99230416e-02
-1.41374397e+00 -5.46989799e-01 7.91570127e-01 -4.86525387e-01
5.80011249e-01 3.96364421e-01 7.19673395e-01 1.28749788e+00
-5.61385095e-01 8.80217552e-01 9.57989573e-01 -3.49707365e-01
9.45065990e-02 -1.10963397e-01 4.37589437e-02 2.30890274e-01
-4.93423492e-02 2.48028100e-01 -9.66814160e-01 -1.59803063e-01
1.03144741e+00 -2.49027401e-01 -1.42396882e-01 -3.55217420e-02
-1.51056051e+00 6.26787245e-01 3.15798849e-01 4.86246854e-01
-5.31486422e-02 5.68147898e-01 6.01728380e-01 5.51440358e-01
4.81527537e-01 1.05975068e+00 -7.28593946e-01 -6.28751695e-01
-1.31879485e+00 2.93429822e-01 5.12162447e-01 8.62993062e-01
3.57142359e-01 4.99929219e-01 -9.29886624e-02 1.22702456e+00
-2.80801922e-01 4.43900824e-01 8.20792913e-01 -1.19303155e+00
3.36966336e-01 3.38399023e-01 -1.91876650e-01 -6.22401416e-01
-4.57627714e-01 -1.29410875e+00 -8.86417687e-01 -5.00739105e-02
5.71065009e-01 -2.60922629e-02 -5.82881987e-01 1.97974849e+00
-2.83940613e-01 5.78111470e-01 1.07416054e-02 8.64294410e-01
8.72106016e-01 3.96925241e-01 -1.65188000e-01 -6.51488602e-02
8.86437297e-01 -1.02581489e+00 -3.13719660e-01 -1.13559306e-01
6.11173451e-01 -8.42641711e-01 1.47006619e+00 7.84638107e-01
-1.04133344e+00 -9.77271438e-01 -1.00414741e+00 5.50118871e-02
-3.36026400e-02 6.79356337e-01 9.53733683e-01 4.64700073e-01
-7.64491439e-01 1.05485010e+00 -5.15981376e-01 -5.38887978e-02
5.84754705e-01 5.80644190e-01 -2.16613978e-01 4.11503077e-01
-9.76071358e-01 1.85017779e-01 5.10523558e-01 -2.65652776e-01
-1.10338128e+00 -1.04378831e+00 -4.22813058e-01 -2.94013657e-02
2.82896310e-01 -4.32781756e-01 1.62319112e+00 -1.24240065e+00
-1.40190101e+00 9.24281716e-01 2.32684314e-01 -6.03659332e-01
3.76036257e-01 -3.99181455e-01 -4.92822081e-01 -1.93178967e-01
2.03108534e-01 8.37729931e-01 8.52982163e-01 -9.52376127e-01
-6.10756993e-01 -1.02822751e-01 -1.24634914e-01 -4.27821577e-02
-6.12197697e-01 -2.69424289e-01 -3.07593375e-01 -1.32290125e+00
2.97562405e-02 -1.14736283e+00 1.63692068e-02 -4.03020680e-01
-3.96480680e-01 -7.73192942e-02 1.56649172e-01 -3.06001425e-01
1.45336568e+00 -2.53968692e+00 2.07860604e-01 2.75803208e-02
-1.62771225e-01 6.48635626e-02 -6.24794245e-01 1.52688757e-01
-4.49581504e-01 1.47811621e-01 -4.02004123e-01 -1.81883946e-01
-4.15761881e-02 -9.34578478e-03 -6.85893178e-01 9.96513218e-02
-6.08250089e-02 1.07480252e+00 -8.24826062e-01 -7.11810812e-02
-1.04051463e-01 2.09035501e-01 -8.09118986e-01 -2.14046668e-02
-2.37678841e-01 7.08922505e-01 5.12999296e-02 6.46154642e-01
1.39493361e-01 -1.80329695e-01 -3.63371670e-02 -2.32367188e-01
5.37637696e-02 5.59257030e-01 -1.30526114e+00 2.32195425e+00
-5.30035615e-01 5.28243244e-01 -4.24973279e-01 -8.57545078e-01
1.02448428e+00 2.64098644e-01 7.00295329e-01 -5.41537225e-01
-9.24241468e-02 3.93410057e-01 1.80113688e-01 -7.11330548e-02
4.05080378e-01 -2.80109733e-01 -4.35994714e-01 8.06919396e-01
3.79575491e-01 -1.36211678e-01 2.35583514e-01 1.64254140e-02
1.11530459e+00 3.36486608e-01 4.53594960e-02 -2.47233845e-02
7.19057098e-02 3.02585270e-02 6.44885659e-01 8.26954961e-01
1.71376362e-01 6.83943272e-01 4.05310392e-02 -3.26126754e-01
-9.74039316e-01 -1.09602344e+00 -2.79600948e-01 1.56232548e+00
-4.65960115e-01 -8.31734121e-01 -2.96051472e-01 -6.14677429e-01
1.69577435e-01 5.24412811e-01 -5.34150720e-01 -1.43178895e-01
-4.35602605e-01 -7.12064028e-01 1.15695143e+00 7.42913187e-01
3.08909714e-01 -1.16676903e+00 -1.32868454e-01 3.15738022e-01
-1.93520486e-01 -7.13817537e-01 -4.31645721e-01 4.32040513e-01
-1.15398705e+00 -9.76013303e-01 -5.98047733e-01 -8.73924434e-01
1.40889026e-02 1.18616022e-01 1.25156891e+00 -3.65025043e-01
-2.39489421e-01 3.28678608e-01 -4.33659315e-01 -6.46132588e-01
-3.43267858e-01 6.53815269e-01 3.57911974e-01 -1.92633912e-01
1.13032416e-01 -1.07769120e+00 -3.59265387e-01 3.72740746e-01
-6.15666986e-01 -4.56594042e-02 4.23611194e-01 8.78531516e-01
7.46575654e-01 7.82531649e-02 1.16178977e+00 -8.46514642e-01
4.56068754e-01 -2.26395801e-01 -3.29828858e-01 -1.99605852e-01
-5.96961856e-01 -1.28446072e-01 5.97907066e-01 -8.63772571e-01
-4.20777500e-01 2.82278210e-01 -1.48435742e-01 -4.95738089e-01
-8.68381467e-03 4.33107048e-01 2.33996257e-01 2.99298972e-01
1.11675632e+00 8.18810612e-02 -2.39171550e-01 -9.90098476e-01
5.58484256e-01 8.58343244e-01 1.01230550e+00 -7.42915332e-01
9.92647767e-01 2.81079799e-01 -2.18517348e-01 -5.17147481e-01
-1.25461292e+00 -4.69065338e-01 -6.41831160e-01 -1.15253724e-01
4.12427217e-01 -1.26949871e+00 -7.40310431e-01 2.19461545e-01
-4.84225124e-01 -6.21530950e-01 -9.90882277e-01 6.63067997e-01
-8.73284578e-01 8.59155059e-02 -6.33709431e-01 -8.50280106e-01
-4.54293579e-01 -5.28356075e-01 1.19083464e+00 -1.97381631e-01
-6.73236310e-01 -6.34660542e-01 3.87940764e-01 4.89914626e-01
1.20972544e-01 -7.51647279e-02 8.24622273e-01 -7.50832379e-01
-3.21116865e-01 -8.76375064e-02 1.26397207e-01 5.38450003e-01
2.49915212e-01 -2.72330731e-01 -1.56144524e+00 -3.20896000e-01
-4.40028131e-01 -8.55859816e-01 1.16834986e+00 2.72680521e-01
1.55522597e+00 -1.99892651e-02 2.02071235e-01 7.35008240e-01
9.16639626e-01 -8.49376840e-04 3.36906970e-01 4.35309798e-01
6.40822530e-01 2.18280151e-01 7.67516375e-01 4.93198991e-01
-3.26047838e-02 7.60431349e-01 1.89478457e-01 -7.14883134e-02
-3.96858811e-01 -6.38799727e-01 4.88211393e-01 1.24466443e+00
-2.82522947e-01 2.04122663e-01 -6.45192683e-01 5.69874048e-01
-1.85154176e+00 -1.01870072e+00 6.40893728e-03 2.35601497e+00
1.14632928e+00 4.18442488e-03 5.13384342e-01 8.96517932e-01
4.13209856e-01 -4.28269319e-02 -6.57800436e-01 -5.51506178e-03
-2.48710603e-01 7.31103539e-01 1.15388639e-01 -2.75910068e-02
-1.38304830e+00 1.12094510e+00 7.20787859e+00 1.01587570e+00
-1.06581771e+00 1.65179312e-01 2.52725631e-01 -5.94725013e-01
-1.67591363e-01 -1.21399447e-01 -4.74577487e-01 2.25442812e-01
1.09863305e+00 -1.32322405e-02 7.98900366e-01 8.35022748e-01
1.42497346e-01 4.39516515e-01 -1.35033751e+00 1.34073770e+00
-1.45722404e-01 -1.31132066e+00 7.03434497e-02 -4.57734205e-02
8.62485468e-01 9.99757126e-02 2.70337135e-01 7.32080102e-01
1.69869617e-01 -1.11799920e+00 7.54595280e-01 3.10339868e-01
1.16784286e+00 -8.79469097e-01 4.18373972e-01 2.84255624e-01
-1.11749125e+00 -3.52605224e-01 -3.62713575e-01 -4.15367723e-01
-2.77891487e-01 4.97375369e-01 -8.44295084e-01 5.06999016e-01
6.87860310e-01 1.25290942e+00 -7.70072043e-01 9.69208896e-01
-9.11323279e-02 1.07431507e+00 -1.64780691e-01 4.05264795e-01
-1.40844837e-01 1.15575440e-01 3.23592663e-01 1.03230286e+00
4.08712924e-01 -4.68262970e-01 1.10904835e-01 7.94679105e-01
-4.08835709e-01 3.31593782e-01 -3.67109150e-01 -5.26224077e-01
4.66628969e-01 1.26201570e+00 -3.84789467e-01 -2.07092881e-01
-1.06649788e-03 7.98657417e-01 3.64951015e-01 1.40725553e-01
-6.72057867e-01 -1.63879424e-01 5.43908238e-01 1.24351501e-01
1.14445195e-01 -3.03549096e-02 -5.17169893e-01 -1.27138555e+00
-2.39664838e-01 -1.24040997e+00 5.42017221e-01 -8.13057184e-01
-1.56128609e+00 5.02332032e-01 -3.53616685e-01 -1.74180233e+00
-4.63107079e-01 -4.42264408e-01 -3.55410874e-01 5.56398869e-01
-1.18311846e+00 -1.16969657e+00 -1.46507949e-01 6.19086862e-01
5.58374465e-01 -6.88625455e-01 1.25620711e+00 6.27767146e-01
-2.84337491e-01 9.88783777e-01 6.10807352e-02 3.86317044e-01
1.14993548e+00 -1.13623214e+00 4.02728587e-01 2.14661047e-01
1.19413316e+00 4.78760600e-01 2.25316525e-01 -3.78146648e-01
-1.14498162e+00 -1.24109471e+00 4.00111198e-01 -4.96539623e-01
7.45366752e-01 -3.93662542e-01 -6.94023788e-01 8.10480952e-01
-2.32738331e-01 -3.28862756e-01 1.35427296e+00 7.89831758e-01
-6.97951615e-01 -3.53745192e-01 -5.58557630e-01 4.43616688e-01
1.42205334e+00 -7.60056317e-01 -6.50709987e-01 2.24260971e-01
5.88429570e-01 -1.01510867e-01 -1.12523425e+00 7.20984101e-01
8.42207849e-01 -5.98859966e-01 8.69239867e-01 -7.54930317e-01
4.24559593e-01 -4.02681828e-01 -2.92708576e-01 -1.24811614e+00
-4.43691343e-01 -5.35157859e-01 -1.99687202e-02 1.43595159e+00
6.76776171e-01 -1.14323005e-01 8.62485290e-01 -2.94898272e-01
-1.91703275e-01 -4.63622689e-01 -7.56770849e-01 -1.04666138e+00
1.62778750e-01 -1.13281476e+00 4.43114817e-01 1.34458983e+00
3.54627520e-02 5.78651845e-01 -5.25316179e-01 -2.91012645e-01
4.89246279e-01 5.38396358e-01 1.17438042e+00 -1.66721678e+00
-9.82978523e-01 -3.76166970e-01 -4.25691158e-01 -8.46032143e-01
2.48570815e-01 -1.60558116e+00 -2.17467666e-01 -1.11709774e+00
3.91814351e-01 -8.38849127e-01 -7.04526484e-01 8.84019256e-01
1.71196654e-01 1.08912897e+00 3.66902083e-01 5.27200580e-01
-5.36158800e-01 3.85853142e-01 1.10749710e+00 -4.50685114e-01
-4.71782565e-01 2.18870819e-01 -9.55799103e-01 7.32103884e-01
8.65856647e-01 -5.15083015e-01 -4.43750292e-01 -2.88234591e-01
3.11238199e-01 -2.65814304e-01 2.14497179e-01 -1.26207316e+00
-1.59622684e-01 3.56542170e-01 5.20545065e-01 -4.05195087e-01
5.50456703e-01 -4.50226039e-01 2.75659353e-01 1.42479017e-01
-7.97569633e-01 -3.14204693e-01 4.28781569e-01 4.18755651e-01
-2.73699582e-01 -9.57568064e-02 5.98238051e-01 4.65790369e-02
-4.09126073e-01 1.67991951e-01 -5.46179675e-02 3.28735232e-01
4.68917310e-01 -1.94448549e-02 6.19221516e-02 -5.61805248e-01
-1.12289357e+00 -2.42804408e-01 2.12566644e-01 6.47905469e-01
1.75696388e-01 -1.61443949e+00 -8.33356738e-01 2.18645439e-01
2.77348965e-01 -1.63434714e-01 1.13081314e-01 5.48249125e-01
1.18774744e-02 2.71544874e-01 -1.21172965e-01 -7.88398445e-01
-1.14850426e+00 4.46684539e-01 7.78181478e-02 -1.66882455e-01
-5.02637506e-01 8.72311532e-01 1.25630945e-01 -7.62102187e-01
5.22322178e-01 -3.27696800e-01 -1.77928787e-02 2.65886039e-01
4.62079763e-01 3.40131730e-01 -5.75972013e-02 -4.80785251e-01
-1.15307420e-01 7.59123385e-01 2.43713766e-01 -3.30748171e-01
1.45419443e+00 4.80054557e-01 2.10277945e-01 1.10444927e+00
8.88442814e-01 3.43706489e-01 -9.55239713e-01 -5.33291399e-02
9.24917776e-03 -2.63297856e-01 4.33765538e-02 -1.17629087e+00
-1.07919979e+00 8.82162929e-01 7.10284472e-01 -2.87726820e-02
1.27656114e+00 7.34937340e-02 5.74092031e-01 4.70769972e-01
4.27623659e-01 -9.77307200e-01 2.09941432e-01 4.12331879e-01
9.89901185e-01 -9.53257680e-01 -4.57851216e-02 -1.12126380e-01
-6.38042569e-01 9.02375340e-01 4.39353406e-01 -1.08061738e-01
4.72219974e-01 3.16433519e-01 2.20955729e-01 1.30242631e-01
-6.58896983e-01 -3.48740250e-01 4.79560345e-01 4.58814532e-01
7.90769398e-01 1.17337175e-01 -1.19848169e-01 1.22925663e+00
-9.44520712e-01 2.32329905e-01 2.93857396e-01 5.16231596e-01
-9.87476483e-02 -1.36318541e+00 -1.77128926e-01 6.14199638e-01
-5.75330794e-01 -2.18193322e-01 -7.04938352e-01 6.68657482e-01
3.09440494e-01 9.60790157e-01 1.24362275e-01 -8.40940654e-01
3.69212836e-01 3.56493950e-01 6.23201728e-01 -7.12338805e-01
-9.63497818e-01 4.76146370e-01 1.64484069e-01 -2.90319860e-01
-5.81996024e-01 -4.47511226e-01 -1.21854031e+00 1.25706731e-03
-2.08360806e-01 2.04823524e-01 4.98124689e-01 5.39418638e-01
4.13165838e-01 7.49503732e-01 6.57800078e-01 -8.71598840e-01
-5.28413415e-01 -1.27644253e+00 -8.99933815e-01 4.70768034e-01
-5.59302345e-02 -5.39027274e-01 -3.34708631e-01 3.82488668e-02] | [15.793618202209473, 5.262299060821533] |
22e82438-bd28-45c3-a30b-abd08a0415e4 | the-analysis-of-online-event-streams | 2203.09619 | null | https://arxiv.org/abs/2203.09619v1 | https://arxiv.org/pdf/2203.09619v1.pdf | The Analysis of Online Event Streams: Predicting the Next Activity for Anomaly Detection | Anomaly detection in process mining focuses on identifying anomalous cases or events in process executions. The resulting diagnostics are used to provide measures to prevent fraudulent behavior, as well as to derive recommendations for improving process compliance and security. Most existing techniques focus on detecting anomalous cases in an offline setting. However, to identify potential anomalies in a timely manner and take immediate countermeasures, it is necessary to detect event-level anomalies online, in real-time. In this paper, we propose to tackle the online event anomaly detection problem using next-activity prediction methods. More specifically, we investigate the use of both ML models (such as RF and XGBoost) and deep models (such as LSTM) to predict the probabilities of next-activities and consider the events predicted unlikely as anomalies. We compare these predictive anomaly detection methods to four classical unsupervised anomaly detection approaches (such as Isolation forest and LOF) in the online setting. Our evaluation shows that the proposed method using ML models tends to outperform the one using a deep model, while both methods outperform the classical unsupervised approaches in detecting anomalous events. | ['Hajo A. Reijers', 'Xixi Lu', 'Suhwan Lee'] | 2022-03-17 | null | null | null | null | ['activity-prediction', 'activity-prediction'] | ['computer-vision', 'time-series'] | [ 4.22494590e-01 1.86707992e-02 4.50954698e-02 -2.37838805e-01
-2.63801813e-01 -2.14140475e-01 7.85498261e-01 1.08090043e+00
-3.01172495e-01 5.87329626e-01 -1.88438799e-02 -6.47538066e-01
-3.98260981e-01 -1.11291802e+00 -3.90861362e-01 -5.30841529e-01
-4.69323397e-01 6.45716131e-01 2.93153822e-01 2.84359753e-01
5.91877937e-01 8.93953264e-01 -1.48104477e+00 3.23414862e-01
8.50616157e-01 1.16753531e+00 -5.81757367e-01 7.07210600e-01
-6.05519056e-01 1.07105434e+00 -8.24478567e-01 -7.14550465e-02
1.05579130e-01 -4.24357355e-01 -6.42006397e-01 6.75741062e-02
-1.11975767e-01 -4.25590158e-01 1.12832509e-01 8.37204397e-01
2.70636263e-03 2.91239887e-01 5.38130999e-01 -1.30523837e+00
3.99630480e-02 4.58790272e-01 -5.79426348e-01 8.33871961e-01
5.32383323e-01 2.09104776e-01 7.71474719e-01 -4.47880805e-01
3.30604315e-01 9.59202290e-01 5.40188372e-01 5.15827656e-01
-1.02328181e+00 -2.85394430e-01 5.65280735e-01 3.44557464e-01
-8.25452685e-01 -2.69845456e-01 5.22424817e-01 -3.63794744e-01
9.66235995e-01 2.15375513e-01 5.12559474e-01 1.33031273e+00
4.28026736e-01 9.66488719e-01 6.81927443e-01 -5.49668789e-01
7.26139307e-01 -3.05543721e-01 2.41875812e-01 4.17534918e-01
4.35184270e-01 7.62414336e-02 -6.42818332e-01 -8.33838344e-01
4.31231797e-01 8.68253648e-01 9.92248058e-02 2.84735113e-02
-1.06121516e+00 7.17254281e-01 -1.32645145e-01 5.47307134e-01
-1.07117724e+00 -8.97775441e-02 5.92067599e-01 3.62126887e-01
6.06623948e-01 5.07819653e-01 -4.85561490e-01 -3.12439531e-01
-8.89623940e-01 2.94889629e-01 1.14094484e+00 3.69358182e-01
2.90920079e-01 3.56394142e-01 -5.24611712e-01 2.97563851e-01
3.51241469e-01 -1.82525404e-02 8.23304772e-01 -4.01316375e-01
2.50480384e-01 9.07589674e-01 1.55586347e-01 -7.87217736e-01
-4.08388317e-01 -2.97370464e-01 -7.15101898e-01 7.30453879e-02
5.81310987e-01 -1.11574575e-01 -7.42501616e-01 1.15400958e+00
1.98610470e-01 5.37049711e-01 6.48120195e-02 2.91250259e-01
-1.56060547e-01 5.76145351e-01 4.11066532e-01 -5.09317458e-01
1.15542197e+00 -5.33696532e-01 -9.53483224e-01 -1.52443469e-01
7.27484345e-01 -5.46621263e-01 6.07944965e-01 7.68755317e-01
-6.75395906e-01 -2.58409947e-01 -7.34435737e-01 9.10831392e-01
-5.18256187e-01 -1.57604888e-01 4.42028880e-01 5.59788108e-01
-5.83609819e-01 1.06125474e+00 -1.34956431e+00 -6.90481722e-01
5.41579485e-01 2.33279675e-01 4.28699702e-03 1.74066216e-01
-8.91792059e-01 5.39199829e-01 7.05255628e-01 1.46139652e-01
-8.47215116e-01 -1.82309866e-01 -6.87613904e-01 1.93799600e-01
6.85401022e-01 -1.66015998e-01 1.41924798e+00 -8.94667447e-01
-1.10400021e+00 2.06616387e-01 -4.21342075e-01 -1.20884860e+00
6.86068356e-01 -6.75778985e-01 -7.41675436e-01 -6.52103033e-03
-1.61248758e-01 -4.10890847e-01 8.46772790e-01 -7.87655890e-01
-1.15282023e+00 -6.00161076e-01 -3.53752345e-01 -3.01597118e-01
-3.59485298e-01 3.33917551e-02 2.96812981e-01 -5.86095452e-01
1.69957235e-01 -5.36527753e-01 -5.10070324e-01 -5.74079931e-01
-7.35730350e-01 -4.43666875e-01 1.15436256e+00 -8.13705981e-01
1.37957370e+00 -2.00614190e+00 -6.22441232e-01 7.67514586e-01
1.33960515e-01 3.62485558e-01 5.83048403e-01 5.99430323e-01
-1.75762619e-03 5.18781506e-02 -3.54052842e-01 -1.20611504e-01
-2.63854980e-01 2.80886471e-01 -5.44445157e-01 3.52153957e-01
5.01276374e-01 4.74872738e-01 -8.15726161e-01 -2.52966702e-01
3.57462674e-01 -2.52029657e-01 -1.52822584e-01 5.41607678e-01
-3.72484684e-01 6.14306152e-01 -7.35726058e-01 9.43190813e-01
1.60675675e-01 -6.98931739e-02 9.31925923e-02 6.97064936e-01
-2.06956923e-01 2.25570157e-01 -1.13890946e+00 9.31258500e-01
-3.59766632e-01 3.71296167e-01 -4.90517110e-01 -1.32239974e+00
1.09587955e+00 4.96038258e-01 6.79760575e-01 -5.70823789e-01
-2.36751750e-01 3.01808149e-01 5.55574298e-02 -4.69101787e-01
1.90472379e-01 1.96432233e-01 7.18795881e-02 8.96682858e-01
-2.68891335e-01 1.07021260e+00 2.81232268e-01 -9.23193321e-02
1.76229000e+00 7.62612522e-02 7.38987446e-01 1.29002839e-01
9.13551033e-01 -4.97575104e-02 6.71480775e-01 1.04249167e+00
-2.38627166e-01 -6.82401955e-02 7.84614623e-01 -8.92689109e-01
-6.17206812e-01 -1.03847420e+00 2.42771149e-01 9.27248359e-01
-3.48795176e-01 -3.11532825e-01 -5.11821389e-01 -1.20387518e+00
1.19014569e-01 1.08836389e+00 -3.23281109e-01 -3.39438349e-01
-6.34758830e-01 -8.59563291e-01 3.37761551e-01 5.76163709e-01
4.85679418e-01 -1.51143122e+00 -9.48645175e-01 7.11986184e-01
2.35839844e-01 -8.81772459e-01 3.29672217e-01 5.29081881e-01
-1.25744200e+00 -1.27521467e+00 -2.68834010e-02 -1.00111097e-01
5.30542135e-01 -4.51206207e-01 9.70919251e-01 3.34501192e-02
-2.58608907e-01 2.12774619e-01 -4.47254062e-01 -8.55168641e-01
-5.77279210e-01 -9.14213732e-02 2.15275034e-01 5.11919916e-01
1.07926190e+00 -5.79099953e-01 -4.89324510e-01 5.60452081e-02
-1.00738072e+00 -8.62334847e-01 7.05501318e-01 5.31046152e-01
5.92285693e-01 2.00590432e-01 7.86012709e-01 -1.30701697e+00
9.77999091e-01 -6.65551007e-01 -5.01752853e-01 2.00904503e-01
-1.03045928e+00 1.68535739e-01 9.44983244e-01 -1.79370672e-01
-1.13843095e+00 -8.47884268e-02 -8.92913118e-02 -3.73094201e-01
-9.66424108e-01 2.63410658e-01 -2.31846720e-02 4.61049348e-01
5.29041767e-01 4.71541613e-01 -3.19225788e-01 -6.10758185e-01
-3.43945414e-01 5.93286932e-01 2.11163089e-01 -3.62454414e-01
5.33090830e-01 6.03880167e-01 3.83788981e-02 -7.40685284e-01
-5.88424265e-01 -7.70061851e-01 -6.13013864e-01 -2.29441181e-01
6.04446292e-01 -3.45759124e-01 -7.76893139e-01 3.55516970e-01
-1.00820947e+00 8.91203955e-02 -5.78549027e-01 3.15990686e-01
-4.48564947e-01 3.61891091e-01 -6.72769785e-01 -1.46885335e+00
-4.28042561e-01 -5.36428392e-01 1.01213717e+00 1.69263065e-01
-5.94708383e-01 -1.05690646e+00 1.95414320e-01 -4.40819301e-02
3.14615935e-01 5.90185583e-01 7.07605362e-01 -2.01362276e+00
-4.51837689e-01 -7.41956949e-01 2.26083159e-01 1.58638507e-01
3.80551755e-01 4.42420132e-02 -8.93682361e-01 -1.20232932e-01
1.55140430e-01 4.19293821e-01 6.56945407e-01 2.05222741e-01
1.56559134e+00 -4.86749977e-01 -4.54570323e-01 1.76510274e-01
1.05406463e+00 6.98685944e-01 5.79414070e-01 6.83381677e-01
5.01312017e-01 6.49812281e-01 8.38523746e-01 8.58480215e-01
-1.32664919e-01 2.55082041e-01 3.89104366e-01 2.68043756e-01
9.69956994e-01 -2.20402464e-01 4.48427349e-01 3.64090383e-01
-3.97808522e-01 -2.91624635e-01 -1.12170589e+00 4.13390845e-01
-2.26047182e+00 -1.30317760e+00 -3.25214595e-01 2.55707121e+00
8.90321359e-02 6.50814116e-01 3.56132507e-01 5.69310069e-01
7.94128895e-01 5.81009164e-02 -5.08413851e-01 -8.49043548e-01
2.14335233e-01 2.76505768e-01 3.32672149e-01 -1.36323720e-02
-1.34553957e+00 4.26231742e-01 5.63971901e+00 3.78467977e-01
-7.16069520e-01 5.00577018e-02 8.04036260e-01 -1.23274699e-01
2.23520204e-01 4.08775248e-02 -6.94820642e-01 6.08714104e-01
1.40682030e+00 4.38178368e-02 -5.56650981e-02 1.03881824e+00
5.89158475e-01 -2.25205407e-01 -1.31208479e+00 5.32907903e-01
-1.93567961e-01 -8.95277977e-01 6.91538081e-02 3.41274917e-01
4.80469227e-01 -3.24593008e-01 -5.05428016e-01 2.50666887e-01
1.43682957e-01 -8.04062724e-01 1.33895740e-01 1.06368971e+00
-4.41509396e-01 -1.09367216e+00 1.03662694e+00 5.00294805e-01
-8.99139881e-01 -5.47705710e-01 1.43935874e-01 -2.47983024e-01
8.15811306e-02 1.03776288e+00 -1.00673687e+00 6.24800801e-01
6.78784311e-01 7.18040943e-01 -3.99979889e-01 1.00616086e+00
-1.47804826e-01 9.19236183e-01 -2.12253049e-01 3.94387310e-03
3.23570907e-01 -9.96210128e-02 7.05610096e-01 9.82153594e-01
4.39468175e-01 -5.41745722e-01 4.42693383e-01 7.50493407e-01
4.08391118e-01 1.00907832e-01 -8.33464622e-01 -2.89142042e-01
2.13357866e-01 9.18516517e-01 -1.00541961e+00 -4.04588252e-01
-4.82068658e-01 1.01715398e+00 -4.92417514e-02 1.92549586e-01
-5.06713510e-01 -4.02349114e-01 5.76537728e-01 4.24499571e-01
2.20664799e-01 -9.66601595e-02 -2.47279525e-01 -9.14416790e-01
2.30713278e-01 -7.87862778e-01 9.49090779e-01 2.16393322e-02
-1.53851104e+00 3.71635824e-01 -2.43151918e-01 -1.31169367e+00
-8.30638230e-01 -5.92541218e-01 -1.36067450e+00 5.37743509e-01
-1.28461063e+00 -5.71424425e-01 -1.67496294e-01 7.26216435e-01
6.62901461e-01 -2.98009396e-01 7.06578314e-01 -1.03095807e-02
-7.64881790e-01 7.57220536e-02 2.05018505e-01 1.34735495e-01
5.24713099e-01 -1.52700019e+00 5.55645287e-01 1.18953633e+00
2.80228019e-01 4.84084547e-01 6.84550524e-01 -9.09884751e-01
-8.88390958e-01 -1.25268233e+00 8.82653236e-01 -4.19223815e-01
7.03961730e-01 3.37076448e-02 -1.41284871e+00 8.28983068e-01
-2.04652250e-01 5.45200706e-02 8.65813553e-01 2.31326953e-01
2.12582096e-01 -8.81640911e-02 -1.18896472e+00 3.53517890e-01
8.81711304e-01 -1.48166671e-01 -6.19932175e-01 2.43079230e-01
1.14472643e-01 8.97379220e-02 -6.25489831e-01 4.08449262e-01
1.60610408e-01 -1.22053838e+00 5.64069688e-01 -1.02472115e+00
2.19047084e-01 -6.28568381e-02 2.26179555e-01 -1.07645380e+00
-2.08190065e-02 -5.37550151e-01 -9.40304399e-01 1.08035588e+00
1.48726687e-01 -9.53735054e-01 8.99198115e-01 4.13519561e-01
1.89508960e-01 -6.95360959e-01 -9.23550606e-01 -6.85800016e-01
-7.24986792e-01 -6.17586493e-01 6.51402473e-01 9.06987488e-01
-1.21563025e-01 -2.07401261e-01 -1.42871439e-01 3.32513481e-01
6.13828599e-01 1.81563109e-01 5.75432062e-01 -1.66033995e+00
-3.25928092e-01 -4.15104896e-01 -6.66932642e-01 -2.77579099e-01
1.91035271e-02 -3.08944076e-01 -1.02507681e-01 -1.21295357e+00
-2.43839234e-01 2.83804350e-02 -9.64387238e-01 4.58679199e-01
-2.12878764e-01 -3.15973550e-01 -3.64412993e-01 3.17504942e-01
-7.47056186e-01 2.91573703e-01 3.30481350e-01 1.77678570e-01
-5.01776457e-01 6.84030116e-01 -2.90665120e-01 1.18256927e+00
1.12191987e+00 -4.92556512e-01 -6.67871833e-02 3.11052769e-01
1.15090713e-01 1.04096124e-03 4.92844671e-01 -1.22037935e+00
2.89268434e-01 -2.08131582e-01 6.67544603e-01 -6.54470444e-01
-3.27071995e-01 -7.74290442e-01 -3.87447685e-01 8.69279802e-01
-4.23261404e-01 3.65501195e-01 -4.93442044e-02 1.23747182e+00
-5.07262290e-01 -2.94836760e-01 2.70182818e-01 -2.90581584e-01
-7.42095411e-01 1.29376277e-01 -9.86334264e-01 -2.72251755e-01
1.42976558e+00 -2.52488405e-01 -8.89593735e-02 -4.10398751e-01
-9.60489213e-01 2.55189121e-01 -1.11241661e-01 3.33972484e-01
6.14589214e-01 -1.06078362e+00 -4.96220976e-01 3.36792171e-01
3.39003950e-01 -9.93273687e-04 1.02235556e-01 9.47736561e-01
-4.96989757e-01 3.14123511e-01 -1.59167573e-01 -5.44224858e-01
-1.06357837e+00 4.83780473e-01 2.29589552e-01 -8.26294422e-01
-6.39603078e-01 2.16396272e-01 -3.77616554e-01 -3.42452489e-02
2.69602954e-01 -1.25269845e-01 -4.06816125e-01 -3.57212909e-02
6.66768909e-01 7.06333697e-01 3.85035485e-01 5.18199848e-03
-4.42537367e-01 -2.99753487e-01 -4.56492186e-01 1.84643805e-01
1.17721224e+00 2.55828202e-01 -1.72444612e-01 8.18613172e-01
2.78745145e-01 1.94263197e-02 -8.62402678e-01 -3.27131510e-01
1.18903720e+00 -4.54397380e-01 -1.85850024e-01 -6.26723528e-01
-7.84523845e-01 8.21064532e-01 6.82929158e-01 9.07857120e-01
1.18859410e+00 -1.80672720e-01 7.50381470e-01 6.33113980e-01
1.29374564e-01 -1.15006101e+00 -2.84299757e-02 4.49489176e-01
2.06144631e-01 -1.09043694e+00 -1.52557507e-01 4.02633324e-02
-4.96794224e-01 1.32510769e+00 7.78631151e-01 -1.67768061e-01
5.63477397e-01 4.55456339e-02 -1.78862929e-01 -2.49252781e-01
-8.62696946e-01 -2.09046945e-01 5.55026010e-02 3.89748186e-01
4.13536221e-01 -2.53412202e-02 -2.02903777e-01 4.52034473e-01
2.53889412e-01 -1.43496305e-01 3.38735670e-01 1.42051673e+00
-4.65822071e-01 -9.94966209e-01 -5.33593595e-01 1.20291722e+00
-9.75993574e-01 2.56710529e-01 -5.83217561e-01 5.61931431e-01
6.48160055e-02 1.02635574e+00 6.15574479e-01 -1.37564853e-01
5.81392348e-01 8.42549205e-01 -2.15157643e-01 -5.13108313e-01
-6.13757730e-01 -1.38981879e-01 1.73558399e-01 -7.85228550e-01
-1.47828892e-01 -7.88781285e-01 -1.04522300e+00 -1.46680802e-01
-1.10500418e-01 6.32487163e-02 4.32091534e-01 1.35198164e+00
3.62360477e-01 8.88726175e-01 6.65391564e-01 -3.02931815e-01
-6.49455011e-01 -1.07775104e+00 -5.94130516e-01 5.82906783e-01
2.40648434e-01 -4.13780659e-01 -3.18351746e-01 -2.34591007e-01] | [7.487866401672363, 2.5852413177490234] |
90cbdd03-8df0-49e9-8401-266d4f58d305 | provable-convergence-guarantees-for-black-box | 2306.03638 | null | https://arxiv.org/abs/2306.03638v1 | https://arxiv.org/pdf/2306.03638v1.pdf | Provable convergence guarantees for black-box variational inference | While black-box variational inference is widely used, there is no proof that its stochastic optimization succeeds. We suggest this is due to a theoretical gap in existing stochastic optimization proofs-namely the challenge of gradient estimators with unusual noise bounds, and a composite non-smooth objective. For dense Gaussian variational families, we observe that existing gradient estimators based on reparameterization satisfy a quadratic noise bound and give novel convergence guarantees for proximal and projected stochastic gradient descent using this bound. This provides the first rigorous guarantee that black-box variational inference converges for realistic inference problems. | ['Robert Gower', 'Guillaume Garrigos', 'Justin Domke'] | 2023-06-04 | null | null | null | null | ['stochastic-optimization'] | ['methodology'] | [-5.48545867e-02 -3.47934440e-02 -2.98852801e-01 -4.65860277e-01
-1.15352404e+00 -5.20735979e-01 5.39729416e-01 -2.80963242e-01
-3.38414133e-01 1.06568718e+00 6.78514838e-02 -4.66084391e-01
-1.20890439e-01 -3.82073581e-01 -9.38896179e-01 -8.52477074e-01
-1.27021849e-01 5.68816245e-01 -3.19586415e-03 1.41103178e-01
3.38297158e-01 2.66567200e-01 -1.06645083e+00 -4.77863520e-01
8.33438277e-01 5.83671331e-01 -5.26423715e-02 1.11496067e+00
-2.59311162e-02 4.02749717e-01 -1.55781507e-01 -6.85003042e-01
2.96669751e-01 -5.36028266e-01 -5.00875652e-01 -3.13145041e-01
9.91730094e-01 -7.65758455e-01 -2.26696059e-02 1.41612804e+00
4.26949918e-01 4.39979255e-01 8.65826905e-01 -1.22265542e+00
-5.75244725e-01 7.50399530e-01 -6.63967073e-01 2.39946634e-01
1.27690703e-01 -1.60855606e-01 1.33265841e+00 -1.20297325e+00
7.51139522e-01 1.39011204e+00 1.16493225e+00 7.40718782e-01
-1.47256494e+00 -2.92423636e-01 5.99459857e-02 -3.05080146e-01
-1.22850382e+00 -5.86804569e-01 5.11925399e-01 -4.79685754e-01
5.57884872e-01 2.60187119e-01 3.76668692e-01 1.39469242e+00
-2.18099952e-01 9.28381085e-01 7.97486901e-01 -1.40410572e-01
4.76382852e-01 3.07151556e-01 2.76584536e-01 9.30541456e-01
4.94116068e-01 1.06527865e-01 -5.59749842e-01 -5.98676205e-01
7.75679410e-01 4.32509445e-02 -1.76527187e-01 -4.48314905e-01
-9.60610688e-01 1.21744466e+00 -9.01200548e-02 -3.90168317e-02
-3.96899544e-02 7.14162350e-01 4.97834444e-01 2.00958654e-01
6.53592110e-01 4.20645066e-02 -4.36734706e-01 -6.24504328e-01
-1.40710044e+00 8.14492047e-01 1.04992759e+00 9.13090348e-01
8.15023899e-01 5.15717804e-01 -1.31125137e-01 6.40287995e-01
7.17600524e-01 1.07159472e+00 -1.39007598e-01 -1.45948470e+00
2.64260978e-01 -4.07476068e-01 4.51236159e-01 -9.26891327e-01
-8.51878449e-02 -1.74243554e-01 -6.39312267e-01 2.45622411e-01
8.23196650e-01 -5.43853879e-01 -4.70596552e-01 2.03141046e+00
5.86176574e-01 3.15412879e-01 -2.66169757e-01 8.50300729e-01
2.11217746e-01 6.09575629e-01 -4.04722393e-01 -4.98536885e-01
6.74434841e-01 -7.23376751e-01 -7.63552666e-01 2.88261697e-02
3.48642260e-01 -4.69397694e-01 1.10691917e+00 2.45612502e-01
-1.33417368e+00 -4.14243676e-02 -1.15277255e+00 -3.94679219e-01
5.64815030e-02 -2.18143165e-01 9.08740342e-01 9.63490069e-01
-9.65640008e-01 9.12445188e-01 -1.21645641e+00 -1.20389909e-01
4.84706163e-01 -6.72668070e-02 -3.45106088e-02 3.12852800e-01
-5.44819891e-01 5.85300565e-01 1.94742642e-02 3.35390061e-01
-1.13818705e+00 -1.29399967e+00 -6.99584842e-01 -9.20389891e-02
3.21202517e-01 -8.88995528e-01 1.33167017e+00 -4.81727988e-01
-1.79486811e+00 5.65549493e-01 -7.15098977e-01 -5.61799645e-01
1.23907924e+00 -4.98879939e-01 3.07669848e-01 1.10907294e-01
1.28437310e-01 -6.00682124e-02 1.23079586e+00 -1.06581068e+00
-4.75700349e-01 -6.65245831e-01 -1.68163180e-01 -2.41783727e-03
-1.53109729e-01 -2.76757777e-02 -1.77182361e-01 -2.82198161e-01
1.24341175e-02 -8.15608621e-01 -3.79083008e-01 2.26606339e-01
-4.71882999e-01 -2.14964882e-01 5.95449746e-01 -3.75100642e-01
1.05688071e+00 -1.96387982e+00 1.23070508e-01 2.82457322e-01
4.63624358e-01 -1.36849388e-01 3.45093697e-01 1.38937533e-01
5.17402351e-01 4.74230707e-01 -5.34885943e-01 -1.00547755e+00
5.87897718e-01 1.59330457e-01 -7.77654827e-01 9.42409933e-01
-2.00560108e-01 8.69912207e-01 -1.17104840e+00 -5.45700967e-01
1.63604334e-01 3.94989520e-01 -9.68487561e-01 1.06492609e-01
-1.16960451e-01 1.44619614e-01 -3.43000978e-01 2.82733172e-01
1.04188395e+00 -2.18421742e-01 2.60962322e-02 1.71623871e-01
-1.43537149e-02 2.29950368e-01 -1.54144502e+00 1.56419325e+00
-1.24960929e-01 8.13122571e-01 7.48513222e-01 -9.90403354e-01
2.69551009e-01 1.83435470e-01 3.71159852e-01 4.46539313e-01
-1.26779392e-01 5.74383914e-01 -7.45471060e-01 -1.80950627e-01
5.72653234e-01 -4.33978558e-01 2.52019793e-01 3.73835683e-01
1.66084096e-01 -5.04723072e-01 3.06386054e-01 3.92632186e-01
7.55996466e-01 3.37866724e-01 -2.75439531e-01 -6.11359239e-01
1.96682606e-02 -3.14365596e-01 6.48703516e-01 1.54544044e+00
-2.48936668e-01 9.43270743e-01 5.81704676e-01 -1.52583301e-01
-1.35676122e+00 -1.66362751e+00 -5.09137213e-01 1.25713575e+00
-3.19798708e-01 -5.44006705e-01 -7.86539555e-01 -5.03277063e-01
3.58958125e-01 6.97309613e-01 -8.36102188e-01 3.64390790e-01
-2.76215702e-01 -9.08428311e-01 6.62841141e-01 4.17522103e-01
1.18607469e-01 -2.51607746e-01 -1.88053995e-01 2.20214248e-01
2.17827827e-01 -1.02057767e+00 -7.91127861e-01 -4.53645438e-02
-1.21531832e+00 -1.01445675e+00 -8.12888622e-01 -2.16813460e-01
4.82383072e-01 8.15155804e-02 1.11904681e+00 -1.49788067e-01
3.21578141e-03 5.22683561e-01 3.52657259e-01 -3.32956642e-01
-5.26995599e-01 -1.79886714e-01 3.06064576e-01 -3.00773740e-01
2.74186850e-01 -7.80503452e-01 -5.05208731e-01 -1.35934979e-01
-5.57914734e-01 -3.56984526e-01 -1.37023389e-01 9.35483515e-01
5.38728178e-01 -3.50056976e-01 3.28767635e-02 -1.05857062e+00
7.35969782e-01 -5.55446386e-01 -1.09279490e+00 1.76797628e-01
-8.77322674e-01 2.60974526e-01 5.22511244e-01 -3.83537680e-01
-1.09673321e+00 -1.65248916e-01 -2.65161276e-01 -5.49593389e-01
4.15976286e-01 4.47239161e-01 5.12698889e-01 1.33416668e-01
7.85656571e-01 2.08517730e-01 1.50793359e-01 -4.71514225e-01
6.60162568e-01 1.94534630e-01 5.59572458e-01 -1.06750488e+00
8.68310213e-01 1.03341746e+00 2.71480858e-01 -9.41601574e-01
-1.14463246e+00 -3.07361484e-01 -3.11416030e-01 1.46794960e-01
6.94406331e-01 -7.32485473e-01 -1.04012382e+00 2.44898751e-01
-1.09801161e+00 -3.97951365e-01 -4.40350652e-01 6.21807158e-01
-9.55596864e-01 5.97352624e-01 -6.93133652e-01 -1.47699451e+00
-3.64426047e-01 -9.42196250e-01 9.54144180e-01 1.74415894e-02
-7.57743269e-02 -1.43011153e+00 4.64715034e-01 -9.88876563e-04
4.68141317e-01 1.35931909e-01 3.70484680e-01 -3.12437087e-01
-3.10952067e-01 4.19550724e-02 -1.94405109e-01 5.45987368e-01
-3.50895256e-01 6.66839361e-01 -7.42452443e-01 -2.06657812e-01
3.65141749e-01 -2.41670236e-01 1.09909689e+00 9.30749059e-01
8.11639607e-01 -4.00542319e-01 -1.03485055e-01 1.04685187e+00
1.46363616e+00 -8.11115146e-01 2.42479220e-01 -3.87544006e-01
8.74807656e-01 8.53077918e-02 -4.98005971e-02 7.60937870e-01
1.67013973e-01 2.72494793e-01 1.70329884e-02 3.38629007e-01
2.63385713e-01 -5.33425391e-01 6.51965141e-01 8.69904876e-01
-1.61210850e-01 3.02628756e-01 -6.54685736e-01 7.10783720e-01
-2.13648033e+00 -1.32151318e+00 -3.69321853e-01 2.26522875e+00
1.16751075e+00 9.22970250e-02 2.80288935e-01 -3.80180240e-01
5.17123222e-01 2.30943531e-01 -6.81645036e-01 -5.58017075e-01
-2.45884746e-01 2.92941123e-01 8.44064593e-01 1.15366077e+00
-8.46122503e-01 8.83609235e-01 8.62915325e+00 1.02121854e+00
-4.72723216e-01 6.83854342e-01 2.32788742e-01 -4.81775612e-01
-6.45152569e-01 1.88480422e-01 -1.16385424e+00 5.14140069e-01
9.48453307e-01 -2.29422316e-01 6.66632414e-01 1.15333438e+00
3.61315489e-01 -6.06808765e-03 -1.19659078e+00 1.04146445e+00
-2.09391430e-01 -1.56058300e+00 -2.10498676e-01 1.39312968e-01
1.12817955e+00 3.57791960e-01 1.56692669e-01 3.06073785e-01
9.02923048e-01 -8.76885951e-01 6.12501621e-01 5.45285344e-01
2.63717592e-01 -7.82636523e-01 2.76107043e-01 2.08737493e-01
-7.46286631e-01 3.03857237e-01 -7.62329638e-01 5.74386455e-02
3.94887656e-01 1.02384448e+00 -3.74278903e-01 -3.93117070e-02
3.99026394e-01 6.47024810e-01 5.90432808e-02 6.81537449e-01
-3.66525531e-01 1.21891212e+00 -1.00282860e+00 -3.00111234e-01
2.87888438e-01 -7.80602098e-01 1.11482012e+00 1.47275519e+00
2.48300180e-01 -3.70804161e-01 -6.27834275e-02 1.56492722e+00
-1.27400205e-01 -7.53944516e-02 -5.85926771e-01 -2.52307028e-01
3.91259968e-01 8.82112741e-01 -1.97550550e-01 -4.01365459e-01
-3.64720196e-01 9.62720394e-01 4.59707171e-01 7.22537458e-01
-9.67032313e-01 -2.20266089e-01 1.15186107e+00 -1.90737411e-01
5.76319516e-01 -6.21079624e-01 -6.38023019e-01 -1.58159196e+00
1.78651780e-01 -5.35763323e-01 1.88695014e-01 -3.60749364e-01
-1.52974367e+00 -3.74444872e-01 1.76758282e-02 -1.29690319e-01
-5.78286469e-01 -7.00303018e-01 -6.68577611e-01 7.42556214e-01
-9.67971921e-01 -8.08390558e-01 3.50504547e-01 5.06766915e-01
2.42330372e-01 9.69653353e-02 5.17832041e-01 6.83098510e-02
-4.87259477e-01 8.38642657e-01 8.35955560e-01 -6.81186467e-02
2.62376755e-01 -1.96027815e+00 3.45304966e-01 1.09282160e+00
4.06664878e-01 1.04058838e+00 1.27661371e+00 -4.76870745e-01
-1.93913996e+00 -4.76708323e-01 5.96248090e-01 -7.56126404e-01
1.06329310e+00 -6.00178123e-01 -7.74288595e-01 8.99487972e-01
-1.77217975e-01 1.71568051e-01 3.91953051e-01 7.25921631e-01
-5.47571123e-01 1.39565796e-01 -1.14143670e+00 7.40883768e-01
1.01576698e+00 -8.82141232e-01 -3.88321906e-01 7.51066506e-01
6.02336466e-01 -3.75575960e-01 -9.47022080e-01 1.35049030e-01
7.58542299e-01 -8.34166944e-01 1.03161645e+00 -8.18175614e-01
3.50458890e-01 -1.30835354e-01 -3.93917561e-01 -9.54335809e-01
1.96535945e-01 -1.29858553e+00 -7.42134333e-01 9.69152033e-01
4.37943935e-01 -6.33118331e-01 9.86888945e-01 8.18381071e-01
1.35412782e-01 -5.43865025e-01 -1.09194899e+00 -8.94667506e-01
5.31260490e-01 -9.28981662e-01 5.39445458e-03 6.80334270e-01
3.07340696e-02 1.23751208e-01 -6.01895690e-01 6.48604929e-02
1.29880869e+00 -1.54741824e-01 9.54927802e-01 -9.38300490e-01
-5.28831542e-01 -8.88041437e-01 -2.22969696e-01 -1.50890231e+00
3.82375687e-01 -7.89049447e-01 2.19338924e-01 -1.22730970e+00
4.94502813e-01 -1.32509202e-01 -1.12518281e-01 -3.24993193e-01
-5.79127371e-01 2.15959936e-01 -9.09220278e-02 1.14881672e-01
-6.76740468e-01 7.41512120e-01 8.38327289e-01 1.57135680e-01
-9.63766649e-02 1.54675663e-01 -4.58568335e-01 8.98835301e-01
4.05857116e-01 -5.18913627e-01 -3.54251355e-01 -3.79680365e-01
6.96131229e-01 -2.02298045e-01 4.28925753e-01 -5.18859625e-01
2.15467498e-01 -2.77705610e-01 -1.31406283e-04 -6.55730247e-01
3.11738759e-01 -1.45750582e-01 -1.98067650e-01 1.59905463e-01
-3.92634153e-01 -2.26831540e-01 -5.74039705e-02 8.54179025e-01
2.23917410e-01 -7.29178905e-01 8.44324589e-01 4.44864333e-02
-1.02468021e-01 4.75038975e-01 -2.53337890e-01 8.75155389e-01
3.21559101e-01 9.98968557e-02 1.94760635e-01 -5.53196251e-01
-6.74236357e-01 -3.07307648e-03 3.52847427e-01 -3.36824417e-01
5.42077899e-01 -1.23325014e+00 -1.24542236e+00 -1.26990259e-01
-4.29677993e-01 -9.26539376e-02 6.74377978e-02 1.00333083e+00
-4.51459974e-01 -3.08169983e-02 6.20908141e-01 -7.79893994e-01
-9.47127104e-01 3.52871984e-01 4.05401677e-01 -3.84601541e-02
-6.44870043e-01 1.25082099e+00 -4.97378968e-02 -5.57499766e-01
2.49070928e-01 -2.89495498e-01 7.75515616e-01 -2.51600593e-01
6.57148421e-01 7.95025885e-01 -4.16013241e-01 -1.05474256e-01
-2.19216332e-01 3.01095247e-01 2.35775243e-02 -7.97523260e-01
1.11186802e+00 -3.16740274e-01 -1.49982423e-01 1.03484011e+00
1.57704449e+00 5.04148602e-01 -1.66243565e+00 -1.97685078e-01
-1.09806985e-01 -5.59110820e-01 1.72450200e-01 -5.28026447e-02
-6.91496730e-01 9.74788725e-01 3.69413555e-01 1.92183182e-01
2.44130924e-01 -1.54404223e-01 7.74036884e-01 8.21446180e-01
5.34847565e-02 -1.42444897e+00 -3.65895659e-01 6.85192049e-01
6.78132713e-01 -1.39803863e+00 3.83911669e-01 -5.08129299e-02
-1.42273515e-01 9.05914187e-01 -1.73202708e-01 -5.82390547e-01
1.01110375e+00 1.86970949e-01 -3.50483149e-01 -1.36536226e-01
-6.12020671e-01 8.00060481e-02 1.90979660e-01 4.61775362e-01
2.61000156e-01 4.00853939e-02 -3.46505493e-01 2.18130916e-01
-4.70830739e-01 -1.23277061e-01 3.12310517e-01 4.50837463e-01
-4.57859159e-01 -5.55086851e-01 1.27596511e-02 3.70039493e-01
-8.79828632e-01 -3.23223591e-01 3.08086257e-02 6.54188216e-01
-7.10623503e-01 9.39192057e-01 -7.91886970e-02 2.75051117e-01
-3.78462672e-01 2.83599738e-02 9.09646273e-01 -2.90801227e-01
-2.15346828e-01 -6.73234016e-02 -1.49803564e-01 -6.66636705e-01
-3.36455792e-01 -1.00430667e+00 -9.70894098e-01 -1.01323402e+00
-4.55660462e-01 3.35555285e-01 1.03186965e+00 1.06859493e+00
1.01509765e-01 -1.72303289e-01 5.57868838e-01 -7.11660564e-01
-1.53876078e+00 -9.83357906e-01 -8.16433311e-01 4.18666959e-01
5.76033354e-01 -3.88128310e-01 -1.00142682e+00 -4.53527607e-02] | [6.905270099639893, 4.034294128417969] |
25b00e98-e8e3-44d9-9fe7-8cd96fb57c9b | trans-svnet-accurate-phase-recognition-from | 2103.09712 | null | https://arxiv.org/abs/2103.09712v2 | https://arxiv.org/pdf/2103.09712v2.pdf | Trans-SVNet: Accurate Phase Recognition from Surgical Videos via Hybrid Embedding Aggregation Transformer | Real-time surgical phase recognition is a fundamental task in modern operating rooms. Previous works tackle this task relying on architectures arranged in spatio-temporal order, however, the supportive benefits of intermediate spatial features are not considered. In this paper, we introduce, for the first time in surgical workflow analysis, Transformer to reconsider the ignored complementary effects of spatial and temporal features for accurate surgical phase recognition. Our hybrid embedding aggregation Transformer fuses cleverly designed spatial and temporal embeddings by allowing for active queries based on spatial information from temporal embedding sequences. More importantly, our framework processes the hybrid embeddings in parallel to achieve a high inference speed. Our method is thoroughly validated on two large surgical video datasets, i.e., Cholec80 and M2CAI16 Challenge datasets, and outperforms the state-of-the-art approaches at a processing speed of 91 fps. | ['Pheng-Ann Heng', 'Qi Dou', 'Yonghao Long', 'Yueming Jin', 'Xiaojie Gao'] | 2021-03-17 | null | null | null | null | ['surgical-phase-recognition'] | ['computer-vision'] | [ 6.35773391e-02 -2.52565518e-02 -4.63501066e-01 -8.59290138e-02
-8.60435307e-01 -5.63239276e-01 4.10609752e-01 6.38361633e-01
-9.65670884e-01 1.65228397e-01 3.19972515e-01 -5.29665172e-01
-6.99943125e-01 -3.74807894e-01 -3.19511533e-01 -8.64972532e-01
-5.23491085e-01 2.91540205e-01 3.51998568e-01 -7.61996862e-03
1.93517908e-01 6.65689766e-01 -1.35862577e+00 3.04168493e-01
4.96859103e-01 1.03123283e+00 2.05869898e-01 8.53794038e-01
5.66281602e-02 7.93226063e-01 -2.22037196e-01 -1.24008201e-01
1.55277446e-01 1.53571114e-01 -7.39551961e-01 -2.47525781e-01
7.65702054e-02 -3.33426625e-01 -7.44238675e-01 4.76862252e-01
7.72085369e-01 -2.40465850e-02 2.74962664e-01 -7.48570263e-01
-1.43731192e-01 3.56372774e-01 -1.65249646e-01 6.82674348e-01
2.28015289e-01 3.51709485e-01 7.62445211e-01 -9.46831703e-01
6.89149737e-01 5.49164176e-01 6.42672837e-01 2.97991782e-01
-9.62015152e-01 -3.20426553e-01 1.89547002e-01 3.88145715e-01
-1.34415209e+00 -2.25950196e-01 6.27335310e-01 -4.37086165e-01
9.98479366e-01 4.45000470e-01 1.02688336e+00 1.10945773e+00
8.77113521e-01 9.88339841e-01 9.29832399e-01 -1.32584274e-01
6.97188675e-02 -1.91884443e-01 8.82684365e-02 9.01211381e-01
-6.80356398e-02 4.66881752e-01 -7.39020705e-01 -1.35229915e-01
8.30270529e-01 5.47792852e-01 -4.02169466e-01 -6.85007930e-01
-1.92465389e+00 6.40740216e-01 7.00436294e-01 4.84917045e-01
-5.63492477e-01 3.60650063e-01 7.15793908e-01 3.22584137e-02
9.03640538e-02 5.12698770e-01 -2.59542435e-01 -6.42141402e-01
-9.78001654e-01 -2.00063154e-01 6.83026552e-01 8.79688025e-01
3.10528666e-01 -4.07532215e-01 -5.23618758e-01 2.95254380e-01
9.98158306e-02 -9.70790610e-02 6.77498400e-01 -5.93765676e-01
1.85018525e-01 6.08579874e-01 -1.21489733e-01 -8.02788675e-01
-9.45168614e-01 -5.71754992e-01 -9.61932480e-01 1.69718154e-02
2.13500187e-01 2.33891368e-01 -8.41276765e-01 1.12146580e+00
4.08773661e-01 5.36650479e-01 -6.53672591e-02 9.41221118e-01
9.74810183e-01 2.67456733e-02 1.79952756e-02 -2.34773830e-01
1.59258068e+00 -1.18118966e+00 -9.35899258e-01 6.58016726e-02
1.08358741e+00 -6.93883419e-01 7.78055131e-01 4.67373997e-01
-9.23524857e-01 -4.17996734e-01 -8.62310886e-01 -1.69908702e-01
-3.73884618e-01 4.55082893e-01 1.03054965e+00 3.09550136e-01
-1.11408424e+00 6.46109104e-01 -1.53410077e+00 -8.80095139e-02
3.27919662e-01 7.76045024e-01 -6.76998556e-01 -1.15880661e-01
-8.98700237e-01 7.64726996e-01 1.85088277e-01 4.67289716e-01
-9.04126227e-01 -1.12070644e+00 -1.00173616e+00 -1.71894550e-01
5.88006616e-01 -7.38242686e-01 1.31045234e+00 -1.71513677e-01
-1.42038691e+00 9.09079790e-01 -1.00937814e-01 -5.32825232e-01
6.59154654e-01 -3.01219195e-01 -3.32096696e-01 5.19973636e-01
-3.73206228e-01 4.43791747e-01 4.34900612e-01 -7.44355321e-01
-4.37997848e-01 -5.64851880e-01 2.91879028e-01 2.55421251e-01
-6.46018922e-01 -2.98406303e-01 -7.33847320e-01 -6.78681672e-01
1.18345395e-01 -1.16850662e+00 -7.19646454e-01 7.06462801e-01
-1.74838379e-01 -7.88057502e-03 4.08168137e-01 -4.31828380e-01
1.56004965e+00 -2.32404518e+00 2.82854021e-01 -3.27486321e-02
6.09873831e-01 1.80406168e-01 2.04446107e-01 4.91338789e-01
-2.29751319e-01 -2.39956081e-01 6.11887313e-02 -5.33254981e-01
-1.81566805e-01 4.22442287e-01 1.01065980e-02 8.24244380e-01
4.50164787e-02 1.03859174e+00 -1.25528431e+00 -8.16136479e-01
5.75654805e-01 6.03195250e-01 -7.13640392e-01 2.57664949e-01
5.39882421e-01 4.22964483e-01 -3.36682588e-01 7.05847144e-01
2.42478475e-01 -2.83558160e-01 2.47124881e-01 -5.13040066e-01
-3.17114562e-01 2.75662005e-01 -6.62394464e-01 2.52408004e+00
-7.45650351e-01 4.82679337e-01 -6.26924485e-02 -7.75832415e-01
4.36855286e-01 3.73896450e-01 1.29521406e+00 -6.35720968e-01
1.73159853e-01 1.67661980e-01 9.23408121e-02 -5.28741062e-01
5.31637847e-01 9.16529670e-02 -1.88873902e-01 7.69077707e-03
1.76841751e-01 2.41817385e-01 1.38689280e-01 1.12250961e-01
1.57012796e+00 -4.96803373e-02 5.26667953e-01 -3.13597172e-01
2.22173333e-01 1.81328937e-01 3.35741907e-01 7.49916852e-01
-5.18825114e-01 5.79537272e-01 5.44237077e-01 -7.37190425e-01
-4.37735885e-01 -1.33226132e+00 -3.47943187e-01 7.58423030e-01
3.94906193e-01 -7.33341336e-01 -9.86068770e-02 -1.02622700e+00
9.26883444e-02 9.46879759e-02 -1.19062138e+00 -3.38969290e-01
-7.75063515e-01 -6.35966539e-01 2.79060274e-01 8.13163936e-01
-3.22814286e-01 -7.57525921e-01 -1.09956753e+00 3.89715254e-01
8.89433548e-02 -1.21137345e+00 -4.74133015e-01 3.65528136e-01
-1.09715652e+00 -1.14085376e+00 -6.80412412e-01 -6.21326327e-01
5.72077811e-01 3.47601712e-01 9.88095224e-01 4.47160751e-02
-1.04684794e+00 4.89703774e-01 -4.54376251e-01 -2.40726694e-01
7.41991699e-02 1.58658400e-01 -5.91668300e-02 -2.23486543e-01
1.19444795e-01 -3.79215896e-01 -1.11391759e+00 1.98215634e-01
-1.02524817e+00 2.47706577e-01 1.01580203e+00 1.14496362e+00
7.89265931e-01 -3.84116590e-01 -2.14585036e-01 -8.24893475e-01
1.67988062e-01 -4.50919628e-01 -4.11614507e-01 1.70579389e-01
-6.54446602e-01 8.09622332e-02 4.89498764e-01 -5.05814850e-01
-3.66318762e-01 2.14500159e-01 -2.61278927e-01 -8.45202744e-01
7.25124627e-02 6.16906881e-01 5.42968690e-01 -1.27061993e-01
4.83500212e-01 3.87548566e-01 3.55962515e-01 -3.36291522e-01
1.35841057e-01 1.92977205e-01 3.79178405e-01 -2.12671131e-01
5.99463463e-01 7.83353090e-01 7.53029659e-02 -5.95958591e-01
-7.54384339e-01 -1.02379608e+00 -7.15192676e-01 -2.46671617e-01
7.17875361e-01 -8.82360339e-01 -1.06914687e+00 7.69942999e-02
-9.47396636e-01 -1.76627040e-01 -5.17056882e-01 9.69819963e-01
-8.05573583e-01 3.33385855e-01 -7.95358419e-01 -2.91476578e-01
-3.15436184e-01 -1.58306408e+00 1.35766566e+00 -1.02835305e-01
-1.15777463e-01 -1.02724564e+00 3.96732330e-01 -2.94494033e-02
3.97444814e-01 3.21206361e-01 4.49741393e-01 -5.37564039e-01
-9.11863923e-01 -3.69924098e-01 -7.23278970e-02 -2.87016779e-01
2.32883826e-01 -7.40404055e-02 -7.29670823e-01 -5.09660244e-01
-2.20327884e-01 1.30633667e-01 9.27253783e-01 4.17630345e-01
1.57929075e+00 -8.99652317e-02 -6.24557436e-01 1.04656851e+00
1.38620508e+00 -1.03482828e-01 5.81802905e-01 4.06810284e-01
4.35828120e-01 3.09864551e-01 9.48571801e-01 7.14535832e-01
3.26460272e-01 7.83181727e-01 7.62421370e-01 -3.34854156e-01
3.69338170e-02 1.64410278e-01 3.64427306e-02 1.05818224e+00
1.20091783e-02 3.98889557e-02 -1.21535575e+00 7.94991612e-01
-1.91318607e+00 -8.07515621e-01 2.06558615e-01 2.11706972e+00
7.93693304e-01 6.42490312e-02 -3.24862003e-01 2.05052033e-01
-1.00339847e-02 5.21233976e-01 -2.38164753e-01 -1.26612216e-01
4.21749860e-01 4.29344833e-01 8.49841595e-01 2.81144917e-01
-1.45797241e+00 4.82439667e-01 6.01130199e+00 6.31407201e-01
-1.16576970e+00 3.09771478e-01 1.85877129e-01 -6.93136811e-01
-6.53341264e-02 -1.21042579e-01 -5.61392665e-01 2.62064606e-01
8.73900294e-01 -1.56463936e-01 -3.63894589e-02 8.91848862e-01
5.38428724e-02 7.77275488e-02 -1.19049430e+00 1.25475812e+00
1.75197333e-01 -1.77050555e+00 -2.75170475e-01 2.84339905e-01
1.78735644e-01 8.57630894e-02 2.89886445e-01 3.46418977e-01
2.59008743e-02 -1.14786923e+00 3.05369794e-01 7.35739946e-01
7.62999773e-01 -4.95083511e-01 9.90900040e-01 6.28360808e-02
-1.24790096e+00 -2.25724369e-01 8.19395781e-02 3.93958390e-01
1.92446694e-01 5.37167251e-01 -9.46044326e-01 9.78708744e-01
7.46638060e-01 1.05235112e+00 -5.70216119e-01 1.34403312e+00
1.76163703e-01 2.18117550e-01 -3.93551916e-01 1.50903329e-01
4.58767474e-01 3.33731413e-01 3.02917749e-01 1.30268610e+00
3.61871481e-01 7.85674751e-02 1.15883030e-01 3.03160083e-02
3.50228250e-01 5.61065786e-03 -6.61150634e-01 -1.46812484e-01
-2.00674571e-02 1.46305370e+00 -5.45931756e-01 -1.06187031e-01
-3.43498111e-01 1.02044833e+00 3.08986008e-01 5.14614722e-03
-1.10354924e+00 -1.85596988e-01 9.23409998e-01 1.43125147e-01
1.30317882e-01 -6.15878701e-01 -1.24330677e-01 -1.10515642e+00
1.60989001e-01 -4.82525617e-01 6.47518635e-01 -2.44161278e-01
-7.57016242e-01 7.25582182e-01 -1.98453486e-01 -1.89524949e+00
-2.69218504e-01 -9.69908118e-01 -3.03379446e-01 4.55490738e-01
-1.84166563e+00 -1.07876432e+00 -7.35053658e-01 7.13096201e-01
4.13053572e-01 2.23847166e-01 1.18616724e+00 4.83106852e-01
-5.08460045e-01 7.47454524e-01 4.10931706e-02 2.48173639e-01
8.51485610e-01 -1.15680385e+00 -2.85145231e-02 4.86858815e-01
6.98774680e-02 8.28269243e-01 4.37936395e-01 -5.25147542e-02
-1.98724556e+00 -9.37528312e-01 6.73286438e-01 -5.45584083e-01
7.66215503e-01 -2.59694397e-01 -5.38588583e-01 6.08527720e-01
2.95521989e-02 6.41964316e-01 1.20255446e+00 9.36137438e-02
-6.55940399e-02 -3.20623159e-01 -6.82299495e-01 5.86538374e-01
1.17638600e+00 -3.86044383e-01 -4.57955629e-01 4.77412909e-01
7.07988977e-01 -9.97979283e-01 -1.42747462e+00 7.56407380e-01
6.63245499e-01 -1.02298224e+00 1.33149898e+00 -5.26931524e-01
4.16119248e-01 -1.39785245e-01 1.21569283e-01 -9.89463568e-01
-2.91514218e-01 -5.13327122e-01 -3.20961297e-01 2.06823915e-01
1.68525696e-01 -4.88020003e-01 8.16140115e-01 1.88812017e-01
-6.54475510e-01 -1.48129904e+00 -1.37898028e+00 -6.16843045e-01
-3.18330169e-01 -3.90115589e-01 1.06405452e-01 7.75666177e-01
1.41497821e-01 -4.51927543e-01 -5.34987897e-02 1.92360103e-01
2.98983425e-01 2.56906807e-01 5.34280598e-01 -8.73192370e-01
-3.23958606e-01 -5.48343062e-01 -8.84464920e-01 -1.03408992e+00
-2.60092080e-01 -8.42860937e-01 -7.82647356e-02 -1.55324018e+00
9.37987566e-02 -4.66278642e-01 -8.72486532e-01 3.75324368e-01
-5.72845899e-02 2.15534389e-01 2.26118341e-02 2.45486051e-01
-7.85646975e-01 6.91747189e-01 1.31928825e+00 -1.00441553e-01
-2.66877618e-02 -2.04536796e-01 -2.96877682e-01 4.17862564e-01
4.36355323e-01 -4.66298074e-01 -3.35446745e-01 -5.82154751e-01
3.07085738e-02 3.94089043e-01 4.45757896e-01 -1.16010404e+00
7.87955105e-01 -5.85561059e-02 2.38363355e-01 -7.66316712e-01
7.67738342e-01 -1.01055861e+00 9.45498794e-02 7.00743854e-01
-2.05026567e-01 3.80719841e-01 4.05013263e-01 7.02947855e-01
-5.86914659e-01 3.88014078e-01 5.63099980e-01 -6.32202812e-03
-7.78516591e-01 8.72690558e-01 -3.20891976e-01 -1.97595686e-01
1.42534721e+00 -3.59929919e-01 -1.41974390e-01 2.79560328e-01
-1.05608022e+00 2.03548387e-01 4.25571710e-01 6.06144845e-01
1.05273128e+00 -1.32581282e+00 -4.59701896e-01 2.47988597e-01
6.17753446e-01 8.47542882e-02 8.57860684e-01 1.92145085e+00
-9.39626396e-01 6.39717042e-01 -1.16488272e-02 -1.06638467e+00
-1.22662544e+00 9.17523801e-01 2.34272435e-01 -9.16105092e-01
-9.89272833e-01 6.82149768e-01 3.36723894e-01 -3.18625480e-01
3.73113573e-01 -4.89859045e-01 -2.49866024e-01 1.13839075e-01
5.61449051e-01 -2.23726928e-01 3.94423813e-01 -1.86881453e-01
-7.99236059e-01 5.66123009e-01 -2.27904394e-01 1.79950520e-01
1.25777817e+00 2.20481589e-01 1.06682494e-01 4.59389687e-01
1.51085579e+00 -1.28866330e-01 -9.64407802e-01 -4.03751343e-01
-1.50864035e-01 -6.80564702e-01 2.53417492e-01 -4.39050615e-01
-1.07207239e+00 9.88145828e-01 9.07659411e-01 -2.85718054e-01
1.20076776e+00 -1.55978933e-01 7.71683991e-01 2.29794905e-01
3.79716635e-01 -7.01797247e-01 2.51052737e-01 2.89269239e-01
7.29699850e-01 -1.30460131e+00 3.01387142e-02 -3.20489466e-01
-6.00767910e-01 1.30357289e+00 4.78281409e-01 -1.26270950e-01
9.12109196e-01 4.93607700e-01 1.65486008e-01 -2.91829914e-01
-9.09066737e-01 -2.54757404e-01 3.47521037e-01 1.38202652e-01
4.83587950e-01 1.47187747e-02 -3.29384953e-01 4.66343492e-01
8.78269076e-02 1.57378212e-01 1.65601656e-01 1.26634264e+00
1.62096575e-01 -8.52507472e-01 1.80679753e-01 4.62343514e-01
-5.37189960e-01 -2.28959665e-01 3.85741085e-01 9.85813081e-01
-1.63359672e-01 4.27238107e-01 9.88627896e-02 -3.93410265e-01
5.42224467e-01 -3.33911359e-01 3.57276708e-01 -5.35958230e-01
-8.73102009e-01 -1.44258484e-01 -1.90165043e-01 -1.36373484e+00
-2.52280235e-01 -4.92398322e-01 -1.16005158e+00 2.09399257e-02
3.19612399e-02 -3.20222452e-02 6.71542704e-01 4.02023852e-01
4.95155364e-01 1.08361185e+00 5.24720848e-01 -8.94696176e-01
-5.97562611e-01 -5.31308293e-01 -3.01494062e-01 2.92096704e-01
6.21512651e-01 -8.69620562e-01 -3.53248656e-01 -3.89916301e-01] | [14.096832275390625, -3.3288676738739014] |
85747b06-2180-4359-a975-117120f06aa4 | matching-text-with-deep-mutual-information | 2003.11521 | null | https://arxiv.org/abs/2003.11521v1 | https://arxiv.org/pdf/2003.11521v1.pdf | Matching Text with Deep Mutual Information Estimation | Text matching is a core natural language processing research problem. How to retain sufficient information on both content and structure information is one important challenge. In this paper, we present a neural approach for general-purpose text matching with deep mutual information estimation incorporated. Our approach, Text matching with Deep Info Max (TIM), is integrated with a procedure of unsupervised learning of representations by maximizing the mutual information between text matching neural network's input and output. We use both global and local mutual information to learn text representations. We evaluate our text matching approach on several tasks including natural language inference, paraphrase identification, and answer selection. Compared to the state-of-the-art approaches, the experiments show that our method integrated with mutual information estimation learns better text representation and achieves better experimental results of text matching tasks without exploiting pretraining on external data. | ['Zhi Yu', 'Jiajun Bu', 'Zhou Yu', 'Xixi Zhou', 'Chengwei Yao', 'Keyue Shi', 'Chengxi Li'] | 2020-03-09 | null | null | null | null | ['mutual-information-estimation', 'paraphrase-identification', 'answer-selection'] | ['methodology', 'natural-language-processing', 'natural-language-processing'] | [ 5.49258173e-01 3.18113305e-02 -2.51101077e-01 -6.66241109e-01
-7.86392868e-01 -1.30052269e-01 6.42468095e-01 6.15124524e-01
-6.40680194e-01 2.25782916e-01 5.47717333e-01 -2.68914670e-01
-1.76460490e-01 -9.38095748e-01 -5.28848112e-01 1.98757481e-02
6.63370311e-01 6.48034513e-01 -4.73193265e-02 -1.79848358e-01
6.45580053e-01 -1.25735939e-01 -1.72237694e+00 1.00788951e+00
1.11577332e+00 9.62533236e-01 3.73027116e-01 6.18273675e-01
-9.02134538e-01 1.25589514e+00 -3.56014937e-01 -6.43358529e-01
6.80387020e-02 -5.61989188e-01 -1.38102078e+00 -5.48046231e-01
5.94185054e-01 -1.72315896e-01 -7.62517512e-01 1.12018442e+00
5.79605043e-01 4.69859660e-01 9.15289581e-01 -9.34630990e-01
-7.79752314e-01 9.29075956e-01 -3.01146120e-01 2.72532731e-01
7.72857547e-01 -1.40236661e-01 1.62419987e+00 -9.09564316e-01
3.29363257e-01 1.27942574e+00 7.43127882e-01 4.59945768e-01
-1.15526557e+00 -7.62096107e-01 -2.03056917e-01 4.41272408e-01
-9.60491598e-01 -5.34301579e-01 7.07099259e-01 -3.39438498e-01
1.49607170e+00 2.25960270e-01 -1.27833292e-01 7.93402433e-01
3.60312700e-01 1.28671587e+00 5.14116585e-01 -7.19768167e-01
-2.04426527e-01 2.33221993e-01 7.57935047e-01 8.07636023e-01
-2.38182507e-02 6.71591088e-02 -5.39178491e-01 -4.08364207e-01
1.76041737e-01 3.18935990e-01 -9.42047909e-02 -9.13211405e-02
-1.02212417e+00 8.79424036e-01 5.24953246e-01 6.86242521e-01
-3.60237062e-02 4.08953130e-02 5.46771944e-01 9.44076657e-01
4.13036019e-01 6.30903423e-01 -4.13303882e-01 9.72683653e-02
-1.02566957e+00 3.11033279e-02 1.08393753e+00 1.03527069e+00
9.75104928e-01 -4.26517338e-01 -5.29884696e-01 1.27340186e+00
4.13767457e-01 4.50888425e-01 1.30288625e+00 -5.77461421e-01
1.04051125e+00 1.25585914e+00 -5.02958179e-01 -1.26565862e+00
-3.58549535e-01 -1.25738934e-01 -1.16507959e+00 -3.08459073e-01
2.89355695e-01 -1.58151627e-01 -4.69566852e-01 1.67245162e+00
-2.81563789e-01 -2.43199423e-01 1.41953051e-01 5.38293183e-01
1.30438256e+00 5.37357450e-01 -1.01463653e-01 1.45684674e-01
1.16772604e+00 -1.04316878e+00 -8.02042723e-01 -5.39367795e-01
1.09446013e+00 -1.07662416e+00 7.67011642e-01 -3.65949720e-01
-1.15763390e+00 -7.43792832e-01 -8.96343470e-01 -3.61435831e-01
-6.18684590e-01 8.45921934e-02 3.95183027e-01 4.90670323e-01
-8.06470811e-01 1.08855712e+00 -1.81899250e-01 -4.91603792e-01
1.38653994e-01 4.26005155e-01 -4.28673416e-01 -7.21640438e-02
-1.58306062e+00 1.04485345e+00 5.05953431e-01 -3.30270708e-01
2.17751116e-01 -5.21221578e-01 -1.13268840e+00 3.85095745e-01
-7.92112350e-02 -1.22308290e+00 1.33277702e+00 -1.22277117e+00
-1.37189138e+00 1.36788249e+00 -4.97868389e-01 -5.65449893e-01
3.26351255e-01 -3.32896948e-01 -2.28749719e-02 -1.06126979e-01
3.38129327e-02 5.86192667e-01 7.42170930e-01 -3.64520788e-01
-2.10008591e-01 -5.09280443e-01 -4.57580149e-01 3.61484468e-01
-5.75887442e-01 3.59107643e-01 -2.45328560e-01 -6.20875835e-01
3.01680982e-01 -3.90936285e-01 -9.45775285e-02 7.80283706e-04
-1.73178285e-01 -6.61175311e-01 3.86314452e-01 -7.85595536e-01
1.36681497e+00 -1.63571382e+00 1.84066594e-01 9.43390932e-03
2.06006482e-01 1.11991569e-01 -2.36476555e-01 7.04226017e-01
-2.64504552e-01 -1.00688830e-01 -3.26672316e-01 -5.94593465e-01
2.68893152e-01 -1.48967072e-01 -3.80958647e-01 2.14838147e-01
1.32633122e-02 1.31687284e+00 -6.51392162e-01 -7.77789831e-01
4.83094268e-02 -9.36534628e-02 -5.25713146e-01 6.34379327e-01
-2.25310162e-01 -3.30535024e-01 -3.27314794e-01 1.68255910e-01
5.72823226e-01 -5.62863946e-01 2.43627250e-01 -1.29181266e-01
3.66181016e-01 7.38299310e-01 -9.89486873e-01 2.09421849e+00
-4.32405233e-01 9.88348901e-01 -2.83716112e-01 -1.21366262e+00
1.11054957e+00 1.59773007e-01 3.83453459e-01 -1.12356222e+00
3.73085737e-01 6.78287819e-02 -3.13317180e-01 -7.64667988e-01
6.87889993e-01 1.54858932e-01 -5.83691411e-02 1.33333063e+00
1.68360546e-01 1.23180993e-01 -8.46584290e-02 4.97835189e-01
1.21323824e+00 -1.89995348e-01 4.30953205e-01 -3.42982560e-01
6.68989360e-01 -1.90419406e-01 8.38758722e-02 1.04512918e+00
-8.16063061e-02 4.15598601e-01 2.10456848e-01 -2.35569611e-01
-9.51011777e-01 -4.88587409e-01 2.38233898e-02 1.50935328e+00
1.38322338e-01 -4.37259078e-01 -5.95993757e-01 -7.11780965e-01
3.86441529e-01 6.03406727e-01 -5.99054754e-01 -5.51613867e-01
-6.78580105e-01 -4.43519503e-01 7.56202996e-01 5.75508237e-01
6.51536703e-01 -1.22424889e+00 -1.02227315e-01 1.50041744e-01
-4.79150027e-01 -6.87455058e-01 -8.38555217e-01 9.33455378e-02
-1.07346833e+00 -1.04608178e+00 -5.23218691e-01 -1.18255377e+00
5.46440125e-01 5.29536545e-01 1.43597686e+00 3.17232996e-01
-4.68457758e-01 3.28233868e-01 -3.15276325e-01 1.83669999e-01
-6.03839397e-01 4.03412580e-01 -4.66975212e-01 -1.40654802e-01
9.12671983e-01 -5.24064124e-01 -5.01445949e-01 3.78237277e-01
-9.30966198e-01 2.39186585e-01 6.95166707e-01 1.06493866e+00
-8.84503871e-03 -3.74931663e-01 5.52971840e-01 -1.13974166e+00
1.00423813e+00 -6.13906384e-01 -2.32089460e-01 5.01328707e-01
-8.55778933e-01 7.64365673e-01 3.77601475e-01 -2.02455729e-01
-1.07447934e+00 -3.29544991e-02 -8.49130899e-02 -1.65707886e-01
-8.30947459e-02 5.48901975e-01 5.08994609e-02 2.74580210e-01
8.56775701e-01 3.88234049e-01 1.26688302e-01 -7.55518496e-01
3.25595260e-01 1.15541971e+00 3.93311858e-01 -5.24286449e-01
3.64436060e-01 1.58427916e-02 -5.31800687e-01 -2.18074530e-01
-9.83738244e-01 -1.09766769e+00 -7.47306705e-01 1.53718591e-01
4.61727917e-01 -6.65558100e-01 -8.50755930e-01 4.12309349e-01
-1.41417062e+00 -1.67607553e-02 1.29123256e-01 2.67088771e-01
-5.46606183e-01 7.86415398e-01 -7.90177107e-01 -5.08907557e-01
-1.08493960e+00 -6.96148872e-01 8.81771326e-01 3.45070660e-02
-4.58308488e-01 -1.12191617e+00 4.00495291e-01 6.56047344e-01
5.58458328e-01 -6.21089280e-01 1.05575418e+00 -1.36074698e+00
-2.64160961e-01 -3.40956330e-01 -8.02381635e-01 3.54158469e-02
4.22163643e-02 -5.10581434e-01 -1.04851270e+00 -2.40452170e-01
1.00561053e-01 -4.83935505e-01 1.36160874e+00 -7.40588829e-02
1.13127112e+00 -5.10519326e-01 -2.24637583e-01 4.98779744e-01
1.23016036e+00 -4.15885389e-01 6.07952118e-01 3.12609047e-01
5.04480362e-01 9.44064081e-01 2.69408375e-01 3.45962912e-01
3.22982728e-01 4.49208915e-01 7.66677111e-02 2.81603605e-01
1.08283600e-02 -4.11854744e-01 2.01722920e-01 1.22487509e+00
7.80710280e-01 -2.40763009e-01 -7.67443955e-01 3.63659412e-01
-2.37774444e+00 -1.21782386e+00 -4.90244851e-02 2.09880781e+00
1.13856316e+00 -1.00836933e-01 -2.30165839e-01 7.12552015e-03
1.10681963e+00 5.41939065e-02 -7.78178871e-01 -3.26414704e-01
-1.26797527e-01 2.44654447e-01 2.36104295e-01 3.94610316e-01
-1.12059939e+00 8.93905282e-01 6.32199478e+00 8.23611796e-01
-4.01026368e-01 -1.32601634e-01 2.60747671e-01 3.37247491e-01
-4.29586887e-01 -1.48130925e-02 -6.49696827e-01 4.35520023e-01
9.61032271e-01 -4.59170312e-01 2.48658478e-01 6.79864407e-01
-5.32562554e-01 1.98969141e-01 -1.47514832e+00 1.15749156e+00
3.70121956e-01 -1.40059423e+00 2.47468784e-01 -4.40135956e-01
6.57379866e-01 1.23111255e-01 -3.19637984e-01 6.46372914e-01
4.79673892e-01 -1.09928954e+00 -4.91738133e-02 9.12192464e-01
2.55537570e-01 -5.16854227e-01 8.72502506e-01 6.91024661e-01
-1.11624801e+00 2.17565000e-02 -7.43183017e-01 -1.10420763e-01
-4.33630109e-01 5.23365140e-01 -5.73814988e-01 5.24358928e-01
2.32993588e-01 1.03137743e+00 -6.47678852e-01 1.03318989e+00
-3.62970158e-02 2.14987248e-01 -1.19088992e-01 -5.72794497e-01
-6.90077916e-02 1.05667589e-02 2.85281956e-01 1.59694719e+00
-2.17796594e-01 -3.24907333e-01 7.78992027e-02 1.21322727e+00
-6.56869233e-01 6.81991100e-01 -7.50342488e-01 -4.23752740e-02
3.64646196e-01 9.42193091e-01 4.17734757e-02 -6.08276963e-01
-3.41695368e-01 1.23117518e+00 7.90536165e-01 3.16758491e-02
-2.01548278e-01 -8.77514482e-01 2.99435019e-01 -4.29358721e-01
-1.99674860e-01 3.26195359e-01 -5.43317556e-01 -1.45395207e+00
1.55286834e-01 -9.68464196e-01 1.02595675e+00 -7.48142183e-01
-1.95641637e+00 3.85650188e-01 -2.50611395e-01 -1.01584864e+00
-8.17797482e-01 -6.27370059e-01 -8.37623298e-01 1.09378898e+00
-1.51924944e+00 -8.46649706e-01 -3.07199448e-01 6.44302845e-01
8.02365482e-01 -6.21317923e-01 9.65714931e-01 3.73829484e-01
-5.90315223e-01 1.13377547e+00 4.68349129e-01 5.56347370e-01
9.36367512e-01 -1.11083245e+00 7.13978708e-01 2.83628434e-01
2.24426717e-01 7.49925494e-01 3.25154752e-01 -6.52557671e-01
-1.45560098e+00 -7.46983051e-01 1.55016387e+00 -3.84300411e-01
8.62517595e-01 -4.25843894e-02 -1.21688807e+00 5.79781711e-01
5.92807353e-01 -6.86914325e-01 7.90724993e-01 2.95162141e-01
-7.86428630e-01 2.06925496e-01 -9.93933260e-01 5.38869143e-01
9.42408860e-01 -1.16081548e+00 -1.19896305e+00 4.80598599e-01
8.70282173e-01 -1.86362918e-02 -7.25405276e-01 3.68201584e-01
6.48064017e-01 -8.47996950e-01 7.59509206e-01 -1.04866385e+00
6.12074137e-01 4.27731514e-01 -1.81851029e-01 -9.20990288e-01
-2.99891263e-01 -5.74230850e-01 1.61846742e-01 1.21146417e+00
5.56444287e-01 -6.96660459e-01 9.00969565e-01 8.69843781e-01
1.79933161e-01 -4.22039390e-01 -6.32119596e-01 -2.97303110e-01
4.25230712e-01 -2.51004964e-01 5.03854573e-01 1.40803576e+00
8.73386502e-01 6.70105577e-01 -1.75542668e-01 -4.62235570e-01
5.76739311e-01 6.85050428e-01 6.56675994e-01 -1.43269086e+00
-4.22056764e-01 -9.54325616e-01 -3.27138603e-01 -1.38300920e+00
8.88907909e-01 -1.62983847e+00 -8.65207911e-02 -1.46431172e+00
9.34491336e-01 1.01778835e-01 -3.70051444e-01 2.07422003e-01
-4.18155402e-01 -3.93290609e-01 1.73717346e-02 5.44694424e-01
-6.39775455e-01 6.11910641e-01 7.35042274e-01 -6.16880238e-01
6.33859113e-02 1.69876948e-01 -4.94189084e-01 4.28504944e-01
9.90204573e-01 -5.93117118e-01 -1.06355518e-01 -7.31154144e-01
5.06274045e-01 1.93646684e-01 -3.52890715e-02 -6.28653347e-01
1.10286057e+00 1.46356508e-01 2.97453403e-01 -8.34383368e-01
1.03383027e-01 -6.32940114e-01 -3.57073665e-01 3.33139658e-01
-1.41400349e+00 2.73807347e-01 9.18651894e-02 4.60302293e-01
-2.57072955e-01 -9.32049036e-01 7.07723439e-01 -3.54760557e-01
-2.99292922e-01 3.90705355e-02 -3.79842341e-01 2.46882856e-01
1.75595969e-01 6.99083358e-02 -5.42426586e-01 -4.53341037e-01
-2.99839646e-01 4.06331241e-01 2.05644086e-01 5.99724650e-01
8.91618252e-01 -1.35328424e+00 -9.93716955e-01 3.72515291e-01
3.77460897e-01 -7.38227844e-01 5.40527739e-02 2.85029501e-01
-4.50214967e-02 6.95254505e-01 -9.77559462e-02 -2.49078721e-01
-1.33785427e+00 3.98589373e-01 5.93239605e-01 -8.47503126e-01
-4.39432204e-01 6.03215098e-01 1.17663555e-01 -1.12831211e+00
5.00482023e-01 5.60133085e-02 -4.69841331e-01 -8.57599732e-03
6.46699548e-01 2.21364126e-01 2.28327140e-01 -3.89950037e-01
-4.29378599e-02 4.67007458e-01 -5.68257868e-01 8.34014192e-02
8.45879078e-01 -1.56924501e-01 -4.49694276e-01 2.51223773e-01
1.89414883e+00 -5.53860307e-01 -1.30302951e-01 -9.67527926e-01
5.07007301e-01 -3.77849072e-01 1.11098804e-01 -4.50181127e-01
-8.22365105e-01 1.10565364e+00 5.31066597e-01 -1.14239253e-01
5.78422129e-01 -1.95708126e-01 1.08271265e+00 1.24109852e+00
-2.86638498e-01 -1.11674726e+00 2.78274506e-01 9.54133809e-01
6.61422431e-01 -1.49698257e+00 -9.82281864e-02 3.51798944e-02
-3.27938467e-01 1.27386260e+00 7.50657737e-01 -8.21943283e-02
6.83293700e-01 -4.76607196e-02 -1.79345936e-01 -2.86191493e-01
-1.23345566e+00 -4.76505876e-01 9.09322560e-01 1.11492045e-01
9.41071272e-01 -3.92431736e-01 -1.15842573e-01 3.52105945e-01
-4.14242558e-02 -3.28045398e-01 -2.79887438e-01 9.55610216e-01
-9.51992452e-01 -1.26088905e+00 1.06655003e-03 8.64067793e-01
-3.99208963e-01 -5.78674853e-01 -9.81071949e-01 1.26791313e-01
-7.39660084e-01 9.92020547e-01 2.57393479e-01 -4.79809880e-01
1.89474896e-01 5.30457795e-01 2.90989310e-01 -5.46185017e-01
-1.16723871e+00 -6.54016554e-01 1.21546388e-01 -3.64380687e-01
-2.09523365e-01 -4.06856447e-01 -1.11076808e+00 -5.41881084e-01
-6.26172125e-01 1.62156254e-01 6.28235042e-01 1.22785425e+00
4.88042772e-01 2.42001384e-01 7.14785576e-01 -2.72962749e-01
-8.52117479e-01 -1.22812080e+00 -1.12518705e-01 8.69538426e-01
-8.97995681e-02 -5.46455160e-02 -3.61834049e-01 -1.49555638e-01] | [11.138785362243652, 8.32764720916748] |
4afd2adb-f386-44e8-b312-865b74051952 | streaming-variational-monte-carlo | 1906.01549 | null | https://arxiv.org/abs/1906.01549v4 | https://arxiv.org/pdf/1906.01549v4.pdf | Streaming Variational Monte Carlo | Nonlinear state-space models are powerful tools to describe dynamical structures in complex time series. In a streaming setting where data are processed one sample at a time, simultaneous inference of the state and its nonlinear dynamics has posed significant challenges in practice. We develop a novel online learning framework, leveraging variational inference and sequential Monte Carlo, which enables flexible and accurate Bayesian joint filtering. Our method provides an approximation of the filtering posterior which can be made arbitrarily close to the true filtering distribution for a wide class of dynamics models and observation models. Specifically, the proposed framework can efficiently approximate a posterior over the dynamics using sparse Gaussian processes, allowing for an interpretable model of the latent dynamics. Constant time complexity per sample makes our approach amenable to online learning scenarios and suitable for real-time applications. | ['Mónica Bugallo', 'Josue Nassar', 'Yuan Zhao', 'Il Memming Park', 'Ian Jordan'] | 2019-06-04 | null | null | null | null | ['variational-monte-carlo'] | ['miscellaneous'] | [-1.01936966e-01 -3.62896204e-01 -2.60541469e-01 -9.48433951e-03
-7.09906697e-01 -5.76869130e-01 8.42584670e-01 6.92929476e-02
-8.98465514e-02 5.59904993e-01 6.00015335e-02 -2.67675847e-01
-3.98775160e-01 -7.29787171e-01 -7.17940688e-01 -7.90515244e-01
-4.51645374e-01 7.28553116e-01 2.66076207e-01 3.24291199e-01
-1.36384010e-01 5.15507698e-01 -1.18524504e+00 -4.05540943e-01
7.45843649e-01 8.13426673e-01 2.46262744e-01 1.00033081e+00
-1.08262256e-01 5.42838871e-01 -4.49739009e-01 -1.09631196e-01
5.94112314e-02 -1.21551089e-01 -6.62912428e-02 3.45219553e-01
1.74897328e-01 -4.74605858e-01 -5.73293030e-01 1.08624434e+00
8.71913508e-02 1.97038412e-01 9.09190059e-01 -1.13686371e+00
-1.64704025e-01 4.60192859e-01 -4.75658119e-01 3.03452492e-01
1.28910601e-01 1.63592130e-01 9.65972900e-01 -5.51034272e-01
2.92435825e-01 1.40540504e+00 6.45274580e-01 1.39714390e-01
-1.75953758e+00 -5.22624254e-01 6.12365544e-01 1.02861878e-02
-1.42457330e+00 -4.72732931e-01 5.30063212e-01 -5.94814956e-01
4.47396606e-01 -2.66297517e-04 1.01569998e+00 1.17839968e+00
5.74290812e-01 9.41949308e-01 1.00418711e+00 2.80857421e-02
6.65365875e-01 -3.82227391e-01 -9.70357936e-03 5.20455897e-01
2.05596343e-01 2.00057209e-01 -5.71968555e-01 -6.34731591e-01
1.21433139e+00 6.67604804e-01 -1.57865748e-01 -3.53482276e-01
-1.02608216e+00 7.76503146e-01 -2.54237920e-01 -2.80209959e-01
-5.72284877e-01 5.54221332e-01 1.68481559e-01 2.76160419e-01
5.27864695e-01 -3.32435936e-01 -4.77784038e-01 -4.04810965e-01
-1.03556037e+00 7.48900115e-01 1.25879729e+00 7.68889725e-01
5.40561080e-01 2.99455732e-01 -3.30401629e-01 2.44455218e-01
6.65734589e-01 1.11692226e+00 -9.46166664e-02 -1.01783454e+00
2.07541466e-01 -2.22299069e-01 7.72642493e-01 -6.82560742e-01
5.39265424e-02 -5.58675706e-01 -1.02371168e+00 4.80856281e-03
7.50301838e-01 -3.47689003e-01 -8.33774447e-01 1.74175048e+00
6.47568166e-01 9.59744751e-01 -1.39521688e-01 3.87470871e-01
-2.26308092e-01 1.10092163e+00 -1.49447948e-01 -7.85530388e-01
1.15252018e+00 -5.76851487e-01 -1.05413485e+00 -9.81862396e-02
-9.07201394e-02 -5.86638987e-01 8.28332543e-01 6.08246922e-01
-1.05292547e+00 -2.77964264e-01 -4.97173011e-01 4.82598960e-01
2.68095016e-01 -1.70376837e-01 5.93874097e-01 4.19762254e-01
-7.18949437e-01 6.86584473e-01 -1.82954216e+00 -5.77822886e-02
1.90514058e-01 8.02799091e-02 2.91864842e-01 -6.06648251e-02
-7.83885479e-01 3.57667536e-01 1.78137884e-01 2.14895099e-01
-1.51582313e+00 -8.72047842e-01 -4.49195325e-01 2.51408458e-01
5.89039087e-01 -7.06480563e-01 1.61599219e+00 -2.39523157e-01
-1.97561681e+00 -8.06156397e-02 -6.94740951e-01 -6.86453998e-01
7.31094241e-01 -4.39172119e-01 -5.42447448e-01 9.84474570e-02
-2.01387778e-01 -3.89537424e-01 1.68526375e+00 -7.43591726e-01
-4.16199863e-01 -1.61995173e-01 -2.40289852e-01 -4.76084352e-02
-1.14813246e-01 -2.21188083e-01 -4.39390957e-01 -6.03684366e-01
8.00162181e-02 -1.00089729e+00 -5.22180438e-01 2.49579668e-01
-6.30556643e-02 -2.71729410e-01 8.44887197e-01 -5.95020473e-01
1.39099610e+00 -1.75078797e+00 1.82517767e-01 3.15035939e-01
1.14553370e-01 9.00455117e-02 2.34041452e-01 9.35469031e-01
4.49495018e-01 -2.63474643e-01 -1.45518199e-01 -5.97066760e-01
1.95261657e-01 4.31383371e-01 -8.76945674e-01 7.21886754e-01
2.65175402e-02 6.13584578e-01 -1.00067019e+00 -1.80231407e-01
3.08763832e-01 4.81794894e-01 -5.65072298e-01 4.41234976e-01
-4.90524769e-01 6.84109926e-01 -6.27791584e-01 3.37827057e-01
5.95316648e-01 -6.15415454e-01 2.35189855e-01 1.38472110e-01
-1.15972765e-01 -1.07326582e-01 -1.67129576e+00 1.33923900e+00
-6.64667785e-01 3.11694503e-01 3.33335280e-01 -9.51182544e-01
6.91344559e-01 4.96094525e-01 3.94333124e-01 2.69629657e-01
-1.56676620e-01 2.96897404e-02 -2.35853732e-01 -3.61487448e-01
1.23039208e-01 -3.95468384e-01 -5.82786575e-02 6.08226120e-01
7.51543045e-02 -2.07482234e-01 1.18117429e-01 8.03943798e-02
9.63777661e-01 8.35912600e-02 4.09638196e-01 -9.81580466e-02
4.76581454e-02 -4.46179599e-01 7.08324611e-01 1.22065735e+00
-3.57418344e-03 -9.18975845e-03 3.01131308e-01 -3.15946609e-01
-1.10342538e+00 -1.66501558e+00 -1.16633736e-01 5.84540069e-01
-1.43946335e-01 -4.30571139e-01 -3.11386317e-01 -2.49186084e-02
1.23736843e-01 3.44582140e-01 -3.21068048e-01 2.10595429e-02
-4.09844726e-01 -7.88936198e-01 1.05076851e-02 3.34239811e-01
-5.51492237e-02 -4.03336465e-01 -3.75725776e-01 6.52060270e-01
1.01100616e-02 -1.03997934e+00 -5.68839073e-01 -2.62056887e-01
-1.25701630e+00 -6.72772706e-01 -4.87678736e-01 -4.59467955e-02
2.26099119e-01 1.28008530e-01 8.84920239e-01 -5.54501176e-01
-6.78182170e-02 6.02814317e-01 2.22095191e-01 -4.01114136e-01
-5.10748625e-01 -2.18790740e-01 3.37530673e-01 4.47969675e-01
-3.39523330e-02 -9.10227060e-01 -5.61060607e-01 1.16425622e-02
-1.03416657e+00 -1.72866434e-01 3.05498272e-01 8.05830359e-01
7.48470306e-01 3.94043177e-01 3.41488808e-01 -4.88180012e-01
5.81095874e-01 -7.51581311e-01 -1.25512719e+00 2.60276705e-01
-4.33376849e-01 3.12409550e-02 7.29386508e-01 -9.11624253e-01
-1.21118677e+00 1.77301262e-02 1.76734850e-01 -8.34587395e-01
-6.60005957e-02 7.52291620e-01 1.56209648e-01 3.45608175e-01
1.80613682e-01 5.00655651e-01 1.91960528e-01 -6.56829119e-01
2.84010857e-01 3.35793614e-01 6.38860166e-01 -8.56934786e-01
8.77870917e-01 9.16721523e-01 2.80949056e-01 -9.50369418e-01
-1.03679144e+00 -5.07896543e-01 -4.11906898e-01 -1.49460331e-01
4.08174127e-01 -1.31368387e+00 -9.87236440e-01 6.61698878e-01
-9.35886681e-01 -4.44051057e-01 -3.17623615e-01 8.24023604e-01
-7.45756567e-01 3.54279220e-01 -8.03771019e-01 -1.46669924e+00
-4.84110266e-02 -6.29021525e-01 1.07851160e+00 2.50418991e-01
-1.05682619e-01 -1.52107823e+00 4.40112144e-01 -2.99022853e-01
1.95582673e-01 2.39460483e-01 5.82011878e-01 -2.71101266e-01
-8.46551776e-01 -6.56767726e-01 1.96522862e-01 5.41759804e-02
1.52492970e-01 3.93393397e-01 -5.02101004e-01 -5.80351412e-01
2.69209653e-01 9.32774171e-02 7.23871887e-01 8.29091072e-01
1.02531314e+00 -4.86809075e-01 -3.84345770e-01 3.23532909e-01
1.39275205e+00 -1.17298812e-01 1.50500402e-01 -4.63374615e-01
5.02186894e-01 9.89881381e-02 4.94928032e-01 1.18902898e+00
2.71438658e-01 5.27279317e-01 1.64050877e-01 5.27579188e-01
5.02759516e-01 -5.58932841e-01 5.93467653e-01 1.06433558e+00
1.58177182e-01 -2.49188736e-01 -7.59200096e-01 7.88467407e-01
-2.15956616e+00 -1.23891342e+00 -5.48805818e-02 2.40135050e+00
8.24500084e-01 1.84523851e-01 3.27677459e-01 -2.41651833e-01
6.36812091e-01 1.68303534e-01 -9.37881887e-01 1.47799566e-01
1.98671460e-01 2.51825809e-01 3.70677263e-01 6.51760519e-01
-1.08800828e+00 6.57388747e-01 7.64464569e+00 9.60733771e-01
-8.19131732e-01 2.30023265e-01 2.55672503e-02 -9.59048420e-02
-1.67856291e-01 2.78448731e-01 -1.16028595e+00 6.73775613e-01
1.41485715e+00 -4.49324399e-01 5.57611883e-01 4.07744348e-01
8.93803716e-01 2.20026765e-02 -8.59722137e-01 8.81920278e-01
-6.34516239e-01 -1.37486637e+00 3.34254391e-02 3.26985240e-01
8.91127765e-01 8.19175094e-02 7.45879337e-02 5.59612773e-02
8.56947660e-01 -6.17936134e-01 6.56936765e-01 1.09528530e+00
2.80954808e-01 -6.54388547e-01 1.50599450e-01 7.85474658e-01
-1.19225359e+00 -3.66342336e-01 -3.67542744e-01 -3.83283854e-01
8.32593083e-01 1.12099659e+00 -5.32245874e-01 1.04193054e-01
5.35059333e-01 9.11586106e-01 2.45099023e-01 1.22675347e+00
-1.83888555e-01 1.23599064e+00 -8.95561814e-01 1.06491886e-01
5.90601377e-02 -5.71060419e-01 9.48576093e-01 9.09009278e-01
7.11887479e-01 -1.89167067e-01 7.38099694e-01 8.34914386e-01
3.41530144e-01 -3.15423667e-01 -3.10681134e-01 -4.65033025e-01
8.36112201e-01 9.88995552e-01 -6.37382090e-01 -5.04368961e-01
-4.24875051e-01 6.32621109e-01 2.93219000e-01 7.10749030e-01
-8.34280729e-01 1.19833186e-01 8.67893636e-01 6.18083291e-02
8.36134076e-01 -8.10397625e-01 2.29492083e-01 -1.67441189e+00
-1.49434745e-01 -6.46679997e-01 2.96101630e-01 -2.84804821e-01
-1.56040454e+00 -1.43603906e-01 2.60126740e-01 -1.15056777e+00
-5.85954905e-01 -2.92695463e-01 -8.54084969e-01 8.00527453e-01
-1.18905890e+00 -8.73961449e-01 2.80101508e-01 7.14322686e-01
5.68381011e-01 5.17053157e-03 7.02590168e-01 -2.16492146e-01
-5.95683336e-01 -7.28786364e-02 7.99915910e-01 -3.61708015e-01
2.31886059e-01 -1.32337701e+00 4.73209471e-01 9.27402675e-01
3.68549496e-01 7.47028887e-01 1.34045684e+00 -8.77960086e-01
-1.73025918e+00 -1.07411432e+00 3.35661262e-01 -2.59430170e-01
1.43574619e+00 -4.04015571e-01 -1.08678138e+00 7.00184524e-01
-1.78405195e-01 8.82398859e-02 6.94255888e-01 2.15487987e-01
-1.41754642e-01 -4.07153182e-02 -7.80132294e-01 5.79898715e-01
7.01428413e-01 -6.63708806e-01 -2.74169713e-01 5.38727224e-01
5.66011012e-01 -3.27737808e-01 -1.16283286e+00 -3.11959796e-02
6.68220341e-01 -3.88040423e-01 9.09409344e-01 -6.48251832e-01
-7.68634379e-02 -4.93860006e-01 -5.82948849e-02 -1.31387508e+00
-2.81401128e-01 -1.37395835e+00 -1.38428223e+00 9.88590896e-01
-1.83709115e-02 -6.30865574e-01 4.75845724e-01 4.62589473e-01
3.34526390e-01 -4.45402920e-01 -9.66284752e-01 -1.03455758e+00
-9.86301079e-02 -6.06830895e-01 4.96564806e-01 3.74221951e-01
-4.18428272e-01 -8.45848992e-02 -8.06264460e-01 7.74313450e-01
1.20074654e+00 4.39865887e-01 8.62882733e-01 -1.43036675e+00
-9.33607221e-01 -2.98208669e-02 -8.01006556e-02 -1.66621208e+00
3.00364554e-01 -3.16910654e-01 6.47406504e-02 -1.09986603e+00
1.91138908e-01 -9.49060321e-02 -3.64825204e-02 -1.90413639e-01
-2.12165669e-01 -9.01448280e-02 -9.76392478e-02 4.53787148e-01
-4.53743607e-01 8.20948064e-01 1.09076691e+00 1.92008287e-01
-1.19603619e-01 7.75459588e-01 -5.41170873e-02 8.18254769e-01
5.48429072e-01 -5.00041306e-01 -5.84624290e-01 -2.03539878e-01
1.40080824e-01 5.75153232e-01 6.44653976e-01 -6.92762136e-01
3.81266564e-01 -5.72894692e-01 4.40010317e-02 -8.56337607e-01
5.71913302e-01 -7.12571859e-01 4.89697248e-01 5.51544547e-01
-1.32570416e-01 -1.02270044e-01 -1.89125650e-02 1.61086762e+00
-8.69948491e-02 -3.08937840e-02 6.35867357e-01 -1.35451676e-02
-2.95002431e-01 8.28863204e-01 -8.73728037e-01 3.90770957e-02
6.63495600e-01 1.39080763e-01 2.35773042e-01 -9.90487933e-01
-1.32867837e+00 3.29490930e-01 1.26652002e-01 -6.99273646e-02
3.82248133e-01 -1.17537594e+00 -5.91400564e-01 6.88532516e-02
-2.82610148e-01 -8.40498358e-02 5.38739443e-01 8.79744530e-01
-9.36134383e-02 1.91693157e-01 4.27105337e-01 -8.87080669e-01
-7.54056811e-01 4.56808925e-01 1.87887281e-01 -5.35565734e-01
-7.52261221e-01 4.08130229e-01 -4.18921076e-02 -1.16508827e-01
5.13799861e-02 -4.69516963e-01 2.75131196e-01 -1.40283689e-01
8.97892773e-01 5.55164993e-01 -4.81750041e-01 -2.63638139e-01
2.00453654e-01 2.61060059e-01 -7.83983022e-02 -3.66443902e-01
1.24074924e+00 -6.10506296e-01 -9.73700173e-03 1.28005195e+00
8.18444967e-01 -2.85027385e-01 -2.13497782e+00 -6.57946944e-01
-4.61986065e-02 -5.34072101e-01 2.02912018e-01 -2.38281697e-01
-6.40178204e-01 8.36190641e-01 4.71053988e-01 5.05425453e-01
6.42980993e-01 -5.53211793e-02 6.76550329e-01 4.46447968e-01
4.04279172e-01 -8.36202741e-01 -1.00458980e-01 4.98153627e-01
4.74927366e-01 -9.19897437e-01 -1.57780088e-02 -3.29292625e-01
-7.23208115e-02 1.02167201e+00 -1.44756272e-01 -5.42448997e-01
1.26548648e+00 3.92581701e-01 -4.12766635e-01 1.30886450e-01
-1.27177668e+00 8.83819461e-02 2.26762429e-01 5.04251242e-01
-3.17213982e-02 3.33876133e-01 1.07438758e-01 3.81818771e-01
1.18307307e-01 1.12908609e-01 5.75075746e-01 8.35285366e-01
-4.42073077e-01 -8.45406473e-01 -3.90259415e-01 4.98217851e-01
-6.07714653e-01 1.13141209e-01 5.63419342e-01 2.99526125e-01
-5.87340415e-01 8.57084811e-01 3.88412774e-01 4.67129916e-01
-2.24920548e-02 8.72968063e-02 5.57648897e-01 -5.22535920e-01
2.46376336e-01 7.71949530e-01 -3.63980174e-01 -4.89939362e-01
-3.65305007e-01 -1.30696857e+00 -8.27155590e-01 -3.79278272e-01
-2.93894708e-01 1.30197644e-01 4.31560993e-01 1.24063480e+00
3.26805949e-01 2.29779020e-01 6.55843377e-01 -9.53170180e-01
-1.23002565e+00 -7.83747733e-01 -9.31606114e-01 -1.90810263e-02
5.98752439e-01 -7.61351168e-01 -4.18562174e-01 2.16750368e-01] | [6.807008743286133, 3.7776670455932617] |
afb60c09-c66d-4a56-8f4a-15f090b720b6 | pose-estimation-and-3d-reconstruction-of | 2107.10898 | null | https://arxiv.org/abs/2107.10898v1 | https://arxiv.org/pdf/2107.10898v1.pdf | Pose Estimation and 3D Reconstruction of Vehicles from Stereo-Images Using a Subcategory-Aware Shape Prior | The 3D reconstruction of objects is a prerequisite for many highly relevant applications of computer vision such as mobile robotics or autonomous driving. To deal with the inverse problem of reconstructing 3D objects from their 2D projections, a common strategy is to incorporate prior object knowledge into the reconstruction approach by establishing a 3D model and aligning it to the 2D image plane. However, current approaches are limited due to inadequate shape priors and the insufficiency of the derived image observations for a reliable alignment with the 3D model. The goal of this paper is to show how 3D object reconstruction can profit from a more sophisticated shape prior and from a combined incorporation of different observation types inferred from the images. We introduce a subcategory-aware deformable vehicle model that makes use of a prediction of the vehicle type for a more appropriate regularisation of the vehicle shape. A multi-branch CNN is presented to derive predictions of the vehicle type and orientation. This information is also introduced as prior information for model fitting. Furthermore, the CNN extracts vehicle keypoints and wireframes, which are well-suited for model-to-image association and model fitting. The task of pose estimation and reconstruction is addressed by a versatile probabilistic model. Extensive experiments are conducted using two challenging real-world data sets on both of which the benefit of the developed shape prior can be shown. A comparison to state-of-the-art methods for vehicle pose estimation shows that the proposed approach performs on par or better, confirming the suitability of the developed shape prior and probabilistic model for vehicle reconstruction. | ['Franz Rottensteiner', 'Max Coenen'] | 2021-07-22 | null | null | null | null | ['3d-object-reconstruction', 'vehicle-pose-estimation', 'object-reconstruction'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 1.70064643e-01 1.18781604e-01 -7.44455904e-02 -5.63487172e-01
-5.51183403e-01 -4.50773031e-01 9.47721601e-01 -2.92174593e-02
-4.85848367e-01 3.93009543e-01 -3.72818023e-01 -2.11672127e-01
-2.23969832e-01 -6.72974408e-01 -1.04445279e+00 -6.62490487e-01
2.86094069e-01 1.17397201e+00 5.26855707e-01 -2.01746479e-01
2.06518143e-01 1.25205088e+00 -2.02492952e+00 -2.33167648e-01
3.93465430e-01 1.06345177e+00 6.35870218e-01 3.29830348e-01
-9.16896239e-02 -6.70734569e-02 -3.84738967e-02 -2.08641052e-01
3.65133077e-01 1.66832969e-01 -4.21267390e-01 5.72044373e-01
3.84530991e-01 -3.57879281e-01 5.04086129e-02 8.55777919e-01
-1.71035286e-02 -1.65256124e-03 9.56778288e-01 -1.21942842e+00
2.73804605e-01 -2.27495477e-01 -1.33562341e-01 -2.98942775e-01
2.09872007e-01 -8.22019801e-02 5.24716794e-01 -1.09943438e+00
9.48219001e-01 1.10292411e+00 7.94435918e-01 2.07019523e-01
-1.28127980e+00 -2.94945180e-01 -1.76512823e-02 4.58469987e-01
-1.49994028e+00 -5.61300218e-01 9.77545738e-01 -6.79963410e-01
8.10474694e-01 1.81418493e-01 6.30932093e-01 9.16791439e-01
-3.03444900e-02 5.49717128e-01 1.02253294e+00 -4.93536562e-01
1.63214564e-01 5.87917387e-01 1.25716655e-02 4.07414049e-01
3.62237096e-01 2.98563242e-01 -7.35244229e-02 7.62504265e-02
6.57107115e-01 -3.89787741e-02 -7.56706437e-03 -1.10256386e+00
-9.43585634e-01 6.42078638e-01 2.55802065e-01 2.25986838e-01
-6.14669383e-01 8.15047920e-02 1.09189309e-01 -2.53442734e-01
3.42444569e-01 -7.11918771e-02 -2.73848861e-01 2.10694090e-01
-8.10870409e-01 5.15988052e-01 6.78593099e-01 1.06332445e+00
1.17626560e+00 1.48736492e-01 2.53464401e-01 4.36689377e-01
9.63582814e-01 9.81095195e-01 -1.07970372e-01 -9.58759487e-01
1.09041087e-01 5.34394622e-01 3.76639545e-01 -1.09743345e+00
-3.08555692e-01 -3.21911722e-01 -4.62970734e-01 5.76718390e-01
4.46909338e-01 4.90223885e-01 -9.46838379e-01 1.37010407e+00
6.72320902e-01 2.80636489e-01 3.25252092e-03 8.41803849e-01
7.04703510e-01 3.89057219e-01 -2.49981403e-01 -8.77269879e-02
1.31063557e+00 -2.11199403e-01 -5.70875049e-01 -2.51592904e-01
4.21366692e-01 -6.67383969e-01 1.77711800e-01 2.86697716e-01
-7.85449266e-01 -7.26695240e-01 -1.02107525e+00 1.29528299e-01
-3.56998771e-01 2.50872970e-01 6.23965003e-02 5.20639122e-01
-7.54911363e-01 3.27006817e-01 -8.91819417e-01 -3.99643242e-01
1.63382679e-01 4.61961746e-01 -7.49345422e-01 1.26837054e-02
-7.79994488e-01 1.56484342e+00 4.66986150e-01 3.58099997e-01
-7.65102267e-01 -4.09697354e-01 -1.10461187e+00 -3.07229996e-01
1.93304271e-01 -7.21262932e-01 9.26622868e-01 -6.84857965e-01
-1.43890023e+00 1.19674873e+00 -2.24776506e-01 -7.06544340e-01
8.08246255e-01 1.73823480e-02 1.81150157e-02 1.64354458e-01
-2.17643883e-02 7.16882706e-01 1.15860069e+00 -1.77172613e+00
-5.30641735e-01 -6.07191980e-01 -1.88660160e-01 4.50023785e-02
2.50555009e-01 -3.52116495e-01 -7.46131361e-01 3.59068513e-02
2.34251603e-01 -1.13841271e+00 -2.05291808e-01 -2.55538300e-02
-1.84608772e-01 -1.08970813e-01 1.09446323e+00 -7.51381159e-01
3.81634504e-01 -1.91754723e+00 3.10050696e-01 5.24889588e-01
2.96920463e-02 2.79757142e-01 1.02397695e-01 2.42700532e-01
-3.34233083e-02 -3.53436679e-01 -2.42151737e-01 -7.32077718e-01
2.04117626e-01 5.02745509e-01 -1.97120130e-01 9.59607661e-01
2.33018935e-01 6.91899061e-01 -3.42225313e-01 -3.33252668e-01
9.27864909e-01 9.86179292e-01 -2.46100858e-01 2.57099479e-01
-2.97400832e-01 4.86795664e-01 -4.86103505e-01 2.46667743e-01
1.06163192e+00 3.16270441e-01 -5.27406065e-03 -4.62583452e-01
-2.31283352e-01 -9.60819200e-02 -1.48671675e+00 1.24511588e+00
-4.15021002e-01 4.39738542e-01 2.83634663e-01 -1.27852190e+00
1.33985841e+00 2.88066268e-01 5.88551581e-01 -4.70955998e-01
4.15090531e-01 4.06945646e-01 -3.57905090e-01 -4.83349860e-01
4.60801691e-01 -2.30459288e-01 1.58881336e-01 -7.37725049e-02
2.97902990e-02 -6.23999774e-01 -5.36960289e-02 -2.77372688e-01
4.45032299e-01 5.97896636e-01 3.04743290e-01 -2.47819558e-01
8.66822243e-01 6.29481524e-02 3.84561777e-01 1.80473208e-01
1.52502254e-01 5.63375652e-01 -2.11196877e-02 -6.03884637e-01
-1.30115116e+00 -7.50984251e-01 -5.53120792e-01 1.42076284e-01
2.68763006e-01 9.33940783e-02 -6.84878170e-01 -5.42959750e-01
2.03938946e-01 7.77456820e-01 -6.09222591e-01 1.20575786e-01
-7.49460638e-01 -3.82247627e-01 2.31564716e-02 3.06292623e-01
1.57876357e-01 -8.23276639e-01 -8.90692890e-01 1.38590828e-01
-1.15332380e-01 -1.46011996e+00 2.40343735e-01 8.68915394e-02
-8.80514741e-01 -9.68714476e-01 -4.77564663e-01 -4.55014259e-01
7.44432926e-01 2.11848468e-01 9.29864645e-01 2.44581159e-02
-4.06368412e-02 6.14932477e-01 -1.97153777e-01 -5.50609946e-01
-8.45468104e-01 -2.41419062e-01 2.53794372e-01 3.59268308e-01
3.28311652e-01 -4.70557213e-01 -1.48044676e-01 5.54423034e-01
-7.95418978e-01 7.62173459e-02 6.40251577e-01 3.82557213e-01
9.14878905e-01 -9.71119627e-02 1.90249816e-01 -5.27741551e-01
-2.33980089e-01 -3.38332146e-01 -1.10228264e+00 -1.82682648e-01
-4.24966514e-01 9.66619700e-02 3.48457068e-01 -2.46067211e-01
-9.83055472e-01 7.45529115e-01 -5.06349862e-01 -7.48662829e-01
-7.49197006e-01 2.37031505e-01 -5.42690694e-01 -2.07648918e-01
3.81529868e-01 1.83267608e-01 4.73163456e-01 -5.71294487e-01
1.19336091e-01 4.17709202e-01 4.41509426e-01 -3.37550163e-01
1.03066504e+00 7.47881711e-01 4.87953454e-01 -1.03919005e+00
-2.35172331e-01 -6.65078580e-01 -1.00943124e+00 -5.50468147e-01
9.09976363e-01 -8.13764572e-01 -5.44818997e-01 3.21949154e-01
-1.41609991e+00 -3.99234071e-02 -2.71607906e-01 6.93158746e-01
-9.89651382e-01 5.21817207e-01 2.30926275e-01 -1.01995885e+00
6.02751002e-02 -1.58589065e+00 1.30637455e+00 -1.65657908e-01
6.02103472e-02 -1.05135584e+00 -8.06565732e-02 5.62662184e-01
1.10088028e-01 3.73914748e-01 5.27219176e-01 -5.29317796e-01
-9.04205322e-01 -5.99562049e-01 -6.66159764e-02 3.39962184e-01
-3.02496761e-01 -4.52149427e-03 -9.53605354e-01 7.90627394e-03
8.87275189e-02 2.61903584e-01 7.29982197e-01 4.97242957e-01
5.21021366e-01 2.28048682e-01 -6.46572351e-01 5.12149334e-01
1.56804848e+00 -1.10076979e-01 6.48116708e-01 4.88290370e-01
6.34225249e-01 9.38483715e-01 8.19524229e-01 2.81141907e-01
5.73808610e-01 1.35597897e+00 1.12384117e+00 2.16758803e-01
-2.08634079e-01 -8.57023895e-02 2.75762439e-01 4.22549188e-01
-2.65875965e-01 1.25338063e-01 -9.44038808e-01 5.75838327e-01
-1.84108210e+00 -7.52143741e-01 -6.37136817e-01 2.43509030e+00
2.03792170e-01 2.83099651e-01 1.60361901e-01 2.55747616e-01
5.59426844e-01 -3.53264362e-01 -2.29454651e-01 -1.06602646e-01
3.09636891e-02 -2.79235095e-01 7.43799210e-01 7.43212521e-01
-1.05973172e+00 6.00335896e-01 5.27313042e+00 7.73037553e-01
-1.05520642e+00 -6.71470240e-02 3.41571681e-02 6.86314225e-01
-3.63272011e-01 1.13922857e-01 -1.39717364e+00 2.25822985e-01
8.62288117e-01 3.00934881e-01 8.09555650e-02 8.25305820e-01
2.77349502e-01 -2.27746084e-01 -1.11374915e+00 9.07900214e-01
2.75149733e-01 -1.26909113e+00 3.74533944e-02 4.51213986e-01
4.22128946e-01 8.92465785e-02 -1.95632100e-01 -1.15223676e-01
-2.82147497e-01 -7.53440440e-01 1.29266787e+00 9.08845246e-01
5.65859795e-01 -6.73635602e-01 1.01780713e+00 8.22120726e-01
-1.12734318e+00 1.95190683e-01 -1.44403338e-01 1.91325232e-01
5.14257491e-01 4.12713051e-01 -1.26169503e+00 8.24879348e-01
6.35105789e-01 6.83546603e-01 -3.84820461e-01 1.10239124e+00
7.21684918e-02 2.54393727e-01 -7.06432998e-01 1.41402394e-01
7.78212324e-02 -2.68937200e-01 1.02539694e+00 1.17483318e+00
5.43003798e-01 -3.09364200e-01 6.40369281e-02 1.07082474e+00
4.80660200e-01 2.52874810e-02 -8.90913785e-01 4.61111724e-01
2.57047892e-01 1.36749613e+00 -9.20551062e-01 -1.90891370e-01
-2.09066018e-01 2.67160892e-01 -4.80648316e-02 1.06884331e-01
-7.46196926e-01 2.40453988e-01 4.71857488e-01 5.85458696e-01
8.13193262e-01 -4.96060938e-01 -2.29097545e-01 -6.87960923e-01
1.34889290e-01 -2.67655849e-01 -4.03255492e-01 -7.58174896e-01
-8.54679227e-01 6.01468861e-01 6.54719472e-01 -1.49408078e+00
-6.76023364e-01 -9.10423934e-01 -2.08912551e-01 9.50199127e-01
-1.70061862e+00 -1.52631032e+00 -2.66588867e-01 3.29718351e-01
3.83417219e-01 5.31257922e-03 6.77891552e-01 4.58738245e-02
-1.00126803e-01 -6.38921037e-02 -1.18462317e-01 -2.84426242e-01
2.14917347e-01 -9.64545429e-01 -1.52633991e-02 7.04794407e-01
4.92600277e-02 1.89138949e-01 1.14348710e+00 -6.70832574e-01
-1.60445881e+00 -1.04020309e+00 9.16257679e-01 -7.35894620e-01
2.75576621e-01 -3.86675686e-01 -9.09283221e-01 5.54870605e-01
-1.59480542e-01 2.09870368e-01 1.04743026e-01 -4.70967472e-01
-5.88258132e-02 -2.33679861e-01 -1.14130759e+00 1.09763205e-01
5.98157108e-01 -2.97969371e-01 -5.84847033e-01 -1.60130933e-02
1.35390997e-01 -4.32538092e-01 -7.20621407e-01 7.70973325e-01
5.99746585e-01 -1.04528785e+00 1.33050025e+00 -2.66829617e-02
-1.24524325e-01 -5.91125965e-01 -3.39630157e-01 -8.93638909e-01
-7.33597428e-02 -1.28376454e-01 -1.14358395e-01 1.03283000e+00
7.75975510e-02 -4.76686835e-01 7.59229541e-01 5.87744296e-01
-3.77814561e-01 -5.29719532e-01 -1.23172987e+00 -7.18410790e-01
-1.86969161e-01 -8.56446087e-01 3.47708195e-01 4.45386201e-01
-7.79008567e-01 1.84893757e-01 -2.25562632e-01 5.06914616e-01
8.03513288e-01 6.23890795e-02 1.24391162e+00 -1.73177743e+00
4.59808856e-02 -4.24314260e-01 -8.79355252e-01 -1.00086153e+00
5.01154602e-01 -7.51497030e-01 3.19055557e-01 -1.41260648e+00
-1.17353529e-01 -5.55477202e-01 2.56149411e-01 -3.12959850e-02
3.34993958e-01 5.28614879e-01 2.73250431e-01 1.77240282e-01
-3.00508171e-01 5.62106967e-01 9.15970504e-01 7.72679821e-02
1.84998866e-02 3.74699295e-01 1.05228655e-01 9.74390149e-01
4.78779793e-01 -3.00545484e-01 -1.41498879e-01 -2.22934783e-01
2.45686859e-01 7.39845559e-02 8.45984578e-01 -8.27316999e-01
5.75088970e-02 1.65661097e-01 2.22163036e-01 -1.12196779e+00
8.68641257e-01 -1.54208672e+00 6.71981096e-01 1.81367755e-01
2.23718256e-01 -1.76482499e-01 2.86904871e-01 6.39713287e-01
-9.73781571e-02 -5.60419142e-01 7.66782284e-01 -1.10496126e-01
-9.21842515e-01 2.76057959e-01 -3.18787456e-01 -6.10930324e-01
1.18326211e+00 -8.67603183e-01 5.40145516e-01 -2.75032848e-01
-7.84203351e-01 -1.49706230e-01 7.06743121e-01 4.54243273e-01
7.65081882e-01 -1.28325450e+00 -7.85619140e-01 4.61042404e-01
2.65828162e-01 1.93454430e-01 2.99362034e-01 1.10984051e+00
-4.75516647e-01 6.24142587e-01 -3.07621002e-01 -1.17387855e+00
-1.42384243e+00 5.61063647e-01 3.99499297e-01 2.35067420e-02
-6.20790362e-01 2.13787317e-01 1.29908726e-01 -6.58927381e-01
3.09749674e-02 -3.93383801e-01 -6.14789844e-01 8.60334858e-02
2.44197533e-01 2.76897967e-01 3.65468949e-01 -1.62819993e+00
-4.94618505e-01 1.07060134e+00 3.81456584e-01 -9.79919285e-02
1.46563685e+00 -3.89105380e-01 6.41354248e-02 3.04044634e-01
1.02161670e+00 -7.79146701e-02 -1.47943723e+00 -2.18111366e-01
-2.38338877e-02 -2.63668746e-01 2.02066407e-01 -2.52611548e-01
-7.38421917e-01 9.45447147e-01 5.79243481e-01 1.96795672e-01
6.74481869e-01 1.18773676e-01 3.30529422e-01 1.96829155e-01
5.94952226e-01 -8.02096963e-01 -4.46716547e-01 4.91970986e-01
1.11386693e+00 -1.22183633e+00 5.42513430e-02 -5.57865143e-01
-2.48938918e-01 1.39779127e+00 1.53055266e-01 -1.93767101e-01
8.14325869e-01 1.71139210e-01 -8.02475065e-02 -4.60411198e-02
-1.48521036e-01 -5.02606094e-01 6.56798065e-01 7.97128081e-01
-2.44631901e-01 4.86539006e-02 -4.77612987e-02 3.48920971e-01
-4.79379902e-03 -1.34144306e-01 2.49977902e-01 6.21713579e-01
-4.33735877e-01 -1.08264446e+00 -7.90908575e-01 -9.88053828e-02
1.16365161e-02 4.89202082e-01 -8.94651189e-02 9.28824544e-01
4.70507741e-01 5.57310462e-01 2.03299657e-01 -2.63018668e-01
4.86719549e-01 9.56027806e-02 6.15720630e-01 -4.36204523e-01
2.66430841e-04 2.83826083e-01 8.45946148e-02 -3.89759898e-01
-8.68873060e-01 -1.02385473e+00 -1.18214047e+00 1.79063052e-01
-4.72092807e-01 -1.50301799e-01 1.35350323e+00 1.21840990e+00
2.52756685e-01 7.91562125e-02 4.11474943e-01 -1.74119091e+00
-4.57377702e-01 -7.52414227e-01 -3.62431973e-01 4.27025110e-01
3.59709918e-01 -1.20246303e+00 -2.40260422e-01 2.94492513e-01] | [7.712561130523682, -2.6407310962677] |
80e7a680-24e3-41a5-98b8-dfae9a863b02 | predicting-clinical-trial-results-by-implicit | 2010.05639 | null | https://arxiv.org/abs/2010.05639v1 | https://arxiv.org/pdf/2010.05639v1.pdf | Predicting Clinical Trial Results by Implicit Evidence Integration | Clinical trials provide essential guidance for practicing Evidence-Based Medicine, though often accompanying with unendurable costs and risks. To optimize the design of clinical trials, we introduce a novel Clinical Trial Result Prediction (CTRP) task. In the CTRP framework, a model takes a PICO-formatted clinical trial proposal with its background as input and predicts the result, i.e. how the Intervention group compares with the Comparison group in terms of the measured Outcome in the studied Population. While structured clinical evidence is prohibitively expensive for manual collection, we exploit large-scale unstructured sentences from medical literature that implicitly contain PICOs and results as evidence. Specifically, we pre-train a model to predict the disentangled results from such implicit evidence and fine-tune the model with limited data on the downstream datasets. Experiments on the benchmark Evidence Integration dataset show that the proposed model outperforms the baselines by large margins, e.g., with a 10.7% relative gain over BioBERT in macro-F1. Moreover, the performance improvement is also validated on another dataset composed of clinical trials related to COVID-19. | ['Songfang Huang', 'Xiaozhong Liu', 'Mosha Chen', 'Chuanqi Tan', 'Qiao Jin'] | 2020-10-12 | null | https://aclanthology.org/2020.emnlp-main.114 | https://aclanthology.org/2020.emnlp-main.114.pdf | emnlp-2020-11 | ['pico'] | ['natural-language-processing'] | [ 3.97796988e-01 2.99301326e-01 -1.05943274e+00 -4.74644363e-01
-1.13098514e+00 -4.78506297e-01 5.00193357e-01 7.82179177e-01
-4.95487809e-01 1.03163016e+00 5.36499262e-01 -8.37002397e-01
-3.69547337e-01 -4.34829742e-01 -9.18824017e-01 -6.92673981e-01
-8.34157094e-02 6.15977883e-01 -3.01822335e-01 4.16168958e-01
1.54462859e-01 1.25556931e-01 -6.40600324e-01 4.85703498e-01
9.09895897e-01 9.07575250e-01 -8.93864185e-02 4.30287689e-01
4.12096739e-01 6.84285164e-01 -4.38302994e-01 -7.40987837e-01
1.78615144e-03 -5.42149901e-01 -4.17043239e-01 -3.94368351e-01
1.10392757e-01 -2.53553063e-01 -2.36077622e-01 8.64786625e-01
7.24163055e-01 -3.67181599e-01 6.64362311e-01 -8.46522152e-01
-3.18415701e-01 9.72699881e-01 -3.69126439e-01 9.80648249e-02
4.09906544e-02 4.33402270e-01 1.06370878e+00 -6.85584903e-01
9.16804373e-01 7.79720783e-01 4.34596688e-01 3.92851621e-01
-1.27683783e+00 -8.26972544e-01 -6.04356714e-02 1.46654742e-02
-1.04309046e+00 -4.07541245e-01 1.88081071e-01 -4.67490226e-01
6.99855983e-01 4.31081325e-01 5.34748435e-01 1.29423749e+00
6.25486434e-01 4.96803522e-01 8.58627081e-01 -1.29085584e-02
4.23298150e-01 -1.40115350e-01 4.57318157e-01 5.50127149e-01
4.93269145e-01 2.71769881e-01 -2.55582869e-01 -5.26449025e-01
1.75053284e-01 1.55502439e-01 -4.34003383e-01 5.68192713e-02
-1.45091355e+00 6.87536418e-01 4.02677357e-01 -1.32229313e-01
-6.19392395e-01 -6.62959367e-03 5.97773731e-01 -6.86701983e-02
3.85257930e-01 2.55552679e-01 -8.14025223e-01 -6.93492666e-02
-9.70707834e-01 4.13951069e-01 7.75182903e-01 7.86049426e-01
-1.47920474e-01 -7.06094921e-01 -5.28439879e-01 7.40868151e-01
2.89209276e-01 5.27222037e-01 5.88054836e-01 -6.63596630e-01
9.07335877e-01 5.27452528e-01 9.78628993e-02 -5.40607154e-01
-4.86554444e-01 -4.95063514e-01 -1.01442254e+00 -3.83272409e-01
3.15288067e-01 -3.41206789e-01 -1.04772639e+00 1.62787223e+00
4.36019331e-01 3.65073681e-01 1.39584735e-01 9.01732683e-01
1.14596426e+00 4.44367647e-01 5.68700731e-01 -5.31554163e-01
1.63631415e+00 -8.33115280e-01 -9.25872266e-01 -9.44124684e-02
1.02120531e+00 -5.97429991e-01 5.65156877e-01 3.76448661e-01
-9.79169726e-01 1.54288813e-01 -1.04033577e+00 2.69729774e-02
-9.18206647e-02 2.87403733e-01 4.20632035e-01 3.49102139e-01
-3.09681773e-01 7.20427752e-01 -7.73230314e-01 8.82191211e-02
8.54977787e-01 2.47179225e-01 -4.16009933e-01 -2.12673023e-01
-1.17174625e+00 8.02143693e-01 4.74999398e-01 2.64175206e-01
-1.00862157e+00 -1.38174891e+00 -7.05321074e-01 2.32921988e-01
5.78108013e-01 -1.16039121e+00 1.14206576e+00 -7.46637257e-03
-1.37613666e+00 6.57011330e-01 -1.65437669e-01 -6.41066492e-01
5.74900150e-01 -1.11585878e-01 -2.95080900e-01 5.87617643e-02
1.94877058e-01 3.64935637e-01 1.53727382e-01 -4.66726750e-01
-3.71338367e-01 -5.17106831e-01 -3.68829548e-01 -1.54647827e-01
8.25928152e-02 1.57256380e-01 -4.25875217e-01 -8.74512196e-01
-3.62859547e-01 -1.02794302e+00 -7.00997293e-01 -7.79546797e-02
-9.43886042e-01 -1.00899510e-01 1.64088964e-01 -8.25899780e-01
1.31303144e+00 -1.93718195e+00 1.34794116e-01 8.91297832e-02
4.52208132e-01 2.87938744e-01 -8.49908665e-02 1.57449782e-01
-3.78416777e-01 5.70344210e-01 -2.34638035e-01 4.43220548e-02
-2.61164606e-01 -9.88549367e-02 -7.18974471e-02 3.93814594e-01
4.07774866e-01 1.22606242e+00 -9.05322433e-01 -6.43417835e-01
-2.20579818e-01 1.43133000e-01 -8.89957309e-01 2.49394357e-01
-4.36678231e-01 5.23807406e-01 -8.91307354e-01 5.20074725e-01
4.90433753e-01 -7.81622767e-01 6.32995069e-01 -1.57036483e-01
1.90772682e-01 7.05173552e-01 -5.49873292e-01 1.67980099e+00
-1.71642259e-01 7.23759830e-02 -1.94552153e-01 -1.07769275e+00
3.32169175e-01 5.98041475e-01 4.93557036e-01 -3.94310087e-01
1.60388380e-01 3.43352377e-01 5.70959270e-01 -7.81174779e-01
-2.18594462e-01 -4.24061328e-01 1.64030716e-02 3.22341055e-01
-2.79747039e-01 1.81469336e-01 -9.36393440e-02 1.93947613e-01
1.41851056e+00 -2.27843791e-01 7.96797991e-01 1.50404163e-02
2.89315134e-01 3.22892442e-02 9.71140862e-01 6.70369446e-01
-1.03775449e-01 4.92822319e-01 7.72991657e-01 -1.47431329e-01
-6.61239266e-01 -8.49774122e-01 -6.26913548e-01 1.43416271e-01
-3.18801105e-01 -6.46308780e-01 -3.03957224e-01 -1.03249323e+00
8.56938884e-02 6.48538768e-01 -7.34384775e-01 -6.31645098e-02
-3.99705678e-01 -1.31972730e+00 4.97096539e-01 3.18410248e-01
1.43023044e-01 -7.55949438e-01 -3.63349676e-01 3.03436935e-01
-1.54761702e-01 -1.25736010e+00 -5.74892402e-01 4.65407781e-02
-1.22325718e+00 -1.47551358e+00 -7.86311388e-01 -3.42256933e-01
4.02308673e-01 -3.69366020e-01 8.84664953e-01 1.95976626e-02
-4.26487982e-01 -5.41880667e-01 -1.65796280e-01 -6.01158023e-01
-4.82934028e-01 -3.01378876e-01 -3.82622063e-01 -2.71240443e-01
2.78708786e-01 -2.65160859e-01 -1.11598194e+00 5.95626161e-02
-7.02264488e-01 1.84069350e-01 9.64976668e-01 1.13593841e+00
7.85788298e-01 -4.15686429e-01 8.50421727e-01 -1.26506388e+00
5.97862661e-01 -8.92752290e-01 -4.87400323e-01 2.97152072e-01
-1.08821881e+00 9.61486250e-02 3.76242250e-01 -3.06381494e-01
-8.44586194e-01 -2.48871475e-01 -1.09090127e-01 -6.02868237e-02
9.36966240e-02 1.21533930e+00 -1.91784129e-01 5.36433160e-01
4.78450209e-01 -2.11457655e-01 1.07119344e-01 -5.37933230e-01
3.51574391e-01 7.77793646e-01 5.39156012e-02 -2.88437486e-01
7.77835771e-02 1.45128369e-01 2.86674261e-01 -3.15638036e-01
-9.75733459e-01 -3.01358551e-01 6.29801229e-02 4.82365906e-01
8.35348189e-01 -9.88252819e-01 -9.61033642e-01 -6.76829889e-02
-1.10997188e+00 -2.92746574e-01 -9.00283363e-03 9.59243715e-01
-2.23640248e-01 2.02270269e-01 -7.19961643e-01 -2.52823591e-01
-7.23298073e-01 -1.57308435e+00 1.03364444e+00 -1.63818628e-01
-4.39967781e-01 -7.91333735e-01 1.83696613e-01 5.41637957e-01
-8.04640502e-02 4.21394676e-01 1.37602532e+00 -1.10839868e+00
-4.64671075e-01 -4.88661319e-01 -2.63576388e-01 3.01080430e-03
1.27545342e-01 -7.74770230e-02 -5.66888213e-01 1.98811218e-01
-7.63034150e-02 -1.95678160e-01 8.22943747e-01 7.91429877e-01
1.46463919e+00 -5.51674366e-01 -7.16903090e-01 4.30245996e-01
1.05131912e+00 1.98780492e-01 4.05560255e-01 4.28037904e-02
3.90061736e-01 4.69248891e-01 7.35387743e-01 3.77320081e-01
2.31564760e-01 9.50768054e-01 1.07164517e-01 1.83812957e-02
-1.59657642e-01 -3.24375421e-01 4.93747294e-02 5.54592729e-01
2.47385856e-02 -2.36825451e-01 -8.04560900e-01 2.88300365e-01
-1.86580086e+00 -6.81292236e-01 -1.98760048e-01 2.17194271e+00
1.46388507e+00 2.46850833e-01 -1.79056078e-02 -2.35632136e-01
4.99441803e-01 -3.12330693e-01 -7.49611557e-01 -1.74383491e-01
-7.55206542e-03 4.11991656e-01 4.46628004e-01 3.21738362e-01
-8.94610226e-01 4.58756924e-01 6.08271313e+00 9.36581433e-01
-1.01385760e+00 2.21671686e-01 1.10292590e+00 -4.87265676e-01
-7.34997466e-02 -1.21030575e-02 -6.66447103e-01 7.87076175e-01
1.41412771e+00 -4.94020075e-01 -2.34884158e-01 3.67556006e-01
8.48258495e-01 7.22511113e-02 -1.50755858e+00 5.59763789e-01
-3.22052538e-01 -1.71656144e+00 -3.17233317e-02 2.81407803e-01
4.76778477e-01 1.83004797e-01 -5.66735864e-02 5.71611561e-02
3.37082386e-01 -1.13611758e+00 4.29196537e-01 3.09340417e-01
8.52434516e-01 -1.79004669e-01 1.11858797e+00 3.79550427e-01
-3.05393010e-01 6.06803596e-02 -4.17218804e-02 3.14568758e-01
2.61827856e-01 1.08897710e+00 -1.30792189e+00 9.63061333e-01
3.37477893e-01 1.15596783e+00 -3.14374894e-01 1.32124484e+00
-3.83808017e-01 1.07049894e+00 1.00343022e-02 -8.59847814e-02
3.14887874e-02 -2.56737649e-01 4.91386771e-01 1.14301491e+00
3.37139606e-01 4.37033266e-01 -1.28621876e-01 8.73928308e-01
-5.43417215e-01 3.20150286e-01 -4.71957535e-01 -4.87537265e-01
4.01813418e-01 1.08130336e+00 -2.52863437e-01 -5.78678489e-01
-2.40284503e-01 2.65878648e-01 8.06890335e-03 2.27906346e-01
-8.33869815e-01 2.20786911e-02 3.54448617e-01 -2.52696611e-02
9.38298628e-02 6.91520214e-01 -4.09096241e-01 -1.14088440e+00
-3.13515053e-03 -1.18286991e+00 6.84328318e-01 -5.60636401e-01
-1.24985278e+00 3.92970443e-01 -1.11940302e-01 -1.25181043e+00
-2.79147446e-01 -5.02444744e-01 -2.49471083e-01 1.06604648e+00
-1.31185055e+00 -8.52596104e-01 2.46942654e-01 -4.68494073e-02
2.88609385e-01 2.23354381e-02 9.04224634e-01 5.02205372e-01
-1.03487468e+00 7.66625166e-01 -6.46628588e-02 2.10715726e-01
8.76785934e-01 -9.43802059e-01 -3.28619555e-02 2.78182268e-01
-2.15055987e-01 9.39657927e-01 5.84948242e-01 -9.06233966e-01
-1.31947875e+00 -1.17007232e+00 1.09854352e+00 -5.30691028e-01
8.94389510e-01 2.30025612e-02 -7.29530334e-01 6.77996457e-01
-3.61953564e-02 6.28688484e-02 1.12027597e+00 2.04911053e-01
-3.42269272e-01 5.50383478e-02 -1.01484776e+00 8.19111884e-01
8.21184754e-01 -1.76375762e-01 -7.72376060e-01 5.96608639e-01
1.02364910e+00 -4.54121351e-01 -1.30317402e+00 7.32939065e-01
5.94828963e-01 -1.18863776e-01 1.07738137e+00 -1.41076815e+00
1.06191611e+00 -6.37559500e-03 5.39742857e-02 -1.17114902e+00
6.42770976e-02 -4.70979929e-01 1.35766687e-02 7.17382073e-01
1.06447268e+00 -5.75747311e-01 7.85602391e-01 6.55485392e-01
2.41510924e-02 -1.19441080e+00 -9.62937474e-01 -4.40014631e-01
2.74971932e-01 -2.82100469e-01 4.05946225e-01 9.24752772e-01
4.14815128e-01 5.38734555e-01 -6.74343407e-02 1.75770119e-01
5.72914541e-01 4.72601026e-01 3.91830385e-01 -1.06063831e+00
-4.15906876e-01 -4.51822728e-01 -1.59564435e-01 -7.53942728e-01
-2.01504789e-02 -1.29118311e+00 -2.15740204e-02 -1.40623045e+00
9.12123442e-01 -6.58323109e-01 -5.50484240e-01 4.71476912e-01
-7.54609883e-01 -1.90335080e-01 -4.32527959e-02 7.46774226e-02
-4.18830842e-01 4.29337800e-01 1.35507250e+00 -2.96727061e-01
-2.43560001e-02 1.06285378e-01 -8.11036110e-01 4.30718869e-01
5.24860024e-01 -9.93130744e-01 -2.28063777e-01 -4.11312878e-02
3.02674353e-01 6.40988588e-01 5.50308168e-01 -1.46802247e-01
7.42752627e-02 -2.02193677e-01 8.85501504e-02 -4.33616847e-01
-1.02505185e-01 -6.71628833e-01 3.47462058e-01 9.10061240e-01
-9.51993644e-01 -5.99526465e-02 3.18878919e-01 1.03858292e+00
-1.85177267e-01 -2.98575640e-01 1.89964414e-01 2.27906317e-01
3.65305305e-01 4.10025507e-01 -7.20639825e-02 1.76223055e-01
7.96708882e-01 5.62635481e-01 -6.77885234e-01 2.18510162e-02
-8.21952045e-01 3.91367555e-01 -1.18303463e-01 2.61324614e-01
4.98731166e-01 -1.04244101e+00 -1.03856850e+00 -3.20694506e-01
2.33558893e-01 6.02971502e-02 3.11025798e-01 1.36929297e+00
-2.56334394e-01 7.99763083e-01 4.45101947e-01 -4.38808829e-01
-1.35594940e+00 7.28738070e-01 6.51214048e-02 -8.25104117e-01
-7.59457767e-01 3.98589641e-01 3.61779630e-01 -2.28600517e-01
1.66565090e-01 -6.96315885e-01 -7.34438151e-02 -1.40788868e-01
5.22262633e-01 -1.07338607e-01 3.17558885e-01 1.07452869e-01
-4.56907541e-01 2.31656462e-01 -3.28159541e-01 -9.72275808e-02
1.67107105e+00 4.73786205e-01 -3.06441933e-01 1.64222270e-01
1.35480618e+00 1.17382795e-01 -7.40589857e-01 -2.07642570e-01
1.86532378e-01 -8.95416513e-02 2.41438448e-01 -1.38517475e+00
-9.88086462e-01 7.32399225e-01 2.88893580e-01 -4.66451764e-01
8.67854297e-01 -9.54744890e-02 4.39866811e-01 2.77333766e-01
-5.35879396e-02 -5.05234718e-01 -2.86769897e-01 -1.94487303e-01
1.04796338e+00 -1.36730909e+00 4.01839674e-01 -6.12546742e-01
-6.51297092e-01 7.51410902e-01 -8.71892199e-02 3.93045336e-01
9.34994996e-01 1.04998775e-01 -1.93195015e-01 -4.75731164e-01
-1.20984256e+00 4.67136890e-01 3.60653043e-01 -3.25139612e-02
5.49416006e-01 4.01317358e-01 -9.12031591e-01 1.20169282e+00
1.21136278e-01 6.91970646e-01 2.61185259e-01 6.46958590e-01
3.44765693e-01 -1.25719547e+00 -2.19614971e-02 7.96455145e-01
-9.11453605e-01 -6.45509124e-01 -1.59824505e-01 7.18938053e-01
-8.87015983e-02 8.33899260e-01 -4.29314584e-01 -9.75939110e-02
4.76993293e-01 9.00562201e-03 1.86513633e-01 -6.05484784e-01
-8.61043334e-01 2.46899381e-01 4.54282522e-01 -6.64250195e-01
-3.93987179e-01 -6.88674271e-01 -1.21687937e+00 1.95548851e-02
-2.60935187e-01 4.50952351e-01 6.66339934e-01 9.80077803e-01
8.32679331e-01 7.82614350e-01 3.46264750e-01 1.03861272e-01
-8.44901085e-01 -1.02918160e+00 -7.16951862e-02 3.14520419e-01
2.60415494e-01 -4.43270624e-01 -1.55291021e-01 1.67870410e-02] | [8.478293418884277, 8.654912948608398] |
a4859433-c3f3-43ab-b23d-800e68aea858 | neobility-at-semeval-2017-task-1-an-attention | 1703.05465 | null | http://arxiv.org/abs/1703.05465v1 | http://arxiv.org/pdf/1703.05465v1.pdf | Neobility at SemEval-2017 Task 1: An Attention-based Sentence Similarity Model | This paper describes a neural-network model which performed competitively
(top 6) at the SemEval 2017 cross-lingual Semantic Textual Similarity (STS)
task. Our system employs an attention-based recurrent neural network model that
optimizes the sentence similarity. In this paper, we describe our participation
in the multilingual STS task which measures similarity across English, Spanish,
and Arabic. | ['Wenli Zhuang', 'Ernie Chang'] | 2017-03-16 | neobility-at-semeval-2017-task-1-an-attention-1 | https://aclanthology.org/S17-2023 | https://aclanthology.org/S17-2023.pdf | semeval-2017-8 | ['cross-lingual-semantic-textual-similarity'] | ['natural-language-processing'] | [-2.84147933e-02 2.87410878e-02 -7.98763707e-02 -3.86542827e-01
-1.15143609e+00 -2.15664104e-01 7.24130094e-01 5.54214835e-01
-9.31710064e-01 2.58157015e-01 7.94794977e-01 -3.11884671e-01
1.26447394e-01 -2.01503798e-01 -5.09608209e-01 1.98956411e-02
4.58013326e-01 7.91335881e-01 -1.21565960e-01 -8.83110404e-01
6.81275547e-01 1.18647754e-01 -1.08668423e+00 8.06771100e-01
8.88359845e-01 8.10161352e-01 2.48802796e-01 5.91046214e-01
1.17026009e-02 9.82093632e-01 -5.76091528e-01 -6.79862380e-01
-3.24007869e-01 -5.22244453e-01 -1.30793583e+00 -9.69547689e-01
6.67111635e-01 4.33919489e-01 -2.52073020e-01 8.32870662e-01
7.34728873e-01 5.40709019e-01 8.34573209e-01 -7.70500183e-01
-1.34233499e+00 1.32255387e+00 2.66605578e-02 6.21546686e-01
5.39500773e-01 -4.28514272e-01 1.42875230e+00 -1.47971642e+00
7.46122003e-01 1.36814940e+00 1.00014079e+00 4.65443432e-01
-7.70203650e-01 -1.92566216e-01 -3.03661287e-01 8.45734715e-01
-1.40304506e+00 -7.25901663e-01 5.34783721e-01 1.59331232e-01
2.25992799e+00 -8.45495670e-05 3.06924254e-01 1.39113486e+00
4.33488399e-01 9.37305331e-01 6.68963671e-01 -6.99024081e-01
-4.71168756e-02 -1.15863957e-01 2.15590134e-01 3.26296747e-01
-3.05458158e-01 -1.54340133e-01 -6.91952467e-01 1.08187564e-01
-1.68375209e-01 -4.86812681e-01 2.71255933e-02 2.08673254e-01
-9.92893338e-01 1.16686010e+00 5.77851593e-01 9.17807460e-01
-3.49360466e-01 -1.65348854e-02 1.06419027e+00 5.94034195e-01
7.47502685e-01 9.54201400e-01 -4.91438240e-01 -1.43719256e-01
-7.20802188e-01 -1.09185666e-01 7.79017568e-01 7.21378982e-01
2.84354687e-02 3.59970957e-01 -3.85820270e-01 1.44156945e+00
1.71803951e-01 5.09937167e-01 1.30467713e+00 -9.59462404e-01
6.17301762e-01 1.40914246e-01 -4.22943264e-01 -1.16457593e+00
-2.81966656e-01 -1.83187515e-01 -4.69301909e-01 -6.04982018e-01
-2.89963692e-01 2.16570675e-01 -4.08430845e-01 1.52674687e+00
-4.23231006e-01 -2.16139302e-01 6.53519452e-01 6.60669625e-01
1.33484340e+00 7.94358492e-01 3.10181707e-01 4.17468138e-02
1.12504935e+00 -1.34260714e+00 -7.71381497e-01 -3.40102315e-01
1.01422501e+00 -1.12632704e+00 1.16528821e+00 -2.11291805e-01
-1.54871976e+00 -5.37427008e-01 -9.92331564e-01 -4.55971777e-01
-9.73634183e-01 3.52544226e-02 -5.39448895e-02 9.71245617e-02
-1.43987417e+00 7.56541550e-01 -4.76294070e-01 -1.06608319e+00
-1.13669120e-01 1.47967553e-02 -2.18378186e-01 4.81016515e-03
-1.67025387e+00 2.04579234e+00 6.76685691e-01 1.10355690e-01
-3.25184554e-01 -4.65908885e-01 -1.14262927e+00 1.35359332e-01
-1.68193832e-01 -5.33648908e-01 1.49346697e+00 -1.05945158e+00
-1.51500416e+00 1.30167210e+00 -3.14859957e-01 -9.63695168e-01
-2.65891939e-01 -3.19793999e-01 -8.66406798e-01 1.30037904e-01
3.62157255e-01 8.64898503e-01 2.67096460e-01 -5.38589478e-01
-1.19350575e-01 -2.14997008e-01 -4.11238700e-01 6.59837842e-01
-4.06325519e-01 7.29429007e-01 -1.82616450e-02 -9.39574599e-01
-2.38251433e-01 -8.26896787e-01 -5.78379519e-02 -8.32966089e-01
-2.82733105e-02 -9.29363608e-01 5.44285119e-01 -9.88829970e-01
1.05543363e+00 -1.96810329e+00 5.31501293e-01 -5.31751551e-02
-4.19140548e-01 2.58723676e-01 -6.94059670e-01 8.15468311e-01
-3.14554572e-01 1.76932439e-01 -2.28299111e-01 -6.22357786e-01
7.56654739e-02 1.22613169e-01 -3.70511621e-01 1.17081232e-01
2.87019312e-01 1.35538626e+00 -7.86342680e-01 -4.97072548e-01
-2.09095165e-01 4.86706316e-01 -1.95984170e-01 1.76970303e-01
1.22187033e-01 -2.54276931e-01 1.63249657e-01 3.88614625e-01
-9.01510343e-02 -4.70606750e-03 1.25561684e-01 -3.33372295e-01
-1.81923226e-01 8.17518592e-01 1.71929374e-02 2.21706724e+00
-7.20201612e-01 6.66351199e-01 -4.95142877e-01 -1.09439242e+00
9.31223691e-01 4.82883990e-01 1.78078622e-01 -1.33994782e+00
3.80024731e-01 6.60298765e-01 -1.00800894e-01 -3.75849128e-01
1.03365660e+00 2.62415707e-02 -4.62795347e-01 7.60197222e-01
2.65003443e-01 -3.36143970e-01 1.77836716e-01 3.79609764e-01
7.19924212e-01 1.53804552e-02 5.12336075e-01 -6.01110518e-01
6.52704239e-01 -4.28033657e-02 -1.50698006e-01 5.70266962e-01
-2.88411736e-01 5.95571697e-01 7.23871067e-02 -4.91097838e-01
-1.16286623e+00 -9.47325885e-01 2.09483691e-02 1.37339604e+00
-3.39923024e-01 -5.61962783e-01 -5.69451153e-01 -6.26136482e-01
-2.55645812e-01 1.10263765e+00 -7.37547576e-01 -7.05013096e-01
-7.27068961e-01 -2.25853547e-01 9.94170249e-01 5.78104317e-01
2.02718958e-01 -1.67076898e+00 -1.24856912e-01 2.91816384e-01
-7.29269445e-01 -1.05877030e+00 -7.78899014e-01 3.11104208e-01
-4.30359840e-01 -6.03224635e-01 -6.02502704e-01 -1.22881269e+00
-9.74258855e-02 -2.65314225e-02 1.41171288e+00 -3.02844316e-01
-6.66109324e-02 2.36458719e-01 -6.97082162e-01 -3.75657082e-01
-5.76940000e-01 7.03078747e-01 1.59191880e-02 -6.91352546e-01
8.16913068e-01 -8.47329274e-02 1.07693300e-01 -1.19657688e-01
-4.96866167e-01 -5.36516190e-01 2.46310756e-01 8.76715660e-01
2.42789105e-01 -8.37325454e-01 9.72881198e-01 -3.18099260e-01
1.14047706e+00 -7.76810050e-01 -1.05408944e-01 6.65532947e-01
-6.17368340e-01 2.45889977e-01 6.30920291e-01 -1.42027577e-02
-8.17459464e-01 -3.34755898e-01 -4.62686896e-01 -2.72683263e-01
2.11853996e-01 1.07946038e+00 4.22372639e-01 4.28254995e-03
8.31855893e-01 -5.76831726e-03 -1.02707237e-01 -3.61833274e-01
4.28489387e-01 9.29029703e-01 6.19232535e-01 -4.54679132e-01
-8.85736477e-03 -3.30932945e-01 -4.39213634e-01 -6.00178301e-01
-1.15529168e+00 -3.13153028e-01 -5.91480017e-01 -4.49403040e-02
9.38363254e-01 -7.84178138e-01 -4.39664245e-01 1.96668521e-01
-1.60875642e+00 -1.02040820e-01 -2.85226882e-01 5.04823685e-01
-5.56602955e-01 1.64640740e-01 -1.00837684e+00 -1.69626679e-02
-1.02596903e+00 -8.03964376e-01 9.61783707e-01 -7.89549500e-02
-8.17815661e-01 -1.38588762e+00 6.45953417e-01 2.43572667e-01
1.03119385e+00 -3.62861335e-01 8.87463748e-01 -1.18423009e+00
2.55144238e-01 2.64272001e-02 -1.82091400e-01 2.34014824e-01
-2.61623710e-01 -1.03623681e-01 -6.46348953e-01 -1.99470192e-01
-1.26848131e-01 -8.77146661e-01 9.19829547e-01 3.74216110e-01
9.20275271e-01 -1.72267631e-01 2.25291148e-01 3.49054843e-01
1.11897004e+00 1.36753879e-02 3.61450970e-01 5.72652340e-01
5.61627567e-01 8.09405148e-01 3.81088078e-01 -5.68859912e-02
6.33759856e-01 9.53066409e-01 -6.52375296e-02 1.31392568e-01
-1.95526451e-01 1.10802352e-02 6.49632335e-01 1.81286585e+00
2.02165127e-01 -4.05292988e-01 -1.10391748e+00 7.42364109e-01
-1.96490479e+00 -9.60267961e-01 -1.65718962e-02 1.56337655e+00
8.40662539e-01 -3.46451491e-01 -8.46226737e-02 -3.14618051e-01
6.02897346e-01 1.67493209e-01 -1.17891699e-01 -1.58527780e+00
-5.85742950e-01 6.57458365e-01 8.22205991e-02 7.19609976e-01
-9.48935807e-01 1.56981444e+00 7.22737837e+00 9.32605982e-01
-9.37305868e-01 4.65048105e-01 2.25214899e-01 -8.23894665e-02
-3.43317240e-01 -4.44156408e-01 -5.71296334e-01 1.76673979e-01
1.75190699e+00 -5.32049119e-01 3.71741295e-01 6.24394536e-01
-2.22995564e-01 1.76741451e-01 -8.73979568e-01 7.00839758e-01
1.00003815e+00 -1.25921822e+00 2.43185237e-01 -7.06145287e-01
7.48812437e-01 9.03226554e-01 2.88463473e-01 6.05707169e-01
4.83307451e-01 -1.31438065e+00 5.10742784e-01 4.19802070e-01
5.14192104e-01 -1.12364256e+00 1.12661672e+00 -2.86293268e-01
-9.79297101e-01 3.12725693e-01 -3.47735822e-01 3.66176605e-01
1.53164834e-01 -1.78503931e-01 -8.13663900e-01 5.11760294e-01
8.74944568e-01 1.43112028e+00 -9.13981020e-01 5.09639978e-01
-2.98867285e-01 2.83175260e-01 1.27371371e-01 -3.26999694e-01
5.69579720e-01 6.87373728e-02 5.00826299e-01 1.68713069e+00
2.99649924e-01 -2.62289405e-01 -1.12104900e-01 6.67458594e-01
-3.82901967e-01 6.79346800e-01 -9.08412576e-01 -2.70921260e-01
4.30823565e-01 9.17067051e-01 -4.29939181e-01 -3.90079558e-01
-1.54002652e-01 1.33457363e+00 8.31898928e-01 2.10859269e-01
-5.54698229e-01 -8.19083869e-01 2.28387177e-01 -6.93535805e-01
3.21394265e-01 1.62059702e-02 -4.07905191e-01 -1.09784973e+00
-2.14654922e-01 -8.31091940e-01 6.19195759e-01 -1.22201943e+00
-1.87335801e+00 1.25820565e+00 -2.67600298e-01 -7.20351577e-01
-6.81938350e-01 -4.48744386e-01 -4.30401087e-01 1.00614309e+00
-1.25870085e+00 -1.11734009e+00 3.11937362e-01 6.93809032e-01
7.07349837e-01 -7.35125303e-01 1.17041731e+00 2.92935491e-01
-4.11434323e-01 9.36711550e-01 2.58177876e-01 2.30640452e-02
1.08405507e+00 -8.97023261e-01 8.89312923e-01 5.51436245e-01
1.87033489e-01 5.41073680e-01 4.59023595e-01 -5.73916376e-01
-6.71049118e-01 -1.02749479e+00 2.21450734e+00 -5.89814961e-01
1.19471860e+00 1.44351602e-01 -1.02282107e+00 6.50682211e-01
1.22598934e+00 -4.49031979e-01 6.41565144e-01 2.54058540e-01
-7.17297435e-01 1.72599182e-01 -8.02511752e-01 7.25980222e-01
7.50377536e-01 -1.32318652e+00 -1.17649817e+00 4.32456523e-01
1.07180429e+00 -1.57603905e-01 -1.09937680e+00 4.86453176e-01
3.69415581e-01 -3.37334961e-01 1.02079523e+00 -1.09515512e+00
8.41849327e-01 2.60575533e-01 -4.93093282e-01 -1.86770856e+00
-2.50213176e-01 -1.30598739e-01 1.81076050e-01 1.00303590e+00
6.45446718e-01 -4.34538245e-01 4.68652099e-02 5.17343404e-04
-7.31938541e-01 -4.94899988e-01 -1.12686622e+00 -9.37810600e-01
5.19107223e-01 -2.46192962e-01 1.84489816e-01 1.35440135e+00
5.39192498e-01 8.69387865e-01 2.86443438e-02 -5.09214759e-01
4.70880233e-02 -3.11570168e-01 -3.19035858e-01 -8.11214507e-01
1.16083831e-01 -9.93751645e-01 -2.43640229e-01 -2.69919753e-01
1.05370104e+00 -1.51365936e+00 1.48338333e-01 -1.47098398e+00
2.09960788e-01 2.66965061e-01 -5.18556237e-01 5.40760040e-01
-1.42203167e-01 3.79660636e-01 1.65857926e-01 1.36399046e-01
-8.77964675e-01 9.06877160e-01 2.32116148e-01 -2.56337225e-01
1.71187997e-01 -6.02785110e-01 -4.11254406e-01 2.19587833e-01
1.17222369e+00 -7.39555478e-01 1.82621419e-01 -9.47197735e-01
4.40201581e-01 -2.03618065e-01 -6.67545758e-03 -6.59790993e-01
3.99517953e-01 2.93058455e-01 -5.50511815e-02 -4.62515861e-01
3.40930343e-01 -4.04540479e-01 -3.17666799e-01 6.73240483e-01
-9.95872319e-01 9.48123038e-01 3.43662620e-01 -1.57194957e-01
-5.80769658e-01 -4.63734537e-01 7.39470720e-01 -1.48029238e-01
-4.79261667e-01 -1.47553965e-01 -8.24690819e-01 4.93533939e-01
4.61280525e-01 2.08616614e-01 -5.10816813e-01 -3.73848736e-01
-6.87690377e-01 1.20256156e-01 2.31544152e-01 1.09532547e+00
6.78377151e-01 -1.72294331e+00 -1.34183574e+00 -1.45768136e-01
5.52244663e-01 -1.04812145e+00 -1.41091803e-02 8.35325897e-01
-1.74644738e-01 9.01553690e-01 -4.23000902e-01 -2.01897681e-01
-1.20073438e+00 4.11828160e-01 4.29985434e-01 -2.14511558e-01
-1.36460796e-01 8.69043946e-01 -6.89117074e-01 -8.65399241e-01
1.42183788e-02 3.22738975e-01 -6.82061911e-01 8.78531039e-02
3.54130477e-01 4.86502379e-01 4.76325035e-01 -1.13556445e+00
-4.05890018e-01 6.38565302e-01 -2.84835041e-01 -4.43131328e-01
1.18979621e+00 -1.63809299e-01 -5.37465811e-01 8.13267946e-01
1.85750604e+00 -4.83240575e-01 1.09786868e-01 -4.32572335e-01
6.30743623e-01 3.13112170e-01 -1.72314923e-02 -8.61155808e-01
-6.98586881e-01 9.00330484e-01 3.05532366e-01 -8.87639448e-02
5.95883787e-01 1.46587223e-01 1.07130086e+00 9.10203457e-01
-2.61884242e-01 -1.64422607e+00 1.84982166e-01 1.70934069e+00
1.20901489e+00 -1.34633458e+00 -3.92727464e-01 4.06213105e-01
-9.63167429e-01 1.27617002e+00 5.63723743e-01 -2.85257995e-01
4.20559883e-01 -1.12360336e-01 -3.64221558e-02 -4.42072093e-01
-9.87515211e-01 -1.03346333e-01 7.43811429e-01 3.72080296e-01
9.38925028e-01 -1.66012987e-01 -5.79399765e-01 3.71822417e-01
-4.85166728e-01 -4.52642113e-01 2.44759962e-01 6.98995292e-01
-4.18598711e-01 -9.14489925e-01 5.36845475e-02 9.68270469e-03
-5.22887707e-01 -8.79098356e-01 -9.81040359e-01 4.35068727e-01
-4.76628065e-01 1.05663073e+00 3.40495080e-01 -2.46293172e-01
3.24894190e-01 2.93164015e-01 3.84609073e-01 -5.53611934e-01
-1.32249141e+00 -4.27527279e-01 5.14015377e-01 -8.54315639e-01
-4.44394797e-01 -7.76338041e-01 -1.15032327e+00 -2.72707224e-01
5.14520928e-02 5.02136886e-01 8.98251235e-01 1.15565038e+00
4.35450107e-01 3.40732396e-01 5.85435867e-01 -5.02596021e-01
-2.93332100e-01 -1.39207530e+00 -4.40354422e-02 1.97640434e-01
-2.21475512e-01 -1.07477251e-02 -3.14769506e-01 -2.49454379e-01] | [10.890318870544434, 9.683730125427246] |
81fcdb91-2ed2-485d-835d-c0fc68a6f1b5 | language-family-relationship-preserved-in-non | null | null | https://aclanthology.org/C14-1183 | https://aclanthology.org/C14-1183.pdf | Language Family Relationship Preserved in Non-native English | null | ['Ryo Nagata'] | 2014-08-01 | language-family-relationship-preserved-in-non-1 | https://aclanthology.org/C14-1183 | https://aclanthology.org/C14-1183.pdf | coling-2014-8 | ['grammatical-error-detection'] | ['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] | [-7.185693264007568, 3.9062962532043457] |
13ab6dbf-8999-438a-86ea-118148dc5b17 | learning-semantic-correspondences-from-noisy | null | null | https://aclanthology.org/2020.coling-main.272 | https://aclanthology.org/2020.coling-main.272.pdf | Learning Semantic Correspondences from Noisy Data-text Pairs by Local-to-Global Alignments | Learning semantic correspondences between structured input data (e.g., slot-value pairs) and associated texts is a core problem for many downstream NLP applications, e.g., data-to-text generation. Large-scale datasets recently proposed for generation contain loosely corresponding data text pairs, where part of spans in text cannot be aligned to its incomplete paired input. To learn semantic correspondences from such datasets, we propose a two-stage local-to-global alignment (L2GA) framework. First, a local model based on multi-instance learning is applied to build alignments for texts spans that can be directly grounded to its paired structured input. Then, a novel global model built upon a memory-guided conditional random field (CRF) layer aims to infer missing alignments for text spans which not supported by paired incomplete inputs, where the memory is designed to leverage alignment clues provided by the local model to strengthen the global model. In this way, the local model and global model can work jointly to learn semantic correspondences in the same framework. Experimental results show that our proposed method can be generalized to both restaurant and computer domains and improve the alignment accuracy. | ['Chin-Yew Lin', 'Jinpeng Wang', 'Feng Nie'] | 2020-12-01 | null | null | null | coling-2020-8 | ['data-to-text-generation'] | ['natural-language-processing'] | [ 5.03378034e-01 4.98897642e-01 -3.71801347e-01 -5.80330014e-01
-1.17015493e+00 -4.56712991e-01 6.19632125e-01 3.85758817e-01
-8.47252235e-02 9.56394076e-01 5.83427489e-01 -1.98223248e-01
-4.58713025e-02 -1.03245795e+00 -9.69134092e-01 -3.80201817e-01
7.12448657e-01 9.02297199e-01 1.64837763e-01 -3.53576064e-01
2.69788951e-01 -4.03770506e-01 -1.27740824e+00 7.93197215e-01
1.22839224e+00 6.03499413e-01 6.22709095e-01 2.55709589e-01
-1.04201698e+00 6.13159239e-01 -2.93037832e-01 -3.65194291e-01
1.18811153e-01 -7.06947386e-01 -1.17134476e+00 4.14021984e-02
1.74448207e-01 -9.76507440e-02 7.16895908e-02 9.40442204e-01
2.55998492e-01 2.76451081e-01 4.07165110e-01 -1.24674761e+00
-6.72616601e-01 1.33592272e+00 -3.83738339e-01 -1.26708329e-01
4.21154231e-01 -2.49849573e-01 1.31382179e+00 -9.56091225e-01
5.75589895e-01 1.38674080e+00 3.86748552e-01 4.68976259e-01
-1.12144351e+00 -5.82254648e-01 4.64487731e-01 -1.10113181e-01
-9.56243157e-01 -3.99893582e-01 7.30157673e-01 -1.54076129e-01
8.77552867e-01 6.15162998e-02 3.32804084e-01 1.00143218e+00
-2.68547118e-01 1.00954890e+00 7.49744058e-01 -6.78587139e-01
1.02089308e-01 5.32125980e-02 1.11973248e-01 5.03753364e-01
4.75876555e-02 -3.26218426e-01 -7.93769419e-01 -1.42965555e-01
6.07274532e-01 -1.79426726e-02 6.70185164e-02 -1.45665966e-02
-1.50850403e+00 7.59605825e-01 3.55444163e-01 4.67233956e-01
-1.87339753e-01 -1.39367312e-01 3.32481742e-01 1.46543771e-01
5.99383891e-01 4.40980881e-01 -7.08564699e-01 -3.02838255e-03
-9.30191159e-01 4.96314406e-01 7.16000736e-01 1.49063420e+00
1.14033389e+00 -5.30889571e-01 -4.89593625e-01 8.84679973e-01
5.83832502e-01 4.71974939e-01 6.29386783e-01 -3.52925450e-01
1.39124823e+00 8.30906987e-01 2.51192451e-01 -8.81678104e-01
-2.40004286e-01 -7.13573396e-02 -7.40928888e-01 -8.03114414e-01
4.67670828e-01 -1.56684026e-01 -6.24958098e-01 1.89660513e+00
5.24565041e-01 4.53563929e-01 3.06369156e-01 7.51348019e-01
8.84354711e-01 6.10014379e-01 1.08031496e-01 -5.11950292e-02
1.22157645e+00 -1.30456567e+00 -5.65943420e-01 -6.02190614e-01
1.04924035e+00 -9.90361869e-01 1.02546775e+00 -3.56258571e-01
-9.89974737e-01 -7.71347344e-01 -6.75645173e-01 -4.19918209e-01
-3.15034598e-01 1.34952724e-01 1.74583107e-01 1.14157729e-01
-6.40164196e-01 5.05787015e-01 -7.01435268e-01 -4.74205077e-01
1.91722140e-01 4.94350353e-03 -1.24418780e-01 -1.82690799e-01
-1.32783496e+00 6.66221142e-01 8.87708306e-01 1.20986044e-01
-3.17878693e-01 -6.68665409e-01 -1.04315662e+00 2.01549046e-02
3.63740295e-01 -9.76039648e-01 1.29393351e+00 -7.00870812e-01
-1.33571768e+00 6.67150140e-01 -6.10207617e-01 -3.67037654e-01
3.70598495e-01 -4.04448777e-01 -2.53323436e-01 -1.57451078e-01
5.28262556e-01 7.61173904e-01 5.32764673e-01 -1.13522351e+00
-9.31045353e-01 -4.82002616e-01 -1.80314451e-01 4.10922527e-01
-1.90001503e-01 -1.71036959e-01 -2.67063737e-01 -6.95032477e-01
4.13241178e-01 -6.17189288e-01 -3.71027410e-01 -4.77627039e-01
-7.63481319e-01 -4.91619051e-01 4.09960210e-01 -8.41159284e-01
1.28783548e+00 -1.59462738e+00 3.78312796e-01 2.06203088e-01
-1.75118476e-01 -4.46190126e-02 -4.49603289e-01 8.24945629e-01
7.15382695e-02 3.39586809e-02 -4.00189757e-01 -6.44707203e-01
2.15261102e-01 2.59305418e-01 -8.02597106e-01 -2.78788447e-01
3.35925519e-01 1.16885698e+00 -1.07117391e+00 -6.08144343e-01
-2.61533931e-02 -1.41061410e-01 -3.69288445e-01 5.86320758e-01
-7.04695046e-01 7.42481649e-01 -6.34290695e-01 3.35573912e-01
6.06388867e-01 -3.56826663e-01 3.96443576e-01 -1.95695341e-01
1.29152477e-01 7.01434851e-01 -1.06118822e+00 2.45745277e+00
-6.53683007e-01 -1.53223008e-01 -3.36385369e-01 -1.21790242e+00
1.33744979e+00 3.43717933e-01 2.61483878e-01 -6.82575643e-01
-2.02951357e-01 4.11171347e-01 -3.76139849e-01 -5.42655706e-01
8.40033948e-01 -1.85313180e-01 -3.25658113e-01 6.79891944e-01
2.01837510e-01 -7.04819188e-02 1.61911532e-01 2.52859712e-01
7.85744011e-01 4.35774446e-01 2.69332230e-01 -1.07130684e-01
5.51281214e-01 1.37721226e-02 6.44310534e-01 6.25729740e-01
3.42063874e-01 3.89269471e-01 2.52945006e-01 -2.19586194e-01
-1.31062961e+00 -8.50452721e-01 1.91232175e-01 1.18460274e+00
4.28288221e-01 -6.22576952e-01 -7.76914001e-01 -9.37829018e-01
-1.02751307e-01 7.08084166e-01 -2.99470633e-01 -5.91592528e-02
-6.99653864e-01 -4.20630276e-01 4.31191385e-01 8.83664727e-01
3.48551154e-01 -1.12674594e+00 -5.16659170e-02 6.88286781e-01
-7.57209182e-01 -1.25310922e+00 -5.48553944e-01 4.96358350e-02
-1.09909701e+00 -8.27168226e-01 -4.37704533e-01 -8.94263804e-01
7.99293101e-01 2.55835533e-01 1.21629655e+00 7.56742135e-02
2.95726359e-01 -1.63616873e-02 -5.91266692e-01 -1.77065417e-01
-5.19567609e-01 4.89060760e-01 -3.40588629e-01 -1.05762847e-01
6.15442097e-01 -5.72781086e-01 -3.12333375e-01 3.16325605e-01
-9.48079944e-01 7.17429876e-01 6.70499861e-01 1.16117764e+00
6.15187883e-01 -4.25814360e-01 9.85466957e-01 -1.10165226e+00
5.49119532e-01 -9.35538232e-01 -3.08980286e-01 5.67255259e-01
-3.54008228e-01 3.77665579e-01 7.24399924e-01 -1.67711005e-01
-1.35143435e+00 8.42039883e-02 -1.37721837e-01 -1.65587455e-01
-3.35892051e-01 8.74499023e-01 -5.81406593e-01 9.70000565e-01
3.33545744e-01 2.72179544e-01 -2.67803162e-01 -7.17368603e-01
7.69683182e-01 7.83664107e-01 6.20417595e-01 -1.04175460e+00
7.21997321e-01 1.16995335e-01 -3.32233489e-01 -3.36477339e-01
-1.23630846e+00 -6.59198403e-01 -9.87037838e-01 3.19342792e-01
7.75483012e-01 -1.00554633e+00 1.94579195e-02 3.44864368e-01
-1.39176559e+00 -3.06165248e-01 -1.95365608e-01 3.49584937e-01
-7.25795150e-01 2.64037549e-01 -4.37271059e-01 -5.94286740e-01
-4.98632163e-01 -8.04656565e-01 1.51402497e+00 3.72777790e-01
-3.15865457e-01 -1.32100272e+00 1.64124787e-01 6.55461967e-01
1.42919734e-01 -1.55523568e-01 1.06890011e+00 -1.04238117e+00
-5.45878112e-01 1.47918463e-01 -3.15015078e-01 -7.47658536e-02
4.30479616e-01 -4.27018106e-01 -7.57160544e-01 -6.71464130e-02
-3.63645136e-01 -4.86641675e-01 7.06031859e-01 -1.24885188e-02
1.01485252e+00 -5.31452119e-01 -5.71472585e-01 1.94518194e-01
1.14636123e+00 -1.44990966e-01 5.09228349e-01 1.51993573e-01
7.77897775e-01 7.93312132e-01 1.01832414e+00 4.69278932e-01
6.64324582e-01 6.53255999e-01 1.69738367e-01 -2.58825943e-02
5.64332306e-02 -1.04883087e+00 2.33405128e-01 1.10147130e+00
5.13151944e-01 -2.61578798e-01 -7.11635530e-01 7.85449505e-01
-2.40058112e+00 -8.88906777e-01 -1.44705445e-01 2.12537718e+00
1.22450733e+00 3.63009647e-02 -1.05590276e-01 -1.04083613e-01
1.00261807e+00 1.22449271e-01 -4.39604163e-01 -1.93657316e-02
7.61819780e-02 2.84249723e-01 -1.28682703e-01 4.33392674e-01
-8.64066184e-01 1.14967191e+00 4.90185213e+00 1.06770682e+00
-5.95302045e-01 6.66833073e-02 4.89005178e-01 3.27419579e-01
-9.59689796e-01 4.06771481e-01 -1.15688682e+00 7.35284567e-01
7.93469548e-01 -3.66121411e-01 2.04683751e-01 7.58655190e-01
2.14046299e-01 1.58321247e-01 -1.34927273e+00 6.88051403e-01
-5.48438765e-02 -1.48397028e+00 2.47274280e-01 -1.89209819e-01
8.44445467e-01 -3.53944629e-01 -2.83756375e-01 3.58489931e-01
5.50273597e-01 -9.22954619e-01 6.60380721e-01 5.85735559e-01
7.14963257e-01 -6.61100984e-01 7.32667565e-01 7.24042773e-01
-1.51519036e+00 1.13733552e-01 -6.16359055e-01 1.22162014e-01
4.38817501e-01 7.44976163e-01 -1.06217623e+00 9.73120272e-01
2.05388904e-01 9.70939815e-01 -1.41338393e-01 6.16041601e-01
-4.04154271e-01 4.04689550e-01 -1.41479731e-01 6.05748817e-02
2.73645431e-01 -3.69870335e-01 3.16750914e-01 9.80417609e-01
5.67082226e-01 4.24231514e-02 5.85128486e-01 1.22053599e+00
-2.26139769e-01 3.51526618e-01 -6.14232123e-01 -1.14824235e-01
8.98400187e-01 1.32584286e+00 -4.71863002e-01 -4.97002512e-01
-4.97265548e-01 1.02595615e+00 7.14298189e-01 1.88103110e-01
-6.27419233e-01 -3.88074428e-01 4.90429521e-01 -1.25244185e-01
1.42127186e-01 2.03515161e-02 -4.34510320e-01 -1.33307588e+00
2.93962747e-01 -6.70465231e-01 5.26686490e-01 -7.28687584e-01
-1.67151463e+00 4.39146429e-01 -1.18233025e-01 -1.21699095e+00
-5.20547032e-01 -3.06625124e-02 -8.52951467e-01 1.27801931e+00
-1.54450905e+00 -1.63545930e+00 -2.77545154e-01 6.39765024e-01
7.74078608e-01 -9.56896320e-02 8.88646066e-01 1.25687355e-02
-4.26098436e-01 6.04352653e-01 -1.32164918e-02 3.72292876e-01
8.54936898e-01 -1.19973409e+00 5.90528488e-01 9.99564171e-01
4.49705333e-01 7.21225858e-01 3.27093750e-01 -9.64441657e-01
-1.04393947e+00 -1.52102518e+00 1.43936718e+00 -4.32052672e-01
6.23627484e-01 -2.78644621e-01 -1.09765589e+00 9.42545116e-01
2.39274666e-01 -2.14311257e-01 8.74695480e-01 2.42811441e-01
-3.86538774e-01 1.28512800e-01 -7.65702248e-01 4.14528280e-01
1.23307705e+00 -6.46475852e-01 -1.00030553e+00 3.61878633e-01
1.16456747e+00 -7.48216093e-01 -8.21467161e-01 9.96181667e-02
9.88552943e-02 -5.11277318e-01 6.41081572e-01 -1.06191111e+00
9.14830446e-01 -3.11738640e-01 -2.73039043e-01 -1.41662848e+00
2.39198399e-03 -6.37856245e-01 -1.02235578e-01 1.76873302e+00
6.56262100e-01 -3.46058726e-01 6.35031223e-01 5.66109776e-01
-5.11096418e-01 -5.33617914e-01 -7.76570439e-01 -6.57801509e-01
3.03880066e-01 -3.87092650e-01 1.14337397e+00 9.98696744e-01
4.79135275e-01 7.72067487e-01 -3.55829209e-01 1.10826485e-01
3.64596128e-01 7.98245549e-01 8.76505971e-01 -1.05266261e+00
-3.68556738e-01 -4.16726694e-02 2.36256182e-01 -1.64084160e+00
4.30772752e-01 -1.41657770e+00 3.16758960e-01 -1.68097222e+00
4.19761211e-01 -9.69955325e-01 -1.41270339e-01 6.87027216e-01
-8.00584018e-01 -2.79374808e-01 1.51652947e-01 2.08258420e-01
-6.32448137e-01 7.90935576e-01 1.13287258e+00 -5.42334914e-02
-2.97725767e-01 -3.35060135e-02 -9.60690975e-01 3.25596422e-01
8.24503481e-01 -5.72427094e-01 -4.61328179e-01 -6.88530266e-01
3.44753534e-01 4.21460271e-01 8.30616578e-02 -6.58108056e-01
4.64348972e-01 -4.29378122e-01 1.65749907e-01 -6.50514424e-01
-1.69498309e-01 -5.81611454e-01 -1.22738048e-01 -4.56420295e-02
-7.59043992e-01 -8.98960158e-02 -6.44291118e-02 6.71542704e-01
-4.20582235e-01 -3.82569730e-01 3.26689750e-01 -1.61874905e-01
-4.89910394e-01 3.26243341e-01 2.44275540e-01 4.59145993e-01
7.22009242e-01 1.49522737e-01 -5.54729462e-01 -8.97243842e-02
-6.55884087e-01 5.62887967e-01 2.01924071e-01 7.10272908e-01
4.58510756e-01 -1.65386164e+00 -8.60496759e-01 1.69362754e-01
4.80132520e-01 6.44896567e-01 2.38232136e-01 7.41570055e-01
2.05183372e-01 5.66362739e-01 -8.42431039e-02 -6.01838529e-01
-9.97029543e-01 4.34834570e-01 -7.07519427e-02 -7.83166468e-01
-4.84694004e-01 7.35039711e-01 7.45495260e-02 -8.31872523e-01
-1.38827279e-01 -4.54271376e-01 -5.04839979e-02 1.01415990e-02
6.29512668e-01 1.42110616e-01 1.31325588e-01 -4.91542786e-01
-3.44721340e-02 3.44075531e-01 -1.87871546e-01 -2.53915757e-01
1.26436305e+00 -4.92538422e-01 -2.62163669e-01 4.15337920e-01
9.29299831e-01 -6.25828207e-02 -9.62420702e-01 -8.66326511e-01
4.04070795e-01 -4.47534591e-01 -5.18016875e-01 -6.63565576e-01
-8.05266142e-01 1.02848494e+00 -2.06356987e-01 -2.17618853e-01
8.84576917e-01 1.46235690e-01 1.28278971e+00 3.33947986e-01
3.25333357e-01 -1.08386135e+00 3.01782012e-01 7.36655653e-01
7.20599651e-01 -1.22257841e+00 -6.74616873e-01 -5.49490869e-01
-6.34235919e-01 1.19482815e+00 9.27900672e-01 3.19432914e-01
2.90182233e-01 1.38545841e-01 -1.11022674e-01 1.58686042e-01
-8.71976733e-01 -3.20327699e-01 2.33668804e-01 6.86815500e-01
4.95651245e-01 9.39218886e-03 -2.83865869e-01 1.22247171e+00
-2.68010736e-01 5.84273674e-02 8.50195810e-02 8.46243739e-01
-5.96548736e-01 -1.64751792e+00 -1.20720856e-01 4.52013612e-01
5.84888458e-02 -4.14427459e-01 -2.45503515e-01 2.12177947e-01
1.92185700e-01 1.07383049e+00 3.55204046e-01 -2.98481017e-01
3.57560404e-02 5.00308573e-01 2.57851422e-01 -1.02512312e+00
-5.18957973e-01 6.81308582e-02 1.18681811e-01 -4.23617452e-01
-5.26627779e-01 -4.99427050e-01 -1.61224866e+00 -9.37335268e-02
-3.63463700e-01 4.33467507e-01 3.24566633e-01 1.35102797e+00
5.09995341e-01 2.99857527e-01 5.91235757e-01 -4.64314669e-01
-5.15921772e-01 -1.14572942e+00 -3.98763597e-01 7.72296906e-01
-8.74958113e-02 -2.68901557e-01 1.19549274e-01 4.15218532e-01] | [11.579787254333496, 8.836142539978027] |
c4054e4f-86ea-407f-a944-200a89cecdb9 | multilingual-models-for-compositional | 1404.4641 | null | http://arxiv.org/abs/1404.4641v1 | http://arxiv.org/pdf/1404.4641v1.pdf | Multilingual Models for Compositional Distributed Semantics | We present a novel technique for learning semantic representations, which
extends the distributional hypothesis to multilingual data and joint-space
embeddings. Our models leverage parallel data and learn to strongly align the
embeddings of semantically equivalent sentences, while maintaining sufficient
distance between those of dissimilar sentences. The models do not rely on word
alignments or any syntactic information and are successfully applied to a
number of diverse languages. We extend our approach to learn semantic
representations at the document level, too. We evaluate these models on two
cross-lingual document classification tasks, outperforming the prior state of
the art. Through qualitative analysis and the study of pivoting effects we
demonstrate that our representations are semantically plausible and can capture
semantic relationships across languages without parallel data. | ['Karl Moritz Hermann', 'Phil Blunsom'] | 2014-04-17 | multilingual-models-for-compositional-1 | https://aclanthology.org/P14-1006 | https://aclanthology.org/P14-1006.pdf | acl-2014-6 | ['learning-semantic-representations', 'cross-lingual-document-classification'] | ['methodology', 'natural-language-processing'] | [-2.75373667e-01 2.45197546e-02 -5.65633059e-01 -5.31995654e-01
-8.09669733e-01 -9.64224696e-01 1.07857418e+00 6.36219740e-01
-6.26362443e-01 4.19553876e-01 9.62039530e-01 -5.16125679e-01
-8.68029296e-02 -6.99644923e-01 -4.14896697e-01 -1.28763184e-01
9.85745192e-02 5.09016156e-01 -1.10430062e-01 -4.21865284e-01
2.61622578e-01 1.91776767e-01 -1.21850419e+00 2.33644649e-01
8.07255864e-01 3.38012904e-01 1.99383199e-01 1.52806759e-01
-5.31695247e-01 5.28519869e-01 -3.83841962e-01 -5.18452525e-01
-1.73480082e-02 -1.58743694e-01 -9.59018946e-01 -1.79403588e-01
6.73293412e-01 -6.36030734e-02 -4.36917722e-01 9.04211879e-01
4.07712042e-01 7.03328475e-02 9.43438232e-01 -9.91119683e-01
-1.71397197e+00 6.42243266e-01 -2.34513864e-01 2.86190659e-01
4.64429438e-01 -4.60688442e-01 1.65219450e+00 -1.03553557e+00
8.43240857e-01 1.55958009e+00 8.58346641e-01 4.29501891e-01
-1.42501974e+00 -4.35024053e-01 3.63414705e-01 2.38738671e-01
-1.32380629e+00 -3.56515288e-01 5.84963202e-01 -6.35359168e-01
1.29560828e+00 -2.70472169e-01 2.28891134e-01 1.28249240e+00
7.52605051e-02 6.25186265e-01 9.96796727e-01 -7.54725873e-01
-5.87233827e-02 1.70220032e-01 3.07773501e-01 6.68380976e-01
3.14165801e-01 4.35365848e-02 -3.75090241e-01 -2.42203638e-01
5.04439771e-01 1.31829947e-01 -1.02985330e-01 -4.87691045e-01
-1.14060867e+00 1.40559912e+00 5.25406361e-01 6.82678223e-01
-4.30261567e-02 1.85099229e-01 5.19175589e-01 2.93276578e-01
7.88401186e-01 5.84194422e-01 -7.07072496e-01 1.20520368e-01
-7.39608109e-01 2.20009968e-01 6.84778214e-01 9.52410579e-01
7.43705630e-01 -2.29758531e-01 2.15488747e-01 1.15443814e+00
4.60103989e-01 4.07947004e-01 8.99960399e-01 -7.25662947e-01
4.22828019e-01 4.65347648e-01 -1.09529585e-01 -1.07085848e+00
-2.95190126e-01 -9.75745395e-02 -1.98304933e-02 -3.19619715e-01
1.60067946e-01 7.35013932e-03 -5.82460463e-01 1.96817732e+00
-3.61967534e-02 -1.98574290e-01 3.80986840e-01 6.08911574e-01
6.36226833e-01 5.13380766e-01 4.24999118e-01 2.42369995e-01
1.45214510e+00 -9.13173437e-01 -6.52387500e-01 -3.03482205e-01
1.21061373e+00 -8.12248349e-01 1.28429973e+00 -2.55237341e-01
-8.17793548e-01 -5.10746360e-01 -1.01282895e+00 -6.72382593e-01
-8.16367984e-01 -4.26189937e-02 7.68006980e-01 5.56540072e-01
-1.23425245e+00 3.94096494e-01 -6.76936567e-01 -8.19078684e-01
1.78014830e-01 -2.10971087e-01 -5.78528285e-01 -2.97051460e-01
-1.38167953e+00 1.43879044e+00 3.17343771e-01 -5.25723100e-01
-5.04254222e-01 -6.96910203e-01 -1.39376032e+00 -5.48935421e-02
-1.96954370e-01 -7.21200526e-01 1.16525006e+00 -1.02778304e+00
-9.24592316e-01 1.17850327e+00 -3.18347573e-01 -3.83799165e-01
-8.80613476e-02 -2.41827354e-01 -5.77600360e-01 6.16528392e-02
5.07516861e-01 8.03956866e-01 2.02117518e-01 -1.12646914e+00
-3.14031422e-01 -4.12818402e-01 1.95710316e-01 3.14611971e-01
-7.63749361e-01 2.64902681e-01 -3.25193778e-02 -9.77542341e-01
-1.51299452e-02 -7.64642298e-01 -1.11234836e-01 1.55621946e-01
2.39338391e-02 -5.14984965e-01 4.94482160e-01 -8.01598966e-01
9.84977603e-01 -2.11613083e+00 3.51902127e-01 -1.94320723e-01
-9.65202376e-02 -2.71630675e-01 -4.99728322e-01 8.23651731e-01
-5.84402867e-02 4.34424639e-01 -2.54987866e-01 -5.16721189e-01
3.85586500e-01 5.94953477e-01 -5.68926275e-01 5.55001915e-01
1.75607622e-01 1.06644440e+00 -9.44224358e-01 -3.79494518e-01
8.08020458e-02 4.39015567e-01 -7.82754779e-01 6.52847439e-02
-1.44252731e-02 -6.19172007e-02 -2.54756570e-01 3.12197953e-01
2.80983120e-01 1.25185037e-02 6.82689786e-01 -1.97970882e-01
-1.02918632e-02 9.27572548e-01 -5.23224413e-01 2.26048660e+00
-9.98096049e-01 6.88388407e-01 -3.59544426e-01 -1.20088232e+00
9.05000687e-01 3.74822170e-01 1.96759813e-02 -9.11348581e-01
-2.27801666e-01 3.19847167e-01 -1.22548185e-01 -4.99238610e-01
6.00208461e-01 -5.88136792e-01 -5.48482180e-01 8.07337880e-01
3.60685468e-01 -2.01401055e-01 -3.41065265e-02 2.70228028e-01
5.76771259e-01 2.27992594e-01 4.89055514e-01 -7.75708020e-01
1.73576489e-01 -1.28268106e-02 3.66459221e-01 3.28120917e-01
-5.48640564e-02 2.49708965e-01 3.45103949e-01 -3.19918156e-01
-1.13394594e+00 -1.58161402e+00 -5.30830383e-01 1.44537330e+00
1.18994325e-01 -6.76787734e-01 -4.17191058e-01 -9.63593602e-01
3.30317229e-01 1.03789091e+00 -8.24313402e-01 -9.69795808e-02
-2.49023646e-01 -3.96408498e-01 7.27800250e-01 9.08792257e-01
-3.27108532e-01 -7.44621456e-01 -1.06122695e-01 8.65095332e-02
-4.69830148e-02 -1.07227230e+00 -5.07016182e-01 1.39534604e-02
-7.95993090e-01 -1.04859948e+00 -3.56009990e-01 -1.11868501e+00
4.95911241e-01 1.94824159e-01 1.32168913e+00 -1.00286655e-01
-2.86022842e-01 6.70189440e-01 -3.93730432e-01 -2.09859669e-01
-3.31302702e-01 -4.67504933e-02 3.26503634e-01 -5.28792679e-01
8.23444128e-01 -6.28492355e-01 -2.50308841e-01 1.90103911e-02
-9.20231521e-01 -2.55323023e-01 2.09612533e-01 8.30302477e-01
3.44407201e-01 -4.39927459e-01 7.86479890e-01 -9.07558441e-01
7.91852891e-01 -9.70176935e-01 -1.64582714e-01 2.97315449e-01
-4.61894661e-01 4.78465080e-01 5.47640622e-01 -2.65135825e-01
-8.95090103e-01 -6.22105956e-01 1.32623315e-02 -1.89458400e-01
-1.54466465e-01 6.82408214e-01 1.97664816e-02 4.85360503e-01
5.97841978e-01 -5.13438880e-03 -4.20795241e-03 -6.91763222e-01
1.01888943e+00 6.86219633e-01 3.82533848e-01 -1.00591803e+00
5.92013597e-01 5.13855696e-01 -5.05449057e-01 -8.30801845e-01
-9.33799088e-01 -2.79754311e-01 -9.34083521e-01 5.19571602e-01
1.00579667e+00 -1.12331057e+00 8.44923407e-02 -2.89115578e-01
-1.32640028e+00 6.60173520e-02 -2.64374107e-01 6.76315725e-01
-3.24511617e-01 4.35756087e-01 -6.40578151e-01 -1.26879260e-01
6.91645443e-02 -8.75787318e-01 1.14346290e+00 -2.81888515e-01
-7.04615772e-01 -1.92015183e+00 4.89439130e-01 2.40679979e-01
4.50733393e-01 -2.74506714e-02 1.47396564e+00 -1.14370561e+00
-1.42804282e-02 5.62297367e-02 -2.53936350e-01 3.79297674e-01
5.09865403e-01 -1.67570919e-01 -7.81188548e-01 -3.86365861e-01
-2.46955559e-01 -6.27229452e-01 7.92188406e-01 1.14742197e-01
7.78831601e-01 -3.11742514e-01 -3.43291700e-01 4.95530963e-01
1.48002458e+00 -3.01085591e-01 1.76368415e-01 5.30236304e-01
7.31804669e-01 1.08503282e+00 1.49287760e-01 8.26804265e-02
8.33389103e-01 6.26488924e-01 -3.75670306e-02 3.67327407e-02
-2.39461660e-01 -6.28021657e-01 4.48480487e-01 1.29411995e+00
4.12629753e-01 -7.48845264e-02 -1.00801337e+00 1.07133365e+00
-1.64504111e+00 -7.99388826e-01 1.06414288e-01 1.88262224e+00
8.30516696e-01 -2.80884802e-01 -2.15820506e-01 -3.99160504e-01
5.53932905e-01 4.54857290e-01 -1.31849453e-01 -9.04972494e-01
-1.00858480e-01 6.19925857e-01 2.75295466e-01 7.64163136e-01
-9.35336232e-01 1.27947867e+00 7.63403130e+00 3.92167687e-01
-7.64746606e-01 4.16207135e-01 1.74073465e-02 -6.35631382e-02
-1.10937262e+00 2.59735405e-01 -5.56817293e-01 1.53019607e-01
1.04706192e+00 -3.41821313e-01 1.47696033e-01 6.39465570e-01
-4.99596372e-02 3.38817716e-01 -1.49206710e+00 6.38520956e-01
4.73997235e-01 -1.35784531e+00 4.04675096e-01 -7.59309381e-02
7.53901362e-01 2.54617006e-01 7.61133507e-02 3.41377020e-01
8.17366779e-01 -1.44085717e+00 7.23806560e-01 2.00807512e-01
7.15863526e-01 -6.66187286e-01 4.96241689e-01 -5.02763800e-02
-9.88442719e-01 3.85679603e-02 -5.17543674e-01 -2.70101041e-01
1.80644646e-01 1.48286954e-01 -5.26730776e-01 6.21452451e-01
5.26227295e-01 1.32870948e+00 -6.80711508e-01 2.00745165e-01
-5.75311422e-01 2.87620217e-01 6.48233369e-02 7.83812627e-02
4.41014498e-01 -8.06844234e-02 1.44555390e-01 1.53898108e+00
3.52768153e-01 -3.49830210e-01 2.89919317e-01 9.27836299e-01
-1.74559176e-01 4.96630996e-01 -9.97110128e-01 -3.14388543e-01
7.53490210e-01 7.33685195e-01 -1.45492002e-01 -3.18185508e-01
-9.28006887e-01 1.00458777e+00 7.80236781e-01 3.98648590e-01
-4.70283359e-01 -3.19287181e-01 1.14803839e+00 -1.14335403e-01
3.41403395e-01 -6.15862846e-01 -1.58933118e-01 -1.32641327e+00
1.45540535e-02 -5.97907186e-01 5.12904465e-01 -6.95916951e-01
-1.92724442e+00 3.56820464e-01 4.82100211e-02 -8.17693949e-01
-4.56297874e-01 -9.05706406e-01 -4.51853007e-01 1.08761919e+00
-1.75196326e+00 -1.44692838e+00 4.46917206e-01 5.57163477e-01
4.98503059e-01 -3.20456386e-01 1.35861969e+00 1.35667399e-01
-1.87760115e-01 6.06384099e-01 3.85742575e-01 2.48050049e-01
8.65796328e-01 -1.23859644e+00 5.74408233e-01 5.70123076e-01
6.73619688e-01 1.05613601e+00 4.25292075e-01 -3.71939331e-01
-1.19423401e+00 -8.20511281e-01 1.57658434e+00 -7.79642522e-01
1.26180780e+00 -5.92060447e-01 -9.24953043e-01 1.11408246e+00
5.70146143e-01 5.76410890e-02 1.29435408e+00 8.05310786e-01
-1.21605718e+00 4.03773576e-01 -8.52219045e-01 5.62022984e-01
1.16237152e+00 -1.27334380e+00 -1.47468901e+00 2.55155563e-01
9.59200978e-01 2.30023280e-01 -9.65087473e-01 -1.67319048e-02
5.15954852e-01 -6.10658646e-01 1.08256042e+00 -1.22446740e+00
6.98237181e-01 4.31948225e-04 -7.28884697e-01 -1.60834146e+00
-2.86777377e-01 5.92839643e-02 2.52136350e-01 1.31642210e+00
4.96068537e-01 -7.28404462e-01 1.02759160e-01 2.40896702e-01
-1.87267512e-01 -3.65116179e-01 -1.00258875e+00 -9.65292394e-01
9.45681036e-01 -4.21402127e-01 5.49734890e-01 1.56759214e+00
3.01863879e-01 4.85790879e-01 1.73447698e-01 1.78027213e-01
4.29308295e-01 2.57013202e-01 2.80705363e-01 -1.17642272e+00
-3.23572755e-01 -6.06327772e-01 -6.92046225e-01 -9.11082149e-01
1.05107284e+00 -1.62847495e+00 -3.38347554e-01 -1.70124197e+00
2.82157302e-01 -3.78458500e-01 -3.75201106e-01 4.59348798e-01
-1.87799539e-02 1.21580705e-01 3.23846936e-02 1.74268737e-01
-3.48450035e-01 8.49359989e-01 5.67132592e-01 -1.81239203e-01
3.46591949e-01 -9.27638471e-01 -1.07255220e+00 7.66265750e-01
6.00610018e-01 -4.99318868e-01 -5.74526787e-01 -1.16240644e+00
2.07859918e-01 -5.47657967e-01 2.86685318e-01 -3.47665370e-01
-1.66038573e-01 -1.14837497e-01 3.74472558e-01 1.84174538e-01
9.76676643e-02 -7.41086245e-01 -4.08434749e-01 2.19196633e-01
-7.09221423e-01 4.64601725e-01 3.39876175e-01 6.62373722e-01
-3.41547370e-01 -2.36742467e-01 4.42930937e-01 1.75099894e-02
-9.68215287e-01 3.34561872e-03 -3.29339594e-01 4.82672572e-01
8.78098905e-01 1.68155637e-02 -4.40522462e-01 -1.50574982e-01
-6.36714280e-01 1.91055909e-01 8.42676580e-01 1.01626575e+00
3.58935088e-01 -1.73888755e+00 -7.35182047e-01 1.49548009e-01
6.75373793e-01 -6.75066948e-01 -1.48111016e-01 2.87767649e-01
-2.63108402e-01 6.38636827e-01 -6.58991039e-02 -3.21731120e-01
-8.63887548e-01 6.56655312e-01 2.47581750e-01 5.25654480e-02
-5.75365365e-01 5.45090258e-01 4.05560136e-01 -1.19452512e+00
-1.48006633e-01 -3.05903941e-01 -1.50759473e-01 2.79589534e-01
2.65408665e-01 -6.30856603e-02 -1.92322761e-01 -9.08001363e-01
-5.95935702e-01 7.23441124e-01 -1.36518553e-01 -3.85705352e-01
1.45946896e+00 -2.71655023e-01 -2.32875109e-01 7.61359870e-01
1.77646613e+00 2.20998019e-01 -7.52447665e-01 -4.48496878e-01
3.84765357e-01 -5.30762613e-01 -7.43254274e-02 -5.80989540e-01
-5.91888785e-01 1.08986616e+00 2.43135035e-01 -1.04665078e-01
3.96405935e-01 5.98536670e-01 7.11157918e-01 2.17484951e-01
1.83982879e-01 -1.04681182e+00 8.45783576e-02 6.87851608e-01
7.80336142e-01 -1.05486786e+00 -9.07289535e-02 -1.18247233e-02
-6.28389776e-01 1.08754885e+00 2.73417115e-01 -3.05921853e-01
7.33937502e-01 1.19380817e-01 1.84016064e-01 -2.55508423e-01
-8.15504193e-01 -1.70425162e-01 2.88057178e-01 8.04234087e-01
1.17317688e+00 1.93567574e-01 -5.24197698e-01 6.26576781e-01
-3.85454684e-01 -5.42970121e-01 2.48300388e-01 9.14790273e-01
-2.57062286e-01 -1.38142693e+00 -5.44303749e-03 2.75862440e-02
-4.09139276e-01 -5.44654548e-01 -4.27882552e-01 9.08841133e-01
-6.41211569e-02 7.81844378e-01 6.41037166e-01 9.74661112e-03
2.94951558e-01 4.59736973e-01 6.06087089e-01 -9.53641176e-01
-1.65140569e-01 -3.00134063e-01 2.39138767e-01 -4.39374357e-01
-5.89217424e-01 -8.49581718e-01 -1.19289041e+00 -1.72665328e-01
5.20009771e-02 2.03695565e-01 7.39075363e-01 1.10965598e+00
5.25090873e-01 2.15160057e-01 4.43258256e-01 -5.34110308e-01
-6.39658272e-01 -8.60072672e-01 -5.57421327e-01 7.36428738e-01
2.18163937e-01 -7.11337805e-01 -3.43726337e-01 -8.64025205e-03] | [10.886643409729004, 9.710125923156738] |
22a4505e-8833-4012-a296-1dbe0ec1ae57 | agent-performing-autonomous-stock-trading | 2306.03985 | null | https://arxiv.org/abs/2306.03985v1 | https://arxiv.org/pdf/2306.03985v1.pdf | Agent Performing Autonomous Stock Trading under Good and Bad Situations | Stock trading is one of the popular ways for financial management. However, the market and the environment of economy is unstable and usually not predictable. Furthermore, engaging in stock trading requires time and effort to analyze, create strategies, and make decisions. It would be convenient and effective if an agent could assist or even do the task of analyzing and modeling the past data and then generate a strategy for autonomous trading. Recently, reinforcement learning has been shown to be robust in various tasks that involve achieving a goal with a decision making strategy based on time-series data. In this project, we have developed a pipeline that simulates the stock trading environment and have trained an agent to automate the stock trading process with deep reinforcement learning methods, including deep Q-learning, deep SARSA, and the policy gradient method. We evaluate our platform during relatively good (before 2021) and bad (2021 - 2022) situations. The stocks we've evaluated on including Google, Apple, Tesla, Meta, Microsoft, and IBM. These stocks are among the popular ones, and the changes in trends are representative in terms of having good and bad situations. We showed that before 2021, the three reinforcement methods we have tried always provide promising profit returns with total annual rates around $70\%$ to $90\%$, while maintain a positive profit return after 2021 with total annual rates around 2% to 7%. | ['Zhangqi Duan', 'Yunfei Luo'] | 2023-06-06 | null | null | null | null | ['q-learning'] | ['methodology'] | [-8.23666632e-01 -1.53875798e-01 -2.13186190e-01 -4.91343550e-02
-3.92329663e-01 -6.16035640e-01 6.75029635e-01 6.86376616e-02
-5.58199704e-01 1.09679973e+00 1.42799588e-02 -6.83028638e-01
1.09126717e-02 -1.02780008e+00 -5.20214081e-01 -3.50326270e-01
-3.43186051e-01 5.51206529e-01 1.15103178e-01 -5.78357697e-01
7.51453638e-01 3.56003135e-01 -1.48073292e+00 -7.16010705e-02
6.69919312e-01 1.30144608e+00 -3.08448151e-02 4.29904312e-01
-3.90744925e-01 1.26679516e+00 -8.44990849e-01 -3.53218943e-01
7.28604078e-01 -4.14837927e-01 -3.72600347e-01 -2.13687122e-01
-4.13495034e-01 -6.68363273e-01 9.06659737e-02 9.36566353e-01
2.77327836e-01 -6.40921295e-02 3.74206692e-01 -1.27665567e+00
-2.96889633e-01 6.47112250e-01 -6.13365352e-01 3.71065646e-01
1.01618208e-01 6.90233052e-01 1.00274563e+00 -4.62987721e-01
3.64423960e-01 8.76049578e-01 4.05387849e-01 3.61882269e-01
-9.62572217e-01 -9.81968164e-01 6.21435270e-02 -1.02468804e-01
-6.05600119e-01 -2.06900567e-01 6.61639750e-01 -3.70665878e-01
1.19406748e+00 -6.62767664e-02 1.29657817e+00 6.45053566e-01
7.38156319e-01 7.42233455e-01 1.35593736e+00 -2.28404656e-01
4.76797581e-01 2.37962782e-01 -2.58280665e-01 1.83592036e-01
2.53840119e-01 5.96056581e-01 -3.07102740e-01 -2.16399059e-01
8.86357903e-01 7.14273676e-02 2.74157822e-01 -1.49319783e-01
-1.17955649e+00 1.09522414e+00 2.46038690e-01 3.57973933e-01
-8.62678885e-01 3.78797531e-01 2.80857980e-01 8.41425121e-01
4.09337670e-01 9.03680146e-01 -6.81788743e-01 -6.94163382e-01
-1.00659370e+00 7.20382035e-01 1.11235356e+00 4.00657803e-01
4.56785679e-01 6.51091516e-01 1.81441978e-01 4.02413040e-01
2.49766260e-01 4.19240206e-01 9.71461773e-01 -1.28468955e+00
2.62770474e-01 3.83827925e-01 7.55590796e-01 -8.72826695e-01
-4.39401358e-01 -3.90276730e-01 -3.51073474e-01 8.13680410e-01
5.83025396e-01 -5.60470045e-01 -6.55482173e-01 1.44913483e+00
-5.96012175e-02 -5.36228297e-03 3.09660405e-01 4.55977976e-01
-6.42173737e-02 8.35254252e-01 -1.11072235e-01 -6.19781494e-01
8.53353202e-01 -7.86570311e-01 -6.13108516e-01 -2.26806536e-01
4.21552032e-01 -6.16806805e-01 9.03884888e-01 5.80478489e-01
-1.53381968e+00 -4.05157298e-01 -1.09558117e+00 9.32315767e-01
-3.54203284e-01 -5.24807334e-01 6.71288729e-01 5.08518159e-01
-1.17990375e+00 1.04209781e+00 -8.31369460e-01 3.12694401e-01
1.42621741e-01 3.41805249e-01 2.59884864e-01 7.81042993e-01
-1.35513866e+00 1.07315040e+00 4.20349479e-01 -2.18569472e-01
-8.48925769e-01 -5.95564961e-01 -2.59404749e-01 2.50941187e-01
4.66905832e-01 -2.13779449e-01 1.70471263e+00 -1.13885117e+00
-1.74571359e+00 3.48202646e-01 5.54609895e-01 -1.11352026e+00
7.65386522e-01 -1.79289088e-01 -5.70813358e-01 -1.69147253e-01
5.70681058e-02 5.16451120e-01 5.87631881e-01 -7.51748204e-01
-9.68850315e-01 -2.93619603e-01 3.90023068e-02 2.06547901e-01
-2.31834546e-01 -3.14648002e-02 4.50616419e-01 -7.85139143e-01
-2.68275797e-01 -8.01584840e-01 -3.44651341e-01 -7.21233249e-01
2.86155999e-01 -3.81401032e-02 5.53831160e-01 -6.79679513e-01
1.11049771e+00 -1.90796053e+00 -5.39646327e-01 3.50624859e-01
-1.79244801e-01 7.84402937e-02 2.42514685e-01 5.10344684e-01
-1.36622176e-01 2.60852218e-01 1.11080877e-01 3.16176236e-01
1.53493017e-01 -5.64943328e-02 -6.06257617e-01 1.36353830e-02
5.06274700e-02 8.69930804e-01 -7.59219170e-01 8.71380195e-02
7.32436180e-02 -3.87135148e-01 -3.83377850e-01 3.07081431e-01
-6.97503388e-01 6.73846230e-02 -3.91473770e-01 6.98404670e-01
2.13980183e-01 -3.27061951e-01 2.44423226e-01 4.83669996e-01
-5.06223857e-01 3.80707979e-01 -1.42259121e+00 8.26042771e-01
-3.96999657e-01 2.94471264e-01 -1.73491091e-01 -9.28306758e-01
1.20754600e+00 2.19543651e-01 8.41260433e-01 -1.29714072e+00
-9.72531829e-03 6.99223995e-01 2.79971212e-01 -7.30200410e-02
5.62636018e-01 -4.49918151e-01 -9.08430666e-03 1.01383913e+00
-3.35014731e-01 -4.20878768e-01 4.32279885e-01 -1.65665999e-01
1.09742486e+00 8.93601179e-02 3.17300498e-01 -2.89271116e-01
8.24365951e-03 2.59399027e-01 6.11189187e-01 3.83611143e-01
-4.42136645e-01 -1.57857686e-01 8.84501398e-01 -7.97665179e-01
-1.08046138e+00 -9.32537496e-01 3.37389380e-01 8.47463727e-01
-8.56053978e-02 8.93655419e-02 -3.79747152e-01 -3.35779369e-01
3.97946507e-01 1.05441606e+00 -3.80306900e-01 -1.26636341e-01
-4.48663354e-01 -7.11297512e-01 4.17401530e-02 5.24213612e-01
9.17167842e-01 -1.62865186e+00 -1.07161319e+00 6.98563457e-01
4.48610842e-01 -4.65291709e-01 -2.14800701e-01 3.13721091e-01
-9.59189057e-01 -9.05801117e-01 -7.85294771e-01 -3.02266836e-01
-7.45046372e-03 -9.31270048e-02 1.37559497e+00 -2.16355741e-01
1.07486792e-01 9.63840634e-02 -9.63157341e-02 -9.55904901e-01
-3.55443686e-01 -1.02134384e-01 1.81342274e-01 -4.16090786e-01
1.37479588e-01 -4.09154654e-01 -7.73750722e-01 2.26183489e-01
-6.98347032e-01 -2.38850147e-01 5.65536678e-01 7.85757303e-01
3.01526636e-01 2.65018106e-01 1.14143336e+00 -6.93995059e-01
9.51109648e-01 -5.34310102e-01 -1.23606920e+00 9.42917541e-02
-1.29193163e+00 1.20410375e-01 4.41708535e-01 -2.33421639e-01
-9.73435223e-01 -2.38881737e-01 1.09360091e-01 -1.84123114e-01
4.83029068e-01 6.50781512e-01 3.80740821e-01 3.46973777e-01
4.03770447e-01 2.20601574e-01 4.93760973e-01 -1.62500173e-01
-1.28830835e-01 4.19223130e-01 8.47751647e-02 -3.20683748e-01
5.49559534e-01 1.78622603e-01 -3.65456134e-01 -1.81177229e-01
-2.63096958e-01 1.29330084e-01 6.92920387e-02 -2.92625546e-01
4.60641026e-01 -8.18321466e-01 -8.30869198e-01 8.80911350e-01
-4.31321800e-01 -7.54098117e-01 -5.30869544e-01 4.68280554e-01
-8.13489497e-01 -2.28218511e-01 -7.95465887e-01 -1.01308477e+00
-4.16586816e-01 -1.04230654e+00 2.24378139e-01 5.39292216e-01
-3.27893376e-01 -9.89225686e-01 3.73743981e-01 8.71824846e-03
9.24988210e-01 3.60288918e-01 7.66701341e-01 -7.85933793e-01
-6.23309135e-01 -2.41765520e-03 8.12864602e-02 4.80184227e-01
2.65395880e-01 1.85753435e-01 -4.77679968e-01 -4.62189525e-01
7.10854381e-02 -4.81908351e-01 5.13951421e-01 6.13295555e-01
8.07239711e-01 -3.48998845e-01 1.73580199e-01 2.00352892e-01
1.28109717e+00 1.07067502e+00 5.85056365e-01 1.04453683e+00
-1.85890585e-01 5.67670405e-01 1.00265586e+00 7.82713056e-01
1.84167296e-01 2.65715569e-01 5.33469260e-01 2.66815454e-01
5.97141743e-01 -1.83412805e-01 7.38955736e-01 5.18032610e-01
1.71774998e-02 2.42727160e-01 -9.48405981e-01 4.33389366e-01
-1.70190299e+00 -1.22897637e+00 6.70076609e-01 2.15086365e+00
9.59380686e-01 9.15267110e-01 7.35703349e-01 -1.16328180e-01
3.59103203e-01 1.74808607e-01 -9.75262225e-01 -6.21486962e-01
2.06032507e-02 2.88535386e-01 6.78444147e-01 9.40048322e-02
-8.82869184e-01 8.01669240e-01 6.95040607e+00 3.45298976e-01
-1.43547845e+00 -5.68029344e-01 1.19272649e+00 -3.30679297e-01
-3.47745627e-01 -1.52097363e-02 -6.29709303e-01 7.71768451e-01
1.52817905e+00 -8.65899563e-01 5.12211621e-01 1.05387294e+00
4.39833641e-01 -2.63983279e-01 -6.43631339e-01 5.37505865e-01
-6.56601906e-01 -1.74685073e+00 -2.88199782e-01 2.67987549e-01
7.74496853e-01 7.59986266e-02 2.06111595e-01 6.57077610e-01
8.81152272e-01 -1.00669694e+00 9.36038435e-01 6.90232694e-01
2.51465291e-01 -1.26577055e+00 7.76780248e-01 4.69583780e-01
-8.44843268e-01 -5.37979782e-01 -4.29010317e-02 -4.51882154e-01
-6.86363652e-02 3.48885894e-01 -7.71955252e-01 8.30089971e-02
7.44953573e-01 5.59971213e-01 -2.01006502e-01 7.15975463e-01
1.21107541e-01 4.84667540e-01 -1.97985649e-01 -5.05872607e-01
5.14599979e-01 -4.81256872e-01 7.91240558e-02 4.08422470e-01
6.16028130e-01 -2.13860790e-03 1.61823377e-01 7.66316414e-01
1.03125926e-02 5.38330190e-02 -5.54798126e-01 -4.56492990e-01
3.70005697e-01 8.13281298e-01 -8.57542574e-01 -3.66064489e-01
-3.44844908e-01 3.20204914e-01 -2.06434369e-01 1.07058808e-01
-6.72434092e-01 -4.20546174e-01 7.53892720e-01 3.10027689e-01
2.70245999e-01 -1.86353162e-01 -3.23323131e-01 -8.70137036e-01
7.15364004e-03 -1.28671312e+00 2.71049440e-01 -6.61048055e-01
-1.02029884e+00 2.67767221e-01 -1.10470936e-01 -1.12784982e+00
-1.05028892e+00 -6.32730782e-01 -7.81620860e-01 7.84285843e-01
-1.50299907e+00 -2.52471149e-01 3.14830154e-01 3.56613576e-01
4.44348037e-01 -8.25117230e-01 4.74312842e-01 -9.53744426e-02
-3.67042333e-01 7.18695000e-02 4.07310784e-01 8.05648044e-02
3.20913166e-01 -1.52956104e+00 4.87390041e-01 4.85874057e-01
2.64731981e-03 3.26622128e-01 6.53993487e-01 -8.31634939e-01
-1.13061774e+00 -6.23215437e-01 2.73447037e-01 1.51460022e-01
1.17390108e+00 2.78924912e-01 -7.86293268e-01 4.99965101e-01
4.07130748e-01 -4.30195242e-01 3.17050368e-01 -2.96679676e-01
1.36828989e-01 -4.42040652e-01 -1.07141268e+00 6.14755988e-01
3.37601155e-01 -7.55998194e-02 -5.47436714e-01 -1.62266307e-02
6.24190509e-01 -3.32771838e-01 -8.27438593e-01 2.16477260e-01
5.43613255e-01 -1.32659507e+00 6.73080862e-01 -5.65218091e-01
4.13889617e-01 1.68379217e-01 8.28262419e-02 -1.64392269e+00
-1.59512669e-01 -9.59073365e-01 -1.87064439e-01 9.16791022e-01
6.53119683e-01 -1.01124287e+00 1.05971682e+00 8.61185431e-01
2.31834278e-01 -8.41726422e-01 -7.44675457e-01 -8.79711866e-01
4.73822683e-01 -1.25089958e-01 1.03755522e+00 7.50916302e-01
3.54406871e-02 -2.16916516e-01 -2.52553523e-01 -5.07300794e-01
4.28478688e-01 5.99851370e-01 7.15134680e-01 -9.11599100e-01
-5.85103810e-01 -9.66649413e-01 -1.59406755e-02 -4.81513828e-01
-9.48836356e-02 -2.06888258e-01 -3.12178016e-01 -1.18955326e+00
-1.97380945e-01 -4.82923001e-01 -7.26515889e-01 3.23041022e-01
8.69886503e-02 -4.44831610e-01 3.60238135e-01 4.42732215e-01
-8.72014314e-02 4.25150722e-01 1.27312887e+00 -9.86021757e-02
-3.92238259e-01 4.65513259e-01 -8.44689131e-01 6.85179830e-01
1.21866071e+00 -6.22691438e-02 -2.91801602e-01 1.01684190e-01
4.85224366e-01 4.46938306e-01 4.64909564e-04 -8.52094650e-01
-3.71272564e-02 -6.82846308e-01 5.00786066e-01 -5.74079275e-01
6.10718243e-02 -5.60184598e-01 2.87109375e-01 1.18748343e+00
-2.24958330e-01 7.84404218e-01 2.28872612e-01 1.62081510e-01
-5.16342819e-01 -1.41666725e-01 6.76776528e-01 -5.88085175e-01
-1.02029073e+00 1.00870691e-01 -4.38548803e-01 1.87062323e-01
1.40552223e+00 -1.40452206e-01 -1.43698364e-01 -8.11871827e-01
-6.09965503e-01 6.59588873e-01 3.54652137e-01 3.49991262e-01
4.89570886e-01 -1.23895872e+00 -6.19218469e-01 9.57049131e-02
-4.40588176e-01 -5.29693067e-01 -1.87947139e-01 3.29173803e-01
-8.44058275e-01 3.20926964e-01 -7.79675901e-01 -6.80958927e-02
-5.66291511e-01 4.56731528e-01 7.14285970e-01 -7.95316219e-01
-3.21135312e-01 4.31923687e-01 -2.71917820e-01 -1.23136774e-01
1.67384058e-01 -4.31205809e-01 -1.80984601e-01 4.98201758e-01
6.10755146e-01 5.04995704e-01 3.30716744e-02 9.41666663e-02
-1.47638604e-01 1.31796584e-01 -3.54862027e-02 -4.84172374e-01
1.72018468e+00 2.86981583e-01 1.15871713e-01 6.27281666e-01
6.77678525e-01 -2.50321567e-01 -1.49397182e+00 1.78420261e-01
5.47964454e-01 -4.02882993e-01 -2.66183075e-02 -8.97035182e-01
-1.30047321e+00 5.46908140e-01 5.36606491e-01 8.63008380e-01
1.07102227e+00 -6.46231830e-01 9.18838620e-01 2.78278708e-01
6.26087785e-01 -1.58200371e+00 3.10412049e-01 4.57705468e-01
8.38628888e-01 -1.19687998e+00 -1.20296188e-01 6.27390265e-01
-8.67718875e-01 1.11351299e+00 6.36645317e-01 -4.36171621e-01
9.21880543e-01 4.78058934e-01 3.56031239e-01 -4.24076729e-02
-1.14650178e+00 6.47442266e-02 -4.01714355e-01 1.11232787e-01
2.69798219e-01 2.26313293e-01 -1.13074295e-01 4.47518170e-01
-4.39119905e-01 2.72967786e-01 6.33060396e-01 1.15238690e+00
-7.25412071e-01 -1.16311395e+00 -3.41149181e-01 7.67586291e-01
-7.22867846e-01 1.95605889e-01 -3.02308410e-01 1.14056492e+00
-3.49346787e-01 6.00706458e-01 3.87446254e-01 -1.80131495e-01
4.42619383e-01 1.20199829e-01 -2.84614805e-02 -3.24987918e-01
-9.01330888e-01 2.83173800e-01 9.31003168e-02 -6.56399965e-01
-1.77852318e-01 -9.73384380e-01 -1.45091987e+00 -6.13400042e-01
1.03566520e-01 2.59036720e-01 5.20264328e-01 7.45386302e-01
2.37008095e-01 3.75379771e-01 1.14297533e+00 -6.84543014e-01
-1.37362885e+00 -7.71547556e-01 -1.05490339e+00 1.78755060e-01
1.21177673e-01 -6.69310808e-01 -3.33113492e-01 -3.15222830e-01] | [4.4478230476379395, 3.9400532245635986] |
ad559eed-af7b-4b00-a313-f73c24430949 | geographical-distance-is-the-new | 2205.08621 | null | https://arxiv.org/abs/2205.08621v1 | https://arxiv.org/pdf/2205.08621v1.pdf | Geographical Distance Is The New Hyperparameter: A Case Study Of Finding The Optimal Pre-trained Language For English-isiZulu Machine Translation | Stemming from the limited availability of datasets and textual resources for low-resource languages such as isiZulu, there is a significant need to be able to harness knowledge from pre-trained models to improve low resource machine translation. Moreover, a lack of techniques to handle the complexities of morphologically rich languages has compounded the unequal development of translation models, with many widely spoken African languages being left behind. This study explores the potential benefits of transfer learning in an English-isiZulu translation framework. The results indicate the value of transfer learning from closely related languages to enhance the performance of low-resource translation models, thus providing a key strategy for low-resource translation going forward. We gathered results from 8 different language corpora, including one multi-lingual corpus, and saw that isiXhosa-isiZulu outperformed all languages, with a BLEU score of 8.56 on the test set which was better from the multi-lingual corpora pre-trained model by 2.73. We also derived a new coefficient, Nasir's Geographical Distance Coefficient (NGDC) which provides an easy selection of languages for the pre-trained models. NGDC also indicated that isiXhosa should be selected as the language for the pre-trained model. | ['Innocent Amos Mchechesi', 'Muhammad Umair Nasir'] | 2022-05-17 | null | null | null | null | ['low-resource-neural-machine-translation'] | ['natural-language-processing'] | [-1.10366553e-01 -2.37018883e-01 -5.83779275e-01 -1.54189020e-01
-1.34716642e+00 -8.19350302e-01 7.96626031e-01 -2.37372052e-02
-6.35204017e-01 1.08630288e+00 4.79704857e-01 -9.69027519e-01
-1.42451391e-01 -6.21177554e-01 -5.82028508e-01 -4.89732563e-01
3.22903961e-01 7.17707217e-01 -2.96119988e-01 -5.96500754e-01
2.75259286e-01 3.72468293e-01 -7.57911861e-01 2.97900021e-01
1.21177852e+00 -5.51685784e-03 5.55523038e-01 1.38366908e-01
-6.72785342e-02 2.19999239e-01 -5.56319475e-01 -6.75302684e-01
3.39583278e-01 -7.83468604e-01 -7.95413852e-01 -3.23947817e-01
1.78609252e-01 -1.46604285e-01 -1.49964700e-02 7.80307531e-01
6.64563417e-01 -5.89157157e-02 6.75947070e-01 -6.29161596e-01
-8.96935761e-01 8.61377239e-01 -3.92338246e-01 3.26774418e-01
2.78367311e-01 -1.14997432e-01 1.09478784e+00 -1.30582809e+00
8.57299089e-01 1.15420234e+00 5.41715622e-01 2.22319856e-01
-9.39913988e-01 -7.06002951e-01 -3.92263591e-01 4.29440215e-02
-1.27381313e+00 -6.07162118e-01 3.88706595e-01 -3.13857704e-01
1.26765239e+00 -8.77598021e-03 3.92574400e-01 7.47019351e-01
3.52616370e-01 2.67895162e-01 1.67920947e+00 -1.10041463e+00
-2.64832050e-01 5.10881007e-01 -7.65438497e-01 3.62724096e-01
4.61287439e-01 3.57962735e-02 -6.70241952e-01 -5.46293855e-02
7.10373759e-01 -4.14884418e-01 6.39436617e-02 1.80198476e-01
-1.47587359e+00 9.90661323e-01 1.73345849e-01 8.72671068e-01
-2.62407899e-01 -6.29091740e-01 4.65602666e-01 6.25621021e-01
8.62329662e-01 4.10292625e-01 -7.26330936e-01 -2.61694878e-01
-9.51836705e-01 -3.74115258e-02 6.31246984e-01 8.10069621e-01
7.34843850e-01 1.49719179e-01 4.41333979e-01 1.17833734e+00
2.64616072e-01 8.15419376e-01 4.96046782e-01 -2.98363149e-01
9.93750453e-01 4.85090643e-01 -8.68034065e-02 -7.46635973e-01
1.33592233e-01 -3.48870814e-01 -4.69329834e-01 -2.34836072e-01
5.26810944e-01 -3.47523540e-01 -8.47113550e-01 1.50538266e+00
5.24370559e-02 -6.79267466e-01 5.23498714e-01 7.77554452e-01
1.87081769e-01 1.05457723e+00 4.74590100e-02 -2.78958470e-01
9.28050697e-01 -8.99660468e-01 -5.03899932e-01 -3.37146610e-01
1.07682300e+00 -1.40970635e+00 1.19497216e+00 7.38629475e-02
-9.80709314e-01 -4.62645918e-01 -9.23312604e-01 1.35883838e-01
-5.49266160e-01 2.58197695e-01 6.65326774e-01 6.50651932e-01
-1.14274037e+00 5.51210463e-01 -6.91891193e-01 -8.80820572e-01
4.86296490e-02 4.47665244e-01 -6.58743203e-01 -4.59462613e-01
-1.28050792e+00 1.59333956e+00 5.37585497e-01 1.13669679e-01
-3.22170764e-01 -3.05949301e-01 -6.88219666e-01 -3.56829017e-01
-7.08124116e-02 -1.88717872e-01 7.13462234e-01 -1.13309693e+00
-1.32156992e+00 8.74152422e-01 -4.15159576e-02 -1.19000547e-01
3.87540519e-01 -3.14148664e-02 -6.94354713e-01 -6.61027851e-03
5.21586895e-01 3.59860927e-01 2.68838942e-01 -9.25571501e-01
-7.52488852e-01 -3.15608531e-01 -3.68808895e-01 6.02133393e-01
-3.25921625e-01 8.19291234e-01 -1.62544087e-01 -6.31026149e-01
1.07188337e-02 -1.02075624e+00 2.52891034e-02 -7.79475391e-01
2.86260843e-01 -1.80536643e-01 5.72916090e-01 -1.25241911e+00
1.16939104e+00 -1.87665641e+00 -1.49974851e-02 1.26918390e-01
-5.71841955e-01 4.40059215e-01 -2.79006004e-01 1.00617266e+00
1.96155831e-01 3.22836637e-01 -1.08314138e-02 2.74097890e-01
-2.89115161e-01 4.39806849e-01 -6.20094836e-02 3.49702477e-01
4.94059056e-01 7.54758716e-01 -1.04703701e+00 -4.93000835e-01
1.14725761e-01 4.89085823e-01 -2.10183218e-01 -1.19404003e-01
2.90290833e-01 5.21030307e-01 -2.32383907e-01 7.81701982e-01
3.85137051e-01 2.66317219e-01 2.87978947e-01 4.74146217e-01
-3.75109226e-01 7.90718317e-01 -5.25140822e-01 1.52828693e+00
-7.64456034e-01 5.98950028e-01 -2.32270747e-01 -5.72231531e-01
1.27306318e+00 4.47347224e-01 3.46738279e-01 -9.07059908e-01
3.80531251e-02 1.13442898e+00 7.02919126e-01 -2.57794827e-01
4.56850588e-01 -4.25947309e-01 -6.70260340e-02 5.18585324e-01
3.56354453e-02 -2.69612432e-01 2.92732835e-01 -2.00609043e-01
5.44639409e-01 5.49857676e-01 1.64846733e-01 -6.61081374e-01
4.27590698e-01 5.43788612e-01 5.95016360e-01 1.19386189e-01
-5.91883920e-02 3.65836799e-01 -1.38232842e-01 -3.24751049e-01
-1.31957674e+00 -7.62036920e-01 -3.25796723e-01 1.08691442e+00
-5.63538313e-01 -3.07018548e-01 -5.70121765e-01 -4.96436298e-01
-3.57483089e-01 7.37872958e-01 -6.09213375e-02 1.67739600e-01
-8.55283380e-01 -8.50399852e-01 6.19287074e-01 2.48444661e-01
3.31920356e-01 -9.37259555e-01 -1.90527037e-01 3.21980447e-01
-4.57847059e-01 -9.70359504e-01 -3.75773638e-01 1.63120002e-01
-1.00961423e+00 -4.35900182e-01 -1.00759029e+00 -1.10117495e+00
5.29551566e-01 2.02301964e-01 8.74052167e-01 -2.21820176e-01
3.17248136e-01 -2.66620368e-01 -4.65031385e-01 -5.38829744e-01
-8.84342492e-01 5.47369361e-01 2.91117460e-01 -4.75514978e-01
8.11711729e-01 -1.89436063e-01 2.83670370e-02 3.63574296e-01
-5.22406757e-01 5.95072545e-02 8.54172111e-01 9.23000097e-01
4.24075752e-01 -1.12313647e-02 1.09294522e+00 -6.66873217e-01
5.77103138e-01 -5.56579053e-01 -2.90350765e-01 3.80932659e-01
-7.54784644e-01 -2.69402832e-01 6.32694304e-01 -4.65969354e-01
-1.15265512e+00 -3.31733942e-01 -4.36353981e-02 2.97473818e-01
1.47479475e-01 9.22912955e-01 -1.80408075e-01 -3.22602107e-03
5.94078660e-01 9.31154117e-02 -1.32342026e-01 -3.59925270e-01
1.01219594e-01 1.17243993e+00 1.77086055e-01 -8.04073155e-01
7.75792301e-01 -2.11898372e-01 -3.55579793e-01 -9.77375150e-01
-3.14460814e-01 -3.75884831e-01 -1.04822516e+00 5.75076491e-02
6.67009115e-01 -1.14430797e+00 3.23664069e-01 2.20061719e-01
-9.15119410e-01 -5.27557492e-01 2.81286836e-01 1.07276893e+00
-2.94866174e-01 -4.30750009e-03 -6.13531351e-01 -5.95756948e-01
-3.56130570e-01 -1.13040864e+00 6.02690101e-01 -5.55767789e-02
-4.74722207e-01 -1.36857307e+00 3.23030978e-01 5.98733485e-01
5.53859055e-01 6.18624389e-02 1.31791198e+00 -8.67951989e-01
-1.95403755e-01 -9.13416669e-02 -1.54024422e-01 3.52391750e-01
6.47723496e-01 -5.73841520e-02 -5.04341781e-01 -5.45690358e-01
-1.48984537e-01 -3.64706248e-01 1.37817651e-01 -3.30355503e-02
-1.73101291e-01 -2.12172940e-01 1.04148649e-02 2.17122242e-01
1.55231214e+00 3.97303909e-01 4.81403589e-01 7.55062044e-01
4.87726331e-01 5.40766418e-01 7.49531746e-01 -2.12947175e-01
4.98422593e-01 5.56974053e-01 -4.09583479e-01 -3.38892907e-01
-1.48003846e-01 -4.45318818e-01 8.77322793e-01 1.75540960e+00
-4.01357889e-01 5.41319773e-02 -1.36513972e+00 8.37545335e-01
-1.46228015e+00 -5.64823866e-01 -1.48230031e-01 2.28244567e+00
1.08399844e+00 -8.97529945e-02 7.61390850e-02 -2.10838944e-01
5.82317829e-01 -1.54814318e-01 4.97337840e-02 -1.02725911e+00
-2.97460645e-01 5.41635454e-01 6.38299406e-01 5.87178051e-01
-6.57655358e-01 1.62921536e+00 6.43027353e+00 7.52422035e-01
-1.27194011e+00 7.72635862e-02 4.19629723e-01 2.03382760e-01
-2.05313161e-01 2.44479463e-01 -8.55105162e-01 2.72624999e-01
1.57574749e+00 -5.09157538e-01 4.46203411e-01 4.13437247e-01
5.19883156e-01 -1.00108303e-01 -8.11000109e-01 4.17627454e-01
2.24445701e-01 -9.09044981e-01 2.06366196e-01 3.58856469e-01
1.19824350e+00 4.76995587e-01 -9.92419124e-02 2.74542749e-01
3.91283661e-01 -9.94963408e-01 3.46399575e-01 -1.79627672e-01
1.09179127e+00 -8.87025237e-01 9.49733078e-01 4.21110988e-01
-8.11119735e-01 3.90037417e-01 -7.27605164e-01 -3.14461082e-01
1.49109522e-02 5.74038438e-02 -1.29356301e+00 9.52894092e-01
4.24937248e-01 5.97757936e-01 -5.31386614e-01 5.83698153e-01
-2.38388538e-01 9.47333157e-01 -1.94538578e-01 1.23595968e-02
6.88951612e-01 -6.78797066e-01 1.38599426e-01 1.41422856e+00
7.78893769e-01 -2.61678666e-01 1.14015952e-01 1.43275380e-01
-1.09179132e-02 9.34122443e-01 -1.04528499e+00 -4.57869947e-01
4.14081275e-01 9.05343890e-01 -7.08274782e-01 -1.36553556e-01
-8.62495482e-01 9.69386399e-01 1.73918471e-01 1.93313330e-01
-4.61009204e-01 -4.04089808e-01 2.64334321e-01 1.37779862e-01
1.02646891e-02 -5.47686160e-01 -3.67387384e-01 -1.06970477e+00
-4.96674366e-02 -1.38526642e+00 1.96321502e-01 -5.67982256e-01
-1.08401966e+00 8.31291020e-01 3.38964127e-02 -1.09783340e+00
-5.51616192e-01 -6.15279675e-01 -2.67896235e-01 1.65748072e+00
-1.36140060e+00 -1.65690815e+00 6.62892401e-01 4.41116810e-01
6.99796677e-01 -5.51577210e-01 1.07922292e+00 4.72677290e-01
-2.66334683e-01 5.90267777e-01 4.17392761e-01 2.59449899e-01
1.16785920e+00 -7.92218566e-01 3.24107379e-01 1.05345023e+00
3.58115315e-01 8.65123272e-01 2.68305928e-01 -9.01719749e-01
-1.38620818e+00 -8.87306929e-01 1.84054303e+00 -5.15976250e-01
9.67670858e-01 -1.65232137e-01 -8.18418622e-01 7.26285040e-01
6.15945280e-01 -8.23525429e-01 9.73309755e-01 1.84386581e-01
-8.78754631e-02 6.87765479e-02 -9.39642966e-01 7.35519052e-01
5.96560121e-01 -6.95107877e-01 -7.50798881e-01 3.73724222e-01
3.95660758e-01 -5.79796061e-02 -1.05356658e+00 1.42142028e-01
4.78685200e-01 -3.07192236e-01 5.41203439e-01 -6.92254007e-01
5.67965686e-01 -1.55521601e-01 -3.86137724e-01 -1.62623990e+00
-3.44266027e-01 -5.30061364e-01 8.07260036e-01 1.40579736e+00
9.65624809e-01 -6.92472756e-01 3.92736018e-01 3.85418534e-01
-8.29598159e-02 -5.29969335e-01 -8.68256092e-01 -9.66025889e-01
8.44254494e-01 -9.25328583e-02 3.98058355e-01 1.26599312e+00
1.09013706e-01 7.53943503e-01 -3.25660348e-01 -1.07905634e-01
8.97349119e-02 -1.71757564e-01 6.59970522e-01 -9.18303013e-01
-1.01936489e-01 -2.52013505e-01 -1.02626249e-01 -3.89799595e-01
1.61536619e-01 -1.32860434e+00 -1.44970447e-01 -1.74176800e+00
1.23159580e-01 -7.07918584e-01 -8.55660290e-02 6.62832558e-01
-1.30347565e-01 4.41602081e-01 2.34343231e-01 5.71459651e-01
3.03598732e-01 1.23208351e-01 1.41466618e+00 1.87686533e-01
-4.64784980e-01 -6.11186512e-02 -9.25711632e-01 4.81221199e-01
1.10858774e+00 -5.81389189e-01 -3.05631697e-01 -8.35411131e-01
2.31528252e-01 -1.16302706e-01 -4.49228734e-01 -5.90700507e-01
-5.14976457e-02 -5.21746755e-01 4.25846249e-01 -4.28887784e-01
1.83926523e-02 -6.51485801e-01 2.33223334e-01 3.81355226e-01
-5.13372719e-02 6.92898214e-01 4.12686676e-01 -3.19201261e-01
-2.76304036e-01 -2.25576237e-01 4.70359415e-01 -2.13057905e-01
-5.07065296e-01 -3.71330827e-02 -4.11952525e-01 -4.06512320e-02
6.74137175e-01 -2.94966131e-01 -8.79877508e-02 -1.26972094e-01
-2.36183047e-01 -1.35545954e-01 7.75973022e-01 5.53836107e-01
2.52402812e-01 -1.39153230e+00 -1.22379553e+00 2.25867242e-01
1.47792220e-01 -4.54953462e-01 -4.31233078e-01 8.16188514e-01
-7.52758265e-01 6.55245662e-01 -6.21198475e-01 -2.15384945e-01
-1.00965750e+00 1.51136592e-01 -1.44255176e-01 -3.38961333e-01
-3.95637512e-01 2.57890433e-01 -3.61106068e-01 -7.89372385e-01
-4.21244830e-01 9.47237462e-02 9.28759947e-02 2.97387950e-02
1.73410073e-01 2.95350879e-01 3.08632851e-01 -1.24443030e+00
-3.52852464e-01 4.37058419e-01 -1.99614957e-01 -7.03651905e-01
1.40365839e+00 -1.61432281e-01 -3.27251256e-01 5.79957485e-01
1.03394353e+00 5.07467806e-01 -4.95691836e-01 -2.18844607e-01
2.72115260e-01 -6.06308937e-01 -4.92651425e-02 -1.24736416e+00
-4.54297334e-01 8.70814979e-01 1.88563302e-01 -4.99661356e-01
9.94747043e-01 -2.50446409e-01 6.29367292e-01 3.42298925e-01
7.05772519e-01 -1.30338025e+00 -5.16235113e-01 8.76910746e-01
7.25104749e-01 -1.15364707e+00 -9.05257538e-02 -5.97903058e-02
-7.97249734e-01 1.04868639e+00 2.30639726e-01 9.63091329e-02
3.42052132e-01 1.10528678e-01 6.07505620e-01 3.53461623e-01
-5.35576224e-01 -2.33292088e-01 2.16602385e-01 6.74600065e-01
9.51615930e-01 3.26262414e-01 -8.65184903e-01 2.62697637e-01
-4.94708568e-01 -8.10439959e-02 3.77260357e-01 7.98621476e-01
-2.88711160e-01 -1.75619805e+00 -5.41972816e-01 2.55491734e-01
-8.40580285e-01 -5.06769121e-01 -6.15085363e-01 1.18531072e+00
1.57157376e-01 9.92467225e-01 -1.27719104e-01 -3.93124312e-01
-4.20717224e-02 3.97816241e-01 5.41249692e-01 -8.03152561e-01
-7.98401952e-01 6.14007652e-01 3.11397403e-01 2.47769535e-01
-4.43436205e-01 -7.08364844e-01 -8.92254353e-01 -5.18336713e-01
-5.22226036e-01 4.70763743e-01 9.86230433e-01 9.65182900e-01
1.97025537e-01 -1.38965011e-01 6.91097558e-01 -4.98192459e-01
-4.83247906e-01 -1.22973609e+00 -2.17088446e-01 1.80007797e-02
-3.21410745e-01 -1.57209769e-01 -1.13377254e-02 1.18026972e-01] | [11.384538650512695, 10.369169235229492] |
9b492bf5-7317-40be-b0dd-42622efc1d7c | defusionnet-defocus-blur-detection-via | null | null | http://openaccess.thecvf.com/content_CVPR_2019/html/Tang_DeFusionNET_Defocus_Blur_Detection_via_Recurrently_Fusing_and_Refining_Multi-Scale_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Tang_DeFusionNET_Defocus_Blur_Detection_via_Recurrently_Fusing_and_Refining_Multi-Scale_CVPR_2019_paper.pdf | DeFusionNET: Defocus Blur Detection via Recurrently Fusing and Refining Multi-Scale Deep Features | Defocus blur detection aims to detect out-of-focus regions from an image. Although attracting more and more attention due to its widespread applications, defocus blur detection still confronts several challenges such as the interference of background clutter, sensitivity to scales and missing boundary details of defocus blur regions. To deal with these issues, we propose a deep neural network which recurrently fuses and refines multi-scale deep features (DeFusionNet) for defocus blur detection. We firstly utilize a fully convolutional network to extract multi-scale deep features. The features from bottom layers are able to capture rich low-level features for details preservation, while the features from top layers can characterize the semantic information to locate blur regions. These features from different layers are fused as shallow features and semantic features, respectively. After that, the fused shallow features are propagated to top layers for refining the fine details of detected defocus blur regions, and the fused semantic features are propagated to bottom layers to assist in better locating the defocus regions. The feature fusing and refining are carried out in a recurrent manner. Also, we finally fuse the output of each layer at the last recurrent step to obtain the final defocus blur map by considering the sensitivity to scales of the defocus degree. Experiments on two commonly used defocus blur detection benchmark datasets are conducted to demonstrate the superority of DeFusionNet when compared with other 10 competitors. Code and more results can be found at: http://tangchang.net
| [' Albert Zomaya', ' Lizhe Wang', ' Xinwang Liu', ' Xinzhong Zhu', 'Chang Tang'] | 2019-06-01 | null | null | null | cvpr-2019-6 | ['defocus-blur-detection', 'defocus-estimation'] | ['computer-vision', 'computer-vision'] | [ 2.15739626e-02 -6.12770915e-01 1.99076608e-01 -4.57561225e-01
-2.62975663e-01 -4.83512193e-01 2.99655944e-01 -6.26525134e-02
-2.26958498e-01 7.30065703e-01 5.39068758e-01 2.11734965e-01
-1.22428454e-01 -5.20071328e-01 -5.46521902e-01 -8.90442729e-01
1.72479331e-01 -3.30088466e-01 5.17488658e-01 1.02414645e-01
4.71186966e-01 5.66063762e-01 -1.51545715e+00 3.69373590e-01
1.13512015e+00 1.09945047e+00 5.12269795e-01 4.92855012e-01
1.38599917e-01 7.42717385e-01 -5.00849783e-01 1.80883080e-01
-4.88078967e-02 -1.80479944e-01 -2.50136584e-01 -1.39740705e-02
5.10696292e-01 -7.27003157e-01 -5.21203697e-01 1.52054906e+00
5.99652708e-01 3.67032811e-02 4.05007452e-01 -7.43705928e-01
-5.76515794e-01 2.86020607e-01 -7.91199148e-01 7.41907477e-01
2.78332442e-01 3.73607308e-01 5.37110150e-01 -1.07870448e+00
1.87640429e-01 1.27505481e+00 3.63923848e-01 1.45982087e-01
-7.01747298e-01 -7.46807992e-01 1.00052550e-01 4.46163982e-01
-1.21417880e+00 -5.82467735e-01 7.34755039e-01 -3.91231924e-01
6.53956056e-01 1.22464322e-01 4.36077267e-01 7.28440464e-01
4.58986193e-01 7.54676819e-01 9.95043635e-01 5.67282885e-02
-4.66448329e-02 -1.87813506e-01 4.10308689e-01 9.03603911e-01
4.82725143e-01 2.89134651e-01 -3.70276690e-01 1.65923551e-01
1.06657159e+00 3.22610289e-01 -9.75317240e-01 -2.59645998e-01
-1.33774507e+00 5.66670358e-01 1.19688308e+00 4.73193794e-01
-4.54861522e-01 1.31441101e-01 1.30229950e-01 -2.05252506e-02
4.81905907e-01 6.25089347e-01 -3.68161857e-01 -1.34469360e-01
-7.40024984e-01 1.61129117e-01 3.50124210e-01 7.18060195e-01
1.15794611e+00 -2.92925179e-01 -6.49934769e-01 8.52932990e-01
1.84304163e-01 4.90839094e-01 3.51001740e-01 -7.67396033e-01
3.62722486e-01 7.07022011e-01 3.52999628e-01 -1.08163035e+00
-5.29898286e-01 -7.26375997e-01 -9.31175113e-01 1.40971048e-02
1.24567918e-01 -3.95158887e-01 -1.07455921e+00 1.48892701e+00
8.21404159e-02 7.01732397e-01 -6.36476800e-02 1.65792358e+00
9.88679707e-01 5.68795681e-01 -3.24653000e-01 -2.21703306e-01
1.50461531e+00 -1.21911216e+00 -7.44744837e-01 -2.77064055e-01
7.33865276e-02 -9.98367250e-01 6.74701333e-01 1.03859724e-02
-1.08073032e+00 -8.27805400e-01 -1.03576696e+00 -2.48676762e-01
-2.29833826e-01 2.61419922e-01 6.13961577e-01 8.74915570e-02
-1.10328925e+00 2.05917165e-01 -6.86204970e-01 1.77844483e-02
7.73905635e-01 2.39992678e-01 5.42020686e-02 -4.14530724e-01
-1.38036156e+00 7.11879313e-01 3.22554678e-01 4.76414740e-01
-9.78867114e-01 -7.80315340e-01 -1.08273852e+00 3.81637722e-01
2.25781932e-01 -8.65808487e-01 1.04238212e+00 -5.43763697e-01
-9.82971013e-01 3.00318539e-01 -3.25287938e-01 -1.63543046e-01
4.83321995e-01 -4.18132424e-01 -2.94300079e-01 2.23994896e-01
1.50125742e-01 5.38329065e-01 1.11057079e+00 -1.01360905e+00
-1.05679739e+00 -2.97301590e-01 2.60905564e-01 4.77848679e-01
-5.79231456e-02 -5.89145944e-02 -4.88849908e-01 -5.13591588e-01
1.06395505e-01 -4.77239311e-01 -1.00983620e-01 -1.38542876e-01
-3.64407003e-01 -2.94679739e-02 1.20031726e+00 -3.89650196e-01
1.20414066e+00 -2.45743060e+00 2.18526766e-01 -3.62862736e-01
6.35699272e-01 3.06809694e-01 2.32453402e-02 -1.13214083e-01
1.06576048e-02 -3.47971231e-01 -2.36730650e-01 -1.22480899e-01
-2.42421314e-01 -2.08775580e-01 -1.41544864e-01 4.98788118e-01
3.12131435e-01 1.05221748e+00 -1.17332399e+00 -2.15810165e-01
6.88841343e-01 5.17977893e-01 -4.07373965e-01 4.65451628e-01
-9.85526815e-02 4.54377443e-01 -6.57117009e-01 6.03407204e-01
1.15238059e+00 -3.47297311e-01 -6.69306695e-01 -8.05153191e-01
-3.96177173e-01 -5.13024516e-02 -8.53743255e-01 1.68579400e+00
-4.71803427e-01 6.79848433e-01 1.64641172e-01 -5.00586748e-01
7.87497520e-01 -8.19231793e-02 1.06327221e-01 -3.81478101e-01
3.50995809e-01 3.21244538e-01 4.71134111e-02 -7.28217959e-01
4.30361658e-01 -6.92132413e-02 -8.92063454e-02 1.48336977e-01
-2.30750188e-01 -2.01946825e-01 -1.16493575e-01 -9.21330452e-02
8.73771429e-01 -1.71388090e-01 -3.93309601e-04 -3.95837158e-01
8.49741101e-01 -4.00806904e-01 3.34107876e-01 5.80709219e-01
-5.37505329e-01 8.41265738e-01 1.46309450e-01 -4.53263491e-01
-5.04890859e-01 -1.17205715e+00 -2.61433393e-01 6.30030334e-01
9.02147770e-01 9.95573625e-02 -6.63435102e-01 -8.39639425e-01
2.26436242e-01 3.96950811e-01 -7.12947130e-01 -5.06371558e-01
-3.02741021e-01 -5.82222402e-01 -4.10909541e-02 4.23804909e-01
1.09052563e+00 -1.00093138e+00 -8.46759439e-01 2.52939481e-02
-2.45947078e-01 -1.04722178e+00 -8.07984710e-01 1.28336847e-01
-4.36306298e-01 -1.08089781e+00 -1.13679183e+00 -9.20461833e-01
6.62194192e-01 7.62498677e-01 5.49987316e-01 3.53023782e-02
-3.24328929e-01 -2.69534767e-01 -2.60835681e-02 -1.91065118e-01
8.79112072e-03 1.74425244e-02 -1.34629577e-01 1.67750776e-01
3.90439808e-01 -3.79844844e-01 -1.16407382e+00 2.40006000e-01
-8.32708359e-01 1.94362566e-01 7.65689552e-01 8.38425934e-01
1.70368627e-01 4.17053133e-01 3.30249757e-01 -5.73259234e-01
7.14812338e-01 -3.80425334e-01 -6.09045029e-01 -8.39275345e-02
-1.36287898e-01 4.68999334e-03 4.37231898e-01 -1.04674838e-01
-1.28616929e+00 -4.19331163e-01 9.11913216e-02 -8.87225389e-01
-4.38896000e-01 4.50050473e-01 -1.98529422e-01 -5.10277674e-02
4.25350785e-01 1.71484068e-01 -2.31898546e-01 -6.09594464e-01
1.89963341e-01 8.56790781e-01 6.95664346e-01 -4.17959951e-02
6.30703032e-01 3.20947945e-01 -3.18819821e-01 -4.54587519e-01
-1.27436066e+00 -4.27065909e-01 -3.66101116e-01 -1.52409792e-01
9.40323234e-01 -1.04280090e+00 -6.67417347e-01 1.06730139e+00
-1.24456704e+00 2.02326581e-01 1.41420737e-01 6.28765464e-01
-1.73681408e-01 4.04985070e-01 -8.18805397e-01 -3.98697048e-01
-2.42588833e-01 -1.50709939e+00 1.10328341e+00 9.93762732e-01
4.03353661e-01 -8.60456526e-01 -3.15102607e-01 3.13228965e-01
6.60990775e-01 1.36297092e-01 7.80962944e-01 -1.43266931e-01
-8.61652851e-01 2.56020159e-01 -1.00554538e+00 3.14848065e-01
8.73686314e-01 -2.47491091e-01 -1.08489823e+00 -4.51175898e-01
3.27521265e-01 -1.17873922e-02 1.37509263e+00 7.71882355e-01
1.09049189e+00 -9.48104039e-02 -3.86931896e-01 9.46621358e-01
1.46324778e+00 2.36301735e-01 3.53687644e-01 1.48219466e-01
9.06658411e-01 2.00350478e-01 7.53891170e-01 4.34852570e-01
2.37086132e-01 3.32952052e-01 6.26376569e-01 -4.12615985e-01
-4.08388197e-01 1.42829390e-02 3.99541594e-02 4.93965894e-01
2.91092008e-01 1.81117803e-02 -6.38039947e-01 6.54729486e-01
-1.80362892e+00 -7.53403485e-01 4.05321456e-02 1.78654957e+00
7.74944842e-01 2.30429664e-01 -2.71151751e-01 -2.40477383e-01
1.10970700e+00 5.49342215e-01 -9.14846778e-01 -3.67947333e-02
-1.61306396e-01 -1.61148645e-02 2.49340251e-01 5.57862759e-01
-1.22981906e+00 9.63718355e-01 4.86353970e+00 6.08699977e-01
-1.39672434e+00 -9.92555842e-02 5.93657434e-01 -5.16064353e-02
-1.89178213e-01 -1.38584659e-01 -6.10241294e-01 8.30270410e-01
2.12217629e-01 -8.89178589e-02 6.89941525e-01 4.45180148e-01
3.77533704e-01 -2.38458112e-01 -8.14161718e-01 1.17637658e+00
2.25104112e-03 -1.20399082e+00 5.13003245e-02 -2.74790406e-01
8.75220478e-01 3.27539742e-01 1.07440524e-01 6.92379028e-02
5.85050806e-02 -9.61320519e-01 4.60174859e-01 8.65806162e-01
7.71415293e-01 -9.27137733e-01 1.09565091e+00 2.65392333e-01
-1.14967167e+00 -2.54742056e-01 -6.03371680e-01 -9.61924642e-02
4.08278741e-02 1.03909898e+00 -6.67930841e-01 6.99275136e-01
8.36432695e-01 1.19739962e+00 -5.02909243e-01 1.48957634e+00
-2.04539329e-01 1.47959158e-01 1.91357993e-02 2.01484021e-02
3.25312763e-01 -5.30427098e-02 5.82348645e-01 1.35046232e+00
4.67282832e-01 2.74399552e-03 -9.45665464e-02 1.12569380e+00
-8.52255821e-02 -5.72222888e-01 -2.22243160e-01 1.70110777e-01
3.68195117e-01 1.30311251e+00 -3.01008403e-01 -3.70068640e-01
-4.49371129e-01 1.07490349e+00 2.57108718e-01 6.43355072e-01
-9.93860126e-01 -7.85795629e-01 1.02169192e+00 -1.99811503e-01
5.34010291e-01 5.16956262e-02 -2.46725455e-01 -1.47943223e+00
-1.60319343e-01 -5.52624464e-01 2.90864646e-01 -1.13154924e+00
-1.22674298e+00 6.84269190e-01 -1.83882937e-01 -1.02231729e+00
4.07670766e-01 -4.44942266e-01 -6.39381111e-01 1.28924727e+00
-2.09309983e+00 -7.60284364e-01 -1.07807481e+00 5.60482025e-01
8.39997351e-01 1.62737846e-01 2.44075030e-01 1.52073145e-01
-6.47415996e-01 1.35938928e-01 -7.56435618e-02 3.51541974e-02
8.06527793e-01 -1.18388283e+00 2.35472217e-01 9.78493273e-01
-4.05232579e-01 6.51898384e-01 5.82128823e-01 -5.28063297e-01
-9.67427671e-01 -1.30681014e+00 5.78286171e-01 -8.58885869e-02
5.20783961e-01 -2.47703731e-01 -1.02985752e+00 2.05086052e-01
2.04118237e-01 2.23168507e-01 -1.79003820e-01 -3.66902888e-01
-9.51434001e-02 -3.05214316e-01 -9.50560331e-01 3.83606046e-01
9.30174947e-01 -6.04722559e-01 -7.94810772e-01 9.92851481e-02
9.23160136e-01 -8.79675150e-01 -5.82541406e-01 7.96928883e-01
2.82343835e-01 -1.34567702e+00 7.77759552e-01 6.00418001e-02
6.16460919e-01 -6.78959489e-01 2.03493521e-01 -1.60828602e+00
-7.57008851e-01 -5.96633494e-01 -6.37085214e-02 9.86481309e-01
7.02086315e-02 -8.60021234e-01 4.06908154e-01 -4.04811138e-03
-5.33540070e-01 -7.75445461e-01 -7.00018227e-01 -1.89790905e-01
-3.74093741e-01 1.70580104e-01 9.46317077e-01 8.30113411e-01
-2.68814981e-01 5.03488839e-01 -8.60694349e-02 4.24157679e-01
5.43517351e-01 6.81566656e-01 2.47065529e-01 -9.29971635e-01
4.96020168e-02 -5.15931666e-01 -2.93218702e-01 -1.56595445e+00
-1.17115535e-01 -6.31709218e-01 2.22950652e-01 -1.66634643e+00
5.76332211e-01 -3.57977450e-01 -8.05171847e-01 2.20144033e-01
-7.66450882e-01 2.52117455e-01 3.74321500e-03 2.79484034e-01
-4.21988457e-01 6.74765110e-01 1.93133807e+00 -2.54836351e-01
-6.58840463e-02 3.04866396e-02 -8.85906339e-01 6.76366687e-01
3.94739270e-01 -4.35128212e-02 -5.26156366e-01 -7.85024583e-01
-4.71582204e-01 1.07146837e-01 5.15684426e-01 -1.21962667e+00
3.14557105e-01 -1.22158892e-01 9.04468298e-01 -7.31990397e-01
1.87114060e-01 -7.18388796e-01 -1.75233141e-01 4.26950783e-01
-1.35121107e-01 -3.90242308e-01 3.78874272e-01 5.40587783e-01
-3.61146957e-01 -9.85262543e-02 9.97134387e-01 -4.89517003e-02
-7.44557917e-01 4.82368588e-01 -7.52378860e-03 -9.81620327e-02
8.55816245e-01 -9.66853574e-02 -8.23706806e-01 -2.50542521e-01
-3.82024795e-01 3.04788262e-01 6.95614457e-01 6.38605833e-01
8.59432936e-01 -1.05836713e+00 -6.90765560e-01 6.63857996e-01
4.13185842e-02 3.22189540e-01 5.31333923e-01 9.74432290e-01
-5.29886186e-01 5.66022038e-01 -2.16199219e-01 -8.04067254e-01
-9.42948937e-01 5.90874493e-01 6.96325600e-01 1.58819705e-01
-3.63011390e-01 1.20667374e+00 1.00474513e+00 3.88677478e-01
1.22191228e-01 -9.23342943e-01 -3.05809587e-01 -3.14737380e-01
5.70952535e-01 9.90111083e-02 -1.52597111e-02 -4.59113806e-01
-5.38110435e-01 6.78002357e-01 -3.88948053e-01 4.34044272e-01
1.29144776e+00 -4.30994928e-01 -2.36945480e-01 5.68489283e-02
1.36585581e+00 1.24217108e-01 -1.71126640e+00 -3.28796297e-01
-4.78811145e-01 -7.83146739e-01 3.53992105e-01 -9.94693339e-01
-1.37094092e+00 1.09069419e+00 7.06302941e-01 1.32296626e-02
1.59129369e+00 3.73352692e-02 9.19457376e-01 -1.45624042e-01
6.65298849e-02 -3.36321831e-01 9.91163328e-02 5.98611534e-01
9.11907256e-01 -9.33916867e-01 -1.80548519e-01 -4.43695933e-01
-3.45286012e-01 1.11396718e+00 7.85250306e-01 -2.57542133e-01
6.23883426e-01 4.07331318e-01 8.67302045e-02 -2.72762746e-01
-4.63617295e-01 -1.64712161e-01 4.46982145e-01 4.35803592e-01
1.47359416e-01 -1.96143791e-01 -7.85231888e-02 6.49858952e-01
3.71510424e-02 7.93162361e-02 3.92251283e-01 5.42892277e-01
-8.40586364e-01 -4.42766100e-01 -2.55928099e-01 4.68432933e-01
-2.21262470e-01 -2.63472766e-01 -1.87162414e-01 2.40574852e-01
5.26879430e-01 9.61502731e-01 1.26822919e-01 -3.31796974e-01
3.13521981e-01 -6.78882897e-01 3.74479920e-01 -6.41593218e-01
-4.66991752e-01 3.39909881e-01 -1.43833280e-01 -5.02181053e-01
-2.37912744e-01 -4.75708365e-01 -1.19211543e+00 2.64183898e-02
-5.40108144e-01 1.90645292e-01 2.02650413e-01 8.04907739e-01
4.48115796e-01 7.06073523e-01 9.17962432e-01 -9.12267089e-01
-3.23689967e-01 -1.10481393e+00 -4.79362011e-01 4.15921837e-01
1.01271927e+00 -8.92787814e-01 -7.54248559e-01 -2.98652470e-01] | [11.344001770019531, -2.7346229553222656] |
018ba0bb-9587-49b9-abdc-98a49c78fd1d | nilm-as-a-regression-versus-classification | 2010.16050 | null | https://arxiv.org/abs/2010.16050v1 | https://arxiv.org/pdf/2010.16050v1.pdf | NILM as a regression versus classification problem: the importance of thresholding | Non-Intrusive Load Monitoring (NILM) aims to predict the status or consumption of domestic appliances in a household only by knowing the aggregated power load. NILM can be formulated as regression problem or most often as a classification problem. Most datasets gathered by smart meters allow to define naturally a regression problem, but the corresponding classification problem is a derived one, since it requires a conversion from the power signal to the status of each device by a thresholding method. We treat three different thresholding methods to perform this task, discussing their differences on various devices from the UK-DALE dataset. We analyze the performance of deep learning state-of-the-art architectures on both the regression and classification problems, introducing criteria to select the most convenient thresholding method. | ['David Gómez-Ullate', 'Daniel Precioso'] | 2020-10-28 | null | null | null | null | ['non-intrusive-load-monitoring', 'non-intrusive-load-monitoring', 'non-intrusive-load-monitoring'] | ['knowledge-base', 'miscellaneous', 'time-series'] | [ 3.01697224e-01 -4.71007228e-02 -1.13812052e-01 -6.94802880e-01
-5.25655508e-01 -4.49519038e-01 6.96547627e-01 9.53641832e-02
-2.56137788e-01 5.25342226e-01 -3.37545127e-02 -1.52448863e-01
1.70906503e-02 -8.33021462e-01 -2.72271067e-01 -1.12457132e+00
1.31301954e-01 5.71628034e-01 -2.74041951e-01 1.26622524e-02
-1.97053015e-01 4.26798284e-01 -1.37488353e+00 1.88172068e-02
6.42655909e-01 1.64182568e+00 -1.44699499e-01 3.23623061e-01
1.52733639e-01 8.91809523e-01 -1.18519688e+00 -2.26322606e-01
1.76768359e-02 -3.11671883e-01 -6.96477413e-01 -5.84482849e-02
2.21107334e-01 -2.99870908e-01 1.48616016e-01 1.00423968e+00
7.37184107e-01 -7.21292198e-02 7.97413290e-01 -1.68219161e+00
-2.30166912e-01 9.43277001e-01 -1.75038636e-01 3.58390450e-01
5.66445470e-01 -1.66420132e-01 8.14221740e-01 1.85282111e-01
-1.35799333e-01 7.29434669e-01 7.58354008e-01 2.12001875e-01
-1.68298149e+00 -6.86632156e-01 -2.97071904e-01 5.77453196e-01
-1.32733464e+00 -3.99121374e-01 1.03307402e+00 -4.72078145e-01
1.36770535e+00 5.33838987e-01 6.65072620e-01 1.39745295e+00
1.00697920e-01 6.60122097e-01 1.25104249e+00 -1.77583262e-01
6.74873590e-01 9.84550565e-02 2.62790889e-01 1.49095654e-01
3.72620910e-01 -3.39196771e-01 -2.12770984e-01 -3.02900821e-01
-5.92564344e-02 -1.22552507e-01 -1.13076888e-01 -1.66028783e-01
-8.14299583e-01 7.95083940e-01 2.85789549e-01 8.15619767e-01
-4.29793864e-01 8.17789882e-02 6.38594866e-01 5.12824476e-01
5.73752820e-01 3.06895256e-01 -7.12888479e-01 -5.69536015e-02
-1.21936584e+00 -9.83642563e-02 1.21182537e+00 3.55566531e-01
6.39894068e-01 2.36780629e-01 4.55175340e-02 4.36945826e-01
1.27796009e-01 3.88523608e-01 6.05929911e-01 -5.50323665e-01
3.66900861e-01 6.03843153e-01 1.88290656e-01 -7.26493657e-01
-1.09752262e+00 -1.70474332e-02 -1.27291942e+00 2.21432865e-01
2.43360236e-01 -5.24504304e-01 -4.78386790e-01 1.68131518e+00
6.60488382e-02 1.00236505e-01 -3.59556258e-01 2.48571187e-01
6.63423479e-01 7.26905346e-01 3.59360665e-01 -7.44530439e-01
1.05283153e+00 -2.01897964e-01 -1.12762189e+00 1.96016937e-01
4.20402050e-01 -4.84862290e-02 6.79970026e-01 7.21612871e-01
-7.19508588e-01 -4.22825634e-01 -1.40788269e+00 -4.35138680e-02
-9.04286563e-01 1.61656916e-01 2.62429625e-01 9.60453570e-01
-9.37852263e-01 8.87945116e-01 -8.86776567e-01 -3.79658073e-01
3.90736043e-01 7.61134565e-01 -9.61245522e-02 6.57705426e-01
-1.08171380e+00 1.34688783e+00 3.40171158e-01 9.95559543e-02
-4.10223544e-01 -3.77538353e-01 -8.41608822e-01 2.70356566e-01
9.53369290e-02 -1.84592381e-01 1.19719315e+00 -9.38718081e-01
-1.64467812e+00 8.99592996e-01 1.50190726e-01 -7.52560556e-01
7.09095359e-01 -7.98619986e-02 -6.32665694e-01 -2.92635381e-01
6.15803301e-02 1.77772552e-01 1.14435279e+00 -7.60197341e-01
-6.46053314e-01 -4.39394981e-01 -4.21734929e-01 -4.36414093e-01
-1.60161331e-01 -5.41251674e-02 2.51718700e-01 -4.30835456e-01
-2.42079630e-01 -5.61005950e-01 2.24944770e-01 -7.41468489e-01
-6.63803339e-01 -8.41970801e-01 1.15746677e+00 -9.70635235e-01
1.16852200e+00 -2.12959433e+00 -1.20571025e-01 8.71741399e-02
6.16046637e-02 2.32956618e-01 2.95397222e-01 1.48643434e-01
-4.13175315e-01 -1.30076572e-01 -1.24244004e-01 -7.09379315e-01
5.76591790e-01 2.51743019e-01 -1.36412084e-01 9.55986559e-01
-8.88028070e-02 1.05359459e+00 -5.98265350e-01 -2.73238748e-01
1.00415528e+00 5.80930650e-01 2.53654242e-01 1.75689071e-01
-1.40148506e-01 3.41170311e-01 -1.10630997e-01 5.73965251e-01
4.76965189e-01 -3.00307460e-02 4.04435724e-01 -6.97932184e-01
1.94819495e-02 5.93557417e-01 -1.05078757e+00 1.30968308e+00
-7.82647669e-01 8.25042844e-01 5.96939623e-02 -1.68020284e+00
9.10593987e-01 6.09660268e-01 1.04518509e+00 -8.31872821e-01
8.21355462e-01 -8.40092748e-02 -2.45709017e-01 -4.65076476e-01
-6.26215860e-02 2.78924480e-02 -5.77563882e-01 3.30697924e-01
3.84696215e-01 6.97429255e-02 1.64783493e-01 -6.79878414e-01
1.15839171e+00 -2.32473761e-01 7.31674373e-01 -5.34180582e-01
4.05091375e-01 -4.24949199e-01 3.87915343e-01 7.07905471e-01
-1.74451366e-01 -1.79794118e-01 4.66528535e-01 -6.46020353e-01
-6.57109976e-01 -8.30080628e-01 -4.55056399e-01 1.25477302e+00
-2.36842707e-01 -7.67828524e-02 -7.81001747e-01 -6.83630943e-01
1.48840427e-01 1.00257051e+00 -8.02261531e-01 -7.04590976e-02
-5.13386309e-01 -1.27794576e+00 2.82347590e-01 4.80292916e-01
6.88752115e-01 -1.24707890e+00 -1.13704133e+00 4.77689236e-01
-1.47564992e-01 -9.95903790e-01 1.20681874e-01 1.22054148e+00
-4.03070301e-01 -1.10200071e+00 -1.73295498e-01 -6.15334988e-01
1.23146877e-01 -5.09080648e-01 1.49518943e+00 -3.03178340e-01
-2.62518406e-01 7.65221938e-02 -1.25445962e-01 -5.72368979e-01
-3.72741282e-01 4.80363339e-01 1.63161412e-01 1.13111228e-01
8.82702887e-01 -1.15497673e+00 -1.40447333e-01 2.80170768e-01
-4.85571235e-01 -4.90591288e-01 2.67272174e-01 2.09002912e-01
2.56632835e-01 5.60373664e-01 5.58018982e-01 -5.62487483e-01
4.86960620e-01 -4.68770266e-01 -1.02343988e+00 3.52728143e-02
-5.87482214e-01 -6.41127229e-02 9.43691552e-01 -6.40843391e-01
-4.80342954e-01 4.40662920e-01 -3.99863064e-01 -7.15908855e-02
-5.07336438e-01 -2.35947013e-01 -8.36946845e-01 1.36663631e-01
2.00177073e-01 9.98784676e-02 -7.13470459e-01 -7.69273877e-01
3.02706599e-01 8.16922367e-01 7.58343399e-01 -2.09567338e-01
4.91425365e-01 4.12035316e-01 1.97497427e-01 -8.18166554e-01
-7.00572729e-01 -4.07230288e-01 -9.47987735e-01 1.85882691e-02
1.16062760e+00 -6.23084247e-01 -1.42230976e+00 6.59155846e-01
-9.05684114e-01 -5.57544708e-01 -6.98452771e-01 -1.81638226e-01
-4.72229451e-01 -4.17886563e-02 -4.48957622e-01 -9.64963377e-01
-6.29285872e-01 -9.68111932e-01 1.20102298e+00 -2.54737493e-02
-6.16312623e-01 -9.75772262e-01 4.31651622e-02 2.82284409e-01
4.28979635e-01 7.23636985e-01 1.00028312e+00 -9.02694225e-01
2.37400353e-01 -5.31361043e-01 5.04246019e-02 5.23689806e-01
4.47299570e-01 -4.40209180e-01 -1.23459852e+00 -4.12750661e-01
6.62483692e-01 -1.10574380e-01 6.51493132e-01 3.96461785e-01
1.27706516e+00 -5.56772113e-01 -3.69990826e-01 6.33563101e-01
1.56380093e+00 2.26124093e-01 4.49770153e-01 1.54990017e-01
5.37748396e-01 7.41358399e-02 -2.43467927e-01 4.30802196e-01
2.07687527e-01 8.87577295e-01 3.87145579e-01 -7.71460757e-02
2.78243393e-01 7.97739252e-02 3.29951167e-01 7.12640643e-01
2.38971278e-01 -3.32050443e-01 -5.51847517e-01 2.77894169e-01
-1.87229967e+00 -9.08194661e-01 -2.79814620e-02 1.83590138e+00
6.47793949e-01 1.62854820e-01 5.52406073e-01 1.04903269e+00
5.17428458e-01 2.92920798e-01 -7.84897029e-01 -3.98151040e-01
-6.02428541e-02 3.15511674e-01 5.83136022e-01 1.98751733e-01
-1.50590551e+00 1.70222633e-02 7.19626665e+00 6.82623327e-01
-1.24705887e+00 5.74934185e-01 6.31067336e-01 -1.73309073e-01
6.44303679e-01 -9.00675178e-01 -6.00803494e-01 9.82119620e-01
1.20523834e+00 9.88551229e-02 5.64304054e-01 1.01474202e+00
2.39518121e-01 -4.64858145e-01 -1.80364084e+00 1.38910270e+00
1.71051379e-02 -7.27623463e-01 -4.10873681e-01 5.26170395e-02
4.79698122e-01 2.80528754e-01 -2.63469011e-01 3.26919854e-01
4.28476095e-01 -9.51664686e-01 5.62209427e-01 3.95557970e-01
6.30640984e-01 -4.49178725e-01 9.37136590e-01 2.75735974e-01
-1.24847579e+00 -3.77357781e-01 1.72612101e-01 -3.25730324e-01
-3.69956344e-02 7.93853581e-01 -4.35425729e-01 2.96426684e-01
1.02925146e+00 6.15430295e-01 -6.09131992e-01 5.33976495e-01
-1.68576106e-01 6.34805620e-01 -8.24373722e-01 -4.44693975e-02
-2.68863916e-01 -4.02620047e-01 -3.64331044e-02 1.28752530e+00
5.98435812e-02 -2.39489093e-01 4.12212834e-02 8.15132320e-01
-1.42607912e-01 -1.58210829e-01 -3.68505716e-01 6.75941765e-01
2.27111384e-01 1.36494684e+00 -7.43011773e-01 -3.40171158e-01
-4.42973316e-01 1.07623494e+00 1.99580174e-02 -4.04181778e-02
-7.80750692e-01 -1.13333315e-01 6.05027556e-01 4.71312329e-02
4.92678523e-01 1.31895557e-01 -4.69781727e-01 -1.05008459e+00
7.33479783e-02 -4.18503106e-01 4.27601039e-01 -7.00169444e-01
-1.57240891e+00 1.29184648e-01 7.88665414e-02 -7.08630979e-01
-4.77383643e-01 -5.65935433e-01 -8.65988851e-01 4.69724983e-01
-1.13964617e+00 -8.93716753e-01 -4.11288768e-01 7.13597119e-01
3.16932827e-01 1.39817417e-01 9.60086226e-01 4.45988953e-01
-8.18887413e-01 2.21570149e-01 3.15944761e-01 3.92792761e-01
5.75768948e-02 -1.86139476e+00 3.18552941e-01 5.36679327e-01
1.38616875e-01 -4.93077964e-01 7.74924934e-01 -1.63392022e-01
-1.17498636e+00 -1.14412379e+00 7.68308222e-01 -7.24969685e-01
6.26958430e-01 -7.52669454e-01 -5.44960260e-01 8.13926935e-01
2.96550661e-01 -1.34149864e-01 6.82840466e-01 9.11174342e-03
1.03165936e-02 -6.44007564e-01 -1.52070057e+00 -1.40331820e-01
4.63786304e-01 -5.45073211e-01 -8.24659765e-01 3.61622334e-01
1.41945094e-01 2.12047949e-01 -1.12678456e+00 3.58678192e-01
3.09727281e-01 -1.00234377e+00 6.86361015e-01 5.66131659e-02
-4.97954816e-01 -2.38297373e-01 -3.88200313e-01 -1.29696429e+00
-2.15703070e-01 -8.90394628e-01 -5.53132713e-01 1.59567904e+00
7.44078821e-03 -6.69164419e-01 5.04169166e-01 5.09330630e-01
5.88418841e-01 -4.24959362e-01 -1.47190118e+00 -6.81675911e-01
-1.41836792e-01 -6.39461517e-01 9.86442804e-01 1.01647568e+00
9.33918357e-02 7.09408820e-01 -2.75098324e-01 1.11877985e-01
5.62168419e-01 -6.89511821e-02 3.06283057e-01 -1.59567940e+00
1.18193217e-02 -6.69605672e-01 -6.12227798e-01 -5.33207953e-01
3.65168273e-01 -6.16394162e-01 4.34158668e-02 -1.29730320e+00
-1.36323884e-01 -8.43827799e-02 -4.97342587e-01 8.25025320e-01
3.76638979e-01 5.28273046e-01 8.91223028e-02 -2.90579218e-02
-6.24810278e-01 1.46145999e-01 -8.99194479e-02 -6.20791495e-01
-1.93200931e-01 2.90344179e-01 -2.91949838e-01 8.43419135e-01
1.02068579e+00 -2.89561421e-01 -3.17975543e-02 -3.99165973e-02
3.72121811e-01 -7.82570615e-02 2.73886681e-01 -1.32592547e+00
-9.74649116e-02 3.05689782e-01 8.01758647e-01 -8.02690625e-01
3.47769678e-01 -1.25982320e+00 3.38132679e-01 4.92068976e-01
-4.45323298e-03 -2.56023668e-02 8.44389945e-03 3.00425217e-02
2.26668805e-01 -1.36520028e-01 8.02904189e-01 6.68515563e-02
-4.79741067e-01 -2.04418764e-01 -5.09387612e-01 -3.73147011e-01
8.98685694e-01 1.71065822e-01 -5.37439436e-02 -3.13037872e-01
-1.10216904e+00 2.58027077e-01 5.96446469e-02 3.53405386e-01
-3.11654657e-01 -1.27404869e+00 -3.88692021e-01 2.17360973e-01
-3.48228216e-01 -1.29407480e-01 -4.55154777e-01 7.76929200e-01
1.18254147e-01 4.09782648e-01 -5.06338710e-03 -7.02229083e-01
-1.08618796e+00 9.36060429e-01 7.24559188e-01 -4.75751609e-01
-3.21720660e-01 4.76711169e-02 -2.82041520e-01 -1.31783988e-02
5.67503035e-01 -8.96919906e-01 -3.24970365e-01 6.86262608e-01
5.46561718e-01 8.64175677e-01 7.50455022e-01 -6.14844263e-01
-4.33193594e-01 4.32626486e-01 5.74965298e-01 5.36783278e-01
1.52153802e+00 -1.28835797e-01 -2.60508657e-01 8.98254514e-01
1.46302831e+00 -6.99375987e-01 -7.92472363e-01 1.38233259e-01
2.91374266e-01 4.70418870e-01 2.97556043e-01 -1.09884548e+00
-1.21936429e+00 6.89893782e-01 1.01125777e+00 1.21162498e+00
1.59268332e+00 1.13138907e-01 7.74469197e-01 5.32064021e-01
4.37400162e-01 -1.33177137e+00 -6.27376854e-01 5.16011789e-02
5.89208543e-01 -1.20560920e+00 1.07468791e-01 -6.16756715e-02
1.54444948e-01 1.04724967e+00 3.97092104e-02 -1.01844355e-01
1.04022384e+00 6.87469959e-01 -4.09108289e-02 -1.25447512e-01
-3.70513499e-01 -2.76573241e-01 4.19973508e-02 9.45721030e-01
9.53656361e-02 4.60207999e-01 -1.14036456e-01 8.11169505e-01
-4.98981357e-01 1.48453057e-01 4.55322340e-02 6.33050621e-01
-3.57334577e-02 -7.82475948e-01 -4.47621822e-01 8.89937341e-01
-5.72671592e-01 3.89097273e-01 -3.07104617e-01 5.85078180e-01
7.20792174e-01 1.36671281e+00 4.04399455e-01 -4.12095726e-01
4.60330397e-01 4.04622167e-01 4.25118357e-01 -1.55923471e-01
-8.06408286e-01 -1.02160677e-01 -2.91955713e-02 -6.59431756e-01
-6.60279453e-01 -7.77054965e-01 -5.74867547e-01 -4.37240899e-01
-3.72081637e-01 -2.08495915e-01 8.35582554e-01 1.28908741e+00
-2.53482133e-01 6.15784168e-01 9.69098449e-01 -1.36865067e+00
-4.67419297e-01 -1.25860012e+00 -8.95139396e-01 6.54707730e-01
5.58890224e-01 -5.32086790e-01 -4.45974857e-01 -5.25736902e-03] | [16.058664321899414, 7.576294422149658] |
6a44a7ab-6ff3-47f7-b92b-30be126efedd | a-nonparametric-bayesian-approach-for-spoken | 1606.05967 | null | http://arxiv.org/abs/1606.05967v1 | http://arxiv.org/pdf/1606.05967v1.pdf | A Nonparametric Bayesian Approach for Spoken Term detection by Example Query | State of the art speech recognition systems use data-intensive
context-dependent phonemes as acoustic units. However, these approaches do not
translate well to low resourced languages where large amounts of training data
is not available. For such languages, automatic discovery of acoustic units is
critical. In this paper, we demonstrate the application of nonparametric
Bayesian models to acoustic unit discovery. We show that the discovered units
are correlated with phonemes and therefore are linguistically meaningful. We
also present a spoken term detection (STD) by example query algorithm based on
these automatically learned units. We show that our proposed system produces a
P@N of 61.2% and an EER of 13.95% on the TIMIT dataset. The improvement in the
EER is 5% while P@N is only slightly lower than the best reported system in the
literature. | ['Joseph Picone', 'Amir Hossein Harati Nejad Torbati'] | 2016-06-20 | null | null | null | null | ['acoustic-unit-discovery'] | ['speech'] | [ 1.46434620e-01 8.48228484e-02 7.50556961e-02 -6.38016105e-01
-1.36420059e+00 -5.55074215e-01 6.49879694e-01 2.54830390e-01
-7.16748297e-01 7.62768567e-01 2.27531388e-01 -2.41312861e-01
2.55979151e-01 -4.00637865e-01 -5.48811972e-01 -6.74766183e-01
-4.40356024e-02 5.53443491e-01 5.34373820e-01 1.63886324e-01
2.26647452e-01 2.91415513e-01 -1.71034038e+00 1.22380391e-01
7.13411510e-01 1.10300756e+00 5.72950125e-01 9.11450922e-01
-2.09900081e-01 1.11204967e-01 -8.83765876e-01 6.48997575e-02
-1.35133743e-01 -3.43380421e-01 -4.04489577e-01 -1.72280699e-01
2.80135244e-01 -1.84817418e-01 1.89704984e-01 9.39040601e-01
6.19715214e-01 4.89619792e-01 8.93476665e-01 -5.91132760e-01
-1.84104308e-01 8.63819718e-01 -1.29292145e-01 1.34082049e-01
2.10917130e-01 -4.78705019e-01 1.14326131e+00 -1.45563066e+00
1.18691109e-01 1.22353566e+00 2.67500043e-01 5.65328836e-01
-1.11982310e+00 -6.11359954e-01 -7.28779510e-02 9.40657854e-02
-1.83020878e+00 -1.14002752e+00 4.90128279e-01 -2.63928294e-01
1.40598321e+00 2.77553797e-01 2.18840074e-02 7.61771679e-01
-4.26022261e-01 7.00313866e-01 9.17138040e-01 -1.03528285e+00
5.06657600e-01 4.35150653e-01 1.40848726e-01 4.04631495e-01
3.71558033e-02 1.09578550e-01 -8.75551105e-01 -2.04840470e-02
4.85566854e-01 -5.57824612e-01 -1.02596834e-01 4.29373801e-01
-9.49472070e-01 8.43935251e-01 -1.23764575e-01 7.95768559e-01
-4.08937424e-01 1.34653747e-01 2.40128994e-01 1.00198388e-01
6.08224571e-01 2.14280024e-01 -4.48149472e-01 -5.22558510e-01
-1.19695735e+00 -3.03805262e-01 9.21200454e-01 1.05273414e+00
5.62073231e-01 5.22380948e-01 1.41033217e-01 1.50984013e+00
5.24584055e-01 8.07017088e-01 4.83702183e-01 -7.18325436e-01
1.37158945e-01 -8.02376568e-02 5.78622799e-03 -3.34725350e-01
1.92253869e-02 -2.46648908e-01 -3.05938751e-01 -2.79797554e-01
2.90732741e-01 -1.66915610e-01 -9.78764355e-01 1.65066397e+00
7.20198601e-02 8.07835609e-02 2.51685679e-01 4.74555969e-01
6.15176380e-01 1.02960062e+00 -7.07010180e-02 -5.72889864e-01
1.27847528e+00 -6.54808700e-01 -9.47471738e-01 -3.40589762e-01
4.24840868e-01 -1.11396551e+00 1.03761113e+00 5.74866295e-01
-9.19489086e-01 -5.01070738e-01 -9.65023458e-01 2.16981232e-01
-2.75836527e-01 3.20947170e-01 2.38078311e-01 1.08627892e+00
-1.04665458e+00 8.39415416e-02 -7.71595657e-01 -5.23749053e-01
-2.19106421e-01 4.30938989e-01 -1.09688096e-01 2.08386406e-01
-1.14757216e+00 7.09870040e-01 4.67565566e-01 -7.27662072e-02
-8.70461345e-01 -2.70296872e-01 -7.19442189e-01 -1.35252746e-02
3.37628931e-01 2.57251024e-01 1.58720922e+00 -5.14818907e-01
-1.77056813e+00 5.69219530e-01 -5.56765676e-01 -5.57970405e-01
3.98173754e-04 -2.90955901e-01 -7.39406288e-01 1.37162328e-01
-2.93757260e-01 5.29630005e-01 7.07790077e-01 -1.15224159e+00
-1.00555444e+00 5.30605856e-03 -6.26753747e-01 8.74850750e-02
-4.97976452e-01 3.10769647e-01 -5.98693192e-01 -5.43332458e-01
7.78814629e-02 -7.80085504e-01 1.80489510e-01 -4.59572285e-01
-3.43700260e-01 -5.59012175e-01 4.89322484e-01 -7.03927457e-01
1.49073994e+00 -2.30950260e+00 -3.71452481e-01 4.86171365e-01
-4.57772374e-01 3.25613827e-01 2.25984603e-01 4.20674860e-01
3.95251870e-01 8.65051299e-02 -3.22268486e-01 -5.88658392e-01
1.93379745e-01 5.16186416e-01 -5.17986953e-01 1.90149829e-01
-1.33437693e-01 3.71302485e-01 -7.65615761e-01 -4.42421943e-01
2.90109307e-01 5.72116315e-01 -4.76421952e-01 3.45060945e-01
-5.80559298e-02 1.20239556e-02 -3.77969556e-02 7.05787480e-01
2.85682499e-01 5.03688276e-01 1.77055046e-01 1.00738972e-01
-3.49984914e-01 7.89809644e-01 -1.07486582e+00 1.47572255e+00
-5.70321202e-01 7.79700756e-01 6.93478286e-02 -8.62665355e-01
1.18656814e+00 6.02781773e-01 -2.65795365e-02 -3.43704462e-01
-1.30595127e-02 7.34208822e-01 2.29903728e-01 -1.30035639e-01
5.73495328e-01 -4.15217578e-01 1.17804797e-03 2.28894517e-01
3.12441379e-01 -2.67643929e-01 5.23512773e-02 -1.49446607e-01
8.34683239e-01 -4.34110582e-01 2.98927367e-01 -3.16404372e-01
4.29178983e-01 -4.15493339e-01 3.92720312e-01 9.75695550e-01
-1.80788502e-01 5.30420482e-01 -1.93234645e-02 3.25825185e-01
-8.72349977e-01 -1.26963627e+00 -4.97702330e-01 1.43221807e+00
-4.32674825e-01 -4.30533499e-01 -8.25116038e-01 -1.20485216e-01
-3.80660236e-01 1.07071733e+00 -6.28634989e-02 1.42404348e-01
-5.50639153e-01 -5.23482263e-01 9.05322969e-01 4.85864520e-01
3.86055708e-02 -1.22608161e+00 -1.65661499e-01 4.48777318e-01
-2.17608884e-01 -1.51049340e+00 -5.00165939e-01 5.17932475e-01
-7.45919466e-01 -2.55386710e-01 -7.56735921e-01 -9.82181430e-01
4.92106587e-01 -1.67899609e-01 8.96223903e-01 -4.23082829e-01
5.55487014e-02 1.74542174e-01 -5.90938330e-01 -5.83425939e-01
-7.04636514e-01 1.80579927e-02 6.60589933e-01 8.57974589e-02
6.54610574e-01 -4.40544814e-01 -2.97630638e-01 3.60082686e-01
-7.12264001e-01 -5.16050160e-01 5.75486898e-01 8.74220550e-01
7.50088990e-01 -2.47863606e-02 8.96986127e-01 -6.65226161e-01
5.81001401e-01 -9.81989726e-02 -6.70722485e-01 9.85715985e-02
-8.10770750e-01 2.20173839e-02 2.46485770e-01 -5.66119850e-01
-1.10254347e+00 2.42720410e-01 -3.41587871e-01 -1.65202528e-01
-4.38652486e-01 4.82639909e-01 -1.28269583e-01 3.68717700e-01
7.24234879e-01 3.50467741e-01 -3.55583429e-01 -7.94378459e-01
2.82697618e-01 1.29596579e+00 5.96907139e-01 -4.61402982e-01
3.44779760e-01 8.27364698e-02 -6.27657473e-01 -1.59096587e+00
-3.86510640e-01 -8.24250340e-01 -4.16942418e-01 4.49831970e-02
7.43243456e-01 -8.07601213e-01 -2.62309074e-01 1.24655239e-01
-1.00015545e+00 -1.74044222e-01 -2.68279076e-01 9.92175996e-01
-5.72052836e-01 3.56694162e-01 -4.73127127e-01 -1.49886930e+00
-4.58048373e-01 -9.34007645e-01 1.02511704e+00 -3.53504866e-02
-4.71336454e-01 -6.68119133e-01 1.21080972e-01 2.38749564e-01
4.50205624e-01 -5.32539725e-01 6.59489691e-01 -1.16662657e+00
-1.99971631e-01 -3.12529087e-01 1.56685844e-01 7.26278841e-01
4.84522641e-01 -1.25541436e-02 -1.56231225e+00 -9.31294188e-02
-7.03977495e-02 -2.13201955e-01 7.26230681e-01 4.65334445e-01
6.56405330e-01 -1.62676573e-01 -2.47625276e-01 -9.43466425e-02
9.36034679e-01 7.66575992e-01 3.20078492e-01 -3.95930380e-01
2.69582599e-01 6.64538205e-01 5.83637118e-01 3.88009101e-01
-5.90042509e-02 6.90707624e-01 -1.80793926e-01 3.61055166e-01
-7.62719214e-02 -1.57549202e-01 8.08899462e-01 1.64221060e+00
2.41876438e-01 -3.83182615e-01 -1.06003475e+00 9.68909323e-01
-1.59259498e+00 -6.26580298e-01 2.49171540e-01 2.43316770e+00
1.15437031e+00 3.87504339e-01 8.45872685e-02 2.09552810e-01
7.77635217e-01 -1.52730703e-01 -2.38590911e-02 -6.41242146e-01
-3.74577306e-02 5.50515056e-01 3.07434767e-01 9.63759661e-01
-9.60193217e-01 1.18368793e+00 6.65226316e+00 1.08150685e+00
-9.12357271e-01 1.83985457e-01 3.11855644e-01 -4.80353869e-02
-1.32229716e-01 -1.51476473e-01 -1.03827941e+00 2.48656839e-01
1.71272945e+00 -5.64860320e-03 3.31341028e-01 7.50864923e-01
3.10145736e-01 -3.39645416e-01 -1.18866420e+00 1.11198878e+00
1.52119845e-01 -6.17589593e-01 -2.13888466e-01 1.67136073e-01
4.20584947e-01 2.13414580e-01 -1.31075075e-02 2.75502414e-01
1.18880518e-01 -1.01995492e+00 5.24561167e-01 3.82700920e-01
9.63565230e-01 -1.00217354e+00 5.66678345e-01 5.38881898e-01
-1.06086659e+00 3.86711031e-01 -4.75596339e-01 2.65465587e-01
2.33489692e-01 6.23048067e-01 -1.68966281e+00 5.65033741e-02
4.66965884e-01 8.31416100e-02 1.80167723e-02 1.05948150e+00
-2.65645981e-01 1.33626068e+00 -9.61489141e-01 -5.60242712e-01
2.23490149e-01 6.50705118e-03 5.54268718e-01 1.67951298e+00
6.62743330e-01 -5.53812459e-02 7.99632818e-03 5.68902016e-01
-1.53488055e-01 3.75249505e-01 -5.38300991e-01 -2.45152891e-01
8.11053097e-01 8.73551309e-01 -6.23809993e-01 -5.01232564e-01
-2.97846228e-01 9.07353580e-01 2.38778372e-03 3.48637789e-01
-3.88882369e-01 -6.18049622e-01 5.44864178e-01 -2.40087673e-01
5.87368369e-01 -4.16984618e-01 2.33650059e-02 -8.70342374e-01
-4.95176874e-02 -6.64021373e-01 -2.72447634e-02 -4.44865823e-01
-1.22027493e+00 7.07245708e-01 6.68903440e-02 -8.45807076e-01
-1.00904143e+00 -6.37844324e-01 -2.06799313e-01 1.00536251e+00
-1.20510328e+00 -6.56642258e-01 3.76331896e-01 1.72408730e-01
8.95462573e-01 -3.06849182e-01 1.31765044e+00 3.24830085e-01
-2.53798097e-01 8.46701920e-01 4.80646610e-01 2.34221026e-01
8.27521026e-01 -1.41223133e+00 3.97406012e-01 8.56057584e-01
7.65169501e-01 7.44369626e-01 7.64753461e-01 -4.34673369e-01
-9.90898669e-01 -7.30222464e-01 1.36491489e+00 -3.81876022e-01
7.43269861e-01 -8.50242913e-01 -1.09164798e+00 3.15883696e-01
2.70042360e-01 -2.52936989e-01 9.68595982e-01 3.53361160e-01
-3.67690444e-01 -2.06266925e-01 -9.26925123e-01 3.96206945e-01
5.31989574e-01 -9.32097852e-01 -8.19955707e-01 1.27762005e-01
7.96806633e-01 1.88627820e-02 -7.17249632e-01 2.82282531e-01
6.14211738e-01 -5.52187204e-01 8.10750961e-01 -8.12043101e-02
-4.77060914e-01 -3.27878565e-01 -8.84510875e-01 -1.30759513e+00
2.34582990e-01 -5.80173790e-01 -1.08100310e-01 1.56844807e+00
9.45612192e-01 -4.78499204e-01 4.19556022e-01 4.17226374e-01
-3.22751433e-01 -2.02011019e-01 -1.47071838e+00 -1.04521501e+00
-4.59669456e-02 -1.06696606e+00 7.87153617e-02 6.10754788e-01
1.54383585e-01 5.17988205e-01 -3.58591825e-01 2.70696133e-01
4.78221387e-01 -1.86433062e-01 3.06881011e-01 -9.93162513e-01
-4.86312419e-01 -2.91026294e-01 -1.38929635e-01 -1.27498794e+00
1.18076056e-01 -5.54748237e-01 7.91647613e-01 -1.10971963e+00
-2.43500292e-01 -3.76160204e-01 -6.31417811e-01 4.48073179e-01
1.06010817e-01 1.73933029e-01 -3.61553691e-02 1.50732072e-02
-2.50235647e-01 4.94235575e-01 3.63662183e-01 -4.40044478e-02
-3.78674269e-01 1.30247682e-01 -2.76643842e-01 7.48912454e-01
9.49846447e-01 -4.83212501e-01 -4.10561055e-01 -2.43142024e-01
-3.61741900e-01 -2.15749249e-01 -4.49273944e-01 -9.27234590e-01
3.71922135e-01 6.76635504e-02 -9.36698318e-02 -8.06112647e-01
8.45727205e-01 -6.17181420e-01 -2.52964914e-01 2.27227598e-01
-2.88037837e-01 -3.35458755e-01 2.34188080e-01 5.68846941e-01
-4.45173025e-01 -5.65470159e-01 6.30293548e-01 6.62153810e-02
-7.09501088e-01 -2.72797108e-01 -8.78760993e-01 -8.73984694e-02
4.59631562e-01 2.74679512e-02 1.70552731e-01 -5.59950709e-01
-6.15199804e-01 -2.86889344e-01 -1.19953245e-01 4.38522607e-01
7.31848180e-01 -1.05392754e+00 -6.91359282e-01 2.25161076e-01
3.85716617e-01 -2.61672109e-01 -2.80139029e-01 3.95176619e-01
-3.02766114e-01 7.51971483e-01 5.04195273e-01 -6.41330540e-01
-1.39605832e+00 7.41044581e-02 9.27073061e-02 3.80244613e-01
-1.04563050e-01 1.04784155e+00 1.96440622e-01 -3.35255176e-01
6.48805737e-01 -6.14957273e-01 -1.36284307e-01 2.64863640e-01
6.13911629e-01 1.68695763e-01 3.39312166e-01 -7.79409766e-01
-6.56312525e-01 1.51503906e-01 -1.62908603e-02 -9.46954012e-01
1.13134205e+00 -2.08040461e-01 5.62374294e-02 1.09464097e+00
1.09614778e+00 3.25467497e-01 -7.13183880e-01 -4.62206393e-01
3.82223248e-01 -2.01773927e-01 2.74165332e-01 -7.92928338e-01
-3.86390507e-01 1.07594168e+00 8.41214001e-01 3.35741103e-01
1.00180912e+00 1.59679338e-01 7.07541764e-01 6.80298030e-01
3.34247977e-01 -1.40845942e+00 -1.37381211e-01 9.22345579e-01
8.35614741e-01 -1.28367627e+00 -3.97048444e-01 -2.95918494e-01
-4.05772984e-01 9.35457647e-01 2.77620167e-01 1.96322948e-01
8.71389568e-01 3.98411125e-01 2.49152973e-01 1.56109378e-01
-8.19296360e-01 -4.91400689e-01 5.23108006e-01 4.95607764e-01
7.50845134e-01 3.56708884e-01 -3.11400533e-01 4.68590677e-01
-3.67119521e-01 -5.20840287e-01 2.84743309e-01 8.15025330e-01
-1.10960674e+00 -1.25310540e+00 -5.01255393e-01 2.44748101e-01
-5.67734182e-01 -5.69620848e-01 -4.31560099e-01 3.34867179e-01
-3.25821340e-01 1.51308489e+00 1.38702556e-01 -2.31543437e-01
2.77234972e-01 6.50324225e-01 1.10676058e-01 -9.69690859e-01
-2.46603757e-01 8.94229591e-01 4.10138756e-01 -2.38801874e-02
-4.49408591e-01 -5.61187029e-01 -1.43930399e+00 1.25871688e-01
-6.89404547e-01 4.04679209e-01 9.90542054e-01 8.50371778e-01
6.89566806e-02 3.02067518e-01 5.76848030e-01 -4.56510127e-01
-3.33044350e-01 -1.38457429e+00 -6.86630547e-01 -1.61147863e-01
3.91700834e-01 -4.35935467e-01 -5.42157233e-01 1.63365722e-01] | [14.461377143859863, 6.681459903717041] |
a65f1d3a-f6ad-45f3-9120-55c9e94255ba | data-efficient-visuomotor-policy-training | 2007.13134 | null | https://arxiv.org/abs/2007.13134v2 | https://arxiv.org/pdf/2007.13134v2.pdf | Data-efficient visuomotor policy training using reinforcement learning and generative models | We present a data-efficient framework for solving visuomotor sequential decision-making problems which exploits the combination of reinforcement learning (RL) and latent variable generative models. Our framework trains deep visuomotor policies by introducing an action latent variable such that the feed-forward policy search can be divided into three parts: (i) training a sub-policy that outputs a distribution over the action latent variable given a state of the system, (ii) unsupervised training of a generative model that outputs a sequence of motor actions conditioned on the latent action variable, and (iii) supervised training of the deep visuomotor policy in an end-to-end fashion. Our approach enables safe exploration and alleviates the data-inefficiency problem as it exploits prior knowledge about valid sequences of motor actions. Moreover, we provide a set of measures for evaluation of generative models such that we are able to predict the performance of the RL policy training prior to the actual training on a physical robot. We define two novel measures of disentanglement and local linearity for assessing the quality of latent representations, and complement them with existing measures for assessment of the learned distribution. We experimentally determine the characteristics of different generative models that have the most influence on performance of the final policy training on a robotic picking task. | ['Mårten Björkman', 'Ville Kyrki', 'Petra Poklukar', 'Ali Ghadirzadeh', 'Danica Kragic'] | 2020-07-26 | null | null | null | null | ['safe-exploration'] | ['robots'] | [ 2.30151579e-01 3.08742702e-01 -3.48175883e-01 -1.12585254e-01
-4.62157726e-01 -5.26452422e-01 1.11149383e+00 -6.87691346e-02
-5.97765267e-01 8.90580297e-01 2.58931369e-01 -2.90772915e-01
-4.88714367e-01 -7.94453442e-01 -8.70980322e-01 -1.17869735e+00
-1.42656922e-01 8.37142110e-01 2.95527373e-02 -1.44711575e-02
2.62610346e-01 7.11223483e-01 -1.69037843e+00 7.10360929e-02
8.31333995e-01 8.55038643e-01 6.78067744e-01 8.06444228e-01
1.13068841e-01 7.90570319e-01 -4.28675830e-01 5.11019409e-01
3.29822719e-01 -6.28617942e-01 -8.28263819e-01 1.41753241e-01
-2.11878791e-01 -2.79917479e-01 -1.60455585e-01 9.31444466e-01
3.09417456e-01 3.60276759e-01 1.01874995e+00 -1.12444830e+00
-4.30507451e-01 8.45798492e-01 6.59909248e-02 -1.25885829e-02
1.66768745e-01 8.26819956e-01 7.33176291e-01 -2.46680856e-01
7.78238356e-01 1.44434965e+00 6.61588311e-02 5.62469184e-01
-1.47759771e+00 -3.69920522e-01 1.75181851e-01 1.77609906e-01
-7.27200747e-01 -3.87124777e-01 3.89875799e-01 -7.88855910e-01
1.03106761e+00 -2.88030356e-01 8.09271038e-01 1.45730746e+00
4.61048543e-01 8.10001552e-01 1.44440544e+00 -5.76075196e-01
8.83005917e-01 -1.37752980e-01 -1.83751240e-01 5.88076711e-01
-1.61177099e-01 9.08653975e-01 -4.94786829e-01 -1.67593002e-01
1.07086611e+00 -1.37411639e-01 1.88064314e-02 -1.01058400e+00
-1.12038004e+00 8.69585276e-01 2.65381753e-01 4.72935319e-01
-8.34192932e-01 5.73376715e-01 3.13820928e-01 4.38637912e-01
-4.49831635e-02 7.04712152e-01 -3.74681413e-01 -3.07431519e-01
-5.85632026e-01 5.53256273e-01 5.93192637e-01 8.34050655e-01
7.45115042e-01 2.17058316e-01 -5.09891987e-01 5.07921278e-01
3.21002394e-01 5.12899280e-01 8.31265092e-01 -1.15273607e+00
1.78907722e-01 3.33053648e-01 2.23405153e-01 -1.94850892e-01
-1.90924495e-01 -2.78686643e-01 -1.33054629e-01 9.31122363e-01
4.35064167e-01 -2.22407669e-01 -1.18122721e+00 1.90271711e+00
2.07987651e-01 -1.28609061e-01 2.14110598e-01 8.26662600e-01
-3.14029343e-02 3.07229280e-01 3.62729907e-01 -3.33508223e-01
7.13781238e-01 -7.78918326e-01 -6.22983813e-01 -5.32152802e-02
5.02152205e-01 -4.06997323e-01 1.02027583e+00 5.43631017e-01
-1.18077719e+00 -7.44349420e-01 -8.74392807e-01 3.62164021e-01
-1.68591306e-01 2.36767858e-01 6.01524472e-01 1.80986285e-01
-1.08422804e+00 1.17515802e+00 -1.24375522e+00 -2.22995013e-01
3.08757335e-01 4.71429497e-01 -2.26101980e-01 2.68160284e-01
-8.27192724e-01 1.16747499e+00 8.20160866e-01 -1.97850585e-01
-1.73660719e+00 -1.61086351e-01 -6.93822026e-01 1.32249549e-01
4.10869211e-01 -6.71025634e-01 1.31465197e+00 -1.09485197e+00
-2.12137008e+00 3.15963179e-01 2.71315664e-01 -5.56148171e-01
6.92162812e-01 -2.70164937e-01 1.64726943e-01 -2.16111988e-02
8.17020312e-02 7.72833228e-01 1.20177627e+00 -1.23040867e+00
-6.98728263e-01 -3.17870498e-01 -7.28254672e-04 3.00855607e-01
4.81465943e-02 -4.68861878e-01 -1.49029180e-01 -3.58206540e-01
-1.76942348e-01 -1.12194371e+00 -3.94688308e-01 -2.71398872e-01
-1.98820591e-01 -4.41114783e-01 4.51252192e-01 -3.10194314e-01
6.97732627e-01 -1.97975993e+00 9.55445766e-01 1.88854501e-01
-1.27597168e-01 1.21363319e-01 -3.95542592e-01 5.81060052e-01
-1.31606847e-01 -3.76103193e-01 -1.43636599e-01 -2.42473140e-01
9.77237150e-02 5.66484630e-01 -5.61105669e-01 4.86248553e-01
5.55055514e-02 9.36357439e-01 -1.23203790e+00 -7.91254044e-02
4.67771232e-01 2.10901484e-01 -4.75188583e-01 5.05023539e-01
-7.75048614e-01 7.93589532e-01 -6.59285307e-01 4.89182286e-02
-5.33912219e-02 2.42787436e-01 4.50638652e-01 3.83869261e-01
-2.27404684e-01 4.75786507e-01 -1.01135325e+00 1.91964984e+00
-4.42681521e-01 3.86254460e-01 -4.05240953e-01 -8.51973116e-01
9.50745225e-01 2.97270775e-01 4.56346214e-01 -6.52452409e-01
3.36950451e-01 1.27342820e-01 3.62820216e-02 -6.35645807e-01
2.83824921e-01 -2.00944915e-01 -6.24910630e-02 5.27852952e-01
6.55856907e-01 -1.70058101e-01 2.41338447e-01 -9.14732516e-02
1.10580969e+00 1.02809691e+00 3.58697593e-01 -3.37520570e-01
1.26629367e-01 -6.87376037e-02 9.15853009e-02 8.42574179e-01
-4.27477211e-02 -1.28182277e-01 8.79895091e-01 -2.13195220e-01
-1.12605500e+00 -1.16906333e+00 1.82820082e-01 1.17928064e+00
-1.70136124e-01 -4.14760523e-02 -6.26338243e-01 -7.01019287e-01
2.20069230e-01 1.11601877e+00 -1.07689226e+00 -4.71341014e-01
-5.11205316e-01 -2.51219392e-01 2.01722324e-01 6.24242544e-01
-1.44737229e-01 -1.68461299e+00 -1.07474744e+00 1.95780665e-01
2.99371064e-01 -6.64430201e-01 5.36228046e-02 8.13943207e-01
-1.13669038e+00 -9.77220237e-01 -4.63285208e-01 -4.39478815e-01
6.50072992e-01 -3.80315244e-01 7.95250058e-01 -4.23248649e-01
1.25227282e-02 4.51036304e-01 -2.80509859e-01 -1.83495522e-01
-8.99731517e-01 -2.59846896e-02 3.13328534e-01 -3.53732765e-01
2.06602335e-01 -7.15515912e-01 -4.43887979e-01 8.19603577e-02
-8.64042640e-01 2.57047024e-02 1.02233636e+00 9.02344704e-01
5.76456070e-01 -3.17754567e-01 3.42183977e-01 -5.25487185e-01
8.89241755e-01 -4.35395718e-01 -7.85247564e-01 1.61019683e-01
-6.72420144e-01 9.44735348e-01 6.04066014e-01 -8.66017818e-01
-8.25365424e-01 3.27409506e-01 7.04552084e-02 -7.61590540e-01
-4.57378030e-01 3.06020409e-01 7.44486302e-02 3.64897072e-01
8.53500068e-01 3.66378605e-01 2.33217865e-01 -4.80147421e-01
6.81465626e-01 2.64106125e-01 4.94629502e-01 -8.14133823e-01
5.10216653e-01 1.11890219e-01 2.96965707e-02 -6.88133776e-01
-2.63341695e-01 -2.09035829e-01 -8.09829056e-01 -3.23035747e-01
8.83568168e-01 -4.15672272e-01 -9.54732478e-01 3.30347240e-01
-8.26538742e-01 -9.36443746e-01 -8.88676405e-01 6.76568806e-01
-1.43441999e+00 1.26016751e-01 -2.68663257e-01 -9.96587396e-01
7.80203640e-02 -1.44377887e+00 9.45006311e-01 -3.12536955e-02
-3.17567140e-01 -8.00837040e-01 5.12801349e-01 -3.85417908e-01
2.16113329e-01 2.03610048e-01 1.05968404e+00 -4.96430963e-01
-6.45264804e-01 2.56999165e-01 2.63219804e-01 3.45341891e-01
6.00447208e-02 -2.01183900e-01 -8.45624983e-01 -3.87561023e-01
-2.81563122e-02 -5.64792812e-01 8.63664448e-01 6.01199508e-01
7.34641850e-01 -2.74349958e-01 -3.22673172e-01 3.76755297e-01
1.38298738e+00 2.99150825e-01 6.10084176e-01 2.37182572e-01
3.15927386e-01 5.13754547e-01 6.59878612e-01 2.37145767e-01
-3.00678074e-01 7.69062817e-01 5.47200501e-01 4.16545868e-01
7.21600130e-02 -5.36236346e-01 5.90891242e-01 1.82371721e-01
-3.58528972e-01 4.85802479e-02 -5.93802810e-01 4.93147016e-01
-2.08831000e+00 -9.68224168e-01 4.46825445e-01 2.31938291e+00
7.61152208e-01 3.36634636e-01 4.54751194e-01 8.23377818e-02
1.46452665e-01 -7.90331960e-02 -7.55987346e-01 -5.63285887e-01
5.63554764e-01 3.21710855e-01 5.03438056e-01 6.27169907e-01
-7.59374499e-01 1.17322218e+00 6.73457050e+00 7.60670722e-01
-1.16199493e+00 -6.43960536e-02 7.18578845e-02 -1.31137380e-02
-2.90684491e-01 -2.50125434e-02 -7.41918504e-01 3.99857044e-01
9.32930171e-01 9.06752348e-02 7.69560635e-01 9.97454822e-01
3.27944160e-01 -2.83348411e-01 -1.31265378e+00 4.87714291e-01
-4.32246774e-01 -1.09837341e+00 4.50956449e-02 3.42497230e-01
5.85584760e-01 1.49023786e-01 1.31038144e-01 5.82879424e-01
8.86802197e-01 -1.03402543e+00 9.72395360e-01 7.79029191e-01
6.58045053e-01 -5.99548578e-01 2.32814059e-01 6.87897563e-01
-5.22020996e-01 -4.22026873e-01 -2.99253225e-01 -7.08821788e-03
-5.72816543e-02 2.30332091e-02 -1.02440906e+00 2.33588755e-01
1.50450587e-01 3.91613096e-01 -1.64509341e-01 7.59391129e-01
-7.00473070e-01 4.98680145e-01 -1.58989012e-01 -3.24749053e-01
5.39901495e-01 -1.95968419e-01 7.51518369e-01 9.06196475e-01
1.12600945e-01 -2.58894026e-01 1.89495102e-01 1.16132283e+00
4.56960469e-01 -1.05729364e-01 -7.22510815e-01 -3.20999146e-01
9.09517929e-02 8.29816043e-01 -6.62272513e-01 -1.44527212e-01
3.71480852e-01 8.57081890e-01 7.19625533e-01 5.11654794e-01
-4.39549804e-01 3.25344145e-01 5.66431701e-01 -2.41695046e-01
6.37935340e-01 -4.83986080e-01 -9.36386734e-02 -6.52442455e-01
-3.20661932e-01 -8.36938918e-01 1.55771552e-02 -6.34335220e-01
-7.59677351e-01 5.64652681e-01 2.44384557e-01 -1.03909576e+00
-1.16090238e+00 -5.58362901e-01 -3.61577749e-01 1.11301529e+00
-1.19303143e+00 -7.51880586e-01 7.24202916e-02 4.70556706e-01
4.77368921e-01 -3.87890220e-01 1.06330454e+00 -4.25436705e-01
-1.50271788e-01 1.72050089e-01 2.35306472e-01 -2.93581456e-01
2.72527039e-01 -1.51122975e+00 1.18253708e-01 5.74952722e-01
1.10153571e-01 5.74940622e-01 1.07342124e+00 -6.35257840e-01
-1.18035746e+00 -6.60606384e-01 2.71174401e-01 -4.03339475e-01
5.22969604e-01 -2.13654637e-01 -6.13349795e-01 7.65792727e-01
5.28749498e-03 -4.24479395e-01 1.44483551e-01 2.14670911e-01
1.49620980e-01 2.58845180e-01 -7.61680722e-01 5.51574051e-01
8.54238927e-01 -3.29533160e-01 -8.09455872e-01 1.56526268e-01
4.88514394e-01 -4.32707429e-01 -5.52212536e-01 6.29344210e-02
6.97321355e-01 -7.28734195e-01 7.54551351e-01 -9.91012514e-01
5.24253786e-01 -2.22708061e-02 1.26266703e-01 -1.63277352e+00
-4.91682351e-01 -6.86263144e-01 -3.22289377e-01 5.29721737e-01
1.65685073e-01 -5.00116587e-01 5.50321341e-01 1.70565784e-01
-1.32531583e-01 -9.41029429e-01 -7.85940886e-01 -7.46843517e-01
1.78570405e-01 -3.37680519e-01 2.75379062e-01 5.16516209e-01
1.57786757e-02 1.95918173e-01 -2.51410902e-01 -1.00709260e-01
4.75736678e-01 2.20302090e-01 7.98155189e-01 -7.98855424e-01
-7.63020754e-01 -6.33525074e-01 -3.75165224e-01 -1.13331711e+00
2.60108262e-01 -8.61455798e-01 4.33084100e-01 -1.40623438e+00
-1.13615669e-01 -4.28969920e-01 -4.44390804e-01 6.17258728e-01
6.46873638e-02 -5.52236021e-01 1.18502013e-01 2.92184025e-01
-4.49411541e-01 7.68058598e-01 1.43359911e+00 2.25574404e-01
-5.79717696e-01 1.33503690e-01 -2.14287996e-01 5.40518939e-01
6.37593150e-01 -6.06843293e-01 -6.37277246e-01 -7.30760396e-02
3.12413536e-02 4.05691892e-01 3.60668242e-01 -9.39909935e-01
-8.92311484e-02 -5.59173226e-01 4.56446201e-01 -3.19877744e-01
9.61394385e-02 -7.10229218e-01 -5.61256669e-02 8.98794591e-01
-7.67064393e-01 -2.46956199e-01 1.72484368e-01 6.50741398e-01
2.72433400e-01 -3.42941791e-01 7.69272804e-01 -5.39353769e-03
-7.64760256e-01 1.13419771e-01 -6.97711527e-01 -3.40616077e-01
1.09555793e+00 3.14153656e-02 2.90693566e-02 -9.85094607e-02
-1.13048935e+00 1.37438998e-01 4.29615498e-01 6.00162864e-01
3.33826482e-01 -1.14639962e+00 -5.13107896e-01 4.14176881e-01
-9.87690166e-02 -3.96635830e-01 -7.65233263e-02 5.56812346e-01
-2.08339259e-01 5.71084976e-01 -7.65920758e-01 -7.06786990e-01
-9.49745417e-01 7.49742329e-01 4.21848476e-01 -5.29001236e-01
-6.29922807e-01 4.96175349e-01 -4.57268860e-03 -2.94542849e-01
2.87935913e-01 -5.81208110e-01 -3.32112223e-01 -1.02371633e-01
1.02983713e-01 2.66841382e-01 -3.08293104e-01 -2.62697399e-01
6.41945153e-02 -1.66672282e-02 1.35031372e-01 -5.82113683e-01
1.29790938e+00 2.58067340e-01 3.36372793e-01 5.89903951e-01
7.49529719e-01 -4.20049757e-01 -1.88532281e+00 -1.85000941e-01
-1.47227108e-01 -3.00561279e-01 2.05985568e-02 -1.03860688e+00
-4.38288599e-01 7.51617908e-01 6.72508717e-01 3.20290588e-03
8.47510338e-01 -1.38847381e-02 2.17467099e-01 3.80731583e-01
6.42092168e-01 -1.17523050e+00 2.62855828e-01 5.31391203e-01
1.10781896e+00 -7.74689794e-01 -7.13427514e-02 3.10887486e-01
-6.97368622e-01 1.10309505e+00 2.35683694e-01 -4.10955071e-01
3.06240320e-01 2.27045402e-01 -1.28370985e-01 -2.84640551e-01
-9.27344918e-01 -5.52719355e-01 5.30845046e-01 6.90922379e-01
-5.77105349e-03 2.74824202e-01 -3.80776286e-01 8.82806629e-02
-1.89637899e-01 4.09679323e-01 6.41410192e-03 9.63057637e-01
-5.53141177e-01 -1.20804024e+00 2.85473131e-02 2.14209169e-01
8.99758488e-02 3.71966779e-01 -2.07624614e-01 7.94162273e-01
7.05085322e-02 3.26430023e-01 -6.11026175e-02 -3.10277641e-01
2.69470215e-01 3.72751087e-01 1.00212467e+00 -6.68677866e-01
-3.99762750e-01 1.64017811e-01 -1.66665912e-01 -8.57585967e-01
-2.03054875e-01 -9.02896523e-01 -1.26937628e+00 1.66192353e-01
-1.33444190e-01 1.62580222e-01 8.68154585e-01 1.30132937e+00
1.28383681e-01 8.31740916e-01 4.22590256e-01 -1.14640176e+00
-1.12329817e+00 -1.21462190e+00 -4.90040481e-01 3.77889901e-01
2.91823685e-01 -1.03986669e+00 -1.69438943e-01 -4.09195200e-02] | [4.213240623474121, 1.7313885688781738] |
33589420-876d-4844-8902-dac302788ab6 | statistical-and-machine-learning-ensemble | 1909.08573 | null | https://arxiv.org/abs/1909.08573v2 | https://arxiv.org/pdf/1909.08573v2.pdf | Statistical and machine learning ensemble modelling to forecast sea surface temperature | In situ and remotely sensed observations have potential to facilitate data-driven predictive models for oceanography. A suite of machine learning models, including regression, decision tree and deep learning approaches were developed to estimate sea surface temperatures (SST). Training data consisted of satellite-derived SST and atmospheric data from The Weather Company. Models were evaluated in terms of accuracy and computational complexity. Predictive skill were assessed against observations and a state-of-the-art, physics-based model from the European Centre for Medium Weather Forecasting. Results demonstrated that by combining automated feature engineering with machine-learning approaches, accuracy comparable to existing state-of-the-art can be achieved. Models captured seasonal patterns in the data and qualitatively reproduce short-term variations driven by atmospheric forcing. Further, it demonstrated that machine-learning-based approaches can be used as transportable prediction tools for ocean variables -- the data-driven nature of the approach naturally integrates with automatic deployment frameworks, where model deployments are guided by data rather than user-parametrisation and expertise. The low computational cost of inference makes the approach particularly attractive for edge-based computing where predictive models could be deployed on low-power devices in the marine environment. | ["Fearghal O'Donncha", 'Bei Chen', 'Stefan Wolff'] | 2019-09-18 | null | null | null | null | ['automated-feature-engineering'] | ['methodology'] | [-3.65090221e-01 -2.88545668e-01 8.96470696e-02 -4.93256241e-01
-6.02231383e-01 -7.05333889e-01 7.58337438e-01 2.37422362e-01
-1.26208425e-01 7.86771417e-01 7.51900300e-02 -1.00935578e+00
-5.03216743e-01 -1.03767121e+00 -2.73047477e-01 -9.24051762e-01
-7.50420570e-01 3.36558044e-01 -1.08688764e-01 -6.15095317e-01
2.91085958e-01 7.50311852e-01 -1.76667249e+00 -2.66498506e-01
8.88058186e-01 1.08057749e+00 1.72419950e-01 9.08713043e-01
-2.04110146e-01 4.12081450e-01 -1.24458641e-01 5.12095988e-01
4.08477396e-01 -2.11369842e-01 -1.14668272e-01 -6.12099946e-01
3.45378816e-01 -1.72955785e-02 2.37522587e-01 3.98238629e-01
4.51600879e-01 1.49413586e-01 4.19148922e-01 -9.39137399e-01
1.45826876e-01 -1.20361559e-01 3.86560448e-02 3.58760476e-01
-5.90438843e-02 4.06893075e-01 7.79864907e-01 -7.78726459e-01
2.16525927e-01 8.28725576e-01 1.29069424e+00 1.85930938e-01
-1.40657580e+00 -5.87223351e-01 -6.56925095e-03 -3.25724393e-01
-1.21949208e+00 -7.66640127e-01 2.53035694e-01 -7.88820982e-01
1.45224690e+00 6.58590138e-01 1.15361202e+00 1.95183501e-01
6.39838696e-01 -1.38551012e-01 1.41984808e+00 -4.97927576e-01
6.06412649e-01 3.30082327e-01 -3.63375515e-01 4.24147457e-01
2.94915825e-01 6.99989796e-01 -5.98997891e-01 -3.08463007e-01
4.16705161e-01 -2.16793288e-02 -9.96725187e-02 1.57029659e-01
-6.80687726e-01 8.71709406e-01 4.19416308e-01 -4.69036913e-03
-5.08505940e-01 1.51478931e-01 1.23504527e-01 5.84689617e-01
1.26939619e+00 6.71063662e-01 -1.44882989e+00 -3.18631858e-01
-1.49219167e+00 4.19318348e-01 1.12497497e+00 6.08552158e-01
8.57476830e-01 6.58700168e-01 6.95703447e-01 3.82029355e-01
1.16681492e+00 1.04267037e+00 2.88919955e-01 -8.42726707e-01
-8.11959878e-02 4.10611868e-01 2.44402155e-01 -9.26610947e-01
-8.00394654e-01 -4.80803102e-01 -6.32815182e-01 6.11755848e-01
4.19875011e-02 -7.96292186e-01 -9.22885776e-01 9.32253659e-01
4.63521153e-01 1.78259954e-01 4.14034694e-01 9.64884400e-01
7.01095283e-01 1.09409547e+00 1.38711125e-01 -1.06122792e-01
9.68130708e-01 -5.44681072e-01 -4.59470958e-01 -3.84905994e-01
9.11879539e-01 -4.69642997e-01 4.06185806e-01 1.24743529e-01
-6.87339962e-01 -3.74544740e-01 -8.51538777e-01 4.88663226e-01
-7.94603050e-01 -2.93064654e-01 9.02291954e-01 6.68228984e-01
-1.55060375e+00 9.18111265e-01 -1.20759952e+00 -5.04074454e-01
6.19143955e-02 3.09067667e-01 -1.16979942e-01 5.33968449e-01
-1.06730711e+00 1.22010469e+00 -6.96653947e-02 4.92355227e-01
-9.66820240e-01 -1.05973434e+00 -9.91926432e-01 3.57262790e-02
-7.37346828e-01 -6.29294991e-01 1.21164024e+00 -1.05857408e+00
-1.60797393e+00 4.41546559e-01 -2.67738253e-01 -4.80956972e-01
2.47938141e-01 -1.91719875e-01 -9.14101601e-01 -1.30007163e-01
-1.56443998e-01 2.01858997e-01 5.36322236e-01 -1.09532022e+00
-8.83973360e-01 -2.00855851e-01 -4.51667488e-01 2.31048018e-01
-2.04846933e-02 1.51475474e-01 2.71532118e-01 -1.52149528e-01
2.58882046e-01 -9.25077081e-01 -6.96363747e-01 3.34653497e-01
2.34707639e-01 1.28340125e-01 1.05827284e+00 -7.26455271e-01
1.10343778e+00 -1.86433661e+00 -5.50409973e-01 4.44149613e-01
-3.03331345e-01 2.12664843e-01 2.09825560e-01 9.89502251e-01
2.71835506e-01 4.82619017e-01 -8.76236185e-02 -1.85934976e-01
-2.19529614e-01 6.52016521e-01 -4.56147105e-01 7.19388247e-01
4.85163361e-01 2.05715701e-01 -8.40851486e-01 -1.22620814e-01
6.03193879e-01 5.68900943e-01 -1.83198288e-01 3.23323488e-01
-3.72723341e-01 8.22877645e-01 -3.98737371e-01 5.80943108e-01
6.67027473e-01 6.03094026e-02 2.81503320e-01 5.66119671e-01
-9.49650109e-01 7.05461919e-01 -1.01372421e+00 1.35195661e+00
-8.60789239e-01 1.05471075e+00 5.17948210e-01 -6.38770640e-01
1.28873432e+00 4.39176649e-01 -1.25057504e-01 -6.70636296e-01
-3.99560452e-01 5.81405938e-01 5.53554529e-03 -8.85918081e-01
4.97568786e-01 -3.57687414e-01 4.44338590e-01 1.76457137e-01
-3.13087702e-01 -5.55565000e-01 -5.32953262e-01 -3.50856990e-01
3.35622251e-01 8.12388897e-01 7.79126137e-02 -1.16554487e+00
1.89167738e-01 6.38440669e-01 4.88260984e-01 4.71431881e-01
3.45756888e-01 3.12920719e-01 -3.07249483e-02 -1.10978711e+00
-9.97738779e-01 -3.22237492e-01 -8.19180846e-01 1.31677628e+00
-3.06930929e-01 -2.86612809e-01 -2.55616128e-01 8.13785717e-02
2.63190150e-01 6.23110712e-01 -6.92019880e-01 4.51285005e-01
-1.10878922e-01 -9.92809772e-01 3.58954757e-01 4.12138700e-01
4.33191322e-02 -6.16970479e-01 -8.34912956e-01 5.53079069e-01
6.37786508e-01 -4.62831587e-01 7.04403996e-01 5.74876964e-01
-1.54691851e+00 -4.94521350e-01 -1.47371620e-01 -3.43503475e-01
5.15151620e-01 -1.67849451e-01 1.30645621e+00 2.81717628e-01
3.59781943e-02 1.64091624e-02 -2.03553066e-01 -9.30753231e-01
-4.34204161e-01 -1.13417327e-01 2.04338163e-01 -4.14406359e-01
2.57562548e-01 -8.06105733e-01 -9.14331436e-01 3.61086071e-01
-4.97525871e-01 -7.42141297e-03 2.94577688e-01 4.65679497e-01
4.48444128e-01 -2.83188939e-01 5.11602938e-01 -4.92132634e-01
-3.94283459e-02 -9.29711461e-01 -1.16053712e+00 -2.03984082e-01
-1.08760500e+00 -3.82262021e-02 4.98073041e-01 1.62558049e-01
-9.35330868e-01 2.32393503e-01 -1.30247474e-01 -8.18055682e-03
-4.17223662e-01 1.17147088e+00 3.90049785e-01 -4.24825311e-01
7.01880276e-01 4.09080297e-01 3.05709362e-01 -7.88384318e-01
-5.55519648e-02 8.75231624e-01 -5.34731448e-02 -4.02102262e-01
6.20590925e-01 4.37728316e-01 2.47545168e-01 -1.19858360e+00
-3.05060863e-01 -2.51742363e-01 -5.25201201e-01 -2.86258876e-01
4.01119262e-01 -1.52369106e+00 -3.25801000e-02 6.16587639e-01
-6.61760092e-01 -7.84053206e-01 -6.50891066e-02 7.10359693e-01
8.52008685e-02 -3.95574778e-01 -9.65966359e-02 -1.36908031e+00
-7.39799619e-01 -6.07021749e-01 1.04018009e+00 5.06191194e-01
-1.84903756e-01 -1.72611785e+00 6.18077159e-01 -3.39755058e-01
1.33396924e+00 5.25038660e-01 3.80791932e-01 -7.04687715e-01
-3.00602615e-01 -3.01489174e-01 7.52638653e-02 1.02594338e-01
2.24475302e-02 7.22113073e-01 -1.32568944e+00 -3.19816977e-01
-3.68291199e-01 -7.21240707e-04 4.48537469e-01 4.93575186e-01
5.68597794e-01 -5.16323507e-01 -3.03344846e-01 9.00179148e-01
1.71592760e+00 -1.86975926e-01 2.41815209e-01 7.43393183e-01
2.06131220e-01 5.81542730e-01 4.59237128e-01 6.72885060e-01
4.18610752e-01 1.85766011e-01 5.77631593e-01 -4.72764134e-01
5.43767989e-01 1.32213756e-01 -5.99024110e-02 7.23499537e-01
-4.61400658e-01 1.87375292e-01 -1.55778360e+00 6.51279688e-01
-1.64963293e+00 -7.02956021e-01 -7.58535683e-01 2.23744893e+00
7.01902390e-01 5.26355170e-02 -3.65646362e-01 -2.69778162e-01
4.34498824e-02 1.23895951e-01 -3.33517373e-01 -9.11236227e-01
1.65398434e-01 2.56029665e-01 1.06110156e+00 4.80141789e-01
-1.01479542e+00 6.85077488e-01 7.16703272e+00 -6.06532358e-02
-1.64954782e+00 1.02455951e-02 4.65982765e-01 2.29661874e-02
-3.92999947e-01 8.51405337e-02 -1.05040359e+00 3.89974833e-01
1.97947216e+00 2.16862583e-03 1.57117933e-01 9.01154697e-01
1.17110682e+00 -2.35969037e-01 -6.26325011e-01 6.37709582e-03
-7.24344134e-01 -1.78079486e+00 -4.59612995e-01 1.51904866e-01
9.22513127e-01 9.77542996e-01 -4.29904997e-01 1.84648499e-01
5.77248573e-01 -1.04875600e+00 5.92466235e-01 9.96072173e-01
9.69969332e-01 -4.97875094e-01 8.81791234e-01 3.72856706e-01
-1.33142912e+00 5.49116880e-02 -2.69278973e-01 -1.06283009e+00
-2.62347490e-01 6.30498588e-01 -6.37372851e-01 4.88595784e-01
1.24471831e+00 7.68422425e-01 -1.90495029e-01 1.15006495e+00
9.90719646e-02 1.17200077e+00 -1.02580404e+00 -2.22164735e-01
4.34570432e-01 -1.17198974e-01 2.48223558e-01 1.49982738e+00
4.96748447e-01 2.31340438e-01 -1.84187759e-02 3.48346174e-01
5.78058243e-01 -2.36066822e-02 -8.75463188e-01 3.02886814e-01
5.87907195e-01 1.37883079e+00 -1.34883299e-01 -2.31955901e-01
-4.31718558e-01 -1.98114460e-04 -3.43311459e-01 3.28449160e-01
-2.72828698e-01 -2.07969010e-01 1.26220739e+00 3.93003464e-01
3.04376125e-01 -6.10487163e-01 -5.13015389e-01 -6.64440334e-01
-5.39097726e-01 -6.00603640e-01 -1.30999297e-01 -8.93810689e-01
-1.09741819e+00 5.33490896e-01 -1.06049716e-01 -1.59195817e+00
-4.11079049e-01 -5.69630504e-01 -1.15591371e+00 1.63013315e+00
-2.03548050e+00 -1.04880559e+00 -3.29727232e-01 2.04764716e-02
2.38171965e-01 -3.18017118e-02 1.56184280e+00 -2.18840107e-01
-3.60716939e-01 -2.18471214e-01 1.14963818e+00 -1.56243905e-01
1.70526087e-01 -1.32349873e+00 7.90767968e-01 6.66811883e-01
-4.08416599e-01 5.36631048e-01 8.90327156e-01 -5.00250876e-01
-1.99280822e+00 -1.30628073e+00 9.99227703e-01 -1.91384047e-01
1.08364689e+00 -8.71836022e-03 -9.52379048e-01 5.11850119e-01
3.55154395e-01 6.59498036e-01 1.03268421e+00 1.82593584e-01
2.45155931e-01 -4.72621560e-01 -9.89278555e-01 4.44407836e-02
2.04492927e-01 -2.81184196e-01 -4.16418463e-01 3.23603511e-01
1.15058459e-01 -5.38729846e-01 -1.11580348e+00 4.41735208e-01
8.16467166e-01 -6.26501203e-01 2.42161974e-01 -8.85961354e-01
3.44049096e-01 -4.46207225e-01 -1.32297531e-01 -1.66035283e+00
-3.37588161e-01 -7.33815253e-01 -2.32216567e-02 9.48997855e-01
9.95514929e-01 -9.10577953e-01 3.48586857e-01 1.00794184e+00
-2.86664933e-01 -5.44019461e-01 -1.11072373e+00 -6.39324903e-01
3.99523079e-01 -4.70886469e-01 5.79154015e-01 1.12751102e+00
-9.94728208e-02 -1.44940898e-01 -1.02888107e-01 9.27824140e-01
3.35034251e-01 3.73916686e-01 8.57593358e-01 -1.76364374e+00
-1.00457378e-01 -1.57226458e-01 -4.68250126e-01 -4.35678929e-01
-3.60597372e-01 -1.90612346e-01 2.43844315e-01 -1.60952163e+00
-7.90992975e-01 -8.51875782e-01 -1.82961226e-01 7.87743211e-01
1.80652633e-01 7.48250633e-02 -3.18205029e-01 3.66562515e-01
3.09183210e-01 2.93162256e-01 8.05856764e-01 4.32178468e-01
-4.44073051e-01 3.97357568e-02 -1.35677069e-01 6.46418333e-01
1.01460898e+00 -6.13370538e-01 7.03533217e-02 -6.42419577e-01
5.90041459e-01 1.20049886e-01 2.01537237e-01 -1.07577288e+00
3.39066237e-01 -4.47621763e-01 4.68462348e-01 -2.19148263e-01
-7.97003731e-02 -1.08158207e+00 5.37119985e-01 5.24063945e-01
1.33600086e-01 4.89417128e-02 9.70358551e-01 2.65859067e-01
-2.19932914e-01 8.58285055e-02 5.36379099e-01 -8.11695531e-02
-9.08287823e-01 5.66090792e-02 -9.00835633e-01 -3.56654823e-01
8.89168620e-01 -1.68202609e-01 -4.36788380e-01 -2.51952648e-01
-5.61519146e-01 4.87823308e-01 5.19059658e-01 1.17112428e-01
2.39851877e-01 -4.33106214e-01 -1.16181672e+00 7.44206131e-01
2.36101579e-02 1.86761916e-01 -9.15650278e-02 8.10776174e-01
-1.25980818e+00 1.76788941e-01 1.82985310e-02 -8.01829100e-01
-9.39450979e-01 -2.40274444e-01 9.81293619e-01 2.77807832e-01
-6.92162156e-01 7.92903781e-01 -4.80842501e-01 -8.36794794e-01
-3.09225172e-01 -4.88366485e-01 5.86011633e-03 1.28614172e-01
6.82057738e-01 1.27889346e-02 2.48585448e-01 -3.19580644e-01
-4.31056350e-01 5.42040646e-01 6.68532908e-01 -9.01033506e-02
1.91210890e+00 -2.32499495e-01 3.94300632e-02 7.38997996e-01
8.86747897e-01 -1.51273936e-01 -1.68713868e+00 -5.42001724e-02
-3.62532437e-02 -2.79819041e-01 9.87058163e-01 -1.15786612e+00
-1.06101251e+00 9.40425694e-01 8.28235269e-01 5.18778622e-01
9.95017886e-01 -5.06342828e-01 2.52941221e-01 5.01524687e-01
8.06151405e-02 -8.78274798e-01 -1.33124077e+00 4.37196225e-01
6.58757269e-01 -1.44446886e+00 3.82262528e-01 3.64724696e-01
-3.39789271e-01 1.50744700e+00 1.36736393e-01 -3.13625224e-02
1.26065898e+00 7.21093595e-01 6.76971257e-01 -1.66582257e-01
-1.03335369e+00 -7.81601891e-02 1.16743483e-01 4.87182289e-01
5.07522404e-01 1.45482868e-01 9.20151174e-03 1.77528813e-01
-1.65279210e-01 2.09396839e-01 1.91786811e-01 1.15126538e+00
-6.74280465e-01 -6.10602200e-01 -5.41242480e-01 6.68066621e-01
-3.81962419e-01 -6.64403379e-01 1.63841560e-01 6.90841258e-01
-9.47674140e-02 1.07982242e+00 5.59350729e-01 -1.53957516e-01
-1.26346290e-01 6.14554323e-02 -4.29036498e-01 -4.84522343e-01
-9.63619769e-01 3.80322933e-02 6.73323870e-01 -2.98764557e-01
-4.64389652e-01 -8.69518578e-01 -1.02405548e+00 -5.93062997e-01
-3.54786485e-01 6.85591519e-01 1.54335582e+00 9.55729365e-01
7.66498744e-01 1.57656342e-01 1.07968426e+00 -1.57071698e+00
-2.33861744e-01 -1.19887865e+00 -8.08861196e-01 -5.91933131e-01
7.44283855e-01 -1.92695782e-01 -7.52878726e-01 8.49296004e-02] | [6.505917549133301, 3.0412042140960693] |
72a9a378-ec63-432f-afb0-3d98764f016a | lie-sensor-a-live-emotion-verifier-or-a | 2102.11318 | null | https://arxiv.org/abs/2102.11318v1 | https://arxiv.org/pdf/2102.11318v1.pdf | Lie-Sensor: A Live Emotion Verifier or a Licensor for Chat Applications using Emotional Intelligence | Veracity is an essential key in research and development of innovative products. Live Emotion analysis and verification nullify deceit made to complainers on live chat, corroborate messages of both ends in messaging apps and promote an honest conversation between users. The main concept behind this emotion artificial intelligent verifier is to license or decline message accountability by comparing variegated emotions of chat app users recognized through facial expressions and text prediction. In this paper, a proposed emotion intelligent live detector acts as an honest arbiter who distributes facial emotions into labels namely, Happiness, Sadness, Surprise, and Hate. Further, it separately predicts a label of messages through text classification. Finally, it compares both labels and declares the message as a fraud or a bonafide. For emotion detection, we deployed Convolutional Neural Network (CNN) using a miniXception model and for text prediction, we selected Support Vector Machine (SVM) natural language processing probability classifier due to receiving the best accuracy on training dataset after applying Support Vector Machine (SVM), Random Forest Classifier, Naive Bayes Classifier, and Logistic regression. | ['Santosh Kumar Bharti', 'NirmalKumar Patel', 'Falguni Patel'] | 2021-02-11 | null | null | null | null | ['emotional-intelligence'] | ['natural-language-processing'] | [ 1.15411632e-01 2.81526655e-01 -3.87084633e-01 -6.12388372e-01
-1.60391077e-01 -6.22800291e-01 5.32680213e-01 2.46784724e-02
4.10147272e-02 6.81007564e-01 3.04757785e-02 -3.13219279e-01
4.76532698e-01 -4.19872940e-01 -1.66038617e-01 -5.16884685e-01
2.22545400e-01 -4.01556045e-02 -6.41216934e-01 -4.09351662e-02
4.46865976e-01 5.52899718e-01 -1.46524668e+00 8.37506235e-01
5.27141213e-01 1.86081469e+00 -9.26671624e-01 7.96617448e-01
-2.01079324e-01 2.06458020e+00 -7.40660608e-01 -1.31329417e+00
-1.30418837e-01 -5.28574407e-01 -6.94214821e-01 -2.12187797e-01
-4.01529297e-02 -3.20558965e-01 4.24645357e-02 1.16406631e+00
2.02095002e-01 -4.65954214e-01 6.64819539e-01 -2.09978390e+00
-8.67562592e-01 5.39715350e-01 -3.46387088e-01 -6.58294708e-02
6.70325875e-01 6.38530403e-02 6.68239772e-01 -6.53406680e-01
6.30689085e-01 1.02702069e+00 9.76345181e-01 7.17541933e-01
-6.45080566e-01 -9.86895919e-01 -4.86227244e-01 2.88658053e-01
-9.82478321e-01 -6.18191719e-01 1.02945709e+00 -7.47362852e-01
7.39099681e-01 4.58726078e-01 5.99667728e-01 1.43066931e+00
6.67687237e-01 7.42847681e-01 1.41779339e+00 -3.16962637e-02
3.64797384e-01 9.12488997e-01 1.10716544e-01 7.47107923e-01
-3.53506327e-01 -6.65420964e-02 -6.55034482e-01 -4.78332967e-01
-2.51101911e-01 1.48442864e-01 9.26548764e-02 4.75825667e-01
-5.03167272e-01 1.00386178e+00 3.03584158e-01 1.22670777e-01
-6.47704601e-01 -8.03419724e-02 8.29594493e-01 7.63897002e-01
7.26417661e-01 2.42683142e-01 -3.50298971e-01 -5.40109754e-01
-8.13733459e-01 -2.12541863e-01 1.31774151e+00 3.02942306e-01
5.02842247e-01 -3.93909998e-02 7.54917972e-04 5.53566992e-01
4.44010645e-01 4.58678722e-01 7.30127037e-01 -6.78207040e-01
-1.59569338e-01 9.69902396e-01 -1.56629793e-02 -1.78288066e+00
-2.28505984e-01 3.05227451e-02 -7.91843951e-01 5.83516598e-01
-4.47091926e-03 -6.40835226e-01 -1.78567603e-01 1.12546813e+00
2.55883515e-01 1.48789376e-01 1.76516265e-01 8.20570588e-01
1.03872526e+00 7.40035117e-01 2.98918039e-01 -4.49505985e-01
1.51149380e+00 -6.31227612e-01 -1.13878644e+00 1.83343261e-01
7.52685666e-01 -9.83452380e-01 6.17943406e-01 8.37938666e-01
-4.96293426e-01 3.45030352e-02 -9.37999904e-01 2.73499973e-02
-6.56632364e-01 2.68715620e-01 7.86889613e-01 1.10321939e+00
-4.05068278e-01 8.29792261e-01 -4.53095399e-02 2.53389385e-02
9.13088262e-01 3.98479313e-01 -6.32066965e-01 4.17070121e-01
-1.13230038e+00 9.45779145e-01 -2.56034583e-01 3.74225199e-01
-3.31807852e-01 -3.32617849e-01 -8.19063485e-01 -4.46346253e-02
-3.07427078e-01 1.03796422e-01 1.10676050e+00 -2.11722159e+00
-1.77909970e+00 1.32742548e+00 -1.20147038e-03 -5.29918313e-01
5.11789143e-01 -8.90122280e-02 -9.35758710e-01 6.75960109e-02
-3.35025191e-01 3.03683937e-01 1.32057011e+00 -7.99874663e-01
-2.84722149e-01 -4.55684781e-01 -4.81233865e-01 -3.05198163e-01
-2.65179276e-01 7.30807185e-01 9.08242404e-01 -1.68760255e-01
-1.85709968e-01 -6.60470784e-01 5.69686949e-01 -3.75934318e-02
-5.46632528e-01 -4.54734564e-01 1.56870866e+00 -1.01865804e+00
1.05464017e+00 -2.27428937e+00 -6.51734471e-01 2.13730022e-01
5.32549500e-01 2.17113137e-01 3.55802923e-01 1.82154685e-01
-2.47941941e-01 2.95637965e-01 3.31288218e-01 -1.06724270e-01
2.26911947e-01 -1.98875908e-02 -4.90664899e-01 7.89377272e-01
3.61628443e-01 8.92794847e-01 -5.14699459e-01 -4.93641883e-01
1.83436312e-02 4.70954984e-01 -6.05247319e-02 2.01704964e-01
2.89213825e-02 9.73308533e-02 -3.69271904e-01 9.50299740e-01
7.58515418e-01 7.33714178e-02 -8.02417751e-03 -2.95800239e-01
-8.50223303e-02 1.23580836e-01 -6.55330837e-01 3.37403655e-01
-3.99010211e-01 1.10787392e+00 2.61549562e-01 -7.76576817e-01
1.50892019e+00 4.22371119e-01 3.18677753e-01 -5.12428343e-01
1.08085310e+00 1.01437084e-01 -2.90976346e-01 -1.07032776e+00
2.96592265e-01 -5.33226907e-01 -2.92036057e-01 6.92559361e-01
-7.99687132e-02 2.58893639e-01 -6.85438752e-01 2.46434305e-02
9.17484581e-01 -2.09089503e-01 3.13060045e-01 3.56226623e-01
6.29650056e-01 -1.69605628e-01 1.96190357e-01 2.01082379e-01
-9.48707879e-01 2.44328529e-02 1.30599344e+00 -6.56800687e-01
-5.77229440e-01 -2.91314542e-01 3.22487243e-02 1.30614567e+00
-1.85582563e-01 3.38479951e-02 -5.61083734e-01 -1.01119459e+00
1.29654800e-05 7.03741252e-01 -6.56912684e-01 -3.25017810e-01
7.70037025e-02 -2.42090479e-01 8.97157252e-01 -9.47937742e-02
5.72582543e-01 -1.33167982e+00 -5.33264220e-01 -1.91858903e-01
7.78077124e-03 -7.42724895e-01 1.21938631e-01 1.03020012e-01
-1.95778206e-01 -1.25292206e+00 -2.11675659e-01 -7.54179955e-01
1.89466804e-01 -4.14255112e-01 5.11729300e-01 2.82373512e-03
-1.56750172e-01 -1.99280214e-02 -5.19251108e-01 -6.59381390e-01
-7.41090834e-01 -3.05322558e-01 -3.87505465e-03 7.24614143e-01
7.45619953e-01 -4.01798129e-01 -3.25025350e-01 1.28698781e-01
-4.80604082e-01 -3.45202893e-01 2.15201959e-01 6.13033295e-01
-3.46003234e-01 -6.45536557e-02 6.99170172e-01 -1.00445199e+00
9.96495128e-01 -9.45531309e-01 -9.70936473e-03 8.83843824e-02
-5.82197249e-01 -4.09721136e-01 6.94836199e-01 -6.14174306e-01
-1.16798186e+00 1.53655216e-01 -3.04854810e-01 -3.58324260e-01
-3.48293155e-01 2.64790893e-01 1.46510139e-01 -2.54357338e-01
7.74263144e-01 -7.03478903e-02 3.24424922e-01 1.03578545e-01
2.80404001e-01 1.65327013e+00 2.33283728e-01 2.07252234e-01
1.17017932e-01 3.90310079e-01 -4.65613186e-01 -5.18078327e-01
-6.75985813e-01 -3.65645438e-01 -6.98610842e-02 -8.80284309e-01
8.23399842e-01 -6.77438140e-01 -1.89225292e+00 8.87362897e-01
-1.44127071e+00 2.99901813e-01 3.27029467e-01 3.02957416e-01
-1.45867184e-01 1.55841470e-01 -8.67866457e-01 -1.37228560e+00
-6.29348218e-01 -5.90222836e-01 6.72819674e-01 3.52357596e-01
-9.37549531e-01 -6.44945204e-01 -1.53660357e-01 5.61658025e-01
4.86850888e-01 5.96163511e-01 5.35498917e-01 -1.12496269e+00
4.02683198e-01 -1.04641461e+00 -3.46123248e-01 6.52228534e-01
-1.25855640e-01 5.92956662e-01 -1.20657778e+00 6.08226359e-01
5.14127314e-01 -8.96238267e-01 3.72196138e-01 -2.20054150e-01
8.96579325e-01 -1.07335055e+00 5.37074469e-02 1.58013538e-01
1.04414117e+00 2.53051370e-01 8.96719873e-01 -3.62373404e-02
3.57443988e-01 8.60183358e-01 3.20303530e-01 9.11011100e-01
6.41955957e-02 1.02983169e-01 7.35136628e-01 1.65225044e-01
5.67892730e-01 -2.19619066e-01 7.72367716e-01 2.07200468e-01
3.38284045e-01 -9.08325091e-02 -4.89215732e-01 -2.26555437e-01
-1.65185118e+00 -1.34732401e+00 -5.25579453e-01 1.40375876e+00
7.31600344e-01 -2.40570121e-02 1.27534106e-01 3.75324458e-01
9.18409467e-01 5.41164689e-02 -4.05026317e-01 -1.43694735e+00
-2.24615350e-01 2.12372299e-02 3.49090725e-01 2.22998545e-01
-1.06778026e+00 7.91519284e-01 5.09832144e+00 6.70178473e-01
-1.77976310e+00 3.07456136e-01 1.29385245e+00 -4.96906713e-02
4.59329188e-02 -3.79645824e-01 -2.09510699e-01 6.12168550e-01
1.02099764e+00 4.68764193e-02 4.27979469e-01 1.43768942e+00
3.43721807e-01 -1.52825460e-01 -8.24584067e-01 1.21121275e+00
5.13950706e-01 -1.28976285e+00 -5.44506550e-01 -4.19992357e-01
4.41274494e-01 -2.99253613e-01 1.26588807e-01 5.03160834e-01
4.40119021e-02 -1.31627154e+00 8.41488302e-01 6.11242056e-01
4.90035474e-01 -5.60904324e-01 1.06585479e+00 3.22757542e-01
-2.51623988e-01 -2.88570136e-01 4.26445231e-02 -7.98147440e-01
-6.87732249e-02 4.96350706e-01 -7.60791183e-01 -2.52312094e-01
4.74474221e-01 8.01704526e-01 -2.13627055e-01 2.99125642e-01
-2.51684964e-01 7.01056004e-01 2.49995589e-02 -7.89806962e-01
-2.56982129e-02 9.53006148e-02 1.42967775e-01 1.20529139e+00
-7.18492270e-02 4.97979932e-02 -5.28891027e-01 8.18906665e-01
-4.56782788e-01 3.90136778e-01 -5.87767124e-01 -5.71698368e-01
1.31909028e-01 1.73895872e+00 -5.16126037e-01 -2.33052611e-01
-1.05332173e-01 1.28779244e+00 6.07725419e-02 -2.88225175e-03
-1.00730598e+00 -5.85264325e-01 4.06289279e-01 -1.24543747e-02
-1.72097147e-01 5.68454981e-01 -8.19324255e-01 -9.15354908e-01
-1.34590730e-01 -7.24702418e-01 1.65894404e-01 -1.04494941e+00
-1.54839909e+00 5.73769271e-01 -9.91127193e-01 -8.76120687e-01
3.47845182e-02 -7.81449914e-01 -1.01694334e+00 6.76864564e-01
-1.07040536e+00 -9.31420147e-01 -3.75036389e-01 5.11373818e-01
4.01033126e-02 -2.05614924e-01 7.86660254e-01 1.61351189e-01
-6.14876807e-01 5.60421586e-01 -3.13302428e-01 4.86031085e-01
5.05707562e-01 -5.80060601e-01 -6.07831895e-01 1.34803340e-01
-3.70229393e-01 2.19681606e-01 6.34567857e-01 -5.99891782e-01
-1.26917088e+00 -7.94677079e-01 1.44201875e+00 -2.94615805e-01
1.01365376e+00 -2.14923292e-01 -4.20610428e-01 4.71369743e-01
2.23907471e-01 -4.80753221e-02 1.14393497e+00 -3.39024104e-02
-5.83176970e-01 -1.21730410e-01 -1.78535628e+00 1.91749647e-01
-2.77703367e-02 -7.20773458e-01 -2.87216038e-01 5.40402472e-01
2.79886484e-01 -2.34247223e-02 -7.73645759e-01 -2.05816492e-01
1.09236300e+00 -1.32122123e+00 1.31016046e-01 -9.92683828e-01
1.24513078e+00 2.25638166e-01 4.66286466e-02 -5.97199023e-01
3.01096976e-01 -8.28510165e-01 -8.31060782e-02 1.56417286e+00
4.94371772e-01 -6.71437144e-01 1.05180371e+00 1.06097579e+00
3.75158072e-01 -7.72221565e-01 -9.49794471e-01 -1.10739656e-01
-2.94187337e-01 -7.83970296e-01 2.88831979e-01 1.58174717e+00
7.14704216e-01 4.11198914e-01 -9.54062521e-01 -2.83320725e-01
1.03669511e-02 -1.78931877e-01 5.66330612e-01 -1.05472481e+00
1.13973930e-01 -6.35269523e-01 -7.60144413e-01 -2.76001215e-01
4.97491598e-01 -8.21678042e-01 -9.95779335e-02 -6.62048221e-01
5.24975779e-03 -1.99762121e-01 1.09211974e-01 7.25361943e-01
5.23991764e-01 5.02531886e-01 4.52199019e-02 1.06016681e-01
-6.86965346e-01 3.68442327e-01 6.11363292e-01 -1.00994147e-01
-1.21228106e-01 4.26776826e-01 -9.12707269e-01 8.92751515e-01
9.24031496e-01 -5.60481191e-01 2.69423991e-01 5.88817537e-01
9.34722245e-01 1.57975391e-01 6.69686377e-01 -4.68319058e-01
6.24505505e-02 -9.65833366e-02 4.78384942e-01 -2.09519237e-01
1.14034846e-01 -9.58730519e-01 2.67161857e-02 5.85415602e-01
-5.97117305e-01 -2.75424868e-01 -2.36315548e-01 2.52146453e-01
-2.28723541e-01 -3.90400022e-01 8.58401477e-01 9.37572643e-02
-4.31734860e-01 -1.03582971e-01 -7.09060311e-01 -5.18207788e-01
1.30216408e+00 -4.59603727e-01 -5.84435999e-01 -7.89476871e-01
-7.95578599e-01 -8.64377916e-02 4.16194238e-02 3.75041097e-01
6.95247352e-01 -1.09303308e+00 -5.50467730e-01 4.73929308e-02
4.41813022e-02 -1.07944298e+00 2.75430441e-01 1.08579683e+00
-6.63180768e-01 -1.98267758e-01 -3.09430599e-01 -1.31796032e-01
-1.70032096e+00 3.44351649e-01 4.37013537e-01 2.39603698e-01
9.61073115e-02 9.11796927e-01 -7.45385408e-01 -2.84498781e-01
2.86975026e-01 5.07588722e-02 -5.18126011e-01 4.83231217e-01
7.39013076e-01 6.22281909e-01 1.45503040e-02 -1.07974732e+00
-4.34005588e-01 -1.71509147e-01 1.09632827e-01 3.28268051e-01
1.03287542e+00 9.10408944e-02 -6.54799640e-01 4.88175809e-01
1.74149334e+00 5.35086393e-02 -5.90693176e-01 3.96245718e-01
-2.02106893e-01 -2.64720559e-01 9.55533087e-02 -1.16965723e+00
-9.06982839e-01 6.71979308e-01 5.15636981e-01 7.00535774e-01
9.75173593e-01 -4.45958227e-03 8.40639949e-01 1.53229997e-01
-3.61697495e-01 -1.17657304e+00 -1.14957104e-02 2.16813684e-01
8.54114056e-01 -1.51477408e+00 -2.15712249e-01 -3.87109995e-01
-1.23284781e+00 1.57098949e+00 4.12729025e-01 -2.27080379e-02
9.74614620e-01 2.50659227e-01 3.17744762e-01 -3.82960141e-01
-7.40845621e-01 4.46350873e-01 -5.31805726e-03 5.00261903e-01
4.21808898e-01 2.13884279e-01 -4.23735261e-01 1.33255529e+00
-4.75342631e-01 3.99859041e-01 5.26330829e-01 5.35602868e-01
-5.28370380e-01 -2.40980521e-01 -3.11145335e-01 6.60068035e-01
-9.49378371e-01 -1.05000352e-02 -1.04937625e+00 1.40886158e-01
5.98992229e-01 1.33518136e+00 9.66793820e-02 -8.86071980e-01
-1.88458622e-01 3.23742181e-01 -2.25012571e-01 -5.08459248e-02
-1.12402260e+00 -6.64337456e-01 4.22043294e-01 -3.88818860e-01
-3.62865001e-01 -4.46330845e-01 -1.13655078e+00 -8.45341265e-01
-3.69507551e-01 3.94043252e-02 1.23044312e+00 1.19191754e+00
3.47151011e-01 -1.80132166e-01 1.25375128e+00 -3.14282328e-01
-2.84153014e-01 -9.65547860e-01 -6.02447093e-01 6.46363974e-01
8.92448649e-02 -2.52752304e-01 -7.02393174e-01 1.99552864e-01] | [12.19404125213623, 6.313703536987305] |
336a0d99-c932-4d8f-a4c4-04c7764e7a79 | a-new-family-of-constitutive-artificial | 2210.02202 | null | https://arxiv.org/abs/2210.02202v2 | https://arxiv.org/pdf/2210.02202v2.pdf | A new family of Constitutive Artificial Neural Networks towards automated model discovery | For more than 100 years, chemical, physical, and material scientists have proposed competing constitutive models to best characterize the behavior of natural and man-made materials in response to mechanical loading. Now, computer science offers a universal solution: Neural Networks. Neural Networks are powerful function approximators that can learn constitutive relations from large data without any knowledge of the underlying physics. However, classical Neural Networks ignore a century of research in constitutive modeling, violate thermodynamic considerations, and fail to predict the behavior outside the training regime. Here we design a new family of Constitutive Artificial Neural Networks that inherently satisfy common kinematic, thermodynamic, and physic constraints and, at the same time, constrain the design space of admissible functions to create robust approximators, even in the presence of sparse data. We revisit the non-linear field theories of mechanics and reverse-engineer the network input to account for material objectivity, symmetry, and incompressibility; the network output to enforce thermodynamic consistency; the activation functions to implement physically reasonable restrictions; and the network architecture to ensure polyconvexity. We demonstrate that this new class of models is a generalization of the classical neo Hooke, Blatz Ko, Mooney Rivlin, Yeoh, and Demiray models and that the network weights have a clear physical interpretation. When trained with classical benchmark data for rubber, our network autonomously selects the best constitutive model and learns its parameters. Our findings suggests that Constitutive Artificial Neural Networks have the potential to induce a paradigm shift in constitutive modeling, from user-defined model selection to automated model discovery. Our source code, data, and examples are available at https://github.com/LivingMatterLab/CANN. | ['Ellen Kuhl', 'Kevin Linka'] | 2022-09-15 | null | null | null | null | ['model-discovery'] | ['miscellaneous'] | [ 1.11577928e-01 6.71136156e-02 -4.14518923e-01 -2.12724403e-01
-1.23968266e-01 -3.06474805e-01 1.54669270e-01 -2.31590301e-01
-1.53539613e-01 9.76858139e-01 -2.13804860e-02 -2.17176408e-01
-6.41447365e-01 -8.60929132e-01 -9.88241076e-01 -9.79383647e-01
-1.04004636e-01 5.94857395e-01 7.72710592e-02 -3.32898796e-01
8.42395872e-02 6.98865771e-01 -1.19826686e+00 -1.10274076e-01
9.76729929e-01 9.84261096e-01 -6.16523288e-02 5.19503415e-01
5.83325744e-01 5.20678997e-01 1.42519698e-01 2.14657083e-01
2.37296000e-01 -9.36682895e-02 -8.41677845e-01 -2.16158241e-01
1.08100981e-01 -3.75715137e-01 -6.29802048e-01 7.77114213e-01
4.21298891e-01 4.96047109e-01 9.77585495e-01 -1.00972652e+00
-1.30980742e+00 5.89817047e-01 -1.68820381e-01 -2.07658574e-01
-1.43315271e-01 3.57169628e-01 9.11354721e-01 -7.19168067e-01
6.75960243e-01 9.61496651e-01 1.04184198e+00 9.12524402e-01
-1.30359161e+00 -2.02863142e-01 -1.66904792e-01 -1.94702283e-01
-1.16587090e+00 -2.39415571e-01 1.10272920e+00 -3.84405851e-01
9.08538163e-01 3.37378949e-01 7.62593031e-01 1.18720055e+00
6.00919545e-01 4.65760320e-01 9.01512921e-01 -3.72515291e-01
4.44855541e-01 -2.21103594e-01 5.50341979e-02 6.77836478e-01
4.55769986e-01 4.09639031e-01 -1.73097074e-01 -4.14121777e-01
1.15078020e+00 -1.32756636e-01 -3.80562335e-01 -5.05896568e-01
-7.04930365e-01 7.45773196e-01 3.35306048e-01 2.56482363e-01
-2.55635411e-01 6.66911721e-01 1.88338563e-01 8.18981901e-02
3.64987612e-01 9.77309346e-01 -6.72044814e-01 9.77448076e-02
-6.26524866e-01 6.86485171e-01 9.92958963e-01 5.62070131e-01
5.83331168e-01 3.28151345e-01 6.02545440e-01 9.43051338e-01
3.84394407e-01 5.08617222e-01 4.24720168e-01 -1.51461625e+00
-1.12845950e-01 4.27667975e-01 5.91049949e-03 -1.04601514e+00
-6.23121977e-01 -4.83342230e-01 -8.26321602e-01 1.97155237e-01
4.16125715e-01 -3.10248494e-01 -9.99602437e-01 1.67610896e+00
3.40480894e-01 -2.10955054e-01 -6.19749725e-02 1.09449697e+00
5.90138018e-01 4.89724666e-01 -5.54520823e-02 9.18656737e-02
6.53197765e-01 -5.21530092e-01 -3.30434680e-01 -2.05180079e-01
5.87535918e-01 -4.03055370e-01 1.04439259e+00 2.84078717e-01
-1.34629226e+00 -2.37909496e-01 -1.10922158e+00 -1.15892619e-01
-5.33480108e-01 -1.29927456e-01 9.95429814e-01 1.91964477e-01
-8.82441163e-01 1.35342026e+00 -1.12179065e+00 -4.82039936e-02
2.26605624e-01 6.69259846e-01 -1.17393874e-01 1.85711339e-01
-1.41408408e+00 1.00474977e+00 3.70955199e-01 7.09012747e-01
-5.84045351e-01 -8.53919089e-01 -5.80501199e-01 -5.70899248e-02
1.49452075e-01 -9.68002975e-01 9.35695231e-01 -8.27288866e-01
-1.56426477e+00 4.83141422e-01 3.50921303e-01 -1.83757305e-01
4.08254921e-01 -8.32409784e-02 7.70508219e-03 1.09424576e-01
-4.91612613e-01 5.78982949e-01 4.09267813e-01 -1.62167192e+00
2.76633203e-01 1.95242956e-01 -3.48412320e-02 -9.22776535e-02
-3.76668781e-01 -5.25406063e-01 2.92633865e-02 -7.50369608e-01
2.92729765e-01 -1.17112839e+00 -4.47201341e-01 1.65852770e-01
-4.39438999e-01 -8.30217004e-02 7.22853363e-01 -4.97604787e-01
7.59926856e-01 -1.76481950e+00 3.83757949e-01 6.61206424e-01
3.31507027e-01 -1.09258942e-01 -3.57767963e-03 4.22164977e-01
1.26065416e-02 3.57106984e-01 -7.32686877e-01 2.79852420e-01
5.01586571e-02 5.55707991e-01 -6.92704022e-02 6.71309769e-01
7.68443570e-02 9.76251543e-01 -7.65705943e-01 -2.95236260e-01
1.19686283e-01 7.65280783e-01 -7.50432849e-01 -8.58622417e-02
-4.81544614e-01 2.81708658e-01 -6.32910013e-01 6.08661592e-01
4.49245751e-01 -2.95840859e-01 2.09460959e-01 -3.69298309e-01
3.83408964e-02 -1.65045485e-02 -1.09589183e+00 1.17991567e+00
-2.80901402e-01 5.15969455e-01 3.79492670e-01 -1.18936062e+00
8.77235055e-01 3.61651957e-01 1.08365357e+00 -4.36333001e-01
4.18584853e-01 6.66650116e-01 1.44285873e-01 -6.93015337e-01
2.57888168e-01 -1.80310234e-01 3.83189842e-02 4.31310564e-01
-1.42912686e-01 -5.65221429e-01 2.26328075e-01 -1.90069258e-01
8.98744285e-01 5.08040965e-01 -4.16625202e-01 -7.02959061e-01
7.92215317e-02 9.80270505e-02 6.00371063e-01 6.43474102e-01
9.33512449e-02 5.03803432e-01 2.72375256e-01 -8.03183854e-01
-1.48950446e+00 -9.84724760e-01 -4.53424513e-01 8.62226903e-01
1.69875070e-01 2.99231917e-01 -7.41721451e-01 3.44847828e-01
4.69511688e-01 3.45614552e-01 -8.86494040e-01 -5.02460063e-01
-1.08782721e+00 -1.06331563e+00 4.61410254e-01 7.29613781e-01
1.99145839e-01 -1.21685553e+00 -5.45048416e-01 2.41718501e-01
2.76842356e-01 -8.25356424e-01 -1.49781689e-01 4.82889324e-01
-1.05614519e+00 -1.06518090e+00 -4.06604826e-01 -8.02963555e-01
8.13440561e-01 -5.63878596e-01 9.06064391e-01 5.31864941e-01
-3.42633635e-01 2.52449334e-01 1.26887009e-01 -5.32349795e-02
-6.90122128e-01 2.20913321e-01 4.07264560e-01 -5.03016710e-01
-2.68221021e-01 -1.04313111e+00 -6.42828763e-01 2.64133632e-01
-1.06194389e+00 8.55683722e-03 4.52651292e-01 8.09691131e-01
6.75184906e-01 8.10148418e-02 6.79793477e-01 -7.97152281e-01
8.10851872e-01 -4.04313922e-01 -3.06619525e-01 6.40826970e-02
-7.99737096e-01 3.19060773e-01 8.51023614e-01 -6.78134799e-01
-9.00734663e-01 9.23551917e-02 -1.58197731e-01 -4.34502244e-01
-4.44380678e-02 8.12481105e-01 -8.46759509e-03 -5.07080257e-01
7.74486959e-01 -2.61938442e-02 3.36408466e-02 -3.84904265e-01
3.16300154e-01 2.33651102e-01 6.25648499e-01 -1.55494249e+00
8.13961625e-01 4.65758204e-01 3.77947509e-01 -6.79289699e-01
-4.45013523e-01 1.72901317e-01 -7.14789450e-01 -3.21559250e-01
4.36240286e-01 -1.21961564e-01 -1.01706588e+00 4.89483237e-01
-8.87844026e-01 -7.92970181e-01 -5.13601899e-01 3.40626627e-01
-7.88534582e-01 1.27647787e-01 -9.74739373e-01 -5.85357964e-01
-3.96771818e-01 -1.10928774e+00 5.75786710e-01 2.62462914e-01
-3.91217798e-01 -1.37846351e+00 -1.10883648e-02 1.48976281e-01
7.00808883e-01 7.43357897e-01 1.14706516e+00 -1.60348982e-01
-3.64083588e-01 -2.01077059e-01 7.17578903e-02 4.16230232e-01
9.04849991e-02 5.54913700e-01 -7.81298101e-01 -1.73380181e-01
-4.79347818e-02 -4.50526327e-01 8.72910202e-01 8.73456657e-01
1.30448949e+00 -5.42867899e-01 -2.56826431e-01 7.91297019e-01
1.49454343e+00 3.23572695e-01 3.89873177e-01 2.59791762e-01
8.21204066e-01 5.18691719e-01 -4.51438338e-01 2.02110857e-02
-1.15226759e-02 2.06513032e-01 4.43935484e-01 -1.42533377e-01
1.63536772e-01 -8.81904066e-02 -1.45710632e-01 9.98390913e-01
-6.47366405e-01 -1.01800956e-01 -8.94062340e-01 4.24769253e-01
-1.83747745e+00 -8.77713084e-01 1.55472741e-01 1.85128307e+00
1.19256449e+00 5.12822270e-01 -9.46709067e-02 8.57564732e-02
3.79137784e-01 -7.89347067e-02 -8.63392353e-01 -4.95679975e-01
-2.92254508e-01 1.91193342e-01 8.42264652e-01 6.55908763e-01
-9.51847196e-01 5.26076615e-01 7.08695889e+00 4.67152596e-01
-1.38525426e+00 -1.91410303e-01 6.63175046e-01 -2.93254852e-03
-5.48913896e-01 -8.83041844e-02 -3.70286256e-01 3.76461625e-01
9.21064258e-01 -4.52256761e-02 6.75555348e-01 7.56802440e-01
4.75279421e-01 1.22896180e-01 -1.00750697e+00 1.87104106e-01
-3.45892280e-01 -1.66207695e+00 -2.20251456e-01 5.85207306e-02
8.71538281e-01 4.25774492e-02 -1.39155658e-02 -1.03640340e-01
4.51559156e-01 -1.17697978e+00 6.82488561e-01 8.48705888e-01
5.13858438e-01 -3.58310044e-01 4.43025172e-01 1.09791771e-01
-9.11676466e-01 -1.26739204e-01 -4.84271199e-01 -1.85537517e-01
2.94721365e-01 5.51028192e-01 -1.39531404e-01 9.08426568e-02
4.84703988e-01 5.38131595e-01 -7.16541857e-02 7.86155164e-01
1.21263973e-01 6.85033143e-01 -6.96799457e-01 -2.23534018e-01
2.66754538e-01 -2.32494324e-01 4.41518694e-01 7.79089093e-01
-6.46261051e-02 1.98356181e-01 1.33282974e-01 1.31799591e+00
-1.90164506e-01 -1.64807037e-01 -4.40175653e-01 -7.25668594e-02
4.23398852e-01 8.15283239e-01 -8.30544829e-01 3.94102633e-02
7.98638016e-02 1.25742212e-01 1.38906077e-01 8.09418917e-01
-7.81433225e-01 -1.76741287e-01 3.62144709e-01 3.18199039e-01
-1.21892475e-01 -5.61276913e-01 -7.22565353e-01 -8.46110404e-01
1.01258673e-01 -4.82354820e-01 -3.22878920e-02 -8.01177859e-01
-1.35248125e+00 1.59447551e-01 3.01559567e-01 -5.81640303e-01
-1.02720000e-01 -1.08364785e+00 -6.20199621e-01 5.49649596e-01
-1.07309401e+00 -1.10924840e+00 -5.63015714e-02 6.86148927e-02
4.23100553e-02 2.55352288e-01 5.86017191e-01 2.91881114e-01
-5.02637565e-01 1.26248389e-01 5.65131545e-01 1.82023585e-01
3.24779898e-01 -1.11422062e+00 4.03089263e-02 2.73056954e-01
-5.64089417e-01 8.93128514e-01 1.16577768e+00 -8.49969208e-01
-1.65138805e+00 -6.80254161e-01 3.79191428e-01 -1.62914917e-01
8.57966602e-01 -1.91147432e-01 -1.26323342e+00 4.43241090e-01
-1.38832733e-01 1.74117610e-01 2.88064569e-01 3.62976454e-04
-1.29528910e-01 -3.50800566e-02 -1.02758992e+00 5.84471285e-01
9.75402117e-01 -2.46830359e-01 -1.26388788e-01 5.92992783e-01
5.44524133e-01 -5.86521745e-01 -1.37771881e+00 6.60489023e-01
1.08293974e+00 -5.00602841e-01 1.19651103e+00 -8.35760951e-01
7.83136666e-01 1.34353429e-01 1.62681192e-02 -1.10761726e+00
-3.70133758e-01 -6.32103086e-01 -3.10301274e-01 6.21415854e-01
6.27435029e-01 -7.21691012e-01 8.58226299e-01 1.33042598e+00
-5.08737981e-01 -1.63419163e+00 -8.06355953e-01 -8.35888445e-01
7.70812809e-01 -2.18856812e-01 3.81813884e-01 1.02136981e+00
-1.11862563e-01 -2.35959411e-01 -2.86916196e-01 -9.49471593e-02
5.83291590e-01 2.23691687e-02 7.58344680e-02 -1.35260141e+00
-2.84697443e-01 -7.46813953e-01 1.25782371e-01 -6.47115231e-01
3.73285919e-01 -7.31906652e-01 1.56535372e-01 -1.58236682e+00
-1.54727772e-01 -8.23434114e-01 -1.24562219e-01 5.80523551e-01
3.04298908e-01 4.23014313e-02 -3.20958555e-01 4.57179874e-01
2.79877186e-01 4.54454452e-01 1.63094306e+00 -2.25115821e-01
-4.11501080e-01 -3.76258790e-01 -5.92543185e-01 8.47595513e-01
1.05446029e+00 -2.84539998e-01 -3.43470961e-01 -5.45262277e-01
5.01494884e-01 -1.35646179e-01 6.24245822e-01 -8.87357295e-01
1.85470492e-01 -8.33267987e-01 5.59683323e-01 1.41753882e-01
3.13678473e-01 -8.26565981e-01 4.17028099e-01 5.22505581e-01
-5.85489810e-01 -1.20049313e-01 2.05700263e-01 2.04198465e-01
3.41069728e-01 -3.52722347e-01 9.47397947e-01 -1.82563648e-01
-2.32845992e-01 3.48235965e-01 -3.77684414e-01 3.13583426e-02
4.78147656e-01 -3.98086220e-01 -5.53000152e-01 -1.25578552e-01
-8.62028480e-01 1.31606981e-01 7.01051056e-01 7.99530149e-02
4.85181779e-01 -1.18162477e+00 -3.46997231e-01 1.11213200e-01
-6.00582004e-01 2.27704868e-01 2.35257179e-01 5.78472674e-01
-8.47279787e-01 1.62891492e-01 -2.79676467e-01 -2.75697052e-01
-2.95091510e-01 2.67227709e-01 9.84380782e-01 2.59255189e-02
-4.91418540e-01 6.74374521e-01 -2.23320365e-01 -5.33679724e-01
-2.05207258e-01 -5.50277650e-01 3.57753098e-01 -5.62463522e-01
-4.44968790e-01 4.85977799e-01 -1.22209959e-01 -5.01320124e-01
-1.41992435e-01 7.01597333e-01 4.39519197e-01 2.16140434e-01
1.70309329e+00 2.13233814e-01 -2.51903355e-01 3.03346008e-01
1.08856201e+00 -2.79292196e-01 -1.31575847e+00 7.32133314e-02
-2.21374556e-01 1.38315573e-01 5.86749725e-02 -5.93471944e-01
-1.34922278e+00 5.25075972e-01 7.45724812e-02 3.37057501e-01
7.82320201e-01 -4.21199128e-02 9.86806929e-01 6.60346806e-01
-2.09866211e-01 -1.57344532e+00 -4.45200838e-02 5.27380586e-01
1.05343282e+00 -8.21129084e-01 4.11367327e-01 -3.64106834e-01
1.17855959e-01 1.40300655e+00 8.14746320e-01 -5.48724353e-01
1.00935459e+00 7.02457309e-01 -2.16525361e-01 -4.06738460e-01
-6.08038425e-01 4.89257365e-01 2.66369760e-01 3.30589801e-01
5.72404921e-01 2.23480631e-02 -2.73676544e-01 3.69353563e-01
-2.56359667e-01 7.28524029e-02 4.37645465e-01 1.11123025e+00
-5.23341060e-01 -9.50045884e-01 -3.08158457e-01 7.32388318e-01
-2.25302950e-01 1.11248694e-01 -2.30000541e-01 9.34008658e-01
2.68166885e-02 3.02050531e-01 -2.10751608e-01 -2.10095212e-01
2.38680661e-01 3.80990840e-02 5.31446338e-01 -1.90715730e-01
-3.18067223e-01 7.04158396e-02 5.06117828e-02 -1.76143795e-01
-5.92249572e-01 -4.59385246e-01 -1.59282506e+00 -5.82959533e-01
-3.86131346e-01 2.21731633e-01 4.56869751e-01 1.11582792e+00
1.99620292e-01 4.42913622e-01 2.66816467e-01 -1.20620263e+00
-5.67976654e-01 -8.13103616e-01 -4.39010769e-01 4.18481886e-01
2.56442577e-01 -1.04538822e+00 -5.46486080e-01 3.75123978e-01] | [6.380465030670166, 3.4253063201904297] |
e875b797-b93b-4d05-b351-c6ee7d23446f | investigation-on-combining-3d-convolution-of | 1903.04176 | null | http://arxiv.org/abs/1903.04176v2 | http://arxiv.org/pdf/1903.04176v2.pdf | Investigation on Combining 3D Convolution of Image Data and Optical Flow to Generate Temporal Action Proposals | In this paper, several variants of two-stream architectures for temporal
action proposal generation in long, untrimmed videos are presented. Inspired by
the recent advances in the field of human action recognition utilizing 3D
convolutions in combination with two-stream networks and based on the
Single-Stream Temporal Action Proposals (SST) architecture, four different
two-stream architectures utilizing sequences of images on one stream and
sequences of images of optical flow on the other stream are subsequently
investigated. The four architectures fuse the two separate streams at different
depths in the model; for each of them, a broad range of parameters is
investigated systematically as well as an optimal parametrization is
empirically determined. The experiments on the THUMOS'14 dataset show that all
four two-stream architectures are able to outperform the original single-stream
SST and achieve state of the art results. Additional experiments revealed that
the improvements are not restricted to a single method of calculating optical
flow by exchanging the formerly used method of Brox with FlowNet2 and still
achieving improvements. | ['David Münch', 'Michael Arens', 'Patrick Schlosser'] | 2019-03-11 | null | null | null | null | ['temporal-action-proposal-generation'] | ['computer-vision'] | [ 3.24535817e-01 -7.53147602e-02 -3.32802571e-02 -4.84098047e-02
7.61271045e-02 -2.29427502e-01 1.10618854e+00 -3.78974259e-01
-7.67734408e-01 6.35091841e-01 4.19238865e-01 -2.66134460e-02
-1.97825789e-01 -4.75543499e-01 -3.13930571e-01 -6.29310071e-01
-4.02532548e-01 7.29255751e-02 7.80095398e-01 -3.69748324e-01
4.97241646e-01 5.91466129e-01 -1.81802368e+00 6.13839328e-01
3.04922968e-01 1.07558608e+00 -1.56218216e-01 1.05369771e+00
-3.11138660e-01 1.06915176e+00 -5.45789301e-01 -3.08896542e-01
5.79709888e-01 -6.43097341e-01 -1.04950643e+00 2.58953184e-01
7.87439466e-01 -6.66810870e-01 -5.15397370e-01 6.21975780e-01
3.38443726e-01 3.40776503e-01 3.15402329e-01 -1.19660103e+00
-1.96495757e-01 4.20062304e-01 -7.17715546e-02 7.12517023e-01
5.35132647e-01 5.07211447e-01 7.08410740e-01 -8.21170092e-01
8.29208791e-01 1.50002909e+00 5.55044532e-01 6.29208267e-01
-1.12824154e+00 -2.65994281e-01 3.54668170e-01 7.04583764e-01
-1.06252670e+00 -6.49072886e-01 5.25224447e-01 -4.26196903e-01
1.25076163e+00 -5.57332672e-02 9.67126250e-01 1.20446992e+00
1.94457561e-01 1.01503074e+00 9.08011138e-01 -2.90514827e-01
2.32616290e-01 -2.41096467e-01 -2.11844191e-01 7.24650562e-01
-2.86028460e-02 3.83447587e-01 -8.95368576e-01 1.36656478e-01
1.00678420e+00 -1.63316324e-01 -2.74497330e-01 -3.82277042e-01
-1.65840721e+00 6.84426188e-01 4.58650470e-01 6.49669230e-01
-5.62981367e-01 3.79610568e-01 6.72162116e-01 2.97211587e-01
3.36362362e-01 2.92713255e-01 -3.12883049e-01 -4.77692544e-01
-1.07654643e+00 5.22035360e-01 7.43006945e-01 6.24985933e-01
5.22138178e-01 2.43434235e-01 -4.89390314e-01 1.70771182e-01
2.75563240e-01 -7.87287056e-02 6.89045727e-01 -1.36745405e+00
4.65627730e-01 4.95467395e-01 3.68504435e-01 -7.41273880e-01
-4.36056495e-01 -8.35799351e-02 -5.46239734e-01 3.87714982e-01
7.13058591e-01 -7.30155781e-02 -9.27641869e-01 1.37372422e+00
2.76708901e-01 5.99356353e-01 1.39361069e-01 1.10508215e+00
6.23708904e-01 4.47739989e-01 1.44750699e-01 -9.98269618e-02
1.08574653e+00 -1.40785444e+00 -5.70838034e-01 1.35617703e-01
5.52467823e-01 -6.96129978e-01 5.46188116e-01 6.05907857e-01
-1.25623965e+00 -9.06191349e-01 -1.01388133e+00 -9.82075110e-02
-3.99260372e-01 7.64136612e-02 6.58942521e-01 5.22633135e-01
-1.41348517e+00 1.01944590e+00 -9.16921437e-01 -5.42358279e-01
4.92454082e-01 2.30428830e-01 -4.10258651e-01 1.11287730e-02
-1.20160282e+00 8.68589103e-01 3.99095297e-01 4.53033864e-01
-1.07995665e+00 -5.40075004e-01 -6.40895784e-01 2.09948160e-02
7.42110983e-02 -9.14843857e-01 1.31834424e+00 -1.31546593e+00
-1.95021284e+00 5.10081887e-01 -1.27787992e-01 -9.82676148e-01
1.04797208e+00 -3.48957509e-01 -1.49812967e-01 8.02841842e-01
-5.99876009e-02 1.17416561e+00 9.72422481e-01 -6.44921005e-01
-7.91837811e-01 3.32606472e-02 3.36993307e-01 1.61424652e-01
-2.13174760e-01 2.15349793e-01 -8.73405114e-02 -5.95777452e-01
-2.43593201e-01 -7.52502024e-01 -4.20848846e-01 4.09721315e-01
-2.29233271e-03 -2.30769113e-01 8.05142283e-01 -2.40286544e-01
1.07531750e+00 -1.87674320e+00 3.64636034e-01 -3.39631468e-01
1.93670369e-03 7.68470705e-01 -4.09092367e-01 4.31507587e-01
-1.67822316e-01 -1.98944986e-01 -1.20813519e-01 -5.13340950e-01
-1.51869252e-01 2.82073110e-01 -1.51743352e-01 4.57886547e-01
4.87922072e-01 7.38666475e-01 -1.16629052e+00 -5.24788618e-01
6.31027639e-01 6.49870038e-01 -5.18062532e-01 2.14425802e-01
-1.47904262e-01 6.42490864e-01 -3.45246702e-01 2.69616306e-01
4.53353524e-01 -7.05069378e-02 -4.22422551e-02 1.60761569e-02
-4.28459883e-01 2.98690766e-01 -1.30010605e+00 2.05364633e+00
-1.23679847e-01 6.30117595e-01 -2.40059450e-01 -8.75496447e-01
7.61911333e-01 6.72341645e-01 8.21994245e-01 -5.12092590e-01
1.15850262e-01 1.03902653e-01 1.37560576e-01 -6.91554368e-01
5.98412097e-01 -2.12560833e-01 5.03209531e-01 5.31621456e-01
4.11134958e-01 2.68654048e-01 8.66567969e-01 9.92703289e-02
1.32413447e+00 7.71754503e-01 1.46963194e-01 -6.02526255e-02
1.09740293e+00 -1.06378861e-01 3.74296069e-01 6.46137655e-01
-8.06182563e-01 8.22645843e-01 5.08046210e-01 -9.43321824e-01
-9.11646485e-01 -8.76416862e-01 2.80187994e-01 9.26176071e-01
-3.80885005e-02 -4.75701392e-01 -5.43057084e-01 -8.06060791e-01
-1.94161654e-01 3.40394557e-01 -6.64106667e-01 1.35282680e-01
-8.79414380e-01 -3.36408138e-01 6.87842667e-01 7.16793776e-01
7.30631709e-01 -1.38684702e+00 -1.39541006e+00 3.55113417e-01
-1.33904427e-01 -1.34528816e+00 -1.57390341e-01 -1.94474205e-01
-1.11410785e+00 -1.33968425e+00 -1.05696189e+00 -2.01208398e-01
3.67288858e-01 2.77396649e-01 9.17811573e-01 6.66808560e-02
-2.91174561e-01 5.21636903e-01 -6.73618853e-01 -5.72435409e-02
-4.25901234e-01 -1.83227286e-01 -1.06179327e-01 5.05088687e-01
2.64277995e-01 -5.08170843e-01 -8.61239076e-01 1.62840500e-01
-1.06175160e+00 -4.17517424e-02 4.87135440e-01 5.44180036e-01
4.68074940e-02 -3.78158867e-01 3.59783500e-01 -3.03055376e-01
2.74908096e-01 -2.44577855e-01 -4.86486614e-01 -1.25765458e-01
-2.90836543e-01 9.45891961e-02 6.74216688e-01 -3.79267722e-01
-1.31231368e+00 2.64572352e-01 -1.78278014e-02 -8.29003334e-01
-4.78708595e-01 -1.20733313e-01 4.65512186e-01 -1.39111772e-01
5.32932937e-01 1.43878266e-01 2.54927725e-01 -4.31979269e-01
5.06885588e-01 1.00724749e-01 3.82638067e-01 -1.44315138e-01
4.01931018e-01 8.64050984e-01 1.10585444e-01 -6.72031283e-01
-4.06832397e-01 -5.79462588e-01 -1.04775000e+00 -6.44812942e-01
9.47088301e-01 -5.57053745e-01 -6.97957516e-01 9.23786223e-01
-1.49264026e+00 -3.23330730e-01 -5.41581213e-01 6.83101475e-01
-9.45807576e-01 5.83774328e-01 -6.89434588e-01 -7.70600677e-01
-1.08303554e-01 -1.23679161e+00 1.00592208e+00 2.20520943e-01
-1.53864622e-01 -9.98150229e-01 1.41195655e-01 1.37606069e-01
6.57535493e-01 1.46298319e-01 3.33580971e-01 -4.41998154e-01
-7.57956803e-01 -1.74244633e-03 -1.32055998e-01 6.01504266e-01
-6.57609254e-02 1.38404414e-01 -8.82776916e-01 -2.28842244e-01
-1.61544442e-01 -2.10102707e-01 1.10813498e+00 4.43238258e-01
7.93885767e-01 7.93801472e-02 2.36791209e-03 5.50290883e-01
1.20393169e+00 1.30586758e-01 9.06972945e-01 4.54385251e-01
3.50741416e-01 6.64120734e-01 4.28200364e-01 5.30654907e-01
8.23951066e-02 6.64955199e-01 6.93353534e-01 1.75056562e-01
-4.07402903e-01 -4.84418161e-02 7.00284481e-01 3.28223318e-01
-6.45940900e-01 -1.24738842e-01 -4.84843343e-01 5.93782246e-01
-1.92942238e+00 -1.31684327e+00 -1.92579597e-01 2.07721114e+00
3.45092416e-01 1.81412295e-01 5.03281176e-01 2.91801184e-01
5.50838113e-01 6.72200203e-01 -2.42136285e-01 -5.24704099e-01
-5.48069887e-02 3.69940311e-01 3.65284890e-01 4.53949034e-01
-1.32800651e+00 9.05378401e-01 7.09304476e+00 2.28246599e-01
-1.10802567e+00 3.33692990e-02 1.32233158e-01 -3.11148882e-01
1.61753833e-01 1.05128296e-01 -7.70321608e-01 3.91608089e-01
1.09907079e+00 6.76230416e-02 1.54675692e-01 2.84477115e-01
5.21641314e-01 -3.07301521e-01 -1.01440358e+00 6.55547678e-01
4.78551053e-02 -1.50062752e+00 1.55301794e-01 2.86276788e-02
5.99145591e-01 1.96121596e-02 -2.47697726e-01 5.26047498e-02
-1.03888318e-01 -7.02838242e-01 8.34779382e-01 8.89002383e-01
3.15535754e-01 -3.38349491e-01 6.47364855e-01 5.90319037e-02
-1.25551283e+00 -4.60737348e-01 -1.58060655e-01 -3.22159708e-01
6.11524165e-01 1.22869082e-01 -5.87747753e-01 5.73386788e-01
6.73130989e-01 1.19012642e+00 -5.68762302e-01 1.41905379e+00
-3.06854248e-01 4.01389629e-01 -1.93636805e-01 4.51809131e-02
8.14184070e-01 -3.00641283e-02 8.32997441e-01 1.32011187e+00
1.19241059e-01 -9.22787338e-02 -1.33269772e-01 7.83512294e-01
4.30438697e-01 -1.99850082e-01 -4.99060065e-01 9.43704098e-02
-1.91392586e-01 1.23122966e+00 -6.52615547e-01 -5.48412442e-01
-4.71248806e-01 9.70416784e-01 -4.32727486e-02 4.11288142e-01
-7.54320621e-01 -1.77314907e-01 7.42137671e-01 1.23817801e-01
7.46990442e-01 -4.11293834e-01 1.52356476e-01 -1.13917029e+00
-1.71531156e-01 -4.99697864e-01 4.22312468e-01 -7.89852262e-01
-9.46226537e-01 6.75331175e-01 3.15502137e-01 -1.55850720e+00
-5.81546605e-01 -8.02456319e-01 -6.30857766e-01 6.80494905e-01
-1.83355689e+00 -9.12355065e-01 -2.75205553e-01 7.93289542e-01
7.95297742e-01 -1.50278687e-01 7.27228820e-01 3.15724790e-01
-3.17599535e-01 1.81665257e-01 -3.49756867e-01 -2.53530480e-02
5.46552241e-01 -1.03641927e+00 5.30942023e-01 1.13382328e+00
1.49324849e-01 3.25356692e-01 5.15626848e-01 -2.68440604e-01
-1.03772771e+00 -7.86441743e-01 1.02740109e+00 -1.50999472e-01
7.61031091e-01 1.13565832e-01 -8.20902824e-01 6.04511321e-01
5.07623672e-01 2.56765217e-01 2.44957790e-01 -5.65543473e-01
-1.27816245e-01 -4.11799736e-02 -9.59794879e-01 5.05753696e-01
1.27430975e+00 -1.31072095e-02 -7.09432662e-01 1.02684632e-01
4.91894394e-01 -3.13536257e-01 -7.68083751e-01 2.40787193e-01
5.70696175e-01 -1.66859317e+00 9.37440574e-01 -8.94709826e-01
5.53761065e-01 -3.60260218e-01 1.62231386e-01 -9.75191534e-01
-1.60259336e-01 -9.33090568e-01 -3.41427147e-01 6.25373900e-01
-9.33615118e-02 -5.42117417e-01 9.99666989e-01 1.02870904e-01
-3.24327946e-01 -7.09833145e-01 -1.11751020e+00 -6.09732091e-01
-8.22064057e-02 -4.02878672e-01 2.08629131e-01 3.11722308e-01
-2.74317473e-01 1.07418910e-01 -3.92286181e-01 -4.62712914e-01
3.93583983e-01 -1.50925532e-01 7.31996536e-01 -1.02395630e+00
-2.13846564e-01 -7.33832359e-01 -8.88376772e-01 -1.29381645e+00
1.40560329e-01 -5.72763860e-01 3.44631486e-02 -1.32230473e+00
-3.60686809e-01 -1.96928810e-02 -3.72548878e-01 3.94878268e-01
3.50662172e-02 1.99708268e-01 6.34013355e-01 1.53763071e-01
-5.86604536e-01 6.58885539e-01 1.33711481e+00 1.51485845e-01
-5.86837456e-02 1.72169432e-01 -1.00041844e-01 6.91417813e-01
5.25548160e-01 -7.79294148e-02 -5.22303998e-01 -5.16686976e-01
-2.55703449e-01 -2.05951706e-02 6.85247898e-01 -1.35129499e+00
2.27205783e-01 2.71320995e-02 2.52668321e-01 -5.77261627e-01
5.45772731e-01 -5.81698954e-01 -2.99257278e-01 6.72780216e-01
-4.09309059e-01 1.98523402e-01 1.14183493e-01 7.44947255e-01
-2.80584961e-01 -2.03345627e-01 8.75419855e-01 -4.71368432e-01
-1.20219219e+00 1.87469244e-01 -6.64614916e-01 -2.53184319e-01
1.04146314e+00 -5.83058417e-01 -3.08563054e-01 -1.60284832e-01
-1.03704178e+00 7.25575387e-02 9.33097526e-02 6.69337273e-01
7.60077298e-01 -1.27392280e+00 -7.60810494e-01 3.32263023e-01
-1.56491876e-01 -2.11340412e-01 2.78197348e-01 1.21358109e+00
-7.11683273e-01 7.77166665e-01 -7.67189860e-01 -7.27564454e-01
-1.00964868e+00 4.13681477e-01 7.10792184e-01 -4.01167095e-01
-8.74546111e-01 8.15154612e-01 -8.49806741e-02 4.06898856e-02
2.67941892e-01 -3.80186826e-01 -4.38933820e-01 1.13220304e-01
6.61360204e-01 6.80889666e-01 -1.02974363e-01 -8.26957345e-01
-2.68290430e-01 4.55204308e-01 1.74051002e-01 -3.09236556e-01
1.25814593e+00 -9.44246277e-02 2.49852743e-02 3.90700996e-01
1.03942883e+00 -7.60581970e-01 -1.77287030e+00 -1.39581159e-01
1.38696032e-02 -7.80166507e-01 -1.91674501e-01 -4.68096673e-01
-1.32234418e+00 1.08083868e+00 5.85250914e-01 2.71159649e-01
1.09359717e+00 -4.19846594e-01 7.73984373e-01 1.05134130e-01
2.88178116e-01 -9.43176031e-01 4.13304895e-01 5.92805088e-01
8.66257727e-01 -8.36338103e-01 -1.07917026e-01 -6.76215738e-02
-4.96666849e-01 1.66588056e+00 6.46063626e-01 -3.71894538e-01
4.33680415e-01 -1.63241372e-01 1.33934226e-02 -4.62943017e-02
-9.58712757e-01 -4.37426358e-01 1.24496885e-01 5.81548870e-01
4.79780555e-01 -4.51578528e-01 -6.22077823e-01 -1.94225714e-01
2.85016149e-01 3.91840726e-01 8.43099594e-01 1.07376873e+00
-2.61110693e-01 -1.09600616e+00 -3.46970350e-01 5.89058697e-02
-2.90426731e-01 1.92814916e-01 -2.43563220e-01 8.90778005e-01
3.01400959e-01 8.65886211e-01 2.61472583e-01 -1.11316331e-01
5.12519300e-01 2.07124248e-01 7.25088596e-01 -4.07881945e-01
-9.19427216e-01 -1.40547693e-01 1.79586425e-01 -1.13234198e+00
-1.28354836e+00 -9.02019978e-01 -1.06135690e+00 -1.80260837e-01
4.49326336e-02 -1.59785137e-01 4.73027200e-01 1.06676412e+00
4.29968387e-01 4.49313700e-01 4.43269789e-01 -1.49189699e+00
-5.41320086e-01 -1.10610461e+00 -3.56859893e-01 3.88156414e-01
4.77082402e-01 -8.30546558e-01 -3.72113973e-01 3.21578681e-01] | [8.249382019042969, 0.14921210706233978] |
a788cbc6-7500-4d86-a916-384efeda7bf9 | discovering-latent-concepts-learned-in-bert-1 | 2205.07237 | null | https://arxiv.org/abs/2205.07237v1 | https://arxiv.org/pdf/2205.07237v1.pdf | Discovering Latent Concepts Learned in BERT | A large number of studies that analyze deep neural network models and their ability to encode various linguistic and non-linguistic concepts provide an interpretation of the inner mechanics of these models. The scope of the analyses is limited to pre-defined concepts that reinforce the traditional linguistic knowledge and do not reflect on how novel concepts are learned by the model. We address this limitation by discovering and analyzing latent concepts learned in neural network models in an unsupervised fashion and provide interpretations from the model's perspective. In this work, we study: i) what latent concepts exist in the pre-trained BERT model, ii) how the discovered latent concepts align or diverge from classical linguistic hierarchy and iii) how the latent concepts evolve across layers. Our findings show: i) a model learns novel concepts (e.g. animal categories and demographic groups), which do not strictly adhere to any pre-defined categorization (e.g. POS, semantic tags), ii) several latent concepts are based on multiple properties which may include semantics, syntax, and morphology, iii) the lower layers in the model dominate in learning shallow lexical concepts while the higher layers learn semantic relations and iv) the discovered latent concepts highlight potential biases learned in the model. We also release a novel BERT ConceptNet dataset (BCN) consisting of 174 concept labels and 1M annotated instances. | ['Hassan Sajjad', 'Jia Xu', 'Nadir Durrani', 'Firoj Alam', 'Abdul Rafae Khan', 'Fahim Dalvi'] | 2022-05-15 | discovering-latent-concepts-learned-in-bert | https://openreview.net/forum?id=POTMtpYI1xH | https://openreview.net/pdf?id=POTMtpYI1xH | iclr-2022-4 | ['novel-concepts'] | ['reasoning'] | [ 1.59118742e-01 4.62591171e-01 -3.62610042e-01 -5.79066575e-01
5.01113236e-01 -7.54661977e-01 8.56450200e-01 7.24332511e-01
-6.94677711e-01 6.69103444e-01 3.88165534e-01 -3.04234654e-01
-4.73545343e-01 -9.60268855e-01 -8.65787506e-01 -4.58117634e-01
-5.91441572e-01 8.19473863e-01 2.31552467e-01 -4.23574954e-01
2.25113794e-01 2.40306020e-01 -1.78003085e+00 5.94720066e-01
9.11799073e-01 6.92270160e-01 2.04692662e-01 1.04488134e-01
-5.23347855e-01 6.53995812e-01 -5.20757973e-01 -2.53003985e-01
7.57485479e-02 -3.97997439e-01 -1.02202058e+00 -4.93881524e-01
3.82764071e-01 -3.58273163e-02 2.31316030e-01 1.10780954e+00
-1.20424971e-01 9.11712945e-02 9.01817441e-01 -1.03132832e+00
-1.06588900e+00 1.42364240e+00 8.99942417e-04 3.63433272e-01
-6.87007084e-02 -1.71038568e-01 1.31156456e+00 -6.30761683e-01
8.81570101e-01 1.48528910e+00 9.04466569e-01 5.95346570e-01
-1.43253374e+00 -7.05414593e-01 6.25410020e-01 1.38339877e-01
-1.39937770e+00 -2.25998774e-01 5.67826152e-01 -7.13850856e-01
1.08943474e+00 -2.12550148e-01 8.08408499e-01 1.17980289e+00
7.18983784e-02 4.05769438e-01 1.08991432e+00 -7.28582025e-01
3.64721954e-01 2.88833231e-01 4.56964701e-01 5.74859142e-01
5.50155580e-01 4.23554868e-01 -8.04513752e-01 4.04615365e-02
5.18867552e-01 -2.40732476e-01 2.61571437e-01 -3.38292450e-01
-1.13584113e+00 1.07524812e+00 6.39687002e-01 9.91213083e-01
-1.43094569e-01 1.29513085e-01 5.73075533e-01 2.17758030e-01
4.79757428e-01 6.26714230e-01 -1.05984163e+00 1.51882574e-01
-7.85346866e-01 2.37177927e-02 6.58933282e-01 8.21776628e-01
1.18617070e+00 1.01611324e-01 3.60532254e-01 9.19094741e-01
4.20540154e-01 1.48362458e-01 1.05211186e+00 -6.68788493e-01
1.22942403e-01 6.63834751e-01 -4.89305973e-01 -1.04574668e+00
-4.82531130e-01 -4.43364263e-01 -3.86237890e-01 -3.08659375e-01
1.04020767e-01 6.30472302e-02 -9.12980974e-01 2.15854430e+00
-2.19153419e-01 -1.11187793e-01 2.64843225e-01 3.09205472e-01
1.09922957e+00 4.11460042e-01 7.41349220e-01 -1.17018655e-01
1.13616264e+00 -5.95083237e-01 -4.07318205e-01 -6.37461722e-01
8.57084632e-01 -1.77769467e-01 1.11148918e+00 1.25710085e-01
-7.08184779e-01 -9.70988929e-01 -9.15253580e-01 -5.27938344e-02
-1.07831013e+00 -1.90996587e-01 1.12823129e+00 6.05193496e-01
-1.27769732e+00 6.07306063e-01 -4.96984035e-01 -9.63863313e-01
4.08327311e-01 4.21972364e-01 -2.16899768e-01 3.32928747e-01
-1.39767694e+00 8.21923673e-01 1.48659456e+00 -2.39439443e-01
-8.28342736e-01 -5.62835217e-01 -8.91614079e-01 2.09931642e-01
4.07689571e-01 -6.45376563e-01 8.03011954e-01 -1.41050851e+00
-1.20976055e+00 1.21717048e+00 4.01207469e-02 -6.73865020e-01
-2.39273071e-01 -2.71009821e-02 -5.01886427e-01 2.08298936e-01
2.26670325e-01 1.19169354e+00 2.54540026e-01 -1.55689335e+00
-8.95239532e-01 -1.02157928e-01 3.00569683e-01 1.13128461e-01
-6.32290602e-01 -4.58147563e-02 2.43386909e-01 -9.03588533e-01
3.08599532e-01 -7.49519944e-01 7.09858388e-02 -3.37079346e-01
-2.72864904e-02 -4.47465062e-01 3.52199852e-01 -2.93107897e-01
1.06297672e+00 -1.93784857e+00 6.58275113e-02 1.91052496e-01
3.91116202e-01 2.74457093e-02 -5.82789518e-02 4.59653705e-01
-2.43551433e-01 5.85500181e-01 -3.06856304e-01 -2.02835370e-02
2.23281220e-01 7.97578037e-01 -5.68931282e-01 3.98540497e-02
-5.33355922e-02 1.15204048e+00 -1.07019269e+00 -3.75089109e-01
1.07146446e-02 -1.29255932e-02 -8.33590567e-01 -2.81006247e-01
-4.92035598e-01 1.84303969e-01 -4.26138379e-03 5.70682585e-01
2.49054492e-01 -9.37960148e-02 6.63224995e-01 -1.36541680e-01
-2.08750412e-01 7.01178491e-01 -5.78014076e-01 1.63100600e+00
-2.86480755e-01 7.57501185e-01 -5.32683969e-01 -1.40036201e+00
9.70825672e-01 5.61379567e-02 1.74060374e-01 -3.99045497e-01
2.14756966e-01 3.60020131e-01 5.37962735e-01 -4.14641589e-01
4.20223147e-01 -6.75455391e-01 -1.33307651e-01 3.02189708e-01
7.21359730e-01 3.01559240e-01 4.04864788e-01 6.47238642e-02
5.38048029e-01 2.88415760e-01 5.65950096e-01 -9.25300658e-01
1.66268736e-01 8.62697512e-02 7.08576977e-01 8.06545854e-01
-1.84458233e-02 -1.70237765e-01 5.47387123e-01 -4.78057861e-01
-7.12690830e-01 -1.30451190e+00 -5.16793728e-01 1.58867490e+00
-1.32794723e-01 -5.39208174e-01 -2.92181790e-01 -6.46295249e-01
1.08529143e-01 8.71803284e-01 -1.10664225e+00 -3.82653892e-01
-3.89861345e-01 -6.60236061e-01 6.31597161e-01 6.69304371e-01
3.10890049e-01 -1.38506901e+00 -7.00719476e-01 1.74291253e-01
2.95443814e-02 -8.60920072e-01 3.27672333e-01 6.18691206e-01
-1.23603904e+00 -7.40746021e-01 1.17625609e-01 -1.15051234e+00
9.29643512e-01 -3.01775455e-01 1.38389492e+00 2.46368185e-01
9.41035822e-02 2.37025887e-01 -7.09678233e-01 -5.86522460e-01
-4.60852027e-01 2.87589163e-01 3.97350460e-01 -3.73699725e-01
1.00543404e+00 -7.35821009e-01 -6.94993287e-02 -3.93734202e-02
-1.11234558e+00 -4.53469474e-05 5.08733690e-01 7.91991711e-01
2.46885955e-01 2.90478855e-01 7.58836150e-01 -1.24347281e+00
4.81425017e-01 -7.53456175e-01 -2.78220415e-01 3.09816778e-01
-6.42047763e-01 1.06661193e-01 4.02432472e-01 -5.67573965e-01
-9.74465549e-01 -4.14556354e-01 -5.67284748e-02 2.46930793e-01
-6.67804003e-01 1.09932029e+00 3.20392214e-02 3.23816299e-01
8.16471696e-01 1.22038037e-01 -3.67809176e-01 -5.26553214e-01
7.33676732e-01 2.53471464e-01 3.25887829e-01 -1.16594172e+00
8.31467271e-01 5.36354959e-01 -3.65676016e-01 -9.52026606e-01
-1.11525142e+00 -2.82049477e-01 -1.45277739e+00 1.26494318e-01
9.29354012e-01 -6.83170617e-01 -5.34524262e-01 5.67830727e-02
-1.03107905e+00 -4.65465039e-01 -6.59475803e-01 5.99188745e-01
-3.59986812e-01 1.57033667e-01 -6.91719353e-01 -6.15638793e-01
8.40484425e-02 -5.79266310e-01 6.08967364e-01 1.71183869e-01
-7.53758967e-01 -1.57630730e+00 -6.60440400e-02 -2.56256610e-01
2.94506937e-01 1.05217658e-01 1.68806660e+00 -9.24600363e-01
-1.29311979e-01 5.13878584e-01 -7.17329001e-03 2.41951466e-01
2.68894404e-01 -1.15168445e-01 -1.03372538e+00 -3.21974725e-01
-3.84606645e-02 -5.33134162e-01 1.07642043e+00 3.18985045e-01
9.19033170e-01 -5.02650976e-01 -5.41720271e-01 6.99956059e-01
1.47623718e+00 3.26125145e-01 2.29443386e-01 4.26232606e-01
4.11355227e-01 1.25758219e+00 8.20924714e-02 -1.04180448e-01
3.20034623e-01 1.78752154e-01 1.20838834e-02 3.48410845e-01
2.30885401e-01 -5.05746603e-01 4.65719134e-01 1.30778778e+00
-2.33104592e-03 5.30028194e-02 -1.33099198e+00 7.03543723e-01
-1.75867081e+00 -6.85600877e-01 2.49704018e-01 1.81917143e+00
1.19982326e+00 4.83321279e-01 -1.12403728e-01 -2.03993171e-01
3.38350534e-01 -1.77173823e-01 -4.24206585e-01 -4.85153586e-01
-3.88306051e-01 4.95809913e-01 1.64367959e-01 4.79319960e-01
-8.99327159e-01 1.58841527e+00 6.98782539e+00 5.50629735e-01
-1.00734222e+00 1.35719210e-01 3.18524569e-01 3.19141328e-01
-5.47446668e-01 3.49261075e-01 -8.50122273e-01 9.24527273e-02
9.39221799e-01 -2.27410972e-01 1.71755865e-01 7.61259198e-01
-4.22606677e-01 -7.08194897e-02 -1.36727571e+00 4.97149736e-01
3.26906294e-01 -1.30999005e+00 6.29321337e-01 1.25732511e-01
6.14667416e-01 9.81091633e-02 3.96268107e-02 7.37815678e-01
6.85047805e-01 -1.26260984e+00 1.06206441e+00 4.38258410e-01
8.33167911e-01 -4.64782149e-01 5.70897460e-01 3.36404592e-01
-9.88782704e-01 -3.65303457e-01 -7.03979969e-01 -4.42956746e-01
-1.13003217e-01 1.79575443e-01 -7.27823675e-01 3.05630773e-01
1.00387931e+00 9.95641530e-01 -1.05176890e+00 3.82287681e-01
-6.39939308e-01 9.90497828e-01 -2.89951384e-01 -3.79876345e-02
3.30034673e-01 -5.14665470e-02 2.49759376e-01 1.17483795e+00
6.67367056e-02 -3.52131575e-02 1.29841998e-01 1.28692043e+00
1.19513258e-01 2.67123103e-01 -6.02343023e-01 -4.05141920e-01
4.76892859e-01 7.58019149e-01 -1.30839932e+00 -4.19489920e-01
-3.32853913e-01 4.93006378e-01 3.99665415e-01 4.54309523e-01
-4.34869349e-01 -4.67398539e-02 4.64564055e-01 1.76684126e-01
2.43676767e-01 -3.88843417e-01 -3.33119661e-01 -1.08817816e+00
-4.89120990e-01 -4.52160001e-01 5.42527318e-01 -6.39276087e-01
-1.71410406e+00 6.02280021e-01 6.15414321e-01 -5.11138439e-01
-2.53615350e-01 -9.48174655e-01 -2.08630279e-01 6.69092417e-01
-1.36505282e+00 -1.28220081e+00 4.35253829e-02 2.80683935e-01
2.58674890e-01 -5.95549762e-01 1.08001125e+00 8.13546330e-02
1.92828272e-02 4.05497640e-01 6.00561835e-02 2.96839386e-01
4.08212990e-01 -1.17309391e+00 3.60259712e-01 6.80139720e-01
5.71232200e-01 1.40285015e+00 4.32186127e-01 -9.15862143e-01
-2.22961351e-01 -6.62433922e-01 1.30453551e+00 -6.21232867e-01
8.36970031e-01 -7.77631879e-01 -9.41884577e-01 1.11560380e+00
1.65845454e-01 -4.17695642e-01 1.28671098e+00 7.31035292e-01
-5.52190483e-01 -6.64987564e-02 -7.51233280e-01 5.30850947e-01
1.40124428e+00 -6.15618408e-01 -1.47546446e+00 1.12525217e-01
1.10132742e+00 2.43113771e-01 -4.08856422e-01 4.38640714e-01
4.42176431e-01 -7.83383012e-01 9.15179670e-01 -1.04288161e+00
4.16409284e-01 -1.27191201e-01 -3.65154803e-01 -1.33929133e+00
-5.56546688e-01 1.80222429e-02 1.19429022e-01 1.34086096e+00
5.72271585e-01 -8.19836199e-01 4.79694575e-01 1.95790038e-01
-2.26637885e-01 -4.95798439e-01 -7.74571598e-01 -7.81889856e-01
6.25690222e-01 -4.20814633e-01 5.95356345e-01 1.50469208e+00
5.88963665e-02 6.07753336e-01 3.03769082e-01 -2.46012270e-01
2.66189933e-01 4.65715714e-02 2.13633969e-01 -1.82827652e+00
-1.78992093e-01 -6.09191298e-01 -3.67408037e-01 -7.75530934e-01
5.23229897e-01 -1.56169617e+00 -1.21015444e-01 -1.51524329e+00
1.65578261e-01 -8.21171403e-01 -5.69514275e-01 9.08749402e-01
6.90773651e-02 -1.03646189e-01 1.39637455e-01 3.24876249e-01
-4.07448053e-01 4.72961158e-01 3.85757506e-01 -4.63359896e-03
-8.78899321e-02 -7.98376799e-01 -1.10397732e+00 1.15691566e+00
8.82633388e-01 -7.04520881e-01 -6.27018929e-01 -6.48071349e-01
9.11682725e-01 -9.15950716e-01 2.92380482e-01 -8.82052898e-01
5.95199987e-02 -2.07480475e-01 3.85522276e-01 -2.34688938e-01
-5.41123189e-02 -9.42575634e-01 -2.22291946e-01 7.79430509e-01
-5.51958263e-01 -1.51561961e-01 5.19128740e-01 3.87911856e-01
-2.21767098e-01 -6.01741433e-01 5.95033288e-01 -4.89895493e-01
-1.18489933e+00 -6.88841715e-02 -5.26991844e-01 2.72797167e-01
7.14465261e-01 -4.19515222e-01 -2.92891204e-01 1.41608372e-01
-9.24684644e-01 -1.59697086e-02 3.88056666e-01 8.81789386e-01
2.64094681e-01 -1.34860992e+00 -4.55148071e-01 1.73967272e-01
4.17114824e-01 -1.07204482e-01 -5.19583113e-02 2.30495125e-01
-5.14831543e-01 5.97858846e-01 -4.80863333e-01 -6.38789952e-01
-7.91917682e-01 6.74820721e-01 2.66561359e-01 -6.91386536e-02
-4.41211104e-01 1.13434064e+00 7.43696749e-01 -8.28481257e-01
8.59691352e-02 -6.29383445e-01 -6.00165367e-01 6.93245471e-01
-1.31598726e-01 -2.48021469e-01 -4.84606713e-01 -9.63496447e-01
-3.26734692e-01 4.69668418e-01 -8.65125842e-03 4.11173329e-02
1.70856440e+00 -1.60228014e-01 -6.16044760e-01 1.17846298e+00
9.28365350e-01 -8.88207629e-02 -7.27234423e-01 -4.63883907e-01
6.67369723e-01 -6.16727211e-02 -4.39990759e-01 -9.68062639e-01
-6.01833761e-01 9.81684029e-01 4.68245029e-01 -6.53058514e-02
8.32020462e-01 3.07399243e-01 3.74050498e-01 4.03016478e-01
4.82056111e-01 -1.32914329e+00 2.18291506e-01 1.02423620e+00
7.18746960e-01 -7.37131000e-01 -1.28630936e-01 -3.13787311e-01
-3.29868674e-01 1.20219886e+00 7.07662046e-01 7.18305781e-02
8.99504364e-01 -9.42356437e-02 2.16766950e-02 -4.28781480e-01
-7.32996643e-01 -4.25163448e-01 2.78800070e-01 7.20747352e-01
6.82313621e-01 1.25336617e-01 -5.76607108e-01 9.63244915e-01
-7.96217203e-01 -1.60161108e-01 2.46715099e-01 7.53964305e-01
-5.28928101e-01 -1.09884226e+00 7.85903111e-02 5.10861397e-01
-1.81677863e-01 -5.11465967e-01 -5.41992307e-01 8.03795099e-01
1.15451443e+00 6.11474454e-01 3.01808357e-01 -2.87332833e-01
1.76444706e-02 5.28608024e-01 3.71231735e-01 -1.34155774e+00
-6.80485845e-01 -2.82790899e-01 3.87804955e-02 -1.31183555e-02
-7.93610871e-01 -6.34234369e-01 -1.46619511e+00 3.77525352e-02
-2.32590303e-01 3.20316404e-01 3.55659872e-01 1.18566632e+00
1.04638496e-02 5.84912658e-01 -3.24721754e-01 -4.36834484e-01
-2.06156254e-01 -1.15697420e+00 -5.62395394e-01 6.60048485e-01
-9.37618539e-02 -9.03244317e-01 -3.59234393e-01 4.96329725e-01] | [10.45700454711914, 9.080110549926758] |
0d63747c-dc02-4c91-8cf8-ae2117322133 | the-limits-of-word-level-differential-privacy | 2205.02130 | null | https://arxiv.org/abs/2205.02130v1 | https://arxiv.org/pdf/2205.02130v1.pdf | The Limits of Word Level Differential Privacy | As the issues of privacy and trust are receiving increasing attention within the research community, various attempts have been made to anonymize textual data. A significant subset of these approaches incorporate differentially private mechanisms to perturb word embeddings, thus replacing individual words in a sentence. While these methods represent very important contributions, have various advantages over other techniques and do show anonymization capabilities, they have several shortcomings. In this paper, we investigate these weaknesses and demonstrate significant mathematical constraints diminishing the theoretical privacy guarantee as well as major practical shortcomings with regard to the protection against deanonymization attacks, the preservation of content of the original sentences as well as the quality of the language output. Finally, we propose a new method for text anonymization based on transformer based language models fine-tuned for paraphrasing that circumvents most of the identified weaknesses and also offers a formal privacy guarantee. We evaluate the performance of our method via thorough experimentation and demonstrate superior performance over the discussed mechanisms. | ['Florian Kerschbaum', 'Benjamin Weggenmann', 'Justus Mattern'] | 2022-05-02 | null | https://aclanthology.org/2022.findings-naacl.65 | https://aclanthology.org/2022.findings-naacl.65.pdf | findings-naacl-2022-7 | ['text-anonymization'] | ['natural-language-processing'] | [ 9.73592550e-02 1.22241490e-01 -3.94698381e-02 -3.26304346e-01
-7.30004013e-01 -9.36105907e-01 6.00027561e-01 4.97484177e-01
-5.35602331e-01 8.41426671e-01 5.36099970e-01 -4.17693526e-01
-2.01811016e-01 -6.79714978e-01 -4.53236401e-01 -6.22634828e-01
1.47176757e-01 8.65612328e-02 3.51381004e-02 -3.42940718e-01
3.24335426e-01 6.96063697e-01 -1.18678653e+00 4.10163915e-03
9.56565201e-01 5.18223166e-01 -6.17440939e-01 2.91898578e-01
-1.23241775e-01 4.30470049e-01 -6.09794736e-01 -1.22459090e+00
5.57114303e-01 -8.75591710e-02 -7.30624735e-01 -2.96360165e-01
3.28378230e-01 -3.94951671e-01 -6.12196267e-01 1.16563261e+00
6.25526011e-01 -1.75836518e-01 2.67582178e-01 -1.52592146e+00
-9.33457077e-01 6.87584341e-01 -6.26579702e-01 3.96746472e-02
3.89892191e-01 -1.89561248e-01 1.08755946e+00 -6.09485209e-01
3.78723711e-01 9.97082412e-01 7.55386889e-01 6.44008756e-01
-1.26906037e+00 -7.26355970e-01 3.33039425e-02 -8.50931853e-02
-1.50765514e+00 -6.42299592e-01 5.88756263e-01 -2.06028104e-01
6.75741732e-01 6.06553733e-01 2.34147340e-01 1.13225460e+00
1.59817204e-01 8.27886343e-01 9.34750497e-01 -4.02122468e-01
2.01126441e-01 7.63400257e-01 3.22337389e-01 4.99493778e-01
7.35000908e-01 -2.74538755e-01 -4.42170322e-01 -1.00779271e+00
1.45891085e-01 -1.78876221e-02 -5.87233126e-01 -7.50049472e-01
-7.39026546e-01 8.88785243e-01 -7.32469046e-03 2.88068920e-01
1.39731411e-02 1.49053372e-02 7.50470757e-01 3.07190865e-01
6.65068865e-01 4.59640503e-01 -2.62229353e-01 4.40701693e-02
-7.79287219e-01 3.83480281e-01 9.78308022e-01 9.37336981e-01
5.37055552e-01 -2.12266460e-01 -1.68726847e-01 4.34254527e-01
1.41301572e-01 1.82787120e-01 6.23921752e-01 -5.36194563e-01
8.30796957e-01 4.95477229e-01 2.02689961e-01 -1.49207473e+00
1.22433029e-01 -2.41651461e-01 -7.37123609e-01 -8.10691193e-02
3.68039817e-01 -1.82543188e-01 -1.94449216e-01 1.94324887e+00
1.00344308e-01 -3.13729286e-01 2.50235796e-01 4.60534096e-01
6.71755970e-02 3.68666649e-01 1.83182791e-01 -2.28072014e-02
1.31966734e+00 -4.30263788e-01 -9.46229875e-01 2.83915866e-02
7.13986337e-01 -6.16340101e-01 9.86450374e-01 1.26243189e-01
-1.02431977e+00 1.30160138e-01 -1.10226047e+00 -2.30499387e-01
-4.56276715e-01 -1.87642947e-01 7.43648410e-01 1.51878417e+00
-1.05266404e+00 5.98174095e-01 -6.66841507e-01 -3.97064984e-01
4.64967638e-01 6.17635608e-01 -6.91689312e-01 3.74944389e-01
-1.41358352e+00 6.52325988e-01 1.13409907e-01 -1.28775746e-01
-1.97809502e-01 -8.08041215e-01 -8.03928554e-01 3.97422940e-01
1.28110781e-01 -5.41033745e-01 9.79846358e-01 -6.61722302e-01
-1.28385913e+00 9.82036650e-01 -8.90895873e-02 -9.21278954e-01
9.11103666e-01 -2.33949527e-01 -4.56149608e-01 -2.26304196e-02
-3.60404253e-02 1.52062863e-01 5.60120285e-01 -1.06167841e+00
-5.29529631e-01 -6.36101127e-01 1.94354087e-01 -1.31300800e-02
-1.23487079e+00 2.26170465e-01 -5.23663275e-02 -1.02118194e+00
-1.66808486e-01 -8.83229733e-01 -3.49987298e-01 1.33492187e-01
-5.13858020e-01 9.87242460e-02 9.38449442e-01 -6.06048465e-01
1.52721465e+00 -2.39679599e+00 -1.35072023e-01 2.02563658e-01
2.86498427e-01 5.01224101e-01 2.22665086e-01 8.84381235e-01
-4.24381308e-02 6.92009509e-01 -4.09023732e-01 -6.99524224e-01
3.26199025e-01 1.71335116e-02 -7.89838195e-01 8.64456415e-01
-5.95458820e-02 7.09691107e-01 -5.82992792e-01 -2.30022356e-01
6.57806396e-02 4.50556755e-01 -6.19287848e-01 -4.22209837e-02
1.43715903e-01 -1.94969594e-01 -5.16154826e-01 2.08354622e-01
9.86312866e-01 2.33180061e-01 4.05666590e-01 8.20334256e-02
-2.02832464e-02 2.28630140e-01 -1.21328819e+00 1.23507166e+00
-2.17117041e-01 4.28646058e-01 2.54388392e-01 -7.69184470e-01
7.70247042e-01 4.70348626e-01 3.90334755e-01 -1.44933999e-01
1.70233041e-01 4.52269688e-02 -3.25258493e-01 -3.48428756e-01
9.85727251e-01 -1.79366305e-01 -3.56500685e-01 8.41752172e-01
-2.58924216e-01 2.50610918e-01 -2.47491151e-01 4.66387004e-01
9.50824142e-01 -3.35514724e-01 3.47450912e-01 -3.99653226e-01
6.98385000e-01 -4.06088501e-01 5.86257160e-01 6.55359447e-01
-4.45248932e-01 5.53715348e-01 7.35327721e-01 -1.34776518e-01
-1.15125155e+00 -7.73865819e-01 5.73854111e-02 7.51329899e-01
1.52288899e-01 -6.31263554e-01 -8.90590250e-01 -7.89232433e-01
4.08834040e-01 8.21458280e-01 -5.95424235e-01 -3.78085762e-01
-4.13133681e-01 -8.01192462e-01 1.16282439e+00 3.83580416e-01
3.44940215e-01 -4.21802282e-01 -3.00981611e-01 -9.62263420e-02
-1.49048075e-01 -1.13206017e+00 -6.91094100e-01 -1.16295844e-01
-6.79004490e-01 -8.38170111e-01 -4.13987398e-01 -4.66403604e-01
7.12056398e-01 5.38223684e-01 4.16637689e-01 -6.53786063e-02
-1.01067349e-01 4.78318453e-01 -1.94493547e-01 -4.58302796e-01
-5.71582377e-01 2.90616810e-01 3.36096942e-01 5.05576074e-01
5.09221435e-01 -6.71872377e-01 -3.35352004e-01 6.76954538e-02
-1.24773395e+00 -4.21611756e-01 1.35026693e-01 5.93659222e-01
1.74658358e-01 8.34213793e-02 6.09302104e-01 -1.30962527e+00
1.11738896e+00 -5.60710251e-01 -5.09581327e-01 2.78815329e-01
-8.89433682e-01 1.34142727e-01 1.00173771e+00 -1.10232681e-01
-1.07557619e+00 -9.44984406e-02 -2.02638566e-01 -1.84226707e-01
5.48022538e-02 6.60875738e-02 -5.62593102e-01 -3.48018676e-01
4.96743321e-01 2.78387964e-01 2.32365921e-01 -6.12373531e-01
5.34349084e-01 9.45097208e-01 4.00467306e-01 -6.09845877e-01
1.07785320e+00 8.30339074e-01 -2.44207978e-01 -6.54297471e-01
-4.09234792e-01 -2.57282138e-01 -3.11235845e-01 5.37751257e-01
3.91918093e-01 -8.18664074e-01 -7.95072794e-01 3.69540483e-01
-1.00233376e+00 5.95446587e-01 -3.79954517e-01 1.80929050e-01
-3.45328897e-01 1.09499061e+00 -5.93036175e-01 -9.40498650e-01
-5.42494297e-01 -9.60505903e-01 5.02563417e-01 -3.90837602e-02
-4.85782713e-01 -9.00998652e-01 -8.84945393e-02 3.35956782e-01
4.21837568e-01 3.33495319e-01 1.06973815e+00 -1.00823259e+00
-2.90122420e-01 -7.38875389e-01 -2.75960825e-02 5.21179378e-01
3.47659469e-01 -2.18840800e-02 -9.49849188e-01 -5.64706981e-01
2.58377522e-01 1.47495624e-02 3.64954650e-01 -1.96103007e-01
1.12057912e+00 -6.92840099e-01 -2.45035186e-01 7.04186022e-01
1.44290745e+00 -1.03005938e-01 6.94513738e-01 4.95572716e-01
4.98876065e-01 7.09729791e-01 1.25083134e-01 7.05458641e-01
1.73540339e-01 4.61394668e-01 2.46856570e-01 5.48726842e-02
4.29613888e-01 -4.93501008e-01 3.00547987e-01 5.57823837e-01
4.76155251e-01 -5.42112648e-01 -4.20320004e-01 7.65554488e-01
-1.75808585e+00 -9.99313354e-01 -8.01141486e-02 2.31938910e+00
6.98805451e-01 -9.44629461e-02 1.64950162e-01 3.71421665e-01
7.42290914e-01 3.77382427e-01 -3.47632200e-01 -8.28345060e-01
-4.52989042e-01 1.54608905e-01 1.02830267e+00 4.04183239e-01
-9.27672565e-01 7.96690285e-01 6.90405655e+00 4.60806072e-01
-7.44203746e-01 1.16666155e-02 4.10009921e-01 -6.69659451e-02
-8.17660511e-01 2.45565057e-01 -7.05222547e-01 5.72226703e-01
8.23434591e-01 -8.98290455e-01 3.06963623e-01 6.91713572e-01
9.55730602e-02 3.86672169e-01 -1.06727600e+00 7.28076637e-01
1.41397804e-01 -1.18883669e+00 2.95294791e-01 1.83648080e-01
4.35255706e-01 -4.13717240e-01 3.61802101e-01 -2.68985212e-01
3.53174955e-01 -8.21123064e-01 6.69180870e-01 1.67513058e-01
5.13520360e-01 -1.11096954e+00 6.00606501e-01 2.06159532e-01
-5.66229343e-01 -1.46179274e-01 -4.68560636e-01 -9.70865563e-02
6.47071423e-03 5.22219837e-01 -3.66371870e-01 8.39977562e-01
5.36812842e-01 2.66166866e-01 -5.54660439e-01 6.61205888e-01
-2.05568656e-01 5.38765609e-01 -2.19129294e-01 -6.39813095e-02
1.06735177e-01 -3.08319092e-01 6.18044734e-01 1.21345758e+00
1.96960911e-01 -1.14404336e-01 -3.53438556e-01 7.58117676e-01
-3.77068460e-01 3.80244017e-01 -7.53208995e-01 -2.51629233e-01
6.93091929e-01 8.80337656e-01 -2.33795628e-01 4.68834713e-02
-4.57446605e-01 1.13919914e+00 3.09568435e-01 2.12869823e-01
-6.34710908e-01 -7.23805368e-01 1.28066397e+00 2.52849668e-01
1.59884945e-01 -1.95620641e-01 -4.24596876e-01 -1.34050226e+00
5.76161683e-01 -1.11221182e+00 4.15250838e-01 -2.77636256e-02
-1.30738938e+00 3.70754272e-01 -1.59207329e-01 -1.07055998e+00
2.59441994e-02 -2.44824424e-01 -2.47886539e-01 9.78575706e-01
-1.39033198e+00 -8.60411942e-01 3.09273869e-01 5.28703928e-01
-9.93041173e-02 -2.23916724e-01 1.04666376e+00 4.59201783e-01
-6.51011348e-01 1.31390119e+00 6.92076325e-01 1.19628586e-01
6.72049701e-01 -9.75320935e-01 8.04861426e-01 1.10494077e+00
1.67762283e-02 1.21221268e+00 8.34977567e-01 -4.58298296e-01
-1.55149293e+00 -9.84562993e-01 1.37096608e+00 -4.91206080e-01
7.61372268e-01 -8.42637062e-01 -1.05046999e+00 9.32686329e-01
1.53287023e-01 -2.89108247e-01 9.79020119e-01 -9.73804742e-02
-5.38795412e-01 -1.81823924e-01 -1.58244050e+00 8.56070399e-01
8.50969315e-01 -8.11806500e-01 -5.79241931e-01 2.48770759e-01
7.92391002e-01 -1.28272623e-01 -6.47458792e-01 -1.20158158e-01
5.40429115e-01 -1.17781115e+00 8.07578623e-01 -7.11370230e-01
4.97670211e-02 -2.74316490e-01 -1.53330967e-01 -9.25932705e-01
-2.06552461e-01 -1.05845737e+00 4.57431227e-02 1.86119127e+00
2.00011969e-01 -1.15316093e+00 1.04302168e+00 1.03337097e+00
5.02091348e-01 -3.94469261e-01 -1.04690289e+00 -8.01334560e-01
2.87629157e-01 -1.76142126e-01 1.05968142e+00 1.07415354e+00
1.86715856e-01 -7.50067532e-02 -7.51136243e-01 3.04937810e-01
6.33251250e-01 -1.44914076e-01 8.03672850e-01 -1.00618482e+00
3.09387129e-03 -1.10037334e-01 -5.99907815e-01 -6.88781261e-01
4.40140814e-01 -9.60833609e-01 -5.69604158e-01 -1.01745820e+00
2.04839930e-01 -2.32884049e-01 -3.68139476e-01 3.08568984e-01
-1.53882876e-02 4.10495363e-02 4.00226526e-02 1.09574206e-01
-5.62816001e-02 6.41521037e-01 3.94526541e-01 -7.77631253e-02
5.54065369e-02 1.09962888e-01 -1.49287188e+00 4.63272572e-01
8.52944732e-01 -6.52514219e-01 -5.41908264e-01 -6.51675999e-01
2.18211100e-01 -2.86514193e-01 2.39901394e-01 -7.32924461e-01
2.18401074e-01 1.37084335e-01 -1.79728866e-01 -1.04701854e-01
6.51528910e-02 -1.16180897e+00 1.71607003e-01 3.49744469e-01
-6.20139897e-01 3.99358571e-01 1.01327367e-01 8.70802879e-01
-1.56158090e-01 -2.68435776e-01 7.53611743e-01 2.61453483e-02
-1.72123238e-01 2.78975427e-01 -4.34165865e-01 5.02302386e-02
1.22034323e+00 -3.28641772e-01 -1.88276112e-01 -3.98358703e-01
-2.26711214e-01 2.05182403e-01 8.88123035e-01 5.84722817e-01
3.51256520e-01 -1.06669307e+00 -6.33140624e-01 1.57804385e-01
1.27872467e-01 -7.48628497e-01 5.49301952e-02 4.36818004e-01
-4.28892344e-01 4.95163441e-01 -7.57728592e-02 1.85256228e-01
-1.47993624e+00 8.38390768e-01 1.26309291e-01 -1.12691738e-01
-8.07597578e-01 5.28202355e-01 1.38226405e-01 -3.40720356e-01
4.42622215e-01 -2.95132026e-02 1.74247742e-01 -4.61648293e-02
5.02975523e-01 3.84916276e-01 1.69706881e-01 -6.27472818e-01
-4.89784926e-01 1.54683158e-01 -3.40229422e-01 -1.10511128e-02
1.24327171e+00 -5.31459272e-01 -3.56609136e-01 -6.48208186e-02
1.42382967e+00 6.21789157e-01 -6.98770165e-01 -3.98925006e-01
2.65324354e-01 -8.40516508e-01 -3.04432064e-01 -2.60804057e-01
-8.79468918e-01 7.43189573e-01 3.57466698e-01 3.46356034e-01
9.88968551e-01 -5.70190668e-01 1.07760489e+00 2.48090014e-01
3.00837219e-01 -9.68602896e-01 -5.97896695e-01 1.69540644e-01
3.92235607e-01 -7.41430640e-01 2.88578272e-01 -6.06940508e-01
-6.34007692e-01 7.65296936e-01 2.45154306e-01 -1.11803457e-01
3.72105300e-01 4.52874988e-01 9.30330604e-02 2.68943995e-01
-4.72499996e-01 5.07884562e-01 -4.23793703e-01 6.39066517e-01
3.34449381e-01 -4.10670489e-02 -9.27224159e-01 6.95367813e-01
-3.27713907e-01 -1.74555242e-01 8.67303133e-01 9.68164384e-01
3.26476097e-02 -1.53261673e+00 -2.61981398e-01 1.81860551e-01
-1.06987226e+00 -2.18173251e-01 -6.05511189e-01 6.91733897e-01
-2.90163398e-01 8.46062243e-01 -2.33004764e-01 -3.56506675e-01
5.04432738e-01 1.28992423e-01 6.07462227e-02 -2.65503466e-01
-9.89102304e-01 -6.62480652e-01 8.96134749e-02 -2.83820033e-01
-1.40815079e-02 -7.58383930e-01 -8.13731492e-01 -8.96931052e-01
-3.24990451e-01 5.70303917e-01 4.96411443e-01 6.45224750e-01
7.73311079e-01 -4.68874983e-02 9.77792859e-01 -3.35875191e-02
-1.13978434e+00 -3.13650072e-01 -8.58758569e-01 6.12184525e-01
3.23542476e-01 -6.93440959e-02 -6.13918722e-01 -1.68916166e-01] | [6.122953414916992, 6.969371795654297] |
b2e52979-1705-4d11-b729-75d85224258d | mic-masked-image-consistency-for-context | 2212.01322 | null | https://arxiv.org/abs/2212.01322v2 | https://arxiv.org/pdf/2212.01322v2.pdf | MIC: Masked Image Consistency for Context-Enhanced Domain Adaptation | In unsupervised domain adaptation (UDA), a model trained on source data (e.g. synthetic) is adapted to target data (e.g. real-world) without access to target annotation. Most previous UDA methods struggle with classes that have a similar visual appearance on the target domain as no ground truth is available to learn the slight appearance differences. To address this problem, we propose a Masked Image Consistency (MIC) module to enhance UDA by learning spatial context relations of the target domain as additional clues for robust visual recognition. MIC enforces the consistency between predictions of masked target images, where random patches are withheld, and pseudo-labels that are generated based on the complete image by an exponential moving average teacher. To minimize the consistency loss, the network has to learn to infer the predictions of the masked regions from their context. Due to its simple and universal concept, MIC can be integrated into various UDA methods across different visual recognition tasks such as image classification, semantic segmentation, and object detection. MIC significantly improves the state-of-the-art performance across the different recognition tasks for synthetic-to-real, day-to-nighttime, and clear-to-adverse-weather UDA. For instance, MIC achieves an unprecedented UDA performance of 75.9 mIoU and 92.8% on GTA-to-Cityscapes and VisDA-2017, respectively, which corresponds to an improvement of +2.1 and +3.0 percent points over the previous state of the art. The implementation is available at https://github.com/lhoyer/MIC. | ['Luc van Gool', 'Haoran Wang', 'Dengxin Dai', 'Lukas Hoyer'] | 2022-12-02 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Hoyer_MIC_Masked_Image_Consistency_for_Context-Enhanced_Domain_Adaptation_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Hoyer_MIC_Masked_Image_Consistency_for_Context-Enhanced_Domain_Adaptation_CVPR_2023_paper.pdf | cvpr-2023-1 | ['synthetic-to-real-translation'] | ['computer-vision'] | [ 2.18637213e-01 7.00483993e-02 -1.06805615e-01 -6.10642970e-01
-1.04043436e+00 -6.27869248e-01 6.65299594e-01 -1.62068591e-01
-2.60181189e-01 6.78619146e-01 -3.56593311e-01 -1.35000408e-01
3.73745143e-01 -7.05726266e-01 -9.29499626e-01 -7.19292998e-01
2.22436488e-01 4.62020516e-01 4.53105092e-01 -2.88047940e-02
-2.19104126e-01 1.97204113e-01 -1.38794184e+00 4.83063698e-01
1.00682974e+00 1.16033840e+00 4.61240113e-01 5.53610682e-01
2.24696938e-02 4.92971569e-01 -5.62945068e-01 -2.27572709e-01
4.32604790e-01 -4.15698498e-01 -6.43052995e-01 5.07115543e-01
6.71415925e-01 -1.36496469e-01 -1.03164569e-01 1.07931197e+00
3.09258431e-01 4.94494252e-02 8.10342371e-01 -1.17793894e+00
-8.63795102e-01 5.03879339e-02 -8.97576034e-01 9.96534377e-02
-8.17603841e-02 3.52021009e-01 6.32752359e-01 -1.10188127e+00
7.90713310e-01 9.95810747e-01 5.97288370e-01 6.09184444e-01
-1.40846467e+00 -7.47785747e-01 5.04211307e-01 2.82602966e-01
-1.60542130e+00 -3.67248356e-01 5.12362897e-01 -5.81842422e-01
6.91635847e-01 2.78474241e-01 3.12693954e-01 1.27019882e+00
-1.72582030e-01 7.56885588e-01 1.38732684e+00 -4.31514233e-01
3.67994547e-01 3.05589825e-01 -1.17278084e-01 3.57603550e-01
1.16698205e-01 4.08584774e-02 -3.66128087e-01 2.17605561e-01
4.06173885e-01 -1.79583758e-01 -1.26237929e-01 -4.37266707e-01
-1.06254959e+00 5.52093804e-01 6.39911175e-01 -7.36240819e-02
-3.08463305e-01 -3.19136232e-01 1.09640270e-01 1.09688975e-01
6.87904060e-01 1.61079749e-01 -5.24335265e-01 4.27745581e-01
-9.08790290e-01 1.05619743e-01 4.00686890e-01 9.48168576e-01
9.00304317e-01 2.42829934e-01 -1.62119135e-01 1.14035535e+00
2.44063601e-01 9.37120736e-01 3.85104626e-01 -7.15247810e-01
4.02971894e-01 4.29948241e-01 2.48799056e-01 -7.67837465e-01
-2.30190113e-01 -5.63811541e-01 -9.28168416e-01 4.07070816e-01
7.98256040e-01 1.60790645e-02 -1.64379776e+00 1.85300601e+00
6.30273581e-01 3.72884780e-01 2.67741233e-01 1.03227293e+00
6.43637955e-01 8.93399596e-01 6.87986836e-02 6.61185980e-02
1.21276104e+00 -1.03190708e+00 -2.22190693e-01 -7.85510123e-01
2.02382788e-01 -8.23308289e-01 1.20959485e+00 1.79722205e-01
-6.28626049e-01 -8.09373438e-01 -9.83740628e-01 2.02806354e-01
-5.93161404e-01 3.12009454e-01 1.64294958e-01 4.39869851e-01
-9.50249016e-01 1.05198480e-01 -6.06887817e-01 -7.14298368e-01
5.77273190e-01 -2.05467567e-02 -4.13772523e-01 -3.71275663e-01
-1.02558184e+00 7.16090560e-01 4.61123466e-01 6.75638020e-02
-1.15025628e+00 -7.02654958e-01 -8.66651952e-01 -3.75731677e-01
4.01148647e-01 -5.08662462e-01 1.21941757e+00 -1.46675789e+00
-1.11391640e+00 1.32771885e+00 -7.32806250e-02 -6.29954994e-01
6.72513545e-01 -2.45461352e-02 -6.08286083e-01 -1.12403043e-01
5.06467879e-01 1.00076747e+00 1.08358026e+00 -1.50967407e+00
-8.37196648e-01 -2.79622585e-01 -2.38219991e-01 1.67362094e-01
5.83244450e-02 -2.49631330e-01 -7.66163170e-01 -7.86813378e-01
9.02046040e-02 -1.07716167e+00 -9.01715755e-02 1.62198097e-01
-5.39144337e-01 1.19304061e-01 9.19118702e-01 -8.59817207e-01
6.63909018e-01 -2.36535192e+00 -1.29287884e-01 1.99948832e-01
-5.76969236e-02 3.09872955e-01 -3.54857892e-01 -7.31858686e-02
-1.12995535e-01 -2.59026766e-01 -6.90144420e-01 -1.91741168e-01
-1.23126388e-01 2.88399428e-01 -4.49915737e-01 4.76501524e-01
3.18190426e-01 6.94437325e-01 -7.44396448e-01 -3.08191538e-01
2.79377162e-01 3.18737209e-01 -1.10803008e-01 4.39009875e-01
-4.84557092e-01 7.42533088e-01 -1.68439344e-01 7.26768672e-01
1.01407647e+00 -4.74433750e-01 2.87520677e-01 -6.44911779e-03
4.06782441e-02 -1.11103663e-02 -1.23396516e+00 1.61174595e+00
-4.35659885e-01 7.00656414e-01 8.22186768e-02 -1.04260480e+00
9.52846766e-01 -3.22074965e-02 1.39267802e-01 -1.14292407e+00
-2.51152366e-01 3.68810594e-01 -8.38540271e-02 -3.03254038e-01
2.56075859e-01 1.72761425e-01 -1.15347393e-01 1.96148306e-01
-1.97975840e-02 3.91136436e-03 4.05433849e-02 9.08409804e-02
6.87389076e-01 2.45954737e-01 2.59696007e-01 -1.87937856e-01
3.12582165e-01 3.86954337e-01 8.64956379e-01 8.16749454e-01
-2.77242899e-01 1.10728347e+00 1.20595723e-01 -4.88290131e-01
-1.19550753e+00 -1.30813742e+00 -2.35186562e-01 9.91520703e-01
2.46461391e-01 2.57264346e-01 -8.89541507e-01 -7.98751652e-01
-2.99118347e-02 6.92849100e-01 -9.02391613e-01 -4.29896228e-02
-2.99412876e-01 -9.55707610e-01 3.18001062e-01 4.85359192e-01
9.14049804e-01 -9.96522129e-01 -2.29844972e-01 6.70991689e-02
-3.19150805e-01 -1.42940283e+00 -6.16013050e-01 7.36806244e-02
-3.63804489e-01 -9.63521183e-01 -8.83621395e-01 -7.95287728e-01
7.90108383e-01 3.54546905e-01 1.15608954e+00 -3.62866193e-01
-3.67033780e-01 1.71594590e-01 -1.83036640e-01 -3.09032261e-01
-4.05080497e-01 -8.35913941e-02 -1.48791717e-02 5.24475992e-01
2.14409605e-01 -3.94645095e-01 -7.74303734e-01 6.17356181e-01
-8.23265910e-01 3.50340515e-01 5.97027540e-01 8.68006945e-01
1.02990401e+00 -2.38784373e-01 5.27321577e-01 -9.83963370e-01
-1.24421202e-01 -6.44979835e-01 -7.18577027e-01 4.75375712e-01
-6.29295707e-01 -1.85334876e-01 5.07148802e-01 -6.80128455e-01
-1.26444077e+00 4.24053967e-01 1.74530998e-01 -5.76565683e-01
-5.30669034e-01 1.58241376e-01 -3.36066842e-01 1.36662766e-01
9.33845043e-01 3.83497298e-01 -3.82044345e-01 -3.85245800e-01
4.60983932e-01 6.86964869e-01 8.48275423e-01 -5.45967937e-01
9.61965978e-01 6.90386474e-01 -4.44236487e-01 -6.62597537e-01
-1.02604055e+00 -5.37321210e-01 -6.14378214e-01 -1.64116457e-01
9.85089600e-01 -1.28032255e+00 1.44529015e-01 7.54296184e-01
-9.84950781e-01 -8.91224504e-01 -2.73750752e-01 2.98007816e-01
-3.76857817e-01 6.19671308e-02 -1.53067261e-01 -5.21783769e-01
-2.04700187e-01 -8.89310002e-01 9.64091837e-01 4.22405303e-01
-1.31794393e-01 -7.24321187e-01 -8.83469507e-02 3.78692389e-01
3.38683635e-01 3.45869631e-01 6.43545151e-01 -5.09094238e-01
-6.13019168e-01 -2.39465088e-02 -5.50494432e-01 5.47725320e-01
3.40465903e-01 2.53908895e-03 -1.38578856e+00 -2.02201843e-01
-3.00110221e-01 -2.84588367e-01 8.23558688e-01 5.45057297e-01
1.13776422e+00 -2.59300649e-01 -3.19556981e-01 6.60786271e-01
1.32234097e+00 3.34966362e-01 5.78278899e-01 4.20678556e-01
6.81216717e-01 5.49223185e-01 9.37877059e-01 3.33212256e-01
5.08782744e-01 8.73316288e-01 4.43258524e-01 -4.43343371e-01
-7.04006910e-01 -3.31432015e-01 3.40498179e-01 3.59684259e-01
3.22172582e-01 -2.31009260e-01 -1.11489904e+00 1.00681460e+00
-1.86314261e+00 -7.33324885e-01 -9.93028805e-02 2.37305379e+00
8.23913753e-01 1.10865973e-01 1.74753875e-01 -2.80196398e-01
9.80908692e-01 9.35428888e-02 -9.36945975e-01 -1.07034266e-01
-3.99516463e-01 -5.23616597e-02 5.55012167e-01 4.34491664e-01
-1.45361066e+00 1.01856756e+00 5.11620998e+00 1.12780142e+00
-1.41299438e+00 3.14121336e-01 1.20882320e+00 1.05750211e-01
-2.46977638e-02 -2.01242015e-01 -5.55274487e-01 5.49325764e-01
7.31192648e-01 1.93076581e-01 3.60283047e-01 1.00772095e+00
1.43300369e-01 -1.99953079e-01 -8.09083045e-01 9.00183976e-01
6.51755231e-03 -1.29963946e+00 -1.38767108e-01 -1.83155268e-01
1.26223266e+00 2.98425496e-01 3.74210954e-01 2.96400875e-01
3.84647459e-01 -9.34681654e-01 8.95034611e-01 3.03686678e-01
1.16078603e+00 -4.09123063e-01 5.49123824e-01 2.29179665e-01
-1.20376098e+00 1.88938618e-01 -3.62089306e-01 2.15575799e-01
-1.78732589e-01 4.86179292e-01 -1.00925195e+00 4.78674293e-01
1.08143437e+00 6.90784752e-01 -6.72761559e-01 9.16989625e-01
-3.19980264e-01 7.40694821e-01 -3.38035822e-01 4.72303182e-01
3.24605495e-01 -1.13732107e-01 4.15259153e-01 1.28692472e+00
2.34665975e-01 -1.22449897e-01 1.88698485e-01 9.46391463e-01
-1.37369633e-01 -7.03480560e-04 -5.19864261e-01 4.25705403e-01
4.06990796e-01 1.14544415e+00 -9.09243643e-01 -4.26854879e-01
-4.95857805e-01 1.28801668e+00 2.53535628e-01 6.72047138e-01
-1.07997954e+00 -2.05735937e-01 7.33667374e-01 2.93460429e-01
5.09972811e-01 6.78166151e-02 -4.32165861e-01 -1.05083811e+00
2.86150366e-01 -9.23580945e-01 4.88989234e-01 -9.51588273e-01
-1.45707738e+00 7.84388900e-01 5.25663979e-03 -1.48899221e+00
-1.91416830e-01 -5.31950772e-01 -6.98705077e-01 9.30264592e-01
-1.57589388e+00 -1.34944177e+00 -5.14037371e-01 6.30614102e-01
7.09102452e-01 -1.02670461e-01 6.92579806e-01 2.95833915e-01
-6.65266216e-01 5.99064529e-01 3.66618484e-01 2.87861735e-01
9.89684701e-01 -1.17175436e+00 6.31349862e-01 1.25278938e+00
2.77048349e-01 -5.27117848e-02 4.52182084e-01 -4.24235523e-01
-7.46387720e-01 -1.72102118e+00 6.72380209e-01 -4.65969235e-01
5.09086311e-01 -5.19080698e-01 -1.21927178e+00 5.65192342e-01
1.02669485e-01 5.44690251e-01 1.88593179e-01 -1.83649227e-01
-8.30174267e-01 -4.79396880e-01 -1.19319892e+00 5.86220801e-01
7.74257779e-01 -5.33583879e-01 -1.56876236e-01 4.44840938e-01
5.51628470e-01 -6.13221109e-01 -5.02555370e-01 4.56955284e-01
2.79573202e-01 -8.68218660e-01 9.80446398e-01 -3.69240046e-01
1.64546713e-01 -7.67117679e-01 -4.14326161e-01 -1.29944217e+00
-1.48245245e-01 -2.84608692e-01 2.47373372e-01 1.28828609e+00
7.62427926e-01 -6.22176111e-01 6.35586679e-01 5.75601041e-01
-1.79740652e-01 -4.65825379e-01 -1.03436744e+00 -1.04520249e+00
8.29562321e-02 -5.17421901e-01 5.19463956e-01 1.23434997e+00
-7.01120615e-01 2.00335130e-01 -4.09254700e-01 6.69364035e-01
6.49931848e-01 4.02523190e-01 8.49843562e-01 -8.49945903e-01
-2.51903206e-01 -2.17040420e-01 -2.25604236e-01 -8.32186222e-01
6.39126985e-04 -9.15903032e-01 2.55163431e-01 -1.58367884e+00
2.50019252e-01 -6.81440532e-01 -3.85749727e-01 7.44301438e-01
-2.10461676e-01 6.50749981e-01 1.11583389e-01 3.33582789e-01
-6.60911798e-01 3.95145178e-01 1.02402401e+00 -4.43958700e-01
-1.58644691e-01 3.45336460e-02 -4.20640498e-01 6.48337126e-01
1.13420391e+00 -6.03583276e-01 -3.11279744e-01 -5.36291480e-01
-3.89431149e-01 -2.20709831e-01 5.99081337e-01 -1.09175909e+00
-6.87220916e-02 -3.61108720e-01 6.06892824e-01 -3.26345682e-01
2.33740658e-01 -6.30695403e-01 2.52073497e-01 1.92627624e-01
-7.76470527e-02 -2.53227890e-01 5.12687802e-01 6.53113663e-01
-3.04598987e-01 1.89228132e-01 1.22049010e+00 1.29488632e-01
-1.26724720e+00 3.01131934e-01 -1.86444819e-01 3.87044013e-01
1.14695466e+00 -2.74164945e-01 -4.64097738e-01 -3.34377140e-01
-6.72100484e-01 1.09816566e-01 5.15324771e-01 5.96931934e-01
5.19932032e-01 -1.19651330e+00 -9.85610783e-01 2.77436763e-01
6.11528814e-01 2.77871788e-01 6.22877359e-01 6.97275043e-01
-3.14098090e-01 1.23416372e-01 -2.01545835e-01 -9.48625505e-01
-1.17312074e+00 4.47567642e-01 5.70321798e-01 -4.86639962e-02
-6.10268772e-01 7.78960705e-01 7.88313687e-01 -5.50132036e-01
1.29123986e-01 -2.72159010e-01 3.28228176e-01 -2.36871302e-01
4.61860597e-01 -1.10380322e-01 1.31601065e-01 -7.62902915e-01
-4.25859988e-01 5.96832871e-01 -5.90750501e-02 3.69224511e-02
1.01188076e+00 -2.70472497e-01 3.07194591e-01 3.52335513e-01
9.37615991e-01 -1.02734394e-01 -1.87184489e+00 -5.44183850e-01
-1.01611853e-01 -4.27980125e-01 -2.46589839e-01 -1.24754810e+00
-1.26077676e+00 8.46568465e-01 1.01292419e+00 -1.05097756e-01
1.28064752e+00 2.31363043e-01 4.13061768e-01 5.49921067e-03
1.66319415e-01 -1.06283343e+00 8.35691318e-02 2.87442088e-01
9.05057609e-01 -1.69654620e+00 -2.75350422e-01 -3.04402560e-01
-1.01413906e+00 6.42273068e-01 9.00790751e-01 9.26487222e-02
4.54649508e-01 4.27562892e-02 5.48668683e-01 1.53748319e-01
-5.01361251e-01 -4.49398696e-01 5.59458792e-01 1.03082728e+00
-5.69068678e-02 2.85170585e-01 3.42476606e-01 4.22571570e-01
2.27573901e-01 -4.28175598e-01 2.69739389e-01 3.87982011e-01
-2.53215939e-01 -8.18441749e-01 -5.44024527e-01 1.46764025e-01
-2.53424317e-01 -1.45080209e-01 -3.17851663e-01 9.63769019e-01
3.39568257e-01 9.53397274e-01 2.02469930e-01 -2.60119915e-01
3.61783087e-01 4.05625552e-02 1.57103926e-01 -5.59317708e-01
-9.47927237e-02 1.48878977e-01 -8.26129392e-02 -5.24755657e-01
-3.78086776e-01 -7.08815217e-01 -1.15027714e+00 -5.34161702e-02
8.67214799e-02 -1.93191141e-01 6.28344715e-01 7.00708807e-01
4.81085032e-01 3.65391821e-01 6.06891870e-01 -7.22278357e-01
-5.79134747e-02 -9.23313379e-01 -5.19310892e-01 5.47084928e-01
3.39394063e-01 -5.14604151e-01 -1.38282299e-01 3.53806287e-01] | [9.769550323486328, 1.3514660596847534] |
7a9876b9-a848-4aba-b9c4-5c35cbfb468a | nighttime-smartphone-reflective-flare-removal | 2303.15046 | null | https://arxiv.org/abs/2303.15046v1 | https://arxiv.org/pdf/2303.15046v1.pdf | Nighttime Smartphone Reflective Flare Removal Using Optical Center Symmetry Prior | Reflective flare is a phenomenon that occurs when light reflects inside lenses, causing bright spots or a "ghosting effect" in photos, which can impact their quality. Eliminating reflective flare is highly desirable but challenging. Many existing methods rely on manually designed features to detect these bright spots, but they often fail to identify reflective flares created by various types of light and may even mistakenly remove the light sources in scenarios with multiple light sources. To address these challenges, we propose an optical center symmetry prior, which suggests that the reflective flare and light source are always symmetrical around the lens's optical center. This prior helps to locate the reflective flare's proposal region more accurately and can be applied to most smartphone cameras. Building on this prior, we create the first reflective flare removal dataset called BracketFlare, which contains diverse and realistic reflective flare patterns. We use continuous bracketing to capture the reflective flare pattern in the underexposed image and combine it with a normally exposed image to synthesize a pair of flare-corrupted and flare-free images. With the dataset, neural networks can be trained to remove the reflective flares effectively. Extensive experiments demonstrate the effectiveness of our method on both synthetic and real-world datasets. | ['Chen Change Loy', 'Chongyi Li', 'Shangchen Zhou', 'Yihang Luo', 'Yuekun Dai'] | 2023-03-27 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Dai_Nighttime_Smartphone_Reflective_Flare_Removal_Using_Optical_Center_Symmetry_Prior_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Dai_Nighttime_Smartphone_Reflective_Flare_Removal_Using_Optical_Center_Symmetry_Prior_CVPR_2023_paper.pdf | cvpr-2023-1 | ['flare-removal'] | ['computer-vision'] | [ 8.26032460e-01 -6.33490503e-01 4.02892709e-01 -2.33175918e-01
-2.79058099e-01 -8.24680090e-01 3.46278101e-01 -7.40272462e-01
3.15325797e-01 7.17003286e-01 2.15142369e-01 1.61211461e-01
7.34753683e-02 -5.57554007e-01 -8.47626567e-01 -8.09190452e-01
4.45205927e-01 -4.83387411e-01 2.87618816e-01 -1.82109207e-01
4.53567803e-01 4.87669319e-01 -1.94891989e+00 2.01197088e-01
1.28078866e+00 9.00215626e-01 5.11495113e-01 6.37010932e-01
2.14875430e-01 8.65097463e-01 -9.24130619e-01 5.02861338e-03
7.05965519e-01 -3.58025104e-01 1.26641288e-01 3.60698104e-01
1.31146765e+00 -6.74781680e-01 -1.64150640e-01 1.30461597e+00
3.14078838e-01 1.54682562e-01 3.00663888e-01 -1.05866027e+00
-7.97100306e-01 -1.52691260e-01 -1.01939404e+00 2.30788589e-01
7.42593765e-01 4.62997913e-01 4.11261529e-01 -8.83792043e-01
1.67201385e-01 1.01251268e+00 9.17329669e-01 1.91646755e-01
-8.54954302e-01 -8.09880316e-01 -1.77347392e-01 -3.54296528e-02
-1.08213019e+00 -7.33388722e-01 9.77097034e-01 -4.55607593e-01
6.63938761e-01 3.50182652e-01 7.51843572e-01 6.03902221e-01
5.58400869e-01 3.99625719e-01 1.43713546e+00 -4.02512252e-01
-2.15384051e-01 -1.74097732e-01 1.39477938e-01 4.70447868e-01
6.10734582e-01 4.93618518e-01 -2.91098684e-01 -4.41546477e-02
5.28070509e-01 3.92432839e-01 -1.23354530e+00 1.69740543e-02
-7.33019471e-01 4.68378514e-01 5.49994051e-01 -6.35403246e-02
-1.53281540e-01 -1.25368893e-01 -5.47173083e-01 -1.76377930e-02
2.98680395e-01 5.00213385e-01 -1.82276025e-01 3.96434903e-01
-7.74325788e-01 2.56104052e-01 4.15880769e-01 5.92708886e-01
1.03930330e+00 1.47868037e-01 2.97705848e-02 9.42706704e-01
3.58667135e-01 1.23370945e+00 1.28536597e-01 -9.30841923e-01
7.09722610e-03 6.38416052e-01 5.50872564e-01 -1.14371777e+00
-3.20475399e-01 -1.23947389e-01 -7.74252117e-01 4.92919981e-01
1.54727697e-01 -2.77905971e-01 -1.25478637e+00 1.22554731e+00
1.33476883e-01 6.80468917e-01 1.40158264e-02 1.51704645e+00
8.51853132e-01 7.35071957e-01 -8.54607284e-01 -6.35131478e-01
9.51277256e-01 -7.89302766e-01 -1.11637950e+00 -7.88679779e-01
-4.05606553e-02 -1.23338032e+00 1.04531527e+00 5.43333888e-01
-1.15569019e+00 -4.05655801e-01 -1.06799972e+00 4.02823128e-02
1.16304569e-01 5.62466346e-02 5.70506454e-01 3.58678043e-01
-9.04415905e-01 1.36181056e-01 -2.68018693e-01 4.86099496e-02
1.34318843e-01 -1.95823044e-01 8.90520141e-02 -4.71937150e-01
-8.77770901e-01 7.08748519e-01 -2.28545383e-01 5.05188942e-01
-5.81475556e-01 -8.81439865e-01 -7.94503033e-01 -1.97808713e-01
4.49734479e-01 -6.11059964e-01 9.41086054e-01 -1.33710420e+00
-1.24383783e+00 6.68074191e-01 -5.45767009e-01 -1.83969364e-01
-4.31059487e-02 -5.13852298e-01 -7.73802280e-01 1.86164498e-01
4.60942276e-02 -7.47363567e-02 1.36015809e+00 -1.80342650e+00
-7.83961654e-01 -2.06725284e-01 3.20442766e-02 2.82651454e-01
3.05802166e-01 4.71316017e-02 -2.60930300e-01 -3.59850675e-01
3.50003913e-02 -9.64348018e-01 1.46150276e-01 -2.86770999e-01
-4.71050531e-01 4.19411451e-01 1.08699155e+00 -3.28193516e-01
1.17869723e+00 -2.08172560e+00 -5.49895167e-01 -1.91109210e-01
1.82824105e-01 7.19979107e-01 -2.95072328e-02 1.41525030e-01
-1.19900256e-01 -3.86984527e-01 -4.47706699e-01 2.20321506e-01
-4.64471757e-01 1.26448184e-01 -7.25209713e-01 5.70079923e-01
2.61694044e-01 5.32692313e-01 -8.50070477e-01 1.70032769e-01
3.15568715e-01 5.76739430e-01 -1.75375238e-01 2.79099256e-01
-2.14142188e-01 3.42402637e-01 -1.20615222e-01 6.64213955e-01
1.37112606e+00 1.09627031e-01 -3.57262313e-01 -3.47578466e-01
-4.55048263e-01 -1.51166454e-01 -1.22300601e+00 1.01438665e+00
-3.28698337e-01 7.78269291e-01 9.33519006e-02 1.70695484e-02
1.14168847e+00 -2.95054495e-01 1.64771408e-01 -7.73045838e-01
1.24807671e-01 1.98493004e-01 -2.46986330e-01 -7.64603853e-01
4.85530198e-01 -1.91878587e-01 4.45708066e-01 2.84260780e-01
-6.73601627e-01 -2.14830577e-01 1.11633176e-02 -2.04908833e-01
1.09668517e+00 1.06727876e-01 8.26865062e-02 3.21897194e-02
5.80022275e-01 -2.26937681e-01 9.37561512e-01 7.29207456e-01
-5.27420826e-02 1.14598060e+00 -2.95864135e-01 -5.64995229e-01
-3.90212715e-01 -1.21347570e+00 -2.14151636e-01 5.80878496e-01
5.49666286e-01 -5.26792966e-02 -6.73103690e-01 -3.23846459e-01
1.03701828e-02 7.66738594e-01 -1.57081410e-01 -2.51474619e-01
-5.80234528e-01 -7.33098745e-01 -1.96968600e-01 1.55990925e-02
7.89965630e-01 -1.05794525e+00 -9.72634673e-01 -2.42990360e-01
-2.53449738e-01 -9.50333536e-01 -5.04995584e-01 -3.35536093e-01
-1.75245777e-01 -1.63175988e+00 -3.18275660e-01 -4.13808376e-01
6.46391511e-01 1.40341842e+00 1.26421666e+00 2.89629191e-01
-7.22674549e-01 2.54947782e-01 -3.35476011e-01 -7.73334980e-01
-1.32035464e-01 -8.96001041e-01 -1.00828886e-01 4.17211771e-01
3.73754919e-01 -3.68831277e-01 -1.12750530e+00 3.11203480e-01
-1.08680582e+00 -1.29466094e-02 4.59713727e-01 5.12717426e-01
4.93163109e-01 4.38069344e-01 2.53243476e-01 -8.14500034e-01
5.14893770e-01 -1.74635544e-01 -9.72724020e-01 -3.83478552e-02
-3.02552521e-01 -5.49368560e-01 6.36811078e-01 -3.95339847e-01
-1.48847055e+00 1.20202273e-01 4.36965227e-01 -8.26920450e-01
-2.39489108e-01 -7.45452428e-03 -2.15820938e-01 -3.67720783e-01
8.37399006e-01 1.94379203e-02 -2.47911766e-01 -2.87701100e-01
1.07719317e-01 7.84369171e-01 8.14491391e-01 1.72799647e-01
1.01620281e+00 1.02621198e+00 6.52771518e-02 -9.77731287e-01
-1.66995835e+00 -5.27783036e-01 -8.31840113e-02 -4.65722442e-01
5.25308728e-01 -9.48542535e-01 -6.62718773e-01 7.88840532e-01
-9.96278524e-01 -2.40793049e-01 -8.99586303e-04 4.01019424e-01
-2.10639581e-01 3.53057265e-01 -8.71138871e-02 -9.95845973e-01
-2.80260742e-01 -9.41453934e-01 1.11152923e+00 9.39841628e-01
4.10405099e-01 -6.00748539e-01 3.84393096e-01 3.82898152e-01
2.86496758e-01 2.27243185e-01 3.55858296e-01 5.51497400e-01
-8.66712749e-01 -8.55043009e-02 -4.02412474e-01 4.73999918e-01
6.29515886e-01 3.61892879e-01 -1.29745829e+00 -2.91725248e-01
3.59441787e-01 1.63913250e-01 1.06751835e+00 6.77914441e-01
5.07782817e-01 -9.95212570e-02 -3.24105352e-01 1.01594138e+00
1.75471413e+00 4.04993415e-01 9.55255449e-01 3.65705758e-01
8.56719375e-01 5.30477822e-01 8.90460849e-01 4.34112787e-01
-1.76411439e-02 4.64453995e-01 6.79761946e-01 -4.91274267e-01
-4.56939369e-01 1.72077522e-01 4.37397957e-01 2.45210797e-01
2.21571445e-01 -5.23700893e-01 -5.58386445e-01 4.10662293e-01
-1.63600481e+00 -1.20132673e+00 -7.32806027e-01 2.31596231e+00
7.56585062e-01 -3.67656916e-01 -4.31388348e-01 -1.72778398e-01
8.49555612e-01 2.30895042e-01 -5.91242075e-01 -8.59606117e-02
-5.98994315e-01 2.08254922e-02 6.01993442e-01 8.44589889e-01
-9.54088509e-01 8.06115091e-01 6.41130781e+00 9.16521996e-02
-1.47352755e+00 -2.98384845e-01 -1.72640920e-01 -3.68470490e-01
-5.00979662e-01 1.21699817e-01 -9.51200485e-01 7.70765245e-01
4.64767694e-01 1.59993637e-02 7.82435179e-01 4.04189825e-02
7.24338233e-01 -7.42212117e-01 -5.31174004e-01 1.18103635e+00
7.09801972e-01 -8.77175093e-01 -9.48411450e-02 -2.33281165e-01
9.38109875e-01 1.66500062e-02 -2.35022213e-02 -3.92502517e-01
2.70585269e-01 -9.16652858e-01 2.09909990e-01 1.31798959e+00
5.50498486e-01 -7.06509531e-01 3.90615493e-01 1.51352048e-01
-8.29321921e-01 -1.39139652e-01 -5.42080462e-01 -1.96737394e-01
3.33383739e-01 1.08474362e+00 -6.59005880e-01 3.36523950e-01
9.73729432e-01 9.58855808e-01 -6.41494930e-01 1.44709051e+00
-5.54306984e-01 3.43903005e-01 -4.18121099e-01 5.97448289e-01
-1.92388907e-01 -8.60682964e-01 8.50749373e-01 8.13483059e-01
5.06503105e-01 -7.67947547e-03 1.23059206e-01 8.66222620e-01
1.05852455e-01 -4.44687188e-01 -9.63025510e-01 3.94541204e-01
3.90856713e-01 1.44885051e+00 -1.54154629e-01 1.56945646e-01
-6.14697814e-01 8.88719082e-01 -3.42247128e-01 6.45488262e-01
-8.34307313e-01 -4.67875600e-01 9.47211921e-01 4.45302963e-01
1.16580658e-01 1.71661139e-01 1.93808436e-01 -1.10489726e+00
2.83882529e-01 -6.84829295e-01 2.10203193e-02 -1.77392268e+00
-1.25021183e+00 4.41078037e-01 -2.53218710e-01 -1.39527023e+00
1.19215429e-01 -4.43850845e-01 -7.85579503e-01 9.52562869e-01
-2.18095613e+00 -1.13088977e+00 -9.28367913e-01 7.90147007e-01
2.20925495e-01 1.90748245e-01 4.45028245e-01 2.64667332e-01
-5.25580585e-01 -9.78517905e-02 3.12202305e-01 -3.11202288e-01
1.22815967e+00 -1.07472217e+00 -2.08728269e-01 1.40828705e+00
-3.35607082e-02 4.69570607e-01 9.92722332e-01 -7.75955975e-01
-1.56261802e+00 -1.32531738e+00 3.63889605e-01 -4.92037922e-01
4.30377752e-01 -3.09261624e-02 -1.19230080e+00 6.98443174e-01
3.51336271e-01 1.29887626e-01 2.51176596e-01 -3.89801681e-01
-1.08128153e-01 -4.87772077e-01 -8.71792316e-01 6.91013396e-01
9.77221131e-01 -1.21164843e-01 -7.42570937e-01 6.27696276e-01
4.35499579e-01 -4.16594207e-01 -2.68128514e-01 7.82700121e-01
2.16398284e-01 -1.65866411e+00 9.66273069e-01 2.78906345e-01
2.02149048e-01 -8.04509342e-01 4.70791832e-02 -1.48498726e+00
-1.92706510e-01 -1.06699646e+00 -1.71747983e-01 1.38423705e+00
-1.69688314e-01 -7.97913074e-01 3.01794201e-01 6.36373311e-02
-4.89420742e-01 -1.42010435e-01 -1.37388483e-01 -3.64989132e-01
-5.13538539e-01 -1.28165826e-01 5.43846190e-01 8.86336863e-01
-7.79181778e-01 4.34785068e-01 -5.36583841e-01 6.59773350e-01
7.87066460e-01 7.28789270e-01 1.11551976e+00 -1.48786843e+00
1.84191801e-02 -4.85919975e-02 9.69409272e-02 -8.00350130e-01
1.13205396e-01 -2.96804458e-01 6.24635339e-01 -1.61908102e+00
7.98045993e-02 -1.38633892e-01 2.88333781e-02 2.43549913e-01
-5.67275941e-01 5.46575069e-01 5.95213547e-02 4.70499367e-01
-2.99344957e-01 2.89481401e-01 1.39693308e+00 1.80061862e-01
-5.18240809e-01 -7.48189539e-02 -7.67276168e-01 1.40627933e+00
5.89619398e-01 -1.48628533e-01 -5.38857996e-01 -6.11385703e-01
6.29664958e-01 -1.75601840e-01 5.19710898e-01 -1.14828658e+00
9.10711959e-02 -4.44139868e-01 5.84221423e-01 -7.16264427e-01
3.29194218e-01 -8.52164567e-01 3.79295915e-01 -1.28872588e-01
3.50237399e-01 -4.02976602e-01 1.19380876e-01 6.35327458e-01
-1.39728442e-01 -4.14315701e-01 1.01675284e+00 -2.14547589e-01
-5.51806211e-01 1.25368059e-01 -2.44801193e-01 -4.42734733e-02
9.92016792e-01 -4.45189357e-01 -1.01964700e+00 -3.14063698e-01
5.86760119e-02 4.88005169e-02 1.04605389e+00 3.50710750e-01
8.14620197e-01 -9.45851088e-01 -5.76829910e-01 6.58431768e-01
1.16100498e-01 3.05628944e-02 4.33656871e-01 8.18010151e-01
-7.29555011e-01 9.55121666e-02 3.85483764e-02 -5.09171367e-01
-1.41281748e+00 5.34218490e-01 7.18776226e-01 5.45848131e-01
-9.26393867e-01 4.40267146e-01 7.23985493e-01 3.09034318e-01
-1.38604164e-01 -1.09296188e-01 -4.04638112e-01 -1.44403279e-01
1.02235031e+00 2.67980009e-01 7.12546706e-02 -6.95903957e-01
-2.32781675e-02 1.11659682e+00 3.50932106e-02 5.57008445e-01
1.07856739e+00 -6.24005079e-01 -2.65124291e-01 5.10004997e-01
7.91837335e-01 8.33294988e-01 -1.44803810e+00 -1.74215600e-01
-7.80993462e-01 -1.11879158e+00 4.01550710e-01 -8.28963339e-01
-1.29470563e+00 6.03562653e-01 6.29150808e-01 3.75453353e-01
1.70120108e+00 -4.15274113e-01 9.71996546e-01 -1.13803428e-02
1.33012325e-01 -7.40364611e-01 7.28468644e-04 3.60815585e-01
8.57667327e-01 -1.05863714e+00 -2.55760085e-02 -6.92712128e-01
-4.92561340e-01 1.02878428e+00 5.58967829e-01 -3.48032534e-01
3.51865411e-01 4.80519891e-01 4.41433221e-01 -3.35320085e-01
-4.84841764e-01 -5.34338474e-01 2.97340423e-01 7.85501122e-01
-7.46666715e-02 -4.43476737e-01 3.99459638e-02 2.22743049e-01
-1.95195273e-01 1.16876952e-01 1.06905222e+00 9.52679634e-01
-8.59858215e-01 -4.09137189e-01 -9.29063082e-01 5.07871866e-01
-4.40413535e-01 -1.41038641e-01 -5.85752726e-01 3.27922255e-01
3.36150676e-01 1.29544103e+00 2.88427830e-01 -8.87923595e-03
4.17787910e-01 -2.69389868e-01 5.46291053e-01 -4.68364656e-01
-3.50321591e-01 5.67449391e-01 -2.19217449e-01 -7.48281419e-01
-6.36867821e-01 -4.52057511e-01 -9.79666948e-01 -7.92359784e-02
-5.21962762e-01 -2.53025442e-01 2.41625115e-01 5.69646060e-01
4.51930940e-01 3.87684464e-01 8.58916461e-01 -8.58036101e-01
2.02597320e-01 -7.36847281e-01 -9.14739549e-01 7.40914583e-01
8.31632078e-01 -8.99687171e-01 -1.16602600e+00 2.52135605e-01] | [10.697420120239258, -3.0583784580230713] |
f0680a0f-88f3-4c23-bb90-67c5fdd6266e | reasoning-through-memorization-nearest | 2201.05575 | null | https://arxiv.org/abs/2201.05575v3 | https://arxiv.org/pdf/2201.05575v3.pdf | Reasoning Through Memorization: Nearest Neighbor Knowledge Graph Embeddings | Previous knowledge graph embedding approaches usually map entities to representations and utilize score functions to predict the target entities, yet they typically struggle to reason rare or emerging unseen entities. In this paper, we propose kNN-KGE, a new knowledge graph embedding approach with pre-trained language models, by linearly interpolating its entity distribution with k-nearest neighbors. We compute the nearest neighbors based on the distance in the entity embedding space from the knowledge store. Our approach can allow rare or emerging entities to be memorized explicitly rather than implicitly in model parameters. Experimental results demonstrate that our approach can improve inductive and transductive link prediction results and yield better performance for low-resource settings with only a few triples, which might be easier to reason via explicit memory. Code is available at https://github.com/zjunlp/KNN-KG. | ['Huajun Chen', 'Xu Cheng', 'Yongheng Wang', 'Xiang Chen', 'Xin Xie', 'Ningyu Zhang'] | 2022-01-14 | null | null | null | null | ['knowledge-graph-embeddings', 'knowledge-graph-embeddings'] | ['graphs', 'methodology'] | [-5.07642210e-01 5.93935788e-01 -8.63918960e-01 -2.45536506e-01
-4.13971394e-01 -5.81212342e-01 7.40153611e-01 6.42464399e-01
-4.57421839e-01 9.83049810e-01 4.79588002e-01 -3.17367375e-01
-2.89114654e-01 -1.32568574e+00 -8.72160316e-01 -1.51113942e-01
-1.78518206e-01 7.72407055e-01 1.96295142e-01 -1.86973155e-01
-1.82696551e-01 1.26962796e-01 -1.12637854e+00 1.27490059e-01
8.74832869e-01 5.30020595e-01 -2.74787813e-01 4.11493272e-01
-3.37794721e-01 1.00256073e+00 2.15903595e-02 -9.57827508e-01
-7.46558607e-02 8.59031174e-03 -9.38371956e-01 -7.92587698e-01
3.47245574e-01 -1.05653241e-01 -7.48901784e-01 8.16046298e-01
3.49922597e-01 3.48116994e-01 6.85416281e-01 -1.45386183e+00
-1.71518183e+00 1.09384894e+00 -6.49458915e-02 2.17640474e-01
4.54908520e-01 -4.18301880e-01 1.45373487e+00 -1.45327771e+00
9.31285203e-01 1.03909647e+00 7.51992702e-01 5.03434777e-01
-1.53629673e+00 -6.61213398e-01 1.14577867e-01 5.70380509e-01
-1.73005509e+00 -2.47091696e-01 5.14105618e-01 -3.30875069e-01
1.36501634e+00 5.06488569e-02 6.98581815e-01 8.66795123e-01
-2.98127294e-01 6.50319636e-01 3.56851459e-01 -4.49447632e-01
7.05192238e-03 4.48330432e-01 3.33812952e-01 9.51588213e-01
7.49975741e-01 -1.48938060e-01 -6.77335560e-01 -4.15481210e-01
5.05938888e-01 1.23287745e-01 -3.02134484e-01 -7.53212810e-01
-1.18488503e+00 1.16208673e+00 1.00808036e+00 4.48839039e-01
-3.79023880e-01 4.76121157e-01 1.44909024e-01 1.31985709e-01
5.92703164e-01 5.70538223e-01 -5.46430051e-01 1.06837757e-01
-4.63528603e-01 1.35485858e-01 9.57947671e-01 1.04753518e+00
1.18094027e+00 -2.57715553e-01 -1.97108269e-01 7.88000166e-01
3.08351964e-01 3.47933143e-01 3.40672404e-01 -4.80151981e-01
4.48305935e-01 8.64008725e-01 2.03870624e-01 -1.24303412e+00
-1.07048012e-01 1.90890860e-04 -4.03463185e-01 -5.61811984e-01
6.21521883e-02 -1.28269106e-01 -8.50120246e-01 1.64844453e+00
4.34774309e-01 6.75819337e-01 1.83529705e-01 6.93761110e-01
9.82973278e-01 7.70065725e-01 5.89587510e-01 1.95512950e-01
1.17579985e+00 -9.98415411e-01 -6.35480940e-01 -2.29491800e-01
1.23031676e+00 -3.60512406e-01 9.62036669e-01 -4.38925833e-01
-8.02441061e-01 -5.99083826e-02 -5.87785840e-01 -3.73099983e-01
-1.08398306e+00 1.08023249e-01 1.11449909e+00 2.84083724e-01
-1.27984357e+00 4.68193442e-01 -8.15191805e-01 -5.06445229e-01
3.86981845e-01 3.27482432e-01 -5.73345661e-01 -1.76947325e-01
-1.80567360e+00 1.03056955e+00 9.66732800e-01 -2.13184506e-01
-1.00624867e-01 -1.18807220e+00 -1.26160944e+00 3.21269155e-01
3.53754699e-01 -9.27686930e-01 6.47736192e-01 -2.65345067e-01
-1.01255774e+00 7.01109529e-01 -2.66941428e-01 -4.87238765e-01
-2.32583076e-01 -3.06049645e-01 -7.08572626e-01 -1.34041950e-01
-8.85572936e-03 7.07025170e-01 2.43750408e-01 -1.10167289e+00
-2.57572472e-01 -8.33978951e-02 2.57635355e-01 1.52306870e-01
-6.96633279e-01 -4.47291791e-01 -4.83445883e-01 -4.99060929e-01
-1.16554476e-01 -8.49891961e-01 -5.20207062e-02 3.59760225e-02
-4.00022417e-01 -6.50129080e-01 5.58665931e-01 -5.30783653e-01
1.55675149e+00 -1.89597797e+00 1.93820849e-01 3.73259932e-01
3.76800418e-01 4.52803880e-01 -8.42071921e-02 7.19176531e-01
-9.29401740e-02 3.32272708e-01 2.35206127e-01 2.43622158e-02
3.38242829e-01 3.45970005e-01 -5.61580718e-01 -5.50761633e-03
3.15461367e-01 1.48730743e+00 -1.28314614e+00 -5.74893653e-01
2.17830092e-02 7.51362324e-01 -6.46474540e-01 -6.73550963e-02
-2.22485989e-01 -3.28901947e-01 -5.83030581e-01 6.40310407e-01
2.46504128e-01 -7.11462557e-01 4.96849597e-01 -4.10360068e-01
4.59925801e-01 4.31058556e-01 -9.26288545e-01 1.51954103e+00
-3.68973732e-01 4.62707609e-01 -6.55622482e-01 -9.05070305e-01
8.11318099e-01 2.27667198e-01 1.98913395e-01 -4.95824635e-01
-2.05961406e-01 1.36477515e-01 -4.69441324e-01 -2.57915139e-01
8.91752422e-01 1.07491367e-01 -6.15835153e-02 2.51804143e-01
3.76405030e-01 3.99369448e-01 2.07569256e-01 7.20177591e-01
1.13693023e+00 -8.41572061e-02 3.89252484e-01 1.91870511e-01
1.14802323e-01 1.42951444e-01 5.09129107e-01 6.43102288e-01
1.33998454e-01 -1.14331357e-01 2.94237226e-01 -3.75871927e-01
-8.88565779e-01 -1.35880470e+00 -4.19251435e-02 1.19365430e+00
2.69416064e-01 -8.41727495e-01 -1.03449814e-01 -8.99348676e-01
6.15251541e-01 1.13767791e+00 -7.03078806e-01 -5.26926458e-01
-3.18297654e-01 -4.58257616e-01 5.86054981e-01 8.01761985e-01
-1.74490005e-01 -8.59327912e-01 4.23063695e-01 2.13536814e-01
-1.20431386e-01 -8.40754569e-01 -2.74091363e-01 7.88410567e-03
-7.74276674e-01 -8.36951971e-01 -4.41304743e-01 -8.55071306e-01
9.33379054e-01 -2.97349133e-02 1.31206155e+00 2.01492980e-01
-8.88798013e-02 6.80951536e-01 -4.06529337e-01 -1.13788038e-01
3.85932215e-02 3.42197865e-01 1.51409134e-01 -2.55095512e-01
9.83825684e-01 -5.98656118e-01 -6.32197261e-01 -1.14693306e-01
-7.55605102e-01 -1.45361379e-01 3.23359966e-01 9.21216190e-01
7.03266621e-01 -2.85652131e-01 6.41445100e-01 -1.18165886e+00
6.55095398e-01 -1.05941951e+00 -3.35608006e-01 7.83911824e-01
-1.15300047e+00 3.31237078e-01 3.60739827e-01 -5.32067358e-01
-6.56310856e-01 -2.64167756e-01 4.35056627e-01 -6.02794290e-01
3.06012124e-01 9.79775667e-01 2.47986421e-01 -1.28080145e-01
6.62850797e-01 -9.72714834e-03 -5.24069667e-01 -3.24182510e-01
1.13089168e+00 2.16024563e-01 1.88389063e-01 -7.11148083e-01
1.04231799e+00 1.90403000e-01 -3.34195614e-01 -2.42590457e-01
-9.40997481e-01 -5.44160426e-01 -3.28657389e-01 1.60701707e-01
6.44772887e-01 -1.20268524e+00 -5.14231801e-01 -4.13494051e-01
-9.13024724e-01 -2.68324196e-01 -4.50406760e-01 8.29427838e-01
-6.52421489e-02 9.51067358e-02 -8.00966620e-01 -4.03409868e-01
-4.15434748e-01 -1.82865605e-01 7.28820860e-01 3.35940510e-01
-3.42113644e-01 -1.57100379e+00 3.29165995e-01 1.35565579e-01
4.56041396e-01 -7.10705891e-02 1.18421614e+00 -9.05478776e-01
-7.19282568e-01 -3.57705861e-01 -4.44857538e-01 -2.47199342e-01
1.11134402e-01 -5.43815270e-02 -5.41014314e-01 -1.67482391e-01
-1.13271391e+00 -4.62994784e-01 1.06407416e+00 -1.23030506e-01
9.91031826e-01 -5.88074684e-01 -8.85677338e-01 4.41139787e-01
1.59317064e+00 -4.05302733e-01 6.24704599e-01 2.37960100e-01
1.09029090e+00 1.20674416e-01 5.38611710e-01 3.76862019e-01
1.01897395e+00 5.86992979e-01 -6.13546856e-02 2.08575591e-01
-1.97528467e-01 -1.00034773e+00 2.70912498e-01 1.18572485e+00
-1.49192542e-01 -3.80498707e-01 -1.06242895e+00 1.02723074e+00
-2.00981259e+00 -1.13065791e+00 1.31221384e-01 1.99212432e+00
1.27598941e+00 -1.88982219e-01 -2.88626283e-01 -4.12540644e-01
7.13669538e-01 -2.28433423e-02 -4.85247970e-01 -2.80620426e-01
-8.53913128e-02 3.54048312e-01 6.82467997e-01 7.82935202e-01
-8.60119700e-01 1.34297276e+00 5.56621695e+00 7.57062197e-01
-8.69279981e-01 2.43681461e-01 8.24244469e-02 -9.01280195e-02
-9.07909453e-01 3.65874976e-01 -9.97209191e-01 4.20456469e-01
1.01066351e+00 -7.50742555e-01 4.70468104e-01 8.37733686e-01
-5.58036089e-01 2.94936001e-01 -1.19354057e+00 8.81778896e-01
-1.88589878e-02 -1.65297818e+00 4.29397583e-01 -1.65920526e-01
7.83307135e-01 9.28972438e-02 -6.19614534e-02 8.44320416e-01
8.01867664e-01 -9.35842216e-01 1.74940079e-01 9.40173268e-01
5.97060025e-01 -5.67218363e-01 4.81931835e-01 1.03979670e-02
-1.35994756e+00 6.46947026e-02 -5.66499233e-01 1.82248771e-01
1.18961640e-01 6.88443720e-01 -1.30861425e+00 6.51392281e-01
5.97203910e-01 9.07591879e-01 -7.33149767e-01 7.37838924e-01
-5.53255439e-01 7.31079698e-01 -3.72742057e-01 -3.46872121e-01
-4.78478633e-02 -9.70432088e-02 2.51745552e-01 1.15250981e+00
5.71770549e-01 2.10896954e-01 5.74571937e-02 9.85945344e-01
-6.64498568e-01 2.87976503e-01 -8.60395849e-01 -5.65340817e-01
1.00243771e+00 1.17031515e+00 -4.06586915e-01 -4.33265388e-01
-5.30637324e-01 8.61052155e-01 1.12994146e+00 6.34878576e-01
-9.40423250e-01 -5.72219849e-01 6.49574399e-01 9.53870714e-02
6.52020276e-01 -7.93387964e-02 2.20624000e-01 -1.52397668e+00
1.07662408e-02 -8.12220797e-02 7.44393468e-01 -8.94305944e-01
-1.62614083e+00 2.75734961e-01 -6.16519116e-02 -8.17882061e-01
-2.89959550e-01 -3.86036515e-01 -3.75627011e-01 7.54864693e-01
-1.60136509e+00 -1.31873345e+00 2.66735591e-02 5.54333031e-01
-3.71215820e-01 3.41082551e-02 1.19432318e+00 5.62011957e-01
-4.05645937e-01 9.15484428e-01 2.29815304e-01 4.01200384e-01
7.66049027e-01 -1.31027269e+00 2.88683295e-01 3.23273063e-01
6.27674401e-01 9.54633534e-01 4.39073652e-01 -8.49442840e-01
-1.26501966e+00 -1.34033871e+00 1.65187156e+00 -8.51076305e-01
1.05556071e+00 -2.94161737e-01 -1.20812917e+00 1.24402833e+00
8.62273201e-02 6.08582735e-01 1.19609284e+00 8.71996701e-01
-8.95227373e-01 7.23955855e-02 -9.44667816e-01 6.97963893e-01
1.11800623e+00 -9.59772587e-01 -7.78419971e-01 3.26318949e-01
8.66600871e-01 -2.23608926e-01 -1.37490487e+00 2.60101825e-01
2.68276751e-01 -1.25335872e-01 1.28173101e+00 -1.10912812e+00
2.44568110e-01 -4.71572906e-01 -1.09479629e-01 -1.39207482e+00
-6.25824571e-01 -6.21574633e-02 -9.88806486e-01 1.30199647e+00
9.80216265e-01 -8.41919124e-01 8.24422896e-01 9.87273097e-01
3.97061974e-01 -9.67784703e-01 -6.73791587e-01 -7.39599288e-01
-7.89410546e-02 5.63145392e-02 7.60082304e-01 1.63457811e+00
4.12952542e-01 3.92790794e-01 -1.75334930e-01 5.00850737e-01
4.98973310e-01 3.54582161e-01 4.24630314e-01 -1.35865247e+00
-5.38516566e-02 -4.60990258e-02 -8.14618945e-01 -5.59202731e-01
5.48244476e-01 -1.63319337e+00 -5.21606863e-01 -1.84512854e+00
2.37390459e-01 -7.91650832e-01 -7.92133152e-01 9.59528804e-01
-5.01216531e-01 1.82074621e-01 -9.48458239e-02 4.36467025e-03
-9.77710485e-01 8.04352939e-01 7.10062504e-01 -2.16082126e-01
-1.53423175e-01 -6.74835145e-01 -7.50219703e-01 4.15277272e-01
8.62646580e-01 -7.07042038e-01 -5.62208772e-01 -6.02812827e-01
7.29254305e-01 -3.00156087e-01 5.46190500e-01 -5.19920588e-01
5.47322631e-01 4.26851474e-02 5.19167721e-01 -2.50037789e-01
3.68190140e-01 -6.89427912e-01 3.52799267e-01 2.25892011e-02
-5.67171693e-01 -1.16326690e-01 1.91862449e-01 8.83188128e-01
-2.07797542e-01 -8.74017254e-02 7.76658356e-02 2.83782870e-01
-1.04309261e+00 5.27968168e-01 1.84364095e-01 2.77210027e-01
9.86867130e-01 7.10632131e-02 -6.40188336e-01 -2.81634510e-01
-1.00554204e+00 4.06895578e-01 3.76844734e-01 6.49874330e-01
7.13716626e-01 -2.04568958e+00 -3.81132901e-01 -1.83684513e-01
5.68601668e-01 -3.08141917e-01 1.56738445e-01 5.96039534e-01
-2.41287008e-01 5.26851952e-01 1.90262958e-01 -1.15729170e-02
-8.78325462e-01 7.44695544e-01 8.57775956e-02 -3.49874735e-01
-3.88332784e-01 9.86594081e-01 -2.36872748e-01 -8.27327073e-01
-2.53617186e-02 -1.07301973e-01 -2.22216085e-01 1.53624728e-01
2.97748893e-01 3.01537067e-01 -2.15157598e-01 -4.97457266e-01
-4.69683677e-01 3.00688475e-01 -3.28232735e-01 3.45366985e-01
1.39791918e+00 -6.37910841e-03 -1.82685792e-01 4.44052994e-01
1.28300691e+00 1.02486432e-01 -5.98870099e-01 -8.44488502e-01
2.40400627e-01 -4.79976058e-01 1.41345322e-01 -7.40083516e-01
-9.14900064e-01 3.97261083e-01 2.74346620e-01 9.62788314e-02
5.38047910e-01 4.44758773e-01 8.81258190e-01 8.81644309e-01
4.42329377e-01 -1.00295842e+00 -2.81439722e-01 3.99702311e-01
5.65693617e-01 -1.14346528e+00 1.40294239e-01 -4.55396205e-01
-5.26609778e-01 8.18310976e-01 6.85693324e-01 5.25621399e-02
1.01413047e+00 -2.23972738e-01 -1.68355718e-01 -4.73450840e-01
-9.44179535e-01 -3.27956975e-01 4.78576213e-01 5.32841802e-01
5.17116249e-01 3.29695284e-01 -1.43181235e-01 7.69796848e-01
-2.76950374e-02 1.40663087e-01 1.38734147e-01 8.04703355e-01
-2.15032890e-01 -1.23266137e+00 2.85934150e-01 7.14629531e-01
-9.16711316e-02 -4.90259498e-01 -4.15685564e-01 6.52901888e-01
1.56180654e-03 4.81972665e-01 4.47995774e-02 -4.55726147e-01
2.92546093e-01 5.14834046e-01 2.37541199e-01 -6.71535492e-01
-1.78709209e-01 -8.72300684e-01 2.53042877e-01 -6.10671818e-01
-2.68696308e-01 -3.66600871e-01 -1.49083459e+00 -6.98240399e-01
-5.39671779e-01 4.25517589e-01 2.08182216e-01 4.22079206e-01
1.04412174e+00 1.58849090e-01 2.00693965e-01 -2.79321343e-01
-2.20777720e-01 -5.86189985e-01 -5.87112010e-01 4.73434985e-01
-1.04265027e-01 -8.42318654e-01 -2.23962724e-01 -2.81316065e-03] | [8.880461692810059, 8.008665084838867] |
96b1fd5b-9e51-48f7-8bc4-e31a28be2e04 | a-new-outlier-removal-strategy-based-on | 2205.07404 | null | https://arxiv.org/abs/2205.07404v1 | https://arxiv.org/pdf/2205.07404v1.pdf | A New Outlier Removal Strategy Based on Reliability of Correspondence Graph for Fast Point Cloud Registration | Registration is a basic yet crucial task in point cloud processing. In correspondence-based point cloud registration, matching correspondences by point feature techniques may lead to an extremely high outlier ratio. Current methods still suffer from low efficiency, accuracy, and recall rate. We use a simple and intuitive method to describe the 6-DOF (degree of freedom) curtailment process in point cloud registration and propose an outlier removal strategy based on the reliability of the correspondence graph. The method constructs the corresponding graph according to the given correspondences and designs the concept of the reliability degree of the graph node for optimal candidate selection and the reliability degree of the graph edge to obtain the global maximum consensus set. The presented method could achieve fast and accurate outliers removal along with gradual aligning parameters estimation. Extensive experiments on simulations and challenging real-world datasets demonstrate that the proposed method can still perform effective point cloud registration even the correspondence outlier ratio is over 99%, and the efficiency is better than the state-of-the-art. Code is available at https://github.com/WPC-WHU/GROR. | ['Ming Huang', 'Hao Wu', 'Jicheng Dai', 'Hong Xie', 'Pengcheng Wei', 'Li Yan'] | 2022-05-16 | null | null | null | null | ['point-cloud-registration'] | ['computer-vision'] | [-3.87971401e-01 -3.40321481e-01 1.37245417e-01 9.77697317e-03
-6.21907473e-01 -3.22660685e-01 1.34735122e-01 2.94141144e-01
-5.44834062e-02 1.83536083e-01 -2.99609601e-01 2.44626980e-02
-3.93445075e-01 -7.54239917e-01 -6.03641808e-01 -5.92447996e-01
-1.08731367e-01 6.35922968e-01 4.22960997e-01 -1.93453193e-01
6.14807427e-01 9.29519117e-01 -1.36475265e+00 -5.78069568e-01
1.24111664e+00 8.29711854e-01 1.57245532e-01 8.13547149e-02
1.33976880e-02 -2.67264307e-01 -3.98615807e-01 -1.06203124e-01
5.26794851e-01 2.29514632e-02 -2.70046860e-01 6.54943958e-02
1.97739527e-01 -1.15207464e-01 -2.33018309e-01 1.23921549e+00
6.27904177e-01 1.43597528e-01 4.10512537e-01 -1.49742544e+00
-2.18225062e-01 -1.18297271e-01 -8.58188331e-01 -1.48105323e-02
4.94469702e-01 -5.58912084e-02 5.49354255e-01 -1.30997562e+00
5.41644812e-01 8.95541072e-01 7.99113572e-01 4.17313091e-02
-8.69192481e-01 -9.44530547e-01 -4.77023609e-02 1.35582060e-01
-2.09934974e+00 -2.86463737e-01 1.00240886e+00 -4.07451898e-01
5.93098402e-01 4.32689637e-01 6.99879944e-01 2.51518369e-01
3.01314354e-01 4.57091480e-02 5.76339424e-01 -6.73048273e-02
1.90333247e-01 -5.62543154e-01 3.37427631e-02 5.94503522e-01
8.01969528e-01 9.68176499e-02 -4.58231360e-01 -7.11275816e-01
1.00358522e+00 4.52686667e-01 -3.63264292e-01 -4.41885054e-01
-1.34030485e+00 5.05578518e-01 7.07303107e-01 2.45687991e-01
-3.09369564e-01 3.66237015e-02 3.17840502e-02 2.72626668e-01
3.37857336e-01 1.13300070e-01 -1.71793029e-01 9.82743353e-02
-8.88964593e-01 3.33450973e-01 5.18315732e-01 1.38914788e+00
9.78168309e-01 -1.86597779e-01 2.52070069e-01 5.77787340e-01
5.39418578e-01 6.66022897e-01 -2.65670270e-02 -6.92656219e-01
5.12521267e-01 8.43861878e-01 9.92603898e-02 -1.52260351e+00
-4.05350000e-01 -3.59876961e-01 -1.08949494e+00 2.39513859e-01
1.08833238e-01 1.96458548e-01 -7.75593460e-01 1.05285704e+00
6.33490324e-01 7.60996222e-01 -3.87052506e-01 9.48222697e-01
6.76595151e-01 4.84409839e-01 -3.25463295e-01 -4.27939773e-01
1.02686357e+00 -3.17830443e-01 -6.18127465e-01 7.44014755e-02
4.06955868e-01 -1.20523596e+00 6.37778640e-01 3.87585126e-02
-8.91075492e-01 -3.14786673e-01 -1.07419717e+00 2.05198005e-01
1.55293554e-01 -3.90891954e-02 2.44128734e-01 1.12456575e-01
-8.63614917e-01 7.21234500e-01 -9.15019929e-01 -3.16585511e-01
2.25817010e-01 6.55972898e-01 -4.71246123e-01 -1.84940785e-01
-6.89300239e-01 5.49553573e-01 2.20633328e-01 4.01768118e-01
-5.43092266e-02 -6.11661375e-01 -5.56807220e-01 -1.87930748e-01
3.18528861e-01 -8.41114521e-01 6.90396905e-01 -3.11995506e-01
-1.14872158e+00 6.48770511e-01 -3.35736126e-01 2.14186355e-01
5.72668314e-01 -1.17274128e-01 -3.15100372e-01 -5.76447928e-03
3.36015791e-01 1.44459205e-04 5.41097045e-01 -1.38723099e+00
-3.12729806e-01 -5.76209784e-01 -6.45044029e-01 2.94957101e-01
1.66411176e-01 -8.55431333e-03 -7.66807318e-01 -4.69103634e-01
1.17607081e+00 -1.04021406e+00 -5.02501965e-01 1.42409280e-01
-2.59761304e-01 -2.62536287e-01 9.31735992e-01 -4.81697023e-01
1.19013798e+00 -2.30354261e+00 -1.15467928e-01 9.12409782e-01
3.83289278e-01 -4.74435166e-02 4.93925437e-02 6.58761799e-01
-6.12437055e-02 1.04012070e-02 -2.88413018e-01 -9.77668390e-02
-4.08683002e-01 -4.87351641e-02 3.92182022e-02 1.07108629e+00
-1.67582467e-01 4.29518759e-01 -1.00109088e+00 -4.13187206e-01
4.45710838e-01 3.87626737e-01 -3.95283401e-01 1.53815776e-01
3.64825130e-01 6.77123070e-01 -7.34834909e-01 9.28864062e-01
1.24240482e+00 -1.28572151e-01 -1.29011571e-01 -2.66607046e-01
-1.61419243e-01 -2.62931436e-01 -1.71137953e+00 1.59783220e+00
1.07872941e-01 1.33445024e-01 -6.06912524e-02 -3.92435938e-01
1.41290855e+00 1.23560466e-01 9.72502172e-01 -3.49326819e-01
1.55468717e-01 5.84440947e-01 8.31057876e-03 -2.03519210e-01
4.74133641e-01 1.74595818e-01 1.23451259e-02 3.64759937e-02
-4.81398702e-01 -3.55987549e-01 -2.50447541e-01 7.25759491e-02
1.09163404e+00 -1.32462785e-01 4.73718137e-01 -5.14621735e-01
6.04984164e-01 -3.38217542e-02 8.39409947e-01 3.26751024e-01
-5.38226850e-02 9.36614037e-01 7.33049363e-02 -2.77200818e-01
-1.04539752e+00 -1.09474802e+00 -2.36537114e-01 6.23485073e-02
9.57021654e-01 -5.35248756e-01 -3.94225597e-01 -8.15899074e-02
3.15541983e-01 2.84192026e-01 2.46705152e-02 -2.03687966e-01
-7.68084526e-01 -4.04433459e-01 4.55794223e-02 1.62305892e-01
3.36484820e-01 -5.77826321e-01 -1.02413848e-01 6.84119314e-02
-1.35588631e-01 -9.42562878e-01 -5.68150222e-01 -5.76088011e-01
-1.29425895e+00 -1.23228526e+00 -2.63281643e-01 -8.25410604e-01
1.29956210e+00 6.14813745e-01 8.49710763e-01 8.42841983e-01
4.00392078e-02 6.25611190e-03 -3.19042116e-01 -7.98978060e-02
-9.59158242e-02 -4.70408127e-02 3.24532211e-01 -1.79215133e-01
2.60177195e-01 -8.32851112e-01 -7.80534804e-01 7.01516032e-01
-6.78788483e-01 -2.31331632e-01 3.96221817e-01 5.70205331e-01
9.09376323e-01 6.66718408e-02 2.18055487e-01 -3.49128783e-01
4.94683713e-01 -4.13431555e-01 -9.59744215e-01 -7.76917636e-02
-6.97575986e-01 -2.91757584e-01 4.05518234e-01 -1.72749981e-01
-3.61910075e-01 2.85942703e-01 -1.31792665e-01 -7.52872765e-01
7.31767118e-02 3.23163450e-01 -2.30832875e-01 -5.66937149e-01
3.98918539e-01 -1.00410178e-01 -3.25220451e-02 -4.53334033e-01
1.15369730e-01 5.59111714e-01 5.01845658e-01 -6.13618910e-01
1.38476896e+00 6.22071087e-01 4.01455194e-01 -7.54179478e-01
-5.53081930e-02 -9.60232019e-01 -7.92189658e-01 -4.12453413e-01
2.47715488e-01 -9.42742288e-01 -9.15396512e-01 3.85918319e-01
-1.30551112e+00 4.54525083e-01 -6.04457520e-02 5.51407218e-01
-3.25511038e-01 7.22779274e-01 -3.50435227e-01 -6.78365827e-01
-4.24636513e-01 -1.19951487e+00 1.10676730e+00 1.21901736e-01
2.13789269e-02 -5.32646716e-01 9.82126780e-03 -1.29086316e-01
9.16729569e-02 7.54033267e-01 4.78051424e-01 -3.39415610e-01
-9.95234013e-01 -5.96952438e-01 -1.40233368e-01 -3.34175587e-01
1.84987426e-01 3.87551546e-01 -4.28528726e-01 -4.69988227e-01
1.28565609e-01 7.07128167e-01 2.28962436e-01 2.03122735e-01
9.45865870e-01 -1.00192390e-01 -5.92874587e-01 7.58642614e-01
1.78883564e+00 3.30753028e-02 8.49668503e-01 3.07685256e-01
9.00933802e-01 1.23423912e-01 9.27066028e-01 5.45669198e-01
1.83519647e-01 7.92162418e-01 6.54074073e-01 3.82171236e-02
2.22643733e-01 -2.41934776e-01 2.04938687e-02 1.48137701e+00
-3.10603440e-01 1.09703280e-02 -1.26806033e+00 4.57038045e-01
-2.18262768e+00 -7.11255670e-01 -7.85601616e-01 2.64312339e+00
2.94542283e-01 4.84124608e-02 -2.14531749e-01 1.28408223e-01
1.14359760e+00 -9.08656493e-02 -2.87785619e-01 4.36142161e-02
1.50504053e-01 -1.09758168e-01 8.04408550e-01 6.29871428e-01
-6.43393338e-01 7.87670553e-01 5.50226736e+00 7.08285868e-01
-7.71973133e-01 5.81295639e-02 -3.81967165e-02 1.88988417e-01
-2.10113525e-01 3.65020990e-01 -6.08000517e-01 5.85998833e-01
4.34466779e-01 -4.07723725e-01 1.87960565e-01 8.05867493e-01
3.12836856e-01 -7.66545832e-02 -6.95070028e-01 1.37837338e+00
-1.88861061e-02 -1.27448881e+00 -3.65731567e-02 2.10804105e-01
6.50995910e-01 1.97701767e-01 -4.07876879e-01 -4.63705599e-01
-1.40715137e-01 -6.48851216e-01 3.52671802e-01 7.57205069e-01
8.32440853e-01 -8.96770537e-01 8.55543971e-01 3.92944634e-01
-1.60269666e+00 2.77923524e-01 -6.62216544e-01 -2.23733694e-03
2.40396231e-01 8.71465981e-01 -5.48345208e-01 1.10410047e+00
6.96808279e-01 7.51409352e-01 -4.29692417e-01 1.56760395e+00
-6.48976043e-02 7.17914030e-02 -7.60957599e-01 2.47272819e-01
-2.89460599e-01 -8.20369482e-01 1.03141761e+00 5.89634299e-01
8.10404539e-01 3.88403922e-01 3.83476555e-01 5.34138560e-01
1.58193670e-02 4.31972355e-01 -7.21980572e-01 6.72604799e-01
1.05777586e+00 1.01895940e+00 -9.23783600e-01 3.67812254e-02
-3.13271105e-01 8.04613650e-01 1.93485931e-01 1.61366448e-01
-7.88727343e-01 -3.05180639e-01 7.91103780e-01 4.94317651e-01
-1.52541190e-01 -6.12765312e-01 -4.98386532e-01 -1.08720624e+00
4.34017569e-01 -5.93753695e-01 1.17187954e-01 -6.71267092e-01
-1.22315574e+00 5.50053060e-01 -1.63191296e-02 -1.97497118e+00
2.53021181e-01 -2.85888404e-01 -8.04736078e-01 7.44052351e-01
-1.10922980e+00 -7.89218724e-01 -8.71590555e-01 7.01508820e-01
1.08214326e-01 5.18004075e-02 5.49577951e-01 5.25349438e-01
-4.34427232e-01 4.17506337e-01 2.79140979e-01 -2.77027171e-02
6.83301687e-01 -6.13606334e-01 4.95381147e-01 1.09607100e+00
-2.62827843e-01 7.84287035e-01 7.87714541e-01 -1.14297926e+00
-1.55907154e+00 -8.43980551e-01 5.95016122e-01 -2.57924467e-01
5.18469334e-01 -2.03460276e-01 -9.79624569e-01 5.04681528e-01
-3.48804355e-01 1.96527392e-01 2.54036784e-01 -1.89634308e-01
7.09639564e-02 -2.53301382e-01 -1.30237806e+00 6.85999990e-01
1.25898385e+00 -7.53062814e-02 -4.75344151e-01 4.76427764e-01
6.39658272e-01 -9.28924024e-01 -1.02643967e+00 8.22565734e-01
2.70859361e-01 -8.27122808e-01 1.03870022e+00 2.24875718e-01
-2.05013424e-01 -1.01817715e+00 -3.66816148e-02 -9.24102128e-01
-4.64590043e-01 -7.69191444e-01 4.96288948e-02 1.18351138e+00
-6.50388673e-02 -8.67405415e-01 7.12530136e-01 4.23034132e-01
-2.34075472e-01 -7.38685369e-01 -1.26150548e+00 -9.01143909e-01
-3.93687427e-01 -2.37788007e-01 9.04562891e-01 1.07976258e+00
-2.05823511e-01 -7.44999349e-02 -6.10096417e-02 9.08073306e-01
9.41940784e-01 4.46247496e-02 1.08208263e+00 -1.66078937e+00
3.43938351e-01 -1.51320964e-01 -9.42813873e-01 -8.60063672e-01
-1.40206233e-01 -7.94727981e-01 -6.54941192e-04 -1.61313319e+00
-1.02842003e-01 -8.63289475e-01 -4.83572967e-02 9.30391848e-02
-1.91576436e-01 5.58002628e-02 9.73018110e-02 8.73512506e-01
-2.85490245e-01 5.10555506e-01 1.04880679e+00 3.34686130e-01
-4.43306625e-01 1.75001964e-01 -2.64582098e-01 6.99663222e-01
9.12549496e-01 -7.56813765e-01 -5.92339337e-02 -4.07375932e-01
2.24052593e-01 1.54472828e-01 2.08125100e-01 -1.23171282e+00
6.39580667e-01 -2.81732649e-01 2.19606072e-01 -9.12292004e-01
2.56692052e-01 -1.34908390e+00 8.71962905e-01 5.49693763e-01
6.48148179e-01 6.10241532e-01 7.35890716e-02 7.15440214e-01
-1.94219306e-01 -6.78638816e-02 6.47392392e-01 1.41461447e-01
-4.80859995e-01 7.94898570e-01 3.15558463e-01 -3.51740897e-01
1.25162578e+00 -5.14273405e-01 -3.13440293e-01 -2.18418375e-01
-3.50302994e-01 3.67701292e-01 1.15481925e+00 2.82656372e-01
1.01721430e+00 -1.63081229e+00 -8.11686754e-01 4.23306346e-01
4.23432291e-01 3.97888213e-01 1.23454347e-01 8.96934867e-01
-9.34709609e-01 -2.74228960e-01 -2.46884227e-02 -1.02974105e+00
-1.21857917e+00 3.13265413e-01 2.82773376e-01 1.52163148e-01
-7.85870194e-01 4.34560508e-01 -2.39573583e-01 -3.62124622e-01
-2.15325654e-01 -2.02365145e-01 3.51303935e-01 -4.58733857e-01
4.68516238e-02 5.68207383e-01 2.64607608e-01 -1.01821232e+00
-7.20071197e-01 1.28959024e+00 2.92324811e-01 2.77532041e-01
1.03004849e+00 -1.88640207e-01 -4.40514088e-01 3.20222625e-03
1.04455578e+00 4.59631741e-01 -8.19348037e-01 -3.08485553e-02
-2.39254944e-02 -1.02650869e+00 -1.26222938e-01 7.26408884e-02
-1.06181943e+00 2.53923267e-01 6.02755547e-01 -4.41538803e-02
9.75344062e-01 -1.13802530e-01 8.38089585e-01 1.15939699e-01
7.36065507e-01 -8.12364399e-01 -3.20305884e-01 2.95107245e-01
1.00091505e+00 -1.12238204e+00 6.27728224e-01 -1.03993511e+00
-1.10888243e-01 1.04236269e+00 6.15548372e-01 -7.62494981e-01
9.24244463e-01 1.56275213e-01 -7.01888949e-02 -4.56800818e-01
-5.71839958e-02 -2.68391985e-02 2.40114123e-01 3.63809317e-01
9.70560461e-02 1.61386337e-02 -6.04661763e-01 1.06729425e-01
-2.63597816e-01 -2.54343778e-01 3.79157573e-01 8.89622688e-01
-4.88240868e-01 -1.10946989e+00 -7.39243865e-01 4.75331932e-01
-1.05495267e-01 8.15585330e-02 -3.05753481e-02 7.90525436e-01
-9.42801312e-02 8.77146602e-01 1.61710531e-01 -5.06482542e-01
6.68298125e-01 -3.68032753e-01 1.80012479e-01 -4.76931512e-01
-3.37161243e-01 2.36466005e-01 -3.76363218e-01 -7.55311847e-01
-3.24651957e-01 -6.43136501e-01 -1.45089436e+00 -7.05446303e-01
-7.19951868e-01 3.20573896e-01 6.49923265e-01 5.69801569e-01
7.31289208e-01 -1.84961669e-02 9.45316195e-01 -9.55166340e-01
-1.55184388e-01 -5.82575023e-01 -6.05434120e-01 5.36408305e-01
1.61188915e-01 -7.81791270e-01 -5.18486261e-01 -3.71588349e-01] | [7.721286773681641, -2.8823838233947754] |
6d49248c-f959-4dc8-a448-2abe44846f81 | ido-vfi-identifying-dynamics-via-optical-flow | 2305.10198 | null | https://arxiv.org/abs/2305.10198v2 | https://arxiv.org/pdf/2305.10198v2.pdf | IDO-VFI: Identifying Dynamics via Optical Flow Guidance for Video Frame Interpolation with Events | Video frame interpolation aims to generate high-quality intermediate frames from boundary frames and increase frame rate. While existing linear, symmetric and nonlinear models are used to bridge the gap from the lack of inter-frame motion, they cannot reconstruct real motions. Event cameras, however, are ideal for capturing inter-frame dynamics with their extremely high temporal resolution. In this paper, we propose an event-and-frame-based video frame interpolation method named IDO-VFI that assigns varying amounts of computation for different sub-regions via optical flow guidance. The proposed method first estimates the optical flow based on frames and events, and then decides whether to further calculate the residual optical flow in those sub-regions via a Gumbel gating module according to the optical flow amplitude. Intermediate frames are eventually generated through a concise Transformer-based fusion network. Our proposed method maintains high-quality performance while reducing computation time and computational effort by 10% and 17% respectively on Vimeo90K datasets, compared with a unified process on the whole region. Moreover, our method outperforms state-of-the-art frame-only and frames-plus-events methods on multiple video frame interpolation benchmarks. Codes and models are available at https://github.com/shicy17/IDO-VFI. | ['Yibo Zhang', 'Boyi Wei', 'Yuzhen Li', 'Wenzhuo Li', 'Jing Jin', 'Hanxiao Liu', 'Chenyang Shi'] | 2023-05-17 | null | null | null | null | ['event-based-optical-flow', 'video-frame-interpolation'] | ['computer-vision', 'computer-vision'] | [ 1.32117663e-02 -4.24710035e-01 -2.68606633e-01 -1.92657590e-01
-4.46973979e-01 -2.76870787e-01 3.90426874e-01 -2.13548183e-01
-3.24281007e-01 9.03845489e-01 1.96724340e-01 -1.50462225e-01
1.44027919e-01 -7.64256001e-01 -6.20764434e-01 -6.28607929e-01
2.55389549e-02 -1.09844238e-01 5.96733749e-01 2.13697124e-02
2.31636018e-01 3.96064550e-01 -1.47932541e+00 2.84276903e-01
9.09071803e-01 1.12586963e+00 2.42966190e-01 7.95608401e-01
-2.51010768e-02 1.33984208e+00 -1.76299155e-01 -9.48469341e-02
3.38075995e-01 -6.91901147e-01 -6.13994002e-01 9.98732895e-02
5.17336369e-01 -8.39008629e-01 -7.79314160e-01 6.90674424e-01
4.29706872e-01 3.80991071e-01 1.59395725e-01 -1.25200820e+00
-3.46737862e-01 1.74148992e-01 -5.71710765e-01 5.40801167e-01
7.03320563e-01 4.91496712e-01 6.04241669e-01 -1.00572908e+00
9.06167626e-01 1.13296437e+00 4.30972427e-01 5.74826777e-01
-1.15432835e+00 -7.64138401e-01 5.02214655e-02 5.18350005e-01
-1.23386884e+00 -6.52474523e-01 7.45355546e-01 -4.58492190e-01
6.56947851e-01 2.92772442e-01 8.22965205e-01 7.54990876e-01
2.01299503e-01 6.48493409e-01 6.25226021e-01 -1.63664311e-01
-2.55747195e-02 -5.46796739e-01 -2.37119213e-01 7.30077624e-01
-1.16379865e-01 3.44472110e-01 -7.89403379e-01 1.18950851e-01
1.26943600e+00 4.99598533e-02 -7.78630197e-01 8.65878910e-02
-1.63997388e+00 4.44832087e-01 3.25501710e-01 6.32696971e-03
-5.70731997e-01 2.61127830e-01 4.38034952e-01 -5.16826706e-03
4.87595707e-01 -1.73989087e-01 -1.06224351e-01 -3.40098143e-01
-1.27304304e+00 3.53288442e-01 4.90928769e-01 9.81488764e-01
8.46911490e-01 3.09961110e-01 -4.75607604e-01 4.18409914e-01
1.70434728e-01 2.49827027e-01 2.11024731e-01 -1.56120396e+00
4.74664062e-01 1.63387865e-01 4.83642727e-01 -1.08164334e+00
-9.89755541e-02 -4.44433838e-02 -1.02704489e+00 2.71684527e-01
6.36014521e-01 -1.21564649e-01 -6.83382869e-01 1.45232880e+00
5.73073864e-01 1.01304078e+00 -1.87130153e-01 1.20631564e+00
7.58318484e-01 1.06059158e+00 5.02335802e-02 -6.19846582e-01
1.16583955e+00 -1.06808150e+00 -9.50613618e-01 1.31208047e-01
2.99710661e-01 -1.05231130e+00 6.62824094e-01 3.12890977e-01
-1.64744520e+00 -9.16919768e-01 -7.06634283e-01 -3.45154166e-01
4.51313525e-01 3.43063548e-02 3.43223751e-01 1.36795893e-01
-1.05540240e+00 4.79373932e-01 -1.00850570e+00 1.47910550e-01
4.74416375e-01 2.52985477e-01 -1.19816475e-01 -1.42902941e-01
-1.08771122e+00 3.59215766e-01 3.43281239e-01 2.28668272e-01
-9.18861032e-01 -1.08896339e+00 -9.23953116e-01 -5.27089015e-02
2.43023962e-01 -9.37045097e-01 1.08864188e+00 -9.64639664e-01
-1.52767110e+00 2.45684326e-01 -8.00551534e-01 -4.97849792e-01
7.72080898e-01 -1.77087858e-01 -2.46343568e-01 4.58247542e-01
3.57055515e-02 9.19124901e-01 8.26266706e-01 -1.07946289e+00
-1.05221570e+00 1.61012799e-01 2.46238530e-01 2.22101405e-01
9.13752019e-02 6.23751208e-02 -6.52719259e-01 -8.68990958e-01
-2.53524948e-02 -8.32770526e-01 -1.65395141e-01 4.66161013e-01
-5.93545325e-02 -7.19789714e-02 1.07231283e+00 -8.34970653e-01
1.52683222e+00 -2.09576464e+00 8.33218731e-03 -3.55981112e-01
3.68860483e-01 4.61452931e-01 1.40525103e-01 -4.32427377e-02
6.20180368e-02 -2.15716526e-01 -1.48431674e-01 -3.48766983e-01
-5.94549835e-01 8.92783180e-02 -2.67419666e-01 4.49640304e-01
1.39807642e-01 7.05392182e-01 -1.24347806e+00 -7.13988543e-01
8.64556611e-01 9.71710563e-01 -7.99174905e-01 2.82585651e-01
-4.84177098e-02 8.91851842e-01 -1.44150376e-01 5.90878129e-01
7.85145223e-01 -2.22079322e-01 -5.90501167e-02 -5.42723119e-01
-3.60156685e-01 1.62413090e-01 -1.30979121e+00 1.83942223e+00
-4.92319942e-01 9.01510596e-01 -7.63969272e-02 -6.31226838e-01
6.24524415e-01 5.25606573e-01 8.22107673e-01 -5.18962860e-01
1.11576878e-01 1.53077215e-01 -1.87420830e-01 -3.58449966e-01
6.86056912e-01 2.31870577e-01 4.80677903e-01 -1.64439287e-02
-9.79628786e-02 3.53732407e-01 6.76413238e-01 1.42561659e-01
8.34711790e-01 6.09328866e-01 1.75739616e-01 -7.78563619e-02
9.42220449e-01 -2.60060668e-01 1.07883883e+00 3.31656873e-01
-4.53123629e-01 9.21796620e-01 2.53153116e-01 -7.55363107e-01
-1.07241094e+00 -1.11117315e+00 -7.92105868e-02 6.35329962e-01
6.36296093e-01 -5.00840127e-01 -6.26687109e-01 -2.06704706e-01
-3.61729562e-01 4.76080239e-01 -2.76600003e-01 1.79272696e-01
-9.38243151e-01 -4.19592828e-01 2.21454963e-01 3.52132559e-01
6.50815427e-01 -1.00015724e+00 -7.20219731e-01 7.17020392e-01
-8.34235728e-01 -1.42963409e+00 -9.12592530e-01 -6.03536487e-01
-9.43548322e-01 -1.12505913e+00 -8.36384535e-01 -7.34145224e-01
6.45117521e-01 5.19731224e-01 1.19370067e+00 2.28044480e-01
-1.24671705e-01 -2.90829390e-02 -2.66713977e-01 3.03868175e-01
-3.68686944e-01 -3.67923200e-01 -1.73660666e-01 3.82225543e-01
-5.81926433e-03 -6.17403924e-01 -1.22855008e+00 5.74146092e-01
-9.93872166e-01 5.50591946e-01 -5.76503724e-02 8.55768681e-01
8.36479425e-01 -2.23977014e-01 3.06242198e-01 -3.84878635e-01
7.60685876e-02 -4.23945427e-01 -7.43480682e-01 -3.07313837e-02
-2.75919139e-01 -2.43699759e-01 7.91906595e-01 -3.92410278e-01
-1.31980538e+00 1.20113574e-01 -8.97058193e-03 -7.77933240e-01
9.59467143e-02 -1.07853748e-01 8.97432342e-02 -3.54213384e-03
2.96326339e-01 3.50831956e-01 -1.56759545e-01 -4.07486744e-02
1.99332029e-01 2.66918033e-01 8.74798596e-01 -5.16928971e-01
5.22317827e-01 8.76274407e-01 1.18384492e-02 -6.55594051e-01
-5.02920032e-01 -4.28951979e-01 -4.26771551e-01 -6.86171949e-01
1.05086076e+00 -1.22585380e+00 -9.67996001e-01 6.56414509e-01
-1.50511301e+00 -3.06683391e-01 -2.17034116e-01 7.05646694e-01
-5.68736613e-01 5.82007170e-01 -1.02829862e+00 -5.14988065e-01
-4.24043834e-01 -1.53329492e+00 9.79686141e-01 3.94927233e-01
6.65101781e-02 -9.26832974e-01 -1.21142119e-01 3.15923303e-01
2.54138321e-01 4.48648512e-01 7.52994865e-02 5.24208486e-01
-1.07931459e+00 2.80089706e-01 -3.86583388e-01 1.97060853e-01
2.96608984e-01 3.32569867e-01 -6.97346568e-01 -2.76923686e-01
-1.81913480e-01 2.03935161e-01 7.07574069e-01 7.83423305e-01
1.27957547e+00 -2.93379515e-01 -3.18765789e-02 1.00827706e+00
1.59214556e+00 4.11348581e-01 1.08103693e+00 1.02129139e-01
9.42547679e-01 1.67551845e-01 6.51780605e-01 7.53538489e-01
5.37035704e-01 8.70608687e-01 4.46582854e-01 3.61673790e-03
-5.63433588e-01 -1.33462980e-01 5.06565094e-01 8.21395516e-01
-6.16829813e-01 -4.70247865e-01 -5.31088531e-01 6.32454693e-01
-1.91976440e+00 -1.39493036e+00 -5.28197408e-01 2.10433197e+00
8.40437114e-01 1.24662951e-01 1.01933002e-01 2.16823921e-01
9.38991904e-01 3.77889395e-01 -3.33109617e-01 3.01980157e-03
-1.79201454e-01 1.47702217e-01 5.15131831e-01 9.09017980e-01
-9.78372335e-01 7.58136868e-01 4.82128286e+00 9.54944968e-01
-1.13649035e+00 1.80407986e-01 8.53099167e-01 -2.36015707e-01
-5.61230294e-02 1.73167974e-01 -8.11527431e-01 9.25900698e-01
8.67471695e-01 -1.06054224e-01 5.23810923e-01 3.48985672e-01
9.33493555e-01 -2.83836663e-01 -7.86563814e-01 1.15114498e+00
-2.35906079e-01 -1.94578147e+00 8.09498206e-02 -1.79317564e-01
9.18031454e-01 -2.26414368e-01 -4.07347381e-01 -2.82025188e-01
-8.11142102e-02 -6.28799975e-01 9.75811422e-01 7.03937769e-01
8.90003800e-01 -6.65544569e-01 4.02910143e-01 6.79434882e-03
-1.80151165e+00 2.00476006e-01 -1.26433074e-01 -2.36386299e-01
8.17872167e-01 7.80586183e-01 -1.72739774e-01 6.97535872e-01
7.08451509e-01 1.13646030e+00 -1.99070454e-01 1.09696245e+00
-2.01094151e-01 6.34767592e-01 -2.51694471e-01 6.74124360e-01
8.82812366e-02 -3.96199942e-01 5.85845470e-01 1.10474777e+00
4.71270353e-01 2.66174495e-01 2.17881843e-01 7.28388011e-01
7.00939819e-02 -2.51002401e-01 -1.94432095e-01 5.61885238e-01
6.36623919e-01 1.14483762e+00 -6.50070965e-01 -6.62360907e-01
-6.49278462e-01 1.03767550e+00 -1.81311443e-01 4.30527449e-01
-1.22367609e+00 -2.26307765e-01 9.19228911e-01 3.41917604e-01
1.98752448e-01 -3.68107527e-01 -1.51640773e-01 -1.48281312e+00
2.01955020e-01 -5.41036248e-01 2.25519747e-01 -9.69810486e-01
-6.56503916e-01 5.79974473e-01 -1.84586376e-01 -1.77697206e+00
-4.22412455e-01 -4.35810164e-02 -4.88177389e-01 7.69078076e-01
-1.83412230e+00 -8.73765051e-01 -8.09517205e-01 8.95306885e-01
9.26034570e-01 3.25898618e-01 2.23610878e-01 7.97084212e-01
-4.45090830e-01 3.14663500e-01 2.25836840e-02 2.16377050e-01
7.45371819e-01 -7.62867332e-01 3.39754343e-01 1.16385543e+00
-1.48308143e-01 1.95960417e-01 5.47009289e-01 -5.37842751e-01
-1.26151431e+00 -1.28391719e+00 8.72144401e-01 -5.42593002e-02
3.43682379e-01 6.65992647e-02 -9.67438042e-01 5.11785448e-01
3.04439604e-01 7.39300072e-01 4.73004393e-02 -9.35959816e-01
9.37770829e-02 -2.66478628e-01 -9.52052772e-01 5.82139790e-01
1.14699578e+00 -2.84648210e-01 -4.89119031e-02 9.99911129e-02
7.82652199e-01 -8.28944445e-01 -1.03028345e+00 2.50601262e-01
5.01496911e-01 -1.33893919e+00 1.27820635e+00 2.11154930e-02
8.62999320e-01 -8.56156647e-01 8.25410411e-02 -7.75843740e-01
-3.42554420e-01 -1.12392604e+00 -4.33188558e-01 1.26005948e+00
-1.96357206e-01 -3.88878405e-01 7.50315011e-01 3.25091690e-01
-1.92635760e-01 -6.98143661e-01 -9.36776340e-01 -5.67822993e-01
-5.50466895e-01 -5.21143198e-01 3.50353301e-01 8.21651757e-01
-4.35202211e-01 -1.10259086e-01 -5.97859383e-01 8.61135200e-02
8.91540587e-01 -5.52084483e-02 6.84493780e-01 -6.76275134e-01
-1.64802462e-01 -2.50521719e-01 -5.02146006e-01 -1.42862070e+00
9.31620970e-02 -5.46919584e-01 -1.03696428e-01 -1.41633153e+00
-1.46407455e-01 -3.51821542e-01 -6.06634207e-02 1.89240441e-01
-4.01125789e-01 6.02998435e-01 4.52230603e-01 3.82693410e-01
-4.88128513e-01 4.85699415e-01 1.55094242e+00 1.33301362e-01
-2.96370417e-01 -2.93585747e-01 3.08685005e-02 7.31866241e-01
6.16517603e-01 -1.56915396e-01 -4.10550594e-01 -6.29630864e-01
-1.54233441e-01 7.65387237e-01 6.90930307e-01 -1.28633904e+00
3.02422374e-01 -3.15184206e-01 4.45415646e-01 -6.45657659e-01
3.85516971e-01 -6.57762766e-01 5.87082863e-01 5.18893301e-01
-9.55505222e-02 2.18278110e-01 7.04059899e-02 5.45627058e-01
-4.81129080e-01 9.98739451e-02 9.93115962e-01 -1.62828416e-02
-9.95340466e-01 6.78689361e-01 -3.38599175e-01 1.76145867e-01
1.07738888e+00 -3.77680719e-01 -2.65970558e-01 -4.07513738e-01
-6.15927517e-01 6.29627854e-02 5.85115016e-01 3.30631942e-01
7.25384295e-01 -1.55917561e+00 -8.40882003e-01 1.39764398e-01
-3.26455742e-01 2.81927109e-01 5.75748682e-01 9.75338280e-01
-1.02685452e+00 1.91126093e-01 -2.79828101e-01 -8.52420032e-01
-1.11410832e+00 4.05532956e-01 2.64717042e-01 -2.37270817e-01
-7.69964516e-01 6.03315651e-01 4.25504476e-01 3.83745730e-01
-3.35922949e-02 -4.53970432e-01 1.17822193e-01 -8.87866020e-02
7.70217657e-01 7.84702241e-01 -1.73176467e-01 -9.16356087e-01
-2.14708880e-01 5.20824432e-01 3.16973656e-01 1.35162234e-01
1.00759912e+00 -4.62610394e-01 7.17375726e-02 2.74102967e-02
1.18216515e+00 -1.90602571e-01 -1.96922565e+00 -8.64043385e-02
-5.38714647e-01 -9.61365938e-01 1.15228057e-01 -1.85048327e-01
-1.50961471e+00 6.87115729e-01 5.08455396e-01 -1.74182490e-01
1.58111227e+00 -5.24561524e-01 1.37387156e+00 -5.11032403e-01
3.88408363e-01 -6.28589571e-01 -1.86839148e-01 2.52494007e-01
5.24178565e-01 -1.09674907e+00 6.30229488e-02 -6.80091023e-01
-2.55229145e-01 1.26214409e+00 5.51718891e-01 -2.99360871e-01
4.47779298e-01 2.91113317e-01 -2.28483945e-01 4.56402123e-01
-7.84768283e-01 6.33721128e-02 1.42997369e-01 4.28037703e-01
6.08372927e-01 -2.31485754e-01 -6.12343967e-01 -1.44526035e-01
1.86773002e-01 7.14086235e-01 7.46021330e-01 6.73299670e-01
-1.10835552e-01 -9.26173568e-01 -5.10184944e-01 1.70827165e-01
-4.77156937e-01 -1.24562182e-01 7.65842676e-01 5.99606812e-01
2.47427002e-01 1.04371595e+00 2.44947121e-01 -1.17331922e-01
5.97907342e-02 -4.50704217e-01 4.66807812e-01 1.27739549e-01
-4.03789848e-01 2.98787355e-01 -1.32573424e-02 -1.01286471e+00
-8.93848956e-01 -6.55101359e-01 -1.43362176e+00 -6.84945822e-01
-6.29013851e-02 -7.00924620e-02 1.27468631e-01 6.22763276e-01
5.13047993e-01 7.26389885e-01 6.14428222e-01 -1.33204186e+00
2.19822198e-01 -4.74020243e-01 -8.23543891e-02 6.16592169e-01
5.79995990e-01 -4.96976882e-01 -4.35091138e-01 7.71466672e-01] | [10.719619750976562, -1.495254635810852] |
05522907-6a3e-4b89-b6d1-7a261ae8658b | lvlm-ehub-a-comprehensive-evaluation | 2306.09265 | null | https://arxiv.org/abs/2306.09265v1 | https://arxiv.org/pdf/2306.09265v1.pdf | LVLM-eHub: A Comprehensive Evaluation Benchmark for Large Vision-Language Models | Large Vision-Language Models (LVLMs) have recently played a dominant role in multimodal vision-language learning. Despite the great success, it lacks a holistic evaluation of their efficacy. This paper presents a comprehensive evaluation of publicly available large multimodal models by building a LVLM evaluation Hub (LVLM-eHub). Our LVLM-eHub consists of $8$ representative LVLMs such as InstructBLIP and MiniGPT-4, which are thoroughly evaluated by a quantitative capability evaluation and an online arena platform. The former evaluates $6$ categories of multimodal capabilities of LVLMs such as visual question answering and embodied artificial intelligence on $47$ standard text-related visual benchmarks, while the latter provides the user-level evaluation of LVLMs in an open-world question-answering scenario. The study reveals several innovative findings. First, instruction-tuned LVLM with massive in-domain data such as InstructBLIP heavily overfits many existing tasks, generalizing poorly in the open-world scenario. Second, instruction-tuned LVLM with moderate instruction-following data may result in object hallucination issues (i.e., generate objects that are inconsistent with target images in the descriptions). It either makes the current evaluation metric such as CIDEr for image captioning ineffective or generates wrong answers. Third, employing a multi-turn reasoning evaluation framework can mitigate the issue of object hallucination, shedding light on developing an effective pipeline for LVLM evaluation. The findings provide a foundational framework for the conception and assessment of innovative strategies aimed at enhancing zero-shot multimodal techniques. Our LVLM-eHub will be available at https://github.com/OpenGVLab/Multi-Modality-Arena | ['Ping Luo', 'Yu Qiao', 'Siyuan Huang', 'Fanqing Meng', 'Meng Lei', 'Shuo Liu', 'Peng Gao', 'Kaipeng Zhang', 'Wenqi Shao', 'Peng Xu'] | 2023-06-15 | null | null | null | null | ['visual-question-answering-1', 'image-captioning', 'instruction-following'] | ['computer-vision', 'computer-vision', 'natural-language-processing'] | [-1.56736761e-01 3.97472968e-03 2.78478917e-02 -1.88550174e-01
-1.08335495e+00 -7.76178896e-01 6.54225349e-01 1.39823735e-01
-6.60171270e-01 3.59324545e-01 1.92360058e-01 -5.75662732e-01
1.36046007e-01 -4.99810487e-01 -8.97903204e-01 -3.48293632e-01
3.60063851e-01 4.80871916e-01 6.65701404e-02 -4.31411088e-01
2.47455612e-01 -3.61931100e-02 -1.78437364e+00 6.40699089e-01
8.92680824e-01 9.72014427e-01 3.82138550e-01 7.84736812e-01
-3.45011890e-01 1.22603345e+00 -6.94793642e-01 -8.75689268e-01
3.47672105e-02 -1.57365158e-01 -8.76535654e-01 -1.55260032e-02
1.19080627e+00 -5.51962197e-01 -2.43047848e-01 8.36679339e-01
8.54879737e-01 2.19333157e-01 4.01652813e-01 -1.76811481e+00
-1.02980244e+00 2.82702923e-01 -2.40475446e-01 1.54075176e-01
6.95803046e-01 9.21906054e-01 1.02205646e+00 -1.19528794e+00
7.74809837e-01 1.44878983e+00 4.66133863e-01 5.73637128e-01
-1.10336578e+00 -6.07126892e-01 1.27582490e-01 5.20571887e-01
-1.19930673e+00 -5.90730250e-01 4.68486190e-01 -3.30072433e-01
1.09104168e+00 2.87647635e-01 4.86049145e-01 1.34252679e+00
4.70353588e-02 1.12542808e+00 1.25359774e+00 -3.27189803e-01
-2.28739460e-03 3.80926639e-01 2.80197617e-02 1.10196829e+00
-1.93465594e-02 4.11989503e-02 -9.40574050e-01 2.19227523e-01
3.99202913e-01 -4.44060981e-01 -3.35348934e-01 -4.40726131e-01
-1.40072966e+00 7.53168106e-01 5.30153275e-01 -6.94987401e-02
-5.09754270e-02 2.92986482e-01 4.82034653e-01 3.75335187e-01
-1.47779966e-02 5.38912892e-01 -1.52086064e-01 -2.16012046e-01
-6.25746608e-01 3.39645118e-01 5.25713980e-01 1.13302755e+00
8.44965100e-01 1.02343127e-01 -5.72602034e-01 8.66236210e-01
5.57776153e-01 6.52599633e-01 2.96021581e-01 -1.33860266e+00
7.73285329e-01 6.75293326e-01 -7.68044144e-02 -8.58408391e-01
-3.11418355e-01 -2.04622895e-01 -1.75981432e-01 3.75926942e-01
5.81743836e-01 -1.85349695e-02 -8.14831614e-01 1.83201635e+00
1.13631152e-01 -4.21655923e-01 2.88998008e-01 9.93429184e-01
1.91927052e+00 6.00218832e-01 4.21671242e-01 3.27309757e-01
1.50656724e+00 -1.46484518e+00 -6.17863536e-01 -3.78292859e-01
7.01849461e-01 -8.34962070e-01 1.81742239e+00 2.18971819e-01
-1.32852995e+00 -6.73543513e-01 -7.50428796e-01 -4.36703265e-01
-6.89869881e-01 -8.93850699e-02 6.00409687e-01 6.07185662e-01
-1.31650794e+00 -2.01539248e-01 -1.99549064e-01 -4.30016875e-01
5.04431963e-01 1.04417019e-01 -5.27993083e-01 -5.10237634e-01
-9.22539413e-01 1.11061251e+00 1.16610095e-01 -8.91588554e-02
-1.35128391e+00 -7.43056357e-01 -1.16057086e+00 -1.20920792e-01
3.80588382e-01 -8.78943741e-01 1.38048089e+00 -9.37971711e-01
-8.79358590e-01 1.07283759e+00 -4.50729653e-02 -1.31033435e-01
4.29954201e-01 9.37964693e-02 -2.87913412e-01 3.94156307e-01
1.99138522e-01 1.55437398e+00 6.75949216e-01 -1.60775495e+00
-3.34836751e-01 -2.69319832e-01 5.53330779e-01 6.78896964e-01
-3.64681512e-01 -2.48260990e-01 -7.37135172e-01 -2.53445357e-01
-2.87589908e-01 -7.26878226e-01 1.15794756e-01 -2.46714591e-03
-2.44246721e-01 -3.30568373e-01 6.79703951e-01 -5.57453990e-01
9.50948656e-01 -1.83261669e+00 1.91552825e-02 -4.46751326e-01
3.75307411e-01 1.76141709e-01 -6.93572581e-01 6.93751931e-01
1.45015150e-01 1.72723114e-01 1.53724536e-01 -6.00232184e-01
2.07036182e-01 4.44454700e-02 -1.72095194e-01 6.47712722e-02
1.49202123e-01 1.44403899e+00 -8.39261413e-01 -7.92788088e-01
3.82231116e-01 3.51497352e-01 -5.27672350e-01 1.92605019e-01
-5.16070306e-01 2.29660258e-01 6.74829856e-02 1.13033164e+00
5.20862937e-01 -2.83731878e-01 -3.68839502e-01 -5.23206532e-01
-1.83834180e-01 -2.80865967e-01 -5.89998782e-01 1.99077821e+00
-3.78106475e-01 1.04888082e+00 -4.49880883e-02 -3.63185763e-01
3.97669762e-01 1.38230100e-01 6.24225177e-02 -1.18206084e+00
4.59273979e-02 -8.78913403e-02 -1.55903995e-01 -9.46527481e-01
7.20184743e-01 -4.62318175e-02 -4.53233793e-02 1.40912414e-01
4.78570253e-01 -3.39128524e-01 3.39820445e-01 6.47622406e-01
1.06038702e+00 2.71571696e-01 -2.36799613e-01 1.81061238e-01
3.30489963e-01 3.77953887e-01 -9.08961520e-02 9.48267460e-01
-6.46601498e-01 5.84480166e-01 3.55399847e-01 -5.05789071e-02
-6.60718262e-01 -1.26824498e+00 3.30465995e-02 1.50066745e+00
1.96209118e-01 -4.92185503e-01 -7.06831992e-01 -7.09948123e-01
-1.23141691e-01 9.40437794e-01 -3.16234827e-01 -3.05144005e-02
1.34227708e-01 -3.62741441e-01 9.24061537e-01 3.28055322e-01
7.63617754e-01 -1.15063548e+00 -7.03967154e-01 -2.91399926e-01
-7.22117484e-01 -1.43733406e+00 -2.49684513e-01 -1.12067118e-01
-4.31561172e-01 -1.10245180e+00 -8.06685865e-01 -9.30665791e-01
6.20585740e-01 5.00502229e-01 1.56275368e+00 9.10136849e-02
-3.22527677e-01 1.27599406e+00 -3.37830395e-01 -5.78226209e-01
-3.48752528e-01 -4.52481538e-01 -2.09901154e-01 -3.57637405e-01
4.28524762e-01 2.21374556e-01 -7.65517712e-01 3.71111870e-01
-8.54275286e-01 3.65456909e-01 6.41830862e-01 7.10560918e-01
4.32214588e-01 -4.43934321e-01 6.20848596e-01 -2.74354875e-01
8.91959310e-01 -3.59802902e-01 -4.33913141e-01 4.91401166e-01
-5.80567777e-01 -2.14962542e-01 1.03216633e-01 -4.82698858e-01
-1.03331268e+00 -3.52602690e-01 -1.66618392e-01 -6.93482339e-01
-2.67952412e-01 5.69514632e-01 -2.01794997e-01 -3.26938540e-01
7.26239622e-01 2.67063290e-01 8.62885192e-02 -7.47616887e-02
5.94500303e-01 5.52101493e-01 5.94169617e-01 -7.09499896e-01
5.38253486e-01 3.38751078e-01 -3.67802262e-01 -8.42510998e-01
-5.80472112e-01 -4.01697218e-01 -1.57406945e-02 -7.47956038e-01
1.14368534e+00 -1.30054712e+00 -1.10456634e+00 4.20693725e-01
-1.07325995e+00 -6.22227073e-01 -1.75776124e-01 2.09974319e-01
-6.39666855e-01 3.73078942e-01 -5.07884920e-01 -6.42479718e-01
-2.25913152e-01 -1.60757124e+00 1.14377666e+00 3.79512757e-01
-1.86119273e-01 -9.96438682e-01 -5.16733015e-03 1.41817653e+00
2.66085118e-01 -1.31524488e-01 1.06842732e+00 -3.37430149e-01
-7.57361829e-01 1.27889942e-02 -5.75871170e-01 3.52631748e-01
-5.64339101e-01 -1.17029756e-01 -1.19096351e+00 -3.27185452e-01
-3.88540804e-01 -1.17958140e+00 7.53345191e-01 1.08831346e-01
8.85643899e-01 2.95314956e-02 8.72104615e-02 4.28426266e-01
1.43238842e+00 -2.72069667e-02 6.68558419e-01 3.95167291e-01
8.21697533e-01 7.58573890e-01 6.70024931e-01 5.92942722e-02
1.10891151e+00 4.69886720e-01 8.73424649e-01 -2.29994535e-01
-5.78894317e-01 -3.55371952e-01 6.86818600e-01 5.70204675e-01
2.56832000e-02 -4.40331221e-01 -1.24295866e+00 6.04222119e-01
-1.64384210e+00 -9.82539117e-01 -2.28376046e-01 2.03402019e+00
6.48972273e-01 -2.31355309e-01 -8.95723477e-02 -5.24010122e-01
2.33575016e-01 1.38140768e-01 -5.73004246e-01 -4.74204689e-01
-3.51687521e-01 -2.09696978e-01 1.36561111e-01 4.14021701e-01
-8.02852094e-01 1.11051118e+00 6.06505346e+00 8.82415712e-01
-7.10582554e-01 3.75580698e-01 6.49426699e-01 -3.19726825e-01
-5.27413368e-01 -6.17319047e-02 -7.87891746e-01 2.19432965e-01
8.18998754e-01 7.06314072e-02 3.12878519e-01 5.80655813e-01
9.86375660e-02 -4.67336297e-01 -1.14188325e+00 1.29320395e+00
5.99980891e-01 -1.36312044e+00 4.16899502e-01 -7.50932470e-02
5.56537330e-01 3.62895966e-01 4.57396954e-01 8.93869221e-01
7.35525861e-02 -1.26018262e+00 9.44155931e-01 4.96645838e-01
7.96736836e-01 -5.60561478e-01 6.42758906e-01 2.68259078e-01
-8.87135446e-01 -2.13445183e-02 -1.78675756e-01 7.27746263e-02
6.17491528e-02 -2.38260522e-01 -7.57565081e-01 2.21282303e-01
9.27206993e-01 2.49720633e-01 -1.15767193e+00 9.23779011e-01
-1.43233570e-03 2.60282755e-01 8.21854770e-02 7.40916356e-02
4.23720956e-01 -3.80623750e-02 5.58056951e-01 1.09310770e+00
7.82169625e-02 -2.89518923e-01 3.47934775e-02 1.02838087e+00
-2.03566641e-01 1.67078182e-01 -8.48607898e-01 -2.28342444e-01
2.57723510e-01 1.37380195e+00 -3.36356431e-01 -1.66338339e-01
-9.44278598e-01 8.35678995e-01 4.43971455e-01 6.95701718e-01
-1.00188696e+00 9.33601335e-02 4.79320228e-01 -8.46757181e-03
-7.99200237e-02 -7.95947760e-03 -1.69790074e-01 -1.16312099e+00
-1.35069460e-01 -1.37395120e+00 4.29681331e-01 -1.56282771e+00
-1.21504796e+00 5.46134651e-01 1.05409078e-01 -1.07937860e+00
3.00985388e-02 -8.11702371e-01 -4.58802789e-01 6.15425766e-01
-1.45203626e+00 -1.51261926e+00 -7.08463550e-01 9.18106973e-01
9.52893913e-01 -3.34707201e-01 8.31173360e-01 2.17198446e-01
-5.91460288e-01 8.31083059e-01 -4.25407678e-01 -5.24134859e-02
1.03510439e+00 -1.08045948e+00 -1.40605062e-01 5.78824639e-01
2.06409916e-01 5.42024434e-01 6.49147511e-01 -5.45448959e-01
-1.76152170e+00 -8.77435029e-01 5.62122583e-01 -1.10059631e+00
6.48066938e-01 -1.38449460e-01 -6.48376405e-01 6.77742481e-01
8.29289317e-01 -2.22022444e-01 8.80431116e-01 -6.04024902e-02
-4.88298029e-01 1.19897306e-01 -1.06208313e+00 1.03593481e+00
9.02824342e-01 -6.80531621e-01 -5.72672129e-01 4.82857287e-01
7.59941101e-01 -3.98655176e-01 -8.19528699e-01 3.44406962e-01
4.45531994e-01 -1.16534722e+00 1.20679688e+00 -5.75452089e-01
9.05443549e-01 -1.66259661e-01 -4.86652315e-01 -1.12368453e+00
6.48105443e-02 -2.48899683e-01 -9.70187336e-02 1.21932185e+00
4.69286352e-01 -3.06448042e-01 5.48724174e-01 7.96256542e-01
-2.58521825e-01 -8.37000608e-01 -8.89937520e-01 -5.87632537e-01
7.58836567e-02 -7.50747144e-01 5.67776747e-02 8.32788944e-01
-1.37119189e-01 4.80483055e-01 3.67616862e-02 6.94716275e-02
7.47666121e-01 -2.65092522e-01 9.40806687e-01 -5.66188812e-01
-1.88961387e-01 -6.22021973e-01 -3.80731851e-01 -8.20089817e-01
1.41515866e-01 -1.02568817e+00 -1.71792850e-01 -1.85785484e+00
4.51840103e-01 -5.14495447e-02 -1.95968106e-01 6.70930445e-01
-2.45741159e-02 5.71252882e-01 6.25398278e-01 8.53725448e-02
-1.14998055e+00 6.15969658e-01 1.49287391e+00 -4.46442425e-01
1.23283140e-01 -6.20738387e-01 -8.07776451e-01 7.96324670e-01
5.65003037e-01 1.13939881e-01 -6.15696013e-01 -7.44644463e-01
3.78235728e-01 3.28007638e-02 9.53707278e-01 -9.42295134e-01
2.17331067e-01 -1.95881240e-02 3.89633656e-01 -7.25101054e-01
6.60892129e-01 -6.13000870e-01 -1.82456166e-01 1.34941325e-01
-3.73583317e-01 2.86676437e-01 5.37809491e-01 3.62516195e-01
-2.79973030e-01 -2.92668849e-01 4.13757682e-01 -2.98948765e-01
-1.33325624e+00 7.85428286e-03 -4.08471823e-01 3.00201207e-01
1.05557168e+00 -2.88461089e-01 -1.09668934e+00 -6.59266114e-01
-5.02300382e-01 6.61602318e-01 5.70910633e-01 7.27778614e-01
1.01629710e+00 -1.28508151e+00 -6.51942492e-01 -6.89617321e-02
7.81208515e-01 -2.50475734e-01 5.43948770e-01 9.47280407e-01
-5.38823724e-01 4.35899049e-01 -2.64079601e-01 -7.00674176e-01
-1.55792630e+00 4.93455052e-01 3.06904465e-01 -5.28620332e-02
-2.34700277e-01 1.08251369e+00 4.18987066e-01 -5.81197083e-01
4.41094935e-01 -2.95729134e-02 -1.94986492e-01 2.38210097e-01
4.08295035e-01 3.45415026e-01 -1.13812961e-01 -8.68750453e-01
-2.89731324e-01 2.75169313e-01 3.51562612e-02 -4.30933803e-01
6.91652179e-01 -3.44742715e-01 -4.10582796e-02 4.58461225e-01
1.01756299e+00 -1.72013298e-01 -1.08736432e+00 1.02312751e-01
-5.26030779e-01 -2.46299312e-01 1.27258614e-01 -1.20593178e+00
-7.65088320e-01 9.87309635e-01 8.27188015e-01 -2.60676652e-01
9.92253780e-01 2.38165945e-01 5.32986879e-01 5.95384538e-01
3.90727699e-01 -1.04341280e+00 6.53335631e-01 5.34227252e-01
1.14592671e+00 -1.87537515e+00 -2.30229497e-01 4.01815884e-02
-1.12824559e+00 6.55082166e-01 1.23490477e+00 6.87711477e-01
-4.51790802e-02 -1.39816225e-01 4.65505660e-01 -4.49015677e-01
-9.69176471e-01 -5.48647523e-01 4.18721408e-01 7.13845432e-01
4.92947936e-01 -2.44394392e-02 7.01465309e-02 5.17159104e-01
-1.41339913e-01 -2.31668293e-01 3.93744767e-01 8.16658854e-01
-2.83035547e-01 -6.90704346e-01 -5.49361229e-01 2.77594656e-01
-8.63345414e-02 -4.46405768e-01 -4.50080544e-01 8.52835536e-01
2.22303033e-01 1.38528621e+00 -5.64034563e-03 -3.46954346e-01
3.94076347e-01 2.82102138e-01 7.06442475e-01 -5.43475211e-01
-7.31583297e-01 -2.07870677e-01 3.37414920e-01 -7.65101731e-01
-2.37479940e-01 -3.57717484e-01 -1.20655727e+00 -2.54524171e-01
-2.83103087e-03 -3.83286357e-01 7.98152566e-01 7.10834622e-01
3.81913990e-01 5.26908159e-01 -2.00542375e-01 -7.86321700e-01
-2.50600964e-01 -7.32753694e-01 5.22193871e-02 5.74318528e-01
8.59235078e-02 -6.38372302e-01 -2.68126488e-01 -2.85244957e-02] | [10.912623405456543, 1.6095521450042725] |
76a906e6-34e4-4fb7-a1d2-ed915eb83186 | learned-3d-shape-representations-using-fused | 1904.04297 | null | http://arxiv.org/abs/1904.04297v1 | http://arxiv.org/pdf/1904.04297v1.pdf | Learned 3D Shape Representations Using Fused Geometrically Augmented Images: Application to Facial Expression and Action Unit Detection | This paper proposes an approach to learn generic multi-modal mesh surface
representations using a novel scheme for fusing texture and geometric data. Our
approach defines an inverse mapping between different geometric descriptors
computed on the mesh surface or its down-sampled version, and the corresponding
2D texture image of the mesh, allowing the construction of fused geometrically
augmented images (FGAI). This new fused modality enables us to learn feature
representations from 3D data in a highly efficient manner by simply employing
standard convolutional neural networks in a transfer-learning mode. In contrast
to existing methods, the proposed approach is both computationally and memory
efficient, preserves intrinsic geometric information and learns highly
discriminative feature representation by effectively fusing shape and texture
information at data level. The efficacy of our approach is demonstrated for the
tasks of facial action unit detection and expression classification. The
extensive experiments conducted on the Bosphorus and BU-4DFE datasets, show
that our method produces a significant boost in the performance when compared
to state-of-the-art solutions | ['Munawar Hayat', 'Naoufel Werghi', 'Bilal Taha', 'Stefano Berretti'] | 2019-04-08 | null | null | null | null | ['action-unit-detection', 'facial-action-unit-detection'] | ['computer-vision', 'computer-vision'] | [ 4.78721052e-01 -4.16409075e-02 -6.85771331e-02 -4.93754268e-01
-9.29921269e-01 -1.84937045e-01 5.60889423e-01 -5.22015914e-02
-2.42299184e-01 3.46218616e-01 -9.45243612e-02 4.92263049e-01
-1.71113864e-01 -9.85046744e-01 -7.06581235e-01 -9.04793262e-01
-1.07786402e-01 3.52147460e-01 -4.98483777e-02 -3.78218174e-01
2.40484893e-01 9.02446449e-01 -2.19967103e+00 5.69250941e-01
3.94557983e-01 1.57916260e+00 -1.42881170e-01 2.89654911e-01
-1.01813443e-01 4.38109010e-01 -1.73962072e-01 -2.65659213e-01
3.41583252e-01 -1.53313339e-01 -7.88008511e-01 4.16423470e-01
1.04131579e+00 -3.31297398e-01 -5.40607348e-02 8.23572874e-01
4.05435950e-01 -3.54366340e-02 7.06037104e-01 -9.12080050e-01
-5.07599235e-01 -3.08836341e-01 -8.20030570e-01 -8.58404040e-02
5.37285864e-01 -4.02827442e-01 7.42413282e-01 -1.08622849e+00
6.87711000e-01 1.49581563e+00 6.50241792e-01 5.07659912e-01
-1.33777201e+00 -6.56595945e-01 -1.83376819e-01 -7.47891515e-02
-1.57827711e+00 -4.88445193e-01 1.07594216e+00 -3.26807410e-01
8.13860536e-01 2.38331914e-01 7.59730279e-01 1.01952815e+00
3.39814723e-01 6.95156813e-01 1.31560946e+00 -4.97736424e-01
4.50360142e-02 -1.77863404e-01 -2.89881438e-01 1.20624804e+00
-2.08827734e-01 8.69720131e-02 -8.96416724e-01 -2.73011088e-01
1.01473355e+00 3.13181914e-02 -7.48041272e-02 -6.95169806e-01
-7.68173158e-01 7.94608593e-01 2.13322550e-01 4.46594983e-01
-4.12226856e-01 2.62840718e-01 3.66045415e-01 3.49318326e-01
1.17156971e+00 -1.25323772e-01 -3.48134011e-01 -6.42332211e-02
-8.55901837e-01 3.85499358e-01 5.72166502e-01 6.29513979e-01
1.18516445e+00 -7.77953193e-02 -4.62161843e-03 8.15104365e-01
3.66221696e-01 4.89910632e-01 3.32046509e-01 -9.13113296e-01
-3.42211090e-02 8.70441735e-01 -2.60420144e-01 -1.15194631e+00
-3.29489052e-01 3.34410034e-02 -6.02031589e-01 6.35033607e-01
9.84204784e-02 2.43408337e-01 -8.84323955e-01 1.78765416e+00
5.38311541e-01 2.57057518e-01 -9.56516191e-02 7.07015157e-01
8.90463710e-01 2.51840860e-01 3.09548597e-03 -1.72350630e-02
1.24166656e+00 -5.82865953e-01 -7.43299961e-01 2.07990885e-01
5.72219908e-01 -7.22625554e-01 7.14593589e-01 4.50010777e-01
-1.20178235e+00 -7.23414540e-01 -9.87576842e-01 -2.55932301e-01
-4.96720314e-01 3.32095355e-01 7.23810136e-01 5.14028549e-01
-1.29974067e+00 8.11702311e-01 -8.35233331e-01 -4.56283033e-01
8.78342032e-01 8.02151263e-01 -8.40555727e-01 1.29053229e-02
-6.69185400e-01 8.06731880e-01 8.06344673e-02 -1.25447810e-01
-5.77807903e-01 -5.93225121e-01 -1.03462040e+00 -1.57430321e-01
-1.00649454e-01 -6.04146540e-01 9.92463231e-01 -1.38363719e+00
-1.77943718e+00 1.38688922e+00 -1.49663612e-01 1.24163061e-01
1.43694058e-01 -2.13535428e-01 -1.62264466e-01 4.91222054e-01
-1.33638352e-01 8.01318109e-01 1.25895452e+00 -1.04915404e+00
-5.65448225e-01 -7.77090132e-01 7.07270131e-02 1.75388336e-01
-4.62719947e-01 -5.33122085e-02 -2.02652171e-01 -6.70760572e-01
2.44004384e-01 -7.31451869e-01 2.38747954e-01 6.39265418e-01
3.18127036e-01 -3.38823050e-01 1.19683671e+00 -4.48888212e-01
7.60699272e-01 -2.33940077e+00 5.68951368e-01 1.91859022e-01
1.34275004e-01 1.66657239e-01 -2.78661638e-01 2.04267412e-01
-2.39486992e-01 -1.54475272e-01 -1.81486353e-01 -5.72874784e-01
-1.15085036e-01 2.91616350e-01 -1.92397833e-02 7.62712777e-01
5.45319736e-01 9.50959742e-01 -7.87364781e-01 -4.65131074e-01
4.38579768e-01 9.51446235e-01 -4.61179942e-01 2.12008268e-01
6.05336484e-03 5.54795325e-01 -6.55905008e-01 7.89074481e-01
7.59389162e-01 8.78261775e-02 -7.82139320e-03 -4.37231809e-01
-1.96503382e-02 -1.76507473e-01 -9.74448979e-01 2.32099438e+00
-5.87671220e-01 3.98905635e-01 2.56398052e-01 -1.28192890e+00
1.12565351e+00 5.24066567e-01 7.46614039e-01 -9.25584793e-01
4.15506959e-01 2.32126936e-01 -7.73769498e-01 -5.13445318e-01
1.61214933e-01 -3.54817390e-01 2.07161903e-03 4.09620613e-01
5.58549881e-01 -1.95585445e-01 -3.25707346e-01 -4.71489370e-01
7.76302218e-01 6.70757353e-01 1.96772158e-01 -5.41196764e-01
8.17773998e-01 -4.95606065e-01 2.09381774e-01 8.22109506e-02
5.62235788e-02 4.50479567e-01 2.45517343e-01 -6.92645431e-01
-6.68290854e-01 -1.06953871e+00 -3.71686965e-01 1.23560512e+00
-3.97259146e-02 -2.67698288e-01 -1.00296330e+00 -5.86681783e-01
2.27369681e-01 8.18811581e-02 -1.06966829e+00 -2.50654429e-01
-4.81746048e-01 -3.72784466e-01 4.40081745e-01 5.20237386e-01
6.50527298e-01 -8.58775020e-01 -7.01363027e-01 -7.44106472e-02
1.21766925e-01 -9.29510474e-01 -2.23714605e-01 5.84881008e-03
-1.06406105e+00 -1.02256739e+00 -5.20541131e-01 -9.37356949e-01
6.13756478e-01 8.75880495e-02 8.97134721e-01 1.34352252e-01
-6.55232191e-01 6.29850328e-01 -3.29900473e-01 -1.60902545e-01
-5.25208302e-02 -3.10249925e-01 -1.34642139e-01 7.11186767e-01
3.49783897e-01 -7.28740871e-01 -4.09589916e-01 -7.04855248e-02
-1.11663139e+00 -5.52851520e-03 4.38082248e-01 8.60318244e-01
8.79354417e-01 -3.20153743e-01 4.85346138e-01 -6.60585046e-01
2.05391496e-01 -1.46052212e-01 -4.69828874e-01 1.35770679e-01
-2.43899792e-01 1.01619497e-01 1.43032342e-01 -2.04108998e-01
-1.04374719e+00 3.91633064e-01 -1.63237035e-01 -7.83474982e-01
-4.25283879e-01 1.71966240e-01 -1.68632954e-01 -7.51807630e-01
3.59879971e-01 9.37962309e-02 4.38414335e-01 -7.02003360e-01
3.07236403e-01 5.93998969e-01 4.22815263e-01 -6.77563906e-01
5.04474044e-01 9.36496556e-01 3.21121544e-01 -8.56813073e-01
-7.79683709e-01 -2.54703492e-01 -1.22396684e+00 -2.83053964e-01
9.37428653e-01 -7.79918134e-01 -6.86148942e-01 7.31860161e-01
-1.05538642e+00 2.61030719e-02 -2.47842163e-01 2.09446430e-01
-9.41708505e-01 2.97824174e-01 -4.29340899e-01 -6.15571260e-01
-3.69088829e-01 -1.14339066e+00 1.92614210e+00 -8.23711902e-02
-1.18885010e-01 -1.15005600e+00 1.42505839e-01 4.03093606e-01
2.85331875e-01 9.22077775e-01 1.02311313e+00 -2.68897384e-01
-3.54418039e-01 -2.97627807e-01 -1.62993729e-01 4.62599963e-01
3.40218693e-01 -3.63845266e-02 -1.35726678e+00 -3.18342090e-01
-4.54171561e-02 -6.88731432e-01 7.25903511e-01 8.70732367e-02
1.36017203e+00 1.55688494e-01 -2.39843145e-01 7.71085382e-01
1.46033978e+00 -1.02893323e-01 4.05915946e-01 -7.94526264e-02
5.47632754e-01 5.85681915e-01 5.65326989e-01 6.12636447e-01
8.53549242e-02 8.83549273e-01 4.94160026e-01 -3.85109246e-01
-2.14732096e-01 -3.20425145e-02 2.29928434e-01 5.68767607e-01
-4.46815223e-01 1.90631166e-01 -5.05701303e-01 2.74103135e-01
-1.75718021e+00 -7.42456496e-01 2.76914746e-01 2.13195133e+00
6.00274682e-01 -4.96721715e-01 -9.70210955e-02 2.56107599e-01
4.08405393e-01 1.43037990e-01 -2.24168077e-01 -6.70665741e-01
-6.78026527e-02 1.15169299e+00 4.41376455e-02 3.04309100e-01
-1.34949875e+00 9.48000431e-01 6.74349737e+00 9.77992475e-01
-1.31743181e+00 2.84383506e-01 4.05372649e-01 -1.21216394e-01
2.31177676e-02 -5.32292128e-01 -2.48528048e-01 4.03939709e-02
8.31119657e-01 1.39775202e-01 3.97218198e-01 6.83488071e-01
-1.40861481e-01 -5.87577820e-02 -1.28839016e+00 1.33780026e+00
4.93665487e-01 -1.38136959e+00 2.39215836e-01 1.93493366e-01
5.64098358e-01 -1.60341889e-01 2.45064601e-01 -3.84420156e-02
-2.08123237e-01 -1.12741995e+00 7.24115074e-01 8.03983927e-01
1.09385502e+00 -9.68588114e-01 5.55724740e-01 -7.88208321e-02
-1.33505952e+00 1.52630836e-01 -1.17196754e-01 -2.18302265e-01
-3.86788040e-01 1.44865647e-01 -3.05874795e-01 7.53935516e-01
7.94066489e-01 9.14041519e-01 -4.63752031e-01 3.71374160e-01
9.48888063e-02 9.93923917e-02 -2.65645921e-01 3.57970327e-01
1.38792530e-01 -2.32865527e-01 2.51817137e-01 1.07312882e+00
2.55852073e-01 1.57380421e-02 1.37644440e-01 8.69754136e-01
-8.12624469e-02 3.78253192e-01 -8.44888449e-01 1.97181455e-03
-1.79003999e-01 1.47645462e+00 -5.87190568e-01 -1.63145170e-01
-6.48719907e-01 1.13910699e+00 5.69254160e-01 1.16237231e-01
-5.81056893e-01 -1.10910878e-01 7.61552811e-01 -5.77206463e-02
5.20395875e-01 -1.23635672e-01 7.33153545e-04 -9.02497530e-01
-9.77352727e-03 -5.75372875e-01 2.53203154e-01 -6.57774806e-01
-1.23401594e+00 7.01626241e-01 6.55748323e-02 -1.13960564e+00
-1.84205666e-01 -9.68608975e-01 -3.09126347e-01 8.32360506e-01
-1.54168820e+00 -1.72165573e+00 -4.70347703e-01 8.93365264e-01
2.18350172e-01 -1.70245379e-01 1.40642679e+00 3.41117531e-01
-1.96560621e-01 6.51128471e-01 6.35263622e-02 -7.55979568e-02
4.86667544e-01 -8.94229472e-01 -6.93713129e-02 1.14604309e-01
2.49272749e-01 8.84215906e-02 1.34056211e-01 -1.82276383e-01
-1.65721369e+00 -1.00247395e+00 5.71352541e-01 -3.58812422e-01
2.52081126e-01 -4.78973478e-01 -8.27562094e-01 4.80508596e-01
1.53593302e-01 3.84335876e-01 9.00597513e-01 -4.04273830e-02
-4.64941323e-01 -2.53873467e-01 -1.38958645e+00 1.38190880e-01
1.06090331e+00 -7.41764367e-01 -6.38643444e-01 2.91215539e-01
1.45472810e-01 -3.83583188e-01 -1.24949038e+00 7.12275743e-01
8.19606245e-01 -1.12748206e+00 7.37180710e-01 -4.70519334e-01
3.09102595e-01 -4.16190252e-02 -4.50761318e-01 -1.07850695e+00
-2.07835943e-01 -4.83555645e-01 -1.08141690e-01 8.72915983e-01
-1.75238177e-01 -4.21804875e-01 6.90623760e-01 9.43861529e-02
-8.08676183e-02 -1.18639266e+00 -1.33323061e+00 -3.83070618e-01
5.18874973e-02 -2.70700306e-01 5.51448643e-01 8.89265001e-01
-1.97215036e-01 -2.03619562e-02 -7.46807829e-02 -6.97944686e-02
4.87601608e-01 4.25861746e-01 5.41486084e-01 -1.45466983e+00
1.17760003e-01 -2.78932780e-01 -9.70566034e-01 -5.91802180e-01
7.37207949e-01 -1.12389600e+00 -2.94720829e-01 -1.10413277e+00
1.42712981e-01 -1.60349771e-01 -3.82509291e-01 7.91076660e-01
4.51896697e-01 8.34526300e-01 -7.39110634e-02 -1.52214944e-01
-3.45065564e-01 7.22857833e-01 1.32420850e+00 1.15993796e-02
1.35146782e-01 -4.79176879e-01 -3.54962170e-01 8.76932025e-01
4.23928469e-01 -7.79137909e-02 -2.80121863e-01 -5.49063563e-01
-2.51547337e-01 -9.91109684e-02 6.52004898e-01 -9.71531332e-01
-1.96555153e-01 8.50030109e-02 5.47653198e-01 -5.09460747e-01
8.12601626e-01 -8.93930972e-01 9.60645154e-02 7.30739087e-02
-6.23020679e-02 -1.92414150e-01 5.26426971e-01 4.26592082e-01
-3.38584632e-01 1.12185903e-01 1.02505338e+00 -1.30384294e-02
-7.06562638e-01 5.30218363e-01 -1.85658380e-01 -3.90735716e-01
1.18259466e+00 -5.54037511e-01 2.27957517e-01 1.35832969e-02
-7.76854813e-01 -3.66344243e-01 6.51494980e-01 5.33037245e-01
7.83616364e-01 -1.63135159e+00 -5.45067966e-01 8.81269634e-01
2.02067256e-01 -2.33955815e-01 2.25233257e-01 8.24308634e-01
-4.57326353e-01 2.53913790e-01 -7.26873994e-01 -7.78373361e-01
-1.42401135e+00 2.78831661e-01 4.62142199e-01 4.35538888e-02
-5.41966319e-01 8.59924495e-01 4.17994112e-01 -3.79413784e-01
9.25140977e-02 -3.32011044e-01 -1.72205970e-01 1.56840593e-01
4.33635026e-01 1.87563121e-01 4.11810875e-01 -1.15900373e+00
-2.86634147e-01 1.16797495e+00 1.36271343e-01 9.11024436e-02
1.45712531e+00 -5.21622878e-03 -4.36917156e-01 5.07039547e-01
1.69793487e+00 -2.62275159e-01 -1.05332077e+00 -2.73639590e-01
-2.00219855e-01 -6.47467792e-01 4.97655779e-01 -4.87299055e-01
-1.38760424e+00 9.72484589e-01 1.01066267e+00 -3.82338792e-01
1.47769415e+00 2.05033436e-01 6.21347725e-01 1.27879307e-01
6.24622881e-01 -9.49572325e-01 2.77225524e-01 2.16896623e-01
1.14001477e+00 -9.96802807e-01 -4.47248481e-02 -5.57074547e-01
-1.13135405e-01 1.37613213e+00 4.52950835e-01 -4.64238882e-01
8.85815263e-01 2.40230113e-01 1.72967035e-02 -5.20741940e-01
-5.74004650e-01 -2.25025535e-01 5.71162760e-01 4.61370021e-01
6.05801165e-01 -1.64231554e-01 -7.22540244e-02 2.15472952e-01
1.80021122e-01 9.87562165e-02 -2.05398709e-01 1.28037930e+00
-2.63804883e-01 -1.26248693e+00 -1.72186151e-01 3.34197313e-01
-5.34045160e-01 2.56648391e-01 -5.12760699e-01 9.25146461e-01
4.52438116e-01 4.73572105e-01 4.63808477e-01 -4.44547683e-01
3.41670454e-01 3.44455749e-01 1.10621262e+00 -6.57817066e-01
-4.44501430e-01 2.34448969e-01 -1.52048588e-01 -1.11112523e+00
-1.02691531e+00 -6.23045564e-01 -1.13795328e+00 -3.22090052e-02
-1.88453764e-01 -3.11921626e-01 5.92541099e-01 1.05034006e+00
6.10881925e-01 2.87620425e-01 7.71653771e-01 -1.37819970e+00
-2.46058539e-01 -7.83117533e-01 -8.67260575e-01 6.82550490e-01
4.07392263e-01 -1.42057192e+00 -6.69317394e-02 1.66166767e-01] | [13.494872093200684, 1.2702604532241821] |
b505571c-99fb-4a45-b282-189a643f837d | on-solutions-of-the-distributional-bellman | 2202.00081 | null | https://arxiv.org/abs/2202.00081v3 | https://arxiv.org/pdf/2202.00081v3.pdf | On solutions of the distributional Bellman equation | In distributional reinforcement learning not only expected returns but the complete return distributions of a policy are taken into account. The return distribution for a fixed policy is given as the solution of an associated distributional Bellman equation. In this note we consider general distributional Bellman equations and study existence and uniqueness of their solutions as well as tail properties of return distributions. We give necessary and sufficient conditions for existence and uniqueness of return distributions and identify cases of regular variation. We link distributional Bellman equations to multivariate affine distributional equations. We show that any solution of a distributional Bellman equation can be obtained as the vector of marginal laws of a solution to a multivariate affine distributional equation. This makes the general theory of such equations applicable to the distributional reinforcement learning setting. | ['Denis Spiegel', 'Ralph Neininger', 'Julian Gerstenberg'] | 2022-01-31 | null | null | null | null | ['distributional-reinforcement-learning'] | ['methodology'] | [-4.96371299e-01 -7.88726099e-03 -3.90634060e-01 -3.75193805e-01
-5.51543415e-01 -8.68309855e-01 2.35776380e-01 8.04535896e-02
-8.69742215e-01 1.15737414e+00 1.80926675e-03 -6.65241361e-01
-7.66764998e-01 -8.01590621e-01 -7.80605912e-01 -1.13756025e+00
-1.95452631e-01 5.90493917e-01 -3.71484965e-01 -2.31171474e-01
1.27263784e-01 6.77268863e-01 -1.16847765e+00 -4.78786379e-01
7.54126251e-01 8.51871610e-01 1.05497502e-01 1.17476308e+00
-2.36125171e-01 6.98088050e-01 -8.81588161e-01 -1.46718025e-01
4.48575616e-01 -5.10781586e-01 -4.64392751e-01 -1.30767509e-01
3.49972665e-01 -5.80679536e-01 -1.39665619e-01 1.47805929e+00
5.53028047e-01 6.16269827e-01 1.13609624e+00 -1.66066480e+00
-1.00476694e+00 4.23000187e-01 -7.51277506e-01 6.04140818e-01
-9.17920321e-02 -9.44786333e-03 1.15823078e+00 -8.89213160e-02
2.65225708e-01 1.25380290e+00 4.68794227e-01 7.11096406e-01
-1.53624523e+00 -2.67458051e-01 1.21882312e-01 -2.69534975e-01
-7.46119261e-01 -5.25554530e-02 1.96691841e-01 -5.29919505e-01
7.81556189e-01 -2.47454736e-02 4.37760830e-01 7.62710869e-01
7.07377791e-01 7.49120653e-01 1.16899788e+00 -2.53810018e-01
7.17656910e-01 5.97823113e-02 2.70119697e-01 5.78702211e-01
4.54805523e-01 5.51627278e-01 1.92670390e-01 -3.27600926e-01
1.02243054e+00 9.73967984e-02 4.79484946e-02 -5.97078085e-01
-3.74247164e-01 1.27997506e+00 8.38883221e-02 6.21938705e-03
-7.93859482e-01 5.73171377e-01 1.26989409e-01 5.64194560e-01
1.95700973e-01 2.05550075e-01 -3.87698591e-01 -1.81903109e-01
-3.15945297e-01 9.35405314e-01 1.05591047e+00 1.10736382e+00
6.63228691e-01 3.79020542e-01 -4.12058532e-01 3.85577381e-01
3.62496495e-01 1.39995575e+00 1.97513014e-01 -1.20863450e+00
3.12858850e-01 -1.67916343e-01 5.96425772e-01 -3.77837896e-01
-4.16576803e-01 -4.07653123e-01 -2.44798452e-01 4.39092636e-01
9.49836314e-01 -7.36938179e-01 -4.63575840e-01 2.16942906e+00
2.80705601e-01 -1.07568704e-01 3.31331819e-01 5.48547685e-01
-7.36485794e-02 6.74637794e-01 1.61488838e-02 -6.40461922e-01
7.39353240e-01 5.04572839e-02 -7.32033610e-01 3.15619141e-01
4.18139547e-01 -5.87632358e-01 9.98843193e-01 6.60991147e-02
-1.10176969e+00 1.02290818e-02 -5.08751094e-01 5.17085969e-01
1.07843801e-01 -3.49215388e-01 1.05764814e-01 6.41622901e-01
-9.76410091e-01 6.96949542e-01 -5.82746685e-01 -2.77599897e-02
2.16773212e-01 3.55038822e-01 1.42179564e-01 2.68383026e-01
-8.75055254e-01 9.13029134e-01 -1.11196533e-01 -1.45003274e-01
-1.30862141e+00 -9.23232436e-01 -6.40495300e-01 1.11562237e-01
4.58483875e-01 -3.96249473e-01 2.03552866e+00 -6.97377980e-01
-1.48055995e+00 1.69622123e-01 1.14848100e-01 -5.49497545e-01
5.08689702e-01 -3.01611163e-02 6.87406212e-02 1.60240475e-02
1.85478598e-01 -1.07246131e-01 7.86987960e-01 -1.01119411e+00
-7.96346605e-01 -3.49268019e-01 4.16052997e-01 1.33221805e-01
-7.14989603e-02 -2.26169258e-01 8.48127365e-01 -3.63962382e-01
-7.92436600e-01 -9.19996500e-01 -2.75292367e-01 -4.61015224e-01
-4.01229598e-02 -5.16188741e-01 4.88729984e-01 -2.70545393e-01
8.16993117e-01 -2.04278159e+00 1.05167143e-01 2.52409697e-01
-2.89293408e-01 -2.77362615e-01 -4.04641449e-01 4.22424942e-01
-9.06846859e-03 -1.02843784e-01 -1.99940100e-01 5.41087985e-02
6.55306220e-01 7.35449791e-01 -7.29973972e-01 1.05293190e+00
2.13566492e-03 5.49824059e-01 -9.52506185e-01 -1.30061001e-01
-1.01530544e-01 8.90317708e-02 -9.98711646e-01 4.55568790e-01
-5.14219642e-01 -7.20514823e-03 -7.41560936e-01 4.56804708e-02
8.46287727e-01 5.57102025e-01 -1.67064324e-01 6.65846348e-01
-2.71394938e-01 -3.47032547e-02 -1.06356466e+00 7.34289527e-01
-6.25098169e-01 1.25722855e-01 2.15703875e-01 -1.27571666e+00
8.57010007e-01 1.35351792e-01 6.55864000e-01 -5.49783945e-01
4.03344750e-01 1.58908010e-01 1.57870486e-01 -6.56777680e-01
4.95791614e-01 -1.19287848e+00 -2.34230459e-01 9.03863728e-01
3.29347730e-01 -1.34840310e-01 1.81412876e-01 1.64988339e-01
6.68463469e-01 1.09252170e-01 1.40817580e-03 -9.53160048e-01
1.18921861e-01 -3.54368210e-01 4.08973604e-01 8.69467020e-01
-3.47847968e-01 -1.93950653e-01 1.13031125e+00 -6.93517104e-02
-1.05348325e+00 -1.57062650e+00 -1.73089430e-01 1.17424893e+00
-2.38061860e-01 3.51840369e-02 -7.73711503e-01 -5.86444139e-01
6.15951955e-01 1.01857197e+00 -8.13818872e-01 -3.21768939e-01
-1.60599470e-01 -7.09824145e-01 3.80768299e-01 5.28484046e-01
1.15383163e-01 -1.02893615e+00 -9.66121078e-01 5.40303737e-02
2.27369353e-01 -4.62638021e-01 -7.26474941e-01 4.43073064e-01
-8.52233648e-01 -1.08682597e+00 -8.91685069e-01 -2.36990392e-01
4.39476281e-01 -2.07791820e-01 6.94047451e-01 -7.20266819e-01
-5.34860268e-02 1.20854449e+00 3.39036465e-01 -8.60322714e-01
-4.73276198e-01 -3.78175706e-01 4.76245046e-01 8.96169320e-02
3.20566565e-01 -3.98976415e-01 -4.71204102e-01 1.37315094e-01
-1.01298511e+00 -1.24641049e+00 1.44793615e-01 5.34752309e-01
6.81730926e-01 2.47829691e-01 9.11625922e-01 -5.61992764e-01
1.34846330e+00 -6.18510604e-01 -1.29100716e+00 1.13906138e-01
-5.69076300e-01 6.01477802e-01 7.50715435e-01 -5.94568312e-01
-1.33928168e+00 -1.81479067e-01 1.98619604e-01 -3.51863384e-01
-3.38855013e-02 4.84698683e-01 3.81591097e-02 4.26009715e-01
6.41430497e-01 1.66562781e-01 3.44316542e-01 -4.12151098e-01
5.35367370e-01 4.48203623e-01 5.41915238e-01 -1.28203058e+00
6.01189196e-01 2.01447278e-01 4.14127260e-01 -1.05601597e+00
-8.91426802e-01 -3.24074864e-01 -4.04870778e-01 -3.04846406e-01
7.35861719e-01 -2.16821626e-01 -1.22915196e+00 9.76375490e-02
-8.92301202e-01 -8.03046167e-01 -1.02475643e+00 6.28156841e-01
-1.41395366e+00 -1.12897065e-02 -4.16622728e-01 -1.57875276e+00
2.99971879e-01 -7.85587907e-01 5.06682277e-01 3.50413650e-01
1.52728796e-01 -1.44345927e+00 7.35214829e-01 -4.18956906e-01
5.04943967e-01 -9.05076489e-02 1.01728284e+00 -3.85724992e-01
-1.74069908e-02 -4.87133227e-02 2.12104380e-01 6.88998759e-01
-5.68502545e-02 2.06042342e-02 -4.10724074e-01 -3.54694903e-01
5.22972524e-01 -4.29811999e-02 6.89924836e-01 1.08887696e+00
4.09896225e-01 -7.24784672e-01 3.51152629e-01 3.30044657e-01
1.52847660e+00 3.85744542e-01 1.37791470e-01 2.79832453e-01
2.42493168e-01 6.71551168e-01 5.38847029e-01 7.99195588e-01
2.55989134e-01 -1.00747511e-01 4.77543920e-01 6.41344965e-01
7.72669613e-01 -3.72997493e-01 8.55611682e-01 2.01167479e-01
-3.27590592e-02 1.93788946e-01 -6.00217938e-01 5.47134936e-01
-1.92679310e+00 -1.49456155e+00 -6.24474585e-02 2.52249599e+00
7.81787455e-01 -2.95780957e-01 8.62757266e-01 -3.76135111e-01
7.33428836e-01 -3.81381482e-01 -9.23291206e-01 -1.05534101e+00
-4.51734243e-03 4.75339830e-01 1.06128311e+00 9.91169095e-01
-7.25363910e-01 3.39538127e-01 7.75755358e+00 4.64501560e-01
-7.48851895e-01 4.40904014e-02 1.45988747e-01 -2.86060542e-01
-5.65064490e-01 -1.15753792e-01 -8.15996647e-01 2.34019667e-01
1.11082482e+00 -7.48063564e-01 5.39782047e-01 9.96458411e-01
1.85594708e-01 -3.23608756e-01 -1.01045108e+00 6.48289680e-01
-3.84172380e-01 -7.78071642e-01 -3.57526690e-01 7.08458424e-01
1.06125748e+00 -4.86548990e-02 5.48506200e-01 4.52232271e-01
1.03050315e+00 -9.82928097e-01 6.27049863e-01 6.40816689e-01
3.09230804e-01 -1.39297974e+00 6.14479959e-01 5.19121706e-01
-6.31054401e-01 -3.40553582e-01 -9.23528850e-01 -3.95903140e-01
-1.40941247e-01 2.31744170e-01 -3.86498421e-01 1.08786866e-01
2.29046226e-01 3.68400455e-01 2.52090871e-01 1.00712955e+00
-4.89163473e-02 5.17498374e-01 -4.02126402e-01 -2.91135609e-01
5.83411992e-01 -6.56483233e-01 3.58204812e-01 1.05634344e+00
5.14532208e-01 1.28764957e-01 1.02836095e-01 1.03596330e+00
6.32138848e-02 2.63997972e-01 -7.26904154e-01 -1.38627291e-01
5.78886494e-02 9.18789625e-01 -3.41932952e-01 -1.85544297e-01
-2.41036177e-01 3.39663267e-01 3.28481793e-01 6.22131944e-01
-7.94996202e-01 -3.02500904e-01 1.22738922e+00 -8.94331932e-02
4.51965749e-01 -1.23273797e-01 8.41796175e-02 -9.70829308e-01
-2.63879687e-01 -3.21202368e-01 8.32086682e-01 -1.92754045e-01
-1.51018572e+00 1.24094009e-01 4.44868356e-01 -6.39683247e-01
-5.88726938e-01 -7.71616340e-01 -8.38025451e-01 9.96689975e-01
-1.33409798e+00 9.53075383e-03 6.25924826e-01 1.13650453e+00
-9.93926972e-02 -3.93414259e-01 6.52806580e-01 -1.63669765e-01
-4.25906569e-01 2.95829266e-01 6.92544460e-01 -2.64550745e-01
2.87590712e-01 -1.82675600e+00 -4.99043226e-01 4.87046897e-01
-1.14973277e-01 5.05316019e-01 1.15403008e+00 -4.60055321e-01
-1.54890072e+00 -9.05949950e-01 3.27881336e-01 -1.68678448e-01
1.24026656e+00 2.54666120e-01 -5.81784427e-01 9.75748539e-01
3.37118149e-01 -1.60808846e-01 5.94316244e-01 2.70002298e-02
-1.83351457e-01 -1.47847841e-02 -1.26535666e+00 3.06957930e-01
3.37194294e-01 -5.13960719e-01 -4.69070733e-01 2.25622177e-01
3.29385281e-01 -6.00905791e-02 -7.08252072e-01 -2.65991986e-01
4.23274517e-01 -7.18716085e-01 5.78857958e-01 -1.20466495e+00
2.26934955e-01 2.73011671e-03 -4.89721239e-01 -1.89272082e+00
-1.95625156e-01 -9.82718587e-01 1.20255820e-01 9.41778779e-01
8.18779022e-02 -9.45593476e-01 3.94955963e-01 6.19578719e-01
1.49359271e-01 -6.57691121e-01 -1.16684616e+00 -1.41437125e+00
1.07133126e+00 -2.86533743e-01 4.61110413e-01 5.78903735e-01
2.87771881e-01 1.28986210e-01 -2.03208938e-01 2.72875000e-02
1.06132865e+00 2.70864218e-01 3.80714625e-01 -8.43191206e-01
-5.07545173e-01 -7.89407909e-01 -3.83329205e-02 -8.89779687e-01
6.16675079e-01 -1.05865669e+00 3.36690784e-01 -1.33509803e+00
1.96222544e-01 -5.10059416e-01 -5.84718108e-01 -7.16544837e-02
2.03592718e-01 -5.37414432e-01 3.92273009e-01 -2.72661567e-01
-2.53382683e-01 7.24341035e-01 1.25065279e+00 1.54994637e-01
-1.76401034e-01 5.27767003e-01 -7.30633914e-01 6.06421173e-01
1.28007698e+00 -6.94576919e-01 -8.21011841e-01 -2.15614781e-01
2.31915712e-01 3.63614142e-01 1.55593276e-01 -2.40131140e-01
2.96575483e-03 -8.12713802e-01 -2.86731850e-02 -1.86030984e-01
-1.94177315e-01 -5.60474873e-01 -3.18672836e-01 7.83850074e-01
-6.67649031e-01 4.68658447e-01 1.91883221e-01 7.08340764e-01
1.83712780e-01 -8.43587399e-01 1.01448226e+00 9.49309170e-02
1.97857432e-02 4.21251506e-01 -7.87024498e-01 5.44781744e-01
1.15133607e+00 4.18581367e-01 -2.02412248e-01 -8.28927815e-01
-7.90399969e-01 3.89538437e-01 1.33307174e-01 -1.88692525e-01
5.25124013e-01 -1.33242095e+00 -6.38606727e-01 4.42995839e-02
-4.43732142e-01 -2.43847311e-01 1.18216567e-01 8.69238794e-01
-1.05366915e-01 2.56809592e-01 -5.02080023e-01 -2.28827238e-01
-8.29595745e-01 8.83163035e-01 8.53861332e-01 7.56397769e-02
-2.12765843e-01 5.65399885e-01 4.29006517e-01 -4.23232734e-01
3.07122946e-01 -5.78125000e-01 1.69191405e-01 1.62830055e-01
6.02939785e-01 4.49655235e-01 -6.34898782e-01 -5.66810071e-01
-1.55935049e-01 4.04014796e-01 2.39800423e-01 -6.84009910e-01
1.49137616e+00 2.97094113e-03 -7.59882033e-02 5.38780451e-01
1.26071060e+00 -3.65917943e-02 -1.56700635e+00 -9.62906331e-02
-1.12205096e-01 -3.84921879e-01 -1.57921538e-01 -5.92653036e-01
-1.04785752e+00 9.33886886e-01 3.10792565e-01 6.05924428e-01
6.53627038e-01 1.29226267e-01 4.07129347e-01 3.15022796e-01
9.97930318e-02 -1.12067473e+00 -2.05861550e-04 6.44825578e-01
8.52375329e-01 -6.51235878e-01 -3.96506965e-01 4.38031793e-01
-7.00314760e-01 1.20105052e+00 3.04762512e-01 -8.99366438e-01
9.68451738e-01 7.62005866e-01 -1.20190077e-01 1.23957671e-01
-8.25749159e-01 -4.22419548e-01 -5.06825820e-02 9.12510633e-01
2.35901400e-01 4.22236145e-01 -3.01381201e-01 5.59134126e-01
-4.12436008e-01 -7.90119171e-02 9.30264473e-01 9.18953359e-01
-6.97343171e-01 -1.07675517e+00 -4.16450143e-01 3.17731172e-01
-5.95178246e-01 3.14311504e-01 -6.41884133e-02 5.37677586e-01
-4.12803382e-01 7.26264119e-01 3.84278595e-01 1.98218688e-01
5.80055416e-01 4.96820882e-02 1.11419964e+00 -4.45895016e-01
-3.52267660e-02 2.52375126e-01 -5.64572155e-01 -1.69619843e-01
-1.18456766e-01 -1.09732878e+00 -1.38099670e+00 -7.79682338e-01
-1.26188904e-01 4.84212220e-01 5.16239762e-01 8.27408791e-01
-3.31250876e-01 4.58113015e-01 6.17929697e-01 -3.17118198e-01
-1.92445576e+00 -7.79548943e-01 -1.17057657e+00 2.19236165e-01
7.68150985e-01 -5.51710367e-01 -4.71165746e-01 -1.95532173e-01] | [4.137731075286865, 2.590224266052246] |
9a444c05-f6a4-4c44-ab1c-dad7ff3206fc | visual-analogy-deep-learning-versus | 2105.07065 | null | https://arxiv.org/abs/2105.07065v1 | https://arxiv.org/pdf/2105.07065v1.pdf | Visual analogy: Deep learning versus compositional models | Is analogical reasoning a task that must be learned to solve from scratch by applying deep learning models to massive numbers of reasoning problems? Or are analogies solved by computing similarities between structured representations of analogs? We address this question by comparing human performance on visual analogies created using images of familiar three-dimensional objects (cars and their subregions) with the performance of alternative computational models. Human reasoners achieved above-chance accuracy for all problem types, but made more errors in several conditions (e.g., when relevant subregions were occluded). We compared human performance to that of two recent deep learning models (Siamese Network and Relation Network) directly trained to solve these analogy problems, as well as to that of a compositional model that assesses relational similarity between part-based representations. The compositional model based on part representations, but not the deep learning models, generated qualitative performance similar to that of human reasoners. | ['Hongjing Lu', 'Alan Yuille', 'Keith J. Holyoak', 'Shuhao Fu', 'Qing Liu', 'Nicholas Ichien'] | 2021-05-14 | null | null | null | null | ['visual-analogies'] | ['computer-vision'] | [-7.38703385e-02 1.97150201e-01 2.19395697e-01 -3.37292969e-01
-9.84211266e-02 -7.56773889e-01 8.69248033e-01 4.55884904e-01
-5.23483515e-01 6.72581732e-01 4.02455181e-01 -5.54540694e-01
-5.70693791e-01 -9.20839906e-01 -7.10864246e-01 -7.53563941e-02
2.69593537e-01 1.15707552e+00 1.39812917e-01 -6.07889771e-01
7.70282328e-01 8.71029437e-01 -1.29369307e+00 7.47128606e-01
9.55787599e-01 8.24317217e-01 -2.32998356e-02 3.34221870e-01
-3.48223597e-01 1.25048435e+00 -3.59666824e-01 -6.85924470e-01
3.64856958e-01 -6.03007615e-01 -1.19387782e+00 -5.62323868e-01
1.15769303e+00 -4.68703806e-01 -6.51375711e-01 9.64068949e-01
2.97130663e-02 3.68139803e-01 9.31397736e-01 -1.15724862e+00
-1.63958967e+00 5.06352842e-01 -3.65332782e-01 7.24041224e-01
8.87633920e-01 2.84209967e-01 1.05560052e+00 -7.54891038e-01
4.79305297e-01 1.54562366e+00 1.03230703e+00 1.95724130e-01
-1.72483277e+00 -7.04398572e-01 -7.12178871e-02 7.16030777e-01
-1.02428985e+00 7.82732591e-02 4.57725406e-01 -3.94788265e-01
1.34671021e+00 5.99618256e-02 1.00994539e+00 9.35410440e-01
4.48793024e-01 3.44128639e-01 1.49685848e+00 -3.01399738e-01
2.88302034e-01 -1.34534210e-01 3.04516107e-01 5.62358856e-01
4.80591685e-01 5.60031056e-01 -2.36256465e-01 -2.13917643e-01
1.26767671e+00 6.50321320e-02 -7.11696371e-02 -3.86147261e-01
-1.41525996e+00 7.12489128e-01 1.10360515e+00 6.52473450e-01
-5.36075652e-01 1.66633055e-01 5.46348058e-02 5.20287991e-01
-3.78261268e-01 1.57609653e+00 -2.65509278e-01 5.99555373e-01
-7.23824143e-01 8.41795862e-01 7.93611109e-01 6.70367301e-01
6.59720004e-01 1.18770435e-01 -3.64916325e-01 4.70973343e-01
9.43127573e-02 2.63193190e-01 7.67125309e-01 -1.43212795e+00
5.03213644e-01 7.02572942e-01 3.88126597e-02 -1.31568909e+00
-7.16994822e-01 -2.58649021e-01 -5.28308749e-01 5.23264885e-01
9.60660875e-01 3.54557484e-01 -6.98339045e-01 1.68878901e+00
-1.48364738e-01 -2.91922867e-01 1.06802471e-01 1.14053237e+00
8.26891363e-01 4.78061199e-01 3.97153199e-01 2.53824174e-01
1.40468323e+00 -8.26982975e-01 -5.26175976e-01 -6.30541265e-01
4.67131078e-01 -6.11408830e-01 1.32718194e+00 1.40546978e-01
-1.39796162e+00 -1.04612863e+00 -1.21049964e+00 -6.47327125e-01
-6.70829713e-01 -6.27303660e-01 8.45698357e-01 2.70515680e-01
-1.32443762e+00 1.19699872e+00 -1.44408494e-01 -5.17977178e-01
5.63861072e-01 2.76846886e-01 -4.83237445e-01 -1.19354188e-01
-1.35265338e+00 1.87506092e+00 4.45758969e-01 -3.95039648e-01
-5.15902936e-01 -8.00170183e-01 -8.19277883e-01 3.63478541e-01
-1.73390821e-01 -1.08114505e+00 1.31633282e+00 -1.15440536e+00
-1.07796884e+00 1.16649783e+00 7.15514123e-02 -4.52034444e-01
2.72915870e-01 -7.96435922e-02 -3.10770154e-01 2.41645649e-01
2.78183192e-01 7.48557925e-01 3.58904630e-01 -1.18364978e+00
-4.17817347e-02 -4.92922485e-01 3.47073436e-01 1.67864114e-01
1.28269181e-01 -2.50093073e-01 6.41906381e-01 -8.94265234e-01
4.37088788e-01 -9.59919751e-01 -1.09561577e-01 6.39281392e-01
2.15631977e-01 -4.23905730e-01 2.03822494e-01 -8.89127672e-01
2.44994089e-01 -1.79514802e+00 1.47534326e-01 1.66049138e-01
3.06400388e-01 2.24001825e-01 -5.11352181e-01 2.84609079e-01
-6.90345585e-01 6.87222183e-02 -1.54795814e-02 5.00577152e-01
4.21741635e-01 2.28762358e-01 -5.19815624e-01 2.02098414e-01
2.78585386e-02 1.44189227e+00 -1.17765343e+00 -4.07799959e-01
1.09031513e-01 -1.16475038e-01 -7.31543541e-01 1.01046674e-01
-1.20773308e-01 -4.22256067e-04 2.10073486e-01 1.82439074e-01
5.47694445e-01 -1.98533729e-01 1.56603649e-01 -2.87898242e-01
2.57805735e-01 4.33757633e-01 -6.06975436e-01 1.70823431e+00
-4.17174041e-01 7.87679613e-01 -7.07843840e-01 -9.96724308e-01
8.74873817e-01 4.10261273e-01 -2.87621170e-01 -1.29901505e+00
1.33006334e-01 2.84458637e-01 7.09421873e-01 -3.88884664e-01
3.95231992e-01 -7.40749896e-01 1.17496647e-01 7.76400208e-01
7.43629634e-02 -8.54194999e-01 -3.75211872e-02 1.66319743e-01
9.76396143e-01 2.03459293e-01 5.69859922e-01 -4.27385926e-01
1.50672957e-01 3.65213186e-01 4.64214571e-02 9.72958803e-01
-1.95699200e-01 5.24965584e-01 4.87466604e-01 -1.01134348e+00
-1.31277013e+00 -1.79520094e+00 -1.39711713e-02 8.52530062e-01
1.91145897e-01 9.10642743e-02 -3.89709592e-01 -1.75541684e-01
2.64402002e-01 1.48941648e+00 -9.83013690e-01 -6.11789286e-01
-4.39489841e-01 -1.90303817e-01 4.30125833e-01 1.06320310e+00
5.36889970e-01 -1.41130507e+00 -6.33546054e-01 -1.14574265e-02
1.35652259e-01 -9.98594761e-01 8.22727755e-02 3.24025787e-02
-1.00478625e+00 -1.04482841e+00 -6.45268261e-01 -8.94123197e-01
4.51605350e-01 8.23951662e-02 1.52858377e+00 2.74690419e-01
-2.58706033e-01 3.42752695e-01 4.57674786e-02 -2.52503961e-01
-4.69527841e-01 -4.47995335e-01 1.29441142e-01 -8.58660877e-01
5.06176412e-01 -9.61506546e-01 -4.19191360e-01 4.30745393e-01
-6.55294240e-01 4.83906306e-02 8.06599736e-01 7.65518665e-01
2.12296054e-01 -4.23033595e-01 6.13027215e-01 -3.56510639e-01
9.98535872e-01 -4.21991378e-01 -4.18758571e-01 3.17448318e-01
-3.46074700e-01 3.32056195e-01 8.25689137e-01 -6.07213020e-01
-1.13532162e+00 -2.52321780e-01 2.31277585e-01 -4.37956363e-01
-1.50495946e-01 2.53680676e-01 2.47242391e-01 -3.55481394e-02
1.13836753e+00 -1.77713707e-02 -8.01287815e-02 -1.36210501e-01
6.07890248e-01 -4.00571451e-02 8.27961445e-01 -1.09281170e+00
8.66234004e-01 1.99759483e-01 2.34004170e-01 -3.31058741e-01
-9.35638428e-01 2.33196199e-01 -6.97524965e-01 1.61380753e-01
1.18504441e+00 -6.80154026e-01 -8.50953400e-01 -4.53324690e-02
-1.56658518e+00 -3.68557990e-01 -6.25284255e-01 7.68399358e-01
-8.38869274e-01 1.29120886e-01 -7.63117373e-01 5.47156716e-03
1.72773764e-01 -7.78372824e-01 3.64718825e-01 1.25685081e-01
-1.18989718e+00 -1.02265358e+00 2.44962618e-01 4.12968576e-01
5.57655931e-01 1.66374832e-01 1.60602367e+00 -1.03030598e+00
-3.62998605e-01 -1.86391532e-01 -8.37317705e-01 -1.05497412e-01
-9.21038445e-03 -4.86638755e-01 -8.56313169e-01 1.92388758e-01
1.39947936e-01 -7.43788123e-01 6.52804971e-01 -4.21499871e-02
1.20573783e+00 -2.62345403e-01 -2.62903839e-01 2.45276690e-01
1.05980062e+00 3.69295031e-01 9.32704687e-01 3.25642884e-01
5.12454867e-01 6.78686976e-01 -1.35012064e-02 -1.50608867e-01
3.29633385e-01 6.25842988e-01 5.71853779e-02 1.50185630e-01
-2.08276570e-01 -4.76003528e-01 -3.06248009e-01 1.55644327e-01
-3.10189664e-01 2.86578238e-01 -1.18560266e+00 4.44745034e-01
-1.58524728e+00 -1.38259876e+00 -1.92000970e-01 1.77064371e+00
8.06283474e-01 1.40014052e-01 -9.49899852e-02 1.77866012e-01
4.69653159e-01 -1.29755333e-01 -6.59276247e-01 -9.24068987e-01
-5.97932748e-02 6.03439510e-01 -2.04919308e-01 3.60041022e-01
-5.11167586e-01 8.52554679e-01 7.79745674e+00 3.53237599e-01
-4.51380581e-01 -2.05181226e-01 3.87653083e-01 -6.73845336e-02
-3.69112939e-01 1.08842283e-01 1.21377483e-01 -1.00971766e-01
9.24841702e-01 -2.97052413e-01 7.60258675e-01 6.67329848e-01
-4.01554734e-01 3.26334871e-02 -1.99626124e+00 9.02137041e-01
1.63551763e-01 -1.50556111e+00 5.72398365e-01 -3.45816791e-01
8.88980508e-01 -3.93834382e-01 1.99434042e-01 6.10097945e-01
7.28314757e-01 -1.55084193e+00 8.42702687e-01 6.49474919e-01
2.50791878e-01 -2.75612533e-01 5.12952328e-01 9.84434560e-02
-6.63525343e-01 -2.67030120e-01 -6.17780089e-01 -8.96346986e-01
-2.38778383e-01 -1.58291683e-01 -6.10455155e-01 4.85136844e-02
8.08237612e-01 3.47686946e-01 -7.96558022e-01 8.91961694e-01
-4.06618893e-01 -1.85030282e-01 3.09186913e-02 -1.08169340e-01
1.11708112e-01 1.76566988e-01 3.05656623e-03 7.01995432e-01
1.52164206e-01 5.91531456e-01 -2.59403795e-01 1.63401091e+00
7.06418604e-02 -1.97645545e-01 -7.47045159e-01 2.23979615e-02
6.95497334e-01 8.70714188e-01 -6.44624770e-01 -3.10769886e-01
-4.40928370e-01 9.47465241e-01 7.69126415e-01 4.23648119e-01
-8.11854601e-01 -4.20496255e-01 4.28182900e-01 3.46026212e-01
4.82506417e-02 -2.09237151e-02 -6.65511906e-01 -7.26499081e-01
-2.06883803e-01 -9.91386771e-01 4.11409527e-01 -1.73428559e+00
-1.81830728e+00 5.01491547e-01 1.81933135e-01 -8.39211762e-01
-2.72536457e-01 -1.14138079e+00 -9.31138158e-01 1.07488513e+00
-8.55671883e-01 -8.73760164e-01 -4.17068392e-01 7.45344818e-01
-6.91147447e-02 -2.43551001e-01 9.42714691e-01 -2.70819634e-01
2.74338901e-01 5.12122989e-01 -3.59712660e-01 3.64952445e-01
4.80057836e-01 -1.14864981e+00 6.57850385e-01 5.41091822e-02
3.71633410e-01 8.91524553e-01 7.88498521e-01 -3.32555652e-01
-8.42469931e-01 -2.64432788e-01 8.80567074e-01 -8.67915154e-01
8.45330834e-01 1.92138076e-01 -1.25233734e+00 9.75320935e-01
4.89489615e-01 1.28927261e-01 6.16735101e-01 8.93958211e-02
-1.03531396e+00 2.41342828e-01 -1.27871203e+00 8.87544692e-01
1.35897863e+00 -9.57147896e-01 -2.02648950e+00 3.92123908e-01
6.47953808e-01 -2.59405345e-01 -7.89878249e-01 2.05218513e-02
7.88542449e-01 -1.15688801e+00 1.36357415e+00 -1.34044635e+00
1.04485881e+00 -1.25586316e-01 -4.44770366e-01 -1.56929588e+00
-9.76362050e-01 1.83810055e-01 3.05991501e-01 5.26638150e-01
2.91818112e-01 -7.63058722e-01 3.43853176e-01 9.07586813e-01
3.29344459e-02 -3.84774774e-01 -8.89650762e-01 -1.03736925e+00
7.59992599e-01 2.28996366e-01 6.42581105e-01 1.36129236e+00
3.83010119e-01 5.39765537e-01 5.19491315e-01 -1.38526529e-01
4.43753809e-01 4.96471345e-01 4.23256278e-01 -1.67338669e+00
-2.86139876e-01 -9.43455875e-01 -6.60732567e-01 -4.65684503e-01
6.64479792e-01 -1.29656887e+00 -4.08295095e-01 -1.90074253e+00
1.78891778e-01 -9.78019387e-02 -3.86295840e-02 4.56587017e-01
-2.31223647e-02 1.24817885e-01 5.71955860e-01 2.82883406e-01
-1.80317342e-01 3.63585860e-01 1.52090800e+00 -1.78403065e-01
1.66666210e-01 -6.35466099e-01 -1.05723560e+00 1.08498669e+00
7.56859362e-01 -3.10132325e-01 -3.39753509e-01 -5.87843180e-01
5.00747442e-01 2.38062963e-01 1.03772235e+00 -1.38563359e+00
2.30996624e-01 -2.38529369e-01 1.39431369e+00 -1.18712164e-01
3.96229774e-01 -8.70360196e-01 1.13772787e-01 7.29833961e-01
-8.03198099e-01 7.36480713e-01 8.25472713e-01 2.16536567e-01
3.42614502e-02 -2.18860522e-01 8.81872833e-01 -5.29698074e-01
-6.93287015e-01 -2.98772156e-01 -4.66040492e-01 4.06024694e-01
8.65120888e-01 -3.88451159e-01 -7.12687910e-01 -4.81207281e-01
-8.87772024e-01 -5.61692305e-02 5.48727989e-01 3.20847273e-01
7.07431138e-01 -1.73282862e+00 -6.57749593e-01 -2.54816860e-01
2.34210845e-02 -6.40696824e-01 -2.15528570e-02 4.82857615e-01
-8.11933994e-01 4.15561527e-01 -1.02492046e+00 4.72224578e-02
-5.35154462e-01 1.24545860e+00 6.60423517e-01 -8.74419957e-02
-4.61470813e-01 7.33706117e-01 4.61424887e-01 -6.82614505e-01
-3.09737891e-01 -5.50499201e-01 6.10796851e-04 -2.30717555e-01
5.09278476e-01 3.12516093e-01 -3.75359356e-01 -4.02666420e-01
-3.50141943e-01 7.58808792e-01 -1.49854347e-01 -1.27881050e-01
1.08299363e+00 4.94507939e-01 -1.95442572e-01 4.59208071e-01
8.98379743e-01 -4.84779119e-01 -7.57996202e-01 -3.10324043e-01
-3.72326784e-02 -4.25897062e-01 -3.37218136e-01 -1.09986782e+00
-6.64391279e-01 1.05374801e+00 3.85172367e-01 -6.19868785e-02
5.18612742e-01 1.49722338e-01 3.84741217e-01 1.08049595e+00
1.61355466e-01 -8.11335623e-01 6.96491063e-01 5.54552972e-01
1.54207528e+00 -1.01722217e+00 3.80453914e-01 6.42262772e-02
-5.04681110e-01 1.11013007e+00 9.86797273e-01 -7.36506641e-01
3.64998758e-01 -2.79986233e-01 -2.14550272e-01 -3.67171377e-01
-4.75410044e-01 1.56722486e-01 4.81850386e-01 7.13457525e-01
4.36773181e-01 -2.76963979e-01 8.05985704e-02 6.54638946e-01
-6.30147099e-01 1.94445383e-02 2.84396887e-01 6.77940726e-01
-3.54544401e-01 -3.47277582e-01 -6.91757143e-01 4.53112125e-01
2.34878123e-01 -4.04097527e-01 -7.11202204e-01 1.10260975e+00
1.13559607e-02 4.86240149e-01 6.19587183e-01 -2.21965536e-01
5.20056605e-01 1.91970810e-01 1.10500288e+00 -3.55811357e-01
-5.86447537e-01 -9.78334010e-01 -2.34853432e-01 -6.50406241e-01
-2.22973183e-01 -5.36693037e-01 -1.47435892e+00 -6.49035096e-01
3.15197349e-01 -4.18214872e-02 -2.82974560e-02 1.19501674e+00
-8.03254023e-02 5.19288421e-01 -2.76137978e-01 -1.14397156e+00
-6.61627114e-01 -8.14495504e-01 -5.53675473e-01 8.30448210e-01
-7.66785536e-03 -8.55996788e-01 -4.44714606e-01 -3.70665252e-01] | [10.587660789489746, 2.308619499206543] |
8a10ff02-bd1a-480e-a3fa-afa280e78d82 | algorithms-for-learning-graphs-in-financial | 2012.15410 | null | https://arxiv.org/abs/2012.15410v1 | https://arxiv.org/pdf/2012.15410v1.pdf | Algorithms for Learning Graphs in Financial Markets | In the past two decades, the field of applied finance has tremendously benefited from graph theory. As a result, novel methods ranging from asset network estimation to hierarchical asset selection and portfolio allocation are now part of practitioners' toolboxes. In this paper, we investigate the fundamental problem of learning undirected graphical models under Laplacian structural constraints from the point of view of financial market times series data. In particular, we present natural justifications, supported by empirical evidence, for the usage of the Laplacian matrix as a model for the precision matrix of financial assets, while also establishing a direct link that reveals how Laplacian constraints are coupled to meaningful physical interpretations related to the market index factor and to conditional correlations between stocks. Those interpretations lead to a set of guidelines that practitioners should be aware of when estimating graphs in financial markets. In addition, we design numerical algorithms based on the alternating direction method of multipliers to learn undirected, weighted graphs that take into account stylized facts that are intrinsic to financial data such as heavy tails and modularity. We illustrate how to leverage the learned graphs into practical scenarios such as stock time series clustering and foreign exchange network estimation. The proposed graph learning algorithms outperform the state-of-the-art methods in an extensive set of practical experiments. Furthermore, we obtain theoretical and empirical convergence results for the proposed algorithms. Along with the developed methodologies for graph learning in financial markets, we release an R package, called fingraph, accommodating the code and data to obtain all the experimental results. | ['Daniel Perez Palomar', 'Jiaxi Ying', 'José Vinícius de Miranda Cardoso'] | 2020-12-31 | null | null | null | null | ['time-series-clustering'] | ['time-series'] | [-3.27213824e-01 3.29905115e-02 -1.65588409e-01 -9.08310041e-02
3.76139767e-02 -7.40271568e-01 3.82965952e-01 1.96355730e-02
1.54961497e-01 6.37113094e-01 -1.10756405e-01 -7.03594685e-01
-8.99117708e-01 -1.05912912e+00 -4.16093826e-01 -7.07159996e-01
-8.35582137e-01 4.62007821e-01 -7.55286366e-02 -1.53298289e-01
5.03368735e-01 5.14055312e-01 -8.83520663e-01 -4.59695190e-01
7.94956923e-01 1.06306076e+00 -3.36067617e-01 4.13408935e-01
-1.72103360e-01 8.52320194e-01 -6.84972256e-02 -1.07828879e+00
6.84310138e-01 -3.78639668e-01 -3.07528436e-01 4.31653976e-01
5.71812317e-03 -1.71866547e-02 -4.19341266e-01 1.31180441e+00
2.92722225e-01 -2.73853950e-02 9.07845080e-01 -1.47264206e+00
-8.44285190e-01 9.38499212e-01 -1.07850993e+00 4.17207211e-01
-2.29124114e-01 -4.64792028e-02 1.59510767e+00 -8.91775846e-01
3.29696417e-01 9.39526260e-01 6.54413640e-01 -3.05388063e-01
-1.07105350e+00 -5.99004984e-01 1.64386496e-01 2.32555062e-01
-1.18517148e+00 1.91509277e-02 1.27863240e+00 -7.79399276e-01
5.65595269e-01 2.00411543e-01 8.35163951e-01 5.49722016e-01
3.39827508e-01 3.82289112e-01 9.88801479e-01 -4.10560340e-01
1.93457335e-01 -1.36827668e-02 4.41777885e-01 7.91711211e-01
1.03129613e+00 4.86969911e-02 -4.59999651e-01 -4.12387460e-01
9.76580739e-01 -1.39863282e-01 -1.29324555e-01 -8.80333602e-01
-8.65256190e-01 1.23096299e+00 1.83531344e-01 2.00024426e-01
-3.18580270e-01 7.90475905e-02 2.41671473e-01 4.51929450e-01
6.00875378e-01 1.99293375e-01 -2.13192761e-01 3.82291466e-01
-8.55260849e-01 1.30657749e-02 1.19277799e+00 8.74865353e-01
6.64971292e-01 4.44114178e-01 3.66196811e-01 3.43036681e-01
7.30042040e-01 5.18255353e-01 -2.40237955e-02 -7.03246295e-01
7.54682481e-01 7.65451491e-01 -8.23585093e-02 -1.52352452e+00
-5.57737887e-01 -9.38009262e-01 -1.16131210e+00 2.60208279e-01
6.60399318e-01 -3.78049493e-01 -1.25901490e-01 1.34576273e+00
2.00073093e-01 3.32551479e-01 -6.95482269e-02 6.95441961e-01
1.83381081e-01 3.01683247e-01 -3.07991475e-01 -5.76679587e-01
1.19174957e+00 -5.31050801e-01 -6.12540662e-01 2.28063107e-01
3.78352642e-01 -6.17704153e-01 6.62418187e-01 3.67915630e-01
-1.03143024e+00 -1.38219148e-01 -9.37718451e-01 7.86958098e-01
-2.52623200e-01 -1.67327017e-01 1.04090858e+00 9.12380576e-01
-1.01492822e+00 7.69649804e-01 -7.83039749e-01 7.17124296e-03
1.52607739e-01 2.31558189e-01 1.18159428e-01 6.84879899e-01
-1.34759271e+00 6.32969797e-01 3.35023463e-01 4.06938434e-01
-2.74102926e-01 -5.17544687e-01 -5.81445694e-01 1.34341866e-01
3.86231035e-01 -8.54860246e-01 7.40606010e-01 -9.12647009e-01
-1.21741962e+00 6.23959243e-01 5.79426587e-01 -6.85593784e-01
9.44738209e-01 1.86386541e-01 -4.89363343e-01 2.70306557e-01
-2.72326857e-01 -4.06587481e-01 9.43152905e-01 -7.59777009e-01
-1.89669967e-01 -2.69995272e-01 1.66055132e-02 -7.24972561e-02
-5.88947117e-01 -2.24096984e-01 -6.81167543e-02 -1.05145657e+00
7.83264861e-02 -6.64194047e-01 -1.81959018e-01 -2.72594631e-01
-6.62517250e-01 4.90040369e-02 4.08374608e-01 -6.83262587e-01
1.54301929e+00 -1.74198103e+00 2.47729719e-01 1.07231951e+00
4.26844239e-01 -3.31946552e-01 1.72892824e-01 8.05210471e-01
-3.44532311e-01 2.44548663e-01 -3.96516383e-01 1.01748750e-01
3.75926048e-01 -1.87922418e-01 -3.32608640e-01 8.77195239e-01
-1.50219798e-02 7.88862646e-01 -6.81083620e-01 -3.73162925e-01
6.25828877e-02 1.70135275e-01 -2.61752486e-01 2.54134052e-02
1.05518587e-01 2.64273465e-01 -7.17350304e-01 5.73412895e-01
7.68446505e-01 -7.65467167e-01 5.70086062e-01 -1.94176570e-01
6.59866482e-02 -8.06524158e-02 -1.50935113e+00 9.37753558e-01
1.49618849e-01 3.26350093e-01 6.62095100e-02 -1.40266895e+00
9.61229265e-01 4.40644845e-02 5.67649722e-01 -5.49185276e-01
1.40024692e-01 2.52498657e-01 2.10126460e-01 -3.57589453e-01
2.06391830e-02 -2.08184004e-01 1.11919224e-01 8.15264225e-01
-1.89518034e-01 7.05381781e-02 5.80502033e-01 4.31218386e-01
8.73416901e-01 -1.83608562e-01 5.00856400e-01 -6.00770473e-01
6.72739267e-01 -3.57500672e-01 2.05526859e-01 4.03780282e-01
1.58468887e-01 -2.82421783e-02 1.04652250e+00 -3.32844615e-01
-1.12350941e+00 -1.06688404e+00 -1.64887771e-01 4.96362567e-01
-3.27672571e-01 -1.61490828e-01 -6.82404578e-01 -4.75707859e-01
4.28329408e-01 4.47190553e-01 -5.68458557e-01 1.45098135e-01
-3.99025857e-01 -1.19016469e+00 1.06867716e-01 3.01065683e-01
3.65982801e-01 -8.98418963e-01 -3.36019278e-01 4.21874188e-02
5.01981914e-01 -8.55342031e-01 -4.46588695e-01 1.56715605e-02
-1.01811838e+00 -1.48494232e+00 -7.45849431e-01 -4.79304671e-01
7.22318709e-01 3.65366749e-02 1.31353331e+00 2.92307049e-01
-1.65071249e-01 7.12392330e-01 -2.33572230e-01 9.01721232e-03
-2.32560486e-01 1.46803066e-01 3.15046422e-02 4.96155560e-01
-4.02239244e-03 -9.10524607e-01 -6.15168095e-01 2.69093990e-01
-8.90246391e-01 -1.02532588e-01 6.98764145e-01 6.64929390e-01
2.65051097e-01 5.27730048e-01 7.16125906e-01 -1.06784058e+00
8.49261880e-01 -7.40560055e-01 -1.44217658e+00 4.92756695e-01
-9.46632624e-01 2.40897924e-01 4.67547536e-01 -1.64635018e-01
-7.43971944e-01 -2.80632645e-01 5.46541750e-01 -2.08326206e-01
6.53679490e-01 1.18676877e+00 6.62248060e-02 -3.08324665e-01
4.66730520e-02 1.17535152e-01 1.47299111e-01 -2.53551364e-01
4.72133100e-01 1.72056600e-01 2.28882674e-02 -5.01719296e-01
1.23379230e+00 2.60183692e-01 6.30536795e-01 -7.85035968e-01
-6.21293724e-01 -1.03301704e-01 -7.51460195e-01 -4.76930559e-01
4.70584780e-01 -6.04787111e-01 -9.07187641e-01 4.98600692e-01
-6.25397742e-01 -1.30463824e-01 1.54841304e-01 4.41572040e-01
-5.51785588e-01 7.56411850e-01 -8.31333160e-01 -1.30606222e+00
-2.56351560e-01 -5.59775054e-01 2.35855773e-01 -3.04305889e-02
-2.88433637e-02 -1.89374638e+00 2.56596416e-01 2.54390817e-02
2.45346576e-01 5.39818764e-01 1.18507028e+00 -7.39016771e-01
-7.88097084e-01 -1.04179680e-01 -2.17632681e-01 1.79180935e-01
1.96982041e-01 1.73418924e-01 -4.41195607e-01 -4.68412161e-01
1.25554860e-01 2.17313424e-01 7.13806331e-01 5.10856867e-01
7.09225237e-01 -5.00878155e-01 2.81201247e-02 6.01984501e-01
1.51995146e+00 1.70354415e-02 2.95409977e-01 4.18587536e-01
5.74602604e-01 8.57630014e-01 2.28545368e-01 8.48902941e-01
1.98298335e-01 2.78811783e-01 4.74640608e-01 3.87430079e-02
5.17979383e-01 -1.68646976e-01 3.05309951e-01 1.28452325e+00
-2.73042053e-01 -1.23236842e-01 -9.53138709e-01 5.29888310e-02
-2.01444578e+00 -1.05555725e+00 -4.13349539e-01 2.45452762e+00
5.57331681e-01 4.48343277e-01 4.89573628e-01 9.53929350e-02
9.31901276e-01 1.35233372e-01 -5.57668746e-01 2.10351333e-01
-2.68228680e-01 2.09668814e-03 8.30923319e-01 5.52396655e-01
-1.07349586e+00 3.15414369e-01 6.35468578e+00 6.06860042e-01
-7.86484599e-01 -2.99308509e-01 7.77327955e-01 2.16268137e-01
-6.29913867e-01 8.76553208e-02 -5.37104964e-01 5.30971050e-01
8.02806377e-01 -7.04060197e-01 4.94695276e-01 6.38885260e-01
2.71467179e-01 3.75110954e-01 -7.83708453e-01 9.16741669e-01
-2.19693616e-01 -1.23410511e+00 6.77909330e-03 3.84708941e-01
7.70353794e-01 -4.76368189e-01 3.42677176e-01 -1.39818490e-01
3.53255242e-01 -8.08471560e-01 7.33688533e-01 8.23434532e-01
2.89335102e-01 -9.69720006e-01 6.82869136e-01 1.81781560e-01
-1.49993277e+00 -9.58165750e-02 -5.52244425e-01 -2.58408636e-01
2.72860378e-01 1.02695453e+00 -5.36018908e-01 8.81745994e-01
2.10754037e-01 1.07873964e+00 -7.03639388e-01 1.19187069e+00
-2.00119734e-01 7.63703644e-01 -1.27484262e-01 -1.07956290e-01
1.26890376e-01 -1.01916564e+00 3.67462009e-01 8.49257827e-01
6.10809922e-01 5.26848668e-03 -2.28565037e-01 1.01570845e+00
5.25167435e-02 3.74319553e-01 -6.17854178e-01 -4.86292034e-01
2.37966657e-01 1.45796204e+00 -1.21350276e+00 -1.12673789e-01
-8.22165072e-01 3.61034393e-01 2.64447361e-01 6.27975464e-01
-8.49908173e-01 -2.70415902e-01 1.60038292e-01 3.57326210e-01
4.58142877e-01 -4.24995750e-01 -3.22337180e-01 -1.27911830e+00
1.04824975e-01 -7.77025104e-01 6.51581466e-01 -3.69934082e-01
-1.93361664e+00 3.33826095e-01 1.01035118e-01 -1.44598973e+00
-3.41406345e-01 -1.02064872e+00 -8.19475114e-01 7.44593143e-01
-1.56692338e+00 -7.47560859e-01 1.96499661e-01 5.44704914e-01
-8.00395831e-02 -6.37576401e-01 3.06710482e-01 2.66134202e-01
-7.22524524e-01 1.93028718e-01 2.38915622e-01 3.70216370e-01
1.93258643e-01 -1.48713613e+00 5.69513142e-01 8.17421794e-01
4.24280763e-01 6.58790052e-01 7.26565957e-01 -9.24297690e-01
-1.42763901e+00 -7.04854965e-01 5.20333946e-01 -2.18998179e-01
1.77283716e+00 -2.98547119e-01 -8.32126617e-01 7.18874514e-01
4.37515765e-01 -1.58120349e-01 6.92849040e-01 2.28942223e-02
-1.20975889e-01 -1.54245004e-01 -7.74190307e-01 5.11565149e-01
8.65013778e-01 -2.33840898e-01 -4.04108584e-01 4.82365698e-01
3.41636121e-01 5.63932583e-02 -8.98664355e-01 1.51549587e-02
3.26323241e-01 -1.26514482e+00 1.08500481e+00 -6.41310453e-01
-2.13464163e-03 -1.30415261e-01 2.41874516e-01 -1.10384679e+00
-4.39082712e-01 -9.39691663e-01 -4.20729607e-01 1.19636238e+00
5.94005346e-01 -1.05588794e+00 7.64918685e-01 4.39700276e-01
4.26770061e-01 -5.25859237e-01 -7.89406300e-01 -8.23681414e-01
2.15701833e-01 -1.72789991e-01 3.08168560e-01 1.17575252e+00
1.75543547e-01 2.63637573e-01 -4.48038816e-01 1.56890228e-01
1.24418175e+00 2.74309039e-01 4.93203640e-01 -1.54506600e+00
-6.91085517e-01 -8.54978085e-01 -4.23883617e-01 -5.99175394e-01
3.31580341e-01 -9.34732318e-01 -9.02962923e-01 -1.02362633e+00
2.32532531e-01 -2.17591867e-01 -3.85924011e-01 -1.54217154e-01
2.87230983e-02 -2.74162157e-03 1.36541873e-01 3.09950471e-01
-4.81837749e-01 2.53616571e-01 1.11370707e+00 -7.82541409e-02
1.62679866e-01 4.69408393e-01 -5.87413847e-01 7.18821526e-01
8.19736838e-01 -2.16809317e-01 -6.14373267e-01 1.64203957e-01
1.08806694e+00 3.59737873e-01 3.98396373e-01 -4.73881155e-01
2.85979062e-01 -1.67687714e-01 1.64461300e-01 -5.12105763e-01
-1.98212475e-01 -9.53196704e-01 3.94561917e-01 5.79627454e-01
-1.67429239e-01 6.07255459e-01 -1.18816875e-01 6.95820570e-01
-2.40202293e-01 -4.98115033e-01 4.26286221e-01 -1.16208695e-01
-3.40106487e-01 3.63921851e-01 -1.36977792e-01 2.74332851e-01
1.03440487e+00 1.00633927e-01 -2.00545579e-01 -7.97645271e-01
-7.51313508e-01 3.84484291e-01 1.28531963e-01 -5.11962213e-02
3.80862743e-01 -1.43242872e+00 -8.76416206e-01 5.06716482e-02
-4.12878424e-01 -7.97370255e-01 1.15551919e-01 1.20977604e+00
-7.15162933e-01 4.55358267e-01 -1.19834039e-02 -1.98004723e-01
-8.23250771e-01 7.65134692e-01 4.21817511e-01 -5.66047311e-01
-4.33436543e-01 2.47299269e-01 8.52471665e-02 9.14221816e-03
1.16780043e-01 -2.62718111e-01 -3.68396133e-01 4.21036214e-01
2.72837400e-01 6.21244371e-01 -1.25001416e-01 -3.77809554e-01
-1.46611780e-01 9.14146483e-01 4.65876728e-01 2.63522118e-02
1.55591488e+00 -2.88517386e-01 -3.32482010e-01 4.96098399e-01
7.83726096e-01 2.79103488e-01 -1.20901012e+00 -1.61077634e-01
6.22301102e-01 -2.96902508e-01 -2.98190892e-01 -2.66791791e-01
-1.66893673e+00 7.51186252e-01 1.10433154e-01 1.18943822e+00
9.39408600e-01 -1.62307233e-01 2.49151528e-01 3.86576444e-01
4.54129100e-01 -9.29723144e-01 1.41949341e-01 2.39814162e-01
8.76886487e-01 -8.80311668e-01 2.81618029e-01 -6.11752033e-01
-5.70600033e-01 1.53046930e+00 -8.58523250e-02 -5.09651959e-01
1.07164967e+00 2.87175238e-01 -6.01580068e-02 -4.80241597e-01
-4.74571168e-01 2.18840856e-02 4.96533811e-01 2.99682349e-01
5.10855556e-01 6.78672120e-02 -3.51331800e-01 6.16860509e-01
-3.48308802e-01 -4.65464354e-01 5.65954030e-01 4.83228534e-01
-3.49704653e-01 -1.13935328e+00 -3.82253557e-01 4.13758665e-01
-4.70133543e-01 -2.51349330e-01 -4.37544048e-01 1.14656281e+00
-6.05030537e-01 5.44213116e-01 -4.45860066e-02 9.00521204e-02
2.20064253e-01 1.08009040e-01 3.33057433e-01 -1.90849245e-01
-2.46300146e-01 3.53841275e-01 -1.13500401e-01 2.54887268e-02
-6.63613737e-01 -9.23787653e-01 -1.00265610e+00 -5.58995724e-01
-3.79051924e-01 4.14191335e-01 8.49821195e-02 6.96418226e-01
-2.66081607e-03 4.61582661e-01 7.64628649e-01 -6.70589626e-01
-9.56551373e-01 -8.08885396e-01 -1.46711004e+00 1.49813041e-01
-3.65990363e-02 -7.96278059e-01 -8.78012836e-01 -8.05452690e-02] | [5.106787204742432, 4.097522258758545] |
0533ef0e-1077-47a6-929d-4047b507a12a | a-simple-algorithm-for-the-constrained | 2103.02919 | null | https://arxiv.org/abs/2103.02919v1 | https://arxiv.org/pdf/2103.02919v1.pdf | A Simple Algorithm for the Constrained Sequence Problems | In this paper we address the constrained longest common subsequence problem. Given two sequences $X$, $Y$ and a constrained sequence $P$, a sequence $Z$ is a constrained longest common subsequence for $X$ and $Y$ with respect to $P$ if $Z$ is the longest subsequence of $X$ and $Y$ such that $P$ is a subsequence of $Z$. Recently, Tsai \cite{Tsai} proposed an $O(n^2 \cdot m^2 \cdot r)$ time algorithm to solve this problem using dynamic programming technique, where $n$, $m$ and $r$ are the lengths of $X$, $Y$ and $P$, respectively. In this paper, we present a simple algorithm to solve the constrained longest common subsequence problem in $O(n \cdot m \cdot r)$ time and show that the constrained longest common subsequence problem is equivalent to a special case of the constrained multiple sequence alignment problem which can also be solved. | ['S. K. Kim', 'Alfredo De Santis', 'Ngai Lam Ho', 'Francis Yuk Lun Chin'] | 2021-03-04 | null | null | null | null | ['multiple-sequence-alignment'] | ['medical'] | [ 5.40133417e-01 -3.44684005e-01 -9.41122696e-02 -2.53257155e-01
-5.91732442e-01 -8.46355319e-01 -4.52844113e-01 2.85009503e-01
-8.46393645e-01 6.05344296e-01 -6.63066626e-01 -6.20268643e-01
-5.05029321e-01 -8.65116894e-01 -6.26574636e-01 -7.34152496e-01
-6.06526077e-01 1.39006317e-01 1.60930708e-01 -5.63053727e-01
7.10536420e-01 3.50758404e-01 -1.64836776e+00 -3.06884408e-01
5.11502504e-01 7.73016334e-01 5.20157218e-01 7.94923306e-01
-6.64912760e-01 -1.65969715e-01 -5.23620367e-01 -5.29968143e-01
6.86304033e-01 -8.66372526e-01 -1.10927105e+00 -1.81931019e-01
1.83594346e-01 2.25496903e-01 2.87186742e-01 1.34586573e+00
1.52795792e-01 2.56177217e-01 6.18371889e-02 -1.39520311e+00
-9.35537219e-02 6.49323761e-01 -1.03121257e+00 2.31410503e-01
6.65262282e-01 -1.98572502e-02 1.30194771e+00 -4.08488452e-01
8.94685507e-01 9.07945156e-01 4.24284607e-01 1.88761234e-01
-8.64997566e-01 -1.04986167e+00 3.80186588e-01 2.83901449e-02
-1.70445931e+00 1.78021923e-01 4.53575850e-01 -2.57836848e-01
1.42914224e+00 4.63016361e-01 5.06413102e-01 -3.04689139e-01
2.73269564e-01 2.16958940e-01 7.78653741e-01 -9.41046834e-01
1.04152793e-02 -5.68971157e-01 4.42254812e-01 9.40406501e-01
2.20803395e-02 -6.93349838e-02 -3.54054242e-01 -2.44601116e-01
2.77389288e-01 -1.10592715e-01 -1.70131490e-01 2.71697845e-02
-7.55919158e-01 1.01560557e+00 -1.05508290e-01 3.65008026e-01
-2.51944549e-02 3.13191175e-01 1.90237150e-01 5.59345663e-01
-2.16844693e-01 3.30790281e-01 -7.51549482e-01 -3.57906848e-01
-7.15512931e-01 3.80600482e-01 6.44836187e-01 1.48704219e+00
1.25271964e+00 -1.33742094e-01 7.92696774e-01 7.25111485e-01
2.60840096e-02 7.33304679e-01 1.86271951e-01 -1.24457991e+00
6.86074555e-01 7.71610022e-01 1.04918100e-01 -8.71050358e-01
-2.69316941e-01 2.46756151e-01 -4.45482284e-01 6.24960326e-02
5.26362956e-01 7.89523795e-02 -5.27705491e-01 2.05979705e+00
9.41916406e-02 2.71002632e-02 -1.66233137e-01 4.93688166e-01
6.54323176e-02 7.73687303e-01 -4.84444685e-02 -1.05885613e+00
1.34880543e+00 -7.31692433e-01 -2.53160447e-01 -3.50845963e-01
8.01503062e-01 -1.33817315e+00 9.28651750e-01 1.91584975e-01
-1.54819381e+00 -2.06379294e-01 -1.09770358e+00 2.67638624e-01
-1.22531876e-01 -5.34774363e-01 3.48550051e-01 6.60111964e-01
-9.78214383e-01 5.35040557e-01 -3.76782089e-01 -2.10719079e-01
-4.22761768e-01 7.24070489e-01 -2.92506278e-01 -2.86555082e-01
-8.63546729e-01 5.43447554e-01 1.98456094e-01 -2.34266639e-01
-6.68752342e-02 -2.51205266e-01 -8.17347944e-01 -7.29039386e-02
7.72238493e-01 -8.22581574e-02 1.15873575e+00 -8.38837385e-01
-1.09887826e+00 1.25925958e+00 -6.57375455e-01 -2.29386166e-01
5.03640063e-03 2.49399468e-01 -2.16976762e-01 2.48040125e-01
3.93201977e-01 3.27089369e-01 4.93080914e-01 -4.88585472e-01
-9.25739527e-01 -5.49174130e-01 7.81501755e-02 1.36198461e-01
5.14031529e-01 7.19015479e-01 -7.05940962e-01 -5.19391835e-01
6.88970149e-01 -1.25007451e+00 -7.13751912e-01 -2.92426407e-01
-2.03686267e-01 -6.02638908e-02 1.72397226e-01 -4.13791090e-01
1.54087317e+00 -2.23459506e+00 3.03934932e-01 7.18876600e-01
-1.75293088e-01 2.61422783e-01 -3.68423671e-01 8.89606059e-01
-4.49644804e-01 3.89884114e-01 -7.30499625e-01 1.96930349e-01
-2.03128949e-01 4.10614252e-01 -6.47823736e-02 3.66408139e-01
-4.71247107e-01 1.78102940e-01 -6.79674923e-01 -3.42908084e-01
-2.34665319e-01 -2.48324975e-01 -4.09048408e-01 1.05836973e-01
-5.30868173e-01 -2.62039751e-01 -3.02418023e-01 5.48692584e-01
9.47063982e-01 -3.60306650e-02 7.19294965e-01 5.59591770e-01
-6.85704529e-01 -3.15878447e-03 -1.41407883e+00 1.64368129e+00
-1.46879941e-01 8.00446942e-02 3.25111330e-01 -1.33107316e+00
1.32124436e+00 8.74359310e-02 7.14913130e-01 -9.18970585e-01
9.41810757e-02 5.26568830e-01 -6.74223751e-02 -5.44116795e-01
6.03912771e-01 -6.68516517e-01 -4.49913174e-01 9.37920153e-01
-5.75405061e-01 -3.49074602e-01 6.18713498e-01 1.30940035e-01
1.29689074e+00 -1.49986669e-01 4.75210816e-01 -4.64470178e-01
9.47015285e-01 -8.18324909e-02 1.01040876e+00 8.00468862e-01
-4.06410158e-01 1.49453551e-01 7.57891238e-01 -2.60580659e-01
-1.25560915e+00 -8.17320168e-01 2.84131378e-01 1.18145096e+00
4.10226315e-01 -5.64550042e-01 -7.62322187e-01 -3.03397536e-01
-2.13115335e-01 3.44341785e-01 1.05903968e-01 2.41843194e-01
-1.26723313e+00 -4.81708676e-01 2.38681093e-01 2.89146930e-01
5.05459845e-01 -1.10011685e+00 -1.02504671e+00 5.97850084e-01
-2.48036459e-01 -8.12508166e-01 -1.01704681e+00 2.72076428e-01
-9.07089114e-01 -1.33502519e+00 -5.98268151e-01 -1.26839590e+00
7.88217902e-01 5.36709428e-01 8.39422286e-01 4.57428336e-01
-7.36621857e-01 -1.07877888e-02 -4.71272349e-01 -1.14036866e-01
2.82638688e-02 -1.49019703e-01 -1.57649685e-02 -7.25991011e-01
4.35049891e-01 -7.58216083e-01 -6.14974558e-01 6.52729094e-01
-9.06098902e-01 -4.87683743e-01 3.69441450e-01 4.99856591e-01
1.00928783e+00 2.37895012e-01 3.49332780e-01 -6.98829710e-01
3.73652279e-01 -1.20240740e-01 -9.33247149e-01 2.98729509e-01
-7.26454139e-01 1.54273495e-01 8.99522066e-01 -9.32209492e-02
-3.80851746e-01 2.15706080e-01 -5.35899699e-01 -5.44758551e-02
3.34483951e-01 8.33394647e-01 -1.43685699e-01 8.08525681e-02
2.41565213e-01 3.61517012e-01 1.78431213e-01 -5.84447801e-01
1.80789545e-01 3.74050766e-01 4.45177972e-01 -5.41839540e-01
4.54419523e-01 3.60726237e-01 2.45650038e-01 -6.38307452e-01
9.49123949e-02 -6.49209619e-01 -4.79088753e-01 3.41475338e-01
4.52204704e-01 -4.47172105e-01 -1.16346407e+00 1.67060539e-01
-8.24879825e-01 -8.53951797e-02 2.13212576e-02 2.73723722e-01
-8.10473502e-01 9.26203847e-01 -9.50989053e-02 -8.00594985e-01
-6.30119860e-01 -1.37804306e+00 4.74263042e-01 -3.85783464e-02
-4.66223031e-01 -1.38138354e-01 1.91598475e-01 3.76203582e-02
-1.62330158e-02 2.06527680e-01 1.41405511e+00 -1.57830209e-01
-7.03610301e-01 -2.89675862e-01 -2.12878764e-01 1.93952456e-01
-4.03769091e-02 2.43235841e-01 3.60446811e-01 -5.79374015e-01
-7.77803641e-03 3.11530650e-01 1.52758107e-01 3.50757837e-01
6.86260760e-01 -3.95585328e-01 -5.37096143e-01 4.42742020e-01
1.88531899e+00 1.12938011e+00 6.72094703e-01 5.08908689e-01
-1.61729634e-01 5.85595787e-01 8.79146874e-01 4.68953520e-01
-2.41580401e-02 6.04131818e-01 1.18968613e-01 5.46018720e-01
6.55708790e-01 2.73818970e-01 1.80175841e-01 8.61690402e-01
1.13506742e-01 -1.08666420e-02 -1.04895854e+00 6.20943129e-01
-1.53676379e+00 -9.28113937e-01 -4.37350683e-02 2.17027450e+00
7.62503326e-01 7.83550814e-02 3.32394391e-01 3.93889397e-01
9.23134327e-01 9.16572139e-02 -3.76058728e-01 -1.23164904e+00
-8.16201121e-02 1.16909492e+00 6.62251353e-01 7.06111789e-01
-4.26154464e-01 6.40319109e-01 6.13868809e+00 7.56870925e-01
-9.61531043e-01 -3.43006194e-01 1.68856263e-01 -4.49247867e-01
-3.93382311e-01 2.00043604e-01 -7.91875362e-01 7.46552944e-01
9.96042907e-01 -3.38028222e-01 4.35664356e-01 8.70309830e-01
-3.80461998e-02 -5.38150370e-01 -8.29263031e-01 1.22995710e+00
-3.85849625e-02 -1.40771234e+00 -3.17665666e-01 7.06084371e-02
5.64244151e-01 -3.93867314e-01 8.61408189e-02 -1.56130433e-01
5.12597561e-01 -7.64678657e-01 4.16675568e-01 -3.03527683e-01
8.41113210e-01 -1.27799201e+00 2.69897997e-01 5.39859354e-01
-1.74233878e+00 -8.19221139e-02 -4.04293895e-01 -1.39881313e-01
4.57491815e-01 2.01146096e-01 -1.68518767e-01 5.40362477e-01
9.12911296e-01 -3.60087417e-02 3.95995468e-01 6.25184536e-01
-3.50096561e-02 -1.82942212e-01 -4.11788911e-01 -1.63936049e-01
6.74549520e-01 -6.39864802e-01 2.97050327e-01 1.06489897e+00
8.13127995e-01 9.38609719e-01 3.53744239e-01 5.35212979e-02
4.35936786e-02 5.07734179e-01 -4.45831150e-01 -1.29942507e-01
5.95566511e-01 4.96286362e-01 -9.25887764e-01 -1.32576779e-01
-4.73224014e-01 8.12657416e-01 1.29423082e-01 1.33133791e-02
-7.11934924e-01 -1.04292130e+00 1.02387786e+00 -1.44049421e-01
6.40358686e-01 -5.54413140e-01 2.26420201e-02 -6.92968786e-01
1.75872698e-01 -1.08618188e+00 9.91268456e-01 -4.60989505e-01
-5.38391650e-01 6.27285421e-01 -1.24975629e-01 -1.07722080e+00
-1.88362479e-01 -4.05203313e-01 -3.69280934e-01 1.30133140e+00
-1.16879880e+00 -2.60527372e-01 5.27767062e-01 7.60756314e-01
3.93982053e-01 -8.95660650e-03 9.69442666e-01 -1.89338818e-01
-4.72619861e-01 6.88046694e-01 3.13810349e-01 -2.48502031e-01
2.01803744e-01 -6.38103783e-01 2.87070513e-01 1.01921594e+00
-2.18800187e-01 1.19351029e+00 8.80638242e-01 -5.15537500e-01
-1.55912161e+00 -4.24037695e-01 1.35312355e+00 4.04718429e-01
2.68649042e-01 8.35674629e-02 -5.14140725e-01 8.39275777e-01
3.18518311e-01 -5.41267276e-01 1.32976925e+00 -2.43716732e-01
-5.28079450e-01 -8.69716182e-02 -1.45072460e+00 6.18686438e-01
1.22781909e+00 -4.12148118e-01 -2.40841791e-01 -1.32979453e-01
7.13987410e-01 -2.98057824e-01 -8.51675689e-01 3.12177658e-01
6.96115613e-01 -9.13028240e-01 9.69049871e-01 -1.46332234e-01
1.77187309e-01 -3.36833775e-01 -6.20674670e-01 -5.22456765e-01
-2.83223875e-02 -9.74348128e-01 6.67766452e-01 5.78659892e-01
5.83989680e-01 -4.39818978e-01 1.07406223e+00 4.45991665e-01
-2.22948864e-01 -6.86240017e-01 -1.36451995e+00 -1.09042048e+00
4.60315704e-01 -5.14039159e-01 8.09175730e-01 7.21590936e-01
4.82544899e-01 -1.71247184e-01 -1.74333036e-01 -6.31851610e-03
4.24611032e-01 8.47359836e-01 5.06776869e-01 -6.80718899e-01
-4.91534352e-01 -3.40790749e-01 -5.38849831e-02 -1.16343629e+00
1.00636199e-01 -7.52376854e-01 6.21595532e-02 -9.24528480e-01
1.83141932e-01 -8.01799059e-01 -2.46067122e-02 1.66152209e-01
3.61118913e-02 5.10844290e-02 4.06588763e-01 1.33868307e-01
-2.50010550e-01 -1.70469761e-01 8.89112711e-01 5.10755144e-02
-1.02678619e-01 -1.77562207e-01 -5.86499274e-01 5.55266321e-01
6.23419702e-01 -3.97150248e-01 -2.84921110e-01 -6.80383265e-01
5.51526010e-01 8.34184587e-01 -6.47342682e-01 -2.38030612e-01
1.03286691e-01 -6.16823137e-01 -6.01675391e-01 -7.08990395e-01
1.55281246e-01 -6.57390177e-01 4.59394038e-01 1.01183009e+00
-1.45856410e-01 9.92340922e-01 2.50239164e-01 2.66650081e-01
-1.21378675e-01 -7.41566002e-01 9.09640431e-01 -6.11204982e-01
-9.12733734e-01 3.67496073e-01 -5.58658540e-01 -3.70912224e-01
1.36548662e+00 -4.60969239e-01 -9.03858468e-02 -1.81591213e-01
-7.29563534e-01 3.36724401e-01 5.43136716e-01 4.22888212e-02
7.21267462e-01 -7.16753125e-01 -1.78895384e-01 2.70653844e-01
7.84855410e-02 -3.60466391e-02 4.89116341e-01 3.78727227e-01
-1.11509180e+00 5.41392803e-01 -1.92255661e-01 -4.00571436e-01
-1.78375185e+00 9.68718648e-01 -2.90457830e-02 -4.29583579e-01
-2.87030071e-01 1.36541545e+00 -6.07454777e-01 -1.72046408e-01
1.37581751e-01 -1.77861586e-01 4.22282338e-01 -1.64661095e-01
4.88340884e-01 3.29197496e-01 -2.64929533e-01 -6.82047606e-01
-6.73381329e-01 1.18595338e+00 -3.15814286e-01 -5.33702314e-01
1.06835532e+00 -3.32675517e-01 -8.21070552e-01 -1.80936113e-01
1.65780973e+00 -3.66501231e-03 -5.87678432e-01 -2.31171191e-01
1.67458877e-01 -6.50792718e-01 -1.05209243e+00 -2.28675425e-01
-1.00125182e+00 6.05909050e-01 4.66459870e-01 1.39709760e-03
1.40262544e+00 -7.39047825e-02 1.02220786e+00 3.03837359e-01
9.07746911e-01 -1.24074388e+00 6.24944381e-02 7.04345286e-01
2.43749022e-01 -5.10462701e-01 -6.78495094e-02 -6.87742472e-01
-3.09222966e-01 1.00248814e+00 4.33378100e-01 -1.18169717e-01
5.87680399e-01 3.34561318e-01 3.93518023e-02 7.16610765e-03
-4.69323605e-01 -1.32088780e-01 -6.49689734e-01 5.88449799e-02
4.29775923e-01 -1.47558615e-01 -1.48529685e+00 3.26707423e-01
-4.86972421e-01 -1.44915916e-02 4.72815275e-01 1.80321407e+00
-9.43201184e-01 -2.03375697e+00 -3.28864098e-01 -5.36931120e-02
-5.56563675e-01 -1.47110969e-01 -8.80578905e-02 5.96343875e-01
3.80145460e-01 9.27984357e-01 1.56612128e-01 -3.39211404e-01
3.32159489e-01 3.85285795e-01 6.12957001e-01 -4.33093429e-01
-6.79995477e-01 4.23514187e-01 -1.64246812e-01 -4.75922346e-01
-3.65500450e-01 -6.27963603e-01 -1.90022993e+00 -8.21146369e-01
-1.69418514e-01 5.65905094e-01 7.33227909e-01 7.68702328e-01
1.90422267e-01 -2.88211137e-01 7.43063211e-01 5.48603274e-02
-2.62028396e-01 -3.35190654e-01 -8.94280255e-01 2.33814061e-01
-1.17826708e-01 -6.63851202e-02 -4.46444079e-02 1.29136741e-01] | [6.430612564086914, 4.7293548583984375] |
7d3225e9-d0b9-460e-ad4c-b22ea6769e31 | exploring-spatial-temporal-variations-of | 2306.16031 | null | https://arxiv.org/abs/2306.16031v1 | https://arxiv.org/pdf/2306.16031v1.pdf | Exploring Spatial-Temporal Variations of Public Discourse on Social Media: A Case Study on the First Wave of the Coronavirus Pandemic in Italy | This paper proposes a methodology for exploring how linguistic behaviour on social media can be used to explore societal reactions to important events such as those that transpired during the SARS CoV2 pandemic. In particular, where spatial and temporal aspects of events are important features. Our methodology consists of grounding spatial-temporal categories in tweet usage trends using time-series analysis and clustering. Salient terms in each category were then identified through qualitative comparative analysis based on scaled f-scores aggregated into hand-coded categories. To exemplify this approach, we conducted a case study on the first wave of the coronavirus in Italy. We used our proposed methodology to explore existing psychological observations which claimed that physical distance from events affects what is communicated about them. We confirmed these findings by showing that the epicentre of the disease and peripheral regions correspond to clear time-series clusters and that those living in the epicentre of the SARS CoV2 outbreak were more focused on solidarity and policy than those from more peripheral regions. Furthermore, we also found that temporal categories corresponded closely to policy changes during the handling of the pandemic. | ['Galletti Martina', 'Anslow Michael'] | 2023-06-28 | null | null | null | null | ['time-series'] | ['time-series'] | [-1.69201240e-01 2.22633541e-01 -1.51320547e-01 -1.81459576e-01
-5.68571016e-02 -5.84726214e-01 8.93443167e-01 1.24302816e+00
-5.80667675e-01 3.92348677e-01 1.16648448e+00 -5.60842574e-01
-5.14228106e-01 -7.25928128e-01 -3.33669364e-01 -4.93282795e-01
-5.87264299e-01 1.56335726e-01 -2.52557188e-01 -4.86319244e-01
3.94597173e-01 4.30268586e-01 -1.08792102e+00 2.39708021e-01
6.86735511e-01 3.36090744e-01 7.97163025e-02 2.78136224e-01
-1.83742002e-01 6.06968045e-01 -7.81380594e-01 -1.01120420e-01
-2.12395266e-01 -3.28967571e-01 -6.25423610e-01 -1.66777864e-01
-3.99522960e-01 2.67373860e-01 1.13437913e-01 7.75831044e-01
3.76754552e-01 2.88018197e-01 7.60365129e-01 -1.05210364e+00
-5.60307741e-01 6.34480536e-01 -5.84120393e-01 7.23405182e-01
6.23131931e-01 -1.02315694e-01 6.40046120e-01 -3.87563258e-01
1.08655608e+00 1.36760545e+00 1.03249550e+00 -1.17676541e-01
-1.22980464e+00 -4.12431896e-01 2.57804006e-01 -1.58791214e-01
-1.30769622e+00 -3.52916479e-01 7.23980725e-01 -9.34526086e-01
8.31629276e-01 3.96003783e-01 7.99383402e-01 1.08361673e+00
9.75224078e-02 -2.02011943e-01 1.11370373e+00 -1.96564823e-01
1.57896340e-01 4.18652475e-01 1.48021787e-01 -3.37282084e-02
1.32003829e-01 -9.38870460e-02 -9.21430513e-02 -7.51536191e-01
1.75600164e-02 3.53276223e-01 -2.16157258e-01 3.24876398e-01
-1.44901311e+00 1.25982821e+00 5.94798565e-01 1.06427300e+00
-9.22564566e-01 -4.30898279e-01 5.71414530e-01 3.22597980e-01
1.06098878e+00 4.23645914e-01 -4.61684793e-01 -3.23047578e-01
-8.87664258e-01 5.80349658e-03 6.58929050e-01 7.43965060e-02
4.21359181e-01 -5.92126727e-01 -4.81322454e-03 5.23489773e-01
1.20465331e-01 4.45089936e-01 2.38601834e-01 -4.11407620e-01
3.30100447e-01 6.01287305e-01 1.44176364e-01 -1.82452166e+00
-9.45006132e-01 -3.55629116e-01 -6.61622465e-01 -4.03407991e-01
3.97263825e-01 -6.46272957e-01 -2.04974368e-01 1.79248548e+00
4.26923096e-01 -6.10370524e-02 -3.25041302e-02 5.23578703e-01
3.80654991e-01 7.65069425e-01 5.94304144e-01 -8.55806231e-01
1.53919041e+00 9.68162566e-02 -8.63630414e-01 2.17206627e-01
9.71238434e-01 -6.49014831e-01 8.02923560e-01 -1.31800234e-01
-8.15477014e-01 -3.64023775e-01 -4.41332459e-01 4.84844774e-01
-7.00465143e-01 -7.02114105e-01 2.15700448e-01 6.93072677e-01
-1.24947965e+00 4.63056743e-01 -3.29627991e-01 -1.37153316e+00
2.12408714e-02 -2.73501992e-01 -1.22280747e-01 6.40706122e-01
-1.40745068e+00 7.51563489e-01 4.10750479e-01 -2.96524316e-01
9.11806822e-02 -6.55548215e-01 -4.93139327e-01 -3.56926948e-01
7.23585486e-02 -3.71047378e-01 4.75492984e-01 -1.13783610e+00
-7.64841735e-01 1.02291465e+00 -1.99630812e-01 -4.51483637e-01
8.32886621e-02 3.39759767e-01 -1.17512691e+00 1.99712396e-01
2.93704927e-01 2.57064432e-01 3.48195851e-01 -1.01262820e+00
-5.21959603e-01 -5.38610101e-01 -1.16545735e-02 -1.75185278e-01
-5.94438493e-01 9.06412303e-01 5.02409279e-01 -8.81944835e-01
-2.86201507e-01 -7.03219116e-01 -3.34086657e-01 -8.26503277e-01
-7.12333098e-02 -4.36605960e-01 7.98974514e-01 -6.46401823e-01
1.71939003e+00 -2.30663061e+00 -1.32186756e-01 5.20072818e-01
3.85646969e-01 -1.95899665e-01 2.96323180e-01 1.19579935e+00
-3.06542832e-02 7.08889365e-01 1.10448778e-01 2.90419728e-01
-6.47339895e-02 -7.12824762e-02 -5.62892497e-01 8.08929503e-01
2.56078374e-02 7.26580977e-01 -1.23789716e+00 -3.09281707e-01
-9.42025930e-02 5.19412816e-01 -6.08124256e-01 -2.24341720e-01
9.46241394e-02 6.30288720e-01 -2.66330540e-01 1.57889605e-01
4.07248795e-01 -9.69086364e-02 3.05957586e-01 1.10683396e-01
-9.21881437e-01 3.05061072e-01 -3.03683013e-01 9.82378960e-01
-8.89845472e-03 8.91673446e-01 1.79877896e-02 -8.37556720e-01
8.93485904e-01 5.84370196e-01 6.42491996e-01 -7.54435897e-01
3.55731308e-01 -2.64991611e-01 1.51350170e-01 -8.60560060e-01
5.56830049e-01 -3.14041317e-01 -4.18778926e-01 7.94881642e-01
-6.55984640e-01 5.48440516e-02 -3.45469937e-02 1.24803960e-01
7.06385374e-01 -5.77792525e-01 6.92047954e-01 -7.81431139e-01
2.31795818e-01 1.91137090e-01 4.74534541e-01 3.57023865e-01
-4.51799899e-01 -3.14445398e-03 7.43292809e-01 -3.31488341e-01
-8.14492822e-01 -7.60511220e-01 -2.92519003e-01 1.16043258e+00
-6.25406429e-02 -4.89050686e-01 -7.35310316e-01 -2.35401392e-01
-2.34253541e-01 1.11354923e+00 -1.07162654e+00 1.22280873e-01
-3.44745219e-01 -9.04151499e-01 2.40725622e-01 -1.08923435e-01
2.62582332e-01 -1.15661144e+00 -1.04156017e+00 1.80229053e-01
-3.60703200e-01 -6.30409241e-01 -3.03781122e-01 -3.16524595e-01
-5.48502266e-01 -9.05857563e-01 -4.90093648e-01 -4.53631580e-01
6.70108318e-01 1.86507404e-01 8.42268169e-01 -7.84222931e-02
7.91077614e-02 4.21585739e-01 -6.34700298e-01 -8.22611630e-01
-4.76390421e-01 -1.08972251e-01 2.87465096e-01 6.11973330e-02
5.97673059e-01 -4.45358008e-01 -4.95334536e-01 1.62515432e-01
-9.71911967e-01 -3.26487452e-01 -1.99337319e-01 -3.69905680e-02
-2.39130795e-01 2.96602607e-01 9.10448253e-01 -6.84755385e-01
1.14968669e+00 -1.61481714e+00 7.39563704e-02 -1.25632405e-01
-2.68840402e-01 -7.46481299e-01 2.74901211e-01 -4.90834147e-01
-8.00560176e-01 -8.43528330e-01 2.51526237e-01 4.77683216e-01
-5.14072359e-01 1.06649220e+00 6.05983436e-01 5.88161767e-01
7.84851789e-01 -2.54260153e-01 1.21800564e-01 -5.64497560e-02
1.00136064e-01 9.60089326e-01 -6.09260648e-02 -1.53296068e-01
7.71054327e-01 7.83487201e-01 -5.99551976e-01 -1.44129276e+00
-3.72859806e-01 -6.34352624e-01 -4.63806301e-01 -6.74856424e-01
1.15416825e+00 -6.52635217e-01 -1.02363098e+00 1.05453655e-01
-9.35951114e-01 -3.69344562e-01 -1.24810807e-01 6.08070433e-01
-1.26149416e-01 1.08220927e-01 -2.28159755e-01 -9.84297037e-01
1.53093934e-01 -4.22604322e-01 4.48128998e-01 6.41672406e-03
-1.13273883e+00 -1.78358352e+00 6.81155264e-01 -4.84869583e-03
7.27227926e-01 9.15637136e-01 1.09222150e+00 -9.85595167e-01
5.38718462e-01 1.30155757e-01 -8.05815533e-02 -6.56536162e-01
6.14939451e-01 4.83077317e-02 -5.75293183e-01 -4.38376755e-01
4.39757586e-01 1.24559380e-01 2.13687718e-01 3.37943256e-01
4.29497033e-01 -7.03174770e-01 -6.60125434e-01 -6.24640770e-02
1.05220068e+00 6.06970131e-01 2.51141906e-01 5.49327374e-01
1.92429841e-01 1.45945895e+00 5.05119860e-01 6.46880627e-01
4.97690499e-01 7.32800424e-01 -6.58535287e-02 -3.09524953e-01
5.92490315e-01 -4.52775173e-02 6.69533491e-01 7.56665051e-01
-4.18718979e-02 -2.50757992e-01 -1.29896057e+00 9.55927968e-01
-1.63797021e+00 -1.32528174e+00 -3.78548801e-01 2.06662893e+00
3.92412275e-01 -1.21158719e-01 7.74953783e-01 -1.02106459e-01
8.25845063e-01 4.50020462e-01 1.48735255e-01 -6.82686210e-01
-1.09267727e-01 -4.50570554e-01 2.23653719e-01 5.64278543e-01
-7.54429281e-01 3.93218786e-01 6.82047224e+00 2.86051661e-01
-1.06790006e+00 2.28359327e-01 8.07829082e-01 -4.57581952e-02
-6.06096148e-01 -2.16180772e-01 -7.17372373e-02 5.64740658e-01
1.43071949e+00 -3.75490248e-01 1.13096602e-01 8.48487690e-02
1.02316236e+00 -1.83549389e-01 -4.58155483e-01 3.64191264e-01
4.76683974e-02 -1.01356006e+00 -3.60752583e-01 3.84138346e-01
6.93003356e-01 8.75229202e-03 8.25771615e-02 -8.67642909e-02
1.22204207e-01 -6.51715636e-01 7.33300507e-01 4.68227178e-01
5.34369111e-01 -7.41255105e-01 4.74061072e-01 2.95513511e-01
-9.17754710e-01 -4.01152708e-02 3.54819074e-02 -4.79517579e-01
5.03026605e-01 7.62674987e-01 -1.06886744e+00 2.49651909e-01
6.75759733e-01 7.85876691e-01 -3.59147251e-01 5.68707466e-01
2.61571735e-01 8.42052639e-01 -8.31217319e-02 -1.54452026e-01
4.54594761e-01 -4.75612700e-01 9.14124668e-01 1.46996462e+00
4.34530646e-01 1.92844048e-01 -2.14830071e-01 5.48905551e-01
4.32235807e-01 3.33782673e-01 -1.12827647e+00 -5.09264767e-01
3.34545642e-01 9.32777822e-01 -1.31216788e+00 -3.47943664e-01
-4.11222965e-01 3.48547786e-01 1.87406436e-01 4.95810658e-01
-7.68936872e-01 -3.06816161e-01 6.81881487e-01 6.76372707e-01
6.71964213e-02 -1.80239975e-01 -5.06942980e-02 -7.09225297e-01
-5.08624911e-01 -6.05388999e-01 4.03194636e-01 -3.84528846e-01
-1.15838051e+00 6.68039083e-01 4.59756464e-01 -8.05393219e-01
-1.31390318e-01 8.55356380e-02 -8.95265460e-01 5.72583556e-01
-9.01700497e-01 -6.67142987e-01 6.94399104e-02 5.42367816e-01
3.72422859e-02 2.13613018e-01 9.18198049e-01 5.09907193e-02
-3.97319943e-01 -2.46525723e-02 1.91339388e-01 -2.20375285e-01
4.71809417e-01 -8.63961160e-01 3.20107788e-01 5.62669873e-01
-3.82941216e-01 9.07895744e-01 9.39522326e-01 -1.15464640e+00
-4.78445292e-01 -9.61519539e-01 1.74043071e+00 -2.85275668e-01
1.31851542e+00 -4.95961249e-01 -6.57916009e-01 6.35034621e-01
6.65542305e-01 -8.61234784e-01 1.09839058e+00 3.36447865e-01
3.87006253e-02 5.07933080e-01 -1.43157554e+00 1.00022995e+00
1.08944225e+00 -6.98200822e-01 -9.66606438e-01 5.93347669e-01
8.90853763e-01 4.98142868e-01 -9.85895038e-01 2.23071165e-02
2.86912769e-01 -1.02510297e+00 7.30472028e-01 -6.94934070e-01
-1.03209238e-03 -7.47744814e-02 -3.26458551e-02 -1.45317352e+00
-6.76821411e-01 -1.06717563e+00 7.35695601e-01 1.37888587e+00
2.41818205e-01 -1.14737451e+00 -4.87886779e-02 2.57707536e-01
1.84187338e-01 -1.63489908e-01 -6.90258145e-01 -2.99420744e-01
1.27948597e-01 -3.67893994e-01 6.22202814e-01 1.63988459e+00
5.87240338e-01 -1.80070996e-01 -1.31237116e-02 6.43445775e-02
8.80003870e-02 -1.20564371e-01 3.98281902e-01 -1.31100154e+00
1.13234818e-01 -6.34106576e-01 -3.44386011e-01 5.95746785e-02
-4.09192145e-02 -6.45674706e-01 -4.48992282e-01 -1.21957338e+00
1.24233417e-01 -3.46589118e-01 -1.47997871e-01 -1.04164749e-01
5.63380755e-02 1.03121974e-01 3.08665216e-01 2.13602975e-01
-1.83824673e-01 -2.31584590e-02 6.27984345e-01 1.68835625e-01
-7.89341569e-01 -2.21844494e-01 -9.51848686e-01 7.63810813e-01
1.04182839e+00 -4.71803695e-01 -3.84916246e-01 1.41354755e-01
9.50376391e-01 4.78206277e-02 3.45846564e-01 -7.23130524e-01
5.64348325e-02 -5.78743815e-01 -1.48346230e-01 -5.01479268e-01
-1.57069325e-01 -1.07652545e+00 7.16712117e-01 8.78893495e-01
-4.10648972e-01 5.02227128e-01 3.68609548e-01 3.68952900e-01
-1.27641603e-01 2.34608635e-01 2.42431462e-01 2.56768703e-01
-1.22674584e-01 -1.72452003e-01 -1.11975372e+00 2.01262712e-01
1.46888089e+00 -3.56688768e-01 -3.29753578e-01 -5.75817764e-01
-8.38839531e-01 3.34385522e-02 6.33069456e-01 5.24842620e-01
2.13951156e-01 -1.20089555e+00 -8.65376711e-01 -8.44721645e-02
-3.07427812e-03 -8.27639699e-01 2.36417979e-01 1.55419719e+00
-3.28459024e-01 5.61286449e-01 -1.79116905e-01 -1.95995465e-01
-1.11779428e+00 9.60689723e-01 1.92615669e-02 1.95742566e-02
-4.19731766e-01 3.02516341e-01 3.73454154e-01 -3.68550867e-01
-5.17631471e-02 -4.87559915e-01 -7.74737418e-01 1.09415197e+00
6.30860686e-01 6.58674359e-01 -3.30589801e-01 -1.28040636e+00
-6.00509524e-01 6.17041588e-01 3.78035694e-01 -1.68422148e-01
1.40211880e+00 -5.45535207e-01 -4.27500218e-01 9.44157362e-01
1.48760295e+00 4.39488649e-01 -4.92453933e-01 -1.52854193e-02
4.42324698e-01 -1.71002433e-01 -1.67171359e-01 -5.30338764e-01
-5.53405821e-01 4.15021062e-01 3.54325682e-01 1.22569382e+00
1.00881636e+00 1.69274092e-01 6.27447605e-01 -1.48860216e-01
1.82870001e-01 -1.00589848e+00 -4.26494837e-01 4.23863560e-01
1.09820211e+00 -6.84593737e-01 -3.87585253e-01 -1.59146503e-01
-6.42079234e-01 9.22236204e-01 -4.77835983e-01 1.56536758e-01
1.02980995e+00 -1.75967827e-01 -2.13621417e-04 -7.93107688e-01
-4.29033458e-01 -1.12102993e-01 1.55528396e-01 6.47291541e-01
6.56301379e-01 5.08054674e-01 -7.99262285e-01 1.77804932e-01
-2.70535290e-01 -2.53694206e-01 5.05610406e-01 6.93055212e-01
-3.25061828e-01 -4.93696123e-01 -7.05690145e-01 2.69623190e-01
-6.49986804e-01 -7.92061538e-02 -8.62264812e-01 9.73166883e-01
3.47176194e-01 1.28132045e+00 7.70953655e-01 -5.17203510e-01
1.41211122e-01 4.67526652e-02 -2.82212168e-01 -4.22768950e-01
-1.15356886e+00 6.98438555e-04 3.62433583e-01 -4.06611979e-01
-8.66799533e-01 -1.22597349e+00 -1.20602000e+00 -7.13953733e-01
5.68149202e-02 5.93262970e-01 5.39021909e-01 8.91357481e-01
5.74254930e-01 4.59808826e-01 7.62324870e-01 -6.31380379e-01
4.03468519e-01 -9.80381787e-01 -4.83179480e-01 6.48170769e-01
4.82769459e-01 -2.80252397e-01 -6.40120506e-01 -6.78163543e-02] | [8.567516326904297, 9.780698776245117] |
b9eb5e17-cc1a-4040-86f1-379babb4a1f4 | learning-object-bounding-boxes-for-3d | 1906.01140 | null | https://arxiv.org/abs/1906.01140v2 | https://arxiv.org/pdf/1906.01140v2.pdf | Learning Object Bounding Boxes for 3D Instance Segmentation on Point Clouds | We propose a novel, conceptually simple and general framework for instance segmentation on 3D point clouds. Our method, called 3D-BoNet, follows the simple design philosophy of per-point multilayer perceptrons (MLPs). The framework directly regresses 3D bounding boxes for all instances in a point cloud, while simultaneously predicting a point-level mask for each instance. It consists of a backbone network followed by two parallel network branches for 1) bounding box regression and 2) point mask prediction. 3D-BoNet is single-stage, anchor-free and end-to-end trainable. Moreover, it is remarkably computationally efficient as, unlike existing approaches, it does not require any post-processing steps such as non-maximum suppression, feature sampling, clustering or voting. Extensive experiments show that our approach surpasses existing work on both ScanNet and S3DIS datasets while being approximately 10x more computationally efficient. Comprehensive ablation studies demonstrate the effectiveness of our design. | ['Bo Yang', 'Sen Wang', 'Ronald Clark', 'Andrew Markham', 'Niki Trigoni', 'Qingyong Hu', 'Jianan Wang'] | 2019-06-04 | learning-object-bounding-boxes-for-3d-1 | http://papers.nips.cc/paper/8899-learning-object-bounding-boxes-for-3d-instance-segmentation-on-point-clouds | http://papers.nips.cc/paper/8899-learning-object-bounding-boxes-for-3d-instance-segmentation-on-point-clouds.pdf | neurips-2019-12 | ['3d-instance-segmentation-1'] | ['computer-vision'] | [ 2.64250338e-01 2.40036771e-01 -1.65428579e-01 -6.08581185e-01
-7.11801648e-01 -3.29100728e-01 6.81229770e-01 2.95039177e-01
-4.12702948e-01 1.39141336e-01 -3.70069265e-01 -7.27434635e-01
1.68813542e-02 -8.56168687e-01 -1.20362687e+00 -2.50223547e-01
-3.02162290e-01 9.06640053e-01 5.75445473e-01 1.15122609e-01
4.01429981e-01 1.33047593e+00 -1.58005118e+00 6.38743788e-02
3.62263471e-01 1.43676078e+00 -5.56109194e-03 6.78362668e-01
-3.72518212e-01 3.41721475e-01 -2.06780627e-01 -1.60826638e-01
6.12255037e-01 6.18269920e-01 -7.58968174e-01 1.07331149e-01
7.37043381e-01 -2.97843486e-01 -9.12177861e-02 6.63692176e-01
2.50386059e-01 1.17540672e-01 7.39571095e-01 -1.04361665e+00
-3.97388190e-02 2.89035022e-01 -9.71371591e-01 -1.02226481e-01
-1.35801703e-01 1.21023245e-01 9.19366717e-01 -1.27210975e+00
1.98603690e-01 1.25170040e+00 1.19828010e+00 2.31093749e-01
-1.31940460e+00 -4.66840953e-01 2.35848367e-01 -3.80507857e-01
-1.19768870e+00 -2.78972298e-01 7.27994502e-01 -5.79009354e-01
1.41298091e+00 6.04087532e-01 6.68669105e-01 4.50587660e-01
-9.48887169e-02 8.17328453e-01 8.14972818e-01 -4.94538188e-01
3.52337450e-01 -2.19073117e-01 2.01569751e-01 7.64247596e-01
3.82059196e-04 6.02976680e-02 -1.04139253e-01 -3.55190217e-01
1.27426410e+00 2.18592167e-01 1.29346550e-01 -7.25693107e-01
-1.25884247e+00 6.59625769e-01 8.04414690e-01 -1.71880811e-01
-4.25846279e-01 4.54070210e-01 1.40965790e-01 9.70088467e-02
8.84646952e-01 1.57109693e-01 -7.36471713e-01 2.20361799e-01
-1.29574955e+00 3.75783980e-01 6.97731853e-01 1.16359150e+00
1.04360354e+00 -2.35972568e-01 2.86327656e-02 6.95658028e-01
5.56430101e-01 3.06419522e-01 -6.89476207e-02 -9.46633041e-01
4.86894995e-01 8.69476020e-01 2.32434407e-01 -7.66896605e-01
-6.66898966e-01 -2.51599580e-01 -6.06964886e-01 7.26981163e-01
-4.96994667e-02 -1.75355710e-02 -1.44023311e+00 9.35156047e-01
6.15873456e-01 4.64407355e-01 -4.98791426e-01 7.20206261e-01
9.79296505e-01 6.25444353e-01 5.51562868e-02 5.21951735e-01
1.11148822e+00 -9.71943080e-01 3.20391595e-01 -5.12434304e-01
2.90091842e-01 -5.59865654e-01 7.21866608e-01 1.20202705e-01
-1.45805275e+00 -6.78653896e-01 -1.12531435e+00 -2.18973517e-01
-4.23136562e-01 -8.13054666e-02 8.78789008e-01 5.49395621e-01
-1.10330093e+00 8.96195471e-01 -1.23271513e+00 -1.35234902e-02
9.45520759e-01 8.00556123e-01 -2.22917140e-01 3.98021996e-01
-2.82580882e-01 8.32197189e-01 4.13871199e-01 -4.64773364e-02
-5.55582166e-01 -1.06421983e+00 -8.48006904e-01 6.33242801e-02
1.32613271e-01 -8.09553206e-01 1.66887724e+00 -5.86991072e-01
-1.51324427e+00 1.18193793e+00 -4.38459158e-01 -5.61312258e-01
6.68274462e-01 -3.88225853e-01 1.39018968e-01 -1.39649108e-01
1.42345309e-01 1.27188218e+00 8.61989975e-01 -1.40123558e+00
-1.02688456e+00 -3.52120787e-01 -1.08366273e-01 1.35563761e-01
2.70943373e-01 -1.57138973e-01 -8.45785022e-01 -2.85479069e-01
7.00433016e-01 -7.00240850e-01 -7.87136674e-01 3.23281288e-01
-8.01760018e-01 -4.03982490e-01 7.19438732e-01 -2.99242049e-01
5.85705757e-01 -2.12803984e+00 -4.08631414e-01 6.03505969e-01
4.09401059e-01 -1.01146296e-01 2.52854049e-01 2.84540683e-01
-1.41670585e-01 2.82462686e-01 -6.10554874e-01 -6.58523500e-01
1.40540123e-01 1.41162097e-01 -4.01890635e-01 6.29871130e-01
5.61736524e-01 8.51830900e-01 -5.51568449e-01 -5.13353527e-01
7.27747917e-01 4.94289637e-01 -5.55797637e-01 -4.70285229e-02
-5.15335798e-01 2.33593881e-01 -3.78817618e-01 9.93389666e-01
9.58854616e-01 -4.74547893e-01 -4.85802412e-01 -9.86455567e-03
-4.08911675e-01 3.97231162e-01 -1.09730947e+00 1.77577782e+00
-3.99232388e-01 5.46372533e-01 1.31242827e-01 -8.72106135e-01
1.03948390e+00 1.80679299e-02 6.29595757e-01 -3.78105789e-01
7.10264668e-02 3.26756954e-01 -5.58845401e-01 -1.56445712e-01
4.75252301e-01 -5.17357104e-02 2.10205447e-02 8.29429179e-02
-1.13556040e-02 -3.92619371e-01 -2.15396568e-01 -4.01860982e-01
1.21318305e+00 3.62643898e-01 1.54842421e-01 -7.85013512e-02
2.76272595e-01 3.43522459e-01 2.38996670e-01 9.87361372e-01
-3.72961573e-02 7.76047111e-01 2.40840837e-01 -7.45046496e-01
-1.19683993e+00 -1.26499140e+00 -4.11050320e-01 8.83277416e-01
3.60061735e-01 -1.29165441e-01 -5.74477673e-01 -6.59913957e-01
4.99244601e-01 6.39045119e-01 -3.63542676e-01 5.62824667e-01
-8.14373076e-01 -3.10463399e-01 3.96804035e-01 7.39273310e-01
3.95047218e-01 -1.03865921e+00 -8.40605021e-01 3.07274401e-01
5.28636992e-01 -9.95283186e-01 4.97074425e-02 6.75196230e-01
-1.47030592e+00 -7.24025428e-01 -5.08810997e-01 -9.15334284e-01
7.93263316e-01 2.13889197e-01 1.66274345e+00 5.98172694e-02
-2.01446816e-01 1.58040613e-01 8.80687833e-02 -8.70575249e-01
2.81378757e-02 1.95000589e-01 -3.01084995e-01 -4.13102567e-01
6.61014736e-01 -8.64917994e-01 -7.09336400e-01 1.73812523e-01
-4.64189678e-01 1.37485027e-01 9.88042176e-01 4.04515564e-01
1.16177619e+00 -1.68983132e-01 -1.46059529e-03 -7.35764861e-01
1.71271473e-01 -3.90712708e-01 -8.19400907e-01 -2.31247529e-01
-2.24163145e-01 -3.08937103e-01 1.93940729e-01 4.23836187e-02
-5.30522883e-01 5.97241640e-01 -5.41560054e-01 -6.39622808e-01
-7.73904622e-01 1.57043606e-01 2.44010329e-01 -4.67679828e-01
4.72454816e-01 6.18267916e-02 -1.76612362e-01 -8.33260834e-01
4.70769078e-01 6.44278228e-01 6.40514672e-01 -3.94564897e-01
1.17666161e+00 9.85426545e-01 1.17270052e-01 -1.06978083e+00
-5.08545578e-01 -6.21399820e-01 -1.15671563e+00 -3.78739238e-02
6.75623894e-01 -8.84864390e-01 -8.47914517e-01 1.25624359e-01
-1.40681171e+00 -2.12047994e-01 -4.22644585e-01 2.15477407e-01
-6.05034947e-01 1.13308430e-01 -5.99402547e-01 -8.99826109e-01
-6.52294517e-01 -9.52934384e-01 1.47895622e+00 4.73553054e-02
-2.50369042e-01 -7.08978415e-01 -1.43342186e-02 1.97439402e-01
6.30914643e-02 5.12958407e-01 9.37282979e-01 -6.83451116e-01
-9.45005476e-01 -5.41320622e-01 -3.92834008e-01 1.99775234e-01
-2.80516356e-01 1.16421014e-01 -1.03168845e+00 -1.01951130e-01
-2.74777234e-01 -9.24120545e-02 8.01843941e-01 6.10061169e-01
1.62309170e+00 -1.63081706e-01 -7.60916710e-01 8.20051432e-01
1.49152827e+00 -9.45046637e-03 2.63338894e-01 4.31589872e-01
8.01375568e-01 3.03173810e-01 3.49786997e-01 1.73462436e-01
3.51225764e-01 3.02427024e-01 8.78899813e-01 -6.43389821e-01
2.71205992e-01 -1.36879608e-01 -1.93115234e-01 3.68197262e-01
-1.26945302e-01 8.01917836e-02 -1.17907667e+00 7.07950056e-01
-1.76127756e+00 -5.04234731e-01 -3.34971964e-01 1.96296024e+00
3.64559770e-01 5.27960479e-01 5.13443761e-02 1.85304880e-01
4.62464958e-01 1.51605129e-01 -6.46537900e-01 -4.39772129e-01
3.59279156e-01 6.11804843e-01 9.07606781e-01 2.97181755e-01
-1.58977342e+00 1.00900149e+00 6.97482491e+00 6.72300935e-01
-9.03352380e-01 -1.21048130e-01 6.71615064e-01 -2.59153306e-01
-6.36014622e-03 -4.20054421e-02 -7.00567484e-01 2.53118545e-01
5.66206276e-01 6.39146030e-01 1.77378431e-02 1.33056366e+00
6.32265396e-03 5.21592200e-02 -1.30007756e+00 8.92804921e-01
-4.74637628e-01 -1.71835983e+00 -1.51440084e-01 -8.02529790e-03
3.67736310e-01 9.79523599e-01 -1.98795572e-01 2.54608423e-01
4.11284238e-01 -1.16993165e+00 9.76081789e-01 3.18260610e-01
6.94486439e-01 -9.27002668e-01 3.61922055e-01 5.43109357e-01
-1.00356793e+00 1.82092980e-01 -4.18333054e-01 -1.35303020e-01
1.32976994e-01 8.19846213e-01 -1.18192554e+00 3.84547561e-01
1.05418742e+00 4.16182786e-01 -3.74874979e-01 1.33051252e+00
-6.16011396e-02 6.39153421e-01 -9.43323135e-01 1.92854758e-02
6.82877719e-01 -9.52345133e-02 6.46690845e-01 1.38234973e+00
1.16443157e-01 -1.76523522e-01 3.87680233e-01 1.00830007e+00
-8.15725103e-02 -1.63956508e-01 -5.80032527e-01 5.23308933e-01
7.56622493e-01 1.13849568e+00 -1.10053873e+00 -2.14004382e-01
-5.29701591e-01 6.36136830e-01 3.46112162e-01 1.88013166e-01
-6.65665269e-01 -3.95902246e-01 6.92975760e-01 2.44539693e-01
8.14030707e-01 -4.04569924e-01 -9.99593973e-01 -4.50468808e-01
1.79630697e-01 -2.37385005e-01 3.77561827e-03 -6.24070644e-01
-1.42974484e+00 5.34590542e-01 -8.29833299e-02 -1.29959321e+00
1.17538869e-02 -8.41574788e-01 -7.93158352e-01 8.63441765e-01
-1.71865737e+00 -1.39161563e+00 -2.65302539e-01 4.04381841e-01
5.61370373e-01 1.33701473e-01 5.95207930e-01 2.00047359e-01
-4.18871313e-01 1.41996756e-01 -3.45001556e-02 2.10355848e-01
-5.96800484e-02 -1.46351790e+00 1.30789649e+00 6.06833518e-01
2.29144856e-01 3.80556494e-01 4.55878019e-01 -7.23641992e-01
-1.20779967e+00 -1.21035492e+00 5.88326752e-01 -6.98004425e-01
4.79128152e-01 -4.89572406e-01 -8.30606759e-01 9.93382573e-01
-3.42911988e-01 1.52127206e-01 4.33682531e-01 3.00796539e-01
5.23959845e-02 2.42918152e-02 -1.13712692e+00 5.96963346e-01
1.12592447e+00 -1.14795588e-01 -6.67310536e-01 3.50571305e-01
7.52458572e-01 -9.34502840e-01 -9.44071770e-01 8.49456489e-01
3.13317031e-01 -1.03378069e+00 1.40955627e+00 -4.74401414e-01
4.07886416e-01 -3.42025220e-01 -5.63410074e-02 -8.49947155e-01
-5.21921575e-01 -5.39496601e-01 -4.39642608e-01 6.73833728e-01
6.47884369e-01 -3.95629972e-01 1.38196480e+00 5.85581779e-01
-6.67811275e-01 -1.19564617e+00 -1.19209898e+00 -5.77372670e-01
1.49983168e-01 -1.01943183e+00 9.37053561e-01 6.42172635e-01
-5.96291959e-01 2.81671993e-02 1.41094521e-01 7.02843308e-01
8.52777362e-01 3.65978390e-01 1.06036687e+00 -1.57760310e+00
-4.56570685e-02 -6.64359570e-01 -3.99037570e-01 -1.54953694e+00
-1.73548356e-01 -8.47279787e-01 1.54371619e-01 -1.71714687e+00
-4.85003442e-01 -8.43788028e-01 -1.20999187e-01 7.03720033e-01
1.72132865e-01 3.49369526e-01 -1.47601157e-01 5.00713050e-01
-4.70858425e-01 2.34070107e-01 8.84164929e-01 -1.60606608e-01
-4.46203351e-01 4.12232965e-01 -3.69282603e-01 1.14985275e+00
7.45271325e-01 -4.87141371e-01 -2.64063496e-02 -6.34396732e-01
3.47279273e-02 -9.20403749e-02 6.43563747e-01 -1.27467823e+00
2.93975264e-01 -5.82687818e-02 8.36768508e-01 -1.55921018e+00
6.53506815e-01 -1.06945741e+00 -1.57366127e-01 2.16039181e-01
2.28162944e-01 -5.84470071e-02 4.33647066e-01 4.99536037e-01
1.18240044e-01 -2.19736919e-01 7.09540904e-01 -3.73119175e-01
-7.47793376e-01 6.08172774e-01 1.42158596e-02 -5.55831850e-01
1.08056056e+00 -8.10935795e-01 2.05328032e-01 1.81409150e-01
-5.19535542e-01 4.50764477e-01 4.38441813e-01 2.09830970e-01
6.22005999e-01 -1.01482177e+00 -6.87844455e-01 3.67180407e-01
-3.45561579e-02 1.02609420e+00 4.23252732e-02 5.64853251e-01
-9.59800899e-01 4.49403912e-01 2.40928695e-01 -1.30578601e+00
-9.17642236e-01 3.28782231e-01 3.65866363e-01 -2.22721156e-02
-1.26712763e+00 1.20185864e+00 -1.38119325e-01 -7.67632186e-01
3.98375154e-01 -7.02600062e-01 9.41973850e-02 -3.91414851e-01
1.80520892e-01 2.23588988e-01 4.42214698e-01 -3.41496527e-01
-4.08714265e-01 7.04443097e-01 -1.06750891e-01 -1.08437791e-01
1.67231333e+00 3.30809444e-01 -1.09573841e-01 4.92475271e-01
8.60010803e-01 -2.70904541e-01 -1.55700648e+00 -1.83243871e-01
3.36727113e-01 -3.22158992e-01 3.12991202e-01 -6.97233200e-01
-9.04470563e-01 6.96030974e-01 3.59127134e-01 4.34015870e-01
7.64159381e-01 1.88583791e-01 8.56383443e-01 5.69379747e-01
2.99820065e-01 -7.65938520e-01 -5.12108028e-01 5.63001633e-01
7.95955479e-01 -1.17742157e+00 1.91843584e-01 -4.78130162e-01
1.61409099e-02 1.05748117e+00 6.07704639e-01 -5.99721193e-01
8.94477129e-01 4.00209367e-01 7.17004165e-02 -6.29846036e-01
-4.62460309e-01 -5.56397326e-02 4.40999568e-01 4.69475836e-01
8.99601653e-02 -5.69324978e-02 4.39677864e-01 1.77647233e-01
-4.75817621e-01 -1.45893499e-01 -1.23199202e-01 1.15653849e+00
-6.13186657e-01 -6.56039953e-01 -6.16396844e-01 7.61152267e-01
-1.60286024e-01 3.82639281e-02 -3.76844317e-01 1.06414640e+00
9.56981182e-02 5.26127040e-01 7.16885507e-01 -4.38898087e-01
5.63275337e-01 8.29083622e-02 2.53803402e-01 -6.22933686e-01
-6.59839094e-01 4.54762578e-02 -1.33597776e-01 -6.70998454e-01
-2.64347166e-01 -4.51944202e-01 -1.30353367e+00 -3.51821452e-01
-3.33163232e-01 -2.15340510e-01 1.02707613e+00 8.34538460e-01
5.92631996e-01 3.54288518e-01 6.50254130e-01 -1.76789939e+00
-2.80906975e-01 -9.42975223e-01 -2.73867190e-01 -9.30100456e-02
6.17047727e-01 -5.24018884e-01 -2.31315032e-01 -1.84706315e-01] | [7.941258907318115, -3.4245734214782715] |
4712df51-768c-4c25-b5a6-6b9bd21bc747 | interpretable-multimodal-emotion-recognition | 2208.11868 | null | https://arxiv.org/abs/2208.11868v2 | https://arxiv.org/pdf/2208.11868v2.pdf | Interpretable Multimodal Emotion Recognition using Hybrid Fusion of Speech and Image Data | This paper proposes a multimodal emotion recognition system based on hybrid fusion that classifies the emotions depicted by speech utterances and corresponding images into discrete classes. A new interpretability technique has been developed to identify the important speech & image features leading to the prediction of particular emotion classes. The proposed system's architecture has been determined through intensive ablation studies. It fuses the speech & image features and then combines speech, image, and intermediate fusion outputs. The proposed interpretability technique incorporates the divide & conquer approach to compute shapely values denoting each speech & image feature's importance. We have also constructed a large-scale dataset (IIT-R SIER dataset), consisting of speech utterances, corresponding images, and class labels, i.e., 'anger,' 'happy,' 'hate,' and 'sad.' The proposed system has achieved 83.29% accuracy for emotion recognition. The enhanced performance of the proposed system advocates the importance of utilizing complementary information from multiple modalities for emotion recognition. | ['Balasubramanian Raman', 'Sarthak Malik', 'Puneet Kumar'] | 2022-08-25 | null | null | null | null | ['multimodal-emotion-recognition', 'multimodal-emotion-recognition'] | ['computer-vision', 'speech'] | [ 4.93885756e-01 1.86528698e-01 2.61093318e-01 -7.86021292e-01
-9.41580415e-01 -3.14998418e-01 7.48269737e-01 1.36192232e-01
-2.75423348e-01 5.35310447e-01 4.61937219e-01 4.91401851e-02
-1.40693039e-01 6.20405525e-02 1.30207941e-01 -8.60561371e-01
4.66726273e-02 1.03306167e-01 -4.72408503e-01 -2.59444922e-01
5.34956753e-01 5.04392087e-01 -1.89717185e+00 7.52679825e-01
6.99337840e-01 1.45601630e+00 4.15400080e-02 9.32683051e-01
-2.60839492e-01 1.06950915e+00 -8.27305317e-01 -2.66067028e-01
-2.74121553e-01 -5.57276130e-01 -7.52429366e-01 5.19387662e-01
-4.45312075e-03 6.50490448e-03 -3.42638977e-02 9.59973454e-01
5.61995924e-01 2.48553380e-01 9.34399247e-01 -1.56395817e+00
-5.90453804e-01 1.06423423e-01 -6.13380492e-01 2.75664866e-01
6.16390467e-01 -2.49669552e-01 6.30766213e-01 -1.03263569e+00
4.32167113e-01 1.33201611e+00 2.09197551e-01 4.45672899e-01
-6.19229615e-01 -5.19633651e-01 -7.29150474e-02 2.92179734e-01
-1.45918953e+00 -7.78205693e-01 9.62678492e-01 -3.93927485e-01
1.13175440e+00 6.29191101e-01 4.13315564e-01 7.77520955e-01
1.87435806e-01 1.04056776e+00 1.46803427e+00 -5.97086966e-01
1.24565549e-01 8.07405949e-01 2.89505303e-01 8.14343691e-01
-7.43445277e-01 -7.21573532e-02 -6.30939305e-01 -2.09372804e-01
3.07112224e-02 -1.70272693e-01 -9.22622252e-03 3.22468460e-01
-8.27642798e-01 5.68385601e-01 1.82095245e-01 4.16707247e-01
-5.25615871e-01 -2.92684793e-01 5.35237968e-01 3.25711995e-01
6.44707382e-01 -1.73382908e-01 -2.12575734e-01 -2.83274710e-01
-6.36597574e-01 -4.38907295e-01 4.60407376e-01 6.26916051e-01
4.63913113e-01 9.11196470e-02 1.23933017e-01 1.21560800e+00
6.40283227e-01 5.49016595e-01 5.24980545e-01 -8.59132290e-01
2.18611822e-01 7.49157846e-01 -3.24686915e-02 -1.28605628e+00
-4.50170219e-01 5.78989834e-02 -8.05177689e-01 3.34689438e-01
-4.11284775e-01 -2.30573416e-01 -9.29580331e-01 1.56735122e+00
1.38917878e-01 -2.50343263e-01 9.08074081e-01 8.55870306e-01
1.26422179e+00 8.69538367e-01 3.48288983e-01 -4.10221785e-01
1.71214724e+00 -7.84281015e-01 -1.01753604e+00 5.39541580e-02
2.15109661e-01 -9.73371267e-01 6.03913963e-01 7.03688145e-01
-9.00677204e-01 -6.69637084e-01 -9.07454312e-01 2.17258155e-01
-6.85197353e-01 7.42411852e-01 5.95203280e-01 7.88218558e-01
-1.07585955e+00 -1.63734511e-01 -1.26819611e-01 -4.72809285e-01
2.05200568e-01 6.18959546e-01 -6.65196836e-01 4.46877927e-01
-1.05008793e+00 1.02280760e+00 2.42936254e-01 3.66382718e-01
-6.95251167e-01 2.49253839e-01 -7.87612200e-01 -2.70148758e-02
-1.91804647e-01 -2.52361774e-01 8.98569524e-01 -1.75561404e+00
-1.51777661e+00 9.18354094e-01 -5.35576701e-01 8.39760974e-02
-1.47748023e-01 1.77692428e-01 -1.02767253e+00 6.79136634e-01
-2.98646420e-01 9.28140163e-01 1.10710883e+00 -1.40545118e+00
-8.61510515e-01 -5.64438462e-01 -2.83635318e-01 7.26787806e-01
-3.42713952e-01 5.73738933e-01 -6.32325113e-02 -4.63604063e-01
4.11529988e-01 -6.95518613e-01 1.77320123e-01 -3.79611164e-01
-1.40979633e-01 -3.70479763e-01 1.25070953e+00 -7.20343471e-01
1.06095064e+00 -2.42145872e+00 1.58846185e-01 3.93830150e-01
1.23237237e-01 1.51120231e-01 -2.89157573e-02 3.25110435e-01
-4.94421840e-01 9.90746841e-02 -5.19425310e-02 -4.42434520e-01
1.04550389e-03 2.20955417e-01 -2.55861739e-03 2.35528260e-01
4.56640631e-01 4.61860538e-01 -4.80092347e-01 -6.99012816e-01
5.38451433e-01 7.57379770e-01 -8.19856003e-02 3.28096867e-01
4.60040927e-01 3.65355194e-01 -5.00880599e-01 1.00242865e+00
4.61750686e-01 2.00118870e-01 6.65787086e-02 -5.18278956e-01
-2.88258586e-03 -5.36911845e-01 -9.93399262e-01 1.16071188e+00
-3.36748034e-01 7.30276763e-01 2.79600829e-01 -1.03535390e+00
1.41332507e+00 9.01873231e-01 2.90185064e-01 -5.13026416e-01
5.42844355e-01 1.86274260e-01 -1.92309707e-01 -1.05310678e+00
5.58337688e-01 -4.15191829e-01 -4.09212738e-01 2.81494081e-01
2.68765420e-01 2.81235166e-02 -2.69048244e-01 1.54392421e-01
7.05003858e-01 -1.85306862e-01 3.55171382e-01 -3.47909145e-02
1.08881140e+00 -1.61555186e-01 1.95163205e-01 3.51050168e-01
-7.26456821e-01 2.13876635e-01 4.73350585e-01 -2.61097372e-01
-6.56821191e-01 -9.47113156e-01 6.52354658e-02 1.12468600e+00
1.14740342e-01 1.06720991e-01 -5.42095184e-01 -6.03794217e-01
-5.10809124e-01 5.13394058e-01 -7.97459126e-01 -1.93572462e-01
2.03051567e-01 -6.67891204e-01 7.07410514e-01 3.34663063e-01
6.85897112e-01 -1.19924128e+00 -4.76642549e-01 -1.51819557e-01
-3.14788997e-01 -8.14876795e-01 -2.62902509e-02 1.96233705e-01
-3.34732503e-01 -6.65166914e-01 -3.91452909e-01 -9.49177086e-01
9.76083159e-01 -1.71289537e-02 6.01484179e-01 -1.22953907e-01
-2.84576535e-01 7.92210400e-01 -8.01164567e-01 -4.57524061e-01
-4.10183966e-01 -4.91706729e-01 1.69896603e-01 5.64440608e-01
2.45422080e-01 -3.44382524e-01 -2.94245899e-01 1.77315235e-01
-8.64052057e-01 1.18647786e-02 7.40071058e-01 8.56552958e-01
2.67831832e-01 4.22639191e-01 9.44418728e-01 -3.85997951e-01
1.07041943e+00 -5.19172728e-01 1.94999397e-01 4.86980230e-01
-2.09286392e-01 -2.55311579e-01 2.28899285e-01 -1.96980625e-01
-1.63986218e+00 2.47561947e-01 -2.82216191e-01 -2.86202282e-01
-8.16058815e-01 4.69046324e-01 -1.60759449e-01 -1.92978069e-01
1.78722993e-01 2.58794278e-01 1.95618898e-01 3.70329176e-03
2.58490860e-01 1.42751408e+00 6.33120000e-01 -4.73600805e-01
-2.27623899e-02 2.42215902e-01 -3.98293018e-01 -1.14383245e+00
-3.00285757e-01 -5.03131866e-01 -3.80994111e-01 -8.26940179e-01
1.05580127e+00 -9.94338393e-01 -9.65085983e-01 4.81433839e-01
-1.06098509e+00 6.81291401e-01 2.24780604e-01 5.49457252e-01
-3.67049962e-01 3.92231852e-01 -5.41009545e-01 -1.67150640e+00
-5.90245783e-01 -1.32549405e+00 1.18080986e+00 5.45643508e-01
-4.76982534e-01 -6.80118740e-01 -3.58185768e-01 7.01391220e-01
1.85196027e-01 1.46973684e-01 7.62236834e-01 -7.49606252e-01
2.02730685e-01 -3.67436409e-01 -4.43906039e-01 5.95644414e-01
1.75648287e-01 3.13019753e-01 -1.16632485e+00 8.25009793e-02
8.15294757e-02 -7.11797178e-01 5.62309623e-01 1.12759270e-01
7.22273290e-01 -9.98404324e-02 -1.51592651e-02 -3.46416212e-03
1.07848477e+00 1.11046517e+00 5.87808967e-01 -2.24181086e-01
2.80752540e-01 8.40107679e-01 7.98408329e-01 5.07079184e-01
2.41953298e-01 2.82939851e-01 2.18233019e-01 -2.89923668e-01
3.57993305e-01 3.82758915e-01 6.44455791e-01 1.20046401e+00
-2.20551454e-02 -2.82583714e-01 -6.36972845e-01 2.34113559e-01
-1.66730690e+00 -1.01272988e+00 2.48594105e-01 1.36514747e+00
2.28170201e-01 -1.86094075e-01 -6.77622557e-02 3.32926333e-01
7.51643181e-01 5.58339655e-02 -2.47856885e-01 -1.15627277e+00
-2.74853975e-01 1.01596445e-01 -3.49901140e-01 6.26132727e-01
-1.25903881e+00 7.43513882e-01 6.25452709e+00 8.24850857e-01
-1.23791134e+00 -6.31610006e-02 1.01760161e+00 2.53469020e-01
-1.72191903e-01 -3.63153160e-01 -2.42525086e-01 2.40783855e-01
1.05496931e+00 1.25225587e-02 3.02400082e-01 7.65157580e-01
3.32058042e-01 -5.28129637e-01 -3.80821705e-01 1.12229514e+00
7.02920139e-01 -8.15913677e-01 8.25572684e-02 -3.62801015e-01
1.77140579e-01 -3.83110166e-01 2.39989609e-01 2.33387843e-01
-1.40351310e-01 -9.55193102e-01 6.01403654e-01 7.34718144e-01
7.07557201e-01 -1.00422049e+00 8.74504209e-01 2.12661803e-01
-1.11942291e+00 -1.85205892e-01 1.62465319e-01 -1.13731008e-02
-1.10443220e-01 -6.01542108e-02 -9.02652502e-01 6.06353164e-01
6.16267502e-01 4.40527648e-01 -5.54537475e-01 1.87734708e-01
1.37948856e-01 2.72397935e-01 -3.42083991e-01 -2.74765521e-01
1.83218986e-01 -2.49955505e-01 2.97753185e-01 1.35414934e+00
4.07959342e-01 6.35848463e-01 -1.18748762e-01 3.12247425e-01
3.00243169e-01 4.65811580e-01 -5.74489832e-01 -3.26204658e-01
2.85314202e-01 1.60909569e+00 -9.43204761e-01 -6.85679376e-01
-2.99778402e-01 9.75107789e-01 -8.02835822e-02 2.89512992e-01
-7.75328040e-01 -4.49126720e-01 3.25265735e-01 -8.83333683e-01
3.02837510e-02 2.00993389e-01 -1.47290036e-01 -8.11872303e-01
-2.91544944e-01 -8.91368389e-01 5.03623843e-01 -1.43630302e+00
-9.48231339e-01 1.25690496e+00 -1.00367710e-01 -1.08955336e+00
-2.65456259e-01 -6.49324834e-01 -5.87712526e-01 8.34102035e-01
-8.35943878e-01 -1.25493491e+00 -1.73857912e-01 7.01721787e-01
6.54926002e-01 -5.65782845e-01 1.14745975e+00 1.85928643e-01
-8.15554559e-01 4.32761878e-01 -2.18438238e-01 -1.09038107e-01
3.91541749e-01 -9.82257664e-01 -7.09697962e-01 5.04949510e-01
-1.23763934e-01 3.45135957e-01 7.38646567e-01 -3.21911961e-01
-1.10888875e+00 -4.01633143e-01 8.85810077e-01 -1.26370803e-01
3.63187253e-01 9.84609574e-02 -4.95364130e-01 3.25562567e-01
7.79559970e-01 -3.82417738e-01 1.19161725e+00 5.62308962e-03
-1.29419491e-01 -2.04718739e-01 -1.53077269e+00 3.02444935e-01
1.10218681e-01 -7.09612608e-01 -7.33536422e-01 -6.63526282e-02
2.72720426e-01 -1.37514090e-02 -1.11998236e+00 4.22902465e-01
6.76748872e-01 -1.04701245e+00 7.64405310e-01 -3.84405375e-01
5.32825887e-01 -2.62480050e-01 -5.10781944e-01 -1.10875571e+00
5.65452315e-02 -2.21741423e-01 3.54414850e-01 1.48499739e+00
5.40525138e-01 -5.51560700e-01 4.73654807e-01 6.77348495e-01
-1.88929141e-01 -7.83026814e-01 -1.07593822e+00 1.83873400e-01
-7.48441815e-01 -4.55245823e-01 2.55633295e-01 9.88303781e-01
4.74854320e-01 5.10747492e-01 -5.61264455e-01 2.73910880e-01
2.47521147e-01 -7.53044058e-03 2.06846029e-01 -7.71069050e-01
2.02207774e-01 -4.15962696e-01 -5.80023587e-01 -2.63442039e-01
2.74156600e-01 -7.07900047e-01 -9.01749507e-02 -1.07152355e+00
2.78889209e-01 -7.45786801e-02 -4.97813404e-01 6.10080838e-01
-1.71587430e-02 3.60246539e-01 1.21149816e-01 8.63789115e-03
-6.56334996e-01 5.74090838e-01 8.93927276e-01 -7.16011822e-02
-6.98300973e-02 -3.22267741e-01 -6.60836935e-01 7.31966317e-01
9.88544166e-01 2.65447851e-02 -6.27460778e-01 2.45153010e-02
-2.61265486e-01 6.83509648e-01 2.12447390e-01 -7.68330038e-01
1.43127337e-01 -3.41772027e-02 5.80373108e-01 -7.03971028e-01
9.10388470e-01 -9.71436858e-01 2.39342034e-01 1.85035154e-01
-5.19317806e-01 2.64534563e-01 2.41524488e-01 3.08839709e-01
-7.32254446e-01 -7.55607709e-02 7.08207607e-01 6.17087483e-02
-1.24975586e+00 -2.80252546e-01 -7.66977310e-01 -7.63555229e-01
1.12145877e+00 -3.57966572e-01 -2.41875663e-01 -6.63862884e-01
-1.39667845e+00 3.28399893e-03 -2.39255369e-01 5.49614191e-01
1.29492140e+00 -1.38764048e+00 -4.38292950e-01 3.23870391e-01
2.59537846e-01 -8.74826252e-01 6.88230336e-01 9.32750821e-01
-2.73373693e-01 1.55351073e-01 -5.49931586e-01 -4.95592922e-01
-2.04193091e+00 2.57715106e-01 1.43014058e-01 1.20731540e-01
-7.18432060e-03 7.96410859e-01 -2.26573013e-02 -3.22754800e-01
2.00198025e-01 8.34429562e-02 -7.25343287e-01 4.55311537e-01
4.14065033e-01 9.90152881e-02 -1.72212660e-01 -1.40902305e+00
-5.06538332e-01 4.66987729e-01 3.14107514e-03 -4.43531424e-01
9.65958297e-01 -5.98723650e-01 -3.40558946e-01 5.42118132e-01
1.42513311e+00 -3.28612387e-01 -4.66694564e-01 4.87128757e-02
-1.78051278e-01 -2.04463944e-01 1.15004718e-01 -1.30067706e+00
-1.04334760e+00 9.11520958e-01 1.16108358e+00 2.54209012e-01
1.84776676e+00 -4.74295957e-04 2.90227294e-01 7.41325393e-02
5.67110330e-02 -1.27999675e+00 1.18600130e-01 2.07373142e-01
1.06881428e+00 -1.41954017e+00 -2.14220166e-01 -3.99786234e-01
-1.14922726e+00 1.17454720e+00 4.26778793e-01 2.95279503e-01
6.61029220e-01 1.94374517e-01 5.13173401e-01 -3.72909755e-01
-1.01706314e+00 -1.66170612e-01 3.72032523e-01 3.92700344e-01
4.57696527e-01 3.26172173e-01 -2.56132990e-01 7.40851045e-01
1.22444242e-01 -1.41684949e-01 1.39084548e-01 8.40051055e-01
-6.13986671e-01 -7.59261250e-01 -6.81200862e-01 4.14370209e-01
-2.66054153e-01 7.48158768e-02 -7.75039017e-01 5.60308397e-01
3.47022533e-01 1.25371635e+00 -4.37223576e-02 -9.18999791e-01
1.10925786e-01 3.23511958e-01 2.35917732e-01 -1.11676171e-01
-6.27791762e-01 3.53365332e-01 4.13140208e-01 -1.93451673e-01
-9.04182971e-01 -6.23286009e-01 -1.19906509e+00 -8.87933150e-02
-1.29158631e-01 5.11198938e-01 8.26803088e-01 9.90395606e-01
3.61896574e-01 5.85157096e-01 9.15199995e-01 -7.92055905e-01
3.34450789e-03 -9.12858486e-01 -5.59818864e-01 5.21675944e-01
1.84347659e-01 -6.27270937e-01 -5.33568621e-01 2.38434285e-01] | [13.25274658203125, 5.2092413902282715] |
87a0062b-8940-425c-b2bd-8cc517c9345d | reclor-a-reading-comprehension-dataset-1 | 2002.04326 | null | https://arxiv.org/abs/2002.04326v3 | https://arxiv.org/pdf/2002.04326v3.pdf | ReClor: A Reading Comprehension Dataset Requiring Logical Reasoning | Recent powerful pre-trained language models have achieved remarkable performance on most of the popular datasets for reading comprehension. It is time to introduce more challenging datasets to push the development of this field towards more comprehensive reasoning of text. In this paper, we introduce a new Reading Comprehension dataset requiring logical reasoning (ReClor) extracted from standardized graduate admission examinations. As earlier studies suggest, human-annotated datasets usually contain biases, which are often exploited by models to achieve high accuracy without truly understanding the text. In order to comprehensively evaluate the logical reasoning ability of models on ReClor, we propose to identify biased data points and separate them into EASY set while the rest as HARD set. Empirical results show that state-of-the-art models have an outstanding ability to capture biases contained in the dataset with high accuracy on EASY set. However, they struggle on HARD set with poor performance near that of random guess, indicating more research is needed to essentially enhance the logical reasoning ability of current models. | ['Zi-Hang Jiang', 'Yanfei Dong', 'Weihao Yu', 'Jiashi Feng'] | 2020-02-11 | null | https://openreview.net/forum?id=HJgJtT4tvB | https://openreview.net/pdf?id=HJgJtT4tvB | iclr-2020-1 | ['logical-reasoning-question-ansering', 'logical-reasoning-reading-comprehension'] | ['natural-language-processing', 'natural-language-processing'] | [ 8.84790868e-02 5.14203548e-01 -3.44583809e-01 -6.73176944e-01
-5.60281754e-01 -3.53935242e-01 3.37660074e-01 7.18390465e-01
-3.45196635e-01 6.59999490e-01 3.64070714e-01 -8.15878212e-01
-3.31342518e-01 -9.43227530e-01 -5.28415978e-01 -1.10256724e-01
6.35733604e-01 4.26919490e-01 1.80827811e-01 -5.86637795e-01
7.83648312e-01 1.38717353e-01 -1.41546905e+00 4.47988600e-01
1.43067706e+00 6.88055336e-01 -5.41624576e-02 6.28153324e-01
-3.89875203e-01 1.54794085e+00 -5.24880409e-01 -8.31775725e-01
-2.85454988e-01 -6.07162118e-01 -1.25556159e+00 -5.18197894e-01
6.91506922e-01 -2.89693654e-01 -1.74799398e-01 1.22825801e+00
6.51045144e-01 1.85790285e-02 7.38118827e-01 -9.05605793e-01
-1.14619696e+00 1.09690845e+00 -3.87446433e-01 4.19495881e-01
7.77663946e-01 -1.11757606e-01 1.01144218e+00 -7.00010180e-01
2.55493730e-01 1.35576892e+00 5.97294211e-01 6.34483397e-01
-7.98964381e-01 -8.60765457e-01 3.22336078e-01 4.67489451e-01
-1.14823663e+00 -2.63756305e-01 1.02341974e+00 -4.71315205e-01
6.07375324e-01 2.74502337e-01 3.87465477e-01 1.32536840e+00
3.20134789e-01 1.09134185e+00 1.80270839e+00 -6.17262840e-01
1.87806383e-01 1.97463855e-01 1.29816353e+00 7.15920687e-01
6.45246267e-01 -4.06627893e-01 -6.69813156e-01 1.71663940e-01
2.99711734e-01 -1.10139556e-01 -4.43050981e-01 -1.15249589e-01
-9.37044322e-01 5.95094323e-01 2.75794566e-01 1.30699024e-01
-1.28291592e-01 -5.41990340e-01 -7.65494704e-02 6.31043077e-01
1.50596589e-01 5.69698989e-01 -5.27189195e-01 -3.41433764e-01
-8.53580832e-01 5.57925224e-01 9.05060053e-01 9.77918684e-01
1.34667456e-01 -3.08584452e-01 -5.21349132e-01 9.48456585e-01
2.77355850e-01 6.67866528e-01 7.81042397e-01 -4.85090226e-01
8.16508591e-01 1.23486137e+00 -2.30726257e-01 -1.16585612e+00
-5.81632137e-01 -6.96908116e-01 -9.03008759e-01 -1.29509956e-01
7.27504671e-01 1.95002586e-01 -6.76259160e-01 1.65300274e+00
-2.48242378e-01 -1.76913962e-01 2.60970861e-01 6.38809621e-01
1.21019900e+00 5.71768992e-02 2.47657537e-01 1.58237442e-01
1.58764350e+00 -8.91057909e-01 -8.25544178e-01 -5.24692059e-01
8.08220506e-01 -5.94568491e-01 1.42466211e+00 7.67421186e-01
-1.50001395e+00 -6.46912992e-01 -1.22519016e+00 -3.92245680e-01
-4.51089621e-01 7.16581792e-02 5.59991121e-01 9.40607667e-01
-6.71281993e-01 1.79427519e-01 -4.58011985e-01 -1.41890459e-02
7.95108676e-01 -4.57202829e-02 -5.15666418e-02 -6.24516308e-01
-1.34041953e+00 9.18334186e-01 2.36565948e-01 -6.86711892e-02
-7.89596200e-01 -8.84482443e-01 -8.27041388e-01 3.28780860e-01
3.51872593e-01 -5.77621937e-01 1.25011647e+00 -6.19498551e-01
-1.27604234e+00 1.15290558e+00 -9.08808038e-02 -2.53463775e-01
7.84621716e-01 -5.36515415e-01 -2.65633434e-01 -1.31450981e-01
-1.08205555e-02 2.86133707e-01 5.41855633e-01 -1.08102155e+00
-2.85005242e-01 -4.50885296e-01 2.25439221e-01 2.64132060e-02
-5.82916021e-01 -2.52056748e-01 4.76496406e-02 -6.43932998e-01
4.05207276e-01 -6.98159873e-01 4.02398199e-01 -4.96682405e-01
-4.89171028e-01 -7.48587191e-01 8.00834298e-02 -6.07528865e-01
1.63752186e+00 -1.67340493e+00 7.46493340e-02 -1.43700177e-02
5.15390098e-01 4.69780892e-01 -7.82875940e-02 1.51261657e-01
-2.66280502e-01 5.18490791e-01 1.16616458e-01 -9.99838188e-02
1.98098570e-01 -1.35243580e-01 -5.96092343e-01 -3.58137153e-02
1.62650749e-01 9.35889781e-01 -1.14344430e+00 -7.01626360e-01
-2.48888001e-01 -1.77476525e-01 -7.06829250e-01 4.07200962e-01
-9.58573893e-02 9.22283679e-02 -7.80877292e-01 7.90813863e-01
8.22402894e-01 -1.75385892e-01 -5.64310290e-02 3.27088714e-01
3.43426079e-01 6.73741579e-01 -8.66275370e-01 1.43654430e+00
-1.14557646e-01 5.10050952e-01 -3.32208246e-01 -9.58637238e-01
1.07679713e+00 -5.36119156e-02 -2.54644394e-01 -9.91201937e-01
3.02442789e-01 9.89743471e-02 5.59878290e-01 -8.67248595e-01
5.09175658e-01 -2.13786185e-01 -1.30783260e-01 3.94652694e-01
-2.85672039e-01 -1.65596828e-01 2.11653276e-03 2.41586655e-01
1.07314301e+00 -1.17206082e-01 1.21267036e-01 -5.19075274e-01
9.99343336e-01 -1.06546491e-01 2.77770638e-01 1.11204278e+00
-2.95054823e-01 4.74494040e-01 6.82501078e-01 -2.23947346e-01
-4.44162697e-01 -9.97612357e-01 -2.71031648e-01 1.18924463e+00
1.55743375e-01 -4.32388783e-01 -7.85323024e-01 -8.00085783e-01
-8.17351937e-02 1.12218416e+00 -5.20688236e-01 -5.80238700e-01
-4.25243944e-01 -6.76637769e-01 9.42252994e-01 7.22165823e-01
1.11863577e+00 -7.44545937e-01 -4.32989478e-01 -2.88002193e-02
-4.28192526e-01 -1.08324456e+00 4.09526855e-01 -7.83994980e-03
-8.98503780e-01 -1.15908039e+00 -3.50889802e-01 -8.16896677e-01
7.23086894e-01 1.87807873e-01 1.61804760e+00 6.48533106e-01
1.01173535e-01 2.51169801e-01 -3.79076153e-01 -1.05999994e+00
-3.03787351e-01 3.71690363e-01 -1.69684201e-01 -6.13320172e-01
1.03145051e+00 -5.32107390e-02 -2.84135550e-01 1.66403562e-01
-7.62810469e-01 1.58588886e-01 6.15223110e-01 7.60984480e-01
3.50372456e-02 3.00799012e-01 6.92858517e-01 -8.40380371e-01
1.21356738e+00 -5.91579020e-01 -4.06849645e-02 4.85284239e-01
-8.93000901e-01 1.78877577e-01 5.02626657e-01 -4.03395146e-01
-1.21943367e+00 -1.03391373e+00 -2.34265044e-01 2.27270409e-01
-1.30158037e-01 7.09218860e-01 -5.37777925e-03 1.91966116e-01
7.66809464e-01 2.67936528e-01 -3.74772042e-01 -4.55013514e-01
-2.23186567e-01 8.23807597e-01 3.89445901e-01 -1.17979765e+00
4.91065800e-01 -1.41866729e-02 -7.63420612e-02 -6.09072804e-01
-1.72623134e+00 -8.39076266e-02 -5.16365170e-01 -1.27302796e-01
5.96303463e-01 -8.72062743e-01 -7.49621391e-01 6.98053181e-01
-7.14131176e-01 -4.44335550e-01 2.09101006e-01 4.30385530e-01
-1.47073343e-01 2.42009372e-01 -4.68724430e-01 -8.92470300e-01
-5.63743599e-02 -1.01874733e+00 7.46263325e-01 4.88383263e-01
-5.88746369e-01 -1.19272232e+00 -3.39912586e-02 1.30867648e+00
4.51661229e-01 -2.56165087e-01 1.42272568e+00 -1.05102015e+00
-3.90690923e-01 -3.21657449e-01 -2.13554054e-01 3.21573406e-01
-2.88690716e-01 -3.05900663e-01 -1.02719009e+00 2.02849554e-03
2.04996914e-01 -8.97436917e-01 9.20868099e-01 4.50040959e-02
1.61454809e+00 1.89262405e-01 -9.01703686e-02 1.96126789e-01
1.05811346e+00 -2.86794543e-01 7.90484428e-01 5.50036669e-01
6.38885856e-01 4.70999390e-01 6.31060958e-01 5.41687794e-02
9.82510507e-01 -7.99479857e-02 1.65906921e-01 3.91788304e-01
-1.56031594e-01 -5.67819655e-01 2.79522032e-01 8.63676906e-01
-8.25385824e-02 -4.77506906e-01 -1.50322104e+00 5.05495310e-01
-1.51187861e+00 -7.89197266e-01 -4.05982941e-01 1.97955358e+00
1.16974771e+00 5.50263941e-01 -5.96893013e-01 4.53036934e-01
3.35777514e-02 6.06732331e-02 -3.48923922e-01 -6.44279182e-01
-8.96222293e-02 2.88637459e-01 2.44880803e-02 5.06596923e-01
-6.92598760e-01 9.11040783e-01 6.38111830e+00 5.73539138e-01
-8.16800475e-01 -2.33578756e-01 8.31912994e-01 2.67109126e-01
-6.73288882e-01 -3.42929304e-01 -1.12113273e+00 2.14107499e-01
1.02241850e+00 1.67107835e-01 5.66505035e-03 7.48727083e-01
5.01629189e-02 -3.99847984e-01 -1.23586035e+00 9.16303813e-01
5.12782812e-01 -6.73043430e-01 2.28825077e-01 -2.71373063e-01
9.20376241e-01 -5.86915970e-01 1.22234091e-01 1.07737827e+00
4.30562675e-01 -1.65102375e+00 3.85433048e-01 9.35065329e-01
5.73477289e-03 -4.90195483e-01 1.21292686e+00 9.20807183e-01
-1.53312847e-01 -2.51105666e-01 -5.61120808e-01 -7.97914743e-01
-5.86644113e-01 5.12209833e-01 -6.27381206e-01 3.05296123e-01
6.94672704e-01 5.71758270e-01 -1.38520062e+00 6.92451775e-01
-5.83734214e-01 9.69241858e-01 8.42727646e-02 -4.51347381e-01
5.30336164e-02 8.63965675e-02 4.55934368e-02 9.89395082e-01
8.59859809e-02 3.03134173e-01 1.30660638e-01 9.13741350e-01
-3.18802357e-01 1.85906440e-01 -3.88782561e-01 -2.04671130e-01
3.91790271e-01 8.01212668e-01 -4.38446313e-01 -5.43916762e-01
-3.56783599e-01 4.82956499e-01 7.38817871e-01 3.44193876e-02
-7.42879510e-01 -1.47157624e-01 3.76219511e-01 3.35358009e-02
-3.74317318e-01 -1.55009687e-01 -1.08549082e+00 -1.37828600e+00
5.66823147e-02 -1.29597902e+00 5.67748785e-01 -1.11343050e+00
-1.39123881e+00 6.99517801e-02 1.53827608e-01 -6.46326125e-01
1.28295258e-01 -9.18457210e-01 -4.69756544e-01 9.20221865e-01
-1.83683133e+00 -9.46920335e-01 -1.06750822e+00 4.68893260e-01
8.03978026e-01 -2.71371186e-01 6.60796523e-01 -3.43833715e-02
-6.97605014e-01 9.13333952e-01 -3.28956872e-01 5.75469255e-01
1.02195621e+00 -1.40878081e+00 -2.30455279e-01 7.41875887e-01
-3.10738891e-01 1.03483832e+00 6.77329123e-01 -6.48557246e-01
-1.32463992e+00 -4.95811701e-01 1.27047420e+00 -1.15340769e+00
5.72372675e-01 -3.37529965e-02 -1.25856805e+00 8.57911289e-01
3.96431923e-01 -6.48512423e-01 9.70005870e-01 3.59610438e-01
-4.73343372e-01 1.04034968e-01 -9.61116970e-01 9.71517801e-01
9.33920920e-01 -3.80409509e-01 -1.54049492e+00 1.14399530e-01
1.39590561e-01 -6.50222421e-01 -8.89997780e-01 5.08806229e-01
3.71208370e-01 -1.10074174e+00 9.66170311e-01 -1.00428784e+00
1.12160933e+00 2.13085502e-01 1.07412726e-01 -1.22984338e+00
-2.43682817e-01 -3.35606262e-02 -2.73432750e-02 1.24850297e+00
2.91231275e-01 -5.47039807e-01 7.39997387e-01 8.89859140e-01
-1.30568333e-02 -1.01500690e+00 -3.04969043e-01 -2.95510948e-01
8.68063450e-01 -5.62313616e-01 6.45228922e-01 1.22564089e+00
3.16671640e-01 5.41579723e-01 2.32516617e-01 1.37547210e-01
5.63830853e-01 9.97096300e-02 6.39344990e-01 -1.53751063e+00
1.64175570e-01 -7.68841565e-01 -1.52182475e-01 -1.32939112e+00
5.74360669e-01 -9.30841029e-01 -2.87915200e-01 -1.50029123e+00
5.57412386e-01 -4.35648739e-01 -3.20594221e-01 5.57468653e-01
-8.30420613e-01 -9.31971297e-02 -2.24835292e-01 -2.29704991e-01
-5.82385898e-01 4.67989892e-01 1.61152577e+00 -2.48156101e-01
1.70148998e-01 -1.35462791e-01 -1.34579504e+00 1.05675030e+00
1.00017786e+00 -1.69909254e-01 -8.67312729e-01 -5.81507504e-01
7.13382483e-01 -2.02375457e-01 3.84294838e-01 -1.01871693e+00
2.30630472e-01 -2.70202786e-01 6.45660162e-01 -5.47695816e-01
-1.29163638e-01 -4.97931004e-01 -8.27025592e-01 2.24294052e-01
-1.02979505e+00 2.57181257e-01 1.21633910e-01 2.80023843e-01
-2.86254764e-01 -6.76138937e-01 6.42631054e-01 -2.15905786e-01
-5.01071095e-01 -3.43759477e-01 -1.87515765e-01 6.86242759e-01
6.57236993e-01 1.15535289e-01 -7.12692201e-01 -4.32526648e-01
-5.14758170e-01 5.23865342e-01 2.22412229e-01 7.80643344e-01
6.82695866e-01 -1.02061009e+00 -8.87710035e-01 2.81519532e-01
1.30413324e-01 3.26840162e-01 3.01396936e-01 9.75170910e-01
-5.99902213e-01 6.95377707e-01 -2.80145645e-01 -4.18793976e-01
-1.43596423e+00 4.05966818e-01 3.04598421e-01 -5.88836968e-01
-3.17000985e-01 9.00403321e-01 -2.63158739e-01 -6.17685556e-01
2.60161400e-01 -7.26002038e-01 -6.16385221e-01 -2.47081998e-03
8.36126804e-01 4.05061990e-01 1.75391376e-01 -1.17640875e-01
-1.20594315e-01 4.79211450e-01 -2.62840420e-01 4.53873098e-01
1.09817493e+00 -5.77380555e-03 -2.20742300e-01 6.16523564e-01
6.31539226e-01 2.54226148e-01 -5.07131875e-01 -1.38561442e-01
1.49431854e-01 -2.81304389e-01 2.44651929e-01 -1.25406265e+00
-4.60369349e-01 1.21940529e+00 3.44990522e-01 1.18750513e-01
8.27666998e-01 -2.93776393e-01 3.98717701e-01 6.73549771e-01
2.62944490e-01 -1.04421413e+00 2.92541593e-01 1.01399219e+00
6.79547489e-01 -1.60395479e+00 1.73207477e-01 -4.36772585e-01
-5.56942940e-01 1.20697796e+00 1.02572155e+00 2.42086016e-02
4.81641620e-01 3.74607109e-02 1.40374303e-01 -2.74079949e-01
-6.61371708e-01 1.05588824e-01 4.73321468e-01 2.24087432e-01
9.73298430e-01 3.24662365e-02 -4.50889736e-01 1.18825150e+00
-8.13935399e-01 6.50123879e-02 6.98488653e-01 9.15661275e-01
-5.20768702e-01 -9.18710947e-01 -5.68081141e-01 5.23407459e-01
-5.09899914e-01 -3.68628383e-01 -6.44555390e-01 7.90750563e-01
-1.18362427e-01 1.14519179e+00 -2.06270322e-01 -1.91166043e-01
4.77971435e-01 7.68394828e-01 5.99542975e-01 -4.45342630e-01
-6.76642179e-01 -8.93951595e-01 1.86706066e-01 -2.81397820e-01
-1.28696725e-01 -4.64416921e-01 -1.16357970e+00 -4.14326906e-01
2.35356260e-02 -4.60464926e-03 2.11936966e-01 1.29129469e+00
5.49017964e-03 8.94342244e-01 -1.59971431e-01 1.86508801e-02
-1.10006618e+00 -1.34646809e+00 -1.85522556e-01 6.73656344e-01
2.34590486e-01 -6.84035301e-01 -3.90169472e-01 -2.32653007e-01] | [10.076277732849121, 7.659755229949951] |
8c8c9ce6-85a6-404d-b160-f8993a5e8c4e | reconstructing-vehicles-from-orthographic | 2206.08789 | null | https://arxiv.org/abs/2206.08789v1 | https://arxiv.org/pdf/2206.08789v1.pdf | Reconstructing vehicles from orthographic drawings using deep neural networks | This paper explores the current state-of-the-art of object reconstruction from multiple orthographic drawings using deep neural networks. It proposes two algorithms to extract multiple views from a single image. The paper proposes a system based on pixel-aligned implicit functions (PIFu) and develops an advanced sampling strategy to generate signed distance samples. It also compares this approach to depth map regression from multiple views. Additionally, the paper uses a novel dataset for vehicle reconstruction from the racing game Assetto Corsa, which features higher quality models than the commonly used ShapeNET dataset. The trained neural network generalizes well to real-world inputs and creates plausible and detailed reconstructions. | ['Robin Klippert'] | 2022-06-14 | null | null | null | null | ['object-reconstruction'] | ['computer-vision'] | [ 1.23319559e-01 2.87156254e-01 -8.96882149e-04 -4.41365629e-01
-4.24779147e-01 -2.10939810e-01 6.66377008e-01 -5.42648077e-01
-2.49360114e-01 8.19705427e-01 7.01184273e-02 -6.10886253e-02
-2.44158342e-01 -1.20284200e+00 -1.22275662e+00 -1.67363420e-01
3.65276635e-01 9.63460743e-01 3.34197819e-01 -3.99611980e-01
6.24275804e-01 1.24783349e+00 -1.70105946e+00 5.80124378e-01
2.05133289e-01 1.07590878e+00 3.65090854e-02 7.15669394e-01
-6.26743808e-02 7.73555517e-01 -4.62472588e-01 -6.60178840e-01
7.47388303e-01 -3.09652332e-02 -5.20505786e-01 1.72189832e-01
1.17684352e+00 -1.01772726e+00 -6.45080209e-01 4.75498348e-01
2.27128953e-01 -5.81361279e-02 9.36713219e-01 -1.21774364e+00
-6.56942546e-01 1.35746300e-01 -4.50684488e-01 -1.49179041e-01
1.06109969e-01 1.42511696e-01 4.23849314e-01 -1.13884163e+00
1.11334372e+00 1.41805959e+00 1.05497003e+00 5.21577656e-01
-1.10284686e+00 -6.44227445e-01 -3.03883612e-01 2.17992306e-01
-1.10323620e+00 -2.85137564e-01 9.79036331e-01 -3.90698403e-01
1.24404943e+00 -1.29467383e-01 1.02520394e+00 1.08251417e+00
3.25543731e-01 7.26111293e-01 1.12122357e+00 -4.54342037e-01
1.43007427e-01 5.98368701e-03 -2.82478124e-01 8.56002510e-01
3.68845612e-01 7.64845729e-01 -5.23766994e-01 1.41048893e-01
1.55413544e+00 -9.70649049e-02 8.89475923e-03 -9.83142555e-01
-8.15916836e-01 9.22455549e-01 3.37128133e-01 -2.33489305e-01
-3.88041228e-01 6.01118982e-01 3.05088848e-01 -1.42339841e-02
3.21320325e-01 2.05003530e-01 -3.16702962e-01 -9.55888107e-02
-1.13488388e+00 5.70508838e-01 6.24513507e-01 1.14875138e+00
7.83559144e-01 6.37608051e-01 7.52585769e-01 5.34371853e-01
2.76301980e-01 4.51640248e-01 -3.23911272e-02 -1.61486256e+00
4.01354969e-01 3.69153798e-01 3.63837630e-02 -1.08814561e+00
-2.98948079e-01 -2.02435330e-01 -3.77831906e-01 1.22539651e+00
3.57314259e-01 1.71837270e-01 -9.96112287e-01 9.84365761e-01
6.68144822e-02 -2.57730037e-01 4.20263000e-02 8.48768115e-01
1.19898951e+00 4.45769548e-01 -6.64567828e-01 4.27790046e-01
5.99033475e-01 -1.06731319e+00 -5.03549516e-01 -1.80507258e-01
-6.33220077e-02 -3.85945171e-01 5.23720622e-01 1.01455986e+00
-1.62822962e+00 -6.89816654e-01 -1.54204965e+00 -2.69252926e-01
-3.61342311e-01 2.15897053e-01 5.16282499e-01 4.46602821e-01
-1.07197976e+00 9.04369116e-01 -5.14855981e-01 -1.71976179e-01
6.17550552e-01 3.18290859e-01 -8.17601621e-01 -2.18452029e-02
-5.81817746e-01 1.29940999e+00 3.80778044e-01 3.33801359e-02
-1.06061924e+00 -6.73370838e-01 -9.05058920e-01 -4.01437432e-01
9.40701813e-02 -7.27409124e-01 1.09868801e+00 -1.22415137e+00
-1.99343216e+00 9.02721465e-01 2.40388438e-01 -5.92585385e-01
9.12458658e-01 -3.53585750e-01 -5.86520694e-02 5.17120361e-01
-2.63201684e-01 1.20316863e+00 9.12304461e-01 -1.77211690e+00
-3.61265570e-01 -3.66259068e-01 3.22221279e-01 3.55837941e-02
4.89765584e-01 -4.75741059e-01 -4.71512554e-03 -5.63434184e-01
2.97766328e-01 -5.52490234e-01 -1.29267707e-01 5.47198057e-01
3.69800739e-02 2.26767957e-01 7.89028347e-01 -9.18575525e-01
3.25060248e-01 -1.62284756e+00 2.29488552e-01 2.24546060e-01
6.72279596e-02 1.83302939e-01 -1.66314095e-01 4.42249030e-01
-1.07883550e-01 -2.98720747e-01 -2.04668343e-01 -3.20052356e-01
-6.46017566e-02 4.50663149e-01 -2.75169641e-01 6.94041491e-01
9.66213867e-02 9.92235661e-01 -4.25044030e-01 -1.88118130e-01
5.05540311e-01 4.82886881e-01 -4.55468208e-01 -1.62337288e-01
-3.14645827e-01 9.31239799e-02 1.58601090e-01 5.73130965e-01
1.22591662e+00 5.35186887e-01 -3.04098558e-02 -3.21490705e-01
-1.50283352e-01 -2.26524070e-01 -1.19796073e+00 1.90038514e+00
-5.16132295e-01 1.02691042e+00 -4.69430350e-02 -1.09028828e+00
1.38109112e+00 5.00547178e-02 3.46139222e-01 -7.29374349e-01
2.71998137e-01 3.16342235e-01 -4.89068329e-01 -5.11186600e-01
9.16891992e-01 -3.07988197e-01 3.30008924e-01 1.32983550e-01
2.77598739e-01 -7.32722640e-01 -9.97220501e-02 -1.00363426e-01
7.29827523e-01 1.10907078e+00 2.81929016e-01 -1.47793397e-01
3.90280902e-01 3.26075524e-01 3.09159845e-01 1.96794063e-01
3.11192330e-02 1.20810521e+00 1.83192253e-01 -1.16404402e+00
-1.74051559e+00 -1.38645887e+00 -7.30121359e-02 1.84591085e-01
1.88922226e-01 3.07608843e-02 -7.22029030e-01 -6.30416334e-01
1.93730310e-01 9.17818427e-01 -8.51599216e-01 2.99140483e-01
-8.85605872e-01 1.75615922e-01 7.37669289e-01 1.03671217e+00
6.65027916e-01 -1.15201938e+00 -1.00901866e+00 4.18559685e-02
7.30409995e-02 -1.06536615e+00 2.69494355e-01 -1.32700922e-02
-1.22236979e+00 -1.16500187e+00 -5.92812359e-01 -6.18263304e-01
6.77618086e-01 1.90962046e-01 1.40900278e+00 -2.28966594e-01
-2.06208855e-01 6.67815149e-01 -5.13726585e-02 -5.10367572e-01
-5.48646033e-01 -4.32287186e-01 -1.56501085e-01 -3.80648285e-01
2.98863322e-01 -6.85077190e-01 -2.67195523e-01 2.04650566e-01
-8.17788899e-01 3.60002130e-01 6.41890645e-01 8.21662605e-01
7.66867936e-01 -1.71011582e-01 3.94117057e-01 -3.08919936e-01
9.44661051e-02 -1.75444603e-01 -8.82046700e-01 -1.02697752e-01
-1.77474752e-01 5.85287288e-02 6.57877207e-01 -1.68384433e-01
-1.19374418e+00 2.56439447e-01 -2.76226431e-01 -8.81137967e-01
-5.62982261e-01 -1.46602005e-01 -6.37107342e-03 -2.55732477e-01
8.19941521e-01 1.89820319e-01 2.81838030e-01 -3.10043305e-01
5.24392724e-01 2.73981929e-01 6.42869532e-01 -3.46579701e-01
7.28510618e-01 1.05522025e+00 3.14221859e-01 -1.07257891e+00
-2.14455053e-01 9.32096562e-04 -1.20397913e+00 -6.35886252e-01
7.32888818e-01 -6.86923325e-01 -5.41630030e-01 4.31804895e-01
-1.57205951e+00 -4.77670282e-01 -6.26485169e-01 5.09291947e-01
-1.38512361e+00 1.77140072e-01 -4.46610451e-01 -6.69672191e-01
-7.24294558e-02 -1.11642838e+00 1.10219479e+00 1.10364944e-01
-7.87217021e-02 -9.43624377e-01 2.04605266e-01 5.30568302e-01
1.52906090e-01 5.59125066e-01 5.13237536e-01 7.36410543e-03
-1.04729676e+00 4.85587167e-03 -1.46109328e-01 6.15656495e-01
-4.44852293e-01 2.22882345e-01 -9.98331010e-01 1.43305510e-02
-1.42772347e-01 -5.03055692e-01 7.75214016e-01 6.74206018e-01
9.91751075e-01 -3.15695703e-02 -9.70134288e-02 8.99214089e-01
1.77429378e+00 1.84042931e-01 1.16599393e+00 7.38443017e-01
5.26586592e-01 7.56705642e-01 4.98079777e-01 2.11275876e-01
3.59663874e-01 6.07535958e-01 9.10254240e-01 1.28626317e-01
-4.22245264e-01 -5.56860447e-01 2.88343370e-01 5.43789923e-01
-4.54157919e-01 1.52693465e-01 -7.02094793e-01 5.17117083e-01
-1.51538754e+00 -1.22495139e+00 -2.97464103e-01 1.82004237e+00
1.63923249e-01 1.84768990e-01 1.13709934e-01 1.99301258e-01
3.77397865e-01 -1.12620160e-01 -4.90459293e-01 -1.02799034e+00
-2.96987087e-01 5.52357852e-01 7.76929855e-01 4.70242649e-01
-8.23546767e-01 6.81458414e-01 7.63166142e+00 9.10713077e-01
-9.79502380e-01 -5.13093434e-02 3.42892170e-01 -8.95111337e-02
-4.49265391e-01 -1.22732133e-01 -6.51543319e-01 -2.22133160e-01
8.26305449e-01 3.00436974e-01 4.15982336e-01 1.08425832e+00
-3.81034538e-02 -4.88953143e-01 -9.79042768e-01 1.00598466e+00
6.64674759e-01 -1.96122098e+00 2.05594882e-01 1.57508478e-01
8.64275753e-01 1.46999285e-02 1.86250964e-03 -5.24200127e-02
2.84481019e-01 -1.30734909e+00 1.34471416e+00 8.54857385e-01
9.74926531e-01 -1.19273794e+00 4.69086230e-01 5.20971358e-01
-1.02515340e+00 -1.46645993e-01 -6.09755933e-01 -4.81200889e-02
2.25143850e-01 -3.06806602e-02 -8.82078469e-01 8.74959886e-01
6.32590532e-01 7.36264944e-01 -5.99541545e-01 9.83383536e-01
-2.86200717e-02 2.13586129e-02 -3.07421982e-01 2.28511825e-01
3.01936507e-01 -4.15854931e-01 5.11015236e-01 9.80817318e-01
5.29798925e-01 -1.54192895e-01 -2.72442639e-01 1.14985561e+00
2.33319357e-01 -8.23070556e-02 -1.28660905e+00 4.27824855e-01
8.14881250e-02 1.08014667e+00 -7.40483463e-01 -2.90535688e-01
-3.13158005e-01 6.99239671e-01 3.51432890e-01 2.71224290e-01
-9.36806262e-01 -2.51383752e-01 1.94365412e-01 6.07596159e-01
6.62192702e-01 -3.51928532e-01 -5.49141467e-01 -5.66399217e-01
3.18209380e-02 -6.18709087e-01 -2.02800915e-01 -1.35697043e+00
-1.09416401e+00 6.02721512e-01 4.67018068e-01 -1.61436832e+00
-3.39398772e-01 -1.09259510e+00 -4.23601776e-01 6.59093022e-01
-1.24552143e+00 -1.50895083e+00 -4.22527045e-01 1.74323127e-01
9.14869308e-01 -6.00709081e-01 7.38404632e-01 3.01071871e-02
1.64050087e-01 2.48978794e-01 1.09017901e-01 1.48838647e-02
1.90435067e-01 -9.78543401e-01 7.07834303e-01 4.49196309e-01
6.73413696e-03 1.27365351e-01 5.46546042e-01 -6.45592034e-01
-1.28362381e+00 -8.42397988e-01 -4.79305349e-02 -4.63315129e-01
3.11705880e-02 1.00488126e-01 -6.23513699e-01 8.21632087e-01
5.01966119e-01 1.05165929e-01 1.21427111e-01 -7.90599227e-01
-1.88533157e-01 -1.01863392e-01 -1.56671917e+00 3.55904996e-01
1.01512814e+00 -3.41283649e-01 -8.90089810e-01 -3.35970461e-01
1.72113746e-01 -7.73264349e-01 -5.41907012e-01 4.93757695e-01
1.03434920e+00 -1.66666865e+00 1.13019526e+00 -4.25885797e-01
9.36388016e-01 -1.85202137e-01 -3.79502475e-01 -1.45996773e+00
-8.22216794e-02 -2.37807319e-01 -2.10531920e-01 5.38189232e-01
-9.78695229e-03 -3.85392219e-01 1.23104107e+00 1.90050229e-01
-3.63279104e-01 -8.66102874e-01 -9.78943765e-01 -7.62192786e-01
4.68765140e-01 -4.48430270e-01 6.85847163e-01 5.10511637e-01
-6.04550004e-01 -1.36541143e-01 -2.52825409e-01 -2.04703808e-01
8.72415364e-01 -1.17738716e-01 1.21517515e+00 -1.33355808e+00
-2.02788234e-01 -3.40998113e-01 -7.62992442e-01 -9.63892281e-01
3.27290058e-01 -8.57045591e-01 -8.73621777e-02 -1.84582746e+00
-3.02128822e-01 -1.42849818e-01 6.86607957e-01 -1.78902104e-01
8.38903785e-01 5.58511078e-01 1.53052509e-01 -3.62931371e-01
-1.15450621e-01 6.14079356e-01 1.50345480e+00 -5.80163673e-02
2.29166761e-01 -2.01559678e-01 -8.47320259e-02 1.28444886e+00
8.20487618e-01 -2.90649444e-01 -4.37750369e-01 -5.41271806e-01
3.10355246e-01 1.59231484e-01 7.15582132e-01 -1.24780560e+00
-1.73565090e-01 -8.84367079e-02 9.03124213e-01 -1.45083320e+00
1.02570772e+00 -1.03007281e+00 7.09155262e-01 4.18725729e-01
-9.61418450e-02 5.78005672e-01 4.95408624e-01 5.01561821e-01
-6.95204735e-02 -6.65459394e-01 9.06674206e-01 -6.77205384e-01
-8.65247011e-01 9.75475311e-02 -2.97351897e-01 -1.41370982e-01
1.13607037e+00 -1.18317962e+00 -3.12443823e-01 -4.16668922e-01
-5.76268137e-01 -4.74887997e-01 8.31264853e-01 3.42908978e-01
1.46765375e+00 -1.81328237e+00 -7.28305280e-01 5.54629028e-01
-2.32931778e-01 1.22828193e-01 2.33116388e-01 2.94764698e-01
-1.52387035e+00 4.10195708e-01 -1.00943995e+00 -7.12661266e-01
-1.10534847e+00 2.58657068e-01 6.38862550e-01 1.46453574e-01
-1.02301335e+00 5.66157281e-01 1.84870302e-03 -7.45715559e-01
-3.90339307e-02 -2.01727808e-01 -2.67029643e-01 -1.97757632e-01
2.97951519e-01 7.98664093e-01 1.58164263e-01 -9.82533753e-01
1.93146765e-01 8.98729444e-01 3.83754849e-01 -4.69281852e-01
1.65777910e+00 2.11974964e-01 1.35116071e-01 3.63636613e-01
1.05849862e+00 -1.30292535e-01 -1.78928995e+00 1.59729257e-01
-4.30794209e-01 -7.79454112e-01 3.68634686e-02 -6.09903157e-01
-1.31848216e+00 1.13564956e+00 4.87345487e-01 -2.68054694e-01
6.46803439e-01 -5.02094448e-01 6.34466767e-01 4.72423583e-01
6.96927428e-01 -1.26672983e+00 4.45517540e-01 5.53456724e-01
1.48180318e+00 -8.87829959e-01 2.40252540e-01 -4.21402119e-02
-6.77869976e-01 1.89528489e+00 8.14289212e-01 -7.14948058e-01
4.69937682e-01 6.08061552e-01 -1.61100980e-02 -4.59359586e-01
-3.47917974e-01 2.99047738e-01 1.93134546e-01 1.03634822e+00
-1.25258923e-01 -6.85766041e-02 1.00577518e-01 2.36102849e-01
-4.29663986e-01 2.25103945e-01 1.02489483e+00 9.20608580e-01
-3.75810325e-01 -8.12799752e-01 -7.14366913e-01 2.76840866e-01
1.63261630e-02 3.72348040e-01 -3.01021814e-01 1.26925099e+00
3.06428075e-01 2.52246380e-01 3.12430978e-01 -1.71587080e-01
5.21759748e-01 -1.26143277e-01 1.23120642e+00 -2.17596859e-01
-4.02961999e-01 -2.94560045e-01 2.83000290e-01 -6.74164116e-01
-5.87719977e-01 -5.70087969e-01 -1.11676705e+00 -4.77900863e-01
-1.55473173e-01 -5.21799266e-01 8.16029608e-01 6.30086124e-01
4.91168164e-02 4.84574437e-01 2.99077034e-01 -1.69543159e+00
-3.45615417e-01 -7.74512887e-01 -4.77337748e-01 1.77480087e-01
2.00790957e-01 -1.03811550e+00 -2.21153554e-02 1.39429253e-02] | [8.938570022583008, -3.119022846221924] |
e208114d-5412-42d5-a446-1a22067f7c5d | pre-training-image-language-transformers-for | 2209.04372 | null | https://arxiv.org/abs/2209.04372v1 | https://arxiv.org/pdf/2209.04372v1.pdf | Pre-training image-language transformers for open-vocabulary tasks | We present a pre-training approach for vision and language transformer models, which is based on a mixture of diverse tasks. We explore both the use of image-text captioning data in pre-training, which does not need additional supervision, as well as object-aware strategies to pre-train the model. We evaluate the method on a number of textgenerative vision+language tasks, such as Visual Question Answering, visual entailment and captioning, and demonstrate large gains over standard pre-training methods. | ['Anelia Angelova', 'Weicheng Kuo', 'AJ Piergiovanni'] | 2022-09-09 | null | null | null | null | ['visual-entailment'] | ['reasoning'] | [ 4.57764775e-01 3.81565571e-01 8.88826475e-02 -5.49986660e-01
-8.48034501e-01 -5.60400188e-01 1.12669241e+00 -5.22771925e-02
-5.57578981e-01 4.33218598e-01 4.10505831e-01 -7.27407098e-01
6.18215144e-01 -4.88006651e-01 -1.15857947e+00 -1.44658938e-01
6.30026758e-01 8.12110662e-01 3.89487058e-01 -1.26199424e-01
1.91756219e-01 2.31397688e-01 -1.35532081e+00 6.38241589e-01
4.40779477e-01 1.12456071e+00 4.59709972e-01 8.66077840e-01
-4.03225899e-01 1.33417392e+00 -4.10877943e-01 -7.29750991e-01
-5.32202125e-02 -3.94088149e-01 -1.12061167e+00 4.32061762e-01
1.06613648e+00 -5.76100051e-01 -3.97821188e-01 7.06099033e-01
2.18095586e-01 -2.13486075e-01 7.38020957e-01 -1.31090093e+00
-1.16290963e+00 2.45733649e-01 -4.83310163e-01 3.45818430e-01
2.63947099e-01 4.71919745e-01 1.05889094e+00 -1.21506059e+00
6.26902282e-01 1.55187321e+00 6.50464296e-01 7.11429894e-01
-1.41348290e+00 -2.31900081e-01 1.85235634e-01 3.49555135e-01
-7.31912911e-01 -7.63106048e-01 5.21349132e-01 -7.06598341e-01
1.30078006e+00 -2.68986285e-01 6.21754885e-01 1.38209224e+00
-1.25540107e-01 1.16593742e+00 1.29864430e+00 -7.29764938e-01
-8.60692188e-02 4.59869891e-01 1.08124577e-02 9.47371721e-01
-7.04142451e-02 2.95355231e-01 -4.40900892e-01 1.56979952e-02
7.12635458e-01 -3.46811742e-01 -2.28045747e-01 -4.77958083e-01
-1.16517663e+00 8.72328281e-01 4.78350252e-01 -8.10429603e-02
-2.32583895e-01 6.47296131e-01 4.58442688e-01 9.64452848e-02
4.53309566e-01 4.15394187e-01 -3.09028417e-01 3.83685231e-01
-1.03339481e+00 1.63494796e-01 5.82992494e-01 1.02792406e+00
9.10421073e-01 4.62541133e-02 -6.87308669e-01 6.83085561e-01
5.69746494e-01 6.75920784e-01 2.37762153e-01 -1.10089660e+00
6.77888989e-01 2.63591051e-01 7.98130035e-02 -3.27094525e-01
-3.51278931e-02 -1.88975334e-02 -4.12345588e-01 2.56522238e-01
3.60000640e-01 1.18769854e-01 -1.62155080e+00 1.66021204e+00
2.64798775e-02 3.01866103e-02 1.84675813e-01 6.52714193e-01
1.13049042e+00 6.33599579e-01 3.64439070e-01 1.90161660e-01
1.45482135e+00 -1.55729806e+00 -5.99972844e-01 -7.26818562e-01
4.69815910e-01 -7.66292274e-01 1.35264802e+00 1.62743833e-02
-1.35806656e+00 -6.23866796e-01 -6.01539671e-01 -6.54469669e-01
-4.07692313e-01 1.20076835e-01 5.75626552e-01 2.75382251e-01
-1.62205184e+00 1.37698486e-01 -6.39408708e-01 -5.73935866e-01
7.78563559e-01 8.77923891e-02 -3.13475430e-01 -3.96091580e-01
-7.38908112e-01 1.32245719e+00 2.60548860e-01 -2.31410041e-01
-1.44855905e+00 -7.75134563e-01 -1.14419019e+00 -1.54371127e-01
3.07834893e-01 -1.28105474e+00 1.71460247e+00 -1.39293325e+00
-1.24908292e+00 1.49005067e+00 -2.92529911e-01 -7.18548119e-01
5.09986401e-01 -1.74744546e-01 1.57471120e-01 4.63399261e-01
5.60345873e-02 1.37457013e+00 1.23978436e+00 -1.40445888e+00
-3.99709731e-01 -2.14998916e-01 3.36679548e-01 4.17864323e-01
-1.79777667e-01 2.40649045e-01 -6.90244555e-01 -3.94913882e-01
-5.54637194e-01 -7.17410266e-01 -1.49903640e-01 4.25722420e-01
-3.29390377e-01 -2.92701513e-01 9.44253981e-01 -8.07149410e-01
1.75148576e-01 -1.73199904e+00 -5.84728159e-02 -3.61703128e-01
3.09737831e-01 3.09849173e-01 -5.76157570e-01 3.29760671e-01
-9.99084488e-02 3.65339927e-02 -1.84575468e-01 -8.14789474e-01
-3.57286856e-02 4.97187436e-01 -7.83300936e-01 1.61174625e-01
6.65469825e-01 1.60771370e+00 -7.76925087e-01 -9.68734324e-01
5.35804331e-01 2.90008247e-01 -4.55540091e-01 5.13264477e-01
-9.06936705e-01 2.06424102e-01 -1.91118509e-01 6.50887847e-01
2.36258119e-01 -6.68395698e-01 -2.48126745e-01 -4.31487978e-01
2.25743651e-02 5.27051449e-01 -1.89001322e-01 1.76192462e+00
-6.03165627e-01 1.06589985e+00 1.20717533e-01 -1.27468288e+00
4.21959102e-01 2.76703358e-01 -2.01016152e-03 -9.71439064e-01
1.31669104e-01 -2.30012834e-01 -4.96382356e-01 -8.66157293e-01
2.48716459e-01 -2.32067183e-01 3.83723736e-01 4.11801696e-01
5.33465087e-01 -4.66331124e-01 3.71972769e-02 4.33285028e-01
8.99526596e-01 6.24474943e-01 1.63130984e-01 -1.96667574e-02
2.57928818e-01 4.03513730e-01 -2.39207625e-01 8.29945326e-01
-2.32907251e-01 7.27577984e-01 2.38420978e-01 -1.81492940e-01
-1.21568751e+00 -9.57189918e-01 9.92990434e-02 1.27948320e+00
-1.35288378e-02 -3.15576673e-01 -6.79296136e-01 -9.52514231e-01
-6.08372502e-02 9.34869409e-01 -8.43527079e-01 1.87481809e-02
-2.96170473e-01 -2.96122998e-01 4.64675695e-01 8.60457957e-01
4.76887971e-01 -1.32627761e+00 -6.03326321e-01 -2.39154741e-01
-2.72168577e-01 -1.75943065e+00 -4.61064577e-01 1.89186260e-01
-7.35315084e-01 -8.51320326e-01 -8.01375329e-01 -1.06302989e+00
6.56331658e-01 4.41522807e-01 1.76848972e+00 2.86787152e-01
-1.62288502e-01 1.05483937e+00 -1.28537342e-01 -5.79326212e-01
-5.76365650e-01 -2.72787720e-01 -6.55536532e-01 -3.48727852e-01
2.72046506e-01 -2.61482984e-01 -4.38669950e-01 -1.14832334e-02
-7.80426145e-01 4.71817166e-01 8.11285317e-01 8.92983854e-01
5.46075821e-01 -9.51655209e-01 1.81132674e-01 -6.42148435e-01
4.19125587e-01 -1.49023265e-01 -5.34724116e-01 6.87083364e-01
-5.17206967e-01 1.76914856e-01 3.86792719e-01 -4.98289347e-01
-8.44248116e-01 1.71596557e-01 -1.41936883e-01 -8.27202916e-01
-2.23497972e-01 2.95460224e-01 -1.16825797e-01 -2.84305543e-01
5.75749159e-01 4.00037289e-01 1.46662891e-01 -1.97800756e-01
8.63384843e-01 4.26268548e-01 8.06988299e-01 -4.97485816e-01
9.67753410e-01 5.13417661e-01 -1.97107479e-01 -8.75655532e-01
-1.04686749e+00 -3.62784505e-01 -4.57860172e-01 1.07714059e-02
1.24247420e+00 -1.08902895e+00 -4.54250604e-01 2.02202722e-01
-1.56446064e+00 -6.87773347e-01 -4.08940792e-01 1.39626265e-01
-7.25050449e-01 5.33010483e-01 -3.54684293e-01 -5.21588206e-01
-5.91963649e-01 -9.55949545e-01 1.41953099e+00 1.50514310e-02
1.37522146e-01 -1.19097745e+00 -6.11546822e-03 7.09050059e-01
3.42265606e-01 -2.43563112e-02 1.01275980e+00 -6.69914663e-01
-8.62271965e-01 3.34525198e-01 -7.99342692e-01 5.38369417e-01
-3.69820118e-01 -3.14727515e-01 -1.20679927e+00 -1.80003151e-01
-1.79835707e-01 -1.31384671e+00 1.19385934e+00 2.80273825e-01
1.10048687e+00 -1.90548196e-01 -4.62681502e-01 5.61060309e-01
1.39358842e+00 -2.15643957e-01 7.45712638e-01 2.67851084e-01
8.96827638e-01 7.40914047e-01 2.79118210e-01 -2.19455108e-01
9.74266291e-01 6.16312385e-01 7.83735752e-01 -6.49062812e-01
-5.21884620e-01 -5.52328050e-01 2.35058188e-01 1.08447835e-01
2.01334402e-01 -2.89736509e-01 -1.00231612e+00 9.05396760e-01
-1.77925932e+00 -9.90235388e-01 1.59779355e-01 1.72019219e+00
8.96936893e-01 -8.28656331e-02 2.10920528e-01 -3.27797890e-01
3.09618443e-01 1.52131900e-01 -6.37102664e-01 -4.05315250e-01
2.05696537e-03 4.17518020e-01 5.57625592e-01 5.73175192e-01
-1.14559495e+00 1.46832395e+00 7.80030251e+00 3.59109730e-01
-1.00431406e+00 3.06157351e-01 6.80406451e-01 3.42441171e-01
-5.68825960e-01 2.58421063e-01 -6.05276883e-01 -5.20683117e-02
8.13942492e-01 1.07609577e-01 3.60284209e-01 7.37127125e-01
-1.17121540e-01 -2.75778547e-02 -1.32322240e+00 1.10011077e+00
6.96885586e-01 -1.54278719e+00 3.75615001e-01 -1.69044971e-01
3.64610285e-01 5.09766102e-01 -1.13159634e-01 4.49364990e-01
5.05909145e-01 -1.22696793e+00 1.00541878e+00 2.61024207e-01
8.75607789e-01 3.43752205e-02 3.27831656e-01 9.54037234e-02
-8.00864220e-01 1.80860907e-01 -3.06730211e-01 1.93927914e-01
1.35730818e-01 2.36070439e-01 -1.07583153e+00 2.57836163e-01
7.41327465e-01 8.43532979e-01 -9.07709122e-01 8.03449810e-01
-3.76971066e-01 6.50187373e-01 -2.71621972e-01 6.66429028e-02
2.63072759e-01 1.68418273e-01 2.12057203e-01 1.29539216e+00
-1.81876332e-01 -1.71072736e-01 2.71744460e-01 1.13320100e+00
-1.91808879e-01 -1.51951686e-01 -7.37664461e-01 -2.29967713e-01
8.50678831e-02 1.14447677e+00 -3.80912900e-01 -4.94996220e-01
-9.27159190e-01 1.15969598e+00 5.26382864e-01 6.20135546e-01
-9.28982019e-01 -5.69865368e-02 2.53089786e-01 2.34853297e-01
7.43340135e-01 -2.48559520e-01 -1.37554165e-02 -1.24111187e+00
-8.38781968e-02 -8.70721579e-01 2.74903446e-01 -1.53057420e+00
-1.17198193e+00 6.46128178e-01 1.04562826e-01 -6.94143236e-01
-5.45432746e-01 -9.37441349e-01 -5.91765761e-01 7.87921429e-01
-2.13683963e+00 -1.93879557e+00 -4.87783521e-01 9.23633695e-01
5.28532684e-01 3.04146633e-02 5.85659504e-01 2.17610337e-02
-2.32495405e-02 4.70409214e-01 -5.49971163e-01 9.56517756e-02
6.20392859e-01 -1.30237150e+00 7.03639925e-01 7.73801327e-01
5.43399036e-01 1.89794540e-01 7.76825666e-01 -2.50305355e-01
-1.41564453e+00 -1.27654159e+00 8.79963636e-01 -8.85657847e-01
6.87925696e-01 -6.33428693e-01 -6.69988751e-01 1.18129289e+00
8.86407197e-01 8.18978250e-02 2.53784388e-01 -3.43576595e-02
-6.58617139e-01 1.33308187e-01 -8.36098969e-01 5.97847462e-01
1.07742453e+00 -9.21006024e-01 -1.01847315e+00 6.80396259e-01
9.38576102e-01 -3.29285681e-01 -2.81026453e-01 2.93873698e-01
2.67486691e-01 -6.71298206e-01 1.38716578e+00 -8.25274527e-01
6.70562446e-01 -8.11630413e-02 -1.80726737e-01 -1.02259851e+00
-9.56932083e-02 -4.18586224e-01 -7.59162754e-02 1.21139443e+00
4.97365326e-01 -2.87410647e-01 6.42609537e-01 3.85437161e-01
-1.76865384e-01 -4.21689361e-01 -6.40264988e-01 -4.35757071e-01
3.17321308e-02 -3.75653148e-01 -5.40949851e-02 6.61596954e-01
-5.07038951e-01 9.33691978e-01 -3.42154860e-01 -1.34345949e-01
6.50926709e-01 1.38211265e-01 9.10585582e-01 -8.25353742e-01
-4.44529712e-01 -2.22304136e-01 -2.88200527e-01 -1.25584733e+00
3.92735958e-01 -1.05843675e+00 9.33438912e-02 -2.20999026e+00
3.70529860e-01 2.01783821e-01 1.06971310e-02 9.92307663e-01
-7.90518522e-02 4.62561727e-01 2.35603839e-01 2.51608044e-01
-9.10251081e-01 6.82020128e-01 1.25676274e+00 -3.51616263e-01
1.85972646e-01 -4.42667812e-01 -7.90077090e-01 6.63623095e-01
4.19602454e-01 -1.97743624e-01 -6.39516711e-01 -9.28743482e-01
7.17435107e-02 -3.23942825e-02 9.33333337e-01 -6.81150675e-01
-1.86210475e-03 -1.23261623e-01 4.48026597e-01 -5.96379578e-01
5.56000113e-01 -6.74497902e-01 -5.40950835e-01 2.37544939e-01
-4.86450195e-01 2.50941128e-01 4.77855831e-01 5.28926551e-01
-7.69254565e-02 -7.02534392e-02 8.74230206e-01 -3.45185608e-01
-7.01136589e-01 2.50518769e-01 -2.41920367e-01 4.17449504e-01
7.90634036e-01 -2.01059088e-01 -6.95136666e-01 -5.69199800e-01
-4.64040130e-01 4.17947829e-01 5.81492960e-01 4.69639510e-01
8.26931238e-01 -1.06510651e+00 -7.28556037e-01 -1.95927501e-01
4.67145562e-01 -2.29062617e-01 -2.37085849e-01 7.65904903e-01
-3.44692856e-01 5.71987987e-01 -1.74362853e-01 -8.50910604e-01
-1.31674552e+00 9.84997690e-01 5.06474078e-01 -1.65912181e-01
-7.22451627e-01 7.73134947e-01 6.13620639e-01 -1.76850006e-01
3.17618102e-01 -4.49575216e-01 -2.43650675e-01 -2.95414686e-01
5.40647686e-01 -4.63658184e-01 -1.69850230e-01 -4.33605582e-01
-3.91043454e-01 5.80705822e-01 -3.05690289e-01 -4.35811579e-01
9.71083224e-01 -3.26253772e-01 -5.97477295e-02 2.80220300e-01
1.05213904e+00 -5.94387770e-01 -1.16757977e+00 -4.71811771e-01
1.26127852e-02 -1.12648897e-01 2.59871930e-01 -1.02436149e+00
-8.75369728e-01 1.25823164e+00 4.52966750e-01 -3.72636057e-02
1.01846159e+00 5.53260863e-01 6.43027544e-01 6.40901268e-01
-2.55584121e-02 -8.17576528e-01 5.25883436e-01 5.84094942e-01
1.09939623e+00 -1.42300594e+00 -1.87922746e-01 -2.10084990e-01
-8.05639148e-01 7.87676215e-01 7.80236959e-01 -5.42276837e-02
5.01416743e-01 1.56181827e-01 1.79508343e-01 -3.02394569e-01
-1.09988117e+00 -7.41520345e-01 3.49965096e-01 8.58946085e-01
2.89691031e-01 -5.10195494e-01 2.51812130e-01 2.51176842e-02
3.00673842e-02 2.31267452e-01 3.35737854e-01 8.69302094e-01
-2.39283636e-01 -8.70712996e-01 -2.24943295e-01 3.79616797e-01
-3.21480066e-01 -6.73202097e-01 -7.59041369e-01 8.03651094e-01
-1.13670692e-01 9.32374775e-01 1.00445338e-01 -1.20006107e-01
9.00427401e-02 1.57896549e-01 1.06411994e+00 -6.80655062e-01
-3.37915778e-01 -1.21404521e-01 4.44086492e-01 -6.38953447e-01
-7.49218047e-01 -3.18146199e-01 -9.46437359e-01 1.63648933e-01
-3.01914185e-01 -1.90723121e-01 7.19580948e-01 1.31338239e+00
3.62611085e-01 3.50966573e-01 1.05894739e-02 -8.59842181e-01
-4.39417988e-01 -8.82953346e-01 1.64036810e-01 5.12369215e-01
5.54819882e-01 -4.91123676e-01 -2.68369287e-01 5.57562411e-01] | [10.859192848205566, 1.724151611328125] |
61d98e75-5d25-4188-91b4-dcd74271b805 | bn-htrd-a-benchmark-dataset-for-document | 2206.08977 | null | https://arxiv.org/abs/2206.08977v1 | https://arxiv.org/pdf/2206.08977v1.pdf | BN-HTRd: A Benchmark Dataset for Document Level Offline Bangla Handwritten Text Recognition (HTR) and Line Segmentation | We introduce a new dataset for offline Handwritten Text Recognition (HTR) from images of Bangla scripts comprising words, lines, and document-level annotations. The BN-HTRd dataset is based on the BBC Bangla News corpus, meant to act as ground truth texts. These texts were subsequently used to generate the annotations that were filled out by people with their handwriting. Our dataset includes 788 images of handwritten pages produced by approximately 150 different writers. It can be adopted as a basis for various handwriting classification tasks such as end-to-end document recognition, word-spotting, word or line segmentation, and so on. We also propose a scheme to segment Bangla handwritten document images into corresponding lines in an unsupervised manner. Our line segmentation approach takes care of the variability involved in different writing styles, accurately segmenting complex handwritten text lines of curvilinear nature. Along with a bunch of pre-processing and morphological operations, both Hough line and circle transforms were employed to distinguish different linear components. In order to arrange those components into their corresponding lines, we followed an unsupervised clustering approach. The average success rate of our segmentation technique is 81.57% in terms of FM metrics (similar to F-measure) with a mean Average Precision (mAP) of 0.547. | ['Mohammad Khairul Islam', 'Riya Pal', 'Mitu Paul', 'Nazifa Tabassum', 'Md. Ataur Rahman'] | 2022-05-29 | null | null | null | null | ['handwritten-line-segmentation'] | ['computer-vision'] | [ 1.42766386e-01 -3.76812488e-01 -2.86918897e-02 -5.32085359e-01
-5.69907844e-01 -1.02585316e+00 7.66145170e-01 2.59357374e-02
-4.67581987e-01 4.80785310e-01 -7.09700808e-02 -4.50665474e-01
-9.99542177e-02 -7.22806811e-01 -4.16338265e-01 -7.60423839e-01
5.26548862e-01 8.28516483e-01 2.15408370e-01 -9.62684825e-02
9.48559642e-01 9.29327011e-01 -8.73050690e-01 4.34710592e-01
7.80972958e-01 5.94420016e-01 2.50940412e-01 9.57184315e-01
-5.67205966e-01 8.81842315e-01 -9.65521038e-01 -7.55015910e-01
1.00433119e-01 -7.07022130e-01 -9.47983205e-01 9.93955135e-01
1.34247258e-01 -3.94307494e-01 -3.02164406e-01 8.61263573e-01
2.50351012e-01 -4.31535579e-02 1.03044200e+00 -5.16840041e-01
-7.04597890e-01 9.60046113e-01 -9.02828157e-01 -1.24892764e-01
-2.98441853e-02 -2.24624574e-01 7.91732848e-01 -1.10440314e+00
7.14180946e-01 7.27383494e-01 5.69967031e-01 3.00054520e-01
-6.56304777e-01 -3.54839452e-02 -3.89275789e-01 8.45749825e-02
-1.22888494e+00 -3.37823063e-01 7.31754363e-01 -6.22069657e-01
6.04635835e-01 3.87845904e-01 3.47050518e-01 6.58522785e-01
6.07041679e-02 8.66096377e-01 1.09907115e+00 -1.04330671e+00
3.86741981e-02 2.84889579e-01 6.06884241e-01 7.13244855e-01
1.14792641e-02 -7.90773094e-01 -1.84302285e-01 2.57700771e-01
9.71946061e-01 -4.03809518e-01 -1.45315573e-01 1.69400916e-01
-1.30569875e+00 8.20426285e-01 -1.74399808e-01 6.64604247e-01
-1.80145323e-01 -3.55685711e-01 4.71041322e-01 1.26935497e-01
9.41069871e-02 5.12644887e-01 -1.04766466e-01 -9.30395424e-02
-1.28336620e+00 -5.12423515e-02 9.48408425e-01 1.13135600e+00
4.32278991e-01 -8.57156068e-02 -3.73591006e-01 1.33835793e+00
2.32407153e-01 4.41708684e-01 7.03836262e-01 -4.83837992e-01
7.76362181e-01 4.72695142e-01 1.67589724e-01 -1.17055631e+00
-1.94134697e-01 1.22662624e-02 -7.91853130e-01 -4.17176746e-02
7.24326134e-01 -4.26047519e-02 -1.09089041e+00 4.00880963e-01
-2.22975656e-01 -8.98356259e-01 5.47657954e-03 8.59534979e-01
4.25790757e-01 9.07690704e-01 -5.24405718e-01 -2.53283083e-01
1.45499086e+00 -1.26754642e+00 -8.88572395e-01 1.82726994e-01
5.58532178e-01 -1.44228530e+00 1.01292360e+00 9.15562868e-01
-8.96359026e-01 -5.88091373e-01 -1.05897558e+00 -1.32463992e-01
-2.82016575e-01 1.09617710e+00 2.46392265e-01 7.88679183e-01
-6.54768646e-01 3.67235601e-01 -6.79933310e-01 -4.10442889e-01
3.52081001e-01 6.97497055e-02 -3.61315370e-01 8.25770646e-02
-4.03885186e-01 7.67770112e-01 5.39864957e-01 5.29934406e-01
-4.48655754e-01 1.68945462e-01 -3.00802231e-01 -4.68403548e-01
3.25169414e-01 1.92829251e-01 1.05536079e+00 -9.92215395e-01
-1.75495887e+00 1.13333726e+00 -1.85413174e-02 -2.87063450e-01
1.05665982e+00 -1.69549718e-01 -6.33058846e-01 3.10955524e-01
-1.07238851e-01 2.96957493e-01 9.02652323e-01 -1.19718444e+00
-6.29720688e-01 -3.28036875e-01 -6.45571470e-01 1.19667612e-01
-2.08874345e-01 3.44948322e-01 -9.55102265e-01 -1.00065494e+00
4.27293450e-01 -6.84577584e-01 5.10414056e-02 -4.36770052e-01
-8.31117213e-01 -2.14170530e-01 1.13761079e+00 -1.32156682e+00
1.14880991e+00 -1.84828460e+00 -1.25006989e-01 8.05557251e-01
-4.30704877e-02 3.32997262e-01 2.52775010e-02 5.57800353e-01
3.90354663e-01 -1.69481281e-02 -3.81755084e-01 8.58214945e-02
-1.26556665e-01 2.71512959e-02 -5.07002175e-01 4.35891002e-01
1.21273920e-01 7.16706276e-01 -4.48989779e-01 -6.18407428e-01
4.79816971e-03 1.47240594e-01 1.45930767e-01 1.34439543e-01
-2.42408901e-01 3.14187735e-01 -2.87149429e-01 7.24698901e-01
5.50463557e-01 1.49291784e-01 2.12920114e-01 -1.23941980e-01
-4.10601586e-01 -1.99748084e-01 -1.05167913e+00 1.04894507e+00
-1.86900556e-01 1.06375086e+00 -4.22530323e-01 -8.59598577e-01
1.77438903e+00 5.01493327e-02 1.35221452e-01 -4.93008852e-01
3.43649387e-01 6.09467030e-01 -7.89203048e-02 -7.29144096e-01
1.06955647e+00 4.16061521e-01 -5.95530681e-03 6.03388190e-01
1.12121522e-01 -4.85801369e-01 7.21345663e-01 -8.01662505e-02
4.55949515e-01 8.62494782e-02 3.88180614e-02 -1.82772800e-01
7.68757939e-01 3.92986804e-01 -6.36324137e-02 8.55675220e-01
2.49165207e-01 9.33972895e-01 6.32303357e-01 -4.70615417e-01
-1.66747570e+00 -4.56025362e-01 -2.26502925e-01 7.92982638e-01
-2.93310016e-01 -6.12608343e-02 -1.12166309e+00 -5.41110039e-01
-4.89685744e-01 6.07402980e-01 -2.90525019e-01 5.77800930e-01
-6.44215167e-01 -7.26506472e-01 9.29734707e-01 5.64351797e-01
9.36753273e-01 -1.10648906e+00 -3.09297085e-01 3.35413188e-01
5.84740080e-02 -1.17290914e+00 -5.72007239e-01 2.96961784e-01
-7.16313839e-01 -8.60554099e-01 -1.37437642e+00 -1.24993348e+00
1.08538616e+00 -4.44628038e-02 4.80245322e-01 -1.15016624e-01
-3.36658657e-01 1.61183149e-01 -7.73996353e-01 -2.49819830e-01
-7.20224380e-01 7.63435513e-02 -2.23068893e-01 1.46886036e-01
3.36986482e-01 3.72480750e-01 -6.08375967e-02 7.60844767e-01
-8.62761438e-01 -9.25831124e-03 6.62527561e-01 8.35768521e-01
5.51426411e-01 3.19222122e-01 2.32070349e-02 -1.06615543e+00
7.82606661e-01 1.52671933e-01 -8.07113469e-01 6.90096021e-01
-2.25567639e-01 -2.52497673e-01 7.50960708e-01 -4.10641700e-01
-1.16300535e+00 1.48648739e-01 -2.98674069e-02 1.60562783e-01
-3.20513725e-01 5.55763006e-01 1.17093585e-02 -2.84170937e-02
6.27006412e-01 7.38956690e-01 -1.82756260e-01 -4.53767955e-01
4.00608659e-01 1.51298702e+00 1.05312026e+00 -5.10749459e-01
5.51548064e-01 6.29405528e-02 -3.95199567e-01 -1.36081004e+00
-5.60008049e-01 -5.15929759e-01 -1.30275989e+00 -3.67606997e-01
8.67827535e-01 -2.72952497e-01 -3.26896042e-01 1.31192970e+00
-1.27842236e+00 -4.70045328e-01 5.86770251e-02 3.67158353e-01
-3.99304628e-01 7.47186422e-01 -9.32774305e-01 -8.67064893e-01
-2.40097865e-01 -8.96751165e-01 7.45847166e-01 4.02636230e-01
-1.88698128e-01 -8.34992349e-01 -1.14806250e-01 6.05826199e-01
-1.47011459e-01 -2.37136409e-02 1.01083219e+00 -9.09489214e-01
-3.17071714e-02 -4.75317240e-01 -2.79086292e-01 7.02109635e-01
1.94683105e-01 6.61121845e-01 -6.91609502e-01 1.21570587e-01
-1.72314659e-01 -2.65575886e-01 7.37777591e-01 -4.94134016e-02
1.19684219e+00 -3.76358181e-01 1.82371855e-01 2.36548185e-01
1.21678388e+00 6.53506041e-01 8.70269179e-01 4.57441956e-01
1.00466549e+00 6.96606696e-01 6.69535816e-01 6.20010734e-01
-1.27021179e-01 4.66183037e-01 -5.80790102e-01 1.32478159e-02
-2.16513388e-02 -4.66706157e-02 3.71828973e-01 1.18536103e+00
-4.44495738e-01 -4.75723177e-01 -1.34758830e+00 3.62086803e-01
-1.46146953e+00 -8.30882370e-01 -7.82593668e-01 1.98457265e+00
1.02570987e+00 2.34998137e-01 2.87820041e-01 6.46044493e-01
9.94436324e-01 -1.17464624e-01 -1.07318297e-01 -8.19790542e-01
-4.76352960e-01 -1.24470089e-02 7.55685747e-01 3.93479079e-01
-1.19042766e+00 1.17660332e+00 5.71069813e+00 9.50096309e-01
-1.33442664e+00 -5.49161196e-01 8.49918664e-01 8.49054337e-01
1.64334133e-01 -2.42798701e-01 -8.99576545e-01 3.38234395e-01
5.43446779e-01 2.25180641e-01 2.88353115e-01 5.60672998e-01
1.67464197e-01 -2.50370920e-01 -6.80137992e-01 8.77857924e-01
4.07129884e-01 -1.29181278e+00 2.76248246e-01 -1.72761172e-01
1.07913554e+00 -5.06250322e-01 -2.06538290e-01 -3.27988774e-01
7.84213245e-02 -9.89039660e-01 1.10666835e+00 6.72668397e-01
8.69685292e-01 -6.76923752e-01 9.26699638e-01 3.24598372e-01
-6.47510767e-01 2.11551920e-01 -4.40584630e-01 2.39620864e-01
-1.38259575e-01 4.99763489e-01 -1.14715254e+00 5.19573689e-01
2.62315720e-01 4.40226436e-01 -6.81904256e-01 1.02726388e+00
-3.46446067e-01 8.52199495e-01 -8.21664557e-02 -4.18799102e-01
6.05071902e-01 -6.29654229e-01 -4.78152074e-02 1.65030456e+00
3.01662564e-01 8.55514128e-03 -2.27558851e-01 5.52578807e-01
-2.47137547e-02 5.19865751e-01 -6.50679022e-02 -5.69470882e-01
3.84324312e-01 1.23660529e+00 -1.68701744e+00 -3.87852460e-01
-1.22970641e-01 1.45898116e+00 -1.71509758e-01 3.20820600e-01
-6.05454922e-01 -1.24609292e+00 -6.40065551e-01 -1.24761507e-01
5.17914057e-01 -6.63427532e-01 -5.52992761e-01 -8.04150105e-01
8.10982361e-02 -1.09179091e+00 -4.06964421e-02 -6.30947113e-01
-1.19906795e+00 7.43701756e-01 -4.41009432e-01 -9.64547694e-01
-3.19547579e-02 -1.06863904e+00 -5.56287467e-01 7.52080679e-01
-7.41151750e-01 -1.06247771e+00 -3.80361229e-01 3.96061331e-01
8.62273157e-01 -6.56941652e-01 6.57304764e-01 -4.72223721e-02
-7.36265063e-01 5.50508976e-01 7.13683844e-01 9.45672691e-01
6.02178574e-01 -1.23853469e+00 2.97121763e-01 8.77391994e-01
5.95997214e-01 4.02487040e-01 4.80378509e-01 -6.44438088e-01
-1.17365897e+00 -9.99403775e-01 8.62467706e-01 -2.21326321e-01
6.37777507e-01 -3.47449332e-01 -8.52505744e-01 5.57968616e-01
6.60426617e-02 -3.61718923e-01 5.62877655e-01 -4.83767152e-01
-1.65772274e-01 7.85031095e-02 -8.63804162e-01 4.83097464e-01
2.39111543e-01 -4.21660125e-01 -6.78223789e-01 4.52800751e-01
-2.31984288e-01 -3.99882197e-01 -7.91604996e-01 -2.66989291e-01
5.76727331e-01 -7.57154524e-01 3.10134947e-01 -1.81332245e-01
7.76773214e-01 -3.70471656e-01 1.08401537e-01 -8.00960302e-01
-9.82029960e-02 -5.75167298e-01 3.40237588e-01 1.66249883e+00
4.74552959e-01 -2.71396875e-01 6.90430164e-01 1.61554500e-01
6.50287652e-03 -4.02529180e-01 -2.05702618e-01 -7.11049616e-01
-7.08002001e-02 -2.16644377e-01 2.88934588e-01 1.12180328e+00
-1.16489924e-01 9.54854488e-02 -3.22709978e-01 -1.39707431e-01
5.29453456e-01 7.79669285e-02 7.26592720e-01 -7.83448517e-01
-2.21947581e-01 -8.15838277e-01 -3.43220204e-01 -1.11683404e+00
-1.37434646e-01 -6.68309033e-01 2.68304676e-01 -1.41563630e+00
6.19510524e-02 -4.23315465e-01 3.64505142e-01 4.94028419e-01
2.63168097e-01 5.33676147e-01 2.20742926e-01 6.86835110e-01
-2.63310313e-01 -3.82491797e-02 1.03998232e+00 -2.76176482e-01
-3.43353510e-01 1.31816834e-01 -6.48537427e-02 7.95115650e-01
9.11038399e-01 -1.22025147e-01 5.06874286e-02 -5.47950268e-01
-1.65467501e-01 -6.83167716e-03 3.86947468e-02 -7.46780932e-01
4.07122374e-01 -9.26546827e-02 8.22384953e-01 -9.67353582e-01
-3.41769427e-01 -5.32606304e-01 -1.16265126e-01 3.80755872e-01
-4.22963113e-01 -7.36569613e-02 -7.51016214e-02 1.16991997e-02
-3.61630470e-01 -9.37333405e-01 6.58004105e-01 -5.72725534e-02
-8.79694462e-01 -2.99016774e-01 -7.37150967e-01 -3.87654454e-01
1.15530920e+00 -6.05046809e-01 -2.31264114e-01 -9.66728479e-02
-4.87513572e-01 -1.91132382e-01 3.63190025e-01 7.58142993e-02
5.15475094e-01 -9.37794268e-01 -9.33794320e-01 1.68903232e-01
2.52473094e-02 -2.11820215e-01 -7.64493346e-02 6.90252483e-01
-1.65728533e+00 5.29706001e-01 -5.16462326e-01 -5.27339816e-01
-1.26026595e+00 3.78281102e-02 1.02836872e-02 2.06556614e-03
-5.24058759e-01 7.36899018e-01 -6.78158224e-01 -1.45026654e-01
3.08516204e-01 -1.82870120e-01 -3.56962323e-01 3.74336958e-01
5.40541530e-01 5.75320661e-01 3.05698395e-01 -7.20111012e-01
-7.81189203e-02 1.02043128e+00 -3.38775963e-01 -4.05575722e-01
1.15195298e+00 1.15295410e-01 -3.47716719e-01 6.23703182e-01
8.52993846e-01 5.13198256e-01 -9.90476787e-01 -8.64870474e-03
4.71166015e-01 -3.69343370e-01 -2.49683291e-01 -1.00207400e+00
-7.53043532e-01 7.73892999e-01 8.97487551e-02 2.36535907e-01
9.14344668e-01 -2.03983083e-01 4.55360115e-01 8.68694007e-01
3.62529308e-01 -1.65621650e+00 -2.29295082e-02 7.23830760e-01
9.92590606e-01 -1.03344631e+00 1.31354630e-02 -2.02588350e-01
-1.03537703e+00 2.00857639e+00 2.35126093e-01 -2.77125686e-02
-4.76881629e-04 3.78326476e-01 3.86489421e-01 1.17536657e-01
1.42400429e-01 1.63503364e-01 4.10052806e-01 4.13273484e-01
6.16276443e-01 3.57094221e-02 -6.46026075e-01 5.65833271e-01
-4.37120020e-01 -1.68860152e-01 8.60031366e-01 9.50488329e-01
-6.18902802e-01 -1.05171466e+00 -7.87408471e-01 4.85937089e-01
-4.88546342e-01 2.46370941e-01 -7.92573631e-01 6.40356064e-01
-1.29179731e-01 7.92352974e-01 1.66229174e-01 -3.19573253e-01
1.87791705e-01 -3.00788842e-02 6.55320764e-01 -3.41609001e-01
-4.37325746e-01 2.97190189e-01 1.64812878e-02 3.07204127e-01
-3.25686902e-01 -5.70412993e-01 -1.26636171e+00 -1.62533343e-01
-4.24560845e-01 2.11931229e-01 8.73385429e-01 8.38784158e-01
-4.47697937e-01 2.20951438e-01 8.66036236e-01 -6.20473683e-01
-5.20352662e-01 -1.08839011e+00 -9.19426322e-01 2.68275559e-01
-1.82778612e-01 1.13625236e-01 6.85984045e-02 7.04571486e-01] | [11.830108642578125, 2.6056110858917236] |
50163adc-b3a4-4623-b941-dbac4a38d518 | 190506229 | 1905.06229 | null | https://arxiv.org/abs/1905.06229v2 | https://arxiv.org/pdf/1905.06229v2.pdf | Toward Standardized Classification of Foveated Displays | Emergent in the field of head mounted display design is a desire to leverage the limitations of the human visual system to reduce the computation, communication, and display workload in power and form-factor constrained systems. Fundamental to this reduced workload is the ability to match display resolution to the acuity of the human visual system, along with a resulting need to follow the gaze of the eye as it moves, a process referred to as foveation. A display that moves its content along with the eye may be called a Foveated Display, though this term is also commonly used to describe displays with non-uniform resolution that attempt to mimic human visual acuity. We therefore recommend a definition for the term Foveated Display that accepts both of these interpretations. Furthermore, we include a simplified model for human visual Acuity Distribution Functions (ADFs) at various levels of visual acuity, across wide fields of view and propose comparison of this ADF with the Resolution Distribution Function of a foveated display for evaluation of its resolution at a particular gaze direction. We also provide a taxonomy to allow the field to meaningfully compare and contrast various aspects of foveated displays in a display and optical technology-agnostic manner. | ['David Luebke', 'Rachel Albert', 'Michael Stengel', 'Kaan Aksit', 'Josef Spjut', 'Trey Greer', 'Jonghyun Kim', 'Ben Boudaoud'] | 2019-05-03 | null | null | null | null | ['foveation'] | ['computer-vision'] | [ 2.12880388e-01 -1.37487993e-01 3.55480969e-01 -2.29684561e-01
-1.56555831e-01 -7.98788726e-01 3.08557719e-01 -1.31200746e-01
-3.71859848e-01 3.88508767e-01 3.03169370e-01 -7.14891791e-01
-1.19533524e-01 -3.11293781e-01 -1.29122213e-01 -3.28733623e-01
2.91180342e-01 -4.84927922e-01 3.88913721e-01 -5.32541648e-02
2.58668870e-01 6.32935822e-01 -2.06599045e+00 5.14780842e-02
6.81928873e-01 1.24750829e+00 5.32124579e-01 9.88637447e-01
3.37439626e-01 3.79656672e-01 -8.56191635e-01 1.35371029e-01
1.07496887e-01 -3.67144853e-01 -2.31830135e-01 1.03394262e-01
8.23203862e-01 -5.05316675e-01 -5.97097399e-03 7.40893483e-01
9.32289660e-01 -2.10227836e-02 5.85534275e-01 -1.07220149e+00
-4.43358988e-01 -6.52408421e-01 -5.75641751e-01 6.31949961e-01
7.72838056e-01 3.24834615e-01 4.13114518e-01 -5.46606064e-01
1.85085610e-01 8.76587570e-01 2.17723623e-01 3.59412819e-01
-1.28489983e+00 -3.19528282e-01 7.07746744e-02 -2.19896942e-01
-1.35076582e+00 -7.40603387e-01 1.82083264e-01 -7.34702945e-01
1.28375697e+00 4.03386414e-01 6.78830147e-01 2.80333370e-01
6.77716315e-01 -1.73868477e-01 1.20602715e+00 -6.48330927e-01
1.64071113e-01 3.63470018e-01 1.21786751e-01 3.45324576e-01
4.75670218e-01 -1.09660760e-01 -6.91098750e-01 -1.58262119e-01
8.32701802e-01 -3.03396255e-01 -8.26885700e-01 -3.81948471e-01
-5.70167184e-01 2.02187568e-01 1.80899471e-01 1.66535918e-02
-4.11281049e-01 -1.80738792e-01 -9.48007405e-02 1.72541857e-01
8.81716013e-02 5.84321082e-01 -1.10305898e-01 -4.10829067e-01
-4.98882413e-01 1.93726525e-01 5.99066138e-01 9.10951257e-01
2.85223782e-01 -7.30881393e-02 -2.27667615e-01 4.84795183e-01
3.76594037e-01 6.24832690e-01 2.85868675e-01 -9.89754140e-01
6.33099824e-02 5.82443833e-01 6.14899218e-01 -4.83485371e-01
-6.00308478e-01 6.94635734e-02 -1.49636462e-01 1.17601705e+00
5.00322342e-01 -2.92621106e-01 -7.65841544e-01 1.55006468e+00
1.56072602e-01 -6.73305750e-01 -2.33297825e-01 1.03692544e+00
5.20018578e-01 1.11354433e-01 -9.73641127e-02 -4.99909431e-01
1.43347168e+00 -1.22205950e-01 -8.01299930e-01 -1.42953485e-01
3.22112471e-01 -1.01848817e+00 1.59465992e+00 3.93096805e-01
-1.27173865e+00 -5.04965007e-01 -1.45325804e+00 -1.91211894e-01
-1.10419728e-02 -2.05509901e-01 1.61832511e-01 9.55530882e-01
-1.48632860e+00 -8.08067098e-02 -7.99106419e-01 -6.34788573e-01
-5.36057679e-03 4.98776108e-01 7.44254589e-02 3.90525460e-01
-2.79035032e-01 1.27759743e+00 -3.47472996e-01 -3.25988084e-01
2.86327690e-01 -5.50949574e-01 -6.09056950e-01 2.32241839e-01
3.56372111e-02 -9.49583650e-01 1.45407701e+00 -5.35778105e-01
-1.36445367e+00 1.00616837e+00 -7.31192350e-01 -1.15570545e-01
4.71715331e-02 -1.99942291e-01 -6.01998925e-01 8.61532707e-03
-3.24387014e-01 2.79732138e-01 8.97326291e-01 -9.08225954e-01
-8.83373141e-01 -4.94002640e-01 2.10103691e-02 4.98647153e-01
-7.02691600e-02 2.51314849e-01 -3.48921686e-01 2.17067048e-01
-8.17083642e-02 -8.85057747e-01 4.44839507e-01 7.24964440e-02
-2.08693355e-01 -2.26697147e-01 8.46488416e-01 -1.63867235e-01
1.66202569e+00 -2.27460027e+00 -4.30949479e-01 9.98356864e-02
8.56908262e-01 2.14419663e-01 3.46331000e-01 1.44840986e-01
4.38059606e-02 -1.46439016e-01 5.32627523e-01 1.05615696e-02
-1.29700437e-01 -3.99853975e-01 -2.65112609e-01 4.72815782e-01
-4.04521450e-02 5.99567831e-01 -2.14029610e-01 -1.00136079e-01
2.84484059e-01 5.83881855e-01 -3.34166497e-01 2.26757258e-01
3.62277746e-01 3.45560759e-01 -8.28737244e-02 3.59854549e-01
6.30646646e-01 -4.71114069e-01 1.52526498e-01 -7.32468888e-02
-7.63171852e-01 2.99329221e-01 -8.30733836e-01 1.05324185e+00
-3.18715125e-01 1.13789356e+00 2.82359391e-01 2.53786683e-01
8.09176683e-01 4.58216630e-02 7.32811019e-02 -1.16501713e+00
2.26302639e-01 6.99853823e-02 4.79966819e-01 -5.58687806e-01
7.25138485e-01 9.40991044e-02 3.43591034e-01 6.78256571e-01
-6.56630218e-01 -1.24841973e-01 -1.32476553e-01 1.11880386e-02
9.58817422e-01 4.44135210e-03 3.74587417e-01 -4.25971389e-01
9.08605829e-02 -4.07785624e-01 -1.64876491e-01 3.53435725e-01
-3.66357774e-01 5.99025667e-01 2.45896786e-01 -1.35084435e-01
-1.08829713e+00 -1.15214777e+00 -4.26195443e-01 1.14803970e+00
1.05229929e-01 -4.94717151e-01 -8.03433299e-01 3.53800744e-01
5.05319983e-02 6.04649723e-01 -2.39680961e-01 2.28825822e-01
-2.19165877e-01 -1.91692859e-01 -9.01146606e-02 3.66990566e-01
1.64419517e-01 -5.34817219e-01 -2.13454914e+00 -2.65111893e-01
3.69121611e-01 -8.38088751e-01 -4.58037823e-01 8.77272561e-02
-4.20570284e-01 -1.07076144e+00 -7.08023787e-01 -4.72307950e-01
4.14041281e-01 5.49774647e-01 1.04853272e+00 -1.26849696e-01
-3.56496930e-01 9.40745950e-01 3.89565416e-02 -7.14059234e-01
6.11106232e-02 -3.86489809e-01 4.52610761e-01 -3.22173148e-01
5.97863972e-01 -4.09512430e-01 -1.06040037e+00 3.53021413e-01
-5.15009761e-01 -5.07616885e-02 1.54027253e-01 9.09252241e-02
3.68636072e-01 -1.17384106e-01 7.72288302e-03 -1.41421646e-01
1.11705971e+00 4.72517833e-02 -7.73066163e-01 1.36542618e-01
-1.01194823e+00 -1.29842564e-01 2.79693902e-01 -4.03286815e-01
-7.85772920e-01 -3.04138690e-01 4.61339355e-01 -2.92114168e-01
-8.30730423e-02 -7.61099458e-02 -1.73225179e-01 -2.01797664e-01
9.69573915e-01 -1.45114198e-01 2.85901010e-01 -2.58512169e-01
1.32461876e-01 1.08496916e+00 7.08303392e-01 -4.53593582e-02
1.03833385e-01 4.07698244e-01 1.68465436e-01 -9.11542714e-01
-2.22918078e-01 -2.59127915e-01 -3.71006310e-01 -3.96532595e-01
7.81203330e-01 -6.70018256e-01 -1.44176209e+00 2.83188224e-01
-9.39403415e-01 -2.23589048e-01 -1.85682088e-01 5.85728824e-01
-5.69292784e-01 -2.54873157e-01 -6.90669008e-03 -1.21952438e+00
-2.59377211e-01 -1.16711581e+00 9.79380190e-01 8.16327929e-01
-6.12078488e-01 -7.11839437e-01 -6.39778674e-02 -8.59767124e-02
8.12588036e-01 2.66422391e-01 1.11060119e+00 1.08175524e-01
-3.74379843e-01 -1.31446183e-01 -3.30833405e-01 7.66916871e-02
4.71521020e-01 1.20326476e-02 -1.36367929e+00 -4.90299255e-01
1.57890126e-01 -5.31646013e-02 2.47817114e-02 9.73535597e-01
4.90065843e-01 -9.96068679e-03 -4.04452473e-01 6.46227360e-01
1.32506108e+00 6.16402626e-01 6.43737614e-01 2.93069243e-01
1.96576044e-01 7.01355219e-01 3.26927871e-01 5.03556013e-01
2.04244986e-01 9.00058150e-01 8.26840475e-02 -1.90151840e-01
-2.08435610e-01 -1.04660736e-02 -1.30182967e-01 2.07464173e-01
6.44217292e-03 -4.34445769e-01 -9.28321362e-01 -2.48621926e-02
-1.28573322e+00 -5.75090826e-01 -5.02202027e-02 2.73887396e+00
3.93574685e-01 2.12398708e-01 4.93682206e-01 -5.06726578e-02
6.32747054e-01 -3.23368013e-01 -9.82418001e-01 -6.35172546e-01
2.10502911e-02 1.66807815e-01 5.25344133e-01 5.92105269e-01
-4.60387707e-01 3.02364975e-01 8.09284210e+00 -4.66651320e-02
-1.53559983e+00 -2.78521150e-01 4.06151235e-01 -7.25141108e-01
-1.03261620e-01 -2.67279923e-01 -7.75072455e-01 6.38776660e-01
1.14123929e+00 -7.25391865e-01 5.56806624e-01 4.14175093e-01
5.06666660e-01 -7.39294291e-01 -1.34999573e+00 1.50472200e+00
-2.32551172e-02 -7.87617028e-01 -4.25717950e-01 6.04158938e-01
-3.22141200e-02 -3.04656029e-01 5.34097612e-01 -3.87027889e-01
-3.69070649e-01 -1.31965411e+00 4.50827479e-01 7.57524192e-01
1.64358270e+00 -5.53660214e-01 6.33598194e-02 7.41414651e-02
-1.02379000e+00 -2.41600610e-02 -1.21346600e-01 -5.28894544e-01
1.13716029e-01 -1.53662965e-01 -7.11263478e-01 -2.86371917e-01
9.40089345e-01 -2.73689210e-01 -6.93302989e-01 1.20118642e+00
3.98064017e-01 -4.04429995e-02 -3.03810924e-01 -7.68415481e-02
-3.80698204e-01 1.15361521e-02 5.36869884e-01 5.95821142e-01
4.36402321e-01 2.32926562e-01 -5.38159609e-01 6.86206102e-01
5.28344929e-01 -1.55329987e-01 -6.99531198e-01 2.63241470e-01
9.14833248e-01 9.95573044e-01 -3.13060075e-01 8.05366486e-02
-6.94799185e-01 7.48722792e-01 -1.26412332e-01 5.67549944e-01
-3.83698851e-01 -5.82964063e-01 8.31073105e-01 9.27487254e-01
-1.35091683e-02 -3.59351188e-01 -4.25749928e-01 -2.99300730e-01
2.22836792e-01 -6.24602258e-01 -2.81108730e-02 -1.44585216e+00
-6.49235249e-01 9.00445640e-01 5.34196086e-02 -1.29557455e+00
-5.57577670e-01 -7.89035499e-01 -2.78781652e-01 1.82835174e+00
-8.82090092e-01 -5.65882564e-01 -7.63440609e-01 7.56804109e-01
7.87389204e-02 -1.00218929e-01 8.16663504e-01 -1.18850894e-01
-8.29831809e-02 7.29339182e-01 -1.66759610e-01 -6.18474424e-01
9.28546250e-01 -1.18359292e+00 2.60071874e-01 7.07545280e-01
-6.19977176e-01 9.55047786e-01 9.61748481e-01 -3.62236500e-01
-1.47649086e+00 -3.18458647e-01 4.64422882e-01 -9.36064661e-01
1.99929744e-01 -1.92677334e-01 -7.64558613e-01 4.05701905e-01
4.74271029e-01 -3.89752924e-01 8.38734508e-01 2.40378097e-01
-7.37554431e-02 -1.38149679e-01 -9.90895748e-01 8.34085405e-01
7.75649846e-01 -6.26809895e-01 -2.93671370e-01 -1.25110626e-01
5.08016884e-01 -5.66427290e-01 -5.19769847e-01 3.76086272e-02
9.94472921e-01 -1.42759383e+00 3.51942718e-01 -7.05354214e-02
-3.76207888e-01 -3.69295955e-01 -2.15154156e-01 -1.09993231e+00
-6.67340219e-01 -1.07652104e+00 -1.10248968e-01 7.90199041e-01
1.32226691e-01 -1.13596082e+00 5.14044106e-01 1.40251255e+00
-2.54987516e-02 -6.84908628e-01 -8.20744812e-01 -5.98055065e-01
-3.52195740e-01 -2.36015216e-01 4.86236572e-01 1.61328942e-01
4.91249889e-01 5.23608029e-01 9.94496346e-02 6.42348677e-02
4.03821379e-01 -1.83334455e-01 8.48024905e-01 -1.22541332e+00
-2.06413940e-01 -5.40789783e-01 -7.05891967e-01 -1.31473804e+00
-7.68900275e-01 -2.33725950e-01 -2.96408653e-01 -1.31188381e+00
-1.03390065e-03 -1.36701316e-01 1.50852546e-01 -1.16859578e-01
-3.24097387e-02 -2.15270258e-02 2.41320416e-01 5.51392019e-01
-5.85121140e-02 -1.34411857e-01 1.15862834e+00 6.78922713e-01
-8.21243286e-01 1.56311601e-01 -1.11036479e+00 4.52030599e-01
4.41427886e-01 1.92324668e-01 -7.11512029e-01 -4.03421223e-01
3.51996243e-01 1.10948309e-01 2.29680896e-01 -9.89943504e-01
6.50645375e-01 2.43228730e-02 5.26461244e-01 -4.27807927e-01
4.42644894e-01 -8.25540304e-01 1.76568285e-01 1.81988239e-01
1.87352188e-02 6.58098817e-01 6.08544767e-01 2.61466831e-01
2.13420123e-01 2.51376957e-01 9.57460761e-01 3.56440187e-01
-5.10438502e-01 -2.27932677e-01 -5.71595132e-01 -7.69543424e-02
9.86495495e-01 -9.01466250e-01 -8.04263234e-01 -7.30682254e-01
-5.69418132e-01 9.21579357e-03 1.20090878e+00 2.66373098e-01
3.98678452e-01 -9.19701397e-01 -1.11566469e-01 6.49533391e-01
1.89672977e-01 -2.24752933e-01 1.30285889e-01 9.15619135e-01
-4.58785921e-01 8.47902060e-01 -4.29281622e-01 -7.52233267e-01
-1.63701868e+00 4.98608500e-01 6.61429524e-01 5.43892860e-01
-6.29281759e-01 6.25246584e-01 4.78605807e-01 8.52999806e-01
2.06723288e-01 -3.28432649e-01 -1.44911274e-01 -2.74898410e-01
9.92366433e-01 4.66935128e-01 9.58527252e-02 -5.81570804e-01
-4.03968185e-01 6.00009561e-01 1.04412533e-01 -3.57180476e-01
5.75352967e-01 -7.77300894e-01 2.26065174e-01 6.50437772e-01
5.25317371e-01 4.32099164e-01 -1.47064781e+00 -4.96276394e-02
-4.72067595e-01 -5.75553477e-01 1.03819683e-01 -8.70252907e-01
-2.17034459e-01 5.45058608e-01 1.26552069e+00 6.49613917e-01
1.57331491e+00 -1.89814381e-02 1.95079595e-01 -1.94146812e-01
2.99765736e-01 -9.84859884e-01 -1.23197652e-01 6.78555444e-02
7.49691963e-01 -7.60639071e-01 6.65701255e-02 -3.05619836e-01
-6.65259957e-01 9.43527639e-01 7.17252254e-01 3.66295725e-01
6.47362649e-01 7.66152084e-01 2.60178387e-01 -2.74595410e-01
-8.85548055e-01 -4.06827360e-01 4.96447176e-01 9.45609391e-01
7.20841467e-01 -5.96599951e-02 -2.49829218e-01 2.71630466e-01
-5.96379697e-01 4.98426855e-02 4.53294098e-01 1.12555897e+00
-7.18070328e-01 -5.80351353e-01 -5.19177735e-01 7.04511940e-01
-3.56114596e-01 -8.26668218e-02 -3.88287783e-01 6.34786904e-01
6.69299886e-02 1.45517731e+00 7.24309087e-01 -3.43670487e-01
7.33077168e-01 8.51453245e-02 8.81307244e-01 -6.10619605e-01
-1.24300040e-01 4.52514917e-01 -2.60811001e-01 -6.03326142e-01
-7.29749501e-02 -3.23276997e-01 -8.96072268e-01 -4.98201162e-01
-6.22338690e-02 -3.16771358e-01 7.68459618e-01 4.66718554e-01
7.58350074e-01 5.58744669e-01 2.08347887e-01 -7.83006370e-01
-1.55148432e-01 -9.96457636e-01 -1.01850367e+00 1.66467935e-01
6.94292843e-01 -6.21789932e-01 -3.31289738e-01 -7.04395398e-02] | [14.070602416992188, 0.10945285111665726] |
01c25cb7-1bd4-4c97-956a-f4ba9a0c7698 | hawk-an-industrial-strength-multi-label | 2301.06057 | null | https://arxiv.org/abs/2301.06057v1 | https://arxiv.org/pdf/2301.06057v1.pdf | Hawk: An Industrial-strength Multi-label Document Classifier | There are a plethora of methods and algorithms that solve the classical multi-label document classification. However, when it comes to deployment and usage in an industry setting, most, if not all the contemporary approaches fail to address some of the vital aspects or requirements of an ideal solution: i. ability to operate on variable-length texts and rambling documents. ii. catastrophic forgetting problem. iii. modularity when it comes to online learning and updating the model. iv. ability to spotlight relevant text while producing the prediction, i.e. visualizing the predictions. v. ability to operate on imbalanced or skewed datasets. vi. scalability. The paper describes the significance of these problems in detail and proposes a unique neural network architecture that addresses the above problems. The proposed architecture views documents as a sequence of sentences and leverages sentence-level embeddings for input representation. A hydranet-like architecture is designed to have granular control over and improve the modularity, coupled with a weighted loss driving task-specific heads. In particular, two specific mechanisms are compared: Bi-LSTM and Transformer-based. The architecture is benchmarked on some of the popular benchmarking datasets such as Web of Science - 5763, Web of Science - 11967, BBC Sports, and BBC News datasets. The experimental results reveal that the proposed model outperforms the existing methods by a substantial margin. The ablation study includes comparisons of the impact of the attention mechanism and the application of weighted loss functions to train the task-specific heads in the hydranet. | ['Arshad Javeed'] | 2023-01-15 | null | null | null | null | ['document-classification'] | ['natural-language-processing'] | [ 3.12115222e-01 -1.49368510e-01 -1.08414508e-01 -4.87856328e-01
-6.86560988e-01 -2.60208905e-01 6.20460093e-01 3.35397273e-01
-5.64571083e-01 4.66598302e-01 1.79793030e-01 -4.33789760e-01
-3.67100179e-01 -4.10055012e-01 -6.17462754e-01 -7.96341300e-01
3.82979959e-02 5.52966356e-01 1.23668991e-01 -2.19466627e-01
5.83819687e-01 1.90183625e-01 -1.77449477e+00 5.07874310e-01
7.70922542e-01 1.35095465e+00 3.84050876e-01 5.73948681e-01
-5.14384091e-01 8.35534096e-01 -6.44529283e-01 -5.63424528e-01
1.35666609e-01 2.03888807e-02 -8.15542877e-01 -1.85376838e-01
5.77645659e-01 2.53489129e-02 7.94696733e-02 8.05357814e-01
7.95209527e-01 -3.97413261e-02 6.28396094e-01 -1.25275576e+00
-8.16767514e-01 6.04526222e-01 -7.00388908e-01 3.21180582e-01
-5.87794892e-02 4.81555089e-02 1.22905886e+00 -8.99714530e-01
4.61341321e-01 1.17347407e+00 7.15343833e-01 5.75125992e-01
-1.01237035e+00 -4.97728705e-01 4.30629164e-01 3.06221604e-01
-9.03806448e-01 -3.81797522e-01 6.38827562e-01 -4.62397605e-01
1.25854397e+00 2.86988795e-01 2.05902100e-01 1.33434808e+00
6.58587992e-01 6.64159954e-01 9.09107149e-01 -4.38667446e-01
1.95442453e-01 4.53437418e-01 5.33438563e-01 4.85009432e-01
2.82979399e-01 -2.64551997e-01 -7.94654489e-01 -2.74113584e-02
-1.07816219e-01 1.38918489e-01 -1.42725194e-02 -2.93908745e-01
-8.87629747e-01 7.48179793e-01 2.14063749e-01 4.06970233e-01
-3.65111157e-02 -2.13494133e-02 9.12513196e-01 4.36232448e-01
7.79880166e-01 3.72999072e-01 -6.37458444e-01 -3.49259637e-02
-9.08372641e-01 2.29384705e-01 6.79075360e-01 8.16653252e-01
5.08946359e-01 2.45275125e-02 -3.52408797e-01 9.64456260e-01
2.76192993e-01 4.45099585e-02 9.41983700e-01 -5.52514315e-01
9.51843679e-01 7.44788349e-01 -1.72317848e-01 -1.04474699e+00
-6.29756391e-01 -7.39609063e-01 -8.15305293e-01 1.08711831e-01
8.81021544e-02 4.87909056e-02 -8.68940413e-01 1.70223296e+00
2.44181335e-01 -3.21378499e-01 -3.55574526e-02 6.14083052e-01
7.91314185e-01 6.39539242e-01 1.97687641e-01 3.11096311e-02
1.30688417e+00 -1.20643818e+00 -7.52168715e-01 -5.51872492e-01
7.14665055e-01 -6.79794967e-01 1.35723662e+00 3.86387527e-01
-1.06720114e+00 -5.34932792e-01 -1.21288157e+00 -4.23694938e-01
-7.47334838e-01 6.96400777e-02 2.32101023e-01 5.07223308e-01
-1.03875875e+00 7.78912842e-01 -4.48925793e-01 -4.00640011e-01
2.01112300e-01 3.97072762e-01 -1.44184083e-01 9.04723182e-02
-1.33942246e+00 1.25607216e+00 3.69765610e-01 1.34505495e-01
-5.41062295e-01 -6.85916901e-01 -6.00584686e-01 2.78100729e-01
-2.78788097e-02 -6.57925189e-01 1.24226058e+00 -1.06127512e+00
-1.33008301e+00 7.33978152e-01 -4.49803211e-02 -5.44596791e-01
6.70586944e-01 -4.11542177e-01 -4.12601411e-01 -2.18700141e-01
-8.01709965e-02 5.02605796e-01 7.68894613e-01 -1.03571153e+00
-7.92936683e-01 -4.97887880e-01 -1.81557775e-01 4.33362901e-01
-9.09329712e-01 -6.13005608e-02 -1.51125833e-01 -5.17696261e-01
-2.01735601e-01 -6.71868801e-01 1.21171936e-01 -2.06909254e-01
-5.00175118e-01 -5.02429605e-01 1.02108848e+00 -6.33395076e-01
1.40979147e+00 -2.09019899e+00 1.20636329e-01 -1.67861387e-01
6.87683076e-02 2.49351248e-01 -2.96985269e-01 7.17559934e-01
-1.86290681e-01 2.38255411e-01 7.50770569e-02 -7.93137848e-01
1.60740316e-01 -2.15146551e-03 -3.81657451e-01 4.22325373e-01
1.86198130e-01 6.34576201e-01 -5.54741621e-01 -4.73402292e-01
-1.54039264e-02 5.07005095e-01 -3.05535793e-01 1.44528756e-02
-2.84092695e-01 -3.31578329e-02 -2.16914982e-01 4.30072069e-01
3.96222204e-01 -3.21001261e-01 1.78800836e-01 -8.30255449e-02
-2.62159616e-01 5.16772926e-01 -9.47869182e-01 1.62211466e+00
-5.57952523e-01 4.78450805e-01 -8.08721334e-02 -1.00788236e+00
8.99110079e-01 3.30071479e-01 2.58623540e-01 -9.69329059e-01
8.41021016e-02 3.13982248e-01 6.08758256e-03 -5.05900621e-01
7.47443974e-01 -1.39965817e-01 -1.24125322e-03 6.03553236e-01
1.42717987e-01 2.32971191e-01 2.24532291e-01 2.23684102e-01
1.08618438e+00 5.67921847e-02 3.43223512e-02 -2.27325693e-01
5.10805845e-01 -1.05454281e-01 5.01420498e-01 6.27438128e-01
-9.55805406e-02 4.01243001e-01 3.99065942e-01 -7.12046742e-01
-1.16109931e+00 -5.50693810e-01 -1.15400650e-01 1.60533285e+00
-1.31276205e-01 -3.51557195e-01 -4.79380250e-01 -6.64543033e-01
1.81858599e-01 9.18224514e-01 -7.56292403e-01 -3.92209738e-01
-4.78096068e-01 -8.70805919e-01 3.33511204e-01 2.94697464e-01
2.97796667e-01 -1.03205967e+00 -7.38590658e-01 1.97804376e-01
-1.08410597e-01 -7.03715146e-01 -3.52703065e-01 6.32141232e-01
-8.74210894e-01 -7.48052597e-01 -3.70521039e-01 -7.48845279e-01
2.27762386e-01 3.59862238e-01 1.13017941e+00 -1.16577394e-01
-3.39303970e-01 3.62260677e-02 -2.93930590e-01 -5.97939849e-01
-2.09410220e-01 5.36661327e-01 -1.10370312e-02 -9.49873850e-02
4.99281645e-01 -5.42350531e-01 -5.63490570e-01 1.87894002e-01
-9.96935844e-01 7.71313682e-02 7.84694731e-01 9.68623996e-01
3.21018159e-01 2.47587766e-02 6.79706872e-01 -1.06160498e+00
9.16929781e-01 -6.18334591e-01 -1.86035126e-01 3.98646683e-01
-1.05932367e+00 2.07856774e-01 9.59125340e-01 -3.81962329e-01
-9.15202260e-01 -2.46793658e-01 -9.92094129e-02 -2.18635365e-01
1.89120039e-01 3.68136227e-01 4.67093810e-02 2.82515973e-01
5.45804501e-01 2.93351412e-01 -1.00906886e-01 -6.42754376e-01
1.56682342e-01 8.82094264e-01 2.76572630e-02 -3.72285426e-01
4.92200494e-01 2.07251832e-01 -2.24325687e-01 -4.29281920e-01
-1.11440277e+00 -3.97148639e-01 -4.43150818e-01 -9.13740695e-02
5.85358918e-01 -6.61254644e-01 -5.72717488e-01 3.71051729e-01
-1.18664706e+00 -8.04192349e-02 -9.50882733e-02 -9.01582018e-02
-4.21203226e-01 1.72872856e-01 -7.19345510e-01 -8.07655156e-01
-9.35504436e-01 -1.03936636e+00 9.05931413e-01 5.27695306e-02
-2.07288563e-01 -8.58621538e-01 1.42147511e-01 4.73811299e-01
7.10566342e-01 -5.71984872e-02 1.34483171e+00 -1.01384509e+00
-1.32210717e-01 -2.48780161e-01 -1.95866913e-01 5.60463369e-01
-9.76475552e-02 -3.40282987e-03 -1.21218336e+00 -4.95034039e-01
-4.00562584e-02 -5.35042286e-01 1.04601681e+00 9.48241651e-02
1.23356640e+00 -4.76894647e-01 -2.95342833e-01 3.97982389e-01
1.43248093e+00 1.26188710e-01 4.00217474e-01 7.90587366e-01
4.49234396e-01 8.80903184e-01 3.63223672e-01 3.09117943e-01
4.61532444e-01 5.29352427e-01 5.96201599e-01 1.93345889e-01
-8.59705359e-02 -2.41019070e-01 3.80500168e-01 1.16342044e+00
4.75112915e-01 -4.45617169e-01 -9.84225452e-01 4.27308887e-01
-1.88718867e+00 -8.78843844e-01 4.37822938e-02 1.98823297e+00
8.31927717e-01 5.01836956e-01 -2.02713326e-01 3.52947474e-01
6.33867681e-01 3.59523445e-01 -8.06239247e-01 -8.49194705e-01
-1.32128065e-02 -1.12076029e-01 3.17161947e-01 2.06589311e-01
-1.11681819e+00 5.46016097e-01 5.78919744e+00 8.30655992e-01
-1.31347811e+00 3.13829392e-01 8.26921821e-01 -5.16689301e-01
-1.92813590e-01 -3.46166342e-01 -1.04427552e+00 6.72845840e-01
1.25549281e+00 -9.45345238e-02 8.07235837e-02 8.46162975e-01
1.30608007e-01 1.11067496e-01 -1.09454942e+00 7.42828965e-01
2.44115502e-01 -1.19523156e+00 8.12347308e-02 -1.07831769e-01
3.49598646e-01 2.35750675e-01 4.92876112e-01 5.23821473e-01
1.26176134e-01 -1.01296604e+00 1.13012075e+00 3.53493571e-01
5.93925893e-01 -7.63335168e-01 7.84597278e-01 4.55906570e-01
-5.82158148e-01 -5.40340722e-01 -4.09289449e-01 -2.19897702e-02
-8.26605633e-02 8.28544557e-01 -7.98119783e-01 4.82842833e-01
9.03310657e-01 5.22976279e-01 -7.37192392e-01 7.72997856e-01
1.45749167e-01 4.80958939e-01 -8.02623332e-02 -3.40112120e-01
3.11860174e-01 4.29027602e-02 1.95328832e-01 1.33975780e+00
4.35769022e-01 -6.93644047e-01 -1.59240231e-01 5.17113626e-01
-1.02915496e-01 4.30259556e-01 -5.49295008e-01 5.46051934e-02
3.51688176e-01 1.41161799e+00 -3.15272868e-01 -1.52097791e-01
-5.14740705e-01 8.45472157e-01 7.60718882e-01 1.63279831e-01
-7.14505196e-01 -5.60273290e-01 5.34122407e-01 1.18758045e-01
3.41326922e-01 7.78951719e-02 -5.81654787e-01 -8.19947183e-01
2.58142501e-01 -9.17132080e-01 3.84582311e-01 -4.04658169e-01
-1.41402924e+00 8.98631036e-01 -4.32858825e-01 -9.65619266e-01
-7.99851194e-02 -6.35692835e-01 -5.88024259e-01 8.69641602e-01
-1.77888811e+00 -1.04801059e+00 -1.64046846e-02 1.13099009e-01
6.43669307e-01 -1.90169543e-01 8.55990112e-01 6.01163685e-01
-9.80070055e-01 5.74257851e-01 5.23162961e-01 -2.74460435e-01
9.16030765e-01 -1.23770142e+00 3.61009598e-01 4.91942972e-01
-8.19506347e-02 6.00305498e-01 8.86818767e-01 -3.48784536e-01
-1.27719986e+00 -1.21702886e+00 1.34453535e+00 -3.84269655e-01
5.61897933e-01 -7.39078224e-01 -8.40089262e-01 5.28275490e-01
3.78979832e-01 -2.14782640e-01 8.13462198e-01 2.20094010e-01
-5.32432675e-01 -4.79612678e-01 -9.81502712e-01 3.95708948e-01
7.07339346e-01 -3.51354092e-01 -4.96523947e-01 4.83702719e-01
7.51009047e-01 -2.29352370e-01 -5.02227187e-01 1.84275538e-01
6.86512887e-01 -1.04415977e+00 8.04279268e-01 -7.51180172e-01
7.74360001e-01 -2.54224800e-03 -7.12254420e-02 -1.25435925e+00
-4.37571019e-01 -2.99528450e-01 -1.93967849e-01 1.38707435e+00
6.72497571e-01 -5.92963040e-01 5.32360375e-01 4.42423373e-01
-2.78370380e-01 -1.14583421e+00 -1.00299859e+00 -6.21600330e-01
3.50644171e-01 -2.18874082e-01 4.49979961e-01 8.10487807e-01
-8.20904002e-02 6.17670596e-01 -4.62312281e-01 -3.10288310e-01
3.59253377e-01 1.18319072e-01 3.71998399e-01 -1.32647955e+00
-1.56697109e-01 -5.51472306e-01 -5.34783527e-02 -7.03513145e-01
7.98554197e-02 -1.07678795e+00 -1.09201737e-01 -1.68829620e+00
2.65206367e-01 -4.09872741e-01 -6.66925371e-01 4.95809823e-01
-1.50367210e-03 -1.00513019e-01 2.56119400e-01 2.28851691e-01
-7.06495821e-01 4.89925563e-01 9.06935394e-01 -3.35227728e-01
8.58180672e-02 -1.34142563e-01 -8.88220489e-01 3.82631123e-01
1.01429844e+00 -5.96176147e-01 -3.30002755e-01 -8.07788908e-01
5.24791777e-01 -3.29557210e-01 3.21028158e-02 -9.11589384e-01
4.95134622e-01 6.56580254e-02 4.00700539e-01 -5.66354811e-01
2.87599951e-01 -6.89232945e-01 -1.22346312e-01 4.84920859e-01
-7.46754348e-01 5.00187159e-01 1.14129908e-01 6.26422524e-01
-1.03182964e-01 -4.10357714e-01 6.90158010e-01 -9.59976539e-02
-4.61322755e-01 2.64702830e-02 -2.77976364e-01 5.67559451e-02
9.53739464e-01 -1.39545247e-01 -5.80373645e-01 -1.20158881e-01
-3.15314323e-01 4.10932660e-01 1.61473751e-01 7.34388113e-01
3.76099616e-01 -1.15997922e+00 -5.15496135e-01 1.88552782e-01
1.67983860e-01 -2.31875509e-01 2.72989750e-01 7.78134644e-01
-2.60234892e-01 7.86315620e-01 -6.47745654e-02 -2.82491684e-01
-1.12207878e+00 5.88773131e-01 3.19542289e-01 -7.81949520e-01
-4.55712706e-01 7.36238658e-01 -3.43093015e-02 -5.57535112e-01
7.31929243e-01 -2.00631022e-01 -2.85553604e-01 3.44889373e-01
5.26641846e-01 2.72432417e-01 5.83853543e-01 -3.80772442e-01
-2.95688152e-01 2.16574371e-01 -5.34379900e-01 1.70152292e-01
1.59105051e+00 -2.70904154e-01 -1.37730598e-01 8.17677081e-01
1.28236592e+00 -5.16282260e-01 -9.31308448e-01 -3.40784580e-01
4.52492803e-01 -9.33922455e-02 1.27980471e-01 -1.09196365e+00
-9.09754992e-01 1.22850442e+00 6.95029676e-01 3.03665102e-01
9.07515407e-01 -3.67285758e-01 8.83688033e-01 3.76313627e-01
1.81291476e-01 -1.51840901e+00 1.98641047e-01 7.40655661e-01
7.47025669e-01 -1.02954793e+00 -1.30183116e-01 1.46779194e-01
-6.69389069e-01 1.23422599e+00 6.48818910e-01 2.93834507e-01
4.33708936e-01 2.75676489e-01 1.17046110e-01 -1.73814103e-01
-1.31580400e+00 3.28785151e-01 2.62517184e-01 1.14220783e-01
7.41759777e-01 -3.11216146e-01 -3.60908598e-01 6.06198490e-01
-1.37282118e-01 -1.78667367e-01 1.98951825e-01 8.89015615e-01
-5.85258842e-01 -1.03802371e+00 -1.19107611e-01 7.73927331e-01
-7.00473487e-01 -1.66678324e-01 -3.02528977e-01 5.07741272e-01
2.74815947e-01 8.37019801e-01 2.13614795e-02 -3.80799651e-01
3.49687994e-01 6.20578229e-01 -2.96353106e-03 -5.29597104e-01
-9.93884265e-01 -2.63194472e-01 -4.68042158e-02 -3.09422314e-01
-2.56679207e-01 -5.08293271e-01 -8.88921499e-01 -2.95577198e-01
-3.03399056e-01 6.76767230e-02 9.32480276e-01 7.88079858e-01
6.79491103e-01 6.71337068e-01 4.79875416e-01 -6.40255749e-01
-1.03493357e+00 -1.20070708e+00 -5.84831119e-01 4.07577455e-01
4.34865385e-01 -6.23966455e-01 -4.57106709e-01 -3.06170821e-01] | [10.398131370544434, 7.849839687347412] |
d9633e26-f479-43e0-b7a7-8cba562d87e6 | satellite-image-search-in-agoraeo | 2208.10830 | null | https://arxiv.org/abs/2208.10830v1 | https://arxiv.org/pdf/2208.10830v1.pdf | Satellite Image Search in AgoraEO | The growing operational capability of global Earth Observation (EO) creates new opportunities for data-driven approaches to understand and protect our planet. However, the current use of EO archives is very restricted due to the huge archive sizes and the limited exploration capabilities provided by EO platforms. To address this limitation, we have recently proposed MiLaN, a content-based image retrieval approach for fast similarity search in satellite image archives. MiLaN is a deep hashing network based on metric learning that encodes high-dimensional image features into compact binary hash codes. We use these codes as keys in a hash table to enable real-time nearest neighbor search and highly accurate retrieval. In this demonstration, we showcase the efficiency of MiLaN by integrating it with EarthQube, a browser and search engine within AgoraEO. EarthQube supports interactive visual exploration and Query-by-Example over satellite image repositories. Demo visitors will interact with EarthQube playing the role of different users that search images in a large-scale remote sensing archive by their semantic content and apply other filters. | ['Volker Markl', 'Begüm Demir', 'Jorge-Arnulfo Quiané-Ruiz', 'Marcela Charfuelan', 'Holmer Hemsen', 'Eleni Tzirita Zacharatou', 'Pavel Dushev', 'Ahmet Kerem Aksoy'] | 2022-08-23 | null | null | null | null | ['content-based-image-retrieval'] | ['computer-vision'] | [-5.40670276e-01 -5.40889323e-01 -6.30545150e-03 -2.74756700e-01
-8.43882442e-01 -8.44829142e-01 5.88835299e-01 4.54054207e-01
-8.06138039e-01 3.26384515e-01 2.59757251e-01 -4.17863339e-01
-3.11803609e-01 -1.26423955e+00 -1.60446897e-01 -6.13211274e-01
-7.80665755e-01 5.42926788e-01 4.66734141e-01 -4.79843259e-01
1.87560290e-01 1.05947495e+00 -2.11490512e+00 -2.78258715e-02
4.35500205e-01 1.00516975e+00 5.62690198e-01 6.70025170e-01
5.50733358e-02 -4.56671268e-02 -2.97659695e-01 3.43609005e-02
5.24590433e-01 1.84699088e-01 -7.23963976e-01 -6.82948351e-01
5.19589365e-01 -8.09373200e-01 -4.29355472e-01 9.91866231e-01
8.94820869e-01 2.97531545e-01 2.82530248e-01 -1.33454394e+00
-3.17743838e-01 1.02921352e-01 -4.39126901e-02 5.38740277e-01
4.22889024e-01 2.25618966e-02 1.26197052e+00 -8.90909493e-01
8.05479646e-01 1.07439470e+00 7.54231811e-01 3.71988267e-02
-1.11111236e+00 -5.35378814e-01 -6.11108065e-01 4.48470145e-01
-1.87955344e+00 -4.09369886e-01 5.49176857e-02 -2.00831592e-01
1.11483276e+00 7.54914224e-01 7.19416320e-01 2.09302425e-01
-5.78318201e-02 2.12265968e-01 8.13091516e-01 -2.05269635e-01
3.47782969e-01 1.31042555e-01 -2.13527650e-01 6.11644983e-01
1.65762939e-02 1.25108257e-01 -4.81266290e-01 -7.06478834e-01
4.61588711e-01 4.81314622e-02 -3.20808530e-01 -2.98410267e-01
-1.01876998e+00 1.02298617e+00 1.06568384e+00 1.97583348e-01
-3.05293202e-01 1.05422974e-01 3.80115658e-01 5.59450269e-01
3.10970515e-01 4.04238343e-01 1.36490921e-02 3.59473564e-02
-1.29317117e+00 4.99764442e-01 8.10033441e-01 5.03202438e-01
1.11228299e+00 -5.22954583e-01 3.77891660e-01 5.95358610e-01
5.93768418e-01 7.36195326e-01 5.04574656e-01 -8.87250066e-01
1.29843876e-01 3.60953778e-01 3.16204369e-01 -1.40444243e+00
-4.01510537e-01 -6.88773841e-02 -5.56753099e-01 5.19830585e-01
-9.84483883e-02 5.56939006e-01 -5.06658614e-01 1.06636178e+00
7.03277707e-01 -3.58034760e-01 -9.99871194e-02 1.22246873e+00
7.42882073e-01 8.22306097e-01 -4.54891985e-03 5.65485716e-01
1.57427490e+00 -2.40755871e-01 -3.75679314e-01 3.26534174e-02
8.34780753e-01 -5.72859108e-01 1.08636856e+00 1.47968143e-01
-6.33949578e-01 -2.45141566e-01 -1.10216761e+00 -2.33737409e-01
-1.02423930e+00 -1.38157248e-01 4.22319382e-01 3.99309009e-01
-1.52129507e+00 6.81849718e-01 -9.06343877e-01 -8.00179899e-01
1.59166113e-01 2.73080021e-01 -5.30706406e-01 1.10616378e-01
-1.48913956e+00 8.91507506e-01 5.30590117e-01 -1.27279192e-01
-8.60337257e-01 -4.31864768e-01 -7.13386476e-01 4.65339720e-01
-3.07301998e-01 -1.41596064e-01 8.35333824e-01 -1.07710548e-01
-5.52762628e-01 1.02169192e+00 2.89485037e-01 -5.73017180e-01
2.06918150e-01 -1.89059794e-01 -6.00528300e-01 5.97062826e-01
3.07154119e-01 1.00688672e+00 3.91270429e-01 -7.63110399e-01
-7.57266700e-01 -5.59251726e-01 -4.63037640e-02 2.12611169e-01
-9.25408304e-01 2.03945309e-01 -1.88189194e-01 -4.48134214e-01
2.83923775e-01 -9.15994763e-01 -2.67846882e-01 9.92725968e-01
2.87738264e-01 -1.34249613e-01 1.19987500e+00 -6.92079604e-01
1.37264812e+00 -2.31967759e+00 -3.16428512e-01 5.11212766e-01
6.46479875e-02 1.50347948e-01 -8.97995904e-02 9.39184785e-01
1.83463961e-01 3.41345847e-01 -8.93545896e-02 -7.29072094e-02
1.90637454e-01 2.52569646e-01 -4.41024750e-01 7.08808959e-01
-3.83941710e-01 3.00057709e-01 -1.05631924e+00 -6.47418320e-01
1.88891232e-01 4.62475449e-01 -1.97551638e-01 7.60206059e-02
-1.23476416e-01 -9.21252146e-02 -4.67563033e-01 8.20430756e-01
9.81148660e-01 -3.47996175e-01 8.77371356e-02 -8.56060162e-02
-6.12237275e-01 3.11608523e-01 -1.14823067e+00 1.70910776e+00
-4.22400832e-01 9.08587515e-01 2.51105845e-01 -4.44783419e-01
9.76345658e-01 8.47911015e-02 1.75825372e-01 -8.05350363e-01
-4.93913859e-01 5.55375934e-01 -8.17112923e-01 -3.30248088e-01
1.02260888e+00 2.46576801e-01 7.60602504e-02 6.59254193e-01
-4.06068325e-01 -1.35349423e-01 1.11974187e-01 3.59146118e-01
8.45248282e-01 -1.30425125e-01 1.96898207e-01 -5.11927664e-01
3.65711629e-01 3.35489720e-01 -1.20097838e-01 7.82544196e-01
-3.95716280e-02 5.38068831e-01 -2.84227908e-01 -9.31994021e-01
-1.31183004e+00 -8.96461487e-01 -8.07931483e-01 1.31505013e+00
1.48825511e-01 -6.42042756e-01 -1.15249172e-01 -2.54985660e-01
3.79952580e-01 2.56658375e-01 -3.65670651e-01 1.66801319e-01
-1.07108340e-01 -5.46697021e-01 6.30713224e-01 1.66173473e-01
7.32849002e-01 -1.03251004e+00 -1.11204898e+00 1.60286739e-01
-3.13516557e-01 -5.91779470e-01 -1.48774892e-01 -5.77185676e-02
-8.54073465e-01 -6.98868990e-01 -7.46017158e-01 -5.06482005e-01
2.56569237e-01 6.88894391e-01 1.07098591e+00 2.29127884e-01
-6.73936009e-01 2.92508394e-01 -5.97007036e-01 1.85658202e-01
-1.93328455e-01 3.35519612e-01 9.77433994e-02 -5.46191335e-01
4.22278404e-01 -7.96448410e-01 -1.19232070e+00 4.97485965e-01
-1.32453036e+00 -5.49782991e-01 1.55460328e-01 4.64609772e-01
4.41195130e-01 9.31706354e-02 -1.44374818e-01 2.48903379e-01
3.29202652e-01 -8.41376364e-01 -1.10757184e+00 2.14297444e-01
-7.09689379e-01 1.86861992e-01 1.01976164e-01 1.41086951e-01
-3.43734533e-01 1.11092634e-01 -1.28286511e-01 -6.22448996e-02
7.39231415e-04 9.82321441e-01 1.40984312e-01 -4.43887115e-01
8.77332628e-01 5.08465707e-01 9.95939672e-02 -8.49666834e-01
3.01633000e-01 1.33332050e+00 5.89073956e-01 -6.87486604e-02
7.60330498e-01 9.40071404e-01 -6.77418336e-02 -8.86230171e-01
-3.19779336e-01 -7.73111641e-01 -4.12213504e-01 -1.18301705e-01
7.17039406e-01 -1.25784683e+00 -7.23114967e-01 1.36832252e-01
-7.62724876e-01 1.65986437e-02 2.08528452e-02 3.27248216e-01
-1.86022684e-01 5.40597618e-01 -7.86220878e-02 -5.62297761e-01
-6.87840581e-01 -7.40138292e-01 1.25768578e+00 1.76023111e-01
4.73313555e-02 -6.09506309e-01 5.13335466e-01 -7.27795884e-02
6.57993078e-01 1.81623429e-01 4.68084186e-01 -5.62404692e-01
-8.23631942e-01 -3.42178941e-01 -4.60000634e-01 -6.71204031e-02
-1.32061511e-01 -2.85380810e-01 -7.91284561e-01 -8.18279862e-01
-4.98051643e-01 -4.65146601e-01 7.80447602e-01 -3.17459315e-01
1.01661909e+00 -3.56579840e-01 -4.32158589e-01 8.52081716e-01
1.79453540e+00 -2.91265130e-01 8.57369781e-01 1.33475029e+00
4.28473689e-02 4.80102181e-01 7.72660673e-01 1.11103463e+00
4.95497644e-01 7.83639848e-01 9.27414775e-01 -3.76268178e-02
3.52241814e-01 -7.63133839e-02 1.35698944e-01 1.02933206e-01
2.31123716e-01 -2.65228927e-01 -1.25367224e+00 9.15200889e-01
-1.57913530e+00 -9.81975973e-01 3.63932326e-02 2.47464752e+00
7.25046396e-01 -3.75481158e-01 4.20328900e-02 -3.21307242e-01
6.06028557e-01 4.66909200e-01 -2.10928828e-01 3.83266062e-02
-1.62987441e-01 -6.47345558e-02 9.42111254e-01 3.12208027e-01
-1.27732837e+00 6.88557506e-01 6.18039989e+00 7.32044876e-01
-1.22879672e+00 1.74434837e-02 7.31985122e-02 -1.07421890e-01
-4.96492267e-01 7.50491768e-02 -8.80057096e-01 3.77298027e-01
1.17957795e+00 -1.69173062e-01 2.93735147e-01 9.86888766e-01
1.44885957e-01 -2.84397662e-01 -4.94367301e-01 8.82154763e-01
-3.04778844e-01 -1.49852061e+00 1.46938026e-01 5.64783871e-01
6.86445311e-02 7.77189672e-01 -5.41246571e-02 -2.84759581e-01
4.05890912e-01 -8.42429817e-01 6.78885460e-01 4.72846091e-01
1.00881362e+00 -7.81173646e-01 3.82452339e-01 1.06923506e-01
-1.47418392e+00 -1.07710473e-01 -7.13415265e-01 1.90722734e-01
1.60506703e-02 3.10747206e-01 -6.89968228e-01 2.49168798e-01
1.54284370e+00 4.59710091e-01 -6.68094218e-01 1.52215052e+00
3.20839107e-01 1.61809862e-01 -1.04894900e+00 6.88036755e-02
4.54526752e-01 3.24793980e-02 6.74907684e-01 1.18731713e+00
8.50763440e-01 6.88911229e-02 9.76769552e-02 4.51994449e-01
1.43942863e-01 3.89722572e-03 -8.50306511e-01 -8.96636993e-02
9.37956095e-01 1.30501676e+00 -5.79914987e-01 -2.37093836e-01
-2.32273281e-01 9.40858543e-01 -5.83984982e-03 -7.82755315e-02
-4.16851729e-01 -8.52209687e-01 8.10634673e-01 3.67766678e-01
6.25861228e-01 -7.26064980e-01 4.49439079e-01 -9.22345996e-01
-9.57580879e-02 -8.90817881e-01 6.84606254e-01 -9.81616259e-01
-7.59812295e-01 7.29507327e-01 1.40201375e-01 -1.61986184e+00
-3.81664410e-02 -2.77687103e-01 -2.44156808e-01 8.63747835e-01
-1.54987538e+00 -9.47804153e-01 -6.25719488e-01 6.45374954e-01
-1.54150590e-01 -5.17901629e-02 1.19022310e+00 4.42073017e-01
2.32797697e-01 1.78083345e-01 9.17670190e-01 -3.18788551e-02
6.57656252e-01 -1.07598424e+00 8.25731397e-01 3.80425900e-01
2.07239777e-01 5.50155103e-01 7.58187056e-01 -6.28767669e-01
-1.33084893e+00 -6.80636644e-01 1.11397243e+00 6.59564394e-04
9.80143249e-01 -4.71245438e-01 -1.13908267e+00 2.40481123e-01
9.50677469e-02 2.36580953e-01 7.94522524e-01 -3.58611554e-01
-4.76730138e-01 -3.31731647e-01 -1.26021945e+00 4.40257877e-01
6.25567198e-01 -9.73657370e-01 -3.73081803e-01 6.70313179e-01
4.17695194e-01 -4.69787940e-02 -1.19542360e+00 -2.52511315e-02
7.63777316e-01 -1.06769633e+00 1.29011750e+00 -2.45347634e-01
-8.32029209e-02 -7.14135885e-01 -4.93768334e-01 -8.55330229e-01
-2.49501377e-01 -3.68160307e-01 3.82033318e-01 9.00689960e-01
9.49646905e-02 -5.84854364e-01 4.54738677e-01 5.36696076e-01
3.35578740e-01 8.60004053e-02 -1.20628202e+00 -9.55806315e-01
-3.96362305e-01 4.60669980e-04 1.11989379e+00 7.98910141e-01
-5.83413839e-02 -4.77400631e-01 -3.33277434e-01 7.01775134e-01
7.29035497e-01 3.65690827e-01 5.23216307e-01 -1.58784914e+00
-4.26823720e-02 6.90146815e-03 -7.42256522e-01 -7.95111299e-01
-4.12472367e-01 -8.37928832e-01 -5.25749847e-02 -1.26307380e+00
-5.31887915e-03 -6.98420882e-01 -1.39697388e-01 6.66557848e-01
3.47266346e-01 6.29625440e-01 2.21393883e-01 1.07862878e+00
-6.44653022e-01 6.31395936e-01 4.56558198e-01 -3.35930176e-02
1.09974235e-01 -3.90213132e-01 -2.01978087e-02 1.44643649e-01
6.49154067e-01 -8.37193370e-01 1.56423464e-01 -5.97531438e-01
7.64326036e-01 1.22111244e-02 8.60605896e-01 -1.36360848e+00
4.97538298e-01 2.44922712e-01 2.21476376e-01 -8.89733553e-01
1.62095100e-01 -1.10838747e+00 4.49938565e-01 7.32221842e-01
-1.77853674e-01 3.50223482e-01 1.48450747e-01 4.33809221e-01
-3.52830529e-01 -3.71498108e-01 5.56679726e-01 -2.18932375e-01
-1.11442447e+00 3.10588866e-01 -5.05844355e-01 -5.10133326e-01
5.31171739e-01 -1.17373332e-01 -3.13651025e-01 -5.55846274e-01
-8.76119733e-01 4.53860313e-01 8.89824748e-01 1.95297480e-01
6.18069112e-01 -1.38342488e+00 -6.05454922e-01 2.19780311e-01
5.11757016e-01 -3.57751817e-01 4.17656451e-02 2.78396189e-01
-1.23048091e+00 3.37206215e-01 -4.79907632e-01 -5.60379088e-01
-1.41768408e+00 2.56072342e-01 4.15217698e-01 8.59143659e-02
-8.02928030e-01 6.11649513e-01 -5.49808562e-01 -4.81152475e-01
2.25813329e-01 2.71144122e-01 -1.01525582e-01 4.73683298e-01
1.05720770e+00 4.36712921e-01 2.53255606e-01 -5.20954013e-01
-6.38085723e-01 3.07722390e-01 1.71861991e-01 -5.60342848e-01
1.60340834e+00 -5.99212348e-01 -3.59881669e-01 1.89008228e-02
1.33817041e+00 -3.51808995e-01 -1.03754449e+00 -2.30510965e-01
1.63293600e-01 -9.15196180e-01 4.77071434e-01 -4.38265920e-01
-7.92451143e-01 7.58572638e-01 1.43788528e+00 4.24563855e-01
1.04919076e+00 -4.21547219e-02 8.61205637e-01 9.64417100e-01
4.79440570e-01 -9.28230405e-01 -3.89427245e-01 2.37739593e-01
9.58034635e-01 -1.29182029e+00 3.77660155e-01 4.57947373e-01
-8.21528956e-02 1.29902184e+00 2.66388636e-02 1.01801746e-01
8.70966315e-01 -7.63770100e-03 2.66936898e-01 -4.80220020e-01
-3.78263950e-01 -4.76109862e-01 1.74292829e-02 5.40508866e-01
-5.35955876e-02 -2.09115580e-01 -2.83420295e-01 -4.69586879e-01
-1.76962852e-01 -4.70789149e-02 5.37802242e-02 1.23171294e+00
-1.07852578e+00 -1.17082536e+00 -8.08271110e-01 7.08937123e-02
-1.49340495e-01 -5.65036118e-01 -9.78138447e-02 6.15718544e-01
-3.87119949e-01 5.27828813e-01 3.74781013e-01 -2.23461211e-01
-1.61430761e-01 -3.25317919e-01 -2.71012098e-01 -9.40731466e-02
-5.89813113e-01 -3.11008275e-01 2.37554274e-02 -7.15088785e-01
-2.18473271e-01 -7.14240611e-01 -1.10424554e+00 -4.10883725e-01
-1.01275407e-01 6.32763028e-01 1.19650519e+00 2.88749129e-01
7.34452903e-01 -7.16724694e-01 7.33907998e-01 -1.04211664e+00
-8.36782694e-01 -6.34872019e-01 -8.36939275e-01 1.15873612e-01
5.44068277e-01 -2.80900538e-01 -5.27268410e-01 -2.62765616e-01] | [7.763664245605469, -1.9286174774169922] |
6ad47fa9-a1b9-4102-aad5-7cb6b29d653c | adaptive-temporal-encoding-network-for-video | 1808.00661 | null | http://arxiv.org/abs/1808.00661v2 | http://arxiv.org/pdf/1808.00661v2.pdf | Adaptive Temporal Encoding Network for Video Instance-level Human Parsing | Beyond the existing single-person and multiple-person human parsing tasks in
static images, this paper makes the first attempt to investigate a more
realistic video instance-level human parsing that simultaneously segments out
each person instance and parses each instance into more fine-grained parts
(e.g., head, leg, dress). We introduce a novel Adaptive Temporal Encoding
Network (ATEN) that alternatively performs temporal encoding among key frames
and flow-guided feature propagation from other consecutive frames between two
key frames. Specifically, ATEN first incorporates a Parsing-RCNN to produce the
instance-level parsing result for each key frame, which integrates both the
global human parsing and instance-level human segmentation into a unified
model. To balance between accuracy and efficiency, the flow-guided feature
propagation is used to directly parse consecutive frames according to their
identified temporal consistency with key frames. On the other hand, ATEN
leverages the convolution gated recurrent units (convGRU) to exploit temporal
changes over a series of key frames, which are further used to facilitate the
frame-level instance-level parsing. By alternatively performing direct feature
propagation between consistent frames and temporal encoding network among key
frames, our ATEN achieves a good balance between frame-level accuracy and time
efficiency, which is a common crucial problem in video object segmentation
research. To demonstrate the superiority of our ATEN, extensive experiments are
conducted on the most popular video segmentation benchmark (DAVIS) and a newly
collected Video Instance-level Parsing (VIP) dataset, which is the first video
instance-level human parsing dataset comprised of 404 sequences and over 20k
frames with instance-level and pixel-wise annotations. | ['Liang Lin', 'Qixian Zhou', 'Ke Gong', 'Xiaodan Liang'] | 2018-08-02 | null | null | null | null | ['human-parsing'] | ['computer-vision'] | [ 4.63901907e-01 -8.46436247e-02 -2.30406567e-01 -3.89262646e-01
-7.98941493e-01 -4.99422550e-01 4.80010390e-01 -7.53976852e-02
-5.32707810e-01 4.34713840e-01 9.93865207e-02 -1.45783544e-01
2.76038289e-01 -8.32058191e-01 -8.68590772e-01 -5.14539838e-01
-1.90555602e-01 -1.18382432e-01 5.69057107e-01 1.16885282e-01
-1.37187064e-01 2.23583624e-01 -1.35498428e+00 2.82105744e-01
5.10199666e-01 1.15625060e+00 1.29753143e-01 9.38400686e-01
-9.99850929e-02 9.46229517e-01 -5.23920834e-01 -4.78958875e-01
4.06689793e-01 -6.11364067e-01 -1.05420256e+00 4.67630923e-01
6.55230463e-01 -7.06786096e-01 -4.46842283e-01 9.37142432e-01
2.87354857e-01 3.63481939e-01 -5.87281249e-02 -1.23703897e+00
-5.16668260e-01 5.42886615e-01 -7.55279183e-01 5.84311426e-01
5.32867134e-01 5.22479713e-01 1.00035179e+00 -4.37510818e-01
7.03949273e-01 1.39242649e+00 5.63672841e-01 4.92058456e-01
-7.97033906e-01 -6.93287313e-01 7.27280974e-01 9.08834189e-02
-8.59253585e-01 -1.45622581e-01 6.43985868e-01 -3.55743915e-01
8.22854161e-01 -9.82116610e-02 1.00966132e+00 1.05440354e+00
1.42386407e-01 1.09442186e+00 8.06524217e-01 -3.92394178e-02
5.05418070e-02 -6.99989915e-01 4.94571447e-01 9.79328990e-01
-3.32471728e-02 -2.21638177e-02 -4.89855617e-01 4.16377842e-01
1.12913668e+00 1.20401837e-01 -2.29524061e-01 1.38131171e-01
-1.46188605e+00 6.50137603e-01 4.75603998e-01 2.89110802e-02
-6.33514404e-01 4.55544859e-01 5.78865409e-01 -1.12404682e-01
2.54422426e-01 -2.74622124e-02 -4.91333097e-01 -3.03753048e-01
-1.16328228e+00 2.07691222e-01 5.17541647e-01 1.08232749e+00
7.26760447e-01 -2.03218986e-03 -6.70756042e-01 6.17671788e-01
2.82408595e-02 2.09205419e-01 3.54938209e-01 -1.47288752e+00
5.45876920e-01 6.59460843e-01 -4.07744274e-02 -9.66227770e-01
-3.91457856e-01 -4.97801183e-03 -7.52215445e-01 -2.90094018e-01
5.62189758e-01 -2.89995849e-01 -1.21496868e+00 1.86702108e+00
4.94824469e-01 6.63822591e-01 -1.35422617e-01 1.06080425e+00
1.14631259e+00 8.37058604e-01 5.62604070e-01 -3.43745023e-01
1.90064847e+00 -1.41221666e+00 -5.85878313e-01 -2.49050289e-01
2.29260370e-01 -3.58322978e-01 8.46628726e-01 7.81018808e-02
-1.37421501e+00 -1.00598240e+00 -7.53554821e-01 -2.84893394e-01
-1.82338789e-01 -5.11795580e-02 7.95682967e-01 3.51617217e-01
-9.46645021e-01 5.35328567e-01 -1.03374708e+00 -3.75101805e-01
5.56628764e-01 1.25839055e-01 -3.35930705e-01 -4.24962305e-02
-1.15554750e+00 5.78159355e-02 3.97343397e-01 4.10743058e-01
-8.18830967e-01 -7.67115116e-01 -1.14193916e+00 -4.80105318e-02
5.30059457e-01 -8.90670478e-01 1.19851971e+00 -1.17313969e+00
-1.32522368e+00 7.34418392e-01 -3.94199759e-01 -6.12810493e-01
6.36939526e-01 -3.65237504e-01 -1.93097934e-01 5.73307276e-01
3.33611429e-01 1.18204105e+00 8.85572016e-01 -8.89071643e-01
-1.02346838e+00 -1.04118474e-01 4.55795556e-01 6.32995889e-02
2.89629344e-02 2.45411828e-01 -1.27380812e+00 -7.94562876e-01
-5.69292605e-02 -8.41454208e-01 -1.36553779e-01 -2.46651456e-01
-5.96195102e-01 -3.87403965e-01 8.25326025e-01 -1.01114798e+00
1.41295540e+00 -1.99365020e+00 2.77739525e-01 -1.40055001e-01
2.42605120e-01 1.25592142e-01 -3.15785646e-01 -1.51594833e-01
-1.43769339e-01 2.07646042e-01 -2.48801976e-01 -4.83201534e-01
-2.77765781e-01 3.02783221e-01 2.18207929e-02 1.80224538e-01
4.75502998e-01 1.34421039e+00 -9.78846073e-01 -7.09284127e-01
2.66608357e-01 6.12215042e-01 -6.31099463e-01 2.62945861e-01
-2.51316279e-01 5.02062857e-01 -6.10714495e-01 8.77723932e-01
3.68391007e-01 -4.98761624e-01 5.48518933e-02 -4.77339923e-01
-7.56186098e-02 -2.24221479e-02 -1.01335740e+00 1.88039029e+00
-9.29914787e-02 5.46731591e-01 -9.83726457e-02 -8.35277855e-01
3.66666764e-01 2.82610297e-01 8.72725546e-01 -8.85830522e-01
1.57527834e-01 -3.40739906e-01 -3.34473312e-01 -4.80184883e-01
7.15918839e-01 4.36713547e-01 -3.80033791e-01 1.76685959e-01
8.85410532e-02 4.07009095e-01 5.90864480e-01 2.66987205e-01
1.21156359e+00 5.70549846e-01 2.59615421e-01 -6.61923066e-02
6.50484562e-01 -2.22403273e-01 1.02399373e+00 7.26109445e-01
-7.87724316e-01 8.38080406e-01 6.57201886e-01 -8.03776741e-01
-8.15306783e-01 -8.24577093e-01 2.31708348e-01 1.29536819e+00
4.53929901e-01 -5.25792122e-01 -1.12247813e+00 -8.10485303e-01
-3.37463826e-01 1.88414723e-01 -7.48489082e-01 2.65480876e-01
-1.10702372e+00 -6.20551884e-01 5.02931595e-01 8.13311517e-01
1.08538485e+00 -1.32614124e+00 -9.79397953e-01 4.97742623e-01
-5.51266074e-01 -1.77270269e+00 -8.68380368e-01 -3.78491394e-02
-6.79884672e-01 -1.28835762e+00 -8.57523143e-01 -8.28204453e-01
4.82371688e-01 1.81570843e-01 1.37959492e+00 1.88601449e-01
-3.44883084e-01 5.62312186e-01 -5.25282621e-01 3.04000288e-01
4.49746400e-02 -2.03954708e-02 -3.84470791e-01 3.05269379e-02
1.53804138e-01 -3.07430565e-01 -9.70399261e-01 3.49550575e-01
-8.97512376e-01 3.38749647e-01 3.85518223e-01 7.15317845e-01
9.89354849e-01 5.07646147e-03 2.91340172e-01 -8.27520669e-01
1.34409055e-01 -2.45573252e-01 -5.06525099e-01 3.47615272e-01
1.21633515e-01 -2.88859576e-01 5.01231253e-01 -3.72553736e-01
-9.36032951e-01 1.55949920e-01 -9.17188376e-02 -4.82508391e-01
-2.13923737e-01 5.09677939e-02 -1.18591972e-01 2.58254737e-01
-2.14411672e-02 3.33145201e-01 -3.90545279e-01 -1.09279208e-01
4.27943796e-01 1.03305109e-01 1.06581151e+00 -8.29301417e-01
4.91733134e-01 3.01281840e-01 -3.05655688e-01 -5.10838747e-01
-1.06166780e+00 -3.60290706e-01 -7.56036878e-01 -3.01942110e-01
1.69598925e+00 -1.12948143e+00 -8.51281166e-01 6.25942886e-01
-1.20469975e+00 -7.13733554e-01 -2.11720526e-01 8.02925304e-02
-6.17777526e-01 5.62216520e-01 -1.09357536e+00 -4.28081900e-01
-4.89973545e-01 -1.42748034e+00 1.52550173e+00 6.18685246e-01
-9.36314017e-02 -8.23715806e-01 -5.01127899e-01 5.53330541e-01
-9.39632058e-02 5.24753809e-01 6.15846753e-01 -2.82061726e-01
-8.36036861e-01 2.05355436e-02 -4.92500454e-01 1.84798479e-01
4.51175235e-02 1.89881846e-01 -6.82993591e-01 -2.17323601e-01
-4.33743328e-01 -8.90562236e-02 9.39625025e-01 7.43382096e-01
1.48374295e+00 -1.30578801e-01 -1.52184889e-01 8.19769442e-01
1.17022491e+00 2.80471087e-01 6.74377978e-01 3.91573459e-01
1.05183089e+00 3.78091305e-01 6.33173585e-01 3.00463140e-01
6.89201951e-01 6.09934568e-01 1.54689938e-01 -2.21017793e-01
-3.44402194e-01 -2.46235132e-01 3.78708810e-01 5.39772928e-01
-3.90782624e-01 -2.28405610e-01 -5.39295197e-01 3.63639474e-01
-1.87034011e+00 -1.22269619e+00 1.54726610e-01 1.86108899e+00
6.17297888e-01 1.81468293e-01 5.53302586e-01 -1.14565164e-01
1.00347960e+00 3.95304918e-01 -6.14646494e-01 -6.53838590e-02
-8.38460401e-02 1.77342117e-01 4.47547883e-01 3.28640670e-01
-1.65162802e+00 1.22812605e+00 5.66799831e+00 6.41281843e-01
-8.74450326e-01 -5.87570891e-02 1.00138497e+00 -8.94506201e-02
1.27898797e-01 -1.43708900e-01 -9.24789131e-01 7.12081492e-01
7.75343180e-01 3.64124030e-01 4.20087010e-01 5.58143437e-01
4.96149749e-01 -1.38643980e-01 -1.08127105e+00 9.85518277e-01
-2.18088582e-01 -1.25370753e+00 1.65553048e-01 -6.69639260e-02
5.75813115e-01 -3.52297425e-01 -2.03261361e-01 4.84992504e-01
1.85235441e-01 -9.14078236e-01 9.60914791e-01 4.29358035e-01
6.24455333e-01 -7.61160612e-01 5.04692137e-01 -6.05895109e-02
-1.93030834e+00 -6.43533319e-02 -2.14858390e-02 9.87684205e-02
7.42474318e-01 3.45339149e-01 -1.37832269e-01 6.36619568e-01
1.10023916e+00 1.05016053e+00 -6.64706647e-01 5.75249791e-01
-1.67731553e-01 6.55895948e-01 -1.65752992e-01 5.68972230e-01
5.34153223e-01 -2.73036152e-01 1.70473397e-01 1.49743176e+00
5.71614243e-02 5.13740361e-01 6.70007527e-01 7.59469807e-01
-1.39697105e-01 -3.19782972e-01 -5.02909012e-02 -2.76079718e-02
4.24684584e-01 1.49902129e+00 -1.24246633e+00 -6.95165157e-01
-6.59685671e-01 1.33774030e+00 4.25885133e-02 5.77174723e-01
-1.29387295e+00 -9.73147601e-02 9.00256097e-01 -1.21493958e-01
6.19641125e-01 -2.58383065e-01 3.16386111e-02 -1.21228373e+00
2.01912075e-02 -7.72872984e-01 6.78241014e-01 -6.78091168e-01
-1.13144541e+00 6.81555271e-01 -2.03127563e-02 -9.29992795e-01
-2.23399192e-01 -3.12474072e-01 -5.50435483e-01 5.54778814e-01
-1.66197801e+00 -1.29000020e+00 -5.66636741e-01 8.75733733e-01
1.09108770e+00 3.24229479e-01 2.27175653e-01 3.58264416e-01
-1.18346882e+00 6.71166420e-01 -8.53072405e-01 8.22166443e-01
3.51669818e-01 -1.14737391e+00 7.31694520e-01 1.27330232e+00
-7.09836930e-02 5.52458823e-01 2.38311887e-01 -8.20376992e-01
-1.24654651e+00 -1.43996942e+00 5.29043674e-01 -3.82271618e-01
4.55890089e-01 -1.74992025e-01 -8.31317365e-01 8.34348202e-01
1.26567319e-01 4.19537991e-01 3.22590381e-01 -3.83936644e-01
-2.73381770e-01 7.51746371e-02 -8.48878324e-01 4.92673397e-01
1.38461828e+00 -4.22640145e-01 -4.20011789e-01 1.23827413e-01
1.18753028e+00 -6.53701067e-01 -1.07618237e+00 5.13818741e-01
5.86505532e-01 -1.11655164e+00 1.24123108e+00 -4.27975804e-01
8.00074100e-01 -4.33726370e-01 -1.31840646e-01 -5.58459401e-01
-4.24392253e-01 -6.93424344e-01 -1.51423842e-01 1.43519711e+00
-9.64790285e-02 -2.18606740e-01 8.73957634e-01 8.12504590e-01
-1.26043245e-01 -8.93038094e-01 -6.41600013e-01 -4.78729784e-01
-1.97419137e-01 -6.54429138e-01 4.66723472e-01 5.39971709e-01
-5.53916574e-01 1.77714825e-02 -4.89233583e-01 2.07393318e-01
5.62313855e-01 2.01513797e-01 6.93931222e-01 -7.51799166e-01
-3.78245682e-01 -3.61808151e-01 -4.40791875e-01 -1.45421779e+00
3.39816630e-01 -5.20300686e-01 2.12903365e-01 -1.41682434e+00
3.91010404e-01 -1.01979733e-01 -2.04561368e-01 6.18710101e-01
-7.33868182e-01 4.32704061e-01 6.00694716e-01 1.43857345e-01
-1.08252728e+00 1.66976690e-01 1.54369032e+00 -1.80702899e-02
-3.09585899e-01 -2.60556668e-01 -5.83286941e-01 7.97218859e-01
2.80158550e-01 -8.96199197e-02 -2.95513034e-01 -4.34137285e-01
-2.19003573e-01 4.70777035e-01 7.03992665e-01 -1.04720974e+00
3.27728420e-01 -1.09203532e-01 6.15925908e-01 -5.69863021e-01
1.98020518e-01 -5.03692925e-01 8.75776932e-02 3.43547940e-01
-3.16053331e-01 3.79471332e-01 -1.22534344e-02 7.97756612e-01
-3.38017404e-01 3.16183180e-01 7.53364146e-01 -2.99208641e-01
-1.54813087e+00 8.11742365e-01 -5.93653172e-02 2.13260546e-01
1.10898817e+00 -5.44765472e-01 -1.90720677e-01 -1.80958375e-01
-7.38588691e-01 4.45925683e-01 3.55873376e-01 5.90464532e-01
4.79434818e-01 -1.09517252e+00 -4.57503676e-01 6.16759621e-02
-1.94078699e-01 3.07378352e-01 6.09388351e-01 8.53219151e-01
-4.93958771e-01 2.23469868e-01 -1.42188907e-01 -8.86970580e-01
-1.12748921e+00 4.17385161e-01 3.80237669e-01 -2.77872026e-01
-1.10726225e+00 1.00769758e+00 4.52436745e-01 3.35769285e-03
1.07016496e-01 -6.49488866e-01 -2.22533673e-01 1.21995121e-01
6.86342418e-01 2.78658837e-01 -3.10626596e-01 -8.35934997e-01
-2.58041114e-01 7.51489580e-01 5.45791313e-02 1.31008789e-01
1.06157434e+00 -4.03537959e-01 5.37391193e-02 1.25088945e-01
1.25864267e+00 -4.44055885e-01 -2.02620196e+00 -1.88232452e-01
-2.57168323e-01 -3.63590777e-01 -2.41067335e-01 -5.20125151e-01
-1.62802541e+00 6.19999588e-01 3.52790564e-01 -6.45228720e-04
1.43205619e+00 -5.33500463e-02 1.43305075e+00 -1.20024487e-01
3.96354437e-01 -9.20988679e-01 3.10692042e-01 4.61939275e-01
2.34871298e-01 -1.20019948e+00 -2.03930482e-01 -6.02778435e-01
-6.33120239e-01 1.18921590e+00 7.38800526e-01 -2.11984932e-01
2.79516935e-01 2.19002619e-01 -1.12534352e-01 -5.90012819e-02
-4.92557526e-01 -2.88173378e-01 2.92611063e-01 4.34588999e-01
4.79079157e-01 -4.67992015e-02 -1.88075036e-01 9.43652987e-01
-9.18251351e-02 1.28857628e-01 3.68718386e-01 7.95571625e-01
-2.38165453e-01 -8.08674932e-01 -1.52695209e-01 3.03414702e-01
-5.68750501e-01 9.33466759e-03 1.72464564e-01 9.49320138e-01
2.85747916e-01 9.79670227e-01 2.56238818e-01 -2.88189083e-01
2.58650780e-01 -9.09115002e-03 2.88876921e-01 -3.28284502e-01
-8.12658370e-01 1.95370212e-01 -5.61640896e-02 -1.28621590e+00
-8.74934912e-01 -7.60786891e-01 -1.60316133e+00 -3.10977519e-01
2.11598352e-01 -2.49468327e-01 6.42646849e-02 1.01604915e+00
2.36735001e-01 1.00580251e+00 2.88009822e-01 -1.20530057e+00
2.44761243e-01 -4.03307706e-01 -2.88209438e-01 7.67895758e-01
2.54600823e-01 -4.42243695e-01 1.11761382e-02 4.98508126e-01] | [9.194022178649902, -0.0267354566603899] |
2e3c9def-7c66-4510-a562-02c8260b029c | multi-scale-interaction-for-real-time-lidar | 2008.09162 | null | https://arxiv.org/abs/2008.09162v2 | https://arxiv.org/pdf/2008.09162v2.pdf | Multi-scale Interaction for Real-time LiDAR Data Segmentation on an Embedded Platform | Real-time semantic segmentation of LiDAR data is crucial for autonomously driving vehicles, which are usually equipped with an embedded platform and have limited computational resources. Approaches that operate directly on the point cloud use complex spatial aggregation operations, which are very expensive and difficult to optimize for embedded platforms. They are therefore not suitable for real-time applications with embedded systems. As an alternative, projection-based methods are more efficient and can run on embedded platforms. However, the current state-of-the-art projection-based methods do not achieve the same accuracy as point-based methods and use millions of parameters. In this paper, we therefore propose a projection-based method, called Multi-scale Interaction Network (MINet), which is very efficient and accurate. The network uses multiple paths with different scales and balances the computational resources between the scales. Additional dense interactions between the scales avoid redundant computations and make the network highly efficient. The proposed network outperforms point-based, image-based, and projection-based methods in terms of accuracy, number of parameters, and runtime. Moreover, the network processes more than 24 scans per second on an embedded platform, which is higher than the framerates of LiDAR sensors. The network is therefore suitable for autonomous vehicles. | ['Dengxin Dai', 'Yun Liu', 'Xieyuanli Chen', 'Shijie Li', 'Juergen Gall', 'Cyrill Stachniss'] | 2020-08-20 | null | null | null | null | ['real-time-3d-semantic-segmentation'] | ['computer-vision'] | [-1.85082421e-01 -1.64792314e-01 -9.80565883e-03 -4.70213234e-01
-1.54797733e-01 -1.41927645e-01 3.57739091e-01 1.35178104e-01
-9.06667113e-01 4.39338237e-01 -1.04207575e+00 -6.07543707e-01
-8.74618143e-02 -1.46277344e+00 -7.92100191e-01 -4.81501341e-01
2.42207721e-01 9.65138197e-01 1.25788522e+00 -1.93010181e-01
8.42740908e-02 9.12954152e-01 -2.07932782e+00 -2.86426634e-01
1.08637297e+00 1.32776737e+00 4.53100711e-01 2.86876559e-01
-5.80377698e-01 -2.93795049e-01 -2.25530952e-01 -2.35638306e-01
2.99990356e-01 1.95778921e-01 3.68228480e-02 -1.98678643e-01
4.02032435e-01 -3.56634110e-01 -2.45655894e-01 1.02432764e+00
2.95023531e-01 8.11677799e-02 1.81578666e-01 -1.51785421e+00
4.45116729e-01 -8.44776854e-02 -6.26551509e-01 -1.22398689e-01
-3.52402121e-01 6.25877529e-02 4.87382919e-01 -1.13729882e+00
3.29776436e-01 1.24898970e+00 7.27406442e-01 2.61428118e-01
-9.54074919e-01 -9.32270110e-01 1.04184836e-01 7.15700030e-01
-1.56689620e+00 -3.58752906e-01 6.74272239e-01 -2.78127380e-02
6.91139758e-01 2.27406457e-01 9.18615460e-01 3.32468659e-01
3.64240631e-02 2.45941862e-01 1.02156842e+00 4.36076596e-02
4.29380268e-01 1.99484870e-01 -5.48268743e-02 5.55684805e-01
7.62592733e-01 3.54955718e-02 -4.80352193e-01 8.81377310e-02
7.40860641e-01 3.52828801e-01 7.87982047e-02 -5.70632577e-01
-1.00359964e+00 9.89748597e-01 5.91733754e-01 6.45331815e-02
-3.70006382e-01 2.27190003e-01 3.11028063e-01 -1.31562829e-01
3.76454651e-01 -2.32578218e-01 -1.79591119e-01 -1.41350463e-01
-1.12058461e+00 2.06049517e-01 5.20787001e-01 1.00975120e+00
1.14981496e+00 -1.89920634e-01 6.22053802e-01 5.91638744e-01
3.43927026e-01 9.91287053e-01 6.32204190e-02 -1.03699946e+00
6.49283230e-01 8.79180670e-01 -9.69350263e-02 -1.13278949e+00
-7.12218225e-01 1.76378060e-02 -1.02223611e+00 7.33864903e-01
2.10771605e-01 1.71671212e-01 -7.23994493e-01 1.01075959e+00
7.74511814e-01 1.47490606e-01 1.25104725e-01 7.46647775e-01
7.94830203e-01 8.16400111e-01 -1.33546785e-01 -9.25740972e-02
1.43948817e+00 -9.81727004e-01 -6.42731071e-01 -2.73913354e-01
5.36786854e-01 -7.76548326e-01 9.46358800e-01 2.83436239e-01
-9.62318540e-01 -5.73683739e-01 -1.25130177e+00 -1.82115119e-02
-5.46708941e-01 3.03894952e-02 5.89546561e-01 6.14787698e-01
-1.00390601e+00 5.25698364e-01 -1.11130798e+00 -3.01579922e-01
3.82407546e-01 4.97492492e-01 -7.49568120e-02 -6.32305443e-02
-8.13335836e-01 1.00639069e+00 3.53394210e-01 2.84270138e-01
-1.06678434e-01 -6.35920525e-01 -6.59795225e-01 -8.61208048e-03
4.49557096e-01 -3.56533796e-01 7.95420706e-01 -2.61638999e-01
-1.61111283e+00 3.56090188e-01 -3.91446322e-01 -6.24048650e-01
6.56803906e-01 -5.97491860e-02 -2.07715601e-01 3.75945628e-01
-2.95391446e-03 9.21393335e-01 5.98579168e-01 -1.04670835e+00
-9.09889579e-01 -4.93207335e-01 -1.08379893e-01 1.40236571e-01
-3.28400880e-01 -2.66683370e-01 -5.60719073e-01 2.55972773e-01
5.49212396e-01 -1.16613579e+00 -5.24701893e-01 5.95291972e-01
1.14484407e-01 -3.25950623e-01 1.65626073e+00 -1.10899717e-01
7.74228096e-01 -2.25165820e+00 -4.31162894e-01 3.72180879e-01
2.29406998e-01 6.30328655e-01 2.04029441e-01 5.70398793e-02
6.15119517e-01 -6.33702427e-02 -1.64859921e-01 -2.64993310e-01
-1.77294120e-01 5.69132268e-01 4.41467464e-02 4.59178686e-01
-1.99972034e-01 5.19241691e-01 -4.53747660e-01 -9.09951270e-01
7.27342904e-01 7.84881294e-01 -3.81062597e-01 -2.53274977e-01
-9.85838324e-02 2.00208247e-01 -5.01710773e-01 3.56767565e-01
1.30436742e+00 1.21824659e-01 -3.07917967e-02 -2.63755858e-01
-6.75396442e-01 1.19621925e-01 -1.33832979e+00 1.24618554e+00
-6.79112434e-01 6.78944767e-01 3.60652089e-01 -8.26108992e-01
1.18310511e+00 -3.39151956e-02 5.90477467e-01 -6.92135811e-01
-9.61297099e-03 7.50239313e-01 -2.17229232e-01 -7.60598481e-02
6.39488220e-01 1.03184931e-01 1.72967881e-01 1.01847768e-01
-5.52639306e-01 -6.00798726e-01 3.75441104e-01 -5.72939441e-02
8.40890646e-01 -1.46618560e-01 -3.08097899e-02 -1.49153024e-01
6.84309840e-01 1.00277461e-01 8.00879955e-01 2.47991249e-01
1.54327322e-02 1.66019857e-01 1.95932686e-01 -5.68869114e-01
-1.08376288e+00 -8.15200150e-01 -4.69196767e-01 2.86671013e-01
8.18528175e-01 -3.69507819e-01 -6.99479640e-01 -2.50671089e-01
1.22006871e-01 4.47258204e-01 2.54929006e-01 3.85824651e-01
-8.16601455e-01 -3.85264248e-01 2.35832766e-01 5.95733762e-01
1.06755626e+00 -5.78980982e-01 -1.23327875e+00 3.17902803e-01
-8.96433592e-02 -1.65356147e+00 9.98324454e-02 -2.07243755e-01
-1.34905219e+00 -9.39632595e-01 -2.54751176e-01 -3.57910365e-01
6.62524402e-01 7.19165146e-01 7.23824203e-01 2.40672827e-01
2.86935624e-02 -1.10821798e-01 -8.40935633e-02 -7.75098383e-01
2.98653040e-02 1.44258440e-01 1.62841216e-01 -6.67038187e-02
4.91540641e-01 -8.17999959e-01 -6.68516874e-01 7.47718811e-01
-7.43658423e-01 2.20132276e-01 7.18608618e-01 3.98442477e-01
1.10413587e+00 4.86101300e-01 1.99047942e-03 -4.68265712e-01
-5.79431988e-02 -4.07388471e-02 -1.42244065e+00 -2.56723464e-01
-7.17445552e-01 -1.79860443e-01 5.78781486e-01 -2.59340554e-01
-7.50257432e-01 4.49915409e-01 -1.70516014e-01 -5.15271127e-01
3.36550339e-03 6.34123236e-02 -1.66724309e-01 -5.72095394e-01
2.56450474e-01 -3.15989479e-02 1.95333034e-01 -3.28119755e-01
1.77736431e-01 5.94839036e-01 2.58146614e-01 -2.02724114e-01
9.91325259e-01 9.90023971e-01 7.40017354e-01 -1.23565960e+00
-2.08066359e-01 -5.49372494e-01 -7.89204180e-01 -4.33759898e-01
8.33266377e-01 -7.62481928e-01 -8.35366011e-01 3.84785205e-01
-1.41305852e+00 -1.24335185e-01 -2.81816632e-01 7.48497009e-01
-5.09583354e-01 2.57309616e-01 -1.51035339e-01 -7.73905575e-01
-2.31473088e-01 -1.40200222e+00 9.05370295e-01 3.78822088e-01
2.18683913e-01 -5.02521157e-01 -4.52080637e-01 2.64951676e-01
5.50012410e-01 1.80259705e-01 5.24171114e-01 -9.96839162e-03
-1.09650099e+00 -1.57447308e-01 -6.79624140e-01 4.36748797e-03
-3.55388582e-01 2.47357726e-01 -7.50091136e-01 -1.20158950e-02
1.23684138e-01 2.25581378e-01 5.67441523e-01 2.94711411e-01
1.17083406e+00 3.37728076e-02 -6.80767179e-01 6.46543622e-01
1.65704513e+00 1.52305216e-01 6.61579311e-01 4.29115057e-01
5.58088005e-01 7.49532402e-01 1.00976074e+00 8.37069601e-02
5.65569043e-01 9.78210807e-01 7.89388835e-01 -2.25839823e-01
1.24900110e-01 6.40073195e-02 3.13312300e-02 1.06708550e+00
-2.87645131e-01 1.11659281e-01 -9.69582498e-01 4.31649357e-01
-2.01814079e+00 -7.32198417e-01 -8.09886932e-01 2.26658559e+00
2.89163202e-01 4.45420325e-01 5.24386242e-02 3.48672539e-01
6.38558447e-01 3.68361697e-02 -5.49996495e-01 -5.24130702e-01
1.37766153e-01 2.46886224e-01 1.09718990e+00 3.16594511e-01
-7.77029991e-01 9.00700629e-01 5.00222826e+00 8.96436155e-01
-1.11292768e+00 4.58053738e-01 1.04594521e-01 -2.13854045e-01
-4.61461581e-02 5.95381446e-02 -1.21739793e+00 5.87455392e-01
1.22383177e+00 2.57954560e-02 9.18981284e-02 1.21396840e+00
4.46088761e-01 -5.39434671e-01 -7.25840271e-01 1.16319847e+00
-2.40614012e-01 -1.27379572e+00 -2.06267208e-01 4.63213414e-01
3.20667684e-01 4.87862110e-01 -2.55163103e-01 -1.82085007e-01
-1.32433817e-01 -5.78079939e-01 7.17026234e-01 1.95685163e-01
6.89818382e-01 -1.01772642e+00 8.73491883e-01 6.22465789e-01
-1.64137614e+00 1.64773911e-01 -9.03735697e-01 7.86715075e-02
5.20859718e-01 1.09420240e+00 -2.82181770e-01 3.54666978e-01
1.03497481e+00 2.48466760e-01 -3.27502906e-01 1.05503654e+00
-1.14902936e-01 1.40875652e-01 -1.07063460e+00 -2.83064812e-01
2.45968416e-01 -6.21930897e-01 4.71267641e-01 8.58457625e-01
7.53562212e-01 8.45165253e-02 2.75543809e-01 6.75828218e-01
2.63221502e-01 1.04058258e-01 -7.42573619e-01 4.47567046e-01
6.91983521e-01 1.43366098e+00 -1.09222949e+00 -4.75453317e-01
-6.44144118e-01 3.55493605e-01 -1.10201351e-01 -3.42668742e-02
-1.01452971e+00 -5.75642407e-01 8.32801521e-01 6.05620265e-01
3.86605799e-01 -8.01708937e-01 -5.37748158e-01 -6.91515267e-01
3.90682936e-01 -2.99107105e-01 -2.93523163e-01 -4.56191897e-01
-6.16543591e-01 7.01421559e-01 1.07663959e-01 -1.53578722e+00
1.56683967e-01 -6.30712628e-01 -3.88739765e-01 5.36297679e-01
-1.91245377e+00 -8.43390584e-01 -7.31927752e-01 7.15600610e-01
5.92875421e-01 2.62255013e-01 4.46243167e-01 6.06130302e-01
-4.00048345e-01 1.17871158e-01 6.57647178e-02 -4.01765853e-01
1.52617499e-01 -6.99876428e-01 2.97376901e-01 8.56545806e-01
-2.00699523e-01 1.62693769e-01 4.35712934e-01 -5.58663368e-01
-1.47559559e+00 -1.17358756e+00 8.20302486e-01 1.22374050e-01
3.92343074e-01 -3.19572985e-01 -8.94224405e-01 6.82443753e-02
-1.59758702e-01 2.95597374e-01 2.59608388e-01 -2.97180235e-01
9.15329605e-02 -7.12154806e-01 -1.18119216e+00 4.35231537e-01
1.03602302e+00 -1.14376865e-01 -2.19745352e-03 4.22249734e-01
6.90060794e-01 -5.28567195e-01 -6.93130195e-01 5.64598083e-01
3.98078322e-01 -1.11517417e+00 9.79413211e-01 5.84662855e-01
-1.92118138e-01 -6.61176920e-01 -8.15017521e-02 -7.61155128e-01
9.25090685e-02 -2.54340827e-01 1.32763296e-01 9.35526550e-01
3.26709270e-01 -1.32798719e+00 1.05002403e+00 4.72886950e-01
-1.17294192e-01 -8.15703869e-01 -1.35661900e+00 -1.12922740e+00
-4.11798835e-01 -6.99050248e-01 9.49997902e-01 2.64798969e-01
-5.02023876e-01 1.70092896e-01 2.85122663e-01 3.32787752e-01
9.22036409e-01 2.46734202e-01 1.10308135e+00 -1.77429175e+00
3.56250525e-01 -3.95616829e-01 -7.87470281e-01 -1.01716685e+00
-1.92359053e-02 -3.92152518e-01 6.08878061e-02 -1.58125937e+00
-3.52005243e-01 -1.15858650e+00 3.95794243e-01 2.16554761e-01
3.91635746e-01 4.20757055e-01 2.53307492e-01 5.43016195e-01
-2.69613743e-01 4.43922669e-01 9.83839869e-01 1.42894546e-02
-2.89893508e-01 2.80135989e-01 1.48225456e-01 1.15891969e+00
9.60854769e-01 -5.64285934e-01 -4.86114830e-01 -5.31553805e-01
2.23967478e-01 -1.27721265e-01 3.16389889e-01 -1.46566558e+00
6.43987477e-01 -2.51822203e-01 1.97674017e-02 -1.34804535e+00
9.97540176e-01 -1.36636531e+00 3.53478640e-01 6.94551766e-01
6.20227039e-01 3.28372300e-01 1.44247547e-01 4.68238056e-01
-3.31142277e-01 -2.45633215e-01 1.00319076e+00 -4.91407365e-02
-4.99439001e-01 5.48550785e-01 -2.19721034e-01 -6.06015921e-01
1.42776322e+00 -6.98040903e-01 -1.22495078e-01 -1.27256304e-01
-2.45763227e-01 2.54150450e-01 5.87576747e-01 7.59479105e-02
8.00628066e-01 -1.31488287e+00 -4.21533853e-01 2.00415626e-01
-1.95665449e-01 7.13780463e-01 1.78203687e-01 9.99184906e-01
-1.11973381e+00 6.43267572e-01 -1.82703331e-01 -1.18530619e+00
-1.45072842e+00 3.65529150e-01 -5.11575975e-02 -5.29849902e-03
-9.09978688e-01 3.75075072e-01 -3.78546305e-02 -4.62761223e-01
-8.85471329e-02 -3.82965326e-01 -1.10494174e-01 1.04834341e-01
3.76950592e-01 8.61039281e-01 3.28740448e-01 -8.86714637e-01
-4.75386292e-01 1.19339848e+00 4.02712405e-01 -1.54339850e-01
1.14790416e+00 -1.77343488e-01 -3.20519000e-01 2.02215895e-01
9.81045008e-01 -2.94743627e-01 -1.06675708e+00 -1.11655749e-01
-2.08364159e-01 -6.41050756e-01 3.32418054e-01 3.27111185e-01
-1.40503645e+00 1.18366635e+00 6.46599770e-01 1.50526866e-01
1.03545570e+00 -2.87172586e-01 1.26073086e+00 4.80002820e-01
9.01723802e-01 -1.41281617e+00 -3.44599038e-01 3.50949615e-01
4.31621909e-01 -1.01191461e+00 3.71823251e-01 -1.14363921e+00
-2.50411537e-02 1.21523464e+00 7.88561761e-01 -3.73403984e-03
8.12938750e-01 3.72380465e-01 -1.26206756e-01 -1.43389970e-01
-3.31069022e-01 -2.39643514e-01 -2.23819405e-01 5.65386117e-01
-4.12965834e-01 1.73109174e-01 -6.38878822e-01 -1.56509042e-01
-3.76138240e-01 -5.22931144e-02 3.51355344e-01 8.36696088e-01
-7.46431887e-01 -1.10650945e+00 -6.41499937e-01 4.24275070e-01
1.20647676e-01 2.88939834e-01 2.58810073e-01 8.20870757e-01
3.05728883e-01 1.02667868e+00 3.84980381e-01 -2.64164239e-01
5.24592936e-01 -2.40027115e-01 2.46519178e-01 -2.33500972e-01
5.38959578e-02 -1.95059106e-01 -1.26711547e-01 -1.04697144e+00
-5.05263150e-01 -5.11035800e-01 -1.52747917e+00 -7.82311797e-01
-5.30688465e-01 1.22991137e-01 1.56902647e+00 5.69610596e-01
5.35438538e-01 1.53181031e-01 6.10180020e-01 -1.22528851e+00
-1.61404297e-01 -5.36357343e-01 -3.33712488e-01 -4.23786730e-01
-2.63408214e-01 -8.58217537e-01 -2.58944541e-01 -4.66791421e-01] | [8.097898483276367, -2.642282485961914] |
b7e1ac4f-5066-475c-9d07-33fdbcb89764 | automated-audio-captioning-an-overview-of | 2205.05949 | null | https://arxiv.org/abs/2205.05949v2 | https://arxiv.org/pdf/2205.05949v2.pdf | Automated Audio Captioning: An Overview of Recent Progress and New Challenges | Automated audio captioning is a cross-modal translation task that aims to generate natural language descriptions for given audio clips. This task has received increasing attention with the release of freely available datasets in recent years. The problem has been addressed predominantly with deep learning techniques. Numerous approaches have been proposed, such as investigating different neural network architectures, exploiting auxiliary information such as keywords or sentence information to guide caption generation, and employing different training strategies, which have greatly facilitated the development of this field. In this paper, we present a comprehensive review of the published contributions in automated audio captioning, from a variety of existing approaches to evaluation metrics and datasets. We also discuss open challenges and envisage possible future research directions. | ['Wenwu Wang', 'Mark D. Plumbley', 'Xubo Liu', 'Xinhao Mei'] | 2022-05-12 | null | null | null | null | ['audio-captioning'] | ['audio'] | [ 6.48716748e-01 3.12899381e-01 -1.06354743e-01 -2.85069168e-01
-1.54332185e+00 -5.05224109e-01 7.49755561e-01 -1.26353034e-03
-2.18373790e-01 1.10627043e+00 7.84617126e-01 1.69854328e-01
1.55149683e-01 -3.91515672e-01 -7.98825026e-01 -3.31918329e-01
-1.02514185e-01 5.30904233e-01 -1.68436840e-01 -1.55626789e-01
1.30230278e-01 3.01831998e-02 -1.58486378e+00 8.38462710e-01
4.49641824e-01 1.03872919e+00 7.58202896e-02 7.66653776e-01
-1.29664734e-01 7.68532515e-01 -8.93779635e-01 -7.09354579e-01
-6.24340102e-02 -8.69751215e-01 -1.05625343e+00 -1.63756281e-01
2.87975818e-01 -3.08148074e-03 -2.33082503e-01 5.59886396e-01
1.02420664e+00 1.10555805e-01 4.35065150e-01 -1.26853609e+00
-6.63308442e-01 1.10830176e+00 1.69029199e-02 7.48379529e-02
7.72589982e-01 -8.74197297e-03 1.19497335e+00 -8.18933666e-01
5.98203421e-01 9.79743540e-01 5.18611610e-01 8.75840306e-01
-1.16668785e+00 -6.67137980e-01 -1.96833774e-01 5.78300774e-01
-1.48338616e+00 -6.38203144e-01 1.10078347e+00 -3.11320156e-01
8.42326283e-01 3.86660457e-01 3.17441612e-01 1.77928782e+00
-3.07307601e-01 1.01816010e+00 9.30649996e-01 -6.54864848e-01
2.09655259e-02 2.83032417e-01 -3.78520221e-01 7.56705552e-03
-2.06178799e-01 -1.88100472e-01 -8.14837456e-01 -2.41272554e-01
4.23563808e-01 -9.35036957e-01 -4.07441258e-01 3.12708169e-02
-1.57298064e+00 7.83221126e-01 6.68870062e-02 4.99270439e-01
-4.82881546e-01 2.05873158e-02 8.11020792e-01 3.24776530e-01
4.60810244e-01 8.60299826e-01 -2.22833216e-01 -4.69601035e-01
-1.02322912e+00 5.14879048e-01 7.41156220e-01 1.00732124e+00
3.06226969e-01 1.64907277e-01 -3.88775736e-01 1.03824484e+00
-9.91692021e-03 2.10157156e-01 5.20233810e-01 -9.51069534e-01
9.57685888e-01 2.14947574e-02 1.70111269e-01 -8.24574888e-01
-2.41003886e-01 -2.74172187e-01 -9.15483057e-01 -4.28879052e-01
1.23653956e-01 -5.34938097e-01 -5.10751069e-01 1.70485353e+00
-2.39502743e-01 1.84813440e-01 3.09054166e-01 1.04988623e+00
1.17233229e+00 9.23555911e-01 4.49505299e-02 -1.94231808e-01
1.16290045e+00 -9.66274381e-01 -8.66292357e-01 -1.26632750e-01
2.21713722e-01 -1.03267503e+00 1.00930917e+00 4.42821205e-01
-1.27723420e+00 -6.07014954e-01 -9.33933914e-01 -8.12644437e-02
-2.51707077e-01 1.97097555e-01 3.50710750e-01 4.01108652e-01
-1.19777441e+00 2.89600253e-01 -5.38054049e-01 -3.30240399e-01
1.72163308e-01 2.12431058e-01 -2.30174407e-01 1.71153739e-01
-1.78480017e+00 7.13662982e-01 5.70090055e-01 6.26055673e-02
-9.34001505e-01 -5.59908032e-01 -7.10973859e-01 4.53630500e-02
1.23532690e-01 -8.12784433e-01 1.77575767e+00 -1.08922005e+00
-1.70911717e+00 6.48513496e-01 3.37950028e-02 -9.23139453e-01
4.59927499e-01 -1.81955636e-01 -5.86774945e-01 4.20536458e-01
-5.69039434e-02 1.25390172e+00 6.73121989e-01 -1.08694458e+00
-4.45686549e-01 1.10675618e-01 2.95176864e-01 4.51478332e-01
-5.37650406e-01 4.55934644e-01 -2.60071486e-01 -9.41419184e-01
-6.08092070e-01 -9.86205816e-01 -5.00734448e-02 -4.05701816e-01
-3.87171596e-01 -3.05515885e-01 4.04001594e-01 -6.92031562e-01
1.45338297e+00 -1.94093108e+00 3.54596704e-01 -3.21196407e-01
-2.98174053e-01 2.85327524e-01 -2.86412954e-01 8.63207161e-01
-4.17760238e-02 1.60170197e-01 -3.72637898e-01 -4.83853757e-01
1.06791705e-01 -1.70452312e-01 -7.21120119e-01 -7.96475485e-02
4.85279918e-01 7.35906184e-01 -8.20969582e-01 -4.76331383e-01
-5.53587228e-02 8.22024107e-01 -2.24694416e-01 4.60768908e-01
-4.35684949e-01 4.88045543e-01 -2.44901896e-01 4.08980846e-01
1.74819127e-01 -1.20881079e-02 -1.96324632e-01 6.78289274e-04
-8.74792114e-02 8.93691063e-01 -7.90053606e-01 1.84363937e+00
-6.83857143e-01 1.06709480e+00 4.29431088e-02 -1.05255044e+00
1.04146242e+00 1.10741341e+00 5.13559878e-01 -6.04733169e-01
1.64140269e-01 3.35619867e-01 -1.72542140e-01 -6.02482080e-01
6.89961672e-01 -2.39352539e-01 -2.21178278e-01 5.11377335e-01
1.49470389e-01 -1.43393785e-01 3.58201653e-01 -1.13405742e-01
7.70869792e-01 2.06595361e-01 1.82480916e-01 3.00539076e-01
7.55512774e-01 1.79378927e-01 -5.28743044e-02 5.26043415e-01
-1.62254676e-01 1.18554640e+00 2.67428666e-01 -2.94505268e-01
-1.11238182e+00 -7.27876306e-01 -1.73885301e-02 9.72437441e-01
-3.42811704e-01 -4.72048491e-01 -8.89705002e-01 -4.19231534e-01
-6.67832613e-01 6.09492064e-01 -5.83010137e-01 -3.35830189e-02
-7.12930977e-01 -4.17639166e-01 9.29366827e-01 4.85087723e-01
4.75254804e-01 -1.61319673e+00 -3.80822241e-01 4.16154563e-01
-9.20597315e-01 -1.32919288e+00 -2.27939770e-01 -3.05759981e-02
-6.97561324e-01 -5.28427958e-01 -1.18690252e+00 -1.01968861e+00
6.02420308e-02 3.02819535e-02 1.24240816e+00 -4.79160637e-01
4.05921079e-02 4.35578167e-01 -7.40438223e-01 -6.39825702e-01
-6.65674925e-01 7.43365526e-01 -6.92265034e-02 1.47648200e-01
4.29504842e-01 -6.61720276e-01 -2.40289629e-01 1.02253959e-01
-9.44013894e-01 2.99333721e-01 7.49167919e-01 5.76222122e-01
5.49384654e-01 -5.58722258e-01 1.04533529e+00 -4.24637467e-01
1.16513443e+00 -6.16147339e-01 -8.50881189e-02 4.84046862e-02
-1.20415241e-01 -4.60386686e-02 7.77658939e-01 -3.77850413e-01
-9.42265809e-01 -1.35118514e-02 -4.90637243e-01 -4.47976470e-01
-6.49145961e-01 8.48424256e-01 -7.60664046e-02 3.24018955e-01
7.07933187e-01 3.71708870e-01 -7.73304179e-02 -4.88374710e-01
2.82803953e-01 1.21327126e+00 7.05497146e-01 -6.00452363e-01
5.92987597e-01 7.75830373e-02 -4.09154624e-01 -9.17499661e-01
-9.05935526e-01 -3.26159298e-01 -3.78236890e-01 -3.63775134e-01
8.27135444e-01 -9.85518336e-01 -1.00572981e-01 1.13612145e-01
-1.56222892e+00 -1.00824267e-01 -1.31059721e-01 6.48907006e-01
-9.73772585e-01 9.30625200e-02 -5.19500494e-01 -7.18674421e-01
-6.75960064e-01 -1.13995433e+00 1.20521140e+00 1.07741527e-01
-7.39556789e-01 -7.95551896e-01 4.10636067e-01 5.83739161e-01
5.41824341e-01 2.99459606e-01 5.00780046e-01 -8.45482111e-01
-4.34762597e-01 -2.20437765e-01 9.56572816e-02 2.93480754e-01
-7.27366358e-02 -2.62562841e-01 -1.20429850e+00 -5.52281290e-02
-2.00375900e-01 -7.04860449e-01 5.10739982e-01 1.97311103e-01
1.02566957e+00 -4.29411530e-01 -4.93710786e-02 1.61191359e-01
8.55189741e-01 2.09773362e-01 6.40943229e-01 5.69475889e-01
2.93447435e-01 9.19256508e-01 6.90209150e-01 3.96735549e-01
3.02155524e-01 9.22341228e-01 3.38965148e-01 8.59770775e-02
-2.70595551e-01 -3.29107881e-01 5.15495479e-01 1.04242706e+00
-1.10165700e-01 -5.24250686e-01 -8.34666669e-01 7.95897841e-01
-1.67939520e+00 -1.06954432e+00 6.16510920e-02 1.84481430e+00
9.67903972e-01 -5.24165221e-02 3.80115420e-01 3.70433539e-01
9.47250128e-01 2.58292258e-01 -2.50453055e-01 -5.32635450e-01
-1.32898226e-01 1.94800720e-01 -3.09110492e-01 2.44529650e-01
-1.11134577e+00 5.85631311e-01 6.62323570e+00 5.23203552e-01
-1.20676720e+00 5.07136844e-02 4.86852318e-01 -1.24941863e-01
-2.79604256e-01 -3.69577318e-01 -4.98180270e-01 4.43352342e-01
1.41107714e+00 -5.03754497e-01 3.56439382e-01 6.72491014e-01
3.71618599e-01 4.28800523e-01 -1.21965814e+00 1.09387970e+00
4.76340294e-01 -1.23671174e+00 2.53132284e-01 -3.02140176e-01
6.61843121e-01 1.61103368e-01 1.72937632e-01 4.06711251e-01
-2.06563175e-01 -8.87116671e-01 7.35750020e-01 3.23408574e-01
8.41254115e-01 -6.66918516e-01 9.32050049e-01 -3.11481450e-02
-1.04513109e+00 1.16736613e-01 7.93809369e-02 -3.41226220e-01
4.71507549e-01 2.01410756e-01 -1.06516027e+00 7.90146947e-01
6.20872080e-01 5.24201870e-01 -3.62243056e-01 1.30648673e+00
-3.08540881e-01 9.07153249e-01 -1.20768815e-01 -2.04122767e-01
3.42078835e-01 1.53513268e-01 6.92000628e-01 1.41199684e+00
4.96730387e-01 -3.05714667e-01 -1.20214395e-01 7.62210011e-01
-2.73741245e-01 4.51339394e-01 -5.14160097e-01 -3.35249692e-01
3.55075002e-01 1.03280318e+00 -3.60655457e-01 -2.88735777e-01
-4.96179372e-01 7.22297966e-01 2.74722371e-02 2.75777370e-01
-9.59770858e-01 -6.33889854e-01 5.87943792e-01 1.34432346e-01
-2.96006590e-04 -9.83053073e-02 2.66866963e-02 -9.91505742e-01
2.74229288e-01 -9.45784867e-01 4.68873143e-01 -1.05789983e+00
-1.09066129e+00 1.16867876e+00 3.88494134e-01 -1.60257709e+00
-8.02105069e-01 -1.93905160e-01 -4.77334380e-01 6.37162924e-01
-1.47374392e+00 -9.23448682e-01 -1.19906634e-01 3.17445129e-01
9.44905579e-01 -2.80659109e-01 1.05533123e+00 5.35698891e-01
-2.91846156e-01 6.49574578e-01 -1.08061299e-01 7.14316890e-02
1.03043616e+00 -8.46387029e-01 3.96050811e-01 6.08882487e-01
4.38724965e-01 2.88203657e-01 1.03020966e+00 -6.37751669e-02
-9.06653821e-01 -1.11930013e+00 1.41597927e+00 -2.58371770e-01
7.40592659e-01 -4.39080358e-01 -9.01995122e-01 4.40269113e-01
9.30497050e-01 -4.48880047e-01 9.68543172e-01 -2.69021362e-01
-6.34765327e-02 -9.76093337e-02 -7.02707052e-01 6.32475019e-01
8.08765292e-01 -6.05086565e-01 -7.03750908e-01 3.67969990e-01
9.19867218e-01 -4.39037949e-01 -7.76152611e-01 3.34657490e-01
3.84361744e-01 -6.67097211e-01 6.57748938e-01 -5.75565577e-01
7.23003805e-01 -1.49161398e-01 8.32674131e-02 -1.30658424e+00
-6.72487020e-02 -1.03339016e+00 -1.29701927e-01 1.55040109e+00
6.40082896e-01 -1.26157522e-01 6.86020553e-01 2.80647278e-01
-5.53526103e-01 -6.49088860e-01 -8.89021575e-01 -5.33048511e-01
-4.63380106e-03 -5.18358409e-01 5.99385500e-01 7.02973485e-01
2.47803271e-01 7.02981651e-01 -7.56660879e-01 -1.54951155e-01
2.64225677e-02 8.97257403e-02 6.66024923e-01 -1.09104288e+00
-6.38605133e-02 -5.06794631e-01 -3.06209683e-01 -8.79197419e-01
3.98374766e-01 -9.21654999e-01 1.31801948e-01 -1.70183194e+00
-7.51519296e-03 1.69854596e-01 -1.70593143e-01 4.87623632e-01
9.72410589e-02 7.75926709e-01 1.78998545e-01 7.22650886e-02
-6.62185013e-01 6.40887380e-01 9.84813809e-01 -4.03115213e-01
-1.97961926e-01 1.39625475e-01 -7.96220362e-01 2.84574181e-01
1.18841183e+00 -3.82084846e-01 -4.40450191e-01 -6.09547853e-01
3.30227256e-01 1.38832152e-01 1.57289907e-01 -1.27020121e+00
2.60551833e-02 1.50551632e-01 -6.81300014e-02 -4.22296196e-01
5.77671170e-01 -6.55606329e-01 1.62513182e-01 -6.21921243e-03
-9.91210639e-01 3.47836494e-01 2.71502137e-01 3.26515675e-01
-8.98522258e-01 -3.06231558e-01 3.93787742e-01 -8.21819529e-02
-3.00493121e-01 7.00378492e-02 -6.27912641e-01 1.05656162e-01
8.28555346e-01 7.72951096e-02 1.13764390e-01 -8.62602890e-01
-8.39231431e-01 7.63233081e-02 -1.07657276e-01 9.57604945e-01
4.19754148e-01 -1.55747962e+00 -1.20527303e+00 -2.76616096e-01
3.26795220e-01 -2.79809922e-01 6.40472844e-02 6.51992261e-01
-1.67996272e-01 9.91028249e-01 -3.24302703e-01 -4.42701310e-01
-1.20608854e+00 4.78242457e-01 5.87705430e-03 -2.06244528e-01
-2.24580318e-01 5.80805957e-01 -2.08694771e-01 -6.06456473e-02
4.71962601e-01 -1.67819243e-02 -6.49183512e-01 2.74282634e-01
6.42004251e-01 2.10837752e-01 2.70816892e-01 -9.02827382e-01
-7.86588192e-02 2.08580971e-01 -7.04968423e-02 -4.50572968e-01
1.28832448e+00 -2.26212814e-01 1.74882382e-01 5.26659548e-01
1.20807946e+00 -3.41682702e-01 -6.97629333e-01 -5.40348738e-02
1.36565147e-02 3.12119052e-02 -1.93755999e-01 -8.70129466e-01
-6.98524177e-01 1.16893220e+00 3.11439991e-01 4.66438770e-01
1.25979745e+00 1.16480432e-01 8.74036551e-01 2.82543302e-01
3.54167789e-01 -9.87599909e-01 1.39939144e-01 5.62827110e-01
1.32116520e+00 -1.08531010e+00 -5.65033138e-01 -1.28602058e-01
-7.22748041e-01 1.13131094e+00 6.23730123e-01 1.93436444e-01
-5.73521517e-02 9.00951624e-02 2.41818398e-01 3.03530663e-01
-8.49160731e-01 -1.31376699e-01 3.25633883e-01 7.52982140e-01
8.79764140e-01 -1.63953125e-01 -4.90437448e-01 8.56015086e-01
-6.25243068e-01 1.18214048e-01 5.82838237e-01 5.64523160e-01
-1.07080974e-01 -1.45681608e+00 -4.77721542e-01 2.29073524e-01
-7.37321079e-01 -3.88499387e-02 -9.20924425e-01 5.58926463e-01
-7.74113312e-02 9.52809632e-01 -6.27302825e-02 -3.67011219e-01
3.18551898e-01 3.44209343e-01 3.13702315e-01 -7.00710297e-01
-6.67598009e-01 2.04999298e-02 4.00404662e-01 -1.61740080e-01
-8.04890931e-01 -7.76637316e-01 -7.40384758e-01 2.83155113e-01
-1.14757381e-01 7.38988876e-01 6.39926434e-01 6.97502553e-01
5.87352753e-01 4.81993884e-01 4.67871308e-01 -1.18219972e+00
-2.74375767e-01 -1.30374765e+00 5.31274527e-02 9.04866159e-02
2.59387314e-01 -3.60045075e-01 -2.83962607e-01 3.92988801e-01] | [15.274275779724121, 4.844076156616211] |
6e95c9b4-b127-49c8-a3bf-3ee251265231 | tractable-epistemic-reasoning-with-functional | 1403.0034 | null | https://arxiv.org/abs/1403.0034v4 | https://arxiv.org/pdf/1403.0034v4.pdf | Tractable Epistemic Reasoning with Functional Fluents, Static Causal Laws and Postdiction | We present an epistemic action theory for tractable epistemic reasoning as an extension to the h-approximation (HPX) theory. In contrast to existing tractable approaches, the theory supports functional fluents and postdictive reasoning with static causal laws. We argue that this combination is particularly synergistic because it allows one not only to perform direct postdiction about the conditions of actions, but also indirect postdiction about the conditions of static causal laws. We show that despite the richer expressiveness, the temporal projection problem remains tractable (polynomial), and therefore the planning problem remains in NP. We present the operational semantics of our theory as well as its formulation as Answer Set Programming. | ['Manfred Eppe'] | 2014-03-01 | null | null | null | null | ['epistemic-reasoning'] | ['miscellaneous'] | [ 9.81011391e-02 1.21327817e+00 -3.18154126e-01 7.36028794e-03
-5.44692218e-01 -8.41858685e-01 9.69889045e-01 5.02138697e-02
-7.87955523e-02 9.08346713e-01 7.88500249e-01 -8.49036098e-01
-7.32013583e-01 -1.14292824e+00 -7.05500424e-01 -4.06172216e-01
-4.38563615e-01 5.07476211e-01 7.04782128e-01 -3.56201679e-01
-1.21156708e-03 3.62204373e-01 -1.16253841e+00 1.74016699e-01
5.24996281e-01 6.36592507e-01 -3.27195138e-01 4.36014205e-01
1.19679317e-01 1.92044806e+00 -2.64296401e-02 -3.77638102e-01
7.86135644e-02 -4.34293270e-01 -1.40303504e+00 -3.85722250e-01
-2.20659569e-01 -4.74058241e-01 -5.66060066e-01 1.15975749e+00
-1.21753365e-01 1.85619965e-01 2.12719560e-01 -1.64989936e+00
-6.80922091e-01 1.40980446e+00 2.87827873e-03 -1.28770322e-01
9.81375813e-01 3.63452047e-01 1.39058948e+00 -9.40148458e-02
6.66879177e-01 1.52677178e+00 5.51260054e-01 5.75887203e-01
-1.22859621e+00 2.82925546e-01 3.36909592e-01 2.91200340e-01
-1.09356391e+00 -5.52274406e-01 1.76885739e-01 -3.51119190e-01
1.11719811e+00 5.68314731e-01 8.84673059e-01 6.78458452e-01
2.84887969e-01 6.47321880e-01 1.37578821e+00 -7.46513069e-01
4.61951673e-01 -2.61281461e-01 4.69205797e-01 6.16347373e-01
1.97798431e-01 7.52689838e-01 -4.82799798e-01 -6.36845410e-01
5.02473533e-01 -3.18339974e-01 -2.78713495e-01 -6.76388964e-02
-9.83573496e-01 8.44100595e-01 3.94344144e-03 1.16226293e-01
-3.16738486e-01 8.44945431e-01 2.31500626e-01 3.49794596e-01
-2.41161913e-01 5.84678829e-01 -4.02600408e-01 -1.16693757e-01
-1.42602354e-01 7.97344029e-01 1.16694057e+00 7.87890196e-01
3.17742646e-01 -2.08791047e-01 -3.79724830e-01 -3.65750343e-01
7.66465068e-01 5.69987237e-01 -3.47147435e-01 -1.85098445e+00
3.58878337e-02 4.01806891e-01 7.17507958e-01 -6.51505888e-01
-2.68427610e-01 3.39073747e-01 7.83426389e-02 2.44252250e-01
6.03504717e-01 -1.85915321e-01 -4.82604176e-01 2.17219114e+00
3.36417913e-01 8.88668448e-02 2.65622258e-01 4.90306079e-01
1.22983448e-01 8.42593491e-01 3.46885353e-01 -8.42706621e-01
1.06876385e+00 -4.41487491e-01 -1.04936934e+00 -1.45413280e-01
6.29044473e-01 8.55403095e-02 9.04083967e-01 2.79534519e-01
-1.38661337e+00 6.85207725e-01 -8.40464711e-01 -9.54961106e-02
7.98605308e-02 -1.11473310e+00 1.35416126e+00 5.18066287e-01
-1.24277103e+00 3.17905575e-01 -9.59911466e-01 5.52214757e-02
4.18627299e-02 2.21498117e-01 -1.94585174e-01 1.36550078e-02
-1.66767991e+00 1.23109353e+00 5.11779249e-01 1.26258895e-01
-1.24736607e+00 -5.01690507e-01 -8.18017662e-01 2.24560693e-01
1.13266659e+00 -5.08400679e-01 1.81381786e+00 -4.19039220e-01
-1.71124709e+00 4.55348343e-01 1.67753309e-01 -7.68505514e-01
5.60830891e-01 -4.03206535e-02 -4.00182277e-01 2.99379617e-01
-1.03695998e-02 8.24556574e-02 3.02145034e-02 -8.78174067e-01
-5.76012790e-01 -4.49427038e-01 1.25364280e+00 2.02454448e-01
5.01924098e-01 6.35035276e-01 1.25259504e-01 3.86616796e-01
1.60068393e-01 -9.98460352e-01 -5.72638273e-01 -1.40417263e-01
-4.93404180e-01 -2.65822738e-01 4.40140814e-01 -2.38272920e-02
1.40750360e+00 -1.81547999e+00 2.47517124e-01 4.19998795e-01
1.92745849e-01 -4.77139264e-01 2.82743990e-01 7.98087835e-01
-6.48567975e-02 3.39949012e-01 -1.11737490e-01 3.89112353e-01
6.61378086e-01 4.13723379e-01 -8.78098667e-01 7.64272332e-01
-2.55445302e-01 1.14869499e+00 -8.76318336e-01 -6.36386693e-01
2.09130228e-01 -2.73250461e-01 -7.49337018e-01 -8.35440238e-04
-1.05488443e+00 -2.17071678e-02 -6.33786082e-01 4.70530450e-01
1.42967805e-01 -3.59985419e-02 8.50355387e-01 4.68338877e-01
-4.81703460e-01 7.24139571e-01 -1.32302225e+00 1.03746915e+00
-1.20741189e-01 9.37432051e-02 9.52252373e-02 -3.07212383e-01
-1.27689674e-01 8.44218254e-01 2.79784262e-01 -4.56123024e-01
1.33089840e-01 2.28997413e-02 3.18783335e-02 -5.33563852e-01
1.19795442e-01 -8.04638088e-01 -6.04843915e-01 8.11734617e-01
-4.96448785e-01 -3.68137270e-01 -4.93674958e-03 5.44326067e-01
1.18290234e+00 4.64366794e-01 6.50031507e-01 -5.58072269e-01
3.80046248e-01 3.69479984e-01 7.46331394e-01 7.49924421e-01
-2.57857919e-01 -4.42640662e-01 1.29904842e+00 -3.28748703e-01
-5.97092032e-01 -9.86923695e-01 -1.46684989e-01 1.02592313e+00
3.79737467e-01 -7.30403006e-01 -6.61715746e-01 -4.10568058e-01
-1.75431103e-01 1.28579211e+00 -7.86817491e-01 -1.66689456e-01
-4.77932423e-01 -3.03779721e-01 8.80127966e-01 5.93709588e-01
1.91599265e-01 -9.25680101e-01 -1.07129204e+00 2.56621428e-02
-2.14115858e-01 -6.43193543e-01 1.82372272e-01 3.57130647e-01
-5.64505458e-01 -1.18442345e+00 6.35831475e-01 2.35054389e-01
2.10195169e-01 -3.23792934e-01 7.31696844e-01 6.31823614e-02
5.29216766e-01 5.40067375e-01 -1.45656556e-01 -4.21263546e-01
-5.26166201e-01 -8.46493304e-01 -1.56227171e-01 -5.35771668e-01
1.29936621e-01 -5.06994069e-01 -7.50827119e-02 2.85097566e-02
-7.49742329e-01 -7.40262493e-02 -2.73017019e-01 5.47644615e-01
5.46204388e-01 4.13957179e-01 3.27835754e-02 -1.08074796e+00
7.17494190e-01 -4.29035485e-01 -1.03962469e+00 6.36233807e-01
-5.04592180e-01 1.65018842e-01 4.15025145e-01 2.66761463e-02
-1.69229352e+00 -2.14448884e-01 3.20283413e-01 2.26766154e-01
1.79940239e-01 1.03542709e+00 -4.95359480e-01 2.05460489e-01
9.09383237e-01 -2.84873456e-01 -1.42397091e-01 1.13456093e-01
7.15749085e-01 1.22973621e-01 5.40683806e-01 -1.36820829e+00
6.44481003e-01 7.06262052e-01 4.36097234e-01 -1.17269345e-01
-1.00120997e+00 3.26417327e-01 -2.02071384e-01 -3.83013368e-01
8.52210224e-01 -6.29996896e-01 -1.38223934e+00 -1.94712095e-02
-9.87912953e-01 -8.54855180e-01 -7.53155589e-01 3.01164597e-01
-1.22416675e+00 2.82256037e-01 -8.25273871e-01 -1.43609881e+00
3.38366240e-01 -9.92113411e-01 5.48922658e-01 -2.78916150e-01
-5.17281830e-01 -8.73418927e-01 1.59434661e-01 6.46574125e-02
9.76998126e-04 4.96075213e-01 1.02144718e+00 -4.66137290e-01
-7.83191025e-01 -2.09050521e-01 1.40953660e-01 -4.00194049e-01
-4.52608377e-01 3.16584826e-01 -7.70355463e-01 3.91911387e-01
2.09059983e-01 -3.08788896e-01 1.09808855e-01 4.84304160e-01
4.48533684e-01 -7.75106311e-01 -2.58816659e-01 -9.85810682e-02
1.51756847e+00 2.85474062e-01 1.14644349e+00 4.04498875e-01
-7.24737570e-02 6.95822597e-01 8.88546467e-01 4.28986728e-01
4.18772429e-01 5.46708941e-01 5.22781551e-01 7.81483114e-01
1.40258759e-01 -5.04446983e-01 4.02414113e-01 -2.29491919e-01
-5.67204893e-01 -6.83450624e-02 -1.06682873e+00 4.09309149e-01
-2.35970640e+00 -1.61829734e+00 -4.99373645e-01 1.89408088e+00
1.21377957e+00 7.32424632e-02 2.71080673e-01 1.08153351e-01
4.42620426e-01 -1.90263763e-02 -6.83411434e-02 -5.33277273e-01
-2.25853920e-02 -2.54895668e-02 6.47654414e-01 1.17543733e+00
-6.20429039e-01 9.55681384e-01 8.17944050e+00 2.59320259e-01
-3.16138238e-01 5.78651071e-01 -8.39828476e-02 -5.44286482e-02
-9.11057651e-01 9.18278337e-01 -3.14648211e-01 1.96444467e-01
1.16820931e+00 -5.06257355e-01 7.08228528e-01 7.44055927e-01
1.33808568e-01 -5.33610761e-01 -1.29788506e+00 -6.17820546e-02
-6.78229630e-01 -1.43224168e+00 -4.04990047e-01 1.23293541e-01
5.49445927e-01 -4.18239504e-01 -2.34202027e-01 1.23382136e-01
1.24694633e+00 -9.47824717e-01 1.60809052e+00 4.27662492e-01
6.37835205e-01 -8.00911844e-01 4.41310346e-01 3.89628947e-01
-7.34410822e-01 -7.18914270e-01 3.84847634e-02 -8.12990665e-01
6.36952698e-01 3.66116911e-01 -6.94733411e-02 3.36654037e-01
3.29827219e-01 -5.94630055e-02 4.22471702e-01 6.64936721e-01
-1.07529628e+00 4.42056358e-01 -6.39386833e-01 1.46273956e-01
2.73500174e-01 -1.63164556e-01 5.86004496e-01 4.82149571e-01
-1.04391769e-01 8.62490416e-01 2.98869517e-02 1.02339339e+00
5.39110839e-01 -5.69221675e-01 -5.16276896e-01 -7.69126266e-02
6.90820456e-01 3.83781195e-01 -3.81526381e-01 -3.14897090e-01
-2.38264665e-01 2.34543905e-02 1.85146734e-01 2.27606460e-01
-1.05726218e+00 1.72402054e-01 4.16645646e-01 2.60755587e-02
-3.22110951e-01 -1.67327091e-01 -2.89988190e-01 -1.01760411e+00
-3.08530420e-01 -8.09743762e-01 8.37854624e-01 -8.37466002e-01
-7.11412787e-01 -5.26280291e-02 7.70566702e-01 -4.09302622e-01
-3.55914325e-01 -3.86224896e-01 -5.49596429e-01 5.49289346e-01
-9.54234898e-01 -1.28399682e+00 6.31274283e-01 7.67939270e-01
-2.58567244e-01 7.50203133e-01 1.00755310e+00 -3.48812073e-01
-3.79700333e-01 -2.66806692e-01 -6.43663585e-01 -5.70682704e-01
2.70499685e-03 -1.25760126e+00 -2.60281384e-01 1.29308808e+00
-5.90323448e-01 1.07203245e+00 1.20139694e+00 -8.53801727e-01
-1.98340166e+00 -6.92718089e-01 9.68230307e-01 -6.92822039e-01
1.30110061e+00 2.06104472e-01 -4.29786503e-01 1.66662681e+00
3.02133560e-01 -3.11800361e-01 4.66574043e-01 6.09461188e-01
-5.58016777e-01 2.12894440e-01 -1.14129114e+00 1.06006551e+00
1.20140684e+00 -7.54227042e-01 -1.23071051e+00 5.40977776e-01
1.06119728e+00 -4.51356918e-01 -7.93207169e-01 1.93144232e-01
4.80557799e-01 -8.74737561e-01 5.73841631e-01 -7.87206054e-01
3.32942635e-01 -7.77306259e-01 -4.86572206e-01 -5.00401795e-01
-5.02531230e-01 -1.11021936e+00 -4.87992138e-01 7.04317629e-01
5.30968070e-01 -9.09758806e-01 1.72991425e-01 1.68808079e+00
-2.93530911e-01 -3.10226053e-01 -1.08066666e+00 -7.78117478e-01
1.37511477e-01 -9.01386619e-01 7.64198303e-01 8.57279718e-01
1.23022008e+00 1.49765879e-01 -3.26554954e-01 4.11076933e-01
4.82814699e-01 4.25818026e-01 9.08954740e-02 -1.08268476e+00
-7.18347073e-01 -1.41624480e-01 -2.60033980e-02 -3.31838965e-01
3.96758348e-01 -7.04791784e-01 3.98133874e-01 -1.56430948e+00
2.37105891e-01 -4.45934832e-01 1.66507721e-01 1.20497787e+00
3.42496961e-01 -3.86301279e-01 1.11759141e-01 4.48887572e-02
-6.96287811e-01 1.98105887e-01 1.09158599e+00 1.19997762e-01
-1.57969251e-01 -6.40172511e-02 -9.64939296e-01 8.59777927e-01
4.52445000e-01 -3.57234865e-01 -7.70305634e-01 -1.76537663e-01
1.30948400e+00 7.89758205e-01 6.19580746e-01 -4.34403390e-01
4.22727704e-01 -1.24829197e+00 -7.11672187e-01 -2.77534723e-02
2.64950879e-02 -7.18403757e-01 8.17369521e-01 6.97871864e-01
-5.57246983e-01 -4.47780222e-01 8.19848999e-02 5.08258104e-01
5.50823845e-02 -3.90725374e-01 4.44067955e-01 -1.20461009e-01
-8.41512978e-01 -4.34363149e-02 -8.04564655e-01 -1.32082939e-01
1.25456691e+00 3.02546266e-02 -7.97119796e-01 -4.89312202e-01
-1.00315857e+00 2.38581270e-01 5.67324162e-01 -5.73435605e-01
2.83407986e-01 -8.60418439e-01 -9.11587998e-02 -8.09876859e-01
-6.66689500e-02 -1.98124796e-01 3.74863774e-01 1.28253496e+00
-4.32815164e-01 6.00768685e-01 4.02870402e-03 2.63787866e-01
-6.64329767e-01 1.03641200e+00 5.08845866e-01 -4.87855971e-02
-6.59378767e-01 5.98644078e-01 1.13049440e-01 -8.75651389e-02
-3.06027960e-02 -2.40594849e-01 1.18044518e-01 -3.89664114e-01
7.05894887e-01 4.24003154e-01 -5.01599669e-01 -1.61169186e-01
-7.26605952e-01 -1.64057687e-01 5.38350046e-01 -7.87651777e-01
1.25363457e+00 -3.17614228e-01 -8.41999412e-01 3.34587067e-01
1.07356638e-01 3.39992255e-01 -9.87320662e-01 9.95360091e-02
4.82640527e-02 -2.32999533e-01 2.02333909e-02 -9.45808887e-01
-3.33266675e-01 1.37181938e-01 -3.70196134e-01 5.51482379e-01
9.07210171e-01 4.70423102e-01 1.07984602e-01 5.70759118e-01
7.58302748e-01 -1.12984538e+00 -6.93109632e-01 6.02868497e-01
9.31093931e-01 -1.20335527e-01 3.35423678e-01 -8.02229583e-01
-4.72067803e-01 7.68153191e-01 2.95830280e-01 1.23981193e-01
3.18111509e-01 9.98825669e-01 -4.48096782e-01 -6.46981895e-01
-1.24122560e+00 -1.93268478e-01 -6.55772686e-01 4.55305874e-01
-4.32892889e-02 4.73636478e-01 -5.68993032e-01 6.32775009e-01
-1.34428948e-01 2.68513530e-01 9.83628571e-01 1.25627291e+00
-4.66088653e-01 -9.91571188e-01 -6.23240709e-01 -1.94127932e-02
-6.15463853e-01 -1.05580404e-01 -4.06285793e-01 9.86910224e-01
-1.21035784e-01 1.23840666e+00 -2.38710105e-01 -8.52562860e-02
1.52475566e-01 6.88130967e-03 9.12609696e-01 -5.76335609e-01
-2.23801196e-01 -1.42458543e-01 8.01336646e-01 -9.43293631e-01
-6.47278965e-01 -7.71005332e-01 -1.63823378e+00 -6.77680016e-01
-4.90770340e-01 3.11279327e-01 -5.65701490e-03 1.15853238e+00
-2.55525976e-01 2.99558848e-01 2.11682990e-01 -1.60822228e-01
-9.27246571e-01 -4.40684736e-01 -7.53941596e-01 -1.75931245e-01
9.59858000e-02 -6.86865985e-01 -5.08381426e-01 2.48725284e-02] | [8.486250877380371, 6.540452003479004] |
7d4e160e-ffe7-4ef5-9767-c148f2f211d0 | open-set-likelihood-maximization-for-few-shot | 2301.08390 | null | https://arxiv.org/abs/2301.08390v2 | https://arxiv.org/pdf/2301.08390v2.pdf | Open-Set Likelihood Maximization for Few-Shot Learning | We tackle the Few-Shot Open-Set Recognition (FSOSR) problem, i.e. classifying instances among a set of classes for which we only have a few labeled samples, while simultaneously detecting instances that do not belong to any known class. We explore the popular transductive setting, which leverages the unlabelled query instances at inference. Motivated by the observation that existing transductive methods perform poorly in open-set scenarios, we propose a generalization of the maximum likelihood principle, in which latent scores down-weighing the influence of potential outliers are introduced alongside the usual parametric model. Our formulation embeds supervision constraints from the support set and additional penalties discouraging overconfident predictions on the query set. We proceed with a block-coordinate descent, with the latent scores and parametric model co-optimized alternately, thereby benefiting from each other. We call our resulting formulation \textit{Open-Set Likelihood Optimization} (OSLO). OSLO is interpretable and fully modular; it can be applied on top of any pre-trained model seamlessly. Through extensive experiments, we show that our method surpasses existing inductive and transductive methods on both aspects of open-set recognition, namely inlier classification and outlier detection. | ['Ismail Ben Ayed', 'Céline Hudelot', 'Pablo Piantanida', 'Antoine Toubhans', 'Myriam Tami', 'Etienne Bennequin', 'Malik Boudiaf'] | 2023-01-20 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Boudiaf_Open-Set_Likelihood_Maximization_for_Few-Shot_Learning_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Boudiaf_Open-Set_Likelihood_Maximization_for_Few-Shot_Learning_CVPR_2023_paper.pdf | cvpr-2023-1 | ['few-shot-image-classification', 'open-set-learning'] | ['computer-vision', 'miscellaneous'] | [ 3.49328876e-01 3.02980930e-01 -2.39953920e-01 -4.73718494e-01
-1.34387696e+00 -5.78073561e-01 4.28931773e-01 3.52918804e-01
-1.86619535e-01 6.47150040e-01 -6.99765831e-02 7.62529895e-02
-2.89541423e-01 -5.16640723e-01 -1.07370710e+00 -7.58206069e-01
1.21014483e-01 8.79970014e-01 1.06769226e-01 8.68538991e-02
1.65257186e-01 1.78331539e-01 -1.64760005e+00 2.95127392e-01
9.86048698e-01 1.05568695e+00 -5.50145030e-01 4.89510566e-01
1.65680140e-01 7.84248233e-01 -2.54725963e-01 -5.27357996e-01
3.53642344e-01 -1.07308574e-01 -5.45052350e-01 6.81868732e-01
7.44394839e-01 5.49472347e-02 -2.41412949e-02 1.00181115e+00
3.84827822e-01 3.73272508e-01 7.97204673e-01 -1.43660891e+00
-6.17362738e-01 2.03450903e-01 -5.06790638e-01 -1.81325171e-02
4.44107264e-01 -4.33458276e-02 1.34790015e+00 -1.52896297e+00
5.70034921e-01 1.05954182e+00 8.32264304e-01 1.31014273e-01
-1.76860869e+00 -1.93856597e-01 3.05575222e-01 9.53717306e-02
-1.42448711e+00 -6.61947787e-01 5.84880888e-01 -6.47504449e-01
7.00499177e-01 4.79155958e-01 3.98294330e-01 1.05538988e+00
-3.16689610e-01 9.29849863e-01 8.16882610e-01 -5.49698532e-01
5.43411434e-01 2.65986472e-01 2.83198655e-01 7.91398823e-01
3.49062979e-01 5.70551418e-02 -6.84948325e-01 -7.07164288e-01
3.64566624e-01 2.88724601e-01 -2.02209651e-01 -1.23240507e+00
-1.19714034e+00 9.08567846e-01 2.17722252e-01 -1.14572316e-01
-1.92540050e-01 -1.06241733e-01 3.31285328e-01 5.12385786e-01
8.15433443e-01 2.31865078e-01 -3.33317816e-01 2.29261488e-01
-9.15974379e-01 9.57494527e-02 9.34186101e-01 9.31888700e-01
1.09190536e+00 -2.46746615e-01 -1.70005634e-01 1.01401079e+00
5.86071491e-01 1.53907225e-01 3.35159451e-01 -8.36748838e-01
2.67320961e-01 7.60104835e-01 1.85218021e-01 -5.49511015e-01
9.06497911e-02 -3.75863820e-01 -3.92576188e-01 4.92866859e-02
4.71358269e-01 6.93636537e-02 -1.01002336e+00 1.57859159e+00
5.71888685e-01 5.46442747e-01 6.23552091e-02 6.63409829e-01
4.27269608e-01 4.51778561e-01 -2.90199846e-01 -5.28583229e-01
8.46754730e-01 -9.03497338e-01 -3.63349169e-01 -1.89032197e-01
1.00086689e+00 -5.86516559e-01 9.92847860e-01 5.32079220e-01
-9.12778556e-01 -1.59749463e-02 -9.94561672e-01 -1.45404533e-01
-2.03378290e-01 7.74045214e-02 4.77187663e-01 5.36700070e-01
-7.19862998e-01 6.35504425e-01 -9.24605608e-01 -2.94108421e-01
5.29124379e-01 3.58703613e-01 -4.67444301e-01 -2.06513345e-01
-5.94769955e-01 6.68751001e-01 3.14419389e-01 1.89056367e-01
-8.77338290e-01 -6.61359787e-01 -1.01746845e+00 -1.06201783e-01
1.10943055e+00 -6.87500119e-01 1.17863357e+00 -9.27427948e-01
-1.22777271e+00 1.15790308e+00 -2.36732125e-01 -3.74598533e-01
6.58806622e-01 -5.13494015e-01 -1.05299599e-01 -7.54480138e-02
2.69494772e-01 2.85041928e-01 1.03529823e+00 -1.27804077e+00
-2.28107005e-01 -6.28757000e-01 -2.57273644e-01 2.02415466e-01
-2.62950450e-01 -1.21583022e-01 -4.93950278e-01 -6.53938055e-01
4.63316888e-01 -1.06996262e+00 -4.15420294e-01 8.42698887e-02
-7.41751313e-01 -2.78590232e-01 5.66332936e-01 -2.40864173e-01
1.11228240e+00 -2.12481999e+00 2.32503459e-01 5.72695971e-01
3.52933466e-01 8.64349306e-02 1.49129808e-01 4.41580832e-01
-3.18191826e-01 -2.52479553e-01 -5.48991144e-01 -5.35381854e-01
1.87555075e-01 4.87699091e-01 -6.57136977e-01 7.86006093e-01
2.60832459e-01 8.41563940e-01 -9.75534976e-01 -5.84266782e-01
1.35614321e-01 -1.02295093e-02 -8.43293428e-01 2.17946723e-01
-5.20130813e-01 3.15064579e-01 -4.12685812e-01 8.83986115e-01
3.21048617e-01 -3.12569261e-01 -1.55540556e-01 1.45646781e-01
1.60154611e-01 2.54623815e-02 -1.55629528e+00 1.73263204e+00
-2.21818388e-01 2.29171231e-01 -2.15340853e-01 -1.13473916e+00
9.53494608e-01 3.49312037e-01 5.48124790e-01 1.00013852e-01
-3.65962461e-02 5.63945174e-01 -2.65807241e-01 -4.85174835e-01
1.54753342e-01 -3.10611248e-01 -8.47754776e-02 4.33006734e-01
3.99334371e-01 3.24689932e-02 1.43932477e-01 3.56199652e-01
1.06837773e+00 2.81194717e-01 5.21315873e-01 6.98985439e-03
3.18933934e-01 -8.97557959e-02 7.30108380e-01 1.01173162e+00
-1.08548030e-01 7.71852851e-01 6.17107749e-01 -2.69059092e-01
-8.18928063e-01 -1.41432178e+00 -2.93685257e-01 1.23224676e+00
-1.19334936e-01 -4.60707694e-01 -4.74405557e-01 -9.21387076e-01
1.05810367e-01 5.56848049e-01 -5.91192663e-01 -1.62302807e-01
-2.06026167e-01 -7.16891229e-01 2.56715328e-01 2.42266625e-01
-3.54335397e-01 -7.57172108e-01 -1.19687863e-01 3.11008189e-02
-6.32282123e-02 -7.23728001e-01 -3.51550609e-01 5.40633321e-01
-1.00091302e+00 -1.01100028e+00 -4.34792548e-01 -6.69541299e-01
7.45278716e-01 2.64333617e-02 1.17518282e+00 -9.13716704e-02
-3.17111015e-01 5.82808316e-01 -2.06242800e-01 -3.99724662e-01
-1.99782237e-01 -1.95923150e-01 3.36121142e-01 5.88755190e-01
5.70702791e-01 -6.00014985e-01 -1.20337658e-01 3.32253963e-01
-9.21759188e-01 -3.18971485e-01 4.37316000e-01 1.13253558e+00
9.70488667e-01 -5.50037146e-01 6.57974958e-01 -1.24914503e+00
2.70163976e-02 -7.72007406e-01 -4.78708446e-01 4.46029216e-01
-6.45920455e-01 4.19794954e-02 3.80178958e-01 -3.82455766e-01
-7.61843264e-01 2.07149044e-01 1.82683468e-01 -9.19196129e-01
-1.91901281e-01 3.11740190e-01 -3.26937586e-01 -4.29601781e-02
8.63561511e-01 -1.01511739e-01 -1.20062128e-01 -4.87521380e-01
4.73993152e-01 6.24918997e-01 6.58321559e-01 -6.13736451e-01
9.73652780e-01 7.28289247e-01 -2.42772788e-01 -8.56975257e-01
-1.43648791e+00 -9.71062303e-01 -8.29305947e-01 -9.37700346e-02
4.34637040e-01 -9.74652112e-01 -2.83165514e-01 1.24290392e-01
-7.96498775e-01 7.50987083e-02 -7.78502226e-01 5.19956052e-01
-8.62091482e-01 5.03136277e-01 -2.23592848e-01 -8.47444713e-01
1.08594671e-01 -9.29803252e-01 1.41576290e+00 -2.23620564e-01
-2.12400287e-01 -1.00066078e+00 3.76564413e-01 4.72972780e-01
-3.82365197e-01 3.19313675e-01 5.68661273e-01 -1.33581913e+00
-6.03040516e-01 -6.53179169e-01 1.78170085e-01 4.24458891e-01
-1.96670711e-01 -8.03901330e-02 -1.32426572e+00 -3.86896372e-01
1.65047511e-01 -7.60791659e-01 1.10048795e+00 1.78352430e-01
1.02371132e+00 -2.64225185e-01 -3.53534400e-01 5.98492265e-01
1.26581895e+00 -4.40924406e-01 4.68229681e-01 1.29776374e-01
8.33037615e-01 5.98941743e-01 7.13138521e-01 5.29848456e-01
1.13196388e-01 6.50322199e-01 3.46556276e-01 -2.28203442e-02
6.13997281e-01 -3.88478458e-01 4.85271364e-01 4.90297467e-01
2.75617748e-01 -1.49591580e-01 -7.00810134e-01 5.24159610e-01
-2.19545150e+00 -8.90489697e-01 -2.05236033e-01 2.69197822e+00
7.91876912e-01 1.70744359e-01 1.47589073e-01 2.18816698e-01
6.86039269e-01 2.69883424e-02 -7.50384450e-01 -1.42308936e-01
-4.85406257e-02 8.66089612e-02 2.18438581e-01 6.36755526e-01
-1.29646993e+00 7.42932439e-01 5.88776541e+00 8.36806953e-01
-5.56391060e-01 6.89895377e-02 6.80165052e-01 -4.80965048e-01
-4.20125186e-01 4.42665070e-01 -7.16990829e-01 1.71931714e-01
7.50217259e-01 -4.10708226e-02 3.31160218e-01 9.25732315e-01
6.86419616e-03 -1.42868057e-01 -1.79172051e+00 9.53856111e-01
3.84313613e-01 -1.00412738e+00 -9.11356285e-02 7.29549080e-02
8.57310534e-01 1.48382932e-01 -5.35802729e-03 4.69394803e-01
3.34310025e-01 -8.05338144e-01 6.46915793e-01 6.11195505e-01
5.02521694e-01 -3.49537671e-01 3.80108625e-01 5.09346068e-01
-6.57953024e-01 -1.15400866e-01 -2.68619388e-01 8.97896364e-02
-2.18110848e-02 8.46262515e-01 -8.31652641e-01 4.31164831e-01
4.80548114e-01 9.07422066e-01 -4.33811933e-01 1.22281766e+00
-1.40784934e-01 7.31247306e-01 -6.66685700e-01 3.72601330e-01
3.82980742e-02 -4.95426059e-01 9.74422038e-01 7.86664605e-01
1.52254567e-01 -1.03527680e-01 6.16554320e-01 9.00778830e-01
-1.48272976e-01 2.31784418e-01 -9.72624302e-01 2.64049709e-01
3.14713866e-01 1.05210757e+00 -5.94413221e-01 -4.44253206e-01
-4.98653084e-01 9.59854722e-01 5.65902054e-01 3.11842561e-01
-5.92772722e-01 -7.38449320e-02 3.19813997e-01 -6.38498217e-02
4.98807788e-01 2.28049949e-01 -2.86157399e-01 -1.64467645e+00
5.61264455e-01 -6.48358226e-01 9.07821596e-01 -5.67091048e-01
-1.55026615e+00 6.54877489e-03 -1.43614411e-01 -1.51069546e+00
-2.75534451e-01 -4.60492104e-01 -6.06065512e-01 4.34187919e-01
-1.27768958e+00 -9.76941049e-01 1.54972196e-01 4.59417552e-01
4.91932482e-01 1.28149018e-01 7.73005009e-01 1.58353016e-01
-8.81407857e-01 5.79904616e-01 4.24276173e-01 -1.99079245e-01
8.26478660e-01 -1.50832021e+00 -4.40623835e-02 1.05731618e+00
6.22024834e-01 5.94034016e-01 6.63407385e-01 -6.21181667e-01
-1.22518325e+00 -1.27061832e+00 9.07788873e-01 -8.66644621e-01
9.05475616e-01 -5.65092146e-01 -1.15088189e+00 1.13952875e+00
-5.72647631e-01 5.87736309e-01 1.04710042e+00 3.83022726e-01
-5.59973896e-01 -1.47832539e-02 -9.74411905e-01 4.80452687e-01
9.86744165e-01 -5.50330281e-01 -7.87226975e-01 6.52417958e-01
6.37716889e-01 -3.12660903e-01 -7.90889502e-01 2.92490333e-01
1.67700723e-01 -9.71666873e-01 1.05999899e+00 -9.21046793e-01
1.50327459e-01 -2.61364132e-01 -3.77806455e-01 -1.17064142e+00
4.16834764e-02 -9.35790181e-01 -4.89504546e-01 1.14618134e+00
4.58873600e-01 -5.53045392e-01 8.76743019e-01 7.17628777e-01
-3.04455429e-01 -9.43939209e-01 -1.13704896e+00 -8.94245267e-01
-2.78034508e-01 -7.08506525e-01 -5.67021407e-02 8.91365469e-01
2.00673640e-01 5.20527482e-01 -4.94032949e-01 3.89268547e-01
8.67718816e-01 1.52398586e-01 8.56397510e-01 -1.43393362e+00
-6.18085265e-01 1.45253301e-01 -4.49086219e-01 -9.61467147e-01
9.06086639e-02 -1.06270683e+00 2.24552199e-01 -9.99848127e-01
2.92851061e-01 -5.06492794e-01 -2.86121666e-01 6.28563821e-01
-2.88401157e-01 5.21811247e-01 -1.82820082e-01 5.09097636e-01
-1.10069764e+00 7.89996028e-01 6.85126364e-01 1.33816451e-01
-3.85824680e-01 2.33444422e-01 -6.73678100e-01 1.07992899e+00
2.01518342e-01 -6.79752052e-01 -2.15462640e-01 -1.21675968e-01
2.89210290e-01 5.32450266e-02 5.75273395e-01 -8.06224883e-01
1.71559766e-01 1.28426366e-02 1.73811331e-01 -4.21455622e-01
4.04206604e-01 -7.30412483e-01 -2.55981833e-01 4.71149459e-02
-5.29635847e-01 -4.03738856e-01 -2.92052627e-01 1.18113410e+00
-7.54961744e-02 -3.83287281e-01 7.40541399e-01 1.02563873e-01
-4.11809117e-01 4.38512295e-01 -1.11104988e-01 3.50459099e-01
1.26941049e+00 -4.71401989e-01 7.39671439e-02 -1.48417041e-01
-1.15388155e+00 2.90959507e-01 5.84747970e-01 2.05776498e-01
5.29600739e-01 -1.44116974e+00 -6.70662999e-01 3.09109867e-01
6.88560069e-01 3.49558622e-01 -2.04997919e-02 1.28158045e+00
5.95790111e-02 2.19644994e-01 5.31585336e-01 -9.24500704e-01
-1.09124732e+00 6.92130804e-01 3.03687900e-01 -2.28456765e-01
-6.31089926e-01 8.62617791e-01 2.41866454e-01 -7.90222049e-01
4.67769563e-01 -1.95408911e-01 1.47210285e-01 9.28685218e-02
2.26451322e-01 4.83824521e-01 4.16315019e-01 -5.91499925e-01
-2.37994611e-01 2.33977973e-01 -2.44605064e-01 3.93915996e-02
1.37758517e+00 -5.60046360e-02 -2.16491878e-01 1.06736839e+00
1.19983470e+00 4.08444665e-02 -1.19179749e+00 -7.85379708e-01
3.37949902e-01 -6.89106584e-01 -2.23959148e-01 -5.05356669e-01
-3.75850290e-01 5.03949344e-01 3.68273109e-01 1.25571042e-01
7.71195948e-01 3.28393877e-01 4.59649861e-01 7.10773528e-01
2.51891553e-01 -1.10470831e+00 2.73656964e-01 3.71130049e-01
6.41524851e-01 -1.55406463e+00 -2.37913076e-02 -6.18788183e-01
-5.23572266e-01 8.67093682e-01 4.47748721e-01 -3.41066778e-01
5.01886249e-01 -1.03559241e-01 -4.34961580e-02 -4.48568732e-01
-9.24118578e-01 -1.60234272e-01 4.48529869e-01 3.63257855e-01
2.27578685e-01 -7.70917349e-03 1.45183459e-01 3.72395396e-01
1.17288768e-01 -6.04282171e-02 4.55678135e-01 9.66256678e-01
-5.66266954e-01 -8.29701424e-01 -6.99152231e-01 8.49918723e-01
-2.89811045e-01 -1.39318720e-01 -4.62156653e-01 3.01915318e-01
1.29151762e-01 7.99682021e-01 -2.97702905e-02 -5.07197753e-02
2.90535331e-01 3.94502103e-01 2.91152507e-01 -1.10377514e+00
-1.15000233e-01 2.49089107e-01 7.20906034e-02 -6.37318075e-01
-2.19968095e-01 -1.03793287e+00 -8.35755348e-01 3.11791718e-01
-8.21986318e-01 3.05990934e-01 1.40208006e-01 1.20253968e+00
1.77957952e-01 3.18889394e-02 7.69838035e-01 -8.43999624e-01
-1.06432939e+00 -8.19590032e-01 -6.01891577e-01 4.65451658e-01
4.76687372e-01 -7.22407937e-01 -6.13491237e-01 1.35289907e-01] | [9.592618942260742, 3.1508677005767822] |
d5894520-fc74-41fe-96b2-62b03ea94c4c | the-effectiveness-of-mae-pre-pretraining-for | 2303.13496 | null | https://arxiv.org/abs/2303.13496v1 | https://arxiv.org/pdf/2303.13496v1.pdf | The effectiveness of MAE pre-pretraining for billion-scale pretraining | This paper revisits the standard pretrain-then-finetune paradigm used in computer vision for visual recognition tasks. Typically, state-of-the-art foundation models are pretrained using large scale (weakly) supervised datasets with billions of images. We introduce an additional pre-pretraining stage that is simple and uses the self-supervised MAE technique to initialize the model. While MAE has only been shown to scale with the size of models, we find that it scales with the size of the training dataset as well. Thus, our MAE-based pre-pretraining scales with both model and data size making it applicable for training foundation models. Pre-pretraining consistently improves both the model convergence and the downstream transfer performance across a range of model scales (millions to billions of parameters), and dataset sizes (millions to billions of images). We measure the effectiveness of pre-pretraining on 10 different visual recognition tasks spanning image classification, video recognition, object detection, low-shot classification and zero-shot recognition. Our largest model achieves new state-of-the-art results on iNaturalist-18 (91.3%), 1-shot ImageNet-1k (62.1%), and zero-shot transfer on Food-101 (96.0%). Our study reveals that model initialization plays a significant role, even for web-scale pretraining with billions of images. | ['Ishan Misra', 'Rohit Girdhar', 'Ross Girshick', 'Christoph Feichtenhofer', 'Piotr Dollár', 'Armand Joulin', 'Aaron Adcock', 'Vaibhav Aggarwal', 'Haoqi Fan', 'Kalyan Vasudev Alwala', 'Quentin Duval', 'Mannat Singh'] | 2023-03-23 | null | null | null | null | ['video-classification', 'zero-shot-transfer-image-classification', 'video-recognition', 'action-classification', 'few-shot-image-classification'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [ 3.50868762e-01 -2.44537815e-01 -4.46338147e-01 -3.71029198e-01
-5.55838883e-01 -3.43053222e-01 7.26639330e-01 -2.64722496e-01
-8.44019234e-01 2.33016133e-01 -1.98257029e-01 -3.48439991e-01
2.69516885e-01 -5.71146667e-01 -1.22251582e+00 -3.67181152e-01
-4.53328937e-02 4.56881106e-01 4.56620336e-01 -2.54452646e-01
4.53108270e-03 4.90197763e-02 -1.81727517e+00 6.38514638e-01
4.89787102e-01 1.31000793e+00 3.01864237e-01 9.89824831e-01
1.65416703e-01 7.78325379e-01 -4.88218874e-01 -2.19929531e-01
4.55604732e-01 -2.86824167e-01 -7.11345375e-01 2.28417605e-01
1.01491904e+00 -6.23822033e-01 -3.98507535e-01 9.09279823e-01
4.39161390e-01 5.66084124e-02 7.22882330e-01 -1.37671423e+00
-1.01424325e+00 5.97769916e-01 -6.87339246e-01 5.17822206e-01
-3.02277118e-01 6.61543012e-01 7.27852106e-01 -1.10426307e+00
6.75172150e-01 1.25447512e+00 7.76719511e-01 8.24976504e-01
-1.24546397e+00 -7.29715765e-01 -2.87386030e-02 2.36907810e-01
-1.16345274e+00 -5.79117060e-01 6.02238663e-02 -5.72447181e-01
1.51915467e+00 -1.20974831e-01 6.64264977e-01 1.53372645e+00
1.27873138e-01 7.23797977e-01 1.00029111e+00 -5.76275468e-01
2.16183439e-01 3.45527641e-02 5.30837119e-01 7.66938090e-01
2.53852457e-01 3.22466880e-01 -6.67966723e-01 2.96798646e-01
8.73593926e-01 -4.36186492e-02 6.92039132e-02 -1.93839088e-01
-1.18188143e+00 8.85878742e-01 5.00914276e-01 6.47846609e-02
-5.00166081e-02 4.71279174e-01 5.45101643e-01 6.33308291e-01
2.69800931e-01 4.25315410e-01 -5.23551643e-01 -1.57519922e-01
-9.50641930e-01 -2.27400750e-01 7.45131791e-01 1.00823545e+00
7.11564720e-01 5.01461625e-01 -1.48578301e-01 9.77962255e-01
4.25521545e-02 6.81138396e-01 8.51653159e-01 -8.79813194e-01
5.11441529e-01 2.87168384e-01 -1.80853084e-01 -4.60118860e-01
-1.26519129e-01 -4.31996346e-01 -8.10374558e-01 2.82787710e-01
4.58604515e-01 -1.67018488e-01 -1.84233689e+00 1.73066247e+00
2.47171130e-02 5.37753165e-01 2.59952843e-01 8.61335516e-01
1.09015596e+00 7.85792410e-01 2.95187026e-01 1.75708219e-01
1.28780043e+00 -1.56483960e+00 9.43338848e-04 -5.90308845e-01
6.33617103e-01 -4.53056872e-01 1.42681777e+00 1.29590407e-01
-9.47237670e-01 -9.12307382e-01 -1.06195068e+00 3.24790105e-02
-5.38751602e-01 -7.78132454e-02 6.48211539e-01 6.01450741e-01
-1.09326184e+00 5.56308687e-01 -7.55923092e-01 -7.89622605e-01
6.67281091e-01 4.64072645e-01 -4.58567798e-01 -4.20556188e-01
-9.05991554e-01 9.75549757e-01 4.55993742e-01 -3.35232794e-01
-1.52690744e+00 -1.08627725e+00 -7.61628449e-01 1.97839066e-01
2.78956890e-01 -8.03858817e-01 1.24464917e+00 -1.31202543e+00
-1.23939228e+00 1.06365407e+00 2.77054518e-01 -8.29071641e-01
3.85543406e-01 -1.55194134e-01 -3.42491060e-01 1.17353521e-01
-1.94881216e-01 1.43006516e+00 1.14882803e+00 -9.94788706e-01
-5.90014875e-01 -8.62855539e-02 7.76816308e-02 6.56552762e-02
-7.44953752e-01 6.24359325e-02 -6.67711496e-01 -4.29988444e-01
-5.26488483e-01 -8.81725013e-01 1.06739113e-02 2.92443559e-02
8.19774941e-02 -5.37295686e-03 9.03970301e-01 -3.91441882e-01
7.19132304e-01 -2.15606570e+00 2.58412119e-02 -1.90371960e-01
1.07564211e-01 5.85496902e-01 -9.36492503e-01 3.50286663e-02
-1.94189265e-01 1.94345430e-01 -1.61476418e-01 -4.02916759e-01
-8.20749551e-02 3.85330051e-01 -2.28487283e-01 1.86780810e-01
2.91615516e-01 1.35514164e+00 -5.78165889e-01 -1.83405787e-01
3.42266709e-01 3.71777117e-01 -6.13454521e-01 2.63779104e-01
-4.52298522e-02 -1.76756442e-01 1.73884764e-01 8.06258440e-01
3.91502917e-01 -7.64136016e-01 -3.30988765e-01 -2.23426774e-01
1.91754997e-01 -4.66565460e-01 -7.34775662e-01 1.84635031e+00
-2.51702368e-01 6.03026986e-01 -1.47789627e-01 -1.08895445e+00
5.46220899e-01 8.58335756e-03 2.68965602e-01 -8.15930426e-01
2.01511964e-01 4.72520590e-02 3.17489564e-01 -3.66662592e-01
2.27253780e-01 -4.81958762e-02 3.54434401e-02 1.67979747e-01
7.96058595e-01 1.76079154e-01 5.44366360e-01 2.33425647e-01
1.15847063e+00 -2.28015006e-01 1.14852816e-01 -2.26846337e-01
-1.00532204e-01 1.28001049e-01 1.64798483e-01 1.05604684e+00
-1.63503796e-01 7.27348328e-01 1.57601178e-01 -5.42363286e-01
-1.38044667e+00 -1.07670438e+00 -1.14413843e-01 1.50884497e+00
8.93015563e-02 -2.85757095e-01 -1.00528407e+00 -6.96119606e-01
2.19360605e-01 5.37613869e-01 -1.04155123e+00 -4.27958518e-01
-9.16627273e-02 -9.73664880e-01 5.26940644e-01 8.36614430e-01
7.81252503e-01 -1.18761897e+00 -8.21575165e-01 -1.96486879e-02
4.11142886e-01 -1.38947701e+00 -5.38639307e-01 3.66856009e-01
-8.61217618e-01 -1.04784870e+00 -7.84007251e-01 -8.34546924e-01
6.99544072e-01 4.98874515e-01 1.22416055e+00 1.62867963e-01
-6.82748437e-01 5.56801617e-01 -4.36799228e-01 -4.16500658e-01
-3.32014233e-01 1.32006347e-01 2.32796520e-01 -1.90149441e-01
3.65398616e-01 -4.52133894e-01 -4.31688458e-01 4.77118224e-01
-9.43547547e-01 2.75082290e-01 7.41218626e-01 1.11758304e+00
4.23374027e-01 -5.06552458e-01 5.04041314e-01 -8.33491206e-01
2.44685367e-01 -4.60266411e-01 -5.98075449e-01 5.04050791e-01
-6.95629060e-01 -9.18844715e-02 6.13081813e-01 -9.92764771e-01
-6.69302225e-01 -2.60620303e-02 9.01138708e-02 -1.15238893e+00
-2.86670685e-01 3.13840747e-01 3.10726553e-01 -4.69230890e-01
9.92781639e-01 2.63022780e-01 -5.22660017e-02 -4.00929093e-01
5.24294138e-01 4.51714873e-01 6.31855249e-01 -3.08586806e-01
7.62927890e-01 1.70843914e-01 -3.11183035e-01 -9.42557216e-01
-7.67166197e-01 -3.98647457e-01 -5.96730888e-01 -6.79975376e-02
8.78611624e-01 -1.23216856e+00 -3.82918239e-01 8.05222631e-01
-8.35861444e-01 -9.34975624e-01 -5.07339895e-01 4.16216999e-01
-3.83877367e-01 -6.99658021e-02 -8.62388194e-01 -2.69259423e-01
-6.53214872e-01 -1.11137927e+00 8.47337067e-01 2.03827649e-01
2.36559920e-02 -7.16826260e-01 -7.33079314e-02 3.39568108e-01
6.41478539e-01 -4.84852456e-02 7.52413034e-01 -4.58730191e-01
-3.00167680e-01 4.48358012e-03 -6.60642385e-01 5.87503076e-01
-2.62354374e-01 -1.34834781e-01 -9.92789805e-01 -5.05398393e-01
-4.74962026e-01 -1.14529347e+00 1.39271295e+00 4.25394505e-01
1.03714228e+00 -1.38787612e-01 -5.76450042e-02 1.01779222e+00
1.45265055e+00 -4.32455465e-02 7.57340789e-01 3.29890668e-01
8.20795715e-01 1.79372966e-01 2.64771193e-01 2.09588364e-01
1.70258716e-01 4.49790299e-01 3.66936445e-01 -2.76118606e-01
-5.41023254e-01 -2.29396805e-01 5.53816378e-01 8.11235726e-01
-1.15018681e-01 -1.69701606e-01 -1.00586188e+00 4.28872824e-01
-1.79667222e+00 -9.02367532e-01 4.15642411e-01 1.97299635e+00
6.13014758e-01 2.24505022e-01 2.46690139e-01 -3.79864722e-01
5.66765606e-01 6.28213882e-02 -1.03013599e+00 -2.76094019e-01
-1.86356544e-01 1.38739377e-01 6.88455820e-01 1.50240958e-01
-1.19321477e+00 1.34200633e+00 7.13295507e+00 9.39534247e-01
-1.36745453e+00 3.73734593e-01 9.43575978e-01 -4.51119214e-01
2.49636769e-01 -3.46938580e-01 -1.07131159e+00 4.63087916e-01
1.32169414e+00 -8.83957930e-03 6.15782857e-01 1.23527849e+00
-3.69784862e-01 -1.24346390e-01 -1.30127513e+00 1.22695851e+00
3.82171094e-01 -1.64030910e+00 2.86887646e-01 2.55592708e-02
1.07116640e+00 8.76413643e-01 2.23447219e-01 7.85982370e-01
3.01746428e-01 -1.21451235e+00 8.56689632e-01 1.85171098e-01
1.13707793e+00 -4.23245341e-01 5.39996803e-01 2.98512489e-01
-1.05240905e+00 -2.94899017e-01 -6.66424274e-01 2.61846054e-02
-1.77120924e-01 1.04075439e-01 -4.57046777e-01 -1.42965361e-01
9.68727887e-01 7.80308187e-01 -9.50294316e-01 9.33902264e-01
8.30493420e-02 8.44706833e-01 -4.90774781e-01 1.70011073e-01
3.84667337e-01 2.82164097e-01 -5.07223234e-02 1.43186200e+00
2.85293847e-01 1.82017148e-01 3.51285040e-01 5.09790361e-01
-5.45459092e-01 -1.93205073e-01 -4.70919400e-01 -2.23951593e-01
1.74142405e-01 1.17811131e+00 -7.69960821e-01 -7.34913051e-01
-5.78666568e-01 1.03130448e+00 5.92501640e-01 5.25198877e-01
-9.52282071e-01 -1.29942656e-01 5.21334708e-01 1.89647228e-02
6.41523540e-01 -1.41536489e-01 -1.13092363e-01 -1.44178843e+00
-2.72326261e-01 -1.07064950e+00 4.48219270e-01 -7.96670377e-01
-1.31516087e+00 5.62100708e-01 9.64836106e-02 -9.04761076e-01
-1.91576853e-01 -9.86867309e-01 -6.41206622e-01 3.92687261e-01
-1.46258378e+00 -1.31249082e+00 -4.68131274e-01 5.95100522e-01
8.09534371e-01 -5.13449430e-01 9.39203382e-01 2.46890575e-01
-8.85132253e-01 8.24946404e-01 1.11383691e-01 2.53850222e-01
6.52304053e-01 -9.32117105e-01 6.48907661e-01 8.11966181e-01
3.11655968e-01 3.74029756e-01 4.15081412e-01 -3.08879644e-01
-1.71556747e+00 -1.11347723e+00 6.08803630e-02 -4.27220762e-01
8.63889575e-01 -4.58004445e-01 -9.28489566e-01 7.34461963e-01
3.74808073e-01 7.61091113e-01 5.08805692e-01 1.21153943e-01
-9.10590529e-01 -3.48971069e-01 -9.23032284e-01 3.04841280e-01
1.25129902e+00 -4.36984628e-01 -4.56314236e-01 1.93135455e-01
7.33223975e-01 -3.06158811e-01 -9.05733168e-01 4.38702792e-01
8.18579495e-01 -6.85367882e-01 1.13277459e+00 -9.74752545e-01
7.11292863e-01 1.89556301e-01 -3.26678276e-01 -1.23551786e+00
-4.52135444e-01 -1.88541532e-01 -5.38813829e-01 8.11576843e-01
6.13410294e-01 -3.38034630e-01 9.05424476e-01 6.05277717e-01
-4.39125113e-02 -7.94625401e-01 -7.19374001e-01 -1.12128925e+00
-4.91292328e-02 -3.10564518e-01 3.68422680e-02 9.29431856e-01
-2.33569339e-01 6.73357010e-01 -6.54433429e-01 -3.01574588e-01
7.09832788e-01 9.95867327e-03 9.80220795e-01 -9.56855893e-01
-5.80241084e-01 -3.42688262e-01 -4.65611398e-01 -9.27252948e-01
1.20649487e-02 -8.00256133e-01 4.12691906e-02 -1.30071199e+00
5.89951694e-01 -1.47403702e-01 -6.68120086e-01 9.60137308e-01
-4.91074510e-02 5.60614645e-01 5.96063614e-01 3.29035580e-01
-8.78411770e-01 4.18674231e-01 1.04198396e+00 -3.94919246e-01
1.37488022e-02 -4.92878556e-01 -4.03477818e-01 5.34171581e-01
6.54632390e-01 -4.37849343e-01 -5.64734936e-01 -7.87766993e-01
-4.51438248e-01 -3.83163810e-01 4.82733727e-01 -1.34695423e+00
1.41412556e-01 -1.90163285e-01 6.27353787e-01 -1.06124312e-01
3.80450547e-01 -5.24100900e-01 -2.12027028e-01 6.79846168e-01
-3.93609256e-01 1.60554182e-02 5.43336689e-01 5.55130064e-01
-8.43448157e-04 -1.87564611e-01 1.07948101e+00 -1.81473389e-01
-1.36810601e+00 5.33993065e-01 -8.25745687e-02 2.39516929e-01
1.03991079e+00 -2.18867883e-01 -7.10697114e-01 -9.64687690e-02
-5.26240647e-01 1.53098956e-01 4.42069679e-01 7.04270601e-01
6.27151072e-01 -1.16384459e+00 -6.97025537e-01 2.83007264e-01
3.92216116e-01 -3.49879950e-01 2.84565240e-01 6.31529450e-01
-3.97041768e-01 2.79613465e-01 -6.55196726e-01 -8.90445411e-01
-1.33968651e+00 7.03376412e-01 1.04889028e-01 -1.19480342e-01
-5.75572491e-01 1.09874535e+00 4.02781069e-01 -8.51523355e-02
4.84831363e-01 -4.43503469e-01 1.39798805e-01 -5.37947007e-02
8.17347825e-01 1.58362538e-01 -1.10798411e-01 -4.30584937e-01
-2.30735540e-01 6.76893055e-01 -3.39491874e-01 1.38345107e-01
1.43921888e+00 3.38485241e-01 4.50937986e-01 4.03960288e-01
1.37883222e+00 -1.02030385e+00 -1.69301164e+00 -1.05897680e-01
-3.32861125e-01 -1.52054742e-01 1.87675759e-01 -8.51548374e-01
-1.17352045e+00 1.21764302e+00 8.96756649e-01 -2.89900273e-01
9.04393315e-01 -7.94800818e-02 7.64622808e-01 7.68795729e-01
4.92937565e-01 -1.13758016e+00 4.96255189e-01 7.99229622e-01
7.53532350e-01 -1.72356498e+00 -2.04306826e-01 -5.64167723e-02
-6.85861230e-01 8.28295767e-01 1.11155570e+00 -9.79226083e-02
5.16833723e-01 4.39104199e-01 3.93103957e-02 7.68790096e-02
-1.05491364e+00 -1.97470173e-01 4.45650369e-01 5.45122445e-01
2.22839490e-02 -1.02244250e-01 3.36665899e-01 4.46317047e-01
1.35296479e-01 2.58372843e-01 2.07155481e-01 8.47243130e-01
-6.93572938e-01 -6.97112322e-01 -3.39823589e-02 6.81388617e-01
-1.38469730e-02 -4.70840931e-01 -1.60850853e-01 8.08146000e-01
4.96891476e-02 6.15063965e-01 2.53938913e-01 -6.47708952e-01
2.38546878e-01 1.37156978e-01 7.02416241e-01 -7.72857904e-01
-6.08275115e-01 -1.02464803e-01 1.10560646e-02 -6.20028496e-01
-3.02182615e-01 -1.38442457e-01 -9.39883709e-01 -3.84323031e-01
-3.13329756e-01 -2.54067898e-01 6.21315420e-01 8.41860414e-01
4.81274158e-01 4.20985430e-01 2.11491957e-01 -1.03976130e+00
-7.64104843e-01 -1.17146838e+00 -3.65994096e-01 5.17423809e-01
7.42449164e-02 -6.10620856e-01 -3.03956419e-01 3.60949844e-01] | [9.837603569030762, 1.9590505361557007] |
668ebbf8-2701-4cab-9d46-1f72681b6aa9 | analyzing-the-vulnerabilities-in-splitfed | 2307.03197 | null | https://arxiv.org/abs/2307.03197v1 | https://arxiv.org/pdf/2307.03197v1.pdf | Analyzing the vulnerabilities in SplitFed Learning: Assessing the robustness against Data Poisoning Attacks | Distributed Collaborative Machine Learning (DCML) is a potential alternative to address the privacy concerns associated with centralized machine learning. The Split learning (SL) and Federated Learning (FL) are the two effective learning approaches in DCML. Recently there have been an increased interest on the hybrid of FL and SL known as the SplitFed Learning (SFL). This research is the earliest attempt to study, analyze and present the impact of data poisoning attacks in SFL. We propose three kinds of novel attack strategies namely untargeted, targeted and distance-based attacks for SFL. All the attacks strategies aim to degrade the performance of the DCML-based classifier. We test the proposed attack strategies for two different case studies on Electrocardiogram signal classification and automatic handwritten digit recognition. A series of attack experiments were conducted by varying the percentage of malicious clients and the choice of the model split layer between the clients and the server. The results after the comprehensive analysis of attack strategies clearly convey that untargeted and distance-based poisoning attacks have greater impacts in evading the classifier outcomes compared to targeted attacks in SFL | ['Raj Mani Shukla', 'Aysha Thahsin Zahir Ismail'] | 2023-07-04 | null | null | null | null | ['data-poisoning', 'handwritten-digit-recognition', 'federated-learning'] | ['adversarial', 'computer-vision', 'methodology'] | [-2.67948121e-01 -2.44186148e-01 3.25684279e-01 -2.80929148e-01
-6.35964751e-01 -1.02684760e+00 6.95372283e-01 2.62411088e-01
-4.06026900e-01 6.75825775e-01 -4.34271805e-02 -5.38571179e-01
-5.15331388e-01 -4.51777548e-01 -2.39228144e-01 -1.06551433e+00
-3.13008815e-01 2.18644306e-01 1.89180031e-01 2.80013740e-01
3.38173687e-01 1.00269389e+00 -1.01973641e+00 6.58745229e-01
3.34822983e-01 9.14319634e-01 -8.75938535e-01 7.70518839e-01
-1.20143965e-01 1.11539531e+00 -1.10280645e+00 -7.62729585e-01
7.98062384e-01 -4.94840086e-01 -7.55643547e-01 -3.69369119e-01
-4.59102318e-02 -2.87776262e-01 -1.50094717e-03 9.92564082e-01
1.07425570e+00 -3.79332274e-01 3.36126834e-01 -1.95660865e+00
2.45714262e-01 8.74794066e-01 -4.39796180e-01 3.16939414e-01
2.79156536e-01 2.11688280e-01 4.00250018e-01 -4.40083653e-01
4.14907187e-01 1.24369025e+00 8.03592622e-01 5.27553558e-01
-1.09461379e+00 -9.25822794e-01 -2.69734234e-01 2.22551391e-01
-1.34285843e+00 -3.59580666e-01 8.69003415e-01 -2.96331018e-01
5.84193528e-01 5.77113807e-01 8.20030198e-02 1.12963676e+00
6.13561749e-01 4.26495701e-01 1.48616719e+00 -5.12696445e-01
6.60817266e-01 5.44156253e-01 7.33763501e-02 3.13599497e-01
6.65929019e-01 2.58049965e-01 -6.28573120e-01 -1.33107185e+00
2.81938732e-01 1.69252262e-01 -8.10365602e-02 -5.38206875e-01
-7.12255597e-01 8.70935440e-01 -9.44806486e-02 2.99732655e-01
-3.69187444e-01 -1.90056324e-01 9.49611366e-01 5.42141199e-01
4.31736290e-01 3.44867915e-01 -7.25856304e-01 2.87840329e-02
-6.46394074e-01 9.30538550e-02 1.32711864e+00 2.86030501e-01
2.44838238e-01 5.49238920e-02 -5.66591807e-02 1.12191491e-01
2.88388312e-01 6.64261952e-02 4.92667764e-01 -6.24888539e-01
3.49370480e-01 4.95966047e-01 7.48748779e-02 -9.79560196e-01
-9.58262309e-02 -3.57723176e-01 -5.44043601e-01 5.38669944e-01
4.99094248e-01 -8.34665775e-01 -4.18581776e-02 1.26499927e+00
6.55346990e-01 8.64525586e-02 4.08488601e-01 5.10075033e-01
2.87789375e-01 2.59065896e-01 2.10387751e-01 -2.26030484e-01
7.43460953e-01 -5.43118060e-01 -6.96042597e-01 6.27055645e-01
9.08622086e-01 -7.73277581e-01 3.58603984e-01 7.77296066e-01
-5.88464081e-01 6.64341077e-02 -8.85673583e-01 8.17016482e-01
-5.31746745e-01 -3.43010485e-01 5.44055581e-01 1.51808512e+00
-7.02955365e-01 4.76942986e-01 -8.08010638e-01 -2.99258769e-01
6.19451225e-01 4.73369181e-01 -3.87598366e-01 6.37082092e-04
-1.00193202e+00 9.01613057e-01 -1.66129217e-01 -3.56165320e-01
-1.04286981e+00 -6.08878076e-01 -8.66518840e-02 -1.33181304e-01
1.21596389e-01 -1.28901601e-01 9.00899291e-01 -1.13785553e+00
-1.17151427e+00 5.62583923e-01 6.75174594e-01 -9.51334953e-01
9.29939568e-01 1.48887992e-01 -4.11962926e-01 1.53802842e-01
-2.70863324e-01 -2.31994972e-01 8.86094630e-01 -1.13074291e+00
-6.28664434e-01 -6.31768823e-01 -1.61852971e-01 -1.32074684e-01
-4.20977861e-01 5.57471097e-01 6.96794033e-01 -4.73336279e-01
-5.19220948e-01 -5.66409171e-01 -8.31587166e-02 6.91465586e-02
-3.85008723e-01 -1.50170073e-01 1.45996749e+00 -3.70918423e-01
1.15862238e+00 -2.22776341e+00 -3.61898720e-01 4.93121088e-01
1.91382572e-01 5.51349819e-01 1.19382039e-01 9.58064854e-01
-6.38089627e-02 2.00016737e-01 -1.09822927e-02 -1.75449356e-01
-1.67339131e-01 1.52246714e-01 -2.47807369e-01 9.12490308e-01
-4.57560062e-01 4.00554210e-01 -4.23045218e-01 -4.77983922e-01
6.99179992e-02 4.38698024e-01 -3.50593895e-01 3.08938801e-01
-7.69320549e-03 5.18833756e-01 -5.11822402e-01 7.25186169e-01
6.07713819e-01 1.66151404e-01 3.33581209e-01 -1.76489085e-01
-5.17814718e-02 -2.97040403e-01 -1.28802478e+00 8.00715804e-01
-2.10055290e-03 1.93751186e-01 2.33913630e-01 -5.91762900e-01
1.03728104e+00 7.94460297e-01 7.31845319e-01 2.12855916e-02
3.43545169e-01 1.57190949e-01 1.88789740e-01 -4.82966244e-01
-3.98835480e-01 5.13569750e-02 -7.60807171e-02 8.46747100e-01
-9.99879986e-02 5.25366485e-01 -6.66419685e-01 2.17641383e-01
1.55397773e+00 -3.40799481e-01 4.95868266e-01 -3.60333771e-01
6.60052836e-01 -1.46670818e-01 6.28738403e-01 9.07779038e-01
-8.47569823e-01 -1.21028066e-01 6.69828773e-01 -5.75448453e-01
-5.70306778e-01 -7.08097756e-01 2.54531622e-01 7.91881144e-01
-9.42711756e-02 -3.36761922e-01 -9.04930770e-01 -1.39056742e+00
3.02736610e-01 6.02963030e-01 -3.83538574e-01 -3.20392340e-01
-1.13453269e-01 -6.57449365e-01 1.34891450e+00 1.08276792e-01
9.37275529e-01 -7.65530050e-01 -1.05618596e+00 -3.54807898e-02
3.30389977e-01 -6.25080109e-01 -1.09888487e-01 3.64388555e-01
-7.35461771e-01 -1.55779052e+00 -1.44405156e-01 -1.46169975e-01
4.53930914e-01 -2.19053522e-01 4.75364029e-01 8.09238851e-03
-6.49445474e-01 8.77221406e-01 -4.52725470e-01 -8.20336461e-01
-7.48273492e-01 2.55166320e-03 -1.85100604e-02 7.41828799e-01
5.84073663e-01 -4.27825481e-01 -4.69703466e-01 3.67712259e-01
-9.18037415e-01 -7.15039909e-01 2.73299903e-01 5.19951224e-01
-6.01761900e-02 2.42315084e-01 7.43190408e-01 -1.32247102e+00
1.03847814e+00 -6.54931009e-01 -5.55944979e-01 4.02605593e-01
-9.45504963e-01 -1.56158850e-01 8.00209463e-01 -6.64164066e-01
-9.72104132e-01 2.62179434e-01 2.18348518e-01 -7.11641729e-01
-4.22113329e-01 9.33264345e-02 -3.75965655e-01 -5.61061859e-01
9.47823167e-01 -5.26024178e-02 1.80898696e-01 -6.30123675e-01
-4.01368141e-02 9.76984560e-01 -1.21175990e-01 -3.67179662e-01
5.16525865e-01 6.31427646e-01 3.56196426e-02 -6.29815638e-01
-1.71189442e-01 -2.38287687e-01 -3.01720589e-01 -3.84441346e-01
5.81313252e-01 -3.30271244e-01 -8.87769580e-01 7.84206867e-01
-1.00899613e+00 -2.54965480e-02 -8.78174677e-02 3.65745753e-01
-1.86477795e-01 3.94583136e-01 -5.45473278e-01 -1.20003521e+00
-8.96423817e-01 -7.86755204e-01 2.66658753e-01 -4.39633839e-02
-2.54239231e-01 -9.12965000e-01 2.00493619e-01 3.18447262e-01
9.16668892e-01 8.29835057e-01 1.00558531e+00 -1.68718028e+00
-1.42929390e-01 -6.69091880e-01 4.65612888e-01 4.05473799e-01
2.69460499e-01 3.84924226e-02 -1.00130785e+00 -6.92047417e-01
4.66279447e-01 -2.37250969e-01 -1.53645113e-01 -2.97654957e-01
7.04005361e-01 -6.91712677e-01 -1.23571910e-01 3.00082386e-01
1.59279037e+00 5.44257939e-01 5.04870176e-01 2.96966851e-01
3.02314520e-01 4.48962390e-01 4.25287098e-01 9.06247020e-01
-2.87435353e-01 2.64687300e-01 4.95992243e-01 1.22246705e-01
2.26281211e-01 6.51593059e-02 4.38071728e-01 1.28677294e-01
2.87125617e-01 2.18053032e-02 -1.08070648e+00 1.30132303e-01
-1.71078515e+00 -8.87619972e-01 1.45106548e-02 2.36705971e+00
5.68891168e-01 -3.88294868e-02 3.16002220e-01 5.11795580e-01
8.84659708e-01 -1.43723562e-01 -6.16770744e-01 -8.66137922e-01
9.30750445e-02 2.59577841e-01 7.14492679e-01 1.35016367e-01
-1.10643017e+00 4.52899009e-01 5.77456760e+00 6.33511484e-01
-1.29965758e+00 4.83103812e-01 5.27164936e-01 -1.66815832e-01
2.95219362e-01 1.29352838e-01 -5.42240679e-01 4.94464159e-01
1.15527415e+00 -3.39109868e-01 2.49021500e-01 1.04372501e+00
7.71193281e-02 1.23692714e-01 -9.82386410e-01 9.99078453e-01
1.47325667e-02 -1.18550754e+00 -7.04497173e-02 8.22801739e-02
3.63233536e-01 6.79876357e-02 -3.33794206e-01 3.30245122e-03
4.33668107e-01 -6.82812214e-01 3.33999097e-01 4.76404458e-01
2.35546991e-01 -1.15318489e+00 8.39364231e-01 4.98542935e-01
-6.65068090e-01 -6.00601315e-01 -1.13235759e-02 9.31327716e-02
-4.13576543e-01 2.17310935e-01 -5.42198241e-01 4.83247101e-01
8.34813774e-01 -7.55643696e-02 -5.51630139e-01 1.05308843e+00
1.37863666e-01 9.11598265e-01 -2.96779662e-01 -2.06691042e-01
4.38412093e-02 3.05928022e-01 6.66669726e-01 9.27852750e-01
-3.90538067e-01 -3.47313955e-02 4.55499999e-02 4.61977094e-01
2.00301960e-01 1.37688011e-01 -8.31615865e-01 7.45234340e-02
7.53596306e-01 1.15932572e+00 -5.97107828e-01 6.46237880e-02
-3.71332318e-01 7.79260635e-01 -1.42835993e-02 1.27020106e-01
-6.44227982e-01 -4.72430497e-01 5.72806239e-01 1.56108007e-01
1.32742301e-01 -1.99533999e-03 -6.02812879e-02 -6.82646155e-01
-2.87650734e-01 -1.48153949e+00 9.30001080e-01 -1.96124867e-01
-1.67022765e+00 7.46602595e-01 1.08580947e-01 -9.84011829e-01
2.73991693e-02 -3.60658690e-02 -6.83939457e-01 6.93742573e-01
-7.88692594e-01 -1.09307480e+00 -6.65007830e-02 1.27774179e+00
2.62411863e-01 -6.66584194e-01 1.21178210e+00 2.36576542e-01
-5.99065244e-01 7.99908340e-01 3.29215974e-01 2.74937034e-01
7.32373416e-01 -5.81438482e-01 -3.94826561e-01 9.12791371e-01
4.81758043e-02 5.12162864e-01 5.62521279e-01 -7.35346377e-01
-1.44497049e+00 -1.07685041e+00 5.32403529e-01 -3.73000294e-01
2.78044909e-01 -4.05906260e-01 -6.42967343e-01 5.59548199e-01
3.88956964e-01 1.72750100e-01 1.12303317e+00 -5.28450608e-01
-4.71281350e-01 -4.54860151e-01 -2.16984534e+00 2.13637441e-01
1.16145842e-01 -4.33088630e-01 -2.07742378e-01 3.19927484e-01
1.96696877e-01 1.94208458e-01 -1.00431609e+00 7.99484774e-02
3.19677770e-01 -1.28645682e+00 4.87857550e-01 -8.50124002e-01
-3.02757770e-01 -2.17153132e-02 -3.33498031e-01 -7.05048144e-01
8.01011249e-02 -9.47638214e-01 -2.12486669e-01 1.52557957e+00
9.68250260e-02 -1.30667436e+00 8.66893113e-01 9.67031896e-01
8.37749898e-01 -7.50956833e-01 -1.12720335e+00 -6.86477125e-01
8.31896067e-02 -4.63762879e-03 6.57047272e-01 1.19215000e+00
-9.59932655e-02 -2.00599909e-01 -2.87230313e-01 2.95015514e-01
1.05725026e+00 -1.55509070e-01 9.38203692e-01 -1.11004996e+00
-4.28525299e-01 2.71558076e-01 -6.72988892e-01 4.14441049e-01
-2.36381844e-01 -6.94790959e-01 -8.00948024e-01 -7.57997811e-01
-1.24603786e-01 -4.58001137e-01 -3.72855932e-01 9.05134559e-01
5.20752192e-01 -2.04669029e-01 4.25425589e-01 5.31874657e-01
-2.88825333e-01 -5.98879531e-02 2.84814715e-01 1.18193746e-01
-7.07596242e-02 3.81174386e-01 -3.94043773e-01 5.07174492e-01
1.05904794e+00 -1.03770936e+00 -4.94889885e-01 1.50101438e-01
-2.46338814e-01 1.36489987e-01 4.03131336e-01 -8.56687248e-01
6.11296654e-01 -5.81056699e-02 2.87975520e-01 -2.91639477e-01
-3.00274372e-01 -1.43376708e+00 4.95665014e-01 9.55064893e-01
-5.93573153e-01 6.37067631e-02 -7.75999483e-03 6.33593202e-01
6.16132654e-02 -3.11555434e-02 1.02598691e+00 -1.96657926e-01
-2.08623447e-02 2.65632067e-02 -6.79585278e-01 -2.64972389e-01
1.64348626e+00 -1.57830447e-01 -3.65903944e-01 -3.51988494e-01
-7.59871185e-01 -9.33026448e-02 1.82938501e-01 3.55597228e-01
5.31030715e-01 -8.49906087e-01 -5.78950346e-01 4.08363163e-01
-3.31186235e-01 -6.65449262e-01 -1.60563990e-01 8.54765952e-01
-5.52423358e-01 2.35457286e-01 -3.50082159e-01 -1.91941604e-01
-1.90432072e+00 8.02420318e-01 8.68819654e-01 -3.05831373e-01
-5.24892688e-01 4.87809658e-01 -5.97697794e-01 -3.47035080e-01
8.14104080e-01 5.25023282e-01 1.30018711e-01 -1.83504716e-01
2.85955578e-01 9.91709590e-01 2.76253641e-01 -1.85998693e-01
-8.16732705e-01 -2.93086618e-01 -1.50681496e-01 4.97096255e-02
1.13622093e+00 1.24639407e-01 -3.32107335e-01 2.44604826e-01
1.17925620e+00 2.57157572e-02 -8.74577820e-01 -4.53329012e-02
2.83751577e-01 -5.62959075e-01 -3.97490198e-03 -1.19144440e+00
-1.14719021e+00 5.54798067e-01 1.10417879e+00 2.70465046e-01
1.08286369e+00 -4.91317779e-01 3.84172440e-01 2.48200923e-01
8.84789467e-01 -7.57365763e-01 -1.67712942e-01 -1.72133014e-01
6.02216244e-01 -6.30706370e-01 1.60091575e-02 -8.80284682e-02
-7.36770630e-01 1.02365696e+00 4.36076105e-01 -1.62476391e-01
1.06922328e+00 7.40303874e-01 3.61956209e-01 -2.25382447e-01
-8.63875389e-01 7.84957707e-01 -4.97161448e-01 6.38018310e-01
4.44761850e-02 -2.44945675e-01 -3.95581752e-01 8.42958331e-01
4.75987703e-01 8.05996172e-03 3.78149778e-01 1.49413562e+00
2.54251678e-02 -1.35802567e+00 -8.42847407e-01 5.00924706e-01
-8.92001569e-01 3.80853444e-01 -1.16686785e+00 6.70204937e-01
2.07776859e-01 1.13365972e+00 -5.16264796e-01 -4.88144487e-01
5.24025485e-02 4.63378847e-01 1.20567143e-01 -1.60931692e-01
-1.58649325e+00 -1.48673460e-01 -3.88847403e-02 -5.52650630e-01
-3.11686844e-01 -4.98223215e-01 -1.04878116e+00 -3.48781854e-01
-6.00255430e-01 5.40220797e-01 9.20619786e-01 6.39542162e-01
7.07616389e-01 -1.36912921e-02 1.12630022e+00 -2.96072990e-01
-1.25222814e+00 -5.95510781e-01 -9.19355810e-01 2.68152475e-01
4.48225550e-02 -2.22616747e-01 -8.24846566e-01 -3.49051952e-01] | [5.604349613189697, 7.054699897766113] |
56a487bb-3092-4466-8e6b-5a0feb93117e | anatomically-aware-dual-hop-learning-for | 2303.17593 | null | https://arxiv.org/abs/2303.17593v1 | https://arxiv.org/pdf/2303.17593v1.pdf | Anatomically aware dual-hop learning for pulmonary embolism detection in CT pulmonary angiograms | Pulmonary Embolisms (PE) represent a leading cause of cardiovascular death. While medical imaging, through computed tomographic pulmonary angiography (CTPA), represents the gold standard for PE diagnosis, it is still susceptible to misdiagnosis or significant diagnosis delays, which may be fatal for critical cases. Despite the recently demonstrated power of deep learning to bring a significant boost in performance in a wide range of medical imaging tasks, there are still very few published researches on automatic pulmonary embolism detection. Herein we introduce a deep learning based approach, which efficiently combines computer vision and deep neural networks for pulmonary embolism detection in CTPA. Our method features novel improvements along three orthogonal axes: 1) automatic detection of anatomical structures; 2) anatomical aware pretraining, and 3) a dual-hop deep neural net for PE detection. We obtain state-of-the-art results on the publicly available multicenter large-scale RSNA dataset. | ['Marius Leordeanu', 'A Mohamed Ali', 'Jonathan Sperl', 'Puneet Sharma', 'Lucian Itu', 'Saikiran Rapaka', 'Florin Condrea'] | 2023-03-30 | null | null | null | null | ['pulmonary-embolism-detection'] | ['medical'] | [ 1.07299440e-01 -5.62098995e-02 -2.22373217e-01 1.46966696e-01
-1.19270837e+00 -4.70778346e-01 2.25654811e-01 4.18488979e-01
-6.23641610e-01 6.61163747e-01 1.56689644e-01 -8.76343429e-01
-2.41857708e-01 -6.88546658e-01 -3.53250802e-01 -6.18190467e-01
-2.80004352e-01 1.20306981e+00 5.34070134e-01 4.53761667e-01
-2.53059894e-01 1.21114480e+00 -7.26598620e-01 1.73857942e-01
7.53532887e-01 1.03676152e+00 -1.02752998e-01 9.84440029e-01
2.77215764e-02 1.08274400e+00 -4.73728120e-01 -2.30592027e-01
2.15646207e-01 -6.49295270e-01 -9.30510819e-01 -5.27161658e-01
2.02783436e-01 -5.40158868e-01 -5.53294599e-01 5.31700969e-01
1.09875321e+00 -2.54600793e-01 8.33696187e-01 -5.52808166e-01
-4.68394458e-01 3.04413766e-01 -5.05546510e-01 1.07305753e+00
-1.73809245e-01 1.80432811e-01 7.84787893e-01 -8.35880816e-01
1.95222735e-01 6.85841322e-01 9.02836621e-01 4.47232604e-01
-7.60884941e-01 -5.34532845e-01 -7.29725838e-01 3.13663989e-01
-9.69350159e-01 -1.30116045e-01 3.51673633e-01 -4.92107123e-01
7.47010052e-01 1.63454250e-01 1.01742232e+00 9.08359528e-01
5.04641354e-01 3.82141441e-01 9.20156419e-01 -1.87690839e-01
-2.14650676e-01 -2.55898237e-01 -1.81445852e-01 9.06746805e-01
4.20065790e-01 6.44474089e-01 4.25527915e-02 -4.82071877e-01
1.16483641e+00 1.46666944e-01 -3.62791836e-01 -4.46239024e-01
-1.28609467e+00 7.70229638e-01 5.27141035e-01 4.18205798e-01
-7.35549033e-01 2.17378914e-01 4.99572843e-01 5.12770861e-02
-1.67145208e-01 5.27966499e-01 -1.09297030e-01 -2.93760091e-01
-9.78432834e-01 -1.33536726e-01 8.27784538e-01 5.14838472e-02
-2.10379720e-01 2.40410045e-02 -5.97651005e-01 6.26523793e-01
1.22016124e-01 5.69964707e-01 5.79086661e-01 -8.17911685e-01
2.30354682e-01 1.25477135e-01 2.78226826e-02 -7.10847974e-01
-6.97498620e-01 -9.93961453e-01 -1.03935170e+00 2.32633829e-01
6.20084882e-01 -2.00665340e-01 -7.38204062e-01 1.29204881e+00
6.01341166e-02 4.36590642e-01 -2.92124867e-01 1.00029004e+00
1.11214280e+00 9.50604007e-02 4.30748492e-01 -1.31668031e-01
1.53198028e+00 -1.11458111e+00 -3.63604307e-01 -5.96478069e-03
4.18695658e-01 -4.46575642e-01 6.02623105e-01 2.65557617e-01
-1.25466967e+00 -2.28003502e-01 -9.14993048e-01 -2.39064340e-02
1.41156428e-02 4.46166754e-01 6.11299872e-01 8.91000390e-01
-8.30071509e-01 5.75087667e-01 -1.04247916e+00 -1.84805363e-01
1.08502007e+00 3.86495382e-01 -2.26267219e-01 3.43768984e-01
-1.18378460e+00 1.14756799e+00 -1.00703500e-02 -9.41086933e-02
-1.24887955e+00 -1.12206745e+00 -1.76219016e-01 5.96401632e-01
6.80360317e-01 -1.26419139e+00 1.23920417e+00 -1.12808898e-01
-1.38576984e+00 1.18463409e+00 5.92225380e-02 -8.66272151e-01
8.76342297e-01 -4.19151515e-01 -2.76065439e-01 8.37363660e-01
-6.27086535e-02 3.81573319e-01 9.34044480e-01 -6.21194541e-01
-8.78490686e-01 -2.15580672e-01 -5.34281373e-01 7.67466007e-03
-2.94061620e-02 1.67901427e-01 -1.35143161e-01 -7.06486940e-01
-2.01453283e-01 -6.87266946e-01 -6.53662443e-01 3.87373835e-01
-1.23164400e-01 -2.17868745e-01 4.17058736e-01 -6.84387743e-01
1.16970563e+00 -1.63290071e+00 -1.53582752e-01 2.96367835e-02
1.14941192e+00 8.56920063e-01 2.11778149e-01 -1.71428725e-01
-3.85170519e-01 2.24275678e-01 -2.34974757e-01 2.20817208e-01
-3.65872204e-01 -1.68248624e-01 -1.24508798e-01 3.97574157e-01
1.47559062e-01 1.38789558e+00 -8.80614579e-01 -6.10211551e-01
5.70342541e-01 5.10085285e-01 -3.22884440e-01 2.85918325e-01
4.24705327e-01 7.76971340e-01 -5.06112099e-01 3.57141495e-01
3.17538202e-01 -9.45053160e-01 -6.61498383e-02 4.07657325e-02
1.32109538e-01 2.66501993e-01 -5.06268024e-01 1.24307382e+00
-2.45850980e-01 4.37624067e-01 -1.63982913e-01 -1.16699290e+00
5.51238418e-01 7.28719592e-01 9.77182448e-01 -3.99095923e-01
5.25559723e-01 3.40935141e-01 5.40020168e-01 -5.31561792e-01
-2.81484842e-01 -4.99661148e-01 4.58047748e-01 4.61696923e-01
-2.51959532e-01 2.95990169e-01 -6.33235499e-02 6.10886291e-02
1.60481834e+00 -4.29747105e-01 5.92930794e-01 -1.68139771e-01
8.34698021e-01 1.02509372e-01 2.88884342e-01 1.17512631e+00
-7.50832617e-01 8.46543670e-01 6.82427824e-01 -7.00760245e-01
-8.02370846e-01 -1.59364915e+00 -3.12045455e-01 4.46992099e-01
-2.53054559e-01 -7.77811855e-02 -3.20363313e-01 -9.40275252e-01
-5.19165359e-02 1.11642599e-01 -4.79649276e-01 -1.56444982e-01
-1.00356388e+00 -9.46667254e-01 9.04754281e-01 8.14370155e-01
6.89728320e-01 -1.04277861e+00 -1.14252830e+00 5.26958704e-01
-2.29004681e-01 -1.13853407e+00 3.94485798e-03 4.27542068e-02
-1.06662560e+00 -1.56994045e+00 -1.44736731e+00 -7.75905073e-01
1.21228285e-01 1.48074627e-01 1.40402877e+00 1.86886683e-01
-9.75649655e-01 2.84045696e-01 -6.03594817e-02 -5.08871257e-01
-5.27693510e-01 9.03958827e-02 -3.31327915e-01 -3.27170938e-01
3.75442415e-01 -5.88734686e-01 -1.18118298e+00 -2.33057551e-02
-2.69946516e-01 -4.62554067e-01 1.18437850e+00 8.06004107e-01
7.84381986e-01 -1.02979653e-01 6.10976577e-01 -1.15568674e+00
7.82021940e-01 -5.16839504e-01 -3.33745807e-01 1.43354148e-01
-6.30451798e-01 -4.95503813e-01 7.21780598e-01 -9.75550264e-02
-8.44184577e-01 -1.49001852e-01 -5.32760382e-01 -5.11203349e-01
-4.59872127e-01 3.81869555e-01 6.93170905e-01 -2.49063984e-01
9.10377860e-01 1.78250715e-01 1.39229484e-02 -3.65710407e-01
4.64046048e-03 2.69128472e-01 7.94511139e-01 -2.61199862e-01
6.39299333e-01 7.01996267e-01 6.05792344e-01 -7.53604054e-01
-9.12148654e-01 -5.17903090e-01 -6.25615835e-01 6.00737445e-02
1.16023850e+00 -6.53881967e-01 -6.23015285e-01 3.26623432e-02
-8.24163735e-01 5.26111461e-02 -5.09374499e-01 6.33180857e-01
-3.71100992e-01 9.07504737e-01 -7.30699062e-01 -1.73734605e-01
-9.64807332e-01 -1.10609853e+00 7.19668508e-01 3.92548293e-02
-2.24656254e-01 -9.95411336e-01 3.29723328e-01 1.41624495e-01
7.68732429e-01 2.70623803e-01 1.24912250e+00 -7.85019994e-01
-5.28948605e-01 -4.89558160e-01 -7.22091734e-01 2.07437739e-01
6.21710196e-02 -2.55188406e-01 -6.39509797e-01 -1.75211549e-01
-5.83091490e-02 -6.44995226e-03 8.99706304e-01 8.22879314e-01
1.53630829e+00 4.49541867e-01 -6.57067716e-01 8.22748840e-01
9.58441675e-01 3.21720660e-01 5.36992967e-01 1.46970615e-01
6.26077294e-01 -3.52718346e-02 5.14986217e-02 3.71263504e-01
-9.40667465e-02 1.83616415e-01 3.98841977e-01 -4.28731620e-01
-4.68786269e-01 1.38054388e-02 -6.41732574e-01 3.56077522e-01
-5.01545489e-01 -2.73261398e-01 -1.32525039e+00 5.00228524e-01
-1.31636894e+00 -8.49548519e-01 -4.04722691e-01 2.24344349e+00
2.54044712e-01 -2.09154729e-02 2.34938487e-01 -4.69766930e-02
4.97177243e-01 7.98198804e-02 -4.73593801e-01 8.43540952e-02
1.57398105e-01 9.99850690e-01 4.20561492e-01 5.52644692e-02
-1.35043156e+00 4.43206131e-01 7.48809910e+00 7.00776160e-01
-1.28922439e+00 5.39281726e-01 5.45607865e-01 9.98027399e-02
2.02114820e-01 -2.82448649e-01 -3.01414311e-01 4.01064813e-01
8.14001024e-01 -1.48372859e-01 -2.06445053e-01 6.06432080e-01
6.82389550e-03 -1.19859856e-02 -9.33597445e-01 1.12836969e+00
-2.61032432e-01 -1.75922227e+00 9.01997685e-02 -7.64580909e-03
1.35292351e-01 3.30129802e-01 1.99118063e-01 1.78845823e-01
8.57786164e-02 -1.35588610e+00 -2.91715890e-01 2.44015723e-01
1.13391697e+00 -5.28745055e-01 9.33578849e-01 1.56648159e-01
-9.82984185e-01 9.43481773e-02 -3.41532558e-01 2.79861689e-01
3.33285540e-01 6.28334284e-01 -8.26581717e-01 3.82023305e-01
8.34825635e-01 7.07110703e-01 -4.40053135e-01 1.49860287e+00
-5.94735146e-01 9.39765632e-01 -2.62873709e-01 1.56321093e-01
1.45182580e-01 6.36284277e-02 8.08821738e-01 9.75997150e-01
4.14732218e-01 3.72757286e-01 -3.18586007e-02 8.74129236e-01
-1.58535212e-01 2.54304051e-01 -3.33766252e-01 -1.32451564e-01
3.56268704e-01 1.34747362e+00 -8.51820409e-01 -6.37535691e-01
-3.83674324e-01 7.69015133e-01 3.53104472e-01 1.43100202e-01
-9.84578431e-01 -1.76288709e-01 -3.66756991e-02 3.37311447e-01
1.71120837e-01 8.60590190e-02 -3.44013780e-01 -1.00482833e+00
-3.21092755e-01 -5.34031212e-01 9.18056309e-01 -3.77575219e-01
-1.16989481e+00 5.38049936e-01 -5.25099516e-01 -1.24062121e+00
-1.13488309e-01 -7.94312358e-01 -8.45412433e-01 1.00950384e+00
-1.70215023e+00 -5.66568434e-01 -4.24004048e-01 3.28286022e-01
1.31527156e-01 -2.93546915e-01 1.00196195e+00 6.60654724e-01
-3.84227395e-01 4.92639422e-01 -5.29635809e-02 3.46971244e-01
8.65284562e-01 -1.44300997e+00 8.13638419e-02 6.05152726e-01
-2.21976459e-01 4.20549542e-01 1.55943006e-01 -5.69420338e-01
-8.67245018e-01 -9.45382774e-01 8.97084534e-01 -6.18326545e-01
5.83879173e-01 6.42196238e-01 -8.20736349e-01 7.21170723e-01
-9.42908078e-02 3.63420874e-01 9.29603040e-01 -2.33729631e-01
-1.15768373e-01 1.04127705e-01 -1.14719367e+00 4.46195513e-01
8.09746921e-01 -2.23432496e-01 -6.92793608e-01 4.91616845e-01
1.09203262e-02 -6.75057590e-01 -1.21743059e+00 7.74086714e-01
7.00878859e-01 -1.18054104e+00 1.44707751e+00 -8.52155864e-01
7.43022978e-01 1.03358746e-01 6.54787838e-01 -9.21266675e-01
-8.81234765e-01 -3.96175772e-01 -3.68415445e-01 1.94938600e-01
2.70028919e-01 -7.77007699e-01 1.24234664e+00 8.75690207e-02
-2.14941174e-01 -1.05366111e+00 -1.05634058e+00 -4.38076168e-01
6.36654973e-01 -1.53194830e-01 2.35285938e-01 6.06794000e-01
-4.23555791e-01 3.52771103e-01 -2.71472842e-01 1.67921409e-01
6.97450995e-01 1.79806069e-01 3.32471162e-01 -1.51503396e+00
-5.63678086e-01 -1.00107515e+00 -1.93787917e-01 -1.00453675e+00
-2.75881022e-01 -1.15365195e+00 -4.12470341e-01 -1.79589474e+00
5.01098394e-01 -3.68636668e-01 -5.66138029e-01 8.46171007e-02
-4.98114496e-01 3.02301615e-01 -1.71761900e-01 4.06537741e-01
-3.54636490e-01 1.06093585e-01 1.45152545e+00 1.78631514e-01
7.80740157e-02 3.21802586e-01 -6.30066514e-01 8.88307810e-01
8.13120842e-01 -6.87500834e-01 -1.41317323e-01 -2.56743312e-01
-1.37442932e-01 6.20046437e-01 5.52559614e-01 -1.25720012e+00
9.42660645e-02 3.76969248e-01 8.29715252e-01 -7.37050772e-01
1.09310606e-02 -7.99201906e-01 -1.91175789e-01 1.25286925e+00
-1.61072910e-01 -1.99594006e-01 1.68503627e-01 5.66873372e-01
-1.52262568e-01 -4.20145728e-02 1.23142397e+00 -4.88800883e-01
-4.08859044e-01 8.33924115e-01 -9.46938276e-01 4.39163625e-01
8.85270059e-01 -2.67473906e-02 -2.81521618e-01 -2.98867434e-01
-1.03068280e+00 1.54259056e-01 -2.66950071e-01 9.61093977e-02
6.55346394e-01 -9.42326307e-01 -8.40678573e-01 1.16384462e-01
-2.20616058e-01 -2.14969479e-02 4.68901634e-01 1.47853732e+00
-1.22934556e+00 7.24912524e-01 -2.18888432e-01 -8.04351985e-01
-1.16702890e+00 5.66480935e-01 8.73785138e-01 -9.53104138e-01
-1.22864020e+00 9.81928408e-01 3.39430481e-01 -7.00103641e-02
1.84940770e-01 3.93686024e-03 -3.43035042e-01 -3.58079374e-01
5.09800494e-01 6.18270636e-01 1.21275418e-01 -2.81060100e-01
-2.98081368e-01 5.42531252e-01 -1.27687022e-01 2.53002584e-01
8.76269877e-01 2.23928601e-01 -4.27114628e-02 -2.32246816e-01
7.34014809e-01 -9.17851403e-02 -5.50515175e-01 -3.24308753e-01
-1.25143185e-01 -3.44248325e-01 1.68802977e-01 -1.02380621e+00
-1.20836484e+00 1.57788050e+00 8.86547148e-01 -7.99972489e-02
9.13337886e-01 1.29500896e-01 1.19419670e+00 3.75744581e-01
1.27747813e-02 -3.92453134e-01 2.02081189e-01 2.48564318e-01
5.39164424e-01 -1.27979803e+00 8.72678161e-02 -6.25897586e-01
-2.71080941e-01 1.25045848e+00 4.26715672e-01 -2.66462505e-01
8.58667076e-01 2.97932595e-01 2.60984838e-01 -4.29960668e-01
-2.77089298e-01 -2.09763259e-01 2.42957145e-01 6.52519703e-01
5.51250637e-01 2.31735766e-01 -2.88348019e-01 7.77481496e-01
2.07009837e-02 1.01672970e-01 2.87954092e-01 7.89379120e-01
-7.33579218e-01 -8.92399013e-01 -1.44558460e-01 9.76606965e-01
-1.28904021e+00 -2.17678607e-01 -3.88573825e-01 8.85707557e-01
-1.05485909e-01 4.37838018e-01 -2.44844288e-01 3.80913407e-01
3.25164646e-01 -1.29392240e-02 6.23796582e-01 -4.94756550e-01
-6.61461592e-01 1.60009921e-01 -5.63214393e-03 -4.66391474e-01
-1.95538148e-01 -4.73797292e-01 -1.33636773e+00 -7.56768063e-02
1.31314009e-01 2.10814804e-01 -8.79666805e-02 9.76034760e-01
3.67810696e-01 1.14757764e+00 1.49011329e-01 -2.77043462e-01
-6.12735033e-01 -6.80586934e-01 -6.05505168e-01 2.34121233e-01
3.21394235e-01 -6.33959889e-01 -3.69079471e-01 -2.91288376e-01] | [15.17671012878418, -2.00669527053833] |
26fdd151-b8de-4567-b1e5-7041b73040a0 | learning-progressive-point-embeddings-for-3d | null | null | http://openaccess.thecvf.com//content/CVPR2021/html/Wen_Learning_Progressive_Point_Embeddings_for_3D_Point_Cloud_Generation_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Wen_Learning_Progressive_Point_Embeddings_for_3D_Point_Cloud_Generation_CVPR_2021_paper.pdf | Learning Progressive Point Embeddings for 3D Point Cloud Generation | Generative models for 3D point clouds are extremely important for scene/object reconstruction applications in autonomous driving and robotics. Despite recent success of deep learning-based representation learning, it remains a great challenge for deep neural networks to synthesize or reconstruct high-fidelity point clouds, because of the difficulties in 1) learning effective pointwise representations; and 2) generating realistic point clouds from complex distributions. In this paper, we devise a dual-generators framework for point cloud generation, which generalizes vanilla generative adversarial learning framework in a progressive manner. Specifically, the first generator aims to learn effective point embeddings in a breadth-first manner, while the second generator is used to refine the generated point cloud based on a depth-first point embedding to generate a robust and uniform point cloud. The proposed dual-generators framework thus is able to progressively learn effective point embeddings for accurate point cloud generation. Experimental results on a variety of object categories from the most popular point cloud generation dataset, ShapeNet, demonstrate the state-of-the-art performance of the proposed method for accurate point cloud generation. | ['DaCheng Tao', 'Baosheng Yu', 'Cheng Wen'] | 2021-06-19 | null | null | null | cvpr-2021-1 | ['point-cloud-generation', 'object-reconstruction'] | ['computer-vision', 'computer-vision'] | [ 3.02008409e-02 1.31393030e-01 2.75610209e-01 -7.70073608e-02
-8.90492439e-01 -5.11291564e-01 9.42793965e-01 5.01728393e-02
-1.02855898e-01 6.68598473e-01 -3.84454697e-01 -2.83502638e-01
4.32429463e-03 -1.28588808e+00 -1.18104780e+00 -6.08119369e-01
1.74125105e-01 1.00928366e+00 4.57009710e-02 -4.23365057e-01
3.03310752e-01 1.15275264e+00 -1.74677885e+00 -3.43184590e-01
1.06762075e+00 8.80115449e-01 3.28770965e-01 4.92490619e-01
-4.71838027e-01 4.71150607e-01 -5.18678367e-01 -4.92879927e-01
4.53858852e-01 -3.31090428e-02 -3.46021116e-01 7.33771771e-02
5.01610339e-01 -2.89944291e-01 -3.12698781e-01 9.30907786e-01
3.56140554e-01 1.93395287e-01 1.02378941e+00 -1.44273448e+00
-8.35837543e-01 7.94531703e-02 -3.57244939e-01 -4.07905251e-01
-2.68782601e-02 3.16612095e-01 6.97345614e-01 -1.14294231e+00
5.94543874e-01 1.33961308e+00 6.00004137e-01 6.54053569e-01
-1.09218740e+00 -8.13892901e-01 -9.68136936e-02 -1.16920471e-01
-1.42574704e+00 -1.19223677e-01 1.20707917e+00 -6.22178495e-01
7.25685537e-01 1.19953444e-02 8.49131048e-01 9.86838937e-01
2.73350209e-01 4.39942241e-01 7.78079867e-01 -1.22670583e-01
5.08255064e-01 3.03360503e-02 -5.53391755e-01 4.12443697e-01
2.92619586e-01 2.70437807e-01 -9.93915182e-03 -3.13218325e-01
1.24927032e+00 3.02963406e-01 -7.66418949e-02 -7.65768468e-01
-1.19844246e+00 1.09487641e+00 9.99435663e-01 6.49599656e-02
-5.44668853e-01 5.71430683e-01 -2.27344465e-02 -4.23795059e-02
5.02927542e-01 4.59530622e-01 5.76089807e-02 2.50271708e-02
-9.21727479e-01 9.01321769e-01 5.16942382e-01 1.36860108e+00
1.02406847e+00 4.86466229e-01 3.72072961e-03 6.82915390e-01
5.51397741e-01 7.99045444e-01 1.79667160e-01 -8.23693812e-01
5.07138968e-01 5.21874249e-01 3.46426785e-01 -9.40814734e-01
1.09899253e-01 -5.21408558e-01 -1.07816982e+00 7.91162431e-01
-6.25801906e-02 1.86593622e-01 -1.21314108e+00 1.42282283e+00
3.76272082e-01 6.68657541e-01 1.97913423e-01 7.26290703e-01
7.85307586e-01 8.11993361e-01 3.03837238e-03 3.23491633e-01
9.35701013e-01 -5.85929930e-01 -2.35962629e-01 -1.66554913e-01
5.87227233e-02 -5.25680006e-01 8.08227003e-01 -2.51302794e-02
-1.19002342e+00 -8.35804820e-01 -1.17126763e+00 -1.62978008e-01
-1.93342119e-01 -2.15283915e-01 3.83521885e-01 2.32248396e-01
-9.58672464e-01 6.34886146e-01 -1.01238954e+00 1.05525054e-01
8.76648366e-01 1.61233991e-01 -3.31704021e-01 -1.53992742e-01
-9.05176938e-01 8.47760379e-01 3.55530649e-01 6.18378408e-02
-1.15946531e+00 -1.05021071e+00 -1.08268952e+00 -2.78823264e-02
-2.53027081e-01 -1.19691277e+00 9.54455018e-01 -4.92899567e-01
-1.42183101e+00 8.21038961e-01 -4.31225263e-02 -6.13034308e-01
7.26907492e-01 -5.38999625e-02 8.77473280e-02 -3.39755602e-02
2.07358241e-01 1.08807886e+00 1.17652655e+00 -1.88351607e+00
-4.63878185e-01 -4.73623931e-01 -1.82109028e-02 2.59870082e-01
2.39851713e-01 -8.22004676e-01 -8.13966915e-02 -5.80827594e-01
2.24047586e-01 -9.85321879e-01 -5.41165471e-01 2.20814794e-01
-3.21464658e-01 -2.94032931e-01 9.14552748e-01 -3.04261595e-01
1.45578548e-01 -1.97011042e+00 2.62602121e-01 7.91547149e-02
3.45422983e-01 9.44697335e-02 -1.57881111e-01 4.31958228e-01
-1.82288885e-01 1.06301367e-01 -5.10561168e-01 -6.58541083e-01
5.59357442e-02 3.45626354e-01 -6.51164889e-01 4.10155207e-01
8.03616524e-01 1.12594092e+00 -1.06167471e+00 -1.14399880e-01
6.98462367e-01 8.84809911e-01 -6.35115027e-01 3.21261823e-01
-4.48686451e-01 7.33794451e-01 -5.27445078e-01 5.46047211e-01
1.00567329e+00 1.38444617e-01 -6.50370061e-01 2.84808930e-02
-4.48867492e-02 -2.26313304e-02 -9.00657117e-01 1.97096384e+00
-6.99387431e-01 4.47471350e-01 -3.14662278e-01 -7.69157767e-01
1.48100555e+00 1.36767074e-01 3.96230340e-01 -2.03609452e-01
3.86513434e-02 3.48288119e-01 -3.28770429e-01 -2.34017149e-02
6.76921368e-01 -5.62528968e-01 -1.13972023e-01 -1.14434287e-01
3.47958989e-02 -1.20264089e+00 -3.28048825e-01 2.75035370e-02
7.19012499e-01 2.90160060e-01 -1.10647708e-01 8.60936418e-02
4.61905390e-01 1.04167201e-01 2.71828115e-01 4.95633334e-01
2.36824870e-01 1.18092656e+00 2.08739623e-01 -4.42143708e-01
-1.53868365e+00 -1.42025363e+00 -1.33792818e-01 4.76401411e-02
2.22779050e-01 1.05549805e-01 -5.00897110e-01 -3.93838435e-01
3.10938060e-01 1.05177164e+00 -4.62619960e-01 -2.47231752e-01
-6.06753290e-01 -1.00666046e-01 3.87267262e-01 5.90358615e-01
3.87167722e-01 -1.11879790e+00 -4.50304002e-01 2.84289807e-01
1.07551441e-01 -9.70456541e-01 1.41267434e-01 -1.41846895e-01
-1.07269561e+00 -8.04202318e-01 -9.24697638e-01 -7.36039281e-01
8.44256461e-01 3.27554256e-01 1.15200973e+00 9.97930765e-02
-3.20241563e-02 1.60199091e-01 -2.44173482e-01 -7.28008151e-01
-6.76097751e-01 -6.56234249e-02 -8.48218575e-02 -1.68555170e-01
1.33203298e-01 -8.43127728e-01 -5.66543043e-01 -3.68176848e-02
-1.13926876e+00 -1.68667641e-02 5.78622401e-01 8.67900014e-01
9.49090421e-01 1.23067409e-01 4.63074058e-01 -6.19146943e-01
5.40006101e-01 -6.85789466e-01 -8.65084529e-01 -3.20767999e-01
-1.83221444e-01 7.83822536e-02 6.93291247e-01 -1.83603927e-01
-6.49838924e-01 1.40167654e-01 -7.86719918e-01 -1.20633590e+00
-2.45026514e-01 1.17729068e-01 -8.45462903e-02 -2.26595536e-01
7.20994651e-01 5.27948141e-01 9.81121883e-02 -2.73114413e-01
6.88209772e-01 3.01981002e-01 7.39769459e-01 -6.27787650e-01
1.57279956e+00 7.06434488e-01 2.45027110e-01 -8.54000807e-01
-1.54834867e-01 -3.05304796e-01 -7.30829418e-01 -1.16669416e-01
9.13377106e-01 -1.09510410e+00 -4.12842065e-01 4.53904271e-01
-1.35263920e+00 -7.28307515e-02 -6.71167612e-01 1.95893109e-01
-9.55769241e-01 1.83196083e-01 2.14617010e-02 -6.95612311e-01
-4.00797784e-01 -1.38744330e+00 1.58308649e+00 2.02539653e-01
1.48022205e-01 -7.82995641e-01 1.86329931e-01 4.60530482e-02
1.39498755e-01 9.13096428e-01 8.17117631e-01 -1.28248096e-01
-1.02144802e+00 -3.90721858e-01 -9.86431316e-02 4.39110756e-01
1.12285949e-01 1.20427586e-01 -9.37120080e-01 -3.38859916e-01
1.27327768e-02 -1.58971772e-01 5.88079035e-01 2.50177324e-01
1.21489787e+00 1.47380279e-02 -3.30092460e-01 9.11961555e-01
1.71963608e+00 2.71074846e-02 9.40150321e-01 1.93410769e-01
9.68872786e-01 2.67862678e-01 4.93629634e-01 3.24928850e-01
3.59670520e-01 5.77291012e-01 1.05407286e+00 4.65232171e-02
-7.48592168e-02 -7.97764897e-01 -1.86375096e-01 5.49244523e-01
-2.18287230e-01 -4.93342318e-02 -9.27835524e-01 8.19194853e-01
-1.60207486e+00 -8.46168280e-01 -2.04131782e-01 2.16547060e+00
3.95805597e-01 1.66609421e-01 -1.17630318e-01 2.80460984e-01
4.92789805e-01 1.19415998e-01 -6.10986829e-01 -1.60090581e-01
2.40933254e-01 6.52583838e-01 3.37713420e-01 3.58457744e-01
-7.24018216e-01 9.81771886e-01 5.14200830e+00 8.33352089e-01
-1.17036045e+00 6.69784173e-02 1.60800755e-01 2.49114916e-01
-7.39721894e-01 -1.85306370e-02 -6.76944196e-01 5.14713824e-01
5.33732474e-01 -2.43538633e-01 1.48842752e-01 1.21264458e+00
-2.69547720e-02 4.87739563e-01 -9.95201230e-01 1.32473540e+00
-9.80169550e-02 -1.71827567e+00 5.13126612e-01 1.52374133e-01
8.84003401e-01 1.90643281e-01 1.47179395e-01 2.68535525e-01
2.59851515e-01 -1.09886599e+00 1.07209516e+00 6.72983289e-01
8.82848740e-01 -1.10765851e+00 5.02819717e-01 6.97860420e-01
-1.07977080e+00 3.36657524e-01 -8.12967718e-01 1.87194020e-01
2.80258834e-01 6.19417191e-01 -1.06512880e+00 7.24435508e-01
5.12026846e-01 7.68420279e-01 -2.60759443e-01 1.13021851e+00
-3.54427099e-01 9.52244177e-02 -3.28201085e-01 7.85010085e-02
4.39226985e-01 -4.20220047e-01 9.48685467e-01 6.91556454e-01
6.72812283e-01 -8.93028975e-02 -1.76503450e-01 1.56106877e+00
-2.65562207e-01 -2.12844968e-01 -1.14331317e+00 6.27983138e-02
7.15000749e-01 1.20267224e+00 -4.37220871e-01 -1.04953319e-01
4.74344566e-02 6.94257975e-01 4.26673651e-01 1.90348491e-01
-8.14919829e-01 -3.74888361e-01 9.94254589e-01 3.20229590e-01
4.32858974e-01 -6.81927383e-01 -3.14915985e-01 -9.49749589e-01
-4.18421365e-02 -4.62818623e-01 -4.09144282e-01 -1.10163486e+00
-1.33307719e+00 8.46363902e-01 4.51662354e-02 -1.68301666e+00
-5.84419727e-01 -3.60700339e-01 -7.41422296e-01 1.36927724e+00
-1.76493347e+00 -1.48958766e+00 -7.41351604e-01 5.09885550e-01
6.90390348e-01 -1.62695676e-01 7.30662346e-01 5.74097596e-02
8.72814879e-02 2.42931992e-01 3.59032489e-02 -1.18647039e-01
7.08519965e-02 -1.33130634e+00 1.07990468e+00 6.93490982e-01
2.07246408e-01 5.12586832e-01 6.86226666e-01 -5.89821339e-01
-1.39045227e+00 -1.63612378e+00 2.53508091e-01 -6.73088729e-01
2.40047351e-01 -4.02530491e-01 -1.02317989e+00 5.77388108e-01
-1.55497879e-01 1.91957310e-01 4.20636982e-02 -5.33601165e-01
-7.40300789e-02 -3.33550498e-02 -1.39694524e+00 5.38759291e-01
8.77354205e-01 -3.13667774e-01 -7.01720655e-01 3.21374834e-01
8.07019293e-01 -7.24084675e-01 -8.39076877e-01 5.72574437e-01
1.84668198e-01 -8.26434195e-01 1.28378963e+00 -2.93459058e-01
7.73251891e-01 -5.25766075e-01 1.86902774e-03 -1.67357183e+00
-4.64088559e-01 -4.56019700e-01 -4.41026837e-02 9.89153266e-01
-1.71105396e-02 -7.14579463e-01 9.91823316e-01 9.94098410e-02
-4.93553728e-01 -8.39600146e-01 -9.79950249e-01 -7.16568530e-01
5.53704143e-01 -4.79120404e-01 1.16273165e+00 7.07350016e-01
-1.01949370e+00 4.19734716e-02 -5.99395782e-02 4.05579209e-01
9.50182855e-01 7.77792782e-02 1.28356469e+00 -1.24809682e+00
-7.38961473e-02 -4.29233283e-01 -8.89081478e-01 -1.23387611e+00
3.25892061e-01 -8.17518413e-01 2.33360589e-01 -1.82822359e+00
-4.58942741e-01 -1.08600163e+00 2.79592931e-01 -1.66017637e-02
-1.76806778e-01 3.54369938e-01 2.74252862e-01 3.60993862e-01
3.00797075e-01 1.03404832e+00 1.40765572e+00 -3.80046397e-01
-4.36563939e-02 2.23880038e-01 -6.70298159e-01 5.14831007e-01
5.52414715e-01 -4.06227380e-01 -6.23347163e-01 -6.46445870e-01
1.59142256e-01 -5.33283390e-02 7.95144856e-01 -1.24141181e+00
-7.35227764e-02 -1.32151425e-01 4.58685219e-01 -1.09886837e+00
7.45505452e-01 -8.45093489e-01 4.78150547e-01 3.65734339e-01
2.37881228e-01 4.61957557e-03 3.20706666e-01 7.15876579e-01
-3.06723684e-01 -1.95336908e-01 9.71063316e-01 -2.75196046e-01
-5.43656468e-01 7.95378804e-01 3.33451509e-01 -1.30158737e-01
1.12707913e+00 -3.63786876e-01 -2.41202451e-02 -3.14257413e-01
-4.68803376e-01 -1.28066645e-03 9.18411314e-01 6.93907976e-01
1.11839890e+00 -1.73990202e+00 -9.28284049e-01 4.96167511e-01
2.95083374e-01 9.78377581e-01 2.13723063e-01 5.60691133e-02
-1.11215997e+00 1.45848200e-01 -2.77958661e-01 -1.11247027e+00
-5.26754916e-01 4.78586912e-01 4.15679336e-01 8.69591385e-02
-9.01195228e-01 9.75989759e-01 3.94184530e-01 -5.89803159e-01
-2.65290201e-01 -5.42041659e-01 4.13530692e-02 -4.76129591e-01
1.96788579e-01 1.24943733e-01 2.38777995e-01 -8.30075622e-01
-5.94872981e-02 7.83449352e-01 3.27542692e-01 5.08533232e-03
1.39843345e+00 3.11199844e-01 5.63099198e-02 3.67918193e-01
1.07726109e+00 -1.64076269e-01 -1.41239154e+00 1.38243437e-01
-6.04308605e-01 -7.35454738e-01 4.34440076e-02 -5.47839664e-02
-1.00783598e+00 1.05840683e+00 3.89709592e-01 1.17008194e-01
5.84459662e-01 -9.07316208e-02 9.23243165e-01 5.27244620e-02
8.04111183e-01 -2.32706070e-01 -1.03244662e-01 3.81462157e-01
1.32032561e+00 -1.08686411e+00 -1.55741200e-01 -3.51553887e-01
-4.16623443e-01 9.57430065e-01 4.84414309e-01 -8.64473283e-01
5.57394207e-01 -1.37052229e-02 -2.08609715e-01 -2.95022815e-01
-4.78862166e-01 -3.57941315e-02 2.12833524e-01 9.55551565e-01
-6.82217255e-02 1.20775484e-01 2.28835180e-01 9.99591053e-02
-5.56812882e-01 -1.05255671e-01 3.08894932e-01 7.59956598e-01
-3.14161122e-01 -1.04593003e+00 -5.80221474e-01 3.09207857e-01
2.54985213e-01 1.30255967e-01 -9.70484391e-02 8.63978982e-01
2.27849379e-01 4.81200963e-01 3.88232738e-01 -3.24505776e-01
6.06709003e-01 -2.41857111e-01 5.94530284e-01 -7.70830214e-01
-1.54751793e-01 -3.67157340e-01 -5.32580614e-01 -2.58329183e-01
-1.08480260e-01 -5.13892055e-01 -1.22022986e+00 -2.52467990e-01
-1.48361102e-01 1.26614481e-01 1.08457375e+00 6.11268640e-01
4.71489310e-01 5.53275824e-01 8.17121744e-01 -1.61069071e+00
-6.19794726e-01 -9.09680426e-01 -4.94104892e-01 6.31101012e-01
3.79531384e-01 -1.01990008e+00 -3.59140873e-01 -1.36271507e-01] | [8.733473777770996, -3.6424381732940674] |
7f9989c0-1816-4d83-982e-58645900da04 | exploring-deep-models-for-practical-gait | 2303.03301 | null | https://arxiv.org/abs/2303.03301v2 | https://arxiv.org/pdf/2303.03301v2.pdf | Exploring Deep Models for Practical Gait Recognition | Gait recognition is a rapidly advancing vision technique for person identification from a distance. Prior studies predominantly employed relatively small and shallow neural networks to extract subtle gait features, achieving impressive successes in indoor settings. Nevertheless, experiments revealed that these existing methods mostly produce unsatisfactory results when applied to newly released in-the-wild gait datasets. This paper presents a unified perspective to explore how to construct deep models for state-of-the-art outdoor gait recognition, including the classical CNN-based and emerging Transformer-based architectures. Consequently, we emphasize the importance of suitable network capacity, explicit temporal modeling, and deep transformer structure for discriminative gait representation learning. Our proposed CNN-based DeepGaitV2 series and Transformer-based SwinGait series exhibit significant performance gains in outdoor scenarios, \textit{e.g.}, about +30\% rank-1 accuracy compared with many state-of-the-art methods on the challenging GREW dataset. This work is expected to further boost the research and application of gait recognition. Code will be available at https://github.com/ShiqiYu/OpenGait. | ['Shiqi Yu', 'Yongzhen Huang', 'Saihui Hou', 'Chao Fan'] | 2023-03-06 | null | null | null | null | ['gait-recognition', 'person-identification'] | ['computer-vision', 'computer-vision'] | [-1.73626587e-01 -8.94815445e-01 -1.51871189e-01 -3.06929141e-01
-5.66152215e-01 -2.35174432e-01 2.56154239e-01 -2.37414047e-01
-3.81734550e-01 6.36178792e-01 1.35884389e-01 9.99441221e-02
-5.99272549e-03 -7.54843593e-01 -2.25471467e-01 -7.34859467e-01
-4.39055800e-01 4.61168945e-01 8.71620979e-03 -4.89249438e-01
-5.04659340e-02 2.27213144e-01 -1.55820644e+00 -4.20102254e-02
5.40380657e-01 1.13447630e+00 -2.14162618e-01 7.36704588e-01
3.64897549e-01 8.87337506e-01 -5.69604218e-01 -6.54756367e-01
1.30441576e-01 -2.16967866e-01 -4.12557364e-01 -1.13621831e-01
6.27881348e-01 -4.75898236e-01 -1.10643005e+00 7.42778301e-01
1.00423932e+00 5.16107529e-02 4.28572655e-01 -1.38628888e+00
-7.62707531e-01 -6.55457005e-02 -4.05450553e-01 8.20761681e-01
6.09572589e-01 5.32183349e-01 8.53822470e-01 -7.47505903e-01
2.25271583e-01 7.51223564e-01 1.25685203e+00 4.48353827e-01
-7.29511082e-01 -6.53326511e-01 9.47854891e-02 8.57560694e-01
-1.44771814e+00 -4.22256678e-01 8.34393442e-01 -3.09261352e-01
1.36278975e+00 2.12887168e-01 1.01636863e+00 1.88431621e+00
2.51207262e-01 9.99639750e-01 8.08171630e-01 2.77262018e-03
-1.25884181e-02 -7.43909299e-01 2.51755267e-01 9.32677627e-01
5.81826210e-01 1.52662769e-01 -8.38263750e-01 1.15714170e-01
7.42564619e-01 1.06523104e-01 -9.31979567e-02 -1.41016185e-01
-1.06516135e+00 4.91437495e-01 6.34540915e-01 4.67019439e-01
-4.41033274e-01 5.90827584e-01 7.95267880e-01 3.52817148e-01
3.90524536e-01 -2.03574702e-01 -1.90655127e-01 -8.47385347e-01
-1.13631237e+00 4.81829166e-01 4.38868791e-01 5.86233318e-01
2.42507190e-01 5.61677098e-01 -2.45691419e-01 7.52057135e-01
1.65971309e-01 8.84490848e-01 4.96408015e-01 -5.66860139e-01
7.84028947e-01 2.32297093e-01 -2.83466876e-02 -1.17210042e+00
-5.16342461e-01 -7.16647089e-01 -1.03487647e+00 -3.11682701e-01
5.07939577e-01 -1.95790809e-02 -1.02668059e+00 1.51666415e+00
-1.66160703e-01 4.66896415e-01 -2.19207928e-01 9.89790857e-01
8.58783603e-01 2.02203229e-01 4.51777168e-02 3.01189393e-01
1.32423830e+00 -8.62058818e-01 -5.37101388e-01 -5.49364150e-01
7.15621352e-01 -4.35881674e-01 9.32092011e-01 3.25597703e-01
-7.07503259e-01 -8.34241569e-01 -1.20110059e+00 9.21465755e-02
-2.70285308e-01 5.24687469e-01 7.39324868e-01 1.08647871e+00
-1.23537886e+00 8.23365331e-01 -1.30325830e+00 -9.01179194e-01
5.92118800e-01 4.31149036e-01 -4.24143761e-01 -2.69520611e-01
-1.21018910e+00 7.35154510e-01 1.29882805e-02 6.85402989e-01
-7.86880612e-01 -1.51808530e-01 -9.49803889e-01 -1.60307303e-01
-8.39984119e-02 -1.00507462e+00 9.93406832e-01 -2.21369162e-01
-1.17243433e+00 8.96749496e-01 -1.64999977e-01 -7.90683210e-01
6.87038183e-01 -5.06366789e-01 -5.47113419e-01 7.26917163e-02
3.93153787e-01 1.24828435e-01 7.55425394e-01 -6.08866513e-01
-3.53408217e-01 -5.25898039e-01 -3.09540600e-01 -4.74343961e-03
-5.44812441e-01 -1.49894627e-02 -5.04161894e-01 -9.19693053e-01
-6.37955824e-03 -1.02321994e+00 -1.42901793e-01 -1.77562699e-01
-3.01720351e-01 -2.12547243e-01 7.67571867e-01 -1.00307631e+00
1.26149666e+00 -1.71588576e+00 1.43771358e-02 -8.65996704e-02
1.93771243e-01 4.65660840e-01 6.51188418e-02 5.33428192e-01
6.41624108e-02 -2.87107408e-01 -2.45060865e-02 -7.17058361e-01
2.63618261e-01 2.66865611e-01 1.04018301e-02 7.71727562e-01
1.11108422e-01 1.10256898e+00 -8.30315888e-01 -2.57185102e-01
6.01351440e-01 5.91846824e-01 -3.23549479e-01 -9.02503543e-03
5.12252390e-01 4.67068046e-01 -5.47074497e-01 1.20545053e+00
5.74701428e-01 -2.30523586e-01 9.54579636e-02 -2.46590272e-01
7.58642405e-02 1.62421092e-01 -7.91195035e-01 1.69072032e+00
-1.09866977e-01 8.86621058e-01 -3.68326038e-01 -1.43371427e+00
9.67281282e-01 2.28347912e-01 8.09126139e-01 -9.39140797e-01
3.80181044e-01 3.59686583e-01 -8.36859196e-02 -5.50900340e-01
4.83982354e-01 1.68407634e-01 -3.19817722e-01 8.47309232e-02
2.09149286e-01 5.43999493e-01 3.98323745e-01 -2.57508963e-01
1.31676292e+00 3.27631891e-01 8.87918621e-02 -6.96465299e-02
4.36577111e-01 -1.76959574e-01 5.54232836e-01 8.62095416e-01
-1.05448699e+00 7.17098117e-01 9.32808686e-03 -8.41213465e-01
-9.03618515e-01 -1.22160602e+00 2.06324190e-01 9.36218202e-01
2.10863426e-02 -6.52349055e-01 -6.04237080e-01 -3.91155273e-01
-3.92004289e-02 -8.87707472e-02 -7.17015505e-01 -3.21195692e-01
-1.10485673e+00 -1.06595051e+00 1.23498642e+00 1.23846602e+00
1.32665503e+00 -7.69146800e-01 -6.89657509e-01 2.87912875e-01
-7.30681717e-01 -1.28795433e+00 -4.55322474e-01 -1.37648415e-02
-8.36392641e-01 -1.04301524e+00 -1.03581369e+00 -8.87946248e-01
6.98906407e-02 2.55435675e-01 1.08541870e+00 1.31803036e-01
-3.48041624e-01 4.85984981e-01 -3.75789791e-01 4.78856713e-02
5.50113618e-01 1.81938633e-01 5.10857940e-01 -9.37428102e-02
7.73240864e-01 -7.80948579e-01 -8.60119462e-01 2.41059765e-01
2.73765152e-04 -4.07107919e-01 2.68647701e-01 1.01669288e+00
2.32178763e-01 -3.40388380e-02 2.54945457e-01 2.31429726e-01
5.00411689e-01 -2.23911643e-01 1.23083912e-01 2.79270876e-02
-2.65115589e-01 -1.16218053e-01 5.72149098e-01 -2.97403961e-01
-5.40751874e-01 -2.80926824e-01 -5.48764050e-01 -6.18774831e-01
-1.64691672e-01 5.31566024e-01 9.34547000e-03 -2.31858730e-01
5.57520211e-01 7.39932895e-01 -2.34803945e-01 -4.01862800e-01
-2.06787348e-01 3.22616577e-01 7.82890558e-01 -7.38740742e-01
8.07118118e-01 5.76053500e-01 1.74212561e-03 -8.92875016e-01
-4.69887733e-01 -6.01620376e-01 -7.46926367e-01 -4.81908441e-01
7.89967060e-01 -1.15653372e+00 -8.43646407e-01 1.14228785e+00
-6.14664674e-01 -4.18144912e-01 5.17426655e-02 3.03413302e-01
-6.45804644e-01 6.73380554e-01 -8.84278715e-01 -6.78890288e-01
-6.02060556e-01 -9.73146379e-01 1.19827282e+00 2.64760166e-01
-3.42185080e-01 -9.57187355e-01 1.79967597e-01 6.70253873e-01
5.40225089e-01 4.46220934e-01 6.44433647e-02 -3.29404175e-01
-3.02899390e-01 -4.58422631e-01 -3.44642140e-02 9.10095125e-03
8.86278376e-02 -1.57602876e-01 -1.05939460e+00 -6.08613133e-01
-3.39490891e-01 -3.48235220e-01 1.15220392e+00 5.81742465e-01
8.50763083e-01 -2.55371500e-02 -3.62176597e-01 9.42452431e-01
1.12862027e+00 1.51936308e-01 8.82072031e-01 7.62699425e-01
1.01297247e+00 3.41698602e-02 4.65636671e-01 4.83070046e-01
7.07683086e-01 9.58380163e-01 1.04448803e-01 1.10532425e-01
-2.30511516e-01 -1.94960475e-01 5.19150198e-01 8.63304079e-01
-7.85049617e-01 -2.59976566e-01 -1.15912497e+00 5.62346935e-01
-2.04109478e+00 -1.41832685e+00 -6.03108704e-02 1.96035814e+00
-3.14614289e-02 3.87347817e-01 6.11789346e-01 5.48554182e-01
5.49969673e-01 4.77560490e-01 -4.77792442e-01 -4.42945622e-02
-2.70464450e-01 5.24338484e-01 5.40956497e-01 -1.63940474e-01
-1.72294772e+00 8.62455308e-01 5.80348587e+00 6.81979716e-01
-1.18776345e+00 1.11220524e-01 4.01366144e-01 -7.75982961e-02
5.85104823e-01 -5.43221235e-01 -6.29743516e-01 6.15584195e-01
1.06859970e+00 -5.21415845e-03 1.03009613e-02 8.42335045e-01
3.16732377e-01 2.62885213e-01 -7.75722384e-01 1.58527064e+00
1.09731354e-01 -1.13419569e+00 -1.48467392e-01 2.22577780e-01
2.08615750e-01 1.49233758e-01 4.03804094e-01 5.04523158e-01
-1.68337286e-01 -1.04138005e+00 7.44640291e-01 3.53577167e-01
7.47099817e-01 -6.55091345e-01 1.00649965e+00 -2.16186017e-01
-2.15937448e+00 -2.33691156e-01 -2.07004428e-01 -4.56863523e-01
2.03318909e-01 1.04487196e-01 -2.96181232e-01 8.50775778e-01
1.18715584e+00 1.52985764e+00 -9.50350940e-01 1.15485764e+00
9.94221121e-03 7.73325861e-01 -2.64311463e-01 -7.24698603e-02
3.55120212e-01 -8.86769295e-02 4.42391068e-01 1.40704870e+00
6.13446653e-01 -2.61775881e-01 3.42953712e-01 3.36778462e-01
2.17408881e-01 -3.22818100e-01 -7.73461401e-01 -9.65893716e-02
2.10659042e-01 8.15015912e-01 -7.79188156e-01 -3.61744732e-01
-3.44720364e-01 1.33776450e+00 9.92689058e-02 2.70766467e-01
-1.11999047e+00 -1.74833536e-01 1.22860098e+00 5.89282699e-02
4.40166205e-01 -6.52159333e-01 -4.00421560e-01 -1.61941278e+00
3.73356819e-01 -7.01605082e-01 6.20510757e-01 -5.46667099e-01
-1.41274881e+00 5.97116351e-01 -2.20936090e-01 -1.70421553e+00
-4.10447568e-01 -9.46536958e-01 -8.28520775e-01 5.04894257e-01
-1.03047228e+00 -1.46429539e+00 -6.60106182e-01 8.51416111e-01
5.54531634e-01 -3.25346470e-01 7.95620084e-01 7.74598420e-01
-9.39286590e-01 1.07836223e+00 1.84414104e-01 8.12305689e-01
5.15970647e-01 -1.06090713e+00 1.13361180e+00 1.35498798e+00
6.36613220e-02 6.14482939e-01 7.83390343e-01 -6.13605082e-01
-1.57938159e+00 -1.04726326e+00 7.64779150e-01 -6.31337404e-01
7.28499174e-01 -3.05943877e-01 -5.47736943e-01 6.97278440e-01
-7.86856283e-03 9.16630030e-02 6.34317100e-01 6.94710612e-02
-3.46254677e-01 -1.98452339e-01 -1.02095711e+00 6.80730641e-01
1.71613383e+00 -6.06508613e-01 -4.47050571e-01 1.09157503e-01
-2.01156158e-02 -3.01801056e-01 -8.00375283e-01 4.99690205e-01
8.87569547e-01 -1.10559666e+00 1.51572168e+00 -4.40660775e-01
1.17715165e-01 -2.62795120e-01 -2.93009460e-01 -1.05980599e+00
-6.20743573e-01 -3.29381704e-01 -5.83245158e-01 7.32405603e-01
-3.18749815e-01 -6.88051403e-01 1.25754011e+00 1.57144025e-01
-2.38680705e-01 -6.28997684e-01 -1.25990546e+00 -1.12548971e+00
-8.20665658e-02 -6.13920391e-01 2.69313127e-01 7.43817866e-01
-2.08412886e-01 1.82011053e-02 -8.64680588e-01 6.53753802e-02
1.02715194e+00 -1.62456930e-01 8.29410136e-01 -1.18897855e+00
-1.10814385e-01 -4.90742475e-01 -1.34017670e+00 -9.70216870e-01
-1.17919333e-01 -5.93905270e-01 -2.29404271e-01 -1.67698586e+00
-5.57936579e-02 -2.71274149e-03 -4.49249983e-01 5.23897350e-01
-1.33376151e-01 8.56245637e-01 2.73993641e-01 1.85786769e-01
-6.51309311e-01 8.46095920e-01 9.63652492e-01 -5.21745086e-01
1.71184897e-01 -3.09358984e-02 -4.33960170e-01 4.54273254e-01
1.02012384e+00 -2.38285121e-03 -2.06244335e-01 -3.38894635e-01
-3.42028916e-01 -3.20300788e-01 8.19161236e-01 -1.82006609e+00
2.47820571e-01 2.25015819e-01 8.96909773e-01 -7.01841831e-01
7.71276236e-01 -2.54148155e-01 1.55111507e-01 8.93840134e-01
3.24149430e-01 4.83210236e-01 1.70503780e-01 4.69172984e-01
-2.34283969e-01 5.79429448e-01 3.69992495e-01 -1.51856825e-01
-1.25169921e+00 6.11062884e-01 -5.46920359e-01 2.56490130e-02
6.21699631e-01 -8.40562701e-01 -4.18855220e-01 -4.07121599e-01
-8.39825094e-01 3.26559283e-02 2.01739222e-01 5.71866393e-01
9.26492512e-01 -1.67882586e+00 -6.45985961e-01 2.04527333e-01
2.79856771e-01 -7.26524413e-01 5.38437724e-01 9.55948949e-01
-6.78699553e-01 6.24531507e-01 -6.58473670e-01 -6.94360197e-01
-1.23719120e+00 1.18088774e-01 6.72189534e-01 -5.46227038e-01
-9.03930068e-01 8.41089010e-01 -2.88884163e-01 -2.49566078e-01
1.13057427e-01 -2.54428536e-01 -1.09415144e-01 -1.84763246e-03
4.50072497e-01 6.14547253e-01 3.22889268e-01 -8.94871891e-01
-9.37389135e-01 7.55592525e-01 6.95341527e-02 1.83193564e-01
1.20630753e+00 -1.26173673e-02 4.29681718e-01 1.89137563e-01
1.13398957e+00 -7.11824238e-01 -1.26867223e+00 9.69001800e-02
-7.27658253e-03 -4.40133721e-01 -2.80065656e-01 -3.66567492e-01
-1.12749243e+00 9.33321714e-01 1.11022842e+00 -1.78988695e-01
1.16795897e+00 -4.90491688e-01 1.17837512e+00 5.55328846e-01
9.29732323e-01 -1.01637733e+00 4.48453456e-01 7.25794554e-01
5.84074497e-01 -1.19687462e+00 -2.12566495e-01 -8.60899091e-02
-3.93271804e-01 1.16864729e+00 6.83953404e-01 -4.26926225e-01
5.55714726e-01 9.26036015e-02 -6.87292814e-02 -2.42118344e-01
-3.67366076e-01 -5.27518392e-01 1.52718246e-01 1.01345563e+00
5.48185289e-01 1.72714472e-01 5.18329777e-02 5.75661361e-01
-5.28336406e-01 2.18207866e-01 9.38406810e-02 1.17786026e+00
-1.52811617e-01 -8.92038047e-01 -3.77950639e-01 5.68935752e-01
-3.28618705e-01 -5.55795431e-02 -7.93793350e-02 7.59854317e-01
2.48993281e-02 9.69642341e-01 -1.39169365e-01 -9.33512449e-01
3.93559903e-01 2.88767461e-02 7.79776394e-01 1.97194647e-02
-5.02880692e-01 -3.06779593e-01 4.33380514e-01 -7.90466785e-01
-4.85167801e-01 -8.74518037e-01 -6.34953976e-01 -8.36814463e-01
3.71739388e-01 -2.14856356e-01 -5.24196140e-02 9.73485172e-01
1.84426770e-01 5.64852834e-01 1.95552632e-01 -1.22863185e+00
-2.74858683e-01 -9.88732278e-01 -4.89633560e-01 2.92836428e-01
2.61587977e-01 -1.08270121e+00 -3.11773876e-03 3.28373946e-02] | [14.29710578918457, 1.4279143810272217] |
e9454fa3-3a4b-401c-9645-0bfead3d1d84 | visualizing-relation-between-de-motivating | 2306.12118 | null | https://arxiv.org/abs/2306.12118v2 | https://arxiv.org/pdf/2306.12118v2.pdf | Visualizing Relation Between (De)Motivating Topics and Public Stance toward COVID-19 Vaccine | While social media plays a vital role in communication nowadays, misinformation and trolls can easily take over the conversation and steer public opinion on these platforms. We saw the effect of misinformation during the COVID-19 pandemic when public health officials faced significant push-back while trying to motivate the public to vaccinate. To tackle the current and any future threats in emergencies and motivate the public towards a common goal, it is essential to understand how public motivation shifts and which topics resonate among the general population. In this study, we proposed an interactive visualization tool to inspect and analyze the topics that resonated among Twitter-sphere during the COVID-19 pandemic and understand the key factors that shifted public stance for vaccination. This tool can easily be generalized for any scenario for visual analysis and to increase the transparency of social media data for researchers and the general population alike. | ['Hamed Alhoori', 'Ashiqur Rahman'] | 2023-06-21 | null | null | null | null | ['misinformation'] | ['miscellaneous'] | [ 2.90199444e-02 3.52019936e-01 -1.18897790e-02 1.00233763e-01
1.59360049e-03 -7.15333283e-01 5.76710343e-01 1.00918794e+00
-5.22105217e-01 8.01404417e-01 7.43798971e-01 -8.53844225e-01
1.56130612e-01 -9.36953604e-01 -3.29913467e-01 -7.87899077e-01
-2.33516678e-01 3.06390822e-01 4.63452265e-02 -7.79138088e-01
2.89279372e-01 3.08480084e-01 -1.17321110e+00 2.75243223e-01
5.92885852e-01 -6.28982186e-02 3.23293746e-01 5.81262350e-01
-4.31144625e-01 6.73613131e-01 -1.13277042e+00 1.81364596e-01
-2.86593944e-01 -1.99328825e-01 -5.91650248e-01 -3.92767191e-01
-3.26705843e-01 -9.39370915e-02 2.69012123e-01 8.16431463e-01
6.01837754e-01 -3.79199922e-01 3.07605356e-01 -1.10174847e+00
5.98220080e-02 6.25446498e-01 -6.46468878e-01 6.93043530e-01
7.24963307e-01 1.22008853e-01 3.10590547e-02 -2.93877423e-01
1.01975155e+00 1.52010512e+00 7.08916664e-01 3.18020344e-01
-9.49484706e-01 -7.05127656e-01 1.60282746e-01 -2.01055557e-01
-9.40485179e-01 2.44358271e-01 5.79476476e-01 -1.05164337e+00
2.62738824e-01 7.35938489e-01 9.00045395e-01 1.13605297e+00
3.35715324e-01 6.69941530e-02 1.27001119e+00 3.78132500e-02
-1.63632184e-01 4.01624709e-01 1.68362662e-01 2.05524787e-01
4.49953049e-01 -3.26579183e-01 -4.21197891e-01 -7.73632646e-01
2.01882899e-01 4.33782309e-01 -5.89598119e-01 5.90094388e-01
-1.31273198e+00 1.11862314e+00 3.65100086e-01 6.84926450e-01
-7.08353817e-01 -3.68040115e-01 4.02112126e-01 2.44004726e-01
7.66132712e-01 5.36723614e-01 -3.46912563e-01 -4.56537157e-01
-4.68268156e-01 3.25216681e-01 6.94145381e-01 -1.17449351e-01
6.54626369e-01 -4.62118596e-01 -6.79678023e-02 1.43177122e-01
3.20927382e-01 1.01593077e+00 -1.16514303e-01 -1.49266511e-01
2.11834982e-01 8.70117188e-01 4.04313445e-01 -1.76654053e+00
-9.44324195e-01 -4.06071663e-01 -4.70631272e-01 -3.83561328e-02
5.98516226e-01 -9.23104763e-01 -5.09799838e-01 1.55881917e+00
8.29891980e-01 -3.14516425e-01 -3.13645154e-01 7.83498168e-01
1.13328815e+00 7.89600909e-01 3.14837843e-01 -2.39657760e-01
1.82040954e+00 1.91791341e-01 -1.22973669e+00 -8.48186202e-03
8.84293377e-01 -9.41832721e-01 7.72079408e-01 -2.58139640e-01
-6.16102099e-01 1.86853707e-02 -7.19820499e-01 5.43891311e-01
-8.54121208e-01 -7.15078950e-01 -3.51887904e-02 8.53169441e-01
-8.66952419e-01 3.28372031e-01 -5.65722525e-01 -8.43853056e-01
3.41862500e-01 4.02899384e-02 2.63007209e-02 3.55273783e-01
-1.47974300e+00 9.51667845e-01 2.87810564e-01 -9.65613052e-02
-3.35863322e-01 -7.82173514e-01 -3.90322328e-01 -7.39965260e-01
3.59813333e-01 -4.82517093e-01 7.37801075e-01 -3.83707315e-01
-8.15950632e-01 9.08996224e-01 8.60086232e-02 -1.31575406e-01
5.04420698e-01 -7.60184303e-02 -4.48355734e-01 8.75168666e-02
1.48546681e-01 3.65962803e-01 2.15270653e-01 -1.21164179e+00
-5.77820957e-01 -4.92582262e-01 2.05553174e-01 -1.27465546e-01
-1.31256178e-01 5.16092122e-01 3.97596598e-01 -5.31478882e-01
-4.31744188e-01 -8.96695793e-01 -2.06499487e-01 -6.30014062e-01
-4.48348910e-01 -5.75836338e-02 1.61193502e+00 -8.19354177e-01
1.38391936e+00 -1.95810187e+00 -3.61104280e-01 8.05764124e-02
4.57546562e-01 3.68379086e-01 3.58386815e-01 1.25168598e+00
3.22195709e-01 4.70796973e-01 1.78755462e-01 3.75445843e-01
-2.69799948e-01 8.98421407e-02 -2.67137051e-01 7.19733894e-01
-1.71117857e-01 5.17514467e-01 -1.32101095e+00 -1.46238238e-01
-1.35720208e-01 8.49540830e-01 -4.82879460e-01 2.45112717e-01
-1.47609577e-01 1.32347715e+00 -7.26223886e-01 8.88023078e-02
9.39977705e-01 -4.77440804e-01 3.64312083e-01 1.62067279e-01
-8.21755111e-01 5.24443984e-01 -5.35951555e-01 6.05334938e-01
-2.51481384e-02 8.60545754e-01 4.37094957e-01 -4.16816115e-01
8.30887616e-01 4.29098219e-01 5.11202931e-01 -3.46440524e-01
5.76852620e-01 -2.69656062e-01 -3.35884467e-02 -7.82557666e-01
6.24914289e-01 -1.43391415e-01 -1.49699017e-01 1.07686639e+00
-1.00964975e+00 2.68267065e-01 -1.33788317e-01 3.39816213e-01
4.84083325e-01 -4.34063524e-01 2.69560158e-01 -6.59140527e-01
2.81037927e-01 3.28325927e-01 1.07041150e-01 4.32297617e-01
-4.38198261e-02 -3.16300616e-02 6.87208235e-01 -8.35635304e-01
-3.71530116e-01 -5.02938569e-01 -1.27440859e-02 1.15086305e+00
1.23506851e-01 -3.77847910e-01 -8.16754222e-01 -4.60917294e-01
-2.86762416e-01 6.75825179e-01 -8.15028846e-01 4.47148681e-01
-4.57743227e-01 -1.03997242e+00 1.33043572e-01 -4.92588311e-01
3.73669326e-01 -9.27942216e-01 -1.26180148e+00 2.60535330e-01
-7.27940321e-01 -7.58598685e-01 -2.45153204e-01 -3.75780195e-01
-3.78020555e-01 -1.31465995e+00 -6.42680109e-01 -4.05551344e-01
1.05747652e+00 5.72756052e-01 6.73962116e-01 2.68591821e-01
-1.32417902e-01 2.56509870e-01 -2.76172668e-01 -1.15012121e+00
-8.40420365e-01 3.79163586e-03 -1.37853086e-01 -1.89466611e-01
2.00494349e-01 -2.94180214e-01 -9.24291372e-01 4.18278456e-01
-1.09846711e+00 3.03804159e-01 -4.81662929e-01 -3.94616313e-02
-4.44862574e-01 -2.35005081e-01 6.23861790e-01 -1.11059546e+00
8.80167782e-01 -1.10422540e+00 -2.69782722e-01 -2.12737873e-01
-1.02775693e-01 -4.74403709e-01 1.11382827e-01 -1.48042291e-01
-7.44068921e-01 -8.23404133e-01 -9.23274234e-02 7.55940795e-01
-7.21766278e-02 6.92233205e-01 5.11398375e-01 3.01645339e-01
6.06328547e-01 -1.84773922e-01 2.32414708e-01 -5.19648969e-01
4.40723598e-02 9.50333595e-01 -2.71872461e-01 1.97300151e-01
7.45707393e-01 9.51751351e-01 -4.31707710e-01 -1.26351464e+00
-6.44510925e-01 -4.48149323e-01 3.56498696e-02 -7.94422865e-01
1.14837611e+00 -7.40675271e-01 -1.24486959e+00 4.75263059e-01
-1.49634457e+00 -2.35007748e-01 3.31184685e-01 2.37038210e-01
5.14590800e-01 5.54288104e-02 -3.37906778e-01 -8.15569043e-01
-3.24151844e-01 -8.13186407e-01 4.38565165e-01 3.51657629e-01
-6.30225837e-01 -1.50097072e+00 5.27635396e-01 2.95228839e-01
1.26202631e+00 8.83335888e-01 6.45088673e-01 -8.08606684e-01
-2.04524383e-01 7.35412017e-02 -2.58898854e-01 -6.34644508e-01
6.23448491e-01 -3.40369232e-02 -8.02700818e-01 -3.58032197e-01
1.26676053e-01 5.37304431e-02 -1.39775733e-02 1.97496787e-01
3.51078689e-01 -9.76712704e-01 -5.75161755e-01 -1.78197637e-01
8.87943864e-01 3.90480608e-01 4.11446124e-01 5.85047245e-01
3.25014979e-01 1.29050994e+00 2.88948447e-01 8.14820170e-01
6.85784340e-01 5.30184805e-01 5.61846256e-01 -5.77367604e-01
4.04798537e-01 -2.47745559e-01 2.84902006e-01 8.23274136e-01
-2.07674265e-01 -1.51589960e-01 -1.21725786e+00 2.73654372e-01
-1.46475339e+00 -9.04045761e-01 -5.55606782e-01 2.00789595e+00
7.33504832e-01 -1.08138204e-01 4.90763992e-01 -1.17725246e-01
8.78328979e-01 5.25526404e-01 2.69251019e-01 -3.32680225e-01
5.06889410e-02 -5.44656217e-01 3.66967231e-01 8.31172466e-01
-6.99231267e-01 2.84748554e-01 6.48462915e+00 2.19321221e-01
-1.79331672e+00 4.30627793e-01 6.06986463e-01 3.33819091e-01
-6.63266122e-01 -1.38133779e-01 -7.77251601e-01 4.15790081e-01
8.90476763e-01 -2.76868284e-01 1.47165684e-02 1.28469378e-01
1.03420949e+00 -3.33792120e-01 -1.40695184e-01 4.99186337e-01
-1.97916716e-01 -1.65264130e+00 -3.58340085e-01 1.86709017e-01
5.99403679e-01 1.99915320e-01 -1.62527367e-01 -2.71030456e-01
5.70589304e-02 -6.95582509e-01 1.26709759e-01 4.14221495e-01
4.20741558e-01 -3.87151808e-01 7.01773584e-01 5.72238863e-01
-7.20591784e-01 3.80109727e-01 2.34906614e-01 -4.96504039e-01
6.40661597e-01 5.78380108e-01 -1.59511340e+00 3.26786526e-02
3.56666565e-01 2.86013097e-01 -3.72766644e-01 6.04942262e-01
-5.30889519e-02 5.26092947e-01 -6.25450723e-03 -4.38767761e-01
3.64174783e-01 -3.25451745e-03 9.74280179e-01 1.49553537e+00
8.09783116e-02 -7.50613213e-02 -3.29137295e-02 4.02285099e-01
5.62581718e-01 3.12030584e-01 -6.73579574e-01 -5.75652838e-01
9.60892588e-02 1.02875495e+00 -1.10920548e+00 -2.36855850e-01
-2.86519974e-02 2.20900729e-01 -3.21760267e-01 3.89642864e-01
-8.07907701e-01 -5.77580743e-02 5.99338233e-01 8.84075165e-01
-1.63180262e-01 -1.66425645e-01 1.11950882e-01 -5.10316312e-01
-6.20350301e-01 -1.02105963e+00 2.34262392e-01 -2.10034713e-01
-6.17169201e-01 7.17100441e-01 2.57827282e-01 -7.53008127e-01
1.35840103e-01 -8.91415700e-02 -8.60703290e-01 7.24821568e-01
-1.27970481e+00 -7.23994493e-01 -1.99777216e-01 2.17106104e-01
7.01387003e-02 4.62689191e-01 8.10401618e-01 2.02601135e-01
-4.02653754e-01 -1.63702488e-01 -2.14889914e-01 -1.16497688e-01
3.42076808e-01 -5.96827686e-01 3.03882033e-01 3.22041720e-01
-7.70101428e-01 8.07295978e-01 1.46957052e+00 -1.04107237e+00
-1.23029447e+00 -8.09813619e-01 1.38016248e+00 -2.80426949e-01
9.52749729e-01 -5.33396959e-01 -7.50071347e-01 3.46309513e-01
8.32121253e-01 -7.48548806e-01 9.55078483e-01 -7.46009052e-02
8.74935240e-02 4.22714055e-01 -1.07552338e+00 9.61580753e-01
4.81119841e-01 -3.14220548e-01 -4.92198706e-01 7.94832885e-01
7.25223064e-01 -2.68326521e-01 -5.21802664e-01 3.70615758e-02
3.81333381e-01 -1.06816673e+00 5.87822020e-01 -6.00500882e-01
-1.11590467e-01 -3.05705816e-01 3.27022344e-01 -1.34071183e+00
1.45477399e-01 -1.28217947e+00 8.24960053e-01 7.20804751e-01
4.98999447e-01 -1.25975537e+00 4.00464416e-01 2.74867386e-01
3.22838515e-01 -4.52670872e-01 -6.01305544e-01 1.68195684e-02
-3.01452547e-01 -1.96231648e-01 6.26822174e-01 1.13136387e+00
4.82317954e-01 8.56637508e-02 -5.20956099e-01 2.55705893e-01
1.70985714e-01 -1.58238292e-01 1.06657267e+00 -1.10728276e+00
9.89732239e-03 -1.33230880e-01 -4.15479913e-02 -4.31089908e-01
-9.26080942e-01 -4.63529885e-01 -4.72207934e-01 -1.56261492e+00
1.54515326e-01 -3.60909581e-01 1.98188692e-01 1.47489995e-01
-1.06333047e-01 1.26510948e-01 3.79571289e-01 4.64060307e-02
-2.72899449e-01 -9.70624387e-02 1.58652496e+00 7.94954151e-02
-5.73094010e-01 -1.40638009e-01 -9.92868245e-01 6.35141969e-01
8.99908662e-01 -8.91253412e-01 -2.05088824e-01 1.71819448e-01
1.13567126e+00 1.32421821e-01 1.70574591e-01 -4.52019304e-01
2.00426474e-01 -3.22483033e-01 -1.47697642e-01 -8.66738915e-01
-8.31700638e-02 -6.60575390e-01 7.79190898e-01 1.37913609e+00
5.23342267e-02 4.98055577e-01 4.81138080e-01 3.23840320e-01
1.80688333e-02 3.02915961e-01 4.57237810e-01 1.03248373e-01
3.21101487e-01 -1.31518647e-01 -1.17529619e+00 4.33195800e-01
1.07055306e+00 5.46638668e-03 -1.16163027e+00 -7.29406238e-01
-6.41963243e-01 2.46393412e-01 4.38091427e-01 5.81221521e-01
3.36544812e-01 -7.20219433e-01 -1.06406629e+00 7.22639114e-02
-2.33241245e-01 -4.41034764e-01 5.37339807e-01 1.32008445e+00
-8.25754166e-01 4.41402465e-01 -3.00180882e-01 -5.76254845e-01
-1.50132060e+00 2.82842189e-01 6.93370774e-02 -1.62391543e-01
-4.41559047e-01 4.62018728e-01 3.57940793e-01 -3.56907070e-01
1.16270393e-01 -7.69173503e-02 -7.64671266e-01 6.95353687e-01
1.18741882e+00 8.13173115e-01 -3.31595749e-01 -8.47628295e-01
-7.15731144e-01 3.24450612e-01 1.99523911e-01 1.08971380e-01
1.28012538e+00 -3.89513314e-01 -4.59495962e-01 6.27663136e-01
1.13853526e+00 6.39340401e-01 -6.42965674e-01 1.29487813e-01
-1.49758086e-01 -5.66909432e-01 -3.57318997e-01 -7.45301545e-01
-7.21417725e-01 6.46919072e-01 3.47959995e-01 1.14067245e+00
5.31938374e-01 1.32914275e-01 6.94611430e-01 -1.44805893e-01
4.56541106e-02 -6.02048397e-01 -1.41410023e-01 3.07434589e-01
1.07456326e+00 -9.96228933e-01 -2.74242051e-02 -4.03888136e-01
-6.12845004e-01 1.10008740e+00 -1.36115566e-01 4.99138087e-01
1.03666520e+00 4.49008375e-01 7.82674432e-01 -9.08988416e-01
-5.66212237e-01 1.19919688e-01 -1.51902348e-01 6.55053794e-01
6.18056297e-01 3.63389254e-01 -8.67553055e-01 -2.98459325e-02
-5.05395494e-02 -1.21018179e-01 7.43531048e-01 9.31186974e-01
-8.18985939e-01 -1.11097646e+00 -1.01104426e+00 3.32461670e-02
-8.98089528e-01 2.66866326e-01 -5.10252178e-01 8.00290763e-01
4.29665089e-01 1.33702457e+00 4.41429988e-02 -1.97579309e-01
7.28780851e-02 -2.25943714e-01 -3.11027646e-01 -3.24292749e-01
-9.52815056e-01 3.18272486e-02 5.75837731e-01 -1.58023238e-01
-5.84085226e-01 -6.25265002e-01 -1.30503082e+00 -6.97380185e-01
-4.79029976e-02 6.17980182e-01 1.02543819e+00 8.26356173e-01
4.30543154e-01 3.27489853e-01 6.55043304e-01 -5.99076092e-01
2.81765401e-01 -1.01154423e+00 -1.82891846e-01 2.92422593e-01
5.64625561e-01 -2.99969763e-01 -6.05970144e-01 -2.94420093e-01] | [8.491903305053711, 9.75222110748291] |
522b5360-150b-4914-8b8f-c79c793dec7b | mind-the-gap-polishing-pseudo-labels-for | 2207.08185 | null | https://arxiv.org/abs/2207.08185v2 | https://arxiv.org/pdf/2207.08185v2.pdf | Mind the Gap: Polishing Pseudo labels for Accurate Semi-supervised Object Detection | Exploiting pseudo labels (e.g., categories and bounding boxes) of unannotated objects produced by a teacher detector have underpinned much of recent progress in semi-supervised object detection (SSOD). However, due to the limited generalization capacity of the teacher detector caused by the scarce annotations, the produced pseudo labels often deviate from ground truth, especially those with relatively low classification confidences, thus limiting the generalization performance of SSOD. To mitigate this problem, we propose a dual pseudo-label polishing framework for SSOD. Instead of directly exploiting the pseudo labels produced by the teacher detector, we take the first attempt at reducing their deviation from ground truth using dual polishing learning, where two differently structured polishing networks are elaborately developed and trained using synthesized paired pseudo labels and the corresponding ground truth for categories and bounding boxes on the given annotated objects, respectively. By doing this, both polishing networks can infer more accurate pseudo labels for unannotated objects through sufficiently exploiting their context knowledge based on the initially produced pseudo labels, and thus improve the generalization performance of SSOD. Moreover, such a scheme can be seamlessly plugged into the existing SSOD framework for joint end-to-end learning. In addition, we propose to disentangle the polished pseudo categories and bounding boxes of unannotated objects for separate category classification and bounding box regression in SSOD, which enables introducing more unannotated objects during model training and thus further improve the performance. Experiments on both PASCAL VOC and MS COCO benchmarks demonstrate the superiority of the proposed method over existing state-of-the-art baselines. | ['Wei Wei', 'Yuxuan Sun', 'Lei Zhang'] | 2022-07-17 | null | null | null | null | ['semi-supervised-object-detection'] | ['computer-vision'] | [ 3.24411869e-01 4.81349438e-01 -3.04219306e-01 -4.51106280e-01
-7.33155131e-01 -7.99246311e-01 5.92536569e-01 1.45133123e-01
-5.10186493e-01 5.72825730e-01 -1.54864684e-01 9.36326906e-02
9.93663371e-02 -6.91869259e-01 -9.32324708e-01 -9.65815961e-01
3.30412865e-01 5.68109989e-01 6.30858183e-01 2.94833213e-01
-2.61330634e-01 2.95567274e-01 -1.71149457e+00 3.85671765e-01
9.26351607e-01 1.28202331e+00 3.80307674e-01 2.02347353e-01
-1.65573493e-01 5.23882806e-01 -6.09784901e-01 -6.01398408e-01
3.97873491e-01 -2.64470540e-02 -5.59081137e-01 3.97077858e-01
9.19612944e-01 -2.74741083e-01 -2.34740481e-01 1.28467739e+00
3.50240842e-02 -2.04220135e-02 7.47974336e-01 -1.24225950e+00
-5.69445491e-01 8.22390556e-01 -6.07836425e-01 -2.86925673e-01
-2.68577516e-01 2.13243470e-01 1.38395596e+00 -1.06694949e+00
3.36917937e-01 1.33731568e+00 5.75782597e-01 5.96634507e-01
-1.80488300e+00 -1.03665960e+00 4.58183438e-01 -5.13071492e-02
-1.49728954e+00 -2.28280112e-01 7.80782163e-01 -5.11452556e-01
4.03574467e-01 1.17883056e-01 3.02424401e-01 1.09968948e+00
-4.99665886e-01 1.14234483e+00 9.90900993e-01 -3.36495340e-01
3.65029648e-02 5.31345546e-01 3.53240609e-01 7.96211004e-01
4.71349388e-01 1.00450777e-02 -3.05988908e-01 -2.83537079e-02
2.99971759e-01 9.96578038e-02 -2.85213500e-01 -1.03599763e+00
-1.30949080e+00 5.74600816e-01 7.72134066e-01 -1.16935752e-01
-7.98502043e-02 -4.58267368e-02 3.15350115e-01 -2.33837858e-01
4.86918926e-01 5.07176757e-01 -7.10944891e-01 4.12026107e-01
-9.96955454e-01 2.47526556e-01 4.90523219e-01 1.15959561e+00
9.52990294e-01 -2.53973186e-01 -5.74668467e-01 9.49818790e-01
4.15500611e-01 5.32402575e-01 1.96206495e-01 -5.78595698e-01
6.24651611e-01 9.37299669e-01 2.83664286e-01 -6.38388455e-01
-2.02255294e-01 -4.95281488e-01 -4.84922469e-01 2.07815468e-01
8.12953591e-01 4.67927493e-02 -1.33195817e+00 1.94583023e+00
6.46030843e-01 3.24697524e-01 1.39653534e-01 8.30051601e-01
9.02814150e-01 4.48056966e-01 2.30765328e-01 2.30100229e-01
1.58693540e+00 -1.29045141e+00 -3.28989178e-01 -4.50326234e-01
7.59544253e-01 -5.29990315e-01 1.19128144e+00 3.76231641e-01
-4.22335267e-01 -7.78430581e-01 -1.07890558e+00 1.36503190e-01
-3.39917809e-01 5.72541475e-01 2.84062088e-01 5.53846836e-01
-3.93773943e-01 5.55701792e-01 -8.58094394e-01 1.50923684e-01
9.82826650e-01 4.20914352e-01 -3.21602941e-01 -1.95972547e-01
-9.41384017e-01 6.59942627e-01 9.74001765e-01 2.17066467e-01
-1.06139934e+00 -8.23887587e-01 -8.87512863e-01 9.29159671e-02
8.94313395e-01 -2.08581239e-01 1.36270189e+00 -9.30771828e-01
-1.14033222e+00 1.01514435e+00 2.41288528e-01 -3.84230524e-01
5.08639216e-01 -2.29125246e-01 -2.02795669e-01 4.88492958e-02
1.44581631e-01 1.07075691e+00 9.70390022e-01 -1.40678406e+00
-1.06660986e+00 -3.05166602e-01 2.51513645e-02 1.58267900e-01
-3.47794592e-01 -4.50454265e-01 -3.90869558e-01 -4.76795018e-01
1.72814757e-01 -1.00936139e+00 -5.97049892e-02 3.13345522e-01
-7.48221755e-01 -7.04052627e-01 8.48842442e-01 -2.23415941e-01
7.55731940e-01 -2.36683822e+00 7.40106851e-02 -1.26159579e-01
4.37617838e-01 6.52674377e-01 -2.50003040e-01 -2.51029074e-01
4.02566716e-02 -7.23058432e-02 -3.02051067e-01 -6.09679699e-01
2.05759287e-01 5.53393543e-01 -5.02878189e-01 3.89384091e-01
5.37487090e-01 6.67943478e-01 -1.18875265e+00 -4.67705250e-01
2.36126348e-01 2.81555712e-01 -4.11894679e-01 4.20346737e-01
-4.45556074e-01 2.95270771e-01 -3.00807804e-01 5.18785715e-01
9.28992152e-01 -1.18789397e-01 -9.00670514e-02 -2.20110193e-01
3.44096065e-01 3.98107439e-01 -1.36477709e+00 1.33900571e+00
-5.00580847e-01 3.67804825e-01 4.98565733e-02 -9.77603674e-01
9.75320935e-01 1.84792966e-01 1.77576989e-02 -8.92485753e-02
2.80609857e-02 3.64892095e-01 -5.07085398e-02 -1.34461120e-01
2.74945915e-01 -2.45312646e-01 -1.02651887e-01 2.13426188e-01
4.80619490e-01 -2.98522681e-01 2.34833091e-01 -2.40415912e-02
6.97399020e-01 4.62041020e-01 1.40113488e-01 -1.82728052e-01
6.47690952e-01 6.95263818e-02 6.73687160e-01 7.82303452e-01
-3.33018184e-01 6.09715462e-01 4.91700500e-01 -3.56546581e-01
-8.80405366e-01 -1.29161441e+00 -6.02393091e-01 1.20532393e+00
2.47166455e-01 -1.91309541e-01 -8.30840766e-01 -1.43594921e+00
2.53419966e-01 7.10839152e-01 -6.30153298e-01 -2.08297908e-01
-3.61807615e-01 -6.60209954e-01 6.96729124e-01 6.50264978e-01
3.73812675e-01 -1.03934550e+00 -4.15066779e-01 2.38194853e-01
-7.01590627e-02 -1.38122928e+00 -4.67287183e-01 6.02967739e-01
-6.62772655e-01 -1.14108562e+00 -5.58083057e-01 -7.34757364e-01
7.88291335e-01 3.16958010e-01 9.77754354e-01 -2.48401925e-01
-9.51640978e-02 -1.70953438e-01 -2.48703286e-01 -4.31636304e-01
-3.60372037e-01 -1.63446758e-02 1.47300854e-01 3.03773999e-01
5.31925976e-01 -2.82096505e-01 -4.24829036e-01 6.00352824e-01
-9.31487560e-01 2.69126803e-01 7.33969212e-01 1.09120810e+00
7.21451044e-01 8.21023956e-02 4.09997910e-01 -1.25083268e+00
-4.77428138e-02 -2.57950336e-01 -9.89642203e-01 8.98137540e-02
-6.34823740e-01 3.31494182e-01 6.48489475e-01 -9.04558003e-01
-1.09017789e+00 3.64028424e-01 2.79581230e-02 -6.22599304e-01
-4.36016440e-01 4.19994779e-02 -4.53038156e-01 1.73698649e-01
6.88947260e-01 3.63699696e-03 -1.45716950e-01 -6.91357374e-01
6.15914583e-01 7.86065340e-01 8.27743471e-01 -5.40367961e-01
9.80175197e-01 4.61746901e-01 -1.64107233e-01 -3.20554614e-01
-1.58673525e+00 -6.26174271e-01 -7.58077323e-01 2.39992812e-01
7.96580136e-01 -1.05606139e+00 -3.51808518e-01 5.04742622e-01
-1.07328725e+00 -3.49699348e-01 -4.24221724e-01 4.25534606e-01
-2.40841180e-01 2.88317114e-01 -3.77746165e-01 -6.77605867e-01
-1.06097251e-01 -1.28237534e+00 1.36751401e+00 2.56265968e-01
-1.16354972e-01 -6.46614492e-01 -2.81181604e-01 4.92216408e-01
-1.99838966e-01 1.91445291e-01 8.40597749e-01 -1.12490916e+00
-5.68665326e-01 -4.49397624e-01 -6.23024821e-01 6.74360633e-01
9.64623764e-02 -1.94271654e-01 -1.27222371e+00 -1.09016918e-01
-2.85243243e-01 -5.98778605e-01 9.97622192e-01 -3.81713137e-02
1.16266382e+00 -3.45383912e-01 -5.62994480e-01 4.88022178e-01
1.13967144e+00 -4.26002145e-01 -4.25591059e-02 5.88490218e-02
9.50585127e-01 8.81895423e-01 8.88003647e-01 1.87830731e-01
1.45708993e-01 7.84616828e-01 7.47311115e-01 -1.65475402e-02
-4.41031009e-01 -6.17733836e-01 2.06846237e-01 2.23126501e-01
4.78688300e-01 -1.41213015e-01 -8.32516015e-01 6.42778575e-01
-1.77084494e+00 -4.25638080e-01 -2.22353056e-01 2.22585464e+00
1.07816303e+00 4.41501260e-01 1.01126023e-02 1.34695500e-01
8.94841194e-01 1.35379463e-01 -6.02069974e-01 2.46369556e-01
8.82175118e-02 7.36858547e-02 5.01326799e-01 1.82198212e-01
-1.44744098e+00 1.01405096e+00 4.35864401e+00 1.17995465e+00
-8.27287734e-01 9.10514221e-02 4.79319543e-01 5.31516932e-02
1.08039178e-01 -6.67694658e-02 -1.34434903e+00 3.65355968e-01
5.15073121e-01 3.45301360e-01 2.71430582e-01 1.47684920e+00
-1.50004238e-01 -3.04923616e-02 -1.47461069e+00 8.26087058e-01
-1.09532461e-01 -9.26615000e-01 -8.74395296e-03 -9.05157998e-02
8.23161304e-01 6.20434061e-02 8.37961677e-03 6.74269795e-01
4.08585161e-01 -6.56247616e-01 1.16986799e+00 -8.17748159e-02
8.65122080e-01 -5.64443052e-01 7.83021927e-01 6.26714289e-01
-1.04071581e+00 -2.67679036e-01 -5.81916571e-01 5.36458343e-02
-1.83738694e-01 6.58864319e-01 -1.12101090e+00 2.68455625e-01
6.26352012e-01 4.33898002e-01 -7.09078133e-01 9.82047915e-01
-8.55496764e-01 7.52140582e-01 -4.26316649e-01 1.93086686e-03
3.47330332e-01 6.88940585e-02 4.35602576e-01 1.16718340e+00
-1.16770856e-01 -2.21568584e-01 3.09747070e-01 1.28345478e+00
-3.16850364e-01 -2.02840865e-01 -1.91876605e-01 4.25807796e-02
4.99228776e-01 1.61185014e+00 -8.07571650e-01 -6.57492459e-01
-4.33301806e-01 5.47075152e-01 5.95212460e-01 2.95039117e-01
-8.67713928e-01 -3.79289597e-01 5.78498006e-01 5.30987419e-02
5.10552645e-01 1.69136137e-01 -3.84232014e-01 -1.21901214e+00
8.45446065e-02 -7.01702178e-01 3.99118334e-01 -5.14970839e-01
-1.39147282e+00 5.55612147e-01 1.81658193e-01 -1.17142844e+00
1.66083038e-01 -9.00589228e-01 -2.52529204e-01 9.08403039e-01
-1.49544811e+00 -1.37069833e+00 -3.60737056e-01 1.95482373e-01
5.23759544e-01 1.96189404e-01 6.65653408e-01 1.02866322e-01
-7.42184341e-01 6.43242359e-01 -6.68639317e-02 3.80825222e-01
8.54385078e-01 -1.49871552e+00 1.87463909e-01 9.79505837e-01
5.29926658e-01 2.58682549e-01 6.17540598e-01 -4.85048592e-01
-6.93407238e-01 -1.56019890e+00 6.52263463e-01 -5.45817792e-01
6.44189298e-01 -9.42323983e-01 -1.01940572e+00 5.32338858e-01
-4.27763104e-01 5.47983289e-01 3.77732337e-01 2.17684880e-01
-7.46034324e-01 -3.17129344e-01 -1.01940691e+00 6.11339033e-01
8.92534673e-01 -3.70867610e-01 -8.33858311e-01 4.61558044e-01
9.65476394e-01 -4.66505617e-01 -2.81300217e-01 7.14702785e-01
3.51223648e-01 -5.89003623e-01 9.66233909e-01 -4.79704708e-01
3.52932841e-01 -6.47883773e-01 -7.43199978e-03 -1.19632423e+00
-5.85401244e-02 -9.75345001e-02 -3.28175157e-01 1.48396504e+00
4.70520109e-01 -4.29709494e-01 9.99379396e-01 6.45900548e-01
-2.68143177e-01 -8.41901600e-01 -8.28145862e-01 -9.51636910e-01
-1.37877345e-01 -5.14219701e-01 7.09059596e-01 6.08555436e-01
-2.38153160e-01 5.21731079e-01 -1.93100393e-01 4.12123024e-01
7.73158371e-01 3.24796289e-01 8.64091873e-01 -1.32392061e+00
-5.58686435e-01 -3.53587717e-01 -3.83543253e-01 -1.33297062e+00
2.90457308e-01 -9.12116110e-01 6.33877695e-01 -9.72957432e-01
3.26945186e-01 -9.45613444e-01 -3.61446500e-01 9.47441459e-01
-5.52924395e-01 7.89235711e-01 1.49564266e-01 1.98537037e-01
-6.65399492e-01 5.87901413e-01 1.34238136e+00 -2.37202376e-01
-1.16136692e-01 2.52022058e-01 -5.74671507e-01 9.67289507e-01
2.65449584e-01 -7.66348839e-01 -2.71457583e-01 -2.92963386e-01
-2.38794237e-01 -2.66352683e-01 5.27951777e-01 -7.28860557e-01
-1.40743583e-01 -1.18485675e-03 1.97562039e-01 -5.46242058e-01
3.03518683e-01 -1.04155195e+00 -4.30741310e-01 3.32585305e-01
-2.25058660e-01 -8.83364022e-01 5.68072200e-02 8.56420517e-01
-1.98352918e-01 -4.84877288e-01 1.03714442e+00 1.33566618e-01
-6.82915032e-01 2.93557078e-01 1.23788483e-01 -3.21250111e-02
1.01327491e+00 -1.91033855e-01 -1.48650125e-01 2.85088956e-01
-6.60792589e-01 2.71908760e-01 3.02236199e-01 3.53082746e-01
1.78194493e-01 -1.26082349e+00 -3.54242086e-01 2.90710986e-01
5.17050266e-01 8.02750766e-01 1.79677282e-03 6.36643291e-01
-1.20426521e-01 3.68007869e-01 5.62526546e-02 -9.10795748e-01
-1.12413061e+00 9.88412380e-01 2.60643721e-01 -4.69943076e-01
-4.47918296e-01 1.11872458e+00 8.48524928e-01 -7.61857152e-01
5.45624673e-01 -6.81776822e-01 -1.02417432e-01 1.53932124e-01
4.76002127e-01 2.69437116e-02 2.99988072e-02 -5.94246209e-01
-2.92227417e-01 3.70933264e-01 -2.51877874e-01 3.41521472e-01
1.16317129e+00 1.68154575e-03 2.47728646e-01 3.45917284e-01
9.19043124e-01 -5.06330878e-02 -1.91763461e+00 -6.84554279e-01
2.26377875e-01 -4.20896739e-01 -5.14002666e-02 -7.75381982e-01
-7.66851068e-01 1.06158555e+00 4.18616712e-01 -8.45522583e-02
9.17274237e-01 1.85457230e-01 5.68125486e-01 3.16246063e-01
3.19158852e-01 -7.64235079e-01 2.02301189e-01 2.09994450e-01
4.19139385e-01 -1.35838175e+00 -2.33775198e-01 -8.65240753e-01
-5.28638482e-01 9.80438411e-01 8.47223103e-01 -2.48488799e-01
3.68201226e-01 6.21687435e-02 4.26474288e-02 3.12420130e-02
-3.63670141e-01 -3.34087670e-01 6.85352981e-01 5.55074513e-01
6.64116442e-02 2.37750679e-01 1.14534415e-01 9.98460770e-01
7.54382759e-02 -3.89329940e-01 2.07620561e-01 4.61716264e-01
-4.90453064e-01 -1.11154389e+00 -4.87765402e-01 4.78349566e-01
-3.02945733e-01 -1.38656333e-01 -2.21952692e-01 6.92013443e-01
6.15035892e-01 7.25541472e-01 -7.15994388e-02 -2.00085416e-01
3.99777085e-01 2.14774355e-01 2.80133009e-01 -1.17629242e+00
-2.82515734e-01 -6.85398234e-03 6.77449768e-03 -1.99008375e-01
-2.18401670e-01 -5.04011631e-01 -1.06736696e+00 3.51814657e-01
-9.63393211e-01 1.75689027e-01 2.48319417e-01 1.03172207e+00
6.71172217e-02 4.43570971e-01 3.57696801e-01 -9.54771280e-01
-1.03992450e+00 -1.22489691e+00 -6.27231359e-01 6.42799795e-01
3.36621016e-01 -1.18467104e+00 -4.00861561e-01 -3.73284034e-02] | [9.218660354614258, 1.3630975484848022] |
96aebd50-ad1c-4a6b-b2ab-17ee024467f5 | learning-from-adjective-noun-pairs-a | null | null | https://aclanthology.org/2022.coling-1.590 | https://aclanthology.org/2022.coling-1.590.pdf | Learning from Adjective-Noun Pairs: A Knowledge-enhanced Framework for Target-Oriented Multimodal Sentiment Classification | Target-oriented multimodal sentiment classification (TMSC) is a new subtask of aspect-based sentiment analysis, which aims to determine the sentiment polarity of the opinion target mentioned in a (sentence, image) pair. Recently, dominant works employ the attention mechanism to capture the corresponding visual representations of the opinion target, and then aggregate them as evidence to make sentiment predictions. However, they still suffer from two problems: (1) The granularity of the opinion target in two modalities is inconsistent, which causes visual attention sometimes fail to capture the corresponding visual representations of the target; (2) Even though it is captured, there are still significant differences between the visual representations expressing the same mood, which brings great difficulty to sentiment prediction. To this end, we propose a novel Knowledge-enhanced Framework (KEF) in this paper, which can successfully exploit adjective-noun pairs extracted from the image to improve the visual attention capability and sentiment prediction capability of the TMSC task. Extensive experimental results show that our framework consistently outperforms state-of-the-art works on two public datasets. | ['Jiajun Chen', 'ShuJian Huang', 'Xinyu Dai', 'Siyu Long', 'Zhen Wu', 'Fei Zhao'] | null | null | null | null | coling-2022-10 | ['aspect-based-sentiment-analysis'] | ['natural-language-processing'] | [ 1.30555600e-01 -3.43884736e-01 -1.64005294e-01 -3.51187497e-01
-4.80651081e-01 -5.26643097e-01 5.35685539e-01 2.39456128e-02
-1.41984239e-01 3.26253086e-01 5.89689314e-01 -1.17799386e-01
3.70762616e-01 -4.69204813e-01 -5.24823546e-01 -9.65007603e-01
5.69261789e-01 -8.35067704e-02 1.89583197e-01 -4.08988237e-01
6.08726799e-01 2.35035811e-02 -1.38509476e+00 9.86684322e-01
6.50777996e-01 1.41140366e+00 2.17139453e-01 1.41230687e-01
-6.26921594e-01 1.04206431e+00 -6.25259161e-01 -6.96082711e-01
-2.48160332e-01 -5.43072104e-01 -5.05199671e-01 3.73508722e-01
3.66445661e-01 7.08590448e-02 5.10174455e-03 1.39937520e+00
4.19212937e-01 -2.75657028e-01 8.70483935e-01 -1.43342805e+00
-1.09431708e+00 2.96516389e-01 -1.13819790e+00 2.11401269e-01
3.37417334e-01 6.93500414e-02 9.90970850e-01 -1.22038126e+00
5.30700743e-01 1.36322427e+00 1.57396108e-01 3.46419513e-01
-6.57549858e-01 -7.00279415e-01 7.95142531e-01 4.16008443e-01
-1.03562808e+00 -3.06622595e-01 1.09246600e+00 -2.75916934e-01
5.29262185e-01 2.05832645e-01 9.79935944e-01 1.11134171e+00
4.08386230e-01 1.09250641e+00 1.44606197e+00 -2.25815892e-01
1.06705531e-01 5.97081244e-01 7.30324760e-02 6.30067587e-01
-2.75176857e-03 -3.45705301e-01 -7.29969203e-01 2.31301948e-01
1.13395892e-01 1.28335729e-01 -5.70101917e-01 -4.95639116e-01
-1.26635802e+00 7.21112907e-01 7.40383148e-01 2.06468925e-01
-2.89509624e-01 -1.78386137e-01 5.47508538e-01 7.18501210e-02
5.62462866e-01 -2.56506633e-03 -4.09458995e-01 1.94104627e-01
-4.35721248e-01 -1.80095479e-01 3.76483798e-01 8.16820502e-01
7.74709105e-01 4.96164523e-02 -2.76181668e-01 6.23777449e-01
6.92601085e-01 7.75416195e-01 7.21023321e-01 -1.91235274e-01
7.93703258e-01 1.19814920e+00 2.07536928e-02 -1.54483175e+00
-4.15140241e-01 -2.41499066e-01 -6.53356373e-01 2.33612135e-01
5.91034964e-02 -7.19372109e-02 -8.40320587e-01 1.37268794e+00
4.80446965e-01 -1.92892313e-01 3.05255741e-01 1.31845808e+00
1.32646108e+00 8.32871675e-01 3.18974584e-01 -3.66251707e-01
1.73236477e+00 -1.17685974e+00 -8.02374840e-01 -7.91654527e-01
3.99078339e-01 -9.30616379e-01 1.22779810e+00 7.27752075e-02
-7.80044854e-01 -5.45900881e-01 -1.10241926e+00 2.46787183e-02
-6.04877710e-01 1.49020672e-01 6.39194608e-01 5.51056802e-01
-7.26044059e-01 -2.29055315e-01 -8.69049728e-02 -9.42257419e-02
4.45651948e-01 -4.59695272e-02 -4.51193571e-01 -2.29189470e-01
-1.03183901e+00 7.67344654e-01 2.25405440e-01 4.89789069e-01
-5.53239167e-01 -4.54649389e-01 -1.04125154e+00 -6.49514375e-03
2.54633069e-01 -4.01366681e-01 7.45760083e-01 -1.83814740e+00
-1.02358377e+00 8.73948038e-01 -3.77467662e-01 3.59298706e-01
5.09270802e-02 1.49924353e-01 -8.00795436e-01 1.91080913e-01
2.77849227e-01 6.95352435e-01 9.79255080e-01 -1.78654456e+00
-8.27097118e-01 -8.52745652e-01 1.59212127e-01 7.55984843e-01
-5.87256193e-01 1.18782423e-01 -8.68286967e-01 -4.76048708e-01
1.35678872e-02 -5.44845939e-01 -1.50112836e-02 -1.51376441e-01
-4.62938130e-01 -1.67302608e-01 1.02108705e+00 -2.98956603e-01
1.00321090e+00 -2.23488617e+00 1.56752795e-01 -2.32775300e-03
7.77909607e-02 9.49989855e-02 -3.33833992e-01 5.52655905e-02
-1.53107688e-01 6.61418773e-03 -1.16273716e-01 -3.05463374e-01
-3.17629755e-01 -1.03652462e-01 -6.01825535e-01 4.60495561e-01
2.09451541e-01 1.17732906e+00 -8.17033172e-01 -9.56078947e-01
7.63495192e-02 4.89309520e-01 -3.10050277e-03 2.32294314e-02
-1.12651207e-01 4.20919031e-01 -6.19084835e-01 9.99926865e-01
7.66800582e-01 -1.66451946e-01 1.23374321e-01 -7.66508818e-01
-1.37431463e-02 -1.53555527e-01 -9.05784369e-01 1.30345035e+00
-3.64600360e-01 6.84905171e-01 -7.50816241e-02 -1.07747078e+00
9.45329845e-01 1.30751073e-01 5.67510240e-02 -9.62478340e-01
3.52904141e-01 -1.72610991e-02 -1.66001350e-01 -7.44329691e-01
4.57914650e-01 -3.38382483e-01 -2.22238883e-01 2.40208253e-01
-1.61561131e-01 -2.27084085e-02 6.52612671e-02 1.30634159e-01
2.18884796e-01 1.13998335e-02 1.59462973e-01 -1.09842949e-01
9.00250971e-01 2.43694395e-01 6.37720823e-01 2.15981573e-01
-3.64034086e-01 6.39161825e-01 9.72645283e-01 -4.32920605e-01
-6.10162556e-01 -6.63626730e-01 3.59217860e-02 1.04563081e+00
9.32495236e-01 -2.36745134e-01 -3.91901910e-01 -1.14118516e+00
-3.47230047e-01 4.80246276e-01 -1.00144804e+00 -1.79936901e-01
-1.31253496e-01 -7.54105866e-01 -1.69870302e-01 6.57005072e-01
4.58402932e-01 -1.46732509e+00 -3.24338704e-01 -2.10124344e-01
-3.60179454e-01 -1.12571383e+00 -6.95700765e-01 -1.62429944e-01
-5.90060055e-01 -1.08044887e+00 -8.65417898e-01 -1.23016751e+00
1.00863552e+00 7.20785022e-01 1.10842180e+00 1.12089686e-01
2.06667915e-01 4.45743412e-01 -4.39294130e-01 -6.69249475e-01
1.87386632e-01 -3.77544075e-01 -3.36332113e-01 5.78671038e-01
6.51906252e-01 8.06454718e-02 -9.27595079e-01 2.92395562e-01
-8.27752292e-01 1.47680402e-01 6.82250619e-01 7.11239159e-01
8.58860075e-01 8.36136471e-03 4.73323703e-01 -8.90816569e-01
6.74757719e-01 -4.29708213e-01 -3.58792782e-01 5.07990897e-01
-4.88292336e-01 -4.47202057e-01 6.45405293e-01 -4.41016883e-01
-1.16383564e+00 4.63315286e-02 2.37433091e-01 -3.99999529e-01
-1.10181205e-01 7.81480908e-01 -4.03168261e-01 1.31331116e-01
3.86672691e-02 5.42197049e-01 -1.18532576e-01 6.08889759e-02
2.79109329e-01 8.32018971e-01 2.58435428e-01 -6.56801760e-02
3.46234679e-01 7.05464005e-01 -1.00429870e-01 -4.25001025e-01
-1.31961215e+00 -4.67736036e-01 -3.62844378e-01 -5.64966977e-01
1.02107382e+00 -1.11167479e+00 -7.44024158e-01 4.02272463e-01
-1.21415436e+00 2.28206828e-01 2.72507548e-01 2.51128584e-01
-2.44040862e-01 5.03827274e-01 -2.66716182e-01 -9.20320451e-01
-5.33267200e-01 -1.47721922e+00 1.34189391e+00 6.20453179e-01
1.69619024e-01 -9.27334666e-01 -9.93624926e-02 5.33282161e-01
1.54840708e-01 -3.86038832e-02 9.69081521e-01 -3.20987254e-01
-2.87616551e-01 -2.64584541e-01 -6.49615884e-01 7.39427879e-02
-9.56470370e-02 9.45941731e-02 -9.70201969e-01 -1.00632139e-01
5.22772223e-02 -3.38102221e-01 9.92557347e-01 2.19245657e-01
1.12449551e+00 -1.67152241e-01 -3.89125556e-01 3.58633310e-01
1.43118238e+00 2.61429489e-01 7.19368219e-01 3.53020459e-01
8.79981101e-01 1.02869856e+00 1.00282121e+00 5.19592054e-02
6.33969486e-01 5.94573677e-01 8.47633183e-01 -4.33094144e-01
3.09098195e-02 -1.48527771e-01 6.71756864e-01 8.42192590e-01
8.50018114e-02 -2.17258051e-01 -5.42392433e-01 7.39250898e-01
-2.00466013e+00 -8.65589917e-01 -3.26355994e-01 1.70482635e+00
5.26022077e-01 1.51035368e-01 -1.04143329e-01 -1.74933761e-01
7.87224472e-01 5.19258499e-01 -6.00133121e-01 -3.94887120e-01
-3.68478298e-01 -3.93721759e-01 4.26212214e-02 8.24191123e-02
-1.24722552e+00 8.31197858e-01 5.19617414e+00 9.28406477e-01
-1.43621802e+00 8.86621885e-03 1.04771745e+00 1.19225055e-01
-5.31652212e-01 -1.43324256e-01 -7.30053902e-01 6.63081765e-01
2.95113325e-01 1.44931987e-01 -3.94857116e-03 1.07993400e+00
-5.28405085e-02 -8.93356130e-02 -5.02906561e-01 1.01453722e+00
6.77399457e-01 -1.06882811e+00 3.60046774e-01 -1.17998362e-01
8.03865790e-01 -4.24583077e-01 4.24698114e-01 3.74870270e-01
-3.75462979e-01 -8.98842216e-01 8.76281679e-01 5.43261349e-01
7.56808996e-01 -9.53997910e-01 1.05586183e+00 -5.58297299e-02
-1.48444104e+00 -1.37863830e-01 -6.74067318e-01 1.81117952e-01
1.07371539e-01 4.64239925e-01 -1.96298197e-01 3.82190615e-01
9.93521333e-01 1.07370472e+00 -8.43980074e-01 7.07751691e-01
-4.95383769e-01 2.52558649e-01 4.72658753e-01 -5.01246333e-01
2.46703118e-01 -1.64937273e-01 3.50215495e-01 9.88588333e-01
1.09497897e-01 -9.45777521e-02 -1.03091765e-02 6.11160517e-01
4.17403094e-02 5.18907249e-01 -6.00038052e-01 -6.46256208e-02
-4.40386720e-02 1.78632545e+00 -9.07685757e-01 -2.92850196e-01
-7.99989998e-01 1.15682340e+00 3.95714402e-01 5.80768764e-01
-8.46473217e-01 -3.75961244e-01 4.27247494e-01 -3.21937174e-01
5.44055343e-01 2.77819246e-01 -4.10066396e-01 -1.39460731e+00
2.11254314e-01 -9.26654458e-01 3.35247070e-01 -1.38814580e+00
-1.36546350e+00 8.40365231e-01 -5.39230764e-01 -1.63333118e+00
3.59259814e-01 -9.56373394e-01 -8.71577621e-01 8.17310691e-01
-1.74248147e+00 -1.45633793e+00 -4.10323650e-01 7.97841549e-01
6.23600543e-01 -1.89076185e-01 6.34382844e-01 -6.30409569e-02
-6.36631429e-01 4.05243754e-01 -2.42877170e-01 1.06641300e-01
6.87718153e-01 -1.10555065e+00 -2.44160846e-01 7.69870758e-01
8.46661702e-02 4.49962765e-01 4.04092193e-01 -5.63410938e-01
-1.52515113e+00 -1.01941383e+00 8.54098260e-01 -4.79365647e-01
6.71571434e-01 -2.70777851e-01 -7.32444823e-01 3.58594030e-01
5.51768184e-01 -5.09364307e-02 6.73495829e-01 4.73652668e-02
-5.18449903e-01 -2.68736035e-01 -6.67932510e-01 9.23053443e-01
5.14916718e-01 -4.85350043e-01 -5.87998152e-01 1.48713619e-01
6.32340133e-01 -2.02562600e-01 -4.55356836e-01 4.79354352e-01
5.74270368e-01 -1.22028470e+00 9.32129323e-01 -5.71006417e-01
9.99747097e-01 -7.29111075e-01 -2.29665592e-01 -1.36423564e+00
-8.76332540e-03 4.16031241e-01 -1.90846976e-02 1.41070580e+00
6.59093618e-01 -3.41741979e-01 5.10919929e-01 2.72969186e-01
2.73880269e-02 -1.05149043e+00 -7.34622002e-01 1.24836363e-01
-2.09462643e-01 -3.72235209e-01 5.98150492e-01 8.11855972e-01
2.36088932e-01 8.88304353e-01 -3.61638516e-01 2.89569139e-01
1.21586286e-01 7.86415637e-01 6.13679111e-01 -8.60496402e-01
1.44122452e-01 -5.85770130e-01 -4.50773895e-01 -9.68167663e-01
1.23789877e-01 -4.77911353e-01 4.72765639e-02 -1.75159645e+00
5.61022580e-01 -1.11670252e-02 -5.46169043e-01 3.71834457e-01
-6.42687261e-01 7.11776018e-01 2.77502537e-01 1.70109585e-01
-1.07403111e+00 8.05300832e-01 1.71078241e+00 -6.19363964e-01
6.74403086e-02 -2.88270593e-01 -1.30484891e+00 8.19660783e-01
5.91996729e-01 -1.49625257e-01 -3.83984149e-01 -2.96669781e-01
6.40774310e-01 6.09438773e-03 1.92270741e-01 -4.31379288e-01
1.50649771e-01 -2.84269869e-01 8.58803689e-01 -8.57878506e-01
3.66019696e-01 -1.00130880e+00 -5.21281838e-01 3.21446300e-01
-1.16078727e-01 2.37436116e-01 1.69753671e-01 8.57873440e-01
-5.07210076e-01 -1.19186260e-01 6.06878102e-01 -2.20055375e-02
-1.18619740e+00 2.18416989e-01 -2.89705753e-01 -7.74259716e-02
1.05291069e+00 -8.90995115e-02 -7.11677492e-01 -4.29721922e-01
-2.64897883e-01 3.83900017e-01 4.23146397e-01 7.25884736e-01
9.17885184e-01 -1.46974993e+00 -5.01542032e-01 1.75594270e-01
8.03100348e-01 -3.13804656e-01 6.91403449e-01 1.11340594e+00
-1.33053243e-01 2.30955273e-01 -6.26303703e-02 -6.26571178e-01
-1.33334613e+00 9.62424397e-01 3.46047431e-01 -4.23660241e-02
-1.99673563e-01 7.75631964e-01 8.79410326e-01 -3.41705710e-01
-2.81545054e-02 1.63108736e-01 -1.04321706e+00 4.13085520e-01
7.90045440e-01 -2.20303744e-01 -2.02831969e-01 -1.26265848e+00
-6.10448182e-01 1.06161642e+00 -1.30846575e-01 1.34135067e-01
9.13453281e-01 -5.40925443e-01 -2.46029168e-01 4.74232137e-01
1.11030102e+00 2.11443216e-01 -1.03766870e+00 -8.58195219e-03
-5.41057944e-01 -4.43855405e-01 5.20476811e-02 -8.40725183e-01
-1.53188872e+00 1.03645062e+00 6.07849538e-01 2.05908254e-01
1.31132710e+00 1.21005520e-01 4.81300026e-01 -2.78568873e-03
3.75631973e-02 -9.55192387e-01 1.47406012e-01 2.73881793e-01
9.33988571e-01 -1.64006472e+00 -2.87829395e-02 -3.57938141e-01
-1.52831888e+00 1.08184767e+00 9.18100655e-01 1.18159726e-01
5.39713442e-01 -2.77924836e-01 6.76723659e-01 -5.72523177e-01
-5.70901573e-01 -3.45137209e-01 7.54220724e-01 6.75120115e-01
4.72407550e-01 -1.36963770e-01 -3.29133719e-01 8.05253685e-01
1.68475315e-01 -6.53181314e-01 3.39456290e-01 6.01484895e-01
-3.64888459e-01 -5.74170172e-01 -4.55593199e-01 2.92960793e-01
-4.59690332e-01 -3.23208123e-01 -5.98567843e-01 5.64141572e-01
1.00954898e-01 1.32262695e+00 1.07056037e-01 -4.08945382e-01
4.01467711e-01 -2.46139899e-01 1.53747154e-02 -2.48044342e-01
-3.89741033e-01 3.54878664e-01 -1.24541998e-01 -5.39644241e-01
-8.40630949e-01 -3.29735070e-01 -1.06650186e+00 1.70627534e-01
-3.13406795e-01 1.51474983e-01 6.28098309e-01 9.33604181e-01
3.30416560e-01 5.47614932e-01 7.68782496e-01 -6.81203604e-01
1.63930684e-01 -6.98771536e-01 -7.20961511e-01 7.06524432e-01
2.79996008e-01 -7.91765571e-01 -3.22970718e-01 3.20688868e-03] | [10.86031723022461, 1.7689217329025269] |
b821611a-b943-42e6-b944-7ced3625bf45 | reasoning-based-software-testing | 2303.01302 | null | https://arxiv.org/abs/2303.01302v1 | https://arxiv.org/pdf/2303.01302v1.pdf | Reasoning-Based Software Testing | With software systems becoming increasingly pervasive and autonomous, our ability to test for their quality is severely challenged. Many systems are called to operate in uncertain and highly-changing environment, not rarely required to make intelligent decisions by themselves. This easily results in an intractable state space to explore at testing time. The state-of-the-art techniques try to keep the pace, e.g., by augmenting the tester's intuition with some form of (explicit or implicit) learning from observations to search this space efficiently. For instance, they exploit historical data to drive the search (e.g., ML-driven testing) or the tests execution data itself (e.g., adaptive or search-based testing). Despite the indubitable advances, the need for smartening the search in such a huge space keeps to be pressing. We introduce Reasoning-Based Software Testing (RBST), a new way of thinking at the testing problem as a causal reasoning task. Compared to mere intuition-based or state-of-the-art learning-based strategies, we claim that causal reasoning more naturally emulates the process that a human would do to ''smartly" search the space. RBST aims to mimic and amplify, with the power of computation, this ability. The conceptual leap can pave the ground to a new trend of techniques, which can be variously instantiated from the proposed framework, by exploiting the numerous tools for causal discovery and inference. Preliminary results reported in this paper are promising. | ['Stefano Russo', 'Roberto Pietrantuono', 'Luca Giamattei'] | 2023-03-02 | null | null | null | null | ['causal-discovery'] | ['knowledge-base'] | [ 2.13245749e-01 3.98144513e-01 -3.99211556e-01 -2.66786486e-01
-2.82667458e-01 -5.68916202e-01 7.62589753e-01 2.26825476e-01
-5.65620838e-03 7.77567923e-01 -2.95883685e-01 -7.80328929e-01
-5.38357019e-01 -1.12877464e+00 -8.32563698e-01 -5.77524662e-01
-2.54000604e-01 7.41214633e-01 4.68368024e-01 -2.97101110e-01
6.13445461e-01 2.83066958e-01 -2.01433563e+00 5.83273061e-02
1.12298942e+00 8.74613822e-01 9.73349437e-02 5.45592785e-01
-9.46287736e-02 7.06915498e-01 -2.70453006e-01 -1.75395265e-01
-5.03818356e-02 -5.90963244e-01 -7.17932463e-01 -1.73475355e-01
-3.28133166e-01 5.08570001e-02 2.21578747e-01 1.18808413e+00
1.62510015e-02 -3.70777220e-01 -1.16620725e-02 -1.47905827e+00
-3.83385569e-01 8.22090149e-01 -1.71994105e-01 -1.05251089e-01
6.59954011e-01 4.02160108e-01 6.90406859e-01 -4.48435068e-01
4.10422772e-01 1.08014882e+00 3.60314220e-01 4.42842990e-01
-9.85881925e-01 -3.97198170e-01 2.91847169e-01 5.75891376e-01
-1.29095829e+00 -1.84049875e-01 8.86115432e-01 -6.48815870e-01
6.62333488e-01 5.96136928e-01 8.95482361e-01 1.10731268e+00
3.04399401e-01 6.71836853e-01 1.56326997e+00 -7.11883664e-01
7.06248105e-01 4.89231259e-01 -8.83526579e-02 6.91139519e-01
5.05603969e-01 4.78449017e-01 -6.48340523e-01 -3.39288563e-01
5.36115587e-01 -5.57403490e-02 -1.55533522e-01 -5.53369224e-01
-8.35754395e-01 6.76124394e-01 -1.63968742e-01 3.63097608e-01
-3.43508452e-01 1.66591331e-01 2.25212887e-01 4.90431100e-01
6.83134794e-02 8.14669311e-01 -7.51841843e-01 -5.81507146e-01
-1.05249596e+00 1.65153325e-01 1.19673097e+00 4.84252393e-01
6.96088910e-01 1.75398327e-02 1.05147898e-01 -3.50043853e-03
6.39513552e-01 2.13265136e-01 6.23282015e-01 -4.42843586e-01
-1.03291474e-01 1.10554183e+00 -1.51961595e-02 -5.19767880e-01
-3.09488922e-02 -4.26810205e-01 -3.56191754e-01 6.74989998e-01
2.38598078e-01 -6.38540760e-02 -7.69180179e-01 1.68436885e+00
4.20762241e-01 4.01959687e-01 -1.63162500e-01 7.09316492e-01
-3.03090423e-01 2.43153170e-01 -2.42718861e-01 -4.65285808e-01
1.00690675e+00 -5.67585170e-01 -5.62402129e-01 -8.38565305e-02
3.96120876e-01 -3.25710654e-01 1.31760061e+00 1.07451189e+00
-1.02002430e+00 -3.75533253e-01 -1.48277342e+00 8.85740817e-01
-4.67972100e-01 -5.12495399e-01 8.56508732e-01 8.87511432e-01
-1.12095273e+00 7.20938385e-01 -1.10108709e+00 -1.27221912e-01
1.83326825e-01 3.83154064e-01 -2.43361406e-02 -2.55100895e-03
-1.22138572e+00 8.90628457e-01 3.19428384e-01 3.49415809e-01
-1.29058981e+00 -4.67296600e-01 -5.09812355e-01 -3.99759710e-02
1.04169273e+00 -5.05892634e-01 1.08870244e+00 -8.34166706e-01
-1.67597485e+00 4.25027311e-01 1.48428217e-01 -5.47585130e-01
8.07693124e-01 -3.70324492e-01 -5.87938786e-01 -2.92211860e-01
-1.22483112e-01 -1.11669078e-01 8.59538555e-01 -1.31282628e+00
-3.17738593e-01 -4.88062888e-01 2.07239345e-01 -4.65155244e-01
-2.10271224e-01 -1.26103938e-01 1.49774045e-01 -8.42362940e-02
1.64930567e-01 -8.75493824e-01 -5.22937834e-01 -3.67927492e-01
-5.18800259e-01 -2.36494690e-01 8.32480431e-01 -6.38939291e-02
1.47119677e+00 -1.78889346e+00 1.36332616e-01 4.57080334e-01
1.41485065e-01 1.86988145e-01 2.78379828e-01 6.27237856e-01
-1.35780245e-01 4.19338197e-01 -1.35494843e-01 2.20723301e-01
3.64896148e-01 2.38975897e-01 -3.46813470e-01 2.85091728e-01
3.41253489e-01 7.94266105e-01 -9.95506704e-01 -4.64732349e-01
1.60831392e-01 7.30616003e-02 -6.68035984e-01 2.05433071e-01
-8.64430666e-01 5.83941638e-01 -9.72749949e-01 7.07932591e-01
2.29490116e-01 -2.62636065e-01 3.96850944e-01 4.05373633e-01
-4.43449050e-01 2.66567439e-01 -1.38521218e+00 1.45592093e+00
-5.92880368e-01 2.32655942e-01 -3.58289689e-01 -1.15865827e+00
9.65328693e-01 4.03644949e-01 9.38621163e-02 -6.88728511e-01
-3.41159217e-02 4.27838564e-01 2.78812438e-01 -8.06322455e-01
-6.94402978e-02 -2.29197755e-01 -1.67361185e-01 5.42079389e-01
-3.83286893e-01 -1.98974431e-01 7.85802975e-02 -2.77206630e-01
1.35868812e+00 5.19086540e-01 5.35073102e-01 -2.41279185e-01
7.89738655e-01 2.74206072e-01 6.28160119e-01 7.59617209e-01
1.71029091e-01 -1.73401549e-01 7.77123928e-01 -5.80851853e-01
-7.10502982e-01 -1.06086826e+00 -5.66172414e-02 7.26567924e-01
2.92635914e-02 -2.69829750e-01 -6.66179776e-01 -7.84805775e-01
-1.40291885e-01 9.69060004e-01 -5.66482008e-01 -3.72575641e-01
-3.98425102e-01 -4.67235565e-01 2.89583892e-01 1.85911164e-01
3.60121131e-01 -1.05555677e+00 -1.15534794e+00 3.63609672e-01
3.71801287e-01 -3.93171281e-01 4.53466624e-01 4.91502225e-01
-1.02980864e+00 -9.81291234e-01 1.56864643e-01 9.63206664e-02
3.57873231e-01 -3.45873564e-01 1.29278696e+00 3.59212726e-01
-4.23770845e-01 2.86292851e-01 -6.06766880e-01 -5.87496161e-01
-6.67942107e-01 -3.47799450e-01 9.83634591e-02 -3.55554223e-02
3.28152448e-01 -1.07458305e+00 -5.01150370e-01 3.23630452e-01
-9.06078577e-01 -2.43376493e-01 9.63495851e-01 8.58745217e-01
4.40474927e-01 5.81872761e-01 5.85400403e-01 -9.16506290e-01
6.83504879e-01 -6.24031425e-01 -8.82969141e-01 5.52976787e-01
-1.36001444e+00 5.97691238e-01 6.33517802e-01 -6.58843517e-01
-9.26131070e-01 -3.75621676e-01 2.73888379e-01 -2.29112864e-01
-2.88670480e-01 8.55245352e-01 -3.78936857e-01 7.51065835e-02
6.32224202e-01 3.01607996e-01 -2.72185266e-01 -1.78506568e-01
3.36123079e-01 4.40322340e-01 3.41610283e-01 -9.35826838e-01
9.72105324e-01 2.66891807e-01 1.48094937e-01 -1.73388973e-01
-4.32332367e-01 -1.62457246e-02 -4.56341803e-01 -4.14600521e-01
6.76385105e-01 -1.31733254e-01 -8.84047329e-01 6.21668994e-02
-7.51147926e-01 -5.00175178e-01 -5.07261097e-01 3.10883790e-01
-5.69639623e-01 1.69849973e-02 8.21815804e-02 -1.29037249e+00
2.11938798e-01 -1.14717829e+00 8.12463045e-01 3.20150316e-01
-3.48893106e-01 -8.25819492e-01 4.77386475e-01 1.93202868e-01
4.21330631e-01 5.05882382e-01 1.15307462e+00 -8.83928597e-01
-9.78918195e-01 -4.77109998e-01 3.34073514e-01 2.65397519e-01
-9.91808847e-02 2.30826050e-01 -9.45569098e-01 4.47289161e-02
5.00448167e-01 -2.62110621e-01 1.53817400e-01 -8.73022079e-02
1.10097826e+00 -3.87772083e-01 -2.59406567e-01 4.82254475e-02
1.52864480e+00 5.34350812e-01 6.68549299e-01 4.43966776e-01
1.25676036e-01 6.89590037e-01 9.38548267e-01 7.68792033e-01
1.79446921e-01 6.49739563e-01 7.80209363e-01 1.93543509e-01
2.88738042e-01 -4.36054856e-01 2.78829604e-01 5.93074858e-01
-2.28440121e-01 -7.44639784e-02 -1.34309614e+00 2.99492836e-01
-2.01168299e+00 -9.50462937e-01 -3.17645706e-02 2.46091962e+00
9.42869127e-01 6.50150001e-01 -2.15893894e-01 4.86031801e-01
3.60039264e-01 -2.40285873e-01 -6.43733263e-01 -6.06847644e-01
3.13213617e-01 3.51470709e-01 9.38404948e-02 2.58622289e-01
-4.76120234e-01 5.64668894e-01 5.39167881e+00 5.67311108e-01
-1.19509482e+00 7.08707944e-02 3.57667089e-01 2.79967070e-01
-6.99011624e-01 5.32551706e-01 -3.94259810e-01 3.52320611e-01
1.25870383e+00 -2.77812809e-01 6.27092898e-01 1.05422890e+00
3.40735763e-01 -4.62098956e-01 -1.49049449e+00 5.17956018e-01
-1.78751230e-01 -1.11455965e+00 -3.09729248e-01 1.89457938e-01
5.79630852e-01 -2.54495412e-01 8.81942064e-02 2.72404581e-01
5.49664617e-01 -1.20107532e+00 9.55424905e-01 7.51821697e-01
3.80408674e-01 -5.55970848e-01 8.65043759e-01 7.35691905e-01
-7.56914675e-01 -4.76613075e-01 6.62806556e-02 -4.96754974e-01
1.16193714e-03 8.60053599e-01 -9.83790219e-01 6.62872374e-01
6.30824089e-01 -4.74837273e-02 -5.76605022e-01 1.01069474e+00
-6.18577242e-01 6.31195784e-01 -3.12252715e-02 -4.94706839e-01
1.88445225e-01 -1.91090569e-01 4.76048529e-01 6.62693262e-01
4.51787710e-01 -2.32973099e-01 7.26507604e-02 1.20162189e+00
5.92263103e-01 -3.31604213e-01 -6.87319160e-01 -2.68457055e-01
4.78294581e-01 9.66247380e-01 -6.98690832e-01 -3.68369401e-01
-1.37874991e-01 5.14906824e-01 -6.18263297e-02 -2.48923171e-02
-1.02875710e+00 4.89490665e-02 3.58039737e-01 3.28081608e-01
-5.92515692e-02 -2.62062877e-01 -4.05078411e-01 -7.67666459e-01
1.78712383e-01 -1.24828708e+00 4.26007248e-02 -6.76958025e-01
-9.23471808e-01 5.71227014e-01 2.43270233e-01 -1.03352690e+00
-6.24272525e-01 -4.22734678e-01 -7.08806574e-01 5.85636675e-01
-1.36866772e+00 -6.18054211e-01 -2.29235902e-01 5.00785947e-01
4.56346929e-01 1.28592640e-01 7.20160604e-01 -1.55905103e-02
-2.98500001e-01 2.18876213e-01 -5.05566657e-01 -5.85107505e-01
4.50253278e-01 -1.33473623e+00 -3.73032875e-02 8.03486824e-01
1.67937160e-01 8.22026014e-01 1.05633307e+00 -7.12373793e-01
-2.00729418e+00 -4.40084696e-01 6.89571738e-01 -5.12186944e-01
1.23247087e+00 -4.86963868e-01 -8.64048660e-01 3.47708404e-01
-1.44383637e-02 -1.94031939e-01 4.65885580e-01 4.10252213e-01
-3.82375389e-01 -1.74708664e-01 -1.04476655e+00 8.18829656e-01
1.03298390e+00 -4.93917972e-01 -8.14243257e-01 2.61186272e-01
7.73910284e-01 4.18980122e-02 -6.98564947e-01 6.75940752e-01
5.80043554e-01 -1.19897234e+00 5.07518947e-01 -5.92285335e-01
6.08870946e-02 -5.81068873e-01 -5.30840680e-02 -1.05517304e+00
1.97527438e-01 -1.10098946e+00 -1.36863559e-01 1.09288764e+00
5.31947136e-01 -8.93623352e-01 7.67203987e-01 6.82963431e-01
-2.23094508e-01 -1.00006068e+00 -9.52106655e-01 -8.68442953e-01
-2.58580714e-01 -7.09500432e-01 7.52329707e-01 9.17532802e-01
4.62431878e-01 9.94736794e-03 1.45466834e-01 3.02982748e-01
4.54484850e-01 3.10170650e-01 6.93985462e-01 -1.53131914e+00
-8.77141833e-01 -4.62724417e-01 -4.80690271e-01 -3.49091321e-01
-7.97720999e-02 -4.24540907e-01 2.34305128e-01 -1.04547608e+00
5.13562150e-02 -7.00061560e-01 -3.41101110e-01 5.25972188e-01
1.00294590e-01 -3.91682386e-01 -2.83451557e-01 -5.83590120e-02
-6.00874543e-01 2.13785738e-01 1.12914288e+00 1.12145036e-01
-2.61444896e-01 1.58662766e-01 -6.94568813e-01 8.52127314e-01
8.53818357e-01 -4.20070112e-01 -8.40721369e-01 1.53536245e-01
9.73072827e-01 4.91920263e-01 5.79827964e-01 -1.10933173e+00
4.88453686e-01 -4.23458070e-01 -1.76245004e-01 -2.66439199e-01
-1.37636393e-01 -7.34945059e-01 5.66649675e-01 8.92186224e-01
-1.92410126e-01 2.43362233e-01 -1.05303966e-01 6.05160832e-01
-3.68907988e-01 -3.80913496e-01 3.94792557e-01 -1.24775402e-01
-7.38743961e-01 -5.03889704e-03 -2.60174394e-01 -2.51378417e-01
1.11328638e+00 -4.21272457e-01 -3.22406232e-01 -5.38292378e-02
-7.45422721e-01 9.08666179e-02 4.24224347e-01 2.90314525e-01
5.47779739e-01 -7.89811611e-01 -3.56329590e-01 2.57714123e-01
4.63425703e-02 -1.46787435e-01 1.42639086e-01 1.07429421e+00
-1.18030600e-01 3.62142682e-01 -1.11442983e-01 -6.07206583e-01
-7.32088566e-01 8.80317867e-01 5.36579639e-02 -5.89536369e-01
-3.07488084e-01 6.49591029e-01 -1.20188102e-01 -8.86996388e-02
-1.11435503e-02 -2.75083274e-01 6.67411461e-02 -2.18492135e-01
3.24538141e-01 3.24994415e-01 1.45287707e-01 6.91824138e-01
-5.07639289e-01 2.77871728e-01 3.33667696e-01 -3.55697125e-01
1.41212118e+00 2.19731152e-01 -3.78231913e-01 7.64960825e-01
2.83033997e-01 1.82019919e-01 -1.14919341e+00 2.30529383e-01
6.33756578e-01 -3.64665419e-01 -3.81125920e-02 -1.27863204e+00
-4.45096016e-01 7.42223918e-01 4.86790925e-01 8.57502759e-01
1.24088573e+00 8.23655725e-02 -2.25008652e-02 2.89991051e-01
1.01368427e+00 -1.01475239e+00 2.27511436e-01 8.03228319e-02
5.90615273e-01 -1.09190595e+00 -1.86851963e-01 -1.98635533e-01
-2.52225995e-01 1.13940620e+00 5.41879773e-01 -1.06240369e-01
4.86896694e-01 8.51680398e-01 -2.95662194e-01 -3.23274404e-01
-1.19318092e+00 3.16264890e-02 -1.04195110e-01 4.85971093e-01
2.53950268e-01 2.23099992e-01 -3.23219419e-01 5.49683154e-01
-1.40955627e-01 2.44848922e-01 5.49969018e-01 9.89913523e-01
-6.08895898e-01 -1.55058432e+00 -5.70218980e-01 1.94033384e-01
-3.94831859e-02 -3.71304415e-02 -3.95462751e-01 1.17090595e+00
4.96032923e-01 1.15117610e+00 -4.07528222e-01 -5.33334196e-01
1.97573587e-01 1.84028223e-01 4.76899087e-01 -5.61950386e-01
-1.51497111e-01 -2.04299778e-01 6.39707446e-02 -1.00957811e+00
-2.25460708e-01 -8.67068768e-01 -1.08441043e+00 -2.55167503e-02
-3.05412382e-01 4.56630230e-01 9.80167150e-01 1.07410002e+00
4.62937467e-02 7.35996604e-01 6.67828321e-01 -3.62792164e-01
-8.59760344e-01 -5.61575413e-01 -1.90799922e-01 -3.39530915e-01
7.15956884e-03 -8.71913850e-01 -4.45875704e-01 -3.31583351e-01] | [5.455946445465088, 2.7785487174987793] |
edd0033c-c978-4f86-a296-8709f5ac6e9d | learning-discrete-structures-for-graph-neural | 1903.11960 | null | https://arxiv.org/abs/1903.11960v4 | https://arxiv.org/pdf/1903.11960v4.pdf | Learning Discrete Structures for Graph Neural Networks | Graph neural networks (GNNs) are a popular class of machine learning models whose major advantage is their ability to incorporate a sparse and discrete dependency structure between data points. Unfortunately, GNNs can only be used when such a graph-structure is available. In practice, however, real-world graphs are often noisy and incomplete or might not be available at all. With this work, we propose to jointly learn the graph structure and the parameters of graph convolutional networks (GCNs) by approximately solving a bilevel program that learns a discrete probability distribution on the edges of the graph. This allows one to apply GCNs not only in scenarios where the given graph is incomplete or corrupted but also in those where a graph is not available. We conduct a series of experiments that analyze the behavior of the proposed method and demonstrate that it outperforms related methods by a significant margin. | ['Massimiliano Pontil', 'Luca Franceschi', 'Xiao He', 'Mathias Niepert'] | 2019-03-28 | null | null | null | null | ['music-genre-recognition'] | ['music'] | [-1.04027696e-01 1.84257090e-01 -1.98262542e-01 -1.98361561e-01
-6.71892911e-02 -6.43500865e-01 4.29770201e-01 4.42435563e-01
-2.79890209e-01 7.25660801e-01 -2.75235951e-01 -5.21786273e-01
-3.63379061e-01 -1.03343666e+00 -1.05372560e+00 -5.29030323e-01
-4.34177727e-01 7.58173406e-01 -2.31144149e-02 -3.58754247e-02
-2.45032072e-01 6.84998155e-01 -9.54541504e-01 -3.86682779e-01
7.33307779e-01 6.64333344e-01 1.82217911e-01 6.52043700e-01
-1.75410621e-02 8.06903660e-01 -3.01031917e-01 -3.94241542e-01
3.33966464e-01 -5.92044480e-02 -5.75004637e-01 1.83289364e-01
2.26160944e-01 -2.25679010e-01 -8.90676975e-01 1.30818462e+00
1.86007410e-01 1.58467799e-01 5.37887514e-01 -1.37827575e+00
-3.13861609e-01 7.93610156e-01 -5.15951633e-01 1.25289649e-01
-7.20923245e-02 1.14724040e-01 1.09791243e+00 -2.58324772e-01
5.26200771e-01 1.01046348e+00 6.36742890e-01 1.88109815e-01
-1.26191020e+00 -5.56256831e-01 3.86602074e-01 1.81589067e-01
-1.43605435e+00 -2.90017426e-02 1.04920602e+00 -1.59049481e-01
5.77857971e-01 -1.53051078e-01 6.65866554e-01 1.01474321e+00
6.75000176e-02 5.17543614e-01 6.68178916e-01 -9.02828798e-02
3.34168196e-01 -2.45937407e-01 2.16826424e-01 1.05700541e+00
7.62555718e-01 5.21545187e-02 -2.08884791e-01 -2.53450960e-01
9.02568281e-01 2.15108469e-01 -2.29604140e-01 -6.22374356e-01
-8.57767999e-01 9.62970078e-01 8.94244432e-01 1.54919818e-01
-5.62339485e-01 4.77419555e-01 3.39148074e-01 4.20477003e-01
3.38752836e-01 2.02536255e-01 -9.58559811e-02 1.09214321e-01
-7.06909716e-01 1.11895971e-01 1.12440097e+00 8.07793081e-01
9.71607983e-01 4.09469940e-02 2.85201490e-01 5.39150596e-01
2.41032422e-01 2.43024006e-01 2.16053724e-02 -5.55917978e-01
5.85817039e-01 6.81241512e-01 -7.61483563e-03 -1.46907318e+00
-5.86453378e-01 -6.19290650e-01 -1.34108448e+00 -1.10260077e-01
6.25294566e-01 -4.07643318e-01 -9.98076618e-01 1.77230179e+00
9.74237397e-02 5.65451324e-01 -1.90621838e-01 7.45600343e-01
7.52600253e-01 5.23316383e-01 -2.53220767e-01 4.89218011e-02
7.63782561e-01 -6.19196653e-01 -4.33925480e-01 -4.50613201e-01
5.67139030e-01 -1.73442826e-01 7.51061857e-01 2.37479195e-01
-6.82441115e-01 -1.94846004e-01 -9.94230926e-01 2.02610344e-01
-2.18693420e-01 -6.86957240e-02 7.27356374e-01 4.98636335e-01
-1.10400414e+00 8.44653368e-01 -1.10776651e+00 -2.64672309e-01
4.73734856e-01 5.40589809e-01 -6.26664460e-01 -5.56124091e-01
-9.06393230e-01 4.47063118e-01 4.90802974e-01 5.84175527e-01
-1.05046439e+00 -2.66853631e-01 -1.05301344e+00 5.11444509e-01
6.70344472e-01 -4.82731372e-01 8.64926040e-01 -8.24712873e-01
-9.45035815e-01 3.95127565e-01 1.05833232e-01 -7.23465383e-01
4.64658707e-01 2.06412435e-01 -5.71845099e-02 1.26663208e-01
-3.48725438e-01 1.58259735e-01 6.69478655e-01 -1.06424606e+00
-1.13822065e-01 -4.19260204e-01 5.52377224e-01 7.97780603e-02
-3.20892423e-01 -5.45897484e-01 -6.65710747e-01 -3.96686137e-01
1.86501279e-01 -1.04534805e+00 -4.63998884e-01 7.18248785e-02
-8.28331947e-01 -1.47092775e-01 9.26300228e-01 -6.45415187e-01
9.55392480e-01 -2.03705955e+00 1.48194909e-01 6.25438392e-01
5.66726148e-01 3.05251092e-01 -2.79975712e-01 6.25756383e-01
1.21617511e-01 1.79921731e-01 -3.04720104e-01 -3.64258468e-01
-1.70004070e-02 6.78364575e-01 9.02826935e-02 7.16965139e-01
-4.81786067e-03 9.18015122e-01 -8.43580902e-01 -6.77938163e-02
1.24378473e-01 5.29287815e-01 -4.14774716e-01 2.66311198e-01
-5.63043773e-01 3.42831284e-01 -5.41738868e-01 4.09646392e-01
5.87535083e-01 -7.68560708e-01 6.33389950e-01 5.09603545e-02
5.16367197e-01 6.15715496e-02 -1.44703090e+00 1.36333978e+00
-3.77455771e-01 7.01048732e-01 2.67553687e-01 -1.42187786e+00
8.93377602e-01 1.43861860e-01 4.22504634e-01 -3.15707982e-01
2.15701327e-01 -3.94662581e-02 2.68627703e-01 -2.50585228e-01
1.35886580e-01 -2.98587270e-02 1.68526798e-01 5.26974738e-01
1.02720186e-01 1.46221891e-01 5.26359677e-01 2.36676350e-01
1.60514200e+00 -4.64649230e-01 1.49504811e-01 -5.23367859e-02
1.13318957e-01 -3.38463366e-01 5.32704532e-01 9.70244765e-01
9.74091291e-02 4.58912611e-01 9.80263412e-01 -5.49443543e-01
-9.31739151e-01 -8.91859710e-01 3.68207753e-01 4.20433074e-01
7.93725476e-02 -3.30589384e-01 -5.90054274e-01 -8.49755943e-01
-2.99571585e-02 2.16559961e-01 -5.62150538e-01 -1.49720535e-01
-4.36384797e-01 -6.93252981e-01 1.95562631e-01 4.46420044e-01
4.18323308e-01 -9.37181354e-01 -3.90662365e-02 2.22629696e-01
-7.79645592e-02 -1.45542598e+00 -2.31129736e-01 2.02230856e-01
-9.38280165e-01 -1.57994020e+00 -3.23742807e-01 -8.10048521e-01
1.06491828e+00 2.38665760e-01 1.05837381e+00 4.69489485e-01
4.95061686e-04 2.28721350e-01 -1.31383032e-01 -1.43820554e-01
-3.95752132e-01 3.18193287e-01 -2.63062656e-01 2.09583238e-01
9.86179262e-02 -9.63340282e-01 -2.61880100e-01 -1.46165723e-02
-1.09766591e+00 -9.69623849e-02 5.00484049e-01 8.71510088e-01
4.43065703e-01 6.27267897e-01 4.68412369e-01 -1.37045598e+00
6.88680589e-01 -6.96794748e-01 -1.04379463e+00 1.31912917e-01
-4.62567270e-01 1.49947301e-01 9.91505861e-01 -3.37936014e-01
-4.70908582e-01 4.13910821e-02 -2.68968455e-02 -7.00374424e-01
-1.98004767e-01 1.13073456e+00 -1.77926704e-01 -2.45691687e-01
4.73116487e-01 -9.15038288e-02 4.16794419e-02 -5.33765793e-01
1.84815094e-01 1.89420387e-01 3.39693010e-01 -4.14297909e-01
9.70517278e-01 4.14725900e-01 4.24595028e-01 -7.61801481e-01
-5.65175474e-01 -3.97079110e-01 -4.81471807e-01 -2.42330693e-02
2.67756552e-01 -8.06566358e-01 -6.66114748e-01 4.82189029e-01
-8.89152586e-01 -6.23736799e-01 -2.58361679e-02 5.23006916e-01
-3.18518460e-01 4.94210303e-01 -7.26896226e-01 -5.41669428e-01
-2.62489133e-02 -9.78473008e-01 5.44431925e-01 2.72498608e-01
2.57045388e-01 -1.33890760e+00 -2.18191028e-01 -8.89832675e-02
2.58081377e-01 5.36714792e-01 1.06417692e+00 -8.98675442e-01
-8.13203514e-01 -5.55426538e-01 -3.17417353e-01 2.56487370e-01
2.39727214e-01 2.89692376e-02 -4.37565148e-01 -5.73119462e-01
-3.75868946e-01 -2.24805474e-01 6.89382493e-01 5.74405849e-01
1.17741632e+00 -3.45852971e-01 -2.26648584e-01 6.25539839e-01
1.63755524e+00 -1.81341544e-01 2.99527794e-01 -7.54668266e-02
1.00234568e+00 4.41385865e-01 9.40710306e-03 3.25739115e-01
5.17616510e-01 2.72813588e-01 8.94168019e-01 -1.63468406e-01
1.54124349e-01 -5.13339579e-01 1.40034398e-02 7.65200555e-01
4.25659977e-02 -6.90259576e-01 -9.67751026e-01 6.30651116e-01
-1.99902892e+00 -5.43807387e-01 -5.41617759e-02 2.11877847e+00
3.68822366e-01 1.85655981e-01 7.11315572e-02 4.35907096e-02
9.12645102e-01 3.50021690e-01 -7.32525229e-01 -1.05369657e-01
-1.71400383e-01 2.02728212e-01 6.23336017e-01 4.22144443e-01
-9.79254603e-01 7.12405264e-01 5.84899569e+00 3.15585792e-01
-1.04699135e+00 -3.47640187e-01 5.21249831e-01 2.18379021e-01
-2.10815787e-01 1.63512275e-01 -3.20353121e-01 3.73334050e-01
8.73274982e-01 -1.98837414e-01 6.82981908e-01 6.60503745e-01
2.84273755e-02 1.44428033e-02 -1.22948825e+00 9.44050014e-01
-1.82677254e-01 -1.29878998e+00 -1.94171160e-01 3.10179085e-01
7.04135418e-01 3.74614775e-01 -1.45075783e-01 1.98479712e-01
8.42709959e-01 -1.24542451e+00 1.90526485e-01 3.46718043e-01
4.91077572e-01 -8.20655942e-01 7.52069473e-01 7.36911237e-01
-1.09894192e+00 -1.24555133e-01 -5.09329498e-01 -1.58924296e-01
9.62359458e-02 9.13421452e-01 -9.38385189e-01 6.38706803e-01
5.39315999e-01 8.25665951e-01 -4.18678939e-01 1.40312040e+00
-4.98485506e-01 8.52370262e-01 -6.99033439e-01 -3.24282572e-02
3.84802669e-01 -3.89202178e-01 5.47185659e-01 7.53247201e-01
3.40696305e-01 -1.03715420e-01 3.71058434e-01 8.68632138e-01
-6.12780094e-01 -9.22131613e-02 -9.46968853e-01 -4.29841936e-01
3.62035424e-01 1.19790649e+00 -9.18416798e-01 -2.71050334e-02
-6.34241343e-01 7.48824477e-01 7.68363297e-01 6.58222854e-01
-5.23067176e-01 -2.03328133e-01 4.55785096e-01 7.48914182e-02
4.87335563e-01 -6.18205607e-01 1.52792096e-01 -1.13563049e+00
1.72053680e-01 -7.68368065e-01 5.80621183e-01 -4.82515007e-01
-1.48694336e+00 5.15172601e-01 -8.91260877e-02 -6.39891863e-01
-3.33802909e-01 -5.77654064e-01 -6.07531309e-01 7.20856667e-01
-1.59268188e+00 -1.09154797e+00 -3.40172827e-01 6.48682415e-01
-8.98803025e-02 1.96000119e-03 6.57182574e-01 2.97489047e-01
-5.93739569e-01 3.93941015e-01 2.36579791e-01 5.66555560e-01
1.03836127e-01 -1.24052215e+00 5.80158949e-01 1.03752279e+00
5.70828497e-01 4.14520174e-01 6.39730453e-01 -6.74533069e-01
-1.61105859e+00 -1.16842532e+00 4.95639414e-01 2.98774302e-01
7.47942626e-01 -6.23455405e-01 -1.10616112e+00 9.79556739e-01
-1.62289858e-01 6.79267466e-01 2.74310172e-01 3.42114806e-01
-2.86816716e-01 -9.39503014e-02 -1.04686606e+00 5.43766558e-01
8.92092288e-01 -5.46163142e-01 1.05780907e-01 4.68892753e-01
4.83593702e-01 -5.95920086e-01 -7.04185009e-01 3.11570197e-01
2.00592175e-01 -7.95715868e-01 6.73896790e-01 -6.36150002e-01
6.70631826e-02 -2.38051176e-01 3.43777537e-02 -1.66442549e+00
3.38577945e-03 -6.06285214e-01 -3.49625707e-01 8.83830667e-01
3.86705905e-01 -7.76274204e-01 1.06376302e+00 5.63640237e-01
3.16921771e-01 -5.58135033e-01 -9.07324374e-01 -7.53054321e-01
-1.38502240e-01 -3.59003276e-01 7.22962737e-01 9.81064260e-01
-4.10364240e-01 3.19030792e-01 -4.53526229e-01 5.49807370e-01
6.42653823e-01 1.31448552e-01 8.33481014e-01 -1.42769432e+00
-4.17885363e-01 -1.76951453e-01 -7.19328940e-01 -9.51067805e-01
5.50059915e-01 -8.75658274e-01 -8.89383405e-02 -1.68509126e+00
9.42826122e-02 -5.53606272e-01 -2.56531894e-01 4.93645996e-01
2.80912761e-02 -2.46001668e-02 4.30275276e-02 -2.13436544e-01
-4.02841300e-01 4.21426535e-01 9.87798512e-01 -2.21993148e-01
-1.74962550e-01 3.69827121e-01 -5.86426675e-01 6.41736329e-01
8.18847835e-01 -5.16343951e-01 -6.32413149e-01 -4.68902051e-01
3.77995431e-01 2.89239854e-01 5.84918439e-01 -8.86662126e-01
5.10362387e-01 -2.10983939e-02 -2.33673304e-02 -3.50675702e-01
9.04100314e-02 -1.07423973e+00 3.68376046e-01 2.35355049e-01
-6.33700266e-02 6.38721362e-02 4.82014380e-02 1.02860248e+00
-2.48750806e-01 -2.47720525e-01 5.52507758e-01 -1.30926639e-01
-4.17116284e-01 9.48468804e-01 3.94144505e-02 3.46204713e-02
7.07447708e-01 1.36234879e-01 -2.43483588e-01 -8.33116055e-01
-8.07686329e-01 3.59135538e-01 5.13249516e-01 2.08181322e-01
6.04919434e-01 -1.17631400e+00 -5.74167311e-01 2.70462602e-01
-9.47364196e-02 4.97111320e-01 1.79851502e-01 5.81439853e-01
-5.67584991e-01 1.13006704e-01 -9.70264431e-03 -4.36777204e-01
-1.03952706e+00 5.73449790e-01 3.90804917e-01 -4.73267734e-01
-9.80740488e-01 5.37256777e-01 4.55074236e-02 -5.65414369e-01
4.98990506e-01 -2.86850542e-01 -1.28834993e-01 -4.20383215e-01
1.72608584e-01 3.42870392e-02 1.67717129e-01 -6.50879860e-01
7.43767843e-02 7.26001859e-02 -4.71598245e-02 3.36941272e-01
1.57808375e+00 -8.79837796e-02 -2.12732702e-01 3.80559504e-01
1.15109694e+00 -2.88358092e-01 -1.34156871e+00 -6.91976190e-01
-4.95092794e-02 -3.36473942e-01 7.11955316e-03 -2.61479735e-01
-1.56585205e+00 9.06546175e-01 1.04260713e-01 5.73712230e-01
9.80207860e-01 4.53641526e-02 7.30476081e-01 6.90103412e-01
2.54674703e-01 -7.17227876e-01 -1.37961790e-01 4.79471534e-01
5.29575348e-01 -1.16582775e+00 -1.02341317e-01 -4.66641366e-01
-1.01496041e-01 1.12104166e+00 3.57591897e-01 -4.63568449e-01
9.46668386e-01 1.95305288e-01 -2.94167846e-01 -4.10775155e-01
-7.55198240e-01 -1.40105337e-01 7.59149343e-02 8.33850861e-01
-4.58838306e-02 1.55955851e-01 1.15780011e-01 3.99549276e-01
-1.09603994e-01 -2.19391376e-01 7.64443278e-01 7.03899205e-01
-4.76841442e-02 -1.06911302e+00 -2.12434232e-02 7.82127976e-01
-4.94006425e-01 4.85976087e-03 -3.60461771e-01 8.12277019e-01
-3.86478186e-01 9.02370751e-01 -3.17173339e-02 -1.53147876e-01
1.78866282e-01 -4.04573560e-01 3.86151582e-01 -6.65693164e-01
-1.85296744e-01 -1.39221013e-01 9.84335020e-02 -6.06158137e-01
-2.26660579e-01 -3.39939892e-01 -1.12854636e+00 -6.12554252e-01
-2.65945137e-01 1.03805311e-01 6.98053360e-01 1.00751650e+00
2.73680896e-01 5.01118660e-01 5.54997683e-01 -5.23009360e-01
-4.06188965e-01 -7.13294923e-01 -1.02699876e+00 1.80537984e-01
5.97430825e-01 -5.36444366e-01 -4.08495814e-01 -2.96671033e-01] | [6.910409450531006, 6.160907745361328] |
1096b07f-59cc-4ee8-b704-8986ecb60dc5 | mdssd-multi-scale-deconvolutional-single-shot | 1805.07009 | null | https://arxiv.org/abs/1805.07009v3 | https://arxiv.org/pdf/1805.07009v3.pdf | MDSSD: Multi-scale Deconvolutional Single Shot Detector for Small Objects | For most of the object detectors based on multi-scale feature maps, the shallow layers are rich in fine spatial information and thus mainly responsible for small object detection. The performance of small object detection, however, is still less than satisfactory because of the deficiency of semantic information on shallow feature maps. In this paper, we design a Multi-scale Deconvolutional Single Shot Detector (MDSSD), especially for small object detection. In MDSSD, multiple high-level feature maps at different scales are upsampled simultaneously to increase the spatial resolution. Afterwards, we implement the skip connections with low-level feature maps via Fusion Block. The fusion feature maps, named Fusion Module, are of strong feature representational power of small instances. It is noteworthy that these high-level feature maps utilized in Fusion Block preserve both strong semantic information and some fine details of small instances, rather than the top-most layer where the representation of fine details for small objects are potentially wiped out. The proposed framework achieves 77.6% mAP for small object detection on the challenging dataset TT100K with 512 x 512 input, outperforming other detectors with a large margin. Moreover, it can also achieve state-of-the-art results for general object detection on PASCAL VOC2007 test and MS COCO test-dev2015, especially achieving 2 to 5 points improvement on small object categories. | ['Zhimin Gao', 'Pei Lv', 'Rui Ma', 'Xiaoheng Jiang', 'Mingliang Xu', 'Bing Zhou', 'Lisha Cui'] | 2018-05-18 | null | null | null | null | ['small-object-detection'] | ['computer-vision'] | [-4.36130539e-02 -1.85021713e-01 1.81999415e-01 -2.19771296e-01
-7.80315816e-01 -2.07963467e-01 4.18998003e-01 1.20628826e-01
-7.82450795e-01 4.29502130e-01 -6.75512478e-02 3.71874034e-01
1.65589601e-01 -8.54626358e-01 -7.55429804e-01 -1.01042295e+00
9.75663289e-02 -4.80121374e-02 1.18067443e+00 -7.53888041e-02
1.48719540e-02 4.15932804e-01 -1.74194205e+00 7.48803914e-01
5.79889655e-01 1.39874613e+00 8.25864077e-01 4.19712245e-01
-2.94018723e-02 5.34471095e-01 -6.81832373e-01 -1.53645560e-01
1.26040563e-01 -5.02179228e-02 -1.91788033e-01 -2.55733460e-01
7.21032321e-01 -5.25730312e-01 -4.27532017e-01 1.37267387e+00
7.35075355e-01 -1.33777395e-01 5.82945943e-01 -9.77166235e-01
-6.58762097e-01 4.31225479e-01 -9.02797639e-01 7.78184533e-01
-9.45382565e-03 3.23964864e-01 7.94221520e-01 -1.43949723e+00
4.12663192e-01 1.35447598e+00 5.42246699e-01 3.85520071e-01
-8.52671742e-01 -9.74432945e-01 1.57759517e-01 4.58679914e-01
-1.65540695e+00 -2.75324732e-01 3.21736336e-01 -3.09090823e-01
1.11879086e+00 5.99607974e-02 4.56605136e-01 8.93395185e-01
3.28805149e-01 1.15592325e+00 1.05879903e+00 -4.89617605e-03
-4.64052558e-02 4.56078291e-01 3.32228959e-01 5.82839072e-01
5.17506897e-01 9.18295905e-02 -5.92527628e-01 2.95033693e-01
7.15169549e-01 4.26691532e-01 -1.60743430e-01 -1.39278144e-01
-1.21438396e+00 6.66597903e-01 1.15249431e+00 2.94453293e-01
-3.74233425e-01 -2.24148065e-01 1.83374852e-01 -4.16478887e-02
1.94224596e-01 -9.94541571e-02 -3.69735211e-01 4.61149484e-01
-8.86816442e-01 1.02721646e-01 1.18309230e-01 1.00731790e+00
7.04080522e-01 3.54777761e-02 -7.49029279e-01 7.22955048e-01
4.02850538e-01 7.22977281e-01 5.99687099e-01 -4.07309264e-01
6.89616680e-01 8.61273825e-01 9.72130448e-02 -6.84556305e-01
-5.76207757e-01 -8.92765820e-01 -8.65886867e-01 3.94667208e-01
4.70505744e-01 1.02698028e-01 -1.11553609e+00 1.47259665e+00
3.14225107e-01 5.71694747e-02 2.28920773e-01 1.23326397e+00
1.30020905e+00 7.29732156e-01 1.93931803e-01 -7.56537840e-02
1.79138231e+00 -1.03235102e+00 -4.71374452e-01 -5.06456554e-01
2.82022864e-01 -6.92242146e-01 9.44162250e-01 1.20626874e-01
-9.46264923e-01 -1.04167116e+00 -1.21444619e+00 -8.19660351e-02
-4.39208329e-01 6.31938696e-01 5.06067276e-01 4.50459749e-01
-7.84102380e-01 3.55748147e-01 -7.05173492e-01 -4.24057424e-01
9.05652523e-01 3.03711265e-01 -3.58532041e-01 -2.97997504e-01
-1.05381846e+00 9.38373029e-01 6.57488048e-01 2.37967178e-01
-9.54268634e-01 -7.06411541e-01 -7.50843525e-01 3.53948534e-01
4.40849215e-01 -4.00366664e-01 9.19558942e-01 -4.09045607e-01
-7.73224890e-01 7.01866329e-01 -6.36331290e-02 -5.97817183e-01
4.72749770e-01 -1.73479468e-01 -4.06231821e-01 2.01562107e-01
2.73977309e-01 9.54628289e-01 1.11099839e+00 -8.37587714e-01
-1.22190475e+00 -6.96221590e-01 -1.34210050e-01 3.25640053e-01
-5.64694643e-01 2.92197466e-01 -4.83959496e-01 -5.32483935e-01
4.43915501e-02 -5.02027035e-01 -9.50072519e-03 2.11193457e-01
-2.96020240e-01 -4.33063596e-01 1.15003645e+00 -3.61219853e-01
1.06962490e+00 -2.30855513e+00 -1.21962830e-01 -3.96740377e-01
3.47275615e-01 6.17553711e-01 -1.35249898e-01 -1.25747383e-01
2.47016639e-01 -3.12779367e-01 -4.62781154e-02 -1.98460743e-01
-1.01346232e-01 -2.42255062e-01 -3.07209700e-01 4.89086449e-01
5.25796711e-01 1.05061948e+00 -6.61127150e-01 -4.35681432e-01
5.57860315e-01 6.02702081e-01 -2.61009425e-01 -6.75282106e-02
3.20694894e-01 8.51862691e-03 -6.76957488e-01 8.63153338e-01
1.12950706e+00 -2.56604463e-01 -5.35580873e-01 -5.39121032e-01
-3.81524563e-01 -1.50573254e-01 -1.58786631e+00 1.52186263e+00
-1.42509654e-01 6.10637665e-01 1.12294085e-01 -5.07422388e-01
7.26920366e-01 1.71891078e-02 -2.81393528e-02 -8.70000899e-01
1.94431961e-01 3.48321587e-01 2.02514470e-01 -1.80296287e-01
4.01342273e-01 -1.68800354e-02 -3.08056325e-02 -1.62739888e-01
2.78352588e-01 1.63652256e-01 9.30709392e-02 2.69039840e-01
9.57766950e-01 -2.50558704e-01 4.05714989e-01 -2.73568958e-01
6.25344813e-01 -1.65079728e-01 4.86866564e-01 9.26836193e-01
-4.51411784e-01 8.02338004e-01 1.40050963e-01 -1.87026218e-01
-8.19193780e-01 -1.33293498e+00 -6.13278985e-01 1.07515228e+00
5.91672719e-01 -2.30426624e-01 -6.81268394e-01 -6.05914950e-01
1.98820859e-01 4.20571417e-01 -7.23173201e-01 -3.48289758e-01
-4.68603015e-01 -1.10020232e+00 5.20663679e-01 9.56262290e-01
8.29138696e-01 -1.37801421e+00 -9.05644536e-01 2.93130845e-01
2.42269501e-01 -1.36505139e+00 -4.01332110e-01 4.75720733e-01
-6.75797880e-01 -1.08647037e+00 -9.44764197e-01 -8.98331881e-01
5.45197785e-01 8.23406994e-01 6.03850663e-01 -3.26090157e-01
-8.15919518e-01 -2.42554456e-01 -1.85312986e-01 -7.37584174e-01
8.65532160e-02 -2.08059266e-01 4.48273905e-02 6.36918694e-02
5.72950959e-01 8.73353519e-03 -8.51134360e-01 5.01837194e-01
-7.93514132e-01 -1.08558893e-01 1.10402024e+00 8.82069886e-01
6.29228950e-01 1.17770240e-01 6.53311133e-01 -3.53289217e-01
9.25831869e-02 -2.26546451e-01 -5.87980807e-01 -3.23574096e-02
-1.79743826e-01 -2.29473412e-01 7.33898878e-01 -4.02108490e-01
-1.12589025e+00 1.17493691e-02 -1.12840869e-01 -4.71934468e-01
-1.64295331e-01 -3.33504826e-01 -2.76693821e-01 -1.83773637e-01
7.93384135e-01 4.69655156e-01 -3.12618226e-01 -7.24126577e-01
2.17067427e-03 7.50913560e-01 5.12890458e-01 7.90874958e-02
7.41614819e-01 6.16356850e-01 -1.59449711e-01 -8.29164147e-01
-1.13104773e+00 -7.59360313e-01 -5.99156618e-01 -4.71452600e-04
1.01177633e+00 -1.43665564e+00 -5.36667526e-01 8.88440609e-01
-1.00890219e+00 9.44575593e-02 -4.18681592e-01 4.83511627e-01
-5.24071567e-02 -1.59220412e-01 -6.02911115e-01 -5.70733368e-01
-4.40644950e-01 -1.23798823e+00 1.51218224e+00 7.00864494e-01
5.66774666e-01 -2.96946436e-01 -6.46331370e-01 1.19127728e-01
4.46963161e-01 -9.92936343e-02 3.90266031e-01 -5.35258114e-01
-6.80548728e-01 -4.56145644e-01 -9.82501566e-01 4.40336764e-01
-1.31727874e-01 -3.51080090e-01 -1.38721251e+00 -5.32255769e-01
2.98779234e-02 -4.43911582e-01 1.46069169e+00 6.06009662e-01
1.03014815e+00 2.28390783e-01 -6.46471798e-01 5.18680573e-01
1.62264252e+00 -4.38283570e-02 4.46659297e-01 2.06388429e-01
8.59019160e-01 3.10723394e-01 8.45480263e-01 4.03390974e-01
1.82041585e-01 6.54650807e-01 4.53674346e-01 -1.31967783e-01
-7.19570696e-01 -5.02503775e-02 4.05659288e-01 1.41961932e-01
6.34469241e-02 -2.87973899e-02 -6.32864952e-01 5.02014756e-01
-1.63997269e+00 -9.23100770e-01 -3.21987271e-01 2.07957482e+00
3.89241546e-01 5.41117251e-01 1.33373350e-01 5.04228957e-02
8.99532378e-01 1.60648212e-01 -8.15565288e-01 2.34948859e-01
-5.43051362e-01 1.36570066e-01 7.41539359e-01 -6.96773380e-02
-1.29789293e+00 9.81511950e-01 5.27798033e+00 1.18179619e+00
-9.12015855e-01 3.93249840e-01 2.75995135e-01 -3.48129451e-01
4.87958848e-01 -5.14400005e-01 -1.61538255e+00 4.87445086e-01
6.38453126e-01 9.81153101e-02 -1.58767477e-01 1.01437378e+00
1.00316249e-01 -2.71450460e-01 -7.50025749e-01 1.05642712e+00
-1.06219873e-01 -1.24315751e+00 -8.49927366e-02 -1.41841799e-01
7.20134616e-01 3.71913612e-01 1.80281207e-01 4.74317372e-01
-1.15606174e-01 -8.59981239e-01 9.95996475e-01 4.11312170e-02
6.94259524e-01 -6.56647801e-01 8.89165103e-01 6.06450975e-01
-1.58658469e+00 -6.18899107e-01 -9.29591417e-01 1.52125299e-01
-1.71225965e-02 4.78608489e-01 -4.35205191e-01 2.82707512e-01
1.06049907e+00 6.83300436e-01 -8.85011017e-01 1.42125285e+00
-2.64944673e-01 2.84767210e-01 -3.98281187e-01 -8.52656811e-02
4.36730951e-01 3.30733806e-01 5.20926952e-01 1.38595378e+00
3.61264855e-01 9.80022103e-02 9.95760411e-02 8.20954561e-01
-1.17009237e-01 -1.00615017e-01 -3.58617485e-01 4.82796818e-01
3.62267494e-01 1.62247562e+00 -9.07927990e-01 -5.81888735e-01
-5.79733074e-01 8.91267538e-01 2.21870452e-01 2.58035153e-01
-9.41017628e-01 -5.39726436e-01 6.35957181e-01 1.60885200e-01
8.95803690e-01 3.64035480e-02 -1.82728469e-01 -1.14030266e+00
1.93963990e-01 -5.21102369e-01 4.88193661e-01 -5.93939841e-01
-8.81726563e-01 6.64799094e-01 -1.30688369e-01 -1.18581975e+00
2.99969196e-01 -9.01040316e-01 -6.44970238e-01 1.01318157e+00
-1.78343654e+00 -1.23705876e+00 -6.12621844e-01 6.92243397e-01
7.88312018e-01 -1.53406724e-01 4.53491598e-01 4.45400983e-01
-7.51953065e-01 7.26865947e-01 2.73054749e-01 -1.99510553e-03
7.10298836e-01 -1.08294761e+00 2.65636623e-01 9.85729277e-01
1.50451928e-01 3.23416084e-01 4.61699843e-01 -4.84582543e-01
-1.32271850e+00 -1.36407971e+00 3.35655391e-01 -4.24873203e-01
4.29847956e-01 -6.73319101e-01 -1.06342578e+00 2.36800924e-01
-2.15907753e-01 6.91242218e-01 -1.00379467e-01 -4.24692363e-01
-4.18701172e-01 -2.70368725e-01 -1.28696179e+00 2.36933865e-02
9.60300565e-01 -3.54934633e-01 -6.92667961e-01 3.23059857e-01
7.62177408e-01 -3.21896583e-01 -5.57730019e-01 5.60159564e-01
4.66854066e-01 -1.04815578e+00 1.24071276e+00 -2.52447546e-01
4.02763002e-02 -6.83224678e-01 -3.18046480e-01 -1.05820537e+00
-5.67271054e-01 4.52845953e-02 -2.77882099e-01 1.12752771e+00
1.57754526e-01 -4.58081305e-01 6.95199549e-01 -8.06896091e-02
-2.88471192e-01 -5.04122376e-01 -9.35195804e-01 -8.60663116e-01
-2.16835022e-01 -3.23502034e-01 2.82647222e-01 2.47454032e-01
-5.92473030e-01 4.46917593e-01 1.12794712e-02 4.64175105e-01
9.39434707e-01 2.96202451e-01 3.10078323e-01 -1.14456928e+00
-1.62778944e-01 -4.22524601e-01 -7.32505858e-01 -9.66654480e-01
-3.40294003e-01 -7.55221307e-01 7.63181597e-02 -1.54732680e+00
6.04680419e-01 -8.22317824e-02 -7.46039510e-01 4.03099746e-01
-4.67579365e-01 8.64291430e-01 4.18175280e-01 2.46885642e-01
-8.63337398e-01 5.23619533e-01 1.40431976e+00 -1.93367302e-01
5.72021818e-03 1.38871530e-02 -6.77010477e-01 8.41622889e-01
4.06668842e-01 -5.24760783e-01 -1.10045910e-01 -3.50796223e-01
-5.10568976e-01 -2.52241075e-01 7.41589129e-01 -1.43372893e+00
3.45642298e-01 2.40657285e-01 1.05321681e+00 -9.86362815e-01
4.40933943e-01 -5.69737196e-01 -3.82225633e-01 7.54278362e-01
1.53890653e-02 -6.20324016e-01 5.04569411e-01 6.55787349e-01
-3.36528569e-01 -7.02149747e-03 1.32693470e+00 -1.94694160e-03
-1.33910513e+00 4.39164013e-01 -7.28920177e-02 -1.31980143e-02
1.27909064e+00 -3.93419206e-01 -4.46175635e-01 5.27194560e-01
-6.24432325e-01 2.48660356e-01 1.12596996e-01 6.96245611e-01
8.35295498e-01 -1.33263433e+00 -7.63542533e-01 4.33128238e-01
3.30258399e-01 2.49859408e-01 7.03970611e-01 8.44287217e-01
-3.55152301e-02 6.07103050e-01 -5.28617680e-01 -7.13840246e-01
-1.25739849e+00 6.47081554e-01 2.12759227e-01 1.15710527e-01
-8.98059845e-01 1.36433387e+00 1.01934266e+00 1.67655200e-01
1.86094254e-01 -5.67503452e-01 -3.89021397e-01 3.31125945e-01
1.21088123e+00 3.33303154e-01 1.09993286e-01 -7.76910603e-01
-7.27866709e-01 7.47535110e-01 -2.71560133e-01 5.62672019e-01
1.42699850e+00 -1.41998082e-01 3.04730743e-01 1.87124655e-01
1.23113441e+00 -2.96409398e-01 -1.67817593e+00 -5.27109444e-01
-3.95621091e-01 -5.08452058e-01 3.06726575e-01 -6.84583843e-01
-1.02442157e+00 1.21668780e+00 1.03988779e+00 -4.18138839e-02
1.07825100e+00 3.49025011e-01 7.03163087e-01 2.12958336e-01
2.82549083e-01 -9.43010151e-01 1.33305401e-01 3.42925012e-01
8.35556269e-01 -1.63667810e+00 -3.15985270e-02 -3.04508775e-01
-6.05494440e-01 9.73003387e-01 8.65417063e-01 -3.39857399e-01
2.96131313e-01 4.40811157e-01 -3.66020411e-01 -1.95084080e-01
-5.66435277e-01 -5.77310443e-01 7.02951431e-01 5.53160429e-01
4.42881957e-02 1.15185663e-01 8.72913077e-02 9.18929935e-01
1.95937961e-01 -3.09894383e-01 2.46565565e-01 5.18444359e-01
-1.18807960e+00 -3.17815274e-01 -5.65839827e-01 6.96530044e-01
-3.52570027e-01 -2.50914872e-01 -1.69945657e-02 9.12080944e-01
4.95220810e-01 7.82697916e-01 2.12906599e-01 -2.13551477e-01
5.88245511e-01 -3.31456751e-01 4.68043655e-01 -6.93375885e-01
-5.11898875e-01 2.08587378e-01 -2.66429663e-01 -6.95917308e-01
-1.74695887e-02 -6.94056392e-01 -1.45683610e+00 6.88430071e-02
-5.68505347e-01 -2.54576713e-01 5.65560997e-01 7.46733546e-01
1.28052145e-01 8.20498109e-01 3.27151269e-01 -1.03272569e+00
-7.41792858e-01 -1.16784871e+00 -8.91524255e-01 1.40820757e-01
3.51277292e-01 -9.93699312e-01 -1.91491351e-01 -4.02783632e-01] | [8.806475639343262, -0.48596686124801636] |
0f193f1d-be47-4b6e-a896-51a59b252df5 | end-to-end-conversational-search-for-online | 2109.05460 | null | https://arxiv.org/abs/2109.05460v1 | https://arxiv.org/pdf/2109.05460v1.pdf | End-to-End Conversational Search for Online Shopping with Utterance Transfer | Successful conversational search systems can present natural, adaptive and interactive shopping experience for online shopping customers. However, building such systems from scratch faces real word challenges from both imperfect product schema/knowledge and lack of training dialog data.In this work we first propose ConvSearch, an end-to-end conversational search system that deeply combines the dialog system with search. It leverages the text profile to retrieve products, which is more robust against imperfect product schema/knowledge compared with using product attributes alone. We then address the lack of data challenges by proposing an utterance transfer approach that generates dialogue utterances by using existing dialog from other domains, and leveraging the search behavior data from e-commerce retailer. With utterance transfer, we introduce a new conversational search dataset for online shopping. Experiments show that our utterance transfer method can significantly improve the availability of training dialogue data without crowd-sourcing, and the conversational search system significantly outperformed the best tested baseline. | ['Yaohui Jin', 'Hao He', 'Tong Zhao', 'Wei Chen', 'Nasser Zalmout', 'Pascual Martinez-Gomez', 'Xin Luna Dong', 'Jun Ma2', 'Liqiang Xiao'] | 2021-09-12 | null | https://aclanthology.org/2021.emnlp-main.280 | https://aclanthology.org/2021.emnlp-main.280.pdf | emnlp-2021-11 | ['conversational-search'] | ['natural-language-processing'] | [-1.04988776e-01 6.32768571e-01 -1.43330336e-01 -7.55950451e-01
-9.77801979e-01 -7.55007982e-01 8.57713521e-01 -8.65619779e-02
-2.33274817e-01 4.10610825e-01 6.16646469e-01 -2.33191773e-01
8.57607424e-02 -5.29816091e-01 -2.87249953e-01 -8.13707933e-02
4.45407838e-01 1.33508873e+00 9.28363428e-02 -1.03860760e+00
2.40145281e-01 -3.43936920e-01 -1.35073411e+00 5.68492711e-01
9.22280967e-01 1.15035594e+00 4.41864520e-01 6.24341965e-01
-6.19284809e-01 8.92805636e-01 -5.23594797e-01 -8.53775084e-01
2.52751678e-01 -3.28349233e-01 -1.29272211e+00 2.17238411e-01
2.98632801e-01 -6.70393288e-01 -3.69571522e-02 5.01498878e-01
5.93006015e-01 4.57262516e-01 2.18303517e-01 -1.18077314e+00
-1.10656631e+00 1.01521456e+00 3.16312462e-01 -4.30308193e-01
8.44480932e-01 3.15821737e-01 1.33669281e+00 -8.61379206e-01
8.47755253e-01 1.55738723e+00 4.60556418e-01 8.20129633e-01
-1.36971092e+00 -4.53812093e-01 1.98450729e-01 -4.53313952e-03
-8.39734077e-01 -5.11874437e-01 7.38269091e-01 -2.67385632e-01
1.30387223e+00 1.93749979e-01 2.31918082e-01 1.59598196e+00
-4.12683249e-01 1.40824449e+00 6.06932461e-01 -3.47302914e-01
9.57928672e-02 6.84630096e-01 3.52768749e-01 4.41091299e-01
-8.07695210e-01 -1.67070404e-01 -7.77207851e-01 -3.51356030e-01
3.38036925e-01 -1.20472707e-01 1.26325667e-01 -4.03968006e-01
-9.29259479e-01 1.18267274e+00 1.57083020e-01 4.30942737e-02
-3.06411386e-01 -4.49035168e-01 5.88642538e-01 6.41011417e-01
3.97522300e-01 9.35321331e-01 -8.94399643e-01 -4.34161752e-01
-2.17607319e-01 5.00673950e-01 1.77824116e+00 1.55143011e+00
6.05850518e-01 -4.74477053e-01 -2.31716111e-01 1.48296607e+00
2.60163993e-01 6.31230533e-01 6.77047551e-01 -1.32469356e+00
4.53460127e-01 7.88808763e-01 3.38730425e-01 -5.94205618e-01
-3.25578928e-01 1.26361087e-01 -1.72792390e-01 -5.31776428e-01
4.17685419e-01 -2.48350278e-01 -4.09266114e-01 1.57585907e+00
2.80215710e-01 -5.86277664e-01 4.13195997e-01 8.64646614e-01
1.27949893e+00 3.65199357e-01 2.65995950e-01 1.40330255e-01
1.60625648e+00 -1.59485948e+00 -8.37594211e-01 -3.95567536e-01
7.26309478e-01 -9.15639460e-01 1.49869454e+00 3.21385652e-01
-1.20362031e+00 -5.34227371e-01 -6.12761199e-01 -3.94220978e-01
-5.75339317e-01 -1.21917926e-01 8.23350906e-01 5.35429597e-01
-9.05710161e-01 5.47348149e-02 -4.19344336e-01 -6.54695451e-01
-1.56512871e-01 3.01140219e-01 -2.36580968e-02 1.93047430e-02
-1.34791410e+00 7.77819097e-01 2.99880039e-02 -2.39092425e-01
-4.34957504e-01 -6.59976900e-01 -9.91271496e-01 2.16532707e-01
7.77610958e-01 -6.82500780e-01 2.18499398e+00 -6.80842161e-01
-2.17265010e+00 5.14327765e-01 -1.16234839e-01 -4.96370554e-01
2.98207641e-01 -4.17704910e-01 -3.29125285e-01 -3.29051726e-02
2.51259685e-01 9.21808243e-01 4.03619528e-01 -1.25113344e+00
-6.00138187e-01 -2.32086897e-01 2.36788496e-01 4.66991037e-01
-2.15838596e-01 -1.93946827e-02 -6.30947590e-01 -6.04632199e-02
-1.75376579e-01 -1.01760340e+00 -2.59323955e-01 -4.61585552e-01
-4.21250194e-01 -9.38171208e-01 7.03899562e-01 -4.60167646e-01
7.15492010e-01 -2.12271714e+00 -1.07571699e-01 -1.51850134e-01
-1.76594686e-02 1.47943601e-01 -3.90094668e-01 9.86480236e-01
6.57253504e-01 2.91373264e-02 2.45338842e-01 -7.61009455e-01
5.59895635e-01 2.21324980e-01 -4.54108864e-01 -5.15313685e-01
1.39163420e-01 1.30392408e+00 -1.00985742e+00 -5.71123540e-01
1.40366942e-01 6.65895417e-02 -7.31784523e-01 5.65391123e-01
-8.76727343e-01 2.45494053e-01 -7.26375043e-01 6.93013430e-01
3.82580131e-01 -3.78869146e-01 2.82664031e-01 9.43876877e-02
2.43227005e-01 6.52624905e-01 -6.77541912e-01 2.09353805e+00
-9.58836317e-01 1.98623732e-01 5.61810017e-01 -3.82829666e-01
1.17109799e+00 2.50410020e-01 4.82136220e-01 -1.12120330e+00
-7.30280532e-03 1.70622587e-01 -3.69475126e-01 -7.27590978e-01
9.40604866e-01 2.55526006e-01 -6.77855551e-01 8.17351162e-01
2.69193083e-01 -5.50011158e-01 -1.33439824e-01 3.96503210e-01
1.09894049e+00 -1.71060205e-01 -1.74507439e-01 -3.66716236e-02
4.85197634e-01 3.79688114e-01 6.71718195e-02 1.07978117e+00
-2.34459594e-01 1.19256437e-01 1.18088186e-01 -3.57059240e-01
-5.84700525e-01 -9.51932609e-01 1.21467359e-01 2.04142976e+00
3.87439340e-01 -4.93959069e-01 -6.11234486e-01 -8.64287198e-01
3.00169498e-01 9.15476978e-01 -2.52894700e-01 1.77610163e-02
-4.55016643e-01 2.40306314e-02 4.87994701e-01 4.00660306e-01
8.15383732e-01 -1.49559438e+00 -2.41686031e-01 4.85199124e-01
-4.23756123e-01 -1.39055729e+00 -8.28666687e-01 -1.75342094e-02
-3.57366323e-01 -7.68305302e-01 -4.42163229e-01 -1.13898289e+00
-7.77331144e-02 3.38777184e-01 1.62579811e+00 -8.49162340e-02
-1.18339546e-01 8.21730793e-01 -6.14076912e-01 -3.47867340e-01
-8.93084824e-01 6.31945670e-01 -2.79334486e-01 -3.38550597e-01
9.24686730e-01 -1.39644757e-01 -8.64578903e-01 9.48713958e-01
-5.31484306e-01 -1.57969311e-01 5.23133218e-01 1.07301390e+00
-2.46343851e-01 -4.27707791e-01 1.06523693e+00 -9.92094815e-01
1.41851056e+00 -5.25171280e-01 -2.31088758e-01 3.47528666e-01
-9.68038261e-01 3.04442048e-02 2.53442377e-01 -4.15874392e-01
-1.63712299e+00 3.01956803e-01 -2.37945005e-01 1.55209854e-01
-3.43571395e-01 5.10626435e-01 4.97493893e-02 3.62202734e-01
7.40801513e-01 8.10035914e-02 3.96264225e-01 -7.26509571e-01
9.62876856e-01 1.26226008e+00 6.52747691e-01 -6.87913120e-01
1.11646771e-01 -1.66936666e-02 -8.55756700e-01 -7.18050003e-01
-6.03262842e-01 -1.04393482e+00 -4.38348889e-01 -4.02502082e-02
7.46328652e-01 -7.21801102e-01 -1.47316778e+00 1.53131530e-01
-1.14696252e+00 -6.40839815e-01 -1.69017360e-01 1.18426263e-01
-6.67448521e-01 3.24293137e-01 -1.11096907e+00 -1.01496685e+00
-6.17681623e-01 -1.19754839e+00 1.47419870e+00 5.81033565e-02
-6.71040833e-01 -9.38852608e-01 2.30674148e-02 1.01047373e+00
8.26339960e-01 -5.95988691e-01 8.50053787e-01 -1.44701922e+00
-4.90272284e-01 -2.80277491e-01 -1.60550233e-02 3.80618386e-02
2.52148416e-02 -6.53894544e-01 -8.90272677e-01 1.04647554e-01
1.22380137e-01 -1.12246597e+00 5.03147185e-01 -1.55626595e-01
5.27135849e-01 -4.83997196e-01 -2.32067868e-01 -1.30652115e-01
7.69948423e-01 3.37749153e-01 2.09668398e-01 2.95186251e-01
1.07010745e-01 1.24392927e+00 8.31303656e-01 4.54525590e-01
1.03585565e+00 1.01139772e+00 -3.36005422e-03 -5.36667220e-02
1.05080508e-01 -5.06248653e-01 1.89533785e-01 6.01846397e-01
6.74667537e-01 -2.89769828e-01 -6.20980740e-01 5.00794768e-01
-2.20138192e+00 -8.13641310e-01 2.90461063e-01 1.54677927e+00
1.21018636e+00 -1.30820721e-01 3.82864714e-01 -7.06284046e-01
2.91358232e-01 -1.34775639e-01 -8.78391743e-01 -6.90789282e-01
9.64768231e-02 1.02734469e-01 1.26988485e-01 7.01036632e-01
-1.02029622e+00 1.39866936e+00 6.39269495e+00 4.61014032e-01
-5.17376184e-01 1.32233724e-01 2.81274498e-01 3.83244604e-02
-3.09998065e-01 -2.02567264e-01 -8.41589987e-01 1.83866218e-01
9.36948299e-01 -2.20171753e-02 6.49327040e-01 1.31506479e+00
-2.70436198e-01 5.17057888e-02 -1.48531878e+00 8.09667826e-01
1.26634315e-01 -1.10593903e+00 -2.97956347e-01 -6.26148805e-02
3.25293809e-01 3.43810357e-02 6.91924393e-02 1.15066469e+00
1.16743135e+00 -8.68478298e-01 3.31606418e-01 7.78682679e-02
1.53061017e-01 -1.66478708e-01 9.02396858e-01 6.15737736e-01
-8.88790309e-01 -1.39528707e-01 1.99605212e-01 1.70103505e-01
6.46295667e-01 -3.10947061e-01 -1.51786482e+00 4.23566662e-02
5.98228157e-01 3.47358942e-01 -4.24320608e-01 2.73025364e-01
2.33212516e-01 3.97968106e-02 -3.42892557e-01 -5.67691684e-01
2.52790540e-01 -2.12902188e-01 2.72396833e-01 1.32821405e+00
-1.63692310e-01 8.16828981e-02 7.19913661e-01 1.16668844e+00
-1.57288790e-01 2.86819100e-01 -6.25380516e-01 -3.61662991e-02
6.79106236e-01 1.31674206e+00 -4.98186611e-02 -2.79948056e-01
-6.33090734e-01 1.20254290e+00 1.06922805e-01 3.37658733e-01
-2.36706004e-01 -1.99042276e-01 5.99406540e-01 -3.31808418e-01
3.27766210e-01 3.72071750e-02 -8.59211981e-02 -8.64806354e-01
7.96044692e-02 -1.31137192e+00 4.18232322e-01 -6.14223421e-01
-1.86466181e+00 7.08392441e-01 -6.65058270e-02 -6.36268437e-01
-1.07801545e+00 -3.49756718e-01 -3.42115432e-01 8.03893805e-01
-1.20127451e+00 -1.37404311e+00 -2.86455989e-01 4.82487410e-01
1.45852220e+00 -3.60867769e-01 1.13886058e+00 5.66533320e-02
-1.48187459e-01 7.89024413e-01 -2.44285509e-01 1.48917720e-01
1.00766480e+00 -1.39685857e+00 7.18597889e-01 -3.70289832e-01
-1.85568646e-01 9.61826026e-01 6.60008907e-01 -6.09305322e-01
-1.76334655e+00 -4.74559069e-01 9.90821362e-01 -8.82470310e-01
8.76009226e-01 -8.82617295e-01 -8.09170306e-01 6.32451296e-01
8.06263924e-01 -8.72151434e-01 9.07550275e-01 7.52264500e-01
-3.38850111e-01 2.74667084e-01 -1.22957981e+00 6.64589524e-01
1.07981288e+00 -7.87415028e-01 -7.75981486e-01 5.66820562e-01
1.20647013e+00 -4.32653040e-01 -1.17687428e+00 2.27067426e-01
6.56178176e-01 -7.19691932e-01 1.14643955e+00 -5.83406806e-01
2.48611137e-01 5.25340736e-01 -2.13248849e-01 -1.35832787e+00
1.99150011e-01 -1.02126992e+00 1.31223217e-01 1.55964756e+00
1.04748392e+00 -7.10947990e-01 9.20683861e-01 1.61735857e+00
-2.49561027e-01 -4.97963578e-01 -6.52577817e-01 -6.65621221e-01
3.58417593e-02 -1.50156617e-01 8.87895882e-01 7.47083664e-01
8.40763748e-01 1.13099778e+00 -5.21873832e-02 -3.90886545e-01
4.02038656e-02 3.58815938e-01 1.29709923e+00 -1.11574960e+00
-5.83570659e-01 -4.47413564e-01 3.43630552e-01 -1.81467378e+00
4.48187679e-01 -7.84721196e-01 4.94454533e-01 -1.36133134e+00
-8.79412666e-02 -5.60534775e-01 4.08510298e-01 3.16665798e-01
6.42759725e-02 -4.24941212e-01 2.88258284e-01 4.45555598e-01
-9.80157077e-01 7.16453373e-01 1.27985799e+00 -2.94606417e-01
-8.50301504e-01 1.50283217e-01 -9.08718884e-01 2.79680640e-01
6.07119620e-01 2.33510360e-01 -6.43269300e-01 -3.49025398e-01
5.93666770e-02 4.06263560e-01 -1.03441387e-01 -1.92391813e-01
6.06076300e-01 1.57911330e-01 -3.02331120e-01 -5.17913640e-01
7.97399998e-01 -7.56092012e-01 -3.61902237e-01 -1.00121975e-01
-1.12459791e+00 -2.17778590e-02 1.37198851e-01 6.57556951e-01
-2.31318474e-01 -2.70256609e-01 -9.76455584e-02 -4.03072745e-01
-7.36102283e-01 -2.03741595e-01 -4.08709377e-01 1.60842668e-02
5.97350776e-01 1.93183258e-01 -7.72218049e-01 -1.01612532e+00
-7.88984954e-01 9.43263590e-01 2.55467087e-01 1.01079071e+00
1.79251686e-01 -1.10185170e+00 -5.64800560e-01 2.55636722e-01
5.72870374e-01 -1.32828042e-01 -2.23952327e-02 3.08304518e-01
6.26920387e-02 8.26322079e-01 1.65751919e-01 -6.91840351e-01
-1.08497345e+00 6.36193812e-01 6.30545542e-02 -2.02411413e-01
-3.87825966e-01 8.47223759e-01 3.92373651e-03 -1.15870154e+00
5.38535833e-01 -1.99028671e-01 -3.05216819e-01 -1.20068043e-02
2.51027018e-01 7.99737796e-02 -5.31173758e-02 -3.91923517e-01
4.53426130e-02 6.35506287e-02 -5.41161656e-01 -6.17788613e-01
1.02821469e+00 -6.38053596e-01 3.32648069e-01 1.15399711e-01
1.15479612e+00 -1.98541448e-01 -1.08678484e+00 -8.05441797e-01
4.46123362e-01 -3.19820434e-01 -3.08720917e-01 -1.29296219e+00
-4.78660911e-01 4.01269466e-01 3.06902885e-01 9.15553510e-01
6.88225746e-01 6.36233389e-01 1.41058576e+00 1.07028949e+00
5.20142496e-01 -1.48881662e+00 4.33895111e-01 7.68541873e-01
1.06400764e+00 -1.74307001e+00 -6.46159232e-01 -6.75775051e-01
-1.39830399e+00 9.22547519e-01 9.11218345e-01 3.39747518e-01
5.46008229e-01 8.93256813e-03 5.83535671e-01 -6.16987884e-01
-1.06625473e+00 -4.00039345e-01 7.86381289e-02 5.88490725e-01
4.54315096e-01 -6.24199919e-02 -1.81238092e-02 9.62029278e-01
-5.18948555e-01 -1.23780213e-01 -4.78636324e-02 9.25089836e-01
-1.61898851e-01 -1.37249613e+00 2.52825826e-01 2.78142571e-01
-4.60918760e-03 -2.35711098e-01 -1.14149916e+00 7.01005638e-01
-6.31898999e-01 1.67482293e+00 1.19602285e-01 -5.69163263e-01
7.25194931e-01 6.56431377e-01 -4.25768085e-02 -7.15905905e-01
-1.08697748e+00 8.96924809e-02 8.83007228e-01 -7.74368286e-01
-1.02609552e-01 -6.47182882e-01 -1.03843760e+00 -2.25278065e-01
-6.16223156e-01 4.46256727e-01 7.49198616e-01 6.74471200e-01
9.75975811e-01 -2.08013102e-01 6.39375925e-01 -6.35785580e-01
-1.08441198e+00 -1.30660582e+00 -1.56716838e-01 7.82684982e-01
1.36061534e-01 -5.46734512e-01 -2.75128752e-01 4.36641872e-02] | [12.441850662231445, 7.90454626083374] |
107c8d32-89bb-426d-984f-05437e81467a | smaller3d-smaller-models-for-3d-semantic | 2305.03188 | null | https://arxiv.org/abs/2305.03188v1 | https://arxiv.org/pdf/2305.03188v1.pdf | Smaller3d: Smaller Models for 3D Semantic Segmentation Using Minkowski Engine and Knowledge Distillation Methods | There are various optimization techniques in the realm of 3D, including point cloud-based approaches that use mesh, texture, and voxels which optimize how you store, and how do calculate in 3D. These techniques employ methods such as feed-forward networks, 3D convolutions, graph neural networks, transformers, and sparse tensors. However, the field of 3D is one of the most computationally expensive fields, and these methods have yet to achieve their full potential due to their large capacity, complexity, and computation limits. This paper proposes the application of knowledge distillation techniques, especially for sparse tensors in 3D deep learning, to reduce model sizes while maintaining performance. We analyze and purpose different loss functions, including standard methods and combinations of various losses, to simulate the performance of state-of-the-art models of different Sparse Convolutional NNs. Our experiments are done on the standard ScanNet V2 dataset, and we achieved around 2.6\% mIoU difference with a 4 times smaller model and around 8\% with a 16 times smaller model on the latest state-of-the-art spacio-temporal convents based models. | ['Erik Harutyunyan', 'Alen Adamyan'] | 2023-05-04 | null | null | null | null | ['3d-semantic-segmentation'] | ['computer-vision'] | [-5.53435683e-01 -1.01631820e-01 2.16470689e-01 -3.30916107e-01
-2.41695136e-01 -1.44363597e-01 4.59944367e-01 -2.65704319e-02
-4.89180237e-01 3.52754831e-01 -1.16045969e-02 -4.61044639e-01
-2.61924773e-01 -1.10782957e+00 -1.00520253e+00 -3.43350142e-01
-6.21756554e-01 8.76762092e-01 5.01322627e-01 -2.16328233e-01
7.77137652e-02 1.07104611e+00 -1.55174077e+00 3.16158056e-01
5.22864997e-01 1.63393795e+00 -5.06433062e-02 4.98954207e-01
-6.45150185e-01 9.06000197e-01 -3.45653534e-01 -2.99858689e-01
5.69673717e-01 4.26577151e-01 -6.94228232e-01 -3.99145514e-01
8.32915485e-01 -1.47775471e-01 -7.86946535e-01 6.61603451e-01
6.26962245e-01 1.84386194e-01 3.76869529e-01 -9.19049084e-01
-6.05134308e-01 4.84668642e-01 -3.89803946e-01 4.26499337e-01
-2.08777294e-01 1.64722070e-01 5.63989878e-01 -1.05968833e+00
3.45049649e-01 1.48084867e+00 1.07691097e+00 1.90590695e-01
-1.01418078e+00 -7.42500126e-01 8.55840072e-02 1.68659627e-01
-1.56357455e+00 -3.67465675e-01 5.87824523e-01 -3.31481010e-01
1.86143970e+00 9.45222005e-02 1.00153458e+00 6.56518936e-01
3.77559900e-01 5.80066323e-01 9.04230237e-01 -1.79049551e-01
1.21172681e-01 -1.80084899e-01 -5.63761070e-02 9.97783005e-01
4.16309759e-02 1.49709553e-01 -8.26845825e-01 3.17303166e-02
1.12151051e+00 7.42447972e-02 8.71746540e-02 -4.40112174e-01
-1.10250735e+00 6.73256397e-01 8.44676197e-01 2.00104728e-01
-3.57610613e-01 6.45030141e-01 2.33390257e-01 2.65936524e-01
9.74661171e-01 3.18058491e-01 -6.23983145e-01 -1.06487699e-01
-1.13185787e+00 5.46437562e-01 1.00947654e+00 1.06177759e+00
7.69571424e-01 2.00506493e-01 -9.67739522e-02 7.02833831e-01
3.82383227e-01 5.78638375e-01 8.51187762e-03 -9.89233673e-01
6.79249763e-01 6.68459415e-01 -2.68430650e-01 -1.08844125e+00
-6.00315034e-01 -6.52374804e-01 -1.08212781e+00 5.04610240e-01
-8.28372082e-04 4.00048681e-02 -1.36020029e+00 1.23332715e+00
3.71606320e-01 5.86252868e-01 -3.70883048e-01 9.05788720e-01
1.14541090e+00 5.27699947e-01 -4.15411592e-02 6.18733883e-01
8.60547960e-01 -9.56692994e-01 -1.36874169e-01 -7.99818784e-02
6.01265430e-01 -6.75915837e-01 8.30995083e-01 2.43767932e-01
-1.42583644e+00 -4.55191970e-01 -9.16715562e-01 -3.83759379e-01
-5.17851651e-01 -2.65391886e-01 1.03343034e+00 5.08746743e-01
-1.59555161e+00 1.02034259e+00 -1.25490749e+00 -2.26183861e-01
7.91550517e-01 5.65579295e-01 -2.66116977e-01 -1.47714049e-01
-1.03806722e+00 1.13798463e+00 1.53322697e-01 2.46074513e-01
-1.06675482e+00 -1.30503988e+00 -6.95472896e-01 2.40417615e-01
-4.00766209e-02 -1.01581037e+00 9.80826497e-01 -9.46460217e-02
-1.32210207e+00 8.34075928e-01 -2.43363958e-02 -7.22169638e-01
2.95031160e-01 -2.85859466e-01 -9.48358327e-02 -1.03461467e-01
-1.60668135e-01 7.98529446e-01 5.60845554e-01 -8.47514629e-01
-3.24843228e-01 -4.93805379e-01 2.80496150e-01 2.68938869e-01
-1.12155497e-01 -2.90155083e-01 -6.26381338e-01 -3.28432202e-01
4.00572360e-01 -8.36705208e-01 -5.23350418e-01 4.39739108e-01
-3.41771066e-01 -1.56668395e-01 7.22167134e-01 -4.14294630e-01
7.62884140e-01 -2.10722136e+00 1.38909698e-01 3.99443179e-01
6.88452065e-01 1.28588140e-01 -1.16724037e-01 2.11186886e-01
1.23894557e-01 3.06035250e-01 -1.26347646e-01 -6.85075521e-01
1.12904189e-02 5.09337783e-01 1.01604722e-01 3.75718415e-01
3.18090856e-01 8.13296080e-01 -4.83674973e-01 -3.93200547e-01
3.70527595e-01 8.53727281e-01 -8.00271630e-01 -5.57357594e-02
-2.51478672e-01 1.11285895e-02 -6.26814544e-01 7.66691089e-01
7.75085866e-01 -5.89785039e-01 -2.83577502e-01 -4.14493978e-01
-3.41130197e-01 5.08375645e-01 -1.09129870e+00 2.06645513e+00
-5.54859400e-01 4.65172142e-01 1.90565616e-01 -1.11107099e+00
8.93044889e-01 1.11290663e-01 6.47936046e-01 -8.22184086e-01
1.00580141e-01 2.44006708e-01 -1.83091611e-01 -3.14716041e-01
4.56995577e-01 3.45055237e-02 4.45771307e-01 1.13024883e-01
1.55846640e-01 -5.63309908e-01 -2.52064671e-02 2.29236022e-01
1.37835705e+00 1.73515096e-01 -6.37187481e-01 -4.51264590e-01
1.22251948e-02 1.65318027e-01 1.15346730e-01 7.20374346e-01
1.36118561e-01 5.96560657e-01 4.46993381e-01 -8.50169241e-01
-1.11476541e+00 -9.64786470e-01 -1.22725822e-01 6.09860003e-01
-1.83271188e-02 -2.37483263e-01 -2.99149394e-01 -2.80598044e-01
4.63024944e-01 5.62911332e-01 -5.15810013e-01 8.12742412e-02
-7.24314094e-01 -7.42115438e-01 6.51791811e-01 5.42006671e-01
7.30140507e-01 -7.02335596e-01 -3.34325284e-01 1.81514710e-01
5.01971424e-01 -1.11299813e+00 1.96813303e-03 4.65902299e-01
-1.38488257e+00 -8.14922631e-01 -4.38779414e-01 -6.12741470e-01
4.10104305e-01 2.78667212e-01 1.54789960e+00 9.23612192e-02
6.41825572e-02 1.93022370e-01 -5.41133694e-02 -5.08551061e-01
2.28032619e-01 3.77632856e-01 -3.90419364e-02 -4.93736267e-01
1.61585033e-01 -1.08191562e+00 -6.62774682e-01 5.39988577e-02
-7.48287082e-01 1.30615428e-01 5.97218037e-01 4.45709199e-01
9.23422873e-01 2.00206786e-02 -2.76429802e-01 -7.21842706e-01
2.61954904e-01 -4.93132412e-01 -6.62007332e-01 -7.01911002e-02
-5.49705088e-01 2.93169916e-01 4.23611492e-01 -1.94727108e-01
-7.63257384e-01 -4.81666215e-02 -3.53895068e-01 -1.13501000e+00
7.34952390e-02 6.70754611e-01 3.23628008e-01 -7.85709679e-01
6.93121493e-01 -7.26805106e-02 1.40880972e-01 -7.19654202e-01
4.76376474e-01 -1.23004213e-01 1.02937639e-01 -7.61651218e-01
8.17978442e-01 6.44006312e-01 3.98552388e-01 -6.84769809e-01
-8.12930882e-01 -1.41292796e-01 -4.96967971e-01 -3.84093225e-02
6.67544007e-01 -1.05601203e+00 -7.25503683e-01 5.73832452e-01
-1.28988051e+00 -6.61977768e-01 -4.52622563e-01 5.42714536e-01
-3.00797522e-01 -3.18510905e-02 -7.36448884e-01 -3.00016820e-01
-5.49891174e-01 -1.25321794e+00 1.07616818e+00 -1.25640230e-02
3.76594007e-01 -8.90615404e-01 1.58145037e-02 1.56122804e-01
9.95752513e-01 2.10793480e-01 1.20940745e+00 -3.03355008e-01
-1.17764592e+00 -4.25992124e-02 -6.27382278e-01 3.99161041e-01
-4.98392284e-01 -1.41790763e-01 -8.82700086e-01 2.05312874e-02
-1.01802357e-01 -1.44004107e-01 8.20431292e-01 6.27844155e-01
1.55645740e+00 -9.48439986e-02 -4.47658509e-01 1.25735664e+00
1.46256435e+00 -1.10649429e-01 6.88448310e-01 5.70841953e-02
1.14367437e+00 1.46008253e-01 -1.14041507e-01 1.95986822e-01
6.81980729e-01 3.89281631e-01 8.95410776e-01 -1.87895164e-01
-6.01866007e-01 -1.26515492e-03 -2.58219957e-01 1.23753834e+00
-5.06061971e-01 -1.76579550e-01 -1.18785810e+00 5.66077054e-01
-1.67239201e+00 -7.07803249e-01 -2.21275389e-01 1.73703706e+00
5.86326003e-01 4.14867043e-01 -3.65794390e-01 -1.84758738e-01
1.77175969e-01 3.98496866e-01 -5.84197342e-01 -4.70869869e-01
-9.66699198e-02 8.76904309e-01 9.78216946e-01 2.27887139e-01
-9.84855711e-01 9.20187235e-01 6.82497835e+00 7.94201434e-01
-1.32067907e+00 2.13955805e-01 7.68758297e-01 -5.84277153e-01
-3.46193373e-01 -2.17030361e-01 -9.09086823e-01 8.75508562e-02
1.05181730e+00 1.01618968e-01 5.89738309e-01 9.85359013e-01
1.54164463e-01 1.04493208e-01 -1.16219342e+00 1.07530916e+00
-6.84637949e-02 -1.88585722e+00 1.12622365e-01 9.28839818e-02
8.07417929e-01 9.80765462e-01 5.41959144e-02 4.74024922e-01
2.94431776e-01 -1.26917577e+00 8.30251813e-01 7.35261559e-01
7.48280823e-01 -5.89790523e-01 5.83727479e-01 9.68845040e-02
-1.22380757e+00 4.94094878e-01 -5.83541095e-01 -2.52565980e-01
4.11430709e-02 9.84959245e-01 -5.99582016e-01 3.19956809e-01
1.31086743e+00 7.81415880e-01 -4.16810751e-01 1.18638968e+00
2.19553992e-01 4.29175347e-01 -1.00696325e+00 -3.08664531e-01
5.59755325e-01 2.12137047e-02 4.13086861e-01 8.88130486e-01
5.40711403e-01 7.51004517e-02 -5.22247702e-02 1.03371000e+00
-4.01875943e-01 -2.20851630e-01 -5.31533957e-01 1.02792367e-01
4.15674061e-01 8.73998225e-01 -4.28341836e-01 -1.16750650e-01
-5.88082790e-01 4.34557974e-01 5.58821321e-01 3.62280071e-01
-8.12843859e-01 -3.20029080e-01 1.00786507e+00 4.30821359e-01
5.39769292e-01 -7.70078719e-01 -6.60484612e-01 -9.39302146e-01
6.58641681e-02 -4.37568873e-01 1.56052234e-02 -8.14361572e-01
-1.39680183e+00 5.77813148e-01 3.17008682e-02 -8.11747074e-01
3.03186536e-01 -8.83815527e-01 -3.72607499e-01 1.06888378e+00
-1.79870033e+00 -1.03573930e+00 -3.70643497e-01 8.02082479e-01
1.53064400e-01 -9.95215252e-02 7.14596987e-01 8.36284816e-01
-1.60073757e-01 3.80183160e-01 -4.99143964e-03 2.27570906e-02
3.89139950e-02 -1.02650058e+00 8.98404121e-01 3.49667609e-01
9.78625864e-02 2.60191202e-01 1.68408096e-01 -3.96817416e-01
-1.85864449e+00 -1.04945767e+00 9.30256248e-01 -5.36557794e-01
7.15875089e-01 -3.96631181e-01 -9.31134999e-01 6.26261652e-01
-2.58587360e-01 4.75668132e-01 1.58313870e-01 3.42231631e-01
-2.46236265e-01 -3.12818795e-01 -1.11740446e+00 4.75767225e-01
1.48674703e+00 -3.78839403e-01 -1.91153660e-01 5.86984217e-01
1.18850243e+00 -1.04445434e+00 -1.31057978e+00 5.32323360e-01
4.80605245e-01 -9.84030485e-01 1.43474174e+00 -5.20767689e-01
4.36462015e-01 -1.25763133e-01 -2.62978405e-01 -1.13846958e+00
-4.74621594e-01 -1.18398391e-01 -3.81073445e-01 5.92970133e-01
5.66053689e-01 -5.67656159e-01 1.20985603e+00 6.14766955e-01
-7.60665655e-01 -1.14271688e+00 -1.28195882e+00 -6.15660489e-01
1.33007422e-01 -8.45064759e-01 9.06790018e-01 8.41579199e-01
-7.77108192e-01 5.21525219e-02 2.15787608e-02 2.79992193e-01
7.67962456e-01 -1.72114119e-01 5.73121011e-01 -1.17687309e+00
-9.83485803e-02 -5.16643703e-01 -5.89960337e-01 -1.06897151e+00
-1.30080685e-01 -1.04776418e+00 -3.41641635e-01 -1.79172683e+00
-3.52982998e-01 -1.12962854e+00 -3.13900441e-01 6.16235435e-01
4.72855002e-01 7.41181523e-02 1.77113935e-01 1.89206898e-01
-4.43041086e-01 7.09566176e-01 1.54303479e+00 -3.11386496e-01
-9.43640769e-02 -2.67789066e-01 -2.41635978e-01 6.23979151e-01
5.64948440e-01 -4.88692969e-01 -3.14221323e-01 -1.44992113e+00
3.73880506e-01 -6.63247928e-02 6.99080169e-01 -1.32837486e+00
4.75587904e-01 3.06174494e-02 5.26375711e-01 -9.59765136e-01
7.47506440e-01 -8.77886951e-01 3.95257324e-01 2.61281312e-01
9.64075699e-02 3.45778406e-01 6.69788182e-01 2.85235912e-01
-2.08652794e-01 2.49954164e-01 5.26167512e-01 -4.95394409e-01
-7.39872992e-01 9.85552132e-01 2.22659335e-01 4.86241654e-02
5.09893298e-01 -1.09827533e-01 -3.05828780e-01 5.28006665e-02
-7.33723700e-01 2.61469275e-01 9.96380821e-02 1.51396558e-01
7.59472251e-01 -1.30938733e+00 -7.10195780e-01 2.38390282e-01
-3.84908617e-01 6.45883024e-01 4.14942771e-01 7.73666978e-01
-1.21752179e+00 4.97325152e-01 -1.54457450e-01 -8.59392166e-01
-5.72826326e-01 2.27476135e-01 6.55527055e-01 -3.72196496e-01
-7.48124719e-01 1.28438187e+00 -1.55018792e-01 -6.51872277e-01
4.54192013e-01 -6.93691015e-01 1.62041649e-01 -5.00237346e-01
1.33248493e-02 3.52804273e-01 6.05820358e-01 -1.62317783e-01
-5.01471937e-01 6.14106059e-01 1.86653629e-01 3.44060272e-01
1.73468709e+00 3.16385120e-01 -2.54807353e-01 1.21707059e-01
1.18379807e+00 -5.90402663e-01 -1.07115722e+00 -3.42530876e-01
-4.62857306e-01 -4.08707947e-01 6.61537945e-01 -6.06882870e-01
-1.66232204e+00 1.06479216e+00 7.17228055e-01 2.29394257e-01
8.26325059e-01 1.16032541e-01 1.00699604e+00 4.80082393e-01
5.83989143e-01 -7.01147556e-01 -3.05059373e-01 1.02854395e+00
8.45801115e-01 -9.96593654e-01 2.12310612e-01 -3.25698286e-01
1.12424351e-01 8.38070571e-01 6.11950099e-01 -4.92498428e-01
1.34361875e+00 4.19870764e-01 -2.71908253e-01 -7.63343275e-01
-6.90758407e-01 -9.16993096e-02 3.07189882e-01 5.67400813e-01
3.56843919e-01 -5.61650321e-02 2.68768430e-01 2.54819673e-02
-3.55030715e-01 8.91756192e-02 -1.61431506e-01 7.68291414e-01
-1.35168701e-01 -7.52269447e-01 -1.46624088e-01 9.28898275e-01
-3.63304764e-01 -4.41669613e-01 2.47623429e-01 9.14264500e-01
1.90727502e-01 3.67303282e-01 2.87849993e-01 -5.54648757e-01
7.18818128e-01 -2.03685626e-01 6.98731363e-01 -5.65340221e-01
-7.43541479e-01 -3.54496211e-01 1.44182682e-01 -9.57560658e-01
-6.24739885e-01 -3.02482486e-01 -1.16023552e+00 -1.02702618e+00
1.55671258e-02 -2.72337496e-01 1.08598697e+00 7.28639364e-01
7.25853920e-01 6.29042387e-01 1.53318807e-01 -1.31732118e+00
-3.30608964e-01 -7.22884715e-01 -5.90899348e-01 -5.48343349e-04
-8.73311833e-02 -7.97288299e-01 -2.75292605e-01 -4.58844572e-01] | [7.937671184539795, -3.6251511573791504] |
981cfe80-2a84-494f-87b1-9665d073ead9 | teaching-the-pre-trained-model-to-generate | 2305.12463 | null | https://arxiv.org/abs/2305.12463v1 | https://arxiv.org/pdf/2305.12463v1.pdf | Teaching the Pre-trained Model to Generate Simple Texts for Text Simplification | Randomly masking text spans in ordinary texts in the pre-training stage hardly allows models to acquire the ability to generate simple texts. It can hurt the performance of pre-trained models on text simplification tasks. In this paper, we propose a new continued pre-training strategy to teach the pre-trained model to generate simple texts. We continue pre-training BART, a representative model, to obtain SimpleBART. It consistently and significantly improves the results on lexical simplification, sentence simplification, and document-level simplification tasks over BART. At the end, we compare SimpleBART with several representative large language models (LLMs). | ['Xiaojun Wan', 'Wei Xu', 'Renliang Sun'] | 2023-05-21 | null | null | null | null | ['lexical-simplification'] | ['natural-language-processing'] | [ 4.26546931e-01 4.29577023e-01 6.31404668e-02 -4.44579095e-01
-7.35589802e-01 -2.82930583e-01 5.87819099e-01 1.61721319e-01
-5.69833040e-01 8.35442662e-01 4.01005626e-01 -6.84589624e-01
2.13162512e-01 -5.25764525e-01 -7.29066670e-01 -1.41091123e-01
5.44424653e-01 6.65087402e-01 -7.94673860e-02 -5.09944499e-01
8.33963975e-02 4.05121505e-01 -1.04429519e+00 4.40613687e-01
1.58427775e+00 -6.53980970e-02 5.24092317e-01 8.00958633e-01
-2.62675583e-01 9.83347952e-01 -1.10968471e+00 -8.58210564e-01
7.92155713e-02 -5.15268445e-01 -6.99643135e-01 -3.24738115e-01
7.97284782e-01 -4.05386090e-01 -6.16112232e-01 8.65227938e-01
5.73295891e-01 3.45734119e-01 8.71721208e-01 -6.64709330e-01
-7.17388630e-01 1.44368565e+00 -2.59237587e-01 2.34113127e-01
2.29184508e-01 6.71020672e-02 8.13082159e-01 -7.79043198e-01
9.36866641e-01 1.73410213e+00 8.76771808e-01 8.39628756e-01
-1.20190644e+00 -6.66091681e-01 2.75924712e-01 -1.18425034e-01
-1.27179503e+00 -6.59546614e-01 4.18229371e-01 -2.70181410e-02
1.45486403e+00 5.75118303e-01 5.70290804e-01 7.66610026e-01
5.17874122e-01 9.11645651e-01 4.89797533e-01 -7.13855803e-01
-3.21280152e-01 -1.49068803e-01 2.28701934e-01 4.91997182e-01
6.17433071e-01 -2.15683341e-01 -2.34451190e-01 1.87519372e-01
2.57237256e-01 -1.83100581e-01 -8.79849643e-02 2.18687907e-01
-8.47986460e-01 7.18224943e-01 6.50937781e-02 4.27030504e-01
-5.36016934e-02 6.96842074e-02 6.05916619e-01 5.48266172e-01
5.10123909e-01 1.02350557e+00 -4.70378429e-01 -9.10262689e-02
-1.17722988e+00 5.02778292e-01 8.16607356e-01 1.41709733e+00
5.75103521e-01 6.12670660e-01 -6.58343911e-01 9.59803998e-01
-3.86746764e-01 7.00329125e-01 5.55722237e-01 -7.53653347e-01
8.48741889e-01 4.02621269e-01 -2.49571681e-01 -5.13709188e-01
-3.86623442e-01 -3.59774560e-01 -1.18084812e+00 -3.31760287e-01
1.78140223e-01 -3.73689175e-01 -1.08485484e+00 1.61767507e+00
-1.20288253e-01 -4.33394849e-01 3.39554884e-02 1.43596590e-01
9.01747048e-01 8.77106369e-01 2.59488046e-01 -4.37415361e-01
8.15144300e-01 -1.31684864e+00 -1.23556960e+00 -4.18048710e-01
1.23037350e+00 -1.08724487e+00 1.53332210e+00 3.95806730e-01
-1.40114570e+00 -6.59581125e-01 -1.11202288e+00 -7.81906486e-01
-4.53007579e-01 3.13134491e-01 5.21184862e-01 4.71082985e-01
-9.25977468e-01 9.47110057e-01 -5.18177032e-01 -1.83717459e-01
4.83820319e-01 2.00359076e-01 -1.67101592e-01 -3.60661507e-01
-1.31872606e+00 1.35469079e+00 4.67596561e-01 -2.11460322e-01
-7.44638443e-01 -1.08140182e+00 -1.02698064e+00 2.50366628e-01
2.47993246e-01 -1.02835941e+00 1.79636240e+00 -5.62108517e-01
-1.66066265e+00 5.36970258e-01 -5.29682994e-01 -6.05687618e-01
5.65388560e-01 -7.22256184e-01 -1.69622988e-01 -4.18445855e-01
-7.09665418e-02 7.86840558e-01 1.01011586e+00 -9.66952801e-01
-4.29268867e-01 2.56553799e-01 -2.74081714e-03 2.79345840e-01
-2.06774741e-01 1.26059473e-01 -3.29328895e-01 -1.17501426e+00
-6.19730473e-01 -8.69074702e-01 -2.82928407e-01 -7.63487220e-01
-8.90658855e-01 -4.12742406e-01 5.81145048e-01 -9.35650945e-01
1.90733707e+00 -1.85250902e+00 3.47452879e-01 -1.93797559e-01
2.11818472e-01 8.56214762e-01 -4.97229755e-01 6.41106129e-01
-9.41076875e-02 7.21895218e-01 -2.38140672e-01 -1.07794750e+00
5.06857820e-02 4.55648005e-01 -5.20068944e-01 -4.11758274e-02
1.84782431e-01 1.14577568e+00 -7.46048570e-01 -6.63884282e-01
3.87721151e-01 2.02576786e-01 -9.27473903e-01 1.27985269e-01
-3.45351666e-01 1.33344054e-01 4.17799428e-02 1.67702824e-01
5.71624100e-01 3.91683996e-01 1.02119587e-01 1.45746499e-01
-1.40918158e-02 7.90237188e-01 -5.43085635e-01 1.59777689e+00
-8.64211679e-01 1.03858793e+00 -3.77182186e-01 -5.51219583e-01
5.34429967e-01 1.59865469e-01 -2.89552152e-01 -7.27974713e-01
1.98829994e-01 5.38729429e-02 4.80649501e-01 -2.97982007e-01
9.17809844e-01 -1.73802614e-01 -1.63425565e-01 5.21610141e-01
1.21673808e-01 -8.50879490e-01 8.36286366e-01 5.47746241e-01
7.78402686e-01 -1.70720160e-01 3.80885065e-01 -2.15652257e-01
5.24707913e-01 -1.75960213e-01 3.21120203e-01 1.13066983e+00
3.87315273e-01 6.34242058e-01 5.03975272e-01 -2.82982647e-01
-1.40401733e+00 -7.56106973e-01 1.23164304e-01 1.40698850e+00
-4.00098115e-01 -1.21890044e+00 -1.14710259e+00 -7.09107220e-01
1.23402327e-02 1.62190032e+00 -3.50442648e-01 -3.96466047e-01
-1.21532178e+00 -4.19245541e-01 8.05904150e-01 2.68286914e-01
7.01375529e-02 -1.11204493e+00 -7.00061172e-02 9.49990600e-02
-2.06170112e-01 -9.86974359e-01 -8.29964280e-01 2.33476847e-01
-9.24454808e-01 -6.05415940e-01 -4.87875640e-01 -8.98344636e-01
7.65148282e-01 1.73803821e-01 1.35021126e+00 4.23689842e-01
-7.41886348e-02 -5.36793411e-01 -3.29111695e-01 -9.45170701e-01
-9.82403994e-01 8.80431831e-01 3.47803980e-02 -8.21697295e-01
5.47501594e-02 -3.83182764e-01 7.02669322e-02 -1.02683753e-01
-1.16966844e+00 4.19387758e-01 6.76593065e-01 7.80315161e-01
4.32856679e-01 1.33029213e-02 4.35904175e-01 -1.40903711e+00
1.12450409e+00 1.78970799e-01 -2.16325328e-01 3.53678226e-01
-6.19925976e-01 2.83787727e-01 1.32700443e+00 -7.06882000e-01
-1.18924356e+00 -4.06979352e-01 -5.30380964e-01 -1.16396978e-01
8.07489008e-02 5.83111048e-01 -2.68850952e-01 1.07961878e-01
6.60469115e-01 7.60350227e-02 -3.36442709e-01 -9.37161863e-01
6.95687771e-01 4.75994289e-01 3.72685939e-01 -4.96041000e-01
1.18402255e+00 -1.22616455e-01 -3.33498359e-01 -8.87019992e-01
-1.25009775e+00 1.10975534e-01 -7.48043299e-01 4.06363010e-01
2.63508022e-01 -7.07646668e-01 1.14556573e-01 4.62848514e-01
-1.50186849e+00 -8.50029230e-01 -5.58814049e-01 1.21106148e-01
-2.58071989e-01 5.28964162e-01 -7.42662191e-01 -3.26725721e-01
-8.05349231e-01 -9.60327625e-01 9.07250106e-01 1.84158772e-01
-7.90571511e-01 -1.03872156e+00 7.54042938e-02 -3.54737975e-02
5.99882007e-01 -2.25147605e-01 1.27832770e+00 -7.39139557e-01
-6.34912848e-02 -2.39004195e-01 1.23733349e-01 6.45773053e-01
9.96058136e-02 3.59998137e-01 -6.29245162e-01 -2.60105669e-01
-8.04218724e-02 -2.00325623e-01 9.67975438e-01 3.76060009e-01
1.29980624e+00 -6.58525109e-01 -2.06237435e-01 1.01423168e+00
7.08092511e-01 7.13454112e-02 9.12258983e-01 2.75372744e-01
1.03582847e+00 2.81255424e-01 4.89057630e-01 1.24821424e-01
2.54097044e-01 4.55244362e-01 -3.82681519e-01 -2.59817362e-01
-6.77210808e-01 -5.94829261e-01 4.43056285e-01 1.42817473e+00
2.21051753e-01 -5.14745951e-01 -7.53303707e-01 2.48950228e-01
-1.52679980e+00 -8.66569400e-01 -1.62838221e-01 1.62934554e+00
1.48378944e+00 4.87201333e-01 -3.14880133e-01 1.80490583e-01
5.71119666e-01 2.01801807e-01 -4.48194623e-01 -1.12921083e+00
-3.76312554e-01 5.44921398e-01 2.89836705e-01 1.11411691e+00
-9.78763402e-01 1.70058620e+00 7.09477234e+00 1.42086601e+00
-9.15748239e-01 -1.45384306e-02 4.02442753e-01 -3.81196916e-01
-5.62193155e-01 -5.80962524e-02 -1.14910305e+00 2.58970946e-01
1.13935840e+00 -7.75701165e-01 5.65988660e-01 6.63814902e-01
4.37899530e-01 1.55173331e-01 -1.33572555e+00 7.46059179e-01
2.12725282e-01 -1.09810507e+00 1.04230833e+00 -4.45038795e-01
1.01875114e+00 -2.01652586e-01 -8.59019160e-02 1.05898809e+00
3.80532086e-01 -1.36733687e+00 7.99345732e-01 4.71036345e-01
9.38703299e-01 -9.30014849e-01 5.90064287e-01 7.15991497e-01
-7.16051102e-01 2.02619746e-01 -7.29616761e-01 -4.48939323e-01
3.87676984e-01 6.68175519e-01 -9.32429910e-01 3.93941760e-01
-3.61893326e-05 6.25774622e-01 -1.08213758e+00 9.23647583e-01
-7.44254827e-01 7.35453725e-01 -3.37482065e-01 -2.85055637e-02
7.90564157e-03 -3.38158980e-02 6.39404774e-01 1.70665383e+00
1.99162155e-01 3.54696512e-02 -1.50207788e-01 4.88585263e-01
-4.55373526e-01 2.88947344e-01 -5.98366976e-01 -1.86790898e-01
5.65386653e-01 9.35203135e-01 -1.88787624e-01 -8.64946663e-01
-1.18981645e-01 7.74726033e-01 5.61128259e-01 4.42063272e-01
-7.26528764e-01 -1.10622096e+00 5.14415026e-01 -7.74083659e-02
2.52648950e-01 -2.61532843e-01 -7.86575437e-01 -1.41743946e+00
-2.00954691e-01 -1.38618350e+00 1.63490660e-02 -7.83584118e-01
-9.19385970e-01 6.76676631e-01 1.83000058e-01 -6.90631390e-01
-5.04143000e-01 -2.91847914e-01 -7.74867773e-01 1.07388353e+00
-1.51581097e+00 -9.66331363e-01 6.08710833e-02 3.11821699e-01
1.00985193e+00 -3.38590406e-02 7.58107066e-01 3.22035372e-01
-9.99189079e-01 1.20812833e+00 2.70379364e-01 2.12723911e-02
9.59341288e-01 -1.28678441e+00 1.16470361e+00 1.22825241e+00
3.56764942e-02 1.12261176e+00 1.09837806e+00 -9.16555285e-01
-1.02765656e+00 -1.25443876e+00 1.51351261e+00 -5.57123005e-01
4.29524064e-01 -5.56684434e-01 -1.03740740e+00 9.94558096e-01
4.38203990e-01 -8.54533017e-01 4.24367726e-01 1.38625681e-01
-1.17584988e-01 -1.49229601e-01 -7.20935822e-01 1.19554520e+00
1.00711763e+00 -3.66389900e-01 -1.10861778e+00 3.66139501e-01
1.25691903e+00 -7.51663446e-01 -5.29136121e-01 2.65984833e-01
2.14143366e-01 -4.25817490e-01 7.86231935e-01 -7.99610138e-01
5.59818625e-01 2.32287303e-01 4.09348965e-01 -1.83956003e+00
-3.63228410e-01 -1.32481647e+00 -2.68152893e-01 1.31966329e+00
6.58582568e-01 -3.01374972e-01 4.37488794e-01 4.61617589e-01
-6.89853787e-01 -4.90681767e-01 -5.80431104e-01 -8.16227913e-01
8.15459728e-01 -3.79472911e-01 7.54676104e-01 6.63606346e-01
2.44813263e-02 8.06597650e-01 -5.33966303e-01 -5.73113322e-01
2.79827088e-01 -2.83970028e-01 1.12257004e+00 -1.09191656e+00
2.83543933e-02 -6.97922945e-01 4.83914793e-01 -1.37276149e+00
3.48936379e-01 -1.05344808e+00 5.64324558e-02 -1.64117599e+00
5.34161646e-03 -2.44587868e-01 3.28487009e-01 5.59772909e-01
-6.79859757e-01 -1.62414193e-01 5.64975739e-01 7.80590251e-02
-5.97235620e-01 7.46924162e-01 1.69408822e+00 -2.01395497e-01
-4.40841347e-01 6.78269845e-03 -1.03092360e+00 7.62922168e-01
1.05272651e+00 -7.87017584e-01 -4.55032468e-01 -1.01520252e+00
2.24163979e-01 -3.35067987e-01 -4.16606575e-01 -8.07661295e-01
-5.30074462e-02 -1.07076868e-01 2.63597250e-01 -6.22777462e-01
-5.34249097e-02 -2.53528357e-01 -1.38813481e-01 5.08619130e-01
-6.72900379e-01 3.14005077e-01 5.90660036e-01 -6.55006394e-02
-5.52169681e-02 -6.43069267e-01 8.09300721e-01 -1.38558066e-02
-2.73101404e-02 6.69520721e-02 -6.83110714e-01 4.66941357e-01
3.79043162e-01 8.24497417e-02 -4.32801723e-01 -3.91312391e-01
-4.50048953e-01 3.04351807e-01 5.75098157e-01 3.87137055e-01
3.51652205e-01 -9.35391784e-01 -7.67596304e-01 1.73984751e-01
-4.16807532e-01 3.68161976e-01 8.81973431e-02 5.36220729e-01
-8.87954712e-01 8.75188351e-01 3.54275778e-02 -3.61695886e-02
-1.35141575e+00 5.67767918e-01 2.80224353e-01 -7.77028680e-01
-6.59861684e-01 7.00361133e-01 1.79896742e-01 -6.07725263e-01
4.53462541e-01 -9.09618258e-01 -8.32749307e-02 -1.42474607e-01
7.41473377e-01 3.34143341e-01 3.12537402e-01 -2.50245959e-01
1.66801482e-01 2.53965735e-01 -7.45049894e-01 1.57509565e-01
1.21451569e+00 -2.91094601e-01 -1.82441875e-01 2.58443207e-01
1.03400934e+00 2.96759486e-01 -6.42830491e-01 -1.55197293e-01
4.89527136e-02 -1.08467974e-01 -1.15806788e-01 -8.53872061e-01
-5.70163012e-01 1.15019584e+00 -3.58559340e-01 9.64175463e-02
9.55454826e-01 -4.73249465e-01 1.24877000e+00 1.11227274e+00
-1.30509257e-01 -1.11861849e+00 -1.80683345e-01 1.33866823e+00
1.06156397e+00 -8.41405392e-01 3.00506592e-01 -4.09036934e-01
-7.01618195e-01 1.03250992e+00 6.87610269e-01 -1.19702615e-01
2.63624132e-01 4.76309031e-01 1.36163272e-03 2.98314303e-01
-1.00560760e+00 -2.41698176e-02 3.47233802e-01 5.37390351e-01
6.71359956e-01 -9.33382958e-02 -5.81459761e-01 7.22132742e-01
-1.01416945e+00 -1.40547454e-01 7.01959431e-01 7.35890687e-01
-5.38826346e-01 -1.26263642e+00 -1.93709627e-01 6.18192911e-01
-5.66521347e-01 -7.15966105e-01 -7.01200545e-01 1.06766987e+00
1.91399492e-02 7.12505937e-01 -8.66248012e-02 -2.39930496e-01
7.09652662e-01 3.04087162e-01 5.08513212e-01 -1.04239857e+00
-8.85326266e-01 2.77082007e-02 4.19744462e-01 -7.82492310e-02
3.09266329e-01 -4.63057816e-01 -1.04526162e+00 -8.65713894e-01
-2.97619402e-01 1.67710707e-01 2.19800934e-01 1.00487351e+00
1.90629989e-01 8.28027368e-01 1.45755529e-01 -1.05280352e+00
-6.99941337e-01 -1.50234890e+00 -2.47004926e-01 5.10341048e-01
4.23163295e-01 -1.80730745e-01 -3.68743420e-01 2.12724343e-01] | [11.14171028137207, 10.307579040527344] |
a54ef58a-fd0b-4c6f-bdd2-682faee3c21e | envgan-adversarial-synthesis-of-environmental | 2104.07326 | null | https://arxiv.org/abs/2104.07326v1 | https://arxiv.org/pdf/2104.07326v1.pdf | EnvGAN: Adversarial Synthesis of Environmental Sounds for Data Augmentation | The research in Environmental Sound Classification (ESC) has been progressively growing with the emergence of deep learning algorithms. However, data scarcity poses a major hurdle for any huge advance in this domain. Data augmentation offers an excellent solution to this problem. While Generative Adversarial Networks (GANs) have been successful in generating synthetic speech and sounds of musical instruments, they have hardly been applied to the generation of environmental sounds. This paper presents EnvGAN, the first ever application of GANs for the adversarial generation of environmental sounds. Our experiments on three standard ESC datasets illustrate that the EnvGAN can synthesize audio similar to the ones in the datasets. The suggested method of augmentation outshines most of the futuristic techniques for audio augmentation. | ['Suresh K', 'Aswathy Madhu'] | 2021-04-15 | null | null | null | null | ['environmental-sound-classification', 'sound-classification'] | ['audio', 'audio'] | [ 4.50451195e-01 1.69716805e-01 5.91753483e-01 1.86967269e-01
-7.79609919e-01 -5.40073276e-01 5.97260535e-01 -6.16349816e-01
-1.02487184e-01 9.99414742e-01 2.58538902e-01 -2.98943460e-01
2.39895210e-01 -1.08802772e+00 -5.85303187e-01 -1.00326419e+00
1.47042572e-01 1.63025588e-01 -1.06286794e-01 -5.93643665e-01
-1.38437912e-01 2.85240442e-01 -1.57688689e+00 2.41926551e-01
5.63700020e-01 7.35725939e-01 -1.86907593e-02 1.02549767e+00
-3.09926331e-01 7.57889926e-01 -1.28395128e+00 -6.57703519e-01
4.48194087e-01 -9.99152362e-01 -4.48358953e-01 -4.54089820e-01
1.24065220e-01 -1.42757803e-01 -4.81216431e-01 1.02343631e+00
1.18372035e+00 1.78993389e-01 6.12511992e-01 -1.63098133e+00
-6.97305918e-01 9.97093558e-01 -1.83134720e-01 3.09634563e-02
1.63823679e-01 2.25951836e-01 7.10049033e-01 -4.62655306e-01
2.74365693e-01 1.00850964e+00 7.48268723e-01 1.01901722e+00
-8.98016095e-01 -9.62859988e-01 -6.11150384e-01 -1.23443007e-01
-1.08367884e+00 -4.13173437e-01 1.12151933e+00 -2.40569130e-01
4.59257066e-01 5.04420400e-01 7.97517300e-01 1.67242682e+00
-1.24827817e-01 5.76156437e-01 1.29616666e+00 -5.16195714e-01
3.36734682e-01 1.70298234e-01 -7.80812979e-01 1.75360262e-01
-4.23443057e-02 5.78297973e-01 -7.45236039e-01 -6.93171024e-02
8.27981830e-01 -5.97351134e-01 -1.18768416e-01 4.25200462e-01
-9.28643167e-01 7.81333148e-01 2.68269449e-01 2.89806455e-01
-3.31260204e-01 5.81618369e-01 4.98505563e-01 5.29689252e-01
3.82160991e-01 8.12419415e-01 -1.22407833e-02 -3.53202015e-01
-8.47049713e-01 5.19739866e-01 5.86136401e-01 7.96477020e-01
1.33422375e-01 1.24471962e+00 5.65562360e-02 7.66168773e-01
-9.32251960e-02 6.62387848e-01 8.11479509e-01 -8.32347512e-01
3.61607105e-01 -2.86816712e-02 -6.46158010e-02 -1.02320683e+00
-4.05050479e-02 -6.88568354e-01 -1.16204441e+00 4.37427342e-01
3.09504658e-01 -5.33412516e-01 -8.97823393e-01 1.80771434e+00
1.47624582e-01 5.55896342e-01 3.74735147e-01 6.02358162e-01
1.14471185e+00 8.64357293e-01 -1.74183026e-02 2.23757014e-01
9.43958819e-01 -8.77758622e-01 -8.33897412e-01 -1.43116236e-01
-1.77267522e-01 -8.05015326e-01 1.07689083e+00 5.15019655e-01
-7.38052547e-01 -9.13924634e-01 -1.05940831e+00 3.29437047e-01
-4.59911168e-01 -2.86164999e-01 7.89320230e-01 1.31037354e+00
-9.04630363e-01 3.62265319e-01 -3.33078384e-01 1.59945309e-01
4.03873295e-01 2.52557963e-01 -1.51114672e-01 6.07900321e-01
-1.44979095e+00 4.55550015e-01 2.56661654e-01 1.27697974e-01
-1.27737224e+00 -7.62493134e-01 -4.11725879e-01 -3.48789133e-02
9.63519290e-02 -7.03894556e-01 1.15680468e+00 -1.19467413e+00
-1.94600797e+00 4.53419238e-01 6.03151500e-01 -7.54238605e-01
7.92711318e-01 -3.02019507e-01 -7.94894755e-01 -2.08439857e-01
-1.44846976e-01 7.03466296e-01 1.17173696e+00 -1.28078043e+00
-3.42064738e-01 2.01159447e-01 -1.36436895e-01 -1.19970322e-01
-4.53666329e-01 -8.97000730e-02 2.79211581e-01 -1.21119845e+00
-3.79475176e-01 -8.88005078e-01 -3.94990414e-01 -4.95491326e-01
-7.35597551e-01 2.14121610e-01 8.24554741e-01 -4.63302106e-01
9.09768760e-01 -2.15535760e+00 -3.94451618e-02 1.19806477e-03
-2.27699727e-01 4.04297024e-01 -2.85165548e-01 6.13683522e-01
-2.73209333e-01 4.81793553e-01 -3.67079049e-01 -3.44674349e-01
5.32337204e-02 1.38777494e-01 -9.51874912e-01 5.70605882e-03
1.89919680e-01 8.51917624e-01 -8.92341435e-01 -9.48181823e-02
2.29028761e-02 7.37506807e-01 -5.20249546e-01 4.08996224e-01
-3.27502012e-01 9.55770791e-01 -3.23029250e-01 3.81455272e-01
4.30877566e-01 5.54322124e-01 -2.12499216e-01 3.30482125e-01
-2.11346447e-02 1.15565591e-01 -1.16511858e+00 1.54389286e+00
-5.16236603e-01 7.97234058e-01 -1.72091439e-01 -8.42484355e-01
1.19494665e+00 6.66420460e-01 3.68525773e-01 -4.90574419e-01
1.53198525e-01 3.41932893e-01 1.48347825e-01 -4.91894662e-01
4.22943771e-01 -5.93237996e-01 -3.01879108e-01 3.09903085e-01
1.47469128e-02 -7.48144209e-01 -2.23044619e-01 -1.26158625e-01
7.99666643e-01 3.38627934e-01 8.90963599e-02 2.06292063e-01
3.21497232e-01 -1.57699227e-01 3.82638901e-01 6.67547524e-01
1.96541429e-01 8.99976254e-01 2.32384145e-01 -3.55175972e-01
-1.27773178e+00 -9.02079284e-01 3.00531536e-01 7.96389401e-01
-3.87466162e-01 -1.86208159e-01 -1.06188631e+00 -4.41229314e-01
-2.12008983e-01 7.56288826e-01 -5.19908905e-01 -2.52189517e-01
-6.53747857e-01 -7.66760170e-01 1.44133651e+00 5.02202570e-01
7.16892242e-01 -1.63256383e+00 -6.27597272e-01 3.50176096e-01
-1.88319623e-01 -9.75129008e-01 3.39264758e-02 2.10866496e-01
-5.65001309e-01 -6.83964193e-01 -7.32615590e-01 -6.02759659e-01
1.14084870e-01 -1.81170434e-01 1.00830829e+00 -1.31568611e-01
-2.99531847e-01 2.01480716e-01 -5.47995448e-01 -1.34939182e+00
-1.31801581e+00 1.74383536e-01 1.55761093e-01 2.25495268e-02
-8.29680935e-02 -1.09613049e+00 -4.69974637e-01 -4.88424767e-03
-1.16002178e+00 -3.58542427e-02 3.03272486e-01 8.39908421e-01
3.79986316e-01 3.86321187e-01 1.34782004e+00 -1.04730523e+00
8.55240941e-01 -5.47970176e-01 -2.96472937e-01 -2.09677190e-01
-2.47514784e-01 -1.49013072e-01 1.06783402e+00 -4.24097449e-01
-1.00704479e+00 -9.58058685e-02 -8.32433999e-01 -2.59599417e-01
-4.94718373e-01 2.09045455e-01 -2.03639895e-01 9.82333049e-02
8.51772249e-01 2.27811649e-01 -2.22551405e-01 -5.92672110e-01
3.46203953e-01 8.75243306e-01 8.66722286e-01 -5.31211257e-01
1.14179695e+00 2.70879209e-01 2.23288447e-01 -1.06873035e+00
-5.00272453e-01 3.54924709e-01 -2.17945755e-01 -2.17868909e-01
8.35310221e-01 -7.27530777e-01 -4.04862344e-01 7.68088877e-01
-8.00454795e-01 -3.87946934e-01 -1.01714838e+00 4.88263257e-02
-6.80151343e-01 1.32053271e-01 -1.81862190e-01 -9.98924434e-01
-5.31909943e-01 -9.41125453e-01 7.14747429e-01 2.10286692e-01
-1.09948449e-01 -7.15685368e-01 3.55617702e-01 2.30588168e-01
7.84866691e-01 8.51587296e-01 7.87715912e-01 -6.08150959e-01
-3.25347006e-01 -3.52068752e-01 6.45645440e-01 8.15325558e-01
2.78505176e-01 1.42199561e-01 -1.52209568e+00 9.26771611e-02
2.19245955e-01 -2.19962820e-01 7.48615801e-01 -6.18113158e-03
1.21281052e+00 -2.91143715e-01 3.50765020e-01 6.64327621e-01
1.21902323e+00 6.02890611e-01 1.03268683e+00 3.35646659e-01
6.08349919e-01 1.90151468e-01 2.85737842e-01 3.29046041e-01
-2.53688902e-01 5.20425618e-01 7.46837258e-01 -1.83994308e-01
-5.86671233e-01 -6.82311833e-01 4.23085332e-01 9.27214742e-01
-4.84992206e-01 -5.66777408e-01 -7.85369456e-01 5.00015855e-01
-1.12334645e+00 -1.19370377e+00 -1.47220269e-01 1.89286149e+00
8.05305362e-01 -7.00641647e-02 2.40724161e-01 9.00740743e-01
6.30700767e-01 3.27555239e-01 -2.97443718e-01 -5.59596896e-01
-3.34577292e-01 9.03860271e-01 -9.56934616e-02 6.43413961e-02
-1.18577528e+00 8.38886619e-01 6.54728794e+00 8.27535748e-01
-1.28597379e+00 1.06809683e-01 5.56763947e-01 1.20714165e-01
-4.54342753e-01 -4.20308650e-01 -4.14444685e-01 7.13907123e-01
1.06912875e+00 -4.04642001e-02 5.15673459e-01 8.96975577e-01
-5.02362959e-02 3.27223510e-01 -7.40771234e-01 7.88999677e-01
-2.66576678e-01 -1.22657657e+00 3.21722746e-01 -1.97965339e-01
1.10565686e+00 -2.82319635e-01 5.93192399e-01 2.91522503e-01
2.76870668e-01 -1.34787190e+00 8.74701858e-01 3.71046990e-01
1.04568839e+00 -1.10218489e+00 8.68821025e-01 1.83248550e-01
-7.67642617e-01 -1.13304839e-01 -2.63822377e-01 -1.17831184e-02
2.17066556e-01 4.12284255e-01 -1.02559984e+00 5.64046562e-01
6.10005260e-01 -2.85194814e-02 -2.41123170e-01 9.07493114e-01
-4.03656721e-01 1.14135385e+00 -1.77722827e-01 -1.27362488e-02
2.58042216e-01 -1.39088273e-01 9.62905347e-01 1.07419693e+00
7.51021862e-01 -1.04377732e-01 -4.44469333e-01 7.26997674e-01
-3.29589635e-01 1.58231333e-01 -1.01500785e+00 -4.08388257e-01
3.74562621e-01 9.38125312e-01 -4.68142033e-01 -1.49722889e-01
2.89710984e-02 8.13489377e-01 -3.40657383e-01 1.84907719e-01
-1.01941633e+00 -4.72655356e-01 6.12735152e-01 3.11862100e-02
9.92961973e-02 -2.07701027e-02 -4.13529634e-01 -6.84722245e-01
-2.83764988e-01 -1.30825925e+00 3.72820385e-02 -8.12762618e-01
-1.19055796e+00 9.97307599e-01 -4.31255698e-01 -1.51121390e+00
-5.15414834e-01 -2.37970129e-01 -8.70200515e-01 8.33534181e-01
-1.30479395e+00 -1.18760097e+00 -4.35272604e-01 4.78183359e-01
6.11570418e-01 -7.23476708e-01 1.31439984e+00 3.62543374e-01
-1.96582764e-01 6.69861376e-01 1.09740146e-01 1.67276815e-01
5.74312687e-01 -1.26142550e+00 7.43259668e-01 8.75446558e-01
5.89768827e-01 1.72238156e-01 1.06507123e+00 -4.66922849e-01
-1.09913421e+00 -1.12549210e+00 3.35759908e-01 -2.02106774e-01
4.69229132e-01 -3.67046416e-01 -6.24814808e-01 3.60872179e-01
5.98194242e-01 -3.02067965e-01 8.14083993e-01 -5.69981158e-01
-2.13926405e-01 -2.40642682e-01 -1.30387437e+00 6.25980377e-01
9.21919823e-01 -3.79021496e-01 -2.31980637e-01 -2.71949861e-02
7.66715050e-01 -4.29460466e-01 -6.99783146e-01 3.67294818e-01
3.52077812e-01 -1.08523023e+00 1.05660486e+00 -8.53670835e-01
8.65350485e-01 -2.73434967e-01 -2.36359686e-01 -1.73626709e+00
1.57061383e-01 -1.19246995e+00 2.64732651e-02 1.70144618e+00
3.02892566e-01 -5.41656077e-01 8.22575390e-01 8.95235389e-02
-4.42799062e-01 -3.08161318e-01 -9.58705902e-01 -7.04918504e-01
3.09353471e-01 -6.27194822e-01 1.20774460e+00 9.20521617e-01
-7.38509297e-01 7.96681643e-02 -9.16425645e-01 -2.62314007e-02
4.56702709e-01 -1.35921374e-01 1.15143263e+00 -1.08467305e+00
-5.26673198e-01 -2.86640763e-01 -5.24644375e-01 -2.69830346e-01
4.21785656e-03 -8.00989866e-01 6.76974207e-02 -1.18410623e+00
-3.96658093e-01 -5.07342279e-01 -2.34800994e-01 3.30747962e-01
-1.63267881e-01 7.70882428e-01 3.95549595e-01 -3.08752596e-01
4.46496606e-01 6.40266836e-01 1.39187503e+00 -1.09220050e-01
-1.34042114e-01 3.72219831e-01 -7.84360230e-01 7.07665145e-01
1.29291606e+00 -7.48498261e-01 -5.45763314e-01 -2.47168481e-01
5.82033873e-01 1.13537408e-01 5.28235376e-01 -1.35571730e+00
-7.13056996e-02 7.35679418e-02 2.18841031e-01 -2.25678861e-01
4.06323254e-01 -7.52167404e-01 8.29318285e-01 3.91495645e-01
-3.34277600e-01 -7.83855170e-02 3.41310680e-01 5.01331568e-01
-5.52856326e-01 -2.23649442e-01 8.44877779e-01 -2.45180324e-01
-4.79023725e-01 7.06640482e-02 -4.52381462e-01 2.85105050e-01
7.70256281e-01 -3.25712189e-02 -2.30212674e-01 -7.19026923e-01
-7.75837481e-01 -5.78334689e-01 1.14213917e-02 5.66777647e-01
6.00107610e-01 -1.44320083e+00 -9.05601859e-01 2.09961846e-01
-5.31511426e-01 2.89920717e-02 3.44152808e-01 2.76767224e-01
-5.80009639e-01 5.94212227e-02 -5.10044634e-01 -6.87868521e-02
-1.15422690e+00 4.31569993e-01 3.74569118e-01 -1.22001283e-01
-6.65340006e-01 9.81782436e-01 3.90615761e-02 -2.61639506e-01
1.30285963e-01 1.24377348e-01 -2.28572667e-01 -2.24450342e-02
4.03445452e-01 6.11378908e-01 -1.58571713e-02 -2.98812509e-01
2.02687979e-01 1.03990711e-01 6.08615994e-01 -2.88112521e-01
1.76790369e+00 3.70851964e-01 -1.11688524e-02 5.28639734e-01
7.83544779e-01 4.56238985e-01 -8.55604529e-01 2.26070538e-01
-4.51927096e-01 -3.51202607e-01 -2.53482163e-01 -8.35241616e-01
-1.40474153e+00 1.14932406e+00 5.48045337e-01 7.83001840e-01
1.31776333e+00 -4.91612762e-01 9.61115539e-01 1.60316572e-01
5.03421187e-01 -7.53967047e-01 2.97025412e-01 2.75523067e-01
1.18639028e+00 -7.47412145e-01 -4.51250464e-01 -3.08047175e-01
-8.13848972e-01 1.03334188e+00 3.56692970e-01 -2.20728040e-01
4.19460207e-01 9.15173173e-01 3.49434882e-01 1.75099149e-01
-3.81345093e-01 5.75764403e-02 6.76742271e-02 9.29358840e-01
3.00764829e-01 8.24870169e-02 8.84697586e-03 8.33107173e-01
-9.17219818e-01 -1.80691406e-01 7.88625598e-01 5.38659453e-01
-1.04565628e-01 -1.26365006e+00 -5.15334725e-01 1.93111554e-01
-1.01778996e+00 -1.95735827e-01 -5.17568290e-01 8.14815938e-01
3.41160893e-01 8.06764901e-01 -1.40770793e-01 -6.08233929e-01
2.44056910e-01 3.23756635e-01 2.35249162e-01 -3.67895514e-01
-9.67130065e-01 -1.43667432e-02 2.33308390e-01 -7.36300424e-02
-5.76909482e-01 -4.12366480e-01 -1.06327045e+00 -1.84597790e-01
-1.45577595e-01 3.44420969e-01 8.23598564e-01 2.76980758e-01
1.62071496e-01 9.73721027e-01 8.75915706e-01 -6.37083948e-01
-5.57365537e-01 -9.35090542e-01 -6.68291390e-01 3.45938861e-01
1.51917145e-01 -1.76040098e-01 -5.09003043e-01 1.56495348e-01] | [15.547972679138184, 5.962237358093262] |
45222772-c394-4651-a474-6df7718c5235 | correcting-semantic-parses-with-natural | 2305.19974 | null | https://arxiv.org/abs/2305.19974v1 | https://arxiv.org/pdf/2305.19974v1.pdf | Correcting Semantic Parses with Natural Language through Dynamic Schema Encoding | In addressing the task of converting natural language to SQL queries, there are several semantic and syntactic challenges. It becomes increasingly important to understand and remedy the points of failure as the performance of semantic parsing systems improve. We explore semantic parse correction with natural language feedback, proposing a new solution built on the success of autoregressive decoders in text-to-SQL tasks. By separating the semantic and syntactic difficulties of the task, we show that the accuracy of text-to-SQL parsers can be boosted by up to 26% with only one turn of correction with natural language. Additionally, we show that a T5-base model is capable of correcting the errors of a T5-large model in a zero-shot, cross-parser setting. | ['Preethi Raghavan', 'Parag Pravin Dakle', 'Parker Glenn'] | 2023-05-31 | null | null | null | null | ['text-to-sql', 'semantic-parsing'] | ['computer-code', 'natural-language-processing'] | [ 2.69374937e-01 6.18447423e-01 3.22427712e-02 -7.96481550e-01
-1.49911094e+00 -5.60268223e-01 2.91024148e-01 4.87739652e-01
-3.20532650e-01 3.38335127e-01 6.46894455e-01 -6.24871016e-01
3.18849653e-01 -7.26119578e-01 -9.86967981e-01 4.15853232e-01
4.60047811e-01 7.17859745e-01 4.24653411e-01 -5.05066931e-01
2.19452277e-01 -1.00712374e-01 -1.25414670e+00 8.98179710e-01
9.00939047e-01 7.08519101e-01 2.69405723e-01 8.62461507e-01
-9.99682367e-01 1.08855271e+00 -5.41371107e-01 -7.79120743e-01
-8.98725837e-02 -2.28385970e-01 -1.24963844e+00 -4.37619865e-01
4.43075091e-01 -1.62379026e-01 1.19385384e-01 1.04653680e+00
2.38324061e-01 -3.39884847e-01 1.51984811e-01 -8.27347875e-01
-8.68169904e-01 7.92754531e-01 -8.85658115e-02 7.80668631e-02
7.06034482e-01 -8.49350542e-02 1.32814324e+00 -9.91864145e-01
6.68426991e-01 1.64724028e+00 6.75264180e-01 5.90915501e-01
-1.18426251e+00 -6.07815683e-02 1.42296419e-01 -2.87436038e-01
-7.52506196e-01 -7.83441067e-01 1.32085040e-01 -2.44974986e-01
1.92288470e+00 1.55986965e-01 -1.76233724e-01 9.12696958e-01
1.11280464e-01 6.91910386e-01 5.28764904e-01 -7.48711646e-01
2.61762291e-02 1.43467128e-01 4.75559235e-01 9.60729897e-01
1.94077775e-01 -2.42107242e-01 -6.85120761e-01 -1.06197178e-01
1.12090126e-01 -2.94200242e-01 2.58595228e-01 -9.01289731e-02
-6.16602540e-01 1.01401782e+00 1.14239715e-01 1.42064571e-01
-1.60292506e-01 3.92609358e-01 7.14144766e-01 2.40392640e-01
5.56959927e-01 8.61715794e-01 -1.05941617e+00 -5.82027376e-01
-5.24320543e-01 2.17796221e-01 1.07654250e+00 1.38329899e+00
6.59828842e-01 -4.53231223e-02 3.93132120e-02 7.93869734e-01
1.74939454e-01 4.45511967e-01 4.62413192e-01 -1.02869761e+00
8.00413907e-01 7.74708092e-01 1.73860952e-01 -5.65798104e-01
-2.98354387e-01 -3.49377505e-02 1.45862922e-01 -2.49339983e-01
4.42282617e-01 3.12546976e-02 -7.75203168e-01 1.69343650e+00
-4.43407930e-02 -5.09537697e-01 3.25492263e-01 3.09793264e-01
5.11690021e-01 4.63740289e-01 7.04239666e-01 -8.05909634e-02
1.57018101e+00 -8.18575382e-01 -5.97889125e-01 -8.86873960e-01
1.22275484e+00 -8.93022656e-01 1.58758986e+00 -3.76088731e-02
-1.37242472e+00 -4.33072835e-01 -6.91840351e-01 -5.74604809e-01
-3.70445549e-01 -1.64233759e-01 6.55784726e-01 6.14666164e-01
-1.13343847e+00 4.85758424e-01 -1.01396215e+00 -8.47448051e-01
8.51722807e-02 -4.91808988e-02 -1.18426703e-01 -1.34466365e-01
-1.06679165e+00 1.15670717e+00 3.37798685e-01 -4.17212218e-01
-3.33069265e-01 -8.54809225e-01 -8.70172322e-01 2.62991816e-01
6.51845753e-01 -8.15097511e-01 1.81105161e+00 -7.76096165e-01
-1.12698209e+00 9.59100425e-01 -7.76994944e-01 -4.71045434e-01
3.28693092e-01 -7.85822570e-01 -2.89319456e-01 -4.18683216e-02
5.91866314e-01 5.12854636e-01 3.24901670e-01 -8.57351184e-01
-7.58090556e-01 -5.20157754e-01 5.12547158e-02 -2.01700367e-02
-1.55164972e-01 4.70284849e-01 -3.99421662e-01 -3.49596918e-01
-9.40192695e-05 -5.71609139e-01 -2.60066897e-01 -4.56777334e-01
-2.35137388e-01 -3.30499947e-01 4.52168524e-01 -1.06939781e+00
1.05510485e+00 -2.20935988e+00 -1.26046449e-01 -2.75633812e-01
-6.53772876e-02 1.93889260e-01 -3.53302985e-01 5.00903070e-01
-7.34298676e-02 6.75263107e-01 -2.11294457e-01 -2.48530388e-01
3.34397666e-02 3.16953540e-01 -7.39247501e-01 -4.78412777e-01
8.22464466e-01 1.27692533e+00 -9.68549609e-01 -4.04976755e-01
-8.16491991e-02 8.03591609e-02 -9.79916036e-01 4.18968022e-01
-7.84180284e-01 -3.93569469e-02 -6.09348893e-01 6.04606807e-01
5.23202717e-02 -2.89573163e-01 2.87470043e-01 7.33324364e-02
5.39604798e-02 1.12598920e+00 -5.92024982e-01 1.78225350e+00
-6.28988087e-01 2.40885586e-01 -9.02421698e-02 -8.21691394e-01
8.38692605e-01 1.86441794e-01 1.84672475e-01 -1.20883012e+00
-1.63365081e-01 4.02421087e-01 -5.05965650e-01 -7.71885991e-01
7.09888637e-01 -3.32341999e-01 -5.49877644e-01 3.01897496e-01
1.82044223e-01 -2.58212656e-01 9.39377025e-03 4.12963480e-01
1.46180427e+00 2.16757476e-01 1.71924070e-01 -1.32753506e-01
8.62515867e-02 5.38864851e-01 3.80386919e-01 1.01007009e+00
7.99859315e-02 4.91252244e-01 7.79184103e-01 -4.22080278e-01
-1.21405697e+00 -1.00551641e+00 2.91861981e-01 1.66946089e+00
-3.72283608e-01 -6.97797179e-01 -1.02225149e+00 -9.14532423e-01
1.74245253e-01 1.48325515e+00 -1.82624564e-01 -4.91176099e-01
-7.32184768e-01 -6.23305678e-01 7.99693644e-01 8.14689457e-01
3.37786563e-02 -8.75088513e-01 -6.39829338e-01 5.98680675e-01
-4.09267962e-01 -1.57134187e+00 -1.28400877e-01 4.95195240e-01
-1.07489896e+00 -1.03812611e+00 2.11271390e-01 -6.63275898e-01
2.19657749e-01 1.08106704e-02 1.64466572e+00 1.48041070e-01
-2.28960440e-01 5.24640560e-01 -4.65313256e-01 -4.63873476e-01
-9.86202419e-01 2.54567206e-01 -4.65794623e-01 -7.54679680e-01
8.68018210e-01 -1.71916887e-01 5.73888943e-02 -6.28871545e-02
-7.66574502e-01 4.94894572e-02 3.60285372e-01 6.71355307e-01
3.32364321e-01 -4.58351046e-01 6.60412550e-01 -1.37670875e+00
7.07103610e-01 -3.53540570e-01 -4.96100336e-01 4.90042746e-01
-8.51766765e-01 6.33252084e-01 6.91041172e-01 2.54891038e-01
-1.37075508e+00 1.10174403e-01 -4.07311380e-01 2.28434745e-02
-2.83396374e-02 6.14086032e-01 -9.59010348e-02 3.67491335e-01
8.12012136e-01 -1.26229808e-01 -1.45735696e-01 -6.90638006e-01
8.23570669e-01 5.01791954e-01 7.40718007e-01 -7.37463593e-01
4.29285437e-01 -1.66882593e-02 -4.87442255e-01 -5.62220335e-01
-1.21275282e+00 -2.79499948e-01 -4.58338082e-01 4.20258999e-01
1.04915178e+00 -9.36071813e-01 -3.76573503e-01 -4.95814160e-03
-1.47299337e+00 -1.65842250e-01 -4.68174428e-01 -9.29120556e-02
-5.28613865e-01 3.92134398e-01 -8.38598847e-01 -5.57663143e-01
-3.05147737e-01 -9.89642680e-01 1.23851240e+00 -5.77344149e-02
-5.56798100e-01 -8.66142094e-01 -9.90418568e-02 5.78735054e-01
6.09728754e-01 -4.49438542e-01 1.68946660e+00 -1.03131580e+00
-5.93062103e-01 -1.01031214e-01 -4.29465294e-01 2.80621082e-01
-1.01983845e-01 -1.87112033e-01 -9.91862833e-01 1.58570990e-01
2.16323599e-01 -5.74746072e-01 7.36052155e-01 6.05579987e-02
8.44456136e-01 -2.74453610e-01 -9.05862451e-02 3.16255033e-01
1.49574578e+00 1.19590633e-01 5.07875621e-01 2.61703998e-01
6.05461597e-01 7.25050151e-01 5.59364796e-01 1.57135338e-01
6.74641788e-01 4.09466773e-01 2.67508715e-01 2.04686493e-01
-2.58458376e-01 -7.78219104e-01 2.67208815e-01 9.09678936e-01
6.49878860e-01 -1.97395772e-01 -1.22612536e+00 5.60096443e-01
-1.71372223e+00 -4.57061678e-01 -3.25668007e-01 2.08841825e+00
8.90905261e-01 4.17791247e-01 -3.08501333e-01 -5.52752852e-01
4.34477091e-01 5.51264547e-02 -3.76732349e-01 -8.90808880e-01
-7.92985335e-02 5.33320904e-01 6.75419927e-01 9.09992695e-01
-7.80753374e-01 1.59622395e+00 7.47916603e+00 2.64167756e-01
-7.30041325e-01 2.51278967e-01 4.13906813e-01 2.13726640e-01
-5.98399758e-01 4.36085045e-01 -1.03845799e+00 2.18259111e-01
1.67552161e+00 -3.08576196e-01 4.68255907e-01 1.07714999e+00
5.72188646e-02 -2.18173385e-01 -1.40690362e+00 5.02426267e-01
1.95752252e-02 -1.15419650e+00 2.25933626e-01 -5.79680324e-01
1.78177401e-01 2.24573433e-01 -4.60182905e-01 7.01493382e-01
7.72535801e-01 -1.03905153e+00 5.61040580e-01 3.28945607e-01
7.13212252e-01 -3.75158072e-01 5.82363665e-01 2.47698784e-01
-9.78187680e-01 -1.49173915e-01 -3.93618166e-01 -1.14413992e-01
2.97255665e-01 3.44890356e-01 -1.00978732e+00 4.37840670e-01
6.06092572e-01 4.43490624e-01 -8.54894936e-01 3.06968629e-01
-2.33569801e-01 4.91149187e-01 -1.42384499e-01 5.97660430e-02
2.01204270e-02 2.13963464e-01 1.44185588e-01 1.39786220e+00
2.28056252e-01 5.11057638e-02 1.09061979e-01 8.90660822e-01
-2.32331797e-01 6.84319735e-02 -6.44869149e-01 -4.31090295e-01
4.97372627e-01 6.12158060e-01 -3.25848401e-01 -6.48335278e-01
-6.67191565e-01 1.05049372e+00 7.07171261e-01 2.49931365e-01
-5.06675899e-01 -2.78158575e-01 5.98933756e-01 6.96889609e-02
2.62267679e-01 -5.17156497e-02 -7.12536752e-01 -1.26027334e+00
4.22636539e-01 -1.04929757e+00 5.42918146e-01 -1.13635409e+00
-1.08135486e+00 3.17704856e-01 -2.87721425e-01 -1.88823149e-01
-6.50343716e-01 -6.17379963e-01 -2.66748816e-01 8.71552408e-01
-1.50631297e+00 -7.45948017e-01 -7.00161746e-03 -3.60762589e-02
7.81919956e-01 2.42780540e-02 1.22602367e+00 3.01703215e-01
-1.94092080e-01 4.44058657e-01 -4.81737554e-02 1.56763256e-01
6.98123097e-01 -1.26801276e+00 1.23552668e+00 1.02725458e+00
4.25782800e-02 6.40186787e-01 7.41287053e-01 -7.81480610e-01
-1.65723264e+00 -1.22932279e+00 1.68367231e+00 -9.08120692e-01
7.62418747e-01 -3.63053828e-01 -1.26807594e+00 1.16715205e+00
1.64139345e-02 -3.68496388e-01 2.78136104e-01 3.15602392e-01
-7.47374475e-01 1.37431204e-01 -9.66442287e-01 3.70247722e-01
1.11522353e+00 -1.08387256e+00 -1.03996897e+00 4.46206182e-01
1.41942441e+00 -4.23011869e-01 -6.39605343e-01 2.30076894e-01
2.20964476e-01 -7.64515460e-01 9.02587116e-01 -1.24726200e+00
7.53790498e-01 1.89803988e-01 -4.68147069e-01 -1.07924569e+00
2.18435824e-02 -4.54528749e-01 2.48439744e-01 1.26206243e+00
5.63042164e-01 -5.13736606e-01 8.58367026e-01 1.06174052e+00
-3.13151002e-01 -2.12012842e-01 -8.16147208e-01 -5.33399880e-01
2.42810503e-01 -7.25191712e-01 3.52269560e-01 6.31209850e-01
1.23009980e-01 7.53201246e-01 1.06684156e-01 1.60596464e-02
2.66171575e-01 -1.70816973e-01 6.10643029e-01 -1.24990380e+00
-3.21512997e-01 -1.78935483e-01 2.08888836e-02 -1.13571250e+00
2.00953633e-01 -9.78866696e-01 3.40442449e-01 -1.52978218e+00
2.54696757e-01 -3.15211028e-01 -4.11600806e-03 5.80458581e-01
-2.61756390e-01 -4.51412797e-01 2.43483767e-01 5.08819930e-02
-7.22482264e-01 3.40663910e-01 3.40816647e-01 1.35788396e-01
-2.10008379e-02 -4.35018450e-01 -1.12570596e+00 6.38097584e-01
5.27193248e-01 -8.11432779e-01 -2.29240373e-01 -1.24459803e+00
5.90949714e-01 4.64077353e-01 1.12844042e-01 -6.63210690e-01
1.36870176e-01 -1.86642143e-03 -1.84738934e-01 -3.15900564e-01
9.23423097e-03 -4.51190442e-01 -2.69823462e-01 4.12971348e-01
-6.98674381e-01 5.85169077e-01 3.61095995e-01 5.17707705e-01
-2.69972563e-01 -3.98047209e-01 6.89480722e-01 -5.27531743e-01
-7.76784241e-01 -3.73934180e-01 -3.14971358e-01 8.27066600e-01
4.39670265e-01 2.03299165e-01 -6.16634607e-01 -1.42486244e-01
-6.00156963e-01 2.41843477e-01 4.56574082e-01 6.96196795e-01
4.64942425e-01 -6.91621363e-01 -4.96015638e-01 1.95629075e-01
3.61497551e-01 -8.28543156e-02 -2.46877924e-01 4.36004639e-01
-5.59740365e-01 8.65272403e-01 5.43152541e-02 -3.81824702e-01
-1.04305089e+00 4.33601081e-01 4.41880792e-01 -3.28493029e-01
-5.08910358e-01 8.92688334e-01 1.42877787e-01 -7.12560475e-01
2.42868379e-01 -4.55274642e-01 2.36143366e-01 -2.44314343e-01
3.00488234e-01 4.52534780e-02 5.35481572e-01 -1.07771195e-01
-4.57299173e-01 2.67638057e-01 -3.11009258e-01 -8.30410123e-02
1.40834618e+00 -1.49944410e-01 -1.48945332e-01 4.06172097e-01
1.17351532e+00 2.21362300e-02 -7.85788298e-01 -3.04606676e-01
8.75616372e-01 -2.77064532e-01 -2.98707515e-01 -1.12648559e+00
-5.09850144e-01 1.04672158e+00 3.00193906e-01 1.34263709e-01
8.01944137e-01 2.31106400e-01 1.01997554e+00 7.39224970e-01
3.03829759e-01 -1.19162667e+00 3.14825587e-02 1.00570917e+00
4.81073439e-01 -1.33671474e+00 -5.37881732e-01 -5.63783944e-01
-6.06245458e-01 1.15352702e+00 5.42460680e-01 1.29744530e-01
2.03755185e-01 5.13565302e-01 1.36143595e-01 -3.98208678e-01
-1.05331504e+00 -4.32069739e-03 -3.97105366e-01 4.31988716e-01
8.12153697e-01 -3.61606926e-02 -1.17535539e-01 8.06136370e-01
-2.45601401e-01 -8.97206459e-03 5.43740571e-01 1.15094995e+00
-7.81987846e-01 -1.26163065e+00 3.99776995e-02 4.53303277e-01
-7.04134405e-01 -6.51547551e-01 -5.20283222e-01 6.15891397e-01
-5.46886683e-01 1.17517710e+00 6.33685812e-02 -2.10286498e-01
7.60684967e-01 8.44552696e-01 8.34419727e-02 -1.08458948e+00
-7.54873335e-01 -1.83884025e-01 6.66680276e-01 -9.90313709e-01
1.26159161e-01 -5.18678248e-01 -1.50022268e+00 -1.54595658e-01
-1.75330583e-02 1.87970757e-01 7.74148345e-01 1.05882168e+00
6.97660744e-01 5.76599836e-01 1.45463988e-01 1.34626672e-01
-9.61172402e-01 -7.28134990e-01 -6.78262860e-02 6.17957711e-01
4.97295195e-03 -8.52020830e-02 -1.32157952e-01 2.34382346e-01] | [9.947724342346191, 7.954697608947754] |
4f4c6f01-9728-4c75-92db-a110a616e32a | quark-a-gradient-free-quantum-learning | 2210.01311 | null | https://arxiv.org/abs/2210.01311v1 | https://arxiv.org/pdf/2210.01311v1.pdf | Quark: A Gradient-Free Quantum Learning Framework for Classification Tasks | As more practical and scalable quantum computers emerge, much attention has been focused on realizing quantum supremacy in machine learning. Existing quantum ML methods either (1) embed a classical model into a target Hamiltonian to enable quantum optimization or (2) represent a quantum model using variational quantum circuits and apply classical gradient-based optimization. The former method leverages the power of quantum optimization but only supports simple ML models, while the latter provides flexibility in model design but relies on gradient calculation, resulting in barren plateau (i.e., gradient vanishing) and frequent classical-quantum interactions. To address the limitations of existing quantum ML methods, we introduce Quark, a gradient-free quantum learning framework that optimizes quantum ML models using quantum optimization. Quark does not rely on gradient computation and therefore avoids barren plateau and frequent classical-quantum interactions. In addition, Quark can support more general ML models than prior quantum ML methods and achieves a dataset-size-independent optimization complexity. Theoretically, we prove that Quark can outperform classical gradient-based methods by reducing model query complexity for highly non-convex problems; empirically, evaluations on the Edge Detection and Tiny-MNIST tasks show that Quark can support complex ML models and significantly reduce the number of measurements needed for discovering near-optimal weights for these tasks. | ['Zhihao Jia', 'Heyang Huang', 'Zhuoming Chen', 'Zhihao Zhang'] | 2022-10-02 | null | null | null | null | ['edge-detection'] | ['computer-vision'] | [ 1.10948339e-01 -9.70929712e-02 -2.77243644e-01 -1.71267271e-01
-1.14997625e+00 -4.12526339e-01 3.70471835e-01 1.63284287e-01
-6.67715907e-01 6.33869886e-01 -2.87030041e-01 -6.24506056e-01
1.45704998e-02 -1.02473736e+00 -7.83316851e-01 -8.18676293e-01
5.79903722e-02 6.06070995e-01 1.82654709e-01 -2.78320044e-01
5.66029072e-01 1.56913221e-01 -1.09300697e+00 1.66352138e-01
1.03722227e+00 7.60637820e-01 -4.24952693e-02 8.25122714e-01
2.73203492e-01 2.50592381e-01 3.14901546e-02 -5.07987261e-01
5.00161529e-01 -6.14277184e-01 -5.29700518e-01 -4.24783081e-01
5.38727045e-01 -1.77755266e-01 -8.74180496e-01 1.39719093e+00
7.11357892e-01 2.31432348e-01 5.80104411e-01 -8.74427557e-01
-8.14201772e-01 6.21060669e-01 -1.89380422e-01 2.82621056e-01
-4.85564396e-02 4.53944981e-01 1.81902015e+00 -6.96934521e-01
4.69857961e-01 1.00503135e+00 5.90147853e-01 4.99954164e-01
-1.89636791e+00 -5.56519091e-01 -4.98239756e-01 5.97320437e-01
-1.76380944e+00 -5.58668792e-01 3.64741415e-01 -6.98229596e-02
1.39056444e+00 1.89275146e-01 6.17830515e-01 5.16528010e-01
4.96218532e-01 6.08526528e-01 1.16297221e+00 -6.44733310e-01
5.73117971e-01 -5.42387217e-02 2.87844449e-01 1.21105671e+00
2.09852263e-01 4.24744248e-01 -9.04202759e-01 -3.62773359e-01
3.04462045e-01 -2.67727673e-01 -2.14504465e-01 -2.48717695e-01
-1.35021734e+00 1.09177399e+00 5.28436005e-01 -9.34552774e-02
-1.80490166e-01 5.85772216e-01 3.40039998e-01 4.55305189e-01
8.00286755e-02 6.81642115e-01 -2.76644558e-01 -1.70110494e-01
-9.01686966e-01 2.81638652e-01 9.53364909e-01 7.57662058e-01
1.33126390e+00 -2.98533171e-01 -2.07412988e-01 1.24436282e-01
2.94028848e-01 9.23529267e-01 5.15213013e-02 -9.94873285e-01
2.82056451e-01 1.91415295e-01 7.13221952e-02 -4.41016525e-01
-6.70094550e-01 -4.23971713e-01 -6.68556094e-01 -1.49753243e-01
3.91553998e-01 -2.37438474e-02 -7.67368257e-01 1.72538245e+00
3.42895091e-01 1.74155980e-01 1.61854416e-01 8.37265491e-01
4.90263611e-01 5.02618849e-01 -4.31642622e-01 -1.99771538e-01
1.24271607e+00 -8.19949448e-01 -5.23082137e-01 -2.26242304e-01
1.24011815e+00 -5.08290052e-01 1.14352024e+00 5.09354532e-01
-7.84717500e-01 5.88367134e-02 -1.43165553e+00 -3.83962512e-01
-1.53358147e-01 -1.94301710e-01 1.09616673e+00 1.06882453e+00
-1.00650442e+00 1.15509808e+00 -1.11014831e+00 -1.34718539e-02
3.06353569e-01 6.91859066e-01 -9.96589959e-02 -1.73535630e-01
-1.18134093e+00 8.16188157e-01 4.77520078e-01 1.01816647e-01
-5.31889141e-01 -5.05195856e-01 -6.63584352e-01 5.61702363e-02
4.92490619e-01 -9.98649478e-01 1.10989714e+00 -3.21547836e-01
-1.77868700e+00 6.03342772e-01 -2.51696587e-01 -6.29074872e-01
1.85171530e-01 8.75252038e-02 -9.55257937e-02 1.77100167e-01
-2.19027922e-01 4.03672695e-01 5.88548303e-01 -2.57553071e-01
-2.45700642e-01 -4.63982701e-01 2.45939389e-01 2.56983966e-01
-1.85524344e-01 -3.95860940e-01 -4.41416740e-01 2.84767807e-01
4.91828829e-01 -1.37061512e+00 -4.54079777e-01 -4.84369636e-01
-6.16326213e-01 -1.53792977e-01 9.44313556e-02 2.94562448e-02
1.32332838e+00 -1.96248293e+00 3.48980576e-01 4.17324364e-01
4.83630508e-01 -1.01744041e-01 -1.53160691e-01 5.91793895e-01
4.15164381e-01 1.96654364e-01 -2.19568893e-01 -3.35976630e-01
5.47090352e-01 2.89680362e-01 -8.93920287e-02 7.94820428e-01
-3.05049401e-02 1.31819892e+00 -7.99777389e-01 -5.01806200e-01
1.14144705e-01 1.11937962e-01 -1.35898137e+00 -3.96965265e-01
-2.16192186e-01 2.99615234e-01 -5.40073752e-01 4.22158629e-01
6.66054785e-01 -6.61515534e-01 3.67550373e-01 -5.39652169e-01
-1.82443216e-01 6.62980258e-01 -1.25018525e+00 1.76902854e+00
-3.32265645e-01 6.34645641e-01 5.20830899e-02 -9.73060310e-01
4.54556458e-02 -2.52182186e-01 2.36327678e-01 -9.12371278e-01
1.51607260e-01 5.06950319e-01 2.09928423e-01 -1.97470322e-01
5.79541326e-01 -4.96810853e-01 -1.13542080e-01 5.74626088e-01
4.05654609e-02 -5.43375850e-01 2.96811104e-01 5.27381599e-01
1.29905593e+00 -3.62597220e-02 1.17475847e-02 -4.19704109e-01
1.42572090e-01 -9.78024229e-02 6.13306820e-01 1.16043866e+00
-2.94956863e-01 3.14776868e-01 4.60557282e-01 -2.52395153e-01
-1.02596116e+00 -1.32951069e+00 -6.12151504e-01 1.09998250e+00
4.82605785e-01 -9.93995667e-01 -7.13264823e-01 -3.29868764e-01
5.22000082e-02 8.22429538e-01 -3.15094560e-01 -4.44578499e-01
-2.26984069e-01 -1.49811172e+00 4.82695162e-01 -1.00757234e-01
2.58647859e-01 -4.55413342e-01 -3.28017116e-01 2.44590521e-01
1.79351456e-02 -1.24029517e+00 -5.18442512e-01 5.08367836e-01
-9.36234295e-01 -9.50882971e-01 1.84102416e-01 1.37594128e-02
4.80366558e-01 1.07111603e-01 9.76574659e-01 -3.70449908e-02
-4.91192013e-01 2.62736768e-01 -9.05456096e-02 -5.83500937e-02
-3.88633519e-01 1.81047767e-01 4.03492093e-01 -2.06433862e-01
6.20482385e-01 -4.32787091e-01 -7.41810858e-01 -1.24998882e-01
-7.67810047e-01 1.24146231e-01 8.16313386e-01 1.22910810e+00
6.56124532e-01 -1.43681258e-01 1.88534990e-01 -8.55712056e-01
4.03732806e-01 9.95243713e-03 -9.92062867e-01 2.22013369e-01
-9.73027945e-01 7.54111171e-01 6.26848459e-01 -4.16057706e-02
-3.71555507e-01 2.99400724e-02 -1.05409160e-01 -8.34972877e-03
5.94249964e-01 6.52828872e-01 3.86407852e-01 -5.74744165e-01
1.02542925e+00 2.50937849e-01 -1.31645501e-01 5.63576212e-03
6.21654034e-01 5.43728828e-01 3.73378880e-02 -6.81566298e-01
8.21572959e-01 5.39726853e-01 5.97114921e-01 -1.14249539e+00
-1.02015162e+00 -2.42919028e-01 -6.18215859e-01 3.18830043e-01
9.88162994e-01 -8.95071864e-01 -1.18077338e+00 1.01583138e-01
-8.52313221e-01 -3.11318487e-01 -9.89375636e-02 8.98497283e-01
-5.07769704e-01 6.87141657e-01 -1.03877413e+00 -9.89396513e-01
-3.63004535e-01 -1.43634760e+00 1.20100582e+00 1.74889699e-01
3.58988345e-01 -7.88428903e-01 -6.29937947e-02 3.98716927e-01
2.54595578e-01 -3.10430378e-01 9.35951173e-01 -2.93420970e-01
-1.13947320e+00 -3.08755606e-01 -3.42368394e-01 2.09464058e-01
-4.81224090e-01 -1.54090062e-01 -9.44464982e-01 -3.97810429e-01
-1.54500663e-01 -5.37480652e-01 1.12206066e+00 2.79253960e-01
6.68303847e-01 -3.95408459e-03 -1.72778755e-01 9.14274991e-01
1.32365453e+00 -4.92257327e-01 4.26630318e-01 -5.86889833e-02
8.04385006e-01 -1.06267102e-01 1.22066535e-01 4.26861972e-01
5.39533913e-01 7.86544502e-01 2.85405070e-01 2.81191885e-01
3.77638847e-01 -1.12087086e-01 5.06208777e-01 1.15827417e+00
7.73828253e-02 3.19978148e-01 -6.99203491e-01 -1.01179928e-01
-1.83158565e+00 -1.00042164e+00 -4.27395254e-01 2.53334451e+00
1.01250994e+00 3.13783497e-01 -2.05847770e-01 -3.14867616e-01
3.26695472e-01 -5.18204607e-02 -8.51294696e-01 -4.27438974e-01
-1.32170662e-01 6.69453979e-01 1.01501489e+00 6.64491773e-01
-1.06726682e+00 1.17264748e+00 6.13107681e+00 9.75562334e-01
-9.74169850e-01 5.29738307e-01 2.03806520e-01 -3.73568207e-01
-3.68372858e-01 5.24565697e-01 -9.33765411e-01 5.05667590e-02
1.14134789e+00 -2.43961409e-01 1.12739241e+00 5.56841850e-01
-1.09534629e-01 -2.54439324e-01 -1.46250427e+00 1.43321955e+00
-3.74430865e-01 -1.62572324e+00 -3.44247878e-01 5.14819443e-01
7.89513826e-01 9.00041461e-01 7.15804007e-03 5.45336843e-01
-7.11323787e-03 -9.71903384e-01 5.77939332e-01 3.28382492e-01
6.99891448e-01 -7.56848454e-01 5.23331821e-01 5.80729723e-01
-9.27139103e-01 -1.14028946e-01 -7.98521876e-01 -3.04943472e-01
-1.56066958e-02 7.52168179e-01 -5.16468406e-01 4.95147109e-01
4.33660030e-01 2.77194619e-01 -4.59960878e-01 7.13249087e-01
-1.18015297e-02 5.69171548e-01 -9.64269340e-01 -4.97402132e-01
4.11483169e-01 -9.40818906e-01 4.56909120e-01 9.39701021e-01
5.98790497e-02 9.57272053e-02 3.09621334e-01 1.23151886e+00
-3.33531618e-01 3.06031168e-01 -4.98947084e-01 -4.71161574e-01
4.22010034e-01 1.16336524e+00 -6.06158972e-01 -1.63799480e-01
-4.65578496e-01 1.14472234e+00 3.95459503e-01 2.18058035e-01
-8.77659023e-01 -4.93955225e-01 3.66603822e-01 -3.37364793e-01
1.84048384e-01 -5.89003503e-01 -2.84324378e-01 -1.79194355e+00
-1.91422641e-01 -7.78760850e-01 8.33424479e-02 -1.80264175e-01
-1.08792555e+00 6.70191944e-02 -5.09278953e-01 -5.84190428e-01
-3.17935646e-02 -8.63582194e-01 -3.73078167e-01 8.75414193e-01
-1.36559451e+00 -7.57972062e-01 3.04821610e-01 2.72675216e-01
-2.49015689e-01 2.61094123e-01 9.36806977e-01 3.22775751e-01
-8.15301716e-01 6.58556461e-01 5.99974871e-01 -2.38765880e-01
5.85289180e-01 -1.53538406e+00 4.43910807e-01 8.02978337e-01
5.49869597e-01 9.16748643e-01 7.66500592e-01 -4.30253744e-01
-2.52389336e+00 -5.98572016e-01 7.28518248e-01 -6.04733169e-01
9.09356058e-01 -7.20829487e-01 -4.21998352e-01 4.78600353e-01
-1.98063299e-01 8.81737918e-02 8.18397403e-01 5.11987865e-01
-4.30576295e-01 -5.67986257e-02 -8.50975215e-01 7.75655746e-01
9.63808179e-01 -1.17973411e+00 -1.96041524e-01 9.76725221e-01
5.19393504e-01 -3.73014897e-01 -7.62176514e-01 2.88840145e-01
6.07832968e-01 -1.07316303e+00 8.47289681e-01 -6.04749143e-01
-4.21803668e-02 -1.84404433e-01 -5.10208547e-01 -9.89406049e-01
-2.01328740e-01 -1.06069553e+00 -2.70171225e-01 1.11970097e-01
6.61219776e-01 -8.62062693e-01 8.46513212e-01 7.68182337e-01
-2.60335654e-01 -7.90871561e-01 -1.35047090e+00 -7.37925529e-01
3.34293634e-01 -7.58083522e-01 1.57682687e-01 7.14109898e-01
2.47821063e-01 7.54047513e-01 -2.00875685e-01 3.28216732e-01
1.03744972e+00 3.71503025e-01 7.67761588e-01 -8.75883758e-01
-9.83953953e-01 -7.43998289e-01 -4.72599804e-01 -1.36580062e+00
3.74667682e-02 -1.45556402e+00 -1.36287594e-02 -1.04867995e+00
4.11452740e-01 -2.41426200e-01 -3.31073940e-01 1.63765460e-01
-1.52438581e-01 4.63247925e-01 1.12408109e-01 2.16184989e-01
-9.32733715e-01 7.67884016e-01 1.00857294e+00 -8.24567750e-02
-1.70296565e-01 -1.61281720e-01 -3.94511044e-01 4.38418239e-01
3.12786549e-01 -5.23116231e-01 -4.25993949e-02 -2.10978225e-01
1.01743650e+00 -8.84110332e-02 2.74495661e-01 -8.82668197e-01
5.07270694e-01 1.99288651e-02 -7.09470175e-03 -1.91452652e-01
3.30977142e-01 -3.81185338e-02 -4.05282617e-01 7.82882869e-01
-4.50290702e-02 -2.89830089e-01 -8.48330334e-02 6.07168734e-01
2.59122998e-01 -3.64092618e-01 9.52276886e-01 1.58638880e-01
-4.24884677e-01 5.08098006e-01 7.86986351e-02 5.36270216e-02
6.44339442e-01 1.91702977e-01 -2.88019776e-01 -1.37651831e-01
-6.55403018e-01 2.54045337e-01 5.89345872e-01 -3.49797219e-01
2.92790830e-01 -1.12255192e+00 -5.03909051e-01 7.90850818e-02
1.36559173e-01 -3.19299012e-01 3.37536186e-01 1.49033308e+00
-7.25489497e-01 5.13548493e-01 4.74997193e-01 -7.58371949e-01
-6.23777151e-01 3.62201393e-01 5.16864598e-01 -2.12884337e-01
-5.57358980e-01 9.77278173e-01 2.67380685e-01 -6.81774378e-01
-2.65797347e-01 -2.55261302e-01 7.18657136e-01 -3.14827383e-01
2.56702036e-01 3.03017050e-01 1.56819105e-01 -3.80774409e-01
-4.07822549e-01 4.27151650e-01 -7.76338112e-03 -2.89484411e-01
1.04110682e+00 7.76084363e-02 -3.06591481e-01 4.98165399e-01
1.32544768e+00 4.95503992e-02 -9.21176732e-01 -3.57293963e-01
1.09957546e-01 -4.45260145e-02 6.13200784e-01 -4.16160315e-01
-3.71208936e-01 1.19582951e+00 6.69674158e-01 1.88193262e-01
7.49250352e-01 -1.01478226e-01 1.01558816e+00 1.38492656e+00
7.02415109e-01 -1.41530228e+00 -2.18437448e-01 7.26188004e-01
4.09252122e-02 -1.39687777e+00 1.91981703e-01 -1.46942049e-01
-1.71798795e-01 1.18317211e+00 1.71435207e-01 -1.19303361e-01
5.90920568e-01 -1.23678729e-01 -5.61622322e-01 -3.63647997e-01
-7.42544651e-01 -3.70591789e-01 4.48683441e-01 -1.61078200e-01
2.91919827e-01 3.33856225e-01 -2.85534292e-01 3.67593288e-01
-2.68012792e-01 -1.77386045e-01 5.67060232e-01 7.65474200e-01
-5.45364261e-01 -1.32933271e+00 -1.18823880e-02 7.44841516e-01
-3.69162560e-01 -8.09918821e-01 -4.94644716e-02 4.68716651e-01
-2.31362984e-01 1.04253244e+00 -2.04964802e-01 -5.66140413e-01
2.50965841e-02 2.06722423e-01 1.12689340e+00 -7.80983686e-01
-2.35835791e-01 8.33609328e-02 -6.26809895e-02 -8.08291495e-01
6.97221160e-02 -5.87915778e-01 -1.56293809e+00 -7.46853650e-01
-9.52654839e-01 2.64210522e-01 9.11311209e-01 1.26893079e+00
6.80520833e-01 1.36247441e-01 4.75743562e-01 -8.62800062e-01
-1.50020206e+00 -7.78388321e-01 -6.34549797e-01 2.29130755e-03
1.98625714e-01 -4.69870359e-01 -5.61959684e-01 -6.37399137e-01] | [5.643320560455322, 4.881498336791992] |
f8e81403-1ef3-41c5-ac4e-e11c7a91026b | decoding-urban-health-nexus-interpretable | 2306.11847 | null | https://arxiv.org/abs/2306.11847v2 | https://arxiv.org/pdf/2306.11847v2.pdf | Decoding Urban-health Nexus: Interpretable Machine Learning Illuminates Cancer Prevalence based on Intertwined City Features | This study investigates the interplay among social demographics, built environment characteristics, and environmental hazard exposure features in determining community level cancer prevalence. Utilizing data from five Metropolitan Statistical Areas in the United States: Chicago, Dallas, Houston, Los Angeles, and New York, the study implemented an XGBoost machine learning model to predict the extent of cancer prevalence and evaluate the importance of different features. Our model demonstrates reliable performance, with results indicating that age, minority status, and population density are among the most influential factors in cancer prevalence. We further explore urban development and design strategies that could mitigate cancer prevalence, focusing on green space, developed areas, and total emissions. Through a series of experimental evaluations based on causal inference, the results show that increasing green space and reducing developed areas and total emissions could alleviate cancer prevalence. The study and findings contribute to a better understanding of the interplay among urban features and community health and also show the value of interpretable machine learning models for integrated urban design to promote public health. The findings also provide actionable insights for urban planning and design, emphasizing the need for a multifaceted approach to addressing urban health disparities through integrated urban design strategies. | ['Ali Mostafavi', 'Chenyue Liu'] | 2023-06-20 | null | null | null | null | ['causal-inference', 'interpretable-machine-learning', 'causal-inference'] | ['knowledge-base', 'methodology', 'miscellaneous'] | [ 1.88215151e-02 5.07188216e-02 -8.70427191e-01 1.48362210e-02
-6.41196430e-01 3.03304493e-01 4.38875318e-01 9.99408007e-01
-3.48001122e-01 5.25175452e-01 1.29352498e+00 -1.00307858e+00
-3.17474365e-01 -1.43692720e+00 -5.20950258e-01 -5.04560649e-01
3.74876037e-02 -3.38523120e-01 -3.95692229e-01 -1.94330946e-01
9.43704844e-02 3.18976879e-01 -1.07183993e+00 -2.28779048e-01
1.22768366e+00 1.33042574e-01 1.47425085e-01 3.09698880e-01
2.09268436e-01 4.84910101e-01 -7.13106990e-02 2.80258626e-01
-9.21867490e-02 -3.13283622e-01 -2.52910048e-01 -2.43695170e-01
-2.48192139e-02 -4.01788592e-01 -3.56673181e-01 5.31214952e-01
4.77627575e-01 5.19511625e-02 8.41576874e-01 -1.07636547e+00
-6.11710191e-01 4.90497231e-01 -2.77739197e-01 1.26083747e-01
2.34970719e-01 3.73583227e-01 8.74528766e-01 -3.02806824e-01
9.80771482e-02 1.03154302e+00 7.97471344e-01 -6.91961572e-02
-1.09762299e+00 -7.74309218e-01 4.62139279e-01 1.40653238e-01
-1.26106727e+00 -5.41113555e-01 4.05172467e-01 -8.11164677e-01
6.73721552e-01 6.20584309e-01 1.10297120e+00 5.15834093e-01
4.45101112e-01 3.10078800e-01 9.83263135e-01 -5.80278635e-01
2.55044580e-01 1.10336635e-02 2.47318104e-01 8.91538024e-01
7.43363023e-01 4.44768816e-01 -1.98341399e-01 -3.52961600e-01
2.42007039e-02 6.12707198e-01 1.09903492e-01 2.21306697e-01
-9.34007943e-01 9.30923939e-01 9.68838811e-01 1.84011534e-01
-4.22444075e-01 5.02370656e-01 3.63083705e-02 -4.91162211e-01
7.64393449e-01 1.72873005e-01 -2.42289957e-02 3.69676709e-01
-6.01825774e-01 8.09117407e-02 2.82757163e-01 1.74650982e-01
7.76273549e-01 -1.82128713e-01 -1.80247113e-01 6.50709331e-01
9.38802540e-01 1.12061119e+00 -1.67239845e-01 -5.58859289e-01
2.97139078e-01 7.66136944e-01 -6.74032643e-02 -1.24593222e+00
-7.91945338e-01 -3.12515229e-01 -4.29340154e-01 -1.47377804e-01
7.78429359e-02 -5.14353275e-01 -6.83592558e-01 1.55662489e+00
2.99380094e-01 2.04452816e-02 -3.85609716e-01 3.65601182e-01
5.67769289e-01 3.48160237e-01 1.06689262e+00 9.90096480e-02
1.21024859e+00 -3.68390173e-01 -4.04227853e-01 -2.36668393e-01
7.68487632e-01 -1.46172971e-01 9.78943586e-01 -6.12810671e-01
-4.82470423e-01 -3.68009061e-01 -5.61858416e-01 4.08608347e-01
-7.17928231e-01 -3.00395817e-01 5.59867740e-01 1.08812129e+00
-7.00102210e-01 8.25743675e-02 -1.12750435e+00 -8.66213143e-01
5.61076939e-01 -1.21134810e-01 1.38996005e-01 -1.31078735e-01
-1.12003851e+00 8.99906218e-01 -1.50227770e-01 -1.08211040e-01
-4.56983626e-01 -9.74615932e-01 -1.06316268e+00 1.15205511e-01
-2.74176329e-01 -8.62453103e-01 5.03904819e-01 -2.77936935e-01
-5.30274987e-01 1.91450194e-01 -2.18957677e-01 -9.45728421e-02
2.13473663e-02 2.08025542e-03 -7.03432441e-01 -3.14341038e-01
8.17276716e-01 4.84786987e-01 -3.79651636e-01 -9.81875122e-01
-9.21164334e-01 -6.74230635e-01 -3.87938976e-01 2.78348446e-01
-7.25507021e-01 -4.18510027e-02 3.51848751e-01 -3.73015553e-01
-1.61889896e-01 -8.71932983e-01 -7.87554622e-01 -1.92761526e-01
-5.93028963e-01 -2.75622774e-02 1.59176573e-01 -8.25179577e-01
1.56991541e+00 -1.96474946e+00 -7.34862387e-01 5.40863991e-01
7.19249323e-02 -4.01049018e-01 1.01340137e-01 5.62800050e-01
4.65997964e-01 6.00252688e-01 -4.63865139e-02 2.75443882e-01
-3.18389982e-02 -2.33062450e-02 2.96787709e-01 6.01819098e-01
1.49071664e-01 8.88973773e-01 -1.18501031e+00 -4.11538452e-01
8.52921009e-01 3.65625560e-01 -6.57083809e-01 -3.80677700e-01
7.74626210e-02 3.15552354e-01 -7.45278835e-01 5.78677833e-01
3.92928809e-01 -1.14987060e-01 2.80696690e-01 3.77676874e-01
-7.63096690e-01 4.85495746e-01 -5.38833678e-01 6.66533768e-01
-5.29263079e-01 7.86126912e-01 -2.41882131e-01 -4.54985708e-01
6.71091437e-01 -2.41195806e-03 7.32007205e-01 -8.93030763e-01
-1.55038923e-01 -7.48007298e-02 -9.59397033e-02 -9.11723554e-01
2.85496265e-01 -2.55172133e-01 -1.80945411e-01 4.09856677e-01
-1.21604133e+00 1.66735351e-01 -1.83894709e-01 -1.19184010e-01
1.17151761e+00 -4.72727537e-01 5.34008920e-01 -6.59814537e-01
1.41716272e-01 4.48183835e-01 4.69368696e-01 4.75187361e-01
-3.78790408e-01 -1.62447557e-01 1.28414437e-01 -2.64154494e-01
-5.85746169e-01 -1.23147821e+00 -5.36659002e-01 1.26060379e+00
2.37429485e-01 -2.60668397e-01 -1.89652458e-01 -1.67577147e-01
2.97574729e-01 1.13953912e+00 -6.80465519e-01 -2.95201808e-01
-3.88808221e-01 -1.31249857e+00 4.06101346e-01 4.13735867e-01
7.48652756e-01 -2.70068824e-01 -6.26314998e-01 5.47182076e-02
-3.75871420e-01 -3.68683070e-01 -1.97257340e-01 -2.54241139e-01
-6.61489964e-01 -1.44732571e+00 -3.51494402e-01 -5.13649702e-01
8.22148025e-01 5.58311224e-01 7.55980313e-01 3.74888092e-01
-2.54444927e-01 6.16594613e-01 -1.39421195e-01 -7.92583287e-01
-3.93209189e-01 2.15364903e-01 -4.03476596e-01 -5.60471117e-01
4.83102918e-01 -3.95824045e-01 -9.66536522e-01 2.96222270e-01
-3.68003279e-01 1.23637609e-01 3.66331369e-01 2.69158214e-01
3.61878365e-01 2.13284105e-01 7.47690141e-01 -7.16829896e-01
3.12407166e-01 -1.31160331e+00 -1.13861553e-01 4.95573282e-02
-9.99203444e-01 -3.10345888e-01 5.37007581e-03 9.82056186e-02
-1.00608253e+00 -1.19711392e-01 -2.21143991e-01 9.66682315e-01
-3.60105157e-01 7.37998843e-01 -2.84532428e-01 2.90585428e-01
9.02778804e-01 -4.12376642e-01 -9.09711644e-02 -5.82691506e-02
1.25819653e-01 4.28388089e-01 -2.77472049e-01 -2.77327508e-01
7.67733335e-01 7.82926023e-01 8.53493530e-03 -1.17544496e+00
-4.92737442e-01 -6.29273355e-01 -2.18562558e-01 -5.52795410e-01
1.14446747e+00 -1.27314949e+00 -4.91561353e-01 6.18771128e-02
-1.82264671e-01 -7.61591673e-01 1.42233754e-02 9.05930281e-01
1.06764726e-01 -3.17476869e-01 -2.68263847e-01 -7.77005553e-01
-2.73312144e-02 -7.66757309e-01 6.11054361e-01 3.01485330e-01
-4.71636921e-01 -1.29958439e+00 4.49526936e-01 6.15881085e-01
6.73249245e-01 9.01381791e-01 1.40423083e+00 2.05509469e-01
-6.18124783e-01 1.51724756e-01 -3.38631541e-01 -3.88920367e-01
6.45318210e-01 8.94727558e-02 -6.17839396e-01 1.83743328e-01
-7.06153750e-01 2.74920106e-01 9.24652636e-01 9.32297289e-01
6.35826588e-01 -4.61955518e-01 -1.05724943e+00 2.26619154e-01
1.55047166e+00 1.98792189e-01 3.78169417e-01 5.35991371e-01
7.06545532e-01 6.16907597e-01 2.99479008e-01 4.06113774e-01
1.08804262e+00 2.83971369e-01 4.73835915e-01 -7.20372200e-01
-2.17220157e-01 -3.59008044e-01 1.17682345e-01 1.05713464e-01
1.05415575e-01 -1.95335951e-02 -1.47883463e+00 8.49293053e-01
-1.31371677e+00 -1.22790420e+00 -4.05401021e-01 2.04661274e+00
3.63379568e-01 -7.94435292e-02 3.55385780e-01 -9.09490436e-02
4.90713835e-01 3.08312416e-01 -3.11996460e-01 -1.31125435e-01
1.39090791e-01 -2.02654935e-02 8.80672157e-01 6.04674160e-01
-9.46014822e-01 3.70000213e-01 7.16815090e+00 1.17552705e-01
-9.64009285e-01 9.94294807e-02 1.09556746e+00 -5.75641841e-02
-1.03269506e+00 2.91335192e-02 -5.92108727e-01 3.71810436e-01
1.14722085e+00 -1.46099240e-01 7.12738708e-02 5.46871960e-01
1.24742472e+00 -4.63647187e-01 -4.12600070e-01 -1.62248731e-01
-3.74813765e-01 -1.43629348e+00 -5.71016312e-01 5.50637722e-01
9.74100411e-01 3.89453650e-01 4.17368300e-03 2.08706558e-01
6.82375550e-01 -9.03397202e-01 5.49620867e-01 7.53902495e-01
5.69465458e-01 -7.56077111e-01 2.06106797e-01 1.15098976e-01
-1.45199764e+00 -4.09402400e-01 1.85158893e-01 -4.41538215e-01
1.12668961e-01 8.20000589e-01 -1.13037694e+00 1.61524594e-01
5.38123131e-01 5.09455740e-01 -7.06273437e-01 9.17674065e-01
-9.90717486e-02 1.03088343e+00 -3.47474724e-01 -2.66381055e-01
1.47536308e-01 1.63813844e-01 1.34441126e-02 1.13905883e+00
3.97962183e-01 1.50681689e-01 2.09876716e-01 7.61717319e-01
2.12530270e-01 2.27847815e-01 -7.70263553e-01 1.36510715e-01
8.64325285e-01 7.00779796e-01 -7.74148047e-01 1.15883037e-01
-6.72408164e-01 -1.31833479e-01 -9.83867347e-02 4.10400391e-01
-7.95468807e-01 9.92072280e-03 9.24804389e-01 7.12475181e-01
-2.79970586e-01 -1.18131854e-01 -6.84834301e-01 -3.57113123e-01
-7.10273683e-01 -2.71532714e-01 4.37322199e-01 -1.86758012e-01
-7.99317658e-01 -5.47560811e-01 2.73513556e-01 -4.96186018e-01
2.48177782e-01 -4.82824333e-02 -1.08056271e+00 6.40306175e-01
-1.64768755e+00 -1.19638705e+00 -4.60153222e-01 8.92622620e-02
1.46435186e-01 3.49137068e-01 6.73512399e-01 3.71522635e-01
-9.63676929e-01 1.34217009e-01 5.90935767e-01 7.83001781e-02
1.22293912e-01 -8.81060183e-01 3.12400609e-01 6.47253335e-01
-5.62692821e-01 3.47678125e-01 1.86598599e-01 -1.14112914e+00
-9.35111105e-01 -1.56663156e+00 1.12867033e+00 -3.70012713e-03
3.70915830e-01 7.56417066e-02 -5.10474183e-02 5.08078337e-01
1.94580015e-02 -7.59754300e-01 1.35093176e+00 6.65513575e-01
1.47529736e-01 -2.62515306e-01 -1.30674851e+00 9.72070813e-01
8.20048571e-01 -3.27746809e-01 9.46495831e-02 4.83086914e-01
7.63803601e-01 2.75657654e-01 -9.85634923e-01 2.27968961e-01
6.41678095e-01 -8.85729074e-01 9.21766758e-01 -1.69184968e-01
5.01569927e-01 9.35601294e-02 -2.62080520e-01 -1.18078506e+00
-7.66517043e-01 2.62289077e-01 6.54302120e-01 8.49099576e-01
7.62827933e-01 -6.99919343e-01 7.28068352e-01 1.28550410e+00
-1.69746369e-01 -7.26213455e-01 -7.37473071e-01 -1.69020623e-01
3.69270891e-01 -5.31195104e-01 1.04251611e+00 7.90676534e-01
-5.19816428e-02 -1.28574207e-01 2.03046188e-01 4.62482750e-01
4.69232917e-01 -5.86770251e-02 6.28167033e-01 -9.67167258e-01
2.88507819e-01 -5.70171297e-01 -2.89051291e-02 -1.28355131e-01
9.66038704e-02 -6.67434812e-01 -1.82544097e-01 -2.03711820e+00
5.10760188e-01 -1.03579187e+00 -2.50924885e-01 6.44633949e-01
-6.62401199e-01 -2.44130403e-01 -5.70812225e-02 -4.92030121e-02
4.34906408e-02 5.29850543e-01 7.50242949e-01 -5.81302345e-01
-6.57256663e-01 3.17736357e-01 -1.22152722e+00 3.50724339e-01
9.67068434e-01 -5.05289257e-01 -3.66575539e-01 -4.21823651e-01
1.07165061e-01 -1.56770363e-01 7.05414116e-01 -1.10387421e+00
5.76732941e-02 -9.79525864e-01 4.72741902e-01 -5.54035902e-01
-1.25873953e-01 -9.00958240e-01 5.39458811e-01 1.21507430e+00
-5.70813179e-01 -1.43779889e-01 2.70849198e-01 6.90417051e-01
5.20385921e-01 4.50519472e-01 4.65824574e-01 8.55836198e-02
-2.82445759e-01 1.82522148e-01 -1.04797137e+00 -1.97438255e-01
1.11019647e+00 -2.60766655e-01 -4.72830832e-01 -1.35319009e-01
-4.24128592e-01 4.33181971e-01 5.83711445e-01 8.07002336e-02
2.85675704e-01 -1.41889548e+00 -7.94117272e-01 6.73302189e-02
9.13190246e-02 -4.67149258e-01 3.65902662e-01 7.58196175e-01
-3.50631326e-01 4.73354518e-01 -9.05486196e-03 -9.38348621e-02
-8.42879295e-01 1.16635136e-01 5.06428778e-01 -6.07841685e-02
-3.34747076e-01 2.73208320e-01 6.64963424e-02 -3.57710063e-01
-7.48146102e-02 -3.44407529e-01 6.86098402e-03 -5.73276877e-02
2.59517312e-01 1.35959828e+00 -3.04937601e-01 -4.56208259e-01
-4.92814690e-01 1.37408242e-01 7.77351856e-01 6.92478120e-02
1.30762863e+00 -4.83935565e-01 3.23235765e-02 2.78437287e-01
9.45501506e-01 3.37496340e-01 -8.84728670e-01 6.02004156e-02
6.18298948e-02 -3.49375039e-01 3.01960766e-01 -6.73419058e-01
-7.41862416e-01 2.07301974e-01 8.80763292e-01 1.49734780e-01
1.13521802e+00 3.30514573e-02 7.02252030e-01 1.37122333e-01
-1.81189612e-01 -1.09712243e+00 -2.64610499e-01 -1.17053688e-01
3.79166245e-01 -1.33870101e+00 2.50622869e-01 -1.03660434e-01
7.59252580e-03 6.04520202e-01 3.76599252e-01 2.99432397e-01
1.26408625e+00 4.07161703e-03 -1.87102586e-01 -1.57281145e-01
-4.30761963e-01 -4.88047510e-01 3.31724039e-03 5.98587751e-01
5.86565375e-01 9.00141478e-01 -2.95076787e-01 5.86370118e-02
-2.16460414e-02 -2.28165522e-01 1.10500306e-01 6.22034013e-01
-9.56689000e-01 -8.63749146e-01 -8.31589580e-01 1.15220547e+00
-4.80789952e-02 -3.63357842e-01 -2.78877586e-01 1.14419019e+00
5.28544188e-01 1.46128321e+00 1.13034576e-01 -2.13443205e-01
3.10737938e-01 -1.13157883e-01 -2.74710864e-01 -5.30922532e-01
-5.88721693e-01 -2.41509810e-01 3.62664491e-01 -1.35053113e-01
-3.95018905e-01 -8.78465831e-01 -1.26195788e+00 -8.35154235e-01
-2.76644975e-01 -9.71304774e-02 7.32584536e-01 9.41607952e-01
4.82304662e-01 5.79940796e-01 7.99505770e-01 -2.71864653e-01
2.33570471e-01 -7.07773805e-01 -3.11812669e-01 4.40894067e-03
3.51889133e-01 -3.44499469e-01 -1.01171717e-01 -3.18049431e-01] | [6.505660057067871, 1.9556492567062378] |
bcd29956-3084-46ea-85a4-920e90606d1e | pmp-net-point-cloud-completion-by-transformer | 2202.09507 | null | https://arxiv.org/abs/2202.09507v3 | https://arxiv.org/pdf/2202.09507v3.pdf | PMP-Net++: Point Cloud Completion by Transformer-Enhanced Multi-step Point Moving Paths | Point cloud completion concerns to predict missing part for incomplete 3D shapes. A common strategy is to generate complete shape according to incomplete input. However, unordered nature of point clouds will degrade generation of high-quality 3D shapes, as detailed topology and structure of unordered points are hard to be captured during the generative process using an extracted latent code. We address this problem by formulating completion as point cloud deformation process. Specifically, we design a novel neural network, named PMP-Net++, to mimic behavior of an earth mover. It moves each point of incomplete input to obtain a complete point cloud, where total distance of point moving paths (PMPs) should be the shortest. Therefore, PMP-Net++ predicts unique PMP for each point according to constraint of point moving distances. The network learns a strict and unique correspondence on point-level, and thus improves quality of predicted complete shape. Moreover, since moving points heavily relies on per-point features learned by network, we further introduce a transformer-enhanced representation learning network, which significantly improves completion performance of PMP-Net++. We conduct comprehensive experiments in shape completion, and further explore application on point cloud up-sampling, which demonstrate non-trivial improvement of PMP-Net++ over state-of-the-art point cloud completion/up-sampling methods. | ['Zhizhong Han', 'Yu-Shen Liu', 'Wen Zheng', 'Pengfei Wan', 'Yan-Pei Cao', 'Peng Xiang', 'Xin Wen'] | 2022-02-19 | null | null | null | null | ['point-cloud-completion'] | ['computer-vision'] | [-1.16198938e-02 -6.43905550e-02 1.29925936e-01 1.70931611e-02
-6.16848648e-01 -6.06865704e-01 4.80450571e-01 -2.74052560e-01
3.27299327e-01 4.00965005e-01 -8.28481764e-02 -2.06039146e-01
-6.13097101e-02 -1.26108098e+00 -1.15404558e+00 -4.04131740e-01
7.45495141e-04 1.13924098e+00 2.92351041e-02 -3.70630115e-01
1.18703321e-01 1.17825985e+00 -1.31657732e+00 1.77473545e-01
1.00747013e+00 6.75180197e-01 5.15550077e-01 4.09389228e-01
-4.11258429e-01 -4.07101698e-02 -2.80343443e-01 -3.25548798e-01
5.06602764e-01 2.33825967e-01 -5.36503971e-01 6.46342933e-02
3.04874539e-01 -4.97551858e-01 -5.15683591e-01 6.84184670e-01
2.55124956e-01 -3.70804034e-02 1.01908517e+00 -1.39138424e+00
-1.06005633e+00 2.50392526e-01 -6.78436279e-01 -6.02692783e-01
2.08245188e-01 1.89153194e-01 7.34418273e-01 -1.34324908e+00
7.80491173e-01 1.24554312e+00 9.26828027e-01 5.32599866e-01
-1.23721170e+00 -8.01466644e-01 6.39288276e-02 -3.51547182e-01
-1.50791395e+00 1.21090710e-01 1.29703200e+00 -4.06568259e-01
7.29530931e-01 3.98003966e-01 1.03256392e+00 8.47055197e-01
1.90825850e-01 6.70235097e-01 3.71918231e-01 1.60137168e-03
7.91363791e-02 -3.94740492e-01 -3.90355200e-01 4.06574816e-01
2.42096826e-01 2.05666050e-01 -4.79806103e-02 -4.44019824e-01
1.43989277e+00 5.27210534e-01 -2.18872353e-01 -7.47446597e-01
-1.28779960e+00 5.28061032e-01 8.33460629e-01 3.18960808e-02
-4.37894881e-01 3.69633049e-01 -7.02315290e-03 -1.34718018e-02
5.88393867e-01 -9.81464284e-04 -6.01296723e-01 -5.98852746e-02
-9.29136395e-01 6.67747855e-01 5.36008179e-01 1.61266100e+00
1.10698414e+00 1.66080594e-01 1.35381725e-02 6.85406089e-01
4.57287610e-01 8.12623262e-01 -5.84681332e-02 -9.47506130e-01
8.32094491e-01 1.06683373e+00 1.16975814e-01 -1.41220844e+00
-1.76393345e-01 -4.76134807e-01 -1.49099851e+00 3.49340796e-01
-1.67127520e-01 1.68802917e-01 -8.94254923e-01 1.31583905e+00
3.93708229e-01 4.11597192e-01 -1.58566847e-01 7.23232627e-01
7.44541228e-01 1.07024014e+00 -3.75354379e-01 -6.21783063e-02
9.27012146e-01 -6.24903560e-01 -1.29021555e-01 1.90346614e-01
3.26016307e-01 -7.04199195e-01 9.76991951e-01 1.48878604e-01
-1.24711120e+00 -7.65402317e-01 -7.83572555e-01 -1.66789532e-01
1.24948353e-01 1.46482959e-01 6.23556077e-01 8.97124708e-02
-8.76103282e-01 1.06428599e+00 -9.55554008e-01 2.98171341e-01
7.04326272e-01 4.70382810e-01 -3.17178190e-01 -2.66708821e-01
-5.66574216e-01 2.71855444e-01 -2.18399223e-02 2.72345394e-01
-8.54604959e-01 -1.34930134e+00 -6.66334569e-01 1.02775812e-01
-1.15425013e-01 -1.14805377e+00 9.14799571e-01 -3.29413474e-01
-1.19564176e+00 6.21529460e-01 -2.17400596e-01 -4.50822450e-02
5.85757613e-01 -1.16346389e-01 -1.13815777e-01 -1.01034135e-01
1.67521492e-01 1.05306017e+00 8.98837328e-01 -1.91729653e+00
-2.71182239e-01 -3.42410326e-01 -2.09486946e-01 2.19075501e-01
1.37327224e-01 -5.99973381e-01 -6.17268085e-01 -7.07413912e-01
6.26315475e-01 -1.07366395e+00 -4.02932227e-01 3.06777507e-01
-4.68961567e-01 -4.82028306e-01 1.16260433e+00 -5.44206679e-01
7.50858545e-01 -2.02218485e+00 8.66978243e-02 2.47746557e-01
3.66708010e-01 4.07731496e-02 -2.58221626e-01 7.11920381e-01
-1.80392474e-01 2.79095680e-01 -6.30265951e-01 -6.72379434e-01
6.22797832e-02 4.33489978e-01 -8.59833539e-01 1.60105228e-01
4.17436033e-01 1.08882213e+00 -7.90090024e-01 -1.81273088e-01
3.68472576e-01 6.77168906e-01 -7.95877755e-01 8.61288980e-02
-5.14218628e-01 5.63860714e-01 -6.03175282e-01 9.45412219e-01
1.45149374e+00 -1.89657584e-01 -5.00350296e-01 -3.09261620e-01
-1.15261570e-01 -2.10047171e-01 -1.15912616e+00 2.12789559e+00
-5.50267041e-01 1.03324912e-01 -1.65425539e-01 -5.51391542e-01
1.34188461e+00 2.28572398e-01 7.52774239e-01 -2.29435995e-01
-3.13633889e-01 2.86286175e-01 -4.14916992e-01 -9.45949703e-02
7.04558313e-01 -6.82450458e-02 1.62450656e-01 2.42919147e-01
-4.32184726e-01 -7.49288559e-01 -4.25242484e-01 2.49830768e-01
8.31904411e-01 6.11531436e-01 -2.56253839e-01 7.79623762e-02
3.41981411e-01 2.22893327e-01 6.54075980e-01 2.52159625e-01
4.38793123e-01 1.24755001e+00 2.93131053e-01 -7.03785419e-01
-1.53159368e+00 -1.32335281e+00 -1.62744597e-01 2.49254391e-01
4.13536668e-01 -2.33274207e-01 -4.15470749e-01 -3.31557959e-01
2.84837335e-01 6.54149234e-01 -3.95545185e-01 2.61836648e-02
-9.66351271e-01 -4.57682014e-01 2.39505485e-01 5.84102452e-01
2.13721380e-01 -1.12333477e+00 -1.43204376e-01 3.11487436e-01
-3.34407389e-02 -7.87692189e-01 -5.51707685e-01 -3.12470227e-01
-1.48482549e+00 -9.54681814e-01 -8.33941579e-01 -8.55918109e-01
1.10699332e+00 3.70450765e-01 1.27003002e+00 4.38067317e-01
2.96725295e-02 2.08114833e-01 -2.71989524e-01 -2.55039811e-01
-3.44721884e-01 1.80115163e-01 1.80296227e-02 -1.40803546e-01
3.03970277e-02 -1.26120448e+00 -7.28645504e-01 3.55132252e-01
-1.03514671e+00 4.24952865e-01 6.66084528e-01 6.66761100e-01
1.04029548e+00 8.20020866e-03 2.32021049e-01 -4.12817329e-01
3.36584955e-01 -4.44250375e-01 -5.11271298e-01 -1.72146931e-02
-1.32446155e-01 -1.08134285e-01 6.96595013e-01 -3.59106213e-01
-7.43194163e-01 2.89775550e-01 -4.24031258e-01 -1.26686144e+00
-2.29028892e-02 1.89456105e-01 -2.73154050e-01 -1.92873091e-01
4.71132308e-01 6.17947221e-01 -1.38028935e-01 -7.01119483e-01
3.67477685e-01 2.49222681e-01 6.13849938e-01 -8.80908132e-01
1.52439547e+00 7.19504476e-01 3.79059553e-01 -6.34424210e-01
-2.03759342e-01 -2.14171559e-01 -9.19393361e-01 -4.94817011e-02
4.51372653e-01 -9.83935237e-01 -9.05074716e-01 2.80978501e-01
-1.71001649e+00 -6.24453239e-02 -4.08784568e-01 1.46077827e-01
-7.19103873e-01 4.08897817e-01 -3.93106282e-01 -6.32720947e-01
-5.48982382e-01 -1.00108933e+00 1.53640413e+00 -1.50837734e-01
1.40231177e-01 -7.40434587e-01 2.42991462e-01 -3.49042974e-02
5.17203622e-02 6.59233868e-01 1.03020179e+00 -6.76099136e-02
-1.30183923e+00 -3.25374901e-01 -1.47924140e-01 1.48341164e-01
1.92929417e-01 2.43209168e-01 -6.37267590e-01 -3.89939845e-01
-5.15781790e-02 2.17608243e-01 5.09257436e-01 3.69553179e-01
1.56649888e+00 -3.97110581e-01 -6.38499081e-01 1.06861985e+00
1.58741462e+00 -1.34396926e-01 7.57293463e-01 -1.02197841e-01
1.12263107e+00 3.28268498e-01 3.43321234e-01 5.94293058e-01
5.01416504e-01 6.52755201e-01 9.76237178e-01 7.19338432e-02
-1.43838838e-01 -9.30873811e-01 -1.15651608e-01 1.29707515e+00
-3.78268093e-01 1.23882681e-01 -1.02650607e+00 5.49917221e-01
-1.63312399e+00 -7.34761596e-01 -4.43829149e-01 2.11477041e+00
4.52265799e-01 -1.76631764e-01 -2.16105968e-01 1.32816583e-02
7.56834269e-01 -2.12399233e-02 -5.68762422e-01 2.11059272e-01
5.51173799e-02 2.20767528e-01 1.70074657e-01 3.43216836e-01
-5.54134011e-01 7.76099861e-01 5.24561930e+00 1.12169790e+00
-9.28180218e-01 -9.67175290e-02 3.65683168e-01 1.80781499e-01
-8.83030415e-01 2.68132806e-01 -7.04202592e-01 5.68107963e-01
1.00403100e-01 -1.49731487e-01 4.39560860e-01 1.13819289e+00
1.91967234e-01 7.15512455e-01 -1.10524487e+00 1.26500583e+00
-1.29159436e-01 -1.73481953e+00 6.03948414e-01 2.63769507e-01
9.99805748e-01 1.68698207e-01 -7.61104971e-02 2.90535867e-01
2.05417499e-01 -1.06602013e+00 6.41757190e-01 7.76463449e-01
1.05340707e+00 -9.45226610e-01 2.11796016e-01 7.94151843e-01
-1.45716107e+00 3.07686239e-01 -1.00599408e+00 7.74439797e-02
4.39015865e-01 8.76428545e-01 -7.35099673e-01 1.12767529e+00
4.79060411e-01 9.48727906e-01 -1.39891595e-01 1.11271155e+00
1.09638842e-02 1.92076802e-01 -5.43173552e-01 2.46072188e-01
9.29620564e-02 -6.32689297e-01 7.96045303e-01 6.03086948e-01
9.01545405e-01 2.53824532e-01 2.54389495e-01 1.38695252e+00
-2.62176007e-01 -8.76745284e-02 -9.27125514e-01 2.87699044e-01
7.96448112e-01 1.31996989e+00 -4.85160112e-01 -1.13951810e-01
-5.27694859e-02 7.63392746e-01 3.57723683e-01 2.75531590e-01
-7.00847268e-01 -5.53670563e-02 6.55569017e-01 3.99431527e-01
3.22511852e-01 -5.60955167e-01 -5.60527325e-01 -1.04641640e+00
2.83054680e-01 -3.20772558e-01 -2.86693186e-01 -1.20555735e+00
-1.37538981e+00 6.91201448e-01 -6.76025152e-02 -2.03591919e+00
1.09726004e-01 -4.17707264e-01 -8.78438652e-01 9.67790902e-01
-1.42409778e+00 -1.68027842e+00 -5.85448980e-01 4.90672112e-01
6.36904061e-01 -7.62636885e-02 6.70798182e-01 1.51588529e-01
-1.30664796e-01 2.24851787e-01 3.98719348e-02 -2.20253505e-02
5.85701689e-02 -9.92990553e-01 1.03869450e+00 4.23140645e-01
1.39837980e-01 6.67472064e-01 2.28760496e-01 -1.18839979e+00
-1.76470590e+00 -1.52293849e+00 5.05990386e-01 -6.95641935e-01
3.13331708e-02 -2.33454466e-01 -9.64615226e-01 7.30538011e-01
-3.34499687e-01 1.66227371e-01 3.81389484e-02 -2.14083359e-01
-8.11105445e-02 -1.05010547e-01 -1.03132641e+00 7.29610145e-01
1.45878410e+00 -1.70801491e-01 -6.89503968e-01 4.93394047e-01
1.22733784e+00 -8.49681735e-01 -1.00172150e+00 7.06853449e-01
2.08355486e-01 -6.94296241e-01 1.40165985e+00 -3.33337694e-01
8.90147984e-01 -5.57623148e-01 -1.38457179e-01 -1.37302589e+00
-5.38083136e-01 -6.55646265e-01 -2.84144998e-01 1.19120741e+00
9.73804966e-02 -3.70058686e-01 1.24861920e+00 3.94302398e-01
-7.40680993e-01 -1.14190471e+00 -9.19741869e-01 -7.72173226e-01
4.29361969e-01 -6.25940979e-01 1.42403579e+00 9.56103146e-01
-6.24565899e-01 -1.32066667e-01 -3.89481038e-01 4.38234419e-01
6.91735446e-01 4.47450817e-01 1.20167363e+00 -1.41049922e+00
-6.44885376e-02 -1.76262647e-01 -3.01497757e-01 -1.61069477e+00
3.46376486e-02 -1.06106949e+00 -2.00059921e-01 -1.68361831e+00
-1.62029162e-01 -1.12635410e+00 3.70880693e-01 3.65498811e-01
1.19438283e-01 2.00919300e-01 2.81074435e-01 7.40887046e-01
2.41599932e-01 1.10741985e+00 1.84304047e+00 -2.21555814e-01
-3.35113883e-01 2.40782142e-01 -3.41856688e-01 6.96998894e-01
3.98545682e-01 -5.53091407e-01 -4.18100655e-01 -8.75335336e-01
4.68491584e-01 4.21315461e-01 4.95570868e-01 -1.09691560e+00
3.04629683e-01 -3.42366189e-01 6.72387719e-01 -1.54326355e+00
6.75750494e-01 -1.14609623e+00 8.14976513e-01 3.00370425e-01
2.55552769e-01 4.04567122e-01 1.28323272e-01 6.42814875e-01
-3.09295510e-03 -1.87189609e-01 2.37546757e-01 -3.71902883e-01
-1.58693850e-01 1.16191328e+00 5.45979857e-01 -3.35927308e-01
9.19141591e-01 -5.28987169e-01 -3.08142025e-02 1.86815280e-02
-4.95166600e-01 2.35368669e-01 8.04056346e-01 4.68599737e-01
1.09408247e+00 -1.95633197e+00 -8.03946555e-01 4.58897233e-01
-9.71648551e-04 1.13388920e+00 5.26029587e-01 2.74286985e-01
-9.66377676e-01 1.59514636e-01 -3.10487766e-03 -9.93924141e-01
-6.14270926e-01 5.60182571e-01 1.07826352e-01 1.49845630e-02
-1.17385578e+00 5.19919753e-01 3.51986885e-01 -1.07351458e+00
-2.38244370e-01 -6.26356542e-01 2.65010446e-01 -4.79301125e-01
8.19722489e-02 3.77252907e-01 8.69332626e-02 -5.20114839e-01
-2.70433575e-02 1.08562148e+00 1.40262514e-01 1.57703161e-01
1.61612010e+00 2.64809489e-01 -2.18854010e-01 5.95164075e-02
1.06028104e+00 7.03778490e-02 -1.48253489e+00 -3.72686461e-02
-7.12931216e-01 -7.22052038e-01 -3.47975463e-01 -3.25366199e-01
-1.12410903e+00 8.10843289e-01 2.33839862e-02 -4.21211198e-02
7.91788936e-01 -1.73077255e-01 1.10567856e+00 2.79490650e-01
7.13252544e-01 -5.12173355e-01 2.16414817e-02 4.81717527e-01
1.50139248e+00 -7.99040973e-01 5.13870921e-03 -7.59642780e-01
-4.23374176e-01 1.19768453e+00 6.72877431e-01 -6.05389535e-01
7.63416588e-01 -3.17406990e-02 -4.72070128e-01 -3.17396969e-01
-5.15983999e-01 4.52050418e-01 3.18536192e-01 8.09632540e-01
-2.62947649e-01 1.99167624e-01 2.49364242e-01 4.28150386e-01
-6.05678499e-01 6.67328909e-02 3.10961008e-01 6.69914961e-01
-1.73916802e-01 -1.12812757e+00 -6.94964707e-01 4.42520082e-01
4.26604062e-01 -2.02209111e-02 1.34636387e-02 6.96470082e-01
2.46378496e-01 1.71674058e-01 3.54382128e-01 -4.42167133e-01
4.80299175e-01 -3.15816343e-01 2.59518743e-01 -5.87693930e-01
-2.21021712e-01 -1.62544310e-01 -6.01358473e-01 -3.08499426e-01
-2.16038525e-02 -6.34162009e-01 -1.30489361e+00 -5.27890861e-01
-4.99628298e-02 6.01727851e-02 6.72659934e-01 5.45318365e-01
6.91284359e-01 3.91201347e-01 1.04806602e+00 -1.49291480e+00
-5.89346170e-01 -7.95293391e-01 -4.63268667e-01 6.06184304e-01
1.72348797e-01 -5.86424351e-01 -1.65039480e-01 -6.08460791e-02] | [8.426887512207031, -3.594987392425537] |
d1f658a8-7ce7-4e6f-92fa-f7db57efa4df | masked-path-modeling-for-vision-and-language | 2305.14268 | null | https://arxiv.org/abs/2305.14268v1 | https://arxiv.org/pdf/2305.14268v1.pdf | Masked Path Modeling for Vision-and-Language Navigation | Vision-and-language navigation (VLN) agents are trained to navigate in real-world environments by following natural language instructions. A major challenge in VLN is the limited availability of training data, which hinders the models' ability to generalize effectively. Previous approaches have attempted to address this issue by introducing additional supervision during training, often requiring costly human-annotated data that restricts scalability. In this paper, we introduce a masked path modeling (MPM) objective, which pretrains an agent using self-collected data for downstream navigation tasks. Our proposed method involves allowing the agent to actively explore navigation environments without a specific goal and collect the paths it traverses. Subsequently, we train the agent on this collected data to reconstruct the original path given a randomly masked subpath. This way, the agent can actively accumulate a diverse and substantial amount of data while learning conditional action generation. To evaluate the effectiveness of our technique, we conduct experiments on various VLN datasets and demonstrate the versatility of MPM across different levels of instruction complexity. Our results exhibit significant improvements in success rates, with enhancements of 1.32\%, 1.05\%, and 1.19\% on the val-unseen split of the Room-to-Room, Room-for-Room, and Room-across-Room datasets, respectively. Furthermore, we conduct an analysis that highlights the potential for additional improvements when the agent is allowed to explore unseen environments prior to testing. | ['Nanyun Peng', 'Feng Gao', 'Zi-Yi Dou'] | 2023-05-23 | null | null | null | null | ['action-generation', 'navigate', 'vision-and-language-navigation'] | ['computer-vision', 'reasoning', 'robots'] | [ 3.40532094e-01 1.09507732e-01 2.76701227e-02 -5.10874927e-01
-7.64777184e-01 -5.93301713e-01 6.25479758e-01 -2.36170664e-02
-8.36891353e-01 8.88557196e-01 7.33435825e-02 -7.49031961e-01
2.11645141e-01 -8.20156217e-01 -1.09409463e+00 -6.32227719e-01
-8.10026675e-02 3.79738390e-01 3.24842155e-01 -4.43804711e-01
3.03038359e-01 3.40408027e-01 -1.62796724e+00 -7.70053566e-02
1.20565140e+00 4.69917893e-01 7.71075487e-01 7.98902333e-01
-2.00325139e-02 9.93903697e-01 -5.28067648e-01 2.90510893e-01
3.89299154e-01 -4.64599520e-01 -5.71468592e-01 -5.78810945e-02
4.05464619e-01 -4.81279254e-01 -3.49257499e-01 7.89851964e-01
2.96884388e-01 6.52002811e-01 3.88021052e-01 -1.16733801e+00
-2.95474231e-01 3.46892089e-01 -3.37015688e-01 -5.27272671e-02
4.17536825e-01 6.83099210e-01 6.97233438e-01 -6.36458158e-01
6.16598308e-01 1.17307687e+00 2.63675928e-01 9.55257535e-01
-1.25640976e+00 -5.78567684e-01 4.03975904e-01 -4.15154546e-02
-9.10433471e-01 -4.84263003e-01 4.50082034e-01 -2.92043447e-01
1.18196464e+00 -1.86700076e-01 2.73650289e-01 1.35724771e+00
2.64187127e-01 8.11102211e-01 1.11298490e+00 -4.72103298e-01
4.23969775e-01 -2.69785449e-02 2.32543852e-02 1.03198278e+00
1.66106865e-01 4.65464979e-01 -5.51406622e-01 2.57382751e-01
6.37951612e-01 -2.11527869e-01 -3.69525433e-01 -6.57361269e-01
-1.19836295e+00 7.73775637e-01 6.64902210e-01 -2.05648839e-01
-2.72316813e-01 1.16729900e-01 2.02558398e-01 4.66484688e-02
-2.28075996e-01 6.42686546e-01 -4.14787829e-01 -4.11651880e-01
-6.10925138e-01 2.26479217e-01 7.62273312e-01 1.12028801e+00
8.98182869e-01 2.31878027e-01 2.92809233e-02 5.23635387e-01
3.75124395e-01 3.76214296e-01 5.66868305e-01 -1.21581066e+00
8.32684755e-01 4.67400551e-01 1.94524854e-01 -5.55741668e-01
-5.06969810e-01 -1.76699713e-01 -4.39084858e-01 7.40729630e-01
5.25255203e-01 -3.56367767e-01 -1.38943326e+00 2.12801886e+00
3.12285095e-01 5.01426272e-02 5.07907331e-01 6.95419848e-01
5.61347544e-01 6.24845862e-01 1.81888103e-01 1.61001667e-01
8.09306622e-01 -1.59573197e+00 -4.94046718e-01 -8.34841967e-01
8.50793779e-01 -3.56118470e-01 1.54602742e+00 3.25910330e-01
-8.09943914e-01 -6.63815856e-01 -1.17107952e+00 -1.94797348e-02
-2.40711451e-01 -6.78525344e-02 5.61547399e-01 3.76608789e-01
-1.20579970e+00 3.81899089e-01 -1.16473353e+00 -4.39534545e-01
3.38119775e-01 3.73337299e-01 -5.27908981e-01 -4.91983443e-01
-7.18717813e-01 7.62500405e-01 4.73587304e-01 5.79389073e-02
-1.57609260e+00 -3.91360044e-01 -1.25874627e+00 -6.21442571e-02
4.85150486e-01 -4.75714803e-01 1.31960261e+00 -3.68240297e-01
-1.52619004e+00 3.09253722e-01 -3.13696414e-01 -4.46081102e-01
4.83753949e-01 -3.74814004e-01 -1.28275186e-01 -2.21860841e-01
3.47952187e-01 9.80193794e-01 2.18961418e-01 -1.62634385e+00
-8.03944111e-01 -2.85869747e-01 3.91184568e-01 4.68809813e-01
9.58130285e-02 -7.31151879e-01 -5.44265807e-01 -1.47494107e-01
2.07477391e-01 -1.19823194e+00 -5.56382656e-01 -2.45461926e-01
-4.14843738e-01 -4.35619568e-03 5.44488847e-01 -4.13131118e-01
8.48113477e-01 -2.15132737e+00 8.14322457e-02 -2.72576809e-02
-8.34040046e-02 -3.18307988e-02 -5.36563575e-01 3.04694384e-01
4.55101758e-01 -7.47729763e-02 -3.62722158e-01 -6.80098593e-01
-1.64536595e-01 6.28334522e-01 -2.01954424e-01 1.44389058e-02
-1.08573874e-02 7.02093601e-01 -9.90838230e-01 -2.85608143e-01
1.45320147e-01 2.40993083e-01 -9.32886839e-01 3.61739814e-01
-5.10007620e-01 9.53295767e-01 -4.41306472e-01 5.03220379e-01
3.88505340e-01 -4.91407402e-02 1.83466509e-01 3.54381412e-01
-2.96503484e-01 4.78663832e-01 -8.77059460e-01 2.10816288e+00
-8.53261471e-01 6.81934834e-01 1.26369163e-01 -7.15205133e-01
8.15389514e-01 -1.48456484e-01 -1.79416090e-02 -1.02543902e+00
-1.26601294e-01 2.37725750e-01 1.52664661e-01 -4.62979704e-01
7.63172448e-01 2.05130041e-01 -2.40999624e-01 2.51562804e-01
-1.83853582e-01 -1.99285895e-03 2.21127316e-01 8.00683796e-02
1.36446810e+00 4.96354580e-01 1.58813559e-02 -4.53315116e-02
4.01644558e-01 5.08847356e-01 4.91031289e-01 1.17497957e+00
-4.02379394e-01 3.08249205e-01 1.62402123e-01 -2.87745208e-01
-8.04234266e-01 -1.21370113e+00 2.37604693e-01 1.27589762e+00
3.37847322e-01 -2.30803728e-01 -7.17035472e-01 -8.46150339e-01
-2.66698539e-01 1.24671209e+00 -5.12327433e-01 -2.01668605e-01
-9.85360086e-01 -5.63750386e-01 3.79688203e-01 6.61870122e-01
6.75270438e-01 -1.40351224e+00 -1.08696234e+00 7.36651719e-02
-3.78385544e-01 -1.19556510e+00 -3.23732138e-01 5.05714893e-01
-8.43728960e-01 -1.01931632e+00 -1.19800568e-02 -9.81673658e-01
9.42005932e-01 4.11445022e-01 1.07171249e+00 2.51564831e-02
-1.82620600e-01 4.14527506e-01 -1.60273895e-01 -8.47138241e-02
-5.37470818e-01 1.63495094e-01 1.06168814e-01 -5.42826772e-01
2.50796169e-01 -4.25869167e-01 -5.35616696e-01 2.85095036e-01
-4.90830392e-01 1.94793850e-01 7.45110631e-01 1.12841928e+00
6.29328966e-01 3.45453359e-02 3.67667675e-01 -7.78871179e-01
6.42542660e-01 -4.39935535e-01 -8.38880062e-01 -7.92587250e-02
-7.88639963e-01 4.43258882e-01 7.36889482e-01 -3.98127526e-01
-1.18213749e+00 2.37856552e-01 4.02983539e-02 -8.55930224e-02
-4.61219668e-01 3.49769503e-01 -3.12312335e-01 4.15634140e-02
7.28220344e-01 4.19360340e-01 -1.16391554e-01 -1.61942661e-01
4.43195492e-01 1.96119517e-01 6.71842039e-01 -8.36328804e-01
7.87753761e-01 3.59140247e-01 -5.68872578e-02 -6.23932004e-01
-6.35712624e-01 -2.16734856e-01 -3.58302981e-01 7.41708055e-02
1.05202079e+00 -9.13561761e-01 -7.56396651e-01 2.23407879e-01
-7.72327125e-01 -1.17004561e+00 -2.11563945e-01 6.54250622e-01
-6.90959811e-01 -9.62788463e-02 -4.20510352e-01 -7.28713930e-01
1.15926020e-01 -1.60889006e+00 6.67881727e-01 5.78454435e-01
-1.42691568e-01 -8.87589276e-01 5.87329678e-02 6.09779000e-01
4.41482067e-01 1.07886083e-01 9.57382143e-01 -5.46201944e-01
-9.92695689e-01 3.54484804e-02 2.30868161e-02 9.67852771e-02
2.08360210e-01 -5.11173844e-01 -8.69655907e-01 -5.09224594e-01
-1.61777198e-01 -7.37583220e-01 8.37024331e-01 1.79663792e-01
1.05249512e+00 -1.37123883e-01 -5.17780721e-01 6.09688580e-01
1.42912292e+00 5.75801373e-01 4.23545569e-01 8.06816816e-01
7.14967132e-01 6.72264278e-01 7.86225080e-01 1.79292917e-01
6.64670408e-01 3.25920910e-01 8.33096921e-01 1.83082625e-01
-3.73794697e-02 -5.71837068e-01 5.36964357e-01 5.33272505e-01
1.68187812e-01 -5.15873671e-01 -9.34622526e-01 7.03813493e-01
-1.71236062e+00 -6.08515084e-01 2.26819977e-01 2.31025171e+00
7.69554973e-01 4.38967466e-01 -3.68186653e-01 -2.68778890e-01
1.77246138e-01 1.69450298e-01 -8.90360713e-01 -2.78676808e-01
2.00961515e-01 1.73826143e-01 5.28191745e-01 9.70507920e-01
-9.10445988e-01 1.27161121e+00 5.84227037e+00 2.89462656e-01
-8.97159338e-01 -2.65485078e-01 3.31385404e-01 -1.32630214e-01
-1.46337509e-01 9.97039452e-02 -1.06711078e+00 1.52662978e-01
1.02461100e+00 2.11533368e-01 7.07414567e-01 9.86878276e-01
3.88523281e-01 -5.34643888e-01 -1.38091636e+00 5.61714172e-01
-2.08273605e-02 -1.02478826e+00 -1.34184480e-01 7.40478709e-02
7.35221684e-01 2.23423332e-01 2.17153788e-01 9.33215201e-01
8.95427227e-01 -1.18626201e+00 5.30661047e-01 1.89733297e-01
3.88238966e-01 -6.16203547e-01 4.75639790e-01 6.94009602e-01
-9.97555315e-01 -2.40879193e-01 -1.27860054e-01 -2.74507642e-01
1.91500112e-01 -1.96200505e-01 -1.29460990e+00 3.52539301e-01
6.58615649e-01 2.61739969e-01 -5.50349057e-01 8.45370412e-01
-5.54691792e-01 4.49262053e-01 -2.39776179e-01 -1.27201214e-01
5.88840187e-01 -2.44992748e-01 3.66023153e-01 8.42824519e-01
3.28644812e-01 -4.87796664e-02 6.42700791e-01 8.77152562e-01
5.02948128e-02 -3.34561586e-01 -7.06110775e-01 2.39979982e-01
5.35104573e-01 9.35953975e-01 -5.43230057e-01 -1.01548210e-01
-3.67550373e-01 1.05622065e+00 5.64668834e-01 7.21520305e-01
-8.61017048e-01 -1.88026980e-01 7.76348650e-01 -2.65254378e-02
1.99152172e-01 -6.32886171e-01 -2.43707478e-01 -6.79708958e-01
1.03943266e-01 -7.96528816e-01 7.85423890e-02 -7.04096615e-01
-7.63328016e-01 7.30282962e-01 -1.19193815e-01 -1.03708017e+00
-5.40620208e-01 -6.40897632e-01 -5.51957786e-01 9.73722935e-01
-1.73720253e+00 -8.02531719e-01 -8.09073508e-01 4.54354435e-01
1.02377903e+00 -2.58716583e-01 8.81833076e-01 -2.10493188e-02
-6.83150887e-01 6.17515564e-01 -4.40196618e-02 6.54444471e-02
6.95258260e-01 -1.32832646e+00 4.87143368e-01 9.70455468e-01
-1.59529820e-01 7.98601747e-01 8.11796427e-01 -6.30767226e-01
-1.41209853e+00 -1.23396254e+00 2.73129791e-01 -5.68704903e-01
3.40260059e-01 -3.50395501e-01 -7.32841551e-01 9.91233051e-01
1.23748243e-01 1.82389636e-02 5.46284199e-01 -7.66265318e-02
-3.07175457e-01 2.35399961e-01 -1.03248453e+00 1.12752056e+00
1.20294774e+00 -2.31245384e-01 -5.00943840e-01 7.03963861e-02
1.00775492e+00 -6.54642642e-01 -2.50543088e-01 2.54144102e-01
3.55613977e-01 -1.06604612e+00 8.41557860e-01 -6.25754058e-01
5.47559559e-01 -4.42962497e-01 -4.74453449e-01 -1.40040827e+00
2.18189997e-03 -3.68701011e-01 9.82432291e-02 8.92738640e-01
6.88228488e-01 -6.49629354e-01 1.06877923e+00 6.43052340e-01
-4.69957709e-01 -7.04557061e-01 -6.61950231e-01 -6.93482995e-01
6.80650920e-02 -4.11319107e-01 3.67693007e-01 5.87748945e-01
-2.47546360e-01 4.13423359e-01 -1.45281941e-01 5.56011081e-01
6.74733043e-01 -1.16887540e-01 1.15848505e+00 -8.30346644e-01
-3.15822929e-01 -2.06684723e-01 5.89313433e-02 -1.55124712e+00
3.61841261e-01 -6.95314050e-01 7.53544509e-01 -1.81022668e+00
-1.09658994e-01 -7.65375912e-01 -2.00921729e-01 5.47346711e-01
-1.59152687e-01 -3.87798935e-01 1.31238744e-01 5.08160181e-02
-6.72963142e-01 6.43274188e-01 1.43501282e+00 -1.03453666e-01
-6.08869433e-01 -1.60801336e-02 -4.91206884e-01 6.68890059e-01
9.42252219e-01 -1.83354735e-01 -9.90716875e-01 -7.99459577e-01
-3.75801772e-01 1.47910360e-02 2.51419544e-01 -1.32773077e+00
4.19847876e-01 -3.11359733e-01 3.41224879e-01 -6.04516208e-01
6.29417658e-01 -7.32524097e-01 -4.02791172e-01 4.91437405e-01
-3.38261247e-01 2.57370949e-01 5.47931194e-01 7.64793813e-01
-1.15351081e-02 -2.21559778e-01 5.33132553e-01 -3.33756328e-01
-1.23723233e+00 -8.03110525e-02 -4.83874530e-01 1.94249079e-01
1.06394589e+00 -3.70464981e-01 -3.93065065e-01 -4.13468480e-01
-5.51593840e-01 7.76475191e-01 7.01222658e-01 5.42700708e-01
7.23676801e-01 -1.04755580e+00 -3.02467555e-01 5.03541887e-01
1.31527662e-01 6.50830328e-01 2.54281133e-01 4.78468984e-01
-6.85571253e-01 4.63022053e-01 -2.33354464e-01 -6.64183974e-01
-9.35092688e-01 6.03160918e-01 2.75166005e-01 -3.69517297e-01
-6.11111343e-01 8.76997948e-01 4.55638528e-01 -9.38663065e-01
4.26274002e-01 -4.62163299e-01 -1.60691679e-01 -5.12028635e-01
3.07038218e-01 2.31784880e-01 -2.44973645e-01 -3.17103744e-01
-2.47466832e-01 2.61168957e-01 -2.48270318e-01 -3.92392814e-01
1.09847724e+00 -1.60526261e-01 2.95467257e-01 4.15895879e-01
8.78891885e-01 4.12948579e-02 -1.87746930e+00 -3.39520536e-02
-7.80735584e-03 -1.99680746e-01 -2.16200545e-01 -9.76570308e-01
-6.40208721e-01 8.35489571e-01 6.08628809e-01 -3.35388660e-01
8.51167202e-01 -2.45603517e-01 6.18634105e-01 6.65716052e-01
7.20438242e-01 -8.90879214e-01 3.88401777e-01 7.67946780e-01
5.40852487e-01 -1.68314183e+00 -3.66192579e-01 -1.09516196e-01
-5.83572865e-01 7.70265818e-01 1.30878842e+00 1.70582801e-01
8.55858028e-02 2.39018962e-01 2.58056164e-01 7.13615073e-03
-7.55810559e-01 -2.02150811e-02 -2.14928284e-01 8.34518194e-01
1.67959720e-01 4.61910740e-02 3.03331524e-01 2.69158244e-01
-4.52420890e-01 -2.71768451e-01 7.96891451e-01 1.43065190e+00
-6.30210817e-01 -9.43062961e-01 -3.10243547e-01 7.44588673e-02
1.73726454e-01 -1.17183939e-01 7.48867169e-02 1.00829852e+00
-4.54665497e-02 1.15084350e+00 1.36960344e-02 -3.96784931e-01
4.80893433e-01 8.73983204e-02 2.95385003e-01 -7.50119686e-01
-4.71077561e-02 -2.04580382e-01 7.41342008e-02 -9.12401915e-01
-5.51092885e-02 -5.23815930e-01 -1.73890436e+00 1.83533016e-03
1.14290826e-02 1.33492798e-01 6.88422740e-01 7.73849249e-01
3.67938638e-01 9.94710267e-01 3.20862174e-01 -9.45187747e-01
-7.15021431e-01 -7.50237346e-01 7.78051838e-02 1.48868456e-01
6.49789393e-01 -7.38900006e-01 -4.12789494e-01 -4.44896892e-02] | [4.495048522949219, 0.6228921413421631] |
957763f2-6be8-4655-8165-ae5a8752d424 | learning-to-see-moving-objects-in-the-dark | null | null | http://openaccess.thecvf.com/content_ICCV_2019/html/Jiang_Learning_to_See_Moving_Objects_in_the_Dark_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Jiang_Learning_to_See_Moving_Objects_in_the_Dark_ICCV_2019_paper.pdf | Learning to See Moving Objects in the Dark | Video surveillance systems have wide range of utilities, yet easily suffer from great quality degeneration under dim light circumstances. Industrial solutions mainly use extra near-infrared illuminations, even though it doesn't preserve color and texture information. A variety of researches enhanced low-light videos shot by visible light cameras, while they either relied on task specific preconditions or trained with synthetic datasets. We propose a novel optical system to capture bright and dark videos of the exact same scenes, generating training and groud truth pairs for authentic low-light video dataset. A fully convolutional network with 3D and 2D miscellaneous operations is utilized to learn an enhancement mapping with proper spatial-temporal transformation from raw camera sensor data to bright RGB videos. Experiments show promising results by our method, and it outperforms state-of-the-art low-light image/video enhancement algorithms.
| [' Yinqiang Zheng', 'Haiyang Jiang'] | 2019-10-01 | null | null | null | iccv-2019-10 | ['video-enhancement', 'miscellaneous'] | ['computer-vision', 'miscellaneous'] | [ 4.44398284e-01 -7.06339598e-01 9.21165943e-02 -3.43564540e-01
-1.89108655e-01 -4.94077921e-01 4.31039304e-01 -9.68353450e-01
-3.55956316e-01 7.72842646e-01 -9.55245495e-02 -5.99587373e-02
3.01621705e-01 -8.25061858e-01 -7.68518984e-01 -1.04514194e+00
3.93495858e-01 -6.53640032e-01 3.93707007e-01 -4.11481887e-01
1.59008831e-01 1.24903865e-01 -1.74025786e+00 4.13920432e-01
5.58376372e-01 1.02457154e+00 1.90980971e-01 8.18864167e-01
6.03319146e-02 1.04623365e+00 -2.70026803e-01 -6.64517522e-01
9.03721094e-01 -3.31822157e-01 -1.00710444e-01 4.68893915e-01
7.81059027e-01 -1.01187325e+00 -7.76063085e-01 1.61733723e+00
5.82666218e-01 -4.16401327e-02 3.09783723e-02 -1.31673086e+00
-1.20571673e+00 -2.40576014e-01 -8.92170370e-01 4.98353451e-01
7.51653016e-01 8.84349048e-01 2.39289999e-01 -9.19907928e-01
4.76135075e-01 1.00007200e+00 7.38236785e-01 6.89694226e-01
-7.06800878e-01 -8.97650778e-01 -2.45660141e-01 4.32304591e-01
-1.04742444e+00 -6.09132707e-01 9.38544154e-01 1.66692920e-02
7.62048841e-01 -2.07520486e-03 8.56871486e-01 1.28526497e+00
5.28209031e-01 4.01105940e-01 1.55198336e+00 -2.36908048e-01
-1.48423566e-02 -7.91427586e-03 -1.45052850e-01 8.18128824e-01
7.81169415e-01 7.65051305e-01 -4.59393650e-01 1.34165317e-01
7.92796075e-01 6.56676471e-01 -9.16134894e-01 -3.84078473e-02
-1.01060200e+00 1.39123410e-01 2.22896799e-01 1.74255818e-01
-2.82180071e-01 6.63715228e-02 8.14164802e-02 3.44844699e-01
2.85911113e-01 -1.66643575e-01 -2.87065119e-01 6.41831830e-02
-5.79662979e-01 -2.70731866e-01 2.97027230e-01 1.03533494e+00
9.58243906e-01 1.50588706e-01 -5.09362668e-03 4.88423884e-01
4.52225208e-01 9.80871797e-01 2.10124791e-01 -1.02105939e+00
4.50123191e-01 2.87469029e-01 3.52189362e-01 -9.84372556e-01
-2.22891167e-01 -1.62021980e-01 -1.00019979e+00 7.51774728e-01
1.48570180e-01 -3.00604910e-01 -8.33871186e-01 1.24631119e+00
3.12714815e-01 4.00748581e-01 3.68375659e-01 1.44078100e+00
9.01746631e-01 9.25528705e-01 -1.57194704e-01 -6.86149478e-01
1.07607555e+00 -8.44425499e-01 -1.00410199e+00 -2.66822278e-01
-6.32620975e-03 -1.00260413e+00 1.05581701e+00 8.19263697e-01
-1.44610751e+00 -9.45079029e-01 -9.07791853e-01 -1.89665779e-01
-1.45400658e-01 1.55585393e-01 7.01703548e-01 9.71624196e-01
-1.01435661e+00 1.67308316e-01 -5.62397122e-01 -4.71497513e-02
2.62305439e-01 -4.71148975e-02 -5.15157044e-01 -6.69762552e-01
-1.15998113e+00 6.72315955e-01 3.74536574e-01 5.37974715e-01
-1.18055010e+00 -6.27765179e-01 -7.80814409e-01 -2.88554847e-01
5.42062104e-01 -8.18071544e-01 6.51815891e-01 -1.03662813e+00
-1.69357228e+00 8.80801380e-01 1.56199768e-01 1.21323541e-01
3.58423233e-01 -1.74384758e-01 -7.06250906e-01 2.93654948e-01
-2.86944419e-01 1.41767085e-01 1.09215748e+00 -1.45703697e+00
-7.92553842e-01 -3.48272681e-01 3.51922601e-01 2.11292222e-01
-4.14188027e-01 2.07894862e-01 -6.52276516e-01 -3.15052003e-01
-8.12327042e-02 -5.92058539e-01 2.39416678e-02 3.25360119e-01
-5.97315505e-02 3.30501735e-01 1.29846430e+00 -7.98171461e-01
8.40367615e-01 -1.90073848e+00 -4.69521314e-01 -4.19854969e-01
1.37315422e-01 7.62335300e-01 -1.90118030e-01 -5.00709042e-02
2.39105504e-02 -4.73058492e-01 -1.65171102e-01 3.91526073e-02
-3.70427281e-01 1.05490111e-01 -2.48656303e-01 7.70004690e-01
-1.40045043e-02 7.74249911e-01 -1.19142199e+00 -4.45486754e-01
7.71655738e-01 9.05410767e-01 -4.04735953e-01 4.19868827e-01
1.86263304e-02 4.43189234e-01 -3.83684300e-02 1.00789189e+00
1.16322041e+00 1.39176026e-01 -2.36221924e-01 -6.47291780e-01
-1.76888824e-01 -5.61506927e-01 -1.12402904e+00 1.66989279e+00
-5.63214362e-01 9.17979538e-01 2.90527821e-01 -4.26331103e-01
1.00745308e+00 1.38467371e-01 4.09380257e-01 -9.04635370e-01
4.32295561e-01 2.95715667e-02 -5.54017603e-01 -1.30943716e+00
5.15954375e-01 -3.48836690e-01 6.05852425e-01 1.52035460e-01
-2.04328477e-01 -8.23363960e-02 1.32034853e-01 -1.76438212e-01
9.18254197e-01 6.03224754e-01 -1.08860224e-01 2.80265361e-01
6.82476699e-01 -4.36034441e-01 7.59150624e-01 3.43909740e-01
-2.81255573e-01 7.17789173e-01 -2.04987839e-01 -6.23289168e-01
-1.07604790e+00 -1.00752413e+00 -7.32489973e-02 7.20194340e-01
8.05641532e-01 1.31184697e-01 -5.33436477e-01 -3.98647189e-01
-4.96776968e-01 5.02688766e-01 -2.77511030e-01 -3.25273603e-01
-4.83896464e-01 -1.05667961e+00 3.00168902e-01 2.03534335e-01
1.30430222e+00 -7.15126276e-01 -8.21488857e-01 -2.03912050e-01
-1.29519925e-01 -1.54434240e+00 -3.89393061e-01 -1.45139605e-01
-5.24018407e-01 -1.51728559e+00 -7.62644649e-01 -6.97549403e-01
5.88649988e-01 1.12346530e+00 9.57771540e-01 7.54967630e-02
-5.09428084e-01 3.71467412e-01 -3.67661893e-01 -6.00540228e-02
-2.39689842e-01 -8.20437491e-01 1.16701916e-01 5.43914974e-01
6.10409260e-01 -2.91241527e-01 -9.79263484e-01 3.11492532e-01
-1.30420172e+00 4.70380560e-02 7.44034231e-01 8.63764286e-01
2.42580190e-01 3.18607628e-01 -2.06108361e-01 -4.79413837e-01
3.04472864e-01 -3.27283703e-02 -8.25904787e-01 4.43776786e-01
-5.17385900e-01 -3.65772724e-01 9.40240622e-01 -2.86670059e-01
-1.65443015e+00 1.93872433e-02 4.82567549e-02 -8.58610809e-01
-3.31567705e-01 -3.69758427e-01 -4.70781267e-01 -5.73435128e-01
7.03766644e-01 4.85070676e-01 -2.49299496e-01 -1.27168551e-01
1.89259693e-01 9.47486639e-01 8.97853255e-01 -1.65207744e-01
9.91716027e-01 9.20436084e-01 7.23401308e-02 -8.31572056e-01
-8.19593608e-01 -4.02644545e-01 -5.32667577e-01 -5.78172684e-01
1.01337814e+00 -1.03659010e+00 -8.71605337e-01 7.43846297e-01
-1.18888080e+00 -2.87706763e-01 -8.21146183e-03 6.91400349e-01
-3.60246956e-01 8.12121153e-01 -8.85992527e-01 -6.71276271e-01
-2.59314358e-01 -1.10534251e+00 1.20353651e+00 6.35809600e-01
8.63302290e-01 -8.17583382e-01 1.47797763e-02 4.48678106e-01
4.50442970e-01 2.65830368e-01 2.05111519e-01 7.76841283e-01
-1.01057434e+00 -1.35477399e-02 -6.41164243e-01 7.84937441e-01
2.59568751e-01 1.82590425e-01 -1.19039869e+00 -4.64983046e-01
2.37535432e-01 -1.99189484e-01 9.01506424e-01 5.38012803e-01
1.05481601e+00 -2.09705785e-01 -8.61954987e-02 1.40446818e+00
1.83235180e+00 3.20851654e-01 1.06476617e+00 2.72609890e-01
7.58460820e-01 2.95362145e-01 7.03768849e-01 3.51678133e-01
-1.19848941e-02 6.16756678e-01 8.17710638e-01 -4.79743093e-01
-3.69903564e-01 1.35633573e-01 7.99891829e-01 3.42819870e-01
-4.75469053e-01 -9.58245397e-02 -4.44139987e-01 1.75264522e-01
-1.38606608e+00 -1.41849601e+00 -4.24093008e-01 2.12985444e+00
8.42881978e-01 -2.00166013e-02 -2.94435710e-01 9.87639725e-02
7.83626318e-01 2.47474670e-01 -4.19101596e-01 1.01208851e-01
-4.89236772e-01 -1.61357641e-01 8.14128637e-01 3.70124102e-01
-9.20153201e-01 6.56686723e-01 5.99498796e+00 5.23160994e-01
-1.26759684e+00 2.12671131e-01 5.76082885e-01 -1.82535633e-01
-8.13509151e-02 -1.01138979e-01 -7.60491550e-01 6.99787021e-01
4.22276020e-01 2.36360684e-01 6.27465010e-01 6.03314042e-01
4.11748081e-01 -1.59489736e-01 -6.89966023e-01 1.65442741e+00
6.10063016e-01 -1.02152491e+00 -1.19079471e-01 -1.11585848e-01
1.19298005e+00 -1.83390886e-01 3.08921695e-01 -1.78694665e-01
-6.02477901e-02 -5.65309346e-01 4.37849075e-01 8.19621146e-01
1.11956310e+00 -4.72314686e-01 7.48789012e-01 7.89795220e-02
-1.05020225e+00 -3.04570317e-01 -8.62622142e-01 -9.96198319e-03
2.94751018e-01 6.09080613e-01 -8.28327984e-02 5.07813752e-01
9.83155012e-01 1.09820747e+00 -6.99022233e-01 1.06323957e+00
-3.39243174e-01 2.75596589e-01 -3.31690237e-02 2.28522405e-01
1.21598855e-01 -6.33390486e-01 3.99284363e-01 1.17777777e+00
2.98173547e-01 3.99216652e-01 1.02889156e-02 8.55941772e-01
1.64688259e-01 -4.83612090e-01 -9.66518521e-01 4.43134338e-01
-4.48525362e-02 1.51643384e+00 -3.64664048e-01 -3.75906587e-01
-8.28002214e-01 1.15976560e+00 -7.26648688e-01 6.43289566e-01
-1.17804992e+00 -6.66190147e-01 5.29808760e-01 3.76926363e-02
-2.52053231e-01 -1.95203170e-01 2.37891041e-02 -1.49767125e+00
1.43671721e-01 -7.23070681e-01 1.07424915e-01 -1.75247371e+00
-1.24064052e+00 5.99665582e-01 -1.03578866e-01 -1.68082845e+00
1.58708468e-01 -9.80583608e-01 -5.28374493e-01 6.34253144e-01
-2.25352144e+00 -1.17024100e+00 -1.18626416e+00 1.09909391e+00
4.84075874e-01 -2.28207380e-01 6.35082722e-02 6.01480424e-01
-5.49998343e-01 2.07889184e-01 2.74186701e-01 1.51191458e-01
9.43171024e-01 -6.33383453e-01 -6.26367256e-02 1.42368948e+00
-3.06950390e-01 4.12538826e-01 5.84117651e-01 -4.19747710e-01
-1.88913453e+00 -1.23273301e+00 2.10149527e-01 -3.19137245e-01
1.13798581e-01 1.04710020e-01 -6.27241135e-01 5.91576397e-01
6.10456824e-01 5.59618652e-01 2.33414456e-01 -7.54899144e-01
-1.09554298e-01 -4.87834156e-01 -1.23321283e+00 5.24815261e-01
1.18103957e+00 -4.66410309e-01 -5.28164744e-01 3.46768051e-01
6.30821705e-01 -3.12811702e-01 -4.83249575e-01 4.41252053e-01
5.22523761e-01 -1.63533664e+00 1.24314046e+00 -1.46993071e-01
5.69767058e-01 -7.66024888e-01 -1.36830926e-01 -9.84520197e-01
-1.98208869e-01 -8.09958160e-01 3.41904424e-02 1.02435601e+00
-3.70884061e-01 -5.77436388e-01 6.80059433e-01 5.04169941e-01
-2.94651806e-01 -2.76473433e-01 -3.66710037e-01 -6.38454556e-01
-7.27017999e-01 -3.84393543e-01 4.78291422e-01 9.95156288e-01
-4.82955307e-01 8.36488605e-02 -7.98491359e-01 3.73944342e-01
1.15539992e+00 6.27488866e-02 8.59554231e-01 -7.87656486e-01
-3.38978529e-01 -2.57274918e-02 -3.80525172e-01 -8.24523926e-01
-1.09530054e-01 -4.41523015e-01 6.36153743e-02 -1.40327311e+00
3.73002589e-01 7.91189447e-02 -1.26078382e-01 2.53185719e-01
-3.46022904e-01 9.76463258e-01 6.74628243e-02 5.72701283e-05
-7.71426857e-01 5.15932560e-01 1.37569177e+00 -2.57720649e-01
-2.39410266e-01 -3.05603296e-01 -3.51167709e-01 7.47003913e-01
4.93013680e-01 1.00099586e-01 -4.17364448e-01 -7.45867133e-01
3.76597166e-01 -8.60483106e-03 7.54274905e-01 -1.20896029e+00
2.50076294e-01 -5.34325719e-01 7.36620665e-01 -2.61378735e-01
5.06227851e-01 -1.26015735e+00 3.63627076e-01 4.26815957e-01
-6.73866197e-02 -2.72767097e-02 -1.27779339e-02 6.23338342e-01
-1.22168407e-01 -1.61361381e-01 1.12168348e+00 -3.62113535e-01
-1.12147200e+00 4.75155234e-01 -1.77599583e-02 -1.39136255e-01
1.30472684e+00 -5.88740528e-01 -9.15286183e-01 -2.04712659e-01
-5.28079793e-02 -3.12000453e-01 8.55030119e-01 2.48546243e-01
1.09127390e+00 -1.25678372e+00 -6.11465693e-01 7.36613393e-01
4.32001576e-02 -3.11117768e-01 4.97328252e-01 7.81996965e-01
-1.01747608e+00 9.61952433e-02 -7.95120418e-01 -5.42495787e-01
-1.33013463e+00 8.92577767e-01 5.10956407e-01 4.58496362e-01
-7.55641699e-01 5.18040121e-01 1.69038743e-01 1.34286940e-01
1.15337484e-01 -3.28849077e-01 3.57978903e-02 -5.74907064e-01
1.02804947e+00 6.16640568e-01 1.39473472e-02 -6.24583423e-01
-7.68441260e-02 1.04225016e+00 5.15426815e-01 2.50010222e-01
1.39167798e+00 -4.94939953e-01 -3.02939583e-02 -3.77231017e-02
1.21235025e+00 -1.60849139e-01 -1.73236239e+00 -6.31975830e-02
-8.18229258e-01 -1.36250079e+00 4.26815480e-01 -4.44144249e-01
-1.53082168e+00 1.04484212e+00 9.65346396e-01 1.17130741e-01
1.76429534e+00 -5.30033708e-01 1.13358748e+00 4.87774283e-01
6.50241077e-01 -1.09619153e+00 2.94097185e-01 6.75469190e-02
3.50464910e-01 -1.57772517e+00 5.72202988e-02 -3.45200866e-01
-5.06946981e-01 1.46166563e+00 6.72143638e-01 -1.09049613e-02
3.71337175e-01 4.85737801e-01 2.01630160e-01 5.65414093e-02
-7.65843213e-01 -2.96801597e-01 -1.45959616e-01 1.13514912e+00
1.95703223e-01 -6.84015930e-01 -1.53930355e-02 7.48354867e-02
4.76055712e-01 4.48382586e-01 8.53251636e-01 7.18597770e-01
-5.31649888e-01 -6.15994930e-01 -8.64373207e-01 -2.45987028e-02
-5.35714626e-01 1.22260908e-03 5.78711703e-02 5.40180683e-01
6.12465501e-01 1.14161468e+00 -2.40674376e-01 -5.65309048e-01
1.98667198e-01 -3.77790242e-01 5.94322741e-01 -4.42351475e-02
-3.15524191e-01 5.52198254e-02 -4.73774076e-01 -7.92591989e-01
-9.40590501e-01 -4.09735590e-01 -7.32202291e-01 -4.84448582e-01
-2.67157257e-01 -3.90379012e-01 6.14615798e-01 4.71807361e-01
1.09709334e-02 3.50916624e-01 8.82976234e-01 -1.15193284e+00
-1.90151975e-01 -6.62742913e-01 -7.39353955e-01 7.76632011e-01
5.92254579e-01 -5.90170860e-01 -4.97625768e-01 6.96363091e-01] | [10.698983192443848, -2.367225170135498] |
2a80d19b-a69c-4413-8a7e-a1ff689a92a5 | automatic-punctuation-restoration-with-bert | 2101.07343 | null | https://arxiv.org/abs/2101.07343v1 | https://arxiv.org/pdf/2101.07343v1.pdf | Automatic punctuation restoration with BERT models | We present an approach for automatic punctuation restoration with BERT models for English and Hungarian. For English, we conduct our experiments on Ted Talks, a commonly used benchmark for punctuation restoration, while for Hungarian we evaluate our models on the Szeged Treebank dataset. Our best models achieve a macro-averaged $F_1$-score of 79.8 in English and 82.2 in Hungarian. Our code is publicly available. | ['Judit Ács', 'Bence Bial', 'Attila Nagy'] | 2021-01-18 | null | null | null | null | ['punctuation-restoration'] | ['natural-language-processing'] | [-3.60715181e-01 2.95842320e-01 1.27315745e-01 -4.15345967e-01
-1.49757528e+00 -4.48873132e-01 2.79821996e-02 2.24926829e-01
-8.10055494e-01 9.92311239e-01 7.12089181e-01 -4.25990373e-01
3.65363330e-01 -4.61663157e-01 -6.60443962e-01 -3.72387052e-01
-6.68387637e-02 2.72768438e-01 1.80469394e-01 -6.66158557e-01
-5.44311479e-02 -3.88378091e-02 -8.68467689e-01 4.18598622e-01
9.70586121e-01 3.65777582e-01 2.69438803e-01 5.81350923e-01
-2.26813242e-01 5.97505569e-01 -1.03758276e+00 -9.09501910e-01
-8.80336687e-02 -3.92576605e-01 -1.45977998e+00 -1.47700727e-01
8.02513659e-02 2.94801235e-01 -3.96840423e-01 1.16022992e+00
7.24060476e-01 -4.84651234e-03 2.73610026e-01 -4.40943182e-01
-3.79252553e-01 1.79372180e+00 6.63135573e-02 3.49974602e-01
5.31288981e-01 -2.14175925e-01 1.22776175e+00 -8.85083497e-01
8.94231379e-01 1.25529015e+00 6.62983239e-01 6.21555686e-01
-1.09611845e+00 -3.73723507e-01 6.89465494e-04 3.20029289e-01
-1.58685315e+00 -7.88354635e-01 5.38166642e-01 2.26429328e-01
1.28219354e+00 1.98721632e-01 2.86424994e-01 8.69844973e-01
1.70087263e-01 1.24134564e+00 9.70462024e-01 -9.15287971e-01
3.97509597e-02 2.29549920e-03 1.15675248e-01 4.92334723e-01
-1.28637239e-01 -3.23307216e-01 -8.18794966e-01 1.66721076e-01
3.18811119e-01 -8.57647300e-01 -4.87077802e-01 5.90655506e-01
-1.16061401e+00 5.76281130e-01 1.20956197e-01 5.09342492e-01
-3.92164469e-01 5.69146164e-02 6.27448618e-01 5.57086647e-01
3.80585879e-01 4.26488072e-01 -9.52743888e-01 -4.41098452e-01
-9.26496208e-01 4.97054309e-01 9.10462439e-01 1.17028201e+00
7.89534211e-01 2.94254541e-01 -1.70093551e-01 1.21887863e+00
-6.80163596e-03 4.46557730e-01 7.42214382e-01 -6.11160755e-01
6.88929737e-01 -1.13912582e-01 2.77925402e-01 -1.16084591e-01
-3.25456113e-01 -2.95899928e-01 -4.23588157e-01 -4.64397967e-01
6.46361470e-01 -3.83808434e-01 -1.01841044e+00 1.69819784e+00
2.81647705e-02 -2.94489652e-01 4.00404483e-01 5.47065020e-01
9.37731266e-01 6.20386779e-01 3.04902792e-01 -2.96639651e-01
1.54863620e+00 -9.67528462e-01 -1.28363872e+00 -3.05720925e-01
8.78538370e-01 -1.41542435e+00 1.30140245e+00 5.36981940e-01
-1.45992291e+00 -2.84117043e-01 -7.77757227e-01 -2.34928995e-01
-3.23565811e-01 -1.17758978e-02 1.84185416e-01 8.77144575e-01
-1.07626307e+00 7.46494174e-01 -7.70124555e-01 -2.65709579e-01
-1.41887397e-01 -2.78384192e-03 -2.32610852e-01 1.34094149e-01
-1.78973651e+00 9.16435719e-01 5.00843883e-01 -1.15634121e-01
-4.51879025e-01 -4.16064382e-01 -9.58878815e-01 1.51876444e-02
-3.65143381e-02 4.89332974e-02 1.96973145e+00 -6.05208755e-01
-1.70326018e+00 1.09140921e+00 -3.41842294e-01 -9.17566717e-01
4.73848701e-01 -5.06243169e-01 -7.82357097e-01 -9.76710767e-02
1.18465446e-01 2.26677373e-01 2.20732048e-01 -8.37999582e-01
-7.92257011e-01 -4.34481315e-02 -5.15280008e-01 2.40959048e-01
-1.27147837e-02 4.85112846e-01 -4.29160327e-01 -1.07960367e+00
4.14956361e-01 -5.29963315e-01 -2.49211997e-01 -1.16339672e+00
-5.39252877e-01 -3.33211392e-01 4.14579421e-01 -1.26722229e+00
1.74778187e+00 -1.95181406e+00 -2.82972723e-01 5.76896034e-02
-5.00739872e-01 2.95454592e-01 -3.23869437e-02 4.31452870e-01
-8.35787505e-02 4.40169722e-01 -2.58997917e-01 -5.43370724e-01
1.88034892e-01 2.02354059e-01 -2.10917830e-01 2.43783355e-01
2.08646506e-01 8.88586402e-01 -8.28020632e-01 -4.28090781e-01
6.88624382e-02 1.94831252e-01 -5.72277904e-01 1.08746111e-01
-7.52808340e-03 -4.04034182e-02 -1.38516843e-01 8.73209655e-01
4.30371612e-01 5.06239057e-01 2.82093406e-01 3.31476569e-01
-3.44217122e-01 1.21899426e+00 -1.17191422e+00 1.90486765e+00
-6.20594680e-01 1.92925975e-01 2.17972949e-01 -6.18525863e-01
9.61208880e-01 6.13102913e-01 -6.79990202e-02 -6.54254794e-01
7.58106261e-02 4.45529848e-01 -1.92894593e-01 -1.32881075e-01
1.20768642e+00 -3.39157790e-01 -8.12152326e-01 9.14403424e-02
3.73723954e-01 -7.10542738e-01 2.78739423e-01 1.75032109e-01
1.23738623e+00 -1.59814563e-02 4.14932489e-01 -6.03011608e-01
2.45227069e-01 1.28049254e-01 8.08005393e-01 7.57311344e-01
-4.49409366e-01 8.95311594e-01 2.47276455e-01 -2.00530559e-01
-9.61123466e-01 -1.24550617e+00 -2.54336894e-01 1.10605955e+00
-3.16439718e-01 -9.74316418e-01 -1.38748240e+00 -6.42478287e-01
-3.74870956e-01 1.09144366e+00 -1.87879249e-01 2.24036664e-01
-1.03115141e+00 -6.58207715e-01 1.03557968e+00 1.12000801e-01
5.37995398e-01 -1.54377472e+00 3.00450586e-02 8.17840517e-01
-8.91951382e-01 -1.16553378e+00 -7.53718734e-01 6.43499255e-01
-6.54547453e-01 -6.50890589e-01 -5.14714718e-01 -1.25903916e+00
-2.62260616e-01 -1.47597671e-01 1.33531713e+00 -1.83392927e-01
-7.80919716e-02 3.24226432e-02 -6.88086748e-01 -5.70701361e-01
-7.65640020e-01 4.19431657e-01 5.23818843e-02 -5.88705003e-01
3.93888056e-01 -3.44615906e-01 -1.82420686e-02 2.17026830e-01
-7.10318685e-01 -6.03516877e-01 2.59899229e-01 8.44184816e-01
7.36872256e-01 -1.48388833e-01 8.07018578e-01 -1.06499052e+00
3.04449886e-01 -1.80042773e-01 -5.51449299e-01 5.23175783e-02
-2.00901091e-01 7.48926029e-02 7.50784278e-01 1.21936820e-01
-1.34752798e+00 1.64564118e-01 -8.00311327e-01 2.43684098e-01
-1.65494695e-01 6.25665903e-01 -5.84746361e-01 3.50976020e-01
7.95542479e-01 2.41728388e-02 -7.87768841e-01 -1.02472484e+00
4.21155035e-01 8.79235685e-01 1.00451708e+00 -7.01323748e-01
3.11593056e-01 -8.05230290e-02 -8.85327876e-01 -1.20480192e+00
-8.42002213e-01 -6.51468515e-01 -5.47579467e-01 1.44644156e-01
6.50710166e-01 -1.01109719e+00 -2.10899457e-01 5.97021639e-01
-1.27989841e+00 -4.80233967e-01 -4.67755973e-01 2.73672014e-01
-7.12680101e-01 6.22314394e-01 -1.04322863e+00 -5.47036946e-01
-6.68575585e-01 -7.38810360e-01 7.07019091e-01 1.60361603e-01
-5.58119357e-01 -9.75102961e-01 1.10320881e-01 1.77461445e-01
2.08918720e-01 -1.90594822e-01 6.89321697e-01 -6.47136748e-01
6.02353923e-02 8.78995657e-02 3.21072400e-01 5.24208128e-01
2.54990786e-01 -3.14732432e-01 -9.61766541e-01 6.96610138e-02
-1.05268024e-01 -3.05576295e-01 9.10797179e-01 4.80856955e-01
8.22746515e-01 -3.42215061e-01 1.30191863e-01 4.25577849e-01
1.08620286e+00 -6.74722865e-02 1.00048685e+00 8.00294220e-01
1.35140732e-01 4.23951864e-01 8.23785365e-01 6.89332545e-01
5.59254229e-01 4.02203560e-01 -1.07280493e-01 3.68858315e-02
-4.24117178e-01 -5.56772232e-01 8.38143110e-01 1.42343915e+00
3.86343926e-01 -2.66685486e-01 -1.18175209e+00 1.25039339e+00
-1.70021975e+00 -8.02760601e-01 -3.10496300e-01 2.10495281e+00
1.57892859e+00 2.73464620e-01 -1.95556488e-02 3.42611134e-01
9.93761957e-01 3.16213109e-02 3.99547696e-01 -7.73274720e-01
-6.31927550e-01 1.00717962e+00 4.36182708e-01 7.85851896e-01
-1.42039573e+00 2.04957867e+00 7.35729647e+00 1.02622974e+00
-6.17073894e-01 4.59450409e-02 4.97380823e-01 1.55381203e-01
-1.81156576e-01 8.76219124e-02 -1.04360664e+00 2.47768402e-01
1.61685169e+00 -5.03378808e-01 2.63817877e-01 7.21505523e-01
3.79986823e-01 -1.41642407e-01 -6.74276948e-01 8.40134799e-01
-3.80185127e-01 -1.24193776e+00 -2.19807371e-01 -5.20746112e-01
7.50763118e-01 3.05716455e-01 -2.38198265e-01 4.71538663e-01
9.20479655e-01 -9.23604310e-01 1.18395007e+00 2.07330585e-02
6.08507037e-01 -1.11353862e+00 8.74916017e-01 1.45886345e-02
-1.01091695e+00 7.24857211e-01 -3.32652748e-01 1.89138353e-01
5.01344264e-01 7.12383330e-01 -1.10755324e+00 6.91299081e-01
1.05089056e+00 3.76756817e-01 -3.83584380e-01 1.15915537e+00
-8.54948580e-01 1.30669248e+00 -4.89514679e-01 6.94248378e-02
2.88910598e-01 1.60049945e-01 6.39643908e-01 2.11711073e+00
1.76659226e-01 2.02198774e-01 1.68561429e-01 3.54363590e-01
-1.78148434e-01 5.90644360e-01 1.52084762e-02 1.61311686e-01
6.32219434e-01 1.00584900e+00 -5.19902349e-01 -4.61456746e-01
9.48972814e-03 1.04153490e+00 4.57322598e-01 2.29160592e-01
-6.44127309e-01 -1.04894006e+00 8.21655750e-01 -2.10129872e-01
4.88450408e-01 -1.84997395e-01 -3.39487433e-01 -1.23955107e+00
-1.29085751e-02 -9.97427106e-01 6.93295538e-01 -4.30687726e-01
-1.09320796e+00 7.85271049e-01 -3.44726145e-01 -7.00141966e-01
-2.85179436e-01 -3.67959797e-01 -5.11484802e-01 9.47107315e-01
-1.75637376e+00 -7.83540547e-01 4.40007150e-01 5.30963182e-01
6.27879381e-01 2.33794358e-02 1.01979363e+00 2.44885728e-01
-5.14323354e-01 9.49488997e-01 2.90198714e-01 5.49157262e-01
1.02856481e+00 -1.54468417e+00 1.04495788e+00 1.03994882e+00
3.26684654e-01 3.60425800e-01 1.08727670e+00 -4.86415505e-01
-8.93189609e-01 -1.15271795e+00 1.84625733e+00 -2.34856561e-01
8.53034377e-01 -3.39933038e-01 -1.07769501e+00 8.47681344e-01
4.12038565e-01 -3.55919957e-01 4.58502293e-01 2.77590960e-01
-1.11134991e-01 2.48307899e-01 -1.16055632e+00 6.21657908e-01
1.00280678e+00 -4.45554316e-01 -1.00025380e+00 2.66643435e-01
8.47483814e-01 -6.50937915e-01 -1.04378521e+00 -4.77874316e-02
7.14292228e-02 -7.41697729e-01 4.36126709e-01 -4.78464395e-01
9.63736605e-03 -1.10186011e-01 -3.40207249e-01 -1.67291784e+00
-1.72059044e-01 -1.41524756e+00 6.66072786e-01 1.50132108e+00
8.03908825e-01 -3.07080626e-01 6.72082603e-01 1.77024931e-01
-7.13007450e-01 2.00069547e-01 -1.31511080e+00 -9.24301386e-01
5.06396234e-01 -7.14906991e-01 3.65267813e-01 8.45996499e-01
6.77871108e-01 1.99358001e-01 -1.95461854e-01 1.83866341e-02
3.81603599e-01 -3.21388990e-01 3.36547941e-01 -6.32402182e-01
-1.53531864e-01 -2.63527304e-01 -5.16478531e-02 -7.11158514e-01
4.37385768e-01 -8.90617013e-01 5.96310675e-01 -1.25600696e+00
-2.50265390e-01 -3.42985451e-01 -2.65190244e-01 8.51058900e-01
-2.23456740e-01 2.33571127e-01 1.17998704e-01 -3.36020112e-01
-5.49093008e-01 6.68533742e-01 6.79609418e-01 7.68874539e-03
-5.27582288e-01 -7.02731311e-02 -5.23012042e-01 6.49940491e-01
1.16131616e+00 -6.10151947e-01 7.14572608e-01 -4.64097738e-01
-2.82292753e-01 -2.78919637e-02 -2.78664321e-01 -9.07683492e-01
-1.53279647e-01 -2.81183198e-02 -1.53188407e-02 -5.90285897e-01
3.91783491e-02 3.71871777e-02 -2.14183196e-01 3.87353837e-01
-3.89705822e-02 1.62728101e-01 6.67416155e-01 -6.58982918e-02
-6.26474977e-01 -3.43057990e-01 8.88197720e-01 -2.34349072e-01
-7.49488592e-01 -2.84166068e-01 -8.34668636e-01 5.51280141e-01
2.16026217e-01 1.57778457e-01 -6.70958310e-02 -4.49204594e-01
-1.01509237e+00 1.38788730e-01 3.36117208e-01 3.60410094e-01
3.92059326e-01 -1.10313189e+00 -1.09187853e+00 1.03722373e-02
-7.65540376e-02 -2.06951350e-01 -7.94463754e-02 3.19864303e-01
-6.87560499e-01 3.78623784e-01 1.80176198e-01 -5.81226498e-03
-1.22066569e+00 5.87654635e-02 3.15511733e-01 -4.16710794e-01
-4.72229987e-01 9.05402541e-01 -2.45363936e-01 -5.46956003e-01
3.31339955e-01 -4.23488081e-01 6.10378012e-02 -8.92169625e-02
6.10535800e-01 3.43324900e-01 8.50802243e-01 -7.53857195e-01
-6.33707106e-01 -1.95078123e-02 -3.34504277e-01 -6.08418941e-01
1.19240212e+00 -2.28835449e-01 -1.35221988e-01 3.15703154e-01
8.04685295e-01 4.58791047e-01 -6.32695436e-01 -4.11653340e-01
7.28762090e-01 8.66501704e-02 1.20522857e-01 -1.03776956e+00
-4.20600265e-01 5.08605182e-01 8.20532069e-02 3.02900421e-03
9.73469257e-01 6.94015846e-02 1.00721741e+00 6.19409680e-01
3.91344875e-01 -1.71236622e+00 -4.22344327e-01 1.48883057e+00
7.72435188e-01 -8.82287979e-01 -3.58539611e-01 -5.22909820e-01
-7.72887111e-01 8.70927572e-01 1.49071977e-01 -1.01555713e-01
5.62042236e-01 4.48028237e-01 4.30594414e-01 3.19306999e-01
-6.44699812e-01 -3.56617928e-01 -2.09424794e-01 5.27498364e-01
1.01611030e+00 3.94444197e-01 -7.01802492e-01 1.19479251e+00
-1.08399320e+00 -4.82898802e-01 9.82592523e-01 9.26896453e-01
-6.93206966e-01 -1.55889845e+00 -6.76676452e-01 -2.59470135e-01
-1.04874182e+00 -6.67866230e-01 -6.02433085e-01 8.91086698e-01
-4.03295159e-01 1.18766665e+00 -1.10394500e-01 -1.32058099e-01
6.48339212e-01 5.32644987e-01 3.77825677e-01 -8.65992486e-01
-1.19622707e+00 6.51476920e-01 6.99109554e-01 -3.56884331e-01
-8.37316066e-02 -1.04742432e+00 -1.57355607e+00 -4.60635811e-01
-2.80421704e-01 5.91048837e-01 5.68007648e-01 7.27246463e-01
-6.27741367e-02 2.89769441e-01 4.97591585e-01 -3.77926737e-01
-4.35568780e-01 -1.24892485e+00 -9.40000653e-01 2.96023041e-01
9.58338901e-02 3.20241749e-02 -3.89048219e-01 2.78838366e-01] | [14.25218391418457, 7.130068302154541] |
1c859a47-427f-455d-9342-43df9e5dab3c | deep-joint-face-hallucination-and-recognition | 1611.08091 | null | http://arxiv.org/abs/1611.08091v1 | http://arxiv.org/pdf/1611.08091v1.pdf | Deep Joint Face Hallucination and Recognition | Deep models have achieved impressive performance for face hallucination
tasks. However, we observe that directly feeding the hallucinated facial images
into recog- nition models can even degrade the recognition performance despite
the much better visualization quality. In this paper, we address this problem
by jointly learning a deep model for two tasks, i.e. face hallucination and
recognition. In particular, we design an end-to-end deep convolution network
with hallucination sub-network cascaded by recognition sub-network. The
recognition sub- network are responsible for producing discriminative feature
representations using the hallucinated images as inputs generated by
hallucination sub-network. During training, we feed LR facial images into the
network and optimize the parameters by minimizing two loss items, i.e. 1) face
hallucination loss measured by the pixel wise difference between the ground
truth HR images and network-generated images; and 2) verification loss which is
measured by the classification error and intra-class distance. We extensively
evaluate our method on LFW and YTF datasets. The experimental results show that
our method can achieve recognition accuracy 97.95% on 4x down-sampled LFW
testing set, outperforming the accuracy 96.35% of conventional face recognition
model. And on the more challenging YTF dataset, we achieve recognition accuracy
90.65%, a margin over the recognition accuracy 89.45% obtained by conventional
face recognition model on the 4x down-sampled version. | ['Shengyong Ding', 'Wei Xu', 'Junyu Wu', 'Hongyang Chao'] | 2016-11-24 | null | null | null | null | ['face-hallucination'] | ['computer-vision'] | [ 3.53122413e-01 4.78744447e-01 6.95549995e-02 -5.40858924e-01
-6.32286072e-01 9.60212275e-02 4.40995932e-01 -9.50735211e-01
-1.40344277e-01 5.68994045e-01 1.25496119e-01 1.70214012e-01
1.35551870e-01 -7.53196180e-01 -8.39146435e-01 -7.02570856e-01
2.08686963e-01 7.60456026e-02 -6.74945295e-01 2.85265684e-01
-1.61399618e-01 7.74397016e-01 -1.75998247e+00 3.24398369e-01
6.46712780e-01 1.54636037e+00 -8.32249224e-02 2.74455816e-01
3.34407419e-01 7.88053393e-01 -7.19590008e-01 -3.86576414e-01
4.07958448e-01 -5.40811419e-01 -3.73133302e-01 5.47708094e-01
6.98713601e-01 -5.72358549e-01 -7.80683160e-01 9.31733131e-01
6.80853367e-01 1.67003781e-01 5.76453447e-01 -1.38089287e+00
-1.13652742e+00 1.88892901e-01 -5.75018108e-01 -3.91558290e-01
2.09970251e-01 3.75819862e-01 3.79068255e-01 -1.61266470e+00
4.26985711e-01 1.46376252e+00 6.18818998e-01 9.88145113e-01
-1.41513848e+00 -1.03252649e+00 -5.73723316e-01 6.01986889e-03
-1.74810493e+00 -9.43960071e-01 4.75429147e-01 -2.52947897e-01
8.05667400e-01 4.50535268e-02 1.53312415e-01 1.18925071e+00
2.83502601e-02 4.88693863e-01 1.02881229e+00 -1.81912839e-01
-1.15802847e-01 3.41120839e-01 -3.32720429e-01 8.90108287e-01
-8.22970197e-02 3.65849644e-01 -6.77314818e-01 5.47958612e-02
1.00280821e+00 1.64185703e-01 -6.82921529e-01 -4.06070671e-04
-7.47555137e-01 7.25601971e-01 5.56447029e-01 -1.58544064e-01
-4.27632153e-01 5.35904020e-02 -1.11576594e-01 4.86016035e-01
3.92459780e-01 3.07526976e-01 2.95854509e-02 3.72479320e-01
-1.05874038e+00 -1.68668613e-01 5.09990871e-01 7.16571093e-01
6.77108705e-01 5.31619012e-01 -4.46736902e-01 1.46965110e+00
4.91869956e-01 6.84276402e-01 4.49484289e-01 -9.73205149e-01
1.85048804e-01 3.72887492e-01 -1.74126416e-01 -7.82208025e-01
1.18700322e-02 -6.56195045e-01 -1.15102756e+00 3.49286377e-01
6.79840371e-02 -1.43629178e-01 -1.28042603e+00 1.88125682e+00
-8.06603804e-02 3.80804181e-01 3.22724164e-01 1.04305136e+00
1.02660298e+00 8.26891541e-01 -8.08520094e-02 -2.63698041e-01
1.08553660e+00 -1.12705088e+00 -5.96261263e-01 7.82642812e-02
-2.18866300e-03 -7.40368426e-01 1.08530617e+00 2.17941493e-01
-1.29474795e+00 -9.23489094e-01 -1.21515059e+00 8.92506987e-02
-6.11408800e-02 5.86543143e-01 1.63913786e-01 5.63436389e-01
-1.11630619e+00 5.76965630e-01 -2.75185823e-01 -3.45806837e-01
9.75806773e-01 5.24665236e-01 -6.60133421e-01 -3.98096234e-01
-8.74025345e-01 7.93382943e-01 -1.43413141e-01 3.38460088e-01
-1.58090091e+00 -8.17091465e-01 -9.02198613e-01 4.61853594e-01
1.58097282e-01 -3.91696602e-01 1.04667711e+00 -9.51648593e-01
-1.51204801e+00 8.51675272e-01 -2.27115571e-01 -2.96885282e-01
4.17777002e-01 2.68998761e-02 -7.96199501e-01 1.76607873e-02
-2.77842488e-02 7.56831408e-01 9.82046247e-01 -1.39259160e+00
4.93263677e-02 -6.14913046e-01 -5.82755685e-01 3.92661765e-02
-5.55461109e-01 -1.88272208e-01 -4.04989958e-01 -4.63002563e-01
-8.08134452e-02 -7.01196074e-01 3.22437555e-01 5.56593597e-01
-5.51275373e-01 5.57119176e-02 1.26421094e+00 -7.79516339e-01
6.88391209e-01 -2.51182365e+00 -1.81040838e-01 2.02531531e-01
2.37676457e-01 6.12231553e-01 -6.85946286e-01 1.00484371e-01
-3.48435551e-01 -8.63471404e-02 -1.89841181e-01 -7.75276005e-01
-1.75169662e-01 1.37914149e-02 -6.40667140e-01 5.34462094e-01
5.36522925e-01 9.77578402e-01 -3.81096423e-01 1.05247281e-01
-1.30991302e-02 9.98422503e-01 -2.33465791e-01 6.21937990e-01
1.52692825e-01 1.36562616e-01 1.33115128e-01 9.04022753e-01
1.05434930e+00 -1.88264936e-01 5.60591519e-02 -2.32363358e-01
4.12959486e-01 -2.02617943e-01 -7.24058270e-01 1.38955724e+00
-7.18574941e-01 7.77387023e-01 -4.81196083e-02 -8.21617067e-01
1.50448298e+00 4.77833599e-01 1.70674957e-02 -8.62267435e-01
1.66048147e-02 2.22285196e-01 -2.32604563e-01 -2.97526658e-01
1.32218227e-01 -4.41473275e-01 4.09194976e-01 3.29102725e-01
4.60306078e-01 3.14175993e-01 -5.43292820e-01 -2.82360613e-01
7.75956392e-01 -1.59706950e-01 -1.65679649e-01 3.42412814e-02
3.84615004e-01 -5.48324823e-01 3.73682439e-01 5.89629233e-01
1.73258677e-01 1.00280631e+00 4.67276126e-01 -2.48659357e-01
-9.02098894e-01 -1.28409445e+00 -3.65487576e-01 4.90981787e-01
-7.64424726e-02 1.33548573e-01 -5.77573359e-01 -5.94594300e-01
1.70603245e-01 5.35751700e-01 -7.85606384e-01 -6.39202714e-01
-2.24384010e-01 -6.28383696e-01 1.00135016e+00 5.67202151e-01
1.04385090e+00 -1.25930834e+00 -1.76895648e-01 -2.92208582e-01
1.14663884e-01 -1.09948599e+00 -6.31038129e-01 -3.03150713e-01
-4.88168627e-01 -7.36653924e-01 -9.97610211e-01 -8.11989009e-01
7.74204135e-01 3.08068752e-01 6.44136727e-01 2.14510635e-01
-5.51213086e-01 -1.10076545e-02 1.31025448e-01 -1.21379439e-02
-1.18327297e-01 -3.80721569e-01 8.61524567e-02 6.22740030e-01
1.28510147e-01 -5.02615035e-01 -5.04537344e-01 5.27410865e-01
-8.09372306e-01 -2.61720233e-02 5.94803214e-01 1.18902147e+00
4.23076004e-01 -2.46764481e-01 6.89926386e-01 -4.68972951e-01
4.03849125e-01 -4.61396635e-01 -3.90096068e-01 3.26960444e-01
-5.93506694e-01 -1.04576960e-01 5.82644939e-01 -4.69831645e-01
-1.13383818e+00 -3.57713476e-02 -5.13273738e-02 -9.28038895e-01
1.11586452e-01 2.37000614e-01 -5.59551954e-01 -2.18735874e-01
5.89609921e-01 5.12406468e-01 5.80889523e-01 -4.48310614e-01
-1.52588366e-02 8.22635174e-01 8.13216090e-01 -1.26373351e-01
7.68818140e-01 3.32178354e-01 2.50446890e-03 -9.52901840e-01
-6.61436021e-01 1.00698121e-01 4.65497561e-02 -2.61620581e-01
6.64911091e-01 -1.13020754e+00 -7.93774188e-01 5.10121822e-01
-1.04264247e+00 -3.44762892e-01 -3.83821338e-01 3.95167917e-01
-4.79042917e-01 -2.86707934e-02 -5.37546575e-01 -9.26467836e-01
-4.17456746e-01 -1.16004848e+00 1.24750221e+00 3.95052314e-01
1.98348358e-01 -4.94994432e-01 -3.52726579e-01 3.91270846e-01
5.84236920e-01 3.66409510e-01 7.67120481e-01 -2.51077771e-01
-4.67456728e-01 -1.78704202e-01 -7.91374385e-01 7.06843257e-01
-1.73064396e-02 -2.92035133e-01 -1.54121959e+00 -3.08622479e-01
-2.40005761e-01 -7.74084806e-01 1.12973392e+00 -6.46587908e-02
1.49297512e+00 -3.87600988e-01 -1.06145486e-01 9.73098576e-01
1.43682730e+00 2.06318811e-01 1.21244907e+00 -4.49300855e-01
5.20198643e-01 5.79119325e-01 2.77795970e-01 3.67317498e-01
-8.39862898e-02 9.14176047e-01 3.96816909e-01 -2.76641428e-01
-6.95148051e-01 -6.54994190e-01 5.99749386e-01 -2.56559765e-03
7.92190284e-02 -2.43622452e-01 -6.42673969e-01 2.61353880e-01
-1.68427277e+00 -1.08304107e+00 3.91071856e-01 2.24611521e+00
5.20693541e-01 -5.12749970e-01 -1.97970599e-01 -1.42746314e-01
5.43244481e-01 8.57639834e-02 -7.87137270e-01 -2.15402275e-01
-3.40865463e-01 5.48707843e-01 9.15393140e-03 6.25891864e-01
-5.84555209e-01 8.06960404e-01 5.43809462e+00 7.81062901e-01
-1.48356915e+00 1.47037625e-01 1.03086448e+00 -2.86016315e-01
-8.16195235e-02 -5.42259932e-01 -6.13723516e-01 2.10283250e-01
8.48277092e-01 -3.15290727e-02 7.18708873e-01 7.52557814e-01
1.71517596e-01 2.82751352e-01 -1.09410119e+00 1.32899308e+00
5.62398612e-01 -1.22678173e+00 3.37747365e-01 4.02792126e-01
6.87183261e-01 -2.48192698e-01 6.05080724e-01 4.65552986e-01
-1.11264020e-01 -1.92680693e+00 4.89151210e-01 7.11184561e-01
1.64127100e+00 -7.17474639e-01 6.55039012e-01 4.52472530e-02
-7.42366731e-01 -2.22897734e-02 -5.50878286e-01 2.91490883e-01
-1.16249651e-01 5.34788609e-01 -8.94650459e-01 1.19468309e-01
5.29730678e-01 5.40715873e-01 -3.77063453e-01 7.12883174e-01
-9.37965214e-02 4.22227263e-01 -1.40893124e-02 3.74027908e-01
-1.20908707e-01 4.57169265e-02 2.12429687e-01 8.48496377e-01
6.11773372e-01 4.46484014e-02 -3.03777307e-01 1.37017584e+00
-7.02621698e-01 -4.06972200e-01 -6.77450359e-01 -1.24054626e-02
4.59122479e-01 1.21933389e+00 6.18767999e-02 -2.60172963e-01
-8.77007693e-02 1.25830185e+00 3.18399578e-01 7.24512279e-01
-9.93462741e-01 -5.46947002e-01 8.66457045e-01 2.65125096e-01
2.04088539e-01 9.58293378e-02 -2.10298762e-01 -1.16732717e+00
1.09555893e-01 -8.20291758e-01 -1.36505708e-01 -9.85067487e-01
-1.10320973e+00 1.05771494e+00 -3.97676647e-01 -8.41854453e-01
-1.67644382e-01 -6.27428651e-01 -6.03154004e-01 1.23314548e+00
-1.36238027e+00 -1.11737502e+00 -7.72052824e-01 7.43665755e-01
4.82471228e-01 -6.05789363e-01 9.33741510e-01 3.98540556e-01
-8.23185265e-01 1.27029729e+00 -1.05580837e-02 1.99350834e-01
4.69558865e-01 -5.12118399e-01 8.23065862e-02 5.57169259e-01
-2.67047863e-02 5.56581676e-01 2.73186535e-01 -2.57941395e-01
-1.43288422e+00 -1.44870341e+00 8.89967024e-01 1.06874360e-02
-3.90563309e-02 -5.19519508e-01 -9.59297717e-01 4.87028509e-01
1.42128430e-02 4.33540583e-01 5.96319079e-01 -3.40407908e-01
-7.23060787e-01 -3.00629616e-01 -1.69142973e+00 4.51519847e-01
1.17149091e+00 -7.56723166e-01 -3.72621194e-02 1.27009675e-01
4.80258167e-01 1.09217063e-01 -8.03393364e-01 5.67293346e-01
8.50827634e-01 -8.08979332e-01 8.80143285e-01 -4.38081205e-01
7.55681992e-01 -2.80750275e-01 -4.62632000e-01 -1.28156841e+00
-2.70519387e-02 -3.87829334e-01 -4.39356193e-02 1.17272031e+00
6.30663812e-01 -7.24847198e-01 8.52826118e-01 4.61664259e-01
9.46519077e-02 -1.01406384e+00 -9.14321542e-01 -9.25316036e-01
-2.77181476e-01 -3.95598002e-02 6.74877167e-01 7.49589920e-01
-2.77974069e-01 4.54674274e-01 -6.47656083e-01 7.75325671e-02
8.59004676e-01 5.46039902e-02 6.47553444e-01 -8.84006858e-01
-3.33406746e-01 -1.75132543e-01 -2.75395542e-01 -9.65455234e-01
4.48875517e-01 -9.30219293e-01 7.23620281e-02 -1.08092439e+00
4.51831520e-01 -1.15443870e-01 1.32142073e-02 8.38899195e-01
3.08211535e-01 8.78016651e-01 2.96146929e-01 1.63756803e-01
-1.93505008e-02 1.17532361e+00 1.35549915e+00 -1.78226277e-01
5.23165166e-02 -2.86109209e-01 -7.15892911e-01 2.71523327e-01
7.34364688e-01 -9.86878797e-02 -4.19601798e-01 -5.01244545e-01
-7.08443880e-01 5.54008126e-01 7.35317826e-01 -1.07719314e+00
-4.40956019e-02 5.24586290e-02 9.81056631e-01 2.12052241e-02
9.27073658e-01 -5.63094079e-01 3.65705758e-01 3.91020089e-01
-4.72359031e-01 -4.13255483e-01 2.30240896e-01 2.79834300e-01
-3.66252095e-01 1.74892545e-01 1.17142653e+00 2.51332223e-01
-3.27938288e-01 5.65802515e-01 -1.61219668e-02 -5.12991428e-01
9.25885916e-01 -3.90847415e-01 -5.09771526e-01 -6.58398211e-01
-9.46683288e-01 -2.24883407e-02 3.12700719e-01 5.61296582e-01
1.36412716e+00 -1.59810483e+00 -8.74489665e-01 8.48695815e-01
1.18867658e-01 -4.81088847e-01 1.88742235e-01 5.13674021e-01
-6.80475757e-02 2.13301092e-01 -4.22108352e-01 -2.03671411e-01
-1.33186340e+00 3.31928790e-01 5.92212498e-01 9.10645351e-02
-3.13084960e-01 8.17633152e-01 3.64796549e-01 -4.00828451e-01
3.91020834e-01 2.93261617e-01 4.70153578e-02 -3.29105377e-01
8.06617975e-01 3.45783710e-01 -6.52628988e-02 -7.25873590e-01
-2.26692125e-01 4.90313351e-01 -2.96598412e-02 -1.41473606e-01
1.32506490e+00 2.03975528e-01 -7.34381825e-02 -9.45067406e-02
1.74651432e+00 -4.87379342e-01 -1.32862592e+00 -1.69916481e-01
-5.89913607e-01 -6.71404362e-01 5.46962693e-02 -1.14698315e+00
-1.59876180e+00 1.02983010e+00 9.29861784e-01 -2.50526249e-01
1.40789878e+00 -4.14785743e-03 7.69395947e-01 1.71213686e-01
1.84526056e-01 -4.36943799e-01 3.40001851e-01 3.23803872e-01
1.37329555e+00 -1.06696379e+00 -3.35865200e-01 -4.20055389e-01
-5.38891494e-01 9.24008727e-01 1.03976882e+00 -1.75883591e-01
5.69034040e-01 6.53956681e-02 -1.04359567e-01 -1.67824417e-01
-8.03918660e-01 9.52174962e-02 3.08154196e-01 5.99536479e-01
2.09640294e-01 9.78496447e-02 3.00317973e-01 5.52372336e-01
-1.12311922e-01 2.45285556e-01 2.87736088e-01 1.48910493e-01
-2.88119495e-01 -6.25639260e-01 -3.94448042e-01 5.11344075e-01
-1.12506226e-01 3.07371076e-02 -4.08783674e-01 5.11176765e-01
1.12723358e-01 9.56163943e-01 1.87613845e-01 -7.39146709e-01
5.30903637e-01 -5.75451665e-02 5.34321964e-01 -5.55609226e-01
-2.55552471e-01 -2.75373310e-01 -7.67868981e-02 -6.56564236e-01
1.19410880e-01 -2.14287952e-01 -1.02927411e+00 -5.04576683e-01
-2.03281254e-01 -8.90285801e-03 5.38963318e-01 5.71779907e-01
5.30587018e-01 1.59405470e-01 9.96862471e-01 -7.73330271e-01
-8.90394092e-01 -1.03282142e+00 -9.20951605e-01 2.31134236e-01
5.77452540e-01 -5.70266485e-01 -3.50791276e-01 -1.21210262e-01] | [12.784052848815918, 0.0031386511400341988] |
8add0f82-9b40-4f3a-81fa-a609377d4e0a | uni-fedrec-a-unified-privacy-preserving-news | 2109.05236 | null | https://arxiv.org/abs/2109.05236v1 | https://arxiv.org/pdf/2109.05236v1.pdf | Uni-FedRec: A Unified Privacy-Preserving News Recommendation Framework for Model Training and Online Serving | News recommendation is important for personalized online news services. Most existing news recommendation methods rely on centrally stored user behavior data to both train models offline and provide online recommendation services. However, user data is usually highly privacy-sensitive, and centrally storing them may raise privacy concerns and risks. In this paper, we propose a unified news recommendation framework, which can utilize user data locally stored in user clients to train models and serve users in a privacy-preserving way. Following a widely used paradigm in real-world recommender systems, our framework contains two stages. The first one is for candidate news generation (i.e., recall) and the second one is for candidate news ranking (i.e., ranking). At the recall stage, each client locally learns multiple interest representations from clicked news to comprehensively model user interests. These representations are uploaded to the server to recall candidate news from a large news pool, which are further distributed to the user client at the ranking stage for personalized news display. In addition, we propose an interest decomposer-aggregator method with perturbation noise to better protect private user information encoded in user interest representations. Besides, we collaboratively train both recall and ranking models on the data decentralized in a large number of user clients in a privacy-preserving way. Experiments on two real-world news datasets show that our method can outperform baseline methods and effectively protect user privacy. | ['Xing Xie', 'Yongfeng Huang', 'Chuhan Wu', 'Fangzhao Wu', 'Tao Qi'] | 2021-09-11 | null | https://aclanthology.org/2021.findings-emnlp.124 | https://aclanthology.org/2021.findings-emnlp.124.pdf | findings-emnlp-2021-11 | ['news-generation'] | ['natural-language-processing'] | [-0.17897792 -0.07581278 -0.6179729 -0.6684934 -0.9235289 -0.66471857
0.41909876 0.04734858 -0.16199298 0.6233099 0.86244285 -0.01878562
0.1297399 -1.1967126 -0.71191037 -0.7522254 0.485189 0.3391466
0.21453911 -0.24393813 0.20857628 0.19259854 -1.5213549 0.8135222
0.6970169 1.2588408 -0.22781214 0.1296038 0.1366349 0.85775614
-0.3628686 -0.6336235 0.46456653 -0.13971648 -0.7458713 -0.28275284
-0.0516166 -0.83257127 -0.5847052 0.99842495 0.5621729 0.8461334
0.11005379 -1.0412505 -1.2634641 0.9191208 -0.29298574 -0.0767063
0.33305013 -0.49219152 1.1346185 -0.5012356 0.60285026 0.8232122
0.23126017 0.83573586 -1.000776 -0.84126765 0.5855692 -0.02899089
-1.0462655 -0.6858962 0.6205774 -0.08424204 0.39792758 1.0306844
0.5014182 1.1934215 -0.0782458 1.0915656 0.7362937 0.3311568
0.5108018 0.7577837 0.54324394 -0.00479973 0.20030451 0.12146949
-0.38998985 -0.92905957 0.4607987 0.9492329 -0.5205831 -0.4212196
-0.6961697 1.0508955 0.3408655 -0.08920268 -0.41694316 -0.36745995
0.49987012 0.65913755 0.9830947 0.30343005 -0.6710005 0.3342787
-0.592969 0.38441995 1.06461 1.0425313 0.42334357 -0.35601786
-0.39106354 0.90230197 0.5014093 0.23896958 0.79454297 -0.36318162
0.34588948 0.43134853 0.50933266 -1.2092482 -0.0403947 -0.5345499
-0.79404056 -0.41227087 0.04680246 -0.35849068 -0.29167837 1.4510552
0.63968563 0.5601041 0.31631362 1.0489385 0.9851056 1.1519436
-0.15524168 -0.58555925 1.4980265 -1.1002376 -0.5007786 0.29716504
0.46412796 -0.40515977 0.8415471 0.40810886 -0.91251427 -0.02105786
-0.48987278 -0.2894468 -0.28648758 0.21868387 0.509249 0.6590525
-0.69935125 0.37177482 -0.6869942 -0.1635673 0.39591432 0.5518219
-0.24417774 0.05377859 -1.2640479 0.15355718 -0.01810582 -0.2763366
-0.6369257 -0.7076469 -0.44296703 0.4476017 0.50829095 -0.7130487
1.3317145 -0.53356546 -1.8631423 0.2595134 -0.02149608 -0.4313234
0.5118214 -0.39354524 -0.7835195 -0.4348639 -0.21937115 -0.4887856
0.6548143 -1.3272631 -1.0780565 -0.5240819 0.04444795 0.14844005
-0.5009665 0.370948 -0.5392826 -0.7453393 0.09936894 -0.7591891
-0.3399227 -0.44145465 -0.41545242 -0.0423374 0.89577496 -0.77447814
1.2573696 -2.1297379 -0.32379624 0.74280775 0.13899201 0.27284387
0.13222206 0.40413377 0.12382917 0.00477944 0.50589633 -0.50018626
0.04942153 0.00784794 -1.1445643 0.4789872 -0.95204496 0.5763893
-0.6850271 0.15729894 -0.13465878 0.50319093 -0.98504496 0.44043285
-0.18621664 0.40456724 -1.1143764 0.3241184 0.8330325 -0.20289971
0.1626401 0.09767542 -0.06763699 0.64300114 -1.2665299 1.091735
-0.4979357 -0.3672014 0.15008518 -0.74148184 1.0593915 0.5524163
0.5430291 -0.39100933 0.2691389 -0.23131812 -0.8569053 -0.195619
0.879702 0.33279145 -0.09310609 1.2833143 -0.3539761 0.89188105
-0.579983 0.22319116 0.681976 -0.4450387 0.24319094 -0.00579284
0.6457254 -0.5410589 0.87541443 0.78610116 0.24388552 0.50167567
0.17025231 -0.7243438 -0.42580688 -0.64034086 0.10556599 1.8837904
0.33783638 -0.7138458 -0.5012909 -1.1729908 0.01648642 0.9884157
-0.5746925 -0.42999348 -0.23042254 -0.60354877 -0.10208727 0.3249869
0.2699476 -0.85579413 0.12933165 0.27438962 -0.2435311 -0.20030573
-1.0088519 -0.4293908 -0.59740126 -0.6825491 -0.49509102 -0.41057873
0.9464838 0.87541866 0.41539413 -0.10445157 0.5333244 0.30107695
-0.57929265 -0.26853615 -0.02543357 0.1095708 0.3192618 0.7770253
0.47203222 -0.66151106 -0.8212477 0.7328884 -1.0108712 -0.28947526
0.11138405 0.67862886 0.89159787 0.03859386 0.70332634 -1.6262236
0.92395943 -1.0764413 -0.75457734 0.5622464 -0.96405137 -0.2666146
1.2035629 -0.5237988 -1.3583091 0.03889731 -0.2211347 0.06977855
-0.04497195 0.31971028 -0.64171267 0.10614718 0.5529416 0.5005455
-0.23972403 -0.90389806 0.6181653 1.0161313 0.25982264 -0.42666748
0.55461377 0.43633783 -0.88744634 -0.20030183 -0.8339796 -0.69847125
0.29772723 0.31242752 0.30214646 -0.9502294 -0.91279924 0.19363335
-0.6603644 0.48112917 -0.52693343 0.6748148 -0.40159163 0.17609097
-0.7275475 -0.7882352 -0.86704767 -0.9659954 0.6115023 0.37633964
0.18435283 -0.54013765 0.2711804 0.5849654 0.5517294 -0.32352108
0.30648822 -1.6265982 -0.53419304 -0.77901536 -0.15064658 0.12234524
0.15349866 -0.41847643 -0.9112027 -0.4602016 0.316159 -0.05816806
0.68848187 0.14441483 1.4889476 -1.3797777 -0.2981504 0.7871856
0.93081635 0.21860863 0.63793373 0.18064591 0.54269475 0.3965902
0.54522616 0.855325 0.593898 0.5359744 0.09357771 0.30886608
0.79764956 -0.6509153 0.35077575 0.6230067 -0.1799564 -0.4279976
0.11854647 -0.03807386 -2.4207864 -1.3105627 0.39250603 2.5036585
0.8363864 -0.47980398 0.35133576 -0.4253526 0.70770717 0.27352408
-0.69750273 -0.12882753 0.35536107 -0.32765 0.5880951 0.09085115
-1.0269217 0.72137755 4.671841 0.53401256 -1.0773724 0.2763338
0.63257074 -0.47627878 -0.73558754 -0.01936169 -0.9197818 0.71732837
0.75200695 -0.7238833 0.6401464 1.6728956 0.15317525 0.6837265
-0.8065152 0.84133697 0.10044322 -1.5851 0.06920774 0.2660549
0.88410765 -0.04193291 0.29858935 0.33752698 0.85194415 -0.3895008
0.318433 0.75643855 0.22293858 -1.2394159 0.64694905 0.5704933
-0.8961788 -0.30096158 -0.8971386 0.3996655 0.17907158 0.5062021
-0.3422189 0.55547297 0.72229475 0.62284195 -0.00592597 1.1404041
-0.00927441 0.5973056 -0.4361498 -0.08594269 -0.3145506 -0.4452548
0.21692638 0.6467325 0.49061775 0.5953381 0.2763085 0.2712993
-0.5609605 0.75653017 -0.42032474 0.29200155 0.6029549 1.2897474
-0.22281142 -0.32108864 -0.47282687 0.97109777 0.22993295 0.33889684
-0.6331242 -0.23731875 0.8883139 0.23727621 0.3180528 0.43157068
0.22427452 -1.7703619 -0.02152694 -0.9190715 0.9306557 -0.2547988
-1.5450774 0.5609723 -0.44569966 -1.2500399 -0.11394046 0.19718213
-0.88112754 0.6452016 -1.0138471 -1.2502965 0.11038059 1.2698905
0.08778454 -0.35768166 1.084809 0.3470492 -0.7507108 1.2837684
1.0024749 0.01239396 0.7737305 -0.67283356 0.42837214 0.65921533
0.21771134 1.2832043 0.33965468 -0.6947792 -1.3870971 -1.5270517
0.8328696 -0.21701409 0.22469693 -0.52902025 -0.98881245 1.0914135
-0.42605492 0.20783421 1.4642727 0.29051796 -0.68459505 -0.39545432
-1.6191037 0.59870154 0.68511784 -0.59812105 -0.48189417 0.5924948
1.0374305 -0.33334866 -0.83381313 -0.2145477 0.75747883 -0.6075673
0.8875886 -1.2272675 -0.10182326 -0.32954547 -0.09886683 -1.193051
-0.6499873 -1.0931188 -0.52534395 1.5734832 0.49674672 -1.1417078
1.0558534 1.4899359 0.17771141 -0.49259612 -0.66564375 -0.24278398
-0.5055266 -0.14197887 1.310104 1.0467167 0.24949405 0.07939554
-1.0817691 0.46701175 0.2890449 0.64788765 1.016967 -1.2817849
-0.41295916 0.09327931 0.18392472 -1.2393479 0.04479216 -0.8099366
-0.2806807 -1.4207526 0.01082547 -0.48824412 -0.72896904 0.4903244
0.08439415 -0.3387303 -0.19155316 0.53073657 -0.88534725 0.5394174
0.76941407 0.28935298 -0.5861131 1.2910664 -1.6890563 0.5534458
0.7595368 -0.60112387 -0.80275816 -0.11265156 0.14666414 0.12494156
-0.18784264 -0.43130037 0.544213 -0.5002133 0.27937907 -0.47569248
0.14602338 -1.1872746 0.21476188 -0.05707371 -0.9007056 -0.36323237
-0.5703629 1.0600734 0.18620299 0.14027908 0.49849665 -0.13424703
-0.20252064 0.9632538 -0.12937683 -0.34462923 1.1335008 0.37042502
-0.681205 -0.88764745 -0.80847895 0.44600484 0.39956018 0.5804876
0.5488141 -1.342434 -0.52668715 0.4800325 0.24071771 -0.17308025
0.6437238 0.23928621 0.28253582 0.01754672 0.16525628 0.40217093
-1.2392842 1.023806 -0.16777664 -0.2244204 -0.69065434 1.0294238
0.26651368 -0.59386754 0.5635617 -0.05465535 -0.48050117 0.07171571
1.4402732 0.43765518 -0.07180413 -0.61120987 -0.03900702 -0.1084864
-0.82722807 0.11392932 1.6014143 -0.4574699 -0.15841569 -0.17603372
1.2963107 0.6857551 -0.9484743 -0.7081797 -0.5113457 -0.8243813
0.09604973 -0.65180564 -1.676746 0.08759277 0.34196874 0.3508524
1.1362836 -0.13360137 1.3839579 0.6252856 0.42312276 -1.1824684
-0.7220017 0.12392178 0.7429116 -0.96971244 0.12344367 -0.16311653
-1.1129183 0.6863309 0.537763 -0.17085701 1.0545974 -0.15636843
0.02184245 0.19311243 -0.82965153 0.4950889 0.3647553 0.49653533
0.03205352 0.11618759 -0.2689885 1.9829149 -0.23587547 0.12208955
0.2680238 0.78990376 -0.47065306 -1.2855831 -0.09241732 0.8830576
-0.7413004 0.03325878 -0.33618936 -0.15851504 -0.27167177 0.99687004
-0.15209575 -0.92292184 0.565335 -0.12659726 -0.6462698 -0.64687276
-1.0902536 -0.01817701 0.10128947 -0.8620483 0.2970031 -0.6303247
-0.93680865 -0.57656926 -0.59305656 0.87492377 0.4799139 0.55260295
0.9172626 -0.16887605 1.3616552 -0.25278437 -0.9276345 -0.31921867
-0.86744535 0.23223041 0.17506142 -0.14571433 -0.30756506 -0.21796109] | [5.919621467590332, 6.4227142333984375] |
85b345dd-dc50-4c53-9c73-4198a6e87015 | subword-mapping-and-anchoring-across | 2109.04556 | null | https://arxiv.org/abs/2109.04556v1 | https://arxiv.org/pdf/2109.04556v1.pdf | Subword Mapping and Anchoring across Languages | State-of-the-art multilingual systems rely on shared vocabularies that sufficiently cover all considered languages. To this end, a simple and frequently used approach makes use of subword vocabularies constructed jointly over several languages. We hypothesize that such vocabularies are suboptimal due to false positives (identical subwords with different meanings across languages) and false negatives (different subwords with similar meanings). To address these issues, we propose Subword Mapping and Anchoring across Languages (SMALA), a method to construct bilingual subword vocabularies. SMALA extracts subword alignments using an unsupervised state-of-the-art mapping technique and uses them to create cross-lingual anchors based on subword similarities. We demonstrate the benefits of SMALA for cross-lingual natural language inference (XNLI), where it improves zero-shot transfer to an unseen language without task-specific data, but only by sharing subword embeddings. Moreover, in neural machine translation, we show that joint subword vocabularies obtained with SMALA lead to higher BLEU scores on sentences that contain many false positives and false negatives. | ['Andrei Popescu-Belis', 'Giorgos Vernikos'] | 2021-09-09 | null | https://aclanthology.org/2021.findings-emnlp.224 | https://aclanthology.org/2021.findings-emnlp.224.pdf | findings-emnlp-2021-11 | ['cross-lingual-natural-language-inference'] | ['natural-language-processing'] | [-1.00270741e-01 -1.17979825e-01 -5.52616239e-01 -4.68886465e-01
-1.20681489e+00 -9.56275225e-01 6.94758058e-01 3.32792312e-01
-8.02329242e-01 1.14058423e+00 3.56363714e-01 -3.52111250e-01
1.76391467e-01 -6.32999837e-01 -9.55660462e-01 -4.22913164e-01
2.79240668e-01 6.34024382e-01 3.64769921e-02 -6.11742318e-01
-2.85099417e-01 6.60420880e-02 -1.27477098e+00 2.98685223e-01
8.44625056e-01 1.69667587e-01 2.58336902e-01 2.71481574e-01
-6.99718416e-01 6.40889183e-02 -3.90398204e-01 -8.94809306e-01
3.45615745e-01 -3.93551022e-01 -5.77390671e-01 -5.30503273e-01
8.76868486e-01 -2.47260220e-02 2.04862237e-01 1.29148376e+00
3.20857763e-01 -2.64356524e-01 5.27381301e-01 -1.01184142e+00
-1.09714246e+00 9.94439542e-01 -2.90569365e-01 1.06852107e-01
2.76960313e-01 -4.14920747e-02 1.61019957e+00 -1.12225425e+00
8.85003746e-01 1.17545509e+00 7.74755239e-01 6.49482906e-01
-1.61632073e+00 -1.05342448e+00 1.01485781e-01 -3.07552796e-03
-1.67555726e+00 -5.69698632e-01 4.79034156e-01 -3.40703875e-01
1.23796403e+00 2.08929256e-01 3.01410705e-01 1.50303030e+00
3.75839025e-01 5.49136221e-01 1.12139809e+00 -8.43463242e-01
-4.70888615e-02 4.37599540e-01 3.33139390e-01 6.33565784e-01
7.15008318e-01 -3.67787853e-02 -5.99093616e-01 -4.51367423e-02
2.80559361e-01 -3.63600582e-01 -2.48664841e-01 -4.08233434e-01
-1.53103518e+00 9.13404465e-01 6.56237081e-02 7.89167702e-01
-8.36486891e-02 4.57862467e-02 5.92717111e-01 6.90889001e-01
4.41331506e-01 6.12244785e-01 -6.42751813e-01 2.99900174e-01
-7.64489174e-01 1.56771928e-01 6.46929383e-01 1.28540397e+00
1.09080982e+00 -3.34061474e-01 -5.58298454e-02 9.53207374e-01
1.82828352e-01 8.61090779e-01 8.71160150e-01 -2.37107500e-01
6.08659446e-01 3.79290253e-01 -4.51888144e-02 -6.40203357e-01
-2.04043075e-01 -1.48342997e-01 -5.94852686e-01 -2.72295773e-01
2.92396396e-01 -1.99047308e-02 -6.79485083e-01 2.20344353e+00
8.20056424e-02 -1.55854166e-01 4.24134880e-01 6.29242361e-01
5.55136859e-01 4.15433615e-01 1.66827053e-01 -9.34941843e-02
1.65787792e+00 -7.29807258e-01 -8.42127085e-01 -4.23133433e-01
9.32528794e-01 -9.62275803e-01 1.37789726e+00 1.71884477e-01
-6.97299063e-01 -6.17507875e-01 -1.22909260e+00 -4.18939352e-01
-8.22351038e-01 -1.71578992e-02 3.40289503e-01 7.55293608e-01
-9.67252791e-01 1.74745113e-01 -5.99728048e-01 -6.14207029e-01
-4.80925180e-02 2.03427836e-01 -7.86096931e-01 -2.36307174e-01
-1.52329016e+00 1.20670831e+00 5.76974034e-01 -3.40871185e-01
-5.45107365e-01 -6.56521618e-01 -1.13575292e+00 -1.66295424e-01
2.08599404e-01 -4.64800507e-01 8.73048961e-01 -1.05744958e+00
-1.10217464e+00 1.14514899e+00 -3.42323542e-01 -4.91839767e-01
8.33561420e-02 -3.28610629e-01 -6.16596937e-01 -3.99030685e-01
3.94867927e-01 6.59147263e-01 5.52677095e-01 -9.37498927e-01
-4.09717947e-01 -1.81456715e-01 -1.59688681e-01 -3.92334163e-02
-6.11451566e-01 6.76169023e-02 -3.42488945e-01 -8.75458062e-01
-2.98051178e-01 -8.71483564e-01 1.30161569e-01 -1.76042870e-01
-5.24288081e-02 -4.59935337e-01 2.00081423e-01 -6.61784410e-01
1.24370551e+00 -1.88280177e+00 3.92154574e-01 7.53050596e-02
-5.13357855e-02 3.48896474e-01 -5.58274448e-01 6.55950665e-01
2.12527201e-01 1.85097009e-01 -2.44644776e-01 -4.46792215e-01
1.72106698e-01 6.91345096e-01 -5.29083073e-01 3.20633918e-01
3.81082624e-01 1.11731064e+00 -1.10840762e+00 -4.92918253e-01
3.28051113e-02 4.44870859e-01 -4.15660411e-01 -7.68814683e-02
-1.81554884e-01 1.85736209e-01 -2.04946116e-01 3.84642363e-01
4.74761546e-01 6.17447309e-02 6.07354939e-01 -2.74328053e-01
-7.54937828e-02 6.51680589e-01 -7.74238884e-01 1.99074888e+00
-9.25075769e-01 4.75543797e-01 -3.37961465e-01 -8.34209025e-01
1.03220522e+00 5.00434756e-01 -5.25470413e-02 -6.67900801e-01
9.19924378e-02 7.78436422e-01 -1.58523321e-01 -3.42737973e-01
5.52393854e-01 -4.58506972e-01 -4.05429482e-01 4.46969479e-01
7.35153198e-01 -5.20782173e-02 3.19469810e-01 -7.30619654e-02
8.16666067e-01 2.11092889e-01 7.92425871e-01 -5.11634648e-01
6.75265253e-01 4.20678072e-02 4.29398119e-01 5.58620095e-01
2.96363491e-04 2.57314950e-01 2.87594795e-02 -4.07242805e-01
-1.36459994e+00 -1.19013166e+00 -2.25593835e-01 1.18649852e+00
8.73269979e-03 -5.55207729e-01 -5.74957728e-01 -7.66001940e-01
7.18116239e-02 7.50607312e-01 -6.43305719e-01 -4.43548448e-02
-7.93835282e-01 -3.77283245e-01 9.86043394e-01 3.30965102e-01
-3.33334506e-02 -9.11748469e-01 -1.12531833e-01 3.26680541e-01
-2.79098719e-01 -1.45391846e+00 -6.83693528e-01 3.32323134e-01
-4.66894776e-01 -8.62125635e-01 -7.50114918e-01 -1.03066134e+00
6.23776734e-01 2.82884210e-01 1.39555311e+00 -1.03701778e-01
-1.16869966e-02 8.60877484e-02 -4.43275601e-01 -2.75244534e-01
-8.10609221e-01 3.62584770e-01 5.40574968e-01 -1.37086317e-01
1.13260269e+00 -4.98295933e-01 1.22683398e-01 1.76963806e-01
-8.89541924e-01 -1.36473803e-02 6.24172270e-01 9.70043838e-01
7.53768206e-01 -7.87754714e-01 6.92561567e-01 -9.39793646e-01
6.37038767e-01 -4.32123542e-01 -9.36309397e-01 6.72892630e-01
-5.39318562e-01 6.19285643e-01 9.16635215e-01 -5.88656068e-01
-4.59086001e-01 -2.59503335e-01 -1.16001785e-01 -3.29329044e-01
-3.90417762e-02 3.77193272e-01 -5.97282499e-02 1.47178501e-01
7.10358977e-01 2.79393196e-01 -1.29586935e-01 -5.00783920e-01
7.95363188e-01 7.94131696e-01 3.85472387e-01 -7.04710960e-01
8.07488203e-01 1.62723541e-01 -2.67892629e-01 -6.72958791e-01
-9.47618604e-01 -6.10705674e-01 -8.40358019e-01 2.05491289e-01
1.17912698e+00 -1.21886170e+00 -1.13281444e-01 -1.22886233e-01
-1.41740322e+00 1.28086582e-01 -1.71026945e-01 7.14505613e-01
-3.01428169e-01 9.94469821e-02 -2.56809175e-01 -3.50169271e-01
-5.23653984e-01 -1.21595335e+00 1.26490068e+00 -3.32101822e-01
-6.61734104e-01 -1.14163828e+00 5.47571957e-01 1.42577931e-01
3.59853655e-01 -2.20971718e-01 1.05739331e+00 -1.00390816e+00
-3.82692665e-01 -5.09399828e-03 -5.82456887e-02 5.40964007e-01
4.75008488e-01 -5.50056159e-01 -9.38695073e-01 -5.13141036e-01
-3.64853621e-01 -4.23943251e-01 5.10775864e-01 -1.14680424e-01
3.28253120e-01 -2.53499269e-01 -2.50619531e-01 4.26469654e-01
1.62800932e+00 -3.47446561e-01 1.51623800e-01 2.65892446e-01
6.88810289e-01 5.14467537e-01 3.27518314e-01 -2.70388633e-01
5.29251695e-01 9.04625416e-01 -1.40046343e-01 9.90849286e-02
-2.96724051e-01 -4.89329338e-01 7.52756357e-01 1.68400455e+00
4.06027675e-01 -6.36369810e-02 -9.52317119e-01 1.09519970e+00
-1.46628499e+00 -6.70906782e-01 6.51454031e-02 2.32371521e+00
1.27574527e+00 6.01671413e-02 -2.40293473e-01 -3.29645336e-01
5.64173639e-01 4.59387116e-02 -1.39031574e-01 -5.58107197e-01
-5.41967273e-01 7.15808868e-01 9.03161705e-01 8.95393074e-01
-8.22346330e-01 1.37420630e+00 5.68129587e+00 8.87684882e-01
-1.02492666e+00 6.78517699e-01 -1.15696304e-01 1.73914671e-01
-8.30216944e-01 1.22171408e-02 -1.22420466e+00 2.76381493e-01
8.73540044e-01 -3.37310582e-01 5.73484540e-01 5.79922915e-01
-3.47272933e-01 3.43221128e-01 -1.22595561e+00 7.88131654e-01
5.31994581e-01 -1.25190902e+00 3.89893711e-01 -1.99265257e-01
7.55068004e-01 5.50085366e-01 -3.77324462e-01 4.69997555e-01
6.77870750e-01 -7.87341177e-01 6.22338355e-01 1.31573632e-01
1.03836882e+00 -5.99969685e-01 8.22928846e-01 1.46498859e-01
-1.30613887e+00 4.64071125e-01 -6.04469478e-01 2.10892349e-01
1.87969983e-01 4.02523994e-01 -8.34207237e-01 7.18083501e-01
2.58269012e-01 5.22877514e-01 -4.18345392e-01 2.84088016e-01
-6.05749369e-01 3.10510010e-01 -3.24115515e-01 -2.51779944e-01
2.99248099e-01 -2.14534938e-01 5.74692726e-01 1.45968878e+00
4.12592143e-01 -5.73509693e-01 1.97370827e-01 7.39950001e-01
-4.38011438e-01 6.65835559e-01 -1.05580044e+00 -1.83357131e-02
6.17613673e-01 9.37394321e-01 -2.39427939e-01 -4.85787362e-01
-7.97483921e-01 1.09348297e+00 7.42793918e-01 1.89263761e-01
-6.60991073e-01 -3.47633004e-01 1.09455621e+00 -1.92308769e-01
3.17484051e-01 -3.91869009e-01 -1.66746065e-01 -1.54186356e+00
2.26327434e-01 -1.02017367e+00 1.77380621e-01 -4.84141201e-01
-1.57495010e+00 8.71825278e-01 8.65314528e-02 -1.32255387e+00
-2.46801674e-01 -7.10995913e-01 2.23895553e-02 9.32871521e-01
-1.77037883e+00 -1.47828829e+00 1.82343632e-01 4.95973319e-01
7.15822458e-01 -4.43626732e-01 1.33844543e+00 5.57307303e-01
-1.50845330e-02 1.00620031e+00 7.50013292e-02 2.82013923e-01
1.15520275e+00 -1.09031403e+00 9.13338721e-01 9.43847001e-01
8.24118555e-01 1.01679397e+00 7.24858284e-01 -6.44995451e-01
-1.37132192e+00 -1.28338397e+00 1.71319139e+00 -7.03796268e-01
1.10775006e+00 -8.59359205e-01 -1.01738286e+00 9.00619030e-01
3.12041581e-01 1.24920070e-01 9.82766747e-01 4.21887636e-01
-8.57630968e-01 -5.50014488e-02 -7.85448968e-01 7.60537028e-01
9.55129087e-01 -9.27322567e-01 -1.07447028e+00 3.76618087e-01
9.96122420e-01 2.86840145e-02 -7.56168067e-01 1.99398175e-01
7.99680412e-01 -2.64472067e-01 7.60873914e-01 -8.81789923e-01
2.03529596e-01 -2.76631564e-01 -6.33279681e-01 -1.41942346e+00
3.59498896e-03 -4.17713791e-01 3.01075339e-01 1.18672502e+00
7.00893402e-01 -8.17198813e-01 4.42986935e-02 -1.01616286e-01
1.07297122e-01 -2.80473977e-01 -1.41870296e+00 -1.25169623e+00
5.46825588e-01 -5.35252333e-01 8.59240353e-01 1.26384735e+00
3.36770825e-02 6.92812562e-01 -5.07704139e-01 2.23876521e-01
4.24916595e-01 4.83476510e-03 9.20188785e-01 -1.05335081e+00
-2.13247284e-01 -1.76621869e-01 -3.93680692e-01 -9.70114946e-01
5.79885721e-01 -1.47409153e+00 1.31034628e-01 -1.10356498e+00
8.45702142e-02 -5.08074939e-01 -4.73831862e-01 6.42249107e-01
-2.17655674e-01 6.38599992e-01 1.54850677e-01 1.78558469e-01
-3.71886104e-01 4.98362750e-01 8.13572705e-01 -1.79098561e-01
4.54671457e-02 -8.12280476e-01 -4.70739484e-01 5.15800893e-01
5.76135993e-01 -8.71356368e-01 -1.47194937e-01 -9.36652720e-01
5.58471203e-01 -4.74775195e-01 -3.38652194e-03 -7.26682127e-01
1.90892446e-04 -5.86597286e-02 -1.86954498e-01 -2.00553089e-01
1.25082090e-01 -9.97793853e-01 6.66992739e-02 3.27428013e-01
-3.68858337e-01 4.37317669e-01 3.43164444e-01 2.85568923e-01
-2.14263454e-01 -2.19869390e-01 5.66982567e-01 1.71903856e-02
-4.69889879e-01 -9.19053257e-02 -1.98024467e-01 2.20122650e-01
8.22107017e-01 1.20781891e-01 -1.33866683e-01 1.22091599e-01
-3.94659281e-01 1.17404580e-01 5.07570863e-01 9.76017594e-01
1.81477472e-01 -1.64015579e+00 -9.17533100e-01 5.79332054e-01
8.77281606e-01 -5.09797394e-01 -3.72258514e-01 6.99624538e-01
-2.46908262e-01 6.62818313e-01 -2.95189589e-01 -5.07411718e-01
-1.33362079e+00 5.23690164e-01 1.33618549e-01 -4.33815658e-01
-4.92155880e-01 7.38017619e-01 1.19912244e-01 -1.04178166e+00
-1.54839575e-01 -5.93759716e-01 1.14753738e-01 2.01870814e-01
4.69086349e-01 -3.05706710e-01 3.18626046e-01 -9.13132489e-01
-4.54323173e-01 8.76664162e-01 -3.22728455e-01 -1.92582682e-01
1.04513955e+00 -2.03461945e-01 -1.25024438e-01 7.79093325e-01
1.26075125e+00 5.43681204e-01 -3.50735843e-01 -6.62854135e-01
2.35224023e-01 -2.51834899e-01 -1.86698288e-01 -5.38072407e-01
-4.57376838e-01 7.27927864e-01 4.79913414e-01 -1.68442622e-01
7.28209794e-01 -1.33903222e-02 1.12358785e+00 5.47357559e-01
7.93809712e-01 -9.27534640e-01 -5.56340873e-01 7.58609056e-01
6.90790415e-01 -1.34783065e+00 -1.21300936e-01 -1.07557453e-01
-4.79032815e-01 9.36571836e-01 9.23733562e-02 -7.19468966e-02
4.26397681e-01 1.81020454e-01 5.05941808e-01 -1.60860608e-03
-7.31895328e-01 -2.61750609e-01 5.46320081e-01 4.34787333e-01
6.41184509e-01 3.19915563e-01 -7.89760530e-01 5.35891116e-01
-3.41361046e-01 -9.87695456e-02 1.10372424e-01 6.44319355e-01
-1.72046870e-01 -1.78236938e+00 -2.16934308e-01 1.39179863e-02
-5.05314231e-01 -6.69005275e-01 -2.38892689e-01 8.42858493e-01
2.70360678e-01 5.66357732e-01 -3.75378877e-02 -1.84843421e-01
1.36379927e-01 4.45880681e-01 6.22227609e-01 -7.80213058e-01
-5.79140663e-01 -2.83678770e-01 1.42491534e-01 -3.24770302e-01
-4.03048635e-01 -3.17565978e-01 -9.15811658e-01 2.90220454e-02
-4.68113780e-01 1.61447778e-01 7.73577094e-01 1.13443506e+00
2.86165118e-01 2.59255588e-01 2.03978837e-01 -3.79219621e-01
-5.97491026e-01 -1.00235331e+00 -1.02084003e-01 5.59529364e-01
5.12171328e-01 -6.05509639e-01 -1.71726793e-01 1.10757083e-01] | [11.055376052856445, 10.0125150680542] |
45f6db7f-3b39-488c-8f9c-3f219be9db0e | optimality-of-variational-inference-for-1 | null | null | http://proceedings.neurips.cc/paper/2021/hash/a5e308070bd6dd3cc56283f2313522de-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/a5e308070bd6dd3cc56283f2313522de-Paper.pdf | Optimality of variational inference for stochasticblock model with missing links | Variational methods are extremely popular in the analysis of network data. Statistical guarantees obtained for these methods typically provide asymptotic normality for the problem of estimation of global model parameters under the stochastic block model. In the present work, we consider the case of networks with missing links that is important in application and show that the variational approximation to the maximum likelihood estimator converges at the minimax rate. This provides the first minimax optimal and tractable estimator for the problem of parameter estimation for the stochastic block model with missing links. We complement our results with numerical studies of simulated and real networks, which confirm the advantages of this estimator over current methods. | ['Olga Klopp', 'Solenne Gaucher'] | 2021-12-01 | null | https://openreview.net/forum?id=iYzsR0JNaa2 | https://openreview.net/pdf?id=iYzsR0JNaa2 | neurips-2021-12 | ['stochastic-block-model'] | ['graphs'] | [ 3.72470655e-02 3.04924428e-01 -5.93142688e-01 -4.58405949e-02
-5.58695614e-01 -2.65169173e-01 3.12484801e-01 -9.22902822e-02
-2.50021279e-01 1.23748064e+00 -2.42812574e-01 -6.32264555e-01
-7.78720617e-01 -6.02907538e-01 -7.74888694e-01 -7.33238399e-01
-6.14691556e-01 5.32936394e-01 2.18144562e-02 6.78159818e-02
-2.95298770e-02 3.87197345e-01 -6.54998899e-01 -7.01755226e-01
6.72401667e-01 7.83845127e-01 7.82280695e-03 6.12635016e-01
2.32677311e-01 3.40753078e-01 -3.05941820e-01 -4.96828318e-01
3.25801551e-01 -3.42682868e-01 -4.79778677e-01 5.85834049e-02
9.55177993e-02 -3.57569963e-01 -5.11854589e-01 1.15993893e+00
3.92415851e-01 3.06053068e-02 8.61677766e-01 -1.67466283e+00
1.38090372e-01 8.78796339e-01 -8.59274924e-01 4.92795885e-01
-2.26151310e-02 -3.51056337e-01 1.27808535e+00 -5.78665197e-01
6.29904568e-01 1.31987441e+00 9.12705541e-01 4.18700933e-01
-1.62673366e+00 -6.88271642e-01 -2.75771972e-02 -1.43611223e-01
-1.76803207e+00 -5.74498236e-01 6.96287692e-01 -4.25686806e-01
2.26544887e-01 1.63140506e-01 3.60951722e-01 1.08464074e+00
9.97627527e-02 6.41819954e-01 7.27100432e-01 -2.49773413e-01
2.91748166e-01 9.78737772e-02 2.21304595e-01 5.58182895e-01
7.86041737e-01 1.45354956e-01 -2.69759625e-01 -7.01109111e-01
1.01468599e+00 -8.96133408e-02 -1.44526914e-01 -7.20452666e-01
-8.00663888e-01 1.16953635e+00 1.50112242e-01 -1.68209195e-01
-6.56574965e-01 6.58775747e-01 5.66332862e-02 2.87693918e-01
7.88895667e-01 -3.36136729e-01 -1.13954157e-01 -1.54845461e-01
-1.14110911e+00 1.76164001e-01 1.23700595e+00 1.06981397e+00
3.47526848e-01 1.36928901e-01 8.09039176e-02 6.75211191e-01
8.55296373e-01 8.60512316e-01 -7.92537510e-01 -1.40824366e+00
6.55330479e-01 -2.06652582e-01 3.85308594e-01 -7.93239534e-01
-2.82534778e-01 -9.43706810e-01 -1.11763740e+00 -3.50698382e-02
4.57382321e-01 -6.95984304e-01 -2.38497198e-01 2.06513286e+00
2.97230154e-01 2.25348547e-01 -1.61971107e-01 5.40841520e-01
4.99006882e-02 7.07400620e-01 -1.52059987e-01 -1.02338767e+00
6.16723955e-01 -4.72871959e-01 -9.54010844e-01 1.11165993e-01
5.10402501e-01 -3.83531511e-01 1.13323137e-01 3.22010040e-01
-1.24125803e+00 9.44225118e-02 -7.54590929e-01 7.27164328e-01
3.30502748e-01 -4.06692803e-01 3.36549342e-01 8.23071063e-01
-1.18362081e+00 6.63426638e-01 -8.47967982e-01 -5.62342525e-01
4.51929450e-01 3.26632351e-01 -1.27317291e-02 7.72512611e-03
-1.03071761e+00 5.38489401e-01 -4.36196290e-02 6.09966993e-01
-9.21214819e-01 -7.71013796e-01 -6.26870155e-01 2.49711573e-01
6.20090306e-01 -4.49305773e-01 1.07041776e+00 -7.68733144e-01
-1.12662303e+00 5.17241620e-02 -4.42549020e-01 -6.71898544e-01
9.45159674e-01 1.70472652e-01 1.28561705e-01 1.45109951e-01
1.75679788e-01 -1.07490577e-01 7.58889019e-01 -1.20635259e+00
-1.10529974e-01 -1.12065999e-02 -6.98445505e-03 -2.43538171e-01
-2.97684856e-02 -1.66686267e-01 -2.89002687e-01 -2.92219698e-01
2.61596255e-02 -8.11495543e-01 -3.84159178e-01 3.42173517e-01
-5.35653949e-01 -1.16859652e-01 5.22126198e-01 -5.97331166e-01
1.20003879e+00 -1.88100171e+00 1.72590673e-01 9.51611340e-01
2.07344726e-01 -3.51637691e-01 -7.11290538e-03 9.25693333e-01
8.94531235e-02 4.51607972e-01 -4.03933972e-01 -4.97881442e-01
8.77452865e-02 2.43770063e-01 1.34745277e-02 1.11036146e+00
-3.46581459e-01 4.21570748e-01 -7.28729844e-01 -7.79579461e-01
1.65367629e-02 2.69332707e-01 -5.51446021e-01 -1.00522719e-01
2.80010909e-01 -2.02755891e-02 -7.00958133e-01 1.52850673e-01
7.42066443e-01 -4.74024773e-01 6.67428195e-01 1.01129256e-01
1.35730982e-01 -3.72089714e-01 -1.46412504e+00 9.83534575e-01
-3.94553930e-01 8.55793357e-01 9.62757349e-01 -1.30440891e+00
2.56748587e-01 5.76393962e-01 7.38531470e-01 -6.45791516e-02
8.21503103e-02 1.06447555e-01 4.03032713e-02 -2.48504207e-01
-2.42941529e-01 -3.81542116e-01 1.41432405e-01 5.47912419e-01
-1.20318934e-01 2.16761291e-01 4.04643506e-01 6.85432136e-01
8.50082040e-01 -3.52968425e-01 3.51492614e-01 -7.34793961e-01
1.77581787e-01 -4.79945421e-01 5.15195608e-01 1.22907197e+00
-2.38129854e-01 8.47672150e-02 1.18684375e+00 3.46480995e-01
-1.27496958e+00 -1.32938230e+00 -3.78506541e-01 3.98571044e-01
-1.10828634e-02 -2.02345356e-01 -7.85682380e-01 -3.21963698e-01
2.24291250e-01 4.05767322e-01 -6.48310304e-01 2.84321845e-01
-1.40811846e-01 -8.67975950e-01 3.89741331e-01 1.71087667e-01
1.17179334e-01 -1.48828626e-01 1.65712357e-01 3.12692195e-01
-2.40524948e-01 -1.27389669e+00 -2.92416632e-01 -1.25277832e-01
-1.03619087e+00 -1.33235502e+00 -8.26730430e-01 -2.95500249e-01
6.98747575e-01 5.57856485e-02 7.39360273e-01 1.24880023e-01
-1.11039832e-01 5.60209334e-01 1.72964245e-01 1.88206419e-01
-5.54623663e-01 1.69525683e-01 2.22755164e-01 1.41171053e-01
-4.31527585e-01 -5.78776658e-01 -4.20905948e-01 3.68881464e-01
-7.28532374e-01 -2.63180465e-01 3.69915456e-01 6.77346766e-01
2.55082518e-01 2.25417569e-01 6.56950951e-01 -8.95102322e-01
9.22729731e-01 -1.08606017e+00 -1.04856277e+00 7.63494670e-02
-6.80576801e-01 1.16652567e-02 3.13355178e-01 -1.30867615e-01
-9.08824444e-01 -2.29441762e-01 2.28971079e-01 -8.55961740e-02
3.46961707e-01 8.61300528e-01 8.80617425e-02 -7.40632787e-02
1.59869909e-01 -1.67194337e-01 3.42690825e-01 -4.55325097e-01
1.40268892e-01 6.16208076e-01 -2.22792625e-02 -5.26891410e-01
7.80013382e-01 7.14469731e-01 7.73351908e-01 -1.22465372e+00
-5.78984916e-01 -3.30151737e-01 -4.35384899e-01 -5.09987593e-01
3.18949401e-01 -7.42486596e-01 -9.06244338e-01 2.87855446e-01
-1.02018034e+00 -4.35186028e-01 -1.83828101e-01 5.52230597e-01
-8.87186885e-01 6.64119244e-01 -6.70104027e-01 -1.54285908e+00
1.63554847e-01 -7.19464958e-01 4.43655163e-01 -1.15461007e-01
-1.06125809e-01 -1.58047020e+00 1.34963229e-01 1.16484202e-01
3.91722500e-01 3.57377291e-01 6.34420753e-01 -2.31291056e-01
-5.90877950e-01 -3.29091847e-01 -2.94091761e-01 2.22114190e-01
-2.79139966e-01 4.78521943e-01 -3.00910592e-01 -5.22896051e-01
-3.37977439e-01 2.42684454e-01 6.54871464e-01 1.03851378e+00
7.01871574e-01 -4.40134227e-01 -5.39052606e-01 3.96068424e-01
1.71702588e+00 -4.11776423e-01 2.36361712e-01 -2.36951604e-01
2.43124574e-01 6.63619041e-01 2.86805183e-01 7.56043971e-01
1.21584304e-01 5.62048197e-01 5.90883315e-01 2.23736361e-01
2.39801884e-01 -3.23565871e-01 3.35095525e-01 6.28716409e-01
2.23492101e-01 -7.29977787e-01 -8.26923907e-01 7.19697356e-01
-2.16251063e+00 -1.01640821e+00 -4.70770866e-01 2.23403859e+00
5.60651004e-01 2.58446574e-01 4.04537171e-01 5.07745855e-02
1.13258851e+00 5.56479301e-03 -2.91249514e-01 -2.17849135e-01
-7.97170773e-03 -1.83681116e-01 1.04578221e+00 1.00720823e+00
-8.25376332e-01 3.82786810e-01 8.18067741e+00 9.56920445e-01
-1.60452023e-01 1.45999089e-01 2.00587004e-01 -2.76864827e-01
-4.32172008e-02 2.02304900e-01 -5.74491203e-01 6.26984775e-01
1.35944748e+00 -2.22618878e-01 4.81428921e-01 3.83320063e-01
9.91411269e-01 -2.76116252e-01 -8.47597182e-01 5.55245876e-01
-5.80986559e-01 -1.19297051e+00 -3.99083823e-01 7.56732643e-01
9.70046103e-01 1.88356385e-01 -8.27983469e-02 -4.02589113e-01
5.43403924e-01 -7.57222235e-01 4.08444703e-01 7.49002814e-01
4.57491338e-01 -8.90055537e-01 8.24304163e-01 3.70507270e-01
-7.72460401e-01 -2.87362672e-02 -4.03444946e-01 -1.80754736e-01
7.72469997e-01 1.00278461e+00 -4.61211950e-01 2.27313414e-01
2.32814074e-01 8.62101197e-01 1.44964546e-01 1.49852514e+00
5.85237099e-03 1.11982763e+00 -9.71445918e-01 -1.78174078e-01
1.28981575e-01 -4.77903575e-01 1.08413303e+00 8.81603062e-01
2.04398498e-01 -2.48205855e-01 -1.04459129e-01 9.49812770e-01
-9.93591845e-02 -1.01094143e-02 -5.95522583e-01 6.77003637e-02
6.09142959e-01 8.87933135e-01 -6.99296415e-01 -8.03628936e-02
-1.46061480e-01 3.35569173e-01 1.77257076e-01 9.03291166e-01
-6.57880604e-01 -3.86519670e-01 7.00213492e-01 2.34466150e-01
5.13063014e-01 -4.74731952e-01 -6.85992837e-02 -7.41309345e-01
1.78954080e-02 -2.77966499e-01 2.08533809e-01 -3.26502204e-01
-1.28531516e+00 -2.92273104e-01 4.51171160e-01 -6.34347022e-01
-5.53663552e-01 -4.92631167e-01 -7.25391865e-01 7.18947291e-01
-1.14966607e+00 -5.30046880e-01 3.31093520e-01 2.73184180e-01
4.87774909e-02 2.08111539e-01 3.21099907e-02 2.27273807e-01
-9.70308006e-01 3.98903012e-01 1.11537826e+00 6.85570389e-02
6.03349917e-02 -1.15879250e+00 4.20973338e-02 1.07337379e+00
-1.71938524e-01 4.71959740e-01 1.29561007e+00 -5.77853262e-01
-1.23776126e+00 -6.39156282e-01 5.19099295e-01 -4.58257236e-02
1.04387581e+00 -5.06412029e-01 -4.90316868e-01 7.56897271e-01
-4.74326871e-02 3.91448848e-02 5.86046875e-01 9.87577960e-02
1.59642726e-01 -1.56831667e-02 -1.22629619e+00 4.71839428e-01
9.65519309e-01 -2.71566540e-01 2.69190073e-01 5.57088733e-01
2.28569746e-01 2.30038524e-01 -1.05645704e+00 -2.37914063e-02
5.96577168e-01 -6.73078179e-01 8.02328110e-01 -6.30760252e-01
-1.24866277e-01 9.28637162e-02 -1.78926706e-01 -1.24265087e+00
-3.95146273e-02 -1.21944380e+00 -3.20900649e-01 1.20628440e+00
4.88189638e-01 -7.54656255e-01 9.50055122e-01 2.79929042e-01
8.33758533e-01 -4.49053228e-01 -1.46532106e+00 -1.07809615e+00
2.29848027e-01 -5.42808354e-01 -8.32330137e-02 4.85012561e-01
1.39253005e-01 8.98556560e-02 -5.27798295e-01 1.64859831e-01
1.58580637e+00 -3.61559510e-01 5.13103247e-01 -1.40189147e+00
-2.06616580e-01 -3.77064139e-01 -3.60950083e-01 -1.20665669e+00
7.31264532e-01 -4.34659481e-01 8.47348869e-02 -1.48853493e+00
5.61278582e-01 -3.87610286e-01 5.75536229e-02 -3.50335628e-01
1.66265473e-01 8.42652619e-02 -6.88620731e-02 1.20418027e-01
-6.42046988e-01 6.14745915e-01 8.05913389e-01 4.03634794e-02
2.23944232e-01 6.89539194e-01 -4.07604963e-01 5.38627625e-01
7.38538802e-01 -7.93525040e-01 -5.22559345e-01 -9.35491696e-02
5.78679085e-01 5.01723170e-01 6.73517823e-01 -4.04476613e-01
1.84036016e-01 -1.11459702e-01 -2.12144136e-01 -5.14209211e-01
4.08024371e-01 -9.86502051e-01 1.83973804e-01 6.74493849e-01
-3.68455946e-01 -1.69872090e-01 -1.27750382e-01 1.31069267e+00
2.47365966e-01 -6.01331770e-01 6.76449478e-01 3.87654722e-01
4.89546843e-02 4.05301124e-01 -7.31235206e-01 3.77102196e-01
8.87156487e-01 2.80694966e-03 -4.30827774e-02 -1.32710910e+00
-1.06996047e+00 5.95451176e-01 9.44313779e-02 -2.86215395e-01
4.12047267e-01 -1.10693979e+00 -1.02102518e+00 -2.11296454e-01
-5.56960881e-01 -5.75456202e-01 7.85749927e-02 1.25855494e+00
-3.88711065e-01 3.83035392e-01 3.81002188e-01 -5.95045388e-01
-1.15854931e+00 3.91328126e-01 6.29569709e-01 -4.52356823e-02
-8.54742229e-02 3.02397400e-01 -1.17626809e-01 -7.17538297e-02
1.59914941e-01 1.22949891e-01 2.78450459e-01 4.23458777e-03
3.42820655e-03 1.05370176e+00 -5.67148626e-01 -6.40072525e-01
-2.23325402e-01 2.31866211e-01 2.12149352e-01 -5.75647771e-01
1.27481592e+00 -8.80492747e-01 -4.00234759e-01 5.05788386e-01
1.37687099e+00 -5.70541658e-02 -1.40400589e+00 -3.64658624e-01
8.98242556e-03 -1.95213109e-01 2.83345371e-01 -6.97962493e-02
-1.20451438e+00 7.33562887e-01 2.77125388e-01 8.30219209e-01
2.71857023e-01 1.72314778e-01 1.49680600e-01 2.58866042e-01
3.20034295e-01 -9.66585696e-01 -5.98248899e-01 1.90236390e-01
4.93496597e-01 -8.74480844e-01 2.81963378e-01 -6.68642044e-01
1.44601017e-01 8.21159184e-01 -4.92432006e-02 -4.58061039e-01
1.22434843e+00 1.39475673e-01 -4.17466372e-01 2.72106356e-03
-8.79553735e-01 4.16255258e-02 -1.08528754e-03 6.96826816e-01
2.05683053e-01 3.62629220e-02 -5.76568961e-01 1.13657348e-01
1.68012172e-01 -4.07309741e-01 9.83818352e-01 7.63077497e-01
-3.86634529e-01 -8.69619608e-01 -1.23212524e-01 5.10155559e-01
-7.41617918e-01 -8.86212662e-02 -2.57650614e-01 1.08843696e+00
-8.10943127e-01 1.25131333e+00 7.22838789e-02 5.26662588e-01
-6.24059699e-04 -1.89767867e-01 5.35383940e-01 -7.99787492e-02
6.55243993e-02 2.43959516e-01 8.72843564e-02 -4.21581179e-01
-5.20712972e-01 -1.06162226e+00 -6.30649030e-01 -9.28555489e-01
-7.66719341e-01 5.77915967e-01 5.85064650e-01 9.48576748e-01
9.66827869e-02 3.71994168e-01 7.55176485e-01 -4.18885291e-01
-1.04355252e+00 -8.87538671e-01 -1.06479836e+00 -1.08022988e-01
6.51849151e-01 -4.97427016e-01 -8.10794234e-01 -3.41583610e-01] | [6.891709804534912, 5.160587310791016] |
bebcf91a-89f8-4003-8a56-57cb10c26be4 | dependency-aware-prototype-learning-for-few | null | null | https://aclanthology.org/2022.coling-1.205 | https://aclanthology.org/2022.coling-1.205.pdf | Dependency-aware Prototype Learning for Few-shot Relation Classification | Few-shot relation classification aims to classify the relation type between two given entities in a sentence by training with a few labeled instances for each relation. However, most of existing models fail to distinguish multiple relations that co-exist in one sentence. This paper presents a novel dependency-aware prototype learning (DAPL) method for few-shot relation classification. Concretely, we utilize dependency trees and shortest dependency paths (SDP) as structural information to complement the contextualized representations of input sentences by using the dependency-aware embedding as attention inputs to learn attentive sentence representations. In addition, we introduce a gate controlled update mechanism to update the dependency-aware representations according to the output of each network layer. Extensive experiments on the FewRel dataset show that DAPL achieves substantially better performance than strong baselines. For reproducibility, we will release our code and data upon the publication of this paper at https://github.com/publicstaticvo/DAPL. | ['Xiaoyan Zhao', 'Min Yang', 'Tianshu Yu'] | null | null | null | null | coling-2022-10 | ['few-shot-relation-classification', 'relation-classification', 'few-shot-relation-classification'] | ['methodology', 'natural-language-processing', 'natural-language-processing'] | [ 2.77161505e-02 4.48400885e-01 -4.76311207e-01 -7.76102960e-01
-6.88302517e-01 -4.56936359e-01 6.30080879e-01 8.10004950e-01
-3.35626990e-01 7.67514884e-01 4.88035321e-01 -4.06263858e-01
6.51606778e-03 -9.37559485e-01 -4.48864281e-01 -3.18268180e-01
-1.44633606e-01 4.19864029e-01 1.84677228e-01 -4.43180203e-01
-4.77038585e-02 3.03678274e-01 -1.06240129e+00 5.80579698e-01
6.38161957e-01 8.40910316e-01 6.89319819e-02 6.70126438e-01
-1.34088084e-01 9.16761100e-01 -5.16603529e-01 -6.87403619e-01
-8.08808953e-02 -5.12296319e-01 -1.35610509e+00 -4.18944836e-01
3.35508026e-02 -1.25413924e-01 -6.22049749e-01 8.93744171e-01
4.57483560e-01 5.34732580e-01 5.86050510e-01 -9.17590261e-01
-1.05616713e+00 1.14930642e+00 -2.59040296e-01 8.95405054e-01
4.37269092e-01 -1.33326873e-01 1.59799695e+00 -1.08413804e+00
7.91306078e-01 1.17040324e+00 3.28085750e-01 4.61093277e-01
-1.21825218e+00 -3.27584416e-01 4.34432864e-01 6.88530445e-01
-1.32765520e+00 -5.34114838e-01 8.01581025e-01 -2.08753839e-01
1.73798585e+00 1.28642857e-01 3.63104612e-01 1.23773372e+00
2.59105086e-01 6.79895222e-01 4.74602282e-01 -5.19945621e-01
1.75601587e-01 -2.45048921e-03 9.57699299e-01 5.65131605e-01
2.14301363e-01 -2.65011221e-01 -4.65391964e-01 -3.22541408e-02
5.96481934e-02 -1.81557283e-01 -3.11110020e-01 3.33276577e-03
-8.48456264e-01 8.67306292e-01 7.91355669e-01 5.94311476e-01
-3.35192382e-01 -1.62026107e-01 6.18823886e-01 4.62055475e-01
8.28358948e-01 4.53818142e-01 -6.89355552e-01 3.08274198e-02
-2.10397795e-01 -1.61255524e-01 8.23268890e-01 9.43903565e-01
7.49728143e-01 -5.53036809e-01 -7.35020101e-01 9.66530621e-01
1.26765460e-01 -2.21822113e-01 5.23311317e-01 -4.13492113e-01
7.93089092e-01 8.51748705e-01 -3.04965824e-01 -8.80480647e-01
-4.52549905e-01 -2.26313025e-01 -7.18481123e-01 -4.51851398e-01
-5.76386824e-02 -2.05122769e-01 -7.03999877e-01 1.50340569e+00
3.76644284e-01 1.47331432e-01 4.08138365e-01 6.23858392e-01
1.46230924e+00 7.53298223e-01 2.96904594e-01 -3.58573496e-01
1.50665724e+00 -1.18220270e+00 -8.34584594e-01 -5.05755842e-01
1.21305704e+00 -3.49932879e-01 9.81412232e-01 -3.75516236e-01
-7.59823442e-01 -4.98622656e-01 -1.10435045e+00 -4.52153772e-01
-6.79456532e-01 -7.06764013e-02 8.21773529e-01 2.38141462e-01
-6.87176764e-01 6.89215839e-01 -8.27914178e-01 -6.43759191e-01
5.97924471e-01 6.16541579e-02 -4.03878033e-01 -4.55375575e-03
-1.79554582e+00 1.11016846e+00 6.49759710e-01 2.23978534e-02
-3.52061301e-01 -6.28222346e-01 -1.33836329e+00 3.38405699e-01
3.65986794e-01 -4.17607516e-01 1.28486240e+00 -4.19830084e-01
-1.14503336e+00 9.68719721e-01 -4.52189893e-01 -6.30508423e-01
-1.97998196e-01 -3.13230306e-01 -4.49852467e-01 1.47411272e-01
1.11626551e-01 3.49563837e-01 6.07671067e-02 -8.55141938e-01
-3.87578934e-01 -1.50540218e-01 3.80285621e-01 2.32277811e-01
-2.87815571e-01 2.52756685e-01 -2.90306658e-01 -3.55366766e-01
-2.40217596e-02 -5.04477501e-01 -1.67693734e-01 -3.77700269e-01
-7.17306852e-01 -8.26400161e-01 4.68064994e-01 -3.32827270e-01
1.39373219e+00 -2.05754590e+00 1.38513461e-01 -3.24645460e-01
7.48243108e-02 4.56312686e-01 -3.17880213e-01 5.90890944e-01
-5.62713325e-01 1.47304833e-01 -2.71506369e-01 -3.90151352e-01
-2.44094774e-01 3.98897678e-01 -2.77980626e-01 2.76294976e-01
6.31621003e-01 1.34301162e+00 -1.34592533e+00 -4.84495282e-01
-1.11460779e-02 4.07532126e-01 -1.22991815e-01 2.87184149e-01
-2.84234136e-02 5.13043888e-02 -2.94459730e-01 4.39333707e-01
2.66692698e-01 -3.75112504e-01 5.06395280e-01 -3.08409840e-01
2.88656592e-01 1.00112414e+00 -5.54622650e-01 1.67358494e+00
-5.13883770e-01 6.52013183e-01 -5.38059175e-01 -1.06056559e+00
9.02766764e-01 6.03141189e-01 3.33193503e-02 -4.92582530e-01
4.84041512e-01 -1.64172098e-01 2.23719269e-01 -4.43918318e-01
4.03258562e-01 -1.03185341e-01 -2.16497481e-01 3.61217648e-01
5.16256630e-01 1.68539315e-01 5.10151923e-01 5.38144290e-01
1.40176237e+00 -8.53565559e-02 8.97875726e-01 1.29854187e-01
4.79418576e-01 -1.92439735e-01 7.23415196e-01 5.50230503e-01
-3.09820414e-01 4.33620483e-01 7.71067977e-01 -4.49568957e-01
-5.67505717e-01 -1.02107596e+00 -2.36726880e-01 1.06533825e+00
6.04077205e-02 -6.99175537e-01 -1.79148570e-01 -1.01826787e+00
-2.39993080e-01 1.05231810e+00 -8.89853001e-01 -3.74144465e-01
-4.92933035e-01 -6.78039551e-01 2.24813506e-01 7.72944570e-01
3.23378712e-01 -1.27416539e+00 -2.91256785e-01 1.43651247e-01
-2.05545858e-01 -1.13038015e+00 -2.65901238e-01 7.52777696e-01
-6.53804183e-01 -1.04234660e+00 -1.57029733e-01 -1.06138349e+00
6.61246777e-01 1.64236635e-01 1.28620052e+00 1.46664858e-01
-2.16615677e-01 -1.18200906e-01 -7.76039183e-01 -4.81206626e-02
-2.28798613e-01 3.61199319e-01 -1.27854124e-01 -2.65194744e-01
8.59941602e-01 -6.44762039e-01 -3.24487448e-01 -3.43686819e-01
-3.27020437e-01 -1.01512536e-01 3.78964633e-01 9.23613429e-01
5.25682509e-01 -1.40655652e-01 5.75542927e-01 -1.44521189e+00
9.31183934e-01 -8.47614348e-01 -2.58238800e-02 4.24208194e-01
-5.92537999e-01 1.41433775e-01 6.06063902e-01 -2.01165855e-01
-1.10345256e+00 -1.48307979e-01 -1.97564125e-01 -1.77942798e-01
-3.14000934e-01 8.97291303e-01 -1.18369401e-01 5.92736542e-01
7.11661577e-01 -1.85226068e-01 -5.67788899e-01 -3.08165640e-01
6.82782412e-01 6.18705809e-01 3.90242279e-01 -2.91381687e-01
5.32731771e-01 1.44097814e-02 -3.20595801e-01 -4.19536352e-01
-1.58750999e+00 -5.17680824e-01 -1.08958530e+00 1.19601689e-01
9.08823729e-01 -7.94873714e-01 -3.11778098e-01 -7.54497722e-02
-1.44913685e+00 -3.41345966e-01 -4.48546737e-01 4.46537107e-01
-1.52777553e-01 5.74210919e-02 -9.73716199e-01 -6.38733208e-01
-5.48000157e-01 -5.11505187e-01 5.95277309e-01 3.71831089e-01
-4.68362063e-01 -1.15744317e+00 3.39113981e-01 2.05835387e-01
-4.81785983e-02 2.65119970e-02 9.70810413e-01 -1.13898623e+00
-7.65969828e-02 -4.57716197e-01 -1.60522416e-01 1.12309702e-01
4.21318769e-01 -7.22489785e-03 -1.07171643e+00 -4.33431678e-02
-1.65937990e-01 -5.17437398e-01 1.16313517e+00 1.46821380e-01
7.68691421e-01 -2.85144448e-01 -6.41655147e-01 3.85612190e-01
1.11839950e+00 1.50630504e-01 3.89147311e-01 1.18171506e-01
7.24301219e-01 6.15209281e-01 8.55828285e-01 2.59721965e-01
5.59469044e-01 4.68697339e-01 -6.53589610e-03 2.25209877e-01
-2.33228445e-01 -1.70055792e-01 8.36180747e-02 9.32734430e-01
-2.92767622e-02 -4.37711447e-01 -9.82017219e-01 6.32747948e-01
-1.91755462e+00 -1.15655422e+00 -1.16678916e-01 1.85745525e+00
1.08846700e+00 4.63269532e-01 -3.96796316e-01 -2.87481230e-02
7.99407423e-01 4.64125842e-01 -3.85815471e-01 -7.42254138e-01
-1.03585616e-01 3.75091404e-01 1.22298293e-01 7.96807408e-01
-1.36936355e+00 1.18441224e+00 5.47024965e+00 3.97363514e-01
-7.90717781e-01 4.53480005e-01 7.78702199e-01 -1.99759677e-01
-2.48055354e-01 1.34705588e-01 -8.12594831e-01 8.99829119e-02
1.33748341e+00 -5.31349659e-01 -1.61594570e-01 8.00122619e-01
-4.84574735e-02 -3.95072140e-02 -1.30113268e+00 5.88255227e-01
6.89401403e-02 -1.42151165e+00 -2.06818968e-01 -3.57192576e-01
5.65454662e-01 3.16697299e-01 -3.88875991e-01 6.93180859e-01
4.86094654e-01 -7.72491276e-01 1.41538575e-01 2.95420885e-01
5.56671381e-01 -5.93064785e-01 1.00417066e+00 -3.17928009e-03
-1.40858710e+00 4.63868491e-02 -5.32035232e-01 -4.88734812e-01
3.12477231e-01 5.07399023e-01 -7.94631779e-01 7.56310582e-01
4.27275002e-01 1.07522142e+00 -6.22502923e-01 6.70603693e-01
-9.26308274e-01 6.30185664e-01 6.82386309e-02 -2.48916194e-01
1.87089387e-02 2.34264001e-01 3.08413327e-01 1.37215257e+00
-2.20867544e-01 4.81809616e-01 2.44726837e-02 7.00292468e-01
-3.91751975e-01 2.69935846e-01 -7.00272799e-01 -9.56284106e-02
7.80724108e-01 1.44360566e+00 -7.00914681e-01 -4.37144727e-01
-6.81406856e-01 1.02833426e+00 1.08981204e+00 3.15574676e-01
-6.38493538e-01 -6.24145925e-01 7.11659968e-01 -3.83251041e-01
4.60540563e-01 -1.37452018e-02 -2.80406088e-01 -1.29266608e+00
-1.00083100e-02 -4.12664115e-02 6.80873334e-01 -6.12016201e-01
-1.57404721e+00 9.94840264e-01 -7.42583647e-02 -9.70148921e-01
-2.53297925e-01 -3.70387405e-01 -9.90497291e-01 9.37850118e-01
-1.47986352e+00 -9.53516901e-01 -6.92163929e-02 8.09661895e-02
7.35543728e-01 1.96971800e-02 1.27606308e+00 7.91456774e-02
-9.20060754e-01 6.40374541e-01 -2.94097304e-01 6.43817723e-01
6.13434136e-01 -1.23445070e+00 8.04249883e-01 9.59734797e-01
4.76537824e-01 7.04323769e-01 5.30567646e-01 -6.40621483e-01
-8.21561515e-01 -1.18269956e+00 1.61050224e+00 -4.65118319e-01
9.12025928e-01 -4.28788722e-01 -1.23871195e+00 1.02841127e+00
5.19161046e-01 4.71741915e-01 1.03092730e+00 7.63468206e-01
-6.62283123e-01 -7.66370669e-02 -8.66283476e-01 4.11999762e-01
1.08783615e+00 -7.37967968e-01 -1.09992075e+00 5.09589374e-01
1.08260942e+00 -2.53866911e-01 -9.38247859e-01 3.66514862e-01
4.94384952e-02 -5.36827922e-01 7.23755062e-01 -9.16263282e-01
6.64424658e-01 -2.06827335e-02 -1.03747338e-01 -1.39686334e+00
-6.50571167e-01 -1.09264061e-01 -6.28652453e-01 1.51972401e+00
9.45592046e-01 -5.64895093e-01 4.64087814e-01 6.23895407e-01
-1.59123659e-01 -1.24188912e+00 -9.21199739e-01 -8.07640374e-01
1.42516270e-01 -2.83424675e-01 3.47112298e-01 1.16934729e+00
6.96949542e-01 1.15098560e+00 4.07815985e-02 2.75708973e-01
1.82108626e-01 2.89923996e-01 2.46983707e-01 -1.02647233e+00
-3.57047737e-01 -2.66166836e-01 -5.29787540e-01 -6.82822704e-01
5.88071465e-01 -1.34897506e+00 2.05618918e-01 -1.95163441e+00
4.54127371e-01 -1.98629737e-01 -6.74807787e-01 8.16664815e-01
-7.31023610e-01 1.06784746e-01 -2.35259756e-02 7.39619955e-02
-7.80676782e-01 6.74566686e-01 8.70960474e-01 -3.73954803e-01
-2.42145032e-01 -1.39632851e-01 -8.14770937e-01 3.04901093e-01
1.08673751e+00 -5.85582137e-01 -6.08029366e-01 -3.25130194e-01
8.40711966e-02 -9.07143578e-02 -2.13247798e-02 -7.10858703e-01
2.03642085e-01 -1.29322615e-03 2.56307840e-01 -4.97373790e-01
3.68441731e-01 -2.71993965e-01 -4.18275118e-01 3.42712820e-01
-7.78510511e-01 -3.07222962e-01 1.13596646e-02 5.36862850e-01
-2.86289483e-01 -4.18479502e-01 5.79696119e-01 -6.20458799e-04
-8.89121950e-01 3.52073520e-01 -2.81259626e-01 2.70478368e-01
1.04817438e+00 3.75960529e-01 -7.03048170e-01 -1.72096685e-01
-9.22909498e-01 3.67918283e-01 -1.27843469e-02 6.79932654e-01
7.16817975e-01 -1.22039163e+00 -7.44443476e-01 -9.00741592e-02
5.21919131e-01 6.27807574e-03 2.08072931e-01 6.29501522e-01
-1.51449993e-01 4.12297845e-01 1.46185383e-01 6.35564849e-02
-1.46271431e+00 8.76403511e-01 2.21387982e-01 -3.67048472e-01
-7.62997389e-01 1.35049093e+00 -5.91798797e-02 -3.49537194e-01
6.25107214e-02 -3.57213944e-01 -5.43456078e-01 2.68896729e-01
8.28935802e-01 -1.70013502e-01 8.88258666e-02 -6.33783519e-01
-6.96408808e-01 -5.83734550e-02 -5.61009765e-01 1.75174981e-01
1.38242078e+00 -9.57118422e-02 -1.86839402e-01 1.00878894e+00
1.28808630e+00 -3.75017881e-01 -9.09093738e-01 -5.42915940e-01
3.32514554e-01 -2.18997061e-01 -6.12782799e-02 -6.12103343e-01
-9.03110325e-01 7.38699913e-01 9.93866250e-02 3.59548599e-01
8.43612254e-01 7.11684585e-01 6.17408812e-01 6.39281094e-01
7.54164085e-02 -7.94484973e-01 -1.24005064e-01 9.08098936e-01
7.93126881e-01 -1.31139207e+00 6.27721921e-02 -7.26141155e-01
-6.90499306e-01 9.86773729e-01 8.34420979e-01 -7.69938827e-02
8.83306861e-01 2.65481353e-01 1.22678854e-01 -2.47700244e-01
-1.39510608e+00 -4.77755487e-01 1.83551043e-01 5.45892715e-01
1.06020296e+00 1.00768186e-01 -6.43522203e-01 7.81812727e-01
-4.02282290e-02 -3.82193506e-01 4.43302959e-01 1.04225326e+00
-3.33792716e-01 -1.28367662e+00 4.28248703e-01 5.99797487e-01
-1.73217699e-01 -6.33775830e-01 -3.88171792e-01 2.90765613e-01
-1.68283194e-01 1.15815330e+00 2.71507293e-01 -4.84885633e-01
3.62325370e-01 4.12047029e-01 3.38606596e-01 -1.33484662e+00
-5.75415492e-01 -5.18602252e-01 4.96973157e-01 -4.37030494e-01
-3.09266627e-01 -5.84525287e-01 -1.54180956e+00 -1.13218583e-01
-3.10824007e-01 2.63127863e-01 2.14524679e-02 1.05246902e+00
4.43545997e-01 7.94954598e-01 5.28397918e-01 -5.28520465e-01
-1.72543570e-01 -1.30311739e+00 -4.79474545e-01 4.58599597e-01
1.21093571e-01 -6.65953338e-01 -2.72545487e-01 -1.05684251e-01] | [9.355977058410645, 8.668853759765625] |
5c8bb8e6-07ee-42f3-beaf-f7291510dd5c | spatio-temporal-encoding-improves | 2010.14184 | null | https://arxiv.org/abs/2010.14184v2 | https://arxiv.org/pdf/2010.14184v2.pdf | Spatio-temporal encoding improves neuromorphic tactile texture classification | With the increase in interest in deployment of robots in unstructured environments to work alongside humans, the development of human-like sense of touch for robots becomes important. In this work, we implement a multi-channel neuromorphic tactile system that encodes contact events as discrete spike events that mimic the behavior of slow adapting mechanoreceptors. We study the impact of information pooling across artificial mechanoreceptors on classification performance of spatially non-uniform naturalistic textures. We encoded the spatio-temporal activation patterns of mechanoreceptors through gray-level co-occurrence matrix computed from time-varying mean spiking rate-based tactile response volume. We found that this approach greatly improved texture classification in comparison to use of individual mechanoreceptor response alone. In addition, the performance was also more robust to changes in sliding velocity. The importance of exploiting precise spatial and temporal correlations between sensory channels is evident from the fact that on either removal of precise temporal information or altering of spatial structure of response pattern, a significant performance drop was observed. This study thus demonstrates the superiority of population coding approaches that can exploit the precise spatio-temporal information encoded in activation patterns of mechanoreceptor populations. It, therefore, makes an advance in the direction of development of bio-inspired tactile systems required for realistic touch applications in robotics and prostheses. | ['Nitish V. Thakor', 'Nathan F. Lepora', 'Andrei Nakagawa', 'Anupam K. Gupta'] | 2020-10-27 | null | null | null | null | ['texture-classification'] | ['computer-vision'] | [ 8.76894653e-01 -4.97921467e-01 5.52325487e-01 1.13882385e-01
-2.47188255e-01 -5.55691421e-01 5.30729592e-01 2.27224872e-01
-8.69458377e-01 9.59107041e-01 -1.42556489e-01 2.99132824e-01
-3.39110076e-01 -7.33592391e-01 -7.97144711e-01 -1.16334152e+00
-3.34834129e-01 7.25130644e-03 7.04058588e-01 -2.58494645e-01
7.86860645e-01 4.15890217e-01 -2.03694320e+00 7.05718040e-01
3.54588360e-01 1.48367512e+00 6.04145765e-01 7.69144773e-01
7.04957247e-02 9.66348276e-02 -4.53559577e-01 2.70537585e-01
2.99392462e-01 -5.37568152e-01 -7.83233047e-02 -5.80454171e-01
8.76915157e-02 3.79084140e-01 -7.00197518e-02 7.84879982e-01
7.71856725e-01 -2.74939626e-01 1.01853621e+00 -7.29977965e-01
-6.98941946e-01 2.88114041e-01 -2.37573534e-01 3.22865129e-01
1.54340193e-01 2.56282181e-01 4.96975750e-01 -5.54167330e-01
7.79735506e-01 8.18591237e-01 6.61637545e-01 5.79403520e-01
-1.52063084e+00 -3.73838842e-01 -3.58800828e-01 4.04385617e-03
-1.11213362e+00 -3.03259969e-01 5.13896644e-01 -4.12786424e-01
1.34886289e+00 3.54552865e-02 9.04648304e-01 1.17241871e+00
1.19766510e+00 8.67091119e-02 1.56123412e+00 -4.08938080e-01
7.24377394e-01 -5.20447135e-01 -3.54763389e-01 2.31763005e-01
3.73615801e-01 3.67499262e-01 -9.44476247e-01 -1.75125927e-01
1.38256669e+00 -2.11399794e-01 2.24678740e-02 -2.24566117e-01
-1.17867291e+00 1.98956728e-01 4.26707298e-01 6.54248893e-01
-6.91199780e-01 7.50494301e-01 3.46998394e-01 3.04054677e-01
-1.85011193e-01 7.05953538e-01 -2.87777364e-01 -6.03632271e-01
-3.65734547e-01 2.03936711e-01 7.87215650e-01 3.05219293e-01
4.63066190e-01 -5.27119935e-02 -1.26560807e-01 8.98119271e-01
-1.65999040e-01 8.00561965e-01 6.73559606e-01 -1.09730136e+00
-1.14410430e-01 5.29914975e-01 3.25775743e-02 -8.25570762e-01
-2.98802853e-01 2.89302319e-01 -8.01649153e-01 8.53308678e-01
6.26291454e-01 3.80632281e-02 -1.10067308e+00 1.60224438e+00
-3.82849097e-01 -1.86359331e-01 -2.21674833e-02 8.69390428e-01
-8.18486661e-02 2.87030965e-01 2.58808494e-01 4.34544794e-02
1.22115469e+00 2.87798554e-01 -6.02749407e-01 -1.75323367e-01
-4.46081609e-02 -5.19649625e-01 7.20461309e-01 2.72839636e-01
-9.07254219e-01 -2.88602769e-01 -1.05571353e+00 5.04607141e-01
-6.13667369e-01 -2.10721195e-01 7.28430092e-01 4.17149276e-01
-1.03553605e+00 8.29067647e-01 -9.13737118e-01 -7.20083296e-01
4.56922412e-01 6.47313595e-01 -4.51510280e-01 3.50734353e-01
-8.89241815e-01 8.52123260e-01 1.20614149e-01 -1.01256356e-01
-2.53744304e-01 -2.78393745e-01 -2.01154366e-01 -1.22442156e-01
-4.18841690e-01 -3.41932744e-01 9.17466521e-01 -8.71461213e-01
-1.60701668e+00 8.77424657e-01 -7.62799382e-03 -5.28999209e-01
7.18056187e-02 3.55404496e-01 2.33205408e-02 1.60126433e-01
-1.89621180e-01 7.21825659e-01 6.50502622e-01 -1.22664392e+00
-3.06050092e-01 -5.53635240e-01 -8.44229996e-01 -1.02149367e-01
-1.13534465e-01 -2.72725761e-01 2.92592883e-01 -5.95684826e-01
1.74814865e-01 -9.08631980e-01 -1.79998428e-01 2.97437578e-01
2.12024286e-01 8.70676190e-02 5.89303911e-01 -9.20777470e-02
7.77712584e-01 -2.43235540e+00 2.52716690e-01 4.85916078e-01
-1.58412024e-01 1.20867699e-01 -2.72331119e-01 9.68743920e-01
4.28311825e-01 -1.20571256e-01 -5.54627895e-01 5.32489836e-01
-2.96623051e-01 3.51511002e-01 -2.82215267e-01 2.73041725e-01
5.19229174e-01 9.58002388e-01 -7.13493586e-01 -1.27661154e-01
2.37790287e-01 5.57115734e-01 -1.76105291e-01 -3.06496471e-01
-7.03540817e-02 4.48883533e-01 -4.46881086e-01 8.11756968e-01
4.75816905e-01 1.49676010e-01 9.20341238e-02 3.19803990e-02
-4.71778691e-01 1.01416774e-01 -7.78980613e-01 1.61040092e+00
-3.44632357e-01 6.32867396e-01 9.68211070e-02 -7.74683654e-01
1.27996731e+00 2.63687432e-01 6.08843803e-01 -1.52186954e+00
2.71414846e-01 7.85458148e-01 4.75833595e-01 -2.98708975e-01
1.01240993e-01 -3.95721197e-01 -2.22565591e-01 2.25827262e-01
-6.92134548e-04 -4.09916073e-01 -7.39220083e-02 -4.83926266e-01
1.57599854e+00 1.53611191e-02 -2.00394522e-02 -4.77308899e-01
-2.75730819e-01 -1.33061722e-01 8.91546309e-02 1.03600025e+00
-2.48740166e-01 6.55218482e-01 1.57345533e-01 -1.80031031e-01
-1.11669350e+00 -1.66363657e+00 -5.06229639e-01 9.80369687e-01
3.68654490e-01 3.46867412e-01 -3.98319870e-01 5.16563296e-01
4.50069427e-01 7.11381957e-02 -1.02068114e+00 -2.90789753e-01
-5.55066407e-01 -7.56858110e-01 9.39720690e-01 5.62526643e-01
3.66638184e-01 -1.75023997e+00 -1.68921936e+00 7.17639804e-01
3.59705448e-01 -8.34766984e-01 3.87801141e-01 1.08347571e+00
-1.01009774e+00 -4.84166235e-01 -1.03604984e+00 -8.59256804e-01
3.03223640e-01 -2.03639239e-01 3.43855113e-01 -3.47899169e-01
-1.00776207e+00 5.12536645e-01 -4.66358900e-01 -7.37604439e-01
1.50380716e-01 -1.59112662e-01 -5.11093205e-03 -3.81312788e-01
3.94240111e-01 -1.09466231e+00 -6.18232012e-01 2.77555734e-01
-9.26217735e-01 -2.67139643e-01 8.68918359e-01 9.01383817e-01
8.12312722e-01 -3.50285500e-01 7.30719507e-01 -2.46738032e-01
9.92949307e-01 -2.06578791e-01 -3.28206122e-02 -2.05437448e-02
-6.34749159e-02 4.47531551e-01 4.30959970e-01 -7.65040159e-01
-7.61176348e-01 1.76778540e-01 2.12967560e-01 4.11673933e-01
-2.56025076e-01 3.25832427e-01 6.90249562e-01 -4.49978977e-01
1.02845693e+00 2.44314402e-01 1.53448522e-01 -5.91060743e-02
-2.87334770e-01 7.90453255e-01 6.62338376e-01 -6.72676802e-01
-9.39435363e-02 7.45097220e-01 1.77345484e-01 -1.06133866e+00
4.95177180e-01 -3.39592487e-01 -4.79027450e-01 -3.05543214e-01
6.57241523e-01 -3.22435677e-01 -7.70625234e-01 1.15268159e+00
-1.06350946e+00 -8.19704890e-01 -3.44694555e-01 3.11634183e-01
-1.02345943e+00 -1.23896480e-01 -7.74505615e-01 -1.10858226e+00
-1.87835231e-01 -5.98205447e-01 1.19114637e+00 2.85820901e-01
-4.24764276e-01 -5.72523355e-01 4.68799770e-01 -6.00226104e-01
9.52917874e-01 4.41247880e-01 9.20703888e-01 1.05872497e-01
-3.00375700e-01 -3.19234103e-01 -2.07385346e-01 -1.26181349e-01
3.06169003e-01 -1.65800527e-02 -1.04262292e+00 5.00706807e-02
-3.74486119e-01 -6.87998712e-01 1.19664264e+00 6.09234273e-01
9.27764416e-01 3.01013172e-01 -4.16050524e-01 -2.46489812e-02
1.71132064e+00 5.20213902e-01 1.05784845e+00 6.17338680e-02
1.77946500e-03 8.89074624e-01 1.56500533e-01 5.20795047e-01
-3.36946100e-01 6.34533644e-01 2.15417415e-01 1.35035694e-01
-1.44156218e-01 -8.61532465e-02 1.60544887e-01 4.83483493e-01
-7.57463872e-01 -8.02358165e-02 -8.13081563e-01 7.27619767e-01
-1.68806279e+00 -9.56084967e-01 9.46710929e-02 2.40418029e+00
7.96505332e-01 1.10794887e-01 -1.77908450e-01 1.93292856e-01
5.89695752e-01 -4.07480657e-01 -7.97672033e-01 -9.60007131e-01
-5.83966494e-01 9.55783606e-01 8.18740189e-01 -1.45872841e-02
-4.94414717e-01 6.89266562e-01 6.29264784e+00 5.26011467e-01
-1.52540457e+00 -2.45129332e-01 -1.54589885e-03 -6.71954229e-02
-2.20241293e-01 -4.80382234e-01 -6.42383322e-02 7.00691223e-01
9.10522223e-01 1.77546918e-01 6.75679266e-01 1.89733624e-01
4.97850813e-02 -8.46979320e-01 -7.90317059e-01 9.44894135e-01
-4.08322662e-01 -1.23104513e+00 -1.43078893e-01 1.68653607e-01
4.90188599e-01 2.98313498e-01 2.17717141e-01 -3.54449362e-01
1.02328472e-01 -1.20345819e+00 6.87365115e-01 8.22364628e-01
9.26976383e-01 -1.80294350e-01 6.38944805e-01 -1.84608195e-02
-1.29262447e+00 -3.26129764e-01 -5.31884313e-01 -5.71868658e-01
1.24531947e-01 4.06450987e-01 -3.97083968e-01 -1.78793207e-01
1.06297719e+00 2.88674384e-01 -1.64165348e-01 1.42921329e+00
4.48302418e-01 2.78882384e-01 -7.26167381e-01 -7.07088351e-01
-3.69367152e-02 1.75157636e-01 1.10233188e-01 1.24770069e+00
5.93645036e-01 1.33976683e-01 -7.23562837e-01 7.77430475e-01
4.51400667e-01 -1.24220602e-01 -1.04911458e+00 -4.19365853e-01
6.82347775e-01 9.06148612e-01 -1.10515201e+00 1.87222600e-01
2.17700101e-04 1.18748498e+00 1.15926936e-01 5.01873732e-01
-1.66498050e-01 -6.84090614e-01 5.04792213e-01 1.35557875e-01
6.27440155e-01 -5.26331365e-01 -1.04513216e+00 -3.23565990e-01
1.89001516e-01 -2.34576628e-01 -2.75622755e-01 -6.38377845e-01
-1.33790016e+00 3.87253195e-01 -4.30991620e-01 -8.92998815e-01
-1.22706041e-01 -1.01117039e+00 -6.97348714e-01 9.33100522e-01
-8.29026461e-01 -6.59247279e-01 -1.17863134e-01 2.38287807e-01
4.73785251e-02 2.08625168e-01 1.20157027e+00 -2.46488452e-01
3.28580737e-01 1.97481871e-01 1.97260737e-01 -3.06954712e-01
6.03807151e-01 -9.57588553e-01 3.15442950e-01 1.05641484e-01
-1.82980433e-01 6.19827628e-01 6.81109905e-01 -7.34505355e-01
-1.30825877e+00 -4.17627692e-01 4.72754985e-01 -2.38869905e-01
5.74688017e-01 -6.20886564e-01 -9.41672146e-01 -2.94625610e-01
1.12424329e-01 -2.40755022e-01 6.13381028e-01 -3.54967326e-01
-1.94446668e-01 3.64577435e-02 -1.22566366e+00 5.74429691e-01
1.10151255e+00 -6.95252240e-01 -5.22350252e-01 -5.63414931e-01
-8.84799957e-02 1.74643055e-01 -7.76577234e-01 5.58260918e-01
1.38784659e+00 -9.96020675e-01 6.67679965e-01 3.41657661e-02
3.23633343e-01 -1.34523019e-01 -2.93348432e-01 -1.12344527e+00
-3.27305079e-01 -1.05367906e-01 5.29626012e-01 5.44719875e-01
4.61724073e-01 -9.07038093e-01 5.65212011e-01 3.11162829e-01
9.72873997e-03 -7.40432620e-01 -1.60794675e+00 -7.56981373e-01
1.98737934e-01 -5.12233190e-02 -3.98269594e-01 2.47630090e-01
5.27509272e-01 -2.88621157e-01 1.65238962e-01 -4.28512216e-01
3.80420059e-01 -1.53839871e-01 -1.13060348e-01 -1.33800530e+00
-2.94096053e-01 -5.51964641e-01 -8.89992595e-01 -4.17657435e-01
-5.03575623e-01 -5.66799998e-01 7.31629670e-01 -1.47773230e+00
5.89771494e-02 -4.99771416e-01 -8.79790664e-01 4.92778540e-01
2.57581085e-01 5.43959618e-01 -3.40914242e-02 2.04863027e-01
-9.22812968e-02 2.51232922e-01 1.26914322e+00 1.64288715e-01
-3.27690214e-01 -3.34336281e-01 -1.06786802e-01 1.44536585e-01
9.05066848e-01 -3.55663806e-01 1.16549362e-03 -2.72381306e-01
2.78261989e-01 -1.08214378e-01 7.96823204e-01 -1.58966506e+00
4.65324730e-01 -2.19739415e-02 5.74402630e-01 6.14178404e-02
4.87766951e-01 -6.41082168e-01 1.43599331e-01 8.94321561e-01
-4.30245131e-01 9.60334167e-02 5.16259313e-01 7.61603534e-01
-1.01793818e-01 2.06560597e-01 6.72052741e-01 -3.15611869e-01
-9.95667338e-01 -3.69249433e-01 -1.35444355e+00 -3.12956512e-01
9.30208266e-01 -9.71896410e-01 -3.27909231e-01 2.43504152e-01
-3.41575354e-01 -5.87993145e-01 8.43865871e-01 2.51140326e-01
7.41066396e-01 -1.09044850e+00 -2.07156181e-01 4.58307385e-01
3.27938080e-01 -7.58095264e-01 3.43644291e-01 6.60271525e-01
-4.47738707e-01 2.67716527e-01 -1.38221157e+00 -6.12148166e-01
-7.29688406e-01 -4.10106890e-02 3.43300074e-01 3.13145071e-01
-1.70909226e-01 1.05700684e+00 -6.63071871e-02 -3.10554518e-03
6.33251574e-03 -5.66173553e-01 9.41923037e-02 -2.63072550e-01
1.76400334e-01 1.01396389e-01 -2.14785188e-02 -1.01199634e-01
-5.02837896e-01 1.01855385e+00 3.30095977e-01 -6.45575523e-01
1.17750335e+00 1.23872928e-01 -1.01221703e-01 1.12854850e+00
6.99935019e-01 -3.58947575e-01 -1.53008938e+00 4.60079908e-01
1.93645190e-02 -8.75201523e-02 -4.44364130e-01 -1.22797620e+00
-4.03539270e-01 1.06916320e+00 9.91078436e-01 2.42698357e-01
1.16639113e+00 3.33994664e-02 7.05494344e-01 4.53036457e-01
1.18952346e+00 -1.37052548e+00 3.33853990e-01 4.87873822e-01
8.69297385e-01 -7.03864336e-01 -6.78748965e-01 -3.20425600e-01
-4.62733358e-01 1.18431091e+00 4.07419682e-01 -6.77145958e-01
4.90659237e-01 9.05455172e-01 1.23154759e-01 -2.07219675e-01
-7.96711326e-01 -2.69800127e-01 -1.57513276e-01 1.11368513e+00
4.78132486e-01 4.00584072e-01 -5.30783772e-01 2.94871151e-01
2.05027238e-01 4.57114547e-01 2.08648995e-01 1.35183179e+00
-8.88400078e-01 -1.09837115e+00 -9.43973884e-02 8.20831418e-01
-1.92391023e-01 -5.85618615e-02 -7.75604308e-01 3.81880969e-01
1.85329497e-01 5.38314939e-01 3.84048074e-01 -7.14714766e-01
3.75071764e-01 6.98556602e-02 1.02558756e+00 -4.65237945e-01
-5.07377982e-01 -1.03589252e-01 -2.05152243e-01 -4.80220675e-01
-3.56270730e-01 -5.25411069e-01 -1.71120632e+00 1.58579811e-01
3.68612409e-02 -2.64258236e-01 1.13707829e+00 6.10281944e-01
5.50274134e-01 4.96179074e-01 -6.83251768e-02 -1.32651079e+00
-3.26599568e-01 -8.51009429e-01 -9.07265663e-01 3.70372713e-01
4.39831689e-02 -8.68401170e-01 -2.38112882e-01 1.31004736e-01] | [8.285067558288574, 2.4269487857818604] |
fe2d27f4-e2fd-4c61-aa10-a1140e29e236 | adaptive-pd-control-using-deep-reinforcement | 2305.16979 | null | https://arxiv.org/abs/2305.16979v1 | https://arxiv.org/pdf/2305.16979v1.pdf | Adaptive PD Control using Deep Reinforcement Learning for Local-Remote Teleoperation with Stochastic Time Delays | Local-remote systems allow robots to execute complex tasks in hazardous environments such as space and nuclear power stations. However, establishing accurate positional mapping between local and remote devices can be difficult due to time delays that can compromise system performance and stability. Enhancing the synchronicity and stability of local-remote systems is vital for enabling robots to interact with environments at greater distances and under highly challenging network conditions, including time delays. We introduce an adaptive control method employing reinforcement learning to tackle the time-delayed control problem. By adjusting controller parameters in real-time, this adaptive controller compensates for stochastic delays and improves synchronicity between local and remote robotic manipulators. To improve the adaptive PD controller's performance, we devise a model-based reinforcement learning approach that effectively incorporates multi-step delays into the learning framework. Utilizing this proposed technique, the local-remote system's performance is stabilized for stochastic communication time-delays of up to 290ms. Our results demonstrate that the suggested model-based reinforcement learning method surpasses the Soft-Actor Critic and augmented state Soft-Actor Critic techniques. Access the code at: https://github.com/CAV-Research-Lab/Predictive-Model-Delay-Correction | ['Saber Fallah', 'Luc McCutcheon'] | 2023-05-26 | null | null | null | null | ['model-based-reinforcement-learning'] | ['reasoning'] | [-1.02461748e-01 2.93259621e-01 -1.08547360e-01 2.72233337e-01
-5.59473097e-01 -4.90580678e-01 3.38931888e-01 2.21979752e-01
-4.54550654e-01 8.96075308e-01 -6.20689511e-01 -5.01730621e-01
-6.73122108e-01 -5.99379241e-01 -6.54625535e-01 -9.86430109e-01
-1.72179386e-01 5.10169268e-01 1.99774384e-01 -5.05310416e-01
1.22274712e-01 7.26469874e-01 -9.20535207e-01 -6.66632354e-01
6.13574147e-01 6.24523222e-01 6.56033218e-01 8.31198990e-01
5.77442110e-01 5.22392154e-01 -4.87314433e-01 6.58346295e-01
4.05161530e-01 -1.45279691e-01 -3.68179739e-01 -2.34862193e-01
-8.62695277e-01 -1.95504308e-01 -5.26780486e-01 8.20722401e-01
7.03324616e-01 6.29944444e-01 5.23852408e-01 -1.42919970e+00
6.10388741e-02 4.96541411e-01 -6.76446676e-01 2.05982462e-01
3.89855236e-01 5.08274913e-01 2.28413597e-01 -2.53738195e-01
1.66695356e-01 1.26100826e+00 5.35235226e-01 3.46021026e-01
-1.02983427e+00 -6.76279485e-01 3.36504906e-01 9.66842622e-02
-1.55863810e+00 -2.30980232e-01 5.97130775e-01 -3.95774633e-01
8.79569709e-01 -2.61095643e-01 4.14994061e-01 1.00282347e+00
9.17962968e-01 2.03571990e-01 6.94212615e-01 -3.48010272e-01
2.32371613e-01 -4.06523496e-01 -4.73788261e-01 4.48741645e-01
1.92877576e-01 5.55193365e-01 1.15126289e-01 5.64281382e-02
1.24278772e+00 1.51335835e-01 -8.15158989e-03 -2.34484568e-01
-1.53922641e+00 5.23391783e-01 4.63621050e-01 1.95905969e-01
-8.52313101e-01 4.47228730e-01 4.98048633e-01 6.05647862e-01
-1.81143746e-01 6.27818882e-01 -5.12427032e-01 -1.65260360e-01
3.28253359e-02 1.64911881e-01 6.74709201e-01 1.26917458e+00
1.84515774e-01 5.64083576e-01 2.85738975e-01 5.04844427e-01
4.06841010e-01 6.16597474e-01 2.32521832e-01 -1.09278560e+00
5.13153970e-01 2.80371718e-02 4.84371483e-01 -8.86502385e-01
-8.28267932e-01 -2.34295517e-01 -1.03836489e+00 6.68863356e-01
3.72004777e-01 -7.05038488e-01 -3.36345255e-01 1.39547932e+00
5.75385034e-01 -6.48678988e-02 2.52777964e-01 8.20506155e-01
-3.91253233e-01 9.25092876e-01 -1.44047648e-01 -5.93163967e-01
9.14846003e-01 -6.93687677e-01 -9.01799619e-01 -8.96404162e-02
2.62049705e-01 -7.38749802e-01 6.23894036e-01 3.99973124e-01
-9.91525054e-01 -5.35559177e-01 -1.21081853e+00 7.90966868e-01
5.50568290e-02 6.02562651e-02 -1.37093410e-01 -2.43177544e-02
-6.39889777e-01 8.12551916e-01 -1.20897233e+00 -2.41482690e-01
-4.13304895e-01 7.36410975e-01 -4.74890359e-02 4.86403912e-01
-1.34654617e+00 1.33782351e+00 4.24680501e-01 4.36443120e-01
-1.01396406e+00 -4.69915003e-01 -5.31564534e-01 -3.33525300e-01
8.75394583e-01 -5.24311364e-01 1.59098315e+00 -3.61639857e-01
-2.27482128e+00 -3.74500424e-01 4.01786983e-01 -2.16949090e-01
5.81678689e-01 -1.23476274e-01 -3.08988243e-01 1.83124527e-01
-1.55310854e-01 1.13065511e-01 8.78272831e-01 -1.11558425e+00
-5.20995736e-01 -2.12471783e-02 2.24765949e-02 4.72880483e-01
1.76470622e-01 -4.18051988e-01 1.02265283e-01 -4.02818441e-01
-6.90360591e-02 -1.34647620e+00 -7.56232023e-01 5.51827773e-02
-1.18087105e-01 -1.31702483e-01 7.64322042e-01 -2.03961104e-01
8.58839691e-01 -1.96688867e+00 1.47978976e-01 4.07547802e-01
-3.21528643e-01 3.19518983e-01 5.38213849e-02 9.16615427e-01
-4.44221497e-02 -4.43718016e-01 3.06152523e-01 4.41677302e-01
-1.47718325e-01 2.83820868e-01 -2.13060342e-02 6.62217677e-01
8.99739861e-02 3.40434700e-01 -9.53814328e-01 -8.05384815e-02
6.05004430e-01 3.86863470e-01 -1.31804213e-01 2.15838626e-01
-9.58369076e-02 1.03807449e+00 -9.75880265e-01 3.33101779e-01
1.62096232e-01 1.43695325e-01 3.19461107e-01 1.68177694e-01
-3.96606684e-01 -1.92023233e-01 -1.45603728e+00 1.26696801e+00
-7.86031187e-01 2.18852699e-01 7.05478549e-01 -1.02458048e+00
1.22892380e+00 7.16804802e-01 7.98503458e-01 -8.59825432e-01
7.00712025e-01 3.82727295e-01 -1.11750893e-01 -4.85398412e-01
1.88408822e-01 -3.99812572e-02 -1.28887042e-01 2.70436049e-01
-4.84956414e-01 -5.86316764e-01 -1.42518952e-01 -1.80458248e-01
1.21505451e+00 1.57177493e-01 4.19925213e-01 -2.37063840e-01
8.60949516e-01 -5.52110039e-02 6.58284903e-01 3.54080796e-01
-4.57923084e-01 -3.29744101e-01 2.15168595e-01 1.56702846e-02
-1.11635232e+00 -8.34999502e-01 4.64995712e-01 8.49540114e-01
6.14917159e-01 2.29285136e-01 -8.98240432e-02 -4.63236384e-02
2.71787256e-01 5.75843930e-01 -5.75349405e-02 -6.42559707e-01
-6.20430827e-01 -5.73288575e-02 3.33173424e-01 5.68285584e-01
-1.24042600e-01 -6.01878703e-01 -6.62375212e-01 7.77665794e-01
2.37179577e-01 -8.82056713e-01 -4.72092628e-01 3.83164316e-01
-7.30431259e-01 -1.10434949e+00 -5.26814282e-01 -7.32494473e-01
7.31407702e-01 2.72968203e-01 1.64975330e-01 -4.41151522e-02
-1.83256537e-01 4.80973244e-01 -3.11092645e-01 -3.80972862e-01
-6.89082325e-01 -2.36357898e-02 6.34398460e-01 -5.31291783e-01
-6.11010253e-01 -5.35167336e-01 -4.94907558e-01 8.45662653e-01
-5.48149467e-01 -2.59698272e-01 6.12854838e-01 9.11239684e-01
5.24661481e-01 3.64296019e-01 1.05814493e+00 -6.11259565e-02
9.06701207e-01 -4.44718331e-01 -1.28138840e+00 -4.67921905e-02
-6.80908859e-01 4.96293604e-02 9.45497990e-01 -7.91398823e-01
-1.09760213e+00 4.07787025e-01 7.13246465e-02 -3.58886659e-01
2.14942619e-01 4.39264685e-01 2.38538131e-01 -2.07922861e-01
5.25248766e-01 -1.50134951e-01 6.84318244e-01 7.76430070e-02
1.74654335e-01 7.36361265e-01 4.17528182e-01 -6.03534877e-01
9.18863118e-01 -4.43310067e-02 3.95946056e-01 -5.99080026e-01
8.41312706e-02 -5.35112083e-01 -4.70393479e-01 -4.55921024e-01
3.92570615e-01 -1.05446970e+00 -1.49445450e+00 3.99320275e-01
-1.08182418e+00 -7.00354934e-01 -1.15051351e-01 8.72198462e-01
-8.14742327e-01 1.37924790e-01 -7.18371272e-01 -1.10662270e+00
-6.31815642e-02 -1.14659333e+00 5.77814758e-01 2.74047166e-01
-2.27809936e-01 -9.33810592e-01 1.24097623e-01 -1.86835140e-01
6.44645870e-01 4.44550484e-01 5.68148851e-01 -1.18668824e-01
-3.12131554e-01 -4.01011586e-01 4.68148887e-01 1.12174280e-01
8.88408870e-02 1.22673549e-01 -2.74881512e-01 -8.79217267e-01
-2.32561514e-01 -2.00275645e-01 -1.98757201e-01 3.31762552e-01
4.91145313e-01 -1.93020582e-01 -4.97353077e-01 -2.32440636e-01
1.41759121e+00 8.66560757e-01 2.07283616e-01 6.14282429e-01
3.23333800e-01 4.44340914e-01 1.28690171e+00 9.36100543e-01
3.42686534e-01 6.06407762e-01 8.09500694e-01 2.23919213e-01
5.70279241e-01 2.97115892e-02 5.50137639e-01 8.96491826e-01
-4.40635420e-02 -1.18565038e-01 -8.26440334e-01 4.35453892e-01
-2.09905601e+00 -6.31451249e-01 -5.67191020e-02 2.29937744e+00
7.63909221e-01 2.38325164e-01 -9.69650596e-02 2.63438940e-01
1.13782334e+00 -3.59770685e-01 -7.00359702e-01 -5.51764786e-01
7.08745182e-01 -5.08127034e-01 9.39675450e-01 7.02848256e-01
-6.62533939e-01 4.98619735e-01 5.41948223e+00 6.42880201e-01
-1.36398757e+00 -2.30204254e-01 -1.07530728e-01 -3.51667516e-02
3.95768970e-01 -4.62855585e-02 -4.23297584e-01 4.07491177e-01
1.28256154e+00 -5.84403336e-01 7.47887850e-01 8.79127622e-01
9.60918963e-01 -2.77558595e-01 -9.77423549e-01 5.35782874e-01
-6.44394100e-01 -7.46317208e-01 -8.75017166e-01 -7.63826296e-02
6.25159919e-01 -1.10151492e-01 -1.30417630e-01 3.16043824e-01
4.86811012e-01 -4.02909130e-01 8.85299444e-01 4.98726010e-01
5.17044485e-01 -1.05245590e+00 7.36257434e-01 7.41455257e-01
-1.24796796e+00 -6.28171980e-01 -2.38084927e-01 -2.86495417e-01
5.91311097e-01 4.91630852e-01 -1.03050661e+00 7.03385592e-01
1.36020720e-01 3.67374748e-01 1.76651090e-01 9.37749207e-01
-3.09567302e-01 6.20814338e-02 -4.50324357e-01 -3.45177174e-01
1.44946903e-01 -1.89316049e-01 7.24883497e-01 4.09184158e-01
4.01497632e-01 1.85240656e-01 4.76940602e-01 4.21405613e-01
5.51932812e-01 -4.31470662e-01 -7.68708527e-01 6.61675781e-02
7.76501536e-01 1.05910981e+00 -5.41654408e-01 4.61636819e-02
1.64997116e-01 4.50155944e-01 -1.30137905e-01 1.98035657e-01
-1.12579525e+00 -9.68114197e-01 3.76110733e-01 -1.36836529e-01
-1.28241673e-01 -1.03166807e+00 1.32565647e-01 -2.48806745e-01
-1.93565801e-01 -5.58032274e-01 -1.14442125e-01 -7.86186576e-01
-9.55463171e-01 4.04168904e-01 4.51451801e-02 -1.77944291e+00
-6.98353529e-01 -2.31347963e-01 -4.40973967e-01 5.92796683e-01
-1.43296278e+00 -5.05815983e-01 6.69538826e-02 7.84917414e-01
5.40124118e-01 -1.20314449e-01 6.24086976e-01 3.07523966e-01
-7.35459626e-01 -7.47974068e-02 6.35516047e-01 -4.25475508e-01
9.77497637e-01 -8.54254365e-01 -1.20588727e-01 6.50266588e-01
-1.04995704e+00 4.22834992e-01 9.86033916e-01 -5.66710711e-01
-1.86479545e+00 -1.12605667e+00 2.58155644e-01 2.70824432e-01
1.00635517e+00 1.26968130e-01 -5.37302494e-01 6.41219974e-01
2.45723605e-01 2.15330464e-03 -1.15170836e-01 -4.71277446e-01
4.21515852e-01 -4.03279334e-01 -1.08818638e+00 6.13958359e-01
3.63561958e-01 -3.12580854e-01 -2.32073128e-01 3.79803896e-01
7.48842835e-01 -7.10744977e-01 -1.05024159e+00 2.55204350e-01
4.40099776e-01 -6.14776388e-02 6.77180231e-01 9.66091231e-02
-8.78013968e-02 -5.38844287e-01 2.77875245e-01 -1.84949124e+00
-4.13855940e-01 -1.14876115e+00 -4.12392393e-02 7.88926661e-01
3.21854413e-01 -1.11925161e+00 3.64372432e-01 2.55968511e-01
-2.99942702e-01 -5.02855301e-01 -1.11983216e+00 -1.28640473e+00
-1.70993075e-01 -7.60373920e-02 2.38391638e-01 5.99838018e-01
4.00597155e-01 1.89217627e-01 -2.36886829e-01 8.61134350e-01
3.82402390e-01 -4.53881174e-01 6.54910624e-01 -7.19091773e-01
-2.87965178e-01 -1.60766006e-01 -2.33654991e-01 -6.79292500e-01
1.42015994e-01 -2.88994491e-01 7.30173230e-01 -1.55794585e+00
-6.72802448e-01 -7.07360268e-01 -3.87723684e-01 4.84839648e-01
2.13591427e-01 -2.69127250e-01 9.88347307e-02 2.02576712e-01
-6.16670430e-01 7.44431555e-01 1.58327508e+00 -1.64806128e-01
-5.99530339e-01 4.42310840e-01 -1.06982671e-01 4.74881113e-01
1.13048875e+00 -5.14015079e-01 -7.26084411e-01 -4.96255547e-01
7.34414682e-02 9.73415613e-01 9.50110853e-02 -1.31183887e+00
7.20807493e-01 -5.83318889e-01 -1.43750817e-01 -2.48924911e-01
1.73805863e-01 -1.29838920e+00 4.08917427e-01 1.15821803e+00
-4.66814578e-01 4.42379832e-01 2.60066837e-01 1.06920075e+00
-6.14067726e-02 1.04626387e-01 8.42992723e-01 2.60489553e-01
-4.13165122e-01 6.70969561e-02 -1.02622747e+00 -5.05577743e-01
1.75623071e+00 -7.44298324e-02 -1.03051484e-01 -4.96712595e-01
-7.83852041e-01 8.61760676e-01 -7.57368701e-03 5.13835609e-01
4.52997237e-01 -1.09347093e+00 -1.69597164e-01 -4.82579507e-02
-3.19219142e-01 1.88287999e-02 2.62205303e-01 1.21187580e+00
-5.94232142e-01 4.14007306e-01 -5.94115078e-01 -6.04197860e-01
-1.08837223e+00 4.40830648e-01 2.29220361e-01 -5.97972795e-03
-4.21790838e-01 3.38137239e-01 -5.76032221e-01 -4.73087043e-01
2.43040413e-01 -4.62970287e-01 3.07898708e-02 -2.74001449e-01
3.34571116e-02 7.36776471e-01 -5.84211610e-02 -8.46440569e-02
-3.56211782e-01 5.76118886e-01 1.04235023e-01 -1.33580357e-01
1.39370346e+00 -5.15369117e-01 2.05722958e-01 4.28408355e-01
6.91112578e-01 -4.63662922e-01 -1.76214921e+00 -1.67808570e-02
-5.95347136e-02 -1.68375120e-01 3.92989330e-02 -5.27271152e-01
-9.33916330e-01 4.44391459e-01 5.20360529e-01 1.23014174e-01
8.14001083e-01 -6.47903204e-01 5.94079792e-01 6.30313754e-01
8.03975821e-01 -1.47630966e+00 4.94348675e-01 9.99336600e-01
9.08686340e-01 -9.23083305e-01 7.88646489e-02 -3.41442853e-01
-8.08160543e-01 1.33599532e+00 6.07573748e-01 -5.54872572e-01
6.00036144e-01 5.58892965e-01 1.23033836e-01 2.86561847e-01
-1.01194155e+00 2.49152765e-01 -4.40569311e-01 6.24607384e-01
4.20527048e-02 7.06954226e-02 -3.55277568e-01 2.34879047e-01
5.81667542e-01 -5.67607768e-02 8.83630812e-01 1.55696034e+00
-5.73161900e-01 -1.10713291e+00 -6.73666716e-01 -1.92669243e-01
-3.44705358e-02 7.55392909e-01 4.28055108e-01 9.03641105e-01
-2.69674569e-01 1.14862192e+00 -2.78938487e-02 -3.18252087e-01
6.89899623e-01 -4.09418732e-01 3.46889913e-01 -3.68718803e-01
-4.71806079e-01 4.15780216e-01 7.16366842e-02 -7.38044560e-01
-1.83410257e-01 -3.40486556e-01 -1.83428288e+00 -4.46839005e-01
-6.22315645e-01 9.87205207e-02 1.03542686e+00 7.36538529e-01
3.72019917e-01 1.07935882e+00 1.10538888e+00 -1.33331704e+00
-1.32373142e+00 -6.25020385e-01 -6.49536729e-01 -5.14719069e-01
4.96705085e-01 -8.54706049e-01 -2.20078677e-01 -2.60239840e-01] | [4.785290241241455, 1.7817612886428833] |
883a565f-1452-4cc4-aae7-9f9e8f5459e4 | learning-environment-aware-control-barrier | 2303.04313 | null | https://arxiv.org/abs/2303.04313v1 | https://arxiv.org/pdf/2303.04313v1.pdf | Learning Environment-Aware Control Barrier Functions for Safe and Feasible Multi-Robot Navigation | Control Barrier Functions (CBFs) have been applied to provide safety guarantees for robot navigation. Traditional approaches consider fixed CBFs during navigation and hand-tune the underlying parameters apriori. Such approaches are inefficient and vulnerable to changes in the environment. The goal of this paper is to learn CBFs for multi-robot navigation based on what robots perceive about their environment. In order to guarantee the feasibility of the navigation task, while ensuring robot safety, we pursue a trade-off between conservativeness and aggressiveness in robot behavior by defining dynamic environment-aware CBF constraints. Since the explicit relationship between CBF constraints and navigation performance is challenging to model, we leverage reinforcement learning to learn time-varying CBFs in a model-free manner. We parameterize the CBF policy with graph neural networks (GNNs), and design GNNs that are translation invariant and permutation equivariant, to synthesize decentralized policies that generalize across environments. The proposed approach maintains safety guarantees (due to the underlying CBFs), while optimizing navigation performance (due to the reward-based learning). We perform simulations that compare the proposed approach with fixed CBFs tuned by exhaustive grid-search. The results show that environment-aware CBFs are capable of adapting to robot movements and obstacle changes, yielding improved navigation performance and robust generalization. | ['Amanda Prorok', 'Guang Yang', 'Zhan Gao'] | 2023-03-08 | null | null | null | null | ['robot-navigation'] | ['robots'] | [-2.71196775e-02 1.04241572e-01 -2.92885602e-01 -1.31772414e-01
-5.28079808e-01 -7.94521153e-01 4.10715789e-01 1.47224665e-01
-7.80869484e-01 7.47274697e-01 -1.63572311e-01 -5.21876097e-01
-6.46274030e-01 -8.54327321e-01 -9.30267513e-01 -8.84188235e-01
-6.52645767e-01 1.64318159e-01 3.72868448e-01 -8.49586189e-01
5.55904746e-01 6.94164038e-01 -1.43272972e+00 -5.35419226e-01
1.09643853e+00 7.07027853e-01 4.64085281e-01 8.00077975e-01
4.94630605e-01 3.48819226e-01 -3.07425380e-01 2.32247338e-01
6.77621067e-01 -1.76667139e-01 -5.71675599e-01 -2.60065079e-01
-1.82586893e-01 -5.27568758e-02 -2.17935160e-01 1.16589427e+00
2.79258579e-01 6.98528409e-01 6.58517718e-01 -1.58458066e+00
-3.02082092e-01 3.52859408e-01 -4.48014021e-01 -6.52917400e-02
-1.66672453e-01 3.13565135e-01 7.56025851e-01 3.99663821e-02
3.78132701e-01 1.31415737e+00 4.87848282e-01 7.02804148e-01
-1.29916370e+00 -6.01334274e-01 6.43754065e-01 -2.20090132e-02
-1.32605457e+00 -1.72624379e-01 3.85410190e-01 -2.87834555e-01
8.86133969e-01 7.15205595e-02 3.73525560e-01 1.07456899e+00
9.39502358e-01 7.66291469e-02 7.65494406e-01 -1.41315833e-01
7.00336516e-01 -8.74926075e-02 -2.79586136e-01 7.00876713e-01
5.81416190e-01 3.89116138e-01 -6.95521906e-02 -1.58794299e-01
6.85218215e-01 -2.42811903e-01 -2.39204064e-01 -1.01422775e+00
-9.40985262e-01 8.95202160e-01 6.65329576e-01 -1.91885874e-01
-3.97569448e-01 4.91515517e-01 4.58728909e-01 5.22626519e-01
-3.29227716e-01 7.54425108e-01 -4.93949920e-01 -2.96828281e-02
-5.94170354e-02 5.80007911e-01 7.60518014e-01 9.61945534e-01
8.32052886e-01 3.20461512e-01 1.85099512e-01 5.74412227e-01
2.39033863e-01 6.42610908e-01 5.50958276e-01 -1.25886881e+00
6.09468341e-01 3.73769403e-01 3.71024638e-01 -1.25314009e+00
-7.92774677e-01 -3.49718869e-01 -6.01962388e-01 7.66785502e-01
2.32459813e-01 -3.14028829e-01 -8.82395685e-01 2.29147530e+00
3.64769846e-01 -2.98913658e-01 4.04953629e-01 7.79643238e-01
-4.02263194e-01 6.16870880e-01 3.86617966e-02 -1.72319785e-02
8.70288610e-01 -5.92383862e-01 -5.01345515e-01 -4.72259283e-01
9.61388230e-01 2.57507395e-02 1.35137045e+00 3.65743130e-01
-5.46481669e-01 -9.89900082e-02 -1.38357413e+00 5.44063032e-01
-2.87537754e-01 -3.08201402e-01 1.34400412e-01 7.02909768e-01
-1.19380701e+00 4.94730055e-01 -1.03558791e+00 -6.67941570e-01
-1.27284378e-01 8.33140254e-01 -1.86150879e-01 3.47916335e-01
-1.21260917e+00 1.20348144e+00 5.72388172e-01 1.39970854e-01
-1.19833326e+00 -2.59939849e-01 -8.43583941e-01 -1.02296650e-01
7.79494226e-01 -5.58870196e-01 1.23581076e+00 -6.23045027e-01
-1.85936379e+00 4.87772562e-02 3.59973729e-01 -6.14000916e-01
4.67968762e-01 -5.58268577e-02 -8.84266570e-02 -9.43585262e-02
9.75458920e-02 6.24623716e-01 6.81228578e-01 -1.49628007e+00
-8.79500329e-01 -1.91873774e-01 5.04553735e-01 5.73606014e-01
-5.16945839e-01 -6.15798116e-01 -4.81489114e-02 -6.97102323e-02
-5.49465716e-02 -1.30315089e+00 -9.98549640e-01 -2.30890766e-01
-2.05597237e-01 2.19658419e-01 7.72918642e-01 -8.73380899e-02
9.41273868e-01 -1.86503506e+00 3.16783994e-01 4.91601169e-01
-3.35906833e-01 -1.32754833e-01 -4.23958600e-01 5.28418422e-01
5.15729368e-01 -3.41446558e-03 -2.73316652e-01 -3.07547767e-02
-2.15851106e-02 7.36803532e-01 -3.59802902e-01 6.42369509e-01
-5.94550297e-02 4.48883474e-01 -9.33021307e-01 -4.28812578e-02
4.01632376e-02 1.41402140e-01 -1.01452005e+00 3.56810950e-02
-3.64009589e-01 6.12011194e-01 -7.66561568e-01 2.28146136e-01
3.40293735e-01 2.51662105e-01 4.25871491e-01 1.70147255e-01
-4.50504780e-01 -1.75153822e-01 -1.22184145e+00 1.26301467e+00
-8.38769436e-01 -2.79470434e-04 4.38404351e-01 -1.14515078e+00
1.05556142e+00 -1.52019948e-01 3.41885120e-01 -8.56445074e-01
3.27329099e-01 4.13826972e-01 5.46661541e-02 -3.06320965e-01
7.19268322e-01 8.68756250e-02 -6.53953433e-01 5.62805235e-02
-3.25621396e-01 -2.72967100e-01 2.69835852e-02 -7.34753758e-02
1.11015391e+00 1.37718737e-01 4.20142710e-01 -7.78328121e-01
7.76570439e-01 2.08969533e-01 8.08742464e-01 8.94053817e-01
-3.95581782e-01 -1.33681446e-01 6.44276500e-01 -1.72870696e-01
-9.82227683e-01 -8.67996871e-01 3.16210747e-01 1.16158164e+00
8.26007128e-01 1.04880996e-01 -7.12661564e-01 -4.25025046e-01
1.06785491e-01 7.12495863e-01 -7.24363804e-01 -5.67182302e-01
-9.99547839e-01 -5.63117862e-01 4.09599930e-01 3.75184506e-01
4.57897246e-01 -7.31078744e-01 -1.38589442e+00 3.31238449e-01
1.13704473e-01 -8.27406406e-01 -5.61927438e-01 6.89938486e-01
-6.01574481e-01 -1.00649500e+00 -2.85929918e-01 -6.85376287e-01
7.78894424e-01 2.59323239e-01 2.33295545e-01 -4.42846864e-02
6.44049495e-02 5.51992238e-01 -1.68862373e-01 1.96920950e-02
-3.77630889e-01 2.78227031e-01 5.58597684e-01 -2.13465288e-01
-6.08397484e-01 -6.70950532e-01 -5.49144626e-01 6.08021915e-01
-9.54046547e-01 -3.91810209e-01 6.00124657e-01 8.66098285e-01
5.38700104e-01 2.79898047e-01 6.92412198e-01 -4.77614492e-01
9.24191415e-01 -5.72319508e-01 -1.33843648e+00 1.20550856e-01
-9.05155957e-01 5.07220268e-01 1.05742514e+00 -5.96268535e-01
-9.25330579e-01 1.62856624e-01 1.41711816e-01 -5.37160113e-02
1.43900290e-01 3.81952792e-01 -3.19842696e-01 -4.92320567e-01
6.83403015e-01 -6.09833980e-03 1.46532491e-01 9.67023820e-02
5.34871280e-01 3.31945479e-01 6.79049671e-01 -9.97624278e-01
7.26656616e-01 3.75932842e-01 4.61687237e-01 -5.78201830e-01
-2.01220021e-01 -1.63194805e-01 -3.32641095e-01 -2.90873438e-01
6.91314876e-01 -5.75561404e-01 -1.23747540e+00 1.72920227e-01
-7.64309406e-01 -7.65785396e-01 3.43208797e-02 2.28275225e-01
-1.18223333e+00 1.47355556e-01 -2.26754263e-01 -1.17872882e+00
-1.34005234e-01 -1.29145360e+00 6.04173839e-01 3.28172326e-01
3.99891064e-02 -8.17677319e-01 2.16326624e-01 -3.26732576e-01
7.26113975e-01 5.07659674e-01 9.23020780e-01 -3.69729221e-01
-6.71768486e-01 -7.73670094e-04 1.64720744e-01 -6.13282174e-02
6.17437921e-02 -2.91980386e-01 -3.85361016e-01 -6.95211887e-01
-7.24703372e-02 -2.06141800e-01 3.40728343e-01 3.42249542e-01
7.40211546e-01 -7.42822766e-01 -5.13375998e-01 5.47955036e-01
1.76173770e+00 7.80140638e-01 2.55886912e-01 1.18262446e+00
4.64144945e-01 7.99434900e-01 8.21713805e-01 4.83094573e-01
7.05679119e-01 7.03560114e-01 1.10871792e+00 5.46225667e-01
5.63971400e-01 -2.51277059e-01 6.13240719e-01 2.81247765e-01
2.25647047e-01 -3.44458699e-01 -9.73314881e-01 5.79118609e-01
-2.24017549e+00 -5.00167251e-01 2.55625784e-01 2.35971737e+00
4.05894399e-01 3.16233903e-01 2.43421823e-01 -2.31546685e-01
5.60923755e-01 -2.09039912e-01 -8.91874313e-01 -8.25338602e-01
1.83194026e-01 -4.59562808e-01 1.29117346e+00 8.63560736e-01
-8.77026439e-01 8.37521017e-01 5.86191940e+00 4.01891053e-01
-1.19869483e+00 -2.63683319e-01 2.64755845e-01 1.42072976e-01
-2.66361386e-01 7.59340674e-02 -8.78315926e-01 8.91721174e-02
1.17676091e+00 -3.79469305e-01 6.03400290e-01 1.03813386e+00
5.60920179e-01 -1.33607417e-01 -8.33274066e-01 2.84173071e-01
-3.33464026e-01 -8.56257617e-01 -2.22959161e-01 1.33631453e-01
6.17868006e-01 1.08546680e-02 2.25517660e-01 5.20476103e-01
8.48573625e-01 -6.77583039e-01 9.35400546e-01 2.39325225e-01
2.84855217e-01 -1.11262488e+00 5.89724064e-01 5.55539191e-01
-1.33819282e+00 -7.53324568e-01 -1.68795884e-01 -4.62167114e-02
2.77316868e-01 -2.08919212e-01 -6.93033218e-01 5.75967908e-01
6.56894147e-01 1.57161027e-01 -2.06135109e-01 1.06193757e+00
-4.74408269e-02 -6.43181130e-02 -3.88099164e-01 -5.16296327e-01
6.47651196e-01 -3.72531950e-01 6.19274795e-01 8.53437185e-01
3.83212447e-01 -1.26641929e-01 5.79426587e-01 6.20498538e-01
4.56646293e-01 -3.76619361e-02 -7.84709454e-01 3.13446134e-01
5.16357183e-01 9.49974418e-01 -8.38260353e-01 2.52876461e-01
2.44579092e-02 6.02741838e-01 3.81631643e-01 4.90105748e-01
-8.82645905e-01 -6.00786984e-01 1.05664980e+00 -1.83544114e-01
3.42669159e-01 -7.06763625e-01 -3.87426078e-01 -4.22165662e-01
-2.06170589e-01 -7.84794390e-01 2.91526467e-01 -1.65174380e-01
-8.98813188e-01 6.85651720e-01 1.57508180e-01 -1.32914114e+00
-5.48175991e-01 -7.05784023e-01 -3.82513940e-01 5.45250833e-01
-1.61628997e+00 -7.97753334e-01 -7.66827120e-03 5.39630651e-01
3.02806348e-01 9.62191299e-02 5.84980547e-01 5.44753019e-03
-5.39669693e-01 5.01598001e-01 8.79238918e-02 -5.28490663e-01
4.21428055e-01 -1.10684168e+00 2.46323943e-01 9.26272452e-01
-7.25917637e-01 8.03481817e-01 1.18659782e+00 -7.74612427e-01
-1.75186729e+00 -1.17614424e+00 5.80141172e-02 -7.38912374e-02
8.65048945e-01 -3.31792414e-01 -7.88522124e-01 4.72830415e-01
-1.52922064e-01 -1.78367093e-01 -8.90298784e-02 -1.81146964e-01
-2.94119120e-01 -1.89593226e-01 -1.25259626e+00 1.18585491e+00
9.65482414e-01 -4.68033627e-02 -4.47827578e-02 -1.08757071e-01
9.05623436e-01 -3.98326367e-01 -4.79016542e-01 6.85113192e-01
6.03375137e-01 -4.79037970e-01 8.09162319e-01 -5.03634274e-01
-2.74750859e-01 -5.97599983e-01 -4.58105803e-01 -1.58103347e+00
-2.64028370e-01 -6.96803868e-01 2.75613397e-01 6.71034396e-01
4.72646028e-01 -9.15776730e-01 7.25228667e-01 5.21614790e-01
-3.77576232e-01 -7.26045310e-01 -1.11456263e+00 -1.37983561e+00
3.09771001e-01 -9.14234743e-02 4.02824074e-01 4.55564827e-01
1.31189033e-01 -2.26650620e-03 -5.53604662e-01 9.24071431e-01
4.91095543e-01 -2.08498374e-01 9.24483478e-01 -7.96784997e-01
-2.34979630e-01 -6.33248448e-01 -2.55349636e-01 -7.95463383e-01
2.84347057e-01 -4.92598146e-01 8.84976685e-01 -1.37029910e+00
-3.31338406e-01 -7.71996498e-01 -3.19581836e-01 5.57122469e-01
1.53231099e-01 -4.60277170e-01 1.33864656e-02 5.58629073e-02
-5.39462328e-01 7.40889430e-01 1.01727533e+00 2.02103355e-03
-6.18440270e-01 4.41367850e-02 -5.92905045e-01 4.60101336e-01
1.13569295e+00 -3.51401538e-01 -9.27534521e-01 -4.37524498e-01
2.50169069e-01 1.67483613e-01 7.66689926e-02 -1.10601568e+00
4.29243743e-01 -6.84267700e-01 -4.05470759e-01 -2.22886816e-01
2.45153368e-01 -9.43812549e-01 5.02163433e-02 1.07815111e+00
-4.46418583e-01 5.51251888e-01 3.75864476e-01 1.07121038e+00
1.74282104e-01 -1.82889640e-01 8.93778861e-01 1.39223918e-01
-8.47994745e-01 2.44337142e-01 -8.77218068e-01 -1.46536082e-01
1.21417189e+00 -9.28922221e-02 -2.34470353e-01 -5.18028259e-01
-1.80699408e-01 8.10387850e-01 5.46597362e-01 5.81148565e-01
4.46351945e-01 -1.00916374e+00 -7.43421819e-03 1.20576456e-01
1.78759083e-01 -1.85001388e-01 5.44551685e-02 5.96074224e-01
-4.21681166e-01 4.69555199e-01 -5.30215323e-01 -3.81442040e-01
-7.82073259e-01 6.26074433e-01 5.78870058e-01 -1.72375098e-01
-4.85096544e-01 5.87542713e-01 2.41726756e-01 -7.96175778e-01
4.20923173e-01 -4.29520369e-01 -1.81360796e-01 -4.39595491e-01
1.07439838e-01 2.46663034e-01 -9.46213752e-02 -3.09842885e-01
-3.78223628e-01 5.87941647e-01 9.94973853e-02 -4.16088223e-01
1.05662143e+00 -6.39497876e-01 2.85020858e-01 1.09940931e-01
8.76147747e-01 -2.52217591e-01 -1.47145188e+00 2.82296538e-01
4.38604832e-01 -2.92640150e-01 -1.04518704e-01 -4.26338792e-01
-7.35237420e-01 2.42684469e-01 6.47360623e-01 6.47622198e-02
9.60333765e-01 -5.74174881e-01 5.68410218e-01 8.83627713e-01
9.36707556e-01 -1.26610065e+00 2.76085973e-01 8.36266756e-01
6.48159206e-01 -8.39553833e-01 -4.81360555e-01 -1.42567664e-01
-5.77525616e-01 1.01676333e+00 1.06335795e+00 -3.34499508e-01
3.26600999e-01 3.61736089e-01 4.32451740e-02 2.14672834e-01
-7.05350459e-01 -1.62418801e-02 -1.96570709e-01 7.88908958e-01
-3.89612645e-01 2.74410825e-02 -4.31504518e-01 3.17523062e-01
-1.85043618e-01 -4.96557534e-01 7.98483372e-01 1.17547345e+00
-9.56658959e-01 -1.17605042e+00 -3.57959449e-01 9.42993164e-02
1.24035463e-01 4.34770674e-01 1.70597836e-01 9.40610588e-01
-1.48736358e-01 1.03897238e+00 -3.23987633e-01 -4.67709392e-01
6.10523641e-01 -4.58129823e-01 3.23110342e-01 -2.38849923e-01
-1.73305094e-01 -2.63852954e-01 1.52686119e-01 -7.90771961e-01
-4.17052656e-02 -4.98006701e-01 -1.60169363e+00 -2.72882879e-02
-2.34406009e-01 2.09656820e-01 7.65871286e-01 7.70989656e-01
2.91448712e-01 5.65071285e-01 8.01127255e-01 -7.03295350e-01
-1.08547795e+00 -2.11541831e-01 -3.23613524e-01 -2.21475959e-01
6.40214741e-01 -1.08713031e+00 -4.23321486e-01 -4.62853253e-01] | [4.785665512084961, 1.6936392784118652] |
450ebb2d-5454-45db-b2fb-156699e2b99e | towards-robust-and-semantically-organised | 2205.02309 | null | https://arxiv.org/abs/2205.02309v1 | https://arxiv.org/pdf/2205.02309v1.pdf | Towards Robust and Semantically Organised Latent Representations for Unsupervised Text Style Transfer | Recent studies show that auto-encoder based approaches successfully perform language generation, smooth sentence interpolation, and style transfer over unseen attributes using unlabelled datasets in a zero-shot manner. The latent space geometry of such models is organised well enough to perform on datasets where the style is "coarse-grained" i.e. a small fraction of words alone in a sentence are enough to determine the overall style label. A recent study uses a discrete token-based perturbation approach to map "similar" sentences ("similar" defined by low Levenshtein distance/ high word overlap) close by in latent space. This definition of "similarity" does not look into the underlying nuances of the constituent words while mapping latent space neighbourhoods and therefore fails to recognise sentences with different style-based semantics while mapping latent neighbourhoods. We introduce EPAAEs (Embedding Perturbed Adversarial AutoEncoders) which completes this perturbation model, by adding a finely adjustable noise component on the continuous embeddings space. We empirically show that this (a) produces a better organised latent space that clusters stylistically similar sentences together, (b) performs best on a diverse set of text style transfer tasks than similar denoising-inspired baselines, and (c) is capable of fine-grained control of Style Transfer strength. We also extend the text style transfer tasks to NLI datasets and show that these more complex definitions of style are learned best by EPAAE. To the best of our knowledge, extending style transfer to NLI tasks has not been explored before. | ['Maunendra Sankar Desarkar', 'Suvodip Dey', 'Sharan Narasimhan'] | 2022-05-04 | null | https://aclanthology.org/2022.naacl-main.34 | https://aclanthology.org/2022.naacl-main.34.pdf | naacl-2022-7 | ['text-style-transfoer'] | ['natural-language-processing'] | [ 5.17691791e-01 1.98654443e-01 2.86282122e-01 -5.46810985e-01
-8.57890069e-01 -8.66649032e-01 1.13120103e+00 -2.47386340e-02
-4.80653971e-01 7.66144156e-01 6.77738845e-01 -1.74718842e-01
6.06717989e-02 -9.06500638e-01 -7.04318047e-01 -7.97740102e-01
2.56905377e-01 6.39093041e-01 -1.03974715e-01 -5.27453959e-01
1.93341359e-01 3.67628783e-01 -1.21694732e+00 4.59503412e-01
7.85621583e-01 4.44062561e-01 3.54727991e-02 8.00279856e-01
-3.87009054e-01 3.82292181e-01 -9.60567236e-01 -4.40619588e-01
1.58090129e-01 -7.73553371e-01 -1.02911115e+00 -3.21372487e-02
5.53578615e-01 -1.07743703e-01 -2.20506340e-01 8.28421533e-01
5.01977563e-01 2.65940815e-01 1.12967062e+00 -9.05643582e-01
-1.18039727e+00 4.91905153e-01 -6.87822625e-02 -2.85689253e-02
4.26930398e-01 1.29030913e-01 1.03769386e+00 -8.72596383e-01
8.73013973e-01 1.63709712e+00 8.12030315e-01 8.72185767e-01
-2.00978971e+00 -3.77788275e-01 -2.51218081e-02 -5.01480997e-01
-1.16817582e+00 -3.98154825e-01 8.20249498e-01 -3.03945035e-01
9.99098659e-01 4.43941772e-01 2.74927348e-01 1.73630941e+00
2.67048091e-01 7.25213826e-01 1.35815656e+00 -6.53765082e-01
3.59298587e-01 3.29149961e-01 -2.34280393e-01 2.59186000e-01
-2.78025001e-01 7.61632249e-02 -5.07569969e-01 -1.49631441e-01
6.75014257e-01 -1.97545692e-01 -1.31178111e-01 -2.13799968e-01
-1.46296740e+00 1.17918921e+00 3.71846408e-01 5.77091336e-01
-2.22757477e-02 9.53480229e-02 5.98763883e-01 7.62723446e-01
8.09670627e-01 8.25454235e-01 -3.91667008e-01 -1.90994307e-01
-1.11275709e+00 4.70754802e-01 9.36346233e-01 1.12017119e+00
8.00817788e-01 1.79972962e-01 -4.45595533e-01 9.94796872e-01
-4.41224389e-02 4.24020141e-01 6.63976729e-01 -9.09514368e-01
3.61801028e-01 2.80833006e-01 4.51166518e-02 -7.51093984e-01
-3.98518592e-02 -1.51030108e-01 -9.05641496e-01 4.66256529e-01
5.87443054e-01 -1.91332564e-01 -9.36183035e-01 2.00055218e+00
-2.60349095e-01 -1.23725437e-01 1.83394775e-01 4.92963821e-01
6.25217080e-01 8.05737853e-01 2.50153512e-01 1.57633647e-01
1.36530900e+00 -4.99601096e-01 -5.75711071e-01 -2.80410528e-01
6.98419392e-01 -9.98648465e-01 1.74479425e+00 6.20472580e-02
-1.13073647e+00 -7.20539093e-01 -9.97091472e-01 -3.25605184e-01
-7.14357734e-01 -3.25825989e-01 4.49262559e-01 6.63574040e-01
-1.36461747e+00 9.24416244e-01 -4.66444045e-01 -5.61121225e-01
3.08063030e-01 1.21494666e-01 -4.35723245e-01 1.71815604e-01
-1.48659408e+00 9.70324039e-01 3.22238863e-01 -4.56341267e-01
-5.62017739e-01 -9.11073029e-01 -1.04654288e+00 -2.04686493e-01
-1.51989326e-01 -1.07303798e+00 1.02802300e+00 -1.21343625e+00
-1.53305614e+00 1.10539782e+00 -3.05258065e-01 -4.54318821e-01
5.53120077e-01 -8.90410766e-02 -2.87184298e-01 -4.89626639e-02
2.16658801e-01 1.21153998e+00 1.04533160e+00 -1.33763075e+00
-1.34671256e-01 -1.68984219e-01 5.85082583e-02 2.82145560e-01
-5.19032657e-01 6.81937262e-02 1.68646887e-01 -1.37123907e+00
-3.02256435e-01 -8.71125042e-01 -3.13776769e-02 5.63778169e-02
-1.62451670e-01 -2.82617033e-01 8.44339490e-01 -4.80647922e-01
1.02068913e+00 -2.09452438e+00 4.34314370e-01 -7.37493560e-02
1.07353188e-01 4.15967628e-02 -3.55738491e-01 6.23767018e-01
-6.24544695e-02 4.56520438e-01 -4.68361497e-01 -6.39585674e-01
3.12686443e-01 5.91197371e-01 -5.43674648e-01 8.62067714e-02
4.54185247e-01 1.22818816e+00 -9.06679809e-01 -3.49833310e-01
6.81839436e-02 5.67463875e-01 -6.70155823e-01 2.46021792e-01
-1.96416050e-01 4.14672613e-01 3.35126068e-03 1.80586055e-02
2.52375126e-01 1.89767778e-01 -2.89583027e-01 -1.55827463e-01
1.32455364e-01 3.28782171e-01 -1.01365590e+00 1.98717713e+00
-8.30000341e-01 7.91845560e-01 -1.20719075e-01 -6.94795310e-01
1.21219814e+00 3.10663491e-01 6.02942146e-03 -3.15749079e-01
-1.49423167e-01 1.38039649e-01 -1.91392645e-01 -1.45619392e-01
5.81793904e-01 -8.10042262e-01 -3.98830533e-01 5.72014809e-01
3.93168598e-01 -6.43668771e-01 4.89280634e-02 2.84026086e-01
1.13516784e+00 2.71393776e-01 -5.79565810e-03 -5.08074880e-01
2.86417842e-01 -2.21711189e-01 2.21730828e-01 9.21326578e-01
-1.91912651e-02 9.81886446e-01 4.88338500e-01 -2.38018528e-01
-1.64184558e+00 -1.37276292e+00 -2.78294027e-01 1.27804935e+00
-2.97551274e-01 -2.05860391e-01 -9.43371952e-01 -7.53270149e-01
3.28674936e-03 1.29320729e+00 -1.07312262e+00 -3.24016601e-01
-3.92556429e-01 -4.07814234e-01 8.90400052e-01 7.06321716e-01
1.33731499e-01 -1.51244187e+00 -2.12672893e-02 2.05515996e-01
-1.57512605e-01 -7.10674584e-01 -7.44239032e-01 3.70155632e-01
-7.73837507e-01 -2.84670144e-01 -1.04459393e+00 -7.93577850e-01
5.20877361e-01 -2.99994200e-01 1.56423593e+00 -3.36799234e-01
-2.36851741e-02 3.33086997e-01 -4.91910785e-01 -4.74563986e-01
-9.21873987e-01 1.14814326e-01 1.50850832e-01 -1.34892881e-01
6.38260424e-01 -7.21885562e-01 -2.51921624e-01 9.16173160e-02
-1.20061362e+00 -1.53128847e-01 4.40336645e-01 1.16651165e+00
2.75759339e-01 -2.30646044e-01 6.37785673e-01 -1.10166061e+00
1.05325687e+00 -2.93749154e-01 1.39148071e-01 6.68671653e-02
-3.51783961e-01 4.41649705e-01 7.24078357e-01 -5.54498076e-01
-1.02246368e+00 -3.98245990e-01 -2.87860990e-01 -2.49745429e-01
-8.27945232e-01 -7.29967430e-02 -3.28552872e-01 2.42252201e-01
1.06079853e+00 2.55666256e-01 3.26862514e-01 -5.23377717e-01
7.08672762e-01 6.18529201e-01 3.98574769e-01 -8.11294615e-01
9.98023033e-01 4.67101157e-01 -2.12418631e-01 -8.97512138e-01
-6.81004465e-01 -1.94985390e-01 -9.27766860e-01 3.92691970e-01
1.03305233e+00 -6.43050611e-01 -9.49375704e-02 2.08165526e-01
-1.29230762e+00 -5.16109526e-01 -9.79470909e-01 1.19070239e-01
-9.29432154e-01 4.07396704e-01 -7.66110897e-01 -5.05815566e-01
-3.35803032e-01 -1.00774908e+00 1.30771494e+00 -2.60724038e-01
-9.55279768e-01 -1.54489195e+00 1.67644083e-01 -7.82874078e-02
4.85176146e-01 2.46843517e-01 1.07557559e+00 -6.91032350e-01
1.33571103e-01 6.34893999e-02 -4.26556021e-02 6.31678522e-01
3.39032590e-01 -1.59839168e-01 -1.16435826e+00 -2.76726663e-01
1.77093685e-01 -3.05673957e-01 9.40978825e-01 1.94575310e-01
7.72937000e-01 -3.38677377e-01 3.20348926e-02 6.20935142e-01
1.15988982e+00 -6.98113739e-02 8.68605018e-01 3.19461644e-01
7.92682350e-01 8.21452796e-01 5.88539355e-02 3.50014418e-02
-2.37013474e-02 7.19475806e-01 -2.06092238e-01 -3.03389043e-01
-2.84491867e-01 -3.79109681e-01 5.01038790e-01 7.67888069e-01
2.45317489e-01 -2.36396104e-01 -6.75825417e-01 7.33436763e-01
-1.38931668e+00 -1.02214158e+00 1.77836522e-01 1.83134568e+00
1.23587120e+00 4.32484597e-01 9.86427888e-02 1.93989426e-01
4.90777194e-01 2.03629971e-01 -3.11365217e-01 -1.10380924e+00
-3.33030164e-01 6.01097226e-01 2.67121792e-01 7.59701729e-01
-1.05573487e+00 1.11360157e+00 6.09414005e+00 1.11608338e+00
-6.85561538e-01 6.31362945e-02 6.59545600e-01 5.29076681e-02
-8.58746946e-01 -6.28169775e-02 -6.66570723e-01 5.75728238e-01
1.02462220e+00 2.47205421e-02 4.91646856e-01 6.41093373e-01
2.08721057e-01 3.35361212e-01 -1.43712807e+00 7.05466628e-01
1.00971453e-01 -1.28934026e+00 3.73881340e-01 -1.13972574e-01
8.53072047e-01 -2.99368173e-01 2.81723917e-01 3.47844005e-01
5.37399113e-01 -1.37073565e+00 7.40842342e-01 7.21905828e-01
1.09872735e+00 -8.75493288e-01 5.52009702e-01 2.67739117e-01
-7.37408936e-01 2.34391138e-01 -4.90232348e-01 -7.54420385e-02
2.11321905e-01 4.03529316e-01 -6.50014400e-01 2.97450393e-01
4.99192357e-01 5.13817191e-01 -7.69358158e-01 2.06266731e-01
-2.30333537e-01 7.18190968e-01 -1.99836656e-01 -1.04524404e-01
4.85832006e-01 -1.88665792e-01 7.32412696e-01 1.61031377e+00
1.10100850e-01 -1.81575462e-01 -1.08332053e-01 1.24344814e+00
1.02174394e-01 2.90601179e-02 -9.83045399e-01 1.20495306e-02
3.13129187e-01 8.92894924e-01 -5.10938346e-01 -3.63180965e-01
-3.17376077e-01 1.46203065e+00 1.09901622e-01 5.37760019e-01
-3.45180869e-01 -6.86479747e-01 8.71649384e-01 1.18434392e-01
2.68532813e-01 -2.43082061e-01 -8.37086439e-01 -1.11557627e+00
-1.63486347e-01 -9.06579196e-01 3.91100273e-02 -9.04412031e-01
-1.86336792e+00 8.17627966e-01 -5.97654171e-02 -1.08234453e+00
-5.91555476e-01 -5.81511199e-01 -8.06289196e-01 1.49700618e+00
-9.81358409e-01 -1.32684803e+00 1.14592850e-01 4.34112757e-01
8.04241478e-01 -3.84800673e-01 1.41298890e+00 -2.55343020e-01
8.93198922e-02 8.05481076e-01 2.94048548e-01 2.67894179e-01
9.87713397e-01 -1.88104951e+00 9.24894273e-01 6.10082924e-01
3.88355523e-01 9.05234635e-01 1.00873196e+00 -7.02843010e-01
-7.52405465e-01 -1.14014208e+00 1.19111502e+00 -9.78375256e-01
6.56517208e-01 -7.97188759e-01 -1.16424286e+00 5.69444299e-01
5.59936821e-01 -2.87635028e-01 7.65265942e-01 2.98957914e-01
-4.45961624e-01 1.98431551e-01 -1.13281274e+00 8.07729065e-01
1.06262064e+00 -8.77906144e-01 -1.01852059e+00 2.56421119e-01
8.54116440e-01 5.40379919e-02 -1.03556681e+00 2.51086265e-01
2.78785855e-01 -8.56780827e-01 9.89114761e-01 -8.63227487e-01
6.28030717e-01 -7.54637420e-02 -1.64465606e-01 -1.69898462e+00
-3.98369253e-01 -6.39567316e-01 2.13006094e-01 1.68311596e+00
3.96462888e-01 -4.29899395e-01 6.55253232e-01 4.37809706e-01
-3.35474573e-02 -7.23282635e-01 -6.89525366e-01 -7.92563140e-01
8.23958755e-01 -2.84302741e-01 6.03141904e-01 9.65902030e-01
-3.72418612e-01 4.69530821e-01 -1.21230192e-01 -4.42499369e-01
4.65363443e-01 -1.91922396e-01 6.30635440e-01 -1.06228328e+00
-3.86109442e-01 -7.31485665e-01 -3.80956650e-01 -9.12100434e-01
4.66851473e-01 -1.16983366e+00 -1.49739176e-01 -1.29678226e+00
-1.57014802e-01 -3.85641456e-01 1.08128320e-02 1.90308496e-01
-2.56488323e-01 4.57381606e-01 1.90817624e-01 1.69880614e-01
-1.85975060e-01 6.47209585e-01 1.16029680e+00 -1.71338767e-01
-1.23187356e-01 -6.34762421e-02 -7.66153812e-01 7.30748892e-01
9.10824358e-01 -3.24785590e-01 -4.96379137e-01 -3.99518549e-01
4.79425490e-02 -4.22492623e-01 5.34140289e-01 -7.55332053e-01
-2.14864478e-01 4.02129665e-02 7.47105002e-01 -1.00945599e-01
2.45594770e-01 -5.41067004e-01 -1.52547255e-01 1.08319014e-01
-8.91819537e-01 2.08337624e-02 1.66743413e-01 4.80080485e-01
-1.28481790e-01 -3.11610132e-01 9.16921914e-01 -3.10322136e-01
-5.20798683e-01 -9.33270276e-05 -4.24071193e-01 5.22803247e-01
5.67610323e-01 -4.35707152e-01 1.72652364e-01 -4.78947490e-01
-9.16528821e-01 -3.96224767e-01 7.61217952e-01 5.35120487e-01
2.56233305e-01 -1.69112861e+00 -9.74622190e-01 3.90134662e-01
1.54353350e-01 -9.56834331e-02 -4.47347090e-02 1.69434428e-01
-2.84119248e-01 6.03427887e-02 -1.44729316e-01 -5.86510360e-01
-9.10964251e-01 5.28232813e-01 1.62410855e-01 -2.08197519e-01
-7.51625478e-01 1.15241361e+00 4.44389403e-01 -7.92122900e-01
2.35766452e-02 -1.60362080e-01 8.60624462e-02 2.55615324e-01
4.22254950e-01 1.24452822e-01 -2.01739129e-02 -6.97607994e-01
-1.37167173e-02 4.92968649e-01 -1.84414104e-01 -4.75567967e-01
1.20923162e+00 -2.58024931e-01 -2.51648389e-02 8.86885583e-01
1.49790812e+00 5.19948825e-02 -1.38719094e+00 -3.34144741e-01
1.25280814e-03 -3.27034473e-01 -2.76771843e-01 -7.52608359e-01
-2.68048167e-01 1.20924246e+00 3.89904588e-01 4.42407817e-01
8.55871797e-01 7.17638358e-02 8.19023013e-01 1.43606171e-01
-4.50647548e-02 -1.16393566e+00 2.44067535e-01 7.05491960e-01
1.17058229e+00 -1.00445127e+00 -4.49096173e-01 5.45243658e-02
-9.51943278e-01 1.05857205e+00 3.08508635e-01 -3.44485283e-01
4.99425054e-01 4.59533304e-01 3.36931199e-01 1.02247976e-01
-5.95305681e-01 -4.83173802e-02 4.94285882e-01 8.42160821e-01
6.37916088e-01 1.25545427e-01 -3.52570899e-02 2.69247949e-01
-7.56417036e-01 -4.98191535e-01 2.10140690e-01 6.32778347e-01
-3.00714850e-01 -1.37472880e+00 -4.02370423e-01 4.68556374e-01
-3.43408018e-01 -3.83392125e-01 -4.63826150e-01 6.16978526e-01
4.13837619e-02 7.90172338e-01 2.90759623e-01 -2.33225495e-01
3.45320463e-01 5.90922713e-01 3.70077670e-01 -8.33535969e-01
-6.01809978e-01 -1.75635722e-02 -5.89952543e-02 -2.14742064e-01
-9.42780003e-02 -6.82233393e-01 -8.65369439e-01 -4.03243452e-01
1.16733208e-01 2.75477283e-02 5.66147566e-01 9.95029747e-01
1.97635651e-01 5.81757069e-01 5.43264449e-01 -1.04897821e+00
-7.34094620e-01 -1.25475860e+00 -5.82573771e-01 1.11145103e+00
2.39632592e-01 -4.44160610e-01 -4.69424337e-01 3.00378352e-01] | [11.45560073852539, 9.701375961303711] |
7bc923d7-d609-42fb-bce9-fe1023d7999d | copilot-human-collision-prediction-and | 2210.01781 | null | https://arxiv.org/abs/2210.01781v2 | https://arxiv.org/pdf/2210.01781v2.pdf | COPILOT: Human-Environment Collision Prediction and Localization from Egocentric Videos | The ability to forecast human-environment collisions from egocentric observations is vital to enable collision avoidance in applications such as VR, AR, and wearable assistive robotics. In this work, we introduce the challenging problem of predicting collisions in diverse environments from multi-view egocentric videos captured from body-mounted cameras. Solving this problem requires a generalizable perception system that can classify which human body joints will collide and estimate a collision region heatmap to localize collisions in the environment. To achieve this, we propose a transformer-based model called COPILOT to perform collision prediction and localization simultaneously, which accumulates information across multi-view inputs through a novel 4D space-time-viewpoint attention mechanism. To train our model and enable future research on this task, we develop a synthetic data generation framework that produces egocentric videos of virtual humans moving and colliding within diverse 3D environments. This framework is then used to establish a large-scale dataset consisting of 8.6M egocentric RGBD frames. Extensive experiments show that COPILOT generalizes to unseen synthetic as well as real-world scenes. We further demonstrate COPILOT outputs are useful for downstream collision avoidance through simple closed-loop control. Please visit our project webpage at https://sites.google.com/stanford.edu/copilot. | ['Leonidas J. Guibas', 'Yanchao Yang', 'Kaichun Mo', 'Despoina Paschalidou', 'Davis Rempe', 'Bokui Shen', 'Boxiao Pan'] | 2022-10-04 | null | null | null | null | ['synthetic-data-generation', 'synthetic-data-generation'] | ['medical', 'miscellaneous'] | [-1.17695756e-01 -1.11264825e-01 2.36288354e-01 -3.66844654e-01
-4.22627032e-01 -4.66548473e-01 3.51738006e-01 -4.51280206e-01
-3.25370997e-01 4.85597134e-01 3.37161928e-01 7.44539201e-02
1.84489340e-01 -5.33586681e-01 -9.51407731e-01 -1.98286891e-01
-7.20947459e-02 7.26201594e-01 3.15257758e-01 -5.16394138e-01
6.01789057e-02 3.93152714e-01 -1.60861933e+00 2.74430782e-01
5.01565993e-01 5.59111953e-01 6.04991913e-01 1.06595528e+00
6.39887512e-01 5.22042394e-01 -2.71354407e-01 -2.29392543e-01
3.62222105e-01 -1.64539024e-01 -6.61214471e-01 3.91771272e-02
3.84882718e-01 -5.11151493e-01 -5.62441468e-01 5.08030474e-01
6.73407376e-01 5.07323503e-01 4.35434520e-01 -1.62777770e+00
-2.12204531e-01 -2.17346400e-02 -3.88422519e-01 3.11674654e-01
9.81812119e-01 3.77889335e-01 6.38263166e-01 -9.78178203e-01
1.00339293e+00 1.41450417e+00 5.66930652e-01 7.54466474e-01
-8.62295866e-01 -5.17946601e-01 2.30674371e-01 3.80128145e-01
-1.27344453e+00 -4.04762119e-01 7.92457700e-01 -4.90908265e-01
1.25718319e+00 2.64727652e-01 1.08724141e+00 1.44203532e+00
7.81915367e-01 8.47255647e-01 3.87565762e-01 -6.78193346e-02
1.34633943e-01 -4.37547207e-01 -4.32866544e-01 6.68421626e-01
2.50223339e-01 -9.49091685e-04 -7.49884307e-01 -1.63468942e-02
1.07090151e+00 2.01880738e-01 -2.37026408e-01 -1.17194521e+00
-1.58371007e+00 5.62292874e-01 6.37468100e-01 -3.36993098e-01
-4.01189297e-01 5.39503753e-01 3.63795042e-01 -7.07282797e-02
3.24906856e-01 2.76672989e-01 -4.31767911e-01 -2.67239124e-01
-1.49142668e-01 7.90926635e-01 3.87434691e-01 1.18984294e+00
3.79952848e-01 -1.21389084e-01 1.77900970e-01 4.26160365e-01
2.30019897e-01 6.74326241e-01 1.92656845e-01 -1.39023483e+00
7.02379584e-01 3.61541331e-01 4.55116928e-01 -1.02633440e+00
-8.03076625e-01 -1.69588879e-01 -4.28615898e-01 1.22375272e-01
6.23619407e-02 -1.37575015e-01 -7.93557584e-01 1.67011487e+00
8.11662495e-01 2.05132812e-01 -1.20793946e-01 1.47413957e+00
4.16204333e-01 3.79616112e-01 3.21776494e-02 3.94081593e-01
1.37111485e+00 -1.00303817e+00 -3.70971918e-01 -4.67244297e-01
6.45232797e-01 -4.30338204e-01 9.70524371e-01 3.78405333e-01
-1.03429246e+00 -6.98973179e-01 -8.07590604e-01 -2.35145196e-01
-3.86561379e-02 -8.04708228e-02 6.65419996e-01 6.13693334e-02
-8.20434093e-01 3.24730724e-01 -1.40683126e+00 -6.25937462e-01
5.07758893e-02 2.71400660e-01 -6.67275310e-01 -5.74071743e-02
-1.12910128e+00 8.68386328e-01 2.39614129e-01 1.47750318e-01
-9.90524113e-01 -3.43285501e-01 -1.39579809e+00 -3.06244552e-01
3.91065001e-01 -1.55451870e+00 1.48299348e+00 -3.82346511e-01
-1.13865376e+00 7.06477165e-01 -2.49108702e-01 -2.75232047e-01
5.59514880e-01 -8.89329791e-01 -8.05545226e-02 2.08869413e-01
4.37564403e-01 7.85532236e-01 4.84134644e-01 -1.23388076e+00
-6.26761794e-01 -6.03332162e-01 1.76478401e-02 8.42082143e-01
3.33415627e-01 -3.33461791e-01 -7.44860113e-01 -5.81686616e-01
4.13920909e-01 -1.25294375e+00 -4.84696239e-01 1.83984488e-01
-4.46417749e-01 1.36021711e-02 7.05738246e-01 -3.71077240e-01
5.34690142e-01 -1.80836070e+00 5.77404678e-01 -8.04171935e-02
9.70147699e-02 -2.57084489e-01 4.29411232e-02 4.39828604e-01
1.43322483e-01 -3.61958236e-01 1.68641135e-01 -5.58008075e-01
-1.34523228e-01 2.13765129e-01 -2.06950322e-01 5.63411057e-01
1.55016547e-02 1.08496118e+00 -1.23547733e+00 -4.67035264e-01
6.17387295e-01 6.04408860e-01 -9.61167812e-01 3.41650575e-01
-9.59262624e-02 8.72564673e-01 -5.06366372e-01 6.78461194e-01
4.17107970e-01 -7.53020346e-02 -2.74353530e-02 -1.95315897e-01
1.22479610e-01 9.32422429e-02 -1.12416863e+00 2.40058208e+00
-3.99736524e-01 4.40481246e-01 -1.78323202e-02 -5.78582644e-01
5.49496770e-01 1.53342664e-01 6.02026522e-01 -6.20511174e-01
2.82989502e-01 -1.45138323e-01 -4.70678091e-01 -6.21905625e-01
8.70298445e-01 3.32612805e-02 -4.52644289e-01 1.55239314e-01
-6.83076605e-02 -2.80520707e-01 -1.24209799e-01 4.09147620e-01
1.08197343e+00 7.66717672e-01 1.68738350e-01 2.43920594e-01
2.86018521e-01 3.80886167e-01 7.06519008e-01 5.82168639e-01
-4.74783003e-01 9.04260695e-01 -7.96212628e-02 -7.06354856e-01
-1.26614618e+00 -1.61003768e+00 2.51990050e-01 9.01722968e-01
7.11962461e-01 -4.66347069e-01 -5.53182900e-01 -3.84135902e-01
1.18770100e-01 5.01148641e-01 -3.15233558e-01 -2.75405020e-01
-8.03398907e-01 -1.59991458e-01 1.17003284e-01 8.37519825e-01
2.72081703e-01 -1.00146580e+00 -1.29078948e+00 1.30369261e-01
-7.42876053e-01 -1.32200432e+00 -5.18384278e-01 -1.89880773e-01
-6.46638095e-01 -1.21292651e+00 -7.24904835e-01 -5.72819293e-01
5.37547946e-01 6.90945387e-01 1.09988117e+00 -3.17932844e-01
-5.71856797e-01 8.52745593e-01 -3.80762428e-01 -3.12838674e-01
6.15232661e-02 -1.38661802e-01 6.64833724e-01 -4.23337042e-01
2.26392746e-01 -4.95446414e-01 -9.90401506e-01 4.39185202e-01
-2.02913657e-01 4.93407398e-01 1.62958115e-01 6.17234468e-01
7.08754301e-01 -4.04042333e-01 4.30556573e-02 -3.27050865e-01
2.53886789e-01 -5.51182687e-01 -3.25474530e-01 -1.31539762e-01
1.14388593e-01 -2.90413797e-01 4.04733747e-01 -2.75566965e-01
-9.89597261e-01 5.24662077e-01 -2.05611978e-02 -8.51929605e-01
-1.70486763e-01 1.02361299e-01 -2.10620552e-01 2.68907666e-01
6.38814211e-01 5.84174581e-02 -1.37634560e-01 -1.73682600e-01
5.25822401e-01 4.18319881e-01 1.08454180e+00 -5.19482791e-01
7.06387520e-01 7.96017587e-01 -1.22082844e-01 -5.72591066e-01
-4.38118935e-01 -6.52704716e-01 -8.60254705e-01 -4.89924341e-01
1.13349581e+00 -1.36919618e+00 -1.09500706e+00 3.33065718e-01
-1.22078466e+00 -4.35435951e-01 -1.64175220e-02 7.18366444e-01
-1.17785168e+00 5.17429292e-01 -5.47396600e-01 -8.84724915e-01
-2.73058534e-01 -1.10320604e+00 1.63930428e+00 1.08804211e-01
-5.56904078e-01 -5.73083460e-01 3.09221119e-01 5.32879412e-01
-1.40684068e-01 4.43650365e-01 8.59251991e-02 3.99770923e-02
-7.13998497e-01 -1.62805170e-02 2.30319366e-01 -3.48617494e-01
-1.29487005e-03 -3.07228863e-01 -5.48548520e-01 -4.23157394e-01
-2.45630667e-01 -3.11086714e-01 5.66101313e-01 4.92759556e-01
9.61845577e-01 1.21102408e-01 -6.90755725e-01 8.22359383e-01
9.02887464e-01 9.84457880e-02 5.06991625e-01 5.44673026e-01
1.05871701e+00 6.24710739e-01 1.15958869e+00 6.49955213e-01
9.83057380e-01 1.14932632e+00 7.34871089e-01 1.46196872e-01
1.10191852e-01 -5.53018451e-01 3.94212425e-01 5.82951605e-01
-1.13863453e-01 -4.00902003e-01 -1.05181897e+00 5.69922030e-01
-2.03773618e+00 -1.06151092e+00 3.01732123e-02 2.06880236e+00
9.51343104e-02 1.91321611e-01 1.38342977e-01 -2.16205418e-01
5.68245292e-01 -7.90103897e-02 -7.77362823e-01 -3.87547851e-01
4.05699372e-01 -2.77268320e-01 3.39708686e-01 5.04040420e-01
-1.31421840e+00 1.12011147e+00 5.45279741e+00 -5.28379856e-03
-1.00512064e+00 -1.63034759e-02 1.12810090e-01 -3.57165396e-01
-1.56757936e-01 -1.66387647e-01 -5.33808589e-01 2.83636183e-01
6.48218572e-01 -1.43057974e-02 2.68819541e-01 1.09252763e+00
3.90311927e-01 -2.44988233e-01 -1.14888632e+00 1.32098210e+00
1.63659692e-01 -8.79342198e-01 -2.42345259e-02 2.02321284e-03
4.20566231e-01 3.36575091e-01 -4.03026827e-02 2.23664343e-01
3.88568431e-01 -6.35244906e-01 8.91688526e-01 6.45000517e-01
5.42248487e-01 -6.78648412e-01 3.69534612e-01 6.02462590e-01
-1.50989103e+00 -1.26830608e-01 -3.68823498e-01 -3.52968186e-01
8.01311135e-01 1.21086836e-01 -9.42521214e-01 6.47889316e-01
9.75002110e-01 7.33965576e-01 -2.54540831e-01 9.58580673e-01
-9.81056467e-02 -2.45026231e-01 -3.74326140e-01 3.00413758e-01
-1.23252243e-01 -2.44984031e-02 8.83284688e-01 7.92230964e-01
5.06195426e-01 3.32578927e-01 2.40869477e-01 6.40184104e-01
3.63979012e-01 -3.00071865e-01 -1.01949990e+00 6.05064571e-01
3.88865083e-01 1.10519004e+00 -6.47364557e-01 -3.56270313e-01
-7.47665688e-02 1.61259735e+00 4.01110709e-01 3.87505621e-01
-1.14023411e+00 -1.98653817e-01 1.04360294e+00 1.80625632e-01
1.52263939e-01 -7.61393368e-01 1.45049840e-01 -1.61042309e+00
1.91799343e-01 -4.61446881e-01 2.47130781e-01 -1.44499159e+00
-8.61394763e-01 5.66756725e-01 2.59848475e-01 -1.55561697e+00
-6.29849136e-01 -5.13118923e-01 -4.35964853e-01 5.70555151e-01
-8.90905082e-01 -1.14951694e+00 -8.76159132e-01 7.77824223e-01
8.31159472e-01 1.73044294e-01 5.32078385e-01 3.19258380e-03
-3.18627447e-01 2.21865594e-01 -3.90039057e-01 -3.79316621e-02
7.62139499e-01 -1.10930276e+00 1.15625644e+00 6.70886636e-01
1.27170548e-01 5.56552291e-01 8.54737401e-01 -9.85334873e-01
-1.72657669e+00 -1.11630285e+00 6.14955306e-01 -1.16966927e+00
1.45591244e-01 -6.56224489e-01 -3.70450079e-01 9.96978402e-01
-3.15372646e-01 2.55415201e-01 2.34438285e-01 -1.22732930e-01
-6.18821904e-02 1.31572202e-01 -1.01675653e+00 9.29008186e-01
1.75614142e+00 -4.38502103e-01 -6.58388257e-01 2.27875516e-01
6.94359243e-01 -1.00771832e+00 -6.23557210e-01 4.98537272e-01
8.97642314e-01 -1.01757550e+00 1.28932750e+00 -5.48777103e-01
3.08836311e-01 -3.59263897e-01 -3.53667825e-01 -1.37443280e+00
-3.96640092e-01 -3.93283010e-01 -2.99275845e-01 4.41344053e-01
-2.09841490e-01 -4.16710109e-01 1.17886162e+00 6.82633519e-01
-4.33728784e-01 -6.16325617e-01 -1.02571893e+00 -6.67371154e-01
-4.15978760e-01 -6.16239548e-01 2.58546531e-01 4.65739340e-01
1.18375376e-01 2.40459576e-01 -5.23022890e-01 4.44167972e-01
7.37224817e-01 -1.32059092e-02 1.47509432e+00 -9.47543740e-01
-9.28601921e-02 2.49609679e-01 -8.31674993e-01 -1.32478714e+00
3.68264653e-02 -6.38436854e-01 2.28266150e-01 -1.62575245e+00
1.53846830e-01 -2.73464024e-01 6.82545230e-02 1.12933025e-01
-1.98639899e-01 3.88070971e-01 3.96076053e-01 2.48904735e-01
-8.63493204e-01 8.90196741e-01 1.42936075e+00 1.48740903e-01
-2.60607898e-01 -1.63092703e-01 -2.60401219e-01 8.28571320e-01
7.76730716e-01 -2.53561825e-01 -6.43435597e-01 -6.86488450e-01
1.02101646e-01 3.99379194e-01 8.80751014e-01 -1.44945896e+00
2.54185587e-01 -2.77013898e-01 7.29824901e-01 -8.99237812e-01
1.05113304e+00 -7.18604028e-01 3.96143466e-01 6.47713363e-01
7.70064592e-02 5.30032814e-01 9.22757387e-02 6.92115664e-01
2.00003237e-01 5.63798308e-01 1.01086065e-01 -4.49168563e-01
-1.05831134e+00 3.87087435e-01 -3.37262392e-01 1.85394399e-02
1.28409290e+00 -4.39592183e-01 -1.71908692e-01 -7.12243378e-01
-9.36794043e-01 5.77124596e-01 8.82551074e-01 8.22432339e-01
1.01077259e+00 -1.34038341e+00 -4.28093463e-01 3.22492123e-01
2.83940047e-01 3.75241667e-01 6.21498466e-01 6.21978521e-01
-7.54920781e-01 4.75405127e-01 -4.84822541e-01 -1.09657729e+00
-1.23576939e+00 5.35180867e-01 3.38529080e-01 1.80812314e-01
-9.18285608e-01 7.73165286e-01 5.90323567e-01 -7.14388728e-01
-7.19523877e-02 -3.19307566e-01 6.90288469e-02 -5.78413546e-01
2.85793751e-01 3.67082149e-01 -1.73676401e-01 -1.06974828e+00
-5.39738059e-01 7.42348492e-01 1.82359323e-01 -3.13818723e-01
1.06002939e+00 -5.79403043e-01 5.06381631e-01 3.15765083e-01
9.59812701e-01 -2.52371460e-01 -1.60781956e+00 2.84832001e-01
-5.05693436e-01 -7.52141833e-01 -4.51053590e-01 -2.56972969e-01
-6.99916720e-01 6.86469913e-01 6.02643728e-01 -5.42775154e-01
8.62158835e-01 1.86441958e-01 1.05201447e+00 4.83733773e-01
1.06505334e+00 -9.05081511e-01 4.32026803e-01 5.23800850e-01
1.05070448e+00 -1.09753144e+00 2.77810246e-02 -5.11760950e-01
-9.84707355e-01 7.44397938e-01 1.12183118e+00 -3.05089682e-01
2.82470375e-01 6.21026866e-02 7.96154737e-02 -4.04909611e-01
-9.17863369e-01 -4.69509065e-02 1.14672467e-01 8.82062554e-01
6.54200017e-02 4.95775342e-02 1.36558652e-01 4.69886929e-01
-5.24103880e-01 -9.69452411e-03 5.15298426e-01 1.22201204e+00
-3.13327283e-01 -8.25458050e-01 -5.30117631e-01 8.81031007e-02
-4.93178628e-02 4.13204908e-01 -1.58421755e-01 6.96021974e-01
1.29295647e-01 7.52032220e-01 3.70156735e-01 -7.58143246e-01
6.54001236e-01 -1.77625567e-01 6.02691591e-01 -6.43040359e-01
-2.10117519e-01 -7.72690680e-03 8.66581500e-02 -1.20170593e+00
-2.63440698e-01 -7.73811460e-01 -1.50615084e+00 -1.97102383e-01
-2.87385583e-02 -7.76449144e-02 5.93122303e-01 5.77117741e-01
6.37855351e-01 4.67684060e-01 3.32249165e-01 -1.71374261e+00
-2.90720791e-01 -5.06106973e-01 -2.31128484e-01 6.63948298e-01
3.92826140e-01 -1.09023082e+00 2.13397387e-02 7.10281655e-02] | [6.996034622192383, -0.8411203026771545] |
76414b65-a929-4d3b-ba9a-d46ee455d1fc | function-prediction | 2307.02173 | null | https://arxiv.org/abs/2307.02173v2 | https://arxiv.org/pdf/2307.02173v2.pdf | Function Prediction | While many good textbooks are available on Protein Structure, Molecular Simulations, Thermodynamics and Bioinformatics methods in general, there is no good introductory level book for the field of Structural Bioinformatics. This book aims to give an introduction into Structural Bioinformatics, which is where the previous topics meet to explore three dimensional protein structures through computational analysis. We provide an overview of existing computational techniques, to validate, simulate, predict and analyse protein structures. More importantly, it will aim to provide practical knowledge about how and when to use such techniques. We will consider proteins from three major vantage points: Protein structure quantification, Protein structure prediction, and Protein simulation & dynamics. There are still huge gaps in understanding the molecular function of proteins. This raises the question on how we may predict protein function, when little to no knowledge from direct experiments is available. Protein function is a broad concept which spans different scales: from quantum scale effects for catalyzing enzymatic reactions, to phenotypes that manifest at the organism level. In fact, many of these functional scales are entirely different research areas. Here, we will consider prediction of a smaller range of functions, roughly spanning the protein residue-level up to the pathway level. We will give a conceptual overview of which functional aspects of proteins we can predict, which methods are currently available, and how well they work in practice. | ['K. Anton Feenstra', 'Sanne Abeln', 'Jose Gavaldá-Garciá', 'Katharina Waury', 'Olga Ivanova', 'Hans de Ferrante', 'Qingzhen Hou', 'Annika Jacobsen', 'Bas Stringer'] | 2023-07-05 | null | null | null | null | ['protein-structure-prediction'] | ['miscellaneous'] | [ 3.94529462e-01 -1.37012014e-02 -1.39527142e-01 -6.21602312e-02
-3.21062952e-01 -6.86111152e-01 1.13686889e-01 4.29581910e-01
-2.06554770e-01 1.31669831e+00 -1.46745712e-01 -6.51798606e-01
-9.40288380e-02 -3.61698985e-01 -7.54769146e-01 -1.22828329e+00
-9.57064703e-02 4.07259315e-01 3.93499523e-01 -6.06064022e-01
2.39043310e-01 7.06073105e-01 -1.42382300e+00 8.74587223e-02
8.79915893e-01 3.60177070e-01 5.11033297e-01 8.04621398e-01
-4.64675464e-02 6.24278963e-01 -4.07489568e-01 -4.04039770e-01
-3.25990230e-01 -8.38081658e-01 -9.66632426e-01 -1.31914526e-01
-2.58472770e-01 3.71365666e-01 2.50094056e-01 6.25142455e-01
5.28669834e-01 -1.87911779e-01 5.73674142e-01 -5.19438684e-01
-7.32660592e-01 -7.57638738e-02 -1.85579553e-01 1.37420446e-01
5.91961682e-01 4.51683581e-01 8.40734303e-01 -5.73883235e-01
8.99112105e-01 1.07890320e+00 6.60021961e-01 6.16489768e-01
-1.63335490e+00 8.23171064e-02 -2.35338926e-01 4.07672942e-01
-9.70505416e-01 -2.66665578e-01 3.00144643e-01 -5.84450722e-01
1.51800668e+00 2.01664299e-01 7.23666430e-01 7.55228341e-01
7.40568578e-01 6.62786141e-02 1.22290123e+00 -4.36850458e-01
2.76849896e-01 -1.81923807e-01 3.05089325e-01 5.46318054e-01
3.32755089e-01 2.27526218e-01 -7.19711006e-01 -4.64268804e-01
2.00984538e-01 -7.03337491e-02 -3.15322250e-01 -7.50152469e-01
-1.03950500e+00 6.51948392e-01 -1.32294223e-01 4.64933485e-01
-4.58755493e-01 -1.66407265e-02 4.69658166e-01 2.45999843e-01
9.14989412e-02 4.58173901e-01 -1.18845797e+00 -2.68873751e-01
-5.90314209e-01 5.65956712e-01 8.73781562e-01 4.47017819e-01
6.89815998e-01 -3.66262943e-01 4.78963256e-01 6.54827297e-01
3.71622801e-01 2.60628432e-01 3.04211944e-01 -7.40587890e-01
-2.84782559e-01 2.94960558e-01 3.39428008e-01 -4.03197736e-01
-5.81874311e-01 3.69428575e-01 -4.16827649e-01 4.63562429e-01
7.76279092e-01 3.98805961e-02 -5.72673738e-01 1.69068527e+00
4.11010027e-01 -2.94208646e-01 1.71759158e-01 6.95638657e-01
6.70653403e-01 6.37439311e-01 4.81332928e-01 -8.76953125e-01
1.57076824e+00 -5.48336148e-01 -5.72500885e-01 1.98160142e-01
8.78007770e-01 -9.14013207e-01 6.13340259e-01 4.94760752e-01
-1.31195903e+00 -2.74989337e-01 -1.11243963e+00 -3.72547537e-01
-7.45088398e-01 -1.61035106e-01 7.07109809e-01 7.10602701e-01
-9.16165233e-01 1.15950954e+00 -1.14994097e+00 -9.36177850e-01
-2.65414745e-01 4.60638374e-01 -4.77137029e-01 3.73731911e-01
-1.37877703e+00 1.58635509e+00 4.88528162e-01 -3.20541620e-01
-3.29144776e-01 -7.15563774e-01 -3.75556111e-01 -1.63805202e-01
1.61143783e-02 -7.65217304e-01 1.03493202e+00 -3.18972200e-01
-1.45726740e+00 1.21055210e+00 -7.09810972e-01 -4.27860051e-01
-1.75990418e-01 3.71006101e-01 -1.39711514e-01 1.19809180e-01
-2.87875414e-01 3.28404516e-01 -1.48154140e-01 -9.00513589e-01
-2.77380705e-01 -6.13199234e-01 -3.56747792e-03 2.68679857e-01
4.29409057e-01 3.46959978e-01 3.43209773e-01 -2.32247576e-01
1.98328361e-01 -8.00341964e-01 -4.62812722e-01 7.96299800e-03
-6.78857788e-02 -3.69245470e-01 4.82890278e-01 -4.93086666e-01
8.43785703e-01 -1.66777372e+00 5.04541874e-01 -1.38419896e-01
3.80486310e-01 2.55263180e-01 3.61628145e-01 1.18214369e+00
-5.70536315e-01 1.60803825e-01 -3.49386364e-01 6.85106397e-01
-1.89682499e-01 8.50689486e-02 -5.12298495e-02 6.43010199e-01
1.04707070e-02 1.06122172e+00 -6.19725287e-01 3.17623727e-02
4.24086154e-01 7.30709553e-01 -2.73459107e-01 -6.69774190e-02
-2.53280252e-01 4.93148506e-01 -3.81303042e-01 6.26461089e-01
7.79302776e-01 -4.17801499e-01 7.23319888e-01 -2.02901185e-01
-4.76134151e-01 5.21310329e-01 -5.59384644e-01 1.24336839e+00
3.22622091e-01 7.25043118e-02 3.14412117e-02 -1.32356071e+00
8.37727189e-01 3.97218585e-01 7.82937229e-01 -3.15989316e-01
-2.70575076e-01 2.51151264e-01 5.25294006e-01 -4.17622030e-01
-1.34754434e-01 -5.85166037e-01 3.41671705e-01 1.87408254e-01
-1.15621835e-01 1.24878548e-02 3.92389029e-01 5.46668433e-02
1.21261632e+00 7.38947809e-01 8.50340426e-01 -6.32917047e-01
8.78730536e-01 4.48532462e-01 4.16281372e-01 1.97174132e-01
-6.25199914e-01 1.69384256e-01 8.92279863e-01 -6.60378039e-01
-1.39068270e+00 -8.02883148e-01 -5.07552803e-01 1.30986035e+00
2.08699763e-01 -8.42734993e-01 -1.16905832e+00 1.68565422e-01
-1.32164374e-01 2.35842858e-02 -4.16099399e-01 -1.86857685e-01
-5.41748166e-01 -1.58100998e+00 1.71604097e-01 -3.78620513e-02
-1.81567371e-01 -1.17695057e+00 -6.09583795e-01 4.78550851e-01
1.60974726e-01 -7.44475424e-01 1.24177277e-01 8.47580492e-01
-1.08956933e+00 -1.48460162e+00 -4.49734956e-01 -5.27269781e-01
2.00433344e-01 2.42621422e-01 1.13611019e+00 2.59885907e-01
-5.65920889e-01 -1.19888723e-01 -1.38119966e-01 -4.39882487e-01
-6.86550856e-01 -2.41836667e-01 2.33559832e-01 -7.85263419e-01
9.36916471e-01 -9.13141966e-01 -7.21510649e-01 2.71452993e-01
-6.58086956e-01 2.40247902e-02 5.35846114e-01 8.49190235e-01
8.60505879e-01 -2.54766941e-01 3.33388627e-01 -9.35816169e-01
6.21735334e-01 -1.37382627e-01 -3.33516121e-01 5.52200973e-01
-8.00964296e-01 3.06053877e-01 6.40381634e-01 9.02100056e-02
-7.81593859e-01 2.92690575e-01 -5.66565275e-01 6.06903315e-01
-4.75653827e-01 3.04000974e-01 -4.20992583e-01 -2.35198081e-01
8.32098961e-01 5.44934630e-01 4.26391989e-01 -5.30765772e-01
1.54395863e-01 2.18181774e-01 -3.15467268e-02 -7.21987724e-01
3.57354671e-01 3.00797015e-01 3.08769554e-01 -1.00557137e+00
-4.16713774e-01 -4.81302172e-01 -9.34121966e-01 1.53502896e-01
9.07619834e-01 -3.09413075e-01 -1.41521978e+00 2.87579119e-01
-9.95264113e-01 -1.59600943e-01 1.12476073e-01 2.71977365e-01
-1.12091684e+00 8.87374282e-01 -7.91749358e-01 -6.27847254e-01
-3.67700368e-01 -1.54257536e+00 1.12603438e+00 -1.54017648e-02
-2.56754428e-01 -1.08562446e+00 3.09161723e-01 4.43538040e-01
1.83679581e-01 4.82443303e-01 1.27629089e+00 -3.42434525e-01
-2.93132991e-01 3.85063976e-01 2.56010473e-01 -1.46718010e-01
1.11640789e-01 2.77035981e-01 -6.07749164e-01 -1.98094189e-01
6.12053871e-02 -4.38379258e-01 7.44233489e-01 5.17860591e-01
8.65659475e-01 1.30107105e-01 -6.47975028e-01 2.38687962e-01
1.49049425e+00 4.46125209e-01 9.15875673e-01 4.09655601e-01
2.75198400e-01 8.93635809e-01 7.80318201e-01 2.38164281e-03
-2.57039964e-01 7.19076455e-01 2.50201315e-01 7.59969056e-02
1.17091507e-01 1.06316499e-01 4.88140225e-01 6.99772477e-01
-8.09464872e-01 -3.11213937e-02 -8.77934277e-01 -1.04935728e-01
-1.80801153e+00 -1.14584005e+00 -5.52138090e-01 2.21909189e+00
1.24354482e+00 6.22515902e-02 4.52578902e-01 -5.85177680e-03
6.26498699e-01 -3.37854177e-01 -9.28083301e-01 -7.38614142e-01
-2.58113235e-01 3.80354464e-01 5.43975472e-01 5.27486861e-01
-7.61564672e-01 9.12128329e-01 8.20287228e+00 7.62975931e-01
-9.64947939e-01 -1.90534536e-02 4.17018056e-01 2.27814242e-01
7.45973513e-02 4.15190935e-01 -8.00036371e-01 3.73225480e-01
1.59601700e+00 -1.92850962e-01 2.57163048e-01 7.29310215e-01
7.97471881e-01 -2.60063708e-01 -9.05999422e-01 6.14942849e-01
-5.39039731e-01 -1.42016757e+00 -3.14596325e-01 4.71681416e-01
1.84850782e-01 6.53214753e-02 -4.62674648e-01 -8.01316723e-02
1.03989780e-01 -1.10240388e+00 8.77867192e-02 5.88887513e-01
4.01973784e-01 -6.77587092e-01 5.41738391e-01 5.61660230e-01
-1.00915825e+00 5.26724875e-01 -9.70613658e-01 -2.82486945e-01
2.67334074e-01 7.06579208e-01 -3.73624831e-01 2.96852589e-01
5.46065867e-01 6.10149443e-01 -1.38752505e-01 7.47534513e-01
-6.39907420e-02 3.15808862e-01 -1.24733515e-01 -4.38008577e-01
-1.13629825e-01 -7.63150811e-01 1.09400980e-01 1.01350260e+00
-1.66652471e-01 6.09093070e-01 -9.88428071e-02 7.47297347e-01
3.62787127e-01 3.30376863e-01 -2.96598941e-01 -2.75381178e-01
1.54949510e-02 1.20073509e+00 -6.93205655e-01 -2.10256204e-01
-6.35601103e-01 9.01257813e-01 3.29312593e-01 2.45891094e-01
-8.20595801e-01 -2.62077183e-01 1.25683343e+00 2.98828393e-01
7.36938640e-02 -2.25374624e-01 1.40527084e-01 -9.59813297e-01
-3.09931457e-01 -8.46666813e-01 -3.42085585e-02 -7.76881516e-01
-1.10338414e+00 -1.81991220e-01 -2.70011365e-01 -3.53061080e-01
1.14678450e-01 -1.34363818e+00 -2.15681270e-01 1.05046344e+00
-1.16164064e+00 -6.84673667e-01 4.12004411e-01 -4.21663597e-02
1.91570729e-01 3.90408307e-01 1.14339828e+00 -1.07243791e-01
-4.39218879e-01 1.37095302e-01 9.46729720e-01 -6.32735312e-01
7.23492265e-01 -1.47607493e+00 4.82480675e-01 1.38690129e-01
-6.08865321e-01 1.22079849e+00 1.18969047e+00 -6.47835433e-01
-1.60179973e+00 -5.22604942e-01 9.55350637e-01 -7.50765145e-01
6.74066186e-01 -3.22837830e-01 -1.10874844e+00 3.36819172e-01
1.45425186e-01 -5.08489430e-01 1.19609225e+00 1.46574944e-01
2.81654447e-02 4.66665059e-01 -1.22446620e+00 3.64257395e-01
1.00494814e+00 -3.30121279e-01 -5.64611733e-01 7.83614039e-01
4.77609843e-01 -2.93520957e-01 -1.46334529e+00 3.83470118e-01
8.52494299e-01 -1.29030824e+00 1.20298934e+00 -1.03074050e+00
2.34482870e-01 -3.97800654e-01 1.34635374e-01 -7.89647400e-01
-6.01580679e-01 -6.01534307e-01 -2.28112206e-01 4.48021382e-01
5.59827149e-01 -6.81767404e-01 8.99857998e-01 7.62630820e-01
-9.91005003e-02 -1.24507141e+00 -6.63309753e-01 -6.43958449e-01
6.95144594e-01 2.10819349e-01 2.79260457e-01 6.83227181e-01
7.65072465e-01 4.86911654e-01 -7.75957629e-02 -4.86519754e-01
5.22745192e-01 1.29118919e-01 4.73747492e-01 -1.20564973e+00
-3.68376076e-01 -2.96908677e-01 -7.41424918e-01 -9.59609091e-01
-1.64684340e-01 -6.48948729e-01 -3.57832164e-01 -1.47174394e+00
6.18903339e-01 4.67191517e-01 -1.26472399e-01 1.06774621e-01
-9.78507251e-02 -1.27869118e-02 -3.62106442e-01 3.11603576e-01
-3.87004554e-01 1.27837807e-01 1.34155548e+00 2.89973527e-01
9.68984514e-02 -2.63132840e-01 -9.00130272e-01 4.81845737e-01
1.00362611e+00 -3.00166517e-01 -3.05965692e-01 5.95214009e-01
3.81162316e-01 -1.21854529e-01 5.84557420e-03 -6.03429735e-01
-2.27374569e-01 -6.48348629e-01 3.41229707e-01 -6.33832633e-01
3.20112944e-01 -5.06469131e-01 3.72139990e-01 8.42957377e-01
-9.32167247e-02 -3.57072204e-02 4.41505499e-02 5.90260267e-01
1.71631142e-01 -3.24130505e-01 1.19404018e+00 -6.00237012e-01
-4.11026716e-01 1.53281987e-01 -8.79978538e-01 -2.39835575e-01
1.30789101e+00 -5.90098739e-01 -3.25640589e-01 6.95479438e-02
-1.28161883e+00 -4.86869477e-02 9.89499390e-01 -2.44891927e-01
9.07521546e-02 -7.19860613e-01 -1.75771087e-01 -3.61119390e-01
1.72173321e-01 -7.44297028e-01 2.24176839e-01 1.06691110e+00
-1.06660867e+00 1.21853578e+00 -1.60854131e-01 -4.91086572e-01
-1.50149095e+00 1.04993582e+00 7.60160863e-01 -1.05926529e-01
-1.87026799e-01 4.56794143e-01 4.60329443e-01 -5.71285546e-01
-3.15201938e-01 -3.82315144e-02 -2.38665029e-01 -3.96620393e-01
5.33545136e-01 2.16237187e-01 2.19552740e-01 -7.96281695e-01
-4.15726721e-01 7.58433580e-01 -5.70393167e-02 5.69999874e-01
1.33032477e+00 -3.69165987e-01 -7.57262230e-01 3.97439927e-01
9.51218128e-01 -3.99211794e-01 -8.86421025e-01 2.02850670e-01
2.27581874e-01 1.59876838e-01 -5.43958664e-01 -8.29624832e-01
-1.58501253e-01 9.18831408e-01 6.41426921e-01 2.55039692e-01
8.78474772e-01 2.93734580e-01 7.52254069e-01 6.95876896e-01
3.46826762e-01 -1.06635749e+00 -4.50816572e-01 4.67292488e-01
5.42493701e-01 -8.94676685e-01 2.12825418e-01 -6.23692930e-01
-1.30962119e-01 1.22395301e+00 4.18157429e-01 1.54515743e-01
6.06909037e-01 2.83214867e-01 -1.88546613e-01 -4.71225441e-01
-1.10279286e+00 -3.52514356e-01 -2.16794312e-01 7.27739215e-01
1.38833654e+00 3.25261801e-02 -1.25144088e+00 2.83517897e-01
-2.37046257e-01 -1.69865057e-01 5.33756137e-01 1.12689936e+00
-1.10839891e+00 -1.93063354e+00 -5.10803461e-01 2.07147673e-01
-8.03346097e-01 -2.01913610e-01 -1.14420879e+00 4.88272041e-01
4.84321788e-02 8.21645200e-01 -8.08527291e-01 4.59328517e-02
2.18661606e-01 6.41128540e-01 8.56840551e-01 -4.52820092e-01
-2.51849502e-01 7.04985252e-03 9.41680446e-02 -4.90616471e-01
-9.15528357e-01 -7.63439894e-01 -1.66943312e+00 -8.80455136e-01
-6.30600929e-01 6.96192920e-01 7.38923967e-01 9.72794354e-01
4.82905716e-01 3.45227003e-01 -1.00637740e-02 -6.41755700e-01
-2.70451546e-01 -7.19320178e-01 -8.65830243e-01 2.23632917e-01
1.75336879e-02 -7.06501484e-01 -1.23366565e-01 3.54937881e-01] | [4.7344512939453125, 5.279197692871094] |
e90df107-201b-4904-ace9-b3e1bc69e6f8 | bridging-the-gap-point-clouds-for-merging | 2112.02039 | null | https://arxiv.org/abs/2112.02039v2 | https://arxiv.org/pdf/2112.02039v2.pdf | Bridging the Gap: Point Clouds for Merging Neurons in Connectomics | In the field of Connectomics, a primary problem is that of 3D neuron segmentation. Although deep learning-based methods have achieved remarkable accuracy, errors still exist, especially in regions with image defects. One common type of defect is that of consecutive missing image sections. Here, data is lost along some axis, and the resulting neuron segmentations are split across the gap. To address this problem, we propose a novel method based on point cloud representations of neurons. We formulate the problem as a classification problem and train CurveNet, a state-of-the-art point cloud classification model, to identify which neurons should be merged. We show that our method not only performs strongly but also scales reasonably to gaps well beyond what other methods have attempted to address. Additionally, our point cloud representations are highly efficient in terms of data, maintaining high performance with an amount of data that would be unfeasible for other methods. We believe that this is an indicator of the viability of using point cloud representations for other proofreading tasks. | ['Jingpeng Wu', 'Dmitri B. Chklovskii', 'Jules Berman'] | 2021-12-03 | null | null | null | null | ['point-cloud-classification'] | ['computer-vision'] | [ 2.18764409e-01 1.29797742e-01 1.88302591e-01 -1.24057807e-01
-7.52498686e-01 -6.69037104e-01 3.17699641e-01 4.70737576e-01
-4.48828369e-01 5.53658724e-01 -3.33184570e-01 -4.81354356e-01
5.19352779e-02 -6.24638557e-01 -1.12367058e+00 -5.32405496e-01
2.12293491e-01 6.66821897e-01 3.75464231e-01 -4.88544144e-02
6.22345984e-01 7.52117395e-01 -1.33407891e+00 1.75933689e-01
8.42015088e-01 7.92797923e-01 2.22799242e-01 2.85312712e-01
-4.16236728e-01 1.27661213e-01 -6.33984864e-01 -4.74470615e-01
2.38077953e-01 -1.47445530e-01 -9.72413898e-01 4.52048369e-02
6.63993180e-01 -3.85466665e-01 3.72106135e-02 1.09024370e+00
2.56839693e-01 -2.17390671e-01 7.58108318e-01 -1.36337221e+00
-3.82523984e-01 1.74972013e-01 -7.25263715e-01 3.38533461e-01
-2.04475686e-01 6.95137158e-02 7.58605063e-01 -7.76100516e-01
6.48887634e-01 9.86589909e-01 8.98304045e-01 4.62302864e-01
-1.48058426e+00 -6.81813776e-01 2.11206824e-01 -1.08927161e-01
-1.12984359e+00 -2.16842264e-01 6.47987425e-01 -7.56613135e-01
1.07899106e+00 -1.76662896e-02 9.53636050e-01 7.01384366e-01
1.36559308e-01 6.54127598e-01 8.86925042e-01 -1.86694652e-01
3.90176237e-01 -3.99807036e-01 1.44153595e-01 5.64667106e-01
4.73532617e-01 -1.88980460e-01 -1.05224222e-01 -3.49998251e-02
1.13665783e+00 2.69484371e-01 -1.25238914e-02 -4.65225279e-01
-1.31936073e+00 6.89313889e-01 6.82326078e-01 3.65057856e-01
-1.64089605e-01 4.49955553e-01 2.11211786e-01 -6.23013116e-02
6.23714268e-01 6.01585329e-01 -5.57692647e-01 -4.66022044e-02
-1.40663660e+00 6.08581424e-01 2.93571681e-01 9.38177168e-01
5.88116884e-01 -2.13332579e-01 3.69285285e-01 6.98981881e-01
1.83727816e-01 -3.93636078e-02 1.25423297e-01 -1.29671562e+00
3.73297006e-01 8.90499234e-01 -1.06737822e-01 -1.03078473e+00
-6.98246360e-01 -4.94726896e-01 -6.20289981e-01 7.24525392e-01
9.25423205e-01 9.82430279e-02 -1.13569355e+00 1.43753910e+00
1.16253585e-01 1.24914773e-01 -3.43373209e-01 8.85616541e-01
4.99380201e-01 4.06448692e-01 -1.25843510e-02 1.32897645e-01
1.14628065e+00 -6.22938931e-01 -3.16503495e-01 -1.31089166e-01
6.91358864e-01 -5.31027734e-01 7.53949344e-01 5.59623420e-01
-1.34544110e+00 -2.33845308e-01 -1.10268629e+00 -4.07569289e-01
-3.68769079e-01 5.31046912e-02 7.22589433e-01 3.51546139e-01
-1.19986856e+00 1.06712019e+00 -1.04853475e+00 -4.75401431e-01
1.14312673e+00 5.00419557e-01 -4.23170835e-01 2.10411221e-01
-4.67600435e-01 1.00327098e+00 1.89826712e-01 -5.61031736e-02
-5.35931766e-01 -9.23747063e-01 -4.14170563e-01 2.27450341e-01
-1.15104765e-01 -6.33675039e-01 1.14851642e+00 -8.32994044e-01
-7.08609581e-01 1.13837349e+00 -8.56737792e-02 -4.32274878e-01
5.37755072e-01 1.73912942e-01 1.86436966e-01 3.79994035e-01
2.51738876e-01 1.22818136e+00 4.57401305e-01 -1.39549875e+00
-5.99795341e-01 -9.02423322e-01 -1.67137519e-01 -1.68938227e-02
1.59788355e-02 -1.66870967e-01 -3.61704260e-01 -5.45244992e-01
6.80150867e-01 -8.78751636e-01 -3.02565068e-01 7.15467274e-01
-2.08829567e-01 -1.97583541e-01 5.73793292e-01 -6.54165626e-01
4.15544182e-01 -2.03982997e+00 -8.49109888e-02 5.92828207e-02
6.38263881e-01 2.78378636e-01 -2.98175775e-02 5.28814830e-02
-8.15043226e-02 6.66891575e-01 -5.57929337e-01 -3.09253603e-01
-4.60298330e-01 2.59854198e-02 -1.28697544e-01 6.30110979e-01
4.18916345e-01 8.27038229e-01 -7.33491778e-01 -4.20825779e-01
-7.14168996e-02 4.09824669e-01 -5.44957161e-01 -5.32592535e-01
-2.96082407e-01 4.71312672e-01 -2.86752820e-01 6.47109866e-01
8.26585710e-01 -2.57102519e-01 -6.65149689e-02 -5.22014275e-02
-2.79809207e-01 1.72209144e-01 -8.09852004e-01 2.09995961e+00
1.50723651e-01 7.84990847e-01 1.50691390e-01 -1.22621083e+00
8.56570244e-01 1.28632382e-01 8.79035890e-01 -4.43943024e-01
1.52825773e-01 4.16469693e-01 1.41759902e-01 -3.06662083e-01
2.52607584e-01 -2.20129550e-01 2.90158182e-01 2.91593552e-01
1.67121977e-01 -3.96270007e-01 2.48581693e-02 8.75170082e-02
1.14732492e+00 3.66948456e-01 -2.16227159e-01 -3.81583512e-01
-2.85476714e-01 5.40188015e-01 4.78687733e-01 4.61215258e-01
-4.69202280e-01 1.14990163e+00 8.52588832e-01 -3.13723952e-01
-1.58244669e+00 -9.70770061e-01 -3.02729219e-01 4.78357941e-01
1.83771878e-01 -1.36508003e-01 -9.38702166e-01 -4.67893362e-01
2.02398196e-01 4.72481132e-01 -4.64536607e-01 7.71514773e-02
-3.96110266e-01 -5.02131939e-01 6.44696712e-01 6.78149164e-01
2.17278078e-01 -8.03587258e-01 -8.19193482e-01 3.68805289e-01
-6.24938644e-02 -9.94960725e-01 8.77337232e-02 2.29744136e-01
-1.48562300e+00 -1.12642539e+00 -1.06867170e+00 -8.71142507e-01
8.89631987e-01 5.20271719e-01 9.56088483e-01 3.95619452e-01
-3.80367219e-01 7.45391920e-02 -1.48396969e-01 -7.68199086e-01
-5.50523587e-02 2.43123751e-02 -3.60482931e-01 -6.05899155e-01
3.70121181e-01 -7.63425231e-01 -6.76683664e-01 7.91464970e-02
-8.52477431e-01 5.26052304e-02 4.33513016e-01 6.88240349e-01
9.12464857e-01 1.97526556e-03 5.58243752e-01 -7.72300363e-01
3.77898663e-01 -3.49568784e-01 -6.29520893e-01 -1.30387858e-01
-6.54299736e-01 -1.99411258e-01 3.49824935e-01 -9.88004729e-02
-4.19780672e-01 5.90587631e-02 -2.35722497e-01 -4.66742545e-01
-3.57597172e-01 2.78583407e-01 1.92711353e-01 -2.08256245e-01
6.56157434e-01 -2.47671157e-02 4.35886562e-01 -6.02329910e-01
8.48554745e-02 3.94401848e-01 4.62131768e-01 -3.24479818e-01
4.49178517e-01 8.75828862e-01 2.03696817e-01 -7.62069762e-01
-5.21494806e-01 -5.31208515e-01 -9.29003060e-01 -8.78249332e-02
9.53426480e-01 -7.21280158e-01 -6.41125739e-01 4.02097940e-01
-1.44837201e+00 -3.64273578e-01 -1.41939536e-01 2.62338102e-01
-7.02511549e-01 3.53993058e-01 -6.58684134e-01 -5.10168433e-01
-1.45896137e-01 -1.24901330e+00 1.18325531e+00 1.68740645e-01
-1.02382787e-01 -7.36828864e-01 -2.12896898e-01 4.73757029e-01
8.92568976e-02 4.32027102e-01 1.20233703e+00 -5.79069138e-01
-6.75559759e-01 -1.76550001e-01 -3.77764940e-01 1.50275463e-03
-2.22541675e-01 1.93372637e-01 -8.52985561e-01 -1.51430190e-01
-1.02891035e-01 -3.21316034e-01 1.04015517e+00 6.00286901e-01
1.61003745e+00 1.46251470e-01 -7.17551768e-01 6.17583692e-01
1.37822747e+00 2.32576728e-01 7.25425661e-01 4.45023477e-01
5.55338860e-01 7.27978766e-01 1.19703531e-01 -8.59340951e-02
3.41108501e-01 5.39645970e-01 6.89966738e-01 -5.55211008e-01
-1.58040851e-01 -1.81438431e-01 -3.85191232e-01 2.71326602e-01
8.92101415e-03 -2.80883797e-02 -1.20159769e+00 6.82135403e-01
-1.75120354e+00 -9.31282282e-01 -5.34169614e-01 1.88760757e+00
5.15005887e-01 3.68961811e-01 3.80659074e-01 3.98627788e-01
5.52352071e-01 -3.86137038e-01 -7.86865652e-01 -3.29015702e-01
1.85859611e-03 1.53509691e-01 5.60700059e-01 -2.94882935e-02
-9.42049623e-01 9.05015945e-01 6.95236158e+00 4.17430550e-01
-1.10256946e+00 -1.76105034e-02 7.64012873e-01 -8.25203583e-02
-3.42318714e-01 7.93017074e-02 -5.18832028e-01 5.26527524e-01
6.29767537e-01 3.22902232e-01 4.14063007e-01 5.77077389e-01
2.43153766e-01 -2.27071211e-01 -1.09060478e+00 1.04404581e+00
-1.36533648e-01 -1.70304525e+00 -8.23540464e-02 2.91203409e-01
5.25760114e-01 2.57609159e-01 -8.93198512e-03 -6.83216378e-02
3.34462849e-03 -1.24462211e+00 8.31215739e-01 3.26578230e-01
7.02264369e-01 -6.49271667e-01 4.28003281e-01 5.06879628e-01
-5.81356227e-01 1.00991890e-01 -7.64322519e-01 -7.98379034e-02
-4.36065979e-02 7.11967170e-01 -7.91476429e-01 2.29960620e-01
8.80912125e-01 5.78975916e-01 -6.92288280e-01 1.69095695e+00
2.42455695e-02 4.59333241e-01 -4.61594969e-01 1.45424500e-01
2.16639623e-01 -2.42376834e-01 3.42378736e-01 8.62407267e-01
4.83507484e-01 6.47590756e-02 -1.84774324e-01 1.40159035e+00
-1.49271101e-01 -1.07644469e-01 -8.15769374e-01 -1.22662865e-01
4.41877335e-01 1.23860216e+00 -1.27101624e+00 -1.60540715e-01
-4.47517991e-01 7.62386560e-01 6.06841624e-01 2.00208426e-01
-4.14708376e-01 -1.12500638e-01 6.52626157e-01 2.87062019e-01
3.79489183e-01 -5.19309342e-01 -9.91747200e-01 -9.13052619e-01
8.26078057e-02 -5.45107007e-01 -1.40957031e-02 -9.67557669e-01
-1.16799903e+00 1.59653381e-01 -2.56498903e-01 -1.09982991e+00
1.89873189e-01 -8.14618647e-01 -5.74251175e-01 7.38768578e-01
-1.33261025e+00 -1.03781140e+00 -1.82754412e-01 2.81287670e-01
3.71493071e-01 2.10388154e-01 5.84446490e-01 2.88335145e-01
-3.13048124e-01 3.26676607e-01 2.24540800e-01 1.67218536e-01
3.34337354e-01 -1.21385419e+00 6.69431686e-01 6.15454495e-01
2.48141930e-01 5.51673472e-01 5.37241817e-01 -7.05700040e-01
-1.04946911e+00 -8.81723642e-01 6.40038192e-01 -4.77863729e-01
3.60154122e-01 -5.56922853e-01 -1.03155637e+00 7.89630413e-01
-2.67334878e-01 3.06575093e-02 5.54541051e-01 1.06391676e-01
-2.16470137e-01 1.58649191e-01 -1.19916570e+00 6.12491667e-01
1.09971941e+00 -1.48935661e-01 -5.80459356e-01 4.23451275e-01
5.39171159e-01 -4.32323962e-01 -7.73051262e-01 1.88309401e-01
5.02745867e-01 -9.50523734e-01 1.00253403e+00 -5.92218876e-01
6.57231867e-01 -2.82393008e-01 1.29836455e-01 -1.37547481e+00
-3.61200958e-01 -1.48220435e-02 2.75384843e-01 1.14664412e+00
5.54540038e-01 -5.08465588e-01 1.24717844e+00 6.81551576e-01
-5.54202616e-01 -8.45952988e-01 -1.10515308e+00 -8.51986051e-01
9.37777102e-01 -2.65450656e-01 6.07842088e-01 8.83239806e-01
7.97165260e-02 1.80128634e-01 4.11346644e-01 -1.26393899e-01
7.47247696e-01 4.68122140e-02 5.25379121e-01 -1.64315617e+00
3.03359479e-01 -8.59549582e-01 -7.04874992e-01 -9.31790113e-01
1.20943531e-01 -1.15160990e+00 -4.02808702e-03 -2.03549910e+00
1.83190838e-01 -6.72050416e-01 2.00990457e-02 6.45471096e-01
5.27347624e-03 3.13278019e-01 3.41335177e-01 5.26106358e-01
-2.27266699e-01 2.13801101e-01 1.44241571e+00 -2.65012622e-01
-2.79662330e-02 -1.84608892e-01 -8.32915366e-01 8.30820978e-01
9.48682845e-01 -4.50305670e-01 -1.97550461e-01 -7.90231228e-01
-7.27792084e-02 -5.52750900e-02 7.54779935e-01 -1.16557682e+00
2.37065628e-01 1.45735025e-01 9.07061279e-01 -9.56127405e-01
3.88541758e-01 -8.08297753e-01 3.16142254e-02 2.74044991e-01
-3.56239170e-01 5.01589589e-02 3.40808988e-01 3.95288050e-01
4.76542823e-02 -2.90408194e-01 7.89877057e-01 -4.26925212e-01
-3.34129184e-01 3.90149325e-01 -3.59950215e-01 -8.47619995e-02
1.00162733e+00 -5.31710565e-01 -5.75857103e-01 2.11117845e-02
-6.57253683e-01 1.59826711e-01 1.00811255e+00 -4.24560159e-02
7.11085498e-01 -9.69063818e-01 -4.41393554e-01 7.82283321e-02
-6.38976544e-02 4.49887246e-01 -4.77285720e-02 8.68814766e-01
-8.19482625e-01 4.64828283e-01 -5.13492405e-01 -8.49642873e-01
-9.01080370e-01 5.60619712e-01 4.31914657e-01 2.06569120e-01
-1.01858783e+00 7.99810290e-01 1.06407739e-01 -3.37539405e-01
1.01868235e-01 -5.43741524e-01 -2.02476725e-01 5.29921465e-02
1.73911333e-01 2.67829150e-01 4.40975159e-01 -4.55327928e-01
-3.67953777e-01 5.25374532e-01 -4.35495526e-02 -5.45877479e-02
1.51159489e+00 9.14966464e-02 -7.60706812e-02 3.71657252e-01
1.16785693e+00 -5.12149334e-01 -1.56405354e+00 3.05762261e-01
-3.87175158e-02 -3.22843313e-01 1.49866700e-01 -7.62180984e-01
-1.08955729e+00 1.30799878e+00 6.29996121e-01 2.27417216e-01
6.89063668e-01 6.42043650e-02 9.10011411e-01 -3.48133296e-02
4.01368678e-01 -9.95925605e-01 -6.07314184e-02 3.52552682e-01
7.65441656e-01 -1.18402016e+00 1.49873972e-01 -3.81848603e-01
-2.32885361e-01 1.20802188e+00 6.08928859e-01 -3.68477970e-01
4.83613610e-01 1.98493972e-01 -6.13967739e-02 -5.25500894e-01
-5.36451161e-01 -2.35868350e-01 -1.30824581e-01 9.41431582e-01
4.49487746e-01 -1.77653909e-01 -2.26189092e-01 3.00984293e-01
-2.26717710e-01 1.72113478e-01 5.82670689e-01 9.64626253e-01
-6.48896873e-01 -8.47163677e-01 -3.75958294e-01 8.50490808e-01
-6.12298012e-01 5.31715415e-02 -5.76681554e-01 7.90499687e-01
2.08049923e-01 8.13164949e-01 5.22288620e-01 -1.41440704e-01
3.13632250e-01 -1.42096905e-02 7.25944221e-01 -6.26246452e-01
-4.96904105e-01 2.20760360e-01 -3.57588351e-01 -4.71632510e-01
-1.73884064e-01 -7.93460488e-01 -1.73780930e+00 -2.61210024e-01
-2.80435741e-01 -3.20213586e-01 1.11130393e+00 9.85724568e-01
4.62701768e-01 4.73278373e-01 8.98097306e-02 -1.01241100e+00
-3.84896696e-01 -5.73851883e-01 -5.91772377e-01 4.77569133e-01
2.53467381e-01 -9.31503236e-01 -1.45742819e-01 2.83070444e-03] | [14.259784698486328, -3.1247987747192383] |
94507a7f-a204-4a7e-9f4b-ca600b3e8dd8 | twitter-sentiment-analysis-lexicon-method | 1507.00955 | null | http://arxiv.org/abs/1507.00955v3 | http://arxiv.org/pdf/1507.00955v3.pdf | Twitter Sentiment Analysis: Lexicon Method, Machine Learning Method and Their Combination | This paper covers the two approaches for sentiment analysis: i) lexicon based
method; ii) machine learning method. We describe several techniques to
implement these approaches and discuss how they can be adopted for sentiment
classification of Twitter messages. We present a comparative study of different
lexicon combinations and show that enhancing sentiment lexicons with emoticons,
abbreviations and social-media slang expressions increases the accuracy of
lexicon-based classification for Twitter. We discuss the importance of feature
generation and feature selection processes for machine learning sentiment
classification. To quantify the performance of the main sentiment analysis
methods over Twitter we run these algorithms on a benchmark Twitter dataset
from the SemEval-2013 competition, task 2-B. The results show that machine
learning method based on SVM and Naive Bayes classifiers outperforms the
lexicon method. We present a new ensemble method that uses a lexicon based
sentiment score as input feature for the machine learning approach. The
combined method proved to produce more precise classifications. We also show
that employing a cost-sensitive classifier for highly unbalanced datasets
yields an improvement of sentiment classification performance up to 7%. | ['Tomaso Aste', 'Tharsis T. P. Souza', 'Olga Kolchyna', 'Philip Treleaven'] | 2015-07-03 | null | null | null | null | ['twitter-sentiment-analysis'] | ['natural-language-processing'] | [ 1.72599480e-02 -1.16740294e-01 -7.60861412e-02 -7.30579674e-01
-5.43026865e-01 -6.68399990e-01 8.37550104e-01 8.01785231e-01
-6.76358819e-01 7.66229272e-01 5.01146495e-01 -2.75201619e-01
7.46645108e-02 -8.96190763e-01 1.11553133e-01 -4.98731166e-01
3.99340749e-01 4.38837200e-01 1.10541321e-01 -9.87628162e-01
8.04896414e-01 2.91202605e-01 -2.08182764e+00 7.46967673e-01
3.84577841e-01 1.22623205e+00 -4.80173856e-01 6.14195049e-01
-5.81681073e-01 9.15381551e-01 -8.46067071e-01 -6.42791092e-01
1.03759632e-01 -2.25302473e-01 -6.45565748e-01 -9.75729674e-02
-5.19966744e-02 6.50941908e-01 8.44075680e-01 7.92730510e-01
6.71269476e-01 -4.18176278e-02 1.13890243e+00 -1.13704848e+00
-5.68233989e-02 6.16972744e-01 -5.54246128e-01 2.40848169e-01
7.63569832e-01 -5.70588350e-01 7.87567973e-01 -8.07973444e-01
6.29251182e-01 9.04634535e-01 1.02726591e+00 2.46910509e-02
-6.73494637e-01 -7.07887232e-01 -3.89734879e-02 -9.74369422e-02
-8.75902593e-01 -2.13375464e-01 6.40203118e-01 -6.91386521e-01
1.12502635e+00 4.33110237e-01 6.62658811e-01 5.09088993e-01
3.16297829e-01 4.01380599e-01 1.86008132e+00 -7.86798537e-01
2.31996268e-01 1.18702316e+00 7.11076736e-01 4.65173751e-01
3.38140845e-01 -3.23310673e-01 -7.20504284e-01 -3.10673952e-01
-5.20998716e-01 -2.30194733e-01 4.07561123e-01 2.61886060e-01
-7.00180829e-01 1.23231387e+00 -2.12065838e-02 6.37104750e-01
-4.44981128e-01 -4.61805969e-01 1.08043885e+00 6.96582258e-01
1.09645391e+00 7.33051658e-01 -9.34049368e-01 9.15662106e-03
-8.12625051e-01 2.25340366e-01 1.12428260e+00 4.86102939e-01
6.28983676e-01 -1.31605729e-01 5.92531972e-02 9.29350555e-01
5.06370544e-01 5.56702852e-01 1.02408719e+00 1.41288951e-01
1.58083111e-01 1.13886821e+00 -2.45123636e-02 -1.44798386e+00
-8.65584910e-01 -6.83791116e-02 -1.20985761e-01 1.73199639e-01
9.50047448e-02 -6.53404355e-01 -6.16841376e-01 1.05574071e+00
5.21659195e-01 -5.96658349e-01 4.11738753e-01 1.68494999e-01
1.33230746e+00 5.33913672e-01 4.03199673e-01 -3.48899603e-01
1.59143484e+00 -7.29523718e-01 -9.38154161e-01 2.28379861e-01
1.24154139e+00 -1.29023242e+00 8.16927314e-01 6.14895463e-01
-6.31924391e-01 -3.59533221e-01 -1.03546286e+00 5.13505876e-01
-1.46673381e+00 9.71783772e-02 8.67274821e-01 1.56245399e+00
-8.69548857e-01 4.93632466e-01 -2.37280056e-01 -6.11610532e-01
1.67994261e-01 7.97906339e-01 -5.25321782e-01 7.59729266e-01
-1.22196579e+00 1.22667789e+00 9.05041918e-02 -5.12175322e-01
4.30166066e-01 -2.21210256e-01 -8.65025342e-01 -4.68929887e-01
-3.61298740e-01 -1.37672588e-01 1.13629508e+00 -1.57101142e+00
-1.48207963e+00 1.41376448e+00 -1.30855665e-01 -2.69577265e-01
1.63507760e-01 7.44741857e-02 -6.08478189e-01 -9.31080133e-02
2.69865841e-01 1.49258569e-01 5.41772902e-01 -1.01755571e+00
-8.92603397e-01 -4.33659613e-01 -1.54462025e-01 2.63585657e-01
-6.17311656e-01 5.46961188e-01 3.90332520e-01 -3.90786529e-01
-8.16581100e-02 -1.03627443e+00 -1.53782979e-01 -9.85361338e-01
-2.91644722e-01 -4.90606487e-01 7.73816228e-01 -2.30848446e-01
1.22644949e+00 -1.70552826e+00 -4.53391284e-01 6.21592224e-01
-1.75621554e-01 2.96220660e-01 5.11876404e-01 7.68472075e-01
-2.69110233e-01 3.50140244e-01 1.63651064e-01 -4.22593117e-01
-3.69668528e-02 -1.26023665e-01 -1.73429385e-01 3.73843104e-01
-1.16128981e-01 4.72446948e-01 -6.34297192e-01 -6.59717798e-01
4.23813224e-01 5.04774451e-01 -3.01803559e-01 -1.35548875e-01
1.39430150e-01 1.55665293e-01 -5.71029425e-01 6.40393496e-01
5.18814087e-01 1.45328045e-01 1.47423565e-01 -2.81804115e-01
-2.82645285e-01 2.80897856e-01 -1.08018553e+00 7.47522414e-01
-6.51583850e-01 6.27352118e-01 -6.58499122e-01 -8.13041151e-01
1.45948112e+00 3.22730452e-01 3.96462858e-01 -4.38243717e-01
1.03035069e+00 4.18407559e-01 -3.80846798e-01 -6.41261101e-01
6.12658024e-01 -6.18415892e-01 -4.99650300e-01 5.63066661e-01
1.34867400e-01 -4.27373707e-01 4.12694961e-01 -4.25357558e-02
4.38224047e-01 -6.02012910e-02 7.62419164e-01 -6.54001415e-01
1.12614262e+00 3.87231380e-01 -2.21735254e-01 5.10514498e-01
-3.11337531e-01 1.15223251e-01 6.75748885e-01 -5.66698313e-01
-5.40513873e-01 1.20416016e-03 -2.93272287e-01 1.54184401e+00
-1.25301674e-01 -5.81933200e-01 -7.43731976e-01 -8.23303342e-01
-1.84606567e-01 4.72231358e-01 -9.64828134e-01 1.33846879e-01
-5.85257597e-02 -1.59795177e+00 2.16006473e-01 2.39491463e-02
1.38790622e-01 -1.15750670e+00 -3.81035060e-01 1.09146591e-02
1.59822896e-01 -7.39865124e-01 4.11039084e-01 5.32594919e-01
-8.02965164e-01 -1.00089455e+00 -8.94157216e-02 -7.40773261e-01
4.41226512e-01 -1.27873361e-01 1.05511808e+00 2.35408679e-01
1.86722577e-01 3.22445989e-01 -1.09113085e+00 -1.25727963e+00
-5.89157879e-01 4.89589661e-01 -9.70648602e-02 1.39437959e-01
1.16956067e+00 -2.29442015e-01 -2.01188087e-01 2.58117151e-02
-7.32065082e-01 -3.43791515e-01 7.99645111e-02 4.47546780e-01
1.26064032e-01 4.77501564e-02 9.43182528e-01 -1.73959100e+00
1.06621408e+00 -7.88338900e-01 -2.13542715e-01 -2.07087785e-01
-8.78909945e-01 -4.55328040e-02 4.20937568e-01 5.35675557e-03
-8.51197839e-01 2.27218598e-01 -5.77823997e-01 8.34489465e-01
-3.58310603e-02 6.92694664e-01 3.86911929e-01 -4.46472794e-01
1.07996321e+00 -1.23772040e-01 1.13118887e-01 -2.24513203e-01
-1.38716456e-02 1.18588734e+00 -4.17858809e-01 -1.35841578e-01
2.31864005e-01 5.64595938e-01 -7.67196063e-04 -7.09169626e-01
-1.09178030e+00 -9.26997066e-01 -5.89601040e-01 -5.44777751e-01
6.64602041e-01 -8.80965114e-01 -8.28775227e-01 7.16032088e-01
-7.13211834e-01 2.05378190e-01 1.50363585e-02 2.99587727e-01
-3.24510545e-01 -1.47266120e-01 -2.17905834e-01 -1.05099165e+00
-7.07452834e-01 -9.95637238e-01 1.01020694e+00 3.56605291e-01
-7.40483344e-01 -1.32060266e+00 4.71785992e-01 5.35698771e-01
5.85240781e-01 6.98052406e-01 5.37717760e-01 -1.43608344e+00
7.59918511e-01 -8.71652305e-01 2.37372622e-01 2.83963859e-01
1.12923078e-01 2.23133758e-01 -1.25644767e+00 3.10554951e-01
6.44448251e-02 -3.13432693e-01 7.71491468e-01 1.20334513e-02
5.77111900e-01 -3.40449810e-02 -2.97095388e-01 -4.22534570e-02
1.75468469e+00 2.78200448e-01 4.83067214e-01 9.69452560e-01
2.71195263e-01 1.01454139e+00 9.40176308e-01 4.16298300e-01
4.83928084e-01 3.77429843e-01 1.09363727e-01 1.06824804e-02
3.63375962e-01 2.18847573e-01 3.95122945e-01 9.13418353e-01
-4.57802378e-02 -2.12645590e-01 -9.40571189e-01 3.68477285e-01
-1.63970494e+00 -7.41541743e-01 -7.88621485e-01 1.77684057e+00
8.19307387e-01 3.86264116e-01 4.23693985e-01 8.81863356e-01
5.66457748e-01 -1.75327994e-02 3.15378308e-01 -1.35097063e+00
-3.71112704e-01 6.29115999e-01 5.89323759e-01 5.72126746e-01
-1.35902071e+00 1.07828927e+00 6.30575752e+00 7.10096836e-01
-1.38908601e+00 3.85721862e-01 6.49061501e-01 1.47298053e-01
-6.91989390e-03 -4.36371922e-01 -1.10150814e+00 3.15232098e-01
1.22469318e+00 -1.04995511e-01 -3.98765385e-01 9.83248115e-01
3.53947747e-03 -6.20788157e-01 -2.54063815e-01 6.93523049e-01
4.53727841e-01 -1.21601105e+00 -1.20476233e-02 -3.62123698e-01
1.02515888e+00 1.16815560e-01 -1.23323351e-01 3.84109557e-01
-3.71426381e-02 -8.23569834e-01 6.07911766e-01 3.04100484e-01
3.18115950e-01 -9.41796780e-01 1.37989223e+00 -6.12383820e-02
-8.81983936e-01 1.19085819e-01 -1.98742282e-03 -6.73445106e-01
-3.17958593e-02 6.04473650e-01 -7.06809819e-01 2.71110028e-01
6.62114143e-01 8.97636652e-01 -8.69226933e-01 6.23602808e-01
-3.95063944e-02 4.65903699e-01 -2.99805850e-01 -9.28439617e-01
3.68362218e-01 1.10003557e-02 2.39060268e-01 1.78443933e+00
-7.25959837e-02 -2.63891697e-01 -1.88075930e-01 -3.84987056e-01
4.28879887e-01 1.27658546e+00 -7.63878286e-01 5.22923395e-02
1.00792311e-02 1.67192698e+00 -1.45670843e+00 -6.41887069e-01
-4.44629341e-01 4.08673674e-01 -1.40189156e-01 -3.59415352e-01
-3.03169370e-01 -5.80485702e-01 2.38343000e-01 3.95664759e-02
4.59234007e-02 3.93474609e-01 -8.18544388e-01 -8.69813204e-01
-4.28743899e-01 -9.17688429e-01 4.39750701e-01 -6.05277419e-01
-1.22569335e+00 1.22711635e+00 -1.20042771e-01 -1.03693283e+00
-1.49950609e-01 -1.10832965e+00 -5.10899127e-01 7.31703639e-01
-1.31928968e+00 -1.03360534e+00 -2.14501664e-01 4.04621512e-01
1.95580095e-01 -5.81838369e-01 1.10401201e+00 2.86312789e-01
-1.84839755e-01 2.67392427e-01 -2.68196571e-03 -4.88145612e-02
8.54390502e-01 -1.37367308e+00 -1.39778823e-01 1.92394510e-01
-2.76544511e-01 4.44935292e-01 9.29636419e-01 -5.41083872e-01
-6.98650897e-01 -8.68050516e-01 1.67660904e+00 -9.32890415e-01
8.00490856e-01 -2.63645440e-01 -1.81552023e-01 3.67640585e-01
4.27576751e-01 -6.27745450e-01 1.63811111e+00 1.91227362e-01
-1.80300802e-01 -1.82890985e-02 -1.44797111e+00 2.97534466e-01
-5.13299443e-02 -2.86187261e-01 -6.88131154e-01 6.97129011e-01
1.84244409e-01 -1.58769444e-01 -8.58288169e-01 3.89676183e-01
7.78050840e-01 -1.32562733e+00 5.08746028e-01 -7.05705106e-01
6.45101547e-01 -2.78289944e-01 -2.24189192e-01 -1.23329699e+00
4.95114267e-01 -3.45562518e-01 6.63408697e-01 1.20066452e+00
1.01787043e+00 -8.66263568e-01 6.09458029e-01 2.91590184e-01
3.89564544e-01 -9.47480798e-01 -3.14158767e-01 -1.56900585e-02
1.54676855e-01 -6.44900858e-01 3.90735656e-01 1.11666191e+00
4.39800501e-01 7.78755009e-01 -1.32602841e-01 -5.27077615e-01
-1.02336481e-01 -1.07562039e-02 7.85772860e-01 -1.40090024e+00
3.06255281e-01 -6.10774338e-01 -9.33978558e-01 2.54698634e-01
1.15437262e-01 -8.89883220e-01 -4.29405391e-01 -1.28223729e+00
-4.36577909e-02 -4.65888202e-01 -2.78360546e-01 2.47196004e-01
-1.79967955e-01 9.03720438e-01 1.12385824e-01 -1.98139071e-01
-5.18405199e-01 -1.61640614e-01 7.72701085e-01 2.83877939e-01
-2.17386603e-01 1.96786672e-01 -1.34711003e+00 9.30059731e-01
1.00293362e+00 -9.13269341e-01 -2.45324105e-01 4.26906258e-01
1.15969121e+00 -6.58113718e-01 -3.79283607e-01 -7.85827518e-01
3.12502682e-02 1.88315883e-01 3.08118612e-01 -4.96359318e-01
5.15676439e-02 -7.56966650e-01 -4.24533784e-02 5.50983191e-01
-3.22311342e-01 2.19026893e-01 3.85193914e-01 2.16937676e-01
-3.92685145e-01 -6.57531321e-01 7.20412433e-01 -1.87827334e-01
-5.24433255e-01 -3.43798161e-01 -8.31867576e-01 -2.23629937e-01
1.08269691e+00 -3.96936119e-01 -5.05460240e-02 -4.16444927e-01
-8.15981388e-01 -6.56494647e-02 2.73372799e-01 4.55542415e-01
-2.68451065e-01 -8.67308319e-01 -5.78025162e-01 5.16961962e-02
3.90756637e-01 -9.01454568e-01 -1.83387190e-01 9.58433330e-01
-8.00646424e-01 3.68832350e-01 -2.72495925e-01 -2.59276092e-01
-1.70694840e+00 1.97157875e-01 3.75992447e-01 -4.33816463e-01
3.95360619e-01 9.92215335e-01 -4.55231994e-01 -8.15213561e-01
-1.40072018e-01 -1.80642642e-02 -1.53092086e+00 1.03584254e+00
6.41729295e-01 2.65566081e-01 5.99404156e-01 -1.29091096e+00
-5.41302621e-01 9.08234358e-01 1.38623901e-02 -1.86285630e-01
1.40147865e+00 -5.94507568e-02 -4.35575485e-01 8.84381950e-01
1.26023304e+00 4.67975557e-01 1.47557452e-01 1.77993432e-01
3.52288544e-01 -4.22127210e-02 1.19829185e-01 -9.82976317e-01
-9.24190223e-01 5.08144915e-01 5.73576808e-01 9.21526313e-01
1.24741948e+00 -4.28673238e-01 2.82060534e-01 2.74908364e-01
1.66629165e-01 -1.37023783e+00 -3.37206215e-01 6.75507009e-01
5.32659709e-01 -1.67880177e+00 2.65463203e-01 -5.86958587e-01
-1.00473607e+00 1.43128204e+00 1.67366013e-01 -4.06727493e-01
1.41484749e+00 3.10788780e-01 6.69636548e-01 -5.21217644e-01
-5.98977566e-01 -5.42091131e-01 2.93732464e-01 4.36086506e-01
9.67763841e-01 1.98146626e-01 -1.21153510e+00 8.79027188e-01
-8.32893133e-01 -8.57983530e-02 7.18428671e-01 1.08760917e+00
-5.70662260e-01 -1.34779751e+00 -4.02979761e-01 8.19322288e-01
-1.15801084e+00 -2.39035010e-01 -8.79990399e-01 7.71212757e-01
2.71770239e-01 1.48886847e+00 -1.87566712e-01 -6.81880355e-01
4.71804142e-01 3.06623459e-01 1.11943588e-01 -8.32038343e-01
-1.68968546e+00 -2.17788443e-01 6.37904704e-01 -1.49292201e-01
-1.23985624e+00 -8.14151347e-01 -8.98901820e-01 -4.04602498e-01
-7.28200495e-01 5.36359727e-01 1.44433713e+00 1.18386602e+00
1.02226503e-01 2.35560432e-01 8.70419383e-01 -7.70861149e-01
1.45473361e-01 -1.11939728e+00 -5.79028189e-01 4.80659813e-01
1.81258947e-01 -7.69312441e-01 -6.43299460e-01 2.51233369e-01] | [11.084571838378906, 6.819433212280273] |
a23a8b50-aafa-4b61-a838-e660f7ae125a | aggretriever-a-simple-approach-to-aggregate | 2208.00511 | null | https://arxiv.org/abs/2208.00511v2 | https://arxiv.org/pdf/2208.00511v2.pdf | Aggretriever: A Simple Approach to Aggregate Textual Representations for Robust Dense Passage Retrieval | Pre-trained language models have been successful in many knowledge-intensive NLP tasks. However, recent work has shown that models such as BERT are not ``structurally ready'' to aggregate textual information into a [CLS] vector for dense passage retrieval (DPR). This ``lack of readiness'' results from the gap between language model pre-training and DPR fine-tuning. Previous solutions call for computationally expensive techniques such as hard negative mining, cross-encoder distillation, and further pre-training to learn a robust DPR model. In this work, we instead propose to fully exploit knowledge in a pre-trained language model for DPR by aggregating the contextualized token embeddings into a dense vector, which we call agg*. By concatenating vectors from the [CLS] token and agg*, our Aggretriever model substantially improves the effectiveness of dense retrieval models on both in-domain and zero-shot evaluations without introducing substantial training overhead. Code is available at https://github.com/castorini/dhr | ['Jimmy Lin', 'Minghan Li', 'Sheng-Chieh Lin'] | 2022-07-31 | null | null | null | null | ['passage-retrieval'] | ['natural-language-processing'] | [-6.33611158e-02 -1.44595414e-01 -3.09348524e-01 -2.99509555e-01
-1.59407175e+00 -5.64747930e-01 7.40158677e-01 5.19396007e-01
-7.21050680e-01 6.28707945e-01 7.74689317e-01 -2.90300906e-01
1.12653218e-01 -6.88533247e-01 -7.44388640e-01 -2.40518674e-01
5.20495847e-02 5.34264982e-01 -4.51097675e-02 -3.98098111e-01
1.31606042e-01 1.17766678e-01 -1.20129251e+00 4.63740081e-01
8.15961659e-01 5.43898165e-01 2.96766490e-01 7.67133355e-01
-3.06333035e-01 1.10851634e+00 -5.35102129e-01 -4.56399113e-01
1.11034632e-01 -2.36946627e-01 -8.99808109e-01 -4.10213143e-01
1.51563972e-01 -7.00024843e-01 -6.45694673e-01 6.39530778e-01
7.34192908e-01 5.99507689e-01 7.11587667e-01 -5.56334615e-01
-1.04258883e+00 1.03678536e+00 -5.38386703e-01 3.52180302e-01
3.70426387e-01 -1.09333165e-01 1.63782108e+00 -1.38575852e+00
7.25391924e-01 1.11534464e+00 4.13773835e-01 5.38527369e-01
-1.23893392e+00 -4.85280067e-01 1.44037262e-01 2.42730856e-01
-1.55737031e+00 -6.19062841e-01 6.84958518e-01 -2.25122467e-01
1.68376064e+00 1.48767278e-01 4.53957349e-01 1.20535457e+00
-3.34544450e-01 1.34081388e+00 5.47244668e-01 -6.92788601e-01
2.48130448e-02 1.62299901e-01 4.64549631e-01 5.07706940e-01
2.22311556e-01 -1.88480407e-01 -3.37837338e-01 -2.41723448e-01
5.33723474e-01 2.06320003e-01 -1.94508433e-01 1.54897049e-02
-8.39589715e-01 1.06728542e+00 3.89040887e-01 4.07454848e-01
-5.28182924e-01 1.36866510e-01 6.53800786e-01 2.71688312e-01
6.47658825e-01 8.22770953e-01 -4.27003503e-01 -4.03040141e-01
-1.09975958e+00 2.86340058e-01 7.73126423e-01 8.61763895e-01
9.30057704e-01 -1.81758821e-01 -5.29778600e-01 1.40730870e+00
2.20279858e-01 4.21697259e-01 5.87745368e-01 -7.94528902e-01
6.02440119e-01 2.99572110e-01 1.87251821e-01 -7.16183662e-01
6.85063228e-02 -4.19096053e-01 -3.83139044e-01 -6.70315623e-01
-2.01117229e-02 -1.77970484e-01 -1.01697421e+00 1.63342655e+00
-1.24396095e-02 4.88202162e-02 2.06566423e-01 7.76109695e-01
6.53766274e-01 1.18228734e+00 2.71647006e-01 1.00120686e-01
1.10794604e+00 -1.33651018e+00 -6.26736403e-01 -4.16707665e-01
1.21157932e+00 -9.12861526e-01 1.37137425e+00 1.19504884e-01
-1.33419049e+00 -2.38548771e-01 -9.24403965e-01 -7.41872013e-01
-3.09561819e-01 1.82894528e-01 4.47590888e-01 2.66436394e-02
-9.90424275e-01 4.62899864e-01 -8.02089155e-01 -2.74650156e-01
3.98136020e-01 -2.86572501e-02 -3.61308277e-01 -6.74481332e-01
-1.43232214e+00 1.10727441e+00 2.79784024e-01 -1.16817057e-01
-1.04894185e+00 -1.00047851e+00 -8.49913001e-01 3.10077697e-01
3.18471551e-01 -7.15297341e-01 1.50249791e+00 -6.05903327e-01
-1.29608524e+00 6.81808233e-01 -3.82038891e-01 -4.84090269e-01
1.17681459e-01 -7.74379015e-01 -6.54003695e-02 1.46556869e-01
-1.28279418e-01 6.14989340e-01 5.38328111e-01 -1.06245697e+00
-3.72102261e-01 -6.88828602e-02 1.72412768e-01 4.89676386e-01
-4.90879267e-01 2.53267765e-01 -7.82729328e-01 -7.26328313e-01
-4.07339156e-01 -5.72233558e-01 -3.08864832e-01 -4.59100991e-01
-2.15817407e-01 -5.98904371e-01 2.90010095e-01 -9.79341447e-01
1.63842177e+00 -2.07789493e+00 1.43027052e-01 1.12405360e-01
1.68573707e-01 6.13063395e-01 -6.82324886e-01 8.18347573e-01
1.76709414e-01 3.18097889e-01 2.24352274e-02 -7.55920172e-01
1.41074792e-01 4.35062706e-01 -6.33927584e-01 1.26805648e-01
2.79726356e-01 1.08292747e+00 -1.18721962e+00 -5.33603489e-01
6.63096607e-02 6.92718387e-01 -8.85029554e-01 4.64844465e-01
-4.34849620e-01 -1.97702244e-01 -5.28111994e-01 4.55375224e-01
2.26017445e-01 -3.28200877e-01 6.88027292e-02 2.95389257e-02
2.05104314e-02 1.01345754e+00 -6.66286707e-01 1.96764386e+00
-7.71740556e-01 5.95308542e-01 -1.33892268e-01 -1.04510140e+00
7.36835241e-01 3.53726566e-01 2.74442792e-01 -9.87238109e-01
7.56216347e-02 2.08565801e-01 -3.63736898e-01 -4.11683470e-01
1.01933527e+00 -1.24852240e-01 -1.45636231e-01 7.14318693e-01
2.34700143e-01 -2.19874177e-02 3.17303866e-01 6.88233674e-01
1.42594039e+00 -4.75939810e-02 9.10908952e-02 9.48769301e-02
9.37401503e-02 5.36973216e-02 4.28231359e-01 1.01964235e+00
1.47973672e-01 7.37347007e-01 1.32592201e-01 6.91539869e-02
-1.34986544e+00 -1.09631550e+00 1.17558800e-01 1.39021194e+00
-3.32728386e-01 -7.34037340e-01 -3.86247784e-01 -4.57167387e-01
9.32299942e-02 9.29010987e-01 -4.20470625e-01 -3.26186985e-01
-7.43869603e-01 -4.10208225e-01 6.58127308e-01 6.57264888e-01
2.40710285e-02 -9.22689021e-01 5.47994561e-02 4.95035619e-01
-4.51709449e-01 -1.02254927e+00 -5.21504879e-01 2.17497453e-01
-5.98639846e-01 -4.39290106e-01 -8.94859374e-01 -6.09619021e-01
5.12076378e-01 5.30565917e-01 1.44146299e+00 2.35534102e-01
-4.34923708e-01 4.97457743e-01 -8.47848892e-01 -2.93538183e-01
-1.89844668e-01 3.06676924e-01 -1.98001973e-03 -4.75681990e-01
7.39778996e-01 -4.73243386e-01 -7.60176003e-01 -8.27938914e-02
-9.58160639e-01 -9.97369438e-02 7.47001529e-01 9.63184893e-01
7.85852075e-01 -4.21750158e-01 6.66929662e-01 -7.97662199e-01
1.03715277e+00 -7.29046047e-01 -1.98638707e-01 5.00204563e-01
-4.92369145e-01 2.05208570e-01 4.85963464e-01 -3.92611057e-01
-9.63316858e-01 -6.02201939e-01 -4.17709887e-01 -5.64766824e-01
2.67657548e-01 9.18358386e-01 2.86351979e-01 5.49522579e-01
7.02030540e-01 2.21681371e-01 -3.14578950e-01 -7.63351738e-01
7.44535029e-01 7.29337633e-01 1.77862883e-01 -9.51797426e-01
6.82711124e-01 1.24236815e-01 -8.46451581e-01 -8.92611027e-01
-1.23358274e+00 -9.49598074e-01 -1.08093716e-01 1.82243600e-01
6.35405421e-01 -1.37808752e+00 1.46776259e-01 -9.77816060e-02
-1.22603953e+00 -6.23598158e-01 -6.29481256e-01 5.91313422e-01
-2.79720336e-01 3.68664294e-01 -9.39514875e-01 -5.45348048e-01
-8.26114178e-01 -6.82228744e-01 9.78470564e-01 9.26634669e-02
-4.04269516e-01 -9.95579064e-01 5.87104321e-01 4.61991340e-01
6.53996408e-01 -5.18687308e-01 9.08771813e-01 -9.37723935e-01
-4.24422234e-01 -5.95129311e-01 -4.02548969e-01 5.93908966e-01
-1.85519025e-01 -8.35963264e-02 -9.55121577e-01 -3.63864213e-01
-3.62421572e-01 -8.28187048e-01 1.11871517e+00 8.92313942e-02
9.94199753e-01 -5.39776266e-01 -2.04026386e-01 2.43743867e-01
1.44411087e+00 -1.92921698e-01 6.70116067e-01 1.26125365e-01
7.05412269e-01 2.79307038e-01 6.00550473e-01 4.57546175e-01
4.77428645e-01 4.98610467e-01 -2.72083133e-01 4.61319787e-03
-1.56289563e-01 -7.06784368e-01 5.51391602e-01 1.39817917e+00
7.27268159e-02 -3.92949790e-01 -9.35063839e-01 9.34141219e-01
-1.76153147e+00 -9.70928311e-01 5.86552799e-01 1.95736933e+00
1.27858484e+00 -1.78546101e-01 -3.06144595e-01 -3.01902592e-01
2.10406259e-01 2.49198705e-01 -2.79107273e-01 -6.09970570e-01
3.53496918e-03 5.99104464e-01 2.54037857e-01 7.83153772e-01
-7.22562432e-01 1.19794667e+00 5.32092667e+00 1.06232738e+00
-8.83824289e-01 3.64720762e-01 2.46143118e-01 -4.91409153e-01
-7.69747853e-01 8.21885392e-02 -9.54154909e-01 1.15353711e-01
1.12431741e+00 -3.50986749e-01 4.49484974e-01 7.44017541e-01
2.13706478e-01 6.76421374e-02 -1.12653303e+00 7.88566530e-01
3.28148335e-01 -1.42891562e+00 4.29830164e-01 -1.74502432e-01
6.88458443e-01 6.18777990e-01 -5.59467152e-02 1.05254364e+00
6.16429090e-01 -8.84868741e-01 4.69372213e-01 3.08807015e-01
5.76340318e-01 -4.43450958e-01 5.58796704e-01 3.91326457e-01
-8.57107997e-01 8.84661749e-02 -7.10421801e-01 2.35355459e-02
2.24901915e-01 6.72232747e-01 -9.32141304e-01 4.68395203e-01
3.13248605e-01 7.94810474e-01 -4.48716134e-01 7.77275920e-01
-3.73122662e-01 8.06331992e-01 -3.63233000e-01 -1.13424912e-01
4.91806358e-01 4.54085134e-02 4.23746496e-01 1.62966776e+00
2.71501452e-01 1.59608915e-01 1.14097841e-01 6.84687495e-01
-4.71693158e-01 3.28125864e-01 -5.55257618e-01 -5.63634396e-01
7.24337041e-01 1.12371278e+00 2.37730350e-02 -6.94626629e-01
-5.82862079e-01 8.66367579e-01 8.54416966e-01 6.36575222e-01
-6.54100299e-01 -5.39223373e-01 6.78947270e-01 4.04140651e-02
5.05955279e-01 -2.54333347e-01 3.81844081e-02 -1.63021672e+00
5.09203561e-02 -7.06885636e-01 3.96129310e-01 -6.99666858e-01
-1.34080076e+00 3.51206154e-01 -1.04727760e-01 -7.53338397e-01
-3.98459643e-01 -2.80003637e-01 -4.71477479e-01 1.04431319e+00
-1.91155756e+00 -9.81982827e-01 2.00370401e-01 4.90179718e-01
7.23526716e-01 5.21133468e-02 9.90480483e-01 5.78950405e-01
-6.58683658e-01 7.72975564e-01 3.72809678e-01 3.66359502e-01
8.61012995e-01 -1.11522293e+00 2.51483262e-01 7.74303854e-01
3.01953167e-01 1.02538109e+00 5.63717961e-01 -4.31034565e-01
-1.51589072e+00 -1.15647078e+00 1.34974313e+00 -5.23472548e-01
7.98179090e-01 -4.46497917e-01 -1.19943237e+00 8.52198660e-01
3.15481365e-01 -7.92821422e-02 1.00568473e+00 5.48768163e-01
-8.08606684e-01 1.90319475e-02 -6.10097826e-01 6.06478155e-01
6.29284441e-01 -1.02103639e+00 -9.97348726e-01 4.49540287e-01
8.48272681e-01 -4.73437794e-02 -1.02627397e+00 1.06423669e-01
4.07586783e-01 -2.37395212e-01 1.12077320e+00 -7.76259303e-01
7.33517528e-01 6.49831817e-02 -3.52462769e-01 -1.18726695e+00
-3.42698097e-01 -3.72210503e-01 -3.91729802e-01 1.24997997e+00
6.82123065e-01 -2.71432817e-01 3.80164713e-01 6.71600759e-01
-2.68337637e-01 -8.83058369e-01 -5.80554426e-01 -7.94530988e-01
4.22762275e-01 -5.77476561e-01 2.83770084e-01 9.56828356e-01
3.86645943e-01 6.21416390e-01 -2.34857708e-01 -9.99687463e-02
3.69935781e-01 -9.72912386e-02 7.02057004e-01 -6.46713138e-01
-4.51367617e-01 -3.20835561e-01 1.56172156e-01 -1.44972920e+00
2.74185061e-01 -1.17832983e+00 2.95275748e-01 -1.91519964e+00
5.12761116e-01 -6.56593859e-01 -5.94308376e-01 6.45543814e-01
-3.58712167e-01 3.78668010e-02 1.01710893e-01 2.74269164e-01
-9.35177565e-01 6.82286978e-01 9.16807652e-01 -1.86896279e-01
-2.85727710e-01 -6.30217135e-01 -7.93662906e-01 2.82646507e-01
8.08773875e-01 -5.94211400e-01 -4.07045573e-01 -1.06833553e+00
3.81776243e-01 5.80167398e-02 4.90180962e-02 -4.67247128e-01
2.28406474e-01 1.58187434e-01 9.85021740e-02 -4.08185989e-01
5.99748075e-01 -3.10854703e-01 -4.28198069e-01 4.21978459e-02
-7.47557461e-01 -3.17230672e-02 2.21404761e-01 4.51783925e-01
-4.10937726e-01 -4.12727535e-01 5.07390559e-01 -3.35847080e-01
-6.32659793e-01 2.92939246e-01 -3.02838057e-01 4.80061114e-01
5.01691222e-01 3.12737495e-01 -4.48759198e-01 -4.01391715e-01
-4.90415603e-01 4.70896244e-01 2.94165522e-01 3.66222501e-01
6.33547485e-01 -1.14583147e+00 -8.17851782e-01 -1.69720218e-01
2.15681836e-01 1.16554357e-01 4.86953884e-01 5.11842847e-01
-2.95196861e-01 7.30940700e-01 3.49606276e-01 -1.57149687e-01
-9.60802257e-01 5.08365095e-01 -1.86126139e-02 -8.55622232e-01
-7.67705560e-01 1.04560220e+00 2.37752334e-03 -2.78455913e-01
3.36932182e-01 -2.50872701e-01 2.43749134e-02 1.74478963e-01
6.97509468e-01 6.69345409e-02 2.09315419e-02 -2.18521312e-01
-1.02244496e-01 2.21434832e-01 -7.89421201e-01 -3.49923342e-01
1.45992494e+00 -1.33090302e-01 -3.52249364e-03 3.46330792e-01
1.82486403e+00 -3.86576727e-02 -8.24399948e-01 -6.66438639e-01
1.96092930e-02 -2.55524993e-01 3.15391988e-01 -8.26908648e-01
-5.64784825e-01 1.18713737e+00 -1.04980931e-01 -2.14261502e-01
8.33081782e-01 1.72080934e-01 1.19247913e+00 8.89859140e-01
7.52096996e-02 -1.28545749e+00 2.23695412e-01 7.96226621e-01
8.49170148e-01 -1.06809461e+00 -1.29747614e-01 2.55293489e-01
-6.31147683e-01 6.17178559e-01 4.35479909e-01 -2.80982345e-01
5.98460376e-01 8.31910893e-02 -2.60316785e-02 4.04783728e-04
-1.16393292e+00 -2.51526892e-01 1.69077963e-01 1.89034969e-01
5.78078151e-01 1.35008860e-02 -2.39286378e-01 6.72326326e-01
9.21925821e-04 2.40140840e-01 3.36143076e-01 1.17490101e+00
-5.99654257e-01 -1.14817309e+00 1.63115874e-01 5.71727037e-01
-5.98665893e-01 -5.65062642e-01 -1.52208939e-01 4.61967498e-01
-5.15469551e-01 6.70243740e-01 2.40908563e-01 -2.83328027e-01
3.23543131e-01 3.31280917e-01 4.39401597e-01 -1.06009865e+00
-4.70049351e-01 8.95246491e-02 3.88418436e-01 -4.98886824e-01
-7.27528557e-02 -4.74495977e-01 -1.02174139e+00 -2.87063420e-01
-3.41243982e-01 5.13979256e-01 4.31078315e-01 9.52472210e-01
5.65815032e-01 3.33779126e-01 4.87290800e-01 -5.82150996e-01
-7.98863351e-01 -1.09981656e+00 -2.54007161e-01 4.23965037e-01
2.73880333e-01 -2.55213797e-01 -4.12876636e-01 -1.99201643e-01] | [11.425405502319336, 7.737371444702148] |
c10d634e-64e4-4030-b5f4-4bc82343eb8a | difffacto-controllable-part-based-3d-point | 2305.01921 | null | https://arxiv.org/abs/2305.01921v2 | https://arxiv.org/pdf/2305.01921v2.pdf | DiffFacto: Controllable Part-Based 3D Point Cloud Generation with Cross Diffusion | While the community of 3D point cloud generation has witnessed a big growth in recent years, there still lacks an effective way to enable intuitive user control in the generation process, hence limiting the general utility of such methods. Since an intuitive way of decomposing a shape is through its parts, we propose to tackle the task of controllable part-based point cloud generation. We introduce DiffFacto, a novel probabilistic generative model that learns the distribution of shapes with part-level control. We propose a factorization that models independent part style and part configuration distributions and presents a novel cross-diffusion network that enables us to generate coherent and plausible shapes under our proposed factorization. Experiments show that our method is able to generate novel shapes with multiple axes of control. It achieves state-of-the-art part-level generation quality and generates plausible and coherent shapes while enabling various downstream editing applications such as shape interpolation, mixing, and transformation editing. Project website: https://difffacto.github.io/ | ['Leonidas J Guibas', 'Ke Li', 'Shi-Min Hu', 'Jiahui Huang', 'Mikaela Angelina Uy', 'Kiyohiro Nakayama'] | 2023-05-03 | null | null | null | null | ['point-cloud-generation'] | ['computer-vision'] | [-1.63313687e-01 1.83808982e-01 2.27348894e-01 1.74481124e-02
-4.68003303e-01 -9.69237208e-01 9.65224862e-01 -1.02021270e-01
4.15725619e-01 5.25354564e-01 2.12363780e-01 -1.63696647e-01
-7.71652013e-02 -1.15333807e+00 -8.89535546e-01 -4.86649990e-01
2.16246650e-01 8.50521445e-01 -3.83465886e-02 -3.87751311e-01
8.50959346e-02 9.07503605e-01 -1.61619318e+00 1.58492476e-01
1.07416773e+00 6.11559451e-01 2.50724614e-01 6.30735219e-01
-3.07324857e-01 6.02738336e-02 -4.41637814e-01 -5.03279448e-01
5.97420692e-01 -1.79129958e-01 -4.22448903e-01 2.27792561e-01
2.92616010e-01 2.30151489e-02 5.73470667e-02 6.42367721e-01
6.07059121e-01 -5.62044829e-02 9.39789116e-01 -1.55804884e+00
-9.98878956e-01 4.05928403e-01 -4.90497828e-01 -6.95474625e-01
3.10486913e-01 3.39947701e-01 7.56538630e-01 -9.43238854e-01
9.79702950e-01 1.32008505e+00 5.49164355e-01 7.37091601e-01
-1.72016609e+00 -4.71593857e-01 9.61724371e-02 -5.70251822e-01
-1.37444556e+00 -1.96578607e-01 9.40675735e-01 -6.68131888e-01
4.59108561e-01 3.71640593e-01 1.04057443e+00 9.67272282e-01
1.50351912e-01 7.96528757e-01 8.40913296e-01 -2.72963434e-01
3.18740308e-01 -5.65718152e-02 -6.63664699e-01 3.64906043e-01
1.57595456e-01 -1.06408782e-01 -4.60629553e-01 -3.95860821e-01
1.39826667e+00 1.32922605e-01 -2.17549652e-01 -9.14061725e-01
-1.48948419e+00 6.38758898e-01 4.89275843e-01 1.77842185e-01
-3.54305327e-01 4.19508457e-01 -1.60602942e-01 -1.38544375e-02
6.94030285e-01 4.88261640e-01 -3.91934693e-01 -1.39995575e-01
-9.41095650e-01 8.12494874e-01 7.39323676e-01 1.37578082e+00
7.54042387e-01 -9.91424918e-03 -3.74959111e-01 7.74868190e-01
3.35589468e-01 7.01928794e-01 -1.03318103e-01 -1.16716766e+00
1.54832318e-01 8.04236472e-01 2.84914434e-01 -8.62905025e-01
1.98681820e-02 -2.46141881e-01 -1.06493640e+00 6.47671044e-01
1.65641308e-01 -9.13552791e-02 -1.11951411e+00 1.62696576e+00
4.84890074e-01 7.83363879e-02 -3.84768516e-01 6.15544200e-01
6.24262691e-01 5.08301914e-01 -1.04090884e-01 1.60513476e-01
1.04567254e+00 -6.60364151e-01 -3.15431118e-01 2.81714350e-01
2.29722448e-02 -9.72179830e-01 1.05444610e+00 3.31055433e-01
-1.35740674e+00 -4.50930446e-01 -7.25498319e-01 -1.02218606e-01
-1.55315384e-01 -4.75992039e-02 7.40922272e-01 4.29705322e-01
-1.26466632e+00 7.79230714e-01 -9.20498848e-01 -8.92004594e-02
7.06693888e-01 1.95221886e-01 -4.18259233e-01 -8.61128420e-03
-5.50785959e-01 4.32190031e-01 -1.01425394e-01 -1.73949480e-01
-6.58506572e-01 -1.32839382e+00 -7.35080361e-01 -3.86985298e-03
4.69180606e-02 -1.69030595e+00 1.07058775e+00 -5.36297977e-01
-1.69620514e+00 6.70923829e-01 -2.14236706e-01 9.62898135e-03
9.78164673e-01 -1.87699527e-01 1.36992782e-01 -1.04756817e-01
1.42828017e-01 1.18806553e+00 1.17936492e+00 -1.67509079e+00
-1.36878222e-01 -9.36930403e-02 -4.31384854e-02 9.63305086e-02
-3.76216844e-02 -4.74627137e-01 -4.83794391e-01 -9.60673273e-01
1.32609010e-01 -1.07940745e+00 -5.13890445e-01 5.90043545e-01
-5.59656322e-01 -2.66527474e-01 8.64066601e-01 -2.09666848e-01
8.49111974e-01 -1.96545768e+00 3.74271363e-01 2.97622412e-01
3.56493086e-01 -1.09026738e-01 -1.08237825e-01 8.21727514e-01
-4.21770215e-02 5.20561397e-01 -5.54863989e-01 -8.14233363e-01
3.42359155e-01 1.74903348e-01 -4.65880513e-01 -1.25575773e-02
3.81368637e-01 1.06156993e+00 -8.98513436e-01 -2.03835025e-01
4.07212049e-01 1.07833517e+00 -8.72890532e-01 1.37191549e-01
-6.62859559e-01 8.24429095e-01 -4.08442557e-01 6.35070324e-01
1.08668971e+00 -1.57354161e-01 -1.11537717e-01 -1.42431706e-01
-7.60882646e-02 -1.19890712e-01 -1.24346817e+00 2.14902258e+00
-5.53880751e-01 2.11267561e-01 1.03791639e-01 -1.82602316e-01
1.01359940e+00 3.30304682e-01 6.60561860e-01 4.46120240e-02
-1.03743605e-01 1.93104014e-01 -2.47715637e-01 1.55387655e-01
6.28696680e-01 -2.61565298e-01 2.53638793e-02 5.12970507e-01
-1.30609736e-01 -1.04421353e+00 1.69147834e-01 4.44808245e-01
8.33499730e-01 6.98679149e-01 -5.19888029e-02 -3.23707670e-01
1.23124741e-01 -1.92367643e-01 1.93526566e-01 2.65369445e-01
5.62293828e-01 1.25342536e+00 4.36000943e-01 -3.30681026e-01
-1.18019354e+00 -1.46310234e+00 -3.67603637e-02 5.14829457e-01
-3.98891047e-02 -5.54657400e-01 -6.95791781e-01 -4.25553262e-01
3.80906582e-01 8.29648614e-01 -7.66736269e-01 2.19541311e-01
-3.71479064e-01 -3.79857451e-01 1.76210716e-01 4.14090693e-01
6.83505088e-02 -1.04003942e+00 -2.11985096e-01 1.31925389e-01
-2.98445150e-02 -7.03577280e-01 -6.02217257e-01 -3.50502312e-01
-1.03504145e+00 -7.83165991e-01 -9.76766407e-01 -4.27790940e-01
1.10190201e+00 2.31257528e-01 1.46768987e+00 1.71593744e-02
-2.16354892e-01 4.75832492e-01 -1.59183055e-01 -4.10874546e-01
-7.25225747e-01 7.67901391e-02 -1.17042504e-01 4.67561521e-02
-4.54289705e-01 -1.23657525e+00 -9.25512135e-01 2.08751202e-01
-1.25575411e+00 4.77802724e-01 3.63067955e-01 4.36714500e-01
9.42006588e-01 -1.94418922e-01 3.53359610e-01 -6.47693813e-01
8.25573385e-01 -2.91896224e-01 -4.43787456e-01 7.98234046e-02
-3.95721048e-01 2.44432718e-01 5.88022411e-01 -3.22998434e-01
-9.93685246e-01 1.28781676e-01 -1.07001223e-01 -8.16343963e-01
-3.40405017e-01 8.31502490e-03 -3.19665730e-01 3.86761315e-02
5.15509069e-01 1.51150495e-01 1.78159177e-01 -6.21950150e-01
1.02942932e+00 1.03989474e-01 3.05844426e-01 -9.82842386e-01
1.31845629e+00 7.10158348e-01 1.69167131e-01 -7.17281878e-01
2.77609546e-02 -4.80265506e-02 -8.00823212e-01 -1.89034566e-01
6.99130595e-01 -8.55160058e-01 -6.18423820e-01 5.38434982e-01
-1.30037689e+00 -3.85409683e-01 -6.38349891e-01 -2.76338845e-01
-8.64599884e-01 9.37456414e-02 -3.00091624e-01 -5.79672754e-01
-4.46541458e-01 -1.08294892e+00 1.48516464e+00 9.14897621e-02
-3.58286291e-01 -8.00870836e-01 2.60314673e-01 -6.23989217e-02
4.72520113e-01 7.63112009e-01 7.46653616e-01 3.15415829e-01
-9.80686903e-01 -1.39887378e-01 7.61621669e-02 8.35679621e-02
4.54964101e-01 5.65659642e-01 -7.56733179e-01 -1.77831709e-01
-5.33145010e-01 1.91138282e-01 5.14279366e-01 4.40633774e-01
1.11334074e+00 -2.12862775e-01 -5.34432769e-01 7.35665143e-01
1.33921111e+00 -3.84459674e-01 7.38040566e-01 -1.59758925e-01
9.22607124e-01 3.61531138e-01 9.69586521e-02 4.84423965e-01
4.44687009e-01 7.69164443e-01 4.44983602e-01 2.24339440e-02
-3.95838439e-01 -7.18196034e-01 -2.49494538e-02 7.59751856e-01
-5.25293887e-01 -2.64267147e-01 -7.69377530e-01 5.96367598e-01
-1.83186710e+00 -8.69214535e-01 -1.69178665e-01 2.12025428e+00
6.84199750e-01 -1.58443406e-01 1.40584677e-01 -1.13498196e-01
6.34660840e-01 -2.33424176e-02 -4.30165231e-01 -2.69979507e-01
7.81317651e-02 4.80339170e-01 1.37197539e-01 3.31588209e-01
-7.56317317e-01 7.67749190e-01 5.72370481e+00 9.33669746e-01
-1.10569954e+00 -6.97887018e-02 6.62293673e-01 -1.33147806e-01
-1.08102024e+00 -9.10245031e-02 -4.10684586e-01 4.36283737e-01
2.48855740e-01 -3.84556353e-01 2.82429844e-01 8.48764420e-01
4.09227014e-01 1.93516478e-01 -9.40111458e-01 9.08295155e-01
-2.51773804e-01 -1.78484118e+00 5.34999371e-01 2.18857437e-01
1.15464437e+00 -1.94443569e-01 1.33377567e-01 -9.38118994e-02
5.60527027e-01 -1.04701269e+00 1.05819917e+00 9.63759780e-01
1.00047231e+00 -8.74526441e-01 -5.60100526e-02 5.12285948e-01
-1.20059669e+00 4.84830499e-01 -2.08445013e-01 1.29785925e-01
5.08764565e-01 8.95709574e-01 -7.43677914e-01 8.11546147e-01
3.86863559e-01 6.85164034e-01 -3.85838330e-01 1.11635554e+00
-3.14599782e-01 1.94458127e-01 -5.77652752e-01 2.82504350e-01
-2.43148729e-01 -4.98040348e-01 7.52403200e-01 8.12804282e-01
7.85453737e-01 -8.83831456e-02 9.73529741e-02 1.48349500e+00
-2.47563750e-01 -1.25634167e-02 -7.80301690e-01 4.14757282e-02
5.33083737e-01 1.30552936e+00 -8.91726434e-01 -5.63215427e-02
1.33332908e-01 1.06700051e+00 2.61095881e-01 2.35544860e-01
-6.81133986e-01 -4.72371578e-02 9.35657501e-01 6.44815505e-01
4.66659695e-01 -6.39344037e-01 -6.68976307e-01 -1.13856089e+00
1.11228906e-01 -5.58120668e-01 -3.27261120e-01 -1.01804328e+00
-1.48901641e+00 5.96787930e-01 8.85508806e-02 -1.59749365e+00
-1.76482663e-01 -3.37038398e-01 -8.69844317e-01 9.66493487e-01
-1.10147619e+00 -1.66254592e+00 -2.81994492e-01 3.93527925e-01
3.74769479e-01 1.56946078e-01 8.88677239e-01 5.96455820e-02
-4.68030125e-02 3.31825227e-01 -7.56546259e-02 -3.17971438e-01
6.92873478e-01 -1.28769624e+00 9.49171484e-01 6.53789222e-01
1.97657436e-01 7.99251437e-01 6.23359025e-01 -7.44235933e-01
-1.48393643e+00 -1.10841620e+00 4.92287725e-01 -8.17892492e-01
2.47735664e-01 -5.16852081e-01 -5.24095118e-01 3.23730081e-01
2.11571977e-01 3.02040260e-02 3.46166939e-01 -2.92718738e-01
-2.80752152e-01 1.15919739e-01 -1.26138771e+00 8.28788817e-01
1.38302338e+00 -2.02540919e-01 -1.78541750e-01 1.35650441e-01
8.53848398e-01 -5.31328380e-01 -9.41014528e-01 4.11691636e-01
5.56533098e-01 -1.03294241e+00 1.16933167e+00 -2.08523244e-01
6.90659165e-01 -7.32423842e-01 1.12843893e-01 -1.67905986e+00
-5.04653275e-01 -1.05768180e+00 -1.67287663e-01 1.35001171e+00
3.68396938e-01 -4.97630984e-01 8.76328349e-01 6.95126534e-01
-3.18214983e-01 -8.96701157e-01 -6.76234365e-01 -5.90412855e-01
3.83752406e-01 -5.81005633e-01 1.11861241e+00 7.79107690e-01
-4.02373642e-01 -7.69034252e-02 -1.10800043e-01 -5.28681986e-02
6.03698254e-01 5.03397584e-01 1.22370327e+00 -1.31614769e+00
-3.10685158e-01 -6.98229134e-01 -2.80765682e-01 -1.16496706e+00
-1.64400116e-01 -1.03335810e+00 -1.27297267e-01 -1.97004914e+00
-1.21828057e-01 -8.16202760e-01 3.12632859e-01 3.28889757e-01
-1.03317231e-01 4.52058285e-01 4.94264990e-01 1.40947446e-01
6.92492723e-02 8.75600636e-01 1.83103836e+00 -3.04757301e-02
-2.33154297e-01 1.42442793e-01 -9.11316752e-01 4.95064348e-01
7.74780571e-01 -2.25858882e-01 -5.07511020e-01 -5.22537172e-01
3.50444406e-01 -9.58227366e-02 5.15459418e-01 -9.67894614e-01
-1.71544343e-01 -2.29956344e-01 3.72683167e-01 -6.63180172e-01
4.00167525e-01 -6.92629874e-01 9.01689947e-01 4.87777740e-02
1.63179666e-01 1.43165767e-01 3.64079773e-01 5.38618684e-01
3.35497931e-02 2.88319111e-01 5.16798019e-01 -1.97929516e-01
-9.28474218e-02 6.92690670e-01 7.60308700e-03 -7.74588585e-02
9.04933095e-01 -2.44910374e-01 -6.61111325e-02 -3.83900821e-01
-8.43615592e-01 3.17482091e-02 1.12696862e+00 7.21613467e-01
6.03949249e-01 -1.69256234e+00 -7.95870721e-01 4.09998238e-01
1.63616810e-03 5.57086051e-01 3.14914614e-01 3.86724919e-01
-6.47432804e-01 -9.11928937e-02 -1.76674977e-01 -7.60923743e-01
-8.60984445e-01 2.64468372e-01 1.10682070e-01 -1.05479427e-01
-7.31287599e-01 8.41401517e-01 3.66991311e-01 -8.07711005e-01
-4.25724387e-01 -4.68312621e-01 3.15288633e-01 -1.48028776e-01
2.25286737e-01 2.66791254e-01 5.54899126e-02 -4.93450761e-01
-6.67135492e-02 7.32177377e-01 4.29365963e-01 -1.80483162e-01
1.52929699e+00 1.86902732e-01 -2.44649842e-01 2.33701006e-01
6.73082471e-01 3.80363435e-01 -1.55865657e+00 4.29038376e-01
-6.26763642e-01 -6.53763294e-01 -2.78427988e-01 -7.30867565e-01
-1.08862150e+00 7.76161075e-01 7.66141340e-02 3.26571077e-01
7.38262296e-01 1.49534093e-02 7.86348403e-01 -1.73270017e-01
8.06052625e-01 -4.59858477e-01 -1.30027607e-01 4.37498420e-01
1.45376539e+00 -6.76287591e-01 -5.29252701e-02 -7.99447656e-01
-4.21754569e-01 1.03331518e+00 2.67888367e-01 -3.70877415e-01
9.12055075e-01 4.58315402e-01 -1.08981850e-02 -1.58853978e-01
-6.85519040e-01 1.07079156e-01 4.82656986e-01 9.10417020e-01
4.96185839e-01 3.35958481e-01 -2.36261636e-02 3.44949305e-01
-4.99211252e-01 1.40927389e-01 3.63738567e-01 6.74032927e-01
4.63632159e-02 -1.72094667e+00 -3.56723130e-01 3.39356512e-01
3.06607317e-02 1.63571425e-02 -2.85235852e-01 6.37731671e-01
3.12084407e-01 4.59375471e-01 4.45654653e-02 -1.63563579e-01
4.17493999e-01 1.49685163e-02 6.34860814e-01 -6.61035478e-01
-4.65042293e-01 1.15770631e-01 -2.66289383e-01 -5.34693360e-01
-3.63515645e-01 -7.26978779e-01 -1.05443132e+00 -4.84819233e-01
-8.30261707e-02 3.10360324e-02 6.25748277e-01 3.39311153e-01
1.09238064e+00 4.54308957e-01 5.16444564e-01 -1.57562196e+00
-5.11113030e-04 -7.92192757e-01 -4.36858714e-01 6.44518912e-01
-6.22840077e-02 -8.39281082e-01 -1.88234657e-01 3.14974129e-01] | [8.96389389038086, -3.6081926822662354] |
8be18371-d69a-4340-a3ca-b4a7e3966ead | data-augmentation-for-intent-classification-3 | null | null | https://aclanthology.org/2022.konvens-1.1 | https://aclanthology.org/2022.konvens-1.1.pdf | Data Augmentation for Intent Classification of German Conversational Agents in the Finance Domain | null | ['Martin Rückert', 'Christian Stab', 'Martin Riedl', 'Sophie Rentschler'] | null | null | null | null | konvens-ws-2022-9 | ['intent-classification'] | ['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] | [-7.260087490081787, 3.7425098419189453] |
450bc44a-5054-48ab-8932-93ea0afb00e7 | latent-cognizance-what-machine-really-learns | 2110.15548 | null | https://arxiv.org/abs/2110.15548v1 | https://arxiv.org/pdf/2110.15548v1.pdf | Latent Cognizance: What Machine Really Learns | Despite overwhelming achievements in recognition accuracy, extending an open-set capability -- ability to identify when the question is out of scope -- remains greatly challenging in a scalable machine learning inference. A recent research has discovered Latent Cognizance (LC) -- an insight on a recognition mechanism based on a new probabilistic interpretation, Bayesian theorem, and an analysis of an internal structure of a commonly-used recognition inference structure. The new interpretation emphasizes a latent assumption of an overlooked probabilistic condition on a learned inference model. Viability of LC has been shown on a task of sign language recognition, but its potential and implication can reach far beyond a specific domain and can move object recognition toward a scalable open-set recognition. However, LC new probabilistic interpretation has not been directly investigated. This article investigates the new interpretation under a traceable context. Our findings support the rationale on which LC is based and reveal a hidden mechanism underlying the learning classification inference. The ramification of these findings could lead to a simple yet effective solution to an open-set recognition. | ['Tatpong Katanyukul', 'Jiradej Ponsawat', 'Pisit Nakjai'] | 2021-10-29 | null | null | null | null | ['sign-language-recognition', 'open-set-learning'] | ['computer-vision', 'miscellaneous'] | [ 8.24258327e-01 5.12722254e-01 -5.27920783e-01 -5.63914716e-01
-5.29537022e-01 -4.54325765e-01 7.55872965e-01 -3.68695199e-01
-3.57447052e-03 6.21163785e-01 3.71747702e-01 -7.08077550e-01
-8.17225635e-01 -6.23002052e-01 -6.81362748e-01 -7.90864348e-01
2.34096840e-01 5.13521850e-01 1.47571966e-01 2.70090938e-01
5.49309552e-01 4.12811399e-01 -2.18559289e+00 5.12239218e-01
6.16420984e-01 8.34330738e-01 2.75054723e-02 6.35692954e-01
-2.72272527e-01 8.38675320e-01 -3.45124453e-01 -3.97880495e-01
2.24528372e-01 -3.43927056e-01 -8.86154771e-01 -6.12976328e-02
8.31988156e-01 -2.74261922e-01 -5.58581464e-02 1.04031849e+00
1.28731176e-01 -4.15720969e-01 8.80592167e-01 -1.20415616e+00
-8.54774356e-01 5.40273905e-01 2.28405580e-01 1.26814961e-01
4.83021349e-01 2.16996387e-01 1.14482164e+00 -5.60039878e-01
6.97015047e-01 1.17854714e+00 7.87113726e-01 6.84616387e-01
-1.09574151e+00 -4.41762924e-01 6.34594634e-02 4.65470523e-01
-1.27956641e+00 -2.72101074e-01 4.95636344e-01 -6.82317615e-01
7.04089999e-01 5.96037567e-01 7.79803574e-01 1.43374062e+00
1.16972387e-01 8.28103840e-01 1.98761094e+00 -9.60422337e-01
1.90121189e-01 1.44392923e-01 8.36807847e-01 8.39305103e-01
7.17728615e-01 3.98285419e-01 -8.16453099e-01 -4.30142181e-03
7.02658892e-01 -2.69095719e-01 -2.20622867e-01 -1.05433740e-01
-8.52809072e-01 5.36464572e-01 -4.23826799e-02 4.12674904e-01
-1.10639542e-01 2.69130208e-02 -8.14482197e-03 3.78909200e-01
1.96266789e-02 3.61732692e-01 -4.54199404e-01 -4.02572274e-01
-9.06733930e-01 -4.00230661e-02 1.31097543e+00 4.50335890e-01
4.91879851e-01 -2.39384666e-01 -1.55676916e-01 2.81993270e-01
7.10586905e-01 6.62715614e-01 2.74067193e-01 -8.23068738e-01
-1.60493240e-01 8.20905983e-01 -1.34320915e-01 -7.60495067e-01
-3.36196199e-02 -4.61522788e-01 -3.04282367e-01 5.86400867e-01
9.93370175e-01 4.77578998e-01 -8.31438541e-01 1.71788573e+00
8.22874680e-02 1.43869102e-01 1.31027117e-01 8.62130761e-01
6.49447620e-01 4.65377271e-02 -9.37832668e-02 -2.43432261e-02
1.61388552e+00 -4.31323916e-01 -6.68472350e-01 1.82594433e-01
3.62431139e-01 -4.16023642e-01 1.16046119e+00 6.34786189e-01
-3.93826485e-01 -2.20793456e-01 -1.04793584e+00 -1.03856727e-01
-4.91263509e-01 1.83809742e-01 8.32131267e-01 1.06924188e+00
-6.05534315e-01 1.54632181e-01 -6.23357534e-01 -6.67701244e-01
5.57602823e-01 2.53199995e-01 -3.01975846e-01 -1.53731823e-01
-8.66301298e-01 1.30413234e+00 2.85990417e-01 2.10491762e-01
-6.40752375e-01 -4.97259051e-01 -4.42972243e-01 -1.86297715e-01
5.88025272e-01 -8.96835029e-01 9.62882102e-01 -9.08498824e-01
-1.61966538e+00 9.98916745e-01 -5.51213086e-01 -3.68129194e-01
5.44270575e-01 -2.46791199e-01 -5.02622008e-01 1.64850041e-01
2.19298285e-02 2.06465766e-01 1.03961623e+00 -1.31016517e+00
-3.32111597e-01 -5.06710410e-01 -7.42783770e-02 -2.82263547e-01
-1.37134641e-01 1.02925852e-01 1.99258536e-01 -4.02472585e-01
5.61197579e-01 -7.97542751e-01 2.93741405e-01 2.78404236e-01
-1.84351146e-01 -5.55801630e-01 7.68555105e-01 -5.26950002e-01
1.26386786e+00 -1.96869886e+00 -1.35204911e-01 3.48164350e-01
3.82371664e-01 3.29613388e-01 2.69455582e-01 3.02895516e-01
-1.03320479e-01 2.52115637e-01 -2.09247753e-01 1.77764714e-01
3.43991667e-01 7.19455242e-01 -9.66056406e-01 3.68042350e-01
7.68099874e-02 9.94322538e-01 -6.23580277e-01 -4.55914378e-01
1.52406186e-01 2.90150702e-01 -3.77420336e-01 -4.13023792e-02
-3.64846259e-01 1.60522565e-01 -4.23703343e-01 8.03094506e-01
4.88484591e-01 -4.45692658e-01 1.75200179e-01 -1.06348947e-01
-3.17454785e-01 3.80211204e-01 -1.17218316e+00 1.18117940e+00
2.65110940e-01 8.23681235e-01 -5.64020455e-01 -8.87434602e-01
1.00429773e+00 7.55521879e-02 -2.59536970e-02 -1.98366910e-01
-4.52071689e-02 4.76150870e-01 1.92664459e-01 -1.03761959e+00
1.60391077e-01 -2.17541739e-01 1.53932065e-01 5.50228655e-01
1.36556206e-02 3.86977881e-01 -1.27164334e-01 -1.24768756e-01
1.00329316e+00 4.75478172e-01 5.46326816e-01 -4.43894029e-01
4.53176588e-01 2.62563862e-02 3.63384366e-01 1.33414459e+00
-3.30151051e-01 4.82163727e-01 5.80589473e-01 -4.06257480e-01
-7.24956632e-01 -1.26703405e+00 -7.32707500e-01 8.09517145e-01
-1.70925200e-01 -3.74629110e-01 -5.97055256e-01 -6.51482761e-01
9.50590000e-02 7.76483297e-01 -7.88408518e-01 2.55053267e-02
-1.86454445e-01 -5.75878322e-01 1.09447050e+00 4.80264544e-01
4.29831833e-01 -7.16724753e-01 -9.68634486e-01 -3.13218564e-01
-1.06785916e-01 -9.60725307e-01 2.81825811e-01 3.11745759e-02
-8.26191127e-01 -1.25285721e+00 -3.26464772e-01 -4.42402899e-01
7.03015625e-01 -2.51691461e-01 5.87330878e-01 1.20067775e-01
-3.20080429e-01 8.07381809e-01 -2.67027527e-01 -7.15906322e-01
-6.00420237e-01 -1.43624440e-01 2.05229551e-01 1.56689122e-01
1.00186265e+00 -5.91029227e-01 -1.52140334e-01 3.74001950e-01
-7.06262469e-01 3.03164944e-02 7.66506076e-01 8.17608416e-01
2.55274959e-02 -3.14051241e-01 4.77627486e-01 -5.22089481e-01
3.80087167e-01 -2.22382024e-01 -4.02153194e-01 6.42230809e-01
-9.92443264e-01 4.18096036e-01 -1.44149810e-01 -4.61186886e-01
-1.24446654e+00 -4.14019287e-01 1.61282450e-01 -2.92490840e-01
-5.54934502e-01 5.48149288e-01 -2.51984417e-01 4.49980237e-02
6.21754408e-01 5.92980444e-01 3.97810638e-01 -4.46919709e-01
3.33459556e-01 7.89714277e-01 4.87570882e-01 -8.77050638e-01
6.68213010e-01 6.78444922e-01 2.63222665e-01 -1.02709091e+00
-1.21072495e+00 -4.88919467e-01 -7.96850741e-01 -3.35420191e-01
8.13024342e-01 -4.75738019e-01 -9.03569579e-01 4.02266651e-01
-1.17820573e+00 4.89006676e-02 -4.53749597e-01 7.19946086e-01
-5.04793406e-01 5.99822581e-01 -1.12749822e-01 -1.22166562e+00
2.02394590e-01 -8.31470251e-01 7.93509722e-01 1.37555405e-01
-5.87794006e-01 -9.22743738e-01 1.78864107e-01 8.61332774e-01
1.44707829e-01 -2.26739049e-01 9.49358642e-01 -9.21270430e-01
-9.36291993e-01 -4.36646864e-02 -3.72500479e-01 3.69924873e-01
-7.16328025e-02 1.51737437e-01 -1.32093632e+00 1.98362783e-01
1.02940351e-01 -3.70222092e-01 1.02136397e+00 5.42787611e-02
9.43988562e-01 -4.08501059e-01 -2.34059170e-01 3.78038734e-01
1.36276495e+00 -9.77229178e-02 9.30950046e-01 2.12509379e-01
2.45004132e-01 6.38330758e-01 7.35923201e-02 2.15087131e-01
2.55521923e-01 4.67406392e-01 1.53836533e-01 4.94102538e-01
-5.22255301e-01 -4.25699562e-01 4.76454884e-01 7.57668555e-01
-6.90265656e-01 1.25429109e-01 -9.74626541e-01 2.25653902e-01
-1.79989147e+00 -1.42963350e+00 -3.64385933e-01 2.16315556e+00
7.98927248e-01 1.89776883e-01 -2.73038805e-01 3.42757195e-01
5.37754059e-01 -7.26970509e-02 -3.41446847e-01 -1.90691978e-01
-4.90636349e-01 1.85965285e-01 1.35775149e-01 5.22551775e-01
-5.80632210e-01 6.96062446e-01 7.67124844e+00 8.86675358e-01
-1.04990017e+00 5.25858216e-02 -4.51542400e-02 4.19701338e-01
-5.07123113e-01 4.72620279e-01 -1.31029344e+00 1.83733746e-01
6.54130280e-01 2.30148822e-01 7.33891279e-02 7.12647676e-01
-1.72727317e-01 -3.94283831e-01 -1.43523932e+00 8.48386049e-01
5.55913687e-01 -1.40379417e+00 5.93769908e-01 6.04274750e-01
3.55148464e-01 -3.13448101e-01 1.72013536e-01 3.30774546e-01
-3.26902792e-02 -1.14475942e+00 8.12120557e-01 1.20805204e+00
6.19107068e-01 2.10315958e-01 4.94026363e-01 5.01248956e-01
-6.87179685e-01 -3.09772462e-01 -7.32696652e-02 -6.39813244e-01
-1.48786470e-01 3.81623775e-01 -8.66939008e-01 4.08373296e-01
4.18090761e-01 3.07536006e-01 -6.17608190e-01 8.46875668e-01
-5.43028057e-01 1.17622364e+00 -3.92954707e-01 -1.70071542e-01
-6.58844337e-02 -6.05953448e-02 1.01690876e+00 1.13627446e+00
3.35805900e-02 1.27543092e-01 -2.18610257e-01 1.45005596e+00
6.91888273e-01 -3.15389246e-01 -7.06837475e-01 -1.70795277e-01
1.35988355e-01 6.96618319e-01 -7.31668591e-01 -2.21676260e-01
-2.72744179e-01 6.04351521e-01 2.65864164e-01 2.34444290e-01
-7.04694510e-01 1.98899552e-01 2.77788401e-01 2.46235598e-02
3.68473113e-01 -1.66749448e-01 -8.10977042e-01 -1.37634218e+00
3.40332240e-01 -7.78099477e-01 4.38753992e-01 -6.59245372e-01
-1.48864603e+00 -6.14710189e-02 3.66512835e-01 -9.56029713e-01
-4.19249982e-02 -1.11609721e+00 -4.51503992e-01 7.11177826e-01
-1.35243559e+00 -1.48043394e+00 -1.58219695e-01 4.80744213e-01
2.17076555e-01 -3.46846581e-01 1.22375691e+00 -1.54282346e-01
-3.98869514e-01 5.76477945e-01 -1.03489086e-01 9.19197425e-02
4.42850679e-01 -1.19317162e+00 -3.97060275e-01 8.56519341e-01
2.86711365e-01 1.07680440e+00 6.81229293e-01 -7.15752542e-01
-1.42041969e+00 -3.76628488e-01 1.26272631e+00 -1.25076985e+00
7.51517236e-01 -1.48162857e-01 -8.99567842e-01 6.48303509e-01
-1.34872779e-01 -1.28784537e-01 9.96553838e-01 4.78422463e-01
-8.92398715e-01 1.09886378e-01 -9.07485604e-01 5.70595980e-01
1.22873962e+00 -9.20319498e-01 -1.60287035e+00 -2.79153064e-02
2.72256017e-01 1.38967456e-02 -7.92013943e-01 3.68526518e-01
1.20598435e+00 -9.62040126e-01 7.27221489e-01 -9.23123538e-01
3.39798748e-01 -3.39287013e-01 -5.98083377e-01 -3.74779612e-01
-1.18640259e-01 -5.92903435e-01 -4.36704278e-01 1.07169104e+00
2.86447436e-01 -1.07184899e+00 5.72246552e-01 8.47457647e-01
-5.19769676e-02 -7.77211726e-01 -1.18750262e+00 -9.74678874e-01
-6.80151656e-02 -9.94533181e-01 2.27056876e-01 9.13163126e-01
1.17245033e-01 1.71246067e-01 -6.71890974e-02 3.49288672e-01
8.76096129e-01 1.55788168e-01 7.74487793e-01 -1.70829260e+00
-2.98187196e-01 -5.70151150e-01 -7.57682860e-01 -9.97235537e-01
2.89495617e-01 -1.04229736e+00 -3.03261578e-01 -1.26012599e+00
3.47740144e-01 -3.45818639e-01 -3.04858983e-01 5.69184244e-01
1.03085376e-01 -6.39446378e-02 2.36529335e-01 5.03804505e-01
-4.47103262e-01 1.97226599e-01 1.01834846e+00 6.02896921e-02
2.24711388e-01 -3.30994837e-02 -8.11980963e-01 1.00392210e+00
6.25680089e-01 -4.55956340e-01 -2.82435626e-01 -5.28483912e-02
4.23497319e-01 -4.03740853e-01 1.03745747e+00 -9.33377922e-01
4.90891218e-01 -1.06740572e-01 5.35191484e-02 -4.73417759e-01
2.56416708e-01 -8.31260562e-01 1.41732574e-01 6.25967205e-01
-3.34555805e-01 -7.62397408e-01 -2.40936391e-02 7.46681750e-01
1.63825005e-01 -5.65270126e-01 3.18300754e-01 -3.61858793e-02
-1.05590022e+00 -2.82069623e-01 -5.48058808e-01 1.97326597e-02
8.71329010e-01 -9.36114311e-01 -4.71439391e-01 8.71896297e-02
-9.78820801e-01 -4.15430307e-01 1.69658795e-01 3.64650816e-01
6.37240052e-01 -1.00590467e+00 -5.28189898e-01 4.99502033e-01
3.52448702e-01 -5.66452503e-01 2.62668300e-02 9.80027497e-01
-2.44279921e-01 6.67806149e-01 -2.08814740e-01 -1.03914809e+00
-1.30099285e+00 1.28627464e-01 3.88896435e-01 -1.51127722e-04
-8.83052528e-01 6.12214148e-01 -1.46454051e-01 -3.91237825e-01
3.86172682e-01 -5.36837995e-01 1.32453796e-02 -3.88057381e-02
4.12512958e-01 4.46198940e-01 -4.33231056e-01 -5.18072546e-01
-2.98677742e-01 7.24456012e-01 3.60823721e-02 -2.17017472e-01
1.05062389e+00 -4.94836606e-02 -3.83545756e-01 1.01921368e+00
5.83632529e-01 -6.93732873e-02 -9.56844807e-01 -2.24709064e-01
1.41874760e-01 -4.97114629e-01 -7.72609785e-02 -1.16825867e+00
-6.83714822e-02 6.53595686e-01 7.11136103e-01 -1.72044169e-02
4.43610162e-01 2.68028885e-01 7.70099610e-02 7.28503823e-01
4.34733331e-01 -9.13516223e-01 -6.56004921e-02 6.41794324e-01
9.61626530e-01 -1.32736194e+00 1.41207367e-01 -4.52988029e-01
-1.03225537e-01 1.41835892e+00 4.52521294e-01 1.66195452e-01
9.42214429e-01 2.28205845e-01 -1.27401203e-01 -4.44866598e-01
-7.17862666e-01 -2.86222517e-01 5.73592424e-01 7.20336914e-01
2.10045591e-01 1.71745375e-01 -4.04403329e-01 8.20248604e-01
-3.07521909e-01 5.77803314e-01 3.58800143e-01 8.43400121e-01
-4.93851334e-01 -8.46972764e-01 -6.25474870e-01 3.86243761e-01
-1.15880512e-01 7.50489756e-02 -5.34812391e-01 7.61270165e-01
5.69504321e-01 8.17673266e-01 -3.62646058e-02 -3.16808134e-01
-1.99739069e-01 6.71208143e-01 9.37121749e-01 -4.09180552e-01
-8.54958519e-02 -4.66872573e-01 1.58645928e-01 -2.89632648e-01
-8.59126985e-01 -9.44817960e-01 -9.11126018e-01 -1.34040499e-02
-6.22988999e-01 -4.10159528e-02 6.11688495e-01 1.53109336e+00
1.54412746e-01 1.31552607e-01 -1.97899848e-01 -1.20040908e-01
-9.87889349e-01 -9.32720363e-01 -4.96927202e-01 8.97501931e-02
1.55880317e-01 -8.20577681e-01 -5.59644818e-01 1.70663193e-01] | [8.838802337646484, 6.435564994812012] |
78d9fcc1-0bd5-4715-9654-785be8aea563 | cclap-controllable-chinese-landscape-painting | 2304.04156 | null | https://arxiv.org/abs/2304.04156v2 | https://arxiv.org/pdf/2304.04156v2.pdf | CCLAP: Controllable Chinese Landscape Painting Generation via Latent Diffusion Model | With the development of deep generative models, recent years have seen great success of Chinese landscape painting generation. However, few works focus on controllable Chinese landscape painting generation due to the lack of data and limited modeling capabilities. In this work, we propose a controllable Chinese landscape painting generation method named CCLAP, which can generate painting with specific content and style based on Latent Diffusion Model. Specifically, it consists of two cascaded modules, i.e., content generator and style aggregator. The content generator module guarantees the content of generated paintings specific to the input text. While the style aggregator module is to generate paintings of a style corresponding to a reference image. Moreover, a new dataset of Chinese landscape paintings named CLAP is collected for comprehensive evaluation. Both the qualitative and quantitative results demonstrate that our method achieves state-of-the-art performance, especially in artfully-composed and artistic conception. Codes are available at https://github.com/Robin-WZQ/CCLAP. | ['Shiguang Shan', 'Jinfeng Bai', 'Zhilong Ji', 'Jie Zhang', 'Zhongqi Wang'] | 2023-04-09 | null | null | null | null | ['chinese-landscape-painting-generation'] | ['computer-vision'] | [ 2.93297946e-01 -2.13057995e-01 1.64916769e-01 -2.37787277e-01
-6.84150696e-01 -6.15073323e-01 7.64449239e-01 -5.72956622e-01
2.73667455e-01 8.06637406e-01 3.85177344e-01 1.51217490e-01
9.34052765e-02 -1.30328572e+00 -3.01971853e-01 -8.87694299e-01
5.76769054e-01 1.47256911e-01 -1.87478568e-02 -2.90632933e-01
1.79913044e-01 4.34967637e-01 -1.32386482e+00 4.37400401e-01
1.10027111e+00 7.80055583e-01 4.12078083e-01 7.17712283e-01
-3.47268075e-01 7.41478384e-01 -8.17429960e-01 -5.04372001e-01
4.70423885e-02 -7.92580366e-01 -3.21240723e-01 1.56292439e-01
9.03068706e-02 -3.65231663e-01 -3.87041360e-01 9.23997521e-01
7.35140800e-01 -1.12811804e-01 6.99707747e-01 -1.50507033e+00
-1.28066075e+00 5.87164164e-01 -5.44738531e-01 -3.56932670e-01
2.07201004e-01 4.45521802e-01 7.34050035e-01 -6.31865919e-01
9.41934168e-01 1.23726201e+00 4.38588075e-02 9.04793203e-01
-1.23692024e+00 -9.02302623e-01 1.93228230e-01 -1.43969670e-01
-1.27182150e+00 -3.11599880e-01 1.29258561e+00 -2.61318147e-01
1.33877099e-01 3.54890823e-01 7.20008850e-01 1.37470758e+00
4.26026024e-02 1.03521907e+00 1.27580810e+00 -1.78226843e-01
4.30921704e-01 2.24962234e-02 -6.05897129e-01 3.00566137e-01
-1.29446149e-01 -8.90518427e-02 -5.46098292e-01 6.48181662e-02
1.16538405e+00 -1.09050974e-01 -2.26235062e-01 -1.53400406e-01
-1.00205553e+00 7.16607630e-01 5.62349379e-01 1.54414475e-01
-3.05617929e-01 3.60632241e-01 -8.73632636e-03 4.16755537e-03
6.91734195e-01 1.28481850e-01 1.34740651e-01 -1.70376509e-01
-9.64383483e-01 4.32469517e-01 5.54123580e-01 1.46758389e+00
4.83228207e-01 2.23192513e-01 -6.44624829e-01 8.38931501e-01
1.81361780e-01 7.67900169e-01 2.08403170e-01 -7.98257113e-01
5.47673821e-01 6.99153304e-01 1.82344705e-01 -1.01893902e+00
1.94495618e-01 -2.45911926e-01 -1.14787316e+00 4.63373214e-01
-3.96490656e-02 -3.73034120e-01 -9.52659726e-01 1.64569485e+00
1.51646152e-01 -3.87297720e-02 -6.54734224e-02 8.93346786e-01
9.57298517e-01 8.27086985e-01 1.89733505e-01 4.11214195e-02
1.08314633e+00 -1.20819914e+00 -8.62030566e-01 1.10542260e-01
-1.74495995e-01 -1.09381318e+00 1.56517839e+00 1.87016219e-01
-1.22340405e+00 -6.47036910e-01 -8.86463583e-01 -1.17175706e-01
-2.48194918e-01 4.27891940e-01 5.25337815e-01 7.53385901e-01
-9.59902883e-01 2.63356507e-01 -5.21750152e-01 -7.29809701e-02
7.95547187e-01 -1.89607248e-01 1.23708293e-01 5.82312942e-02
-9.49799538e-01 2.74243802e-01 3.94097120e-02 1.91540852e-01
-1.11355984e+00 -7.01378524e-01 -3.72690737e-01 1.94025040e-02
7.49419704e-02 -8.04720223e-01 1.13185358e+00 -1.02136087e+00
-1.98024952e+00 5.16300261e-01 -1.39344335e-01 1.54183969e-01
1.11466289e+00 -2.33736292e-01 -4.84647095e-01 6.61316290e-02
1.58017650e-01 9.59040403e-01 8.34183574e-01 -1.55500472e+00
-5.01016915e-01 9.75211710e-02 1.13069005e-01 1.06418200e-01
-4.50706273e-01 -3.04698676e-01 -7.21417367e-01 -1.03321707e+00
-3.21416587e-01 -8.12053680e-01 -1.16022073e-01 1.84478313e-01
-7.83919394e-01 -1.72942244e-02 1.07656467e+00 -3.56058598e-01
1.32217526e+00 -2.13738894e+00 2.25513354e-01 1.28547460e-01
6.76726550e-02 8.75963364e-03 -3.93135279e-01 7.86910415e-01
1.11199468e-01 5.14683604e-01 -3.26136291e-01 -5.73294580e-01
2.17145443e-01 -7.33307675e-02 -7.27665186e-01 -7.03341514e-02
4.43221390e-01 1.12350261e+00 -9.59983706e-01 -5.29350400e-01
2.28162766e-01 8.47699523e-01 -2.37412870e-01 2.21899658e-01
-5.92929065e-01 7.16872752e-01 -7.48998642e-01 7.64645696e-01
8.72697115e-01 -1.62028462e-01 4.26696762e-02 -6.70208782e-02
-2.22609773e-01 -2.97897398e-01 -9.69201922e-01 1.88656318e+00
-4.69286591e-01 7.68156052e-01 -2.57743776e-01 2.26809025e-01
1.22552443e+00 2.00437397e-01 8.54671150e-02 -7.27613747e-01
1.25747353e-01 4.08112593e-02 -5.45786619e-01 -3.09785396e-01
6.37833595e-01 -4.53583822e-02 2.32264102e-02 5.37032664e-01
-3.35279793e-01 -5.49418747e-01 4.24484044e-01 2.82461017e-01
8.30538869e-01 5.58226049e-01 -4.70222652e-01 -1.31749406e-01
4.60762829e-01 -1.61317185e-01 2.22173512e-01 4.51264203e-01
1.94479018e-01 1.08001828e+00 6.92573786e-01 -1.14618018e-01
-1.29507971e+00 -1.32134712e+00 1.25566512e-01 6.81632459e-01
2.88575381e-01 -3.59567463e-01 -9.75219011e-01 -3.13854039e-01
-2.74056673e-01 8.27975214e-01 -7.64697969e-01 2.23643798e-02
-3.84722501e-01 -5.29371560e-01 4.94631737e-01 4.19227242e-01
1.04450572e+00 -1.63923848e+00 -5.62289238e-01 -4.29358557e-02
-1.56429499e-01 -6.61350369e-01 -7.26561487e-01 -7.06201434e-01
-6.06103897e-01 -8.15612853e-01 -1.29148340e+00 -7.43233323e-01
9.08449769e-01 9.97192487e-02 1.00058687e+00 9.93302763e-02
-3.12007457e-01 1.41361818e-01 -4.46039647e-01 -4.58492428e-01
-3.47489417e-01 3.25318843e-01 -6.72587752e-01 2.76763946e-01
-1.81905970e-01 -6.91695571e-01 -1.05830967e+00 3.08607578e-01
-1.25778317e+00 6.53650641e-01 7.15557694e-01 5.60970962e-01
6.81249917e-01 2.06908450e-01 5.02723575e-01 -9.47861671e-01
1.12819588e+00 -2.27332577e-01 -6.68514252e-01 3.26423138e-01
-3.85036916e-01 -1.12403944e-01 5.30619919e-01 -5.97221971e-01
-1.64393342e+00 2.90476177e-02 4.82923910e-02 -2.71573514e-01
-1.72684848e-01 2.59106368e-01 -5.24174273e-01 2.54974037e-01
4.67386842e-01 3.72020930e-01 -4.14115012e-01 -4.92050320e-01
7.14981377e-01 5.68758845e-01 3.65957797e-01 -6.32488906e-01
1.01957059e+00 7.74484873e-01 -1.46763831e-01 -6.18681788e-01
-5.30974150e-01 3.80290478e-01 -5.05015492e-01 -6.07698977e-01
9.72416520e-01 -8.13393354e-01 -5.37041426e-01 7.53830373e-01
-1.04528356e+00 -7.63712525e-01 -4.20529515e-01 -2.58300096e-01
-5.57947278e-01 2.60298736e-02 -5.06996512e-01 -9.76484835e-01
-5.46110213e-01 -1.03385568e+00 1.34759891e+00 7.05803573e-01
-6.07466064e-02 -7.39597857e-01 1.73771352e-01 1.07204288e-01
6.86368525e-01 7.46243417e-01 7.98260152e-01 6.60324991e-01
-1.03316379e+00 -1.40384138e-01 -4.05689836e-01 5.38235381e-02
4.93352383e-01 4.21516240e-01 -1.09982896e+00 -1.36925682e-01
-3.04312080e-01 -1.37705833e-01 7.15896726e-01 2.17609629e-01
1.13547981e+00 -1.84567302e-01 -1.65067643e-01 6.97814822e-01
1.58224690e+00 3.68798703e-01 1.11816978e+00 2.45276123e-01
6.85221612e-01 4.37793106e-01 3.17853898e-01 4.26652014e-01
1.07637592e-01 4.44725037e-01 1.40483707e-01 -2.62609333e-01
-5.67832649e-01 -7.86659181e-01 1.38676345e-01 5.45849442e-01
-1.73057377e-01 -6.38958395e-01 -6.48727477e-01 4.67725575e-01
-1.76548684e+00 -1.03138328e+00 -1.16126075e-01 1.76327312e+00
8.74180436e-01 2.11808234e-01 -3.96514386e-02 -1.63460359e-01
7.49299586e-01 3.43457341e-01 -7.39535391e-01 -8.44461769e-02
-4.42030013e-01 1.76272333e-01 2.82929447e-02 3.04450780e-01
-7.59324610e-01 1.08344674e+00 5.55246067e+00 1.15553021e+00
-1.07890856e+00 8.69935602e-02 1.06895649e+00 -3.06999236e-01
-8.68774474e-01 -6.70878822e-03 -5.65063119e-01 8.07425201e-01
2.44137213e-01 -3.79399151e-01 4.19363618e-01 7.46119261e-01
3.21971565e-01 -2.73234472e-02 -5.15421450e-01 9.59004760e-01
1.89086087e-02 -1.50469232e+00 4.23616886e-01 -3.25700715e-02
1.21580303e+00 -5.16431272e-01 5.38281798e-01 -1.11701600e-01
2.46670648e-01 -1.02280974e+00 9.76749718e-01 1.01775622e+00
1.14431489e+00 -8.69859040e-01 2.62134969e-01 9.67475586e-03
-1.38314068e+00 1.43806040e-01 -3.21341634e-01 1.70853466e-01
4.31181729e-01 4.98117477e-01 -1.31383196e-01 5.79552472e-01
5.24149179e-01 6.53621256e-01 -7.20489621e-01 8.57301712e-01
-8.66412520e-01 5.72516739e-01 8.26977491e-02 -2.07025319e-01
1.56921558e-02 -4.93907094e-01 3.59413803e-01 9.67447579e-01
4.20265049e-01 2.31137499e-02 -2.25897312e-01 1.55678809e+00
-1.56321779e-01 2.06944905e-02 -4.16490346e-01 -2.15685263e-01
4.81142402e-01 1.55712843e+00 -8.91966581e-01 -3.10679168e-01
1.09413184e-01 1.16440260e+00 1.75966159e-01 5.77311099e-01
-9.72516954e-01 -5.22290349e-01 3.97058487e-01 3.10550220e-02
1.77414581e-01 -2.90204793e-01 -6.02176785e-01 -1.06434381e+00
1.08232513e-01 -6.07227206e-01 -1.10456482e-01 -1.38870120e+00
-1.44004035e+00 9.48137879e-01 -1.14267744e-01 -1.23767292e+00
2.52362192e-01 -1.85135186e-01 -1.14553833e+00 1.15009522e+00
-1.37538266e+00 -1.55872738e+00 -9.75073278e-01 4.64622557e-01
6.18176520e-01 -1.95544004e-01 7.39448965e-01 1.92847878e-01
-5.40215433e-01 4.00973290e-01 1.38480917e-01 1.35651365e-01
5.99556804e-01 -1.11892653e+00 6.35637641e-01 7.28846431e-01
-2.52037775e-03 3.93467963e-01 4.55404073e-01 -9.57044303e-01
-1.13070142e+00 -1.20417261e+00 5.25110662e-01 -4.33298856e-01
3.70518833e-01 -6.64010048e-01 -4.79554713e-01 2.34021366e-01
5.61910272e-01 -6.42581224e-01 4.47968036e-01 -3.80613178e-01
-5.07384203e-02 -1.68467924e-01 -9.56596255e-01 1.17921329e+00
1.16396558e+00 -4.13547009e-01 4.28345129e-02 2.54772544e-01
6.58219993e-01 -4.68157142e-01 -5.89695692e-01 -5.46446517e-02
6.43133819e-01 -8.63311172e-01 7.45192885e-01 -1.32686287e-01
1.00594056e+00 -5.50779343e-01 8.15069154e-02 -1.17056656e+00
-4.40783024e-01 -8.22814822e-01 2.34824643e-01 1.79069722e+00
4.14719790e-01 -2.41027936e-01 1.00128198e+00 7.04407275e-01
1.58021003e-01 -5.94396591e-01 -4.14612710e-01 -6.51746333e-01
-1.26200505e-02 -1.58768967e-01 1.10922837e+00 5.87121189e-01
-5.63681304e-01 3.21482807e-01 -4.94974554e-01 -2.40767956e-01
5.42151153e-01 5.74875414e-01 1.00223172e+00 -6.19923651e-01
-7.37634376e-02 -7.16876090e-01 9.95688960e-02 -7.76845574e-01
-2.27732807e-01 -6.15843832e-01 -1.62591055e-01 -2.00249338e+00
1.92369685e-01 -5.09868026e-01 1.97918154e-02 3.22494060e-01
-2.01024458e-01 4.24305171e-01 5.11137903e-01 2.64614582e-01
-3.31548989e-01 9.98332441e-01 1.81954527e+00 -3.13705266e-01
-4.28219825e-01 -1.60110950e-01 -8.53536367e-01 3.36944312e-01
1.21311307e+00 -2.99583942e-01 -7.27796316e-01 -6.86155915e-01
2.62876838e-01 -1.33182928e-01 3.51015896e-01 -9.74057436e-01
7.67734125e-02 -2.90929526e-01 7.63879895e-01 -6.45134389e-01
4.40407544e-01 -4.95163023e-01 6.56020105e-01 4.34746444e-01
-5.00712693e-01 1.52480692e-01 1.49022788e-01 5.13233960e-01
-1.42739266e-01 6.09844886e-02 6.13913715e-01 -4.33302671e-02
-5.59304118e-01 4.86615330e-01 -6.57665730e-02 -1.88564509e-02
9.46988761e-01 9.07808468e-02 -6.54554605e-01 -6.80492997e-01
-3.93967271e-01 1.34376392e-01 6.17324531e-01 6.95209205e-01
8.11963737e-01 -1.82634687e+00 -8.94558191e-01 1.26013204e-01
4.90880571e-02 1.19204834e-01 6.54350519e-01 1.81048602e-01
-7.57273614e-01 1.30426288e-02 -4.25193787e-01 -2.25109607e-01
-1.07429838e+00 4.45222795e-01 -4.58209254e-02 -6.80049732e-02
-6.08832777e-01 7.24478066e-01 4.32702661e-01 -4.09873724e-02
-3.10502723e-02 -7.67653361e-02 -5.07550612e-02 -1.64959989e-02
6.58563852e-01 3.18252504e-01 -5.28127134e-01 -4.60700154e-01
1.29432976e-02 5.90831637e-01 2.78344274e-01 -3.81982625e-01
1.28817058e+00 7.95088708e-02 -8.74702558e-02 2.39706293e-01
7.11463630e-01 2.23055303e-01 -1.64058661e+00 3.16903055e-01
-4.73345548e-01 -7.67157197e-01 -1.62300020e-01 -1.12303090e+00
-1.41147017e+00 9.65099037e-01 6.25455439e-01 7.03023449e-02
1.45824957e+00 -2.73750722e-01 7.49697924e-01 -1.86066285e-01
2.78389245e-01 -9.60220337e-01 3.30060363e-01 -4.60369848e-02
1.38577986e+00 -7.51641035e-01 -7.74259046e-02 -3.48972917e-01
-9.01926279e-01 9.31019366e-01 6.44690812e-01 -2.98544705e-01
5.19957483e-01 1.97449371e-01 3.31066251e-01 -1.37786478e-01
-6.53500259e-01 1.65051743e-02 3.19063604e-01 6.14744842e-01
3.79944861e-01 2.57589936e-01 -1.81043178e-01 3.94789696e-01
-4.49826539e-01 2.69280970e-01 4.28724766e-01 7.98809588e-01
2.47593746e-02 -1.33937764e+00 -1.27712741e-01 1.80186018e-01
5.55616692e-02 -1.28320992e-01 -7.00985372e-01 7.01905191e-01
1.40180439e-01 1.11891961e+00 -8.74112919e-03 -2.28145197e-01
3.69451374e-01 -3.27353954e-01 3.24399441e-01 -3.25019896e-01
-3.86844814e-01 3.76314968e-01 -1.46833867e-01 -3.73875022e-01
-4.36482787e-01 -3.77535969e-01 -8.74080658e-01 -4.97980952e-01
-8.81170332e-02 8.20057020e-02 5.93507886e-01 2.60909498e-01
5.02078772e-01 8.60663474e-01 6.21914923e-01 -8.52775991e-01
1.83364257e-01 -9.49730814e-01 -6.91774964e-01 4.82770622e-01
-9.77753475e-02 -3.27257961e-01 -1.48329055e-02 3.55006099e-01] | [11.610400199890137, -0.38211458921432495] |
067feb5f-b020-4e39-b91b-58fdc4982934 | sentence-pair-embeddings-based-evaluation | null | null | https://aclanthology.org/2022.lrec-1.646 | https://aclanthology.org/2022.lrec-1.646.pdf | Sentence Pair Embeddings Based Evaluation Metric for Abstractive and Extractive Summarization | The development of an automatic evaluation metric remains an open problem in text generation. Widely used evaluation metrics, like ROUGE and BLEU, are based on exact word matching and fail to capture semantic similarity. Recent works, such as BERTScore, MoverScore and, Sentence Mover’s Similarity, are an improvement over these standard metrics as they use the contextualized word or sentence embeddings to capture semantic similarity. We in this work, propose a novel evaluation metric, Sentence Pair EmbEDdings (SPEED) Score, for text generation which is based on semantic similarity between sentence pairs as opposed to earlier approaches. To find semantic similarity between a pair of sentences, we obtain sentence-level embeddings from multiple transformer models pre-trained specifically on various sentence pair tasks such as Paraphrase Detection (PD), Semantic Text Similarity (STS), and Natural Language Inference (NLI). As these sentence pair tasks involve capturing the semantic similarity between a pair of input texts, we leverage these models in our metric computation. Our proposed evaluation metric shows an impressive performance in evaluating both abstractive and extractive summarization models and achieves state-of-the-art results on the SummEval dataset, demonstrating the effectiveness of our approach. Also, we perform the run-time analysis to show that our proposed metric is faster than the current state-of-the-art. | ['Ivan Garibay', 'Ramya Akula'] | null | null | null | null | lrec-2022-6 | ['sentence-embeddings', 'sentence-embeddings', 'extractive-summarization'] | ['methodology', 'natural-language-processing', 'natural-language-processing'] | [ 3.90749067e-01 -5.16683348e-02 -5.38838878e-02 -3.81788433e-01
-1.12397540e+00 -5.08554518e-01 1.02638471e+00 8.48266125e-01
-4.94806081e-01 6.65583909e-01 9.54758406e-01 -1.06974676e-01
-1.61670640e-01 -8.46420825e-01 -3.64232212e-01 -1.54677957e-01
3.63730073e-01 4.58167374e-01 1.49413124e-01 -6.89466894e-01
9.08423126e-01 -1.85771197e-01 -1.46133375e+00 3.02514493e-01
1.28300560e+00 8.36914539e-01 2.28007168e-01 9.51310694e-01
-4.42541152e-01 5.88382185e-01 -9.10620809e-01 -5.88148713e-01
-1.23173028e-01 -8.21865976e-01 -1.28697896e+00 -5.82929671e-01
5.03615022e-01 1.20031141e-01 -2.10275456e-01 9.57374632e-01
7.31039405e-01 4.10084933e-01 7.92844892e-01 -1.03661823e+00
-1.03924024e+00 9.11619723e-01 -1.48605347e-01 4.66293097e-01
9.26149726e-01 -2.93771863e-01 1.54872262e+00 -1.02278960e+00
5.80334902e-01 1.25016856e+00 6.69986665e-01 5.47331750e-01
-9.01975989e-01 -2.28429258e-01 -2.27779731e-01 4.39506471e-01
-9.06628013e-01 -2.84813046e-01 7.55504906e-01 -7.67677203e-02
1.20000374e+00 4.60725218e-01 2.33359903e-01 1.14860404e+00
2.90588468e-01 9.25570667e-01 7.51548707e-01 -6.31740570e-01
6.58636987e-02 -1.61320735e-02 3.26482594e-01 4.87062216e-01
1.57031998e-01 -4.15362090e-01 -6.03586614e-01 -1.36832789e-01
8.41480494e-02 8.37893318e-03 -3.18223566e-01 8.49343538e-02
-1.51018727e+00 1.16322660e+00 4.09534216e-01 6.13646924e-01
-1.95456579e-01 -1.51738683e-02 8.00500453e-01 4.02605176e-01
6.13201320e-01 1.07729149e+00 -2.03931361e-01 -5.02491236e-01
-1.21329010e+00 5.56357026e-01 7.81638086e-01 7.53111601e-01
3.83207858e-01 -3.50016654e-01 -8.71291935e-01 1.04156625e+00
9.62325260e-02 3.30097079e-01 1.33812332e+00 -5.75720608e-01
7.92745829e-01 8.69464636e-01 -1.63881689e-01 -1.18324518e+00
-1.66478053e-01 -2.41667524e-01 -8.64353418e-01 -4.85042691e-01
-7.22574741e-02 2.33428720e-02 -5.42021811e-01 1.65234232e+00
-7.37718958e-03 1.61093429e-01 6.41721725e-01 5.90416610e-01
1.27914572e+00 6.95494115e-01 -2.31547013e-01 -7.28011206e-02
1.20995080e+00 -1.35366440e+00 -6.44129813e-01 -2.36348525e-01
9.97984052e-01 -1.14309788e+00 1.21755517e+00 -1.85671017e-01
-1.33271849e+00 -7.33874083e-01 -1.18116271e+00 -3.75739783e-01
-5.30146241e-01 -1.30690962e-01 2.42816463e-01 2.32805029e-01
-1.07853484e+00 1.16819632e+00 -4.34015632e-01 -7.15272605e-01
-2.71105058e-02 -1.77331299e-01 -1.91036344e-01 -4.89270277e-02
-1.57047272e+00 1.16674066e+00 4.81666356e-01 -4.06085283e-01
-3.33492965e-01 -7.65692890e-01 -1.03609061e+00 2.79719323e-01
-1.64889336e-01 -1.17392743e+00 1.38871455e+00 -4.75233257e-01
-1.34429491e+00 7.87292957e-01 -2.42907092e-01 -7.79942691e-01
1.41809627e-01 -2.23133206e-01 -4.41159815e-01 2.70676047e-01
3.56922239e-01 5.80752432e-01 4.33529198e-01 -6.53999746e-01
-5.93374312e-01 -1.86930820e-01 1.08785681e-01 5.39041817e-01
-6.78251982e-01 1.58158675e-01 1.61390930e-01 -8.70031238e-01
-1.12649366e-01 -4.98354465e-01 -1.29092172e-01 -4.44588780e-01
-4.14874762e-01 -7.63514221e-01 5.83062708e-01 -7.42165625e-01
1.61427665e+00 -1.76311553e+00 2.88142622e-01 -3.34908724e-01
-1.42307160e-03 4.08487827e-01 -4.71523911e-01 9.59009647e-01
1.18094780e-01 1.81545705e-01 -4.75300729e-01 -5.03094018e-01
2.18505830e-01 -1.79518521e-01 -4.74700689e-01 -1.84877187e-01
3.20197642e-01 1.05484557e+00 -1.28220129e+00 -5.93085170e-01
3.99856232e-02 4.20909300e-02 -4.04040337e-01 3.65460068e-01
4.34211269e-02 -1.81879058e-01 -3.39059234e-01 9.60825980e-02
2.38445580e-01 -2.50567850e-02 -1.40702426e-01 -7.19349310e-02
1.77018821e-01 8.11336398e-01 -6.45612359e-01 2.25608397e+00
-8.04942667e-01 5.64688265e-01 -7.86811411e-01 -1.04634631e+00
1.11323380e+00 3.26381326e-01 1.20112590e-01 -7.44094610e-01
4.05396931e-02 3.06550682e-01 -1.93305850e-01 -3.73861849e-01
1.26019919e+00 -9.38635096e-02 -4.33984697e-01 7.98675179e-01
1.64309859e-01 -5.08328557e-01 5.80133677e-01 6.10914052e-01
1.43763840e+00 -1.71364635e-01 5.21228254e-01 -2.30292365e-01
7.22075582e-01 -7.40153342e-02 1.55805126e-02 7.16739118e-01
-1.39674637e-02 8.85183156e-01 3.04155380e-01 -8.97334889e-03
-1.08750713e+00 -1.13455617e+00 1.05371892e-01 9.40642774e-01
2.98018664e-01 -8.43661010e-01 -9.05453026e-01 -8.23877156e-01
-7.22681135e-02 1.07268655e+00 -5.09837866e-01 -5.48445821e-01
-5.19205868e-01 -6.40560210e-01 6.67255461e-01 7.40667999e-01
2.53152043e-01 -1.05475974e+00 -3.56028020e-01 3.34293902e-01
-5.35125911e-01 -1.01261485e+00 -8.40254843e-01 -3.61857474e-01
-9.55935955e-01 -8.64976227e-01 -6.99279606e-01 -9.56894219e-01
3.64180207e-01 3.30833197e-01 1.30672920e+00 -7.82194585e-02
-1.09455287e-01 3.09889913e-01 -8.48216057e-01 -2.59874672e-01
-6.17006063e-01 3.75614166e-01 1.10587608e-02 -2.57550657e-01
3.42838556e-01 -4.58641797e-01 -6.81188464e-01 -1.18015222e-01
-1.07577503e+00 8.29392753e-04 5.39505064e-01 1.02268028e+00
2.91473478e-01 -4.05184418e-01 9.92629945e-01 -6.13667786e-01
1.53438401e+00 -5.86102307e-01 3.84810716e-01 5.46639383e-01
-7.44230747e-01 5.34834027e-01 9.88955975e-01 -4.89064716e-02
-8.29495251e-01 -8.40753555e-01 -1.80932775e-01 -1.02380164e-01
7.12625459e-02 7.26641119e-01 2.74565995e-01 4.74553585e-01
7.43524849e-01 4.52356696e-01 2.13586798e-04 -4.41074908e-01
4.92358357e-01 1.06853545e+00 5.80884099e-01 -3.11824977e-01
6.78806245e-01 5.66545501e-02 -1.88920125e-01 -5.12698114e-01
-1.21949792e+00 -7.53036797e-01 -3.83220017e-01 2.64960557e-01
6.86972916e-01 -6.74228728e-01 -2.10564613e-01 5.88277727e-02
-1.48753858e+00 3.72341454e-01 -4.75047797e-01 4.09460574e-01
-5.29675484e-01 7.79181540e-01 -3.80131811e-01 -4.10174668e-01
-1.11545634e+00 -7.59283245e-01 1.31873798e+00 2.18996346e-01
-7.20474064e-01 -1.39938879e+00 4.73352969e-01 5.26697993e-01
6.67860866e-01 1.48236930e-01 9.58813548e-01 -1.10300565e+00
1.00650504e-01 -3.45359504e-01 -1.34439245e-01 5.41110992e-01
2.62422472e-01 -1.52773917e-01 -5.30977368e-01 -2.10905954e-01
4.41343524e-03 -3.37523997e-01 1.02622914e+00 8.37000459e-02
1.04229391e+00 -3.75352979e-01 -8.99571031e-02 1.67178303e-01
1.18532753e+00 -3.32004964e-01 5.95349014e-01 2.46882126e-01
4.81471360e-01 4.61355031e-01 8.16502631e-01 4.67265248e-01
5.68598628e-01 7.16055155e-01 7.34445080e-02 2.39750564e-01
-2.95921206e-01 -5.87270916e-01 5.86061239e-01 1.56046343e+00
2.30267256e-01 -4.01582152e-01 -6.05928063e-01 6.60402715e-01
-2.05084109e+00 -1.24425352e+00 -1.12609170e-01 2.20311952e+00
1.07528532e+00 7.96475857e-02 -7.42712840e-02 3.28552306e-01
6.57226622e-01 3.76158059e-01 -5.63544258e-02 -1.03977656e+00
-7.43044540e-02 6.54919624e-01 1.13879472e-01 4.74492103e-01
-8.29340518e-01 9.31865513e-01 5.69217634e+00 1.06688547e+00
-7.18482494e-01 1.58213019e-01 1.98700175e-01 -1.98925495e-01
-6.05441689e-01 2.78192628e-02 -7.38669395e-01 6.34072244e-01
1.21443534e+00 -8.71457815e-01 1.28474995e-01 6.65197253e-01
1.42860189e-01 1.03243411e-01 -1.28205538e+00 8.73307109e-01
8.02507341e-01 -1.49163580e+00 3.81759465e-01 -5.73103011e-01
8.12811852e-01 -1.20383129e-01 -2.10068926e-01 4.40112770e-01
2.62114316e-01 -9.54557002e-01 4.78659093e-01 5.41581690e-01
3.98774594e-01 -7.25217164e-01 1.03183234e+00 1.33250713e-01
-1.04453743e+00 9.27702636e-02 -4.43126231e-01 -1.91784531e-01
4.15259600e-01 5.79250991e-01 -8.94847870e-01 9.50334489e-01
2.08635792e-01 9.78686988e-01 -8.56221378e-01 8.77825499e-01
-3.56444925e-01 3.02460730e-01 1.73935384e-01 -5.81033289e-01
4.42817152e-01 -2.10136510e-02 6.52685821e-01 1.25512516e+00
5.93647063e-01 -4.38179433e-01 7.13233789e-03 7.27233648e-01
-3.21105570e-01 4.23731923e-01 -6.19150579e-01 -1.86968744e-01
6.14751935e-01 1.23816168e+00 -2.09187776e-01 -5.20300746e-01
-1.70857280e-01 1.29839194e+00 3.67674708e-01 -2.08384007e-01
-7.10495949e-01 -1.08569813e+00 6.92483425e-01 -2.57441610e-01
1.11154348e-01 -6.76944032e-02 -3.48746270e-01 -1.14595008e+00
3.31284970e-01 -5.89290500e-01 3.79077822e-01 -6.48872912e-01
-1.55079985e+00 6.16021156e-01 -4.86031845e-02 -1.29213309e+00
-5.45759082e-01 -2.40766585e-01 -1.19462848e+00 7.40111649e-01
-1.41762626e+00 -9.18635190e-01 -2.21689016e-01 2.61901826e-01
1.02007961e+00 -2.26582512e-01 9.50751424e-01 -9.28974827e-04
-4.15365547e-01 8.11344504e-01 3.42362791e-01 1.86417662e-02
7.97367096e-01 -1.48772061e+00 7.84614682e-01 8.77867520e-01
3.60588312e-01 5.56440353e-01 7.56803274e-01 -3.86539668e-01
-1.07639563e+00 -1.15419161e+00 1.69495082e+00 -6.21145844e-01
8.13013613e-01 -1.09427035e-01 -6.81622028e-01 1.45644128e-01
5.16755939e-01 -5.57505310e-01 8.62454116e-01 -7.86439981e-03
-1.89386800e-01 2.70232148e-02 -9.30508316e-01 6.94259048e-01
1.17719889e+00 -4.54617172e-01 -1.34242404e+00 5.68991184e-01
1.16095388e+00 -1.78148597e-01 -9.68774796e-01 2.14815810e-01
3.44026893e-01 -7.87309945e-01 8.47150564e-01 -9.43164110e-01
1.19417751e+00 -9.86579619e-03 -2.16246352e-01 -1.86169744e+00
-2.07000509e-01 -5.20857811e-01 -1.10528618e-01 1.45296645e+00
5.37768781e-01 -4.58015978e-01 2.66414225e-01 1.64743304e-01
-5.21267414e-01 -8.61985445e-01 -8.31665337e-01 -1.03252482e+00
3.48841518e-01 -8.50628316e-02 8.58979106e-01 7.87158906e-01
5.56349754e-01 9.47359085e-01 1.10943764e-02 -4.25405324e-01
2.44230047e-01 3.35840434e-01 5.40655553e-01 -1.12610447e+00
-6.50458857e-02 -7.56296158e-01 -6.53127789e-01 -7.76844025e-01
4.64801699e-01 -1.39057517e+00 -2.13355571e-02 -2.08056164e+00
2.97058821e-01 7.90959671e-02 -3.72807890e-01 1.34590045e-02
-5.72839439e-01 4.62974384e-02 2.48672426e-01 -5.58761768e-02
-7.37431049e-01 1.01575172e+00 1.13541341e+00 -2.86290556e-01
1.18427135e-01 -2.88273901e-01 -9.88313138e-01 4.47612315e-01
8.34093273e-01 -4.83267754e-01 -3.90075922e-01 -6.60213351e-01
1.36412725e-01 -1.39549375e-01 1.82005495e-01 -1.14653003e+00
2.98055083e-01 5.64130321e-02 -9.42533836e-02 -3.66847038e-01
8.68521482e-02 -1.42539084e-01 -6.30822480e-01 3.69542807e-01
-9.22862291e-01 6.25264227e-01 -3.09738904e-01 4.60531443e-01
-6.65646315e-01 -7.93913662e-01 5.34177005e-01 2.01238450e-02
-4.26132768e-01 4.98316847e-02 3.30010355e-02 6.28659427e-01
7.58159161e-01 -1.32629558e-01 -4.30139810e-01 -2.86396474e-01
-8.56538564e-02 2.70295054e-01 3.20881814e-01 8.60534608e-01
8.28492224e-01 -1.55606997e+00 -1.19412351e+00 -2.49359965e-01
4.30643648e-01 -1.54710591e-01 9.73795429e-02 7.09008813e-01
-3.93309146e-01 5.73418677e-01 7.61942938e-02 -2.64198065e-01
-1.35393298e+00 2.50492007e-01 -1.53812900e-01 -7.89594412e-01
-3.14715862e-01 7.39534914e-01 -3.23464364e-01 -6.03769243e-01
-1.25980660e-01 -4.74111915e-01 -3.30384701e-01 1.62889913e-01
6.38170123e-01 6.49027169e-01 2.59312451e-01 -4.75995749e-01
-2.48898908e-01 4.76350397e-01 -2.79743999e-01 -1.10863052e-01
1.14868975e+00 -1.69680826e-02 -1.87188208e-01 3.11559886e-01
1.52041984e+00 -2.71444678e-01 -1.21389732e-01 -3.59199494e-01
1.81029782e-01 -3.79146367e-01 -1.06291845e-01 -5.83938897e-01
-4.24060047e-01 1.00171053e+00 2.76725143e-01 3.31173301e-01
8.24127555e-01 -8.53487104e-03 1.53161037e+00 5.59815466e-01
-7.61213228e-02 -1.41382933e+00 4.46405143e-01 8.69370997e-01
1.02098179e+00 -1.20448756e+00 -1.50936842e-01 7.26511627e-02
-7.70845115e-01 1.14537191e+00 4.96249706e-01 -1.77782848e-01
9.53075588e-02 -2.54320681e-01 -3.34427387e-01 4.94041294e-03
-9.05399203e-01 -1.82970718e-01 5.49458027e-01 3.09350938e-01
7.72412181e-01 3.02417036e-02 -9.39533472e-01 6.00222409e-01
-8.71525466e-01 -1.57576606e-01 5.44874787e-01 7.90183008e-01
-6.54110849e-01 -1.20467055e+00 1.37621149e-01 7.41865098e-01
-3.33497554e-01 -4.05342907e-01 -6.22288287e-01 2.02797800e-01
-4.41697747e-01 1.19255924e+00 2.28779942e-01 -4.15401459e-01
4.09166634e-01 1.89689532e-01 5.67413151e-01 -1.01121616e+00
-8.22581887e-01 -9.73287404e-01 1.87523782e-01 -2.59317815e-01
-2.87128419e-01 -4.63778108e-01 -1.13134718e+00 -3.24380159e-01
-3.62142354e-01 3.90789926e-01 6.41265273e-01 1.12986934e+00
5.85802436e-01 4.49842125e-01 9.42023218e-01 -6.00795627e-01
-1.03715575e+00 -1.43066835e+00 -2.51165807e-01 9.58555996e-01
-2.41209697e-02 -3.13759476e-01 -4.20985729e-01 -2.31519371e-01] | [11.601219177246094, 9.178308486938477] |
9fdf994a-5e0e-439a-97c3-be99b3862bd2 | quantitative-day-trading-from-natural | null | null | https://aclanthology.org/2021.naacl-main.316 | https://aclanthology.org/2021.naacl-main.316.pdf | Quantitative Day Trading from Natural Language using Reinforcement Learning | It is challenging to design profitable and practical trading strategies, as stock price movements are highly stochastic, and the market is heavily influenced by chaotic data across sources like news and social media. Existing NLP approaches largely treat stock prediction as a classification or regression problem and are not optimized to make profitable investment decisions. Further, they do not model the temporal dynamics of large volumes of diversely influential text to which the market responds quickly. Building on these shortcomings, we propose a deep reinforcement learning approach that makes time-aware decisions to trade stocks while optimizing profit using textual data. Our method outperforms state-of-the-art in terms of risk-adjusted returns in trading simulations on two benchmarks: Tweets (English) and financial news (Chinese) pertaining to two major indexes and four global stock markets. Through extensive experiments and studies, we build the case for our method as a tool for quantitative trading. | ['Rajiv Ratn Shah', 'Shivam Agarwal', 'Arnav Wadhwa', 'Ramit Sawhney'] | 2021-06-01 | null | null | null | naacl-2021-4 | ['stock-prediction'] | ['time-series'] | [-8.69389474e-01 -3.26728761e-01 -6.21600032e-01 -5.02084531e-02
-6.99180245e-01 -8.79467070e-01 8.98737907e-01 2.08003953e-01
-4.96677279e-01 9.07498658e-01 4.13265646e-01 -4.73321229e-01
3.17491516e-02 -1.33821642e+00 -6.25664413e-01 -3.11620802e-01
-4.95927602e-01 8.55920434e-01 2.78655082e-01 -4.10302103e-01
6.38192475e-01 3.58803421e-01 -9.97341394e-01 -8.70535672e-02
5.59079826e-01 1.47006571e+00 -2.91790158e-01 2.89703399e-01
-5.58951080e-01 1.39806843e+00 -6.91504598e-01 -6.34149611e-01
5.97518742e-01 -2.45391592e-01 -4.04623628e-01 -2.31050029e-01
-5.91907382e-01 -5.64466000e-01 -2.15048403e-01 9.60642457e-01
1.98826298e-01 -2.35899702e-01 5.50634205e-01 -1.19307160e+00
-7.99485505e-01 1.18039417e+00 -7.36232579e-01 5.69702208e-01
-2.48884737e-01 1.60713837e-01 1.62725627e+00 -5.17381728e-01
5.29383659e-01 8.64860177e-01 5.30811489e-01 -1.67574853e-01
-9.99911427e-01 -8.25452566e-01 2.60269105e-01 -2.13963777e-01
-5.45806348e-01 3.23872715e-02 7.36557782e-01 -5.19200325e-01
9.93504345e-01 1.09473690e-01 1.19948542e+00 1.08805466e+00
6.03865027e-01 1.03911078e+00 1.09896493e+00 2.08795175e-01
3.24625134e-01 1.09774992e-01 -2.11163372e-01 1.30812386e-02
3.28908563e-01 6.73354864e-01 -6.24203146e-01 -4.37077522e-01
6.56077027e-01 1.28823057e-01 3.04957181e-01 1.24224618e-01
-1.38804758e+00 1.44900513e+00 2.81626731e-01 3.79054785e-01
-9.02502835e-01 3.65576118e-01 4.27502602e-01 9.55472052e-01
1.02330923e+00 7.87636161e-01 -7.93282986e-01 -5.39256692e-01
-1.13789237e+00 8.22196007e-01 1.37030745e+00 5.12683272e-01
2.72461653e-01 3.58797312e-01 -6.69289157e-02 2.95087934e-01
4.23771352e-01 7.43988872e-01 7.59979844e-01 -7.69031703e-01
4.90231931e-01 4.22024846e-01 5.88090003e-01 -1.12130976e+00
-4.84289497e-01 -6.05948210e-01 -3.12768400e-01 2.23681703e-01
5.48870921e-01 -6.40736520e-01 -3.12921375e-01 1.18310916e+00
-9.48054343e-02 -1.46935238e-02 2.26426497e-01 5.98148108e-01
1.49055526e-01 9.33187962e-01 -1.46811292e-01 -5.30981064e-01
9.69894469e-01 -8.76444221e-01 -6.69523597e-01 -2.69676387e-01
3.18795264e-01 -7.03823388e-01 5.50964236e-01 4.22429085e-01
-1.27796352e+00 1.40131310e-01 -7.65954912e-01 6.56100035e-01
-5.01847565e-01 -5.08448601e-01 6.67263508e-01 2.78862193e-02
-7.91093171e-01 8.85372818e-01 -7.74529159e-01 5.20488381e-01
3.91242355e-01 1.57405809e-01 6.31243825e-01 1.03719330e+00
-1.56383550e+00 1.00550115e+00 2.53230572e-01 -7.22507536e-02
-5.81163287e-01 -8.95136416e-01 -1.47449926e-01 1.28425494e-01
5.91264307e-01 -1.55605525e-01 1.66876185e+00 -9.67108846e-01
-1.63128710e+00 4.42394316e-01 5.42289138e-01 -1.24392831e+00
1.23645079e+00 -4.24424410e-01 -4.78357166e-01 -6.20036945e-02
2.58208625e-02 8.93442929e-02 6.91549599e-01 -7.33173072e-01
-9.68139470e-01 1.68265179e-02 -1.05203621e-01 -6.82740435e-02
-1.82880417e-01 1.87093362e-01 2.59356111e-01 -1.34002340e+00
-2.83088267e-01 -6.77857697e-01 -3.55465889e-01 -5.84586740e-01
-3.10669273e-01 -1.05142750e-01 6.23482823e-01 -6.22941196e-01
1.31422114e+00 -1.63887608e+00 -1.77154005e-01 3.99908304e-01
1.68614924e-01 -2.26790279e-01 2.94662178e-01 8.17019999e-01
2.33787671e-01 4.63022113e-01 1.78240072e-02 1.21267088e-01
5.18253624e-01 -5.66175543e-02 -1.13118410e+00 4.14504886e-01
2.49560222e-01 1.45903003e+00 -7.10810304e-01 -2.08855066e-02
-2.09635094e-01 -1.85717359e-01 -3.22783709e-01 1.36897549e-01
-8.56060028e-01 8.64997432e-02 -9.08823252e-01 7.69584298e-01
9.56881940e-02 -6.37193024e-01 -6.09808303e-02 4.96645659e-01
-4.45132136e-01 3.72412443e-01 -1.00896394e+00 5.01524508e-01
-1.56255364e-01 4.92150813e-01 -2.83385009e-01 -7.82054365e-01
9.34625864e-01 6.06275685e-02 7.67177284e-01 -1.09998786e+00
2.41171166e-01 6.09969139e-01 -4.31767339e-03 5.00122793e-02
6.14134789e-01 -3.91777337e-01 -3.47571909e-01 1.31225109e+00
-5.94838023e-01 -2.46245995e-01 2.63418794e-01 -5.74536286e-02
9.18197036e-01 -1.86019555e-01 2.05009356e-01 -3.26486975e-01
-1.08944938e-01 7.00619668e-02 5.17034888e-01 7.21168280e-01
-1.00860618e-01 -5.19852620e-03 1.14786005e+00 -7.67428696e-01
-9.77120340e-01 -8.44802439e-01 2.80147325e-02 9.65054154e-01
-1.24329478e-01 6.87243938e-02 -3.40202153e-01 -5.08980274e-01
6.72860324e-01 9.90973294e-01 -6.80954039e-01 2.55242884e-01
-4.48358089e-01 -1.32163060e+00 2.69974377e-02 3.69776905e-01
3.15927356e-01 -1.52783358e+00 -7.94331312e-01 7.98066318e-01
2.41285250e-01 -8.33187163e-01 -3.74262035e-01 1.98536530e-01
-6.39673233e-01 -1.12015748e+00 -8.27744126e-01 -1.58029526e-01
-5.85113913e-02 -2.94390619e-01 1.51484072e+00 -2.31042534e-01
3.66598874e-01 1.83351606e-01 -2.06108555e-01 -1.17429817e+00
-5.59028924e-01 2.05122024e-01 -1.15311541e-01 1.99251547e-01
4.57869500e-01 -4.34176743e-01 -7.18274951e-01 2.29642600e-01
-7.72908151e-01 -3.22587669e-01 5.53150058e-01 7.46523023e-01
6.27875447e-01 2.02009961e-01 1.10658753e+00 -1.00437164e+00
1.10010362e+00 -8.08358967e-01 -1.57648623e+00 1.86918467e-01
-1.17181945e+00 2.20408991e-01 3.80097806e-01 -4.98126268e-01
-7.24217594e-01 -5.61167002e-01 2.41987735e-01 -8.35042000e-02
6.43310487e-01 1.03388381e+00 8.21065485e-01 2.58901119e-01
7.02689365e-02 4.46134418e-01 1.08342670e-01 -2.02507570e-01
-3.94738726e-02 2.83738405e-01 4.72144037e-03 -1.63700119e-01
1.09911048e+00 3.66515428e-01 -3.92079234e-01 -2.65501767e-01
-8.54754746e-01 1.02023184e-01 -2.56678045e-01 -1.89478889e-01
4.67499793e-01 -9.38283741e-01 -8.78105402e-01 8.57415974e-01
-6.54031813e-01 -6.49910688e-01 -5.55433154e-01 5.06264150e-01
-6.19526267e-01 -2.94786751e-01 -1.10434055e+00 -9.85689044e-01
-5.33674061e-01 -9.20366824e-01 6.39051139e-01 1.41491488e-01
-3.60927045e-01 -1.28780651e+00 5.32625139e-01 -3.03894207e-02
9.66951907e-01 5.50534010e-01 8.21730435e-01 -1.25723755e+00
-9.41203177e-01 -4.66104954e-01 -5.65782189e-02 1.61202952e-01
4.60136570e-02 2.14176461e-01 -6.60949647e-01 2.28643585e-02
1.72331765e-01 -4.99001414e-01 8.84783089e-01 6.21685982e-01
6.82576120e-01 -6.85810208e-01 1.01104088e-01 3.32849920e-01
1.24984062e+00 4.38147873e-01 2.12444156e-01 9.56165314e-01
1.08622506e-01 6.38596416e-01 8.23668242e-01 9.63226318e-01
3.21097404e-01 7.75623471e-02 4.18312103e-01 1.92474425e-01
1.02804506e+00 -3.13041866e-01 5.83665967e-01 9.38514769e-01
8.58400762e-03 -2.24165395e-01 -9.04729903e-01 3.27082843e-01
-1.81806815e+00 -1.22853446e+00 3.15266520e-01 1.74348462e+00
1.00963330e+00 8.01032066e-01 6.71705008e-01 -2.26070762e-01
1.07112691e-01 7.74511755e-01 -9.33150649e-01 -6.77729994e-02
-3.80899489e-01 -1.15708284e-01 9.88952458e-01 2.88400203e-01
-1.10490918e+00 9.30372000e-01 7.00209427e+00 5.48938394e-01
-1.34587610e+00 -2.41238549e-01 1.17318320e+00 -4.88410681e-01
-7.08247662e-01 -3.47210556e-01 -8.76460314e-01 7.51618803e-01
1.41706645e+00 -6.98569775e-01 4.77906972e-01 7.28764892e-01
2.62546211e-01 5.24599925e-02 -7.63786852e-01 5.93581021e-01
-6.78080022e-01 -2.02278399e+00 -1.65612563e-01 3.52869719e-01
6.80738807e-01 6.17792666e-01 4.61139798e-01 3.66146326e-01
8.96690726e-01 -9.24590409e-01 1.19402969e+00 8.54829490e-01
-1.22199073e-01 -1.06522548e+00 7.33292460e-01 3.32856297e-01
-8.49824965e-01 -2.80864954e-01 -1.05957225e-01 -6.66361749e-02
2.60708332e-01 5.85947454e-01 -4.06413972e-01 4.68881540e-02
6.66949391e-01 1.02888143e+00 -1.33019775e-01 4.82342601e-01
1.99748622e-03 8.36680651e-01 -3.01690876e-01 -6.54521585e-01
8.02897036e-01 -5.35853446e-01 3.07714462e-01 7.07604885e-01
4.18482125e-01 -1.04114763e-01 4.54139598e-02 1.13257933e+00
-3.85938108e-01 2.03260541e-01 -5.73646307e-01 -8.06622922e-01
3.45641196e-01 7.66643524e-01 -9.73463356e-01 -1.99801758e-01
-8.36301506e-01 3.17736983e-01 8.17799792e-02 3.14471185e-01
-7.71905839e-01 -7.43912235e-02 6.04858994e-01 2.62834787e-01
4.59565908e-01 -1.67740509e-01 -3.34400356e-01 -1.20550454e+00
1.95175521e-02 -1.03746223e+00 2.78538465e-01 -2.57463545e-01
-1.74401307e+00 3.99083555e-01 -2.34086230e-01 -1.34684265e+00
-6.89883769e-01 -5.63517988e-01 -7.51057982e-01 5.21016359e-01
-2.01861453e+00 -3.62034798e-01 7.47948647e-01 2.77945548e-01
2.41024926e-01 -7.80440509e-01 2.88717568e-01 -1.88364446e-01
-3.82294387e-01 -7.60383625e-03 6.98807001e-01 4.65902746e-01
2.41796732e-01 -1.57911420e+00 1.03723943e+00 5.16933680e-01
5.14173269e-01 9.20830220e-02 6.56150162e-01 -8.93556714e-01
-1.34822035e+00 -9.15312767e-01 8.30051780e-01 -4.19028670e-01
1.88250184e+00 -1.94580391e-01 -8.86258781e-01 6.97476506e-01
5.33624351e-01 -1.97181270e-01 6.17795587e-01 -1.59768507e-01
-9.74946395e-02 -2.42075875e-01 -7.67554581e-01 5.94586909e-01
1.76413417e-01 -2.19632104e-01 -6.10206723e-01 5.55351615e-01
9.36089218e-01 -2.35442176e-01 -1.05249512e+00 -8.22070912e-02
5.31918168e-01 -1.08366323e+00 7.58986294e-01 -5.75772643e-01
5.20738721e-01 4.51735467e-01 1.83981016e-01 -1.48654664e+00
9.19608027e-02 -1.10732746e+00 -4.39228684e-01 7.94962108e-01
8.78470421e-01 -1.05871475e+00 8.72156978e-01 6.68083489e-01
5.26894152e-01 -9.38672781e-01 -7.98024714e-01 -6.56376660e-01
5.15338838e-01 -3.29651088e-01 1.11537063e+00 9.98820066e-01
-1.85049266e-01 -1.16653867e-01 -5.19608319e-01 -3.82666379e-01
5.53844690e-01 9.36106503e-01 4.89639074e-01 -1.29574466e+00
-3.73580247e-01 -1.07084072e+00 2.32654512e-01 -5.22220731e-01
1.66076824e-01 -4.23157603e-01 -2.81992525e-01 -8.94425452e-01
-1.65963754e-01 -3.52487624e-01 -6.31020248e-01 1.19684316e-01
3.17286402e-01 -1.96155339e-01 1.95748270e-01 4.26034242e-01
-3.92136306e-01 8.04279804e-01 1.16470933e+00 -2.60177016e-01
-3.33405554e-01 4.60731804e-01 -8.20404053e-01 7.37493575e-01
9.87086356e-01 -4.92320687e-01 -1.22170731e-01 8.62087868e-03
1.03664386e+00 2.97208309e-01 1.06264666e-01 -3.69468272e-01
4.97184098e-02 -7.41765738e-01 3.40892255e-01 -9.25953448e-01
-1.37001440e-01 -5.66545606e-01 -3.80919091e-02 8.06649983e-01
-5.63593388e-01 6.69648647e-01 2.96208840e-02 7.83272028e-01
-7.24031985e-01 7.70868137e-02 4.95528877e-01 -3.60713035e-01
-5.13565421e-01 6.49599314e-01 -6.03489101e-01 6.73429847e-01
1.16604900e+00 2.29997456e-01 -3.10252398e-01 -7.82568574e-01
-3.46459508e-01 6.64438367e-01 1.41117930e-01 4.98957068e-01
3.38382572e-01 -1.25540578e+00 -9.90877092e-01 -5.59703186e-02
-2.37211347e-01 -3.70799899e-01 -3.35273057e-01 5.58208466e-01
-8.87142420e-01 5.49134910e-01 -7.00728744e-02 2.59081591e-02
-2.42160648e-01 4.77783799e-01 6.10906184e-01 -7.74444222e-01
-5.95485687e-01 6.37607396e-01 -9.43243355e-02 -1.43086657e-01
2.87391275e-01 -8.64401340e-01 -2.90792316e-01 9.76071417e-01
6.93119228e-01 1.67062923e-01 -2.37165242e-01 -2.98954695e-01
6.21089675e-02 1.72335148e-01 -1.17017269e-01 -3.79906863e-01
2.00188637e+00 9.45490599e-02 2.84334589e-02 1.03643107e+00
9.53019440e-01 1.14261523e-01 -1.41061103e+00 -4.17722702e-01
8.35589349e-01 -2.68741727e-01 1.08150013e-01 -6.57462418e-01
-1.55666244e+00 3.76306415e-01 -1.27050832e-01 9.47781801e-01
6.45902395e-01 -1.70491591e-01 1.14474380e+00 4.74819064e-01
1.83438301e-01 -1.60370219e+00 1.52531743e-01 5.53174436e-01
1.05252635e+00 -1.33851326e+00 -1.76418498e-02 5.38150787e-01
-1.08103073e+00 1.09942579e+00 4.92481664e-02 -5.18888354e-01
1.44047093e+00 6.87605560e-01 3.50431502e-01 -3.21805179e-01
-1.18056881e+00 2.59181827e-01 1.63961038e-01 -1.97171167e-01
1.58247352e-01 6.12311140e-02 -6.67057335e-02 8.39532256e-01
-5.88159084e-01 -3.81161310e-02 6.68263316e-01 1.07390475e+00
-4.92844492e-01 -9.24644113e-01 -5.44801950e-02 1.00157964e+00
-1.15806508e+00 -1.39687732e-01 -5.24186015e-01 1.08532894e+00
-8.71556401e-01 3.56287360e-01 3.79273981e-01 -1.93421960e-01
1.85986459e-01 -2.39081308e-02 -4.03448433e-01 -2.61239082e-01
-9.76754725e-01 3.69637936e-01 -1.54541597e-01 -6.03888035e-01
-3.73983622e-01 -1.08982027e+00 -1.21911371e+00 -7.26556361e-01
1.21701360e-01 2.02549264e-01 1.70217559e-01 1.05242395e+00
7.97680467e-02 3.47102940e-01 9.70010698e-01 -5.31812847e-01
-1.32447660e+00 -6.58222973e-01 -1.28324461e+00 -1.93160400e-02
6.03374064e-01 -5.82737446e-01 -4.04587954e-01 -4.29466635e-01] | [4.467689037322998, 4.129909992218018] |
167089ec-aa4f-4aa8-8d0b-e03568873a1f | adaptive-navigation-scheme-for-optimal-deep | 1906.04888 | null | https://arxiv.org/abs/1906.04888v1 | https://arxiv.org/pdf/1906.04888v1.pdf | Adaptive Navigation Scheme for Optimal Deep-Sea Localization Using Multimodal Perception Cues | Underwater robot interventions require a high level of safety and reliability. A major challenge to address is a robust and accurate acquisition of localization estimates, as it is a prerequisite to enable more complex tasks, e.g. floating manipulation and mapping. State-of-the-art navigation in commercial operations, such as oil & gas production (OGP), rely on costly instrumentation. These can be partially replaced or assisted by visual navigation methods, especially in deep-sea scenarios where equipment deployment has high costs and risks. Our work presents a multimodal approach that adapts state-of-the-art methods from on-land robotics, i.e., dense point cloud generation in combination with plane representation and registration, to boost underwater localization performance. A two-stage navigation scheme is proposed that initially generates a coarse probabilistic map of the workspace, which is used to filter noise from computed point clouds and planes in the second stage. Furthermore, an adaptive decision-making approach is introduced that determines which perception cues to incorporate into the localization filter to optimize accuracy and computation performance. Our approach is investigated first in simulation and then validated with data from field trials in OGP monitoring and maintenance scenarios. | ['Andreas Birk', 'Sören Schwertfeger', 'Christian A. Mueller', 'Arturo Gomez Chavez', 'Qingwen Xu'] | 2019-06-12 | null | null | null | null | ['point-cloud-generation'] | ['computer-vision'] | [ 1.94395036e-01 -1.23181321e-01 6.29903018e-01 -3.90037924e-01
-5.37420094e-01 -6.10716343e-01 2.16090664e-01 6.55281663e-01
-9.78310227e-01 7.69151509e-01 -2.77573556e-01 -1.36502311e-01
-6.42672837e-01 -1.00859082e+00 -7.64648199e-01 -8.02899778e-01
-3.95868719e-01 6.87018394e-01 5.21970093e-01 -6.82837367e-01
6.44587934e-01 8.68722498e-01 -1.83087468e+00 -4.28923637e-01
1.10030282e+00 8.61766160e-01 8.48959744e-01 5.59045196e-01
-1.76068217e-01 5.94336428e-02 -3.37827742e-01 -2.61334032e-02
3.04050237e-01 -1.82007626e-02 -2.44154394e-01 -2.08356336e-01
-4.81942520e-02 -3.25814843e-01 4.00398910e-01 1.11005664e+00
6.60975218e-01 3.88557732e-01 5.50927639e-01 -8.74151051e-01
5.34522235e-01 1.85348690e-01 -1.93791643e-01 -2.70919234e-01
2.74049729e-01 -5.16311228e-02 3.62479031e-01 -1.14150977e+00
3.38798612e-01 1.07321358e+00 8.23082983e-01 2.16362804e-01
-1.00411236e+00 -5.75236738e-01 -1.37166366e-01 1.13899440e-01
-1.31364655e+00 -3.48829180e-01 4.96886820e-01 -5.39283216e-01
5.15656233e-01 -1.01760991e-01 7.84549594e-01 4.28828239e-01
5.54476380e-01 -6.43870458e-02 9.76267159e-01 -5.12730539e-01
6.96354270e-01 6.05647778e-03 -3.64255816e-01 4.72679526e-01
7.93265045e-01 8.00892785e-02 -8.26532423e-01 -2.28862911e-02
6.30607426e-01 -1.43038660e-01 -5.51619887e-01 -8.14364970e-01
-7.54481196e-01 7.86655009e-01 3.65808874e-01 7.77368322e-02
-8.10006559e-01 9.50651020e-02 1.20323770e-01 8.33646059e-02
5.66080511e-02 4.67833191e-01 -3.81287873e-01 -4.17125404e-01
-6.12996697e-01 3.61818671e-01 1.10649526e+00 1.05087984e+00
1.05779672e+00 -6.49654269e-02 6.28679454e-01 5.79299867e-01
9.24095511e-01 8.56497228e-01 8.23038667e-02 -9.49171007e-01
3.88806880e-01 4.41300809e-01 4.71783996e-01 -1.06239629e+00
-7.47645259e-01 -1.92048535e-01 -4.41294163e-01 9.17400360e-01
1.58745617e-01 -2.63572365e-01 -8.70695591e-01 1.08138192e+00
3.59091640e-01 -2.08651766e-01 4.53712523e-01 8.56538773e-01
4.42465425e-01 4.49145675e-01 -1.66339606e-01 6.99067637e-02
1.30278468e+00 -1.09938301e-01 -7.03492999e-01 -4.52828526e-01
4.51799273e-01 -6.85055673e-01 3.07584494e-01 5.26129544e-01
-7.62184620e-01 -3.04436207e-01 -1.33902371e+00 3.26044500e-01
-4.93206173e-01 1.15485959e-01 2.95825869e-01 5.00330329e-01
-1.01115429e+00 7.49004900e-01 -1.28427029e+00 -5.74807107e-01
1.79454461e-02 5.70983887e-01 -5.40815175e-01 -1.01594806e-01
-1.01763701e+00 1.43062449e+00 2.47865438e-01 6.67631388e-01
-1.08949053e+00 -3.59387785e-01 -1.18122005e+00 -1.39566943e-01
7.53442496e-02 -3.04232448e-01 1.07986248e+00 -1.67711601e-01
-1.68443584e+00 8.47830549e-02 -1.71454936e-01 -4.65559065e-01
4.99079138e-01 -5.89070141e-01 1.90214276e-01 4.20634717e-01
1.57133117e-01 4.02427047e-01 5.92576146e-01 -1.79089427e+00
-9.14495647e-01 -4.25186813e-01 1.84598379e-02 4.73306924e-01
5.27855046e-02 -3.30553889e-01 -2.47964367e-01 -7.84377288e-03
8.64552796e-01 -1.02859688e+00 -6.35831058e-01 3.70218128e-01
1.99060112e-01 2.75613338e-01 4.56583977e-01 -6.29181445e-01
5.49061716e-01 -1.89156330e+00 3.52059990e-01 6.94122076e-01
-3.62774551e-01 -9.71277356e-02 3.74528915e-01 8.96370053e-01
5.41412592e-01 -2.17461631e-01 -2.90802568e-01 -6.82130456e-01
-7.03311861e-02 4.52026039e-01 1.01984918e-01 8.29575896e-01
3.41608971e-02 -5.55878803e-02 -9.80075359e-01 -1.99769840e-01
5.63409805e-01 4.95037764e-01 -4.28276807e-01 1.84950113e-01
1.67175919e-01 6.75868750e-01 -4.23429281e-01 7.26066530e-01
9.43330288e-01 5.69266379e-01 1.47751078e-01 -2.01858595e-01
-1.02102709e+00 -1.41675305e-02 -1.47714448e+00 1.92496717e+00
-7.19153762e-01 3.78055751e-01 7.93955743e-01 -6.68616414e-01
1.27284062e+00 1.19884737e-01 3.32697690e-01 -2.13263988e-01
2.71630198e-01 7.53086925e-01 -1.37012556e-01 -6.88764870e-01
1.01662552e+00 -1.35529026e-01 2.26431377e-02 -3.64703715e-01
-7.81864375e-02 -7.17501938e-01 -8.16313026e-04 -9.21119973e-02
8.54972422e-01 5.75372756e-01 1.19983099e-01 -7.21802413e-01
5.73964000e-01 3.91672522e-01 5.92482090e-01 6.11413360e-01
1.83097705e-01 5.32296598e-01 1.42144680e-01 -1.24689348e-01
-6.49809897e-01 -7.74492502e-01 -3.03316116e-01 5.72136879e-01
8.55845630e-01 1.30310372e-01 -4.69476789e-01 -5.06085902e-02
1.55528650e-01 4.04920816e-01 -4.60733473e-01 8.48586038e-02
-5.86346328e-01 -5.04165292e-01 1.62812278e-01 3.43916893e-01
2.26659074e-01 -7.05291927e-01 -1.31369889e+00 5.40786207e-01
1.14929661e-01 -8.38398457e-01 5.00231743e-01 7.14144766e-01
-9.32713568e-01 -1.03187227e+00 -4.37927961e-01 -5.82134783e-01
8.86456370e-01 3.59775454e-01 5.89651346e-01 1.89793631e-01
2.80548725e-02 2.65081435e-01 -9.27069604e-01 -7.20253229e-01
-3.35120112e-01 -2.20187455e-01 3.41352254e-01 -2.31714100e-01
5.86434547e-03 -5.44459999e-01 -5.75979590e-01 5.28930783e-01
-6.64446473e-01 -4.16452825e-01 9.67294574e-01 7.06556499e-01
5.34184992e-01 6.27458394e-02 2.43942201e-01 -3.17764223e-01
1.40727401e-01 -5.00623763e-01 -1.15620029e+00 -1.49421632e-01
-2.23656759e-01 -6.13277294e-02 2.10577443e-01 6.35479856e-03
-7.85265982e-01 5.62495649e-01 -4.39896047e-01 -1.23483941e-01
-5.66847064e-02 9.51434433e-01 -1.31487548e-01 -8.24925661e-01
5.93066275e-01 6.39411137e-02 4.29945469e-01 -5.10810971e-01
-7.47041851e-02 7.27670252e-01 5.32071114e-01 -4.44540620e-01
1.06972957e+00 7.27338791e-01 4.01135892e-01 -1.20980155e+00
-9.39300507e-02 -6.48733675e-01 -1.00691450e+00 -5.12098491e-01
7.28581369e-01 -9.52078640e-01 -6.47668362e-01 5.13136864e-01
-1.18042529e+00 -2.37288818e-01 1.55206263e-01 8.07505786e-01
-3.09836566e-01 3.18451166e-01 2.08093803e-02 -1.15184700e+00
-1.70558691e-01 -1.37952232e+00 1.10340822e+00 5.17757833e-01
1.36554837e-01 -7.48197913e-01 1.37826473e-01 -2.07930997e-01
5.52826464e-01 3.68479311e-01 4.58124280e-02 -3.09393913e-01
-7.98597574e-01 -3.01647604e-01 2.06480578e-01 1.03613861e-01
7.72562474e-02 -1.53532609e-01 -6.16148114e-01 -4.72581178e-01
-9.16252509e-02 -1.30356953e-01 3.98288816e-01 2.03680381e-01
2.08479896e-01 2.25459844e-01 -4.90996093e-01 5.33854723e-01
1.62206411e+00 3.22251439e-01 6.18843019e-01 7.77182162e-01
3.29424173e-01 9.52848136e-01 1.43457878e+00 6.53265059e-01
5.57936668e-01 6.88993573e-01 1.19327593e+00 2.47645840e-01
5.63043177e-01 9.15562082e-03 2.06230879e-01 5.83117545e-01
-3.93556058e-01 6.09185407e-03 -1.04093909e+00 6.46682680e-01
-1.80682421e+00 -5.40654600e-01 -2.56778598e-01 2.44581223e+00
3.38071644e-01 1.34932876e-01 -6.58572495e-01 3.12510699e-01
4.81153339e-01 -3.79831940e-01 -3.43398154e-02 -1.08265646e-01
3.83417070e-01 1.33528203e-01 1.08509099e+00 7.56338239e-01
-9.59221542e-01 6.37256861e-01 4.54878283e+00 1.01879805e-01
-1.09137738e+00 -1.25733867e-01 -5.85315466e-01 5.78306377e-01
-1.14287682e-01 1.17123157e-01 -9.89203811e-01 2.99556434e-01
6.83102608e-01 2.87409008e-01 1.06889762e-01 1.01054823e+00
3.83457810e-01 -6.48954809e-01 -6.92381144e-01 5.91426611e-01
4.49802205e-02 -1.05028367e+00 -4.58073169e-01 1.86792925e-01
3.42514932e-01 1.70419455e-01 -6.68047011e-01 -1.65410955e-02
6.96272999e-02 -4.60285544e-01 1.28358567e+00 6.30195856e-01
3.45537126e-01 -7.85591722e-01 1.46007347e+00 4.69553769e-01
-1.41132522e+00 -9.94857550e-02 -5.66475093e-01 -1.73911989e-01
7.36180663e-01 3.86592329e-01 -8.81098807e-01 9.47791576e-01
1.10374033e+00 2.04554304e-01 -4.12610769e-02 1.65155804e+00
-3.96686018e-01 -6.70620725e-02 -7.44353354e-01 -5.01848936e-01
2.09518984e-01 -4.26686734e-01 6.55286312e-01 8.73038888e-01
9.44267571e-01 1.66998804e-01 3.30929130e-01 3.97576302e-01
4.92040247e-01 4.53056134e-02 -6.34471059e-01 6.06335580e-01
9.46204424e-01 1.29310298e+00 -7.64295101e-01 1.09379537e-01
-2.12638870e-01 4.63218451e-01 1.21438392e-02 -4.35112417e-02
-5.51234007e-01 -8.28157425e-01 6.57957256e-01 1.81889266e-01
3.03114235e-01 -8.19252551e-01 -1.74749196e-01 -5.35686016e-01
-6.88583404e-02 -2.45495871e-01 -2.61052549e-01 -5.60975373e-01
-6.79874063e-01 6.22023940e-01 2.79686004e-01 -1.83685946e+00
-2.47499779e-01 -7.78905809e-01 -3.62825274e-01 1.12364054e+00
-1.81704724e+00 -8.93458247e-01 -1.04816651e+00 1.37844821e-02
3.79876703e-01 2.13313371e-01 8.21134031e-01 2.86005050e-01
8.29390064e-02 -1.42705187e-01 1.69011280e-01 -1.67108640e-01
3.59542340e-01 -1.30831873e+00 1.03488835e-02 1.06953788e+00
-5.36270618e-01 6.52970433e-01 1.18418145e+00 -1.02109706e+00
-1.87687051e+00 -8.52876663e-01 4.20519263e-01 -8.87459703e-03
6.95344448e-01 -2.75865912e-01 -8.53572786e-01 3.70362431e-01
-2.70942599e-01 5.57339657e-03 2.89727718e-01 -3.12881202e-01
6.84717357e-01 -1.26878172e-01 -1.21191955e+00 3.55076909e-01
5.60151935e-01 -2.22933702e-02 -6.45418167e-01 -4.17933101e-03
1.46422893e-01 -8.97918940e-01 -9.29719687e-01 7.77091384e-01
7.98156977e-01 -7.70493388e-01 7.26994395e-01 4.34494376e-01
-1.30973980e-01 -7.88861990e-01 -2.55194187e-01 -1.41889739e+00
1.52447864e-01 -5.03368020e-01 4.99773413e-01 1.09712636e+00
3.54284942e-01 -7.50393867e-01 7.75998890e-01 2.28822947e-01
-7.14955628e-01 -4.88380581e-01 -1.35082591e+00 -5.91776907e-01
-3.22403073e-01 -3.28754455e-01 1.32752672e-01 2.45142817e-01
-1.58022895e-01 -3.43650341e-01 -1.67970926e-01 1.24200547e+00
8.49430919e-01 -2.37525374e-01 1.05386829e+00 -1.53772497e+00
2.36092642e-01 1.45574898e-01 -7.72596598e-01 -8.49864423e-01
-1.68574631e-01 -9.62804556e-02 1.23216057e+00 -1.89445269e+00
-7.32315540e-01 -8.21307600e-01 2.42010504e-01 1.98040530e-01
2.32919678e-01 2.30188042e-01 -3.45468731e-03 1.43622056e-01
-1.02794290e-01 8.03688169e-01 6.34581745e-01 3.02343637e-01
-4.54047322e-01 1.89570397e-01 -3.33574787e-02 9.48259652e-01
5.26100695e-01 -5.49262285e-01 -8.29252079e-02 -5.47244132e-01
3.02914709e-01 2.11771652e-01 2.06474498e-01 -1.54656923e+00
7.35696197e-01 6.89402893e-02 6.29525781e-02 -8.45262110e-01
7.61602342e-01 -1.37079489e+00 1.81413293e-01 8.02788079e-01
4.26672101e-01 1.28912911e-01 2.46677980e-01 7.46844947e-01
-3.29135567e-01 -9.61878896e-01 7.89932013e-01 -2.06585974e-01
-8.89274001e-01 -2.30538279e-01 -6.05015993e-01 -6.71930134e-01
1.08904183e+00 -2.50133634e-01 -2.37123236e-01 -2.90903866e-01
-6.41817749e-01 4.85074848e-01 6.82681143e-01 1.17643222e-01
9.70925391e-01 -7.08533823e-01 -6.10396624e-01 2.22099110e-01
1.12704553e-01 5.45573056e-01 1.93858743e-01 8.87196958e-01
-1.37443149e+00 -1.62062213e-01 -2.83095807e-01 -8.67371082e-01
-1.13054705e+00 -2.12058380e-01 3.95937771e-01 2.58889347e-01
-5.65088451e-01 7.92330801e-01 -4.60983723e-01 -4.43376899e-01
1.67891979e-01 -4.57786113e-01 -5.36998272e-01 1.89216420e-01
5.16223431e-01 5.08186519e-01 3.18645239e-01 -7.60511696e-01
-5.73432863e-01 1.06286228e+00 5.18932402e-01 -4.15851116e-01
1.38818979e+00 -3.88317376e-01 -1.37863681e-01 1.94785073e-01
4.59325343e-01 2.92712688e-01 -1.45286953e+00 3.24974447e-01
1.71958596e-01 -5.97153664e-01 1.03520148e-01 -3.71897757e-01
-5.56813896e-01 7.27101684e-01 7.14044392e-01 7.99861997e-02
7.56905437e-01 -2.57575780e-01 1.15611389e-01 7.26537943e-01
1.26426733e+00 -9.83004153e-01 -3.21610749e-01 7.25268900e-01
1.09436488e+00 -1.13898623e+00 2.92615861e-01 -4.99116302e-01
-4.05029684e-01 1.32083714e+00 3.69145334e-01 -2.52408415e-01
7.10281551e-01 5.97391248e-01 3.83949578e-01 -1.30974934e-01
-1.05864860e-01 -3.34823221e-01 -2.36155316e-01 5.39868057e-01
-1.78193077e-01 -3.75625879e-01 -4.09627914e-01 5.16063988e-01
-1.63931027e-01 -2.02030614e-01 6.93273008e-01 1.72393501e+00
-1.12152445e+00 -8.78045976e-01 -8.15340340e-01 9.07339677e-02
-2.37949356e-01 1.45559371e-01 3.86115164e-01 9.05949950e-01
2.35932693e-01 9.35039580e-01 -6.22371286e-02 -4.49360520e-01
6.48329198e-01 -3.90904844e-01 1.55966014e-01 -6.53246224e-01
-3.97059470e-01 -6.05172589e-02 2.75289685e-01 -5.60012877e-01
-5.98764539e-01 -9.35235322e-01 -1.58357167e+00 4.07504261e-01
-5.55444479e-01 7.36012757e-01 1.70360696e+00 8.76104236e-01
1.02884836e-01 4.96811301e-01 4.51371729e-01 -1.91300225e+00
-5.71976602e-01 -1.19058347e+00 -5.68717003e-01 -2.31091976e-01
2.46239930e-01 -1.27424908e+00 -6.41440094e-01 -2.68733174e-01] | [7.4569091796875, -1.8099533319473267] |
1d5dd07f-5217-43c1-8e7b-225849a3f569 | introducing-two-vietnamese-datasets-for | 1804.05388 | null | http://arxiv.org/abs/1804.05388v2 | http://arxiv.org/pdf/1804.05388v2.pdf | Introducing two Vietnamese Datasets for Evaluating Semantic Models of (Dis-)Similarity and Relatedness | We present two novel datasets for the low-resource language Vietnamese to
assess models of semantic similarity: ViCon comprises pairs of synonyms and
antonyms across word classes, thus offering data to distinguish between
similarity and dissimilarity. ViSim-400 provides degrees of similarity across
five semantic relations, as rated by human judges. The two datasets are
verified through standard co-occurrence and neural network models, showing
results comparable to the respective English datasets. | ['Sabine Schulte im Walde', 'Ngoc Thang Vu', 'Kim Anh Nguyen'] | 2018-04-15 | introducing-two-vietnamese-datasets-for-1 | https://aclanthology.org/N18-2032 | https://aclanthology.org/N18-2032.pdf | naacl-2018-6 | ['vietnamese-datasets'] | ['natural-language-processing'] | [ 3.84186320e-02 -9.21030492e-02 -7.13240921e-01 -5.24561822e-01
-1.38049632e-01 -7.20012486e-01 9.90123987e-01 8.10487390e-01
-9.76542830e-01 5.99221230e-01 7.50746489e-01 -1.14476256e-01
-5.57990789e-01 -6.06754243e-01 3.00743401e-01 8.95113777e-03
1.25145987e-01 8.05358291e-01 -2.06024632e-01 -9.23418343e-01
6.25278354e-01 1.94652006e-01 -1.43376839e+00 4.86715823e-01
1.05657327e+00 7.30406165e-01 1.85567245e-01 -1.05058394e-01
-2.35631585e-01 5.57895362e-01 -7.32763171e-01 -8.81076872e-01
1.56809509e-01 -1.71887383e-01 -1.21005726e+00 -4.61936444e-01
5.57556987e-01 1.93481699e-01 -1.97639376e-01 9.08775508e-01
3.97474408e-01 4.60575908e-01 8.97110581e-01 -1.12868011e+00
-1.45418024e+00 8.77156734e-01 1.06827416e-01 5.94704032e-01
1.02507877e+00 -3.75027210e-01 1.68607259e+00 -1.11704898e+00
1.18761313e+00 1.36459076e+00 8.51003885e-01 2.62206763e-01
-1.31223249e+00 -7.32363880e-01 -4.22769368e-01 8.25723767e-01
-1.49489510e+00 -3.16728622e-01 3.00095707e-01 -3.66090953e-01
1.79662371e+00 1.03993945e-01 6.85197532e-01 1.37958336e+00
-1.22286744e-01 6.89285845e-02 1.47539520e+00 -3.09531003e-01
1.22675924e-02 3.98437202e-01 1.74096510e-01 9.09766462e-03
2.50536352e-01 1.85831606e-01 -7.28517175e-01 -1.70760155e-01
6.10948324e-01 -3.00313920e-01 -5.75144961e-02 1.91358864e-01
-1.43451846e+00 1.01974738e+00 6.89251721e-01 9.27556515e-01
-3.64545614e-01 -5.36167681e-01 6.95726991e-01 7.35765576e-01
3.74196798e-01 1.21257448e+00 -3.50783199e-01 -2.64316469e-01
-4.96960789e-01 3.15961093e-01 8.25008333e-01 1.05597436e+00
5.05349815e-01 1.18877694e-01 -5.44292293e-02 1.52531803e+00
7.38323387e-03 3.90925258e-01 1.14005816e+00 -8.77462924e-01
4.77691561e-01 5.92277467e-01 -1.11405015e-01 -1.57496405e+00
-4.20416713e-01 -7.69893229e-02 -4.98818874e-01 -3.93301338e-01
4.03841324e-02 4.57994252e-01 -2.95680761e-01 1.70608330e+00
-8.17302465e-02 -1.03908584e-01 3.06714803e-01 7.74622500e-01
1.42265117e+00 3.21670890e-01 4.82206315e-01 -2.00717926e-01
1.41479158e+00 -6.45241559e-01 -7.87864506e-01 -5.70382118e-01
9.27346349e-01 -7.90208876e-01 1.43958879e+00 1.74609032e-02
-1.13821423e+00 -8.49405169e-01 -1.04325581e+00 -3.61305565e-01
-1.13734961e+00 -3.87566805e-01 6.07069075e-01 4.88479942e-01
-1.00733697e+00 8.26061070e-01 1.76052585e-01 -7.15679228e-01
1.34423440e-02 6.20774440e-02 -8.08292150e-01 1.50665745e-01
-2.00415015e+00 1.89677358e+00 1.06289315e+00 -2.95679510e-01
-3.14640164e-01 -7.47740328e-01 -1.16669631e+00 -1.36975512e-01
-1.31505728e-01 -3.98760647e-01 7.38938630e-01 -9.82040405e-01
-9.44127381e-01 1.70848453e+00 -7.68331513e-02 -2.87509710e-01
-2.03168560e-02 4.81293872e-02 -1.28848398e+00 1.04937300e-01
7.55608261e-01 6.72109127e-01 2.13372335e-01 -9.71447170e-01
-4.68429536e-01 -7.05030411e-02 3.20570081e-01 6.00130022e-01
-6.81075037e-01 5.59706271e-01 3.06793779e-01 -9.58472192e-01
-3.48904729e-02 -6.15592420e-01 1.64322108e-01 -5.34230649e-01
-4.11484651e-02 -6.00291550e-01 1.93995610e-01 -8.72188032e-01
1.37400413e+00 -1.89773560e+00 1.03466213e-01 3.44054490e-01
1.97755948e-01 3.45231861e-01 -5.07326007e-01 7.50190198e-01
-6.24148965e-01 4.17465180e-01 -5.41016571e-02 1.02197871e-01
1.33006349e-01 4.73896712e-01 1.64142419e-02 2.49271929e-01
2.79818565e-01 1.12552488e+00 -1.24115169e+00 -5.58503807e-01
1.03200234e-01 1.35316133e-01 -3.44177783e-01 -1.31629184e-01
5.95735312e-01 -1.59513075e-02 2.12151423e-01 5.31386971e-01
1.76083386e-01 1.33764014e-01 5.08394301e-01 -2.42134184e-01
2.53423333e-01 1.00496018e+00 -7.04848647e-01 1.74667084e+00
-7.50443697e-01 6.00436330e-01 -5.81105292e-01 -1.12752318e+00
1.20785713e+00 3.44251782e-01 9.08866990e-03 -1.26853979e+00
1.83511227e-01 4.20432925e-01 3.99863183e-01 -5.07048905e-01
1.00159085e+00 -5.26806235e-01 -2.86007643e-01 2.03925282e-01
1.80026367e-01 -4.70450938e-01 5.04102588e-01 2.60198563e-01
6.70918345e-01 -1.05914116e-01 9.00969267e-01 -7.64744222e-01
4.87927467e-01 1.65183738e-01 5.23564160e-01 2.85332650e-01
-1.78036585e-01 1.61170829e-02 1.53432429e-01 -2.11324528e-01
-1.23230934e+00 -1.53571892e+00 -5.98919988e-01 9.99946594e-01
2.48863205e-01 -8.07936549e-01 -3.15944880e-01 -4.33810920e-01
8.96370709e-02 1.23572600e+00 -6.81568384e-01 -4.23256934e-01
-4.14679170e-01 -7.43908957e-02 7.92267919e-01 7.27574348e-01
1.89076527e-03 -1.32287490e+00 -3.58917564e-01 7.70505937e-03
-4.12611634e-01 -1.24033141e+00 -2.28903502e-01 1.11626096e-01
-5.63897848e-01 -9.73737419e-01 -1.93604156e-01 -1.08689713e+00
5.69920763e-02 2.81583488e-01 1.86889529e+00 2.36583024e-01
-3.06099534e-01 -5.20236902e-02 -6.65593147e-01 4.65292782e-02
-5.26540697e-01 -7.49257673e-03 4.71161962e-01 -9.65123475e-01
1.03289366e+00 -6.04382515e-01 -5.65183572e-02 3.53175730e-01
-7.37909615e-01 -3.81952584e-01 1.48493350e-01 1.01641881e+00
1.27986595e-01 -2.15370387e-01 8.08196843e-01 -8.16485763e-01
9.31334198e-01 -1.02855349e+00 1.06540121e-01 3.28025855e-02
-7.99066901e-01 -4.18777674e-01 6.61229372e-01 -4.43298757e-01
-9.28573847e-01 -9.58937645e-01 -1.13330334e-01 1.17372856e-01
-1.52726009e-01 8.55547667e-01 9.96200442e-02 1.29846826e-01
1.04519224e+00 -2.19489112e-01 -3.01007539e-01 -2.46984035e-01
6.52243018e-01 8.27655733e-01 7.46662915e-01 -7.62471437e-01
5.97796738e-01 9.19943079e-02 -3.37019324e-01 -9.98319328e-01
-7.91767478e-01 -6.87658668e-01 -1.06605208e+00 1.55354753e-01
7.06877112e-01 -7.61834025e-01 -6.49905920e-01 -2.91944206e-01
-1.12998450e+00 1.60880700e-01 -4.46517467e-01 6.40491426e-01
-5.07784605e-01 2.17294753e-01 -7.48457789e-01 -9.43842228e-04
-1.00949243e-01 -7.58951485e-01 6.34172082e-01 -8.69135186e-02
-1.35726416e+00 -1.60240066e+00 9.37516540e-02 4.78633851e-01
3.54700595e-01 1.62962779e-01 1.41595149e+00 -1.33028817e+00
6.85637832e-01 2.70064250e-02 -3.89727771e-01 3.33064765e-01
3.34263146e-01 -3.13512176e-01 -6.30202293e-01 -1.72767028e-01
-5.12843765e-02 -6.39011860e-01 5.56329429e-01 -1.21404044e-01
5.42465448e-01 -2.21807376e-01 -1.74663350e-01 3.20711464e-01
1.41884601e+00 2.96186030e-01 4.98418093e-01 6.12650573e-01
6.96989775e-01 1.02087486e+00 7.46295989e-01 1.66698277e-01
5.58826685e-01 7.51494825e-01 -1.92511797e-01 2.76380908e-02
-9.57027748e-02 -2.95281589e-01 4.53667380e-02 1.26995456e+00
-6.66107908e-02 3.09416447e-02 -1.12630582e+00 8.57635140e-01
-1.41047227e+00 -9.96490419e-01 -4.35989857e-01 1.94876969e+00
9.16866302e-01 -8.57885182e-02 8.92601684e-02 2.62948364e-01
6.68794692e-01 4.06723529e-01 -9.01049525e-02 -1.03348815e+00
-6.21544898e-01 7.39186287e-01 5.43203577e-02 5.19640684e-01
-6.36417627e-01 1.53326058e+00 8.00716305e+00 9.46964443e-01
-5.68575621e-01 2.14982972e-01 7.52684996e-02 2.90850215e-02
-5.87621450e-01 -1.16779236e-02 -2.55859047e-01 3.12317818e-01
1.18585086e+00 -5.54412603e-01 3.07242125e-01 6.04003847e-01
-6.30924627e-02 8.72043222e-02 -1.27362752e+00 8.62950802e-01
4.33626652e-01 -1.08419609e+00 4.26164597e-01 -5.30440092e-01
7.52680302e-01 -2.91042011e-02 -2.30383612e-02 1.73766971e-01
5.56735516e-01 -1.32892060e+00 5.58118939e-01 5.00886478e-02
1.08751917e+00 -8.79823625e-01 8.21495295e-01 -2.57687896e-01
-1.03935289e+00 8.67495611e-02 -6.32009089e-01 -4.79162693e-01
2.34697133e-01 7.26958215e-02 -6.43454969e-01 5.75679958e-01
7.66789436e-01 1.16224647e+00 -7.57716060e-01 3.22661817e-01
-5.28159499e-01 7.28164688e-02 1.90307915e-01 -3.90460044e-02
3.68093014e-01 -2.75140166e-01 4.27134037e-01 1.45827079e+00
9.62031484e-02 -1.05784386e-01 5.78460060e-02 9.60470736e-01
1.44201964e-01 5.92360616e-01 -7.98799157e-01 -3.28542888e-01
1.05621576e+00 1.04489410e+00 -6.37557924e-01 -5.81217051e-01
-4.58751351e-01 1.22308624e+00 6.70230806e-01 1.66186169e-01
-6.01918101e-01 -6.77152872e-01 9.55604196e-01 -9.76335183e-02
-3.81402582e-01 -2.57786423e-01 -5.10852218e-01 -9.43603158e-01
-2.65209377e-01 -8.57861638e-01 5.58280945e-01 -9.45589185e-01
-1.99868846e+00 6.89639688e-01 3.24758112e-01 -9.21552420e-01
-2.94727057e-01 -9.10711527e-01 -3.21552634e-01 1.14469838e+00
-1.13067639e+00 -9.50351119e-01 -2.94499218e-01 5.70113063e-01
3.43921244e-01 -3.37773055e-01 1.20315242e+00 4.46295172e-01
-3.88523340e-02 5.96430063e-01 -2.38708854e-01 1.14171132e-01
1.00347257e+00 -1.15422738e+00 8.64697754e-01 2.44900167e-01
3.10416341e-01 9.66773629e-01 5.76398790e-01 -8.56849730e-01
-4.41492885e-01 -5.01205087e-01 1.72571111e+00 -5.69128275e-01
1.18377936e+00 3.44651081e-02 -1.00193298e+00 4.70283836e-01
4.68587577e-01 -5.57821751e-01 1.15942073e+00 4.46398377e-01
-8.26639473e-01 3.83872896e-01 -1.15952373e+00 6.01622522e-01
1.74387038e+00 -1.22177887e+00 -1.48129463e+00 4.73676205e-01
6.50907159e-01 3.35604772e-02 -1.33957231e+00 3.12106013e-01
5.21379590e-01 -8.34185898e-01 1.29949951e+00 -1.32024395e+00
8.12379122e-01 7.61253089e-02 -4.60863411e-01 -1.98862982e+00
-7.96482027e-01 -5.84337302e-02 6.39958024e-01 1.24311507e+00
6.06819749e-01 -6.71726942e-01 1.83161870e-01 3.02314252e-01
-2.15551794e-01 -1.81165382e-01 -1.03415930e+00 -1.19627249e+00
4.22627866e-01 -4.33223516e-01 6.25685990e-01 1.93022561e+00
7.30463564e-01 7.68481612e-01 1.80497333e-01 -5.12297273e-01
-9.64935943e-02 -2.78362840e-01 -1.28362924e-01 -1.42026651e+00
1.08510666e-01 -9.12351251e-01 -7.51733541e-01 -2.23727927e-01
9.84632194e-01 -1.57027793e+00 -4.48189259e-01 -1.56464434e+00
1.61531150e-01 -5.21891475e-01 -2.50002027e-01 2.71669209e-01
-2.67017096e-01 5.75208783e-01 5.51717952e-02 1.75527528e-01
-8.54648948e-02 4.18422282e-01 6.66765988e-01 1.26559306e-02
1.42178118e-01 -8.25239539e-01 -8.40715349e-01 8.82301331e-01
7.75996745e-01 -3.32453758e-01 -5.08837640e-01 -6.50424898e-01
6.02917254e-01 -4.28400844e-01 2.24239156e-01 -6.11887038e-01
-1.21327333e-01 -3.76368582e-01 2.72917271e-01 8.58861730e-02
4.13268954e-01 -7.94486165e-01 1.42794505e-01 4.64928895e-01
-7.25243211e-01 9.70160425e-01 1.11516304e-01 -9.86596569e-03
-5.18889725e-01 -4.57740188e-01 6.77025437e-01 -2.26418912e-01
-1.18641269e+00 -2.56662339e-01 -3.04556817e-01 6.51302040e-01
9.50548649e-01 -4.55260009e-01 -5.84035993e-01 -2.80962408e-01
-6.87579155e-01 -5.38348407e-02 5.38672209e-01 1.16607916e+00
7.26599991e-01 -2.00763679e+00 -1.00179577e+00 -7.98665509e-02
7.51595914e-01 -9.98745024e-01 -1.13992998e-02 4.83682334e-01
-3.88549715e-01 3.72994095e-01 -6.81232035e-01 -6.78033277e-04
-1.09543800e+00 8.13564539e-01 -5.56440242e-02 1.38327733e-01
-1.86349884e-01 7.79751658e-01 -2.61831582e-01 -6.40390933e-01
-4.09636080e-01 3.59238051e-02 -6.84508085e-01 6.35494888e-01
3.20794135e-01 6.95020914e-01 -1.59792811e-01 -1.43148279e+00
-4.24325407e-01 4.65964437e-01 1.29638672e-01 -1.95733890e-01
8.98257792e-01 -1.63459063e-01 -5.03217995e-01 8.36231172e-01
1.40079665e+00 -2.40742877e-01 3.39616001e-01 -4.21727151e-01
5.30912101e-01 -8.01134408e-01 -4.20340687e-01 -8.77471507e-01
-3.48628521e-01 6.66960120e-01 2.13002861e-01 2.13654935e-01
6.49684489e-01 -7.62015954e-02 7.26515174e-01 4.40960616e-01
1.60642862e-01 -1.56524086e+00 7.99151212e-02 9.64852214e-01
1.03485823e+00 -1.19492638e+00 -1.49027392e-01 -5.15076935e-01
-1.03032517e+00 8.13919127e-01 6.37410462e-01 -2.34827250e-01
4.13275778e-01 -8.92015547e-02 1.88721418e-01 -3.29249799e-01
-5.10535121e-01 -2.52076298e-01 7.17268229e-01 1.07192874e+00
9.59962308e-01 2.27305159e-01 -1.13771069e+00 3.98757309e-01
-9.56938446e-01 -7.16340959e-01 2.77392060e-01 3.40918630e-01
-1.75067801e-02 -1.11025512e+00 7.87482336e-02 5.66413105e-01
-1.95350707e-01 -7.32009888e-01 -9.40014482e-01 9.58041191e-01
2.10179895e-01 1.16813791e+00 3.97923917e-01 -4.51362401e-01
4.46280122e-01 -9.98284146e-02 4.02951360e-01 -7.47810721e-01
-1.10979950e+00 -7.45618761e-01 8.92299235e-01 -6.52908266e-01
-5.82309067e-01 -4.44331825e-01 -1.09052110e+00 -4.17781919e-01
-1.46026254e-01 2.82274306e-01 4.47155595e-01 1.11903989e+00
1.28768161e-01 2.38786787e-01 3.17051351e-01 -5.36833882e-01
-2.82494515e-01 -1.28644001e+00 -7.44848967e-01 1.31216490e+00
-5.22782981e-01 -8.99626493e-01 -4.25571173e-01 -2.27774888e-01] | [10.583477973937988, 9.176840782165527] |
3a6836a5-9162-418b-b5c2-4734952a0ec7 | trwp-text-relatedness-using-word-and-phrase | null | null | https://aclanthology.org/S15-2016 | https://aclanthology.org/S15-2016.pdf | TrWP: Text Relatedness using Word and Phrase Relatedness | null | ['Evangelos Milios', 'Aminul Islam', 'Md Rashadul Hasan Rakib'] | 2015-06-01 | null | null | null | semeval-2015-6 | ['phrase-relatedness'] | ['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] | [-7.068718910217285, 3.949030637741089] |
faa132e4-03cb-4bd7-ad9c-4b29f65bbc95 | team-ainlpml-mup-in-sdp-2021-scientific | null | null | https://aclanthology.org/2022.sdp-1.36 | https://aclanthology.org/2022.sdp-1.36.pdf | Team AINLPML @ MuP in SDP 2021: Scientific Document Summarization by End-to-End Extractive and Abstractive Approach | This paper introduces the proposed summarization system of the AINLPML team for the First Shared Task on Multi-Perspective Scientific Document Summarization at SDP 2022. We present a method to produce abstractive summaries of scientific documents. First, we perform an extractive summarization step to identify the essential part of the paper. The extraction step includes utilizing a contributing sentence identification model to determine the contributing sentences in selected sections and portions of the text. In the next step, the extracted relevant information is used to condition the transformer language model to generate an abstractive summary. In particular, we fine-tuned the pre-trained BART model on the extracted summary from the previous step. Our proposed model successfully outperformed the baseline provided by the organizers by a significant margin. Our approach achieves the best average Rouge F1 Score, Rouge-2 F1 Score, and Rouge-L F1 Score among all submissions. | ['Asif Ekbal', 'Kartik Shinde', 'Guneet Singh Kohli', 'Sandeep Kumar'] | null | null | null | null | sdp-coling-2022-10 | ['scientific-article-summarization', 'extractive-summarization', 'document-summarization'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 3.12512636e-01 4.01096433e-01 -1.32649273e-01 -2.87311655e-02
-1.61483443e+00 -7.10250556e-01 5.82337260e-01 6.03660285e-01
-3.36985946e-01 1.08760452e+00 8.95755827e-01 -6.29539937e-02
1.14737429e-01 -3.22656423e-01 -7.16493249e-01 -3.66033345e-01
3.07714999e-01 4.48587328e-01 4.03360166e-02 -3.27782519e-02
1.00948071e+00 3.13185871e-01 -9.75682497e-01 5.30591249e-01
1.37565446e+00 4.74414021e-01 5.32019973e-01 1.18275678e+00
-2.93880343e-01 5.46395957e-01 -1.14842820e+00 -2.07546845e-01
-2.95047343e-01 -7.81650305e-01 -9.11765218e-01 -9.76396352e-02
5.17476737e-01 -7.48205855e-02 2.50885473e-03 8.23763430e-01
6.29938245e-01 2.85652518e-01 8.89154017e-01 -5.68592131e-01
-1.57404736e-01 1.12753880e+00 -8.41099620e-01 4.53189045e-01
5.76490462e-01 -1.76848471e-01 1.27706289e+00 -9.18415368e-01
8.98011327e-01 1.07870030e+00 7.54261538e-02 4.91008341e-01
-8.79367828e-01 -2.95982569e-01 -8.63398388e-02 -6.65267557e-02
-7.91375339e-01 -5.09476662e-01 7.44494975e-01 -2.07448512e-01
9.91800368e-01 4.99262780e-01 4.77472752e-01 7.90827811e-01
6.01601303e-01 1.06574118e+00 4.32906061e-01 -5.39833784e-01
2.58829862e-01 -2.15974320e-02 5.69534242e-01 3.64517093e-01
3.44580710e-01 -1.01230323e+00 -7.59945631e-01 -2.58406043e-01
-6.86362982e-02 -5.71756661e-01 -1.75152436e-01 4.85605448e-01
-1.33791327e+00 4.94846821e-01 -1.25717446e-01 4.52241004e-01
-7.12177992e-01 -1.99484840e-01 7.28300035e-01 3.94610763e-02
7.39926219e-01 9.45107639e-01 -1.92807540e-01 -1.59440503e-01
-1.73749101e+00 5.38209915e-01 9.05948877e-01 9.37936366e-01
1.46886513e-01 -1.38226032e-01 -1.07576108e+00 8.84640932e-01
-4.07067724e-02 4.07329202e-01 3.90218854e-01 -1.02430511e+00
9.76210833e-01 5.30318975e-01 1.87027261e-01 -7.98887551e-01
-1.73824549e-01 -7.16705978e-01 -8.39221776e-01 -7.03892350e-01
-2.66234159e-01 -4.35989499e-01 -5.95340967e-01 1.12781179e+00
-5.47068380e-02 -2.48706296e-01 3.93824100e-01 3.36556971e-01
1.32130325e+00 1.17072654e+00 -5.95287681e-02 -7.30715990e-01
1.40315521e+00 -1.27803385e+00 -8.45048785e-01 -8.93678442e-02
4.21911776e-01 -1.09117520e+00 6.79245472e-01 3.57720643e-01
-1.44237530e+00 -5.00117540e-01 -1.34064579e+00 -2.03232810e-01
1.04445904e-01 8.60736251e-01 -1.03498407e-01 4.82353196e-02
-1.05107808e+00 6.88746512e-01 -6.41516745e-01 -3.67017210e-01
2.80054897e-01 -6.79814117e-03 -2.56791651e-01 4.24759053e-02
-7.22511768e-01 6.78154469e-01 4.18012947e-01 -1.31509170e-01
-7.88957179e-01 -8.85906458e-01 -5.59267342e-01 4.40659672e-01
1.84436232e-01 -8.86373401e-01 1.35671568e+00 -2.74776146e-02
-1.37363350e+00 5.62020481e-01 -6.08841538e-01 -6.92074716e-01
5.07743418e-01 -4.40971226e-01 -7.58636510e-03 4.55198824e-01
3.85015458e-01 4.76106614e-01 3.38489234e-01 -9.69686925e-01
-7.33350694e-01 -2.01280355e-01 -3.26286972e-01 3.63902390e-01
-1.95613027e-01 1.51160568e-01 -5.04837990e-01 -6.10032022e-01
-1.99784547e-01 -5.48803151e-01 -9.79641676e-02 -1.14131868e+00
-1.07917488e+00 -5.45300424e-01 5.35762489e-01 -1.19257343e+00
1.57498837e+00 -1.71256256e+00 3.06914747e-01 -1.33259982e-01
2.02604949e-01 2.60652572e-01 -3.50174695e-01 8.31283748e-01
2.59112298e-01 3.53310227e-01 -1.21400222e-01 -6.65855706e-01
-1.51013851e-01 -5.85482597e-01 -6.78280115e-01 -6.62282854e-02
5.51924594e-02 7.20604897e-01 -1.07534432e+00 -8.03446651e-01
-2.56469369e-01 -9.31382924e-02 -2.76391000e-01 3.99474174e-01
-4.03356791e-01 3.75299692e-01 -6.48913860e-01 1.64148182e-01
5.64529121e-01 -1.71388388e-02 -1.11719873e-02 -9.32872966e-02
-3.81444335e-01 6.43745720e-01 -6.39424860e-01 1.74010229e+00
-3.85695398e-01 8.15854073e-01 -1.99251473e-01 -7.76810110e-01
9.48442757e-01 2.77415812e-01 4.30195004e-01 -1.24366879e-01
-1.32091284e-01 3.21044326e-01 -2.86072165e-01 -3.75218332e-01
1.17375243e+00 3.88741344e-01 -4.16189998e-01 5.89324951e-01
2.65715003e-01 -3.55770558e-01 9.83516216e-01 1.03142452e+00
1.10612214e+00 6.20888025e-02 4.78277534e-01 -3.13490868e-01
9.77380455e-01 1.66960523e-01 4.12397981e-01 8.87832820e-01
1.89540967e-01 8.08465600e-01 8.09390426e-01 5.25332950e-02
-1.02394021e+00 -5.56738675e-01 3.17730665e-01 6.77419960e-01
-3.17259014e-01 -7.64261186e-01 -9.55900550e-01 -8.39940310e-01
-3.83381307e-01 1.36758232e+00 -3.60363990e-01 -1.22248761e-01
-6.98318720e-01 -5.40311992e-01 5.64651370e-01 2.02618375e-01
5.03333867e-01 -1.14030290e+00 -3.41488481e-01 1.35388747e-01
-8.23980033e-01 -9.32632387e-01 -9.70037878e-01 -1.94945157e-01
-7.95356750e-01 -6.94510758e-01 -9.52182055e-01 -6.89529598e-01
5.40653646e-01 9.74659175e-02 8.65050137e-01 -3.60634506e-01
1.62277967e-01 -1.45734146e-01 -3.04275662e-01 -6.23321474e-01
-7.22274303e-01 6.59494102e-01 -2.56971151e-01 -3.35732996e-01
2.74307989e-02 -1.19395375e-01 -3.66423190e-01 -5.09506524e-01
-5.09035408e-01 2.56001562e-01 6.41406715e-01 5.37553728e-01
5.46666563e-01 -2.34959677e-01 9.90597188e-01 -7.96570122e-01
1.25799096e+00 -1.39238194e-01 -1.62450716e-01 5.80745399e-01
-3.53221923e-01 2.06393957e-01 7.73945987e-01 2.33777333e-02
-1.35111248e+00 -3.95987988e-01 -4.91609760e-02 2.92207867e-01
1.93431929e-01 7.30844676e-01 -2.02977628e-01 6.38929546e-01
4.56434280e-01 6.16579771e-01 -2.96254367e-01 -5.49046874e-01
2.65098691e-01 1.03084838e+00 6.43003166e-01 -3.29475999e-01
4.78509456e-01 -1.13699399e-01 -1.55007571e-01 -1.06441498e+00
-1.15838301e+00 -5.65272987e-01 -6.09795272e-01 -1.81378916e-01
7.57969081e-01 -8.25299919e-01 -3.65776390e-01 1.65079802e-01
-1.66611981e+00 1.92774430e-01 -3.32609504e-01 3.59636903e-01
-3.49207908e-01 6.66158438e-01 -7.09649742e-01 -6.60061955e-01
-1.36022484e+00 -8.80840540e-01 1.24663734e+00 6.70286238e-01
-7.40988374e-01 -5.16856253e-01 3.03489357e-01 6.90430105e-01
1.17711045e-01 1.79402113e-01 9.42497015e-01 -1.07817376e+00
-1.30260602e-01 -3.15885574e-01 -9.27389190e-02 3.49063575e-01
2.50598006e-02 2.62750536e-01 -5.07681489e-01 -1.64161608e-01
-1.89725146e-01 -1.42434388e-01 1.32090342e+00 7.29917347e-01
8.86686385e-01 -4.50754762e-01 -3.73993069e-01 4.27969582e-02
9.02054667e-01 7.88459852e-02 4.15031016e-01 2.28237882e-01
5.09577096e-01 5.12060881e-01 7.62955785e-01 5.34439266e-01
3.66728991e-01 2.49877214e-01 -2.52118021e-01 3.71388286e-01
-3.72801930e-01 -3.53729397e-01 5.23045421e-01 1.50493741e+00
1.71780139e-01 -6.80280626e-01 -5.72183430e-01 7.12634206e-01
-1.78613770e+00 -1.07716990e+00 -2.09131137e-01 2.03149295e+00
8.44228625e-01 2.08539501e-01 -4.09248285e-03 -1.70208722e-01
7.45907009e-01 4.45459455e-01 -1.74342141e-01 -8.25654387e-01
-1.74904749e-01 -7.57981092e-02 1.63201317e-01 6.01432621e-01
-7.73695111e-01 9.09000993e-01 5.85984898e+00 1.01551580e+00
-7.50087559e-01 -2.00486124e-01 5.35115123e-01 -3.75144035e-01
-4.68685150e-01 7.03559145e-02 -1.07168293e+00 4.83003497e-01
1.27011073e+00 -9.79938328e-01 -2.22462803e-01 4.03908014e-01
6.70932710e-01 -3.68110985e-01 -9.40447867e-01 3.73058319e-01
4.81080323e-01 -1.60965419e+00 5.12116134e-01 -5.71843423e-02
9.47389305e-01 -1.23618349e-01 -4.53359723e-01 3.48595083e-01
1.76937029e-01 -5.37577093e-01 5.73941529e-01 8.33758593e-01
5.48765421e-01 -9.47764635e-01 8.99319410e-01 5.96150935e-01
-7.03539431e-01 3.02407712e-01 -4.57995892e-01 3.42689484e-01
2.72355378e-01 9.71346915e-01 -1.06596684e+00 1.17345023e+00
5.83191356e-03 8.26847672e-01 -5.27863801e-01 1.11369717e+00
-3.76935095e-01 8.75754178e-01 4.53897230e-02 -3.37021500e-01
1.67825982e-01 -2.11441889e-01 1.07157397e+00 1.56266665e+00
5.31658232e-01 -8.69565308e-02 1.71128921e-02 5.60853124e-01
-7.17450917e-01 4.26715732e-01 -2.19119847e-01 -3.55112940e-01
4.36987728e-01 1.43225098e+00 -6.85491145e-01 -7.16107905e-01
2.54349679e-01 9.95726049e-01 1.41324878e-01 2.05884159e-01
-4.26704317e-01 -9.73489642e-01 -2.89012253e-01 -2.48289660e-01
3.04216832e-01 1.30109608e-01 -5.87205648e-01 -1.26167285e+00
1.76487461e-01 -7.26468503e-01 2.92406559e-01 -7.87699461e-01
-5.85177481e-01 6.86483204e-01 -1.83711722e-01 -9.60991323e-01
-3.83170933e-01 3.84262174e-01 -1.15828931e+00 9.54736412e-01
-1.11455142e+00 -1.03187072e+00 6.41702563e-02 -4.81622219e-01
1.15040457e+00 -4.28798676e-01 6.36788309e-01 -2.19526246e-01
-8.35085690e-01 2.98380047e-01 3.81210417e-01 -1.47439808e-01
9.22096074e-01 -1.32726824e+00 4.48054224e-01 1.24327731e+00
-1.39903814e-01 7.36734271e-01 1.03446531e+00 -1.14063811e+00
-1.07330418e+00 -1.09295607e+00 1.73577785e+00 -1.66687086e-01
3.23697358e-01 -9.94064286e-03 -6.18771255e-01 4.25054640e-01
7.77961910e-01 -9.33327198e-01 5.55990994e-01 -6.93584010e-02
1.45846888e-01 -1.22267216e-01 -8.54011059e-01 4.83354867e-01
4.63172257e-01 -4.48233373e-02 -9.87795472e-01 4.83491033e-01
9.13098276e-01 -4.38389391e-01 -7.65335321e-01 1.46290898e-01
4.69898224e-01 -1.62993029e-01 6.10524297e-01 -4.73701626e-01
1.21407080e+00 -1.61194861e-01 2.60242373e-01 -1.55720913e+00
-2.79919446e-01 -7.86617994e-01 -2.05611378e-01 1.74313033e+00
7.36724257e-01 -1.20701306e-02 4.97797579e-01 1.44660175e-01
-4.79639590e-01 -6.19484723e-01 -6.19668603e-01 -4.07352477e-01
1.15541451e-01 2.04264864e-01 2.23003894e-01 1.83714703e-01
4.09416527e-01 8.55983675e-01 -3.06463391e-01 -3.19833666e-01
6.77338839e-01 4.78480637e-01 8.34040403e-01 -9.93770361e-01
2.67567579e-02 -5.51755548e-01 3.63596052e-01 -1.00352132e+00
2.61929959e-01 -9.99572992e-01 2.17606544e-01 -2.28681684e+00
7.96487451e-01 3.83346349e-01 -1.00733303e-01 6.02061599e-02
-6.40196800e-01 -2.89619654e-01 3.14493775e-01 3.72579694e-01
-9.76715088e-01 5.06798685e-01 1.25269353e+00 -4.56032962e-01
-4.62875724e-01 1.76548004e-01 -1.23012137e+00 2.94327050e-01
8.19290876e-01 -4.91111994e-01 -1.94306254e-01 -2.19619825e-01
-9.02305692e-02 2.74925798e-01 -2.09568828e-01 -9.15911496e-01
4.50748771e-01 -4.61844280e-02 2.70993680e-01 -1.29707396e+00
-1.49442852e-01 2.01119244e-01 -2.39543468e-01 3.27917457e-01
-9.95532572e-01 -1.89353488e-02 2.94978291e-01 3.13809037e-01
-2.01791599e-01 -4.81298745e-01 3.33098769e-01 -4.29887399e-02
5.31566665e-02 -1.87731758e-01 -5.49974084e-01 2.56887913e-01
6.52359486e-01 2.43435115e-01 -5.65774798e-01 -3.59126270e-01
-3.53078872e-01 5.99295735e-01 1.21782854e-01 2.93880969e-01
5.29654622e-01 -6.80541933e-01 -1.41437519e+00 -4.01298523e-01
-1.66016102e-01 -7.72990584e-02 4.81912524e-01 8.18744659e-01
-5.82132161e-01 8.76116395e-01 -4.47533503e-02 -1.24552876e-01
-1.41760194e+00 9.39472020e-03 -3.24271649e-01 -9.53797221e-01
-5.44984698e-01 6.51599824e-01 -1.52967468e-01 -4.25277427e-02
5.10799214e-02 -2.05656275e-01 -5.88512838e-01 3.98939580e-01
9.60872233e-01 7.25669026e-01 3.45699966e-01 -4.34145838e-01
-2.95331359e-01 1.92361489e-01 -6.72509253e-01 -3.74760151e-01
1.54563940e+00 -1.06279366e-01 -3.81334513e-01 4.28653121e-01
1.07208300e+00 4.79906738e-01 -5.77575982e-01 2.06528138e-02
1.45588949e-01 3.17307442e-01 1.16129689e-01 -9.81770933e-01
-4.33774918e-01 7.14406908e-01 -4.78167117e-01 5.59486635e-02
7.63108790e-01 -1.16823383e-01 9.84712720e-01 3.64423305e-01
-2.22582877e-01 -1.28168666e+00 -8.90541077e-02 6.57308042e-01
1.24326241e+00 -7.55687118e-01 5.34732521e-01 -1.73781246e-01
-7.98739791e-01 1.09955192e+00 3.67390901e-01 -4.46131490e-02
-1.04887933e-01 -3.89786586e-02 -2.94016242e-01 -1.74612869e-02
-9.81885612e-01 4.00722176e-01 6.51760578e-01 4.63253595e-02
6.71663225e-01 4.53457534e-02 -1.10333157e+00 8.50876331e-01
-4.42637265e-01 4.83144261e-02 1.04703891e+00 5.72254658e-01
-7.70655870e-01 -8.62123787e-01 -1.07762225e-01 7.07330823e-01
-8.21835935e-01 -5.40630668e-02 -7.58306444e-01 1.31315559e-01
-5.81464827e-01 1.16681087e+00 -1.07691929e-01 -1.73114225e-01
5.07036924e-01 1.25800744e-01 2.48983145e-01 -9.03409898e-01
-9.12581086e-01 2.64590293e-01 4.30324137e-01 -3.67849320e-02
-2.02670768e-01 -8.32566619e-01 -1.23073769e+00 -1.21527195e-01
-8.22930709e-02 8.52577865e-01 7.62383640e-01 8.21952820e-01
6.65766299e-01 9.28977489e-01 6.44499838e-01 -8.51008058e-01
-6.05188191e-01 -1.53387237e+00 -2.47186780e-01 8.35794304e-03
3.27105701e-01 2.02724338e-01 -2.19813049e-01 1.35363430e-01] | [12.5426607131958, 9.534754753112793] |
5fd269d7-0607-4176-b7aa-396e39835716 | illumination-estimation-based-on-bilayer | null | null | http://openaccess.thecvf.com/content_cvpr_2013/html/Li_Illumination_Estimation_Based_2013_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2013/papers/Li_Illumination_Estimation_Based_2013_CVPR_paper.pdf | Illumination Estimation Based on Bilayer Sparse Coding | Computational color constancy is a very important topic in computer vision and has attracted many researchers' attention. Recently, lots of research has shown the effects of using high level visual content cues for improving illumination estimation. However, nearly all the existing methods are essentially combinational strategies in which image's content analysis is only used to guide the combination or selection from a variety of individual illumination estimation methods. In this paper, we propose a novel bilayer sparse coding model for illumination estimation that considers image similarity in terms of both low level color distribution and high level image scene content simultaneously. For the purpose, the image's scene content information is integrated with its color distribution to obtain optimal illumination estimation model. The experimental results on real-world image sets show that our algorithm is superior to some prevailing illumination estimation methods, even better than some combinational methods. | ['Houwen Peng', 'Weiming Hu', 'Weihua Xiong', 'Bing Li'] | 2013-06-01 | null | null | null | cvpr-2013-6 | ['color-constancy'] | ['computer-vision'] | [ 3.44581515e-01 -1.01234794e+00 -5.75355776e-02 -3.04337859e-01
-2.43504852e-01 -3.10052454e-01 3.88604075e-01 -2.29067251e-01
-1.29549265e-01 7.25038886e-01 1.92952082e-01 1.47146851e-01
3.28604355e-02 -5.07315814e-01 -2.55112261e-01 -1.14595532e+00
5.35052061e-01 -2.56507009e-01 2.69863099e-01 -5.17883524e-02
5.87519526e-01 3.50098550e-01 -1.71805584e+00 6.19832948e-02
1.07684696e+00 9.34395850e-01 3.84142548e-01 3.74690115e-01
-4.01086211e-01 8.41764033e-01 -2.56462514e-01 -2.29511976e-01
3.11255455e-01 -7.92655468e-01 -2.04870597e-01 4.98311251e-01
4.38368767e-01 -1.17849745e-01 -1.49890091e-02 1.67680061e+00
6.30334437e-01 1.71423495e-01 5.78811288e-01 -1.07592142e+00
-6.93676710e-01 -4.93881181e-02 -1.00846088e+00 3.20883214e-01
3.57261568e-01 1.89787913e-02 5.43941498e-01 -9.74319935e-01
2.24053696e-01 1.26612353e+00 2.99755156e-01 -4.16650921e-02
-9.69455779e-01 -5.16945541e-01 2.24466354e-01 6.03752792e-01
-1.51044810e+00 -3.70634735e-01 1.37250769e+00 -1.15603924e-01
3.31004143e-01 1.95623994e-01 6.60707116e-01 2.72859573e-01
2.25941017e-01 7.93045163e-01 1.75431073e+00 -6.94961548e-01
2.04536483e-01 4.01001245e-01 5.56833968e-02 8.90058339e-01
4.93366450e-01 -1.31152654e-02 -3.96048993e-01 -5.64523786e-02
7.87248313e-01 -2.51229983e-02 -6.42882884e-01 -4.29765522e-01
-9.40522134e-01 5.89548886e-01 3.81899804e-01 3.92954677e-01
-3.81264091e-01 -1.40190989e-01 4.85325977e-02 -1.44091278e-01
3.24102730e-01 -3.25375274e-02 -1.29652932e-01 2.79382527e-01
-6.81366384e-01 -1.59915954e-01 4.16862577e-01 7.86383986e-01
1.11073303e+00 3.58745784e-01 -5.35635911e-02 1.21997166e+00
4.61587578e-01 7.82648385e-01 3.30758691e-01 -9.53387022e-01
1.09489553e-01 5.06593883e-01 1.86506942e-01 -1.47643209e+00
-7.98620954e-02 -4.26659584e-01 -1.05528879e+00 3.43356133e-01
2.48473495e-01 -9.83409733e-02 -6.90634072e-01 1.46803176e+00
4.26436514e-01 3.11460048e-01 5.40379845e-02 1.22730160e+00
7.63681829e-01 8.94775748e-01 -2.17837274e-01 -6.96830451e-01
1.37689459e+00 -7.79969871e-01 -1.03581500e+00 -3.02912295e-02
-1.62927240e-01 -1.45041823e+00 8.31833005e-01 7.74378777e-01
-1.08791494e+00 -8.76337469e-01 -9.66825247e-01 1.06637269e-01
-1.61766917e-01 3.36885095e-01 7.60472953e-01 8.90922785e-01
-9.51590657e-01 -8.30591992e-02 -2.20494688e-01 -3.12559873e-01
1.99473679e-01 9.02468860e-02 4.78676744e-02 -6.94097400e-01
-9.50905621e-01 7.87822664e-01 2.64510483e-01 2.54187465e-01
-3.01497608e-01 -3.70969158e-03 -7.34846950e-01 -2.16764152e-01
2.15672940e-01 -5.17324150e-01 7.37893522e-01 -1.27990985e+00
-1.51404464e+00 4.27796125e-01 -5.78282833e-01 3.11546475e-01
1.36120290e-01 1.53725103e-01 -6.42942965e-01 2.27002025e-01
-1.83895856e-01 2.51564741e-01 1.03583992e+00 -1.66547465e+00
-6.67529404e-01 -3.41798216e-01 -1.70218423e-01 6.24055684e-01
-2.91713446e-01 6.88221529e-02 -9.18766260e-01 -2.75016189e-01
4.15037990e-01 -5.69693446e-01 -1.71961501e-01 9.59147811e-02
-2.53552794e-01 4.99380268e-02 9.09207523e-01 -5.19122720e-01
1.06559730e+00 -2.07836151e+00 3.00033726e-02 2.92357773e-01
-2.08400756e-01 2.52162635e-01 -1.78862318e-01 2.35961169e-01
2.47018486e-02 -4.47954893e-01 -3.00881773e-01 -4.90460545e-03
-3.58293921e-01 1.41461670e-01 1.78879574e-01 7.97426045e-01
-1.77819759e-01 3.21869731e-01 -9.12954867e-01 -1.01003456e+00
6.93693697e-01 7.83149123e-01 -4.58203644e-01 2.14362323e-01
2.13821232e-01 4.71263736e-01 -4.64199543e-01 1.02303076e+00
1.26119459e+00 4.31384072e-02 1.40585508e-02 -7.61064708e-01
-3.94872457e-01 -5.14498174e-01 -1.54916883e+00 1.35903525e+00
-4.19441283e-01 6.21694326e-01 2.12963060e-01 -8.83557379e-01
9.77894783e-01 1.50712386e-01 5.36233962e-01 -7.89312243e-01
3.79594326e-01 1.21379495e-01 6.71412796e-02 -6.05634272e-01
4.22895700e-01 -9.32291895e-02 5.33487141e-01 1.54575139e-01
-2.77613103e-01 -4.07325149e-01 4.12852079e-01 -2.72411164e-02
2.38536537e-01 1.74999788e-01 5.05013883e-01 -1.50128558e-01
1.06197488e+00 -2.52243966e-01 8.39997232e-01 3.98113996e-01
-4.03567940e-01 6.34394944e-01 3.48588228e-02 -2.11354017e-01
-8.37808847e-01 -8.14025342e-01 -3.93544942e-01 7.44690835e-01
8.94829094e-01 -4.70681712e-02 -6.72582388e-01 -1.82136938e-01
-2.43813753e-01 6.23733342e-01 -4.02396262e-01 7.08602816e-02
-3.86035621e-01 -1.02059209e+00 -1.37215927e-01 1.01983987e-01
1.11455011e+00 -8.47945869e-01 -3.20183158e-01 -3.00325677e-02
-4.35017139e-01 -9.33757722e-01 -5.51374018e-01 -2.36097738e-01
-7.04760373e-01 -1.11795068e+00 -8.29642951e-01 -9.20305669e-01
1.04239738e+00 1.12312686e+00 9.39470649e-01 1.47759795e-01
-4.83617663e-01 4.55281287e-01 -5.01236975e-01 -4.43272561e-01
2.52071749e-02 -5.55406809e-01 -1.69555143e-01 4.93788421e-01
2.63088286e-01 -3.78286064e-01 -8.94468486e-01 5.00543475e-01
-8.80186200e-01 1.09159820e-01 8.66747320e-01 7.75819659e-01
6.13923192e-01 6.10683918e-01 1.22458704e-01 -7.54037082e-01
4.52421278e-01 -2.61890650e-01 -7.36713767e-01 3.20136964e-01
-5.39890647e-01 -2.47753024e-01 6.76147401e-01 -1.87849119e-01
-1.75844598e+00 2.93412637e-02 3.38189192e-02 -1.88998654e-01
-1.11115202e-01 3.17934781e-01 -4.87959832e-01 -5.58555484e-01
1.77425519e-01 9.10940468e-01 -9.59669501e-02 -3.16476047e-01
2.32681677e-01 6.01969838e-01 4.72205400e-01 -5.14351904e-01
7.15102851e-01 4.91888374e-01 1.25122398e-01 -9.99842644e-01
-6.06168151e-01 -6.19427264e-01 -5.00622988e-01 -5.75848281e-01
7.30857611e-01 -8.33536208e-01 -7.82391250e-01 6.89405620e-01
-1.00250220e+00 2.27716595e-01 2.71123797e-01 6.60186827e-01
-1.56304240e-01 7.11631000e-01 -2.70896494e-01 -1.01097703e+00
3.65799107e-02 -1.34647536e+00 8.34833980e-01 7.32967138e-01
5.47542274e-01 -1.09341061e+00 1.40268533e-02 2.52197862e-01
5.42429686e-01 -8.63201618e-02 7.52205014e-01 4.89296734e-01
-7.87751138e-01 -5.84359914e-02 -7.39341557e-01 5.03071368e-01
5.07849276e-01 1.90015584e-01 -9.03947830e-01 -1.67098135e-01
2.50901759e-01 2.99185049e-03 7.26775587e-01 6.66332781e-01
1.31542838e+00 -8.29594284e-02 -3.69705558e-02 8.14484656e-01
2.14505100e+00 4.47753102e-01 9.51309204e-01 1.07930899e-01
7.43007302e-01 3.97471100e-01 7.51929641e-01 5.16374826e-01
1.16026007e-01 5.53946555e-01 5.13095498e-01 -5.03520608e-01
-3.39026570e-01 7.98511133e-02 1.64755180e-01 1.03509283e+00
-2.75091738e-01 -2.72277266e-01 -3.00375491e-01 3.43060255e-01
-1.66241527e+00 -1.13713479e+00 -2.12459490e-01 2.12181687e+00
7.79043317e-01 -2.70004034e-01 -2.26838201e-01 1.54913723e-01
9.55145597e-01 3.17000240e-01 -4.98567760e-01 -4.29581136e-01
-4.59628642e-01 -1.99302062e-01 4.45862085e-01 4.15956765e-01
-8.60681772e-01 6.37311697e-01 6.29459906e+00 9.46288943e-01
-1.11016452e+00 -1.29215315e-01 6.17202222e-01 3.79118443e-01
-3.48646194e-01 5.76804690e-02 -5.89248419e-01 7.29549766e-01
-7.42484406e-02 -9.11794305e-02 6.01161361e-01 4.81149167e-01
3.37477982e-01 -7.45000422e-01 -3.93079937e-01 1.37940109e+00
5.11346936e-01 -8.22627425e-01 1.36988342e-01 -1.51710257e-01
1.09995317e+00 -5.29074311e-01 2.32982472e-01 -1.77927315e-01
1.42640159e-01 -5.90402901e-01 2.87385345e-01 7.62638092e-01
6.07937396e-01 -8.60086501e-01 8.56346011e-01 1.71406671e-01
-1.42660379e+00 4.34369408e-02 -6.64367318e-01 3.18979323e-02
1.34741575e-01 6.57696605e-01 -2.12424353e-01 7.03593493e-01
4.52242166e-01 8.83890927e-01 -5.93206823e-01 1.38514841e+00
-3.13420683e-01 5.32013297e-01 -1.15589976e-01 -8.23463276e-02
2.00748324e-01 -7.56220460e-01 3.23857963e-01 1.11435962e+00
2.37064064e-01 4.10912246e-01 2.67888397e-01 5.60994446e-01
3.40235054e-01 5.37996888e-01 -4.67549801e-01 4.29840416e-01
3.73446733e-01 1.49096477e+00 -8.63887072e-01 -4.16538090e-01
-8.11212957e-01 9.40530479e-01 -1.27792478e-01 6.41958117e-01
-8.23626637e-01 -4.67997253e-01 3.62610102e-01 -3.91675591e-01
1.73857555e-01 -1.67309508e-01 -3.05250257e-01 -1.22212589e+00
-2.81268388e-01 -8.31707299e-01 2.01871067e-01 -1.27190065e+00
-1.15697920e+00 3.01561117e-01 -1.65793598e-01 -1.49197602e+00
4.37511563e-01 -5.37517309e-01 -8.65764558e-01 1.05687428e+00
-2.04157805e+00 -9.94739413e-01 -6.17848694e-01 1.00883782e+00
7.73021340e-01 -6.58285618e-02 4.76199120e-01 2.17453092e-01
-7.88099289e-01 2.58731902e-01 4.58932877e-01 -1.74187049e-01
8.65523458e-01 -1.14318252e+00 -6.40904367e-01 1.20359361e+00
-6.13306724e-02 6.00894213e-01 8.94664705e-01 -4.03220445e-01
-1.55786383e+00 -7.75125802e-01 4.96965498e-01 2.42035225e-01
9.36490595e-02 2.62723923e-01 -7.22453296e-01 8.27762708e-02
6.58253491e-01 -6.47454262e-02 6.52029753e-01 -8.50921348e-02
-3.10642365e-02 -5.12072742e-01 -1.18910515e+00 5.81766546e-01
4.95541811e-01 -2.06404239e-01 -3.59010875e-01 1.27618879e-01
3.29594910e-02 -4.39430214e-02 -5.47195256e-01 3.51273268e-01
5.39817512e-01 -1.31085873e+00 1.04240501e+00 3.35139573e-01
2.84821481e-01 -8.81117940e-01 -4.25692588e-01 -1.22334063e+00
-5.76198697e-01 -1.83738440e-01 2.69170552e-01 1.26167071e+00
-3.09567600e-01 -7.33292282e-01 4.13267881e-01 2.73848861e-01
-7.73946231e-04 -4.60770369e-01 -2.95581102e-01 -4.26328838e-01
-6.21907890e-01 -1.16925120e-01 3.69597822e-01 9.03699160e-01
-3.16934764e-01 3.48416567e-01 -6.41973972e-01 3.34630311e-01
1.18878829e+00 5.30255795e-01 7.36172736e-01 -9.81341660e-01
-1.10362485e-01 -5.11076152e-01 -2.35601932e-01 -8.21436346e-01
-8.88540000e-02 -4.10625428e-01 1.59305811e-01 -1.71137941e+00
5.89272499e-01 -2.82708466e-01 -5.95813274e-01 7.40383044e-02
-5.54151475e-01 5.94477534e-01 4.13808413e-02 2.64510155e-01
-6.61897361e-01 5.65051436e-01 1.63672090e+00 -2.03174755e-01
-6.94969147e-02 -3.43170110e-03 -8.18313241e-01 8.99559796e-01
7.26740301e-01 4.49962504e-02 -5.95407546e-01 -3.42969865e-01
-2.78053954e-02 -3.79617438e-02 2.28481919e-01 -1.00197423e+00
2.67080873e-01 -6.11313701e-01 7.73621678e-01 -8.29202950e-01
3.98062080e-01 -9.90713239e-01 1.47629557e-02 3.43513340e-01
-1.81289613e-02 -2.27713168e-01 -4.50154468e-02 6.44302189e-01
-5.37950695e-01 -1.35898754e-01 1.19823170e+00 -2.42123172e-01
-1.26278031e+00 3.36890936e-01 -2.98894346e-01 -2.65407264e-01
8.36113930e-01 -5.33294141e-01 -2.05626920e-01 -4.40620095e-01
-9.10620019e-02 -3.59840319e-02 6.09890521e-01 -4.50230427e-02
8.44334006e-01 -1.42339432e+00 -7.03246653e-01 3.67244214e-01
1.28730208e-01 -4.45871115e-01 4.78632182e-01 9.12179053e-01
-5.86596131e-01 1.95698306e-01 -4.08910275e-01 -6.29395664e-01
-1.43388486e+00 5.95080912e-01 9.06393304e-02 2.64087677e-01
-2.30670094e-01 8.65666509e-01 4.51643318e-01 2.12647483e-01
-1.72857586e-02 -4.03802991e-02 -4.86741275e-01 -1.90535918e-01
6.35042191e-01 4.60241288e-01 -3.27808738e-01 -9.04297173e-01
-3.12974453e-01 1.23402131e+00 1.78162381e-01 3.03955171e-02
9.54119146e-01 -6.88177466e-01 -4.80059475e-01 3.74572486e-01
1.34495640e+00 2.63456672e-01 -1.24304342e+00 -5.62971056e-01
-5.95416844e-01 -1.22898507e+00 4.29692119e-01 -8.64482641e-01
-1.35731685e+00 8.44826877e-01 8.15641224e-01 5.77698760e-02
1.85478508e+00 -4.92950588e-01 5.72769165e-01 9.28773805e-02
5.08278310e-01 -1.18345094e+00 2.18168840e-01 2.14277983e-01
6.89651668e-01 -1.52455163e+00 5.36671579e-01 -6.41783416e-01
-6.04730070e-01 1.10831308e+00 6.48706079e-01 -1.38599306e-01
5.32933652e-01 3.38800550e-02 1.62827954e-01 5.38899750e-02
-4.30301607e-01 -6.44020438e-01 3.49100769e-01 6.27180636e-01
5.77833295e-01 -2.16737866e-01 -5.19258261e-01 -9.59740573e-05
3.51298898e-01 -1.87827215e-01 4.75923449e-01 6.25046074e-01
-9.16526914e-01 -1.10245109e+00 -8.70925546e-01 2.36375883e-01
-3.04143161e-01 -1.77165121e-01 2.50336132e-03 4.69453663e-01
3.80915582e-01 1.20401871e+00 -1.84396654e-01 1.97598208e-02
-9.79532138e-04 -3.73661101e-01 7.38697529e-01 -2.74326175e-01
1.39407709e-01 5.15562475e-01 -3.36316854e-01 -4.85100597e-01
-8.73261690e-01 -8.06562066e-01 -7.55676925e-01 -1.68909699e-01
-5.82294285e-01 2.34631494e-01 8.80326390e-01 5.97855985e-01
-2.46754929e-01 4.05798912e-01 1.09901583e+00 -9.75923896e-01
1.98478788e-01 -6.67148054e-01 -9.82869148e-01 4.42020893e-01
2.22034663e-01 -7.88917840e-01 -4.43264216e-01 3.54554296e-01] | [10.647780418395996, -2.552541732788086] |
982dc34b-e682-42ed-9a0f-1ce7e79eb3ac | ensemble-prosody-prediction-for-expressive | 2304.00714 | null | https://arxiv.org/abs/2304.00714v1 | https://arxiv.org/pdf/2304.00714v1.pdf | Ensemble prosody prediction for expressive speech synthesis | Generating expressive speech with rich and varied prosody continues to be a challenge for Text-to-Speech. Most efforts have focused on sophisticated neural architectures intended to better model the data distribution. Yet, in evaluations it is generally found that no single model is preferred for all input texts. This suggests an approach that has rarely been used before for Text-to-Speech: an ensemble of models. We apply ensemble learning to prosody prediction. We construct simple ensembles of prosody predictors by varying either model architecture or model parameter values. To automatically select amongst the models in the ensemble when performing Text-to-Speech, we propose a novel, and computationally trivial, variance-based criterion. We demonstrate that even a small ensemble of prosody predictors yields useful diversity, which, combined with the proposed selection criterion, outperforms any individual model from the ensemble. | ['Simon King', 'Mark Gales', 'James Leoni', 'Alexandra Torresquintero', 'Tomás Gomez Ibarrondo', 'Christopher G. R. Wallis', 'Zack Hodari', 'Devang S Ram Mohan', 'Vivian Hu', 'Tian Huey Teh'] | 2023-04-03 | null | null | null | null | ['prosody-prediction', 'expressive-speech-synthesis', 'speech-synthesis'] | ['natural-language-processing', 'speech', 'speech'] | [ 3.42117012e-01 -7.94981942e-02 -1.11122690e-01 -4.09173936e-01
-1.07132840e+00 -6.29439592e-01 4.34355080e-01 -1.12192757e-01
-2.12151870e-01 8.38294685e-01 4.92433339e-01 -2.68598765e-01
-2.64194105e-02 -3.01594824e-01 -2.22020611e-01 -7.15962052e-01
2.08101705e-01 5.96955955e-01 8.61981958e-02 -5.38941443e-01
4.24067564e-02 3.00575614e-01 -1.66688538e+00 3.64019096e-01
6.95924640e-01 7.29634166e-01 3.53661448e-01 9.09237146e-01
-2.34628752e-01 9.20361400e-01 -8.68092418e-01 -2.46299922e-01
6.98426515e-02 -1.01015913e+00 -4.54099953e-01 -1.66871414e-01
7.54864216e-02 4.67765853e-02 2.24610567e-01 9.20077145e-01
8.87919545e-01 2.77813345e-01 7.95843422e-01 -7.64862359e-01
-2.08554655e-01 1.28480065e+00 7.79289380e-02 2.85907865e-01
2.68219054e-01 -8.37298706e-02 1.25543272e+00 -7.24350274e-01
4.17948365e-01 1.13848841e+00 6.43178403e-01 8.00762475e-01
-1.67503154e+00 -7.03729630e-01 6.20613657e-02 -9.60623026e-02
-1.22808123e+00 -9.06266987e-01 1.21166062e+00 -3.87551695e-01
1.00554097e+00 6.53809786e-01 5.36563993e-01 1.34399033e+00
-2.09677164e-02 7.26558149e-01 1.04013097e+00 -6.62493289e-01
2.31695145e-01 3.82182628e-01 -1.05985425e-01 1.08232409e-01
-1.67663798e-01 9.57806557e-02 -7.70054519e-01 -3.51418763e-01
1.81194976e-01 -7.90835023e-01 -1.45769566e-01 -9.34492722e-02
-9.49111462e-01 9.51972663e-01 -3.73723984e-01 5.46021521e-01
-4.66957957e-01 8.43139278e-05 5.66940069e-01 3.84675503e-01
5.87537527e-01 6.22325838e-01 -5.16375840e-01 -4.56591398e-01
-1.07122695e+00 5.23362756e-01 1.10665798e+00 5.00323296e-01
2.43099868e-01 5.98773360e-01 8.62033740e-02 1.42223704e+00
3.20775919e-02 2.99892157e-01 5.66871822e-01 -9.20118868e-01
3.07222307e-01 1.78169996e-01 -4.22026813e-02 -5.05343914e-01
-3.43851149e-01 -2.55434632e-01 -4.60065186e-01 1.46728560e-01
2.68264681e-01 -4.85821187e-01 -5.22172034e-01 1.89750075e+00
1.82770386e-01 -1.59674257e-01 1.71313897e-01 6.65222526e-01
5.37351549e-01 7.50136316e-01 6.78606257e-02 -7.73443162e-01
7.29432881e-01 -6.37411118e-01 -7.56875277e-01 1.07033616e-02
5.34416318e-01 -8.95163655e-01 9.18056726e-01 7.25851834e-01
-1.37176859e+00 -5.25602698e-01 -9.82845604e-01 3.45767140e-01
4.27115373e-02 5.31779453e-02 2.32026890e-01 1.01160944e+00
-9.70824480e-01 7.31223524e-01 -5.51687539e-01 1.34333977e-02
-1.17970981e-01 4.83391702e-01 -4.16195393e-02 6.89366698e-01
-9.41164434e-01 1.01339591e+00 6.09346390e-01 -3.26534539e-01
-4.60177690e-01 -5.24308980e-01 -5.90277195e-01 4.06642817e-02
2.06341952e-01 -6.45618677e-01 1.67205107e+00 -1.56328273e+00
-2.05950046e+00 4.51998681e-01 -2.41168857e-01 -5.90146661e-01
3.78760040e-01 -9.79048982e-02 -5.02332687e-01 -2.27073759e-01
-4.96013284e-01 5.05070925e-01 9.75888431e-01 -1.35354936e+00
-5.33542335e-01 2.93896962e-02 -4.88911808e-01 3.56423199e-01
-2.35850528e-01 5.19433916e-01 1.68526769e-01 -8.61597061e-01
-1.24913454e-01 -9.81415331e-01 -2.00462475e-01 -6.04743063e-01
-5.00334084e-01 -3.29875737e-01 5.26462674e-01 -6.21704280e-01
1.68219411e+00 -1.93449163e+00 4.13844287e-01 2.32913345e-01
-1.50115877e-01 1.77939221e-01 5.61507046e-02 5.54726124e-01
-2.19638214e-01 1.95739791e-01 -2.21442640e-01 -6.44856095e-01
1.20805427e-01 2.33369425e-01 -5.34993112e-01 -7.09774196e-02
1.56995714e-01 3.12016934e-01 -6.50391638e-01 -4.63090658e-01
2.13494718e-01 5.02396107e-01 -7.13312924e-01 2.43976772e-01
-4.51579899e-01 5.76392770e-01 -2.14910477e-01 3.59743834e-01
1.21904887e-01 3.26012611e-01 5.41535258e-01 1.22178525e-01
-2.63193041e-01 8.23434651e-01 -1.11644781e+00 1.14504862e+00
-5.74630260e-01 5.92762649e-01 -3.36819813e-02 -8.65811348e-01
1.28847122e+00 6.70575798e-01 5.45694172e-01 -2.67668545e-01
2.30796650e-01 6.77722812e-01 7.73201644e-01 -3.10706943e-01
6.40434206e-01 -5.84154665e-01 -4.41052057e-02 4.38886583e-01
2.17354342e-01 -3.36134523e-01 2.02877179e-01 -4.97106761e-01
7.62654185e-01 6.67603761e-02 5.55835605e-01 -2.80791253e-01
5.86017132e-01 -1.25988066e-01 5.47034383e-01 7.02042282e-01
-5.34795299e-02 6.86765492e-01 3.34522933e-01 -2.81118900e-01
-1.38008726e+00 -8.58662307e-01 -1.16495468e-01 1.50416160e+00
-4.78189617e-01 -3.64178151e-01 -9.00823772e-01 -3.76281589e-01
-2.84268379e-01 1.19566810e+00 -1.36783555e-01 1.25224531e-01
-9.36722338e-01 -6.62392139e-01 7.90050030e-01 3.26562017e-01
-2.18333349e-01 -1.46561813e+00 -4.43166405e-01 6.21104062e-01
-1.07393481e-01 -7.92490900e-01 -5.64070642e-01 6.17993593e-01
-7.48033166e-01 -4.28338140e-01 -6.04136467e-01 -6.64200246e-01
-1.05951481e-01 -3.59043896e-01 1.20282114e+00 -1.62000746e-01
2.92858303e-01 1.89280704e-01 -5.21432340e-01 -7.26747930e-01
-1.31852579e+00 2.03976259e-01 2.02239498e-01 4.49485891e-02
2.77650356e-01 -8.20281327e-01 -1.61820173e-01 1.61597684e-01
-6.03432894e-01 2.26036347e-02 2.77946383e-01 9.93952155e-01
5.07353127e-01 -6.20788559e-02 1.00511909e+00 -7.43575692e-01
1.01743209e+00 -4.53241587e-01 -2.20007554e-01 8.49882215e-02
-2.55536616e-01 9.66913775e-02 9.06201184e-01 -8.73163819e-01
-9.89951968e-01 2.72055924e-01 -6.65829539e-01 -5.02132118e-01
-2.74782002e-01 5.21572471e-01 -2.41038069e-01 2.03461275e-01
7.08710730e-01 4.34955955e-01 1.09523661e-01 -6.10574603e-01
2.65824765e-01 8.10822427e-01 3.78857881e-01 -5.93433619e-01
2.62564838e-01 -2.04915658e-01 -4.40507054e-01 -1.30705607e+00
-4.83237177e-01 -2.38296941e-01 -4.79319125e-01 -2.29332596e-01
6.69286668e-01 -6.94954634e-01 -4.11696345e-01 2.51028031e-01
-1.17860448e+00 -3.86222988e-01 -3.11332822e-01 5.18266022e-01
-9.85351384e-01 1.36809513e-01 -2.46993572e-01 -1.48141098e+00
-4.50138628e-01 -1.19415629e+00 8.90003681e-01 -1.05678119e-01
-9.01398897e-01 -9.04679775e-01 4.11330372e-01 1.27392426e-01
5.13880849e-01 4.79175784e-02 9.49327707e-01 -1.25460756e+00
5.43316547e-03 1.75618216e-01 6.51970208e-01 6.49267375e-01
1.75588846e-01 2.48407930e-01 -1.14049220e+00 5.95972547e-03
1.24895655e-01 -3.38387847e-01 8.05470169e-01 5.59388340e-01
1.09827101e+00 -4.14363623e-01 1.88180730e-02 3.28581482e-01
7.81148553e-01 7.06411362e-01 2.43480816e-01 2.77707316e-02
3.31650794e-01 6.46999002e-01 1.74681008e-01 4.11028922e-01
1.88373879e-01 9.34525967e-01 -5.82643002e-02 3.36515635e-01
-1.98612250e-02 -2.79875547e-01 7.40606785e-01 1.34729052e+00
-1.33267611e-01 -5.63677371e-01 -8.68987381e-01 5.25294065e-01
-1.54439700e+00 -1.31749582e+00 2.37225875e-01 2.17373657e+00
1.01942790e+00 3.00790638e-01 7.04565883e-01 4.33987975e-01
6.11305356e-01 2.78202862e-01 -4.03224617e-01 -8.24388146e-01
-2.86729902e-01 4.66406614e-01 -1.02969650e-02 5.02392054e-01
-1.06166399e+00 8.27934146e-01 7.10135698e+00 9.50880527e-01
-1.25458467e+00 -3.02507840e-02 4.45552886e-01 -3.61662745e-01
-6.95534766e-01 -1.39765695e-01 -9.39050198e-01 5.71179867e-01
1.32572699e+00 -4.22886044e-01 5.78309715e-01 8.35051239e-01
4.22791243e-01 3.27270001e-01 -1.09298658e+00 9.63472724e-01
1.36916175e-01 -1.16829622e+00 1.67013034e-02 -8.96315053e-02
6.70331776e-01 -4.29824181e-02 3.77140269e-02 3.55292290e-01
2.64602035e-01 -1.09245789e+00 1.06825888e+00 3.25601339e-01
3.24622780e-01 -9.72802997e-01 4.34231848e-01 5.98170400e-01
-9.75149155e-01 -2.59711772e-01 -1.82254892e-02 1.35763243e-01
4.51006174e-01 1.85426980e-01 -1.09265351e+00 2.82073915e-01
2.56012857e-01 1.91861719e-01 -6.74705356e-02 7.76185513e-01
2.95606226e-01 1.03937185e+00 -5.06362796e-01 -4.81584400e-01
5.14234044e-02 -3.91883105e-02 9.87002194e-01 1.45535862e+00
6.37254059e-01 -4.45700735e-02 1.80694222e-01 7.90982246e-01
2.09868282e-01 5.39684951e-01 -7.68948972e-01 -1.88267499e-01
6.78862691e-01 7.45581985e-01 -4.62042451e-01 -1.69540420e-01
-7.38553107e-02 5.95614851e-01 2.85908610e-01 1.28761247e-01
-5.30589402e-01 6.76111653e-02 7.55100548e-01 -5.83072230e-02
3.96011621e-01 -2.33007729e-01 -4.79976594e-01 -7.88274229e-01
-1.27422944e-01 -1.32901478e+00 3.40265870e-01 -5.89450538e-01
-1.29984224e+00 1.14295983e+00 2.19843075e-01 -1.08175981e+00
-9.88227844e-01 -4.62583482e-01 -7.85635293e-01 1.10062313e+00
-8.87212217e-01 -8.44722092e-01 3.43082666e-01 2.48932287e-01
1.05634439e+00 -5.74550033e-01 1.00450718e+00 9.63517092e-03
-5.14905214e-01 5.27213752e-01 2.54676789e-02 -2.70726651e-01
4.80239064e-01 -1.39664483e+00 2.55679756e-01 5.37823319e-01
5.00745058e-01 4.64918971e-01 1.24153733e+00 -4.21012700e-01
-8.91128719e-01 -7.78472364e-01 1.09159875e+00 -4.15961236e-01
6.88282311e-01 -2.97630548e-01 -1.00415301e+00 3.75497609e-01
4.27133709e-01 -5.91634035e-01 9.92106020e-01 2.16555476e-01
-5.31236194e-02 -1.35948015e-02 -6.58066273e-01 6.78034723e-01
7.81632543e-01 -5.72203815e-01 -7.53694355e-01 -1.31646186e-01
6.81717873e-01 -3.14314008e-01 -7.25525558e-01 3.95787984e-01
7.01034367e-01 -1.21288967e+00 7.20687747e-01 -6.34460151e-01
1.94963276e-01 -5.61308190e-02 -5.56907475e-01 -1.64908504e+00
-2.93313675e-02 -9.38712060e-01 -1.32997602e-01 1.41356492e+00
8.19064319e-01 -4.16355282e-01 5.85100174e-01 3.40421706e-01
-3.12503457e-01 -7.16632366e-01 -1.01179898e+00 -8.01619112e-01
3.85270476e-01 -7.47205317e-01 5.22664309e-01 5.77716589e-01
4.28901501e-02 5.22321999e-01 -6.42487764e-01 -1.97007909e-01
1.52102187e-01 -2.14570872e-02 7.88799345e-01 -1.20759022e+00
-6.85455620e-01 -9.66866910e-01 -1.46592349e-01 -9.71566498e-01
3.72756273e-01 -8.15940857e-01 2.63371199e-01 -8.47794294e-01
-1.67576045e-01 -5.28793752e-01 -3.68724167e-01 2.26610228e-01
-1.74029008e-01 -1.24575704e-01 4.81497377e-01 1.54138748e-02
-6.06215298e-02 5.58643341e-01 8.08387458e-01 9.82906893e-02
-7.96840370e-01 3.82829368e-01 -6.73222542e-01 7.84811735e-01
1.21005416e+00 -5.27830482e-01 -5.05018592e-01 1.33109823e-01
-1.94918346e-02 3.90539408e-01 5.41240312e-02 -8.94168079e-01
-5.39981388e-02 -2.28697807e-01 6.61471337e-02 -6.68547034e-01
7.12603211e-01 -4.96937215e-01 5.42360842e-01 1.08401760e-01
-6.22347414e-01 4.58628647e-02 2.05352113e-01 1.83129549e-01
-3.44683647e-01 -5.48978209e-01 7.98566401e-01 -1.27626546e-02
-3.87719870e-01 -1.45032942e-01 -4.81304228e-01 -1.90237556e-02
7.28910387e-01 -1.45257428e-01 2.29074121e-01 -6.07470572e-01
-9.14097667e-01 -2.75427520e-01 1.28800511e-01 5.02380371e-01
4.01569277e-01 -1.18255925e+00 -9.61469233e-01 1.71660140e-01
-1.23369239e-01 -4.29313600e-01 1.30465358e-01 6.97478592e-01
-1.92543119e-01 2.58290857e-01 5.64162657e-02 -5.09195328e-01
-1.62753320e+00 2.75553733e-01 5.74861825e-01 -2.53168821e-01
-5.05718648e-01 7.07172692e-01 -1.61428630e-01 -4.15040493e-01
1.16259307e-01 -1.84963971e-01 -3.56546253e-01 2.48598024e-01
4.13587540e-01 2.00335875e-01 -4.46392177e-03 -8.97005439e-01
-1.28702641e-01 1.85092971e-01 1.11981831e-01 -4.86843437e-01
1.32878339e+00 -5.12380712e-02 3.19099665e-01 1.06245553e+00
8.41561437e-01 2.97852397e-01 -1.03644800e+00 2.21517950e-01
2.74691314e-01 -1.20997146e-01 -3.29179093e-02 -7.29154944e-01
-5.00118554e-01 8.39599252e-01 1.47454605e-01 6.80220187e-01
1.15377593e+00 -2.19922498e-01 5.83389997e-01 3.81034046e-01
2.16659188e-01 -1.13862991e+00 -2.13138610e-01 7.12562203e-01
1.19497776e+00 -9.76336122e-01 -3.05437297e-01 -1.04027435e-01
-1.12904191e+00 1.26708889e+00 3.48953843e-01 -6.72650784e-02
6.56085193e-01 4.22436029e-01 2.96646923e-01 4.04841244e-01
-1.26789546e+00 -2.61775970e-01 2.82188147e-01 6.06701016e-01
7.82083333e-01 1.41743541e-01 -3.73127848e-01 6.99757576e-01
-8.23333919e-01 -1.94092467e-01 3.59782308e-01 3.76197278e-01
-6.08804524e-01 -1.45067513e+00 -3.45712543e-01 4.52147573e-01
-5.70593536e-01 -1.42017528e-01 -6.06705427e-01 6.15604877e-01
6.58035204e-02 1.03138268e+00 1.96293332e-02 -5.51448286e-01
3.56606245e-01 6.85628057e-01 3.57310444e-01 -5.16592503e-01
-9.95783746e-01 4.90966618e-01 7.00119913e-01 -5.87301813e-02
-4.37981248e-01 -8.11862886e-01 -7.82100379e-01 -8.47731009e-02
-4.10095394e-01 1.77656934e-01 6.02636099e-01 7.77881026e-01
-8.60916171e-03 4.53128278e-01 7.79039800e-01 -1.05770683e+00
-9.80371356e-01 -1.16172361e+00 -5.91814458e-01 1.83009923e-01
4.58757639e-01 -6.41449273e-01 -4.21926886e-01 6.92508891e-02] | [14.90961742401123, 6.645076274871826] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.